This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Aepfelbacher, M. Articles by Weber, P. C. Search for Related Content PubMed PubMed Citation Articles by Aepfelbacher, M. Articles by Weber, P. C.

(Arteriosclerosis, Thrombosis, and Vascular Biology. 1997;17:1623-1629.)
© 1997 American Heart Association, Inc.

Articles

Bacterial Toxins Block Endothelial Wound Repair

Evidence That Rho GTPases Control Cytoskeletal Rearrangements in Migrating Endothelial Cells

Martin Aepfelbacher; Markus Essler; Elisabeth Huber; Motoyuki Sugai; ; Peter C. Weber

From the Institut für Prophylaxe und Epidemiologie der Kreislaufkrankheiten (M.A., M.E., E.B., P.C.W.), University of Munich, Germany; and the Department of Microbiology (M.S.), Hiroshima University, Japan.

Correspondence to Martin Aepfelbacher, Institut für Prophylaxe der Kreislaufkrankheiten, Pettenkoferstr. 9, 80336 München, Germany.

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Abstract We investigated the effect of bacterial toxins that modify and inactivate Rho GTP-binding proteins on the migratory response of endothelial cells to wounding. C3-transferase from Clostridium botulinum, EDIN from Staphylococcus aureus, and toxin A from Clostridium difficile blocked migration of human umbilical vein endothelial cells (HUVECs) in an in vitro wound repair assay. Migrating HUVECs expressed actin microspikes (maximum at 10 minutes after wounding), ruffles (maximum at 12 hours), and fibers (maximum at 24 hours), and within these actin structures, vinculin-containing focal complexes/adhesions were formed. C3-Transferase ADP ribosylated RhoA, RhoB, and RhoC in HUVECs and abolished the formation of actin stress fibers/focal adhesions but had no effect on expression of microspikes, ruffles, or the associated vinculin-containing focal complexes. Similar results were obtained with EDIN and toxin A. These results indicate that endothelial cells migrating into a wounded area express distinct combinations of actin/vinculin structures in a spatially and temporally coordinated manner. The GTPase Rho selectively controls the formation of actin fibers/focal adhesions that occurs 2 to 24 hours after wounding. A mechanism is proposed by which Rho-specific bacterial toxins could influence vascular repair, angiogenesis, or atherosclerosis.

Key Words: vascular endothelium • bacterial toxins • cell migration • ADP ribosylation • Rho GTPase

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Endothelial cell migration is a critical event in organ development, tumor angiogenesis, and the repair of injured blood vessels.^{1 2 3} Wound repair models using EC cultures revealed a complex response that involves migration and spreading of cells at the wound edge, as well as proliferation of cells distal to the wound.⁴ Sholley et al⁵ demonstrated that cell migration and spreading was responsible for wound repair within the first 24 hours, followed by cell proliferation after 36 hours. Changes in actin organization seem to be important for the EC responses, because cytochalasin B, an inhibitor of actin polymerization, blocked postinjury migration of ECs⁶ as well as rapid lamellipodia-mediated wound closure.⁷ In addition, tyrosine kinase inhibitors could reduce EC wound repair by interfering with the formation of focal adhesion sites,⁸ where actin stress fibers are connected to extracellular integrin receptors via multimolecular protein complexes.⁹ Hence, migration of ECs during wound repair seems to require a finely tuned balance between actin filament polymerization/depolymerization and integrin receptor activation/deactivation.¹⁰

It has become well established that Ras-related GTPases of the Rho family organize the actin cytoskeleton and focal adhesion formation.¹¹ The family of Rho-like proteins in mammalian cells consists of RhoA, RhoB, RhoC, RhoD, RhoE, and RhoG; Rac1 and Rac2; TC10; TTF, and two CDC42Hs isoforms.¹² Rho GTPases bind and hydrolyze GTP, and their cycle between a GDP-bound inactive and a GTP-bound active state is regulated by GTPase-activating proteins, guanine nucleotide exchange factors, and guanine nucleotide dissociation inhibitors.¹³ RhoA, RhoB, and RhoC can specifically be inactivated by bacterial toxins from C. botulinum and S. aureus, which act as ADP ribosyltransferases,¹⁴ or toxins A and B from C. difficile, which act as glucosyltransferases.¹⁵ ADP ribosylation and glucosylation take place at Asn⁴¹ and Thr³⁷, respectively, in the putative effector region of Rho and presumably block the interaction of Rho with downstream targets.¹¹ By use of these toxins, Rho was identified as a regulator of cell contraction, adhesion, division, and motility.¹¹ The molecular function of Rho thereby seems to be to induce the formation of stress fibers and focal adhesion sites.¹⁶

Given the increasing number of naturally occurring bacterial toxins that inactivate one or more Rho GTPases, the question of their pathogenetic role arises. When these toxins enter the bloodstream, they will likely affect ECs and thus could potentially modulate inflammation, angiogenesis, and/or atherogenesis.

We demonstrate here that the ADP ribosyltransferases from C. botulinum and S. aureus as well as the glucosyltransferase from C. difficile inhibit EC migration and wound repair by deactivating Rho and blocking actin filament and focal adhesion formation.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Materials and Toxins
RPMI-1640, L-glutamine, fetal calf serum, and all other materials not specifically indicated were from Sigma. Anti-RhoA and anti-RhoB antibodies were purchased from IC Chemikalien and tested with recombinant glutathione S-transferase fusion proteins of RhoA¹⁷ on Western blots.

The C3 exoenzyme was expressed as a glutathione S-transferase fusion protein in E. coli strain NM522. The pGEX2T-C3 exoenzyme expression vector was a generous gift from Dr L. Feig (Tufts University, Boston, Mass). Expression of the fusion protein was induced by isopropyl thiogalactopyranoside (Boehringer), bacteria were lysed by sonication, and the fusion protein was purified with glutathione Sepharose 4B beads (Pharmacia) and cleaved with thrombin (Boehringer) as described in detail.¹⁷ Purity of the C3 exoenzyme was tested on SDS/12% PAGE, and protein concentration was determined with the BCA method (Pierce), using BSA as standard. EDIN was purified exactly as described.¹⁸ C. difficile toxin A was a generous gift from Drs F. Hofmann and I. Just, Universität Freiburg, Freiburg, Germany.

HUVEC Culture and Migration Assay
HUVECs were obtained by cannulating segments of human umbilical cord veins followed by incubation with -chymotrypsin (Sigma) for 30 minutes and gentle irrigation. The irrigant was collected and harvested cells were plated onto collagen-coated (collagen G consisting of 90% collagen I and 10% collagen III, from Biochrom) plastic culture flasks. HUVECs were cultured in endothelial growth medium (Promo Cell) containing EC growth supplement/Heparin and 10% fetal calf serum at 5% CO₂ and 37°C in a humidified atmosphere. The identity of HUVECs was verified by cobblestone morphology and positive staining for factor VIII. HUVECs were used between passages 2 and 4. For wound repair experiments, they were plated onto collagen-coated six-well dishes (Falcon) at a density of 1x10⁴/cm² and grown to confluence with medium changes every 2 to 3 days. After reaching confluence, HUVECs were cultured for 3 to 5 more days until a homogenous, quiescent monolayer with the typical cobblestone morphology developed. Thereafter, toxins were added for indicated time periods. Wounds were made by dragging a sterile pipette tip across the monolayer to create a cell-free path 1 mm wide. Quantification of cell migration was done under phase contrast on an inverted microscope with a 10x objective and a 10x ocular lens equipped with a 1-mmx1-mm grid consisting of 16 squares. At indicated times after wounding, the grid was positioned over the wound path and the number of cells that had migrated into the wounded area was determined. To relocate the grid exactly, two 1- to 2-cm-long fine scratches forming a cross line were made in the center of the wound path with a diamond pencil. Each value represents the mean of three different experiments with a minimum of 500 counted cells per experiment.

Microinjection
Cells were microinjected with C3-transferase as described elsewhere.¹⁶ Briefly, recombinant C3-transferase at 25 µg/mL in a buffer containing (in mmol/L) 50 Tris, pH 7.5; 50 NaCl; 5 MgCl2; and 0.1 DTT was injected into the cytoplasm of HUVECs with an Eppendorf microinjector. In some experiments, efficiency of microinjection was tested by coinjection with mouse IgG, followed by FITC-conjugated goat anti-mouse IgG and fluorescence microscopy.

ADP Ribosylation, 2D Gel Electrophoresis, and Western Blot
HUVECs were trypsinized, pelleted by centrifugation (150g, 5 minutes), and washed once in PBS (pH 7.4). Then they were resuspended at 2x10⁶/mL in ice-cold buffer containing 250 mmol/L sucrose; 1 mmol/L EDTA; 5 mmol/L MgCl₂; 10 mmol/L HEPES (pH 7.4); 1 mmol/L DTT; 0.1 mmol/L GTP; 1 µg/mL leupeptin, pepstatin, and aprotinin; and 2 mmol/L PMSF and sonicated for 10 seconds at 50 W with a Labsonic U (B. Braun). This homogenate (100 µL) was added to 50 µL of a solution containing 150 mmol/L Tris-HCL (pH 8), 30 mmol/L thymidine, 3 mmol/L DTT, 5 mmol/L MgCl₂, 0.3 mmol/L GTP, and 0.5 µg recombinant C3 exoenzyme on ice. The reaction was started by adding 1 µCi [³²P]NAD (800 Ci/mmol, NEN) to the sample; then it was incubated for 15 minutes in a 37°C water bath and precipitated for 45 minutes at -20°C with 6 vol ice-cold acetone. Precipitates were analyzed by SDS/12% PAGE (Mini-Protean II, Bio-Rad) and autoradiography with Kodak X-OMAT films. For 2D gel electrophoresis, acetone-precipitated proteins were dissolved in 50 µL of 9 mol/L urea, 2% Nonidet P-40, 2% mercaptoethanol, 0.8% ampholytes (pH 3 to 10), and 0.1% bromophenol blue and then subjected to isoelectric focusing on a 4% polyacrylamide gel containing 4.4% ampholytes, pH 3 to 10, and 0.7% ampholytes, pH 5 to 7 (Biolyte, Bio-Rad), essentially as described by O'Farrell.¹⁹ The pIs of the [³²P]ADP-ribosylated proteins were determined in pH gradient calibration curves obtained with a mixture of pI markers (4.6, 5.4, 5.9, 6.6, 6.8, 7.2; Sigma). The second dimension was performed with SDS/12% PAGE, using the Protean II xi system (Bio-Rad). For Western blot, proteins were electrophoretically transferred to Immobilon polyvinylidene difluoride membranes (Millipore) for 90 minutes at 195 mA in 20 mmol/L Tris/192 mmol/L glycine. Membranes were blocked for 1 hour in a solution containing 20 mmol/L Tris-HCL (pH 7.4)/150 mmol/L NaCl/0.3% Tween/10% calf serum; then they were incubated for 1 hour with mouse monoclonal or rabbit polyclonal anti-RhoA or rabbit polyclonal anti-RhoB antibody (all 1:2000, IC Chemikalien). After incubation for 45 minutes with peroxidase-labeled sheep anti-mouse or donkey anti-rabbit IgG, respectively (1:5000; Amersham), blots were developed using enhanced chemiluminescence (ECL, Amersham) and then exposed to X-Omat AR film (Kodak).

Northern Blot
Total cellular RNA from HUVECs was separated on 1% agarose gels, transferred onto Nytran N membranes (Schleicher & Schuell) with a vacuum blotter using 10x SSC and fixed under UV. Prehybridization was performed in 5x SSPE, 50% formamide containing 2x Denhardt's solution, 0.1% SDS, and 0.1 mg/mL salmon sperm DNA for at least 2 hours at 45°C. Hybridization with the oligonucleotides was done in the same solution overnight. Filters were washed three times for 30 minutes each in 0.2x SSPE and 0.1% SDS. For ³²P labeling, 20 pmol of specific oligonucleotides GGTTTCACAAGACAAGGCAACCA (RhoA), TTCTGGGAGCCGTAGCGCTTC (RhoB), and CGACGCTTGTTCTTGCGGACC (RhoC) were incubated with 100 µCi [³²P]gATP (Amersham), and 20 U T4 polynucleotide kinase (Pharmacia) for 1 hour at 37°C. After addition of 6 µg carrier tRNA, the phosphorylated oligonucleotides were precipitated twice with 4 mol/L ammonium acetate and 100% ethanol. Labeling efficiencies were similar for the RhoA, RhoB, and RhoC oligonucleotides. The number of radioactive counts used for hybridizations and the exposure times of the blots were identical for RhoA, RhoB, and RhoC to ensure comparability of band intensities. Specificities of the oligonucleotide probes were tested with rhoA DNA.

Immunofluorescence
Glass coverslips (13.7 mmx13.7 mm, No. 1, Menzel Gläser) were sonicated in dH₂O, soaked for 2 hours in ether/ethanol (1:1, vol/vol), and dried. These coverslips were coated for 2 hours or overnight with 100 µg/mL collagen G (Biochrom) and washed twice with PBS. HUVECs (4x10⁴ cells) were plated onto the collagen-coated coverslips and grown to confluence, typically for 5 to 7 days. To label F-actin, cells were fixed for 10 minutes with 3.7% paraformaldehyde in PBS containing 1 mmol/L Mg²⁺ and 0.1 mmol/L Ca²⁺, permeabilized in cold acetone (-20°C) for 5 minutes, and air dried. Coverslips were washed three times, incubated for 20 minutes with rhodamine phalloidin (Molecular Probes, 1:20 in PBS), and mounted in Mowiol (Calbiochem) containing 0.2% p-phenylenediamine (Sigma) as antifading agent. The number of microspikes per cell and the percentage of cells showing membrane ruffles were determined by evaluating at least 500 cells at the wound edge, using a 40x objective (total magnification x400). Actin fibers were quantified by using a 40x objective (total magnification x400) and counting clearly discernible fibers that had a length of at least 50% of the diameter of the cell in one microscopic plane.

For vinculin staining, cells were fixed for 10 minutes with 3.7% paraformaldehyde in PBS containing 1 mmol/L Mg²⁺ and 0.1 mmol/L Ca²⁺, permeabilized in extraction buffer containing 60 mmol/L PIPES, 25 mmol/L HEPES (pH 6.1), 10 mmol/L EGTA, 3 mmol/L MgCl₂, and 0.1% Triton for 5 minutes, blocked for 20 minutes with 10% normal goat serum, reacted with mouse monoclonal anti-vinculin hVIN-1 (Sigma, 1:2000) for 1 hour, and finally incubated with FITC-labeled goat anti-mouse IgG (1:200, Sigma). All steps were performed at room temperature with three washes in PBS/2% BSA between antibody incubations. Fluorescence microscopy was performed with a Zeiss Axiophot and micrographs were recorded on Kodak T-Max 400 panchromatic films.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Bacterial Toxins Block Endothelial Wound Repair In Vitro
We tested the ADP ribosyltransferases from Clostridium botulinum (C3-transferase) and Staphylococcus aureus (EDIN) as well as the monoglucosyltransferase from Clostridium difficile (toxin A) in an in vitro model of EC responses to mechanical wounding. Confluent and quiescent monolayers of HUVECs cultured on collagen-coated surfaces were wounded by removing cells on a path 1 mm wide with a pipette tip. This procedure leads to wound repair that is dependent on EC migration only, because proliferation of ECs in response to wounding does not start before 36 hours and occurs mainly in cells distal to the wound.^{4 5 8} Under our conditions, cells continuously migrated into the cell-free area from both wound edges until wound closure was reached at 36 hours (Fig 1A and 1B). When C3-transferase (3 µg/mL) was present 24 hours before wounding and during wound repair, wound closure could not be reached (Fig 1C and 1D). Likewise, addition of S. aureus EDIN (10 µg/mL, 24 hours pretreatment) or C. difficile toxin A (15 ng/mL, 14 hours pretreatment) blocked wound repair with a similar efficiency as C3-transferase (data not shown). Quantification and dose dependency of the C3-transferase effect revealed that within 36 hours, 481±100 cells migrated into 1 mm² of wounded area. This value was reduced to 330±46, 293±36, and 248±16 cells in the presence of 1, 3, and 10 µg/mL C3-transferase, respectively (Fig 2A). Time courses revealed that independent of the presence of C3-transferase, 50% of cells entered the wounded area in the first 12 hours and 50% in the following 24 hours (Fig 2A).

View larger version (57K): [in this window] [in a new window]   Figure 1. Effect of C3-transferase on EC wound repair in vitro. Confluent monolayers of HUVECs cultured on collagen were not treated (A and B) or were pretreated with C3-transferase from C. botulinum (3 µg/mL, 24 hours, C and D) and wounded with a pipette tip. Photographs were taken under phase contrast at 12 hours (A and C) or 36 hours (B and D). Bar=100 µm.

View larger version (13K): [in this window] [in a new window]   Figure 2. Effect of toxins on cell migration and in situ ADP ribosylation of Rho. A, Confluent monolayers of HUVECs cultured on collagen were not treated (control) or were pretreated with indicated concentrations of C3-transferase for 24 hours and wounded with a pipette tip. At indicated times after wounding, the number of cells that had migrated into 1 mm2 of wounded area was determined. Each value represents the mean of three different experiments performed in triplicate. SDs were between 10% and 20% of values. B, Confluent HUVECs were cultured for 24+36 hours in the presence of indicated concentrations of C3-transferase (C3), lysed, and subjected to in vitro [32P]ADP ribosylation. The degree of in situ ADP ribosylation can be estimated as the difference between band intensities of control and C3-pretreated cells. An experiment representative of three similar ones is shown. C, Confluent HUVECs were not treated (CTRL) or were treated with C3-transferase (C3; 3 µg/mL, 24 hours), EDIN (10 µg/mL, 24 hours), or toxin A (ToxA; 30 ng/mL, 12 hours), lysed, and subjected to in vitro [32P]ADP ribosylation. A representative experiment is shown.

To demonstrate unambiguously that intracellularly acting C3-transferase blocks EC migration, we microinjected C3 into cells at the wound edge and the adjacent two to three cell rows behind the wound edge. Within 12 hours, 221±32 control cells and 118±20 C3-injected cells migrated into 1 mm² of area (n=2; mean±SD). Thus, at 12 hours, the degree of inhibition of migration obtained with microinjection (50%) was very similar to what was seen with external addition of C3. Microinjection was less effective at longer times after wounding, possibly due to shorter half-life of C3 in the cytosol.

These data indicate that different bacterial toxins known to modify and inactivate Rho GTP-binding proteins are effective inhibitors of EC migration and wound repair in vitro.

Identification of Toxin Targets in ECs
To identify toxin target proteins in HUVECs, we incubated cell lysates with [³²P]NAD⁺ and C3-transferase. SDS/PAGE and autoradiography revealed one [³²P]ADP-ribosylated band that migrated at 23 kD (Fig 2B). In lysates of cells pretreated with increasing concentrations of C3-transferase, incorporation of [³²P]ADP ribose into the Rho substrate was dose dependently reduced, indicating ADP ribosylation in situ (Fig 2B). We conclude that inhibition of cell migration correlates with in situ ADP ribosylation of Rho.

Pretreatment of HUVECs with EDIN (10 µg/mL, 24 hours) or toxin A (30 ng/mL, 12 hours) reduced [³²P]ADP ribosylation of the Rho substrate to a similar degree as C3-transferase (3 µg/mL, 24 hours; Fig 2C). Together, these data suggest that different bacterial toxins are able to ADP ribosylate (C3, EDIN) or glucosylate (toxin A) Rho in intact HUVECs.

In 2D gels of the in vitro [³²P]ADP-ribosylation substrates, three [³²P]ADP-ribosylated protein spots of different intensities and pIs, designated A, B, and C were detected (Fig 3). The intensities of all three spots were diminished in lysates from C3- or toxin A–pretreated cells (data not shown), indicating that they all were subject to in situ modification. The most abundant spot, A, consisted of one major and one minor spot with pIs of 6.3 and 6.2, respectively. These protein spots most likely represent RhoA because they (1) have the calculated pI of RhoA (6.2), (2) comigrate with RhoA present in platelets and monocytic cells,²⁰ and (3) react with specific monoclonal and polyclonal antibodies to RhoA in Western blots (Fig 3). Spot B, with a pI of 5.2, likely represents RhoB because (1) its pI is similar to the calculated pI of RhoB (5.3) and (2) it reacted with a specific polyclonal antibody to RhoB (Fig 3). The least abundant spot, C, likely represents RhoC because its pI of 6.7 resembles the calculated pI of RhoC (6.8). So far we were unable to detect Rac1 or Rac2 in HUVECs by Western blot (Aepfelbacher et al, unpublished data, 1997).

View larger version (62K): [in this window] [in a new window]   Figure 3. Two-dimensional gel analysis of C3 [32P]ADP-ribosylation substrates. Lysates of HUVECs were subjected to in vitro [32P]ADP ribosylation, run on 2D gel, and analyzed by autoradiography ([32P]ADP) or Western blot using anti-RhoA or anti-RhoB antibodies. Three [32P]ADP-ribosylated protein spots of different intensities and pIs, designated A (pI 6.2 and 6.3), B (pI 5.2), and C (pI 6.7) were detected. The [32P]ADP-ribosylated spot A comigrated with RhoA and the [32P]ADP-ribosylated spot B comigrated with RhoB in Western blot. A representative experiment is shown.

To further confirm the presence of RhoA, RhoB, and RhoC in HUVECs, we analyzed mRNA expression. Oligonucleotide probes specific for rhoA, rhoB, or rhoC genes hybridized to distinct bands in Northern blot analysis (Fig 4). The rhoC signal in the Northern blot was considerably weaker than the rhoA and rhoB signals. This finding is consistent with the low abundance of spot C (which likely represents the RhoC protein) in the 2D gel.

View larger version (76K): [in this window] [in a new window]   Figure 4. HUVEC expression of mRNA for RhoA, RhoB, and RhoC. Northern blots using oligonucleotide probes specific for RhoA, RhoB, or RhoC were performed with total HUVEC RNA (see "Methods"). The oligonucleotide probes hybridized to distinct, single bands. An experiment representative of two similar ones is shown.

Taken together, these data indicate that RhoA, RhoB, and most likely RhoC are in vivo substrates of C3-transferase, EDIN, and toxin A in HUVECs. Due to its availability as recombinant protein, most remaining experiments were performed with C3-transferase.

C3-Transferase Selectively Affects Actin Stress Fibers and Focal Adhesion Sites in Migrating ECs
Rho family GTPases have been implicated in the formation of distinct filamentous actin structures,²² namely microspikes/filopodia (CDC42Hs), membrane ruffles/lamellipodia (Rac) and stress fibers (Rho). To test whether such structures are expressed in migrating HUVECs, we stained cells with rhodamine-phalloidin before and at various intervals after wounding. As demonstrated in Fig 5, microspikes, membrane ruffles, and stress fibers were largely absent from untreated and C3-treated cells in the confluent monolayer (A and B). C3 slightly reduced actin at the cell-to-cell contacts and induced few gaps between adjacent cells, but in contrast to fibroblasts¹¹ and similar to monocytes²³ and neurons,²⁴ C3 caused flattening of most cells rather than rounding. Approximately 5% to 10% of cells in the confluent monolayer were elongated and contained stress fibers, even in the presence of C3-transferase. We think that these cells take up external C3 inefficiently, because microinjection of C3 into these particular cells was able to abolish stress fibers (not shown). Within 10 minutes after wounding, nearly all of the cells at the wound edge expressed microspikes that pointed into the wounded area (Fig 5C). Approximately seven microspikes per cell were expressed 10 minutes after wounding; thereafter, this number decreased to approximately one at 24 hours. Microspikes were also found in cells pretreated with C3-transferase (Fig 5D). Quantification of microspikes in control and C3-treated cells at 10 minutes after wounding revealed 6.5±1.1 and 6.8±2 microspikes per cell, respectively (mean±SD, n=3, 500 cells counted per experiment). The membrane ruffles that developed in cells at the wound edge were oriented toward the direction of migration (Fig 5E). The appearance of membrane ruffles occurred in two phases: between 10 minutes and 4 hours after wounding, 20% of cells showed ruffles. This percentage increased to 70% between 4 and 12 hours. Ruffles were also found in C3-treated cells (Fig 5F), and quantification at 6 hours after wounding demonstrated that the percentage of cells showing ruffles was 33±11 in control and 31±8 in C3-treated cells (mean±SD, n=3, 500 cells counted per experiment). Finally, the number of actin filaments per cell increased dramatically in HUVECs at the wound edge (compare Fig 5G with 5A). Stress fibers were hardly detectable until 2 hours after wounding, and thereafter their number increased to a maximum after 12 to 24 hours. C3-Transferase almost completely abolished the formation of actin fibers (compare Fig 5H with 5G). Similar to C3-transferase, EDIN and toxin A were able to prevent stress fiber formation (data not shown). Microinjection of C3 produced essentially identical results as external addition of C3 (data not shown).

View larger version (151K): [in this window] [in a new window]   Figure 5. Effect of C3 on actin and vinculin structures in migrating HUVECs. Control (A, C, E, G, I, K, and M) or C3-treated (B, D, F, H, J, L, and N) HUVECs on collagen-coated glass coverslips were stained for actin (A through H) or vinculin (I through N). Actin staining was performed before wounding (A and B), 15 minutes after wounding to visualize microspikes (C and D), 4 hours after wounding to visualize ruffles (E and F), or 12 hours after wounding for fibers (G and H). Vinculin staining was performed before wounding (I and J), 15 minutes after wounding to visualize focal complexes in microspikes (K and L), or 12 hours after wounding to visualize focal adhesions (M) or focal complexes at the cell periphery (N). Cells at the wound edge are shown. The wounded area is at the right, except for K and L, in which it is at the bottom. Representative photographs are shown. Photographs C and D were overexposed to visualize microspikes optimally. The strong nuclear signal in N is due to spurious unspecific binding of the anti-vinculin antibody to the nucleus. Bar=10 µm.

We next investigated the effect of toxins on vinculin reorganization during wound repair. By immunostaining with anti-vinculin antibodies, a few tear-shaped focal adhesions were detected at the ventral surface of cells in the intact or C3-treated monolayer (Fig 5I and 5J). Vinculin staining was mainly concentrated at the cell boundaries in cell-to-cell contacts. On wounding of the monolayer, cells at the wound edge showed a rapid (within 10 minutes) redistribution of vinculin into microspikes (Fig 5K). C3-Transferase did not prevent relocation of vinculin to microspikes (Fig 5L). The formation of vinculin-containing focal adhesions paralleled the generation of stress fibers. They were first seen after 2 to 4 hours and were maximally expressed at 12 hours (Fig 5M). C3-Transferase completely inhibited the formation of the tear-shaped focal adhesions (Fig 5N), which by double immunostaining were found to be mainly located at the tip of stress fibers. Interestingly, however, C3-transferase did not abolish vinculin accumulation in punctate-like focal complexes at the cell periphery, indicating that these were assembled independent of Rho.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
This study demonstrates that C3-transferase from C. botulinum, EDIN from S. aureus, and toxin A from C. difficile, all known to inactivate Rho GTP-binding proteins in situ, prevent migration and wound repair of human ECs. C3-Transferase specifically ADP ribosylates and inactivates RhoA, RhoB, and RhoC¹⁴ and thus greatly facilitated the elucidation of Rho function in a variety of cells. Because extracellular concentrations of up to 30 µg/mL are necessary for effective uptake into the cytosol, it seems unlikely that the C3-transferase in this form is of pathophysiological importance. EDIN, originally described as epidermal cell differentiation inhibitor,¹⁸ is a C3-like ADP ribosyltransferase that was shown to be effective at concentrations of 10 to 100 ng/mL in keratinocytes.²⁵ However, in our study, a concentration of at least 3 µg/mL was needed to efficiently inhibit EC wound repair. C. difficile toxin A, considered a causative agent for pseudomembranous colitis,¹⁵ blocked EC wound repair at a concentration of 5 to 15 ng/mL. These concentrations were chosen because they caused a minimum of rounded or arborized ECs (20%), allowing for a valid and quantitative evaluation of actin structures.

The 2D gel electrophoresis of the [³²P]ADP-ribosylation substrates and the Northern blots suggested that RhoA, RhoB, and RhoC are expressed in HUVECs and that RhoA is by far the most abundant Rho isoform. The reduced in vitro [³²P]ADP ribosylation in lysates of C3- or toxin A–pretreated cells indicated that RhoA, RhoB, and RhoC can all be ADP ribosylated by C3-transferase or glucosylated by toxin A, respectively, in intact HUVECs. The functions of Rho proteins in cytoskeletal organization have mostly been attributed to RhoA, because it was identified as the sole ADP-ribosylation substrate in fibroblasts,²⁶ human blood platelets,²⁷ lymphocytes,²⁸ neuronal cells,²⁴ and monocytic cell lines.^{20 29} Because RhoA and RhoC are 98% homologous on the level of amino acid sequence, they most likely are redundant with respect to function. In contrast, RhoB is only 88% homologous to RhoA and RhoC and is inducible by growth factors³⁰ and DNA-damaging agents in quiescent fibroblasts.³¹ Whether RhoB is also involved in actin organization is controversial.^{31 32} Hence, the toxin effects described herein can most likely be attributed to inhibition of RhoA.

HUVECs at the wound edge expressed microspikes, lamellipodia/ruffles, and actin fibers, all of which have been implicated in cell migration.¹⁰ In quiescent fibroblasts stimulated with bradykinin or microinjected with activated CDC42Hs, a hierarchical cascade exists whereby CDC42Hs induces filopodia and activates Rac, Rac in turn produces lamellipodia and activates Rho, and Rho finally stimulates formation of actin fibers.²² This sequence of events takes place within 30 minutes after stimulation of cells. In contrast, HUVECs at the wound edge expressed microspikes and some membrane ruffles (in 20% of cells) within 30 minutes after wounding, but formation of actin fibers and the majority of membrane ruffling occurred between 2 and 24 hours.

Actin structures in migrating HUVECs were closely associated with vinculin-containing focal complexes. We found that C3 inhibited neither the rapid vinculin relocation to microspikes nor the formation of punctate-like focal complexes at the cell periphery, which is consistent with its lacking effect on microspike or ruffle formation. Consequently, C3 completely prevented the formation of focal adhesions, which are located at the end of stress fibers. Hence, the wound repair response of ECs seems to involve the temporally and spatially controlled formation of at least three distinct combinations of actin/vinculin structures that most probably are controlled by individual Rho family GTPases (summarized in Fig 6). In the early phase (10 minutes to 2 hours after wounding), cells express actin microspikes and ruffles, potentially through activation of CDC42Hs and Rac, respectively. In the late phase (2 to 24 hours after wounding), Rho induces formation of actin filaments/focal adhesions. Further experiments should analyze the role of Rac and CDC42Hs in wound repair and determine the hierarchy and temporal and spatial coordination of Rho, Rac, and CDC42Hs activities in migrating ECs.

View larger version (18K): [in this window] [in a new window]   Figure 6. Schematic diagram of the actin structures found in the early (protrusive) and late (migratory) phases of endothelial wound repair.

In summary, EC migration is crucial for the repair of injured blood vessels,¹ angiogenesis,³ and atherogenesis³³ and depends on assembly of specific actin structures and focal adhesions (see References 8 and 34^{8 34} and the present study). Our data suggest that Rho family GTP-binding proteins control the complex cytoskeletal reorganizations in migrating ECs and should promote further investigations about the influence of Rho-inhibiting bacterial toxins on tumor angiogenesis, inflammation, and atherogenesis.

	Selected Abbreviations and Acronyms

 

2D = two-dimensional EC = endothelial cell HUVEC = human umbilical vein EC PAGE = polyacrylamide gel electrophoresis

	Acknowledgments

 
This work was supported in part by a grant from the Deutsche Forschungsgemeinschaft (Ae11/2-1), August Lenz Stiftung und Wilhelm Sander-Stiftung. The pGEX2T-C3 exoenzyme expression vector was a generous gift from Dr L. Feig, Tufts University, Boston, Mass. C. difficile toxin A was provided by Drs F. Hofmann and I. Just, Universität Freiburg, Freiburg, Germany. We thank Barbara Böhlig for expert technical assistance, Manfred Schliwa for help with immunofluorescence microscopy, and Wolfgang Siess for discussions.

Received October 2, 1996; accepted January 16, 1997.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Schwartz SM, Haudenschild CC, Eddy EM. Endothelial regeneration, I: quantitative analysis of initial stages of endothelial regeneration in rat aortic intima. Lab Invest. 1978;38:568-580.[Medline] [Order article via Infotrieve]
2. Haudenschild CC, Schwartz SM. Endothelial regeneration, II: restitution of endothelial continuity. Lab Invest. 1979;41:407-418.[Medline] [Order article via Infotrieve]
3. Folkman J, Shing Y. Angiogenesis. J Biol Chem. 1992;267:10931-10934.[Free Full Text]
4. Coomber BL, Gotlieb AI. In vitro endothelial wound repair: interaction of cell migration and proliferation. Arteriosclerosis. 1990;10:215-222.[Abstract/Free Full Text]
5. Sholley MM, Gimbrone MA, Cotran RS. Cellular migration and replication in endothelial regeneration: a study using irradiated endothelial cultures. Lab Invest. 1977;36:18-25.[Medline] [Order article via Infotrieve]
6. Selden SC, Schwartz SM. Cytochalasin B inhibition of endothelial proliferation at wound edges in vitro. J Cell Biol. 1979;81:348-354.[Abstract/Free Full Text]
7. Wong MKK, Gotlieb AI. In vitro reendothelialization of a single-cell wound: role of microfilament bundles in rapid lamellipodia-mediated wound closure. Lab Invest. 1984;51:75-81.[Medline] [Order article via Infotrieve]
8. Romer LH, McLean N, Turner CE, Burridge K. Tyrosine kinase activity, cytoskeletal organization, and motility in human vascular endothelial cells. Mol Biol Cell. 1994;5:349-361.[Abstract]
9. Burridge K, Fath K, Kelly T, Nuckolls G, Turner C. Focal adhesions: transmembrane junctions between the extracellular matrix and the cytoskeleton. Annu Rev Cell Biol. 1988;4:487-525.
10. Huttenlocher A, Sandborg RR, Horwitz AF. Adhesion in cell migration. Curr Opin Cell Biol. 1995;7:697-706.[Medline] [Order article via Infotrieve]
11. Hall A. Small GTP-binding proteins and the regulation of the actin cytoskeleton. Annu Rev Cell Biol. 1994;10:31-54.
12. Ridley AJ. Rho: theme and variations. Curr Biol. 1996;6:1256-1264.[Medline] [Order article via Infotrieve]
13. Boguski MS, McCormick F. Proteins regulating ras and its relatives. Nature. 1993;366:643-654.[Medline] [Order article via Infotrieve]
14. Aktories K, Just I. In vitro ADP-ribosylation of Rho by bacterial ADP-ribosyltransferases. Methods Enzymol.1995;256:184-195.
15. Aktories K, Just I. Monoglucosylation of low-molecular mass GTP-binding Rho proteins by clostridial cytotoxins. Trends Cell Biol. 1995;5:441-443.[Medline] [Order article via Infotrieve]
16. Ridley AJ, Hall A. The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell. 1992;70:389-399.[Medline] [Order article via Infotrieve]
17. Smith DB, Johnson KS. Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione S-transferase. Gene. 1988;67:31-40.[Medline] [Order article via Infotrieve]
18. Sugai M, Enomoto T, Hashimoto K, Matsumoto K, Matsuo Y, Ohgai H, Hong YM, Inoue S, Yoshikawa K, Suginawa H. A novel epidermal cell differentiation inhibitor (EDIN): purification and characterization from Staphylococcus aureus. Biochem Biophys Res Commun. 1990;173:92-98.[Medline] [Order article via Infotrieve]
19. O'Farrell PH. High resolution two-dimensional electrophoresis of proteins. J Biol Chem. 1975;250:4007-4012.[Abstract/Free Full Text]
20. Aepfelbacher M, Essler M, Luber De Quintana K, Weber PC. ADP-ribosylation of the GTP-binding protein RhoA blocks cytoplasmic division in human myelomonocytic cells. Biochem J. 1995;308:853-858.
21. Dillon ST, Feig LA. Purification and assay of recombinant C3-transferase. Methods Enzymol. 1995;256:174-184.[Medline] [Order article via Infotrieve]
22. Nobes CD, Hall A. Rho, Rac, and Cdc42 GTPase regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. Cell. 1995;81:53-62.[Medline] [Order article via Infotrieve]
23. Aepfelbacher M, Essler M, Huber E, Czech A, Weber PC. Rho is a negative regulator of human monocyte spreading. J Immunol. 1996;157:5070-5075.[Abstract]
24. Jalink K, van Corven EJ, Hengeveld T, Morii N, Narumiya S, Moolenaar WH. Inhibition of lysophosphatidate- and thrombin-induced neurite retraction and neuronal cell rounding by ADP ribosylation of the small GTP-binding protein Rho. J Cell Biol. 1994;126:801-810.[Abstract/Free Full Text]
25. Sugai M, Hashimoto K, Kikuchi A, Inoue S, Okumura H, Matsumoto K, Goto Y, Ohgai H, Moriishi K, Syuto B, Yoshikawa K, Suginaka H, Takai Y. Epidermal cell differentiation inhibitor ADP-ribosylates small GTP-binding proteins and induces hyperplasia of epidermis. J Biol Chem. 1992;267:2600-2604.[Abstract/Free Full Text]
26. Yamamoto M, Marui N, Sakai T, Morii N, Kozaki S, Ikai K, Imamura S, Narumiya S. ADP-ribosylation of the rhoA gene product by botulinum C3 exoenzyme causes Swiss 3T3 cells to accumulate in the G₁ phase of the cell cycle. Oncogene. 1993;8:1449-1455.[Medline] [Order article via Infotrieve]
27. Nemoto Y, Namba T, Teru-uchi T, Ushikubi F, Morii N, Narumiya S. A rho gene product in human blood platelets. J Biol Chem. 1992;267:20916-20920.[Abstract/Free Full Text]
28. Lang P, Guizani L, Vitté-Mony I, Stancou R, Dorseuil O, Gacon G, Bertoglio J. ADP-ribosylation of the ras-related, GTP-binding protein rhoA inhibits lymphocyte-mediated cytotoxicity. J Biol Chem. 1992;267:11677-11680.[Abstract/Free Full Text]
29. Aepfelbacher M. ADP-ribosylation of Rho enhances adhesion of U937 cells to fibronectin via the 5ß1 integrin receptor. FEBS Lett. 1995;363:78-80.[Medline] [Order article via Infotrieve]
30. Jähner D, Hunter T. The ras-related gene rhoB is an immediate-early gene inducible by v-fps, epidermal growth factor, and platelet-derived growth factor in rat fibroblasts. Mol Cell Biol. 1991;11:3682-3690.[Abstract/Free Full Text]
31. Fritz G, Kaina B, Aktories K. The ras-related small GTP-binding protein rhoB is immediate-early inducible by DNA-damaging treatments. J Biol Chem. 1995;270:25172-25177.[Abstract/Free Full Text]
32. Adamson P, Paterson H, Hall A. Intracellular localization of the p21rho proteins. J Cell Biol. 1992;119:617-627.[Abstract/Free Full Text]
33. Ross R, Glomset JA. The pathogenesis of atherosclerosis. N Engl J Med. 1976;295:369-377.[Medline] [Order article via Infotrieve]
34. Lampugnani MG, Corada M, Caveda L, Breviario F, Ayalon O, Geiger B, Dejana E. The molecular organization of endothelial cell to cell junctions: differential association of plakoglobin, ß-catenin, and -catenin with vascular endothelial cadherin (VE-cadherin). J Cell Biol. 1995;129:203-217.[Abstract/Free Full Text]

This article has been cited by other articles:

				S. Y. Schubert, A. Benarroch, J. Ostvang, and E. R. Edelman Regulation of Endothelial Cell Proliferation by Primary Monocytes Arterioscler. Thromb. Vasc. Biol., January 1, 2008; 28(1): 97 - 104. [Abstract] [Full Text] [PDF]

				P. A. Singleton, R. Salgia, L. Moreno-Vinasco, J. Moitra, S. Sammani, T. Mirzapoiazova, and J. G. N. Garcia CD44 Regulates Hepatocyte Growth Factor-mediated Vascular Integrity: ROLE OF c-Met, Tiam1/Rac1, DYNAMIN 2, AND CORTACTIN J. Biol. Chem., October 19, 2007; 282(42): 30643 - 30657. [Abstract] [Full Text] [PDF]

				Hardev Ramandeep Singh Girn, S. Ahilathirunayagam, A. I. D. Mavor, and S. Homer-Vanniasinkam Reperfusion Syndrome: Cellular Mechanisms of Microvascular Dysfunction and Potential Therapeutic Strategies Vascular and Endovascular Surgery, September 1, 2007; 41(4): 277 - 293. [Abstract] [PDF]

				K. Aase, M. Ernkvist, L. Ebarasi, L. Jakobsson, A. Majumdar, C. Yi, O. Birot, Y. Ming, A. Kvanta, D. Edholm, et al. Angiomotin regulates endothelial cell migration during embryonic angiogenesis Genes & Dev., August 15, 2007; 21(16): 2055 - 2068. [Abstract] [Full Text] [PDF]

				M. Oktem, I. Esinler, D. Eroglu, N. Haberal, N. Bayraktar, and H. B. Zeyneloglu High-dose atorvastatin causes regression of endometriotic implants: a rat model Hum. Reprod., May 1, 2007; 22(5): 1474 - 1480. [Abstract] [Full Text] [PDF]

				A. Gojova and A. I. Barakat Vascular endothelial wound closure under shear stress: role of membrane fluidity and flow-sensitive ion channels J Appl Physiol, June 1, 2005; 98(6): 2355 - 2362. [Abstract] [Full Text] [PDF]

				S. Chifflet, J. A. Hernandez, and S. Grasso A possible role for membrane depolarization in epithelial wound healing Am J Physiol Cell Physiol, June 1, 2005; 288(6): C1420 - C1430. [Abstract] [Full Text] [PDF]

				D. E. Voth and J. D. Ballard Clostridium difficile Toxins: Mechanism of Action and Role in Disease Clin. Microbiol. Rev., April 1, 2005; 18(2): 247 - 263. [Abstract] [Full Text] [PDF]

				H. R. Evans, J. M. Sutton, D. E. Holloway, J. Ayriss, C. C. Shone, and K. R. Acharya The Crystal Structure of C3stau2 from Staphylococcus aureus and Its Complex with NAD J. Biol. Chem., November 14, 2003; 278(46): 45924 - 45930. [Abstract] [Full Text] [PDF]

				I. Adini, I. Rabinovitz, J. F. Sun, G. C. Prendergast, and L. E. Benjamin RhoB controls Akt trafficking and stage-specific survival of endothelial cells during vascular development Genes & Dev., November 1, 2003; 17(21): 2721 - 2732. [Abstract] [Full Text] [PDF]

				D. GRATZINGER, S. CANOSA, B. ENGELHARDT, and J. A. MADRI Platelet endothelial cell adhesion molecule-1 modulates endothelial cell motility through the small G-protein Rho FASEB J, August 1, 2003; 17(11): 1458 - 1469. [Abstract] [Full Text] [PDF]

				G. P. van Nieuw Amerongen, P. Koolwijk, A. Versteilen, and V. W.M. van Hinsbergh Involvement of RhoA/Rho Kinase Signaling in VEGF-Induced Endothelial Cell Migration and Angiogenesis In Vitro Arterioscler. Thromb. Vasc. Biol., February 1, 2003; 23(2): 211 - 217. [Abstract] [Full Text] [PDF]

				M. Weis, C. Heeschen, A. J. Glassford, and J. P. Cooke Statins Have Biphasic Effects on Angiogenesis Circulation, February 12, 2002; 105(6): 739 - 745. [Abstract] [Full Text] [PDF]

				T. Yamaguchi, T. Hayashi, H. Takami, M. Ohnishi, T. Murata, K. Nakayama, K. Asakawa, M. Ohara, H. Komatsuzawa, and M. Sugai Complete Nucleotide Sequence of a Staphylococcus aureus Exfoliative Toxin B Plasmid and Identification of a Novel ADP-Ribosyltransferase, EDIN-C Infect. Immun., December 1, 2001; 69(12): 7760 - 7771. [Abstract] [Full Text] [PDF]

				A. D. Verin, A. Birukova, P. Wang, F. Liu, P. Becker, K. Birukov, and J. G. N. Garcia Microtubule disassembly increases endothelial cell barrier dysfunction: role of MLC phosphorylation Am J Physiol Lung Cell Mol Physiol, September 1, 2001; 281(3): L565 - L574. [Abstract] [Full Text] [PDF]

				G. P. van Nieuw Amerongen and V. W.M. van Hinsbergh Cytoskeletal Effects of Rho-Like Small Guanine Nucleotide-Binding Proteins in the Vascular System Arterioscler. Thromb. Vasc. Biol., March 1, 2001; 21(3): 300 - 311. [Abstract] [Full Text] [PDF]

				Y Yamada, M Sugai, M Woo, N Nishida, and T Sugimoto Acquired subglottic stenosis caused by methicillin resistant Staphylococcus aureus that produce epidermal cell differentiation inhibitor Arch. Dis. Child. Fetal Neonatal Ed., January 1, 2001; 84(1): 38F - 39. [Abstract] [Full Text]

				J. M. Carbajal, M. L. Gratrix, C.-H. Yu, and R. C. Schaeffer Jr. ROCK mediates thrombin's endothelial barrier dysfunction Am J Physiol Cell Physiol, July 1, 2000; 279(1): C195 - C204. [Abstract] [Full Text] [PDF]

				L. Moldovan, N. I. Moldovan, R. H. Sohn, S. A. Parikh, and P. J. Goldschmidt-Clermont Redox Changes of Cultured Endothelial Cells and Actin Dynamics Circ. Res., March 17, 2000; 86(5): 549 - 557. [Abstract] [Full Text] [PDF]

H. Lee, E. J. Goetzl, and S. An Lysophosphatidic acid and sphingosine 1-phosphate stimulate endothelial cell wound healing Am J Physiol Cell Physiol, March 1, 2000; 278(3): C612 - C618. [Abstract]