Immunohistochemistry and confocal microscopy
For c-Fos immunofluorescence, series of sections were collected from -1.6 to -2.0 mm i.e. the AP intermediate BA. Using the same custom-written Fiji macro as described for the BL/6 mouse experiment, the numbers and threshold gray values for CTB+, tdTom+, c-Fos+, CTB+/tdTom+, CTB+/c-Fos+ and CTB+/tdTom+/c-Fos+cells/neurons were determined. The numbers and densities of cells/neurons were calculated as the sum of [mean of both hemispheres per section] for 5 sections (-1.6, -1.7, -1.8, -1.9, -2.0). In addition, for c-Fos, the integrated density of each c-Fos+ BA-NAc neuron was calculated as area x mean gray value.
The TRAPing parameters used were optimized by pilot study, in which reporter expression by BA cells (not specifically BA-NAc neurons ) was quantified. A maximal efficacy of about 30% tdTom labeling of c-Fos labeling was realized, i.e. if a stimulus (e.g. SR, SA, footshock) resulted in an average of 100 c-Fos+ cells, it resulted in an average of 30 tdTom+ cells. Protocol changes, including higher 4-OHT dose, signal amplification with tdTomato immunostaining, did not increase TRAPing efficacy.