Nuclear SUN1 stabilizes endothelial cell junctions via microtubules to regulate blood vessel formation

Endothelial cells line all blood vessels, where they coordinate blood vessel formation and the blood-tissue barrier via regulation of cell-cell junctions. The nucleus also regulates endothelial cell behaviors, but it is unclear how the nucleus contributes to endothelial cell activities at the cell periphery. Here, we show that the nuclear-localized linker of the nucleoskeleton and cytoskeleton (LINC) complex protein SUN1 regulates vascular sprouting and endothelial cell-cell junction morphology and function. Loss of murine endothelial Sun1 impaired blood vessel formation and destabilized junctions, angiogenic sprouts formed but retracted in SUN1-depleted sprouts, and zebrafish vessels lacking Sun1b had aberrant junctions and defective cell-cell connections. At the cellular level, SUN1 stabilized endothelial cell-cell junctions, promoted junction function, and regulated contractility. Mechanistically, SUN1 depletion altered cell behaviors via the cytoskeleton without changing transcriptional profiles. Reduced peripheral microtubule density, fewer junction contacts, and increased catastrophes accompanied SUN1 loss, and microtubule depolymerization phenocopied effects on junctions. Depletion of GEF-H1, a microtubule-regulated Rho activator, or the LINC complex protein nesprin-1 rescued defective junctions of SUN1-depleted endothelial cells. Thus, endothelial SUN1 regulates peripheral cell-cell junctions from the nucleus via LINC complex-based microtubule interactions that affect peripheral microtubule dynamics and Rho-regulated contractility, and this long-range regulation is important for proper blood vessel sprouting and junction integrity.

Abstract Endothelial cells line all blood vessels, where they coordinate blood vessel formation and the blood-tissue barrier via regulation of cell-cell junctions. The nucleus also regulates endothelial cell behaviors, but it is unclear how the nucleus contributes to endothelial cell activities at the cell periphery. Here, we show that the nuclear-localized linker of the nucleoskeleton and cytoskeleton (LINC) complex protein SUN1 regulates vascular sprouting and endothelial cell-cell junction morphology and function. Loss of murine endothelial Sun1 impaired blood vessel formation and destabilized junctions, angiogenic sprouts formed but retracted in SUN1-depleted sprouts, and zebrafish vessels lacking Sun1b had aberrant junctions and defective cell-cell connections. At the cellular level, SUN1 stabilized endothelial cell-cell junctions, promoted junction function, and regulated contractility. Mechanistically, SUN1 depletion altered cell behaviors via the cytoskeleton without changing transcriptional profiles. Reduced peripheral microtubule density, fewer junction contacts, and increased catastrophes accompanied SUN1 loss, and microtubule depolymerization phenocopied effects on junctions. Depletion of GEF-H1, a microtubule-regulated Rho activator, or the LINC complex protein nesprin-1 rescued defective junctions of SUN1-depleted endothelial cells. Thus, endothelial SUN1 regulates peripheral cell-cell junctions from the nucleus via LINC complexbased microtubule interactions that affect peripheral microtubule dynamics and Rho-regulated contractility, and this long-range regulation is important for proper blood vessel sprouting and junction integrity.

Editor's evaluation
Endothelial cells lining the inner side of blood vessels constitute the blood-tissue barrier via regulation of cell-cell junctions. The cell nucleus regulates endothelial cell behaviors, but it is unclear how the nucleus contributes to endothelial cell activities at the cell membrane. This study for the first time demonstrates that nuclear membrane protein SUN1 stabilizes endothelial cell-cell junctions far from the nucleus via regulation of microtubule dynamics and Rho GEF-H1 signaling, revealing longrange cellular communication important for vascular development and endothelial barrier function.

Introduction
Blood vessels form and expand via sprouting angiogenesis, a dynamic process whereby endothelial cells migrate from pre-existing vessels to form new conduits (Carmeliet and Jain, 2011;Wacker and Gerhardt, 2011;Kushner and Bautch, 2013;Bautch and Caron, 2015). During angiogenesis, endothelial cell-cell junctions destabilize and rearrange to allow for repolarization and migration toward pro-angiogenic cues (Esser et al., 1998;Dejana, 2004;Blum et al., 2008). Specifically, the endothelial cell adherens junction protein VE-cadherin is required for vascular sprouting and viability (Carmeliet et al., 1999;Montero-Balaguer et al., 2009;Sauteur et al., 2014;Szymborska and Gerhardt, 2018). As vessels mature, endothelial cell junctions stabilize and form a functional barrier that regulates egress of fluid and oxygen; barrier dysfunction leads to increased permeability and severe disease (Claesson-Welsh, 2015;Rho et al., 2017;Claesson-Welsh et al., 2021). Thus, regulation of endothelial cell junction stability is important developmentally and for vascular homeostasis.
The nucleus is usually found far from the cell periphery and junctions, yet it is important for functions critical to angiogenesis and vascular remodeling, such as polarity, migration, and mechanotransduction (Tkachenko et al., 2013;Guilluy et al., 2014;Graham et al., 2018), and perturbations of some nuclear membrane proteins affect transcriptional profiles May and Carroll, 2018;Carley et al., 2021). The linker of the nucleoskeleton and cytoskeleton (LINC) complex is comprised of both SUN (Sad1p, UNC-84) and KASH (Klarsicht, ANC-1, Syne/Nesprin Homology) proteins ) that function as a bridge between the nucleus and the cytoskeleton, and also link through subnuclear lamin filaments to chromatin (Haque et al., 2006). SUN proteins localize to the inner nuclear membrane and bind KASH proteins, or nesprins, from their C-terminus and lamins at their N-terminus, thus providing a structural link from the nuclear cortex to the cellular cytoskeleton (Padmakumar et al., 2005;Haque et al., 2006;McGee et al., 2006;Stewart-Hutchinson et al., 2008). Nesprins are long spectrin-rich proteins localized to the outer nuclear envelope that bind SUN proteins via their C-terminus while N-terminally interacting indirectly with microtubules (via various motor proteins such as dynein and kinesin) and intermediate filaments (via plectins), and directly with actin via calponin homology domains (Ketema et al., 2007;Meyerzon et al., 2009;Zhang et al., 2009;Fridolfsson et al., 2010;Starr and Fridolfsson, 2010). Two mammalian SUN proteins are ubiquitously expressed, and based on functional consequences of SUN manipulations it has been posited that SUN1 regulates microtubule-based functions while SUN2 coordinates actin regulation (Zhu et al., 2017). However, in vitro binding studies do not reveal a SUN-nesprin specificity to account for this bias (Stewart-Hutchinson et al., 2008;Ostlund et al., 2009), so how complexes are assembled and sorted in cells is unclear. It is well established that the LINC complex integrates external inputs sensed by focal adhesions, such as substrate stiffness, to regulate transcription (Carley et al., 2022), but how the LINC complex relays signals from the nucleus to the cell periphery is less understood.
The LINC complex functions in cultured endothelial cells, as knockdown (KD) of nesprin-3 leads to impaired endothelial polarity under flow (Morgan et al., 2011), while nesprin-2 and lamin A regulate proliferation and apoptosis in endothelial cells exposed to shear stress (Han et al., 2015). Depletion of nesprin-1 alters nuclear tension (Chancellor et al., 2010;Anno et al., 2012) and force application to isolated nuclei via nesprin-1 alters stiffness (Guilluy et al., 2014), while KD of nesprin-1 or nesprin-2 leads to reduced collective endothelial cell migration (Chancellor et al., 2010;King et al., 2014). Recent work showed compromised matrix adhesion and barrier function of cultured endothelial cells using a dominant negative KASH (Denis et al., 2021). However, less is understood about the roles of the SUN proteins in endothelial cell function or how the LINC complex functions in vessels in vivo (Salvador and Iruela-Arispe, 2022).
The LINC complex is required for viability in vivo. Loss of both mammalian Sun genes is embryonic lethal due to impaired neuronal nuclear migration required for proper neuronal differentiation Zhang et al., 2009). Global Sun2 loss affects epidermal nuclear positioning and cell adhesion, leading to alopecia (Stewart et al., 2015), while global loss of both Sun genes impairs epidermal differentiation due to altered integrin signaling (Carley et al., 2021). A role for the LINC complex in mechanotransduction in vivo is suggested by findings that perturbations in mechanically active skeletal and cardiac muscle affect function (Zhang et al., 2005;Zhang et al., 2010;Lei et al., 2009;Banerjee et al., 2014;Stroud et al., 2017;Zhou et al., 2017;van Ingen and Kirby, 2021). However, while global deletion of multiple LINC components that disrupt the entire complex highlight its importance Zhang et al., 2009;Carley et al., 2021), these studies do not reveal functions of individual LINC components in specific tissues. Whether the SUN proteins cell autonomously regulate the vascular endothelium, which is also mechanically active due to outward pressure and shear stress from blood flow, has not been explored.
Mutations in LMNA (lamin A/C) cause a premature aging syndrome linked to cardiovascular defects (Capell and Collins, 2006). The LINC complex protein SUN1 overaccumulates in the nucleus in this disease (Chen et al., 2012), and cellular defects due to the LMNA mutation are rescued by reduced levels of SUN1 protein, highlighting a potential function for SUN1 in the disease pathology (Chen et al., 2012;Chang et al., 2019). Here, we present an in-depth analysis of how the LINC complex component SUN1 affects blood vessel development and function in endothelial cells in vivo. We found that Sun1 cell autonomously regulates blood vessel sprouting and junction properties in vivo, consistent with SUN1 regulating endothelial cell functions via adherens junction activity. In primary endothelial cells, nuclear SUN1 coordinates peripheral microtubule dynamics that in turn regulate peripheral RhoGEF activation and junction stability. Thus, nuclear SUN1 that resides far from cell junctions regulates cell-cell communication and blood vessel sprouting via a novel microtubule-based integration pathway from the nucleus to the cell periphery.

Results
The nuclear LINC protein SUN1 regulates vascular development The LINC complex is important for cell migration (Chancellor et al., 2010;King et al., 2014;Denis et al., 2021), and blood vessel formation involves extensive endothelial cell migration; thus, we hypothesized that the LINC complex regulates angiogenic sprouting. Because mutations in endothelial cell LMNA causative for human cardiovascular disease are associated with expression changes in the LINC protein SUN1 (Chen et al., 2012), and because Sun1 has not been functionally analyzed in the vascular endothelium in vivo, we first asked whether SUN1 is required for vascular development. Utilizing a mouse line carrying a conditional Sun1 allele that we generated from Sun1 tm1a 'knockout first' mice ( Figure 1-figure supplement 1A Skarnes et al., 2011), Sun1 fl/fl mice were bred to Sun1 fl/+ ;Cdh5-CreERT2/+ mice to generate Sun1iECKO (inducible endothelial cell knockout) mice with both endothelial cell-selective and temporal control over Sun1 excision. Examination of lung DNA, which is rich in endothelial cells, revealed appropriate excision in vivo (Figure 1-figure supplement  1C).
The retinal vasculature of Sun1iECKO pups injected with tamoxifen at P (postnatal day) 1-3 and harvested at P7 had significantly reduced radial expansion relative to littermate controls ( Figure 1A-C, Figure 1-figure supplement 1D), consistent with a role for Sun1 in vascular development. Sun1iECKO retinas also had increased density at the vascular front ( Figure 1-figure supplement  1E), consistent with defects in sprouting that prevent expansion and thus increase density (Hellström et al., 2007;Ricard et al., 2012;Angulo-Urarte et al., 2018). Since vessel densities in the plexus ahead of arteries and veins exhibit heterogeneity, we measured by area and found increased density in the plexus ahead of both arteries and veins in Sun1iECKO retinas ( Figure 1B and D). We did not find differences in nuclear shape between Sun1iECKO and control retinas, though we did observe  The online version of this article includes the following source data and figure supplement(s) for figure 1:   some nuclear crowding in Sun1iECKO retinas (Figure 1-figure supplement 1F), consistent with the increase in vascular density. However, there was no significant difference in nuclear area relative to vessel area between Sun1iECKO and control retinas (Figure 1-figure supplement 1F-G).
Because adherens junction dynamics regulate vascular sprouting (Sauteur et al., 2014), this mutant phenotype suggested that endothelial cell-cell junctions were affected by loss of Sun1. VE-cadherin localization, a readout of junction integrity (Huveneers et al., 2012;Bentley et al., 2014;Wylie et al., 2018;Vion et al., 2020), was significantly less linear and more punctate in Sun1iECKO vessels, indicating increased adherens junction turnover and junction instability ( Figure 1E-F). Dextran injection was used to functionally evaluate the effects of Sun1 loss on endothelial cell junction integrity in vivo, and Sun1iECKO mice had increased signal in the surrounding tissue compared to controls ( Figure 1-figure supplement 1H), suggesting increased vascular leakage in the postnatal retina. Together, these data indicate a specific role for SUN1 in angiogenic sprouting and endothelial cell-cell junctions in vivo.
Nuclear SUN1 is required for sprouting angiogenesis Sun1 loss disrupts vascular development in the postnatal mouse retina (Figure 1), but this tissue is not amenable to long-term live image analysis. To query dynamic aspects of angiogenic sprouting, which occurs via regulated changes in endothelial adherens junction stability (Sauteur et al., 2014;Angulo-Urarte et al., 2018;Wylie et al., 2018), we utilized a 3D sprouting model (Nakatsu and Hughes, 2008) coupled with temporal image acquisition. Reduced levels of endothelial cell SUN1 via siRNA KD ( In contrast, SUN1 KD sprouts retracted more often, and many mutant sprouts collapsed partially or completely ( Figure 2D-E, Video 2). SUN1 KD sprouts also showed a more diffuse VE-cadherin junction pattern ( Figure 2F-G), similar to those of Sun1iECKO mice and indicative of disorganized junctions. Thus, SUN1 is required for proper vascular sprout dynamics and morphology, and reduced sprout length and branching are likely downstream of excess sprout retractions and perturbed junctions in SUN1-depleted vessels. The LINC complex is important for mechanotransduction in muscle fibers with high mechanical loads (van Ingen and Kirby, 2021), and sprouting angiogenesis is regulated by mechanical forces arising from blood pressure and blood flow (Huang et al., 2003). To determine whether SUN1 also regulates sprouting dynamics under laminar flow in vivo, we analyzed embryonic zebrafish using a Tg(fli:LifeAct-GFP) reporter that labels the endothelial actin cytoskeleton. Zebrafish have two sun1 genes, sun1a and sun1b; the SUN domain of Sun1b is more homologous to human SUN1, and sun1b is more highly expressed in cardiovascular tissue (Bastian et al., 2021), so this gene was chosen for manipulation. Sun1b depletion in zebrafish embryos via morpholino (MO) injection led to significantly increased numbers of shorter endothelial cell filopodia at 33-34 hpf (hours post fertilization) in the inter-segmental vessels (ISVs) that sprout from the dorsal aorta and connect to the dorsal longitudinal anastomotic vessel (DLAV) ( Figure 3A-C). Like the morphants, fish carrying a point mutation in the sun1b gene leading to a premature stop codon (sun1b sa33109 , see Materials and Methods for details) had shorter filopodia, although filopodia numbers were unchanged in the mutant background ( Figure 3D-F). Because increased filopodia are typically seen in actively migrating and sprouting endothelial cells (DeLisser, 2011), these changes are consistent with Sun1b regulating endothelial cell activation in developing zebrafish vessels exposed to physiological flow forces.
Next, sun1b morphant fish were imaged from 26 to 36 hpf to determine the effects of Sun1b depletion on vascular sprouting in vivo. In controls, the ISVs sprouted toward the dorsal plane and connected to the DLAV between 32 and 36 hpf ( Figure 3G-H, Video 3). In contrast, numerous ISVs either failed to reach the DLAV or made aberrant connections in sun1b morphant fish ( Figure 3G-H, Video 4). Staining for the tight junction protein ZO-1 revealed less linear and more abnormally shaped junctions in sun1b morphant fish, independent of ISV-DLAV anastomosis ( Figure 3I-J). These results complement the 3D sprouting analysis and show that the nuclear LINC complex protein SUN1 is SUN1 stabilizes endothelial cell-cell junctions and regulates junction integrity SUN1 loss or depletion in mouse, zebrafish, and 3D sprouting models resulted in abnormal endothelial cell-cell junctions and sprouting behaviors. Thus, we examined more rigorously the hypothesis that SUN1 regulates endothelial cell junction stability and morphology. Primary human umbilical vein endothelial cells (HUVEC) in confluent monolayers had more serrated cell-cell junctions without altered levels of VE-cadherin protein expression after SUN1 depletion, indicative of destabilized junctions ( Figure 4A, Figure 4-figure supplement 1A-B). Destabilized endothelial cell adherens junctions are associated with impaired junction integrity, so we measured electrical resistance across endothelial monolayers using real time cell analysis (RTCA) that provides an impedance value positively correlated with junction integrity. SUN1 KD endothelial cells had reduced electrical resistance compared to controls ( Figure 4B-C), indicative of functional consequences to endothelial junctions and consistent with the increased dextran leakage in vivo. Endothelial cells in vivo are exposed to blood flow that remodels endothelial junctions (Seebach et al., 2000;Yang et al., 2020). SUN1-depleted HUVEC exposed to laminar shear stress for 72 hr elongated and aligned properly, but adherens junctions were more serrated and allowed for significantly more matrix exposure as assessed by a biotin labeling assay (Dubrovskyi et al., 2013) under both static and flow conditions ( Figure 4D-F, Figure 4figure supplement 1C-D). These data are consistent with the finding that a KASH dominant negative construct disrupts endothelial electrical impedance (Denis et al., 2021) and support the hypothesis that nuclear SUN1 regulates junction integrity via effects on endothelial cell junction stability and morphology. Dysfunctional cell-cell junctions can result from abnormal junction formation or the inability of formed junctions to stabilize. To determine how two-tailed unpaired t-test. For all graphs, boxes represent the upper quartile, lower quartile, and median; whiskers represent the minimum and maximum values.
The online version of this article includes the following figure supplement(s) for figure 2: Figure supplement 1. SUN1 is nuclear localized in endothelial cells and does not regulate proliferation.  , such that junctions with a linear VE-cadherin signal (stable) had a higher value than those with more serrated patterns (destabilized). No significant difference between SUN1 KD and control junctions was seen at early times post-washout, indicating that SUN1 depletion does not affect adherens junction formation ( Figure 4G-H). However, later times post-washout revealed a significant increase in serrated junctions and gaps between endothelial cells in SUN1 KD monolayers relative to controls ( Figure 4G-H). Consistent with these findings, SUN1 KD endothelial cells also had increased VE-cadherin internalization at steady state, consistent with actively remodeling Quantification of average filopodia length shown in (A). n=39 ROIs (15 fish, NT MO) and 56 ROIs (20 fish, sun1b MO) compiled from three replicates. ****, p<0.0001 by Student's two-tailed unpaired t-test. (D) Representative images of zebrafish embryos at 34 hpf with indicated genotypes; anterior to left. Tg(fli:LifeAct-GFP) (green, vessels). Insets show ISVs with filopodia, outlines highlighted to show filopodia. Scale bar, 20 µm. (E) Quantification of filopodia number shown in (D). n=27 ROIs (9 fish, sun1b +/+ ) and 30 ROIs (10 fish, sun1b sa33109 ) compiled from two replicates. ns, not significant by Student's two-tailed unpaired t-test. (F) Quantification of average filopodia length shown in (D). n=27 ROIs (9 fish, sun1b +/+ ) and 30 ROIs (10 fish, sun1b sa33109 ) compiled from two replicates. **, p<0.01 by Student's two-tailed unpaired t-test. (G) Stills from Video 3 and Video 4 showing ISV sprouting from 26 to 36 hpf in zebrafish embryos with indicated morpholino treatment; anterior to left. Tg(fli:LifeAct-GFP) (green, vessels). White arrowhead points to ISV that does not extend or connect to dorsal longitudinal anastomotic vessel (DLAV). Yellow arrow points to ISV that extends but does not connect to DLAV. Scale bar, 20 µm. (H) Quantification of ISV connection to DLAV shown in (G). n=32 ISVs (6 fish, NT MO) and 36 ISVs (6 fish, sun1b MO) compiled from two replicates. p<0.05 by χ 2 analysis. (I) Representative images of zebrafish embryos at 34 hpf with indicated morpholino treatments; anterior to left. Tg(fli:LifeAct-GFP) (green, vessels); ZO-1 (white junctions). Outlines highlighted to show junction shapes. Scale bar, 10 µm. (J) Quantification of junction morphology shown in (I). n=80 junctions that reach DLAV and 68 junctions that do not reach DLAV (21 fish, NT MO) and 47 junctions that reach DLAV and 37 junctions that do not reach DLAV junctions (22 fish, sun1b MO). ns, not significant; *, p<0.05; **, p<0.01 by two-way ANOVA with Tukey's multiple comparisons test. For all graphs, boxes represent the upper quartile, lower quartile, and median; whiskers represent the minimum and maximum values.

SUN1 regulates microtubule localization and dynamics in endothelial cells absent effects on gene transcription
We next considered how SUN1 that resides in the nuclear membrane regulates cell behaviors at the cell periphery. SUN1 may affect cell junctions directly via the cytoskeleton, or SUN1 may indirectly affect junctions downstream of gene expression regulation. Because SUN1 is reported to regulate gene transcription and RNA export in other cell types May and Carroll, 2018), we asked whether endothelial cell transcriptional profiles were altered by SUN1 depletion. To our surprise, RNASeq analysis of HUVEC under both static and flow conditions revealed essentially no significant changes in RNA profiles except for SUN1 itself (Table 1, Figure 5-figure supplement 1), while in the same experiment we documented extensive expression changes in both up and downregulated genes between control HUVEC in static vs. laminar flow conditions, as we and others have shown (Conway et al., 2010;Liu et al., 2021;Maurya et al., 2021;Ruter et al., 2021). These findings are similar to a re-analysis of HeLa cell data (data not shown; Li et al., 2017) that revealed few significant changes in gene expression following SUN1 depletion. Thus, although the LINC complex is important for nuclear communication that can affect gene expression, SUN1 is not required for this communication in endothelial cells. We hypothesized that SUN1 directly regulates cell junction stability via the cytoskeleton. SUN1 has a functional relationship with the microtubule cytoskeleton (Zhu et al., 2017), and microtubule dynamics regulate endothelial cell-cell junctions (Sehrawat et al., 2008;Sehrawat et al., 2011;Komarova et al., 2012). Thus, we asked whether the observed adherens junction defects following SUN1 depletion resulted from changes in the microtubule cytoskeleton. Microtubule depolymerization via nocodazole treatment phenocopied SUN1 KD and destabilized adherens junctions in control monolayers but did not exacerbate the junction defects seen with SUN1 KD ( Figure 5A-B). These findings suggest that endothelial cell junction defects are downstream of microtubule perturbations induced by SUN1 depletion.
We next investigated how SUN1 affects microtubule localization in endothelial cell monolayers and found that significantly fewer microtubules reached the cell periphery or surrounded the nucleus in t-test. (D) Representative images of HUVEC with indicated siRNAs plated on biotinylated fibronectin and exposed to 15 dyn/cm 2 shear stress for 72 hr then treated with streptavidin. Endothelial cells were stained for DAPI (cyan, DNA), streptavidin (green), and VE-cadherin (     The changes in peripheral microtubule localization were accompanied by significantly fewer microtubule-junction contacts, while α-tubulin levels around the MTOC were not affected ( Figure 5C-D and F). Microtubule dynamics were assessed via tip tracking using mCherry-labeled tip protein EB3, which decorated the microtubule lattice and concentrated at growing microtubule tips ( Figure 5G). This labeling pattern can occur with EB overexpression but does not affect growth rate (Komarova et al., 2005), so the decorated lattice was used to assess microtubule catastrophe and shrinkage. Microtubules in SUN1-depleted cells had increased shrinkage rates (a more negative value) coupled with increased catastrophe rate and time spent shrinking ( Figure 5G-K, Video 5 and Video 6 ), consistent with elevated microtubule depolymerization and impaired microtubule dynamics downstream of SUN1 loss, while the microtubule growth rate was unchanged. Taken together, these results indicate that loss of SUN1 impairs microtubule localization and dynamics, and these changes associate with destabilized endothelial cell junctions.

SUN1 regulates endothelial cell contractility
Microtubule dynamics communicate with the actin cytoskeleton to regulate cell-cell junctions (Verin et al., 2001;Birukova et al., 2004;Birukova et al., 2006), and cellular changes in actomyosin contractility contribute to junction destabilization (Rauzi et al., 2010;Huveneers et al., 2012). We hypothesized that actomyosin mis-regulation induced by SUN1 depletion contributes to endothelial cell junction destabilization and found that SUN1 depletion led to ectopic stress fibers, distinct radial actin bundles at the periphery, and increased phosphorylated myosin light chain (ppMLC), consistent with increased actomyosin contractility (   blockade of myosin-II ATPase rescued the destabilized cell junctions, ectopic stress fibers, and radial actin structures seen with SUN1 depletion ( Figure 6A-B, Figure 6-figure supplement 1C), suggesting that SUN1 depletion increases contractility. In contrast, thrombin treatment that induces contractility produced disorganized junctions in controls that phenocopied SUN1 depletion but did not contribute to further junction destabilization in SUN1 KD cells, indicating that SUN1 loss induces a maximal contractile state ( Figure 6C-D). Taken together, these data indicate that SUN1 loss results in hypercontractility and destabilized cell-cell junctions.

SUN1 affects endothelial junctions through microtubule-associated Rho GEF-H1
We next wanted to better understand the link between microtubule dynamics, actomyosin contractility, and junction regulation downstream of SUN1. SUN1 silencing increases RhoA activity in HeLa cells (Thakar et al., 2017), and pharmacological inhibition of Rho kinase (ROCK) signaling in SUN1depleted endothelial cells rescued the destabilized cell-cell junctions and prevented the radial actin structures ( Figure 6-figure supplement 2A-C). RhoA signaling and barrier function in endothelial cells is regulated by the microtubule-associated RhoGEF, GEF-H1, which is inactive while bound to microtubules and peaks in activation following microtubule depolymerization and release into the cytosol (Krendel et al., 2002;Birukova et al., 2006;Birkenfeld et al., 2008;Azoitei et al., 2019). We hypothesized that the impaired microtubule dynamics and Rho-dependent hypercontractility observed following SUN1 depletion were linked via GEF-H1. We first examined Rho activation via pulldowns and found increased activated Rho with SUN1 silencing that was abrogated when GEF-H1 was also depleted ( Figure 6-figure supplement 2D, Figure 6-figure supplement 3A), suggesting that GEF-H1 signaling is upstream of elevated RhoA activation in SUN1-depleted cells. Consistent with this idea, GEF-H1 strongly localized to peripheral microtubules in control cells but was significantly less localized and more diffuse in the cytosol in SUN1-depleted endothelial cells ( Figure 6E-F), suggesting its release from microtubules. Furthermore, depletion of GEF-H1 in endothelial cells that were also depleted for SUN1 rescued the destabilized cell-cell junctions and loss of junction integrity observed with SUN1 KD alone ( Figure 6G-J), showing that GEF-H1 is required to transmit the effects of SUN1 depletion to endothelial cell junctions. Interestingly, depletion of GEF-H1 was not sufficient to rescue the loss of microtubule density at the periphery in SUN1-depleted cells ( Figure 6-figure supplement 3B-C), indicating that GEF-H1 acts downstream of changes in the microtubule cytoskeleton in SUN1-depleted cells. Thus, nuclear SUN1 normally promotes microtubule-GEFH1 interactions to regulate actomyosin contractility and endothelial cell-cell junction stability via RhoA, providing a novel linkage from the LINC complex to cell junctions via the microtubule cytoskeleton.

SUN1 exerts its effects on endothelial junctions through nesprin-1
Since SUN1 binds nesprins to interact with the cytoskeleton, we considered whether SUN1-nesprin interactions were involved in SUN1 regulation of microtubule dynamics and endothelial cell junction stability. The KASH protein nesprin-1 modulates tight junction protein localization under laminar shear stress (Yang et al., 2020) and regulates microtubule dynamics at the nucleus in muscle syncytia (Gimpel et al., 2017). Co-depletion of SUN1 with nesprin-1 in endothelial cells rescued the effects of SUN1 depletion on junction morphology and functional destabilization measured by matrix biotin labeling ( Figure 7A and D, Figure 7-figure supplement 1A), while nesprin-1 depletion alone did not affect these parameters (Figure 7-figure supplement 1B-C). This rescue extended to the cytoskeleton, as co-depletion of SUN1 and nesprin-1 rescued the decreased peripheral microtubule density and microtubule-GEF-H1 contacts seen with SUN1 depletion ( Figure 7B-C and E-F), while nesprin-1 depletion alone did not have an effect (Figure 7-figure supplement 1D-G). Thus, the LINC complex protein nesprin-1 is required to transmit the effects of SUN1 depletion to endothelial cell junctions.

Discussion
The nucleus compartmentalizes and organizes genetic material, but how the nucleus directly communicates with other organelles and the cytoskeleton to regulate cell behaviors is poorly understood.
Here, we show for the first time that the nuclear LINC complex protein SUN1 regulates angiogenic sprouting and endothelial cell-cell junction integrity via long-distance regulation in vivo, and that these effects go through the microtubule cytoskeleton to regulate microtubule dynamics and actomyosin contractility. Our data is consistent with a model in which SUN1 stabilizes peripheral microtubules that coordinate activity of the microtubule-associated Rho exchange factor GEF-H1. GEF-H1 becomes activated to stimulate Rho signaling upon release from microtubules, and we posit that peripheral microtubules normally regulate local GEF-H1 activity to maintain appropriate microtubulejunction interactions and actomyosin contractility in endothelial cells, leading to cell-cell junctions that remodel to support angiogenic sprouting and maintain vessel barrier integrity (Figure 8). Loss of SUN1 reduces peripheral microtubules, resulting in GEF-H1 over-activation, elevated RhoA signaling, and increased contractility, leading to destabilized endothelial cell adherens junctions that impair blood vessel formation and function in fish and mice. The LINC complex is further implicated in endothelial junction regulation by our finding that the KASH protein nesprin-1 is required for the effects of SUN1 loss on microtubules, GEF-H1, and junction integrity. Thus, we describe a specific role for SUN1 as a critical mediator of communication between the endothelial cell nucleus and the cell periphery via microtubule regulation of GEF-H1. We investigated the role of SUN1 in endothelial cell and vascular function; however, the molecular machinery whereby SUN1 regulates cell junctions is not specific to endothelial cells. Thus, SUN1 likely influences cell-cell junctions via similar mechanisms in other cell and tissue types. We show that the LINC complex protein SUN1 is required for endothelial adherens junction stabilization and proper blood vessel formation and function in vivo. The adherens junctions of expanding retinal vessels are destabilized in mice lacking endothelial Sun1, and this cellular phenotype is accompanied by increased leak, reduced radial expansion, and increased network density, indicating that destabilized junctions contribute to the vessel network perturbations. Interestingly, mice with global knockout of Sun1 are viable Zhang et al., 2009) and zebrafish injected with MO against sun1b resemble controls by 72 hpf (data not shown), suggesting that the developmental disruption of the vascular network does not perdure. This may be due to partial redundancy with SUN2 Zhang et al., 2009) preventing more profound vascular effects. However, the significant vascular defects in vessels lacking Sun1 reveal non-redundant functions for SUN1 in vascular development. We show that depletion of SUN1 leads to impaired junction morphology and integrity as measured by increased biotin matrix labeling and decreased impedance in cultured cells and increased dextran leak in the postnatal retina. Because junction morphology and integrity are integral to maintenance of the endothelial barrier, these findings suggest that SUN1 depletion leads and 149 junctions (SUN1 KD, blebbistatin) compiled from three replicates. ns, not significant; ****, p<0.0001 by two-way ANOVA with Tukey's multiple comparisons test. (C) Representative images of HUVEC with indicated siRNAs and indicated treatments. Endothelial cells were stained for DAPI (cyan, DNA) and VE-cadherin (white, junctions). Insets show junctions. Scale bar, 20 µm. (D) Quantification of VE-cadherin line scans for treatments shown in (C). n=75 junctions (NT, vehicle), 70 junctions (NT, thrombin), 71 junctions (SUN1 KD, vehicle), and 73 junctions (SUN1 KD, thrombin) compiled from three replicates. ns, not significant; ****, p<0.0001 by two-way ANOVA with Tukey's multiple comparisons test. (E) Representative images of HUVEC with indicated siRNAs. Endothelial cells were stained for DAPI (blue, DNA), α-tubulin (red, microtubules), and GEF-H1 (cyan). Insets show α-tubulin and GEF-H1 colocalization. Scale bar, 20 µm. (F) Quantification of free GEF-H1 normalized to α-tubulin associated GEF-H1 shown in (E).      to impaired barrier function. Mutations in other genes that affect endothelial cell junction integrity, such as Smad6, Pi3kca, and Yap/Taz, also perturb retinal angiogenesis (Angulo-Urarte et al., 2018;Neto et al., 2018;Wylie et al., 2018). Live image analysis of active vessel sprouting showed that SUN1 regulates filopodia dynamics and anastomosis in zebrafish and sprout dynamics in mammalian endothelial cell sprouts, and these vessels also had abnormal junctions in the absence of SUN1. Altered sprout dynamics are found in other scenarios where VE-cadherin is absent or abnormal such as Wnt inhibition, loss of VE-cadherin, PI3-kinase inhibition, and excess centrosomes (Sauteur et al., 2014;Sauteur et al., 2017;Kushner et al., 2016;Angulo-Urarte et al., 2018;Hübner et al., 2018; Buglak et al., 2020). The LINC complex regulates endothelial cell aggregation into tube-like structures in Matrigel (King et al., 2014;Denis et al., 2021), consistent with our findings that highlight a central role for SUN1 in blood vessel formation in vivo. Unlike genes encoding components of junctions and signaling effectors, SUN1 functions at significant cellular distances from cell junctions.
How does SUN1 regulate cell-cell junctions from a distance? The importance of the LINC complex in transducing signals from the cell periphery and outside the cell to the nucleus to affect nuclear envelope properties and gene transcription is well studied (Carley et al., 2022), but how information goes from the nucleus to the cell periphery is less well understood. The LINC complex is important in nucleus-to-cytoplasm communication, but most published studies assess manipulations that affect the LINC complex globally Graham et al., 2018;Carley et al., 2021;Denis et al., 2021), and how individual components contribute is not well understood. Ueda et al., 2022, found that SUN1 regulates focal adhesion maturation in non-endothelial cells in vitro via effects on the actin cytoskeleton (Ueda et al., 2022), and we found that SUN1 regulates endothelial cell-cell junctions through microtubules, indicating that SUN1 is important for signaling from the nucleus to the cell periphery by regulating cytoskeletal organization at several levels. Since LINC complex manipulations are reported to affect focal adhesions and cell migration (King et al., 2014;Denis et al., 2021;Ueda et al., 2022), it is likely that vessel dysfunction following SUN1 manipulation is complex. However, SUN1 is thought to be important in microtubule-associated LINC complex functions (Zhu et al., 2017), and disruption of microtubules or their dynamics destabilizes adherens junctions in both endothelial and non-endothelial cells (Komarova et al., 2012;Vasileva and Citi, 2018). Our data show that SUN1 depletion impairs both microtubule dynamics and microtubule localization. Specifically, elevated rates

Loss of SUN1
Normal Condition of microtubule catastrophe and shrinkage absent changes in growth rate likely account for reduced peripheral microtubule density and microtubule-junction contacts. Endothelial junction destabilization was phenocopied by microtubule depolymerization, consistent with our model that the influence of SUN1 on microtubule dynamics and localization is important for the stabilization of cell junctions.
Here, we show that SUN1 regulates cell junctions via the microtubule-regulated Rho activator GEF-H1. GEF-H1 specifically modulates RhoA signaling at endothelial cell adherens junctions to influence VE-cadherin internalization (Juettner et al., 2019). Microtubule depolymerization releases GEF-H1 to activate RhoA signaling and elevate actomyosin contractility (Krendel et al., 2002;Birkenfeld et al., 2008;Azoitei et al., 2019) in non-endothelial cells, while direct manipulation of GEF-H1 via depletion or blockade attenuates agonist-induced endothelial barrier dysfunction (Birukova et al., 2006). Our work revealed that peripheral GEF-H1 localization was more diffuse following endothelial SUN1 silencing, suggesting its activation with SUN1 loss, and Rho-kinase inhibition or GEF-H1 depletion rescued SUN1 depletion-induced RhoA activation and destabilization of endothelial cell junctions. These findings support that altered microtubule dynamics downstream of SUN1 depletion promote the release and activation of peripheral GEF-H1, and this activation destabilizes cell junctions and likely affects vascular barrier function.
Thus, SUN1 regulation of microtubule dynamics is linked to its regulation of junction stability, although exactly how SUN1 influences microtubule dynamics and function is unclear. SUN1 resides in the nuclear envelope and alters gene transcription in other cell types May and Carroll, 2018). Our transcriptional analysis of endothelial cells under flow did not reveal significant transcriptome changes with SUN1 depletion, indicating that SUN1 regulation of endothelial junctions occurs via its role in LINC complex interactions with the cytoskeleton. Nesprin-1, a KASH protein that functions in LINC complex-cytoskeletal interactions, is shown here to mediate the effects of SUN1 loss on endothelial cell junctions, including junction morphology and function and peripheral microtubule density, suggesting that SUN1 normally sequesters nesprin-1 to prevent formation of ectopic complexes. SUN1 may normally prevent abnormal or unstable nesprin-1 LINC complexes that promote microtubule depolymerization, and SUN1 loss allows these complexes to form. SUN1 may compete for nesprin-1 binding with other nuclear envelope proteins that interact with nesprin-1, such as SUN2 or nesprin-3 (Stewart-Hutchinson et al., 2008;Taranum et al., 2012;Yang et al., 2020). This idea is consistent with the finding that SUN1 antagonizes SUN2-based LINC complexes that promote RhoA activity in HeLa cells, although microtubule localization changes were not reported (Thakar et al., 2017).
Our finding that SUN1 regulates endothelial cell junction integrity and blood vessel sprouting has implications for diseases associated with aging, as vascular defects underlie most cardiovascular disease. Children with Hutchinson-Gilford progeria syndrome (HGPS) have a mutation in the LMNA gene encoding lamin A/C, resulting in accumulation of an abnormal lamin protein called progerin; these patients age rapidly and die in their early to mid-teens from severe atherosclerosis (De Sandre-Giovannoli et al., 2003;Eriksson et al., 2003;Olive et al., 2010). Progerin has increased SUN1 affinity that leads to SUN1 accumulation in HGPS patient cells (Haque et al., 2010;Chen et al., 2012;Chen et al., 2014;Chang et al., 2019). Endothelial cells also accumulate SUN1 in HGPS mouse models (Osmanagic-Myers et al., 2019), and loss of Sun1 partially rescues progeria phenotypes in mouse models and patient cells (Chen et al., 2012;Chang et al., 2019). Thus, nuclear membrane perturbations affecting SUN1 cause disease, and here we find that nuclear SUN1 regulates microtubules to affect both the microtubule and actin cytoskeletons. These effects are transmitted to endothelial cell-cell junctions far from the site of SUN1 localization to influence endothelial cell behaviors, blood vessel sprouting, and barrier function.

Microscopy
Unless otherwise stated, all imaging was performed as follows: confocal images were acquired with an Olympus confocal laser scanning microscope and camera (Fluoview FV3000, IX83) using 405, 488, 561, and 640 nm lasers and a UPlanSApo 40× silicone-immersion objective (NA 1.25), UPlanSApo 60× oil-immersion objective (NA 1.40), or UPlanSApo 100× oil-immersion objective (NA 1.40). Imaging was performed at RT for fixed samples. Images were acquired with the Olympus Fluoview FV31S-SW software and all image analysis, including Z-stack compression, was performed in Fiji (Linkert et al., 2010;Schindelin et al., 2012). Any adjustments to brightness and contrast were performed evenly for images in an experiment.

Mouse retinas
Tamoxifen-injected mice were sacrificed at P7, eyes were collected, fixed in 4% PFA for 1 hr at RT, then dissected and stored at 4°C in PBS for up to 2 weeks (Chong et al., 2017). Retinas were permeabilized in 0.5% Triton X-100 (T8787, Sigma) for 1 hr at RT, blocked for 1 hr at RT in blocking solution (0.3% Triton X-100, 0.2% BSA (A4503, Sigma)), and 5% goat serum (005-000-121, Jackson ImmunoResearch), then incubated with VE-cadherin antibody (anti-mouseCD144, 1:100, 550548, BD Pharmingen) or ERG antibody (1:500, ab196149, Abcam) in blocking solution overnight at 4°C. Samples were washed 3×, then incubated with Isolectin B4 AlexaFluor 488 (1:100, I21411, Thermo Fisher) and goat anti-rat AlexaFluor 647 (1:500, A21247, Life Technologies) for 1 hr at RT. Retinas were mounted with Prolong Diamond Antifade mounting medium (P36961, Life Technologies) and sealed with nail polish. Images were obtained using either a UPlanSApo 10× air objective (NA 0.40) or UPlanSApo 40× siliconeimmersion objective (NA 1.25). Percent radial expansion was calculated by dividing the distance from the retina center to the vascular front by the distance from the retina center to the edge of the tissue with four measurements per retina. Vascular density was measured by imaging a 350 µm × 350 µm ROI at the vascular edge. Fiji was used to threshold images, and the vessel area was normalized to the area of the ROI (n=4 ROI/retina, chosen at 2 arteries and 2 veins). Junctions were measured by taking the ratio of the mean intensity of the junction and the mean intensity of the area immediately adjacent to the junction. Mean intensity was measured via line scans in Fiji. Sixteen junctions/retina were measured.

Retina blood vessel permeability
Tamoxifen-injected mice were anesthetized at P7 in isoflurane for 5 min. The abdomen was opened, and the diaphragm was cut. One-hundred µl of 5 mg/ml 10 kDa Dextran-Texas Red (D1863, Invitrogen) in PBS was injected into the left ventricle of the heart. Eyes were immediately collected and fixed in 4% PFA for 1 hr at RT, then dissected and stained as described above. Leak was determined by making a mask of the vessel area using the isolectin channel, then assessing the dextran signal outside the vessel.

3D sprouting angiogenesis assay
The 3D sprouting angiogenesis assay was performed as previously described (Nakatsu and Hughes, 2008;Nesmith et al., 2017). 48 hr following siRNA KD, HUVEC were coated onto cytodex 3 microcarrier beads (17048501, GE Healthcare Life Sciences) and embedded in a fibrin matrix by combining 7 µl of 50 U/ml thrombin (T7201-500UN, Sigma) with 500 µl of 2.2 mg/ml fibrinogen (820224, Fisher) in a 24-well glass-bottomed plate (662892, Grenier Bio). The matrix was incubated for 20 min at RT followed by 20 min at 37°C to allow the matrix to solidify. EGM-2 was then added to each well along with 200 µl of normal human lung fibroblasts (CC2512, Lonza) at a concentration of 2×10 5 cells/ml. At day 7 of sprouting, fibroblasts were removed via trypsin treatment (5X-trypsin for 3 min at 37°C), and samples were fixed in 4% PFA for 15 min at RT. 0.5% Triton X-100 in DBPS was added to the wells, and incubation was overnight at 4°C. After rinsing 3× in DPBS, samples were blocked (5% goat serum [005-000-121, Jackson ImmunoResearch], 1% BSA [A4503, Sigma], and 0.3% Triton X-100 [T8787, Sigma]) overnight at 4°C. Samples were rinsed 3× in DPBS then anti-VE-cadherin antibody (1:1000, 2500S, Cell Signaling) in blocking solution was added for 24 hr at 4°C. Samples were rinsed 3× 10 min in 0.5% Tween 20 then washed overnight at 4°C in 0.5% Tween. Samples were rinsed 3× in DPBS, then DAPI (0.3 µM, 10236276001, Sigma) and AlexaFluor488 Phalloidin (1:50, A12379, Life Technologies) in blocking solution were added to the wells, and incubation was overnight at 4°C prior to rinsing 3× in DPBS. For whole bead analysis, images were acquired in the Z-plane using a UPlanSApo 20× oilimmersion objective (NA 0.58) and processed in Fiji. Average sprout length was measured by tracing each sprout from base (bead) to tip, then averaging lengths per bead. Branching was measured by counting total branch points and normalizing to total sprout length per bead using the AnalyzeSkeleton plugin (Arganda-Carreras et al., 2010). For junctions, images were acquired in the Z-plane using a UPlanSApo 60× oil-immersion objective (NA 1.40). Junctions were measured by taking the ratio of the mean intensity of the junction and the mean intensity of the area immediately adjacent to the junction. Mean intensity was measured via line scans in Fiji.
Live imaging on HUVEC sprouts was performed between days 4 and 6.5 of sprouting. Images were acquired on an Olympus VivaView Incubator Fluorescence Microscope with a UPLSAPO 20× objective (NA 0.75) and ×0.5 magnification changer (final magnification ×20) with a Hamamatsu Orca R2 cooled CCD camera at 30 min intervals for 60 hr at 37°C. Images were acquired using the MetaMorph imaging software and analyzed in Fiji. Sprouts were considered to have 'retracted' if they regressed toward the bead for at least three imaging frames (1.5 hr).

Zebrafish imaging
Dechorionated embryos were incubated in ice-cold 4% PFA at 4°C overnight or RT for 2 hr. Embryos were permeabilized in 0.5% Triton X-100 in PBST (DPBS+0.1% Tween 20) for 1 hr at RT then blocked (PBST+0.5% Triton X-100+1% BSA+5% goat serum+0.01% sodium azide) for 2 hr at RT. Anti-ZO1 primary antibody (1:500, 33-9100, Thermo Fisher) was added overnight at 4°C. Embryos were rinsed in PBST overnight at 4°C. Goat anti-mouse AlexaFluor647 secondary antibody (1:1000, A21236, Life Technologies) was added overnight at 4°C. Embryos were washed 3× in PBST for 30 min then overnight in PBST at 4°C. Embryos were rinsed in PBS and a fine probe was used to de-yolk and a small blade to separate the trunk from the cephalic region. Samples were mounted using Prolong Diamond Antifade mounting medium (P36961, Life Technology) and the coverslip was sealed with petroleum jelly. Imaging was at RT using a UPlanSApo 20× oil-immersion objective (NA 0.58) or a UPlanSApo 60× oil-immersion objective (NA 1.40) with an additional 3× magnification, for a total magnification of 180×. Filopodia length was measured by drawing a line from the filopodia base to the tip. Filopodia number was measured by counting the number of filopodia and normalizing to the total vessel length. Filopodia in at least three areas per fish were measured. Junctions were analyzed by drawing a line along the junction then normalizing to the shortest distance between the two ends of the junction. At least three junctions were measured per fish.
For live imaging of zebrafish, fish were dechorionated and then anesthetized with 1× Tricaine in E3 for 5 min. Fish were embedded in 0.5% agarose in 1× Tricaine-E3 medium on a glass-bottomed plate in a stage-top incubator (TOKAI HIT, WSKM) at 28.5°C. Images were acquired using a UPlanSApo 40× air objective (NA 0.95) every 15 min for 10-15 hr. ISVs that did not reach the DLAV or connected at non-consistent intervals were considered to have a missing or aberrant DLAV connection if at least one ISV posterior to the scored ISV made a normal connection.

EdU labeling
HUVEC were labeled with EdU using the Click-It EdU Kit 488 (Invitrogen, C10337) and fixed according to the manufacturer's instructions. Cells positive for EdU labeling were counted and compared to total cell number to obtain percent positive.

Junction analysis
Endothelial cell adherens junctions were quantified in monolayers using Fiji to generate 15 µm line scans of VE-cadherin signal parallel to the cell junctions. VE-cadherin signal was integrated to obtain the area under the curve. Linear junctions with consistent VE-cadherin signal thus had a large area under the curve, while more serrated junctions had reduced area under the curve (Figure 4-figure  supplement 1E). Measurements were performed on at least 9 cells per field of view, with 3-6 fields of view per condition.

Real time cell analysis
Barrier function was assessed using the xCELLigence Real-Time Cell Analyzer (RTCA, Acea Biosciences/Roche Applied Science) to measure electrical impedance across HUVEC monolayers seeded onto microelectrodes. HUVEC were seeded to confluency on the microelectrodes of the E-plate (E-plate 16, Roche Applied Science). Electrical impedance readings were taken every 2 min for 5 hr. The percent change in cell index was obtained at the 5 hr timepoint using the following formula: (Cell Index SUN1 -Cell Index NT )/ABS(Cell Index NT ).

Flow experiments
Flow experiments were performed using an Ibidi pump system (10902, Ibidi) as described . HUVEC were seeded onto fibronectin coated Ibidi slides either µ-Slide I Luer I 0.4 mm (80176, Ibidi) or µ-Slide Y-shaped (80126, Ibidi) in flow medium (EBM-2 with 10% FBS and 1× Antibiotic-antimycotic). HUVEC were exposed to laminar shear stress for 30 min at 5 dyn/cm 2 , followed by 30 min at 10 dyn/cm 2 , and finally for 72 hr at 15 dyn/cm 2 . Alignment was measured by taking the ratio of the longitudinal axis to the transverse axis relative to the flow vector. At least 10 cells were measured per condition per experiment. Vascular permeability in vitro was determined using the biotin matrix labeling assay as described below.

Biotin matrix labeling assay
Labeling of biotinylated matrix was assessed as described (Dubrovskyi et al., 2013). Briefly, fibronectin was biotinylated by incubating 0.1 mg/ml fibronectin with 0.5 mM EZ-Link Sulfo-NHS-LC-Biotin (A39257, Thermo Fisher) for 30 min at RT. Glass chamber slides were coated with 5 µg/ml biotinylated-fibronectin and HUVEC were seeded on top. At confluency, HUVEC were treated with 25 µg/ml Streptavidin-488 (S11223, Invitrogen) for 3 min then immediately fixed. For quantification, Fiji was used to threshold the streptavidin signal, and the streptavidin area was measured and normalized for total area for at least three fields of view per experiment.

Junction reformation assay
The EDTA junction reformation assay was performed as previously described (Wright et al., 2015). Briefly, HUVEC were treated with 3 mM EDTA (EDS-100G, Sigma-Aldrich) for 1 hr at 37°C. EDTA was then washed out 3× with DPBS, incubation at 37°C in EGM-2 was continued, and cells were fixed at 0, 20, 40, and 60 min intervals.

Microtubule analysis
For nuclear microtubule analysis, an ROI was drawn in Fiji over the nucleus. Fiji was used to threshold α-tubulin signal, and the area of the signal was measured within the ROI. This was performed on 8 cells per field of view, with 3-6 fields per condition per experiment. For high-resolution microtubule analysis, highresolution confocal images were acquired with a Zeiss 880 Confocal with AiryScan FAST microscope with GaAsP detector and camera (Zeiss) using a Plan-Apo 63× oil-immersion objective (NA 1.40) and 488 and 561 nm lasers. Imaging was performed at RT, and images were acquired with the Zeiss 880 software and with the AiryScan detector. Images were then processed with AiryScan. All image analysis, including Z-stack compression, was performed in Fiji. For microtubule density, an ROI was drawn at the MTOC and the cell periphery. Fiji was used to threshold α-tubulin signal, and the area of the signal was measured within the ROI. This was performed on at least 10 cells per condition per experiment. For junction analysis, the number of contacts between α-tubulin and VE-cadherin was counted and normalized to the junction length. At least 15 junctions were measured per condition per experiment.

Microtubule tip tracking
HUVEC were infected with an EB3-mCherry Lentivirus (Kushner et al., 2014) for 24 hr to visualize microtubule comets. Following infection, HUVEC were incubated with siRNAs for NT or SUN1. For live imaging, cells were incubated at 37°C in a stage-top incubator (TOKAI HIT, WSKM). Images were acquired on an Andor XD spinning disk confocal microscope based on a CSU-X1 Yokogawa head with an Andor iXon 897 EM-CCD camera. A 561 nm laser and FF01-607/36 emission filter were used. Images were acquired with a 470 ms exposure and 32 ms readout time for 2 min (240 frames) using a UPlanSApo 60× silicone-oilimmersion objective (NA 1.30) and Metamorph software. Microtubules at the cell periphery were tracked using the Manual Tracking plugin in Fiji and were tracked for at least 60 frames (30 s). Track information was acquired from the x and y coordinates using a custom algorithm in Visual Basic in Excel provided by Dr Dan Buster at the University of Arizona. Ten microtubule tracks were measured per cell.

GEF-H1 analysis
For GEF-H1 localization analysis, an ROI was drawn at the periphery of the cell, and the α-tubulin signal was used to create a mask. Mean signal intensity was then measured for GEF-H1 within the α-tubulin mask and outside of it, and a ratio was taken. At least 9 cells were analyzed per condition per experiment.
Samples were resolved on polyacrylamide gels in the presence of SDS. Resolved gels were transferred onto nitrocellulose membranes, blocked with 5% BSA in Tris-buffered saline (25 mM Tris, pH 7.6, 150 mM NaCl) plus 0.1% Tween-20 (TBST) and incubated with anti-RhoA primary antibody (1:1000, 2117, Cell Signaling) overnight at 4°C with gentle rocking. Blots were washed extensively in TBST, then incubated with species-appropriate HRP-conjugated secondary antibody (111-035-144, Jackson ImmunoResearch) for 1 hr at RT. Blots were again washed in TBST and developed using an enhanced chemiluminescent reagent (Thermo Fisher Scientific) and X-ray film.

Statistics
Student's two-tailed unpaired t-test was used to determine statistical significance in experiments with two groups and one-way ANOVA with Tukey's multiple comparisons test was used in experiments with three or more groups. For thrombin, blebbistatin, Y-27632, and nocodazole experiments, two-way ANOVA with Tukey's multiple comparisons test was used to determine statistical significance. Χ 2 was used for categorical data. For box and whisker plots, boxes represent the upper quartile, lower quartile, and median; whiskers represent the minimum and maximum values. Statistical tests and graphs were made using the Prism 9 software (GraphPad Software). The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.