A study of biological role of reactive oxygen species in cellular response in stress

Abstract

When proteins are unable to fold properly in the endoplasmic reticulum (ER), the

resultant formation of misfolded proteins causes stress of the ER. Cells with ER stress

often have a higher abundance of reactive oxygen species (ROS). Previous studies

suggest that ROS could aggravate ER stress by further disrupting the ER protein

folding process. More recent studies suggest that the unfolded protein response

signaling pathways activated by ER stress could lead to the production of ROS. Such

studies lead to the hypothesis that ER stress could be promoted by ROS, and vice

versa. The aim of the present study is to test the above hypothesis by studying how

ROS could be generated in ER-stressed cells. This is followed by investigating if ROS

could increase or decrease the level of ER stress in cells. Finally, the extent of ER

stress induced cell death in the presence and absence of ROS is assessed.

The treatment of HeLa cells with tunicamycin (Tm), a common ER-stress

inducing agent, resulted in the elevation of intracellular ROS that could be detected

with the ROS-reactive probe dichlorodihydrofluorescein (DCF), but not

dihydroethidium which is relatively specific towards superoxide anion. The

Tm-induced elevation of ROS could be prevented by co-incubation of cells with thiol

reductants such as dithiothreitol and N-acetylcysteine but not with the free radical

scavenger ascorbate. The tunicamycin-induced elevation of ROS level could also be

prevented by the over-expression of catalase in HeLa. These data is consistent with

the idea that hydrogen peroxide is a major form of ROS produced in Tm-treated cells.

In addition to elevation of ROS level, HeLa cells treated with tunicamycin also

resulted in the phosphorylation of PERK and eIF2α, and the splicing of XBP-1. In the

presence of cycloheximide to inhibit protein synthesis so as to deplete protein

substrates for folding in the ER, tunicamycin-induced ER stress was greatly

minimized as was evident by the absence of both the phosphorylation of PERK and

splicing of XBP-1. However, the phosphorylation of eIF2α and elevation of

DCF-detectable ROS remained unaffected. The cycloheximde-resistant

phosphorylation of eIF2α could be prevented when cells were co-treated with thiol

reductants, or upon the over-expression of catalase. These data suggest that the

production of ROS in Tm-treated cells does not require the presence of ER stress as a

prerequisite. Furthermore, the ROS so produced could induce phosphorylation of

eIF2α without the need to cause ER stress in the first place.

The quenching of ROS through the use of thiol reductants, or the over-expression

of catalase, had no effect on inhibition of protein synthesis in cells treated with

tunicamycin. However, the extent of cell death was significantly increased. The data

obtained in this study is not consistent with the idea that ROS is a downstream

product of ER stress, capable of inducing more ER-stress by a feedback mechanism.

Therefore, a mutually enhancing effect between ER stress and ROS may not exist.

The ROS found in stressed cells may serve to extend cellular survival under the

condition of continuous stress.