Molecular Medicine |
From the Department of Molecular Physiology and Biological Physics (L.J., A.V.S.), Department of Medicine (B.R.W.), Cardiovascular Division, University of Virginia, Charlottesville; Department of Cell Biology and Anatomy (M.S.K.), Medical University of South Carolina, Charleston; Department of Cell and Molecular Physiology and the Neuroscience Center (C.A.O.), University of North Carolina at Chapel Hill, Chapel Hill.
Correspondence to Avril V. Somlyo, PhD, University of Virginia, Department of Molecular Physiology and Biological Physics 1300 Jefferson Park Avenue P.O. Box 800736 Jordan Hall, Charlottesville, VA 22908-0736. E-mail avs5u{at}virginia.edu
| Abstract |
|---|
|
|
|---|
-actin, was increased by angiotensin II, regulated by actin dynamics, upregulated by myocardin and expressed in the neointima of injured aorta. Overall, the data suggest that the function of LPP and palladin is context dependent, that they play a critical role in cytoskeletal remodeling, respond to signals induced by vascular injury as well as signals that induce smooth muscle cell hypertrophy, such as angiotension II.
Key Words: lipoma preferred partner migration palladin smooth muscle vascular injury
| Introduction |
|---|
|
|
|---|
SM
-actin (SMA) and SM myosin heavy chain (MHC), expression are down regulated by alterations in the actin cytoskeleton through inhibition of the RhoA/ROCK signaling pathway.6 LPP expression is also downregulated by C3 or ROCK inhibition.2 Therefore we determined whether expression of these focal adhesion proteins as with the cytoskeletal proteins actin and myosin are also regulated by changing the ratio of globular to filamentous actin (G:F) and are thus sensitive to cytoskeletal dynamics such as occur during phenotypic switching.
Angiotensin II (Ang II) plays a major role in the hypertrophy of the vessel wall that occurs in chronic hypertension and also increases contractile protein synthesis and hypertrophy in cultured aortic SMCs.7 Myocardin and paired-related homeobox gene-1 protein (Prx1) have been shown to contribute to Ang II-induced expression of SMA.7 Prx1/Prx2 combined null mutant mice display vascular abnormalities of the great vessels8 and both genes are induced with vascular disease.9 Prx1 genes are regulated by SM adhesion and regulate expression of the extracellular matrix protein tenascin-C (TN-C).9 Both Prx1 and TN-C expression levels are reduced in focal adhesion kinase (FAK) null cells where FAK induced expression of Prx1 is thought to promote TN-C-dependent fibroblast migration.10 We have previously reported that LPP is markedly decreased in FAK null cells and both LPP and cell migration are rescued by inducible expression of FAK.2 Therefore in the present study we tested the hypothesis that FAK regulates expression of Prx1, which in turn regulates expression of palladin, as well as LPP and that expression of these focal adhesion proteins is responsive to Ang II, thus playing a role in SMC migration and Ang II-induced hypertrophy.
| Materials and Methods |
|---|
|
|
|---|
Cell Lines, Transient Transfections, Immunoprecipitation, and FAK Phoshorylation
HEK-293, HIVS-125, FAK null or FAK wild-type cells were transfected with calcium phosphate, lipofectamine or electroporation as described in the online data supplement.
In Vitro Pull Down Assays
Protein-protein interactions were followed by affinity chromatography using glutathione sepharose beads to pull down GST or myc tagged LPP and palladin constructs (see the online data supplement available at http://circres.ahajournals.org).
Cell Migration and Spreading Assay
Cell migration assays are described previously,2 cell spreading as a measure of cell migration is detailed in the online data supplement.
Lamellipodia Purification
Lamellipodia were allowed to extend through 3.0 µm pores of a transwell membrane and isolated (see the online data supplement).
Rat Aorta Injury Model
Vascular injury was performed in Sprague Dawley rats (7 to 9w) with a balloon catheter introduced through an arteriotomy in the external carotid artery (see the online data supplement).
| Results |
|---|
|
|
|---|
Tissue Distribution of LPP and Palladin
Both LPP and palladin a 90 to 92kDa doublet are highly and selectively expressed in mouse smooth muscle containing tissues such as ileum, stomach, vas deferens and blood vessels as well as in cultured (R518) SMCs (Figure 1). LPP in the lung originates from SMCs in the airway.2 Unlike LPP, both a 140 and the 90 to 92kDa isoforms of palladin are ubiquitously expressed in embryonic and neonatal tissue but like LPP, palladin (only the 90 to 92 kDa isoform) is most abundant in SMCs of adult tissue.10 The lack of palladin but not LPP expression in the lung could relate to the well known different mechanical characteristics and responses to injury of pulmonary SMCs.
|
LPP and Palladin Interacting Domains
Evidence for the interaction of these two proteins was obtained by in vitro pull down assays using GST-tagged full length and C-terminal domains of LPP bound to glutathione beads and recombinant myc-tagged full length, N-terminal and C-terminal palladin. Full length and N-terminal palladin bound with full length as well as C-terminal LPP (Figure 2B).
|
LPP and palladin interaction was also assessed in vivo. GFP-LPP, GFP-C terminal LPP (412 to 612 aa), myc-palladin, myc-N terminal palladin, myc-C terminal palladin were transiently coexpressed in HEK-293 cells. Both full length and N terminal palladin, but not C terminal palladin, were immunoprecipitated with the GFP antibody (Figure 2A). Both full length and the C-terminus of LPP, but not N-terminal LPP (data not shown), were pulled down by myc antibody. Neither protein was found using an IgG control. Consistent with the in vitro data, the results suggest that the in vivo interaction site resides in the C terminal 200 aa of LPP and the N-terminal 1 to 326aa of palladin.
As LPP can be structurally and functionally divided into a pre-LIM proline rich amino terminal and carboxy terminal consisting of three zinc finger LIM domains,11 we further defined which of the three LIM domains of LPP interacts with the palladin protein. Four GFP C-terminal LPP fragments, representing LIM 1 (aa 412 to 471), LIM 2 (aa 471 to 531), LIM 3 (aa 531 to 612), and LIM1 and 2 (aa 412 to 531) were synthesized. These constructs were cotransfected with full-length myc-palladin in HEK-293 cells for 72 hour. The cleared supernatant from cell lysates were incubated with myc antibody, and then applied to protein G sepharose beads, and Western blots were performed. LPP LIM domains 1 and 2 but not 3 interact with the N-terminus of palladin (Figure 2C). As reported previously, palladin contains proline-rich motifs FPLPPP and FPPPP2,12 highly conserved in human and mouse, and known to function as protein interaction domains. To test whether these motifs represent functional LPP binding sites, palladin proline-rich motifs were mutated and binding to LPP examined. The FPPPP, which begins at aa 82, was mutated (P83D, P84D and P85D) either independently or in concert with the FPLPPP, which begins at aa102 and was mutated at P105D, P106D, and P107D. Mutated GFP-palladin constructs were cotransfected with myc-LPP in HEK-293 cells and the interaction tested by immunoprecipitation. Mutated palladin still bound LPP (Figure 2D). Both domains mutated in concert and a GFP antibody was used to immuoprecipitate the GFP tagged palladin mutants and again LPP was found to coimmunoprecipitate (Figure 2D). The IgG controls were negative. These results suggest that the FPPPP and FPLPPP domains of palladin do not bind LPP, or if so the binding is weak. Because these mutations did not abolish palladins N-terminus interaction with LPP, it suggests that the major interaction domain is upstream of aa 81.
LPP and Palladin Colocalize at Focal Adhesions in Cultured Cells and Tissue
Overexpressed dsRed-palladin colocalized with GFP-LPP to cell adhesions (Figure 3A), but in addition palladin was distributed with periodicity along stress fibers as previously described.13The mutated polyproline GFP-palladin constructs described above gave the same distribution pattern (data not shown), supporting the immunoprecipitation results (Figure 2). Besides the distribution along stress fibers, palladin also differed from LPP expression because it did not translocate to the nucleus of rat aortic SMCs and DU17 fibroblasts (supplemental Figure I in the online data supplement available at http://circres.ahajournals.org) with H1152 inhibition of Rho kinase, unlike the translocation of LPP.1 LPP11 but not palladin has a nuclear export signal. Thus, although these partner proteins bind and colocalize at focal adhesions, they appear to also have distinct expression patterns and likely functions.
|
An association of LPP and palladin was also found in tissues where a punctuate pattern of palladin and LPP staining occurred at the plasma membrane dense body sites (Figure 3B). Further evidence for this association was the colocalization of LPP, palladin and
-actinin when human iliac SMCs in suspension, lacking focal adhesions, spread and began to attach, focal adhesions formed (monitored by
-actinin labeling). Both LPP and palladin colocalized at these sites within 20 minutes of plating before stress fiber formation (Figure 3C).
Regulation of the Expression of LPP and Palladin by Actin Dynamics
Jasplakinolide treatment to stabilize filamentous (F) actin, led to a significant increase (P<0.001) in protein expression levels of both LPP and palladin at 15 hour, while actin depolymerization with latrunculin B treatment decreased (P<0.001) expression (Figure 4A). Immunofluorescence studies at 5 to 10 minutes showed only slight changes with some loss of palladin periodicity along stress fibers and a slight increase in cytosolic LPP signal but LPP and palladin remained colocalized at focal adhesions. By 30 minutes LPP and palladin associated with large bundles or clumps of F actin at the cell periphery or perinuclear region (Figure 4B). Depolymerization of actin by latrunculin B resulted in a loss of palladin periodicity with the loss of stress fibers, LPP and palladin remained colocalized to focal adhesions, but by 30 minutes both proteins were diffuse in the cytoplasm and at small focal adhesions at the cell periphery (Fig 4C). Treatment of R518 SMCs with the Rho kinase inhibitor, H1152 (0.5 µmol/L) resulting in a loss of stress fibers and focal adhesions as expected1 and a weak nuclear distribution of LPP. A dramatic increase in nuclear LPP occurred in DU17 fibroblasts (supplemental Figure I). These experiments suggest that both protein expression and the cellular distribution of LPP and palladin can be regulated by actin dynamics.
|
Regulation of LPP and Palladin by FAK and Prx1
In our previous study, we have shown that expression of LPP was markedly reduced in FAK null fibroblasts, whereas the other focal adhesion proteins, vinculin and paxillin were elevated.2 In the present study, like LPP, the expression of palladin as well as Prx1a and Prx1b proteins are also decreased in FAK null cells and these proteins can be induced to near normal wild-type levels when FAK expression is turned on using a tetracycline repression system6 (Figure 5A). With tetracyclin, some Prx1 protein was expressed compared with wild-type cells, likely reflecting incomplete repression by the tetracycline or perhaps a more stable Prx1 message. Interestingly, the expression of both LPP and palladin is also increased when Prx1a or 1b proteins are overexpressed in the FAK null cells (Figure 5B). Thus the expression of both LPP and palladin is regulated, at least partially by FAK and the transcription factors Prx1a and 1b.
|
Ang II Induces the Expression of LPP and Palladin
It has been shown that Ang II treatment increases SMA and SM-MHC expression in rat aortic SM cells (R518) through myocardin and Prx1 pathways.7 LPP has been proposed as a SM specific marker.2 To examine whether LPP and palladin expression were also regulated by Ang II, protein expression was measured by Western blotting. Both LPP and palladin, like SMA, were increased
2.5-fold after Ang II treatment for 8 hour, expression of GAPDH was unchanged indicating the selective upregulation of protein expression (Figure 6A). The expression of FAK tyr 397 was maximal at 5 to 10 minutes and returned to basal levels by 60 minutes (data not shown) consistent with a previous study.14 Expression of myocardin using an adenovirus vector increased palladin (41%), LPP (27%) and SMA (28%) but not GAPDH expression (Figure 6B) suggesting that LPP and palladin share transcriptional regulatory mechanisms common to other SMC-selective gene expression programs.
|
Both LPP and Palladin Enhance Cell Migration and Cell Spreading
HIVS-125 cells were transfected by electroporation with EGFP-human LPP, EGFP-palladin or control EGFP plasmids as previously described,1 seeded onto Transwell culture inserts and the number of cells that migrated to the lower chamber in response to EGF, used as chemoattractant, were counted. The total number of migrating cells was significantly greater in cells transfected with EGFP-LPP (2.0-fold increase P<0.01) and in cells transfected with EGFP-palladin (1.8-fold increase P<0.05) than in cells transfected with EGFP alone. (Figure 6C, left panel)
To further test the function of LPP and palladin in cell migration, EGFP-LPP, EGFP-palladin were transfected into nonmotile FAK null embryonic fibroblasts. Both LPP and palladin increased cell spreading (Figure 6C middle panel). Expression of LPP or palladin in FAK null fibroblasts also increased EGF stimulated cell motility followed by time-lapse video imaging from 10 to 28 µm/h (P<0. 05) for LPP and from 10 to 35 µm/h (P<0.05) for palladin (Figure 6C right panel). As a control, expression of FRNK (FAK related nonkinase), another focal adhesion protein did not change cell motility.
Both LPP and Palladin Are Expressed in Isolated Lamellipodia
Lamellipodia have extensively branched arrays of growing actin filaments, oriented with their plus (barbed) ends toward the plasma membrane. Lamellipodia growth is a dynamic process involving changes in focal adhesions and actin filaments. Based on the ability of LPP and palladin to increase cell motility,2,15 we proposed that both LPP and palladin would be detected in lamellipodia. Using the Boyden chambers with 3 µm pores pure lamellipodia fractions were obtained and indeed both LPP and palladin protein were detected by western blot (Fig 6D) and immunofluorescence (supplemental Figure II).
Expression of Palladin and LPP Following Rat Aortic Artery Injury
LPP and palladin antibodies strongly labeled SMCs of the tunica media of normal and of vessels 7 days post injury. Both LPP and palladin were expressed in the SMCs, which also express SMA and migrate into the neointima (Figure 7). LPP and palladin were detected in the muscle layers of microvessels in the adventitia but not in the adjacent tissue. The high expression of LPP and palladin in SMCs and their ability to enhance migration in in vitro assays as well as their presence in SMCs in the neointima suggest a role for these molecules in cytoskeletal remodeling of SM when they transition to migratory cells.
|
| Discussion |
|---|
|
|
|---|
Palladin is a key regulator of actin organization and localizes in focal adhesions, cell-cell junctions, dorsal and circular ruffles, and periodically along stress fibers where it colocalizes with
-actinin, and is required for the maintenance of normal stress fibers and focal adhesions in cultured cells.13,16 Inactivation of palladin leads to embryonic lethality because of severe defects of cranial neural tube closure and germination of liver and intestine.17 Because of their early death, the importance of palladin in the development of the vasculature is not known but its ability to enhance cell migration and its presence in dedifferentiated SMCs suggests a significant role.
The additional periodic localization of palladin but not LPP along stress fibers likely reflects their different interaction domains such as the IgC2 repeats present in palladin11 but not LPP or the ability of the 3 tandem LIM domains of LPP to bind different partners. Note that palladin binds LIM1 and 2 of LPP but not 3 and that structural studies have shown that adjacent LIM domains can have different binding surfaces.18 Vinculin is another protein with an FPPPP domain like palladin but is found only at focal adhesions and surface dense bodies yet is known to bind VASP (for Ena/vasodilator-stimulated phosphoprotein), at its EVH1 domain where LPP also binds.19 VASP which inhibits actin capping localizes to both focal adhesions and periodically along stress fibers and has been shown to interact with
-actinin, profilin, palladin and LPP, yet LPP segregates to focal adhesions. Thus, multiple interaction domains with different binding affinities must control localization. In addition, focal adhesions are considered to be dynamic structures with their proteins displaying different dynamics20 and their assembly and disassembly regulated by signaling pathways. Interestingly, in FAK null cells, defective in cell migration, paxillin, vinculin and talin are not decreased at focal adhesions, unlike LPP and palladin.2 The ability of LPP or palladin to rescue migration in these cells suggests that they play a role in the disassembly process and this remains to be explored.
Overexpression or downregulation of LPP respectively enhanced or decreased migration of SMCs2 and we now show that increased expression of palladin also significantly increases migration of human iliac vein SMCs; thus both of these proteins may be important for the migration of smooth muscle cells following vessel injury. Futhermore, palladin is upregulated in astrocytes, closest to the wound in injured rat cortex, where it has been suggested to function as a critical scaffold that directs localization, assembly and organization of other cytoskeletal proteins.13 We have previously shown that SMCs which migrated into the neointima in stent injured pig coronary vessels express LPP as well as smoothelin and SMA which are not detectable in advential fibroblasts consistent with their arising from the vessel media.2 As only one time point was possible in this pig model and palladin localization was not examined, a more extensive study following LPP and palladin expression in relation to SMA, over time, following aorta injury in the rat was investigated in this study. As shown in Figure 7, both LPP and palladin, like SMA were expressed in the media of the vessels, and appeared in the neointima following 7 days after injury. Thus, we suggest that these molecules, which play a structural role at the dense body sites where actin filaments are anchored to the cell membrane, respond to the signals generated by injury and are important for the turnover of focal adhesion complexes and the ability of the SMCs to become migration competent. Taken together, we propose that complementation of LPP and palladin, has the potential to play a significant role in processes involving SMC migration and in the definition of the SMC phenotype in the normal state as well as in the transition on injury.
FAK plays an important role in promoting cell invasion, and embryonic cells from FAK null mice display reduced cell motility with enhanced expression of vinculin, talin, or paxillin and increased focal adhesion contact formation.21 Interestingly, both LPP and palladin protein levels were decreased in FAK null cells and inducible expression of FAK in fibroblasts derived from FAK null embryos lead to marked expression of LPP and palladin, and enhanced cell spreading. Neither a transient nor maintained decrease in RhoA/ROK signaling accounts for the reduced LPP expression as GTP·RhoA content was actually increased in FAK null cells (data not shown), as reported by others.22 Surprisingly, induced expression of FAK using a tetracycline repression system in FAK-null fibroblasts led to a marked expression of LPP and palladin. FAK has been reported to induce expression of the homeodomain protein Prx1,10 which has been shown to increase SRF binding to CArG elements before the recruitment of myocardin in SMCs in response to Ang II stimulation7 Thus, the expression of both LPP and palladin can be regulated through FAK and Prx1.
Ang II plays an important role in remodeling of the arterial wall during development of vascular hypertrophy associated with hypertension. Inhibition of Ang II down regulates the SMC medial hypertrophy in a hypertension model.23 Ang II can increase protein synthesis and induce cellular hypertrophy in cultured aortic SMCs24 and can activate SMC marker gene expression through increased [Ca2+]I which leads to increased myocardin expression,25 as well as Prx1. Prx1 dramatically increases SRF binding to the highly conserved degenerate CArG elements found in the promoter region of most SMC marker genes.26,27 Nucleotide analysis of LPP and palladin promoters indicate the presence of CArG elements. However, Prx1 may also directly regulate transcription as analysis of the LPP and palladin promotor regions reveal phylogenetically conserved Prx1/2 binding sites (see online data supplement). In this study, we have shown that the expression of LPP and palladin proteins were up regulated by Ang II in a dose dependent manner and that overexpression of myocardin increases the expression of palladin and LPP. This coupled with the ability of Prx1 to upregulate LPP and palladin expression is consistent with their playing a role in Ang II-induced hypertrophy such as occurs in hypertension. In addition, Ang II as well as injury leads to increases in vessel wall tenascin-C (TN-C), the antiadhesive and growth promoting extracellular matrix protein which destabilizes the cell-matrix interactions promoting cell migration.28 As FAK expression increases Prx1 which has been shown to increase TN-C expression,15 we propose that the activation of the FAK/Prx1 pathway by Ang II or injury promote TN-C expression, which destabilizes the extracellular matrix setting the stage for cell migration and hypertrophy (Figure 8). Ang II also increases expression of LPP and palladin needed for the adhesion as well as the focal adhesion turnover necessary for cell migration or the growth promoting activities of hypertrophy.
|
Inhibition of the RhoA pathway in fibroblasts and SMCs significantly decreases the activity of SM specific promoters and this activity was shown to be sensitive to the extent of actin polymerization.6,29 Rho kinase inhibitors decreased LPP expression in SMCs or A404 cells differentiated by retinoic acid,1,2 which may be because of the change in actin polymerization or other RhoA mediated signaling pathways. In the present study of rat aortic SMCs, both LPP and palladin expression increased or decreased respectively by increasing or decreasing F to G actin ratios. Miralles et al30 have shown that the SRF cofactor, MRTF-A a member of the Myocardin Related Transcription Factors is regulated by actin dynamics translocating from the cytoplasm to the nucleus where it regulates SMC specific promotor activity.31 Therefore we expect that the reponsiveness of LPP and palladin expression to actin dynamics may also be mediated through MRTFs.
Taken together, we show that LPP and palladin proteins, highly enriched in SM tissues, respond to environmental cues switching from a static structural role in the contractile phenotype to an active role in the adhesion turnover of migrating SMCs following vessel injury or vasculogenesis and in the signaling mechanisms mediating SM hypertrophy, such as Ang II.
| Acknowledgments |
|---|
Sources of Funding
Supported by NIH grants PO1 HL48807 and PO1 HL 19842.
Disclosures
None.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
Related Article:
This article has been cited by other articles:
![]() |
M. M.R. Petit, H. Lindskog, E. Larsson, P. Wasteson, E. Athley, S. Breuer, M. Angstenberger, D. Hertfelder, E. Mattsson, A. Nordheim, et al. Smooth Muscle Expression of Lipoma Preferred Partner Is Mediated by an Alternative Intronic Promoter That Is Regulated by Serum Response Factor/Myocardin Circ. Res., July 3, 2008; 103(1): 61 - 69. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. D. S. Dixon, D. K. Arneman, A. S. Rachlin, N. R. Sundaresan, M. J. Costello, S. L. Campbell, and C. A. Otey Palladin Is an Actin Cross-linking Protein That Uses Immunoglobulin-like Domains to Bind Filamentous Actin J. Biol. Chem., March 7, 2008; 283(10): 6222 - 6231. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. L. Jones Move On!: Smooth Muscle Cell Motility Paired Down Circ. Res., March 30, 2007; 100(6): 757 - 760. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Research Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2007 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |