The function and integrity of the vascular system requires

(1)

Spreading, and Blood Vessel Wall Stability

Sabu Abraham, Naoko Kogata, Reinhard Fa¨ssler, Ralf H. Adams

Abstract—Growth, maturation, and integrity of the blood vessel network require extensive communication between the endothelial cells, which line the vascular lumen, and associated mural cells, namely vascular smooth muscle cells and pericytes. Pericytes extend long processes, make direct contact with the capillary endothelium, and promote vascular quiescence by suppressing angiogenic sprouting. Vascular smooth muscle cells are highly contractile, extracellular matrix–secreting cells that cover arteries and veins and provide them with mechanical stability and elasticity. In the damaged blood vessel wall, for example in atherosclerotic lesions, vascular smooth muscle cells lose their differentiated state and acquire a highly mitotic, so-called “synthetic” phenotype, which is thought to promote pathogenesis. Among other factors, extracellular matrix molecules and integrin family cell–matrix receptors may regulate this phenotypic transition. Here we show that the inactivation of the gene encoding the integrin␤1 subunit (Itgb1) with a Cre-loxP approach in mice leads to mural cell defects and postnatal lethality. Integrin␤1–deficient vascular smooth muscle cells display several hallmarks of the synthetic phenotype: Cell proliferation is enhanced, whereas differentiation and their ability to support blood vessels are compromised. Similarly, mutant pericytes are poorly spread but present in larger numbers. Our analysis of this mutant model shows that integrin ␤1–mediated cell–matrix adhesion is a major determinant of the mural cell phenotype. (Circ Res. 2008;102:562-570.)

Key Words: integrin 䡲 adhesion 䡲 blood vessel 䡲 vascular smooth muscle cell 䡲 pericyte

T

he function and integrity of the vascular system requires reciprocal interactions between the endothelial layer of the vessel wall and the more peripherally located vascular smooth muscle cells (VSMCs) and pericytes (PCs). PCs associate tightly with the endothelial cells (ECs) of capillaries, small venules, and immature blood vessels so that both cell types are in direct contact, coupled by junctions, and enclosed by a single basement membrane layer.^1–3By contrast, VSMCs are found on more mature and larger caliber blood vessels, lack direct EC contact, and surround the outer basement membrane surface in one or several sheets.^1,3,4 Fully differentiated VSMCs are highly contractile and help to provide the vasculature with mechanical stability and elasticity. Smooth muscle cells (SMCs) are also a major source of the matrix in the vessel wall.^4,5

The close relationship between matrix proteins and blood vessel morphogenesis is highlighted by the phenotypes of knockout mice lacking extracellular matrix (ECM) components. Loss of fibronectin or collagen IV results in lethality around midgestation because of defects in the embryonic heart and vasculature.^{6 – 8}The alternatively spliced EIIIA and EIIIB regions of fibronectin are essential for normal vascular remodeling, VSMC association, and embryonic survival.⁹

Targeted inactivation of the Lama4 gene (laminin ␣4) is compatible with embryonic angiogenesis and cardiac development, but mutant capillaries have basement membrane defects and are dilated and fragile.¹⁰Binding to these matrix substrates is mediated by integrin receptors, which, in turn, control cellular responses such as adhesion, spreading, motility, proliferation, and survival. Integrins function as heterodimers consisting of ␣ and ␤ subunits. Gene targeting experiments have uncovered that complexes containing the integrin molecules␣4, ␣5, ␣7, ␣V, and ␤8 are essential for vascular morphogenesis.^11,12 Because these integrins play important roles in many different tissues and cell types, the mutant phenotypes may reflect primary defects in ECs, PCs/VSMCs, or other cell populations. Alterations in the local matrix and/or integrin expression are also thought to promote a phenotypic switch of VSMCs in response to vessel wall injury or in atherosclerosis. Affected SMCs acquire a mitotic, poorly differentiated (“synthetic”) phenotype and share features with immature embryonic VSMCs.^4,13–15

To gain better insight into integrin function in the vessel wall, we inactivated the Itgb1 gene encoding integrin␤1 in mural cells. Integrin␤1 is a particularly promiscuous subunit that can partner with 11 distinct␣ chains and thereby mediate

Original received July 9, 2007; resubmission received November 14, 2007; revised resubmission received December 12, 2007; accepted January 3, 2008.

From the Cancer Research UK London Research Institute (S.A., N.K., R.H.A.), Vascular Development Laboratory, London, UK; Max-Planck-Institute of Biochemistry (R.F.), Department of Molecular Medicine, Martinsried, Germany; Max-Planck-Institute for Molecular Biomedicine (R.H.A.), Department of Tissue Morphogenesis, and University of Mu¨nster, Faculty of Medicine, Mu¨nster, Germany.

Correspondence to Ralf H. Adams, Vascular Development Laboratory, Cancer Research UK London Research Institute, 44 Lincoln’s Inn Fields, London WC2A 3PX, United Kingdom. E-mail [email protected]

Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/CIRCRESAHA.107.167908 562

Downloaded from http://ahajournals.org by on February 5, 2022

(2)

binding to a diversity of matrix substrates including collagen I and IV, laminins, vitronectin, and fibronectin. This versa- tility may explain why the global inactivation of the Itgb1 gene in mice is incompatible with survival beyond embryonic day (E)5.5.¹¹To circumvent this early developmental block, conditional Itgb1 mutants permitting tissue-specific gene targeting with the Cre-loxP approach have been generated previously.¹⁶ We combined these mice with Pdgfrb-Cre transgenics,¹⁷ a strain of mice expressing Cre recombinase under the control of a Pdgfrb (the gene for platelet-derived growth factor receptor␤) genomic DNA fragment. Because of Pdgfrb-Cre expression in PCs and VSMCs of the skin and other tissues,¹⁷ this gene-targeting strategy allowed us to investigate the function of the integrin␤1 subunit in mural cells without disrupting its expression in the endothelium.

Materials and Methods

Animal Models

Mice carrying a loxP-flanked Itgb1 gene (Itgb1^lox/lox) and Pdgfrb-Cre have been reported previously.^16,17 For the generation of mutants, Pdgfrb-Cre Itgb1^lox/^⫹double heterozygotes were bred to Itgb1^lox/lox mice in a mixed 129⫻C57BL/6 genetic background. Cre-negative littermates were used as controls. H-2KbtsA58 immorto¹⁸or Rosa26 – enhanced yellow fluorescent protein (EYFP) Cre reporter trans- genes¹⁹were introduced for the isolation of immortalized VSMCs or flow cytometric experiments, respectively. All animal experiments complied with the relevant laws and were approved by the Cancer Research UK Animal Ethics Committee.

Analysis of Tissues

For immunofluorescence on sections, tissue samples were fixed overnight in 4% paraformaldehyde at 4°C and embedded in paraffin for sectioning. Microtome sections were blocked with 0.1% BSA and 1.5% goat serum in PBS (30 minutes) before overnight incubation with primary antibody diluted in blocking solution. Primary antibodies were rat monoclonal anti–integrin␤1 (Chemicon, 1:100), rabbit polyclonal anti-fibronectin (Sigma, 1:200), anti– collagen IV (Chemicon, 1:200), anti–phospho-histone H3 (Upstate, 1:100), and goat polyclonal anti–smoothelin B (Santa Cruz Biotechnology, 1:100). After washing, samples were incubated in secondary antibody (anti-rabbit Alexa Fluor-488 or Fluor-546, anti-rat Alexa Fluor-488, Molecular Probes, 1:500) and counterstained with 4⬘,6- diamidino-2-phenylindole (DAPI) (Sigma, 1:1000). Fluorescence was visualized using a Leica DM IRBE light microscope.

For whole-mount staining, skin samples were fixed overnight in 4% paraformaldehyde, washed with PBS, blocked in 1% goat serum in PBS containing 0.1% Tween 20 (3 hours at room temperature), followed by overnight incubation with primary antibodies (diluted in blocking solution) at 4°C. Samples were washed 3 times (1 hour each) in PBS, incubated for 3 hours at room temperature with secondary antibodies diluted in blocking solution, and washed as before. Primary antibodies were rat anti-mouse platelet endothelial cell adhesion molecule-1 MEC13.3 (Pharmingen, 1:100), mouse anti-human ␣-smooth muscle actin (SMA) (Clone1A4, Sigma, 1:400), polyclonal rabbit anti-desmin (Abcam, 1:200), rat monoclonal anti-endomucin (gift from Dietmar Vestweber, Max-Planck- Institute for Molecular Biomedicine, Mu¨nster, Germany), and rabbit anti-fibronectin (Sigma, 1:400). A Zeiss LSM510 Meta was used for confocal microscopy. Sample analysis by electron microscopy has been described previously.¹⁷

Flow Cytometry

Mesenteric tissue from E17.5 or postnatal day (P) 2 Pdgfrb-Cre Itgb1^lox/loxRosa26-EYFP or control mice was incubated for 2 hours at 37°C in 2 mL of PBS containing 400 U/mL collagenase (GIBCO).

After the addition of 10 mL of medium (10% FCS in DMEM), cells

were dispersed with a Pasteur pipette, sieved (pore size, 100␮m), collected by centrifugation (1000 rpm, 5 minutes) and washed with 5 mL of medium. Cells were resuspended in 500␮L of DMEM containing 2% FCS, incubated with integrin ␤1 antibody for 45 minutes and anti-rat secondary antibody conjugated to Alexa Fluor- 647 (Molecular Probes). Propidium iodide (Invitrogen) staining selected live cells, which were analyzed with a FACSCalibur system (BD Biosciences).

Smooth Muscle Cell Isolation, Culture, and Analysis

The isolation of SMCs from adult aortas, as well as the culture, verification, transfection, and staining of these cells; the analysis of fibronectin fibrillogenesis; and video microscopic and automatic cell shape analysis are described in the online data supplement, available at http://circres.ahajournals.org.

Results

Targeting of Itgb1 in Mural Cells

Itgb1^Pdgfrb-Cre mutants, generated by breeding Pdgfrb-Cre transgenic mice¹⁷ into a background of animals carrying a loxP-flanked version of the Itgb1 gene (Itgb1^lox/lox),¹⁶ were obtained at E17.5 almost at the expected Mendelian ratio (18.8% instead of 25%) predicted for our breeding scheme (see Materials and Methods; Figure 1B). Even though these embryos were of normal size and appeared healthy, their skin contained visibly dilated blood vessels (Figure 1A). After birth, a gradually increasing fraction of mutants died and only 3 survivors were obtained beyond P10 (Figure 1B). Immu- nofluorescence with anti–integrin ␤1 antibodies on skin sections from P2 animals validated our genetic approach.

Whereas the anti–integrin ␤1 signal labels both endothelial cells and VSMCs in control littermates, specific staining is only visible in the Itgb1^Pdgfrb-Creendothelium but absent from

␣-SMA–positive cells (Figure 1C). Residual mural integrin

␤1 expression in mutant survivors at later postnatal stages suggests incomplete gene inactivation, and we therefore excluded these animals from our study. Flow cytometric analysis showed that integrin␤1 is gradually depleted during development and that the majority of Pdgfrb-Cre–positive cells in postnatal mutants (74.2% at P1) has lost the protein (Figure 1D).

Vascular Smooth Muscle Cell Defects in Itgb1 Mutants

Analysis of the Itgb1^Pdgfrb-Credermal vasculature by whole-mount immunofluorescence revealed the presence of vascular aneurysms, local distensions of blood vessels, affecting both arteries and veins already in embryos (Figure 2A and 2B). These aneurysms correspond with regions with poor VSMC coverage, suggesting that this phenotype is linked to mural cell defects. At P2, mutant VSMCs show a highly rounded, button-like morphology. Because of poor spreading, individual cells are often separated by gaps so that Itgb1^Pdgfrb-Cre arteries and veins lack continuous SMC coverage (Figure 2C and 2D). Staining with anti–␣-SMA antibodies shows that the cytoskeleton of control VSMCs is aligned into a sheet of parallel fibers that surrounds the endothelium tightly. By contrast, Itgb1 mutant mural cells are highly disorganized and fail to align with neighboring cells.

Moreover,␣-SMA is no longer polarized to one surface of the VSMCs in the absence of integrin␤1 (Figure 2E).

(3)

On the ultrastructural level, mutant mural cells fail to associate with the subendothelial basement membrane, appear rounded and poorly spread, and are frequently located in some distance from the endothelium. As a consequence, Itgb1 vessel walls are not covered by a layer of tightly packed VSMCs and instead appear loosely organized (Figure 2F).

Integrin␤1 Controls the Spreading of PCs

Because PCs play critical roles in the stabilization of capillary beds^2,3 and are targeted by the Pdgfrb-Cre transgene,¹⁷ we

analyzed the morphology PCs in Itgb1 mutants. Whole- mount staining with antibodies directed against desmin, an intermediate filament protein and PC marker, allows the visualization of the PCs that cover the dermal vasculature with an extensive lattice of fine processes (Figure 3A).

Desmin-positive PCs are also present in the Itgb1^Pdgfrb-Creskin, and their number is even significantly increased compared with control littermates (Figure 3A and 3E). However, mutant cells lack the characteristic slender and stretched morphology of normal PCs, and their processes appear short (Figure 3B).

Figure 1. Gene targeting of Itgb1 in mural cells. A, Phenotype of freshly isolated E17.5 embryos. The bottom images show higher magnifications of insets.

Note dilated blood vessels in the Itgb1^Pdgfrb-Creskin (right). B, Number of Itgb1^Pdgfrb-Cremice (mutants) among the total of embryos/pups analyzed at indicated stages. Percentage (%) of mutant vs total mice relative to the expected ratio (25% for this breeding scheme) indicates lethality of Itgb1 mice. C, Dou- ble immunofluorescence with antibodies against integrin␤1 (green) and ␣-SMA (red) on skin sections from P2 animals, confirming the absence of integrin␤1 in VSMCs but not the endothelium of Itgb1^Pdgfrb-Cremutants (right). Bottom images show individual channels of insets at higher magnification. Nuclei were stained with DAPI. Arrows indicate mutant or control VSMCs. D, Flow cytometric analyses of integrin␤1 expression cells isolated from Pdgfrb-Cre Itgb1^lox/lox ROSA26-YFP Cre reporter mice at E17.5 or P1. Cells from ROSA26-YFP (without anti–integrin␤1 antibody) and Pdgfrb- Cre ROSA26-YFP double transgenic embryos at E17.5 were used as controls, as indicated. Numbers in quadrants (pur- ple boxes) indicate fractions of integrin

␤1–expressing (top) and YFP-positive (Cre-expressing) integrin␤1–negative (bottom right quadrant) cell populations relative to total cells. Scale bars: 1 mm (A) and 20␮m (C).

(4)

Several lines of evidence suggest that loss of integrin ␤1 affects the interaction of PCs with ECs. In electron micrographs, mutant PCs have a round morphology, fail to wrap around dermal capillaries, and areas of endothelial-PC contact are small (Figure 3D). Whereas control PCs show no

appreciable ␣-SMA immunofluorescence, mutant PCs are SMA-positive, similar to what has been reported previously for the poorly attached PCs covering the tumor vasculature.²⁰ Furthermore, distended capillary diameters in the Itgb1^Pdgfrb-Cre skin suggest that mutant PCs, despite their presence in greater

Figure 3. Defective morphology and association of Itgb1 PCs. A through C, Whole-mount staining of control and Itgb1^Pdgfrb-Credermal vasculature with anti-desmin (green), anti–␣-SMA (red or blue), and/or anti-endomucin (red) antibodies as indicated. Arteries (a) and veins (v) are indicated in A. Desmin- positive PCs lack long cellular processes (arrows in B) and are more abundant (asterisks label cell bodies in B). C, Channel with anti–␣-SMA signal of images shown in B. D, Electron micro- graphs of dermal capillaries. Itgb1^Pdgfrb-Cre mutants lack PC processes seen around the endothelium in controls (black arrows). PCs (pc), endothelial (ec), and blood cells (bc) are indicated. E, Quanti- tative analysis showing larger capillary diameters (5.0⫾0.4 vs 6.7⫾0.7␮m;

*P⫽0.011 by 2-tailed t test assuming unequal variances) and increased PC numbers (48.5⫾6.2 vs 61.8⫾2.6 per 0.04 mm²area; *P⫽0.017 by 2-tailed t test assuming unequal variances) in P2 Itgb1^Pdgfrb-Crehead skin. Values are⫾SD and based on 4 mutant and control samples, respectively. Scale bars: 50␮m (A), 20␮m (B), and 2 ␮m (D).

Figure 2. VSMC defects in Itgb1 mutants.

A through E, Confocal images of E17.5 (A and B) or P2 skins (C through E) after whole-mount immunostaining with antibodies detecting ECs (platelet endothelial cell adhesion molecule-1, green) and/or VSMCs (␣-SMA, red). B and D, Higher-magnification images of insets in A and C, respectively. Arteries (a) and veins (v) are labeled. Arrows indicate an- eurysm in A and B and rounded VSMCs in D and E. F, Electron micrographs of dermal blood vessels. Black arrows point at loosely associated Itgb1^Pdgfrb-Cre VSMCs. Endothelial (ec) and blood cells (bc) are labeled. Scale bars: 50␮m (A and C), 20␮m (E), and 2 ␮m (F).

(5)

numbers, are not supporting the endothelium sufficiently (Figure 3E). In contrast to other mutant models in which the loss or defective association of mural cells leads to edema formation and hemorrhaging,^17,21 we did not find extrava- sated red blood cells in Itgb1 mutants. The angiogenic growth of the vascular network, as judged by the size of the vascularized area (Figure IA in online data supplement), the number of blood vessel branch points, and endothelial proliferation (data not shown), was also not significantly altered when mural cells had lost integrin␤1 expression.

Integrin␤1 Controls the Shape, Adhesion, and Motility of Cultured Mural Cells

To investigate the role of integrin␤1 in cultured mural cells, we isolated aortic smooth cell cells from adult Itgb1^lox/lox homozygous mice, which also carried the H-2Kb-tsA58 immorto transgene, allowing the inducible expression of a temperature-sensitive SV40 T antigen.¹⁸Transient expression of Cre recombinase in these cells yielded integrin ␤1–

deficient (Itgb1^KO) SMCs. Deletion of the loxP-flanked region in the Itgb1 gene of these cells was verified by genotyping PCR (data not shown), and immunofluorescence confirmed the absence of integrin ␤1 protein (Figure 4A and 4B).

Mock-transfected cells were used as controls (Figure 4A and 4B). Visualization of the focal adhesions by anti-paxillin antibody staining and of the actin cytoskeleton shows that loss of Itgb1 in cultured VSMCs phenocopies morphological changes observed in vivo. Itgb1^KO SMCs are round, poorly spread, and lack cellular protrusions seen in control cells (Figure 4A and 4C through 4E). Absence of integrin ␤1 expression does not prevent the formation of focal adhesions, even though they are short and disorganized in comparison with control cells (Figure 4A and 4B). As expected, reexpres- sion of green fluorescent protein–tagged integrin ␤1 in Itgb1^KOSMCs restores their morphology, so that they resem- ble control cells (supplemental Figure II).

Consistent with the binding affinities of ␤1-containing integrin heterodimers, attachment and spreading on collagen I and fibronectin matrix substrates is delayed in Itgb1^KO SMCs (Figure 5A and 5B). Similarly, cell motility and the persistence of migration are significantly reduced in the absence of integrin␤1 (Figure 5C and 5D and supplemental Figure I).

ECM Protein Expression in Itgb1 Blood Vessels Previous work has indicated that integrins are not only necessary for cell–matrix adhesion but, at least in some tissues, for the proper expression and deposition of ECM proteins.²² Evaluation of laminin, fibronectin, and collagen IV by immunofluorescence shows that all these matrix proteins are expressed in Itgb1^Pdgfrb-Cre dermal blood vessels (Figure 6A through 6C and supplemental Figure III). How- ever, we noted that Itgb1 ␣-SMA–positive cells protrude through the matrix layer into the surrounding dermis and are no longer fully ensheathed by collagen IV, as is the case for control blood VSMCs (Figure 6A and 6B). Whereas fibronectin is still located within the basement membrane that separates the endothelium from the VSMC layer, its assembly into long parallel fibrils is defective in Itgb1^Pdgfrb-Cremutants

Figure 4. Characterization of cultured integrin␤1–deficient VSMCs. A, Anti–integrin␤1 (red) and anti-paxillin immunofluorescence of control and integrin␤1–deficient (Itgb1^KO) aortic SMCs, as indicated. B, Details of individual channels of insets in A. Arrows indicate integrin␤1–positive focal adhesions (upper left). Note focal adhesion morphology is abnormal in Itgb1^KO cells (bottom right). C, Organization of the actin cytoskeleton (phalloidin, green) in control and mutant cells. Focal adhesions are labeled by anti-paxillin staining (red). Arrows indicate cortical actin in Itgb1^KOVSMCs. D, Example images of phalloidin- stained control (left) and Itgb1^KO(right) cells from automatic cell shape analysis (see Materials and Methods). Automatically iden- tified and analyzed cells are in green and rejected cells are in orange. E, Histogram of shape factor data. Height of bars indicates cell numbers. Larger shape factor of integrin␤1–deficient cells indicates decreased complexity of cell outlines, that is, more rounded cell shapes. Scale bars: 10␮m (A, C, and E).

(6)

(Figure 6C and 6D). Confocal analysis reveals that mutant blood vessels lack long fibronectin fibrils and the ECM protein is instead predominantly accumulated in a circular fashion underneath VSMCs (Figure 6D). Similarly, cultured Itgb1^KOSMCs are only capable of limited fibronectin fibrillogenesis (Figure 6E).

Proliferation and Differentiation of Itgb1 VSMCs To gain a better understanding of the defects in mutant vessel walls, we evaluated whether the loss of integrin␤1 has any effect on the survival, proliferation, or differentiation of mural cells. TUNEL staining shows no overlap between apoptotic (TUNEL-positive) nuclei and anti–␣-SMA antibody signal in both control and mutant skin sections (data not shown), arguing against an essential role of integrin␤1 in the protection of mural cells against cell death. By contrast, VSMC proliferation, assayed with antibodies against the mitotic marker phospho-histone H3, is increased in Itgb1^Pdgfrb-Cre mutants (Figure 7A through 7C). Thus, changes in the morphology of Itgb1 VSMCs are accompanied by upregu- lated proliferation reminiscent of the SMC behavior seen in response to vascular injury.

The loss of late differentiation markers is another feature of the synthetic VSMC phenotype.^4,15We therefore analyzed the

expression of smoothelin B, an actin-binding protein that is a late VSMC differentiation marker linked to SMC contractil- ity.²³ Whereas smoothelin B is abundantly present in

␣-SMA–positive cells of control vessels, such staining is missing in the Itgb1 dermal vasculature (Figure 7D and 7E).

The expression of various transcriptional factors with established roles in the regulation of the smooth muscle differentiation program is also reduced and/or delayed in cultured integrin ␤1–deficient SMCs (supplemental Figure IV). All these findings together are consistent with a failure of VSMCs to acquire a fully differentiated and functional phenotype in the absence of integrin␤1.

Discussion

Regulation of Mural Cell Function by Integrin␤1 Our findings demonstrate that integrin␤1 is a critical regu- lator of mural cell morphology and function in vivo and in vitro. In the skin of Itgb1^Pdgfrb-Cremutants, VSMCs are poorly spread, lack the normal alignment with neighboring cells, and support the vasculature insufficiently. This leads to poorly organized blood vessel walls and the formation of aneurysms.

Similarly, mutant PCs, albeit present in increased numbers, fail to extend the long processes that are characteristic for this cell type and lack proper interactions with the endothelium.

Figure 5. Regulation of cell adhesion and motility by integrin␤1. A, Diagram showing the spreading of control and Itgb1^KOcells on collagen I or fibronectin substrates. Fraction of spread cells at indicated times were determined by video microscopy (see Materials and Methods). B, Still images of video microscopy showing areas containing spread (black arrowheads) and unspread (white arrowheads) control and mutant cells 90 minutes after plat- ing. C, Diagrams showing tracks from control (n⫽16) and Itgb1^KO(n⫽13) cells with starting points shifted to the ori- gin. Distances are indicated (in␮m).

Similar tracks were recorded in 2 fur- ther repeat experiments but are not shown in the figure. D, Graphic repre- sentation of migration speeds (␮m/h) of control (left, n⫽16) and Itgb1^KOcells (right, n⫽13). Circles correspond to median speeds, boxes to 50%, and whiskers to 80% of values. Migration of mutant cells (14.0⫾1.4␮m/h) is significantly reduced compared with control cells (24.6⫾2.2␮m/h). ***P⬍0.001 by ANOVA.

(7)

Ectopic expression of␣-SMA by integrin ␤1–deficient PCs is highly reminiscent of PC defects in the deregulated vasculature of tumors²⁰and most likely reflects changed gene expression in response to the loose association of these cells

with the capillary network. The dilation of capillaries in the Itgb1^Pdgfrb-Creskin is also consistent with insufficient support by mural cells. Even though we were unable to establish that these vascular defects were directly responsible for the lethality of the mutant mice, some form of cardiovascular insufficiency or the rupture of dilated and fragile blood vessels are the most likely cause of death. Embryonic and perinatal growth and development is accompanied by a Figure 6. Matrix protein expression in Itgb1^Pdgfrb-Creblood ves-

sels. A through E, Immunofluorescence with antibodies directed against fibronectin or collagen IV (col IV) and␣-SMA (red), as indicated. Sectioned P2 skins (A through C), confocal images of whole-mount samples (D) or immunostaining of cultured Itgb1^KO and control SMCs (E) are shown. B, Higher-magnification images of insets in A. Arrows point to rounded VSMCs protrud- ing through the collagen IV layer. D, Long fibronectin fibers (arrows) in the proximity of␣-SMA–positive VSMCs are present in control but not in Itgb1^Pdgfrb-Credermal blood vessels. E, Itgb1^KO cells display limited fibrillogenesis of plated fibronectin compared with control SMCs, which are capable of assembling long fibers (arrows). Nuclei in A through C and E were stained with DAPI (blue). Scale bars: 20␮m (A, C, and D) and 10 ␮m (E).

Figure 7. Loss of integrin␤1 affects VSMC proliferation and differentiation. A, B, D, and E, Immunofluorescence on P2 skin sections with antibodies against phosphorylated histone H3 (p-histone H3, green), smoothelin B (green), and␣-SMA (red) as indicated. Nuclei were stained with DAPI (gray). Staining of red blood cells (*) is unspecific. B, The isolated green channel of mutant image in A. Arrows point at proliferating VSMCs in (A and B) and at control or Itgb1 VSMCs in E. C, Quantitation of p-histone H3–stained mutant and control blood vessel section.

Bars represent proportion of control or Itgb1 vessels with a cer- tain VSMC proliferation index (mitotic VSMCs/total VSMCs).

Ninety-five percent of control sections contain no phosphorylated histone H3–stained VSMCs. E, Green channel (smoothelin B) of images shown in D. Scale bars: 20␮m (A through E).

(8)

significant increase in blood pressure,^24,25and the severity of the mural cell defects suggests that the Itgb1^Pdgfrb-Cremutant vasculature may be too weak to cope with these stronger hemodynamic forces.

Because many of the known molecular players controlling mural cell biology are linked to integrins, it is worthwhile to compare the different mutant phenotypes. For example, signaling by platelet-derived growth factor (PDGF) B and its receptor PDGFR␤ is essential for the proliferation, chemo- tactic guidance, and association of PCs/VSMCs. Studies in cultured cells have established that this pathway synergisti- cally cooperates with integrins.^26,27 However, this crosstalk involves integrin␣v␤3 rather than complexes with ␤1, and we found that Itgb1^Pdgfrb-Cre mutants do not recapitulate the dramatic reduction of PCs/VSMCs caused by the inactivation of the Pdgfb or Pdgfrb genes.³

We have shown previously that that mural cells require ephrin-B2, a small transmembrane protein and ligand for Eph family receptor tyrosine kinases, for their correct association with capillaries and small caliber arteries and veins.¹⁷Even though cultured ephrin-B2– deficient aortic SMCs display prominent focal adhesion defects, the accelerated but random motility of these cells appears very distinct from the compa- rably static Itgb1 KO cells. Thus, mural cell-specific integrin

␤1 mutants fall into a separate phenotypic category charac- terized by compromised PC and VSMCs spreading, migration, and differentiation.

Redundancy Versus Functional Specificity of Integrin Receptors

Several mutant mice lacking individual integrin ␣ subunits develop vascular defects, suggesting that these molecules may partner with␤1 in PCs and VSMCs. Integrin ␣5␤1 is strongly expressed in SMCs, and the knockout of the Itga5 gene encoding the ␣5 subunit reproduces the very severe vascular defects seen in embryos lacking fibronectin, the major ligand of␣5␤1.^11,28The midgestation lethality of Itga5 embryos is likely to reflect a combination of problems affecting the heart, the endothelium, and, possibly, mural cells. Tissue-specific loss-of-function studies will be required to identify the roles of integrin ␣5 in individual cell types.

Similar to␣5, the majority of Itga4 (integrin ␣4) mutants die at midgestation, which has been attributed to failed cho- rioallantoic fusion.²⁹A fraction of mutants surviving up to E14.5 displays defective distribution of PCs/VSMCs in the cranial vasculature.³⁰Adhesion and motility of cultured Itga4 cells was also reduced, similar to Itgb1^KOSMCs.^30,31It also has been shown that the laminin receptor integrin ␣7␤1 is expressed in SMCs, and cerebrovascular hemorrhaging in the knockout mice has been attributed to VSMC defects.^32,33

Given that integrin␤1 forms functional complexes with all the subunits mentioned above, mediates adhesion to fibronectin, laminin, and collagen substrates, and also controls fibronectin fibrillogenesis in the vessel wall, it is surprising that VSMC survival or proliferation are not disrupted in Itgb1^Pdgfrb-Cre mutants. Similarly, mutant PCs are compromised in their ability to spread and support ECs but are actually present in increased numbers. These data suggest that other integrins, such as ␣v␤3 or ␣v␤5, provide sufficient adhesion for

necessary promitotic and antiapoptotic signals. Indeed, the integrins␣v␤3 or ␣v␤5 are expressed by SMCs and can bind a set of ECM molecules that partially overlaps with ␤1- containing receptor complexes.³⁴

Integrin␤1 and the Synthetic VSMC Phenotype Changes in the matrix environment or the expression, cell surface presentation, binding, or signaling properties of integrins may cause or contribute to pathological changes in the vasculature. For example, it has been shown that laminin and type I or type IV collagens help to maintain a differentiated, contractile phenotype of cultured arterial SMCs, whereas the cell attachment (RGD) motif of fibronectin has the opposite effect.^35–38 Experimentally induced vascular injury reduces the local expression and activity of␤1 integrins in vascular cells, whereas␣v␤3 or ␣v␤5 and the corresponding matrix substrates are upregulated.^{5,39 – 41}Moreover,␤3 integrins fa- vor a poorly differentiated and highly motile SMC phenotype, and knockout mice are protected against pathological VSMC migration and neointima formation.^{5,40 – 44}

The sum of these findings suggests that different integrin complexes may have opposite roles in the regulation of the SMC phenotype. The defects in Itgb1^Pdgfrb-Cre mutant mice directly confirms that integrin ␤1 is essential for VSMC morphology, differentiation, and function and may provide useful leads for future research investigating tissue repair and vascular regeneration processes.

Acknowledgments

We thank A. Compagni and P. Lindblom for preliminary character- ization of the Itgb1 mutant phenotype, D. Vestweber for reagents, D.

Zicha for help with the analysis of Itgb1 cells, and N. Hogg for critically reading the manuscript.

Sources of Funding

This work was funded by Cancer Research UK and the European Molecular Biology Organization Young Investigator Programme (R.H.A.).

Disclosures

None.

References

1. Jain RK. Molecular regulation of vessel maturation. Nat Med. 2003;9:

685– 693.

2. Bergers G, Song S. The role of pericytes in blood-vessel formation and maintenance. Neuro-oncol. 2005;7:452– 464.

3. Armulik A, Abramsson A, Betsholtz C. Endothelial/pericyte interactions.

Circ Res. 2005;97:512–523.

4. Owens GK, Kumar MS, Wamhoff BR. Molecular regulation of vascular smooth muscle cell differentiation in development and disease. Physiol Rev. 2004;84:767– 801.

5. Moiseeva EP. Adhesion receptors of vascular smooth muscle cells and their functions. Cardiovasc Res. 2001;52:372–386.

6. Poschl E, Schlotzer-Schrehardt U, Brachvogel B, Saito K, Ninomiya Y, Mayer U. Collagen IV is essential for basement membrane stability but dispensable for initiation of its assembly during early development.

Development. 2004;131:1619 –1628.

7. George EL, Georges-Labouesse EN, Patel-King RS, Rayburn H, Hynes RO. Defects in mesoderm, neural tube and vascular development in mouse embryos lacking fibronectin. Development. 1993;119:1079 –1091.

8. George EL, Baldwin HS, Hynes RO. Fibronectins are essential for heart and blood vessel morphogenesis but are dispensable for initial specifi- cation of precursor cells. Blood. 1997;90:3073–3081.

(9)

9. Astrof S, Crowley D, Hynes RO. Multiple cardiovascular defects caused by the absence of alternatively spliced segments of fibronectin. Dev Biol.

2007;311:11–24.

10. Thyboll J, Kortesmaa J, Cao R, Soininen R, Wang L, Iivanainen A, Sorokin L, Risling M, Cao Y, Tryggvason K. Deletion of the laminin alpha4 chain leads to impaired microvessel maturation. Mol Cell Biol.

2002;22:1194 –1202.

11. Bouvard D, Brakebusch C, Gustafsson E, Aszodi A, Bengtsson T, Berna A, Fassler R. Functional consequences of integrin gene mutations in mice.

Circ Res. 2001;89:211–223.

12. Sheppard D. In vivo functions of integrins: lessons from null mutations in mice. Matrix Biol. 2000;19:203–209.

13. Schwartz SM, Virmani R, Rosenfeld ME. The good smooth muscle cells in atherosclerosis. Curr Atheroscler Rep. 2000;2:422– 429.

14. Manabe I, Nagai R. Regulation of smooth muscle phenotype. Curr Ath- eroscler Rep. 2003;5:214 –222.

15. Orlandi A, Bochaton-Piallat ML, Gabbiani G, Spagnoli LG. Aging, smooth muscle cells and vascular pathobiology: implications for athero- sclerosis. Atherosclerosis. 2006;188:221–230.

16. Potocnik AJ, Brakebusch C, Fassler R. Fetal and adult hematopoietic stem cells require beta1 integrin function for colonizing fetal liver, spleen, and bone marrow. Immunity. 2000;12:653– 663.

17. Foo SS, Turner CJ, Adams S, Compagni A, Aubyn D, Kogata N, Lindblom P, Shani M, Zicha D, Adams RH. Ephrin-B2 controls cell motility and adhesion during blood-vessel-wall assembly. Cell. 2006;124:

161–173.

18. Jat PS, Noble MD, Ataliotis P, Tanaka Y, Yannoutsos N, Larsen L, Kioussis D. Direct derivation of conditionally immortal cell lines from an H-2Kb-tsA58 transgenic mouse. Proc Natl Acad Sci U S A. 1991;88:

5096 –5100.

19. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, Costantini F. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol. 2001;1:4.

20. Morikawa S, Baluk P, Kaidoh T, Haskell A, Jain RK, McDonald DM.

Abnormalities in pericytes on blood vessels and endothelial sprouts in tumors. Am J Pathol. 2002;160:985–1000.

21. Betsholtz C, Lindblom P, Gerhardt H. Role of pericytes in vascular morphogenesis. EXS. 2005;(94):115–125.

22. Geiger B, Bershadsky A, Pankov R, Yamada KM. Transmembrane crosstalk between the extracellular matrix– cytoskeleton crosstalk. Nat Rev Mol Cell Biol. 2001;2:793– 805.

23. van Eys GJ, Niessen PM, Rensen SS. Smoothelin in vascular smooth muscle cells. Trends Cardiovasc Med. 2007;17:26 –30.

24. Phoon CK, Aristizabal O, Turnbull DH. 40 MHz Doppler characterization of umbilical and dorsal aortic blood flow in the early mouse embryo.

Ultrasound Med Biol. 2000;26:1275–1283.

25. Jones JE, Jose PA. Neonatal blood pressure regulation. Semin Perinatol.

2004;28:141–148.

26. Woodard AS, Garcia-Cardena G, Leong M, Madri JA, Sessa WC, Languino LR. The synergistic activity of alphavbeta3 integrin and PDGF receptor increases cell migration. J Cell Sci. 1998;111(pt 4):469 – 478.

27. Schneller M, Vuori K, Ruoslahti E. Alphavbeta3 integrin associates with activated insulin and PDGFbeta receptors and potentiates the biological activity of PDGF. EMBO J. 1997;16:5600 –5607.

28. Yang JT, Rayburn H, Hynes RO. Embryonic mesodermal defects in alpha 5 integrin-deficient mice. Development. 1993;119:1093–1105.

29. Yang JT, Rayburn H, Hynes RO. Cell adhesion events mediated by alpha 4 integrins are essential in placental and cardiac development.

Development. 1995;121:549 –560.

30. Grazioli A, Alves CS, Konstantopoulos K, Yang JT. Defective blood vessel development and pericyte/pvSMC distribution in alpha 4 integrin- deficient mouse embryos. Dev Biol. 2006;293:165–177.

31. Garmy-Susini B, Jin H, Zhu Y, Sung RJ, Hwang R, Varner J. Integrin alpha4beta1-VCAM-1-mediated adhesion between endothelial and mural cells is required for blood vessel maturation. J Clin Invest. 2005;115:

1542–1551.

32. Velling T, Collo G, Sorokin L, Durbeej M, Zhang H, Gullberg D. Distinct alpha 7A beta 1 and alpha 7B beta 1 integrin expression patterns during mouse development: alpha 7A is restricted to skeletal muscle but alpha 7B is expressed in striated muscle, vasculature, and nervous system. Dev Dyn. 1996;207:355–371.

33. Yao CC, Breuss J, Pytela R, Kramer RH. Functional expression of the alpha 7 integrin receptor in differentiated smooth muscle cells. J Cell Sci.

1997;110:1477–1487.

34. Serini G, Valdembri D, Bussolino F. Integrins and angiogenesis: a sticky business. Exp Cell Res. 2006;312:651– 658.

35. Hedin U, Bottger BA, Luthman J, Johansson S, Thyberg J. A substrate of the cell-attachment sequence of fibronectin (Arg-Gly-Asp-Ser) is sufficient to promote transition of arterial smooth muscle cells from a contractile to a synthetic phenotype. Dev Biol. 1989;133:489 –501.

36. Hedin U, Bottger BA, Forsberg E, Johansson S, Thyberg J. Diverse effects of fibronectin and laminin on phenotypic properties of cultured arterial smooth muscle cells. J Cell Biol. 1988;107:307–319.

37. Yamamoto M, Yamamoto K, Noumura T. Type I collagen promotes modulation of cultured rabbit arterial smooth muscle cells from a con- tractile to a synthetic phenotype. Exp Cell Res. 1993;204:121–129.

38. Thyberg J, Hultgardh-Nilsson A. Fibronectin and the basement membrane components laminin and collagen type IV influence the phenotypic properties of subcultured rat aortic smooth muscle cells dif- ferently. Cell Tissue Res. 1994;276:263–271.

39. Koyama N, Seki J, Vergel S, Mattsson EJ, Yednock T, Kovach NL, Harlan JM, Clowes AW. Regulation and function of an activation- dependent epitope of the beta 1 integrins in vascular cells after balloon injury in baboon arteries and in vitro. Am J Pathol. 1996;148:749 –761.

40. Slepian MJ, Massia SP, Dehdashti B, Fritz A, Whitesell L. Beta3- integrins rather than beta1-integrins dominate integrin-matrix interactions involved in postinjury smooth muscle cell migration. Circulation. 1998;

97:1818 –1827.

41. Stouffer GA, Hu Z, Sajid M, Li H, Jin G, Nakada MT, Hanson SR, Runge MS. Beta3 integrins are upregulated after vascular injury and modulate thrombospondin- and thrombin-induced proliferation of cultured smooth muscle cells. Circulation. 1998;97:907–915.

42. Kappert K, Blaschke F, Meehan WP, Kawano H, Grill M, Fleck E, Hsueh WA, Law RE, Graf K. Integrins alphavbeta3 and alphavbeta5 mediate VSMC migration and are elevated during neointima formation in the rat aorta. Basic Res Cardiol. 2001;96:42– 49.

43. Thyberg J, Blomgren K, Roy J, Tran PK, Hedin U. Phenotypic modulation of smooth muscle cells after arterial injury is associated with changes in the distribution of laminin and fibronectin. J Histochem Cytochem. 1997;45:837– 846.

44. Choi ET, Khan MF, Leidenfrost JE, Collins ET, Boc KP, Villa BR, Novack DV, Parks WC, Abendschein DR. Beta3-integrin mediates smooth muscle cell accumulation in neointima after carotid ligation in mice. Circulation. 2004;109:1564 –1569.