Endothelin-1 Mediates the Systemic and Renal Hemodynamic Effects of GPR81 Activation

Abstract

GPR81 (G-protein-coupled receptor 81) is highly expressed in adipocytes, and activation by the endogenous ligand lactate inhibits lipolysis. GPR81 is also expressed in the heart, liver, and kidney, but roles in nonadipose tissues are poorly defined. GPR81 agonists, developed to improve blood lipid profile, might also provide insights into GPR81 physiology. Here, we assessed the blood pressure and renal hemodynamic responses to the GPR81 agonist, AZ'5538. In male wild-type mice, intravenous AZ'5538 infusion caused a rapid and sustained increase in systolic and diastolic blood pressure. Renal artery blood flow, intrarenal tissue perfusion, and glomerular filtration rate were all significantly reduced. AZ'5538 had no effect on blood pressure or renal hemodynamics in Gpr81^-/- mice. Gpr81 mRNA was expressed in renal artery vascular smooth muscle, in the afferent arteriole, in glomerular and medullary perivascular cells, and in pericyte-like cells isolated from kidney. Intravenous AZ'5538 increased plasma ET-1 (endothelin 1), and pretreatment with BQ123 (endothelin-A receptor antagonist) prevented the pressor effects of GPR81 activation, whereas BQ788 (endothelin-B receptor antagonist) did not. Renal ischemia-reperfusion injury, which increases renal extracellular lactate, increased the renal expression of genes encoding ET-1, KIM-1 (Kidney Injury Molecule 1), collagen type 1-α1, TNF-α (tumor necrosis factor-α), and F4/80 in wild-type mice but not in Gpr81^-/- mice. In summary, activation of GPR81 in vascular smooth muscle and perivascular cells regulates renal hemodynamics, mediated by release of the potent vasoconstrictor ET-1. This suggests that lactate may be a paracrine regulator of renal blood flow, particularly relevant when extracellular lactate is high as occurs during ischemic renal disease.

Keywords: blood pressure; hypoxia; ischemia-reperfusion-injury; pericytes; renal blood flow.

Figure 1.

Effect of AZ′5538 infusion on blood pressure (BP) and renal hemodynamics. Mice were infused intravenously with vehicle (5% mannitol; n=6, open squares) or AZ′5538 (1 µmol/[kg·min]; n=6 black circles) for 15 minutes. Systolic blood pressure (A; SBP), diastolic blood pressure (B; DBP), peak BP responses (C), and heart rate (D). E, Renal artery blood flow (RBF) in separate mice infused with vehicle (n=6) or AZ′5538 (n=6). F, Glomerular filtration rate (GFR) before and after infusion with AZ′5538 (n=8), compared to vehicle (n=8). Data are mean±SD, and for C and F, individual datapoints are shown. Statistical comparisons were made by 2-way ANOVA for the main effects of treatment, of time and of the interaction (see text for P values) and by unpaired t test (C and F), with P values as shown.

Figure 2.

Blood pressure and hemodynamic effects of AZ′5538 are GPR81 (G-protein-coupled receptor 81) dependent. Gpr81^−/− mice (open circles; n=6) and wild-type littermates (closed squares; n=6) were infused intravenously with vehicle (5% mannitol) or AZ′5538 (1 µmol/[kg·min]; n=6 black circles) for 15 minutes. Systolic blood pressure (SBP, A), diastolic blood pressure (DBP, B), renal artery blood flow (RBF, C), renal vascular resistance (D), perfusion of the renal cortex (E), and perfusion of the renal medulla (F) ΔTPU= change in Total Perfusion Units. All data are mean±SD from baseline. Statistical comparisons were made by 2-way ANOVA for the main effects of genotype, time, and of the interaction (see text for P values).

Figure 3.

Gpr81 (G-protein-coupled receptor 81) expression in artery and renal cells. A, RNA was extracted and reverse transcribed from C57Bl/6JCrl mouse vessels. Negative controls were samples where reverse transcription enzyme, or RNA, were left out at the cDNA conversion step. All samples underwent end point polymerase chain reaction (PCR) before gel electrophoresis. n=4 for all vessel types, pooled from 2 mice for renal arteries and mesenteric arteries. B, Defined populations of cells were isolated by fluorescence-activated cell sorting (FACS) from kidney taken from n=4 mice. Quantitative PCR was used to measure Gpr81 in PDGFRβ⁺ (pericytes), CD31⁺ (endothelial cells), F4/80⁺ (macrophages), and LTL⁺ (tubule) cell groups. Expression was significantly >0 in PDGFRβ⁺ cells (P<0.001).

Figure 4.

Representative figures of Gpr81 in situ hybridization. Positive Gpr81 mRNA expression shown by red punctuated dots. Expression found in wild-type (WT) mouse glomeruli of the kidney cortex (A) where the star indicates an arteriole and medulla (C). No staining was seen in Gpr81^−/− mouse kidney tissues (B and D). Receptor expression also seen in smooth muscle cells of the WT mouse aorta (E) and renal artery (F) where staining is indicated with arrows. Scale bars are 50 μm (A–D) and 20 μm (E and F).

Figure 5.

Role of the endothelin system. A, The change in plasma endothelin-1 in C57Bl/6J mice infused with either vehicle (open square; n=8) or AZ′5538 (1 µmol/[kg·min] black square; n=8) for 15 min. Individual data points and mean±SD are shown. In separate experiments, C57Bl/6J was infused intravenously with vehicle or an endothelin receptor antagonist for 25 min before treatment with AZ′5538 (1 µmol/[kg·min] for 15 min). The change in systolic blood pressure over baseline is shown. B, Bosentan; C, BQ123; and D, BQ788. Data are mean±SD, analyzed by 2-way ANOVA. Statistical comparisons were made by 2-way ANOVA for the main effects of treatment, time, and of the interaction (see text for P values).

Figure 6.

Gpr81^−/− mice have reduced injury following renal ischemia-reperfusion. Renal ischemia-reperfusion injury or a sham operation was performed on Gpr81^−/− (n=4/6) and wild-type mice (n=6/4). One week later, the renal expression of the following genes was measured by quantitative polymerase chain reaction: (A) Havcr1 (encoding KIM-1 [kidney injury molecule 1]); (B) Col1a1 (encoding collagen type 1 α1); (C) Tnf (encoding TNF-α [tumor necrosis factor-α], ); (D) Ccl2 (encoding MCP-1 [monocyte chemotactic protein 1]); (E) Cxcl1 (encoding C-X-C motif chemokine ligand 1); and (F) Edn1 (encoding preproendothelin-1). Expression is normalized to housekeepers; individual data points and group mean±SD are shown. Statistical comparisons were made by 1-way ANOVA with Holm-Sidak test for planned comparisons with P values as indicated.