Inhibition of smooth muscle β-catenin hinders neointima formation after vascular injury

Abstract

Objective

Smooth muscle cells (SMCs) contribute to neointima formation after vascular injury. Although β-catenin expression is induced after injury, whether its function is essential in SMCs for neointimal growth is unknown. Moreover, although inhibitors of β-catenin have been developed, their effects on SMC growth have not been tested. We assessed the requirement for SMC β-catenin in short-term vascular homeostasis and in response to arterial injury, and investigated the effects of β-catenin inhibitors on vascular SMC growth.

Approach and Results

We used an inducible, conditional genetic deletion of β-catenin in SMCs of adult mice. Uninjured arteries from adult mice lacking SMC β-catenin were indistinguishable from controls in terms of structure and SMC marker gene expression. After carotid artery ligation, however, vessels from mice lacking SMC β-catenin developed smaller neointimas, with lower neointimal cell proliferation and increased apoptosis. SMCs lacking β-catenin showed decreased mRNA expression of Mmp2, Mmp9, Sphk1 and S1pr1 (genes that promote neointima formation), higher levels of Jag1 and Gja1 (genes that inhibit neointima formation), decreased Mmp2 protein expression and secretion, and reduced cell invasion in vitro. Moreover, β-catenin inhibitors PKF118-310 and ICG-001 limited growth of mouse and human vascular SMCs in a dose-dependent manner.

Conclusions

SMC β-catenin is dispensable for maintenance of the structure and state of differentiation of uninjured adult arteries, but is required for neointima formation after vascular injury. Pharmacologic β-catenin inhibitors hinder growth of human vascular SMCs. Thus inhibiting β-catenin has potential as a therapy to limit SMC accumulation and vascular obstruction.

Graphical abstract

Introduction

The smooth muscle cell (SMC) is a key participant in adult vascular remodeling due to its remarkable phenotypic plasticity¹. This flexibility allows the SMC to switch —in response to environmental cues triggered by injury or disease— from a contractile, differentiated, and quiescent state in normal vessels to a secretory, proliferative, and migratory form in injured vessels, usually associated with a reduction in SMC-selective marker gene expression^{1, 2}. SMC phenotypic plasticity is recognized as a major factor in the pathogenesis of prevalent cardiovascular diseases such as atherosclerosis, hypertension, restenosis, and saphenous vein graft disease^{2, 3}. Current models of vascular SMC phenotypic modulation are based on the integration of several complex environmental factors: increased growth factor and cytokine activity, disruption of cell-extracellular matrix interactions, neuronal influences, and mechanical forces¹. The in vivo molecular mechanisms that underlie this process, however, are not fully elucidated.

The protein β-catenin plays a dual function in the cell: it works as a transcriptional coactivator in the canonical Wnt signaling pathway as well as a structural component of the cadherin-catenin complex that mediates cell-cell adhesion⁴. β-catenin is known to play critical roles during development, adult homeostasis, and disease, particularly in cancer biology⁵. Interestingly, studies performed in the last 15 years suggest that β-catenin may also be a key regulator of SMC biology during adult vascular remodeling. β-Catenin protein levels increase in rat carotid arteries 7 days after balloon injury; this expression decreases by day 14 and is almost absent by day 28⁶. Overexpression of a degradation-resistant β-catenin inhibits apoptosis of vascular SMCs in culture and activates cyclin D1, and this effect is lost after expressing a dominant negative version of T cell factor 4 (Tcf4, also known as Tcf7l2); moreover, expression of this dominant negative Tcf-4 reduces the G1 to S transition of the cell cycle in vascular SMCs⁶. On the other hand, overexpression of N-cadherin, inhibitor of β-catenin and Tcf (ICAT, also known as Ctnnbip1), or a dominant negative Tcf-4 reduces proliferation of vascular SMCs, associated with decreased cyclin D1 expression and increased p21 (also known as Cdkn1a) levels⁷. Other cell culture studies support the idea that Wnt4 acting on frizzled class receptor 1 (Fzd1) activates β-catenin signaling and vascular SMC proliferation⁸. Carotid artery ligation in mice increases β-catenin signaling, which is evident 3 and 28 days after ligation in the media and intima, respectively, and vascular injury also induces Wnt4 and cyclin D1 expression, while loss of one Wnt4 allele in mice (Wnt4^+/−) reduces neointima formation and nuclear β-catenin expression⁸. In human aortic SMCs in culture, oxidized low-density lipoprotein (oxLDL) promotes β-catenin stabilization, nuclear translocation, and cyclin D1 expression, and β-catenin is required for oxLDL-induced proliferation⁹. Moreover, β-catenin expression correlates with proliferation markers in human atherosclerotic plaques⁹

More recent studies have complemented the earlier observations mentioned above, further supporting the idea that SMC β-catenin has a critical role in adult vascular remodeling. Several studies associate decreased vascular remodeling to factors that may decrease canonical Wnt signaling: 1) genetic loss of transglutaminase 2 reduces vascular SMC proliferation and migration, inhibits platelet derived growth factor receptor (PDGFR)/Akt1 and β-catenin activation, and attenuates neointima formation after carotid artery ligation¹⁰; 2) Wnt2^+/− and WNT1-inducible-signaling pathway protein 1 knockout (Wisp1^−/−) mice exhibit reduced intimal thickening after carotid artery ligation¹¹; 3) knockdown of kindlin-2 (also known as Fermt2) reduces carotid intimal hyperplasia after balloon injury in rats, and suppresses in vitro Wnt3a-induced vascular SMC proliferation and migration and expression of β-catenin target genes cyclin D1 and c-myc¹²; 4) Emodin, a plant-derived anthraquinone, inhibits carotid intimal hyperplasia after balloon injury associated with reduction of Wnt4, Dvl-1, and β-catenin protein levels, and seems to require microRNA-126 for its action¹³; 5) the orphan nuclear receptor Nur77 (also known as Nr4a1) opposes angiotensin II-induced vascular SMC proliferation, migration and phenotypic switching by attenuating β-catenin signaling¹⁴; and 6) the long noncoding RNA-growth arrest-specific 5 (GAS5) regulates hypertension induced vascular remodeling, while interacting with β-catenin and limiting its nuclear translocation in endothelial cells and SMCs in vitro¹⁵.

Other reports associate increased canonical Wnt signaling with enhanced SMC activities involved in vascular remodeling. These studies show that 1) treatment of vascular SMCs in culture with recombinant Wnt2b, Wnt4, Wnt5a, or Wnt9a, but not Wnt11, increases β-catenin protein levels and cell proliferation¹⁶; 2) treatment with recombinant Wnt2 increases Wisp1 mRNA levels and induces SMC migration¹¹; 3) balloon injury of carotid arteries in diabetic rats results in neointimal hyperplasia associated with increased Wnt4, Dvl-1, β-catenin, and cyclin D1 expression, and reduced p21 levels, a phenotype suppressed by up-regulation of microRNA-24¹⁷; 4) aortas from rats fed a high-fat diet have increased numbers of SMCs and lipid droplets, associated with higher mRNA levels for Wnt3a, β-catenin, Tcf-4, and cyclin D1, suggesting a role for Wnt/β-catenin signaling in hyperlipidemia-induced SMC proliferation¹⁸; and 5) in a model of Angiotensin II-induced arterial hypertension, complement C1q activates β-catenin signaling, which is required for vascular SMC proliferation¹⁹.

Overall these studies are consistent with a role for the canonical Wnt signaling pathway in SMC biology during adult vascular remodeling; however, the absence of in vivo studies using a SMC-specific, β-catenin loss of function approach, particularly in the response to vascular injury (for instance after carotid artery ligation or balloon injury), limits conclusions as to the direct and essential nature of β-catenin’s involvement in this context. Moreover, whether or not SMC β-catenin is essential during adult vascular remodeling has therapeutic implications. Inhibitors of β-catenin have been developed²⁰, so pharmacological inhibition of β-catenin function is feasible; this strategy would be ineffective if the biological role of β-catenin in adult SMC biology is redundant. On the contrary, if SMC β-catenin is essential in adult vascular remodeling, pharmacologically targeting β-catenin would have potential as a novel therapy for cardiovascular disease. We have recently shown that SMC β-catenin is required in vivo during mammalian development, since its loss precludes arterial wall formation and embryonic survival²¹. Here we have used a tamoxifen-inducible and tissue-specific genetic approach in the mouse to delete SMC β-catenin in adulthood, which has allowed us to test if it is required in the response to vascular injury. These studies show that SMC β-catenin is dispensable for the maintenance of uninjured adult vessels, but is required for neointimal formation after vascular injury. Moreover, β-catenin is required for expression of a set of genes reported to promote SMC invasion and neointimal growth, including matrix metallopeptidase 2 (Mmp2), and is necessary for SMC invasion in vitro; this complements the pro-proliferative and pro-survival roles of SMC β-catenin previously reported^{6–9, 16, 18, 21}. Finally, we found that inhibitors of β-catenin effectively reduced growth of mouse and human vascular SMCs in culture.

Material and Methods

Materials and Methods are available in the online-only Data Supplement.

Results

Adult vascular remodeling is associated with induced β-catenin expression in SMCs

β-catenin is expressed in and required by vascular SMCs during developmental artery formation²¹, so we wondered if it was still expressed in these cells in mature, uninjured arteries of adult mice. We harvested uninjured carotid arteries from 8–9 week old WT mice and evaluated β-catenin and SM22α expression by immunofluorescence. We found a clear signal for β-catenin in the endothelial layer but could not detect β-catenin expression in the media of normal arteries (Figure 1A). Then, we induced vascular remodeling in 8–9 week old mice by carotid artery ligation²², and found β-catenin expression not only in endothelium but also in SM22α⁺ cells in the neointima (Figure 1B). To complement these studies, we also evaluated β-catenin expression by Western blotting in lysates isolated from carotid arteries, and found increased β-catenin levels in ligated vessels compared to uninjured arteries (Figure 1C and Figure IA online-only Data Supplement). In addition, when we isolated SMCs from uninjured carotid arteries or aortas and cultured them under standard growth-promoting conditions, we observed β-catenin expression in these primary SMCs at cell-cell contacts, and in the cytosol and nucleus (Figure 1D and Figure IB online-only Data Supplement). Therefore, quiescent SMCs within an uninjured adult artery do not express detectable levels of β-catenin, in contrast to SMCs in active vascular remodeling or under growth-promoting conditions in culture.

Figure 1. β-Catenin expression is induced in SMCs during adult vascular remodeling.

A, Immunostaining for β-catenin and SM22α (SMC marker) in uninjured carotid arteries. B, Immunostaining for β-catenin and SM22α in carotid arteries 21 days after ligation. Dotted lines mark the internal elastic lamina. In A and B, arteries collected from adult control mice. Representative images of 6 mice per group. DAPI, nuclear staining. Squares indicate the regions magnified on the right. L, lumen. Left scale bar, 200 μm. Right scale bar, 50 μm. C, Western blotting for indicated proteins in total arterial lysates isolated from a pool of 3 uninjured or 3 ligated carotid arteries from control mice, 14 days after ligation. D, Immunostaining for β-catenin and SM22α in SMCs isolated from carotid arteries of adult control mice. DAPI, nuclear staining. Scale bar, 20 μm.

β-Catenin is not required in SMCs for maintaining the structure or differentiation state of arteries in adulthood

To test the significance of SMC β-catenin expression during vascular remodeling in adulthood, we pursued a loss-of-function approach in the mouse. Since loss of SMC β-catenin causes embryonic lethality²¹, we had to overcome developmental demise by crossing Myh11-CreER^T2 (tamoxifen-inducible SMC-selective Cre) mice²³ with β-catenin^flox/flox mice²⁴. Seven to eight week old Myh11-CreER^T2;β-catenin^flox/flox mice were injected with either tamoxifen or vehicle to obtain smooth muscle β-catenin knockout (iSMβCKO) or control mice, respectively. Tamoxifen induced Cre-mediated recombination in arteries and rendered a β-catenin (Ctnnb1) null allele (Figure 2A). Moreover, this recombination event was specific to SMC-rich blood vessels — the null allele was not detected in other tissues such as brain, liver, lung or tail (Figure IIA online-only Data Supplement). The presence of the null allele correlated with a significant reduction in β-catenin protein expression in total arterial lysates from a pool of 3 aortas per group (Figure 2B). Reduction in β-catenin protein expression in aortas was still evident 28 and 35 days after the first tamoxifen injection (Figure IIB online-only Data Supplement). In addition, β-catenin protein levels were also reduced in the bladder (an organ enriched in SMCs) of iSMβCKO mice, but were not affected in the tail (not particularly enriched in SMCs) or in the lung (an organ enriched in endothelial cells) (Figure IIC online-only Data Supplement). These findings are consistent with effective, SMC-selective inactivation of β-catenin. Then, we tested if loss of β-catenin in SMCs in adulthood had any repercussions for overall mouse health or on the structure of uninjured arteries. We followed iSMβCKO and control mice for 12 weeks after the first tamoxifen injection, and did not see lethality or morbidity in unchallenged animals (data not shown). Moreover, carotid arteries from iSMβCKO and control mice harvested 3, 4, and 5 weeks after the first tamoxifen injection did not look different in histological analysis (Figure 2C). In addition, immunostaining of carotid arteries from both groups showed clear expression of β-catenin in the endothelium, but did not detect β-catenin in SMCs within medial layers (Figure 2D). Finally, expression of SM22α, a marker of smooth muscle differentiation, appeared equal in iSMβCKO and control arteries (Figure 2D). These observations suggest that expression of β-catenin is minimal in SMCs of uninjured adult arteries, and that SMC β-catenin is dispensable for at least 5 weeks in mice for maintenance of SMC differentiation state and the structure of the arterial wall under normal conditions.

Figure 2. Loss of SMC β-catenin does not affect the structure or differentiation state of uninjured adult arteries.

A, PCR for indicated β-catenin alleles. DNA isolated from aortas of Myh11-CreER^T2;β-catenin^flox/flox mice 21 days after vehicle (Control) or tamoxifen (iSMβCKO) injection. HZ, Myh11-CreER^T2;β-catenin^WT/flox mouse injected with tamoxifen. B, Western blotting for indicated proteins in total arterial protein lysates isolated from a pool of 3 aortas per mouse group 21 days after vehicle or tamoxifen injection. C, Representative H&E stain of carotid arteries isolated from control or iSMβCKO adult mice 21 (control n=8, iSMβCKO n=16), 28 (control n=13, iSMβCKO n=15), and 35 (control n=6, iSMβCKO n=6) days after first injection of vehicle or tamoxifen, respectively. Squares indicate the regions magnified on the right. Scale bar, 100 μm. Scale bar of magnified images, 50 μm. D, Representative immunostaining for β-catenin and SM22α in uninjured arteries 21 days after first vehicle or tamoxifen injection. n=6 for both control and iSMβCKO. DAPI, nuclear staining. L, lumen. Scale bar, 50 μm.

Loss of β-catenin in SMCs restrains neointima formation after arterial ligation injury

Since SMC β-catenin expression is induced during adult vascular remodeling (Figure 1B), we wondered if SMC β-catenin expression was indeed essential for neointima formation. We performed carotid artery ligation injury in iSMβCKO and control mice seven days after the first tamoxifen injection, and evaluated β-catenin expression and neointima formation. By immunostaining, β-catenin expression was induced in SMCs within the neointima of injured arteries of control mice (Figure 3A), as seen before (Figure 1A); in contrast, β-catenin expression was abrogated in SMCs of iSMβCKO mice but readily apparent in the endothelial cell layer (Figure 3A). The latter was present in both uninjured and injured arteries lining the lumen of the vessel, and did not look different between control and iSMβCKO animals (Figure IIIA online-only Data Supplement). Interestingly, we found a significant reduction in neointima formation in iSMβCKO compared to control mice 14 days after injury (Figure 3B). We also treated Myh11-CreER^T2;β-catenin^WT/WT mice with the same tamoxifen and carotid artery ligation protocols (tamoxifen control group), and did not observe a reduction in neointima formation (Figure 3B). Thus, the decrease in neointima formation seen in iSMβCKO animals was not due to off-target, β-catenin-independent, effects of tamoxifen or CreER^T2, but rather, resulted from genetic inactivation of β-catenin in vascular SMCs. This reduction in neointima formation seemed not to be a simple delay in injury response, but more likely a reduction in its magnitude, because smaller neointimas were still seen 21 days after ligation (Figure 3C). These observations show that β-catenin expression in SMCs is required for neointima formation after injury in adulthood. To assess relevance to clinical cardiovascular disease, we looked for β-catenin expression in human arteries undergoing vascular remodeling after injury. We evaluated diseased post-mortem human coronary arteries proximate to sites of prior stent placement, and found β-catenin expression that coincided in some areas with smooth muscle α-actin (SMA, also known as ACTA2, a marker of SMCs) (Figure 3D and Figure IIIB online-only Data Supplement). These observations support the idea that β-catenin may contribute to clinical vascular pathogenesis.

Figure 3. Loss of SMC β-catenin reduces neointima formation after arterial ligation.

A, Immunostaining for β-catenin and SM22α in carotid arteries harvested 14 days after ligation. Arrowheads delimit the neointimal growth. DAPI, nuclear staining. Scale bar, 25 μm. B, Left: H&E stain of carotid arteries 14 days after ligation. Right: Neointimal growth expressed as the intima/media ratio. n=8 for control, n=16 for iSMβCKO, and n=6 for Tx Control (Myh11-CreER^T2;β-catenin^WT/WT mice injected with tamoxifen). C, Left: H&E stain of carotid arteries 21 days after ligation. Right: Neointimal growth expressed as the intima/media ratio. n=13 for control, n=15 for iSMβCKO. In B and C, arrows delimit the neointimal growth. Scale bar, 50 μm. *, p<0.05, by two-tailed t test. Data shown as mean±s.e.m. D, Representative immunohistochemistry for β-catenin and smooth muscle α actin (SMA), and H&E and Movat stains of 3 restenotic human coronary arteries.

SMC β-catenin promotes proliferation and prevents apoptosis in the neointima

In previous studies, we found that pro-proliferative and anti-apoptotic functions of SMC β-catenin are essential, in part by restraining p53 activity, for artery formation during embryogenesis²¹, so we wondered if β-catenin induced during adult vascular remodeling also promotes proliferation and survival of SMCs. We evaluated cell proliferation by measuring the expression of a marker of mitosis, phosphorylated histone H3 (pHH3), by immunohistochemistry, and assessing the expression of Ki67, a marker of cell proliferation, by immunofluorescence. We found a reduced percentage of pHH3⁺ cells or Ki67⁺ cells in the neointima of iSMβCKO compared to controls (Figure 4A and Figure IVA online-only Data Supplement). We also evaluated apoptosis by expression of cleaved caspase 3 and a TUNEL assay, and found an increased percentage of caspase 3+ cells or TUNEL⁺ cells in the neointima of iSMβCKO mice (Figure 4B and Figure IVB online-only Data Supplement). These findings support a pro-proliferative and pro-survival function of β-catenin in SMCs during vascular remodeling in adulthood. We could not detect cells expressing markers of the monocyte/macrophage lineage or neutrophils by immunostaining for colony stimulating factor 1 receptor (Csfr1) or CD68 within the vessel wall 14 days after injury (Figure IVC-IVE online-only Data Supplement), suggesting that inflammatory infiltrates are not a hallmark of this particular model of vascular injury.

Figure 4. Loss of SMC β-catenin decreases cell proliferation and increases apoptosis in the neointima.

A, Left: immunohistochemistry for a marker of mitosis, phospho-Histone H3 (pHH3), in carotid arteries 14 days after ligation. Arrows indicate pHH3⁺ cells. L, lumen. Scale bar, 25 μm. Right: percentage of neointimal cells positive for pHH3. n=9 for control. n=8 for iSMβCKO. *, p<0.05, by one-tailed t test. Data shown as mean±s.e.m. B, Left: Immunostaining for cleaved caspase 3, marker of apoptosis, and smooth muscle α actin (SMA) in carotid arteries 14 days after ligation. Arrowheads indicate cleaved caspase 3+ cells. Dotted line marks the internal elastic lamina. L, lumen. Scale bar, 25 μm. Right: percentage of neointimal cells positive for cleaved caspase 3. n=4 for control and n=3 for iSMβCKO. **, p<0.01, by two-tailed t test. Data shown as mean±s.e.m.

Loss of β-catenin in SMCs results in a gene expression pattern consistent with a quiescent, non-invasive, and apoptosis-prone phenotype

We previously found that vascular SMCs lacking β-catenin show defective growth in culture, are arrested in G₀/G₁ phase of the cell cycle, and are prone to cell death; this phenotype is associated with increases in p53 acetylation, transcriptional activity, and expression of the p53 target genes p21 (Cdkn1a) and Bax²¹. Moreover, loss of p53 in SMCs in vivo suppresses the effect of β-catenin inactivation, and substantially restores arterial wall formation²¹; this rescue effect, although significant, is not complete, suggesting that additional p53-independent mechanisms operate downstream of β-catenin. Consistent with this idea, we found that SMCs lacking β-catenin (Figure 5A) have decreased expression of several genes positively associated with vascular remodeling, including matrix metallopeptidase 2 (Mmp2), matrix metallopeptidase 9 (Mmp9) and sphingosine-1-phosphate signaling components —sphingosine kinase 1 (Sphk1) and sphingosine-1-phosphate receptor 1 (S1pr1) (Figure 5B). Interestingly, these genes have been shown to promote vascular SMC migration or invasion^25–32, and neointima formation after vascular injury^{27, 33–35}. S1pr1 also promotes vascular SMC proliferation³⁴. In contrast, loss of β-catenin resulted in higher expression of jagged 1 (Jag1) and connexin 43 (Gja1) (Figure 5C), which have been shown to promote a contractile SMC phenotype³⁶ and to prevent neointima formation³⁷. On the other hand, we did not find differences in expression of platelet derived growth factor receptor β (PDGFRβ), tissue inhibitor of metalloproteinase 1 (Timp1), Timp2, or N-cadherin between vascular SMCs lacking β-catenin and control cells (Figure VA-VC online-only Data Supplement). Altogether, these observations indicate that β-catenin is required in SMCs to promote a proliferative, anti-apoptotic, and invasive gene expression signature that favors neointima formation after vascular injury.

Figure 5. Arterial SMCs require β-catenin to express a gene set that promotes neointima formation.

A, Western blotting for β-catenin in mouse aortic SMCs (MASMCs) isolated from β-catenin^flox/flox mice and transduced with GFP-expressing or Cre-expressing adenovirus to generate control or β-catenin KO cells, respectively. B and C, qRT-PCR for Axin2, known β-catenin target gene, and genes that promote vascular SMC migration/invasion and neointima formation (B), or genes involved in direct cell-cell communication that inhibit neointima formation (C). Total RNA isolated from MASMCs. Rps13 was used as housekeeping control. n=6 biological replicates. *, p<0.05; **, p<0.01; ****, p<0.0001, by two-tailed t test. Data shown as mean±s.d. D, Representative Western blotting for indicated proteins in total cell lysates from MASMCs control and KO (left), and respective densitometric analysis of Mmp2 protein levels normalized to loading control (right). n=5 for control and n=4 for KO. **, p<0.01, by two-tailed t test. Data shown as mean±s.e.m. E, Left: Representative images of invading MASMCs in purple. Right: quantification of cell invasion by dissolving stained cells and reading the optical density of the dye/solute mixture at 560 nm; n=6 for both groups; *, p<0.05 by two-tailed t test. Data shown as mean±s.e.m.

Moreover, SMCs lacking β-catenin also showed decreased Mmp2 protein expression measured by Western blotting (Figure 5D), and rendered a conditioned media with decreased levels of Mmp2, which were measured by assessing the proteolytic activity (gelatinase activity) of the conditioned media by zymography (Figure VD and VE, online-only Data Supplement). These findings indicate that β-catenin is necessary in vascular SMCs for expression and secretion of Mmp2. It has been shown that Mmp2 is required for migration of vascular SMCs through a basement membrane barrier³⁰, and also necessary for human vascular SMC migration and invasion in vitro^{31, 32}. In previous transwell assays without a barrier, we found no differences in migration between SMCs lacking β-catenin and control cells²¹, so this time we assessed vascular SMC invasion, and found a reduced ability of β-catenin-deficient SMCs for crossing a basement membrane barrier in vitro (Figure 5E). Thus, β-catenin promotes expression and secretion of Mmp2 and cell invasion in vascular SMCs. As movement of SMCs from the media to the intima is a key feature of the response to vascular injury, it is possible that this pro-cell invasion function of β-catenin (Figure 5E) plus its pro-proliferative and anti-apoptotic activities (Figure 4 and Figure IVA and IVB online-only Data Supplement) contribute to neointima formation.

Inhibitors of β-catenin prevent growth of mouse and human arterial SMCs in culture

Since genetic inhibition of SMC β-catenin prevents SMC investment during artery formation²¹, and restrains neointimal growth after arterial ligation (Figure 3B and 3C), we hypothesized that pharmacological inhibition of β-catenin would inhibit growth of arterial SMCs. We treated mouse aortic SMCs (MASMCs) with increasing concentrations of validated β-catenin inhibitors, ICG-001³⁸ or PKF118-310³⁹—these inhibitors effectively reduce β-catenin/TCF transcriptional activity assessed with a TOPflash reporter assay in arterial SMCs²¹. Both inhibitors opposed MASMC growth in a dose-responsive manner (Figure 6A and Figure VIA online-only Data Supplement). Notably, ICG-001 and PKF118-310 also inhibited human coronary artery SMC (HCASMC) growth (Figure 6B and Figure VIB online-only Data Supplement). PKF118-310 appeared more potent than ICG-001 in both human and mouse SMCs (Figure 6A and 6B). At higher concentrations of PKF118-310, cell number actually decreased, consistent with induction of cell death. The inhibitory effect on SMC population growth observed with higher doses of ICG-001 and PKF118-310 seemed to be stronger than that observed with genetic deletion of β-catenin in vascular SMCs, which we have previously reported²¹, suggesting that these chemicals might mediate additional β-catenin-independent mechanisms that block SMC population growth as their concentration increases. To test this idea, we assessed the effect of these inhibitors on growth of control and β-catenin deficient MASMCs within the same assay. We found that β-catenin-deficient SMCs consistently exhibited impaired growth compared to control SMCs when both groups were treated with only vehicle (Figure VIC online-only Data Supplement). We also observed that treatment with 0.1 μM of PKF118-310 or 1 μM of ICG-001 only inhibited growth of control cells but did not affect that of β-catenin-deficient MASMCs (Figure VIC online-only Data Supplement), indicating a β-catenin specific effect. In contrast, treatment with 0.5 μM of PKF118-310 or 10 μM of ICG-001 affected growth of both control and β-catenin deficient MASMCs (Figure VIC online-only Data Supplement), indicating additional β-catenin-independent inhibitory mechanisms. Thus, there is a concentration threshold at which these β-catenin inhibitors gain additional inhibitory mechanisms; in our hands in MASMCs in culture, that threshold seems to be between 0.1 and 0.5 μM for PKF118-310 and between 1 and 10 μM for ICG-001. Overall our observations show that known pharmacological inhibitors of β-catenin limit growth of mouse and human vascular SMCs in culture.

Figure 6. Inhibitors of β-catenin hinder mouse and human vascular SMC growth.

A, Cell population growth of mouse aortic SMCs (MASMCs) in the presence of increasing concentrations of β-catenin inhibitors, ICG-001 or PKF118-310, or vehicle control. n=4 for 0.5 μM PKF118-310; n=16 for other indicated treatments and concentrations. B, Cell population growth of human coronary artery SMCs (HCASMCs) treated as in A. n=16. In A and B, *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001, by two-way ANOVA. Data shown as mean±s.e.m.

Discussion

These studies provide the first direct evidence that adult SMC β-catenin is not essential in vascular SMCs for the maintenance of the arterial wall under normal conditions in adulthood, but is required for neointimal formation after vascular injury. Although previous reports have shown that β-catenin expression and its transcriptional activity are enhanced after vascular injury^{6, 8}, whether or not β-catenin is actually essential for neointima formation had not been tested. The CreER^T2 driver we have used relies on regulatory elements for the most specific SMC marker gene, Myh11, with tamoxifen induction providing temporal control; β-catenin inactivation through this genetic intervention inhibits neointima formation after arterial ligation, in association with decreased cell proliferation and increased apoptosis. Together with our previous observations that SMC β-catenin is essential for arterial wall formation during embryogenesis²¹, these results define β-catenin as a key factor in vascular SMCs required for the formation and maturation of the arterial wall but dispensable for the maintenance of the mature state once it has been achieved. When the adult arterial wall is perturbed by injury or disease, however, β-catenin again becomes an essential promoter of SMC proliferation and survival. This differential requirement for β-catenin in normal vs. injured arteries in adulthood suggests that a therapeutic strategy aimed at inhibiting β-catenin function in SMCs should have a specific effect on injured vessels —that is, it would modulate remodeling of injured vessels but would not affect normal vasculature.

During adult homeostasis, the main function of SMCs in mature arteries is contraction, which regulates the vessel tone and diameter important for the conduction of blood to different organs, and the regulation of blood flow, vascular resistance and blood pressure. This function requires the expression of proteins involved in the contractile function and an appropriate extracellular matrix that confers the necessary mechanical properties. Disruption of those proteins and matrix would affect vascular homeostasis. We have shown that within this setting, characterized by quiescent non-migratory SMCs, SMC β-catenin is hardly expressed and not required. In contrast, SMC β-catenin is expressed and required in settings characterized by enhanced SMC proliferation and migration, such as during developmental formation and maturation of the arterial wall²¹ or the response to vascular injury (Figure 3). How SMC β-catenin expression is induced and maintained during development, how it is repressed after the vessel wall matures, and how it is induced after vascular injury are not fully understood.

Our studies of vascular injury in adulthood also support the idea that the requirement for β-catenin in vascular remodeling originates with cells that express Myh11, i.e., differentiated medial SMCs. According to our study design, tamoxifen activation of Myh11-CreER^T2-mediated recombination was completed prior to arterial injury; neointimal cells arising from a progenitor or stem cell, or by transdifferentiation from a different cell type, would not have expressed Myh11-CreER^T2 at the time of tamoxifen administration and would not lose the ability to express β-catenin. Thus our observations are consistent with the idea that most neointimal cells that accumulate after injury derive from medial SMCs, as opposed to resident or circulating progenitors or via transdifferentiation from another mature cell type⁴⁰.

We found that β-catenin expression is induced in the SMC-rich neointima after injury and is associated with enhanced cell proliferation and reduced apoptosis. We have recently shown that β-catenin represses p53 activity in vascular SMCs in vivo during artery formation²¹. Interestingly, both loss- and gain-of-function studies indicate that p53 opposes neointima formation after vascular injury due to its anti-proliferative and pro-apoptotic effects^41–46. It has also been suggested that mitogen-induced inactivation of p53 in vascular SMCs precedes the initiation of proliferation and migration of these cells from the media⁴⁷. As SMC phenotype during vascular remodeling after injury may resemble that found during embryogenesis³, it is very likely that β-catenin-mediated repression of p53 in SMCs is also operative in the context of vascular injury — this would explain in part the pro-proliferative and anti-apoptotic functions of β-catenin. However, our recent developmental studies also showed that inhibition of p53 was not the only β-catenin-dependent mechanism during artery formation, because the loss of p53 resulted in a partial suppression of the phenotype observed with loss of β-catenin²¹, suggesting that β-catenin may play roles beyond regulation of cell cycle and apoptosis in SMCs. Here we have shown that loss of β-catenin in primary arterial SMCs reduces the expression of a set of genes that promote migration/invasion or extracellular matrix remodeling, and enhance neointima formation after injury —Mmp2, Mmp9, Sphk1 and S1pr1^25–35; at the same time, β-catenin loss increases the expression of genes that promote a contractile phenotype of SMC and inhibit neointima formation —Jag1 and Gja1^{36, 37}. We have also shown that β-catenin promotes Mmp2 protein expression and secretion, and vascular SMC invasion in vitro. Altogether, these findings, plus observations in previous studies from our group and others, argue that β-catenin expression in SMCs supports a gene expression signature that broadly affects SMC phenotype and favors neointima formation after injury.

While evaluating β-catenin protein expression in normal and injured arteries, we consistently saw strong β-catenin expression in the endothelial layer. β-Catenin function has been studied in endothelial cell biology and shown to be important for angiogenesis, arterial specification, and blood brain barrier formation^{48, 49}. Interestingly, a recent report shows that genetically induced, endothelial specific deletion of β-catenin in adult mice results in systemic depletion of endothelial β-catenin but disrupts the endothelial barrier function only in the central nervous system (CNS) circulation, leading to neurological deficits and lethality within 2 weeks⁵⁰. This study suggest that with the exception of the CNS circulation, endothelial β-catenin function in adult vascular beds is dispensable for maintaining vascular homeostasis. The authors, however, evaluated only the pulmonary circulation as a reference, and did not examine other vascular beds in detail. As far as we know, no studies have evaluated the requirement of endothelial β-catenin in the response to vascular injury, but it is possible that β-catenin is also dispensable in this setting for vessels outside of the CNS. If this is the case, and keeping in mind our findings that SMC β-catenin is required for neointima formation after vascular injury, it suggests that local inhibition of β-catenin in arteries outside the CNS, for instance in the coronary circulation, may mediate a cell type-specific inhibitory effect in the vessel wall —that is, β-catenin inhibition would block growth of SMCs but would not affect endothelial homeostasis.

We have also shown for the first time that inhibitors of β-catenin function are effective growth inhibitors of cultured mouse aortic SMCs and human coronary SMCs. The two inhibitors we tested target distinct β-catenin interactions — with TCF4 for PFK118-310, and with CREB binding protein (CBP) for ICG-001. Disrupting either interaction proved to be effective. These strategies inhibit different aspects of the transcriptional function of β-catenin, but do not affect its cell adhesion function, which is consistent with genetic studies showing that the β-catenin’s signaling function is essential, while its cell-adhesion function is not sufficient for artery formation²¹. As SMC growth plays a key role in the pathogenesis of atherosclerosis, restenosis after angioplasty and stent placement, vein graft disease, and transplant arteriosclerosis^{1, 51}, inhibition of β-catenin could serve as a therapeutic strategy especially for those pathologies. Our studies also show that, depending on the concentration, these inhibitors can have β-catenin independent effects. Overall our observations with β-catenin inhibitors offer a strong rationale to study the efficacy and toxicity of this kind of pharmacological compounds in animal models of vascular injury.

In conclusion, we have demonstrated that β-catenin is dispensable in SMCs to maintain the structure and differentiation state of the arterial wall in uninjured adult vessels, but is essential for neointima formation after vascular injury. We have also shown that β-catenin promotes cell proliferation and prevents apoptosis in the neontima and is required for expression of a gene signature in SMCs that supports functions associated with neointima formation, including SMC invasion. Finally, pharmacological inhibition of β-catenin hinders growth of mouse and human SMCs in culture. Together, these observations support the idea that inhibition of β-catenin has potential as a therapeutic strategy in cardiovascular disease associated with intimal thickening.

Highlights.

• β-catenin is dispensable in SMCs to maintain the structure and differentiation state of uninjured adult arteries

• β-catenin is required in SMCs for neointima formation after vascular injury

• β-catenin inhibitors hinder growth of mouse and human vascular SMCs

Abbreviations

SMCs

smooth muscle cells

iSMβCKO

inducible smooth muscle β-catenin knockout

MASMCs

mouse aortic smooth muscle cells

HCASMCs

human coronary artery smooth muscle cells

TUNEL

terminal deoxynucleotidyl transferase dUTP end labeling

pHH3

phospho histone H3

CNS

central nervous system

DAPI

4,6-diamidino-2-phenylindole