The Role Of Krüppel-like Factor 5 In Abdominal Aortic Aneurysm Formation And The Underlying Mechanisms

Abdominal aortic aneurysms(AAAs) is a chronic inflammatory disease characterized by remodeling of the aortic wall, and frequently lead to high morbidity and mortality due to vascular dissections and ruptures. Although AAA formation is a multifactorial process, involving the infiltration of macrophages, release of proinflammatory cytokines and proteases, elastin breakdown, vascular smooth muscle cell(VSMC) apoptosis, and increase of collagen turnover, macrophages contribute importantly to the pathogenesis of AAAs. It has been demonstrated that monocytes/macrophages activated by chronic inflammatory states, including atherosclerosis, oxidative stress, angiotensin II(Ang II), and inflammatory cytokines, infiltrate the vessel wall, release proteases, including elastase and metalloproteinases, and degrade extracellular matrix components, including collagen and elastin. Simultaneously, infiltrating macrophages secrete inflammatory cytokines in the media and adventitia of aneurysmatic vessels, such as tumor necrosis factor(TNF)-α, interferon-γ, interleukin-1β, and interleukin-6, all of which exacerbate inflammatory responses.The infiltration of macrophages into the vessel wall requires highly coordinated reorganization of actin cytoskeletal structures to create membrane protrusion called podosome. It has been known that both human and murine podosomes contain many of the components structural and signaling proteins commonly found in focal adhesions, such as cortactin, Tks5, vinculin, paxillin, and talin. Podosomes appear as small dots(punctate staining) with a lifetime around 2-10 min, but in response to cytokines, individual podosomes can accumulate in ring-like structures called rosettes that can last for hours, allowing for sustained ECM digestion. In addition, podosomes contain several signaling proteins including Src, FAK, PI(3)-kinase, p190 Rho GAP and Cdc42. Rho regulation has been shown to be a key intermediate in podosome formation. However, there is currently little understanding of the mechanism of macrophage migration in the context of podosome formation.Myo9b, a member of myosin class IX, is a unique actin-based motor protein that contains a Rho GAP domain, which, like other Rho GAPs, is inhibitory to Rho signaling. It has been known that Myo9 b localizes to dynamic F-actin at the extending cell front and regulates Rho signaling. Thereby it controls cell polarity and cell migration. Although Myo9 b has been shown to be expressed in osteoclasts and act as a critical regulator of podosome patterning and osteoclast function, the link between Myo9 b and aneurysm development has not yet been elucidated.Krüppel-like factor 5(KLF5) is DNA-binding transcriptional regulator that regulates a number of cellular processes, including development, differentiation, proliferation, and apoptosis. In the cardiovascular system, KLF5 is a target for angiotensin II signaling and an essential regulator of cardiovascular remodeling. Although KLF5 has also been shown to be highly expressed in large and giant unruptured cerebral cneurysms, the specific contribution of KLF5 to podosome formation and macrophage infiltration during AAA formation remains unknown.In this study, we tested whether and how KLF5 modulates macrophage infiltration during AAA formation by regulating podosome formation. We found that KLF5/Myo9b/Rho A pathway activated by inflammation is a significant aspect of the KLF5-induced podosome formation and macrophage infiltration during AAA formation.PartⅠ KLF5 is involved in AAA formation by promoting the migration of macrophagesObjective: To investigate the role of KLF5 in migration of macrophages as well as in AAA formation.Methods: 1 Real-time q RT-PCR and immunohistochemistry were performed to examine the expression of KLF5 in human AAA tissues. Confocal fluorescence microscopy was used to evaluate macrophages,VSMCs, and KLF5 in human AAA tissue. 2 AAA model was estabolished in C57BL/6 mice, and morphology of injured vessel was assessed by hematoxylin/eosin staining. QRT-PCR and immunohistochemistry were performed to examine the expression of KLF5 in murine AAA tissues. Confocal fluorescence microscopy was used to evaluate macrophages, VSMCs, and KLF5 in murine AAA tissue. 3 Macrophage-specific KLF5 knockout mice(KLF5-/- mice) were generated by crossing KLF5-flox mice and Lys M-cre mice. The experimental AAA model in WT and KLF5-/- mice was induced by Ca PO4 treatment. 4 HE and elastic Van Gieson(EVG) staining were performed to examine medial layer and elastin of vessel. 5 Double immunofluorescence colocalization was performed to evaluate KLF5 and macrophage.Results:1 KLF5 is upregulated in human AAAThe level of KLF5 m RNA was significantly higherin human AAA than in normal aortic tissues via real-time q RT-PCR. KLF5 staining was markedly increased in human AAA tissue versus controls via immunohistochemistry. To determine which cell types specifically expressed KLF5, confocal fluorescence microscopy was used to evaluate macrophages, VSMCs, and KLF5 in human AAA tissues.KLF5 colocalized primarily to intramural macrophages and SMCsin AAA lesions.Among all MAC2-positive cells, the percentages of KLF5-positive macrophages were significantly higher in AAA than that of controls. Furthermore, the percentages of KLF5-positive SMCs were significantly higher in AAA than that of controls.2 KLF5 is upregulated in experimental AAAWe next induced aneurysm in C57BL/6 mice by applying calcium phosphate(Ca PO4) onto the mouse infrarenal aortas periadventitially, an established AAA model. Although no differences in aortic dilation were present at day 3, aortic dilation was obvious at day 7and most prominent 14 and 21 days after treatment with Ca PO4. HE staining showed luminal dilation, decreased smooth muscle layers, and wall thinning in mouse aortas exposed to Ca PO4. In mouse aortas KLF5 m RNA was expressed in a low level under basal conditions. KLF5 expression was slightly increased in the aorta treated by Ca PO4 at 3 days, however, KLF5 expression at 7 and 14 days increased significantly in the Ca PO4-treated aortic wall by 2.39 ± 0.35-foldand 2.14±0.28-fold, respectively, over the controls. By immunohistochemistry, KLF5 expression progressively increased after Ca PO4 application as aortic diameter increased. Confocal immunofluorescence staining of KLF5, F4/80 and SM α-actin demonstrated that KLF5(red) was expressed in both macrophages and VSMCs in our experimental AAA induced by Ca PO4, consistent with the results obtained from human AAA tissue.3 Macrophage-specific knockout of KLF5 attenuates macrophage infiltration and aneurysm formationBecause KLF5 was upregulated in the macrophages infiltrated into AAA tissue and that macrophage infiltration was markedly increased in human and experimental AAA tissues, we sought to determine whether there was a regulatory relationship between KLF5 expression and macrophage infiltration.To do this, macrophage-specific KLF5 knockout mice(KLF5-/-mice) were generated by crossing KLF5-flox mice and Lys M-cre mice. The experimental AAAs in WT and KLF5-/- mice was induced by Ca PO4 treatment.KLF5-/- mice had >30 % absolute reduction in maximal aortic dilation compared with WT mice at 14 days after Ca PO4 treatment. HE staining showed that aortas from KLF5-/- mice had normal architecture and a well preserved medial layer. Elastin degradation was significantly less in KLF5-/- mice than that in WT mice, as documented by EVGstaining. The media thickness shown by HE staining confirmed that macrophage-specific knockout of KLF5 could effectively protect against medial SMC loss. Co-immunostaining with anti-KLF5 and anti-F4/80 antibodies showed that KLF5-positive macrophages infiltrated into the aortic wall were markedly less in KLF5-/- mice than in WT mice.PartⅡ KLF5-media Myo9 b regulatespodosome rosette formation in macrophages and is involved in the macrophagemigrationObjective: To elucidate whether KLF5-dependent regulation of Myo9 baffects podosome rosette formation and macrophagemigration.Method: 1 Cell scratch migration assay was performed for WT and KLF5-/- macrophages. 2 Using a Boyden chamber transwell migration assay, with macrophages seeded in the upper chamber and Ca PO4-stimulated VSMCs in the lower chamber, we measured the total number of WT and KLF5-/- macrophages migrating through a porous membrane to underside of the membrane. 3 4 Dual immunofluorescence staining was performed to confirm expression of KLF5 in WT and KLF5-/-macrophages. 5 Time-lapse imaging was used to clarifyforward-and-backward movement and migrating rate of WT and KLF5-/- macrophages.1 KLF5-/- macrophages are defective in their migration ability in vitroTo further determine the role of KLF5 in macrophage migration and infiltration, marrow-derived macrophage(BMMs) from WT and KLF5-/- mice was used to observe the effects of KLF5 deletion on macrophage migration. Cell scratch migration assay showed that KLF5-/- BMMs had less migration into the cell-denuded gap than did WT BMMs. Boyden chamber transwell migration assayshowed that the number of KLF5-/-BMMs migrating across the membrane was significantly lower than that of WT BMMs. Scanning electron microscopy showed that KLF5-/-BMMs formed fewer and shorter pseudopodial projections when passing through the hole to the other side of membrane compared with those in WT BMMs. To investigate whether forward-and-backward movement is affected by genetic deletion of KLF5, we performed a chemotaxis assay for macrophages using time-lapse imaging on the culture dishes.The results showed that WT BMMs exhibited a marked migration toward the high concentration of TNF-α as TNF-α-treated time was increased, whereas the movement of KLF5-/-BMMs was minimal.2 KLF5 mediates myo9 b expression via direct binding to the TCE site-1 in the Myo9 b promoterTo define which genes regulated by KLF5 are responsible for macrophage migration and infiltration, we performed RNA expression profiling by microarray analysis of AAA and normal abdominal aortas from WT and KLF5-/- mice. The results revealed that the expression of motility-related genes was markedly different between AAA tissues and normal aorta. Of these, Myo9a(2.67-fold) and Myo9b(3.22-fold) were the most down-regulated genes in experimental AAA tissue of Ca PO4-treated KLF5-/- mice. Gene ontology(GO) analysis showed that the higher enriched GOs targeted by genetic deletion of KLF5 were myosin filaments and myosin complex. QRT-PCR was used to validate that Myo9 a, Myo9 b and Macf1 were significantly down-regulated in AAA tissue of KLF5-/-mice compared with WT mice, consistent with microarray data. To further identify whether KLF5 directly regulates Myo9 a and Myo9 b gene expression, we infected Raw264.7 with Ad-KLF5 or transfected macrophages with si RNA to KLF5(si-KLF5) to overexpress or knock down KLF5 expression. The results of Western blot and q RT-PCR showed that overexpression of KLF5 significantly increased the basal and TNF-α-induced Myo9 a and Myo9 b expression at the transcription and translation levels. Conversely, knockdown of KLF5 by si-KLF5 blocked TNF-α-induced Myo9 a and Myo9 b expression and also reduced its basal expression.To examine whether KLF5 regulates the expression of Myo9 b gene through interacting with the Myo9 b promoter,we constructed a Myo9 b promoter-luciferase reporter construct containing the sequence between-1396 to +1 bp of mouse Myo9 b gene and performed a luciferase assay in 293 A cells. The results indicated that overexpression of KLF5 concentration-dependently increased the Myo9 b promoter activity. Truncating the Myo9 b promoter sequence from-1396 to-382 bp did not affect the activation of the promoter by KLF5. However, further deletion between-382 and-287 bp dramatically decreased the activation of this reporter by KLF5, indicating that the TCE site 1 localized at the-319 bp is critical for KLF5-mediated transcriptional activation of Myo9 b. Oligonucleotide pull-down assayshowed KLF5 bound only to the TCE site-1 of the Myo9 b promoter.3 KLF5 deletionreduces podosome rosette formation in macrophages through down-regulating Myo9 b expressionThe formation of podosomes was visualized by co-immunostaining cells for filamentous actin(F-actin) and cortactin, and colocalization of F-actin and cortactin is used to confirm the presence of podosomes in BMMs. Podosome rosettes could hardly be seen in BMMs under normal culture conditions(data not shown), but stimulation with PDBu promoted podosome rosette formation. However, rosette structure in KLF5-/-BMMs could not be observed regardless of the presence of PDBu. In addition, immunofluorescence cell staining with anti-Myo9 b antibody revealed colocalization of Myo9 b with F-actin and cortactin in the podosome rosettes. The number of BMMs containing podosome rosettes was significantly higher in TNF-α-treated cells than in TNF-α-untreated BMMs. Knockout of KLF5 significantly reduced the formation of podosome rosettes regardless of TNF-α treatment. Conversely, the number of BMMs carrying individual podosomes was not influenced in WT and KLF5-/-BMMs treated or not with TNF-α. Myo9 b also colocalized with vinculin and Tks5 upon PDBu and TNF-α treatment, consistent with the above observations of Myo9 b colocalization with F-actin and cortactin. Myo9 a was not colocalized with cortactin and F-actin confirmed the specificity of anti-Myo9 b antibody staining. The defect of podosome formation in KLF5-/-BMMs could be rescued by overexpression of KLF5 as measured by immunofluorescence staining for podosome markers. Myo9 b could hardly be detected through immunostaining with anti-Myo9 b antibody in KLF5-/-BMMs, and that enforced expression of KLF5 in KLF5-/- BMMs increased markedly Myo9 b level. Staining of mouse and human AAA tissues for F-actin and cortactin showed the formation of podosome rosette structures remarkably similar to the ones observed in vitro. These structures were co-stained for the podosome proteins cortactin and F-actin and were not observed in the aortas of KLF5-/- littermates and human control.PartⅢ KLF5-dependent regulation of Myo9 b mediates podosome rosette formation and promotesmacrophage migration and AAA progression via Rho A signaling pathwayObjective: To explore the effect of KLF5-dependent regulation of Myo9b on Rho A signaling pathway.Method: 1 Western blot was used to detect Myo9 b gene expression after Myo9 b wasknockdowned in Raw264.7. Immunofluorescence staining was used to detect podosome rosette formation in macrophages. 2 Boyden chamber transwell migration assay for macrophages. 3Raw264.7 were transfected with si-Myo9 b or si-NC and then KLF5 overexpression. Immunofluorescence staining was performed to detect podosome rosette formation in macrophages. 4 In response to TNF-α, GTPase pull-down and Western blot were used to detect whether TNF-α-induced podosomes and cell migration are related to activation of Cdc42, Rac1 and Rho A in macrophages. 5 The effect of KLF5 knockdown or overexpression on the levels of GTP-bound Rho A, Rac1 and Cdc42 was examined by Westedrn blotting. 6 WT and KLF5-/- macrophages were treated with Rho A specific inhibitor and podosome formation and cell migration were examined. 7 Myo9 b in human AAA tissues was localized and quantified by immunofluorescence staining and q RT-PCR.1 Macrophage-specific knockdown of Myo9 b reduces podosome formation and migration of macrophages induced by TNF-αIn response to TNF-α, the expression level of Myo9 b was markedly attenuated in si-Myo9b-transfected macrophages compared with cells transfected with si-NS. PDBu-elicited formation of the podosome rosettes was significantly decreased in the si-Myo9b-transfected cells regardless of TNF-αstimulation. The results of transwell migration assay showed that the number of macrophages that had migrated through the membrane was significantly less in Myo9b-knocked down cells than in the si-NS-transfected cells regardless of TNF-α treatment. Furthermore, KLF5 overexpression promoted the formation of podosome rosettes in the si-NS-transfected macrophages, and that this effect was abrogated upon Myo9 b knockdown by si RNA.2 Rho A-GTP level increases after knockdown or deletion of KLF5 inmacrophagesThe levels of GTP-bound Cdc42, Rac1 and Rho A were markedly increased within 15 min after TNF-α stimulation in Raw264.7. Among the three small GTPases, the level of GTP-bound Rho A reached a maximum at 30 min and subsequently decreased to base level by 6 h. Importantly, Myo9 b expression was progressively increased as the active GTP-Rho A levels decreased. The effect of KLF5 knockdown or overexpression on the levels of GTP-bound Rho A, Rac1 and Cdc42 in Raw264.7 was investigated. The results showed that overexpression of KLF5 obviously upregulated Myo9 b expression and decreased the level of GTP-bound Rho A in Raw264.7 but did not affect the Rac1 and Cdc42 activation. In contrast, knockdown of KLF5 markedly downregulated Myo9 b expression and increased the level of GTP-bound Rho A in Raw264.7. Next, we used the KLF5-/- BMMs to investigate whether podosome formation and cell migration were affected by Rho A signaling. As expected, when KLF5-/-BMMs were treated with Rho A specific inhibitor, the formation of podosomes induced by PDBu was obviously increased. Instead, inhibition of Rho A signaling by Rhosin slightly increased podosome formation in WT BMMs. Boyden chamber assay showed that the number of macrophages that had migrated through the membrane was significantly increased in KLF5-/-BMMs treated with Rhosin compared with Rhosin-untreated cells.3 Myo9 b is upregulated in human AAA and is localized to macrophages Immunofluorescence staining for Myo9 bshowed that Myo9 b expression was markedly increased in human aneurysm tissues compared with the normal tissues, and that Myo9 b was localized at the MAC2-positive macrophages within human AAA tissues as evidenced by confocal immunohistochemistry performed on aortic tissue sections. Consistent with immunostaining results, real-time q RT-PCR analyses showed that Myo9 b m RNA levels were significantly increased in the AAA samples compared with non-aneurysm tissues.To determined whether there was a correlation between Myo9 b expression and the AAA expansion rates.All patients had the initial diagnosis of an AAA confirmed with a CT scan, the AAA expansion rate was calculated as the change in the maximal anterior-posterior(AP) diameter of AAA during the whole observation period transformed to annual units.Correlation analysis revealed that Myo9 b m RNA expression in abdominal aneurysmal wall was positively correlated with AAA size but, despite a clear trend, KLF5 expression was not significantly associated with AAA size, probably due to more complicated function.Conclusion:1 KLF5 promotes macrophagemigration and AAA formation.2 Myo9 b is a direct target of KLF5 in macrophages.3 KLF5 promotes podosome rosette formation and migrationin macrophage and Myo9 b expressing.4 Myo9 b promotes podosome rosette formation, macrophagemigration and AAA progression via negatively regulating Rho A signaling.