Background

Deciphering how subcellular structures change in space and time is often key to understanding biological functions (Kholodenko et al., 2010). Thus, developing approaches to visualize, track, and measure these spatiotemporal changes has been a critical area of research in cell biology. In the past decades, advances in fluorescent protein engineering (Lippincott-Schwartz and Patterson, 2003; Giepmans et al., 2006) and optical microscopy (Tanaami et al., 2002; Stephens and Allan, 2003; Hell, 2007; Chen et al., 2014; Oreopoulos et al., 2014) have allowed scientists to observe subcellular structures with unprecedented spatiotemporal resolution. Post-imaging processing and analysis are then applied to quantify and interpret the data. For example, a popular Fiji plugin, TrackMate, can be used to visualize and analyze the motion of objects (Tinevez et al., 2017; Ershov et al., 2021), including the tracking of hundreds of centrosomes in the same spectral channel simultaneously (Aydogan et al., 2018, 2020, 2022; Alvarez-Rodrigo et al., 2019). CellProfiler (Lamprecht et al., 2007; Stirling et al., 2021), another open-source software tool, can be used to quantify data from biological images, particularly in a modular and high-throughput manner. However, TrackMate cannot track on multiple spectral channels simultaneously and is not optimal for tracking objects that are not perfectly circular. CellProfiler, while versatile and packed with extensive features, is not usually used for analyzing subcellular structures. Imaris (Bitplane, Belfast, UK) is a popular software capable of tracking objects in three dimensions with a user-friendly interface, but this and other commercially available image processing software (e.g., NIS Elements by Nikon, Tokyo, Japan) are not free to users, especially for the advanced features such as particle tracking. To overcome these limitations, we have developed TRACES (Tracking of Active Cellular Structures), a customizable and semi-automated quantification pipeline capable of simultaneously detecting, tracking, and quantifying up to three fluorescently labeled object types at single-cell resolution. TRACES is built on open-source platforms and freely accessible to users. The use of intensity- and size-based thresholding to segment objects in TRACES also makes it possible to identify and track objects that are not circular (an advantage over TrackMate). In addition, as the tracking and quantification algorithm of TRACES is written in Python and implemented using Jupyter Notebook, it can be easily modified or improved to fit users’ needs. While several available tools are optimized for tracking entire cells, TRACES is developed specifically for tracking subcellular structures at single-cell resolution, including tracking micron-sized objects, such as the centrosome. In this protocol, we use the analysis of condensation of pericentrin (PCNT) proteins and the movement of the resulting “condensates” toward the centrosome (Jiang et al., 2021) as an example, hereby demonstrating the TRACES workflow. This workflow involves image acquisition and segmentation, object identification and tracking, and data quantification and visualization.

One limitation of TRACES is its inability to track dividing cells, as the current tracking algorithm assumes one nucleus object per cell. We hope to implement object tracking features for dividing cells in the future. Moreover, while the Python algorithm is capable of tracking on multiple spectral channels simultaneously, the objects of interest in each channel need to be identified manually in the initial image segmentation step in Fiji. We hope to integrate an automated workflow of image segmentation into our future pipeline.

In sum, our TRACES method is a free and semi-automated pipeline for detecting, tracking, and measuring multiple fluorescently labeled cellular structures in up to three spectral channels simultaneously at single-cell level. If these cellular objects can be segmented and distinguished from one another, their relationships and properties (e.g., distance, size) can then be quantified using the TRACES method. Therefore, we envision that TRACES is not limited to analyzing centrosomes, their related structures, and nuclei, the three objects exemplified in this protocol. TRACES can be applied to analyzing any distinct fluorescent objects in up to three spectral channels in the cell, and will thus be a useful tool for the cell biology community, facilitating quantitative image analysis in other biological contexts.