DeutschClear Cookie - decide language by browser settings
Lewis, Laurence (2016): A neural circuit for resolving sensory conflict in drosophila. Dissertation, LMU München: Graduate School of Systemic Neurosciences (GSN)



Animal habitats are highly complex and encode a surfeit of potentially meaningful information relevant to an animal’s immediate and future survival. Animals must sense and decode this complex environmental information in order to initiate an optimal behavioural response capable of meeting its survival requirements. Some environmental stimuli may elicit innate or learned attraction or aversion, while others may trigger courtship behaviour or grooming. However, the same stimuli may have different meaning to the animal depending on its internal state and the presence or absence of other stimuli. Thus, sensory cues often conflict with one another, requiring the animal to decide on which sensory objects to respond to. Like all animals, the fruit fly Drosophila melanogaster must resolve this sensory conflict within its olfactory environment. An ecologically relevant example of this is the conflict between the strongly aversive odour carbon dioxide (CO2), which is emitted by fermenting fruit, Drosophila’s primary food source, and vinegar volatiles, which include many attractive products of fermentation. It has been shown that when starved, the fruit fly Drosophila melanogaster is capable of overcoming its innate aversion to CO2 to approach the conflicting food odour vinegar (Bräcker et al., 2013). This implies a neural mechanism whereby a neural representation of vinegar is able to suppress the innate capability of CO2 to drive aversion behaviour. It was found that a structure in the fly brain commonly associated with olfactory learning, the mushroom body (MB), is required for processing innate CO2 aversion in a starvation dependent manner (Bräcker et al., 2013). This suggested a possible common neural substrate for both learned and innate sensory processing and behaviour execution. To examine a possible role for the Drosophila MB in weighing conflicting olfactory inputs I conducted a systematic olfactory behavioural screen in which I tested responses to CO2, vinegar, and the conflicting CO2 plus vinegar. During testing MB neurons were thermogenetically inactivated via the targeted expression of UAS-Shibirets1 to the three primary populations of neurons composing the MB: MB output neurons (MBON), MB neuromodulatory input neurons, and MB Kenyon cell (KC) interneurons. The behavioural screen recapitulated data previously obtained in our lab identifying the α’/β’ KCs as playing the dominant role in representing CO2 (Bräcker et al., 2013). The screen also identified MBONs innervating the β’2 region of the MB horizontal lobe as being required for CO2 avoidance. Subsequent optogenetic behavioural experiments demonstrated that the same MBONs were sufficient to drive aversive behaviour, providing a putative MB circuitry for mediating CO2 aversion. Additional behavioural experiments and calcium imaging revealed that dopaminergic neurons (DAN) innervating the same β’2 anatomical compartment represent vinegar, and that their activation is sufficient to reduce CO2 aversion. This observation suggests DAN inhibition of the flow of CO2 olfactory information from the presynaptic KCs to the postsynaptic MBONs. Indeed, these findings are consistent with our current understanding of MB neuromodulation by DANs, which are known to represent contextual information in the formation of associative memory and impinge on the KC-MBON synapse. Critically, it is demonstrated here that MBONs respond more weakly to a combination of vinegar and CO2 than they do to CO2 alone. Taken together my findings suggest that the β’2 MBONs are sufficient to mediate CO2 aversion, and that this sufficiency is dampened by vinegar responsive β’2 DANs, thus allowing flies to overcome their innate CO2 aversion in the context of appetitive vinegar stimulus. The implication of my work is that the processing of sensory information to drive immediate behavioural responses is carried out by the same structure responsible for forming lasting associations, the MB. The economy of energy and neural substrates available for performing the processing tasks required of brains often results in a convergence of multiple functions onto few brain structures. In the present case the inherent similarities between the tasks of immediate and lasting modulation of behaviours supports this parsimony.