Different Sensitivity of Normal and Malignant Cells to HSP90 Inhibitors.

Despite of great progress in cancer treatment over the past decades, there is still strong need for new, more efficient and specific therapeutics. Drugs that blocks activity of heat shock proteins, particularly heat shock protein 90 (HSP90), are the newcomers with potential to fulfill this expectation. HSP90 is a molecular chaperone crucial for correct protein folding, assembly and transportation across cellular compartments. Recent data show also essential role of HSP90 in facilitating malignant transformation. HSP90 serves as a biochemical buffer for the genetic instabilities characteristic of most cancers and enhances cancer cells survival under stress conditions. Thus, HSP90 inhibitors, like Geldanamycin (GA), seem to be attractive anticancer agents. One of the GA derivatives, 17-allylaminogeldanamycin (17AAG), has already entered clinical trials with promising preliminary results.

Although HSP90 is highly expressed in most cells, geldanamycins seem to selectively kill cancer cells. However, the direct comparison of GA effect on normal and cancer cells has not been done yet. This prompted us to investigate the influence of GA and its analogs (17AAG, 17DMAG, 17AEP and 17DMAP) on normal, bone marrow populating cells and cancer cells. We choose normal bone marrow cells because one of the most life-threatening adverse effects of conventional chemotherapy and radiotherapy is damage of bone marrow cells and its environment. Thus, we asked if GA and its derivatives affect CD34⁺, mesenchymal and endothelial cells. We used in this study multiple myeloma (MM) cell lines and rhabdomyosarcoma cell lines (RMS).

We found that GA and its analogs had negative influence on both MM and RMS cell lines. GA exhibited strong anticancer effect at concentrations ranging from 1 nM to 1000 nM, unfortunately causing also strong toxicity toward all normal cells tested. 17AAG, which is already in clinical trials, had surprisingly low inhibitory influence on proliferation of cancer cells (MM and RMS) at 100 nM concentration and lower, with strong inhibitory effect seen only at 1000 nM. However, at the same time 17AAG had little toxicity toward normal cells. New GA derivatives, 17DMAG, 17AEP and 17DMAP inhibited proliferation of cancer cells to almost the same extend as GA, and they exerted only mild negative effects on clonogenicity of CD34⁺, as measured by BFU-E and CFU-GM colony formation assay and proliferation of mesenchymal cells. Endothelial cells growth, important for tumor progression, was restrained by half with 100 nM of GA and its analogs except 17AAG. We also noticed that HSP90 inhibitors were able to block expression of genes responsible for angiogenesis (VEGF, HIF-1 a and IL-8). Finally, it is also worth stressing that 17DMAG, 17AEP and 17DMAP are water-soluble while GA and 17AAG must be dissolved in DMSO, agent known with neuronal side effects observed during bone marrow transplantation procedures.

Different level of toxicity between normal and malignant cells could be due to different level of HSP90 expression in these cells. We performed western blot analysis of HSP90 expression in normal and cancer cells. We found lower expression of this protein in CD34⁺, endothelial and mesenchymal cells in comparison to MM and RMS cell lines. This could be one of the reasons for different sensitivity of normal and tumor cells to geldanamycins.

Based on our results we postulate that GA derivatives, particularly 17AEP, have a potential to become specific, powerful drugs against various tumors with little unwanted side effects toward normal cells.