Getting to the heart of intracellular glucocorticoid regeneration: 11β-HSD1 in the myocardium

Corticosteroids influence the development and function of the heart and its response to injury and pressure overload via actions on glucocorticoid (GR) and mineralocorticoid (MR) receptors. Systemic corticosteroid concentration depends largely on the activity of the hypothalamic–pituitary–adrenal (HPA) axis, but glucocorticoid can also be regenerated from intrinsically inert metabolites by the enzyme 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1), selectively increasing glucocorticoid levels within cells and tissues. Extensive studies have revealed the roles for glucocorticoid regeneration by 11β-HSD1 in liver, adipose, brain and other tissues, but until recently, there has been little focus on the heart. This article reviews the evidence for glucocorticoid metabolism by 11β-HSD1 in the heart and for a role of 11β-HSD1 activity in determining the myocardial growth and physiological function. We also consider the potential of 11β-HSD1 as a therapeutic target to enhance repair after myocardial infarction and to prevent the development of cardiac remodelling and heart failure.


Introduction
The major physiological adrenocorticosteroid hor mones, glucocorticoids (cortisol in most animals and corticosterone in rats and mice) and mineralocorticoids (aldosterone) are vital for normal cardiovascular function. They regulate blood pressure (Hunter & Bailey 2015) and vascular tone (Ullian 1999, Dover et al. 2007, Hadoke et al. 2009), as well as heart rhythm and contractility (Lefer 1967, 1968, Penefsky & Kahn 1971, OuvrardPascaud et al. 2005, CruzTopete et al. 2016. Under pathological conditions, rapid corticosteroid release in response to hypothalamicpituitary-adrenal (HPA) axis activation is an early response to cardiovascular insult, including after myocardial infarction (MI). In the acute period, postMI corticosteroids are cardioprotective and suppress the early inflammatory response to injury (Libby et al. 1973, Skyschally et al. 2004). However, sustained excessive glucocorticoid release from the adrenals, for example in Cushing's syndrome, is associated with detrimental cardiac outcomes, including myocardial ischaemia and hypertrophy, and in rare cases, with loss of cardiomyocytes and dilated cardiomyopathy (Peppa et al. 2009, Shibusawa et al. 2015, Frustaci et al. 2016. Chronic stress associates with cardiovascular disease (Kumari et al. 2011), as does pharmacological glucocorticoid treatment (Wei et al. 2004, Souverein et al. 2004. Corticosteroids influence cell behaviour primarily through the intracellular activation of glucocorticoid (GR) and mineralocorticoid (MR) receptors. In the heart, activation of both GR and MR has an impact on cardiac development, physiology and pathophysiology (reviewed in Oakley &Cidlowski 2015 andRichardson et al. 2016). GR and MR are highly related and belong to the nuclear receptor family of transcriptional regulators (Biddie et al. 2010), although some corticosteroid actions may be mediated by nonclassical signalling through cellsurface receptors (Mihailidou & Funder 2005, Samarasinghe et al. 2012. The GR binds glucocorticoids with 10 to 30fold lower affinity than the MR, but is relatively selective for glucocorticoid compared with mineralocorticoid binding (Mihailidou & Funder 2005, Samarasinghe et al. 2012. The MR has high affinity for both mineralocorticoids and glucocorticoids, but as glucocorticoids typically circulate at levels 100fold higher than those of mineralocorticoids, the MR is likely to be constitutively occupied by glucocorticoids even at daily nadir levels. Downstream signalling may differ depending on which ligand occupies the receptor and, depending on prevailing conditions, glucocorticoids can behave as agonists or antagonists at the MR (reviewed in Funder 2012). Access of mineralocorticoids to the MR depends on specific mechanisms that regulate the intracellular availability of adrenocorticosteroid hormones (Funder et al. 1988). These include the extent of protein binding in the plasma and in the heart itself (Bolton et al. 2014, Schafer et al. 2015 and the activity of transporters that actively extrude steroids from the cell. However, glucocorticoids also undergo intracellular metabolism, and the capacity of individual cells to metabolise glucocorticoids is a critical factor in determining the extent (and indeed selectivity) of MR and GR activation (Funder et al. 1988, Odermatt & Kratschmar 2012, Chapman et al. 2013a).

11β-HSD enzymes
11βhydroxysteroid dehydrogenase (11βHSD) catalyses the intracellular interconversion of the glucocorticoids Figure 1 Adrenocorticosteroids, 11β-HSD1 and the heart. Hypothalamic-pituitary-adrenal (HPA) axis-derived glucocorticoids (cortisol in man, corticosterone in rats and mice) and the mineralocorticoid -aldosterone -compete for binding to cardiac glucocorticoid (GR) and mineralocorticoid receptors (MR). Glucocorticoids are also regenerated within the heart from circulating inert metabolites (cortisone in man and 11-dehydrocorticosterone (11-DHC) in rats and mice) by 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1), when it is expressed alongside hexose-6-phosphate dehydrogenase (H6PDH). As the glucocorticoid concentration ([cort]) normally far exceeds that of aldosterone ([aldo]), GR and MR are usually occupied by glucocorticoids. 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2) inactivates glucocorticoids, but unlike other MR target tissues, there is normally little or no dehydrogenase activity in the heart. Inactivation of glucocorticoids by 11β-HSD2 activity elsewhere in the body, largely in the kidney, generates the inert precursors that, on entering the circulation, become available for glucocorticoid regeneration and GR/MR activation in cells that express 11β-HSD1. In the healthy heart, 11β-HSD1 immunoreactivity (inset panel) is localised to vascular smooth muscle (VSM), fibroblasts (F) and also cardiomyocytes (CM) in a fixed section from mouse left ventricle. The 11β-HSD1 antibody is a sheep anti-mouse polyclonal generated in house (De Sousa Peixoto et al. 2008). Staining is notably absent in endothelial cells (EC) lining the vascular wall. A negative control of the same section generated using sheep IgG is shown in the inset panel. From S J McSweeney, PhD thesis, 2010 (McSweeney 2010). 11β-HSD1 expression can be increased by glucocorticoids, by pro-inflammatory Cytokines and in ageing in other tissues and this is also likely to be the case in the heart. cortisol and corticosterone with their inert 11keto forms (cortisone and 11dehydrocorticosterone (11DHC), respectively). 11βHSD type 2 (11βHSD2) is a high affinity (nanomolar KM), lowcapacity NADdependent dehydrogenase (Albiston et al. 1994, Brown et al. 1996 that inactivates the glucocorticoids (Fig. 1). 11βHSD1 shares less than 30% homology with 11βHSD2 (Lakshmi & Monder 1988, Agarwal et al. 1989) and is more widely distributed than 11βHSD2. Under normal conditions, it is coexpressed at the luminal border of the endoplasmic reticulum alongside hexose6phosphate dehydrogenase (H6PDH) (Atanasov et al. 2008) that provides the cosubstrate NADPH (Bujalska et al. 2005). This drives the oxoreductase activity of 11βHSD1, reactivating the glucocorticoids (Atanasov et al. 2004, Bujalska et al. 2005, Lavery et al. 2006, Chapman et al. 2013a (Fig. 1). 11βHSD1 has low affinity for glucocorticoids (micromolar KM) relative to 11βHSD2, but where H6PDH is genetically deleted (Lavery et al. 2006) or where cells are disrupted so that NADPH cannot be generated in close proximity to 11βHSD1, it switches from predominantly oxoreductase to dehydrogenase activity and this inactivates rather than regenerates glucocorticoids (Atanasov et al. 2004). Activity of 11βHSD1 (largely in the liver) and 11βHSD2 (largely in the kidney) determines active and ketoisoform concentrations in the systemic circulation, but the variability in the activity of 11βHSD enzymes at the level of target cells adds a cellspecific dimension to the control of steroid action (Chapman et al. 2013a). To date, there has been relatively little focus on the capacity of the myocardium or of its component cells to metabolise corticosteroids.

11β-HSD2 and MR activation in the heart
In mineralocorticoid target tissues, 11βHSD2 is expressed alongside the MR, where its activity reduces the availability of glucocorticoid, permitting aldosterone to compete for binding to the MR (Edwards et al. 1988, Funder et al. 1988). The clinical efficacy of MR antagonists after MI (Pitt et al. 2003) and in heart failure (Pitt et al. 1999, Zannad et al. 2011, 2012 are attributed, at least in part, to blockade of MR activation in the heart. A recent preclinical study in adrenalectomised rats has shown that mineralocorticoids regulate cardiac electrical function through the activation of MR (CruzTopete et al. 2016). However, the weight of evidence suggests that there is normally little or no 11βHSD2 activity in the heart (Walker et al. 1991, Whorwood et al. 1992, except perhaps in vascular endothelium (Walker et al. 1991, Smith et al. 1996. This was confirmed in a recent in vivo study on patients under going diagnostic coronary angiography that failed to find any evidence for inactivation of the stable isotope tracer 9,11,12,12[2H]4cortisol across the human heart (Iqbal et al. 2014). Furthermore, infusion of the MR antagonist canrenoate in the same patients resulted in the elevation of cortisol collected from the coronary sinus. These data support the view that the cardiac MR is normally occupied by glucocorticoid, rather than by aldosterone. However, this situation may change in pathological conditions, including hypoxia (Klusonova et al. 2009), in which there is evidence for enhancement of 11βHSD2 activity and expression (Chai et al. 2010, Chapman et al. 2013a, that may allow aldosterone to compete for binding to the MR. Selective overexpression of 11βHSD2 in cardiomyocytes leads to MRdependent hypertrophy and fibrosis in mice (Qin et al. 2003), showing potential for a role in cardiac pathology. Global 11βHSD2 knockout mice have cardiac hypertrophy and fibrosis but, as these mice are also hypertensive (Kotelevtsev et al. 1999), any phenotype that is specifically due to lack of 11βHSD2 in the heart has proved impossible to dissect. Generation of mice with targeted 11βHSD2 deletion in cardiomyocytes and other cardiac cells would provide a more definitive answer regarding the role of dehydrogenase activity in cardiac physiology and pathology.

Glucocorticoid regeneration by 11β-HSD1 in the heart
11βHSD1 is more widely expressed in the body than 11βHSD2, but tissue expression of 11βHSD1 is often reported to be low. This is also the case in the heart or at least in whole ventricular homogenates (Walker et al. 1991. However, immunoreactive 11βHSD1 can be localised to cardiomyocytes, to interstitial and adventitial fibroblasts (Brereton et al. 2001) and to vascular smooth muscle in murine heart ( Fig. 1) (McSweeney 2010). We have found that targeted deletion of 11βHSD1 in cardiomyocytes and vascular smooth muscle cells leads to a significant reduction in Hsd11b1 gene expression in the mouse heart, confirming its expression in these cells . Although endothelial cells in culture are reported to have 11βHSD1 activity (Brem et al. 1998), previous studies have concluded that the 11βHSD enzyme in aortic endothelial cells is primarily a dehydrogenase, with oxoreductase activity limited to the smooth muscle cell compartment of the vascular wall (Walker et al. 1991, Dover et al. 2007. Immunostaining suggests that this is also the case in the coronary vessels R4 Review 58 1 : R4 Review 58 1 :

R4 Review
g a gray and others 11β-HSD1 and the heart 58 1 : R4 Review ( Fig. 1). Fibroblasts are among the most abundant cells in the heart, and transcriptomic analysis has revealed that expression of Hsd11b1 is high in cardiac fibroblasts relative to fibroblasts elsewhere in the body (Furtado et al. 2014), consistent with a potentially important role here.
Myocardial expression of H6PDH (Brereton et al. 2001, GomezSanchez et al. 2008, and H6PDH immunoreactivity in cells identified as fibrocytes (Gomez Sanchez et al. 2008), suggests that NADPH will be available to support oxoreductase activity in the heart. However, conversion of 11DHC to corticosterone is comparatively low in whole ventricular homogenate from mice and rats (Walker et al. 1991, Sheppard & Autelitano 2002. Similarly, although HSD11B1 mRNA is present in human heart (Chai et al. 2010), cortisol is not significantly regenerated from cortisone infused in vivo into the hearts of patients undergoing elective coronary angiography (Iqbal et al. 2014). Thus, although there is plentiful evidence that myocardial cells express 11βHSD1, reactivation of glucocorticoid is apparently limited under physiological conditions. Nevertheless, our phenotyping studies in mice with genetically induced 11βHSD1 deficiency show that 11βHSD1 activity has a role in determining normal postnatal growth of the heart ( Fig. 2 and below), as well as Ca 2+ handling and diastolic function (White 2016). 11β-HSD1 in postnatal growth of the myocardium GR activation during a prenatal surge in plasma glucocorticoid is essential for the structural and functional maturation of the heart before birth (RogZielinska et al. 2014(RogZielinska et al. , 2015, and consequently, genetic deletion of GR from the heart and smooth muscle results in increased perinatal mortality (RogZielinska et al. 2013(RogZielinska et al. , 2014. 11βHSD1 is not expressed to any extent in the mouse heart in the prenatal period (Speirs et al. 2004), and neither global (Kotelevtsev et al. 1997) nor cardiomyocyte and vascular smooth muscle cellspecific ) deletion of Hsd11b1 has any influence on perinatal mortality. However, a role for 11βHSD1 in the maturation of the postnatal heart is indicated by our observation that hearts from adult mice with global deficiency of 11βHSD1 are smaller than those of wildtype mice ( compared with control wild-type mice (filled bars). Reduction in heart size was confirmed when post-mortem heart weight was compared with body weight (B). LV dimensions (left ventricular end-diastolic area, solid red (Hsd11b1 −/− ) and black (control wild type), and left ventricular end-systolic area, hatched red (Hsd11b1 −/− ) and black (control wild type)) were not influenced by the loss of 11β-HSD1 from late gestation until 6 weeks of age (C). In adult hearts, there was no effect of 11β-HSD1 deficiency (red) on the distribution of cardiomyocyte crosssectional area (D), but the length of individual cardiomyocytes (E) isolated from Hsd11b1 −/− hearts was significantly less than that isolated from control wild-type hearts (black).*P < 0.05, ***P < 0.005, n = 5-8.
from proliferative to hypertrophic growth . MR activation in the same period promotes proliferation (Feng et al. 2013). Proliferative neonatal cardiomyocytes can reactivate glucocorticoids through 11βHSD1 oxoreductase activity during the period that they retain proliferative capacity (Lister et al. 2006). However, the hearts of globally 11βHSD1deficient mice have normal ventricular dimensions from E18.5 to at least P40 (Fig. 2B). Mice with targeted deletion of 11βHSD1 in cardiomyocytes and vascular smooth muscle also have phenotypically normal hearts . Therefore, loss of cardiomyocyte corticosteroid receptor activation in the perinatal period does not account for the smaller size of hearts in adult 11βHSD1 mice. Interestin gly, further analysis revealed that although cardiomyocyte crosssectional area is normal in adult hearts of 11βHSD1 deficient mice (Fig. 2C), cardiomyocyte length is reduced ( Fig. 2D; White 2016). Cardiomyocytes are influenced by physical interactions and paracrine signalling from other cells in their environment. Blood volume and blood pressure are not modified in mice with global 11βHSD1 deficiency, so altered in vivo haemodynamic stress is not likely to underlie the reduction in cardiomyocyte length (White 2016). Further investigation is required to find whether the loss of 11βHSD1 in other myocardial cells, for example, resident macrophages or fibroblasts can indirectly influence cardiomyocyte size during postnatal growth. In terms of clinical relevance, it is notable that a single nucleotide polymorphism in the HSD11B1 gene is also associated with reduced left ventricular mass (Rahman et al. 2011), although the underlying mechanism was not investigated.

11β-HSD1 and myocardial pathology
Although cellular 11βHSD1 expression and activity may normally be low in the heart, both can change rapidly in response to external stimuli (reviewed in Chapman et al. 2013b). Glucocorticoid itself promotes 11βHSD1 expression (Sai et al. 2008, Morgan et al. 2014, and there is good evidence that 11βHSD1 is the major regulator of the tissuespecific effects of circulating glucocorticoid excess (Morgan et al. 2014). The proinflammatory cytokines IL1 and TNFα are also key activators of 11βHSD1 gene expression, notably in fibroblasts (Hardy et al. 2006, Stegk et al. 2009, Ahasan et al. 2012, Chapman et al. 2013b, Esteves et al. 2014). These cytokines can act alone or synergistically with glucocorticoid (Kaur et al. 2010, Ahasan et al. 2012, to effect an increase in the capacity for intracellular glucocorticoid regeneration by 11βHSD1. After trauma or injury, proinflammatory conditions that potently increase HPA axis activity (and thereby plasma 11DHC/cortisone levels (Harris et al. 2001)), increased 11βHSD1 activity is likely to be a significant contributor to intracellular glucocorticoid levels and thus GR and MR activation. The generation of mice lacking 11βHSD1 (Kotelevtsev et al. 1997) and more recently, the development of specific 11βHSD1 inhibitors (Hughes et al. 2008, Wheelan et al. 2014, Sooy et al. 2015, have provided the opportunity to identify such roles for amplification of intracellular glucocorticoid signalling by 11βHSD1 in cardiac pathophysiology.

11β-HSD1, MI and heart failure
MI occurs most commonly after the formation of a clot on a ruptured atherosclerotic plaque, leading to ischaemic cardiomyocyte death in the area served by the affected coronary artery. As the adult heart has only very limited ability to regenerate, MI is followed by a period of wound healing that results in the formation of a scar to maintain myocardial integrity (Video 1; Frangogiannis 2014). The extent of cardiomyocyte loss during ischaemia and expansion of injury during infarct repair are critical determinants of subsequent structural remodelling and functional deterioration leading to heart failure (Di Bella et al. 2013).

Video 1
Coronal view of mouse heart collected 7 days after induction of myocardial infarction and imaged by optical projection tomography, as we have described in (Zhao et al. 2015). Progressive optical sectioning through the heart reveals the extensive, thinned infarct area of the left ventricle. In 11β-HSD1 deficient mice the length of the infarcted area is reduced and the ventricle wall is less thinned , White 2016 (Garcia et al. 2013, Kipari et al. 2013, of 11βHSD1 is associated with reduction in plaque size independently of changes in lipid availability. Plaque size reduction in ApoE/11β HSD1 double knockout mice was attributed to loss of 11βHSD1 in bone marrowderived cells (Kipari et al. 2013). These cells are recruited to the arterial wall during plaque formation, where they accumulate lipid and contribute to the determination of plaque stability. Neutrophils, monocyte/macrophages and T cells express 11βHSD1  (Zhang & Daynes 2007, Zhang et al. 2005, and expression is increased during their mobilisation and activation (reviewed in Chapman et al. 2013b). Intracellular generation of glucocorticoid in monocyte/macrophages and neutrophils regulates their recruitment to sites of inflammation (Tiganescu et al. 2011), phagocytic potential (Speirs et al. 2004) and the release of proinflammatory molecules (Zhang & Daynes 2007). 11βHSD1 inhibition may therefore be of therapeutic benefit in preventing the development of complex plaques that are vulnerable to rupture causing MI.
Mice with global deficiency of 11βHSD1 also have reduced structural and functional remodelling after MI, induced by coronary artery ligation, and do not go on to develop heart failure . Thus, 11βHSD1 availability has an impact not only on the development of coronary artery disease that leads to MI but also on the response of the myocardium after MI. Increased plasma corticosterone, derived from HPA axis activation in the immediate postinfarct period, reduces cardiomyocyte death and infarct size and there is evidence supporting roles for both GR (HafeziMoghadam et al. 2002) and MR (Mihailidou et al. 2009, Fraccarollo et al. 2011 in determining this outcome. However, in 11βHSD1 deficient mice, plasma corticosterone is increased to the same extent after MI as in wildtype mice and infarct injury is unchanged . Therefore, differences in the immediate response to MI do not underlie the subsequent reduction in cardiac remodelling in the absence of 11βHSD1. However, periinfarct angiogenesis is enhanced in 11βHSD1deficient mice during repair (Small et al. 2005), and this is associated with reduced infarct expansion in this period and with eventual scar size . Glucocorticoids suppress angiogenesis, and prevention of intracellular glucocorticoid regeneration in 11βHSD1deficient mice promotes angiogenesis in different in vitro and in vivo experimental models (Small et al. 2005). Smooth muscle cells are the main site of intravascular 11βHSD1 activity (Hadoke et al. 2013), but targeted deletion of Hsd11b1 in cardiomyocytes and vascular smooth muscle cells failed to enhance angiogenesis or improve outcome after MI . Engagement of extravascular proangiogenic mechanisms is therefore indicated. Increased angiogenesis in 11βHSD1deficient mice after MI is preceded by increased content of reparative 'M2' macrophages during the proliferative phase of wound healing . These 'M2' macrophages are essential for infarct repair (Shiraishi et al. 2016) and can release growth factors and proangiogenic mediators that increase neovascularisation during wound healing, as well as for promoting the formation of a mature collagen scar. Enhancement of macrophage polarisation in mice lacking 11βHSD1 is therefore likely to be a key mechanism underlying improved outcome after MI (Fig. 3). Increased 'M2' polarisation of macrophages could occur due to loss of 11βHSD1 in the inflammatory cells themselves, as proposed in atherosclerosis (Kipari et al. 2013), resulting in the modification of inflammatory cell recruitment and phenotype (Frangogiannis 2014).

Figure 3
Distribution and actions of 11β-HSD1 in physiology and post-MI. In the healthy heart (panel A), 11β-HSD1 is expressed in cardiomyocytes (cm), vascular smooth muscle (VSMC) and fibroblasts (fib). Deletion of 11β-HSD1 (panel B) results in smaller adult hearts and mild diastolic dysfunction (identified by increased deceleration time on Doppler mitral valve (MV) ultrasound). After MI (panel C), 11β-HSD1 is additionally expressed in myofibroblasts (myofib) and recruited inflammatory cells (T cells, neutrophils (neut), monocyte (mono) and macrophages (mac)). 11β-HSD1 deficiency after MI (panel D) results in increased early neutrophil (neut) recruitment, increased acquisition of pro-repair 'M2' macrophage (mac) phenotype, relative to pro-inflammatory 'M1' macrophage phenotype and increased angiogenesis, leading to increased vessel density.
Macrophage MR knockout (Usher et al. 2010), or GR activation (Gratchev et al. 2008), results in polarisation towards an 'M2' phenotype, and MR antagonists increase 'M2' macrophage polarisation in the heart when given after MI (Fraccarollo et al. 2008). Loss of glucocorticoid regenerative capacity in macrophages could therefore conceivably underlie changes in the activation status by preventing the activation of MR. However, promotion of periinfarct angiogenesis after MI is enhanced when GR is blocked (Small et al. 2005), supporting a more indirect mechanism that involves GR, rather than MR activation.
Neutrophil recruitment is increased in the hearts of 11βHSD1deficient mice after the induction of MI   (Fig. 3). This results at least in part from loss of 11βHSD1 in neutrophils themselves, where its activity restrains recruitment to inflamed tissues through modulation of adhesion molecule availability at the cell surface (Cavalcanti et al. 2006, Tiganescu et al. 2011, Coutinho et al. 2016. Loss of 11βHSD1 activity in the heart itself is also likely to contribute. Expression of neutrophil chemoattractants, including CXCL5/LIX (Smith & Herschman 1995), IL6 (Hardy et al. 2006) and CXCL2/MIP2α (Uhlenhaut et al. 2013) is repressed by glucocorticoid, and these chemoattractants are expressed by cells in the heart that express 11βHSD1, including fibroblasts. Mast cell degranulation after myocardial injury also enhances neutrophil recruitment, and as degranulation is increased in the absence of 11βHSD1 (Coutinho et al. 2013), these cells may also play a role in increasing neutrophil recruitment in 11βHSD1deficient mice. In the postinfarct heart, neutrophils phagocytose necrotic myocytes and break down the intracellular matrix during the initial inflammatory phase of repair. However, they also promote the transition to resolution by releasing gelatinaseassociated lipocalin (Horckmans et al. 2016). Efferocytosis, or macrophage engulfment of apoptotic cells, including neutrophils (Wan et al. 2013), will also enhance the acquisition of an 'M2' phenotype. Early enhancement of neutrophil recruitment may thus be an essential prerequisite for the promotion of repair in the mice lacking 11βHSD1.
Several lines of evidence point to cardiac fibroblasts, particularly their high expression of 11βHSD1 (Furtado et al. 2014), as a key site for intracellular regeneration of glucocorticoid in the healing myocardial infarct (Chen & Frangogiannis 2013). 11βHSD1 activity in stromal cells elsewhere, including synovial fibroblasts, is promoted by the action of corticosteroids and inflammatory mediators (Hardy et al. 2006, Ahasan et al. 2012. In the skin, 11βHSD1 is active in dermal fibroblasts (Tiganescu et al. 2011) and global 11βHSD1 knockout or pharmacological inhibition promotes skin wound healing (Tiganescu et al. 2013). Fibroblasts are abundant in the heart and have multiple potential roles during myocardial inflammation and repair that follows MI (Porter & Turner 2009). In addition to the production of collagen and other matrix proteins that determine scar integrity, they secrete molecules, including cyto kines, prostaglandins (Shinde & Frangogiannis 2014, Fernando et al. 2015, Turner 2016) and microRNA (Fang & Yeh 2015), capable of regulating inflammation (Porter & Turner 2009, Chen & Frangogiannis 2013, van Nieuwenhoven & Turner 2013, Treg recruitment (Frangogiannis 2014), angiogenesis (Newman et al. 2011), cardiogenesis (Furtado et al. 2014) and hypertrophy (Abonnenc et al. 2013, Cartledge et al. 2015. The role of cardiac fibroblast 11βHSD1 in regulating the cellular secretome and on the response to MI and other pathological challenges merits further investigation.

Therapeutic potential of 11β-HSD1 inhibition in myocardial disease
Pharmacological inhibitors of 11βHSD1 have already been developed for use in metabolic disease and for prevention of cognitive decline (Anderson & Walker 2013, Sooy et al. 2015. Survival after MI is increasing, thanks to prompt and efficient intervention to restore perfusion to the ischaemic myocardium (BHF Heart Stats, 2015). However, patients survive with damage to their myocardium that spreads to the periinfarct area during wound healing. Thus, although the introduction of coronary revascularisation has been associated with a reduction in early postMI deaths, there has been an increase in the 5year incidence of HF (Velagaleti et al. 2008). Preclinical data are highly supportive of the therapeutic potential of 11βHSD1 inhibitors in the prevention of atherosclerosis (HermanowskiVosatka et al. 2005, Garcia et al. 2013, Kipari et al. 2013, as well as acutely following MI, to prevent the infarct expansion that promotes the development of heart failure . Ventricular hypertrophy is a feature of the adaptive ventricular remodelling that follows MI and in pressure overload associated with hypertension. Gordon and coworkers (2014) showed that an inhibitor of 11βHSD1 was able to reverse established hypertrophy and associated dysfunction in a mouse model of perfusion deficit induced cardiac remodelling. Interestingly, this outcome was independent of any influence on neovascularisation or perfusion. Inhibition of 11βHSD1 may therefore have additional benefits on postMI remodelling beyond the prevention of infarct expansion. In clinical trials of 11βHSD1 inhibitors in diabetes, improvement in blood glucose and other metabolic end points failed to reach significance when compared with existing therapies, but the evidence for a modest benefit was highly consistent (Anderson & Walker 2013). Such benefits in metabolic outcomes are likely to be advantageous in patients with cardiovascular disease. Together, these data demonstrate the potential for 'repurposing' of 11βHSD1 inhibitors already developed for clinical use in other therapeutic areas. None of the trials of these inhibitors in diabetes have raised any question on the effectiveness of target engagement or on toxicology. Improved understanding of the cellular mechanisms through which 11βHSD1 inactivation enhances angiogenesis prevents infarct expansion, and adverse ventricular remodelling will help to justify the clinical development of 11βHSD1 inhibitors for MI therapy, as well as providing biomarkers to monitor the effects of 11βHSD1 inhibition.
MR antagonists are already used effectively in the treatment of heart failure (Pitt et al. 1999, Zannad et al. 2011, 2012, by preventing MR activation in cardiomyocytes (Fraccarollo et al. 2011) and in macrophages (Fraccarollo et al. 2008). Preclinical studies have revealed the potential benefits of blocking GR in cardiovascular disease (Small et al. 2005, Oakley & Cidlowski 2015), but essential physiological roles of GR activation preclude the use of GR antagonists in vivo. With their distinct profile of action, 11βHSD1 inhibitors offer an alternative means of regulating GR and MR activation in specific cells. 11βHSD1 inhibitors might provide an alternative to MR antagonists in patients who do not tolerate the effects of MR antagonism on K + handling (Pitt et al. 2008). Alternatively, they may have additional or complementary benefits when combined with MR antagonists, and this should be tested in models of cardiovascular disease.
In this review, we have considered 11βHSD1 in the context of its capacity for intracellular regeneration of active glucocorticoids from inert metabolites. However, 11βHSD1 has other actions that may contribute to the physiological and pathophysiological outcomes reported previously. 11βHSD1 is a multifunctional carbonyl reductase that also converts 11 and 7oxosterols into the respective 7hydroxylated forms and also catalyses the reduction of nonsteroidal xenobiotics (Mitic et al. 2013b, Odermatt & Klusonova 2015. Studies with selective inhibitors and genetically modified mice provide further evidence for the effects of 11βHSD1 in the regulation of cellular cholesterol flux and on bile acid homeostasis (Mitic et al. 2013a). Oxysterols influence vascular function (Mitic et al. 2013a), and modulation of genes involved in cellular cholesterol flux is associated with disease severity in patients after MI (Suresh et al. 2014). These mechanisms may therefore be engaged alongside prevention of intracellular glucocorticoid regeneration when 11βHSD1 is genetically depleted or inhibited and could contribute to any beneficial effects of drug intervention in cardiovascular pathology.

Conclusion
Studies over the past three decades have demonstrated roles for the 11βHSD enzymes in metabolic, neurological and inflammatory diseases, but roles in the heart have been relatively unexplored. Recent studies have shown that 11βHSD1 is expressed in myocardial cells, including notably cardiac fibroblasts. Global 11βHSD1 deficiency has effects on cardiac growth and physiological function that are more subtle than those associated with GR deletion (Oakley & Cidlowski 2015), or adrenalectomy (CruzTopete et al. 2016), but nevertheless of physiological relevance. Importantly, genetic or pharmacological depletion of 11βHSD1 also prevents adverse cardiac remodelling and the development of heart failure after MI , Gordon et al. 2014). In the skin, 11βHSD1 expression is increased during ageing (Tiganescu et al. 2011) and 11βHSD1 inhibition prevents ageinduced structural defects (Tiganescu et al. 2013). If 11βHSD1 expression is similarly increased in the heart, the influence of 11βHSD1 on myocardial physiology and pathophysiology is likely to be even more pronounced in the aged heart. Novel mass spectrometry imaging has recently permitted accurate localisation of 11βHSD1 'intracrinology' in sections of mouse adrenal gland and brain (Cobice et al. 2013). Application of mass spectrometry imaging to the heart will be of significant value to understand cellular 11βHSD1 activity and how it is changed in ageing and disease. Whether the key intracellular mechanisms of 11βHSD1 are dependent or independent of glucocorti coid regeneration, preclinical studies have started to reveal novel therapeutic potential for drugs that inhibit 11βHSD1. Given that 11βHSD1 expression and activity can be rapidly increased by inflammatory mediators and by glucocorticoids, there are likely to be further applications for these drugs in cardiovascular disease,