Jan 09, 2026
A live cell biosensor protocol for high-resolution screening of therapy-resistant cancer cells
Peer-reviewed method
- Viral D. Oza1,2,
- Colin S. WIlliams1,2,
- Jessica S. Blackburn1,2
- 1Department of Molecular and Cellular Biochemistry, University of Kentucky, Lexington, KY, USA;
- 2Markey Cancer Center, University of Kentucky, Lexington, KY, USA
- PLOS ONE Lab ProtocolsTech. support email: [email protected]
External link: https://doi.org/10.1371/journal.pone.0343016
Protocol Citation: Viral D. Oza, Colin S. WIlliams, Jessica S. Blackburn 2026. A live cell biosensor protocol for high-resolution screening of therapy-resistant cancer cells. protocols.io https://dx.doi.org/10.17504/protocols.io.eq2ly4d7qlx9/v1
Manuscript citation:
Oza VD, Williams CS, Blackburn JS (2026) A live cell biosensor protocol for high-resolution screening of therapy-resistant cancer cells. PLOS One 21(2). doi: 10.1371/journal.pone.0343016
License: This is an open access protocol distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Protocol status: Working
We use this protocol and it's working
Created: July 08, 2025
Last Modified: January 09, 2026
Protocol Integer ID: 221895
Keywords: therapy resistance, cell death, biosensor, GEDI, genetically encoded, empirical gedi threshold calibration, live cell biosensor protocol, need for empirical gedi threshold calibration, expressing cancer cell line, resistant cancer cell, encoded death indicator, cell death through calcium influx, resistant cancer cells the genetically, cancer cell, death annotation, fluorescence biosensor, cancer cell line, death indicator, vivo validation, radiation screening, radiation screening in heterogeneous population, gedi ratio value, cell resolution, cell morphological data in parallel, cell fate, transcriptomic profiling of resistant clone, transcriptomic profiling, fluorescence, gedi, resolution screening of therapy, cell time, imaging parameter, induced gedi
Funders Acknowledgements:
National Institutes of Health
Grant ID: R37CA227656
National Institutes of Health
Grant ID: F99CA294265
Kentucky Pediatric Cancer Research Trust
Grant ID: PON27282400002665
National Institutes of Health
Grant ID: P30CA177558
Abstract
The Genetically Encoded Death Indicator (GEDI)
is a ratiometric, dual-fluorescence biosensor that enables real-time detection
of cell death through calcium influx. Originally developed for use in
neurodegeneration models, GEDI can be applied to cancer cells to quantify
therapy-induced death at single-cell resolution. This protocol details how to generate
GEDI-expressing cancer cell lines, empirically determine stress-induced GEDI
thresholds using radiation or chemotherapeutic agents, and perform time-resolved
imaging and image analysis to track cell fate. This workflow is optimized for
high-throughput drug and radiation screening in heterogeneous populations and
is especially useful for identifying chemo- and radio-resistant subclones. Key limitations
include the need for empirical GEDI threshold calibration for each treatment
condition and careful standardization of imaging parameters. The protocol
outputs include GEDI ratio values, single-cell time-of-death annotations, and
whole-cell morphological data in parallel, which can be linked to downstream
applications such as FACS-based isolation of live or dying subpopulations,
transcriptomic profiling of resistant clones, or in vivo validation using
xenografts or organotypic slice culture.
Guidelines
General considerations:
For best signal to noise ratios set the GC150 exposure to 2X that of mApple exposure
Sparse seeding of cells will enable greater accuracy of tracking algorithms
Name Image files with relevant metadata for straightforward dataframe manipulation downstream
For access to representative ImageJ output and instructions regarding downstream input into R see the github repository "GEDI_Cancer"
Materials
Biological Materials
1. H3K27M-pDMG cell line SF8628 (SCC127, Millipore Sigma)
2. HEK293T celll line (12022001, Sigma)
Reagents
1. 0.25% Trypsin (25300062, Life Technologies)
2. DMEM 1X with 4.5g/L Glucose (D6546, Sigma), L-Glutamine (07100, Stem Cell Technologies), & Sodium Pyruvate (07000, Stem Cell Technologies)
3. FBS (S11150H, R&D Systems)
4. PBS 1X (PBD500-CS, Sigma)
5. Lipofectamine (L30000015, Thermo Scientific)
6. Lenti-X (631231, Takara Bio)
7. HSBSS (55037C, Sigma)
8. Trypan Blue (C10283, Thermo Scientific)
9. pLenti-CMV-Puro destination vector (17452, Addgene)
10. psPAX2 (12260, Addgene)
11. pMD2.g (12259, Addgene)
12. Invitrogen Gateway technology (11791020, Life Technologies)
13. TransIT reagent (2300, Mirus Bio)
14. Lenti-X (631231, Takara Bio)
15. Caspase-Glo 3/7 Reagent (PAG8091, VWR)
16. Cell Titer-Glo Luminescent Cell Viability Assay Reagent (PAG7572, VWR)
17. Optimem (31985070, Fisher Scientific)
Cell culture media
1. DMEM 1X with 4.5g/L Glucose
2. 10% FBS
3. 1X L-Glutamine
4. 1X Sodium Pyruvate
Laboratory Supplies and equipment
1. Biotek Lionheart
2. BD FACS Aria
3. 100mm tissue culture dish
4. Hemocytometer cell counting chamber
5. Computer with at least 7.9Gb memory
6. FACS tubes (60819-295, VWR)
7. FACS strainer (23A00A612, Thomas Scientific)
8. 15mL conicals
9. 96 well black wall plates (29444, VWR)
10. Biotek Synergy LX Multi-Mode Reader
11. Luminescent Filter Cube (1505003, Biotek)
12. 96 well white walled plates (29444, VWR)
13. LookOut Mycoplasma PCR Detection Kit (MP0035, Sigma)
Protocol materials
pLenti CMV Puro DEST addgeneCatalog #17452
Gateway LR Clonase II Enzyme mixThermo FisherCatalog #11791020
EcoR1-HFNEBCatalog #R3101S
Caspase 3/7 GloVWR International (Avantor)Catalog #PAG8091
0.25% Trypsin Life TechnologiesCatalog #25300062
Lenti-XTakara Bio Inc.Catalog #631231
psPAX2addgeneCatalog #12260
pMD2.gaddgeneCatalog #12259
TransITMirus BioCatalog #2300
OptimumFisher ScientificCatalog #31985070
1x PBS (Phosphate Buffered Saline)Merck MilliporeSigma (Sigma-Aldrich)Catalog #PBD500-CS
HSBSSMerck MilliporeSigma (Sigma-Aldrich)Catalog #55037C
Cell Titer GloVWR International (Avantor)Catalog #29444
FBSR&D SystemsCatalog #S11150H
1X L-GlutamineSTEMCELL Technologies Inc.Catalog #07100
1X Sodium PryvateSTEMCELL Technologies Inc.Catalog #07000
HEK 293TATCCCatalog #ATCC CRL-3216)
DMEM 1X w 4.5g/L GlucoseMerck MilliporeSigma (Sigma-Aldrich)Catalog #D6546
Troubleshooting
Safety warnings
Ensure proper handling when using lentivirus and lentiviral infected materials.
Ethics statement
All human cell lines were obtained from recognized repositories.
Part I: Prepare GEDI lentivirus, transduce and sort cells
1w 6d 0h 50m
To generate GEDI lentivirus, the pMe:GC150-p2A-mApple construct (gift from the Finkbeiner Lab, Gladstone Institutes) was cloned into a pLenti CMV Puro DEST addgeneCatalog #17452 vector using Gateway LR Clonase II Enzyme mixThermo FisherCatalog #11791020 , according to manufacturer’s instructions. Clones were selected based off an EcoR1-HFNEBCatalog #R3101S restriction digest and confirmed through whole plasmid sequencing.
3d
Start culture of HEK 293TATCCCatalog #ATCC CRL-3216) cells in DMEM 1X w 4.5g/L GlucoseMerck MilliporeSigma (Sigma-Aldrich)Catalog #D6546 with 10%FBSR&D SystemsCatalog #S11150H , 1X L-GlutamineSTEMCELL Technologies Inc.Catalog #07100 , and 1X Sodium PryvateSTEMCELL Technologies Inc.Catalog #07000 .
Lentivirus was produced by transfecting
5m
Mix the two solutions together and pipette up and down. Incubate 00:20:00 at room temperature. Add total mixture drop wise to 10cm plate. Change media next day.
20m
Viral supernatant was collected at 48:00:00 and 72:00:00 post-transfection. Viral supernatant was spun 00:05:00 at 300 x g. Supernatant was collected and filtered through 0.45 micron filter. Viral supernatant concentrated using Lenti-XTakara Bio Inc.Catalog #631231 Concentrator and stored in 10 µL aliquots at -80 °C .
2d 0h 5m
Seed cells at 10000 cells per well and test a range of lentivirus doses ranging from 0.25 µL - 2 µL per well.
Check for expression and toxicity using brightfield and a filter set amenable to mApple excitation/emission. We used a Lionheart FX imager with the following LEDs and filter cube sets:
Equipment
RFP filter cube
NAME
filter cube
TYPE
Biotek
BRAND
PN:1225103
SKU
Equipment
GFP filter cube
NAME
filter cube
TYPE
Biotek
BRAND
PN:225101
SKU
Equipment
523nm LED cube
NAME
LED cube
TYPE
Biotek
BRAND
PN:1225003
SKU
Equipment
465nm LED cube
NAME
LED cube
TYPE
Biotek
BRAND
PN: 1225001
SKU
Equipment
Lionheart FX Imager
NAME
Microscope
TYPE
Biotek
BRAND
After lentivirus dose has been determined, transduce cells in a 10cm dish at ~50 % confluency
After 120:00:00 , start feeding cells and prepare test cells for sorting using a
Equipment
iCyt-Sony Cell Sorter
NAME
cell sorter
BRAND
5d
For cells to be sorted, repeat the following steps (Part 2: 15-18).
During resuspension, resuspend cells in HSBSSMerck MilliporeSigma (Sigma-Aldrich)Catalog #55037C warmed to 37 °C
Dilute cells to
Set up a negative control cell population from non-transduced cells. Set up a positive control by resuspending 5.0 * 105- 1.0 * 106 GEDI transduced cells in 1 mL and heat shock at 42 °C for 00:10:00 , then strain through FACS strainer into FACS tube and place on ice.
10m
Optional: For a secondary positive control spike in NaN3 for a total concentration of 2 % (v/v) for 00:10:00 before sorting begins.
10m
During sorting, gate for cells with high mApple expression and low GC150 expression. Set a negative gate on unstained cells (low mApple, low GC150) and a positive gate on dead cells (high mApple, high GC150).
Let cells recover for several days after sorting, at least 72:00:00 before using downstream experiments.
3d
Part II: Determine a lethal dose of drug of interest or x-ray radiation using Caspase 3/7 Glo or Cell Titer Glo
10m
When cells are ~70 % confluency in a 10cm dish wash with1x PBS (Phosphate Buffered Saline)Merck MilliporeSigma (Sigma-Aldrich)Catalog #PBD500-CS , then aspirate PBS.
Add 2 mL of 0.25% Trypsin Life TechnologiesCatalog #25300062 directly to plate and
incubate in cell culture incubator for 00:05:00
5m
Quench Trypsin with 4 mL of freshly warmed
media and transfer to 15mL conical tube.
Spin 15mL conical tube for 00:05:00 at 300 x g and resuspend cell pellet in 1 mL of fresh media.
5m
Seed cells at a concentration adequate for downstream experiments.
Drug Treatment
4h
Seed cells at a concentration of lower 5000 cells per well for drug treatment using steps from Part 2: 15-18. Total plating volume will be 50 µL per well.
Wait at least 04:00:00 after plating or until cells have attached back to the plate.
4h
Add 50 µL of freshly prepared 2X drug directly to the well and pipette up/down 3X
Proceed to Caspase 3/7 or Cell Titer Glo step.
Radiation Treatment
4h
Seed cells at a concentration of lower 5000 cells per well for radiation treatment using steps from Part 2: 14-18. Total plating volume will be 100 µL per well.
Wait at least 04:00:00 after plating or until cells have attached back to the plate.
4h
Bring 2 plates to the irradiator in Styrofoam housing.
Place plate getting doses > 0Gy into irradiator and adjust settings to achieve 8Gy exposure.
On a XRAD 225XL beam parameters are as follows, 13.3mA, SSD = 40, 225kV. For 8Gy, set the dose to 718.6.
Equipment
XRAD 225XL
NAME
Precision X-Ray
BRAND
Proceed to Caspase 3/7 or Cell Titer Glo step.
Cell Titer Glo
12m
Take tissue culture plate out and cool to room temperature before adding reagent.
At appropriate timepoint prepare Cell Titer GloVWR International (Avantor)Catalog #29444 (CTG) reagent at a
With the 100 µL of media in the well, add 100 µL of CTG to each well, pipette up/down 3X and shake, covered in foil, on orbital shaker for 00:10:00
10m
Let stand for 00:02:00 at room temperature and then read in plate reader with Luminescence detection mode.
Equipment
Biotek Synergy LX Multi-Mode Reader
NAME
Biotek
BRAND
Equipment
Luminescent Filter Cube
NAME
Biotek
BRAND
1505003
SKU
2m
Caspase 3/7 Glo
30m
At appropriate timepoint prepare Caspase 3/7 GloVWR International (Avantor)Catalog #PAG8091 reagent in a total volume for 100 µL per well.
With the 100 µL of media in the well, add 100 µL of Caspase 3/7 Glo reagent to each well and pipette up/down 3X. Cover and incubate at 37 °C for at least 00:30:00 , can go up to 2 hours if necessary.
30m
Read in plate reader with Luminescence detection mode.
Part III: Determine GEDI threshold for cell death
Seed cells at a concentration of lower 2500 cells per well or lower to allow adequate cell tracking using steps from Part 2: 15-18.
Image GEDI expressing cells before stress stimulus.
Imaging of GEDI is done using filter sets described above, Part 1: 4.
Use filter sets described in Part 1: 5 for imaging. Exposure times will vary, recommended exposure settings are 150ms for mApple and 300ms for GC150 with a low-medium Gain setting.
Setup a Gen5 protocols recommended with the following considerations:
Determine optimal mApple exposure without saturating pixels.
Set GC150 exposure to twice that of mApple.
Set magnification to 20X for best signal to noise ratio.
Capture 10 x 10 montage of each well.
Use the following data reduction steps for post image processing. Background Subtraction (Rolling point-point spread function), Deconvolution (Gaussian), Stitching (linear blend without downsizing)
Induce Radiation and/or Drug Stress to cells.
Add appropriate amount of 2X drug in a 100uL volume and add to 100uL of volume in the plate.
If irradiating, transport plate to irradiator in Styrofoam housing, expose to appropriate amount of x-ray and return to cell culture incubator or the microscope.
Continue imaging GEDI expressing cells at timed intervals after stress induction.
Image J Data Analysis
Quantify the GEDI ratio per object (cell) in each image by running a script (attached) with these basic principles:
1) Threshold objects based on the mApple intensity. (FIJI script)
2) Use these object masks to measure the fluorescence intensity of the mApple and GC150 signal. (FIJI script)
FIJI Script:
//This FIJI script is designed to take RFP and GFP images from GEDI data and output 1 CSV
//quantifying the fluorescent intensities from both channels
//It is specifically written for images of SF8628 cells that have been exposed to 25 Gy and seeded //in 96 well plates at a 2500 cell per well density
//Designate where/what directory/folder the CSV output will be saved
#@ File(label = "Output directory", style = "directory") output
//Make a list of images so it can be called by an integer value
list = getList("image.titles");
//The list integers start from 0, so double check which images are opening first by showing the //array, this will dictate what image in the list you are using below
Array.show(list);
//Select the open image window by its integer value from the list to do further processing on, in //this case we are generating cell masks from the RFP channel which is the second image in our //image list, remember the list starts with 0 and the first image in the order that we opened it is
// GFP, so to select the 2nd image that is open we would select the list integer of "1" to designate selecting the RFP image.
selectWindow(list[1]);
//In this case we first want to duplicate the image to threshold and leave the original untouched //because we will be overlaying the mask
//on top of it to extract fluorescence intensity data
//Command to duplicate the open image
run("Duplicate...", " ");
//Manually set the threshold, the minimum value is one that should be tested/iterated for what //works best to capture your cell mask of interest
setThreshold(1139, 65535);
setOption("BlackBackground", true);
run("Convert to Mask");
///Analyze all the particles in your image to generate ROIs, we will use these ROIs to place on the //original image and measure the fluorescent intensity
//of the desired channel
run("Analyze Particles...", "size=150-Infinity display clear include summarize add");
//The parameters under analyze particles can be changed to exclude a certain size, this will //need to be tested depending on the size of your object and the magnification used,
//and what you are interested in. In our test case, we have measured our cells and want the //minimum cutoff for particle counting to be 150.
//If there are artifacts in your image that you can avoid using, the size or feret diameter filter this is //a good place to implement that
//You need to set the types of measurements that will be extracted after measuring the image //mask, these can be changed, the best way is to record a macros
//and select specifically what you want measured and then paste that macros line here
run("Set Measurements...", "area mean standard min feret's integrated area_fraction display redirect=None decimal=3");
//Clear the results becasue we do not want mask measurements in the final CSV file
run("Clear Results");
//Select the original image, in this case it is the RFP image
selectWindow(list[1]);
//Overlay the ROIs from the ROI manager onto the image of interest
run("From ROI Manager");
//Now measure all of our ROIs on our new image
roiManager("Measure");
//Now we select the GFP image, in this case its the first image that we opened, so the list index is //"0"
selectWindow(list[0]);
//Again, overlay the ROIs onto the selected GFP image "0" and measure them
run("From ROI Manager");
roiManager("Measure");
//We want the title of the image in our filename, in this case the title of the image has the expt //details that we can parse out later with an R script
//so we use the following command to save the title of the image into a variable called fileName that we can use to name our CSV of the results file
fileName = getTitle();
saveAs("Results", output + fileName + ".csv");
//Close all image windows
close("*")
//Clear results
run("Clear Results");
3) Divide the GC150 signal by the mApple signal to get the GEDI ratio. (EXCEL or R script)
Use the following equation to determine the GEDI ratio threshold for cell death where “mean GEDI ratio dead” is the mean GEDI ratio from cells that have been given a stress stimulus according to Part 1: step 10 and mean GEDI ratio live is the GEDI ratio of the aforementioned cells before the stress stimulus:
GEDI ratio threshold = [(mean GEDI ratio dead) − (mean GEDI ratio live)] ∗ 0.25 + [mean GEDI ratio live]
4) Bind CSVs into one file and by following the attached R Markdown File. (R script)
BIND_FIJI_CSVs.html615KB Part IV: Cell tracking with TrackMate
Open TrackMate. Must have hyperstack open. Screenshots of steps found below.
Use the loG detector for spot detection adjusting the spot threshold accordingly.
Use the LAP tracker to generate tracks.
Screenshots of steps used for TrackMate.
Check calibration settings and crop settings for user's liking. Click next.
Identify that the LoG detector has been chosen. Click next.
Identify the optimal user-definded tracking settings by previewing the image. Once settings identified click next.
Once detection is complete. Click next.
Adjust thresholding of spots if necessary
User can set filters to image if there are spots that need to be removed that cannot be removed from adjusting the threshold.
Identify that the LAP tracker has been chosen. Click next.
Identify the optimal user-definded tracking settings by previewing the image. The max distance will need iteration to achieve optimal tracking before manual curation.
Identify the optimal user-definded tracking settings by previewing the image. Once settings identified click next.
Once detection is complete. Click next.
User can set filters to image if tracks need to be removed.
Representative image with spots and tracks.
Click the spots button.
Export file as a CSV. The CSV file will be used to identify the GEDI ratio by mean channel 1 divided by mean channel 2. The channel being divided depends on which channel is GC150. For our experiment, GC150 was channel 1.
Using annotated tracks and GEDI ratios generate survival curves based off the experimental parameters of interest.
Use the intensities quantified by the spot detector to make a new column titled GEDI ratio for the dataset of interest.
Ensure each track is validated either manually or using the numerous filters that are available in TrackMate. Type of filters to use will be highly dependent on user generated images.