Trajectories of polystyrene beads driven by a carpet of ameboid cells

Four MATLAB .mat files containing tracks of polystyrene beads driven by a monolayer carpet of Dictyostelium discoideum cells. This dataset contains data for four different bead sizes (for details see table below):

• 29 µm
• 46 µm
• 61 µm
• 100 µm

There is one file for each bead size.

Data Structure

Each of the files contains a structure `data_mult[colloid_size]` with three fields:

• `position`: xy-position of the bead in pixels (1.56 px = 1 µm)
•  `time`:  time of the position in frames (1 frame = 15 s)
•  `dT`: time between frames (15s)

A Jupyter notebook is provided for loading the data in Python.

Experimental Setup

This description is mostly identical to the supplementary information of https://doi.org/10.48550/arXiv.2508.12976 .

Cell Culture

Dictyostelium discoideum AX2-214 wild-type cells were cultured in untreated polystyrene cell culture flasks (T-75 Standard, Sarstedt AG & Co. KG, Nümbrecht, Germany) in 10ml  HL5 medium including glucose (Formedium Ltd. Norfolk House, Norfolk, England), supplemented with 100 units/ml Penicillin G soidum salt and 0.1mg/ml Streptomycin sulfate (Penicillin-Streptomycin 10X Solution, Biological Industries, Beit Haemek, Israel) at 21°C. The growth medium was exchanged every two to three days. The cells were diluted regularly to avoid confluency. The cell culture was renewed every 4 weeks at the latest. 

In preparation for the experiments, the cells were harvested and suspended in fresh nutrient medium. The cell density was adjusted to 10^6 cells/ml. The experiment was performed using a chambered coverslip with individual wells and an imprinted 500 µm location grid (ibidi GmbH, Gräfelfing, Germany) with 1.0 cm^2 of growth area per well. 298 µl of the prepared cell suspension was transferred to each of the wells → approx. 30 cells per 100 µm × 100 µm. The cells were given 20 min to sediment and attach to the bottom of the wells.

Beads

As cargo, monodispersed spherical polystyrene beads were used (microParticles GmbH, Berlin, Germany). 20 min after the cells had been introduced into the wells, 2 µl of bead suspension (for concentration see table below) were added in each well just below the surface of the nutrient medium. The slide was slightly tilted to distribute the particles as evenly as possible. The measurements were started between 20 - 60 minutes after adding the polystyrene beads.

The experiments were carried out with 4 different bead sizes:

Imaging

The images were aquired with 

• OLYMPUS XM10 CCD camera (OLYMPUS XM10, resolution: 1376 × 1032 pixels, 8 bit grey scale)
• OLYMPUS IX71 bright field microscope (Olympus Deutschland GmbH, Hamburg, Germany), 10x objective (Olympus UPLanFL N 10x/0.30na).
• Exposure time: 10 ms
• Frame interval: 15 s
• Total duration: 4 hours

Using a computer controlled motorized stage (Märzhäuser Wetzlar GmbH & Co. KG, Wetzlar, Germany), several different positions were recorded during a single run of the experiment.
The focal plane was adjusted such that polystyrene particles appeared as black circles with a bright spot in the middle. 

Data processing

The images captured by the microscopy setup were processed using a custom MATLAB code.

1. Stage correction: Using the location grid of the wells as fixed points, a stage correction was performed to remove any jitter from small inaccuracies of the stage placements. 
2. Cropping: The images were cropped to exclude the grid since the groves of the grid might influence the cell behavior.
3. Bead Tracking:The positions of the bead centroids were tracked and combined into bead trajectories. Touching beads were not tracked; a collision of two beads ended each of their tracked trajectories. A circularity criterion was defined that excluded beads with insufficient area within the frame. 

Camera or stage errors could lead to individual positions missing from the dataset. Such an error occurred about once every 1000 images. Similarly, there are missing positions in the trajectory of a bead which temporarily (<= 3 frames) violated the circularity criterion due to its proximity to a frame boundary.

The number of trajectories N for each of the datasets is given in the following table: