Feeding the Clock

See allHide authors and affiliations

Science  16 Oct 2009:
Vol. 326, Issue 5951, pp. 378-379
DOI: 10.1126/science.1181278

In mammals, sleeping, feeding, and most other physiological processes are influenced by a circadian system and therefore display daily oscillations. These rhythms are generated by self-sustained and cell-autonomous molecular clocks that exist in virtually all cell types. A “master clock” located in the brain's suprachiasmatic nucleus (SCN) can synchronize these peripheral clocks, but for many organs, feeding-fasting rhythms are the dominant zeitgebers (timing cues) (1, 2). On page 437 of this issue, Lamia et al. propose a molecular mechanism through which metabolic cycles may interact with the circadian clockwork circuitry. They show that an enzyme that responds to nutrient availability—adenosine monophosphate–activated protein kinase (AMPK)—directly phosphorylates the core clock protein cryptochrome 1 (CRY1), thereby marking it for degradation (3).

The core of the mammalian circadian clock is thought to constitute a network of interconnected transcriptional and translational feedback loops that control the oscillatory expression of genes encoding clock components and specific output genes. CRY proteins had previously been described as targets of Fbxl3 (4, 5), a ubiquitin ligase complex that marks proteins for degradation in the proteasome. However, regulation of CRY-Fbxl3 interactions has remained unknown. Lamia et al. identified highly conserved amino acid sequences in CRY1 as potential phosphorylation sites of AMPK. AMPK is activated upon its phosphorylation by protein kinases such as liver kinase B1 (LKB1) or calcium-calmodulin–dependent protein kinase kinase β, which sense the AMP/ATP ratio, a direct readout of the cell's metabolic state. By using a site-directed mutation analysis of CRY1, serines at position 71 and 280 were identified as critical AMPK phosphorylation sites. When these serines were phosphorylated, the affinity between CRY1 and Fbxl3 increased. In vitro phosphorylation assays and in vivo loss-of-function experiments in cultured mouse embryonic fibroblasts revealed that AMPK is both necessary and sufficient to induce the phosphorylation-dependent degradation of CRY1.

Coupled cycles.

The AMP/ATP ratio transmits information about the metabolic state of the cell to LKB1, which triggers a cascade of phosphorylation (P) events that eventually targets the clock protein CRY1 for degradation.

To scrutinize the putative role of AMPK as a food sensor, the authors monitored the expression of circadian clock output genes in mouse embryonic fibroblasts that expressed or lacked AMPK and were cultured in different glucose concentrations. The amplitude of circadian oscillations in gene expression correlated to glucose concentrations only in wild-type cells, but not in the absence of AMPK. In mouse liver, the accumulation and nuclear localization of AMPK, as well as the phosphorylation of known AMPK target proteins, oscillated in a circadian manner. Thus, perturbation of nutrient availability—and consequently, of AMPK activity—alters output of the circadian clock.

Although AMPK is an attractive candidate for coupling metabolic and circadian cycles, additional regulators are likely involved. Thus, the ratio of oxidized nicotinamide adenine dinucleotide phosphate (NADP+) to its reduced form (NADPH)—which, like the AMP/ATP ratio, constitutes a diagnostic signature of a cell's metabolic state—has been proposed to affect circadian gene expression through diverse mechanisms. At least in vitro, the binding of the heterodimeric core clock transcription factors CLOCK-BMAL1 and NPAS2-BMAL1 to their cognate DNA sequences (so-called E-boxes) is enhanced by NADPH and impaired by NADP+ (6). The transcriptional regulatory protein peroxisome proliferator–activated receptor γ (PPARγ) coactivator 1α (PGC-1α), a well-known mediator of glucose and lipid metabolism, has been proposed to be another important player in connecting metabolism to circadian gene expression. This transcriptional coactivator associates with nuclear receptors of the ROR family and thereby modulates the transcription of the clock genes Bmal1 and Reverbα. Finally, the NAD+-dependent protein deacetylase sirtuin 1 influences the stability and activity of the core clock components PER2 and BMAL1, respectively (7, 8).

Why are metabolic processes under tight circadian control? A simple explanation arises from the necessity to separate incompatible enzymatic processes within the same cell. Because complete spatial separation of anabolic and catabolic processes is frequently impossible, these have to be gated to different time windows. This necessity is well illustrated by the temporal sequestration of oxidative and reductive phases in yeast by an ultradian respiratory clock. For example, DNA is replicated exclusively in the reductive phase, when the concentration of genotoxic reactive oxygen species generated by mitochondrial respiration is minimal (9). In a yeast mutant in which the reductive phase is too short to allow for the completion of DNA synthesis, the mutation rate increases dramatically (10). In mammals, the master pacemaker in the SCN is phase-entrained primarily by light-dark cycles and thus cannot readily adapt to altered feeding rhythms. Hence, when food availability changes, nutrient-dependent synchronization cues must dominate the more direct signals from the SCN to maintain proper homeostasis of metabolism in peripheral tissues (1). This could explain the multitude of metabolic phase entrainment cues that synchronize the circadian core clock machinery in metabolically active peripheral organs. A major challenge will be to understand how the multiple nutrient-dependent inputs are integrated so as to maintain coherence between the metabolic state of the organism and the circadian system.


View Abstract

Stay Connected to Science

Navigate This Article