|Home | About | Journals | Submit | Contact Us | Français|
Sorafenib is a RAF inhibitor approved for several cancers, including hepatocellular carcinoma (HCC). Inhibition of RAF kinases can induce a dose-dependent “paradoxical” upregulation of the downstream mitogen-activated protein kinase (MAPK) pathway in cancer cells. It is unknown whether “paradoxical” ERK activation occurs after sorafenib therapy in HCC, and if so, if it impacts the therapeutic efficacy. Here, we demonstrate that RAF inhibition by sorafenib rapidly leads to RAF dimerization and ERK activation in HCCs, which contributes to treatment evasion. The transactivation of RAF dimers and ERK signaling promotes HCC cell survival, prevents apoptosis via downregulation of BIM and achieves immunosuppression by MAPK/NF-kB-dependent activation of PD-L1 gene expression. To overcome treatment evasion and reduce systemic effects, we developed CXCR4-targeted nanoparticles to co-deliver sorafenib with the MEK inhibitor AZD6244 in HCC. Using this approach, we preferentially and efficiently inactivated RAF/ERK, upregulated BIM and down-regulated PD-L1 expression in HCC, and facilitated intra-tumoral infiltration of cytotoxic CD8+ T cells. These effects resulted in a profound delay in tumor growth. Thus, this nano-delivery strategy to selectively target tumors and prevent the paradoxical ERK activation could increase the feasibility of dual RAF/MEK inhibition to overcome sorafenib treatment escape in HCC.
The efficacy of targeted therapy with kinase inhibitors in cancer is often limited by rapid treatment evasion. Treatment resistance may develop either due to additional mutations, by alternate mode of activation of the same pathway or alternative oncogenic pathways, or by dynamic reprogramming of the kinome1,2,3. One such mechanism is the “paradoxical” activation of MAP kinase (MAPK) pathway (RAF/MEK/ERK) by RAF inhibitors leading to adverse effects4. The use of RAF inhibitors such as vemurafenib or sorafenib in BRAF-wild-type cancers has been shown to lead to “paradoxical” activation of ERK in cutaneous squamous cell carcinoma, lung cancer or melanoma, likely due to a direct effect of the drug on RAF dimerization2,4,5.
Of the RAF inhibitors, the most widely used drug is sorafenib. Sorafenib is the standard therapy for advanced hepatocellular carcinoma (HCC) worldwide, and is also approved for advanced renal and thyroid cancers6,7. Sorafenib inhibits VEGFR and PDGFR tyrosine kinases, which is thought to exert anti-angiogenic and anti-fibrotic effects in HCC. We have previously shown that these effects are thwarted by treatment-induced hypoxia, which leads to CXCR4 upregulation in HCC and stromal cells and mediates metastasis progression and PD-L1-mediated immunosuppression8,9. On the other hand, sorafenib was developed as a cytotoxic agent to inhibit RAF kinases and downstream mitogen-activated protein kinases (MAPKs)10,11. However, HCCs only rarely harbor somatic mutation of the MAPK pathway such as KRAS or BRAFV600E activating mutations12. This prompted us to examine whether paradoxical activation of MAPK pathway occurs after sorafenib treatment in HCC. The RAF/MEK/ERK signal transduction cascade is widely considered promote tumor progression. RAF/MEK/ERK pathway can regulate the activity and expression of members of the BCL-2 protein family to promote cell survival13. In addition, activation of RAF/MEK/ERK signaling may directly upregulate PD-L1 expression and promote an immunosuppressive tumor microenvironment14,15.
Recent efficacy studies and FDA-approval for a dual RAF/MEK inhibition approach in BRAF-mutant melanoma is supporting the clinical relevance of this escape mechanism, but significant toxicity concerns remain16,17. The importance of paradoxical activation in limiting therapeutic efficacy in RAF-wild type cancers remains unknown. Here, we examined the relevance of ERK activation in HCCs treated with sorafenib (the only approved systemic therapy for this disease), and developed a safer nanoparticle-based multi-drug delivery system to overcome treatment resistance.
We found a moderate and variable cytotoxicity after sorafenib treatment in a panel of HCC cell lines, consistent with previous studies and with the transient and moderate responses typically seen in HCC patients18,19. The BRAF-wild type HCC cells showed differential sensitivity, with IC50 values ranging from 2μM (a clinically relevant concentration18) to over 5μM (Fig. 1a). Of note, BRAFV600 mutant HCC cells (SK-Hep-1) were more sensitive to sorafenib (IC50=0.5μM). We first evaluated the effects of sorafenib on RAF/ERK and MAP kinase p38 (p38MAPK) activation — two relevant targets of this drug in HCC cells20,21. As expected, sorafenib efficiently inhibited p38MAPK activity in all cell lines (Fig. 1b and Table S1). However, all BRAF-wild type HCC cell lines tested showed undetectable to low levels of CRAF and ERK phosphorylation at baseline (Fig. 1b). Moreover, sorafenib treatment induced a rapid and sustained increase in ERK activation in BRAF-wild type but not in BRAFV600E mutant HCC cells (Fig. 1b and Table S1). The robust ERK1/2 phosphorylation was more substantial in sorafenib-resistant HCC cell lines (Fig. 1a,b). Similar to the ERK activation seen in HCC cells in vitro, we found that sorafenib increased ERK activation in orthotopic Hep3B xenografts and in two murine models of spontaneous HCC (in Mst1−/−Mst2F/− mice and chemically-induced carcinogenesis) (Fig. 1c–e)22,23.
To test whether this effect is due to inefficient BRAF inhibition by sorafenib, we repeated the experiment using supra-physiological concentrations of sorafenib (4–16μM)24. At these high concentrations, sorafenib inhibited CRAF and ERK activity similarly to PLX4720—a potent and specific BRAF inhibitor—in BRAF-wild type HCC cells (Fig. 1f). We also examined if ERK activation was due to RAF heterodimerization in BRAF-wild type HCC cells—as previously shown with BRAF inhibitors in other cancers4. Indeed, after treating HCC cells with sorafenib and evaluating the presence of BRAF in CRAF immunoprecipitate, we found that BRAF heterodimerized with CRAF (Figs 1g and S1).
Next, we determined whether the increased ERK phosphorylation after sorafenib treatment mediates HCC cell viability and tumor growth. We found that blocking ERK activation with AZD6244 significantly reduced HCC cell viability both in vitro and in vivo when combined with sorafenib treatment (Fig. 2a,b). Furthermore, we examined whether knocking down CRAF or ERK expression—and thus preventing sorafenib-induced RAF dimer transactivation and consequent ERK activation—could also affect viability of BRAFWT HCC. Downregulation of CRAF alone did not affect HCC cell viability or primary tumor growth (Fig. 2c,d). However, when siRNA-induced CRAF knockdown was combined with sorafenib, the knock down of CRAF reduced ERK activation, decreased HCC cell viability, triggered apoptosis and caused a significant delay in HCC growth (Figs 2c–j and S2).
We further dissected how paradoxical ERK activation modulates HCC cell apoptosis. ERK mediates ubiquitination and degradation of Bim—a pro-apoptosis molecule—leading to resistance to chemotherapeutic drugs25,26. We found that sorafenib treatment alone led to Bim phosphorylation and degradation (Fig. 3a,b). However, Bim degradation was prevented when the HCC cells were treated with sorafenib in combination with AZD6244 both in vitro and in vivo, leading to apoptosis in HCC cells (Fig. 3c–e). Collectively, these data show that ERK activation and Bim degradation may mediate the rapidly acquired resistance to sorafenib in BRAF-wild type HCC.
To overcome this evasion mechanism and reduce systemic toxicities, we developed tumor-targeted nanoparticles (TTNPs), with the structure shown in Fig. 4a, to co-deliver sorafenib with a MEK inhibitor into HCC. To this end, we loaded sorafenib and the MEK inhibitor AZD6244 into NPs developed as previously described with several modifications27,28. The average diameters of drug-loaded TTNPs determined by DLS were 139.7±9.7nm, with poly-dispersity indexes (PDIs) of 0.425±0.057. The efficacy of sorafenib and AZD6244 encapsulation in the NPs was approximately 70%. To more selectively target HCC, CTCE-9908, a peptide antagonist for CXCR4, was conjugated to NPs as a targeting ligand (CTCE-NPs)29,30. We examined the cellular uptake of CTCE-NPs containing a tracer molecule C6 using both murine HCC (HCA-1) and human HCC (Mahlavu and Hep3B) cell lines. As shown in Fig. 4b–d, the uptake of C6 was greater in HCC cells treated with targeted CTCE-NPs than in cells treated with non-targeted NPs modified with scramble peptides (SC-NPs) (Fig. S3). The uptake of CTCE-NPs was competitively inhibited by addition of free CTCE-9908 peptides in a dose-dependent manner (Fig. S3), indicating that the cellular uptake was ligand dependent. Furthermore, CTCE-NPs loaded with sorafenib and AZD6244 exerted more potent cytotoxic effects on HCC cells than the combination of the free agents, unloaded CTCE-NPs, loaded SC-NPs or CTCE-NPs loaded with each agent alone (Figs 4e and S4). In addition, co-delivery of sorafenib and AZD6244 by CTCE-NPs prevented the paradoxical activation of ERK and increased the expression of Bim (Figs 4f,g and S4). These results demonstrate a synergistic cell-killing effect when using CTCE-NPs loaded with sorafenib and AZD6244.
Next, we evaluated the in vivo effects of CTCE-NPs loaded with sorafenib and AZD6244 on ERK activation as well as on tumor environment in orthotopic murine HCC models. As shown in Fig. 5a, increased tumor uptake of CTCE-NPs or non-targeted SC-NPs was observed compared with that of free drugs due to the EPR (enhanced permeability and retention) effect. Moreover, the enhanced intracellular uptake of CTCE-NPs was observed, indicating that NPs presenting a CXCR4 antagonist CTCE9908 peptide can increase tumor uptake in the murine HCC model (Figs 5b and S5). Furthermore, co-delivery of sorafenib and AZD6244 using CTCE-NPs significantly prevented ERK activation induced by sorafenib (Figs 5c and S6). Of note, this effect was achieved while maintaining a potent anti-angiogenic effect of sorafenib, as demonstrated by the decrease in tumor microvascular density (MVD) (Fig. 5d,e).
Finally, we evaluated the treatment-induced changes in programmed death-1-ligand 1 (PD-L1) expression, an immune checkpoint which could suppress cytotoxic CD8+ T cell proliferation and activation, resulting in immunosuppression and treatment resistance31. We examined surgical specimens of HCC from patients with recurring tumors after sorafenib treatment and observed profound ERK activation and increased PD-L1 expression in recurrent HCC (Figs S7 and S8). We further found that sorafenib treatment directly induced PD-L1 expression via ERK-mediated NF-kB activation in murine and human HCC cells in a dose dependent manner (at low doses), which was prevented by MEK or NF-kB inhibition (Figs 6a–c and S9). Moreover, co-delivery of sorafenib and AZD6244 using CTCE-NPs significantly reduced PD-L1 expression and increased cytotoxic T cell accumulation and activation in HCC in vivo (Fig. 6d–f). Treatment with CTCE-NPs loaded with sorafenib and AZD6244 induced potent tumor regression and increased overall survival in orthotopic HCC model in immunocompetent mice (Fig. 6g,h). Furthermore, the preclinical safety study indicated that the serum levels of hepatotoxicity makers such as aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP) and γ-Glutamyltransferase (γ-GT) remained the same as the untreated control animals after treatment of sorafenib and MEK inhibitors in CTCE-NPs (Table S2). We also examined the histology of the various organs of mice treated with sorafenib and AZD6244 in CTCE-NP by H&E staining. We detected no histological changes in the lungs, liver, spleen, kidneys or heart of C3H mice 24hr after treatment (Fig. S10).
In summary, we demonstrate that rapid ERK activation is a mediator of sorafenib resistance in HCC (Fig. 7). Co-delivery of sorafenib and MEK inhibitors via CXCR4-targeted NPs overcomes the cell autonomous mechanism of resistance to sorafenib in HCC as well as inhibits angiogenesis and converts the immunosuppressive microenvironment to an immunostimulatory microenvironment.
Multiple targeted therapies that entered phase III clinical trials in HCC failed to demonstrate superiority or non-inferiority to sorafenib. Thus, sorafenib remains standard of care for patients with advanced stage HCC. The poor understanding of sorafenib’s cell-autonomous mechanisms of action in HCC prompted us to examine whether the effect of sorafenib on HCC cell viability is dependent on RAF inhibition. Recently, the use of RAF inhibitors such as vemurafenib or sorafenib has been shown to paradoxically activate MAPK pathway in BRAF-wild type lung cancer or melanoma, due to induction of RAF (BRAF or CRAF) dimerization2,4,32. The clinical responses observed in patients with HCC patients (i.e., a transient delay in tumor growth followed by progression) is consistent with the moderate and highly variable cytotoxicity of sorafenib that we observed in various murine and human HCC cell lines. It suggests that inhibition of non-mutated cell-proliferation pathways (i.e., BRAF-wild type) may result in compensatory activation of downstream or other pathways that might become the rescue pathway for cell survival. Our data show RAF inhibition with knocking down BRAF/CRAF or using RAF inhibitors only moderately affected cell viability in BRAF-wild type HCCs. Moreover, partial inhibition of RAF using low doses of sorafenib or the specific BRAF inhibitor PLX4720 resulted in rapid MAPK activation due to BRAF heterodimerization as a cell-autonomous mechanism of resistance to sorafenib in BRAF-wild type HCCs.
We further studied how ERK activation mediates HCC cell survival by dissecting the signaling downstream of MAPK pathway. Bim is a member of the Bcl-2 homology 3 (BH3)-only subgroup of the Bcl-2 family. Bim induces apoptosis by blocking the activity of the anti-apoptotic members of the Bcl-2 family. It has been previously shown that ERK mediates ubiquitination and degradation of Bim through the Bim phosphorylation, which leads to resistance to chemotherapeutic drugs25,26. We show that the rapid activation of the MAPK pathway – induced by sorafenib treatment – leads to Bim degradation and protection from apoptosis in HCC cells. When CRAF knockdown or MEK inhibition—mediated by CRAF siRNAs or MEK blockade—was combined with sorafenib, we found reduced ERK activation, increased Bim and apoptosis induction and thus decreased HCC cell viability.
Next, we evaluated whether the sorafenib-induced paradoxical activation of ERK contributed to the changes of PD-L1 expression in HCC. We have previously shown that sorafenib increases PD-L1 expression in HCC in vivo and others have linked PD-L1 upregulation with MAPK activation in melanomas resistant to BRAF inhibition8,33. In this study, we observed the increased PD-L1 expression in HCC from patients with recurring tumors after sorafenib treatment. BRAF-wild type HCC cells showed enhanced ERK activation and PD-L1 expression while treated with sorafenib. The increased PD-L1 expression is transcriptionally modulated by NF-kB. MEK inhibition counteracts sorafenib resistance associated with MAPK activation and PD-L1-mediated immunosuppression. Combination of sorafenib and MEK inhibitors revealed dual therapeutic effects with apoptosis induction and reduction in PD-L1 expression, leading to increased accumulation and activation of tumor-infiltrating cytotoxic T cells. The targeted therapy in combination with immunotherapy not only overcomes resistance, but also provides long-term clinical benefit in HCC patients.
To overcome this evasion mechanism and reduce systemic toxicities, we developed CXCR4-targeted nanoparticles to co-deliver sorafenib with the MEK inhibitor AZD6244 into HCC. We have previously demonstrated that CXCR4 antagonists on the surface of targeted NPs display dual functions, serving both as tumor-targeting ligands and as inhibitors of CXCR4/SDF1α to suppress the accumulation and activation of stromal cells with tumor-promoting properties28. In this study, we utilized a peptide inhibitor of CXCR4—CTCE-9908 peptide to modified NPs. The multifunctional NPs loaded with sorafenib and MEK inhibitor achieve potent MAPK inhibition, induced significantly increased cell apoptosis, suppressed PD-L1 expression and inhibited immunosuppression in HCC.
These findings were consistent among various human and murine HCC cells in vitro, and were reproduced in vivo using both orthotopically implanted and chemically-induced murine HCC tumors. Molecular profiling of human HCCs has shown that that most tumors harbor no RAF-activating mutations and are indeed BRAF-wild type34. Moreover, consistent with our in vitro data obtained from human HCC cell lines, sorafenib treatment activates the RAF/ERK pathway and increases PD-L1 expression in patients with HCC.
Collectively, these data demonstrate that sorafenib can cause rapid RAF dimerization and ERK activation, leading to resistance to sorafenib in BRAF-wild type HCC cells. Adding a MEK inhibitor to sorafenib is clinically feasible35, but this strategy may be limited by the specific pharmacokinetics of these agents, and equally important by adverse systemic effects. We addressed this challenge by using a CXCR4-targeted nano-delivery strategy to target HCCs and co-deliver sorafenib with a MEK inhibitor. We showed that treatment with sorafenib and MEK inhibitors co-delivered by tumor-targeted NPs PEG-PLGA revealed more potent tumor growth inhibition effect when compared to the free-form drug cocktail without causing unwanted toxicity, indicating the CTCE-NP formulation could increase the therapeutic effects and therapeutic window of combined sorafenib and MEK inhibitors. This approach may be useful in safely preventing a key cell autonomous mechanism of resistance to sorafenib in HCC as well as inhibiting angiogenesis and reprogramming the immune microenvironment to promote anti-tumor immunity.
We used the murine HCC cell line HCA-1 and HCC1 derived from Mst1−/−Mst2F/− mice and the human HCC cell lines Hep3B, JHH-7, SK-Hep-1, HLF, SNU-449 and SNU-423 (purchased from ATCC, Manassas, VA). Hep3B cells were stably transduced with secreatable Gaussia luciferase (Gluc) gene by using a retroviral vector provided by Dr. Bakhos Tannous, Massachusetts General Hospital for in vivo xenograft studies. Hep3B cells were maintained in MEM-alpha medium, HLF and JHH-7 cells were maintained in DMEM/F12 medium, HCC1, SNU-449, SNU-423 and SK-Hep-1 were maintained in RPMI medium. All media were supplemented with 10% fetal bovine serum and were purchased from Invitrogen (Carlsbad, CA). The CTCE peptide (KGVSLSYRCRYSLSVGK) and scrambled peptide (SC) (LYSVKRSGCGSRKVSYL) were synthesized and purified (95% purity) by Kelowna International Scientific Inc. (Taipei, Taiwan). siRNA against ERK1/2 and CRAF and control siRNA were purchased from Dharmacon (Lafayette, CO). Sorafenib, AZD6244 and PLX4720 were obtained from MGH Pharmacy or purchased from Selleck Chemicals (Houston, TX). Dimethyl sulfoxide (DMSO), cholesterol, TPGS and coumarin 6 (C6) were purchased from Sigma-Aldrich (St. Louis, MO). Ethanol was obtained from Alfa Aesar (Ward Hill, MA). 1,2-Dioleoyl-sn-glycero-3-phosphocholine (DOPC) was purchased from Avanti Polar Lipids (Alabaster, AL). PLGA (50/50, inherent viscosity: 0.17dl/g) was purchased from Green Square Materials Incorporation (Taoyuan, Taiwan). To evaluate NF-kB activity, HCC cells were seeded in 96-well plates and transfected with NF-kB-driven luciferase reporter construct from Agilent technologies (Austin, TX).
Nanoparticles were prepared via single-step nanoprecipitation as previously described with modifications. Briefly, 0.75mg of PLGA, 0.15mg of drug cocktail, 0.375mg of TPGS, 0.04mg of DOPC, 0.04mg cholesterol and 0.08mg DSPE-PEG(2000)-maleimide in 40μl of DMSO were mixed as the oil phase and each time 10μl of oil phase was added to 280μl of deionized water dropwise under gentle stirring. The NPs were self-assembled with continuous stirring for 30minutes at room temperature. The ratio of the volumes of the oil and water phases was fixed at 1/7 (v/v) throughout all of the experiments. For peptide conjugation, peptide was reduced using immobilized TCEP disulfide reducing gel (Thermo Scientific) according to manufacturer’s recommendations. Twenty-five μl of 1mg/ml of CTCE-9908 peptides or scramble peptides were added into the emulsion where it reacted with DSPE-PEG(2000)-maleimide. Four hours later, the unreacted maleimide groups were quenched by adding l-cysteine. The solution was centrifuged at 16,220rpm for 30min at 25°C to collect the NPs, and the NPs were then resuspended in water in the volume equal to the emulsion for characterization. The same procedure was used to synthesize C6-loaded NPs.
The size and morphology of the NPs were characterized by transmission electron microscopy (TEM; H-7500, Hitachi High-Tech, Tokyo, Japan). The NPs were stained on dried formvar-coated 100-mesh copper grids at room temperature. All grids were further dried again for two days before imaging. The particle size and surface charge were examined using a Zetasizer (3000HS, Malvern Instruments Ltd., Worcestershire, UK) at room temperature.
For orthotopic human HCC model, Gluc-Hep3B cells were orthotopically implanted in livers of 7–8-week-old male nude mice in a 10μl Matrigel solution. HCC tumors in Mst1–/–Mst2F/– transgenic mice were induced by i.v. injection of Cre-adenoviruses. Male C3H mice were administration with carbon tetrachloride (CCl4) (16% [v/v] in olive oil, 100-ml gavage, 3 times per week) in combination with 5% ethanol in drinking water for 24 weeks to create a chemically induced murine HCC model. For murine orthotopic HCC model, HCA-1 cells were orthotopically implanted in the liver of 7- to 8-week-old male mice in a 100μl Matrigel solution.
All experiments were performed in accordance with relevant guidelines and regulations. All animals received humane care, in compliance with the “Guide for the Care and Use of Laboratory Animals” published by the National Academy of Sciences, and all study procedures and protocols were approved by the Animal Research Committee of Massachusetts General Hospital and National Tsing-Hua University. Free-form sorafenib was administered daily by gavage at a dose of 50mg/kg body weight to mice with established tumors. Free-form AZD6244 was administered twice daily by gavage at a dose of 25mg/kg body weight. For in vivo gene silencing, siRNA in LPD formulations were intravenously injected in orthotopic HCC-bearing mice daily at a dose of 1.2mg of siRNA/kg for 4 days. Tumors were collected 2hr after the last injection.
C6 was formulated in the NPs as a tracer molecule, with a final weight ratio of C6 to PLGA as 1/150. HCA-1, Mahlavu or Hep3B cells (10,000 cells per well) were seeded in a 12-well plate (Costar, IL, USA) and incubated for 12hours. The cells were then treated with different formulations containing C6 at 37°C for 4hours. The cells were washed with PBS, fixed with 4% paraformaldehyde (PFA) for 10minutes and counterstained with mounting solution (4′,6-diamidino-2-phenylindole or DAPI, Vector Laboratories, Burlingame, CA). The cellular uptake of C6-loaded NPs was examined and quantified using a confocal microscope (LSM-780, Carl Zeiss, Germany).
To perform a CTCE-9908 peptide competitive assay, cells were prepared as previously described. Before treatment with C6-loaded CTCE-NPs, the cells were treated with free CTCE-9908 peptide at a final concentration of 0μg/ml, 20μg/ml, or 50μg/ml for 10minutes. The cellular uptake of C6-loaded NPs was examined and quantified using a confocal microscope.
Cell viability was assessed using the MTT assay. Cells (10,000 cells per well) were seeded into 96-well plates, allowed to adhere overnight, and exposed to a range of drug concentrations. After 48 or 72hr, 10μl of 5mg/ml MTT dissolved in PBS was added to each well and incubated for 3hr in 37°C. The medium was aspirated, and 100μl of DMSO was added to each well. Absorbance was read at 570nm.
To analyze the effect of downregulation of CRAF or ERK on in vitro cytotoxic effects of sorafenib, cells were transfected with siRNA using Lipofectamine 2000 reagent (Invitrogen). Twelve hours after transfection of siRNA (50nM), media was replaced with fresh media. The trasfected cells were treated with sorafenib 24hr after transfection. After 24hr of sorafenib (2μM) exposure, apoptosis was detected by staining the cells with Annexin V and propidium iodide solution followed by flow cytometry analysis. In addition, cell viability was assessed using the MTT assay 48hours after sorafenib treatment.
Cells were lysed in lysis buffer RIPA for 30min on ice and the supernatant was collected after centrifugation at 12,000rpm. Cell lysate were separated on a 10% acrylamide gel and transferred to a PVDF membrane. Membranes were blocked for 1hr in 5% skim milk and then incubated overnight with polyclonal antibodies against p-CRAF, CRAF, p-ERK, ERK, p-p38, p38, p-AKT, AKT, p-IκBα, Bim, p-Bim (Cell Signaling, Danvers, MA) or β-actin (Sigma-Aldrich, St. Louis, MO).
HCC1 and Hep3B cell immunoprecipitation was performed using extracts prepared in lysis buffer [0.5% Nonidet P-40, 100mM NaCl, 10mM EDTA, 20mM Tris-HCl (pH 8.0), containing 1g/ml leu-peptin, 1g/ml aprotinin, 1g/ml pepstatin, 0.5mM dithio-threitol, 0.2mM sodium vanadate, 100nM microcysteine]. The extracts were sonicated on ice and clarified by centrifugation at 10,000rpm for 15min. BRAF and CRAF was co-immunoprecipitated using the anti-BRAF monoclonal antibody prebound to Protein G beads for Hep3B. CRAF was co-immunoprecipitated using the anti-BRAF monoclonal antibody (Cell Signaling, Danvers, MA) prebound to sepharose A beads for HCC1 cells. Immobilized immune complexes were washed three times, eluted in sample buffer, resolved by SDS-PAGE (10% gels), and transferred to nitrocellulose for immunoblot analysis. Membranes were blocked in 5% milk diluted in PBS for 1hr, after which they were incubated with BRAF or CRAF primary antibodies diluted in 5% milk/PBS overnight at 4°C. Membranes were washed in PBST (PBS with 0.1% Tween-20) three times and then incubated for 1hr with a secondary antibody. Membranes were washed four times and then developed by an enhanced chemiluminescence system according to the manufacturer’s instructions (PerkinElmer).
Frozen sections of Hep3B tumors were stained by using TACSTM TdT Kit (R&D Systems, Minneapolis, MN) according to the manufacturer’s recommendations. The apoptotic cells were counted in four randomly selected visual fields for each sample. The apoptotic index was calculated as the fraction of apoptotic nuclei.
Frozen tumor sections (7–8μm thick) were immunostained with primary antibodies against CRAF, p-ERK, vWF (Dako, Denmark), PD-L1 (Abcam Inc., Cambridge, MA) Granzyme B (Abcam Inc., Cambridge, MA) and CD8a (BD Biosciences, San Jose, CA) and further stained with fluorescent secondary antibodies. The sections were counterstained with DAPI (Vector Laboratories, Burlingame, CA). Samples were imaged by using Zeiss LSM 780 confocal microscope and quantified using 5–10 randomly selected fields per sample.
Cationic liposomes composed of DOTAP and cholesterol (1:1 molar ratio) were prepared by thin film hydration followed by membrane extrusion to reduce the particle size. To prepare LPD, 18μl of protamine (2mg/mL), 140μl of deionized water, and 24μl of a mixture of siRNA and calf thymus DNA (2mg/mL) were mixed and kept at room temperature for 10min before adding 120μl of cationic liposome (10mM). LPD stand at room temperature for 10min before the addition of DSPE-PEG. LPD was then mixed with 40μl of DSPE-PEG (17mg/mL) and kept at 50–60°C for 10min.
HCC tissue was obtained through tumor biopsies of patients who had undergone liver resection at the department of Surgery at the Far Eastern Memorial Hospital, Taiwan and at Fundeni Clinical Institute, Bucharest, Romania. All experiments were performed in accordance with relevant guidelines and regulations and all study procedures and protocols were approved by the Institutional Review Boards of Far Eastern Memorial Hospital (FEMH No. 104192F) and Fundeni Clinical Institute. Written informed consent was obtained from all patients prior to the operation.
All statistical analyses were performed by student t-test. Data were considered statistically significant when p value was less than 0.05.
How to cite this article: Chen, Y. et al. Overcoming sorafenib evasion in hepatocellular carcinoma using CXCR4-targeted nanoparticles to co-deliver MEK-inhibitors. Sci. Rep. 7, 44123; doi: 10.1038/srep44123 (2017).
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank Christina Koppel and Carolyn Smith for outstanding technical support and Dr. Gregory Y. Lauwers (Dept. of Pathology, MGH) for useful advice. This study was supported by NIH grant P01-CA080124. DGD’s work was supported through NIH grants R01-CA159258, R21-CA139168 and Proton Beam/Federal Share Program, and the American Cancer Society grant 120733-RSG-11–073–01-TBG. YC’s work has been supported by Ministry of Science and Technology (MOST 103-2221-E-007-032-MY2, MOST 104-2628-B-007-001-MY3), and Chang Gung Memorial Hospital-National Tsing Hua University Joint Research Grant (104N2744E1). IP was supported by a Norway-Romania joint project (grant 4SEE P4/30.06.2014). RKJ’s work was supported through NIH grants R35 CA197743 and Proton Beam/Federal Share Program.
R.K.J. received consultant fees from Ophthotech, Puretech, SynDevRx, and XTuit. R.K.J. owns equity in Enlight, Ophthotech, SynDevRx, and XTuit and serves on the Board of Directors of XTuit and the Boards of Trustees of Tekla Healthcare Investors, Tekla Life Sciences Investors, Tekla Healthcare Opportunities Fund, and Tekla World Healthcare Fund. D.G.D. received research grants from Merrimack, Leap, Bristol-Meyers-Squibb and Bayer. No reagents or funding from these companies were used in these studies. There is no significant financial or other competing interest in the work. All remaining authors have no potential conflicts to report.
Author Contributions Yunching Chen – performed experiments, supervised study and edited the manuscript text. Ya-Chi Liu - prepared Figures 4–6. Ts-Ting Lin - prepared Figures 4–6. Rakesh R. Ramjiawan - prepared Figures 1 and 3. Yun-Chieh Sung - prepared Figures 5–6. Chih-Chun Chang - prepared Figure 6. Kuo-Shyang Jeng - provided clinical study, prepared Figure S6. Chiung-Fang Chang – provided clinical material, prepared Figure S6. Chun-Hung Liu - drew Figures 4a and 7. Dong-Yu Gao - prepared Figure 5. Fu-Fei Hsu - prepared Figure 6. Annique M. Duyverman - established HCC tumor model. Shuji Kitahara - prepared Figure S8. Peigen Huang - established HCC tumor model. Simona Dima provided study materials. Irinel Popescu provided study materials. Keith T. Flaherty - supervised study. Andrew X. Zhu - supervised study. Nabeel Bardeesy - established HCC tumor model. Rakesh K. Jain - supervised study. Cyril H. Benes - supervised study. Dan G. Duda - supervised study and edited the manuscript text. All authors reviewed and approved the manuscript.