Deep learning-based image-analysis identifies a DAT-negative subpopulation of dopaminergic neurons in the lateral Substantia nigra.

Abstract Here we present a deep learning-based image analysis platform (DLAP), tailored to autonomously quantify numbers of neuronal subtypes in defined areas, and of fluorescence signals, derived from RNAscope probes or immunohistochemistry, in defined cellular compartments. As proof-of-principle, we utilized DLAP to analyse subtypes of dopaminergic midbrain neurons in mouse and human brain-sections. These neurons modulate complex behaviour like voluntary movement, and are differentially affected in Parkinson’s and other diseases. DLAP allows the analysis of large cell numbers from different species, and facilitates the identification of small cellular subpopulations, based on differential mRNA- or protein-expression, and anatomical location. Using DLAP, we identified a small subpopulation of dopaminergic midbrain neurons (~5%), mainly located in the very lateral Substantia nigra (SN), that was immunofluorescence-negative for the plasmalemma dopamine transporter (DAT), with ~30% smaller cell-bodies, and a ~4-fold higher likelihood of calbindin-d28k co-expression. These results have important implications, as DAT is crucial for dopamine-signalling, and its expression is commonly used as marker for dopaminergic SN neurons.