Skipping Breakfast and Late Meals Disrupt Body Clocks and Increase Liver Fat and Fat Tissue in Rats on a High-Cholesterol Diet

The animal study was approved by the Animal Care Committee of Nagoya University (Approval No. 2014110401) and conducted in compliance with the Rules and Regulations of the Guide for the Care and Use of Laboratory Animals, Nagoya University. All surgical procedures were performed under isoflurane anesthetization, and all efforts were made to minimize suffering. Fifty-five 5-week-old male Wistar rats were purchased from Japan SLC (Shizuoka, Japan). The rats were housed in individual wire-bottomed cages and maintained under a 12 h light cycle [zeitgeber time (ZT) 0-ZT 12]. The rats were fed a commercial stock diet (Lab MR Breeder, Nosan Co., Yokohama, Japan) for 3 days, followed by a basal diet for 10 days. The composition of the basal diet is presented in Table 2. Nine days after the rats were obtained, temperature data loggers (KN laboratories Inc., Osaka, Japan) were implanted into their intraperitoneal cavity. The rats were allowed to recover for 4 days after the surgery. Thirteen days after the rats were obtained, they were divided into two groups (n = 27 in the control group, n = 28 in the DMT group) by matching body weight. The rats were then fed a high-cholesterol diet for 16–17 days during the experimental period. As rats are nocturnal animals, they consume ~80% of their daily food intake during ZT 12-24 (33). The rats in the control group (n = 27) and DMT groups (n = 28) were provided access to a high-cholesterol diet from ZT 12 to ZT 24 and ZT 16 to ZT 4, respectively. In the current study, the rats were provided a sufficient portion of the diet at once before the feeding time and free access to the diet was provided during the feeding period (12 h). The control group was fed a diet on ZT 12, and the DMT group was fed a diet on ZT 16. In the DMT protocol, we fixed the feeding duration at 12 h in both groups, because it has been reported that duration of the feeding affects lipid metabolism (13). The 8-h feeding schedule dramatically improved the high-fat diet-induced abnormal lipid metabolism (13). We applied the DMT protocol as a one kind of breakfast skipping model, although the feeding time was delayed by 4 h in the DMT protocol. Moreover, individuals who skipped breakfast are related to late dinner at night. This habit was also associated with metabolic syndrome in Japanese population (34). Water was provided ad libitum throughout the experimental period. The composition of a high-cholesterol diet is presented in Table 2. We added 0.25% sodium cholate to the diet (Table 2), as sodium cholate drastically increased the cholesterol absorption to increase serum total cholesterol levels in rats (35). A high-cholesterol diet with cholate induces fatty liver and hypercholesterolemia. The food was removed at ZT 24 in the control group and at ZT 4 in the DMT group. After the food was removed, the body weight, and total daily food intake were calculated. To analyze the pattern of their food intake during the feeding period, we measured the amount of their food intake three times at 4 h intervals. In order to avoid influencing the food intake, the food was carefully removed from each cage, measured, and immediately returned. Three or four rats from each group were sacrificed at an interval of 4 h from ZT 2 on day 16 to next ZT 2 on day 17. The rats were sacrificed by decapitation without anesthetization to minimize the influence of stress on the serum corticosterone levels. The livers were harvested and frozen immediately using liquid nitrogen. The frozen samples were stored at −80°C.

The spontaneous activities of the control and DMT rats (n = 3) were monitored in separated individual transparent-plastic cages using a LOCOMO LS-8 system (Melquest, Toyama, Japan) for 1 day. The transparent-plastic cage was placed inside the sensor panel (28.0 × 44.0 cm), and the photobeam interruptions were counted by the interface when the rat moved in the cage. The activities were counted every 10 min for 24 h. The rats were allowed to acclimatize to the measuring environment for 1 h before the actual measurement.

The body temperature of the control (n = 6) and DMT rats (n = 7) was measured using a temperature data logger, as described above. Body temperature was recorded every 10 min continuously during the experimental period with a resolution of 0.1°C. The collected data were analyzed at an interval of 60 min using the Rh Manager program (KN laboratories Inc., Osaka, Japan).

Liver lipids were extracted by using the method described by Folch et al. (36). Total liver lipids were measured gravimetrically. Hepatic cholesterol, triglycerides, and phospholipids in the lipid extract were measured enzymatically (T-CHO and TG-EN; Kainos Laboratories, Tokyo, Japan, Phospholipids C-Test, Wako Pure Chemical Industries, Osaka, Japan). Serum glucose, total cholesterol, triglyceride, non-esterified fatty acids (NEFA), and bile acids were measured enzymatically by using commercial kits (Glucose CII-test, Triglyceride E-test, Cholesterol C-test, NEFA C-test, TBA-test; Wako Pure Chemical Industries, Osaka, Japan). The insulin and corticosterone levels were measured using rat-specific enzyme immunoassay kits [Rat insulin ELISA kit (#M1101); Morinaga Institute of Biological Science, Yokohama, Japan, Corticosterone ELISA kit (#EC3001-1); Assaypro, MO, USA].

Total RNA was extracted from the liver and epididymal adipose tissue using the method described by Chomczynski and Sacchi (37). After the RNA quality was assessed by using northern blotting, cDNA was synthesized using a High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). Real-time quantitative PCR was performed using the Power SYBR™ Green PCR Master Mix (Thermo Fisher Scientific) and analyzed using the StepOnePlus Real-Time PCR System (Thermo Fisher Scientific). The primer sequences are provided in Supplementary Table 1. The mRNA levels were normalized relative to the mRNA levels of apolipoprotein E (Apo E) in the liver and 18S RNA in the epididymal adipose tissue.

The oscillations in body temperature, serum lipids, hormones, and gene expression of hepatic clock and lipid metabolism were analyzed by using JTK_CYCLE implemented on R (38). The peak time and amplitude of the rhythms were analyzed using JTK_CYCLE for a period of 24 h. Rhythmicity was confirmed at p < 0.05.

The results were expressed as mean ± standard error of the mean (SEM). Statistical differences between the control and DMT groups were analyzed using the Student's t-test (Figures 1, 2D).

We established a DMT protocol for one kind of breakfast skipping model in which rats consumed a diet after a 4 h delay during the active phase (29). The control rats ate 6.2 g (35.2%), 4.4 g (25.0%), and 7.0 g (39.8%) at ZT 12-16, ZT 16-20, and ZT 20-24, respectively. The DMT rats ate 8.0 g (45.7%), 6.7 g (38.3%), and 2.8 g (16.0%) at ZT 16-20, ZT 20-24, and ZT 24-4, respectively (Figure 1A). During ZT 16-20, amount of the food intake of the DMT rats was higher than that of the control rats (p < 0.001), whereas it did not change during ZT 20-0 (Figure 1A). The total food intake during the experimental period in the control and DMT groups was almost equal (p = 0.11) (Figure 1A, inset). Body weight gain was similar in both groups (Figure 1B).

There was no significant difference in the liver weight between the two groups (Figure 1C), but the epididymal adipose tissue weight increased in the DMT group (Figure 1D). It is known that the consumption of a high-cholesterol diet with cholate induces fatty liver in rats (35). The hepatic total lipid, triglycerides, and cholesterol levels increased in response to DMT (Figures 1E–G). However, the hepatic phospholipid levels did not change (Figure 1H).

The circadian change in body temperature is controlled by the hypothalamic thermoregulatory center, and the rhythm of the locomotive activity is regulated by the hypothalamus (39, 40). We monitored the daily rhythm of the body temperature and locomotive activity in the DMT rats (Figure 2). As rats are nocturnal, their body temperature is elevated when the light was turned off (ZT 12) and decreased when the light was turned on (ZT 24). However, the body temperature of the DMT rats increased gradually at ZT 12 and sharply at ZT 16 when they started feeding (Figure 2A). Furthermore, the diurnal rhythm of the body temperature was delayed by 0.5 h in response to DMT (Figure 2A and Supplementary Table 2). The diurnal variation in locomotive activity was also similar between the two groups (Figures 2B,C and Supplementary Table 2). There was no difference in the total locomotive activity between the groups (Figure 2D). These data suggest that locomotive activity is primary controlled by light conditions. We speculated that as a stimulus to body temperature, the timing of feeding was equally important as light.

In the DMT group, the diurnal rhythm of serum glucose and bile acids did not change (Figures 3A,E and Supplementary Table 2), whereas that of serum NEFA was delayed by 2 h (Figure 3D and Supplementary Table 2). Moreover, the diurnal rhythm of serum total cholesterol and insulin was delayed by 4 h in the DMT rats (Figures 3B,F and Supplementary Table 2). In contrast, the diurnal rhythm of serum triglyceride was advanced by 2 h in the DMT rats (Figure 3C and Supplementary Table 2). The level of serum corticosterone, secreted by the adrenal glands, is controlled by the adrenocorticotropic hormone secreted from the pituitary gland (41). Reportedly, corticosterone level is dependent on the timing of eating (42). However, in this study, the diurnal rhythm of serum corticosterone appeared to be delayed in DMT rats, but the change was not significant (Figure 3G and Supplementary Table 2).

We demonstrated that DMT rats fed a high-fat diet clearly showed a delay in clock gene expression (29). Therefore, we expected that the clock gene expression would be delayed in response to DMT, even in rats fed a high-cholesterol diet. However, the expression of CLOCK, PER2, and DEC2 did not change (Figures 4B,D,H and Supplementary Table 3). In addition, the oscillations of REV-ERBβ, RORα, E4BP4, DBP, and HLF expression were also unaltered (Figures 4J–N and Supplementary Table 3). The expression of clock genes including CLOCK, PER2, DEC2, REV-ERBβ, RORα, E4BP4, DBP, and HLF, did not change. However, the other half of the clock genes including BMAL1, PER1, CRY1, CRY2, DEC1, REV-ERBα, and TEF, were delayed by 2 h in response to DMT (Figures 4A,C,E–G,I,O and Supplementary Table 3). These results showed that the circadian rhythms of hepatic clock gene expression were relatively consistent in DMT rats fed a high-cholesterol diet.

We measured the expression of genes related to lipid metabolism. The oscillation of lipogenic genes such as ACLY, ACACA, FAS, ELOVL6, and SREBP1 was delayed by 4 h in the DMT group (Figures 5A–E and Supplementary Table 4). The expression of SREBP2 gene, which related to cholesterol synthesis, and FXR gene, which related to bile acids synthesis, was delayed by 4 h in the DMT group (Figures 5J,M). The expression of fatty acid oxidation-related genes, such as PPARα, CPT-1α, and ACOX1, was delayed by 2 h in the DMT group (Figures 5F–H and Supplementary Table 4). Moreover, the oscillation of FGF21, controlled by PPARα, was also delayed by 2 h in the DMT group (Figure 5I). The oscillation of HMG-CoAS remained unaltered (Figure 5K and Supplementary Table 4), however, HMG-CoAR, which related to cholesterol synthesis, and SHP, CYP7A1, and CYP8B1, which are related to bile acids synthesis was delayed by 2 h in the DMT group (Figures 5L,N–P and Supplementary Table 4). These results showed that the circadian oscillation of hepatic lipid metabolism was delayed by 2–4 h in response to DMT.

In the present study, DMT with a high-cholesterol diet increased epididymal adipose tissue weight (Figure 1D). We evaluated the circadian oscillation of clock genes in the epididymal adipose tissue. The peaks of BMAL1, PER2, and CRY1 expression were not changed in response to DMT (Supplementary Figures 1a,d,e and Supplementary Table 5), whereas the peaks of PER1, E4BP4, and DBP expression were delayed by 2 h in response to DMT (Supplementary Figures 1c,g,h and Supplementary Table 5). Meanwhile, the peak of CLOCK expression was advanced by 2 h in response to DMT (Supplementary Figure 1b and Supplementary Table 5). The oscillation of HMG-CoAS remained unaltered (Supplementary Figure 1f and Supplementary Table 5). These results showed that the circadian rhythm of clock gene expression in epididymal adipose tissue was slightly altered by DMT.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.