Simultaneous analyses of urinary eicosanoids and related mediators identified tetranor-prostaglandin E metabolite as a novel biomarker of diabetic nephropathy

Yoshifumi Morita,Makoto Kurano,Eri Sakai,Motoji Sawabe,Junken Aoki,Yutaka Yatomi

doi:10.1016/j.jlr.2021.100120

Abstract

Diabetic nephropathy is a major complication of diabetes mellitus, and thus novel biomarkers are desired to evaluate the presence and progression of diabetic nephropathy. In this study, we sought to identify possible metabolites related to diabetic nephropathy among urinary eicosanoids and related mediators. Using liquid chromatogram-tandem mass spectrometry, we optimized the lipid extraction from urine using the Monospin C18 as a solid-phase extraction cartridge and measured the urinary lipid mediators in 111 subjects with type 2 diabetes mellitus as well as 33 healthy subjects. We observed that 14 metabolites differed significantly among the clinical stages of nephropathy. Among them, levels of tetranor-prostaglandin E metabolite (tetranor-PGEM), an arachidonic acid metabolite, were significantly higher in subjects with stage 1 nephropathy than in healthy subjects and increased with the progression of nephropathy. We also observed that levels of maresin-1, a docosahexaenoic acid metabolite, and leukotriene B4-ethanolamide, an arachidonoyl ethanolamide metabolite, were significantly lower in subjects with stage 3–4 nephropathy than in healthy subjects and those with stage 1–2 nephropathy. Finally, using a comprehensive analysis of urinary eicosanoids and related mediators, we concluded that tetranor-PGEM was capable of discriminating clinical stages of nephropathy and thus useful as a novel biomarker for diabetic nephropathy.

Highlights

Most of the analytes in the urine specimens containing the standard solutions were obtained with a high precision and a coefficient of variation (CV) of 2.4%–17.6% for Strata-X, 0.9%–60.3% for Oasis PRiME HLB, and 0.6%–9.9% for Monospin C18
Sex, eGFR, HbA1c, μAlb, and the metabolites detected in urinary samples as variables, the orthogonal projection to latent structures (OPLS)-DA model showed that 9,10-DiHOME, 13,14-dihydro-15-keto-PGE2, 13,14-dihydro-15-ketotetranor-PGE2, 8-iso-PGE2, PGE2, and tetranor-PGEM were selected as significant contributing variables, following HbA1c and μAlb, for discriminating subjects with stage 1 nephropathy from the control group (Fig. 3A and supplemental Fig. S4)
Validation of the solid-phase extraction (SPE) procedures for lipid extraction showed that the use of the Monospin C18 resulted in greater precision and higher recovery rates (Table 1 and supplemental Table S3)

Summary

Introduction

To validate the SPE procedures, the repeatability and recovery rates were evaluated using three pooled urine samples with the addition of the following standard solutions: 10 ng/ml (final concentration) of tetranor-PGEM, 6-keto-PGF1α, TXB2, PGF2α, 8-iso-PGF2α, PGE2, PGD2, 15-keto-PGE2, PGF2α, 11-dehydro-TXB2, PGA2, LTB4, LTD4, 15-HETE, 12-HETE, 5-HETE, 11,12-EET, and PAF, and 100 ng/ml of AA. The levels of LTB4-EA, a metabolite of AEA, and maresin-1, a DHA derivative, were significantly lower in the subjects with stage 3–4 nephropathy than in the control group and the subjects with stage 1 or 2 nephropathy (Fig. 1G, H).

Results

Conclusion