Osteosarcoma is a malignant bone tumor in children and adolescents characterized by intrinsic therapeutic resistance. The IGF2 is expressed at elevated levels in osteosarcoma after treatment with chemotherapy, prompting an examination of its functional contributions to resistance. We found that continuous exposure to IGF2 or insulin in the absence of serum created a dormant growth state in osteosarcoma cells that conferred resistance to various chemotherapeutic drugs in vitro. Mechanistic investigations revealed that this dormant state correlated with downregulation of downstream signaling by the IGF1 receptor, heightened cell survival, enhanced autophagy, and the presence of extracellular glutamine. Notably, inhibiting autophagy or depleting glutamine was sufficient to increase chemotherapeutic sensitivity in osteosarcoma xenografts in mice. Clinically, we confirmed that IGF expression levels were elevated in human osteosarcoma specimens from patients who received chemotherapy. Together, our results suggest that activation of IGF or insulin signaling preserves the survival of osteosarcoma cells under chemotherapeutic stress, providing a drug-resistant population that may engender minimal residual disease. Attenuating this survival mechanism may help overcome therapeutic resistance in osteosarcoma. Cancer Res; 74(22); 6531-41. Ó2014 AACR.

Osteosarcoma is the most common type of primary bone tumor, novel therapeutic agents for which are urgently needed. To identify such agents, we screened a panel of approved drugs with a mouse model of osteosarcoma. The screen identified simvastatin, which inhibited the proliferation and migration of osteosarcoma cells in vitro Simvastatin also induced apoptosis in osteosarcoma cells in a manner dependent on inhibition of the mevalonate biosynthetic pathway. It also disrupted the function of the small GTPase RhoA and induced activation of AMP-activated protein kinase (AMPK) and p38 MAPK, with AMPK functioning upstream of p38 MAPK. Inhibitors of AMPK or p38 MAPK attenuated the induction of apoptosis by simvastatin, whereas metformin enhanced this effect of simvastatin by further activation of AMPK. Although treatment with simvastatin alone did not inhibit osteosarcoma tumor growth in vivo, its combination with a fat-free diet induced a significant antitumor effect that was enhanced further by metformin administration. Our findings suggest that simvastatin induces apoptosis in osteosarcoma cells via activation of AMPK and p38 MAPK, and that, in combination with other approaches, it holds therapeutic potential for osteosarcoma. Mol Cancer Ther; 16(1); 182-92. ©2016 AACR.

Tumors comprise heterogeneous cell types including cancer stem cells (CSC), progenitor cells, and differentiated cells. Chemoresistance is a potential cause of relapse and a key characteristic of CSC, but the development of novel therapeutic approaches for targeting these cells has been limited. We previously established osteosarcoma-initiating (OSi) cells by introducing the gene for c-Myc into bone marrow stromal cells of Ink4a/Arf knockout mice. These OSi cells are composed of two distinct clones: highly tumorigenic cells (AX cells), similar to bipotent committed osteochondral progenitor cells, and tripotent cells of low tumorigenicity (AO cells), similar to mesenchymal stem cells. Here we show that depolymerization of the actin cytoskeleton induces terminal adipocyte differentiation and suppresses tumorigenesis in chemoresistant OSi cells. In contrast to AX cells, AO cells were highly resistant to conventional chemotherapeutic agents such as doxorubicin and were thus identified as chemoresistant cells. Inhibition of Rho-associated coiled-coil containing protein kinase (ROCK) elicited terminal adipocyte differentiation in chemoresistant AO cells through negative regulation of the transcriptional coactivator megakaryoblastic leukemia 1 associated with actin depolymerization. The clinically administered ROCK inhibitor fasudil significantly suppressed growth in vitro and tumorigenicity in vivo of chemoresistant AO cells as well as of OSi cells. Our findings thus suggest a new therapeutic strategy based on the induction of trans-terminal differentiation via modulation of actin cytoskeleton dynamics for therapyresistant osteosarcoma stem cells. Significance: These findings suggest that induction of trans-terminal differentiation through regulation of actin dynamics is a potential novel therapeutic approach for targeting chemoresistant stem-like tumor cells.

Osteosarcoma is the most common type of primary bone tumor, and novel therapeutic approaches for this disease are urgently required. To identify effective agents, we screened a panel of Food and Drug Administration (FDA)‐approved drugs in AXT cells, our newly established mouse osteosarcoma line, and identified calcitriol as a candidate compound with therapeutic efficacy for this disease. Calcitriol inhibited cell proliferation in AXT cells by blocking cell cycle progression. From a mechanistic standpoint, calcitriol induced endoplasmic reticulum (ER) stress, which was potentially responsible for downregulation of cyclin D1, activation of p38 MAPK, and intracellular production of reactive oxygen species (ROS). Knockdown of Atf4 or Ddit3 restored cell viability after calcitriol treatment, indicating that the ER stress response was indeed responsible for the anti‐proliferative effect in AXT cells. Notably, the ER stress response was induced to a lesser extent in human osteosarcoma than in AXT cells, consistent with the weaker suppressive effect on cell growth in the human cells. Thus, the magnitude of ER stress induced by calcitriol might be an index of its anti‐osteosarcoma effect. Although mice treated with calcitriol exhibited weight loss and elevated serum calcium levels, a single dose was sufficient to decrease osteosarcoma tumor size in vivo. Our findings suggest that calcitriol holds therapeutic potential for treatment of osteosarcoma, assuming that techniques to diminish its toxicity could be established. In addition, our results show that calcitriol could still be safely administered to osteosarcoma patients for its original purposes, including treatment of osteoporosis.

Osteosarcoma is the most common high‐grade malignancy of bone, and novel therapeutic options are urgently required. Previously, we developed mouse osteosarcoma AXT cells that can proliferate both under adherent and nonadherent conditions. Based on metabolite levels, nonadherent conditions were more similar to the in vivo environment than adherent conditions. A drug screen identified MEK inhibitors, including trametinib, that preferentially decreased the viability of nonadherent AXT cells. Trametinib inhibited the cell cycle and induced apoptosis in AXT cells, and both effects were stronger under nonadherent conditions. Trametinib also potently decreased viability in U2OS cells, but its effects were less prominent in MG63 or Saos2 cells. By contrast, MG63 and Saos2 cells were more sensitive to PI3K inhibition than AXT or U2OS cells. Notably, the combination of MAPK/ERK kinase (MEK) and PI3K inhibition synergistically decreased viability in U2OS and AXT cells, but this effect was less pronounced in MG63 or Saos2 cells. Therefore, signal dependence for cell survival and crosstalk between MEK–ERK and PI3K–AKT pathways in osteosarcoma are cell context‐dependent. The activation status of other kinases including CREB varied in a cell context‐dependent manner, which might determine the response to MEK inhibition. A single dose of trametinib was sufficient to decrease the size of the primary tumor and circulating tumor cells in vivo. Moreover, combined administration of trametinib and rapamycin or conventional anticancer drugs further increased antitumor activity. Thus, given optimal biomarkers for predicting its effects, trametinib holds therapeutic potential for the treatment of osteosarcoma.

<p>Supplementary Figure Legends.</p>

<p>(A) Immunoblot analysis of AMPK and AKT in AXT cells treated with simvastatin for 12 h. The phosphorylation level of AMPK at Ser485 was not affected by simvastatin, while treatment of 1 μg/ml of insulin for 1 h (Ins) slightly enhanced phosphorylation level of AMPK at Ser485/Ser491. (B) Phase-contrast microscopy of AXT cells exposed to the indicated reagents for 20 h. (C, D) Immunoblot analysis of caspase-3 cleavage and either AMPK (B) or p38 MAPK (C) phosphorylation in AXT cells exposed to the indicated concentrations of 5-iodotubercidin or BIRB796, respectively, in the additional presence of 5 μM simvastatin for 20 h. (E) Immunoblot analysis of AMPK, p38 MAPK, and ACC phosphorylation in AXT cells treated with AICAR for 22 h. (F) Viability of AXT cells exposed to compound C in the additional presence of AICAR for 2 days. (G) Phase-contrast microscopy of AXT cells exposed to C3 transferase for 12 h.</p>

<p>(A) Phase-contrast microscopy of AXT cells treated with 10 μM simvastatin for 24 h in the absence or presence of mevalonate, GGPP, or FPP. (B) Representative migratory AXT cells stained with crystal violet after exposure to simvastatin for 12 h. (C, D) Flow cytometric analysis of DNA content in AXT cells treated with 5 μM simvastatin for 24 h in the absence or presence of mevalonate, GGPP, or FPP (C). The size of the sub-G1 fraction of cells was also determined (D).</p>