Funding Acknowledgements Research Articles

Targeting the histone methyltransferase SETD7 rescues diabetes-induced impairment of angiogenic response by transcriptional repression of semaphorin 3G

Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): Swiss National Science Foundation (n. 310030_197557), the Swiss Heart Foundation (n. FF19045) Background Peripheral artery disease (PAD) is highly prevalent in patients with diabetes (DM) and associates with a poor prognosis (1). Revascularization strategies failed to improve outcome, highlighting the need to develop new strategies to promote blood vessel growth (2). Histone modifications have emerged as key modulators of gene expression (3), however their role in angiogenic response in DM remains poorly understood. The present study investigates the role of chromatin remodelling in DM-related impairment of angiogenic response. Methodology Primary human aortic endothelial cells (HAECs) were exposed to normal glucose (NG, 5 mM) or high glucose (HG, 25 mM) levels for 48 hours. Unbiased gene expression profiling was performed by RNA sequencing (RNA-seq). Cell migration and tube formation were employed to study angiogenic properties in HAECs. Levels of the histone methyltransferase SETD7 and its chromatin signature at histone 3 on lysine 4 (H3K4me1) were investigated by Western blot and chromatin immunoprecipitation (ChIP). Pharmacological blockade of SETD7 was achieved by using the selective inhibitor (R)-PFI-2 while the inactive enantiomer (S)-PFI-2 was used as a control. Mice with streptozotocin-induced DM were orally treated with (R)-PFI-2 or vehicle and underwent hindlimb ischemia by femoral artery ligation. Blood flow was analysed at 24 hours, 7 and 14 days by laser Doppler imaging. Our experimental findings were also translated in endothelial cells (ECs) and gastrocnemius muscle samples obtained from DM patients with PAD. Results RNA-seq in HG-treated HAECs unveiled the histone methyltransferase SETD7 as the top-ranking transcript. SETD7 upregulation was associated with increased H3K4me1 levels as well as with impaired HAECs migration and tube formation. Both SETD7 silencing and inhibition by (R)PFI-2 rescued hyperglycemia-induced impairment of HAECs migration and tube formation, while SETD7 overexpression blunted the angiogenic response. RNA-seq and ChIP assays showed that SETD7-induced H3K4me1 enables the transcription of the angiogenesis inhibitor semaphorin-3G (SEMA-3G) by increasing chromatin accessibility to PPARγ. Moreover, SEMA-3G overexpression mimicked the impairment of angiogenic response observed during hyperglycemia. In DM mice with hindlimb ischemia, (R)-PFI-2 improved limb perfusion by suppressing SEMA-3G. Finally, RNAseq in vascular specimens from DM patients with PAD confirmed the upregulation of SETD7/SEMA3G signalling, whereas SETD inhibition restored angiogenic properties in ECs from DM patients. Conclusion SETD7 is a novel epigenetic target to boost neovascularization in DM patients with PAD.

SGK1 inhibition normalizes action potential duration in transgenic LQT2 rabbits but not in LQT1, suggesting a novel gene-specific therapeutic approach in long QT syndrome

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): SNSF (Swiss National Science Fundation) Background Increasing evidence suggests that secondary remodeling in Na+ and Ca2+ homeostasis may further enhance the propensity to develop arrhythmias in K-channelopathies LQT1 and LQT2 syndrome. In this context, the Serum and Glucocorticoid regulated Kinase-1 (SGK-1) was recently identified as a major regulator of sodium channels in cardiomyocytes. It is mainly activated in pathological conditions, leading to increased INaLate, prolongation of action potential duration (APD) and increased susceptibility to ventricular arrhythmias, recapitulating LQTS phenotype. SGK1 inhibition might hence be a novel therapeutic target also in genetic LQTS subtypes. Methods In cardiomyocytes (CM) isolated from wild type (WT), LQT1 and LQT2 adult rabbits patch-clamp experiments were performed to record AP with SGK1-inhibitor(SGK1-inh, 3 µM) or DMSO vehicle control. Results In the perforated patch configuration at 37°C, 3 μM of SGK1-inh shortened APD90 in LQT2 CMs at 1 Hz (DMSO-control [ms]: 477.3 vs treated cells[ms]: 362.8 p < 0.0001) and at 2 Hz (DMSO-cells: 325.7 vs treated cells:295.4,p <0.05). In contrast, in WT and LQT1 CMs, no effect on APD90 was observed neither at 1 Hz (WT, DMSO-cells: 329.2 vs treated cells: 301.6 p=ns; LQT1, DMSO-cells: 429.8 vs treated cells: 431.6, p=ns) nor at 2 Hz (WT, DMSO-cells:256.8 vs treated cells:242 p=ns; LQT1,DMSO cells:310.4 vs treated cells: 315.8,p=ns). Importantly, the SGK1-inhibitor effect in LQT2 rabbit CMs was so pronounced that it normalized APD to the wildtype level (LQT2+SGK1-Inh: 362.8 vs WT:329.2, ps=ns). Furthermore, short-term-variability in APD was markedly reduced by SGK1-Inh in LQT2, indicating an antiarrhythmic effect. Conclusions SGK1-inhibition exerts a beneficial APD-shortening effect only in LQT2, indicating a gene-specific efficacy. The underlying mechanism is still unclear. We hypothesize that the SGK1 pathway/INaLate activation in LQT1 might not be as active as in LQT2. We are currently conducting more experiments on potential genotype differences in SGK1 activity and SGK1-mediated INaLate inhibition to further elucidate the differences.Effects of SGK1-inh on AP characteristic

AAV-mediated gene delivery to the cardiac conduction system

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): NWO open technology program Biopace #18485 Background Gene therapy for cardiac arrhythmias has so far mostly focused on modifying the arrhythmogenic substrate in working myocardium or on creating biological pacemakers with viral vectors that did not selectively target the cardiac conduction system (CCS). As a result, CCS targeting by viral vectors is a relatively unexplored area. Aim In the current project, we explored the basic requirements for CCS targeting using adeno-associated virus (AAV) vectors. Methods Regular AAV9 and AAV-Myo4A capsid-pseudotyped AAV2 vectors (hereafter called AAV2/9 and AAV2/Myo4A vectors, respectively) expressing enhanced green fluorescent protein (eGFP) from a CMV or TNNT2 promoter, respectively, and an AAV2/9 vector expressing eGFP from a human GJA5 regulatory element were systemically injected in mice. After 4 weeks, hearts were harvested, fixed with paraformaldehyde and stained for TNNT2, eGFP, HCN4 and DNA to determine CCS transduction. Results Transduction of the sinoatrial and atrioventricular (AV) node using AAV2/9 vectors was very poor, while transduction was visible in the atrioventricular (i.e. His) bundle and working myocardium. In contrast, the transduction rate of the AV node and bundle using AAV2/Myo4A vectors were similar to that of the working myocardium. The human GJA5 regulatory element was able to restrict eGFP expression by AAV2/9 vectors to the AV bundle in 2/3 animals but the transduction efficiency was low. A third animal showed no transduction of any tissue, either indicative of the low transduction efficiency or of a failed injection. Conclusion This study not only provides proof-of-principle for CCS-targeted gene therapy but also shows that CCS transduction can be improved through AAV capsid engineering and be restricted to (parts of) the CCS by using specific transcriptional control elements.

Role of ADAR2-mediated innate immune responses in vascular inflammation and atherosclerosis

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Deutsche Forschungsgemeinschaft (DFG) Background Adenosine-to-inosine (A-to-I) RNA editing is an epitranscriptional regulator of RNA metabolism, catalyzed by the adenosine deaminases acting on RNA-1/-2 (ADAR1/2). Whether ADAR2 play a role in atherosclerotic processes remains elusive. Aim We examined the role of ADAR2 in vascular endothelial cell (EC) immunomodulatory responses and its biological relevance in murine and human atherosclerosis. Methods The role of ADAR2 in atherosclerosis was determined in a double ADAR2 and ApoE knockout mice fed with Western diet for 20 weeks. Three mouse genetic models (global and inducible vascular endothelial cell (iEC)-restricted ADAR2 and IL6ST knockout) were used to determine the role of ADAR2 in IL-6-related innate immunity. Expression levels of ADAR2 were measured in peripheral blood mononuclear cells derived from 533 individuals at risk for atherosclerotic vascular disease. Structural vascular measurements were used, including intima-media thickness (IMT), carotid wall maximum thickness (maxWT), and the number of atheromatous plaques by carotid and femoral artery ultrasonography as surrogate markers of subclinical arterial disease. Study participants were followed for major adverse cardiovascular events (MACE). Results Genetic ablation of ADAR2 resulted in a reduction of atherosclerotic plaque burden and necrotic core in the aorta and brachiocephalic trunk as well as reduced leukocyte infiltration in atherosclerotic plaques without affecting cholesterol levels in mice. Blood analysis revealed decreased pro-inflammatory cytokine levels including IL-6 after treatment of ADAR2/APOE KO mice compared to littermate control mice. Intravital imaging of IL-6-inflamed vascular endothelium showed that rolling and adhesion of leukocyte subsets to vascular wall were impaired by 2-fold in ADAR2-/- or iEC-ADAR2-/- KO mice. In vivo and ex vivo histological examination revealed that endothelial ADAR2 controls endothelial vascular adhesion molecule expression induced by gp130 signaling. ADAR2-deficient transcriptome revealed a selective upregulation of a conserved group of miRNAs targeting the IL6ST mRNA. Multilayered rescue studies including miR-inhibitors restored IL6ST levels after ADAR2 deficiency. In humans, increased ADAR2 expression was independently associated with the presence of CAD, higher C-reactive protein and mean carotid IMT after adjustment for traditional risk factors. Increased ADAR2 at baseline was prospectively associated with accelerated progression of subclinical atherosclerosis reflected by the increased number of carotid plaques at follow-up (OR=3.74, P<0.05). Finally, high ADAR2 levels were independently associated with higher MACE incidence across a median follow-up period of 54 months. Conclusion Posttranscriptional control of vascular inflammation by ADAR2 is integral for the development and progression of atherosclerosis in mice and humans. Our findings suggest that ADAR2 may serve as a promising therapeutic target for atherosclerosis.Role of ADAR2 in atherosclerosis

In vivo calcium and voltage mapping of the zebrafish heart

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Research Foundation Flanders (FWO) Purpose Inherited cardiac arrhythmias are characterized by a disturbance in the normal electrical functioning of the heart due to a dysfunction of cardiac ion channels, accessory proteins and/or structural components. This complex process is difficult to model in vitro, while in vivo mouse models differ substantially from humans with regards to cardiac electrophysiology. Thus, there is a need for alternative disease models in cardiac arrhythmia research. The zebrafish heart shows remarkable similarity to humans in its electrocardiogram and action potential (AP) morphology. Due to the optical translucency of zebrafish larvae, it is possible visualize the cardiac calcium and voltage dynamics across their entire hearts in vivo by use of genetically encoded calcium and voltage indicators (GECI and GEVI). We combined the fast kinetics of a next generation voltage reporter (Ace2N-mNeon) with a spectrally compatible calcium sensor (R-GECO) to create a sensitive high throughput method for in vivo characterization of the cardiac AP. Methods Ace2N-mNeon and R-GECO were expressed under control of a myocardial specific promoter via the Tol2kit system. The effect of sensor expression on normal physiology was assessed by patch clamp experiments. Fluorescence recordings were obtained with a Leica SP8 DLS microscope and a custom-built light sheet microscope. Our zebrafish sensor line was validated by experiments with quinidine drug exposure, as well as a novel model for long QT syndrome (LQTS) with a knock-in (KI) mutation in the cacna1c gene. Results Sensor expression did not significantly alter cardiac electrophysiology. The tg(myl7:Ace2N-mNeon/R-GECO) line demonstrated alterations in the heart rate, depolarization and repolarization induced by quinidine exposure. We observed a significantly prolonged ventricular repolarization in the LQTS KI line. Three-dimensional optical maps of AP characteristics across the entire zebrafish heart exhibited the physiological characteristics of the zebrafish cardiac electrophysiology, with activation initiated at the atrial venous pole and propagated towards the ventricular outflow tract, prolonged repolarization in the ventricles compared to the atria and slowed depolarization in the atrio-ventricular canal. Conclusion By expressing Ace2N-mNeon and R-GECO in the zebrafish heart, we were able to accurately detect drug- and mutation induced alterations in the cardiac AP, as well as AP characteristics across the entire heart, without compromising normal cardiac electrophysiology. Our transgenic line will provide a promising assay for future studies of cardiac arrhythmia.

MiRNA signature related to atherosclerotic lesion induced shear stress modifications in familial hypercholesterolemia patients with subclinical atherosclerosis: a bioinformatics systems biology study

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Institute of Health Carlos III (ISCIII) Background Patients with familial hypercholesterolemia (FH), even when appropriately treated, are at risk of premature clinical atherosclerotic cardiovascular disease (ASCVD) presentation. Although atherosclerosis (AT) is directly related to hemodynamic disturbances affecting the vascular geometry, there is scarce information on whether there is miRNA-mediated/flow-mediated modulation of atherosclerosis in FH. Purpose To identify a differential plasma miRNA signature related to fluid shear stress (FSS) in well controlled FH-patients with asymptomatic atherosclerosis. Methods FH-patients (N=49) with subclinical atherosclerosis, identified by computed tomographic angiography, and a reference cohort of non-FH subjects (N=32), both from a national FH-Registry, were included in the study. Coronary plaque volume and phenotype was analyzed by means of the QAngio CT imaging system. miRNAs associated to FSS were defined in silico by bioinformatics (KEGG, Cytoscape, GTEX portal, miRWalk, miRBase tools) and validated in plasma samples by RT-qPCR using TaqMan technology. Results By in silico analysis, 7 miRNAs targeting genes involved in features of functions (ROS production, oxidative stress, leukocyte transendothelial/migration-inflammation and vasodilation) of shear stress-induced cellular dysfunction and atherosclerosis (KEGG, hsa05418) were identified by miRWalk and miRBase databases. Of the 7 miRNAs, a 6-miRNA signature showed significantly lower plasma levels in FH-patients than in the non-FH group (P<0.01 for: miR-30e-5p, miR-378a-3p, let-7d-3p, and miR-550-5p; P=0.03 for miR-324-3p). In the FH-group, miR-324-3p, miR-30e-5p, miR-378a-3p and let-7d-3p significantly discriminated patients stratified by the plaque volume of non-calcified lesions (fibro-lipidic phenotype), using the median plaque volume as cut-off value. And even further, a high percentage (65-70%) of FH-subjects with low levels of the 4 miRNA had plaque volumes above the median. Interestingly, this miRNA-signature did not differentiate FH-patients stratified by the volume of calcified plaques. Conclusion FH-patients show a significantly lower content of an atherosclerosis/shear stress signature of 4 miRNAs targeting genes involved in subclinical atherosclerosis progression in FH.

Phosphoinositide 3-Kinase C2alpha controls cardiac contractility through regulation of beta2-adrenergic receptor recycling

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Doctoral fellowship Introduction Phosphoinositide 3-kinase C2α (PI3KC2α) is a ubiquitously expressed class II PI3K isoform, which has been previously shown to be involved in the control of vesicular trafficking. However, its role in the myocardium has not been investigated. The primary aim of this study is to elucidate the mechanism by which mechanism PI3KC2α controls cardiac pathophysiology. Methods Cardiomyocyte-specific PI3KC2α knockout (PI3KC2α KO) animals were generated by crossing mice expressing a tamoxifen-inducible Cre recombinase under the control of the αMHC promoter with PI3KC2αflox/flox mice. Zebrafish embryos were injected with a PI3KC2α (PI3KC2α morphants) or a control morpholino (control morphants) at 1-cell stage. HEK293 with stable overexpression of GFP-tagged β2-AR were generated and transfected with either scramble or PI3KC2α siRNAs. Results Our findings indicate that PI3KC2α KO mice display a reduced cardiac contractility compared to wildtype animals. Similarly, PI3KC2α morphants zebrafish exhibit lower heart rate and fractional shortening compared to controls, both at baseline and after isoproterenol (ISO) stimulation. A similar unresponsiveness to ISO was found in vivo in PI3KC2α KO mice where chronic treatment with the β-adrenergic receptor (β-AR) agonist failed to induce the classical β-AR-mediated remodeling, characterized by an increase of the left ventricular mass and of cardiomyocyte area. These findings suggest that PI3KC2α plays a critical role in the regulation of β-AR signaling. In agreement, cAMP levels failed to increase in response to ISO treatment in PI3KC2α morphants. Furthermore, silencing of PI3KC2α in HEK293-GFP-β2-AR cells resulted in an increased GFP-β2-AR internalization compared to control cells, both at baseline and after ISO stimulation. Interestingly, preliminary findings indicate that overexpression of constitutively active Rab11Q70L induced a redistribution of GFP-β2-AR at the plasma membrane in PI3KC2α-silenced cells. Conclusion Overall, we identify a key role for PI3KC2α in the control of cardiac contractility through the regulation of β2-AR trafficking to the plasma membrane through a Rab11-dependent mechanism.

Sirt1 in focus: unveiling molecular insights and therapeutic prospects in calcific aortic stenosis with sglt2i inhibitors

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): Fondazione Gigi & Pupa Ferrari ONLUS Background Calcific aortic stenosis (AS) affects 3% of older adults and lacks medical treatment. The well-known deacetylase Sirtuin1 (SIRT1) could be involved in many pathways linked to AS progression. Sodium-glucose co-transporter 2 inhibitors (SGLT2i), a class of antihyperglycemic agents, have been shown to reduce cardiovascular events and hospitalization for heart failure (possibly via SIRT1), but their possible benefits in AS are not yet known. Purpose Our study aims to uncover if SIRT1 plays a key role in AS progression and if SGLT2i can potentially slow the valve fibro-calcification processes, potentially via SIRT1 modulation. Methods RNA sequencing of control (n=27) and AS (n=39) aortic valves was performed. SIRT1 knockdown (SIRT1-) and overexpressing (SIRT1+) valve interstitial cells (VIC) were generated by CRISPR/Cas9. A calcification assay with or without dapagliflozin (dapa 1uM as SGLT2i) was employed to evaluate valve endothelial cells (VEC) and VIC crass-talk on calcification potential. Patient data were collected from the Lombardy regional healthcare utilization database and diabetic patients who were taking sulphonylureas (sul; not affecting SIRT1) and SGLT2i (acting on SIRT1) were selected. Patients were matched 1:1 by age, sex, and multisource comorbidity score. Kaplan-Maier analysis was used to generate time-to-event curves for aortic valve replacement (AVR) cumulative incidence at 2 years and Cox regression analysis was used to account for confounding effects. Results Whole tissue RNA-seq showed that SIRT1 could be a master regulator of multiple pathways linked to AS. Functional analysis and validations on mutant VIC revealed that inflammation, extracellular-matrix remodeling, and calcification mechanisms are directly regulated by SIRT1. Calcium assays unveiled an increment of calcification in SIRT1- VICs (+50.8%, p<0.01), while SIRT1+ VICs deposited less calcium (-23.6%; p<0.01) compared to wild-type ones. VIC exposed to dapa-treated VEC medium showed reduced calcification (-35.2%, p<0.01). In the dapa-treated VEC medium, we found an increase in NO levels (+19.9%, p<0.01) and VIC treated with an NO donor (detanonate 50uM) showed reduced calcification (-23.2%, p<0.05), while VIC treated with a SIRT1 inhibitor (EX527 25uM) showed an increased calcification (+68.1%, p<0.01). To corroborate these findings, we retrospectively enrolled 233,546 diabetic patients, and after matching, we obtained two groups (sul and SGLT2i) with 37,081 patients each. Adjusted Cox regression analysis revealed that patients taking SGLT2i had a lower AVR incidence than patients taking sulp with a hazard ratio of 0.65 (95%CI 0.47-0.89) resulting in a 35% risk reduction. Conclusion Our data highlights SIRT1 as a key regulator of fibro-calcific processes involved in AS and SGLT2i exhibit potential in slowing AS progression, via SIRT1 modulation. This study unveils promising molecular insights and suggests SGLT2i as a potential therapeutic avenue for AS.

Long-range gene editing of LMNA

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): Hartstichting (Dutch Heart Foundation) Background Mutations in the LMNA gene commonly affect the cardiac muscle tissue, which leads to the development of a severe dilated cardiomyopathy with life-threatening arrhythmias. The lack of response to conventional heart failure therapy in these patients indicates the urgent need for new therapies. Gene editing has the potential to become a curative treatment for these patients by correcting the genetic cause of the disease. Prime editing is a promising new gene editing technology, which uses a prime editing guide RNA (pegRNA) to target a specific genetic locus and creates edits using a reverse-transcriptase enzyme linked to a Cas9-nickase. Purpose Conventional prime-editing technology targets one mutation per pegRNA. However, with more than 600 disease-causing mutations described in LMNA, the feasibility of applying this technology to each mutation is limited. Therefore, we aim to target not a single, but a set of mutations with one single pegRNA. Methods The pegRNAs are optimised by selecting the optimal target regions, improving their stability and 3D structure. The in vitro editing efficiency is assessed using the fluoPEER reporter plasmid, transfected into HEK293T cells. Using flow cytometry, expression of mCherry fluorescent protein is detected following successful editing of the target mutation. Results Flow-cytometry analysis of treated cells revealed an editing efficiency of 12% with a long-range pegRNA targeting a full exon, compared to 15% with a conventional short pegRNA targeting the same mutation <10bp downstream of the Cas9 nick site. This editing efficiency using the same long-range pegRNA was retained at 9% when editing at a distant position >100bp downstream of the nick site. To increase the editing efficiency, we improved the stability of the pegRNAs by the addition of different 3’ structural motifs. This resulted in an increased editing efficiencies of 15% and 17% for nearby and distant mutations, respectively. Conclusion We show that it is possible to edit different mutations within a range of up to 120 bp with a single pegRNA. By optimisations of the pegRNA 3D structure and optimal selection of target locations we aim to further improve the editing efficiency of this system. This work could eventually provide a prime editing therapy able to cure disease in a subset of patients carrying different mutations in the same genetic region. For LMNA-associated diseases this would substantially reduce the number of pegRNAs that have to be developed to cover all mutations, greatly improving the feasibility of prime editing therapy.

MicroRNA-26b as novel regulator of cardio-metabolic diseases

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): IZKF Increasing evidence has shown that microRNAs (miRs) are fundamental players in cardio-metabolic diseases. MiR-26b has been associated with the development of metabolic dysfunction-associated steatohepatitis (MASH) and atherosclerosis, although causality has not been demonstrated yet. Therefore, we aimed to determine the causal role of miR-26b in MASH and atherosclerosis. Apoe-/-Mir-26b-/- mice were fed a 4- or 12-week western type diet (WTD). Plasma and liver samples were subjected to various assays to investigate the effects on lipid metabolism, inflammation and fibrosis. Additionally, miR-26b mimic-loaded lipid nanoparticles (LNPs) were used to rescue the phenotype. A 1.5-fold increase in hepatic lipid levels was observed in Apoe-/-Mir26b-/- mice, while plasma cholesterol/triglyceride levels remained unchanged. These results were further confirmed by an increase in steatosis. Additionally, the proinflammatory cytokines TNF-α, IL-6 and CXCL1 were significantly elevated, which coincided with increased macrophage infiltration into the liver. Moreover, loss of miR-26b resulted in more hepatic fibrosis. Finally, a treatment with miR-26b mimic-loaded LNPs reduced hepatic lipid levels, rescuing the observed phenotype. Additionally, Apoe-/-Mir26b-/- mice demonstrated a striking 10-fold increase in atherosclerotic lesion size. Consistent with a more advanced plaque phenotype, collagen content and relative necrotic core area were significantly increased in plaques from mice lacking miR-26b. Intriguingly, this increased atherosclerotic lesion size could also be rescued by repetitive injections of miR-26b mimic-loaded LNPs. Overall, our results demonstrate a protective role of miR-26b in cardio-metabolic diseases, thereby opening doors to future research and potential new treatment options using LNP technology.

The mechanistic role of connexin 43 in maturation of hiPSC cardiomyocytes

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – EU funding. Main funding source(s): ECR (European research council) and ReNew (the Novo Nordisk Foundation Center for Stem Cell Medicine supported by Novo Nordisk Foundation grants) Introduction Cardiomyocytes (CMs) differentiated from human induced pluripotent stem cells (hiPSCs) are functionally immature and this limits the application of these cells to cardiovascular research. Maturation can be improved by aggregating hiPSC-CMs into three-dimensional (3D) microtissues together with hiPSC-cardiac fibroblasts (hiPSC-CFs) and hiPSC-endothelial cells (hiPSC-ECs). These cardiac microtissues showed improved structural (increase of sarcomere length and organisation) and functional (more mature action potential profile) properties. This effect was only observed when the cardiac microtissues contained hiPSC-CFs. Since connexin 43 (CX43), the most ubiquitous connexin isoform in the heart, was significantly upregulated in hiPSC-CMs from cardiac microtissues, we hypothesise that mechanisms underlying hiPSC-CMs maturation involves the coupling of hiPSC-CMs with hiPSC-CFs through this protein. Purpose The aim of this study is to unravel the role of CX43 in the maturation of hiPSC-CMs in the context of 3D multicellular cardiac tissues. Methods We used the CRISPR/Cas9 technology to generate a CX43 knock-out (KO) hiPSC line. We designed two sgRNAs annealing to exon 2 of GJA1 gene, encoding CX43; indels generated by the non-homologous end joining after the Cas9 double strand break will result in a non-functional protein. Cells were differentiated in vitro into CMs, epicardial cells (EPI) and then CFs and characterized for the expression of cell-specific markers and compared with the wild-type parental control cells. Results and conclusions We screened 31 hiPSC clones and selected those in which both GJA1 alleles contained a deletion. Copy number variation analysis by ddPCR of the flanking regions confirmed that the deletion was specific and confined. The lack of GJA1 expression in the KO clones was confirmed by qPCR, while the lack of the CX43 protein was confirmed by immunofluorescence and Western Blot. This did not affect CX43-KO hiPSCs pluripotency state shown by the presence of pluripotent markers such as NANOG, OCT3/4, SSEA4. Additionally, CX43-KO hiPSCs were able to differentiate into the three germ layers. CX43-KO hiPSCs were successfully differentiated into hiPSC-CMs, -EPI, and -CFs. CX43-KO hiPSC-CMs and controls showed comparable gene expression of cardiac troponin T. Since immunofluorescent staining showed that CX43-KO hiPSC-EPI expressed the epicardial markers WT1 and ZO1, we continued to differentiate them into hiPSC-CFs, which expressed markers such as VIM and COL1A1, similar to the control. Differentiated CX43-KO hiPSC-CMs and CFs will be assembled into cardiac microtissues. Functional and structural assays that include patch-clamp and immunofluorescence will be performed on these microtissue to assess the effect of the CX43 KO on the maturation of hiPSC-CMs. Our results show that the lack of CX43 expression does not impede hiPSCs to differentiate into various cell types originating from the mesoderm germ layer.

Arrhythmogenic cardiomyopathy related phenotypes are detected and pharmacologically modulated in Galectin-3 zebrafish models

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Italian Ministry of University and Research Introduction Arrhythmogenic Cardiomyopathy (AC) is a rare inherited cardiac disorder characterized by fibro-fatty replacement and progressive loss of the ventricular myocardium, causing life-threatening arrhythmias, syncope and sudden cardiac death in young and athletes. About 50% of AC cases carry one or more mutations in genes encoding for desmosomal proteins. However, mutations in non-desmosomal genes have been identified in 1-3% of AC patients, such as Galectin-3 (Gal-3). Purpose For a better understanding of Gal-3 role in AC pathogenesis, we have generated a stable Gal-3 knock-out (KO) zebrafish line, able to reproduce cardiac abnormalities and cell signaling dysregulation characteristic of AC, on which environmental factors and candidate drugs can be tested. Methods The zebrafish Gal-3 mutant line was generated by CRISPR/Cas9 strategy and then phenotypically characterized. Phenotyping included heart activity characterization, by pyHeart4Fish software, and gene expression analysis using Real Time PCR and signaling pathway reporters. Immunofluorescence analysis was performed using an antibody against L-plastin, a leukocyte marker for the detection of inflammatory cells, and Acridine Orange/Ethidium Bromide (AO/EB) staining to check for apoptotic and necrotic events. Transmission Electron Microscopy (TEM) analysis was carried out to detect possible ultrastructural alterations. A pathway-directed pharmacological treatment was applied to try to rescue pathological phenotypes. Results The Gal-3 KO zebrafish line, analysed at larval stage, presented structural abnormalities, pericardial effusion and/or hemopericardium associated with reduced contractility, bradycardia and presence of arrhythmia events. In addition to Wnt/β-catenin signaling dysregulation, as seen in other AC forms, we also observed cell death and inflammation. Specifically, the analysis of the L-plastin inflammatory marker showed remarkable infiltration of inflammatory cells in the cardiac region of Gal-3 mutants, whereas AO/EB staining demonstrated elevated number of apoptotic and necrotic cells. Moreover, TEM analysis revealed "pale", disorganized and delocalized desmosomes in mutant adult hearts. Of note, pharmacological treatment at larval stage, using a Wnt/β-catenin signaling agonist, partially rescued a set of AC cardiac phenotypes (contractility, bradycardia, arrhythmia). Conclusions Overall, the Gal-3 KO zebrafish line confirms in zebrafish the strict connection between Gal-3 mutation and AC-like cardiac alterations. Moreover, cell death and inflammation, known to be involved in the remodelling of the myocardium, were observed. The pharmacological rescue of the Wnt/β-catenin signaling partially recovered a set of cardiac phenotypes, confirming this pathway as a key modulator of AC progression. In conclusion, our results strongly support the usefulness of the Gal-3 zebrafish mutant to clarify Gal-3 involvement in AC, offering future perspectives for AC treatment.

Cardiomyocyte Notch1 signaling is essential for protection of the maternal heart from peripartum stress

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Deutsche Forschungsgemeinschaft Background Peripartum cardiomyopathy (PPCM) is a life-threatening disease in women without known cardiovascular disease and is characterized by a sudden onset of heart failure (HF) before, under or after delivery. Previous studies suggest impaired endothelial to cardiomyocyte crosstalk as a major pathomechanism of the disease. Purpose Analyses of the role of Notch1 signaling for the endothelial-cardiomyocyte crosstalk in PPCM. Methods and results Serum from PPCM patients with acute HF reduced Notch1 expression in induced pluripotent stem cell derived cardiomyocytes (-20%, P<0.05) compared to serum from postpartum-matched healthy controls (PP-Ctrl). Beside miRNA-146a previously described to be up-regulated in PPCM plasma (6.8-fold, P<0.01), additional miRNAs targeting Notch1 signaling (miR-30a-3p, miR-34a, miR-200b-3p, miR-200c-3p, miR-429-3p, miR-449a) were up-regulated in PPCM plasma. Overexpression of these miRNAs in cardiomyocytes (CM) reduces Notch1 protein levels. Mice with a CM-restricted knockout of Notch1 (N1-KO) are born at the expected Mendelian ratio and nulli pari (NP) N1-KO female mice do not show a distinct cardiac phenotype at the age of 3 to 4 months compared to WT female mice (%FS WT: 41±5 vs N1-KO: 37±6; n= 9-12). However, they develop mild HF at the age of 6 months (%FS WT: 44±4 vs N1-KO: 31±8; n=11-14; P<0.01). Mated N1-KO females displayed normal cardiac function and LV dimensions in the last week of the first pregnancy. After 3 pregnancies and nursing periods, N1-KO developed severe HF (%FS WT PP: 30±5 vs N1-KO PP: 16±5; n=8-12; P<0.01), and show enhanced mortality (P<0.01). N1-KO-NP and N1-KO-PP displayed significant increase in the cardiac expression of the stress markers ANP and BNP compared to respective WT-NP and WT-PP. RNA-Seq analysis of N1-KO-PP revealed 1380 transcripts (adj. P<0.01) differently regulated genes, among them down-regulation of ErbB4, a known driver in PPCM onset. SiRNA-mediated knockdown of Notch1 in CM were associated with a decrease in ErbB4 protein expression. Treatment with bromocriptin (BR) was not able to improve survival and to rescue cardiac function and LV dimensions in postpartum N1-KO females (WT: 37±4, N1-KO: 23±10; n=7-8). Moreover, RNA-Seq analysis revealed that BR had no effect on the dysregulated cardiac mRNA. In conclusion, the pathophysiology of PPCM is associated with reduced cardiac Notch1 expression, which results in part from 16-kDa-prolactin mediated upregulation of miRNAs targeting Notch1. The reduced cardiac Notch1 expression contributes to PPCM as indicated with severe postpartum HF and high mortality in N1-KO mice. Inhibition of prolactin secretion with BR failed to rescue the development of PPCM in N1-KO mice suggesting that deregulated Notch1 is downstream of the 16-kDa prolactin-miRNA pathophysiology.

Preoperative examination of inflammatory biomarkers in relation to the prognosis of postoperative complications after heart transplantation

Abstract Funding Acknowledgements Type of funding sources: Public hospital(s). Main funding source(s): Semmelweis University Heart transplantation (HTX) is the definitive therapy for end-stage heart failure. Although complications have drastically decreased with modern immunosuppressive therapy, surgical and anesthetic techniques, significant complications such as graft failure, acute rejection or infection must still be expected. The purpose of the study is to examine cytokine pathways linked to inflammation from direct preoperative serum samples, which can serve as prognostic factors in relation to septic events and mortality in the early postoperative period. The study was conducted using the blood samples of 102 patients who underwent HTX. Blood samples were taken from the patients immediately before HTX. The amounts of a total of 32 cytokines and related proteins were determined using LEGENDplex assays. After that, we correlated the biomarker amounts determined in the serum samples with the end points we indicated. In patients undergoing HTX who developed septic events during postoperative intensive care, the preoperative MMP2 serum concentration was significantly higher compared to other patients undergoing HTX (n=23, p=0.04). Independent predictors of mortality in the intensive care unit (n=5) were preoperative MRP8/14 (AUC=0.8), NGAL (AUC=0.75), IGFBP-4 (AUC=0.73), cystatin C (AUC=0.7) and the MMP2 (AUC=0.74) serum concentrations. Patients transplanted from left ventricular mechanical circulatory support (LVAD) (n=11) were examined as a separate group, where early postoperative mortality was significantly higher (n=4, p=0.02), the independent predictor of which was the concentration of IGFBP-4 (p=0.04 , AUC=0.73). Based on our results, it can be said that the differences in the preoperative concentration of certain inflammatory biomarkers that can be measured in the serum show correlations with the postoperative septic events and mortality of patients undergoing HTX.

Untangling the inflammatory responses of liver x receptor (LXR) activation in human macrophages

Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): Rembrandt Institute for Cardiovascular Science (grant 2022) The regulation of cholesterol and fatty acid homeostasis is essential and the Liver X Receptor (LXR) plays a crucial role in this process. When activated, LXR promotes cholesterol efflux and suppresses the transcription of pro-inflammatory genes in mouse macrophages, making it a promising therapy against atherosclerosis. Studies indeed have shown that LXR agonism is effective in pre-clinical atherosclerotic mouse models. However, the limited data on human macrophages suggests that LXR activation may have additional pro-inflammatory effects in humans, and the mechanisms contributing to this are not yet fully understood. To investigate the inflammatory properties of LXR activation in human macrophages, monocytes were isolated from buffy coats of healthy donors and differentiated into macrophages using M-CSF. The macrophages were exposed to synthetic and endogenous LXR ligands, such as GW3965, T0901317, and desmosterol, respectively, followed by stimulation with lipopolysaccharide (LPS) for six hours. We used qPCR and ELISA to analyze experiments and plan to supplement our data with RNA-Seq and multiplex assays. Preliminary results show that LXR activation reduces LPS-induced activation of interferon-mediated genes like CXCL9 and CXCL10, similar to mice. However, it promotes the transcription of NF-kB target genes like IL1B and IL12B upon LPS activation. These findings suggest that LXR activation in human macrophages not solely suppresses inflammatory responses in macrophages and that LXR activation has different effects on the regulation of genes mediated by interferon or NF-kB. Our research aims to further uncover the regulatory mechanisms behind LXR activation in human macrophages as central regulators in atherosclerosis.

Human in vitro disease modeling of ATTR cardiac amyloidosis

Abstract Funding Acknowledgements Type of funding sources: Private company. Main funding source(s): Pfizer Alnylam Transthyretin amyloid cardiomyopathy (ATTR-CM) is a progressive disease with deposition of TTR amyloid fibrils in the myocardium and cardiac dysfunction with dismal prognosis. In ATTR-CM, either genetic mutations (variant; v) or ageing (wild-type; wt) leads to TTR instability. Late diagnosis and a lack of human disease models limit disease understanding and therapeutic discovery. This study aimed to model ATTR-CM in vitro by exposing various cardiac cell types and multicellular 3D cardiac spheroids to TTR fibrils, characterizing disease phenotypes and providing insights into the cellular mechanisms of disease progression. Human recombinant TTR (WT, V30M, V122I) and acid induced fibrils were generated and characterized. Seeding hiPSC-cardiomyocytes (CMs) and endothelial cells (ECs) on TTR fibrils resulted in cell viability reduction. Confocal microscopy revealed the extracellular attachment of TTR fibrils, leading to sarcomere disorganization, prolonged calcium transients and disrupted electromechanical coupling in hiPSC-CMs, and declined migration capacity of ECs showing aberrant cell morphology. Similarly, cardiac spheroids incorporating TTR fibrils resulted in decreased cell viability and prolonged calcium transients with lower amplitude. Our models show the discovery of early cellular ATTR-CM phenotypes, providing a novel approach to study the disease progression, potentially facilitating the identification of novel therapeutic targets and biomarkers.Cytotoxicity of TTR fibrils on hiPS-CMs

Asporin expression is associated with human atherosclerotic plaque integrity

Abstract Funding Acknowledgements Type of funding sources: Other. Main funding source(s): This work was supported by the Swedish Research Council, the Swedish Heart and Lung Foundation, Skåne University Hospital funds Background Calcification serves as a clinical marker of atherosclerotic plaque burden. Recent studies have also shed light on its intricate association with plaque vulnerability, likely depend on calcification structures and the surrounding environment. Asporin (ASPN) is a member of the small-leucine-rich proteoglycans that contains a unique and well-conserved stretch of aspartate (D) residues, which mimic the role of glycosaminoglycans. Via its D-repeat, ASPN can interact with calcium and several other extracellular matrix molecules, thus influencing collagen mineralization and TGF-β signaling. ASPN expression has been shown to be higher in asymptomatic compared to symptomatic atherosclerotic plaques, suggesting a potential protective role. However, the role of ASPN in maintaining plaque integrity remains unknown. Purpose Here we aimed to explore the role of ASPN in atherosclerosis, by investigating if ASPN was expressed in human carotid plaques and by determining if ASPN is associated with risk for postoperative cardiovascular events and death. Methods ASPN and TGF-β protein levels were measured by ELISA and Milliplex in 176 human plaque tissue homogenates obtained from the Carotid Plaque Imaging Project (CPIP) biobank. Plaque tissue sections were stained for ASPN, destabilizing (CD68, oil red O, glycophorin A) and stabilizing plaque components (Movat, α-actin). Plaque tissue RNA expression levels were assessed in 82 plaques using bulk RNA sequencing. Plaque levels of phosphate, elastin and collagen were measured using ELISA. Results ASPN protein levels were higher in plaques from asymptomatic patients compared to symptomatic patients (p=0.011). ASPN RNA levels correlated positively to RNA levels of the smooth muscle cell markers ACTA2, MYH11 and TAGLN. Moreover, protein levels correlated with plaque levels of phosphate (r=0.258, p=0.001), elastin (r=0.423, p<0.001), collagen (r=0.248, p=0.002) and stability factors like the TGF-β family members, TGF-β1 (r=0.176, p=0.035), TGF-β2 (r=0.581, p<0.001) and TGF-β3 (r=0.396, p<0.001). Furthermore, plaque ASPN protein concentration, correlated inversely with a histological plaque vulnerability index (r=-0.265, p=0.0029), which is ratio between the plaque area of destabilizing (macrophages, lipids, intra-plaque hemorrhage) and stabilizing plaque components (collagen and smooth muscle cells). Additionally, above median ASPN levels were associated with lower risk for postoperative cardiovascular events (p=0.008) and death (p=0.009). Conclusion Our findings suggest that elevated levels of ASPN might have a crucial role in maintaining plaque stability, affecting plaque calcification.

Ectopic Pitx2 expression and sinoatrial node dysfunction in mice recapitulating a novel cardiac syndrome identified in unrelated families with deletions in a gene desert on chromosome 4q25

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Dutch Research Councin/NWO Introduction Several unrelated families presenting with a complex, heterogeneous cardiac syndrome including sinoatrial node dysfunction and atrial fibrillation were found to share overlapping deletions in a 1.5 Megabase pair gene desert on chromosome 4q25. These overlapping regions contain evolutionarily conserved CTCF binding sites that modulate the separation of topologically associating domains containing the transcription factor-encoding gene PITX2 and noncoding RNA genes. Aim To understand mechanistically how the deletion in the gene desert causes sinoatrial node dysfunction and atrial arrhythmias. Methods Mice lacking the orthologue of the minimally overlapping genomic region deleted in affected patients were generated. Electrophysiology, morphology, tissue composition, and transcriptomes were analysed in this mouse model. Results These mice recapitulated major morpho-functional features of the syndrome, including sinoatrial node dysfunction and atrial arrhythmogenesis . Out of the 15 genes flanking the gene desert, only Pitx2 expression was deregulated in the sinoatrial node. During development, cardiac Pitx2c expression is normally restricted to the left side where it drives left sinus venosus and left atrial morphogenesis and prevents the formation of a left-sided sinoatrial node. The PITX2 locus has also been associated with atrial fibrillation in GWAS. Immunohistochemistry together with 3D volume analysis showed that Pitx2c was ectopically expressed in half of the pacemaker cardiomyocytes of the developing and postnatal sinoatrial node while its expression in the left atrium was maintained. Key transcription factors that drive pacemaker cell differentiation, Tbx3, Isl1 and Shox2, were lost where Pitx2c was ectopically expressed. Moreover, a conserved long non-coding RNA (Playrr) previously linked to Pitx2 regulation in the developing dorsal mesentery, was downregulated in the sinoatrial node of homozygous mutants. The downregulation of the pacemaker gene program additionally coincided with the activation of the atrial myocardial gene program in half of the pacemaker cardiomyocytes. Conclusion We generated a mouse model of a novel human cardiac syndrome including sinoatrial node dysfunction associated with the deletion of critical CTCF binding sites in a gene desert on chromosome 4q25. This deletion induced ectopic Pitx2c expression in the sinoatrial node alongside a substantial loss of pacemaker cardiomyocyte identity and a gain in a working atrial myocardium-like phenotype in the sinoatrial node of affected mice.Loss of pacemaker cardiomyocyte identitySinoatrial node dysfunction in mutants

Sex differences in cardiac ferroptosis after acute myocardial infarction in wild type and aryl hydrocarbon receptor knockout mice

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – EU funding. Main funding source(s): MICIN/Agencia Estatal de Investigación (DOI: 10.13039/50110001103) (NextGeneration, EU/PRTR) Background and Purpose Ferroptosis is a new regulated programmed cell death pathway, driven by iron-dependent phospholipid peroxidation, that plays an important role in the development of cardiovascular diseases. The aryl hydrocarbon receptor (AhR) is a ligand-activated transcription factor of the basic Helix-Loop-Helix-PER-ARNT-SIM superfamily. Upon ligand binding, AhR translocates into the nucleus and dimerizes with the AhR nuclear translocator; the heterodimer then binds to dioxin-response elements located in the promoter region of receptor-regulated genes, most of which are involved in xenobiotic metabolism. Beyond its role in metabolizing exogenous toxins as part of an adaptive chemical response, it is well known that AhR is activated by endogenous ligands and it has been postulated that AhR has important functions in physiological and/or pathological conditions. Here we studied sex differences in cardiac ferroptosis after acute myocardial infarction (AMI) in mice and the implication of defective AhR signalling on this process. Methods and Results AMI was induced in 8–10-weeks-old males and females AhR+/+ and AhR-/- mice by ligation of the left anterior descending coronary artery. Animal care and experimental procedures followed the principles of animal care (NIH publication Nº85-23 revised 1985) and the guidelines for ethical care of experimental animals of the European Union (2010/63/EU). Survival analysis showed that 16% of AhR+/+male mice and 14% of AhR+/+ female mice died 72 h after AMI surgery whereas the mortality was 100% in AhR-/-male mice and only 17% in AhR-/- female mice group. The treatment of AhR-/- AMI male mice with ferrostatin (1 mg/kg i.p) increased survival to 40%. Immunofluorescence analysis of cardiac staining levels of 8-hydroxy-2’-deoxyguanosine revealed higher oxidative stress in AhR+/+AMI male mice compared with AhR+/+AMI females, 24 h after surgery. The absence of AhR in AMI males showed an increased oxidative stress that was similar to that obtained in AMI AhR+/+ males but significantly greater than that observed in AhR-/- AMI females. Finally, analysis of differential expressions of classical genes related to ferroptosis in AhR-/- AMI male and AMI female mice, 24 h after surgery, showed, only in males, a reduced expression of ferritin heavy chain 1 and ferroportin, and an increased expression of phospholipase A2. Conclusion Our results show that loss of AhR in males but not in females favours the ferroptosis process associated with myocardial ischemia, highlighting a possible role for this nuclear receptor in cardiac remodelling and survival after AMI.

Increased perivascular fibrosis and pro-fibrotic cellular transition in intramyocardial blood vessels in patients with myocardial infarction

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): China Scholarship Council Background Myocardial infarction (MI) is pathologically characterized by irreversible myocardial necrosis caused by ischemia. Intramyocardial vascular dysfunction is believed to increase the risk of MI. In this study, we investigated intramyocardial blood vessels in deceased patients during the early and chronic phases of MI to explore the involvement of perivascular fibrosis and pro-fibrotic cellular transitions in the pathological process of MI. Methods Left ventricular tissue (LV) was obtained from the infarction area of autopsied patients with early-phase MI (3-6 hours old; n=23), chronic-phase MI (5-14 days old; n=12), and from non-infarcted controls (n=14). We quantified perivascular fibrosis using EVG staining and fibroblast activation protein (FAP) expression using immunohistochemistry. Co-localization of smooth muscle actin (SMA) or CD31 with fibroblast marker S100A4 indicative of transition of smooth muscle cells (SMC) and endothelial cells (EC) toward fibroblast-like cells in intramyocardial vessels was quantified using immunofluorescent microscopy. Results Compared to controls (61.03%), perivascular fibrosis was significantly elevated in MI patients (early-phase 77.20%; chronic-phase 82.58%)(p<0.0001), with the highest level observed in the infarct cores of chronic-phase patients (95.18%). Compared to controls, FAP expression was 3.5-fold higher in early-phase- and 15.3-fold higher in chronic-phase MI patients (p<0.05), wherein FAP expression again was highest in the infarct cores. Furthermore, in both early- and chronic-phase MI, SMC- and EC fibroblast-like cellular transition was increased. The fraction of SMA+S100A4+ vessels was significantly higher in early-phase MI (31.96%) and chronic-phase MI (21.9%: border zone; 37.25%: infarct core) compared to controls (8.95%)(p<0.05). Similarly, the fraction of CD31+S100A4+ area was significantly increased in early-phase MI (10.14%) and chronic-phase MI (8.31%: border zone 10.15%: infarct core) compared to controls (2.67%)(p<0.05). Conclusion We observed increased perivascular fibrosis, fibroblast activation and fibroblast-like cellular transition toward a pro-fibrotic phenotype of vascular cells in intramyocardial blood vessels in MI patients, that may contribute to the development of MI. The further increase of FAP expression and perivascular fibrosis, particularly in the infarct core, in the chronic-phase suggest that MI itself can exacerbate fibroblast activation and perivascular fibrosis, theoretically increasing the risk of reinfarction.FAP expression and perivascular fibrosispro-fibrotic vascular cellular transitio