Combination of AKT inhibitor ARQ 092 and sorafenib potentiates inhibition of tumor progression in cirrhotic rat model of hepatocellular carcinoma

The prognosis of patients with advanced hepatocellular carcinoma (HCC) is very poor. The AKT pathway is activated in almost half of HCC cases and in addition, long term exposure to conventional drug treatment of HCC, sorafenib, often results in over-activation of AKT, leading to HCC resistance. Therefore, it is important to assess the safety and the efficacy of selective allosteric AKT inhibitor ARQ 092 (Miransertib) in combination with sorafenib. Here, we demonstrated in vitro that the combination of ARQ 092 with sorafenib synergistically suppressed proliferation, promoted apoptosis, and reduced migration. To test the effect of the combination in vivo, rats with diethylnitrosamine-induced cirrhosis and fully developed HCC were randomized and treated with vehicle, sorafenib, ARQ 092 or the combination of ARQ 092 with sorafenib; (n=7/group) for 6 weeks. Tumor progression, size of tumors and the mean tumor number were significantly reduced by the combination treatment compared to the control or single treatments. This effect was associated with a significant increase in apoptotic response and reduction in proliferation and angiogenesis. Sirius red staining showed a decrease in liver fibrosis. Moreover, treatments improved immune response in blood and in tumor microenvironment. Thus, the combination of ARQ 092 with sorafenib potentiates inhibition of tumor progression and gives the possibility of therapeutic improvement for patients with advanced HCC.


Cell viability assay
Cell viability was determined by MTT (Sigma-Aldrich) ( ) during 48h. CI values were calculated with an implementation of the Median-Effect approach of Chouand Talalay using CompuSyn software as described previously (2). CI=1 indicates an additive effect, CI>1 indicates antagonism, CI<1 indicates a synergistic effect and CI <0.1 indicates very strong synergistic effect (3).
As a negative control, cells were incubated in the same medium with 1 % DMSO without drug. Each sample was analyzed on three replicates and all experiments were repeated three times.

Cell migration
Cell migration was assessed on 4 cell lines by wound-healing assay and cells were then treated with IC20 and IC50 concentrations of Sorafenib or ARQ 092 or with Combination IC50/200 and Combination IC50/10. Images were captured every hour by time-lapse microscopy at 37°C, 5% CO 2 with Zeiss AxioVert 100M connected to a MicroMAX B/W (6.7x6.7 μm, -15°, ~ 3 im/s) camera using MetaMorph ® software (MetaMorph Inc., USA). The width of the wound was quantified at 24 h by software ImageJ. Data are presented as relative percentage of closed-wound. The experiments were performed in duplicates and repeated at least three times. Cell velocity was determined via tracking cells by time-lapse microscopy. Cells were then treated with IC20 and IC50 concentrations of ARQ 092 or sorafenib and images of cells were captured every hour by timelapse microscopy with Zeiss AxioVert 100M connected to a MicroMAX B/W camera. To estimate the velocity, 4 different areas per sample were captured and 5 cells per area were tracked using the Manual Tracking plugin for Image J.

MRI studies
Imaging study was performed with a 4.7 Tesla MR Imaging system (BioSpec 47/40 USR, Bruker Corporation, Germany) and Transmit/Receive Volume Array Coil for rat body 8x2 (Bruker Corporation, Germany) in the Grenoble MRI facility IRMaGE. Rats were fitted in ventral decubitus position and anesthetized with isoflurane inhalation (Forane ® , Abbott, USA). Breathing was continuously monitored to maintain a respiratory rate between 35 and 45 breaths per minute and body temperature was maintained around 37°C.
We used Turbo rapid acquisition with relaxation enhancement T2-weighted (Turbo-RARE T2) sequence (repetition time (TR): 1532.9 msec, echo time (TE): 27.4 msec, flip angle (FA): 180°) with a field of view (FOV) of 55 x 55 mm, 20 slices, a thickness and a slice separation of 2 mm, and were realized with a respiratory triggered acquisition to reduce artefacts. MRI parameters adjustment and image acquisition were realized by using Paravision 5.1 software.
A morphological analysis was realized based on the TurboRARE T2 sequences and according to the Response Evaluation Criteria in Solid Tumors (RECIST) criteria. Ten liver tumors were selected and measured on MRI1, 2 and 3. Estimated tumor size corresponded to the sum of the largest diameter of these 10 lesions. For each rat, MRI1 was considered as the baseline (i.e.: 0%) and tumor progression corresponded to the comparison between MRI2 or 3 and the baseline, (i.e: "(tumor size MRI2/3 -tumor size MRI1 )/tumor size MRI1 ).
To detect proliferating cells, paraffin-embedded sections were incubated overnight at 4°C with the primary anti-Ki67 antibody (Rabbit, clone SP6, Thermofisher scientific, USA), followed by incubation with the peroxidase-conjugated bovine anti-rabbit IgG (Jackson ImmunoResearch, USA). DAB was used as the chromogen for Ki67 immunodetection. For Ki67 + cells, data are presented as positive cell nuclei per area (highpower fields; 20x magnification).
Apoptotic cells were analysed by ApoBrdU-IHC DNA Fragmentation Assay Kit (Biovision, USA) and methyl green solution was used for counter staining the cells. Data are presented as apoptotic cells per area (highpower fields; 20x magnification).
To detect vascularisation, paraffin-embedded sections were blocked by 10% donkey serum and then incubated overnight at 4°C with anti-rat CD34 antibody (Goat, AF4117, R&D Systems; Minneapolis, USA), followed by incubation with Alexa 647-conjugated donkey anti-goat IgG (Life Technologies, Carlsbad, CA, USA). Images were captured using ApoTome microscope (Carl Zeiss, Germany) equipped with a camera AxioCam MRm and collected by AxioVision software. Positive area was quantified using ImageJ software (NIH, USA) on 15 randomly selected fields/section (10x magnification).
Collagen was detected on paraffin-embedded sections with picro-sirius red stain solution (Sigma-Aldrich) and staining was subsequently quantified by MetaMorph ® software in 10 randomly selected fields/ section (10x magnification).
Oil Red O staining was performed on 7μm cryosections, prepared from formalin pre-fixed liver samples. Sections were stained with freshly prepared Oil Red O in isopropanol. Oil Red O staining provides chromogenic as well as fluorescent signals, therefore we used red channel to detect staining. Images were captured using ApoTome microscope (Carl Zeiss, Germany) equipped with a camera AxioCam MRm, collected by AxioVision software and quantified using ImageJ software (NIH, USA). For Oil Red O + liver area, data are presented as Oil Red O positive area in percent of total tissue area. Six random areas per each liver section were analysed.

Real-time polymerase chain reaction (qPCR)
Total RNA was extracted from frozen rat liver tissue samples. RNA purification was performed with RNeasy Mini Kit ® (Qiagen, USA). Reverse transcription was realized with Transcriptor First Strand cDNA Synthesis Kit ® (Life science, Roche), and amplification reactions were performed in a total volume of 20μL by using a Thermocycler sequence detector (BioRad CFX96, USA) with qPCR kit Mesa Green qPCR MasterMix Plus for SYBR Assay ® (Eurogentec, Belgium).
GADPH was used as housekeeping gene. Primers were designed with Primer 3 software (version 4.0.0) and verified on BLAST. Oligonucleotide sequences were synthetized by Eurofins Genomics ® in 0.01μmol scale, with a Salt Free level of purification. Every analysis was done in duplicates.

Flow cytometric analysis
Cells were recovered from liver tissue by mechanical disruption and whole blood samples were used in case of blood analyses. Cells were immunostained for flow cytometric analysis without any stimulation.