Nicotinic acid promotes sleep through prostaglandin synthesis in mice

Nicotinic acid has been used for decades for its antiatherogenic properties in humans. Its actions on lipid metabolism intersect with multiple sleep regulatory mechanisms, but its effects on sleep have never been documented. For the first time, we investigated the effects of acute systemic administration of nicotinic acid on sleep in mice. Intraperitoneal and oral gavage administration of nicotinic acid elicited robust increases in non-rapid-eye movement sleep (NREMS) and decreases in body temperature, energy expenditure and food intake. Preventing hypothermia did not affect its sleep-inducing actions suggesting that altered sleep is not secondary to decreased body temperature. Systemic administration of nicotinamide, a conversion product of nicotinic acid, did not affect sleep amounts and body temperature, indicating that it is not nicotinamide that underlies these actions. Systemic administration of monomethyl fumarate, another agonist of the nicotinic acid receptor GPR109A, fully recapitulated the somnogenic and thermoregulatory effects of nicotinic acid suggesting that they are mediated by the GPR109A receptor. The cyclooxygenase inhibitor indomethacin completely abolished the effects of nicotinic acid indicating that prostaglandins play a key role in mediating the sleep and thermoregulatory responses of nicotinic acid.


Results
Systemic administration of nicotinic acid elicits robust non-rapid-eye movement sleep (nReMS) increases. Intraperitoneal (ip) injection of nicotinic acid brought about marked, dose-dependent increases in NREMS (Fig. 1, Table 1). Intraperitoneal administration of 50 mg/kg nicotinic acid did not affect sleep-wake activity, body temperature or motor activity. The effects of 100 mg/kg nicotinic acid were manifested after a latency of one hour; starting from h 2, NREMS increased 41% above baseline for 4 hours (NREMS amounts in h 2-5, baseline: 82.7 ± 7.3 min, treatment: 116.8 ± 9.6 min, p < 0.01). Administration of 250 mg/kg nicotinic acid elicited robust and long-lasting increases in NREMS; NREMS was elevated 120% above baseline in the 2-11 h time block (baseline: 179.0 ± 7.1 min, treatment: 394.6 ± 20.4 min, p < 0.001). The effects of oral administration of 1 g/kg nicotinic acid were similar to those of the highest ip dose. NREMS increased by 86% in the 2-11 h time block, then returned to baseline in the latter half of the recording period (baseline: 197.9 ± 8.1 min, treatment: 368.9 ± 25.0 min, p < 0.001). NREMS increases were accompanied by decreased electroencephalographic (EEG) slow-wave activity (SWA) and suppressed motor activity (Fig. 1, Table 1).
Nicotinic acid had only minor effects on rapid-eye movement sleep (REMS) without any apparent dose-dependency. Oral gavage administration of 1 g/kg or ip injection of 100 mg/kg nicotinic acid did not affect REMS, while 50 and 250 mg/kg had a slight, but statistically significant effect (Table 1, Fig. 1). After 50 mg/kg nicotinic acid, REMS was elevated in h 21-23, while after 250 mg/kg, it was slightly below baseline in h 12 and 20.
On the recovery days following the ip administration of 250 mg/kg or the gavage administration of 1 g/kg nicotinic acid, sleep-wake activity returned to baseline with no significant rebound responses (for ip 250 mg/ kg nicotinic acid, wakefulness baseline: 482.2 ± 9.1 and 208.6 ± 7.8 min/12 h during the dark and light periods, Indomethacin blocks the effects of nicotinic acid on sleep and body temperature. To examine the role of endogenous prostaglandins in the sleep and thermoregulatory effects of nicotinic acid, mice were pretreated with 5 mg/kg indomethacin, a cyclooxygenase inhibitor, before the administration of 100 mg/kg nicotinic acid. Indomethacin pretreatment completely abolished the effects of nicotinic acid on sleep, EEG SWA, body temperature and motor activity (Fig. 2, Table 2). Indomethacin, when administered by itself, did not have any significant effect on sleep and body temperature (data not shown).
Decreased body temperature is not the cause of nicotinic acid-induced sleep. To investigate if decreased body temperature, per se, could account for increased sleep in response to nicotinic acid, mice were pretreated with CL-316,243, a β3-AR agonist, to stimulate thermogenesis thereby preventing hypothermia. CL-316,243 pretreatment completely abolished the hypothermic response to 250 mg/kg nicotinic acid but did not attenuate the late increases in body temperature in the second half of the recording period (Table 3, Fig. 3). The NREMS-promoting effects of nicotinic acid were not attenuated by the β3-AR agonist pretreatment, and sleep was increased by 134% in the 2-11 h time block (NREMS amounts in the 2-11 h time block; baseline: 179.0 ± 7.1 min, CL-316,243 + nicotinic acid treatment: 394.6 ± 20.4 min, p < 0.001). Likewise, the motor activity and EEG SWA suppressions after nicotinic acid administration were not affected by the CL-316,243 pretreatment.

Systemic administration of nicotinamide does not affect sleep amounts and body temperature.
To test the possibility that its conversion into nicotinamide underlies the sleep-promoting effects of nicotinic acid, we investigated the effects of nicotinamide on sleep and body temperature. Systemic injection of 100 mg/kg (data not shown) or 250 mg/kg nicotinamide did not have any effect on sleep time and body temperature (Fig. 4, Table 4). The higher dose slightly decreased motor activity and suppressed EEG SWA. nicotinic acid suppresses Vo 2 , respiratory exchange ratio (RER) and feeding. To gain further insight in the action of somnogenic doses of nicotinic acid on metabolism, we investigated the effects of ip and oral administration of nicotinic acid on O 2 consumption (VO 2 ), RER and food intake. Intraperitoneal injection of 50 mg/kg nicotinic acid did not have any effect on these measures, while 250 mg/kg nicotinic acid induced marked and long-lasting suppression of VO 2 , RER and feeding. Similar robust effects were observed after gavage administration of 1 g/kg nicotinic acid (Fig. 6, Table 6).

Discussion
Nicotinic acid was introduced into clinical therapy more than 60 years ago to improve plasma lipid profile, but its effects on sleep and body temperature have never been documented. Our main finding is that systemic injection or oral gavage administration of nicotinic acid induces robust increases in NREMS in mice. The effects were dose-dependent, ip administration of 100 and 250 mg/kg doses increased NREMS 41-120% above baseline for 4-11 h. Oral administration of 1 g/kg nicotinic acid had similar sleep-promoting potency as the 250 mg/kg injected ip. The effects were specific to NREMS, as REMS responses were minor and did not show dose-dependency. Nicotinic acid is part of a group known as B3 complex, which also includes nicotinamide and nicotinamide riboside. After ip administration, 25% of the injected nicotinic acid is converted to nicotinamide in the liver in mice 43 . To determine if the acute sleep-promoting actions of nicotinic acid could be due to its conversion to nicotinamide, we also investigated the effects of bolus injection of nicotinamide. Intraperitoneal injection of 100 and 250 mg/kg nicotinamide did not have any significant effect on sleep amounts or body temperature, thus, it is unlikely that the sleep-promoting and thermoregulatory actions of nicotinic acid are mediated by its conversion to nicotinamide. Further, portions of circulating nicotinic acid and nicotinamide are converted to nicotinamide www.nature.com/scientificreports www.nature.com/scientificreports/ adenine dinucleotide (NAD) in the liver 44 . It has been reported that metabolic signals from the hepatoportal region modulate sleep 17,42 . Since the conversion of nicotinamide to NAD is significantly more efficient than that of nicotinic acid 45 , but nicotinamide lacks sleep-inducing activity, it is unlikely that nicotinic acid-induced increases in liver NAD content are responsible for the sleep-promoting effects.
Nicotinic acid induced significant decreases in body temperature. This can be explained, in part, by increased heat loss through cutaneous vasodilation. Nicotinic acid-induced vasodilation has been described in guinea pigs 46 Table 3. CL-316,243 pretreatment followed by administration of nicotinic acid. NREMS, REMS, body temperature, motor activity EEG SWA: statistical results. www.nature.com/scientificreports www.nature.com/scientificreports/ rats 47 and mice 34 and it is the cause of flushing, the well-known side effect of nicotinic acid megadoses in humans 48 . Vasodilation is a prompt and transient response to nicotinic acid [48][49][50] . In mice, vasodilation induced by subcutaneous injection of 100-300 mg/kg nicotinic acid reaches a peak within 2-3 min, and by 10 min the effect is reduced by 50% 34 . The sleep-inducing and the hypothermic actions of nicotinic acid, however, lasted for ~7-11 hours, thus it is unlikely that they are the direct consequences of the actions of nicotinic acid on cutaneous blood flow.
For thermoregulation, mice rely more on metabolic thermogenesis rather than vasomotor mechanisms 51 , and it is likely that the hypothermia is due to suppressed metabolic heat production. In fact, metabolic heat production (VO 2 ) was suppressed in response to nicotinic acid in our study. Consistent with our finding, 200 mg/kg nicotinic acid attenuates cold-induced increases in thermogenesis in mice 52 and norepinephrine-induced 53 or spontaneous 54 thermogenesis in rats. A major factor contributing to decreased thermogenesis is suppressed feeding in mice. Food intake was reduced by ~75% after 250 mg/kg ip and 1 g/kg oral doses of nicotinic acid in the first 8 hours, when VO 2 and sleep responses were the most pronounced. Similar reductions in feeding were also reported after repeated nicotinic acid treatment in mice 55 . Since naturally occurring NREMS is associated by decreased energy expenditure and body temperature 56 , it is also possible that the hypothermic response is simply the thermic manifestation of enhanced NREMS after nicotinic acid treatment. Similar robust ~4 °C hypothermia was observed during enhanced sleep in response to the chemogenetic stimulation of the ventrolateral preoptic sleep center 57 .
There is a complex relationship between sleep and feeding. For example, sleep loss increases feeding 58,59 and increased food intake induces postprandial sleep 2,3 . Since short-term food deprivation elicits robust increases in wakefulness and motor activity in mice [5][6][7] , it is unlikely that increased sleep and suppressed motor activity after nicotinic acid treatment are regulated energy-saving mechanisms in response to suppressed caloric intake. It is possible that the primary effect of nicotinic acid is increased sleep, and since sleep and feeding are mutually exclusive behaviors, decreased feeding may be the consequence of the general behavioral inactivity. Alternatively, the effects of nicotinic acid on sleep and feeding may be independent and mediated by different mechanisms.
It has been proposed that changes in body temperature may trigger somnogenic brain areas to initiate sleep 60 . To ascertain whether hypothermia per se could be a factor in the sleep-inducing effects of nicotinic acid, we also tested www.nature.com/scientificreports www.nature.com/scientificreports/ the effects in mice after the stimulation of thermogenesis by CL-316,243. CL-316,243 pretreatment completely abolished the hypothermic response, but the NREMS-promoting effects of nicotinic acid were not attenuated indicating that decreased body temperature, by itself, cannot account for the somnogenic effects of nicotinic acid.
Most, but not all, of the effects of nicotinic acid are mediated by GPR109A, the high-affinity nicotinic acid receptor [61][62][63] . To explore the role of the receptor in the sleep and thermoregulatory responses to nicotinic acid, we also used another receptor agonist, MMF 64 . MMF treatment fully recapitulated the effects of nicotinic acid on sleep, EEG SWA, body temperature and motor activity. Further, butyrate, another ligand for the GPR109A receptors as well as for the FFAR2, FFAR3 receptors, has similar marked NREMS-promoting and hypothermic effects as nicotinic acid 42 . These findings are consistent with the notion that the effects on sleep and temperature are mediated by the GPR109A receptor.
Nicotinic acid and MMF are potent stimuli for prostaglandin production via the activation of GPR109A receptors [65][66][67] . Ingestion of nicotinic acid results in 400 to 800-fold increases in PGD 2 plasma levels in humans 33 , and systemic injection of 100-300 mg/kg leads to 1.5-4-fold increases in mice 34 . Since PGD 2 has been identified as one of the most potent endogenous sleep-promoting molecules 35 , we sought to assess the role of prostaglandins in nicotinic acid-induced sleep by using the cyclooxygenase inhibitor indomethacin. Indomethacin pretreatment completely abolished the sleep-promoting as well as thermoregulatory actions of nicotinic acid indicating the involvement of   www.nature.com/scientificreports www.nature.com/scientificreports/ endogenous prostaglandins both in the somnogenic and hypothermic effects. The only known significant sources of nicotinic acid-induced prostaglandin production are Langerhans cells and keratinocytes in the skin 66,67 . The ventral surface of the rostral basal forebrain has been identified as a PGD 2 -sensitive sleep-promoting site, and it is possible that peripherally produced prostaglandins reach this structure. Alternatively, prostaglandins may act on a peripheral target to induce sleep. For example, vagus afferents express EP3 and EP4 prostaglandin receptors 68,69 and prostaglandins activate vagal sensory nerves [70][71][72] . It is known that vagal afferents carry sleep-inducing signals [73][74][75][76] .
Nicotinic acid, unlike nicotinamide, does not readily pass between blood and the cerebrospinal fluid [77][78][79] and the relatively small amounts that enter the brain are promptly converted to nicotinamide 77 . Therefore, it is less likely that centrally-produced prostaglandins underly the sleep-inducing actions of nicotinic acid, although this possibility cannot be ruled out with certainty.
Nicotinic acid has complex actions on lipid metabolism. It is considered an antilipolytic agent because it suppresses FFA release from adipocytes in vitro and causes transient decreases in circulating FFA levels in vivo 31 . However, decreased FFA levels are followed by robust rebound increases 2-4 h after a single bolus treatment in humans [80][81][82] , dogs 83 and rats 83,84 . Since the antilipolytic actions of nicotinic acid are independent of prostaglandins 85 , but the sleep-promoting actions require intact prostaglandin synthesis, it is unlikely that the antilipolytic actions of nicotinic acid directly contribute to its sleep effects. Independent of its lipolytic actions, nicotinic acid   also promotes complex changes in plasma lipid profile, an action that is thought to underlie its antiatherogenic properties 86 . It cannot be ruled out that these changes may also contribute to metabolic signaling for sleep. The extent to which our findings in mice are also applicable to humans is not known. It has been argued that the dose range used in our study corresponds well to the doses used in humans, if the 10-times faster metabolism of mice as compared to humans is considered 87 . The use of nicotinic acid supplements is widespread in the United States, and they are often advertised as sleep aids. In human trials, the most common reason for withdrawal is flushing followed by fatigue/sleepiness [88][89][90][91] . Our findings together with the above observational evidence point to the need of characterizing the effects of nicotinic acid on sleep in humans.

Methods
Animals. All procedures involving the use of animals followed the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All animal husbandry and experimental procedures were carried out in accordance with the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) and approved by the Institutional Animal Care and Use Committee (IACUC) of the Washington State University (protocol number 6031). Breeding pairs of C57BL/6J mice were purchased from The Jackson Laboratories, Inc. and were bred in-house at Washington State University. During the experiments, the animals were housed individually in temperature-controlled (29 ± 1 °C), sound-attenuated environmental chambers on a 12:12-hour light-dark cycle (lights on at 3 AM). Food (Harlan Teklad, Product no. 2016) and water were available unrestricted throughout all experiments.
Surgery. Male mice were surgically instrumented using ketamine-xylazine anesthesia (87 and 13 mg/kg, respectively). For sleep-wake activity recordings three cortical EEG electrodes, placed over the frontal and parietal cortices, and two nuchal electromyographic (EMG) electrodes were implanted. Leads from the EEG and EMG electrodes were anchored to the skull with dental cement. Telemetry transmitters were implanted ip for body temperature and motor activity recordings (Starr Life Sciences Corp., model G2 Emitter). The weight of a G2 Emitter is 1.1 g. The animals were allowed to recover from surgery for at least 10 days before any experimental manipulation started and handled daily to adapt them to the experimental procedures. The body weight of the mice at the time of the experiment was 28.9 ± 0.5 g.
Sleep-wake activity recordings and analyses. The animals were connected to the recording system through a lightweight, flexible tether plugged into a commutator, which was further routed to Grass Model 15 Neurodata amplifier system (Grass Instrument Division of Astro-Med, Inc., West Warwick, RI). The amplified EEG and EMG signals were digitized at 256 Hz and recorded by computer. The high-pass and low-pass filters for EEG signals were 0.5 and 30.0 Hz, respectively. The EMG signals were filtered with low and high cut-off frequencies at 100 and 10,000 Hz, respectively. The outputs from the 12A5 amplifiers were fed into an analog-to-digital converter and collected by computer using SleepWave software (Biosoft Studio, Hersey, PA). Sleep-wake states were scored visually off-line in 10-s segments. The vigilance states were defined as NREMS, REMS and wakefulness according to standard criteria as described previously 92 . Artifact-free EEG epochs were subjected to off-line spectral analysis by fast Fourier transformation. EEG power data in the range of 0.5 to 4.0 Hz during NREMS were used to compute EEG SWA. EEG SWA data were normalized for each animal by using the average EEG SWA across 24 h on the baseline day as 100. telemetry recordings. Core body temperature and locomotor activity were recorded by MiniMitter telemetry system (Starr Life Sciences Corp., model G2 Emitter and ER-4000 receiver) using VitalView software. Temperature and activity values were collected every 1 and 10 min, respectively, throughout the experiment and were averaged over 1-h time blocks. www.nature.com/scientificreports www.nature.com/scientificreports/ experimental procedures. Experiment 1: The effects of systemic administration of nicotinic acid on sleep-wake activity, body temperature and metabolic parameters in mice. Three groups of male mice were habituated to the injection procedure by administering 0.3 ml isotonic saline daily for 7 days 5-10 min before dark onset. On the baseline days, the animals were injected ip with 10 ml/kg isotonic NaCl. On the test day, the animals received 50, 100 or 250 mg/kg nicotinic acid (Millipore Sigma) in a volume of 10 ml/kg (n = 6, n = 8, n = 7, respectively). The treatments were performed 5-10 min before dark onset. Sleep and telemetry recordings started at the onset of the dark phase and continued for 24 h.
Another group of mice (n = 5) was habituated to the gavage procedure by administering 0.3 ml water daily for 7 days 5-15 min before dark onset. After the habituation period, a baseline day was recorded after the oral gavage of 0.3 ml vehicle. The following day, 1 g/kg nicotinic acid was administered in the volume of 0.3 ml. The treatments were performed 5-10 min before dark onset. Sleep and telemetry recordings started at the onset of the dark phase and continued for 24 h.

Experiment 2:
The effects of indomethacin pretreatment on the sleep-wake activity and body temperature-modulating effects of nicotinic acid in mice. A group of mice (n = 7) was habituated to the injection procedure as described above. On the baseline day, the animals were injected ip with 10 ml/kg isotonic NaCl. On the test day, the animals were pretreated with 5 mg/kg indomethacin subcutaneously 30 min prior the beginning of the dark phase, followed by 100 mg/kg nicotinic acid ip injection 5 minutes before dark onset. Sleep and telemetry recordings started at onset of the dark phase and continued for 24 h. Experiment 3: The effects of β3-adrenergic receptor agonist, CL-316,243, pretreatment on the sleep-wake activity and body temperature-modulating effects of nicotinic acid in mice. A group of mice (n = 7) was habituated to the injection procedure as described above. On the baseline days, the animals were injected with 10 ml/kg isotonic NaCl ip. On the test day, the animals were pretreated with subcutaneous 0.2 mg/kg CL-316,243 (Millipore Sigma) 30 min prior the beginning of the dark phase, followed by ip 250 mg/kg nicotinic acid injection 5 minutes before dark onset. Sleep and telemetry recordings started at onset of the dark phase and continued for 24 h.

Experiment 4: The effects of ip administration of nicotinamide in mice.
Two doses of nicotinamide were tested in the same group of mice (n = 6). After the habituation period, the animals received 10 ml/kg isotonic NaCl ip to obtain baseline values. On the test day, the mice were injected with 100 mg/kg nicotinamide ip. Two days later, a new baseline day was recorded followed by the test day of 250 mg/kg nicotinamide. The treatments were performed 5-10 min before dark onset. Sleep and telemetry recordings started at onset of the dark phase and continued for 24 h.

Experiment 5:
The effects of ip administration of monomethyl fumarate in mice. Ten male mice were habituated to the injection procedure by daily administration of isotonic saline 5-10 min before dark onset. On the baseline days, the animals were injected ip with 10 ml/kg isotonic NaCl. On the test day, the animals received 20 or 50 mg/ kg monomethyl fumarate in a volume of 10 ml/kg (n = 8, for both; five mice were injected with both doses of MMF). The treatments were at least two days apart and performed 5-10 min before dark onset. Sleep and telemetry recordings started at onset of the dark phase and continued for 24 h.

Experiment 6:
The effects of nicotinic acid on energy expenditure and food intake in mice. Oxygen consumption (VO 2 , ml/kg/h) and RER were measured via indirect calorimetry; food intake was measured simultaneously by an automated system (Oxymax System, Columbus Instruments, Columbus, OH). Animals were habituated to the calorimetry cages for at least three days before the recordings. On the baseline day, vehicle was administered, on the test days 50 mg/kg (n = 4) or 250 mg/kg (n = 4) nicotinic acid was injected ip, or 1 g/kg nicotinic acid was administered via oral gavage (n = 4). All treatments were performed 5-10 min before the onset of the dark period.
Statistics. Time spent in wakefulness, NREMS and REMS, as well as body temperature and motor activity, VO 2 and RER were calculated in 1-h blocks, and EEG SWA was calculated in 2-h blocks. Two-way repeated measures ANOVA was performed across 24 h between test days and the corresponding baselines (factors: treatment and time, both repeated). When appropriate, Tukey's HSD test was applied post hoc. An α-level of P < 0.05 was considered to be significant.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.