Nicotinamide adenine dinucleotide (NAD) is an essential cosubstrate in biochemical reactions catalyzed by the sirtuins, poly(ADP-ribose) polymerase, and cyclic ADP-ribose cyclase/CD38 (1,2). In mam- mals, cellular NAD levels depend mainly on the salvage pathway, in which NAD is resynthesized from nicotinamide via nicotinamide mononucleotide (NMN) (3,4). Alternatively, NAD can be synthe- sized de novo from tryptophan or synthesized from nicotinamide ribose (NR), a trace nutrient in foods. NR is converted to NMN by NR kinase and then to NAD by nicotinamide mononucleotide ade- nylyltransferase (3,4).
The NAD-dependent protein deacetylase sirtuin-1 (SIRT1) in peripheral metabolic organs has been suggested as a key regulator of cellular meta- bolic processes. SIRT1 deacetylates important metabolic regulators, such as the forkhead box protein O1, peroxisome proliferator-activated recep- tor gamma (PPAR-γ), and PPAR-γ coactivator 1-alpha, thereby regulat- ing their activities or degradation. As in the peripheral organs, SIRT1 in the hypothalamic neurons plays an important role in the regulation of the energy balance and normal circadian behaviors (5-8). Therefore, reduced NAD levels and SIRT1 activity in both peripheral metabolic organs and the hypothalamus could lead to metabolic dysregulation and, in turn, con- tribute to the development of metabolic diseases.
The NAD levels in the liver, skeletal muscle, and white adipose tissues are altered by nutritional conditions. Caloric restriction or fasting increases tissue NAD levels, whereas intake of a high-fat diet (HFD) decreases NAD levels in these metabolic organs (9‒11). Chronic consumption of a high-fat, high-sucrose diet also reduces hypothalamic neuronal NAD content and SIRT1 activity (12). Similarly, SIRT1 expression and activ- ity in the hypothalamic arcuate nucleus (ARC) and suprachiasmatic nucleus (SCN) are decreased with aging. The aging-related reduction in hypothalamic SIRT1 has been suggested as a central mechanism for age-associated weight gain and circadian disorder (12,13).
Given the important roles of NAD and SIRT1 in normal cellular func- tion and physiology, trials of the treatment of tissue NAD depletion in obese and aged mice have been conducted by providing NAD precursors NMN and NR. These trials proved the beneficial metabolic effects of NMN and NR supplementation (11,14‒17). As extracellular NAD can be effectively imported into cells and can rescue cellular NAD deple- tion (18), we investigated the antiobesity effect of chronic NAD sup- plementation in HFD-fed obese mice. HFD feeding disrupts the diurnal rhythms of feeding behavior and locomotor activity (19). Therefore, we also investigated whether NAD supplementation could treat the altered daily rhythm of metabolic behaviors in diet-induced obese (DIO) mice.
NAD was purchased from Sigma-Aldrich (St. Louis, Missouri). FK866 was obtained from Enzo Life Sciences (Farmingdale, New York).
N1 murine hypothalamic neuron cells were obtained from Cedarlane Laboratories (Burlington, North Carolina) and were maintained in DMEM supplemented with 10% fetal bovine serum (FBS) and 5% pen- icillin/streptomycin for 14 passages before use. SH-SY5Y human neu- ron cells were purchased from ATCC (Manassas, Virginia) and were cultured in DMEM containing 10% fetal calf serum (FCS) for five passages.
Male C57BL/6N mice (Orient Bio Inc., Seongnam, Korea) were fed a standard chow diet (12.5% of calories from fat; #38057, Cargill Agri Purina, Inc., Seongnam, Korea) ad libitum, unless otherwise mentioned. To generate DIO mice, 8-week-old mice were fed a 60% HFD (60% of calories from fat; #D12492, Research Diets Co., New Brunswick, New Jersey) for 12 weeks. For chronic NAD treatment, NAD (1 mg/kg) was daily injected intraperitoneally 1 hour before lights-off for 4 weeks. The body weight and food intake were moni- tored daily before NAD injection. The mice were housed in an envi- ronment with controlled temperature (22°C±1°C) and a 12-hour light/ dark cycle from 8 am to 8 pm. All procedures were approved by the Institutional Animal Care and Use Committee of the Asan Institute for Life Sciences (Seoul, Korea).
Monitoring of diurnal behaviors
The daily rhythms of food intake, energy expenditure (EE), and loco- motor activity were determined by using a comprehensive lab moni- toring system (CLAMS; Columbus Instruments Inc., Columbus, Ohio). The animals were placed in the CLAMS chambers for 4 days, 2 days of acclimatization and 2 days of measurement. Data from the 2 days of measurement were used for generating the mean data and statistical analysis. During the CLAMS study, the animals did not receive NAD
injections in order to keep the metabolic chambers sealed. The light/ dark cycles were the same as in the home cages. Food was provided in powder form in the food container of CLAMS cages. One feeding bout was defined as food intake of >_ 0.2 kcal over 15 minutes.
Body temperature measurement
Rectal temperature was measured with a rectal probe connected to a digital thermometer (BAT-12 Microprobe Thermometer; Physitemp, Clifton, New Jersey) at 3 pm to 4 pm at room temperature during the fourth week of NAD treatment.
Conditioned taste aversion study
Mice were habituated to a daily 1-hour period of access to water for 7 days. On the eighth day, immediately following a 1-hour period of sac- charin exposure, the mice received intraperitoneal injections of saline, NAD (1 mg/kg), or lithium chloride (200 mg/kg; Sigma-Aldrich) as a positive control. For the next 3 days, a two-bottle choice test was con- ducted; in this test, the mice were allowed 1-hour access to one bottle of tap water and another bottle of 0.15% saccharin solution. Preference ratio was calculated as the intake of saccharin solution divided by the total intake of water and saccharin solution.
Measurement of blood glucose and cholesterol
An oral glucose tolerance test was performed during the second week of treatment. Glucose (1 g/kg) was orally administered in 5-hour fasted mice (from zeitgeber time (ZT) 1 [9 am] to ZT6 [2 pm]). Plasma glu- cose levels were measured using a glucometer (Accu-Chek; Roche Diagnostic, Mannhein, Germany) at indicated times. Plasma choles- terol concentrations were measured by using a Hitachi automatic ana- lyzer 7180 (Hitachi Ltd., Tokyo, Japan) at the end of treatment.
Measurement of NAD content
For the measurement of hypothalamic NAD content, the mice were sacrificed by decapitation at 2 pm to 3 pm after a 5-hour fast and 20 hours after the last NAD injection. The mediobasal hypothalamus (MBH) was dissected, snap frozen in liquid nitrogen, and stored at −70°C before assay. The hypothalami were thawed in 150 μL of 1M HClO4, lysed, and then neutralized by the addition of 50 μL of 3M K2CO2. After centrifugation (4°C, 13,000g) for 15 minutes, 20 μL of supernatant was loaded onto the high-performance liquid chroma- tography (HPLC) column (Apollo C18, 5 μm, 250×4.6 mm; Alltech, Deerfield, Illinois). HPLC was run at a flow rate of 1 mL/min. NAD was eluted as a sharp peak at 11 minutes. The amounts of NAD were normalized to wet tissue.
For the measurement of cellular NAD contents, cells were washed with phosphate-buffered saline (PBS) twice and then lysed using 100 μL of 1M HClO4 and neutralized by adding 33 μL of 3M K2CO2. After centri- fuging, 25 μL of supernatant was mixed with 175 μL of 50mM K2PO4/ KHPO4 (pH 7.0), and 100 μL of the mixture was loaded onto the HPLC column as described above.
For murine period circadian clock 1 (PER1) gene promoter analysis, N1 murine hypothalamic neuronal cells were transfected with pGL3- mPER1 promoter-luciferase construct (nucleotides −1,803 to +40) and/ or plasmids expressing pcDNA3.1-mBMAL1 (brain and muscle aryl hydrocarbon receptor nuclear translocator-like protein 1; GenBank accession number NM007489), pcDNA3.1-mCLOCK ((circadian lo- comotor output cycles protein kaput; GenBank accession number NM007715), and β-galactosidase (50 ng each). Forty-eight hours after
transfection, the cells were treated with NAD, FK866, or both (10nM each) for 2 hours before harvest.
To study the effect of clock genes on human neuropeptide Y (NPY) and Agouti-related protein (AgRP) promoter activity, SH-SY5Y human neuroblastoma cells were transfected with pGL3-hNPY-luc (pGL3-hNPY-luc; nucleotides −963 to +67) or pGL3-hAGRP-luc (pGL3-hAGRP-luc; nucleotides −1,000 to −1) in the presence or absence of plasmids expressing pSCT1-mPER1 (GenBank accession number NM011065), pSCT1-mPER2 (GenBank accession number NM011066), pcDNA3.1-mBMAL1, and pcDNA3.1-mCLOCK (50 ng each). At 48 hours after transfection, the cells were lysed and extracted for the measurement of luciferase activity by using a luminometer (PerkinElmer, Monza, Italy) and normalized to β-galactosidase activ- ity. The promotor analysis was repeated at least three times for each experiment.
Hypothalamic clock gene and neuropeptide expression
Mice were sacrificed at ZT4 (noon) and ZT14 (10 pm) in the freely fed condition. Hypothalamic SCN and ARC were obtained using the micropunch technique. In brief, mice were sacrificed by cervical dislocation, and whole brains were carefully collected and placed in the Harvard brain matrix (Harvard Apparatus, Inc., Holliston, Massachusetts). Three hypothalamic slices (1-mm thick) were ob- tained using razor blades and placed on a glass slide. SCN and ARC were punched using a 20-gauge blunted metal needle (inner dimeter 0.91 mm). RNA was extracted from hypothalamic tissues to measure clock gene and neuropeptide expression. Real-time polymerase chain reaction (PCR) analysis was performed using the following corresponding primers: NPY (5′-ggactgaccctcgctcta-3′ and 5′-tcgcagagcggagtagta-3′), AgRP (5′-gacgcggagaacgagact-3’,
The data are presented as mean + SEM. Statistical analysis was per- formed by using SPSS Statistics software version 22.0 (IBM Corp., Armonk, New York). The statistical significance among the groups was tested by using one-way analysis of variance (ANOVA) followed by a post hoc Fisher least significant difference test or unpaired t test, as appropriate. Repeated-measures ANOVA followed by a post hoc Bonferroni test was used to analyze the effect of chronic NAD supple- mentation on body weight changes. Significance was set at P<0.05. Daily rhythm variables of metabolic behaviors were obtained using cosinor analysis (cosinor package, R software version 3.4; The R Foundation, Vienna, Austria) (20). Diurnal rhythm parameters (mesor, amplitude, and acrophase depicted in Supporting Information Figure S1) between groups were compared by t test and Wald test followed by Bonferroni post hoc test. P < 0.0167 (0.05 divided by 3) was considered significant. Correlations between NPY and PER1 or AgRP and PER1 expression were tested using linear regression analysis. Results To investigate the effects of chronic NAD supplementation on obe- sity, we injected NAD (1 mg/kg) intraperitoneally into HFD-fed DIO mice for 4 weeks. We determined the dose of NAD (1 mg/kg/d) from a dose-response study conducted prior to the present study (Supporting Information Figure S2). NAD treatment for 4 weeks significantly inhib- ited weight gain in HFD-fed mice (Figure 1A; Supporting Information Figure S3). To further dissect the effect of NAD supplementation in the metabolic parameters affecting body weight, we first measured EE throughout a TABLE 1 Diurnal rhythm analysis of energy expenditure, locomotor activity, and food intake between lean mice and obese mice with or without NAD treatment 24-hour period and analyzed the diurnal pattern of EE. Compared with the lean controls, DIO mice showed a reduction in both daytime and night- time EE (Figure 1B). NAD supplementation in DIO mice significantly increased nighttime EE and tended to increase daytime EE (Figure 1B). Cosinor analysis of EE diurnal rhythms revealed that mesor (a rhythm-ad- justed mean) and amplitude (half the distance from the peak to the trough) values were significantly lower in DIO mice compared with lean mice (Table 1; Figure 1C). EE rhythm mesor values in NAD-treated DIO mice were lower than in lean mice but higher than in DIO mice. However, there was no significant difference in amplitude values between the DIO group and DIO with NAD group and in acrophase (time of the peak over 24-hour period) values between three groups. We next determined whether NAD supplementation might affect physical activity. Similar to EE, DIO mice showed a decrease in locomotor activity compared with lean mice (Figure 1D). The degree of reduction in locomo- tor activity was greater at nighttime than in the daytime. Moreover, diurnal rhythm of locomotor activity was suppressed in DIO mice (Figure 1D-1E). Notably, chronic NAD treatment significantly increased nighttime loco- motor activity in DIO mice and thereby recovered the suppressed diurnal rhythms in the locomotor activity of obese mice (Figure 1D-1E). Cosinor measures of diurnal rhythm of locomotor activity showed a greater reduc- tion in mesor and amplitude values in the DIO group than in the lean group, which was completely recovered by NAD treatment, but no alteration in acrophase values (Table 1; Figure 1E). We also studied the effects of NAD supplementation on other aspects of metabolism. Rectal temperature did not significantly differ between three groups (Supporting Information Figure S4), imply- ing that DIO and NAD supply by itself may not significantly affect body thermogenesis. NAD treatment in DIO mice induced a modest reduction in blood glucose levels during the oral glucose tolerance test (Figure 1F). Fasting plasma cholesterol levels were elevated in obese mice, and this elevation was not improved by NAD supple- mentation (Figure 1G). We next analyzed the feeding patterns in those mice. Both daytime and nighttime food intake was greater in DIO mice than in the lean controls (Figure 2A-2B). Lean mice consumed less than 20% of the total caloric intake during the daytime, which is a physiological sleeping period in rodents (Figure 2C). However, daytime caloric intake significantly increased in DIO mice (Figure 2C), which indicated the disruption of diurnal rhythm in the feeding patterns. Notably, NAD-treated obese mice showed a significant decrease in daytime food intake (Figure 2B); therefore, NAD supplementation partially recovered altered diurnal feeding patterns in obese animals (Figure 2C). On the other hand, NAD-untreated DIO mice showed an increase in the number of feeding bouts in both daytime and nighttime (Figure 2D), which suggested the fragmentation of feeding in obese animals. NAD treatment significantly decreased both daytime and nighttime feeding frequency (Figure 2D). Cosinor analysis of food intake diurnal rhythmicity indicated a signifi- cant rise in mesor values of DIO groups with or without NAD treatment compared with the lean group (Table 1; Figure 2E). Amplitude values showed no significant difference among the three groups. Acrophase values of NAD-treated DIO mice were higher than those of other two groups. We finally tested whether NAD decreases food intake through toxicity and adverse effects. For this, we performed a conditioned taste aversion (CTA) test. Intraperitoneal injection of lithium chloride, an established CTA inducer, decreased saccharine consumption, whereas the NAD (1 mg/kg) injection did not (Supporting Information Figure S5), suggest- ing that NAD did not induce CTA. Hematologic, chemical, and autopsy analysis also demonstrated no abnormality and toxicity in mice receiv- ing NAD (data not shown). A previous study showed that extracellular NAD treatment increased cellular NAD levels and rescued cells from NAD-depletion-induced death (21). Therefore, we investigated whether chronic NAD supple- mentation could influence NAD content in the MBH of mice at the end of the treatment. The MBH NAD contents were significantly lower in DIO mice than in the lean controls, and the obesity-related decrease in hypothalamic NAD was almost normalized by the supply of exogenous NAD (Figure 3A). Given the beneficial effect of exogenous NAD on the metabolic circadian rhythms, we investigated the effect of NAD treatment on transcription of the clock gene PER1 in hypothalamic neurons. As reported previously (22), PER1 promoter activity was significantly stimulated by the coexpres- sion of BMAL1 and CLOCK, which are key transcriptional regulators of clock genes (Figure 3B). To test the effect of altered cellular NAD levels on PER1 transcriptional activity, NAD depletion was induced by chemical Nampt inhibitor FK866 treatment (10nM) and NAD repletion by cotreat- ment of FK866 and NAD (10nM each) in N1 hypothalamic neuronal cells. FK866 treatment had no effect on basal PER1 transcriptional activity, but it significantly decreased CLOCK/BMAL1-stimulated PER1 transcriptional activity (Figure 3B). The cotreatment of NAD prevented FK866-induced suppression in PER1 transcriptional activity (Figure 3B). The simulta- neous measurement of cellular NAD contents confirmed a significant reduction by FK866 and its reversal by cotreatment of FK866 and NAD (Figure 3C). These results implied that exogenous NAD may treat NAD- depletion-associated clock gene dysregulation in hypothalamic neurons. Consistently, exogenous NAD significantly increased CLOCK/BMAL1- stimulated PER1 transcription in FK866-untreated cells (Figure 3D). To further investigate whether NAD treatment may affect hypotha- lamic PER1 mRNA expression in vivo, we collected the hypotha- lamic SCN and ARC at ZT4 (noon) and ZT14 (10 pm) at the end of NAD treatment. Consistent with a previous report (23), SCN PER1 mRNA levels in lean mice had a dramatic diurnal variation with much lower levels at ZT14 than at ZT4, which was well preserved in obese mice regardless of NAD treatment (Figure 3E). ARC PER1 mRNA also showed a diurnal change, albeit in a mild and opposite direction compared with that of SCN in lean mice. Notably, daily fluctua- tion in ARC PER1 expression was blunted in saline-injected DIO mice, which was reversed by NAD supplementation (Figure 3F). It is therefore possible that NAD treatment may improve blunted diurnal rhythms in metabolic behaviors through effects on molecular clocks in the hypothalamic ARC. The perturbation of circadian rhythm leads to metabolic syndrome, obesity, and diabetes (24), as clock genes are responsible for the cir- cadian regulation of the normal expression of metabolically important hormones and enzymes (8). Moreover, we observed that acute NAD treatment repressed the mRNA expression of anabolic neuropeptides NPY and AgRP but had no effect on the expression of the catabolic neu- ropeptide proopiomelanocortin in the mice hypothalamus (Supporting Information Figure S6). We also found that NAD treatment suppressed both NPY and AgRP promotor activities, whereas FK866-induced NAD depletion stimulated them (Supporting Information Figure S7). Therefore, we tested a possible link between clock genes and NPY or AgRP transcription. The promoter activities of NPY and AgRP were robustly inhibited by the overexpression of CLOCK/BMAL1 (Figure 4A-4B). In contrast, NPY promotor activities were stimulated by PER1 overexpression, and AgRP promotor activities were upregu- lated by PER1 and PER2 expression (Figure 4C-4D). exhibited diurnal changes in ARC NPY mRNA levels with elevated levels at midnight, and this daily fluctuation was abolished in DIO mice (Figure 4E). Despite reduced food intake, and in contrast to suppression of NPY transcriptional activity by acute NAD treatment, ARC NPY expression was not decreased by chronic NAD treatment. Instead, NAD treatment improved blunted NPY rhythm in obese mice (Figure 4E). ARC AgRP expression showed similar diurnal changes to those of NPY, although NAD treatment did not significantly rescue obesity-in- duced changes in ARC AgRP diurnal variations (Figure 4F). These data suggest that diurnal rhythms of ARC NPY and AgRP may be affected by obesity and exogenous NAD supply. We finally analyzed the correlation between ARC mRNA expression lev- els of NPY and PER1 or AgRP and PER1 at both ZT4 and ZT14. We found a significant positive correlation between NPY and PER1 and between AgRP and PER1 (Figure 4G-4H). These results substantiate interactions between clock gene PER1 and NPY and AgRP in the hypo- thalamic ARC. Discussion In the present study, we demonstrated that 4-week-treatment of NAD significantly attenuated weight gain in DIO mice without any Considering the robust interactions between clock genes and NPY and AgRP transcription in vitro, we examined diurnal fluctuations in NPY and AgRP mRNA expression in the hypothalamic ARC. Lean mice observable adverse effects. Moreover, chronic NAD supplementation significantly increased EE and locomotor activity and suppressed food intake in DIO mice. It also completely corrected the disrupted diurnal rhythm in locomotor activity and modestly improved glucose intolerance in obese mice. Previous studies have reported the benefi- cial metabolic effects of supplementation with NAD precursors NR and NMN in mice and humans (11,14‒17). Similar to our findings, 12-week supplementation with NR (400 mg/kg/d in drinking water) in HFD-fed mice reduced body weight gain and fat mass and increased EE (14). Twelve-month NMN administration in mice (100-300 mg/kg in drinking water) attenuated aging-associated body weight gain (16). Moreover, intraperitoneal administration of NMN (500 mg/kg/d) for 1 week improved glucose metabolism in diabetic mice (11) and restored mitochondrial functions in the muscles of aged mice (15). It is notable that supplementation with NAD by itself, at a 100-times lower dose compared with those of its precursors NMN and NR (25,26), caused a beneficial metabolic effect; however, the mechanism underlying this phenomenon remains to be addressed in the future.
Accumulated evidence has suggested a strong cross talk between the cel- lular clockwork and metabolic regulation (27). Mutation or knockout of CLOCK, BMAL1, PER1/PER2, and cryptochrome 1/2 (CRY1/CRY2) leads to metabolic disturbances in rodents (28‒31). Conversely, HFD dis- rupts normal circadian expression patterns of metabolic proteins in the hypothalamus, liver, and adipose tissue (19,32‒34). Moreover, there is a clear correlation between the consumption of a calorie-dense diet and disrupted feeding patterns. Rodents on a high-calorie diet have shown a fragmented feeding pattern indicated by small and frequent meals (18,19). In contrast, mice on a chow diet have shown a consolidated large meal pat- tern. In line with this, our study showed that feeding frequency increased throughout the day-night cycle in HFD-fed obese mice. In addition, obese mice exhibited increased caloric intake during the day, which is a phys- iological sleeping period in rodents and might also suggest disrupted or fragmented sleeping in these mice. All these findings suggest that obe- sity or intake of a fat-rich diet disturbs normal diurnal rhythms in feeding behavior.
Previous studies have demonstrated that intracellular NAD biology is closely linked to the cellular clockwork (8,12,35). NAD-dependent deacetylase SIRT1 is an active player in the cellular clockwork. SIRT1 binds to core clock regulators CLOCK/BMAL1 and clock repressor PER2 in a circadian manner, and it deacetylates BMAL1 and PER2 (35,36). On the other hand, levels of NAD and the NAD-biosynthetic enzyme Nampt in hepatocytes show rhythmic fluctuation with the cir- cadian cycle (35). Importantly, we found that NAD supplementation during HFD feeding significantly ameliorated obesity-related changes in day-night patterns in feeding and locomotor activity, indicating the potential usefulness of NAD supplements for the treatment of abnor- mal day-night feeding patterns and physical activity in obese individ- uals. Cosinor analysis of diurnal rhythms of metabolic behaviors also demonstrated that the disrupted diurnal pattern in locomotor activity of DIO mice was completely recovered by chronic NAD supplemen- tation. Interestingly, the effect of NAD treatment on diurnal rhythms of food intake and EE was smaller compared with its effects on the daily rhythm of locomotor activity. Therefore, NAD supply may have differential effects on the diurnal fluctuations of metabolic param- eters, though mechanisms behind this phenomenon are far from our understanding.
As for a possible mechanism of NAD effects on the day-night rhythm of metabolic behaviors, exogenous NAD supply to hypothalamic
neuronal cells potentiated BMAL1/CLOCK-stimulated PER1 tran- scription and reversed NAD-depletion-induced suppression of it. Moreover, chronic NAD supplementation treated DIO-induced bluntness in diurnal oscillation of ARC PER1 expression. These results indicate the hypothalamic ARC as a potential site of action of NAD treatment with respect to diurnal rhythm regulation. In con- trast, daily oscillation of PER1 expression in the SCN, where master clock neurons reside, was not significantly altered by obesity and NAD treatment. These findings are in line with a previous study that showed that the rhythmicity of core clock genes in the SCN was not affected by HFD (34).
In this study, we observed that clock genes mediated regulation of NPY and AgRP transcriptional activity and provided a novel mechanism for the cross talk between clock genes and systemic metabolic regu- lators. Indeed, NPY and AgRP transcriptional activities were potently inhibited by CLOCK/BMAL1 and stimulated by PER1 or PER2. These results indicate that exogenous NAD may control NPY and AgRP tran- scription through the cellular clock machinery. As PERs are not tran- scriptional factors (37), they may indirectly regulate NPY and AgRP transcription through inhibition of BMAL1/CLOCK. In line with the in vitro findings, there was a strong positive correlation between PER1 and NPY and AgRP expression in the hypothalamic ARC. Moreover, obesity- or NAD-induced changes in the diurnal pattern of NPY expres- sion were very similar to those observed in ARC PER1 expression. These findings further suggest the possibility that exogenous NAD may help the recovery of blunted diurnal fluctuations in metabolic behaviors in obese mice by potentiating the interactive regulation of clock genes and NPY and AgRP.
Consistent with a previous report (13), we observed a significant decrease in hypothalamic NAD levels in DIO mice, which was res- cued by NAD treatment. Disrupted circadian rhythms have also been found in previous studies conducted on human obesity, which can cause adverse effects on human health (38); therefore, our findings strongly suggest the therapeutic potential of NAD supplementation in patients with obesity who exhibit altered circadian behaviors.O
© 2018 The Obesity Society
We thank Dr. Joon Seo Lim from the Scientific Publications Team at Asan Medical Center for his editorial assistance in preparing this manuscript and Dr. Min-ju Kim from the Department of Clinical Epidemiology and Biostatistics at Asan Medical Center for diurnal rhythm analysis.