Glia Maturation Factor- Is Preferentially Expressed in Microvascular Endothelial and Inflammatory Cells and Modulates Actin Cytosk
http://www.100md.com
Koji Ikeda, Ramendra K. Kundu, Shoko Ike
参见附件。
the Donald W. Reynolds Cardiovascular Clinical Research Center (K.I., R.K.K., T.Q.), Division of Cardiovascular Medicine, Stanford University School of Medicine, Calif
Department of Clinical Pharmacology (M.K.), Kyoto Pharmaceutical University, Japan
Department of Cardiology and Vascular Regenerative Medicine (K.I., S.I., H.M.), Kyoto Prefectural University of Medicine, Japan.
Abstract
Actin cytoskeleton reorganization is a fundamental process for actin-based cellular functions such as cytokinesis, phagocytosis, and chemotaxis. Regulating actin cytoskeleton reorganization is therefore an attractive approach to control endothelial and inflammatory cells function and to treat cardiovascular diseases. Here, we identified glia maturation factor- (GMFG) as a novel factor in actin cytoskeleton reorganization and is expressed preferentially in microvascular endothelial and inflammatory cells. During mouse embryogenesis, GMFG was expressed predominantly in blood islands of the yolk sac, where endothelial and hematopoietic cells develop simultaneously. In endothelial cells, GMFG was colocalized with F-actin in membrane ruffles and was associated with F-actin assessed by actin co-sedimentation assay. Interestingly, GMFG was phosphorylated at N-terminal serine, and its phosphorylation was enhanced by coexpression of dominant active Rac1 and Cdc42. Furthermore, a pseudophosphorylated form of GMFG (GMFG-S2E) demonstrated higher association with F-actin. Stable expression of GMFG-S2E remarkably enhanced stimulus-responsive lamellipodia and subsequent membrane ruffle formation in HeLa cells presumably through its interaction with Arp2/3 complex. Expression of GMFG enhanced actin-based cellular functions such as migration and tube-formation in endothelial cells. Moreover, we found that GMFG expression was significantly increased in a cardiac ischemia/reperfusion model where inflammation and angiogenesis take place actively. Taken together, our findings define a novel pathway in the regulation of actin-based cellular functions. Regulating GMFG function may provide a novel approach to modulate the pathophysiology of cardiovascular diseases.
Key Words: actin cytoskeleton reorganization actin-based cellular function microvascular endothelial cells inflammatory cells
Introduction
Endothelial cells, composing a single layer inside the blood vessels, play critical roles in a variety of physiological and pathological phenomena such as angiogenesis and atherosclerosis. Because genes specifically or preferentially expressed in endothelial cells are likely to play a crucial role for their unique physiology, we performed a series of microarray studies aimed at isolating unique endothelial cell genes. We successfully identified several genes specifically or preferentially expressed in endothelial cells.1 Among these genes, glia maturation factor- (GMFG) demonstrated a very unique expression pattern: expressed only in microvascular endothelial cells but not by endothelial cells from other vascular beds or nonendothelial cells. These results urged us to further analyze GMFG function in endothelial cells.
GMFG was initially identified as a molecule that was highly similar to glia maturation factor-.2 However, GMFG is expressed neither in brain, neuronal cells, nor glia cells, and its function is entirely unknown. We found the amino acid sequence of GMFG is similar to cofilin, a key regulator of actin cytoskeleton reorganization and thus analyzed GMFG function in actin cytoskeleton reorganization. Actin cytoskeleton reorganization is clearly an essential factor for a large number of cellular processes such as cytokinesis, endocytosis, and chemotaxis.3,4 In response to extracellular stimuli, motile protrusions of plasma membrane, called lamellipodia or filopodia are created by the continuous polarized growth and turnover of actin filaments, leading to actin-based cellular functions.3,4 Deficient actin cytoskeleton reorganization causes impaired endothelial cell migration, reduces macrophage phagocytosis, and results in defective lymphocyte development and activation.5–7 Actin cytoskeleton reorganization is thus pivotal for angiogenesis and immune system function.
Actin dynamics is regulated by an elaborate mechanism that involves many actin-associated molecules. In resting cells, actin monomers are kept from polymerizing by monomer-binding protein such as thymosin-4, twinfilin, and profilin.3,8 Capping proteins such as CapZ and gelsolin are also required to maintain a pool of actin monomers by capping free barbed ends of actin filaments, resulting in blocking the addition of actin monomers to the barbed ends. Extracellular stimuli activate Wiscott–Aldrich syndrome protein (WASP) and WASP family verprolin homologous (WAVE) largely through small GTPases, Rac, and Cdc42. Activated WASP/WAVE further activates the actin-related protein 2/3 (Arp2/3) complex to nucleate actin polymerization by serving as a template for the formation of new actin filaments.3 During the active actin polymerization at the barbed ends, actin-depolymerization takes place as a result of the function of cofilin at the pointed ends, which replenishes the actin monomer pool.4,9,10
Recently, it has been reported that WAVE2 is expressed predominantly in endothelial cells during embryogenesis and that its deficiency causes embryonic lethality because of hemorrhage.5 WAVE2–/– mice demonstrated decreased sprouting and branching of endothelial cells from existing vessels during angiogenesis. Furthermore, the formation of lamellipodia and capillaries was severely impaired in WAVE2–/– endothelial cells. These results indicate that proper regulation of actin cytoskeleton reorganization is critical for endothelial cell migration and angiogenesis.
Here, we report the characterization of GMFG, a novel factor in actin cytoskeleton reorganization, preferentially expressed in microvascular endothelial cells and inflammatory cells. GMFG remarkably enhanced stimulus-responsive lamellipodia formation, and its function was likely regulated via phosphorylation of N-terminal serine under the control of small GTPases. Overexpression of GMFG enhanced actin-based cellular functions such as cell motility and tube-formation in endothelial cells. Regulating GMFG function may provide a unique opportunity to control actin cytoskeleton reorganization in microvascular endothelial and inflammatory cells and, therefore, to modulate the pathophysiology of ischemic and inflammatory diseases.
Materials and Methods
DNA Constructs
Full-length human GMFG and cofilin were obtained by RT-PCR using cDNA prepared from human microvascular endothelial cells. A BamHI site was created before the stop codon by adding 5'-GGATCC-3' at the beginning of reverse primer. Mutagenesis was performed by creating point mutations in the forward primers that cause a missense mutation of serine into alanine or glutamic acid. These cDNAs were subcloned into pCR-blunt II-TOPO vector (Invitrogen) and the nucleotide sequence was validated. Fragments cut out with EcoRI and BamHI were subcloned into p3XFLAG-CMV-14 expression vector (Sigma) to obtain the expression constructs.
Reverse-Transcription PCR
Total RNA was isolated from cells cultured in growth media, followed by cDNA synthesis. Primers used for GMFG amplification were of 100% match for both human and bovine GMFG, 5'- GACTCCCTGGTGGTGTG -3' and 5'-TACAACGAAAGAAAGACAACTT-3'. Twenty-eight PCR cycles with annealing temperature at 57°C were performed for amplification of both GMFG and GAPDH.
In Situ Hybridization
Whole-mount in situ hybridization was performed using digoxigenin-labeled cRNA as previously described.11 After all procedures, embryos were embedded in OCT and snap-frozen by immersion in ethanol cooled on dry ice. Sections were counter-stained with eosin.
Transfection and Immunofluorescence Study
Expression constructs were transfected into bovine aortic endothelial cells (BAECs) or HeLa cells using Lipofectamine 2000 (Invitrogen) according to the recommendations of the manufacturer. Cells were plated into chamber slides 24 hours after transfection and further incubated for 24 hours followed by staining with Alexa Fluor3–phalloidin (Molecular Probes) and FITC-anti–FLAG M2 antibody (Sigma). After mounting, cells were observed under a fluorescent microscope (Axioplan 2; Zeiss).
Cell Culture
BAECs in passage 4 to 7 were used for all experiments. Human microvascular endothelial cells were obtained from Clonetics and cultured in EGM-2MV media. To generate stable transfectants, HeLa cells underwent the selection in the growth media containing 300 μg/mL G418, and pooled clones were used for experiments. Expression of transgenes was confirmed by RT-PCR using vector-specific primers, 5'-AAGCTTGCGGCCGCGAATTC-3' and 5'-TTTGTAGTCAGCCCGGGATCC. The nucleotide sequence of the PCR product was validated by direct sequencing. Recombinant protein expression was confirmed by immunoblotting using anti–FLAG M2 antibody (Stratagene). For epidermal growth factor (EGF) stimulation experiment, HeLa cells were cultured in serum-free media for 24 hours, followed by stimulation with 100 ng/mL or 1000 ng/mL EGF (Sigma) for 5 minutes. Cells were fixed with 4% paraformaldehyde and observed under a phase-contrast microscope (TE 2000 U; Nikon).
Metabolic Labeling
Cells were labeled by incubating with 1 mCi/mL of P32-phosphate (Perkin Elmer) in phosphate-free DMEM containing 10% dialyzed FBS (Invitrogen) for 4 hours. Immunoprecipitated recombinant proteins were run on SDS-PAGE and then transferred onto polyvinylidene difluoride (PVDF) membrane. Phosphorylated proteins were visualized by autoradiography.
Actin Co-Sedimentation Assay
BAECs transiently transfected with cofilin or GMFG were collected in ice-cold PBS and resuspended in binding buffer (10 mmol/L imidazole, ph 7.2, 75 mmol/L KCl, 5 mmol/L MgCl2, 1 mmol/L EGTA, 0.5 mmol/L dithiothreitol [DTT], 0.2 mmol/L ATP, 1 mmol/L NaF, 0.5 mmol/L Na3VO4, and protease inhibitor cocktail). Nonmuscle actin was obtained from Cytoskeleton Inc and polymerized according to the instructions of the manufacturer. Cell lysates were prepared with three bursts of sonication followed by ultracentrifugation at 150 000g, and then the supernatants were incubated in the presence or absence of 5.5 μmol/L F-actin at room temperature for 30 minutes. The reaction mixtures underwent ultracentrifuge at 150 000g for 30 minutes to precipitate F-actin. Supernatants were carefully collected and pellets were resuspended in exactly the same volume of binding buffer as the supernatants. Proteins of interest were detected by immunoblotting, and the amount of sedimentation was assessed by densitometry analysis.
Modified Boyden Chamber Assay and Two-Dimensional Tube-Formation Assay
Modified Boyden chamber assay and tube-formation assay were performed as previously described.12
Arp2/3 Complex–Binding Assay
Arp 2/3 complex in cell lysates was immunoprecipitated with anti-ARPC2 antibody (Upstate). Coprecipitation of GMFG was detected by immunoblotting using anti–FLAG M2 antibody.
Mouse Ischemia/Reperfusion Model
The mice underwent pre-anesthesia with a single intraperitoneal injection of sodium pentobarbital (25 mg/kg). Following endotracheal intubation, a thoracotomy incision through the fourth intercostal space was made. The left anterior descending (LAD) coronary artery was occluded for 30 minutes then reperfused for 24 hours. At the end of the experiments, the chest was opened and the left ventricles were immediately excised followed by snap freezing for RNA extraction.
Statistics
Differences between groups were analyzed using student’s t test; P<0.05 was considered significant. Data are presented as mean±SE as indicated.
Results
Glia Maturation Factor- Is Preferentially Expressed in Microvascular Endothelial and Inflammatory Cells
During a series of microarray studies aimed at isolating unique endothelial cell genes, we identified previously uncharacterized GMFG as being expressed only in microvascular endothelial cells derived from human lung but not by endothelial cells from other vascular beds or nonendothelial cells (Figure 1A). Amino acid sequence of GMFG is highly conserved among species, eg, 95% homology between human and mouse and 98% homology between human and bovine. GMFG was also expressed in dermal microvascular cells as much as in lung microvascular endothelial cells, and minimal expression was observed in hepatoma cells, vascular smooth muscle cells, and endothelial cells from bovine aorta, assessed by RT-PCR (Figure 1B). The primers used were of 100% match for both human and bovine GMFG. In adult mouse tissues, GMFG was expressed abundantly in thymus and spleen as well as in lung (Figure 1C). We found that GMFG was expressed in inflammatory cells such as lymphoblasts, T lymphocytes, and macrophages, a finding that is consistent with its high expression in thymus and spleen (Figure 1D). Mouse endothelial cell lines derived from normal yolk sac (Pro5), endothelioma (EOMA), and microvessels of an SV40T transgenic mouse (py4.1) also expressed GMFG (Figure 1E).
During mouse embryogenesis, GMFG was expressed as early as embryonic day 9.5 (e9.5) (Figure 2A), predominantly in blood islands of the yolk sac, in endothelial and hematopoietic cells, and possibly in the angioblast precursors to these lineages (Figure 2B through 2G). Negative control study using sense RNA probe did not show such a significant hybridization signals (Figure I in the online data supplement, available at http://circres.ahajournals.org). These results suggest that GMFG has a function in embryonic vasculogenesis as well as in hematopoiesis and that its expression might be well preserved during differentiation, which could account for the unique GMFG expression in microvascular endothelial cells and hematopoietic lineage-derived cells in adults.
GMFG Has a Similar Structure to Cofilin and Is Associated With F-Actin
The amino acid sequence of GMFG demonstrated significant similarity to cofilin, a key regulator of actin cytoskeleton reorganization (Figure 3A). Cofilin plays critical roles in actin dynamics through its actin-severing and actin-depolymerizing activities. Therefore, we determined to analyze GMFG function in actin cytoskeleton reorganization. We first examined the subcellular localization of GMFG and cofilin in bovine aortic endothelial cells (BAECs). Recombinant cofilin and GMFG localized preferentially in the F-actin–rich structures in membrane ruffles as well as in nucleus (Figure 3B). Furthermore, native GMFG in human microvascular endothelial cells also demonstrated colocalization with F-actin in membrane ruffles (Figure 3C). These results suggest that GMFG may play a role in actin dynamics at the leading edge of cells. We then examined if GMFG is associated with F-actin using actin co-sedimentation assay. Cell lysates from GMFG or cofilin-expressing BAECs were incubated with F-actin followed by ultracentrifugation to precipitate F-actin. Significant co-sedimentation of GMFG in the pellet was detected by immunoblotting as well as of cofilin, indicating that GMFG is associated with F-actin (Figure 3D).
GMFG Is a Phosphoprotein, and Its Phosphorylation Is Enhanced by Coexpression of Dominant Active Rac and Cdc42
Because cofilin function is tightly regulated by phosphorylation of a single N-terminal serine residue, and GMFG also has N-terminal serines at the second (Ser2) and fourth (Ser4) position (Figure 3A), we explored their possible phosphorylation in GMFG. Expression constructs of wild-type GMFG (GMFG-WT) and GMFG mutants, in which Ser2 and/or Ser4 were replaced with alanine, were transfected into BAECs. Metabolic labeling of BAECs with P32-phosphate revealed that GMFG was phosphorylated in cells, and replacing Ser2 and/or Ser4 with alanine completely abolished phosphorylation (Figure 4A). These results indicate that Ser2 and/or Ser4 are the phosphorylation sites and that both Ser2 and Ser4 are essential for GMFG phosphorylation. Of note, when incubated in serum-free media, phosphorylation of GMFG significantly diminished, suggesting that phosphorylation might be enhanced by soluble factors such as growth factors (Figure 4B).
Because small GTPases play central roles in actin cytoskeleton reorganization, we investigated the effect of small GTPases on GMFG phosphorylation. Coexpression of dominant active Rac1 and Cdc42 significantly enhanced GMFG phosphorylation, whereas dominant active RhoA did not (Figure 4C). These results suggest that GMFG function might be modified by Rac and Cdc42.
GMFG Association With F-Actin Is Regulated by Phosphorylation of N-Terminal Serine
To explore the effect of phosphorylation on GMFG function, we prepared a series of GMFG mutant expression constructs in which Ser2 and/or Ser4 were replaced with alanine (SA) (nonphosphorylated mutant) or glutamic acid (SE) (pseudophosphorylated mutant). Cell lysates of BAECs expressing GMFG-WT or GMFG mutants were incubated with F-actin, and their co-sedimentation with F-actin was detected by immunoblotting. When incubated in the absence of F-actin, only minimal sedimentation was observed except in the cases of GMFG-S2/4A and GMFG-S2A where sedimentation of some insoluble fractions in the absence of F-actin were consistently observed in 3 independent experiments (Figure 4). We subtracted the background sedimentation (in the absence of F-actin) from the co-sedimentation with F-actin and regarded it as the sedimentation associated with F-actin. Interestingly, GMFG-S2/4E and especially GMFG-S2E demonstrated significantly higher association with F-actin than other GMFG mutants or GMFG-WT, suggesting that phosphorylation of Ser2 but not Ser4 enhances GMFG association with F-actin. To exclude the possibility that replacement of N-terminal serine with alanine caused inappropriate protein folding or localization, we examined the subcellular localization of recombinant GMFG-S2/4A, GMFG-S2A, and GMFG-S4A. They demonstrated similar subcellular localization to GMFG-WT, suggesting that replacement of N-terminal serine with alanine did not cause inappropriate protein folding or localization (Figure 5B).We also confirmed that recombinant GMFG mutant of which serine(s) was replaced with glutamic acid(s) demonstrated proper subcellular localization (data not shown). It is intriguing that although the affinity of GMFG and cofilin for F-actin is regulated by phosphorylation of the N-terminal serine, phosphorylation enhances GMFG binding, whereas it reduces cofilin binding to F-actin.
GMFG Is a Novel Factor in Actin Cytoskeleton Reorganization
Because cofilin directly affects actin filament turnover, we investigated the effects of GMFG on actin cytoskeleton reorganization. Somewhat surprisingly, HeLa cells transiently transfected with GMFG-WT or the pseudophosphorylated form of GMFG (GMFG-S2E) demonstrated no significant change in F-actin structure, whereas expression of wild-type cofilin (cofilin-WT) or the active form of cofilin (cofilin-S3A) significantly reduced F-actin structure (Figure 6A). These results suggest that GMFG does not have a significant actin-depolymerizing activity and may have a different function in actin dynamics from cofilin.
We then generated HeLa cells stably expressing cofilin-WT, cofilin-S3A, GMFG-WT, GMFG-S2E, and GMFG-S4E. Because HeLa cells are the most widely used and the most-characterized cells for the research of actin cytoskeleton reorganization, we chose HeLa cells to analyze GMFG function in actin cytoskeleton reorganization. Stable expression of transgenes and recombinant proteins were confirmed by RT-PCR and immunoblotting respectively (data not shown). Large populations (30%) of cells stably expressing cofilin-S3A (HeLa/cofilin-S3A) demonstrated enhanced membrane ruffles formation under the basal culture condition, whereas none of the other stable transfectants demonstrated such a phenotype (Figure 6B and 6C).
On the other hand, when cells were stimulated with a low concentration (100 ng/mL) of EGF following the incubation in serum-free media for 24 hours, HeLa/GMFG-S2E demonstrated significant stimulus-responsive lamellipodia and subsequent ruffles formation (Figure 6D and 6E). HeLa/GMFG-WT also demonstrated less but significant stimulus-responsive lamellipodia formation with 100 ng/mL EGF treatment (Figure 6D and 6E). None of the other transfectants, including HeLa/MOCK, demonstrated such an effect. Stimulation with 1000 ng/mL EGF induced stimulus-responsive lamellipodia and subsequent ruffle formation in all stable transfectants (data not shown). These results suggest a significant role of GMFG in stimulus-responsive actin polymerization. Despite its structural similarity, GMFG appears to play a different role from cofilin in actin cytoskeleton reorganization. It has been reported that twinfilin, another actin-depolymerizing factor domain family protein, does not have actin-depolymerizing activity and functions in actin dynamics in clearly distinct manner from cofilin.8 Moreover, Abp1p, an actin binding protein originally isolated from yeast, also has an actin-depolymerizing factor homology domain, but it binds to Arp2/3 complex and activates the actin nucleation, not actin depolymerization.13
To better understand the cellular mechanisms by which GMFG enhances stimulus-responsive lamellipodia formation, we investigated the possible interaction of GMFG with the Arp2/3 complex, a key regulator of stimulus-responsive actin polymerization. Cell lysates expressing GMFG were incubated with anti–Arp2/3 complex subunit 2 (ARPC2) antibody, and coprecipitation of GMFG with ARPC2 was detected by immunoblotting. To exclude the potential nonspecific interactions from coprecipitation of actin filaments, 40 μmol/L latrunculin A was included in the immunoprecipitation reactions. Immunoblotting confirmed no detectable actin in the pellets (data not shown). GMFG was coprecipitated with ARPC2, suggesting GMFG interaction with Arp2/3 complex (Figure 6F). Of note, more GMFG-S2E was pulled down than GMFG-WT or GMFG-S4E, indicating that phosphorylation of Ser2 enhances GMFG interaction with the Arp2/3 complex (Figure 6F). These results suggest that in response to extracellular stimuli, GMFG might be phosphorylated at Ser2 through Rac and Cdc42, promoting its interaction with Arp2/3 complex as well as with F-actin, which then accelerates stimulus-responsive actin polymerization at the leading edge of cells. However, further analysis is certainly required to conclude its functional relevance with Rac and Cdc42. Given that even pseudophosphorylated form of GMFG required exogenous stimulus to modulate actin cytoskeleton reorganization, GMFG binding to Arp2/3 complex is not sufficient to initiate new actin polymerization.
GMFG Positively Regulates Actin-Based Cellular Functions
We then investigated GMFG functions in actin-based cellular functions in endothelial cells. As shown in Figure 1B, BAECs express only a minimal amount of GMFG, much less than in microvascular endothelial cells. Therefore, we used BAECs to examine the effect of GMFG overexpression on actin-based cellular functions. The majority of cells expressed the transgene 24 hours after transfection, as assessed by GFP-construct transfection (supplemental Figure II). BAECs overexpressing GMFG demonstrated significantly higher motility than mock-transfected BAECs (Figure 7A). Moreover, BAECs overexpressing GMFG demonstrated significantly enhanced tube formation on Matrigel as compared with the mock-transfected cells (Figure 7B). These results suggest that GMFG positively regulates actin-based cellular functions in endothelial cells, presumably through accelerating stimulus-responsive actin polymerization.
Finally, we examined GMFG expression in a cardiac ischemia/reperfusion model because both angiogenesis and inflammation play important roles in the pathophysiology of ischemic cardiovascular diseases. GMFG expression was significantly increased in the ischemia/reperfusion tissues (Figure 7C and 7D). Although further analysis, such as immunohistochemistry to clarify the cells express GMFG, is certainly needed to address GMFG function in the ischemia/reperfusion heart, this result suggests possible GMFG involvement in the pathophysiology of cardiovascular diseases.
Our data define a novel pathway in the regulation of actin cytoskeleton reorganization and actin-based cellular functions (Figure 7E). Given that inflammatory cells and endothelial cells have a close relationship and interact with each other in ischemic cardiovascular diseases, GMFG may play a role in the pathophysiology of such diseases.14,15
Discussion
GMFG is preferentially expressed in human microvascular endothelial cells but not in endothelial cells from other vascular beds. Given that microvascular endothelial cells are the most important cellular players in angiogenesis in adults and expression of GMFG enhanced migration and tube formation in BAECs, GMFG may play a role in angiogenesis in adults. During embryogenesis, GMFG is predominantly expressed in blood islands of the yolk sac, in which hematopoietic and endothelial lineages develop simultaneously in close proximity, suggesting GMFG functions in embryonic vasculogenesis as well as hematopoiesis. This may also well correlate with the unique GMFG expression in microvascular endothelial cells and inflammatory cells seen in adults. In support of this contention, a search of the database of microarray analysis at the National Center for Biotechnology Information revealed that GMFG is expressed in a variety of hematopoietic cells such as T cell, B cell, NK cell, monocyte, and dendritic cell, as well as hematopoietic progenitor cells including CD34+ and CD105+ bone marrow cells, which may give rise to endothelial progenitor cells.16,17
Our findings demonstrated that GMFG plays a significant role in actin cytoskeleton reorganization distinct from cofilin, despite structural similarity. Stable expression of the pseudophosphorylated form of GMFG remarkably enhanced stimulus-responsive lamellipodia and subsequent membrane ruffles formation, presumably through interaction with Arp2/3 complex. In response to extracellular stimuli, active Rac and Cdc42 might activate GMFG via phosphorylation of Ser2 as well as WASP/WAVE. Binding of phosphorylated GMFG to the Arp2/3 complex might accelerate stimulus-responsive actin polymerization cooperating with active WASP/WAVE, resulting in enhanced actin-based cellular functions. However, further analysis is required to verify this hypothesis.
Of note, the pseudophosphorylated form of GMFG did not affect actin cytoskeleton reorganization in HeLa cells under basal culture conditions, although it remarkably enhanced stimulus-responsive lamellipodia formation. These results suggest that binding of GMFG to Arp2/3 complex is not sufficient to initiate new actin polymerization, but once Arp2/3 complex is activated by other activator such as WASP/WAVE, GMFG may enhance Arp2/3 complex-induced actin polymerization. Therefore, GMFG may enhance actin-based cellular functions only in the presence, and under the control, of endogenous stimulators such as growth factors and cytokines. In this regard, GMFG may be a good target for gene therapy because unexpected adverse effect caused by dysregulated transgene function is a considerable concern for the clinical application of gene therapy.
Our results indicate that GMFG is a novel factor in actin cytoskeleton reorganization and that it modulates actin-based cellular function in microvascular endothelial cells, inflammatory cells, and possibly hematopoietic progenitor cells. Because these cells are essential players in angiogenesis, vasculogenesis, and immune system function, and each of them plays specific roles in ischemic cardiovascular diseases, we believe that further analysis of GMFG will provide new insights into the molecular mechanisms of these diseases and that GMFG may be an attractive new target for pharmacotherapeutic agents to treat such diseases.
Acknowledgments
We thank Dr Kiyofumi Asai for providing the anti-sera for GMFG. We thank Dr Yoshimi Takai for providing the expression constructs of dominant active and dominant negative Rho, Rac, and Cdc42. We thank Dr Alex Dunn for helpful suggestions for actin-binding and Arp2/3 complex–binding assays.
Source of Funding
This work was supported by the Donald W. Reynolds Foundation.
Disclosures
None.
Footnotes
Original received March 26, 2006; revision received July 12, 2006; accepted July 13, 2006.
References
Ho M, Yang E, Matcuk G, Deng D, Sampas N, Tsalenko A, Tabibiazar R, Zhang Y, Chen M, Talbi S, Ho YD, Wang J, Tsao PS, Ben-Dor A, Yakhini Z, Bruhn L, Quertermous T. Identification of endothelial cell genes by combined database mining and microarray analysis. Physiol Genomics. 2003; 13: 249–262.
Asai K, Fujita K, Yamamoto M, Hotta T, Moriyama M, Kokubo M, Moriyama A, Kato T. Isolation od novel cDNA (hGMF-gamma) homologous to glia maturation factor-beta gene. Biochim Biophys Acta. 1998; 1396: 242–244. [Order article via Infotrieve]
Pollard TD, Blanchoin L, Mullins RD. Molecular mechanisms controlling actin filament dynamics in nonmuscle cells. Annu Rev Biophys Biomol Struct. 2000; 29: 545–576. [Order article via Infotrieve]
Carlier MF, Ressad F, Pantaloni D. Control of actin dynamics in cell motility. Role of ADF/cofilin. J Biol Chem. 1999; 274: 33827–33830.
Yamazaki D, Suetsugu S, Miki H, Kataoka Y, Nishikawa S, Fujiwara T, Yoshida N, Takenawa T. WAVE2 is required for directed cell migration and cardiovascular development. Nature. 2003; 424: 452–456. [Order article via Infotrieve]
Bamburg JR, Wiggan OP. ADF/cofilin and actin dynamics in disease. Trends Cell Biol. 2002; 12: 598–605. [Order article via Infotrieve]
Zhang J, Shehabeldin A, da Cruz LA, Butler J, Somani AK, McGavin M, Kozieradzki I, dos Santos AO, Nagy A, Grinstein S, Penninger JM. Antigen receptor-induced activation and cytoskeletal rearrangement are impaired in Wiskott-Aldrich syndrome protein-deficient lymphocytes. J Exp Med. 1999; 190: 1329–1342.
Goode BL, Drubin DG, Lappalainen P. Regulation of the cortical actin cytoskeleton in budding yeast by twinfilin, a ubiquitous actin monomer-sequestering protein. J Cell Biol. 1998; 142: 723–733.
Bamburg JR. Proteins of the ADF/cofilin family: essential regulators of actin dynamics. Annu Rev Cell Dev Biol. 1999; 15: 185–230. [Order article via Infotrieve]
Ghosh M, Song X, Mouneimne G, Sidani M, Lawrence DS, Condeelis JS. Cofilin promotes actin polymerization and defines the direction of cell motility. Science. 2004; 304: 743–746.
Hirata KI, Ishida T, Penta K, Rezaee M, Yang E, Wohlgemuth J, Quertermous T. Cloning of an immunoglobulin family adhesion molecule selectively expressed by endothelial cells. J Biol Chem. 2001; 276: 16223–16231.
Ikeda K, Quertermous T. Molecular isolation and characterization of a soluble isoform of activated leukocyte cell adhesion molecule that modulates endothelial cell function. J Biol Chem. 2004; 279: 55315–55323.
Olazabal IM, Machesky LM. Abp1p and cortactin, new "hand-holds" for actin. J Cell Biol. 2001; 154: 679–682.
Nian M, Lee P, Khaper N, Liu P. Inflammatory cytokines and postmyocardial infarction remodeling. Cir Res. 2004; 94: 1543–1553.
Naldini A, Carraro F. Role of inflammatory mediators in angiogenesis. Curr Drug Targets Inflamm Allergy. 2005; 4: 3–8. [Order article via Infotrieve]
Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T, Witzenbichler B, Schatteman G, Isner JM. Isolation of putative progenitor endothelial cells for angiogenesis. Science. 1997; 275: 964–967.
Salven P, Mustjoki S, Alital R, Alitalo K, Rafii S. VEGFR-3 and CD133 identify a population of CD34+ lymphatic/vascular endothelial precursor cells. Blood. 2003; 101: 168–172.
the Donald W. Reynolds Cardiovascular Clinical Research Center (K.I., R.K.K., T.Q.), Division of Cardiovascular Medicine, Stanford University School of Medicine, Calif
Department of Clinical Pharmacology (M.K.), Kyoto Pharmaceutical University, Japan
Department of Cardiology and Vascular Regenerative Medicine (K.I., S.I., H.M.), Kyoto Prefectural University of Medicine, Japan.
Abstract
Actin cytoskeleton reorganization is a fundamental process for actin-based cellular functions such as cytokinesis, phagocytosis, and chemotaxis. Regulating actin cytoskeleton reorganization is therefore an attractive approach to control endothelial and inflammatory cells function and to treat cardiovascular diseases. Here, we identified glia maturation factor- (GMFG) as a novel factor in actin cytoskeleton reorganization and is expressed preferentially in microvascular endothelial and inflammatory cells. During mouse embryogenesis, GMFG was expressed predominantly in blood islands of the yolk sac, where endothelial and hematopoietic cells develop simultaneously. In endothelial cells, GMFG was colocalized with F-actin in membrane ruffles and was associated with F-actin assessed by actin co-sedimentation assay. Interestingly, GMFG was phosphorylated at N-terminal serine, and its phosphorylation was enhanced by coexpression of dominant active Rac1 and Cdc42. Furthermore, a pseudophosphorylated form of GMFG (GMFG-S2E) demonstrated higher association with F-actin. Stable expression of GMFG-S2E remarkably enhanced stimulus-responsive lamellipodia and subsequent membrane ruffle formation in HeLa cells presumably through its interaction with Arp2/3 complex. Expression of GMFG enhanced actin-based cellular functions such as migration and tube-formation in endothelial cells. Moreover, we found that GMFG expression was significantly increased in a cardiac ischemia/reperfusion model where inflammation and angiogenesis take place actively. Taken together, our findings define a novel pathway in the regulation of actin-based cellular functions. Regulating GMFG function may provide a novel approach to modulate the pathophysiology of cardiovascular diseases.
Key Words: actin cytoskeleton reorganization actin-based cellular function microvascular endothelial cells inflammatory cells
Introduction
Endothelial cells, composing a single layer inside the blood vessels, play critical roles in a variety of physiological and pathological phenomena such as angiogenesis and atherosclerosis. Because genes specifically or preferentially expressed in endothelial cells are likely to play a crucial role for their unique physiology, we performed a series of microarray studies aimed at isolating unique endothelial cell genes. We successfully identified several genes specifically or preferentially expressed in endothelial cells.1 Among these genes, glia maturation factor- (GMFG) demonstrated a very unique expression pattern: expressed only in microvascular endothelial cells but not by endothelial cells from other vascular beds or nonendothelial cells. These results urged us to further analyze GMFG function in endothelial cells.
GMFG was initially identified as a molecule that was highly similar to glia maturation factor-.2 However, GMFG is expressed neither in brain, neuronal cells, nor glia cells, and its function is entirely unknown. We found the amino acid sequence of GMFG is similar to cofilin, a key regulator of actin cytoskeleton reorganization and thus analyzed GMFG function in actin cytoskeleton reorganization. Actin cytoskeleton reorganization is clearly an essential factor for a large number of cellular processes such as cytokinesis, endocytosis, and chemotaxis.3,4 In response to extracellular stimuli, motile protrusions of plasma membrane, called lamellipodia or filopodia are created by the continuous polarized growth and turnover of actin filaments, leading to actin-based cellular functions.3,4 Deficient actin cytoskeleton reorganization causes impaired endothelial cell migration, reduces macrophage phagocytosis, and results in defective lymphocyte development and activation.5–7 Actin cytoskeleton reorganization is thus pivotal for angiogenesis and immune system function.
Actin dynamics is regulated by an elaborate mechanism that involves many actin-associated molecules. In resting cells, actin monomers are kept from polymerizing by monomer-binding protein such as thymosin-4, twinfilin, and profilin.3,8 Capping proteins such as CapZ and gelsolin are also required to maintain a pool of actin monomers by capping free barbed ends of actin filaments, resulting in blocking the addition of actin monomers to the barbed ends. Extracellular stimuli activate Wiscott–Aldrich syndrome protein (WASP) and WASP family verprolin homologous (WAVE) largely through small GTPases, Rac, and Cdc42. Activated WASP/WAVE further activates the actin-related protein 2/3 (Arp2/3) complex to nucleate actin polymerization by serving as a template for the formation of new actin filaments.3 During the active actin polymerization at the barbed ends, actin-depolymerization takes place as a result of the function of cofilin at the pointed ends, which replenishes the actin monomer pool.4,9,10
Recently, it has been reported that WAVE2 is expressed predominantly in endothelial cells during embryogenesis and that its deficiency causes embryonic lethality because of hemorrhage.5 WAVE2–/– mice demonstrated decreased sprouting and branching of endothelial cells from existing vessels during angiogenesis. Furthermore, the formation of lamellipodia and capillaries was severely impaired in WAVE2–/– endothelial cells. These results indicate that proper regulation of actin cytoskeleton reorganization is critical for endothelial cell migration and angiogenesis.
Here, we report the characterization of GMFG, a novel factor in actin cytoskeleton reorganization, preferentially expressed in microvascular endothelial cells and inflammatory cells. GMFG remarkably enhanced stimulus-responsive lamellipodia formation, and its function was likely regulated via phosphorylation of N-terminal serine under the control of small GTPases. Overexpression of GMFG enhanced actin-based cellular functions such as cell motility and tube-formation in endothelial cells. Regulating GMFG function may provide a unique opportunity to control actin cytoskeleton reorganization in microvascular endothelial and inflammatory cells and, therefore, to modulate the pathophysiology of ischemic and inflammatory diseases.
Materials and Methods
DNA Constructs
Full-length human GMFG and cofilin were obtained by RT-PCR using cDNA prepared from human microvascular endothelial cells. A BamHI site was created before the stop codon by adding 5'-GGATCC-3' at the beginning of reverse primer. Mutagenesis was performed by creating point mutations in the forward primers that cause a missense mutation of serine into alanine or glutamic acid. These cDNAs were subcloned into pCR-blunt II-TOPO vector (Invitrogen) and the nucleotide sequence was validated. Fragments cut out with EcoRI and BamHI were subcloned into p3XFLAG-CMV-14 expression vector (Sigma) to obtain the expression constructs.
Reverse-Transcription PCR
Total RNA was isolated from cells cultured in growth media, followed by cDNA synthesis. Primers used for GMFG amplification were of 100% match for both human and bovine GMFG, 5'- GACTCCCTGGTGGTGTG -3' and 5'-TACAACGAAAGAAAGACAACTT-3'. Twenty-eight PCR cycles with annealing temperature at 57°C were performed for amplification of both GMFG and GAPDH.
In Situ Hybridization
Whole-mount in situ hybridization was performed using digoxigenin-labeled cRNA as previously described.11 After all procedures, embryos were embedded in OCT and snap-frozen by immersion in ethanol cooled on dry ice. Sections were counter-stained with eosin.
Transfection and Immunofluorescence Study
Expression constructs were transfected into bovine aortic endothelial cells (BAECs) or HeLa cells using Lipofectamine 2000 (Invitrogen) according to the recommendations of the manufacturer. Cells were plated into chamber slides 24 hours after transfection and further incubated for 24 hours followed by staining with Alexa Fluor3–phalloidin (Molecular Probes) and FITC-anti–FLAG M2 antibody (Sigma). After mounting, cells were observed under a fluorescent microscope (Axioplan 2; Zeiss).
Cell Culture
BAECs in passage 4 to 7 were used for all experiments. Human microvascular endothelial cells were obtained from Clonetics and cultured in EGM-2MV media. To generate stable transfectants, HeLa cells underwent the selection in the growth media containing 300 μg/mL G418, and pooled clones were used for experiments. Expression of transgenes was confirmed by RT-PCR using vector-specific primers, 5'-AAGCTTGCGGCCGCGAATTC-3' and 5'-TTTGTAGTCAGCCCGGGATCC. The nucleotide sequence of the PCR product was validated by direct sequencing. Recombinant protein expression was confirmed by immunoblotting using anti–FLAG M2 antibody (Stratagene). For epidermal growth factor (EGF) stimulation experiment, HeLa cells were cultured in serum-free media for 24 hours, followed by stimulation with 100 ng/mL or 1000 ng/mL EGF (Sigma) for 5 minutes. Cells were fixed with 4% paraformaldehyde and observed under a phase-contrast microscope (TE 2000 U; Nikon).
Metabolic Labeling
Cells were labeled by incubating with 1 mCi/mL of P32-phosphate (Perkin Elmer) in phosphate-free DMEM containing 10% dialyzed FBS (Invitrogen) for 4 hours. Immunoprecipitated recombinant proteins were run on SDS-PAGE and then transferred onto polyvinylidene difluoride (PVDF) membrane. Phosphorylated proteins were visualized by autoradiography.
Actin Co-Sedimentation Assay
BAECs transiently transfected with cofilin or GMFG were collected in ice-cold PBS and resuspended in binding buffer (10 mmol/L imidazole, ph 7.2, 75 mmol/L KCl, 5 mmol/L MgCl2, 1 mmol/L EGTA, 0.5 mmol/L dithiothreitol [DTT], 0.2 mmol/L ATP, 1 mmol/L NaF, 0.5 mmol/L Na3VO4, and protease inhibitor cocktail). Nonmuscle actin was obtained from Cytoskeleton Inc and polymerized according to the instructions of the manufacturer. Cell lysates were prepared with three bursts of sonication followed by ultracentrifugation at 150 000g, and then the supernatants were incubated in the presence or absence of 5.5 μmol/L F-actin at room temperature for 30 minutes. The reaction mixtures underwent ultracentrifuge at 150 000g for 30 minutes to precipitate F-actin. Supernatants were carefully collected and pellets were resuspended in exactly the same volume of binding buffer as the supernatants. Proteins of interest were detected by immunoblotting, and the amount of sedimentation was assessed by densitometry analysis.
Modified Boyden Chamber Assay and Two-Dimensional Tube-Formation Assay
Modified Boyden chamber assay and tube-formation assay were performed as previously described.12
Arp2/3 Complex–Binding Assay
Arp 2/3 complex in cell lysates was immunoprecipitated with anti-ARPC2 antibody (Upstate). Coprecipitation of GMFG was detected by immunoblotting using anti–FLAG M2 antibody.
Mouse Ischemia/Reperfusion Model
The mice underwent pre-anesthesia with a single intraperitoneal injection of sodium pentobarbital (25 mg/kg). Following endotracheal intubation, a thoracotomy incision through the fourth intercostal space was made. The left anterior descending (LAD) coronary artery was occluded for 30 minutes then reperfused for 24 hours. At the end of the experiments, the chest was opened and the left ventricles were immediately excised followed by snap freezing for RNA extraction.
Statistics
Differences between groups were analyzed using student’s t test; P<0.05 was considered significant. Data are presented as mean±SE as indicated.
Results
Glia Maturation Factor- Is Preferentially Expressed in Microvascular Endothelial and Inflammatory Cells
During a series of microarray studies aimed at isolating unique endothelial cell genes, we identified previously uncharacterized GMFG as being expressed only in microvascular endothelial cells derived from human lung but not by endothelial cells from other vascular beds or nonendothelial cells (Figure 1A). Amino acid sequence of GMFG is highly conserved among species, eg, 95% homology between human and mouse and 98% homology between human and bovine. GMFG was also expressed in dermal microvascular cells as much as in lung microvascular endothelial cells, and minimal expression was observed in hepatoma cells, vascular smooth muscle cells, and endothelial cells from bovine aorta, assessed by RT-PCR (Figure 1B). The primers used were of 100% match for both human and bovine GMFG. In adult mouse tissues, GMFG was expressed abundantly in thymus and spleen as well as in lung (Figure 1C). We found that GMFG was expressed in inflammatory cells such as lymphoblasts, T lymphocytes, and macrophages, a finding that is consistent with its high expression in thymus and spleen (Figure 1D). Mouse endothelial cell lines derived from normal yolk sac (Pro5), endothelioma (EOMA), and microvessels of an SV40T transgenic mouse (py4.1) also expressed GMFG (Figure 1E).
During mouse embryogenesis, GMFG was expressed as early as embryonic day 9.5 (e9.5) (Figure 2A), predominantly in blood islands of the yolk sac, in endothelial and hematopoietic cells, and possibly in the angioblast precursors to these lineages (Figure 2B through 2G). Negative control study using sense RNA probe did not show such a significant hybridization signals (Figure I in the online data supplement, available at http://circres.ahajournals.org). These results suggest that GMFG has a function in embryonic vasculogenesis as well as in hematopoiesis and that its expression might be well preserved during differentiation, which could account for the unique GMFG expression in microvascular endothelial cells and hematopoietic lineage-derived cells in adults.
GMFG Has a Similar Structure to Cofilin and Is Associated With F-Actin
The amino acid sequence of GMFG demonstrated significant similarity to cofilin, a key regulator of actin cytoskeleton reorganization (Figure 3A). Cofilin plays critical roles in actin dynamics through its actin-severing and actin-depolymerizing activities. Therefore, we determined to analyze GMFG function in actin cytoskeleton reorganization. We first examined the subcellular localization of GMFG and cofilin in bovine aortic endothelial cells (BAECs). Recombinant cofilin and GMFG localized preferentially in the F-actin–rich structures in membrane ruffles as well as in nucleus (Figure 3B). Furthermore, native GMFG in human microvascular endothelial cells also demonstrated colocalization with F-actin in membrane ruffles (Figure 3C). These results suggest that GMFG may play a role in actin dynamics at the leading edge of cells. We then examined if GMFG is associated with F-actin using actin co-sedimentation assay. Cell lysates from GMFG or cofilin-expressing BAECs were incubated with F-actin followed by ultracentrifugation to precipitate F-actin. Significant co-sedimentation of GMFG in the pellet was detected by immunoblotting as well as of cofilin, indicating that GMFG is associated with F-actin (Figure 3D).
GMFG Is a Phosphoprotein, and Its Phosphorylation Is Enhanced by Coexpression of Dominant Active Rac and Cdc42
Because cofilin function is tightly regulated by phosphorylation of a single N-terminal serine residue, and GMFG also has N-terminal serines at the second (Ser2) and fourth (Ser4) position (Figure 3A), we explored their possible phosphorylation in GMFG. Expression constructs of wild-type GMFG (GMFG-WT) and GMFG mutants, in which Ser2 and/or Ser4 were replaced with alanine, were transfected into BAECs. Metabolic labeling of BAECs with P32-phosphate revealed that GMFG was phosphorylated in cells, and replacing Ser2 and/or Ser4 with alanine completely abolished phosphorylation (Figure 4A). These results indicate that Ser2 and/or Ser4 are the phosphorylation sites and that both Ser2 and Ser4 are essential for GMFG phosphorylation. Of note, when incubated in serum-free media, phosphorylation of GMFG significantly diminished, suggesting that phosphorylation might be enhanced by soluble factors such as growth factors (Figure 4B).
Because small GTPases play central roles in actin cytoskeleton reorganization, we investigated the effect of small GTPases on GMFG phosphorylation. Coexpression of dominant active Rac1 and Cdc42 significantly enhanced GMFG phosphorylation, whereas dominant active RhoA did not (Figure 4C). These results suggest that GMFG function might be modified by Rac and Cdc42.
GMFG Association With F-Actin Is Regulated by Phosphorylation of N-Terminal Serine
To explore the effect of phosphorylation on GMFG function, we prepared a series of GMFG mutant expression constructs in which Ser2 and/or Ser4 were replaced with alanine (SA) (nonphosphorylated mutant) or glutamic acid (SE) (pseudophosphorylated mutant). Cell lysates of BAECs expressing GMFG-WT or GMFG mutants were incubated with F-actin, and their co-sedimentation with F-actin was detected by immunoblotting. When incubated in the absence of F-actin, only minimal sedimentation was observed except in the cases of GMFG-S2/4A and GMFG-S2A where sedimentation of some insoluble fractions in the absence of F-actin were consistently observed in 3 independent experiments (Figure 4). We subtracted the background sedimentation (in the absence of F-actin) from the co-sedimentation with F-actin and regarded it as the sedimentation associated with F-actin. Interestingly, GMFG-S2/4E and especially GMFG-S2E demonstrated significantly higher association with F-actin than other GMFG mutants or GMFG-WT, suggesting that phosphorylation of Ser2 but not Ser4 enhances GMFG association with F-actin. To exclude the possibility that replacement of N-terminal serine with alanine caused inappropriate protein folding or localization, we examined the subcellular localization of recombinant GMFG-S2/4A, GMFG-S2A, and GMFG-S4A. They demonstrated similar subcellular localization to GMFG-WT, suggesting that replacement of N-terminal serine with alanine did not cause inappropriate protein folding or localization (Figure 5B).We also confirmed that recombinant GMFG mutant of which serine(s) was replaced with glutamic acid(s) demonstrated proper subcellular localization (data not shown). It is intriguing that although the affinity of GMFG and cofilin for F-actin is regulated by phosphorylation of the N-terminal serine, phosphorylation enhances GMFG binding, whereas it reduces cofilin binding to F-actin.
GMFG Is a Novel Factor in Actin Cytoskeleton Reorganization
Because cofilin directly affects actin filament turnover, we investigated the effects of GMFG on actin cytoskeleton reorganization. Somewhat surprisingly, HeLa cells transiently transfected with GMFG-WT or the pseudophosphorylated form of GMFG (GMFG-S2E) demonstrated no significant change in F-actin structure, whereas expression of wild-type cofilin (cofilin-WT) or the active form of cofilin (cofilin-S3A) significantly reduced F-actin structure (Figure 6A). These results suggest that GMFG does not have a significant actin-depolymerizing activity and may have a different function in actin dynamics from cofilin.
We then generated HeLa cells stably expressing cofilin-WT, cofilin-S3A, GMFG-WT, GMFG-S2E, and GMFG-S4E. Because HeLa cells are the most widely used and the most-characterized cells for the research of actin cytoskeleton reorganization, we chose HeLa cells to analyze GMFG function in actin cytoskeleton reorganization. Stable expression of transgenes and recombinant proteins were confirmed by RT-PCR and immunoblotting respectively (data not shown). Large populations (30%) of cells stably expressing cofilin-S3A (HeLa/cofilin-S3A) demonstrated enhanced membrane ruffles formation under the basal culture condition, whereas none of the other stable transfectants demonstrated such a phenotype (Figure 6B and 6C).
On the other hand, when cells were stimulated with a low concentration (100 ng/mL) of EGF following the incubation in serum-free media for 24 hours, HeLa/GMFG-S2E demonstrated significant stimulus-responsive lamellipodia and subsequent ruffles formation (Figure 6D and 6E). HeLa/GMFG-WT also demonstrated less but significant stimulus-responsive lamellipodia formation with 100 ng/mL EGF treatment (Figure 6D and 6E). None of the other transfectants, including HeLa/MOCK, demonstrated such an effect. Stimulation with 1000 ng/mL EGF induced stimulus-responsive lamellipodia and subsequent ruffle formation in all stable transfectants (data not shown). These results suggest a significant role of GMFG in stimulus-responsive actin polymerization. Despite its structural similarity, GMFG appears to play a different role from cofilin in actin cytoskeleton reorganization. It has been reported that twinfilin, another actin-depolymerizing factor domain family protein, does not have actin-depolymerizing activity and functions in actin dynamics in clearly distinct manner from cofilin.8 Moreover, Abp1p, an actin binding protein originally isolated from yeast, also has an actin-depolymerizing factor homology domain, but it binds to Arp2/3 complex and activates the actin nucleation, not actin depolymerization.13
To better understand the cellular mechanisms by which GMFG enhances stimulus-responsive lamellipodia formation, we investigated the possible interaction of GMFG with the Arp2/3 complex, a key regulator of stimulus-responsive actin polymerization. Cell lysates expressing GMFG were incubated with anti–Arp2/3 complex subunit 2 (ARPC2) antibody, and coprecipitation of GMFG with ARPC2 was detected by immunoblotting. To exclude the potential nonspecific interactions from coprecipitation of actin filaments, 40 μmol/L latrunculin A was included in the immunoprecipitation reactions. Immunoblotting confirmed no detectable actin in the pellets (data not shown). GMFG was coprecipitated with ARPC2, suggesting GMFG interaction with Arp2/3 complex (Figure 6F). Of note, more GMFG-S2E was pulled down than GMFG-WT or GMFG-S4E, indicating that phosphorylation of Ser2 enhances GMFG interaction with the Arp2/3 complex (Figure 6F). These results suggest that in response to extracellular stimuli, GMFG might be phosphorylated at Ser2 through Rac and Cdc42, promoting its interaction with Arp2/3 complex as well as with F-actin, which then accelerates stimulus-responsive actin polymerization at the leading edge of cells. However, further analysis is certainly required to conclude its functional relevance with Rac and Cdc42. Given that even pseudophosphorylated form of GMFG required exogenous stimulus to modulate actin cytoskeleton reorganization, GMFG binding to Arp2/3 complex is not sufficient to initiate new actin polymerization.
GMFG Positively Regulates Actin-Based Cellular Functions
We then investigated GMFG functions in actin-based cellular functions in endothelial cells. As shown in Figure 1B, BAECs express only a minimal amount of GMFG, much less than in microvascular endothelial cells. Therefore, we used BAECs to examine the effect of GMFG overexpression on actin-based cellular functions. The majority of cells expressed the transgene 24 hours after transfection, as assessed by GFP-construct transfection (supplemental Figure II). BAECs overexpressing GMFG demonstrated significantly higher motility than mock-transfected BAECs (Figure 7A). Moreover, BAECs overexpressing GMFG demonstrated significantly enhanced tube formation on Matrigel as compared with the mock-transfected cells (Figure 7B). These results suggest that GMFG positively regulates actin-based cellular functions in endothelial cells, presumably through accelerating stimulus-responsive actin polymerization.
Finally, we examined GMFG expression in a cardiac ischemia/reperfusion model because both angiogenesis and inflammation play important roles in the pathophysiology of ischemic cardiovascular diseases. GMFG expression was significantly increased in the ischemia/reperfusion tissues (Figure 7C and 7D). Although further analysis, such as immunohistochemistry to clarify the cells express GMFG, is certainly needed to address GMFG function in the ischemia/reperfusion heart, this result suggests possible GMFG involvement in the pathophysiology of cardiovascular diseases.
Our data define a novel pathway in the regulation of actin cytoskeleton reorganization and actin-based cellular functions (Figure 7E). Given that inflammatory cells and endothelial cells have a close relationship and interact with each other in ischemic cardiovascular diseases, GMFG may play a role in the pathophysiology of such diseases.14,15
Discussion
GMFG is preferentially expressed in human microvascular endothelial cells but not in endothelial cells from other vascular beds. Given that microvascular endothelial cells are the most important cellular players in angiogenesis in adults and expression of GMFG enhanced migration and tube formation in BAECs, GMFG may play a role in angiogenesis in adults. During embryogenesis, GMFG is predominantly expressed in blood islands of the yolk sac, in which hematopoietic and endothelial lineages develop simultaneously in close proximity, suggesting GMFG functions in embryonic vasculogenesis as well as hematopoiesis. This may also well correlate with the unique GMFG expression in microvascular endothelial cells and inflammatory cells seen in adults. In support of this contention, a search of the database of microarray analysis at the National Center for Biotechnology Information revealed that GMFG is expressed in a variety of hematopoietic cells such as T cell, B cell, NK cell, monocyte, and dendritic cell, as well as hematopoietic progenitor cells including CD34+ and CD105+ bone marrow cells, which may give rise to endothelial progenitor cells.16,17
Our findings demonstrated that GMFG plays a significant role in actin cytoskeleton reorganization distinct from cofilin, despite structural similarity. Stable expression of the pseudophosphorylated form of GMFG remarkably enhanced stimulus-responsive lamellipodia and subsequent membrane ruffles formation, presumably through interaction with Arp2/3 complex. In response to extracellular stimuli, active Rac and Cdc42 might activate GMFG via phosphorylation of Ser2 as well as WASP/WAVE. Binding of phosphorylated GMFG to the Arp2/3 complex might accelerate stimulus-responsive actin polymerization cooperating with active WASP/WAVE, resulting in enhanced actin-based cellular functions. However, further analysis is required to verify this hypothesis.
Of note, the pseudophosphorylated form of GMFG did not affect actin cytoskeleton reorganization in HeLa cells under basal culture conditions, although it remarkably enhanced stimulus-responsive lamellipodia formation. These results suggest that binding of GMFG to Arp2/3 complex is not sufficient to initiate new actin polymerization, but once Arp2/3 complex is activated by other activator such as WASP/WAVE, GMFG may enhance Arp2/3 complex-induced actin polymerization. Therefore, GMFG may enhance actin-based cellular functions only in the presence, and under the control, of endogenous stimulators such as growth factors and cytokines. In this regard, GMFG may be a good target for gene therapy because unexpected adverse effect caused by dysregulated transgene function is a considerable concern for the clinical application of gene therapy.
Our results indicate that GMFG is a novel factor in actin cytoskeleton reorganization and that it modulates actin-based cellular function in microvascular endothelial cells, inflammatory cells, and possibly hematopoietic progenitor cells. Because these cells are essential players in angiogenesis, vasculogenesis, and immune system function, and each of them plays specific roles in ischemic cardiovascular diseases, we believe that further analysis of GMFG will provide new insights into the molecular mechanisms of these diseases and that GMFG may be an attractive new target for pharmacotherapeutic agents to treat such diseases.
Acknowledgments
We thank Dr Kiyofumi Asai for providing the anti-sera for GMFG. We thank Dr Yoshimi Takai for providing the expression constructs of dominant active and dominant negative Rho, Rac, and Cdc42. We thank Dr Alex Dunn for helpful suggestions for actin-binding and Arp2/3 complex–binding assays.
Source of Funding
This work was supported by the Donald W. Reynolds Foundation.
Disclosures
None.
Footnotes
Original received March 26, 2006; revision received July 12, 2006; accepted July 13, 2006.
References
Ho M, Yang E, Matcuk G, Deng D, Sampas N, Tsalenko A, Tabibiazar R, Zhang Y, Chen M, Talbi S, Ho YD, Wang J, Tsao PS, Ben-Dor A, Yakhini Z, Bruhn L, Quertermous T. Identification of endothelial cell genes by combined database mining and microarray analysis. Physiol Genomics. 2003; 13: 249–262.
Asai K, Fujita K, Yamamoto M, Hotta T, Moriyama M, Kokubo M, Moriyama A, Kato T. Isolation od novel cDNA (hGMF-gamma) homologous to glia maturation factor-beta gene. Biochim Biophys Acta. 1998; 1396: 242–244. [Order article via Infotrieve]
Pollard TD, Blanchoin L, Mullins RD. Molecular mechanisms controlling actin filament dynamics in nonmuscle cells. Annu Rev Biophys Biomol Struct. 2000; 29: 545–576. [Order article via Infotrieve]
Carlier MF, Ressad F, Pantaloni D. Control of actin dynamics in cell motility. Role of ADF/cofilin. J Biol Chem. 1999; 274: 33827–33830.
Yamazaki D, Suetsugu S, Miki H, Kataoka Y, Nishikawa S, Fujiwara T, Yoshida N, Takenawa T. WAVE2 is required for directed cell migration and cardiovascular development. Nature. 2003; 424: 452–456. [Order article via Infotrieve]
Bamburg JR, Wiggan OP. ADF/cofilin and actin dynamics in disease. Trends Cell Biol. 2002; 12: 598–605. [Order article via Infotrieve]
Zhang J, Shehabeldin A, da Cruz LA, Butler J, Somani AK, McGavin M, Kozieradzki I, dos Santos AO, Nagy A, Grinstein S, Penninger JM. Antigen receptor-induced activation and cytoskeletal rearrangement are impaired in Wiskott-Aldrich syndrome protein-deficient lymphocytes. J Exp Med. 1999; 190: 1329–1342.
Goode BL, Drubin DG, Lappalainen P. Regulation of the cortical actin cytoskeleton in budding yeast by twinfilin, a ubiquitous actin monomer-sequestering protein. J Cell Biol. 1998; 142: 723–733.
Bamburg JR. Proteins of the ADF/cofilin family: essential regulators of actin dynamics. Annu Rev Cell Dev Biol. 1999; 15: 185–230. [Order article via Infotrieve]
Ghosh M, Song X, Mouneimne G, Sidani M, Lawrence DS, Condeelis JS. Cofilin promotes actin polymerization and defines the direction of cell motility. Science. 2004; 304: 743–746.
Hirata KI, Ishida T, Penta K, Rezaee M, Yang E, Wohlgemuth J, Quertermous T. Cloning of an immunoglobulin family adhesion molecule selectively expressed by endothelial cells. J Biol Chem. 2001; 276: 16223–16231.
Ikeda K, Quertermous T. Molecular isolation and characterization of a soluble isoform of activated leukocyte cell adhesion molecule that modulates endothelial cell function. J Biol Chem. 2004; 279: 55315–55323.
Olazabal IM, Machesky LM. Abp1p and cortactin, new "hand-holds" for actin. J Cell Biol. 2001; 154: 679–682.
Nian M, Lee P, Khaper N, Liu P. Inflammatory cytokines and postmyocardial infarction remodeling. Cir Res. 2004; 94: 1543–1553.
Naldini A, Carraro F. Role of inflammatory mediators in angiogenesis. Curr Drug Targets Inflamm Allergy. 2005; 4: 3–8. [Order article via Infotrieve]
Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T, Witzenbichler B, Schatteman G, Isner JM. Isolation of putative progenitor endothelial cells for angiogenesis. Science. 1997; 275: 964–967.
Salven P, Mustjoki S, Alital R, Alitalo K, Rafii S. VEGFR-3 and CD133 identify a population of CD34+ lymphatic/vascular endothelial precursor cells. Blood. 2003; 101: 168–172.
您现在查看是摘要介绍页，详见ORG附件。