Reactive oxygen species (ROS), produced primarily in cells as a by-product of normal metabolism by the mitochondria, have long been implicated in induction of apoptosis ^{1, 2}. Accordingly, the most abundant non-protein thiol, glutathione (GSH), plays an important role in the regulation of apoptosis due to its role as a substrate of ROS scavenging enzymes ³. N-acetylcysteine (NAC) is an aminothiol and synthetic precursor of intracellular cysteine and GSH and is thus considered an important antioxidant ⁴. NAC has been widely used as a research tool in the field of apoptosis research for investigating the role of ROS in induction of apoptosis.

It is generally assumed that the action of NAC results from its antioxidative or free radical scavenging property as an antioxidant through increasing intracellular GSH levels; however, NAC also possesses a reducing property through its thiol-disulfide exchange activity ^{4, 5}. For example, NAC has been shown to induce cell cycle arrest in hepatic stellate cells by modulating the mitogen-activating protein (MAPK) pathway. This involves the direct interaction with target proteins with cysteine residue or thiol group such as Raf-1, MEK and ERK via a thiol-disulfide exchange reaction ⁶.

In the work by Song et al in this issue of the journal ⁷, the apoptosis-inducing activities and the possible underlying mechanism of several 18β-glycyrrhetinic acid (GA) derivatives, particularly methyl 2-cyano-3,11-dioxooleana-1,12-dien-30-oate (CDODO-Me-11) and methyl 2-cyano-3,12-dioxooleana-1, 12-dien-30-oate (CDODO-Me-12) were studied in human leukemia cells. Both CDODO-Me-12 and CDODO-Me-11 effectively induce apoptosis of leukemia cells including activation of caspase-8 and caspase-9 although CDODO-Me-12 showed greater potency than CDODO-Me-11. Moreover, they down-regulate the levels of anti-apoptosis proteins, c-FLIP, XIAP and Mcl-1. These effects can be reversed or attenuated by either GSH or NAC. However, both agents do not alter the intracellular H₂O₂ levels even though they reduce intracellular levels of GSH. Importantly, they elegantly demonstrate a direct chemical interaction between these two agents and NAC or GSH. Therefore, NAC protects cells from induction of apoptosis by CDODO-Me-12 or CDODO-Me-11 via chemically forming inactive adducts or complexes independent of its antioxidant activity. Accordingly they suggest that CDODO-Me-12 or CDODO-Me-11 may modulate certain key signaling pathways such as NF-κB by directly binding to functional proteins with a cycteine or a thiol group in these pathways, leading to downregulation of c-FLIP, XIAP and Mcl-1 and ultimate induction of apoptosis.

Similar results were also obtained with studies using synthetic triterpenoids such as methyl-2-cyano-3, 12-dioxooleana-1, 9-dien-28-oate (CDDO-Me) ⁸. While some investigators suggested that ROS induction of apoptosis by the synthetic triterpenoids primarily employs NAC as a tool ^{9, 10}, results from our lab suggested that CDDO-Me induced apoptosis in human lung cancer cells by directly interacting with GSH or thiol-containing components independent of ROS generation ⁸. In our study, we found that the thiol antioxidants NAC and GSH, but not nonthiol antioxidants, including butylated hydroxyanisole (BHA), Trolox, mannitol, and Mn(II) tetra(4-benoic acid) porphyrin chloride (MnTBAP), abolished all CDDO-Me-induced biological activities including JNK activation, DR5 upregulation and apoptosis. Moreover, CDDO-Me did not increase ROS generation in the tested systems. In agreement with our finding, CDDO-Me and its related synthetic triterpenoids have been shown to chemically interact with cysteine- or thiol-containing molecules such as GSH to form a reversible adduct ¹¹. In agreement with these findings, several studies have shown that CDDO-Me or its analogues directly interacts with IKKβ at Cys-179, janus-activated kinase-1 (JAC1) at Cys-1077 and signal transducer and activator of transcription-3 (STAT3) at Cys-259 ^12–14.

These studies clearly indicate that NAC can exert its action by directly reacting with another small molecule which has ability to interact with thiol-containing proteins. Thus, caution should be exercised when interpreting data where NAC is employed primarily for its antioxidant effect to demonstrate the involvement of ROS in drug-induced apoptosis. Indeed, the same may be true of other antioxidants when employed as research tools. For example, PS-341 (bortezomib or velcade) was suggested to induce apoptosis by promoting ROS generation primarily using the antioxidant tiron as a tool ^{15, 16}. In our own study, we found that only vitamin C and tiron, but not other antioxidants including NAC, GSH and butylated hydroxyanisole, abolished PS-341’s biological activities including induction of apoptosis and enhancement of TRAIL-induced apoptosis. Moreover, we did not detect increased ROS generation in cells treated with PS-341. In fact we have shown that vitamin C or tiron abolished PS-341’s biological activities via direct interaction and by forming a tight but reversible chemical complex with PS-341 using adjacent hydroxyl groups independent of their antioxidative properties ¹⁷. Given that PS-341 is a dipeptidyl boronic acid, theoretically it can form a chemical complex with small molecules whose structures contain adjacent hydroxyl groups. Indeed, a recent study also shows that the green tea (−)-epigallocatechin gallate (EGCG) and other polyphenols with 1,2-benzenediol moieties protect cells from PS-341-induced apoptosis through the same mechanism ¹⁸.

In summary, we strongly suggest that caution should be exercised when interpreting data where a single antioxidant such as NAC is employed to demonstrate the involvement of ROS in drug-induced apoptosis. When possible, it is highly recommended that multiple antioxidants including thiol and nonthiol antioxidants be employed to confirm ROS generation. Only with the right approaches and right data interpretation could the right conclusions be reached.