Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - 5HTCircuits (Modulation of cortical circuits and predictive neural coding by serotonin)

Teaser

Serotonin (5-HT) is a central neuromodulator implicated in the regulation of many processes and one of the most important targets for psychoactive drugs. It profoundly impacts decision-making and its dysregulation can contribute to altered perception as well as pathological...

Summary

Serotonin (5-HT) is a central neuromodulator implicated in the regulation of many processes and one of the most important targets for psychoactive drugs. It profoundly impacts decision-making and its dysregulation can contribute to altered perception as well as pathological conditions such as depression, anxiety or obsessive-compulsive disorders. Yet 5-HT’s function is not well understood, hence impeding progress towards better treatments. The broad aim of this work is to understand the involvement of 5-HT in decision-making under conditions of uncertainty. Uncertainty arises whenever partial, ambiguous, or contradictory information is present, a common occurrence in real-world situations. We propose that 5-HT neurons respond to increases in uncertainty caused by unexpected events. Such effects are hypothesized to modulate behavior at various time-scales. At an immediate time-scale, 5-HT release may promote persistence into ongoing behaviors when outcomes are uncertain. At longer time scales, it may increase behavioral flexibility in response to changes in the environment. Mechanistically, the control exerted by 5HT over these process may be mediated by its ability to modulate the neural representation of prior expectations at different levels of the brain hierarchy. Additionally, to complement our study of 5-HT and uncertainty at the circuit and brain-wide level, we also investigate the inputs and outputs of 5-HT neurons that allow them to exert their behavioral effects.

Work performed

Our work on serotonin can be divided in 4 complementary but synergistic axes of research.
First (Specific Aim 1, SA1), we studied how the activity of 5-HT neurons correlates with specific events such as the anticipation and reception of uncertain rewards or the processing of expected versus unexpected stimuli. In this line of research, we used innovative imaging techniques to record neural activity in the dorsal raphe nucleus (DRN) where most of the 5-HT neurons innervating the forebrain originate. Recently published results confirmed our hypothesis that the activity of 5-HT neurons reflects more than the affective properties of events and stimuli: they appear to encode a generic surprise signal independent of affective valence (a quantity sometimes termed “unsigned prediction error”).
Second (SA2), we investigated how causal manipulations of 5-HT neurons’ activity influence an array of behaviors related to locomotion, exploration, decision-making and learning. To do so, we used various techniques such as optogenetics (for excitation), chemogenetics (for inhibition) and pharmacological challenges (to unravel the functional role of different 5-HT receptors). Taken together, our results indicate that changes in 5-HT activity have consequences at multiple timescales in various contexts: they tend to modify the persistence of ongoing behaviors (such as actively waiting for predictable rewards) and the speed at which animals learn to predict events based on a variety of cues. Importantly, we further showed that short- and long-term effects often go in opposite directions: for example, activating 5-HT neurons immediately reduces locomotion in mice exploring their environment whereas the repetition of such stimulation over weeks tends to increase baseline locomotion.
Third (SA3), we developed experimental paradigms focused on olfactory processing, which constitutes a very interesting window on the brain function of rodents. Olfaction in rodents is supported by a well-characterized neural network which is densely innervated by serotonin. In particular, mice excel at predicting upcoming events based on the odors they detect in their surroundings. Thus, using robust experimental paradigms, we can investigate how predictive processes are shaped by serotonin signaling. We found that serotonin release in the olfactory cortex was able to suppress the spontaneous neural activity of olfactory neurons, a phenomenon which could translate into a reduced weight of predictable events for the guidance of behavior.
Fourth (SA4), we initiated a series of experiments combining functional magnetic resonance imaging (fMRI) and optogenetics in awake rodents. Because serotonin neurons innervate most of the forebrain, its net effects on behavior depend on the coordinated response of multiple target areas. However, conventional imaging methods in neuroscience are most often limited to 2-dimensional fields of view at a submillimeter scale, hence impeding our ability to analyze the effects of 5HT over brain-wide networks. fMRI in awake rodents entails several highly technical challenges which have been progressively solved. Thus, we are now able to measure whole-brain blood oxygenation levels (BOLD signals, an indirect marker of neural activity) in mice performing simple cognitive tasks, with or without optogenetic stimulation. We believe that this approach will provide an ideal complement to the more local dynamics investigated in the other research axes of this project.
Finally, we emphasize two additional features of our work on serotonin research. On the one hand, we aim at framing most of our studies in a general computational framework able to: (i) track and quantify some of the underlying algorithmic principles that organize individual behaviors ; (ii) draw qualitative predictions which can be used to arbitrate between different theories of 5-HT function. On the other hand, we aim at translating our findings in mice to humans by developing ho

Final results

We are now investigating the brain structures through which serotonergic signals exert their influence on behavior, and more particularly on the way individuals use these signals to learn in uncertain environments.
Many results already obtained in the context of this project go beyond the state-of-the-art of serotonin research. For example, very few labs have developed the technology required to perform calcium imaging experiments probing 5-HT neurons in vivo, in behaving animals. We showed that this is indeed possible and are now developing protocols to image 5-HT axonal firing and 5-HT release in target regions such as the nucleus accumbens, the amygdala or the prefrontal cortex.
Note that the study of 5-HT projections represents an effort to characterize not only the dynamics of 5-HT neurons’ activity but also the brain structures driving this activity and the brain structures mediating the impact of 5-HT release on short- and long-term behavioral changes.
We expect to report the results obtained using a new probabilistic foraging task in two major publications before the end of the project. One will describe in extensive detail the logics of the computational model which inspired the task, as well as the consequences of inhibiting different prefrontal cortex subregions. The other one will describe the impact of DRN 5-HT stimulation on performance and model parameters, as well as the activity of 5-HT neurons evoked by different stages of this task, both in the DRN and in the nucleus accumbens.
Furthermore, we expect to publish in the upcoming year an exhaustive investigation probing the development of the descending pathway linking the prefrontal cortex to the DRN across lifespan, in relationship with the emergence of a patient and nearly optimal behavioral profile in the foraging task mentioned above. These “circuit mapping” experiments which heavily relied on slice electrophysiology go beyond current approaches, in that it combines an extensive characterization of 5-HT neurons activity with behavior across time.
Finally, fMRI experiments in awake rodents remain a highly challenging endeavour for many labs around the world. To our knowledge, no other lab has investigated how 5-HT transmission modulates task-related BOLD signals. This approach is particularly important for translational research, given that fMRI is virtually the most powerful non-invasive neuroimaging technique used in human neuroscience and psychiatry. We expect to publish the paper reporting the effect of 5-HT over brain-wide task-related events in awake mice, together with data acquired in anesthetized mice before the end of the project.