ABSTRACT
Rationale: Nicotinamide adenine dinucleotide (NAD+) is a substrate for sirtuin (SIRT) lysine deacylases. Stimulation of SIRT1 is cardioprotective against ischemia-reperfusion (IR) injury, prompting interest in orally-available NAD+ precursors such as nicotinamide mononucleotide (NMN) as potential cardioprotective agents. While the biological activity of NMN has been largely attributed to SIRT stimulation, NMN effects on metabolism, and any role these may play in cardioprotection, are less well understood. Objective: To investigate potential non-SIRT mechanisms for NMN cardioprotection, with a focus on metabolism. Methods & Results: NMN was protective in perfused mouse hearts (post-IR functional recovery: NMN 42±7% vs. vehicle 11±3%). However, protection was insensitive to the SIRT1 inhibitor splitomicin (recovery 47±8%), and NMN did not impact lysine acetylation in the cytosol where cardiac SIRT1 is located, thus suggesting NMN does not stimulate cardiac SIRT1 activity. Surprisingly, NMN was not protective in hearts perfused without glucose (palmitate as fuel source; recovery 11±4%). Since glycolysis requires NAD+, and is associated with some cardioprotective paradigms, we hypothesized NMN protection may be due to glycolytic stimulation. In primary cardiomyocytes, NMN induced cytosolic and extracellular acidification, and enhanced lactate generation, indicative of increased glycolysis. Finally, since extension of ischemic acidosis into early reperfusion (i.e., acid post-conditioning) is cardioprotective, we hypothesized that NMN delivery at reperfusion may protect, and indeed this was the case (recovery 39±8%). Conclusions: The acute cardioprotective benefit of NMN is mediated via glycolytic stimulation, and this effect of NMN may be worthy of investigation in other situations where NAD+ precursor supplements are of therapeutic interest.
INTRODUCTION
The SIRT family of NAD+ dependent lysine deacylases are important regulators of metabolic health 1,2. SIRT activity is known to decline with age 3, and as such a number of NAD+ biosynthetic precursors are under investigation as nutriceuticals, aimed at ameliorating diseases of aging 4,5. Among these compounds, nicotinamide riboside (NR) and nicotinamide mononucleotide (NMN) are orally bioavailable and are currently the subject of clinical trials (NCT02812238, NCT02835664, NCT02303483, NCT02678611, etc)4,6.
Cardiomyocyte SIRT1 is important in several cardioprotective paradigms, including ischemic preconditioning (IPC) 7–9. Recently it was shown that NMN confers cardioprotection in a mouse model of ischemia-reperfusion (IR) injury, and this protection was lost in Sirt1-/- mice 10. Although thus far the biological activity of NMN has been mostly attributed to the SIRT stimulation11, the NAD+/NADH redox couple is critically important for metabolic pathways including glycolysis and mitochondrial oxidative phosphorylation 12. The acute effects of boosting cellular NAD+ levels on these pathways are poorly understood, and the role that such metabolic perturbations may have in the protective effects of NMN are unknown. Herein, we investigated alternative cardioprotective mechanisms of NMN beyond SIRT1.
METHODS
Chemicals and reagents were from Sigma (St. Louis MO). Wild-type male C57BL6/J mice were bred in-house, handled according to the "NIH Guide" with food and water ad libitum, and used at age 8-12 weeks. Following tribromoethanol anesthesia (200mg/kg ip), hearts were perfused in Langendorff constant flow mode (4 ml/min.) as previously described 9. Krebs Henseleit (KH) buffer was supplemented with 5 mM glucose plus 100 μM palmitate-BSA, or palmitate-BSA alone (glucose-free) 9. NMN was delivered via a port above the perfusion cannula from a 100x stock in KH. Hearts were freeze-clamped in liquid N2 and cardiac metabolites extracted in 80% MeOH for LC-MS/MS based metabolomics as previously described 13. Primary cardiomyocytes were isolated by collagenase digestion, and intracellular pH measured using fluorescence microscopy with the pH sensitive probe BCECF, as previously described 14,15. Cardiomyocyte oxygen consumption and extracellular acidification were measured in a Seahorse™ XF24 analyzer as previously described 14. Statistical significance between groups was determined by AnOvA assuming a normal distribution followed by post-hoc Student’s t-test (significance threshold p<0.05).
RESULTS
We recently showed that the SIRT1 inhibitor splitomicin (Sp) causes mild cellular alkalinization in primary mouse cardiomyocytes 16. In companion experiments, we also tested the effects of the SIRT1 activator NMN, and surprisingly found that 1 mM NMN induced cellular acidification (Figure 1A/B). NAD+ is a substrate for gyceraldehyde-3-phosphate dehydrogenase (GAPDH) in glycolysis, and under normal or hypoxic conditions either the malate/aspartate shuttle or lactate dehydrogenase respectively serve to regenerate NAD+ from NADH, to permit continued glycolysis and prevent reductive stress 17. Given this dependence of glycolysis on NAD+ availability, we hypothesized that NMN-induced acidification may be due to glycolytic stimulation. Figure 1C shows that delivery of 1 mM NMN to perfused mouse hearts resulted in a significant elevation of NAD+ and a similar increase in NADH, the latter suggesting NAD+ reduction by metabolism. NMN also drastically elevated cardiac lactate, with a moderate but significant rise in pyruvate. Furthermore, Figure 1D shows that 1 mM NMN elicited a ~3-fold increase in cardiomyocyte extracellular acidification rate (ECAR), without significantly impacting oxygen consumption rate (OCR). Notably, although SIRT1 is known to regulate glycolysis 18, splitomicin was without effect on NMN-induced ECAR enhancement (Figure 1D). Together these data indicate that NMN stimulates cardiomyocyte glycolysis in a SIRT1 independent manner.
Although enhanced glucose utilization is associated with cardiac pathology and heart failure19–23, more recently parallels have been drawn between cardiac glycolysis and hypoxic survival 13,24. Thus we hypothesized NMN-induced glycolytic stimulation may play a role in the reported cardioprotective benefits of this molecule 10. Figure 2A/B shows that pre-ischemic delivery of 1 mM NMN afforded significant cardioprotection against IR injury (post-IR functional recovery: NMN 42±7% vs. vehicle 11±3%; infarct size: NMN 33±5% vs. vehicle 67±5%; means ± SEM, n=5-6, p<0.05 for both parameters). However, in contrast with previous reports of a requirement for SIRT1 10, we observed no effect of splitomicin on NMN-induced cardioprotection (p=0.64 vs. NMN alone). This result is consistent with the lack of splitomicin effect on NMN-induced glycolytic stimulation (Figure 1D). Notably, splitomicin inhibits cardioprotection by IPC9, suggesting the inhibitor is functional in this system. Furthermore, cardiac SiRT1 is primarily cytosolic 25, and acetyl-lysine western blots revealed no effect of NMN on cytosolic lysine acetylation (Figure 2C). Together these results suggest no role for SIRT1 in NMN-induced cardioprotection. Although this contrasts with reports of a requirement for SIRT1 in NMN protection 10, Sirt-/- mice show enhanced baseline IR injury (~50% greater infarct vs. wild-type) which may over-ride the ability of NMN to protect via other mechanisms.
To test the requirement for glycolysis in NMN-induced cardioprotection, hearts were perfused without glucose (i.e., with palmitate as the sole metabolic substrate). Figure 3A/B shows that such restriction of glucose availability ablates the protective benefit of NMN. Importantly, glucose withdrawal had no significant effect on baseline injury alone (infarct 67±5 % with glucose plus fat, vs. 73±4% with fat only), indicating this loss of NMN cardioprotection was not due to hearts being damaged beyond repair.
A key event in IR injury is the opening of the mitochondrial permeability transition (PT) pore. During ischemia, acid pH maintains the pore in a closed state, but pH recovery upon reperfusion triggers pore opening 26–28. As such, reperfusion with acidic media (i.e., acid postconditioning) is cardioprotective via maintaining PT pore closure into early reperfusion 29. Since acidosis is a key effect of NMN (Figure 1), we hypothesized that NMN delivery at reperfusion may be cardioprotective. Figure 3C/D shows that, indeed, 1 mM NMN at reperfusion was significantly protective (recovery 39±8%, infarct 28±5%, means ± SEM, n=6). Overall, our data suggest that NMN acutely stimulates cardiac glycolysis, which may contribute to its cardioprotective effects.
DISCUSSION & CONCLUSIONS
The heart is conventionally viewed as a “metabolic omnivore”, but under normal conditions the bulk of cardiac ATP requirements are furnished by mitochondrial β-oxidation of fatty acids 19. As such, a switch to favor glucose utilization is typically associated with cardiac pathology and heart failure 19–23. However, several studies have also linked glycolysis to cardioprotection: A screen for compounds that enhance glycolysis yielded hits that were protective in models of cardiac IR injury 24. Glycolysis is also up-regulated in IPC 13,30 and the presence of glucose is necessary for IPC 13. Furthermore, the pH hypothesis of ischemic post-conditioning (IPoC) posits that protection by IPoC is afforded by extending ischemic acidosis into early reperfusion, maintaining PT pore closure 29 Together with our data showing NMN stimulates glycolysis, these findings suggest that NMN-induced cardioprotection may proceed via metabolic acidosis.
While acidosis is a response of all eukaryotes to hypoxia/ischemia, the potential of acidic pH to drive protective molecular signaling events has only recently become appreciated. For example, acidic pH imparts de-novo activities to several metabolic enzymes, resulting in generation of unique metabolites 15,31. In particular, the oncometabolite 2-hydroxyglutarate (2-HG) is generated by lactate and malate dehydrogenases under acidic conditions 15,31. 2-hG is a competitive inhibitor of the α-ketoglutarate dependent dioxygenase family of epigenetic regulators, which includes the JmjC domain-containing histone demethylases, the TET 5-methylcytosine hydroxylases, and the EGLN prolyl-hydroxylases that regulate hypoxia inducible factor (HIF)32,33. As such, acid pH can induce HIF-1α via a pathway that requires 2-HG generation 15.
The observation that NMN rapidly induces metabolic acidosis raises the possibility that at least some of the benefits reported for dietary NMN supplementation 4–6,12 may be attributed to such effects. A wide variety of health benefits are reported for NMN and NR supplementation, including neurological, cardiac, obesity/diabetes-related, and anti-aging 12. In many cases, such benefits may be linked to enhancing glycolysis. For example, NMN treatment increases performance in whole body glucose tolerance testing 34. Similarly, NR enhances stem cell function, with stem-ness being generally associated with a glycolytic metabolic state 5.
Despite the reported benefits of dietary NMN/NR supplementation, our results urge caution regarding the widespread human use of these nutriceuticals. For example, the loss of SIRT1 activity in aging is associated with HIF stabilization and a pseudo-hypoxic metabolic state 35. As such, acute acidosis resulting from NMN supplementation may activate HIF, worsening pseudohypoxia. Similarly, HIF activation and metabolic acidosis 36 are well-known hallmarks of cancer, and indeed the same screen that identified glycolytic stimulators as protective against hypoxia found numerous anti-cancer drugs as glycolytic suppressors 24. Acid pH is also known to promote the reverse reaction of isocitrate dehydrogenase (i.e., reductive carboxylation of α-ketoglutarate to citrate), which is an important driver of lipid biosynthesis in cancer. Together with the apparent ability of NMN to promote stem-ness 5, these findings suggest that NAD+ boosting supplements may promote tumor growth.
Overall our results suggest that the nutriceutical use of NMN and NR should not overlook the classical role that NAD+ plays in glycolysis. GAPDH is one of the most highly and stably expressed proteins in the cell, such that it is often used as a “housekeeping protein” for experimental loading controls. The levels of GAPDH are likely to be several orders of magnitude higher than those of NAD+ consuming signaling enzymes such as SIRTs, PARPs, and CD38. Thus, the biological effects of large-scale and acute elevations in [NAD+] are likely to involve glycolytic stimulation. It remains to be determined whether long term beneficial effects of NAD+ precursors are mediated by repetitive transient metabolic acidosis.
DISCLOSURES
The authors declare no conflicts of interest.
ACKNOWLEDGEMENTS
This work was funded by grants from the US National Institutes of Health: RO1 HL-071158 (to PSB) and R01-HL127891 (to PSB and KWN). We thank Xenia Schafer (URMC Biochemistry) for technical support.