How Omics Data Can Be Used in Nephrology

Integrated multi-omics with machine learning to uncover the intricacies of kidney disease

The development of omics technologies has driven a profound expansion in the scale of biological data and the increased complexity in internal dimensions, prompting the utilization of machine learning (ML) as a powerful toolkit for extracting knowledge and understanding underlying biological patterns. Kidney disease represents one of the major growing global health threats with intricate pathogenic mechanisms and a lack of precise molecular pathology-based therapeutic modalities. Accordingly, there is a need for advanced high-throughput approaches to capture implicit molecular features and complement current experiments and statistics. This review aims to delineate strategies for integrating multi-omics data with appropriate ML methods, highlighting key clinical translational scenarios, including predicting disease progression risks to improve medical decision-making, comprehensively understanding disease molecular mechanisms, and practical applications of image recognition in renal digital pathology. Examining the benefits and challenges of current integration efforts is expected to shed light on the complexity of kidney disease and advance clinical practice.

Derangement in Nicotinamide Adenine Dinucleotide Metabolism is Observed During Acute Kidney Injury Among Male Agricultural Workers at Risk for Mesoamerican Nephropathy

Introduction

Mesoamerican nephropathy (MeN) is a chronic kidney disease (CKD) which may be caused by recurrent acute kidney injury (AKI). We investigated urinary quinolinate-to-tryptophan ratio (Q/T), a validated marker of nicotinamide adenine dinucleotide (NAD+) biosynthesis that is elevated during ischemic and inflammatory AKI, in a sugarcane worker population in Nicaragua with high rates of MeN.

Methods

Among 693 male sugarcane workers studied, we identified 45 who developed AKI during the harvest season. We matched them 1:1 based on age and job category with 2 comparison groups: (i) "no kidney injury," active sugarcane workers with serum creatinine (sCr) <1.1 mg/dl; and (ii) "CKD," individuals no longer working in sugarcane due to their CKD, who had additional 1:1 matching for sCr. We measured urine metabolites using liquid chromatography coupled tandem mass spectrometry (LC-MS/MS) and compared Q/T and other metabolic features between the AKI and comparison groups.

Results

Urine Q/T was significantly higher in workers with AKI than in those with no kidney injury (median interquartile Range [IQR]: 0.104 [0.074-0.167] vs. 0.060 [0.045-0.091], P < 0.0001) and marginally higher than in workers with CKD (0.086 [0.063-0.142], P = 0.059). Urine levels of the NAD+ precursor nicotinamide were lower in the AKI group than in comparison groups.

Conclusion

Workers at risk for MeN who develop AKI demonstrate features of impaired NAD+ biosynthesis, thereby providing new insights into the metabolic mechanisms of injury in this population. Therapeutic use of oral nicotinamide, which may ameliorate NAD+ biosynthetic derangement and fortify against kidney injury, should be investigated to prevent AKI in this setting.

The Effect of L-Carnitine on Critical Illnesses Such as Traumatic Brain Injury (TBI), Acute Kidney Injury (AKI), and Hyperammonemia (HA)

L-carnitine (LC) through diet is highly beneficial for critical patients. Studies have found that acetyl-L-carnitine (ALC) can reduce cerebral edema and neurological complications in TBI patients. It significantly improves their neurobehavioral and neurocognitive functions. ALC has also been shown to have a neuroprotective effect in cases of global and focal cerebral ischemia. Moreover, it is an effective agent in reducing nephrotoxicity by suppressing downstream mitochondrial fragmentation. LC can reduce the severity of renal ischemia-reperfusion injury, renal cast formation, tubular necrosis, iron accumulation in the tubular epithelium, CK activity, urea levels, Cr levels, and MDA levels and restore the function of enzymes such as SOD, catalase, and GPx. LC can also be administered to patients with hyperammonemia (HA), as it can suppress ammonia levels. It is important to note, however, that LC levels are dysregulated in various conditions such as aging, cirrhosis, cardiomyopathy, malnutrition, sepsis, endocrine disorders, diabetes, trauma, starvation, obesity, and medication interactions. There is limited research on the effects of LC supplementation in critical illnesses such as TBI, AKI, and HA. This scarcity of studies highlights the need for further research in this area.

Precision Medicine in Diabetic Kidney Disease: A Narrative Review Framed by Lived Experience

Purpose of review

Diabetic kidney disease (DKD) is a leading cause of chronic kidney disease (CKD) for which many treatments exist that have been shown to prevent CKD progression and kidney failure. However, DKD is a complex and heterogeneous etiology of CKD with a spectrum of phenotypes and disease trajectories. In this narrative review, we discuss precision medicine approaches to DKD, including genomics, metabolomics, proteomics, and their potential role in the management of diabetes mellitus and DKD. A patient and caregivers of patients with lived experience with CKD were involved in this review.

Sources of information

Original research articles were identified from MEDLINE and Google Scholar using the search terms "diabetes," "diabetic kidney disease," "diabetic nephropathy," "chronic kidney disease," "kidney failure," "dialysis," "nephrology," "genomics," "metabolomics," and "proteomics."

Methods

A focused review and critical appraisal of existing literature regarding the precision medicine approaches to the diagnosis, prognosis, and treatment of diabetes and DKD framed by a patient partner's/caregiver's lived experience.

Key findings

Distinguishing diabetic nephropathy from CKD due to other types of DKD and non-DKD is challenging and typically requires a kidney biopsy for a diagnosis. Biomarkers have been identified to assist with the prediction of the onset and progression of DKD, but they have yet to be incorporated and evaluated relative to clinical standard of care CKD and kidney failure risk prediction tools. Genomics has identified multiple causal genetic variants for neonatal diabetes mellitus and monogenic diabetes of the young that can be used for diagnostic purposes and to specify antiglycemic therapy. Genome-wide-associated studies have identified genes implicated in DKD pathophysiology in the setting of type 1 and 2 diabetes but their translational benefits are lagging beyond polygenetic risk scores. Metabolomics and proteomics have been shown to improve diagnostic accuracy in DKD, have been used to identify novel pathways involved in DKD pathogenesis, and can be used to improve the prediction of CKD progression and kidney failure as well as predict response to DKD therapy.

Limitations

There are a limited number of large, high-quality prospective observational studies and no randomized controlled trials that support the use of precision medicine based approaches to improve clinical outcomes in adults with or at risk of diabetes and DKD. It is unclear which patients may benefit from the clinical use of genomics, metabolomics and proteomics along the spectrum of DKD trajectory.

Implications

Additional research is needed to evaluate the role of the use of precision medicine for DKD management, including diagnosis, differentiation of diabetic nephropathy from other etiologies of DKD and CKD, short-term and long-term risk prognostication kidney outcomes, and the prediction of response to and safety of disease-modifying therapies.

Transcriptomic analysis reveals partial epithelial-mesenchymal transition and inflammation as common pathogenic mechanisms in hypertensive nephrosclerosis and Type 2 diabetic nephropathy

Hypertensive nephrosclerosis (HN) and Type 2 diabetic nephropathy (T2DN) are the leading causes of chronic kidney disease (CKD). To explore shared pathogenetic mechanisms, we analyzed transcriptomes of kidney biopsies from patients with HN or T2DN. Total RNA was extracted from 10 μm whole kidney sections from patients with HN, T2DN, and normal controls (Ctrl) (n = 6 for each group) and processed for RNA sequencing. Differentially expressed (log2 fold change >1, adjusted p < 0.05) genes (DEG) and molecular pathways were analyzed, and selected results were validated by immunohistochemistry (IHC). ELISA on serum samples was performed on a related cohort consisting of patients with biopsy-proven HN (n = 13) and DN (n = 9), and a normal control group (n = 14). Cluster analysis on RNA sequencing data separated diseased and normal tissues. RNA sequencing revealed that 88% (341 out of 384) of DEG in HN were also altered in T2DN, while gene set enrichment analysis (GSEA) showed that over 90% of affected molecular pathways, including those related to inflammation, immune response, and cell-cycle regulation, were similarly impacted in both HN and T2DN samples. The increased expression of genes tied to interleukin signaling and lymphocyte activation was more pronounced in HN, while genes associated with extracellular matrix organization were more evident in T2DN. Both HN and T2DN tissues exhibited significant upregulation of genes connected with inflammatory responses, T-cell activity, and partial epithelial to mesenchymal transition (p-EMT). Immunohistochemistry (IHC) further confirmed T-cell (CD4+ and CD8+ ) infiltration in the diseased tissues. Additionally, IHC revealed heightened AXL protein expression, a key regulator of inflammation and p-EMT, in both HN and T2DN, while serum analysis indicated elevated soluble AXL levels in patients with both conditions. These findings underline the shared molecular mechanisms between HN and T2DN, hinting at the potential for common therapeutic strategies targeting both diseases.

Biomarker-Development Proteomics in Kidney Transplantation: An Updated Review

Kidney transplantation (KT) is the optimal therapeutic strategy for patients with end-stage renal disease. The key to post-transplantation management is careful surveillance of allograft function. Kidney injury may occur from several different causes that require different patient management approaches. However, routine clinical monitoring has several limitations and detects alterations only at a later stage of graft damage. Accurate new noninvasive biomarker molecules are clearly needed for continuous monitoring after KT in the hope that early diagnosis of allograft dysfunction will lead to an improvement in the clinical outcome. The advent of “omics sciences”, and in particular of proteomic technologies, has revolutionized medical research. Proteomic technologies allow us to achieve the identification, quantification, and functional characterization of proteins/peptides in biological samples such as urine or blood through supervised or targeted analysis. Many studies have investigated proteomic techniques as potential molecular markers discriminating among or predicting allograft outcomes. Proteomic studies in KT have explored the whole transplant process: donor, organ procurement, preservation, and posttransplant surgery. The current article reviews the most recent findings on proteomic studies in the setting of renal transplantation in order to better understand the effective potential of this new diagnostic approach.

Omics and Artificial Intelligence in Kidney Diseases

Omics applications in nephrology may have relevance in the future to improve clinical care of kidney disease patients. In a short term, patients will benefit from specific measurement and computational analyses around biomarkers identified at various omics-levels. In mid term and long term, these approaches will need to be integrated into a holistic representation of the kidney and all its influencing factors for individualized patient care. Research demonstrates robust data to justify the application of omics for better understanding, risk stratification, and individualized treatment of kidney disease patients. Despite these advances in the research setting, there is still a lack of evidence showing the combination of omics technologies with artificial intelligence and its application in clinical diagnostics and care of patients with kidney disease.

Omics-based biomarkers for diagnosis and prediction of kidney allograft rejection

Kidney transplantation is the preferred treatment for patients with end-stage kidney disease, because it prolongs survival and improves quality of life. Allograft biopsy is the gold standard for diagnosing allograft rejection. However, it is invasive and reactive, and continuous monitoring is unrealistic. Various biomarkers for diagnosing allograft rejection have been developed over the last two decades based on omics technologies to overcome these limitations. Omics technologies are based on a holistic view of the molecules that constitute an individual. They include genomics, transcriptomics, proteomics, and metabolomics. The omics approach has dramatically accelerated biomarker discovery and enhanced our understanding of multifactorial biological processes in the field of transplantation. However, clinical application of omics-based biomarkers is limited by several issues. First, no large-scale prospective randomized controlled trial has been conducted to compare omics-based biomarkers with traditional biomarkers for rejection. Second, given the variety and complexity of injuries that a kidney allograft may experience, it is likely that no single omics approach will suffice to predict rejection or outcome. Therefore, integrated methods using multiomics technologies are needed. Herein, we introduce omics technologies and review the latest literature on omics biomarkers predictive of allograft rejection in kidney transplant recipients.

Surfing the Big Data Wave: Omics Data Challenges in Transplantation

In both research and care, patients, caregivers, and researchers are facing a leap forward in the quantity of data that are available for analysis and interpretation, marking the daunting "big data era." In the biomedical field, this quantitative shift refers mostly to the -omics that permit measuring and analyzing biological features of the same type as a whole. Omics studies have greatly impacted transplantation research and highlighted their potential to better understand transplant outcomes. Some studies have emphasized the contribution of omics in developing personalized therapies to avoid graft loss. However, integrating omics data remains challenging in terms of analytical processes. These data come from multiple sources. Consequently, they may contain biases and systematic errors that can be mistaken for relevant biological information. Normalization methods and batch effects have been developed to tackle issues related to data quality and homogeneity. In addition, imputation methods handle data missingness. Importantly, the transplantation field represents a unique analytical context as the biological statistical unit is the donor-recipient pair, which brings additional complexity to the omics analyses. Strategies such as combined risk scores between 2 genomes taking into account genetic ancestry are emerging to better understand graft mechanisms and refine biological interpretations. The future omics will be based on integrative biology, considering the analysis of the system as a whole and no longer the study of a single characteristic. In this review, we summarize omics studies advances in transplantation and address the most challenging analytical issues regarding these approaches.

Mendelian Randomization Analysis as a Tool to Gain Insights into Causes of Diseases: A Primer

Many Mendelian randomization (MR) studies have been published recently, with inferences on the causal relationships between risk factors and diseases that have potential implications for clinical research. In nephrology, MR methods have been applied to investigate potential causal relationships of traditional risk factors, lifestyle factors, and biomarkers from omics technologies with kidney function or CKD. This primer summarizes the basic concepts of MR studies, highlighting methods used in recent applications, and emphasizes key elements in conducting and reporting of MR studies that are important for interpreting the results.

Metabolomic analysis of uremic pruritus in patients on hemodialysis

Pruritus is a common debilitating symptom experienced by hemodialysis patients. Treatment is difficult because the cause of uremic pruritus is not known. This study addressed the hypothesis that pruritus is caused by solutes that accumulate in the plasma when the kidneys fail. We sought to identify solutes responsible for uremic pruritus using metabolomic analysis to compare the plasma of hemodialysis patients with severe pruritus versus mild/no pruritus. Pruritus severity in hemodialysis patients was assessed using a 100-mm visual analogue scale (VAS), with severe pruritus defined as >70 mm and mild/no pruritus defined as <10 mm. Twelve patients with severe pruritus (Itch) and 24 patients with mild/no pruritus (No Itch) were included. Pre-treatment plasma and plasma ultrafiltrate were analyzed using an established metabolomic platform (Metabolon, Inc.). To identify solutes associated with pruritus, we compared the average peak area of each solute in the Itch patients to that of the No Itch patients using the false discovery rate (q value) and principal component analysis. Dialysis vintage, Kt/V_urea, and serum levels of calcium, phosphorus, PTH, albumin, ferritin, and hemoglobin were similar in the Itch and No Itch patients. Metabolomic analysis identified 1,548 solutes of which 609 were classified as uremic. No difference in the plasma or plasma ultrafiltrate levels of any solute or group of solutes was found between the Itch and No Itch patients. Metabolomic analysis of hemodialysis patients did not reveal any solutes associated with pruritus. A limitation of metabolomic analysis is that the solute of interest may not be included in the metabolomic platform’s chemical library. A role for uremic solutes in pruritus remains to be established.

Handling High-Throughput Omics Data for Systems Genetics Analysis

Omics data are being generated and collected at unprecedented scale. During the last decade, single omics, such as genomics, transcriptomics, proteomics, and metabolomics, have already highlighted pathophysiological pathways underpinning a variety of conditions across all the fields of medicine.In fact, high-throughput data generated by the comprehensive and unbiased analysis of an entire segment of the flow of genetic information (i.e., genetic variants in the case of genomics, or gene expression in transcriptomics) certainly provide a plethora of information and a precious support to dissect the mechanisms involved in complex diseases.Yet the most effective approach, set to fully exploit the potential of such big data, lies in the possibility to integrate various omics to unveil previously unappreciated pathways. This approach is the foundation of Systems Biology and allows to overcome the limitations inherent to single omics and traditional biology analyses.A robust and powerful strategy has been developed to integrate genetics and gene expression data in the framework of Systems Genetics. With this technique the first two layers of the flow of genetic information are integrated and specifically it is possible to pinpoint which genetic variants are associated with gene co-expression networks.Here we present a versatile bioinformatic protocol that can be used to study the Systems Genetics of CTLs, in order to identify genes (also known as master regulators) that influence the activation of biological pathways in these cells in a particular state or condition.

Biomarker discovery and beyond for diagnosis of bladder diseases

Molecular biosignatures of altered cellular landscapes and functions have been casually linked with pathological conditions, which imply the promise of biomarkers specific to bladder diseases, such as bladder cancer and other dysfunctions. Urinary biomarkers are particularly attractive due to costs, time, and the minimal and noninvasive efforts acquiring urine. The evolution of omics platforms and bioinformatics for analyzing the genome, epigenome, transcriptome, proteome, lipidome, metabolome, etc., have enabled us to develop more sensitive and disease-specific biomarkers. These discoveries broaden our understanding of the complex biology and pathophysiology of bladder diseases, which can ultimately be translated into the clinical setting. In this short review, we will discuss current efforts on identification of promising urinary biomarkers of bladder diseases and their roles in diagnosis and monitoring. With these considerations, we also aim to provide a prospective view of how we can further utilize these bladder biomarkers in developing ideal and smart medical devices that would be applied in the clinic.

Changes in Novel AKI Biomarkers after Exercise. A Systematic Review

More than 100 substances have been identified as biomarkers of acute kidney injury. These markers can help to diagnose acute kidney injury (AKI) in its early phase, when the creatinine level is not increased. The two markers most frequently studied in plasma and serum are cystatin C and neutrophil gelatinase-associated lipocalin (NGAL). The former is a marker of kidney function and the latter is a marker of kidney damage. Some other promising serum markers, such as osteopontin and netrin-1, have also been proposed and studied. The list of promising urinary markers is much longer and includes cystatin C, NGAL, kidney injury molecule-1 (KIM-1), liver-type fatty-acid-binding protein (L-FABP), interleukin 18, insulin-like growth factor binding protein 7 (IGFBP-7), tissue inhibitor of metalloproteinases-2 (TIMP-2) and many others. Although these markers are increased in urine for no longer than a few hours after nephrotoxic agent action, they are not widely used in clinical practice. Only combined IGFBP-7/TIMP-2 measurement was approved in some countries as a marker of AKI. Several studies have shown that the levels of urinary AKI biomarkers are increased after physical exercise. This systematic review focuses on studies concerning changes in new AKI biomarkers in healthy adults after single exercise. Twenty-seven papers were identified and analyzed in this review. The interpretation of results from different studies was difficult because of the variety of study groups, designs and methodology. The most convincing data concern cystatin C. There is evidence that cystatin C is a better indicator of glomerular filtration rate (GFR) in athletes after exercise than creatinine and also at rest in athletes with a lean mass lower or higher than average. Serum and plasma NGAL are increased after prolonged exercise, but the level also depends on inflammation and hypoxia; therefore, it seems that in physical exercise, it is too sensitive for AKI diagnosis. It may, however, help to diagnose subclinical kidney injury, e.g., in rhabdomyolysis. Urinary biomarkers are increased after many types of exercise. Increases in NGAL, KIM-1, cystatin-C, L-FABP and interleukin 18 are common, but the levels of most urinary AKI biomarkers decrease rapidly after exercise. The importance of this short-term increase in AKI biomarkers after exercise is doubtful. It is not clear if it is a sign of mild kidney injury or physiological metabolic adaptation to exercise.

Proteomic Analysis of Renal Biomarkers of Kidney Allograft Fibrosis-A Study in Renal Transplant Patients

Renal transplantation is the preferred treatment of end stage renal disease, but allograft survival is limited by the development of interstitial fibrosis and tubular atrophy in response to various stimuli. Much effort has been put into identifying new protein markers of fibrosis to support the diagnosis. In the present work, we performed an in-depth quantitative proteomics analysis of allograft biopsies from 31 prevalent renal transplant patients and correlated the quantified proteins with the volume fraction of fibrosis as determined by a morphometric method. Linear regression analysis identified four proteins that were highly associated with the degree of interstitial fibrosis, namely Coagulation Factor XIII A chain (estimate 18.7, adjusted p < 0.03), Uridine Phosphorylase 1 (estimate 19.4, adjusted p < 0.001), Actin-related protein 2/3 subunit 2 (estimate 34.2, adjusted p < 0.05) and Cytochrome C Oxidase Assembly Factor 6 homolog (estimate -44.9, adjusted p < 0.002), even after multiple testing. Proteins that were negatively associated with fibrosis (p < 0.005) were primarily related to normal metabolic processes and respiration, whereas proteins that were positively associated with fibrosis (p < 0.005) were involved in catabolic processes, cytoskeleton organization and the immune response. The identified proteins may be candidates for further validation with regards to renal fibrosis. The results support the notion that cytoskeleton organization and immune responses are prevalent processes in renal allograft fibrosis.

The Serum Metabolome Identifies Biomarkers of Dietary Acid Load in 2 Studies of Adults with Chronic Kidney Disease

BACKGROUND

Dietary acid load is a clinically important aspect of the diet that reflects the balance between acid-producing foods, for example, meat and cheese, and base-producing foods, for example, fruits and vegetables.

METHODS

We used metabolomics to identify blood biomarkers of dietary acid load in 2 independent studies of chronic kidney disease patients: the African American Study of Kidney Disease and Hypertension (AASK, n = 689) and the Modification of Diet in Renal Disease (MDRD, n = 356) study. Multivariable linear regression was used to assess the cross-sectional association between serum metabolites whose identity was known (outcome) and dietary acid load (exposure), estimated with net endogenous acid production (NEAP) based on 24-h urine urea nitrogen and potassium, and adjusted for age, sex, race, randomization group, measured glomerular filtration rate, log-transformed urine protein-to-creatinine ratio, history of cardiovascular disease, BMI, and smoking status.

RESULTS

Out of the 757 known, nondrug metabolites identified in AASK, 26 were significantly associated with NEAP at the Bonferroni threshold for significance (P < 6.6 × 10-5). Twenty-three of the 26 metabolites were also identified in the MDRD study, and 13 of the 23 (57%) were significantly associated with NEAP (P < 2.2 × 10-3), including 5 amino acids (S-methylmethionine, indolepropionylglycine, indolepropionate, N-methylproline, N-δ-acetylornithine), 2 cofactors and vitamins (threonate, oxalate), 1 lipid (chiro-inositol), and 5 xenobiotics (methyl glucopyranoside, stachydrine, catechol sulfate, hippurate, and tartronate). Higher levels of all 13 replicated metabolites were associated with lower NEAP in both AASK and the MDRD study.

CONCLUSION

Metabolomic profiling of serum specimens from kidney disease patients in 2 study populations identified 13 replicated metabolites associated with dietary acid load. Additional studies are needed to validate these compounds in healthy populations. These 13 compounds may potentially be used as objective markers of dietary acid load in future nutrition research studies.

Combination of the fetal urinary metabolome and peptidome for the prediction of postnatal renal outcome in fetuses with PUV

Most of biomarker panels, extracted from single omics traits, still need improvement since they display a gray zone where prediction is uncertain. Here we verified whether a combination of omics traits, fetal urinary metabolites and peptides analyzed in the same sample, improved prediction of postnatal renal function in fetuses with posterior urethral valves (PUV) compared to individual omics traits. Using CE-MS, we explored the urinary metabolome of 13 PUV fetuses with end stage renal disease (ESRD) and 12 PUV fetuses without postnatal ESRD at 2 years postnatally. This allowed the selection of 24 differentially abundant metabolite features which were modelled into predictive classifiers, alone or in combination with 12 peptides previously identified as predictive of ESRD. Validation in 35 new fetuses showed that the combination of peptides and metabolites significantly outperformed the 24 metabolite features with increased AUC (0.987 vs 0.905), net reclassification improvement (36%) and better sensitivity accuracy (86% vs 60%). In addition, the two trait combination tended to improve, but without reaching statistical significance, the already high performances of the 12 peptide biomarkers (AUC 0.967, accuracy 80%). In conclusion, this study demonstrates the potential of cumulating different omics traits in biomarker research where single omics traits fall short.

SIGNIFICANCE

Although increasingly proposed in disease-diagnosis and -prognosis because of their improved efficacy over single markers, panels of body fluid biomarkers based on single omics analysis still fail to display perfect accuracy, probably due to biological variability. Here, we hypothesized that combination of different omics traits allowed to better capture this biological variability. As proof of concept, we studied the added value of fetal urine metabolites and peptides using CE-MS, starting from the same urine sample, to predict postnatal renal outcome in fetuses with posterior urethral valves. We observed that the prognostic power of combined metabolite and peptide markers was clearly higher than that of metabolites alone and slightly, but non-significantly, improved compared to the peptides alone. To our knowledge, this report is the first to demonstrate that combining multiomics traits extracted from (fetal) urine samples displays clear promise for kidney disease stratification.