Adenosine Signaling Increases Proinflammatory and Profibrotic Mediators Through Activation of a Functional Adenosine 2B Receptor in Renal Fibroblasts

Available online at www.annclinlabsci.org Annals of Clinical & Laboratory Science, vol. 46, no. 4, 2016 339 Adenosine Signaling Increases Proinflammatory and Profibrotic Mediators through Activation of a Functional Adenosine 2B Receptor in Renal Fibroblasts

Patrick F. Wilkinson1, Francis X. Farrell2, Diane Morel1, William Law3, and Suzanne Murphy1

1University of the Sciences in Philadelphia, Philadelphia PA, 2Jefferson College of Health Sciences, Roanoke VA, and 3Purdue University Calumet, Hammond, IN, USA

Abstract. Interstitial renal fibrosis is a major pathophysiological manifestation of patients diagnosed with Chronic Kidney Disease (CKD), Diabetic Nephropathy (DN) and other inflammatory diseases. Adenosine signaling is an innate autocrine and paracrine cellular signaling pathway involving several key mediators that are elevated in the blood and kidneys of patients with DN. In these studies, we hypothesized that extracellular adenosine signals through one or more functional adenosine GPCRs on renal fibroblasts which increases profibrotic and proinflammatory mediators by inducing an activated fibroblast phenotype. Utiliz- ing the renal fibroblast cell line NRK-49F, the presence and relative abundance of adenosine receptors (AR) A1, A2A, A2B, and A3 were quantified by RT-PCR. Under normal homeostatic conditions, only AR1 and AR2B were detected. The functionality of each receptor was then assessed by receptor specific pharmacological agonism and antagonism and assessed for modulation of the GPCR associated secondary messenger molecule, cyclic adenosine monophosphate (cAMP). Agonism of the AR2B receptor resulted in increased intracellular cAMP while agonism of the AR1 receptor inhibited cAMP modulation. Upon direct agonism of the AR2B receptor, transcripts for profibrotic and inflammatory mediators including SMA-α, IL-6, TGF-β, CTGF, and fibronectin were elevated between 2-4 fold. These data indicate that renal fibroblasts express a functional AR1 receptor that inhibits cAMP upon stimulation, leading to a functional AR2B receptor that increases cAMP upon stimulation and also induces an activated fibroblast phenotype resulting in increased fibrotic and inflammatory mediators.

Key words: Adenosine, Adenosine receptor 2B, cAMP, renal fibroblasts, fibrosis, inflammation, myofibroblasts, SMA-a.

Introduction particular, studies have determined that the extent of tubulointerstitial ECM accumulation correlates Renal fibrosis is the pathophysiological outcome of with renal function decline; thus the extent of dam- increased circulating inflammatory proteins which aged tubulointerstitial area is both a marker of kid- induce chronic remodeling of the kidney, leading to ney function and a driver of renal failure [2]. The an increase in extracellular matrix (ECM) produc- majority of tubulointerstitial fibrosis is thought to tion. Accumulation of ECM can result in end stage be caused by myofibroblasts, or activated fibro- renal failure leaving patients dependent on dialysis blasts, that have higher rates of proliferation and and eventual kidney transplantation as their only increased ECM protein production [3,4]. treatment options. Multiple cellular responses such as mesangial cell and fibroblast activation, immu- Increased presence of circulating adenosine corre- nological cell infiltration, epithelial to mesenchy- lates with disease severity in human patients diag- mal transition (EMT), and inflammatory cytokine nosed with Diabetic Nephropathy while signifi- release all contribute to ECM accumulation in re- cantly increased adenosine levels in the renal vein of nal disease [1]. Interstitial renal fibrosis is consid- diabetic rats indicate a strong correlation between ered the hallmark of progressive renal disease. In adenosine and renal failure [5,6]. Extracellular adenosine signals through members of the P1 purino- Address correspondence to Patrick Wilkinson, 1400 McKean Rd, ceptor family which consists of four G- protein Spring House , PA, 19477; e mail: [email protected] coupled receptors (GPCR) named AR1, AR2A,

released from the cell and then converted to adenosine mono-phosphate (AMP) by cell surface en- zyme ecto-phosphodiesterase, followed by conver- sion of AMP to adenosine by ecto-5 nucleosidase.

The profibrotic effects of extracellular adenosine signaling on multiple cell types, including renal mesangial and epithelial cells, have been previously reported. However, the effects of chronic, physio- logically relevant high dose adenosine stimulation on interstitial renal fibroblasts in vitro have not been studied to date. Considering the fact that adenosine receptor activation can lead to increases in TGFβ, fibronectin, collagen I and IV gene and protein expression, adenosine receptors may have a sig- nificant impact on fibroblast activation and ECM deposition in the kidney. In this study, we aim to Figure 1. Quantification of AR1, AR2A, AR2B, and AR3 transcript in NRK-49F. NRK-49F cells, cultured to con- characterize the presence and functionality of ade- fluence were assessed for Adenosine receptor subtype mRNA nosine receptors on renal fibroblasts and quantify expression under basal conditions using quantitative RT- PCR. Error bars represent standard deviation of three the effect of adenosine receptor signaling on mark- separate experiments. All ct values above 36 were consid- ers of myofibroblast transdifferentiation and renal ered undetectable. fibrosis. AR , and AR with each receptor having unique 2B 3 Materials and Methods G protein driven intracellular response dependent on cell type or tissue origin. Chronically elevated Small molecule adenosine receptor agonists and an- levels of adenosine in mice leads to increased fibro- tagonists. 5'-N-Ethylcarboxamidoadenosine (NECA) sis of the lung, liver, and kidney that can be attenu- and adenosine (ADO) were purchased from Sigma ated with adenosine A receptor antagonism, 2B Aldrich. AR1 specific receptor agonists 2-Chloro-N- highlighting the direct influence of adenosine A2B cyclopentyl-2'-methyladenosine (MeCCPA), AR2B re- receptor signaling in fibrotic disease [7-9]. Increased ceptor agonist 2-[[6-Amino-3,5-dicyano-4-[4- adenosine signaling in the kidney may also increase (cyclopropylmethoxy)phenyl]-2-pyridinyl] TGF-β1, a profibrotic cytokine responsible for in- thio]-acetamide (BAY-606583), AR2B receptor antago- creased cell proliferation, matrix formation, and nist 8-[4-[4-(4-Chlorophenzyl)piperazide-1-sulfonyl) epithelial to mesenchymal transition[10]. phenyl]]-1-propylxanthine (PSB603), and AR1 receptor antagonist trans-4-[(2-Phenyl-7H-pyrrolo[2,3-d]pyrim- idin-4-yl)amino]cyclohexanol (SLV302) were purchased Adenosine is a nucleoside synthesized de novo by from Tocris Biosciences. All small molecule reagents nearly all cell types and can be transported to the were reconstituted in 100% DMSO with working con- cell surface and released to operate as a dynamic centrations of DMSO kept below 0.4% in all signaling molecule to induce intracellular and tran- experiments. scriptional changes in renal, cardiac, immune, and lung tissues [11-13]. Extracellular adenosine origi- Cell culture. The rat renal fibroblast cell line NRK-49F nates as intracellular adenosine triphosphates (ATCC CRL-1570) was used in all experiment between (ATP), cyclic adenosine monophosphate (cAMP), the passage numbers 20 and 30. Cells were cultured in or from the polyA tail of messenger RNA and is Dulbecco's modified eagle medium containing GlutaMax and supplemented with 10% fetal bovine se- converted enzymatically to extracellular adenosine rum 100 units/mL of penicillin, 100 µg/mL of strepto- by cell surface enzymes CD39 and CD73[14]. mycin, and 0.25 µg/mL of Fungizone® antimycotic. All Adenosine can be formed intracellularly and trans- cell culture reagents were purchased from Life ported into the interstitial space. Or, cAMP can be Technologies. Induction of renal fibrosis via AR2B 341 mRNA Analysis. NRK-49F cells were cultured to 90% confluency in 48 well dishes stimulated with BAY 606583 for 24 hours, lysed in RLT + 1% β- mercap- to ethanol, and then subjected to Qiagen fast prep RNA isolation kit as per manu- factures instructions. RNA at 5ng/ul final concentration was transcribed to cDNA using qScipt (Quanta) cDNA synthesis kit. Taqman primer probes were designed to detect transcript for αSMA (NM_031004.) CTGF (NM_022266), Fibronectin (NM_019143). TGFB1 (NM_021578, IL-6(NM_012589) Adenosine receptor subtype 1 (NM_017155), Adenosine receptor subtype 2A (NM_053294), Adenosine receptor subtype 2B (NM_017161), and Adenosine receptor subtype A3 (AF102804) were used to quantify relative transcript expression via qRT-PCR. Adenosine receptor expression data was normalized to GAPDH and expressed as percent transcript per GAPDH transcript for quantification of adenosine receptor expression or as fold change as compared to vehicle control for analysis of fibrotic and inflammatory transcripts. Figure 2. Stimulation of Adenosine receptors increases intracellular cAMP. (A) Modulation of intracellular cAMP was assessed 30 min after NRK-49F cells were stimulated with DMSO control, ADO, or NECA. Cyclic AMP quantification. Cells were (B) Cells were stimulated for 30min with DMSO, AR1 agonist plated at 2000 cells per well in 96 well 2’-MeCCPA, or AR2B agonist BAY606583. Error bars represent the stan- white plastic clear bottom tissue culture dard deviation of two replicates per data point. treated plates (Costar 3903) and incu- bated at 37°C, 5% CO2 overnight in growth media. Na3 VO4, and 1 µg/ml leupeptin (Cell Signaling tech- NRK-49F cells were then pretreated with various phar- nologies) containing protease and phosphatase inhibi- macological modulators including AR2B adenosine re- tors (Pierce) by cell scraping and incubation on ice for ceptor agonist (BAY 60-6583) antagonist (PSB 603) or 30min followed by centrifugation at 10,000xg for 10 ARA1 receptor agonist (2-MeCCPA) antagonist (SLV- min. Protein was detected using antibodies directed 320) for 45min prior to stimulation with adenosine or against rat αSMA (clone 1A4) from Sigma Aldrich or NECA for 30 min. Cells were then lysed and analyzed CTGF from Santa Cruz (clone H-55). for cAMP using cAMP LANCE detection kit as per manufacturer’s instructions (Perkin Elmer), quantifying Results luminescence as counts per minute. Data was expressed as % change from unstimulated cells. Adenosine receptors are present on all cell types that have been studied to date. However, the pro- Western Analysis. NRK-49F cells were seeded at file and relative expression of each individual ade- 150,000 per well of a 6 well tissue culture plate and rest- ed in growth media overnight. Cells were then stimu- nosine receptor can vary depending on cell type or lated with increasing doses of BAY606583 for 24hrs tissue of origin. To identify and quantify which (αSMA protein) or 48hrs (CTGF protein). Cells were adenosine receptors are present on interstitial renal lysed in 20 mM Tris-HCl (pH 7.5) 150 mM NaCl, 1 fibroblasts, the cell line NRK-49F was profiled for mM Na2EDTA, 1 mM EGTA, 1% Triton, 2.5 mM so- the detection and relative mRNA expression levels dium pyrophosphate, 1 mM b-glycerophosphate, 1 mM of all four adenosine receptor under basal 342 Annals of Clinical & Laboratory Science, vol. 46, no. 4, 2016

24.3+/-3.9 µM (SD). Treatment of cells with ADO or NECA produced a robust induction of cAMP at 30 minutes, indicating that renal fibroblasts exposed to extracellular adenosine signal through one or both of the ARA1 or AR2B receptors.

Since both AR1 and AR2B receptor subtypes were detected by qRT-PCR, specific pharmacological modulators of AR1 and AR2B receptors were used to assess each receptor’s contribution to the observed increase in cAMP. The AR1 specific receptor agonists 2’-Me-CCPA and AR2B receptor agonist BAY- 606583 were used to identify which percentage of cAMP modulation could be attrib- Figure 3. ARA1 and ARA2B have opposing effects on cAMP modulation in renal fibroblasts. Effects of pharmacological agonism or antagonism of AR and AR on 1 2B uted to the AR1 or AR2B re- cAMP was assessed in the presence of NECA or ADO. (A) Cells were pretreated for ceptor, respectively. 45min with AR agonist 2’-MeCCPA or AR antagonist PSB603 followed by 1 2B Increasing doses of the NECA stimulation for 30min. (B) Cells were pretreated for 45min with AR1 antagonist SLV302 or AR1 agonist 2’-MeCCPA, and then stimulated for 30min 2’-Me-CCPA had no influ- with ADO. Error bars represent the standard deviation of two replicates per data ence on cAMP while the points. AR2B agonist Bay-606583 dose dependently increased conditions. After 24 hours of culture, cells were cAMP maximally 500% above unstimulated cells analyzed by qRT-PCR for AR1, AR2A, AR2B, and with an EC50 of 11.57+/-4.5 µM (SD) (Figure AR3 transcript. Only the mRNAs for the adenos- 2B). These data indicate that the AR2B receptor is ine receptor subtypes ARA1 and ARA2B were de- functional in renal interstitial fibroblasts and able tected in these cells. The AR2B receptor transcript to robustly increase cAMP after specific AR2B re- was more highly expressed than ARA1 (Figure 1). ceptor stimulation.

To assess functionality of AR1 and AR2B receptors To more completely characterize the total effects of detected on renal fibroblasts, cells were stimulated adenosine signaling on renal fibroblasts, cells were with either ADO or the adenosine analogue NECA pretreated with 25µM of the AR2B receptor antago- and assessed for modulation of the GPCR second- nist PSB603 or with the AR1 receptor agonist ary messenger molecule cAMP. After 30 minutes of 2-MeCPPA and then stimulated with increasing stimulation with ADO or NECA a dose dependent doses of NECA (Figure 3A). Antagonism of the increase in cAMP were observed (Figure 2A). AR2B receptor decreased NECA stimulated cAMP ADO increased cAMP maximally 90% above un- while agonism of the AR1 receptor was similarly stimulated cells with an EC50 of 48.9+/-15.3 µM able to suppress NECA modulated cAMP. To con- (SD), and NECA increased cAMP maximally firm that agonism of the AR1 receptor is able to 178% above unstimulated cells with an EC50 of decrease ADO driven cAMP modulation, cells were Induction of renal fibrosis via AR2B 343

Figure 4. Increased inflammatory and fibrotic mediators with AR2B agonism. Effects of AR2B stimulation on inflammatory and fibrotic mediators. Cells were stimulated with BAY-606583 for 24 hours and αSMA (A), fibronectin (B), TGF-B1 (C), CTGF (D), and IL-6 (E) transcript were quantified by qRT-PCR. Western blot analysis was used to quantify αSMA (F) protein in cells stimulated with BAY606583 for 48hrs. Error bars for transcript data represent the standard deviation of three experiments. Error bars for αSMA protein represent standard deviation for five experiments. Statistical significance was calculated using one way ANOVA with a Dunnett’s post test*p<0.05, **p<0.01, ***p<0.001 ****p<.0001 pretreated for 45min with 25µM of either renal fibroblasts, cells were stimulated with the 2-MeCPPA or the specific AR1 receptor antagonist AR2B agonist BAY606583 for 24 hours and chang- SLV302 (Figure 3B). Agonism of the AR1 was es in αSMA, fibronectin, CTGF, IL-6, and TGF-β1 able to decrease native ADO induced cAMP modu- transcript levels were quantified by qRT-PCR lation indicating signaling through the AR1 recep- (Figure 4). An increase in αSMA, fibronectin, tor suppresses ADO associated cAMP signaling. CTGF, IL-6, and TGF-β1 mRNA between 2-4 Antagonism of the AR1 did not increase the po- fold was observed in cells stimulated with tency of ADO driven cAMP modulation, although BAY606583 as compared to vehicle stimulated cells cAMP was more maximally induced with AR1. (Figure 4A). Additionally, a 130% increase in This indicates reduced adenosine signaling may still αSMA protein was observed in cells stimulated sustain a small amount of cAMP signaling through with increasing doses of BAY606583 after 24 hours, AR1. AR1 and AR2B receptors are present and func- confirming the observed increases in αSMA tran- tional on NRK-49F renal fibroblasts and uniquely script (Figure 4B). operate as a G inhibitor (GI) GPCR or G stimula- tory (Gs) GPCR, respectively, due to the opposite Discussion effects on cAMP after stimulation. These data indicate that a high concentration of ad- Increased adenosine signaling has also been report- enosine results in activation of renal fibroblasts to a ed to increase profibrotic and proinflammatory me- myofibroblasts phenotype, resulting in overproduc- diators. To assess the effects of adenosine signaling tion of proinflammatory and profibrotic mediators. on profibrotic and proinflammatory mediators in Although both AR1 and AR2B receptors are present 344 Annals of Clinical & Laboratory Science, vol. 46, no. 4, 2016 on interstitial renal fibroblasts, the dominant ade- AR2B specific small molecule agonists and antago- nosine signaling is driven through the AR2B recep- nist were used to confirm receptor functionality tor which results in increased intracellular cAMP. and response. Similar to data published by Dubey et al, agonism of AR2B receptor in fibroblasts ro- Increased adenosine signaling has been implicated bustly increased cAMP indicating that the adenos- as a profibrotic and proinflammatory mediator in ine driven cAMP response is mediated through both in vitro and in vivo model systems. High mi- AR2B signaling only[17]. Perhaps unique to inter- cromolar levels of adenosine can be released upon stitial renal fibroblasts, it was also observed that tissue injury, where adenosine signaling facilitates agonism of AR1 inhibited ADO and NECA in- tissue repair mechanisms promoting tissue remod- duced cAMP, indicating a suppressive cAMP re- eling [15]. However, there has also been evidence sponse to ADO. Decreased cAMP modulation af- that chronically elevated levels of adenosine can re- ter AR1 receptor agonism in the presence of sult in detrimental tissue remodeling, resulting in adenosine indicates that the AR1 operates as a GI fibrosis, excess scar formation and organ impair- GPCR. These data suggest that the AR1 receptor is ment [8]. It has been reported that increasing se- able to down regulate adenosine signaling to some rum adenosine levels correlates with disease severity extent; however, this suppression of cAMP can be in diabetic nephritis patients [6]. The initial stages overwhelmed under high adenosine concentration of diabetic renal disease have been well character- when the low affinity AR2B receptor is strongly ized and include renal hyperfiltration, mesangial activated. cell hypertrophy and separation of the glomeruli from the Bowman ’s capsule [2]. As Diabetic In renal fibroblasts, the totality of high micromolar Nephrotic disease progresses, renal function de- adenosine concentration is a net increase in cAMP clines, culminating in end stage renal failure, where signaling. Along with the observed increase in the disease pathophysiology is highlighted by inter- cAMP, renal fibroblasts also robustly increase mul- stitial fibrosis indicating extensive tissue remodel- tiple profibrotic and proinflammatory mediators. ing and loss of kidney function [16]. The intracellular cytoskeletal protein αSMA was robustly increased in the presence of high concen- Although multiple studies have identified adenos- tration of AR2B receptor stimulation. αSMA is a ine to be a profibrotic signaling molecule in fibro- marker of an activated fibroblasts phenotype known blasts from skin, lung and liver, to date no data has as a myofibroblast, that are characterized by in- been generated examining the effects of adenosine creased ECM production and increased prolifera- on interstitial renal fibroblasts in vitro. To better tion [1]. In support of this myofibroblast pheno- understand the influence of increased circulating type, we observed increased CTGF, fibronectin, adenosine in renal fibrosis, kidney fibroblasts were TGF-β1, and IL-6 transcript when renal fibroblasts stimulated with high concentrations of adenosine were stimulated with high concentrations of ade- equivalent to levels observed in animal models of nosine. Specific agonism of the AR2B receptor renal failure and patients with end stage renal fail- drove renal fibroblasts transdifferentiation to a ure [6,10]. Adenosine receptor subtype expression more activated myofibroblast cell type, with a cor- profiles change depending on cell type, so our ini- responding increase in the release of multiple fi- tial experiments identified that the AR1 and AR2B brotic and inflammatory mediators that contrib- are the only adenosine receptor subtype transcripts utes to the accumulation of ECM proteins. expressed under basal conditions in NRK-49F renal fibroblasts. High concentrations of native ade- In conclusion our examination of adenosine recep- nosine (or NECA) stimulation of AR1 and AR2B tor signaling in renal fibroblasts confirms the pres- lead to increased intracellular cAMP. To quantify ence and functionality of the AR1 and AR2B recep- the individual contribution of AR1 and AR2B sig- tors and identifies them as GI and GS GPCR, naling on cAMP modulation, individual AR1 and respectively. Under high concentrations of Induction of renal fibrosis via AR2B 345

7. Eldaif, S.M., et al., Attenuation of renal ischemia-reperfusion adenosine, the AR2B is responsible for a profibrotic injury by postconditioning involves adenosine receptor and and proinflammatory response. These data indicate protein kinase C activation. Transpl Int, 2010. 23(2): p. that blockade of the AR2B may be a promising tar- 217-26. get to attenuate adenosine driven intestinal renal 8. Sun, C.X., et al., Role of A2B adenosine receptor signaling in adenosine-dependent pulmonary inflammation and injury. J fibrosis. Clin Invest, 2006. 116(8): p. 2173-2182. 9. Dai, Y., et al., A2B adenosine receptor-mediated induction of Conflict of Interest IL-6 promotes CKD. J Am Soc Nephrol, 2011. 22(5): p. 890-901. Author Patrick Wilkinson is an employee of Janssen R&D. 10. Roa, H., et al., Adenosine mediates transforming growth fac- tor-beta 1 release in kidney glomeruli of diabetic rats. FEBS References Lett, 2009. 583(19): p. 3192-8. 11. Belardinelli, L., J. Linden, and R.M. Berne, The cardiac effects 1. Grande, M.T. and J.M. Lopez-Novoa, Fibroblast activation of adenosine. Prog Cardiovasc Dis, 1989. 32(1): p. 73-97. and myofibroblast generation in obstructive nephropathy. Nat 12. Huang, S., et al., Role of A2a extracellular adenosine receptor- Rev Nephrol, 2009. 5(6): p. 319-28. mediated signaling in adenosine-mediated inhibition of T-cell 2. Dronavalli, S., I. Duka, and G.L. Bakris, The pathogenesis of activation and expansion. Blood, 1997. 90(4): p. 1600-10. diabetic nephropathy. Nat Clin Pract Endocrinol Metab, 2008. 13. Zhou, Y., et al., Alterations in adenosine metabolism and sig- 4(8): p. 444-52. naling in patients with chronic obstructive pulmonary disease 3. Qi, W., et al., The renal cortical fibroblast in renal tubulointer- and idiopathic pulmonary fibrosis. PLoS One, 2010. 5(2): p. stitial fibrosis. Int J Biochem Cell Biol, 2006. 38(1): p. 1-5. e9224. 4. Humphreys, B.D., et al., Fate tracing reveals the pericyte and 14. Jackson, E.K., J. Ren, and Z. Mi, Extracellular 2',3'-cAMP is a not epithelial origin of myofibroblasts in kidney fibrosis. Am J source of adenosine. J Biol Chem, 2009. 284(48): p. Pathol, 2010. 176(1): p. 85-97. 33097-106. 5. Angielski, S., et al., Renal handling and metabolism of adenos- 15. Grenz, A., et al., The reno-vascular A2B adenosine receptor ine in diabetic rats. Contrib Nephrol, 1989. 73: p. 52-7; discus- protects the kidney from ischemia. PLoS Med, 2008. 5(6): p. sion 57-8. e137. 6. Xia, J.F., et al., Ultraviolet and tandem mass spectrometry for 16. Nath, K.A., Tubulointerstitial changes as a major determinant simultaneous quantification of 21 pivotal metabolites in plas- in the progression of renal damage. Am J Kidney Dis, 1992. ma from patients with diabetic nephropathy. J Chromatogr B 20(1): p. 1-17. Analyt Technol Biomed Life Sci, 2009. 877(20-21): p. 17. Dubey, R.K., et al., A(2b) receptors mediate the antimitogenic 1930-6. effects of adenosine in cardiac fibroblasts. Hypertension, 2001. 37(2 Pt 2): p. 716-21.