Cyclooxygenase-2 independent effects of cyclooxygenase-2 inhibitors on oxidative stress and intracellular glutathione content in normal and malignant human B-cells

Abstract

We recently reported that inhibition of Cyclooxygenase-2 (Cox-2) reduced human B-CLL proliferation and survival. Herein, we investigated the mechanisms whereby small molecule Cox-2 selective inhibitors, SC-58125 (a Celebrex analog) and CAY10404 blunt survival of human B-cell lymphomas and chronic lymphocytic leukemia B-cells. SC-58125 and OSU03012 (a Celebrex analog that lacks Cox-2 inhibitory activity) both decreased intracellular glutathione (GSH) content in malignant human B-cells, as well as in Cox-2 deficient mouse B-cells. This new finding supports Cox-2 independent effects of SC-58125. Interestingly, SC-58125 also significantly increased B-cell reactive oxygen species (ROS) production, suggesting that ROS are a pathway that reduces malignant cell survival. Addition of GSH ethyl ester protected B lymphomas from the increased mitochondrial membrane permeability and reduced survival induced by SC-58125. Moreover, the SC-58125-mediated GSH depletion resulted in elevated steady-state levels of the glutamate cysteine ligase catalytic subunit mRNA and protein. These new findings of increased ROS and diminished GSH levels following SC-58125 exposure support novel mechanisms whereby a Cox-2 selective inhibitor reduces malignant B-cell survival. These observations also support the concept that certain Cox-2 selective inhibitors may have therapeutic value in combination with other drugs to kill malignant B lineage cells.

Fig. 1

The Cox-2 selective inhibitor, SC-58125, depletes intracellular GSH in normal and malignant B lineage cells. a Ramos B lymphomas treated with vehicle (DMSO), 10, 20, or 40 μM SC-58125 for 3, 6, 12, and 24 h were analyzed for total intracellular GSH content. A dose-dependent decrease in GSH levels was detected by 10, 20, and 40 μM SC-58125 treated cells after 6, 12, and 24 h. Five μM OSU03012 (which lacks Cox-2 inhibitory activity) decreased GSH by ∼45%. b Untreated or CD40L-stimulated normal human B lymphocytes and B-CLL cells were cultured in the presence and absence of SC-58125 for 24 h. Significantly increased GSH levels were measured in CD40L stimulated cells compared to untreated. SC-58125 induced larger dose-dependent decreases in GSH levels in B-CLL cells, compared to normal human B lymphocytes. c Total GSH content was analyzed in wildtype and Cox-2 deficient mouse B-cells cultured with nothing or LPS in the presence and absence of SC-58125 (20 μM). LPS stimulated wildtype and Cox-2 deficient B-cells treated with SC-58125 showed decreased GSH levels compared to non-drug treated cells. No differences between wildtype and Cox-2 deficient B-cells were detected in untreated or LPS stimulated B-cells. Data are shown as mean ± SEM (n = 3, *P < 0.05). d Ramos B lymphomas were treated with SC-58125 for 6, 12, 18, and 24 h and the percentage of bimane-GSH positive cells was determined by flow cytometry using the thiol reactive dye, monochlorobimane (MCB). A dose-dependent decrease in the percentage of Ramos B lymphomas was seen as early as 6 h, and persisted until 24 h. BSO (100 μM) was used to set gate parameters and show complete GSH depletion. Data shown are representative of three experiments with similar results

Fig. 2

Increased ROS production by SC-58125 treated Ramos B lymphoma cells. a Ramos B lymphoma cells treated with SC-58125 showed an increase in the percentage of DCF positive cells as detected by flow cytometry. Histogram plots show the percentage of DCF positive cells increases with the dose of SC-58125 treated Ramos B lymphomas compared to vehicle control. b An increase in the percentage of cells positive for superoxide anion production was detected by dihydroethidium (HE) staining 24 h after exposure to SC-58125 at doses as low as 5 μM

Fig. 3

Cox-2 independent ROS induction in Ramos B lymphoma cells. DCF fluorescence intensity was analyzed by imaging flow cytometry in Ramos B lymphoma cells either a untreated or treated with b SC-58125 (20 μM) or c OSU03012 (5 μM). 7-AAD (cell impermeable) and Draq5 (cell permeable) dyes were used to analyze for cell viability. SC-58125 and OSU03012 showed an increased percentage of ROS positive cells compared to untreated. SC-58125 induced ROS in a larger proportion of live cells compared to OSU03012. A subset of early apoptotic OSU03012 treated cells (R2 gate-Draq5 and 7-AAD double positive) showed the largest increase in DCF fluorescence

Fig. 4

Increased ROS production by SC-58125 in normal human B lymphocytes. a Normal human B lymphocytes cultured with nothing or CD40L and treated in the presence of SC-58125 for 24 h showed significantly (85%) increased ROS production as detected by DCF compared to vehicle control treated cells. b Significantly increased mean fluorescence intensity of normal human B lymphocytes (cultured with nothing or CD40L) treated with SC-58125 were detected compared to control cells (n = 3, *P < 0.05)

Fig. 5

Increased steady-state GCLC mRNA levels in SC-58125 treated normal and malignant human B-cells. a GCLC mRNA levels were measured in SC-58125 treated normal human B-cells, Ramos B lymphomas and B-CLL cells. PCR products were analyzed on a 1.2% agarose gel. Increased GCLC mRNA expression was detected by normal B-cells treated with SC-58125 for 48 or 72 h. b An ∼4–12-fold induction of GCLC mRNA levels in SC-58125 treated Ramos B lymphomas over untreated controls was detected by real-time RT-PCR. Cycle threshold values were normalized to GAPDH. c Unstimulated or CD40L stimulated B-CLL cells treated with SC-58125 for 24 h showed a less than fivefold induction over non-drug treated samples (n = 3, *P < 0.05)

Fig. 6

Increased GCLC protein expression in Ramos B lymphoma cells by SC-58125. a Ramos B lymphomas treated with SC-58125 increased GCLC protein levels in a dose-dependent manner after 24 h as detected by Western blot. No major changes in GCLM protein expression were seen following treatment with SC-58125. b Densitometry analysis of Western blot in panel A shows significant fourfold increased GCLC expression in SC-58125 treated cells compared to untreated (n = 3, *P < 0.05)

Fig. 7

GSH restores SC-58125 induced mitochondrial alterations in Ramos B lymphoma cells. a SC-58125 dose-dependently reduced mitochondrial dehydrogenase activity in Ramos B lymphomas after 48 h as detected by MTT assay. Addition of GSH ethyl ester (1 mM) significantly restored the dose-dependent SC-58125 mediated reduction in mitochondrial activity. About 0.5 mM GSH ethyl ester showed partial rescue of the reduced mitochondrial activity. b SC-58125 increased mitochondrial membrane permeability in ∼90% of Ramos B lymphomas after 12 h as determined by DIOC₆ incorporation and detected by flow cytometry. Co-incubation of 1 mM GSH ethyl ester and SC-58125 (4 μM) completely restored mitochondrial membrane potential (n = 3, *P < 0.05)