Site-specific Smad3 linker region phosphorylation as a marker in cholangiocarcinoma

Background and Aim:

Transforming growth factor (TGF)-β1 has been implicated before in development and progression of cholangiocellular carcinoma (CCC). The canonical TGF-β pathway involves TGF-β receptor-mediated C-terminal phosphorylation of Smad3 in epithelial homeostasis, but in carcinogenesis TGF-β (and other cytokines) may activate variant serine/threonine kinases to phosphorylate Smad3 at linker sites. Thisin turn shifts Smad3 signaling from cytostatic pSmad3C to an oncogenic pSmad3L pathway. The role of Smad3 linker phosphorylation in cholangiocarcinoma (CCC) is still unclear. We identified presence of significant Smad3 linker region phosphorylation in cholangiocarcinoma tumors and cell lines, and started to delineate its function and cellular outcome.

Methods:

Smad3 C-terminal and linker site phosphorylation was examined upon TGF-β1 stimulation in a panel of human CCC cell lines by immunoblotting using site-specific antibodies against the phosphorylated Smad3 (pSmad3C Ser423/425, pSmad3L T179, pSmad3L S204, pSmad3L S208, pSmad3L S213).

To determine serine/threonine kinases that mediate linker site-specific Smad3 phosphorylation, selected CCC cell lines were treated with the following inhibitors: TGF-b1/LY2157299, JNK/SP600125, ERK MEK1/2/UO126, p38/SB203580, or GSK3β/SB216763 for 30 min before TGF-β1 treatment for 1h. We also analyzed Smad3 stability upon treatment with E738 which inhibits TGF/BMP-signaling by R-Smad depletion., Distribution of phosphorylated Smad3 C-terminal and linker sites was investigated in intra- and extrahepatic cholangiocarcinoma patients by immunohistochemistry and immunoblotting.

Results:

In CCC cell lines, phospho-Smad3 isoform signaling is complex but exists and can be related to TGF-β1 stimulation. Individual Smad3 linker regions are selectively phosphorylated in eCCC and iCCC patients and cell lines. Phosphorylated Smad3 linker ser204 and ser213 were the most prominent linker sites and existed in intra- and extra- hepatic cholangiocarcinoma cell lines. Importantly, in CCSW1 cells, we found a counter-regulation between Smad3 C-terminal and linker site patterns. When Smad3 C-terminal phosphorylation was induced then phospho-Smad3LSer204 and phospho-Smad3LSer213 were mitigated. In line, a subgroup of CCC patients displayed a similar pattern. Phospho-Smad3LSer204 and phospho-Smad3LSer213 positive areas excluded positivity for phospho-Smad3Cser423/425 staining. A kinase inhibitor screen in the CCSW1 cell line suggested the involvement of GSK3β, CDKs as well as p38 and Erk MAPKs in the regulation of Smad3LSer204, and Smad3LSer213 phosphorylation.

Interestingly, CCSW1 cells are significantly less sensitive to E738 and do not show a concentration dependent depletion of Smad2/3.

Conclusions:

We suggest that pSmad3Lser204 and pSmad3Lser213 are predominant in cholangiocarcinoma cells and patients and may antagonize the cytostatic pSmad3Cser423/425, thus switching the R-Smad pathway to a tumorigenic route. To understand the direct influence of Smad3 linker phosphorylation and its impact on the patient's outcome, we currently modulate site-specific Smad3 linker phosphorylation in vitro and delineate downstream effects and cellular outcome. Next, site-specific Smad3 phosphorylation in biopsy specimens will be determined and correlated with clinical parameters of the respective patients.