This regular arrangement suggests a functional specialization of tectal laminae, which may explain how a nascent circuit can readily perform computational tasks while being under construction. We used the binary UAS-Gal4 system to target fluorescent reporter constructs to tectal neurons. We first generated a transgenic line that expresses the transcription factor Gal4 under

control of the panneuronal promoter huC ( Kim et al., 1996). When these fish were crossed with a Tg(UAS:GFP) reporter line, offspring larvae showed green fluorescent protein (GFP) expression throughout the CNS. In the retina we observed fluorescently labeled RGCs, which project http://www.selleckchem.com/products/Dasatinib.html to the superficial layers of the tectal neuropil. In the optic tectum, most Epacadostat chemical structure cell bodies as well as their dendritic arbors were labeled. Consistent with

these findings, we observed GFP-positive layers throughout the tectal neuropil ( Figure 1A). In order to identify DS neurons in the larval zebrafish tectum, we targeted GCaMP3, a genetically encoded Ca2+ indicator (GECI), to tectal neurons by crossing Tg(huC:Gal4) with Tg(UAS:GCaMP3). This obviates the need for dye-loading protocols that could interfere with neural circuit function ( Tian et al., 2009; Del Bene et al., 2010; Dombeck et al., 2010). In the offspring larvae, GCaMP3 was expressed in a similar pattern as GFP ( Figure 1B). Our experimental setup consisted of a custom-built multiphoton microscope and a miniature projector to display

moving bars on a screen that surrounded the imaging chamber (Figure 1C). We imaged neurons in the central region of the cell body layer in the contralateral tectal hemisphere (Figure 1B, dashed box). During visual stimulation with moving bars (eight equally spaced directions that covered 360°; 0° corresponds to a bar moving in the caudorostral [CR] direction), many neurons showed DS Ca2+ responses. Figures 1D1–1D4 show Ca2+ signals from four somata imaged in one experiment. From the peak amplitude of the Ca2+ about transients for each stimulus direction, we calculated the preferred direction (PD) and direction selectivity index (DSI) of all responsive neurons (Figure 1E; see Supplemental Experimental Procedures available online). We defined those neurons with DSI ≥ 0.3 as DS. We observed that all directions were represented in the labeled tectal cell population, although the distribution of PDs was not uniform (p < 0.001, Hodges-Ajne test for circular uniformity; Figure 1F). A large fraction of neurons responded to the CR stimulus (41.3% with PD ∈ [315°, 22.5°]), while other cells exhibited DS for stimuli with a rostrocaudal (RC) component (42.6% with PD ∈ [90°, 270°]). It should be noted that this distribution may represent a subset of all tectal DS neurons because weakly active cells may be missed using Ca2+ imaging.