Neurobiology of food anticipatory circadian rhythms

Abstract

Mistlberger, R.E. Neurobiology of food anticipatory circadian rhythms. Physiol Behav 00(00):000–000, 2011. Circadian rhythms in mammals can be entrained by daily schedules of light or food availability. A master light-entrainable circadian pacemaker located in the suprachiasmatic nucleus (SCN) is comprised of a population of cell autonomous, transcriptionally based circadian oscillators with defined retinal inputs, circadian clock genes and neural outputs. By contrast, the neurobiology of food-entrainable circadian rhythmicity remains poorly understood at the systems and cellular levels. Induction of food-anticipatory activity rhythms by daily feeding schedules does not require the SCN, but these rhythms do exhibit defining properties of circadian clock control. Clock gene rhythms expressed in other brain regions and in peripheral organs are preferentially reset by mealtime, but lesions of specific hypothalamic, corticolimbic and brainstem structures do not eliminate all food anticipatory rhythms, suggesting control by a distributed, decentralized system of oscillators, or the existence of a critical oscillator at an unknown location. The melanocortin system and dorsomedial hypothalamus may play modulatory roles setting the level of anticipatory activity. The metabolic hormones ghrelin and leptin are not required to induce behavioral food anticipatory rhythms, but may also participate in gain setting. Clock gene mutations that disrupt light-entrainable rhythms generally do not eliminate food anticipatory rhythms, suggesting a novel timing mechanism. Recent evidence for non-transcriptional and network based circadian rhythmicity provides precedence, but any such mechanisms are likely to interact closely with known circadian clock genes, and some important double and triple clock gene knockouts remain to be phenotyped for food entrainment. Given the dominant role of food as an entraining stimulus for metabolic rhythms, the timing of daily food intake and the fidelity of food entrainment mechanisms are likely to have clinical relevance.

Introduction

In many animals, including humans and common laboratory rodents, a significant portion of the variance in meal frequency and meal size can be accounted for by time of day, or more precisely, time of biological day, as defined by the circadian clocks that generate daily rhythms of behavior and physiology [1]. Daily feeding rhythms have the interesting property of being self-reinforcing, because food intake is coupled to the circadian clock system as both an output and an input. That is, food intake provides cues and elicits physiological responses that can act as entrainment stimuli (so-called ‘zeitgebers’) controlling the phase and period of circadian clocks in the brain and in peripheral organs. This review will examine progress in understanding the mechanisms by which the timing of food intake coordinates daily rhythms in mammals.

A useful if simplified heuristic by which to introduce the subject matter is to distinguish between light-entrainable and food-entrainable components of the mammalian circadian timekeeping system. At the core of the circadian system is the suprachiasmatic nucleus (SCN), which as the name implies sits atop the optic chiasm at the base of the hypothalamus, nestled among hypothalamic structures famous for their roles in homeostatic regulation and associated appetitive states [2], [3]. The rodent SCN consists of some 10⁴ neurons, many if not all of which have the intrinsic capacity to oscillate with a 24 h periodicity [4], [5]. At the molecular level, circadian oscillations are understood to derive from multiple, interlocking, autoregulatory transcription/translation feedback loops, with positive elements (CLOCK and BMAL1 heterodimers) driving expression of negative elements (two Cryptochrome and three Period genes) that form dimers and feed back to inhibit their own activation. Accessory loops (e.g., involving RORA and Rev-Erbα) and posttranslational modifications (e.g., phosphorylation) help set the period, amplitude and precision of circadian cycling. Posttranslational processes alone may also generate circadian oscillations, without transcription, in at least some mammalian cells [6], a recent discovery to be discussed further below. A subset of SCN neurons receive input from intrinsically photoreceptive retinal ganglion cells, which via intra- and inter-cellular cascades alters clock gene expression in SCN neurons, thereby resetting the transcription–translation oscillator by an amount and direction that depends on the timing of light exposure relative to the phase of the clock gene cycles [7]. The net result is stable entrainment of the SCN oscillator ensemble to daily LD cycles. SCN output signals are both diffusible and synaptic and target several medial hypothalamic structures, including periventricular, paraventricular and dorsomedial hypothalamic nuclei that regulate feeding behavior and metabolism [2], [3], [8]. The details remain to be fully specified (including the basis for species-specific nocturnal, diurnal and crepuscular feeding chronotypes), but daily rhythms of ingestive behavior are thought to emerge from SCN modulation of these hypothalamic circuits. Although many other brain regions also receive retinal input or contain circadian oscillators [9], ablation of the SCN eliminates circadian rhythms of behavior and physiology in animals with unrestricted access to food [10]. The SCN can therefore be viewed as a master circadian pacemaker critical for entrainment of circadian rhythms to LD cycles.

The food-entrainable component of the circadian timekeeping system may well include almost every other circadian oscillator in the brain and in peripheral organs and tissues. Daily rhythms of form and function are ubiquitous in cells and organs, and some of these rhythms, such as adrenal responses to adrenocorticotrophic hormone [11] and migrating motor complexes in the duodenum [12], were shown decades ago to persist in vitro, suggesting the existence of circadian oscillators in non-neural tissues. Identification of mammalian circadian clock genes in the 1990s was soon followed by the discovery that these genes are expressed in tissues throughout the body and that, under appropriate conditions, clocks in tissue explants can oscillate with a circadian period almost indefinitely in vitro [13]. In vivo, most circadian oscillators outside of the SCN are preferentially entrained by feeding time. Thus, if a rat or mouse is limited to one daily meal in the middle of the light period, when nocturnal animals normally eat little, clock gene rhythms in stomach, intestines, pancreas, liver, adrenal gland, heart, lungs, muscle and other tissues are shifted, to realign with the daily rhythm of food intake [13], [14], [15], [16], [17]. Circadian clock gene rhythms in numerous brain areas outside of the SCN are also shifted [18], and a behavioral rhythm of food anticipatory activity emerges, characterized by a monotonic increase in locomotor activity (e.g., wheel running) and food appetitive behaviors (e.g., food bin approaches or lever pressing) beginning several hours prior to mealtime and rising to a peak at mealtime (Fig. 1). Daily rhythms of food anticipatory behavior persist for several cycles if food is withheld (Figs. 1b, 2a), do not emerge if feeding intervals are outside of the circadian range (~ 22–31 h), and respond to shifts of mealtime by gradual rather than immediate resetting, thereby qualifying as true circadian rhythms entrainable by food [19], [20], [21], [22]. During inverted feeding schedules, circadian oscillations within the SCN and a few other tissues (e.g., the pineal gland) remain coupled to the LD cycle [14], [15], [16]. SCN ablation eliminates circadian rhythmicity in free-feeding rats and mice, but food-anticipatory rhythms reappear if food is restricted to one daily mealtime [23], [24], [25] (Fig. 2b).

These observations indicate that daily rhythms of foraging and physiology reflect the coordinated activity of a distributed population of circadian oscillators that are normally coupled to signals provided by daily cycles of food intake, whether these cycles are generated spontaneously by the SCN (during times of bounty) or are forced on the organism by the environment (when food availability is temporally restricted). When LD and food availability schedules are in conflict, only a few tissues remain coupled to LD, presumably to preserve other functions, such as measuring daylength, the basis for annual rhythmicity and dependent on stable photic entrainment of the SCN and pineal gland.

In this broad view of a multioscillator circadian system responsible for timing food seeking behavior and coordinating physiology with regular mealtimes, there is much detail that remains to be specified, at the cellular, systems and behavioral levels of analysis. In the following sections, recent progress toward filling in these details will be highlighted, with primary emphasis on neural and molecular substrates of food anticipatory behavioral rhythms.