Cholangiocarcinoma Malignant Traits Are Promoted by Schwann Cells through TGFβ Signaling in a Model of Perineural Invasion

Abstract

The term cholangiocarcinoma (CCA) defines a class of epithelial malignancies originating from bile ducts. Although it has been demonstrated that CCA patients with perineural invasion (PNI) have a worse prognosis, the biological features of this phenomenon are yet unclear. Our data show that in human intrahepatic CCA specimens with documented PNI, nerve-infiltrating CCA cells display positivity of the epithelial marker cytokeratin 7, lower with respect to the rest of the tumor mass. In an in vitro 3D model, CCA cells move towards a peripheral nerve explant allowing contact with Schwann cells (SCs) emerging from the nerve. Here, we show that SCs produce soluble factors that favor the migration, invasion, survival and proliferation of CCA cells in vitro. This effect is accompanied by a cadherin switch, suggestive of an epithelial-mesenchymal transition. The influence of SCs in promoting the ability of CCA cells to migrate and invade the extracellular matrix is hampered by a specific TGFβ receptor 1 (TGFBR1) antagonist. Differential proteomic data indicate that the exposure of CCA cells to SC secreted factors induces the upregulation of key oncogenes and the concomitant downregulation of some tumor suppressors. Taken together, these data concur in identifying SCs as possible promoters of a more aggressive CCA phenotype, ascribing a central role to TGFβ signaling in regulating this process.

Keywords: EMT; Schwann cells; TGFβ; bile duct cancer.

Figure 1

In situ histomorphological and immunohistochemical study of nerve phenotype in intrahepatic cholangiocarcinoma (iCCA) samples with perineural invasion (PNI+). (A). Hematoxylin and eosin (H&E) stain. Nerves without infiltration (spared) and those infiltrated were present within the same tissue slide. Yellow arrowhead indicates a spared nerve, while the area in the box contains a nerve with PNI, magnified in the right panel. (B). Immunohistochemistry for cytokeratin 7 (CK7) in CCA tumor mass (upper panel) vs. PNI (lower panel). Dotted lines individuate an infiltrated nerve. Histogram shows means and standard deviation of the CK7+ cell area percentage. Statistical significance was assessed with Mann–Whitney U-test.

Figure 2

Three-dimensional perineural invasion assay. The cartoon shows the experimental system at the beginning (day 1) and the end (day 7) of the experiment. On the right, the components of the system are listed. The micrographs (A–E) show details of the experiment. (A) incipient degeneration of sciatic nerve (SN); (B) HuCC-T1 original plating site; (C) degenerated nerve surrounded by cells emerging from it; (D) HuCC-T1 cells (white arrows) approaching the gel drop containing the nerve explant; (E) immunoflorescence staining for GFAP (red) shows activated SCs chains. Nuclei are labelled by DAPI (blue). N = 5.

Figure 3

Migration of HuCC-T1 and Oz cells induced by hSC-conditioned medium. Human cell lines derived from iCCA tumors were seeded in the upper chamber of a transwell. Conditioned medium obtained from hSC primary cultures was placed in the lower chamber and migration assessed for HuCCT-T1 cells (panel A) and for Oz cells (panel B). As shown by the micrographs and related diagrams, hSC-conditioned medium increased the migration of iCCA cells through the insert. Student’s t-test * p ≤ 0.05; ** p ≤ 0.01. N = 3.

Figure 4

Human SC-conditioned medium fosters the invasiveness of HuCC-T1 cells in a Matrigel invasion assay. Micrographs and related diagrams show the higher propensity of HuCC-T1 cells to invade Matrigel when exposed to hSC CM with respect to control CM. Student’s t-test: * = p ≤ 0.05. N = 3.

Figure 5

Cadherin switch in HuCC-T1 cells cocultured with hSCs. Western blot analyses of E-cadherin and N-cadherin expression in HuCC-T1 cells cocultured with hSCs. The figure shows the downregulation of the epithelial marker E-cadherin and the coherent upregulation of the mesenchymal marker N-cadherin. The graphs depict the results of densitometric analyses. Student’s t-test: *** p ≤ 0.001. N = 3.

Figure 6

Cell viability assay in hSC CM-treated HuCC-T1 cells. The graphs show viability according to MTT assay in HuCC-T1 cells treated with hSC CM or control CM. Cell viability is higher in samples treated with hSC CM with respect to control at both conventional and suboptimal cell densities. Student’s t-test: ** p ≤ 0.01; *** p ≤ 0.001. N = 3.

Figure 7

Clonogenic assay in hSC CM-treated HuCC-T1 cells. The graph shows higher percentage of colonies formed with respect to initial seeding in HuCC-T1 cells treated with hSC CM compared to the control. Student’s t-test: **** p ≤ 0.0001. N = 3. The micrographs show the different morphology of the colonies formed by cells treated with hSC CM and control CM.

Figure 8

Proliferation assay in hSC CM-treated HuCC-T1 and Oz cells. The graphs show increased proliferation of both cell lines treated with hSC CM compared to control CM. Student’s t-test: *** p ≤ 0.001 **** p ≤ 0.0001. N = 3.

Figure 9

Migration and invasion assays in the presence of TGFBR1 antagonist treatment. Migration and invasion assays were performed in iCCA cells treating HuCC-T1 or Oz cells with the TGFBR1 antagonist (10 μM SB-431542). The graphs show that the antagonist completely reverts the effects of hSC CM. One-way ANOVA * p ≤ 0.05; ** p ≤ 0.01; **** p ≤ 0.0001. N = 3.