AZD6244 (ARRY-142886) enhances the antitumor activity of rapamycin in mouse models of human hepatocellular carcinoma

Authors

  • Hung Huynh PhD

    Corresponding author
    1. Laboratory of Molecular Endocrinology, Division of Molecular and Cellular Research, Humphrey Oei Institute of Cancer Research, National Cancer Centre, Singapore
    • Laboratory of Molecular Endocrinology, Division of Cellular and Molecular Research, National Cancer Centre of Singapore, Singapore 169610
    Search for more papers by this author
    • Fax: (011) 6562265694


Abstract

BACKGROUND:

The protein kinase B (AKT)/mammalian target of rapamycin (AKT/mTOR) and mitogen activated protein kinase/extracellular regulated kinase kinase/extracellular regulated kinase (MEK/ERK) signaling pathways have been shown to play an important role in hepatocellular carcinoma (HCC) growth and angiogenesis, suggesting that inhibition of these pathways may have therapeutic potential.

METHODS:

We treated patient-derived HCC xenografts with 1) mTOR inhibitor rapamycin (RAPA); 2) MEK inhibitor AZD6244 (ARRY-142886); and 3) AZD6244 plus RAPA (AZD6244/RAPA). Western blotting was used to determine pharmacodynamic changes in biomarkers relevant to angiogenesis, mTOR pathway, and MEK signaling. Apoptosis, microvessel density, and cell proliferation were analyzed by immunohistochemistry.

RESULTS:

We report here that pharmacological inhibition of the MEK/ERK pathway by AZD6244 enhanced the antitumor and antiangiogenic activities of mTOR inhibitor RAPA in both orthotopic and ectopic models of HCC. Such inhibition led to increased apoptosis, decreased angiogenesis and cell proliferation, reduced expression of positive cell cycle regulators, and increase in proapoptotic protein Bim.

CONCLUSIONS:

Our findings indicate that the AZD6244/RAPA combination had antitumor and antiangiogenic effects in preclinical models of human HCC. Given the urgent need for effective therapies in HCC, clinical evaluating AZD6244/RAPA combination seems warranted. Cancer 2010. © 2010 American Cancer Society.

Hepatocellular carcinoma (HCC) is the fifth most common cancer worldwide.1 Surgical resection, percutaneous ablation, and liver transplantation are the only potentially curative therapies for this disease; however, <20% of patients with HCC can undergo surgery because of poor liver function and advanced disease.2-3 Of those having surgery, the 5-year survival rate is limited to 25% to 50%. Recurrence, metastasis, and the development of new primary tumors are the most common causes of mortality for patients with HCC.4-6 Until the advent of sorafenib (Nexavar; Bayer and Onyx Pharmaceuticals), patients with advanced HCC had limited treatment options. Given the increasing incidence of the disease around the world,7 newly effective treatments for advanced HCC are needed.

The rat sarcoma-activated factor (raf)/mitogen activated protein kinase/extracellular regulated kinase kinase/extracellular regulated kinase (Raf/MEK/ERK) and phosphatidylinositide 3 kinase/protein kinase B (AKT)/mammalian target of rapamycin (PI3K/AKT/mTOR) pathways are among the most critical cellular signaling pathways that support hepatocarcinogenesis.8-9 Although Raf-activating mutations were relatively rare events in HCC, Raf kinase was activated in a high percentage of HCC tumors.10 HBV or HCV infection or mitogenic growth factors could activate the Raf/MEK/ERK pathway10-11 and its activation was associated with aggressive tumor behavior.12 Preclinical studies have shown that MEK/ERK inhibition resulted in suppressing HCC growth.13-15 It has been reported that the PIK3CA gene was mutated16 and PTEN was inactivated in high proportion of HCC.17 These led to activation of Akt kinase and the mTOR pathway.17 In HCC, total p70S6K expression was positively correlated with tumor grade, and inversely correlated with tumor size.18 We recently reported that inhibition of the mTOR pathway by RAD00119 or rapamycin/bevacizumab combination resulted in tumor growth inhibition.20 Furthermore, the 26-1004 line, which had a 16 bp deletion in exon 8 of the PTEN gene was more sensitive to rapamycin than other xenograft lines. Although complete remission of lung metastases was observed in a patient on rapamycin after a liver transplant for metastatic HCC21 and tumor-free survival was also noted in another HCC patient given rapamycin after a liver transplant,22 rapamycin clinical trials have turned out to be less successful than expected. A potential mechanism of resistance to mTORC1 inhibitors came about by the discovery of a negative feedback loop in which mTORC1 inhibition led to Akt activation23 or activation of the ERK cascade.24 These observations formed the basis for our hypothesis that targeting the mTOR and ERK pathways in combination would be effective for inhibiting tumorigenicity in HCC.

In the present study, we investigated the consequences of combinational inhibition of the mTOR and MEK/ERK signaling pathways.

MATERIALS AND METHODS

Reagents

Research grade Capsitol was purchased from CyDex, Inc., Lenexa, Kan. Primary antibodies against CD31/platelet endothelial cell adhesion molecule 1, p130 Rb, and Ki-67 were from Lab Vision, Fremont, Calif. Antibodies against cleaved poly (ADP-ribose) polymerase (PARP), Bim, Puma, p21, p27, p70S6K, Akt, S6R, 4EBP1, and phospho-specific antibodies against Rb Ser780 and Ser807/811, ERK1/2 Thr202/Tyr204, p70S6K Thr421/Ser424, and Thr389, S6R Ser235/236, 4EBP1 Thr70, and Akt Ser473 were obtained from Cell Signaling Technology, Beverly, Mass. Antibodies against p110 Rb, cyclin D1, cdk-2, cdk-4, cdc-2, Cyclin A, Cyclin B1, Bax, Bad, Bcl-2, Bcl-xL, ERK1/2, and α-tubulin were from Santa Cruz Biotechnology Inc., Calif.

This study received ethics board approval at the National Cancer Centre Singapore and Singapore General Hospital. All mice were maintained according to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health.

Primary HCCs have previously been used to create xenograft lines.19-20, 25 The following 5 lines (01-0207, 06-0606, 26-1004, 25-0705A, and 5-1318) were used to establish tumors in male severe combined immunodeficiency disease (SCID) mice (Animal Resources Centre, Canning Vale, Western Australia) aged 9 to 10 weeks.

To investigate the antitumor effects of AZD6244 (AstraZeneca, Alderley Park, Macclesfield, UK), rapamycin (RAPA), and AZD6244 plus RAPA, mice bearing indicated tumors (14 per group) were orally administered daily either 200 μL of vehicle, or 25 mg/kg AZD6244, 1 mg/kg RAPA, or RAPA plus AZD6244 for 16 days starting from day 7 after tumor implantation. By this time, the tumors reached the size of approximately 100 mm3. Tumor growth was monitored and tumor volume calculated as described.19 At the end of the study, the mice were sacrificed with body and tumor weights being recorded, and the tumors harvested for analysis.

In Vivo Orthotopic HCC Model

To create an orthotopic model of human HCC, mice were anesthetized with a Ketamine/Diazepam solution (50 mg/kg Ketamine hydrochloride, Rotexmedica, Trittau, Germany, and 5 mg/kg Diazepam, Alantic, I.M.). Liver was exposed after an upper-middle incision. Tumor fragments (containing approximately 5 × 106 cells in 30 μL medium-Matrigel mixture) were implanted into the lobe of the liver by means of a 0.9 × 38 mm needle (20G1½TW, Becton Dickingson) and calibrated pushbutton syringe. The incision was closed using a running suture of 5-0 silk. All mice developed tumors within 7 to 10 days of tumor tissue injection and grew rapidly thereafter. Upon tumor establishment, mice (5-8 per group) were allocated to the following groups: 1) control (vehicle only), 2) AZD6244 (25 mg/kg/day, orally), 3) RAPA (1 mg/kg/day, orally), and 4) AZD6244 /RAPA. They were treated for 16 days.

Western Blot Analysis

Three to 4 independent tumors per group were homogenized. Eighty μg of total proteins per sample were subjected to Western blotting according to standard protocols.19, 25 Blots were incubated with indicated primary antibodies and 1:7500 horseradish peroxidase-conjugated secondary antibodies. All primary antibodies were used at a final concentration of 1 μg/mL. The blots were then observed with a chemiluminescent detection system (Amersham).

Immunohistochemistry

Microvessel density, cell proliferation, and apoptosis were assessed by immunohistochemistry as described.19-20 For apoptosis and cell proliferation assessment, the number of cleaved PARP-positive and Ki-67-positive cells among at least 500 cells per region was counted and then expressed as percentage values. For the quantification of mean vessel density in sections stained for CD31, 10 random 0.159 mm2 fields at 100× magnification were captured for each tumor and microvessels were quantified.

Determination of Plasma Vascular Endothelial Growth Factor Receptor

Animals were sacrificed on Day 16 after treatment and approximately 500-800 μL of blood were collected per mouse. The concentration of vascular endothelial growth factor receptor (VEGF) in the plasma was measured using Endogen Human VEGF ELISA kits (Pierce Biotechnology, Rockford, Ill) according to the protocol of the manufacturer. The sensitivity for VEGF kits is <8.0 pg/mL.

Statistical Analysis

Body weight at sacrifice, final mean weight of tumors, plasma VEGF, mean vessel density, Ki-67 index, and percentage of cleaved PARP-positive cells were analyzed by Student t test. For all statistical analyses, significance was established at P < .05.

RESULTS

Our previous studies showed that the MEK/ERK27 and mTOR19-20 signaling pathways were activated in HCC. In addition, mTORC1 inhibition by RAD001 led to activation of the ERK cascade in RAD001-treated tumors.24 On the basis of the above-mentioned information, we hypothesized that targeted inhibition of the mTOR and ERK signaling pathways would inhibit HCC growth. Therefore, we developed experimental paradigms to test the consequences of inhibiting these pathways in orthotopic and ectopic models of human HCC. To achieve inhibition of ERK signaling, we used AZD6244, the MAP kinase 1 (MEK) inhibitor already being characterized in preclinical studies of HCC.14-15 Although the efficacy of rapalogs as single agents may be limited, RAPA is particularly suitable for use in combination therapy. For the preclinical studies, we chose to use RAPA (rather than 1 of the new derivatives), because it has been used in patients with HCC,21-22 is commercially available, and, therefore, logically more feasible to use in combination with AZD6244. We used phosphorylated S6 ribosomal protein (phospho-S6R) and/or phospho-4EBP1 Thr70, as indicators of the mTOR pathway activity and phosphorylated ERK (phospho-ERK) as an indicator of MEK activity. Figure 1 shows that oral delivery of 1 mg/kg of RAPA or 25 mg/kg AZD6244 was sufficient to inhibit the phosphorylation of mTOR targets (S6R and 4EBP1) and ERK1/2 at Thr202/Tyr204, respectively, confirming our previous reports.14-15, 20 Therefore, we implemented a once-daily dosing schedule using 1 mg/kg of RAPA and 25 mg/kg of AZD6244 for subsequent studies.

Figure 1.

Effects of RAPA and AZD6244 on phosphorylation of S6R, 4EBP1, and ERK in 01-0207 xenograft are shown. Mice bearing 01-0207 tumors (14 mice/group) were treated with vehicle or indicated dose of RAPA or AZD6244 by gavage daily for 16 days. Lysates from vehicle-treated and drug-treated tumors were subjected to Western blotting, as described in the Materials and Methods section. Blots were incubated with the indicated antibodies. Representative Western blots are shown. Densitometric data (fold) are shown below each group. Asterisks (*) indicated significant difference (P < .05, Student t test). Experiments were repeated at least twice with similar results.

To investigate the effects of AZD6244/RAPA combination on the growth of HCC xenografts grown in SCID mice, cohort groups that comprised mice bearing 06-0606, 25-0705A, 01-0207, or 26-1004 tumors were randomly assigned to receive vehicle, or single agents (RAPA or AZD6244), or the combination therapy daily for 16 days. Our previous studies showed that the mTOR and MEK/ERK pathways in these xenograft lines were relatively active.14-15, 19-20 Therapy initiated after tumor establishment (approximately 100 mm3) and lasted for 16 days. Although AZD6244 modestly inhibited the growth rate of all 4 lines at each following up time point (Fig. 2A-D), RAPA significantly suppressed the growth rate of 25-0705A, 26-1004, and 01-0207 lines (P < .05) but not 06-0606. AZD6244/RAPA was markedly inhibitory compared with control or single agents (P < .001). Although AZD6244 modestly reduced final mean weight of tumors, RAPA significantly decreased it compared with control (Table 1; P < .05), which is in agreement with the growth rate results. However, AZD6244/RAPA showed significant reduction in tumor weight when compared with AZD6244 or RAPA alone (Table 1, P < .01). Efficacy of AZD6244/RAPA was further evaluated by comparing the final mean weight of tumors in the drug-treated study arm (T) to that of the control study arm (C). The T/C ratio with a value of <0.42 was considered an active response (Drug Evaluation Branch of the Division of Cancer Treatment, NCI criteria). As shown in Table 1, the T/C values less than 0.21 were observed in all 4 lines treated with AZD6244/RAPA for 16 days, suggesting that this combination was quite active. The treatments were well-tolerated and produced minimal toxicity as determined by body weight loss (Table 1).

Figure 2.

Illustrated are effects of AZD6244, RAPA, and AZD6244/RAPA on the growth rate of patient-derived HCC xenografts. Mice bearing 06-0606, 01-0207, 26-1004, and 25-0705A tumors (14 mice/group) were treated with vehicle, AZD6244 (25 mg/kg/day), RAPA (1 mg/kg/day), or AZD6244/RAPA for 16 days, as described in the Materials and Methods section. (A) Mean tumor volume at given time points for lines 01-0207, (B) 06-0606, (C) 26-1004, and (D) 25-0705A are shown. Data were expressed as the mean ± SD. Experiments were repeated twice with similar results.

Table 1. Effects of Rapamycin, AZD6244, and AZD6244, Plus RAPA (AZD6244/RAPA) on Body Weight, Tumor Burden, Microvessel Density, Cell Proliferation, Plasma VEGF Levels, and Apoptosis of Four HCC Lines
Lines of XenograftsTreatmentsBody at Sacrifice, gTumor Weight, mg% T/CMicrovessela DensityKi-67 Index, %Cleaved PARP, %Plasma VEGF, pg/mL
  • RAPA indicates rapamycin; VEGF, vascular endothelial growth factor receptor; HCC, hepatocellular carcinoma; T/C, drug-treated study arm/control study arm; PARP, cleaved poly (ADP-ribose) polymerase.

  • The data were expressed as the mean ± SE (standard error of the mean). Different letters indicate significant difference (P < .05, Student t test).

  • a

    Mean microvessel density of 10 random 0.159 mm2 fields at ×100 magnification.

01-0207Vehicle23.6 ± 1.61349 ± 188A10016.4 ± 5A36.3 ± 5A1.1 ± 0.5A180 ± 37A
AZD624422.4 ± 1.41006 ± 168A74.613.7 ± 3A29.6 ± 4.1A, B4.7 ± 1.1B160 ± 29A
RAPA23.1 ± 0.8566 ± 130B426 ± 2B20 ± 4.3B1.7 ± 0.4A50 ± 15B
AZD6244+RAPA23.1 ± 1.1245 ± 51C18.23.8 ± 3B9 ± 3C8.3 ± 1.1C24 ± 11C
06-0606Vehicle23.1 ± 0.92340 ± 390A10030.1 ± 6A52.2 ± 8A0.9 ± 0.3A461 ± 89A
AZD624422.1 ± 0.81445 ± 179B61.824.2 ± 5A37.1 ± 8A, B6.2 ± 1.5B472 ± 58A
RAPA23.2 ± 0.81670 ± 168B71.416.5 ± 5A, B27.4 ± 7B1.6 ± 0.7A307 ± 45B
AZD6244+RAPA22.2 ± 0.9407 ± 89C17.412.6 ± 4B14.1 ± 4C10.3 ± 1.5B112 ± 45C
26-1004Vehicle23 ± 1.01238 ± 260A10023.5 ± 6A42.6 ± 8A2 ± 0.8A418 ± 88A
AZD624422.4 ± 1.0982 ± 169A79.318 ± 5A, B37.1 ± 8A6.6 ± 1.7B367 ± 79A
RAPA23 ± 0.9370 ± 88B29.97.8 ± 3B, C20.7 ± 6B2.9 ± 0.9A120 ± 35B
AZD6244+RAPA22.4 ± 1.1256 ± 44B20.74 ± 2C11.6 ± 4C12.4 ± 2C100 ± 24B
25-0705AVehicle22.3 ± 0.91165 ± 270A10016.7 ± 4A42.4 ± 7A0.7 ± 0.4A205 ± 45A
AZD624421.5 ± 0.8871 ± 140A74.812 ± 4A31.6 ± 6A4.9 ± 0.9B238 ± 40A
RAPA22.5 ± 1.3470 ± 89B40.36 ± 3A, B19.8 ± 3B3.6 ± 1.1B100 ± 34B
AZD6244+RAPA21.1 ± 0.9232 ± 45C19.94 ± 2B11.5 ± 1.7C11.2 ± 1.3C48 ± 25B

Because mTOR activation was important for VEGF expression28 and RAPA inhibited tumor growth through antiangiogenic activity related to impaired production of VEGF,29 we determined the levels of plasma VEGF in 4 treatment arms by ELISA. As shown in Table 1, VEGF levels in the plasma derived from RAPA-treated animals were significantly lower than those in the plasma derived from vehicle-treated and AZD6244-treated mice (P < .05). A more striking reduction in plasma VEGF was observed in the AZD6244/RAPA combination. Taken together, these data suggest that combined blockade of the mTOR and MEK/ERK pathways resulted in reduction in plasma VEGF.

Next, we evaluated representative tumor sections from 4 treatment arms using immunohistochemistry. Staining results for CD31 (a marker of angiogenesis), Ki-67 (a marker for proliferation), and cleaved PARP (a marker for apoptosis) for 4 treatments are shown in Table 1. RAPA alone and AZD6244/RAPA significantly inhibited tumor cell proliferation in all 4 lines studied compared with the control (Table 1; P < .01). Although AZD6244 had insignificant effect on microvessel density and cell proliferation, it significantly increased apoptosis (P < .05). Interestingly, AZD6244/RAPA therapy showed significant reduction in tumor cell proliferation, microvessel density, and markedly increased apoptosis when compared with control or AZD6244 alone (Table 1; P < .01). The results suggest that AZD6244 produced a predominantly apoptotic effect with a minimal effect on tumor cell proliferation, while RAPA inhibited tumor cell proliferation with modestly antiangiogenesis.

To gain initial insights regarding the mechanistic basis for the combinatorial effects of AZD6244/RAPA, changes in the phosphorylation of key proteins in the VEGFR-2, MEK/ERK, and mTOR pathways and cell cycle regulators as well as apoptotic proteins in 01-0207 xenograft were examined by western blotting. Figure 3A shows that while VEGFR-2 levels were not affected by any treatments, phosphorylation of VEGFR-2 at Tyr951 in AZD6244/RAPA-treated tumors was significantly reduced (P < 0.05), indicating that VEGFR-2 activity was reduced. AZD6244 and AZD6244/RAPA significantly inhibited phosphorylation of ERK at Thr202/Tyr204 (P < .01). As expected, RAPA significantly decreased the levels of phospho-p70S6K Thr389, phospho-4EBP1 Thr70, and phospho-S6R Ser235/236 (P < .01), indicating that the mTOR pathway was inactivated. Although AZD6244 modestly inhibited cdk-2 and cdc-2, RAPA and AZD6244/RAPA induced significant reductions in the levels of p21, cdc-2, cyclin B1, cdk-2, and phospho-Rb Ser780 and Ser807/811 (Fig. 3B; P < .01). In addition, up-regulation of p27 and p130 Rb in RAPA and AZD6244/RAPA-treated tumors was also observed, suggesting that RAPA and AZD6244/RAPA caused cell cycle arrest. Although AZD6244 alone significantly increased the levels of Bim, cleaved caspase 3 and cleaved PARP (P < .01), elevation of these apoptotic markers was most marked when the drugs were used in combination (Fig. 3C). There were no significant alterations in expression of the antiapoptotic proteins Bax, Bcl-2, Bcl-xL, Puma, and Bad after either treatment (Fig. 3C). Similar data were obtained when 06-0606, 25-0705A, and 26-1004 tumors were analyzed (data not shown). The results suggest that AZD6244/RAPA induced apoptosis probably through Bim-dependent mechanisms.

Figure 3.

Effects of RAPA, AZD6244, and AZD6244/RAPA on biomarkers are relevant to (A) angiogenesis, mTOR, and MEK signaling, (B) cell proliferation, (C) and apoptosis in 01-0207 xenograft. Mice bearing 01-0207 tumors were randomized (14 mice/group) and treated with vehicle, AZD6244 (25 mg/kg/day), RAPA (1 mg/kg/day), or AZD6244/RAPA for 16 days. Lysates from vehicle-treated and drug-treated tumors were subjected to Western blot analysis, as described in the Materials and Methods section. Blots were incubated with the indicated antibodies. Representative blots are shown. Densitometric data (fold) are shown below each group. Asterisks (* or **, * and **) indicated significant difference (P < .05, Student t test). Experiments were repeated twice with similar results.

In orthotopic HCC models, all mice developed tumors within 7 to 10 days of tumor tissue injection and grew rapidly thereafter. By using this model, we conducted the 4-arm therapeutic study to compare the effect of AZD6244, RAPA, and AZD6244/RAPA on the growth of HCC xenografts. Figure 4A showed that AZD6244 had little antitumor activity against the 5-1318 and 25-0705A lines. The antineoplastic activity of RAPA was higher than that of AZD6244, and the antitumor effect of the combination was superior to that of either AZD6244 or RAPA alone.

Figure 4.

Effects of RAPA, AZD6244, and AZD6244/RAPA on tumor growth and phosphorylation of mTOR targets and ERK, and apoptosis in HCC xenografts implanted orthotopically are depicted. Mice bearing 25-0705A and 5-1318 tumors were randomized (14 mice/group) and treated with vehicle, AZD6244 (25 mg/kg/day), RAPA (1 mg/kg/day), or AZD6244/RAPA for 16 days. Representative pictures of vehicle-treated and drug-treated tumors are shown in panel A. Lysates from vehicle-treated and drug-treated 5-1315 tumors were subjected to Western blot analysis, as described in the Materials and Methods section. Representative blots are shown. Densitometric data (fold) are shown below each group. Asterisks (* or **, * and **) indicated significant difference (P < .05, Student t test). Experiments were repeated twice with similar results.

We next investigated the phosphorylation status of the mTOR targets and ERK pathway as well as apoptotic markers in 5-1318 tumors. As shown in Figure 4B, RAPA and AZD6244/RAPA induced significant reductions in the levels of phospho-p70S6K Thr389, phospho-S6R Ser235/236, and phospho-4EBP1 Thr70 (P < .001), confirming that the mTOR pathway was inactivated. As expected, AZD6244 alone and AZD6244/RAPA suppressed phosphorylation of ERK at Thr202/Tyr204. Although AZD6244 alone increased the levels of Bim and cleaved PARP, elevation of these apoptotic markers was most marked when the drugs were used in combination (Fig. 4B). Similar data were obtained when 25-0705A tumors were analyzed (data not shown).

Next, we evaluated representative tumor sections from 4 treatment arms using immunohistochemistry. Representative staining results for phospho-S6R Ser235/236, CD31, Ki-67, and cleaved PARP are shown in Figure 5. RAPA alone and AZD6244/RAPA reduced phospho-S6R Ser235/236, blood vessel density, and Ki-67 positive cells compared with control (Fig. 5 and Table 2, P < .01). In addition, apoptosis (cleaved PARP positive cells) was observed in RAPA/AZD6244-treated tumors, and this was markedly increased in the combination compared with RAPA and AZD6244 alone (Fig. 5 and Table 2). The results suggest that AZD6244 produced a predominantly apoptotic effect with a minimal effect on proliferation, while RAPA was predominantly cytostatic, with mild antiangiogenic inhibition and modest apoptotic induction. Interestingly, combination therapy resulted in decreased proliferation, inhibition of angiogenesis, and elevated levels of apoptosis (Table 2). Similar data were obtained when 25-0705A tumors were analyzed (data not shown).

Figure 5.

Effects of RAPA, AZD6244, and AZD6244/RAPA therapies on angiogenesis, cell proliferation, apoptosis and phospho-S6R Ser235/236 in 5-1318 line are depicted. The blood vessels were stained with anti-CD31, proliferative cells stained with anti-Ki-67, apoptotic cells stained with anticleaved-PARP, and cells with activated mTOR stained with phospho-S6R Ser235/236 antibodies. Representative sections of vehicle, AZD6244, RAPA, and AZD6244/RAPA are shown. Experiments were repeated twice with similar results (Magnification, ×400).

Table 2. Effects of RAPA, AZD6244, and AZD6244/RAPA on Body Weight at Sacrifice, Microvessel Density, Cell Proliferation, and Apoptosis of HCC Xenografts Implanted Orthotopically
Lines of XenograftsTreatmentsBody at Sacrifice, gMicrovessela DensityKi-67 Index, %Cleaved PARP, %
  • RAPA indicates rapamycin; HCC, hepatocellular carcinoma; PARP, cleaved poly (ADP-ribose) polymerase.

  • The data were expressed as the mean ± SE (standard error of the mean). Different letters indicated significant difference (P < .05, Student t test).

  • a

    Mean microvessel density of 10 random 0.159 mm2 fields at ×100 magnification.

5-1318Vehicle21.7 ± 1.230.1 ± 6.2A55.9 ± 8.1A0.9 ± 0.4A
RAPA21.4 ± 1.016.4 ± 3B24.2 ± 5B3.4 ± 1.3B
AZD624420.9 ± 0.825.1 ± 5A36.2 ± 7B6.6 ± 2.1B
RAPA/AZD624420.5 ± 0.88.1 ± 3C7.2 ± 3C12.9 ± 2C
25-0705AVehicle22.4 ± 0.920.6 ± 4A46.1 ± 5A0.8 ± 0.4A
RAPA22.3 ± 1.07.2 ± 3B20.2 ± 5B3.8 ± 1.2B
AZD624419.7 ± 0.811.5 ± 3A,B35.3 ± 6B5.3 ± 2B
RAPA/AZD624421.4 ± 1.14 ± 2C10.8 ± 3C8.9 ± 3C

DISCUSSION

HCC is the fifth most common malignancy worldwide.1 Although several new chemotherapies have been explored and refined in the last decade, the mortality rate due to HCC remains unchanged. New therapeutic strategies for this disease are urgently needed. Although many drugs demonstrated preclinical positive results, only some of them are likely to move forward to phase 2 clinical trials.30 One of the challenges the investigators face in preclinical testing of targeted drugs in HCC is the lack of models that recapitulate the human disease. We previously reported the establishment of patient-derived xenografts from HCC tumors.25 We showed that these xenografts exhibit cellular and tissue characteristics that resemble the original tumors, and certain therapies merit further investigation in clinical trials.15, 19-20, 26 One disadvantage of the ectopic xenograft model is that the site of implantation does not reproduce the tumor microenvironment. To circumvent this disadvantage, intact fragments of human HCC tumors are implanted into the liver of immunodeficient mice to create surgical orthotopic models. These orthotopic models of human HCC may serve as better tools for studying HCC drug-targeting processes, particularly with respect to modeling tumor host interactions, angiogenesis, and invasion, and may, therefore, more closely represent clinical disease. In the present study, we use both ectopic and orthotopic models of human HCC to define the functional significance of combinatorial inhibition of the mTOR and MEK/ERK pathways for HCC tumorigenicity.

We previously reported that the mTOR20 and MEK/ERK27 signaling pathways are over-expressed in HCC. On the basis of these findings, we hypothesize that combinatorial inhibition of these pathways would be effective for the treatment of HCC. We now demonstrate that inhibition of the mTOR and MEK/ERK signaling pathways leads to inhibition of tumor growth, reduction in cellular proliferation, suppression of angiogenesis, and increase in apoptosis in both orthotopic and ectopic models of human HCC. Five lines tested have sustained tumor growth inhibition after treatment with AZD6244/RAPA combination. Our observations are in agreement with a recent study by Waugh Kinkade and colleagues, who report that MEK inhibitor PD0325901 plus RAPA synergistically inhibits growth and cell proliferation in androgen-independent and refractory prostate tumors.31 Because HCC is a vascular tissue and one third of HCCs are driven by proliferative signals generated from tyrosine receptor kinases (RTK), Ras/Raf/MEK/ERK, and PI-3K/Akt/mTOR pathways,32-33 the described antitumor and antiangiogenic activities of the AZD6244/RAPA support further investigation of this combination in clinical trials.

Although AZD6244/RAPA inhibits tumor growth in preclinical HCC models, the molecular mechanisms for the crosstalk between RAPA and AZD6244 are not known. It has been proposed that mTORC1 inhibition increases RTK/IRS-1/PI3K activity toward Ras/MAPK, therefore, promoting both Akt activation and ERK phosphorylation in what comprises a dual feedback mechanism.24, 34 In the present study, we observe that MEK inhibition results in elevation of phospho-Akt Ser 473, which is upstream of the mTOR pathway. Despite the observation of up-regulation of phospho-Akt, phosphorylation of 4EBP1, p70S6K, and S6 ribosomal protein remains unchanged (Fig. 3A). The mechanism(s) responsible for this effect remains to be determined.

In this study, we observe that AZD6244/RAPA is more effective in inhibiting cell proliferation and inducing apoptosis than either agent alone. These effects are associated with dephosphorylation of S6 ribosomal protein and p110 Rb, reduction in the expression of positive cell cycle regulators, and increase in proapoptotic protein Bim. In addition, the expression of p27 and p130 Rb was also significantly elevated in response to RAPA and AZD6244/RAPA therapies. It is possible that elevation of p130 Rb and p27 coupled with decreased expression of cdk-2, cdk-6, and dephosphorylation of Rb lead to G1 cell cycle arrest. In the present study, phospho-S6R is reproducibly suppressed by RAPA and AZD6244/RAPA in all xenografts tested, pointing to a mechanism that perturbs formation of the preinitiation complex, which may, in turn, alter translation efficiencies of many mRNAs that regulate cell proliferation. Support of this hypothesis in a previous study35 showed that the conditionally deleted S6 gene in adult liver results in failure of cell proliferation after partial hepatectomy. Up-regulation of the proapoptotic regulator Bim by AZD6244/RAPA may also contribute to its apoptotic activity. Particular functions for Bim have been described in the regulation of apoptosis associated with thymocyte negative selection and after growth factor withdrawal, during which Bim expression is elevated.36-38 It is possible that up-regulation of Bim by AZD6244 and AZD6244/RAPA allows more Bim to bind to Bcl-2 and Bcl-xL, thereby antagonizing their antiapoptotic activities, leading to Bax-dependent apoptogen release, caspase activation, and cell death.

In this study, RAPA and AZD6244/RAPA reduce circulating VEGF concentrations possibly as a reflection of their effect on tumor size. Changes in plasma VEGF cannot distinguish between a blockade of VEGF production by the tumor tissue, which is the source of VEGF or a combination of reduced VEGF production and small tumor size, but the data suggest that plasma VEGF may be a useful marker of efficacy as recently described for the VEGF antibody bevacizumab.39 Because VEGF is an angiogenic, proliferation, and survival factor for vascular endothelial cells and is produced by both cancer cells and stromal cells, creating a microenvironment favorable for tumor growth,40 decrease in plasma VEGF and probably tumor VEGF by RAPA and AZD6244/RAPA, would inhibit endothelial cell proliferation and promote endothelial cell death, leading to decreased tumor microvessel density. Consistent with this hypothesis, RAD001 has recently been shown to inhibit tumor vascularization directly via endothelial and smooth muscle cells and pericytes, and indirectly via VEGF production.19, 41

In summary, our present study provides the evidence that AZD6244/RAPA combination is more effective in suppressing HCC tumor growth, inhibiting angiogenesis and cell proliferation, and inducing apoptosis than either agent alone. Our findings and others31 exemplify how the efficacy of MEK or mTOR inhibitors can be improved in clinics and provide the rationale for combining MEK and mTORC1 inhibitors in the treatment of HCC.

Acknowledgements

We thank Dr. Paul Smith (AstraZeneca Pharmaceuticals, Cheshire, UK) for the AZD6244.

CONFLICT OF INTEREST DISCLOSURES

This work was supported by grants from the Singapore Cancer Syndicate (SCS-AS0032, SCS-HS0021 and SCS-AMS0086) to Huynh Hung.

Ancillary