A Central Role for Nicotinic Cholinergic Regulation of Growth Factor-Induced Endothelial Cell Migration

【摘要】  Objective- An endothelial nicotinic acetylcholine receptor (nAChR) participates in atherogenesis and tumorigenesis by promoting neovascularization. To date, the mechanisms of nAChR-mediated angiogenesis and their relationship to angiogenic factors, eg, VEGF and bFGF, are unknown.

Methods and Results- Nicotine induced dose-dependent human microvascular endothelial cell (HMVEC) migration, a key angiogenesis event, to an extent which was equivalent in magnitude to bFGF (10 ng/mL) but less than for VEGF (10 ng/mL). Unexpectedly, nAChR antagonism not only abolished nicotine-induced HMVEC migration but also abolished migration induced by bFGF and attenuated migration induced by VEGF. Transcriptional profiling identified gene expression programs which were concordantly regulated by all 3 angiogens (nicotine, VEGF, and bFGF), a notable feature of which includes corepression of thioredoxin-interacting protein (TXNIP), endogenous inhibitor of the redox regulator thioredoxin. Furthermore, TXNIP repression by all 3 angiogens induced thioredoxin activity. Silencing thioredoxin by small interference RNA abrogated all angiogen-induced migration while silencing TXNIP strongly induced HMVEC migration. Interestingly, nAChR antagonism abrogates growth factor (VEGF and bFGF)-mediated induction of thioredoxin activity.

Conclusions- Nicotine promotes angiogenesis via stimulation of nAChR-dependent endothelial cell migration. Furthermore, growth factor-induced HMVEC migration, a key angiogenesis event, requires nAChR activation-an effect mediated in part by nAChR-dependent regulation of thioredoxin activity.

Nicotine induces nicotinic acetylcholine receptor (nAChR)-dependent endothelial cell migration (ECM). nAChR antagonism also suppresses bFGF- and VEGF-induced ECM. ECM induced by nicotine, VEGF, or bFGF requires nAChR-dependent induction of thioredoxin activity via repression of the thioredoxin inhibitor, TXNIP. Therefore growth factor-induced ECM, a key angiogenesis event, requires nAChR activation.

【关键词】  nicotine angiogenesis endothelium vascular endothelial growth factor fibroblast growth factor

Introduction

The nicotinic acetylcholine receptor (nAChR) is a pentameric ligand-gated cationic channel. 1 The nAChR was first described in neurons, but has recently been identified in many cell types including endothelial cells (ECs) and vascular smooth muscle cells. 2 Intriguingly, ECs also synthesize and store acetylcholine. 3 Recently, we serendipitously discovered that nAChR activation causes ECs to form capillary tubes in vitro, and promotes angiogenesis in vivo. 4,5

Pathological as well as physiological forms of angiogenesis are mediated by EC nAChRs. For example, by activating the EC nAChR, nicotine accelerates tumor angiogenesis and tumor growth in a murine Lewis lung cancer model. 4 The acceleration of tumor growth by environmental tobacco smoke is also mediated by nAChR-induced angiogenesis. 6 Furthermore, nAChR activation by nicotine stimulates the neovascularization and progression of atherosclerotic plaque. 4 On the other hand, activation of the nAChR in a murine model of diabetic ulceration enhances wound angiogenesis and healing. 7

To date, the mechanisms of nAChR-mediated angiogenesis and their relationship to established angiogenic growth factors, such as VEGF and bFGF, are unknown. We therefore sought to study and compare the effects of nAChR activation on EC migration, a key event in angiogenesis, alongside those induced by VEGF or bFGF. In this article, we report an unexpected observation: pharmacological antagonism of the nAChR fully blocks bFGF-induced EC migration, and substantially suppresses the endothelial response to VEGF. Furthermore, by microarray analysis, we identify gene expression programs which are concordantly regulated by nicotine, VEGF, or bFGF, and confirm the role of one of these genes in the cholinergic component of growth factor-induced endothelial cell migration.

Methods

For Methods, please refer to the Data Supplement, available online at http://atvb.ahajournals.org.

Results

A Cholinergic Component of Growth Factor-Mediated Endothelial Cell Migration

The effects of nicotine on human microvascular endothelial cell (HMVEC) migration were studied using standard wounding migration assays. Figure 1 A depicts typical microphotographs of in vitro HMVEC migration in the presence of vehicle or nicotine. Nicotine stimulated EC migration in a dose-dependent manner with maximal stimulation at 10 -8 mol/L, producing cell migration that was 216±9% that of vehicle-treated ECs ( P <0.001 versus control; Figure 1 B). Stimulation with VEGF or bFGF also induced EC migration with maximal effects at 10 ng/mL (data not shown). Nicotine-induced EC migration at 10- 8 mol/L was equivalent in magnitude to that observed for bFGF (10 ng/mL) but less than for VEGF (10 ng/mL; P =NS for nicotine versus bFGF; P <0.01 for nicotine versus VEGF; Figure 1 C). To further investigate the effects of nAChR-dependent pathways on EC migration, we studied the effect of the nAChR antagonist, hexamethonium (HEX, 10 -4 mol/L), on EC migration induced by nicotine, VEGF, and bFGF. Hexamethonium abrogated nicotine-induced EC migration ( Figure 1 C). Unexpectedly, in addition to abolishing NIC-induced EC migration, HEX abolished migration induced by bFGF ( P =NS versus control) and reduced migration induced by VEGF ( P <0.01 for VEGF+HEX versus VEGF alone; Figure 1 C). Similar results were observed with the nAChR antagonist, mecamylamine (10 -6 mol/L; supplemental Figure I). The nAChR-related effect was dose-dependent: in the case of VEGF, cell migration was attenuated by 26±12%, 43±9%, and 52±12% by HEX concentrations of 10 -8 mol/L, 10 -6 mol/L, and 10 -4 mol/L, respectively ( P <0.05 for trend). Nicotinic receptor activation is known to stimulate EC proliferation. 4 Interestingly, nAChR antagonism also significantly attenuated VEGF- and bFGF-mediated EC proliferation as assessed by bromodeoxyuridine incorporation (supplemental Figure II). The nAChR antagonist-related effects were not the result of cellular toxicity as addition of hexamethonium or mecamylamine alone did not induce cell death as examined via 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide (MTT) assays, nor did mecamylamine or hexamethonium induce apoptosis in VEGF- or bFGF-treated cells as assessed by annexin V staining (data not shown).

Figure 1. Role of nAChR in human microvascular endothelial cell (HMVEC) migration. A, Representative microphotographs of the HMVEC migration assay: after wounding of the monolayer with a razor blade, HMVECs migrate into the denuded area. B, Nicotine stimulates dose-dependent HMVEC migration which is maximal at a nicotine concentration of 10 -8 mol/L ( P <0.001 vs control). C, Effects of hexamethonium (HEX), a nAChR antagonist, on HMVEC migration induced by nicotine (10 -8 mol/L), VEGF (10 ng/mL), and bFGF (10 ng/mL). Values are expressed as a percentage of migrating cells per high-powered field in vehicle-treated wells. ** P <0.001.

Previous in vitro and in vivo data have implicated a central role for the 7-nAChR isoform in mediating nAChR-induced neovascularization. 5 Consistent with this, selective inhibition of the 7-nAChR isoform by -bungarotoxin (10 -9 mol/L) abrogated nicotine effects ( P =NS versus control) and significantly attenuated bFGF and VEGF-induced EC migration by 50±11% and 38±13%, respectively ( P <0.05 for stimulus versus stimulus + -bungarotoxin; Figure 2 A). We also studied HMVEC 7-nAChR mRNA expression in response to nicotine, VEGF, and bFGF over a 24-hour time course. 10-fold downregulation of 7-nAChR expression compared with vehicle-treated conditions ( P <0.05 versus control for all angiogens; Figure 2 B). However, by 6 hours, there was significant upregulation of 7-nAChR expression by nicotine (12±0.8-fold), VEGF (11±1.3-fold), and bFGF (4±0.2-fold; P <0.001 versus control; Figure 2 B). By 24 hours, HMVEC 7-nAChR expression in all conditions had returned to levels similar to control ( P =NS versus control). These data suggest that nicotine, VEGF, and bFGF induce acute stimulation of 7-nAChR with subsequent early downregulation of 7-nAChR expression by negative feedback followed by upregulation at 6 hours. These findings are consistent with a role for nAChR in growth factor signaling pathways.

Figure 2. Role of the 7-nAChR isoform in growth factor-mediated EC migration. A, Effects of -bungarotoxin ( -BT), a specific inhibitor of the 7-nAChR isoform, on EC migration. Coadministration of -BT (10 -9 mol/L) abrogated the migragenic effects of nicotine (10 -8 mol/L; P =NS vs control) and significantly attenuated the effects of bFGF (10 ng/mL) and VEGF (10 ng/mL). B, Effects of nicotine, VEGF, and bFGF on EC 7-nAChR expression at 1, 6, and 24 hours as assessed by relative real-time RT-PCR. * P <0.05; ** P <0.001 compared with control.

Identification of Shared Transcriptional Responses to Nicotine, bFGF, and VEGF by Microarray Analysis

To further evaluate a cholinergic contribution to growth factor-induced EC migration, and to identify commonly regulated genes that may be required for HMVEC migration, we performed microarray analysis of HMVECs after exposure to nicotine, VEGF, or bFGF. At 24 hours after treatment, each of these stimuli induced profound transcriptional changes in HMVECs ( Figure 3 ), resulting in differential expression of a total of 3072 genes uniquely identified by UniGene, as well as 312 expressed-sequence tags, all of which were represented by 4070 nonredundant cDNA clones. To study relationships between gene expression programs induced by nicotine, VEGF, or bFGF, data for all differentially expressed genes at 24 hours were hierarchically clustered by gene and by array, thereby organizing genes and experimental samples on the basis of similarity of expression patterns ( Figure 3A through 3 F). 8 The cluster dendrogram shows that all 3 stimuli induced distinct transcriptional signatures which cluster within 3 distinct groups, but there is a closer relationship between the VEGF- and bFGF-induced expression profiles, which cluster together on the same dendrogram branch ( Figure 3 A).

Figure 3. Hierarchical cluster analysis of transcriptional effects of nicotine, VEGF, or bFGF in human microvascular endothelial cells at 24 hours. A, Overview of the two way (genes against conditions) hierarchical cluster of 15 experiments (each condition was studied in quintuplicate) and 4070 nonredundant cDNA clones with significant change in expression at 24 hours. Data from individual elements or genes are represented in rows and experiments in columns. Red and green denote expression levels greater or less, respectively, than control values. Gray denotes technically inadequate or missing data. The intensity of the color reflects the magnitude of the change from baseline. The dendrogram above the matrix represents similarities in patterns of expression between experimental samples. B through G, Zoom boxes of concordantly expressed gene clusters, whose location are indicated by vertical colored bars adjacent to the dendrogram. Owing to space limitations, only genes discussed in the text are indicated by UniGene symbol.

Within the distinct transcriptional profiles induced by nicotine, VEGF, or bFGF, we identified 6 clusters with concordant gene expression (3 clusters of commonly activated and 3 commonly corepressed genes; Figure 3B through 3 F). The characteristics of these clusters provide insights into shared cellular processes that may be requisite for angiogen-induced cell migration. The first activation cluster ( Figure 3 B), the "migration cluster," was enriched for genes associated with cytokinetic processes including migration-associated G protein signaling (Rho GTPase regulatory proteins and RIN2), integrin binding (ERBB2IP and ADAM9), cell cycle regulation and proliferation (RRM2, MDM2, AHR, MLLT4, and MUTYH), NF- B activation (BCL10 and CASP8AP2), and migration-associated oxidoreductase activity (LOX and ASPH). Significantly, three Rho GTPase activating proteins (GAPs) including ARHGAP5, ARHGAP21, and ARHGAP24 and one Rho guanine nucleotide exchange factor (GEF), ARHGEF7, were concordantly upregulated in this cluster. Rho GEFs and GAPs, by respectively controlling the activation and inactivation of small Rho GTPases (Cdc42 and Rac), regulate the orchestration of cytoskeletal and adhesive changes during cytokinesis. 9

A smaller second coactivation cluster ( Figure 3 C) includes the p21-activated kinase PAK1, an effector for the Rho GTPases Rac and Cdc42, that facilitates cell migration by coordinating formation of new adhesions at the leading edge of the cell with detachment at the trailing edge. 10 Other genes in this activation cluster comprise zinc finger proteins and genes involved in nucleic acid metabolism. Interestingly, all three angiogens induced activation of the VEGF receptor, FLT1, an effect that was stronger for bFGF-treated cells than for nicotine or VEGF ( Figure 3 D, which contains 3 nonredundant cDNA clones for FLT1). In addition to the coinduction of FLT1 by all three stimuli, we found that several isoforms of nAChR subunits are upregulated by VEGF at 24 hours (supplemental Table I), suggesting other potential synergistic interactions between VEGF and cholinergic signaling pathways.

The first repression cluster ( Figure 3 E) contains genes that are strongly downregulated by bFGF, many of which are also concordantly repressed by nicotine and VEGF. A dominant theme among concordantly repressed genes is the downregulation of chemokine genes (principally of CC class) involved leukocyte chemotaxis (CCL2, CCL7, CCL8, CCL20, and CX3CL). Another prominent feature is the robust repression of thioredoxin interacting protein (TXNIP) ( Figure 3 E), a protein that binds and inhibits thioredoxin, a major intracellular antioxidant. Other corepressed genes in this cluster have been implicated in apoptosis (TNFRSF1B, EP300), signal transduction (CD53, SQSTM1), and cell adhesion (ICAM1). The second repression cluster ( Figure 3 F) included two Ephrin receptors: EPHB4, a marker of venous differentiation and EPHA2, an inhibitor of cell migration which suppresses integrin function. Other genes within this group were associated with tumor suppression, microtubular polymerization, and signal transduction. The striking feature of the third cluster of corepressed genes is a strong enrichment for metallothioneins (MT1E, MT1F, MT1G, MT1L, MT1X, MT2A, and MT3; Figure 3 G). Metallothioneins (MTs) comprise a superfamily of small cysteine-rich proteins with high affinity for metal ions and antioxidant activity. By serving as a cellular reservoir for zinc and copper, MTs regulate the function of proteins requiring these metals such as DNA and RNA polymerases, zinc finger transcription factors, and p53. 11 Other genes in this cluster are involved in diverse functions including cell proliferation/apoptosis (MYC, ERCC2), lipid transport (PTLP), calcium ion transport (CACNA1D), and actin binding (COTL1).

In summary, the transcriptional signatures of nicotine, VEGF, and FGF, while distinct, demonstrate many overlapping features. By hierarchical cluster analysis, we have identified a series of shared angiogen-dependent EC transcriptional programs, with implications for understanding shared mechanisms in EC migration/angiogenesis.

Role of Thioredoxin Interacting Protein in the Cholinergic Contribution to Growth Factor-Induced EC Migration

As nAChR antagonism modulates VEGF- and bFGF-dependent EC migration, we hypothesized that some of the transcriptional effects shared by nicotine, VEGF, and bFGF may be nAChR-dependent. Thioredoxin interacting protein (TXNIP), a gene not previously associated with EC migration, was downregulated by all three angiogens. TXNIP is the endogenous inhibitor of thioredoxin. Thioredoxin is a major redox regulator of protein function increasingly implicated in tumorigenesis. 12,13 In microarray data from 3 nonredundant cDNA clones for TXNIP, nicotine, VEGF, and bFGF consistently decreased TXNIP expression-a finding confirmed by RT-PCR, which demonstrated decreased expression by 42±4%, 33±6%, and 26±7% relative to control, respectively ( P <0.001 for all stimuli; Figure 4 A). As in vivo reduction of TXNIP expression in the order of 30% to 40% has been associated 3-fold increases in thioredoxin activity, 14 we hypothesized that TXNIP downregulation may influence thioredoxin activity and play a role in angiogen-mediated EC migration.

Figure 4. Role of nAChR in growth factor-mediated regulation of TXNIP expression and thioredoxin activity in ECs. Effects of nicotine (10 -8 mol/L), VEGF (10 ng/mL), and bFGF (10 ng/mL) with/without coadministration of hexamethomium (10 -4 mol/L) on TXNIP expression (A) and on thioredoxin activity (B) in human microvascular endothelial cells. Values for thioredoxin activity are expressed as a percentage of control (vehicle-treated cells). ** P <0.01.

Using a standard assay for thioredoxin activity, 15 we found that addition of nicotine or VEGF induced thioredoxin activity significantly above vehicle-treated cells ( P <0.001 versus control for all stimuli; Figure 4 B). The addition of bFGF induced a less robust ( P <0.05 versus control) but significant increase in thioredoxin activity ( Figure 4 B). Notably, coadministration of hexamethonium inhibited nicotine-, VEGF-, or bFGF-induced thioredoxin activity ( Figure 4 B; P <0.001 for each stimulus versus stimulus + hexamethomium). Hexamethonium alone had no significant effect on thioredoxin activity. Consistent with these results, nAChR antagonism abrogated nicotine, VEGF, and bFGF-mediated repression of TXNIP mRNA expression ( P =NS versus control for each stimulus + hexamethonium; Figure 4 A). Transfection of small interference RNA (siRNA) against thioredoxin abrogated nicotine-, VEGF-, or bFGF-induced thioredoxin activity and abolished cell migration induced by nicotine, VEGF, or bFGF ( P =NS versus control for each stimulus + siRNA; Figure 5A and 5 B). Furthermore, in the absence of angiogenic stimuli, siRNA against TXNIP significantly stimulated thioredoxin activity ( P <0.0001 versus control; Figure 6 A) and strongly stimulated HMVEC migration ( Figure 6 B, P <0.0001 versus control). These studies indicate that inhibition of TXNIP by cholinergic or growth factor activation promotes endothelial cell migration, via derepression of thioredoxin activity. Furthermore, VEGF- or bFGF-mediated regulation of TXNIP expression is dependent on activation of nAChR ( Figure 4 A).

Figure 5. Effects of gene knockdown of thioredoxin by small interference RNA (siRNA) on induction of thioredoxin activity in human microvascular endothelial cells (HMVECs) by nicotine (10 -8 mol/L), VEGF(10 ng/mL), or bFGF (10 ng/mL) (A) and HMVEC migration induced by nicotine, VEGF, or bFGF (B). Scrambled (randomly arranged) siRNA had no effect on thioredoxin activity or HMVEC migration (data not shown). Values for thioredoxin activity and migration are expressed as a percentage of control (vehicle-treated cells). TXN indicates thioredoxin. ** P =0.01; *** P <0.001.

Figure 6. Effects of gene knockdown of thioredoxin interacting protein (TXNIP) by small interference RNA (siRNA) on thioredoxin activity (A) and human microvascular endothelial cell (HMVEC) migration (B). Scrambled (randomly arranged) siRNA had no effect on thioredoxin activity or HMVEC migration (data not shown). Values for thioredoxin activity and migration are expressed as a percentage of control (vehicle-treated cells). *** P <0.0001.

Discussion

We report a cholinergic contribution to growth factor-induced endothelial cell migration. The salient observations are that: (1) activation of nAChR induces EC migration similar in magnitude to that observed for bFGF or VEGF; (2) antagonism of nAChR markedly attenuates the migragenic effects of bFGF or VEGF on ECs; (3) the nAChR-dependent effects of bFGF and VEGF on EC migration are due, in large part, to activation of the 7-nAChR isoform; (4) nAChR activation induces a transcriptional profile that has many overlapping features to those induced by bFGF or VEGF, particularly for genes involved in EC migration; (5) downregulation of TXNIP with subsequent induction of thioredoxin activity is shown to be important to the migragenic effects exerted by each of the stimuli; (6) antagonism of the nAChR abrogates VEGF- or bFGF-mediated regulation of TXNIP expression. In toto, our findings identify a novel role for the nicotinic cholinergic pathway in growth factor-mediated EC migration, a critical event in angiogenesis.

Previous studies have demonstrated that ECs synthesize, store, and release acetylcholine 16 and express functional nAChRs. 2 Increasing evidence suggests that such nonneuronal nAChRs are involved in the regulation of vital cell functions, such as mitosis, differentiation, organization of the cytoskeleton, cell-cell contact, locomotion, and migration. 17 Thus, acetylcholine, originally identified as a neurotransmitter, may function as an autocrine factor that modulates migration of endothelial cells. We and others have previously shown that exogenous nicotine, at pathophysiologically relevant concentrations, promotes angiogenesis in a number of in vivo settings, including inflammation, wound healing, ischemia, tumor, and atherosclerosis. 4,7,18 Furthermore, inhibition of nAChR, in the absence of exogenous nicotine, reduces the angiogenic response in vitro and in vivo, indicating that there exists an endogenous cholinergic pathway for angiogenesis. 5 In contrast, recent work has shown that VEGF and FGF, originally identified as angiogenic growth factors, exert neurotrophic effects and promote neurogenesis. 19,20 These and other data suggest that there may be interdependence between "vascular" and "neuronal" factors and processes.

In this study, we found that nicotine induced dose-dependent nAChR-mediated EC migration which was maximal at concentrations consistent with those found in moderate smokers (10 -8 mol/L). Surprisingly, coadministration of nAChR antagonists, hexamethonium, or mecamylamine, significantly attenuated the migragenic response of ECs to both VEGF and bFGF. Although several nAChR isoforms exist, we have previously identified a principal role for the 7-nAChR isoform in nAChR-mediated angiogenesis in vitro and in vivo. 5 Consistent with these findings, we now find that VEGF and bFGF both induced changes in 7-nAChR expression consistent with activation of 7-nAChR. Moreover, the 7-nAChR selective antagonist, -bungarotoxin, attenuated VEGF and bFGF-induced EC migration to a similar extent as for the nonselective antagonists, hexamethonium, and mecamylamine. These latter findings suggest that nAChR-dependent pathways, particularly via 7-nAChR activation, are involved in the modulation of growth factor-induced EC migration.

To study the relationship between nicotine and angiogenic growth factors at a genomic level, microarray analysis was performed after HMVEC exposure to nicotine, VEGF, or bFGF. By hierarchical clustering, we found that each stimulus induced distinct but overlapping transcriptional responses, with concordant gene expression being concentrated within six largely functionally coherent gene clusters. A major functional theme among concordantly expressed genes was the coregulation of cell motility-related processes by all three angiogens. In particular, a "migration" cluster of concordantly activated genes was strongly enriched for genes involved in cytokinetic-related processes such as the Rho GTPase cell motility pathways, integrin binding, cell cycle regulation, and NF- B activation. Our findings with regard to activation of Rho GTPase pathways by VEGF and bFGF are consistent with previous studies 21 and reinforce the central role of Rho-related regulation of actomyosin cytoskeletal organization EC migration during angiogenesis. We previously demonstrated that nAChR-dependent endothelial tube formation in vitro is dependent on NF- B activation. 5 The migration cluster included two genes associated with NF- B activation: BCL10, an important activator of NF- B downstream of protein kinase C, and CASP8AP2 (aka FLASH), which coordinates downstream NF- B activity via a TRAF2-dependent pathway. 22 In addition, many genes within the migration cluster have been implicated in oncogenesis (MDM2, ADAM9, BCL10, etc), a finding pathogenetically consistent with the role of angiogenesis in cancer.

The majority of concordantly regulated genes revealed by our microarray analyses have not been previously associated with angiogenesis. These include the p53 inhibitor MDM2 (activated by all three stimuli) and TXNIP, C-C chemokines, and metallothioneins (repressed by all three stimuli). These concordant transcriptional profiles provide further evidence for a cholinergic component of the angiogenic pathways. For example, our findings of coinduction of FLT1 by all three stimuli, and of nAChR subunit induction by VEGF are consistent with interaction between the signaling pathways.

Of the coregulated genes identified by hierarchical clustering, we focused our attention on TXNIP. Originally identified in HL-60 leukemia cells treated with 1,25 dihydroxyvitamin D 3 (and previously known as Vitamin D 3 upregulated protein 1), TXNIP is an endogenous inhibitor of the ubiquitous redox protein thioredoxin. 12 Thioredoxin, a major redox regulator of protein function and signaling via thiol redox control, has been implicated in the regulation of cellular responses to oxidative stress and apoptosis. 23 Thioredoxin selectively regulates the activity of DNA-binding proteins; for example, two transcription factors concordantly regulated by the three stimuli, NF- B, and p53, require thioredoxin reduction for stimulation of DNA binding. 23,24 Increasing evidence implicates TXNIP and thioredoxin in tumorigenesis. Thioredoxin expression is increased in several human primary cancers, whereas TXNIP is strongly downregulated in human tumor tissues. 13,25 Inhibition of thioredoxin signaling with experimental antitumor agents such as PX-12 (1-methylpropyl 2-imidazolyl disulfide) and pleurotin reduces tumor cell production of HIF-1 and VEGF in vitro and inhibits tumor angiogenesis in vivo. 26 We hypothesized that repression of TXNIP may play a role in mediating growth factor-mediated EC migration.

In the present study, we found that nicotine, VEGF, and bFGF stimulated thioredoxin activity-a finding consistent with their common suppression of TXNIP. Gene knockdown of thioredoxin by siRNA reversed the effect of growth factor stimulation and abrogated the effect of nicotine, VEGF, or bFGF on EC migration. These findings are consistent with a critical role for thioredoxin in mediating growth factor-induced EC migration. Interestingly, the increase in thioredoxin activty induced by each of the three stimuli could be blocked by nAChR antagonism. Finally, gene knockdown of TXNIP alone, without addition of growth factors, induced EC migration. Our findings indicate that TXNIP, by regulating of thioredoxin activity, may play an important role in angiogenesis mediated by growth factor receptors or nAChRs. The mechanism whereby thioredoxin mediates EC migration is poorly understood but may involve stimulation of hypoxia-inducible factor-1 (HIF-1 ), a transcription factor that plays a central role in mediating the angiogenic response to hypoxia. Overexpression of thioredoxin in a variety of malignant cells has been shown to induce HIF-1 expression and VEGF production. 27 Induction of HIF-1 in human endothelial cells upregulates the expression of multiple angiogenic factors including the angiopoietins which are potent stimulators of cell migration via Tie-2 signal transduction pathways. 28 Moreover, as cellular redox state is an important determinant of Rho GTPase activity, it is possible that thioredoxin may play a role in Rho-mediated cytoskeletal remodelling during cell migration. 29

In summary, our data show that activation of nicotinic acetylcholine receptors (nAChRs) induces endothelial cell migration. Furthermore, growth factor (VEGF and bFGF)-induced endothelial cell migration involves nAChR activation. By transcriptional profiling we have identified convergent genomic responses of ECs to nicotine, VEGF, and bFGF. Identification of concordantly regulated genes may provide novel insights into molecular processes mediating EC migration and angiogenesis. Indeed, using this approach we found that TXNIP, by regulating thioredoxin activity, is centrally involved in nAChR-mediated EC migration. Our studies provide evidence for a cholinergic contribution to growth factor-induced EC migration. The nAChRs may play an important role in growth factor-induced angiogenesis, and thus may be a target for therapeutic modulation in disorders of pathological or insufficient angiogenesis.

Acknowledgments

We thank Mary Gerritsen, PhD, and Ken Kengatharan, PhD for scientific and editorial comments.

Sources of Funding

Akiko Ishii-Watabe is supported by the Japanese Health Sciences Foundation. Martin K.C. Ng is supported by the National Health and Medical Research Council of Australia. This study was supported by grants from the National Institutes of Health (R01 HL-63685; RO1 HL-75774; R01 CA098303 and P01 AG18784; and PO1AI50153); Philip Morris USA Inc; and the Tobacco Related Disease Research Program (11RT-0147).

Disclosures

Dr Cooke holds equity in Athenagen Inc, which has licensed Stanford University patents for the use of nAChR agonists and antagonists for disorders of angiogenesis. Dr Cooke is an inventor on these patents, and receives royalties from the licenses. A patent is being filed based upon the intellectual property described in this manuscript that may benefit J.P.C., M.K.C.N., E.C., and J.W.

【参考文献】
  Lindstrom J. Nicotinic acetylcholine receptors in health and disease. Mol Neurobiol. 1997; 15: 193-222.

Macklin KD, Maus AD, Pereira EF, Albuquerque EX, Conti-Fine BM. Human vascular endothelial cells express functional nicotinic acetylcholine receptors. J Pharmacol Exp Ther. 1998; 287: 435-439.

Parnavelas JG, Kelly W, Burnstock G. Ultrastructural localization of choline acetyltransferase in vascular endothelial cells in rat brain. Nature. 1985; 316: 724-725.

Heeschen C, Jang JJ, Weis M, Pathak A, Kaji S, Hu RS, Tsao PS, Johnson FL, Cooke JP. Nicotine stimulates angiogenesis and promotes tumor growth and atherosclerosis. Nat Med. 2001; 7: 833-839.

Heeschen C, Weis M, Aicher A, Dimmeler S, Cooke JP. A novel angiogenic pathway mediated by non-neuronal nicotinic acetylcholine receptors. J Clin Invest. 2002; 110: 527-536.

Zhu BQ, Heeschen C, Sievers RE, Karliner JS, Parmley WW, Glantz SA, Cooke JP. Second hand smoke stimulates tumor angiogenesis and growth. Cancer Cell. 2003; 4: 191-196.

Jacobi J, Jang J, Sundram U, Dayoub H, Fajardo LF, Cooke JP. Nicotine accelerates angiogenesis and wound healing in genetically diabetic mice. Am J Pathol. 2002; 161: 97-104.

Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A. 1998; 95: 14863-14868.

Ridley A, Schwartz M, Burridge K, Firtel R, Ginsberg M, Borisy G, Parsons J, Horwitz A. Cell migration: integrating signals from front to back. Science. 2003; 302: 1704-1709.

Kiosses W, Daniels R, Otey C, Bokoch G, Schwartz M. A role for p21-activated kinase in endothelial cell migration. J Cell Biol. 1999; 147: 831-844.

Henkel G, Krebs B. Metallothioneins: zinc, cadmium, mercury, and copper thiolates and selenolates mimicking protein active site features - structural aspects and biological implications. Chem Rev. 2004; 104: 801-824.

Nishiyama A, Matsui M, Iwata S, Hirota K, Masutani H, Nakamura H, Takagi Y, Sono H, Gon Y, Yodoi J. Identification of thioredoxin-binding protein-2/vitamin D(3) up-regulated protein 1 as a negative regulator of thioredoxin function and expression. J Biol Chem. 1999; 274: 21645-21650.

Baker A, Payne CM, Briehl MM, Powis G. Thioredoxin, a gene found overexpressed in human cancer, inhibits apoptosis in vitro and in vivo. Cancer Res. 1997; 57: 5162-5167.

Yoshioka J, Schulze PC, Cupesi M, Sylvan JD, MacGillivray C, Gannon J, Huang H, Lee RT. Thioredoxin-interacting protein controls cardiac hypertrophy through regulation of thioredoxin activity. Circulation. 2004; 109: 2581-2586.

Wang YG, De Keulenaer GW, Lee RT. Vitamin D(3)-up-regulated protein-1 is a stress-responsive gene that regulates cardiomyocyte viability through interaction with thioredoxin. J Biol Chem. 2002; 277: 26496-26500.

Milner P, Kirkpatrick KA, Ralevic V, Toothill V, Pearson J, Burnstock G. Endothelial cells cultured from human umbilical vein release ATP, substance P and acetylcholine in response to increased flow. Proc Biol Sci. 1990; 241: 245-248.

Wessler I, Kirkpatrick CJ, Racke K. The cholinergic ?pitfall?: acetylcholine, a universal cell molecule in biological systems, including humans. Clin Exp Pharmacol Physiol. 1999; 26: 198-205.

Natori T, Sata M, Washida M, Hirata Y, Nagai R, Makuuchi M. Nicotine enhances neovascularization and promotes tumor growth. Mol Cells. 2003; 16: 143-146.

Jin K, Zhu Y, Sun Y, Mao XO, Xie L, Greenberg DA. Vascular endothelial growth factor (VEGF) stimulates neurogenesis in vitro and in vivo. Proc Natl Acad Sci U S A. 2002; 99: 11946-11950.

Yoshimura S, Takagi Y, Harada J, Teramoto T, Thomas SS, Waeber C, Bakowska JC, Breakefield XO, Moskowitz MA. FGF-2 regulation of neurogenesis in adult hippocampus after brain injury. Proc Natl Acad Sci U S A. 2001; 10: 5874-5879.

Soga N, Namba N, McAllister S, Cornelius L, Teitelbaum S, Dowdy S, Kawamura J, Hruska K. Rho family GTPases regulate VEGF-stimulated endothelial cell motility. Exp Cell Res. 2001; 269: 73-87.

Choi Y, Kim K, Kim H, Hong G, Kwon Y, Chung C, Park Y, Shen Z, Kim B, Lee S, Jung Y. FLASH coordinates NF-kappa B activity via TRAF2. J Biol Chem. 2001; 276: 25073-25077.

Yamawaki H, Haendeler J, Berk BC. Thioredoxin. A key regulator of cardiovascular homeostasis. Circ Res. 2003; 93: 1029-1033.

Ueno M, Masutani H, Arai RJ, Yamauchi A, Hirota K, Sakai T, Inamoto T, Yamaoka Y, Yodoi J, Nikaido T. Thioredoxin-dependent redox regulation of p53-mediated p21 activation. J Biol Chem. 1999; 274: 35809-35815.

Kakolyris S, Giatromanolaki A, Koukourakis M, Powis G, Souglakos J, Sivridis E, Georgoulias V, Gatter KC, Harris AL. Thioredoxin expression is associated with lymph node status and prognosis in early operable non-small cell lung cancer. Clin Cancer Res. 2001; 7: 3087-3091.

Welsh SJ, Williams RR, Birmingham A, Newman DJ, Kirkpatrick DL, Powis G. The thioredoxin redox inhibitors 1-methylpropyl 2-imidazolyl disulfide and pleurotin inhibit hypoxia-induced factor 1alpha and vascular endothelial growth factor formation. Mol Cancer Ther. 2003; 2: 235-243.

Welsh SJ, Bellamy WT, Briehl MM, Powis G. The redox protein thioredoxin-1 (Trx-1) increases hypoxia-inducible factor 1alpha protein expression: Trx-1 overexpression results in increased vascular endothelial growth factor production and enhanced tumor angiogenesis. Cancer Res. 2002; 62: 5089-5095.

Yamakawa M, Liu LX, Date T, Belanger AJ, Vincent KA, Akita GY, Kuriyama T, Cheng SH, Gregory RJ, Jiang C. Hypoxia-inducible factor-1 mediates activation of cultured vascular endothelial cells by inducing multiple angiogenic factors. Circ Res. 2003; 93: 664-673.

Nimnual AS, Taylor LJ, Bar-Sagi D. Redox-dependent downregulation of Rho by Rac. Nat Cell Biol. 2003; 5: 236-241.

作者单位：Department of Medicine (M.K.C.N., J.W., E.C., B.-y.W., R.K.-C., A.I.-W., J.P.C.), Division of Cardiovascular Medicine, Stanford University School of Medicine, Stanford, Calif.; and the Department of Cardiology (M.K.C.N.), Royal Prince Alfred Hospital, Sydney, NSW, Australia.