Investigation of the effect of HSP90 inhibitor treatment on tumour cell biology and the bone microenvironment

Abstract

HSP90 is a molecular chaperone that is ubiquitously expressed in cells, participating in the stabilization and activation of over 200 proteins, many of which are essential for cell signaling and adaptive stress responses. To accomplish this, HSP90 forms an ATP-dependent multi-protein dynamic complex termed the HSP90 chaperone machine. Cancer cells are dependent on this machinery to protect numerous mutated and overexpressed oncoproteins from misfolding and degradation. As such, pharmacological inhibitors of HSP90, such as 17- allylamino-17-demethoxy geldanamycin (17-AAG), act as potent anticancer agents in preclinical tumour models. However, their success within the clinical setting has been less pronounced. As with other anticancer agents, intrinsic and acquired drug resistance may significantly limit the utility and efficacy of HSP90 inhibitors. Therefore, the prediction and reversal of HSP90 inhibitor resistance may be a potential way of significantly improving the therapeutic efficacy of these anticancer agents. In addition to resistance, prolonged drug treatment of cancer cells can also result in phenotypic changes that promote their ability to metastasise, whether this occurs with respect to HSP90 inhibitors has yet to be determined. Therefore, one aspect of this study was to investigate the underlying molecular changes that contribute to the acquired resistance to HSP90 inhibitors and the associated phenotypical alterations that occurred. To achieve this we generated MDA-MB-435 and MDA-MB-231 human cancer cell lines that were resistant towards 17-AAG by gradual dose escalation. When cultured in the absence of drug pressure, cells maintained their respective levels of 17- AAG resistance (7–240x), and were cross-resistant towards other benzoquinone ansamycin that are structurally related to 17-AAG. Additionally, the resistant cell lines were also crossresistant to structurally distinct HSP90 inhibitors, such as radicicol and other radicicol-based compounds. Altered expression of NQO1, histone deacetylase 6 (HDAC6) and histone deacetylase 1 (HDAC1) were identified in the resistant cell lines and were considered potential mechanisms of resistance. Consistent with increased HDAC expression, histone 3 acetylation was increased in the resistant cells. Moreover, HDAC inhibition with pharmacological inhibitors significantly re-sensitized resistant cells towards 17-AAG. Similarly, HDAC inhibition dramatically re-sensitized the resistant cells towards other structurally-distinct HSP90 inhibitors. In conclusion, prolonged 17-AAG treatment was found to result in cancer cell resistance towards a spectrum of HSP90 inhibitors to which HDAC upregulation makes a significant contribution. To investigate the effects of chronic exposure to HSP90 inhibitors on cancer cell biology, we examined whether the MDA-MB-435 and MDA-MB-231 resistant cell lines displayed altered growth, survival and metastatic properties. Despite both cell lines being significantly resistant to HSP90 inhibitors, each displayed reduced growth both in vitro and in vivo. Consistent with this, gene expression profiling revealed that genes involved in cell cycle progression were downregulated and markers of cancer stem cells (CSC) associated with slower growth were upregulated in the resistant cells. Despite the decrease in growth, the migratory capacity of the resistant cells was found to be significantly greater than that of the parental cells. This was supported by the increased expression of epithelial to mesenchymal transition (EMT) markers, commonly associated with an increased migratory phenotype. Despite this, the metastatic potential of the resistant cells was actually found to be markedly decreased in vivo, developing less metastatic tumour burden at both visceral and skeletal sites. Interestingly, within the bone, although the tumour burden was significantly lower in mice bearing the resistant cells, x-ray and bone morphometric analysis demonstrated that the extent of osteolytic lesions were similar in both the parental and resistant cell lines, demonstrating an enhanced pro-osteolytic phenotype in the resistant cells. Taken together, these findings show for the first time that acquired resistance towards HSP90 inhibitors results in phenotypical changes at the cellular level that negatively impact upon tumour growth and metastasis, yet paradoxically enhances a pro-osteolytic phenotype, potentially through multiple molecular changes as observed in the resistant cells by gene expression profiling. In relation to bone, 17-AAG has been previously shown to increase bone loss in mouse models through the direct stimulation of osteoclast formation. Heat shock factor 1 (HSF1), the master transcriptional regulator of heat shock response (HSR) associates with HSP90 under normal conditions, maintaining HSF1 in an inactive monomeric state. However, upon 17-AAG binding to the N-terminal ATPase domain of HSP90, or upon cellular stress, HSF1 dissociates from the HSP90 complex and binds to heat shock element (HSE) sites within the promoters of target genes. Hence, the activation of HSF1 by 17-AAG may be involved in the activation of osteoclast formation that subsequently leads to bone loss. Therefore an additional aim of this study was to definitively determine the role of HSF1 in HSP90 inhibitor-induced osteoclast formation. It was determined that HSP90 inhibitors that induced a heat shock response also enhanced osteoclast formation while HSP90 inhibitors that did not, had no impact upon osteoclast formation. Pharmacological inhibition or shRNAmir knockdown of Hsf1 in RAW264.7 cells, as well as the use of Hsf1 null bone marrow cells, demonstrated that enhanced osteoclast formation by 17-AAG was HSF1-dependent. Moreover, ectopic over-expression of Hsf1 enhanced the effect of 17-AAG upon osteoclast formation. Consistent with these findings, protein levels of the essential osteoclast transcription factor, microphthalmia-associated transcription factor (Mitf) were increased by 17-AAG in a HSF1-dependent manner. In addition to HSP90 inhibitors, we also identified ! $! thatgrowth, survival and metastatic properties. Despite both cell lines being significantly resistant to HSP90 inhibitors, each displayed reduced growth both in vitro and in vivo. Consistent with this, gene expression profiling revealed that genes involved in cell cycle progression were downregulated and markers of cancer stem cells (CSC) associated with slower growth were upregulated in the resistant cells. Despite the decrease in growth, the migratory capacity of the resistant cells was found to be significantly greater than that of the parental cells. This was supported by the increased expression of epithelial to mesenchymal transition (EMT) markers, commonly associated with an increased migratory phenotype. Despite this, the metastatic potential of the resistant cells was actually found to be markedly decreased in vivo, developing less metastatic tumour burden at both visceral and skeletal sites. Interestingly, within the bone, although the tumour burden was significantly lower in mice bearing the resistant cells, x-ray and bone morphometric analysis demonstrated that the extent of osteolytic lesions were similar in both the parental and resistant cell lines, demonstrating an enhanced pro-osteolytic phenotype in the resistant cells. Taken together, these findings show for the first time that acquired resistance towards HSP90 inhibitors results in phenotypical changes at the cellular level that negatively impact upon tumour growth and metastasis, yet paradoxically enhances a pro-osteolytic phenotype, potentially through multiple molecular changes as observed in the resistant cells by gene expression profiling. In relation to bone, 17-AAG has been previously shown to increase bone loss in mouse models through the direct stimulation of osteoclast formation. Heat shock factor 1 (HSF1), the master transcriptional regulator of heat shock response (HSR) associates with HSP90 under normal conditions, maintaining HSF1 in an inactive monomeric state. However, upon 17-AAG binding to the N-terminal ATPase domain of HSP90, or upon cellular stress, HSF1 dissociates from the HSP90 complex and binds to heat shock element (HSE) sites within the promoters of target genes. Hence, the activation of HSF1 by 17-AAG may be involved in the activation of osteoclast formation that subsequently leads to bone loss. Therefore an additional aim of this study was to definitively determine the role of HSF1 in HSP90 inhibitor-induced osteoclast formation. It was determined that HSP90 inhibitors that induced a heat shock response also enhanced osteoclast formation while HSP90 inhibitors that did not, had no impact upon osteoclast formation. Pharmacological inhibition or shRNAmir knockdown of Hsf1 in RAW264.7 cells, as well as the use of Hsf1 null bone marrow cells, demonstrated that enhanced osteoclast formation by 17-AAG was HSF1-dependent. Moreover, ectopic over-expression of Hsf1 enhanced the effect of 17-AAG upon osteoclast formation. Consistent with these findings, protein levels of the essential osteoclast transcription factor, microphthalmia-associated transcription factor (Mitf) were increased by 17-AAG in a HSF1-dependent manner. In addition to HSP90 inhibitors, we also identified that other heat shock response inducing agents, such as alcohol, doxorubicin and methotrexate, can also directly increase osteoclast formation in a stress-dependent manner. These results indicate that cellular stress can enhance osteoclast differentiation via HSF1- dependent mechanisms and may significantly contribute to pathological and therapeutic related bone loss. Therefore, this study demonstrates that acquired resistance to a broad spectrum of HSP90 inhibitors in cancer cells is partially mediated by a HDAC-dependent mechanism and is associated with decreased tumour growth and metastasis. (...)