Nitric oxide-donating aspirin inhibits the growth of pancreatic cancer cells through redox-dependent signaling

Abstract

The novel chemopreventive nitric oxide-donating aspirin (NO-ASA) prevents nearly 90% of ductal adenocarcinomas in a animal tumor model. To decipher the mechanism of this effect, we studied in BxPC-3 human pancreatic cancer cells the sequence of signaling events leading from NO-ASA treatment to cell growth inhibition. NO-ASA inhibited the growth of BxPC-3 cells (IC(50) =13 microM), by inhibiting proliferation modestly and inducing apoptosis, necrosis and G(1)/S cell cycle block. At 15 min of treatment with NO-ASA, the intracellular levels of reactive oxygen species (ROS) began increasing (peak at 8h, baseline levels by 24h). ROS activated almost immediately in a time- and concentration-dependent manner the MAPK pathways p38, ERK and JNK (their activation was abrogated by the antioxidant N-acetylcysteine). MAPK activation induced p21(cip-1), which suppressed the levels of cyclin D1 that controls the G(1)/S cell cycle transition. NO-ASA induced COX-2 expression starting 90 min after p21(cip-1) was induced. When COX-2 expression was knocked down using siRNA against cox-2, the expression of p21(cip-1) was induced by NO-ASA, regardless of the level of expression of COX-2, suggesting a marginal, if any, role for COX-2 in the growth inhibitory effect of NO-ASA. These findings along with the temporal sequence of individual changes indicate a signaling sequence that involves ROS-->MAPKs-->p21(cip-1)-->cyclin D1-->cell death. Our findings establish the critical role of ROS as proximal signaling molecules in the action of anticancer compounds and may be useful in designing mechanism-driven approaches to cancer control.

Fig. 1. NO-ASA and its effect on BxPC-3 cell growth

A: The structural components of NO-ASA: the conventional aspirin moiety is linked by a spacer to the NO-donating group (–ONO₂, shaded). B: NO-ASA inhibits the growth of BxPC-3 pancreatic cancer cells, which were treated with various concentrations of NO-ASA for 24 h. Cell growth was assayed using the MTT assay. Data are expressed as percentage of the growth of untreated cells. The results are the average of three independent experiments. Values: mean±SEM. IC₅₀ = 13 μM.

Fig. 2. Effect of NO-ASA on BxPC-3 cytokinetics

A, BxPC-3 cells were treated with three different concentrations of NO-ASA for 24 h and BrdU incorporation was determined as in Methods. Only the highest NO-ASA concentration produced a statistically significant reduction in BrdU incorporation (P<0.04). B: Apoptosis and necrosis of BxPC-3 cells treated with NO-ASA 19.5 μM was determined as in Methods. Data are expressed as fold change over untreated controls. All values: mean±SEM (n=3).

Fig. 3. NO-ASA induces ROS production in BxPC-3 cells

A: Representative flow cytometric histogram of BxPC-3 cells treated with 19.5 μM of NO-ASA for 30 min; ROS levels were measured as in Methods. B: ROS levels at various time points following treatment with NO-ASA 19.5 μM. Values (mean±SEM; n=3) are expressed as fold of untreated controls.

Fig. 4. Effect of NO-ASA on MAPK activation

BxPC-3 cells were treated with NO-ASA 19.5 μM for various times (A) or with various NO-ASA concentrations for 1 h or 8 h (B) and the expression of the phosphoproteins shown was determined by immunoblotting. C: BxPC-3 cells were treated with 20 mM NAC for 4 h prior to treatment with NO-ASA. Loading control: μ-actin

Fig. 5. Effect of NO-ASA on p21 and cyclin D1 expression

BxPC-3 cells were treated with NO-ASA 19.5 μM for the indicated periods of time (A) or with three different concentrations of NO-ASA for 1 h or 8 h (B) or were pretreated with 20 mM NAC (C). The expression of proteins was determined by immunoblotting. Loading control: β-actin

Fig. 6. The role of COX-2 in the effect of NO-ASA on BxPC-3 cells

BxPC-3 cells were treated with NO-ASA 19.5 μM for the indicated periods of time (A) or with three different concentrations of NO-ASA for 8 h (B) and the expression of COX-2 was determined by immunoblotting. C: BxPC-3 cells were transfected with control siRNA or cox-2 siRNA for 24 h and then treated with three concentrations of NO-ASA for 8 h. Expression of COX-2, p21, and cyclin D1 was determined by immunoblotting. Loading control: β-actin. D: The levels of COX-2, p21, and cyclin D1, determined by densitometry, are presented as fold change over untreated controls (representative of two independent experiments giving similar results).

Fig. 7. The signaling cascade initiated by NO-ASA in BxPC-3 cells

Left: A plausible signaling cascade mediating the effect of NO-ASA on BxPC-3 cells. Right: The time periods during which each change occurred and a rough representation of their relative intensity based on the previously shown data. In this scheme, COX-2 does not modulate the expression of p21.