You are when you eat: How food synchronizes liver clocks

Abstract

Food ingestion synchronizes liver circadian gene expression via hormone and nutrient regulation of the mTOR pathway.

Life on Earth is dictated by the 24-hour rotation cycle of the planet, and evolution has selected for endogenous timekeeping systems known as circadian clocks, which anticipate and adapt to predictable environmental fluctuations (1). In mammals, these molecular oscillators operate in each cell of almost every tissue, orchestrating daily, temporal peaks in metabolism, immunity, and behavior. For decades, the prevailing chronological paradigm dictated that the suprachiasmatic nucleus (SCN) in the hypothalamus, synchronized primarily by light cues, acted as the supreme conductor for all clocks in peripheral organs such as the liver (Fig. 1A). However, landmark studies from the groups of Schibler and Menaker (2, 3) demonstrated that when mice are subjected to reverse-phase feeding (eating only during the light period, which is their inactive phase), the liver circadian clock shifts its phase by 12 hours, becoming uncoupled from light and entrained to the availability of food, with the period remaining ∼24 hours. This fundamentally challenged the light-centric dogma, establishing that food intake functions as the dominant environmental synchronizer for the liver as well as other peripheral metabolic organs, but the mechanism by which food controls the liver clock has been a mystery.

Fig. 1. Impact of feeding misalignment on hepatic mTOR signaling and endoplasmic reticulum (ER) homeostasis.

(A) The central clock in the suprachiasmatic nucleus (SCN) synchronizes to the environmental light/dark cycle, generates circadian rhythms, and transmits systemic timing cues (neural, hormonal, and body temperature) to peripheral organs. (B) Different feeding patterns alter molecular signaling in the liver. Left: Feeding aligned with the light/dark cycle delivers nutrient cues (amino acids and insulin) that stimulate rhythmic mTOR signaling, maintaining a synchronized cross-talk with the local transcription-translation feedback loop. Right: Feeding misaligned with the light/dark cycle generates aberrant metabolic regulators, resulting in arrhythmic mTOR signaling and disrupted cross-talk with the molecular clock. (C) Coordinated signaling supports metabolic and ER homeostasis (left), whereas circadian misalignment leads to metabolic dysfunction and ER stress (right). Credit: Ashley Mastin/Science Advances.

A new study from Sahasrabudhe and colleagues in this issue of Science Advances (4) identifies signaling through the mechanistic target of rapamycin (mTOR) pathway as the critical coordinator of these food-driven rhythms in the liver. Notably, mTOR inhibition disrupted only a subset of rhythmic hepatic genes rather than globally eliminating rhythmicity. This partial dependence is reminiscent of studies showing that hepatocyte REV-ERBα/β loss selectively alters subsets of rhythmic liver genes, as well as studies of BMAL1 in which some rhythmic transcriptional and metabolic programs are lost while others persist (5, 6). Together, these findings support a modular view of hepatic circadian regulation, in which feeding, nutrient-sensitive signaling, and the canonical clock machinery control overlapping but incomplete fractions of the rhythmic liver transcriptome.

The liver not only passively anticipates ingestive behavior but, critically, is the primary metabolic clearinghouse responsible for processing, storing, and redistributing incoming nutrients. Thus, when light/dark and fasting/feeding cycles are in conflict, the liver’s temporal alignment must prioritize metabolic demands over solar cycles. This has implications for modern life, which often separates eating from the natural day-night cycle. Shift work, jet lag, late-night snacking, and irregular meals can all expose the liver to food at unexpected times. When these signals disagree, there is dissonance between the synchronizers for the central and peripheral clocks, which can contribute to metabolic diseases, including fatty liver, insulin resistance, and other disorders linked to circadian disruption. The desynchrony mechanism postulated to explain these metabolic perturbations has been buttressed by evidence that matching the length of the light/dark cycle to that of a broken central clock (7) or matching the periods of broken central and peripheral clocks (8) can mitigate the metabolic abnormalities.

But, what is the mechanism by which nutrient availability synchronizes liver clocks? The organismal response to nutrients is largely driven by the secretion of the hormone insulin from pancreatic islets. The insulin signaling pathway involves a series of protein kinase reactions, leading to activation of mTOR, which operates as the cell’s central nutrient-sensing hub, balancing energetic intake with macromolecular synthesis and growth. mTOR activity is also directly regulated by amino acids. In the mouse liver, mTOR activity rises and falls across the day and is closely tied to feeding behavior. The study by Sahasrabudhe et al. (4) shows that mTOR is not just a passive readout of feeding. Rhythmic mTOR activity is both necessary and sufficient to drive many daily rhythms in liver gene expression.

To establish necessity, the authors pharmacologically silenced mTOR signaling by administering its specific inhibitor, rapamycin, to mice subjected to a nighttime feeding regimen (eating only during their active phase). This wiped out more than half of the rhythmic gene expression in the liver without much effect on the core circadian clock. To test whether mTOR was sufficient to create these rhythms, the authors turned to an arrhythmic feeding model. In these continuously grazing mice, systemic nutrient delivery is flattened, resulting in a loss of transcriptomic rhythms and a blunted, non-oscillating mTOR profile. Injecting these mice with a potent, short-acting mTOR inhibitor at a specific time of day artificially forced mTOR activity to drop and reset dynamically, mimicking a natural fasting-feeding cycle. This single, time-locked intervention was sufficient to rescue and reboot the daily oscillations of hundreds of system-driven genes and metabolites, restoring them to a phase and amplitude similar to those of healthy, night-fed controls.

Using these tools, the study classified rhythmic liver genes into regulatory categories. The rhythms of “Clock-driven” genes are primarily controlled by light via the SCN clock, whereas oscillations of “mTOR-driven” genes are largely dependent on feeding/mTOR. Other genes are influenced by both pathways. Functional genomics studies suggested that clock-driven genes rely on core clock components such as REV-ERBs and BMAL1, whereas mTOR-dependent transcripts are regulated by metabolic and stress-responsive transcriptional factors, including CREB1, E2F4, and GABPA, which are not normally considered to be circadian regulators. Intriguingly, overnutrition also induces new circadian rhythms of metabolic transcription factors SREBP1 and PPARα, which then cause oscillations of hundreds of genes independent of the canonical clock, while REV-ERBs and BMAL1 also continue to control the oscillations of canonical clock components (9, 10).

The authors also examined the physiological consequences of the control of hepatic circadian gene expression by the mTOR pathway. mTOR is a major regulator of the endoplasmic reticulum (ER), the cellular factory responsible for folding and processing proteins. When mTOR rhythms are flattened by continuous feeding (Fig. 1B), the expression of essential ER chaperones is attenuated (Fig. 1C), leaving the organ highly vulnerable to proteotoxic stress, which is normally curtailed by the unfolded protein response (UPR). Restoring the mTOR rhythm with time-targeted injections of the short-acting mTOR inhibitor rescued the temporal expression of the UPR machinery, thereby restoring the liver’s homeostatic capacity to manage cellular stress.

The discovery that mTOR acts as an autonomous conductor of nutrient-driven rhythms adds to our understanding of metabolic chronobiology and also raises numerous questions for future investigation. One intriguing observation is that neither disruption of insulin signaling to mTOR by deletion of Tsc1 nor disruption of amino acid sensing (by deleting Raptor) was sufficient to fully abolish the transcriptional rhythms driven by food intake. This resilience suggests a profound level of evolutionary redundancy. Given that the liver has developed multiple, overlapping fail-safes to ensure that mTOR continues to oscillate in response to food, it will be of interest to see whether there are additional inputs to mTOR regulation of circadian gene expression, including perhaps energy sensing via the AMPK pathway or lipid-sensing nuclear receptors such as PPARα, able to sustain mTOR oscillations when insulin secretion or amino acid oscillations are compromised. Determining the extent of this interconnected network would further enhance our understanding of the plasticity of metabolic regulation of circadian gene expression. Similarly, the molecular explanation of how mTOR enlists noncanonical transcription factors to regulate circadian rhythms remains to be elucidated.

Furthermore, the authors noted that the short-acting mTOR inhibitor induced a distinct 5-hour phase delay in the rescued mTOR-driven transcripts, effectively uncoupling their peak expression from the endogenous, clock-driven transcripts. This plasticity underscores that while clock-driven genes are anchored to the rigid, light-entrained SCN pacemaker, system-driven genes are highly malleable and governed strictly by the acute timing of metabolic transitions. Indeed, this observation prompts the question of long-term pathological consequences when clock-driven pathways (e.g., maximizing glycogenolysis at dawn) are permanently out of phase with mTOR-driven pathways (e.g., maximizing protein translation and lipogenesis). This “intramolecular desynchronization” within a single hepatocyte presumably would create a profound chronobiological mismatch, serving as the driver of the cellular dysfunction that could contribute to the mitochondrial dysfunction and steatohepatitis frequently observed in shift workers. The demonstration that a transient, short-half-life chemical inhibitor of mTOR can artificially mimic the transcriptomic and metabolic benefits of a fasting-feeding cycle suggests novel therapeutic possibilities.

More broadly, the present study contributes to our understanding of the bidirectional relationship between circadian clocks and metabolism. The literature contains evidence that much of metabolic physiology displays circadian rhythmicity driven by the circadian clock, while the present study adds to previous work showing that metabolism directly dictates rhythm gene expression. While the authors demonstrate a critical role for mTOR signaling, numerous metabolic signals have also been implicated in regulating circadian clocks. We know that NAD (nicotinamide adenine dinucleotide) levels can alter the function of the clock (11) and that circadian metabolites may drive histone modifications that further affect the transcriptional regulation of core clock genes as well as clock-controlled genes (12). How these all fit together will be a complicated but very interesting subject for future investigations, the results of which will have important implications for understanding and treating metabolic diseases.