Macrophage-derived WNTs in normal cardiac growth and regeneration following injury
Castellan, Raphaël Fabrice Paul
Unlike other regenerative organs such as the liver, the adult mammalian heart does not regenerate tissue lost following injury such as myocardial infarction (MI). Instead a non-contractile fibrous scar develops that in the longer term leads to the development of heart failure (HF). In contrast to the adult, neonatal mammals, including mice and man, retain potent cardiac regenerative capacities and can replace myocardium lost following injury. Understanding the mechanisms underlying scar free repair in the neonate may help in development of new approaches to reduce the impact of myocardial injury in adults. In this thesis MI was induced by coronary artery ligation in mice at post-natal day 1 (P1). Novel electrocardiogram gated high resolution cardiac ultrasound was developed to permit non-invasive confirmation of injury 1 day later and regeneration 21 days later by loss, then restoration, of contractile function. Macrophages (MФ) play important roles in organ growth and homeostasis, and are required for scar-free regeneration of the neonatal mouse heart following MI. WNTs are secreted lipophilic proteins with multiple roles in development. MФ-derived WNTs are essential for scar free tissue regeneration following injury in the kidney, liver, and gut, but their role in the heart is unknown. The primary aim of this thesis was to investigate the role of MФ, and in particular MФ-derived WNTs in determining normal growth of the myocardium from neonate to adult and also in regeneration of the neonatal heart following injury. In wild-type neonatal mouse hearts, Csf1r-expressing cells density (mostly macrophages) was consistent across all time points studied. Three populations of resident cardiac mononuclear phagocytes were identified by flow cytometry: F4/80hi, CD11blo, Ly6C-ve - F4/80lo, CD11bhi, Ly6C-ve - F4/80lo, CD11bhi, Ly6C+ve. F4/80hi, CD11blo, Ly6C-ve cells were hypothesised to correspond to yolk-sac derived mononuclear phagocytes and F4/80lo, CD11bhi, Ly6C-ve - F4/80lo, CD11bhi, Ly6C+ve to foetal liver/bone marrow derived mononuclear phagocytes. Three phases of myocardial growth were identified by ultrasound and histological techniques: hyperplastic (P2-P8, with increased Ki67 and cardiac troponin immunopositive cells), hypertrophic/reorganisation (P8-P21, with increasing cardiomyocyte size and no change in left ventricle wall thickness), and finally hypertrophic solely (P21-P42, with increasing cardiomyocyte size and left ventricle wall thickness). Average coronary vessel size was shown to decrease between P2 and P8 whilst vessel density was increased. The number of α-smooth muscle actin (αSMA) coated vessels greatly increased between P8 and P42, indicating vessel maturation. Throughout all phases cardiac systolic function was maintained at steady state. Diastolic function was however shown to mature from a foetal to an adult pattern between P2 and P8, with reversal of the E:A wave ratio on Doppler ultrasound. In mice globally deficient in MФ due to a germline knock-out of the Csf1r gene (Csf1rnull mice), both body and heart weights were decreased from P7 onwards. The number of proliferating (Ki67+ve) cardiomyocytes at P1 and P7 was unchanged in Csf1r-null mice but there was a trend towards decreased cardiomyocyte size at P7, suggesting an influence on hypertrophic rather than hyperplastic growth of the myocardium. There was also a trend for slowed vascular network maturation, with a delay in the shift from large to smaller vessels in hearts from Csf1r-null mice. In mice with MФ-directed (Csf1r-icre mediated) depletion of Porcupine (Porcn), a gene encoding an enzyme required for WNT acylation and secretion cardiac growth, vascularisation, fibrosis and function were all similar in Cre-ve and Cre+ve animals until P41, when cardiomyocyte size and cardiac systolic function were both significantly increased in Cre+ve animals. However, the underlying mechanism is unknown. In the neonatal mice, Csf1r expressing cells, mostly MФ, were identified in association with regenerating myocardium after induction of MI at P1. Flow cytometry data showed that by P7 the putative resident yolk-sac derived population had mostly disappeared from the heart and was replaced by F4/80lo cells, similar to the pattern reported in the adult. In the regenerating myocardium, Axin2 expression was increased consistent with activation of canonical Wnt signalling. Expression of Wnt5b and Fzd2 receptor, both associated with fibrosis, was significantly increased relative to age matched uninjured hearts. MФ-directed depletion of Porcn did not influence either the functional decrease at day 1 or recovery at day 21 following induction of MI at P1. Coronary re-vascularisation was also unaffected by the genotype. However, retention of intra-myocardial fibrosis (picrosirius red staining) was significantly increased in hearts at day 21 post-MI from mice with MФ-directed depletion of Porcn. MФ-derived WNTs are therefore required for scar-free wound healing in the heart, as they are in the liver and the kidney where they regulate matrix metalloproteinase activity. In summary, novel ECG-gated high-resolution in vivo ultrasound developed in this project has allowed characterisation of cardiac structure and function during early post-natal growth and following injury and regeneration in neonatal mice. The resident MФ population of the heart is established pre-natally, and may play a role in determining maturation of the developing vascular network, although this does not involve MФ-derived Wnt signalling. Following MI, the MФ population may expand from bone marrow cells and MФ accumulate around the regenerating tissue. MФ derived WNTs are not required for regeneration of the neonatal myocardium but do have a role in ensuring scar free wound healing and this merits further investigation.