Background: Study of mitochondrial motility has gained increasing interest in recent years and live cell imaging is crucial to better understand the underlying physiological factors. Different strategies have been employed to track mitochondrial movement, particularly to investigate mitochondrial transport in neuronal axons.

Materials and Methods: This work used an open source automated MATLAB model, developed by Judith Kandel, Philip Chou, and David M. Eckmann, describing mitochondrial motility as a lognormal distribution which provides a quantitative paradigm to assess mitochondrial movement. Differentiated SH‐SY5Y cells were treated with 31.25 nM rotenone and 6.25 μM 6‐hydroxydopamine for 24 h. Mitochondria were then labelled with 25 nM Mitotracker Red and ten‐minute movies of one frame per second were obtained under a Nikon Ti‐E H‐TIRF microscope. Then, the videos were preprocessed in ImageJ as time‐lapse images. First, they were convolved, then converted to the frequency domain using a Fast Fourier Transform (FFT) and subjected to a bandpass filter ranging from 2 to 100 pixels. Finally, the resulting images were manually thresholded to best eliminate the noise. The resulting stacks of images were analysed by the selected model focusing cell projections. After processing and comparing each frame, the model output two histograms of log values of net, the total distances traveled by all mitochondria objects found and the respectively, trajectories. In addition, the TrajPy python package was used to further quantify the trajectories by estimating a set of features: the mean velocity, anisotropy, straightness, efficiency, among others. These features helped to highlight relevant differences between the control and the treated samples.

Results: We found that mitochondria in treated cells appear to travel less in average compared to controls which indicates mitochondrial dynamics is affected by the treatments. This decrease in mobility is more accentuated in rotenone‐treated cells and came along with an increase in trajectories anisotropy and straightness. Strikingly, the treated mitochondria presented more regular movement when compared to the control, which contrasted with their smaller range of velocities.

Conclusions: Moreover, we confirmed that the combination of the selected open source models provides an effective method to quantify mitochondrial motility, thus allowing its application in future studies using different cell treatments and conditions.