Cardioprotection by nicotinamide mononucleotide (NMN): Involvement of glycolysis and acidic pH

Abstract

Stimulation of the cytosolic NAD⁺ dependent deacetylase SIRT1 is cardioprotective against ischemia-reperfusion (IR) injury. NAD⁺ precursors including nicotinamide mononucleotide (NMN) are thought to induce cardioprotection via SIRT1. Herein, while NMN protected perfused hearts against IR (functional recovery: NMN 42 ± 7% vs. vehicle 11 ± 3%), this protection was insensitive to the SIRT1 inhibitor splitomicin (recovery 47 ± 8%). Although NMN-induced cardioprotection was absent in Sirt3^-/- hearts (recovery 9 ± 5%), this was likely due to enhanced baseline injury in Sirt3^-/- (recovery 6 ± 2%), since similar injury levels in WT hearts also blunted the protective efficacy of NMN. Considering alternative cardiac effects of NMN, and the requirement of glycolysis for NAD⁺, we hypothesized NMN may confer protection in part via direct stimulation of cardiac glycolysis. In primary cardiomyocytes, NMN induced cytosolic and extracellular acidification and elevated lactate. In addition, [U-¹³C]glucose tracing in intact hearts revealed that NMN stimulated glycolytic flux. Consistent with a role for glycolysis in NMN-induced protection, hearts perfused without glucose (palmitate as fuel source), or hearts perfused with galactose (no ATP from glycolysis) exhibited no benefit from NMN (recovery 11 ± 4% and 15 ± 2% respectively). Acidosis during early reperfusion is known to be cardioprotective (i.e., acid post-conditioning), and we also found that NMN was cardioprotective when delivered acutely at reperfusion (recovery 39 ± 8%). This effect of NMN was not additive with acidosis, suggesting overlapping mechanisms. We conclude that the acute cardioprotective benefits of NMN are mediated in part via glycolytic stimulation, with the downstream protective mechanism involving enhanced ATP synthesis during ischemia and/or enhanced acidosis during reperfusion.

Keywords: Acidosis; Glycolysis; Ischemia; Lactate; NAD(+); NMN.

Figure 1:. NMN Induced Acute Cardioprotection is Independent of SIRT1.

(A): Cardiac functional data (rate x pressure product) for mouse hearts subjected to 25 min. global ischemia and 60 min. reperfusion. Prior to the onset of ischemia hearts were perfused for 20 min. with vehicle (Ctrl, white circles), 1 mM NMN (black circles), or 1 mM NMN plus 10 μΜ splitomicin (Sp, gray triangles). Perfusion media contained both glucose (5 mM) and fat (palmitate-BSA 100 μΜ) as metabolic substrates. Data shown are means ± SEM, N=5–7 animals per group, *p<0.01 for all points indicated, between Ctrl. and NMN+Sp, #p<0.01 for all points indicated between Ctrl. and NMN.(B): Infarction data for the hearts from panel A, determined by tetrazolium chloride staining. Images above the graph show representative infarct photographs (upper), and pseudo-colored images (lower) used for quantitation by planimetry. In graph, individual data points are shown on the left, and means ± SEM on the right, for each treatment group. *p<0.001 vs. Ctrl. (C): Acetyl-lysine western blot on cardiac fractions from NMN treated hearts. Hearts were perfused for 20 min. with 1 mM NMN under normoxic conditions then separated into fractions (Hmg: homogenate, Nucl: nuclear, Mito: mitochondrial, Cyto: cytosolic) by differential centrifugation. Ponceau S stained membrane (loading control) is shown below compressed vertically to save space. Numbers below the blot indicate densitometric quantitation of acetyl-lysine intensity down whole lanes, normalized to protein loading, for N=4 independent experiments (means ± SEM). Fractionation controls are shown in Figure S1A. (D): Acetyl-lysine western blot on mitochondrial fraction from NMN treated hearts of WT and Sirt3^−/− mice. Center image shows anti-SIRT3 blot, confirming status of knockout animals. Ponceau S stained membrane (loading control) is shown below compressed vertically to save space. (E): Cardiac functional data (rate x pressure product) for Sirt3^−/− mouse hearts subjected to 25 min. global ischemia and 60 min. reperfusion. Prior to the onset of ischemia hearts were perfused for 20 min. with vehicle (Ctrl, white circles) or 1 mM NMN (black circles). Perfusion media contained both glucose (5 mM) and fat (palmitate-BSA 100 μΜ) as metabolic substrates. Data shown are means ± SEM, N=4 animals per group. (F): Infarction data for the hearts from panel E, determined by tetrazolium chloride staining. Images above the graph show representative infarct photographs (upper), and pseudo-colored images (lower) used for quantitation by planimetry. In graph, individual data points are shown on the left, and means ± SEM on the right, for each treatment group.

Figure 2:. NMN Stimulates Cardiomyocyte Glycolysis Leading to Cell Acidification.

(A): Representative fluorescent microscope images of primary adult mouse cardiomyocytes stained with the pH sensitive vital dye BCECF and treated for 40 min. with 1 mM NMN or vehicle control. Images are pseudo-colored according to the pH scale in panel B. (B): Graph shows average pH_i values from 9 independent plates of cells (means ± SEM, *p<0.001 between treatment groups). (C): Simulated IR injury. Cardiomyocytes were subjected to 60 min. similated ischemia followed by 60 min. simulated reperfusion (see methods), with 20 min. prior treatment with 1 mM NMN. Cell death (LDH release assay) is shown as mean ± SEM, N=5 independent cell preparations per group. (D): Oxygen Consumption Rate (OCR) and Extracellular Acidification Rate (ECAR) values measured via Seahorse™ XF analysis on primary adult mouse cardiomyocytes before and 10 min. after addition of vehicle (white bars) or 1 mM NMN (black bars) to the media. Where indicated, splitomicin (Sp, 10 μΜ) was present throughout (gray bars). Data are means ± SD for 7–8 wells per treatment group, on a single XF-24 plate. *p<0.001 between baseline and post drug treated, within a given treatment group. (E): Relative levels of selected metabolites in freeze-clamped perfused mouse hearts treated for 20 min. with 1 mM NMN or vehicle control. Note break in y-axis scale to accommodate lactate. Data are presented as the metabolite level in NMN treated hearts normalized to that in control hearts (means ± SEM, N=7 animals per group, *p<0.05, #p<0.01, between treatment groups). (F): Fractional saturation of metabolites from [U-¹³C]glucose infusion. Following vehicle control (white bars) or NMN treatment (black bars) for 20 min., glucose in KH buffer was replaced with [U-¹³C]glucose for 5 min., followed by freeze-clamp and isotopologue analysis by LC-MS/MS. Data are presented as fractional saturation - i.e., the fraction of the metabolite that is replaced by labeled metabolite within the allotted time, thus equating to metabolic flux from [U-¹³C]glucose to that point in metabolism (means ± SEM, N=5 animals per group, *p<0.05, #p<0.01 between vehicle and NMN groups).

Figure 3:. NMN Induced Acute Cardioprotection Requires Glucose.

(A): Cardiac functional data (rate x pressure product) for mouse hearts subjected to 25 min. global ischemia and 60 min. reperfusion. Prior to the onset of ischemia (black bar at top), hearts were perfused for 20 min. with vehicle (Ctrl, white circles), or 1 mM NMN (black circles). Experiments were performed essentially as in Figure 1A, except that perfusion media contained zero glucose, such that palmitate was the sole metabolic substrate. Data shown are means ± SEM, N=6 animals per group. No significant differences were noted between Ctrl. and NMN groups. (B): Infarction data for the hearts from panel A, determined by tetrazolium chloride staining. Images above the graph show representative infarct photographs (upper), and pseudo-colored images (lower) used for quantitation by planimetry. In graph, individual data points are shown on the left, and means ± SEM on the right, for each treatment group. No significant differences were noted between Ctrl. and NMN groups. (C): Cardiac functional data (rate x pressure product) for mouse hearts subjected to 25 min. global ischemia and 60 min. reperfusion. Prior to the onset of ischemia (black bar at top), hearts were perfused for 20 min. with vehicle (Ctrl, white circles), or 1 mM NMN (black circles). Experiments were performed essentially as in Figure 1A, except that perfusion media contained 5 mM galactose instead of glucose. Palmitate-BSA was still present. Data shown are means ± SEM, N=4 animals per group. No significant differences were noted between Ctrl. and NMN groups. (D): Infarction data for the hearts from panel C, determined by tetrazolium chloride staining. Images above the graph show representative infarct photographs (upper), and pseu-docolored images (lower) used for quantitation by planimetry. In graph, individual data points are shown on the left, and means ± SEM on the right, for each treatment group. No significant differences were noted between Ctrl. and NMN groups.

Figure 4:. NMN is Protective at Reperfusion.

(A): Cardiac functional data (rate x pressure product) for mouse hearts subjected to 25 min. global ischemia and 60 min. reperfusion. At the onset of reperfusion hearts were perfused for 20 min. with vehicle (Ctrl, white circles), or 1 mM NMN (black circles). Perfusion media contained both glucose (5 mM) and fat (palmitate-BSA, 100 μΜ) as metabolic substrates. N.B. Control data in this panel and panel B are the same as those in Figures 1A and B. Data shown are means ± SEM, N=6–7 animals per group. *p<0.01 for all points indicated, between Ctrl. and NMN groups. (B): Infarction data for the hearts from panel A, determined by tetrazolium chloride staining. Images above the graph show representative infarct photographs (upper), and pseudo-colored images (lower) used for quantitation by planimetry. In graph, individual data points are shown on the left, and means ± SEM on the right, for each treatment group. *p<0.001 between Ctrl. and NMN groups. (C): Simulated IR (sIR) injury. Cardiomyocytes were subjected to sIR as in Figure 2C, with delivery of either acidic media, 1 mM NMN, or both at reperfusion. Cell death (LDH release assay) is shown as mean ± SEM, N=5 independent cell preparations per group. N.B. sIR alone data in this panel are the same as those in Figure 2C. *p<0.01 vs. sIR alone. No significant differences noted between acid/NMN/both groups.