Advertisement

Antitumor Activity of a Novel Fibroblast Growth Factor Receptor Inhibitor for Intrahepatic Cholangiocarcinoma

Open ArchivePublished:July 24, 2019DOI:https://doi.org/10.1016/j.ajpath.2019.06.007
      Fibroblast growth factor receptor 2 (FGFR2) might have an important role in the pathogenesis and biology of cholangiocarcinoma (CCA). We examined FGFR expression in CCA tumor specimens obtained from patients and CCA cell lines, and then determined the effects of the novel FGFR inhibitor, derazantinib (DZB; formally, ARQ 087), which is currently in clinical phase 2 trials for intrahepatic CCA. DZB inhibited the growth of CCA cell lines in a dose-dependent manner, and extracellular signal-regulated kinase 1/2 and AKT. It also activated apoptotic and cell growth arrest signaling. DZB reduced the in vitro invasiveness and the expression of key epithelial-mesenchymal transition genes. The in vitro data correlated with the expression of FGFRs in human CCA specimens by immunohistochemistry (FGFR1, 30% positive; and FGFR2, 65% positive) and the CCA cell lines assayed by Western blot analysis. These correlated in vitro studies suggest that FGFR may play an important role in the pathogenesis and biology of CCA. Our findings support the notion that FGFR inhibitors, like DZB, should be further evaluated at the clinical stage as targeted therapy for CCA treatment.
      Cholangiocarcinoma (CCA) forms a highly heterogeneous group of malignant tumors arising from cholangiocytes located in the biliary epithelium. After hepatocellular carcinoma (HCC), CCA is the second most common liver malignancy in the United States and Western Europe. As there is no effective treatment for advanced CCA, it has a poor prognosis.
      • Kumar M.
      • Zhao X.
      • Wang X.W.
      Molecular carcinogenesis of hepatocellular carcinoma and intrahepatic cholangiocarcinoma: one step closer to personalized medicine?.
      • Leyva-Illades D.
      • McMillin M.
      • Quinn M.
      • Demorrow S.
      Cholangiocarcinoma pathogenesis: role of the tumor microenvironment.
      In addition, its survival rate is low, and only 5% of these patients survive for 5 years after tumor development.
      • Shaib Y.H.
      • El-Serag H.B.
      • Davila J.A.
      • Morgan R.
      • McGlynn K.A.
      Risk factors of intrahepatic cholangiocarcinoma in the United States: a case-control study.
      Surgical resection is the treatment for early-stage CCA, and this is frequently followed by chemotherapy or radiotherapy. However, post-resection 5-year survival is only 20% to 30%. In most cases, CCA is diagnosed at an advanced, unresectable stage; and although chemotherapy improves the life quality of these patients, it is only a palliative treatment.
      • Zabron A.
      • Edwards R.J.
      • Khan S.A.
      The challenge of cholangiocarcinoma: dissecting the molecular mechanisms of an insidious cancer.
      • Aljiffry M.
      • Walsh M.J.
      • Molinari M.
      Advances in diagnosis, treatment and palliation of cholangiocarcinoma: 1990-2009.
      • Glimelius B.
      • Hoffman K.
      • Sjoden P.O.
      • Jacobsson G.
      • Sellstrom H.
      • Enander L.K.
      • Linne T.
      • Svensson C.
      Chemotherapy improves survival and quality of life in advanced pancreatic and biliary cancer.
      Most patients with unresectable CCA will undergo a rapid clinical decline and usually die within 12 months from symptom onset. To improve the outlook for patients with CCA, both clinical and bench science advances are imperative.
      • Zabron A.
      • Edwards R.J.
      • Khan S.A.
      The challenge of cholangiocarcinoma: dissecting the molecular mechanisms of an insidious cancer.
      • Aljiffry M.
      • Walsh M.J.
      • Molinari M.
      Advances in diagnosis, treatment and palliation of cholangiocarcinoma: 1990-2009.
      • Glimelius B.
      • Hoffman K.
      • Sjoden P.O.
      • Jacobsson G.
      • Sellstrom H.
      • Enander L.K.
      • Linne T.
      • Svensson C.
      Chemotherapy improves survival and quality of life in advanced pancreatic and biliary cancer.
      • Hubbard S.R.
      • Till J.H.
      Protein tyrosine kinase structure and function.
      • Heldin C.H.
      Dimerization of cell surface receptors in signal transduction.
      • Pawson T.
      • Nash P.
      Protein-protein interactions define specificity in signal transduction.
      The fibroblast growth factor receptors (FGFRs) represent a family of transmembrane proteins with intrinsic tyrosine kinase activity. Binding of fibroblast growth factors (FGFs) is necessary for FGFR activation.
      • Hubbard S.R.
      • Till J.H.
      Protein tyrosine kinase structure and function.
      After ligand binding, FGFR dimerizes, and this leads to transphosphorylation of tyrosine residues and their activation.
      • Hubbard S.R.
      • Till J.H.
      Protein tyrosine kinase structure and function.
      • Heldin C.H.
      Dimerization of cell surface receptors in signal transduction.
      • Pawson T.
      • Nash P.
      Protein-protein interactions define specificity in signal transduction.
      The receptor family has four closely related genes, FGFRs 1 to 4, which are signals in the extracellular signal-regulated kinase (ERK) 1/2, STAT, and AKT effector pathways.
      • Hubbard S.R.
      • Till J.H.
      Protein tyrosine kinase structure and function.
      • Turner N.
      • Grose R.
      Fibroblast growth factor signalling: from development to cancer.
      • Beenken A.
      • Mohammadi M.
      The FGF family: biology, pathophysiology and therapy.
      • Schlessinger J.
      Cell signaling by receptor tyrosine kinases.
      FGFR signaling may play an important role in regulating CCA cell proliferation, survival, migration, invasion, angiogenesis, and tumor progression.
      • Banales J.M.
      • Cardinale V.
      • Carpino G.
      • Marzioni M.
      • Andersen J.B.
      • Invernizzi P.
      • Lind G.E.
      • Folseraas T.
      • Forbes S.J.
      • Fouassier L.
      • Geier A.
      • Calvisi D.F.
      • Mertens J.C.
      • Trauner M.
      • Benedetti A.
      • Maroni L.
      • Vaquero J.
      • Macias R.I.
      • Raggi C.
      • Perugorria M.J.
      • Gaudio E.
      • Boberg K.M.
      • Marin J.J.
      • Alvaro D.
      Expert consensus document: cholangiocarcinoma: current knowledge and future perspectives consensus statement from the European Network for the Study of Cholangiocarcinoma (ENS-CCA).
      Derazantinib (DZB) is a recently developed, orally bioavailable, ATP-competitive inhibitor of FGFRs 1 to 3. DZB has potent in vitro and in vivo inhibitory effects on a variety of FGFR-dependent human cancer cell lines and xenograft tumor models.
      • Yu Y.
      • Hall T.
      • Eathiraj S.
      • Wick M.J.
      • Schwartz B.
      • Abbadessa G.
      In-vitro and in-vivo combined effect of ARQ 092, an AKT inhibitor, with ARQ 087, a FGFR inhibitor.
      • Chila R.
      • Hall G.T.
      • Abbadessa G.
      • Broggini M.
      • Damia G.
      Multi-chemotherapeutic schedules containing the pan-FGFR inhibitor ARQ 087 are safe and show antitumor activity in different xenograft models.
      • Hall T.G.
      • Yu Y.
      • Eathiraj S.
      • Wang Y.
      • Savage R.E.
      • Lapierre J.M.
      • Schwartz B.
      • Abbadessa G.
      Preclinical activity of ARQ 087, a novel inhibitor targeting FGFR dysregulation.
      • Papadopoulos K.P.
      • El-Rayes B.F.
      • Tolcher A.W.
      • Patnaik A.
      • Rasco D.W.
      • Harvey R.D.
      • LoRusso P.M.
      • Sachdev J.C.
      • Abbadessa G.
      • Savage R.E.
      • Hall T.
      • Schwartz B.
      • Wang Y.
      • Kazakin J.
      • Shaib W.L.
      A phase 1 study of ARQ 087, an oral pan-FGFR inhibitor in patients with advanced solid tumours.
      • Mazzaferro V.
      • El-Rayes B.F.
      • Droz Dit Busset M.
      • Cotsoglou C.
      • Harris W.P.
      • Damjanov N.
      • Masi G.
      • Rimassa L.
      • Personeni N.
      • Braiteh F.
      • Zagonel V.
      • Papadopoulos K.P.
      • Hall T.
      • Wang Y.
      • Schwartz B.
      • Kazakin J.
      • Bhoori S.
      • de Braud F.
      • Shaib W.L.
      Derazantinib (ARQ 087) in advanced or inoperable FGFR2 gene fusion-positive intrahepatic cholangiocarcinoma.
      More important, there is clinical proof of concept in intrahepatic CCA (iCCA), and a registrational phase 2 study in iCCA is ongoing for DZB (NCT03230318), which is sponsored by Basilea Pharmaceutica (Basel, Switzerland).
      In this study, we determined the in vitro pharmacologic inhibitory functions of DZB on human CCA cell lines to assess its potential therapeutic use for cancer.

      Materials and Methods

      Human Samples

      A total of 32 formalin-fixed, paraffin-embedded human liver tissue specimens were analyzed for the immunohistochemical expression of FGFR1 and FGFR2. All specimens were obtained after surgical resection and collected in the tissue bank at Humanitas Clinical Institute (Rozzano, Italy) in accordance with informed consent retrieved from patients and local ethics committee approval. The analyzed specimens included 19 samples of CCA, 10 samples of cirrhotic liver (hepatitis C virus positive) with superimposed HCC, together with normal liver tissue (n = 3) as negative controls, which include the surrounding liver in resected metastasis from colorectal cancer (n = 2) and hepatic adenoma (n = 1). Normal liver tissue was obtained after partial hepatectomy for a solitary colorectal cancer metastasis. In these cases, FGFR 1 and 2 expression levels were evaluated in liver tissue distant from the metastasis. The histology of the primary tumors was reviewed by board-certified pathologists (A.D. and L.D.T.) experienced in liver tumors and classified according to the tumor-node-metastasis staging system (Supplemental Table S1).
      For the RT-PCR analysis of Fgfr RNA expression, the nonmalignant tissue counterpart adjacent to CCA tissue was collected at the time of surgery and, when suitable, used as control. Frozen tumor tissues from CCA patients (n = 33) [Humanitas Clinical Institute, n = 13; Institute of Candiolo (Candiolo, Italy), n = 3; and Donostia University Hospital (San Sebastián, Spain), n = 17] were examined. Surrounding normal human tissues (n = 18) were also tested. The RT-PCR cohort was independent from the cohort in which immunohistochemical expression was performed. Correlation analysis between Fgfr gene expression and tumor grade was performed with 26 of 33 cases that had a clear G1/G2/G3 status presented in Supplemental Table S2. Research protocols were approved by the Clinical Research Ethics Committees of the supporting institutions, and all patients signed written consents for the use of their samples for biomedical research.

      Cell Cultures and Reagents

      HUCCT1 and CCLP1 cells from intrahepatic bile duct cancer tissues were a kind gift from Dr. Anthony J. Demetris (University of Pittsburgh, Pittsburgh, PA); cell lines were cultured as described previously.
      • Miyagiwa M.
      • Ichida T.
      • Tokiwa T.
      • Sato J.
      • Sasaki H.
      A new human cholangiocellular carcinoma cell line (HuCC-T1) producing carbohydrate antigen 19/9 in serum-free medium.
      • Han C.
      • Wu T.
      Cyclooxygenase-2-derived prostaglandin E2 promotes human cholangiocarcinoma cell growth and invasion through EP1 receptor-mediated activation of the epidermal growth factor receptor and Akt.
      • Shimizu Y.
      • Demetris A.J.
      • Gollin S.M.
      • Storto P.D.
      • Bedford H.M.
      • Altarac S.
      • Iwatsuki S.
      • Herberman R.B.
      • Whiteside T.L.
      Two new human cholangiocarcinoma cell lines and their cytogenetics and responses to growth factors, hormones, cytokines or immunologic effector cells.
      The human nonmalignant cholangiocyte cell line (normal human cholangiocytes) was provided by Dr. Jesus M. Banales (Donostia University Hospital, San Sebastián, Spain).
      • Banales J.M.
      • Saez E.
      • Uriz M.
      • Sarvide S.
      • Urribarri A.D.
      • Splinter P.
      • Tietz Bogert P.S.
      • Bujanda L.
      • Prieto J.
      • Medina J.F.
      • LaRusso N.F.
      Up-regulation of microRNA 506 leads to decreased Cl-/HCO3- anion exchanger 2 expression in biliary epithelium of patients with primary biliary cirrhosis.
      The primary human hepatocytes were purchased from Sigma (St. Louis, MO). The HCC cell lines (PLC/PRF/5, Huh7, and HepG2) (a kind gift from Dr. Snorri S. Thorgeirsson, Valentina Factor, and Elizabeth A. Conner, Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD) were cultured as described previously.
      • Raggi C.
      • Factor V.M.
      • Seo D.
      • Holczbauer A.
      • Gillen M.C.
      • Marquardt J.U.
      • Andersen J.B.
      • Durkin M.
      • Thorgeirsson S.S.
      Epigenetic reprogramming modulates malignant properties of human liver cancer.
      The stable HUCCT1 cell line overexpressing FGFR2-periphilin 1 (PPHLN1) fusion protein
      • Sia D.
      • Losic B.
      • Moeini A.
      • Cabellos L.
      • Hao K.
      • Revill K.
      • Bonal D.
      • Miltiadous O.
      • Zhang Z.
      • Hoshida Y.
      • Cornella H.
      • Castillo-Martin M.
      • Pinyol R.
      • Kasai Y.
      • Roayaie S.
      • Thung S.N.
      • Fuster J.
      • Schwartz M.E.
      • Waxman S.
      • Cordon-Cardo C.
      • Schadt E.
      • Mazzaferro V.
      • Llovet J.M.
      Massive parallel sequencing uncovers actionable FGFR2-PPHLN1 fusion and ARAF mutations in intrahepatic cholangiocarcinoma.
      was kindly provided by Prof. Vincenzo Mazzaferro (IRCCS Foundation National Cancer Institute, Milan, Italy). DZB was kindly provided by ArQule, Inc. (Burlington, MA). The anti-FGFR inhibitors, TG10052 hydrochloride, ENMD-2076, and ENMD-2076 tartrate were purchased from MedChemExpress (Princeton, NJ).

      Measurement of Cell Viability

      Cell lines were plated in 96-well plates at a concentration of 2000 cells/well and incubated for 2, 24, and 72 hours with different concentrations of ARQ 087 (0.1, 0.3, 1, 3, and 10 μmol/L). The control solution consisted of 0.1% dimethyl sulfoxide. Cell viability was measured by MTT assay (Sigma), as previously described.
      • Petta S.
      • Valenti L.
      • Marra F.
      • Grimaudo S.
      • Tripodo C.
      • Bugianesi E.
      • Camma C.
      • Cappon A.
      • Di Marco V.
      • Di Maira G.
      • Dongiovanni P.
      • Rametta R.
      • Gulino A.
      • Mozzi E.
      • Orlando E.
      • Maggioni M.
      • Pipitone R.M.
      • Fargion S.
      • Craxi A.
      MERTK rs4374383 polymorphism affects the severity of fibrosis in non-alcoholic fatty liver disease.
      Briefly, at the end of treatment, 10 μL of MTT reagent was added to each well, and the plates were incubated at 37°C for an additional 2 hours. The absorbance was measured at 550 nmol/L using the ThermoScientific Multiskan FC microplate reader (Thermo Fisher Scientific, Waltham, MA).
      • Hall T.G.
      • Yu Y.
      • Eathiraj S.
      • Wang Y.
      • Savage R.E.
      • Lapierre J.M.
      • Schwartz B.
      • Abbadessa G.
      Preclinical activity of ARQ 087, a novel inhibitor targeting FGFR dysregulation.

      Western Blot Analysis

      Cells were plated at a density of 106 cells/55-cm2 plate dish. After 4 days, cells were treated with ARQ 087 at different concentrations (0, 0.3, 3, and 10 μmol/L) for 2 hours and then stimulated with FGF7 (R&D Systems, Minneapolis, MN) at 100 ng/mL for 15 minutes. Cells were also stimulated with FGF7 in the absence of ARQ 087 (0-) for use as a control.
      • Hall T.G.
      • Yu Y.
      • Eathiraj S.
      • Wang Y.
      • Savage R.E.
      • Lapierre J.M.
      • Schwartz B.
      • Abbadessa G.
      Preclinical activity of ARQ 087, a novel inhibitor targeting FGFR dysregulation.
      Cells were lysed at 4°C with lysis buffer (1% Triton X-100, 50 mmol/L Tris-HCl, pH 7.4, 150 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L sodium orthovanadate, 2 mmol/L phenylmethylsulfonyl fluoride, and 1 mmol/L each of leupeptin and pepstatin). After 30 minutes of lysis, cellular extracts were centrifuged for 10 minutes at 12,000 × g, and the supernatant was used for Western blot analysis experiments, as detailed previously.
      • Bonacchi A.
      • Romagnani P.
      • Romanelli R.G.
      • Efsen E.
      • Annunziato F.
      • Lasagni L.
      • Francalanci M.
      • Serio M.
      • Laffi G.
      • Pinzani M.
      • Gentilini P.
      • Marra F.
      Signal transduction by the chemokine receptor CXCR3: activation of Ras/ERK, Src, and phosphatidylinositol 3-kinase/Akt controls cell migration and proliferation in human vascular pericytes.
      Antibodies were used for Western blot analysis, according to the manufacturer's recommendations.
      Immunoblots were incubated overnight at 4°C with primary antibody in 1% bovine serum albumin in 1× Dulbecco's phosphate-buffered saline. Rabbit monoclonal anti-FGF receptor 1 XP (number 9740; Cell Signaling Technology, Danvers, MA), rabbit monoclonal anti-FGF receptor 2 (number 11835; Cell Signaling Technology), mouse monoclonal anti–phosphorylated (phospho-)FGFR (Tyr-653/654; number 3476; Cell Signaling Technology), rabbit monoclonal anti–c-Myc (number 5605; Cell Signaling Technology), mouse monoclonal anti–phospho-AKT (Ser473; number 4051; Cell Signaling Technology), rabbit polyclonal anti-AKT (number 9272; Cell Signaling Technology), rabbit polyclonal anti–phospho-ERK1/2 (Thr202/Tyr204; number 9101; Cell Signaling Technology), mouse monoclonal anti-ERK 1/2 (sc-514302; Santa Cruz Biotechnology, Dallas, TX), rabbit polyclonal anti–Bcl-2 (ab7973; Abcam, Cambridge, UK), rabbit monoclonal anti-p27 Kip1 (number 3686; Cell Signaling Technology), rabbit monoclonal anti–poly (ADP-ribose) polymerase (number 9532; Cell Signaling Technology), rabbit polyclonal anti–cyclin D1 (number 2922; Cell Signaling Technology), rabbit polyclonal anti–phospho-p38 (Thr180/Tyr182)-R (sc-17852-R; Santa Cruz Biotechnology), rabbit polyclonal anti-p38α (sc-535; Santa Cruz Biotechnology), rabbit polyclonal anti–Angio-1 (sc-8357; Santa Cruz Biotechnology), and rabbit polyclonal anti-FGF2 (sc-7911; Santa Cruz Biotechnology) antibodies were used. Immunoblots were then incubated with secondary antibody α-rabbit/mouse (1:4000) in 1% bovine serum albumin in 1× Dulbecco's phosphate-buffered saline for 1 hour. Monoclonal anti–β-actin antibody, produced in mouse (A5441; Sigma), or monoclonal antivinculin antibody, produced in mouse (V9131; Sigma), was used as internal control (1:1000) in 1% bovine serum albumin in 1× Dulbecco's phosphate-buffered saline. The signal was quantified by chemiluminescence detection on an Image Quant Las4000 (GE Healthcare Life Sciences, Marlborough, MA), and subsequent analysis was performed with ImageJ software version 1 (NIH; http://imagej.nih.gov/ij).

      Cell Cycle and Apoptosis Analysis

      In total, 100,000 cells/well were seeded in multiwell dishes and exposed to the appropriate conditions. After medium removal, 400 μL of solution containing 50 μg/mL propidium iodide, 0.1% w/v trisodium citrate, and 0.1% Nonidet P-40 were added. Samples were then incubated for 30 minutes at 4°C in the dark, and nuclei were analyzed with a FACSCanto flow cytometer (Becton Dickinson, Franklin Lakes, NJ). Gating was performed to exclude cellular debris. For the quantification of apoptosis, cells were resuspended in antibody-binding buffer (HEPES-buffered saline solution with 2.5 mmol/L CaCl2) and incubated with fluorescein isothiocyanate–labeled annexin-V (Roche Diagnostics, Basel, Switzerland) and propidium iodide for 15 minutes at room temperature in the dark. Flow cytometry was performed using a FACSCanto instrument. Annexin-V–positive/propidium iodide–negative cells were considered early apoptotic, whereas annexin-V–positive/propidium iodide–positive cells were considered late apoptotic; annexin-V–positive cells were defined as total apoptosis.

      Immunohistochemistry

      Paraffin-embedded sections (2 μm thick) of human CCA and HCC and liver metastasis from colorectal cancer were incubated with Antigen Retrieval for 25 minutes at 98°C in Citrate Buffer (pH 6.0) for FGFR1 and EDTA buffer (pH 8.0) for FGFR2 (Bioptica, Milan, Italy). Sections were incubated with FGFR1 (number 9740; Cell Signaling Technology; dilution 1:200) or FGFR2 [FGFR2 antibody (number 11835; Cell Signaling Technology; dilution 1:50] for 1 hour at room temperature, MACH4 (Biocare, Pacheco, CA) for 30 minutes at room temperature, and then diaminobenzidine (Biocare) for 5 minutes at room temperature. In addition, sections were stained with hematoxylin for 1 minute. FGFR1 and FGFR2 detection was accomplished using an automatic immunohistochemical staining machine (Dako, Carpinteria, CA).
      • Hall T.G.
      • Yu Y.
      • Eathiraj S.
      • Wang Y.
      • Savage R.E.
      • Lapierre J.M.
      • Schwartz B.
      • Abbadessa G.
      Preclinical activity of ARQ 087, a novel inhibitor targeting FGFR dysregulation.
      The positive cell proportion was classified into four grades (<1%, 1% to 30%, 31% to 50%, and 51% to 100%). The staining intensity was graded from 0 to 3 (0, no staining; 1, weak staining; 2, moderate staining; and 3, strong staining). After summing two grades, a total staining score was obtained for one section. A section was considered positive if the total staining score was >4. Analysis was performed in a blinded manner by a board-certified pathologist (A.D. or L.D.T.).

      Real-Time Quantitative PCR

      Total RNA was extracted with the RNeasy kit (Qiagen, Düsseldorf, Germany), according to the manufacturer's instructions. RNA concentration and quality were measured using an optical NanoDrop ND1000 spectrophotometer (Thermo Fisher Scientific). Total RNA (500 ng) was transcribed with a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Beverly, MA). Changes in the mRNA expression level of target genes were detected using FAST SYBR-Green PCR Master Mix and the 7900HT Fast Real Time PCR System (Applied Biosystems). RNA isolated from human tissue samples was reverse transcribed using the SuperScript Vilo cDNA Synthesis Kit (Invitrogen, Carlsbad, CA), according to the manufacturer's instructions. Real-time quantitative PCR was performed with the iQ SYBR Green Supermix (Bio-Rad, Milan, Italy), according to the manufacturer's instructions. The mRNA levels of glyceraldehyde-3-phosphate dehydrogenase (Gapdh) were used for normalization. For the CCA cell lines, fold difference (2−ΔΔCT) was calculated using the ΔCT of normal human cholangiocytes; for CCA tissue, surrounding tissues were used as control. All reactions were performed in triplicate. PCR was performed on cDNA with human gene-specific primer pairs for Fgfrs 1-4
      • Rizvi S.
      • Yamada D.
      • Hirsova P.
      • Bronk S.F.
      • Werneburg N.W.
      • Krishnan A.
      • Salim W.
      • Zhang L.
      • Trushina E.
      • Truty M.J.
      • Gores G.J.
      A Hippo and fibroblast growth factor receptor autocrine pathway in cholangiocarcinoma.
      and Fgfr2-Pphln1.
      • Sia D.
      • Losic B.
      • Moeini A.
      • Cabellos L.
      • Hao K.
      • Revill K.
      • Bonal D.
      • Miltiadous O.
      • Zhang Z.
      • Hoshida Y.
      • Cornella H.
      • Castillo-Martin M.
      • Pinyol R.
      • Kasai Y.
      • Roayaie S.
      • Thung S.N.
      • Fuster J.
      • Schwartz M.E.
      • Waxman S.
      • Cordon-Cardo C.
      • Schadt E.
      • Mazzaferro V.
      • Llovet J.M.
      Massive parallel sequencing uncovers actionable FGFR2-PPHLN1 fusion and ARAF mutations in intrahepatic cholangiocarcinoma.

      Chemoinvasion Assays

      Subconfluent CCLP1 cells were treated with vehicle or ARQ 087 at 0.3 and 1 μmol/L for 2 hours and then washed, trypsinized, and resuspended in serum-free medium at a concentration of 1.5 × 105 cells/mL. Chemoinvasion was measured in a modified Boyden chamber equipped with 8-μm pore filters (Millipore Corp., Burlington, MA) coated with Matrigel (150 μg/mL; BD Biosciences, San Jose, CA), as described previously.
      • Rovida E.
      • Di Maira G.
      • Tusa I.
      • Cannito S.
      • Paternostro C.
      • Navari N.
      • Vivoli E.
      • Deng X.
      • Gray N.S.
      • Esparis-Ogando A.
      • David E.
      • Pandiella A.
      • Dello Sbarba P.
      • Parola M.
      • Marra F.
      The mitogen-activated protein kinase ERK5 regulates the development and growth of hepatocellular carcinoma.
      After incubation (24 hours), the cells that invaded the underside of the filters were fixed, stained with Giemsa, mounted, and counted at ×40 magnification. Values for invasion were expressed as the average number of invading cells per microscopic field (×40) over five fields. Each experiment was performed in triplicate.

      Gene Silencing

      All siRNAs used were purchased from Dharmacon (Milan, Italy). Transfection of human CCLP1 was performed with the Amaxa nucleofection technology (Lonza, Basel, Switzerland), as previously described, with 100 nmol/L smart Pool siRNA specific for human Fgfr1 or Fgfr2, in combination, or nontargeting siRNA.
      • Rovida E.
      • Navari N.
      • Caligiuri A.
      • Dello Sbarba P.
      • Marra F.
      ERK5 differentially regulates PDGF-induced proliferation and migration of hepatic stellate cells.

      Statistical Analysis

      GraphPad Prism software version 5 (GraphPad Software, San Diego, CA) was used for data analysis. Error bars represent 1 ± SEM. P value was calculated by t-test. Statistical significance and P value are shown, where relevant.

      Results

      Different FGFRs Are Expressed in Human CCA Cell Lines

      First, the expression of Fgfrs 1-4 in human iCCA cell lines was assessed. Albeit at different levels, both CCLP1 and HUCCT1 cells displayed increased Fgfr1 and Fgfr2 gene and protein expression (Figure 1) compared with normal human cholangiocytes, thereby mirroring the heterogeneity in tumorigenic potential.
      • Banales J.M.
      • Saez E.
      • Uriz M.
      • Sarvide S.
      • Urribarri A.D.
      • Splinter P.
      • Tietz Bogert P.S.
      • Bujanda L.
      • Prieto J.
      • Medina J.F.
      • LaRusso N.F.
      Up-regulation of microRNA 506 leads to decreased Cl-/HCO3- anion exchanger 2 expression in biliary epithelium of patients with primary biliary cirrhosis.
      • Raggi C.
      • Correnti M.
      • Sica A.
      • Andersen J.B.
      • Cardinale V.
      • Alvaro D.
      • Chiorino G.
      • Forti E.
      • Glaser S.
      • Alpini G.
      • Destro A.
      • Sozio F.
      • Di Tommaso L.
      • Roncalli M.
      • Banales J.M.
      • Coulouarn C.
      • Bujanda L.
      • Torzilli G.
      • Invernizzi P.
      Cholangiocarcinoma stem-like subset shapes tumor-initiating niche by educating associated macrophages.
      Gene expression analysis of the HCC cell lines, PLC/PRF/5, Huh7, and HepG2, revealed minimal expression of Fgfr2 compared with Fgfr1 (Supplemental Figure S1). This suggests that Fgfr1 expression can affect HCC development. Collectively, these findings indicate that FGFR2 may have a specific role in the malignant transformation of cholangiocytes.
      Figure thumbnail gr1
      Figure 1Expression of fibroblast growth factor receptor (FGFR) members in cholangiocarcinoma cells. A: Relative expression of transcript-encoding receptors for Fgfrs 1-4 in CCLP1 and HUCCT1 cells. Gapdh was used as an internal control. All mRNA levels are presented as fold changes normalized to 1 [mean expression of normal human cholangiocytes (NHCs) represented as dashed line in the graph]. B: Western blot analysis was used to determine total FGFR1 and FGFR2 in HUCCT1 and CCLP1 cells. Equal loading was evaluated using anti–β-actin antibody. Data are expressed as means ± SEM (A). **P < 0.01, ***P < 0.001 versus NHCs (t-test).

      Evaluation of the Expression of FGFRs in Human CCA Specimens

      Fgfr mRNA expression was also measured by quantitative RT-PCR using archival material of 33 iCCA patients. To further validate the in vitro data, the Fgfr2 gene was found to be the most consistently overexpressed gene in CCA tissues (Figure 2A and Supplemental Tables S1 and S2). Protein overexpression of the FGFR1 and FGFR2 proteins was evident in paraffin-embedded sections of iCCA (n = 19) (Figure 2, B–D) and tissue sections from HCC (n = 10) (Figure 2E), and liver metastases from colorectal cancer (n = 2) and a hepatic adenoma (n = 1) were used as negative controls. Via immunohistochemistry, FGFR2 expression could be demonstrated in 65% of all iCCAs, whereas FGFR1 was revealed in 30% of all CCAs (Supplemental Table S1). Biochemically, FGFR2 immunoreactivity was observed in CCA both in the cytosol and plasma membrane (Figure 2, B and C). HCC showed no immunoreactivity with the antibody to FGFR2 (Figure 2E) and low immunoreactivity toward FGFR1, consistent with a previous study.
      • Mathur A.
      • Ware C.
      • Davis L.
      • Gazdar A.
      • Pan B.S.
      • Lutterbach B.
      FGFR2 is amplified in the NCI-H716 colorectal cancer cell line and is required for growth and survival.
      Normal liver tissue displayed no immunoreactivity for either FGFR1 or FGFR2. Analysis of Fgfr2 gene expression and clinicopathologic data from CCA patients revealed a good correlation between strong Fgfr2 expression and reduction of tumor differentiation (Figure 2F) (n = 26, grade G1 > G2 > G3; the correlation analysis was performed with the 26 cases that had a clear tumor grade of 33 listed) (Supplemental Table S2), as shown by immunohistochemical staining (Figure 2, B and C). Although our immunohistochemical cohort was mainly composed of G3 CCA tissues, it could not confirm the gene expression correlation data. Nonetheless, these results suggest that FGFR2 might be an important player in CCA tumor progression.
      Figure thumbnail gr2
      Figure 2Presence of fibroblast growth factor receptor (FGFR) members in cholangiocarcinoma (CCA) patients. A: Relative Fgfr mRNA expression (real-time quantitative PCR) in CCA tumors compared with surrounding human liver tissue. All mRNA levels are displayed as fold changes normalized to 1 (mean expression of surrounding tissue). Data are dot plots, with lines representing means (P value versus Fgfr2 by one-tailed t-test). Frequency of positive expression is reported as fold >2 per Fgfr1-4 gene. B: FGFR2 immunohistochemical expression in CCA. Strong, cytoplasmic expression level of FGFR2 is observed in CCA glands embedded in a rich fibrous stroma; an entrapped bile duct (asterisk) shows negative staining (FGFR2 immunostaining). C: A moderate, cytoplasmic expression level of FGFR2 is seen in well-differentiated neoplastic glands of a CCA facing the surrounding liver parenchyma; hepatocytes and inflammatory cells are negative (FGFR2 immunostaining). D: A poorly differentiated CCA showing faint to moderate, cytoplasmic expression of FGFR1 (FGFR1 immunostaining). E: A hepatocellular carcinoma (left) and the surrounding liver parenchyma (right) are negative for FGFR2 (FGFR2 immunostaining). F: Correlation between Fgfr2 expression and tumor differentiation grade (analysis of variance test applied to compare well or moderate with poor). Data are expressed as dot plots, with lines representing means. Data are expressed as means ± SEM (A and F). n = 33 CCA tumors (A) and total (F); n = 18 human liver tissue (A); n = 19 (B); n = 8 per group (F). *P < 0.05 versus Fgfr2 (one-tailed t-test); ††P < 0.01 versus poor differentiation grade (analysis of variance test). Original magnification, ×40 (BE). HCC, hepatocellular carcinoma; iCCA, intrahepatic CCA.

      Effect of DZB on CCA Cell Viability

      To verify the inhibitory effect of DZB on cancer cell growth, CCLP1 and HUCCT1 cells were treated with increasing concentrations of DZB (0.1, 0.3, 1, 3, and 10 μmol/L) for 2, 24, and 72 hours (Figure 3 and Supplemental Table S3). An inhibitory drug effect was recognized that was dose dependent on CCA cell viability, suggesting a potential pharmacologic effect in vitro. In fact, pharmacologic effectiveness was found to vary based on the cell line used. We speculated that the different drug responses of CCA cell lines might be due to FGFR's spectrum of expression (Figure 1A). In addition, to further validate the biological significance of FGFR signaling inhibition in CCA cells, three additional commercially available anti-FGFR inhibitors [namely, TG10052 hydrochloride (0.001 to 0.1 μmol/L), ENMD-2076 (0.001 to 1 μmol/L), and ENMD-2076 tartrate (0.001 μmol/L to 0.1 mmol/L)] were tested.
      • Doukas J.
      • Mahesh S.
      • Umeda N.
      • Kachi S.
      • Akiyama H.
      • Yokoi K.
      • Cao J.
      • Chen Z.
      • Dellamary L.
      • Tam B.
      • Racanelli-Layton A.
      • Hood J.
      • Martin M.
      • Noronha G.
      • Soll R.
      • Campochiaro P.A.
      Topical administration of a multi-targeted kinase inhibitor suppresses choroidal neovascularization and retinal edema.
      • Palanki M.S.
      • Akiyama H.
      • Campochiaro P.
      • Cao J.
      • Chow C.P.
      • Dellamary L.
      • Doukas J.
      • Fine R.
      • Gritzen C.
      • Hood J.D.
      • Hu S.
      • Kachi S.
      • Kang X.
      • Klebansky B.
      • Kousba A.
      • Lohse D.
      • Mak C.C.
      • Martin M.
      • McPherson A.
      • Pathak V.P.
      • Renick J.
      • Soll R.
      • Umeda N.
      • Yee S.
      • Yokoi K.
      • Zeng B.
      • Zhu H.
      • Noronha G.
      Development of prodrug 4-chloro-3-(5-methyl-3-{[4-(2-pyrrolidin-1-ylethoxy)phenyl]amino}-1,2,4-benzotria zin-7-yl)phenyl benzoate (TG100801): a topically administered therapeutic candidate in clinical trials for the treatment of age-related macular degeneration.
      • Fletcher G.C.
      • Brokx R.D.
      • Denny T.A.
      • Hembrough T.A.
      • Plum S.M.
      • Fogler W.E.
      • Sidor C.F.
      • Bray M.R.
      ENMD-2076 is an orally active kinase inhibitor with antiangiogenic and antiproliferative mechanisms of action.
      • Wang X.
      • Sinn A.L.
      • Pollok K.
      • Sandusky G.
      • Zhang S.
      • Chen L.
      • Liang J.
      • Crean C.D.
      • Suvannasankha A.
      • Abonour R.
      • Sidor C.
      • Bray M.R.
      • Farag S.S.
      Preclinical activity of a novel multiple tyrosine kinase and aurora kinase inhibitor, ENMD-2076, against multiple myeloma.
      In both CCA cell lines (CCLP1 and HUCCT1), the latter compounds revealed inhibitory results similar to DZB (Supplemental Figure S2).
      Figure thumbnail gr3
      Figure 3Effect of derazantinib (DZB) on cholangiocarcinoma (CCA) viability. Relative cell viability of CCA cell lines at different concentrations of DZB after 2, 24, and 72 hours of treatment. Cell survival was analyzed using MTT assay. All levels are displayed as percentage of vehicle (V) sample. Data are expressed as means ± SEM. n = 5.

      DZB Inhibitory Effect on the FGFR Pathway in CCA Cells

      Using the agonist FGF7, inhibiting FGFR2 phosphorylation by DZB treatment after receptor activation could progressively reduce phosphorylation in a concentration-dependent manner (Figure 4A and Supplemental Figure S3). DZB inhibited the phosphorylation of FGFR1 to FGFR2 in all CCA cell lines in a dose-dependent manner, indicating a possible inhibitory pharmacologic impact on FGFR phosphorylation/activation.
      Figure thumbnail gr4
      Figure 4Derazantinib (DZB) inhibition of the fibroblast growth factor receptor (FGFR) pathway in cholangiocarcinoma cells. A: After treatment with increasing concentrations of DZB (0.1, 0.3, 1, 3, and 10 μmol/L), CCLP1 cells were stimulated with FGF7, an FGFR receptor agonist. In CCLP1 cells, FGFR activation by FGF7 (0+) is shown to result in major phosphorylation relative to the control (0-) in the absence of DZB treatment. B: DZB inhibits ligand-dependent phosphorylation and activation of the downstream signaling pathways of FGFR. Subconfluent CCLP1 cells were treated with indicated concentrations of DZB for 2 hours, followed by stimulation with FGF7 for 15 minutes. Lysates were prepared, and phosphorylated (p-) and total FGFR2, AKT, extracellular signal-regulated kinase (ERK) 1/2, P38, and FGF2 were detected by immunoblot analysis. β-Actin and vinculin were used as loading controls.
      Although the anti–phospho-FGFR antibody does not discriminate among different phosphorylated FGFRs, because the cell lines express high levels of both FGFR1 and FGFR2, this may support the notion that the reduction of the phospho-FGFR signal was mainly due to the inhibition of phospho-FGFRs 1 and 2 (Figure 4A and Supplemental Figure S3). Only transfection experiments could, therefore, clearly reveal which FGFR member would be relevant for CCA pathogenesis.
      The signal transduction pathways activated downstream of FGFR in the CCA cell lines were examined. By inhibiting FGFR phosphorylation, DZB caused the dephosphorylation of further downstream components, such as ERK1/2, AKT, and P38 (Figure 4B). Furthermore, down-regulation of liver cancer stem-like genes (Supplemental Figure S4) and molecules involved in angiogenesis (FGF2, angiopoietin-1 proteins, and Vegf gene) (Figure 4B and Supplemental Figure S4) was observed. This result suggested that DZB has a downstream effect that leads to tumor disruption. Collectively, this result positively correlates with the in vitro antiproliferative activity of DZB in CCA cell lines.

      DZB Induced Cell Cycle Arrest and Apoptosis

      To explain the decrease in cell viability, the effects of DZB were evaluated on cell cycle and apoptotic responses. Treatment with DZB led to cell accumulation in the G0/G1 phase of the cell cycle in a dose-dependent manner. Concomitantly, the percentage of cells in the S phase was found to be reduced (Figure 5A and Table 1). These data were corroborated by the down-regulation of cyclin D1, together with the overexpression of P27 tumor suppressor (Figure 5B). To evaluate whether, besides cell cycle arrest, the reduction of cell number in the culture was due to an increase in apoptosis, CCLP1 cells were treated with 0.3 and 1 μmol/L of DZB for 72 hours. Furthermore, cleaved poly (ADP-ribose) polymerase, BCL2, and gene expression of caspase1 (Casp1) were assessed by immunoblotting analysis. Notably, a 72-hour treatment with the 0.3 μmol/L dose could induce marked early and late apoptosis. To add, the latter was found to increase further with a higher dose of DZB (1 μmol/L), where<10% of cells were still alive (Figure 5, C and D, and Supplemental Table S4). An apoptotic response was observed in CCLP1 cells after treatment with DZB by an increase in cleaved poly (ADP-ribose) polymerase protein and Casp1 gene and the down-regulation of BCL2 (Figure 5, E and F). Overall, these data indicate a substantial reduction in CCA survival after DZB treatment.
      Figure thumbnail gr5
      Figure 5Derazantinib (DZB) induces G1 cell cycle arrest and apoptosis in cholangiocarcinoma cells. A: CCLP1 cells were treated with 0.3 or 1 μmol/L of DZB or vehicle for 72 hours. Cell cycle profiles were measured by flow cytometric analyses. Gating was performed to exclude cellular debris. Representative dot plots are shown. CCLP1 cells were treated with 0.3 or 1 μmol/L of ARQ 087 or vehicle for 72 hours. B: Western blot analysis was performed for cyclin D1 and P27. Vinculin was used as the loading control. C: CCLP1 cells were treated with 0.3 or 1 μmol/L of DZB or vehicle for 72 hours and used in an annexin V/propidium iodide assay to determine the percentage of cells in early or late apoptosis. Representative dot plots are shown. D: Graph bar summarizes percentage of cells alive and in early and late apoptotic phases. E: CCLP1 cells were treated with 0.3 or 1 μmol/L of DZB or vehicle for 72 hours. Western blot analysis was performed for cleaved poly (ADP-ribose) polymerase (PARP) and BCL-2. Vinculin was used as the loading control. F: Casp1 gene expression after a 72-hour treatment with ARQ 087, 1 μmol/L, or vehicle. Gapdh was used as the internal control. Levels of mRNA are displayed as ΔCT. P value versus vehicle sample by t-test. Data are expressed as means ± SEM (D and F). n = 5 (D); n = 3 (F). **P < 0.01 versus relative vehicle (t-test). FITC, fluorescein isothiocyanate; PI-A, propidium iodide–annexin V.
      Table 1Percentage of Cells in G0/G1, S, and G2 Phases
      PhaseVehicleDZB (0.3 μmol/L)DZB (1 μmol/L)
      G0/G114.9 ± 1.335.0 ± 6.6*63.2 ± 5.4**
      S62.4 ± 1.842.3 ± 4.8*24 ± 2.9**
      G2/M22.7 ± 0.522.7 ± 1.912.8 ± 3.5*
      Data are expressed as means ± SEM. n = 5.
      *P ≤ 0.05, **P ≤ 0.01 versus relative vehicle (t-test).
      DZB, derazantinib.

      DZB Inhibits the Invasive Capacity of CCA Cells in Vitro

      To support the experimental results mentioned above, the negative regulation of DZB on the invasiveness of CCA cells was verified using the Matrigel invasion assay. Figure 6A shows that the invasion of CCA cells through the Matrigel decreased with the 2-hour treated cells (either 0.3 or 1 μmol/L) compared with the vehicle controls. Through a quantitative analysis, the 1 μmol/L dose was demonstrated to be more effective at decreasing the invasiveness of these cells compared with 0.3 μmol/L. Tumor cell invasion decreased >30% in cells treated with 0.3 μmol/L and >90% in cells treated with 1 μmol/L (Figure 6, A and B) regardless of cell viability (Supplemental Figure S5). A similar inhibitory effect was observed after treatment with the commercially available anti-FGFR inhibitors, TG10052 hydrochloride, ENMD-2076, and ENMD-2076 tartrate
      • Doukas J.
      • Mahesh S.
      • Umeda N.
      • Kachi S.
      • Akiyama H.
      • Yokoi K.
      • Cao J.
      • Chen Z.
      • Dellamary L.
      • Tam B.
      • Racanelli-Layton A.
      • Hood J.
      • Martin M.
      • Noronha G.
      • Soll R.
      • Campochiaro P.A.
      Topical administration of a multi-targeted kinase inhibitor suppresses choroidal neovascularization and retinal edema.
      • Palanki M.S.
      • Akiyama H.
      • Campochiaro P.
      • Cao J.
      • Chow C.P.
      • Dellamary L.
      • Doukas J.
      • Fine R.
      • Gritzen C.
      • Hood J.D.
      • Hu S.
      • Kachi S.
      • Kang X.
      • Klebansky B.
      • Kousba A.
      • Lohse D.
      • Mak C.C.
      • Martin M.
      • McPherson A.
      • Pathak V.P.
      • Renick J.
      • Soll R.
      • Umeda N.
      • Yee S.
      • Yokoi K.
      • Zeng B.
      • Zhu H.
      • Noronha G.
      Development of prodrug 4-chloro-3-(5-methyl-3-{[4-(2-pyrrolidin-1-ylethoxy)phenyl]amino}-1,2,4-benzotria zin-7-yl)phenyl benzoate (TG100801): a topically administered therapeutic candidate in clinical trials for the treatment of age-related macular degeneration.
      • Fletcher G.C.
      • Brokx R.D.
      • Denny T.A.
      • Hembrough T.A.
      • Plum S.M.
      • Fogler W.E.
      • Sidor C.F.
      • Bray M.R.
      ENMD-2076 is an orally active kinase inhibitor with antiangiogenic and antiproliferative mechanisms of action.
      • Wang X.
      • Sinn A.L.
      • Pollok K.
      • Sandusky G.
      • Zhang S.
      • Chen L.
      • Liang J.
      • Crean C.D.
      • Suvannasankha A.
      • Abonour R.
      • Sidor C.
      • Bray M.R.
      • Farag S.S.
      Preclinical activity of a novel multiple tyrosine kinase and aurora kinase inhibitor, ENMD-2076, against multiple myeloma.
      (Supplemental Figure S2). Under this premise, the alterations in the expression of epithelial-mesenchymal–associated genes were further analyzed in the presence or absence of DZB treatment. DZB caused a reduction of mesenchymal markers [such as vimentin (Vim)] and a concomitant enhanced expression of the epithelial marker, E-cadherin (Cdh1). For β-catenin (Ctnnb1), only a minimal variation in expression was observed (Figure 6C). Altogether, our results indicate that DZB significantly diminished the invasive capability of CCA.
      Figure thumbnail gr6
      Figure 6Reduction of the invasive capacity of cholangiocarcinoma cells by derazantinib (DZB). A: Cultured CCLP1 cells were pretreated with vehicle or the indicated concentrations of DZB for 2 hours. Thus, cells were detached from the culture dish, and invasiveness was measured in modified Boyden chambers in the absence or presence (upper chambers) of 0.3 or 1 μmol/L of DZB for 18 hours. Thereafter, cells were counted and normalized to invaded vehicle. B: RT-PCR analysis was performed for the epithelial-mesenchymal genes, Cdh1, Ctnnb1, Vim, and Snai1. All mRNA levels are displayed as fold changes normalized to 1 (mean expression of vehicle sample). C: Representative images of invaded CCLP1 cells in the presence or absence of DZB. Data are expressed as means ± SEM (A and B). n = 5 (A); n = 3 (B). **P < 0.01, ***P < 0.001 versus vehicle sample (t-test). Original magnification, ×20 (C).

      DZB Specificity in CCA Cells

      To more rigorously examine the role of single components of the FGFR family in CCA aggressiveness, the effects elicited by loss of single FGFR members were compared. To mimic the effects of receptor selective inhibition, CCLP1 cells were treated with a pool of specific Fgfr1- or Fgfr2-targeting siRNAs (siFGFR1, siFGFr2) that completely abolished receptor expression (Supplemental Figure S6). Single and combination siFgfr1 and siFgfr2 silencing significantly reduced cell proliferation, invasion capability (Figure 7, A–C ), and ERK1/2 phosphorylation (Figure 7D). The notable decrease in ERK1/2 activation in the ERK1/2 signal transduction pathway indicates that it is a key downstream signaling protein associated with FGFRs 1 and 2 in human CCA. These results also indicate that FGFR1 and FGFR2 are both critical drivers of CCA malignancy.
      Figure thumbnail gr7
      Figure 7Effect of fibroblast growth factor receptors' (FGFRs') 1 and 2 silencing on cholangiocarcinoma cells. A: Rate of proliferation of FGFR-silenced cells (siFgfr1, siFgfr2, and siFgfr1 + siFgfr2) relative to control (siCrt), as measured by the crystal violet assay. This test was performed 48 and 72 hours after silencing. Results were normalized to that of siCrt. B: Cultured silenced CCLP1 cells were either not pretreated or pretreated with vehicle or the indicated concentrations of derazantinib (DZB) for 2 hours. Thus, cells were detached from the culture dish, and invasiveness was measured in modified Boyden chambers in the absence or presence (upper chambers) of 0.3 μmol/L DZB for 18 hours. Nontreated cells (NTs) are shown. Cells per field (10 fields) are represented in a graph format. C: Fgfr-silenced CCLP1 cells show sensitivity to DZB compared with siCrt. MTT test was performed after 24 hours of ARQ 087 treatment. Results were displayed as absorbance (Abs) readings. D: Immunoblot analysis was used to detect the levels of phosphorylated extracellular signal-regulated kinase (p-ERK) 1/2 proteins after siRNA Fgfr silencing alone and in combination with ARQ treatment for 24 hours. Data are expressed as means ± SEM (AC). n = 3 (A and C); n = 5 (B). **P < 0.01, ***P < 0.001 versus NT siCrt sample (t-test); ††P ≤ 0.01 versus vehicle sample (t-test).
      To verify the possible synergism of DZB treatment on FGFR-depleted cells, efficacy of drug treatment combined with siRNA on CCA cells was considered in vitro. As expected, DZB treatment did not significantly impact survival of FGFR-silenced CCA cells (Figure 7C); however, DZB mainly acted on FGFR2 receptor, as indicated by the effect of a 0.3 μmol/L dose on invasion capability (Figure 7B). Our results highlighted FGFR2 as an excellent target for chemotherapy with DZB, and this might add to the literature, especially because this molecule is currently under clinical investigation (NCT03230318). Appropriately combining the anti-FGFR inhibitor with an siFGFR1-siFGFR2 tool might be a future strategy for CCA treatment.

      DZB Growth Inhibition in FGFR Dysregulated CCA Cells

      Previous studies of several targeted therapies against solid tumors strongly argue that genetic alterations of target molecules (eg, gene amplification, mutations, or fusion gene) are important hallmarks of pathway addiction and drug response.
      • Farshidfar F.
      • Zheng S.
      • Gingras M.-C.
      • Newton Y.
      • Shih J.
      • Robertson A.G.
      • Hinoue T.
      • Hoadley K.A.
      • Gibb E.A.
      • Roszik J.
      • Covington K.R.
      • Wu C.C.
      • Shinbrot E.
      • Stransky N.
      • Hegde A.
      • Yang J.D.
      • Reznik E.
      • Sadeghi S.
      • Pedamallu C.S.
      • Ojesina A.I.
      • Hess J.M.
      • Auman J.T.
      • Rhie S.K.
      • Bowlby R.
      • Borad M.J.
      Cancer Genome Atlas Network; Zhu AX, Stuart JM, Sander C, Akbani R, Cherniack AD, Deshpande V, Mounajjed T, Foo WC, Torbenson MS, Kleiner DE, Laird PW, Wheeler DA, McRee AJ, Bathe OF, Andersen JB, Bardeesy N, Roberts LR, Kwong LN: Integrative genomic analysis of cholangiocarcinoma identifies distinct IDH-mutant molecular profiles.
      In this respect, FGFR2 fusions represent the most recurrent targetable alteration in CCA patients.
      • Banales J.M.
      • Cardinale V.
      • Carpino G.
      • Marzioni M.
      • Andersen J.B.
      • Invernizzi P.
      • Lind G.E.
      • Folseraas T.
      • Forbes S.J.
      • Fouassier L.
      • Geier A.
      • Calvisi D.F.
      • Mertens J.C.
      • Trauner M.
      • Benedetti A.
      • Maroni L.
      • Vaquero J.
      • Macias R.I.
      • Raggi C.
      • Perugorria M.J.
      • Gaudio E.
      • Boberg K.M.
      • Marin J.J.
      • Alvaro D.
      Expert consensus document: cholangiocarcinoma: current knowledge and future perspectives consensus statement from the European Network for the Study of Cholangiocarcinoma (ENS-CCA).
      To add, a new FGFR2-PPHLN1 fusion has recently been described in CCA.
      • Sia D.
      • Losic B.
      • Moeini A.
      • Cabellos L.
      • Hao K.
      • Revill K.
      • Bonal D.
      • Miltiadous O.
      • Zhang Z.
      • Hoshida Y.
      • Cornella H.
      • Castillo-Martin M.
      • Pinyol R.
      • Kasai Y.
      • Roayaie S.
      • Thung S.N.
      • Fuster J.
      • Schwartz M.E.
      • Waxman S.
      • Cordon-Cardo C.
      • Schadt E.
      • Mazzaferro V.
      • Llovet J.M.
      Massive parallel sequencing uncovers actionable FGFR2-PPHLN1 fusion and ARAF mutations in intrahepatic cholangiocarcinoma.
      Hence, the biological activity of DZB was evaluated in a CCA cell line stably overexpressing the FGFR2-PPHLN1 fusion protein.
      • Sia D.
      • Losic B.
      • Moeini A.
      • Cabellos L.
      • Hao K.
      • Revill K.
      • Bonal D.
      • Miltiadous O.
      • Zhang Z.
      • Hoshida Y.
      • Cornella H.
      • Castillo-Martin M.
      • Pinyol R.
      • Kasai Y.
      • Roayaie S.
      • Thung S.N.
      • Fuster J.
      • Schwartz M.E.
      • Waxman S.
      • Cordon-Cardo C.
      • Schadt E.
      • Mazzaferro V.
      • Llovet J.M.
      Massive parallel sequencing uncovers actionable FGFR2-PPHLN1 fusion and ARAF mutations in intrahepatic cholangiocarcinoma.
      Notably, FGFR2-PPHLN1 cells showed enhanced sensitivity to DZB compared with their parental cell line transfected with the empty vector (Figure 8, A and B , and Supplemental Figure S7). In particular, significant inhibition of the migratory capability (Figure 8B) and expression of epithelial-mesenchymal genes (Figure 8C) were observed only in HUCCT1 cells expressing the fusion protein. The above findings support the transforming and oncogenic potential of FGFR2-PPHLN1 fusion protein and the possible efficacy of DZB in the clinical management of CCA patients harboring FGFR2 rearrangements.
      Figure thumbnail gr8
      Figure 8Fibroblast growth factor receptor (FGFR) 2 fusions represent a potential therapeutic target. A: The HUCCT1 cell line transfected to overexpress FGFR2-periphilin 1 (PPHLN1) has increased sensitivity to derazantinib (DZB) compared with its parental cell line transfected with the empty vector. MTT test was performed after a 24-hour DZB treatment. Results were normalized to vehicle (dimethyl sulfoxide). B: DZB (0.1 μmol/L) reduces cell migration only in the HUCCT1 cell line expressing FGFR2-PPHLN1. Cultured FGFR2-PPHNL1 cells were pretreated with vehicle or the indicated concentrations of DZB for 2 hours. Thus, cells were detached from the culture dish, and invasiveness was measured in modified Boyden chambers in the absence or presence (upper chambers) of 0.1 μmol/L DZB for 18 hours. Invaded cells per field (10 fields) are represented in graph. C: Relative expression of transcript-encoding epithelial-mesenchymal genes in HUCCT1 FGFR2-PPHNL1 and relative empty vector cells after DZB treatment. Gapdh was used as the internal control. All mRNA levels are presented as fold changes normalized to 1 (mean expression of empty vector). Data are expressed as means ± SEM (AC). n = 3 (A); n = 5 (B). *P < 0.05, **P < 0.01, and ***P < 0.001 versus empty vector sample (t-test); †††P < 0.001 versus vehicle of the HUCCT1 FGFR2-PPHLN1 cells (t-test).

      Discussion

      Our findings demonstrate the in vitro preclinical activity of DZB in human iCCA. As previously reported,
      • Yu Y.
      • Hall T.
      • Eathiraj S.
      • Wick M.J.
      • Schwartz B.
      • Abbadessa G.
      In-vitro and in-vivo combined effect of ARQ 092, an AKT inhibitor, with ARQ 087, a FGFR inhibitor.
      • Chila R.
      • Hall G.T.
      • Abbadessa G.
      • Broggini M.
      • Damia G.
      Multi-chemotherapeutic schedules containing the pan-FGFR inhibitor ARQ 087 are safe and show antitumor activity in different xenograft models.
      • Hall T.G.
      • Yu Y.
      • Eathiraj S.
      • Wang Y.
      • Savage R.E.
      • Lapierre J.M.
      • Schwartz B.
      • Abbadessa G.
      Preclinical activity of ARQ 087, a novel inhibitor targeting FGFR dysregulation.
      • Papadopoulos K.P.
      • El-Rayes B.F.
      • Tolcher A.W.
      • Patnaik A.
      • Rasco D.W.
      • Harvey R.D.
      • LoRusso P.M.
      • Sachdev J.C.
      • Abbadessa G.
      • Savage R.E.
      • Hall T.
      • Schwartz B.
      • Wang Y.
      • Kazakin J.
      • Shaib W.L.
      A phase 1 study of ARQ 087, an oral pan-FGFR inhibitor in patients with advanced solid tumours.
      DZB inhibited FGFR kinases by an ATP competitive mechanism and delayed FGFR activation by inhibiting its autophosphorylation and the phosphorylated active kinase. By using multiple cell systems' ectopically expressing FGFR receptors, DZB was recently found to inhibit cell proliferation as a functional consequence of FGFR phosphorylation inhibition in a preferential manner, in the order, FGFR2 ≫ FGFR1/FGFR3 ≫ FGFR4.
      Our study showed that treating FGFR-overexpressing CCA cells with DZB affected their signal transduction pathways, thereby impacting tumor cell proliferation, death, and invasiveness. DZB was tested in a phase 1/2 clinical trial (NCT03230318) and demonstrated meaningful clinical benefits for patients with iCCA harboring tumors with a dysregulated FGFR pathway, specifically FGFR2 fusions.
      Our study confirmed that the genetic and protein overexpression levels of FGFRs 1 to 4 are important in iCCA.
      • Rizvi S.
      • Yamada D.
      • Hirsova P.
      • Bronk S.F.
      • Werneburg N.W.
      • Krishnan A.
      • Salim W.
      • Zhang L.
      • Trushina E.
      • Truty M.J.
      • Gores G.J.
      A Hippo and fibroblast growth factor receptor autocrine pathway in cholangiocarcinoma.
      Among FGF receptors, CCA cells expressed high levels of FGFR2, as shown by the immunohistochemical and gene expression data. In contrast, both HCC cells and nonmalignant cholangiocytes (normal human cholangiocyte cells) expressed low levels of FGFR, suggesting that FGFR may have a role in the malignant transformation of tumor cholangiocytes and implying that FGFRs might represent potential CCA-specific markers.
      In vitro studies on malignant CCA cell lines (CCLP1 and HUCCT1) overexpressing various FGFR subtypes revealed the antiproliferative response that occurs toward DZB. DZB dose dependently inhibited proliferation and cell line specificity. Furthermore, its impact on CCA proliferation was found to be a result of the deactivation of ERK1/2 and AKT signaling pathways and activation of apoptotic and cell growth arrest signaling. Notably, invasion was reduced by a 2-hour pretreatment with DZB, as shown by the number of invaded cells through Matrigel devices; and this result was also corroborated by the down-regulation of key epithelial-mesenchymal genes.
      In summary, we have presented a mechanistic description of a novel kinase inhibitor that displays potent in vitro activity in FGFR-driven models, thereby confirming its good drug-like properties. To add, we demonstrated the activity of DZB on CCA cell lines expressing Fgfr mRNA and protein and showed that DZB might be active in tumors expressing FGFR and is independent of FGFR fusion. DZB is currently in clinical trials for CCA. Indeed, data from a phase 1 clinical trial appears to align with the data acquired herein, as tumor response was observed in CCA with amplification, and not only fusions.
      • Papadopoulos K.P.
      • El-Rayes B.F.
      • Tolcher A.W.
      • Patnaik A.
      • Rasco D.W.
      • Harvey R.D.
      • LoRusso P.M.
      • Sachdev J.C.
      • Abbadessa G.
      • Savage R.E.
      • Hall T.
      • Schwartz B.
      • Wang Y.
      • Kazakin J.
      • Shaib W.L.
      A phase 1 study of ARQ 087, an oral pan-FGFR inhibitor in patients with advanced solid tumours.
      DZB can be active in tumors that express both FGFR1 and FGFR2 (and not limited to FGFR2) based on the CCLP1 data. Currently, there are promising data on the antitumor activity of FGFR2 inhibitory compounds, but these might be limited to CCA with FGFR2 fusions (ie, not amplifications).
      • Javle M.
      • Lowery M.
      • Shroff R.T.
      • Weiss K.H.
      • Springfeld C.
      • Borad M.J.
      • Ramanathan R.K.
      • Goyal L.
      • Sadeghi S.
      • Macarulla T.
      • El-Khoueiry A.
      • Kelley R.K.
      • Borbath I.
      • Choo S.P.
      • Oh D.Y.
      • Philip P.A.
      • Chen L.T.
      • Reungwetwattana T.
      • Van Cutsem E.
      • Yeh K.H.
      • Ciombor K.
      • Finn R.S.
      • Patel A.
      • Sen S.
      • Porter D.
      • Isaacs R.
      • Zhu A.X.
      • Abou-Alfa G.K.
      • Bekaii-Saab T.
      Phase II study of BGJ398 in patients with FGFR-altered advanced cholangiocarcinoma.
      Thus, identifying novel compounds with a broader spectrum of activity could be interesting.
      Our future plans include testing the inhibitory effects of DZB on xenografts produced from cultured CCA cell lines injected into immunosuppressed mice. This will be performed to determine the effects of FGFR inhibition on human CCA formation and growth in vivo.

      Acknowledgments

      We thank Dr. Snorri S. Thorgeirsson, Valentina Factor, and Elizabeth A. Conner (Center for Cancer Research, National Cancer Institute, NIH, Bethesda, MD) for providing the hepatocarcinoma cell lines (PLC/PRF/5, Huh7, and HepG2); Dr. Anthony J. Demetris (University of Pittsburgh, Pittsburgh, PA) for cholangiocarcinoma cell lines (HUCCT1 and CCLP1); Dr. Jesus M. Banales for the human nonmalignant cholangiocyte cell line; and Prof. Vincenzo Mazzaferro (IRCCS Foundation National Cancer Institute, Milan, Italy) for the stable HUCCT1 cell line overexpressing fibroblast growth factor receptor 2-PPHLN1 fusion protein.
      C.R. and P.I. designed the study and wrote the manuscript; C.R., K.F., M.P., P.B., M.C., E.F., N.N., A.G., T.H., A.D., L.D.T., M.R., F.M., S.G., E.R., C.P.-N., J.M.B., P.O., A.G., A.E., G.A., M.D.B., S.B., V.M., and F.M. provided materials, performed the experiments, collected the data, and analyzed the results; C.R., G.A., and P.I. supervised the project and critically revised the manuscript; all authors have read and approved the final manuscript.

      Supplemental Data

      Figure thumbnail figs1
      Supplemental Figure S1Expression of fibroblast growth factor receptor (FGFR) members in hepatocellular carcinoma (HCC) cells. Relative expression of transcript-encoding receptors for Fgfr1-4 in HCC cells. Gapdh was used as an internal control. All mRNA levels are presented as fold changes normalized to 1 (mean expression of normal hepatocytes). Data are expressed as means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001 versus normal hepatocytes (t-test).
      Figure thumbnail figs2
      Supplemental Figure S2Effect of anti–fibroblast growth factor receptor inhibitors on cholangiocarcinoma (CCA) cells. A: Relative cell viability of CCA cell lines at different concentrations of TG10052 hydrochloride (0.001 to 0.1 μmol/L), ENMD-2076 (0.001 to 1 μmol/L), and ENMD-2076 tartrate (0.001 μmol/L to 0.1 mmol/L) after 24-, 48-, and 72-hour treatment in HUCCT1 cells. P values versus vehicle sample by t-test. B: The 24-hour treatment for CCLP1 cells. Cell survival was analyzed using MTT assay. All levels displayed as percentage of vehicle sample. C: Reduction of invasive capacity of CCA cells. Cultured CCLP1 cells were pretreated with vehicle or the indicated concentrations of TG10052, ENMD-2076, and ENMD-2076 tartrate for 2 hours. Thus, cells were detached from the culture dish, and invasiveness was measured in modified Boyden chambers in absence or presence (upper chambers) of TG10052, ENMD-2076, and ENMD-2076 tartrate for 18 hours. Thereafter, cells were counted and normalized to invaded vehicle. Data are expressed as means ± SEM (AC). n = 5 (B and C). **P < 0.01, ***P < 0.001 versus relative vehicle sample (t-test).
      Figure thumbnail figs3
      Supplemental Figure S3Derazantinib(DZB) inhibition of fibroblast growth factor receptor (FGFR) pathway. HUCCT1 cells were treated with the indicated concentrations of ARQ 087 for 2 hours, and phosphorylated FGFR (p-FGFR) was assessed by Western blot analyses. β-Actin was used as a loading control.
      Figure thumbnail figs4
      Supplemental Figure S4Effect of derazantinib (DZB) on oncogenesis-associated genes. A: CCLP1 cells were treated with indicated concentrations of DZB for 72 hours. Relative expression levels of oncogenic and cancer stem-like genes, such as CD13 (Anpep), Prom1, Myc, and Sox2, were tested by quantitative RT-PCR. Gapdh was used as an internal control. All mRNA levels displayed as fold changes normalized to 1 (mean expression of vehicle sample). B: Angiopoietin-1 (ANGIO-1) was assessed using Western blot analysis. Vinculin was used as a loading control. C: Relative Vegf gene expression. Gapdh was used as an internal control. All mRNA levels displayed as fold changes normalized to 1 (mean expression of vehicle sample). Data are expressed as means ± SEM (A and C). n = 3 (A and C). *P < 0.05 versus vehicle sample (t-test).
      Figure thumbnail figs5
      Supplemental Figure S5Viability of invaded CCLP1 in presence of derazantinib (DZB). After 2 hours of pretreatment, cells were plated in upper chambers with 0.3 or 1 μmol/L of DZB or vehicle for 18 hours. Thereafter, alive CCLP1 cells were counterstained with crystal violet and normalized to total number of cells. Data are expressed as means ± SEM. n = 5.
      Figure thumbnail figs6
      Supplemental Figure S6Effect of fibroblast growth factor receptor (FGFR) 1 and 2 knockdown on CCLP1 cells. Levels of Fgfr1 (A) and Fgfr2 (B) after siRNA transfection (siFgfr1 and siFgfr2, respectively) were detected by quantitative RT-PCR. Gapdh was used as an internal control. All mRNA levels displayed as fold changes normalized to 1 (mean expression of vehicle sample). Data are expressed as means ± SEM (A and B). n = 4 (A and B). **P < 0.01, ***P < 0.001 versus control targeting siRNA (siCrt) sample (t-test).
      Figure thumbnail figs7
      Supplemental Figure S7Fusion gene validation in HUCCT1 cells. Levels of Fgfr2-Pphln1 were detected by quantitative RT-PCR. Gapdh was used as an internal control. All mRNA levels displayed as fold changes normalized to 1 (mean expression of vehicle sample). Data are expressed as means ± SEM. n = 4. ***P < 0.001 versus empty vector sample (t-test).

      References

        • Kumar M.
        • Zhao X.
        • Wang X.W.
        Molecular carcinogenesis of hepatocellular carcinoma and intrahepatic cholangiocarcinoma: one step closer to personalized medicine?.
        Cell Biosci. 2011; 1: 5
        • Leyva-Illades D.
        • McMillin M.
        • Quinn M.
        • Demorrow S.
        Cholangiocarcinoma pathogenesis: role of the tumor microenvironment.
        Transl Gastrointest Cancer. 2012; 1: 71-80
        • Shaib Y.H.
        • El-Serag H.B.
        • Davila J.A.
        • Morgan R.
        • McGlynn K.A.
        Risk factors of intrahepatic cholangiocarcinoma in the United States: a case-control study.
        Gastroenterology. 2005; 128: 620-626
        • Zabron A.
        • Edwards R.J.
        • Khan S.A.
        The challenge of cholangiocarcinoma: dissecting the molecular mechanisms of an insidious cancer.
        Dis Model Mech. 2013; 6: 281-292
        • Aljiffry M.
        • Walsh M.J.
        • Molinari M.
        Advances in diagnosis, treatment and palliation of cholangiocarcinoma: 1990-2009.
        World J Gastroenterol. 2009; 15: 4240-4262
        • Glimelius B.
        • Hoffman K.
        • Sjoden P.O.
        • Jacobsson G.
        • Sellstrom H.
        • Enander L.K.
        • Linne T.
        • Svensson C.
        Chemotherapy improves survival and quality of life in advanced pancreatic and biliary cancer.
        Ann Oncol. 1996; 7: 593-600
        • Hubbard S.R.
        • Till J.H.
        Protein tyrosine kinase structure and function.
        Annu Rev Biochem. 2000; 69: 373-398
        • Heldin C.H.
        Dimerization of cell surface receptors in signal transduction.
        Cell. 1995; 80: 213-223
        • Pawson T.
        • Nash P.
        Protein-protein interactions define specificity in signal transduction.
        Genes Dev. 2000; 14: 1027-1047
        • Turner N.
        • Grose R.
        Fibroblast growth factor signalling: from development to cancer.
        Nat Rev Cancer. 2010; 10: 116-129
        • Beenken A.
        • Mohammadi M.
        The FGF family: biology, pathophysiology and therapy.
        Nat Rev Drug Discov. 2009; 8: 235-253
        • Schlessinger J.
        Cell signaling by receptor tyrosine kinases.
        Cell. 2000; 103: 211-225
        • Banales J.M.
        • Cardinale V.
        • Carpino G.
        • Marzioni M.
        • Andersen J.B.
        • Invernizzi P.
        • Lind G.E.
        • Folseraas T.
        • Forbes S.J.
        • Fouassier L.
        • Geier A.
        • Calvisi D.F.
        • Mertens J.C.
        • Trauner M.
        • Benedetti A.
        • Maroni L.
        • Vaquero J.
        • Macias R.I.
        • Raggi C.
        • Perugorria M.J.
        • Gaudio E.
        • Boberg K.M.
        • Marin J.J.
        • Alvaro D.
        Expert consensus document: cholangiocarcinoma: current knowledge and future perspectives consensus statement from the European Network for the Study of Cholangiocarcinoma (ENS-CCA).
        Nat Rev Gastroenterol Hepatol. 2016; 13: 261-280
        • Yu Y.
        • Hall T.
        • Eathiraj S.
        • Wick M.J.
        • Schwartz B.
        • Abbadessa G.
        In-vitro and in-vivo combined effect of ARQ 092, an AKT inhibitor, with ARQ 087, a FGFR inhibitor.
        Anticancer Drugs. 2017; 28: 503-513
        • Chila R.
        • Hall G.T.
        • Abbadessa G.
        • Broggini M.
        • Damia G.
        Multi-chemotherapeutic schedules containing the pan-FGFR inhibitor ARQ 087 are safe and show antitumor activity in different xenograft models.
        Transl Oncol. 2017; 10: 153-157
        • Hall T.G.
        • Yu Y.
        • Eathiraj S.
        • Wang Y.
        • Savage R.E.
        • Lapierre J.M.
        • Schwartz B.
        • Abbadessa G.
        Preclinical activity of ARQ 087, a novel inhibitor targeting FGFR dysregulation.
        PLoS One. 2016; 11: e0162594
        • Papadopoulos K.P.
        • El-Rayes B.F.
        • Tolcher A.W.
        • Patnaik A.
        • Rasco D.W.
        • Harvey R.D.
        • LoRusso P.M.
        • Sachdev J.C.
        • Abbadessa G.
        • Savage R.E.
        • Hall T.
        • Schwartz B.
        • Wang Y.
        • Kazakin J.
        • Shaib W.L.
        A phase 1 study of ARQ 087, an oral pan-FGFR inhibitor in patients with advanced solid tumours.
        Br J Cancer. 2017; 117: 1592-1599
        • Mazzaferro V.
        • El-Rayes B.F.
        • Droz Dit Busset M.
        • Cotsoglou C.
        • Harris W.P.
        • Damjanov N.
        • Masi G.
        • Rimassa L.
        • Personeni N.
        • Braiteh F.
        • Zagonel V.
        • Papadopoulos K.P.
        • Hall T.
        • Wang Y.
        • Schwartz B.
        • Kazakin J.
        • Bhoori S.
        • de Braud F.
        • Shaib W.L.
        Derazantinib (ARQ 087) in advanced or inoperable FGFR2 gene fusion-positive intrahepatic cholangiocarcinoma.
        Br J Cancer. 2019; 120: 165-171
        • Miyagiwa M.
        • Ichida T.
        • Tokiwa T.
        • Sato J.
        • Sasaki H.
        A new human cholangiocellular carcinoma cell line (HuCC-T1) producing carbohydrate antigen 19/9 in serum-free medium.
        In Vitro Cell Dev Biol. 1989; 25: 503-510
        • Han C.
        • Wu T.
        Cyclooxygenase-2-derived prostaglandin E2 promotes human cholangiocarcinoma cell growth and invasion through EP1 receptor-mediated activation of the epidermal growth factor receptor and Akt.
        J Biol Chem. 2005; 280: 24053-24063
        • Shimizu Y.
        • Demetris A.J.
        • Gollin S.M.
        • Storto P.D.
        • Bedford H.M.
        • Altarac S.
        • Iwatsuki S.
        • Herberman R.B.
        • Whiteside T.L.
        Two new human cholangiocarcinoma cell lines and their cytogenetics and responses to growth factors, hormones, cytokines or immunologic effector cells.
        Int J Cancer. 1992; 52: 252-260
        • Banales J.M.
        • Saez E.
        • Uriz M.
        • Sarvide S.
        • Urribarri A.D.
        • Splinter P.
        • Tietz Bogert P.S.
        • Bujanda L.
        • Prieto J.
        • Medina J.F.
        • LaRusso N.F.
        Up-regulation of microRNA 506 leads to decreased Cl-/HCO3- anion exchanger 2 expression in biliary epithelium of patients with primary biliary cirrhosis.
        Hepatology. 2012; 56: 687-697
        • Raggi C.
        • Factor V.M.
        • Seo D.
        • Holczbauer A.
        • Gillen M.C.
        • Marquardt J.U.
        • Andersen J.B.
        • Durkin M.
        • Thorgeirsson S.S.
        Epigenetic reprogramming modulates malignant properties of human liver cancer.
        Hepatology. 2014; 59: 2251-2262
        • Sia D.
        • Losic B.
        • Moeini A.
        • Cabellos L.
        • Hao K.
        • Revill K.
        • Bonal D.
        • Miltiadous O.
        • Zhang Z.
        • Hoshida Y.
        • Cornella H.
        • Castillo-Martin M.
        • Pinyol R.
        • Kasai Y.
        • Roayaie S.
        • Thung S.N.
        • Fuster J.
        • Schwartz M.E.
        • Waxman S.
        • Cordon-Cardo C.
        • Schadt E.
        • Mazzaferro V.
        • Llovet J.M.
        Massive parallel sequencing uncovers actionable FGFR2-PPHLN1 fusion and ARAF mutations in intrahepatic cholangiocarcinoma.
        Nat Commun. 2015; 6: 6087
        • Petta S.
        • Valenti L.
        • Marra F.
        • Grimaudo S.
        • Tripodo C.
        • Bugianesi E.
        • Camma C.
        • Cappon A.
        • Di Marco V.
        • Di Maira G.
        • Dongiovanni P.
        • Rametta R.
        • Gulino A.
        • Mozzi E.
        • Orlando E.
        • Maggioni M.
        • Pipitone R.M.
        • Fargion S.
        • Craxi A.
        MERTK rs4374383 polymorphism affects the severity of fibrosis in non-alcoholic fatty liver disease.
        J Hepatol. 2016; 64: 682-690
        • Bonacchi A.
        • Romagnani P.
        • Romanelli R.G.
        • Efsen E.
        • Annunziato F.
        • Lasagni L.
        • Francalanci M.
        • Serio M.
        • Laffi G.
        • Pinzani M.
        • Gentilini P.
        • Marra F.
        Signal transduction by the chemokine receptor CXCR3: activation of Ras/ERK, Src, and phosphatidylinositol 3-kinase/Akt controls cell migration and proliferation in human vascular pericytes.
        J Biol Chem. 2001; 276: 9945-9954
        • Rizvi S.
        • Yamada D.
        • Hirsova P.
        • Bronk S.F.
        • Werneburg N.W.
        • Krishnan A.
        • Salim W.
        • Zhang L.
        • Trushina E.
        • Truty M.J.
        • Gores G.J.
        A Hippo and fibroblast growth factor receptor autocrine pathway in cholangiocarcinoma.
        J Biol Chem. 2016; 291: 8031-8047
        • Rovida E.
        • Di Maira G.
        • Tusa I.
        • Cannito S.
        • Paternostro C.
        • Navari N.
        • Vivoli E.
        • Deng X.
        • Gray N.S.
        • Esparis-Ogando A.
        • David E.
        • Pandiella A.
        • Dello Sbarba P.
        • Parola M.
        • Marra F.
        The mitogen-activated protein kinase ERK5 regulates the development and growth of hepatocellular carcinoma.
        Gut. 2015; 64: 1454-1465
        • Rovida E.
        • Navari N.
        • Caligiuri A.
        • Dello Sbarba P.
        • Marra F.
        ERK5 differentially regulates PDGF-induced proliferation and migration of hepatic stellate cells.
        J Hepatol. 2008; 48: 107-115
        • Raggi C.
        • Correnti M.
        • Sica A.
        • Andersen J.B.
        • Cardinale V.
        • Alvaro D.
        • Chiorino G.
        • Forti E.
        • Glaser S.
        • Alpini G.
        • Destro A.
        • Sozio F.
        • Di Tommaso L.
        • Roncalli M.
        • Banales J.M.
        • Coulouarn C.
        • Bujanda L.
        • Torzilli G.
        • Invernizzi P.
        Cholangiocarcinoma stem-like subset shapes tumor-initiating niche by educating associated macrophages.
        J Hepatol. 2017; 66: 102-115
        • Mathur A.
        • Ware C.
        • Davis L.
        • Gazdar A.
        • Pan B.S.
        • Lutterbach B.
        FGFR2 is amplified in the NCI-H716 colorectal cancer cell line and is required for growth and survival.
        PLoS One. 2014; 9: e98515
        • Doukas J.
        • Mahesh S.
        • Umeda N.
        • Kachi S.
        • Akiyama H.
        • Yokoi K.
        • Cao J.
        • Chen Z.
        • Dellamary L.
        • Tam B.
        • Racanelli-Layton A.
        • Hood J.
        • Martin M.
        • Noronha G.
        • Soll R.
        • Campochiaro P.A.
        Topical administration of a multi-targeted kinase inhibitor suppresses choroidal neovascularization and retinal edema.
        J Cell Physiol. 2008; 216: 29-37
        • Palanki M.S.
        • Akiyama H.
        • Campochiaro P.
        • Cao J.
        • Chow C.P.
        • Dellamary L.
        • Doukas J.
        • Fine R.
        • Gritzen C.
        • Hood J.D.
        • Hu S.
        • Kachi S.
        • Kang X.
        • Klebansky B.
        • Kousba A.
        • Lohse D.
        • Mak C.C.
        • Martin M.
        • McPherson A.
        • Pathak V.P.
        • Renick J.
        • Soll R.
        • Umeda N.
        • Yee S.
        • Yokoi K.
        • Zeng B.
        • Zhu H.
        • Noronha G.
        Development of prodrug 4-chloro-3-(5-methyl-3-{[4-(2-pyrrolidin-1-ylethoxy)phenyl]amino}-1,2,4-benzotria zin-7-yl)phenyl benzoate (TG100801): a topically administered therapeutic candidate in clinical trials for the treatment of age-related macular degeneration.
        J Med Chem. 2008; 51: 1546-1559
        • Fletcher G.C.
        • Brokx R.D.
        • Denny T.A.
        • Hembrough T.A.
        • Plum S.M.
        • Fogler W.E.
        • Sidor C.F.
        • Bray M.R.
        ENMD-2076 is an orally active kinase inhibitor with antiangiogenic and antiproliferative mechanisms of action.
        Mol Cancer Ther. 2011; 10: 126-137
        • Wang X.
        • Sinn A.L.
        • Pollok K.
        • Sandusky G.
        • Zhang S.
        • Chen L.
        • Liang J.
        • Crean C.D.
        • Suvannasankha A.
        • Abonour R.
        • Sidor C.
        • Bray M.R.
        • Farag S.S.
        Preclinical activity of a novel multiple tyrosine kinase and aurora kinase inhibitor, ENMD-2076, against multiple myeloma.
        Br J Haematol. 2010; 150: 313-325
        • Farshidfar F.
        • Zheng S.
        • Gingras M.-C.
        • Newton Y.
        • Shih J.
        • Robertson A.G.
        • Hinoue T.
        • Hoadley K.A.
        • Gibb E.A.
        • Roszik J.
        • Covington K.R.
        • Wu C.C.
        • Shinbrot E.
        • Stransky N.
        • Hegde A.
        • Yang J.D.
        • Reznik E.
        • Sadeghi S.
        • Pedamallu C.S.
        • Ojesina A.I.
        • Hess J.M.
        • Auman J.T.
        • Rhie S.K.
        • Bowlby R.
        • Borad M.J.
        Cancer Genome Atlas Network; Zhu AX, Stuart JM, Sander C, Akbani R, Cherniack AD, Deshpande V, Mounajjed T, Foo WC, Torbenson MS, Kleiner DE, Laird PW, Wheeler DA, McRee AJ, Bathe OF, Andersen JB, Bardeesy N, Roberts LR, Kwong LN: Integrative genomic analysis of cholangiocarcinoma identifies distinct IDH-mutant molecular profiles.
        Cell Rep. 2017; 18: 2780-2794
        • Javle M.
        • Lowery M.
        • Shroff R.T.
        • Weiss K.H.
        • Springfeld C.
        • Borad M.J.
        • Ramanathan R.K.
        • Goyal L.
        • Sadeghi S.
        • Macarulla T.
        • El-Khoueiry A.
        • Kelley R.K.
        • Borbath I.
        • Choo S.P.
        • Oh D.Y.
        • Philip P.A.
        • Chen L.T.
        • Reungwetwattana T.
        • Van Cutsem E.
        • Yeh K.H.
        • Ciombor K.
        • Finn R.S.
        • Patel A.
        • Sen S.
        • Porter D.
        • Isaacs R.
        • Zhu A.X.
        • Abou-Alfa G.K.
        • Bekaii-Saab T.
        Phase II study of BGJ398 in patients with FGFR-altered advanced cholangiocarcinoma.
        J Clin Oncol. 2018; 36: 276-282