Adenosine receptors as emerging therapeutic targets for diabetic kidney disease

Adenosine: formation and metabolism

Adenosine is an endogenous purinergic nucleoside. Intracellular adenosine is synthesized either by the dephosphorylation of adenosine monophosphate (AMP) via intracellular 5’-nucleotidase (5’NTD) or by the hydrolysis of S-adenosylhomocysteine (Fig. 1). Extracellular adenosine is produced by dephosphorylation of ATP via a two-step enzymatic reaction sequence. First, CD39 converts ATP or adenosine diphosphate into AMP. In the second step, CD73 converts AMP into adenosine. Adenosine deaminase (ADA) catabolizes adenosine into inosine, and equilibrative nucleoside transporters (ENTs) carry adenosine across the cell membrane in either direction (Fig. 1). Extracellular adenosine is low under normal conditions (20–300 nM) but rises dramatically upon tissue injury, such as hypoxia and inflammation, due to an increased demand for energy supplied by ATP [25]. Increased adenosine up to a certain level provides protection through increased blood flow via vasodilation and through anti-inflammatory/immune-modulatory cascade [26]. However, a chronically increased level of adenosine may play a role in tissue injuries such as hepatic steatosis, asthma, and fibrosis [8]. In fact, persistent and excessive adenosine exposure under ADA deficiency leads to kidney fibrosis [27]. In addition, adenosine uptake is reduced in kidney proximal tubular cells under diabetic stress resulting from reduced ENT, and ENT deficiency leads to kidney fibrosis [19].

Figure 1.

Adenosine: formation and metabolism.

Intracellular adenosine is generated from SAH hydrolase or 5’NTD and is degraded by ADA and AK. Extracellular adenosine is generated by CD73 and converted to inosine by ADA. ENTs allow adenosine flux through the cell membrane depending on gradient concentration.

ADA, adenosine deaminase; AK, adenosine kinase; AMP, adenosine monophosphate; ATP, adenosine triphosphate; ENT, equilibrative nucleoside transporter; SAH, S-adenosylhomocysteine; 5’NTD, 5’-nucleotidase.

Adenosine receptors: classification and signal transduction

Each AR exhibits differences in affinity for adenosine, G protein coupling, and subsequent intracellular signal transduction (Fig. 2).

Figure 2.

Signaling cascade of each AR.

AR, adenosine receptor; cAMP, cyclic adenosine monophosphate; CREB, cAMP-response element-binding protein; DAG, diacylglycerol; Epac, exchange proteins activated by cAMP; ERK, extracellular signal-regulated protein kinase; GSK-3β, glycogen synthase kinase 3 beta; IP₃, inositol trisphosphate; JNK, c-Jun N-terminal kinase; MAPK, mitogen-activated protein kinase; NO, nitric oxide; PI3K, phosphatidylinositol 4,5-bisphosphate; PLD, phospholipase D.

Distribution of adenosine receptors in the kidney

Data from functional studies, immunohistochemistry, and transgenic AR mice show all four ARs are expressed in the kidney [19,36–39]. As summarized in Fig. 3A, all four ARs are expressed in whole kidney homogenates [38], the glomeruli [19,37,39], the proximal tubules [19,40], and the afferent/efferent arterioles [41]. The loop of Henle, distal convoluted tubules, and collecting duct express A₁AR/A_2AAR [37,38], A₁AR/A_2BAR [37,38], and A₁AR [37], respectively. A recent single-cell transcriptomic analysis study [42] provided a comprehensive atlas of AR expression in the healthy mouse kidney (Fig. 3B).

Figure 3.

Distribution of ARs in the kidney.

(A) Summary of each AR expressed in the nephron. (B) Expression of each AR in healthy mouse kidney delineated by single-cell transcriptomic analysis (https://susztaklab.com).

A-IC, alpha intercalated cell; ALOH, ascending loop of Henle; AR, adenosine receptor; B-IC, beta intercalated cell; CD-PC, collecting duct principal cell; CD-Trans, CD transient cell; CNT, connecting tubule; DC 11b+, CD11b+ dendritic cell; DCT, distal convoluted tubule; DLOH, descending loop of Henle; Endo, endothelial; Fib, fibroblast; GEC, glomerular endothelial cells; Granul, granulocyte; LOH, loop of Henle; Macro, macrophage; Neutro, neutrophil; NK, natural killer cell; pDC, plasmacytoid DC; Podo, podocyte; PT, proximal tubule; Tgd, gamma delta T cell; Treg, regulatory T cell.

Physiological role of kidney adenosine receptors

Adenosine directly acts on the coronary arteries to increase blood flow by four times compared to that of at rest [43]. Interestingly, exogenous adenosine induces a marked but transient reduction in the kidney blood flow [44]. Blockade of endogenous adenosine signaling by theophylline-mediated AR antagonism does not alter basal kidney function; however, the exogenous adenosine-mediated reduction of GFR is inhibited with theophylline by 31.4% [45]. All AR subtypes are expressed in the kidney vasculature, yet levels of A₁AR and A_2BAR are high in the preglomerular microcirculation, whereas A_2AAR and A₃AR levels are low [46,47]. Adenosine is also involved in the regulation of the TGF of the nephron and renin release from juxtaglomerular cells [48,49].

The role of adenosine and adenosine receptors in the diabetic kidney disease

Plasma levels of adenosine and its catabolic product, inosine, are increased in patients with DKD [21,22]. Interestingly, plasma adenosine is not different between healthy and diabetic patients without kidney disease [21]. Our urinary analysis of diabetic patients showed proportional increases in urinary adenosine excretion according to the stage of DKD (Fig. 4A). In addition, urinary excretion of adenosine shows positive correlation with proteinuria or microalbuminuria and negative correlation with creatinine clearance (data not shown). Levels of adenosine in kidney tissue and urine are increased in streptozotocin (STZ)-induced diabetic mice and rats [16,71]. The isolated glomeruli from diabetic rats contains six-fold higher adenosine compared to controls [72]. Adenosine in renal venous plasma but not in renal arterial plasma is significantly increased in diabetic rats compared to controls [23], further suggesting increased synthesis and release of adenosine in diabetic kidneys. Additionally, kidney expression of CD73 transcription and protein are increased in STZ-induced diabetic mice and rats [16,71]. The expression of kidney adenosine kinase (AK) is decreased in diabetic rats, whereas administration of insulin to diabetic rats restores the kidney AK expression [73,74]. The administration of ADA in diabetic rats decreases the urinary excretion of proinflammatory cytokines and increases the anti-inflammatory cytokine [20], suggesting a pathogenic role of increased adenosine in DKD. The ENT, which modulates adenosine uptake, is reduced in diabetic kidneys, contributing to kidney fibrosis [19].

Figure 4.

Urinary excretion of adenosine and renal expression of ARs in diabetic patients.

(A) Urinary excretion of adenosine from normal (n = 10), normoalbuminuria (n = 10), microalbuminuria (n = 10), and overt proteinuria (n = 10) were measured. (B) Immunostaining of ARs was performed from minor glomerulopathy, minimal change disease, and DKD. Original magnification ×400. The data are presented as the mean ± standard error of mean.

AR, adenosine receptor; Cr, creatinine; DKD, diabetic kidney disease.

^**p < 0.01, ^***p < 0.01 vs. control; ^†p < 0.05 vs. normoalbuminuria.

In contrast to the above studies, administration of adenosine, inhibition of cellular uptake of adenosine, and blockade of AK have all been shown to reduce kidney injury in diabetes [75–77]. In addition, CD73 knockout diabetic mice exhibited more severe kidney injury than wild-type diabetic mice [16]. The exact reason for these seemingly contradictory findings is not clear but may be due to experimental conditions such as duration of diabetes and severity of DKD.

We confirmed that expression of each AR is significantly increased in both STZ-induced type-1 diabetes (T1D) and db/db type-2 diabetes (T2D) mouse kidney (Fig. 5A, B). In addition, analysis of the Nephroseq database of human diabetic kidney biopsy samples showed increased AR expression (Fig. 5C). Our immunohistochemical analysis shows that A_2AAR expression is increased in minor glomerulopathy, minimal change disease, and DKD. Interestingly, A₃AR expression is significantly increased only in DKD, and not in minor glomerulopathy or minimal change disease (Fig. 4B).

Figure 5.

The expression of ARs in diabetic kidney.

(A, B) Messenger RNA (mRNA) expressions of ARs were obtained from kidney homogenate of STZ-induced diabetic mice (12 weeks after STZ injury) and 20-week-old db/db mice. (C) Microarray analyses of human kidney biopsy data were provided in the Nephroseq database (nephroseq.org). The data are presented as the mean ± standard error of mean.

AR, adenosine receptor; ERCB, European Renal cDNA Bank; STZ, streptozotocin.

^*p < 0.05 vs. control, db/m, normal.

The involvement of adenosine and ARs in DKD has been also studied using genetically engineered models. Lack of A₁AR in diabetic mice aggravates both hyperfiltration and glomerular injuries [78]. Both A_2AAR or A_2BAR knockout diabetic mice exhibited more severe kidney injury, as indicated by increased albumin excretion compared to wild-type diabetic mice [13,16]. The severity of DKD in A₃AR knockout mice has not been reported, while A₃AR knockout mice are protected against ischemia- or myoglobinuria-induced AKI [79]. Considering the ubiquitous distribution and diverse effects of each AR, future studies with spatiotemporal regulation of AR during the progression of DKD are necessary to understand their exact roles in diabetic kidneys.

The role of NKCC2/adenosine/AR in the renoprotective effect of SGLT2 inhibitors was particularly interesting. SGLT2 inhibitors have recently been shown to prevent the progression of DKD [80,81]. SGLT2 inhibitors reduce proximal reabsorption of sodium chloride and increase sodium chloride concentrations at the macula densa, facilitating NKCC2 transporter which subsequently leads to production of adenosine, activates the TGF mechanism, and reduces hyperfiltration (summarized in section of Tubuloglomerular feedback) associated with DKD [82]. It is speculated that upregulated SGLT2, even in the presence of hyperfiltration in diabetic kidney, will reduce sodium chloride delivered to macular densa, inactivate NKCC2, decrease the adenosine/AR axis in TGF, and maintain hyperfiltration. This speculation is supported by the fact that SGLT2 inhibition in T1D patients increases urinary adenosine excretion [83]. Restoration of TGF as a result of increased extracellular adenosine (due to inhibition of ENT) also has been suggested as a renoprotective mechanism of dipyridamole in STZ-induced diabetic rats [84].

Therapeutic applications of adenosine receptor modulators against diabetic kidney disease

The consumption of more than two cups per day of coffee, which contains the nonselective adenosine antagonist caffeine, for 2 weeks improved eGFR in DKD patients [10], and there is an inverse relationship between caffeine consumption and mortality in CKD patients [85]. In addition, a nonselective AR antagonist, 8-(p-sulfophenyl) theophylline, effectively decreases kidney fibrosis and improves kidney function [86]. These data suggest that well-designed AR antagonists may become novel protective agents against DKD.

With respect to specific AR subtype, A₃AR antagonists consistently prevent experimental DKD [18–20] as summarized in Table 1 and Fig. 6. LJ2698 and MRS1220, the A₃AR antagonists, attenuate fibrosis, oxidative stress, and inflammation in db/db mice [18] and STZ diabetic rats [19,20], respectively. Our previous study showed that LJ1888, a prototype compound of LJ2698, prevents fibrosis in obstructed kidney as well [87]. Inhibition of transforming growth factor-β1 (TGF-β1)–induced extracellular matrix (ECM) upregulation via A₃AR-specific small interfering RNA in renal proximal epithelial cells [87] confirmed the fibrotic effects of A₃AR in the kidney. We have shown that LJ1888 reduced obstruction- or TGF-β1-induced upregulation of lysyl oxidase, which induces cross-linking of ECM, suggesting that A₃AR antagonists may also regulate ECM accumulation via posttranslational regulation [87]. Yet, more studies are needed to confirm the therapeutic effect of A₃AR antagonists against DKD, because decreased proinflammatory cytokines in response to A₃AR agonists and increased proinflammatory cytokines in A₃AR knockout mice under sepsis-induced kidney injury have been reported [88].

Table 1.

Structure and efficacy of AR modulators against DKD

Figure 6.

The role of ARs in diabetic kidney disease.

Each AR has either stimulatory (red dashed arrows) or inhibitory (blue solid inhibitory lines) effects on inflammation, fibrosis, oxidative stress, hyperglycemia, UAE, and GFR in diabetic kidneys.

AR, adenosine receptor; GFR, glomerular filtration rate; UAE, urinary albumin excretion.

Inflammation plays a key role in the development of DKD [1], and A_2AAR acts as a strong anti-inflammatory effector responding to extracellular adenosine [89]. In agreement with this, the A_2AAR agonists ATL146 or CGS21680 attenuate kidney injury including inflammation as well as fibrosis in STZ-induced diabetic rats [13–15]. A_2AAR activation also reduces cytokine and chemokine expression in the kidney tubular epithelial cells under hypoxia and inhibits the macrophage infiltration after ischemia-reperfusion [90,91]. A_2AAR activation, in fact, attenuates kidney fibrosis and suppresses the epithelial-mesenchymal transition in various kidney fibrosis models [92,93].

Since A₃AR antagonists or A_2AAR agonists protect against DKD, it is reasonable to develop AR modulators with dual action as an A₃AR antagonist and an A_2AAR agonist in order to have better protective effects against DKD. We showed for the first time that LJ4459, a newly synthesized dual-acting ligand, prevents the progression of tubulointerstitial fibrosis in obstructed kidneys [94]. It will be interesting to determine whether this dual-acting ligand has better therapeutic efficacy against DKD than an A₃AR antagonist or an A_2AAR agonist.

Opposing pro- and anti-therapeutic roles of A_2BAR in DKD have been reported. The A_2BAR agonist Bay60-6583 attenuates diabetic kidney fibrosis [16]. On the contrary, 1) Bay60-6583 promotes profibrotic activation of kidney fibroblasts [95] and 2) MRS1754, the A_2BAR antagonist, attenuates profibrotic effects along with reduced vascular endothelial growth factor expression in diabetic rats and mice [12,17]. An anti-inflammatory role of A₁AR in ischemia/reperfusion kidney injury has been reported [96–98], but the role of A₁AR modulators in DKD remains to be studied.

Oxidative stress plays a key role in the development and progression of DKD [99]. Diabetic stress including high glucose and lipids increases reactive oxygen species (ROS), and mitochondrial dysfunction and nicotinamide adenine dinucleotide phosphate oxidase (NOX) play key, if not all, roles in increased ROS under diabetic conditions [100]. A₃AR antagonists reduce adriamycin-induced NOS upregulation in the kidney [101], and A₁AR or A₃AR knockout reduces NOX-derived oxidative stress in mice [102,103].

The involvement of adenosine on metabolism should be considered, since hyperglycemia is a key player in DKD [1]. The adenosine system has been suggested to play a role in glucose homeostasis, contributing to the pathophysiology of T1D and T2D [3]. Caffeine reduces plasma glucose and increases pancreatic insulin in STZ diabetic rats [104]. Pharmacological inhibition of A₁AR or knockout of A₁AR in mice enhances basal insulin secretion [105,106]. On the other hand, A_2AAR activation increases insulin secretion and improves beta-cell function by inhibiting the inflammatory response [107,108]. The absence of A₃AR in mice improved a metabolic phenotype including reduced glucose clearance and increased insulin compared to wild-type mice [103] (Fig. 6). Although further mechanistic studies are required, activation of A_2AAR and inhibition of A₁AR or A₃AR improves insulin resistance and hyperglycemia, which may play a role in preventing DKD.

While various AR agonists or antagonists are under clinical trials, according to clinicaltrials.gov, targeting inflammatory, cancer, and cardiovascular diseases, none of the AR modulators are U.S. Food and Drug Administration-approved or under clinical trials against DKD at present.