Quantitative Cell Biology of Neurodegeneration in Drosophila Through Unbiased Analysis of Fluorescently Tagged Proteins Using ImageJ

We developed this method to extract quantitative data from fluorescence-based imaging studies to extract complex and dynamic cell biological processes in Drosophila models of neurodegeneration. The main advantage of this technique is its an adaptable, semi-automated, image analysis workflow for minimizing selection bias, maximizing sampling power, and achieving reproducibility across the field. This approach can be extended for the analysis of other proteinaceous puncta and membrane-bound compartments implicated in neurodegeneration that ultimately will enhance our mechanistic understanding of neurodegenerative diseases.

To keep the lamina intact during the dissection under a stereo microscope slide two forceps under the retina and gently tear through the middle of the eye to remove the retina. Keeping the forceps parallel to the lamina surface pull away any remaining retinal tissue attached to the lamina. Dissect three to five brains for each group and transfer the brains to a microcentrifuge tube containing an appropriate fixative for 15 minutes at room temperature with gentle rocking.

After three 15-minute washes with PBS plus Triton X-100, or PBTx, replace the last wash with the primary antibody of interest diluted in PBTx with 5%normal goat serum for an overnight incubation in an aliquot mixer at four degrees Celsius. The next morning remove the excess antibody with three 15-minute washes in PBTx and incubate the brains in an appropriate secondary antibody diluted in PBTx with 5%normal goat serum for one hour at room temperature. At the end of the incubation wash the brains three times as demonstrated and place a drop of mounting medium between two pieces of clear tape in the center of one microscopy slide per tissue sample.

Place the brains in the mounting media for one to two minutes. When the tissues become transparent use a pipette to carefully remove the mounting medium from each side. To image the stained tissues apply a drop of fresh mounting medium onto the brains and carefully overlay the samples with the a cover glass.

Seal the cover glass edges with rubber cement and air dry the slides for 20 minutes at room temperature. Next view the specimen on the microscope and adjust the image detector controls on the fluorescence microscope to capture the highest signal-to-noise ratio in the highest a dynamic range of the specimen. Once the settings have been optimized view a specimen from another group to confirm the settings will be appropriate across the experiment.

The first specimen can then be imaged taking care to save the image as a file type that can preserve metadata such as acquisition parameters and spatial calibration. To analyze the images open an image in Fiji select View Stack with Hyperstack from the Bio-formats Import Options dialog box and set the color mode to Grayscale. Identify an area based on the marker channels that can be used as a standardized region of interest across specimens using the C scroll bar to view the channels and the Z scroll bar to move through the focal planes.

To simplify the image analysis generate a new image containing the channels and focal planes and select Image and Duplicate. Set the desired channels and slices and change the file name to reflect the selection. Manually select a standardized region of interest using a Selection Tool based on the previously determined selection criteria.

Click Analyze, Tools, Region of Interest Manager, and Add to add the region of interest to the Region of Interest Manager. Then click More and Save and name the file to reflect the region of interest. For pre processing the images click Process in Filters and select an appropriate filter.

Here select filter settings that reduce noise while preserving edges in order to aid in detection of the structures of interest during segmentation. With the pre processed image active in Fiji click SCF, Labeling, and Interactive H_Watershed. From the control panel adjust the Seed dynamics, Intensity threshold, and Peak flooding to optimize the segmentation.

When the segmentation results have been optimized select Export Regions Mask and Export to generate a binary image of the watershed results. Open the saved standardized region of interest. From the Region of Interest Manager select the region of interest identifier to limit the feature extraction to the standardized region of interest in the image.

With the region of interest active on the binary watershed mask select Analyze, Analyze Particles, and Add to Manager to extract the features to add a particle identifier for each structure delineated during the segmentation to the Region of Interest Manager. Then save the added particles taking care that the identifiers are deselected and that all of the particles are saved in a zip file. Now open the pre processed raw image and scroll to the channel used for capturing the structures of interest.

Open the particles obtained from the feature extraction and select Show All in the Region of Interest Manager to overlay an outline of each particle onto the image. Click Analyze and Set Measurements and set the appropriate experimental measurements. Then from the Region of Interest Manager select Measure and copy and paste the results into an appropriate spreadsheet program for compilation and further calculations.

Quantification of the number of semi-automatically segmented RFP huntingtin aggregates from standardized focal planes through the optic lobe of Drosophila reveals a diffuse expression of the non-pathogenic Huntington Q15 expansion in the accumulation of protein aggregates by the disease-causing Huntington Q138 expansion. Size and intensity profiles of all the aggregates from a standardized focal plane of five different optic lobes expressing Huntington Q138 illustrate the robustness and reproducibility of this workflow. When aggregates are over-exposed during image acquisition quantification of the size and number of aggregates is comparable to aggregates captured with ideal exposure.

However the intensity of the over-exposed aggregates becomes a function of size as the saturated pixels exhibit the maximum intensity value. Visualization of mCherry and GFP intensity across several autophagy-related or Atg8 positive structures reveals differences in the reporter where the high intensity of both fluorophores indicates autophagosomes. High mCherry and low GFP indicates fusion with lysosome as GFP is quenched in this acidic compartment.

And finally low mCherry reflects degradation within the lysosome. A scatterplot generated from the mean fluorophore intensity of each semi-automatically segmented mCherry Atg8 positive particle illustrates flux through the autophagy lysosome pathway that can be separated into discrete steps by quadrant analysis. While attempting this procedure it’s important to remember that the user-defined parameters in image analysis steps are highly application dependent.

So be careful to document the workflow to ensure consistency and comparability across specimens.