SIRT3 is required for liver regeneration but not for the beneficial effect of nicotinamide riboside

Abstract

Liver regeneration is critical to survival after traumatic injuries, exposure to hepatotoxins, or surgical interventions, yet the underlying signaling and metabolic pathways remain unclear. In this study, we show that hepatocyte-specific loss of the mitochondrial deacetylase SIRT3 drastically impairs regeneration and worsens mitochondrial function after partial hepatectomy. Sirtuins, including SIRT3, require NAD as a cosubstrate. We previously showed that the NAD precursor nicotinamide riboside (NR) promotes liver regeneration, but whether this involves sirtuins has not been tested. Here, we show that despite their NAD dependence and critical roles in regeneration, neither SIRT3 nor its nuclear counterpart SIRT1 is required for NR to enhance liver regeneration. NR improves mitochondrial respiration in regenerating WT or mutant livers and rapidly increases oxygen consumption and glucose output in cultured hepatocytes. Our data support a direct enhancement of mitochondrial redox metabolism as the mechanism mediating improved liver regeneration after NAD supplementation and exclude signaling via SIRT1 and SIRT3. Therefore, we provide the first evidence to our knowledge for an essential role for a mitochondrial sirtuin during liver regeneration and insight into the beneficial effects of NR.

Keywords: Fatty acid oxidation; Hepatology; Metabolism; Mitochondria; Molecular pathology.

Figure 1. SIRT1 is not required for nicotinamide riboside to enhance liver regeneration.

Sirt1 liver-specific-KO mice (denoted as Sirt1^–/–) and floxed littermates lacking Albumin Cre (denoted as Sirt1^+/+) were subjected to partial hepatectomy (PHx) after 2 weeks of nicotinamide riboside (NR) supplementation. (A) Photographs of regenerating livers 36 hours after PHx. (B) Liver-to-body weight ratios at 24 hours (n = 4–7 per group) and 36 hours (n = 10–14 per group) after PHx. (C) Immunofluorescence showing proliferating hepatocytes (red) detected by EdU and counterstained with DAPI (blue). (D) Quantification of EdU-positive hepatocytes (n = 5 per group). (E) Hepatic lipid content as determined by triglyceride assay. (F) Representative liver sections stained with H&E. (G and H) Hepatic NAD and ATP content in PHx livers prior to and after 36 hours from control or NR-treated mice. High power field, 40× objective, 400× magnification for images in C and F. Data are shown as the mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; 1-way ANOVA followed by Tukey’s post hoc test (B, D, E, G, and H).

Figure 2. NR enhances mitochondrial respiration in regenerating liver and in primary hepatocytes.

Male C57BL6/J mice 10–16 weeks of age were supplemented for 2 weeks with NR at a dose of 500 mg/kg body weight and then were subjected to two-thirds partial hepatectomy (PHx). (A) Mitochondria from livers of H₂O- and NR-treated mice were isolated 24 hours after PHx (n = 6 per group) and were analyzed for fatty acid oxidation capacity using palmitoyl carnitine as a substrate. (B and C) Mitochondrial NAD and NADH content in organelles from control or NR-treated mice at 48 hours after PHx (n = 4–5 per group). Primary hepatocytes were isolated from overnight fasted C57BL6/J WT mice, cultured in glucose-free media with gluconeogenic substrates with or without NR for 5 hours. Mitochondria were then isolated from treated or untreated hepatocytes (n = 4–9 per group). (D) Oxygen consumption of primary hepatocytes. (E and F) Hepatocyte NAD and NADH content after treatment. (G and H) Final glucose and β-hydroxybutyrate concentration in the media. (I) Plasma β-hydroxybutyrate content at the indicated time points after PHx in H₂O- and NR-treated mice (n = 3 per group at 12 hours and n = 8–14 per group in all other time points). (J and K) Blood glucose and plasma lactate levels in prior to and 24 hours after PHx in H₂O- and NR-treated mice (n = 11–14 per group). (L) Mitochondrial NAD and NADH content in treated and untreated hepatocytes (n = 3 per group). (M) State 3-coupled mitochondrial oxygen consumption using pyruvate plus malate as substrates. In vivo data are represented by open circles and squares, and hepatocyte data are denoted by filled circles and squares. Data are shown as the mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; 2-tailed Student’s t test.

Figure 3. Sirt3 is essential for liver regeneration but is dispensable for the beneficial effects of NR.

Liver-specific Sirt3-KO mice (denoted as Sirt3^–/–) or floxed controls (Sirt3^+/+) were subjected to two-thirds PHx and analyzed 48 hours later (n = 5–10 per group). (A) Photographs of regenerating livers. (B) Liver-to-body weight ratios 48 hours after PHx. (C and D) Quantitation of mitosis as determined by counting mitotic figures in hepatocytes in H&E-stained sections under high power. Representation of the images are shown in (D). High power field, 40× objective, 400× magnification for images in D. (E) Hepatic triglyceride content in livers from H₂O- and NR-treated mice prior to and after PHx. (F and G) NAD and NADH content in livers from H₂O- and NR-treated mice prior to and after PHx. (H) ATP content in livers from H₂O- and NR-treated mice prior to and after PHx. Data are shown as the mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; 1-way ANOVA followed by Tukey’s post hoc test (B, C, and E–H).

Figure 4. Sirt3 and NR independently improve mitochondrial function.

(A and B) Primary hepatocytes were isolated from overnight fasted whole-body Sirt3-KO or -overexpressing mice and their respective WT littermates, cultured in glucose-free media with gluconeogenic substrates, and treated with or without NR for 5 hours before measuring oxygen consumption (n = 3 per group). Mitochondria were isolated from NR-treated or -untreated regenerating livers of liver-specific Sirt3-KO mice (n = 4–8 per group). (C and D) Mitochondrial respiration in organelles isolated from regenerating livers. (E and F) Mitochondrial NAD and NADH content 48 hours after PHx. (G) ³H-labeled palmitate oxidation in homogenates from livers 48 hours after PHx. (H) Plasma β-hydroxybutyrate concentration prior to (n = 9–15 per group) and 24 hours (n = 4 per group) after PHx. Data are shown as the mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; 1-way ANOVA followed by Tukey’s post hoc test (A–H).