The central nervous system plays a critical role in regulating energy balance and body fat stores, distinct from its effects on feeding behaviour. In recent years, strong evidence across metazoan systems has emerged to show that different brain regions potently control lipid metabolism in peripheral tissues, independently of changes in feeding behaviour. Thus, discrete neural circuits and neuronal subpopulations regulate body fat metabolism. Although for certain central regulators the neural circuits governing peripheral metabolism are now beginning to be understood, identifying neuroendocrine factors that selectively govern fat metabolism has remained a long-standing challenge. Additionally, the extent to which nutrient sensing by the brain modulates peripheral metabolism via such neuroendocrine factors is poorly understood.
The ancient central nervous system neuromodulator serotonin (5-hydroxytryptamine, 5-HT) is a valuable paradigm for the study of global regulators of animal physiology. In metazoans, 5-HT controls food intake and feeding behaviour, mood, adiposity, locomotion and energy expenditure. In humans, with respect to the control of energy balance, global serotonergic agonists and antagonists have long been administered to suppress appetite, increase energy expenditure, or both. However the limited efficacy of these global serotonin-boosting compounds is often accompanied by substantial side effects, and has mitigated their use. Yet, the importance of 5-HT signalling in stimulating fat metabolism across many species suggests that the identification of selective factors that function downstream of central 5-HT to stimulate fat metabolism would be valuable. A body of evidence has revealed a broad framework for the effects of neuronal 5-HT on whole body energy balance by exploiting the tractability of the C. elegans system. Manipulations of 5-HT levels via exogenous administration or by endogenous control using genetic approaches reveal that, as in mammals, neuronal 5-HT is a potent stimulator of body fat loss and energy expenditure.
In the nematode C. elegans, 5-HT is synthesized by the conserved rate-limiting enzyme tryptophan hydroxylase (TPH-1) in only a few pairs of neurons, and is not present in the intestine or in other metabolic tissues under normal conditions. Loss of the TPH-1 enzyme leads to undetectable levels of neuronal 5-HT and a significant increase in body fat reserves accompanied by decreased energy expenditure. Functional studies reveal that of all the biogenic amine receptors in the genome, a 5-HT-gated chloride channel called MOD-1 is the sole receptor essential for body fat loss. In the intestine, the major metabolic organ for C. elegans and the predominant site for fat metabolism, the rate-limiting enzyme adipocyte triglyceride lipase ATGL-1 is transcriptionally induced by 5-HT signalling and serves to stimulate fat breakdown via hydrolysis of stored triglycerides to fatty acids. RNAi-based screens also revealed that increased 5-HT signalling elicits a cascade of β-oxidation enzymes in the intestine that convert fatty acids to energy in the mitochondria.
Despite its neuronal origins, the metabolic effects of 5-HT occur in the intestine. In our previous work, we deciphered the neural circuit for 5-HTergic fat mobilization in C. elegans. We found that 5-HT synthesis is required in the ADF chemosensory neurons whereas MOD-1 the critical serotonergic channel is necessary and sufficient in a single pair of neurons called URX, which receive direct synaptic input from the ADF neurons. Additionally, we found that octopamine (OA), the invertebrate analogue of adrenaline, provides a permissive cue to maintain 5-HT signalling via regulating tph-1 expression and 5-HT levels in the ADF neurons. Thus, the primary components of the 5-HT pathway: the biosynthetic enzyme and the receptor, reside in the nervous system, and the collective evidence indicates that 5-HTergic regulation of body fat loss occurs indirectly, perhaps via the relay of neuroendocrine factor/s from the nervous system, to the intestine. It has remained a challenge to identify selective neuroendocrine factors that stimulate body fat mobilization downstream of central mediators of energy balance, in any system. Pioneering biochemical approaches were used to identify neuropeptide hormones that communicate in endocrine fashion from the mammalian hypothalamus to control the physiology of stature, reproduction and other aspects of whole animal physiology. Despite these immense advances, biochemical approaches relying on the relative abundance of peptides in the mammalian hypothalamus did not lead to the identification of neuroendocrine hormones that control body weight, and endocrine factors that potently stimulate body fat loss have since remained unknown.
In this study, we identify a secreted neuropeptide ligand and its cognate receptor that constitute the core 5-HT neuroendocrine axis and selectively stimulates body fat loss in C. elegans. The ligand is secreted in proportion to 5-HT circuit functions and the activity of the nutrient sensor AMPK, in the secretory neurons. In the intestine, activation of the receptor promotes fat loss via induction of the ATGL-1 lipase. The broad conservation of this signalling axis suggests that such approaches are valuable in identifying novel and selective neuroendocrine factors that underlie the central control of body fat metabolism.
Our previous work describing the 5-HTergic neural circuit revealed that rather than 5-HT itself, an unknown neuroendocrine factor is released from the nervous system and relayed to metabolic tissues to stimulate body fat loss. The C. elegans intestine is not directly innervated and therefore offers a valuable platform to identify neuroendocrine factors that communicate between the nervous system and the metabolic tissues. The diversity of known mechanisms of neuroendocrine signalling across different species prompted us to use a process of elimination followed by a screen. To begin investigating the nature of this neuroendocrine signal, we first measured the extent to which serotonin-mediated fat loss was dependent upon the release of canonical neurotransmitters (acetylcholine, γ-amino butyric acid and glutamate), versus that of neuropeptidergic signals. In the nervous system, canonical neurotransmitters are localized to clear synaptic vesicles, which require a protein called UNC-13 (MUNC-13 in mammals) for fusion with the plasma membrane at the synapse. On the other hand, neuropeptides and biogenic amine neurotransmitters are localized to dense core vesicles, which require the conserved calcium-dependent activator protein (CAPS) or UNC-31/CAPS in C. elegans. Both unc-13 and unc-31 are broadly expressed in the C. elegans nervous system, and not in other tissues. Thus, loss of unc-13 function leads to a block in the release of the canonical neurotransmitters, whereas loss of unc-31 blocks the release of neuropeptides and biogenic amines. We measured the extent to which unc-13 and unc-31 mutants were essential in promoting 5-HT-mediated fat loss. With respect to body fat content, vehicle-treated unc-13 mutants resembled wild-type animals; however, unc-31 mutants had ∼50% greater body fat than either genotype, suggesting that the contents of dense core vesicles from the nervous system regulate fat stores under basal conditions. As reported previously, 5HT-treated wild-type animals retained approximately 40–50% of the body fat seen in vehicle-treated controls, as did the unc-13 mutants. On the other hand, 5-HT-treated unc-31 mutants fully suppressed serotonergic fat loss and retained as much body fat as the vehicle-treated controls. Thus, a UNC-31/CAPS-dependent secretory process is required for the effects of 5-HT on body fat loss.
