The Role of Hydrogen Sulfide in Renal System

Hydrogen sulfide has gained recognition as the third gaseous signaling molecule after nitric oxide and carbon monoxide. This review surveys the emerging role of H2S in mammalian renal system, with emphasis on both renal physiology and diseases. H2S is produced redundantly by four pathways in kidney, indicating the abundance of this gaseous molecule in the organ. In physiological conditions, H2S was found to regulate the excretory function of the kidney possibly by the inhibitory effect on sodium transporters on renal tubular cells. Likewise, it also influences the release of renin from juxtaglomerular cells and thereby modulates blood pressure. A possible role of H2S as an oxygen sensor has also been discussed, especially at renal medulla. Alternation of H2S level has been implicated in various pathological conditions such as renal ischemia/reperfusion, obstructive nephropathy, diabetic nephropathy, and hypertensive nephropathy. Moreover, H2S donors exhibit broad beneficial effects in renal diseases although a few conflicts need to be resolved. Further research reveals that multiple mechanisms are underlying the protective effects of H2S, including anti-inflammation, anti-oxidation, and anti-apoptosis. In the review, several research directions are also proposed including the role of mitochondrial H2S in renal diseases, H2S delivery to kidney by targeting D-amino acid oxidase/3-mercaptopyruvate sulfurtransferase (DAO/3-MST) pathway, effect of drug-like H2S donors in kidney diseases and understanding the molecular mechanism of H2S. The completion of the studies in these directions will not only improves our understanding of renal H2S functions but may also be critical to translate H2S to be a new therapy for renal diseases.

However, the image of H 2 S has been largely expanded since the revelation of H 2 S as an endogenous neuronal modulator by Kimura's group in Abe and Kimura (1996). Thereafter, the physiological significance of H 2 S has been extensively studied especially in central nervous system (Zhang and Bian, 2014) and cardiovascular system . Emerging evidence has suggested that H 2 S also actively regulates renal function and is implicated in numerous kidney diseases in recent years. Here in this review, recent studies regarded the role of H 2 S in both kidney physiology and diseases will be discussed.

PHYSICAL AND CHEMICAL PROPERTIES OF H 2 S
Hydrogen sulfide exists as a colorless gas with a strong rotten egg smell at room temperature and ambient pressure. The human nose can detect a concentration of 400-fold lower than its toxic level (Wang, 2002), whereas, long term exposure can cause desensitization of olfactory nerves to H 2 S . Distinct from the other gaseous transmitters like NO and CO, H 2 S is a weak acid and hence able to readily dissolve in water. Based on its PKa, it is estimated that there will be 14% H 2 S gas, 86% HS − and a trace of S 2− in physiological condition (pH 7.4, 37 • C; Li et al., 2009). Moreover, H 2 S gas is highly lipophilic which allows it freely to penetrate into the cell membrane of all types and become biologically active (Mathai et al., 2009).

H 2 S GENERATION IN THE KIDNEY
Three traditional H 2 S synthesizing pathways have been identified in mammalians including CSE (EC 4.4.1.1), CBS (EC 4.2.1.22), and 3-MST (EC 2.8.1.2) coupled with CAT (EC 2.6.1.3) pathways. The mechanisms underlying these traditional pathways can be found in our previous review in detail . In short, CSE firstly dimerizes two L-cysteine to L-cystine followed by transforming it into pyruvate, NH 3 and thiocysteine. The resulted thiocystein is then used as a substrate by CSE to react with other thiols to generate H 2 S (Stipanuk and King, 1982). CBS catalyzes the reaction between L-cysteine and homocystenin into cystathinine and H 2 S (Szabo, 2007). However, 3-MST is unable to directly use L-cysteine as a substrate as its counterpart does. L-cysteine has to be transformed into 3-MP by CAT which is then catalyzed by 3-MST into pyruvate and H 2 S (Shibuya et al., 2009). It is worth mentioning that both CSE and CBS require pyridoxal 5 -phosphate as a cofactor to synthesize H 2 S, while 3-MST is dependent on zinc ). In addition, CSE and CBS mainly locate in cytosol yet they can translocate into mitochondria in some oxidative conditions (Fu et al., 2012), whereas 3-MST resides and generates H 2 S in mitochondria (Kimura, 2011).
A fourth H 2 S generation pathway namely DAO/3-MST pathway was discovered recently by Kimura's group (Shibuya et al., 2013). In the study, they showed that kidney lysate can produce 60 times more H 2 S when using D-cysteine as substrate comparing with L-cysteine. Further, the underlying mechanism was studied. Specifically, D-cysteine is transformed into 3-MP by peroxisome located DAO. Due to metabolite exchanges between peroxisome and mitochondria, 3-MP is imported into mitochondria and catalyzed into H 2 S by 3-MST. Since DAO is only located in brain and kidney, this H 2 S generating pathway is believed to exclusively exist in brain and kidney.
Hydrogen sulfide generation is abundant in kidney given the presence of all the above mentioned pathways in this organ (Figure 1). Currently, it is believed that CSE and CBS are the dominated enzymes for H 2 S generation in kidney. The presence of these two enzymes was firstly demonstrated in 1980s by using their inhibitors (Stipanuk and Beck, 1982). Later on, House et al. (1997) suggested that both enzymes were mainly located on renal proximal tubules within the renal cortex by comparison with marker enzymes of known location. This finding was supported by several other studies using different methods (Ishii et al., 2004;Li et al., 2006;Tripatara et al., 2009). However, inconsistent results have been reported regarding the existence of these two enzymes in glomerulus which needs to be resolved (Aminzadeh and Vaziri, 2012;Bos et al., 2013;Yamamoto et al., 2013). In addition, definitive evidence has suggested the presence of 3-MST in kidney (Aminzadeh and Vaziri, 2012;Shibuya et al., 2013;Kimura, 2014;Pan et al., 2015), however, the significance of 3-MST mediated H 2 S generating pathway has not been well-acknowledged in both kidney physiology and diseases due to limited reports. Nevertheless, the revelation of the unique DAO/3-MST pathway in kidney and brain may imply a significant role of 3-MST mediated H 2 S generation in these organs. This will be an interesting area open to explore in the next years.

EFFECT OF H 2 S ON RENAL PHYSIOLOGY H 2 S Effect on Renal Excretory Function
The necessity of H 2 S producing enzymes have long been recognized in the kidney due to their critical effect on homocysteine metabolism (Stipanuk, 2004), however, the effect of H 2 S itself on renal function was not studied until recently. Intra-renal infusion of H 2 S donor NaHS is able to increase GFR, UNa·V and potassium (Uk·V) excretion (Xia et al., 2009;Ge et al., 2014). Moreover, the effect is closely mimicked by the infusion of L-cysteine, an H 2 S generating substrate (Xia et al., 2009). In addition, inhibition of endogenous H 2 S production by AOAA (CBS inhibitor) plus PAG (CSE inhibitor) leads to the decrease of GFR, UNa·V and Uk·V, suggesting that H 2 S regulates renal function in physiological conditions. However, either AOAA or PAG alone fails to produce any effect on renal function implicating a compensatory effect between CBS and CSE on renal regulation which has been confirmed by another study (Roy et al., 2012). Hypothesis concerning the effects of H 2 S on sodium transporters was formed and tested thereafter. The results showed that H 2 S significantly inhibited the activity of NKCC and NKA which may account for its effect on renal function (Figure 2A). Recently, the mechanism of the inhibitory effect of H 2 S on NKA was studied by Zhu's group (Ge et al., 2014). In FIGURE 1 | Endogenous synthesis of H 2 S by four pathways in renal system. (1) CBS mediated H 2 S synthesis by using L-cysteine as the substrate in cytosol; (2) CSE mediated H 2 S synthesis with L-cysteine as the substrate in cytosol; (3) CAT transforms L-cysteine into 3MP which is further catalyzed by 3-MST into H 2 S in mitochondria; (4) Peroxisome resident DAO catalyzes D-cysteine into 3MP which is then exchanged into mitochondria and utilized by 3-MST for the production of H 2 S in mitochondria. CBS, cystathionine β-synthase; CSE, cystathionine γ-lyase; 3-MST, 3-mercaptopyruvate sulfurtransferase; CAT, cysteine aminotransferase; DAO, D-amino acid oxidase; 3MP, 3-mercaptopyruvate.
their study, they showed that NaHS promoted NKA endocytosis by directly activating epidermal growth factor receptor (EGFR) in renal tubular epithelia cells. Moreover, EGFR cys797 mutation fully abolished the effect of H 2 S suggesting a direct interaction between H 2 S and this cysteine residue. Taken together, both endogenous and exogenous H 2 S are able to increase GFR, UNa·V, and Uk·V excretion probably through the inhibitory effect on sodium transporters like NKCC and NKA.
Hydrogen sulfide increases Cl − /HCO 3 − exchanger activity in various tissues like aortic tissues (Liu and Bian, 2010) and vascular smooth muscle cells (Lee et al., 2007). However, the effect of H 2 S on the exchanger in kidney is still unknown. It is likely that H 2 S can also enhance the activity of Cl − /HCO 3 − exchanger in renal system. Considering the critical role of Cl − /HCO 3 − exchanger in regulating the excretion of ions and homeostatic maintenance of physiological pH, it will be of great value to reveal the effect of H 2 S on Cl − /HCO 3 − exchanger activity and the subsequent consequence in kidney.

H 2 S Effect on Renin Release
The RAS is a renovascular hormone system involved in the regulation of plasma sodium concentration and blood pressure (BP). Renin release from JG cells determines RAS activity and the process is known to be regulated by intracellular cAMP (Peters et al., 1993;Gambaryan et al., 1998;Schweda et al., 2007). Moreover, H 2 S was reported to downregulate cAMP level in several cell types (Lim et al., 2008;Yong et al., 2008), implying that H 2 S may modulate renin release. This was demonstrated by our group in Lu et al. (2010). It was found that NaHS inhibited the upregulation of renin mRNA and protein level in a model of renovascular hypertension accompanying with a reduction of intracellular cAMP level. In normoxic condition, H 2 S is metabolized into sulfates in the presence of O 2 ; in hypoxic condition, the shortage of O 2 leads to the accumulation of H 2 S which helps to restore O 2 level by enhancing blood flow and suppressing tubular transport activity in kidney. NKCC, Na + /K + /2Cl − cotransporter; NKA, Na + /K + ATPase; GFR, glomerular filtration rate; UNa·V, urinary sodium; Uk·V, urinary potassium; β1, β1-adrenoceptor; ROS, reactive oxygen species; ATP, adenosine triphosphate; cAMP, cyclic adenosine monophosphate; JG cell, juxtaglomerular cell. This is supported by another study  showing that H 2 S regulates renin degranulation in As4.1 and rat reninrich kidney cells stimulated by isoproterenol, forskolin or 3isobutyl-1-methylxanthine. Further study demonstrated that H 2 S significantly suppressed the stimulated AC activity in these cells. Overexpression of CSE also attenuates isoproterenol-induced renin release suggesting that endogenous H 2 S may also involve in the process. However, the mechanism underlying the inhibitory effect of H 2 S on AC remains to be determined, such as the identification of AC isoform(s) accounting for H 2 S effect and the molecular interaction between H 2 S and the AC isoform(s). Besides AC, ROS was recently reported to be a target of H 2 S for its effect on renin reduction in DN suggesting participation of multiple mechanisms in the process (Xue et al., 2013). The effect of H 2 S on renin activity in normal rats has also been investigated (Lu et al., 2010). The results showed that neither NaHS administration nor inhibition of endogenous H 2 S influenced renin activity implying that H 2 S may only modulate renal activity when RAS is overactivated. The effect of H 2 S on renin activity has been illustrated in Figure 2B.

H 2 S as an O 2 Sensor in the Kidney
Essentially all H 2 S generation is independent of O 2 ; however, the metabolism of H 2 S is a process highly relying on O 2 (Olson, 2015). Accumulating evidence suggests that H 2 S is an O 2 sensor in kidney, especially in medulla ( Figure 2C). The availability of O 2 in renal medulla is lower compared with that in renal cortex resulting in a higher abundance of H 2 S in this region (Koning et al., 2015). Provided mitochondria can use H 2 S as an electron donor for ATP production (Fu et al., 2012;Teng et al., 2013), it will be interesting to hypothesize that H 2 S might be a direct source of energy in renal medulla. During hypoxia, O 2 reduction leads to further accumulation of H 2 S which help to recover O 2 supply by increasing medullary blood flow and inhibition of tubular transport (Beltowski, 2010). In addition, CBS and CSE can translocate into mitochondria and stimulate H 2 S production under hypoxic circumstances (Fu et al., 2012;Teng et al., 2013). The mitochondria derived H 2 S may directly participate in ATP production. Currently, emerging physiological evidence for H 2 S mediated O 2 sensing has also been suggested in various O 2 sensing tissues including cardiovascular system (Olson et al., 2006;Olson and Whitfield, 2010), respiratory system , gastrointestinal tract (Dombkowski et al., 2011)

H 2 S IN ACUTE KIDNEY INJURY
Acute kidney injury (formerly known as acute renal failure) is defined as a syndrome characterized by rapid loss of the kidney's excretory function. It is the clinical manifestation of several disorders that affect the kidney acutely (Bellomo et al., 2012). Here, H 2 S effects in three types of acute kidney injury will be discussed namely renal IRI, obstructive nephropathy, and cisplatin nephrotoxicity.

H 2 S in Renal Ischemia/Reperfusion Injury
Renal IRI is a major cause of acute kidney injury. The pathophysiological mechanism underlying renal IRI is very Frontiers in Pharmacology | www.frontiersin.org complex containing ATP depletion, calcium overload, ROS generation, apoptotic and inflammatory responses et al (Eltzschig and Eckle, 2011). The engagement of endogenous H 2 S in renal IRI has been thoroughly demonstrated in various studies. Specifically, both mRNA and protein levels of CSE and CBS are apparently reduced upon IRI along with the reduction of H 2 S level in kidney and plasma (Xu et al., 2009;Han et al., 2015), although mechanisms underlying IRI caused CSE and CBS reduction are still not revealed. In addition, inhibition of either CSE or CBS by their pharmacological inhibitors severely aggravates renal damage (Tripatara et al., 2008;Han et al., 2015) indicating that the ischemic renal injury might, at least in part, results from the impaired endogenous production of H 2 S. The implication is supported by a recent finding that CSE-deficiency associates with increased renal damage and mortality after renal IRI which might be due to the enhanced production of ROS (Bos et al., 2013). Subsequently, the effect of exogenous H 2 S was extensively studied in various renal IRI scenarios (Tripatara et al., 2008(Tripatara et al., , 2009Bos et al., 2009Bos et al., , 2013Xu et al., 2009;Simon et al., 2011;Hunter et al., 2012;Zhu et al., 2012;Azizi et al., 2015;Han et al., 2015;Ahmad et al., 2016) which have been summarized in Table 1. In most studies, an H 2 S donor, NaHS, was employed and exerted protective effects likely through anti-inflammatory, antiapoptotic, and anti-oxidative responses. Comparing with new generation synthetic H 2 S donors like GYY4137, NaHS is less physiologically accurate H2S producer (Li et al., 2008;Yu et al., 2010). Recently, GYY4137 was shown to attenuate heart damage by inhibiting activation of NF-κB and MAPK signaling in a rat model of myocardial IRI (Meng et al., 2015a,b). Thus, studies are warranted to study whether slow H 2 S donors like GYY4137 can protect kidney form IRI. Besides, it is worth noting that AP39, a mitochondrially targeted donor of H 2 S, was recently found to inhibit intracellular ROS formation caused by glucose oxidase and protect kidney from IRI caused damage in rats (Ahmad et al., 2016). The study implies the importance of mitochondria H 2 S in the pathology of renal IRI which needs to be determined in the future.

H 2 S in Obstructive Nephropathy
Obstructive nephropathy is a type of renal injury caused by obstruction of the genitourinary tract. Renal fibrosis after ureteral obstruction is implicated in the development of obstructive nephropathy (Boor et al., 2010). Hu's group showed that ureteral obstruction impaired endogenous production of H 2 S by reducing the expression level of CBS (Song et al., 2014). Renal fibrosis is attenuated when exogenous H 2 S is administered suggesting an inhibitory effect of H 2 S on renal fibrosis (Song et al., 2014). In cultured kidney fibroblast, NaHS is able to inhibit cell proliferation and block the differentiation into myofibroblasts by suppressing transforming growth-β1-Smad and MAPK signaling pathways (Song et al., 2014). Furthermore, administration of NaHS also prevents the disruption of renal function caused by ureteral obstruction (Jiang et al., 2014;Song et al., 2014;Dursun et al., 2015). A recent study from Sener's group (Lin et al., 2016) showed that H 2 S slow releasing donor GYY4137 mitigated cortical loss, inflammatory damage and tubulointerstitial fibrosis in a rat model of obstructive nephropathy. Taken together, these results suggest a potential use of H 2 S donor as a rescue in obstructive nephropathy.

H 2 S in Cisplatin Nephrotoxicity
Cisplatin is a major therapeutic drug for solid tumors, but causes severe nephrotoxicity. Over 30% of patients receiving high dose cisplatin develop renal dysfunction (Pabla and Dong, 2008). However, effective treatment of cisplatin induced renal failure is still lacking. Extensive research revealed that oxidative stress and inflammatory response are the major driving forces for cisplatin induced renal toxicity (Pabla and Dong, 2008;Peres and da Cunha, 2013). Given the well-known inhibitory effects of H 2 S on oxidative stress and inflammation (Łowicka and Bełtowski, 2006), it is reasonable to hypothesize that H 2 S is protective against cisplatin nephrotoxicity. However, the role of H 2 S is rather controversial due to conflicting data in this field. Coimbra and others (Della Coletta Francescato et al., 2011) firstly showed that cisplatin upregulated the expression level of CSE after 3 days upon cisplatin treatment in a rat model. When PAG was administered with cisplatin, it was found that PAG abolished the upregulation of CSE and rescued cisplatin caused renal damage by suppressing inflammation and apoptosis (Della Coletta Francescato et al., 2011). In contrast to this study, administration of NaHS ameliorates the kidney dysfunction and damage in cisplatin treated rat (Ahangarpour et al., 2014). Liu et al. (2016) recently showed that both CSE and CBS levels were severely decreased upon cisplatin treatment in mouse after 3 days. However, pretreatment with H 2 S slow releasing donor GYY4137 aggravates cisplatin induced renal damage by increasing inflammatory response . However, several defects on the use of GYY4137 need to be pointed out in the study. As they used a rather low dose of GYY4137 (21 mg/kg), it is possible that H 2 S might be not generated sufficiently (Li et al., 2008;Yu et al., 2010;Meng et al., 2015a,b;Lin et al., 2016). In addition, to rule out the possibility that the chemical backbone of GYY4137 molecule has aggravated cisplatin nephrotoxicity, ZYJ1122 (Lee et al., 2011), a structure analog of GYY4127, should be included for the study. Despite the promising protective effect of H 2 S, however, it is still far from the conclusion that H 2 S is protective in cisplatin nephrotoxicity. In future, more studies are still needed to further investigate (1) the change of CSE and CBS expression level in a time course dependant manner; (2) the effect of endogenous H 2 S by using genetic mouse rather than nonspecific CBS/CSE inhibitors; (3) the effect of exogenous H 2 S by using different H 2 S (NaHS, GYY4137, AP39) donors in parallel.

H 2 S IN CHRONIC KIDNEY DISEASE
Chronic kidney disease is a general term for heterogeneous disorders affecting kidney structure and function. In western countries, it is generally associated with old age, diabetes, hypertension, obesity, and cardiovascular disease (Levey and Coresh, 2012). Diabetic nephropathy and hypertensive nephropathy are considered as presumed pathological entities. The role of H 2 S in these two types of CKD will be reviewed below.

H 2 S in Diabetic Nephropathy
Diabetic nephropathy is the number one leading cause of CKD in western countries. Morphologically, DN is characterized by hypertrophy induced kidney growth and excessive accumulation of extracellular matrix proteins, eventually proceeding to fibrosis of glomerular and tubulointerstitial compartments (Cooper, 1998;Dronavalli et al., 2008). Current evidence suggests an active role of H 2 S in the pathogenesis of DN. Plasma H 2 S level in DN patients is significantly lower than that in non-DN patients undergoing chronic hemodialysis . High urinary sulfate concentration, a reflection of high plasma H 2 S level, is associated with reduced risk of renal disease progression in type 2 diabetes (van den Born et al., 2016) and slower decline in Cr 51 -EDTA-assessed GFR in DN patients (Andresdottir et al., 2013). Recent data suggested that the renal expression of H 2 S producing enzyme CBS and CSE is down-regulated in pancreatic CaMTg diabetic mice (Yamamoto et al., 2013), C57BL/KsJ lepr −/− db/db mice (Lee et al., 2012), STZ-diabetic rats (Yuan et al., 2011), and Akita diabetic mice (Kundu et al., 2013). Inhibition of CSE with PAG mimics high glucose-induced glomerular podocyte injury (Liu et al., 2015), implying a contributive role of endogenous H 2 S in DN. The mechanisms underlying CBS and CSE reduction have been also studied. Kundu et al. (2013) showed that MMP-9 was upregulated in Akita diabetic mice along with the reduction of CBS and CSE. When MMP-9 is knocked out, the expression of the two H 2 S producing enzyme namely CSE and CBS shows a trend toward baseline despite hyperglycemia (Kundu et al., 2013) suggesting that MMP-9 regulates CBS and CSE expression in DN ( Figure 3A). Meanwhile, exogenous H 2 S has been proved to be effective in in vitro and in vivo DN models. High glucose caused cell proliferation and collagen formation are attenuated by NaHS in cultured messangial cells (Yuan et al., 2011) and tubular cells (Safar and Abdelsalam, 2015). Moreover, NaHS alone (Yuan et al., 2011;Ahmad et al., 2012;Lee et al., 2012;Kundu et al., 2013;Xue et al., 2013;Zhou et al., 2014;Liu et al., 2015;Safar and Abdelsalam, 2015;Qian et al., 2016) or with losartan (Kaur et al., 2015) ameliorates renal dysfunction and fibrosis formation in various DN related animal models. At least three mechanisms ( Figure 3B) are implicated in H 2 S mediated protective effect in DN: (1) Inhibition of ROS formation by activating Nrf2 pathway. Hyperglycemia induces intracellular ROS which upregulates the expression of MMP-9 (Kundu et al., 2013). MMP-9 plays a major role in diabetic renovascular remodeling. In a STZ induced diabetic rat model, H 2 S was found to reduce high glucose induced oxidative stress by activating the Nrf2 antioxidant pathway and thereby the level of MMP-9 (Zhou et al., 2014); (2) AMPK activation. Matrix protein deposition requires stimulation of protein synthesis. In DN, PI3K/Akt/mTORC1 signaling is activated due to the suppression of AMPK activity (Yuzawa, 2012). The inactivation of AMPK was found to be partially ascribed to the reduction of H 2 S level (Lee et al., 2012). NaHS dose-and time-dependently activates the phosphorylation of AMPK and inhibits the stimulation of mTOR (Lee et al., 2012). As a result, mTOR mediated protein synthesis is inhibited which partially accounts for the protective effect of H 2 S in DN. (3) Stimulation of NO formation. Hyperglycemia upregulates NOX4 expression resulting in the generation of intracellular ROS and laminin γ1, both of which contribute to renovascular remodeling (Gorin et al., 2015). Expression of NOX4 can be attenuated by (1) H 2 S inhibits ROS formation by activating Nrf2 pathway; (2) H 2 S activates AMPK, thereby suppressing PI3K/Akt/mTORC1 signaling and subsequent protein synthesis; (3) H 2 S stimulates NO formation by induction of iNOS expression which inhibits NOX4 level and ROS production. ROS, reactive oxygen species; MMP-9, matrix metalloproteinase-9; AMPK, AMP-activated protein kinase; NO, nitric oxide; iNOS, inducible nitric oxide synthase; mTOR, mechanistic target of rapamycin.
H 2 S (Kamat et al., 2015). L-NAME, a NOS inhibitor, abolished the effect of H 2 S (Feliers et al., 2016). Further study showed that H 2 S upregulates the protein expression of iNOS but not that of eNOS (Feliers et al., 2016). Moreover, iNOS siRNA can block the effect of H 2 S (Feliers et al., 2016) suggesting a role for NO in H 2 S mediated protective effect in DN. Interestingly, NO can also stimulate the activity of CSE and H 2 S production (Nagpure and Bian, 2016) suggesting a cross-talk interaction between these two pathways.

H 2 S in Hypertensive Nephropathy
Hypertensive nephropathy, a result of chronic hypertension, is the second leading cause of CKD worldwide (Hart and Bakris, 2010). Long time of high BP load results in damages of kidney, especially in glomerulus. When kidney cannot function properly, it fails to regulate BP. In turn, BP will rise up which further aggravates renal damage (Bidani and Griffin, 2004). Extensive evidence suggests a role of H 2 S in BP control. Genetic deletion of CSE causes hypertension and diminished endothelium dependent vasorelaxation . Similar with this, inhibition of both CBS and CSE also increases BP in rat (Roy et al., 2012). These data implies that H 2 S is a physiologic regulator of BP. Further studies suggest that H 2 S functions as both an EDHF (Mustafa et al., 2011;Edwards et al., 2012;Tang et al., 2013) and an EDRF (Wang, 2009). The BP lowering effect of exogenous H 2 S has been subsequently determined. Our group first showed that H 2 S donor, NaHS, attenuated BP by inhibiting plasma renin activity in a 2K1C rat model (Lu et al., 2010). Thereafter, NaHS was demonstrated to reduce BP in spontaneous hypertensive rats (Ahmad et al., 2014), angiotensin II treated mice (Al-Magableh et al., 2015), and sFlt transgenic mice (Holwerda et al., 2014). Additionally, renal protective effects of H 2 S have also been reported in hypertensive nephropathy. sFlt overexpression in mouse results in hypertension with proteinuria and glomerular endothelosis, all of which are apparently attenuated by administration of NaHS (Holwerda et al., 2014). Further study suggests an involvement of VEGF in this H 2 S mediated effect (Holwerda et al., 2014). Both H 2 S and tempol were found to alleviate renal dysfunction in a spontaneous hypertensive rat model by suppressing ROS formation (Ahmad et al., 2014). Recently, Jin's group reported that either NaHS or its metabolite sodium thiosulfate attenuated angiotensin II induced proteinuria, renal dysfunction, and structural deterioration in rat (Al-Magableh et al., 2015). Further studies suggested that the renal effects of H 2 S were partially mediated by suppression of epithelial sodium channel (Zhang et al., 2013;Wang et al., 2015). Taken together, H 2 S might be an ideal candidate for the treatment of hypertensive nephropathy.

Role of Mitochondrial H 2 S Pathway in Kidney Diseases
In contrast to the extensive studies on CBS and CSE in various kidney diseases, the role of 3-MST is largely neglected although definitive evidence has demonstrated its abundance in kidney (Aminzadeh and Vaziri, 2012;Shibuya et al., 2013;Kimura, 2014). 3-MST resides in mitochondria and is the major producer of mitochondria derived H 2 S. 3-MST silencing was reported to reduce bioenergetic parameters and H 2 S can serve as an electron donor in mammalian cells (Modis et al., 2013), indicating a possible physiological role of 3-MST/H 2 S pathway in maintaining mitochondrial electron transport and cellular bioenergitics. Renal ischemia is the most common cause of acute kidney injury. Hypoxia leads to the inhibition of mitochondrial respiratory chain by deprivation of O 2 . What will happen if H 2 S is supplemented into the mitochondria when hypoxia occurs? Szabo's group (Ahmad et al., 2016) showed that mitochondrially targeted H 2 S donor AP39 ameliorated renal damage in an ischemia/reperfusion rat model, suggesting possible involvement of 3-MST/H 2 S pathway in the pathogenesis although the change of 3-MST level and activity upon hypoxia was not determined. Additionally, the importance of 3-MST/H 2 S pathway is also suggested by the fact that ROS, the common cause of renal diseases of all types, induces the translocation of CBS and CSE into mitochondria (Fu et al., 2012). Thus, the revelation of the involvement of 3-MST in kidney diseases is of great value.

Targeting DAO/3-MST Pathway for H 2 S Delivery to Kidney
Safety is always a concern when administrating H 2 S systemically due to its well-known toxicity (Guidotti, 2010;Hirose, 2010) which might hamper its development as therapeutics. This safety issue might be diminished by specific delivery of H 2 S to the targeted organs. Unfortunately, such organ specific H 2 S donor is not reported. The discovery of DAO/3-MST pathway might provide a clue about how to deliver H 2 S specifically into the kidney. The pathway uniquely utilizes D-cysteine rather than L-cystein to produce H 2 S and exclusively exists in cerebellum and kidney (Shibuya et al., 2013). When D-cysteine is given, it attenuates IRI in kidney with higher potency than L-cysteine (Shibuya et al., 2013). As H 2 S is broadly renal protective as we have reviewed above, D-cysteine might also ameliorate other renal diseases. Direct administration of D-cysteine induces the generation of H 2 S in both cerebellum and kidney (Shibuya et al., 2013). Nevertheless, it is possible that structural modification of D-cysteine can generate a novel moiety providing D-cysteine to kidney but not cerebellum due to the impermeability of bloodbrain barrier. This hypothesis warrants further investigation.

Test of Drug-Like H 2 S Donors in Kidney Disease
To date, most studies of H 2 S effect in kidney have been largely restricted to the use of NaHS as an H 2 S donor despite of rare exceptions with GYY4137 or AP39. However, NaHS releases H 2 S at an uncontrolled manner (Li et al., 2008) and is unlikely to be a therapeutic agent (Szabo, 2007). Numerous drug-like H 2 S donors have been developed and some of them are under the investigation in clinical trials. For instance, an orally active H 2 S donor SG-1002 is proven to be safe in humans and underwent Phase II study for heart failure (ClinicalTrials.gov identifier: NCT01989208); Antibe Therapeutics are conducting several preclinical or clinical studies with their various H 2 S donors (ATB-346 for osteoarthritis; ATB-352 for Acute pain; ATB-350 for Thrombosis). The information on these H 2 S releasing donors can be found in Wallace and Wang (2015). Testing of these druglike H 2 S donors will not only consolidate the protective effect of H 2 S, but also shed light on the translation of H 2 S as a therapeutic agent for renal diseases.
Understanding the Molecular Mechanism of H 2 S Last, but not least, one should bear in mind that the molecular mechanisms underlying H 2 S effect is still not well-understood. It seems that H 2 S may partially exert its effect as a reducing agent to eliminate ROS (Bruce King, 2013), while its effects on gene expression may be related to specific molecular targets like NF-κB and the ERK pathways (Oh et al., 2006). However, the molecular details are still unclear. Thus, the in-depth portrayal about the interaction between H 2 S and its target protein will be interesting.

CONCLUSION
After recognition as the third gaseous mediator after NO and CO, the biological actions of H 2 S are still expanding. In kidney, it is actively participating in the renal regulation in physiological condition. Due to the regulatory role of kidney in the body, it is possible that H 2 S has far-reaching actions by modulating renal function which needs to be uncovered. Because of the significant role of H 2 S in renal physiology, it is not surprising that dysfunction of H 2 S contributes to the pathogenesis of kidney related diseases. Thereafter, administration of H 2 S mainly with NaHS was proven to rescue kidney damages in animal models with various types of kidney diseases. In the future, drug like H 2 S donors need to be tested to translate H 2 S as a treatment for renal diseases in clinical. Besides, in-depth studies of H 2 S mediated molecular actions are also needed to complete our understanding of the role of H 2 S in both renal physiology and pathology.

AUTHOR CONTRIBUTIONS
All authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication.