Applying multinomial Bayesian modeling to vesicle trafficking

Huntington's disease (HD) is a hereditary neurodegenerative disorder caused by an expanded poly-glutamine stretch in the Huntingtin protein (Htt). Both HD patients and animal models show alterations in the regulated-secretory pathway, evidenced by altered neuropeptide functions and decreased dense-core vesicles (DCV) secretion, respectively. However, Htt´s role in this pathway, and how it becomes affected in HD, remain unclear. We used confocal imaging and bioimage analysis to characterize DCV motility in chromaffin cells. We applied Bayesian statistics to model the probability of each motion type as a function of the experimental conditions and cell regions, using a multinomial logistic regression and fitting different spatial distribution patterns. Under resting conditions, WT-Htt increased DCV motility in the periphery, whereas mHtt induced the same effect but in the central region of the cell. During K+ stimulation, Control cells revealed an increased DCV number near the periphery, with increased DCV motility toward the plasma membrane throughout the cell. The expression of mHtt impaired this last response, reducing the general motility of DCV, and particularly in the direction toward the plasma membrane, an effect absent in WT-Htt-expressing cells. This data contributes to our understanding of Htt biology, showing the importance of Htt activity along the DCV trafficking pathway.