Nicotine induces cell proliferation by β-arrestin–mediated activation of Src and Rb–Raf-1 pathways

Recent studies have shown that nicotine, a component of cigarette smoke, can stimulate the proliferation of non-neuronal cells. While nicotine is not carcinogenic by itself, it has been shown to induce cell proliferation and angiogenesis. Here we find that mitogenic effects of nicotine in non–small cell lung cancers (NSCLCs) are analogous to those of growth factors and involve activation of Src, induction of Rb–Raf-1 interaction, and phosphorylation of Rb. Analysis of human NSCLC tumors show enhanced levels of Rb–Raf-1 complexes compared with adjacent normal tissue. The mitogenic effects of nicotine were mediated via the α₇-nAChR subunit and resulted in enhanced recruitment of E2F1 and Raf-1 on proliferative promoters in NSCLC cell lines and human lung tumors. Nicotine stimulation of NSCLC cells caused dissociation of Rb from these promoters. Proliferative signaling via nicotinic acetylcholine receptors (nAChRs) required the scaffolding protein β-arrestin; ablation of β-arrestin or disruption of the Rb–Raf-1 interaction blocked nicotine-induced proliferation of NSCLCs. Additionally, suppression of β-arrestin also blocked activation of Src, suppressed levels of phosphorylated ERK, and abrogated Rb–Raf-1 binding in response to nicotine. It appears that nicotine induces cell proliferation by β-arrestin–mediated activation of the Src and Rb–Raf-1 pathways.

Tobacco smoke contains a variety of tobacco-specific carcinogens, many of which are derivatives of nicotine that are formed during the curing of tobacco (1). These include molecules like 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and N′-nitrosonornicotine (NNN) (2). Nicotine itself exerts its cellular functions through nicotinic acetylcholine receptors (nAChRs), which are widespread in neurons and neuromuscular junctions (3). Studies in recent years have shown that nAChRs are also present in a variety of non-neuronal tissues, including human bronchial epithelial cells, human endothelial cells, and astrocytes (4–6). These observations suggested that signaling through the nAChRs could have functional roles in non-neuronal cells as well (7). Further, it appears likely that the pathological role of nicotine in human diseases is mediated, at least in part, through its direct effects on non-neuronal cells (6, 8).

The finding that nAChRs are present on non-neuronal cells was followed by the observation that nicotine could induce the proliferation of endothelial cells (4, 9). Further, it was found that nicotine and structurally related carcinogens like NNK could induce the proliferation of a variety of small cell lung carcinoma cell lines (10–12). This led to the hypothesis that nicotine and other tobacco carcinogens might be playing a direct role in the induction and progression of human lung cancers (4, 5, 13). While there is no evidence that nicotine contributes to the induction of tumors, it has been demonstrated that nicotine promotes the growth of solid tumors in vivo, suggesting that nicotine might be contributing to the progression of tumors already initiated (4, 14). Indeed, studies by Song et al. have shown that nAChRs expressed on lung carcinoma form a part of an autocrine-proliferative network that facilitates the growth of neoplastic cells (13, 15); other studies have demonstrated that nicotine can promote the growth of colon, gastric, and lung cancers (4, 5, 13, 16, 17).

It has been found that in non-neuronal tissues nicotine induces the secretion of growth factors like bFGF, TGF-α, VEGF, and PDGF (18), upregulation of the calpain family of proteins (19) as well as COX-2 and VEGFR-2 (20), causing the eventual activation of Raf/MAPK kinase/ERK (Raf/MEK/ERK) pathway (21, 22). Since nAChRs do not have intrinsic tyrosine kinase activity (3), the molecular mechanisms underlying the proliferative signaling remain unclear. Here we demonstrate that nicotine-mediated induction of cell proliferation involves recruitment of β-arrestin to the receptor, which facilitates the activation of Src; this in turn leads to the binding of Raf-1 kinase to Rb, leading to cell cycle entry. IP/Western blot analysis of human non–small cell lung cancer (NSCLC) tumor tissues showed elevated Rb–Raf-1 complexes in tumors relative to adjacent normal lung tissue, suggesting that perhaps the Rb–Raf-1 pathway contributes to the genesis of these tumors. Further, chromatin IP (ChIP) analysis of human NSCLC tumor samples demonstrated increased recruitment of E2F1 and Raf-1 to proliferative promoters like cdc6 and cdc25A. Our results suggest that binding of β-arrestin to nAChRs is an early and critical event in the initiation of nicotine-induced mitogenesis; the subsequent steps resemble growth factor–induced cell proliferation, including activation of Src, association of Rb to Raf-1, inactivation of Rb, and enhanced recruitment of E2F1 and Raf-1 to proliferative promoters. These events can be expected to contribute to the growth and progression of tumors exposed to nicotine through tobacco smoke or cigarette substitutes.

Nicotine induces proliferation of NSCLCs via α₇ -nAChRs. We examined the proliferative effects of nicotine in NSCLC cell lines A549 (human bronchoalveolar carcinoma), NCI-H23, NCI-H441 (lung adenocarcinoma), and NCI-H226 (pleural effusion squamous cell carcinoma) as well as on primary normal human bronchial epithelial cells (NHBEs), small airway epithelial cells (SAECs), human aortic endothelial cells (HAECs), and human microvascular endothelial cells from lung (HMEC-Ls). BrdU incorporation assays showed that 1 μM nicotine, which is equivalent to the amount present in the bloodstream of those who smoke 1 pack a day, stimulated S-phase entry in all the aforementioned cells. Figure 1A shows that the level of BrdU incorporation in quiescent A549 cells was only 12% ± 0.4%; upon treatment with 1 μM nicotine, 89% ± 7.6% of the cells were found to be in S-phase. Similar results were obtained in the other NSCLC cells. The mitogenic effects of nicotine were abrogated by the general nAChR antagonist hexamethonium bromide but were unaffected by atropine, an antagonist of the closely related muscarinic receptor, showing that the proliferative effects of nicotine required nAChR function (Figure 1, A and B). The stimulation of A549 and H23 cells with 1 μM nicotine induced 89% ± 7.6% and 83% ± 7.1% of the cells into S-phase, respectively. When the cells were treated in the presence of hexamethonium the percentage of cells in S-phase dropped to 28% ± 1.7% (in A549 cells) and 34% ± 2.8% (in H23 cells). Further, nicotine-induced proliferation of A549, H23, H441, and H226 cells was significantly ablated by 2 α₇ subunit antagonists, α-bungarotoxin and methylallyl aconitine (MAA), whereas it was unaffected by α-lobeline (α₄β₂ subunit inhibitor) or dihydro β-erythoidine (DHβE; α₃β₂ and α₄β₂ subunit inhibitor) (5), suggesting that α₇ subunits primarily mediate the mitogenic effects of nicotine (Figure 1A). The BrdU assay graphs represent 2 independent experiments, each data point being in duplicate and representing 100 cells from 3 fields. Similar results were obtained in NHBEs and SAECs from normal lungs as well as HMEC-Ls (Figure 1B). The role of α₇-nAChR in the proliferative effects of nicotine was further verified by siRNA techniques (Figure 1, C and D). The densitometric analysis shown in Figure 1D is representative of 3 Western blotting experiments performed with independent sets of lysates. Nicotine stimulation of A549 cells transfected with α₇-nAChR siRNA showed reduced S-phase entry relative to those transfected with control siRNA (Figure 1C), confirming the results with receptor antagonists.

Figure 1

Nicotine induces Rb inactivation and S-phase entry in lung cells. (A) Nicotine induces S-phase entry. Quiescent H23, H441, H226, and A549 cells were treated with 1 μM nicotine for 18 hours in the presence or absence of a panel of nAChR subunit inhibitors. hex, hexamethonium bromide; α-BT, α-bungarotoxin. (B) A similar experiment was performed using normal lung cells, namely NHBEs, SAECs, and HMEC-Ls. (C) Transfection of α₇-nAChR siRNA ablated the mitogenic activity of nicotine on A549 cells, whereas a nontargeting control siRNA had no effect. (D) Western blotting analysis showed that α₇-nAChR siRNA ablated the expression of α₇-nAChR in A549 cells, which was quantitated using densitometric analysis. A nontargeting siRNA sequence was used as a control for all experiments. (E) An in vitro assay using full-length Rb as a substrate to measure cyclin D–associated kinase activity in quiescent and nicotine-stimulated A549 cells. The lower panel shows a Coomassie Blue staining of the gel indicating comparable amounts of substrate in all lanes. (F) A similar assay as in E, showing cyclin E–associated kinase activity using histone H1 as a substrate. (G) Nicotine induces the dissociation of E2F1 from Rb in A549 cells, with concomitant phosphorylation of Rb. WB, Western blot. (H) Nicotine caused increased recruitment of E2F1 on the E2F-responsive cdc6 promoter and cdc25A accompanied by concomitant dissociation of Rb. PCR for the c-Fos promoter was taken as the negative control.

Nicotine induces cyclin/cdk activity and Rb phosphorylation. The retinoblastoma protein plays a central role in regulating cell cycle progression, and the Rb gene is inactivated in a wide variety of cancers (23, 24). Inactivation of Rb by kinases associated with cyclins D and E leads to the activation of E2F-regulated proliferative promoters, facilitating S-phase entry (24). The effect of nicotine stimulation on cyclin/cdk activity as well as Rb function was next examined. Toward this purpose, quiescent A549 cells were stimulated with 1 μM nicotine for different time periods; in vitro kinase assays showed that the kinase activity associated with cyclins D and E was greatly enhanced upon nicotine stimulation (Figure 1, E and F). Furthermore, nicotine stimulation led to the dissociation of E2F1 from Rb as assessed by IP/Western blot analysis (Figure 1G, top panel), correlating with the induction of cyclin D/cdk4 and cyclin E/cdk2 activity as well as Rb phosphorylation (Figure 1G, middle 2 panels). It thus appears that nicotine stimulation affects various components of the cell cycle machinery similar to growth factor stimulation.

ChIP assays were performed to assess the occupancy of E2F1 and Rb on E2F-responsive proliferative promoters (cdc6 and cdc25A) upon nicotine stimulation. It was observed that low levels of E2F1 and high levels of Rb were bound to the cdc6 and cdc25A promoters in quiescent A549 cells. Stimulation with 1 μM nicotine caused the dissociation of Rb from both the promoters, while there were increased amounts of E2F1 bound to them (Figure 1H, top 2 panels). PCR for the c-Fos promoter was used as a negative control; there was no Rb or E2F1 associated with this promoter (Figure 1H, bottom panel). Similarly, significant amounts of the repressive E2Fs, namely E2F4 and E2F5, were bound to cdc6 and cdc25A promoters in quiescent cells. In contrast, low levels of the proliferative E2Fs, namely E2F1–3, were found to be associated with the cdc6 promoter in quiescent A549 cells (data not shown). Among the Rb family members, robust amounts of Rb, p130, and p107 were bound to cdc6 promoter. Nicotine stimulation caused the dissociation of Rb, p107, p130, E2F4, and E2F5 with a concomitant increase in the binding of the proliferative E2Fs E2F1–3 (data not shown). Taken together, it appears that nicotine preferentially induces the association of proliferative E2Fs to mitogenic promoters, thereby stimulating S-phase entry.

Binding of Raf-1 kinase to Rb is necessary for nicotine-mediated induction of cell proliferation. Previous studies had demonstrated that the Raf-1 kinase physically interacts with Rb upon serum stimulation and facilitates cell cycle progression; the binding of Raf-1 to Rb was necessary for the subsequent phosphorylation by cyclins/cdks (25, 26). Since the nicotine derivative NNK has been shown to induce Raf-1 kinase activity (21), we examined whether nicotine promotes the binding of Raf-1 to Rb. Lysates from quiescent A549 cells and primary HAECs or those stimulated with 1 μM nicotine were immunoprecipitated with an anti-Rb antibody, and the presence of Raf-1 was examined by Western blotting. As shown in Figure 2A, quiescent A549 cells and HAECs did not have any detectable Rb–Raf-1 interaction, but stimulation with nicotine induced a robust Rb–Raf-1 interaction. The Rb–Raf-1 interaction was detected up to 2 hours after nicotine stimulation and dissipated within 4 hours. The binding between Rb and Raf-1 was also observed in lung cancer cell lines H23, H441, and H226 (Figure 2B) as well as in primary lung cells such as SAECs, NHBEs, and HMEC-Ls stimulated with 1 μM nicotine (Figure 2C). Nicotine-induced Rb–Raf-1 interaction was abolished by α-bungarotoxin and MAA, suggesting a role for α₇-nAChR in this binding (Figure 2D).

Figure 2

Rb–Raf-1 interaction is induced by nicotine in cultured cells and is elevated in tumor samples. (A) Nicotine induces the binding of Raf-1 to Rb in HAEC and A549 cells as seen by IP/Western blotting. Rb–Raf-1 interaction was also observed in other NSCLC cell lines like H441, H226, and H23 (B) as well as in SAECs, NHBEs, and HMEC-Ls (C). (D) Nicotine-induced Rb–Raf-1 binding is mediated by α₇-nAChRs in A549 cells. Quiescent A549 cells were stimulated with 1 μM nicotine in the presence or absence of 1 μM of the indicated antagonists; IP/Western blot shows the inhibition of the Rb–Raf-1 interaction by α₇-nAChRs inhibitors α-bungarotoxin and MAA. (E) Penetratin–Raf-1 (Pen–Raf-1) peptide conjugate abrogated nicotine-induced S-phase entry in A549 cells (black bars) and HAECs (gray bars) whereas a scrambled peptide conjugate, Pen-Rafscram, had no effect. (F) The penetratin–Raf-1 conjugate abolished Rb–Raf-1 binding in nicotine-stimulated A549 cells and HAECs, whereas the scrambled peptide conjugate (Pen–Rafscram) did not; Rb–Raf-1 interaction was measured by IP/Western blot. (G) NSCLC tumors (N) contained more Rb–Raf-1 complexes than adjacent normal tissue (T). Rb–Raf-1 interaction was assessed by IP/Western blot on nuclear extracts. (H) ChIP assays on human NSCLC tumor samples show that more Raf-1 was present on cdc6 and cdc25A promoters in tumor samples compared with adjacent normal lung tissues.

We had observed that delivery of an 8–amino acid peptide derived from Raf-1 coupled to a carrier peptide (penetratin) could disrupt the Rb–Raf-1 interaction and inhibit cell proliferation induced by growth factors (26). We next evaluated whether the penetratin–Raf-1 peptide conjugate affects nicotine-mediated proliferation of HAECs as well as A549 cells. Stimulation of quiescent HAECs and A549 cells with 1 μM nicotine resulted in robust proliferation, as seen by BrdU incorporation assays (Figure 2E). Nicotine stimulation in the presence of 1 μM penetratin–Raf-1 peptide conjugate resulted in a significant reduction in S-phase entry, while a control scrambled peptide conjugate had no effect (Figure 2E). BrdU incorporation experiments involving penetratin–Raf-1 conjugate were performed twice, and each data point was performed in duplicate. This inhibition correlated with the disruption of the Rb–Raf-1 interaction, as seen by IP/Western blotting (Figure 2F). It thus appears that the Rb–Raf-1 interaction facilitates S-phase entry upon nicotine stimulation as well, as in the case of growth factors. Our results indicate that the binding of Raf-1 kinase to Rb is necessary for nicotine-inducted cell proliferation and that this involves α₇-nAChRs.

Enhanced association of Raf-1 kinase with Rb and proliferative promoters in NSCLC samples. Epidemiological studies indicate that nicotine and tobacco carcinogens are important etiologic agents contributing to the development of lung cancer (27, 28). Given our observation that nicotine induced the binding of Raf-1 to Rb, we examined whether the Rb–Raf-1 interaction was altered in human NSCLC samples. Nuclear extracts from 10 human NSCLC tumors and the adjacent normal tissue from patients were immunoprecipitated with an anti-Rb antibody, and the presence of Raf-1 was examined by Western blotting. As shown in Figure 2G, there were elevated levels of Rb–Raf-1 interaction in 8 of 10 tumors compared with the adjacent normal tissue. This suggests that increased Rb–Raf-1 interaction correlates with oncogenesis. Since Raf-1 reverses Rb-mediated repression of E2F1 transcription, we examined the occupancy of E2F1, Rb, and Raf-1 on 2 E2F-regulated proliferative promoters in human NSCLC tumor samples. Toward this purpose, ChIP assays were performed on 2 NSCLC tumor samples along with their adjacent matched normal lung tissues. As seen in Figure 2H, both E2F1 and Rb were bound to the cdc25A and cdc6 promoters in normal tissues. However, only low levels of Raf-1 were associated with the promoters. In contrast, a significant amount of Raf-1 was bound to both the cdc6 and cdc25A promoters in NSCLC tumor samples. These results suggest that association of Raf-1 with Rb correlates with enhanced expression of proliferative promoters and oncogenesis.

Activation of Src is vital for nicotine-mediated Rb-Raf binding. It has been shown that nicotine can induce calcium channels and activate Src in neuronal as well as certain non-neuronal cells (29, 30); it has also been suggested that EGFR contributes to nAChR signaling (17). The contribution of these molecules to nicotine-induced proliferation of NSCLC cells was examined by BrdU incorporation (Figure 3A). Quiescent A549 cells showed 12.5% ± 1.0% cells in S-phase; upon nicotine stimulation, 86% ± 2.3% cells were found to enter S-phase. The presence of 1 μM of EGFR inhibitor Iressa (gefitinib) abrogated the mitogenic activity of nicotine (reduction of S-phase cells from 86% ± 2.3% to 44% ± 1.3%), whereas 1 μM of the calcium channel blocker nifedipine did not. In contrast, 1 μM of the Src inhibitor PD180 abrogated nicotine-induced proliferation of A549 cells (reduction of cell proliferation from 86% ± 2.3% to 22% ± 0.7%). The graphs shown in Figure 3A represent the average of 2 independent experiments, each data point being performed in duplicate. Next we examined whether these molecules mediate the Rb–Raf-1 interaction upon nicotine stimulation. A549 cells were rendered quiescent by serum starvation and subsequently stimulated with 1 μM nicotine. IP/Western blot analysis of the lysates showed that treatment of cells with 1 μM of EGFR inhibitor Iressa or 1 μM of the calcium channel blocker nifedipine did not affect Rb–Raf-1 binding (Figure 3B). However, the Src inhibitor PD180 ablated the association of Rb and Raf-1, showing that the activation of Src is a vital signaling event facilitating the Rb–Raf-1 interaction (Figure 3B). We observed that nicotine induces phosphorylation of EGFR starting at 4 hours (data not shown), whereas the Rb–Raf-1 interaction precedes this activation; this suggests that induction of EGFR is a secondary event that follows the binding of Raf-1 to Rb.

Figure 3

The mitogenic activity of nicotine requires Src. (A) Nicotine-induced cell proliferation was ablated by 1 μM of the Src inhibitor PD180. The EGFR inhibitor Iressa (1 μM) partially abrogated the mitogenic effects of nicotine, whereas 1 μM nifedipine had no effect. (B) Treatment with 1 μM PD180 ablated Rb–Raf-1 interaction in A549 cells, whereas nifedipine and Iressa had no effect. (C) Nicotine induced phosphorylation of Src in A549 cells. (D) Densitometric analysis was performed to quantitate the results obtained by Western blotting. (E) Nicotine induced the binding of c-Src to β-nAChRs. Quiescent A549 cells were stimulated with 1 μM nicotine for 15 or 30 minutes, and the binding of c-Src to β-nAChRs was assessed by IP/Western blot analysis. (F) The binding between c-Src and nAChRs was inhibited by the Src inhibitor PP2. Quiescent A549 cells were stimulated by 1 μM nicotine for 15 minutes in the presence or absence of 1 μM PP2, and the c-Src–β-nAChR association was assessed by IP/Western blot analysis. (G) Transfection of dominant-negative (DN) Src inhibited Rb–Raf-1 binding, whereas WT or constitutively active Src (v-Src) did not. A549 cells were transfected with the indicated plasmids. Eighteen hours after transfection, cells were rendered quiescent and subsequently stimulated with 1 μM nicotine or serum for 2 hours. Western blot analysis shows the levels of the indicated proteins after transfection.

We also studied the role of Src in nicotine-induced cell proliferation. Stimulation with 1 μM nicotine upregulated the levels of phosphorylated Src in A549 cells, while overall levels of Src remained constant (Figure 3, C and D). Figure 3D represents the densitometric analysis of Western blots performed with independent sets of lysates. We next examined whether nicotine stimulation induces the binding of Src to nAChRs. Quiescent A549 cells were stimulated with 1 μM nicotine, and the binding of nAChR with Src was examined by IP/Western blot analysis (Figure 3E). Previous studies have implicated the binding of β-nAChR with Src in the regulation of neuromuscular junctions as well as acrosome reactions in non-neuronal cells, hence we conjectured that the β-nAChR was probably the subunit that physically interacted with Src in NSCLC cells (31, 32). Therefore, a pan–β-nAChR subunit antibody was used to IP quiescent and nicotine-stimulated A549 lysates, and the presence of Src was examined by Western blotting. Src was found to bind nAChRs within 15 minutes of nicotine stimulation. Interestingly, the nAChR-Src interaction was completely ablated by 1 μM of the Src inhibitor PP2, suggesting that the activation of Src is essential for its binding to nAChRs (Figure 3F). We repeated the IP/Western blot experiments with a pan monoclonal antibody to α-nAChR; Src was detected in the IPs of nicotine-stimulated lysates (data not shown). The α and β subunits of nAChR associate to form a pentamer, explaining the presence of Src in both the IPs.

The role of activated Src in inducing the Rb–Raf-1 proliferative pathway was further examined. A549 cells were transiently transfected with FLAG–Raf-1 and Myc-Rb in the presence of wt Src, dominant negative Src or constitutively activated Src (v-Src). Eighteen hours post transfection, cells were rendered quiescent and stimulated with 1 μM nicotine for 2 hours; binding of Raf-1 to Rb was assessed by IP/Western blot analysis. Figure 3G shows that the Rb–Raf-1 interaction is observed in cells transfected with WT and constitutively activated Src; however, dominant-negative Src inhibited the nicotine-induced Rb–Raf-1 binding. These results suggest that the activation of Src is a crucial signaling event mediating the nicotine-induced Rb–Raf-1 proliferative pathway.

β-arrestin–1 facilitates the recruitment of Src to nAChRs. We next sought to elucidate the molecular mechanisms by which Src is activated upon nicotine stimulation. Data from Luttrell et al. have shown that the scaffolding protein β-arrestin–1 binds to Src and recruits it to the β-adrenergic receptors (33, 34). Given that certain ligands can activate both β-adrenergic and nicotinic receptors, we examined whether β-arrestin–1 mediates the recruitment of Src to nAChRs. The effect of β-arrestin–1 on nicotine-induced proliferation was first assessed by siRNA techniques (35). It was found that transfection of β-arrestin–1 siRNA ablated nicotine-induced S-phase entry but had no effect on serum-induced proliferation of A549 cells (Figure 4A). The treatment of A549 with nicotine caused a significant upregulation (from 10% ± 0.6% to 80% ± 0.2%) of S-phase cells. The presence of β-arrestin–1 siRNA in nicotine-stimulated A549 cells abrogated cell proliferation (decrease in BrdU-positive cells from 80% ± 0.2% to 22% ± 1.3%). Additionally, β-arrestin–1 siRNA caused downregulation of activated Src and phosphorylated ERK in nicotine-stimulated A549 cells; levels of total ERK and total Src remained constant (Figure 4, B and C). The siRNA transfections were performed in duplicate and repeated twice each for both BrdU assays as well as immunoblotting experiments. The Western blotting experiments shown in Figure 4B were performed with independent sets of lysates, and Figure 4C represents the average of 2 experiments. The requirement for β-arrestin–1 in recruiting Src to nAChR was next examined. IP/Western blot analysis showed that Src was not associated with nAChRs in quiescent cells but physically interacted with β-nAChR within 15 minutes of nicotine stimulation (Figure 4D); transfection of β-arrestin–1 siRNA abrogated the β-nAChR–Src interaction, while a control siRNA had no effect. These results suggest that β-arrestin is necessary for mediating the binding of Src to nAChR upon nicotine stimulation.

Figure 4

β-arrestin–1 mediates Src activation upon nAChR signaling. (A) Transfection of β-arrestin–1 siRNA inhibited nicotine-induced, but not serum-induced, S-phase entry in A549 cells. A nontargeting siRNA sequence was used as a control for all experiments. (B) Western blot analysis showing the suppression of β-arrestin in A549 cells upon transfection of β-arrestin–1 siRNA. Ablation of β-arrestin–1 significantly reduced nicotine-induced phosphorylated Src and phosphorylated ERK levels, whereas levels of total Src and ERK remained constant. Western blotting for actin served as the loading control. (C) Densitometric analysis was performed to quantitate the results obtained by Western blotting. (D) Nicotine induced the binding of Src to β-arrestin in A549 cells. Transfection of β-arrestin–1 siRNA abrogated the binding of c-Src to β-arrestin, whereas a nontargeting siRNA sequence had no effect. (E) Nicotine-induced stimulation of quiescent A549 cells for 15 minutes led to the binding of β-arrestin–1 to Src and to β-nAChR. (F) The binding of β-arrestin to nAChRs seen in E was confirmed by IP using β-nAChR antibody and Western blotting for β-arrestin. (G) Transfection of β-arrestin–1 siRNA inhibited Rb–Raf-1 binding in nicotine-stimulated A549 lysates, whereas a non-targeting siRNA sequence had no effect. No effect of β-arrestin–1 siRNA was seen on serum-induced Rb–Raf-1 binding in A549 cells.

Since β-arrestin–1 is a scaffolding protein (34), we examined whether β-arrestin–1 physically interacts with Src as well as nAChRs. Quiescent and nicotine-stimulated A549 lysates were immunoprecipitated with a β-arrestin polyclonal antibody and immunoblotted for endogenous Src and β-nAChR. Figure 4E shows that Src and nAChR were detected in β-arrestin–1 IPs in nicotine-stimulated, but not quiescent, cells. These results were confirmed by performing the IP with a pan–β-nAChR monoclonal antibody and immunoblotting for β-arrestin. It was found that a significant amount of β-arrestin bound to nAChR within 15 minutes of nicotine stimulation (Figure 4F). The role of β-arrestin–1 in mediating the Rb–Raf-1 interaction in nicotine-stimulated A549 cells was next assessed. The transfection of β-arrestin–1 siRNA completely abrogated Rb–Raf-1 binding in nicotine-stimulated A549 cells, but not in serum-stimulated cells (Figure 4G). Taken together, these data suggest that β-arrestin–1 plays a major role in mediating the proliferative effects of nicotine by recruiting Src to the membrane-bound nAChRs, leading to its activation and cell proliferation.

While the genotoxic effects of tobacco carcinogens have long been recognized (1, 36), the contribution of tobacco components to oncogenesis by cell surface receptor signaling is relatively unexplored. Studies from Cooke’s laboratory showed that nicotine and NNK can promote cell proliferation and angiogenesis in a receptor-dependent manner, but the molecules mediating these events have not yet been fully identified (4, 5). Here we showed that nicotine functions like a growth factor and that nAChRs have significant proliferative effects despite lacking intrinsic tyrosine kinase activity. Thus nicotine could activate components of the cell cycle machinery, cause upregulation of cyclin E/cdk2 and cyclin D/cdk4 kinase activity, mediate dissociation of E2F1 from Rb, and induce Rb phosphorylation in NSCLC cells. Furthermore, ChIP assays demonstrated that nicotine caused the binding of E2F1 and Raf-1 to proliferative promoters. Thus the signaling pathways induced by nicotine in NSCLC cells resemble those involved in growth factor stimulation.

A schematic of mitogenic signaling in NSCLCs is shown Figure 5. In quiescent cells, there is no association of β-arrestin and Src to nAChRs; also, E2F1 bound to proliferative promoters is repressed by Rb bound to the promoter (Figure 5A). However, the binding of nicotine to nAChRs causes the recruitment of β-arrestin and Src to the nicotinic receptors, resulting in the activation of the MAPK and Rb–Raf-1 pathways (Figure 5B). These events cause the recruitment of E2F1, Rb, and Raf-1 on E2F-responsive proliferative promoters. The inactivation of Rb by Raf-1 facilitates the dissociation of Rb from the promoters, which — combined with the increased binding of E2F1 — causes transcription of S-phase genes and cell cycle progression.

Figure 5

Schematic depicting the proliferative signaling by nAChRs in NSCLC cells. (A) Quiescent NSCLC cells do not display any association among β-arrestin, Src, and nAChRs. Rb is hypophosphorylated in these resting cells and is associated with E2F transcription factors. Quiescent cells do not show any binding between Raf-1 and Rb. E2F is present on the proliferative promoters but is repressed by Rb. (B) Nicotine stimulation causes the assembly of oligomeric complexes involving β-arrestin, Src, and nAChRs, facilitating the activation of Src. This leads to the activation of Raf-1, which binds to Rb; activation of MAPK and cyclins/cdks also occurs. The activation of Src facilitates the binding of Raf-1 to Rb, and multimeric complexes containing Rb, Raf-1, and E2F1 occupy proliferative promoters. Sustained mitogenic signaling leads to the dissociation of Raf-1 and Rb, while E2F remains bound to the proliferative promoters facilitating S-phase entry. Ablation of α₇-nAChR subunits and β-arrestin–1, inhibition of Src activation, and disruption of the Rb–Raf-1 interaction all blocked nicotine-induced proliferation of NSCLC cells. TS, thymidylate synthase.

Our earlier studies have shown that the binding of the signaling kinase Raf-1 to the Rb protein was a necessary step for cell cycle progression induced by serum and growth factors (25, 26). Binding of Raf-1 to Rb was necessary for its subsequent inactivation by cyclin/cdks; inhibition of the binding of Raf-1 prevented Rb phosphorylation, cell cycle progression, angiogenic tubule formation, and tumor growth in nude mice (26). Exposure of cells to nicotine induces the same interaction, facilitating cell proliferation. Our finding that the levels of Rb–Raf-1 complexes in human NSCLC tumors were greater than those in adjacent normal tissues suggests that the Rb–Raf-1 pathway probably contributes to oncogenesis; the presence of increased levels of Raf-1 on proliferative promoters in human NSCLC tumors supports this hypothesis. Thus it is likely that tumors exposed to nicotine have an added proliferative advantage; indeed, smokers have been found to respond less effectively to chemotherapy and were found to have increased metastasis of breast cancers (37–39). Our recent studies have shown that nicotine by itself could protect NSCLC cells from apoptosis induced by chemotherapeutic drugs by upregulating X-linked inhibitor of apoptosis and survivin (40). Cigarette smoke extract has been shown to promote the growth of gastric, colon, and bladder cancer and to upregulate the invasion of human gastric cancer (13, 16, 17, 41); our unpublished results show that nicotine by itself can promote the proliferation of cell lines derived from pancreatic, breast, and glial tumors as well. It can be imagined that signaling through the nAChRs contributes to these processes.

It appears that β-arrestin–1 and Src kinase are key players in mediating the mitogenic effects of nicotine. The Src family of protein tyrosine kinases has been found to be a critical component of multiple receptor-mediated signaling pathways that regulate proliferation, survival, metastasis, and angiogenesis. Src has been shown to functionally interact with several receptors, like the G protein–coupled receptors (GPCRs), which lack intrinsic tyrosine kinase activity and connect them to growth regulatory pathways like the MAPK cascade (33, 42). We found that the activation of Src is an important upstream signaling event that mediates Rb–Raf-1 binding upon nicotine stimulation. Studies in neuronal systems as well as nAChR-transfected cells have shown that the interaction of Src and its family members regulate the function of neuromuscular junctions and amplitude of whole-cell inward currents (17, 19, 29, 32). Recent evidence also shows that nicotine and NNK cause Src activation in lung cancer cells (19, 29). These observations lend support for a role for Src in smoking-related cancers.

Interestingly, β-arrestin seems to be vital for nicotine-mediated activation of Src and cell proliferation. The proteins in the β-arrestin family have been shown to serve as scaffold proteins that recruit a broad spectrum of signaling molecules to membrane-bound receptors in a highly cocoordinated manner. β-Arrestins are known to regulate endocytotic pathways, connect GPCRs to MAPK pathways, regulate chemotactic migration and motility, and contribute to antiapoptotic signaling (33, 34, 43). Recent studies have implicated the β-arrestin family of proteins in Notch signaling, TLR–interleukin-1 signaling, and regulation of gene expression by facilitating histone acetylation (44–47). Data from Ahn et al. and Fong et al. show that β-arrestin–knockout mice and β-arrestin–1 siRNA–transfected cells demonstrate impaired ERK activation and defective chemotactic responses (35, 48). β-Arrestin–GPCR complexes were also found to be key players in determining the mechanism and functional consequence of ERK activation (49). In addition, recent studies have shown that β-arrestin–1 plays a major role in the metastasis of colorectal cancer (41). Our data showed that β-arrestin is required for nAChR-mediated activation of the MEK/ERK pathway and proliferation of NSCLCs. It appears that β-arrestin acts as a critical link between the mitogenic signals from the nAChRs and the cell cycle machinery, which raises the possibility that nicotine might be promoting tumor progression and metastasis by activating β-arrestin–1. Indeed, it has been reported that there is an association between smoking and increased risk of lung metastasis in breast cancer patients (38, 39).

While tobacco carcinogens can initiate and promote tumorigenesis, the results of the present study raise the possibility that exposure to nicotine, by either cigarette substitutes or nicotine supplements, might confer a proliferative advantage to tumors already initiated. This contention is supported by the findings that nicotine can prevent apoptosis induced by various agents, thus conferring a survival advantage as well (10, 50–52). Elucidation of the signaling events mediated by nAChRs present on non-neuronal cells may open new avenues for targeting cancer therapy, focusing on agents that inhibit the β-arrestin–Src signaling axis or Rb–Raf-1 interaction.

We thank R. Jove, E. Haura, K.L. Guan, G. Bepler, A. Sharma, and R. Lefkowitz for generous gifts of constructs and reagents. This work is supported by NIH grant CA63136 from the National Cancer Institute to S. Chellappan.

Address correspondence to: Srikumar Chellappan, Department of Interdisciplinary Oncology, H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, 12902 Magnolia Drive, Tampa, Florida 33612, USA. Phone: (813) 745-6892; Fax: (813) 745-6748; E-mail: ChellaSP@moffitt.usf.edu .

Nonstandard abbreviations used: ChIP, chromatin IP; DHβE, dihydro β-erythoidine; EGM, endothelial growth medium; GPCR, G protein–coupled receptor; HAEC, human aortic endothelial cell; HMEC-L, human microvascular endothelial cell from lung; MAA, methylallyl aconitine; MEK, MAPK kinase; nAChR, nicotinic acetylcholine receptor; NHBE, normal human bronchial epithelial cell; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNN, N′-nitrosonornicotine; NSCLC, non–small cell lung cancer; SAEC, small airway epithelial cell.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J. Clin. Invest.116:2208–2217 (2006). doi:10.1172/JCI28164.

1. Hecht, S.S. 1999. DNA adduct formation from tobacco-specific N-nitrosamines. Mutat. Res. 424:127-142.
   View this article via: PubMed Google Scholar
2. Schuller, H.M., Orloff, M. 1998. Tobacco-specific carcinogenic nitrosamines. Ligands for nicotinic acetylcholine receptors in human lung cancer cells. Biochem. Pharmacol. 55:1377-1384.
   View this article via: PubMed CrossRef Google Scholar
3. Gotti, C., Clementi, F. 2004. Neuronal nicotinic receptors: from structure to pathology. Prog. Neurobiol. 74:363-396.
   View this article via: PubMed CrossRef Google Scholar
4. Heeschen, C., et al. 2001. Nicotine stimulates angiogenesis and promotes tumor growth and atherosclerosis. Nat. Med. 7:833-839.
   View this article via: PubMed CrossRef Google Scholar
5. Heeschen, C., Weis, M., Aicher, A., Dimmler, S., Cooke, J.P. 2002. A novel angiogenic pathway mediated by non-neuronal nicotinic acetylcholine receptors. J. Clin. Invest. 110:527-536.
   View this article via: JCI CrossRef PubMed Google Scholar
6. Sharma, G., Vijayaraghavan, S. 2002. Nicotinic receptor signaling in nonexcitable cells. J. Neurobiol. 53:524-534.
   View this article via: PubMed CrossRef Google Scholar
7. Wessler, I., Kirkpatrick, C.J., Racke, K. 1998. Non-neuronal acetylcholine, a locally acting molecule, widely distributed in biological systems: expression and function in humans. Pharmacol. Ther. 77:59-79.
   View this article via: PubMed CrossRef Google Scholar
8. Minna, J.D., Fong, K., Zochbauer-Muller, S., Gazdar, A.F. 2002. Molecular pathogenesis of lung cancer and potential translational applications. Cancer J. 8(Suppl. 1):S41-S46.
   View this article via: PubMed Google Scholar
9. Villablanca, A.C. 1998. Nicotine stimulates DNA synthesis and proliferation in vascular endothelial cells in vitro. J. Appl. Physiol. 84:2089-2098.
   View this article via: PubMed Google Scholar
10. Manackjee, R., Minna, J.D. 1990. Opioid and nicotine receptors affect growth regulation of human lung cancer cell lines. Proc. Natl. Acad. Sci. U. S. A. 87:3294-3298.
    View this article via: PubMed CrossRef Google Scholar
11. Cattaneo, M.G., D’Atri, F., Vicentini, L.M. 1997. Mechanisms of mitogen-activated protein kinase activation by nicotine in small-cell lung carcinoma cells. Biochem. J. 328:499-503.
    View this article via: PubMed Google Scholar
12. Minna, J.D. 2003. Nicotine exposure and bronchial epithelial cell nicotinic acetylcholine receptor expression in the pathogenesis of lung cancer. J. Clin. Invest. 111:31-33.
    View this article via: JCI CrossRef PubMed Google Scholar
13. Trombino, S., et al. 2004. Role of the non-neuronal human cholinergic system in lung cancer and mesothelioma: possibility of new therapeutic strategies. Curr. Med. Chem. Anticancer Agents. 4:535-542.
    View this article via: PubMed CrossRef Google Scholar
14. Melkonian, G., Cheung, L., Marr, R., Tong, C., Talbot, P. 2002. Mainstream and sidestream cigarette smoke inhibit growth and angiogenesis in the day 5 chick chorioallantoic membrane. Toxicol. Sci. 68:237-248.
    View this article via: PubMed CrossRef Google Scholar
15. Song, P., et al. 2003. Acetylcholine is synthesized by and acts as an autocrine growth factor for small cell lung carcinoma. Cancer Res. 63:214-221.
    View this article via: PubMed Google Scholar
16. Shin, V.Y., Cho, C.H. 2005. Nicotine and gastric cancer. Alcohol. 35:259-264.
    View this article via: PubMed CrossRef Google Scholar
17. Ye, Y.N., et al. 2004. Nicotine promoted colon cancer growth via epidermal growth factor receptor, c-Src, and 5-lipoxygenase-mediated signal pathway. J. Pharmacol. Exp. Ther. 308:66-72.
    View this article via: PubMed CrossRef Google Scholar
18. Conklin, B.S., Zhao, W., Zhong, D., Chen, C. 2002. Nicotine and cotinine upregulate vascular endothelial growth factor expression in endothelial cells. Am. J. Pathol. 160:413-419.
    View this article via: PubMed Google Scholar
19. Xu, L., Deng, X. 2005. Protein kinase Ciota promotes nicotine-induced migration and invasion of cancer cells via phosphorylation of micro- and m-calpains. J. Biol. Chem. 281:4457-4466.
    View this article via: PubMed CrossRef Google Scholar
20. Shin, V.Y., et al. 2005. Nicotine induces cyclooxygenase-2 and vascular endothelial growth factor receptor-2 in association with tumor-associated invasion and angiogenesis in gastric cancer. Mol. Cancer Res. 3:607-615.
    View this article via: PubMed CrossRef Google Scholar
21. Jull, B.A., Plummer, H.K., Schuller, H.M. 2001. Nicotinic receptor-mediated activation by tobacco-specific nitrosamine NNK of a Raf-1/MAPkinase pathway, resulting in phosphorylation of c-myc in human small cell lung carcinoma and pulmonary neuroendocrine cells. J. Cancer Res. Clin. Oncol. 127:707-717.
    View this article via: PubMed Google Scholar
22. Chu, M., Guo, J., Chen, C.Y. 2005. Long-term exposure to nicotine, via ras pathway, induces cyclin D1 to stimulate G1 cell cycle transition. J. Biol. Chem. 280:6369-6379.
    View this article via: PubMed CrossRef Google Scholar
23. Nevins, J.R. 2001. The Rb/E2F pathway and cancer. Hum. Mol. Genet. 10:699-703.
    View this article via: PubMed CrossRef Google Scholar
24. Stevaux, O., Dyson, N.J. 2002. A revised picture of E2f Transcriptional regulation and Rb function. Curr. Opin. Cell Biol. 14:684-691.
    View this article via: PubMed CrossRef Google Scholar
25. Wang, S., Ghosh, R.N., Chellappan, S.P. 1998. Raf-1 physically interacts with Rb and regulates its function: a link between mitogenic signaling and cell cycle regulation. Mol. Cell. Biol. 18:7487-7498.
    View this article via: PubMed Google Scholar
26. Dasgupta, P., et al. 2004. Disruption of the Rb-Raf-1 interaction inhibits tumor growth and angiogenesis. Mol. Cell. Biol. 24:9527-9541.
    View this article via: PubMed CrossRef Google Scholar
27. Johnson, B.E. 1998. Tobacco and lung cancer. Prim. Care. 25:279-291.
    View this article via: PubMed Google Scholar
28. Sekhon, H.S., Wright, J.L., Churg, A. 1994. Cigarette smoke causes rapid cell proliferation in small airways and associated pulmonary arteries. Am. J. Physiol. 267:L557-L563.
    View this article via: PubMed Google Scholar
29. Charpantier, E., et al. 2005. Alpha7 neuronal nicotinic acetylcholine receptors are negatively regulated by tyrosine phosphorylation and Src-family kinases. J. Neurosci. 25:9836-9849.
    View this article via: PubMed CrossRef Google Scholar
30. Sher, E., et al. 1998. Nicotinic receptors and calcium channels in small cell lung carcinoma. Functional role, modulation, and autoimmunity. Ann. N. Y. Acad. Sci. 841:606-624.
    View this article via: PubMed CrossRef Google Scholar
31. Colledge, M., Froehner, S.C. 1998. Interaction between the nicotinic acetylcholine receptor and Grb2. Implications for signaling at the neuromuscular junction. Ann. N. Y. Acad. Sci. 841:17-27.
    View this article via: PubMed CrossRef Google Scholar
32. Kumar, P., Meizel, S. 2005. Nicotinic acetylcholine receptor subunits and associated proteins in human sperm. J. Biol. Chem. 280:25928-25935.
    View this article via: PubMed CrossRef Google Scholar
33. Luttrell, L.M., et al. 1999. Beta-arrestin-dependent formation of beta2 adrenergic receptor-Src protein kinase complexes. Science. 283:655-661.
    View this article via: PubMed CrossRef Google Scholar
34. Lefkowitz, R.J., Shenoy, S.K. 2005. Transduction of receptor signals by beta-arrestins. Science. 308:512-517.
    View this article via: PubMed CrossRef Google Scholar
35. Ahn, S., Nelson, C.D., Garrison, T.R., Miller, W.E., Lefkowitz, R.J. 2003. Desensitization, internalization, and signaling functions of beta-arrestins demonstrated by RNA interference. Proc. Natl. Acad. Sci. U. S. A. 100:1740-1744.
    View this article via: PubMed CrossRef Google Scholar
36. Hecht, S.S., et al. 2002. Inhibition of lung tumorigenesis in A/J mice by N-acetyl-S-(N-2-phenethylthiocarbamoyl)-L-cysteine and myo-inositol, individually and in combination. Carcinogenesis. 23:1455-1461.
    View this article via: PubMed CrossRef Google Scholar
37. Richardson, G.E., et al. 1993. Smoking cessation after successful treatment of small-cell lung cancer is associated with fewer smoking-related second primary cancers. Ann. Intern. Med. 119:383-390.
    View this article via: PubMed Google Scholar
38. Murin, S., Inciardi, J. 2001. Cigarette smoking and the risk of pulmonary metastasis from breast cancer. Chest. 119:1635-1640.
    View this article via: PubMed CrossRef Google Scholar
39. Murin, S., Pinkerton, K.E., Hubbard, N.E., Erickson, K. 2004. The effect of cigarette smoke exposure on pulmonary metastatic disease in a murine model of metastatic breast cancer. Chest. 125:1467-1471.
    View this article via: PubMed CrossRef Google Scholar
40. Dasgupta, P., et al. 2006. Nicotine inhibits apoptosis induced by chemotherapeutic drugs by up-regulating XIAP and survivin. Proc. Natl. Acad. Sci. U. S. A. 103:6332-6337.
    View this article via: PubMed CrossRef Google Scholar
41. Buchanan, F.G., et al. 2006. Role of beta-arrestin 1 in the metastatic progression of colorectal cancer. Proc. Natl. Acad. Sci. U. S. A. 103:1492-1497.
    View this article via: PubMed CrossRef Google Scholar
42. Ishizawar, R., Parsons, S.J. 2004. c-Src and cooperating partners in human cancer. Cancer Cell. 6:209-214.
    View this article via: PubMed CrossRef Google Scholar
43. Huang, J., Sun, Y., Huang, X.Y. 2004. Distinct roles for Src tyrosine kinase in beta2-adrenergic receptor signaling to MAPK and in receptor internalization. J. Biol. Chem. 279:21637-21642.
    View this article via: PubMed CrossRef Google Scholar
44. Mukherjee, A., et al. 2005. Regulation of Notch signalling by non-visual beta-arrestin. Nat. Cell Biol. 7:1191-1201.
    View this article via: PubMed Google Scholar
45. Kang, J., et al. 2005. A nuclear function of beta-arrestin1 in GPCR signaling: regulation of histone acetylation and gene transcription. Cell. 123:833-847.
    View this article via: PubMed CrossRef Google Scholar
46. Beaulieu, J.M., Caron, M.G. 2005. Beta-arrestin goes nuclear. Cell. 123:755-757.
    View this article via: PubMed CrossRef Google Scholar
47. Wang, Y., et al. 2006. Association of beta-arrestin and TRAF6 negatively regulates Toll-like receptor-interleukin 1 receptor signaling. Nat. Immunol. 7:139-147.
    View this article via: PubMed CrossRef Google Scholar
48. Fong, A.M., et al. 2002. Defective lymphocyte chemotaxis in beta-arrestin2- and GRK6-deficient mice. Proc. Natl. Acad. Sci. U. S. A. 99:7478-7483.
    View this article via: PubMed CrossRef Google Scholar
49. Togho, A., et al. 2003. The stability of the G protein-coupled receptor-beta-arrestin interaction determines the mechanism and functional consequence of ERK activation. J. Biol. Chem. 278:6258-6267.
    View this article via: PubMed CrossRef Google Scholar
50. Mai, H., May, W.S., Gao, F., Jin, Z., Deng, X. 2003. A functional role for nicotine in Bcl2 phosphorylation and suppression of apoptosis. J. Biol. Chem. 278:1886-1891.
    View this article via: PubMed CrossRef Google Scholar
51. Trombino, S., et al. 2004. Alpha7-nicotinic acetylcholine receptors affect growth regulation of human mesothelioma cells: role of mitogen-activated protein kinase pathway. Cancer Res. 64:135-145.
    View this article via: PubMed CrossRef Google Scholar
52. Xin, M., Deng, X. 2005. Nicotine inactivation of the proapoptotic function of Bax through phosphorylation. J. Biol. Chem. 280:10781-10789.
    View this article via: PubMed CrossRef Google Scholar
53. Matsushime, H., et al. 1994. D-type cyclin-dependent kinase activity in mammalian cells. Mol. Cell. Biol. 14:2066-2076.
    View this article via: PubMed Google Scholar
54. Wang, S., Nath, N., Minden, A., Chellappan, S. 1999. Regulation of Rb and E2F by signal transduction cascades: divergent effects of JNK1 and p38 kinases. Embo J. 18:1559-1570.
    View this article via: PubMed CrossRef Google Scholar
55. Derossi, D., Joliot, A.H., Chassaing, G., Prochiantz, A. 1994. The third helix of the Antennapedia homeodomain translocates through biological membranes. J. Biol. Chem. 269:10444-10450.
    View this article via: PubMed Google Scholar
56. Wang, S., Fusaro, G., Padmanabhan, J., Chellappan, S.P. 2002. Prohibitin co-localizes with Rb in the nucleus and recruits N-CoR and HDAC1 for transcriptional repression. Oncogene. 21:8388-8396.
    View this article via: PubMed CrossRef Google Scholar
57. Boyd, K.E., Farnham, P.J. 1999. Coexamination of site-specific transcription factor binding and promoter activity in living cells. Mol. Cell. Biol. 19:8393-8399.
    View this article via: PubMed Google Scholar
58. Forsberg, E.C., Downs, K.M., Bresnick, E.H. 2000. Direct interaction of NF-E2 with hypersensitive site 2 of the beta-globin locus control region in living cells. Blood. 96:334-339.
    View this article via: PubMed Google Scholar