Understanding the Changes in Mitochondrial Morphology through Dynamic and Three-dimensional Fluorescence Micrographs

The aim of our research is to observe changes in mitochondrial networks and investigate how these mitochondrial networks change in response to cellular conditions. We use open source tools like Fiji and Python, combining existing libraries with custom macros and scripts to automate the mitochondrial morphology analysis from complex fluorescence microscopy data. Achieving reliable quantitative data and metrics that describe the dynamics of mitochondrial fission and fusion with their localization in 3D remains a challenge.

Standardization for such data generation is not commonly available. Therefore, limited knowledge exists on cell specific fission and fusion frequency and intracellular localization. We added life detection of mitochondrial fission and fusion to the available research metrics.

And by combining these with mitochondrial structure count, we were able to define a new metric for understanding mitochondrial network dynamics. Our findings allow us to characterize cell specific fission and fusion parameters and for the first time determine whether the mitochondrial system is in equilibrium or shifting. This prevents major misinterpretation of phenotypes in health and disease and provides a clear framework for accurate reporting.

To begin, open the raw file in ImageJ. Adjust the color settings to enhance the visibility of the region of interest but do not set anything. Duplicate the image according to the number of single cells that need to be analyzed.

If multiple cells are present in a field of view, navigate to Analyze, Tools, and use the Synchronized Windows tool and the Freehand Drawing tool to draw a region of interest around a cell of interest, then select Edit and choose Clear Outside to isolate the selected cell. Split the red and blue channels from each other and save the mitochondrial channel as a tif file. To generate a point spread function or PSF, reopen the raw image.

Then open the PSF generator plugin by selecting Plugins, then choosing PSF Generator and selecting the Born wolf 3D Optical Model. Open the image information of the raw image by selecting Image, then choosing Show Information or by pressing I on the keyboard. Scroll to the bottom of the image information window.

Select the voxel size and depth option and change the wavelength to 568 nanometers. Set the Pixelsize XY to 166.1 nanometers, Z-step to 200 nanometers. Set the Size XYZ to match an image resolution of 512 x 512, and configure the Z-stack to contain 10 Z-slices.

Click on Run. Save the PSF as a tif file in its own folder. Navigate to Plugins, select Macros, then choose Edit, followed by Deconvolution_time_lapse_mine.

ijm to access the deconvolution macro. Edit the input and output lines as required and press Run to execute the macro. For image contrast enhancement and blurring, navigate to Plugins, select Macros, choose Edit and open Pre-processing.

ijm to access the pre-processing macro. Perform background subtraction by setting the rolling ball radius to 6. Set the Sigma Filter Plus such that the radius is set to 1 times the scale factor.

The number of pixels used is 2, and the minimum pixel fraction is 0.2, ensuring the plugin is set to be outlier aware. Adjust the CLAHE settings by configuring the block size to 64, histogram bins to 256, maximum slope to 2.5, and gamma to 0.8, then click Run. Open a file of interest that has been modified using the Pre-process.

ijm macros in ImageJ. Navigate to Plugins and select Adaptive Threshold. Set the local threshold to weighted mean and adjust the pixel block size as required.

Click on Preview and adjust the block size to clearly include as many mitochondria as possible. Modify the subtract value for each cell to eliminate the unnecessary background, and take note of the resulting micrograph. To optimize for time, sort images into files according to the applied subtract value.

Now navigate to Plugins, select Macros, choose Edit and open Threshold. ijm to access the thresholding macro. Edit the macro script to define the correct input and output paths, block size and subtract values.

Click Run to execute the macro. Open up to 10 thresholded micrographs that belong to the same treatment condition. Navigate to Image, Stacks, Tools and select Concatenate, then press OK.To remove residual small puncta left behind by thresholding, go to Plugins, followed by Integral Image Filters, and then select Remove Outliers.

Use preview to fine tune the X and Y sizes to eliminate fragments. Save the concatenated file as a TIF file. Finally, navigate to Plugins, Macros, Edit, QuickTest_new.ijm.

Edit the input and output path lines to point to the appropriate directories, click Run and visualize the mitochondrial event localizer or MEL results. Mitochondrial dynamics were tracked over time with red puncta marking fission events and green puncta marking fusion events in both 3D and 2D views. Mitochondrial networks showed treatment specific differences in structure over time with more elongated and interconnected forms in metformin treated cells and highly fragmented networks in metformin CCCP Baf treated cells.

The metformin CCCP Baf group showed significantly higher fission and fusion activity than the control or metformin only groups, suggesting increased mitochondrial remodeling. This group also had a significantly higher mitochondrial count, consistent with enhanced fragmentation. Mitochondrial volume was significantly reduced in the same group, further supporting a shift toward fission.

However, when normalized to mitochondrial number, the metformin only group exhibited the highest relative dynamic activity, suggesting that metformin alone promotes a more active and efficient remodeling network, while co-treatment drives extensive but less efficient structural turnover.