Automated Image-Based Quantification of Neutrophil Extracellular Traps Using NETQUANT

Here we present NETQUANT, a fully automated approach to quantify NETS and immunofluorescence images. This will help researchers in making comparable image analysis from data generated by multiple users and conditions. The advantages of this technique are the ease of use, single-cell data analysis, multiple criteria to evaluate NET formation, and an unbiased analysis of multiple images.

Install and open the analysis software. An up-to-date version can be found on the Zenodo GitHub archive, or the Nordonfelt Lab website. In the setup tab, choose a source folder for analysis by clicking the get path option in the source menu.

Select the folder containing the image sequences to be analyzed. Again, click on the get path option in the target menu, and select the target folder for saving the data following image analysis. Next, when using separate channel images, name the channel folders so that the DNA channel corresponds with the nuclear DNA stain, and the NET channel depicts the neutrophil elastase stain in the images.

Also, name the folder containing the control image files as control. Next, click the load image information button and load the image metadata into the software. Then in the channel order sub-menu, select the correct channel order contained in the images.

This helps to prevent accidental mismatches. Now click on the prepare data button to acquire primary image properties from the raw data and convert the images. The converted images will then appear in the sample types sub-menu.

Click on the sample type menu, select an image from the sub-menu and click on display image data to display the images split into the DNA and Neutrophil Extracellular Trap, or NET channel, respectively. To segment the cells into their respective channels, select the segmentation method tab, and select a method in the method sub-menu for both the DNA and the NET channels. The default method of segmentation is set to adaptive and is the recommended setting.

Other options are available in the software including global edge and chanverse. A watershed option has also been included to help distinguish between closely-placed cells or NETs. Also in the segmentation tab, click on the segment control samples option.

Then select PMA from sample type sub-menu and click on the batch option to begin segmentation of all images included in the data set. Next, select the images in the sample type menu and click on the display image data button to visualize and validate the binary image masks for both the DNA and NET masks generated post-segmentation. Now, in the analysis tab, click on the determine threshold button to analyze the control samples.

Then, on the right hand side, change the sample type to PMA and click the get cell properties button to complete analysis of the stimulated samples. Next, select an image from the sample type sub-menu and click on the display image data button to display the overlay and the number of cells and NET forming cells in the image. Begin by selecting the sample from the sample type sub-menu.

Individual images can be selected from the sample type sub-menu for analysis or the entire batch of images can be analyzed by selecting the batch option. Once selected, click on the analyze NETs button to complete the analysis. Adjust NET criteria manually to yield optimal results for a given sample.

Compare identified NETs with the original images to assess the quality of identification. Once this step has been completed, the NET criteria can be applied across all images in the dataset. Any changes made to the NET criteria are also simultaneously applied to all control samples.

Inspect the data summary in the cell data sub-menu, where the number of images, cell count per image, and percentage of NETs per image are displayed. Enter the output tab to select and view the result outputs. Explore and compare the various data outputs generated from the analysis of control and PMA-treated samples by selecting the form of the output and clicking on the output results button.

Next, launch the results data table CSV file to explore single-cell data and click on the results PDF files to visualize the NET area distribution and DNA to NET ratio in the samples. The red line indicates the threshold value in the graphs. Finally, click on the results bivariate distribution file to determine the NET area versus shape of DNA.

Here, 15 images of control neutrophils and 15 images of PMA-stimulated neutrophils were analyzed in under ten minutes. The total number of cells were counted and the percentage of cells with Neutrophil Extracellular Traps was determined using the automated quantification method described in this article. These results shows that PMA-stimulated neutrophils have a larger area, an increase in nuclear deformation, and a higher DNA to NET ratio than control samples.

When combined into a 3D graph, the PMA-stimulated samples tend to show areas of tighten groupings than the control samples. While attempting the procedure, it's important to remember that although the workflow was tested using multiple donors, it is recommended that the user should decide on the software parameters appropriately based on the individual data set.