Measurements of mitochondrial morphological properties, velocities and intensities in populations or individual mitochondria

This protocol describes how to record and analyze fluorescence images to determine shapes of (visually) individual mitochondria. Generic guidelines are given for setting up confocal or wide-field microscopy for image acquisition and a step-by-step approach to analyze recorded images. Segmented images can be used to measure morphological parameters, such as length, width and roundness, velocities, distances and how these properties change in time. Segmented images can be also used for determining fluorescence intensities over individual mitochondria measured in other fluorescence channels. Results can be graphed to show how do mean properties of a population of mitochondria the view field or bounded by ROIs change in time, or using tracking how properties of individual mitochondria change. Spreadsheet output includes morphological parameters for each mitochondrion in each frame or select parameters as population means or for individual mitochondria.

A simpler protocol using single built-in pipelines is available: Measurements of morphological properties of mitochondria using built-in pipelines.

Equipment

Microscopy requirement:

The assay works with wide-field (epifluorescence),  and confocal microscopy.  The epifluorescence microscope has to be capable recording of z-stacks and low-light level time lapse imaging, e.g. equipped with a fast shutter and high quantum efficiency cooled monochromatic camera. Confocal microscopes trivially work in low-light level mode. Recording of z-volumes is recommended, but the analysis on 2D projection images provides sufficiently reliable results.

Microscopy guidelines:

• Image cell cultures in coverglass bottomed dishes.

• If looking mitochondrial dynamics in mammalian cells, thermostat the sample to 37C using heated environment chamber or heated stage and lens.

• Use the highest available NA (numerical aperture) immersion lens.

• Record sparse z-stacks (1.5-2 µm spacing) in order to reduce photo damage. This spacing will be sufficient to capture all mitochondria in a cell and work with projection images.

• Use at least 0.2 µm/pixel resolution or higher (slight oversampling will help image segmentation)

• Never saturate the image (saturated mitochondria will have a hole after high pass filtering and segmentation)

• Record frames for a time lapse not more often than the biology dictates, excessive number of frames will cause photodamage

• Save recordings in the native file format of the microscope, if compatible with Image Analyst MKII (e.g. czi, lsm, nd2, nd, lif, ... see a list here)

Image analysis in Image Analyst MKII

Tutorial data set: Mitochondrial motion assay / Metamorph (use position 7 only this is included to this zip file!)

A) Segmentation of mitochondrial images

1. Open the recording in Image Analyst MKII using the main toolbar button or File main menu. Alternatively use drag and drop. If loading lsm files, select Zeiss Multi Time Lapse as file type to provide control over projection.
2. In the Multi-Dimensional Open dialog:
   1. In the Open tab select the stage position and channel to be opened.
   2. If the recording has a z-dimension, in the Settings tab select the way z-projection. For this analysis mean or maximum projection can be used.
   3. Note: this method does not use filtering before the z-projection during opening, therefore do not turn spatial filtering on.
   4. Press the Open button.
3. To invoke the pipeline go to the Pipelines main menu and select: Morphological Measurements / Segment mitochondria.
   1. Largest mitochondrion size (width in pixels): Give an approximate width of the largest mitochondria. This value controls the high pass filter to suppress larger than mitochondrial details. It does not act as a classifier for mitochondria, so slightly larger objects will also appear in the image. Decrease this value if smaller mitochondria are lost during segmentation. Increase this value if noise passes through from background. It's value for the tutorial data set is 5.
   2. Sensitivity (top range scaling, percentile): A percentile value, typically between 95-99.99 percentile. Decrease this value to increase sensitivity, if dimmer mitochondria are lost during analysis, or if bright debris is present in the image. Increase this value if noise passes through from background.  It's value for the tutorial data set is 99.5.
   3. Minimum size (area, pixels): Minimum size of mitochondria to detect. This is an object classifier, smaller objects than the specified size will be removed. Decrease this value if smaller mitochondria are lost during the analysis. Increase this value if noise passes through from background.  It's value for the tutorial data set is 5.
4. To process a loaded recording either right-click the image and select "Process this with Segment Mitochondria" or press the  or main toolbar buttons.
5. If needed, adjust the analysis considering point #3 above.
   1. To re-run the analysis use the button in the main toolbar or on the bottom of the Multi-Dimensional Open dialog and select "Clear and Run Pipeline...". This will perform both loading and processing.
6. Open an Excel Data Window using the Tools main menu if tabular data will be saved below.

B) Optional: tracking individual mitochondria and velocity measurements

To measure how morphological (C see below) or intensity properties (D see below)  vary in time over a visually discernible, single mitochondrion, perform tracking after segmentation and then perform measurements. This also allows measuring velocities of individual mitochondria.

1. Duplicate the mitochondrial image before segmentation using the main toolbar button (select "Linked").
2. Segment one of the mitochondrial images as given in section A).
3. Subtract background in in the other image.
4. Using the Tools/Rename dialog change the channel number of the other mitochondrial image to a unique number among the opened channels.
5.  Select the Motion Measurements / Track segments with one intensity channel pipeline and set the parameters as follows:
   • Channel Number of segmented image to track: Set the channel number of the segmented image. Use the Tools/Rename dialog to see it.
   • Channel Number of intensities (helps matching objects):  Set the channel number assigned above to the raw image
   • Maximal distance that a cell travels between consecutive frames (pixels): Increase this distance if fast moving mitochondria are not tracked. Decrease this number if close by segments often swap "color" in tracked images.
   • Minimum track length (frames): Removes segments that are only present in the given number of frames. This filters out objects that weren't successfully tracked.
6. Do not select multiple inputs as Image A. Pipelines take only a single input and seek for other channels based on the above set channel number.
7. Right-click one of the images and select "Process this with ..." or press the  or main toolbar buttons.
8. Proceed to morphological or intensity measurements as above described. For intensity measurements set Time Continuous Segments to Yes.
9. Alternatively measure velocities using the Plotting/Plot Tracking Parameters.
   1. Plot type: this parameter is not applicable here, always velocities of individual mitochondria are returned.
   2. Mean of all traces:
      • Select No to plot data for each mitochondrion. Note: that individual mitochondria in consecutive frames with identical serial number do not correspond to each other, unless the segmented images is tracked first.
      • Select Yes to calculate mean for the whole view field or the constraining ROI. Note: this can be also done in the plot window using the context menu.
   3. Set the normalization and mean calculation parameters (ΔF/F0, Mean of all traces, Advanced normalization, Range markers for means or rates, Calculate rates at markers) as described for this function. The default values are appropriate for this analysis.
   4. Set the data output properties (Y-values only, Place channels into columns, Range of channel numbers to lay out, Plate worksheet output, Place positions into columns, Expect only a single channel) as described for this function. The default values are appropriate for this analysis.
   5. Constrain to ROIs: set this to yes, if constraining the analysis to ROIs.  To this end draw one or more area-type ROI: .
10. Use the File/Save Excel Data main menu item to save the velocity data (the Excel Data Window must be open before analysis using the Tools main menu).

C) Measurement of morphological properties of mitochondria or the distance from a reference ROI

To measure a time course of a mean morphological parameter for a view field or for objects encircled by ROIs:

1. Segment the mitochondrial image as given in section A).
2. If you want to constrain the analysis to ROIs, draw one or more area-type ROI:
3. Select the Plotting/Plot Morphological Parameters of Segments function, and set its parameters as follows:
   1. Morphological parameter: select the morphological parameter of interest: "number", "area", "perimeter", "diameter", "filament length", "branch points", "shape factor", "fiber length", "fiber breadth", "distance from ROI". Note: to measure distance from ROI, draw an ROI before running the pipeline.
   2. Plot type:
      • Select Mean to measure mean parameters in the whole view field or within each ROI.
      • Select Each to record the chosen morphological property for each mitochondrion. Note: that individual mitochondria in consecutive frames with identical serial number do not correspond to each other, unless the segmented image is tracked first (see above, B).
      • Or select other statistical parameters: Sum, Variance, Variance/Mean, SD/Mean to measure how the chosen parameter varies in the population.
   3. Set the normalization and mean calculation parameters (ΔF/F0, Mean of all traces, Advanced normalization, Range markers for means or rates, Calculate rates at markers) as described for this function. The default values are appropriate for this analysis.
   4. Set the data output properties (Y-values only, Place channels into columns, Range of channel numbers to lay out, Plate worksheet output, Place positions into columns, Expect only a single channel) as described for this function. The default values are appropriate for this analysis.
   5. Constrain to ROIs: set this to yes, if constraining the analysis to ROIs.  Note: the  "distance from ROI" measurement cannot be constrained to ROIs.
4. Optionally, select Tools/Excel Data Window to enable recording of data to spreadsheet.
5. Press the main toolbar buttons or right-click the image and select "Process this with...".
6. Use the File/Save Excel Data main menu item to save the velocity data (the Excel Data Window must be open before analysis using the Tools main menu).

D) Measurement of fluorescence intensities of mitochondria in other channels

To measure fluorescence intensities associated with each mitochondrial segments in other channels of the recording:

1. In the Multi-Dimensional Open dialog select and open the channel to be segmented and channels where the fluorescence intensity will be measured.
2. Segment the mitochondrial image as given in section A).
3. Subtract background in the intensity images.
4. If you want to constrain the analysis to ROIs, draw one or more area-type ROI:
5. Select the Plotting/Plot Intensities Corresponding to Segments  function, and set the inputs as follows:
   1. Image A: this is the above segmented image. Note: use the center part of the property bar for this, or the context menu over the Image Windows.
   2. Image B: this is (or these are) the intensity image(s) for readout
6. Then set the parameters as follows:
   1. Plot type:
      • Select Mean to measure mean intensity of all pixels in overlap with a segment.
      • Or select other statistical parameters: Sum, Variance, Variance/Mean, SD/Mean to measure how intensity varies within pixels of a mitochondrion.
   2. Mean of all traces:
      • Select No to plot data for each mitochondrion. Note: that individual mitochondria in consecutive frames with identical serial number do not correspond to each other, unless the segmented images is tracked first (see above, B).
      • Select Yes to calculate mean for the whole view field or the constraining ROI.  Note: this can be also done in the plot window using the context menu.
   3. Set the normalization and mean calculation parameters (ΔF/F0 , Advanced normalization, Range markers for means or rates, Calculate rates at markers) as described for this function. The default values are appropriate for this analysis.
   4. Set the data output properties (Y-values only, Place channels into columns, Range of channel numbers to lay out, Plate worksheet output, Place positions into columns, Expect only a single channel) as described for this function. The default values are appropriate for this analysis.
   5. Constrain to ROIs: set this to yes, if constraining the analysis to ROIs.  The  "distance from ROI" measurement cannot be constrained to ROIs.
7. Optionally, select Tools/Excel Data Window to enable recording of data to spreadsheet.
8. Press the main toolbar button.
9. Use the File/Save Excel Data main menu item to save the velocity data (the Excel Data Window must be open before analysis using the Tools main menu).

Protocol by Akos A. Gerencser 03/22/2016 V1.0        

These or similar approaches were described used in the following papers:

1. Gerencser A. A. and Nicholls D. G. (2008) Measurement of Instantaneous Velocity Vectors of Organelle Transport: Mitochondrial Transport and Bioenergetics in Hippocampal Neurons. Biophys J. 2008 Sep 15;95(6):3079-99.

2. Choi S.W., Gerencser A.A., Nicholls D.G. (2009) Bioenergetic analysis of isolated cerebrocortical nerve terminals on a microgram scale: spare respiratory capacity and stochastic mitochondrial failure. J Neurochem. 2009 May;109(4):1179-91

3. Choi SW, Gerencser AA, Lee DW, Rajagopalan S, Nicholls DG, Andersen JK & Brand MD. Intrinsic bioenergetic properties and stress-sensitivity of dopaminergic synaptosomes. J. Neurosci. 2011 Mar 23;31(12):4524-34