Measurment of mitochondria to cell volume ratio

Note: an updated version of this protocol is in preparation for LSM780 with Definite Focus.

Sample preparation, reagents

• Experimental buffer (EB) in mM: 120 NaCl, 3.5 KCl, 1.3 CaCl₂, 1 MgCl₂, 0.4 KH2PO4, 5 NaHCO₃, 1.2 Na₂SO₄, 20 TES, 15 glucose , pH7.4 at 37°C
• Alternatively use full culture medium and CO₂ control during microscopy.
• Calcein-AM 2mM stock in DMSO
• Mitotracker Red CMX 100mM stock in DMSO
• Load cells with calcein-AM 0.5-1 µM plus Mitotracker Red 25-50 nM for 30 min in EB. Certain cell types need higher dye concentrations, up to 2-µM calcein-AM plus 100-nM Mitotracker Red
• Replace medium over the cultures with dye-free EB. Image at 37°C or alternatively at RT to mitigate mitochondrial movement and dye leakage. If calcein excessively leaks out of the cell, add 500 mM Na-sulfipyrazone.

Image acquisition with Zeiss LSM

The aim is to record best possible quality, slightly oversampled images at evenly spaced z-coordinates from the bottom to the top of the culture. Due to strong photo-toxicity and photo-bleaching it is not possible to use z-stacking, but different x,y-coordinates of the culture are scanned for each image.

• Microscope Settings (for Zeiss LSM 510):
• Microscope and Configuration Control:
  • Lens: Plan-Apochromat 100x/1.4 Oil
  • Single track:
    • Calcein: 488->500-530 (Ch2)
    • Mitotracker Red: 543->560LP (Ch1)
• Scan Control:
  • Optical Zoom: 2x
  • Pixel size: ~0.044um (oversampled, but not that much as for deconvolution)
    Dimensions, scan: 1024x1024, Single plane, 2xLine Average, Scan Speed 5 or 6
  • Pinholes: 1 Airy
  • Gains: ~500V
  • Data depth: 12 bit
  • Laser power: Ar488: ~5-10%; HeNe543:60-100% (increase gain or laser power as advancing for higher sections, to keep uniformly bright images. Stronger Ar488 will increase the crossbleed of calcein into the red channel.) Set the laser power that a decent quality image is recorded, as the image is only once acquired, and not z-stacking is used, it is OK if the view field is bleached / photodamaged after recording.

1. Manual image acquisition:
   1. Using fast XY scan focus to the bottom of the cells, acquire one frame. Always acquire only once one specific view field, the applied laser intensity damages the cells.
   2. Acquire two more bottom plane frames in different areas.
   3. Using fast XY scan focus to the bottom of the cells, the use the stage control dialog to raise the focal plane by 1 mm.
   4. Acquire 3 different view fields at this plane. To advance between fields use arbitrary number of turns of the fine stage control knob and do not align cells to the view field, to avoid biasing the amount of nucleosol/cytosol present in the images (unless whole cells can always fit into the view field in a sparse culture).
   5. Repeat point 3 until getting to the top of the cell culture. Different cell types need different z-step size, look for the cell layer thickness, and record 5-7 planes.
   6. Laser intensities and gains may be increased by advancing in z. If image becomes noisy decrease Scan Speed to 5.
   7. Use the range indicator. DO NOT SATURATE IMAGES! (when having multiple cell types, non-interesting details can be saturated)
2. Automated image acquisition (requiring the Multi Time Lapse module of the Zeiss LSM software)
   1. In multi time series module erase all coordinates an set parameters (Edit Locations / Import From Stage Control).
   2.  (Autofocus and configuration Multi Tracks, Autofocus range 90, Offset 0 (does not matter)
   3. Store lower right corner of each well where the culture is nice.
   4. In Edit Locations set 60um x 60um and 7 x 5 positions and press Multi Grid
   5. Go to first position and switch back and forth between Fixed Location and multiple locations to copy parameters to the newly set locations.
   6. Use the Store/Apply button to Store parameters into registry under the name of “VolumeRatio”.
   7. Adjust volumeratio.xls to same length (number of wells x 35) as the number of locations and copy columns into volumeratio.reg. Save it and double click to enter it into the registry. Offset decreases upwards. The zero is too high to start, check it specifically in the actual dish.
   8. Use the Store/Apply button to Apply parameters into registry under the name of “VolumeRatio”.
   9. Set image database, file name, tweak laser intensities (and save configuration)
   10. Start Time.

Analysis in Image Analyst MKII

Use Image Analyst MKII to determine areas corresponding to mitochondria and the whole cell by using adaptive thresholding.

This protocol assumes that the user is familiar with the following sections of the online manual:

• Spatial filtering in Fourier domain
• Fixing imaging artifacts (Image Stabilizer, Channel Alignment and Spectral Unmixing)
• Using masking to constrain areas to plot or process
• Binarization, Thresholding, Segmentation
• Pipeline

Protocol

• Load the lsm files corresponding to the serial sections of one sample by sorting in Windows Explorer according to Time, multiple selecting and dropping on the Image Analyst.
• For cortical neurons pre-process as follows ("volumeratio preprocess confocal mask and unmix.ips", the steps of the pipeline are detailed below, use the to run the pipeline and proceed to the next step):
  1. Threshold: Mask above Pixel value = 4095 (Threshold from local max/min=None) both images before unmixing, if there were saturated areas.
  2. Subtract Background: at 5 percentile per frame
  3. Blind Spectral Unmixing with NMF: perform automatic unmixing. The UnmixNFM matrix have to be defined in the Preferences/Convolution kernels first as {{1,0.2},{0.01,1}} meaning that we expect mostly calcein crossbleed into the Mitotracker channel.
  4. Inprint Mask: holes created by masking are filled in with this function.
  5. When the pipeline finished running close the copied calcein image, and keep working with the unmixed Mitotracker and the non-unmixed calcein image.

"volumeratio preprocess confocal mask and unmix.ips"

• If the culture is uniform, proceed to next step. If the the mitochondrial volume ratio in a certain cell type is to be measured in a mixed culture:
  1. Work with the calcein image performing the processing steps below by hand:
  2. Subtract Background: at 5 percentile per frame
  3. Threshold: Bottom at Pixel value = 0 (Threshold from local max/min=None).
  4. For each frame of the calcein image draw ROI and perform Apply ROI mask by Outside/0/Current frame only=Yes. Mask only the calcein image.
  5. Save ROIs for later use.
  6. If previously drawn ROIs are loaded the image set can be masked by advancing frames by -> and ROIs by > and repeating the Apply ROI mask for each frame
• Process the unmixed Mitotracker and masked or original calcein image with the Pipeline “volumeratio process confocal cortical-masked.ips” (the steps of the pipeline are detailed below, use the to run the pipeline and proceed to the next step):
  1. The Calcein image is processed as follows:
  2.  Wiener Filter Smooth: adaptive filter for noise removal. Mask width=5, Noise level=0.01. Increase noise level for greater noise suppression or decrease if details are smudged.
  3. Set scaling/color and Write back scaled values: values between 1-90 percentile of each frame is mapped between 0 and 1000 for each frame, so the new intensities in each of the frames will be similar.
  4. Threshold: Mask below Pixel value = 1 (Threshold from local max/min=None): this re-masks the area originally masked when selecting for the cells of interest, so these areas will not contribute into the threshold calculation in the next step:
  5.  Threshold: Above Otsu by frames at 1 (Threshold from local max/min=None): this is an adaptive, but uniform threshold for each frame. 
  6. The Mitotracker Red image is processed as follows:
  7. Set scaling/color and Write back scaled values: values between 0-99.99 percentile of each frame is mapped between 0 and 4095 for each frame, so the new intensities in each of the frames will be similar.
  8. 2D DFT Butterworth BP filter: 80/1.5/2000/100/pixels/.../absolute=No : this is a highpass filter with a cuton at 80 pixels spatial frequency to show mitochondria only. Note that if the resolution of image acquisition is changed a different cuton frequency will be required. See more here and here.
  9. Threshold: Bottom at Pixel value = 0 (Threshold from local max/min=None): this removes negative values created by the high pass filter.
  10. Wiener Filter Smooth: adaptive filter for noise removal. Mask width=3, Noise level=0.01. Increase noise level for greater noise suppression or decrease if details are smudged.
  11. Image Arithmetic: the filtered Mitotracker image is masked with the cytosolic area
  12. Set scaling/color and Write back scaled values: values between 0-99.5 percentile of the whole image series resulting brighter images and saturation of the brightest pixels. The 99.5 is an important empirical value which sets the sensitivity of the method.
  13. Threshold: Mask below Pixel value = 100 (Threshold from local max/min=None): this suppresses background. Increase this value if mitochondria look overflowing.
  14. Threshold: Above Otsu by frames at 1, Threshold from local max/min=Bound Maxima Locally, Determine boundaries at 10(%). This performs the actual locally adaptive thresholding outlining close to the zero crossings of the highpass filtered mitochondrial profiles, therefore reflecting the width of the original fluorescence objects at half maximal intensities, thus the physical size of the mitochondria.
  15. Finally the area of the white (1) pixels is measured by drawing or loading a big ROI encircling the whole the whole image Plot ROIs/ Plot type=Sum.
• Copy Y Data Only by right clicking the plots and paste it into Excel. Sum the mitochondrial and the cytosolic pixels and take their ratio. Multiply the ratio with 0.667 correction factor.

		“volumeratio process confocal cortical-masked.ips” This pipeline was tuned to provide similar results as the EM-.
Original Mitotracker Red image. Shown at 25% zoom. Click to see in original size.	Original calcein image. Shown at 25% zoom. Click to see in original size.	Hand-masked calcein image	binarized calcein image 2.26E+05 pixels
Unmixed Mitotracker Red image	Highpass filtered Mitotracker Red image	Binarized, masked Mitotracker Red image 52581 pixels

• Stereologic correction factor

The factor K_V=0.667 derives from the stereologic correction formula for spheres with equal thickness to the section and truncated at half maximal intensity. Since mitochondria are thinner than the optical thickness, they are blurred by definition to the size of optical thickness, therefore the thickness of mitochondria equals to the section thickness. The correction formula  (Weibel and Paumgartner 1978), where g is the relative section thickness, which is 1 (see above). r is the relative smallest visible cap section, which is 1 where objects are clipped at half maximal intensity. Therefore K_V=2/3. Cells are a lot thicker than the focal plane therefore no correction factor applies for the calcein channel.

Protocol by Akos A. Gerencser 02/11/2010  updated 01/30/2014 V1.1        

Who to cite? This technique has been published here:

1. Gerencser AA, Chinopoulos C, Birket MJ, Jastroch M, Vitelli C, Nicholls DG, Brand MD. Quantitative measurement of mitochondrial membrane potential in cultured cells: calcium-induced de- and hyperpolarization of neuronal mitochondria. J Physiol. 2012 Jun 15;590(Pt 12):2845-71.

We used this technology in the following papers:

1. Birket MJ, Casini S, Kosmidis G, Elliott DA, Gerencser AA, Baartscheer A, Schumacher C, Mastroberardino PG, Elefanty AG, Stanley EG, Mummery CL. PGC-1α and Reactive Oxygen Species Regulate Human Embryonic Stem Cell-Derived Cardiomyocyte Function. Stem Cell Reports. 2013 Dec 12;1(6):560-74.
2. 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
3. Birket MJ, Orr AL, Gerencser AA, Madden DT, Vitelli C, Swistowski A, Brand MD, and Zeng X. A reduction in ATP demand and mitochondrial activity with neural differentiation of human embryonic stem cells. Journal of Cell Science, 2011 24:348-58.

Other references:

1. Integrated stereological and biochemical studies on hepatocytic membranes. II. Correction of section thickness effect on volume and surface density estimates.
   Weibel ER, Paumgartner D. J Cell Biol. 1978 May;77(2):584-97.