Fluorescence microscopy image analysis
automation - time series - physiology

The Unbiased Absolute Membrane Potential Assay in Practice

The assay comprises recording and analysis parts. You perform the recording on an arbitrary live-cell fluorescence microscope setup based on our protocols and publications. You use your own reagents or our reagents that will be available later. You perform image and data analysis in Image Analyst MKII. Purchased software license allows all analysis required to perform the analysis.

 

The Assay Paradigm

Mitochondrial Membrane Potential Assay

Step 1: The potentiometric paradigm

The assay can be performed on any epifluorescence (suitable for low-light level time-lapse imaging), confocal or two-photon system. A prototype experiment is exemplified below using human pancreatic beta-cells (published in (26)). An arbitrary time course recording is followed by a standardized calibration paradigm. The calibrant cocktails have been described in (17), and their preparation in the protocols section.

Step 2: Image Analysis

The mitochondrial membrane potential analysis can now be performed using an interactive protocol in the Primer Window by selecting Assays/Intensity and Ratio Measurements/ Mitochondrial membrane potential assay - worked example. Buttons in the protocol actually operate Image Analyst MKII.

Mitochondrial membrane potential assay - worked example - an interactive protocol
Abstract

This assay performs all image and data processing tasks required for calibration of a TMRM / PMPI (tetramethylrhodamine methyl ester and FLIPR plasma membrane potential assay kit; aka Plasma Membrane Potential Indicator) to millivolts. For TMRM complete calibration is performed, thus all calibration parameters are calculated. For PMPI the rate constant kP is assumed, therefore this protocol is more suitable for following slower changes (e.g. using >=10s acquisition intervals).

The pipeline corrects for misalignment between channels on the frame-by-frame basis, subtracts background, performs spectral unmixing with pre-defined coefficients, and stabilizes the time lapse (registers frames in time). To eliminate background the “Median of pixels below percentile of max projection” with the percentile given at “Background level” is used.

Then ROIs are generated automatically using the nuclear stain as seeds, following the shape of the maximum intensity projection of the summed PMPI+TMRM image. Therefore if a cell moves during the experiment, the ROI will contain the cell for the whole duration of the recording. The calibration is not affected by the amount of background recorded in the ROI.

Prepared TMRM and PMPI images with ROIs are passed to the Membrane Potential Calibration Wizard. Calibration can be performed by manual configuration of the wizard, or using a previously saved configuration file to provide automation.

The Membrane Potential Calibration Wizard implements the calibration technique published by Gerencser et al. in J Physiol. 2012 590:12 2845-71. See more practical details in PLoS One. 2016 Jul 12;11(7):e0159199.

Assay Version 1.1 6/17/2018

Input

Worked example: Download MitoMemPot1.zip ("Calibration of TMRM/PMPI recordings") and unzip contents onto the hard drive.

Sample: Human non-diabetic primary β-cells in a coverglass bottomed 96-well plate (Corning 4580). Published in Endocrinology 2015 Oct;156(10):3496–503.

Fluorophores:

  • Channel (1): TMRM filter set: 580/14 – 594 – 641/75 (150 ms exposure time)
  • Channel (2): PMPI filter set: 500/24 – 520 – 542/27 (75 ms; all filters from Semrock)

 

Microscope: Nikon Ti-Eclipse Perfect Focus System, equipped with a Cascade 512B camera and S-Fluor 20× air lens, a Sutter Lambda LB-LS17 Xe-arc light source and 10-3 filter wheels and an ASI MS-2000 linear encoded motorized stage, controlled by NIS Elements 4.0.

Experimental paradigm:

  • Baseline: in PM (see protocol) 30 frames
  • Addition #1: 16mM glucose ~ 30min (for positions 1 and 2)
  • Addition #2: nothing ~ 30min (for positions 1 and 2)
  • Addition #3: MDC + ASC2 +TTX (see abbreviations in the protocol or in Gerencser et al. in J Physiol. 2012 590:12 2845-71) ~ 20 min
  • Addition #4: stir 5 frames, the volume is 150 µl now
  • Addition #5: +10 µl PMK (supplemented with MDC + ASC2 +TTX)
  • Addition #6: +20 µl PMK
  • Addition #7: -40 µl  PM +40 µl PMK
  • Addition #8: -75 µl  PM +75 µl PMK
  • Addition #9: -100 µl  PM + 80 µl PFA in PMK
Analysis protocol
Experimental Protocol

Link to: Absolute mitochondrial membrane potential measurement in β-cells

Image Processing The pipelines used here can be accessed from the main menu: Pipelines/Intensity Measurements/Applications. Use the buttons below to activate the pipelines and perform the analysis:

 

Membrane Potential Calibration Wizard

1.       Calibration Method tab:

1.1.     Choose the plasma membrane potential calibration method: “Complete with known kP (K-steps)”

1.2.     Choose the mitochondrial membrane potential calibration method: “Complete”

2.       Input/Output tab: the input images have been already selected by the pipeline.

3.       Wizard – Data Ranges tab:

3.1.     Select baseline by pointing the range in one of the graphs on the left, and press the “Select baseline" button or alternatively 1-30 next to the “Select baseline" button.

3.2.     Select complete mitochondrial depolarization in the graph, and press the “Select MDC (K-eq)” button (135-169).

3.3.     Select final complete depolarization in the graph, and press the “Select CDC (zero)” button (222-245).


3.4.     Select the K+-steps, the five segments visible between the mitochondrial and complete depolarization in the top left graph, and press the “Select a K+ step(s)” button.

3.5.     To calculate [K+] during K-steps, enter the K+ concentration in the potentiometric medium (PM), that was 5.3 mM and in the K-based potentiometric medium (PMK), that was 125.3 mM, and the medium volume (150 µl). The K-steps were performed by first adding 10 µl PMK to the assay well. Enter 10 to the second line and “PMK Addition” column in the K-steps table. This was followed by addition of 20 µl (enter it in the next line), then removal of 40 µl and addition of 40ul (in the next line enter 40 to the “PM Removal” column and 40 to the “PMK Addition” column. Finally 75 µl was removed and 75 µl was added (fifth line of the table). Press the “Calculate [K+]ec button”.

K-steps

4.       Wizard – Assumed Parameters tab: No action is required, use the default rate constant for PMPI redistribution kP=0.38 s-1, and assume no significant non-K-permeability of the plasma membrane during K-steps by using the default PN=0.

5.       Wizard – Constants tab: set the cell specific parameters here:

5.1.     VF (mitochondrion:cell volume fraction): use the Measurement of mitochondria:cell volume fraction (VF)  protocol to measure this value using confocal microscopy, assume it based on literature. For the example data set it was measured to be 0.0772 ± 0.0016.

5.2.     VFM (matrix:cell volume fraction): this value affects the results only little (because aR’ –see below- largely cancels its effects). The default is 0.8

5.3.     aR’ (apparent activity coefficient ratio): use the Measurement of apparent activity coefficient ratio for TMRM (aR')  protocol to measure this value using confocal microscopy. For the example data set it was measured to be 0.361 ± 0.005

5.4.     Leave all other constants at their default values.

6.       Press the Calibrate button to perform the calibration. Note: not all cells in the view field can be calibrated, therefore error messages will appear to note this, unless  “Suppress messages” in the bottom of the Calibration Wizard is checked.

7.       To save results

      7.1  For Excel press  and save the contents of the Excel Data Window from the File main menu.

      7.2 Alternatively to record data to a Graphpad Prism file, first click File/Create New or Open Existing Prism File or the button in the Wizard. See more about saving to Excel or Prism here.

       7.3 To directly access data right-click / ”Copy Plot Data” in the graphs on the graphs on the right.

8.       To explore the results:

8.1.     Use the  and  buttons to see the regression analysis used to calculate calibration parameters

8.2.     Use the  button to see only the time course preceding the calibration steps.

8.3.     Use the  button to see only those plasma membrane traces that also were successfully calibrated for mitochondrial membrane potential. Press  again to refresh the results.

8.4.     Right-click / “Calculate Mean” to see mean±SE of all calibrated traces. Note: if the mean calculation is performed on the fluorescence traces on the left, then the mean data will be calibrated.

 

Adjustments

1.  Configuration of the pipeline:

2. Spectral unmixing

The analysis of TMRM/PMPI recordings requires spectral unmixing. See related "Calculation of spectral unmixing coefficients for the ΔψM assay" interactive protocol in the Primer. The spectral unmixing coefficient matrix qualifies the configuration of the microscope, so it has to be re-measured if the relevant microscope configuration changes, but not for different specimens. The ratio of the two (TMRM and PMPI) exposure times affects the coefficients, but a change exposure time can be accounted for by changing exposure correction parameters in the Spectral Unmixing function by editing the pipeline.

Enter the coefficient matrix as calculated.

3. Quality control of the calibration:

3.1.     Check “Calculate the error of calibration” in the Data ranges tab within the Wizard tab and press . Now the predicted error of the calibration is shown. Note: optionally, check the “Suppress messages” on the bottom.

3.2.     Check “Edit all parameters”. Detailed lists of all parameters for PMPI and TMRM calibration, common calibration constants, and error propagation parameters are shown in tabs appearing on the right of the Wizard tab on the top.

3.3.     In the PMPI parameters, set Quality control by propagated error of baseline to “Yes” and press . Traces with larger predicted error than set at the parameter below disappear now.

3.4.     In the TMRM parameters, set Quality control by propagated error of baseline to “Yes” and press . Traces with larger predicted error than set at the parameter below disappear now.

4.    Using the Expert Mode: When the Edit all parameters checkbox is checked in the Data ranges tab within the Wizard tab, all parameters of the calibration algorithms can be directly accessed in the tabs appearing on the right of the Wizard tab. These can be used alternatively to the Wizards tab to enter any of the parameters. Use the “Fill in Range(s)” button to automatically enter a range from the selection made in the right graphs.

5.    Fine tuning the calibration

5.1. Switch to the “Constants” tab, for this the “Edit all parameters” needs to be checked in the “Wizard” tab.

5.2. To accommodate to the temporal resolution and noise of the recording, adjust the “Differentiation kernel width” to 11 frames. Larger width suppresses noise by providing more smoothing, but also smears fast changes.

5.3. Median filter for baseline, fft0 and fp0: if using the “Take maximum of CDC Range” in the “PMPI Parameters” tab or “Take minimum of CDC Range” in the “TMRM Parameters” use this option to suppress noise affecting minimum and maximum calculations.

Output

1.   Output options

  •  Excel: Click  in the Wizard. Save the contents of the Excel Data Window from the File main menu.

  • Alternatively to record data to a Graphpad Prism file, first click File/Create New or Open Existing Prism File or the button in the Wizard. See more about saving to Excel or Prism here.

  • To directly access data right-click / ”Copy Plot Data” in the graphs on the graphs on the right.

2.    Automation

2.1. Save an arbitrary calibration configuration using the  button in the Membrane Potential Calibration Wizard.

2.2. In the Pipeline Parameters (Main Window Parameter Bar; the “Mitochondrial membrane potential assay (TMRM/PMPI)” pipeline is activated) click the “Calibration configuration file name (*.ips)” parameter, and click the button appearing at the end of the line. Select the saved calibration file.

2.3. If using Excel, to automate saving the results give a filename in the “Output Excel Data save file name (*.xlsx)” using the string parser, such as =%LoadBaseName%%LoadPositionNumber:2%.xlsx. See more about the string parser in the Main menu Help/”Help on Expression Evaluation”. Optionally give a path, or set the default path to be used in the Main menu Files/”Set Folder Locations”.

2.4. If using Prism, newly processed recordings will be automatically appended to the open Prism file.

2.5. Use the pull down menu of the  button on the main toolbar or in the bottom of the Multi-Dimensional Open dialog to select “Run pipeline … on all stage positions” or “Run pipeline … on partial plate”.


Alternatively, select from a set of pre-configured pipelines in the main menu of Image Analyst MKII to process fluorescence time-lapse image recordings and automatically draw ROIs to extract fluorescence intensities. See tutorial image data for potentiometric image analysis here, and a video tutorial here.

Pre-configured pipelines in Image Analyst MKII


Human beta-cell with TMRM and PMPI


Left: Microscopic view-field of TMRM (red) and PMPI (green) fluorescence in pancreatic beta-cells.



Human beta-cell line scan

Right: Effect of channel alignment and image stabilization of the time lapse, demonstrated by a line scan


These image processing steps are included into the standard pre-processing pipelines for potentiometric calibrations.



Step 3: Calculation of millivolts using the Membrane Potential Calibration Wizard

The Membrane Potential Calibration Wizard guides the user through the calibration procedure. The user selects the experimental paradigm matching the recording, then points the ranges of calibrant additions on the fluorescence time courses, and finally enters required additional parameters. In the complete calibration paradigm only volumes of K+-based medium additions are required to perform the calibration.  

Eight ΔψP calibration paradigms are available supporting a variety of experimental designs. These differ in what calibration points are used and what assumptions are made. The complete calibration uses no assumption on any specimen-specific parameter.
  • Complete (temporally resolved K-steps)
  • Complete with known kP (K-steps)
  • Baseline & Zero
  • Baseline & K-equilibrium
  • K-equilibrium & Zero
  • K-steps with known [K+]i
  • Zero (fx=0)
  • Baseline & fP0
Three ΔψM calibration paradigms are available, that can be used together with any of the ΔψP calibration paradigms:
  • Complete
  • Complete (known k)
  • Baseline & MDC or CDC (known rest)
Membrane Potential Calibration Wizard

Step 4: Automation

Using pipeline automation the Membrane Potential Calibration Wizard can be executed completely automatically in each stage position or well of a microplate, analyzing thousands of individual cells in a single run. Millivolt calibrated data can be collected in Graphpad Prism from multiple conditions and experiments including pooling technical replicates.

Processed TMRM/PMPI image

As part of the automated pipeline processing, individual cells are identified by an automatic ROI drawing feature.

MMP pipeline

The menu-accessible pipelines are invisible for normal operation, and only key-parameters are shown in the main parameter bar. However, pipelines can be opened for editing and arbitrarily changed.




Development of the unbiased, absolute mitochondrial membrane potential assay has been supported by: SBIR/STTR  NIDA


Theory Overview