| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Physiology and Tumor Biology Program,* Gastrointestinal Research Unit, Mayo Clinic, Rochester, Minnesota; and the Division of Liver Diseases, Mount Sinai School of Medicine, New York, New York
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Hepatic stellate cells (HSCs) are liver-specific pericytes.6 HSC locomotion is integral to a number of fundamental biological and disease processes in liver7,8 and recapitulates the importance of pericytes in other organs as well.9 For example, angiogenesis-based invasion of metastatic liver lesions requires migration of HSCs as does the process of hepatic scarring and fibrosis.10-12 Recent studies have expanded our understanding of HSC migration by elucidating the chemotactic role of platelet-derived growth factor (PDGF) on HSCs.13,14 However, unlike more prototypical migratory cells such as fibroblasts, the mechanisms in HSCs that govern small GTPase function, actin remodeling, and locomotion remain poorly defined in part because of methodological challenges in performing detailed molecular studies in HSCs.15
Nitric oxide (NO), derived from endothelial cells lying adjacent to HSCs, plays an integral role in the remodeling of vascular networks through paracrine signaling pathways some of which are mediated though its downstream targets, guanylate cyclase and protein kinase G (PKG).16 However, the sum influence and signaling effectors of NO signals on cell migration are complex and vary markedly between different cell types. Furthermore, the observed actions of NO on cell migration have not been translated into a rational understanding of how NO interdigitates with small GTPase function and actin-based membrane remodeling.
In this study, we demonstrate, using DNA microinjection of Rho family GTPases in combination with morphological and functional assays in LX2, a cell line derived from human HSCs,13,15,17-19 that 1) Rac promotes migration by causing formation of actin-positive, filopodia spikes; 2) NO and PKG inhibit PDGF- and Rac-driven migration by inhibiting filopodia formation; and 3) Rho confers resistance to NO inhibition of migration and also restores chemotactic responses to PDGF in the absence of functional Rac. Thus, these studies identify novel crosstalk between small GTPases, cytoskeletal structures, and NO signaling pathways that provides counterbalances in the process of cell motility and migration.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
LX2 cells, rat HSCs, human skin fibroblasts (HSFs), NIH 3T3, and rat lung fibroblasts were used in this study. LX2 is a well-characterized cell line derived from human HSCs, which recapitulate many features of the activated HSC phenotype including expression of HSP 47 mRNA, MMP-2, and PDGF receptor ß subunit.15 Unlike primary HSCs, LX2 are amenable to detailed molecular interventions and have been broadly used for this purpose.13,15,17-19 LX2 were cultured in Dulbecco’s modified Eagle’s medium, supplemented with 10% fetal bovine serum, 1 mmol/L L-glutamine, and 100 IU/ml penicillin. HSCs were isolated from rat liver using collagenase digestion and density gradient centrifugation and cultured as we and others have previously described and studied between passages 1 to 3.20,21 HSF, NIH 3T3, and rat lung fibroblasts were grown in minimum Eagle’s medium, supplemented with 10% fetal bovine serum, 1 mmol/L L-glutamine, and 100 IU/ml penicillin.
For adenoviral transduction, cells were washed with phosphate-buffered saline (PBS) and incubated with PBS/0.1% albumin containing 50 multiplicity of infection of adenoviral vector encoding endothelial nitric oxide synthase (AdNOS; kindly provided by Dr. W. Sessa, Yale University, New Haven, CT), PKG (AdPKG, kindly provided by Dr. K. Bloch, Massachusetts General Hospital, Boston, MA), or adenoviral vector encoding green fluorescent protein (AdGFP) for 60 minutes.17,22 The vector solution was then aspirated, and after washing with PBS, complete culture medium was replenished and incubated for 24 hours. The transfection efficiency was confirmed by conventional epifluorescence microscopy and routinely close to 100%. cDNA encoding constitutive active and dominant-negative Rho (RhoQL and dnRho, respectively), constitutive active and dominant-negative Rac (RacQL and dnRac, respectively), and constitutive active and dominant-negative Cdc42 (Cdc42QL and dnCdc42, respectively), originally generated by Dr. J. Silvio Gutkind (National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, MD) by site-directed mutagenesis, were kindly provided by Dan Billadeau (Mayo Clinic, Rochester, MN) and Federico Kalinec (University of California at Los Angeles, Los Angeles, CA).23 Transient transfections were performed in LX2 using a standard calcium phosphate method with efficiency approximating 50%. Co-transfection of adenoviral vectors and plasmid vectors, or multiple plasmid vectors, were performed in some studies as indicated.
Migration Assay
Cellular chemotaxis was measured using modifications of a Boyden transwell migration assay, in which transwells were fitted with polyvinyl/pyrrolidine-free polycarbonate membranes (13-mm diameter, 8-µm pore size; Neuro Probes, Inc., Gaithersburg, MD). Polycarbonate filters were precoated with 50 µg/ml of human type I collagen for 30 minutes at 37°C. The bottom wells of the chamber were filled with 26 µl of serum-free media containing 10 ng/ml of PDGF or vehicle. Wells were covered with the coated membrane sheet, and 20,000 serum-starved cells were added into the upper chamber of each well. Test agents were added alone, or in combination, to the upper chamber of transwells (50 µl) and included the NO donor, sodium nitroprusside (SNP; 50 to 500 µmol/L), and an inhibitor of guanylate cyclase, ODQ (100 µmol/L). The chamber was incubated for 4 hours (37°C, 5% CO2), after which cells from the upper surface of membranes were completely removed with gentle swabbing. The remaining migrated cells on the lower surface of membranes were fixed and stained using Hema3 stain according to the manufacturer’s recommendations (Biochemical Sciences, Inc., Swedesboro, NJ). Membranes were then rinsed with distilled water and mounted onto glass slides. Cellular migration was determined by counting the number of stained cells on membranes in five randomly selected, nonoverlapping high-power fields. Each experiment was repeated three to five separate times with three to five replicate wells for each test condition.
Microscopic Analysis of Plasma Membrane Actin Structures
For phalloidin staining, cells were cultured on tissue culture slides, fixed in 4% paraformaldehyde in phosphate-buffered saline for 30 minutes at 4°C, and permeabilized with 0.1% Triton X-100 in phosphate-buffered saline for 1 minute at room temperature. Filamentous actin was stained with tetramethyl-rhodamine isothiocyanate (TRITC)-phalloidin in phosphate-buffered saline (1 µmol/L) for 20 minutes at room temperature. After mounting, cells were visualized by confocal laser-scanning microscopy. In some studies, cells were scored as positive or negative for lamellipodia or filopodial spikes. Lamellipodia-positive cells were defined as having at least one broad flat actin-positive protrusion, which was aligned parallel to the cell membrane, whereas filopodia-positive cells were defined as having more than 10 thin, actin-positive processes projecting beyond the cellular edge of the plasma membrane.24 Quantitative analyses were performed in a blinded manner from 200 individual cells per each group and from three independent experimental preparations.
For live cell imaging, cells were plated and cultured on 32-mm coverslips and mounted in fresh medium in the 37°C stable temperature control chamber (POC-R; Zeiss, Oberkochen, Germany).25 The chamber was placed on the stage of a microscope (Pasqual LSM 5; Zeiss) and covered. PDGF (10 ng/ml) was added to the chamber and phase contrast images were captured every minute for 25 minutes using a x63 C-apochromat lens. To measure the area of lamellipodia protrusion, cell morphology was traced using Adobe Photoshop (Adobe Systems, Mountain View, CA) from images that were captured every 5 minutes from live cells. The area of lamellipodia protrusion and the starting area of each cell were quantitated using NIH Image, and the ratio of these two values was calculated and expressed as a percentage per cell for each of the conditions with 10 to 15 cells analyzed per experimental group.
DNA Microinjection
For microinjection experiments, cells (104) were plated and cultured on 55-µm2 CELLocate coverslips (Eppendorf Scientific, Inc., Hamburg, Germany). Optimal injections were obtained with sterile femtotips II (diameter of opening of tip, 0.5 µm ± 0.2 µm; Eppendorf Scientific, Inc.). RhoQL, RacQL, or empty plasmid DNA (10 µg/ml) were dialyzed in microinjection buffer (10 mmol/L KH2PO4, pH 7.2, and 75 mmol/L KCl) and Alexa Fluor 488 (hydrazide sodium salt, 200 mmol/L KCl; Molecular Probes, Eugene, OR) was added to all microinjection solutions to enable identification of injected cells. Cells were pressure injected using an Eppendorf InjectMan NI2 and FemotoJet 5247 on a Zeiss Axiovert inverted microscope, with an optimized program adjusted for injection pressure (Pi, 140 hPa; 1 hPa = 0.0145 PSI), compression pressure (Pc, 40 hPa), and time (T, 0.4 second).23 Images were obtained using an intensified charge-coupled device camera attached to a Zeiss Axiovert 35 microscope (Carl Zeiss, Inc.). After injection, cells were allowed to recover 4 to 6 hours at 37°C in a 5% CO2 incubator and then treated with vehicle or 500 µmol/L SNP. Twenty-four hours after injection, cells were fixed in 4% paraformaldehyde in phosphate-buffered saline and stained for actin with TRITC-phalloidin in phosphate-buffered saline (1 µmol/L). Quantitation of filopodia was performed by counting filopodial spikes from captured images of injected cells (n = 5 cells per experimental group).
Rac and Rho Activation Assays and Western Blot
Rac and Rho activation assays were performed as described previously.26
In brief, glutathione S-transferase (GST)-p21-binding domain (GST-PBD) and Rhotekin Rho binding domain (GST-TRBD) fusion constructs, kindly provided by Dr. D. Mukhopadhyay, Mayo Clinic, Rochester, MN were transformed into BL21 (DE3) and induced with IPTG (1 mmol/L) overnight. Bacterial suspensions were prepared, aliquotted, and frozen at –80°C. To prepare the GST-PBD and GST-TRBD beads, each aliquot of frozen bacteria was resuspended in 2 ml of cold PBS, and then 20 µl of 1 mol/L dithiothreitol, 20 µl of 0.2 mol/L phenylmethyl sulfonyl fluoride, and 40 µl of 50 mg/ml lysozyme were added and incubated on ice for an additional 30 minutes. Ten percent Triton X-100 (225 µl), 22.5 µl of 1 mol/L MgCl2, and 22.5 µl of 2000 kU/ml DNase were then added, and the samples were incubated on ice for another 30 minutes. The supernatant was collected and incubated with 100 µl of glutathione-coupled Sepharose 4B beads (Amersham Biosciences) at 4°C for 45 minutes. The beads were then washed (PBS with 10 mmol/L dithiothreitol and 1% Triton X-100) and resuspended in the same buffer to generate a 50% bead slurry. For binding experiments, Rac, Rho, or mock-transfected cells were lysed with a buffer (150 mmol/L NaCl, 0.8 mmol/L MgCl2, 5 mmol/L EGTA, 1% IGEPAL, 50 mmol/L HEPES, pH 7.5, 1 mmol/L phenylmethyl sulfonyl fluoride, 10 µg/ml leupeptin, and 10 µg/ml aprotinin). In some experiments cells were pretreated with 500 µmol/L SNP for 1 hour before lysis. The supernatant was isolated and incubated with 50 µl of GST-PBD or GST-TRBD beads at 4°C for 45 minutes. Bound proteins were washed with a buffer (50 mmol/L Tris-HCl, pH 7.2, 1% Triton X-100, 150 mmol/L NaCl, 10 mmol/L MgCl2, 1 mmol/L phenylmethyl sulfonyl fluoride, 10 µg/ml leupeptin, and 10 µg/ml aprotinin), eluted, separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and analyzed by Western blot with pAb specifically recognizing Rac or Rho (Santa Cruz Biotechnologies, Santa Cruz, CA). For positive control, cell lysates were incubated for 15 minutes at 30°C in the presence of 10 mmol/L EDTA and 100 µmol/L GTP
S. Western blot from cell lysates was also performed using eNOS mAb (Transduction Laboratories, Lexington, KY).
Statistical Analysis
Data are presented as mean ± SE. Data were analyzed using paired and unpaired Student’s t-tests as appropriate.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The small GTPases, Rac, Rho, and Cdc42 mediate signaling pathways, which regulate the formation of distinct actin-based structures essential to cell motility and chemotactic responses. Therefore, we examined the influence of Rho family small GTPases on LX2 migration by transfecting cells with the constitutive active constructs RacQL, RhoQL, Cdc42QL, or the dominant-negative constructs dnRac, dnRho, or dnCdc42, and assessing migration in response to the prototypical chemotactic growth factor, PDGF, using a modified Boyden chamber assay. As seen in Figure 1A
, both LX2 and a prototypical migratory fibroblast cell, HSF, evidenced chemotactic responses to PDGF, although the magnitude of effect was sometimes quantitatively greater in fibroblasts. In transfection experiments performed in LX2, although neither RhoQL nor Cdc42QL influenced basal or PDGF-induced migration, RacQL significantly increased basal migration and migration in response to PDGF compared to cells transfected with empty vector. As seen in Figure 1B
, in experiments using the dominant-negative constructs in LX2, dnRac inhibited PDGF-induced migration whereas neither dnRho nor dnCdc42 influenced PDGF-induced migration. These studies highlight the important role of Rac in PDGF-induced migration of LX2.
|
In fibroblasts, PDGF initiates a PI-3 kinase-dependent pathway that activates Rac and promotes lamellipodia.5
However, the influence of PDGF on actin membrane structures in HSCs is unexplored. Therefore, we performed microscopic analysis of actin membrane structures in response to PDGF to compare and contrast events in fibroblasts and LX2. HSF in serum-free medium have sporadic thin protrusions (Figure 2A
, left). PDGF promotes formation of broad plasma membrane extensions parallel to the substrate, characteristic of lamellipodia (Figure 2A
, right). A larger percentage of fibroblasts evidenced protrusions in response to PDGF in a time-dependent manner (Figure 2A
, far right graph). However, PDGF had a surprisingly different effect in LX2. Similar to fibroblasts, LX2 were relatively devoid of lamellipodia in serum-free control conditions (Figure 2B
, left). However, PDGF stimulated the proliferation of spike-like filopodia structures rather than lamellipodia (Figure 2B
, right). This occurred in a time-dependent manner as evidenced in the graph (Figure 2B
, far right graph) with the majority of cells evidencing numerous filopodial structures after PDGF stimulation. Prominent filopodial structures were also observed in primary rat HSCs in response to PDGF (Figure 2C
; middle panel with arrow pointing to higher magnification image).
|
, left), significant morphological changes in HSCs were observed after 24 hours of RacQL nuclear microinjection. RacQL DNA-injected cells developed numerous actin-positive, extended filopodia spikes reminiscent of that seen in response to PDGF (Figure 2D
, right). This effect was observed in each of the five cells that were studied after microinjection and the graph (far right) depicts computation of the significantly increased number of filopodial structures per cell in LX2 microinjected with RacQL as compared to empty plasmid or RhoQL. In contrast, cells injected with RhoQL appeared to evidence an increase in stress fiber formation compared to control cells (Figure 2D
, middle), although the cellular effects were not as prominent as that observed in cells injected with Rac. NO Inhibits Rac and PDGF-Induced Cell Migration
The stimulatory effect of RacQL overexpression on PDGF-induced HSC chemotaxis and on filopodia formation, in conjunction with the key role of NO on vascular remodeling and cell migration, prompted us to next examine the influence of NO on this chemotactic pathway in HSCs. Although HSCs do not basally express NOS, they are influenced by NO signals from adjacent endothelial cells and do express NOS under specific circumstances.27
Influence of NO signals was studied by complementary approaches, including NOS gene transfer to mimic endogenous intracellularly produced NO and the NO donor SNP to resemble paracrine-derived NO signals. First, LX2 and HSF were treated with the NO donor SNP at varying concentrations at the time of initiation of the migration assay. In both cell types, SNP inhibited PDGF-induced chemotaxis in a concentration-dependent manner with almost complete inhibition of PDGF-induced migration observed at 500 µmol/L SNP (Figure 3A
, left). Qualitatively, similar results demonstrating SNP inhibition of PDGF-induced chemotaxis were obtained in primary isolates of rat HSCs as well as two other representative fibroblast cell types including rat lung fibroblasts and NIH 3T3 cells (Figure 3A
, right). Next, experiments were performed after transduction of cells with AdNOS or AdGFP. Chemotactic response to PDGF was attenuated in both LX2 and HSFs that were transduced with AdNOS, as compared to PDGF chemotaxis in cells transduced with AdGFP, although the magnitude of effect of NO inhibition in these experiments appeared larger in fibroblasts (Figure 3B)
. The detected inhibition of PDGF-induced chemotaxis in response to NO was not secondary to NO-induced apoptosis as evidenced by a lack of increase in DNA fragmentation in cells transduced with AdNOS as compared to cells transduced with AdGFP (percent apoptotic cells: AdGFP, 2.7 ± 0.3%; AdeNOS, 3.0 ± 0.2%; NS).
|
, both SNP as well as AdNOS gene transfer markedly attenuated PDGF-induced migration in cells overexpressing RacQL, reducing chemotactic responses almost to that level observed in control cells. These studies establish an inhibitory effect of NO signaling on PDGF- and Rac-induced LX2 migration. The biological effectors of some NO signaling responses include guanylate cyclase, which generates cGMP in response to NO stimulation. This led us to examine how modulation of the guanylate cyclase pathway affects LX2 migration. As seen in Figure 4A
, the guanylate cyclase inhibitor ODQ did not rescue fibroblasts from the inhibitory effect of NO signals derived from AdNOS on migration responses to PDGF nor did ODQ abrogate the inhibitory effects of NO signals derived from SNP on migration responses to PDGF (Figure 4B)
. The lack of effect of ODQ was corroborated by direct measurements of cellular cGMP levels in LX2 under basal- and SNP-stimulated conditions, which demonstrated that cGMP levels were barely detectable in LX2 cells (data not shown). These results in LX2 are consistent with previous data by Failli and colleagues28
demonstrating that guanylate cyclase expression is down-regulated in highly activated HSCs. We next examined if adenoviral gene delivery of a kinase downstream from NO and guanylate cyclase, PKG, could reconstitute inhibition of PDGF-induced migration in LX2. Owing to the lack of cGMP detected in LX2, these experiments were all performed in the presence of the exogenous PKG substrate, 8-Br-cGMP. As seen in Figure 4C
, migration in response to PDGF was markedly inhibited in HSCs transduced with AdPKG as compared to cells transduced with AdGFP reminiscent of that observed with AdNOS gene transfer. These studies indicate that in LX2, NO inhibits PDGF-induced migration independent of guanylate cyclase but that PKG can mimic the inhibitory effects of NOS overexpression on migration responses.
|
We next sought to examine the mechanism by which NO inhibits migration responses induced by PDGF in HSCs and contrast this to fibroblasts. First, HSFs were transfected with AdGFP or AdNOS and 24 hours later were mounted on a microscope for real-time evaluation of dynamic membrane changes in response to PDGF and NO signals in live cells. As seen in Figure 5A
, in fibroblasts transfected with AdGFP, PDGF increased lamellipodial protrusion area in a time-dependent manner. However, protrusion extension in response to PDGF was attenuated in fibroblasts transduced with AdNOS by 50% compared to cells that were transduced with AdGFP. The attenuation in PDGF-induced lamellipodial protrusion in response to NO signals is also depicted in the representative micrographs of phalloidin-stained cells obtained after 30 minutes of observation. Further complimentary quantitative analyses demonstrated that AdNOS transduction was also associated with a corroborative reduction in the number of cells evidencing lamellipodial protrusions in response to PDGF treatment (data not shown). As PDGF and RacQL induced the proliferation of filopodial spike structures rather than lamellipodia in LX2, we next examined the influence of NO on this process more directly by performing RacQL DNA microinjection studies. In LX2 injected with RacQL DNA, an abundance of filopodial spikes was detected on phalloidin staining of cells (Figure 5B
, top). However, in LX2 injected with RacQL DNA that were treated with SNP, actin-positive filopodia spike formation was dramatically reduced (Figure 5B
, bottom). This effect was observed in each of five cells that were studied and quantitation of the reduction in number of filopodia per cell in LX2 injected with RacQL and incubated with SNP as compared to cells injected with RacQL, is shown in the graph (Figure 5B
, far right). Because signals downstream from Rac are responsible for filopodial spikes in LX2 and are blocked by NO, we hypothesized that NO may inhibit signals either at or downstream from Rac. Therefore, we next examined whether NO directly inhibits Rac activation by pull-down assay using GST-PBD fusion protein that binds selectively to GTP-bound active Rac. LX2 transfected with empty vector or RacQL were treated with SNP or vehicle. As seen in Figure 5C
(left), these analyses demonstrated Rac activation in cells overexpressing RacQL compared to cells treated with empty vector. However, direct inhibition of Rac activity by SNP was not detected. Similar results were obtained in cells transfected with AdNOS used to generate NO signals in place of SNP (data not shown). A parallel protocol was used to examine the influence of SNP on Rho activity except that GST-PBD fusion protein was substituted with GST-TRBD, which binds to GTP-bound Rho. As seen in Figure 5C
(right), although Rho overexpression was associated with increased Rho activity, SNP inhibition of Rho was not detected. Taken together, these studies indicate that although the inhibitory effect of NO signals in fibroblast migration is due to effects on lamellipodia, inhibitory effects of NO in LX2 occur through effects on filopodia formation through signals residing downstream from Rac but not through direct biochemical inhibition of Rac or Rho catalytic activity.
|
In fibroblasts, Rho is frequently activated downstream from Rac,4
however the interplay of Rho and Rac in HSCs is less understood. Because NO signals inhibited Rac-dependent actin structures, we next examined the influence of RhoQL overexpression on NO inhibition of PDGF-induced LX2 chemotaxis. In contrast to the SNP inhibition of migration that was observed in HSCs transfected with RacQL, HSCs transduced with RhoQL were completely protected from SNP inhibition of PDGF-induced migration (Figure 6A)
. Experiments were also performed after co-transfection of HSCs with both RhoQL and AdNOS to assess the influence of Rho activation on locally derived NO inhibitory signals in HSCs. Akin to that observed in the aforementioned experiments using the NO donor SNP, RhoQL rescued cells from the migration inhibition phenotype conferred by AdNOS (Figure 6A)
.
|
. These studies indicate that Rho, can reconstitute PDGF-induced chemotaxis in the absence of functional Rac. |
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Cell protrusion, and subsequent crawling, is driven in part by polymerization of actin within lamellipodia and filopodia, organelles that are regulated by signaling pathways driven through Rho, Rac, and Cdc42.1 Small GTPase regulation of actin cytoskeleton in response to chemotactic growth factors is best characterized in fibroblasts. Previous studies have demonstrated that in fibroblasts, Cdc42 promotes filopodia; Rho, thought to be situated downstream from Rac, promotes stress fiber formation; whereas Rac drives lamellipodia formation.3,4 However, this hierarchical paradigm is less applicable in other cell types. Furthermore, signaling redundancy and cross-talk in these pathways has allowed for looser and sometimes contrary interpretations of these paradigms.5 Overall, the interplay of these pathways has not been studied in pericytes such as those residing in liver, in which migration is essential to a number of pathophysiological conditions including cirrhosis and tumor angiogenesis. In our studies, although Rac-induced lamellipodia drive migration in prototypical motile cells such as fibroblasts,3,4 LX2 were relatively devoid of lamellipodia, and Rac promotes filopodia spikes rather than lamellipodia protrusions as directly evidenced by microinjection of DNA encoding RacQL. An important role for these Rac-dependent spikes in the migration of LX2 is evidenced by the observation that inhibition of filopodia formation by NO signals also inhibits migration responses. A recent study also supports a primary role for Rac-induced filopodia in border cell migration during Drosophila oogenesis.33 What are the reasons for differing interplay of small GTPases and actin membrane structures in different cell types? Interestingly, Rac signals link to filopodia formation in many central nervous system cell types also.34 As HSCs as well as pericytes in other organs maintain several phenotypic similarities to central nervous system type cells, as evidenced by expression of the glial precursor proteoglycan, NG2, neural cell adhesion molecule, N-CAM, and the neuronally enriched, small GTP-ase RhoN, parallels in cellular origin of these cell types may be important in the phenotypic differences detected in these cells compared to fibroblasts and other epithelial type cells.35-38 Therefore, the specialized nature of certain cell types may result in such adaptive responses whereby Rac promotes migration through filopodia formation. It is also possible that the activation state of HSCs may also influence the interplay between small GTPases and actin dynamics as LX2 mimics the phenotype of highly activated HSCs.13 Although Cdc42 and Rho have little noticeable effect on basal migration or PDGF chemotaxis, the prokinetic actions conferred by Rho are evidenced under conditions that counteract PDGF chemotaxis including Rac loss-of-function or augmentation of NO signals. However, biochemical inhibition of Rac or Rho GTPase activity was not detected in the presence of NO suggesting that inhibitory actions of NO on these pathways may occur at the level of a regulatory guanine exchange factor or GTPase activating protein, rather than directly on the catalytic activity of Rac/Rho themselves.5
Anti-angiogenic therapies for cancer have focused in large part on the endothelial cell as a target. Owing in part to its stimulatory effect on endothelial cell permeability and migration,26,39 NO has been implicated in the process of tumor angiogenesis as well.40 Indeed, NO is an essential downstream signal for vascular endothelial growth factor-dependent endothelial cell activation.26,39 However, tumor microvasculature in metastatic and primary liver cancers is comprised of not only endothelial cells but also adjacent pericytes.10,11,41-43 The influence of NO on migration of nonendothelial vascular cell types is varied with reports demonstrating promigratory effects of NO in smooth muscle44 and anti-migratory effects in HSCs.28 Furthermore, few studies have extended the influence of NO on migration to understand the actions of NO signals on specific cell biological correlates of migration such as actin membrane dynamics and small GTPase function. The present studies demonstrate that, contrary to its promigratory effects on endothelial cells and in some studies smooth muscle cells, NO inhibits Rac signals that drive formation of filopodia and ensuing LX2 migration.
Recent studies have highlighted the essential function of the PDGF system in the process of pericyte recruitment to tumor vessels with inhibition of this pathway importantly limiting tumor angiogenesis and growth.43 Therefore, the inhibitory actions of NO on PDGF-dependent pericyte migration may have significant inhibitory effects on tumor vessel morphology, particularly in vascular beds such as liver, which are highly dependent on pericytes for angiogenesis, vascular extension, and remodeling. In this regard, the present studies provide a number of mechanistic advances in the membrane biology and signaling pathways that regulate motility and migration thereby advancing potential avenues for therapeutic interventions aimed at targeting HSCs for tumor anti-angiogenesis and other disease processes.
|
Footnotes |
|---|
Supported by the National Institutes of Health (grants R01 DK59615, R01 DK59388, P50CA102701, and a Pancreatic Cancer SPORE Development Grant to V.S.; and R01DK37340 and R01DK56621 to S.F.) and the Korean Association Study of Liver Disease Glaxo Wellcome Hepatologist Fellowship Fund (to J.S.L.).
Accepted for publication February 11, 2005.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. L. Friedman Hepatic Stellate Cells: Protean, Multifunctional, and Enigmatic Cells of the Liver Physiol Rev, January 1, 2008; 88(1): 125 - 172. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Cetin, C. L. Leaphart, J. Li, I. Ischenko, M. Hayman, J. Upperman, R. Zamora, S. Watkins, H. R. Ford, J. Wang, et al. Nitric oxide inhibits enterocyte migration through activation of RhoA-GTPase in a SHP-2-dependent manner Am J Physiol Gastrointest Liver Physiol, May 1, 2007; 292(5): G1347 - G1358. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Das, M. E. Fernandez-Zapico, S. Cao, J. Yao, S. Fiorucci, R. P. Hebbel, R. Urrutia, and V. H. Shah Disruption of an SP2/KLF6 Repression Complex by SHP Is Required for Farnesoid X Receptor-induced Endothelial Cell Migration J. Biol. Chem., December 22, 2006; 281(51): 39105 - 39113. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Sicklick, S. S. Choi, M. Bustamante, S. J. McCall, E. H. Perez, J. Huang, Y.-X. Li, M. Rojkind, and A. M. Diehl Evidence for epithelial-mesenchymal transitions in adult liver cells Am J Physiol Gastrointest Liver Physiol, October 1, 2006; 291(4): G575 - G583. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q Zhou, M Hennenberg, J Trebicka, K Jochem, L Leifeld, E Biecker, T Sauerbruch, and J Heller Intrahepatic upregulation of RhoA and Rho-kinase signalling contributes to increased hepatic vascular resistance in rats with secondary biliary cirrhosis Gut, September 1, 2006; 55(9): 1296 - 1305. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Liu, J. Hou, X. Hu, X. Wang, Y. Xiao, Y. Mou, and H. De Leon Neuronal Chemorepellent Slit2 Inhibits Vascular Smooth Muscle Cell Migration by Suppressing Small GTPase Rac1 Activation Circ. Res., March 3, 2006; 98(4): 480 - 489. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. E. Perri, D. A. Langer, S. Chatterjee, S. J. Gibbons, J. Gadgil, S. Cao, G. Farrugia, and V. H. Shah Defects in cGMP-PKG pathway contribute to impaired NO-dependent responses in hepatic stellate cells upon activation Am J Physiol Gastrointest Liver Physiol, March 1, 2006; 290(3): G535 - G542. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
