There is great interest in repurposing drugs currently available in the pharmacopeia to other indications like cancer as such an approach would significantly shorten the time needed and cost encumbered to provide new and effective cancer therapeutics. One such drug is the gold compound Auranofin (AF), which is FDA approved as a treatment for rheumatoid arthritis (RA). We demonstrate that AF induces MCL cell death with an LD50 of 1000 nM, 500 nM and 90 nM for Granta and Jeko cell lines, and cells derived from a patient biopsy, respectively. The increased sensitivity of primary MCL patient specimens to AF is confirmed on 4 additional patient samples tested. AF similarly induced DLBCL cell death with an LD50 of 500 nM, 500 nM and 1000 nM for OCI Ly-10, SUDHL-6 and −4 cell lines, respectively. Exposure of Granta cells to an AF concentration that induced cell death resulted in the generation of reactive oxygen species (ROS). Pre-treatment of these cells with N-acetyl-cysteine (NAC) or glutathione (GSH) abrogated both ROS generation and the induction of cell death supporting the notion that AF induces NHL cell death through a redox dependent pathway. Although AF does increase mitochondrial membrane permeability (although not through the classical permeability transition pore), the major mechanism through which AF induces NHL cell death is activation of the extrinsic apoptotic pathway. In this regard, AF induces the activation of caspases 7 and 8 and results in PARP cleavage, all of which are blocked by NAC. Despite the fact that AF is a known inhibitor of thioredoxin reductase (TR), its cytotoxic effect is independent of TR inhibition, as we observe no difference in the response to AF in U266 multiple myeloma cells transfected with an expression vector which results in the over-expression of TR. Given the redox dependence of AF-induced cytotoxicity we hypothesized that inhibition of the glutathione system with buthionine sulphoximine (BSO) would further augment AF induced cell death. To test this hypothesis, Granta cells were exposed to both AF and BSO. Significant synergistic interactions of these drugs were seen when tested in a Laska analysis of synergy. For example, whereas the LD50 for AF alone in Granta cells was 1000 nM, the LD50 for AF in combination with 5 μM BSO was 200nm; for Ly-10 cells, the LD50 for AF±BSO was 400 nm vs. 190 nM, respectively, and for SUDHL-6, 411 nM vs. 185 nM, respectively. Similar results were seen in MCL cells flow cytometrically sorted from patient biopsy specimens. As observed in studies using AF alone, the cytotoxic effects of the combination were blocked with both NAC and GSH. Similar to results with AF alone, the synergistic effects of AF and BSO on NHL cytotoxicity were independent of their effects on TR. Exposure of Granta cells to AF resulted in NF-κB inhibition. NF-κB was further inhibited with concomitant exposure to BSO over-expression of relA in Granta cells, however, had no effect on AF and BSO induced cell death, suggesting the synergistic effects of AF and BSO on NHL cell death may be NF-κB-independent. Finally, we demonstrate that AF and BSO modify free thiols on the plasma membrane. As we have recently shown that the redox agent parthenolide induces NHL death in part by modifying surface protein thiols, AF and BSO may induce NHL cell death through a similar mechanism. In summary, AF induces both MCL and DLCL cell death through a redox-dependent pathway that involves the extrinsic apoptotic pathway. Profound synergy is seen with concomitant depletion of glutathione by BSO. The data presented above, along with the fact that AF is a drug having a favorable safety profile in patients, and is an FDA-approved drug for the treatment of RA makes it an attractive candidate for further study as a single agent and in rational combination with other redox active drugs.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.