Essential Role for Thymosin β4 in Regulating Vascular Smooth Muscle Cell Development and Vessel Wall StabilityNovelty and Significance
Rationale: Compromised development of blood vessel walls leads to vascular instability that may predispose to aneurysm with risk of rupture and lethal hemorrhage. There is currently a lack of insight into developmental insults that may define the molecular and cellular characteristics of initiating and perpetrating factors in adult aneurismal disease.
Objective: To investigate a role for the actin-binding protein thymosin β4 (Tβ4), previously shown to be proangiogenic, in mural cell development and vascular wall stability.
Methods and Results: Phenotypic analyses of both global and endothelial-specific loss-of-function Tβ4 mouse models revealed a proportion of Tβ4-null embryos with vascular hemorrhage coincident with a reduction in smooth muscle cell coverage of their developing vessels. Mechanistic studies revealed that extracellular Tβ4 can stimulate differentiation of mesodermal progenitor cells to a mature mural cell phenotype through activation of the transforming growth factor-beta (TGFβ) pathway and that reduced TGFβ signaling correlates with the severity of hemorrhagic phenotype in Tβ4-null vasculature.
Conclusions: Tβ4 is a novel endothelial secreted trophic factor that functions synergistically with TGFβ to regulate mural cell development and vascular wall stability. These findings have important implications for understanding congenital anomalies that may be causative for adult-onset vascular instability.
- mural cell
- mouse mutants
- smooth muscle differentiation
- vascular biology
- vascular smooth muscle
The development of a functional vasculature is an essential process during embryogenesis, perturbations in which result in fetal lethality or vascular disease after birth. The formation of systemic blood vessels occurs in a stereotypical fashion -endothelial tubes form through a number of mechanisms (angiogenesis, vasculogenesis, or intussusception).1 Endothelial cells then recruit mural cells comprising the subsets of vascular smooth muscle cells (VSMCs) and pericytes to the external wall of the vessel.2–4 These mural cells are required to provide structural support for the blood vessel and probably play a role in maintaining endothelial health and integrity. The establishment of a vessel wall is accomplished either through the differentiation of de novo mural cells from precursor populations or recruitment from a proliferating pool of mature cells. The former is thought to occur chiefly through the actions of endothelial secreted transforming growth factor-beta (TGFβ) and the latter through paracrine platelet-derived growth factor-B (PDGF-B).4 Typically, in the embryo, mural cells originate from the in situ differentiation of mesodermal tissues, which surround endothelial tubes.3,5 The exception to this is in the central nervous system, where blood vessels recruit to their outer layer via the migration of neurectodermal-derived mature mural cells, as typified by the development of the postnatal retinal vasculature.2,5
Consequences of failed mural cell recruitment range widely, depending on the degree of mural cell coverage. Midgestation lethality is seen in Alk5 knockout mice coincident with a failure to differentiate mural cells.6 However, mutants with a partial loss of mural cells can survive to later stages. The PDGF receptor-β–null mouse dies perinatally, probably due to hemorrhage and edema as a result of lack of structural support to blood vessels.7 In contrast, the endothelial-specific PDGF-BB knockout mouse survives into adulthood with reduced mural cell coverage but has deficiencies in renal and cardiac function.8 Nevertheless, in all of these described mutants, some degree of mural cell contribution to blood vessels is still observed, indicating that additional signals are required for mural cell differentiation from progenitor cells.
Thymosin β4 (Tβ4) is a 43–amino acid peptide encoded by the gene Tmsb4x on the X chromosome in mouse. It was initially identified as a G-actin–binding protein with the ability to regulate the cellular availability of actin monomers for the formation of polymeric F-actin.7 In recent times, however, novel functions have been ascribed to Tβ4, based on its ability to affect cell behavior when applied extracellularly or in a paracrine fashion.7 Notably, Tβ4 has been shown to improve cardiac function after ischemic injury.8–10 These cardioprotective effects may be due in part to the ability of Tβ4 to stimulate the differentiation of new coronary vascular cells, including coronary VSMCs, and consequently facilitate the process of neovascularization within the infarcted myocardium.9
Although several groups have shown that exogenous Tβ4 can promote angiogenesis both in vitro and ex vivo,11,12 the role Tβ4 plays in the systemic vasculature in vivo is unknown. We describe a requirement for endothelial Tβ4 in the differentiation of mesoderm-derived mural cells to contribute to the stability of the developing vasculature: a function mediated through synergy with the TGFβ pathway.
An expanded Methods section is available in the Online Data Supplement.
Mice were housed and maintained in a controlled environment, and all procedures were performed in accordance with the Animals (Scientific Procedures) Act 1986, (Home Office, United Kingdom). Global Tβ4 KO mice were generated by deleting exon 2 of the Tmsb4x locus and maintained on a C57Bl6/J background for more than 20 generations. Endothelial-specific Tβ4 knockdown mice were generated by crossing female mice, containing a previously described Tβ4 short hairpin RNA (shRNA)9 targeted to the hypoxanthine phosphoribosyltransferase locus with Tie2-Cre male mice.13 A cardiac-specific Tβ4 shRNA knockdown strain has been previously described,10 using a Nkx2–5-Cre knock-in14 crossed with the Tβ4 shRNA knockdown strain.
Histology, Immunofluorescence, and In Situ Hybridization
Standard histological, immunohistochemical, immunofluorescence, and in situ hybridization were performed on frozen or paraffin embryo sections or on whole-mount fixed embryos or retinas harvested at postnatal day 6, as described in full in the Online Supplement.
Real-time quantitative PCR (qRT-PCR) was performed according to a standard ΔΔCT protocol using SYBR green (Applied Biosystems). Primer sequences are given in the Online Supplement.
10T1/2 cells and A404 cells were maintained under standard conditions. For stimulation experiments, cells were treated for 4 days with PBS vehicle control, 1 μg/mL Tβ4, 2 ng/mL TGFβ, or 1 μg/mL Tβ4 plus 2 ng/mL TGFβ before RNA extraction using Trizol reagent (Invitrogen). For Smad 2 phosphorylation assays, A404 cells were serum-starved (0.5% fetal calf serum) overnight before stimulation, as above, for 20 minutes before protein extraction and standard Western blotting. For coculture experiments, early-passage human umbilical vein endothelial cells (HUVECs) were cultured in 50:50 A404:HUVEC medium (endothelial cell growth medium, PromoCell) for 24 hours; cells were counted, and an equal number of A404 cells were plated in each well. Thereafter, cells were cocultured in A404 medium for 4 days, with or without the addition of a neutralizing rabbit polyclonal anti-Tβ4 antibody (1:250; Immundiagnostik). 10T1/2 cells were transfected with Smad activity luciferase reporter plasmids (SA Biosciences) using Effectene transfection reagent (Qiagen) according to the manufacturer's instructions. After 16 hours, cells were serum-starved overnight (0.5% FCS in DMEM+Glutamax) before stimulation for 6 hours in the presence of 100 ng/mL Tβ4, 2 ng/mL TGFβ, 100 ng/mL Tβ4 plus 2 ng/mL TGFβ, or a PBS vehicle control. Smad activity-dependent firefly luciferase activity was measured by dual luciferase reporter assay (Promega).
Exon Array Analysis
Micro-arrays were performed on Affymetrix Mouse Exon 1.0ST arrays. Raw data were processed with Affymetrix expression console software before being analyzed for gene expression changes in Partek. Statistics were performed in R and pathway analysis performed using the Metacore software (Genego).
Statistical analysis was performed with Graphpad Prism software. Contingency tables were analyzed by χ2 test. Two-tailed, unpaired, nonparametric t tests were used for all other statistical tests.
Tβ4 Is Expressed in the Endothelium of the Developing Embryo
To investigate vascular expression of Tβ4 during development, whole-mount and section in situ hybridization was performed on midgestation embryos. Tβ4 was expressed in the dorsal aorta of the developing embryo from E9.5 onward (Figure 1A, 1B, and 1C). At E10.5, Tβ4 expression appeared relatively ubiquitous at low resolution (Figure 1D) but was detected in all developing blood vessels in the embryo (Figure 1D through 1F), including those within the developing limb bud (Figure 1E) and the intersomitic vessels (Figure 1F). Immunofluorescence analysis revealed that Tβ4 was predominantly expressed in the developing endothelium at E10.5, which was retained at E12.5. Relatively weaker expression was observed in the mural cell population of the dorsal aorta at E12.5, consistent with the onset of a more differentiated smooth muscle cell layer (Figure 1G and 1H).
Loss of Tβ4 Causes Hemorrhage Due to a Defect in Mural Cell Recruitment to Developing Vessels
To determine whether Tβ4 plays an essential role during the development of the systemic vasculature, a global loss of function model of Tβ4 was created by replacing exon 2 of the Tmsb4x gene with a neomycin resistance cassette resulting in a functionally null allele (Online Figure I). Hemizygous null male (Tβ4 −/Y) mice reach adulthood, but in reduced numbers, indicating a degree of embryonic lethality (Table 1). The loss of Tβ4 −/Y embryos was progressive from insignificant lethality before and up to E10.5 (21% versus 25% expected; P=0.482) through to significantly reduced recovery of mutants at E14.5 (16% versus 25% expected; P=3.01×10−4) and up to birth (17% versus 25% expected; P=1.01×10−5), indicating incomplete penetrance (Table 1). When E10.5 Tβ4 −/Y embryos were examined, it was apparent that approximately 10% of Tβ4 −/Y embryos displayed abnormal accumulation of blood in their hearts (Figure 2A and 2B). Mutant embryos also revealed extensive cranial bleeding in the midbrain region compared with +/Y controls (Figure 2C and 2D) and in axial sections there was clear evidence of both pericardial and coelomic cavity hemorrhage (Figure 2E through 2H). Severe vascular hemorrhaging at E10.5 was incompatible with continued survival accounting for the significant loss of Tβ4 −/Y mutant between E10.5 and E14.5 (Table 1). Over the course of the studies, we observed a reduction in embryonic lethality, reflecting an overall increase from approximately 60% to 80% survival of Tβ4 −/Y mutant mice to postnatal stages. This was coincident with a reduced incidence of the more severe hemorrhagic phenotype identified at E10.5, down from 44% to 20%, and reflected in an increased recovery of viable embryos at E14.5. The data in Table 1 represent current embryonic lethality versus survival rates against expected mendelian ratios.
Such hemorrhagic phenotypes can arise through defective mural cell investiture of developing blood vessels.15,16 Thus, Tβ4 −/Y embryos were examined for mural cell defects. Immunofluorescence studies for NG2, a marker that we initially confirmed as specific for mural cells and excluded from the developing endothelium (Figure 3A and 3B), showed that by E10.5 wild-type embryos displayed a dorsal aorta invested with NG2-positive mural cells (Figure 3C). This mural cell coverage was significantly reduced in Tβ4 −/Y embryos (consistent with the hemorrhagic phenotype observed in Figure 2H) in comparison to littermate Tβ4 +/Y controls (Figure 3C through 3F). The mural cell coverage directly correlated with severity of phenotype, observed in Tβ4 −/Y embryos, such that investiture was less significantly reduced in mildly affected embryos, in which the vascular wall appeared intact, as compared with severely affected mutants with evident hemorrhaging (Figure 3F). A reduced mural cell pool, impaired support of the mutant −/Y dorsal aorta, and correlation with severity of phenotype was further illustrated by immunohistochemistry for smooth muscle myosin heavy chain (SM-MHC) in transverse and sagittal sections Figure 3G through 3L). Globally impaired mural cell development in E10.5 Tβ4 −/Y embryos was confirmed by downregulation of a wide panel of mural cell markers in somite matched pairs of E10.5 Tβ4 +/Y and Tβ4 −/Y embryos, quantified by qRT-PCR. Significant changes in gene expression were observed for markers of VSMCs such as smooth muscle α-actin (SMαA) and smooth muscle 22α (SM22α as well as pericyte markers such as endosialin, NG2, CD13, and angiopoietin 1 (Figure 3M). Consistent with the ubiquitous expression of Tβ4 in developing blood vessels, Tβ4 −/Y embryos at E14.5, displayed statistically significant levels of dermal hemorrhage compared with littermate controls (Online Figure IIA through C). Immunostaining of the skin from E14.5 Tβ4 −/Y embryos for SMαA demonstrated a qualitative lack of mural cell-derivatives in comparison to specimens from control Tβ4 +/Y littermates (Online Figure IID and E). We next examined the integrity of the developing endothelium in Tβ4 +/Y versus Tβ4 −/Y aortas (Figure 4A and 4B). Even in selected areas of the walls of the dorsal aortas which appeared to have a disrupted endothelium by gross histological staining (Figure 4A and 4B; black inset boxes), immunofluorescence for the endothelial-specific adhesion molecule, VE-cadherin (Figure 4C and 4D), suggested that the endothelial cell contacts were intact across all genotypes, including both in mildly and severely affected mutants relative to Tβ4 +/Y controls (Figure 4E through 4G). This was supported by immunostaining for zona occludens 1 (ZO-1) (Figure 4H through 4N), a marker of endothelial tight junctions, which revealed an integral endothelium within Tβ4 −/Y aortas (Figure 4J through 4N) comparable to that of littermate controls (Figure 4H and 4I).
To confirm that the mural cell defects observed in Tβ4 −/Y embryos were due to a deficiency of Tβ4 in the endothelium, as suggested by the expression data (Figure 1), rather than a lack of Tβ4 acting cell autonomously in mural cells, we made use of a mouse strain that enables transcription of a Tβ4 shRNA under the control of Cre recombinase with the potential for tissue-specific knockdown of Tβ4.9 By crossing this mouse with a Tie2-Cre strain,13 knockdown of Tβ4 specifically in the developing endothelium was achieved (Figure 5). Initially, we observed a gross hemorrhagic phenotype (including cranial, pericardial, and coelomic cavity bleeding) in 30% of the endothelial-specific Tβ4-knockdown embryos at E10.5 (n=24 embryos analyzed), which was equivalent to the one-third documented for the global Tβ4 −/Y mutants. In addition, we recorded an overall incidence of embryonic lethality in Tie2-Cre Tβ4 shRNA animals of 12% (animals lost in utero between E10.5 and E14.5 and up to P1 according to expected mendelian ratios at birth) compared with 17% for the −/Y knockout animals (Table 1). The reduced incidence of loss in utero in the EC specific Tβ4-shRNA model may reflect both inefficient knockdown via the Cre/shRNA as demonstrated by qRT-PCR (Figure 5) and additional developmental roles for Tβ4 highlighted by global loss of function, such as an essential requirement in epicardial-derived coronary vasculogenesis.9 The latter was confirmed as a contributory factor to the differential embryonic lethality between E14.5 and P1 by analyzing the coronary vasculature of Tβ4 −/Y embryos at E14.5 (Online Figure III). We observed a significant incidence of VSMCs abnormally residing within the epicardial layer and reduced presence in the underlying myocardium, compared with Tβ4 +/Y littermate controls (Online Figure IIIA and B), a phenotype consistent with that described after myocardial shRNA knockdown of Tβ4.9
Serial sections through the developing dorsal aorta of Tie2–Cre Tβ4 shRNA embryos revealed defective NG2-positive mural cell recruitment to the vessel wall in comparison to Cre-only littermate controls (Figure 5A and 5B). The defective recruitment was equivalent to that observed after global knockdown of Tβ4 (Figure 3C through 3E). Knockdown of Tβ4 expression in Tie2–Cre Tβ4 shRNA embryos was confirmed by qRT-PCR on isolated dorsal aortas (Figure 5C) and protein knockdown via immunostaining for Tβ4 with Image J quantification (Online Figure IV); abrogation of Tβ4 mRNA and protein expression was incomplete at approximately 40% of control levels in each case (Figure, 5C; Online Figure IVI). Mural cell density around the wall of the aorta was quantitatively assessed by cell counts across serial sections, which revealed a significant reduction in mural cell coverage after endothelial-specific knockdown of Tβ4 (Figure 5D). Finally, as for the global knockout model, we assessed VE-cadherin expression by immunofluorescence to reveal that the endothelium was intact in Tie2–Cre Tβ4 shRNA aortas (Figure 5E and 5F). Collectively, these data reveal a non–cell autonomous role for endothelial Tβ4 in determining mural cell/VSMC coverage of the systemic vasculature during development.
Endothelial Tβ4 Stimulates Differentiation of Mesodermal Cells to a Mature Mural Cell Phenotype
Mural cell defects in developmental mutants have previously been attributed to either aberrant migration, survival, proliferation, or differentiation of mural cells.2 The postnatal retinal vasculature, consistent with its status as a central nervous system tissue, derives a mural cell layer via the migration of phenotypically mature mural cells along the developing vascular plexus. Whereas Tβ4 is expressed in the primary plexus vasculature of the early postnatal (P6) retina (Online Figure VA and B), P6 retinas from Tβ4 −/Y mice did not show any defect in mural cell coverage when compared with Tβ4 +/Y littermate controls (Online Figure VC through E), suggesting that impaired migration of mural cells is unlikely to be responsible for the defective mural cell coverage of Tβ4 −/Y embryos. In addition, mural cells displayed no overtly abnormal levels of proliferation or apoptosis, as measured by phospho-histone H3 or cleaved caspase 3 immunofluorescence staining, respectively, in either Tβ4 +/Y or Tβ4 −/Y embryos (Online Figure VIA through F), thus ruling out the possibility that the mural cell defects observed were due to defective mural cell hyperplasia or survival.
Blood vessels in mesoderm-derived tissues are thought to recruit mural cells via the in situ differentiation of mesoderm progenitors into mature mural cells.5 Given that mural cell defects in Tβ4 −/Y embryos are present only in mesodermal tissues, such as the aorta and subdermal vessels and not in tissues that derive their mural cells via migration, such as the postnatal retina, we investigated whether Tβ4 acts to induce mesoderm differentiation into a mature mural cell lineage. Initially, we assessed the ability of exogenous Tβ4 to stimulate the differentiation of a P19 embryonal carcinoma cell line known as A404,17 clonally selected for a propensity to differentiate into mural cells. Treatment of A404 mural cell progenitor cells with synthetic Tβ4 resulted in an increased number of SM22α- and SMαA-positive cells in culture compared with vehicle-treated controls (Figure 6A, 6B, 6E, and 6F). TGFβ was used as a positive control in these experiments and resulted in an equivalent increase in the incidence of SM22α- and SMαA-positive cells that was further augmented by the combined addition of both Tβ4 and TGFβ (Figure 6C, 6D, 6G, and 6H). The phenotypic changes in the A404 cells and coincident expression of the smooth muscle differentiation markers in culture were accompanied by a significant upregulation of mural gene expression, as quantified by qRT-PCR. Markers of both VSMCs such as SMαA and SM22α and pericytes, such as endosialin and desmin, were all significantly upregulated (Figure 6I)). These data indicate that Tβ4 can act in a paracrine fashion to stimulate mural cell differentiation from mesoderm.
Tβ4 Stimulates Mural Cell Differentiation by Enhancing the Activity of TGFβ Signaling
To gain greater insight into the molecular mechanisms underlying the role of Tβ4 in mural cell differentiation, gene expression arrays were carried out on E12.5 Tβ4 −/Y and +/Y embryos and compared with array data from Tβ4 +/Y and −/Y adult hearts to further facilitate the identification of Tβ4-dependent gene expression. Metacore software from GeneGo was used to highlight the expression changes of signaling factors in the embryo and adult heart datasets and identify the most likely underlying pathways. Four of the top 5 highlighted pathways included TGFβ (Table 2), previously highlighted as a key molecule involved in mural cell differentiation.17,18 Thus, we hypothesized that Tβ4 may exert its effects on mural cell differentiation by interaction with and/or modulation of the TGFβ pathway.
Activation of the canonical TGFβ pathway in mural cells leads to the transcription of stereotypical TGFβ-responsive genes such as plasminogen activator inhibitor-1 (PAI-1), Id-1, and c-myc.19,20 Levels of the mRNAs encoding these proteins can be used as a read-out of TGFβ pathway activity. Treatment of A404 cells with Tβ4 led to significant upregulation of these TGFβ target genes, comparable to that induced by TGFβ alone, as measured by qRT-PCR, and, moreover, combined Tβ4 and TGFβ acted to further significantly increase expression of all 3 target genes, indicating that Tβ4 can enhance TGFβ signaling in these cells (Figure 6J). We next examined cocultures of A404 progenitors with HUVECs, plus or minus Tβ4 neutralizing antibody. Importantly, Tβ4 antibody alone had no effect on A404 differentiation per se as confirmed by addition to A404 cultures followed by qRT-PCR (Figure 6I and 6J; black hatched bars). HUVECs express Tβ4 at high levels,21 and in the control setting, HUVEC coculture brought about differentiation of the A404 cells, as indicated by changes in cell morphology (adoption of a more elongated and differentiated phenotype) and expression of SMαA (Figure 6K and 6L). However, the addition of Tβ4 antibody significantly blocked the visible differentiation of A404 cells in culture (Figure 6M and 6N) and significantly reduced the expression of both mural cell markers (Figure 6O) and the TGFβ targets PAI-1, Id-1 and c-myc (Figure 6P).
Binding of TGFβ ligands to their cognate receptors leads to phosphorylation of Smad adaptor proteins.22 Treatment of A404 cells with Tβ4 alone resulted in increased levels of phospho-Smad2, to an equivalent level to TGFβ alone (Figure 7A). Most notably, Tβ4 in combination with TGFβ was able to induce higher levels of Smad2 phosphorylation than TGFβ treatment alone (Figure 7A). To investigate whether Tβ4 might be required for TGFβ signaling via Smad2 in vascular development, we assessed phospho-Smad2 levels in global Tβ4 −/Y embryos and observed a significant downregulation of phospho-Smad2 in both mildly and severely affected −/Y mutants compared with littermate (+/Y) controls (Figure 7B).
These data collectively suggest that endothelial Tβ4, acting in a paracrine fashion, regulates mural cell differentiation to VSMCs via synergistic activation of the TGFβ signaling pathway.
Inhibition of the TGFβ Pathway Correlates With the Penetrance of Hemorrhage in Tβ4-Null Aortas and Alterations in Downstream Signaling
To further determine whether TGFβ signaling was impaired with loss of Tβ4 in vivo, we examined aortas at E10.5 from Tβ4-null mutants, with or without evidence of vascular hemorrhage, compared with wild-type littermates, by immunostaining for SMαA and phospho-Smad2 (pSmad2) with Image J quantification (Figure 8A through 8F). Consistent with the previous NG2 and SM-MHC analyses (Figure 3), the VSMC layer was significantly reduced in the mutants with hemorrhaging as compared with those with a mild, nonhemorrhagic phenotype and wild-type littermates (Figure 8A through 8D). This was accompanied by a significant reduction in both the percentage of mural cells positive for pSmad2 and intensity of pSmad2 staining within the aortic wall (Figure 8E and 8F). Interestingly, in the mildly affected Tβ4-null mutants the percentage of pSmad2-positive cells was significantly increased as compared with the wild-type controls, suggesting some form of overcompensation to maintain the integrity of the aorta wall (Figure 8E). To provide further evidence for the ability of Tβ4 to activate the TGFβ/Smad pathway, 10T1/2 cells were transfected with a construct encoding firefly luciferase under the control of a Smad responsive element (SRE). Treatment with Tβ4 stimulated SRE-dependent reporter activity to a level significantly higher than that of PBS alone and in combination with TGFβ stimulated significantly higher activity than TGFβ treatment alone (Figure 8G). Consistent with the in vivo pSmad2 data (Figure 8A through 8F) expression levels of the TGFβ responsive genes, PAI-1, Id-1, and c-myc were found to be significantly downregulated in E10.5 Tβ4 −/Y embryos compared with somite-matched Tβ4 +/Y controls, as additional proof of defective TGFβ downstream signaling in vivo after loss of Tβ4 (Figure 8H).
This study reveals that in the absence of endothelial Tβ4, the secreted signals from the developing endothelium are no longer adequate to induce differentiation of mesodermal progenitor cells to a mural cell phenotype. At a molecular level, this is due to a deficiency in TGFβ signaling in the mesodermal progenitor cell population and manifests as impaired mural investiture of the developing systemic vasculature in general and the aorta in particular.
We initially mapped the vascular expression of Tβ4 predominantly to the developing endothelium. Global knockout of the Tβ4 gene in the developing embryo resulted in a proportion of the resulting E10.5 embryos exhibiting pericardial and coelomic cavity hemorrhage. An explanation for this hemorrhagic phenotype was evident through the observation that dorsal aortas in the Tβ4-null mice had reduced mural cell coverage in comparison to control littermates. The phenotype of the global knockout embryos was incompletely penetrant and subject to apparent compensation, such that we observed mutant embryos with both intact vasculature and those with a more severe hemorrhagic phenotype. This was reflected in survival rates across distinct stages in development. Lethality occurred both between E10.5 and E14,5 due to hemorrhaging, and beyond E14.5 up to birth whereby global knockout embryos revealed epicardium-derived coronary vascular defects, as previously described after shRNA knockdown of Tβ4 in the developing myocardium.9 Interestingly, survival of Tβ4 global knockout mice was in stark contrast to the lethality observed between E14.5 to 16.5 after optimal cardiac-specific knockdown.9 The previous RNAi model revealed a direct correlation between the severity of defective coronary angiogenesis and level of Tβ4 protein, and, moreover, reciprocal Tβ4 gain of function precisely restored the epicardial-derived vascular cell migration and differentiation disrupted by shRNA-induced loss of function. Thus we excluded, to all intents and purposes, the possibility of off-target effects, as can be associated with shRNA gene silencing, increasing the incidence of embryonic lethality. Rather, the differences between our global knockout and knockdown model are probably attributed to the fact that RNAi targeting in vivo, when sufficiently optimal to abrogate expression of the target gene, can result in a more severe phenotype than a corresponding global-null. Genetic ablation via homologous recombination through the germline, leading to complete loss of function from the outset in development, may be partially compensated for by functional orthologues, whereas RNAi-mediated efficient knockdown, occurring rapidly and at a defined developmental stage, may not be permissive for compensation.23 In a recent study, Tβ4 was described as dispensable for murine cardiac development and function after both global and cardiac-specific knockout.24 This contrasts with our findings describing partially penetrant vascular phenotypes after complete loss of Tβ4; however, we acknowledge that over successive generations the incidence of embryonic lethality decreased from 40% at the outset of the study (60% survival) to a current level of 20% loss in utero (80% survival) due to presumptive modifier effects. That said, there is a clear vascular phenotype in our Tβ4-null background both within the developing systemic and coronary vasculature, and therefore differences between our findings herein and that previously described may reflect allelic variation in the targeting strategy or genetic background-dependent events. It should be noted, however, that in the earlier study, the Tβ4 knockout aortas had an apparent reduction in α-SMA+ cell coverage within the vessel wall relative to wild-type controls at E14.524; while not commented at the time, this finding is consistent with our phenotype as described.
Embryonic mural cell defects generally arise as a result of perturbation of one or more of the proliferation, survival, migration, or differentiation of mural cells.3 A proliferation defect was ruled out on the basis of examination of E10.5 dorsal aorta mural cells for the presence of the proliferative marker phospho histone H3 and apoptosis excluded by a lack of cleaved caspase 3 expression. The anatomic location of mural cell defects also provided a clue to the role of Tβ4. The mural cell investiture of the postnatal retina is thought to rely on the migration of phenotypically mature mural cells, along a vascular plexus.2,5 As no defects were observed in the postnatal retina of Tβ4 −/Y mice, it is unlikely that Tβ4 is exerting its effects by reducing the migration of mural cells to their target locations. In contrast, the vascular defects in the Tβ4 −/Y mouse tend to occur in vessels which derive their mural cell coverage from the in situ differentiation of overlying mesoderm, notably the E10.5 dorsal aorta and the E14.5 dermal vasculature. Given that the normal induction of a mural cell layer around the dorsal aorta occurs over a narrow time window in murine development, between E9.5 and E10.5, and that Tβ4 −/Y mice first exhibit mural cell defects at E10.5, we hypothesized that the process of mural cell differentiation is aberrant in Tβ4 −/Y mice. Further insight into the cellular basis of Tβ4-induced mural cell differentiation arose from the observation that endothelial cell (EC)-specific Tβ4 knockdown recapitulated the aortic mural cell defects observed in the global knockout. In this instance, despite overlap in the mural cell phenotype, the embryonic lethality was not as significant as in the global knockout at stages between E10.5 and E14.5, reflecting reduced systemic vessel hemorrhaging due to incomplete knockdown of Tβ4. Beyond E14.5, through to birth, the EC knockdown of Tβ4 had no effect on the coronary vasculature, the cause of later onset lethality in the global knockout embryos. Importantly, a reduction in mural cell differentiation after EC-specific knockdown in this study was complemented by mural cell differentiation from mesodermal progenitors on Tβ4 treatment, further supporting the specificity of the particular shRNA used in the EC targeting, identical to that previously described for the myocardial knockdown and defective coronary vasculature.9
Thus, Tβ4, produced by the developing endothelium, acts in a paracrine fashion to stimulate the differentiation of overlying mesoderm into mature mural cells for vascular support. In support of this, exogenous administration of Tβ4 to an in vitro cell model of mural cell differentiation caused the induction of mature mural cell markers and Tβ4-neutralizing antibody in coculture experiments prevented HUVEC-induced differentiation of A404 mural progenitors. Bioinformatic analysis of gene expression data in the Tβ4 −/Y embryos implicated the TGFβ pathway as a molecular mediator of the defects observed in the Tβ4 −/Y mice. During embryogenesis, TGFβ has been identified as one of the central factors in the formation of a normal mural cell component of the vessel wall,3 and in this study, we implicate TGFβ signaling as an important modulator of Tβ4 function in the developing vasculature. Exogenous Tβ4 was observed to upregulate the expression of TGFβ target genes during Tβ4-induced A404 mural cell differentiation. Moreover, Tβ4 was able to stimulate the activity of a Smad-responsive transcriptional element in transfected 10T1/2 cells and induce higher levels of phosphorylation of Smad2 when used in combination with TGFβ than with TGFβ alone. In vivo expression profiling of E10.5 Tβ4 −/Y embryos revealed a global downregulation of the TGFβ pathway in the absence of Tβ4, and importantly, within the mural cell and VSMC layers of E10.5 Tβ4 mutant aortas, we observed correlative changes in TGFβ signaling via immunostaining for phospho-Smad2 with severity and penetrance of the Tβ4-mutant phenotype. In those Tβ4 −/Y embryos whose vasculature appeared phenotypically normal, we observed a compensatory increase in phospho-Smad2 expression, which may underlie the incomplete penetrance of the vascular phenotype after loss of Tβ4, whereas in the mutants with reduced mural cells and, accompanying hemorrhage, phospho-Smad2 was reduced in the VSMC layer. The implication of signaling downstream of TGFβ in this instance is consistent with a recent study which conditionally inactivated TGFβ type II receptor to reveal an essential role in the VSMC differentiation of the descending aorta such that mutant embryos displayed occasional aneurysms.25
We previously observed that in the developing heart, Tβ4 acts as a secreted myocardial factor on the migration and differentiation of epicardium-derived progenitor cells to form the smooth muscle layer of the coronary vasculature.9 More recently, we have also shown that Tβ4 is a key transcriptional target of the basic helix-loop-helix transcription factor Hand1 and its downregulation is, in part, responsible for the defects seen in the yolk sac vasculature of Hand1-null mutants.26 The present study allows us to expand the role of Tβ4 acting in these highly stereotyped situations to a more generalized function in vascular development. Tβ4 is a significant regulator of VSMC development, functioning as a paracrine signal from myocardial,9 extra-embryonic mesodermal,26 and endothelial lineages (herein), in an analogous manner to components of the Notch pathway.27 Thus, Tβ4 occupies a central role in the formation of a functional circulatory system such that absence or perturbation in Tβ4 function leads to serious deleterious consequences for the growth and stability of blood vessels.
The significance of Tβ4-dependent abnormalities in vascular development is their potential relevance to adult health and disease. Many pathological processes are caused by, or involve, the subversion of normal mural cell development,28 and an insult that results in depletion of medial VSMCs or a loss of functional VSMCs may be critical for adult aortic stability and vascular function. The model we propose (Figure 9) highlights the importance of secreted paracrine factors acting on mural cells and their progenitor populations, specifically in terms of their functional developmental role. Interest in identifying novel candidate molecules that function in these developmental pathways stems from the ability to agonize or antagonize their effects for therapeutic benefit. Thus, our discovery of Tβ4 as a secreted endothelial factor, which stimulates mesodermal progenitor differentiation into mural cells, can be seen in the context of not only a critical role in vascular development but highlights Tβ4 as a possible mediator of postnatal aortic function.
Sources of Funding
This work was supported by a British Heart Foundation MB PhD studentship (to A.R.).
We thank Shalini Jadeja and Marcus Fruttiger (UCL Institute of Opthalmology) for technical assistance with the postnatal retina studies; RegeneRx Biopharmaceuticals Inc, Rockville, MD (http://www.regenerx.com) for provision of clinical grade thymosin β4; and Nick Henriquez (UCL Institute of Neurology) for assistance in processing microarray data.
In May 2012, the average time from submission to first decision for all original research papers submitted to Circulation Research was 12.0 days.
The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.111.259846/-/DC1.
Non-standard Abbreviations and Acronyms
- human umbilical vein endothelial cells
- plasminogen activator inhibitor-1
- platelet-derived growth factor-B
- quantitative reverse transcription polymerase chain reaction
- short hairpin RNA
- smooth muscle alpha-actin
- smooth muscle 22 alpha
- smooth muscle myosin heavy chain
- Smad-responsive element
- thymosin beta 4
- transforming growth factor-beta
- vascular endothelial cadherin
- vascular smooth muscle cell
- Received October 27, 2011.
- Revision received June 11, 2012.
- Accepted June 18, 2012.
- © 2012 American Heart Association, Inc.
- Armulik A,
- Abramsson A,
- Betsholtz C
- Gaengel K,
- Genove G,
- Armulik A,
- Betsholtz C
- Stenzel D,
- Nye E,
- Nisancioglu M,
- Adams RH,
- Yamaguchi Y,
- Gerhardt H
- Manabe I,
- Owens GK
- Velasco S,
- Alvarez-Munoz P,
- Pericacho M,
- Dijke PT,
- Bernabeu C,
- Lopez-Novoa JM,
- Rodriguez-Barbero A
- Grant DS,
- Kinsella JL,
- Kibbey MC,
- LaFlamme S,
- Burbelo PD,
- Goldstein AL,
- Kleinman HK
- Moustakas A,
- Heldin CH
- De Souza AT,
- Dai X,
- Spencer AG,
- Reppen T,
- Menzie A,
- Roesch PL,
- He Y,
- Caguyong MJ,
- Bloomer S,
- Herweijer H,
- Wolff JA,
- Hagstrom JE,
- Lewis DL,
- Linsley PS,
- Ulrich RG
- Banerjee I,
- Zhang J,
- Moore-Morris T,
- Lange S,
- Shen T,
- Dalton ND,
- Gu Y,
- Peterson KL,
- Evans SM,
- Chen J
- Lindblom P,
- Gerhardt H,
- Liebner S,
- Abramsson A,
- Enge M,
- Hellstrom M,
- Backstrom G,
- Fredriksson S,
- Landegren U,
- Nystrom HC,
- Bergstrom G,
- Dejana E,
- Ostman A,
- Lindahl P,
- Betsholtz C
Novelty and Significance
What Is Known?
The stability of developing blood vessels depends on recruitment and differentiation of mural cells to support developing vascular smooth muscle cells.
Impaired support of the vessel wall results in vascular instability and can lead to hemorrhage.
Thymosin β4 (Tβ4) is an actin monomer-binding protein previously implicated in regulating yolk sac and coronary vessel development.
What New Information Does This Article Contribute?
Tβ4 is required for systemic vascular development.
An endothelial source of Tβ4 functions to maintain adequate recruitment and differentiation of mural cells/pericytes to maintain the stability of the developing aorta, cranial, and trunk vessels.
Tβ4 functions with the TGFβ pathway to regulate mural cell development and vascular wall stability.
Vascular instability due to impaired smooth muscle support can lead to aortic aneurysm, with a worldwide prevalence of 5% among the elderly, and, due to rupture and lethal hemorrhage, is associated with a 50% to 80% mortality rate. Until now, most studies have focused on the adult pathology and existing models of the disease have identified only a limited number of causative factors. We reveal a novel congenital mouse model in which either global or endothelial-specific loss of the actin monomer-binding protein thymosin β4 (Tβ4) directly affects TGFβ-induced vascular smooth muscle cell development, predisposing mutant vessels to wall defects ranging from lethal hemorrhage to vascular instability. Although previous studies have shown that Tβ4 plays an important role in smooth muscle contribution to extra-embryonic and coronary vascular beds, our findings reveal a more unifying function of Tβ4 in systemic vessel development per se and suggest the possibility of a new developmental paradigm for adult-onset vascular instability and aneurysm.