The future role of proteomics in the understanding of acute kidney injury

1. Introduction

Acute kidney injury (AKI) is an increasingly frequent clinical condition that occurs in 5% of all hospitalizations, and in up to 26% of critically ill patients [1]. AKI is associated with significant morbidity and mortality in the acute setting, and leads to increased risk of chronic kidney disease (CKD) [2,3]. AKI lacks effective therapeutic options and represents an annual hospitalization cost of $5.4 billion in the United States alone [4]. While research has advanced a number of findings related to the underlying mechanisms of AKI, and novel interventions have shown promise in animal models of AKI, the translation of these findings have yet to have an impact on patient outcomes. Among the reasons for this lack of progress in improving outcomes include (a) a lack of complete understanding of precipitating factors and mechanisms, and (b) a dearth of early diagnostic markers to predict the occurrence of AKI and subsequent outcomes. Advances in transcriptomic and proteomic technologies have led to the discovery of a number of promising diagnostic markers and potential targets for new innovative therapies. The focus of this editorial will be on recent proteomic and metabolomic findings and their potential future use in the field of AKI research.

2. The current state of proteomics in AKI

For proteomic analysis of AKI, urine is the most commonly utilized biofluid as it provides a direct window into the kidney – as 75% of urine proteins are produced in the kidney tissue. Protein analysis of urine does present challenges, however. Urine is a dilute fluid of which the protein concentration can vary with disease status, hydration, time of collection, gender and age. These variations require normalizations, which is fraught with difficulties, especially in the setting of AKI, which is not a steady state condition. However, standard protocols are being developed and improved and can be found in the Human Kidney and Urine Proteome Project web-site (www.hkupp.org) [5].

In the last 20 years, we’ve come from having serum creatinine as the only marker for AKI to having several promising early markers as a result of advances in genomics and proteomics. Proteins like NGAL, KIM-1, IL18 and LFABP have been widely studied over the last decade and provide much earlier notice and a temporal timeline of tubular damage in AKI than does serum creatinine [6]. TIMP2 and IGFBP7 are cell cycle arrest markers that have been approved by the FDA as a clinical test for AKI, however the timing of their increase is later than NGAL and IL18 [7]. New markers are needed that provide more insight into mechanism and for differentiating types of injury. The limitations of traditional methodologies such as 2DGE and MALDI-TOF MS are giving way to newer, more sensitive and reproducible methods, such as capillary electrophoresis mass spectrometry (CE-MS), which are capable of differential peptide identification with high sensitivity in lower sample volumes. Lee et al. 2017 [8] used 2DGE and MALDI to analyze kidney tissue in a rat model of mercury nephrotoxicity and discovered selenium binding protein-1 was markedly upregulated in the renal cortex in an Hg dose dependent manner. Immunohistochemistry of SBP1 showed strong staining in the cytoplasm and membrane of damaged tubular cells in a dose dependent manner. To test whether SBP1 could be used as a urinary marker for nephrotoxic AKI, they measured SBP1 in the urine of mice undergoing the classic cisplatin nephrotoxicity model. Once again, they found SBP1 to be a sensitive and tissue specific biomarker of AKI, as hepatotoxic agent exposure led to no increase in SBP1 levels. These results have since been replicated both in another model of cisplatin induced nephrotoxicity and in models of ischemia reperfusion injury (IRI), indicating that SBP1 may be another tubular marker of AKI in numerous contexts [9]. While SBP1 was elevated in the urine of the IRI group at 9 h post IRI, it is not known whether it is elevated any earlier like NGAL, which peaks at between 2 and 6 h post IRI [10].

Other targeted methods, such as multiplexed screening of cytokines and protein profiles associating with various conditions, allow for screening of a large number of known targets using very small volumes of urine. Multiple platforms, such as flow cytometry–based bead assays, magnetic bead imaging assays, and electrochemiluminescent multi-array technology allow for rapid screening of samples for a large number of analytes in a very-small sample volume. These technologies are useful for screening known markers of disease in new conditions, and further to allow for validation of proteomic findings in a large number of samples, using small volumes. In one such instance, Giquin et al. [11], demonstrated the feasibility of this approach by using protein standard absolute quantification (PSAQ), a targeted LC-MS proteomic method, to multiplex myoinositol oxygenase (MIOX), phosphoenolpyruvate carboxykinase 1 (PCK1), NGAL and LFABP. They used this panel to screen healthy donors or patients with tubular or glomerular AKI. The linearity and accuracy of the assay was excellent for the four markers, and it confirmed the utility of urinary NGAL and LFABP to detect kidney injury.

3. Metabolomics in AKI

Another area with great potential for use in the understanding of AKI is metabolomics and its sub-branch lipidomics. These fields aim to capture a snapshot in time in the cells or biofluid by profiling the individual metabolites and lipid content. Romick-Rosendale et al. [12], performed NMR based metabolic profiling on low birth-weight and very low birth-weight neonates, for which current markers of AKI do not function well. They discovered that carnitine was markedly elevated in the urine of preterm infants with AKI due in part to nephrotoxic antibiotic use. Rao et al. [13], used sequential window acquisition of all theoretical spectra (SWATH)-mass spectrometry to analyze lipid changes in a rat model of IRI AKI. SWATH is a novel method of analyzing all spectra within small sequential mass windows to identify many more low abundance targets than traditional targeted analyses. This method can be applied to both proteomics and metabolomics and allows you to discover an order of magnitude more lipids or proteins that are different between groups. The authors discovered significant increases in two phospholipids at 6 h post injury. The lipids were phosphatidylcholine (PC) O-38:1 and phosphatidylethanolamine (PE) O-42:3. PC O-38:1 remained elevated 24 h post injury, while PE O-42:3 had returned to baseline levels. The use of proteomic imaging technology, specifically, MALDI-imaging MS, allows for pinpointing the location of the lipids in tissue, which is key to begin understanding the potential role for the lipids in the injury pathway.

4. The future

The limitations of the clinical applicability of proteomics to the larger AKI community are that it requires expensive, specialized equipment and in depth specialized training to perform and evaluate. A history of difficulties validating and reproducing proteomic findings still lead to problems with broad acceptance. To overcome these limitations, the field must find ways of streamlining analysis and creating more affordable proteomic platforms so that testing can be performed at even small centers on a global scale. Meanwhile, improved analysis techniques such as SWATH, which allow for more thorough and sensitive findings in a non-targeted manner using existing technology may allow for novel discoveries in the terms of pinpointing the location and types of injury in the kidney. By developing more affordable methods of analysis, larger sample cohorts can be explored, leading to more robust findings and more reproducible results. A search for more specific and detailed information regarding the underlying mechanisms and pathways involved in AKI is the direction that future research must address to keep the fields of proteomics and metabolomics alive in AKI.