Protocol for detection of ferroptosis in cultured cells

Summary

Mammalian cells can die by apoptosis or by one of several non-apoptotic mechanisms, such as ferroptosis. Here, we present a protocol to distinguish ferroptosis from other cell death mechanisms in cultured cells. We describe steps for seeding cells, administering mechanism-specific cell death inducers and inhibitors, and measuring cell death and viability. We then detail the use of molecular markers to verify mechanisms of cell death. This protocol can be used to identify and distinguish ferroptosis in 2D and 3D cultures.

For complete details on the use and execution of this protocol, please refer to Ko, et al. (2019),¹ Magtanong, et al. (2022),² and Armenta, et al. (2022).³

Graphical abstract

Highlights

• •Protocol to distinguish ferroptosis from other forms of cell death
• •Imaging and biochemical techniques to functionally assess ferroptosis induction
• •Distinguish cell death mechanisms using specific cell death inducers and inhibitors

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Mammalian cells can die by apoptosis or by one of several non-apoptotic mechanisms, such as ferroptosis. Here, we present a protocol to distinguish ferroptosis from other cell death mechanisms in cultured cells. We describe steps for seeding cells, administering mechanism-specific cell death inducers and inhibitors, and measuring cell death and viability. We then detail the use of molecular markers to verify mechanisms of cell death. This protocol can be used to identify and distinguish ferroptosis in 2D and 3D cultures.

Before you begin

The following protocol has been optimized for human HT-1080 fibrosarcoma cells, a standard ferroptosis model cell line. However, this protocol can be used to measure cell death in a variety of other adherent cell lines grown in either 2D or 3D culture. The goal of this protocol is to distinguish ferroptosis from apoptosis and a distinct form of non-apoptotic cell death induced by the small molecule caspase-independent lethal 56 (CIL56). The ability to unambiguously distinguish ferroptosis from other forms of cell death can be useful for exploring ferroptosis molecular mechanisms, testing the specificity of new candidate ferroptosis inducers and inhibitors, and exploring similarities and differences between ferroptosis and other forms of cell death.

In this protocol, we evaluate cytotoxicity using two methods: Scalable Time-lapse Analysis of Cell death Kinetics (STACK) and CellTiter-Glo. The STACK method uses fluorescent live and dead cell markers with high throughput time-lapse imaging to directly quantify the number of live and dead cells over time.⁴ Red fluorescent nuclear mKate2 protein is used to mark live cells, and the green fluorescent live-cell impermeable nucleic acid stain SYTOX Green is used to detect dead cells. Note that some in some dying cells plasma membrane integrity can be lost, allowing SYTOX Green uptake, but with mKate2 protein not yet fully lost from the nucleus. These "double-positive" cells are considered dead for the purposes of cell death quantification. Live and dead cell quantifications are then integrated into a ‘lethal fraction’ score, which represents normalized cell death within a population on a scale from 0 to 1. The STACK method requires engineered cell lines and a microscope that may not be readily accessible to all investigators. The second method, CellTiter-Glo, is a luminescence-based cell viability assay that uses a commercially available kit and common detection equipment. Cell viability is inferred by measuring total ATP content, an indicator of metabolically active cells. Either STACK or CellTiter-Glo can be used to evaluate cytotoxicity. However, it is important to note that cell viability is different than cell death, and that metabolic changes within a cell could confound measurements obtained using CellTiter-Glo indepenent of any change in cell viability. STACK allows for more direct detection of cell death.

In this protocol we also present imaging and biochemical methods that can be used to functionally assess whether a given cell death inducer has caused ferroptosis specifically, as opposed to apoptosis, or a distinct form of non-apoptotic cell death triggered by the small molecule CIL56.

Many studies of ferroptosis and other forms of cell death are conducted using cancer cells adherent to plastic surfaces (2D culture). However, growth of cancer cells in non-adherent conditions as spheroids (3D culture) may better recapitulate certain features of tumors in vivo. Thus, we present methods to detect and distinguish ferroptosis in both 2D and 3D cultures.

Generate and culture mKate2-expressing cells

Timing: 2 weeks

These steps describe the generation of cells that stably express nuclear mKate2 (denoted with a superscript ‘N’, i.e., HT-1080^N). Cells expressing mKate2 can be used with the STACK method to analyze cell death and quantify cell death kinetics.

• 1.Generate nuclear mKate2-expressing cells by lentiviral transduction.
• 2.Isolate mKate2-expressing cells with either antibiotic selection or fluorescence-activated cell sorting.

Note: A published protocol describes this viral transduction and cell selection procedure step-by-step.⁵

Note: Cell death responses could be affected by cell confluence and cell seeding density. Thus, it is important to subculture cells in a consistent and regular manner. When culturing HT-1080 or HT-1080^N cells, seed 6,000 cells per cm² every 3 d or 12,000 cells per cm² every 2 d. Aim for 90% confluence before splitting cells, but do not allow cells to ever reach 100% confluence. For other cell lines, these parameters may require optimization to ensure that an appropriate and consistent confluence is achieved prior to cell harvest and seeding.

Optimize cell death measurements

Timing: 4+ days

These steps will ensure that the chosen concentrations of cell death inducers cover a dynamic range of cell death responses and that the image analysis parameters are optimized for STACK.

• 3.
  Determine the optimal range of cell death inducer doses (see problem 1).

  • a.
    For each cell death inducer, treat cells with 10 different doses generated using two-fold serial dilution.

    • i.
      Determine the optimal treatment duration.

      Note: Between 48‒72 h is sufficient for many cell death inducers, but this should be determined experimentally for each chosen compound.

    • ii.
      Adjust the concentration range as needed to ensure the cell death responses span all regions of the sigmoidal dose-response curve when concentrations are plotted in their log₁₀ form.
    • iii.
      Ensure that the chosen range of doses is sufficient to calculate an EC₅₀ value.

      Note: Typically, at least two points at the top and the bottom of the dose-response curve are necessary for an effective curve fit.

• 4.If you are using an S3 Live Cell Analysis Instrument or similar imager, optimize an image analysis job.

Note: Details on image analysis parameters and optimization can be found in the Materials and equipment section as well as in a published STACK protocol.⁵ The analysis job is the defined protocol for automated counting of live mKate2-positive (mKate2⁺) and dead SYTOX-Green positive (SG⁺) “objects” (i.e., cells) for the chosen cell line.

Note: Optimization may require trial and error – adjust image analysis settings as needed to ensure that live and dead cells are counted accurately.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Cleaved Caspase-3 (Asp175) (5A1E) Rabbit mAb #9664	Cell Signaling Technology	Cat#9664T
Tubulin-α Ab-2 Mouse Monoclonal Antibody, Epredia	Fisher Scientific	Cat#MS581P1
IRDye 800CW Donkey anti-Rabbit IgG Secondary Antibody	LI-COR	Cat#926-32213
IRDye 680LT Donkey anti-Mouse IgG Secondary Antibody	LI-COR	Cat#926-68022
Bacterial and virus strains
Incucyte Nuclight Red (Lenti, EF-1 alpha, bleomycin)	Sartorius	Cat#4627
Chemicals, peptides, and recombinant proteins
BODIPY 581/591 C11 (Lipid Peroxidation Sensor)	Thermo Fisher Scientific	Cat#D3861
SYTOX Green Nucleic Acid Stain - 5 mM Solution in DMSO	Thermo Fisher Scientific	Cat#S7020
Erastin2	Cayman Chemical Company	Cat#27087
RSL3	Selleck Chemicals	Cat#S8155
Bortezomib	Fisher Scientific	Cat#NC0587961 (discontinued, see Thermo Scientific Cat#J60378-MA for an alternative)
Caspase-independent lethal 56 (CIL56)	Synthesized	Synthesized as previously described.6 Also available commercially through Cayman Chemical Company (Cat#19287)
Ferrostatin-1	MilliporeSigma	Cat#SML0583
Deferoxamine mesylate salt	MilliporeSigma	Cat#D9533
5-(Tetradecyloxy)-2-furoic acid (TOFA)	MilliporeSigma	Cat#T6575
Quinoline-val-asp-difluorophenoxymethylketone (Q-VD-OPh)	MedChemExpress	Cat#HY-12305
Critical commercial assays
CellTiter-Glo Luminescent Cell Viability Assay	Promega	Cat#G7572 https://www.promega.com/products/cell-health-assays/cell-viability-and-cytotoxicity-assays/celltiter_glo-luminescent-cell-viability-assay/?catNum=G7572
CellTiter-Glo 3D Cell Viability Assay	Promega	Cat#G9683 https://www.promega.com/products/cell-health-assays/cell-viability-and-cytotoxicity-assays/celltiter-glo-3d-cell-viability-assay/?catNum=G9683
Pierce BCA Protein Assay Kit	Thermo Fisher Scientific	Cat#23225 https://www.thermofisher.com/order/catalog/product/23225
Experimental models: Cell lines
HT-1080 cells	ATCC	Cat#CCL-121
HT-1080N cells	Forcina et al.4	N/A
Software and algorithms
Incucyte S3 Software (v2022A)	Sartorius	N/A
BioTek Gen5 Software	Agilent	N/A
Prism 9	GraphPad	N/A
Image Studio (v3.1)	LI-COR	N/A
Other
Incucyte S3 Live Cell Analysis Instrument	Sartorius	Cat#4647
Agilent BioTek Cytation 3 Cell Imaging Multimode Reader	Fisher Scientific	(Discontinued, see Cat#BTCYT5FW for a newer model)
Agilent BioTek Lionheart FX Automated Microscope	Fisher Scientific	Cat#BTLFXWSN
iBlot 2 Gel Transfer Device	Thermo Fisher Scientific	Cat#IB21001 (discontinued, see Cat#IB31001 for newer model)
Costar 6-Well Clear TC-treated Multiple Well Plates, Individually Wrapped, Sterile	Corning	Cat#3516
Corning Costar 12-well Clear TC-treated Multiple Well Plates, Individually Wrapped, Sterile	Fisher Scientific	Cat#07-200-82
Corning Costar 96-Well, Cell Culture-Treated, Flat-Bottom Microplate	Fisher Scientific	Cat#07-200-91
Corning 384-Well, Cell Culture-Treated, Flat-Bottom, Low Flange Microplate	Fisher Scientific	Cat#07-201-013
PrimeSurface 3D Culture Spheroid plates: Ultra-low Attachment (ULA) Plates	S-BIO	Cat#MS-9384UZ
Abgene 384 Well Polypropylene Sample Processing & Storage Plates	Thermo Fisher Scientific	Cat#AB0781
4× Bolt LDS Sample Buffer	Thermo Fisher Scientific	Cat#B0007
10× Bolt Sample Reducing Agent	Thermo Fisher Scientific	Cat#B0009

Materials and equipment

S3 Live Cell Analysis Instrument settings for live cell imaging and STACK

• •
  Settings used to acquire live cell images of 2D cultures:

  • ○
    Scan type: Standard.
  • ○
    Spectral unmixing:

    • -
      Red contributes to green = 8%.
    • -
      Green contributes to red = 0%.
• •
  Settings used to analyze images:

  • ○
    Channel: Green.

    • -
      Object Name: SYTOX Green.
    • -
      Segmentation: Adaptive.
    • -
      Threshold adjustment: 10 GCU.
    • -
      Edge split: On.
    • -
      Edge sensitivity: −5.
    • -
      Hole fill: 0 μm².
    • -
      Adjust size: 0 pixels.
    • -
      Minimum area: 40 μm².
    • -
      Maximum area: 2,000 μm².
  • ○
    Channel: Red.

    • -
      Object name: mKate2.
    • -
      Segmentation: Adaptive.
    • -
      Threshold adjustment: 2 RCU.
    • -
      Edge split: On.
    • -
      Edge sensitivity: −30.
    • -
      Hole fill: 2 μm².
    • -
      Adjust size: 0 pixels.
    • -
      Minimum area: 40 μm².
    • -
      Maximum area: 3,000 μm².
    • -
      Maximum eccentricity: 1.
    • -
      Minimum mean intensity: 10.
  • •
    Channels: Green + Red.

    • ○
      Object name: Yellow.
    • ○
      Minimum area: 100 μm².
    • ○
      Maximum area: 5,000 μm².
    • ○
      Minimum green mean intensity: 10.
    • ○
      Minimum red mean intensity: 10.

Note: Please refer to a published STACK protocol for more detail on adjusting the above parameters to optimize live and dead cell quantification.⁵ Note that optimized detection of live and dead cells may require trial and error and that creating an image analysis job that accurately identifies all live and dead objects may be challenging.

• •
  Settings used to acquire live cell images of 3D cultures:

  • ○
    Scan type: Standard or Spheroid.
  • ○
    Spectral unmixing:

    • -
      Red contributes to green = 12%.
    • -
      Green contributes to red = 0%.

BioTek Cytation 3 settings for CellTiter-Glo

Note: The CellTiter-Glo assay should be compatible with any plate reader that can perform luminescence measurements. The following parameters are optimized for the BioTek Cytation 3 instrument.

• •BioTek Gen5 software was used.
• •Gain: 135.
• •Integration time: 1 s.
• •Read height: 1 mm.

Note: These parameters were used for both 2D and 3D cultures.

LI-COR Odyssey CLx settings for western blotting

• •Image Studio software (version 3.1) was used to acquire images of nitrocellulose blots on a LI-COR Odyssey CLx instrument.
• •Images were acquired in both the 700 nm and 800 nm channels with auto-scaling enabled.
• •Acquisition resolution: 169 μm.
• •Quality: Medium.
• •Focal plane: 0.0 mm.
• •The image was set to automatically flip upon acquisition.
• •Images were quantified using the quantification system built into the Image Studio software.

BioTek Lionheart FX automated microscope settings for C11-BODIPY(581/591)

• •BioTek Gen5 software was used to acquire images of cells probed with C11-BODIPY(581/591) and stained with Hoechst on a BioTek Lionheart FX automated microscope.
• •
  The following filter cubes were used:

  • ○
    DAPI (excitation: 377/50 nm, emission: 447/60 nm).
  • ○
    Texas Red (excitation: 586/15 nm, emission: 647/57 nm).
  • ○
    GFP (excitation: 469/35 nm, emission: 525/39 nm).
• •Upon placement of the above filter cubes, auto-calibration was run using the default settings.
• •A 20× objective was used and the correction collar was set to 1.25 mm.
• •
  The following protocol was run for all wells containing samples:

  • ○
    Plate type: Costar 12-well.
  • ○
    Use lid: No.
  • ○
    Select wells: At runtime.
  • ○
    Temperature: Setpoint 37°C.
  • ○
    Image (20×): DAPI 377,447, Texas Red 586,647, GFP 469,525.
  • ○
    Define beacons: Yes.
  • ○
    Montage: Yes.
  • ○
    Montage: Rows × columns (2 × 2).
  • ○
    Regions of interest per image: 2.
  • ○
    Tile overlap: Auto for stitching.
• •
  The following parameters were used for the various filter cube channels:

  • ○
    DAPI 377,447:

    • -
      Illumination: 5.
    • -
      Integration time: 100.
    • -
      Gain: 0.
    • -
      Autofocus.
  • ○
    Texas Red 586,647:

    • -
      Illumination: 5.
    • -
      Integration time: 100.
    • -
      Gain: 5.
    • -
      Autofocus.
  • ○
    GFP 469,525:

    • -
      Illumination: 5.
    • -
      Integration time: 300.
    • -
      Gain: 10.
    • -
      Autofocus.
• •
  The following image stitching parameters were used for all three channels:

  • ○
    Fusion method: Linear blend.
  • ○
    Crop stitched image to remove black rectangles on the borders: Yes.
  • ○
    Fill gaps between montage tiles with local background color: No.
  • ○
    Downsize full image: No.
• •
  The following image preprocessing parameters were used for all three channels:

  • ○
    Apply image preprocessing: Yes.
  • ○
    Background flattening: Yes.
  • ○
    Auto: Yes.
  • ○
    Rolling ball diameter: 122 μm, 380 pixels.
  • ○
    Priority: Fast speed.
  • ○
    Image smoothing strength: None.
• •
  The following cellular analysis parameters were used:

  • ○
    Primary mask and count (all three channels):

    • -
      Auto: No.
    • -
      Value: 10,000.
    • -
      Background: Dark.
    • -
      Split touching objects: Yes.
    • -
      Fill holes in masks: Yes.
    • -
      Minimum object size: 5 μm.
    • -
      Maximum object size: 100 μm.
    • -
      Include primary edge objects: Yes.
    • -
      Analyze entire image: Yes.
  • ○
    Secondary mask (Texas Red and GFP only):

    • -
      Measure within a primary mask: Yes.
    • -
      Expand primary mask: 3 μm.
  • ○
    Calculated metrics:

    • -
      Cell count.
    • -
      Object mean(Tsf[Stitched[Texas Red 586,647]] (C11-BODIPY(581/591)_non-ox).
    • -
      Object mean(Tsf[Stitched[GFP 469,525]] (C11-BODIPY(581/591)_ox).
    • -
      C11-BODIPY(581/591) oxidation ratio (C11-BODIPY(581/591)_ox/C11-BODIPY(581/591)_non-ox).

Reagents

Note: Prepare the small molecule stock solutions such that the final concentration of DMSO in each well does not exceed 0.5%.

• •10 mM erastin2 solution: add 1 mg erastin2 to 160 μL DMSO.
• •10 mM RSL3 solution: add 5 mg RSL3 to 1.13 mL DMSO.
• •20 mM CIL56 solution: add 10 mg to 1.02 mL DMSO.
• •2 mM bortezomib solution: add 1 mg bortezomib to 1.30 mL DMSO.
• •100 mM quinoline-val-asp-difluorophenoxymethylketone (Q-VD-OPh) solution: add 10 mg Q-VD-OPh to 195 μL DMSO.
• •10 mM ferrostatin-1 solution: add 5 mg to 1.91 mL DMSO.
• •150 mM deferoxamine mesylate (hereafter referred to as deferoxamine) solution: add 100 mg deferoxamine to 1.02 mL DMSO.
• •20 mM 5-(tetradecyloxy)-2-furoic acid (TOFA) solution: add 10 mg to 1.54 mL DMSO.

HT-1080 growth medium recipe

Reagent	Final concentration	Amount
DMEM	88%	440 mL
Fetal bovine serum	10%	50 mL
Penicillin/streptomycin	1%	5 mL
Non-essential amino acids	1%	5 mL
Total	N/A	500 mL

Note: For small molecule storage conditions, keep at −20°C for no longer than is recommended by the manufacturer. Avoid freeze-thaws.

Note: For growth medium storage conditions, keep at 4°C for a maximum of 30 d.

CRITICAL: Some chemicals used in this study are potentially hazardous, as they are designed to modulate mammalian cell death and therefore may be toxic if they contact exposed skin or are ingested. For safest handling, work in a biosafety cabinet, wear eye protection, gloves, and a lab coat, and keep flasks and plates sealed whenever possible. Specific Occupational Health and Safety Administration (OSHA) hazard codes for lethal compounds used in this protocol are provided in table form (Table 1).

Table 1.

OSHA hazard codes for relevant compounds

Compound	OSHA hazard Communication standard codes
Erastin2	Not a hazardous substance or mixture.
RSL3	H302 Acute toxicity, oral. (Category 4) H400 Hazardous to the aquatic environment, acute hazard. (Category 1) H410 Hazardous to the aquatic environment, long-term hazard. (Category 1)
Bortezomib	H301 Acute toxicity, oral. (Category 3) H311 Acute toxicity, dermal. (Category 3) H330 Acute toxicity, inhalation. (Category 2) H372 Specific target organ toxicity, repeated exposure. (Category 1)
CIL56	Not a hazardous substance or mixture.
Ferrostatin-1	H315 Skin corrosion/irritation. (Category 2) H319 Serious eye damage/eye irritation. (Category 2A) H335 Specific target organ toxicity, single exposure; Respiratory tract irritation. (Category 3)
Deferoxamine	Not a hazardous substance or mixture.
TOFA	Not a hazardous substance or mixture.
Q-VD-OPh	H302 Acute toxicity, oral. (Category 4) H315 Skin corrosion/irritation. (Category 2) H319 Serious eye damage/eye irritation. (Category 2A) H335 Specific target organ toxicity, single exposure; Respiratory tract irritation. (Category 3)

Included are hazard classes and categories defined by OSHA for each cell death inducer and inhibitor used in this protocol.

Alternatives: For the assays described in this protocol, alternative cell death inducers or inhibitors could be used to achieve the same effects. It should not be assumed that every cell line will be sensitive to a given cell death inducer; preliminary investigations are always needed to determine whether a cell line of interest is sensitive to cell death triggered by a chosen compound.

• •Ferroptosis induction: Cystine depletion could be used instead of erastin2 treatment. The glutathione peroxidase 4 (GPX4) inhibitors ML162, ML210 or FIN56 could be used in place of RSL3.
• •Ferroptosis inhibition: The lipophilic radical trapping antioxidants ɑ-tocopherol or liproxstatin-1 could be used in place of ferrostatin-1.
• •Apoptosis induction: The BH3 mimetic ABT-737 or the topoisomerase I inhibitor camptothecin could be used instead of bortezomib.
• •Apoptosis inhibition: The pan-caspase inhibitor Z-VAD-FMK could be used in place of Q-VD-OPh.
• •CIL56-induced death induction: Other inducers of this cell death mechanism have yet to be discovered.
• •CIL56-induced death inhibition: The long-chain acyl-CoA synthetase (ACSL) family inhibitor triacsin C could be used as an alternative to TOFA.

Step-by-step method details

Quantification of cell death in 2D models

Timing: 4+ days

This section describes methods to seed cells, treat cells with cell death inducers and inhibitors, and use the STACK and CellTiter-Glo assays to measure cell death and viability, respectively. The results of these assays will demonstrate which cell death inhibitor(s) inhibit death caused by a cell death inducer of interest. Because this protocol focuses on ferroptosis, we used two ferroptosis inducers, erastin2 and RSL3, which act via distinct mechanisms. Note that erastin2 exhibits greater potency and ferroptosis induction than erastin, the analog from which erastin2 was derived.

• 1.
  Day 1: Seed cells (Figure 1A).

  • a.
    Design your plate layout.

    Note: A suggested 384-well plate layout used for the experiments in this protocol is depicted in Figure 2.

  • b.
    Determine the volume of growth medium required to seed the 384-well plate.

    • i.
      Volume = (20 μL per well × number of wells) + overage.
  • c.
    Collect cells from a T-75 flask into a 15 mL conical tube.
  • d.
    Determine the concentration of cells using a manual or automatic cell counter.
  • e.
    Keeping in mind the volume calculated in step 1(b)(i), use growth medium to dilute an appropriate volume of harvested cells to a concentration of 1,500 cells per 20 μL in a sterile trough.

    Note: This cell seeding density has been optimized for HT-1080^N cells and a treatment duration of 48 h.

  • f.
    Using a 384-well compatible multichannel pipette, seed 20 μL of the cell solution per well of a black, clear bottom 384-well plate.

    CRITICAL: If performing cell viability measurements with CellTiter-Glo, leave at least three wells empty. These wells will serve as blanks.

  • g.
    Spin the plate for 3 s at 58 × g to ensure cells adhere to the bottom of the wells.
  • h.
    Place the plate in a sterile 37°C CO₂ tissue culture incubator for 16–32 h.

    CRITICAL: Cell seeding densities could affect cell death sensitivity. It is therefore important to maintain consistency in the incubation time prior to cell death inducer treatment. The incubation should be kept consistent among replicates (e.g., 24 h).

• 2.
  Day 2: Treat cells with cell death inducers and inhibitors and perform live cell imaging (Figures 1A and 1B).

  • a.
    Prepare the cell death inducers.

    • i.
      Determine the volumes of growth medium and cell death inducer required to prepare a solution three times more concentrated than the desired highest dose.

      Note: As an example, our calculations are provided in the table below (Table 2).

      Note: Adjust volumes to avoid pipetting < 0.5 μL.

    • ii.
      For each cell death inducer, volume of growth medium = (number of cell death inhibitors × 100 μL) + overage.
    • iii.
      Volume of each cell death inducer = (concentration of desired highest dose × 3) × (volume of growth medium) ÷ concentration of the stock solution.
    • iv.
      Determine which cell death inducer solution requires the largest volume of stock solution (in this case, erastin2, RSL3, or CIL56) and add an equivalent volume of DMSO for the 3× DMSO solution.
    • v.
      In 1.5 mL Eppendorf tubes, prepare the 3× solutions of each cell death inducer.

      CRITICAL: If there is precipitate in a cell death inducer stock solution, ensure that the compound has been resuspended in the appropriate solvent, according to its known solubility.

    • vi.
      Using a multichannel pipette, add 50 μL of growth medium to columns 3‒11 and 14‒22 of a 384-well storage plate.
    • vii.
      Pipette 100 μL of each 3× cell death inducer solution (except DMSO) into the appropriate wells in columns 2 and 13.
    • viii.
      Using a multichannel pipette, pipette 50 μL of the 3× cell death inducer solution from column 2 into column 3 to achieve a two-fold dilution. Pipette up and down six times to mix.
    • ix.
      Repeat the previous step until you have reached column 11, which will result in a 10-dose serial dilution.
    • x.
      Repeat steps 2(a)(viii) and 2(a)(ix) with new pipette tips for columns 13‒22.
    • xi.
      Add 40 μL of the 3× DMSO solution to the appropriate wells in columns 12 and 23.
  • b.
    Prepare the cell death inhibitors.

    • i.
      Determine the volumes of growth medium and cell death inhibitor required to prepare a solution three times more concentrated than the desired final concentration.

      Note: As an example, our calculations are provided in the table below (Table 3).

      Note: Adjust volumes to avoid pipetting < 0.5 μL.

    • ii.
      For each cell death inhibitor, volume of growth medium = (number of cell death inducers × 11 wells) × (20 μL per well) + overage.
    • iii.
      Volume of each cell death inhibitor = (desired final concentration × 3) × (volume calculated in previous step) ÷ concentration of the stock solution.
    • iv.
      Determine which cell death inhibitor solution requires the largest volume of stock solution (in this case, deferoxamine) and add an equivalent volume of DMSO for the 3× DMSO solution.
    • v.
      If measuring cell death via STACK, make a 60 nM solution of SYTOX Green using growth medium (this is 3× the desired final concentration of 20 nM SYTOX Green).

      Note: Volume of 60 nM SYTOX Green solution = total volume of all cell death inhibitor solutions + overage.

    • vi.
      In small sterile troughs, prepare the 3× solutions of each cell death inhibitor.
  • c.
    Using a multichannel pipette, add 20 μL of each cell death inhibitor solution to the appropriate wells of the assay plate.
  • d.
    Similarly, pipette 20 μL of the cell death inducer solutions from the storage plate to the appropriate wells of the assay plate.
  • e.
    If using both STACK and CellTiter-Glo, add 60 μL 20 nM (1×) SYTOX Green solution per well to three wells without cells.

    Note: These wells will serve as blanks for the CellTiter-Glo cell viability assay.

    • i.
      If using only CellTiter-Glo, add 60 μL growth medium (without SYTOX Green).
    • ii.
      If using only STACK, skip this step.
    • iii.
      To make a 1× SYTOX Green solution, combine 100 μL 60 nM (3×) SYTOX Green solution with 200 μL growth medium.
  • f.
    Spin the assay plate down for 3 s at 58 × g.
  • g.
    Place the assay plate in a sterile 37°C CO₂ tissue culture incubator for 48 h.

    Note: If using cell death inducers not described in this protocol, it may be necessary to extend the treatment duration.

    • i.
      If using a Sartorius S3 Live Cell Analysis Instrument or similar imager to conduct STACK, place the assay plate in the machine housed within an incubator (Figure 1A).

      Note: Set up image acquisition and analysis using the parameters described in the Materials and equipment section as well as a published protocol for STACK.⁵ We recommend image acquisition at 4 h intervals, starting immediately.

      CRITICAL: Set the red channel acquisition time to 800 ms, the green channel acquisition to 400 ms, and the spectral unmixing parameter to 8% of red contributes to green, 0% of green contributes to red.

• 3.
  Day 4+: Measure cell viability and/or death.

  • a.
    If relevant, use the 2D CellTiter-Glo assay to measure cell viability upon completion of the treatment (Figure 1A).

    Note: The day prior, thaw the CellTiter-Glo Luminescent Cell Viability Assay reagent (hereafter referred to as CellTiter-Glo reagent) at 4°C.

    Note: The CellTiter-Glo reagent is relatively expensive. If desired, there are different approaches one can take to reduce the required volume of CellTiter-Glo. First, one could remove a greater volume from each well prior to the addition of CellTiter-Glo. For instance, if 45 μL medium were removed (leaving 15 μL behind), each well would require only 15 μL CellTiter-Glo reagent. This approach maintains the recommended 1:1 ratio of medium to CellTiter-Glo reagent. Alternatively, one could increase the ratio of medium to CellTiter-Glo reagent. For either approach, the method should be validated using an ATP standard curve to ensure that data integrity is not compromised.

    • i.
      30 min before the treatment is complete, warm the plate and CellTiter-Glo reagent to approximately 21°C.

      CRITICAL: If you are performing STACK in addition to CellTiter-Glo, do not remove the plate from the imaging machine until the final scan has been completed.

    • ii.
      Remove 35 μL from each well of the 384-well assay plate, including the three blank (cell-free) wells, leaving approximately 25 μL in each well.
    • iii.
      Add 25 μL CellTiter-Glo reagent to each well, including the three blank wells.
    • iv.
      Place the plate on an orbital shaker for 2 min to lyse the cells.
    • v.
      Place the plate on a benchtop and incubate at approximately 21°C for 10 min.
    • vi.
      Record luminescence using a microplate reader, using settings described in the materials and equipment section.
  • b.
    Analyze the CellTiter-Glo results.

    • i.
      Calculate the mean luminescence signal of the three blank wells.
    • ii.
      Subtract the mean blank signal (calculated in the previous step) from the luminescence signals of each experimental well.
    • iii.
      Calculate the mean luminescence signal of the control wells (wells for which DMSO served as the control for both the cell death inducer and cell death inhibitor).
    • iv.
      Divide the blank-corrected luminescence values of each well by the mean control signal (calculated in the previous step) to obtain the relative cell viability.

      CRITICAL: Look closely at the data to confirm that the cell viability quantifications match the observed cell phenotypes and to look for any potential issues (see problem 2).

    • v.
      Generate dose-response curves for each cell death inducer by log₁₀ transforming the cell death inducer concentrations and using the GraphPad Prism software ‘log(inhibitor) vs. response – Variable slope (four parameters)’ analysis to calculate the nonlinear regression best-fit curve (Figure 3A).
  • c.
    If relevant, analyze the STACK results.

    • i.
      Launch the image analysis using parameters described in the materials and equipment section.
    • ii.
      Copy and paste the mKate2⁺, SYTOX Green⁺, and Yellow⁺ (red and green overlap) object counts per mm² into a spreadsheet.
    • iii.
      Determine the ‘lethal fraction’ for each condition, which represents the proportion of dead cells in a population:

      $$\mathrm{L}\mathrm{e}\mathrm{t}\mathrm{h}\mathrm{a}\mathrm{l}\mathrm{f}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}=1-\frac{{mKate2}^{+}-Yellow}{\left({mKate2}^{+}-Yellow\right)+\max \left(SYTOX{Green}^{+}\right)}$$

      Note: Max(SYTOX Green⁺) refers to the maximum SYTOX Green⁺ count observed between times t = 0 and t = n, where n is the elapsed time in one time course experiment in a given population of cells. It is also possible to compute the lethal fraction at a single timepoint (e.g., 24 h or 48 h) using the SYTOX Green⁺ counts, but it is possible that some dead cells will be unaccounted for due to loss of SYTOX Green signal from long-dead cells.

      CRITICAL: Look at the images (e.g., Figure 3C) to confirm that the cell death quantifications match the observed cell phenotypes and to look for any potential issues (see problems 2 and 3).

  • d.
    Generate dose-response curves for each cell death inducer by log₁₀ transforming the cell death inducer concentrations and using the GraphPad Prism software ‘log(inhibitor) vs. response – Variable slope (four parameters)’ analysis to calculate the nonlinear regression best-fit curve (Figure 3A).
  • e.
    Generate time-course graphs for each cell death inducer at a chosen concentration, with time on the X-axis and lethal fraction on the Y-axis.

    Note: The GraphPad Prism software ‘Plateau followed by one phase association’ analysis can be used to calculate the best-fit curve. The calculated X₀ value represents the death onset (Figure 3B and Table 4).

    Note: For the ‘Plateau followed by one phase association’ analysis parameter X0, set the constraint type to ‘no constraint’.

    Note: Please refer to a published protocol for STACK for more detailed step-by-step instructions.⁵

Figure 1.

Cell viability and cell death assay workflows

(A) Schematic overview of the protocol workflow for 2D (top row) and 3D (bottom row) cultures. Cells are seeded in either flat or U-bottom plates. Cell death inducers and inhibitors are added the following day. Cell death and viability are measured by live cell imaging (including STACK) and CellTiter-Glo, respectively.

(B) Preparation of the cell death inducer dilutions. Prepare 10-point two-fold serial dilutions for each cell death inducer. Transfer 20 μL to the assay plate.

Figure 2.

Plate layout used in this protocol for cell viability and cell death studies

Cells were seeded into the wells colored orange, green, blue, and red. The wells depicted in gray were filled with growth medium only. The opacity of the well colors represents the concentration of the added cell death inducers (orange- erastin2, green- RSL3, blue- CIL56, red- bortezomib). Cells in columns 12 and 23 were treated with DMSO as the vehicle control. Cell death inhibitors (DMSO, deferoxamine, ferrostatin-1, Q-VD-OPh, and TOFA) were added to the indicated wells at a fixed concentration.

Table 2.

Example cell death inducer calculations for STACK and/or CellTiter-Glo

Cell death inducer	Stock concentration	Concentration of desired highest dose	3× concentration	Volume growth medium	Volume cell death inducer
DMSO	N/A	N/A	N/A	1,000 μL	1.5 μL
Erastin2	10 mM	0.005 mM	15 μM	800 μL	1.2 μL
RSL3	10 mM	0.005 mM	15 μM	800 μL	1.2 μL
CIL56	20 mM	0.01 mM	30 μM	800 μL	1.2 μL
Bortezomib	2 mM	0.0004 mM	1.2 μM	1,000 μL	0.6 μL

Included are calculations to compute the required volumes of cell death inducer stock solution and cell growth medium for the STACK and CellTiter-Glo experiments described in this protocol.

Table 3.

Example cell death inhibitor calculations for STACK and/or CellTiter-Glo

Cell death inhibitor	Stock concentration	Desired final concentration	3× concentration	Volume growth medium with SYTOX green	Volume cell death inhibitor
DMSO	N/A	N/A	N/A	2 mL	2 μL
Q-VD-OPh	100 mM	25 μM	75 μM	2 mL	1.5 μL
Ferrostatin-1	10 mM	1 μM	3 μM	2 mL	0.6 μL
Deferoxamine	150 mM	50 μM	150 μM	2 mL	2 μL
TOFA	20 mM	1 μM	3 μM	4 mL	0.6 μL

Included are calculations to compute the required volumes of cell death inhibitor stock solution and cell growth medium for the STACK and CellTiter-Glo experiments described in this protocol.

Figure 3.

Determination of cell death mechanism via cell viability and death assays in 2D cell culture

(A) Dose-response curves depicting HT-1080^N cell viability (left) and cell death (right) after 48 h of treatment with a cell death inducer (erastin2, RSL3, CIL56, bortezomib, or DMSO) and a cell death inhibitor (deferoxamine, ferrostatin-1, Q-VD-OPh, TOFA, or DMSO). Viability measurements were obtained using the CellTiter-Glo assay. Lethal fraction was calculated using STACK.

(B) Lethal fraction kinetics for erastin2 (era2), RSL3, CIL56, and bortezomib (btz) at the concentration used to approximate the EC₉₀. Cells were co-treated with cell death inhibitors or DMSO (vehicle control). Time elapsed indicates hours post-cell death inducer and inhibitor treatment.

(C) Images of HT-1080^N cells after 48 h of treatment with a cell death inducer and cell death inhibitor. Red color indicates mKate2-positive cells. Green color indicates SYTOX Green-positive cells. Scale bar = 50 μm. All data represent mean ± SD, N = 3 biological replicates.

Table 4.

Relevant values calculated from 2D STACK data

Cell death inducer	Cell death inhibitor	EC90 (M)	Death onset (h) at ∼EC90 compound dose	Nearest tested concentration used for death onset calculation
Erastin2	DMSO	3.4 × 10−7	15	3.1 × 10−7
RSL3	DMSO	4.8 × 10−6	3.1	5.0 × 10−6
CIL56	DMSO	1.9 × 10−6	10	2.5 × 10−6
Bortezomib	DMSO	3.8 × 10−8	13	5.0 × 10−8

Data obtained from 2D STACK analysis were used to compute the values provided in the table. The equation EC_F = EC₅₀ × (F ÷ (100-F))^1/H, where H is the Hill slope and F = 90, was used to calculate EC₉₀ values for cells treated with a cell death inducer (erastin2, RSL3, CIL56, or bortezomib) and DMSO as a vehicle control for the cell death inhibitor treatment condition. To approximate the death onset at the EC₉₀ concentration, we chose a concentration (from the dose-response) close to the calculated EC₉₀ value whose cell death induction was consistent with an EC₉₀ dose (i.e., generated 90% of the maximal observed death). The death onset for this approximate EC₉₀ concentration was then calculated as described in the quantification of cell death in 2D models section.

Quantification of cell death in 3D models

Timing: 4+ days

The Dixon lab has previously published methods for measuring cell viability in cells grown in 3D culture.³ This section describes these methods in greater detail.

• 4.
  Day 1: Seed cells (Figure 1A).

  • a.
    Design your plate layout.

    Note: A suggested 384-well plate layout used for the experiments in this protocol is depicted in Figure 2.

  • b.
    Determine the volume of growth medium required to seed the plate.

    • i.
      Volume = (20 μL per well × number of wells) + overage.
  • c.
    Harvest cells from a T-75 flask and obtain cell counts using a manual or automatic cell counter.
  • d.
    Keeping in mind the volume calculated in step 4(b)(i), use growth medium to dilute an appropriate volume of harvested cells to a concentration of 2,500 cells per 20 μL in a sterile trough.

    Note: This cell seeding density has been optimized for HT-1080^N cells and a treatment duration of 48 h.

  • e.
    Use a multichannel pipette to seed 20 μL of the cell solution into the appropriate wells of a 384-well ultra-low attachment U-bottom assay plate (Figure 1A).

    CRITICAL: If performing cell viability measurements with the 3D CellTiter-Glo Viability Assay, leave at least three wells empty, as these wells will serve as blanks.

  • f.
    Spin the plate down at 350 × g for 5 min.
  • g.
    Place the plate in a sterile 37°C CO₂ tissue culture incubator for 16–32 h.

    CRITICAL: Cell seeding densities could affect cell death sensitivity; it is therefore important to maintain consistency in the incubation time prior to cell death inducer treatment. The incubation should be kept consistent among replicates (e.g., 24 h).

• 5.
  Day 2: Treat cells with cell death inducers and inhibitors and perform live cell imaging (Figures 1A and 1B).

  CRITICAL: Prior to treatment with cell death inducers and inhibitors, check cells under the microscope to ensure that they have formed spheroids.

  • a.
    Prepare the cell death inducers and inhibitors as described in the previous section for 2D cell culture (Figure 1B).
  • b.
    Add 20 μL of the 3× cell death inhibitors and 20 μL of the 3× cell death inducers (prepared in 3× SYTOX Green medium) to the assay plate to obtain a final concentration of 1× for all compounds (Figures 1A and 1B).

    • i.
      Omit SYTOX Green if live cell imaging will not be used to assess cell death.
  • c.
    If using 3D CellTiter-Glo to assess cell viability, add 60 μL of 20 nM (1×) SYTOX Green solution per well in three wells without cells.

    Note: These wells will serve as blanks for the 3D CellTiter-Glo assay.

  • d.
    If monitoring cell death by live cell imaging, place the assay plate into the Sartorius S3 Live Cell Analysis Instrument or similar imager housed within a tissue culture incubator (Figure 1A). Otherwise, carefully place the assay plate into a sterile 37°C CO₂ tissue culture incubator for 48 h.

    CRITICAL: For live cell imaging, set the red channel acquisition time to 800 ms, the green channel acquisition to 400 ms, and the spectral unmixing parameter to 12% of red contributes to green, 0% of green contributes to red.

    Note: If using cell death inducers not described in this protocol, it may be necessary to extend the treatment duration.

• 6.
  Day 4+: Measure cell viability using the CellTiter-Glo 3D Cell Viability Assay (Figure 1A).

  Note: The day before, thaw an aliquot of CellTiter-Glo 3D Cell Viability Assay reagent (hereafter referred to as 3D CellTiter-Glo reagent) at 4°C.

  Note: The 3D CellTiter-Glo reagent is relatively expensive. If desired, there are different approaches one can take to reduce the required volume of 3D CellTiter-Glo reagent per well. First, one could remove a greater volume from each well prior to the addition of 3D CellTiter-Glo. For instance, if 40 μL medium was removed (leaving 20 μL behind), each well would require only 20 μL 3D CellTiter-Glo reagent. This approach maintains the recommended 1:1 ratio of medium to 3D CellTiter-Glo reagent. However, removing a greater volume of medium from each well may increase the risk of spheroid disruption. Alternatively, one could increase the ratio of medium to 3D CellTiter-Glo reagent. For either approach, the method should be validated using an ATP standard curve to ensure that data integrity is not compromised.

  • a.
    30 min before the treatment is complete, warm both the 3D CellTiter-Glo reagent and the plate to approximately 21°C.
  • b.
    Remove 30 μL from each well of the 384-well assay plate, including the three blank, cell-free wells. Approximately 30 μL should remain in each well.

    CRITICAL: Be very careful to not disturb or pipette up the spheroids. While spheroids are visible to the naked eye, using a light to illuminate the bottom of the plate can improve spheroid visualization.

  • c.
    Thoroughly mix the 3D CellTiter-Glo reagent by inversion.
  • d.
    Add 30 μL of the 3D CellTiter-Glo reagent to each well of the assay plate, including the three blank wells.
  • e.
    Cover the plate with foil and shake on an orbital shaker for 5 min to lyse the cells.
  • f.
    Incubate the plate on a benchtop at approximately 21°C for 25 min.
  • g.
    Record luminescence using a microplate reader, using the settings described in the materials and equipment section.
  • h.
    Analyze the CellTiter-Glo results as described above for 2D culture (see Figure 4A and problem 2).

Figure 4.

Determination of cell death mechanism in 3D cell culture

(A) Dose-response curves depicting HT-1080^N spheroid cell viability after 48 h of treatment with a cell death inducer (erastin2, RSL3, CIL56 or bortezomib) and a cell death inhibitor (deferoxamine, ferrostatin-1, Q-VD-OPh or TOFA). Viability measurements were obtained using the 3D CellTiter-Glo assay.

(B) Images of HT-1080^N spheroids after 48 h of treatment with a cell death inducer and cell death inhibitor. Red color indicates mKate2-positive cells. Green color indicates SYTOX Green-positive cells. Scale bar = 400 μm. DFO: deferoxamine, Fer-1: ferrostatin-1. All data represent mean ± SD, N = 3 biological replicates.

Validation 1: Western blotting for cleaved caspase-3

Timing: 5+ days

Caspase-3 and caspase-7 are cleaved during apoptosis but not ferroptosis.^7,6 Probing for cleaved caspase-3 via western blot can distinguish apoptotic and non-apoptotic cell death. Q-VD-OPh is a pan-caspase inhibitor that serves as a control because it inhibits apoptotic, but not other forms, of cell death.

Note: The following protocol has been optimized for HT-1080^N cells grown in 2D culture and for a treatment period of 48 h.

CRITICAL: Before conducting validation experiments, it is important to determine the optimal dose and treatment duration for each cell death inducer. Because cell death is a destructive process, it is important to choose a dose and a time point at which the cells are on the verge of dying (to ensure that a given lethal mechanism is fully activated) but have not yet undergone membrane permabilization. To this end, use the kinetic dose-response data obtained from the STACK experiment to define ideal timepoints for sample collection. Specifically, for each cell death inducer, plot the dose-response data at 48 h in GraphPad Prism and fit a nonlinear regression curve (see Quantification of cell death in 2D models section, step 3(d)). Use the resulting computed values (EC₅₀ and Hill slope) to determine the EC₉₀ dose. Next, plot the kinetic data for one or more concentrations near the EC₉₀. Use the GraphPad Prism ‘Plateau followed by one-phase association’ model to generate a model fit to the kinetic data (see Quantification of cell death in 2D models section, step 3(e)). From this model one can extract the X₀ value, which represents the death onset. For the validation experiments, select a time point approximately 30 min before death onset for each cell death inducer. Example values used for the experiments in this protocol are listed in Table 4.

• 7.Seed 150,000 cells in 2 mL growth medium per well of a 6-well plate.

Note: Seed enough wells to account for all conditions. For every apoptosis inducer, include two wells: one for the cell death inducer alone and one for co-treatment with a pan-caspase inhibitor. Additionally, include one well for treatment with the pan-caspase inhibitor alone. If a preliminary experiment results in the detection of cleaved caspase-3 in any wells treated with a cell death inducer not previously known to induce apoptosis, repeat the experiment and include a condition in which cells are co-treated with the cell death inducer and the pan-caspase inhibitor. The results should indicate whether the detected cleaved caspase-3 signal is due to unreported pro-apoptotic activity of the cell death inducer.

• 8.Allow the cells to adhere and proliferate for 2 d total (including the cell death inducer treatment).
• 9.
  Treat the cells with each cell death inducer for a period equal to the death onset minus 30 min (see Table 4 for death onset values).

  • a.
    Aspirate medium from each well and replace with 2 mL of the appropriate cell death inducer (and, if applicable, inhibitor) solution.
  • b.
    Stagger each treatment such that all treatments end 48 h after cell seeding.
• 10.After the treatments are complete, place the plate on ice.
• 11.Remove the growth medium from each well and add 1 mL of ice-cold PBS.
• 12.Remove the PBS and add another 1 mL of ice-cold PBS.
• 13.Scrape the cells into the PBS with a cell scraper.
• 14.Collect the cell solutions into 1.5 mL Eppendorf tubes. Centrifuge at 300 × g for 5 min.
• 15.
  During the centrifugation step, prepare the lysis buffer:

  • a.
    Make enough RIPA buffer (0.1% SDS) with protein and phosphatase inhibitors (5 mM of NaF phosphatase inhibitor and 1:200 P8340 protease inhibitor cocktail) for 100 μL per sample.
• 16.Remove the PBS from the cell pellets.

Pause point: The cell pellets can be frozen at −80°C.

• 17.Resuspend the cell pellets in 100 μL RIPA cocktail (see step 15(a)).
• 18.Sonicate the samples at 60% amplitude (1 s on, 1 s off) for 10 cycles.
• 19.Centrifuge the samples at maximum speed for 15 min at 4°C.
• 20.Transfer the supernatants to new Eppendorf tubes.
• 21.Quantify the protein concentration using a BCA assay.
• 22.Prepare samples containing 25 μg protein each. Add 4× Bolt loading dye and 10× reducing agent to achieve final concentrations of 1×.
• 23.Boil the samples for 5 min at 95°C.
• 24.Load 80% of the sample (20 μg protein) into a polyacrylamide gel.
• 25.Run the gel at 100 V for 1 h 45 min.
• 26.Use the iBlot 2 Gel Transfer Device to transfer the proteins from the gel to a nitrocellulose membrane.
• 27.
  Probe the membrane in a hybridization bag with a 1:1,000 dilution of rabbit anti-cleaved caspase-3 antibody and a 1:2,000 dilution of mouse anti-tubulin antibody in Intercept Blocking Buffer.

  • a.
    Incubate at 4°C on an orbital shaker for 16–32 h.
• 28.The following day, wash the membrane three times (5 min each) at approximately 21°C with 1× TBST.
• 29.
  Add a 1:15,000 dilution each of donkey anti-rabbit IgG antibody and donkey anti-mouse IgG antibody in Intercept Blocking Buffer.

  • a.
    Incubate at approximately 21°C for 1 h on an orbital shaker in a hybridization box.
• 30.Wash the membrane three times (5 min each) at approximately 21°C with 1× TBST.
• 31.Quickly rinse the membrane once with deionized water and store the membrane in deionized water until imaging.
• 32.Image the membrane using a LI-COR imager in the 700 nm and 800 nm channels to detect cleaved caspase-3 and tubulin signals, respectively (see Figure 5A and problem 4).
• 33.Export the images for each channel in black and white.
• 34.Use an image analysis software to measure the cleaved caspase-3 signals for each condition. Normalize these values to the respective tubulin signals by division (Figure 5B).

Alternatives: This western blot protocol was developed using the following reagents and equipment: RIPA lysis buffer, Pierce BCA assay, Bolt gel running system, iBlot 2 dry transfer system, and LI-COR imager. While other methods could also be used to detect cleaved caspase-3, it is important to optimize each step that varies from this established protocol.

Figure 5.

Western blotting for cleaved caspase-3

(A) Representative western blot of HT-1080^N cells treated with cell death inducers, the pan-caspase inhibitor Q-VD-OPh, or the vehicle control DMSO. For the apoptosis inducer bortezomib, cells were either co-treated with bortezomib and DMSO or bortezomib and Q-VD-OPh. Blots were probed for tubulin (loading control) and cleaved caspase-3. Asterisk indicates a non-specific band that occurs in conditions of apoptosis inducer and pan-caspase inhibitor co-treatment, as reported previously.⁸

(B) Quantification of (A). Cleaved caspase-3 signal was normalized to tubulin signal. Asterisk indicates p < 0.05 for a t-test for difference of means between each condition and the bortezomib alone condition (positive control). All data represent N = 3 biological replicates.

Validation 2: C11-BODIPY(581/591) imaging for lipid peroxidation

Timing: 3+ days

Because lipid peroxidation is a hallmark of ferroptosis, C11-BODIPY(581/591) is useful for the identification of ferroptosis. C11-BODIPY(581/591) is a fluorescent fatty acid analog used to detect lipid peroxidation.⁹ Oxidation of the unsaturated portion of the molecule alters its emission wavelength. The unoxidized molecule emits in the red spectrum and the oxidized molecule emits in the green spectrum; this shift can be detected using laboratory techniques such as flow cytometry or fluorescence microscopy, as described in this protocol. Because of overlapping emmissions it is recommended that BODIPY assays be performed using cells that do not express nuclear-localized mKate2.

Note: The following protocol has been optimized for HT-1080 cells grown in 2D culture and for a treatment period of 48 h.

• 35.Seed 50,000 cells in 1 mL growth medium per well of a 12-well plate.

Note: Each condition requires one well. The number of conditions is dependent on the number of cell death inducers and cell death inhibitors used in this experiment. For ferroptosis inducers, include one well with a ferroptosis inhibitor and one without. In addition, reserve a well for the ferroptosis inhibitor alone. If a preliminary experiment results in the detection of a high oxidation ratio in any wells treated with a cell death inducer not previously known to induce ferroptosis, repeat the experiment and include a condition in which cells are co-treated with the cell death inducer and the ferroptosis inhibitor. The results should indicate whether the detected oxidation is due to unreported pro-ferroptotic activity of the cell death inducer.

• 36.Allow the cells to adhere and proliferate for a total of 2 d (including the cell death inducer treatment).
• 37.
  Treat the cells with each cell death inducer for a period equal to the death onset minus 30 min (see Table 4 for death onset values).

  • a.
    Aspirate medium from each well and replace with 1 mL of the appropriate cell death inducer (and, if applicable, inhibitor) solution.
  • b.
    Stagger each treatment such that all treatments end 48 h after cell seeding.
• 38.
  30 min before the treatments are complete, prepare the C11-BODIPY(581/591) and Hoechst solution in PBS.

  • a.
    Prepare enough PBS for 500 μL per condition.
  • b.
    Add C11-BODIPY(581/591) for a final concentration of 1.5 μM.
  • c.
    Add Hoechst for a final concentration of 1 μg/mL.
• 39.Approximately 15 min before the treatments end, remove the growth medium from the wells.
• 40.Add 500 μL of the C11-BODIPY(581/591) and Hoechst solution to each well.

CRITICAL: For these steps, work in a dark, sterile biosafety cabinet (the cabinet light should be off, but lights in the room may be left on). This will ensure sterility and prevent photobleaching of the light-sensitive compounds.

CRITICAL: The timing of the C11-BODIPY(581/591) experiment is crucial. An insufficient cell death inducer treatment duration could result in a lack of oxidized lipid detection in conditions that should induce lipid peroxidation. An excessive cell death inducer treatment duration could lead to cell death. Work reasonably quickly and ensure that the images are obtained at precisely the intended time point (calculated death onset minus 30 min).

• 41.Incubate the plate in a sterile 37°C incubator for 10 min.
• 42.Remove the C11-BODIPY(581/591) and Hoechst solution. Add 500 μL pre-warmed PBS to each well.
• 43.Image cells on a Lionheart instrument or similar imager using Texas Red, DAPI, and GFP filter cubes (see Figure 6A and problem 5).
• 44.Calculate the C11-BODIPY(581/591) oxidation ratio for each condition (Figures 6B and 6C).

Note: See the materials and equipment section for details on image acquisition and calculation of the C11-BODIPY(581/591) oxidation ratio.

Figure 6.

C11-BODIPY(581/591) imaging for lipid peroxidation

(A) Representative images of HT-1080 cells treated with cell death inducers, the lipophilic antioxidant ferrostatin-1 (Fer-1), or the vehicle control DMSO. For the ferroptosis inducers erastin2 and RSL3, cells were either co-treated with the ferroptosis inducer and DMSO or the ferroptosis inducer and ferrostatin-1. Scale bar = 20 μm.

(B) Quantification of the C11-BODIPY(581/591) oxidation ratio for a representative biological replicate depicted in (A).

(C) Mean C11-BODIPY(581/591) oxidation ratio for three biological replicates. Single asterisk indicates p < 0.05 and double asterisks indicate p < 0.01 for a t-test for difference of means between the two conditions for each given comparison.

Expected outcomes

Figure 3, Figure 4, Figure 5, Figure 6 and Table 4 highlight data obtained using the above protocol. The data trends represent expected outcomes for experiments in which apoptosis, ferroptosis, and CIL56-induced death are induced in HT-1080 and HT-1080^N fibrosarcoma cells.

Figures 3 and 4 provide examples of data obtained using the CellTiter-Glo and STACK assays. The lipophilic antioxidant ferrostatin-1 and the iron chelator deferoxamine are well-characterized inhibitors of ferroptosis. Ferrostatin-1 and deferoxamine should therefore maintain HT-1080^N cell viability (Figures 3A and 4A) and suppress cell death (Figures 3A–3C and 4B) in cells treated with the ferroptosis inducers erastin2 and RSL3. It should, however, be noted that deferoxamine-mediated iron chelation can itself cause cell cycle arrest and cell death which can sometimes confound the interpretation of ferroptosis, especially when using bulk metabolic cell viability readouts like CellTiter-Glo.

Q-VD-OPh, a known inhibitor of apoptosis, should maintain cell viability in cells treated with the apoptosis inducer bortezomib (Figure 3A and 4A) and suppress bortezomib-induced cell death (Figures 3A–3C and 4B). In 3D culture, Q-VD-OPh appears less effective at suppressing bortezomib-induced death according to CellTiter-Glo cell viability measurements as compared to live cell imaging. This discrepancy could be attributed to the fact that CellTiter-Glo measures ATP levels as a proxy for cell viability. While live cell imaging clearly demonstrates suppression of bortezomib-induced death by Q-VD-OPh, it could be that Q-VD-OPh does not inhibit metabolic changes induced by bortezomib, which could be reflected in the CellTiter-Glo results. Finally, note that caspase activation is a downstream event in the execution of apoptotic cell death, and that Q-VD-OPh will not revert proteasome inhibition, the accumulation of ubiquitinated proteins or other upstream biochemical changes associated with proteasome inhibition. Thus, prolonged incubation with bortezomib, even in the presence of Q-VD-OPh, will likely eventually result in caspase-independent cell death and membrane permeabilization.

TOFA, an inhibitor of CIL56-induced death, should maintain viability in CIL56-treated cells (Figures 3A and 4A) and suppress death induced by CIL56 (Figures 3A–3C and 4B). As CIL56-induced death is a relatively recently discovered mechanism of cell death, no studies to date have used TOFA to inhibit CIL56-induced death in 3D cultures. Thus, note that 1 μM TOFA, which effectively suppresses CIL56-induced death in 2D culture (Figures 3A–3C), may not be sufficient to inhibit CIL56-induced death in 3D culture (Figure 4B).

The data presented in Figure 5 demonstrate the use of western blotting to quantify levels of cleaved caspase-3. Inducers of apoptotic cell death (but not non-apoptotic cell death pathways) should result in cleavage of caspase-3. As such, a western blot analysis of cells treated with the apoptosis inducer bortezomib 30 min before death onset should show cleaved caspase-3 signal (Figures 5A and 5B). This signal should be abolished when bortezomib-treated cells are co-treated with the pan-caspase inhibitor Q-VD-OPh (Figures 5A and 5B).

Figure 6 illustrates data obtained using the C11-BODIPY(581/591) probe for lipid peroxidation. Ferroptosis is characterized by overwhelming membrane lipid peroxidation, which can be detected using the probe C11-BODIPY(581/591). Ferroptosis inducers should exhibit a high ratio of oxidized (green) to non-oxidized (red) probe signal, whereas non-ferroptotic cell death inducers should exhibit lower green to red signal ratios (Figures 6A–6C). Ferrostatin-1 should suppress lipid peroxidation and thereby lower the green to red ratio in cells treated with the ferroptosis inducers erastin2 and RSL3 (Figures 6A–6C).

Quantification and statistical analysis

Data obtained from STACK-based analysis of HT-1080^N cells grown in 2D culture is provided (Table 4).

Limitations

While this protocol can be applied to diverse cell lines, certain limitations exist. First, some cell lines may be resistant to specific cell death inducers or cell death mechanisms and will therefore be inappropriate for these assays. Second, STACK has not been optimized for use with cells in suspension or spheroids; therefore, this protocol does not describe a method to directly measure cell death in suspension or 3D culture conditions. Finally, this protocol presents validation methods for cells grown only in 2D cell culture due to increased technical difficulty associated with western blotting and C11-BODIPY(581/591) imaging using cells grown in 3D culture.

Troubleshooting

Problem 1

Cells are not dying in response to a cell death inducer (see before you begin, optimize cell death measurements section, step 3).

Potential solution

Confirm that the cell death inducer is functional (i.e., can cause cell death) by testing its efficacy with cell lines known to be sensitive to the specific compound. For example, HT-1080 cells are known to be highly sensitive to the induction of ferroptosis by erastin2.³ It is also possible that the compound is degraded. To avoid this, purchase fresh reagents and avoid freeze-thaws by aliquoting cell death inducers into small volumes sufficient for just one experimental replicate. Alternatively, it could be that the chosen cell line is resistant to death induced by the specific small molecule, or even more generally resistant to the cell death mechanism induced by the small molecule (e.g., ferroptosis).

Problem 2

Unexpected cell death in the outermost wells of the assay plate (see quantification of cell death in 2D models section, step 3(b)(iv) and step 3(c)(iii), and quantification of cell death in 3D models section, step 6(h)).

Potential solution

Growth medium can evaporate more quickly from the edge wells of a plate, which could cause cells to dry out. Evaporation could also lead to an increase in the concentrations of added compounds due to a decrease in volume. If possible, avoid using the outermost wells at the top, bottom, and sides of the assay plate. Instead, fill the edge wells with growth medium or PBS. Additionally, ensure that the tissue culture incubator is properly sealed and that the water bath housed within the incubator is full. This will help ensure proper humidity within the incubator, thereby reducing evaporation.

Problem 3

Loss of nuclear mKate2 detection in some cells (see quantification of cell death in 2D models section, step 3(c)(iii)).

Potential solution

Infrequently, mKate2 expression is lost in a subpopulation of cells within one culture. This is likely due to genetic silencing of the mKate2 transgene. This issue typically occurs over months-long timescales. Should this problem occur, fluorescence-activated cell sorting or antibiotic selection could be used to re-isolate the sub-population of cells with the highest detected red fluorescence (and inferred highest mKate2 expression). Stocks of the sorted cells can then be used immediately or frozen for future use, as needed.

Problem 4

Lack of signal in cleaved caspase-3 western blot (see validation 1: Western blotting for cleaved caspase-3 section, step 31).

Potential solution

Biological or technical reasons could underlie a lack of cleaved caspase-3 signal in the positive control (apoptosis inducer) condition. Possible biological explanations include: the chosen apoptosis inducer dose is too low, the treatment is too short (no cleavage of caspase-3 has yet occurred), or the treatment is too long (causes cell death). Try using a higher concentration of the cell death inducer (such as the EC₉₅ dose rather than the EC₉₀) or selecting a different time point (such 15 min before the calculated death onset instead of 30 min). Technical complications could also underlie the lack of cleaved caspase-3 signal. For example, the primary antibody could have expired. Alternatively, the transfer of proteins to the membrane could have been incomplete, in which case, the parameters of the transfer protocol should be adjusted.

Problem 5

No oxidation of the C11-BODIPY(581/591) probe in cells treated with a ferroptosis inducer (see validation 2: C11-BODIPY(581/591) imaging for lipid peroxidation section, step 42).

Potential solution

Similar to Problem 4, there could be biological or technical explanations for the lack of oxidized C11-BODIPY(581/591) signal in the positive control (ferroptosis inducer) condition. It may be necessary to adjust the concentration (e.g., use the EC₉₅ dose) or duration of the ferroptosis inducer treatment. Technical issues could also underlie this problem. The C11-BODIPY(581/591) probe can degrade after repeated freeze-thaws. To avoid probe degradation, prepare aliquots of C11-BODIPY(581/591) sufficient for just one experimental replicate each. Additionally, ambient light could cause photobleaching of the probe; it may therefore be necessary to work in darker conditions and wrap the assay plate in foil for its transfer from the tissue culture hood to the imager.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Scott Dixon (sjdixon@stanford.edu).

Materials availability

This study did not generate new, unique reagents.

Data and code availability

There are no additional data associated with this protocol.