Enhanced Anticancer Activity of PF-04691502, a Dual PI3K/mTOR Inhibitor, in Combination With VEGF siRNA Against Non-small-cell Lung Cancer.

Lung cancer is the leading cause of cancer deaths in both men and women in the United States accounting for about 27% of all cancer deceases. In our effort to develop newer therapy for lung cancer, we evaluated the combinatory antitumor effect of siRNA targeting VEGF and the PI3K/mTOR dual inhibitor PF-04691502. We analyzed the anticancer effect of siRNA VEGF and PF-04691502 combination on proliferation, colony formation and migration of A549 and H460 lung cancer cells. Additionally, we assessed the combination treatment antiangiogenic effect on human umbilical vein endothelial cells. Here, we show for the first time that the antiangiogenic siRNA VEGF potentiates the PF-04691502 anticancer activity against non-small-cell lung cancer. We observed a significant (P < 0.05) decrease in cell viability, colony formation, and migration for the combination comparing with the single drug treatment. We also showed a significant (P < 0.05) enhanced effect of the combination treatment inhibiting angiogenesis progression and tube formation organization compared to the single drug treatment groups. Our findings demonstrated an enhanced synergistic anticancer effect of siRNA VEGF and PF-04691502 combination therapy by targeting two main pathways involved in lung cancer cell survival and angiogenesis which will be useful for future preclinical studies and potentially for lung cancer patient management.

Introduction

Lung cancer is the leading cause of cancer-related mortality among both men and women, with an estimated 158,040 deaths in 2015 in the United States.[1] Non–small-cell lung cancer (NSCLC) is the most prevalent type of lung cancer, with more than 85% of the cases.[1] Majority of patients newly diagnosed with NSCLC present with locally advanced or metastatic disease beyond the scope of surgical cure.[2] As a result, systemic chemotherapy continues to be the mainstay of treatment to improve the survival and quality of life of patients.[3] In the last years, newer drugs that specifically target molecular pathways in NSCLC have been FDA approved showing promising results for the treatment of lung cancer.[4] Despite recent advances, response rates in NSCLC remain <50% and the 5-year survival rate for stage IV disease is about 1%.[1] To address this problem, attention has been focused on finding novel combinations of anticancer agents to achieve enhanced efficacy with decreased side effects.

Lung cancer has wide molecular heterogeneity and therefore, warrants the targeting of multiple pathways for effective therapy.[5] Tumor angiogenesis is a hallmark of cancer.[6] Currently, newer approaches using antiangiogenic agents in combination with other anticancer drugs have generated clinical interest.[7] Vascular endothelial growth factor (VEGF) is a major angiogenesis inducer that mediates numerous changes within the tumor vasculature, including endothelial cell proliferation, vascular permeability, migration, invasion, and survival.[8] VEGF activates the VEGF receptor (VEGFR) inducing the phosphorylation of several proteins and the activation of different signaling pathways (). A consistent increase in VEGF expression has been observed in solid tumors of different origins.[9] In NSCLC, the VEGF overexpression (61 to 92% of the tumors) is associated with tumor growth, metastasis, and poor survival.[10,11] Anti-VEGF treatment has been proven to have a survival benefit as anticancer therapy to prevent NSCLC progression[12] but clinical trials failed to demonstrate a meaningful clinical benefit to patients. The use of VEGF antibodies has been limited due to poor efficacy and adverse effects.[13,14] Therapeutic antibodies have some functional limitations as inadequate pharmacokinetics, poor oral bioavailability and tissue accessibility, incomplete absorption, nonlinear distribution and elimination, and impaired interactions with the immune system.[15] Small interference RNA (siRNA) represents an emerging therapeutic tool for cancer therapy and may overcome some of these limitations.[16] The advantages of siRNA are its high degree of efficacy and safety. siRNA can potently inhibit gene expression with high sequence selectivity, decreasing the protein expression instead of only sequestering the protein. Opko Health produced the first siRNA therapeutic to reach a phase 3 clinical trial, bevasiranib, targeting VEGF.[17] VEGF-siRNA (siRNA-V) has demonstrated promising in vitro and in vivo antitumor activity,[18,19] and will be an efficient alternative for lung cancer treatment.

The phosphatidylinositol 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) signaling pathway plays a central role in regulating cell growth, proliferation, survival, angiogenesis, metabolism, and motility.[20] The PI3K pathway has been extensively studied in relation to a variety of cancers in recent years.[21] Deregulation of PI3K signaling cascade is known to be involved in lung tumorigenesis and it has been associated with high grade tumors and more advanced disease in NSCLC.[22] Several mechanisms of crosstalk have been reported between VEGF and PI3K pathways.[23] Activation of the PI3K/Akt/mTOR cascade in tumor cells can increase VEGF secretion by both HIF1α-dependent and -independent mechanisms.[24] Several mTOR inhibitors have been FDA approved in recent years and some PI3K inhibitors are already in phase 3 clinical trials. Emerging clinical data show limited single-agent activity of inhibitors targeting PI3K or mTOR at tolerated doses. PI3K/mTOR dual inhibitors target the active sites of both proteins inhibiting the pathway both upstream and downstream of Akt, overcoming the mTORC1–p70S6K–IRS1 negative feedback loop, and blocking PI3K-independent mTOR activation[21,25] (). Dual PI3K/mTOR inhibitors are being tested in preclinical and early-stage clinical studies and have shown promising results in terms of safety and efficacy.[26,27] Some of these agents are beginning to enter phase 2 clinical trials in a variety of diseases.[22] The dual inhibitor PF-04691502 (PF0) potently inhibits PI3K and mTOR in biochemical assays.[28]

Considering the crucial role of VEGF and PI3K/mTOR pathways in lung cancer and the crosstalk connection between them, we hypothesize that the combination of siRNA-V and PF0 will have an enhanced anticancer effect by reducing tumor cell growth, survival, and migration comparing with single-agent treatment. To the best of our knowledge, we demonstrated for the first time that the anticancer activity of PF0 was enhanced in a synergistic manner in combination with VEGF siRNA against lung cancer. We have adopted a unique approach by combining siRNA-V treatment with PF0 to target three important proteins, VEGF, PI3K, and mTOR involved in NSCLC to enhance the therapeutic outcome. PF0 and siRNA-V synergistic combination may improve therapeutically relevant selectivity and enable improved control of complex biological systems offering the possibility of dose reduction and decrease adverse side effects. The results of this study may pave the way toward new therapies for lung cancer.

Results

Dose response cytotoxicity analysis

The in vitro cytotoxicity of PF0 (5 to 5,000 nmol/l) or siRNA-V (5 to 150 nmol/l) was evaluated against A549 cells after 24, 48, and 72 hours treatment (, and Supplementary Figure S1). PF0 and siRNA-V cytotoxicity was found to be dose and time dependent. PF0 IC50 values in A549 cells were 209.77 ± 34.23 nmol/l; 151.45 ± 35.34 nmol/l; 191.60 ± 39.66 nmol/l after 24, 48, and 72 hours treatment respectively ( and Supplementary Figure S1a). siRNA-V has limited effect on A549 cell viability with a maximum decrease of 27% after 72 hours 100 nmol/l siRNA-V treatment (). IC50 values for siRNA-V in A549 cells were 19.04 ± 14.71 nmol/l; 28.49 ± 14.22 nmol/l; 58.89 ± 78.10 nmol/l after 24, 48, and 72 hours of treatment ( and Supplementary Figure S1b). In H460 cells, PF0 (5 nmol/l to 200 µmol/l) or siRNA-V (5 to 200 nmol/l) cytotoxicity was also found to be dose and time dependent (, and Supplementary Figure S2). IC50 values for PF0 in H460 cells were 1,965.5 ± 131.7 nmol/l; 1,080.1 ± 307.1 nmol/l; 936.7 ± 174.6 nmol/l after 24, 48, and 72 hours of treatment respectively ( and Supplementary Figure S2a). Cell viability was decreased a maximum of 42% after 72 hours of 100 nmol/l siRNA-V treatment (). IC50 values for siRNA-V were 26.19 ± 9.18 nmol/l; 37.78 ± 3.71 nmol/l; 40.79 ± 13.73 nmol/l after 24, 48, and 72 hours of treatment against H460 cells ( and Supplementary Figure S2b).

siRNA-V and PF0 combination cytotoxicity assays

We evaluated the combination cytotoxicity effect of siRNA-V (50 nmol/l) and PF0 (5 to 500 nmol/l for A549, and 10 to 5,000 nmol/l for H460 cells) compared to the single agents (). After 24, 48, and 72 hours treatment with siRNA-V and PF0, we observed a significant (P < 0.05) decrease in relative cell viability for the combination compared to single-agent treatments (,). The combination cytotoxicity assays demonstrated that siRNA-V potentiates the anticancer activity of PF0 against A459 and H460 cells.

VEGF silencing effect analysis by quantitative real-time PCR

We analyzed the effect of siRNA-V and PF0 combination and single agents on VEGF mRNA expression in A549 and H460 cells. Cells were treated with siRNA-V (50 nmol/l) and PF0 (200 nmol/l for A549 and 2,000 nmol/l for H460) combination and single agents for 48 hours. PF0 and siRNA-V concentrations were selected based on the previous single-agent cytotoxicity results. VEGF and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA levels were evaluated using a quantitative real-time polymerase chain reaction (PCR). In A549 cells, the siRNA-V treatment showed a significant (P < 0.05) decrease of 62.58 ± 16.60% in VEGF gene expression compared to control (). siRNA-V and PF0 combination treatment showed a significantly (P < 0.01) decreased VEGF expression of 75.06 ± 13.21% compared to control. In H460 cells, we observed a significant (P < 0.001) decrease in the relative VEGF mRNA expression after siRNA-V transfection by 92.22 ± 3.10% as well as after the combination treatment 69.59 ± 7.71% compared to control cells ().

Secreted VEGF protein expression analysis by ELISA

Human VEGF protein concentration was measured by Enzyme-linked immunosorbent assay (ELISA) in A549 and H460 cells after 24, 48, and 72 hours siRNA-V and PF0 combination and single-agent treatments. In A549 and H460 control cells, the VEGF levels increased in a time-dependent manner (,). ELISA showed significantly (P < 0.05) decreased VEGF protein levels after exposure to PF0 in both A549 (200 nmol/l) and H460 (2,000 nmol/l) cells after 24, 48, and 72 hours comparing with control cells. ELISA showed significant (P < 0.01) differences in VEGF protein expression after siRNA-V treatment (50 nmol/l) compared with control samples for A549 and H460 cells after 24, 48, and 72 hours treatment (,). The combination treatment resulted in a significantly (P < 0.05) stronger reduction in the VEGF protein secretion when compared with single-agent treatments in both A549 and H460 cells, with almost undetectable VEGF levels observed after 24, 48, and 72 hours (,).

Western blot analysis of PI3K and mTOR downstream effectors

Western blot analysis was used to examine whether the PF0 treatment and the combination with siRNA-V could effectively inhibit the phosphorylation of target proteins in the PI3K and mTOR pathway. Akt and p70S6K phosphorylation was significantly (P < 0.05) suppressed by PF0 (at 200 nmol/l in A549 and 2,000 nmol/l concentration in H460 cells) as single drug or in combination with 50 nmol/l siRNA-V in both NSCLC cell lines and after 24, 48, and 72 hours treatment (, and Supplementary Figures S3 and S4).

Inhibition of anchorage-dependent colony formation

Colony formation assay was used to investigate long-term effect of PF0 and siRNA-V on proliferation of A549 and H460 cells. We treated cells with siRNA-V (50 nmol/l) and PF0 (200 nmol/l for A549 and 2,000 nmol/l for H460) combination and single agents (–). A549 relative colony-forming ability was reduced after PF0 treatment to 77.50 ± 3.79% and after exposure to siRNA-V to 62.10 ± 12.12% as compared with control (,). The reduction was more remarkable after PF0 was combined with siRNA-V showing a significant (P < 0.05) decrease in colony formation to a 29.83 ± 8.11% as compared with control. For A549 cells, the combination resulted in a significant (P < 0.05) colony formation reduction when compared with PF0 treatment (). The relative colony-forming ability of H460 cells was reduced after PF0 treatment to 61.86 ± 20.48% and after siRNA-V treatment to 73.21 ± 1.97% as compared with controls (,). PF0 and siRNA-V combination showed a significant (P < 0.05) decrease in relative colony formation to 7.97 ± 9.20% as compared with control groups. The combination resulted in a significant (P < 0.01) decrease in the H460 colony formation when compared with single-agent treatment ().

Inhibition of cell migration

The effect of siRNA-V and PF0 in A549 and H460 cell migration was investigated using a monolayer in vitro wound healing assay. The wound widths were imaged and quantified at different time points (–). We observed a significant (P < 0.05) effect of 50 nmol/l siRNA-V on A549 cell migration as compared with scrambled siRNA-treated cells (). A549 and H460 cells also showed significant (P < 0.05) differences between PF0 (200 nmol/l for A549 and 2,000 nmol/l for H460) treated and nontreated cells (,). Furthermore, combination treatment significantly (P < 0.05) decreased A549 and H460 cell migration compared to control cells (,). A549 cells showed that after 16 hours of 50 nmol/l siRNA-V and 200 nmol/l PF0 treatment wound closure was 18.48 ± 1.60% comparing with 47.24 ± 2.09 of the control treated cells. After 40 hours treatment, A549 control cells migrated completely closing the wound while closure was reduced to 51.28 ± 0.96% for combination treated cells (,). After 16 hours combination treatment, H460 wound closure was inhibited to 14.65 ± 2.53% comparing with 26.86 ± 1.76 of the control. Wound closure was 25.01 ± 3.06% after 40 hours combination treatment compared to 51.42 ± 1.47% of the control, and 45.41 ± 6.54% after 64 hours compared to 74.02 ± 3.09% of the control (,). A549 and H460 cell migration was significantly (P < 0.05) reduced after combination treatment compared to single agents ().

Antiangiogenic effect analysis via inhibition of tube formation

Antiangiogenic effect of 50 nmol/l siRNA-V and 100 nmol/l PF0 was analyzed using tube formation assays of human umbilical vein endothelial cells (HUVEC) (–). siRNA-V 6 hours treatment showed a significant (P < 0.01) decrease in tube meshes area to 57.24 ± 3.22% () and segments length to 74.34 ± 1.32% () compared to control. PF0 treatment showed poor organization of HUVEC tube-like structures and a nonsignificant (P < 0.05) decrease in segments length (75.63 ± 13.71%) compared to control (). The combination treatment showed an enhanced effect decreasing progression of angiogenesis with a significant (P < 0.05) decrease in master segments length to 53.34 ± 3.03%, meshes area to 14.72 ± 17.82%, branches to 49.73 ± 4.54% and segments formed to 52.01 ± 6.73% compared with control (–). The combination effect was increased compared to PF0 showing a significant (P < 0.05) decrease of the master segment length from 91.02 ± 10.99 to 53.34 ± 3.03% () and meshes area from 82.64 ± 9.13 to 14.72 ± 17.82% (). The combination treatment also showed a significant (P < 0.05) decrease in total segment length (52.01 ± 6.73%) compared to siRNA-V (74.34 ± 1.32%) ().

Combination index analysis, normalized isobologram, and Fa-CI plot

To test whether the combination treatment with PF0 and siRNA-V induces synergistic cell death in NSCLC, we analyzed drug–drug interactions using normalized isobologram for nonconstant ratio combination design and combination index (CI) analysis. We treated A549 and H460 cells with a fixed dose of siRNA-V (0.1, 0.2, and 0.3 IC50 equivalents) and varying concentrations of PF0 (5 nmol/l to 5 µmol/l for A549, and 5 nmol/l to 100 µmol/l for H460 cells) to evaluate the combination cytotoxicity effect compared to the single agents after 72 hours treatment. We calculated the CI values at 0.5 fraction affected (Fa), indicating 50% of cells inhibited after drug exposure (), and generated normalized isobolograms (,) illustrating the observed synergy. In A549 cells PF0 and siRNA-V combination revealed a moderate synergism at 6 and 12 nmol/l siRNA-V concentration, and slight synergism at 18 nmol/l, with CI values between 0.62 and 0.88 (, ). In H460 cells, PF0 and siRNA-V showed moderate synergism at 4 nmol/l siRNA-V concentration, synergism at 6 nmol/l, and slight synergism at 12 nmol/l, with CI values between 0.51 and 0.81 (, ). Fa versus CI plots were generated for A549 and H460 cells (,) showing the observed synergism at Fa = 0.5, with CI values < 1. The combination cytotoxicity assays demonstrated that siRNA-V synergistically potentiates the anticancer activity of PF0 against A459 and H460 cells.

Discussion

Poor clinical outcome of the current therapies for lung cancer has prompted the need for new strategies. In the last years, newer biological agents specifically targeting molecular pathways in NSCLC have been FDA approved.[4,29] However, the efficacy and safety of these treatments are limited and the 5-year survival rate is still only 15%.[1] Our laboratory is investigating newer combination therapies for lung cancer treatment.[30,31,32] One of the promising treatments for lung cancer are antiangiogenic agents that inhibit tumor growth, proliferation, and metastasis.[7,12] Several signaling molecules governing angiogenesis are of interest, including VEGF and PI3K.[33] The overexpression of VEGF correlates with tumor growth, metastasis,[10,34] and poor prognosis.[11,35] Treatment with anti-VEGF antibodies has been proven to have a survival benefit to prevent NSCLC progression but their use have been limited due to poor efficacy and adverse effects.[13,14] Therapeutic antibodies, including Bevacizumab, have some functional limitations.[15] siRNA represents an emerging tool that is expected to overcome some of these limitations.[36] siRNA can potently inhibit gene expression with high sequence selectivity decreasing protein expression instead of only sequestering the protein.[16] Gene therapy utilizing siRNA to target VEGF has demonstrated promising in vitro and in vivo antitumor activity against variety of cancers,[18,19] and its effect may be further improved in combination with other anticancer agents.

The PI3K/mTOR signaling cascade is one of the most frequently disrupted in human cancer leading to tumor progression and development and FDA-approved drugs are available to inhibit this pathway.[21,37] Efforts are now placed in novel compounds capable of inhibiting both PI3K and mTOR, completely shutting down the PI3K/Akt/mTOR pathway.[38,39] These small molecules block both mTORC1-dependent phosphorylation of p70S6K and mTORC2-dependent phosphorylation of Akt overcoming feedback loops[40] (). PI3K/mTOR dual inhibitors have shown promising results in preclinical and clinical studies[26,27] and some of these agents are now entering phase 2 clinical trials.[22]

Our study explores a unique approach that targets three proteins and two important pathways for lung cancer survival and angiogenesis. In the current study we demonstrated for the first time that PF0 and siRNA-V combination leads to an enhanced anticancer activity against non–small-cell lung cancer cells. We first performed a dose–response study for PF0 and siRNA-V to evaluate the in vitro cytotoxicity effect in NSCLC cells. Previous results had shown that PF0 effectively inhibited tumor cell proliferation and showed antitumor efficacy in NSCLC xenograft models.[28] Our results showed a dose- and time-dependent cytotoxicity of PF0 in both cell lines, with lower IC50 values in A549 (151.45 to 209.77 nmol/l) than in H460 cells (936.7 to 1965.5 nmol/l) (, and Supplementary Figures S1a and S2a). Our study indicates that H460 cells require higher concentrations of PF0 to produce anticancer activity compared with A549 suggesting that H460 cells are less sensitive to PF0 treatment. Therefore, it is possible that H460 cells may be using multiple pathways for survival downstream of PI3K whereas A549 predominantly use the PI3K/mTOR pathway. The antitumor effect of PI3K/mTOR dual inhibitors vary significantly among cancer types, including cell lines,[28,41] which might relate to the complexity of the signaling networks. Previous studies reported highly different IC50 values to PF0 (0.1 to 2 µmol/l) in a wide panel of head and neck carcinoma cell lines suggesting that different factors contribute to the sensitivity of cancer cells to the drug.[42]

siRNA-V cytotoxicity analysis showed a dose-dependent antiproliferative effect against A549 and H460 cells (, and Supplementary Figures S1b and S2b). Our results showed the highest effect of siRNA-V at 72 hours with IC50 values of approximately 60 nmol/l for A549 and 40 nmol/l for H460 cells. siRNA-V treatment showed a higher inhibition of H460 survival as compared to A549 suggesting that H460 cells are more sensitive to siRNA-V treatment. After 72 hours of siRNA-V treatment, cell viability was decreased 27% in A549 cells () and 42% in H460 cells (). Previous reports had shown that VEGF stimulates tumor cell proliferation in NSCLC[43,44] and siRNA-V treatment resulted in tumor suppression in lung cancer animal models.[45]

The analysis of different combinations of PF0 and siRNA-V demonstrated an enhanced therapeutic cytotoxicity against NSCLC. The combination treatment showed a significantly (P < 0.05) higher effect reducing cell survival compared to single-agent therapy after 24, 48, and 72 hours (,). Therefore, this combination therapy could potentially be more effective and have fewer side effects than either agent alone, thereby providing a rationale for combining these drugs as a treatment for NSCLC.

To ensure optimal siRNA-V transfection efficiency and evaluate the effect on VEGF expression, we analyzed mRNA and protein expression levels. Our real-time PCR results showed a significant (P < 0.05) decrease in VEGF expression after 48 hours treatment with siRNA-V or siRNA-V + PF0 in A549 and H460 cells (,). Furthermore, an ELISA showed that secreted human VEGF levels were significantly (P < 0.05) reduced in A549 and H460 cells after 24–72 hours siRNA-V transfection and after PF0 treatment as compared with controls (,). As predicted, PI3K/mTOR inhibitors can decrease VEGF secretion and angiogenesis.[46] Activation of the PI3K/Akt pathway in tumor cells can increase VEGF secretion, both by HIF1α-dependent and -independent mechanisms.[23,24] The combination treatment resulted in almost undetectable levels of secreted VEGF. The 72-hour combination treatment resulted in a significant (P < 0.05) decrease in VEGF expression compared with single-agent treatments suggesting that the combination is highly effective. Our western blot data showed that PF0 inhibits PI3K and mTOR pathways in NSCLC cells after 24–72 hours treatment. PF0 single agent and in combination with siRNA-V significantly (P < 0.05) inhibited the phosphorylation of Akt (S473), a PI3K and mTORC2 downstream target, in A549 and H460 cells (, and Supplementary Figures S3 and S4). PF0 also significantly (P < 0.05) inhibited the phosphorylation of mTORC1 downstream target p70S6K (T389). The decrease in p-Akt/Akt and p-p70S6K/p70S6K ratio after PF0 treatment was significant (P < 0.05) at all the analyzed time points (24–72 hours) and as expected there was no change in total Akt and p70S6K levels (, and Supplementary Figures S3 and S4). Previous studies reported that short-term exposure to PF0 predominantly inhibited PI3K, whereas mTOR inhibition persisted for more than 48 hours.[28] Kinetic studies of PF0 and other PI3K/mTOR dual inhibitors had shown that levels of p-Akt (S473) and p-p70S6K (T389) decrease below control levels after 3 to 72 hours treatment.[47,48,49] This is in agreement with our results showing that phosphorylation of Akt and p70S6K was significantly (P < 0.05) inhibited after 24–72 hours.

To investigate the long-term effect of PF0 and siRNA-V on proliferation of lung cancer cells, an anchorage dependent colony formation assay was performed (,). Results showed a decreased number of colonies after siRNA-V or PF0 treatment. Previous publications reported long-term treatment with PF0 significantly suppressed the in vitro colony-forming abilities of cancer cells.[50,51] siRNA targeting VEGF was also previously reported to be effective to significantly decrease cancer cell colony formation.[52] The combination resulted in a significant (P < 0.05) colony formation decrease when compared with single-agent treatment (,). These results indicated that VEGF knockdown contributed to the suppressive effect of PF0 to a final enhanced effect of the combination therapy on lung cancer colony formation.

Cell migration is an integral part of metastasis that is required at virtually every step of the metastatic cascade.[53] Wound healing assays showed a significant (P > 0.05) effect of siRNA-V on A549 cell migration (,). Previous reports showed that VEGF signaling can promote migration and invasion of cancer cells and siRNA treatment reduces the migration rates.[52,54] PF0 treatment resulted in a significant (P < 0.05) decrease in cell migration in A549 and H460 cells, indicating that PI3K/mTOR pathway plays a central role in tumorigenesis by regulating migration of cancer cells, which is in agreement with previous reports.[37] Migration of both NSCLC cells after combination treatment was significantly (P > 0.05) decreased compared with single agents (after 40 hours treatment in A549 cells and 64 hours in H460 cells) (). Herein, we provided evidence of the siRNA-V and PF0 combination efficacy to inhibit lung cancer cell migration.

To study the possible mechanism involved in the enhanced cytotoxicity of siRNA-V and PF0, we evaluated the antiangiogenic effect in HUVEC tube formation. This assay is critical for evaluating drug antivasculogenic activity in vitro. Tube formation assays are considered representative of the later stages of angiogenesis and they are used as a model of in vivo capillary development.[55] The siRNA-V (50 nmol/l) treatment showed a significant (P < 0.01) decrease in total meshes area and segments length compared to control, which showed a rich meshwork of branching anastomosing capillary-like tubules with multicentric junctions (,,). It is already known that VEGF stimulates endothelial branching morphogenesis.[56] Previous reports showed that VEGF enhanced the ability of HUVEC to organize into tubular networks on Matrigel[57] and that tube formation is reduced by VEGF inhibitors or siRNA.[58,59] PF0 (100 nmol/l) treatment showed a nonsignificant (P < 0.05) decrease in tube segments length compared to control (,). Previous studies showed that higher concentrations of PI3K/mTOR dual inhibitors (1 µmol/l) significantly (P > 0.05) decreased HUVEC tube formation and tumor angiogenesis.[60] siRNA-V and PF0 combination showed the linear structures of the network were clearly disrupted compared to single-agent treatment () presenting a significant (P < 0.05) decrease of the total segment length or total meshes area (,). Compounds that are able to inhibit tube formation are very useful in cancer treatment, where tumors stimulate new blood vessel formation to grow beyond a relatively small size.[12]

CI analyses, isobologram, and Fa-CI plots were used to test whether the PF0 and siRNA-V combination treatment induces synergistic cell death in NSCLC (). We calculated the CIs using the Chou-Talalay method[61] by combining siRNA-V (0.1–0.3 IC50 equivalents) and PF0 (0.5–5 µmol/l for A549, and 0.5–100 µmol/l for H460 cells) treatments for 72 hours. The CIs ranged from 0.51 to 0.88 in A549 and H460 cell lines (), suggesting that the synergistic effect was induced in both adenocarcinoma and large cell carcinoma NSCLC cells. This synergistic effect was contributed due to targeting of multiple proteins: VEGF, PI3K, and mTOR.[40] Therefore, the use of this combination will be beneficial for lung cancer treatment.

Conclusion

Our results revealed the synergistic interaction between VEGF and PI3K/mTOR signaling in NSCLC cell proliferation, which suggest a promising therapeutic benefit of PF0 and siRNA-V combination for NSCLC treatment. The developed combination therapy also showed a decrease in long-term proliferation in A459 and H460 colony formation assays. In vitro wound healing assay showed a decrease in cell migration after siRNA-V and PF0 treatment. The combination therapy also showed an enhanced effect decreasing progression of angiogenesis and tube formation organization. Overall, our results suggest that the developed combination therapy, which preferentially targets tumor proliferation, migration, and angiogenesis, may be effective for the treatment of lung solid tumors and other types of cancers. However, further preclinical studies are warranted to evaluate the anticancer activity of PF0 and siRNA-V combination. Our ongoing combination studies are focused on the use of biocompatible targeted nanoparticles to effectively codeliver siRNA-V and PF0 to NSCLC while sparing normal tissues.

Materials and methods

Cell culture and treatment. We have selected human slow growing A549 adenocarcinoma cells and fast growing H460 large-cell carcinoma cells to evaluate the anticancer effect based on our previous studies.[30,31] The A549 and H460 NSCLC cells were obtained from American Type Culture Collection (Rockville, MD). A549 and H460 cells were grown in F12K and RPMI 1640 medium (Sigma Aldrich, St. Louis, MO) supplemented with 10% fetal bovine serum respectively. Pooled donor HUVEC cells (Lonza, Portsmouth, NH) were cultured in EGM BulletKit medium (Lonza) supplemented with 10% fetal bovine serum. All tissue culture media contained antibiotic antimycotic solution of penicillin (5,000 U/ml), streptomycin (0.1 mg/ml), and neomycin (0.2 mg/ml). All cells were grown at 37 ± 0.5 °C in a humidified 95% air and 5% CO2 atmosphere.

Gene knockdown was achieved using SMART pool ON-TARGETplus siRNA (Dharmacon, Boston, MA) targeting human VEGF. Target RNA sequences: 5′-GCAGAAUCAUCACGAAGUG-3′, 5′-CAACAAAUGUGAAUGCAGA-3′, 5′-GGAGUACCCUGAUGAGAUC-3′, 5′-GAUCAAACCUCACCAAGGC-3′. ON-TARGETplus Non-targeting Pool was used as negative control. Target RNA sequences: 5′-UGGUUUACAUGUCGACUAA-3′, 5′-UGGUUUACAUGUUGUGUGA-3′, 5′-UGGUUUACAUGUUUUCUGA-3′, 5′-UGGUUUACAUGUUUUCCUA-3′. Transfection was performed using DharmaFECT 1 transfection reagent (Dharmacon) in Opti-MEM I reduced serum medium (Thermo Fisher Scientific, Waltham, MA) without antibiotics and following the manufacturer's instructions. PF-04691502 was purchased from Abmole Bioscience (Houston, TX). PF0 was first dissolved in dimethyl sulfoxide to a final dilution of 10 mmol/l and then diluted in the appropriate aqueous culture medium.

For the cytotoxicity studies, A549 or H460 cells were plated in 96-well flat-bottomed tissue culture plates, at a density of 5 × 103 and 6 × 103 cells/well respectively and were allowed to incubate overnight. To evaluate the dose–response effect to the siRNA the cells were transfected with different concentrations of siRNA-V (5 to 150 nmol/l for A549, 5 to 200 nmol/l for H460) and DharmaFECT 1 reagent. To evaluate the dose–response effect to PF0 the cells were treated with various dilutions of PF0 (5 to 5,000 nmol/l in A549 and 5 nmol/l to 200 µmol/l in H460 cells). The IC50 values of siRNA-V and PF0 for A549 and H460 cell lines were calculated using the nonlinear log inhibitor versus normalized response curve fit function from GraphPad PRISM 3.0 software (San Diego, CA).

To analyze the effect of the combination, the cells were treated with various dilutions of PF0 (5 to 500 nmol/l for A549, 10 to 5,000 nmol/l for H460) in the presence of siRNA-V (50 nmol/l) as described previously.[30] The PF0 concentrations for the combination studies were selected based on the previous single-agent cytotoxicity results. 50 nmol/l siRNA-V concentration was selected for the transient transfection of the cells to obtain a high transfection efficiency. The plates were incubated for 24, 48, or 72 hours in a humidified atmosphere of 5% CO2 at 37 ± 0.5 °C. Cell viability in each treatment group was determined by resazurin (Sigma Aldrich) dye assay. Briefly, 15 µl of resazurin (0.1 mg/ml) in PBS (pH 7.4) were mixed with 85 µl of culture medium and the total 100 µl were added to each well. The plate was incubated for 3 hours at 37 ± 0.5 °C. The absorbance was measured at 550/ 580 nm using a microplate reader (BioTek, Winooski, VT).

Combination index analysis, normalized isobologram, and Fa-CI plot. A quantitative assessment of non-constant ratio combinations (fixed-dose combinations) of siRNA-V and PF0 was carried out in vitro to measure the drugs' interaction at 50% effect level. Excel software was used to plot a normalized isobologram and Chou-Talalay method[61] was used to determine the CI values, which express pharmacologic drug interactions, for each cell line combination treatment. Based on the single-agent IC50 determination, each NSCLC cell line was treated for 72 hours with a combination of siRNA-V and PF0 at different concentrations. We fixed the concentration of siRNA-V to 6, 12, and 18 nmol/l for A549, and 4, 8, and 12 nmol/l for H460 cells, corresponding to 0.1, 0.2, and 0.3 IC50 equivalents respectively. The concentration in IC50 equivalent for siRNA-V was calculated by dividing its actual concentration used in the combination treatment from its corresponding single-agent IC50 value. PF0 concentrations ranging from 5 nmol/l to 5 µmol/l, and 5 nmol/l to 100 µmol/l were used to treat A549 and H460 cells, respectively. The concentration in IC50 equivalent of PF0 was calculated by dividing the IC50 of PF0 in combination with siRNA-V from its corresponding single-agent IC50 value. The CI values at 50% effect level were calculated using the CI equation by adding the values of the IC50 equivalents of siRNA-V and PF0 in combination with each other. The line of additivity on the isobologram represents the 50% effect level of each drug. CI > 1.1 is antagonism; CI = 0.9–1.1 is additive; CI = 0.8–0.9 is slight synergism; CI = 0.6–0.8 is moderate synergism; CI = 0.4–0.6 is synergism; CI = 0.2–0.4 is strong synergism. The experiment was performed three independent times in triplicate (n = 9). Fa-CI plot was also generated for A549 and H460 cell lines for the combination of siRNA-V and PF0 using a non-constant drug ratio. CI values were calculated for each experiment at 0.5 Fa, indicating 50% of cells inhibited after drug exposure.

Quantitative real-time PCR. A549 and H460 cells (2 × 105/2.4 × 105) in six-well plates were allowed to incubate overnight and after that were treated with siRNA-V and PF0 combination and single agents. The PF0 and siRNA-V concentrations for the expression analysis and the following colony formation and wound healing assays were selected based on the previous single-agent cytotoxicity results and the observed IC50 values for each cell line. The selected PF0 concentration for A549 treatment was 200 nmol/l (IC50 values 151.45 to 209.77 nmol/l between 24 and 72 hours treatment), and the selected concentration for H460 treatment was 2,000 nmol/l (IC50 values 936.7 to 1,965.5 nmol/l). For the transient transfection of A549 and H460 cells 50 nmol/l siRNA-V concentration was selected (IC50 values 19.04 to 58.89 nmol/l in A549 and 26.19 to 40.79 nmol/l in H460 cells between 24 and 72 hours treatment) to ensure a high transfection efficiency. Total RNA was isolated 48 hours after treatment using High Pure RNA Isolation Kit (Roche, Indianapolis, IN). mRNA was reverse transcribed to cDNA using iScript Reverse Transcription Supermix (Bio-Rad, Hercules, CA). Gene levels were measured by iQ SYBR Green Supermix (Bio-Rad) and CFX384 Touch Real-Time PCR Detection System (Bio-Rad). Predesigned primers (KiCqStart SYBR Green Primers) for gene expression analysis of VEGF and GAPDH were purchased from Sigma Aldrich. Data are shown as percentage in VEGF mRNA expression normalized with GAPDH. Experiments were repeated three times in triplicate, and Bio-Rad CFX Manager Software v3.1 was used for gene pool quantification and statistical analysis. Calculations were performed using the double delta Ct method.[62]

ELISA. A549 and H460 cells (5 × 103/6 × 103 cells/well respectively) in 96-well plates were treated with siRNA-V + PF0 (50 and 200 nmol/l respectively for A549 and 50 and 2,000 nmol/l respectively for H460) and single agents for 24, 48, and 72 hours. Culture medium was collected and secreted VEGF was determined using human VEGF-A ELISA kit (Pierce biotechnology/Thermo Scientific, Rockford, lL) following manufacturer's instructions. Absorbance was measured on a microplate reader (BioTek) at 450 nm minus 550 nm. Inhibition of VEGF secretion was calculated using a standard curve.

Western blot analysis. A549 and H460 cells (2 × 105/2.4 × 105) in six-well plates were allowed to incubate overnight. After that cells were treated with siRNA-V (50 nmol/l) and PF0 (2,00 nmol/l for A549 and 2,000 nmol/l for H460) combination and single agents using the corresponding controls. After 24, 48, and 72 hours of incubation cells were lysed in radioimmunoprecipitation assay buffer (50 mmol/l Tris–HCl pH 7.40, 150 mmol/l NaCl, 2 mmol/l ethylenediaminetetraacetic acid, 1% NP-40, 0.1% sodium dodecyl sulfate) with freshly added protease and phosphatase inhibitors (Sigma Aldrich). Equivalent amounts of protein (20 µg) were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis, followed by electroblotting onto nitrocellulose membrane. Relative protein expression levels of phospho-Akt (S473), Akt, phospho-p70S6K (T389), and p70S6K were determined by Western blot analysis and GAPDH served as a protein loading control. Membranes were immunoblotted with the indicated primary antibodies (all used at a 1:1,000 dilution) at 4 °C overnight. Horseradish peroxidase-conjugated anti-rabbit (1:2,000 dilution) secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was then added at room temperature for 1 hour, and protein signals on membranes were visualized by Clarity Western ECL Chemiluminescent Substrate (Bio-Rad). Primary antibodies used were from Cell Signaling (Danvers, MA) except anti-GAPDH from Santa Cruz Biotechnology.

Anchorage-dependent colony formation assay. Monolayer colony formation assay was performed in the continuous presence of drug or siRNA. A549 and H460 cells were seeded in six-well culture plates (5 × 103/6 × 103) and anchorage-dependent clonogenicity assay was performed. Twenty-four hours after plating A549 and H460 cells were transfected with 50 nmol/l siRNA-V, using nontreated cells and scramble siRNA-treated cells as controls. After that the cells were treated with PF0 (200 nmol/l for A549 and 2,000 nmol/l for H460 cells), single agent and in combination with 50 nmol/l siRNA-V. Cells were maintained in culture for 10 days to let the viable cells propagate to sizable colonies for quantification. Since the drugs were not replenished in the medium, the colony formation is a stringent assessment of the drug effect after single exposure. The colonies were fixed with methanol-acetic acid at 3:1 ratio and stained with 1% crystal violet for 30 minutes at room temperature. The number of colonies formed in each well was photographed under a dissecting microscope Leica MZ10F (Leica Microsystems, Wetzlar, Germany) at 0.8× magnification and counted using the ColonyCounter plugin on ImageJ 1.49 software.

Cell migration wound healing assay. A549 and H460 cells (2 × 105/2.4 × 105) were seeded into six-well plates and after 24 hours were transfected with 50 nmol/l siRNA-V and the appropriate controls. After 6 hours and at approximately 80–90% confluency, the cells were treated with PF0 (200 nmol/l for A549 and 2,000 nmol/l for H460 cells) and appropriate controls. The wound of approximately 1 mm in width was scratched with a 200 μl pipette tip. Wound closure was monitored at different time intervals for the next 40 and 64 hours in A549 and H460 cells respectively and imaged with a Zeiss Axiovert 40 CFL inverted microscope (Carl Zeiss AG, Oberkochen, Germany) using 10× magnification. The open area (scratch) was quantified with TScratch software (ETH Zurich, Swiss).[63]

Endothelial cell tube formation. The antiangiogenic effects of siRNA-V and PF0 single agents and combination were visualized analyzing in vitro inhibition of tube formation of HUVEC as described previously.[31] HUVEC were cultured in the presence or absence of 50 nmol/l siRNA-V and 100 nmol/l PF0 alone or in combination for 6 hours on polymerized Matrigel at 37 °C. Standard Matrigel was allowed to polymerize in 96-well plates and HUVEC were seeded at a density of 1.5 × 104 per well in EGM medium. After 6 hours treatment, endothelial cells were fixed and tube formation was evaluated and photographed in a Zeiss Axiovert 40 CFL inverted microscope (Carl Zeiss AG) using 10× magnification. Quantification of angiogenesis progression was accomplished using the angiogenesis analyzer from ImageJ 1.49 by counting the total master segments length, total meshes area, capillary tube branch points, and total segments formed after 6 hours (end-point assay).

Statistical analysis. The experiments were repeated three times and the resulting data are presented after statistical processing. Data from multiple independent experiments are expressed as mean ± standard deviation. One-way analysis of variance followed by Tukey's Multiple Comparison Test was performed to determine the significance of differences among groups using GraphPad PRISM 3.0 software (San Diego). Differences were considered statistically significant at P < 0.05 in all experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure S1. Cytotoxicity effect of PF0 or siRNA-V in A549 cells after 24 and 48 hours treatment. Figure S2. Cytotoxicity effect of PF0 or siRNA-V in H460 cells after 24 and 48 hours treatment. Figure S3. Densitometric analysis of the Western blots presented in figure 4e (A549 cells) was carried out using Image J software. Figure S4. Densitometric analysis of the Western blots presented in figure 4f (H460 cells) was carried out using Image J software.

Table 1

Combination treatment of PF0 and siRNA-V in A549 and H460 cells for 72 hours