Glutathione and Glutathione Utilizing Enzymes Are Key Determinants in the Sensitivity of Myeloma Cells to Arsenic Trioxide as Well as Its Mechanism of Action.

Arsenic trioxide (ATO, Trisenox) is currently being tested in clinical trials as a single agent or in combination with other agents that have activity in multiple myeloma (MM). We and others have demonstrated that glutathione (GSH) levels can influence the ability of ATO to induce cell death in MM cell lines/patient samples and based on these data have initiated a trial to test the safety and efficacy of the combination of ATO and ascorbic acid for the treatment of refractory/relapsed MM. Therefore we performed expression profiling on 4 MM cell lines treated with ATO over a 48 timecourse. Affymetrix Hu133 2.0 plus arrays were hybridized and up to 7035 out of 54,675 probe sets displayed a change and up to 1546 probes sets displayed 2 fold or greater changes compared to untreated cells at 6 hrs. By only looking at genes that increased in all four cell lines we restricted our search to less than 365 probes sets at any given time point. Interestingly the cells appear to have initiated several pathways that are consistent with an attempt to enhance GSH synthesis. Upregulation of transporters of cysteine (xCT) and glycine (Glyt1) as well as enzymes that convert methionine to cysteine (cystathionase) and serine to glycine (serine hydroxymethyl transferase-1) was observed. Moreover the rate-limiting step of the glutathione salvage pathway gamma-glutamyltransferase is also upregulated. Together this suggests an increase in the building blocks for GSH that can be used for de novo synthesis. The rate limiting step for this reaction is performed by gamma-glutamate cysteine ligase which both the catalytic and modifier subunits are upregulated. GSH can also be regenerated from GSSG by glutathione reductase (GR) in an NADPH-dependent fashion. Both GR and the NADPH generating malic enzyme are also upregulated following treatment with ATO. While consistent with our previous findings the data do not provide much insight as to how the GSH is utilized. The only GSH utilizing enzyme that was observed to be upregulated were the cytosolic and mitochondrial forms of glutaredoxin. Glutathione peroxidase (GPx) activity is not altered by ATO treatment. However GPx baseline expression and activity do correlate with sensitivity of MM cell lines to ATO. We also determined GSTP1 activity in the cells and found that it was expressed in 4/5 MM cell lines tested. In contrast to GPx, GSTP1 baseline expression did not correlate with sensitivity to ATO. However this pattern of expression correlated with our previous findings regarding these cells demonstrating differences in caspase dependence of ATO-induced cell death. The one line that did not express GSTP1, RPMI 8226, also utilizes caspase-independent mechanisms of cell death. Transfection of the GSTP1A allele into these cells could render cells more resistant to ATO-induced apoptosis at concentrations of ATO that are not likely to be achieved in patients. Interestingly transfection of the GSTP1B allele could not render cells more resistant, however like GSTP1A it resulted in inhibition of the caspase-independent pathway. Taken together these data confirm that GSH is an important modulator of ATO therapy and that GPx expression may determine the sensitivity of cells to ATO while GSTP1 can affect the mechanism of action by which ATO-induces apoptosis.