Documentation for Puro-PLA analysis in neurons

Pipeline

• 1. Tiff file extraction from lif files
• 2. Mask creation using Ilastik
• 3. Mask refinement and area detection using Fiji
• 4. Puro-PLA analysis using RS-FISH
• 5. Mask filtering using RS-FISH
• 6. Generating overlay images of the detections
• 7. Collection of all datasets and plotting

This pipeline is part of a publication: Loedige, I. et al. mRNA stability and m6A are major determinants of subcellular mRNA localization in neurons. Mol. Cell (2023) doi:10.1016/j.molcel.2023.06.021.

1. Tiff file extraction from .lif files

Use the Fiji macro open_lif_split_gfp_max.ijm script to open a set of .lif files, split the channels, and resave them as single tiff files. The script also renames the files, for this, the script expects these channels in this order: 1. DAPI, 2. Puro-PLA, 3. Map2-GFP. The script also saves a max projection of the GFP channel, which may be used to create a binary mask.

2. Mask creation using Ilastik

Ilastik can be used to create a binary mask to filter the Puro-PLA detections. Since neurons are pretty flat, masks are created from max projections of the GFP channel. The Autocontex workflow uses pixel classification to segment the cell boundaries. For this, cells are segmented into two rounds. After initial training, similar images can be processed together in batch mode. To export the masks, select the Simple segmentation output format and tiff as the file format.

3. Mask refinement and area detection using Fiji

Use the creating_mask.ijm to create binary masks from the Ilastik segmentation. Using the paintbrush tool, the segmentation can be corrected to remove small background signals or neighboring cells by coloring them in black (with a value of 0). Small corrections on the masks can be performed using white (with a value of 1).

4. Puro-PLA analysis using RS-FISH

One optional step is to prepare the Puro-PLA images using a difference of Gaussian filter using the macro script DoG_filter.ijm. The sigma for the Gaussian blur needs to be adapted for each new set of images.

To detect Puro-PLA spots in the image, the Fiji plugin RS-FISH can be used (info on how to use RS-FISH and how to download the plugin can be found on the RS-FISH GitHub page). The macro RS-FISH_macro.ijm can be used to run RS-FISH in batch mode. Determine the parameters before using one representative image and then run the rest of the images in batch mode.

5. Mask filtering using RS-FISH

With the created mask, the resulting csv files can be filtered using the RS-FISH plugin "Mask filtering". Go to Plugins > RS-FISH > Tools > Mask filtering.

6. Generating overlay images of the detections

To show the RS-FISH detections on the images, one can use the visualization tools within RS-FISH using Fiji or the BigDataViewer. For 2D images, the macro Overlay_generator.ijm can be used to show detections easily. It will also add the detections to the ROI manager so they can be further manipulated in shape, size, and color.

7. Collection of all datasets and plotting

To detect the size of the created masks, run the record_size_of_mask.ijm to create CSV files for each binary mask containing an area measurement. Make sure that files are binary files with values of 0 and 1 and that all have set the same scale.

To create a dataset for plotting, this Python script searches all the files and combines them, but this might depend on the file structure.

Collect_counts_and_mask_sizes.ipynb

To plot from the created database: Plot_Puro-PLA_values.ipynb