Cell Proliferation Assay

Description

Pulse-chase experiments using 5-bromo-2'-deoxyuridine (BrdU), or the more recent EdU (5-etynil-2'-deoxyuridine), enable the identification of cells going through S phase. Furthermore, these DNA synthesis-based methods can be combined with the detection of proliferation-specific proteins to estimate the percentage of cells in other cell cycle phases, thus obtaining a more detailed analysis of the culture proliferation. One of the most commonly used markers is Ki67, which is present within the nucleus of cycling cells during G1, S, G2, and M phases, but not during quiescence (G0). Additionally, phosphohistone 3 (PHH3) can be used to identify those cells that are specifically undergoing mitosis (M phase). By combining a short nucleoside analogue pulse (S phase) with immunocytochemical detection of Ki67 (cycling cells) and PHH3 (M phase), the entire range of cell cycle phases in the sample can be determined. Alternatively, nucleoside pulse-chase may be combined with the detection of other nuclear markers, e.g., antigens associated to specific subpopulations present in the culture, which would allow to estimate the proliferation (S phase) rate of each individual subpopulation.

Our main goal here was to develop a protocol for non-supervised, high throughput image analysis of ex vivo cell proliferation assays based on nucleoside analogue pulse alone or in combination with other nuclear markers. Our assay has been deplyed to be imaged using the high content microscope IN Cell Analyzer 2000 (GE Healthcare), so the script takes as imput datasets acquired using this and other IN Cell Analyzer versions. It consists of an ImageJ macroinstruction which can be easily added and kept to date using the Fiji distribution of ImageJ, as explained above. The workflow segments the individual nuclei and measures the signal of up to three nuclear markers. Moreover, the results table include measurements for post-processing image- and object-quality assessment. The assay must include at least two channels per field-of-view: i) on one hand, the counterstain channel to segment the nuclei; ii) on the other hand, the nucleoside analogue channel to measure the signal of each nucleus.

Please note that the (optional) illumination correction step included in the workflow requires to load a correction function (.tif image per channel). In the Usage section of this README you will find useful information to this aim.

In order to assess the output of the assay it is advisable to use a different software suited to explore high content microscopy data, such as shinyHTM.

Please note that the Cell Proliferation script is based on a publication. The original version is the Cell ProliferationHTS script (outdated). How to cite Cell Proliferation in publications:

Carrillo-Barberà P., Morante-Redolat J.M., Pertusa J.F. (2019) "Cell Proliferation High-Content Screening on Adherent Cell Cultures". In: Rebollo E., Bosch M. (eds) Computer Optimized Microscopy. Methods in Molecular Biology, vol 2040. Humana, New York, NY. DOI: https://doi.org/10.1007/978-1-4939-9686-5_14

Requirements

Fiji
Image dataset following an IN Cell Analyzer file naming convention (note that the NeuroMol update site includes a macroinstruction to turn data acquired with diferent high content microscopes into an IN Cell Analyzer file naming convention dataset)

In order to use a StarDist project, add the following upate sites to your Fiji:

CSBDeep
StarDist

In order to use the Illumination Correction macro, add the following upate sites to your Fiji:

BaSiC

Installation

Start FIJI
Start the ImageJ Updater (Help > Update...)
Click on Manage update sites
Click on Add update site
A new blank row is to be created at the bottom of the update sites list
Type NeuroMol Lab in the Name column
Type https://sites.imagej.net/NeuroMol-Lab/ in the URL column
Close the update sites window
Apply changes
Restart FIJI
Check if NeuroMol Lab appears now in the Plugins dropdown menu (note that it will be hidden at the bottom of the plugin list)

Test Dataset

Download an example image dataset. Please note that the dataset also includes a subfolder containing correction functions, for the illumination correction of each channel, and a pre-established set of parameters that can be loaded into the macro.

Brief description of the dataset:

The example dataset consists in a typical cell proliferation and apoptosis assay. The dataset was generated using methods widely used in fields such as cancer drug discovery: i) EdU (5-etynil-20-deoxyuridine) pulse-chase to label the genomic DNA of cells undergoing S-phase and ii) caspase3 immunocytochemistry to label apoptotic cells. The dataset was acquired within 4 different channels: i) DAPI for nuclei counterstain, ii) Cy3 for EdU, iii) FITC for caspase3 and iv) brightfield.

Usage

Illumination correction (optional)

The NeuroMol Lab collection includes an Illumination Correction macro which allows to generate a corrected dataset, as wells as obtain the illumination correction function for each of the channels. The calculation of the correction functions of a large dataset may require long computation time and corrected images will be saved as a new dataset, doubling the experiment size. For these reasons, it is advisable to just obtain the correction functions and apply them in the Cell Proliferation workflow. To generate the correction function of an image dataset:

Run the Illumination Correction macro (Plugins > NeuroMol-Lab > other macros > Illumination Correction)
Select the Generate correction function mode
Select the directory containing the image dataset (.tif files)
Ok
Correction functions (.tif) will be saved in a new folder created in the parent of the selected directory.

Pre-analysis mode

Run the Cell Proliferation macro (Plugins > NeuroMol-Lab > Cell Proliferation > Cell proliferation)
Select the directory containing the images (.tif files)
Select the type of Project to be applied. Filtering and StarDist are different template workflows for segmentation. Filtering is a faster filter-based approach, although it requires more parameters to be set and is more prone to merge and split objects. StarDist is a deep-learning approach which uses the Versatile (fluorescent nuclei) pre-trained model of this Fiji plugin. StarDist is slower but can perform a much more accurate segmentation if the dataset is reasonably similar to the pre-trained one (object size may be crucial). It is also possible to Load a pre-stablished set of parameters
Check Load function to perform illumination correction based on correction functions
Note that Save ROIs only works within the Analysis mode
Ok
If Load function is checked, a window will prompt to select the folder containing the correction functions
If the Load option (Project) is checked, a window will prompt to load the corresponding file
Adjust the parameters. Know more about the parameters of the workflow on the wiki page (not yet)
Ok
Select the wells to be pre-analysed
Ok
Select the number of random images that you want to test per well (up to 10 if the number of fields-of-view is greater than that number)
Select a feature to classify the objects for visualization (e.g., area, mean gray value, integrated density, circularity, aspect ratio, solidity...). Please note that this parameter (and the two below) will not affect the segmentation, as it is only used to label the obtained objects for visual inspection (see Figure 1)
Select the threshold that will be used to classify the objects according to the selected feature. Segmented objects above and under the selected value will be differently outlined in the pre-analysis output visualization (see Figure 1)
Set the line width of the segmentation outline
Ok
Once the pre-analysis is finished, a stack containing the images for visualization will pop-up

Figure 1. Pre-analysis mode output. The macro generates a stack. Each image shows the merge of the counterstain (blue) and the nucleoside analogue (red) channels. Additionally, outlines represent the segmentation output of nuclei with different colours, depending on the classification output. Objects with feature values less than or equal to the established threshold are outlined in cyan. Conversely, objects with feature values greater the established threshold are outlined in orange. A) Visualization of the mean gray value split at 250 (a.u.). B) Visualization of the solidity split at 0.9. a.u.: arbitrary unit.

Analysis mode

Run the Cell Proliferation macro (Plugins > NeuroMol-Lab > Cell Proliferation > Cell proliferation)
Select the directory containing the images (.tif files)
Select the type of Project to be applied. Filtering and StarDist are different template workflows for segmentation. Filtering is a faster filter-based approach, although it requires more parameters to be set and is more prone to merge and split objects. StarDist is a deep-learning approach which uses the Versatile (fluorescent nuclei) pre-trained model of this Fiji plugin. StarDist is slower but can perform a much more accurate segmentation if the dataset is reasonably similar to the pre-trained one (object size may be crucial). It is also possible to Load a pre-stablished set of parameters
Check Load function to perform illumination correction based on correction functions
Check Save ROIs to save the nuclei ROIs
Ok
If Load function is checked, a window will prompt to select the folder containing the correction functions
If the Load option (Project) is checked, a window will prompt to load the corresponding file
Adjust the parameters. Know more about the parameters of the workflow on the wiki page (not yet)
Ok
Select the wells to be analysed
Ok
A series of new files will be saved within the selected directory: a parameters set file (.txt), a results table file (.csv), a quality control (QC) metrics file (.csv) and, if checked, the ROI files (.zip)

Contributors

Pau Carrillo-Barberà

License

Cell Proliferation in licensed under MIT

paucabar / cell_proliferation_assay Goto Github PK