Final Report Summary - 5HT-OPTOGENETICS (Optogenetic Analysis of Serotonin Function in the Mammalian Brain)
This project aimed to elucidate the behavioural function of the serotonin (5-HT) system by employing optogenetic tools to specifically manipulate 5-HT neurons under precise temporal control in behaving rodents. Optogenetic constructs were delivered with genetically engineered adeno-associated viruses (AAVs) using the promoter Sert (SERT-Cre mice), which is selectively expressed in 5-HT neurons.
To establish causal links between 5HT and its functions, we tested and optimized methods to monitor (genetically encoded fluorescent calcium indicators; GCaMPs), activate (channelrhodopsin-2; ChR2) and silence 5-HT release during behavioral assays. In this process, we developed a novel approach to record activity from genetically identified 5-HT neurons by measuring the global activity of 5-HT neuronal populations over extended time periods. We expressed cre-dependent GCaMP6 in the DRN and, using custom single fiber recording system, we validated this approach by measuring the activity of medium spiny neurons in free behavior (Tecuapetla et al., 2014). To overcome potential movement artifacts during awake recording from deep structures we developed a novel dual-color fluorescence approach using both activity dependent GCaMP (green channel) and an activity independent red fluorescent protein reference (red channel) ) (Matias, 2015)
Using these tools, we assessed a number of leading hypotheses of 5-HT function:
A prominent theory of 5-HT function is the modulation of anxiety and behavioral responses to aversive events. We selectively activated DRN 5-HT neurons in classical assays for anxiety-like behaviors including the open field and conducted detailed quantitative analyses of the effects of 5-HT stimulation on behavior. Surprisingly, we found that DRN activation produces a rapid and robust decrease in locomotion rather than affecting anxiety-like behaviors (Correia, 2015).
The 5-HT dopamine (DA) opponency hypothesis proposing 5-HT carries a negative reward prediction error signal. To test this, we monitored 5-HT DRN activity in a reversal learning task. We discovered that 5-HT neurons responded more to unexpected rewards than punishments, but also responded to worse-than-expected outcomes (Matias, 2015). We also optogenetically activated 5-HT DRN neurons on a series of behavioral assays that assess reinforcement, place preference tasks and a value-based choice task, finding no positive or negative reinforcement effects of 5-HT place (Fonseca et al., 2015).
5-HT is prominently implicated in the control of impulsive actions. We selectively activated DRN 5-HT neurons in a delayed response task that required waiting for a cue to receive a reward (Murakami et al., 2014). We found that 5-HT activation increases the willingness to wait for the delayed rewards, suggesting that 5-HT is sufficient to promote patience (Fonseca et al., 2015).
5-HT is also associated with effects on sensory processing but their nature is poorly understood. We found that releasing 5-HT in olfactory cortex by optogenetic stimulation of 5-HT neurons specifically inhibits spontaneous but not odor-evoked activity (Lottem et al., 2016). In a parallel series of experiments, we also found that 5-HT lessens behavioral responses to mechanosensory stimulation (Dugué et al., 2014).
Taken together, these data led us to propose a novel hypothesis: that 5-HT neurons report the uncertainty or surprise associated with the predictions generated by internal models (priors) and those signals are used to modulate the impact of that information on behavior. This hypothesis formed the basis of a new series of studies that are now beginning.
To establish causal links between 5HT and its functions, we tested and optimized methods to monitor (genetically encoded fluorescent calcium indicators; GCaMPs), activate (channelrhodopsin-2; ChR2) and silence 5-HT release during behavioral assays. In this process, we developed a novel approach to record activity from genetically identified 5-HT neurons by measuring the global activity of 5-HT neuronal populations over extended time periods. We expressed cre-dependent GCaMP6 in the DRN and, using custom single fiber recording system, we validated this approach by measuring the activity of medium spiny neurons in free behavior (Tecuapetla et al., 2014). To overcome potential movement artifacts during awake recording from deep structures we developed a novel dual-color fluorescence approach using both activity dependent GCaMP (green channel) and an activity independent red fluorescent protein reference (red channel) ) (Matias, 2015)
Using these tools, we assessed a number of leading hypotheses of 5-HT function:
A prominent theory of 5-HT function is the modulation of anxiety and behavioral responses to aversive events. We selectively activated DRN 5-HT neurons in classical assays for anxiety-like behaviors including the open field and conducted detailed quantitative analyses of the effects of 5-HT stimulation on behavior. Surprisingly, we found that DRN activation produces a rapid and robust decrease in locomotion rather than affecting anxiety-like behaviors (Correia, 2015).
The 5-HT dopamine (DA) opponency hypothesis proposing 5-HT carries a negative reward prediction error signal. To test this, we monitored 5-HT DRN activity in a reversal learning task. We discovered that 5-HT neurons responded more to unexpected rewards than punishments, but also responded to worse-than-expected outcomes (Matias, 2015). We also optogenetically activated 5-HT DRN neurons on a series of behavioral assays that assess reinforcement, place preference tasks and a value-based choice task, finding no positive or negative reinforcement effects of 5-HT place (Fonseca et al., 2015).
5-HT is prominently implicated in the control of impulsive actions. We selectively activated DRN 5-HT neurons in a delayed response task that required waiting for a cue to receive a reward (Murakami et al., 2014). We found that 5-HT activation increases the willingness to wait for the delayed rewards, suggesting that 5-HT is sufficient to promote patience (Fonseca et al., 2015).
5-HT is also associated with effects on sensory processing but their nature is poorly understood. We found that releasing 5-HT in olfactory cortex by optogenetic stimulation of 5-HT neurons specifically inhibits spontaneous but not odor-evoked activity (Lottem et al., 2016). In a parallel series of experiments, we also found that 5-HT lessens behavioral responses to mechanosensory stimulation (Dugué et al., 2014).
Taken together, these data led us to propose a novel hypothesis: that 5-HT neurons report the uncertainty or surprise associated with the predictions generated by internal models (priors) and those signals are used to modulate the impact of that information on behavior. This hypothesis formed the basis of a new series of studies that are now beginning.