Plasma proteome profiling for AKI biomarker candidates associated with first-line ART in people living with HIV in South Africa

Abstract

With the highest global burden of HIV, South Africa initiates HIV treatment using first-line antiretroviral therapy (ART) regimens in newly diagnosed patients. Antiretroviral (ARV)-associated nephrotoxicity has been observed in -10% of South African patients newly receiving first-line ART, as well as in patients who have extensively received TDF-based ART regimens, and this can progress to acute kidney injury (AKI). To identify potential biomarkers that will improve the detection of ARV-associated nephrotoxicity, proteomic analysis was performed using the sequential window acquisition of all theoretical mass spectra (SWATH-MS) data acquisition method on the plasma of the case group (AKI) and the control group (non-AKI). Data are available via ProteomeXchange with identifier PXD054218. Evaluation of the results identified thirty-four proteins that showed a significant change in abundance, with three proteins showing increased abundance, while thirty-one proteins showed decreased abundance between the AKI and non-AKI groups. Machine learning was also used to evaluate the results and showed twenty ranked proteins that contributed to distinguishing between the AKI and non-AKI groups. The majority of the proteins of significant differential abundance and those ranked by the machine learning model participate in enriched biological processes that correspond to known pathophysiological and cellular mechanisms that contribute to AKI, suggesting that those proteins can serve as potential biomarkers to be further verified and validated for AKI diagnosis. Comparison of the plasma and previously generated urinary proteome profiles showed five proteins with significant differences in abundance that overlapped both proteomes providing evidence for the renal physiological dynamics of these proteins in renal injury and disease. Significance: Improving the detection of ARV-associated nephrotoxicity is challenging because current serum creatinine (srCr) based-equations used to estimate the glomerular filtration rate (eGFR) for assessing kidney function have been reported to overestimate the eGFR in the South African population and changes in srCr measurements also reflect a delayed response to the initial structural injury of the kidney. This hinders the timeous detection of ARV-associated nephrotoxicity which can lead to irreversible renal damage and can compromise the continuation of treatment in PLHIV, thus emphasising the need to introduce novel AKI biomarkers to mitigate ARV-associated nephrotoxicity in at risk PLHIV in South Africa. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=99 SRC="FIGDIR/small/605118v1_ufig1.gif" ALT="Figure 1"> View larger version (18K): org.highwire.dtl.DTLVardef@18743eaorg.highwire.dtl.DTLVardef@17bba7borg.highwire.dtl.DTLVardef@101c78org.highwire.dtl.DTLVardef@1bd2b23_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LIIdentified thirty-four proteins with significant changes in abundance between study groups. C_LIO_LIMachine learning showed fourteen proteins associated with renal disease biology that differentiate the study groups. C_LIO_LIFirst study to analyse plasma proteome of this cohort of PLHIV in South Africa C_LIO_LIFive proteins overlap the plasma and urinary proteome profiles of the study groups. C_LI