Dual Inhibition of Hedgehog and c-Met Pathways for Pancreatic Cancer Treatment

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is one of the most chemotherapy- and radiotherapy-resistant tumors. The c-Met and Hedgehog (Hh) pathways have been shown previously by our group to be key regulatory pathways in the primary tumor growth and metastases formation. Targeting both the HGF/c-Met and Hh pathways has shown promising results in preclinical studies; however, the benefits were not readily translated into clinical trials with PDAC patients. In this study, utilizing mouse models of PDAC, we showed that inhibition of either HGF/c-Met or Hh pathways sensitize the PDAC tumors to gemcitabine, resulting in decreased primary tumor volume as well as significant reduction of metastatic tumor burden. However, prolonged treatment of single HGF/c-Met or Hh inhibitor leads to resistance to these single inhibitors, likely because the single c-Met treatment leads to enhanced expression of Shh, and vice versa. Targeting both the HGF/c-Met and Hh pathways simultaneously overcame the resistance to the single-inhibitor treatment and led to a more potent antitumor effect in combination with the chemotherapy treatment. Mol Cancer Ther; 16(11); 2399-409. ©2017 AACR.

Figure 1. Short-term inhibition of HGF/c-Met or Hh signaling enhances the sensitivity of PDA tumors to gemcitabine in transgenic and orthotopic mouse models of PDA

A. Schematic representation of 1-week treatment regimen in the transgenic (KPC) and orthotopic mouse models of PDA. Day 0 represents the day of the orthotopic implantation of primary pancreatic tumors. Mice in the orthotopic model were subjected to ultrasound on postoperative day 5 to establish baseline tumor data. Daily treatment by oral gavage with inhibitor(s) or vehicle control was initiated on the day after ultrasound. In the KPC mouse model, ultrasound was performed 1 day prior to treatment initiation. In both models gemcitabine was administered bi-weekly by intraperitoneal injection. Second ultrasound was performed on the last day of treatment. Mice from all groups were euthanized on the last day of treatment and the panreata and livers were harvested for analysis. B and C. The KPC (panel B) and orthotopic (panel C) mouse models of PDA were treated with daily Hh and/or HGF/c-Met inhibitors and bi-weekly gemcitabine as shown in Panel A. Tumor volumes were obtained at baseline and on the last day of treatment. The data show tumor volume fold changes calculated as a ratio by comparison of the post-treatment tumor volume to the baseline tumor volume. Data is representative of 1 experiment. Mice that died before the completion of the planned treatment or whose quality of tumor was not adequate for analysis due to necrosis were excluded. ns- not significant *p<0.05, **p<0.01, Gem-gemcitabine (n=8), Hh-Hh inhibitor + Gem (n=9), c-Met- HGF/c-Met inhibitor + Gem (n=7), DMSO-vehicle control (n=14), Hh+c-Met+Gem (n=8). ***p<0.001, ****p<0.0001 (unpaired student t-test).

Figure 2. Combination of Hh and/or HGF/c-Met inhibitor(s) with gemcitabine in short term treatment regimen results in enhancement tumoral cell death in the transgenic (KPC) and orthotopic mouse models

A and B. The KPC (panel A) and orthotopic (panel B) mouse models of PDA were treated with Hh and/or HGF/c-Met inhibitor(s) and bi-weekly gemcitabine as shown in Figure 1A. Primary tumors from both mouse models were harvested on the last day of treatment and subjected to TUNEL staining for apoptotic cells. Representative images for each treatment group in both models are shown. The images are representative of 1 experiment. Mice that died before the completion of the planned treatment or whose quality of tumor was not adequate for analysis due to necrosis were excluded. Gem-gemcitabine (n=8), Hh-Hh inhibitor + Gem (n=9), c-Met- HGF/c-Met inhibitor + Gem (n=7), DMSO-vehicle control (n=14), Hh+c-Met+Gem (n=8). Scale bars, 200 µm.

Figure 3. Prolonged combination treatment of transgenic mouse model (KPC) with Hh and HGF/c-Met inhibitors in combination with gemcitabine leads to reduction in primary tumor volume and increased apoptosis

A. Schematic representation of three-week treatment regimen in the KPC mouse model of PDA. Baseline tumor volume was determined one day before treatment, following by weekly ultrasounds until the last day of treatment. Mice were treated daily by oral gavage with Hh and/or HGF/c-Met inhibitors or vehicle control. Gemcitabine was administered bi-weekly via intraperitoneal injection. B. The change in tumor volume (calculated as ratio between week 3 and baseline tumor volume) is shown. C. Semi-quantification of TUNEL staining for apoptotic cells in KPC mouse model after 3 weeks of treatment (as shown in panel A). The scoring method used score between 0 and 3, where 0 is no positive staining and 3 is high positive staining. The data is representative of 1 experiment. Mice that died before the completion of the planned treatment or whose quality of tumor was not adequate for analysis due to necrosis were excluded. Gem-gemcitabine (n=7), Hh-Hh inhibitor + Gem (n=6), c-Met- HGF/c-Met inhibitor + Gem (n=5), DMSO-vehicle control (n=9), Hh+c-Met+Gem (n=4).. ns- not significant *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (unpaired student t-test).

Figure 4. Prolonged combination treatment of orthotopic mouse model with \Hh and HGF/c-Met inhibitors in combination gemcitabine leads to reduction in primary tumor volume and metastatic burden

A. Schematic representation of two-week treatment regimen in the orthotopic mouse model of PDA. Baseline tumor volume was determined on postoperative day 5, following by weekly ultrasounds until the last day of treatment. Mice were treated daily beginning one day after the first ultrasound by oral gavage with Hh and/or HGF/c-Met inhibitors or vehicle control. Gemcitabine was administered bi-weekly via intraperitoneal injection. In the metastasis study (panel C) livers, lungs, gut and peritoneum were harvested for gross and histological analysis of metastases. B. The change in tumor volume (calculated as ratio between week 3 and baseline tumor volume) is shown. Data is representative of 1 experiment. C. Summary of total numbers of metastases formed (gross and histological) in each group. The data is representative of 1 experiment. Mice that died before the completion of the planned treatment or whose quality of tumor was not adequate for analysis due to necrosis were excluded. Gem-gemcitabine (n=7), Hh-Hh inhibitor + Gem (n=8), c-Met- HGF/c-Met inhibitor + Gem (n=10), DMSO-vehicle control (n=6), Hh+c-Met+Gem (n=8).. ns- not significant *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (unpaired student t-test in panel B and Fisher’s exact test in panel C).

Figure 5. Single c-Met or Hh inhibitor treatment leads to the enhanced expression of the other target and the combination of both c-Met and Hh inhibitors suppresses the expression of both targets more effectively

A and B. Representative IHC images of primary tumors from the transgenic mouse model-KPC stained for expression of Shh (panel A) and c-Met (panel B) showing upregulation of Shh in the c-Met treatment group and vice versa, that can be overcome by combination treatment of Hh and c-Met inhibitors with gemcitabine. C and D. Representative IHC images of primary tumors from the orthotopic mouse model stained for expression of Shh (panel C) and c-Met (panel D) showing upregulation of Shh in the c-Met treatment group and vice versa, that can be overcome by combination treatment of Hh and c-Met inhibitors with gemcitabine. Mice in both models were treated for 1 week (as shown in Figure 1A). The data is representative of 1 experiment. Mice that died before the completion of the planned treatment or whose quality of tumor was not adequate for analysis due to necrosis were excluded. Arrows point to positive staining. Gem-gemcitabine (n=8), Hh-Hh inhibitor + Gem (n=9), c-Met- HGF/c-Met inhibitor + Gem (n=7), DMSO-vehicle control (n=9), Hh+c-Met+Gem (n=8).. Scale bars, 200 µm.

Figure 6. Compensatory overexpression/activation of alternative pathways explains single target resistance in PDA

A. qRT-PCR analysis of c-Met, Shh and GAPDH (control) in KPC tumor cells after c-Met inhibitor treatment and qRT-PCR analysis of Shh, HGF and GAPDH (control) in mCAFs after Hh inhibitor treatment. The gene was normalized to GAPDH and is shown as a fold change. B and C. Representative IHC images of primary tumors from the orthotopic mouse model stained for expression of IGF-1 (panel B) and E-Cadherin (panel C). Mice were treated for 1 week (as shown in Figure 1A). The data is representative of 1 experiment. Mice that died before the completion of the planned treatment or whose quality of tumor was not adequate for analysis due to necrosis were excluded. Gem-gemcitabine (n=8), Hh-Hh inhibitor + Gem (n=9), c-Met- HGF/c-Met inhibitor + Gem (n=7), DMSO-vehicle control (n=9), Hh+c-Met+Gem (n=8). Scale bars, 200 µm.