|Home | About | Journals | Submit | Contact Us | Français|
Integrin-linked kinase (ILK) is located at focal adhesions and links the extracellular matrix (ECM) to the actin cytoskeleton via β1- and β3-integrins. ILK plays a role in the activation of kinases including protein kinase B/Akt and glycogen synthase kinase 3β (GSK3β), and regulates cell proliferation, motility and survival.
To determine the function of ILK in vascular smooth muscle cells (SMCs) in vivo.
SM22Cre+IlkFl/Fl conditional mutant mice were generated in which the integrin-linked kinase (Ilk) gene was selectively ablated in SMCs. SM22Cre+IlkFl/Fl conditional mutant mice survive to birth, but die in the perinatal period exhibiting multiple vascular pathologies including aneurismal dilatation of the aorta and patent ductus arteriosus (PDA). Defects in morphogenetic development of the aorta were observed as early as E12.5 in SM22Cre+Ilk Fl/Fl mutant embryos. By late gestation (E16.5-18.5), striking expansion of the thoracic aorta was observed in ILK mutant embryos. Histological analyses revealed that the structural organization of the arterial tunica media is severely disrupted with profound derangements in SMC morphology, cell-cell and cell-matrix relationships including disruption of the elastic lamellae. ILK deletion in primary aortic SMCs results in alterations of RhoA/cytoskeletal signaling transduced via aberrant localization of myocardin related transcription factor (MRTF)-A repressing the transcription and expression of SMC genes which are required for the maintenance of the contractile SMC phenotype.
These data identify a molecular pathway linking ILK signaling to the contractile SMC gene program. Activation of this pathway is required for morphogenetic development of the aorta and ductus arteriosus during embryonic and postnatal survival.
Vascular smooth muscle cells (SMCs) transduce a variety of signals from a highly structured extracellular matrix (ECM) that modulate cellular differentiation, migration and proliferation.1 Adhesion of vascular SMCs to ECM components determines how cells respond to humoral factors, mechanical forces and developmental cues.2, 3 Vascular SMC-ECM interactions are mediated by integrin cell surface receptors.3 In vascular SMCs, integrins localize to membrane-associated dense plaques, which are structurally similar to the focal adhesions observed in cultured cells.4, 5 Dense plaques transmit forces between the contractile apparatus and the extracellular matrix.6 Since integrin cytoplasmic domains generally lack actin-binding sites and enzymatic activity, signaling is implemented through a series of linker proteins including talin, α-actinin, vinculin, and kinases such as focal adhesion kinase ( FAK) and integrin-linked kinase (ILK).7–9
ILK is a 51-kDa protein originally identified in a search of proteins capable of physically associating with the cytoplasmic domain of the integrin β1 subunit.10 ILK was named based upon in vitro studies demonstrating enzymatic activity of its C-terminal kinase domain,11, 12 which was shown to phosphorylate target proteins including protein kinase B (PKB)/Akt, glycogen synthase kinase (GSK) 3β, and myosin light chain.13 ILK plays a critical role in the organization of the actin cytoskeleton via its association with α and β-parvin, paxillin, and the focal adhesion protein vinculin.14 ILK also physically associates with PINCH 1 and 2,15 which in turn interact with Nck2, a SH2/SH3-adapter protein that associates with cytoskeletal-related signaling molecules including Wiskott-Aldrich syndrome protein (WASP) and Pak (p21 activated serine/threonine kinase). ILK deletion in leukocytes modulates the activationof small GTPases, which are also important for membraneruffling and lamellipodia formation, as well as chemokine-triggeredcell movement.16 These findings suggest strongly that ILK plays a role as a molecular scaffold necessary for maintaining the integrity of the integrin-based cell adhesion and signaling complex.
To examine the function of ILK in vivo, investigators have utilized genetic models in multiple species. Deletion of Ilk gene in Caenorhabditis elegans leads to embryonic demise recapitulating the phenotype observed in β integrin knockouts.17 Similarly, mice harboring a null mutation in the Ilk gene exhibit peri-implantation lethality; as ILK is critical for the polarization of the epiblast.18 Endothelial-specific deletion of the murine Ilk gene inhibits vasculogenesis and also confers embryonic lethality.19 Surprisingly, mice harboring cardiomyocyte-specific deletion of the Ilk gene appear phenotypically normal at birth, but during postnatal development develop dilated cardiomyopathy and heart failure.20 Despite the critical role that focal adhesions and ECM play in modulating the response to vessel wall injury, the role of ILK in vascular SMCs has not been defined. Several studies have shown that ILK promotes calcium/calmodulin-independent contraction via its capacity to phosphorylate myosin light chain in vitro.21, 22 While some in vitro studies have concluded that forced expression of ILK promotes SMC proliferation and migration,23, 24 others have concluded that ILK helps to maintain SMCs in a stationary phenotype in the arterial wall.25 A recent study utilized platelet derived growth factor receptor (PDGFR)–B-driven deletion of ILK in multiple vascular wall cells including dermal pericytes and vascular smooth muscle cells, resulting in embryonic demise associated with local hemorrhage and edema first evident around embryonic day (E) 13.5.26
In the studies described in this report, we have examined the function of ILK in vascular SMCs in the developing vasculature in vivo using ILK conditional mutant mice that were intercrossed with mice expressing Cre recombinase under the transcriptional control of the SMC-restricted SM22α promoter. Loss of ILK in murine vascular SMCs results in perinatal lethality due to defects in arterial morphogenesis and structure. ILK conditional mutant mice survive embryonic development, but die in the perinatal period coincident with re-direction of the fetal circulation. Defects in morphogenetic development of the dorsal aorta and branchial arch arteries are observed as early as E11.5. ILK conditional mutant mice exhibit complex vascular pathology including striking aneurismal dilation of the aorta and patent ductus arteriosus (PDA). Histological analyses revealed that the structural organization of the arterial tunica media in large elastic arteries is severely disrupted in ILK conditional mutant mice. The results identify crucial roles for ILK in SMC homeostasis and structural organization of arteries and suggest a possible role for ILK in acquired and congenital human disease.
ILK conditional mutant mice containing a loxP-flanked (floxed) Ilk gene (IlkFl/Fl), have been previously described in detail.19, 27 To delete Ilk in vivo in SMCs, IlkFl/Fl mice were intercrossed with SM22-Cre transgenic mice expressing Cre recombinase under the transcriptional control of the SMC-restricted SM22α promoter (SM22Cre+).28 Inheritance of the SM22-Cre transgene was determined by PCR using the following oligonucleotide primers: forward, 5′-CTCCTTCCAGTCCACAAACGACC-3′; reverse, 5′-GGCGATCCCTGAACATGTCC-3′. Screening of tail DNA for inheritance of the floxed Ilk gene was performed by PCR as previously reported,19 using two primers, 5′-CCAGGTGGCAGAGGTAAGTA-3′ and 5′-CAAGGAATAAGGTGAGCTTCAGA A -3′, for simultaneous detection of wild-type, floxed, and excised Ilk genomic sequences. All animal experimentation was performed under protocols approved by the Massachusetts General Hospital and University of Pennsylvania IACUC and in accordance with NIH guidelines.
Timed matings were set up between male SM22Cre+IlkFl/+ and female IlkFl/Fl mice in a mixed SV129 and C57BL/6 background. The presence of a vaginal plug was considered day 0 of pregnancy. Embryos harvested between embryonic day E11.5 and E18.5 were fixed in 4% paraformaldehyde for 24–48 hours. After gradient dehydration with ethanol, embryos were embedded in paraffin, sectioned, and stained with hematoxylin and eosin, or immunostained with the following antibodies: anti-ILK-1 (# 3862, Cell Signaling Technology), anti-fibronectin (# sc-20084, Santa Cruz), anti-fibrillin-1(# sc-20084, Santa Cruz), anti-smooth muscle(SM)-α-actin (# A5228, Sigma), anti-SM22α,29 and anti-SM-myosin heavy chain (# BT-562, Biomedical Tech). Elastin was detected by van Gieson staining. To assess cell proliferation, transverse serial sections cut at the level of the aortic arch were prepared from E16.5 control and conditional mutant mice and immunostained with the anti-phospho-histone H3 antibody (#9706L, Cell Signaling Technology). The number of phospho-histone H3-positive cells per section was determined (n > 9 sections). Data are reported as mean number of phospho-histone H3-positive cells per section±S.E.M. Detailed histology and immunohistochemistry protocols are available at www.uphs.upenn.edu/mcrc/histology/histologyhome.html.
For immunocytofluorescence studies, SMCs plated on Lab-Tek chamber slides were washed twice with pre-warmed PBS without Ca2+ and Mg2+, then fixed in 3.7% formaldehyde solution in PBS for 10 minutes, followed by permeabilization for 10 minutes with 0.2% Triton X-100 at room temperature. After permeabilization, cells were immunostained with the following antibodies: rabbit anti-ILK (#I1907, Sigma), rabbit anti-Cre (#69050-3, Novagen), mouse anti-SM-α-actin-FITC (#F3777, Sigma), rabbit anti-SM22α (#Ab10135, Abcam), mouse anti-HA-FITC (#H7411, Sigma), and rabbit anti-MRTF-A.30 For labeling F-actin, rhodamine-phalloidin (#R415, Invitrogen) was used. Nuclei were counterstained with DAPI (Molecular Probes). Immunostained SMCs were visualized on a Nikon ECLIPSE ME600 fluorescence microscope.
Total RNA was isolated from whole length aortas from embryos at E18.5 using mechanical homogenization with a roto-stator and RNeasy columns (Qiagen). After DNaseI digestion, equivalent amounts of RNA from each sample were reverse transcribed using QuantiTect reverse transcription reagents (Qiagen). Real-time qPCR reactions were conducted with the Applied Biosystems 7500/7500 Fast Real-Time PCR System. Amplification plots were analyzed with the 7500 software. Gene expression was normalized to GAPDH as an internal control.
To isolate primary mouse aortic SMCs, aortas were isolated (from the aortic root to the iliac bifurcation) from 6-week old IlkFl/Fl mice. The adventitia was removed and the aorta was excised and the intimal endothelium was removed. The remaining tissue was cut into pieces of approximately 2mm2, and then placed as explants in T25 flasks containing 1mL RPMI 1640 with 20% fetal bovine serum (FBS) (3 aortas per flask).31 Once the SMC outgrowth had become >50% confluent, cells were grown and passaged in RPMI 1640 supplemented with 20% fetal bovine serum, 1% penicillin/streptomycin, and 1% glutamine. Cells from passages three to six were used for experiments. Staining with a FITC conjugated anti-mouse-SM-α-actin (Sigma) mAb confirmed >99% SMC purity, with an isotype-matched IgG control (Sigma) verifying specificity.
Replication defective adenoviruses (RdAV) encoding Cre recombinase (Ad-Cre),32 as well as the adenovirus control (Ad-GFP) have been described previously.33 60%–70% confluent SMCs were transduced with the indicated RdAV at a multiplicity of infection (MOI) of ~100 in RPMI 1640 with 2% FBS for 3 hours at 37°C, and the media with adenovirus was replaced with 3 ml of fresh medium supplemented with 20% FBS. Cells were then cultured and analyzed as detailed in the figure legends. For rescue experiments, SMCs were transduced with an adenovirus encoding wild type ILK (AdWT-ILK) for 3 hours prior to co-transduction with Ad-Cre.
Transient transfection analyses were performed as described previously.32, 33 In brief, 24 hours post RdAV transduction, SMCs were transiently transfected with 200 ng of the indicated luciferase reporter plasmid using Fugene 6 (Roche Applied Science) with a Fugene6/DNA ratio of 3:1. The luciferase reporter plasmids Sm22α.luc, Smα-actin.luc (or, pPIAct.luc) and promoterless control pGL3-Basic plasmid have been described previously.30, 34 48 hours post-transfection, SMCs were harvested, and luciferase activities were measured by Steady-Glo Luciferase Assay System (Promega). Each experiment was repeated at least five times. Data are reported as the mean-fold induction in luciferase activity (relative units) ± S.E.M. To determine the nucleocytoplasmic localization of MRTF-A, primary mouse SMCs were transfected with 500 ng of pcR3.1-HA-MRTF-A expression plasmid encoding HA epitope-tagged-mouse MRTF-A fusion protein.30
The ratio of F-actin to G-actin in smooth muscle cells was analyzed using a commercially available kit according to the manufacturer’s protocol (Cytoskeleton Inc., # BK037). Briefly, 72 hours post RdAV transduction, SMCs cultured in 100mm dishes were lysed with 2 ml of a lysis buffer containing 1 mM ATP to stabilize F-actin. The cell lysates were centrifuged at 100,000 g for 1 hour at 37°C using a Beckman ultracentrifuge equipped with an L-80, SW50.1 rotor. The G-actin containing supernatants were then separated from the F-actin containing pellets and were immediately placed on ice. The pellets were resuspended to the same volume as the supernatants using ice cold water containing 10 μM cytochalasin D and were incubated on ice for 1 hour, and sheared every 15 minutes. Equivalent volumes of the corresponding F- and G-actin fractions were loaded into lanes of an SDS-PAGE gel and analyzed by western blot analysis. The ratio of F-actin versus G-actin was quantified using image analysis software.
Data are expressed as means ± S.E.M. Statistical comparison of means was performed by using the two-tailed unpaired Student t test. P values less than 0.05 were considered statistically significant.
Mice harboring a null mutation in the Ilk gene die at the peri-implantation stage,18 precluding assessment of the function of ILK in vascular SMCs. To examine the function of ILK in SMCs in vivo, conditional mutant mice were generated in which the Ilk gene was selectively ablated in SMCs by intercrossing mice in which the Ilk gene has been conditionally targeted (IlkFl/Fl)35 with SM22Cre transgenic mice that express the Cre-recombinase under the transcriptional control of the SMC-restricted SM22α promoter (SM22Cre+).28 Of note, the gene targeting strategy employed results in excision of Ilk exons 5 through 12 following Cre-mediated recombination generating a true null allele.35 Cre-mediated excision of the loxp-flanked Ilk alleles in SM22Cre+ mice was confirmed using PCR-based protocol designed to amplify and distinguish wild-type, floxed, and excised Ilk simultaneously (Online Figure I).
To confirm that excision of the conditional Ilk allele results in a corresponding loss of ILK protein, immunohistochemical analyses were performed on tissue sections harvested from SM22Cre+IlkFl/Fl mutant and IlkFl/Fl control mice at E11.5 (Fig. 1A–D) and E18.5 (Fig. 1E and 1F), respectively. Of note, we have shown that SM22Cre transgenic mice promote highly efficient Cre-mediated recombination as early as E9.5 in vascular and visceral SMCs and in cardiac myocytes.36 At E11.5 in the control mice, abundant ILK protein (brown stain) was observed in the carotid artery (car) and 3rd (3), 4th (4) and 6th (6) branchial arch arteries (Fig. 1A). High-power magnification revealed that ILK is expressed in both medial SMCs and intimal endothelial cells at this developmental stage prior to remodeling of the branchial arch arteries (Fig. 1C, arrows). By contrast, brown staining indicative of ILK expression was not observed in the tunica media of arteries of SM22Cre+IlkFl/Fl conditional mutant mice at E11.5 (Fig. 1B and 1D). As anticipated in control mice at E18.5, ILK protein (brown stain) was readily observed in aortic intima and tunica media (Ao) (Fig. 1E, arrows). Consistent with the SMC-restricted activity of the SM22α promoter,36 ILK was not observed in the tunica media of the aorta in SM22Cre+IlkFl/Fl mutant embryos, but was observed in intimal endothelial cells (Fig. 1F, arrows). ILK was also expressed abundantly in the muscular layer of the esophagus (eso) (Fig. 1E). Because SM22α promoter-driven expression of Cre recombinase also results in gene deletion in embryonic cardiac myocytes,28 we also examined the expression of ILK in the heart. Consistent with previous reports,20 ILK protein (brown stain) was observed in cardiomyocytes of control embryos (Fig. 1G), but it was not detectable in cardiac myocytes of E14.5 SM22Cre+IlkFl/Fl mutant embryos (Fig. 1H).
To examine the function of ILK in vascular SMCs in vivo, SM22Cre+IlkFl/+ males and females were intercrossed. However, their offspring failed to generate the expected Mendelian frequency of inheritance of all genotypes (Online Table 1). Of 146 offspring analyzed at 2-week of age, no viable offspring were observed that carried both the SM22Cre allele and two copies of the targeted Ilk allele (SM22Cre+IlkFl/Fl). Subsequent analysis of 238 embryos harvested between E11.5 and late gestation (E18.5) revealed the expected Mendelian ratio of SM22Cre+IlkFl/Fl mutant embryos (Online Table 2). However, all SM22Cre+Ilk Fl/Fl mutant pups died between E18.5 and postnatal day (P1) in the immediate postnatal period (i.e. mutant embryos were liveborn). These data demonstrate that expression of ILK in SMCs is required for postnatal survival beyond P1.
To determine the cause of perinatal lethality in SM22Cre+IlkFl/Fl mutant pups, SM22Cre+IlkFl/Fl embryos were harvested from E18.5 through parturition (P1). SM22Cre+IlkFl/Fl pups were born alive, but shortly after birth, all SM22Cre+IlkFl/Fl mutant pups became markedly cyanotic and expired within 2–3 hours (Fig. 2A and 2B). Comparison of histological sections of E18.5 SM22Cre+IlkFl/Fl mutant embryos (Fig. 2F–H) and control littermates (Fig. 2C–E) revealed the extent of derangements in the vascular anatomy of the conditional mutant mice. These included marked enlargement of the aortic arch (Ao) extending through the descending thoracic aorta beyond the insertion site of the ductus areriosus (compare Fig. 2C and 2F and Online Figure II). In all mutant embryos, dilatation of the aorta caused displacement of the trachea and esophagus to the right of the midline (Fig. 2G and 2H and Online Fig. II I and J). At this plane of section, the thoracic aorta filled approximately fifty percent of the chest cavity in some embryos (Fig. 2G and 2H). Moreover, the ductus arteriosus (DA) is grossly enlarged at the level of its insertion into the dilated thoracic aorta (Online Fig. II J). Of note, only mild dilation of the ascending aorta (AAo) was observed extending to the origin of the right carotid artery (Online Fig. II J). However, the medial wall of the AAo was consistently thicker in mutant than control embryos.
Vascular casting studies of seven E18.5 SM22Cre+IlkFl/Fl mutant embryos revealed striking enlargement of the proximal aorta and thoracic aorta (DAo) as well as enlargement of the ductus arteriosus (DA) (compare Online Figure II A and C, and Online Table 3). Mechanistically, it is noteworthy that cardiac outflow tract and great arch artery patterning defects attributable to the cardiac neural crest including persistent truncus arteriosus (PTA), double-outlet right ventricle (DORV), right-sided aortic arch (RSAA) were not observed in the SM22Cre+Ilk Fl/Fl mutant embryos. Consistent with these findings, postmortem analysis of 13 SM22Cre+IlkFl/Fl mutant newborn pups confirmed vascular defects recapitulating those observed in E18.5 SM22Cre+IlkFl/Fl mutant embryos (Online Table 3). In all control pups, anatomic constriction and luminal obliteration of the DA was observed within 2–3 hours of birth (Online Fig. II B). By contrast, the DA of SM22Cre+Ilk Fl/Fl mutant pups was widely patent and dilated at the time of death (Online Fig. II D). Given the timing of their demise it is most likely the mice died from the PDA, which was a consistent finding. However, in some mice aortic dissection was also observed. Together, these data demonstrate that ILK is required for morphogenetic development of the aorta and great vessels required for neonatal survival. The magnitude of the PDA in mutant mice explains, at least in part, the observed cyanosis and lethality of SM22Cre+Ilk Fl/Fl mutant pups.
Because expression of the SM22Cre transgene leads to efficient deletion of conditional alleles in SMCs as well as the embryonic heart,36 the hearts of SM22Cre+IlkFl/Fl mutant pups were also examined. Consistent with the phenotype of MCKCre+IlkFl/Fl mice in which the Ilk gene was specifically ablated in cardiac myocytes20, hearts developed normally in the SM22Cre+Ilk Fl/Fl mutant embryos and were generally indistinguishable from those observed in IlkFl/Fl control littermates, at least through E18.5 (data not shown). These data support the conclusion that the observed vascular phenotype is attributable to ILK-deficiency in vascular SMCs as opposed to a primary cardiac defect. In addition, we examined bronchial smooth muscle cells and found no abnormalities that might contribute to the observed phenotype (data not shown).
To begin to examine the mechanism underlying aneurismal dilation of the aorta observed in SM22Cre+IlkFl/Fl mutant mice, we compared the developmental time course of aortic morphogenesis in control and mutant embryos (Fig. 3). The aorta arises from the left 4th branchial arch artery and the left dorsal aorta through a complex series of morphogenetic steps beginning at mid-gestation [for review see37]. At E11.0, the right and left 3rd (3), 4th (4) and 6th (6) branchial arch arteries arise from the aortic sac and insert into the paired right and left dorsal aortae, respectively. At the level of the bifurcation of the trachea, the paired dorsal aortae merge establishing a single midline dorsal aorta. Beginning at E11.5, the branchial arch arteries begin to asymmetrically remodel which is functionally coupled to remodeling of the left and right dorsal aortae (dAo). As schematically depicted in Fig. 3, by E12.0, the right dorsal aorta and right 6th branchial arch artery have regressed, while the left dorsal aorta enlarges merging with the left 4th branchial arch artery resulting in formation of the distal aspect of the arch of the aorta and the descending thoracic aorta.
At E11.0 prior to branchial arch artery remodeling, patterning of the 3rd, 4th and 6th branchial arch arteries in SM22Cre+Ilk Fl/Fl mutant embryos and IlkFl/Fl control littermates was indistinguishable. By E12.5, in control embryos, the right 4th branchial arch artery has regressed significantly, while the left 4th branchial arch artery (L4BA) has increased in caliber giving rise to the arch of the aorta (Fig. 3A). By contrast, in E12.5 SM22Cre+Ilk Fl/Fl mutant embryos, the right 4th branchial arch artery (R4BA) is approximately equal in caliber to the left 4th branchial arch artery (L4BA) (Fig. 3B). In addition, in E12.5 control embryos, the right 6th branchial arch artery (R6BA) has regressed, while the left 6th branchial arch artery (L6BA) becomes the ductus arteriosus (DA) linking the pulmonary trunk to the descending thoracic aorta (Fig. 3C). However, in SM22Cre+IlkFl/Fl mutant embryos, the right 6th branchial arch artery (R6BA) is approximately equal in caliber to the left 6th branchial arch artery (Fig. 3D). At the level of the distal tracheal cartilage, in control embryos, the right (RdAo) and left descending thoracic aorta (LdAo) appears as a single vessel (Fig. 3E). By contrast, in SM22Cre+Ilk Fl/Fl mutant embryos, the right dorsal aorta merges with the left dorsal aorta forming a grossly enlarged arterial structure representing the descending thoracic aorta (dAo) (Fig. 3F). Enlargement of the descending thoracic aorta is observed in SM22Cre+IlkFl/Fl mutant embryos extending through the thoracic cavity (Online Fig. II K and L). These data demonstrate that prior to re-direction of the fetal circulation, enlargement of the aortic arch and descending thoracic aorta are already observed in SM22Cre+IlkFl/Fl mutant embryos. These data reveal an unanticipated role of ILK in morphogenetic development of the aorta. In this regard it is noteworthy, that the observed defects involve anatomic structures containing both neural crest-derived and mesodermally-derived vascular SMCs populating the aortic arch and descending thoracic aorta beyond the insertion of the DA (or L6BA).
Comparison of hematoxylin and eosin (H&E)-stained sections of thoracic aorta harvested from E13.5 to E18.5 SM22Cre+IlkFl/Fl mutant embryos and Ilk Fl/Fl control embryos revealed marked derangements in the structural organization of the arterial tunica media (Fig. 4A and B). In control embryos at E16.5, SMCs display a spindle-like morphology and are densely packed and circumferentially-oriented around the arterial lumen (Fig. 4A). Each layer of SMCs in the tunica media of the aorta contains an underlying layer of ECM containing elastin (Fig. 4C). By contrast, in SM22Cre+IlkFl/Fl mutants, aortic SMCs appear very heterogeneous with obvious alterations in cell size, morphology and orientation as well as marked alterations of cell-cell contact including obvious gaps between medial SMCs (Fig. 4B). Consistent with these findings, the elastic fibers, which normally underlie each circumferentially-oriented layer of vascular SMCs in the aorta, are not present (Fig 4 D). High-power confocal imaging of fibronectin (FN) expression revealed striking disruption of cell-cell and cell-ECM structural organization that typifies a large elastic artery such as the aorta (compare Fig. 4E and 4F). Interestingly, fibrillin-1 (Fbn-1), which has been implicated in aortic aneurysms observed in Marfan’s disease patients,38, 39 was dramatically up-regulated in the aorta of SM22Cre+Ilk Fl/Fl mutant embryos compared to controls (Fig. 4G and 4H). Taken together, these data demonstrate that ILK expression in vascular SMCs plays a critical role in regulating SMC phenotype and cell-cell relationships in the tunica media of large elastic arteries including the aorta which is required for structural integrity of these arteries.
To determine whether ILK influences the vascular SMC phenotype, immunohistochemical analyses were performed with a panel of antibodies which distinguish contractile and synthetic SMC markers. Immunofluorescence staining of arteries from SM22Cre+Ilk Fl/Fl mutant mice demonstrated marked attenuation of SM-α-actin and SM22α which are both associated with a “contractile SMC phenotype” (Fig. 5A–D). Moreover, expression of SM-myosin heavy chain (SM-MHC), a specific and relatively late marker of the SMC lineage, was dramatically down-regulated in SM22Cre+Ilk Fl/Fl mutant mice (compare Fig. 5E and 5F). Real-time qPCR performed on mRNA harvested from the whole aorta of twelve E18.5 ILK conditional mutant embryos and littermate control embryos revealed significant decreases in expression of SM-α-actin, SM22α, SM-MHC, and Elastin mRNA in SM22Cre+IlkFl/Fl mutant as compared to Ilk Fl/Fl control embryos (Fig. 5G), concordant with the immunhistochemical results. Interestingly, focal patches of SM-MHC positive cells were observed within the mutant tunica media (Fig. 5F). The expression of Fbn-1 as assessed by qPCR was not concordant with the striking Fbn-1 immunostaining (dark brown in Fig. 4G and 4H). To determine whether the down-regulation of genes encoding SMC-restricted contractile proteins is accompanied by an increase in SMC proliferation, transverse sections of the aortic arch prepared from E16.5 control Ilk Fl/Fl and SM22Cre+Ilk Fl/Fl mutant embryos were immunostained for phospho-histone H3 (pH3) and the number of pH3-positive aortic SMCs/section was quantified. As shown in Fig. 5H, the mean number of pH3-positive cells per aortic section in SM22Cre+Ilk Fl/Fl mutant embryos demonstrated a statistically significant increase compared to aortic sections prepared from Ilk Fl/Fl control embryos (p < 0.05).
To determine whether ILK signaling altered the vascular SMC phenotype in a cell autonomous manner, primary aortic SMCs were isolated from IlkFl/Fl mice and transduced with either the Ad-Cre32 or control Ad-GFP replication-defective adenovirus (RdAV). As anticipated, Ad-Cre transduction conferred a dose-dependent increase in Cre protein expression, with concomitant diminution of ILK protein expression after 72 hours (Online Fig. III A and B). Decreased ILK expression was accompanied by decreases in immunofluorescent staining for SM22α and SM-α-actin protein (Fig. 6A), similar to the diminished expression noted in vivo. Moreover, the rich array of cytoskeletal actin fibers that typify contractile SMCs was severely attenuated in Ad-Cre transduced IlkFl/Fl aortic SMCs (Fig. 6A).
To determine whether deletion of ILK signaling directly regulated transcription of the SM22α and SM-α-actin genes, Ad-Cre-transduced IlkFl/Fl SMCs were transiently transfected with luciferase reporter plasmids under the transcriptional control of the SM22α promoter (SM22α.luc) or the SM-α-actin promoter/enhancer (Smα-actin.luc), respectively. Expression of the 441-bp SM22a promoter was reduced by 69%, while expression of the SM-α-actin promoter/enhancer was decreased by 71% compared to luciferase activity observed in un-transduced and Ad-GFP transduced IlkFl/Fl SMCs (Fig. 6B). Taken together, both in vivo and ex vivo experiments demonstrate that SMC-specific deletion of Ilk inhibits transcription and expression of genes encoding SMC contractile proteins. These data also suggest that ILK signaling influences cytoskeletal organization of vascular SMCs in a cell autonomous fashion. Of note, SMC proliferation was not significantly altered in ILK-deficient SMC in vitro, as assessed by thymidine incorporation (data not shown).
Both the SM22α and SM-α-actin promoter/enhancers are regulated by an SRF/myocardin-dependent transcriptional regulatory program.34 In vascular SMCs, SRF-dependent transcription and SMC phenotype is influenced by multiple intracellular signals including myocardin-related transcription factors (MRTFs) [for review see1]. Myocardin related transcription factors (MRTF)-A, is a potent transcriptional coactivator which has been shown to transduce integrin signals from the cytoskeleton to the nucleus influencing expression of genes encoding SMC-contractile proteins.1, 40 In contrast to non-muscle cell lineages, MRTF-A localizes predominantly to the nucleus of serum-starved vascular SMCs.1, 41 To determine whether ILK signaling is transduced to the SMC nucleus at least in part via MRTF-A, Ad-Cre-transduced IlkFl/Fl SMCs were transduced with an expression plasmid encoding HA-tagged MRTF-A. As anticipated, in un-transduced SMCs, the HA-tagged MRTF-A fusion protein (green stain) localized predominantly to the nucleus (Fig. 7A and 7B). By contrast, in Ad-Cre-transduced Ilk conditionally targeted SMCs, HA-MRTF-A fusion protein localized predominantly in the cytoplasm (Fig. 7E and 7F). Consistent with these findings, native MRTF-A localized predominantly to the cytoplasm of Ad-Cre transduced cells (Fig. 7G and 7H), but was observed in the nucleus of un-transduced IlkFl/Fl SMCs (Fig. 7C and 7D). These data are consistent with a model wherein ILK-dependent signals are transduced, at least in part, via MRTF-A to the nucleus modulating SMC phenotype. As anticipated, myocardin was observed exclusively in the nucleus. Of note, the effects of ILK deletion on the localization of MRTF-B were more variable (data not shown), and did not demonstrate the striking findings as seen with MRTF-A translocation.
Previous studies have shown that cytoskeletal actin dynamics control SRF-dependent pathways, at least in part by regulating the subcellular localization of its coactivator MRTF-A.41, 42 Therefore, we examined whether ILK signaling regulates F-actin polymerization and/or F-actin disassembly in vascular SMCs. We first quantified the ratio of filamentous actin (F-actin) to monomeric actin (G-actin) by differential centrifugation and Western blotting. These studies demonstrated that deletion of ILK significantly reduced the amount of F-actin in ILK-deficient cells (Fig. 8A and 8B). IlkFl/Fl SMCs were also transduced with Ad-Cre or left un-treated, and then stained with rhodamine-phalloidin to visualize cytoskeletal actin filaments. As anticipated, in un-transduced or Ad-GFP transduced SMCs, a rich array of stress fibers (Fig. 8C and 8D) was observed. By contrast, F-actin staining of Ad-Cre transduced SMCs demonstrated an absence of organized actin stress fibers, with clumps of staining concentrated at the periphery of cells, which were sometimes retracted (Fig. 8E and 8F). Next, we examined whether overexpression of wild-type ILK (AdWT-ILK) could rescue the cellular pathology observed in the cells rendered ILK-deficient by Ad-Cre. As seen in Fig. 8G, forced expression of ILK rescued the striking F-actin and stress fiber abnormalities as assessed by rhodamine-phalloidin staining.
Prior studies have demonstrated that genes encoding SMC contractile proteins, such as SM22α and SM-α-actin, are regulated by RhoA-mediated actin polymerization.43 We therefore tested whether deletion of ILK in vascular SMCs modulated the activation of RhoA by measuring the active GTP-bound form of the molecule. Indeed, we observed a significant decrease in basal levels of GTP-bound RhoA in SMC transduced with Ad-Cre as compared to un-transduced or Ad-GFP transduced control cells (Online Fig. IV). Taken together, these data are consistent with a molecular model wherein ILK signaling activates RhoA, which in turn influences cytoskeletal dynamics, promoting the nuclear localization of MRTF-A and activation of SRF-dependent SMC contractile genes.
The capacity of vascular SMCs to modulate their phenotype in response to environmental stimuli plays a critical role in the homeostatic response of the cardiovascular system. ILK is ideally positioned in the vascular SMC to transduce extracellular signals received from the extracellular membrane through β-integrin cell surface receptors.44 It plays a role in the phosphorylation of vital substrates in the SMC which regulate cell morphology, migration, contraction, proliferation and survival.13, 45, 46 Therefore, it was of interest to examine the vascular phenotype of mice in which the Ilk gene was selectively ablated in vascular SMCs. Surprisingly, Ilk conditional mutant mice survived to birth, but succumb in the immediate postnatal period coincident with redirection of the fetal circulation. Ilk conditional mutant mice exhibit profound abnormalities in structural organization of great arteries including the aorta and ductus arteriosus. Defects in morphogenetic development of the aorta were observed as early as E12.5. Examination of large elastic arteries in Ilk mutant embryos and pups revealed profound disruption of the tunica media. SMC-specific deletion of ILK alters RhoA activation and cytoskeletal dynamics which is transduced, at least in part, to the nucleus via MRTF-A attenuating the transcription and expression of SMC genes associated with the contractile SMC phenotype. These data demonstrate that ILK plays a critical role in morphogenetic development of large elastic arteries and that ablation of the Ilk gene in vascular SMCs severely disrupts the structural integrity of arterial structures required for postnatal survival.
Prior studies in genetically targeted Ilk mice revealed a critical role for ILK in cardiovascular biology. Endothelial-specific deletion of the murine Ilk gene inhibits vascularization and results in embryonic lethality.19 Targeted ablation of ILK in the murine heart ultimately results in a dilated cardiomyopathy and congestive heart failure.20 With regard to SMC biology, one prior study has demonstrated that angiotensin II increases ILK protein expression and kinase activity in vitro.24 Adenoviral gene transfer experiments using a dominant-negative construct suggested that ILK is necessary for angiotensin II-mediated SMC migration and proliferation in vitro. Other investigators have demonstrated that ILK may modulate calcium-independent myosin-mediated contraction of triton-skinned rat caudal artery preparations. More recently, it was also demonstrated that ILK protein levels were decreased following balloon injury of the rat carotid artery.25 ILK knockdown with a RNA silencing led to augmented cell movement and enhanced wound closure in an in vitro model, which contrasts with the prior published study which employed adenoviral over-expression constructs and found diminished SMC migration.24 The present study thus extends the prior work by specifically defining a role for ILK in the morphogenetic development of the aorta and in the structural integrity of large elastic arteries in vivo. In this regard it is noteworthy that alterations in development of the aorta may result from vascular SMC autonomous defects directly attributable to ILK signaling and/or indirectly via alterations in blood flow resulting from alterations in vascular SMC phenotype/contractility or vascular SMC-ECM interactions.
A recent study evaluated platelet derived growth factor receptor (PDGFR)–B-driven deletion of ILK in multiple vascular wall cells including dermal pericytes and vascular smooth muscle cells, resulting in embryonic demise associated with local hemorrhage and edema first evident around embryonic day E13.5.26 While extensive pathology in the embryonic dermal vasculature was noted, no pathology of the great vessels was described using the PDGFR-B-driven deletion of ILK in mice. Of note, the investigators also described a mechanism by which ILK knockdown in immortalized, murine SMCs activated Rho/ROCK signaling and induced the phosphorylation of myosin light chain, and abnormally enhanced VSMC contraction in vitro and in vivo. However, the observation of down-regulation of contractile SMC genes in the mutant ductus arteriosus (and aorta) of SM22Cre+/IlkFl/Fl mice raises questions about the relevance of this finding in vivo. Undoubtedly, the use of different cell types and promoters contribute to these different findings, but together the data underscore a clear role for ILK in SMC homeostasis.
Our in vivo and in vitro results demonstrate that SMC-specific deletion of ILK represses the transcription and expression of SMC genes which are required for the maintenance of the contractile SMC phenotype. A particularly novel finding in ILK deletion in SMC in vitro relates to the observed effects on MRTF–A, which is believed to transduce Rho GTPase-dependent signaling from the cell membrane to the nucleus, and is critical to maintenance of the contractile SMC phenotype.1 Our present studies are consistent with prior work demonstrating that ILK deletion in leukocytes modulates the activationof small GTPases, which are also important for membraneruffling and lamellipodia formation, as well as chemokine-triggered cell movement. It has also been shown that pharmacological agents that inhibit actin polymerization and/or forced expression of nonpolymerizing actin mutant proteins stimulate the export of MRTF-A.1 However, disruption of MRTF-A signaling alone cannot explain the vascular derangements observed in ILK conditional mutant mice, as mice harboring a null mutation in MRTF-A appear phenotypically normal. This is not surprising as activity of SRF is modulated by multiple overlapping and redundant signaling pathways including signals transduced via the related transcriptional co-activators myocardin and MRTF-A.1
Additional abnormalities observed in the ILK-deficient mice may also be contributing in causal pathways contributing to the vascular pathology. Confocal imaging of the vessel wall demonstrated striking disruption of the extracellular matrix, consistent with prior studies demonstrating a role for ILK in FN matrix assembly.47 Matrix abnormalities could, which in turn, be influencing cell migration and proliferation in the ILK-deficient aortas. The observed changes in elastin staining are similarly notable, since elastin is known to affect actin polymerization through an integrin-independent but a RhoA-dependent pathway.48 It is tempting to speculate that these disruptions of vascular SMC-ECM signaling in SM22Cre+Ilk Fl/Fl mutant embryos may have led to compensatory expression of fibrillin-1 highlighting the complex, but critical role that ECM signaling plays in regulating development of the cardiovascular system.
While our studies provide unequivocal evidence that ILK plays a critical role in the morphogenetic development and structural organization of the aorta required for survival beyond the perinatal period, the role of ILK in maintenance of cardiovascular homeostasis remains to be determined. Given these data and the critical role that ILK plays in transduction of external signals and the phenotypic abnormalities observed in ILK-deficient aortic SMCs, it is tempting to speculate that ILK may be involved directly or indirectly in the pathogenesis of vascular proliferative syndromes including atherosclerosis and/or the pathogenesis of congenital or acquired vascular aneurysms. In support of this hypothesis, we have recently observed that ILK expression is markedly diminished in a mouse model of abdominal aortic aneurysm disease (Online Fig. V). In any case, further studies examining the function of ILK in the postnatal vasculature are warranted.
The molecular pathways that orchestrate the formation of the major arteries remain incompletely defined. To examine the function of ILK in SMCs in vivo, conditional mutant mice were generated in which the Ilk gene was selectively ablated in SMCs by intercrossing mice in which the Ilk gene has been conditionally targeted with SM22Cre transgenic mice that express the Cre-recombinase under the transcriptional control of the SMC-restricted SM22α promoter. The phenotyping of these mice revealed multiple vascular pathologies including aneurysmal dilatation of the aorta and patent ductus arteriosus. These studies define a previously unanticipated molecular pathway linking ILK signaling to the contractile SMC gene program. Activation of this pathway is required for morphogenetic patterning of the aorta and ductus arteriosus during embryonic and postnatal survival. Further, our data suggest that ILK may be involved directly or indirectly in the pathogenesis of vascular proliferative syndromes including atherosclerosis and/or the pathogenesis of congenital or acquired vascular aneurysms.
The authors gratefully acknowledge support from the NIH to REG (R01HL-65584), JDR Jr. (R01HL-96779), and R01HL-094520 and R01HL-102968 to MSP. We also acknowledge the excellent technical assistance of Melinda Pal
Sources of Funding
The authors gratefully acknowledge support from the NIH to REG (R01HL-65584) and R01HL-094520 and R01HL-102968 to MSP.