In Drosophila, T4 and T5 neurons are the first stage along the visual processing chain that respond to visual motion in a direction selective way (Maisak et al. 2013, Serbe et al., 2016). To describe how such direction selective responses in T4 and T5 come about, a so called 3-arm detector model was proposed recently (s. Figure 1). This model relies on the comparison of signals originating from neighboring points in space via a delay-and-compare mechanism. The central arm provides fast excitation to the neuron, while the two flanking arms possess slow dynamics, one amplifying the central signal implementing for stimuli moving along the preferred direction, the other inhibiting the central signal for stimuli moving along the null direction of the neuron (Haag et al., 2016; Arenz et al., 2017).
In order to probe the theoretical model in vivo and map the different model elements to specific cell types in the circuit, nearly all input cells of T4 and T5 have been analyzed, describing their temporal and spatial response properties and their role for direction selectivity (Ammer et al., 2015; Serbe et al., 2016; Arenz et al., 2017). In contrast to the variety of physiological studies, only little is known about neurotransmitter phenotypes and receptors of the different circuit elements. The main inputs to T4 are medulla neurons Mi9, Mi1, Tm3, Mi4, C3 and CT1 (Takemura et al., 2017). Immunohistochemical stainings and RNAseq analysis showed that Mi9 is glutamatergic, Mi1 and Tm3 are cholinergic and Mi4, C3 and CT1 are GABAergic (Pankova et al., 2016; Takemura et al., 2017; Davis et al., 2018; Richter et al., 2018). The presynaptic partners of T5 in the OFF-pathway- Tm1, Tm2, Tm4 and Tm9- are cholinergic (Shinomiya et al., 2014) (s. Figure 2, 3).
Our knowledge about receptors on T4 and T5 dendrites, which receive the repertoire of different neurotransmitters, is rather limited. Two RNAseq. studies (Pankova & Borst, 2016; Davis et al., 2018) showed that T4 as well as T5 express receptor subunits for all main neurotransmitter types, acetylcholine, glutamate and GABA. However, RNAseq. approaches only take into account mRNA levels of genes, omitting protein levels and the distribution of the protein along the neuron. Yet, for the cellular implementation of our model, the organization of transmitter receptors on T4/T5 dendrites plays a crucial role. Since the 3-arm detector proposes enhancing and suppressing inputs to T4/T5, an asymmetric distribution of excitatory vs. inhibitory neurotransmitter receptors along T4/T5 dendrites is expected.
To investigate this hypothesis, we analyze the expression pattern and spatial distribution of receptors on T4/T5 dendrites with the goal to link the anatomical phenotype to the functional role of neurotransmitter receptors in the computation of direction selectivity.