Hydrogen Sulfide, Adipose Tissue and Diabetes Mellitus

Introduction

H₂S, a toxic gas with a rotten egg flavour used as a chemical reagent, was first proposed by Abe and Kimura in 1996 to be an endogenous neuromodulator.¹ Considerable evidence suggests that H₂S is produced endogenously in many parts of the body and participates in various physiological and pathological functions in mammals.² Altered levels of endogenous H₂S and H₂S synthetase expression have been observed in diabetic patients and animals. The blood H₂S levels in diabetic patients were found to be significantly lower than those in age-matched normal control subjects.³ However, YusufM⁴ investigated the increased H₂S production rate in the pancreatic islets and livers of diabetic rats. Zucker diabetic fatty (ZDF) rats have significantly higher levels of pancreatic CSE expression and H₂S production when compared with Zucker fatty (ZF) or Zucker lean (ZL) rats.⁵ Increasing numbers of studies have shown that H₂S is involved in the development of diabetes,⁶ however, its impact on diabetes now appears to be contradictory. The dysfunction of AT plays an important role in the development of diabetes. We will briefly review the effect of H₂S on diabetes by discussing the influence of H₂S on the function of AT.

Hydrogen Sulfide Biosynthesis and Catabolism

H₂S is a colourless, flammable gas with a strong odour of rotten eggs. It is produced by cystathionine β-synthetase (CBS) in the mammalian brain, selectively enhances NMDA (N-methyl-D-aspartate) receptor-mediated responses and facilitates the induction of long-term potentiation in the hippocampus, which was the first proposed biological role of H₂S as a neuromodulator.¹ Several lines of evidence have suggested that H₂S is naturally synthesized by four enzymes: CBS, CSE, 3-mercaptopyruvate sulfurtransferase (3-MST) and cysteine aminotransferase (CAT). It is produced in vivo from L-cysteine (LC) by those enzymes.^6,7 Those enzymes are constitutively present and can be found widely in cells and tissues, and their expression can be induced by a number of disease states.⁸

Four enzymatic pathways for H₂S production have been described thus far. Two of them are catalysed by CBS and CSE, which are pyridoxal 5ʹ-phosphate (vitamin B6)-dependent cytoplasmic enzymes of the trans-sulfuration pathway. The third pathway for H₂S generation requires two enzymes, CAT and 3-MST. In contrast with CBS and CSE, this pathway operates primarily in the mitochondria⁶ (Figure 1). Finally, H₂S in the cerebellum and the kidney can be synthesized from D-cysteine by the synergism of D-amino acid oxidase (which is highly expressed in the peroxisomes of these organs) and 3-MST.⁹ The degradation pathways involved in H₂S consumption are mediated by ethylmalonic encephalopathy protein 1 (ETHE1), mitochondrial sulfide quinone oxidoreductase (SQR) and cysteine dioxygenase (CDO).^10,11

Figure 1 Generalised overview of H2S production in the cell. The endogenous production of H2S occurs via two main pathways-reverse transsulfuration and cysteine oxidation which take place partly inside mitochondria.

Changes in the levels of H₂S or H₂S synthetase expression may contribute to the pathogenesis of many pathophysiological processes, such as neurological systems, inflammation, apoptosis, vascular function, energy metabolism and biogenesis, obesity and ageing.⁶ The pharmacological or genetic inhibition or activation of H₂S production is protective in some models. Therefore, the application of H₂S donors and/or the augmentation of endogenous H₂S or H₂S inhibitors and/or the reduction in endogenous H₂S has attracted much attention as possible therapeutic approaches.

Hydrogen Sulfide and Diabetes

Endogenous H₂S can be synthesized by multiple systems, with the highest rates occurring in the brain, cardiovascular system, liver and kidney.¹¹ At present, studies have shown that almost every mammalian tissue cell can produce a certain amount of H₂S. Under physiological conditions, the H₂S concentration in rat brain tissue is 50–160 μmol/L, and the serum H₂S concentration is approximately 46 μmol/L.¹² Human blood contains a significant amount (10–100 μmol/L) of H₂S.¹³

Altered endogenous H₂S levels, and the expression of H₂S synthetases have been observed in diabetic animals and patients.⁶ YusufM⁴ first reported that a marked increase in H₂S synthesis from exogenous LC in the pancreas and liver is associated with streptozotocin (STZ)-induced diabetes in rat, and the liver pancreatic CBS and CSE mRNA levels were concomitantly increased, with no change in the pancreatic CSE mRNA levels in diabetic rats. However, the plasma H₂S concentrations were similar in diabetic and non-diabetic animals. Therefore, the enhanced formation of H₂S is likely to be a local tissue response to diabetes, which is not reflected in increased circulating plasma levels. Other researchers⁵ have also indicated that the expression of CSE and the production rate of H₂S in pancreatic islets were significantly higher in ZDF rats than in ZL or ZF rats. The serum H₂S levels of ZDF rats were also significantly higher than those of ZF or ZL rats. Increased pancreatic H₂S production underlies the reduced insulin release from the pancreatic β cells of ZDF rats. The activation of K⁺_ATP channels by high levels of endogenous H₂S in pancreatic β cells may be the molecular basis underlying H₂S-suppressed insulin release.⁵ It is well-known that H₂S functions as an endogenous activator of K⁺_ATP channels in β cells.¹⁴ Further research¹⁵ has found that the CSE/H₂S system plays a critical role in regulating cell functions by stimulating β cell apoptosis and inducing K⁺_ATP channel activity and that STZ-induced diabetes is largely mediated by the effects of pancreas-produced H₂S on β cell mass and K⁺_ATP channel activity. A CSE deficiency significantly delayed the development of T1DM induced by STZ.¹⁶

Interestingly, the results of other studies are diametrically opposite. Brancaleone et al¹⁷ found that the plasma H₂S value in NDR (non-obese resistant) mice was approximately 50 μmol/L, whereas NOD (non-obese diabetic) mice exhibited a progressive decrease in plasma H₂S levels that paralleled the disease severity, reaching a 50% reduction in NOD III mice and indicating that endogenous H₂S production is significantly impaired under hyperglycaemic conditions. A further study found that the blood H₂S levels of diabetic patients were significantly decreased when compared with age-matched normal control subjects, which is in accordance with their study of rats.³ Recently, Kunihiro et al¹³ investigated the plasma H₂S levels in Japanese individuals with type 2 diabetes mellitus and found that the patients showed a progressive decrease in the plasma H₂S levels (45.1±15.5 μmol/L versus 54.0±26.4 μmol/L), which is in accordance with poor glycaemic control. There was a significant correlation between the reduction in plasma H₂S levels and the level of HbA1c. They also found that the reduction in plasma H₂S levels was associated with a history of cardiovascular disease in type 2 diabetic patients.

The above study shows that H₂S may play a role in the development of diabetes. Researchers have observed different changes in H₂S levels in different diabetic patients and animals, possibly due to the different stages of diabetes in the diabetic patients studied by different researchers, the different methods of inducing diabetic animals and the use of different animal species.

Hydrogen Sulfide and Adipose Tissue

Diabetes-related obesity and general obesity are characterized by an increase in the AT mass, which can be defined as an increase in the BMI. In addition to the increase in AT mass (eg, increase in the size and number of adipocytes), histological and macroscopic changes and AT dysfunction, including hypersecretion of proinflammatory, proatherogenic and prodiabetic adipocytokines, occurs in obesity.^18–20 Feng et al²¹ first identified the endogenous CSE/H₂S pathway in AT. They detected the rate of H₂S production from visceral AT, including perirenal fat (2.93 ± 0.27 nmol/min/mg protein), epididymal fat (4.76 ± 0.92 nmol/min/mg protein) and brown fat tissue (4.65 ± 0.81 nmol/min/protein). RT-PCR results revealed that CBS and CSE mRNA were expressed in those fat tissues and illuminated the CSE-mediated primary pathway of H₂S generation in AT. Meanwhile, they found H₂S-generation rates in adipocytes and preadipocytes of 2.89±1.34 and 2.17±1.14 nmol/min/mg protein, respectively, and demonstrated that adipocytes also endogenously produce H₂S through a CSE catalytic pathway. Furthermore, they found that H₂S production and CSE protein expression in the epididymal fat pad and in perirenal fat increased in an age-dependent manner. The same year, Fang et al²² first demonstrated that CSE protein expression and endogenous H₂S production in rat PVAT (perivascular adipose tissue) were detectable and that the endogenous H₂S generated by PVAT was predominantly CSE-catalysed.

In recent years, as increasing research has focused on AT-derived H₂S and explored the pathophysiological roles of H₂S in AT,²³ we have uncovered the effects of H₂S on AT inflammation, apoptosis, adipokine secretion, glucose and lipid metabolism and vascular tension.

Hydrogen Sulfide and Inflammation of Adipose Tissue

A series of epidemiological studies have shown that circulating inflammatory markers are closely related to type 2 diabetes mellitus (T2DM) and are risk factors for the development of future T2DM.²⁴ Numerous preclinical and clinical studies now strongly support the view that obesity-induced inflammation plays an important role in the development of insulin resistance and T2DM.^25,26 Although obesity-induced inflammation is similar in many ways to inflammation observed in classical immunity, it is a low-grade form of inflammation that produces much lower levels of circulating cytokines.²⁷ It is also considered to be chronic inflammation, as it requires relatively long dietary therapy before it becomes clearly discernible in AT, which displays the most severe obesity-induced inflammation of all insulin-responsive tissues (as characterized by the increased cytokine/chemokine expression and immune cell infiltration).²⁸ Indeed, AT produces high levels of many inflammatory cytokines in obesity and is therefore considered to be the major inflammatory organ that mediates obesity-induced inflammation. Chronic inflammation and metabolic detrimental factors released by AT into the circulation are associated with some metabolic complications of obesity, such as T2DM and atherosclerosis.²⁹ AT inflammation is frequently observed in obesity and diabetes, is associated with the infiltration of macrophages, and may be caused by the death of adipocytes, the secretion of adipokines, such as TNF-α and IL-6, and adipocyte chemokines, such as monocyte chemo-attractant protein-1 (MCP-1) (Figure 2A).^30,31

The gaseous signalling molecule H₂S has been identified as an important inflammatory mediator. At present, several research groups worldwide are focused on determining the role of H₂S in inflammation. H₂S has different effects on inflammation in different models.^32–34 The observation that an H₂S donor significantly attenuates the production of cytokines triggered by the pro-inflammatory stimuli of primary macrophages³⁵ and macrophage cell lines³⁶ supports the anti-inflammatory action of H₂S in macrophages. These findings are particularly intriguing for data that point to impaired H₂S signalling in type 2 diabetes^3,17,37–39 and show a decrease in the plasma concentration of H₂S in human diabetes and obesity.⁴⁰

The overexpression of CSE and exogenous application of NaHS both increased H₂S concentrations and significantly inhibited high glucose (HG)-induced MCP-1 secretions in mature 3T3-LI adipocytes, indicating that H₂S exhibited anti-inflammatory effects in mature adipocytes.⁴¹ A further study⁴² found that obesity increased the production of H₂S while also increasing the consumption of H₂S likely due to the promotion of H₂S catabolism by radical species (the reactive oxygen and nitrogen species). In addition, H₂S depletion was observed in chronic inflammation models [AT macrophages (ATMs) from obese mice]. In obesity, store-operated Ca²⁺ entry (SOCE) increased concurrently with the consumption of H₂S, and a reduction in the concentration of cellular H₂S promotes an increase in SOCE.⁴³ The activation of the SOCE pathway during pro-inflammatory stimulation provides the Ca²⁺ signals required for the activation of nuclear factor of activated T cells (NFAT) and nuclear factor κB (NFκB), the transcription factors responsible for the expression of cytokine encoding genes.^43,44 During obesity, an increased abundance of M1-associated markers, reduced H₂S bioavailability and increased SOCE were observed in ATMs. The inhibition of SOCE or the prevention of H₂S loss from these cells also limited pro-inflammatory cytokine production. In parallel with the study by Pan,⁴¹ this research implied that H₂S plays an anti-inflammatory role in the ATMs from obese mice (Figure 2B).

Figure 2 (A) Healthy and obese adipose tissue. Obese adipose tissue is characterized by inflammation and progressive infiltration by macrophages as obesity develops. (B) The anti-inflammation effect of H2S on obese adipose tissue.

Those reports suggested that therapeutic strategies designed to maintain H₂S bioavailability may be effective in a number of inflammatory pathological contexts. However, further studies are required to specifically examine the effects of H₂S on AT inflammation in diabetes, obesity and other pathological conditions.

Hydrogen Sulfide and Insulin Resistance of Adipose Tissue

T2DM consists of two independent but related alterations: insulin resistance and dysfunction of the β-cells in the pancreas. AT is an insulin-sensitive organ that mediates glucose uptake and metabolism⁴⁵ and is an important source of methionine metabolism for the synthesis of fatty acids.^46,47 Increasing evidence has shown that excess AT accumulation alters AT secretion profiles,⁴⁸ and only a few adipokines have been shown to regulate insulin sensitivity, including TNF-α, adiponectin, interleukin-6, resistin and leptin (Figure 3A). TNF-α provides a molecular basis for insulin resistance and contributes by inhibiting the expression of genes that are essential for insulin signalling and adipocyte differentiation, including adiponectin.⁴⁹ Single-nucleotide polymorphisms (SNPs) in the promoter region of the adiponectin gene may be associated with the development of insulin resistance, obesity, and T2DM.^50,51

Figure 3 The bilateral regulation of H2S on glucose metabolism and insulin sensitivity in adipocytes.

Researchers have found H₂S-reduced insulin sensitivity in adipocytes, and CSE inhibitors might reverse this effect. A high level of glucose (20 mmol/L) inhibited the H₂S production of adipocytes in a time- and concentration-dependent manner. Under basal and insulin stimulating conditions, H₂S inhibited glucose transport into adipocytes in a concentration-dependent manner. The CSE inhibitors PAG (propargylglycine) and BCA (β-cyano-L-alanine) increased basal and insulin-stimulated adipocyte glucose uptake.²¹ Further research⁵² indicated that the upregulation of the endogenous CSE/H₂S system contributes to TNF-α induced insulin resistance in 3T3-L1 adipocytes, and exogenous H₂S causes a decrease in insulin-stimulated glucose consumption and uptake, which can be blocked by pretreatment with LY294002, a PI3K inhibitor, but not with glibenclamide, a non-selective K⁺_ATP channel inhibitor. Pinacidil, a non-selective K⁺_ATP channel opener, also has no effect on the glucose uptake of adipocytes. Therefore, H₂S might inhibit GLUT4 (glucose transporter 4) translocation through the PI3K (phosphatidylinositol 3-kinase) pathway, rather than through the K⁺_ATP channel pathway, to inhibit glucose uptake by adipocytes.²¹

However, the effect of H₂S on the insulin resistance of adipocytes is controversial. Prasenjit et al⁵³ argued that H₂S plays an important role in vitamin D-induced GLUT4 translocation to the cell surface for insulin-stimulated glucose uptake and utilization in 3T3-L1 adipocytes. Interestingly, they also indicated that LC and H₂S caused an increase in PIP3 (phosphatidylinositol-3,4,5 trisphosphate, a positive regulator of glucose metabolism), Akt (serine/threonine protein kinase) phosphorylation, and glucose utilization in HG (25 mM) treated 3T3-L1 adipocytes.⁵⁴ In addition, H₂S mediated the above effect of LC. H₂S or PIP3 increased GLUT4 activation, glucose utilization and the phosphorylation of IRS1 (insulin receptor substrate 1), Akt and PKCξ/λ (protein kinase C ξ/λ) in HG-treated cells. They provide evidence for a molecular mechanism by which H₂S can upregulate the insulin-signalling pathways by increasing PIP3 and glucose utilization by inhibiting PTEN (Phosphatase and Tensin Homolog) protein and activating the PI3K/AKT/PKCξ/λ pathway in 3T3-L1 adipocytes.⁵⁵

Moreover, Rong et al⁵⁶ found that NaHS treatment may ameliorate insulin resistance by sensitizing insulin receptor (IR)-mediated signals. The phosphorylation of IR, PI3K and Akt increased in adipocytes treated with NaHS. However, the ability of NaHS to activate PI3K was blocked by the IR inhibitor HNMPA (C₁₁H₁₁O₄P). In addition, NaHS directly activated the IRs in a cell-free system, indicating that IRs may act as a direct target molecule for NaHS and that PI3K may act as a downstream element of IR in the actions of NaHS. This hypothesis was further supported by the notion that increased NaHS-induced glucose uptake was blocked by siRNA-mediated IR knockdown or by pretreatment with the IR inhibitor HNMPA or the PI3K inhibitor LY294002. NaHS may react with some amino acid residues of IRs, leading to certain chemical modifications of the residues and to subsequent conformational changes in the IRs. Changes in IR conformation induced by NaHS may promote the trans-autophosphorylation of IRs and, thus, enhance IR activation. The stimulating effects of NaHS on the IRs may have potential clinical relevance. Fasting blood glucose levels were decreased and glucose tolerance was increased in GK rats chronically treated with NaHS at a dose of 30 μmol/kg•day. The activation of the PI3K/Akt pathway was also observed in some tissues in GK rats treated with NaHS. Increased insulin sensitivity and reduced plasma fasting glucose levels were also observed in Wistar rats treated with NaHS.⁵⁶ These data further confirm that the insulin sensitizing effects of exogenous H₂S treatment are also present in subjects without diabetes. Another study⁵⁷ found that H₂S ameliorated insulin resistance by upregulating IRS1 protein, and the CSE/H₂S system might regulate the gene transcription of glucose metabolic enzymes or transporter proteins through a nuclear receptor.

The bilateral regulation of the CSE/H₂S system in glucose metabolism and insulin sensitivity suggested that CSE/H₂S might be an energy balancer⁵⁸ (Figure 3B). Under physiological conditions, the CSE/H₂S system tended to reduce energy consumption, resulting in a slight decrease in glucose utilization, whereas under stress or inflammation, the CSE/H₂S system antagonized injury and increased glucose utilization, resulting in increasing insulin sensitivity. Clarifying the regulation of CSE/H₂S during energy metabolism may help elucidate the complex interactions among glucose, fat and sulfur-containing amino acids in physiology and diseases.

Hydrogen Sulfide and Adipocyte Lipid Metabolism

During AT expansion, adipocytes become either hyperplastic, when their number increases through adipogenesis, or hypertrophic, when their size increases through lipogenesis.⁵⁹ The increased number of adipocytes is primarily determined by the process of adipocyte differentiation, termed adipogenesis, which refers to the conversion of undifferentiated preadipocytes to mature adipocytes. In contrast, lipolysis is one of the most important processes for reducing adipose mass in which triglycerides stored in adipocytes are broken down into glycerol and fatty acids.⁶⁰ The dysregulation of AT metabolism (including lipogenesis and lipolysis) is related to obesity, insulin resistance and diabetes in human and experimental animal models.^19,61

The role of H₂S in the regulation of AT metabolism is complex. Geng et al⁵⁷ found that the H₂S precursor LC or the donor-GYY4137 (morpholin-4-ium 4 methoxyphenyl (morpholino) phosphinodithioate) suppressed lipolysis while PAG induced basal and isoproterenol-stimulated lipolysis in rat epididymal adipocytes. PAG also increased lipolysis in normal chow and high-fat diet (HFD) fed mice by increasing serum glycerol levels without affecting food consumption and then blunting fat deposition and weight gain. GYY4137 reduced lipolysis in HFD mice without increasing fat mass and body weight. Moreover, they found that the PKA-HSL/perilipin 1 pathway may be involved in the lipolytic regulation of H₂S.

Endogenous H₂S was increased after 3T3-L1 differentiation.⁶² The expression levels of the H₂S-synthesizing enzymes CSE, CBS and 3-MST were increased in a time-dependent manner during 3T3-L1 differentiation. GYY4137 and NaHS increased the expression of genes involved in adipogenesis, which include fatty acid binding protein 4 (FABP4/aP2), a key regulator of adipogenesis, and significantly increased the size and number of lipid droplets in mature adipocytes, which also compromised the ability of CL316,243 (a β3-agonist) to promote lipolysis in these cells. In contrast, aminooxyacetic acid (AOAA) and PAG had the opposite effect. Moreover, the inhibition of H₂S production by CBS/CSE inhibitors seriously prevented the induction of adipogenic marker genes,⁶³ PPARγ2 (peroxisome proliferator-activated receptor gamma 2) and its target gene perlipin1, Fabp4 and stearoyl-CoA desaturase-1 (Scd1). In addition, Cai et al⁶⁴ confirmed the upregulation of an endogenous CSE/H₂S system during adipocyte differentiation. Increased H₂S promoted adipogenesis and inhibited lipolysis, resulting in the storage of triglycerides in lipid droplets. More interestingly, using H₂S instead of 3-isobutyl-1-methylxanthine (IBMX) in the differentiation cocktail also upregulated adipogenesis markers and slightly increased the intracellular triglyceride levels. Based on the IBMX mechanism of action, H₂S might directly promote PPARγ expression in part by inhibiting PDE (phosphodiesterase) activity. Therefore, H₂S sulfhydrates PPARγ, enhancing its activity to promote adipogenesis, increase insulin sensitivity and inhibit basal lipolysis, causing a decrease in the circulating free fatty acid level and thereby ameliorating target injury in obesity.⁶⁵

In summary, H₂S increases the expression of adipogenesis-related proteins, increases lipid synthesis, promotes the differentiation of preadipocytes into mature adipocytes and reduces the lipolysis of adipocytes (Figure 4). H₂S synthase inhibitors or siRNAs have the opposite effect.

Figure 4 Generalised overview of the effects of H2S on adipocyte lipolysis and adipogenesis.

Hydrogen Sulfide Is a Vascular Tone Modifier

The contractile force of aortas induced by norepinephrine⁶⁶ and other vasoconstrictors was mitigated by PVAT in a paracrine manner.^67,68 The anti-contractile effect of PVAT was also observed for both large and small vessels in mice, pigs and humans.^69,70 Bioassay experiments showed that the adipocyte derived relaxing factor (ADRF), produced by PVAT, activates the K⁺ channels in vascular smooth muscle cells, exerting its anti-contractile effect on systemic peripheral arteries.⁶⁸

Fang et al²² first demonstrated that the CSE protein was expressed in both the PVAT and aorta and in the adipocytes of PVAT and vascular smooth muscle cells (VSMCs) of aorta media. The rate of endogenous H₂S generation was similar to that of aortic tissues and was primarily catalysed by CSE. The H₂S donor NaHS produces dose-dependent vasorelaxation effects, which are blocked by XE991 (the KCNQ inhibitor). Therefore, H₂S may be a candidate or modulator of ADRF, which activates the K_v channels encoded by KCNQ genes to achieve the paracrine control of vascular tone by PVAT. It is unclear how the PVAT derived H₂S activates KCNQ channels, where it activates the channel directly or produces sulfhydryl radicals [HS(*)/S(^*-)], which, coupled to the formation of superoxide radical anions,⁷¹ are powerful KCNQ channel activators.⁷² Further research confirmed that H₂S generated by PVAT was a releasable vascular relaxation factor, acting in a paracrine manner by opening the K⁺_ATP channel in a NO-, endothelium- and Ca²⁺ channel-independent manner. Inhibitors of CSE, such as BCA (5 mmol/l) and PAG (10 mmol/l), inhibit the anti-contractile effects of PVAT.⁷³

Moreover, the anti-contractile effect of PVAT on rat aortic rings has been demonstrated to be blunted not only by XE991 or 4-AP (4-aminopyridine, the K_v channel inhibitor) but also by glibenclamide. In addition, the dilation of serotonin-preconstricted aortic rings without PVAT induced by NaHS was also abolished by glibenclamide, 4-AP or XE991.⁷⁴ It is unclear whether this phenomenon results from the nonspecific effects of these inhibitors on different types of K⁺ channels or from the fact that both K⁺_ATP and KCNQ channels are necessary for vasodilation induced by PVAT-derived H₂S. Interestingly, the expression levels of CSE in PVAT surrounding mouse aortas is much lower than that in rats, and PVAT also has an anti-contractile effect on mouse aortic rings; this effect is not mediated by H₂S, but it is still inhibited by XE991. These data indicate that H₂S is the ADRF in rat but not mouse aorta, but the H₂S-independent anti-contractile effects of PVAT on mouse aorta are also mediated by KCNQ channels.⁷⁴

The contractions of arteries induced by 5-HT (5-hydroxytryptamine) in diabetic and healthy rats with intact PVAT were significantly different from those in arteries without PVAT. At the highest studied concentrations of 5-HT (10⁻⁵ mol/L), the arteries of diabetic rats with PVAT had significantly stronger contractions than those without PVAT. The different regulatory roles of PVAT in diabetic and control rats could be explained by the decreased levels of H₂S caused by diabetes.⁷⁵

Almost all of the blood vessels are surrounded by a different amount of PVAT. PVAT may have a paracrine function in the regulation of arterial tone, vascular reactivity, and more.⁷⁶ This paracrine adipose-vascular coupling is achieved by the production and function of various ADRFs, which might be a volatile, gaseous mediator, namely, H₂S.⁷⁵ There is evidence that the opening of myocyte K⁺ channels plays a critical role in the paracrine regulation of arterial tone by H₂S. KCNQ (Kv7) channels could represent, at least in part, the subtypes of the voltage-dependent K⁺ (Kv) channels involved.⁷⁷ Additional ion channels, such as Ca²⁺-activated K⁺ (K⁺-Ca²⁺) channels, and cellular mechanisms appear to be involved in the vasoactive effects.⁷⁸ Therefore, PVAT-H₂S-smooth muscle crosstalk in the artery wall (Figure 5) may constitute a therapeutic approach against the harmful effects of diabetes and obesity in different vascular beds.

Figure 5 The mechanisms of the vasorelaxation effect of adipocytes-derived and exogenous H2S.

Hydrogen Sulfide and Adipokines

Adipocytes and AT produce a wide range of hormones and cytokines involved in glucose metabolism (eg, adiponectin, resistin), lipid metabolism (eg, cholesteryl ester transfer protein, CETP; stearoyl-CoA desaturase-1, adipose triglyceride lipase, hormone-sensitive lipase), inflammation [eg, interleukin (IL)-1β, IL-6, IL-8, IL-10, tumour growth factor (TGF)-β, TNF-α, osteopontin (OPN)], the acute-phase and immune response (eg, serum amyloid A, adipsin, PAI-1), blood pressure (eg, angiotensinogen, angiotensin II) and feeding behaviour (eg, leptin), thus affecting the metabolism and function of many organs and tissues, including muscle, liver, vascular and brain tissues^79–84 (Table 1). Approximately 20% of all genes in subcutaneous adipose tissue (SAT) and approximately 30% of genes in visceral adipose tissue (VAT) encode adipokine secretion.⁸⁵ AT also expresses many of the receptors for most of these factors.⁸⁶

Table 1 Multiple Products Called Adipokines Involving Different Biological Processes Expressed and Secreted by White Adipose Tissue May Act at Both the Systemic (Endocrine) and Local (Autocrine and/Or Paracrine) Levels

Adiponectin is an anti-diabetic and anti-atherogenic adipokine. Although the plasma adiponectin levels are reduced in obesity, other plasma adipocytokine levels increase with increases in AT and adipocyte volume.^87,88 H₂S is a recently identified endogenous gasotransmitter. Pan et al⁴¹ indicated that the overexpression of CSE and exogenously supplied NaHS, which both increase the concentration of H₂S could significantly increase the adiponectin secretion that had been depressed by HG (25 mM) in mature 3T3-LI adipocytes, whereas reduced H₂S production may contribute to deregulated adipokine secretion under HG conditions. H₂S may exert anti-diabetic effects by increasing adiponectin levels.

Hydrogen Sulfide and Adipocytes Apoptosis

In several disease states, including autoimmune lipodystrophy,⁸⁹ tumour cachexia,⁹⁰ and HIV patients under highly active anti-retroviral therapy,⁹¹ the apoptosis of adipocytes is accompanied by a loss of AT. In obese individuals, AT is poorly oxygenated,⁹² which may lead to localized hypoxia, promote free fatty acid (FFA) release and inhibit glucose uptake in adipocytes by inhibiting the insulin signalling pathway.⁹³ Hypoxia is associated with ER (endoplasmic reticulum) stress in 3T3-L1 adipocytes, as demonstrated by elevated levels of GRP78 (glucose-regulated protein, 78 kDa) and CHOP (C/EBP homologous protein).⁹⁴ Excessive ER stress induces apoptosis. Adipocyte apoptosis is still poorly understood, as adipocytes are known to be resistant to apoptosis. Studies have reported that weight loss in obese models could be attributed to apoptosis in adipocytes.^95–97 Adipokines such as leptin and TNF-α mediate apoptosis through caspase-dependent pathways and regulate body fat mass.^98,99 In addition, a variety of natural compounds, including xanthohumol, isoxanthohumol and ajoene, have been reported to induce apoptosis in adipocytes.^100,101 Interestingly, these compounds have been independently studied for their anti-obesity effects,^102,103 which means that the anti-obesity effects of some natural compounds may be achieved by inducing the brown fat-like phenotype and apoptosis of white adipocytes. Controlling apoptosis in mature white AT appears to be an ideal strategy for combating obesity and related metabolic syndromes.¹⁰⁴

H₂S has become an effective cytoprotective mediator in various models of tissue and cellular injuries.^105,106 After the ischaemic period, mature adipocytes were treated with NaHS (0, 0.1, 1, 10, 100, and 1000 μmol/L as an H₂S donor), and the late apoptosis level of mature adipocytes decreased in every dose group.¹⁰⁷ In addition, the gene expression level of caspase-3 decreased while that of Bcl-2 increased. These results imply that H₂S has a protective effect on mature adipocytes under ischaemic conditions that is mediated by elevating anti-apoptotic gene expression (Figure 6).

Figure 6 The mechanisms of adipocytes apoptosis in obese adipose tissue and the anti-apoptosis effect of H2S.

The current knowledge regarding the effects of H₂S on adipocyte apoptosis is inadequate, and it is necessary to explore it further and elucidate the roles of hypoxia, ER stress and pro-apoptotic and anti-apoptotic molecules in this process. H₂S may become a new anti-obesity strategy.

The H₂S/CSE System as a Target for Pharmacotherapy

The anti-inflammatory and insulin-sensitizing properties, the protective effects on mature adipocytes, and the beneficial effects on vascular tone of H₂S suggest that increasing H₂S signaling could be a potential new therapeutic strategy for the treatment of diabetes. The application of H₂S donors and the augmentation of endogenous H₂S has attracted the most attention as possible therapeutic approaches for diabetes, even if reducing H₂S could be protective in some instances. However, the available CBS and CSE inhibitors have limited membrane permeability and are not very specific. In addition, there is no 3-MST inhibitor available. CBS, CSE or 3-MST knockout mice exhibit many abnormalities, which precludes considering such an approach in humans. Therefore, in the subsequent section, we will focus on only the approaches aimed at increase H₂S signaling.

In most studies on H₂S, inorganic sulfide salts such as sodium hydrosulfide (NaHS) and sodium sulfide (Na₂S) were used, as they increase the H₂S concentration rapidly and in the short term.¹⁰⁸ NaHS may activate K⁺_ATP channels, protein kinase B/Akt or Nrf2-induced signalling¹⁰⁹ to exert its protective effects. To overcome the limitations of sulfide salts, several organic, slow-releasing H₂S compounds have been synthesized, with GYY4137 being the most popular; however, they may possess H₂S-independent activities mediated by the parent compound.¹¹⁰ LC is widely used in experimental studies to augment endogenous H₂S production. However, LC is not a good H₂S precursor because it is metabolized in many other pathways. In recent years, a new group of H₂S donors has been developed. This group consists of a traditional NSAID that has been structurally modified to release H₂S (S-NSAID). S-NSAIDs possess anti-inflammatory effects and are potentially protective for digestive and circulatory systems.¹⁰⁸ Garlic, originally grown in central Asia, has many medicinal properties based on its antibacterial, antifungal, and antiviral activities and usually acts through the cardiovascular system. Garlic contains approximately 2,000 active compounds, including various sulfur-containing chemicals that may release H₂S. It reduces blood pressure, suppresses blood platelet aggregation and has anti-atherosclerotic properties.¹¹¹ Moreover, garlic lowers cholesterol. Nevertheless, the development of H₂S donors is critical for understanding the biological functions of H₂S. There will be more interesting work arising in this field.

Conclusions and Future Perspectives

The altered expression of H₂S-synthesizing enzymes, as well as endogenous H₂S levels, were observed in diabetic and obese animals. H₂S is synthesized in AT and participates in the regulation of adipose tissue metabolism and function (Table 2). Although increasing research been performed in this field, some effects of H₂S on AT, such as its role in the regulation of insulin sensitivity, are controversial, and its roles in adipokine secretion and adipocyte apoptosis have been incompletely demonstrated. Moreover, the mechanisms through which H₂S acts on AT metabolism, such as glycolipid metabolism, have not been fully clarified; therefore, much work remains to be performed. Because of these controversies, the notion of using H₂S donors or enhancing H₂S signaling to alter AT dysfunction in common metabolic pathologies should be treated with caution.

Table 2 Effects of H2S on Adipose Tissue/Adipocyte Pathophysiological Functions