Redox regulation of cytokeratin 18 protein by NADPH oxidase 1 in preneoplastic human epithelial cells

Abstract

Introduction: A catalytic subunit of NADPH oxidase 1 (Nox1) is implicated to be involved in neoplastic progression in human epithelial cancers. We had previously demonstrated that Nox1 overexpression of immortalized epithelial cells was able to induce the generation of progenitor cells that expressed fetal-type cytokeratins 8 and 18.

Purpose: We aimed to investigate the direct effects and underlying mechanisms of Nox1 on expression of cytokeratin 18 (CK18).

Methods: Immortalized human epithelial GM16 cells with low CK18 were used in Nox1 overexpression experiments. NuB2 cells with high CK18 were used in Nox1 knockdown experiments. Protein expression of CK18, phosphorylated and ubiquitinated CK18 were analyzed by Western blot.

Results: With no effects on the mRNA levels, CK18 protein was increased upon Nox1 overexpression and decreased upon Nox1 knockdown. Treatment with proteasome inhibitor MG132 prevented CK18 degradation and increased CK18 protein indicating translational regulation of CK18. Treatment for NuB2 cells with N-acetyl-L: -cysteine, diphenyleneiodonium, or apocynin decreased CK18 protein levels indicating its regulation involving reactive oxygen species and flavoprotein Nox. It has been known that phosphorylation of CK18 regulates CK18 turnover by ubiquination. Consistently, Nox1 modulated CK18 phosphorylation at ser52. Nox1 knockdown and treatment with diphenyleneiodonium accumulated the levels of ubiquinated CK18 enhancing degradation causing decreased CK18 protein.

Conclusion: We demonstrated that Nox1 was able to induce CK18 stabilization by inhibiting CK18 protein degradation in a phosphorylation-dependent manner. CK18 accumulation induced by Nox1 is consistent with the persistence of fetal-type CK18 protein in many epithelial carcinomas.

Fig. 1

Expression of cytokeratins in model human epithelial cell lines and effects of transfection with Nox1. a Phase contrast pictures showed morphology of the parental GM16 cells in keratinocyte growth medium (KGM) and progenitor NuB2 and NuB1 cells in DMEM. b Expression of CK18 was higher in spindle NuB1 cells compared with cobblestone cell lines including FuB1, NuB2 as well as GM16 cells, which were cultured in KGM [GM16 (KGM)] or adapted to grow in DMEM [GM16 (DMEM)]. c Transfection of GM16 cells with Nox1 plasmid increased CK18 but not CK8 protein in KGM. d Plasmid containing full-length Nox1, but not truncated version of Nox1 (Nox1-tv), increased CK18 protein expression in GM16 cells cultured in KGM

Fig. 2

Post-translational regulation of CK18 protein by Nox1 evidenced by Nox1 overexpression in GM16 cells and Nox1 knockdown in NuB2 cells. For Nox1 overexpression experiments, GM16 cells were nucleofected with 5 μg pcDNA3.1 or pcDNA3.1/Nox1 and harvested after 48 h post-transfection for protein and 6–48 h for mRNA analyses. Transfection of GM16 cells with pcDNA3.1/Nox1 increased Nox1 and CK18 proteins (a), and Nox1 mRNA (c). However, Nox1 overexpression did not alter CK18 mRNA levels (e). For Nox1 knockdown experiments, NuB2 cells were transfected with 100 nM scrambled (Scr.) or Nox1 siRNA (siNox1). Cells were harvested and subjected for protein and mRNA analyses at 48 h and 24 h post-transfection, respectively. Nox1 knockdown decreased Nox1 and CK18 proteins (b), and Nox1 mRNA (d). However, Nox1 knockdown did not alter CK18 mRNA levels (f). Treatment for GM16 cells (g) or NuB2 cells (h) with proteasome inhibitor MG132 at 20 μM for 1 and 3 h increased CK18 protein expression indicating post-translational stabilization of CK18. In (c), (d), (e), and (f), quantitative real-time PCR was performed by TaqMan^® RT-PCR with relative expression (ΔRn) of the target gene versus GAPDH mRNA (data were mean ± SEM, N = 3, *P < 0.001, vs. cont plasmid or Scr.), and the corresponding Insets by conventional RT-PCR using primers stated in Materials and Methods. Cell lysates were prepared and analyzed for Nox1 and CK18 expression. β-Tubulin or GADPH was used as loading control

Fig. 3

Regulation of CK18 protein was ROS-dependent. (a, left panel) ROS generation was increased upon Nox1 overexpression in GM16 cells. GM16 cells (2 × 10⁴) were nucleofected with pcDNA3.1 or pcDNA3.1/Nox1, and ROS generation was determined 48 h later. The generation of ROS was determined by chemiluminescence using L-012. Data were mean ± SEM, N = 10, *P < 0.0001, versus cont plasmid. (a, right panel) ROS generation was inhibited in siNox1-transfected NuB2 cells. NuB2 cells (2 × 10⁴) were transfected with 100 nM scrambled (Scr.) or Nox1 siRNA (siNox1). After 48 h incubation, transfected cells were treated with EGF (100 ng/ml) in serum-free medium for 24 h and ROS generation using L-012 was determined. Data were mean ± SEM, N = 30, *P < 0.0001, versus Scr. Inhibitory effects of CK18 by flavoprotein inhibitor DPI, antioxidant NAC, and Nox inhibitor apocynin indicated redox regulation of CK18. NuB2 cells were treated with (b) 10 μM DPI for 10 h, (c) 2.5 and 5 mM NAC for 24 h, or (d) 1 mM apocynin for 6, 18, and 30 h. CK18 in cell lysates were determined by Western blot. GAPDH was used as a loading control. Normalized band intensities from Western blot data shown in the corresponding right panels were mean ± SE, N = 3 experiments

Fig. 4

Nox1-regulated CK18 protein expression by CK18 ubiquitination in a phosphorylation-dependent manner. a Upon treatment of NuB2 cells with proteasome inhibitor MG132 at 20 μM for 3 h, CK18 expression was increased concomitant with ubiquintination of CK18 (Ub-CK18 detectable at ~53.5 kDa) complex. b Treatment with 10 μM DPI for 10 h decreased CK18 and its phosphorylated protein (pS52CK18). This was concomitant with increased Ub-CK18. c Nox1 knockdown of NuB2 cells increased the levels of CK18 and pS52CK18 concomitant with increased Ub-CK18. d Nox1 overexpression of GM16 cells caused increased CK18 and pS52CK18. In d, right panel, proposed mechanisms for stabilization of CK18 by Nox1 in an ROS-, and CK18 phosphorylation-dependent manner. In human epithelial cells, growth factor ligation leads to Nox1 and subsequent ROS activation. ROS induces increased phosphorylation of CK18 rendering a decrease level of total CK18 that is subjected for ubiquitination. Hence, CK18 degradation is inhibited resulting in CK18 stabilization via Nox1/ROS pathway