Nicotine-induced survival signaling in lung cancer cells is dependent on their p53 status while its down-regulation by curcumin is independent

Even though nicotine-induced proliferation of lung cancer cells has been reported [4], our findings show for the first time that, among the lung cancer cell lines, nicotine induces more proliferation in H1299 lacking p53 especially at lower concentrations (10^-9M -10^-7M) than A549 with active p53 (maximum at 10^-7M -10^-6M). Of interest, pre-treatment with a non-toxic concentration of curcumin significantly reduced (p < 0.001) the viability of both lung cancer cells (Fig 1. A-B).

Nicotine has been shown to activate NF-κB in cancer cells of various origins, including lung [3, 8]. We also have reported earlier that cigarette smoke induces strong activation of NF-κB in H1299 cells [29]. In the present study, H1299 and A549 cells were exposed to various concentrations of nicotine and the status of nuclear translocation of NF-κB was studied. Interestingly, nicotine could activate NF-κB only in H1299 cells (Fig. 2A) while it failed to induce a significant activation of NF-κB in A549 cells at any of the concentrations studied (Fig. 2B). In H1299 cells, lower concentrations (10^-9M-10^-7M) of nicotine induced maximum activation of NF-κB and a time kinetics study using 10^-8 M nicotine showed maximum activation at 30 min of exposure to nicotine (Fig. 2C). In congruence, the phosphorylation pattern of IKK and degradation pattern of IκBα (Fig. 2D) correlated with that of the NF-κB activation, attesting that nicotine induces NF-κB through the classic pathway. The specificity of the NF-κB band was verified by super shift assay and by cold competition (Fig. 2E). The activation of NF-κB induced by nicotine in H1299 cells was completely abolished by treatment with 10 μM curcumin (Fig. 2F).

We compared the effect of nicotine in A549 and H1299 cells in inducing pro-survival molecular entities-Akt, COX-2 and Cyclin D1, which generally get activated in response to NF-κB activation. Though nicotine was ineffective in inducing NF-κB in A549 cells, it effectively induced Akt, COX-2 and Cyclin D1 in both the cells (Fig. 3A-C) indicating that nicotine activates these pro-survival signaling molecules independent of NF-κB. However, there was a marked difference in the expression pattern of all these survival signals between the investigated cell lines. It was noticed that, while nicotine induces activation of these survival molecules in the range of 10^-9M - 10^-7M in H1299 cells, a 10-100 fold high concentration of nicotine was needed to induce the same effect in A549 (Fig. 3A-C). Despite studies reporting the interdependence and cross-talk between Akt and NF-κB [26], phosphorylation of Akt by PI3K pathway independent of NF-κB is also well evidenced [30‐33]. Similarly, while several reports strongly correlate the involvement of NF-κB in Cyclin D1 regulation [30], there are reports which clearly demonstrate the transcription of Cyclin D1, independent of NF-κB, in cell lines including A549 [32, 4]. The protein level expression of Cyclin D1 in A549 was low compared to that in H1299 cells at all the concentrations studied (Fig. 3C). We infer from these results that nicotine-induced expression of these survival signals was severely compromised in the presence of curcumin in both the investigated cell lines (Fig. 3D-F) and hence we draw attention to the protective role of curcumin against survival and proliferative signals induced by nicotine.

The IAP family of proteins involved in apoptosis regulation is generally activated in response to NF-κB activation [33]. Consistent with previous reports indicating that 1 μM nicotine induces the IAP family members, XIAP and survivin [14], we also observed the over-expression of XIAP and survivin in A549 and H1299 cell lines, albeit difference in the concentration of nicotine for effective stimulation of these proteins. However, we observed robust expression of cIAP-1 too in both the cell lines by nicotine (Fig. 4A-C), though Dasgupta et al[14] did not observe it in A549 cells. It was inferred that while nicotine induces activation of these survival signals in the range 10^-9 M-10^-7 M in H1299 cells, it induces the same level of activation in A549 cells at 10^-6 M-10^-4 M range (Fig. 4A-C). Additionally, because Bcl2 is considered as a general inhibitor of apoptosis, we evaluated the role of nicotine in its induction above basal level and tested the effect of curcumin in conferring chemoprotection. As in the case of other survival signals, a 10-100 fold higher concentration of nicotine was necessary to induce Bcl2 in A549 cells, compared to H1299 cells (Fig. 4D) although, curcumin abolished it more or less completely in both the cell lines. Results obtained thus far indicate that in the investigated lung cancer cell lines, curcumin completely inhibits the over-expression of these survival signals (Fig. 4E-H) independent of the p53 status and the concentration of nicotine, strongly suggesting a preventive action of curcumin against nicotine-induced promotion and progression of lung cancer through up-regulation of antiapoptotic molecules.

Over-expression of cyclin D1 and IAPs by nicotine even in the absence of NF-κB led us to speculate the involvement of other pathways regulating the expression of these proteins. Involvement of the MAPK pathway [7] and the induction of cyclin D1 through the MAPK-AP-1 pathway [34] in nicotine-induced cell signaling is well evidenced. Induction of IAPs by MAPKs independent of NF-κB has also been reported [35]. We therefore performed a time kinetic study up to 120 min to investigate the functional role of MAPK in H1299 and A549 cells, by analyzing their phosphorylation status in the presence of nicotine and found that, nicotine induces maximum phosphorylation of all the MAPKs at 30 min (data not shown). Dose response study revealed that all the three MAPKs are phosphorylated in both the cell lines, being maximally activated at 10^-8M concentration of nicotine in H1299 cells (Fig. 5A) and at 10^-6 M in A549 cells (Fig. 5B). However, curcumin treatment (10 μM) inhibited the phosphorylation of these MAPKs by nicotine in both the cells (Fig. 5C-D).

Regulation of MAPK pathway by nicotine through p53 has already been reported [25, 36]. We therefore assessed whether p53 has any regulatory role in the activation of AP-1, which gets translocated to the nucleus in response to the phosphorylation of MAPKs since, up-regulation of AP-1 activity by nicotine has been reported in the literature [37]. Interestingly, though nicotine failed to activate NF-κB in A549 cells, it activated AP-1 very efficiently in both H1299 and A549 cells (Fig. 5E-F) though the pattern of nuclear translocation of AP-1 was in parallel with the phosphorylation pattern of MAPKs in the respective cell lines. As in the case of MAPKs, curcumin treatment down regulated this activation almost completely in both the cells (Fig. 5G-H).

The correlation between NF-κB nuclear translocation and the p53 status of cells is controversial. Some researchers contend that p53 is needed for NF-κB activation [38, 39] while, others have shown that activation of NF-κB can be induced in a p53 independent manner [40]. We noticed differences in the activation pattern of NF-κB, MAPKs and AP-1 in cells harboring p53 versus cells lacking p53. To corroborate whether this distinction is due to the p53 status of the cells, we transfected H1299 cells with p53 WT plasmid and A549 cells with p53 DN plasmids, followed by selection of the clones with maximum p53 expression for further studies. In H1299-p53 WT, clone 6 (Fig. 6A) and in A549-p53 DN, clone 3 was selected (Fig. 6B). Whereas nicotine failed to induce NF-κB nuclear translocation in A549 cells, it caused significant induction of NF-κB in A549-p53 DN cells within the concentration range of 10^-9M - 10^-7M similar to H1299 cells (Fig. 6C). Supporting this observation, inhibition of p53 in A549 cells using 40 μM pifithrin-α, which completely inhibits p53 (Fig. 6E) induced NF-κB nuclear translocation in A549 cells (Fig. 6F). Likewise, nicotine failed to induce NF-κB activation in H1299-p53 WT cells except at a very high concentration of 10^-2 M (Fig. 6D) though, it induced very strong activation of NF-κB in H1299 cells at 10^-9-10^-7 M. This confirms the interdependence between NF-κB and p53. Furthermore, the mode of phosphorylation of MAPKs and the nuclear translocation of AP-1 by nicotine was also seen to be reversed in the transfected cells (Fig. 6G-J), further adding support to our hypothesis.

Because of the difference in the p53 status between the investigated cells, we studied whether p53 can differentially regulate the clonogenicity by nicotine in these cells. While 10^-8M nicotine induced maximum number of clones in H1299 cells, 10^-6M nicotine produced maximum proliferation in A549 cells as indicated by the bigger size of the colonies formed (Fig. 7A-B). Moreover, the number and size of the clones was much higher in H1299 cells compared to A549. As expected, the number of clones by nicotine addition in A549 was escalated when p53 was shut down, whereas those produced in H1299 was significantly reduced (p < 0.005) when p53 was transfected (Fig. 7A-B). However, in both the cells, pretreatment with curcumin (10 μM, 2 h) drastically reduced the number and size of the clones induced by nicotine, clearly indicating that curcumin negatively regulates clonogenicity (Fig. 7A-B).

As part of investigation of this difference in the activation pattern of various survival signals by nicotine in cells with and without p53, we evaluated the effect of nicotine on expression levels of p53 and p21 in A549 parental and A549-p53 DN cells. Interestingly, while lower concentrations (10^-9 M - 10^-7 M) of nicotine induced over-expression of p53 and p21 in A549 cells, it failed to induce significant changes in the expression pattern of p53 and p21 in A549-p53 DN cells (Fig. 7C-D). To check whether p53 is the only molecule regulating the p21 signaling in A549 cells, we also investigated the expression status of p21 in response to nicotine in H1299 cells. We did not find any significant difference in the expression of p21 in H1299 cells (Fig. 7E), implying that p53 is the key regulator of p21 in nicotine-induced signaling events. To confirm this hypothesis, we checked the p21 status in H1299-p53 WT cells in response to nicotine. As in the case of A549 cells, a strong induction of p53 was produced by nicotine in these cells (Fig. 7E) which confirms the regulatory role of p53 in nicotine-induced p21 expression.

We further validated our hypothesis that impaired p53 status leads to proliferation and induction of survival signals in lung cancer cells, by treating two other lung cancer cells harboring mutant p53(Hop-92 and NCI-H522) with nicotine and curcumin. We observed a significant proliferation of these cells by lower concentrations of nicotine (p < 0.001) which is prevented by curcumin pre-treatment (Fig. 8A-B). As in H1299, nicotine induced NF-κB nuclear translocation in these cells too (Fig. 8C-D). Moreover, as we observed in A549-p53 DN cells, nicotine induced NF-κB nuclear translocation when p53 was silenced in A549 cells transfected with p53 siRNA and interestingly, co-transfection of these cells with WT p53 completely inhibited the nuclear translocation of NF-κB (Fig. 8E-F) confirming the role of p53 in regulating nicotine-induced cell signaling.

H1299 cell line was a gift from Prof. BB Aggarwal, MD Anderson, Houston, A549 and NCI-H522 cells were procured from NCCS, Pune, and Hop-92 was obtained from NCI, Bethesda, MD.

Antibodies against p53, IκBα, XIAP and cIAP-1 and phosphospecific antibodies against-IKK, p38, Akt and p42/44 were purchased from Cell Signalling (Beverly, MA) and those against p50, RelA, pJNK, COX-2 and Survivin were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). G418 and pifithrin-α were procured from Calbiochem (San Diego, USA). Lipofectamine was procured from Invitrogen (Carlsbad, CA) and control siRNA and p53 siRNA were procured from Dharmacon, Inc (Chicago, IL). All other chemicals were purchased from Sigma Chemicals (St. Louis, MO).

In all combination treatments 10 μM curcumin was added 2 h before nicotine (10^-8 M in H1299 and 10^-6 M in A549).

Proliferative/cytotoxic effect of nicotine and/or curcumin was determined by MTT assay as described earlier [18].

For [³H] thymidine incorporation assay, cells were seeded in 96 well plates and treated with required concentrations of the drugs. After 18 h of incubation, [³H] thymidine was added (0.5 μCi/well) and further incubated for 6 h. Then the culture medium was removed, the cells were washed twice with phosphate-buffered saline (PBS), treated with 5% trichloroacetic acid, pelleted and supernatant was removed. Cells were then washed with ethanol and solubilized with 0.2N NaOH, contents of each well were mixed with 3 ml scintillation fluid and the radioactivity was counted using a liquid scintillation counter [59]. The experiment was repeated thrice and the error bars indicate standard deviation.

To detect activation of NF-κB and AP-1, EMSA was done and visualized by a Phosphor Imager (Bio-Rad Personal FX). The specificity of the bands were confirmed by super shift [29].

The plasmids were isolated by the standard alkaline lysis method [60].

H1299 Cells were transfected with pcDNA3-p53WT or the empty vector pcDNA3 while A549 cells were transfected with pcDNA3-p53 DN or the empty vector pcDNA3 using the calcium-phosphate transfection kit (Life Technologies, Inc.) according to manufacturer's protocol and selected using G418 (150 μg/ml) [61].

A549 cells were transiently transfected with control siRNA, p53 siRNA and/or WT-p53 plasmid using lipofectamine 2000 reagent according to manufactures protocol (Invitrogen).

The total protein after treatment with nicotine and/or curcumin was electroblotted as described earlier [18] and developed by Enhanced Chemiluminescence (Immobilon Western, Miliipore, Billerica, MA).

The cells were treated with nicotine and/or curcumin for 72 h, as in MTT assay after which the cells were split and seeded in six well plates and clonogenic assay was done as described by Franken et al[62].

Statistical significance was calculated using one-way ANOVA followed by Tukey post-hoc analysis.