New Frontiers: Approaches to Understand the Mechanistic Basis of Renal Toxicity

Abstract

A scientific session entitled “New Frontiers: Approaches to Understand the Mechanistic Basis of Renal Toxicity” focused on novel biomarkers to monitor kidney injury both preclinically and clinically, as well as providing mechanistic insight of the induced injury. Further, the role and impact of kidney membrane transporters in drug-induced kidney toxicity provided additional considerations when understanding kidney injury and the complex role of drug transporters in either sensitivity or resistance to drug-induced injury. The onset of nephropathy in diabetic patients was also presented, focusing on the quest to discover novel biomarkers that would differentiate diabetic populations more susceptible to nephropathy and renal failure. The session highlighted exciting new research areas and novel biomarkers that will enhance our understanding of kidney injury and provide tools for ensuring patient safety clinically.

The scientific session began with a presentation on urinary microRNAs (miRNAs) to monitor kidney injury and the potential to provide mechanistic insight of such injury by Dr. Vishal Vaidya from Pfizer. The second presentation by Dr. Joseph Polli from GlaxoSmithKline described the role and impact of kidney membrane transporters in drug-induced kidney toxicity. The third presentation from Dr. Andrzej Krolewski provided promising data monitoring progressive kidney injury in diabetic nephropathy.

The first speaker, Dr. Vishal Vaidya, discussed the alteration of a number of miRNAs in kidney injury and the efforts to translate research findings to clinical patients in his talk entitled Urinary MicroRNAs as Biomarkers of Kidney Damage. miRNAs are short, noncoding RNAs that regulate a multitude of genes at the posttranscriptional level (Saikumar, Ramachandran, and Vaidya 2014). They have been shown to play a critical role in regulating molecular mechanisms of mammalian biology including the maintenance of normal homeostasis and the regulation of disease pathogenesis. In recent years, there has been substantial progress in innovative techniques to measure miRNAs along with advances in targeted delivery of agents modulating their expression. This has expanded the scope of miRNAs from being important mediators of cell signaling to becoming viable quantitative biomarkers and therapeutic targets (Gerlach and Vaidya 2017). Certainly, the ability to isolate miRNAs from a variety of bodily fluids, including blood and urine, and their apparent remarkable stability in such fluids along with structural conservation across species supports their use as translational biomarkers of disease (Saikumar, Ramachandran, and Vaidya 2014). Dr. Vaidya discussed miR-21 as a prototypical example of a miR serving as both a biomarker of kidney injury (Saikumar et al. 2012; Ramachandran et al. 2013) and a therapeutic target of kidney damage. Upon kidney damage, miR-21 is upregulated in a variety of kidney cells and it has been shown to decrease lipid oxidation, deregulate metabolic processes, and increase the production of reactive oxidation species, all contributing to kidney injury and fibrosis (Chau et al. 2012). It can also be measured in urine supernatant or sediment as a biomarker of kidney injury. Its role in kidney disease pathogenesis led to a clinical trial of a miR-21 inhibitor, RG-012, in patients with Alport syndrome (Gomez, Nakagawa, and Duffield 2016).

Dr. Vaidya then discussed the discovery of miRNAs involved with acute kidney injury (AKI). He first presented data comparing miRNA expression in normal kidneys between humans and mice (Pavkovic and Vaidya 2016), showing a large overlap in the most highly expressed miRNAs, with 12 of the top 20 common between the species (Pavkovic and Vaidya 2016). In particular, miR-10b-5p is highly expressed in the normal kidney of both species. With acute tubular necrosis, the miRNA expression pattern in the kidney changes dramatically, with the majority of miRNAs being downregulated. Target prediction and pathway analysis revealed a variety of affected pathways in AKI, including glucocorticoid receptor, IL-8, PTEN, and STAT3 signaling pathways (Pavkovic and Vaidya 2016). To identify urinary miRNAs as biomarkers for AKI, miRNA profiling was performed on urine samples from patients with AKI and compared with samples from healthy patients. Expression levels of miR-21, 200c, 423, and 4640 were significantly different in AKI patients. Comparing the ability of these miRs to identify AKI, urinary miR-21 individually had the greatest ability to distinguish between disease and control, with the combination of all four having the greatest distinguishing power, with an area under the receiver operator curve of .91 (Ramachandran et al. 2013). While urinary miRNAs are promising as noninvasive biomarkers of kidney disease, one limitation of measuring miRNAs in biofluids by quantitative PCR is determining which normalizer to use, as there are no uniformly accepted miRs to use for this purpose (Pavkovic et al. 2016).

Next, Dr. Vaidya discussed an approach to identify miRNAs specific to certain kidney diseases. In the discovery phase, pooled samples from 10 patients each with diabetic nephropathy and lupus nephritis along with 10 healthy controls were used to identify promising miRs. Samples from a larger number of patients were then used for the confirmation and validation phases, resulting in 5 promising miRNAs detected as altered in diabetic nephropathy (miR-1915-3p, miR-2861, miR-4532, miR-4536-3p, and miR-6747-3p) and 4 for lupus nephritis (miR-1273e, miR-204-5p, miR-30c-5p, and miR-3201). These miRs correlated with glomerular filtration rate and proteinuria and several were associated with histopathologic findings on kidney biopsies, including fibrosis (Cardenas-Gonzalez et al. 2017).

Last, Dr. Vaidya presented data demonstrating that miR-155 knockout mice were more susceptible to cisplatin-induced kidney injury (Pellegrini et al. 2014). Additionally, in both folic acid and unilateral ureteral obstruction models of kidney disease, miR-18a-5p was associated with acute injury, miR-132-3p with intermediate disease, and miR-146b-5p with fibrosis, supporting that these miRNAs are likely playing a distinct mechanistic role in kidney disease progression with respect to regulating injury, inflammation, dedifferentiation/regeneration, and fibrosis (Pellegrini et al. 2016). Overall, this talk summarized the methodology and application for the use of urinary miRNAs to identify and predict kidney damage in humans. Results from several nonclinical and clinical studies highlighted the distinct dysregulated miRNA signatures to differentiate forms of kidney damage such as AKI, lupus nephritis, and diabetic nephropathy.

In the next talk, Dr. Joseph Polli presented The Role of Kidney Membrane Transporters in Drug Development and Impact on Drug-induced Kidney Toxicity. Over the past decade, there have been significant advances in the understanding of membrane transporters and their role in the absorption, distribution, metabolism, excretion, and toxicity (ADMET) of drugs. Transporters can be an effective barrier to drug exposure, a source of drug–drug interactions, or influence organ toxicity through accumulation of a drug into the tissue. Despite the recent advances in characterizing transporter functions and the availability of commercial reagents/assays, insights into the mechanistic role these proteins play in kidney toxicity remains a challenge for drug metabolism and pharmacokinetics (DMPK) and safety scientists. Dr. Polli began with a review on the expression of drug transporters starting from mRNA through recent work quantifying transporter protein levels by mass spectrometry. Most of the major ADMET transporters studied during drug development are expressed in the kidney (>50 different transporters; Bleasby et al. 2006; Prasad et al. 2016), which includes members of the ATP binding cassette (ABC) and solute carrier family (SLC). There are notable tissue-specific expression patterns, as well as species differences. In human kidney, OCT2, OAT1, OAT2, OAT3, Pgp, OATP4C1, and MATE are expressed at high levels, whereas BCRP, OATP1B1, OATP1B3, and BSEP are expressed at much lower amounts. Species differences around BCRP were also highlighted as rat kidney has 40-fold higher BCRP protein expression compared to human kidney (Fallon et al. 2016). Further, the speaker illustrated age- and disease-related changes in expression for several transporters. Recent research demonstrated three age-related (-2 to 850 days postbirth) kidney mRNA expression patterns in rats (Xu et al. 2017). Other work presented highlighted changes in the expression/regulation of transporters by disease states such as AKI (Kunin et al. 2012).

The next portion of the talk provided three case studies investigating the underlying mechanisms of kidney toxicity or drug interactions with respect to transporter involvement. The first was a drug interaction between bupropion and digoxin, where digoxin renal clearance was increased 1.8-fold following dosing with bupropion (He et al. 2014). The mechanism was activation of OATP4C1 on the apical membrane to enhance digoxin renal secretion, and thus overall clearance. The second case study highlighted recent reports of several drugs, such as cobisistat, that increase serum creatinine due to a “drug interaction” with the OCT2/MATE1 transporters in the kidney (Chu et al. 2016). The final example was an investigative safety study in rats to determine the potential role of the organic cation transporter (OCT) transporters in the acute renal toxicity of a series of discovery compounds. The specific compound highlighted had multiple positive charges, a large molecular weight (>500), and clogP near 1.0. However, the clogD at all physiological pH values was negative (ranging from −3.4 to −9.6), suggesting the compound had limited membrane permeability. Interestingly, the compound causes minimal-to-moderate dilation of cortical and medullary tubules, degeneration of the medullary collecting ducts, and mild multifocal degeneration of the proximal convoluted tubules within 24 hr following a single intravenous dose. The investigative study attempted to use cimetidine as an inhibitor of OCT to block the renal uptake of the compound and protect the kidney from toxicity. The work demonstrated that cimetidine pretreatment was unable to alter the toxicity. Thus, it was concluded that OCT transporters were not involved.

Dr. Polli’s takeaway message was that multiple members of the ABC and SLC family are likely to be involved in the ADMET properties of a drug. The interdependency of transporters and their role in drug disposition and toxicity is complex. These proteins can determine the tissue-specific distribution patterns and accumulation in tissues for a drug, which can influence the molecule’s efficacy and/or toxicity. Thus, it is essential to identify key drug transporters and to elucidate their roles in the disposition and toxicity of a compound during drug development. Understanding the mechanistic basis of the toxicity is central to determining how to optimize drug dosing, patient selection, or the decision to select another treatment therapy.

The last speaker, Dr. Andrzej Krolewski, presented a summary of his experience and some of his new findings in predicting renal function decline in patients with diabetic nephropathy in his talk entitled Monitoring Progressive Renal Decline in Diabetes. Based on over 25 years of observational studies performed at the Joslin Diabetes Clinic, he proposed a new model of diabetic nephropathy in type 1 diabetes (T1D). In this model, the predominant clinical feature of both early and late stages of diabetic nephropathy is progressive renal decline, independent of albuminuria (Krolewski 2015). Progressive renal decline (estimated glomerular filtration rate [eGFR] loss >3.5 ml/min/year) is an unidirectional natural progression that develops while patients have normal renal function. It declines progressively at a steady rate until end-stage renal disease (ESRD) is reached, albeit at widely differing rates among individuals. There is an urgent need to develop prognostic tests to identify patients at risk for progression to ESRD early in their disease, when they still have normal renal function. A result of adopting progressive renal decline as a new model of diabetic nephropathy will be the need to develop new interventions to prevent or retard progression to ESRD. A corollary of the latter is development of tools to monitor progressive renal decline so the effectiveness of specific interventions can be evaluated during clinical trials as well as by physicians during routine clinical care.

Dr. Krolewski presented data obtained over the last 5 years to develop such tests and tools. Subjects enrolled in these Joslin Kidney Studies were followed for 5 to 15 years, eGFR slopes were monitored, and time of onset of ESRD was documented. Blood and urine specimens were obtained at baseline and during follow-up in order to identify relevant biomarkers. Two proteomics platforms were utilized—SOMASCAN that measured 560 proteins and OLINK that measured 720 proteins. Several dozen novel proteins were recognized as potential prognostic biomarkers for monitoring progressive renal decline; however, many were strongly intercorrelated and provided redundant information. Ultimately, after vigorous sensitivity testing, a dozen proteins were identified as risk markers associated with development of ESRD and progressive renal decline in T1D subjects. These proteins were also validated in type 2 diabetes subjects. These findings have significant implications for considerations on the mechanisms of progressive renal decline in diabetes.