Effect of Dapagliflozin on Urine Metabolome in Patients with Type 2 Diabetes

Abstract

Context: Inhibitors of sodium-glucose cotransporters-2 have cardio- and renoprotective properties. However, the underlying mechanisms remain indeterminate.

Objective: To evaluate the effect of dapagliflozin on renal metabolism assessed by urine metabolome analysis in patients with type 2 diabetes.

Design: Prospective cohort study.

Setting: Outpatient diabetes clinic of a tertiary academic center.

Patients: Eighty patients with hemoglobin A1c > 7% on metformin monotherapy were prospectively enrolled.

Intervention: Fifty patients were treated with dapagliflozin for 3 months. To exclude that the changes observed in urine metabolome were merely the result of the improvement in glycemia, 30 patients treated with insulin degludec were used for comparison.

Main outcome measure: Changes in urine metabolic profile before and after the administration of dapagliflozin and insulin degludec were assessed by proton-nuclear magnetic resonance spectroscopy.

Results: In multivariate analysis urine metabolome was significantly altered by dapagliflozin (R2X = 0.819, R2Y = 0.627, Q2Y = 0.362, and coefficient of variation analysis of variance, P < 0.001) but not insulin. After dapagliflozin, the urine concentrations of ketone bodies, lactate, branched chain amino acids (P < 0.001), betaine, myo-inositol (P < 0001), and N-methylhydantoin (P < 0.005) were significantly increased. Additionally, the urine levels of alanine, creatine, sarcosine, and citrate were also increased (P < 0001, P <0.0001, and P <0.0005, respectively) whereas anserine decreased (P < 0005).

Conclusions: Dapagliflozin significantly affects urine metabolome in patients with type 2 diabetes in a glucose lowering-independent way. Most of the observed changes can be considered beneficial and may contribute to the renoprotective properties of dapagliflozin.

Trial registration: ClinicalTrials.gov NCT02798757.

Keywords: branched chain amino acids; dapagliflozin; kidney; metabolomics; osmolytes.

© The Author(s) 2021. Published by Oxford University Press on behalf of the Endocrine Society.

Figures

Figure 1.

(A) OPLS-DA and validation (B)of the corresponding OPLS-DA models via permutations tests (n = 100) of the urine metabolic data from patients with type 2 diabetes mellitus before (black squares) and 3 months after (red triangles) administration of dapagliflozin. The regression line of Q2 (goodness of prediction, blue squares) with intercept at −0.306 and R2 (goodness of fit, green circles) regression line intercept at 0.242 indicate that the resulting OPLS-DA model has higher R2 (goodness of fit) and Q2 (goodness of prediction) values in the validation test than the permutated models generated. (C) The corresponding OPLS-DA regression coefficient plot of the urine metabolites, upper (black): higher before and down (red): higher after treatment. Abbreviations: 1MH, 1-methylhistidine; 1MNA, 1-methylnicotinamide; 2HIBA, 2-hydroxyisobutyrate; 3ClTyr, 3-chlorotyrosine; 3HIVA, 3-hydroxyisovalerate; 3IS, 3-indoxylsulfate; 3M2OV, 3-methyl-2-oxovalerate; 3MH, 3-methylhistidine; Ala, alanine; Ans, anserine; Bet, betaine; Car, carnitine; Cho, choline; cisAco, cis-aconitate; Cit, citrate; Cr, creatine; Creat, creatinine; Fum, fumarate; GLA, gluconate; Gluc, glucose; Hip, hippurate; Hpx, hypoxanthine; ILA, indole-3-lactate; Ile, isoleucine; Imid, imidazole; Lys, lysine; Mtl, mannitol; Meth, methanol; N-IVG, N-isovaleroylglycine; N-PAG, N-phenylacetylglycine; Suc, succinate; TG, trigonelline; Thr, threonine; TMAO, trimethylamine N-oxide; transAco, trans-aconitate; Val, valine; *, drug metabolites; Un, unknown.