Lipoxins Regulate the Early Growth Response–1 Network and Reverse Diabetic Kidney Disease
Total Page:16
File Type:pdf, Size:1020Kb
BASIC RESEARCH www.jasn.org Lipoxins Regulate the Early Growth Response–1 Network and Reverse Diabetic Kidney Disease Eoin P. Brennan,1,2 Muthukumar Mohan,1,3 Aaron McClelland,1 Christos Tikellis,1,3 Mark Ziemann,1,4 Antony Kaspi,1,4 Stephen P. Gray,1 Raelene Pickering,1,3 Sih Min Tan,1,3 Syed Tasadaque Ali-Shah,5 Patrick J. Guiry,5 Assam El-Osta,1,4 Karin Jandeleit-Dahm,1,3 Mark E. Cooper,1,3 Catherine Godson,2 and Phillip Kantharidis1,3 1Juvenile Diabetes Research Foundation Danielle Alberti Memorial Centre for Diabetes Complications, Diabetes Division, Baker IDI Heart and Diabetes Institute, Melbourne, Victoria, Australia; 2University College Dublin Diabetes Complications Research Centre, UCD Conway Institute of Biomolecular and Biomedical Research, UCD School of Medicine and Medical Sciences, and 5Centre for Synthesis and Chemical Biology, UCD School of Chemistry and Chemical Biology, University College Dublin, Dublin, Ireland; and 3Department of Diabetes and 4Epigenetics in Human Health and Disease Laboratory, Department of Diabetes, Central Clinical School, Monash University, Clayton, Victoria, Australia ABSTRACT Background The failure of spontaneous resolution underlies chronic inflammatory conditions, including microvascular complications of diabetes such as diabetic kidney disease. The identification of endoge- nously generated molecules that promote the physiologic resolution of inflammation suggests that these bioactions may have therapeutic potential in the context of chronic inflammation. Lipoxins (LXs) are lipid mediators that promote the resolution of inflammation. Methods We investigated the potential of LXA4 and a synthetic LX analog (Benzo-LXA4)astherapeutics 2 2 in a murine model of diabetic kidney disease, ApoE / mice treated with streptozotocin. Results Intraperitoneal injection of LXs attenuated the development of diabetes-induced albuminuria, mesangial expansion, and collagen deposition. Notably, LXs administered 10 weeks after disease onset also attenuated estab- lished kidney disease, with evidence of preserved kidney function. Kidney transcriptome profiling defined a diabetic signature (725 genes; false discovery rate P#0.05). Comparison of this murine gene signature with that of human diabetic kidney disease identified shared renal proinflammatory/profibrotic signals (TNF-a,IL-1b,NF-kB). In diabetic mice, we identified 20 and 51 transcripts regulated by LXA4 and Benzo-LXA4, respectively, and pathway analysis identified established (TGF-b1, PDGF, TNF-a,NF-kB) and novel (early growth response–1 [EGR-1]) networks acti- vatedindiabetes andregulatedby LXs.Inculturedhuman renal epithelial cells, treatment with LXs attenuated TNF-a–driven Egr-1 activation, and Egr-1 depletion prevented cellular responses to TGF-b1 and TNF-a. Conclusions These data demonstrate that LXs can reverse established diabetic complications and support a therapeutic paradigm to promote the resolution of inflammation. J Am Soc Nephrol 29: 1437–1448, 2018. doi: https://doi.org/10.1681/ASN.2017101112 Received October 19, 2017. Accepted January 23, 2018. Diabetes is recognized as a major epidemic, which has increased in incidence by 50% over E.P.B. and M.M. share first authorship. M.E.C., C.G. and P.K. share the past 15 years.1,2 Diabetic microvascular com- senior authorship. plications include kidney disease (nephropathy) Correspondence: Dr. Phillip Kantharidis, Department of Diabetes, as well as neuropathy and retinopathy.3 Diabetic Central Clinical School, Monash University, Clayton, VIC, Australia or Dr. Eoin Brennan, UCD Diabetes Complications Research Centre, kidney disease (DKD) is the leading cause of UCD Conway Institute of Biomolecular and Biomedical Research, UCD ESRD, affecting approximately 30% of patients School of Medicine and Medical Sciences, University College Dublin, with long-standing type 1 and type 2 diabetes, Ireland. Email: [email protected] or [email protected] and is characterized by proteinuria and gradual Copyright © 2018 by the American Society of Nephrology J Am Soc Nephrol 29: 1437–1448, 2018 ISSN : 1046-6673/2905-1437 1437 BASIC RESEARCH www.jasn.org loss of kidney function.4,5 There is now a growing appreci- Significance Statement ation for the role of chronic low-grade inflammation as a common pathogenic mechanism. Evidence from clinical Impaired resolution of inflammation underlies chronic conditions, studies suggests that inflammatory markers including including microvascular complications of diabetes such as diabetic TNF-a, IL-1, IL-6, and monocyte chemoattractant protein kidney disease (DKD). Lipoxins (LXs) are lipid mediators that pro- mote the resolution of inflammation. Here, we investigated the 1 (MCP-1) are elevated in patients with diabetes (type 1 and potential of LXA4 and a synthetic LX analogue (Benzo-LXA4) as type 2), and that their levels may predict the onset and pro- therapeutics in a murine model of DKD (streptozotocin diabetic 2 2 gression of diabetic complications.6–11 ApoE / mouse). The development of diabetes-induced albumin- The failed resolution of inflammation may be a major uria, mesangial expansion, and collagen deposition was attenuated fi fi contributor to the pathogenesis of diabetes, CVD, and as- by LXs. Kidney transcriptome pro ling de ned a diabetic signature, and LX-mediated transcriptome responses. Using human renal 12–15 fi sociated complications. The identi cation of endoge- epithelial cells, we demonstrate that LXs attenuate Egr-1 activation. nous mediators that promote resolution, including lipids These data demonstrate for the first time that LXs can reverse es- and cytokines, provides a template for potential mimicry tablished diabetic complications and support a therapeutic para- and avoiding the complications associated with chronic digm to promote the resolution of inflammation. 12,16–25 anti-inflammatory approaches. Lipoxin A4 (LXA4) fl is an eicosanoid generated during acute in ammatory re- In Vitro Studies fl sponses that promotes the resolution of in ammation via Immortalized human kidney epithelial cells (HK-2; ECACC, the G-protein coupled receptor formyl peptide receptor 2 Porton Down, UK) were cultured at 37°C in a humidified 19,26,27 (ALX/FPR2). Here, we have explored the potential of atmosphere of 95% air/5% CO2, and maintained in DMEM- LXA4 and a synthetic analog, Benzo-LXA4,tomodulate F12 (Sigma-Aldrich, Steinheim, Germany) supplemented with DKD in a murine model (streptozotocin [STZ]-treated 2 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomy- 2/2 ApoE mice). This model combines the genetic deletion cin, 10 ng/ml epidermal growth factor, 36 ng/ml hydrocortisone, of apo E, leading to hyperlipidemia, with STZ-mediated and 3 pg/ml triiodothyronine; and 5 mg/ml insulin, 5 mg/ml b pancreatic islet cell ablation, leading to diabetes, ulti- transferrin, and 5 ng/ml selenium solution (Sigma-Aldrich). Af- mately driving a more severe and accelerated renal in- ter serum restriction for 24 hours, cells were stimulated with ve- jury.28–36 We report that LXs attenuate the development hicle (0.1% ethanol), LXA4 (0.1 nM), or Benzo-LXA4 (1 nM) for of proteinuria and glomerular injury in diabetic mice. 30 minutes and media was removed and replaced with media with Most noteworthy are our data that show that LXs reverse or without TGF-b1(10ng/ml;PromoCellGmbH). established renal damage. We have investigated the under- lying mechanisms for these responses via high-throughput fi sequencing of renal tissue and have identified specific Renal RNA-Seq Pro ling networks of gene expression induced by diabetes and sus- Detailed methods are available on RNA-seq and bioinfor- fl ceptible to regulation by LXs including Egr-1. matic analysis in the Supplemental Material. Brie y, RNA was isolated from kidney Trizol homogenates using the Direct-zol RNA MiniPrep Kit (Zymo Research; n=6 kid- METHODS neys per treatment group). Normalized read counts for all genes in all samples are detailed in Supplemental Table 9. Tran- , In Vivo Preclinical Studies scripts with a false discovery rate (FDR) P value 0.05 were 2 2 fi Six-week-old ApoE / male mice (C57BL/6 background) deemed statistically signi cant. Upstream regulator analysis of were rendered diabetic by five daily intraperitoneal (ip) differentially expressed gene sets was performed using the Inge- injections of STZ (Sigma-Aldrich, St Louis, MO) at a nuity Pathway Analysis Z-score algorithm (Qiagen). Analysis of 2 2 dose of 55 mg/kg. ApoE / mice were administered either promoters for enriched TFBSs was performed using Genomatix Matbase (Genomatix). RNA-seq data are deposited at the Gene ethanol (0.1%), LXA4 (5 mg/kg;Merck,Calbiochem),or Expression Omnibus (GSE107942). Benzo-LXA4 (1.7 mg/kg; synthesized at University College Dublin, Ireland37) twice weekly by ip injection. For the prevention study design, mice were followed for 10 or Statistical Analyses 20 weeks, and were administered ethanol, LXA4,or All statistical analyses were performed utilizing GraphPad Benzo-LXA4 between weeks 1 and 10 or weeks 1 and 20, Prism software. Experiments with only one treatment were respectively. For the intervention study design, mice were assessed by t test. Experiments with multiple treatment followed for 16 weeks, and were administered ethanol, groups were analyzed by one-way ANOVAwith post hoc com- LXA4, or Benzo-LXA4 between weeks 10 and 16. Blood glu- parisons of group means performed by Fisher’s least signifi- cose levels were monitored weekly after STZ injections cant difference method. A P value #0.05 was considered for the duration of the studies