Single Agent and Synergistic Activity of the “First-in-Class” Dual PI3K/BRD4 Inhibitor SF1126 with Sorafenib in Hepatocellular Carcinoma

This article is featured in Highlights of This Issue, p. 2551

Human hepatocellular carcinoma (HCC) is the most common primary malignancy of the liver, which represents the fifth most common cancer and second leading cause of cancer-related mortality worldwide (1, 2). Prognosis of HCC is poor, and curative treatments (resection, liver transplantation, local ablation) can only be applied to a limited number of patients as the diagnosis is often made at an advanced stage of the disease (3). Sorafenib (Nexavar) is the first and only FDA-approved drug that is clinically approved for patients with advanced HCC (4–6). Sorafenib is a tyrosine kinase inhibitor that is known to potently inhibit VEGF receptor tyrosine kinases, viz. KDR and FLT4; serine/threonine kinases of RAF family, viz. c-Raf and B-Raf; and various growth factor receptors, specifically platelet-derived growth factor receptor (PDGFR; ref. 7). However, the increased survival benefit related to sorafenib treatment is limited to about 3 months, emphasizing the need for the development of new treatment strategies (8).

Recent publications indicate that HCC cell activation by different factors is known to increase both Ras/Raf/MAPK and PI3K/AKT/mTOR signaling (9). Sorafenib, the only drug for treatment of patients with advanced HCC inhibits Ras/Raf/MAPK pathway (8) but does not directly inhibit the PI3K/AKT/mTOR pathway, which also plays an important role in HCC proliferation. The PI3K pathway is known to be activated in 30% to 50% of HCC cases (10). In HCC, somatic mutation of PIK3CA, enhancement of Akt and phosphorylated ribosomal protein S6, and decrease of PTEN expression have been reported (11–14). These observations suggest that combined targeting of the PI3K/AKT/mTOR and Ras/Raf/MAPK pathway might provide benefit in the treatment of HCC. In this regard, other laboratories have used PI3K/mTOR inhibitors PI-103 and PKI-587 in combination with sorafenib and found that these drugs can synergistically block Ras/Raf/MAPK and PI3K/mTOR pathways (15, 16). However, these drugs are not slated to enter clinical trials for treatment of cancer patients (17–19), hence there is an urgent need to find a PI3K drug that is in clinical trials and can be used to treat HCC patients.

SF1126, a pan-PI3K inhibitor, has shown antitumor and antiangiogenic activity in a number of xenograft models (20–25). Furthermore, this drug has recently been shown to be safe (no dose-limiting toxicity or hepatotoxicity) and have considerable efficacy in B-cell malignancies and a variety of solid tumors in a phase I clinical trial (26). SF1126 is an RGDS-conjugated LY294002 prodrug, which is designed to exhibit increased solubility and bind to specific integrins within the tumor compartment, resulting in enhanced delivery of the active compound to the tumor vasculature and tumor (27). A recent study has demonstrated that LY294002, the active moiety of SF1126, has been cocrystallized in the active site of BRD4 and inhibits BET bromodomain binding to acetylated lysine-binding sites within chromatin (28). The bromodomain and extraterminal domain (BET) proteins recently emerged as important therapeutic targets in several types of hematopoietic cancers (29–32). Bromodomains are protein motifs that primarily bind to acetylated lysine residues, including those on histone tails (33). The BET family proteins (BRD2, BRD3, BRD4, BRDT) contain two amino-terminal bromodomains and have recently been recognized in the literature as a therapeutic strategy to target c-Myc (32). c-Myc transcription factor is frequently upregulated in a variety of human cancers (34), including liver cancer and targeting c-Myc is considered as a novel approach to treat HCC (35). Recent studies have shown that BRD4 is overexpressed in HCC tissues and it promotes tumor growth and epithelial mesenchymal transition in HCC (36, 37). Recent studies have shown that JQ1 reduced HCC cell growth by suppressing BRD4 using MYC-dependent or -independent pathway (37, 38). Focusing on the novel dual inhibitory activity of SF1126 toward PI3K and BRD4, the aim of this study was to evaluate the antiproliferative effect of SF1126 and sorafenib as single agents and in combination evaluate their effects on the PI3K/AKT/mTOR, Ras/Raf/MAPK, and Myc signaling pathways in preparation for a phase II trial in HCC.

The human HCC cell lines, Hep3B, HepG2, SK-Hep1 were obtained from ATCC. The Huh7 cell line was obtained from Japanese Collection of Research Bio resources. All cell lines were authenticated by short tandem repeat DNA profiling at the respective cell banks and were maintained as recommended by the manufacturers. These cells were cultured in DMEM (Invitrogen) supplemented with 10% FBS, 2 mmol/L glutamine and 1% penicillin–streptomycin at 37°C in a 5% CO₂ atmosphere. SF1126 was provided by SignalRx Pharmaceuticals. JQ1 was a gift from James Bradner (Dana-Farber Cancer Institute, Boston, MA). Antibodies specific for AKT, pAKT, ERK, pERK, pP70S6K, P70S6K, and tri-methyl histone H3 (Lys4) were obtained from Cell Signaling Technology. Normal rabbit IgG, protein A/G agarose beads, and c-Myc antibodies were from Santa Cruz Biotechnology, and anti-BRD4 antibody was obtained from Bethyl Laboratories.

Hep3B, HepG2, SK-Hep1, and Huh7 cells (4 × 10⁴ cells/well) were plated in 96-well plates in DMEM containing 10% FBS and 2 mmol/L l-glutamine. Cells were incubated overnight and treated with DMSO (0.1% final) or different concentrations of SF1126 or sorafenib alone and in combination (1:1) of sorafenib + SF1126 for 48 hours. AlamarBlue (Roche) was added and plates were incubated at 37°C in 5% CO₂ for 6 hours. Fluorescence signals were read as emission at 590 nm after excitation at 560 nm. IC₅₀ values were calculated by plotting fluorescence intensity versus drug concentration. Synergy studies were performed on these cell lines using Chou–Talalay method reported earlier (39) as described in Supplementary Methods.

For apoptosis studies, cells were treated with inhibitor for 24 hours, followed by caspase-3 activity assay using kit (Roche) or staining with Annexin V FITC antibody and propidium iodide (PI) according to the manufacturer's instructions (BD Pharmingen) as reported earlier (25). For cell-cycle analysis, DNA content was analyzed with FACSCalibur flow cytometer (BD Biosciences).

For all Western blots, 2 × 10⁶ cells were plated in 10-cm tissue culture dishes, were allowed to adhere for 36 hours, and were treated with different concentrations of SF1126 (1, 5, 10, 20 μmol/L), sorafenib (1, 10 μmol/L) either alone or in combination (0.5 μmol/L each or 0.25 μmol/L each) for 30 minutes. Whole cell lysates were prepared using RIPA buffer containing protease inhibitor cocktail (Roche Molecular Biochemicals). Clarified lysates were resolved in 10% SDS-PAGE, transferred to PVDF membrane, and probed for different antibodies.

The crystallographic atomic coordinates of BRD4-BD1 cocrystallized with JQ1 (PDB code 3MXF) were obtained from the Protein Data Bank (40). To model the binding of LY294002 and JQ1 at the key acetyl-lysine recognition pocket, the PDB file was imported into LeadIT (BioSolveIT GmbH), all water molecules were kept, residues around JQ1 within a grid of 7 ų were selected and used for in silico docking calculations. The 3D structures of LY294002 and JQ1 (all hydrogens included) were docked using LeadIT's standard parameters. Compounds LY294002 and JQ1 were tested for BRD4-1 and BRD4-2 activity by using Histone H4 peptide (1-21) K5/8/12/16Ac-Biotin as a ligand in alpha screen binding assay. The test was performed in collaboration with Reaction Biology.

Total RNA was extracted using the Qiagen RNAeasy Kit (Qiagen) and reverse transcribed using iscript cDNA Synthesis Kit (Bio-Rad). Amplification of cDNA was performed with 1× SYBR Green supermix (Bio-Rad) on an CFX96 Real time system (Bio-Rad). cDNAs were amplified using specific c-Myc primers and primer sequences will be available upon request.

Huh7 and SK-Hep1 cells were treated with SF1126 (10 μmol/L), JQ1 (1 μmol/L), or vehicle control for 24 hours. Cells were then harvested for chromatin immunoprecipitation (ChIP) assay as previously reported (32) and described in Supplementary Methods. ChIP and input DNA were analyzed by real-time PCR analysis as described before (24) using previous published primers against the MYC transcriptional start site (41) and a negative region upstream of MYC (32). Fold enrichment over control antibody and over the negative region upstream of MYC was determined from duplicate PCR reactions according to the formula |${2^{({C_{\rm{t}}}}}^{{\rm{MYC }} - {\rm{ }}{C_{\rm{t}}}\,{\rm{control \ antibody}})}$| or |${2\^(}^{{C_{\rm{t}}}\,{\rm{MYC}}\, - \,\,{C_{\rm{t}}}\,{\rm{negative}}\,\,{\rm{region}})}$|⁠, respectively.

All procedures involving animals were approved by the University of California San Diego Animal Care Committee. Eight million SK-Hep1 cells or 10 million Huh-7 cells in 100-μL PBS were injected subcutaneously into the right flank of NSG mouse. Tumor dimensions were recorded regularly using Vernier calipers. Tumor volume was measured using the following formula: volume = 0.5 × length × (width)². For SF1126 and sorafenib combination studies, treatment was initiated when tumors reached approximately 50 mm³. After 15 days of implantation of SK-Hep1 and 25 days of Huh-7 tumor implantation, mice were divided into four groups (n = 7–8 mice per group). Mice in group 1 were treated with acidified water (vehicle control), Group 2 treated with SF1126 (50 mg/kg/day), Group 3 treated with sorafenib (25 mg/kg), and Group 4, SF1126 + sorafenib (50 and 25 mg/kg, respectively). Mice in each group were treated for 6 days a week till the termination of the experiment.

The Student t test was used to evaluate differences observed between experimental groups and to compare tumor volume differences between SF1126 treated, combination treatment of SF1126 and sorafenib, and vehicle-treated controls.

To test the effect of SF1126 on HCC cell proliferation, we chose four different established HCC cell lines, Hep3B, HepG2, SK-Hep1, and Huh7. SF1126 as a single agent inhibited proliferation of all the cell lines tested (Fig. 1). The IC₅₀ of SF1126 for Hep3B, HepG2, SK-Hep1, and Huh7 cells was found to be 5.05, 6.89, 3.14, and 2.14 μmol/L, respectively (Fig. 1A–D). Importantly, these IC₅₀ concentrations are well within the pharmacokinetic levels of SF1126 achieved in the human phase I trial (26). Next, we evaluated the potency of the combination of SF1126 and sorafenib in HCC proliferation. All four cell lines were treated with different concentrations of SF1126 and sorafenib. The combination of SF1126 and sorafenib resulted in an increased inhibition of HCC proliferation (Fig. 1). As expected, differences were noted in the sensitivity for each cell line for each single agent and combination treatment. For Hep3B cells, IC₅₀ for sorafenib and combination of SF1126 and sorafenib was found to be 6.52 and 2.74 μmol/L, respectively (Fig. 1A). For HepG2 cells, IC₅₀ for sorafenib and combination of drugs was found to be 8.6 and 1.79 μmol/L, respectively (Fig. 1B). For SK-Hep1 cells, IC₅₀ for sorafenib and combination of SF1126 and sorafenib was found to be 7.6 and 0.52 μmol/L, respectively (Fig. 1C). For Huh 7 cells, IC₅₀s are 8.08 and 0.65 μmol/L for sorafenib alone or in combination with SF1126, respectively (Fig. 1D). Analysis of the enhanced cytotoxic effect in combination versus SF1126 or sorafenib alone was performed using the Chou–Talalay method for calculating the combination index (CI), with values <1 suggesting synergism. Combination of SF1126 and sorafenib in Hep G2, SK-Hep1, and Huh7 cells resulted in CI values of 0.6, 0.7, and 0.5, which is consistent with synergistic inhibition of proliferation. For Hep 3B, the CI value is 1.049 suggests an additive effect of the drug combination (Supplementary Table S1).

To study the growth-inhibitory mechanisms of SF1126 and sorafenib on HCC cells, we evaluated the effect of these inhibitors on cell-cycle progression and apoptosis. Cell-cycle analysis revealed that SF1126 or sorafenib as single-agent treatment resulted in cell-cycle arrest with a proportional increase in G₀–G₁ and a decrease in the number of cells in the S-phase in Hep 3B, Hep G2, SK-Hep1, and Huh7 cells (Fig. 2A–D). The combined targeting of SF1126 and sorafenib on SK-Hep1 cells resulted in substantial increase in the number of cells in G₀–G₁ phase with very few cells in S-phase (Fig. 2A–D).

To determine whether the cell-cycle arrest phenotype was associated with the induction of apoptosis, Annexin-V FITC staining was performed. Hep3B, HepG2, SK-Hep1, and Huh7 cells were incubated with various concentrations of SF1126 and sorafenib for 48 hours and stained with Annexin V and PI. Table 1 and Supplementary Fig. S1A show that SF1126 and sorafenib at 1 and 10 μmol/L concentrations, respectively, dose dependently increase apoptotic cells in Sk-Hep1, Huh-7, and Hep3B cell lines. Interestingly, HepG2 cells are not sensitive to the apoptosis-inducing effects of SF1126 or sorafenib either alone or in combination (Supplementary Fig. S1A; Table 1). Supplementary Fig. S1B shows the Annexin-V FITC analysis of SF1126 and sorafenib-treated SK-Hep1 cells.

To further validate these results, cell death assay and caspase-3 assays were performed on SK-Hep1 and Huh7 cells. Supplementary Fig. S2A and S2B revealed that Huh 7 and SK-Hep1 cells treated with SF1126 either alone or in combination with sorafenib had increased cell death. The results shown in Supplementary Fig. S2C and S2D illustrate caspase-3 activity assay performed on Huh 7 and SK-Hep1 cells suggesting that our combined treatment of SF1126 and sorafenib induce apoptosis in both cells. Taken together, these results suggest that SF1126 and sorafenib inhibited cell proliferation through the induction of G₁ arrest and apoptosis in a dose-dependent manner in all the cell lines tested while differences in sensitivity exist.

SF1126 as a single agent is known to inhibit the PI3K/AKT/mTOR pathway in various cell lines (23, 27, 42). SF1126 at 1 and 10 μmol/L concentrations significantly suppressed phosphorylation of AKT, p70S6K, and 4EBP1 in Hep3B, HepG2, SK-Hep1, and Huh7 cells in a dose-dependent manner as determined by Western blot analysis (Fig. 3). SF1126 supressed phosphorylation of ERK in Hep3B and Huh 7 cells, whereas there was no effect or mild effect of this drug on pERK levels in SK-Hep1 and HepG2 cells, respectively (Fig. 3). As expected, sorafenib showed no effect on phosphorylation of AKT, p70S6K, and 4EBP1 in HepG2, SK-Hep1, and Huh7 cells (Fig. 3B–D). Interestingly, sorafenib increased PI3K signaling in Hep3B cells (Fig. 3A). Similar to this result, a previous report has demonstrated that blockage/inhibition of only one of the main pathways PI3K/AKT/mTOR or Ras/Raf/MAPK, separately, can result in activation of the other pathway (16). We next tested whether sorafenib alone or in combination with SF1126 can block the key enzymes of PI3K/AKT/mTOR and Ras/Raf/MAPK pathway. As shown in Fig. 3, treatment of cells with combination of SF1126 and sorafenib at 0.5 and 2.5 μmol/L suppress phosphorylation of AKT, p70S6K, 4EBP1, and ERK in all the cell lines. These results suggest that combined treatment of SF1126 and sorafenib blocked multiple key enzymes in PI3K/AKT/mTOR and Ras/Raf/MAPK pathway (Fig. 3). This is correlated with greater antitumor activity in vivo.

Herein, we investigated the molecular interaction between LY294002, the active moiety of SF1126, and the BRD4 bromodomain binding domain 1 (BD1; ref. 28). We utilized in silico modeling of BRD4-BD1 with LY294002 compared with JQ1 (PDB code: 3MXF) to map the comparative binding orientation and free energy of binding (ΔG°, kcal/mol) of these two small molecules in the BRD4-BD1 active site (Fig. 4A, left). We discovered that docking LY294002 (BRD4-BD1 IC₅₀ = 5.3 μmol/L) and JQ1 (BRD4-BD1 IC₅₀ = 33 nmol/L) into the binding sites of 3MXF predicted the expected binding conformations of these two compounds as found in their corresponding BRD4-BD1 crystal structures. The predicted binding affinity trend for LY294002 and JQ1 was in accordance when docked in 3MXF (binding scores = −14.808 and −24.956 kcal/mol, respectively). Alpha screen binding assay performed in collaboration with Reaction Biology showed BRD4 activity of LY294002 and JQ1 using K5/8/12/16Ac-Biotin as a ligand (Fig. 4A, bottom) of 5 and 33 nmol/L, respectively, for BD1. Because the literature supports c-MYC as a clinically relevant target for the treatment of HCC (35), we first determined whether SF1126 can block c-Myc expression in HCC cells. Fig. 4B–E showed that SF1126 significantly decreased c-Myc expression in Hep3B, HepG2, SK-Hep1, and Huh7 cell lines. In consistence with these results, SF1126 blocked c-MYC protein levels in all the cell lines as shown by Western blot analysis (Fig. 3). It was previously shown that BET bromodomain inhibitors displace BRD4 from the MYC transcriptional start site, so we next investigated whether SF1126 is able to displace BRD4 from the MYC promoter. Using ChIP with a BRD4 antibody followed by quantitative real-time PCR with primers designed against the MYC transcriptional start site, we observed marked BRD4 localization to the MYC transcriptional start site (>10-fold enrichment in comparison with a nonspecific locus upstream of MYC and >100-fold enrichment in comparison with a nonspecific control antibody). This BRD4 binding to the MYC transcriptional start site was significantly abrogated by treatment with SF1126 by 10-fold in comparison with a nonspecific locus upstream of MYC and by 45-fold in comparison with a nonspecific control antibody (Fig. 4 F and G). The BRD4 inhibitor JQ1 was used as a positive control with similar efficacy. These results demonstrate that SF1126 is a dual PI3K/BRD4 inhibitor that targets multiple signaling nodes dysregulated in HCC.

The antitumor activity of SF1126 in vivo was investigated in a xenograft model using SK-Hep1 and Huh7 cells. Fig. 5A and C illustrate the antitumor efficacy of SF1126 either alone or in combination with sorafenib in SK-Hep1 and Huh-7 cells, respectively. Mice bearing 50 mm³ tumors were treated subcutaneously with SF1126 at 50 mg/kg, six times a week for 3 weeks. The combined treatment of SF1126 (50 mg/kg, six times a week; administered subcutaneously) and sorafenib (administered intraperitoneally; 25 mg/kg, six times a week) significantly blocked SK-Hep1 and Huh 7 HCC tumor growth in vivo (Fig. 5A and C). SF1126 either alone or with sorafenib produced tumor growth inhibition with no evidence of toxicity, as measured by weight loss relative to control animals or drug related lethality (data not shown). Interestingly, the levels of c-MYC, pAKT, and pERK were decreased in tumors excised from the SF1126 or sorafenib or drug combination treated mice at 1 hour after drug administration (Fig. 5B and D). These results suggest that combined SF1126 and sorafenib treatment suppresses HCC tumor growth through the downregulation of c-MYC, PI3K/AKT/mTOR, and Ras/Raf/MAPK pathways in vivo.

This study was focused on the effect of SF1126 alone or in combination with sorafenib in preclinical HCC models. We evaluated HCC proliferation, cell cycle, apoptosis, phosphorylation of the key enzymes in the PI3K/mTOR/Ras/Raf pathways, and antitumor activity in vivo. The conceptual framework for the investigation SF1126 alone or in combination with sorafenib in HCC emerged from the literature in that: (i) PI3K pathway is known to be activated in 30% to 50% of HCC cases (14); (ii) Sorafenib, the only drug FDA approved for use in advanced HCC treatment, does not directly inhibit the PI3K/AKT/mTOR pathway; (iii) the MYC oncogene has been implicated in the pathogenesis and prognosis of HCC. This led to the hypothesis that a novel dual PI3K/BRD4 inhibitor, SF1126 used alone or in combination with sorafenib would demonstrate potent antitumor activity in HCC. Importantly, SF1126 is an IND-enabled drug currently in clinical trials that has shown antitumor and antiangiogenic activity in a number of xenograft models and has recently showed considerable efficacy in B-cell malignancies in a phase I clinical trial (26). A recent report that the active moiety of SF1126, LY294002 can be cocrystalized in the acetyl-lysine binding pocket of BRD4 prompted us to evaluate the effect SF1126 on the potential BRD4–MYC interaction. We observe in Huh7 and SK-Hep1 HCC cells that SF1126 displaced BRD4 coactivator protein from the transcriptional start site of MYC. Our molecular modeling and BRD4 binding studies demonstrate LY294002 binding to the BRD4 BD1 bromodomain. These findings suggest a novel and interesting function of this drug as it targets both PI3K and MYC transcription in HCC via the inhibition of BRD4 interaction with MYC enhancer–promoter elements.

Our cell viability assay experiments indicated that SF1126 or sorafenib, as single agents, inhibited HCC cell proliferation. A combination of SF1126 and sorafenib resulted in synergistic inhibition of HCC cell proliferation in HepG2, SK-hep1, and Huh 7 cells. The synergistic inhibition by the drug combination treatment suggests that both the Ras/Raf/MAPK and PI3K/AKT/mTOR pathways play an important role in HCC cell proliferation. In addition, the results of cell-cycle analysis and apoptosis assays showed that G₁ arrest and apoptosis were induced by SF1126 treatment alone and these effects were augmented in combination with sorafenib. These results suggest that the antiproliferative effects of these agents are due to G₁ arrest and apoptosis in the HCC cell lines and that the combination increased G₁ arrest and apoptosis in vitro.

Western blot experiments demonstrated potent inhibition of key enzymes of both Ras/Raf/MAPK and PI3K/AKT/mTOR pathways when cells were treated with SF1126 and sorafenib. SF1126 monotherapy strongly inhibited phosphorylation of S6K, 4EBP1, and AKT (Ser473). Sorafenib, as single agent, inhibited ERK phosphorylation, but interestingly it stimulated AKT (Ser473) phosphorylation. These results confirm the previous findings that inhibition of one of the canonical pathways may cause stimulation of the pathway in HCC cells (15, 16). This supports the hypothesis that monotherapy in HCC treatment may not be efficacious. Presumably, sorafenib may induce other signals, such as HGF/HGFR, that further stimulates the PI3K/AKT/mTOR pathway. Indeed, sorafenib stimulated HGF secretion in HCC cells promoted c-Met, S6K, and 4EBP1 phosphorylation (43). Increased HCC activation and mobility by HGF were reported to be mediated through PI3K signaling (44).

Recent studies have found that the Ras/Raf/MAPK and PI3K/AKT/mTOR are the dominant signaling pathways activated in HCC cells (45). Several attempts have been made to target both of these pathways as therapy. In one study, everolimus, known rapamycin analogue, and AEE788, EGF/VEGF inhibitor, blocked tumor growth in Huh7 HCC cells in combination compared with everolimus monotherapy (46). In another study, Newell and colleagues used sorafenib and rapamycin to target mTOR and Ras/Raf/MAPK signaling and decreased proliferation and induced apoptosis in HCC cell lines. In a xenograft mouse model, the combination of rapamycin and sorafenib enhanced tumor necrosis and ulceration compared with sorafenib alone (47). Taken together, published reports and unfavorable response of AZD6624, a novel MEK inhibitor in clinical settings (10), further provide evidence to our hypothesis that inhibiting only Ras/Raf/MAPK signaling pathway is not efficacious in HCC treatment.

In this study, we found that SF1126 blocks both the Ras/Raf/MAPK and PI3K/AKT/mTOR pathways in the HCC cell lines. Moreover, SF1126 blocks c-Myc expression and displaces BRD4 from the MYC transcriptional start site in HCC cell lines. Moreover, the active moiety of SF1126, LY294002, blocks BRD4 interaction with acetylated lysine residues Histone H4 peptide (1-21) K5/8/12/16Ac-Biotin, it has been cocrystallized in the active site of BRD4 and we have executed extensive molecular modeling of this chemotype and compared it with JQ1 to develop additional more potent dual inhibitors of PI3K and BRD4 (48). Our data show that SF1126, either alone or in combination with sorafenib, significantly suppressed tumor growth in vivo in SK-Hep1 cells and Huh 7 cells. SF1126 exhibits potent dual inhibition of PI3K and mTOR (mTORC1 and mTORC2). It has high bioavailability and biosafety. SF1126 has been used in preclinical studies to inhibit 50 human cancer types. Finally, several components are in place to rapidly translate our findings into a phase II trial in HCC: (i) SF1126 has completed a phase I clinical trial in humans with good safety profile where no MTD was reached and no hepatotoxicity observed (26); (ii) a well-established GMP manufacturing and formulation plan has been established; (iii), a GMP drug supply and an active IND are in place to rapidly advance SF1126 into a phase II trial in advanced HCC; and (iv) based on the data generated in this report, 3 months ago we submitted a phase II clinical trial protocol to IND#744451 at the FDA to begin a phase II clinical trial of SF1126 in advanced HCC. In conclusion, we demonstrate that SF1126 has potent anti-HCC activity in vitro and in vivo. Moreover, the application of SF1126 in combination with the FDA approved agent Sorafenib shows synergistic antitumor activity in vitro and in vivo, a result that correlates with the inhibition of both canonical Ras/Raf/MAPK and PI3K/AKT/mTOR pathways in vivo. The results suggest the consideration of a phase II clinical trial of SF1126 in advanced HCC and a potential combination phase I trial with sorafenib.

J.K. Sicklick reports receiving a commercial research grant from Novartis Pharmaceuticals, Inc., Foundation Medicine, Inc., and Blueprint Medicines; is a consultant/advisory board member for Sirtex Medical Inc. (SIR-Spheres microspheres Advisory Board). Y. Kono has received speakers bureau honoraria from Bayer and Wako and is a consultant/advisory board member for Bayer. J.R. Garlich has ownership interest (including patents) in SignalRx Pharmaceuticals. G.A. Morales is CEO at Innoventyx, LLC. D.L. Durden has ownership interest (including patents) in and is a consultant/advisory board member for SignalRx Pharmaceuticals. No potential conflicts of interest were disclosed by the other authors.

Conception and design: A.R. Singh, S. Joshi, A.M. Burgoyne, J.K. Sicklick, Y. Kono, D.L. Durden

Development of methodology: A.R. Singh, S. Joshi, A.M. Burgoyne, J.K. Sicklick, D.L. Durden

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J.R. Garlich, D.L. Durden

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A.R. Singh, S. Joshi, A.M. Burgoyne, J.K. Sicklick, D.L. Durden

Writing, review, and/or revision of the manuscript: A.R. Singh, S. Joshi, A.M. Burgoyne, J.K. Sicklick, S. Ikeda, Y. Kono, J.R. Garlich, G.A. Morales, D.L. Durden

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J.R. Garlich, G.A. Morales, D.L. Durden

Study supervision: D.L. Durden

The authors thank Muamera Zulcic for ordering and maintaining the mice needed for the experiment.

This work was supported by NIH grant RO1 CA94233-09, FDA RO1 grant FD-004385, R41 CA192656 (to D.L. Durden), and grants from ALSF (Springboard Grant) and Hyundai Hope on Wheels Foundation, Hope Grant and the Olivia Hudson Foundation NIH K08 CA168999 (to J. K. Sicklick) and P30CA023100 (to S. Lippmann).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.