In the present study we recruited cardiac patients with HFpEF and HFmrEF referred for open-heart surgery, since it has been reported in the literature that EAT accumulation was associated with adverse prognosis in these categories of patients [19].
Over the past decade, both epicardial and perivascular adipose pools have been increasingly acknowledged as modulators of progression of cardiometabolic diseases (associated with chronic inflammation and insulin resistance) and pharmacological therapeutic targets [7, 20]. Both tissues are major sources of pro-inflammatory cytokines and growth factors that exert deleterious effects on both heart and vessels by paracrine signaling via direct diffusion. In particular, EAT isolated from obese, but also from overweight patients was found to have higher expression and levels of pro-inflammatory cytokines (IL-1, IL-6, TNF-α, and IFN-γ) than the paired subcutaneous fat [21]. Of note, in the studied group, BMI was 27.7±4.9 (see Table 1) and only 3 patients presented an BMI over 30 (data not shown). We have to acknowledge as limitation of the study the fact that we did not assess the serum levels of inflammatory cytokines and markers (e.g., CRP).
Moreover, EAT accumulation and its inflammation promote an arrhythmogenic substrate via fibrotic remodeling [22]. Left atrial myocardial fibrosis was positively correlated with the level of proinflammatory and profibrotic cytokines/chemokines, IL‐6, MCP‐1, and TNF‐α, in EAT [23] and myocardial fibrosis is a known substrate for atrial fibrillation. In the present study, 2 patients had atrial fibrillation and one atrial flutter (see Table 1) albeit a causal relation cannot be affirmed.
This pilot study was purported to assess both MAO expression as novel source of oxidative stress in EAT and PVAT samples, as well as the possibility to counteract it by acute exposure to MB. The first major finding is that MAO-A is the predominant isoform expressed in human cardiovascular adipose tissue that can be targeted (at least ex vivo) with submicromolar concentrations of MB. Furthermore, ROS generation assessed in confocal microscopy and spectrophotometry was reduced when the EAT and PVAT samples were acutely incubated with MB, as proof of the antioxidant effect of the redox dye in human adipose tissue. Last but not least, in the presence of serotonin, the MAO-A substrate, ROS production was further increased in both samples, an effect partially reversed by MB.
Besides inflammation, oxidative stress is also increased in the EAT but the sources of ROS are far from being fully elucidated. Salgado-Somoza et al. reported in their pioneering study that EAT produces a higher level of ROS than subcutaneous adipose tissue in patients with CHD. They found, among others, mRNA differences for catalase, glutathione S-transferase P, and protein disulfide isomerase [24]. Since EAT expands from epicardium into the myocardium, following the adventitia of the coronary arteries without any separation from myocardium, it provides harmful signaling via paracrine or vasocrine secretion and the oxidative stress in EAT may induce an increased oxidative stress in both coronary or myocardial tissue [25]. One year later, in EAT harvested from patients with severe stable CHD, Sacks et al. showed an increase mRNAs for 7 molecules involved in oxidative stress and/or oxygen species regulation along with 17 inflammatory adipokines or proteins involved in inflammation. One of the largest increases were reported for NADPH components gp91phox and p47 phox [26].
We have previously reported an increased oxidative stress in visceral adipose tissue (VAT) harvested from obese patients subjected to elective abdominal surgery and that MAO-A was the major isoform overexpressed. The fact that this finding was a particularity of obesity was further proven by the fact that ex vivo inhibition of MAO-A with clorgyline significantly reduced oxidative stress in VAT samples isolated from obese patients and had no effect in those harvested from the non-obese group [27].
The finding that MAO-A is also the predominant isoform in human EAT and PVAT is in line with the pioneering study published by Pizzinat et al. in the late 90s. Indeed, these authors reported that both MAO-A and MAO-B were expressed in human abdominal adipose tissue with MAO-A representing 70-80% of the total enzyme activity; also, the concomitant expression of noradrenaline transporter in human white adipocytes supports their role in the clearance of peripheral catecholamines [28]. Our results confirm the fact that MAO-A is the predominant isoform in the diseased EAT (and PVAT) and its expression was mitigated by MB. We must acknowledge as a limitation of the study the fact that we did not use MAO inhibitors in this study in addition to MB to (indirectly) assess the contribution of the latter to MAO inhibition.
PVAT is the fat depot surrounding most blood vessels, which in health presents anti-inflammatory and anti-contractile properties. At variance, in cardiometabolic pathologies associated with low-grade inflammation, PVAT-derived adipocytes generate various ROS, including superoxide anion and hydrogen peroxide that might signal to the vascular wall, underlying vascular injury. Classic sources of ROS in vascular beds include NADPH oxidase, uncoupled eNOS and dysfunctional ETS at the inner mitochondrial membrane [29, 30]. Moreover, superoxide is able to generate peroxynitrite (ONOO-) in the presence of NO and H2O2 can be converted into the highly reactive hydroxyl radical (•OH) with further oxidation of lipids and DNA, thus leading to cell damage [29]. Interestingly, in healthy mouse mesenteric resistance arteries, it has been reported that PVAT acts as a reservoir for norepinephrine, preventing it from reaching the vessel and causing contraction [31].
We reported here that MAO-A is an important source of H2O2 in PVAT. We have also demonstrated that acute incubation of mesenteric arteries samples harvested form patients undergoing elective abdominal surgery with IL-6 increased MAO-A gene expression, as evidence of the fact that inflammation also potentiate the oxidative stress in the vascular wall [32]. Whether this observation can be recapitulated at the level of EAT and PVAT remains to be determined. Also we have to acknowledge the fact that the study group included more males (20) as compared to female (5); whether gender difference occur in response to MB application is worth further investigation.
The first study reporting that MB is a potent reversible inhibitor of MAO-A was published almost two decades ago by Ramsay et al. These authors reported that MB, at concentrations reported to occur after intravenous administration, completely inhibited MAO-A (and partially MAO-B), due to its action as an oxidizing substrate and an one-electron reductant [33]. This MAO inhibitor effect, also common for other MB analogues [34], has been reported to mediate, at least partially, its antidepressant effects. Of note, the central inhibition of MAO-A by MB has also been linked to serotonin toxicity which may arise only when MB was used in combination with serotonergic drugs [35].
Methylene blue (MB) is known as a mild redox agent, which has been used as an electron carrier to prevent free radicals production and enhance cellular metabolic activity because it will not excessively accumulate in mitochondria and will not compromise the oxidation state of the physiological redox centers [36]. MB can reroute electrons in the ETS directly from NADH to cytochrome c, increasing the activity of complex IV activity and promoting ATP generation, while mitigating oxidative stress and delaying cellular ageing by reversing neuroinflammation [37, 38].
The group of Adam-Vizi performed an elegant study aimed at elucidating the favorable energetic effects of MB in isolated guinea pig brain mitochondria treated with inhibitors of complex I or complex III of ETS. When the flow of electrons was compromised, MB transferred electrons to cytochrome c, increased the rate of ATP production, restored mitochondrial membrane potential, and improved the rate of calcium uptake. In rat heart mitochondria isolated from healthy and 2 months (streptozotocin-induced) diabetic rats, we have also demonstrated that addition of MB (0.1 μmol·L-1) elicited an increase in oxygen consumption of mitochondria energized with complex I and II substrates. In our hands, MB elicited a significant increase in H2O2 release in the presence of complex I substrates (glutamate and malate), but had an opposite effect in mitochondria energized with complex II substrate (succinate) [13].
More recently, the group of Mariana Rosca showed, in isolated diabetic cardiac mitochondria harvested from mice treated orally with MB that the redox agent facilitated NADH oxidation, increased NAD+, the activity of deacetylase sirtuin 3, and reduces protein lysine acetylation. Thus, by providing an alternative route for mitochondrial electron transport, MB alleviated the metabolic inflexibility in the diabetic heart [39].
MB has been extensively studied for its neuroprotective effects in animal models and patients with neurodegenerative diseases, in particular with Alzheimer disease, by targeting several molecular pathways that ultimately protect the brain mitochondria (comprehensively reviewed in ref. [40]). As an antidepressant, MB has been reported to act via various mechanisms. Accordingly, it restores mitochondrial function by acting as an alternative electron acceptor/donor, enhancing mitochondrial respiration, improving energy production and inhibiting the formation of superoxide. Also, MB has been also acknowledged as a non-selective inhibitor of NOS and modulator of the nitric oxide cyclic guanosine monophosphate (NO-cGMP) cascade, which enhances its antidepressant response, since dysfunction of the NO-cGMP cascade is involved in the neurobiology of mood, anxiety and psychosis [41].
Recently, Pluta et al. reported the successful reversal of vasoplegic shock by MB and ascribed the effect to the selective inhibition of iNOS (the inducible form of nitric oxide synthase), which prevented vasodilation in response to the pro-inflammatory cytokines [42].
Pharmacological targeting of the epicardial and perivascular adipose tissues signaling pathways will remain as potential disease modifying approach in cardiometabolic syndromes. Whether MB will finds a place in this scenario remains to be confirmed by larger clinical studies.