1. Patent Search
  2. Details

MODULATION OF MIXED LINEAGE KINASE DOMAIN-LIKE PROTEIN SIGNALING

20230181578 · 2023-06-15

    Inventors

    Cpc classification

    International classification

    Abstract

    The invention is based on a method of modulating the activation or inhibition of Mixed lineage kinase domain-like (MLKL) protein, or a MLKL variant protein, via modulating the intramolecular interaction between the C-terminal helix (Hc) of the psK domain and a hydrophobic groove in the MLKL protein. The invention provides methods and compounds for to selectively target the herein firstly disclosed intramolecular interaction of MLKL protein. Based on the herein disclosed essential intramolecular rearrangement of MLKL, the invention provides small molecules capable of specifically inhibiting mouse or human MLKL. The invention provides uses, including medical applications such as treatments, of MLKL driven conditions including necroptosis, cell trafficking, pathological immune responses and/or inflammation.

    Claims

    1. A method for modulating the activation capacity of necroptosis in a cell, the method comprising the step of contacting the cell with a MLKL modulating compound or MLKL modulating composition, wherein said MLKL modulating compound or MLKL modulating composition when contacted with the cell modulates an intramolecular interaction between the C-terminal helix (Hc) of the psK domain and a hydrophobic groove, wherein the hydrophobic groove is a 3 dimensional structural motif in the MLKL protein, or in a variant MLKL protein, that comprises and/or involves at least one amino acid residue from each of the 4HB domain, the brace region and the psK domain.

    2. The method according to claim 1, wherein the He is located at the C-terminal end of the psK domain, and in the event the MLKL protein, or MLKL variant protein, is human MLKL1 (SEQ ID NO: 3), is between amino acids 400 to 471; and in the event the MLKL protein, or MLKL variant protein, is mouse MLKL2 (SEQ ID NO: 2), is between amino acids 400 to 464.

    3. The method according to claim 1, wherein the hydrophobic groove comprises at least one amino acid residue from each of the 4HB domain, the brace region and the psK domain, and (i) in the event the MLKL protein, or MLKL variant protein, is human MLKL1 (SEQ ID NO: 3), the at least one amino acid residue of the 4HB domain is selected from a sequence between amino acid 80 and 100 of human MLKL; the at least one amino acid residue of the brace region is selected from a sequence between amino acid 120 and 190 from human MLKL; and the at least one amino acid residue of the psK domain is selected from a sequence between amino acid 190 and 471 from human MLKL; or (ii) in the event the MLKL protein, or MLKL variant protein, is mouse MLKL2 (SEQ ID NO: 2), the at least one amino acid residue of the 4HB domain is selected from a sequence between amino acid 20 and 100; the at least one amino acid residue of the brace region is selected from a sequence between amino acid 120 and 190; and the at least one amino acid residue of the psK domain is selected from a sequence between amino acid 190 and 464.

    4. A method for identifying a compound capable of modulating activation of MLKL, the method comprising (iii) bringing into contact an MLKL protein, or an MLKL variant protein, and/or a cell expressing an MLKL protein, or an MLKL variant protein, with a candidate compound, wherein the MLKL protein, or the MLKL variant protein, comprises at least an N-terminal four helix bundle (4HB) domain, a C-terminal pseudo kinase (psK) domain, connected by a brace region; (iv) optionally, activating the MLKL protein, or the MLKL variant protein; (v) determining the intramolecular interaction between the C-terminal helix (He) of the psK domain and the hydrophobic groove of the MLKL protein, or the MLKL variant protein, when contacted with the candidate compound; wherein: (i) an altered intramolecular interaction between He and the hydrophobic groove of the MLKL protein, or the MLKL variant protein, contacted with the candidate compound compared to an MLKL protein, or the MLKL variant protein, not contacted with the candidate compound, and/or (ii) an altered intramolecular interaction between He and the hydrophobic groove of the MLKL protein, or the MLKL variant protein, expressed by a cell contacted with the candidate compound compared to a cell not contacted with the candidate compound, indicates that the candidate compound is a compound capable of modulating the activation of MLKL.

    5. The method according to claim 4, wherein the MLKL protein, or variant MLKL protein, is a constitutively active MLKL protein.

    6. The method according to claim 4, wherein step (ii) is mandatory and comprises phosphorylation of the MLKL protein, or the MLKL variant protein; and/or comprises inducing necroptosis in the cell expressing the MLKL protein, or MLKL variant protein, or alternatively comprises trafficking, pathological immune responses and/or inflammation.

    7. A method for identifying a compound capable of modulating activation of MLKL, wherein the method is selected from: A) a method comprising: (i) bringing into contact a MLKL He domain with a candidate compound, (ii) determining a specific binding of the candidate compound to the MLKL He domain, wherein a specific binding of the candidate compound to the MLKL He domain compared to the binding to an unrelated protein domain indicates that the candidate compound is capable of modulating activation of MLKL; and B) a method comprising: (iii) bringing into contact a candidate compound with a MLKL hydrophobic groove domain which is a protein domain comprising amino acid residues from the 4HB domain, the brace region and the psK domain, (iv) determining a specific binding of the candidate compound to the MLKL hydrophobic groove domain, wherein a specific binding of the candidate compound to the MLKL hydrophobic groove domain compared to the binding to an unrelated protein domain indicates that the candidate compound is capable of modulating activation of MLKL.

    8. (canceled)

    9. The method according to claim 7, wherein the MLKL hydrophobic groove domain and/or the MLKL He domain are provided as full-length MLKL protein, or are provided as test proteins comprising the MLKL hydrophobic groove domain and/or the MLKL He domain, but not comprising the full psK and/or 4HB domain.

    10. A compound or composition for use in the treatment of a disease in a subject, wherein the treatment comprises modulating in the subject the activation of Mixed lineage kinase domain-like (MLKL) protein, or a MLKL variant protein, wherein the MLKL protein, or MLKL variant protein, comprises at least an N-terminal four helix bundle (4HB) domain, a C-terminal pseudo kinase (psK) domain, connected by a brace region, and wherein the method comprises the modulation of the intramolecular interaction between the C-terminal helix (He) of the psK domain and a hydrophobic groove, wherein the hydrophobic groove is a 3 dimensional structural motif in the MLKL protein, or in the variant MLKL protein, that comprises and/or involves at least one amino acid residue from each of the 4HB domain, the brace region and the psK domain.

    11. The compound or composition for use according to claim 15, wherein the He is located at the C-terminal end of the psK domain, and in the event the MLKL protein, or MLKL variant protein, is human MLKL1 (SEQ ID NO: 3), is between amino acids 400 to 471; and in the event the MLKL protein, or MLKL variant protein, is mouse MLKL2 (SEQ ID NO: 2), is between amino acids 400 to 464.

    12. The compound or composition for use according to claim 10, wherein the hydrophobic groove comprises at least one amino acid residue from each of the 4HB domain, the brace region and the psK domain, and (vi) in the event the MLKL protein, or MLKL variant protein, is human MLKL1 (SEQ ID NO: 3), the at least one amino acid residue of the 4HB domain is selected from a sequence between amino acid 80 and 100 of human MLKL; the at least one amino acid residue of the brace region is selected from a sequence between amino acid 120 and 190 from human MLKL; and the at least one amino acid residue of the psK domain is selected from a sequence between amino acid 190 and 471 from human MLKL; or (vii) in the event the MLKL protein, or MLKL variant protein, is mouse MLKL2 (SEQ ID NO: 2), the at least one amino acid residue of the 4HB domain is selected from a sequence between amino acid 20 and 100; the at least one amino acid residue of the brace region is selected from a sequence between amino acid 120 and 190; and the at least one amino acid residue of the psK domain is selected from a sequence between amino acid 190 and 464.

    13. The compound or composition for use according to claim 10, wherein the MLKL modulating compound or MLKL modulating composition is selected from a polypeptide, peptide, glycoprotein, a peptidomimetic, an antigen binding construct and a nucleic acid.

    14. A compound, or composition comprising the compound, wherein the compound is any of the following compounds, as well as derivatives, and solvates, salts, stereoisomers, complexes, polymorphs, crystalline forms, racemic mixtures, diastereomers, enantiomers, tautomers, isotopically labelled forms, prodrugs, and combinations thereof: ##STR00002##

    15. (canceled)

    16. A method for treatment of a disease in a subject, wherein the method comprises modulating in the subject the activation of Mixed lineage kinase domain-like (MLKL) protein, or a MLKL variant protein, wherein the MLKL protein, or MLKL variant protein, comprises at least an N-terminal four helix bundle (4HB) domain, a C-terminal pseudo kinase (psK) domain, connected by a brace region, and wherein the method comprises the modulation of the intramolecular interaction between the C-terminal helix (Hc) of the psK domain and a hydrophobic groove, wherein the hydrophobic groove is a 3 dimensional structural motif in the MLKL protein, or in the variant MLKL protein, that comprises and/or involves at least one amino acid residue from each of the 4HB domain, the brace region and the psK domain.

    17. The method according to according to claim 16, wherein the disease is selected from the group consisting of diseases of the bones, joints, connective tissue and cartilage; muscular diseases; skin diseases; cardiovascular diseases; circulatory diseases; hematological and vascular diseases; diseases of the lung; diseases of the gastro-intestinal tract; diseases of the liver; diseases of the pancreas; metabolic diseases; diseases of the kidneys; viral and bacterial infections; severe intoxications; degenerative diseases associated with the Acquired Immune Deficiency Syndrome (AIDS); disorders associated with aging; inflammatory diseases; auto-immune diseases; dental disorders; ophthalmic diseases or disorders; diseases of the audition tracts; diseases associated with mitochondria; and cancer.

    18. The method according to claim 16, wherein the method comprises administering to the subject a compound, or composition, comprising the compound, wherein the compound is any of the following compounds, as well as derivatives, and solvates, salts, stereoisomers, complexes, polymorphs, crystalline forms, racemic mixtures, diastereomers, enantiomers, tautomers, isotopically labelled forms, prodrugs, and combinations thereof: ##STR00003##

    Description

    BRIEF DESCRIPTION OF THE FIGURES AND SEQUENCES

    [0069] The figures show:

    [0070] FIG. 1: shows that mouse MLKL isoforms have different necroptosis activity. A) Representation of mouse MLKL isoforms; shown sequences correspond to SEQ ID No. 5 and 6. B-C) Necroptosis response of NIH-3T3 MLKL ko cells transfected with one of the isoforms and treated with TSZ. B) Flow cytometry measurements. C) Confocal images. Pictures are representative of at least three independent experiments. Scale bar: 20 m. D) Dead cell population of L929 MLKL ko cells transfected with one of the isoforms and treated with different necroptosis inducers. NT: non-treated, TZ (TNF activation: TNF+zVAD), TSZ (TNF activation: TNF+Smac+zVAD), PZ (TLR3 activation: dsRNA analogue poly(I:C)+zVAD), LZ (TLR4 activation: LPS+zVAD) and DZ (DAI activation: dsDNA analogue poly(A:T)+zVAD). MLKL ko cells were transfected with the different variants of mMLKL-GFP. The dead cell population was calculated from transfected cells. Individual values from at least three independent experiments are shown. E-F) Relative RNA levels of mMLKL1 compared with mMLKL2 in E) different organs from wild type (WT) or MLKL ko mice, F) BMDM and NIH-3T3 cells. Levels are relative to the average of mlkl1_RNA detected in the heart samples E) or in each cell line F). Primers to detect RNA for mMLKL1 were designed to specifically recognize the small sequence that is different to mMLKL2, located near the 3′ end.

    [0071] FIG. 2: shows that the He of mouse MLKL is essential for necroptosis. Model of A) phosphomimetic mMLKL1 (pmMLKL1) and B) phosphomimetic mMLKL2 (pmMLKL2). mMLKL2 structure was refined from the crystal structure (PDB: 4BTF) and mMLKL1 model was based on the refined structure of mMLKL2. Phosphomimetic mutations are represented in red: S228E, S345D, S347D, T349D, and S352D. C) Schematic representation of the Ala insertion mutants; sequence corresponds to SEQ ID No: 7. D) Effect of Ala insertion in the He of mMLKL2 on necroptotic cell death. E) Schematic representation of C-terminal deletion mutants. F) Effect of deletion of different C-terminal segments of mMLKL2 on necroptotic cell death. NIH-3T3 MLKL ko cells were transfected with the different variants of mMLKL-GFP and TSZ-treated. The dead cell population was calculated from transfected cells. Individual values from at least three independent experiments are shown. G) Images from confocal microscopy. Pictures are representative of at least three independent experiments. Scale bar: 20 m.

    [0072] FIG. 3: shows that the accommodation of the He into a novel hydrophobic groove of mMLKL is required for necroptosis. A) 2D projections of PC1 and PC2 eigenvectors. PC1 describes the expansion of the psK domain coupled to the release of the 4HB domain and the rearrangement of the brace helices, while the PC2 describes the motions within the 4HB, the activation loop, and the loops connecting the 4HB, the brace region and the psK domain. B) Representative PC1 structures. More stable structures are represented in dark blue, while similar conformation states are represented in grey and light blue. Arrows indicate the direction of the movement. P indicates phosphorylation. C) Structural representation of the He and the hydrophobic groove of mMLKL2 and mMLKL1. Box: MLKL regions connected by the Hc. D) Structural representation of the Hc/groove of mMLKL2_450R_S454R. E) 2D projections of the mutant. F-G) Necroptotic activity of the mutant. NIH-3T3 MLKL ko cells were transfected with the different variants of mMLKL-GFP and TSZ-treated. F) Dead cell population was quantified by flow cytometry from transfected cells. Individual values from at least three independent experiments are shown. G) Images from confocal microscopy. Pictures are representative of at least three independent experiments. Scale bar: 20 μm. H) Frequency of contacts of the He with different regions of MLKL structure, calculated as average obtained over the whole simulation.

    [0073] FIG. 4: shows that the activation of hMLKL also requires stabilization of the He/groove. A) Representation of human MLKL isoforms. B) Necroptotic activity of different isoforms of hMLKL. HT-29 MLKL ko or HEK cells transfected with one of the isoforms and treated with TSZ. MLKL ko cells were transfected with the different variants of mMLKL-GFP and HEK cells were additionally transfected (as indicated) with hRIP3. C) 2D projections of PC1 and PC2 eigenvectors. Arrows indicate the direction of the movement. D) Representative PC1 structures. The most stable conformation is represented in dark blue, while similar conformation states are represented in grey and light blue. P indicates phosphorylation. E) Structural representation of the He and the hydrophobic groove of hMLKL1. Box: MLKL regions connected by the Hc. F) Representation of insertion and deletion mutants of hMLKL1; sequence corresponds to SEQ ID No: 8. G) Structural representation of the He/groove of hMLKL1_C86D_S467D. H) 2D projections of the mutant. I-J) Necroptotic activity of the mutants. HEK-RIP3 cells were transfected with the different variants of hMLKL1-GFP and TSZ-treated. I) The dead cell population was quantified by flow cytometry from transfected cells. Individual values from at least three independent experiments are shown. J) Images from confocal microscopy. Pictures are representative of at least three independent experiments. Scale bar: 20 μm. K) Frequency of contacts of the He with different regions of MLKL structure, calculated as average obtained over the whole simulation.

    [0074] FIG. 5: shows that the MLKL conformational switch induced by phosphorylation is blocked by alterations in the He/groove. A-D) Necroptotic activity of (A-B) mouse MLKL phosphomutants S345D_S347D or (C-D) human MLKL1_K230M. A and C) Flow cytometry measurements. B and D) Images from confocal microscopy. Cells (NIH MLKL ko for mouse and HEK for human) were transfected with the different variants of MLKL-GFP. The dead cell population was calculated from transfected cells. Individual values from at least three independent experiments are shown. Pictures are representative of at least three independent experiments. Scale bar: 20 μm. E) Model of MLKL auto-activation. Native MLKL fluctuates among different conformations in solution. Phosphorylation stabilizes the active conformation of activatable MLKL through specific interactions that are mediated by the He, which is accommodated in the hydrophobic groove. Correct accommodation of He into the hydrophobic groove is essential to transmit the conformational switch induced by phosphorylation to other regions of MLKL structure.

    [0075] FIG. 6: shows that MBA-h1 and MBA-m1 are new specific inhibitors of human and mouse MLKL that act by blocking the He/groove. A-B) Dose-dependence of the inhibitory effect of MBA-h1 (A) or MBA-m1M (B) on necroptotic cell death in human (HT-29) and mouse (NIH-3T3) cells. C-D) Effect of MBA-h1 (C) or MBA-m1 (D) on MLKL phosphorylation. E-F) Effect of MBA-h1 (E) or MBA-m1 (F) on MLKL translocation to the plasma membrane. HT-29 (C and E) or NIH-3T3 (D and F) wt cells were TSZ-treated in the presence or not of the inhibitors. NSA was used as a control in C and E. G) Inhibitory effect of MBA-m1 in MLKL ko cells transfected with mMLKL2 or the mutant mMLKL2_I84A_F87A. Cells were transfected with the different variants of MLKL-GFP. The dead cell population was calculated from transfected cells. Individual values from at least three independent experiments are shown. H-I) MBA-m1 ameliorates dermatitis in Tnfr1-KO HoipE-KO mice. H) Representative images of mice before and after 2 weeks' treatment with MBA-m1. I) Severity score of dermatitis assessed after and before treatment with the inhibitor. Tnfr1-KO HoipE-KO (n=3). To quantify the sore of region affected, 0 values were assigned to no lesion, and 1 values were assigned when lesions were found in either neck, back, flank or head, being the sum of these elements taken as the final score. To quantify the character of lesions, 0 values were assigned to no lesions, 1 to excoriations one or small punctuate crust, 2 to multiple punctuate crust or coalescing crust, 3 to erosion, ulceration or bleeding. Data are presented as mean values±s.e.m.

    [0076] The sequences show:

    TABLE-US-00001 SEQ ID NOs. 1 - mouse MLKL isoform 1         10         20         30         40 MDKLGQIIKL GQLIYEQCEK MKYCRKQCQR LGNRVHGLLQ         50         60         70         80 PLQRLQAQGK KNLPDDITAA LGRFDEVLKE ANQQIEKFSK         90        100        110        120 KSHIWKFVSV GNDKILFHEV NEKLRDVWEE LLLLLQVYHW        130        140        150        160 NTVSDVSQPA SWQQEDRQDA EEDGNENMKV ILMQLQISVE        170        180        190        200 EINKTLKQCS LKPTQEIPQD LQIKEIPKEH LGPPWTKLKT        210        220        230        240 SKMSTIYRGE YHRSPVTIKV FNNPQAESVG IVRFTFNDEI        250        260        270        280 KTMKKFDSPN ILRIFGICID QTVKPPEFSI VMEYCELGTL        290        300        310        320 RELLDREKDL TMSVRSLLVL RAARGLYRLH HSETLHRNIS        330        340        350        360 SSSFLVAGGY QVKLAGFELS KTQNSISRTA KSTKAERSSS        370        380        390        400 TIYVSPERLK NPFCLYDIKA EIYSFGIVLW EIATGKIPFE        410        420        430        440 GCDSKKIREL VAEDKKQEPV GQDCPELLRE IINECRAHEP        450        460        470 SQRPSVDGRS LSGRERILER LSAVEESTDK KV SEQ ID NOs. 2 - mouse MLKL isoform 2         10         20         30         40 MDKLGQIIKL GQLIYEQCEK MKYCRKQCQR LGNRVHGLLQ         50         60         70         80 PLQRLQAQGK KNLPDDITAA LGRFDEVLKE ANQQIEKFSK         90        100        110        120 KSHIWKFVSV GNDKILFHEV NEKLRDVWEE LLLLLQVYHW        130        140        150        160 NTVSDVSQPA SWQQEDRQDA EEDGNENMKV ILMQLQISVE        170        180        190        200 EINKTLKQCS LKPTQEIPQD LQIKEIPKEH LGPPWTKLKT        210        220        230        240 SKMSTIYRGE YHRSPVTIKV FNNPQAESVG IVRFTFNDEI        250        260        270        280 KTMKKFDSPN ILRIFGICID QTVKPPEFSI VMEYCELGTL        290        300        310        320 RELLDREKDL TMSVRSLLVL RAARGLYRLH HSETLHRNIS        330        340        350        360 SSSFLVAGGY QVKLAGFELS KTQNSISRTA KSTKAERSSS        370        380        390        400 TIYVSPERLK NPFCLYDIKA EIYSFGIVLW EIATGKIPFE        410        420        430        440 GCDSKKIREL VAEDKKQEPV GQDCPELLRE IINECRAHEP        450        460 SQRPSVDGIL ERLSAVEEST DKKV SEQ ID NOs. 3 - human MLKL isoform 1         10         20         30         40 MENLKHIITL GQVIHKRCEE MKYCKKQCRR LGHRVLGLIK         50         60         70         80 PLEMLQDQGK RSVPSEKLTT AMNRFKAALE EANGEIEKFS         90        100        110        120 NRSNICRFLT ASQDKILFKD VNRKLSDVWK ELSLLLQVEQ        130        140        150        160 RMPVSPISQG ASWAQEDQQD ADEDRRAFQM LRRDNEKIEA        170        180        190        200 SLRRLEINMK EIKETLRQYL PPKCMQEIPQ EQIKEIKKEQ        210        220        230        240 LSGSPWILLR ENEVSTLYKG EYHRAPVAIK VFKKLQAGSI        250        260        270        280 AIVRQTFNKE IKTMKKFESP NILRIFGICI DETVTPPQFS        290        300        310        320 IVMEYCELGT LRELLDREKD LTLGKRMVLV LGAARGLYRL        330        340        350        360 HHSEAPELHG KIRSSNFLVT QGYQVKLAGF ELRKTQTSMS        370        380        390        400 LGTTREKTDR VKSTAYLSPQ ELEDVFYQYD VKSEIYSFGI        410        420        430        440 VLWEIATGDI PFQGCNSEKI RKLVAVKRQQ EPLGEDCPSE        450        460        470 LREIIDECRA HDPSVRPSVD EILKKLSTFS K SEQ ID NOs. 4 - human MLKL isoform 2         10         20         30         40 MENLKHIITL GQVIHKRCEE MKYCKKQCRR LGHRVLGLIK         50         60         70         80 PLEMLQDQGK RSVPSEKLTT AMNRFKAALE EANGEIEKFS         90        100        110        120 NRSNICRFLT ASQDKILFKD VNRKLSDVWK ELSLLLQVEQ        130        140        150        160 RMPVSPISQG ASWAQEDQQD ADEDRRAFQM LRRDNEKIEA        170        180        190        200 SLRRLEINMK EIKETLRQSL ESSSGKSPLE ISRFKVKNVK        210        220        230        240 TGSASGCNSE KIRKLVAVKR QQEPLGEDCP SELREIIDEC        250        260 RAHDPSVRPS VDEILKKLST FSK SEQ ID NOs. 5-8: sequences shown in the figures.

    EXAMPLES

    [0077] Certain aspects and embodiments of the invention will now be illustrated by way of example and with reference to the description, figures and tables set out herein. Such examples of the methods, uses and other aspects of the present invention are representative only, and should not be taken to limit the scope of the present invention to only such representative examples.

    [0078] The examples show:

    Example 1: Mouse MLKL Isoforms have Markedly Different Capabilities to Mediate Necroptosis

    [0079] Mouse MLKL has three transcript variants produced by alternative splicing (FIG. 1A). Variant 1 is annotated as the canonical form of MLKL (Uniprot database, Q9D2Y4-1). This transcript is the longest, encoding a protein of 472 amino acid residues (mMLKL1), whereas variant 2 is slightly shorter with 464 residues (mMLKL2). These two isoforms only differ in a sequence of eight amino acids, RSLSGRER, which is located at the C-terminal helix (Hc) of the psK domain of mMLKL1. There is additionally a variant 3 (mMLKL3) that encodes a shorter isoform that lacks the psK domain but contains a short extra sequence at the N-terminus. All three isoforms are generated by alternative 5′ donor sites or exon-skipping alternative splicing mechanisms.

    [0080] To evaluate the necroptotic potential of the different mMLKL isoforms, each of the three isoforms were re-expressed in NIH-3T3 MLKL knock out (ko) cells following induction of necroptosis with a mixture of TNF, a Smac mimetic compound (LCL-161) and the pan-caspase inhibitor zVAD (TSZ), which is the most commonly used necroptosis-inducing cocktail. The inventors quantified cell death by flow cytometry (FIG. 1B) and confocal microscopy (FIG. 1C) using propidium iodide (PI) as a marker of irreversible plasma membrane disruption. As expected, mMLKL3 was intrinsically active, on account of its lack of the psK domain. It had lower activity compared to the 4HB domain alone and was similar to the 4HB+brace mutant, in line with the inhibitory role attributed to the brace region (32, 33).

    [0081] Surprisingly, mMLKL1 and mMLKL2 differed markedly in their capacity to induce necroptosis. Whereas mMLKL1 remained inactive following a stimulus that would normally trigger necroptosis, cells expressing mMLKL2 responded to treatment at the same levels as wild-type (wt), non-transfected NIH-3T3 cells. Specific expression of mMLKL2 in MLKL-deficient cells induced the typical necroptotic phenotype upon treatment: rounding up and detachment prior to plasma membrane breakdown. Similar results were obtained through the activation of the TNF pathway in L929 cells and in MLKL-deficient murine dermal fibroblasts (MLKL-ko MDFs) (not shown) or when necroptosis was triggered via TLR3/4 or DAI (FIG. 1D). Also, these rather different activities of mMLKL1 and mMLKL2 were not the consequence of differential protein expression levels (not shown). Together, these results show that distinct MLKL isoforms have very different necroptosis-inducing potential with mMLKL1 being inactive, mMLKL2 being activatable and mMLKL3 being constitutively active (FIG. 1).

    [0082] The stark difference in necroptotic activity despite the high sequence similarity between mMLKL1 and mMLKL2 raised the question about their distribution in nature. To examine the expression levels of mMLKL1 and mMLKL2 in different tissues and cell lines from mouse, the inventors performed quantitative RT-PCR, with primers that specifically detect the RNA of mMLKL1 or mMLKL2. It was found that necroptosis-inactive mMLKL1 is expressed in the heart, liver, kidney and cecum, as well as in bone-marrow-derived macrophages (BMDMs) and NIH-3T3 cells (FIG. 1E and F). mMLKL1 RNA represented a small fraction compared to the amount obtained for mMLKL2 in most of the cell lines and tissues except in macrophages, and neither isoform was detected in the negative control samples from MLKL ko mice (FIG. 1E).

    Example 2: The C-Terminal Helix (Hc) is a Key Regulatory Element for MLKL Activation

    [0083] The finding that the presence of eight additional amino acids in mMLKL1, as compared to mMLKL2, completely abolishes the capacity of MLKL to induce necroptosis implied that this segment could be important for MLKL regulation. To guide the experimental design for testing the structure-to-function relationship of these extra residues located before the very last C-terminal helix (Hc), molecular dynamics (MD) simulations were used. The inventors built 3D models of mMLKL1 and mMLKL2 based on the crystal structure of full mouse MLKL2 (pdb: 4btf) (21). First, filled gaps were refined in missing regions connecting different secondary structure elements within the 4HB (S79-K94), brace (Y118-Q128), and psK (K351-S358 and V456-V464) domains of the full mMLKL2 structure, since the absence of structural information in the crystal about these segments could mask important interactions of the C-terminus with other protein regions. The segment S79-K94 was modeled based on the α-helix 4 of the human 4HB domain (pdb code: 2msv) (33), while the other unsolved regions were modeled ab initio. Next, the 3D structure of mMLKL1 was modeled based on the refined structure of mMLKL2. Considering that MLKL is activated upon phosphorylation by RIP3, the inventors also modeled phosphomimetic versions of mMLKL1 and mMLKL2 (pmMLKL1 and pmMLKL2) by mutating specific positions within the psK domain (S228E, S345D, S347D, T349D, and S352D) (34, 35). These residues are located opposite to the C-terminal segment of MLKL, in the N-lobe of the psK domain (FIG. 2A and B).

    [0084] Comparative analysis of the structures of mMLKL1 and mMLKL2 phosphomutants predicted that the He of inactive mMLKL1, but not that of mMLKL2, is unfolded due to the presence of the extra eight amino acids. This suggested that the loss of secondary structure in He could be linked, perhaps even causally, to the lack of necroptosis activity of mMLKL1.

    [0085] To assess experimentally whether the effect of this insertion was sequence-specific, different mutants were produced in which Ala-segments of different lengths (2, 4, 6, and 8 Ala) were inserted into mMLKL2 at the equivalent position of the extra eight amino acid sequence of mMLKL1 (FIG. 2C-G). Interestingly, insertion of even the smallest segment of two Ala caused the inactivation of mMLKL2, indicating that the insertion of any sequence before He abolishes mMLKL necroptotic activity (FIG. 2D and G).

    [0086] A previous study suggested that He could act as a suppressor of human MLKL activity by interacting with the α-helix 4 of the 4HB (36). To explore this possibility, the necroptotic activity of He deletion mutants was tested (FIG. 2E). Deletion of the disordered C-terminal tail (Δ457-464) did not impact mMLKL2 activity. Unexpectedly, however, deletion of the He plus the C-terminal tail (Δ448-464) rendered mMLKL2 completely inactive (FIG. 2F and G). These findings support a key role of He in MLKL regulation and imply that, in contrast to previous assumptions (36), He fulfills an activating rather than inhibitory role.

    Example 3: Accommodation of Hc into a Novel Hydrophobic Groove is Required for mMLKL Activity

    [0087] To predict the conformational changes that activate mMLKL2 upon phosphorylation, the inventors compared the MD simulations of mMLKL2 with its phosphomimetic version. To understand how the presence of the eight additional amino acids alters the structural dynamics of the protein, the inventors analyzed this for mMLKL1. An essential dynamics analysis was performed over the structural ensembles obtained through MD simulations. This method allows to predict differences in conformational dynamics among MLKL isoforms and the mutant structures in solution, which cannot be obtained from crystal structures (Amadei et al., 1993). As shown in FIG. 3A and B top panels, the 2D projections of the principal components (PC1) and (PC2) suggested that non-phosphorylated mMLKL2 sampled two equally populated conformational states (denoted as A and B). mMLKL2 phosphorylation stabilized a more compact structure with a significantly reduced conformational subspace (90% of the simulation time remained at the stable state C) (FIG. 3A). The presence of the eight extra residues in mMLKL1 restricted its conformational flexibility and prevented the stabilization of a dominant structure in the phosphomutant (FIG. 3A and B bottom panels). Together, these MD simulations suggested that non-phosphorylated mMLKL1 and mMLKL2 fluctuate between conformational states (FIG. 3A and B). Whereas phosphorylation seemed to stabilize a more compact conformation of mMLKL2, it had no notable effect on the structural dynamics of mMLKL1 (FIG. 3A and B).

    [0088] Inspection of the structure of mMLKL2 predicted that He is accommodated into a hydrophobic groove (FIG. 3C top panel) connecting the He with residues from the 4HB, the brace region and the psK domain. Interestingly, this groove was disrupted in the model of mMLKL1, as indicated by the decrease in the contact frequency of its He with the 4HB and the psK (FIG. 3C bottom panel). In the MD simulations, the He/groove of mMLKL2 underwent larger structural reorganizations upon activation when compared to mMLKL1. mMLKL2 activation induced readjustments in the interactions between the He and the 4HB and psK domains within the hydrophobic groove (FIG. 3C) and decreased the stability of the brace helices. In contrast, a different set of interactions was established in the He/groove of mMLKL1, probably leading to an inactive conformation (FIG. 3C box).

    [0089] To test the hypothesis that He/groove association is essential for mMLKL2 activation, the double-mutant mMLKL2_L450R_S454R was designed. MD simulations predicted that replacement of Leu and Ser by the positively charged Arg would disrupt the He, thereby destabilizing its insertion into the groove and affecting the active conformation of mMLKL2 (FIG. 3D and E). Accordingly, the inventors confirmed experimentally that mMLKL2_L450R_S454R was inactive when re-expressed in mMLKL-deficient cells treated with TSZ (FIG. 3F and G). As before, the lack of activity of mMLKL2_L450R_S454R was neither attributable to an effect of the mutations on the protein expression level nor due to impairment of mMLKL phosphorylation (not shown). These results provided both the experimental validation for the predictions the inventors made about the essential role of the He/groove interaction for the necroptosis-inducing capacity of MLKL, and a satisfactory mechanistic explanation for the marked differences in this capacity between the mMLKL1 and mMLKL2 isoforms.

    Example 4: The Hc/Groove Interaction is Also Essential for Activation of Human MLKL

    [0090] Human MLKL is annotated as two isoforms (hMLKL1 and hMLKL2) that share the N-terminal and the C-terminal sequences, although hMLKL2 lacks a major part of the psK domain (FIG. 4A). An isoform analogous to mMLKL1 has not been identified yet in human or any other organism. The functional characterization of these two human isoforms in MLKL ko HT-29 and HEK cells showed that hMLKL1 could be activated via RIP3 phosphorylation upon exposure to necroptosis stimuli while hMLKL2 was intrinsically active (FIG. 4B), as previously reported (36). Therefore, these two isoforms were functionally homologous to the mMLKL2 and mMLKL3 isoforms (FIG. 1), with hMLKL1 being activatable and hMLKL2 intrinsically active.

    [0091] Human and mouse MLKL orthologues strongly differ in their psK domain (22) and in the underlying mechanism that controls their regulation (32, 37). In fact, both orthologues are unable to induce necroptosis when their host cells are exchanged (24), which raised the question whether He would also be relevant for hMLKL activity. The inventors used MD simulations to generate a structural model of activatable hMLKL1 based on the NMR structure of the human 4HB domain (pdb: 2msv) (33), the crystal structure of its psK (pdb: 4mwi) (22), and the crystal structure of full length mouse MLKL (pdb: 4btf) (21). Phosphomimetic versions of human MLKL1 (phMLKL1) contained the mutations T357E and S358D (38). The MD simulations predicted that human MLKL1 sampled at least two conformational states (denoted as A and B), which were restricted to a single, different conformation in the phosphomutant (C) (FIG. 4C and D), similar to activatable mMLKL2. Furthermore, He of hMLKL1 was also accommodated in the hydrophobic groove (FIG. 4E). The temporal evolution of the structure suggested the formation of relevant interactions connecting He with 4HB, the brace region and the psK domain, thereby leading to increased compactness as well as He/groove and brace stabilization.

    [0092] The role of Hc in hMLKL1 activity was explored experimentally by assessing the necroptosis potential of different mutants (FIG. 4F-J). Either the insertion of the eight amino acid sequence of mMLKL1 or the deletion of a C-terminal segment including the He completely abrogated the activity of hMLKL1 (FIG. 4F, I and J). Conversely, these mutations did not alter the intrinsic activity of hMLKL2. These results demonstrate that the He is also essential for hMLKL1, but not for hMLKL2 activity.

    [0093] Cys86 was identified as a residue positioned at the core of the hydrophobic groove of hMLKL1 in the region comprising the α-helix 4 of the 4HB. The inventors evaluated the effect of exchanging Cys86 (either alone or together with Ser467 in the Hc) by negatively charged Asp (FIG. 4G-J). In the MD simulations, the He/groove interaction in hMLKL1 was disrupted in the resulting mutant hMLKL1_C86D_S467D (FIG. 4G), which had a detrimental impact on its conformational confinement and brace stability (FIG. 4H and S4H). This effect was less dramatic in the simulations of hMLKL1_C86D (data not shown), which may require longer simulation time to destabilize the He/groove. In agreement with the MD predictions, experimentally it was found that the necroptotic activity of hMLKL1 was abolished in both mutants. The defective function of hMLKL1_C86D and hMLKL1_C86D_S467D was not a result of differential hMLKL expression levels or phosphorylation. These results confirmed that, despite the inter-species structural differences, the essential role of the He/groove interaction in MLKL regulation is conserved between the murine and human orthologues.

    Example 5: Hc/Groove Coordinates MLKL Activation Downstream of Phosphorylation

    [0094] We next sought to discern the hierarchy between the He/groove interaction and MLKL phosphorylation in the process of activation. The inventors evaluated experimentally the activity of phosphomimetic variants of mMLKL1 and mMLKL2 generated by substitution of Ser345 and Ser347 by Asp residues (FIG. 5A and B) (35). While introduction of these mutations in mMLKL2 induced stimulus-independent cell death, as expected from previous reports (35), mMLKL1 bearing the same mutations remained incapable of inducing necroptosis. As for the human homologue (hMLKL1), the inventors tested the intrinsically active mutant hMLKL1_K230M harboring a mutation that is located within the ATP-binding pocket imposing the conformational change attained upon phosphorylation (22). For this intrinsically-active mutant, the inventors assessed the effect of (i) the insertion of the extra eight amino acids sequence found in mMLKL1, (ii) the deletion of the C-terminal segment, and (iii) mutations in the He/groove (FIG. 5C and D). Remarkably, all these modifications completely abolished the activity of hMLKL1_K230M, in line with the results obtained with mouse MLKL. Together, these findings demonstrate that altering the He/groove integrity interferes with the molecular switch that is driven by phosphorylation (FIG. 5E) and required for necroptosis induction by both, human and mouse MLKL.

    Example 6: The Hc/Groove Interaction of MLKL can be Selectively Targeted by Small Molecules

    [0095] Based on the conserved relevance of the He/groove for MLKL function, the inventors predicted that it should be possible to find small chemical compounds that inhibit MLKL activity by disrupting the He/groove interaction. Indeed, the structural analysis provided here for the He/groove of activatable MLKL variants affords the ability to rationally screen for and optimize such necroptosis-inhibiting small compounds.

    [0096] By screening the “Tres Cantos Antimalarial Set” (TCAMS) publicly available at the Chembl-NTD database (http://www.ebi.ac.uk/chemblntd) with the Autodock Vina software, the inventors identified putative blockers that target the He/groove of hMLKL1 or mMLKL2. Based on the scoring function of this initial analysis, the inventors selected the best-ranked commercially available compounds for human and mouse MLKL and tested them in necroptosis inhibition assays.

    [0097] Thereby, the inventors identified two compounds, MBA-h1 and MBA-m1, that specifically inhibited necroptosis in human HT-29 (MBA-h1) and mouse NIH-3T3 WT cells (MBA-m1), respectively, in a dose-dependent manner (FIG. 6A and B). The species-specificity was expected considering the large difference in the He/groove interactions between mouse and human MLKL (FIGS. 3 and 4). MBA-h1 inhibited necroptosis in HT-29 MLKL ko re-expressing hMLKL1, whereas MBA-m1 had the same effect in NIH MLKL ko cells containing mMLKL2. These results confirm that both compounds act by inhibiting the activatable variants of human and mouse MLKL. The inhibitory activity of MBA-h1 was observed at concentrations lower than 10 μM, without evidence of toxicity (FIG. 6A). MBA-m1 was also able to inhibit necroptosis at similar concentrations but presented toxicity at higher concentrations (12.5-50 μM) (FIG. 6B).

    [0098] To further characterize the mechanism by which MBA-h1 and MBA-m1 inhibit the necroptosis-inducing capacity of MLKL, the inventors assessed whether they affect MLKL phosphorylation and plasma membrane translocation, two hallmarks of MLKL activation. Western-blot analysis with antibodies against phosphorylated MLKL showed that none of the compounds blocked the phosphorylation of MLKL, implying they indeed act by solely blocking MLKL without affecting the upstream kinases RIP1 and RIP3 (FIG. 6C and D). Instead, both inhibitors potently interfered with MLKL translocation to the plasma membrane, similarly to NSA (FIG. 6E and F). Together, these findings demonstrate that disrupting He/groove interactions with small molecules is a promising strategy to develop selective necroptosis inhibitors based on MLKL targeting.

    [0099] As MBA-m1 is the first specific inhibitor of mouse MLKL the inventors are aware of, the inventors sought to characterize its mechanism of inhibition in more detail. To obtain further evidence that MBA-m1 inhibits necroptosis by specifically targeting the He/groove of mMLKL2, the inventors took a rational approach to design mutations that affect the MBA-m1/MLKL interaction without altering the necroptosis-inducing capacity of mMLKL2. In cell death assays, the inventors confirmed that the mutant mMLKL2_I84A_F87A was as active as mMLKL2, but that it was inert to inhibition by MBA-m1 (FIG. 6G). This result experimentally validates the proposed specific binding mode of MBA-m1 to the He/groove of mMLKL2.

    Example 7: MBA-m1 Ameliorates Dermatitis in Necroptosis-Driven Inflammatory Skin Disease

    [0100] Finally, the inventors aimed to assess the efficacy of inhibiting necroptosis in vivo by specifically targeting the He/groove interaction of MLKL with MBA-m1 in a mouse model with proven necroptosis contribution to disease. the inventors recently showed that mice in which the keratinocyte-specific absence of the LUBAC component HOIP (Hoip.sup.E-KO) is combined with constitutive deficiency for TNFR1 (Tnfr1.sup.KO) suffer from fatal skin inflammation at around day 70 after birth. This phenotype is substantially alleviated by constitutive deficiency in MLKL, showing that MLKL-dependent necroptosis contributes to the pathology (10). At around 30 to 40 days of age, the inventors treated Tnfr1KO;HoipEKO mice, which at this time suffer from mild dermatitis, with MBA-m1 daily for the next three weeks. Strikingly, this treatment significantly ameliorated dermatitis as assessed by a decrease in the severity score of skin lesions at the end of treatment (FIG. 6H and I). These results provide evidence that inhibiting MLKL activity by interfering with the He/groove interaction, in this case with the small molecule MBA-m1, is possible in vivo and that this treatment prevents the progression of a necroptosis-driven pathology (10)

    REFERENCES

    [0101] The references are: [0102] 1. T. Vanden Berghe, A. Linkermann, S. Jouan-Lanhouet, H. Walczak, P. Vandenabeele, Regulated necrosis: the expanding network of non-apoptotic cell death pathways. Nature reviews. Molecular cell biology 15, 135 (February, 2014). [0103] 2. D. Wallach, T. B. Kang, C. P. Dillon, D. R. Green, Programmed necrosis in inflammation: Toward identification of the effector molecules. Science 352, aaf2154 (Apr. 1, 2016). [0104] 3. Z. Su, Z. Yang, L. Xie, J. P. DeWitt, Y. Chen, Cancer therapy in the necroptosis era. Cell death and differentiation 23, 748 (May, 2016). [0105] 4. M. Seehawer et al., Necroptosis microenvironment directs lineage commitment in liver cancer. Nature 562, 69 (October, 2018). [0106] 5. N. Yatim, S. Cullen, M. L. Albert, Dying cells actively regulate adaptive immune responses. Nature reviews. Immunology 17, 262 (April, 2017). [0107] 6. Z. Ying et al., Mixed Lineage Kinase Domain-like Protein MLKL Breaks Down Myelin following Nerve Injury. Molecular cell 72, 457 (Nov. 1, 2018). [0108] 7. S. Zhang, M. B. Tang, H. Y. Luo, C. H. Shi, Y. M. Xu, Necroptosis in neurodegenerative diseases: a potential therapeutic target. Cell death & disease 8, e2905 (Jun. 29, 2017). [0109] 8. W. J. Kaiser et al., Toll-like receptor 3-mediated necrosis via TRIF, RIP3, and MLKL. The Journal of biological chemistry 288, 31268 (Oct. 25, 2013). [0110] 9. H. Nailwal, F. K. Chan, Necroptosis in anti-viral inflammation. Cell death and differentiation 26, 4 (January, 2019). [0111] 10. L. Taraborrelli et al., LUBAC prevents lethal dermatitis by inhibiting cell death induced by TNF, TRAIL and CD95L. Nature communications 9, 3910 (Sep. 25, 2018). [0112] 11. E. J. Petrie, P. E. Czabotar, J. M. Murphy, The Structural Basis of Necroptotic Cell Death Signaling. Trends in biochemical sciences 44, 53 (January, 2019). [0113] 12. R. A. Espiritu, L. Pedrera, U. Ros, Tuning the way to die: implications of membrane perturbations in necroptosis. 29, 201 (2019). [0114] 13. L. Sun et al., Mixed lineage kinase domain-like protein mediates necrosis signaling downstream of RIP3 kinase. Cell 148, 213 (Jan. 20, 2012). [0115] 14. J. Zhao et al., Mixed lineage kinase domain-like is a key receptor interacting protein 3 downstream component of TNF-induced necrosis. Proceedings of the National Academy of Sciences of the United States of America 109, 5322 (Apr. 3, 2012). [0116] 15. M. C. de Almagro, D. Vucic, Necroptosis: Pathway diversity and characteristics. Seminars in cell & developmental biology 39, 56 (March, 2015). [0117] 16. E. J. Petrie, J. M. Hildebrand, J. M. Murphy, Insane in the membrane: a structural perspective of MLKL function in necroptosis. Immunology and cell biology 95, 152 (February, 2017). [0118] 17. Z. Cai et al., Plasma membrane translocation of trimerized MLKL protein is required for TNF-induced necroptosis. Nature cell biology 16, 55 (January, 2014). [0119] 18. X. Chen et al., Translocation of mixed lineage kinase domain-like protein to plasma membrane leads to necrotic cell death. Cell research 24, 105 (January, 2014). [0120] 19. Y. Dondelinger et al., MLKL compromises plasma membrane integrity by binding to phosphatidylinositol phosphates. Cell reports 7, 971 (May 22, 2014). [0121] 20. B. Xia et al., MLKL forms cation channels. Cell research 26, 517 (May, 2016). [0122] 21. J. M. Murphy et al., The pseudokinase MLKL mediates necroptosis via a molecular switch mechanism. Immunity 39, 443 (Sep. 19, 2013). [0123] 22. J. M. Murphy et al., Insights into the evolution of divergent nucleotide-binding mechanisms among pseudokinases revealed by crystal structures of human and mouse MLKL. The Biochemical journal 457, 369 (Feb. 1, 2014). [0124] 23. J. M. Hildebrand et al., Activation of the pseudokinase MLKL unleashes the four-helix bundle domain to induce membrane localization and necroptotic cell death. Proceedings of the National Academy of Sciences of the United States of America 111, 15072 (Oct. 21, 2014). [0125] 24. M. C. Tanzer et al., Evolutionary divergence of the necroptosis effector MLKL. Cell death and differentiation 23, 1185 (July, 2016). [0126] 25. B. Yan et al., Discovery of a new class of highly potent necroptosis inhibitors targeting the mixed lineage kinase domain-like protein. Chemical communications 53, 3637 (Mar. 28, 2017). [0127] 26. W. J. Kaiser et al., RIP3 mediates the embryonic lethality of caspase-8-deficient mice. Nature 471, 368 (Mar. 17, 2011). [0128] 27. P. A. Harris et al., Discovery of Small Molecule RIP1 Kinase Inhibitors for the Treatment of Pathologies Associated with Necroptosis. ACS medicinal chemistry letters 4, 1238 (Dec. 12, 2013). [0129] 28. P. A. Harris et al., Discovery of a First-in-Class Receptor Interacting Protein 1 (RIP1) Kinase Specific Clinical Candidate (GSK2982772) for the Treatment of Inflammatory Diseases. Journal of medicinal chemistry 60, 1247 (Feb. 23, 2017). [0130] 29. T. Xie et al., Structural basis of RIP1 inhibition by necrostatins. Structure 21, 493 (Mar. 5, 2013). [0131] 30. D. Ofengeim, J. Yuan, Regulation of RIP1 kinase signalling at the crossroads of inflammation and cell death. Nature reviews. Molecular cell biology 14, 727 (November, 2013). [0132] 31. W. Declercq, T. Vanden Berghe, P. Vandenabeele, RIP kinases at the crossroads of cell death and survival. Cell 138, 229 (Jul. 23, 2009). [0133] 32. K. A. Davies et al., The brace helices of MLKL mediate interdomain communication and oligomerisation to regulate cell death by necroptosis. Cell death and differentiation 25, 1567 (September, 2018). [0134] 33. L. Su et al., A plug release mechanism for membrane permeation by MLKL. Structure 22, 1489 (Oct. 7, 2014). [0135] 34. M. C. Tanzer et al., Necroptosis signalling is tuned by phosphorylation of MLKL residues outside the pseudokinase domain activation loop. The Biochemical journal 471, 255 (Oct. 15, 2015). [0136] 35. D. A. Rodriguez et al., Characterization of RIPK3-mediated phosphorylation of the activation loop of MLKL during necroptosis. Cell death and differentiation 23, 76 (January, 2016). [0137] 36. K. H. Arnez et al., Analysis of the N-terminal region of human MLKL, as well as two distinct MLKL isoforms, reveals new insights into necroptotic cell death. Bioscience reports 36, eo0291 (2016). [0138] 37. E. J. Petrie et al., Conformational switching of the pseudokinase domain promotes human MLKL tetramerization and cell death by necroptosis. Nature communications 9, 2422 (Jun. 21, 2018). [0139] 38. H. Wang et al., Mixed lineage kinase domain-like protein MLKL causes necrotic membrane disruption upon phosphorylation by RIP3. Molecular cell 54, 133 (Apr. 10, 2014). [0140] 39. E. H. Kim, S. W. Wong, J. Martinez, Programmed Necrosis and Disease: We interrupt your regular programming to bring you necroinflammation. Cell death and differentiation 26, 25 (January, 2019). [0141] 40. M. N. Messmer, A. G. Snyder, A. Oberst, Comparing the effects of different cell death programs in tumor progression and immunotherapy. Cell death and differentiation 26, 115 (January, 2019). [0142] 41. J. K. Rathkey et al., Chemical disruption of the pyroptotic pore-forming protein gasdermin D inhibits inflammatory cell death and sepsis. Science immunology 3, (Aug. 24, 2018). [0143] 42. A. Sali, T. L. Blundell, Comparative protein modeling by satisfaction of spatial restraints. J. Mol. Biol. 234, 779 (1993). [0144] 43. M. S. Madhusudhan, B. M. Webb, M. A. Marti-Renom, N. Eswar, A. Sali, Alignment of multiple protein structures based on sequence and structure features. Protein Eng Des Sel. 22, 569 (Sep. 1, 2009, 2009). [0145] 44. B. J. Bender et al., Protocols for Molecular Modeling with Rosetta3 and RosettaScripts. Biochemistry 55, 4748 (Aug. 30, 2016). [0146] 45. M. J. Abraham et al., GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 1-2, 19 (2015 Sep. 1, 2015). [0147] 46. K. Lindorff-Larsen et al., Improved side-chain torsion potentials for the Amber ff99SB protein force field. Proteins 78, 1950 (2010). [0148] 47. L. J. William, C. Jayaraman, D. M. Jeffry, W. I. Roger, L. K. Michael. (AIP, 1983), vol. 79, pp. 926-935. [0149] 48. H. Mesa-Galloso et al., Disrupting a key hydrophobic pair in the oligomerization interface of the actinoporins impairs their pore-forming activity. Protein Sci 26, 550 (March, 2017). [0150] 49. A. Amadei, A. B. M. Linssen, H. J. C. Berendsen, Essential dynamics of proteins. Proteins: Structure, Function, and Bioinformatics 17, 412 (1993). [0151] 50. J. S. Hub, B. L. de Groot, Detection of functional modes in protein dynamics. PLoS Comput Biol 5, e1000480 (August, 2009). [0152] 51. O. Trott, A. J. Olson, AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem 31, 455 (2010).