Thymosin β4 protects against aortic aneurysm via endocytic regulation of growth factor signaling

Vascular stability and tone are maintained by contractile smooth muscle cells (VSMCs). However, injury-induced growth factors stimulate a contractile-synthetic phenotypic modulation which increases susceptibility to abdominal aortic aneurysm (AAA). As a regulator of embryonic VSMC differentiation, we hypothesized that Thymosin β4 (Tβ4) may function to maintain healthy vasculature throughout postnatal life. This was supported by the identification of an interaction with low density lipoprotein receptor related protein 1 (LRP1), an endocytic regulator of platelet-derived growth factor BB (PDGF-BB) signaling and VSMC proliferation. LRP1 variants have been implicated by genome-wide association studies with risk of AAA and other arterial diseases. Tβ4-null mice displayed aortic VSMC and elastin defects that phenocopy those of LRP1 mutants, and their compromised vascular integrity predisposed them to Angiotensin II–induced aneurysm formation. Aneurysmal vessels were characterized by enhanced VSMC phenotypic modulation and augmented PDGFR-β signaling. In vitro, enhanced sensitivity to PDGF-BB upon loss of Tβ4 was associated with dysregulated endocytosis, with increased recycling and reduced lysosomal targeting of LRP1–PDGFR-β. Accordingly, the exacerbated aneurysmal phenotype in Tβ4-null mice was rescued upon treatment with the PDGFR-β antagonist Imatinib. Our study identifies Tβ4 as a key regulator of LRP1 for maintaining vascular health, and provides insights into the mechanisms of growth factor–controlled VSMC phenotypic modulation underlying aortic disease progression.

Quantification of Tβ4 and LRP1 expression and PDGFRβ activation (phosphorylation at Tyr1021) in human aorta from AAA patients and matched omental artery from the same patients (n=10) and in human tibial arteries (n=4). Expression was determined by immunofluorescence, shown in Figure 4A,C. Association of Tβ4 and LRP1 was assessed across these arteries by proximity ligation assay (PLA; shown in Figure 4D). Significance was calculated using a one-way ANOVA with Tukey's multiple comparison tests * P ≤ 0.05; *** P ≤ 0.001. AAA: abdominal aortic aneurysm; PLA: proximity ligation assay; int: intima; med: media; adv; adventitia. Significance was calculated using one-Way ANOVA with Tukey's multiple comparison tests (B-C). **p≤0.01; ***: p≤0.001; ****p≤ 0.0001. Scale bar: A: 100µm, applies to all images. Figure 8. Representative examples of aortic rupture in AngII-infused mice. A higher incidence of rupture and dissection were observed in mice lacking Tβ4 and, overall, they occurred earlier, with severe disruption of elastin lamellae and aortic dilatation. These samples could not be used for assessment of aortic diameter or elastin integrity ( Figure  5) as the animals died prior to the experimental end point at day 10. Most deaths were due to aortic rupture, although occasionally dissection was detected in the same aorta, an example of which is shown for Tβ4 -/Y (day 8). Scale bar: 200µm, applies to all images. Figure 9. Mice lacking Tβ4 display more severe aneurysmal phenotypes, associated with augmented VSMC phenotypic switching and an apparent increase in apoptosis (in conjunction with Figure 5). Quantification of high and low molecular weight Caldesmon isoforms and contractile marker Calponin by western blotting after 5 days' AngII infusion (A). Immunofluorescence to assess medial layer morphology and VSMC phenotype, with the synthetic marker vimentin along with contractile marker, αSMA (B). Increased expression of the proliferation marker, Ccnd1, was detected in some Tβ4 -/Y aortas, compared with controls (qRT-PCR, C). TUNEL+ apoptotic cells were rarely detected within the medial layer; most aortas lacked TUNEL+ VSMCs, thus these images could not be described as representative. Endothelial (yellow arrows) and adventitial cells were more frequently TUNEL+ than VSCMs in this model (D). Data are presented as mean ± SD, with each data point representing an individual animal. Significance was calculated using a Mann Witney nonparametric test. * P ≤ 0.05; ** P ≤ 0.01. Scale bars: all 50µm. Figure 10. Endocytosis of LRP1 ligands is dysregulated in Tβ4 null aortas. LRP1 ligands, CTGF (A), HTRA1 (B) and PAI-1 (C) were assessed by immunofluorescence and quantified; representative of n=6. Data are presented as mean ± SD, with each data point representing an individual animal. Significance was calculated using a Mann Witney non-parametric test. * P≤ 0.05; ** P ≤ 0.01. int: intima; med: media; adv; adventitia. Scale bars: all 50µm. Figure 11. Despite augmented PDGFRβ signaling, Tβ4 -/Y and Lrp1 -/aortic VSMCs migration rates were reduced, compared with control cells. Migration rates were determined by scratch wound assay for Tβ4 +/Y , Tβ4 -/Y and Lrp1 -/aortic VSMCs, in response to 10ng/ml (A) or 50ng/ml (B) PDGF-BB. Representative images of same region of well at 0, 24, 48 and 72 hours after wounding. Data are presented as mean ± SEM from n=3 experiments, each from an independent VSMC isolation. Significance was calculated using two-way ANOVA with Dunnett's post hoc tests. ***: p≤0.001; ****p≤ 0.0001. Figure 12. siRNA-mediated Tβ4 knockdown. Tmsb4x Knockdown was assessed by qRT-PCR. A minimum of 85% knockdown by qPCR was set as the criterion for inclusion in the experiments in Figure 8. Knockdown data are shown for each of the individual experiments (n=3/4) and the mean qRT-PCR data. Figure 13. Restoration of normal PDGFRβ signaling ameliorates aneurysmal phenotype of Tβ4 null mice (in conjunction with Figure 11). Imatinib attenuated PDGFRβ pathway activity in aortas of AngII-infused Tβ4 +/Y and Tβ4 -/Y mice, confirmed by immunostaining for phospho-LRP1 (Tyr 4507; A) and phospho-PDGFRβ (Tyr1021; B). Western blotting of Calponin1 (contractile marker; C), to further assess VSMC phenotype. All samples were harvested after 8 days' AngII infusion. Data are presented as mean ± SD, with each data point representing an individual animal. Significance was calculated using two-way ANOVA with Tukey's multiple comparison tests. * P≤ 0.05; ** P ≤ 0.01. Scale bars: A, B: 50µm. Figure 14. Pro-inflammatory cytokines are unaltered by Imatinib treatment in the AngII model of aortic aneurysm. Serum from Tβ4 +/Y and Tβ4 -/Y mice were assayed by multiplexed automated ELISA for TNFα (A), IFNγ (B) CCL2 (C) and IL-6 (D) levels after 8 days' AngII infusion. Cytokine levels were not significantly altered regardless of whether mice received Imatinib or vehicle, nor were they altered by genotype (consistent with data in Figure 6). Significance was calculated using two-way ANOVA with Tukey's post hoc tests.  (5).

Supplemental
In vivo blood pressure measurement. One day after the MRI scan (exactly as previously described (6)), animals were anaesthetised with isoflurane (4%) and placed in supine position onto a heating plate with feedback control (Physitemp Instruments, NJ). Animals were kept at 37±1°C while oxygen and anaesthetics (1-2% isoflurane) were supplied via a nose cone (1 L/min). Body temperature and heart rate were recorded for the whole experiment using PowerLab with Chart 5 (ADInstruments, UK). A T-shaped middle-neck incision from mandible to the sternum was made. Blunt dissection was used to expose the left common carotid artery. A 5.0 Mersilk (Ethicon, NJ) suture was used to tie off the distal end of the artery while a micro vascular clip was used to occlude the proximal end of the exposed artery. A small incision near the distal end of the artery was made and a fluid filled tube (heparin; 100 U/mL diluted in 0.9% saline) with an inner diameter of 0.28 mm (Critchley Electrical Products Pty Ltd, Australia) was introduced into the artery and fixed in place with a suture. Following that, the micro vascular clip was removed and the arterial pressure was recorded using the MLT844 pressure transducer (ADInstruments, UK). After stabilisation of the signal for about 5-10 min an average of 1000 heart beats was used for analysis.
Aneurysm model. [8][9][10][11][12] week old adult mice were infused with 1mg/kg/day Angiotensin II (Sigma) or saline by subcutaneous implantation of an osmotic mini pump (Alzet) for 10 days. Animals were regularly monitored and weighed post-surgery until harvest. For the rescue experiment, 12-week old male Tβ4 -/Y and Tβ4 +/Y mice were randomly assigned into groups to be orally gavaged with either 10 mg/kg Imatinib mesylate or sterile water, daily for a period of 10 days, starting two days prior to AngII mini pump implantation.
Multiplexed automated cytokine ELISA Peripheral blood was collected at harvest by cardiac puncture, placed into chilled microcentrifuge tubes coated with 0.5M EDTA and kept on ice.
Samples were centrifuged at 1500 x g for 15 minutes and plasma supernatant collected into a fresh tube and stored at -80ºC. Using 25µl serum per assay, cytokine levels were quantified by multianalyte automated ELISA (Ella technology from ProteinSimple).

Human tissue sampling.
Subjects undergoing open abdominal aortic aneurysm repair were prospectively recruited from the Oxford Abdominal Aortic Aneurysm (OxAAA) study. The study was approved by the Oxford regional ethics committee (Ethics Reference: 13/SC/0250). All subjects gave written informed consent prior to the study procedure. Baseline characteristics of each participant were recorded. During surgery to repair the AAA, a wedge of abdominal omentum containing a segment of omental artery was identified and biopsied en bloc. Isolation of omental artery was performed immediately in the operating theatre. The omental artery segment was cleared of peri-vascular tissue and snapped frozen. Prior to incision of the aortic aneurysm, a marker pen was used to denote the cross section of maximal dilatation according visual inspection. A longitudinal strip of the aneurysm wall along the incision was then excised. The aneurysm tissue was stripped of the peri-vascular tissue and mural thrombus. The tissue at the maximal dilatation was isolated, divided into smaller segments, and snap frozen for subsequent analysis. As a comparison to aneurysm tissue, non-aneurysmal lower limb arteries were obtained from patients who required a major lower limb amputation as treatment for limb ischaemia. Written consent was obtained from each patient at the time of surgery for the retrieval of a small segment of tibial artery (from the amputated limb stump) for research analysis. Immunofluorescence Staining on Cryosections. Frozen sections of descending aorta or aortic sinus (base of the heart) were cut at a thickness of 10 µm, air-dried for 5 minutes and rinsed in PBS. Sections were permeabilized in 0.5% Triton-X100/PBS for 10 minutes, rinsed in PBS and blocked for 1-2 hours (1% BSA, 10% goat serum or donkey serum; 0.1% Triton-X100 in PBS). Sections were incubated with primary antibodies (diluted as below) at 4°C overnight. Sections were washed 5-6 times in 0.1% Triton-X100 in PBS (PBST) then incubated in secondary antibody for 1 hour at RT. Sections were washed 3 times in PBST and, incubated with 300 nmol/µL 4',6-diamidino-2-phenylindole (DAPI) in PBS for 5 minutes and rinsed a further twice in PBS. Slides were mounted with 50% glycerol:PBS and imaging was conducted on an Olympus FV1000 confocal microscope or a Leica DM6000 fluorescence microscope. Apoptosis was assessed using the Click-iT TUNEL Alexa Fluor 546 Imaging Assay kit (from ThermoFisher), according to the manufacturers' instructions.

Histological Sample Preparation
Cell Culture. Primary smooth muscle cells were isolated by enzymatic digestion of murine aortas, as previously described (7). After the first week, fetal bovine serum (FBS) was gradually reduced from 20% to 10% and cells maintained thereafter in DMEM (Gibco) supplemented with 10% FBS, 100 U/ml Penicillin-Streptomycin (Gibco), in a humidified incubator at 37°C and 5% CO2. An immortalised mouse vascular aortic smooth muscle cell line (MOVAS-1) was purchased from ATCC and cultured in DMEM (ATCC) supplemented with 10% FBS, 100 U/ml Penicillin-Streptomycin (Gibco) and 0.2 mg/ml G -418 (Sigma) at 37°C and 5% CO2. Immunoblotting. SiRNA-treated and PDGFβ-stimulated cells were lysed in RIPA buffer (50 mM Tris-HCl pH7.5, 100 mM NaCl, 1% NP-40, 0.1% SDS, 0.5% sodium deoxycholate, PhosSTOP and cOmplete protease inhibitor cocktail (Roche)). Western blots were performed using a standard protocol. Proteins were transferred by semi-dry transfer onto PVDF membrane. After blocking, membranes were probed with primary antibodies overnight. A modified protocol was used for Thymosin β4 Western blotting (8). Proteins were separated on Tris-Tricine gels and cross-linked with 10% glutaraldehyde, before wet overnight transfer onto nitrocellulose membrane. Membranes were crosslinked with UV light (254 nm, 10 min) before blocking and overnight antibody probing. Antibodies used were Rabbit anti-phospho-PDGFRβ following manufacture's protocol. Following Duolink® protocol, cells were co-stained for smooth muscle markers or endocytic compartments using the antibodies described above. Colocalisation was determined, using the ImageJ plug-in JACOP 2.0(9).

NMR Analysis. i) Protein expression and purification.
The pGEX-4T1 vector containing the N-terminus GST-tagged intracellular domain of LRP 1 (residues 4445 to 4544, a kind gift from Petra May, University of Freiburg, Germany)(10) was transformed into E. coli BL21 competent cells. Cultures were grown at 37°C in 15 N-NH4Cl, or 15 N-NH4Cl-13 C glucose minimal medium until an OD600 of 0.9 was reached. Protein over-expression was induced with the addition of 0.5M IPTG, and incubated at 18°C for 12 hours. LRP1-ICD was purified using GST affinity chromatography, followed by cleavage from the GST tag using 12.5 units of thrombin. Protein was then concentrated and purified using a gel filtration HiLoad 16/600 Superdex 75 pg column in a buffer containing 25 mM sodium phosphate pH 6.5, and 150 mM NaCl. Chromatography was performed on an Akta FPLC purification device. Following purification, the protein was evaluated using mass spectrometry to validate its size, followed by concentration for CD and NMR analysis. ii) NMR spectroscopy. LRP1-ICD protein was concentrated to 90 µM for Tβ4 titration analysis, or 300 µM for triple resonance backbone experiments. 1         Additional endogenous controls were used to validate loading. For densitometry, we normalised against β-tubulin or eIF4E, which correlated well with one another. β-actin showed some degradation (common for βactin) and is potentially unreliable due to the role of TB4 in binding actin.