Characterization of MLKL-mediated Plasma Membrane Rupture in Necroptosis

Summary

We report methods for characterization of MLKL-mediated plasma membrane rupture in necroptosis including conventional and confocal live-cell microscopy imaging, scanning electron microscopy, and NMR-based lipid binding.

Abstract

Necroptosis is a programmed cell death pathway triggered by activation of receptor interacting protein kinase 3 (RIPK3), which phosphorylates and activates the mixed lineage kinase-like domain pseudokinase, MLKL, to rupture or permeabilize the plasma membrane. Necroptosis is an inflammatory pathway associated with multiple pathologies including autoimmunity, infectious and cardiovascular diseases, stroke, neurodegeneration, and cancer. Here, we describe protocols that can be used to characterize MLKL as the executioner of plasma membrane rupture in necroptosis. We visualize the process of necroptosis in cells using live-cell imaging with conventional and confocal fluorescence microscopy, and in fixed cells using electron microscopy, which together revealed the redistribution of MLKL from the cytosol to the plasma membrane prior to induction of large holes in the plasma membrane. We present in vitro nuclear magnetic resonance (NMR) analysis using lipids to identify putative modulators of MLKL-mediated necroptosis. Based on this method, we identified quantitative lipid-binding preferences and phosphatidyl-inositol phosphates (PIPs) as critical binders of MLKL that are required for plasma membrane targeting and permeabilization in necroptosis.

Introduction

Identifying genetic components of necroptosis has facilitated the use of animal models to test the implication of necroptosis in physiology and disease^1,2,3,4,5. Knockout of RIPK3 or MLKL in mice had minimal implication in development and adult homeostasis suggesting that necroptosis is not essential for life^3,6. Moreover, certain species do not contain either RIPK3 or MLKL genes, supporting the non-essential role of necroptosis in animals^7,8. On the other hand, challenging knockout animal models with various pathologies induced in the laboratory has revealed an important role of necroptosis in inflammation, innate immunity, and viral infection^9,10,11,12.

Necroptosis can be activated in several ways by signaling through different innate immunity sensors, all of which result in the activation of RIPK3^1,13,14. Active RIPK3 in turn phosphorylates and activates MLKL^{3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18}. The most studied, and perhaps the most complex way, leading to activation of RIPK3 involves death receptor ligation, which bifurcates based on the downstream composition of the signaling complexes to either induce apoptosis or necroptosis¹. Necroptosis ensues when signaling through RIPK1 is favored and results in engagement of RIPK3^19,20. This outcome is easily favored upon pharmacological inhibition or genetic deletion of caspase 8, a putative endogenous inhibitor of necroptosis that keeps necroptosis at bay. RIPK1 binds to and activates RIPK3. Another way to activate necroptosis is through Toll-like receptors TLR3/TLR4 signaling, which engages and activates RIPK3 through TIR-domain-containing adapter-inducing interferon-β (TRIF)²¹. Yet another way to die by necroptosis is by activation of the DNA sensor DAI, which directly engages and activates RIPK3²².

MLKL is a cytosolic protein comprised of an N-terminal helix bundle (NB) domain and a C-terminal pseudokinase domain (psKD) linked by a regulatory brace region³. In normal cells, MLKL is found in the cytosol where it is thought to be in an inactive complex with RIPK3¹⁴. Activation of necroptosis triggers RIPK3 phosphorylation of MLKL in the activation loop of the psKD, and potentially additional sites in the NB and brace^3,15,23. Phosphorylation induces a conformational change in MLKL that results in dissociation from RIPK3¹⁴. Poorly understood conformational changes release the brace from the psKD²⁴. The brace, which contains 2 helices, mediates oligomerization of MLKL into a putative trimer through the C-terminal helix²⁵. The N-terminal helix of the brace inhibits the NB domain, which is essential for membrane permeabilization^24,26. In isolation, NB domain is sufficient to induce plasma membrane permeabilization and necroptosis^16,24,27. The pro-necroptotic activity of NB was reconstituted in mouse embryonic fibroblasts deficient in MLKL (mlkl^-/- MEFs). NB is a lipid binding domain that preferentially engages the phospholipid phosphatidylinositol 4,5 diphosphate (PIP₂). We proposed a stepwise mechanism of activation of MLKL, wherein brace oligomerization facilitates recruitment of MLKL to the plasma membrane via weak interactions of NB with the PIP₂ polar head group²⁴. At the membrane, the NB undergoes regulated exposure of an additional high-affinity binding site for PIP₂, which is masked by the brace in inactive MLKL. Overall, the multiple interactions of NB with PIP₂ destabilize the plasma membrane leading to its rupture, although the molecular mechanism of these events have not been elucidated.

Here we illustrate specific methods used to characterize the function of MLKL as executioner of necroptosis²⁴. In particular, we focus on the most minimal domain of MLKL, the NB and brace (NBB), which is regulated by brace inhibition and can be activated through enforced dimerization to induce plasma membrane rupture and necroptosis. We describe our inducible expression system combined with enforced drug-induced FKBP-mediated dimerization for live-cell imaging, and electron microscopy of cells undergoing necroptosis. Additionally, we illustrate our in vitro NMR analysis of the interactions of NBB with phosphatidylinositols (PIPs).

Protocol

1. Cloning and Cell Line Generation

1. PCR amplify the NBB region, corresponding to amino acid residues 1-140 (NBB₁₄₀), from human MLKL cDNA for in frame standard restriction enzyme-based cloning with the oligomerization domain 2x FK506 binding protein (2xFKBP or 2xFV) and Venus fluorescent protein into the Doxycycline (Dox)-inducible retroviral vector pRetroX-TRE3G to obtain NBB₁₄₀-2xFV-Venus (Table 1, Figures 1A-B).
2. Immortalize primary mlkl^-/- or ripk3^-/- mlkl^-/- mouse embryonic fibroblasts (MEFs) obtained from the respective mice available in the Green laboratory by transient transfection with SV40-large antigen expressing plasmid. Select immortalized cells in ~1–2 weeks, and maintain in DMEM supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 U/mL penicillin and streptomycin, 1 mM sodium pyruvate, and nonessential amino acids at 37 °C and 5% CO₂ in standard tissue culture incubators.
3. Transduce SV40-immortalized mlkl^-/- MEFs with the reverse tetracycline controlled transactivator (rtTA)-containing retrovirus expressed from a plasmid containing the blasticidin resistance gene (our modified plasmid) and treat for 1 week with 5 μg/mL blasticidin to select for cells stably expressing rtTA²⁸.
4. Transduce the rtTA-expressing MEFs with the Dox-inducible retroviral vector containing the chimeric MLKL (pRetroX-NBB₁₄₀-2xFV-Venus-puro) and select using 4 μg/mL puromycin for one week, 2 days after transduction.

2. Live-Cell Microscopy Imaging of MLKL-mediated Necroptosis

1. Plate mlkl^-/- MEFs expressing NBB₁₄₀-2xFV-Venus at a density of 50,000 cells per well (1 mL/well) in 24 well plates, and incubate overnight (Figure 2A). Use live-cell staining for automated cell counting (see Table of Materials).
2. Treat cells for 12 h with Dox (0.5 μg/mL) to induce the expression of NBB₁₄₀-2xFV-Venus.
3. Induce necroptosis with 25 nM FKBP dimerizer (Dim) in addition to 25 nM membrane-impermeable green fluorescent dye (see Table of Materials) in standard DMEM (see 1.2). Cells undergoing necroptosis accumulate the dye as their plasma membrane integrity is compromised by NBB₁₄₀-mediated rupture. Perform assays in triplicate or quadruplicate.
4. To monitor necroptosis using single- or dual-color imaging systems (see Table of Materials), place the vessels in the respective imaging unit housed in a mammalian tissue culture incubator.
5. Open the software (see Table of Materials) and select the tab Schedule Upcoming Scans under the Tasks List.
   NOTE: This protocol is for imaging using single-color systems, but a similar software is available for dual-color systems.
6. Select the "Tray, Vessel, Scan Types", and the "Scan Pattern" in the "Physical Layout" tab and "Fast Fluorescence" in the Properties tab.
7. Add the "Scan Time" by clicking on the "Timeline" window and the total number of time points and frequency of acquisition (typically 1 acquisition every 0.5 h to 1 h for 6 h to 12 h, or every 5 min for fast-kinetics necroptosis) and start image acquisition by selecting "Apply" (red button).
8. Quantify necroptosis using the image analysis software (see Table of Materials) by selecting from the "Tasks Pane", "Object Counting New Analysis with Fixed Segmentation Threshold above the background level", which can be determined by hovering the mouse cursor over background regions in several images used in the analysis.
9. Examine fluorescent objects by selecting Previewing using current image and refine the Threshold levels as needed.
10. Integrate green fluorescence counts by opening the "Launch New Analysis" tab, choose the "Time Range", select the wells to be analyzed, enter the "Analysis Job Name", and press "OK".
11. Access analyzed data from the "Analysis Jobs" tab by double-clicking on the respective job name and exporting from the Graph/Export window for additional analysis in external graphing software (see Table of Materials).
12. Quantitate data as green fluorescence positive (+ve) counts/mm² or normalize the counts by confluence and express the data as green fluorescence +ve counts/confluence.
13. Optionally, at the experimental endpoint, stain cells with 100 nM of the membrane-permeant green fluorescent dye (see Table of Materials). Normalize data to % necroptotic cells as the ratio of green fluorescence/green fluorescence at endpoint.

3. Live-cell Confocal Microscopy Imaging of Plasma Membrane Recruitment and Permeabilization by MLKL

1. Plate mlkl^-/- MEFs expressing NBB₁₄₀-2xFV-Venus at a density of 20,000 cells per well (0.5 mL/well) in a glass bottom 4-well microscopy chamber pretreated with 50 µg/mL fibronectin in PBS at 37 °C for 30 min and incubate for ~12 h (Figure 2B).
2. Treat the cells with Dox (0.5 μg/mL) for ~12 h to induce the expression of NBB₁₄₀-2xFV-Venus.
3. Induce necroptosis by incubating with fresh medium (1 mL/well) containing 25 nM Dim to induce NBB₁₄₀-2xFV-Venus activation and translocation to the plasma membrane.
4. Place the chamber on a spinning disc laser scanning confocal microscope built on an inverted stand equipped with environmental control and a 63 1.4 NA objective and begin imaging acquisition using software of choice (see Table of Materials).
5. Initialize the software, select "Focus" icon and YFP filter configuration (excitation wavelength 515 nm).
6. Locate the field of view and focal plane, and engage the focus maintenance system using the "Definite Focus" tab.
7. To begin acquisition, select the "Capture" tab followed by assignment of filter configuration, appropriate EMCCD camera exposure time and intensification gain, and time-lapse acquisition parameters including interval and length of acquisition.
8. Monitor cellular redistribution of the fluorescent NBB₁₄₀-2xFV-Venus protein during necroptosis by acquiring images every 10 s by a EMCCD camera using a 515 nm laser (Movie 1).
   NOTE: Photobleaching is a concern with fluorescent protein imaging. Venus is one of the more resistant fluorescent proteins to photobleaching²⁹. If fluorescent proteins that are more susceptible to photobleaching are used, image every 30 s or longer to allow recovery of signal between imaging time points.
9. To monitor plasma membrane association of NBB₁₄₀-2xFV-Venus during necroptosis, image every 10 s the sample prepared in 3.3 by total internal reflection fluorescence (TIRF) microscopy using a microscope equipped with an automated TIRF slider and a 100X 1.4 NA objective and an EMCCD camera (Movie 2). Follow steps 3.5–3.7, but adjust the TIRF angle within the TIRF focus tab according to sample and bottom glass thickness of the microscopy chamber.
   NOTE: Plasma membrane markers such as LCK-C-RFP may be used as controls for plasma membrane localization in TIRF analysis.
10. Perform image processing to enhance quality by convolution with the negative normalized second derivative of a Gaussian function (Marr algorithm).

4. Electron Microscopy

1. Plate mlkl^-/- MEFs expressing NBB₁₄₀-2xFV-Venus at a density of 5 million cells in a plasma-coated 150 mm cell culture dish and incubate for 12 h (Figure 3).
2. Treat the cells with Doxycycline (0.5 μg/mL) to induce the expression of NB_1-140-2xFV-Venus for 12 h.
3. Induce NBB₁₄₀-2xFV-Venus activation and translocation to the plasma membrane with 25 nM of homodimerizer for 5 min. This time is established from the kinetics of necroptosis observed in live-cell imaging.
4. Remove the media and fix cells using 10 mL of 2.5% glutaraldehyde, 2% paraformaldehyde in 0.1 M sodium cacodylate buffer pH 7.4 (cacodylate buffer) pre-warmed at 37 °C.
5. Collect the sample by scraping and spin the samples for 10 min at 500 x g. Discard the supernatant.
6. Postfix the sample for 1.5 h in reduced 2% osmium tetroxide with 1.5% potassium ferrocyanide in 0.1 M sodium cacodylate buffer. Osmium tetroxide is a staining agent widely used in TEM to provide contrast. We used TEM as preliminary analysis prior to SEM of cell undergoing necroptosis.
7. Rinse the sample 5 times in ultrapure water for 5 min each at 500 x g, followed by rinses in buffer (see 4.6), water, and ethanol on an automatic processor. Discard the supernatant.
8. For the SEM, stain the previously fixed samples with 1% uranyl acetate and lead aspartate heavy-metal stain³¹.
9. Dehydrate the samples through a graded series of alcohol and propylene oxide solutions.
10. Embed the sample in hard resin³² to obtain a resin block.
11. Cut thick sections of 0.5 μm to determine the correct area of analysis.
12. Coat the samples with an ultra-thin film of electrically-conducting metal (Iridium).
13. Image and analyze the samples (Figure 3). For quantification, a low-magnification image of the resin block surface with exposed cells is scanned at 5 keV using an electron microscope.
14. Visualize cells across individual images captured by SEM or imaging software may be utilized to join multiple fields-of-view into one contiguous image³⁰. Roughly 100 cells may be used to evaluate the fraction of cells undergoing necroptosis in this manner by generating a high-resolution montage in software of choice (see Table of Materials).
    NOTE: Scoring necrotic morphology by SEM compares normal cell features with the progression of pre-necrotic or necrotic phenotypes. These features include plasma membrane alterations including microvilli disappearance, flattening, ruptures in cells ranging from 10 nm to 1 µm in size, or loss of organelle structure across at least 30–40% of the cytosolic cell area.

5. Lipid Binding of MLKL by Nuclear Magnetic Resonance (NMR) Spectroscopy

1. Prepare ¹⁵N-labeled NBB_1-156 by standard protein expression and purification as previously described²⁴.
2. Dissolve 500 µg of 1,2-distearoyl-sn-glycero-3-phosphoinositol ammonium salt (18:0 PI) and 1,2-dioleoyl-sn-glycero-3-phospho-(1'-myo-inositol) ammonium salt (18:1 PI) in 500 µL of ice-cold CHCl₃ each for final concentrations of 1 mg/mL.
3. Sonicate 1 mg/mL 18:0 and 18:1 PI in a sonication water bath for 1 min at room temperature or up to 5 min in an ice-water mixture.
4. Aliquot 1 mg/mL 18:0 and 18:1 PI into clean glass vials for individual NMR titration series (Figure 4). Transfer lipids in organic solvents using glass air-tight syringes.
   1. Aliquot 132 µL of 1 mg/mL 18:1 PI (molar mass = 880.137 g/mol) in CHCl₃ for a final resuspension volume of 1,200 µL at 125 µM concentration.
      NOTE: For 5 mm NMR tubes, prepare serial dilution volumes of >1,000 µL and final NMR sample volumes of >500 µL. For 3 mm NMR tubes, prepare serial dilution volumes of >300 µL and final NMR sample volumes of >150 µL.
5. Aliquot 1 mg/mL porcine brain L-α-phosphatidylinositol-4,5-bisphosphate ammonium salt in 20:9:1 CHCl₃/MeOH/H₂O (brain PI(4,5)P₂) into clean glass vials.
6. Place vials under a gaseous stream of argon (Ar) or nitrogen (N₂) with moderate flow to evaporate organic solvent without splashing against the glass vial walls. Five to thirty min is typically sufficient for volumes of 10–200 µL organic solvent.
7. Arrange glass vial(s) at bottom of a Büchner or vacuum flask using large tweezers, seal with a rubber stopper, and attach plastic tubing connected to a vacuum line. Evacuate flask under mild vacuum overnight to remove residual organics.
8. Overlay lipid film aliquots with inert gas, seal with an airtight lid, and store at -20 °C for use within one month or -80 °C for longer-term storage.
9. Prepare NMR sample buffers without detergent (-Det) containing 20 mM Na_xH_yPO₄ pH 6.8, 10% D₂O and with detergent (+Det) containing 20 mM Na_xH_yPO₄ pH 6.8, 10% D₂O, 0.34 mM n-dodecyl-β-D-maltopyranoside (DDM).
10. Add +Det buffer to glass vial with lipid film to achieve desired concentration and sonicate in water bath for up to 30 min, alternating sonication for 10 min followed by visual inspection.
    NOTE: Sonication may warm sample. Allow to cool to room temperature for 15 min before final protein sample reconstitution.
11. Perform serial dilution of lipid-detergent micelles in +Det buffer using sonication of 1:1 mixtures for the desired concentration series. Starting at 125 µM lipid concentration, three repetitions will yield 62.5 µM, 31.3 µM, and 15.6 µM lipid in 0.34 mM DDM.
12. Prepare control and reference NMR samples of protein and additives in -Det and +Det buffers, respectively. Add protein and additives to each dilution of lipid in +Det buffer.
    NOTE: For ¹⁵N NBB₁₅₆, protein is added to a final concentration of 40 µM and supplemented with 2 mM deuterated dithiothreitol to keep Cys residues reduced.
13. Transfer the appropriate volume of samples to 5 mm or 3 mm NMR tubes (see note in step 5.4).
14. Manually load samples in NMR instrument or arrange in an automation platform such as the SampleCase loader.
15. Acquire composite NMR spectra using standard pulse programs for ¹H proton spectra as quality control followed by 2D ¹H-¹⁵N correlation spectra either as transverse relaxation optimized spectroscopy (TROSY) or heteronuclear multiple quantum coherence (SOFAST-HMQC)³³.
    NOTE: Usage of 3 mm NMR tubes requires 64 scans for 1H-¹⁵N TROSY (~3 h) or ¹H-¹⁵N SOFAST-HMQC (~90 min). Scan numbers are halved for 5 mm NMR tubes.
16. Processed 2D NMR spectra may be added to a CARA repository³⁴ for analysis or other software of choice to the user.
17. Analyze software by annotating 2D NMR resonances for three well-dispersed and resolved peaks for each protein state. Record the peak amplitudes for each of the six peaks for each buffer condition containing NBB₁₅₆ (Figure 5A).
    1. Use CARA to prepare a batch-integration list (Click Main Menu heading "Spectrum | Setup Batch List"…) of all spectra collected and analyze using a single peak assignment file (Click Main Menu heading "Peaks | Import Peaklist" …and choose a user-defined .peaks file with 6 total peaks, 3 for each protein state, defined within).
    2. Tune the settings ("Integrator | Tune Peak Model" …) of X-width 0.06 ppm and Y-width 0.30 ppm (Click Main Menu heading "Integrator | Tune Peak Model" …) before recording the final output in two menu selections (1. Click Main Menu heading "Integrator | Integrate Batch List") (2. Click Main Menu heading "Peaks | Export Integration Table") (Figure 5B).
       NOTE: The backbone amide resonances for residues M21, V53, and L105 have been assigned for the NBB₁₅₆ closed-brace state. The open-brace state was assigned to backbone amide resonances of residues G130 and A141 and the epsilon proton side-chain resonance of W133 using 3D NMR experiments²⁴.
18. Normalize samples for direct comparison from different magnets or sample conditions by averaging the three open-brace amplitudes of the reference samples and calculating a normalization factor relating the two references. Calculate the scaled amplitudes for NBB₁₅₆ resonances using the normalization factor and setting negative amplitudes to zero (Table 2).
    Examples: 1) Comparing samples from different magnets: Magnet A, NBB₁₅₆+DDM and Magnet B, NBB₁₅₆+DDM. 2) Detergent comparison: 0.34 mM DDM and 1.7 mM DDM.
19. Calculate for each resonance the scaled fraction of closed- or open-brace by dividing with the range of minimum to maximum amplitude of all spectra collected containing appropriate references of free NBB₁₅₆ and fully open-brace NBB₁₅₆ in detergent.
20. Plot the normalized NMR amplitudes as a function of lipid concentration and compare the fraction of open-brace NBB₁₅₆ in different conditions (Figure. 5C).
    NOTE: Alternatively, analysis of the fraction of closed-brace conformation against lipid concentration can be performed to compare lipid binding to NBB₁₅₆. We prefer the former analysis because it is quicker and better reports on the fraction fully-engaged by lipids.

Results

Visualizing regulated necroptosis execution in live cells has been possible through inducible expression of a minimal truncated MLKL construct, NBB₁₄₀-2xFV-Venus. This construct maintains the ability to induce plasma membrane permeabilization and is activated through Dim-induced oligomerization of the FKBP cassette (2xFV). We observe and quantify necroptosis by live-cell microscopy imaging, monitoring kinetically (every 5 min) the uptake of a cell impermeable green fluorescence...

Discussion

We provide protocols for techniques that we combined to implicated MLKL as the putative executioner of plasma membrane rupture²⁴. In addition to deciphering the regulatory network that regulates MLKL-mediated necroptosis, these techniques can be used independently to characterize other suitable biological systems. Practically speaking, these techniques are medium- to low-throughput discovery tools.

We have routinely used live-cell imaging of NBB₁₄₀-2xFV-Venus...

Materials

Name	Company	Catalog Number	Comments
Cloning and cell line generation
pRetroX-TRE3G	Clontech	631188
Tet-On transactivator plasmid			Llambi et al., 2016
Mouse Embryonic Fibroblasts (MEFs) mlkl-/-			Dillon et al., 2014
Blasticidin S Hydrochloride	Thermo Fisher Scientific	BP2647100 CAS#3513-03-9
Cell death quantification and live-cell microscopy
Doxycycline	Clontech	631311 CAS# 24390-14-5
B/B Homodimerizer AP20187	Takara	635059 CAS# 195514-80-8
SYTOX Green	Thermo Fisher Scientific	S7020
Syto16	Thermo Fisher Scientific	S7578
NMR
15N Ammonium Chloride	Cambridge Isotope Laboratories	NLM-467-10 CAS# 12125-02-9
Deuterated DTT	Cambridge Isotope Laboratories	DLM-2622-1
Deuterium Oxide	Sigma Aldrich	617385-1 CAS# 7789-20-0
n-Dodecyl-β-D-Maltopyranoside	Anatrace	D310 CAS# 69227-93-6
L-α-phosphatidylinositol-4,5-bisphosphate (Brain, Porcine) (ammonium salt)	Avanti Polar Lipids	840046X CAS# 383907-42-4
1,2-distearoyl-sn-glycero-3-phosphoinositol (ammonium salt) (18:0 PI)	Avanti Polar Lipids	850143 CAS# 849412-67-5
1,2-dioleoyl-sn-glycero-3-phospho-(1'-myo-inositol) (ammonium salt) (18:1)	Avanti Polar Lipids	850149 CAS# 799268-53-4
Specialized Equipment
IncuCyte FLR or ZOOM	Essen BioScience, Inc.		Live-cell microscopy imaging
Helios NanoLab 660 DualBeam	Thermo Fisher Scientific		Electron microscope
Software
IncuCyte 2011A Rev2 v20111.3.4288 (FLR)	Essen BioScience, Inc.	http://www.essenbioscience.com	Imaging analysis
FEI MAPS	Thermo Fisher Scientific	https://www.fei.com/software/maps/	EM analysis
TopSpin v3.2	Bruker BioSpin	http://www.bruker.com	NMR data collection
CARA v1.9.1.7		http://cara.nmr.ch/	NMR data analysis
Slidebook	3i (Intelligent Imaging Innovations)	https://www.intelligent-imaging.com/slidebook	Confocal microscopy