regions identified based on the adjacent cresyl violet-stained sections (Kopnoczky et al., 2005). We used background correction to eliminate potential differences in optical densities across different sections in different experiments. The average of eight background determinations carried out near the brain protein-containing areas of the immunostained nitrocellulose membranes was subtracted from the average pixel densities measured within brain regions. Following background corrections, the average pixel density for the whole region from one animal counted as one ‘n’. Under these conditions, labelling performed on different days produced very consistent results. Data were analysed and plotted using the software Analysis (Soft Imaging Systems, Munster, Germany).
