Overexpression of Nmnat3 efficiently increases NAD and NGD levels and ameliorates age‐associated insulin resistance

Summary Nicotinamide adenine dinucleotide (NAD) is an important cofactor that regulates various biological processes, including metabolism and gene expression. As a coenzyme, NAD controls mitochondrial respiration through enzymes of the tricarboxylic acid (TCA) cycle, β‐oxidation, and oxidative phosphorylation and also serves as a substrate for posttranslational protein modifications, such as deacetylation and ADP‐ribosylation by sirtuins and poly(ADP‐ribose) polymerase (PARP), respectively. Many studies have demonstrated that NAD levels decrease with aging and that these declines cause various aging‐associated diseases. In contrast, activation of NAD metabolism prevents declines in NAD levels during aging. In particular, dietary supplementation with NAD precursors has been associated with protection against age‐associated insulin resistance. However, it remains unclear which NAD synthesis pathway is important and/or efficient at increasing NAD levels in vivo. In this study, Nmnat3 overexpression in mice efficiently increased NAD levels in various tissues and prevented aging‐related declines in NAD levels. We also demonstrated that Nmnat3‐overexpressing (Nmnat3 Tg) mice were protected against diet‐induced and aging‐associated insulin resistance. Moreover, in skeletal muscles of Nmnat3 Tg mice, TCA cycle activity was significantly enhanced, and the energy source for oxidative phosphorylation was shifted toward fatty acid oxidation. Furthermore, reactive oxygen species (ROS) generation was significantly suppressed in aged Nmnat3 Tg mice. Interestingly, we also found that concentrations of the NAD analog nicotinamide guanine dinucleotide (NGD) were dramatically increased in Nmnat3 Tg mice. These results suggest that Nmnat3 overexpression improves metabolic health and that Nmnat3 is an attractive therapeutic target for metabolic disorders that are caused by aging.


Isolation of mitochondria
Isolation of mitochondria from mouse tissues was described elsewhere. In brief, skeletal muscles were excised from hind limb, and were homogenized in Buffer MA (67mM Sucrose, 50mM Tris-HCl pH7.4, 50mM KCl and 10mM EDTA), followed by the centrifugation by the same scheme described above. The pellet was dissolved in Buffer MB (250mM Sucrose, 10mM Tris-HCl pH7.4 and 3mM EGTA) and used as mitochondria.

Real-time quantitative PCR
Total RNAs were extracted from mouse tissues using TRI Reagent (Molecular Research Center, Inc.). cDNA was prepared using ReverTra Ace qPCR RT Master Mix with gDNA Remover (Toyobo, Japan) according to the supplier's protocol. Real-time PCR was carried out using THUNDERBIRD SYBR qPCR Mix (Toyobo) on Thermal Cycler Dice Real Time System II (Takara Bio).
Quantification was done by Delta Delta Ct method, and Rpl13a gene was used as a reference gene.

A glucose tolerance test (GTT) and an insulin tolerance test (ITT)
For the GTT experiments, the mice were fasted for 16 h and then injected with glucose (1g/Kg body weight) intraperitoneally. For the ITT experiments, the mice were fasted for 4 h, and then intraperitoneally injected with human insulin (0.8U/Kg body weight and 0.3U/Kg for HFD and normal chow fed mice, respectively). The blood glucose concentration was measured using an automatic blood glucose meter (NOVA Biomedical). The serum insulin levels were determined by the Mouse Insulin ELISA KIT (AKRIN-031; Shibayagi).

ROS measurement
Amount of ROS was determined using Oxiselect TM In Vitro ROS/RNS Assay Kit (Cell Biolabs Inc.) according to supplier's manual. Tissue lysate were grinded by Multibeads shocker (Yasui Kikai) in PBS with 1% NP40 and 0.1% SDS. For measurement, samples were diluted to ten times with PBS, and ROS amount was calculated using hydrogen peroxide standard curve.

Complex I activity assay
The activity of complex I was measured using MitoCheck® Complex I Activity Assay Kit (Cayman) according to the manufacture's instruction. In this experiment, bovine heart mitochondria were incubated with NADH in the presence or absence of 1mM NGD or 1µM rotenone. The absorbance of NADH (340 nm) was monitored using Varioskan multi plate reader (Thermo). The slope was calculated from linear portion of plotted absorbance. The activity was represented as relative to the value of slope in the control (absence of inhibitor). Tg mice. Rpl13a gene was used as a reference gene, and data are presented as a relative value to WT for each gene (n=5 for each group).

Supplemental Figure 2
Food intake was calculated from 7-week old WT and Nmnat3 Tg mice fed on NC or HFD. Values were average of four mice in each group. gene was used as a reference gene, and data are presented as a relative value to WT for each gene (n=8 for each group).

Supplemental Figure 4
(A and B) Trend chart of oxygen consumption (A) and carbon dioxide production (VCO 2 ) (B) was represented. Data were evaluated using 7-month old female Nmnat3 Tg and wild-type (WT) mice (n=8 for each group). (C) Trend chart of locomotor activity was represented. Data were evaluated using 7-month old female Nmnat3 Tg and wild-type (WT) mice (n=8 for each group).

Supplemental Figure 5
In virto complex I activity assay using bovine heart mitochondria. 1 mM NGD or 1 µM rotenone, a well-known complex I inhibitor, was used for this experiment.
Data were obtained from three independent experiments, and presented as mean ± SD. Single (*) and double (**) asterisk indicated that p-value was less than 0.05 and 0.01, respectively.

Supplemental Figure 6.
Gain-of-function Nmnat3 exhibited the metabolically beneficial effects in vivo.
Schematic of the proposed protection mechanism against aging-associated insulin resistance in Nmnat3 Tg mice.