Male C57BL/6 mice (4-6 week-old; n=51) were purchased from the Experimental Animal Center of Lanzhou Veterinary Research Institute (Lanzhou, China). The animal experiments were approved by the Laboratory Animal Ethics Committee of Northwest Minzu University (Lanzhou, China), which is under the supervision of the Experimental Animal Committee of Lanzhou Veterinary Research Institute. Mice were housed individually in isolated cages (one mouse per cage) at a standardized temperature (18-23˚C) with ~50% humidity and a 12-h light/dark cycle. The mice were fed a standard diet (62% carbohydrate, 26% protein and 12% fat) and had free access to purified water. In addition, the health status of the mice was monitored during daily feeding. Excessive weight loss (20% of the average body weight of mice in the same period), loss of appetite, weakness (inability to eat or drink on their own, inability to stand or barely able stand for 24 h) and seizures (30) were considered humane endpoints, which did not happen in subsequent experiments.

In general, the mice were subjected to overnight fasting prior to sacrifice. However, the present study aimed to specifically investigate the effects of intermittent fasting. Therefore, prior to sacrifice, the intermittent fasting group was subjected to 12 h of fasting, while the control group was maintained with ad libitum feeding. For the 30-day experiment, the mice were randomly divided into two groups (n=10/group) as follows: The ad libitum group (AL) and intermittent fasting group (IF; Fig. 1A). In the 60-day experiment, the mice were randomly divided into three groups (n=8-14/group), including the AL group, IF group and the intermittent fasting followed by ad libitum feeding group (IF&AL; Fig. 1B). In the AL group, the mice had free access to food and water, while in the IF group, food and water were removed for 12 h during the nighttime (from 8:00 p.m. to 8:00 a.m.) on a daily basis (Fig. 1C) (31,32). The food intake and bodyweight of the mice were measured every 10 days until the end of the study. All animal food was purchased from the Double Lion Experimental Animal Feed Technology Co., Ltd. The mice were anesthetized using 4% isoflurane United States Pharmacopoeia (USP) 100% (1,000 ml/min; 4% isoflurane for induction; 500 ml/min; 1.75-2.5% isoflurane for maintenance) and then sacrificed by cervical dislocation. The livers were immediately isolated and weighed. One portion of the liver was fixed in 4% paraformaldehyde and the remaining liver was rapidly frozen in liquid nitrogen and then transferred to a -80˚C freezer.

The livers were processed in a series of stages, including alcohol dehydration, clearing by xylene and then embedding in paraffin. Deparaffinized sections of the liver (4 µm) were rehydrated and stained with H&E according to standard procedures.

The 30-day AL and 30-day IF groups were analyzed by targeted metabolomics. Prior to sacrifice, the animals in the IF group were subjected to 12-h fasting, while the control group was able to feed ad libitum. The livers were immediately isolated and rapidly stored in liquid nitrogen after sacrifice and were finally transferred to a -80˚C refrigerator. The frozen liver tissues were added in cold methanol/acetonitrile/H₂O and vortexed. The sample was homogenized by MP Fastprep-24 Automated Homogenizer (MP Biomedicals) and sonicated by an Ultrasonic Liquid Processors (Scientz JY92-II, Ningbo Scientz Biotechnology Co., Ltd.) in an ice-water bath (30 min each time, 2 times in total), and the mixture was centrifuged for 20 min (14,000 x g, 4˚C). Subsequently, the supernatant was dried in a vacuum centrifuge, re-dissolved in acetonitrile/water (1:1, v/v) and adequately vortexed, and centrifuged for 20 min (14,000 x, 4˚C). The final supernatants were collected for LC-MS/MS analysis.

The LC-MS/MS analyses were performed using an UHPLC (1290 Infinity LC; Agilent Technologies, Inc.) coupled with a QTRAP 5500 apparatus (AB Sciex LLC). For Hydrop Interaction Liquid Chromatography separation, the samples were analyzed using an ACQUITY UPLC BEH Amide column (Waters MS Technologies). The Quality Control sample was set for each interval of a certain number of experimental samples in the sample queue to detect and evaluate the stability and repeatability of the system. In electron spray ionization negative mode, the conditions were set as follows: Source temperature, 450˚C; Ion Source Gas1, 45; Ion Source Gas2, 45; Curtain gas, 30; IonSapary Voltage Floating, 4,500 V; and MS/MS Analysis mode of detection, ion pair.

Peripheral blood samples from the 30-day AL group and 30-day IF group were collected to measure specific biochemical markers. At the time of sample collection, the IF mice had fasted for 12 h, while the control group had ad libitum access to food. Initially, the mice were anesthetized with isoflurane USP 100% (1,000 ml/min; 4% isoflurane for induction; 500 ml/min; 1.75-2.5% isoflurane for maintenance) and subsequently, peripheral blood was immediately collected from the orbital venous sinus. At least 300 µl of blood (maximum 400 µl) was collected from each mouse. The mice were sacrificed immediately by cervical dislocation after blood collection. Mortality was further confirmed by determining the cessation of respiratory and heart function. The serum samples of the mice were harvested by centrifugation (440 x g for 10 min at room temperature) and stored at -20˚C until further analysis. Lilai Biological Co. was commissioned to measure biochemical markers in the blood samples. The concentration levels of glucose, total protein (TP), albumin (ALB) and globulin (GLO), as well as the activity levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), lactate dehydrogenase (LDH) and γ-glutamyl transferase (γ-GT) were measured using an auto-analyzer (Beckman LX-20; Beckman Coulter, Inc.).

Values were expressed as the mean ± standard error of the mean (SEM). Comparisons were performed with the Mann-Whitney U-test between two groups or with the Kruskal-Wallis test with Dunn's post-hoc multiple comparisons test. P<0.05 was considered to indicate a statistically significant difference.

Hierarchical clustering analysis was performed using cluster 3.0 (http://bonsai.hgc.jp/~mdehoon/software/cluster/software.htm) and the Java Treeview 3.0 software (http://jtreeview.sourceforge.net). The principal component analysis was performed using MetaboAnalyst 4.0 software. The Euclidean distance algorithm for similarity measurement and the average linkage clustering algorithm for clustering were selected. The heatmap was used as a visual aid apart from the dendrogram.

During the 30-day experiment, a significant increase in the bodyweight of the mice was noted. The bodyweight exhibited no significant differences between the AL and the IF at the end of the experiment (AL vs. IF, 21.701±1.305 vs. 21.610±1.187, respectively; mean ± SEM; n=10). However, the cumulative food intake of AL mice was higher compared with that of the IF mice. To confirm this result, the duration of fasting was extended to 60 days. A total of three groups were included in the 60-day experiment as follows: The AL, IF and IF&AL (refeeding) groups. The IF&AL group was subjected to 30-day intermittent fasting followed by 30 days of ad libitum access to food. The bodyweight of the mice in these three groups exhibited no significant difference by the end of the experiment (AL vs. IF vs. IF&AL, 26.193±1.680 vs. 26.844±1.136 vs. 26.457±1.779, respectively; mean ± SEM; n=8-14). The bodyweight and food intakes of three different time points were provided in Table I. Overall, the data indicated that 12-h nighttime intermittent fasting did not affect the bodyweight of mice, although it resulted in less cumulative food intake. The bodyweights of the mice subjected to 60-day fasting were significantly increased compared with those at 30 days (30-day AL vs. 60-day AL, 21.701±1.305 vs. 26.193±1.680, respectively; and 30-day IF vs. 60-day IF, 21.610±1.187 vs. 26.844±1.136, respectively; mean ± SEM), while their weight gain was significantly decreased. This may be explained by their food intake, since almost all of the mice in the 60-day experiments consumed less food than the 30-day group. A comparison of the bodyweight and food intakes of three different time points between the 60- and 30-day IF groups was provided in Table II. The maximum percentage of weight loss (before and after fasting) was 7.07%.

The livers were isolated, observed and weighed following animal sacrifice. The livers of IF mice appeared smaller than those of the mice in the AL group (Fig. 1D and F). Subsequently, the wet liver weight was determined and the liver weight/total bodyweight (LW/TBW) ratio was determined. The results indicated that the liver mass of IF mice was significantly lower than that of AL mice (Fig. 1E and G) and the LW/TBW ratio of the mice in the IF group was considerably lower than that of the AL group (Fig. 1H and I). Ad libitum refeeding for 30 days did not recover fasting-induced loss of liver mass. The liver weight of the mice subjected to 60-day intermittent fasting was significantly increased compared with that of the 30-day IF mice, while the LW/TBW ratio was not significantly altered (Fig. 1J and K). This suggested that prolonged intermittent fasting did not alter the effects noted previously.

Subsequently, the morphological changes in the mouse liver tissues were examined. Histological assessment of the liver tissues by H&E staining indicated that the hepatocyte plates were well developed and that the sinusoids were clearly visible in all of the livers (Fig. 2A and B).

Subsequently, the blood biochemical markers associated with liver physiology or function were assessed. The concentration levels of glucose, TP, ALB and GLO and the activity levels of ALT, AST, ALP, LDH and γ-GT were measured. The blood glucose levels were significantly decreased in IF mice compared with those in AL mice (IF vs. AL, 1.391±0.914 vs. 5.726±0.648, respectively; mean ± SEM; n=7; P<0.01; Fig. 3A). The activity levels of ALP were significantly increased in IF mice (IF vs. AL, 171.571±13.465 vs. 152.829±17.130, respectively; mean ± SEM; n=7; P<0.05; Fig. 3B). No significant change was observed for the other parameters tested (Fig. 3C-K). These results suggested that intermittent fasting did not affect the liver function of mice, whereas it was likely to regulate metabolism.

To finally assess the effects of intermittent fasting on liver metabolism, targeted metabolomics was performed by LC-MS/MS analysis. A total of 27 metabolite molecules related to energy metabolism (right column of Fig. 4B) were selected as targets. The principal component analysis provided clear clustering based on the diet patterns of the groups of mice (Fig. 4A). The distribution of the most significantly differentially expressed metabolites was visualized in a heatmap (Fig. 4B); this was used to separate the AL and the IF mice. A >1.5-fold increase was noted for 12 metabolite molecules in the IF mice compared to those in the AL mice (Fig. 5A-L) and the differences in Fig. 5F-I were particularly significant, While the expression of the other molecules did not exhibit any significant differences. The molecules exhibiting increased expression included nicotinamide adenine dinucleotide phosphate (NADP), adenosine triphosphate and succinate, whereas the main molecule exhibiting reduced expression was NADP. The majority of these metabolites are essentially involved in the citric acid cycle and oxidative phosphorylation. It was thus indicated that intermittent fasting enhanced liver metabolism, which may explain the reduction in the liver mass of these mice.