Skip to main content

In vivo pharmacogenetic investigation of 5-HT mechanisms in emotional learning

Periodic Reporting for period 1 - 5HT-DREADD (In vivo pharmacogenetic investigation of 5-HT mechanisms in emotional learning)

Reporting period: 2016-07-01 to 2018-06-30

Serotonin is critical for mood regulation in health and disease and strongly implicated in the vulnerability to anxiety and depression. Importantly, accumulating evidence suggests that the main influence of serotonin is upon emotional learning rather than on mood directly. The objective of this project was to determine the neural mechanisms through which changes in serotonin levels, within defined neuronal circuits, affect learning and memory for aversive emotional experiences. However, the neuronal circuitry involved and how serotonin activity translates into behaviour are yet to be fully elucidated. To explore the effect of specifically manipulating serotonin circuitry on emotionality, I have utilised a chemogenetic approach. This involves inserting new genetic code into certain neurons that leads to the expression of new receptors only in those specific cells. These receptors do not respond to any chemical in the brain but instead are only activated by a drug I then deliver weeks later. These are therefore called DREADDs; Designer Receptors Exclusively Activated by Designer Drugs.
In this project, I introduced these receptors to serotonin cells in the dorsal raphe nucleus (DRN), the key source of serotonin innervation in the brain. However, these are not the only cells in that area as they are controlled by local inhibitory gamma aminobutyric acid (GABA) neurons. I therefore also explored how expressing these receptors and selectively activating these local GABA neurons affected the functionality of the serotonin system and the subsequent effect on behaviour. This selectivity is achieved by delivering the new genetic code to the DRN with viruses containing the information for these receptors (hM3Dq) as well as a fluorescent “tag”, or just this “tag” (mCherry) to act as a control. Different lines of mice then allow for this to either be expressed only in serotonin neurons (SERT-Cre mice) or the GABA neurons (vGAT-Cre mice).
I first ensured that this expression was indeed selective by using immunohistochemistry. This involved visualising the “tag” along with staining for tryptophan hydroxylase (TPH2), a marker for serotonin neurons. Once selective targeting of serotonin and GABA neurons was established (with more than 95% specificity), I tested the functionality of this system by measuring the effect of the DREADD activator, clozapine-N-oxide (CNO), on the expression of the immediate early gene cFos, a marker for recent neuronal activity. I found that CNO significantly increased cFos expression in the DRN in both the SERT-Cre and vGAT-Cre mice in the correct neuronal populations. In addition, CNO increased cFos expression in the basolateral amygdala (BLA) of SERT-Cre mice, suggesting serotonin activation of downstream anxiety-related circuits.
Once serotonin-selective and functional expression of these receptors was confirmed, I tested the effect of manipulating neuronal activity within the DRN in various anxiety models. On the elevated plus maze, I found that increasing the excitability of serotonin neurons through expression and activation of these receptors in the SERT-Cre mice caused an increase in time spent in the open arms. This suggests that it makes them less anxious. vGAT-Cre mice did not show this effect. In two other tests of anxiety (the light and dark box and novelty supressed feeding), however, I saw no changes in the SERT-Cre mice. Surprisingly, vGAT-Cre mice spent a significantly reduced time in the light compartment of the light and dark box while at the same time exhibiting increased exploration of the central area of an open field, suggesting increased and decreased anxiety, respectively. These findings suggest that chemogenetic activation of serotonin neurons reduces anxiety, while chemogenetic activation of GABA neurons in the DRN increases anxiety, at least under certain conditions, and might be doing this independently of serotonin.
I then set out to characterise whether chemogenetically activating serotonin neurons in the raphe would have an impact on the acquisition and retrieval of fear memories. Surprisingly, I found no effects using this system in several variations of fear conditioning, a well-used test to investigate negative emotional memory formation and retrieval. These results suggested that chemogenetically activating serotonin neurons in the DRN is not sufficient to enhance nor disrupt the acquisition and retrieval of emotional fear memories.
Given the lack of a strong behavioural phenotype despite the strong activation observed in the immunohistochemistry experiments, I have recently set out to further validate the functionality of these constructs to assess the extent of the activation of individual neurons within the DRN and whether this leads to network-level effects. To test this, I have measured the effect of CNO on the individual firing patterns of large numbers of neurons in the DRN in anaesthetised SERT-Cre mice by using state of the art multiunit recording with 32-channel silicon probes. This is the first time this method has been applied to characterise the effect of chemogenetically activating discrete populations of neurons within the DRN. This method has the advantage of providing data from a much larger group of neurons than can be achieved with more traditional wire electrodes as each silicon probe has many contacts that each act as an individual point of measurement. This increases the number of neurons that are recorded from and so increases the likelihood of recording from cells expressing the DREADDs as well as other non-DREADD expressing neurons. In preliminary ongoing experiments we have found that CNO seems to elicit opposing effects in neurons within the same animal which depends on the type of neuron being recorded. Thus, slow, regular firing neurons, typically thought to be serotonin neurons, seemed to not respond to CNO treatment while some slow irregular firing units showed an increase in their firing rate when compared to baseline. These differences may be due to not all the neurons recorded expressing DREADD receptors, but could also reflect downstream changes to the activity of the area in response to the altered activity of a small population of neurons that express the inserted receptors. These experiments continue and will also be extended to vGAT-Cre mice. It is therefore expected that a vast increase in the number of subjects analysed will enable us to identify an increase in activation in putative serotonin and/or GABA neurons, something that will be assisted by other ongoing work in the lab to produce a better picture of how these populations of neurons behave electrically.
Together, the data obtained in this project indicate the successful chemogenetic targeting of DRN serotonin neurons to explore the role of serotonin circuitry involved in anxiety and fear behaviour but that the role of DRN GABA neurons may be more important to the control of emotional behaviour than the serotonin neurons themselves. These results will help better understand the mechanisms underlying the control of emotional behaviour and may contribute to identify better candidates for treatments aimed at ameliorating psychiatric conditions in which emotional behaviour is altered.
Schematic representation of the work included in the project.