Enteroendocrine cell types that drive food reward and aversion

Read the full article


Animals must learn through experience which foods are nutritious and should be consumed, and which are toxic and should be avoided. Enteroendocrine cells (EECs) are the principal chemosensors in the GI tract, but investigation of their role in behavior has been limited by the difficulty of selectively targeting these cells in vivo. Here we describe an intersectional genetic approach for manipulating EEC subtypes in behaving mice. We show that multiple EEC subtypes inhibit food intake but have different effects on learning. Conditioned flavor preference is driven by release of cholecystokinin whereas conditioned taste aversion is mediated by serotonin and substance P. These positive and negative valence signals are transmitted by vagal and spinal afferents, respectively. These findings establish a cellular basis for how chemosensing in the gut drives learning about food.

Article activity feed

  1. Evaluation Summary:

    The effective approach for targeting select subsets of enteroendocrine cells described here will be important for neuroscientists, endocrinologists, microbiologists, and other scientists studying nutritional biology. The study reveals new detail of how enteroendocrine cells signal through spinal and vagal sensory neurons to control immediate behavior and to guide learning about potential food sources. Overall, the approach is well thought out and the results are interesting and unexpected.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

  2. Reviewer #3 (Public Review):

    In this manuscript, a range of techniques are used to define transcriptomic diversity amongst EECs and to explore functional differences that result from activation of select EEC populations. A key aspect of the study is the development of an intersectional approach that allows EECs to be selectively manipulated using a chemogenetic strategy. The authors use this approach to infer that activation of various subsets of EECs can reduce food consumption in hungry mice but differentially modulate associative learning, with activation of Tac1-Cre targeted enterochromaffin cells providing a strong aversive signal but activation of CCK-Cre targeted EECs appetitive learning. Moreover, inhibitors, nerve transection and toxin treatment suggest distinct mechanisms and cellular substrates for these effects.

    Some aspects of the study are compelling including the use of Vil-Flp as a means to restrict recombination to the EECs and the taste aversion caused by activating all EECs or Tac1-lineage EECs during conditioning was profound and quite surprising. But the complex temporal effects of EEC activation on feeding and variation shown by controls make some of the results on this far less compelling and this aspect of the study should be toned down or reassessed. There were also some aspects of the data presentation that could be tightened to make the manuscript more accessible to the reader.

  3. Reviewer #2 (Public Review):

    Animals ingest food to replenish energy, a process that begins with ingestion, digestion in the stomach and small intestine, absorption of nutrients by blood capillaries innervating the intestine, and circulation of nutrients to body tissues that utilize or store energy. The entire process of realizing the metabolic consequences of ingestion occurs over minutes to hours, a timescale that is not conducive for guiding moment-to-moment behavioral actions during food ingestion, such as food choices. Moreover, because of the delayed metabolic consequences of ingestion, it is important to predict the caloric load to cease food intake and avoid internal damage caused by overconsumption. How are these predictive signals generated?

    In this manuscript, Bai et al investigate whether enteroendocrine cells (EECs) are involved in relaying nutritive signals to the brain to influence food preference learning and satiety. Notably, EECs in the proximal intestine are known to express chemoreceptors that could sense chemical properties of food that has already been partially digested by the stomach. In turn, it is well-established that EECs release neurotransmitters that can activate peripheral sensory nerve endings, which relay gut-related signals to the brain where they can become integrated with taste. Bai et al make several critical advancements in 1) developing tools for systematic investigation of EEC function, 2) providing some of the first direct evidence showing that activation of EECs suppresses food intake and 3) delineating mechanisms by which EECs modulate feeding behavior.

    •The authors supplement existing single-cell RNA sequencing of the proximal intestine to identify a population of EECs that had not been previously described via sequencing and unsupervised clustering (Cck/Tph1).
    •Development of Villin-Flp mice, which can be bred with Cre driver lines to selectively manipulate EECs.
    •Several subpopulations (Fev, Tac1, Cck, Gcg) suppress food intake when selectively activated using DREADDs.
    •Suppression of food intake by Tac1 EECs requires 5-HT3R activation, whereas suppression of food intake via Cck EECs requires CCKA-R.
    •Activation of Tac1 EECs can produce a conditioned taste aversion, whereas activation of Cck EECs can condition a flavor preference.
    •Both TAC1R and 5-HT3R activation are required for Tac1 EEC activation to condition a taste aversion. This is interesting because TAC1R was not required for suppression of food intake.
    •Blockade of CCKAR abolishes the ability of Cck EECs to condition a flavor preference. This is consistent with published literature regarding CCK and appetition.
    •Conditioned taste aversion induced by activation of Tac1 EECs is partially attenuated by ablating TRPV1 spinal sensory afferents.
    •Conditioned flavor preference induced by activation of Cck EECs is entirely abolished by vagotomy.

    •This is a tidy, well-controlled study without any glaring issues.

  4. Reviewer #1 (Public Review):

    Post-ingestion signals provide important information about the nutritional value of food and warnings of potential toxicity. In the gut, there are a sparse group of cells inside the epithelium, called Enteroendocrine cells (EECs) that serve as detectors of mechanical and chemical stimuli. EECs can be chemosensory, mechanosensory or both. Upon stimulation, they release signaling molecules (e.g. transmitters, peptides and hormones) that communicate directly with afferents in the gut or as circulating factors. Previous work has shown EECs are functionally heterogenous and their signaling promote satiety, malaise, and associative learning. This study uses single cell sequencing to provide details of EEC transcriptomic diversity. Bai and colleagues use the transcriptomic information to devise intersectional genetic strategies to manipulate the function of different collections of EECs. For the most part, they focus on two groups that co-express the gene Vilin and either CCK or Tac1. Generally, chemogenetic stimulation of most types of EECs slows down food consumption, but the authors contend for different reasons. They show that mice learn to prefer flavors when paired with CCK+/Villin+ cell stimulation whereas mice learn to avoid flavors paired with Tac1+/Vilin+ cell stimulation. Bai and colleagues go on to use pharmacological interventions to investigate some of the signaling pathways potentially involved and coarse ablation/transection manipulations to implicate distinct neural pathways.

    Generally, this is an interesting and potentially useful study. The transcriptomic data, particularly when combined with the previous work from the Clevers' lab and others, helps better define different kinds of EECs. The Vilin-Flp strain appears to be a very useful tool for intersectional studies of cells in the digestive system. The perturbation of feeding during chemogenetic stimulation is clear, particularly the large effect seen when CNO is given to mice expressing hM3Dq in Tac1+/Vilin+ cells. Of course, the model that aversive signaling occurs via 5-HT signaling between Tac1+ cells and nociceptive DRG neurons is consistent with recent work from the Julius lab and others. Furthermore, CCK signaling is known to promote satiety associated with palatable and nutritious foods and so the function Bai and colleagues assign these EECs makes sense. I'm less sure about the specificity of the genetic strategy. It is of course difficult to survey the entire body to know that just EECs are being manipulated. Equally, one wonders if EECs expressing Tac1 or CCK are equivalent in the various specialized compartments of the digestive tract. Additionally, the large variation across controls is somewhat surprising, given feeding assays like these in hungry mice tend to be very robust and reproducible across strains. In some cases, the difference across strains seems larger than the manipulations.