The second to last sitting before the BAP conference dinner gave attendees the choice between three excellent short oral sessions. The mechanisms of anxiety and its treatment session included four speakers and covered a wide of interesting approaches to study anxiety, from functional imaging in humans through to gene expression in non-human primates.
First up was Shaun Kit Lung Quah, with his talk titled:
SEROTONIN TRANSPORTER (SLC6A4) EXPRESSION IN THE RIGHT AMYGDALA AND SEROTONIN 2A AND 2C RECEPTOR EXPRESSION RATIO (HTR2A:2C) IN THE RIGHT AMYGDALA AND DORSAL ANTERIOR CINGULATE CORTEX (DACC) ASSOCIATES WITH ANXIOUS BEHAVIOUR –
The aim of this work was to identify differential expression of serotonergic genes within the amygdala and dorsal anterior cingulate cortex (dACC ) that may be associated with anxious behaviour in marmosets. To assess anxious behaviour they took the innovate step of combing all the standard metrics of anxiety in the human intruder test (e.g. time spent in the front of the cage, head and body bobbing, vocalisations) and used principle component analysis to derive a single score for anxious behaviour. They were then able to look for relationships between this score and relative expression of target genes: the serotonin transporter (SLC6A4), and the Serotonin 2A and 2C receptor expression ratio (HTR2A:2C). They found that SLC6A4 expression in the right amygdala only correlated positively with anxious behaviour. In contrast, the Serotonin 2A:2C receptor ratio (HTR2A:2C) in the right amygdala and bilateral dACC correlated negatively with anxious behaviour. A backwards regression model with many predictors revealed that the two gene expression measures in the right amygdala explain an impressive 85% of the variance in anxiety behaviour. In a follow up study serotonin levels in the amygdala were increased directly with citalopram, which was shown to reduce anxious behaviour in the marmosets. These findings add further support to the theory that low serotonin levels in the amygdala leads to anxiety. The authors also tested gene expression in other parts of the prefrontal cortex (medial prefrontal, orbitofrontal, ventrolateral prefrontal) but found no relationship with anxious behaviour. Taken together this work adds further insight into previous findings showing that the functional integrity of the dACC-amygdala circuit, is crucial to pathological anxiety.
The next speaker in this session was Toby Wise, and his talk was titled:
NEURAL SYSTEMS UNDERLYING ACTIVE AVOIDANCE BEHAVIOUR IN MAJOR DEPRESSION
In many ways anxiety can be considered an avoidance disorder, with individuals employing active and passive strategies to escape from anxious feelings. What I especially liked about this talk was that the work acknowledged, and addressed, the comorbidity between symptoms of anxiety in depression. Toby wise employed functional magnetic resonance imaging in humans with a diagnosis of major depressive disorder to assess active avoidance of mild electric shocks, relative to healthy controls. In this task, participants had exert physical effort (squeezing a gripper) to avoid mild electric shock to their foot. In so doing, the aim of this study was to examine the neural mechanisms of anticipation and avoidance of aversive stimuli. Consistent with previous research, the results showed that anticipation of electric shocks activated dorsal anterior cingulate (dACC) and striatal regions. Interestingly, activity in the dACC and supplementary motor area was predictive of later grip squeeze force in the avoidance part of the trial. Avoidance itself recruited the dACC, and the insula – with MTG and striatal activity in the avoidance phase predicting intensity of assessed risk during the task. Hence, overall avoidance is associated with activation of key regions in the anxiety network (c.f. the dACC in the marmoset study before) and relate specifically to metrics of avoidance behaviour in the task. Interestingly, however, there were no group differences between patients with MDD and healthy controls. This may provide some initial tentative evidence that pathological anxiety disorder involves functionally distinct neural mechanisms to mood disorders that include anxiety symptoms. Toby Wise also noted that the mechanisms of passive avoidance, as opposed to active avoidance, may be selectively impaired in MDD.
The third speaker in this session was Franziska Goer, and the title of her talk was:
FMRI INVESTIGATION OF THE EFFECT OF INDUCED ANXIETY ON PAVLOVIAN BIASES DURING REINFORCEMENT LEARNING IN HEALTHY AND ANXIOUS INDIVIDUALS
Keeping with functional MRI in humans, this study examined Pavlovian behavioural biases in pathological anxiety directly. The authors used a common go/no-go reinforcement learning paradigm (i.e. making a go response to gain money is consistent with Pavlovian biases, whereas no-go to win money is inconsistent with Pavolvian biases) under induced anxiety (threat of shock) and neutral (no threat of shock) conditions. They found that threat of shock produced increased activation in the visual cortex in people with pathological anxiety relative to healthy controls. This is interesting considering the suggestion that anxiety may be considered a state in which one is primed for threat – an effect that may be evident even in early visual processing regions. The behavioural data from the study was then treated with computational reinforcement learning models. Following formal comparison procedures, the winning model was shown to simulate the behaviour well. This is reassuring and I am pleased that the authors provided such extensive validation of their model. This approach then allowed for an examination of the neural correlates of a computational parameter that captures Pavlovian influences on behaviour. In healthy controls, the Pavlovian bias is driven by activity in the striatum, but not in people with pathological anxiety. These fascinating results leave open the question of what exactly drives the avoidance bias in patients, which the authors hope to address with follow up work.
The final talk in this session came from Martina Di Simplicio, and was titled:
THE EFFECT OF FOUR WEEKS SSRI ADMINISTRATION ON NONVERBAL BEHAVIOUR IN HIGH AND LOW NEUROTIC HEALTHY VOLUNTEERS
This study examined the effects of experimentally increasing serotonin (via 4 weeks of 20mg citalopram administration) on non-verbal behaviour in individuals with high- and low neurotic traits. A particularly interesting innovation here was to use coded videotaped behaviour during a naturalistic dyadic interaction as the dependant measure of non-verbal behaviour, thus avoiding confounds associated with self-report methods. Dr Simplicio found that there was no drug effect on non-verbal behaviour overall, but a significant effect of citalopram in the high neurotic group on aggressive behaviours, reflecting reduced low-level hostility and assertion following SSRI treatment. Interestingly, this effect correlated with performance in a separate face recognition task in the high neurotic group – suggesting a link to cognitive measures of social processing. There were no significant drug effects on affect, depression and anxiety levels. The finding might represent changes in interactions with other people in response to the drug which may support clinical improvements in symptoms. These results are also especially interesting in the broader context of better understanding the effects that individual differences in personality have on the variable response to antidepressant medication.
Taken together, this was a fascinating session that linked across expression of serotonin receptors in marmoset models of anxiety, the effects of pharmacological manipulations of serotonin on non-verbal behaviour in neuroticism, and fMRI investigations of passive and active avoidance in pathological anxiety and co-morbid major depressive disorder. All four studies converge on the importance of the integrity of the serotonin system, and dACC-amygdala-striatal pathways in supporting active, passive and non-verbal behaviours associated with anxiety.