Abstract
Hepatocellular carcinoma (HCC) is the fifth most common malignancy worldwide. Vascular endothelial growth factor, platelet derived growth factor and the Raf/mitogen‐activated protein kinase/extracellular signal regulated kinase (Raf/MEK/ERK) signalling pathway regulates the growth, neovascularization, invasiveness and metastatic potential of HCC. In this study, we investigated the in vivo antitumour activity and mechanisms of action of sorafenib tosylate on four patient‐derived HCC xenografts. Sorafenib dosed at 50 mg/kg and 100 mg/kg inhibited tumour growth by 85% and 96%, respectively. Sorafenib‐induced growth suppression and apoptosis were associated with inhibition of angiogenesis, down‐regulation of phospho‐platelet‐derived growth factor receptor β Tyr1021, phospho‐eIF4E Ser209, phospho‐c‐Raf Ser259, c‐Raf, Mcl‐1, Bcl‐2, Bcl‐x and positive cell cycle regulators, up‐regulation of apoptosis signalling kinase‐1, p27 and p21. Expression of IGF‐1Rβ and phosphorylation of c‐Raf Ser338, MEK1/2 Ser217/221 and ERK1/2 Thr202/Tyr204 were increased by sorafenib treatment. Phosphorylation of mammalian target‐of‐rapamycin (mTOR) targets (p70S6K, S6R and 4EBP1) was reduced by sorafenib in sorafenib‐sensitive lines but activated in sorafenib‐less‐sensitive 10–0505 xenograft. Sorafenib‐induced phosphorylation of c‐met, p70S6K and 4EBP1 was significantly reduced when 10–0505 cells were co‐treated with anti‐human anti‐HGF antibody, suggesting that treatment with sorafenib leads to increased HGF secretion and activation of c‐met and mTOR targets. Treatment of 10–0505 tumours with sorafenib plus rapamycin resulted in growth inhibition, inhibition of vascular endothelial growth factor receptor‐2 phosphorylation, increased apoptosis and completely blocked sorafenib‐induced phosphorylation of mTOR targets and cyclin B1 expression. These data also provide a strong rationale for clinical investigation of sorafenib in combination with mTOR inhibitors in patients with HCC.
Keywords: hepatocellular carcinoma, vascular endothelial growth factor, angiogenesis, mTOR
Introduction
Hepatocellular carcinoma (HCC) is the fifth most common primary neoplasm worldwide, with approximately 660,000 deaths worldwide annually [1, 2]. Recurrence, metastasis and the development of new primary tumours are the most common causes of mortality for patients with HCC [3, 4]. Surgical resection and liver transplantation are the only proven potentially curative therapy for HCC; however, only 10% to 20% of patients undergo surgery because of poor liver function, metastases or both [5, 6, 7]. A number of chemotherapeutic agents have been evaluated for the treatment of HCC; however, no single or combination chemotherapy regimen has shown to be particularly effective [8]. Thus, novel treatments for advanced HCC are needed.
In addition to being highly angiogenic, human HCC have high expression, and enhanced activity of the mitogen‐activated protein kinase/extracellular signal regulated kinase (MEK/ERK) signalling pathway compared with adjacent non‐neoplastic liver [9]. Although Raf‐activating mutations are relatively rare events in HCC, Raf kinase is overexpressed in a high percentage of HCC patient tumours [10, 11]. Raf/MEK/ERK pathway can be activated by HBV and HCV infection and mitogenic growth factors [11, 12] and its activation is associated with aggressive tumour behaviour [13]. Because vascular endothelial growth factor receptor (VEGFR), platelet‐derived growth factor receptor (PDGFR) and Raf/MEK/ERK signalling cascades play critical role in development and progression of HCC, the use of a multikinase inhibitor to block these signalling cascades could have therapeutic efficacy.
Sorafenib (BAY 43–9006, Nexavar™, Bayer and Onyx Pharmaceuticals, Leverkusen, Germany) is a multikinase inhibitor that has been shown efficacy against a wide variety of tumours in preclinical models including HCC [14, 15]. It blocks tumour cell proliferation by targeting the Raf/MEK/ERK signalling pathway and exerts an anti‐angiogenic effect by targeting the tyrosine kinases VEGF receptor 2, VEGFR‐3, PDGFR‐β, Ret and c‐Kit [14, 16]. In addition to anti‐angiogenesis and inactivation of the Raf/MEK/ERK pathway, inhibition of phospho‐eIF4E, and Mcl‐1 protein may also contribute to its pro‐apoptotic effects [15]. Although sorafenib provides a significant improvement in overall survival in patients with HCC [17], the potential mechanisms of action that lead to the sorafenib‐mediated clinical benefits are largely unknown.
Here, we report the antitumour activity and the possible mechanisms of action of sorafenib in patient‐derived HCC xenografts.
Materials and methods
Reagents
Research grade Capsitol was purchased from CyDex, Inc. (Lenexa, KA, USA). Antibodies against p70S6K, Akt, cleaved caspase‐3, mammalian target‐of‐rapamycin (mTOR), S6R, 4EBP1 and phosphorylation‐specific antibodies against c‐met Tyr1234/1235, ERK1/2 Thr202/Tyr204, p70S6K Thr421/Ser424 and Thr389, ribosomal S6 protein (S6R) Ser235/236 and Ser240/244, 4EBP1 Thr70 and Ser37/46, phospho‐Raf Ser259, phospho‐Raf Ser338, phospho‐MEK Ser217/221, phospho‐p38 Thr180/Tyr184 and c‐Raf were obtained from Cell Signaling Technology (Beverly, MA, USA). The antibodies against phospho‐PDGFR‐β Tyr1021, phospho‐VEGFR‐2 Tyr951, phospho‐c‐kit Tyr568/570, PDGFR‐β, VEGFR‐2, IGF‐1Rβ, HGF, apoptosis signalling kinase‐1 (ASK‐1), MEK1, c‐met, cyclin D1, cyclin B1, cyclin A, Cdk‐2, Cdk‐4, cdc‐2, p21, p27, Cdk‐6, Bcl‐x, Bcl‐2, Bad, Bax and α‐tubulin were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). CD31/platelet endothelial cell adhesion molecule 1, and Ki‐67 antibodies were from Lab Vision (Fremont, CA, USA). The chemiluminescent detection system was supplied by Amersham, Pharmacia Biotech (Arlington Heights, IL, USA).
Effects of sorafenib and sorafenib plus rapamycin on the growth of subcutaneous HCC xenografts
This study received ethics board approval at the National Cancer Centre of Singapore and Singapore General Hospital. All mice were maintained according to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health, USA. They were provided with sterilized food and water ad libitum, and housed in negative pressure isolators with 12‐hr light/dark cycles.
HCC xenografts were implanted in severe combined immunodeficiency disease (SCID) mice (Animal Resources Centre, Canning Vale, Western Australia). Primary HCCs have previously been used to create xenograft lines [18], of which the following four (5–1318 (passage 18), 10–0505 (passage 13), 26–1004 (passage 21), 06–0606 [passage 19]) were used to establish tumours. Tissue fragments (containing approximately 8–10 × 106 cells) were subcutaneously implanted in both flanks of male SCID mice aged 9 to 10 weeks. The 5–1318, 06–0606, and 10–0505 lines were derived from hepatitis B virus‐positive HCC, whereas the 26–1004 and 5–1318 had wild‐type and a mutation in codon 249 of the p53 gene, respectively.
Sorafenib tosylate was dissolved in vehicle (30% Capsitol in water) at an appropriate concentration before treatment. For dose–response experiment, mice bearing the 06–0606 and 10–0505 xenografts were given four doses of sorafenib (10, 30, 50 and 100 mg/kg daily) orally for 12 days. Each treatment group comprised of five mice. To investigate the antitumour effects of sorafenib, mice bearing tumours were orally administered 50 mg/kg sorafenib daily for 12 days. Each treatment group was comprised of 14 animals and each experiment was repeated at least twice. Treatment started on day 7 after tumour implantation. By this time, the HCC xenografts reached the size of approximately 100 mm3. To study the effects of rapamycin plus sorafenib on the growth of 10–0505 xenograft, mice bearing tumours (14 per group) were orally administered either 200 μl of vehicle, or 50 mg/kg of sorafenib, or 1 mg/kg of rapamycin (Rapamune, Wyeth Pharmaceuticals Company, Guayama), or rapamycin plus sorafenib daily for indicated days. Tumour growth was monitored at least twice weekly by Vernier caliper measurement of the length and width of tumour. Tumour volume was calculated as follows: [length × width2×π/6]. At the end of the study, the mice were killed with body and tumour weights being recorded, and the tumours harvested for analysis.
The efficacy of sorafenib in reducing tumour growth was determined by the treatment (T)/control (C) ratio, where T and C are the median weight (mg) of sorafenib‐treated and vehicle‐treated tumours, respectively, on treatment day 12. T/C ratios ≤0.42 are considered an active response according to the Drug Evaluation Branch of the Division of Cancer Treatment, National Cancer Institute criteria.
Western blot analysis
To determine changes in indicated proteins, three to four independent tumours from vehicle‐ and sorafenib‐treated mice (day 12 during treatment) were homogenized separately in lysate buffer as described [18]. A total of 100 μg of proteins from a single tumour were subjected to Western blot analysis as previously described [18]. All primary antibodies were used at a final concentration of 1 μg/ml. The blots were then visualized with a chemiluminescent detection system (Amersham).
Cell culture
The 10–0505, 06–0606, and 26–1004 tumours were finely minced and washed three times with modified Eagle medium (MEM). The minced tissue was incubated with MEM medium containing 5% foetal bovine serum (FBS) and 5 mg/ml collagenase (Roche Diagnostics Corporations, Indianapolis, IN, USA) at 37°C for 12 hrs. Cells were harvested by centrifuging at 800 ×g for 10 min. The cell pellets were washed three times with serum free MEM and allowed to grow in MEM containing 10% FBS. Primary HCC cells were plated at a density of 5.0 × 106 cells per well in MEM containing 10% FBS (growth medium) for 48 hrs. Cells were treated with 3 or 6 μM of sorafenib in serum free MEM in the presence or absence of 5 μg/ml anti‐human hepatocyte growth factor (HGF) antibody (Santa Cruz) for 48 hrs. A total of 2 ml of conditioned medium from vehicle‐ or sorafenib‐treated (without anti‐human antibody) was collected and concentrated using a VIVASPIN 20 (membrane: 10,000 MWCO PES, Vivascience, Ltd., Stonehouse, UK) as described by the manufacturer and secreted HGF in conditioned medium was determined by western blotting.
To determine conditioned medium from sorafenib‐treated 10–0505 cells can independently induce mTOR signalling in the 06–0606 primary cells. The 06–0606 cells were treated with serum‐free MEM containing 10% concentrated conditioned medium (10× concentrated medium) from sorafenib‐treated 10–0505 cells in the presence or absence of 5 μg/ml anti‐human HGF antibody for 48 hrs. The levels of phospho‐c‐met, phospho‐p70S6K, and phospho‐4EBP1 in the cell lysate were determined as described above.
Immunohistochemistry
Tumour tissue samples were processed for paraffin embedding and 5‐μm sections were prepared. After blocking endogenous peroxidase activity and non‐specific staining, the sections were incubated overnight at 4°C with the primary antibodies against CD31, Ki‐67, and cleaved poly (ADP‐ribose) polymerase (PARP) to assess microvessel density, cell proliferation and apoptosis, respectively. Immunohistochemistry was performed with the streptavidin‐biotin peroxidase complex method, according to the manufacturer’s instructions (Lab Vision). For Ki‐67, only nuclear immunoreactivity was considered positive. The number of Ki‐67+ cells among at least 500 cells per region was counted and expressed as percentage values. For the quantification of mean microvessel density in sections stained for CD31, 10 random 0.159 mm2 fields at 100× magnification were captured for each tumour.
Statistical analysis
Body weight, tumour weight at time of killing, mean vessel density, Ki‐67 index and percentage of cleaved PARP+ cells were compared using ANOVA without repeated measures. Experiments were repeated at least three times with similar results.
Results
For the dose–response experiment, mice bearing the 06–0606 and 10–0505 xenografts were orally given 10, 30, 50 and 100 mg/kg sorafenib daily for 12 days. As shown in Fig. 1A, sorafenib treatment inhibited the tumour growth of 06–0606 and 10–0505 xenografts in a dose‐dependent manner (P < 0.01). The growth rate of 06–0606 and 10–0505 xenografts was also significantly reduced by sorafenib (Fig. 1B, P < 0.01). The growth pattern of 06–0606 tumour treated with 50 mg/kg dose sorafenib was quite similar to that of 10–0505 tumour treated 100 mg/kg dose sorafenib, indicating that 10–0505 is less sensitive to serafenib than 06–0606 line (Fig. 1B). The weights of 06–0606 tumours in mice that were treated with sorafenib 50 mg/kg and 100 mg/kg were approximately 13% and 5% of the controls, respectively (Fig. 1C). Progressive weight loss (5–9% of initial body weight) was observed in 100 mg group. No overt toxicity, as defined by weight loss (Fig. 1D), unkempt appearance, mortality and behaviour, was observed in 50 mg group during the course of treatment. As the 50 mg/kg daily dose gave good growth inhibition and minimal toxicity, this dose was used for all subsequent studies. As shown in Table 1, 50 mg dose of sorafenib significantly inhibited tumour growth in mice with lines 5–1318, 26–1004 and 10–0505 (P < 0.01). For 50 mg dose, the T/C ratio, where T and C are the median weight (mg) of sorafenib‐ and vehicle‐treated tumours at the end of the treatment, respectively, for 06–0606, 26–1004, 5–1318, and 10–0505 xenografts was 0.13, 0.10, 0.12 and 0.49, respectively. Because the T/C ratio for 10–0505 xenograft is greater than 0.42, sorafenib was not considered to be active against this line (Drug Evaluation Branch of the Division of Cancer Treatment, NCI criteria).
Figure 1.
Antitumour activity of sorafenib on patient‐derived HCC xenograft line 06–0606. For the dose–response experiment, mice bearing the 06–0606 and 10–0505 xenografts were orally given 10, 30, 50 and 100 mg/kg sorafenib daily for 12 days. Each treatment arm involved five independent tumour‐bearing mice representing the same xenograft line. Percentage of growth inhibition is shown in (A). Mice bearing subcutaneous 06–0606 or 10–0505 tumours were treated with vehicle, 50 mg/kg or 100 mg/kg sorafenib tosylate by gavage daily for indicated time and mean tumour volume ± S.E. at given time‐points is shown in (B). Each treatment arm involved fourteen independent tumour‐bearing mice representing the same xenograft line. (C) The corresponding tumour weight at time of killing and (D) body weight at time of killing are shown. Bars with different letters indicate P < 0.01 as determined by ANOVA.
Table 1.
Changes in body weight, tumour weight, microvessel density (CD‐31), markers of cell proliferation (Ki 67) and apoptosis (cleaved PARP) in mice treated with either vehicle or sorafenib tosylate (50 mg/kg) in four HCC xenograft lines
Lines of xenografts | Treatments | Body weight (g) | Tumour weight (mg) | Ki‐67 index (%) | Cleaved PARP (%) | Microvessel density† |
---|---|---|---|---|---|---|
06–0606 | Vehicle | 22.6 ± 0.9 | 2750 ± 345 | 27 ± 6 | 1 ± 0.5 | 26.8 ± 8 |
Sorafenib | 21.7 ± 0.8 | 410 ± 95* | 8 ± 4* | 14 ± 4* | 5 ± 3* | |
5–1318 | Vehicle | 22.9 ± 0.7 | 2565 ± 247 | 21.6 ± 8 | 1.5 ± 0.6 | 26 ± 7 |
Sorafenib | 22.5 ± 0.9 | 339 ± 84* | 7.5 ± 5* | 12.4 ± 3* | 6.8 ± 4* | |
26–1004 | Vehicle | 23.1 ± 1.0 | 2760 ± 349 | 28.5 ± 8 | 2.7 ± 1 | 20.5 ± 6 |
Sorafenib | 22.5 ± 0.8 | 289 ± 98* | 9 ± 5* | 13.2 ± 5* | 4.5 ± 3* | |
10–0505 | Vehicle | 23.4 ± 0.7 | 1670 ± 196 | 18 ± 8 | 1.2 ± 0.5 | 17 ± 5 |
Sorafenib | 22.6 ± 0.8 | 813 ± 106* | 11 ± 4* | 5.2 ± 1* | 8 ± 4* |
†Mean microvessel density of 10 random 0.159 mm2 fields at 100× magnification.
*P < 0.001 versus vehicle in same xenograft line.
Immunohistochemical analysis revealed that treatment with sorafenib led to decreased blood microvessel density in 06–0606 and 10–0505 xenografts (Fig. 2), which is consistent with inhibition of angiogenesis. The median number of CD31+ endothelial cells and Ki‐67 labelling index in sorafenib‐treated tumours was significantly reduced in all four patient‐derived xenograft lines compared to control (P < 0.01) (Table 1). The percentage of cells stained for cleaved PARP in the same samples was significantly increased in sorafenib‐treated tumours (Fig. 2 and Table 1, P < 0.01), suggesting that sorafenib caused cell death.
Figure 2.
Effects of sorafenib therapy on neovascularization and apoptosis of xenograft lines 06–0606 and 10–0505. Representative pictures of the blood vessels stained with anti‐CD31 and cleaved‐PARP+ cells in vehicle‐ and sorafenib‐treated mice, respectively, are shown. Treatment with sorafenib resulted in significantly decreased vessel density and increased cell death (P < 0.01, ANOVA). Experiments were repeated at least three times with similar results.
We investigated the potential mechanisms of sorafenib in HCC xenografts. As expected, the levels of Mcl‐1 (Figs 3A, B and 6B), phospho‐eIF4E Ser209 (Figs 3B and 6A), phospho‐PDGFR‐β Tyr1021 (Fig. 4) were reduced by sorafenib treatment. Phospho‐VEGFR‐2 Tyr951 was either unchanged or reduced by sorafenib (Fig. 4). Although total c‐Raf and basal levels of phospho‐c‐Raf Ser259 were significantly reduced by sorafenib therapy (Fig. 3A and B), IGF‐1Rβ, phospho‐c‐Raf Ser338, phospho‐MEK1/2 Ser217/221 and phospho‐ERK1/2 Thr202/Tyr204 were elevated (Figs 3 and 4). Sorafenib induced p27 and p21 expression but reduced cyclin D1, Cdk‐2, cdc‐2 and cyclin B1 levels in sorafenib‐treated 06–0606 tumours (Fig. 6A, P < 0.01). There were no significant alterations in the levels of phospho‐c‐kit Tyr568/570 (Fig. 4), Cdk‐6 and cyclin A by sorafenib therapy (data not shown).
Figure 3.
Effects of sorafenib on the phosphorylation of PDGFR‐β, VEGFR‐2 and c‐met, the Raf/MEK/ERK pathway, IGFR‐β and apoptosis in HCC xenograft lines 06–0606, 26–1004 and 10–0505. Mice bearing s.c. 06–0606, 10–0505 or 26–1004 xenografts were treated vehicle or indicated dose of sorafenib by gavage daily for 12 days as described in the section ‘Materials and methods’. Lysates from vehicle‐ and sorafenib‐treated 06–0606 (A), 26–1004 (B), and 10–0505 (C) tumours were subjected to Western blotting described in the section ‘Materials and methods’. Blots were incubated with the indicated antibodies. Representative Western blots show a significant decrease in c‐Raf, phospho‐c‐Raf Ser259, phospho‐eIF4E Ser209, phospho‐VEGFR‐2 Tyr951, phospho‐PDGFR‐β Tyr1021 and Mcl‐1 and a significant increase in cleaved PARP (P < 0.01 for all conditions, anova). Note that IGF‐IRβ, phosphorylation of ERK1/2 and c‐Raf Ser338 in sorafenib‐treated samples was elevated. Sorafenib/RAPA caused a significant increase in apoptosis as determined by the levels of cleaved PARP and completely blocked sorafenib‐induced phosphorylation of c‐met but not IGF‐1Rβ and phospho‐ERK1/2. Importantly, sorafenib/RAPA additively reduced phospho‐VEGFR‐2. Experiments were repeated at least three times with similar results.
Figure 6.
Effects of sorafenib, rapamycin (RAPA), and the combination of sorafenib and RAPA (sorafenib/RAPA) on phosphorylation of mTOR targets, cell cycle regulators and apoptotic regulatory proteins in HCC xenograft lines 06–0606 and 10–0505. Mice bearing s.c. 10–0505 tumours were treated with vehicle, or 50 mg/kg sorafenib, or 1 mg/kg RAPA, or sorafenib plus RAPA orally daily for 12 days. Mice bearing 06–0606 tumours were treated with vehicle, or 50 mg/kg sorafenib for 12 days. Lysates from vehicle and drug‐treated tumours were subjected to Western blotting described in the section ‘Materials and methods’. Blots were incubated with the indicated antibodies. Representative Western blots show that sorafenib caused a significant decrease in phospho‐eIF4E Ser209, cyclin D1, cyclin B1, Cdk‐2 and cdc‐2, and a significant increase in p27 and p21 (A), phospho‐p70S6K Thr421/Ser424 and Thr389, phospho‐4EBP1 Thr70, phospho‐S6R Ser235/236 and Ser242/244 (C) in sorafenib‐sensitive 06–0606 xenografts (P < 0.01 for all conditions, anova). In sorafenib‐less‐sensitive 10–0505 tumours, sorafenib/RAPA inhibited cyclin D1 and Cdk‐2 expression (B). Sorafenib‐induced cyclin B1 expression and phosphorylation of mTOR targets were completed blocked by sorafenib/RAPA (P < 0.01). Experiments were repeated at least three times with similar results.
Figure 4.
Effects of sorafenib on phosphorylation of PDGFR‐β, VEGFR‐2, c‐kit, expression of IGF‐1Rβ, expression of Bcl‐2 family proteins and apoptosis in HCC xenograft lines 5–1318 and 26–1004. Mice bearing s.c. 5–1318 or 26–1004 xenografts were treated vehicle or indicated dose of sorafenib by gavage daily for 12 days as described in the section ‘Materials and methods’. Lysates from vehicle‐ and sorafenib‐treated 5–1318 and 26–1004 tumours were subjected to Western blotting described in the section ‘Materials and methods’. Blots were incubated with the indicated antibodies. Representative Western blots show a significant decrease in phospho‐PDGFR‐β Tyr1021, phospho‐p38 Thr180/Tyr184, c‐Raf, Bcl‐2 and Bcl‐x and a significant increase in IGF‐1Rβ, phospho‐MEK1/2 Ser217/221, phospho‐Raf Ser338, ASK‐1 and cleaved PARP (P < 0.01 for all conditions, anova). Experiments were repeated twice with similar results.
Because apoptosis in mammalian cells is regulated by Bax, Bad, Bcl‐2 and Bcl‐x [19] and overexpression of ASK‐1 is also sufficient to cause apoptosis in a number of cell lines through a mitochondria‐dependent caspase activation pathway [20], we determined whether sorafenib‐induced apoptosis in 26–1004 (wild‐type p53) and 5–1318 (mutant p53) xenografts was also associated with the modulation of these proteins. As shown in Figs 4 and 6B, Bcl‐2 and Bcl‐x levels in sorafenib‐treated tumours were significantly decreased (P < 0.01). Bad and Bax expression was either unaffected or slightly elevated by sorafenib therapy. The expression of ASK‐1 was significantly increased following sorafenib treatment (P < 0.01). The results indicated that sorafenib‐induced apoptosis in HCC xenografts was independent of p53 status and associated with decrease in anti‐apoptotic Bcl‐2 and Bcl‐x proteins and increase in apoptotic ASK‐1 protein.
We next examined the expression of the mTOR signalling pathway in sorafenib‐sensitive 06–0606 and sorafenib‐less‐sensitive 10–0505 xenografts. Although phosphorylation of mTOR targets (p70S6K, S6R and 4EBP1) in sorafenib‐sensitive 06–0606 (Fig. 6C), 26–1004 (Fig. 3B) and 5–1318 (data not shown) xenografts was inhibited by sorafenib therapy, phosphorylation of p70S6K Thr421/Ser424 and Thr389, 4EBP1 Thr70, and S6R Ser235/236 (P < 0.01) was induced (Fig. 6B). Phospho‐c‐met Tyr1234/1235 was also elevated by sorafenib treatment (Fig. 3C). Total c‐met was decreased (Fig. 3C) and total p70S6K, 4EBP1 and S6R were unchanged (Fig. 6D), suggesting mTOR targets and c‐met are activated in 10–0505 xenograft following sorafenib treatment.
To determine whether inhibition of mTOR targets would enhance the antitumour activity of sorafenib in sorafenib‐less‐sensitive xenografts, mice bearing 10–0505 tumours were treated with the combination of sorafenib and rapamycin (sorafenib/RAPA). Figure 5A showed that single‐agent RAPA and sorafenib, and sorafenib/RAPA significantly inhibited the growth rate of 10–0505 xenografts compared with control (P < 0.01). Although no significant differences in growth inhibition were observed between RAPA and sorafenib groups, sorafenib/RAPA significantly caused tumour regression in this xenograft line (P < 0.01). Sorafenib/RAPA therapy completely blocked sorafenib‐induced up‐regulation of phospho‐c‐met Tyr1234/1235 (Fig. 3C), cyclin B1, Cdk‐2, and phosphorylation of mTOR targets (Fig. 6D). The sorafenib/RAPA also reverted sorafenib‐induced suppression of p27, Mcl‐1, Cdk‐4 (Fig. 6B) and phospho‐eIF4E Ser209 (Fig. 6D) without affecting sorafenib‐induced IGF‐1Rβ expression and phosphorylation of ERK1/2 (Fig. 3C). In addition, significantly increased in ASK‐1 levels (Fig. 6B) and apoptosis as determined by the levels of cleaved PARP were observed (Fig. 3C), whereas either drug alone had minimal effect on the levels of this apoptotic marker. Furthermore, this therapy caused a marked decrease in the levels of phospho‐VEGFR‐2 Tyr951 (Fig. 3C) and cyclin D1 (Fig. 6B) achieved by sorafenib/RAPA were greater than either drug alone.
Figure 5.
Effects of sorafenib, rapa‐mycin (RAPA), and sorafenib plus RAPA on growth rate of patient‐derived HCC xenograft line 10–0505. Mice bearing xenograft line 10–0505 were treated with vehicle, or 50 mg/kg sorafenib, or 1 mg/kg RAPA, or sorafenib plus RAPA orally daily for 21 days. Each treatment arm involved 14 independent tumour‐ bearing mice. (A) Mean tumour volume ± S.E. at given time‐points and (B) tumour weight at day 21 during the treatment are shown. Bars with different letters indicate P < 0.01 as determined by anova.
In order to address the mechanism by which sorafenib induced activation of c‐met and mTOR targets, we evaluated the effect of sorafenib on HGF secretion and the phosphorylation levels of c‐met, p70S6K, and 4EBP1in primary 10–0505 and 06–0606 cells. As shown in Fig. 7A , HGF accumulation was found only in the conditioned medium of primary 10–0505 cells but not 06–0606 and 26–1006 cells following sorafenib treatment. Figure 7B showed that in vehicle‐treated 10–0505 cells, sorafenib increased levels of phospho‐c‐met Tyr1234/1235, phospho‐p70S6K Thr421/Ser424 and phospho‐4EBP1 Thr70, which was significantly reduced when cells were co‐treated with anti‐HGF antibody. Furthermore, conditioned medium from sorafenib‐treated 10–0505 cells was able to induce activation of c‐met and mTOR targets in 0606–06 cells, which was reduced in the presence of HGF antibody. Taken together, the data suggest that secreted HGF in conditioned medium of 10–0505 cells is, in part, responsible for c‐met and mTOR activation. Moreover, treatment of 10–0505 tumours with sorafenib plus rapamycin resulted in growth inhibition and completely blocked sorafenib‐induced phosphorylation mTOR targets, suggesting that activation of mTOR pathway by sorafenib treatment is responsible for the lower sensitivity to sorafenib observed in this model.
Figure 7.
Effects of sorafenib on HGF accumulation in conditioned medium and phosphorylation of c‐met Tyr1234/1235, p‐p70S6K Thr421/Ser424 and 4EBP1 Thr70 in primary HCC cells. HGF accumulation was detected in conditioned medium of primary 10–0505 but not 06–0606, and 26–1004 cells following sorafenib therapy (A). Sorafenib also increased c‐met, p70S6K and 4EBP1 phosphorylation, which was significantly reduced in the presence of anti‐human HGF antibody (B). Conditioned medium from sorafenib‐treated 10–0505 cells induced activation of c‐met and mTOR targets in 06–0606 cells, which was also reduced by HGF antibody.
Discussion
Our present study demonstrates that sorafenib alone actively inhibits three of four patient‐derived HCC xenograft lines, confirming previous study [17] and suggesting that at least some patients may be amenable to single agent therapy. Sorafenib‐induced growth suppression is associated with inhibition of angiogenesis, induction of apoptosis and suppression of phospho‐PDGFR‐β, phospho‐eIF4E, Bcl‐x, Bcl‐2 and Mcl‐1 and inactivation of downstream targets of the mTOR pathway. As shown in Table 1, the untreated 10–0505 xenograft line had the lower density of CD31+ endothelial cells than the responsive 06–0606 xenograft line. It is possible that less vessel dense tumours might be less sensitive to sorafenib. However, we observe that the 26–1004 xenograft is more sensitive to sorafenib than 10–0505 line despite both of them having the similar microvessel density (Table 1). This observation suggests that microvessel density may contribute in part but not all the differences in dose–response to sorafenib between 06–0606 and 10–0505 xenografts. Indeed, we observe that sorafenib‐induced growth inhibition also associates with up‐regulation of p27 and p21, and inhibition of cyclin D1, Cdk‐2, Cdk‐4 and cyclin B1. In contrast, sorafenib increased Cdk‐2, cyclin B1 and phospho‐c‐met, and reduction in p27 in sorafenib‐less‐sensitive 10–0505 xenograft. In our present study, sorafenib/RAPA combination not only abolishes sorafenib‐induced mTOR activation, phosphorylation of c‐met, and cyclin B1 expression but also restores the levels of p27. Although activation of MEK/ERK pathway has been shown to contribute to cell proliferation and survival [21], we observe that sorafenib inhibits cell proliferation in the presence of phospho‐ERK1/2 Thr202/Tyr204. Based on the above data, it appears that sorafenib‐mediated inhibition of cyclin D1, cdc‐2, Cdk‐2 and cyclin B1 expression, enhancement of p27 translation, and down‐regulation of phospho‐p70S6K, phospho‐S6R and phospho‐4EBP1 also play an important role in sorafenib‐inhibited tumour growth. These potential kinetic changes would have a negative effect on G1 as well as in the S and G2‐M phases of the cell cycle, the ultimate anti‐proliferative effects of sorafenib on HCC.
In the present study, we are not able to measure the relative contributions of the anti‐angiogenic activity of sorafenib versus its direct antitumour activity. As shown in Figs 3C and 4, inhibition of phosphorylation of PDGFR‐β is detected in the sorafenib‐treated tumours. Significant decrease in phospho‐VEGFR‐2 is only observed in sorafenib/RAPA therapy. It is known that VEGF promotes the proliferation, migration, invasion and survival of endothelial cells [22]. It is possible that the potent anti‐angiogenic effects of sorafenib and sorafenib/RAPA in HCC xenograft may therefore be a result of direct functional impairment of tumour‐vessel‐associated endothelial cells and vascular smooth muscle cells. By disrupting VEGFR signalling, sorafenib/RAPA is able to inhibit VEGF‐driven tubular formation, and endothelial cell migration and sprouting, leading to a striking reduction in tumour growth and microvessel density as observed in sorafenib/RAPA‐treated HCC xenografts. Because PDGF is angiogenic for microvascular sprouting endothelial cells, and especially PDGF‐BB and its receptors recruit pericytes and smooth muscle cells around nascent vessel sprouts [23], the inhibition of PDGFR‐β signalling by sorafenib and sorafenib/RAPA may inhibit pericyte recruitment and attachment to endothelial cells that would confer resistance to VEGFR antagonists on endothelial cells [24]. This suggests that sorafenib or sorafenib/RAPA may potentially be useful in maintaining dormancy of micro‐metastasis and preventing the development of recurrence or metastasis after surgical resection of a primary tumour. Although sorafenib might exert a direct antitumour effect through targets such as c‐kit in addition to the VEGFR and PDGFR‐β, there are compelling results suggesting that the role of c‐kit in HCC is insignificant [25, 26, 27, 28].
The mechanisms underlying the sorafenib‐ and sorafenib/ RAPA‐induced apoptosis in HCC xenografts are not well understood. Previous study shows that sorafenib‐mediated apoptosis in HCC xenografts is associated with inhibition of Mcl‐1 expression [29]. In the present study, we notice that in sorafenib/RAPA treated tumours, apoptosis still occurs despite normal Mcl‐1 levels suggesting that an alternative pathway(s) is able to overcome this anti‐apoptotic pathway. Indeed, we observe two distinct apoptotic pathways in sorafenib and sorafenib/RAPA‐treated tumours. The first one involves the Bcl‐2 family. Figure 4 showed that both sorafenib and sorafenib/RAPA treatment result in no change in expression of pro‐apoptotic Bad and Bax whereas expression of anti‐apoptotic Bcl‐2 and Bcl‐x proteins is inhibited. Thus, there is a shift in the dynamic balance between the outputs of pro‐apoptotic and anti‐apoptotic pathways following these therapies. The second pathway involves ASK‐1, which is an important mediator of apoptotic signalling by a variety of death stimuli such as tumour necrosis factor‐α, Fas activation, oxidative stress and DNA damage [30, 31, 32]. Our in vivo study shows that sorafenib and sorafenib/RAPA increase ASK‐1 expression and reduce c‐Raf levels (Figs. 4 and 6B). Down‐regulation of c‐Raf would enhance apoptotic activity of MST2 [33] and have more ASK‐1 available to induce apoptosis through a mitochondria‐dependent caspase activation pathway [20].
Our study shows that although c‐Raf and phospho‐c‐Raf Ser 259 are inhibited, phosphorylation of c‐Raf Ser338, MEK1/2 and ERK1/2 is induced in four xenograft lines studied. The mechanisms accounting for these phenomena are unknown. Although Raf/MEK/ERK pathway can be activated by HBV [11, 12], no selective activation of the Raf/MEK/ERK in four xenografts is observed regardless of whether they come from HBV+ or HBV– tumours. As shown in Figs. 3 and 4, up‐regulation of IGF‐1Rβ is observed in sorafenib‐treated tumours. Given that the Raf/MEK/ERK pathway can be activated by IGF‐1R and other receptor tyrosine kinases [9, 34], increased expression of IGF‐1Rβ following sorafenib treatment may be responsible for the observed elevation of phospho‐c‐Raf Ser338, phospho‐MEK1/2 and phospho‐ERK1/2.
In the present study, we observe that sorafenib treatment causes activation of p70S6K, S6R and 4EBP1 in sorafenib‐less‐sensitive line 10–0505. In HCC, total p70S6K expression is positively correlated with tumour nuclear grade, and inversely correlated with tumour size [35]. The upstream kinase(s) responsible for activation of downstream targets of mTOR in HCC is (are) still unknown. It has been reported that the PIK3CA gene, encoding p110α of the class IA PI3K, is mutated in approximately 35.6% of HCC cases [36] and that mTOR targets are positively regulated by PI3K/Akt pathway and other RTKs. In the present study, we observe that sorafenib also increases phosphorylation of c‐met in sorafenib‐less‐sensitive 10–0505 xenograft. This raises the possibility that c‐met activation may involve in the activation of mTOR target proteins in 10–0505 tumours. Indeed, treatment of primary 10–0505 cells with sorafenib results in HGF accumulation in the conditioned medium and increased levels of phospho‐c‐met Tyr1234/1235 and mTOR targets, which are significantly reduced when cells are co‐incubated with anti‐HGF antibody. Furthermore, conditioned medium from sorafenib‐treated 10–0505 cells is able to induce activation of c‐met and mTOR targets in sorafenib‐ sensitive 06–0606 cells. This activation is significantly reduced in the presence of HGF antibody, suggesting that treatment with sorafenib leads to increased HGF secretion, which is responsible for c‐met and mTOR activation.
In our present study, sorafenib/RAPA combination not only inhibits tumour growth in sorafenib‐less‐sensitive 10–0505 xenograft but also abolishes sorafenib‐induced mTOR activation, phosphorylation of c‐met, and cyclin B1 and cyclin D1 expression. It has been reported that complete remission of lung metastases was observed in a patient on rapamycin after liver transplant for metastatic HCC [37]. Stippel et al.[38] also reports a tumour‐free survival in another HCC patient given rapamycin after a liver transplant. This patient underwent a bilateral salphingo‐oophorectomy for HCC metastases and immunosuppression was switched to rapamycin monotherapy. Fourteen months after this procedure the patient was asymptomatic with stable liver function. The above clinical data and our present results suggest that sorafenib plus RAPA may be a promising drug combination that can overcome the activation of mTOR targets.
In summary, we have demonstrated that oral delivery of sorafenib causes growth inhibition of patient‐derived HCC xenografts. Our data implicate that induction of apoptosis and inhibition of tumour angiogenesis and cell cycle by sorafenib may be responsible for its antitumour activity in vivo and suggest that sorafenib/RAPA combination may be useful for the treatment of HCC.
Acknowledgements
This work was supported by grants from the Singapore Cancer Syndicate (SCS‐HS0021, SCS−AS0032 and SCS‐AMS0086) to H.H.
References
- 1. Jemal A, Murray T, Ward E, et al . Cancer statistics, 2005. CA Cancer J Clin. 2005; 55: 10–30. [DOI] [PubMed] [Google Scholar]
- 2. Smith RA, Cokkinides V, Eyre HJ. American Cancer Society Guidelines for the Early Detection of Cancer, 2005. CA Cancer J Clin. 2005; 55: 31–44. [DOI] [PubMed] [Google Scholar]
- 3. Nagasue N, Kohno H, Chang YC, et al . Liver resection for hepatocellular carcinoma. Results of 229 consecutive patients during 11 years. Ann Surg. 1993; 217: 375–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Yamamoto J, Kosuge T, Takayama T, et al . Recurrence of hepatocellular carcinoma after surgery. Br J Surg. 1996; 83: 1219–22. [PubMed] [Google Scholar]
- 5. Colombo M. Hepatocellular carcinoma. J Hepatol. 1992; 15: 225–36. [DOI] [PubMed] [Google Scholar]
- 6. Lai EC, Fan ST, Lo CM, et al . Hepatic resection for hepatocellular carcinoma. An audit of 343 patients. Ann Surg. 1995; 221: 291–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Takenaka K, Kawahara N, Yamamoto K, et al . Results of 280 liver resections for hepatocellular carcinoma. Arch Surg. 1996; 131: 71–6. [DOI] [PubMed] [Google Scholar]
- 8. Zhu AX. Systemic therapy of advanced hepatocellular carcinoma: how hopeful should we be Oncologist. 2006; 11: 790–800. [DOI] [PubMed] [Google Scholar]
- 9. Huynh HT, Nguyen TT, Chow PK, et al . Over‐expression of MEK‐MAPK in hepatocellular carcinoma: Its role in tumor progression and apoptosis. BMC Gastroenterol. 2003; 3: 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Karasarides M, Chiloeches A, Hayward R, et al . B‐RAF is a therapeutic target in melanoma. Oncogene. 2004; 23: 6292–8. [DOI] [PubMed] [Google Scholar]
- 11. Tannapfel A, Sommerer F, Benicke M, et al . Mutations of the BRAF gene in cholangiocarcinoma but not in hepatocellular carcinoma. Gut. 2003; 52: 706–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Hwang YH, Choi JY, Kim S, et al . Over‐expression of c‐raf‐1 proto‐oncogene in liver cirrhosis and hepatocellular carcinoma. Hepatol Res. 2004; 29: 113–21. [DOI] [PubMed] [Google Scholar]
- 13. Schmitz KJ, Wohlschlaeger J, Lang H, et al . Activation of the ERK and AKT signaling pathway predicts poor prognosis in hepatocellular carcinoma and ERK activation in cancer tissue is associated with hepatitis C virus infection. J Hepatol. 2008; 48: 83–90. [DOI] [PubMed] [Google Scholar]
- 14. Wilhelm SM, Carter C, Tang L, et al . BAY 43–9006 exhibits broad spectrum oral antitumor activity and targets the RAF/MEK/ERK pathway and receptor tyrosine kinases involved in tumor progression and angiogenesis. Cancer Res. 2004; 64: 7099–109. [DOI] [PubMed] [Google Scholar]
- 15. Liu L, Cao Y, Chen C, et al . Sorafenib blocks the RAF/MEK/ERK pathway, inhibits tumor angiogenesis, and induces tumor cell apoptosis in hepatocellular carcinoma model PLC/PRF/5. Cancer Res. 2006; 66: 11851–8. [DOI] [PubMed] [Google Scholar]
- 16. Carlomagno F, Anaganti S, Guida T, et al . BAY 43–9006 inhibition of oncogenic RET mutants. J Natl Cancer Inst. 2006; 98: 326–34. [DOI] [PubMed] [Google Scholar]
- 17. Llovet J, Ricci S, Mazzaferro V, et al . Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 2008; 359: 378–390. [DOI] [PubMed] [Google Scholar]
- 18. Huynh H, Soo KC, Chow PK, et al . Xenografts of human hepatocellular carcinoma: a useful model for testing drugs. Clin Cancer Res. 2006; 12: 4306–14. [DOI] [PubMed] [Google Scholar]
- 19. Boise LH, Gonzalez‐Garcia M, Postema CE, et al . Bcl‐x, a bcl‐2‐related gene that functions as a dominant regulator of apoptotic cell death. Cell . 1993; 74: 597–608. [DOI] [PubMed] [Google Scholar]
- 20. Hatai T, Matsuzawa A, Inoshita S, et al . Execution of apoptosis signal‐regulating kinase 1 (ASK1)‐induced apoptosis by the mitochondria‐dependent caspase activation. J Biol Chem. 2000; 275: 26576–81. [DOI] [PubMed] [Google Scholar]
- 21. Ballif BA, Blenis J. Molecular mechanisms mediating mammalian mitogen‐activated protein kinase (MAPK) kinase (MEK)‐MAPK cell survival signals. Cell Growth Differ. 2001; 12: 397–408. [PubMed] [Google Scholar]
- 22. Tran J, Rak J, Sheehan C, et al . Marked induction of the IAP family antiapoptotic proteins survivin and XIAP by VEGF in vascular endothelial cells. Biochem Biophys Res Commun. 1999; 264: 781–8. [DOI] [PubMed] [Google Scholar]
- 23. Conway EM, Collen D, Carmeliet P. Molecular mechanisms of blood vessel growth. Cardiovasc Res. 2001; 49: 507–21. [DOI] [PubMed] [Google Scholar]
- 24. Erber R, Thurnher A, Katsen AD, et al . Combined inhibition of VEGF and PDGF signaling enforces tumor vessel regression by interfering with pericyte‐mediated endothelial cell survival mechanisms. FASEB J. 2004; 18: 338–40. [DOI] [PubMed] [Google Scholar]
- 25. Becker G, Schmitt‐Graeff A, Ertelt V, et al . CD117 (c‐kit) expression in human hepatocellular carcinoma. Clin Oncol (R Coll Radiol). 2007; 19: 204–8. [DOI] [PubMed] [Google Scholar]
- 26. Potti A, Ganti AK, Tendulkar K, et al . HER‐2/neu and CD117 (C‐kit) overexpression in hepatocellular and pancreatic carcinoma. Anticancer Res. 2003; 23: 2671–4. [PubMed] [Google Scholar]
- 27. Went PT, Dirnhofer S, Bundi M, et al . Prevalence of KIT expression in human tumors. J Clin Oncol. 2004; 22: 4514–22. [DOI] [PubMed] [Google Scholar]
- 28. Eckel F, von Delius S, Mayr M, et al . Pharmacokinetic and clinical phase II trial of imatinib in patients with impaired liver function and advanced hepatocellular carcinoma. Oncol. 2005; 69: 363–71. [DOI] [PubMed] [Google Scholar]
- 29. Yu C, Bruzek LM, Meng XW, et al . The role of Mcl‐1 downregulation in the proapoptotic activity of the multikinase inhibitor BAY 43–9006. Oncogene. 2005; 24: 6861–9. [DOI] [PubMed] [Google Scholar]
- 30. Ichijo H, Nishida E, Irie K, et al . Induction of apoptosis by ASK1, a mammalian MAPKKK that activates SAPK/JNK and p38 signaling pathways. Science. 1997; 275: 90–4. [DOI] [PubMed] [Google Scholar]
- 31. Chang HY, Nishitoh H, Yang X, et al . Activation of apoptosis signal‐regulating kinase 1 (ASK1) by the adapter protein Daxx. Science. 1998; 281: 1860–3. [DOI] [PubMed] [Google Scholar]
- 32. Saitoh M, Nishitoh H, Fujii M, et al . Mammalian thioredoxin is a direct inhibitor of apoptosis signal‐regulating kinase (ASK) 1. EMBO J. 1998; 17: 2596–606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Chen J, Fujii K, Zhang L, et al . Raf‐1 promotes cell survival by antagonizing apoptosis signal‐regulating kinase 1 through a MEK‐ERK independent mechanism. Proc Natl Acad Sci USA. 2001; 98: 7783–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Lee HS, Huang AM, Huang GT, et al . Hepatocyte growth factor stimulates the growth and activates mitogen‐activated protein kinase in human hepatoma cells. J Biomed Sci. 1998; 5: 180–4. [DOI] [PubMed] [Google Scholar]
- 35. Sahin F, Kannangai R, Adegbola O, et al . mTOR and P70 S6 kinase expression in primary liver neoplasms. Clin Cancer Res. 2004; 10: 8421–5. [DOI] [PubMed] [Google Scholar]
- 36. Lee JW, Soung YH, Kim SY, et al . PIK3CA gene is frequently mutated in breast carcinomas and hepatocellular carcinomas. Oncogene. 2005; 24: 1477–80. [DOI] [PubMed] [Google Scholar]
- 37. Elsharkawi M, Staib L, Henne‐Bruns D, et al . Complete remission of posttransplant lung metastases from hepatocellular carcinoma under therapy with sirolimus and mycophenolate mofetil. Transplantation. 2005; 79: 855–7. [DOI] [PubMed] [Google Scholar]
- 38. Stippel DL, Kasper HU, Schleimer K, et al . Successful use of sirolimus in a patient with bulky ovarian metastasis of hepatocellular carcinoma after liver transplantation. Transplant Proc. 2005; 37: 2185–7. [DOI] [PubMed] [Google Scholar]