Here, we used the Rebus Biosystems platform to quantify expression levels of 31 target genes across three experimental conditions. First, we obtained fresh frozen mouse brain sections from two transgenic lines modeling a neurodegenerative disease and a third wild-type line serving as a control. We mounted one section from each of the three conditions (genotypes) on the same glass coverslip. Sections mounted together corresponded to the same antero-posterior levels of the hippocampus and adjacent somatosensory cortex. We ran nine of these three-section arrays to acquire single mRNA molecule quantitative data for 31 genes using on-system cyclic single-molecule RNA fluorescence in situ hybridization (smFISH) chemistry. Detected mRNA spots were automatically assigned to individual DAPI-stained nuclei to render a single-cell gene-expression matrix. Like with scRNA-seq approaches, cell-type associated genes were used for unbiased classification of single cells into several distinct neuronal and glial cell types. Then, the spatial information was used to further dissect each cell type into distinct anatomical compartments and explore spatial relationships between cells. The expression levels of disease-associated genes were compared across experimental conditions, for each cell type and anatomical structure individually. The combination of single cell gene expression and spatial data allows for fine dissection of tissue heterogenicity, while the balanced tissue-array design minimizes technical variability and enables powerful statistical comparisons