4. Tracking

Single-particle tracking (SPT)

There are a few different techniques to interrogate the dynamic organization of molecules in cells. One is fluorescence recovery after photobleaching (FRAP), based on an ensemble measurement of fluorescent molecules. Fluorescence correlation spectroscopy (FCS) is an alternative and detects molecules while they pass through the confocal volume of a microscope. The capabilities of FCS can be extended for example by varying the size of the confocal volume or performing measurements at different positions.

Conceptionally different, single-particle tracking (SPT) directly follows labeled particles while recording their movement. The resulting trajectory of a particle in its native cellular environment then allows extracting for example the diffusion characteristics or confinement status (i.e. if the particle is trapped or immobilized). A particle here can be anything that can be tracked, a bead, a virus or even a single molecule.

sptPALM is a variant of PALM imaging using photo-activatable/convertible proteins. sptPALM makes good use of the fact that activated molecules stay one for a few frames before they bleach. In a live cell environment, this time can be used to follow the molecule and obtain information about its diffusional state. The example on right shows a short PALM sequence (mEos3) that after localization could then be used for tracking. See the S/N ratio compared with STORM shown here.

The first steps of a possible analysis procedure are described below.

Connecting the dots (scripts can be found here).

1. Following image acquisition, the localization and filtering are performed as described here_

2. Open the script tracking_Crocker_Grier.mand load the csv localization file

This script allows examining the dataset, perform the tracking on a selected region and export the trajectories to interact with other software.

3. The first box shows a scatter and density plot together with the STD sum of the image stack inspect the dataset. As you can see, for a dense dataset the scatter plot is less informative, but the density plot makes it much easier to see interesting areas within the image.

4. Select a region of interest to perform the tracking. This cell renders the dataset with a large pixel size (result similar to density plot, but faster) to identify regions with potential interest.

5. Set up the tracker. The tracking algorithm used here is based on a publicly available SPT routine and can be found here. The main input parameters for the tracker are:

1. Maximum time (in frames) the molecule can be lost (dark time).
2. Minimum length of the trajectory

6. After linking a particle's location and building a trajectory, the next cell will plot the result and give a quantification of the track length, typically an import parameter to evaluate the goodness of the data and the type of post-processing.

7. The final output are two overlay figures showing the STD image with the localizations or trajectories on top. This can be a good quality check and also useful to overlay trajectories over a separate channel (not shown here).