Androgen modulates cardiac fibrosis contributing to gender differences on heart failure

Abstract

Chronic heart failure (HF) is a major health problem throughout the world. Gender has significant effects on the pathophysiology of HF. Low levels of free testosterone are independently associated with increased chronic HF symptoms and mortality. Cardiac fibrosis plays a pivotal role in structural remodeling in HF. Transforming growth factor (TGF)-β, angiotensin II and oxidative stress contribute to the activity/extent of cardiac fibrosis. Androgen deficiency can up-regulate TGF-β expression under angiotensin II stimulation in vivo. In this review, we summarized the potential mechanisms accounting for the effects of androgen on cardiac fibrosis through regulating fibrocytes activity under TGF, which can explain wound healing and cardiac fibrosis in male with acute myocardial injury and chronic HF.

• Androgen
• fibroblasts
• fibrosis
• heart failure
• testosterone
• transforming growth factor-β

Introduction

Heart failure (HF) is a common cause of hospitalization with a poor prognosis in clinical practice [1–4]. Gender plays an important role in the pathophysiology of HF. Female HF patients have more diastolic HF, but male HF is associated with more cardiac ischemia [5–7]. Androgen treatment was shown to improve clinical outcome in male patients with HF [8,9]. HF is associated with increased cardiac fibrosis and structural remodeling [10,11]. Transforming growth factor (TGF)-β, angiotensin (Ang) II, oxidative stress play important roles in the pathophysiology of HF and were also demonstrated to increase the activity of cardiac fibroblasts [12–14]. Estrogen was shown to modulate cardiac fibrosis through inhibition of Ang II-induced collagen synthesis and gel contraction of cardiac fibroblasts [15,16]. Moreover, androgen was shown to modulate the effects of Ang II and possesses an anti-oxidative ability, which may modulate cardiac fibrosis and play a role in the pathophysiology of HF [17,18]. In this review, we summarized the potential mechanisms accounting for the effects of androgen on cardiac fibrosis and HF.

Androgen in HF

In male HF patients, anabolic hormone depletion is common, and reduced levels of androgen and insulin growth factor-1 are related to unfavorable outcomes in men with HF [19]. Wehr et al. [20] proved that low levels of free testosterone are independently associated with increased chronic HF mortality and symptoms [21]. Therefore, androgen deficiency may be responsible for some of the features of advanced HF such as reduced mass of skeletal muscle, abnormal energy handling, cachexia, low mood, depression and fatigue [8,9,22,23]. A randomized, double-blind, placebo-controlled cross-over trial in 2003 reported that administration of testosterone via the buccal route acutely increased cardiac output in men with stable HF [24]. Malkin et al. [8] showed that in men with moderate severity systolic HF, testosterone therapy significantly improved their functional capacity and symptoms, but did not change their left ventricular (LV) ejection fraction. Another trial also showed that long-acting testosterone treatment could improve the functional exercise capacity, skeletal muscle performance, insulin resistance and baroreflex sensitivity, but not the LV ejection fraction, in elderly patients with chronic systolic HF [25]. Moreover, Iellamo et al. [26] found that testosterone therapy improves functional capacity, insulin resistance and muscle strength in women with advanced chronic HF, although the dosage of testosterone used in this study was significantly lower than those in males [8,9,25]. Previous studies revealed that testosterone induce hypertrophy in cardiac myocytes via direct, receptor-specific mechanisms [27] or mammalian target of rapamycin complex 1 pathway [28]. Athletes abusing supraphysiological dose of testosterone may exhibit LV hypertrophy [29,30]. In contrast, there was a trend to a reduction in LV mass index in male patients with moderate severity HF [8]. Therefore, different dosage of testosterone may have inconsistent effects. However, there is no information about the effects of testosterone replacement therapy specified for HF patients with hypertrophic cardiomyopathy. Table 1 summarizes outcomes of clinical trials of testosterone on HF patients. It seems that testosterone supplementation acts only through peripheral mechanisms according to the latter two trials. However, whether testosterone act directly on patient’s LV diastolic function has not yet been proven in a large clinical trial.

Table 1. Clinical outcome of testosterone therapy in chronic HF.

Authors	Patient characteristics	Testosterone therapy regimen	Conclusion
Pugh et al. [24]	Male patients (over 18 years old) with stable chronic HF and objective evidence of impaired LV function	Testosterone 60 mg via the buccal route one tablet per day for two days	Administration of testosterone acutely increased cardiac output, apparently via reduction of the LV afterload
Pugh et al. [9]	Male patients (over 18 years old) with chronic HF of at least six months duration and no other malignant or debilitating disease	Testosterone (20 mg testosterone propionate, 40 mg testosterone phenylpropionate and 40 mg testosterone isocaproate) via an intramuscular injection every two weeks for 12 weeks	Testosterone treatment was safe, well tolerated and led to significant improvement in the physical capacity and symptoms
Malkin et al. [8]	Male patients (over 18 years old) with stable chronic HF of at least six months and at least moderate LV systolic function on 2D trans-thoracic echocardiography	Testosterone 5 mg supplied as an adhesive skin patch preparation with one patch per day for 12 months	Testosterone replacement therapy improved the functional capacity and symptoms without changing LV systolic function in men with moderately severe HF
Caminiti et al. [25]	Elderly male patients with symptomatic HF with NYHA functional class II or III and an LV ejection fraction of <40%	Testosterone undecanoate 1000 mg via an intramuscular injection every six weeks for 12 weeks	Long-acting testosterone therapy improved the exercise capacity, muscle strength, glucose metabolism and baroreceptor cardiac reflex sensitivity in men with moderately severe HF
Iellamo et al. [26]	Female patients with symptomatic HF with NYHA functional class II or III and an LV ejection fraction of <40%	Testosterone 300 μg supplied as an adhesive skin patch preparation with two patches per week for 24 weeks	Testosterone supplementation therapy improves functional capacity, insulin sensitivity and muscle strength in elderly female patients with chronic HF

Myocardial ischemia is an important factor of HF. Androgen deficiency has been documented as a risk factor for a plethora of disease associated with atherosclerosis including the metabolic syndrome, insulin resistance and diabetes [31,32]. Low levels of testosterone are associated with low levels of high-density lipoprotein cholesterol (HDL-C) and high levels of low-density lipoprotein cholesterol (LDL-C) [33,34]. Testosterone therapy was proven to reduce total cholesterol and LDL-C levels [35,36]. However, the effects of testosterone therapy on HDL were controversial [37–41]. Testosterone therapy has been shown to increase HDL-C levels in male patients with metabolic syndrome and hypogonadism [35,37]. In contrast, another trials demonstrated testosterone therapy did not change HDL-C levels in hypogonadal and healthy elderly men [38,39] or even reduced HDL-C levels in healthy men [40,41]. These discrepancies might be caused by a different dosage or a route of androgen administration, or a different hypogonadal state in these patients [42].

Cardiac fibrosis in HF

Adverse LV remodeling is considered an intermediate phenotype of HF [10]. Cardiac fibrosis is characterized by the net accumulation of extracellular matrix (ECM) in the myocardium and the concentration of type I collagen is a major determinant of myocardial diastolic stiffness [11,43]. ECM remodeling is an essential process in cardiac remodeling, hypertensive cardiac hypertrophy and post-infarction healing [44,45]. Post-infarction remodeling is divided into an early phase (within 72 h) and late phase (beyond 72 h) [46]. The infarct zone expansion, which occurs in the early phase, results from degradation of intermyocyte collagen frameworks by serine proteases and activation of matrix metalloproteinases (MMPs) released from neutrophils [47,48]. Infarct expansion causes wall thinning and ventricular dilatation, and enhances diastolic and systolic wall stress [46]. Local Ang II is activated by wall stress and induces cardiac hypertrophy in non-infarcted zones [49]. During late remodeling, TGF-β activates type I and III collagen synthesis and promotes myocardial fibrosis [46]. A fundamental characteristic of hypertensive cardiac remodeling is myocardial stiffness [50]. Myocardial stiffness is mostly due to changes in the composition and arrangement of ECM proteins, including type I and III collagens [51,52]. Compared with normotensive rats, spontaneously hypertensive rats show greater amount of pro-collagen type I in the myocardium [53,54]. Following the development of cardiac hypertrophy, spontaneously hypertensive rats also have more collagen cross-linking in the myocardium [55]. Blood pressure increase is accompanied by increased collagen and total protein synthesis and diminished protein degradation in rodent hearts [56,57]. These evidences suggest that hypertension might contribute to changes in the ECM and fibrosis and lead to HF in the future. In patients with chronic HF, a higher level of cardiac fibrosis is associated with poor outcomes [58]. A higher level of collagen scar formation is correlated with adverse LV function in patients after acute myocardial infarction (MI) [59]. In patients with hypertensive heart disease, fibrosis may also contribute to HF and other cardiac complications [60].

TGF-β and Ang II are likely to be the key driving forces culminating in fibrosis [61]. TGF-β expression is elevated in response to injury [62]. Increased messenger RNA levels of TGF-β1 were observed in cardiac fibroblasts isolated from rats one week after experimental MI [63]. It critically modulates the cardiac fibroblast phenotype and function [64]. TGF-β induces differentiation of cardiac fibroblasts to myofibroblasts, which further stimulate collagen production [12,65]. It also induces fibroblasts to synthesize and contract ECM protein [66,67]. In addition, TGF-β exerts potent matrix-preserving actions by suppressing the activity of MMP and inducing synthesis of protease inhibitors, like tissue inhibitors of metalloproteinases (TIMPs) [63,64]. LV hypertrophy is associated with increased levels of circulating TIMP 1 but decreased levels of circulating MMP 1 in patients with hypertension [44]. Myocardial levels of Ang II increase in a number of pathologies characterized by myocardial remodeling [68]. Ang II can induce proliferation of cardiac myofibroblasts from the non-infarcted zone of rat hearts with MI [13]. Ang II also stimulates the pro-fibrotic effects of cardiac fibroblasts via multiple mechanisms, including increased ECM protein synthesis, decreased MMP activity and increased TIMP activity [69]. Previous studies revealed that Ang II and TGF-β1 do not act independently of one another but act as a network that promotes cardiac fibrosis. Ang II stimulates TGF-β1 gene expression and protein production in cardiac fibroblasts and myofibroblasts which may act as an autocrine/paracrine stimulus to collagen formation [70]. A further study indicated that Ang II was not able to induce cardiac hypertrophy and fibrosis in vivo without TGF-β [71].

Interactions of androgen and TGF-β

A plethora of studies evaluated interactions between androgen and the TGF-β signaling pathway. In different cell lines, androgen has different influences on the TGF-β signal pathway. TGF-β1 may be an important component of the inhibitory complex influencing androgen receptor positive cell proliferation. In adrenalcortical carcinoma cell line, dihydrotestosterone stimulated significant up-regulation of TGF-β1 messenger RNA and protein that reduced proliferation of the adrenocortical cancer cell line in vitro [72]. In ovarian cancer cell line, dihydrotestosterone was able to reverse TGF-β1 growth-inhibitory action. Thus, androgens may induce ovarian cancer progression through decreasing TGF-β receptor levels, thereby allowing ovarian cancer cells to escape TGF-β1 growth inhibition [73]. In prostate cancer cell line, Androgen can block TGF-β responses in prostate epithelial cells through an association of the androgen receptor with SMA and MAD-related protein (Smad)-3, which inhibits the binding of Smad3 to Smad-binding elements in TGF-β responsive promoters. This provides a possible explanation for the positive role of TGF-β in androgen-promoted prostate cancer growth [74]. An important study in 2005 indicated that dehydroepiandrosterone, a type of adrenal androgen, can directly attenuate collagen type I synthesis in vitro in cardiac fibroblasts [75]. However, the underlying mechanism has not yet been elucidated.

Interactions of androgen and angiotensin II

Androgen receptor gene knockout mice exhibit exacerbation of Ang II-induced cardiac fibrosis. Ang II stimulation enhanced cardiac TGF-β1 gene expression and SMA and MAD-related protein (Smad)-2 phosphorylation in androgen receptor gene knockout mice more than in wild-type mice [76]. The same kind of phenomenon was also observed in the aorta after Ang II stimulation. Androgen signaling plays an protective role against Ang II-induced vascular remodeling [17].

Interactions of androgen and oxidative stress

Multiple clinical and experimental evidences support the connection between oxidative stress and the severity of HF [77–79]. Oxidative stress activation was demonstrated in pressure overload LV hypertrophy in mice [80,81]. Previous studies also proved that reactive oxygen species (ROS) derived from NADPH oxidase (Nox)-2 play important roles in the development of interstitial cardiac fibrosis [82–85]. ROS stimulate cardiac fibroblasts proliferation by increasing the production of TGF-β1 [14]. ROS derived from Nox-4 are also involved in TGF-β1-induced myofibroblast differentiation in the kidney and heart [86,87]. Blocking androgen receptor signaling may induce oxidative stress was documented in some experiments. In the rat and human prostate, castration or androgen deprivation may increase the pro-oxidant properties by significantly upregulating NADPH oxidase, and also decrease the anti-oxidant ability by reducing ROS-detoxifying enzymes such as manganese superoxide dismutase, thioredoxin1 and peroxiredoxin [88–91]. Anti-oxidant therapy can alleviate oxidative stress by androgen deprivation therapy in prostate cancer [92]. The androgen receptor counteracts ROS damage to the heart induced by doxorubicin in mice [18]. Therefore, androgen signaling might play a protective role against toxicity from oxidative stress.

Summary

As shown in Figure 1, androgen acting on cardiac fibroblasts elicits anti-fibrotic effects on the heart through down-regulating the signal pathway of TGF-β, blocking the effect of Ang II, and alleviating the pro-fibrotic properties of ROS. Androgen deficiency can promote cardiac fibroblast activation via induction greater TGF-β, Ang II and oxidative stress expression. Therefore, in the initial phase of myocardial injury, androgen might disturb the normal wound healing process. These speculated mechanisms can explain the higher LV wall rupture rate at the acute phase of MI in male mice than in female mice [93–95] being reversed by androgen depletion therapy [93]. On the other hand, chronic androgen deficiency might persistently stimulate the differentiation of cardiac fibroblasts to cardiac myofibroblasts through the over-expression of TGF-β and increase myocardial fibrosis in chronic disease. These effects can explain the phenomenon discovered in clinical studies that androgen deficiency is closely related to the severity and mortality rate of chronic HF. Testosterone supplementation has been documented to be beneficial for HF patients. Testosterone level should be monitored in male HF patients and testosterone therapy can be considered in HF patients.

Figure 1. Mechanisms of the anti-fibrotic effects of androgen on the heart. Androgen suppresses the differentiation of cardiac fibroblasts to cardiac myofibroblasts through blocking the effect of angiotensin II, alleviating the pro-fibrotic properties of oxidative stress and down-regulating the signal pathway of TGF-β. MMP: matrix metalloproteinases; TIMP: tissue inhibitors of metalloproteinases.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

This study was supported by a grant from Taipei Medical University-Wan Fang Hospital Grant no. 100swf05.