Skip to main content
. 2017 Jan 4;37(1):164–183. doi: 10.1523/JNEUROSCI.0997-16.2017

Figure 7.

Figure 7.

Component-motion and pattern-motion selectivity in layer 2/3 vs layer 4 of mouse V1. A, B, Scatter plot of component and pattern partial correlation coefficients calculated for each direction-of-motion-selective pyramidal neuron recorded in layer 2/3 of mouse V1 (A) vs layer 4 of mouse V1 (B), following the study by Smith et al. (2005). The x-axis plots the Rc value for each cell. The y-axis plots the Rp value for each cell. Dotted black line boundaries demarcate statistical significance levels calculated from the Fisher transform (see Materials and Methods; Smith et al., 2005) at p = 0.9. We use the p = 0.9 value here, since this is the value that has generally been used by previous investigators (Movshon and Newsome, 1996; Smith et al., 2005). Direction-tuning data were obtained using the fluorophore GCaMP6s and a stimulation protocol similar to the one used for OGB-1 data collection. A, Results for the layer 2/3 of V1, using 120° CA plaid (square-wave-gratings). The shape of the Rp–Rc distribution and the fraction of cells displaying pattern and component selectivity are essentially similar to the layer 2/3 data obtained using OGB-1 staining (Fig. 4B). Data were collected from four fields of view located at a depth of 130–230 μm. Of 236 cells that displayed direction-of-motion selectivity for the gratings, 23 cells were also pattern-motion selective (∼10%) and 41 cells were also component-motion selective (∼17%). B, Pattern-motion and component-motion selectivity in deep layers of mouse V1. Black filled dots, Results for layer 4 of mouse V1 obtained using the 120° CA plaid. Data were collected from five fields of view located at a depth of between 360 and 460 μm. Of 190 cells that displayed direction-of-motion selectivity for the gratings, 16 cells were also pattern-motion selective (∼8%), and 13 cells were component-motion selective (∼7%). Gray dots show results from 52 putative apical L5/6 dendrites recorded in L4 of mouse V1 (three FOVs, 400–460 μm deep) using the 120° CA plaid. Overall, the percentage of pattern-motion-selective responses detected in deeper layers (6–9%) is commensurate to that seen in L2/3 (9% in the GCaMP6s experiments; 6% in the OGB-1 experiments reported in the text).