Hypoxia-activated prodrug enhances therapeutic effect of sunitinib in melanoma

Abstract

Angiogenesis is a critical step during tumor progression. Anti-angiogenic therapy has only provided modest benefits in delaying tumor progression despite its early promise in cancer treatment. It has been postulated that anti-angiogenic therapy may promote the emergence of a more aggressive cancer cell phenotype by generating increased tumor hypoxia-a well-recognized promoter of tumor progression. TH-302 is a 2-nitroimidazole triggered hypoxia-activated prodrug (HAP) which has been shown to selectively target the hypoxic tumor compartment and reduce tumor volume. Here, we show that melanoma cells grown under hypoxic conditions exhibit increased resistance to a wide variety of therapeutic agents in vitro and generate larger and more aggressive tumors in vivo than melanoma cells grown under normoxic conditions. However, hypoxic melanoma cells exhibit a pronounced sensitivity to TH-302 which is further enhanced by the addition of sunitinib. Short term sunitinib treatment fails to prolong the survival of melanoma bearing genetically engineered mice (Tyr::CreER; BRaf^CA;Pten^lox/lox ) but increases tumor hypoxia. Long term TH-302 alone modestly prolongs the overall survival of melanoma bearing mice. Combination therapy of TH-302 with sunitinib further increases the survival of treated mice. These studies provide a translational rationale for combining hypoxic tumor cell targeted therapies with anti-angiogenics for treatment of melanoma.

Keywords: TH302; hypoxia; melanoma; sunitinib; treatment.

Figure 1. Hypoxia potentiates melanoma growth and metastasis in vivo

Human WM115F melanoma cells were exposed to 1% O₂ for 72 hours. 100 viable cells were injected subcutaneously into the flanks of NOD/SCID mice (n = 4). WM115F cells grown under normoxic conditions were used as control. The mice were followed for 45 days and then sacrificed. All mice that received hypoxia treated melanoma cell developed tumors, while 3 of 4 mice that received 100 normoxic cells developed tumors. (a) Left side: representative xenograft formed by 100 melanoma cells cultured under normoxic conditions. Right side: representative xenograft formed by 100 melanoma cells cultured under hypoxia. (b) 10⁶ of hypoxia-treated melanoma cells were injected subcutaneously, the same number of melanoma cells grown under normoxia were used as controls. H&E staining of a representative xenograft formed by normoxia-treated melanoma cells reveals extensive necrosis in the tumor (Arrow points to areas with tumor necrosis) (25x). (c) H&E staining of a xenograft tumor formed by hypoxia-treated melanoma cells reveals significantly less tumor necrosis. (d-f) Histology of spontaneous metastases developed by hypoxic melanoma cells: lung metastasis (d, 400x); lymphovascular invasion (e, 400x); pancreas metastasis (f, 100x). (g) Tumor volume. Hypoxia-treated melanoma cells form significantly larger tumors than melanoma cells grown under normoxic conditions. ^* indicates p<0.01.

Figure 2. Hypoxia increases melanoma cell resistance to targeted therapeutic agents

(a) WST-1 cell proliferation assays were used to assess cell proliferation in response to chemotherapeutic agents, including PLX-4720 (B-Raf inhibitor), DAPT (Notch inhibitor), Genistein (protein-tyrosine kinase inhibitor), and Ly294002 (PI3-K inhibitor), PLX-4720 (B-Raf inhibitor), DAPT (Notch inhibitor), Genistein (protein-tyrosine kinase inhibitor), Ly294002 (PI3-Kinase inhibitor), MG-132 (Proteasome inhibitor), Rapamycin (mTOR inhibitors), SB2012190 (MAP kinase inhibitor) and U0126 (MEK inhibitor). Hypoxia (red lines) or normoxia-treated (blue lines) melanoma cells were incubated with these agents for 24 hours (n = 3 replicate experiments for each cell line with each drug; ^* indicates p< 0.05 comparing hypoxia-treated to normoxia-treated cells). (b) Relative mRNA expression of VEGFA, Oct4, and Snail in the indicated cell lines grown under normoxic conditions (blue boxes) versus hypoxic conditions (red boxes).

Figure 3. TH-302 increases the inhibitory effect of sunitinib in vitro

(a) Left panel: Sunitinib alone has little effect on the growth of 1205Lu melanoma cells at concentrations less than 500 nM. Cell survival was assessed by WST-1 assay. Right panel: Effects of increasing concentrations of TH-302 together with variable concentrations of sunitinib (0, 0.5μM, 1.5μM) on 1205Lu melanoma cells grown in normoxic and hypoxic conditions. Cell survival was assessed by WST-1 assay. (n = 3 replicate experiments; ^*indicates p<0.05 compared with control). (b) Effects of sunitinib, TH-302 or sunitinib plus TH-302 on 1205Lu cells grown in 3D culture. 1205Lu cells were cultured as spheroids in collagen gel and treated with the indicated concentrations of TH-302, sunitinib or TH-302 plus sunitinib for 72 hours under normoxia or hypoxia. Spheroids were then stained with calcein-AM and imaged with a confocal microscope. There was complete inhibition of melanoma growth with combination of TH-302 and sunitinib under hypoxic conditions.

Figure 4. Short-term sunitinib therapy increases tumor hypoxia and HIF regulated proteins in vivo

Using Tyr::CreER; Braf^CA/+; Pten^lox/lox mice, melanomas were induced by applying 4-hydroxytamoxifen (4-HT) on skin for 3 days. Tumors were allowed to grow for 4 weeks until they were palpable. Melanoma bearing mice were injected with Hypoxyprobe one hour before sacrificing, and melanoma tissue was stained with an anti-pimonidazole antibody, which identified areas of hypoxia. (a, b) Many melanoma cells in the dermis were variably positive for Hypoxyprobe staining. (c, d) Tumor cells were positive for the well-known marker for hypoxia, carbonic anhydrase IX (CAIX). (e) Melanoma was induced in Tyr::CreER; Braf^CA/+; Pten^lox/lox mice. Five mice were used in each group. Tumors were allowed to grow for 3 weeks until palpable tumors were evident. The mice were then treated for one week (short term) with vehicle (DMSO) or sunitinib (See Supplementary Figure 1 for scheme). Expression of HIF-1α, HIF-2α, VEGF-A, Snail1, Oct4 proteins was determined by western blot analysis in melanomas after treatment (n = 3 replicate experiments). β-actin was used as a loading control.

Figure 5. Effects of short term sunitinib and TH-302 therapy in vivo

(a) Schemes of melanoma induction and treatments. Melanoma was induced by applying 4-HT on Tyr::CreER; Braf^CA/+; Pten^lox/lox mice for 3 days. Tumors were allowed to grow for 3 weeks until palpable tumors were evident. The mice were then treated for one week (short term) with vehicle control, TH-302, sunitinib, or TH-302 plus sunitinib as indicated. Twenty mice were used in each group. (b) Effect of short term treatments on mouse survival. Mice were euthanized according to the standard score of body condition. Kaplan-Meier survival analysis demonstrated that short-term sunitinib alone did not affect the lifespan of melanoma bearing mice compared to that of control. In contrast, short term TH-302 modestly increased mouse survival, while addition of sunitinib added little efficacy. (c) Effect of sunitinib and TH-302 on tumor volume. A subset of treated mice was euthanized 40 days after melanoma induction and tumor volume was measured. ^* indicates p<0.05. (d) Quantitative RT-PCR assay for VEGF-A, CD31, Oct4 and Snail1 mRNA expression in melanoma bearing mice treated with short term sunitinib, TH-302 or TH-302 plus sunitinib (n = 3 replicate experiments). Short-term treatment with sunitinib resulted in increased VEGF-A, CD31, Oct4, and Snail1 mRNA expression in melanoma, while treatment with TH-302 or Th-302 plus sunitinib prevented their increased expression. β-actin was used as an internal control. (^*p< 0.01 compared with control).

Figure 6. Effects of long term TH-302 and sunitinib therapy in vivo

(a) Scheme of experiments. Melanoma was induced by applying 4-HT on Tyr::CreER; Braf^CA/+; Pten^lox/lox mice. Twelve mice were used in each group. Tumors were allowed to grow for 3 weeks until palpable tumors were evident. The mice were then treated with vehicle (DMSO), TH-302, sunitinib, or TH-302 plus sunitinib for the indicated time periods. (b) Kaplan-Meier survival analysis demonstrated that TH-302 plus sunitinib (long-term) therapy significantly extended the lifespan of melanoma bearing mice compared with that of long-term TH-302 or sunitinib treatment. Mice were euthanized according to the standard body condition score. (c) Effect of long term therapy on tumor volume. A subset of treated mice were euthanized 40 days after melanoma induction and tumor volume was measured. ^* indicates p<0.05. (d) Quantitative RT-PCR assay of VEGF-A, CD31, Snail1 and Oct4 mRNA expression in sunitinib, TH-302, or TH-302 plus sunitinib long-term treatment mouse groups (n = 3 replicate experiments; ^*indicates p< 0.01 compared with vehicle control). β-actin is used as an internal control. (e) Expression of HIF-1α, HIF-2α, VEGF-A, Snail1 and Oct4 proteins was determined by western blot analysis in melanoma bearing mice treated with long term sunitinib, TH-302 or TH-302 plus sunitinib. β-actin was used as a loading control.

Figure 7. Effects of long term TH-302 and sunitinib prevention regimen in vivo

(a) Scheme of experiments. Melanoma was induced by applying 4-HT for 3 days on Tyr::CreER; Braf^CA/+; Pten^lox/lox mice. Twelve mice were used in each group. On the second day of induction, the mice were treated with vehicle (DMSO), sunitinib, TH-302 or sunitinib plus TH-302 for the time periods indicated. (b) Kaplan-Meier survival analysis demonstrated that prevention scheme with long term TH-302 plus sunitinib treatment significantly extended the lifespan of melanoma-bearing mice compared to TH-302 or sunitinib treatment. Mice were euthanized according to the standard body condition score. (c) Effect of therapies on tumor volume. A subset of treated mice were euthanized 40 days after melanoma induction and tumor volume was measured. ^* indicates p<0.05. (d) Quantitative RT-PCR assay of VEGF-A, CD31, Snail1 and Oct4 mRNA expression in tumors treated with vehicle control, sunitinib, TH-302 or TH-302 plus sunitinib (n=3 replicate experiments; ^*indicates p< 0.01 compared with vehicle control). β-actin was used as an internal control. (e) Expression of HIF-1α, HIF-2α, VEGF-A, Snail1 and Oct4 proteins was determined by western blot analysis in melanoma bearing mice treated with long term sunitinib, TH-302 or TH-302 plus sunitinib. β-actin was used as a loading control.