Cellular and Functional Biomarkers of Renal Injury and Disease

Abstract

Biomarkers, defined as molecules in biological samples that are used as indicators of organ function, can assess exposure to potentially injurious chemicals, effects on organ function, or susceptibility to organ functional decline. The kidneys are frequently exposed to many drugs and chemicals and loss of kidney function is a frequent consequence of diseases such as diabetes. This review summarizes findings reported in 2021 and early-2022 from clinical and experimental animal studies on biomarkers, focusing on five topics: 1) Progression and severity of diabetic kidney disease; 2) acute kidney injury (AKI) and chronic kidney disease (CKD) severity and prognosis; 3) progression of AKI to CKD; 4) renal cell carcinoma (RCC) severity and prognosis; and 5) detection of exposure to environmental chemicals and nephrotoxic drugs.

1. Introduction

The focus of this review is on current efforts to identify and use biomarkers for assessment of kidney function declines due to diseases or exposures to environmental toxicants or therapeutic drugs whose efficacy is dose-limited by nephrotoxicity. First, it is important to provide some definitions for context. Terms or processes to be defined include biomarker, acute kidney injury (AKI), and chronic kidney disease (CKD). It is also important to appreciate the similarities in renal cellular responses to pathological or disease states and exposures to nephrotoxic drugs or other chemicals. Thus, although there will be considerable focus on pathological states such as diabetic kidney disease (DKD), information gained from such work will have relevance and provide insights to biomarkers for exposure to nephrotoxicants.

What is a biomarker? A biomarker is a molecule found in a biological sample, such as urine, plasma or serum, or tissue, that is an indicator of some aspect of a biological response. Biomarkers may include proteins whose levels change in abundance or are covalently or chemically modified (e.g., by oxidation), lipids whose levels change in abundance or whose patterns change, and various low-molecular-weight metabolites that may change in abundance. Here, the focus is on the kidney. There are three types of biomarkers: Biomarkers of effect, biomarkers of exposure, and biomarkers of susceptibility. The most common and readily obtainable biomarkers are those of effect. By definition, these biomarkers are molecules that change in abundance or are modified in response to a change in organ function. Biomarkers of exposure are those that identify or confirm that the tissue of interest has been exposed to a specific chemical of interest. Biomarkers of susceptibility, in contrast to the two other categories of biomarkers, are more diverse and are often used to indicate an enhanced sensitivity to the injurious effects of a pathological condition or chemical exposure. These biomarkers may include things such as genetic polymorphisms, comorbid diseases or chronic conditions, and social or dietary habits, in addition to proteins or metabolites.

The standard or working definition for AKI is based on short-term changes in serum creatinine (SCr) and is an increase of > 0.3 mg/dL within 48 h or > 50% in a 7-d period. Fuentes-Calvo et al. [1] note that the use of SCr to define AKI is primarily associated with the process of glomerular filtration and is not directly linked to renal cellular injury or tissue damage. Thus, although SCr levels may be viewed as a biomarker, it can be said to be more of an operational or functional assessment of kidney function. Whereas many forms of AKI are reversible, there is some evidence that even incidences of relatively mild and reversible AKI can increase risk for both subsequent occurrences of AKI and development of long-term declines in kidney function, known as CKD. Therefore, identification and validation of biomarkers that can indicate AKI, especially in an early phase, are important for improving public health.

CKD can be defined by the Kidney Disease: Improving Global Outcomes (KDIGO) criteria as an abnormality in kidney structure or function that persists for at least 3 months and has health implications. As summarized by Chebotareva et al. [2], the three most common causes of CKD are diabetes mellitus (either Type 1 or Type 2), hypertension, and glomerulonephritis. Major forms of CKD thus include diabetic nephropathy (DN), immunoglobulin A nephropathy (IgAN), lupus nephritis (LN), focal segmental glomerulosclerosis (FSGS), and membranous nephropathy (MN). CKD is a major public health concern, with a global incidence of 11–13%, and is an independent risk factor for death, cardiovascular disease, and end-stage renal disease (ESRD). Thus, identification of biomarkers for CKD, especially those that may be used to define the stage of the disease, are very useful in defining potential treatments and improving outcomes.

The first goal of this review and commentary is to review recent developments in the identification and validation of various biomarkers for detection of either the incidence or severity of various forms of kidney injury or disease and the progression of mild to more serious or chronic disease. The second goal is to review recent developments in the identification and validation of biomarkers for detection and prognosis of various forms of renal cell carcinoma (RCC). Finally, the third goal is to discuss recent work the author and colleagues have done on identifying and validating biomarkers for exposure to diverse environmental pollutants and therapeutic drugs whose efficacy is dose-limited by nephrotoxicity. Studies that have been identified in a literature search of kidney biomarkers include some reviews and meta-analyses, but also primary research studies in both humans and animal models and are all published in calendar years 2021 and early-2022. While the list of studies reviewed is by no means complete, the large number of publications in a 14-month period (January 2021 – February 2022) indicates the interest in and translational importance of work on kidney biomarkers.

In terms of quantitative criteria for validation of biomarkers, there are no established standards specific for function as a biomarker other those noted above for SCr or blood urea nitrogen (BUN), which are routinely measured in clinical settings. For novel experimental biomarkers such as proteins or metabolites, one would typically consider as significant a difference of 1.5- or 2.0-fold (if being more conservative) as a threshold for concluding a potential association between altered concentration and some effect or exposure. Such a threshold is standard in the proteomics and metabolomics fields and is not specific to evaluation of potential biomarkers.

Figure 1 illustrates the underlying mechanisms of various forms of renal injury and the pathways at which various biomarkers may be useful in either identifying exposure, identifying injury, identifying severity of injury, or predicting the risk or likelihood of progression from mild injury to more serious injury, such as CKD or even ESRD.

Figure 1.

Overview of pathways to renal injury and sites for biomarkers.

The scheme illustrates the four major pathways that result in renal injury. The etiologies of renal damage include disease, therapeutic drugs, environmental pollutants, and cancer. Diseases that are major causes of chronic kidney disease (CKD) include diabetic nephropathy (DN), immunoglobulin A nephropathy (IgAN), lupus nephritis (LN), focal segmental glomerulosclerosis (FSGD), and membranous nephropathy (MN). Therapeutic drugs whose efficacy is dose-limited by nephrotoxicity include cisplatin (CDDP), mycotoxins, gentamicin (GEN), and non-steroidal anti-inflammatory drugs (NSAIDs). Environmental pollutants that can result in acute kidney injury (AKI) include organic solvents and heavy metals, among many others. Various forms of renal cell carcinoma (RCC) can also ultimately result in end-stage renal disease (ESRD). Pathways at which biomarkers may be useful in indicating or predicting disease severity are indicated by numbers 1–7. 1 = disease progression to CKD; 2 and 4 = chemical exposures that may result in AKI; 3 = progression of AKI to CKD; 5 = recovery of renal function; 6 = progression of RCC to ESRD.

2. Progression and severity of diabetic kidney disease

One of the most common areas of focus in recent kidney biomarker research involves attempts to identify markers to either determine severity of renal injury in diabetics or prognosis and likelihood of progression from mild to more severe CKD in diabetic patients. The incidence of DKD is such that diabetes is the leading cause of CKD and ESRD, making prevention of disease progression in this population a critical public health issue.

Several clinical studies with plasma and/or urine samples from diabetic patients have used proteomics and metabolomics to identify biomarkers that can be indicators of DKD severity or predictors of progression from mild to more severe CKD. For example, Liu et al. [3] recruited 1,513 participants that included healthy adults, patients diagnosed with type 2 diabetes and early-stage DKD, and patients diagnosed with advanced-stage DKD. As with many biomarker studies, this study found that a combination or battery of biomarkers were the best indicators of disease progression. Here, they report that the combination of α2-macroglobulin, cathepsin D, and CD324 predicted DKD progression. In metabolomic studies, they found that disturbances in galactose and glycerolipid metabolism were the most significantly disturbed metabolic pathways in DKD and concluded that serum concentrations of glycerol-3-galactoside may serve as an independent marker to predict DKD severity.

Using a multi-biomarker algorithm that incorporates plasma biomarkers and clinical parameters, Connolly et al. [4] found that a combination of kidney injury molecule-1 (KIM-1), soluble tumor necrosis factor receptor-1 (sTNFR-1), and soluble tumor necrosis factor receptor-2 (sTNFR-2) were able to produce a sensitive and precise composite risk score to predict progression of kidney function decline in patients with type 2 diabetes and early-stage CKD.

Sauriasari et al. [5] performed a systematic review of 20 research articles that identified protein biomarkers in diabetic patients. Consistent with the finding that a battery of biomarkers is most effective in predicting disease status or likelihood of disease progression, they identified a panel of glomerular biomarkers (angiopoietin like 4 [ANGPTL4], β2-microglobulin, SMAD family member 1, and glypican-5), inflammatory biomarkers (monocyte chemoattractant-1 [MCP-1] and adiponectin), and tubular biomarkers (neutrophil gelatinase-associated lipocalin [NGAL], vitamin D binding protein [VDBP], megalin, sKlotho, and KIM-1) that were promising in diagnosis of DKD. They noted, however, that measurements of albuminuria and estimated glomerular filtration rate (eGFR) had similar prognostic value as did the battery of proteins. Thus, the advantage of using the panel of proteins rather than the more traditional methods of assessing renal function is unclear.

Another patient study using a well-characterized subpopulation of diabetic patients with mild-to-moderate kidney disease in the Chronic Renal Insufficiency Cohort (CRIC) study [6], found that higher plasma levels of KIM-1, TNFR-1, TNFR-2, MCP-1, soluble urokinase-type plasminogen activator receptor (suPAR), and YKL-40 (a heparin- and chitin-binding glycoprotein) were associated with increased risk for progression of DKD. The identification of KIM-1, TNFR-1, and TNFR-2 agrees with the Connolly et al. [4] study. Also, in agreement with these studies, Waijer et al. [7] found that detection of significant increases in plasma concentrations of a combination of KIM-1, TNFR-1, and TNFR-2 were effective biomarkers for CKD progression in patients with type 2 diabetes but normoalbuminuria. Lupusoru et al. [8] also found suPAR concentrations to be associated with duration of diabetes, the occurrence of diabetic retinopathy (another relatively common complication associated with poorly controlled diabetes), renal function, proteinuria, and inflammation. Sánchez-Hidalgo et al. [9] focused their studies on urinary transferrin and found it to be a marker for early endothelial dysfunction in type 2 diabetics who do not yet exhibit nephropathy. In an untargeted, exploratory proteomic study, Fan et al. [10] characterized the urine proteome from diabetic patients with varying degrees of kidney impairment. Patient subgroups included uncomplicated diabetes, DKD patients, and CKD patients without diabetes. A total of almost 3,000 gene products were detected, and use of logistic regression models allowed the authors to differentiate amongst diabetic patients with and without CKD and different stages of CKD.

Metabolomics has also been used to identify patterns of low-molecular-weight, intermediary metabolites in plasma or urine that can serve as sensitive biomarkers. Jin and Ma [11] found that changes in patterns of arginine and tryptophan metabolites were associated with rapid declines in eGFR and increases in albuminuria in type 1 and type 2 diabetic patients. Due to the complexity and variability of the metabolome, it is not surprising that this study identified different metabolites than Liu et al. [3]. Another metabolomic study [12] sought to identify metabolic signatures associated with early-stage DKD, focusing primarily on citric acid cycle and related intermediates. They identified lower concentrations of citric acid cycle organic acids in urine, most notably methyl- and ethylmalonate, as being potential biomarkers for kidney impairment in the early stage of DKD.

Besides proteins and low-molecular-weight metabolites, different RNA species have also drawn attention as potential biomarkers. Lee et al. [13] used the public gene expression omnibus (GEO) repository to identify urinary mRNAs that could predict kidney disease severity and progression in DKD patients with biopsy-proven disease. The authors identified 30 DKD-specific mRNAs and developed a urinary mRNA signature to reflect different degrees of CKD and predict adverse renal outcomes in DKD patients. In recent years, microRNAs (miRNAs) rather than mRNAs have been the more frequent focus of biomarker studies. Liu et al. [14] specifically focused on the relationship between serum concentrations of miR-29a and DKD severity and risk for disease progression in a cohort of 180 type 2 diabetic patients and 180 healthy controls. They found that a combination of serum miR-29a and cystatin C concentrations was associated with occurrence and progression of DKD.

Lei et al. [15] took a different approach in trying to define prognosis for disease progression in more than 1,400 biopsy-proven DKD patients. They compared histologic acute tubular-interstitial injury (hATI) with various tubular injury biomarkers and SCr. While several relationships between parameters and clinically diagnosed AKI were uncovered, the most significant conclusion was that patients with hATI alone were 1.5-times more likely to develop ESRD than those without hATI or AKI; patients with hATI and AKI were 3-times more likely to develop ESRD.

A final recent study on biomarkers for DKD onset or progression that used a completely different approach involved assessment of genetic predisposition [16]. The investigators used a custom target next-generation sequencing 70-gene panel to screen for relevant single nucleotide polymorphisms (SNPs). They identified MYH9 (Myosin p) rd710181 as being inversely associated with DKD risk and RGMA (Repulsive Guidance Molecule BMP Co-Receptor A) rs1969589 CC genotype as correlating with lower albumin-to-creatinine ratios in DKD patients. The authors noted that no single biological pathway stood out as being more affected by genetic variability in DKD, which highlights the complexity of the genetic factors that influence diabetic disease progression.

3. Acute kidney injury (AKI) and chronic kidney disease (CKD) severity and prognosis

As noted above, AKI is defined as a rapid decline in kidney function. AKI may be caused by a large array of chemicals and pathological states and can occur with a range of severity. Whereas clinically used biomarkers are well-established in detecting large declines in kidney function (e.g., as assessed by eGFR, albuminuria, and urine output), there is an urgent need to identify and validate more sensitive biomarkers that can detect losses in kidney function at early stages and prior to these losses becoming large and irreversible. Several recent studies are summarized that have taken an array of approaches and have focused on different classes of molecules as potential biomarkers.

Zhang et al. [17] reviewed the evidence for urinary mitochondrial DNA as a selective and sensitive biomarker for AKI. They conclude that extracellular mitochondrial DNA, which is found in blood, urine, and other tissues, may serve as an indicator of mitochondrial dysfunction and early, sublethal tissue injury. In particular, urinary mitochondrial DNA, which is readily quantified by the polymerase chain reaction and equates with mitochondrial number, may be most closely linked to renal injury.

In addition to the battery of proteins that have been validated by the U.S. Food and Drug Administration (U.S. FDA) as markers for AKI, such as SCr, BUN, KIM-1, NGAL, and cystatin C, among others, some recent work has sought to identify additional, more sensitive protein biomarkers. Although some of the U.S. FDA biomarkers, such as KIM-1, are considered very sensitive and do generally detect mild AKI, others such as SCr, BUN, and cystatin C are mostly useful to identify more advanced AKI.

Sun et al. [18] found that serum concentrations of fibroblast growth factor 23 (FGF23) were associated with AKI and mortality in patients with critical illness. They also performed a meta-analysis of clinical studies from PubMed, Web of Science, EMBASE, and Cochrane of patients with AKI and serum FGF23 measurements. Serum FGF23, most notably the C-terminal fragment, was a sensitive and specific biomarker to predict AKI. Schmidt and colleagues [19] performed a proteomics analysis of plasma from 549 patients with biopsy-confirmed kidney disease and determined association of a panel of putative biomarker proteins with histologic lesions and risk of disease progression, as judged by eGFR decline and initiation of renal replacement therapy (i.e., dialysis). Thirty proteins were significantly associated with AKI progression and another 35 proteins were associated with mortality. The top proteins associated with disease progression were placental growth factor and bone morphogenetic protein while those associated with mortality included TNFR-2 and CUB domain-containing protein-1. It is notable that TNFR-2 was identified in multiple studies described above as a biomarker for progression of DKD.

Urinary proteins have also been the focus of AKI biomarker studies. Woziwodzka et al. [20] assessed the relationship between the urinary proteome and renal tubular injury in multiple myeloma patients. Multiple linear regression analysis found that urinary insulin-like growth factor-binding protein 7 (IGFBP-7) and NGAL monomer were associated with lower eGFR values independent of other urinary proteins. In a review of plasma and urine biomarker proteins in pediatric CKD patients, Sandokji and Greenberg [21] focused on several proteins and reported associations between CKD progression and several plasma proteins, including KIM-1, MCP-1, FGF23, TNFR-1, TNFR-2, suPAR, and chitinase-3-like protein 1, and several urinary proteins, including epidermal growth factor, α1-microglobulin, KIM-1, MCP-1, and chitinase-3-like protein 1. Wendt et al. [22] uniquely focused on urinary complement fragments as potential biomarkers, based on the realization that defective complement activation is often associated with kidney disease. Twenty-three different urinary peptides derived from complement proteins C3, C4, and complement factor B were found to be associated in different ways with renal function. For example, most C3-derived peptides were inversely associated with eGFR whereas most complement factor B-associated peptides were positively associated with eGFR. Such associations may underlie immune system-related processes involved in kidney disease.

A final study that focused on urinary proteins was a systematic review and meta-analysis characterizing the association between urinary uromodulin (UMOD) and AKI [23]. Here, the authors analyzed 11 studies with more than 3,000 subjects and found a strong association between lower urinary UMOD concentrations and higher odds for AKI incidence.

Aomatsu and colleagues [24] focused their efforts at identifying miRNAs as biomarkers for diagnosis of AKI and as aids in development of therapeutics. They performed profiling of miRNA expression in two AKI mouse models, lipopolysaccharide-treated mice that exhibit sepsis and ischemia-reperfusion injured mice. Their profiling identified miR-5100 as being significantly decreased in expression levels in both AKI mouse models. They then administered a miR-5100 mimic to mice and confirmed miR-5100 overexpression in the kidneys and prevention of development of AKI in ischemia-perfusion mice. miR-5100 expression levels were significantly lower in AKI patients than in healthy subjects, suggesting that it may be a diagnostic biomarker in humans. The protective effects of the miR-5100 mimic also suggests that this specific miRNA may also be a therapeutic target for future drug development.

Urinary metabolomic profiles of non-AKI and mild or severe AKI patients showed an inverse correlation with glycine and ethanolamine levels, suggesting that patterns of these metabolites can be used to predict severity of AKI [25].

Two final studies developed novel methods of sample collection and analysis to monitor renal function in patients. In the first study, Beshay et al. [26] monitored creatinine, cystatin C, and urea concentrations in saliva and fingerstick sampling. The authors suggested that these methods, which are simple, cost-effective, and minimally invasive, could fill an important gap in the management of CKD patients. The second study by van Duijl et al. [27] used a liquid chromatography analysis coupled to tandem mass spectrometry in multiple reaction monitoring mode to measure picomolar to nanomolar levels of biomarkers in urine samples. The protein-based biomarker panel included measurements of KIM-1, NGAL, tissue inhibitor of metalloproteinases 2 (TIMP2), IGFBP7, inflammatory chemokine ligand 9, fibrosis marker transforming growth factor β1, nephrin, cubilin, solute carrier family 22 member 2 (SLC22A2; also known as organic cation transporter 2 or OCT2), and UMOD. Besides emphasizing the sensitivity of the method, this paper also illustrates the use of a panel of biomarker proteins, many of which have been identified in other studies, to detect AKI in its early phases.

4. Progression of acute kidney injury (AKI) to chronic kidney disease (CKD)

Incidents of AKI are thought to increase the risks for developing CKD, sometimes much later after recovery from the initial loss of renal function. Understanding the factors that contribute to this susceptibility may provide both improved predictions for the likelihood of someone progressing to CKD or ESRD and identify therapeutic targets. A few recent studies have identified proteins, or more accurately, sets of proteins, that are associated with AKI progression to CKD. For example, Fuentes-Calvo et al. [1] found that high levels of both IGFBP7 and TIMP2 in rat urine were good diagnostic tools for vulnerability to future development of CKD after recovery from AKI due to cisplatin exposure. Likewise, Wilson et al. [28] measured a panel of 11 plasma biomarker proteins in 500 patients 3 months after hospitalization and recovery from AKI. A multivariable model showed that plasma levels of soluble TNFR1, solubleTNFR2, cystatin C, and eGFR were independently associated with kidney disease progression and could clearly distinguish between patients with progressive CKD versus those without CKD.

Govender et al. [29] describe in a scoping review paper the use of different ‘omics approaches to identify biomarkers to detect early-stage CKD, to predict disease progression, and to identify mechanisms that lead from AKI to CKD. They divided the studies (123 papers published from January 2007 to May 2021) into 3 types: Most of the studies involved proteomics, two of them included metabolomics, most of the studies also focused specifically on CKD associated with diabetes, and about 10% of the reviewed studies focused on the potential pathways that lead to CKD. Biomarkers associated with risk or early detection of CKD were SNPs in the MYH9/APO1 and UMOD genes and the metabolite pantothenic acid. Biomarkers that served to predict CKD progression included panthothenic acid, retinoic acid pathway genes, and UMOD gene expression. Their review also emphasized differences amongst various ethnic groups, supporting findings that disease etiologies and incidence varies across ethnic groups.

Daniels et al. [30] quantified an array of serum proteins in AKI patients on dialysis and those who discontinued dialysis to identify biomarkers of kidney functional recovery. The overall conclusions were that kidney regeneration and repair were promoted by or associated with low levels of inflammation-associated proteins (e.g., FGF23, C-X-C Motif Chemokine Ligand 6 [CXCL6]), high levels of chemokines and cytokines for inflammatory tissue repair, and high levels of Wnt-7a, Bruton Agammaglobulinemia Tyrosine Kinase (BTK), c-Myc, ghrelin, platelet-derived growth factor-C, survivin, epidermal growth factor (EGF), and neuroregulin-1. Along the same lines, Sun et al. [31] measured an array of serum metabolites in AKI patients requiring dialysis to try and determine metabolite profiles that could assess risk of mortality or serious CKD. A number of metabolites were identified, such as low levels of two lysophosphocholines in critically ill patients and higher levels of several amino acids and associated metabolites, suggesting muscle wasting.

A retrospective study of 355 nephrology patients who exhibited eGFR values < 60 ml/min per 1.73 m² was conducted to examine the relationship between urinary and serum NGAL to distinguish between patients with AKI and those with CKD [32]. The ratio of urinary to serum NGAL and the fractional excretion of NGAL were shown to discriminate between AKI and CKD, with values being higher in AKI patients.

5. Renal cell carcinoma (RCC) severity and prognosis

Renal cell carcinoma (RCC) is often symptomless until it is at an advanced stage. Although management and therapeutic approaches have improved over the past two decades, many patients do not show good responses. Moreover, while RCC has been biologically characterized and sub-classified, few therapeutic targets are known. The majority of cases of kidney cancer are clear cell RCC (ccRCC), making up approximately 75–80% of all cases, with the remainder being classified as non-clear cell RCC. Gulati and Vogelzang [33] present a review of the state of RCC biomarkers. Their focus is primarily on genetic markers that can correlate with severity of illness and mortality. Historically, they note that programmed death ligand 1 (PD-L1) status was first described in 2004 as being highly associated with mortality risk and discuss other genes, such as von Hippel Lindau (VHL). More recent work has focused on development of gene expression signatures that can correspond to risk. The authors describe a model for predicting RCC prognosis that includes the mutation status of several tumor suppressor or transcriptional regulators: BRCA1 associated protein-1 (BAP1), polybromo-1 protein (PBRM1), tumor protein p53 (TP53), telomerase reverse transcriptase (TERT), lysine demethylase 5C (KDM5C), and SET domain containing 2 / histone lysine methyltransferase (SETD2).

Another approach to biomarkers for RCC progression is described in a study that assessed DNA methylation status [34]. The investigators performed a genome-wide methylation analysis and found that hypermethylation of zinc finger protein 677 (ZNF677; involved in transcriptional regulation) and procadherin 8 (PCDH8; involved in cell adhesion and epithelial-to-mesenchymal transition) in kidney tissue samples was significantly related to poor clinical outcome. A recent study by Gui et al. [35] focused on circular RNAs (circRNAs), some of which have been suggested to function as mediators of tumor progression via miRNA sponging. Expression of circCGST15 was higher in ccRCC tissues than in healthy adjacent kidney tissue and was higher in cell lines derived from RCC than in normal kidney cell lines. A different type of RNA, long non-coding RNA (lncRNA), was the focus of work by Xing et al. [36] and Shu et al. [37], who looked at identifying and validating ferroptosis-related lncRNAs as potential biomarkers to detect and predict the outcome of ccRCC. Three lncRNAs were identified in one study by the group [36] as being significantly correlated with overall survival of ccRCC patients: Double Homeobox A Pseudogene 8 (DUXAP8), Long Intergenic Non-Protein Coding RNA 2609 (LINC02609), and (LUCAT1); the other study [37] identified five lncRNAs as being predictors of overall survival: DOCK8-AS1, SNHG17, RUSC1-AS1, LINC02609, and LUCAT1.

A final study [38] is rather unique in specifically looking at biomarkers for non-clear cell RCC, which accounts for approximately 25% of all RCC cases. The authors enrolled 108 patients in ASPEN, an international randomized phase 2 clinical trial of patients with the various types of non-clear cell RCC who were on therapy with either the mammalian target of rapomycin (mTOR) inhibitor everolimus or the vascular endothelial growth factor (VEGF) receptor inhibitor sunitinib. They focused on tissue biomarkers of mTOR pathway activation (i.e., phospho-S6 and phospho-Akt, c-kit) and VEGF pathway activation (i.e., hypoxia inducible factor-1α [HIF-1α], c-MET). While some associations between levels of activation of these pathway components and overall survival were observed, they could not conclude that any biomarker was appropriately predictive for clinical outcomes.

6. Detection of exposure to environmental chemicals and nephrotoxic drugs

As discussed in the Introduction, there are three types of biomarkers. Thus far, discussion has focused on biomarkers of effect and biomarkers of susceptibility. Biomarkers of exposure, more so than the other two types, may hold the promise of identifying conditions that are known to result in renal functional, biochemical or molecular changes before any significant injury occurs.

In a recent book chapter [39] and review [40], the author and colleagues proposed that mitochondria may serve as sentinels for exposures of the kidneys to environmental toxicants and nephrotoxic drugs. The basis for this hypothesis is that molecular events in renal mitochondria are early steps in the sequelae of events that occur upon exposure of the kidneys to a diverse array of chemicals and drugs. Figure 2 illustrates a schematic representation of this hypothesis. The translational aspect of this hypothesis is that cellular molecules released from exposed renal proximal tubular cells are released into either the tubular lumen or the renal interstitial space and are recovered in either urine or plasma, respectively. The hypothesis further states that the major sources of these molecules will be the mitochondria or plasma membranes. Such molecules may, therefore, be used as biomarkers for the initiating exposure.

Figure 2.

Model for early exposure of renal proximal tubular cells to chemicals resulting in release of putative biomarkers.

The scheme illustrates the hypothesis by Lash and Stemmer [39] and Lash [40] that early exposure of renal proximal tubular cells releases proteins, lipids, RNA species, and low-molecular-weight metabolites from the cells into both the tubular lumen and renal interstitial space. In vivo, the former set of molecules is recovered in urine while the latter is recovered in plasma. Primary sources of these molecules, which may serve as biomarkers of exposure, are mitochondria and the plasma membrane.

Evidence in support of the hypothesis include the predominance of mitochondria-derived proteins in the extracellular media of human proximal tubular cells exposed to a well-characterized mitochondrial and nephrotoxic environmental contaminant [39]. Primary cultures of human proximal tubular cells were exposed to the penultimate nephrotoxic metabolite of the environmental pollutant trichloroethylene. Analysis of proteins recovered in extracellular media by tandem mass tag (TMT) multiplexing and liquid chromatography and tandem mass spectrometry (LC/MS/MS) identified 311 unique proteins, of which 43% were derived from the mitochondrial fraction. Further analysis showed exposure time- and concentration-dependent increases in multiple mitochondrial proteins and keratins. Although the studies and exposures shown in these publications [39, 40] are “proof-of-concept” studies with chemical concentrations that also produce cytotoxicity, more recent studies with exposures to subtoxic concentrations of the trichloroethylene metabolite and other nephrotoxic chemicals have revealed similar recovery of a predominance of mitochondria- and plasma membrane-derived proteins [L.H. Lash, P.M. Mathieu, R. Rosati, and P.M. Stemmer, unpublished data).

As noted in the Introduction (section 1), many of the same biomarkers that have been described for disease states or decreased renal function associated with pathological processes such as ischemia-reperfusion or sepsis, are also useful for AKI or CKD due to exposures to nephrotoxic drugs or environmental pollutants. The critical difference in the studies described here is that the putative biomarkers identified are chosen to indicate exposure to a potentially nephrotoxic drug or chemical rather than to renal injury resulting from that exposure.

7. Conclusions

This review has highlighted recently published (2021 and early-2022) papers that have identified different classes of biomarkers of kidney function. Reviewed studies have included several case-control and case-cohort studies, data from clinical trials, experimental studies with human tissue samples, and studies in animal models of renal disease. Figure 3 presents a schematic summary of how biomarkers can be used at multiple stages of renal injury and disease to inform on nephrotoxicant exposure, disease severity, or renal injury severity or prognosis. Although this figure is conceptually similar to Figure 1, it focuses primarily on the use of biomarkers in assessing renal function whereas Figure 1 is focused primarily on the causes of renal injury and disease.

Figure 3.

Summary scheme of use of biomarkers highlighted in this paper.

Both chemical or drug exposures, disease onset (e.g., diabetic nephropathy), and renal cell carcinoma (RCC) can result in sublethal functional changes in the kidneys. While recovery can occur, continued exposure or disease progression may lead to episodes of acute kidney injury (AKI). Such AKI episodes may progress to chronic kidney disease (CKD) and eventually to end-stage renal disease (ESRD). Biomarkers, which may include proteins, lipids, different RNA species (e.g., miRNAs), and low-molecular-weight metabolites, have been identified to identify exposure or disease status, assess severity of AKI or CKD, and risk for development of CKD or ESRD. Other biomarkers, such as gene expression patterns or specific genetic polymorphisms, have also been identified as predictors of susceptibility.

One of the striking findings from the studies reviewed here is the diversity in the array of biomarkers described. Table 1 lists alphabetically all the biomarkers that were identified and characterized in the reviewed studies. Although some biomarkers were identified in multiple studies, there are numerous unique biomarkers that may suggest these are related more to unique disease or exposure situations rather than as more broadly applicable biomarkers. While proteins are the most common class of biomarkers, changes in concentrations or patterns of metabolites have also been associated with different stages or prognosis of renal disease. In recent years, different classes of RNA molecules, such as miRNAs and lncRNAs, have attracted considerable interest as potential biomarkers of renal function.

Table 1.

Alphabetical list of potential biomarkers used to detect various forms of kidney injury or disease and predict the likelihood or risk of disease progression. The “Source” is the sample type in which each agent has been measured. “Type” is the molecular classification of the biomarker. “Use” is the purpose for which the biomarker has been shown to function.

Biomarker	Source	Type	Use	Reference(s)
α1-microglobulin	Urine	Protein	AKI severity	[21]
α2-macroglobulin	Plasma, Urine	Protein	DKD progression	[3]
Adiponectin	Plasma, Urine	Protein	DKD progression	[5]
ANGPTL4	Plasma, Urine	Protein	DKD progression	[5]
Arginine metabolites	Plasma, Urine	Metabolites	DKD severity	[10]
β2-microglobulin	Plasma, Urine	Protein	DKD severity, AKI severity – progression	[5]
BAP1	Tissue	Gene	RCC progression	[33]
BTK	Serum	Protein	Renal repair	[30]
Cathepsin D	Plasma, Urine	Protein	DKD progression	[3]
c-Myc	Serum	Protein	Renal repair	[30]
CD324	Plasma, Urine	Protein	DKD progression	[3]
circCHST15	Tissue	RNA	RCC	[35]
Chitenase-3-like protein	Plasma, Urine	Protein	AKI severity	[20]
Complement C3, C4, Factor B	Urine	Peptides	AKI severity	[22]
Cubilin	Urine	Protein	AKI severity	[27]
CXCL6	Serum	Protein	Renal repair	[30]
Cystatin C	Plasma or Serum, Saliva	Protein	DKD severity, CKD progression	[14, 26, 28]
DUXAP8	Tissue	RNA	RCC	[36, 37]
EGF and neuroregulin-1	Sereum	Protein	Renal repair	[30]
Ethanolamine	Urine	Metabolite	AKI severity	[25]
FGF23	Serum, Urine	Protein	AKI severity, Renal recovery	[18, 29, 30]
Fibronectin marker transforming growth factor B1	Urine	Protein	AKI severity	[27]
Glycerol-3-galactoside	Serum	Metabolite	DKD severity	[3]
Glycine	Urine	Metabolite	AKI severity	[25]
Ghrelin	Serum	Protein	Renal repair	[30]
IGFBP-7	Urine	Protein	AKI severity	[1, 20, 27]
Inflammatory chemokine ligand 9	Urine	Protein	AKI severity	[27]
KDM5C	Tissue	Gene	RCC prognosis	[33]
KIM-1	Plasma, Urine	Protein	DKD, CKD progression	[4–7, 20, 21, 27]
LINC02609	Tissue	RNA	RCC	[36, 37]
LUCAT1	Tissue	RNA	RCC	[36, 37]
Lysophosphocholines	Serum	Metabolites	CKD severity, progression	[31]
MCP-1	Plasma, Urine	Protein	DKD severity – progression, AKI severity	[5, 6, 20, 21]
Megalin	Plasma, Urine	Protein	AKI severity	[5]
Methyl- and ethylmalonate	Urine	Metabolite	DKD severity	[12]
miR-29a	Serum	miRNA	DKD severity	[14]
miR-5100	Tissue	miRNA	AKI severity – progression	[24]
mtDNA	Urine	DNA	AKI severity	[17]
MYH9	Tissue	Gene polymorphism	DKD severity, CKD progression	[16]
Nephrin	Urine	Protein	AKI severity	[27]
NGAL	Plasma or Serum, Urine	Protein	AKI severity	[5, 20, 27, 32]
Pantothenic acid	Urine	Metabolite	CKD progression	[29]
PBRM1	Tissue	Gene	RCC prognosis	[33]
Platelet-derived growth factor-C	Serum	Protein	Renal repair	[30]
Retinoic acid pathway	Tissue	Genes	CKD progression	[29]
RGMA	Tissue	Gene polymorphism	DKD severity, CKD progression	[16]
SETD2	Tissue	Gene	RCC prognosis	[33]
sKlotho	Plasma, Urine	Protein	AKI severity	[5]
SLC22A2	Urine	Protein	AKI severity	[27]
sTNFR-1, sTNFR-2	Plasma, Urine	Protein	DKD progression, AKI severity	[4, 6, 7, 19–21, 28]
suPAR	Plasma, Urine	Protein	AKI severity	[6, 8, 20, 21]
Survivin	Serum	Protein	Renal repair	[30]
TERT	Tissue	Gene	RCC prognosis	[33]
TIMP2	Urine	Protein	AKI severity	[1, 27]
TP53	Tissue	Gene	RCC prognosis	[33]
Transferrin	Urine	Protein	DKD severity	[9]
Trp metabolites	Plasma, Urine	Metabolites	DKD severity	[10]
UMOD	Urine / Tissue	Protein / Gene	AKI severity – progression / CKD progression	[23, 29]
VDBP	Plasma	Protein	AKI severity	[5]
Wnt-7a	Serum	Protein	Renal repair	[30]
YLK-40	Plasma, Urine	Protein	AKI severity	[6]
ZNF677, PCDH8	Tissue, Urine	Gene – methylation status	RCC prognosis	[34]

Another major conclusion from review of these studies is that biomarkers, even within the same type of molecule, can be used for multiple purposes: To detect exposure to an agent (i.e., chemical or drug), to detect effects on kidney function, to estimate prognosis of renal disease, to estimate risk for renal disease progression (e.g., from AKI to CKD), and to predict susceptibility renal injury from disease or exposures. While diabetes is a common, underlying cause of renal disease, one might wish to identify biomarkers that uniquely predict AKI or CKD that is associated with diabetes and not with other underlying pathologies or chemical exposures. Examination of the uses of putative biomarkers in Table 1 shows that many of those biomarkers shown to be useful for either DKD severity or DKD progression are also useful for other forms of AKI or CKD. There are some biomarkers listed that have only been identified for AKI or CKD associated with diabetes; however, in no case have any of these biomarkers been shown to be specific for DKD. Rather, these biomarkers just happened to be identified in studies using cohorts of diabetic patients or experimental animal models with DKD.

One question that comes to mind is whether a quantitative analysis is possible to specifically correlate the fold elevation or reduction in the plasma, serum, or urine concentration of a biomarker with the severity or degree of loss of renal function. In light of the diverse array of underlying causes of renal injury and the diverse array of putative biomarkers identified, there cannot be any general rule or quantitative guideline. For a given biomarker, such as cystatin C for DKD severity or CKD progression, some quantitative analysis is possible. However, any quantitative guideline for a different biomarker that also indicates CKD severity and AKI severity and progression, such as β2-microglobulin, is likely to be different and unique to that biomarker.

One clear consensus is that a panel of multiple biomarkers, rather than a single biomarker, generally provides the best results in terms of sensitivity and specificity. The application of newer, more sensitive technologies has greatly expanded the array of potential biomarkers that can be identified. Ultimately, validation and correlations with physiological, biochemical, and molecular processes in the kidneys or target cell population will determine how well they can be applied to improve human health.

Abbreviations:

AKI

acute kidney injury

BAP1

BRCA1 associated protein-1

BTK

Bruton Agammaglobulinemia Tyrosine Kinase

BUN

blood urea nitrogen

ccRCC

clear cell renal carcinoma

CDDP

cisplatin

circRNA

circular RNA

CKD

chronic kidney disease

CXCL6

C-X-C Motif Chemokine Ligand 6

DKD

diabetic kidney disease

DN

diabetic nephropathy

DUXAP8

Double Homeobox A Pseudogene 8

EGF

epidermal growth factor

(e)GFR

(estimated) glomerular filtration rate

ESRD

end-stage renal disease

FGF23

fibroblast growth factor 23

FSGS

focal segmental glomerulosclerosis

GEN

gentamicin

GEO

gene expression omnibus

hATI

histologic acute tubular-interstitial injury

IgAN

immunoglobulin A nephropathy

IGFBP-7

insulin-like growth factor-binding protein 7

KDIGO

Kidney Disease: Improving Global Outcomes

KDM5C

lysine demethylase 5C

KIM-1

kidney injury molecule-1

LC/MS/MS

liquid chromatography tandem mass spectrometry

LINC02609

Long Intergenic Non-Protein Coding RNA 2609

lncRNA

long non-coding RNA

miRNA

microRNA

MN

membranous nephropathy

mTOR

mammalian target of rapomycin

NC-RCC

non-clear cell renal carcinoma

NGAL

neutrophil gelatinase-associated lipocalin

NSAIDs

non-steroidal anti-inflammatory drugs

PBRM1

polybromo-1 protein

PCDH8

procadherin 8

PD-L1

programmed death ligand 1

RCC

renal cell carcinoma

SLC22A2 or OCT2

solute carrier family 22 member 2 or organic cation transporter 2

SCr

serum creatinine

SETD2

SET domain containing 2 / histone lysine methyltransferase

SNPs

single nucleotide polymorphisms

sTNFR-1/2

soluble tumor necrosis factor receptor-1/2

suPAR

soluble urokinase-type plasminogen activator receptor

TCA

tricarboxylic acid cycle

TERT

telomerase reverse transcriptase

TIMP2

tissue inhibitor of metalloproteinases 2

TMT

tandem mass tag

TNFR-2

tumor necrosis factor receptor 2

TP53

tumor protein p53

UMOD

uromodulin

U.S. FDA

United States Food and Drug Administration

VEGF

vascular endothelial growth factor

VDBP

vitamin D binding protein

ZNF677

zinc finger protein 677