Adult stem cells maintain tissue homeostasis by continuously replenishing non-functional cells with healthy ones. Most stem cells identified to date are compartmentalized in functionally deterministic niches. From there stem cells are instructed by unique combinations of signals and spatial tensile forces which they translate into a specific behavior. However, how stem cells spatiotemporally coordinate their intrinsic stem cell potential with niche- and systemic cues is still poorly understood. These issues are essential for proper tissue function, since perturbations in the signals that govern stem cell function can cause tissue malfunction such as tumorigenesis and ageing.

We have identified a mechanism of stem cell regulation based on circadian rhythms. The circadian machinery anticipates and synchronizes the daily function of tissues according to the entrainment by natural changes in light and metabolism. We have observed that the molecular clock fine-tunes the behavior of epidermal stem cells by imposing transcriptional oscillations in the expression of stem cell regulatory genes. This ¨circadian transcriptome signature¨ includes genes involved in development, metabolism, drug-metabolism, and cell invasion and motility, among others. Circadian oscillation of these genes provides epidermal stem cells with a spatiotemporal axis for responding to dormancy, activating, and pro-differentiation cues. Interestingly, circadian regulation creates populations of stem cells co-existing within the same niche which are more predisposed than others to respond to activating cues. In addition, the clock imposes a proper timing of epidermal stem cell function by ensuring that they primarily respond to harmful UV radiation during the hours of peak exposure to light, to subsequently and sequentially become responsive to proliferation and pro-differentiation cues in the evening and night.

We hypothesize that this mechanism ensures that the right number of stem cells respond at a given time, avoiding unnecessary proliferation and its associated risk of DNA damage. Also, by timing the function of stem cells this mechanism prevents their proliferation in instances of high risk of DNA damage at the hours of peak UV exposure, but stimulates their activation and differentiation when enough energy supply is available. Notably, forced circadian arrhythmia in epidermal stem cells causes severe tissue ageing and modifies the predisposition to tumorigenesis.

Humans excel at reasoning about novel situations, flexibly combining abstract knowledge and perceptual information from disparate sources in "one-shot" intuitions to predict outcomes of events they have never before directly experienced. To investigate the root of this ability, we study how infants reason in very simple, but completely novel situations which contain potentially conflicting information. Consider a simple situation in which three yellow objects and one blue object bounce inside a container with an opening on its lower side, as in a lottery machine (Figure 1). At a certain point, an occluder hides the container. You have to bet which object will exit first: will it be blue or yellow? The right answer is: it depends.
The scene contains different kinds of information: about the properties of the objects, such as their colors; about the number of objects in the scene; or about the dynamics of their trajectories, collisions and changes in speed. All such potentially conflicting cues can be relevant to form optimal expectations about which object will exit the container. What is crucial is that the most relevant information can change according to subtle changes in how the scene unfolds.

We showed prelinguistic infants such lottery-like scenes, by systematically varying the length of the occlusion prior to the exit of the object. When the occlusion was long, infants were surprised when an object of the less numerous category exited the container, irrespectively of its distance from the exit before the occlusion. This makes sense, because if the occlusion is long, the last position of an object is not relevant to predict what object will exit, as objects keep moving behind the occluder. Instead, when they saw the same situation but the occlusion was very short, infants were surprised when a far object exited, regardless of its color. Also this makes sense, because if an object is far from the exit, it becomes physically impossible for it to travel the distance to the exit during the short occlusion. Thus, infants have rational "gut feelings" about future events.
In our study, we also elaborated a Bayesian ideal observer model which precisely predicts infants' looking times in our experiments and extends to other known results about infant cognition, providing for the first time a unifying explanation of several classic findings. This model shows how powerful pure reasoning capacities could derive from the operation of probabilistic inference mechanisms constrained by abstract principles of how