UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
Filling the gaps
The endothelium in regulating vascular leakage and leukocyte extravasation
Schimmel, L.
Publication date
2018
Document Version
Other version
License
Other
Link to publication
Citation for published version (APA):
Schimmel, L. (2018). Filling the gaps: The endothelium in regulating vascular leakage and
leukocyte extravasation.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
1Department of Plasma Proteins, Molecular Cell Biology Lab, Sanquin Research and Landsteiner
Labo-ratory, Academic Medical Center, University of Amsterdam, Plesmanlaan 125, Amsterdam 1066 CX, The Netherlands. 2Department of Molecular Cytology, Swammerdam Institute of Life Sciences, University of
Amsterdam, Amsterdam, The Netherlands. 3Department of Medical Biochemistry, Amsterdam
Cardio-vascular Sciences, Academic Medical Center, Amsterdam, The Netherlands.
Manuscript in preparation
Lilian Schimmel
1, Janine Arts
1,2, Ivar
Noordstra
1, Niels Heemskerk
1, Nathalie
Reinhard
2, Vivian de Waard
3& Jaap D.
van Buul
1The Rho-GEFs FGD5 and
Tuba differentially regulate
the small GTPase Cdc42 to
control leukocyte
extravasation and vascular
permeability
Summary
The role of multiple small GTPases, including RhoA, Rac1 and Cdc42, in re-gulating endothelial cell (EC) permeability has been extensively investigated, but how ECs prevent vascular leakage during leukocyte transendothelial mi-gration (TEM) is still largely unknown. For leukocyte TEM, multiple GTPases are activated at different timepoints; this requires the tight spatiotemporal control by Guanine nucleotide Exchange Factors (GEFs) and GTPase-acti-vaing proteins (GAPs). However, which Rho-GEFs and -GAPs are involved in regulating basal endothelial permeability, leukocyte-induced vascular lea-kage or leukocyte TEM efficiency remains poorly understood. A short hairpin RNA screen targeting 23 distinct endothelial Rho-GEFs and –GAPs revealed seven regulators of basal endothelial permeability and six regulators of leu-kocyte-induced vascular leakage. Moreover, we elucidated two distinct roles for Cdc42 activation during leukocyte TEM. Firstly, FGD5-mediated activati-on of Cdc42 in ECs upactivati-on TNFα stimulatiactivati-on is required for Intercellular Adhe-sion Molecule-1 (ICAM-1)-rich filopodia formation that mediate PMN adhe-sion. Secondly, Cdc42 activation by the Rho-GEF Tuba appears to mediate the closure of the endothelial pore upon leukocyte diapedesis to prevent vascular leakage. In conclusion, we found involvement of Cdc42 activation by two different GEFs, to support leukocyte TEM in leukocyte adhesion and closing endothelial gap formation after leukocyte passage.
Introduction
During inflammation and immune surveillance, leukocytes cross the endot-helial cell (EC) barrier in order to reach the site of infection. Leukocyte ex-travasation and vascular permeability have been shown to be two separate processes 1–3. Several studies have demonstrated that elevated vascular
permeability precedes leukocyte influx, revealing that leukocyte adhesion and transmigration are not necessarily directly correlated with the evoked permeability change during acute inflammation 4–7. These studies underline
the importance of maintaining vascular permeability both under basal condi-tions and during leukocyte transendothelial migration (TEM). Maintenance of vascular permeability and regulation of leukocyte TEM rely on dynamic re-modeling of the endothelial actin cytoskeleton 8. This actin remodeling is
re-gulated by the Rho family of small GTPases which act as molecular switches that transduce receptor-mediated signaling towards the actin polymerization and remodeling machinery.
We recently showed that the endothelium limits vascular permeability during leukocyte TEM via generation of a F-actin-rich contractile endothelial pore that requires local RhoA activation 8. However, too much RhoA
activa-tion throughout the entire cell results in strong acto-myosin tension and EC contraction, leading to formation of endothelial gaps and increased vascular leakage 9. Interestingly, the two closely-related GTPases RhoB and RhoC
are dispensable for maintaining vascular integrity during leukocyte TEM 10.
Moreover, Rac1 activation is required to maintain stable endothelial cell-cell junctions to prevent basal vascular permeability and, in conjunction with RhoG, is involved in the formation of apical endothelial cup structures to capture leukocytes 11,12. A role for endothelial Cdc42 in leukocyte TEM has
recently been described in the formation of Intercellular Adhesion Molecule-1 (ICAM-1)-rich filopodia via Myosin-X activation 13.
Since multiple GTPases are involved in leukocyte TEM at different intracellular sites, GTPase activation at the right time and place is of gre-at importance. Therefore, specific spgre-atiotemporal regulgre-ation by Rho-Guani-ne nucleotide Exchange Factors (Rho-GEFs) and Rho GTPase Activating Proteins (Rho-GAPs) is essential for proper execution of GTPase-mediated signaling. However, to what extent different Rho-GEFs and Rho-GAPs con-trol GTPase cycling during the maintenance of basal vascular permeability as well as during leukocyte extravasation is not understood. We performed an unbiased short hairpin (sh)RNA screen to examine the involvement of Rho-GAPs and -GEFs in regulating basal EC permeability, vascular leak-age during leukocyte TEM, and leukocyte TEM efficiency. We identified two novel Cdc42-specific endothelial Rho-GEFs, FGD5 and Tuba, to have spe-cific roles during leukocyte TEM. Using flow chambers, a shRNA approach
and high resolution microscopy, we show that FGD5-induced Cdc42 activa-tion regulates TNFα-induced ICAM-1-rich filopodia formaactiva-tion and leukocyte adhesion, whereas Tuba is probably more involved in Cdc42-induced EC gap-closure upon completion of the diapedesis step.
Results
Screen for regulators of vascular permeability and leukocyte extrava-sation
To screen for endothelial Rho-GEFs and -GAPs that are involved in regula-ting leukocyte diapedesis and/or vascular permeability we used simultaneous measurement of calcein red-orange labeled polymorphonuclear leukocytes (PMNs) and 70 kDa FITC-dextran across TNFα-stimulated ECs towards a chemotactic Complement component 5a (C5a) gradient (Figure 1A). Mono-layers of ECs, that were transduced with the enlisted shRNAs (Table 1-3) targeting the different GEFs and GAPs, were cultured on fibronectin-coated Fluoroblock Transwell inserts, allowing detection of only the lower compart-ment with a photo spectrometer. To test whether PMN transmigration kinetics could be detected with the setup, calcein red-orange labeled PMNs were added to each Transwell insert and migration towards the C5a chemokine in the lower compartment was measured each minute for 1 hour. Transmigra-ted PMNs were detecTransmigra-ted 5 minutes after addition and reached a maximum within 35 minutes in control transduced monolayers (Figure 1B).
To investigate EC permeability during leukocyte diapedesis, simulta-neous measurement of FITC-dextran leakage into the lower compartment of the Transwell is measured during PMN transmigration across TNFα-stimula-ted ECs towards C5a. There was minimal FITC-dextran leakage during PMN transmigration across TNF-α-treated control ECs (Figure 1C). This novel screening assay allows to discriminate between changes in basal endothelial permeability, leukocyte-induced endothelial permeability and leukocyte TEM.
TRIO, VAV2, β-PIX, NME1, PLEKH1, CdGAP, and ARAP3 regulate basal endothelial permeability
Using our novel screening assay, we identified five Rho-GEFs and three Rho-GAPs to be involved in the regulation of basal endothelial barrier func-tion. Previous work demonstrated that TRIO-depleted-ECs display unstable VE-cadherin cell-cell junctions, resulting in increased basal endothelial per-meability 11. In agreement with this study, TRIO-deficient ECs showed
impai-red endothelial barrier function, already under basal conditions and conse-quently increased diffusion of FITC-dextran into the lower compartment of the Transwell in the absence or presence of PMN migration, when compa-red to control (Figure 1D,E). Knockdown of Rho-GEFs VAV2, β-PIX, NME1 and PLEKHG1 and the Rho-GAPs CdGAP and ARAP3 revealed a similar FITC-dextran leakage profile as TRIO knockdown, so these GEFs and GAPs are good candidates for the regulation of basal endothelial barrier function (Figure 1E and Table 1).
0 10 20 30 40 50 60 0 2 4 6 8 10 control + PMNshLARG + PMN 0.106 0.1359 shLARG shLARG + PMN shCTRL shCTRL + PMN 0.0 0.5 1.0 1.5 2.5 5.0 4.5 4.0 2.0 0.0 0.5 1.0 1.5 2.5 2.0 HUVEC HUVEC + PMN HUVEC HUVEC + PMN HUVEC HUVEC + PMN *** ** * * * * *** ** * ** Figure 1 Upper compartment Lower compartment Imaging range +C5a FITC 70 kDa-DextranCalcein-redPMNs Transwell membrane ECs 0 10 20 0.104 30 40 50 60 0.0 2.5 5.0 7.5 10.0 12.5 15.0
shTRIO shCdGAPshARAP3
shPLEKHG1 shNME1 shβ-PIX shVAV2 shCTRL shARHGAP11a shBCR shTuba shFGD5 shEct2 shLARG shCTRL Time (min) Time (min) shCTRL + PMN Transmigrated PMNs (a.u.) Transmigrated PMNs (a.u.) 0 10 20 30 40 50 60 Time (min) 0.5 1.0 1.5 2.0 2.5 shCTRL shCTRL + PMN A B C
Dextran leakage (a.u.)
FITC-Dextran 70kDa Leakage Fold Increase FITC-Dextran 70kDa Leakage Fold Increase
0.5 1.0 1.5 2.0 2.5
Dextran leakage (a.u.)
0.5 1.0 1.5 2.0 2.5
Dextran leakage (a.u.)
0 10 20 30 40 50 60 Time (min) 0 10 20 30 40 50 60 Time (min) shCTRL shTRIO H F G I E D shCTRL Tuba Actin shTuba-1 shTuba-2 + 10ng/ml TNFα 250 KDa 130 KDa 55 KDa 35 KDa FGD5 Actin shFGD5-1shFGD5-2 shCTRL + 10ng/ml TNFα 250 KDa 55 KDa 35 KDa
Figure 1. Basal and leukocyte induced endothelial leakage
(A) Schematic representation of a Transwell Fluoroblok insert in the setup for
si-multaneous measurement of FITC-dextran leakage and calcein red-orange labelled PMNs transmigration. ECs were cultured on 3µm pore permeable fibronectin-coa-ted Fluoroblok filters and treafibronectin-coa-ted overnight with TNFα. 70kDa FITC-dextran was added to the top compartment and 0.1 nM C5a was added in the bottom compart-ment. (B) Kinetics of calcein red-orange-labelled PMNs transmigration across the
ECs towards C5a. Data are presented as mean±SEM (dotted lines). (C) Kinetics
of FITC-dextran leakage through ECs under basal conditions (green line) and in presence of transmigrating PMNs (red line). Data are presented as mean±SEM (dot-ted lines). (D) FITC-dextran leakage through TRIO-depleted ECs (blue line). (E)
End point FITC-dextran leakage through ECs transduced with shRNAs resulting in increased basal dextran leakage upon silencing of TRIO, VAV2, β-PIX, NME1, PLEKHG1, CdGAP and ARAP3. (F) FITC-dextran leakage through
LARG, ECT2, FGD5, Tuba, Bcr and ARHGAP11a regulate leukocyte-in-duced vascular leakage
Next to regulating basal endothelial permeability, four Rho-GEFs and two Rho-GAPs were found to be involved in limiting endothelial leakage during PMN extravasation. Silencing of endothelial LARG did not alter basal en-dothelial barrier function or PMN TEM efficiency (Figure 1F-G). However, permeability gradually increased when PMN TEM occurred, indicating that LARG is involved in limiting leakage during TEM. This is in line with previous work, where we found that LARG in combination with Ect2 is involved in limiting vascular leakage during leukocyte TEM through local activation of RhoA 8. In addition to LARG and Ect2, we found the Rho-GEFs FGD5, Tuba and Bcr as potential regulators of leukocyte-induced EC permeability (Figure 1H). Among the Rho-GAPs we found ARHGAP11a and potentially the GAP domain of Bcr to be involved in the regulation of endothelial leakage during PMN diapedesis (Figure 1H and Table 2). Interestingly, the two Rho-GEFs FGD5 and Tuba, both specific for Cdc42, seem to be involved in regulating leukocyte-induced vascular leakage. Since the role for Cdc42 in leukocyte TEM and vascular permeability is largely unknown, additional experiments on the role for Cdc42 activation by FGD5 and Tuba during PMN TEM were performed.
FGD5 regulates Cdc42-induced ICAM-1 filopodia formation
Detailed analysis of the permeability and transmigration kinetics showed that silencing of endothelial FGD5 did not significantly increase basal endothelial permeability, while addition of transmigrating PMNs caused an induction of dextran leakage (Figure 2A). Besides the leukocyte-induced vascular per-meability, depletion of FGD5 also caused a decrease in the number of PMNs that transmigrated (Figure 2B). To further investigate the functional effects of endothelial FGD5 depletion on leukocyte TEM, we included an extra shRNA against FGD5 (Figure 1I) and performed TEM under flow assays 14. Control
compared to control ECs in basal (green line) and upon PMN addition (red line). Data are presented as mean±SEM (dotted lines). (G) Kinetics of calcein red-orange
labelled PMNs transmigration across control (green line) and LARG knockdown (red line) ECs towards C5a. Data are presented as mean±SEM (dotted lines). (H)
End point FITC-dextran leakage through ECs transduced with shRNAs resulting in PMN induced dextran leakage upon silencing of LARG, Ect2, FGD5, Tuba, BCR and ARHGAP11a. (I) Western blot of HUVECs transduced with shCTRL, shFGD5
(two shRNAs) or shTuba (two shRNAs) showing knockdown efficiency of the used shRNAs. Dotted lines indicate removed lane. Comparisons between the indicated conditions were made in Prism Graph-Pad using unpaired T-test. P values for results are: n.s. P>0.05, * P≤ 0.05, ** P≤0.01, *** P≤0.001.
Figure 2 shCTRL shCTRL + PMN shFGD5 shFGD5 + PMN 0 0 20 40 40 30 30 60 80 100 1 2 3 4 5 0 10 20 30 40 50 60 Time (min) 0 0 0 0 0 0.00 0.02 0.04 0.06 0.08 5 15 10 10 10 100 50 10 20 20 20 20 30 40 50 60 Time (min) Time (min) shCTRL shFGD5-1 shFGD5-2 shCTRL shFGD5-1 shFGD5-2 0 5 10 15 20 Time (min)
Dextran leakage (a.u.)
A C D B shCTRL + PMN shFGD5 + PMN Transmigrated PMNs (a.u.) Adhesion shCTRL shFGD5-1 shFGD5-2 # of PMNs % of total cells ICAM1 filopodia/um 2 # of PMNs Transmigration shCTRL shFGD5-1 shFGD5-2
E 20 min after PMN addition G
TEM Adhesive ICAM1 filopodia *** *** F H shCTRL shFDG5-2 25 µm Actin FGD5-mCherry ICAM1 ICAM1 ICAM1 FGD5-mCherry
VE-cadherin ICAM1 Zoom
10 µm + 10 ng/ml TNF α + 10 ng/ml TNF α 10 µm 100 µm
Figure 2. FGD5 regulates Cdc42 induced ICAM-1 filopodia formation
(A) FITC-dextran leakage through FGD5-depleted ECs in basal condition (blue line)
and upon addition of PMNs (purple line) compared to control ECs in basal (green line) and upon PMN addition (red line). Data are presented as mean±SEM (dotted lines). (B) Kinetics of calcein red-orange labelled PMNs transmigration across
con-trol (green line) and FGD5 depleted (red line) ECs towards C5a. Data are presented as mean±SEM (dotted lines). (C) Number of adhesive PMNs over time on control
(green line) and two shFGD5 (red and orange line)-transduced ECs under physiolo-gical flow rates of 0.8 dyn/cm2. (D) Number of transmigrated PMNs over time on
control (green line) and two shFGD5 (red and orange line)-transduced ECs under physiological flow. (E) Percentage of adhesive and transmigrated PMNs 20 minutes
after addition to ECs under physiological flow. (F) Immunofluorescent staining for
F-actin (red), ICAM-1 (green), VE-cadherin (white) and DNA (blue) on control and FGD5 silenced ECs after overnight TNFα stimulation. White box indicates area of ICAM-1 zoom. (G) Quantification of ICAM-1 filopodia/µm2 present on shCTRL
and shFGD5 transduced and TNFα treated ECs. (H) Immunofluorescent staining on
FGD5-mCherry (red) overexpressing HUVECs for VE-cadherin (white) and ICAM-1 (green) shows co-localization for ICAM-ICAM-1 filopodia and FGD5 after TNFα treat-ment.
and FGD5-deficient ECs were cultured in an Ibidi flow chamber and treated overnight with TNFα. Freshly isolated PMNs were injected into the injecti-on port of the flow system and adhesiinjecti-on and transmigratiinjecti-on numbers were quantified over 20 minutes. Depletion of FGD5 resulted in reduced adhesion of PMNs to the EC monolayer within the first 5 minutes after addition (Figure 2C), and subsequently reduced number of PMNs that crossed the EC layer after 20 minutes (Figure 2D). However, the percentage of adhesive cells that were capable of transmigrating was not affected (Figure 2E), indicating that endothelial FGD5 silencing only affects PMN adhesion to the endothelium, and not the transmigration step (i.e. diapedesis) itself.
Immunofluorescent staining for F-actin, ICAM-1 and VE-cadherin re-vealed that ECs treated with TNFα and transduced with shFGD5 showed a decrease in ICAM-1-positive filopodia compared to shCTRL, while VE-cad-herin distribution was unaltered (Figure 2F). Quantification of ICAM-1 filopo-dia showed a significant reduction of filopofilopo-dia/µm2 on shFGD5 ECs when compared to shCTRL cells with two different shRNAs (Figure 2G). Over-expression of mCherry-tagged FGD5 revealed co-localization with ICAM-1 filopodia after TNFα stimulation (Figure 2H).
In conclusion, we found that the Rho-GEF FGD5 regulates forma-tion of ICAM-1-rich filopodia upon TNFα treatment. In combinaforma-tion with re-sults from Kroon and co-workers on Cdc42 depletion and ICAM-1 filopodia formation 13, we suggest that FGD5 regulates Cdc42 activation upon TNFα
Figure 3 A C D B E G F 0 1 2 3 4 0 10 20 30 40 50 60 Time (min)
Dextran leakage (a.u.)
shCTRL shCTRL + PMN shTuba shTuba + PMN 0 20 40 60 80 100 0 10 20 30 40 50 60 Time (min) Transmigrated PMNs (a.u.) shCTRL + PMN shTuba + PMN 20 15 0 0 5 10 15 5 20 10 Time (min) # of PMNs Adhesion shCTRL shTuba-1 shTuba-2 30 0 10 20 0 5 10 15 20 Time (min) # of PMNs Transmigration 40 shCTRL shTuba-1 shTuba-2 0 100 50 shCTRL shTuba-1 shTuba-2 % of total cells
20 min after PMN addition
TEM Adhesive 0.00 0.02 0.04 0.06 0.08 shCTRL shTuba-1 shTuba-2 ICAM1 filopodia/um 2 ICAM1 filopodia * * ICAM1 filopodia in FGD5 depleted ECs (see Figure 2G) shCTRL shFTuba-2
Actin ICAM1 VE-cadherin ICAM1 Zoom
+ 10ng/ml TNF
α
25 µm 10 µm
Figure 3. Tuba regulates Cdc42 activity to prevent vascular permeability (A) FITC-dextran leakage through Tuba depleted ECs in basal condition (blue line)
and upon addition of PMNs (purple line) compared to control ECs in basal (green line) and upon PMN addition (red line). Data are presented as mean±SEM (dotted lines). (B) Kinetics of calcein red-orange labelled PMNs transmigration across
con-trol (green line) and Tuba depleted (red line) ECs towards C5a. Data are presented as mean±SEM (dotted lines). (C) Number of adhesive PMNs over time on control
physiolo-stimulation, prior to and essential for ICAM-1 filopodia formation. These ICAM-1-positive filopodia are essential for proper leukocyte adhesion under inflammatory conditions, hence the disturbed PMN adhesion when FGD5 is absent.
Tuba regulates Cdc42 activity at the site of PMN breaching
Like FGD5, silencing of Tuba in ECs has no effect on basal barrier function, but only provokes leukocyte-induced leakage (Figure 3A). However, in con-trast to FGD5, Tuba depletion did not affect PMN transmigration toward C5a in the Transwell set-up. To exclude redundancy between FGD5 and Tuba, leukocyte transmigration kinetics under physiological flow was assessed. Using two different shRNAs against Tuba (Figure 1I), Tuba-deficient ECs showed no defect in PMN adhesion during the first 5 minutes after PMN ad-dition (Figure 3C). Moreover, transmigration rates of PMNs after 20 minutes is also not compromised upon silencing of Tuba (Figure 3D), which is also reflected in the end point quantification of percentage PMNs that crossed the endothelium, which did not significantly change upon depletion of endothelial Tuba with two different shRNAs (Figure 3E).
Immunofluorescent staining of F-actin, ICAM-1 and VE-cadherin on shCTRL or shTuba transduced ECs treated with TNFα revealed no clear differences in F-actin and VE-cadherin organization. We did find a significant reduction in the number of 1-positive filopodia (Figure 3F) and ICAM-1 filopodia/µm2 cell surface (Figure 3G), between control and Tuba knock-down ECs. Interestingly, despite the reduction in ICAM-1 positive filopodia, there was no difference in the number of adhesive PMNs (Figure 3C). So, endothelial Tuba seems mainly involved in EC barrier function upon PMN TEM, rather than regulating PMN adhesion.
gical flow rates of 0.8 dyn/cm2. (D) Number of transmigrated PMNs over time on
control (green line) and two shTuba (red and orange line) transduced ECs under physiological flow. (E) Percentage of adhesive and transmigrated PMNs 20 minutes
after addition to ECs under physiological flow. (F) Immunofluorescent staining for
F-actin (red), ICAM-1 (green), VE-cadherin (white) and DNA (blue) on control and Tuba silenced ECs after overnight TNFα stimulation. White box indicates area of ICAM-1 zoom. (G) Quantification of ICAM-1 filopodia/µm2 present on shCTRL
and shTuba-transduced and TNFα treated ECs. Level of ICAM-1 filopodia/µm2 upon FGD5 depletion (see Figure 2G) is indicated with dashed line.
Gene name Rho GTPase target Activity Alternative names
TRIO Rac1, RhoG, RhoA GEF ARHGEF23, tga
VAV2 Rac1, RhoA, RhoG, Cdc42 GEF
Β-PIX Rac1, Cdc42 GEF ARHGEF7/COOL-1/Nbla10314/PAK3
NME1 Rac1 (Tiam1), Cdc42 (Dbl) Arf6 GEF NB/AWD/NBS/GAAD/NDKA/NM23/NDPKA/NDPK-A/Nm23-H1
PLEKHG1 RhoA, RhoB, RhoC, Cdc42 GEF D10Ertd733e/Gm521/mKIAA1209
CdGAP Rac1, Cdc42 GAP ARHGAP31/AOS1
ARAP3 RhoA, Arf6 GAP CENTD3/DRAG1
Gene name Rho GTPase target Activity Alternative names
DOCK4 Rac1, Rac2 GEF WUGSC:H_GS034D21.1
DOCK6 Rac1, Cdc42 GEF AOS2/ZIR1
DOCK9 Cdc42 GEF RP11-155N3.2/ZIZ1/ZIZIMIN1
ALS2 Rac1 GEF ALSJ/PLSJ/IAHSP/ALS2CR6
ARHGEF10 RhoA, RhoB, RhoC GEF GEF10
ARHGAP24 Rac1, Cdc42 GAP FilGAP/RC-GAP72/RCGAP72/p73/p73RhoGAP
ARHGAP29 RhoA GAP PARG1/RP11-255E17.1
ARHGAP17 Rac1, Cdc42, RhoA GAP PP367/RICH1/WBP15/MST066/MST110/NADRIN/PP4534/RICH1B/MSTP0
38/MSTP066/MSTP110
DEPDC1B GAP XTP1/BRCC3
RACGAP1 Rac1, Cdc42 GAP CYK4/HsCYK-4/ID-GAP/MgcRacGAP
Gene name Rho GTPase target Activity Alternative names
LARG RhoA, RhoC GEF ARHGEF12/RO2792
Ect2 RhoA, Rac1, Cdc42 GEF ARHGEF31
FGD5 Cdc42 GEF ZFYVE23
Tuba Cdc42 GEF DNMBP/ARHGEF36/RP11-114F7.3
Bcr RhoACdc42 (GAP)(GEF), Rac1(GAP), GEF/GAP
ARHGAP11a RhoA, RhoB, RhoC GAP
Table 1. Rho-GEFs and -GAPs regulating basal endothelial permeability
Table 2. Rho-GEFs and -GAPs regulating leukocyte induced vascular leakage
Table 3. Rho-GEFS and –GAPs not involved in vascular leakage or leukocyte TEM
Discussion
The family of Rho GTPases are widely studied for their role in endothelial permeability and leukocyte TEM, however, which GEFs and GAPs regulate the spatial and temporal activation of these GTPases at endothelial junctions or at sites of leukocyte breaching is poorly understood. Therefore, we per-formed a novel screening assay with shRNAs targeting 23 endothelial Rho-GEFs and -GAPs to identify potential regulators for EC permeability under basal conditions or during PMN diapedesis. We found seven GEFs and/or GAPs, mainly with activity towards Rac1, that are involved in the regulation of basal endothelial junction integrity. We also discovered six GEFs and/or GAPs with activity towards RhoA or Cdc42 that are involved in regulation of leukocyte induced endothelial permeability.
Our novel screening assay identified TRIO, VAV2, β-PIX, NME1, PLEKHG1, CdGAP, and ARAP3 as regulators of basal endothelial perme-ability. A number of these regulators have been described before to regu-late EC junctions. We previously demonstrated that Rho-GEF TRIO-deple-ted-ECs display instable VE-cadherin-based cell-cell junctions, resulting in increased basal endothelial permeability 11. The Rho-GEF VAV2 strengthens
endothelial junctions through Rac1 activation 15. β-PIX plays a pivotal role
in LPS-mediated induction of vascular permeability in vivo 16 and NME1
lo-calization to EC junctions is essential for correct VE-cadherin lolo-calization
17. ARAP3 impairs EC function, since ARAP3-deficient mice and zebrafish
are embryonically lethal due to angiogenesis defects 18,19. Apart from these
known endothelial GEFs and GAPs, we also identified a function in ECs for the Rho-GEF PLEKHG1 and the Cdc42/Rac1 regulator CdGAP, which is not described before. Although nothing is known about CdGAP in ECs, it is involved in repressing E-cadherin expression levels in epithelial cells and thereby promoting destabilization of cell-cell junctions in breast cancer cells
20. This makes CdGAP an interesting target to study in ECs as vascular
per-meability relies on EC cell-cell junction stability. PLEKHG1 on the other hand is involved in mechano-signaling of cyclic stretch in ECs, and the consequent reorientation of ECs in a direction perpendicular to the stretch 21. This
ma-kes PLEKHG1 a potential target for studying the effects of ECs reorientation upon cyclic stretch and how this might affect vascular permeability.
Our novel screening assay identified LARG, ECT2, FGD5, Tuba, Bcr and ARHGAP11a to regulate leukocyte-induced vascular leakage. Earlier work from our group, described in Chapter 3, already showed that the GEFs LARG and Ect2 were involved in leukocyte-induced leakage 8. In addition
to RhoA-specific GEFs, nothing is known about the role of Rho-GAPs Bcr and ARHGAP11a in ECs. In certain cancer cells, ARHGAP11a is expressed during cell-cycle, and thereby making proliferative cancer cells more motile
due to RhoA inhibition 22. However, in vivo ECs don’t proliferate that much
besides angiogenesis, leaving a function for ARHGAP11a in leukocyte TEM to be elucidated. Bcr remains another mystery, as this proteins contains 2 catalytic domains; GEF for RhoA and GAP for Rac1 23. This makes Bcr highly
interesting for regulating leukocyte TEM and vascular permeability by oppo-sing RhoA-Rac1 activation/deactivation, however, it’s mostly been studied in neuronal development 24.
Two Cdc42-specific Rho-GEFs FGD5 and Tuba were identified as po-tential regulators of leukocyte-induced vascular leakage and leukocyte TEM efficiency. This indicates that the small GTPase Cdc42 is spatially and tem-porally regulated during the process of leukocyte TEM. We recently reported on the role of Cdc42 in leukocyte TEM. Together with Myosin-X, Cdc42 is responsible for the formation of ICAM-1-rich filopodia on the apical surface of the inflamed endothelium 13. Reduction of the number of ICAM-1 filopodia
by reducing the expression of Cdc42 results in a decrease of adhesive leu-kocytes to the endothelium, although the percentage leuleu-kocytes that cross the endothelial monolayer is unaltered, similar to what we show here with FGD5-depleted ECs. Our findings presented here argue that the Cdc42-spe-cific Rho-GEF FGD5 is responsible for local Cdc42 activity and formation of ICAM-1 filopodia at the apical surface of the endothelium upon TNFα stimu-lation. Silencing FGD5 results in reduced filopodia and adherent leukocytes. Interestingly, a gain of function experiment, by overexpressing FDG5 failed: no increase in the number of ICAM-1-rich filopodia was detected. However, this is in line with our previous results where overexpression of constitutively active Cdc42 mutant (Q61L) also failed to increase the number of adherent PMNs under physiological flow 13, implicating that under inflammatory
con-ditions, like TNFα, the maximum number of adhesive PMNs is reached, and further increase in ICAM-1-rich filopodia has no additional effect.
Both FGD5 and Tuba depletion resulted in deceased ICAM-1-rich fi-lopodia formation (Figure 2G and 3G). However, a strong defect in adhesion of PMNs was only seen in FGD5-depleted ECs. Silencing Tuba did result in a reduced number of filopodia, however, not to the extent that was detected upon FGD5 silencing. This suggests that the number of filopodia present on Tuba-depleted ECs has reached the critical minimum to facilitate proper PMN adhesion. In conclusion, under normal inflammatory conditions, the number of ICAM-1-rich filopodia is abundantly present on the surface of the endothelium.
Involvement of the small GTPase Cdc42 in the regulation of EC permeability under basal conditions is known to be controlled by VE-cad-herin-based cell-cell junctions 25–27. Our data also implicate a role for Cdc42
in regulating the endothelial cell-cell junction, in particular when a leukocyte has crossed. Then, Cdc42 activity, most likely in conjunction with Rac1 28 is
Functional properties of Tuba have been linked to both Rho GTPa-se signaling specifically at curved membranes. Since Tuba contains a Bin/ amphiphysin/Rvs (BAR) domain instead of the classical plekstrin-homology (PH) domain for its membrane localization. These BAR domains are capable of binding specifically to lipids in curved membranes, and thereby recruiting Tuba to a specific location 29,30. This is of particular interest during
leuko-cyte diapedesis, where EC membranes are deformed by the probing and breaching leukocytes. Although the exact mechanisms and localization of endothelial Tuba to the sites of leukocyte diapedesis are not understood, we speculate that Tuba is recruited towards the formed docking structure via its BAR domain and consequently regulates local Cdc42 activity for pore closer after passage of the leukocyte in order to limit vascular leakage. In epithe-lial cells, Tuba induces formation of linear cell-cell-junctions by activation of Cdc42 which creates junctional surface tension in order to seal the cell-cell junction 31. A similar process could also be present in ECs, where after
paracellular leukocyte TEM, restoration of cell-cell junctions is essential to maintain vascular barrier function. However, further mechanistic studies with BAR domain-deficient Tuba mutants will prove the involvement and impor-tance of such BAR domains in regulating Cdc42 in order to maintain vascular integrity during leukocyte TEM.
In conclusion, we discovered several endothelial Rho-GEFs and -GAPs involved in regulating endothelial permeability both under basal con-ditions or during leukocyte diapedesis and leukocyte TEM efficiency. The number of found Rho-GEFs and -GAPs shows the complexity of the process of leukocyte TEM. This is further underscored by our data showing that two distinct GEFs, FGD5 and Tuba, that regulate the same GTPase, are involved in two different processes during leukocyte TEM; namely adhesion and pore closure. A novel regulatory pathway of Cdc42 through FGD5 upon TNFα sti-mulation is essential for ICAM-1 filopodia formation. Moreover, Cdc42 activa-tion regulated by Tuba is most likely involved in pore closure after leukocyte diapedesis in order to maintain EC barrier.
Experimental Procedures
DNA and RNA constructsshRNA constructs enlisted below were obtained from Sigma Aldrich mission library. FGD5-mCherry (a kind gift of Joachim Goedhart) for overexpression were microporated into HUVECs using Neon Transfection System (Thermo-Fisher) according to manufacturer’s protocol.
shRNA TRCN# Sequence shCtrl shC002 TTGGTGCTCTTCATCTTGTTG shTrio 10561 GCTTCCCAGATGACTACTTTA shARAP3 47344 GTCTGGAGTAATGAGATAGTA shARAP3 47347 GTGATACCTTTGACTGCCATT shARHGAP11A 47281 CGGTATCAGTTCACATCGATA shARHGAP11A 47282 CCTTCTATTACACCTCAAGAA shARHGAP17 47449 GCAGTGATTGAACCCATCATT shARHGAP17 47452 CCCAAGCAGATTACCATAGAA shARHGAP24 48303 GCGTGGACTTTATCCGACAAA shARHGAP24 48307 CCGAGAGAGGAAACACAATAT shARHGEF10 47460 CCTGAACCTTACCTAAATAAT shARHGEF10 47461 CGCGAAACCAAACAAAGTTTA shLARG 47480 GCGTTGCGTAATCATCCAGAA shLARG 47482 GCGAGTATCCAGAGAAGGAAT shB-PIX 47595 CCTGGGATGAGACCAATCTAT shB-PIX 47596 GAAGTTAAGTTCAGCAAACAT shBCR 195792 GTTCCTGATCTCCTCTGACTA shBCR 195722 CCCTCACTGTTGTATCTTGAA shDOCK4 39730 GCCGTGCATGAGAAGTTTGTA shDOCK4 39731 GCTTCGAGTTTCGGCATTGTT shDOCK6 122965 GCCTACATCCAGATCACGTAT shDOCK6 122967 CCTGGTCATCAAGTTGGAGAA shDOCK9 139678 CCAACAGTGACCGGCTTATTA shDOCK9 141047 CGCGAGCTTTCTTAGATGATA shNME1 10055 GCGTACCTTCATTGCGATCAA shNME1 10061 TCCGCCTTGTTGGTCTGAAAT shRACGAP1 47293 CAGGTGGATGTAGAGATCAA shRACGAP1 47296 CATTGATGAATCTGGTTCCAT shALS2 47803 CGACTAAATAAGCAGCCAGAT shALS2 47807 GCTATGCTTCTGGTGAAGTAT shARHGAP29 2149 GCCTCACTGGAATTATCTTTA shARHGAP29 2150 GCAGCTCTCACACACAAGTTT shCdGAP 47638 CCTCAAACATACAACGGCTAA shCdGAP 47640 GCTCTCCAGTAAATCAAAGAA shDEPDC1B 61871 CGTCAAATTAGTCCAGAGGAA shDEPDC1B 61872 GCTGCTAGATTGGTAACGTTT shTuba 148251 CCGAATGCAAGAAAGATTGAT shTuba 148479 CTCCACAACCTAGCAAGTTAT shECT2 47683 CCAGCAATGATAAGCATGTAA shECT2 47684 GCCCGTTGTATTGTACAAGTA shFGD5 48245 CCGGCACCTATTTCTGATGAA shFGD5 48247 CCTGGCACTGACGTTTAAGAA shPLEKHG1 47198 CGGCAACACATTGCATTCTTT shPLEKHG1 47201 GCCTAGCAGTTCTACCATGAT shVAV2 48223 GCCACGATAAATTTGGATTAA shVAV2 48226 CCCGAGATATGAGGGAGCTTT
Antibodies
Monoclonal mouse antibody against actin (Cat#A3853) (1:1000 for WB) and rabbit polyclonal against Tuba (Cat#SAB1410413) (1:1000 for WB) was purchased from Sigma. Monoclonal mouse antibody against FGD5 (Cat# H00152273-B01P) (1:1000 for WB) was purchased from Abnova. Polyclonal rabbit antibody against ICAM-1 (Cat#SC-7891) (1:1000 for WB) was pur-chased from Santa Cruz. Monoclonal mouse antibody against VE-cadherin AF647 (Cat#561567) (1:200 for IF) was purchased from BD Biosciences. Monoclonal mouse antibody against ICAM-1 FITC (Cat#BBA20) (1:200 for IF) was purchased from R&D systems. Alexa Fluor 555 Phalloidin (Cat#P-HDH1) (1:500 for IF) was purchased from Cytoskeleton. Hoechst 33342 (Ca-t#H-1399) (1:25000 for IF) was purchased from Molecular probes. Secon-dary HRP-conjugated goat anti-mouse (Cat#P0447) (1:5000 for WB), swine anti-rabbit (Cat#P0399) (1:5000 for WB) antibodies were purchased from Dako. All antibodies were used according to manufacturer’s protocol.
Cell cultures and treatments
Pooled human umbilical vein endothelial cells (HUVECs) perchased from Lonza (P938 and P1012, Cat # C2419A), were cultured until passage 7 on fibronectin (FN)-coated dishes in EGM-2 medium, supplemented with singlequots (Lonza). Human Embryonic Kidney (HEK)-293T cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM) (Invitrogen), con-taining 10% (v/v) heat-inactivated fetal calf serum, 100 U/ml penicillin and streptomycin and 1x sodium pyruvate (all Invitrogen). All cells were cultured at 37°C and 5% CO2. Lentiviral constructs were packaged into lentivirus in HEK-293T cells by means of third generation lentiviral packaging plasmids (Dull et al., 1998; Hope et al 1990). Lentivirus containing supernatant was harvested on day 2 and 3 after transfection. Lentivirus was concentrated by Lenti-X concentrator (Clontech, Cat# 631232). Cells were transfected with the expression vectors according to manufacturer’s protocol with Trans- IT-LT1 (Myrus) and transduced target cells were selected with Puromycin ( mg/ ml) () after >24 hours and used for assays after >72 hours.
PMN isolation
Polymorphonuclear cells (PMNs) were isolated from whole blood derived from healthy donors in sodium heparin tubes. Whole blood was diluted (1:1) with 10% (v/v) TNC in PBS. Diluted whole blood was pipetted carefully on 12,5 ml Percoll (room temperature) 1.076 g/ml. Tubes were centrifuged (Ro-tanta 96R) at 2000 rpm, slow start, low brake for 20 minutes. After erythrocy-te lysis in an ice-cold isotonic lysis buffer (155 mM NH4CL, 10 mM KHCO3, 0.1 mM EDTA, pH7.4 in Milli-Q (Millipore)), PMNs were centrifuged at 1500 rpm for five minutes at 4°C, incubated once with lysis buffer for 5 minutes on ice, centrifuged again at 1500 rpm for five minutes at 4°C, washed once with
PBS, centrifuged again at 1500 rpm for five minutes at 4°C and resuspended in N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (HEPES)-buffer pH7.4 (20 mM HEPES, 132 mM NaCl, 6 mM KCL, 1 mM CaCL2, 1 mM MgSO4, 1.2 mM K2HPO4, 5 mM glucose (all from Sigma-Aldrich), and 0.4 % (w/v) human serum albumin (Sanquin Reagents) and kept at room tempe-rature for not longer than four hours until use. PMN counts were determined by cell counter (Casey).
FITC-dextran permeability assay
200,000 shRNA transduced and puromycin selected HUVECs were cultured in fibronectin coated 24 well cell culture inserts (Corning FluoroBlok, Fal-con, 3.0 μm pore size Cat# 351151) and treated with 10 ng/ml recombinant Tumor-Necrosis-Factor (TNF)-α (PeproTech) overnight. 30 μg FITC-dextran (70 kDa) (Sigma) in HEPES medium pH7.4 was added to the upper and 0.1 nM C5a (Sigma C-5788) in HEPES medium pH7.4 was added to the lower compartment. Freshly isolated PMNs were labeled and activated with 4 µg calcein red-orange (Molecular probes C34851) per 5*10^6 cells in 1 ml HEPES for 30 min at 37°C and 500,000 labeled PMNs were added to the top compartment. Simultaneous measurement of FITC-dextran and calcein red-orange PMN extravasation was performed for a period of 60 minutes with an interval of 1 minute by an Infinite F200 pro plate reader (TECAN) at 37°C. EX BP 485/9 and EM BP 535/20 was used to measure FITC-dex-tran kinetics. EX BP 570/9 and EM BP 595/20 was used to measure calcein red-orange PMNs transmigration kinetics.
PMN TEM under physiological flow
150.000 shRNA tranduced and puromycin selected HUVECs were cultured per channel in a FN-coated Ibidi µ-slide VI0.4 (Ibidi) the day before the expe-riment was executed and stimulated overnight with 10 ng/ml TNFα (Pepro-tech). Freshly isolated PMNs were resuspended at 1*106 cells/ml in HEPES medium pH7.4 and were activated for 30 minutes at 37ºC. Ibidi flow cham-bers were connected to a perfusion system and exposed to 0.5 ml/minu-te HEPES medium pH7.4 shear flow for 5 minuml/minu-tes (0.8 dyne/cm2). PMNs were subsequently injected into the perfusion system and PMN-endothelial interactions were recorded for 20 minutes with an interval of 5 seconds by a Zeiss Observer Z1 microscope. All live imaging was performed at 37°C in the presence of 5% CO2.
Western blotting
For total cell lysates, cells were washed once with PBS+/+ (1mM CaCl and 0.5 mM MgCl), and lysed with 95°C SDS-sample buffer containing 4% β-me-capto-ethanol. Samples were boiled at 95°C for 5-10 minutes to denature proteins. Proteins were separated on 10% SDS running gel in running buffer
(200 mM Glycine, 25 mM Tris, 0.1% SDS (pH8.6)), transferred to nitrocellu-lose membrane (Thermo Scientific Cat#26619) in blot buffer (48 nM Tris, 39 nM Glycine, 0.04% SDS, 20% MeOH) and subsequently blocked with 5% (w/v) milk (Campina) in Tris-buffered saline with Tween 20 (TBST) for 30 mi-nutes. The immunoblots were analyzed using primary antibodies incubated overnight at 4°C and secondary antibodies linked to horseradish peroxidase (HRP) (Dako, Aligent Tochnologies), after each step immunoblots were was-hed 4x with TBST. HRP signals were visualized by enhanced chemilumines-cence (ECL) (Thermo Scientific) and light sensitive films (Fujifilm).
Confocal laser scanning microscopy and image analysis
Immunofluorescent staining was in general performed on HUVECS cultured on 12 mm glass coverslips coated with 5 µg/ml FN and treated with or wit-hout o/n TNFα (10ng/ml) (Peprotech), washed with PBS+/+(1mM CaCl2 and 0.5mM MgCl2), fixated in 4% PFA (Merck), blocked for 30 min with 2% BSA (Affimetrix) and mounted in Mowiol4-88/DABCO solution. Z-stack image acquisition was performed on a confocal laser scanning microscope (Leica SP8) using a 63x NA 1.4 oil immersion objective.
ICAM1 positive filopodia where quantified using ImageJ 1.51p. Maxi-mum projections of fluorescent images were subjected to Gaussian blur and Unsharp filtering, followed by automated ‘Max entropy’ thresholding. Next, a watershed-based segmentation was applied and particles were detected using particle analysis with a minimal size cut-off of 8 pixels (0,18 µm/pixel). Statistics. Data in graphs are represented as mean ± SEM obtained from 3 independent experiments. Immunofluorescent images show representa-tives. Comparisons between the indicated conditions were made in Prism Graph-Pad using unpaired T-test. P values for results are: n.s. P>0.05, * P≤ 0.05, ** P≤0.01, *** P≤0.001, **** P≤0.0001.
Acknowledgements
We kindly thank Dr. Joachim Goedhart, Department of Molecular Cytology, Swammerdam Institute of Life Sciences, University of Amsterdam, Amster-dam, the Netherlands, for providing us the FGD5-mCherry construct. We sincerely thank Prof. Dr. Peter Hordijk for critically reading the manuscript.
References
1. Wessel, F. et al. Leukocyte extravasation and vascular permeability are each controlled in vivo by different tyrosine residues of VE-ca-dherin. Nat. Immunol. 15, 223–30 (2014). 2. Muller, W. A. Localized signals that regulate
transendothelial migration. Current Opinion in Immunology 38, 24–29 (2016).
3. Schimmel, L., Heemskerk, N. & van Buul, J. D. Leukocyte transendothelial migration: A local affair. Small GTPases 8, (2017). 4. Valeski, J. E. & Baldwin, a L. Effect of early
transient adherent leukocytes on venular per-meability and endothelial actin cytoskeleton. Am. J. Physiol. 277, H569–H575 (1999). 5. Kim, M.-H., Curry, F.-R. E. & Simon, S. I.
Dynamics of neutrophil extravasation and vascular permeability are uncoupled during aseptic cutaneous wounding. Am. J. Physiol. Cell Physiol. 296, C848-56 (2009).
6. Lewis, R. E. & Granger, H. J. Diapedesis and the permeability of venous microvessels to protein macromolecules: the impact of leukotriene B4 (LTB4). Microvasc. Res. 35, 27–47 (1988).
7. Lewis, R. E., Miller, R. a. & Granger, H. J. Acute microvascular effects of the chemotac-tic peptide N-formyl-methionyl-leucyl-phe-nylalanine: Comparisons with leukotriene B4. Microvasc. Res. 37, 53–69 (1989). 8. Heemskerk, N. et al. F-actin-rich contractile
endothelial pores prevent vascular leakage during leukocyte diapedesis through local RhoA signalling. Nat. Commun. 7, (2016). 9. Amerongen, G. P. V. N. et al. Activation of
RhoA by Thrombin in. Circ. Res. 335–340 (2000). doi:10.1161/01.RES.87.4.335 10. Schimmel, L., de Ligt, A., Tol, S., de Waard,
V. & van Buul, J. D. Endothelial RhoB and RhoC are dispensable for leukocyte diape-desis and for maintaining vascular integrity during diapedesis. Small GTPases 0 (2017). doi:10.1080/21541248.2017.1377815 11. Timmerman, I. et al. A local VE-cadherin
and Trio-based signaling complex stabilizes endothelial junctions through Rac1. J. Cell Sci. 128, 3041–54 (2015).
12. van Rijssel, J. et al. The Rho-guanine nucleo-tide exchange factor Trio controls leukocy-te transendothelial migration by promoting docking structure formation. Mol. Biol. Cell
23, 2831–2844 (2012).
13. Kroon, J. et al. Inflammation-Sensitive My-osin-X Functionally Supports Leukocyte Extravasation by Cdc42-Mediated ICAM-1–Rich Endothelial Filopodia Formation. J. Immunol. ji1700702 (2018). doi:10.4049/ jimmunol.1700702
14. Kroon, J., Daniel, A. E., Hoogenboezem, M. & van Buul, J. D. Real-time Imaging of Endo-thelial Cell-cell Junctions During Neutrophil Transmigration Under Physiological Flow. J. Vis. Exp. e51766 (2014). doi:10.3791/51766 15. Schlegel, N. & Waschke, J. CAMP with
other signaling cues converges on Rac1 to stabilize the endothelial barrier - A signaling pathway compromised in inflammation. Cell Tissue Res. 355, 587–596 (2014).
16. Stockton, R. et al. Induction of Vascular Per-meability: BetaPIX and GIT1 Scaffold the Activation of Extracellular Signal-regulated Kinase by PAK. Mol. Biol. Cell 18, 2346– 2355 (2007).
17. Moreno, V. et al. An EMMPRIN-γ-caten-in-Nm23 complex drives ATP production and actomyosin contractility at endothelial junctions. J. Cell Sci. 127, 3768–81 (2014). 18. Gambardella, L. et al. PI3K signaling
through the dual GTPase-activating protein ARAP3 is essential for developmental
angio-genesis. Sci. Signal. 3, ra76 (2010).
19. Kartopawiro, J. et al. Arap3 is dysregulated in a mouse model of hypotrichosis-lymphe-dema-telangiectasia and regulates lymphatic vascular development. Hum. Mol. Genet. 23, 1286–1297 (2014).
20. He, Y. et al. The Cdc42/Rac1 regulator Cd-GAP is a novel E-cadherin transcriptional co-repressor with Zeb2 in breast cancer. On-cogene 36, 3490–3503 (2017).
21. Abiko, H. et al. Rho guanine nucleotide ex-change factors involved in cyclic-stretch-in-duced reorientation of vascular endothelial cells. J. Cell Sci. 128, 1683–95 (2015). 22. Kagawa, Y. et al. Cell cycle-dependent Rho
GTPase activity dynamically regulates can-cer cell motility and invasion in vivo. PLoS One 8, (2013).
23. Chuang, T. H. et al. Abr and Bcr are multi-functional regulators of the Rho GTP-bin-ding protein family. Proc. Natl. Acad. Sci. 92,
10282 LP-10286 (1995).
24. Park, A.-R. et al. Regulation of dendritic ar-borization by BCR Rac1 GTPase-activating protein, a substrate of PTPRT. J. Cell Sci. 125, 4518–4531 (2012).
25. Broman, M. T. et al. Cdc42 regulates adhe-rens junction stability and endothelial perme-ability by inducing α-catenin interaction with the vascular endothelial cadherin complex. Circ. Res. 98, 73–80 (2006).
26. Kouklis, P. Cdc42 Regulates the Restoration of Endothelial Barrier Function. Circ. Res. 94, 159–166 (2004).
27. Amado-Azevedo, J. et al. A CDC42-cente-red signaling unit is a dominant positive re-gulator of endothelial integrity. Sci. Rep. 7, (2017).
28. Martinelli, R. et al. Release of cellular ten-sion signals self-restorative ventral lamel-lipodia to heal barrier micro-wounds. J. Cell Biol. 201, 449–465 (2013).
29. Salazar, M. a. et al. Tuba, a novel protein containing bin/amphiphysin/Rvs and Dbl homology domains, links dynamin to regula-tion of the actin cytoskeleton. J. Biol. Chem. 278, 49031–49043 (2003).
30. Kovacs, E. M., Makar, R. S. & Gertler, F. B. Tuba stimulates intracellular N-WASP-de-pendent actin assembly. J. Cell Sci. 119, 2715–2726 (2006).
31. Otani, T., Ichii, T., Aono, S. & Takeichi, M. Cdc42 GEF Tuba regulates the junctional configuration of simple epithelial cells. J. Cell Biol. 175, 135–146 (2006).
