KLF2 Suppresses TGF-ß Signaling in Endothelium Through Induction of Smad7 and Inhibition of AP-

【摘要】  Objective- The flow-responsive Kruppel-like factor 2 (KLF2) is crucial for maintaining endothelial cell quiescence. Here, we describe its detailed effects on transforming growth factor-ß (TGF-ß) signaling, which normally has proatherogenic effects on endothelium.

Methods and Results- In-depth analysis of genome-wide expression data shows that prolonged lentiviral-mediated overexpression of KLF2 in human umbilical vein endothelial cells (HUVECs) diminishes the expression of a large panel of established TGF-ß-inducible genes. Both baseline and TGF-ß-induced expression levels of plasminogen activator inhibitor 1 (PAI-1) and thrombospondin-1 are greatly diminished by KLF2. Using a combination of ectopic expression, small interfering RNA-mediated knockdown, and promoter activity assays, we show that KLF2 partly inhibits the phosphorylation and subsequent nuclear accumulation of Smad2, thereby suppressing the TGF-ß-induced Smad4-mediated transcriptional activity. This is achieved through TGF-ß-independent induction of inhibitory Smad7. Additionally, a full inhibition of TGF-ß signaling is functionally achieved through a simultaneous suppression of activator protein 1 (AP-1), which is an essential cofactor for TGF-ß-dependent transcription of many genes.

Conclusions- The concerted mechanism by which KLF2 inhibits TGF-ß signaling through induction of inhibitory Smad7 and attenuation of AP-1 activity provides a novel mechanism by which KLF2 contributes to sustaining a quiescent, atheroprotective status of vascular endothelium.

TGF-ß signaling in endothelium is generally considered proatherogenic. The shear stress-induced transcription factor KLF2 inhibits endothelial TGF-ß signaling by inducing the inhibitory Smad7 and suppressing the cofactor AP-1. This mechanism may contribute to the KLF2-mediated atheroprotection of shear stress.

【关键词】  blood flow endothelium gene expression growth factors vascular biology

Introduction

The role of TGF-ß in vascular biology and pathology is still enigmatic, even though it has been the subject of many studies. This is likely because of the complexity of the processes that take place in the vessel wall, involving many cell-types, each of which responds differently to TGF-ß depending on the specific cellular context. For example, TGF-ß signaling in macrophages, smooth muscle cells, and T-lymphocytes is deemed atheroprotective, as it was found to be associated with a decrease in inflammation and vulnerable plaque formation in atherosclerosis. 1-3 TGF-ß signaling in large vessel endothelium, however, promotes apoptosis and increases permeability through mitogen-activated protein kinase (MAPK) p38. 4,5 Furthermore, TGF-ß induces the endothelial oxidized-LDL receptor OLR1, 6 plasminogen activator inhibitor 1 (PAI-1), 7 and monocyte chemotactic protein 1 (MCP-1), 8 all of which are considered proatherogenic. A more detailed understanding of the mechanisms that can control the various aspects of TGF-ß signaling in endothelium are thus of great interest for understanding a variety of vascular processes including atherosclerosis.

TGF-ß signaling occurs through a heteromeric complex of type I and type II TGF-ß receptors. 9 Endothelial cells express one type II receptor and two type I receptors, activin receptor-like kinase (ALK) 5 and ALK1. Stimulation with TGF-ß results in phosphorylation of either ALK1, which leads to phosphorylation of receptor-regulated Smads (R-Smads) Smad1 and Smad5, or ALK5, which leads to phosphorylation of R-Smads Smad2 and Smad3. Phosphorylated R-Smads subsequently translocate to the nucleus with Smad4, where gene expression is regulated through binding to Smad binding elements. Parallel to signaling through Smad proteins, TGF-ß signaling has also been described to occur through MAPKs, probably synergistically with Smad signaling. 10 MAPKs like Jun N-terminal kinase (JNK) and p38 act through activation of activator protein 1 (AP-1), which consists of a homo- or heterodimer of members of the Jun, Fos, musculoaponeurotic fibrosarcoma oncogene homolog (MAF), or activated transcription factor (ATF) families. 11 Known attenuators of TGF-ß signaling are the inhibitory Smads (I-Smads) Smad6 and Smad7, which compete with R-Smads for association with the type I TGF-ß receptor, thereby inhibiting phosphorylation of R-Smads. Moreover, Smad7 induces degradation of the type I receptor by recruiting ubiquitinases. 12 Interestingly, Smad7 is a TGF-ß target gene itself, thus contributing to a negative feedback loop in TGF-ß signaling. Inhibitory Smad7 is known to be endothelial specific and to be induced in vivo by shear stress. 13,14

We have previously identified the transcription factor Kruppel-like factor 2 (KLF2) to be exclusively expressed by endothelial cells exposed to high shear stress. 15 Furthermore, we recently showed that KLF2 affects the expression of vascular tone-regulating genes 16 and establishes a differentiated quiescent endothelial gene expression pattern. 17 Endothelial KLF2 was also shown to inhibit proinflammatory response, 18,19 to suppress prothrombotic function, 20 and to attenuate angiogenesis. 21 Mice lacking KLF2 die in utero because of severe vascular malformations, caused by the inability of endothelial cells to properly organize smooth muscle cells in the vessel wall, resulting in hemorrhage. 22 Here, we report that KLF2 robustly inhibits TGF-ß signaling in endothelial cells by abrogating the phosphorylation of Smad2 in cultured cells and in vivo, and suppressing both Smad3/4- and AP-1-mediated activation of TGF-ß inducible promoters. Using a combination of lentiviral overexpression and small interfering RNA-mediated knockdown, we provide evidence that the inhibition of TGF-ß signaling by KLF2 is mediated through the induction of Smad7 and through the suppression of the active AP-1 component c-Jun.

Methods

A detailed description of the methods is available in the supplemental materials at http://atvb.ahajournals.org.

Cell Culture

Human umbilical vein endothelial cells (HUVECs) were isolated and cultured as previously described. 23 Confluent monolayers were grown from freshly isolated HUVECs and used before the fourth passage.

Lentiviral Overexpression, Western Blotting, and Microscopy

Long-term lentiviral overexpression of KLF2, microscopy, and Western blotting were performed as previously described. 17

Immunohistochemistry and Immunofluorescence

Immunohistochemistry and immunofluorescence was performed essentially as described. 16,17

Luciferase Reporter Constructs and Assay

Firefly luciferase reporter constructs and a GAPDH-Renilla luciferase control construct were transfected into HUVECs by electroporation using the nucleofector system (Amaxa). Luciferase activity was measured with the dual luciferase reporter assay system following the manufacturer?s protocol (Promega).

RNA Silencing by Double-Stranded RNA Oligonucleotides

Double-stranded siRNA oligonucleotides were synthesized by Ambion. Oligofectamine (Invitrogen) was used as transfection agent for introduction of the siRNA oligonucleotides into HUVECs, according to the manufacturer?s instructions.

Real-Time RT- PCR

Real-time RT-PCR was performed as described. 16

Results

KLF2 Abrogates TGF-ß-Induced Gene Expression

Recently, we described the initial analysis and validation of a genome-wide microarray expression profiling study into the general effects of prolonged lentivirus-mediated overexpression of human KLF2 at physiological levels in human umbilical vein endothelial cells (HUVECs). 17 We now performed an in-depth analysis of these expression data probing for affected canonical (signaling) pathways, using the recently published Gene Set Enrichment Analysis (GSEA). 24 One of the highest ranking canonical pathways affected by KLF2 is TGF-ß signaling, which is substantially suppressed (nominal probability value 0.023). Based on the initial default analysis, a custom gene set for endothelial cell TGF-ß targets was constructed containing the core-enriched signaling and downstream effector genes. Figure 1 A shows an overview of the TGF-ß signaling cascade and the core-enriched downstream targets from the GSEA, with superimposed the expression ratio data showing KLF2-upregulated (red) and KLF2-downregulated genes (green). The numbers in the figure indicate fold change in microarray intensity signals of mRNA from KLF2-transduced cells compared with mock-transduced HUVECs. Validation of the KLF2-suppressed basal expression of most of the archetypal TGF-ß downstream genes depicted, most notably: MCP-1, PAI-1, and THBS1 were presented previously. 17

Figure 1. KLF2 represses TGF-ß target gene expression. A, Microarray expression profiling intensity ratios of KLF2- versus mock-transduced HUVECs 17 are superimposed on a customized scheme of the TGF-ß signaling cascade, generated with GenMAPP2.0 to include the downstream genes that were found to be significantly downregulated by KLF2 using Gene Set Enrichment Analysis. 24 1.3-fold and False Discovery Ratio (FDR) probability value <0.001), whereas genes of equal expression are visualized in yellow (when expressed in HUVECs) (absolute ratio 0.001) or gray (no reliable intensity data). Numbers indicate fold change in signal intensities in KLF2- compared with mock-transduced HUVECs. B and C, HUVECs transduced with mock- or KLF2-lentivirus were stimulated with TGF-ß3 for 2 or 24 hours or left unstimulated. RNA levels for PAI-1 and THBS1 were analyzed by real-time RT-PCR (n=3). * P <0.05.

To investigate the mechanism by which KLF2 represses TGF-ß target gene expression, we studied the expression of THBS1 and PAI-1. These genes are known to be induced by TGF-ß, are highly expressed in endothelium, are robustly downregulated by KLF2, and are considered to be proatherogenic. 7,25,26 On stimulation of HUVECs with TGF-ß, mRNA levels of THBS1 and PAI-1 increased approximately 5- and 10-fold, respectively, as measured by RT-PCR ( Figure 1B and 1 C). In KLF2-overexpressing HUVECs, not only the basal expression levels of THBS1 and PAI-1 were significantly lower but also their induction by TGF-ß is severely blunted. These data illustrate that KLF2 suppresses the expression of THBS1 and PAI-1 both in the absence and the presence of exogenous TGF-ß, which suggests that the attenuation of TGF-ß signaling by KLF2 occurs at the receptor level or downstream thereof, and not by altered levels of secreted biologically active TGF-ß isoforms.

KLF2 Increases Smad7 Independent of ALK5-Mediated TGF-ß Signaling

Next, we studied the expression of various genes known to be involved in the regulation of TGF-beta signaling through Smad-proteins by RT-PCR ( Figure 2 A). The expression of 4 of the examined genes is significantly affected by KLF2. TGF-ß1 is 2-fold downregulated, TGF-ß2 is 4-fold downregulated, and the I-Smads Smad6 and Smad7 are upregulated 4- and 3-fold, respectively. Smad7 was reported to be induced in vivo by shear stress, 13 as is KLF2. 16 Long-term laminar shear stress induces Smad7 levels in our cultured endothelial cells to the same levels as observed when overexpressing physiological levels of lentiviral encoded KLF2 ( Figure 2 B and supplemental materials). As Smad7 promoter activity is induced by TGF-ß, the inhibitory Smad7 normally acts as a negative feedback on this signaling cascade. 14 The KLF2-mediated induction of Smad7, however, is independent of ALK5-mediated TGF-ß signaling, as shown by the results in the presence of the specific receptor-blocker SB431542 with both mock- and KLF2-transduced HUVEC ( Figure 2 B). KLF2 induction of Smad7 is confirmed at the protein level by Western blot analysis ( Figure 2 C).

Figure 2. KLF2 regulates genes involved in TGF-ß signaling, most notably Smad7. A, Expression of genes involved in the regulation of TGF-ß signaling were analyzed in mock- and KLF2-transduced HUVECs (n=3). Fold induction by KLF2 is indicated. B, HUVECs were exposed to pulsatile flow (19±12 dynes/cm 2 ) for 7 days (n=3) or kept under static conditions. Mock- and KLF2-transduced HUVECs (n=3) were treated with SB431542 for 16 hours. mRNA levels for Smad7 were measured by real-time RT-PCR. Fold induction compared with controls is indicated. C, Western blot analysis, performed with antibodies raised against Smad7 or -tubulin on cell lysates of mock- and KLF2-transduced HUVECs. Bars indicate quantification of Smad7 levels corrected for -tubulin levels. D, Immunofluorescence images of mock- and KLF2-transduced HUVEC. Smad7 is visualized in red, Hoechst nuclear staining in blue. Bars indicate 10 µm. * P <0.05 compared with their individual controls.

Cytoplasmic localization of Smad7 is required for inhibition of phosphorylation of Smad2, which occurs near the plasma membrane, whereas nuclear localized Smad7 was shown to be inactive. 12 Figure 2 D shows immunofluorescence microscopy images of HUVECs with Smad7 protein in red and nuclear staining in blue. Overexpression of KLF2 is shown to increase active cytoplasmic and plasma membrane localized Smad7 protein, but also to a lesser degree, inactive nuclear localized Smad7. Also visible is the smaller size of KLF2-transduced cells, as we described previously. 17 Taken together, these experiments signify that KLF2 augments active Smad7 protein, independent of TGF-ß signaling.

KLF2-Induced Smad7 Suppresses Smad2 Phosphorylation and Smad3/4-Dependent Transcriptional Activation

To determine whether KLF2-induced Smad7 inhibits TGF-ß signaling by inhibiting phosphorylation of R-Smads, we used specific antibodies raised against phosphorylated Smad2 (P-Smad2) and Smad3 (P-Smad3). Western blot analysis revealed that KLF2-transduced HUVECs contain lower levels of P-Smad2 and P-Smad3 compared with mock-transduced cells, whereas total Smad2 levels are unaltered ( Figure 3 A and supplemental materials). Also, TGF-ß stimulation induces phosphorylation in mock-transduced HUVECs but not in KLF2-transduced cells. On phosphorylation, Smad2 typically translocates to the nucleus to influence transcriptional activity. Immunofluorescence analysis confirmed that P-Smad2 is indeed located predominantly in the nucleus in both mock- and KLF2-transduced cells, but is lower in KLF2-transduced cells ( Figure 3 B and supplemental materials). These results indicate that KLF2 inhibits the TGF-ß signaling pathway in HUVECs upstream of the phosphorylation of Smad2, leading to lower levels of active P-Smad2 in the nucleus.

Figure 3. KLF2 inhibits phosphorylation of Smad2 in a Smad7-dependent manner and suppresses Smad3/4-mediated transcription. A, Western blot analysis was performed with total cell lysates of mock- and KLF2-transduced HUVECs that were treated with 10 ng/mL TGF-ß3 or vehicle for 16 hours. Blots were probed with antibodies raised against P-Smad2 and with antibodies against -tubulin. The bar graph indicates quantified levels of P-Smad2 corrected for -tubulin. B, Immunofluorescence images of HUVECs after stimulation with 10 ng/mL TGF-ß3 for 16 hours. P-Smad2 is visualized in red, Hoechst nuclear staining in blue. Bars indicate 10 µm. C, Cells were transfected by electroporation with luciferase reporter constructs containing the Smad3/4 specific (CAGA) 12 -artificial promoter (n=3). Luciferase activity was measured 16 hours after addition of 10 µmol/L ALK5 inhibitor SB431542 or 10 ng/mL TGF-ß3. D, Western blot analysis was performed with total cell lysates of mock- and KLF2- transduced HUVEC, 24 hours after transfection with double-stranded RNA oligonucleotides designed either to specifically silence Smad7 (siSmad7) or to be nonspecific. Blots were probed with P-Smad2 or -tubulin antibodies. The bar graph illustrates quantified levels of P-Smad2 corrected for -tubulin (n=3). * P <0.05, ** P <0.01, *** P <0.0005.

Next, we used an established Smad3/4-binding and TGF-ß-responsive promoter element 27 to elucidate whether the observed abrogating effect of KLF2 on phosphorylation of R-Smads leads to actual transcriptional suppression. This artificial (CAGA) 12 luciferase reporter is known to be activated only by Smad3/4 binding and can thus be used to specifically determine Smad3/4-mediated TGF-ß signaling. Mock- and KLF2-transduced HUVECs were transfected by electroporation with the (CAGA) 12 luciferase reporter and subsequently treated with the ALK5 inhibitor SB431542, to minimize residual autocrine TGF-ß activity, or with TGF-ß for 16 hours before measuring luciferase activity ( Figure 3 C). As expected, stimulation with TGF-ß greatly induces reporter activity to approximately 18-fold in mock-transduced cells, when compared with SB431542 -treated cells. In contrast, reporter activity in TGF-ß-treated KLF2-transduced cells increases only 5-fold compared with SB431542 -treated KLF2-transduced cells. This implies that KLF2 inhibits TGF-ß signaling in HUVECs partially by specifically diminishing Smad3/4 transcriptional activity.

To determine whether the decrease in phosphorylation of Smad2 by KLF2 is mediated by Smad7, mock- and KLF2-transduced HUVECs were transfected with either double-stranded RNA oligonucleotides to specifically silence Smad7 (siSmad7) or with nonspecific control oligonucleotides ( Figure 3 D). Introduction of siSmad7 reduces Smad7 mRNA in KLF2-transduced cells to about the same levels observed in cells transfected with control oligonucleotides (see supplemental materials). This reduction of Smad7 is sufficient to restore phosphorylation of Smad2 in KLF2-transduced cells to approximately the levels observed in mock-transduced endothelial cells ( Figure 3 D). These data suggest that KLF2 inhibits the TGF-ß signaling cascade by decreasing phosphorylation of Smad2 through induction of Smad7.

Endothelial P-Smad2 Levels Are Diminished In Vivo in Atheroprotected Arterial Regions

We examined sections of a mouse carotid artery collar model 16 and human arteries using immmunohistochemical techniques, to assess whether P-Smad2 levels are also diminished in vivo in endothelium that expresses increased levels of KLF2 ( Figure 4 and supplemental materials). Mouse carotid artery sections and human donor arterial tissue, displaying either an early focal "fatty-streak" or a more advanced acentric lesion, which tend to develop at the low shear stress side of the artery, 16 were stained for KLF2, P-Smad2, HAM56, and CD31. Endothelial cells overlying lesions display diminished levels of KLF2 and enhanced P-Smad2 levels compared with healthy endothelium from the same section. Collectively, these data suggest that high endothelial expression of KLF2 provoked by shear stress is inversely correlated to phosphorylation of Smad2 in vivo.

Figure 4. Endothelial P-Smad2 levels are diminished in regions protected from atherogenesis coinciding with increased KLF2 levels. Previously, we described the induction of KLF2 by shear stress in vivo in a mouse carotid artery collar model 15 (A). Sections from mouse carotid arteries (B and C), sections from the abdominal aorta from a 41-year old female donor (G and H), and sections from an iliac artery from a 49-year old male donor (L and M) were stained for KLF2 (brown/red). A staining for P-Smad2 (D, E, I, J, N, and O) or HAM56 (F and K) was also performed on serial sections. Counterstaining was performed with either hematoxylin, visualizing nuclei in blue (B-F and I-O), or with light-green, showing collagen in green (G and H). Carotid artery sections exposed to low shear stress (proximal to the collar) are visualized in B and D and intracollar sections are shown in C and E. Magnifications are shown of a region where a "fatty-streak" was observed (G and I) and of a healthy region (H and J) in the abdominal aorta section. Furthermore, a region where an atheroma was observed (L and N) and a healthy region (M and O) in the iliac artery section were magnified. Arrow heads indicate positive endothelial cells.

KLF2 Decreases c-Jun Transcriptional Activity on a TGF-ß-Responsive Promoter

An apparent discrepancy is revealed when comparing the partial Smad-dependent effects of KLF2 on an artificial (CAGA) 12 -promoter ( Figure 3 C) to its complete inhibitory effect on endogenous gene expression ( Figure 1 ). Therefore, we studied its effects on an established TGF-ß-responsive core promoter, originally derived from the murine Smad7 gene. This core-promoter fragment contains a synergistic module composed of both a Smad3/4 binding site and an AP-1 binding site, as is frequently observed in TGF-ß-responsive genes. 28 Luciferase activity was measured for the wild-type reporter or reporter constructs containing mutations in either the AP-1 binding site or the Smad binding element (SBE) after incubation with TGF-ß or SB431542 for 16 hours. The activity of the TGF-ß responsive part of the wild-type reporter was indeed induced by TGF-ß in mock-transduced cells compared with SB431542 treatment ( Figure 5 A). In contrast, induction of luciferase activity after TGF-ß stimulation in KLF2-transduced HUVECs is completely abrogated. Notably, baseline levels are similar to mock-transduced cells, indicating that there is no direct effect of KLF2 on the expression of this Smad7 core-promoter. Mutation of the SBE leads to a marked drop in transcriptional activity and loss of TGF-ß responsiveness in both mock- and KLF2-transduced HUVECs, confirming that Smad3/4 binding is essential for both baseline and TGF-ß-induced transcriptional activity ( Figure 5 A).

Figure 5. KLF2 inhibits AP-1 activity and promoter response. A, Luciferase reporter constructs containing either the wild-type TGF-beta responsive part (-613 to +112) of the murine Smad7 promoter or the same part with mutations in either the Smad binding element (SBE mut) or the AP-1 binding element (AP1 mut) were used as a measure for physiological TGF-beta signaling. 28 Luciferase activity was measured 16 hours after addition of 10 µmol/L ALK5 inhibitor SB431542 or 10 ng/mL TGF-ß3 to either mock- or KLF2-transduced HUVEC (n=4). B, Using a transcription factor ELISA, DNA binding activity of nuclear localized P-c-Jun was measured in mock- and KLF2-transduced HUVECs (n=3). C, This panel illustrates immunofluorescence images of HUVECs. P-c-Jun is visualized in red, Hoechst nuclear staining in blue. Bars indicate 10 µm. D, Western blot analysis was performed with antibodies raised against c-Jun or -tubulin on cell lysates of mock- and KLF2-transduced HUVECs. Bars indicate quantification of c-Jun levels corrected for -tubulin levels. (n=3). * P <0.05, ** P <0.01 N.S.: P 0.05

Mutation of the AP-1 binding site equally results in a loss of TGF-ß responsiveness of this core-promoter, confirming the crucial role of AP-1 in Smad-driven transcription. Interestingly, we also noted a significantly decreased basal transcriptional activity of the AP-1 site containing, SBE-mutated construct in the KLF2-transduced cells. We therefore analyzed the levels of active AP-1 in KLF2- and mock-transduced cells. c-Jun is a main component of AP-1, but to be active it needs to be localized in the nucleus in a phosphorylated form (P-c-Jun). 11 This was directly assessed by transcription factor analysis of nuclear extracts, which shows that KLF2 indeed reduces the levels of active nuclear localized P-c-Jun by 80% ( Figure 5 B). In addition, P-c-Jun localization was analyzed by immunofluorescence, showing that KLF2 suppresses both the nuclear-localized amount and the total intracellular amount of P-c-Jun ( Figure 5 C). In marked contrast, total c-Jun protein levels are unaffected by KLF2 ( Figure 5 D), indicating that KLF2 inhibits the phosphorylation, rather than the expression levels of c-Jun.

Discussion

The present study provides evidence that endothelial KLF2 attenuates transcription of many TGF-ß target genes, including PAI-1 and THBS1 ( Figure 1B and 1 C) and completely abrogates transcriptional activity on endogenous TGF-ß-inducible promoter fragments ( Figure 5 A). This is shown to be partially achieved by the induction of Smad7 ( Figure 2 ), which subsequently suppresses the phosphorylation of Smad2 ( Figure 3A and 3 D) and inhibits transcriptional activity of Smad3/4 ( Figure 3 C). Synergistically, KLF2 inhibits the phosphorylation and thus activation of the essential cofactor for TGF-ß signaling AP-1 ( Figure 5 ). These findings are supported by the observation that P-Smad2 levels are indeed higher in vivo at sites of the vasculature that correlate with suppressed endothelial KLF2 levels ( Figure 4 ).

Smad7 has been described to be specifically expressed in endothelium and to be induced by both shear stress and TGF-ß. 13,14 KLF2 is also exclusively expressed by endothelial cells exposed to high shear stress. 15 Because we describe that KLF2 inhibits TGF-ß-dependent transcription, this should lead to a decrease in Smad7 transcription. Here it is shown, however, that KLF2 induces Smad7 independent of ALK5-mediated TGF-ß signaling ( Figure 2 B). KLF2 indeed does not induce the expression of Smad7 through the TGF-ß-responsive part of its promoter in the presence or absence of the ALK5 inhibitor SB431542 ( Figure 5 A). Therefore, KLF2 must act either directly or indirectly via one of its many downstream genes, on a different part of the Smad7 promoter or intergenic regions. Unfortunately, neither a specific consensus binding sequence for KLF2 nor for any of its KLF family members has been defined, except for the core binding sequence (CACCC), 29 which is ubiquitously present in the human genome making it at present difficult to pinpoint a specific putative KLF2 binding site in the Smad7 promoter.

We show that induction of Smad7 is responsible for the suppression of P-Smad2 by KLF2. However, P-Smad2 is not completely absent in KLF2-transduced cells ( Figure 3A, 3B, and 3 E), and this could explain the relatively mild reduction in Smad3/4-dependent transcription ( Figure 3 C). On the contrary, a complete inhibition on target gene expression ( Figure 1B and 1 C) and the full abrogation of TGF-ß responsiveness on the endogenous TGF-ß-responsive promoter piece of Smad7 ( Figure 5 A) by KLF2 were observed. A plausible explanation for this observation is that endogenous TGF-ß-induced gene expression requires not only Smad3/4, but also its cofactor AP-1. 28 In support of this explanation, we show that KLF2 diminishes AP-1 activity ( Figure 5 B) and that mutation of the SBE alone in the Smad7-derived promoter fragment does not abolish decreased transcriptional activity in KLF2-transduced cells compared with control cells ( Figure 5 A). Indeed, promoters of 3 of the most downregulated genes in KLF2-transduced cells, Endothelin-1, MCP-1, and PAI-1 ( Figure 1 A) have all been shown to contain both essential Smad3/4 and AP-1 binding sites. 27,30,31 Because active P-c-Jun is reduced by KLF2, but total c-Jun protein levels are unaffected by KLF2 ( Figure 5 D), it is likely that KLF2 directly or indirectly inhibits JNK to reduce phosphorylation of c-Jun ( Figure 6 ).

Figure 6. KLF2 inhibits TGF-ß signaling via two distinct routes. KLF2-mediated induction of Smad7 leads to the inhibition of phosphorylation of Smad2, which leads to diminished transcriptional activity of Smad4. To establish full abrogation of TGF-ß target genes, KLF2 simultaneously inhibits the activation of the TGF-ß signaling cofactor AP-1.

Previously, we reported that KLF2 establishes endothelial quiescence by directly and indirectly regulating the expression of over a thousand genes. 17 We now show that KLF2 specifically inhibits TGF-ß signaling through a novel concerted mechanism involving both Smad7 and AP-1. Thus, one of the mechanisms is established that explains part of the vast indirect transcriptional regulation that constitutes KLF2-driven endothelial quiescence. Furthermore, this is shown to directly result in a downregulation of a distinct set of TGF-ß-inducible genes, which are considered proatherogenic 4-8 and/or profibrotic 32 ( Figure 1 A). Collectively, these results suggest that healthy flow-induced KLF2 will contribute directly to a quenching of the pathological role of TGF-ß in various vascular processes, including atherosclerosis.

Acknowledgments

We thank Drs. Heuchel and Heldin for providing the Smad7 reporter construct.

Sources of Funding

This study was supported by the Netherlands Heart Foundation (Molecular Cardiology Program grant NHS93.007), the NWO-Genomics Program (grant 050-10-014), the European Vascular Genomics Network (grant LSHM-CT-2003-503254), and by the EC project "Angiotargeting Integrated Project" No.504743.

Disclosures

None.

【参考文献】
  Robertson AK, Rudling M, Zhou X, Gorelik L, Flavell RA, Hansson GK. Disruption of TGF-beta signaling in T cells accelerates atherosclerosis. J Clin Invest. 2003; 112: 1342-1350.

Grainger DJ. Transforming growth factor beta and atherosclerosis: so far, so good for the protective cytokine hypothesis. Arterioscler Thromb Vasc Biol. 2004; 24: 399-404.

Argmann CA, Van Den Diepstraten CH, Sawyez CG, Edwards JY, Hegele RA, Wolfe BM, Huff MW. Transforming Growth Factor-{beta}1 Inhibits Macrophage Cholesteryl Ester Accumulation Induced by Native and Oxidized VLDL Remnants. Arterioscler Thromb Vasc Biol. 2001; 21: 2011-2018.

Hyman KM, Seghezzi G, Pintucci G, Stellari G, Kim JH, Grossi EA, Galloway AC, Mignatti P. Transforming growth factor-beta1 induces apoptosis in vascular endothelial cells by activation of mitogen-activated protein kinase. Surgery. 2002; 132: 173-179.

Goldberg PL, MacNaughton DE, Clements RT, Minnear FL, Vincent PA. p38 MAPK activation by TGF-beta1 increases MLC phosphorylation and endothelial monolayer permeability. Am J Physiol Lung Cell Mol Physiol. 2002; 282: L146-L154.

Minami M, Kume N, Kataoka H, Morimoto M, Hayashida K, Sawamura T, Masaki T, Kita T. Transforming growth factor-beta(1) increases the expression of lectin-like oxidized low-density lipoprotein receptor-1. Biochem Biophys Res Commun. 2000; 272: 357-361.

Sawdey M, Podor TJ, Loskutoff DJ. Regulation of type 1 plasminogen activator inhibitor gene expression in cultured bovine aortic endothelial cells. Induction by transforming growth factor-beta, lipopolysaccharide, and tumor necrosis factor-alpha. J Biol Chem. 1989; 264: 10396-10401.

Wu X, Ma J, Han JD, Wang N, Chen YG. Distinct regulation of gene expression in human endothelial cells by TGF- and its receptors. Microvascular Research. 2006; 71: 12-19.

Heldin CH, Miyazono K, ten Dijke P. TGF-beta signalling from cell membrane to nucleus through SMAD proteins. Nature. 1997; 390: 465-471.

Javelaud D, Mauviel A. Crosstalk mechanisms between the mitogen-activated protein kinase pathways and Smad signaling downstream of TGF-beta: implications for carcinogenesis. Oncogene. 2005; 24: 5742-5750.

Wisdom R. AP-1: one switch for many signals. Exp Cell Res. 1999; 253: 180-185.

Kavsak P, Rasmussen RK, Causing CG, Bonni S, Zhu H, Thomsen GH, Wrana JL. Smad7 binds to Smurf2 to form an E3 ubiquitin ligase that targets the TGF beta receptor for degradation. Mol Cell. 2000; 6: 1365-1375.

Topper JN, Cai J, Qiu Y, Anderson KR, Xu YY, Deeds JD, Feeley R, Gimeno CJ, Woolf EA, Tayber O, Mays GG, Sampson BA, Schoen FJ, Gimbrone MA Jr., Falb D. Vascular MADs: two novel MAD-related genes selectively inducible by flow in human vascular endothelium. Proc Natl Acad Sci U S A. 1997; 94: 9314-9319.

Nakao A, Afrakhte M, Morén A, Nakayama T, Christian JL, Heuchel R, Itoh S, Kawabata M, Heldin NE, Heldin CH, ten Dijke P. Identification of Smad7, a TGFbeta-inducible antagonist of TGF-beta signalling. Nature. 1997; 389: 631-635.

Dekker RJ, van Soest S, Fontijn RD, Salamanca S, de Groot PG, VanBavel E, Pannekoek H, Horrevoets AJ. Prolonged fluid shear stress induces a distinct set of endothelial cell genes, most specifically lung Kruppel-like factor (KLF2). Blood. 2002; 100: 1689-1698.

Dekker RJ, van Thienen JV, Rohlena J, de Jager SC, Elderkamp YW, Seppen J, de Vries CJ, Biessen EA, van Berkel TJ, Pannekoek H, Horrevoets AJ. Endothelial KLF2 links local arterial shear stress levels to the expression of vascular tone-regulating genes. Am J Pathol. 2005; 167: 609-618.

Dekker RJ, Boon RA, Rondaij MG, Kragt A, Volger OL, Elderkamp YW, Meijers JC, Voorberg J, Pannekoek H, Horrevoets AJ. KLF2 provokes a gene expression pattern that establishes functional quiescent differentiation of the endothelium. Blood. 2006; 107: 4354-4363.

Senbanerjee S, Lin Z, Atkins GB, Greif DM, Rao RM, Kumar A, Feinberg MW, Chen Z, Simon DI, Luscinskas FW, Michel TM, Gimbrone MA Jr., Garcia-Cardena G, Jain MK. KLF2 Is a novel transcriptional regulator of endothelial proinflammatory activation. J Exp Med. 2004; 199: 1305-1315.

Lin Z, Hamik A, Jain R, Kumar A, Jain MK. Kruppel-Like Factor 2 Inhibits Protease Activated Receptor-1 Expression and Thrombin-Mediated Endothelial Activation. Arterioscler Thromb Vasc Biol. 2006; 26: 1185-1189.

Lin Z, Kumar A, Senbanerjee S, Staniszewski K, Parmar K, Vaughan DE, Gimbrone MA Jr., Balasubramanian V, Garcia-Cardena G, Jain MK. Kruppel-like factor 2 (KLF2) regulates endothelial thrombotic function. Circ Res. 2005; 96: e48-e57.

Bhattacharya R, Senbanerjee S, Lin Z, Mir S, Hamik A, Wang P, Mukherjee P, Mukhopadhyay D, Jain MK. Inhibition of VPF/VEGF-mediated angiogenesis by the Kruppel-like factor KLF2. J Biol Chem. 2005; 280: 28848-28851.

Kuo CT, Veselits ML, Barton KP, Lu MM, Clendenin C, Leiden JM. The LKLF transcription factor is required for normal tunica media formation and blood vessel stabilization during murine embryogenesis. Genes Dev. 1997; 11: 2996-3006.

Jaffe EA, Nachman RL, Becker CG, Minick CR. Culture of human endothelial cells derived from umbilical veins. Identification by morphologic and immunologic criteria. J Clin Invest. 1973; 52: 2745-2756.

Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, Mesirov JP. From the Cover: Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. PNAS. 2005; 102: 15545-15550.

Freyberg MA, Kaiser D, Graf R, Buttenbender J, Friedl P. Proatherogenic Flow Conditions Initiate Endothelial Apoptosis via Thrombospondin-1 and the Integrin-Associated Protein. Biochem Biophys Res Commun. 2001; 286: 141-149.

Negoescu A, Lafeuillade B, Pellerin S, Chambaz EM, Feige JJ. Transforming growth factors beta stimulate both thrombospondin-1 and CISP/thrombospondin-2 synthesis by bovine adrenocortical cells. Exp Cell Res. 1995; 217: 404-409.

Dennler S, Itoh S, Vivien D, ten Dijke P, Huet S, Gauthier JM. Direct binding of Smad3 and Smad4 to critical TGF beta-inducible elements in the promoter of human plasminogen activator inhibitor-type 1 gene. EMBO J. 1998; 17: 3091-3100.

Brodin G, Ahgren A, ten Dijke P, Heldin CH, Heuchel R. Efficient TGF-beta induction of the Smad7 gene requires cooperation between AP-1, Sp1, and Smad proteins on the mouse Smad7 promoter. J Biol Chem. 2000; 275: 29023-29030.

Feinberg MW, Lin Z, Fisch S, Jain MK. An emerging role for Kruppel-like factors in vascular biology. Trends Cardiovasc Med. 2004; 14: 241-246.

Rodriguez-Pascual F, Redondo-Horcajo M, Lamas S. Functional Cooperation Between Smad Proteins and Activator Protein-1 Regulates Transforming Growth Factor-{beta}-Mediated Induction of Endothelin-1 Expression. Circ Res. 2003; 92: 1288-1295.

Xiao YQ, Malcolm K, Worthen GS, Gardai S, Schiemann WP, Fadok VA, Bratton DL, Henson PM. Cross-talk between ERK and p38 MAPK Mediates Selective Suppression of Pro-inflammatory Cytokines by Transforming Growth Factor-beta. J Biol Chem. 2002; 277: 14884-14893.

Lee HB, Ha H. Plasminogen activator inhibitor-1 and diabetic nephropathy. Nephrology. 2005; 10: S11-S13.

作者单位：Department of Medical Biochemistry (R.A.B., J.O.F., O.L.V., F.W., H.P., A.J.G.H.), Academic Medical Center, University of Amsterdam; the Department of Molecular Cell Biology (E.P., P.t.D.), Leids Universitair Medisch Centrum, Leiden University; and the Division of Biopharmaceutics (E.J.A.v.W., J.K.)