Tandem Tag Assay Optimized for Semi-automated in vivo Autophagic Activity Measurement in Arabidopsis thaliana roots

Materials and Reagents

1. 1.5 ml Eppendorf tubes 
2. Petri dishes (Thermo Fisher Scientific, catalog number: 150239 )
3. Pipette tips
4. 6-well tissue culture plates (Thermo Fisher Scientific, catalog number: 15213338 )
5. Micro cover glass 25 x 50 mm and 25 x 25 mm (VWR, catalog numbers: 48382-136 and 48366-089 )
6. Sealing Film, PVC (Phytotechlab, Product ID: A003) or Parafilm (VWR, catalog number: 52859-079 )
7. Bleach solution (Klorin, Colgate Palmolive)
8. Tween-20, Polysorbate (VWR, catalog number: 97062 )
9. 3D printed seed plating guide (https://www.thingiverse.com/thing:4016917)
10. Murashige and Skoog, MS medium (Duchefa Biochemi, catalog number: M0222 ) liquid medium
11. MES, 2-(N-morpholino) ethanesulfonic acid (Duchefa Biochemi, catalog number: M1503 )
12. Sucrose (Duchefa Biochemi, catalog number: S0809 )
13. KOH (VWR, catalog number: 470302 )
14. Plant agar (Duchefa Biochemi, catalog number: P1001 )
15. DMSO (Sigma-Aldrich, catalog number: 276855 ) or other vehicle
16. AZD8055 (Selleckchem, catalog number: S1555 )
17. Immersion oil (Zeiss, catalog number: 444960 )
18. Transgenic Arabidopsis lines used to establish this protocol were published in Dauphinee et al. (2019)
19. Milli-Q (MQ) water
20. Liquid 0.5x MS medium (see Recipes)
21. Solid 0.5x MS medium (see Recipes)
22. Bleach solution (see Recipes)
23. ConA 1 mM stock in DMSO (see Recipes)
    

Equipment

1. Pipettes for 100-1,000 µl and 1-10 μl
2. Forceps (Dumont, catalog number: 11251-10 )
3. 4 °C fridge
4. Arabidopsis growth cabinet/growth room: 20-22 °C, 50-70% humidity, 150 µM light
5. Confocal Laser Scanning Microscope (CLSM; Zeiss, LSM 800)
   

Software

1. Fiji, the version of ImageJ with included set of plugins (https://fiji.sc/, for this study, we utilized versions 1.51s and 2.0.0-rc-69/1.52i).
2. AuTToFlux repository containing three ImageJ macro and three R script files (https://github.com/jonasoh/AuTToFlux/archive/master.zip):
   1. CalibrateThreshold.ijm
   2. ImageProcessor.ijm
   3. FluorescenceIntensity.ijm
   4. EvaluateCalibration.R
   5. Control-vs-Reporter.R
   6. Flux-vs-Time.R
3. R (https://www.r-project.org, we used 3.5.2 and 3.5.1)
4. RStudio (https://www.rstudio.com/, we used versions 1.1.453 and 1.2.1186).
5. Git (https://git-scm.com/downloads, we used the version 2.24.1)
6. R packages:
   1. dplyr (https://CRAN.R-project.org/package=dplyr)
   2. ggplot2 (https://CRAN.R-project.org/package=ggplot2)
   3. readr (https://CRAN.R-project.org/package=readr)

Procedure

1. Seed sterilization (40 min)
   1. Place ca. 15 μl of seeds of reporter and control lines into 1.5 ml Eppendorf tubes. Generally, the aim is to have at least four-six biological replicates for each treatment and time point planned in the experiment.
   2. Add ca. 1.5 ml of the bleach solution.
   3. Incubate the seeds for ca. 30 min, agitating.
   4. Under sterile conditions, pipette out the bleach solution.
   5. Add sterile water to the seeds. Mix and pipette it out.
   6. Perform the wash with sterile water at least three times to wash out the leftovers of the bleach solution.
   7. Seeds can be vernalized either in the sterile water or as described in the Procesure C.
      
      
      
2. Seed plating (40 min)
   1. Use 1 ml pipette to transfer single seeds onto a Petri dish with solid 0.5x MS medium. Keep ca. 2-5 mm distance between the seeds in the same row to avoid entanglement of roots. Keep ca. 5 cm distance between the rows (Figures 2A-2B). It is recommended to use a seed plating guide for this procedure (Figure 1, Figure 2A, https://www.thingiverse.com/thing:4016917).
   2. Seal the plates with a strip of Parafilm or sealing film.
      
      
      Figure 2. Arabidopsis Tandem Tag autophagic flux assay. A. 3D printed seed plating guide for 9 cm Petri plates. B. Arabidopsis thaliana seeds plated using the seed plating guide (left half) and then grown with the plate positioned vertically for 5 days (right half). C. Arabidopsis seedlings placed in a 6-well plate for treatment. Note that the roots should be submerged gently (white arrow) and should not be floating on the surface (red arrow). D. Seedlings mounted on a glass slide (left) or coverslip (right) for CLSM. E. Approximate beginning of the root differentiation zone F. Images 0, 1 and 2 represent good, bad and ugly quality scans of differentiation zone vacuoles in the reporter line, respectively. G. Vacuole layer masks generated during calibration using the CalibrationThreshold.jim macro. Note that ugly scans F. may lead to complete or partial failure of vacuole detection. H. Sum vacuole areas detected on all analyzed images (2 is the recommended upper threshold value) vs. threshold values generated by the EvaluateCalibration.R script. I. Flux-vs-Time.R script output for vehicle and a compound treatment generated in R using ggplot2. Scale bars = 1.5 cm (B-D); 50 µm (E, F).
      
      
      
3. Vernalization (24-48 h)
   Incubate the plates in the dark at 4 °C for 24-48 h.
   
   
4. Growth on vertical plates (5 days)
   1. Place the plates into growth conditions (150 µM light, 22 °C for 16 h, 0 µM light, 20 °C for 8 h). Make sure the plates are perfectly vertical to ensure root growth on the top of the medium (Figure 2B).
   2. Grow the seedlings on the plates for ca. 5 days, the root length should reach ca. 2 cm.
      
      
      
5. Drug treatment (2-24 h)
   1. Pipette 3 ml of liquid 0.5x MS into wells of a 6-well tissue culture plate. Add the chemical compounds of the required concentration.
   2. Using soft forceps, gently pick seedlings from the plates and transfer into liquid 0.5x MS (not more than 20 seedlings/well).
   3. Gently pipette the medium from the well onto the seedlings in the well to submerge the roots (white arrow, Figure 2C).
   4. Seal the plate with a sealing film or as trip of Parafilm.
   5. Incubate the plate under the same growth conditions for the required amount of time.
      
      
      
6. Mounting the samples (1 min)
   1. Pipette a drop of medium, ca. 50 μl, from a well on the 25 mm x 50 mm cover slip.
   2. Using soft forceps, gently pick seedlings from the well and place it onto the drop of medium making sure that the root is straight. Avoid drying the roots during transfer!
   3. Place up to six seedlings on the cover slip.
   4. Cover the roots with 25 mm x 25 mm cover slip (Figure 2D).
   5. Apply immersion if needed to the objective lens and place the sample on the microscope stage.
      
      
      
7. Setting up scanning parameters for confocal laser microscopy (30 min, required only once)
   1. To estimate optimal range of settings for scanning, it is advisable to perform a pilot experiment using three control treatments with 500 nM AZD 8055 for 4 h, 500 nM ConA for 6 h and corresponding amount of vehicle (e.g., 0.05% DMSO).
   2. Configure settings for sequential scanning of two channels:
      1. Channel 1 for detection of mWasabi: excitation at 488 nm, emission detection range 490-564 nm.
      2. Channel 2 for detection of TagRFP: excitation at 561 nm, emission detection range 564-700 nm.
      3. It is advisable to use the most sensitive detectors available in the system (i.e., GaAsp or HyD detcors for Zeiss or Leica CLSM, respectively).
      4. Using 40x objective is advisable to obtain images most applicable for automated analysis.
      5. Set switching between channels for each frame to minimize the crosstalk between channels and optimize the pinhole size. If possible, use the pinhole of 1 AU for each of the channels.
      6. If possible, use 16-bit resolution to increase the resolution of intensities.
      7. The sample treated with AZD 8055 will have the weakest fluorescence and should be used to adjust laser intensity and the Detector Gain (Master Gain) to the lowest possible values that do not result in oversaturated pixels.
      8. The sample treated with ConA will have the strongest fluorescence, at least in the green channel, and should be used to re-adjust laser intensity and the Detector Gain (Master Gain) to the lowest possible values that do not result in oversaturated pixels.
      9. The sample treated with the DMSO can be used to verify the applicability of the adjusted settings.
      10. Additionally, noise can be decreased by ramping down the scanning speed or increasing the averaging number. In our experience, scanning at the speed of ca 10 seconds per frame produced images with the quality appropriate for further analysis while also resulting in acceptable time for experiment.
      
      
8. Scanning (ca. 30 min per treatment)
   1. The assay is optimized for measurement of autophagic activity in the epidermal root cells. Downstream image analysis relies on automated detection of the vacuoles in the green fluorescent channel. Thus optimally, images should be made in the focal plane where in each cell vacuole is surrounded by clearly visible cytoplasm.
   2. Most CLSM software will have an option for scanning at selected positions that will significantly decrease the time required for the experiment.
   3. Mark the positions at the beginning of the differentiation zone of the root (Figure 2E).
   4. Start fast scanning mode (live scan) and readjust the focal plane for each position to the middle section through the vacuole of the epidermal cells (Figure 2F with and without cortical cytoplasm)
   5. When all positions are readjusted, acquire and save the images. Please note, that automated statistical analysis will use information provided in the names of images. Please use the following rules to introduce the required parameters into the name of your images:
      
      1. separate parameters by underscores (_)
      2. Name the files as follows: line_treatment_seedlingX_imageY (e.g., Reporter_50uM.C12_seedling1_image2.czi)
         Where line and treatment are text variables or strings identifying the lines (either “Reporter” or “Control”) and treatments (i.e., “50uM.C12”) used. One treatment must be named “Vehicle”. X and Y are numbers indicating seedling and image replicates thus representing biological and technical replicates, respectively. Note that “seedling” and “image” are fixed strings and only the numbers X and Y are changed from image to image.

Data analysis

Data needs to be processed prior to the analysis. The processing is done in four steps: (i) conversion of confocal images into .TIF files; (ii) adjustment of the threshold values to optimize recognition of the vacuoles. This step must be performed when using the protocol for the first time and, if needed, can be redone for individual experiments; (iii) quantification of fluorescence intensities in the vacuoles; (iv) quantification of the ratios and plotting of the data as control vs reporter lines, or ratios vs time and concentration.
  We provide demo data (https://github.com/jonasoh/AuTToFlux), which can be used to ensure that the analysis works as expected.

1. Set up the analysis pipeline
   For more detailed instructions see Video 1 (https://youtu.be/6oY3CyvGFPk).
   1. Place all images for processing into a single folder. If several experiments should be processed, place them as subfolders within a single folder.
   2. Download and install Fiji, the version of ImageJ with included set of plugins (https://fiji.sc/) for this study, we utilized versions 1.51s and 2.0.0-rc-69/1.52i).
   3. Download and install R (https://www.r-project.org, we used 3.5.2 and 3.5.1) and RStudio (https://www.rstudio.com/, we used versions 1.1.453 and 1.2.1186).
   4. Either clone the GitHub repository (https://github.com/jonasoh/AuTToFlux) or download it as a zip file. If cloning the repository, RStudio can be used to automate this task: In RStudio, start a new project (File -> New Project -> Version Control -> GIT and input the following URL:
      https://github.com/jonasoh/AuTToFlux. The repository is automatically cloned into the chosen location.
   5. Before running the R scripts for the first time you need to make sure its dependencies are installed. Either use RStudio’s interface for installing packages, and install the packages ggplot2, dplyr, and readr, or install them manually by typing the following command into the console:
      
      install.packages(c("ggplot2", "dplyr", "readr"))
      
      
      
   
   Video 1. Repository version control in R studio
   
   
2. Convert images into TIFF files
   For more detailed instructions see Video 2 (https://youtu.be/nKPq0kNvW_U).
   1. Place all images for processing into a single folder. If several experiments are to be processed, place them as subfolders within a single folder.
   2. Run the Image Processor: launch ImageJ and go to Plugins -> Macros -> Run -> locate ImageProcessor.ijm.
   3. The image processor macro opens all compatible images in the target folder and saves them as .TIF files (with the file extension .tif); original acquisition dates of the images are saved in separate files. Multi-image files from experiments scanned using the “Positions“ function are split so that each image is saved as a separate .TIF file.
      
      
      
      Video 2. Image processing
      
      
3. Select proper threshold values
   For more detailed instructions see Video 3 (https://youtu.be/Wkw3VXFj2is). Copy at least three TIFF images generated by the ImageProcessor into a separate folder. Aim to select images representing the best, the worst, and average quality.
   1. Launch ImageJ and go to Plugins -> Macros -> Run -> locate CalibrateThreshold.ijm. This runs the macro. In the file picker immediately presented, locate the folder with representative images. Macro will record the area sizes corresponding to the vacuoles and also save masks as tiff files containing threshold values in their names, e.g., file named *.thr0-3.tiff would correspond to the mask created with the threshold values (0;3). The results overview will be generated as threshold-overview.tif (Figure 2G).
   2. Launch RStudio.
   3. Open the RStudio project file (AuTToFlux.Rproj).
   4. To estimate what threshold values provide the largest selected vacuole area for all analyzed images, run the EvaluateCalibration.R script (select the R script, click Code -> Source and then choose the folder containing the .CSV files generated in the previous step–note: on non-Windows systems you need to type in the pathname of the folder).
   5. The script will plot average sum vacuole areas selected on all analyzed images vs threshold values. Select the threshold value that corresponds to the largest area (Figure 2H).
   6. Using the information gathered from the threshold-overview and the threshold graph, select an appropriate upper threshold value for marking. The masks should not contain any background areas.
   7. Before proceeding with image thresholding and data analysis, ensure that the calibration folder is removed from the directory containing the processed images.
      
      
      
      Video 3. Treshold callibration
      
      
4. Measure fluorescence intensities
   For more detailed instructions see Video 4 (https://youtu.be/6jYqkYXOpiQ).
   1. Run the threshold macro: launch ImageJ -> Plugins -> Macros -> Run -> locate and open the macro file FluorescenceIntensity.ijm.
   2. Select the folder containing images to be analyzed, previously converted using the image processor macro (step B).
   3. The macro will prompt you to select the threshold value determined earlier in the calibration process. For each image the threshold macro automatically selects areas corresponding to the vacuoles using the GFP channel and saves the masks for the selected areas as TIFF files. The masks are then used to quantify intensities of RFP and GFP fluorescence in the corresponding channels of the image. Ratios of vacuolar RFP/GFP fluorescence intensities are saved to .CSV files with matching filenames.
      
      
      
      Video 4. Fluorescence intensity measurement
      
      
5. Analyze data
   To estimate dose and time-dependent responses the RFP/GFP ratios are plotted vs time using Flux-vs-Time.R. For more detailed instructions see Video 5 (https://youtu.be/0_wDY7RN_hk).
   
   1. Create an info.txt file in the folder in the folder containing the .CSV files for analysis. The initial time of treatment (format YYYY-MM-DD HH:MM, 24-hour time) is user-specified by creating a tab-delimited text file (using any spreadsheet software, e.g., Microsoft Excel) with two columns, which should be saved as info.txt e.g.,
      Treatment     StartTime
      Vehicle          2018-07-06 09:50
      Treatment     2018-07-06 10:00
   2. Launch RStudio.
   3. Locate the R project file in the directory created from the GitHub repository and open it if it is not already opened:
      
      1. In R select Flux-vs-Time.R, click Code -> Source and then choose the folder containing the .CSV files generated in step D.
      2. A summary of the experiment data is generated with ratios normalized to the vehicle and grouped by treatment and seedling number (Figure 2I). To plot the data, we recommend ggplot2 (https://ggplot2.tidyverse.org/), axes (x=Elapsed time, y=Normalized ratios, color=Treatment)
      3. Further statistical analysis will depend on the treatments present in the experiment and can be performed using R or other software, e.g. Origin, JMP.
      4. The data can be also used to estimate IC50 using R, e.g., with the aid of the drc package (Ritz et al., 2015), or other software such as Prism. 
   
   
   
   Video 5. Flux vs. time
   
   To demonstrate autophagy-dependent response, RFP/GFP ratios of control lines are plotted vs ratios of reporter lines. For more detailed instructions see Video 6 (https://youtu.be/PQRZ1oOBgws):
   
   1. Launch RStudio.
   2. Locate the R project file in the directory created from the GitHub repository and open it if it is not already opened:
      1. In R select Control-vs-Reporter.R, click Code -> Source and then choose the folder containing the .CSV files generated in step D.
      2. This script may take as its argument either a folder containing several treatments as subfolders, or a folder without subfolders that contains a single experiment.
      3. The script summarizes RFP/GFP ratios (normalized to vehicle) and groups by line (control or reporter), treatment and seedling number. An unpaired, two-tailed Student’s t-test is used to compare log-transformed normalized means of the control and reporter groups. A summary table named pvals.txt is generated. If P-values derived from permutation (i.e., exact P-values) are desired, uncomment the appropriate sections marked in the R script.

   Note: For both Control-vs-Reporter.R and Flux-vs-Time.R, data is saved both as raw data (as summary-full.txt in the experiment directory) and as per-seedling summary statistics (summary-perseedling.txt), to facilitate analysis using other statistical software.
   

   
   Video 6. Control vs. Reporter

Notes

See Table 1 for further troubleshooting steps.

Table 1. Troubleshooting

Recipes

1. Liquid 0.5x MS medium
   1. 0.5x MS complete with vitamins,10 mM MES, 1% sucrose dissolved in MQ water
   2. Adjust pH to 5.8 using KOH and autoclave for 20’ at 120 °C
   3. Store at 4 °C for max. 2 months
2. Solid 0.5x MS medium
   1. After adjusting pH of the liquid 0.5x MS medium, add 0.8% Plant agar and autoclave for 20 min at 120 °C
   2. Store at 4 °C for max. 2 months. If needed, the medium can be liquified in the microwave
3. Bleach solution
   1. 5% Klorin (final concentration: 2.7 g/L sodium hypochlorite), 0.01% Tween-20 in MQ water
   2. Prepare in 50 ml Falcon tube and store at RT
   3. After ca. 1 month Tween-20 might form precipitates in the solution, however, this does not impact the efficacy of the sterilization
4. AZD 8055 5 mM stock in DMSO
   Store at -20 °C for max. 2 years
5. ConA 1 mM stock in DMSO
   Store at -20 °C for max. 1 year
   

Acknowledgments

This project was supported by Carl Tryggers Foundation (to EAM), Natural Sciences and Engineering Research Council (NSERC) of Canada (to AND) and Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (Formas) under Project Number 2016-20031. This protocol was established for the study carried out by Dauphinee et al. (2019)

Competing interests

The authors declare no competing financial interests.

References

1. Dauphinee, A. N., Cardoso, C., Dalman, K., Ohlsson, J. A., Berglund Fick, S., Robert, S., Hicks, G. R., Bozhkov, P. and Minina, E. A. (2019). Chemical screening pipeline for identification of specific plant autophagy modulators. Plant Physiol 181(3):855-866.
2. Hanamata, S., Kurusu, T., Okada, M., Suda, A., Kawamura, K., Tsukada, E. and Kuchitsu, K. (2013). In vivo imaging and quantitative monitoring of autophagic flux in tobacco BY-2 cells. Plant Signal Behav 8(1): e22510.
3. Huang, R., Xu, Y., Wan, W., Shou, X., Qian, J., You, Z., Liu, B., Chang, C., Zhou, T., Lippincott-Schwartz, J. and Liu, W. (2015). Deacetylation of nuclear LC3 drives autophagy initiation under starvation. Mol Cell 57(3): 456-466.
4. Klionsky, D. J., Abdelmohsen, K., Abe, A., et al. (2016). Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition). Autophagy, 12, 1554-8635 (Online).
5. Zhou, C., Zhong, W., Zhou, J., Sheng, F., Fang, Z., Wei, Y., Chen, Y., Deng, X., Xia, B. and Lin, J. (2012). Monitoring autophagic flux by an improved tandem fluorescent-tagged LC3 (mTagRFP-mWasabi-LC3) reveals that high-dose rapamycin impairs autophagic flux in cancer cells. Autophagy 8(8): 1215-1226.
   6.Ritz, C., Baty, F. and Strebig, J. C. (2015). Dose-response Analysing Using R. PLoS ONE 10(12): e0146021.