Stabilisation of HIF signalling extends epicardial activation and neonatal heart regeneration

doi:10.21203/rs.3.rs-2496938/v1

Download PDF

Article

Stabilisation of HIF signalling extends epicardial activation and neonatal heart regeneration

https://doi.org/10.21203/rs.3.rs-2496938/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

In humans, new-born infants have the ability to regenerate their heart during early life. This is modelled in the mouse, where regenerative capacity is maintained for the first week after birth but lost thereafter. Reactivation of this process holds significant therapeutic potential, however, the molecular pathways that might be targeted to extend neonatal regeneration remain elusive. Here, we explore a role for hypoxia and HIF signalling on the regulation of epicardial activity which is essential for heart development and the response to injury. Hypoxic regions were found in the epicardium from mid-gestation, associating with HIF1α and HIF2α and expression of the epicardial master regulator Wilms’ tumour 1 (WT1). Epicardial deletion of Hif1a reduced WT1 levels in mutant embryos, leading to impaired coronary vasculature. Moreover, targeting of the HIF degradation enzyme PHD, through pharmacological inhibition with clinically approved drugs or epicardial-specific deletion, stabilised HIF and promoted WT1 activity ex vivo. A combination of genetic and pharmacological stabilisation of HIF during neonatal heart injury led to prolonged epicardial activation, increased vascularisation, augmented infarct resolution and preserved function beyond the 7-day regenerative window. Together, these findings suggest pharmacological modulation of HIF signalling may represent a viable therapeutic strategy for treating ischaemic heart disease.

Biological sciences/Stem cells/Regeneration

Health sciences/Cardiology/Cardiovascular biology/Cardiac regeneration

Hypoxia

heart

epicardium

development

regeneration

HIF-1α

HIF-2α

myocardial infarction

EMT

WT1

The epicardium, a mesothelial sheet covering the surface of the heart, plays essential roles during cardiac development and repair. Epicardial development is evolutionally conserved from fish to mammals. In mouse, formation of the epicardium initiates at embryonic day (E) 9.5 and as it expands to envelope the underlying myocardium, a subpopulation of epicardial cells undergo epithelial to mesenchymal transition (EMT) giving rise to epicardium-derived cells (EPDCs). These cells migrate into the sub-epicardial space and colonise the developing myocardium, supporting coronary vessel formation and cardiomyocyte proliferation and compaction ¹. Several transcription factors are activated in the developing epicardium, including Wilms’ tumour 1 (WT1), which is instrumental in the regulation of EMT. Accordingly, when WT1 is deleted in the epicardium cells are unable to migrate into the underlying sub-epicardial region and embryos die between E13.5–E15.5 due to cardiovascular failure. Mutant embryos display oedema, accumulation of blood in the systemic veins, pericardial haemorrhaging and impaired coronary vascular development ². EPDCs secrete a cocktail of mitogens and chemokines, such as insulin growth factor-2 ³, angiopoietin-1 and vascular endothelial growth factor A (VEGF-A) ⁴, to support cardiomyocyte proliferation and regulate coronary vessel patterning. In addition to their paracrine role, EPDCs can directly give rise to many cell types, including fibroblasts ⁵, vascular smooth muscle cells (VSMCs) ⁶ and to a lesser extent endothelial cells (ECs) ⁷. However, EPDC plasticity gradually decreases during embryogenesis and after birth the epicardium becomes quiescent. Embryonic genes, such as Wt1, are downregulated ⁸⁹ and EMT ceases ¹⁰.

The capacity for heart regeneration after injury appears to be species-specific. Zebrafish retain the ability to regenerate their hearts throughout life whereas adult mammals, including humans, fail to regenerate and instead replace necrotic muscle with scar tissue ¹¹. The loss of cardiomyocytes eventually compromises contractility of the remaining myocardium, leading to pathological remodelling, heart failure and death ¹². By contrast, the neonatal mouse can regenerate its heart after myocardial infarction (MI) up to 7 days after birth ¹³ and similar regenerative capacity is evident in new-born humans ¹⁴. In zebrafish, activation of the epicardium is an essential pre-requisite for heart regeneration ¹⁵ and in the adult mouse the epicardium becomes reactivated in response to cardiac injury to contribute to wound healing and tissue repair. Epicardial cells revert to an embryonic-like phenotype with injury; they proliferate, undergo EMT ¹⁶ and differentiate into a default fibroblast fate which in-turn stimulate ECs and the growth of coronary vessels by secreting pro-angiogenic factors, such as VEGF-A and fibroblast growth factor-2 (FGF-2) ¹⁶. Thymosin β4 (Tβ4) pre-treatment enhances EPDC-mediated neovascularization after injury ¹⁷ and induces limited de novo cardiomyocyte formation by differentiation from a progenitor population of epicardial origin ⁸. Upon injury Wt1 expression, a major hallmark of the embryonic epicardial signature, is increased ¹⁸ and its transcriptional activation is epigenetically regulated by SWItch/Sucrose Non Fermentable (SWI/SNF) chromatin–remodelling complexes containing Brahma-related gene 1 (BRG1) and Tβ4 ⁹.

The capacity for tissue regeneration has been associated with the levels of environmental oxygen ¹¹. Zebrafish and neonatal mouse hearts are exposed to relatively low-oxygen, or hypoxic conditions, that favour the maintenance of proliferative competency of several cell types, including cardiomyocytes ¹⁹. However, the increase in environmental oxygen after birth (in mammals) induces cardiomyocyte cell-cycle arrest¹⁹. Fate mapping of hypoxic cells in the adult mouse heart identified a rare population of cycling cardiomyocytes, contributing to new muscle formation²⁰. Moreover, Kocabas and colleagues identified the epicardium and subepicardial space as a hypoxic niche, housing a metabolically distinct progenitor cell population which utilised glycolysis to meet its energy demands²¹. More recently, exposure to hypoxia following MI was shown to induce heart regeneration in adult mice by metabolic reprogramming of cardiomyocytes leading to cell cycle re-entry²². Several oxygen-sensing pathways regulate cellular tolerance to hypoxia, including most notably the hypoxia inducible factor (HIF) family of transcription factors ²³. HIF is a heterodimeric transcription factor, consisting of α and β subunits ²⁴. The latter is constitutively expressed, whereas expression of the α subunit is oxygen-dependent. In human, mouse and rat there are three isoforms of the α subunit, and the best characterized are HIF-1α and HIF-2α. These bind to an identical core consensus in hypoxia response elements (HRE), but transactivate distinct, although partially overlapping, sets of genes ^25,26. Under well-oxygenated conditions (i.e. normoxia), prolyl hydroxylase domain proteins (PHD1-3) hydroxylate specific conserved proline residues within the α subunit creating docking sites for the binding of the von Hippel-Lindau (VHL) protein, tagging it for proteosomal degradation ²⁷. When oxygen levels drop PHD activity is inhibited and the α subunit forms a heterodimer with β, translocates to the nucleus and regulates expression of target genes, such as erythropoietin (EPO) ²⁴, vascular endothelial growth factor, VEGF ²⁸, and glucose transporter 1, GLUT-1 ²⁹. Hypoxia and HIF activation are evident throughout the developing heart at E9.5, but become restricted to the outflow tract (OFT), interventricular septum (IVS) and atrioventricular (AV) cushions, following formation of the coronary vasculature at E14.5 ³⁰. Interestingly, HIF-1α is predominantly expressed in the myocardium, whilst HIF-2α is primarily expressed in ECs ³¹, and inactivation of these individual isoforms result in differential effects on cardiac development. Embryos lacking HIF-1α die by E10.5 due to myocardial hypoplasia, reduced or absent myocardial trabeculation and a severely reduced endocardium ^32,33,34. Conversely, embryos which lack HIF-2α display variable outcomes, including extensive vascular defects, but do not exhibit major cardiac structural abnormalities ^35,36. In addition, loss of HIF-1β (ARNT) is embryonically lethal and causes abnormal cardiac morphogenesis, i.e. hypoplasia and malformed endocardial cushions ³⁷ and disruption of the major oxygen sensor PHD2 results in cardiac abnormalities, such as septal defects, poorly formed trabeculae and a thin myocardial wall ³⁸. Collectively these studies highlight the importance of the tight control of hypoxia/HIF signalling for normal cardiac development and function.

The role of HIF signalling specifically in the epicardium during mammalian heart development and regeneration remains elusive. Stabilization of HIF-1α in the avian embryonic epicardium in vivo and ex vivo inhibits EPDC migration into the myocardium through the disruption of VEGF signalling ⁴⁰. Additionally, hypoxia promotes EPDC differentiation into VSMCs in vitro through non canonical TGF-β signalling ⁴¹. A number of studies have described hypoxia-mediated Wt1 regulation, as potentially relevant for epicardial activation. Wt1 expression is upregulated in the heart and kidneys of rats exposed to hypoxia and an HRE binding site for HIF-1α has been identified within the promoter ⁴². More recently, an additional HRE binding site for HIF-2α has been described in intron 3 of the Wt1 gene using a neuroblastoma cell line ⁴³.

Given the correlative links between hypoxia and epicardial WT1 expression, we sought to investigate whether HIF signalling contributes to epicardial cell activation during heart development and following injury, and as such might represent a therapeutic target for extending the regenerative window¹³. Here, we demonstrate that the developing epicardium is hypoxic. Epicardial-specific deletion of Hif1a disrupted heart development with mutants displaying a significant reduction of WT1 positive cells in the epicardium and EPDCs in the underlying myocardial layer, along with a decrease in the number of coronary vessels, indicative of defective epicardial EMT and vascularisation. In the neonatal mouse, a gradual decrease in WT1 expression, concurrent with a decrease in HIF-mediated signalling, was observed over the first week of life. Global and epicardial-specific inducible genetic targeting of Phd2 alongside pharmacological inhibition of PHDs1-3, utilising drugs already approved for other disease indications (Roxadustat, also known as FG 4592 and Molidustat, otherwise known as BAY 85-3934)^44–47 stabilised HIF signalling in the injured neonatal heart. This was sufficient to maintain activation of the epicardium beyond postnatal day (P) 7, improve cardiac remodelling and preserve function post-MI. These findings provide novel insight into the molecular mechanism regulating epicardial activation in the mouse heart and suggest that modulation of HIF signalling in the dormant epicardium may enhance cardiac repair and regeneration, representing an attractive therapeutic strategy for the treatment of ischaemic heart disease and heart failure.

The epicardium is hypoxic during development

In order to study the role of low oxygen tension in the developing heart, we initially investigated the physiological levels of hypoxia in situ from mid- to late gestation stages, when the epicardium forms and epicardial EMT takes place. Following in utero treatment with the hypoxia marker Pimonidazole-HCl (Hypoxyprobe-1, HP1) ⁴⁸, embryonic hearts were harvested from E12.5 to E18.5 and probed with an anti-pimonidazole fluorescence-conjugated monoclonal antibody (HP1), alongside immunostaining for the spatiotemporal expression of HIF-1α and HIF-2α. At E12.5, when the myocardium is highly trabeculated and the epicardial layer fully established, HP1 staining revealed extensive regions of hypoxia, mostly in areas of dense myocardium, such as the compact wall of the atrioventricular (AV) groove. At this stage, HP1 signal co-localized with the epicardial marker WT1 and both HIF-1α (Supplementary Fig. 1a) and HIF-2α at the apex of the heart, while we observed a strong HIF-1α signal in the AV groove (Fig. 1a). Also, HIF-2α expression was present in the myocardial compartment (Supplementary Fig. 2a). At E14.5, extensive HP1 staining was detected in areas of thicker compact myocardium, such as the interventricular septum (IVS) and co-localization with WT1 and HIF-1α in the epicardium, particularly in the apex (Supplementary Fig. 1b) and AV groove regions, was also observed. HIF-2α and HP1 co-localised in hearts at E14.5, albeit mainly in the myocardium (Fig. 1b, Supplementary Fig. 2b). The analysis of hearts at E16.5 and E18.5 revealed that the HP1 signal became progressively weaker and restricted almost exclusively to WT1-expressing epicardial cells at later stages (Fig. 1c, d). Similarly, HIF-1α localized in discrete regions of the outermost layer of the heart (Supplementary Fig. 1c, d), while HIF-2α immunoreactivity was detected in both the epicardium (apex) and throughout the myocardium at E16.5 (Supplementary Fig. 2c) but restricted to a few isolated myocardial cells by E18.5 (Fig. 1c, d; Supplementary Fig. 2d). Taken together, these data suggest that WT1 + cells comprising the epicardial layer of the developing heart become hypoxic from mid-gestation onwards, preferentially expressing HIF-1α whereas the expression of HIF-2α appeared to be mostly within the forming myocardium.

HIF-1α epicardial deletion disrupts EMT and associated coronary development.

Based on the observation that the epicardium is hypoxic during mid-late gestation and that HIF-1α is expressed in cells localizing to the epicardium, we sought to investigate a functional role for HIF-1α on epicardial development. To target the expression of Hif1α in the epicardium, we crossed mice harbouring a tamoxifen-inducible epicardial-specific CreERT2 driver, Wt1^{CreERT2/+ 49}, with mice in which the exon 2 of Hif1α gene is flanked by LoxP sites (Hif1α^fl/fl). Pregnant Wt1^CreERT2/+;Hif1α^fl/fl females were injected with tamoxifen at E9.5 and E10.5, to target epicardial development, and embryos were collected at E16.5. Deletion of Hif1α did not alter HIF-2α expression in WT1 + cells (Supplementary Fig. 3a-b). Immunostaining against the epicardial markers WT1 and podoplanin (PDPN) revealed a significant reduction in the number of WT1 + cells in hearts from Wt1^CreERT2/+;Hif1α^fl/fl mutants (KO) as compared to littermate controls (CTR) (Fig. 2a-b). Quantification revealed a decrease in the percentage of WT1 + cells in both the epicardial layer (Fig. 2c; Mean ± SEM; CTR: 0.72 ± 0.11; KO: 0.32 ± 0.010; p = 0.0268) and the underlying myocardium (Fig. 2d; Mean ± SEM; CTR: 3.96 ± 1.01; KO: 0.98 ± 0.25; p = 0.0461). Hearts were also probed for the endothelial marker endomucin (EMCN), to mark the endocardium and to outline the myocardial trabeculae. No statistically significant differences were revealed in the myocardial compaction index (Fig. 2e) or in the extent of trabeculation, as assessed by fractal analysis (Fig. 2f), between mutant and control hearts at E16.5. Similarly, immunostaining against Ki67 (a marker of cells undergoing mitosis) revealed no obvious alterations in the proliferation rate of WT1 + cells between mutants and controls at E16.5 (Supplementary Fig. 3c-d). Collectively this data suggests that loss of Hif1α in epicardial cells may result in defective EMT, accounting for impaired migration of EPDCs into the sub-epicardial layer and underlying myocardium of the forming heart.

The epicardium is essential for the coronary vasculature development, with loss of Wt1 leading to absence of the coronary plexus⁵⁰. Whole-mount immunofluorescence staining for EMCN associated with quantification using AngioTool™ software⁵¹, revealed significantly impaired coronary vessel formation following epicardial deletion of Hif1α (Fig. 2g-j). Wt1^CreERT2/+;Hif1α^fl/fl mutant hearts presented with a significant decrease in the total vessel length (microns) (Fig. 2h, Mean ± SEM; CTR: 129229 ± 5509; KO: 81073 ± 10915; p = 0.0170), number of junctions (Fig. 2i, Mean ± SEM; CTR: 2505 ± 63.74; KO: 1457 ± 282.1; p = 0.0223) and end points (Fig. 2j, Mean ± SEM; CTR: 3422 ± 249.5; KO: 2429 ± 218.6; p = 0.0402). During development, epicardial EMT and the migration of EPDCs into the sub-epicardial space is known to support coronary vessel formation. Thus, we sought to investigate whether the lack of vascular complexity observed in the Wt1^CreERT2/+;Hif1α^fl/fl mutant embryos was due to impaired epicardial EMT. To address this, epicardial explants were generated from E11.5 embryos from a Rosa26^+/CreERT2;Hif1α^fl/fl mouse line, in which a tamoxifen-inducible Cre was under the control of the ubiquitous Rosa26 promoter. Immunostaining and quantification analyses (Supplementary Fig. 4a-c) confirmed that tamoxifen treatment was effective in reducing HIF-1α expression (Supplementary Fig. 4b, Mean ± SEM; CTR: 0.99 ± 0.16; KO: 0.46 ± 0.12; p = 0.0397) without affecting HIF-2α levels (Supplementary Fig. 4c; Mean ± SEM; CTR: 2505 ± 63.74; KO: 1457 ± 282.1; p = 0.0223). EPDC-like cells arising from Rosa26^+/CreERT2;Hif1α^fl/fl mutant explants presented a cobblestone-like morphology with comparably fewer phalloidin-labelled stress fibres in their cytoplasm as compared to control explants which revealed a more spindle-like mesenchymal cell morphology (Fig. 2k). In addition, cells from mutant explants were characterised by predominantly membrane localised alpha-smooth muscle actin (α-SMA) staining, demarcating their epithelial cell shape, whereas control cells displayed a more intense cytosolic labelling of filamentous actin as a mesenchymal feature (Fig. 2l). Similarly, in control cells the localisation of the zonula occludens-1 (ZO-1) protein, a structural component of cell-cell adhesion complexes, was predominantly cytoplasmic and/or nuclear, whereas in mutant cells ZO-1 staining was membranous, indicating a more epithelial-like phenotype (Fig. 2m). Consistent with our in vivo findings, epicardial explants derived from Rosa26^+/CreERT2;Hif1α^fl/fl hearts showed a decreased expression of WT1, further supporting impaired epicardial EMT in HIF1α-deficient epicardial cells (Fig. 2n-o; Mean ± SEM; CTR: 725.7 ± 86.93; KO: 490.9 ± 37.65; p = 0.0479).

To test whether WT1 might be a direct target of HIF-1α we interrogated a publicly available dataset of chromatin immunoprecipitation coupled with next generation sequencing (ChIP-seq), generated using an antibody targeting the β-subunit of the heterodimeric α-β HIF complex (i.e. HIF-1β) in the human clear cell renal cell carcinoma line 786-O⁵². The 786-O cell line expresses high levels of WT1 and is a surrogate model for active EMT and cell migration⁵³. ChIP-seq analysis revealed direct binding of HIF within intron 3 of the Wt1 gene (Supplementary Fig. 5), confirming previous findings⁴².

In summary, epicardial-specific deletion of Hif1a significantly reduced the number of WT1 + cells, both in the epicardium and underlying myocardium, affecting EPDC mobilisation and coronary vessel development at E16.5. Epicardial explants, following global loss of Hif1a, revealed a persistent epithelial phenotype which, in combination with the in vivo findings, suggest that HIF regulates epicardial EMT.

Stabilisation of HIF signalling enhances epicardial EMT.

We next sought to investigate whether stabilising HIF signalling under normoxic conditions is sufficient to enhance epicardial EMT. To this end, we established epicardial explants from Rosa26^+/CreERT2;Phd2^fl/fl mice, in which cre-mediated recombination induced loss of Phd2, thus stabilising HIF signalling. Upon tamoxifen treatment, HIF-1α and HIF-2α expression were significantly induced in Rosa26^+/CreERT2;Phd2^fl/fl derived explants (KO) as compared to controls (CTR) (Supplementary Fig. 6a-c; HIF-1a; Mean ± SEM; CTR: 0.26 ± 0.13; KO: 0.86 ± 0.09; p = 0.0061; HIF-2a; Mean ± SEM; CTR: 0.37 ± 0.01; KO: 1.12 ± 0.15; p = 0.0027). The extent and distribution of stress fibres was then visualised with α-SMA staining, revealing an enhanced mesenchymal phenotype in the mutant-derived explants (Fig. 3a). Consistent with this observation, WT1 nuclear fluorescence was significantly increased in mutant (KO) explants as compared to controls (CTR; Fig. 3b-c; Mean ± SEM; CTR: 158.7 ± 48.34; KO: 330.5 ± 27.26; p = 0.0364). Furthermore, the expression level of the EMT transcription factor snail family transcriptional repressor 2 (Snai2) gene was increased in KO as compared to CTR explants (Fig. 3d; Mean ± SEM, CTR: 1 ± 0.025; KO: 1.166 ± 0.036; p = 0.0464).

To complement the genetic studies, we undertook a more therapeutically relevant approach by using pharmacological PHDs inhibitors to stabilise HIF signalling. Epicardial explants were generated from C57BL/6 wildtype embryos harvested at E11.5 and treated with Roxadustat (FG 4592), Molidustat (BAY 85-3934) or DMSO control. Both drugs are clinically approved, orally administered small molecules for the treatment of anaemia in patients with dialysis-dependent chronic kidney disease (CKD)⁴⁷, non-dialysis-dependent CKD^44,46 and in patients with myelodysplastic syndromes⁴⁵. Immunostaining for HIF-1α and HIF-2α was performed (Supplementary Fig. 6d,e) and stabilisation of HIF signalling upon PHDs inhibitor treatment was confirmed (Supplementary Fig. 6f,g). To assess the effect of pharmacological PHD-inhibition on epicardial EMT, the proportion of mesenchymal cells was assessed by morphological evaluation. Staining with phalloidin (Fig. 3e) and α-SMA (Fig. 3f) showed an enhanced spindle-shape morphology with increased stress fibres, suggesting induction of EMT in inhibitor treated-explants. Additionally, immunostaining revealed a significant increase in WT1 expression in treated explants for both drugs, as compared to control (Fig. 3g,h; Mean ± SEM; Control: 1039 ± 134.9; Roxadustat: 2318 ± 438.4; Molidustat: 2500 ± 503.7; p = 0.03). Likewise, PHD inhibitors enhanced the motility of cultured mouse embryonic ventricular epicardial cells (MEC.1 cell line)⁵⁴, compared to control-treated cells (Fig. 3i; Mean ± SEM; Control: 436869 ± 9359; Roxadustat: 378457 ± 14125; p = 0.0266; Molidustat: 372482 ± 20880; p = 0.0477), as assessed by scratch assays, a standardised assay to monitor cell migration following EMT induction⁵⁵.

These data suggest that genetic perturbation and/or pharmacological inhibition of PHDs stabilise HIF signalling in the epicardium ex vivo, leading to an increase in WT1 expression and subsequent stimulation of epicardial EMT.

Decline of HIF signalling is associated with epicardial quiescence in the neonatal heart.

During the latter stages of embryonic heart development through to birth there is a gradual transition from hypoxia to normoxia, which is ongoing across the neonatal regenerative window as defined by post-natal days 1 through to 7 in mice. We predicted that HIF-related molecular pathways are differentially regulated in P1 versus P7 hearts and performed single cell RNA-seq analysis using the 10x Genomics Chromium platform and next generation sequencing to provide unbiased insight into hypoxia-related molecular pathways over this period. Cell clusters were visualised through uniform manifold approximation and projection (UMAP) (Fig. 4a). An epicardial cell cluster (Epi) was identified based on previously described specific epicardial gene expression signatures ⁵⁶ (Fig. 4b). As expected, a cardiomyocyte cell cluster (CM) was underrepresented due to both low survival and size incompatibility with the FACS cell sorting ⁵⁷. Gene Ontology (GO) analysis for biological processes showed an enrichment of hypoxia-related pathways in P1 compared to P7 hearts (Supplementary Fig. 7a), and in the epicardial cell population (Fig. 4c) specifically. In contrast, hypoxia-related pathways were enriched in P7 hearts as compared to P1 hearts in the fibroblast cluster (Supplementary Fig. 7b-d), suggesting a role for hypoxia in the fibrotic response leading to scarring. Accordingly, expression of well-known HIF-induced genes such as Vegfa and pyruvate dehydrogenase kinase 3 (Pdk3) was increased in P1 derived epicardial cells (Fig. 4d). Conversely, the expression of Phd2, encoding for the main suppressor of HIF signalling, was enriched in P7 cells (Fig. 4e). These findings were confirmed by time-course analysis of mRNA levels using ventricle lysates and immunostaining which revealed a marked upregulation of Phd2 cardiac levels from P7 through to adulthood (Fig. 4f, Supplementary Fig. 8a-b). Notably, the increase in Phd2 levels coincided with a clear reduction in WT1 expression in P7 versus P1 hearts, as determined by immunostaining (Fig. 4g).

To determine a causative role in the regulation of epicardial activity, we extended HIF signalling beyond P7 via the use of a genetic gain-of-function approach. Rosa26^+/CreERT2;Phd2^fl/fl neonates were injected with a single-dose of tamoxifen at P2 to induce cre-mediated loss of Phd2 and stabilisation of HIF (Supplementary Fig. 9a). This approach led to a reduction in Phd2 expression levels by over 90% in Rosa26^+/CreERT2;Phd2^fl/fl (KO) hearts compared to littermate controls at P7 (CTR; Phd2^fl/fl) (Mean Fold change ± SEM; CTR: 2.99 ± 0.24, n = 7; KO: 0.25 ± 0.063 n = 5; p < 0.0001; Supplementary Fig. 9b). Additionally, expression of the HIF target, Glut1, was significantly upregulated in KO hearts at P7, suggesting the activation of HIF-mediated transcription (Mean Fold change ± SEM; CTR: 0.78 ± 0.13, n = 7; KO: 1.61 ± 0.20 n = 5; p = 0.005; Supplementary Fig. 9c). Consistent with our hypothesis, epicardial WT1 expression was significantly induced in KO hearts, at both RNA (Mean Fold change ± SEM; CTR 0.84 ± 0.096, n = 7; KO: 1.37 ± 0.086 n = 5; p = 0.003; Supplementary Fig. 9d) and protein level (Mean number of Wt1 + cells ± SEM; CTR: 9.75 ± 1.307, n = 4; KO: 16.7 ± 0.3 n = 2; p = 0.0243; Supplementary Fig. 9e,f). Together, these findings indicate that increased Phd2 levels and accompanying silencing of HIF signalling lead to gradual reduction in WT1 expression and epicardial quiescence after birth and across the neonatal regenerative window.

Activation of HIF signalling improves the response to injury in non-regenerative P7 hearts

Given our findings that HIF signalling modulates epicardial activation, we next sought to determine whether prolonged maintenance of HIF signalling would have a beneficial effect on the response to MI in the postnatal heart, beyond the first 7-days after birth. At first, stabilisation of HIF signalling was induced globally in Rosa26^+/CreERT2;Phd2^fl/fl pups by tamoxifen injection at P2, and MI was induced by permanent ligation of the proximal left anterior descending (LAD) coronary artery in P7 mice (Supplementary Fig. 9a), coincident with loss of regenerative capacity. Immunostaining for WT1 on sections from hearts collected at 4 days post-injury (dpi), revealed a widespread increase in WT1 + cells in the epicardial layer of KO, compared to littermate control hearts (Supplementary Fig. 10a). Notably, this increase was more pronounced in the remote epicardium, than in the infarcted area, suggesting organ-wide epicardial activation analogous to that observed in the regenerating zebrafish heart⁵⁸ (Mean number of WT1 + epicardial cells ± SEM; CTR infarct zone: 51 ± 16.43; KO infarct zone: 34.14 ± 7.567, n = 4, p = 0.327; CTR remote zone: 2.6 ± 1.208; KO remote zone: 15.29 ± 2.90, n = 4, p = 0.0057; Supplementary Fig. 10b,c). Likewise, increased WT1 expression was detected in the myocardium, both in the infarct and remote zone of KO hearts (Mean number of WT1 + cells in myocardial layer ± SEM; CTR infarct zone: 2.22 ± 1.44; KO infarct zone: 52.09 ± 12.61, n = 4, p = 0.0023; CTR remote zone: 2.625 ± 2.485; KO remote zone: 59.2 ± 16.51, n = 4, p = 0.008; Supplementary Fig. 10d,e). WT1 expression in coronary endothelium has previously been described both in the developing and adult heart following MI⁵⁹. We observed extensive co-localization of WT1 and EMCN, supporting an endothelial identity for the observed incidence of myocardial WT1 + cells (Supplementary Fig. 10f). Despite this increase in WT1 + ECs, the neovascularisation response was not substantially affected by stabilisation of HIF signalling, at 4dpi (Supplementary Fig. 10g). Thus, the exact role of the hypoxia-induced WT1 expression in the coronary endothelium remains elusive. In addition, we analysed macrophage infiltration and overall cell proliferation as potential contributors to an improved response in the P7 heart, but neither appeared to be altered in the absence of PHD2 at 4dpi (Supplementary Fig. 10h). Next, ventricular remodelling and scarring were assessed at 21dpi. Histological analysis of control hearts revealed the presence of a transmural scar, extending to the epicardial surface, and evidence of pathological remodelling via left ventricular dilation and a thinned right ventricle (Supplementary Fig. 10i, left panel). In the KO heart a transmural scar was also observed, however the myocardium in the area of risk was comparatively thicker with minimal remodelling of the left ventricle (Supplementary Fig. 10i, right panel); suggesting loss of PHD2 may improve cardiac muscle survival post-MI.

To evaluate the contribution of stabilising HIF signalling specifically in the epicardium, we induced cre-mediated recombination in WT1 expressing cells by injecting Wt1^CreERT2;Phd2^fl/fl pups at P2. MI was then induced by permanent ligation of the proximal LAD coronary artery at P7 and hearts collected 21dpi (Fig. 5a). This approach led to diminished PHD2- and enhanced WT1-immunoreactivity in the epicardial layer of Wt1^CreERT2;Phd2^fl/fl hearts compared to littermate controls at P7 (Phd2^fl/fl) (Supplementary Fig. 11a,b). Immunostaining with the endothelial marker CD31 and smooth muscle marker α-SMA was performed, but no significant difference in the number of vessels was observed between Phd2^fl/fl (CTR) and Wt1^CreERT2;Phd2^fl/fl (KO) hearts at 21dpi (Fig. 5b-c), suggesting that activation of HIF signalling in WT1 expressing cells does not affect the neovascularisation response. To investigate whether epicardial activation of HIF signalling led to improved heart function following MI, we performed cine-MRI analysis at 21 dpi. We observed a marked increase in the ejection fraction (EF) and decrease in the end-diastolic volume (EDV) in KO animals as compared to controls (Fig. 5d-f; EF: CTR 39.50 ± 9.5, KO 63.40 ± 4.63, p = 0.049. EDV; CTR 47.50 ± 3.00, KO 26.40 ± 3.98, p = 0.028). Furthermore, histological analysis by Masson’s trichrome staining revealed a significant reduction of the fibrotic scar in KO animals as compared to controls (Fig. 5g,h; CTR 15.44 ± 1.088, KO 10.34 ± 1.374, p = 0.027). Taken together these findings suggest that activated hypoxia signalling in the epicardium contributes to heart regeneration by reducing scar size and improving heart function.

To further test our hypothesis that prolonged maintenance of HIF signalling has a beneficial effect on the epicardium and response to MI in the postnatal heart, we explored a complementary pharmacological approach recapitulating our gain of HIF-function studies using epicardial explants (Fig. 3). We induced MI by LAD ligation in P7 mice and administered Roxadustat, Molidustat or DMSO control (vehicle) by intraperitoneal injection, immediately after surgery and one week later, to stabilise HIF signalling in the postnatal heart (Supplementary Fig. 12a). Treatment with either PHD inhibitor proved effective in stabilising HIF-1α and HIF-2α, as observed by immunostaining (Supplementary Fig. 12b) and quantification of nuclear intensity (Supplementary Fig. 12c,d) in hearts collected at 9dpi. Moreover, this pharmacological approach revealed a significant increase in epicardial WT1 expression in both the infarct and remote zones, in drug-treated hearts compared to controls (Fig. 6a-c, % of WT1⁺ epicardial cells ± SEM; control infarct zone: 0.34 ± 0.08, n = 4; Roxadustat-treated infarct zone: 0.64 ± 0.13, n = 4; Molidustat-treated infarct zone: 0.50 ± 0.06, n = 4; control remote zone: 0.075 ± 0.02, n = 4; Roxadustat-treated remote zone: 0.16 ± 0.02, n = 4; Molidustat-treated remote zone: 0.25 ± 0.05, n = 4; p = 0.03; Fig. 6d,e). Similarly, an increase in WT1 expression in the myocardium was observed in drug-treated groups and was evidently more pronounced proximal to the infarct region (% of WT1⁺ myocardial cells ± SEM; control infarct zone: 5.02 ± 1.34, n = 4; Roxadustat-treated infarct zone: 14.32 ± 3.09, n = 4; Molidustat-treated infarct zone: 7.09 ± 1.86, n = 4; p = 0.03; control remote zone: 13.88 ± 4.96, n = 4; Roxadustat-treated remote zone: 29.93 ± 6.55, n = 4; Molidustat-treated remote zone: 18.61 ± 6.18, n = 4, Fig. 6f,g). In keeping with our previous findings, using genetic approaches to target Phd2 (Fig. 5; also, Supplementary Figs. 9 and 10), we observed an extensive co-localization of WT1 and EMCN within the myocardial compartment at 9dpi (Supplementary Fig. 12e), supporting an endothelial identity for these cells. In order to determine whether the observed increase in WT1 + cells following treatment with PHD inhibitors promoted neovascularisation, immunostaining against the endothelial marker CD31 and α-SMA was performed on sections from treated and control hearts collected at 9dpi (Fig. 6h-j). No significant difference in the number of vessels was observed between the different experimental groups (Fig. 6k), suggesting no vascular response mediated by WT1 + ECs during the initial stages post-MI. Furthermore, the proliferation of WT1 + cells and cardiomyocytes was assessed by immunostaining for proliferating cell nuclear antigen (PCNA, Supplementary Fig. 12f) and by EdU incorporation (Supplementary Fig. 12g,h), respectively. No significant differences were observed between control and PHD inhibitor-treated animals, suggesting the expansion of WT1 + cells in the epicardium or coronary endothelium was not due to proliferation per se.

In order to analyse the long-term effects of therapeutic HIF-stabilisation after MI, we collected hearts at 21dpi. Interestingly, in contrast to the earlier stage of 9dpi, co-immunostaining and quantification of CD31 and α-SMA revealed an increase in the number of vessels in both treatment groups at this more chronic stage following injury, with a more pronounced effect in the Roxadustat-treated mice (Fig. 7a-d). To investigate whether activation of HIF signalling led to improved heart function following MI, we performed cine-MRI analysis at 21dpi. We observed a marked reduction in left ventricular systolic function following MI in the vehicle-treated group as compared to uninjured hearts and, more importantly, a significant improvement of systolic function in the PHD inhibitor treated-groups (Fig. 7e-h). MRI revealed that cardiac output (CO) and stroke volume (SV) were significantly improved in Roxadustat- and Molidustat-treated animals, with a more prominent effect in the latter group, comparable to uninjured controls (Fig. 7e,f; CO: control 10.53 ± 0.32, Roxadustat 10.14 ± 0.46, Molidustat 8.73 ± 0.38, p = 0.0033. SV; control 24.37 ± 1.53, Roxadustat 22.50 ± 0.96, Molidustat 19.04 ± 0.79, p = 0.0050). Moreover, the end-diastolic volume (EDV) in the drug-treated groups was significantly improved (Fig. 7h; EDV: control 33.38 ± 1.885, Roxadustat 32.26 ± 1.57, Molidustat 25.50 ± 0.82, p = 0.0011). Histological analysis by Masson’s trichrome staining revealed the presence of a fibrotic scar in drug-treated animals, similar to controls (Fig. 7i,j). However, the myocardium in the area of risk was comparatively thicker with minimal dilatation or thinning of the left ventricle. The thickened myocardium was not due to an effect on cardiomyocyte hypertrophy, as determined by (WGA) staining (Fig. 7k,l); indicating preservation of heart muscle and more generally that improved remodelling and function following drug treatment were possible against a background of ensuing fibrosis. Taken together these findings suggest that treatment with the PHD inhibitors Roxadustat and Molidustat is effective in stabilising HIF signalling beyond P7, leading to an increase in epicardial activity and the number of WT1 + cells, augmented neovascularisation, improved outcome and preserved heart function following MI.

In this study, we characterised the levels of hypoxia in the heart during mid-to-late gestation when the epicardium becomes fully formed. At E12.5, hypoxia was restricted to areas of compact myocardium, most significantly in the AV groove, situated close to where the coronary plexus first forms⁶⁰. By E14.5, regions of hypoxia were localised to the IVS, which is perfused later than the muscle of the free wall⁶¹. Finally, by E16.5 and E18.5, when the heart is supplied by the coronaries, hypoxia was restricted to the epicardium. Further spatial expression analysis of HIF-1 isoforms revealed a predominance of the HIF-1α isoform in WT1⁺ cells, whereas HIF-2α appeared to be more localized in the myocardium. Hypoxia is a known promoter of EMT⁶² and in the epicardium this coincided with the presence of WT1 + EPDCs. We confirmed a functional role for HIF signalling in the developing epicardium by stabilisation of HIF signalling in epicardial explant cultures, under physiological oxygen levels. Genetic ablation of Phd2, or chemical inhibition of PHD enzymes with either Roxadustat (FG-4592) or Molidustat (BAY 85-3934), resulted in enhanced EMT and WT1 expression in mutant or treated epicardial explants (Figs. 2 and 3, Supplementary Fig. 5). Moreover, epicardial- specific deletion of Hif1a significantly reduced the number of WT1 expressing cells in both epicardial and myocardial compartments, along with impaired coronary vessel development in E16.5 embryos. The phenotype of epicardial-Hif1a mutants does not recapitulate the more severe phenotype attributable to Wt1 conditional KOs² but this is unsurprising, given WT1 activity within the epicardium is likely mediated through multiple signalling pathways^4,9. Collectively, our findings suggests that HIF-1α induces epicardial EMT and plays an essential role in coronary vessel formation. That said, whilst cre-mediated deletion of HIF-1α, induced at E9.5 and E10.5, predominantly targeted epicardial cells in our hands, expression of WT1 in the coronary endothelium has been reported as early as E11.5⁶³. Thus, aberrant HIF signalling in WT1 + ECs may also contribute to the observed defects on the expansion of the coronary vasculature.

To gain insight into the molecular pathways regulating the post-natal epicardium, we performed single cell RNA sequencing analysis, comparing P1 and P7 stages and focused on the specific gene expression signature of the epicardial cell cluster. GO terms analysis showed an enrichment of hypoxia-related pathways in P1 hearts, and the expression of well-known HIF target genes showed a similar regulation. Importantly, Phd2 expression was enriched in P7 hearts, potentially contributing to epicardial quiescence during the first week of life, concurrent with a decrease in HIF signalling. Complete regeneration, following MI, requires both the replacement of lost cardiomyocytes and the formation of new blood vessels. Given the role of the epicardium during development, in promoting coronary vessel formation and growth of the forming myocardium, the observed epicardial quiescence after birth likely contributes to the loss of regenerative capacity. Following targeting of Phd2, persistent WT1 expression was detected in the epicardium at P7, suggesting that maintaining HIF signalling after birth is sufficient to extend activation of the epicardium beyond the first week of life. Genetic and pharmacological (Roxadustat or Molidustat) stabilisation of HIF signalling beyond P7 proved to be effective in extending the regenerative window post-MI, resulting in reduced pathological remodelling (i.e., reduced scarring and muscle loss), preserved cardiac function and formation of new vessels against a background of fibrotic repair. Interestingly, genetic stabilisation of HIF signalling specifically in WT1-expressing cells significantly reduced scar size (Fig. 5) but did not impact on neovascularization, in contrast to the pharmacological approach that improved both processes. The broader impact of the PHD2 inhibitors, most likely reflects the impact of HIF signalling in cardiovascular cell types other than epicardial, as indicated by our scRNA-seq studies (Fig. 4). Following injury, necrotic cardiomyocytes can trigger the activation of fibroblast and promote fibrosis⁶⁴. The epicardium is an important source of mitogens in the embryonic heart¹ and provides paracrine signals targeting cardiomyocytes in both zebrafish⁶⁵ and mice¹⁶ during regeneration. In our study, we did not observe any change in either cardiomyocyte proliferation or hypertrophy, suggesting an epicardium-mediated role in cardiomyocyte protection via paracrine signalling. Therefore, future studies should aim to dissect the epicardium-myocardium signalling axis at a mechanistic level.

The observation that scarring persisted, albeit reduced, following either genetic perturbation of Phd2 or chemical inhibition of PHDs, and yet there was significant functional improvement is clinically significant, given that fibrotic repair to prevent rupture after MI is considered a barrier to effective tissue regeneration^66–68. The two pharmacological inhibitors utilised in this study, Roxadustat and Molidustat, are orally administered small molecules used for the treatment of several disease indications^44–47 thus, there is genuine potential for repurposing these drugs to treat ischemic heart disease, albeit with a likely requirement for targeted delivery and further mechanistic insight into the observed functional improvement of treated infarcted hearts against a background of fibrosis.

In summary, we show that the epicardium is hypoxic at later stages of development and HIF-1α is involved in epicardial activation and EMT, required to support coronary vessel development. The epicardium becomes quiescent after birth with concomitant loss of HIF-signalling. Genetic or pharmacological stabilisation of HIF after birth is effective in maintaining an activated epicardium, leading to a significant improvement in cardiac remodelling and function after injury, thus representing a potential novel therapeutic target to improve cardiac regeneration following MI.

Acknowledgements

We are very grateful to Prof. Peter J Ratcliffe for technical advice and highly informative scientific discussions throughout the project. We would like to thank the Biomedical Services Unit for animal husbandry and the Micron Oxford Advanced Bioimaging Unit (supported by Wellcome Strategic Awards 091911/B/10/Z and 107457/Z/15/Z) for access to and training in the use of confocal microscopy. This work was funded by the British Heart Foundation (BHF chair award to PRR: CH/11/1/28798; BHF programme grant to PRR: RG/18/33532; BHF Intermediate Basic Science Research Fellowship to JMV: FS/19/31/34158) and supported by the BHF Oxbridge Regenerative Medicine Centre (RM/17/2/33380).

Author Contributions

PRR conceived the original project and sourced funding. PRR and JMV oversaw the project direction and data analyses. EG, ELP, JS, CRS and XS carried out experiments and analysed the data. CDV and MGR preformed surgeries. DP carried out the scRNA-seq experiment and bioinformatic analyses. RPC supervised the scRNA-Seq experiment and computational analyses. RS and DRM, assisted with the analysis of HIF-1β ChIP-seq datasets. TB and CWP provided Rosa^26+/CreERT2;Hif1a^fl/fl and Rosa^26+/CreERT2;Phd2^fl/flmouse strains and provided advice and technical assistance. CAC performed blinded MRI scanning and analyses. EG wrote the main draft of the manuscript. PRR and JMV contributed to editing of the final manuscript.

Competing interests

The authors declare no competing interests.

Simões, F. C. & Riley, P. R. The ontogeny, activation and function of the epicardium during heart development and regeneration. Dev. 145, (2018).
Martínez-Estrada, O. M. et al. Wt1 is required for cardiovascular progenitor cell formation through transcriptional control of Snail and E-cadherin. Nat. Genet. 42, 89–93 (2010).
Li, P. et al. IGF signaling directs ventricular cardiomyocyte proliferation during embryonic heart development. Development 138, 1795–1805 (2011).
von Gise, A. et al. WT1 regulates epicardial epithelial to mesenchymal transition through β-catenin and retinoic acid signaling pathways. Dev Biol 356, 421–431 (2011).
Gittenberger-de Groot, A. C., Vrancken Peeters, M. P. F. M., Mentink, M. M. T., Gourdie, R. G. & Poelmann, R. E. Epicardium-derived cells contribute a novel population to the myocardial wall and the atrioventricular cushions. Circ. Res. 82, 1043–1052 (1998).
Mikawa, T. & Gourdie, R. G. Pericardial mesoderm generates a population of coronary smooth muscle cells migrating into the heart along with ingrowth of the epicardial organ. Dev. Biol. 174, 221–232 (1996).
Katz, T. C. et al. Distinct Compartments of the Proepicardial Organ Give Rise to Coronary Vascular Endothelial Cells. Dev. Cell 22, 639–650 (2012).
Smart, N. et al. De novo cardiomyocytes from within the activated adult heart after injury. Nature 474, 640–644 (2011).
Vieira, J. M. et al. BRG1-SWI/SNF-dependent regulation of the Wt1 transcriptional landscape mediates epicardial activity during heart development and disease. Nat. Commun. 8, (2017).
Zhou, B. & Pu, W. T. Epicardial epithelial-to-mesenchymal transition in injured heart. J. Cell. Mol. Med. 15, 2781–2783 (2011).
Price, E. L., Vieira, J. M. & Riley, P. R. Model organisms at the heart of regeneration. DMM Dis. Model. Mech. 12, (2019).
Porrello, E. R. & Olson, E. N. A neonatal blueprint for regeneration. Stem Cell Res. 13, 556–570 (2014).
Porrello, E. R. et al. Transient Regenerative Potential of the Neonatal Mouse Heart. Science (80-.). 331, 1078–1080 (2011).
Haubner, B. J. et al. Functional Recovery of a Human Neonatal Heart after Severe Myocardial Infarction. Circ. Res. 118, 216–221 (2016).
Wang, J., Cao, J., Dickson, A. L. & Poss, K. D. Epicardial regeneration is guided by cardiac outflow tract and Hedgehog signalling. Nature 522, (2015).
Zhou, B. et al. Adult mouse epicardium modulates myocardial injury by secreting paracrine factors. J Clin Invest 121, 1894–1904 (2011).
Smart, N. et al. Thymosin β4 facilitates epicardial neovascularization of the injured adult heart. Ann. N. Y. Acad. Sci. 1194, 97–104 (2010).
van Wijk, B., Gunst, Q. D., Moorman, A. F. M. & van den Hoff, M. J. B. Cardiac Regeneration from Activated Epicardium. PLoS One 7, (2012).
Puente, B. N. et al. The oxygen-rich postnatal environment induces cardiomyocyte cell-cycle arrest through DNA damage response. Cell 157, 565–579 (2014).
Kimura, W. et al. Hypoxia fate mapping identifies cycling cardiomyocytes in the adult heart. Nature 523, (2015).
Kocabas, F. et al. The hypoxic epicardial and subepicardial microenvironment. J. Cardiovasc. Transl. Res. 5, (2012).
Nakada, Y. et al. Hypoxia induces heart regeneration in adult mice. Nature 541, (2017).
Kaelin, W. G. & Ratcliffe, P. J. Oxygen Sensing by Metazoans: The Central Role of the HIF Hydroxylase Pathway. Mol. Cell 30, 393–402 (2008).
Semenza, G. L. & Wang, G. L. A nuclear factor induced by hypoxia via de novo protein synthesis binds to the human erythropoietin gene enhancer at a site required for transcriptional activation. Mol. Cell. Biol. 12, 5447–5454 (1992).
Hu, C.-J., Wang, L.-Y., Chodosh, L. A., Keith, B. & Simon, M. C. Differential Roles of Hypoxia-Inducible Factor 1α (HIF-1α) and HIF-2α in Hypoxic Gene Regulation. Molecular and Cellular Biology vol. 23 9361–9374 (2003).
Schödel, J. et al. High-resolution genome-wide mapping of HIF-binding sites by ChIP-seq. Blood 117, 207–217 (2011).
Maxwell, P. H. et al. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 399, 271–275 (1999).
Forsythe, J. A. et al. Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol. Cell. Biol. 16, 4604–4613 (1996).
Ebert, B. L., Firth, J. D. & Ratcliffe, P. J. Hypoxia and mitochondrial inhibitors regulate expression of glucose transporter-1 via distinct cis-acting sequences. J. Biol. Chem. 270, 29083–29089 (1995).
Dunwoodie, S. L. The Role of Hypoxia in Development of the Mammalian Embryo. Dev. Cell 17, 755–773 (2009).
Jain, S., Maltepe, E., Lu, M. M., Simon, C. & Bradfield, C. A. Expression of ARNT, ARNT2, HIF1α, HIF2α and Ah receptor mRNAs in the developing mouse. Mech. Dev. 73, 117–123 (1998).
Compernolle, V. et al. Cardia bifida, defective heart development and abnormal neural crest migration in embryos lacking hypoxia-inducible factor-1α. Cardiovasc. Res. 60, 569–579 (2003).
Krishnan, J. et al. Essential role of developmentally activated hypoxia-inducible factor la for cardiac morphogenesis and function. Circ. Res. 103, 1139–1146 (2008).
Ryan, H. E., Lo, J. & Johnson, R. S. HIF-1 alpha is required for solid tumor formation and embryonic vascularization. EMBO J 17, 3005–3015 (1998).
Peng, J., Zhang, L., Drysdale, L. & Fong, G.-H. The transcription factor EPAS-1 hypoxia-inducible factor 2α plays an important role in vascular remodeling.pdf. Proc Natl Acad Sci USA 97, 8386–91 (2000).
Tian, H., Hammer, R. E., Matsumoto, A. M., Russell, D. W. & McKnight, S. L. The hypoxia-responsive transcription factor EPAS1 is essential for catecholamine homeostasis and protection against heart failure during embryonic development. Genes Dev. 12, 3320–3324 (1998).
Adelman, D. M., Gertsenstein, M., Nagy, A., Simon, M. C. & Maltepe, E. Placental cell fates are regulated in vivo by HIF-mediated hypoxia responses. Genes Dev. 14, 3191–3203 (2000).
Takeda, K. et al. Placental but Not Heart Defects Are Associated with Elevated Hypoxia-Inducible Factor α Levels in Mice Lacking Prolyl Hydroxylase Domain Protein 2. Mol. Cell. Biol. 26, 8336–8346 (2006).
Fan, Q. et al. Depletion of Endothelial Prolyl Hydroxylase Domain Protein 2 and 3 Promotes Cardiomyocyte Proliferation and Prevents Ventricular Failure Induced by Myocardial Infarction. Circulation vol. 140 (2019).
Tao, J., Doughman, Y., Yang, K., Ramirez-Bergeron, D. & Watanabe, M. Epicardial HIF signaling regulates vascular precursor cell invasion into the myocardium. Dev Biol 376, 136–149 (2013).
Tao, J., Barnett, J., Watanabe, M. & Ramírez-Bergeron, D. Hypoxia Supports Epicardial Cell Differentiation in Vascular Smooth Muscle Cells through the Activation of the TGFβ Pathway. J. Cardiovasc. Dev. Dis. 5, 19 (2018).
Wagner, K.-D. et al. Oxygen-regulated expression of the Wilms’ tumor suppressor Wt1 involves hypoxia-inducible factor-1 (HIF-1). FASEB J 17, 1364–1366 (2003).
Krueger, K., Catanese, L., Sciesielski, L. K., Kirschner, K. M. & Scholz, H. Deletion of an intronic HIF-2α binding site suppresses hypoxia-induced WT1 expression. Biochimica et Biophysica Acta - Gene Regulatory Mechanisms vol. 1862 71–83 (2019).
Macdougall, I. C., Akizawa, T., Berns, J. S., Bernhardt, T. & Krueger, T. Effects of molidustat in the treatment of anemia in CKD. Clin. J. Am. Soc. Nephrol. 14, (2019).
Henry, D. H. et al. Roxadustat (FG4592; ASP1517; AZD9941) in the Treatment of Anemia in Patients with Lower Risk Myelodysplastic Syndrome (LR-MDS) and Low Red Blood Cell (RBC) Transfusion Burden (LTB). Blood 134, (2019).
Chen, N. et al. Roxadustat for Anemia in Patients with Kidney Disease Not Receiving Dialysis. N. Engl. J. Med. 381, (2019).
Chen, N. et al. Roxadustat Treatment for Anemia in Patients Undergoing Long-Term Dialysis. N. Engl. J. Med. 381, (2019).
Raleigh, J. A., Chou, S. C., Arteel, G. E. & Horsman, M. R. Comparisons among pimonidazole binding, oxygen electrode measurements, and radiation response in C3H mouse tumors. Radiat. Res. 151, (1999).
Zhou, B. et al. Epicardial progenitors contribute to the cardiomyocyte lineage in the developing heart. Nature 454, (2008).
Moore, A. W., McInnes, L., Kreidberg, J., Hastie, N. D. & Schedl, A. YAC complementation shows a requirement for Wt1 in the development of epicardium, adrenal gland and throughout nephrogenesis. Development 126, (1999).
Zudaire, E., Gambardella, L., Kurcz, C. & Vermeren, S. A computational tool for quantitative analysis of vascular networks. PLoS One 6, (2011).
Smythies, J. A. et al. Inherent DNA -binding specificities of the HIF ‐1α and HIF ‐2α transcription factors in chromatin. EMBO Rep. 20, (2019).
Somsuan, K. et al. ARID1A knockdown triggers epithelial-mesenchymal transition and carcinogenesis features of renal cells: Role in renal cell carcinoma. FASEB J. 33, (2019).
Li, P. et al. IGF signaling directs ventricular cardiomyocyte proliferation during embryonic heart development. Development 138, 1795–805 (2011).
Liang, C.-C., Park, A. Y. & Guan, J.-L. In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro. Nat. Protoc. 2, 329–33 (2007).
Bochmann, L. et al. Revealing new mouse epicardial cell markers through transcriptomics. PLoS One 5, (2010).
Cui, M. et al. Dynamic Transcriptional Responses to Injury of Regenerative and Non-regenerative Cardiomyocytes Revealed by Single-Nucleus RNA Sequencing. Dev. Cell 53, (2020).
Lepilina, A. et al. A Dynamic Epicardial Injury Response Supports Progenitor Cell Activity during Zebrafish Heart Regeneration. Cell 127, (2006).
Duim, S. N., Kurakula, K., Goumans, M. J. & Kruithof, B. P. T. Cardiac endothelial cells express Wilms’ tumor-1. Wt1 expression in the developing, adult and infarcted heart. J. Mol. Cell. Cardiol. 81, (2015).
Chen, H. I. et al. The sinus venosus contributes to coronary vasculature through VEGFC-stimulated angiogenesis. Dev. 141, (2014).
Tian, X. et al. Subepicardial endothelial cells invade the embryonic ventricle wall to form coronary arteries. Cell Res. 23, (2013).
Scully, D. et al. Hypoxia promotes production of neural crest cells in the embryonic head. Dev. 143, (2016).
Lupu, I. E., Redpath, A. N. & Smart, N. Spatiotemporal Analysis Reveals Overlap of Key Proepicardial Markers in the Developing Murine Heart. Stem Cell Reports 14, (2020).
Prabhu, S. D. & Frangogiannis, N. G. The biological basis for cardiac repair after myocardial infarction. Circulation Research vol. 119 (2016).
Wills, A. A., Holdway, J. E., Major, R. J. & Poss, K. D. Regulated addition of new myocardial and epicardial cells fosters homeostatic cardiac growth and maintenance in adult zebrafish. Development 135, (2008).
Liang, J. et al. Concise Review: Reduction of Adverse Cardiac Scarring Facilitates Pluripotent Stem Cell-Based Therapy for Myocardial Infarction. Stem Cells vol. 37 (2019).
Simões, F. C. et al. Macrophages directly contribute collagen to scar formation during zebrafish heart regeneration and mouse heart repair. Nat. Commun. 11, (2020).
Koth, J. et al. Runx1 promotes scar deposition and inhibits myocardial proliferation and survival during zebrafish heart regeneration. Dev. 147, (2020).

Mouse Lines

All animal experiments were carried out in accordance with the UK Home Office project license (PPL) 30/2987, PPL30/3155 and PDDE89C84 compliant with the UK animals (Scientific Procedures) Act 1986 and approved by the local Biological Services Ethical Review Process. The following mouse lines were used: Rosa^26+/CreERT2;Hif1a^fl/fl;, Rosa^26+/CreERT2;Phd2^fl/fl. The Wt1^CreERT2/+;Hif1a^fl/fland Wt1^CreERT2/+;Phd2^fl/lfmouse line was generated by crossing Wt1^CreERT2/+⁴⁹ with Hif1a^fl/fl animals¹ for two generations. Genetically modified mouse lines used were kept in a pure C57BL/6 background. Both males and females were used in the study. For timed-mating experiments, 8-12-week-old mice were set up overnight and females checked for vaginal plugs the following morning; the date of a vaginal plug was set as embryonic day (E) 0.5. For tamoxifen-dependent gene activation, 2 doses of 40 mg/kg of body weight of tamoxifen (Sigma) were administered to pregnant dams by oral gavage, at embryonic stages E9.5 and E10.5. For neonate studies, pups were injected intraperitoneally (i.p.) with a single 10 μl dose of 20 mg/ml tamoxifen, at postnatal day (P)2, using a 25-gauge needle³.

5-ethynyl-2'-deoxyuridine (EdU, Thermofisher) was intraperitoneally injected at a dose of 50mg/Kg at the day of surgery (P7) and every other day. Hearts were collected at 9 dpi and the assay was carried out according to manufacturer’s protocol.

Hypoxyprobe

For hypoxia studies, pregnant females were injected i.p. with 1.5 mg Hypoxyprobe (pimonidazole hydrochloride, Hypoxyprobe-1 Inc). After 2 hours, hearts from embryos were harvested and fixed. Pimonidazole is a 2-nitroimidazole that is reductively activated specifically in hypoxic cells and forms stable adducts with thiol groups in proteins, peptides, and amino acids at oxygen levels below 1.3%.

Cell lines

The mouse immortalised embryonic epicardial cell line MEC.1⁵⁴ was purchased from Merck/Millipore (SCC187) and cultured according to manufacturer’ recommendations.

Myocardial Infarction

All surgical and pharmacological procedures were performed in accordance with the Animals (Scientific Procedures) Act 1986 (Home Office, UK). MI was induced by permanent ligation of the left anterior descending (LAD) coronary artery in mice at postnatal day (P) 7, as previously described⁴. Wild type animals received i.p. injection of FG-4592 (Roxadustat, Selleck), BAY 85-3934 (Molidustat, Selleck) (10 mg /kg) or vehicle (DMSO), upon recovery (day 0) and 7 days later. Hearts were harvested at 4- or 9- and 21-days following ligation.

Epicardial explants

Sterile 12-well plates (Fisher Scientific), containing sterile 13-mm diameter coverslips were coated with 0.1% gelatin (Millipore) and allowed to stand for 20 minutes. The gelatin was then replaced by Dulbecco’s Modified Eagle Medium (DMEM) (Sigma) containing 10% FBS (Sigma), 1% Penicillin/Streptomycin (Sigma). Hearts were isolated from E11.5 embryos and the outflow tract and atria were removed. Each ventricle was then cut in half and placed epicardial side down on the coverslip⁵. Explants were kept at 37°C/5% CO₂. 24 hours later, 1 µM of tamoxifen was added to explants. After 48 hours of culture, the explant tissue was carefully peeled off, leaving only the epicardial sheet remaining on the coverslip. For explants prepared from C57BL/6 embryos, 50 µM of FG-4592 (Roxadustat, Selleck), BAY 85-3934 (Molidustat, Selleck) or vehicle (DMSO) was added. After a total of 72 hours in culture, the media was removed, coverslips were washed briefly in ice-cold PBS and then fixed in 4% PFA at room temperature for 15 minutes. Coverslips were subsequently washed in PBS before proceeding with immunostaining.

Histological analysis

Following overnight (O/N) fixation in 2% PFA at 4°C, hearts were washed in PBS and dehydrated by passage through rising concentrations of ethanol. Samples were then washed in Butanol O/N, at RT, before being placed in molten 50:50 butanol: Histoplast paraffin wax (Fisher) at 56°C. After 1 hour the solution was replaced with 100% Histoplast paraffin wax. After several changes of 100% wax, hearts were oriented and embedded into a mould pre-loaded with paraffin. The wax was then rapidly cooled on a bed of ice and stored at 4°C. A microtome was used to cut 10μm-thick sections through the heart. Before staining, sections were first deparaffinised in Histoclear solution (Fisher), followed by rehydration through a decreasing concentration of ethanol. For Masson’s Trichrome Staining, the Masson’s Trichrome kit (Abcam) was used as per the manufacturer’s instructions. Briefly, slides were immersed in Bouin’s solution for 15 minutes at 56°C. Sections were then stained by serial immersion in the following solutions: Weigert’s Iron Haemotoxylin solution (5 minutes); Biebrich scarlet-acid Fuchsin (5 minutes); Aniline Blue (5 minutes); Phosphotungstic/Phosphomolybdic acid solution (5 minutes) and finally 1% Acetic acid solution (2 minutes). Samples were then rinsed in distilled water and dehydrated. Finally, sections were cleared in Xylene and mounted using DPX mounting media.

Immunofluorescence staining

Embryos were harvested at the required embryonic stage, placed in ice-cold PBS (Sigma) and the heart micro-dissected. Similarly, hearts from neonates were removed and washed in ice-cold PBS. Both embryonic and neonatal hearts were fixed for 6 hours in 2% paraformaldehyde (PFA; Santa Cruz Biotechnology) at 4°C and equilibrated in 30% sucrose overnight at 4°C. Hearts were then placed in 50:50 30% sucrose/PBS: Tissue-Tek OCT (VWR) for 30 minutes at room temperature (RT) and embedded in OCT. 10 μm-thick cryosections through the heart were cut. Before use, slides were left to dry for 10 minutes at RT and then washed in PBS for 5 minutes to remove the OCT. Samples were permeabilised with 0.5% (sections) or 0.1% (explants) Triton X-100 in PBS (PBTr) for 10 minutes at RT and subsequently rinsed twice in PBS. Samples were blocked in 10% Serum, 1% bovine serum albumin (BSA, Merck), 0.1% PBTr for 1 hour at RT prior to incubation with the primary antibodies overnight at 4°C. The following day, slides were washed three times for at least 5 min in 0.1% PBTr. Samples were incubated with Alexa Fluor®-conjugated secondary antibodies (1:200 dilution; Invitrogen) and 4',6-Diamidino-2-Phenylindole, Dihydrochloride (DAPI; 0.1 μg/ml, Invitrogen), for 1 hour at RT, protected from light. After final washes, slides were mounted in 50% glycerol in PBS. For wholemount staining, samples were washed in 0.3% PBTr and blocked in 1% BSA (Merck), 0.3% PBTr for at least 2 h. The samples were then incubated with primary antibodies in the blocking solution overnight at 4°C. On the second day, the samples were washed at least five times in 0.3% PBTr and then incubated with secondary antibodies and DAPI (Invitrogen) diluted in PBS overnight at 4°C. The samples were then washed with PBS at least five times the next day and mounted in 50% glycerol in PBS. The following primary antibodies were used: Dylight 549 Mab (Hypoxyprobe-1 Inc, 1:200) antibody to detect hypoxyprobe labelling, HIF-1α (1:100, Novus Biologicals), HIF-2α (1:100, R&D Systems), endomucin (1:50, Santa Cruz Biotech), podoplanin (1:200, Fitzgerald), WT1 (1:100, Abcam), Alexa Fluor™ 488 Phalloidin (1:250, ThermoFischer), actin alpha-smooth muscle Cy3 (1:100, Sigma), Anti-α-Actinin (Sarcomeric) (1:500, Sigma-Aldrich), CD31 (1:100, Abcam), smooth muscle Myosin heavy chain 11 (SM-MHC, 1:100, Abcam), SM22 alpha (1:100, Abcam), Wheat-germ agglutinin (WGA, 1:100, Thermofisher), prolyl hydroxylase domain-2 (PHD2, 1:100, Novus Biologicals).

RNA isolation and qRT-PCR analysis

Total RNA was isolated from frozen ventricles and epicardial explants with Trizol (Invitrogen) using a teflon homogeniser followed by aspiration with a sterile 25-gauge needle and syringe (BD sciences). RNA was transcribed into cDNA utilizing random primers (Promega) with Superscript Reverse Transcriptase III (Life Technologies). Real-time quantitative PCR was performed on a ViiA™ 7 Real-Time PCR System (Applied Biosystems), using SYBR Green mix (Invitrogen). Gene expression was evaluated as DeltaCt relative to control (Atp5b , Sdha and B2M).

Heart dissociation and FACS sorting

Hearts from P1 and P7 mice were harvested and minced into a single cell suspension of FACS sorted live cells. Briefly, for each sample, 3 hearts collected from the same litter were pooled together and finely minced with a scalpel. Tissue was then digested by gentle agitation (180 rpm shaker) in Collagenase II (Gibco,) using a solution of 500 units/ml in HBSS at 37 °C for 45 minutes (P7 hearts) or 20 minutes (P1 hearts). Cell solutions were then passed through a 70 μm filter, washed and incubated in Red Blood Cell lysis buffer (Cell Signaling Technology) for 10 minutes at RT to remove red blood cells. Finally, isolated single cardiac cells were centrifuged, resuspended in 2% FBS in PBS, passed through the filtering cap of the FACS tubes and incubated with 1% 7-AAD viability stain (Invitrogen) for 10 minutes. Approximately 1x10⁵ live cells per sample were sorted using the BD FACSAria Fusion Sorter.

Single cell RNA-Sequencing

FACS sorted cardiac cells viability and concentration were assessed by using an automated cell counter and 1.5x10⁴ cells per sample were loaded onto the 10X Chromium system (10X Genomics) to obtain a target cell recovery of ~6000 cells. Single cell RNA-seq libraries were generated using Single Cell 3 Prime Reagent Kits v1.3 (10x Genomics) according to the manufacturer’s protocol. Sequencing was performed on an Illumina NovaSeq 6000 System operated by the Oxford Genomics Centre at the Wellcome Centre for Human Genetics. Raw sequence reads were aligned against the mouse mm10/GRCm38 reference transcriptome using the Cell Ranger 3.1.0 pipeline (10x Genomics) and processed further using the scRNA-seq analysis R package Seurat (v.3.2.3). Initial filtering removed cells expressing less than 200 genes and genes that were expressed in less than 3 cells. To exclude low quality cells and doublets we filtered out cells with a very high mitochondrial genes content and cells that expressed more than 8000 genes. Based on these criteria, 19211 genes across 5308 and 8734 cells for P1 and P7 samples respectively remained for downstream analysis. Data from the two samples were combined and scaled by regressing out the nUMIs, percentage of mitochondrial gene expression and cell cycle. To correct for batch effect, samples were integrated using the Harmony package (v.1.0). Uniform manifold approximation and projection (UMAP) was performed on the scRNAseq harmonised cell embeddings and unbiased clustering was obtained using the FindCluster function of the Seurat pipeline. Cluster cell types were annotated using a combination of differentially expressed markers, identified using the Seurat FindAllMarkers function and the expression of selected canonical markers for specific cell types. The Epicardial cell cluster was then identified and subset into a new Seurat object with raw reads counts. The standard Seurat pipeline described above was performed and the Model-based Analysis of Single-cell Transcriptomics test (MAST v.1.14.0) was used to analyse the differential gene expression between the P1 and P7 groups in the epicardial cells. Finally, differentially expressed genes were ranked based on both the fold change and p value (avg_logFC * -log10 of the p_val) and the ranked list was submitted to the Metascape platform (www.metascape.org) to identify enriched pathways in the P1 or P7 group, respectively.

Cardiac cine-MRI

Cardiac cine-MRI was performed at 7T using a Varian DDR system. Briefly, mice were anaesthetised with 2% isoflurane in O2, and positioned supine in a custom animal handling system with homeothermic control. Prospectively gated proton cardiac images were acquired with a partial Fourier accelerated spoiled gradient echo CINE sequence (TR 5.9 ms, TE 2.2 ms, 30 kHz bandwidth, 25º FA, approximately 20-30 frames gated to the R wave with a 4 ms postlabel delay; 20% partial acquisition; 4 averages) with a 72 mm volume transmit/4 channel surface receive coil (Rapid Biomedical GmbH) in order to acquire two and four chamber long-axis views and a short axis stack for functional quantification (128x128 matrix; 25.6 mm^2 FOV; 0.7 mm slice thickness, 0.2 mm resolution in-plane). Non-acquired partial Fourier data was reconstructed via the method of projection onto convex sets prior to a simple, cartesian, DFT. Blinded image analysis was performed in ImageJ as described previously⁶.

Wound healing/scratch assay

2 x 10⁴ MEC.1 cells were seeded on a 2 well Culture insert (Ibidi, Germany). The day after, the insert was removed and 50 µM of FG-4592 (Roxadustat, Selleck), BAY 85-3934 (Molidustat, Selleck) or vehicle (DMSO) was added to the culture. Cells were imaged every 2 hours up to 6 hours. The cell-free space area was determined by using the ImageJ software (NIH, Rockville, USA).

Image Analysis

For quantification of immunofluorescence of cryosections, images were captured on an Olympus FluoView 3000 confocal microscope and analysed with Fiji (NIH) and AngioTool⁷. To quantify nuclear fluorescence, a macro was written to identify all DAPI-stained nuclei and to analyse nuclear fluorescence based on the DAPI mask. To localise WT1+ cells in the epicardium and myocardium, cells were manually counted. To quantify the cell size, three independent samples per group with six different fields and positions, two from left and right ventricles, and septum were captured at 40× magnification. ImageJ (National Institutes of Health) was used to quantify the size of each cell.

Statistical Analysis

All data are presented as mean ± standard error of the mean (SEM) or as median, inter-quartile range (IQR) and upper and lower limits. Statistical analysis was performed on GraphPad Prism 8 software. The statistical significance between two groups was determined using an unpaired two-tailed Student’s t-test, these included an F-test to confirm the two groups had equal variances. Among three or more groups, one-way analysis of variance (ANOVA) followed up by Dunnett’s or Tukey’s multiple comparison tests were used. A value of p ≤ 0.05 was considered statistically significant.

Higgins, D. F. et al. Hypoxic induction of Ctgf is directly mediated by Hif-1. Am. J. Physiol. - Ren. Physiol. 287, (2004).
Minamishima, Y. A. et al. Somatic inactivation of the PHD2 prolyl hydroxylase causes polycythemia and congestive heart failure. Blood 111, (2008).
Pitulescu, M. E., Schmidt, I., Benedito, R. & Adams, R. H. Inducible gene targeting in the neonatal vasculature and analysis of retinal angiogenesis in mice. Nature Protocols vol. 5 (2010).
De Villiers, C. & Riley, P. R. A Refined Protocol for Coronary Artery Ligation in the Neonatal Mouse. Curr. Protoc. 1, (2021).
Chen, T. H. P. et al. Epicardial induction of fetal cardiomyocyte proliferation via a retinoic acid-inducible trophic factor. Dev. Biol. 250, (2002).
Carr, C. A. et al. Bone marrow-derived stromal cells home to and remain in the infarcted rat heart but fail to improve function: An in vivo cine-MRI study. Am. J. Physiol. - Hear. Circ. Physiol. 295, (2008).
Zudaire, E., Gambardella, L., Kurcz, C. & Vermeren, S. A computational tool for quantitative analysis of vascular networks. PLoS One 6, (2011).

There is NO Competing Interest.

SupplementaryFigure1.jpg
Supplementary Figure 1
SupplementaryFigure2.jpg
Supplementary Figure 2
SupplementaryFigure3.jpg
Supplementary Figure 3
SupplementaryFigure4.jpg
Supplementary Figure 4
SupplementaryFigure5.jpg
Supplementary Figure 5
SupplementaryFigure6.jpg
Supplementary Figure 6
SupplementaryFigure7.jpg
Supplementary Figure 7
SupplementaryFigure8.jpg
Supplementary Figure 8
SupplementaryFigure9.jpg
Supplementary Figure 9
SupplementaryFigure10.jpg
Supplementary Figure 10
SupplementaryFigure11new.jpg
Supplementary Figure 11
SupplementaryFigure12.jpg
Supplementary Figure 12
SupplementaryFigurelegends.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Stabilisation of HIF signalling extends epicardial activation and neonatal heart regeneration

Status:

Version 1

Abstract

Introduction

Results

The epicardium is hypoxic during development

Activation of HIF signalling improves the response to injury in non-regenerative P7 hearts

Discussion

Declarations

References

Methods

Additional Declarations

Supplementary Files

Status:

Version 1