Molecular Characterization of Human CML Cells with Resistance to the Heat Shock Protein (hsp) 90 Inhibitor 17-Allylamino-Demethoxy Geldanamycin (17-AAG).

Similar to AKT, c-Raf, and Src proteins, Bcr-Abl is a bonafide client protein for the chaperone, the heat shock protein (hsp) 90. Newly synthesized or stress denatured or mutant client proteins require interaction with hsp90 to maintain a mature, stable and functional conformation. Treatment with hsp90 inhibitor 17-AAG (Kosan Biosciences, Hayward, CA) disrupts the chaperone association of Bcr-Abl with hsp90, directing Bcr-Abl to polyubiquitylation and proteasomal degradation. Recently 17-AAG has been shown to have greater affinity for binding to hsp90 from cancer versus normal host cells. 17-AAG mediated down regulation of Bcr-Abl, as well as abrogation of the levels and activity of multiple pro-growth and pro-survival signaling molecules downstream of Bcr-Abl, is responsible for 17-AAG mediated differentiation and apoptosis of imatinib mesylate (IM)-sensitive and IM-refractory CML-BC cells. Indeed, 17-AAG was more active against IM-refractory cells expressing mutant Bcr-Abl. In the present studies we cultured human CML-BC K562 cells in the continuous presence of increasing levels of 17-AAG and isolated K562/AGR cells that are capable of growth in the continuous presence of 3.0 μM 17-AAG. Treatment with 17-AAG (5 μM for 24 hours) induced hsp70 levels in K562 cells. In contrast, treatment of K562/AGR cells with 5 μM of 17-AAG, failed to induce hsp70, and instead downregulated hsp70, hsp27 and p-HSF1 levels. While treatment with 17-AAG did not alter hsp90, it increased the levels of the co-chaperones hsp40, HOP/p60 and cdc37/p50 in K562/AGR cells. Additionally, in contrast to K562 cells, in K562/AGR cells, treatment with 17-AAG (5 μM for 24 hours) failed to attenuate Bcr-Abl, p-AKT, p-STAT5, AKT and c-Raf levels. Although resistant to 17-AAG, K562/AGR cells were as sensitive as K562 cells to apoptosis induced by the hydroxamic acid (HA) analogue histone deacetylase inhibitor (HDI) LAQ824 (100 or 250 nM), which has been previously reported to induce acetylation and inhibit the chaperone function of hsp90. Consistent with this, LAQ824 treatment downregulated Bcr-Abl and the other client proteins AKT and c-Raf in K562 as well as in K562/AGR cells. Importantly, K562/AGR cells were as sensitive as K562 cells to apoptosis induced by IM (0.25 to 2.0 μM for 48 hours) and a novel, aminopyrimidine, Bcr-Abl tyrosine kinase inhibitor AMN107 (20 to 100 nM for 48 hours) (Novartis Pharmaceuticals, Basel, Switzerland). Additionally, K562/AGR cells were collaterally sensitive to high-dose Ara-C (10 to 100 μM). These studies suggest that resistance of Bcr-Abl expressing cells to the hsp90 inhibitor, 17-AAG may be due to its inability to attenuate the levels and thereby activity of pro-growth and pro-survival client proteins, including Bcr-Abl, AKT and c-Raf. These studies also demonstrate that resistance to 17-AAG does not confer cross-resistance to Bcr-Abl TK inhibitors, e.g., IM and AMN107, as well as to the HA-HDI LAQ824.