mechanisms of vessel branching -...

ATVB In FocusDevelopmental Biology in the Vasculature

Series Editor: Mark Majesky, PhD

Mechanisms of Vessel BranchingFilopodia on Endothelial Tip Cells Lead the Way

Frederik De Smet, Inmaculada Segura, Katrien De Bock, Philipp J. Hohensinner, Peter Carmeliet

Abstract—Filopodia, “the fingers that do the walking,” have been identified on endothelial cells at the tip of sproutingvessels for half a century, but the key role of the tip cell in vessel branching has been recognized only in the past fewyears. A model is emerging, whereby tip cells lead the way in a branching vessel, stalk cells elongate the sprout, anda very recently discovered phalanx cell ensures quiescence and perfusion of the newly formed branch. Recent geneticstudies have shed light on the molecular signature of these distinct endothelial phenotypes; this provides a novelconceptual framework of how vessel morphogenesis occurs. Here, we will discuss the molecular candidates thatparticipate in the decision of endothelial cells to adapt these distinct fates and highlight the emerging insights on howthese cells send out filopodia while navigating. (Arterioscler Thromb Vasc Biol. 2009;29:639-649.)

Key Words: endothelial tip cells � stalk cells � phalanx cells � filopodia � sprouting angiogenesis

Much has been learned about the importance of angio-genesis in health and disease, but our understanding of

how vessels branch remains incomplete. An exciting break-through in the field has been the recognition that several typesof specialized endothelial cells (ECs), each with distinctcellular fate specifications, are required to build a functionalbranch.1 The first and, to a certain extent, the most undertak-ing cell in a vessel branch is the “tip cell,” which leads theway. With their continuously searching filopodia, these tipcells sense and respond to guidance cues in their microenvi-ronment, similar to how an axonal growth cone in the nervoussystem2 or an epithelial tip cell in the fruitfly airway system3

explores its surroundings. Even in the sea squirt, tip cells usefilopodia to build a primitive vascular network.4 It is thereforenot surprising that several classes of molecules and principles,used by navigating axons or epithelial cells, are evolutionaryconserved and shared, and even might have been coopted by themigrating endothelial tip cell.1,2 “Stalk cells” trail behind the tipcell and elongate the stalk of the sprout; they proliferate, formjunctions, lay down extracellular matrix, and form a lumen.“Phalanx cells” are the most quiescent ECs, lining vessels oncethe new vessel branches have been consolidated; they form asmooth cobblestone monolayer and are aligned as in a phalanxformation of the ancient Greek soldiers, are covered by peri-cytes, stick to each other via tight junctions, are embedded in athick basement membrane, and stay foot. These cells areengaged in optimizing blood flow, tissue perfusion, and oxygen-ation.5 Tip, stalk, and phalanx ECs each have a specializedfunction in vessel branching. How these cells execute their job

depends, to a substantial extent, on the organization of theircytoskeleton; for instance, for an EC to migrate, it needs to formspike-like filopodia, fan-like lamellipodia, and polarize its actincytoskeleton in the direction of migration.

Endothelial Tip CellsEven though the existence of “filliform processes” at the tipof endothelial sprouts was already observed in the brainalmost half a century ago,6 the concept and importance ofvessel guidance was not fully appreciated until recently.7,8

Classical studies have documented the existence of “seamlessECs” in vivo and postulated the existence of differentendothelial fates (trunk cells with lumen and tip cells withoutlumen) on growing sprout capillaries.9 Electron microscopystudies further described the existence of filopodia at theleading edge of growing capillaries (Figure 1a and 1b).10

Tip cells have now been observed in various models ofsprouting angiogenesis. Key features of this cell are theirlocation at the forefront of vessel branches, highly polarizednature, and numerous filopodia probing the environment,while migrating toward an angiogenic stimulus.7,11 They donot form a lumen and, with some exceptions, proliferateminimally (Figure 1c and 1d).7,12,13 Tip cells have a specificmolecular signature, characterized by the expression ofVEGF receptor (VEGFR) 2, VEGFR3, platelet-derivedgrowth factor (PDGF)-BB, Unc5B, Delta-like ligand-4(Dll4), neuropilin-1 (NRP1), and others (see below).7,14–18

These cells detect gradients of navigatory cues and integratecombinatorial molecular codes into directional migration.

Received January 23, 2009; revision accepted February 23, 2009.From the Vesalius Research Center, VIB, K.U. Leuven, Belgium.Correspondence to Peter Carmeliet, MD, PhD, Vesalius Research Center, VIB, K.U. Leuven, Campus Gasthuisberg, Herestraat 49, B-3000, Leuven,

Belgium. E-mail [email protected]© 2009 American Heart Association, Inc.

Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.109.185165

639

by guest on May 30, 2018

http://atvb.ahajournals.org/D

ownloaded from

http://atvb.ahajournals.org/

Central to their phenotype is the formation of filopodia. Wewill first describe how tip cells are selected in the nascentvessel branch, and explain why ECs do not move as a sheetbut form a sprout in response to an angiogenic signal.

Induction of Tip Cell Formation by VEGFR2and Dll4Genetic studies show that a VEGF gradient is important in theprocess of selection and induction of the endothelial tip cell.7

Via binding to VEGFR2, VEGF induces a signaling cascade,which enables one EC to take the lead and become a tip cell,while its neighbors are prevented from doing so and, instead,are instructed to become stalk cells. Such lateral inhibitionrelies on tip-to-stalk cell communication by Dll4/Notchsignaling (Figure 2a).18–20 ECs express various Notch recep-tors (Notch1, 3, 4) and ligands (Dll1, Dll4, Jagged1,Jagged2).21 After ligand binding, Notch is cleaved intracel-

lularly, generating the Notch intracellular domain (NICD)that acts as a transcriptional regulator.21 Tip cells are exposedto the highest levels of VEGF, which induces the expressionof Dll4 in these cells.21 Dll4 binds to Notch on neighboring(future stalk) ECs and downregulates VEGFR2 signaling; thisdampens the VEGF-induced expression of Dll4 in these cells,thereby establishing a self-reinforcing feedback that allowsthe leading cell to gain and retain its tip position, whilepreventing the follower neighbors from leaving their positionin the stalk. Genetic mosaic studies reveal that Notch regu-lates EC specification by actively suppressing the tip cellphenotype in stalk cells.19

A number of questions about the mechanisms of Dll4 andNotch remain unanswered. For instance, it is unknownwhether Dll4 reverse-signaling prevents the tip cell frombecoming a stalk cell by inactivating Notch signaling in thiscell through internalization of the receptor.22 Another ques-tion is whether Dll4 induces cytoskeletal rearrangements inthe tip cell. Notch ligands such as Dll4, Dll1, and Jagged1contain a PDZ-binding motif, which facilitates interactionswith adaptor proteins, thereby mediating adhesion and migra-tion in the ligand-expressing cell.23 Such adaptor moleculesinclude Dlg1, MAGI proteins, and syntenin, which interactwith the cytoskeleton.23

Lamellipodia and Filopodia Lead the WayThe key job of endothelial tip cells is to navigate, a processthat relies on correct probing of microenvironmental cues,and translating them into a dynamic process of adhesion (atthe front) and deadhesion (at the rear), that ultimately leads tocell movement. Therefore, the tip cell forms lamellipodia andfilopodia. Surprisingly, however, relatively little is knownabout the processes regulating the assembly of these cellularprotrusions in ECs. We will therefore briefly overview somekey insights of this process, as deduced from studying theaxon growth cone, and thereafter discuss our current under-standing of these processes in ECs.

Lamellipodia are short veil-like structures in close prox-imity to the plasma membrane that contain a highly branchedactin network. Filopodia, on the other hand, consist of longspiky plasma membrane protrusions containing tight parallelbundles of filamentous actin (F-actin), which usually extendfrom lamellipodia (Figure 3a).24–26 Filopodia and lamellipo-

Figure 1. Vascular sprouts are guided by endothelial tip cells. a,Original picture showing an embryonic endothelial tip cellobtained by electron microscopy (reprinted with permissionfrom10). b, Confocal micrograph showing filopodia extensions atthe leading tip cell (reprinted with permission from7). c, Schemefrom 1972 proposing several endothelial subtypes in the angio-genic sprout (reprinted with permission from9). d, Schematicrepresentation of a tip cell (green) extending filopodia toward anangiogenic stimulus (red gradient), followed by stalk cells (pur-ple), while phalanx cells (gray) remain quiescent.

TNFα

TNFR

Jagged1 induction

Jagged1 Notch

c Inflammatory cytokines induce tip cell formation

Bradykinin

?

CD13

activeGTPase

a Role of VEGF and Dll4 in tip cell formation

VEGF

VEGFR2

VEGFR3NRP1

Dll4 induction

Dll4 Notch

b Cytoskeletal rearrangements

VEGF

VEGFR2

active GTPase

AmotIntegrins

SRF

Filopodia, Lamellipodia,Cell migration

AngiostatinFGFR

FGF2

Unc5BMMP

Netrin

Figure 2. Molecular pathways of tip cell signaling.a, VEGF signaling via VEGFR2 enables tip cell for-mation (green), whereas Dll4 signaling inducesstalk cells (purple). VEGFR3 is upregulated,whereas Neuropilin 1 (NRP1) is maintained. b, Dif-ferent signaling cascades converge to smallGTPases activation, thereby regulating filopodiaand lamellipodia formation in endothelial cells. c,Inflammatory cytokines (blue gradient), includingtumor necrosis factor � (TNF�) and bradykinin,also induce a tip cell phenotype.

640 Arterioscler Thromb Vasc Biol May 2009



ownloaded from


dia are highly dynamic structures, generated within minutesafter stimulation.27 Both structures are capable of probing theenvironment, thereby sensing the presence of attractive orrepulsive guidance cues.26 An attractive cue will induceF-actin polymerization thereby extending filopodia, whereasa repulsive cue causes depolymerization and retrograde actinflow, resulting in retraction.26 Filopodia and lamellipodia alsoadhere and form focal contact points to connect the cytoskel-eton to the extracellular matrix (ECM).27 This allows stressfibers of actin/myosin filaments to pull the cell toward theseanchors, and induce forward movement.27

F-actin fibers are elongated at their ‘barbed’ or positive endby proteins of the Profilin, Ena/Vasp, and Formin families,which promote polymerization of G-actin into long F-actinstrings in proximity of the plasma membrane (Figure 3b).24,25

Whereas Profilin and Formin directly bind to G-actin andpresent it to the extending fiber, Ena/Vasp functions as ananticapping protein keeping the barbed end clear of inhibitorycap-proteins.25–27 Elongation of these filaments pushes theleading edge forward and promotes cell migration. At the“pointed” or negative end of the actin-filament (pointingtoward the inside of the cell), Cofilin depolymerizes andshortens the actin string.25–27 Branching of F-actin filamentsis mediated by the Actin-related protein-2/3 (ARP2/3) com-plex and members of the Wiscott-Aldrich Syndrome protein(WASP) family.25,26 Myosin X groups these growing fila-

ments into parallel bundles, which are cross-linked by mem-bers of the Fascin protein family.24 Coincidently, severalother proteins, including members of the Inverse (I)-BARdomain such as Insulin receptors substrate p53 (IRSp53),prepare the plasma membrane by inducing bulging of itssurface.24

The main regulators of filopodia and lamellipodia forma-tion are members of the Rho small GTPases, of which RhoA,Rac1, and Cdc42 have been most extensively studied (Figure3c).25,26 These molecules are activated by binding of GTPnucleotides, supplied by Guanine nucleotide Exchange Fac-tors (GEFs). Inactivation occurs via their intrinsic GTPaseactivity, which converts GTP into GDP; this process isstimulated by GTPase Activating Proteins (GAPs). RhoA,Rac1, and Cdc42 are activated in response to stimulation ofmany membrane receptors, including receptor tyrosinekinases and G protein–coupled receptors.26 Cdc42 and Rac1regulate filopodia and lamellipodia formation, respectively,through activation of p21-activated kinase (PAK); RhoA isinvolved in adhesion and forward movement through regula-tion of stress fiber formation via the Rho-associated serine-threonine protein kinase (ROCK; Figure 3d).25–27 PAK andROCK activate LIM-kinase, which, through inhibition ofCofilin, blocks F-actin depolymerization.26 Activated Cdc42and Rac1 also interact with the WASP:ARP2/3 complex,thereby inducing F-actin branching.25 Filopodia extension is

Lamellipodia FilopodiaStress fibers

RhoA Rac1 Cdc42

Exta Cellular Matrix

Focal contact point

ROCK PAK

LIMK

Cofilin

WASP-ARP2/3

de-polymerization

branching

RhoA Rac1 Cdc42active

small GTPase

small GTPase small GTPaseGTP

activeinactive

GDP

+GTP

Guanine nucleotideexchange factor

(GEF)

Pi

GTPaseactivating protein

(GAP)

a

F-actinG-actin

Negative end (-)

Elongation through polymerization

for filopodia extension

Retraction through de-polymerization and

retrograde flowBranching of F-actin to generate meshworks

Profilin, Formin,

Ena/Vasp

Cofilin

ARP2/3 Wasp

Positive end (+)

b

c d

Figure 3. Overview of the filopodia/la-mellipodia machinery during cell migra-tion. a, Schematic representation of fila-mentous actin in filopodia (red),lamellipodia (blue), and stress fibers(green). Cellular protrusions anchor inthe extracellular matrix by focal contactpoints (brown). b, Regulation of filamen-tous (F-actin) and globular (G-actin) actindynamics. c, Guanine nucleotideexchange factors (GEFs) and GTPaseactivating proteins (GAPs) regulate GTP-bound (active) and GDP-bound (inactive)RhoA, Rac1, and Cdc42 states, respec-tively. d, RhoA, Rac1, and Cdc42 down-stream signaling regulates actindynamics.

De Smet et al Mechanisms of Vessel Branching 641



ownloaded from


also regulated by targeting Ena/Vasp proteins to the mem-brane at the leading edge of migrating cells.26 Once extended,filopodia make contact with ECM components such aslaminin, fibronectin, and collagen through integrins on theirsurface that initiate the formation of focal contact points.27 Animportant regulator of this process is the focal adhesionkinase (FAK).28 Hence, signals from integrins and growthfactor receptors are transduced to the cytoskeleton in thesefilopodia.28

Formation and Regulation of Filopodia inEndothelial Tip CellsTo navigate, tip cells become polarized, so that their leadingfront extends filopodia, whereas their rear maintains contactwith trailing stalk cells to avoid branch desintegration. HowVEGF induces these processes is poorly understood. Basedon experiments in fibroblasts and neurons, Rho small GTPaseproteins have been identified as key regulators of cellmigration and morphogenesis.25–27 Increasing evidence sug-gests that these molecules also act downstream of VEGFR2to regulate the formation of filopodia, lamellipodia, and stressfibers, and consolidate adhesion in ECs as well. Indeed, inresponse to VEGF, Cdc42 triggers the formation of filopodiaand regulates cell polarization through microtubule organiza-tion, whereas Rac1, together with PAK, controls lamellipodiaformation in response to VEGF.29,30 Loss of Rac1 results inearly embryonic lethality attributable to defective vesselbranching.31 RhoA induces stress fiber formation and medi-ates EC permeability, migration, and stabilization of capillarytubes in response to VEGF.30,32,33 Shear stress also increasesRhoA activity, and the formation of stress fibers, focaladhesions, and junctional complexes.34 Inhibition of RhoAimpairs EC migration and tube formation, while a dominantnegative RhoA protein inhibits angiogenesis.35 As in othercells, Cdc42 activates Rac1 in ECs.30

For Rac1 to promote EC migration, it needs to be activatedby the GEF Vav2.36 In vitro, silencing of Vav2 impairsVEGF-induced EC migration, which is in line with itsfunction in neurite outgrowth,36 whereas Vav2 knockout micehave cardiovascular defects.37 Wave2, a member of theWASP proteins, is also required for actin reorganization andEC movement38: loss of Wave2 impairs vessel branching andthe formation of EC lamellipodia. Single deficiency of eachof the Ena/VASP proteins, which promote actin polymeriza-tion, results in relatively subtle phenotypes, but triple defi-ciency disrupts vessel integrity and endothelial barrier func-tion.39 Several components of the machinery responsible togenerate and consolidate focal adhesion points in other cellsis also conserved in ECs, and is, in part, mediated byintegrins.28,40 Moreover, FAK regulates EC migration inresponse to VEGF, whereas deficiency of FAK in ECs leadsto vessel defects and regression.40

Genetic studies reveal that Aquaporin-1 (AQP1), a waterchannel protein increasing water permeability in ECs, regu-lates EC migration.41 As a consequence of local actin depo-lymerization and transmembrane ionic fluxes, the cytoplasmadjacent to the leading edge of migrating cells undergoesrapid changes in osmolality. AQPs, polarized to lamellipodia,facilitate osmotic water flow across the plasma membrane in

lamellipodia. Water entry increases local hydrostatic pres-sure, producing cell membrane expansion to form a protru-sion, thereby enhancing lamellipodia dynamics.41 Accordingto this model, actin repolymerization thereafter stabilizes theprotrusion.

Modulation of VEGF Signaling in Tip CellsBesides VEGFR2, tip cells also express other VEGF recep-tors, including VEGFR3 (also known as Flt4) and NRP1. Indevelopment, VEGFR3 is present in endothelia, but becomeslargely restricted to the lymphatic endothelium in adult-hood.16 However, in active vascular endothelia, VEGFR3reappears42 and its expression is mainly confined to filopodialextensions on tip cells at the sprouting front.16 BlockingVEGFR3 reduces the number of sprouts and branch pointsand EC proliferation. VEGFR2 signaling induces VEGFR3expression in tip cells, whereas Notch downregulates itsexpression in stalk cells.16 As is the case for VEGFR2, tipcells express higher levels of VEGFR3 than stalk cells.VEGFR2 and -3 are able to form heterodimers, and maytransmit distinct signals than receptor homodimers.43 Anintriguing possibility is that such VEGFR2/3 heterodimersmay be involved in tip cell functions. VEGFR3 also regulatesvessel integrity, though it is unknown whether this is an effectmediated via tip or, more likely, via stalk cells.42

NRPs function as endothelial receptors for members of theSema3 and VEGF families.2,44 As VEGF receptors, NRPs areessential for cardiovascular development and tumor angio-genesis.44 Both NRP1 and NRP2 are detected in the devel-oping vasculature. The use of an anti-NRP1 antibody engi-neered to bind solely VEGF, and not Sema3s, has revealedthat Sema3s have little effect on VEGF/NRP1-driven vascu-lar development.44 However, Sema3F is a potent inhibitor oftumor angiogenesis, progression, and metastasis.44 NRPseither form receptor complexes with VEGFR1, -2, or -344,45

or induce direct signaling through synectin (also known asNIP), a PDZ domain-containing protein that acts as a scaffoldfor downstream signaling cascades. Genetic studies show thatNRP1 and synectin regulate vessel branching and morpho-genesis.44,46 In the absence of NRP1, sprouts in the embryonichindbrain do not branch and fuse into a vascular plexus,15

while blocking the binding of VEGF to NRP1 preventsvascular remodeling.44 The defective vessel branching andfusion were not caused by the absence of tip cells or filopodiabut were the result of defects in lateral filopodia extension,which is critical for turning and fusion of tip cells.15 HowNRP1 regulates tip cell filopodia and in particular their lateralextensions remains to be further explored. NRP1 inducesfilopodial extension and navigation of tumor cells and neu-rons, though this activity is context-dependent. In tumor cells,filopodia formation in response to NRP1/VEGFR2 signalingis mediated via Cdc42 activation.47

Novel Regulators of Tip Cell FormationOnly a few other molecules have been identified to inducefilopodia formation in ECs (Figure 2c). Inflammation isknown to activate ECs and stimulate vessel branching. Briefexposure to the inflammatory cytokine TNF� primes ECs forangiogenic sprouting by inducing a tip cell phenotype and




ownloaded from


expression of the tip cell markers PDGF-BB and VEGFR2.48

Furthermore, the Notch ligand Jagged1 was induced in tipcells via a NF-kB dependent mechanism, raising the questionwhether Jagged1 might have a similar role in tip cellinduction in pathological (inflammatory) angiogenesis asDll4 has in physiological conditions.48 Another inflammatorymediator, bradykinin, induces filopodia formation in ECsthrough activation of the Cdc42 pathway, a process which ismodulated by the transmembrane peptidase CD13.49

Sphingosine-1-phosphate (S1P), an antiinflammatory phos-pholipid derivative which has angiogenic and chemo-attractant properties,50 has also been involved in the forma-tion lamellipodia on ECs. It is released from activatedplatelets, and exerts its function on ECs via binding to itsS1P-G protein–coupled receptors (S1P1–3). S1P1 signalingis involved in the translocation of the Arp2/3 complexes froman intracellular location to the cellular migratory front,resulting in the formation of lamellipodia.50 This process isdependent on Cdc42 and Rac1 activation.50

Genetic studies in the Drosophila airway system revealthat fibroblast growth factors (FGFs) determine the specifi-cation of an epithelial tip cell.3 Indeed, the FGF homologueBranchless, which binds to its receptor Breathless on trachealcells promotes outgrowth of terminal branches. Airway tipcells respond directly to Branchless and lead branch out-growth, whereas trailing stalk cells receive a secondary signalto follow the lead cells and form a tube. These roles are notgenetically prespecified; rather, there is competition betweencells such that those with the highest FGFR activity take thetip position, whereas those with less FGFR activity assumesubsidiary positions and form the branch stalk. Competitionappears to involve Notch-mediated lateral inhibition thatprevents extra cells from assuming the lead,3 analogous to theendothelial system.1 Whether FGFR levels also determine thespecification of an endothelial tip versus stalk cell remains tobe determined.1 Interestingly, in a mixed EC population, inwhich FGF receptor-1 (FGFR1) was silenced in half of thecells, receptor-expressing cells pioneered migration in re-sponse to FGF2, whereas silenced cells were unable to takethe lead but, nonetheless, were still capable of trailing behindthe leader.51 These findings suggest that FGFRs may regulateendothelial tip cell migration as well (Figure 2b).

Angiomotin (Amot), a membrane-associated protein thatbinds angiostatin, is expressed by activated endothelia, andcontrols EC migration via its PDZ binding motif.52 Trans-genic mice expressing a PDZ-deficient Amot in ECs losetheir response to growth factors, which leads to vesseldefects.52 Amot deficiency in mouse and zebrafish alsoimpairs vessel branching and integrity, with defective filop-odia formation, and stalling of sprouting intersegmentalvessels (ISVs; Figure 2b).52 Although Amot does not affectVEGF-induced EC survival or proliferation, it renders ECsunresponsive to chemotactic stimuli. The subcellular local-ization of Amot in lamellipodia and its ability to bind cellpolarity proteins indicate that it is required for cytoskeletalreorganization during migration.52 Moreover, Syx, a RhoAGEF, was recently identified as an important PDZ interactingpartner of Amot, regulating the activity of RhoA in theleading front of the migrating cell.53

Another novel player is the Serum Response Factor (SRF)transcription factor, which interacts with other transcriptionfactors including members of the ETS and GATA families.54

Late in development, SRF expression becomes confined toECs in small vessels, more precisely in tip and stalk ECs.54

Conditional endothelial SRF-deficient embryos succumb be-cause of reduced vessel branching. This is not attributable toa reduced number of tip cells or to abnormal expression ofNotch1 or Dll4 but to the presence of thinner and fewerfilopodia per tip cell, with a disorganized actin structure at thebase of each filopodium. This abnormal actin organizationalso occurs in stalk cells, suggesting a fundamental role incytoskeletal rearrangements. In vitro, loss of SRF-functionresults in aberrant EC migration and tube-formation (Figure2b) attrituable to decreased F-actin formation. SRF has beenlinked to RhoA signaling, with which it cooperates inregulating transcription of ß-actin. Moreover, loss of SRFreduces VE-Cadherin expression, which could explain theloss of vessel integrity and stability.54 Another regulator ofEC filopodia formation is the Melanoma-associated antigen(MAGE) D1, which was initially identified as a cell surfaceantigen expressed by tumor cells. Overexpression ofMAGE-D1 inhibits EC migration and adhesion, and disruptscytoskeletal rearrangements and lamellipodia formation.55

Not surprisingly, some classical axon guidance moleculesalso regulate tip cell navigation. The Netrin guidance receptorUnc5B is also expressed by endothelial tip cells.17 Binding ofnetrin-1 to Unc5B induces collapse of filopodia in thedeveloping retina (Figure 2b), whereas knockdown of Unc5Binduces ectopic filopodial extensions.17 In Dll4 heterozygous-deficient mice, an increase in Unc5B-positive tip cells wasalso observed.18 The bidirectional signaling system of trans-membrane Eph receptors and ephrin ligands is involved inarterial/venous boundary formation in the embryo.1 Ephrin/Eph family members are also reciprocally expressed betweenblood vessels and their surrounding tissues,1 and complemen-tary ligand/receptor expression patterns provide guidancecues during vascular development, similar to the mechanismdescribed for neuronal development.2 In the process ofspinogenesis, motile dendritic filopodia explore their envi-ronment to contact with the appropriate presynaptic partner,while EphB forward signaling is required for filopodiamotility and synaptogenesis.56 All these observations raise thepossibility that Ephrin/Eph signaling or other moleculesinvolved in this neuronal process might also regulate endo-thelial filopodia formation.

The Invasiveness of the Tip CellSprouting angiogenesis is an invasive process that requiresproteolytic degradation of the extracellular matrix (ECM).Tip ECs express matrix metalloproteinases (MMPs), and theirexpression is regulated by the composition of the surroundingECM they have to invade (Figure 2b). Membrane type-1MMP (MT1-MMP; also known as MMP14) regulates angio-genesis12 and appears indispensable for ECs to form invadingchannels.33,57 There is ample evidence that MT1-MMP ispresent at the leading tip of invading ECs,12,57 while itsexpression is downregulated in stalk cells during vesselstabilization and maturation as a result of an endothelial-




ownloaded from


pericyte crosstalk.12 The role and need for MMPs in vesselbranching may vary. Indeed, vessels in the adult are embed-ded in a thick basement membrane (rich in type IV collagenand laminin), whereas the basement membrane is merely athin layer or even absent during development, which mightexplain why loss of MT1-MMP does not impair vasculardevelopment in the embryo.12 These findings thus suggestthat degradation of the basement membrane is not limitingvessel branching in the embryo, but a more critical hurdle inthe adult.

ECM components are often used by tip cells as scaffold tonavigate.58 Also, on arrest of VEGF-targeted therapy, vesselsregrow alongside ghost tracks of basement membrane.59

Integrin receptors induce cytoskeletal rearrangements, eachwith specific effects; for instance, �5ß1 and �vß3 (bothreceptors of fibronectin) differentially regulate activation ofCofilin and thereby also EC migration (Figure 2b).60 Lami-nins are heterotrimeric protein complexes, which are pro-duced by both tip and stalk cells and deposited as pericellularmatrix.61 In the absence of laminin ß1 or �1 subunits, vesselsprouts fail to form normally.61 When vessels branch, EC tipcells are exposed to different ECM components than thosepresent in the basement membrane; some of these matrixmolecules also stimulate sprout formation.58

Endothelial Stalk CellsA second endothelial subtype, called “stalk cell,” trailsbehind the leading tip cell. Their task is to proliferate,elongate the stalk, form a lumen, and connect to the circula-tion.7,11 In contrast to tip cells, stalk cells do not extendfilopodia.7

Induction of Stalk Cells by Notch and WntAs explained above, lateral inhibition via Dll4/Notch under-lies the selection of a tip cell at the expense of its neighbors,

which are instructed to become stalk cells (Figure 4a).1,19

Notch signaling in stalk cells dampens the migratory responseto VEGF, impairs filopodia extension, and, overall, inhibitsvessel branching by lowering the expression of VEGFR2,VEGFR3, NRP1, and CXCR4.16,62 Hence, inhibition ofNotch signaling increases vessel branching, by promoting aswitch from stalk to tip cell differentiation and stimulatingmigration and proliferation of tip cells, in the mouse retinaand hindbrain,18,19,63 zebrafish embryo,14 and tumor mod-els.20,64,65 Notch-regulated ankyrin-repeat protein (Nrarp) is adownstream target of Notch that counteracts Notch signalingby destabilizing NICD. Recent studies reveal that Nrarp isexpressed in stalk cells at branch points, where it overcomesthe activity of Notch to induce cell cycle arrest and quies-cence.66 At the same time, Nrarp stimulates Wnt signaling instalk cells, which stabilizes the stalk and prevents EC retrac-tion in part via tightening EC junctions.66

Specification of the tip/stalk cell phenotype by Notch islikely more complex. Indeed, Dll4 is not the only Notchligand expressed in sprouting vessels. Notably, loss of Dll4,18

Jagged1, or Dll121 each results in distinct vascular defects,indicating that these 3 ligands are not functionally redundant.Also, expression of these components is nonoverlapping inpostnatal retinal vessel development: Dll4 is the onlyligand expressed in tip cells, whereas Jagged1 and Dll1 arepresent in stalk cells.21 Soluble Jagged1 reduces tip cellnumber, filopodia, and vessels density.19,67 An outstandingquestion is whether Jagged1 instructs cells at the vascularfront to become or remain tip cells, possibly via regulatingNotch signaling in this cell. Differences in Notch signalingby distinct ligands may fine-tune vessel branching, asoccurs in other cellular systems, such as in lymphocytedevelopment.68

Dll4/Notch signaling upregulates the expression of genesthat influence EC migration and adhesion, such as VCAM1,

a b

Wnt

VEGFR2/3NRP1

Notch1

Dll4

Dll1Jagged1

?

Flt1Nrarp

Notch signaling induces a stalk cell phenotype

NICD

Cell-cellcontact

Tip cell

Stalk cellPhalanx cellVEGFEGFL7

“Push” model of vessel sprouting

EGFL7+/+ EGFL7-/-

c

Mechanisms of lumen formation

Intra-cellular Inter-cellular

pinocytosis

coalescence Figure 4. Molecular pathways of stalkcell signaling. a, A stalk cell (purple) isinduced by lateral inhibition through Dll4/Notch signaling, which reduces theexpression levels of VEGFRs and NRPand, via Nrarp, stabilizes Wnt signaling.Other Notch ligands (Dll1 and Jagged1)are also present. b, Schematic repre-sentation of intracellular and intercellularmodels of lumen formation. c, Perivas-cular EGFL7 deposition (blue) in theangiogenic sprout regulates vascularmorphogenesis.




ownloaded from


the RhoGTPase RND1, or the RasGEF RAPGEF5.69 Dll4/Notch also promotes cell adhesion by activating ß1-integrinsthrough NICD transcription-independent mechanisms.70

Further, Notch1 stimulation by Jagged1 induces microtu-bule stabilization in neurons,71 while promoting cell– cellover cell–ECM interactions in fibroblasts72; if this wouldbe also the case in ECs, Notch would hereby promote stalkstabilization and prevent stalk cell retraction. Interestingly,the ECM protein Microfibril-Associated Glycoprotein-2(MAGP-2) also induces actin rearrangements and vesselbranching through antagonizing Notch signaling in ECs.73

Fine Tuning of Stalk Morphogenesis by sFlt1VEGFR1 (Flt1) and its soluble form, sFlt1, act as a trap forVEGF, VEGF-B, and PlGF.74 Both endothelial tip and stalkcells express VEGFR1 during retinal vascular development.7

VEGFR1 and its soluble form are upregulated on Dll4-acti-vated Notch-signaling in stalk cells,69 yet neutralization ofVEGFR1 does not affect vessel branching in vivo.7 Also,�-secretase, which proteolytically activates Notch in stalkcells, cleaves membrane-anchored VEGFR1: release of itsintracellular domain has been proposed to inhibit VEGFR2signaling, while shedding of its extracellular sFlt1 domainmay further dampen VEGF by acting as a trap.75 All theseeffects may potentially reduce or spatially restrict the overallresponse of ECs to VEGF through sequestration of extracel-lular VEGF. Although PlGF may activate ECs via signalingthrough membrane-anchored VEGFR1 in pathological con-ditions,74 it remains to be defined whether this involves aneffect on tip and/or stalk cells.

Stalk Cells Maintain Vessel IntegrityTo elongate the stalk, stalk cells must divide, maintaincontact with the leading tip cell, and form a lumen. Similar totip cells, maintenance of a stalk cell phenotype requirescytoskeletal restructuring. However, the molecular (down-stream) mechanisms of how trailing cells remain in contactwith the leading front remain mostly unexplored. VE-Cadherin is important to maintain cell–cell contacts, as itsabsence induces random nondirectional migration of discon-nected cells.51 Cadherins do not form rigid and fixed struc-tures, but rather exhibit a flow-like movement on reorgani-zation of the cytoskeleton.76 Indeed, when epithelial cellsmove, cadherin clusters rapidly form at the leading edge,while they are absorbed at the rear end. As mentioned above,Nrarp-induced Wnt signaling is also important for stabiliza-tion of the stalk and preventing EC retraction by improvingintercellular junctions.66 FGF signaling has also been impli-cated in vascular homeostasis and integrity through tighten-ing of EC junctions.77 Besides the above described “pull”model, whereby the tip cell pulls stalk cells in the sprout,recent data also provide evidence for a “push” model (Figure4c), whereby dividing stalk cells push the sprout forward.Indeed, in the absence of the vascular-specific secreted factorEGFL7, an ECM protein mainly secreted by stalk cells,newly formed ECs at the base of the stalk accumulate inenlarged, but nonelongating sprouts.78

Robo4 Is Necessary for Stalk Cell StabilizationOriginally discovered in Drosophila neurons as receptors foraxon guidance cues, members of the Roundabout (Robo)receptor family are also expressed by ECs. The Slit/Robosystem in vertebrates consists of three Slit proteins (Slit1–3),which bind to Robo receptors (Robo1–4). Robo4 is structur-ally divergent from other Robo receptors and primarilyexpressed by ECs, in particular in microvessels.79 Robo4activates pathways involved in cell migration,79 includingWASP and Ena/Vasp family members in ECs.80 The role ofRobo4 in EC migration is, however, debated. Some reportsdescribe a role in repulsion,81,82 whereas others suggest apromigratory effect.83,84 Genetic studies further show thatRobo4 stabilizes nascent vessels during postnatal retinalangiogenesis.85 Robo4 is absent from most tip cells, butexpressed by stalk cells, where activation by Slit2 woulddownregulate VEGF signaling and inhibit activation of Src-kinase and Rac1 in this model. Also, Slit2 inhibits ECmigration, tube formation, and permeability, whereas vesselsin the ischemic retina are more leaky and unstable in Robo4mutant mice.85 However, it remains debated whether Slit2 isthe ligand of Robo4. Indeed, overexpression studies showthat Slit2 can be immunoprecipitated with Robo4,81 but othersdid not observe binding of Slit2 to Robo4.83 Recently, Robo4was reported to form heterodimers with Robo1, which in-creases the formation of filopodia in ECs.80 Thus, unravelingthe role of Robo-signaling in tip/stalk cells requires furtherinvestigation.

Stalk Cells and Lumen FormationAfter tip cell induction and stalk elongation, a vessel branchneeds to generate a lumen and initiate blood flow. Most of ourconceptual advances in lumen-formation have been generatedin vitro, by studying the behavior of ECs in two- or three-dimensional gels.33 In these models, lumenogenesis relies onthe formation and coalescence of intracellular pinocyticvacuoles. In zebrafish embryos, intra- and intercellular fusionof vacuoles generates a lumen (Figure 4b).86 However, recentfindings suggest that the lumen in zebrafish vessels is formedby the arrangement of ECs around an intercellular lumen.13

Vacuole formation depends on an interaction of integrinswith the ECM, whereas fusion of vacuoles requires rearrange-ment of the actin and microtubule cytoskeleton.33 EC vacu-oles accumulate in a polarized pattern, adjacent to thecentrosome.87 Both Rac1 and Cdc42, but not RhoA, localizeto vacuole membranes during formation of a lumen.33,87

Downstream signaling includes members of the WASP,PAK, polarity proteins, and protein kinase C (PKC).33,87

However, an inhibitor of ROCK activity (which is down-stream of RhoA) impairs vacuole formation,88 suggesting thatin some conditions, RhoA may regulate lumen formation.Additionally, activation of endothelial RhoA-GTP on loss ofCerebral Cavernous Malformations 2 (CCM2) causes cy-toskeletal changes and impairs lumen formation.89 Once thelumen has formed, additional signals drive EC proliferation,growth, or enlargement, modifying the lumen size. Sucheffects have been reported for different isoforms of VEGF,11

Notch1,67 and laminin.61




ownloaded from


Endothelial Phalanx CellsOnce a vessel branch is formed, ECs become quiescent, fromwhich only 0.01% are dividing in the healthy adult. The keyfunction of vessels is then to supply blood and oxygen totissues. While blood flow itself is already an important factorfor keeping ECs quiescent,34 some of the molecules, deter-mining this endothelial phenotype were recently identified bygenetic studies. Indeed, endothelial-specific heterozygousdeletion of prolyl-hydroxylase domain-2 (PHD2), an oxygensensor, leads to “endothelial normalization” of tumor vessels.In contrast to wild-type tumor vessels (Figure 5a), where ECsare hyper-activated, extend multiple filopodia and loose tightcell-cell contacts, ECs in PHD2 haplodeficient mice werealigned in a smooth tight cobblestone monolayer, resemblingthe “phalanx formation” of ancient Greek soldiers, hencetheir name “phalanx cells” (Figure 5b).5 A shift of tumor ECsto phalanx cells improves cancer perfusion and oxygenation,thereby almost completely preventing metastasis and induc-ing a shift from a malignant to a more benign tumor

behavior.5 Hence, by lowering the activity of an oxygensensor, ECs readjust their shape and phenotype to improveoxygen delivery in case of its shortage.

The molecular mechanisms underlying the endothelialphalanx phenotype remain to be largely explored. Unlike tipcells, phalanx cells extend few filopodia and migrate poorlyin response to VEGF, but form a tight barrier. They resemblestalk cells by depositing a basement membrane and establish-ing junctions, but differ from these cells by their increasedquiescence, and reduced mitogenic response to VEGF. Thetip/stalk cell model of vessel branching suggests that highlevels of VEGF stimulate the induction of a migratory EC tipcell phenotype, while intermediate levels promote the prolif-erative stalk cell phenotype. Genetic studies show that loss ofVEGF in ECs causes widespread EC dysfunction and desin-tegration, indicating that low levels of VEGF are critical forquiescent ECs to survive.90 It is also known that the junctionalmolecule VE-Cadherin shifts the EC response to VEGF fromproliferation and migration to survival and quiescence.91

Interestingly, phalanx cells in PHD2 haplodeficient mice aremore quiescent by expressing elevated levels of VE-Cadherinand sFlt1, which acts as a VEGF trap to counteract the tipcell-inducing activity of VEGF (Figure 5c).5

Additional molecular pathways have been impli-cated in EC quiescence and survival. Activation ofphosphatidylinositol-3-kinase (PI3K) and protein kinase B(PKB)/Akt signaling promotes EC survival in response tovarious signals, including VEGF, FGFs, Ang1 (Figure 5c),and insulin-like growth factor (IGF).77,92–94 Furthermore, inconfluent (quiescent) ECs, Ang1 induces Tie2 translocationto cell–cell contacts and the formation of homotypic Tie2–Tie2 transassociated complexes that include the vascularendothelial phosphotyrosine phosphatase, leading to inhibi-tion of paracellular permeability (Figure 5c).95 Moreover,FGF receptor inhibition disrupts endothelial integrity anddissolves interendothelial junctions.77 Also, bone-morphogenic protein (BMP)-9, and its ALK1 receptor, andpigment epithelium derived factor (PEDF) were recentlyidentified to promote EC quiescence.96,97 Interaction ofthrombospondin (TSP)-1 and -2 with their CD36 receptoralso diverge the proliferative and migratory stimuli of angio-genic factors such as VEGF into survival signals (Figure5c).98 Additionally, a number of molecular pathways havebeen recently shown to induce tumor vessel normalization,such as the formation of perivascular nitric oxide gradients,but whether this involves a shift to a phalanx EC phenotype,remains to be explored.99

ConclusionThe above proposed model of vessel branching is largelybased on seminal insights on tip, stalk, and phalanx cells fromthe last 5 years. These studies have, however, also raised anumber of outstanding questions. For instance, are thesedistinct EC phenotypes interchangeable, and what are the keymolecular determinants of this combinatorial code? Arerelative rather than absolute expression levels, or activationstatus, of membrane receptors more relevant to specify ECphenotypes? How do stalk cells maintain contact with theleading tip cell and trail behind without dissolving the

Figure 5

VEGF

QuiescenceSurvival

Shear stress

Blo

od fl

ow

sFlt

VEGFR2

VEGFR1

PHD2+/-

low O2

FGF2

FGFR

VE-Cadherin

c

a bPhd2+/+ Phd2+/-

Signaling pathways to maintain a phalanx cell phenotype

VE-Cadherin

migration,proliferation

TIE2

TIE2Ang

PEDFBMP9TSP

Figure 5. Molecular pathways of phalanx cell signaling. a and b,Scanning electron microscopic images of the inner wall of tumorvessels grown in wild-type (a) or PHD2�/� (b) mice show“abnormalized” versus “normalized” phalanx ECs. c, Signalingmechanisms regulating phalanx cells quiescence.




ownloaded from


elongating stalk? What is the molecular basis of lumenformation? How do tip cells recognize their counterparts andfuse with each other? What is the molecular code of thephalanx phenotype? Is hypoxia codetermining the specifica-tion of EC fates? What is the full implication of Wnt-signaling? How do endothelial filopodia sense their environ-ment, are guidance signals involved? Future explorations oftip, stalk, and phalanx cell behavior will benefit from im-proved in vivo imaging such as by fluorescence nanos-copy.100 Finally, providing an answer to those questions willbe useful to translate these concepts into the development ofnovel pro- and antiangiogenic therapies.

Sources of FundingP.C. is supported by long-term structural funding (Methusalemfunding by the Flemish Government), Interuniversity Attraction Pole(grant P60/30, funded by Belgian Government, BELSPO), FWOG.0692.09 (Flemish Government), research grant by the Belgian“Foundation against Cancer,” GOA 2006/11 - KU Leuven. P.J.H. isrecipient of a FEBS long-term fellowship. I.S. is recipient of a MarieCurie Fellowship.

DisclosuresNone.

References1. Horowitz A, Simons M. Branching morphogenesis. Circ Res. 2008;103:

784–795.2. Carmeliet P, Tessier-Lavigne M. Common mechanisms of nerve and

blood vessel wiring. Nature. 2005;436:193–200.3. Ghabrial AS, Krasnow MA. Social interactions among epithelial cells

during tracheal branching morphogenesis. Nature. 2006;441:746–749.4. Gasparini F, Longo F, Manni L, Burighel P, Zaniolo G. Tubular

sprouting as a mode of vascular formation in a colonial ascidian(Tunicata). Dev Dyn. 2007;236:719–731.

5. Mazzone M, Dettori D, Leite de Oliveira R, Schmidt T, Lanahan A,Jonckx B, De Smet F, Aragones A, Vinckier S, Loges S, Luttun A,Wyns S, Jordan B, Pisacane A, Gallez A, Lampugnani MG, Dejana ERP, Simons M, Maxwell P, Carmeliet P. Heterozygous deficiency ofPHD2 restores tumor oxygenation and inhibits metastasis via endothelialnormalization. Cell. 2009;136:839–851.

6. Klosovskii BN, Zhukova TP. [Effect of colchicine on remote phases ofgrowing capillaries in the brain.]. Arkh Patol. 1963;35:38–44.

7. Gerhardt H, Golding M, Fruttiger M, Ruhrberg C, Lundkvist A,Abramsson A, Jeltsch M, Mitchell C, Alitalo K, Shima D, Betsholtz C.VEGF guides angiogenic sprouting utilizing endothelial tip cellfilopodia. J Cell Biol. 2003;161:1163–1177.

8. Dorrell MI, Aguilar E, Friedlander M. Retinal vascular development ismediated by endothelial filopodia, a preexisting astrocytic template andspecific R-cadherin adhesion. Invest Ophthalmol Visual Sci. 2002;43:3500–3510.

9. Wolff JR, Bar T. ‘Seamless’ endothelia in brain capillaries duringdevelopment of the rat’s cerebral cortex. Brain Res. 1972;41:17–24.

10. Marin-Padilla M. Early vascularization of the embryonic cerebralcortex: Golgi and electron microscopic studies. J Comp Neurol. 1985;241:237–249.

11. Ruhrberg C, Gerhardt H, Golding M, Watson R, Ioannidou S, FujisawaH, Betsholtz C, Shima DT. Spatially restricted patterning cues providedby heparin-binding VEGF-A control blood vessel branching morpho-genesis. Genes Dev. 2002;16:2684–2698.

12. Yana I, Sagara H, Takaki S, Takatsu K, Nakamura K, Nakao K, KatsukiM, Taniguchi S, Aoki T, Sato H, Weiss SJ, Seiki M. Crosstalk betweenneovessels and mural cells directs the site-specific expression ofMT1-MMP to endothelial tip cells. J Cell Sci. 2007;120(Pt9):1607–1614.

13. Blum Y, Belting HG, Ellertsdottir E, Herwig L, Luders F, Affolter M.Complex cell rearrangements during intersegmental vessel sproutingand vessel fusion in the zebrafish embryo. Dev Biol. 2008;316:312–322.

14. Siekmann AF, Lawson ND. Notch signalling limits angiogenic cellbehaviour in developing zebrafish arteries. Nature. 2007;445:781–784.

15. Gerhardt H, Ruhrberg C, Abramsson A, Fujisawa H, Shima D, BetsholtzC. Neuropilin-1 is required for endothelial tip cell guidance in thedeveloping central nervous system. Dev Dyn. 2004;231:503–509.

16. Tammela T, Zarkada G, Wallgard E, Murtomaki A, Suchting S,Wirzenius M, Waltari M, Hellstrom M, Schomber T, Peltonen R, FreitasC, Duarte A, Isoniemi H, Laakkonen P, Christofori G, Yla-Herttuala S,Shibuya M, Pytowski B, Eichmann A, Betsholtz C, Alitalo K. BlockingVEGFR-3 suppresses angiogenic sprouting and vascular network for-mation. Nature. 2008;454:656–660.

17. Lu X, Le Noble F, Yuan L, Jiang Q, De Lafarge B, Sugiyama D, BreantC, Claes F, De Smet F, Thomas JL, Autiero M, Carmeliet P, Tessier-Lavigne M, Eichmann A. The netrin receptor UNC5B mediatesguidance events controlling morphogenesis of the vascular system.Nature. 2004;432:179–186.

18. Suchting S, Freitas C, le Noble F, Benedito R, Breant C, Duarte A,Eichmann A. The Notch ligand Delta-like 4 negatively regulates endo-thelial tip cell formation and vessel branching. Proc Natl Acad SciU S A. 2007;104:3225–3230.

19. Hellstrom M, Phng LK, Hofmann JJ, Wallgard E, Coultas L, LindblomP, Alva J, Nilsson AK, Karlsson L, Gaiano N, Yoon K, Rossant J,Iruela-Arispe ML, Kalen M, Gerhardt H, Betsholtz C. Dll4 signallingthrough Notch1 regulates formation of tip cells during angiogenesis.Nature. 2007;445:776–780.

20. Noguera-Troise I, Daly C, Papadopoulos NJ, Coetzee S, Boland P, GaleNW, Lin HC, Yancopoulos GD, Thurston G. Blockade of Dll4 inhibitstumour growth by promoting non-productive angiogenesis. Nature.2006;444:1032–1037.

21. Roca C, Adams RH. Regulation of vascular morphogenesis by Notchsignaling. Genes Dev. 2007;21:2511–2524.

22. Matsuda M, Chitnis AB. Interaction with Notch determines endocytosisof specific Delta ligands in zebrafish neural tissue. Development (Cam-bridge, England). 2009;136:197–206.

23. Estrach S, Legg J, Watt FM. Syntenin mediates Delta1-induced cohe-siveness of epidermal stem cells in culture. J Cell Sci. 2007;120(Pt16):2944–2952.

24. Mattila PK, Lappalainen P. Filopodia: molecular architecture andcellular functions. Nat Rev Mol Cell Biol. 2008;9:446–454.

25. Pollard TD, Borisy GG. Cellular motility driven by assembly and dis-assembly of actin filaments. Cell. 2003;112:453–465.

26. Huber AB, Kolodkin AL, Ginty DD, Cloutier JF. Signaling at thegrowth cone: ligand-receptor complexes and the control of axon growthand guidance. Annu Rev Neurosci. 2003;26:509–563.

27. Defilippi P, Olivo C, Venturino M, Dolce L, Silengo L, Tarone G. Actincytoskeleton organization in response to integrin-mediated adhesion.Microscopy Res Tech. 1999;47:67–78.

28. Mitra SK, Hanson DA, Schlaepfer DD. Focal adhesion kinase: incommand and control of cell motility. Nat Rev Mol Cell Biol. 2005;6:56–68.

29. Lamalice L, Houle F, Jourdan G, Huot J. Phosphorylation of tyrosine1214 on VEGFR2 is required for VEGF-induced activation of Cdc42upstream of SAPK2/p38. Oncogene. 2004;23:434–445.

30. Ispanovic E, Serio D, Haas TL. Cdc42 and RhoA have opposing roles inregulating membrane type 1-matrix metalloproteinase localization andmatrix metalloproteinase-2 activation. Am J Physiol Cell Physiol. 2008;295:C600–C610.

31. Tan W, Palmby TR, Gavard J, Amornphimoltham P, Zheng Y, GutkindJS. An essential role for Rac1 in endothelial cell function and vasculardevelopment. Faseb J. 2008;22:1829–1838.

32. van Nieuw Amerongen GP, Koolwijk P, Versteilen A, van HinsberghVW. Involvement of RhoA/Rho kinase signaling in VEGF-inducedendothelial cell migration and angiogenesis in vitro. ArteriosclerThromb Vasc Biol. 2003;23:211–217.

33. Davis GE, Bayless KJ, Mavila A. Molecular basis of endothelial cellmorphogenesis in three-dimensional extracellular matrices. Anatom Rec-ord. 2002;268:252–275.

34. Shiu YT, Li S, Marganski WA, Usami S, Schwartz MA, Wang YL,Dembo M, Chien S. Rho mediates the shear-enhancement of endothelialcell migration and traction force generation. Biophys J. 2004;86:2558–2565.

35. Hoang MV, Whelan MC, Senger DR. Rho activity critically and selec-tively regulates endothelial cell organization during angiogenesis. ProcNatl Acad Sci U S A. 2004;101:1874–1879.

36. Garrett TA, Van Buul JD, Burridge K. VEGF-induced Rac1 activationin endothelial cells is regulated by the guanine nucleotide exchangefactor Vav2. Exp Cell Res. 2007;313:3285–3297.




ownloaded from


37. Sauzeau V, Jerkic M, Lopez-Novoa JM, Bustelo XR. Loss of Vav2proto-oncogene causes tachycardia and cardiovascular disease in mice.Mol Biol Cell. 2007;18:943–952.

38. Yamazaki D, Suetsugu S, Miki H, Kataoka Y, Nishikawa S, Fujiwara T,Yoshida N, Takenawa T. WAVE2 is required for directed cell migrationand cardiovascular development. Nature. 2003;424:452–456.

39. Furman C, Sieminski AL, Kwiatkowski AV, Rubinson DA, Vasile E,Bronson RT, Fassler R, Gertler FB. Ena/VASP is required for endothe-lial barrier function in vivo. J Cell Biol. 2007;179:761–775.

40. Braren R, Hu H, Kim YH, Beggs HE, Reichardt LF, Wang R. Endo-thelial FAK is essential for vascular network stability, cell survival, andlamellipodial formation. J Cell Biol. 2006;172:151–162.

41. Saadoun S, Papadopoulos MC, Hara-Chikuma M, Verkman AS.Impairment of angiogenesis and cell migration by targeted aquaporin-1gene disruption. Nature. 2005;434:786–792.

42. Kubo H, Fujiwara T, Jussila L, Hashi H, Ogawa M, Shimizu K, AwaneM, Sakai Y, Takabayashi A, Alitalo K, Yamaoka Y, Nishikawa SI.Involvement of vascular endothelial growth factor receptor-3 in main-tenance of integrity of endothelial cell lining during tumor angiogenesis.Blood. 2000;96:546–553.

43. Dixelius J, Makinen T, Wirzenius M, Karkkainen MJ, Wernstedt C,Alitalo K, Claesson-Welsh L. Ligand-induced vascular endothelialgrowth factor receptor-3 (VEGFR-3) heterodimerization with VEGFR-2in primary lymphatic endothelial cells regulates tyrosine phosphoryla-tion sites. J Biol Chem. 2003;278:40973–40979.

44. Geretti E, Shimizu A, Klagsbrun M. Neuropilin structure governs VEGFand semaphorin binding and regulates angiogenesis. Angiogenesis.2008;11:31–39.

45. Favier B, Alam A, Barron P, Bonnin J, Laboudie P, Fons P, Mandron M,Herault JP, Neufeld G, Savi P, Herbert JM, Bono F. Neuropilin-2interacts with VEGFR-2 and VEGFR-3 and promotes human endothelialcell survival and migration. Blood. 2006;108:1243–1250.

46. Chittenden TW, Claes F, Lanahan AA, Autiero M, Palac RT, TkachenkoEV, Elfenbein A, Ruiz de Almodovar C, Dedkov E, Tomanek R, Li W,Westmore M, Singh JP, Horowitz A, Mulligan-Kehoe MJ, Moodie KL,Zhuang ZW, Carmeliet P, Simons M. Selective regulation of arterialbranching morphogenesis by synectin. Dev Cell. 2006;10:783–795.

47. Hong TM, Chen YL, Wu YY, Yuan A, Chao YC, Chung YC, Wu MH,Yang SC, Pan SH, Shih JY, Chan WK, Yang PC. Targeting neuropilin1 as an antitumor strategy in lung cancer. Clin Cancer Res. 2007;13:4759–4768.

48. Sainson RC, Johnston DA, Chu HC, Holderfield MT, Nakatsu MN,Crampton SP, Davis J, Conn E, Hughes CC. TNF primes endothelialcells for angiogenic sprouting by inducing a tip cell phenotype. Blood.2008;111:4997–5007.

49. Petrovic N, Schacke W, Gahagan JR, O’Conor CA, Winnicka B,Conway RE, Mina-Osorio P, Shapiro LH. CD13/APN regulates endo-thelial invasion and filopodia formation. Blood. 2007;110:142–150.

50. Lee JF, Ozaki H, Zhan X, Wang E, Hla T, Lee MJ. Sphingosine-1-phosphate signaling regulates lamellipodia localization of cortactin com-plexes in endothelial cells. Histochem Cell Biol. 2006;126:297–304.

51. Vitorino P, Meyer T. Modular control of endothelial sheet migration.Genes Dev. 2008;22:3268–3281.

52. Aase K, Ernkvist M, Ebarasi L, Jakobsson L, Majumdar A, Yi C, BirotO, Ming Y, Kvanta A, Edholm D, Aspenstrom P, Kissil J,Claesson-Welsh L, Shimono A, Holmgren L. Angiomotin regulatesendothelial cell migration during embryonic angiogenesis. Genes Dev.2007;21:2055–2068.

53. Ernkvist M, Persson NL, Audebert S, Lecine P, Sinha I, Liu M,Schlueter M, Horowitz A, Aase K, Weide T, Borg JP, Majumdar A,Holmgren L. The Amot/Patj/Syx signaling complex spatially controlsRhoA GTPase activity in migrating endothelial cells. Blood. 2009;113:244–253.

54. Franco CA, Mericskay M, Parlakian A, Gary-Bobo G, Gao-Li J, PaulinD, Gustafsson E, Li Z. Serum response factor is required for sproutingangiogenesis and vascular integrity. Dev Cell. 2008;15:448–461.

55. Shen WG, Xue QY, Zhu J, Hu BS, Zhang Y, Wu YD, Su Q. Inhibitionof adenovirus-mediated human MAGE-D1 on angiogenesis in vitro andin vivo. Mol Cell Biochem. 300(1–2):89–99, 2007.

56. Kayser MS, Nolt MJ, Dalva MB. EphB receptors couple dendriticfilopodia motility to synapse formation. Neuron. 2008;59:56–69.

57. van Hinsbergh VW, Koolwijk P. Endothelial sprouting and angio-genesis: matrix metalloproteinases in the lead. Cardiovasc Res. 2008;78:203–212.

58. Rhodes JM, Simons M. The extracellular matrix and blood vesselformation: not just a scaffold. J Cell Mol Med. 2007;11:176–205.

59. Mancuso MR, Davis R, Norberg SM, O’Brien S, Sennino B, NakaharaT, Yao VJ, Inai T, Brooks P, Freimark B, Shalinsky DR, Hu-Lowe DD,McDonald DM. Rapid vascular regrowth in tumors after reversal ofVEGF inhibition. J Clin Invest. 2006;116:2610–2621.

60. Danen EH, van Rheenen J, Franken W, Huveneers S, Sonneveld P,Jalink K, Sonnenberg A. Integrins control motile strategy through aRho-cofilin pathway. J Cell Biol. 2005;169:515–526.

61. Jakobsson L, Domogatskaya A, Tryggvason K, Edgar D,Claesson-Welsh L. Laminin deposition is dispensable for vasculogenesisbut regulates blood vessel diameter independent of flow. Faseb J.2008;22:1530–1539.

62. Williams CK, Segarra M, Sierra Mde L, Sainson RC, Tosato G, HarrisAL. Regulation of CXCR4 by the Notch ligand delta-like 4 in endothe-lial cells. Cancer Res. 2008;68:1889–1895.

63. Lobov IB, Renard RA, Papadopoulos N, Gale NW, Thurston G, Yan-copoulos GD, Wiegand SJ. Delta-like ligand 4 (Dll4) is induced byVEGF as a negative regulator of angiogenic sprouting. Proc Natl AcadSci U S A. 2007;104:3219–3224.

64. Ridgway J, Zhang G, Wu Y, Stawicki S, Liang WC, Chanthery Y,Kowalski J, Watts RJ, Callahan C, Kasman I, Singh M, Chien M, TanC, Hongo JA, de Sauvage F, Plowman G, Yan M. Inhibition of Dll4signalling inhibits tumour growth by deregulating angiogenesis. Nature.2006;444:1083–1087.

65. Segarra M, Williams CK, Sierra Mde L, Bernardo M, McCormick PJ,Maric D, Regino C, Choyke P, Tosato G. Dll4 activation of Notchsignaling reduces tumor vascularity and inhibits tumor growth. Blood.2008;112:1904–1911.

66. Phng LK, Potente M, Leslie JD, Babbage J, Nyqvist D, Lobov I, OndrJK, Rao S, Lang RA, Thurston G, Gerhardt H. Nrarp coordinatesendothelial Notch and Wnt signaling to control vessel density in angio-genesis. Dev Cell. 2009;16:70–82.

67. Sainson RC, Aoto J, Nakatsu MN, Holderfield M, Conn E, Koller E,Hughes CC. Cell-autonomous notch signaling regulates endothelial cellbranching and proliferation during vascular tubulogenesis. Faseb J.2005;19:1027–1029.

68. Rutz S, Mordmuller B, Sakano S, Scheffold A. Notch ligands Delta-like1, Delta-like4 and Jagged1 differentially regulate activation of pe-ripheral T helper cells. Eur J Immunol. 2005;35:2443–2451.

69. Harrington LS, Sainson RC, Williams CK, Taylor JM, Shi W, Li JL,Harris AL. Regulation of multiple angiogenic pathways by Dll4 andNotch in human umbilical vein endothelial cells. Microvascular Res.2008;75:144–154.

70. Hodkinson PS, Elliott PA, Lad Y, McHugh BJ, MacKinnon AC, HaslettC, Sethi T. Mammalian NOTCH-1 activates beta1 integrins via the smallGTPase R-Ras. J Biol Chem. 2007;282:28991–29001.

71. Ferrari-Toninelli G, Bonini SA, Bettinsoli P, Uberti D, Memo M. Micro-tubule stabilizing effect of notch activation in primary cortical neurons.Neuroscience. 2008;154:946–952.

72. Lindner V, Booth C, Prudovsky I, Small D, Maciag T, Liaw L. Membersof the Jagged/Notch gene families are expressed in injured arteries andregulate cell phenotype via alterations in cell matrix and cell-cell inter-action. Am J Pathol. 2001;159:875–883.

73. Albig AR, Becenti DJ, Roy TG, Schiemann WP. Microfibril-associateglycoprotein-2 (MAGP-2) promotes angiogenic cell sprouting byblocking notch signaling in endothelial cells. Microvascular Res. 2008;76:7–14.

74. Fischer C, Mazzone M, Jonckx B, Carmeliet P. FLT1 and its ligandsVEGFB and PlGF: drug targets for anti-angiogenic therapy? Nat Rev.2008;8:942–956.

75. Boulton ME, Cai J, Grant MB. gamma-Secretase: a multifaceted reg-ulator of angiogenesis. J Cell Mol Med. 2008;12:781–795.

76. Kametani Y, Takeichi M. Basal-to-apical cadherin flow at cell junctions.Nat Cell Biol. 2007;9:92–98.

77. Murakami M, Nguyen LT, Zhang ZW, Moodie KL, Carmeliet P, StanRV, Simons M. The FGF system has a key role in regulating vascularintegrity. J Clin Invest. 2008;118:3355–3366.

78. Schmidt M, Paes K, De Maziere A, Smyczek T, Yang S, Gray A, FrenchD, Kasman I, Klumperman J, Rice DS, Ye W. EGFL7 regulates thecollective migration of endothelial cells by restricting their spatial dis-tribution. Development (Cambridge, England). 2007;134:2913–2923.

79. Legg JA, Herbert JM, Clissold P, Bicknell R. Slits and Roundabouts incancer, tumour angiogenesis and endothelial cell migration. Angio-genesis. 2008;11:13–21.




ownloaded from


80. Sheldon H, Andre M, Legg JA, Heal P, Herbert JM, Sainson R, SharmaAS, Kitajewski JK, Heath VL, Bicknell R. Active involvement of Robo1and Robo4 in filopodia formation and endothelial cell motility mediatedvia WASP and other actin nucleation-promoting factors. Faseb J. 2009;23:513–522.

81. Park KW, Morrison CM, Sorensen LK, Jones CA, Rao Y, Chien CB,Wu JY, Urness LD, Li DY. Robo4 is a vascular-specific receptor thatinhibits endothelial migration. Dev Biol. 2003;261:251–267.

82. Seth P, Lin Y, Hanai J, Shivalingappa V, Duyao MP, Sukhatme VP.Magic roundabout, a tumor endothelial marker: expression and sig-naling. Biochem Biophys Res Commun. 2005;332:533–541.

83. Suchting S, Heal P, Tahtis K, Stewart LM, Bicknell R. Soluble Robo4receptor inhibits in vivo angiogenesis and endothelial cell migration.Faseb J. 2005;19:121–123.

84. Kaur S, Castellone MD, Bedell VM, Konar M, Gutkind JS, Ram-chandran R. Robo4 signaling in endothelial cells implies attractionguidance mechanisms. J Biol Chem. 2006;281:11347–11356.

85. Jones CA, London NR, Chen H, Park KW, Sauvaget D, Stockton RA,Wythe JD, Suh W, Larrieu-Lahargue F, Mukouyama YS, Lindblom P,Seth P, Frias A, Nishiya N, Ginsberg MH, Gerhardt H, Zhang K, Li DY.Robo4 stabilizes the vascular network by inhibiting pathologic angio-genesis and endothelial hyperpermeability. Nat Med. 2008;14:448–453.

86. Kamei M, Saunders WB, Bayless KJ, Dye L, Davis GE, Weinstein BM.Endothelial tubes assemble from intracellular vacuoles in vivo. Nature.2006;442:453–456.

87. Koh W, Mahan RD, Davis GE. Cdc42- and Rac1-mediated endotheliallumen formation requires Pak2, Pak4 and Par3, and PKC-dependentsignaling. J Cell Sci. 2008;121:989–1001.

88. Somlyo AV, Phelps C, Dipierro C, Eto M, Read P, Barrett M, Gibson JJ,Burnitz MC, Myers C, Somlyo AP. Rho kinase and matrix metallopro-teinase inhibitors cooperate to inhibit angiogenesis and growth of humanprostate cancer xenotransplants. Faseb J. 2003;17:223–234.

89. Whitehead KJ, Chan AC, Navankasattusas S, Koh W, London NR, LingJ, Mayo AH, Drakos SG, Marchuk DA, Davis GE, Li DY. The cerebral

cavernous malformation signaling pathway promotes vascular integrityvia Rho GTPases. Nat Med. 2009.

90. Lee S, Chen TT, Barber CL, Jordan MC, Murdock J, Desai S, Ferrara N,Nagy A, Roos KP, Iruela-Arispe ML. Autocrine VEGF signaling isrequired for vascular homeostasis. Cell. 2007;130:691–703.

91. Wallez Y, Vilgrain I, Huber P. Angiogenesis: the VE-cadherin switch.Trends Cardiovasc Med. 2006;16:55–59.

92. Makinde T, Agrawal DK. Intra and extravascular transmembrane sig-nalling of angiopoietin-1-Tie2 receptor in health and disease. J Cell MolMed. 2008;12:810–828.

93. Conti E, Carrozza C, Capoluongo E, Volpe M, Crea F, Zuppi C,Andreotti F. Insulin-like growth factor-1 as a vascular protective factor.Circulation. 2004;110:2260–2265.

94. Olsson AK, Dimberg A, Kreuger J, Claesson-Welsh L. VEGF receptorsignalling - in control of vascular function. Nat Rev Mol Cell Biol.2006;7:359–371.

95. Saharinen P, Eklund L, Miettinen J, Wirkkala R, Anisimov A, Wind-erlich M, Nottebaum A, Vestweber D, Deutsch U, Koh GY, Olsen BR,Alitalo K. Angiopoietins assemble distinct Tie2 signalling complexes inendothelial cell-cell and cell-matrix contacts. Nat Cell Biol. 2008;10:527–537.

96. David L, Mallet C, Keramidas M, Lamande N, Gasc JM, Dupuis-GirodS, Plauchu H, Feige JJ, Bailly S. Bone morphogenetic protein-9 is acirculating vascular quiescence factor. Circ Res. 2008;102:914–922.

97. Pollina EA, Legesse-Miller A, Haley EM, Goodpaster T, Randolph-Habecker J, Coller HA. Regulating the angiogenic balance in tissues.Cell Cycle (Georgetown, Tex). 2008;7:2056–2070.

98. Silverstein RL, Febbraio M. CD36-TSP-HRGP interactions in the reg-ulation of angiogenesis. Curr Pharmaceut design. 2007;13:3559–3567.

99. Kashiwagi S, Tsukada K, Xu L, Miyazaki J, Kozin SV, Tyrrell JA, SessaWC, Gerweck LE, Jain RK, Fukumura D. Perivascular nitric oxidegradients normalize tumor vasculature. Nat Med. 2008;14:255–257.

100. Folling J, Bossi M, Bock H, Medda R, Wurm CA, Hein B, Jakobs S,Eggeling C, Hell SW. Fluorescence nanoscopy by ground-statedepletion and single-molecule return. Nat Meth. 2008;5:943–945.




ownloaded from


CarmelietFrederik De Smet, Inmaculada Segura, Katrien De Bock, Philipp J. Hohensinner and PeterMechanisms of Vessel Branching: Filopodia on Endothelial Tip Cells Lead the Way

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 2009 American Heart Association, Inc. All rights reserved.

Greenville Avenue, Dallas, TX 75231is published by the American Heart Association, 7272Arteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/ATVBAHA.109.1851652009;29:639-649; originally published online March 5, 2009;Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/29/5/639World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at: Reprints:

document. Question and AnswerPermissions and Rightspage under Services. Further information about this process is available in the

which permission is being requested is located, click Request Permissions in the middle column of the WebCopyright Clearance Center, not the Editorial Office. Once the online version of the published article for

can be obtained via RightsLink, a service of theArteriosclerosis, Thrombosis, and Vascular Biologyin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:



ownloaded from

http://atvb.ahajournals.org/content/29/5/639

http://www.ahajournals.org/site/rights/

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://atvb.ahajournals.org//subscriptions/


mechanisms of vessel branching -...

Documents