XB-ART-57299Nat Commun 2020 Jun 19;111:3060. doi: 10.1038/s41467-020-16823-3.
Show Gene links Show Anatomy links
Distinct pseudokinase domain conformations underlie divergent activation mechanisms among vertebrate MLKL orthologues.
The MLKL pseudokinase is the terminal effector in the necroptosis cell death pathway. Phosphorylation by its upstream regulator, RIPK3, triggers MLKL''s conversion from a dormant cytoplasmic protein into oligomers that translocate to, and permeabilize, the plasma membrane to kill cells. The precise mechanisms underlying these processes are incompletely understood, and were proposed to differ between mouse and human cells. Here, we examine the divergence of activation mechanisms among nine vertebrate MLKL orthologues, revealing remarkable specificity of mouse and human RIPK3 for MLKL orthologues. Pig MLKL can restore necroptotic signaling in human cells; while horse and pig, but not rat, MLKL can reconstitute the mouse pathway. This selectivity can be rationalized from the distinct conformations observed in the crystal structures of horse and rat MLKL pseudokinase domains. These studies identify important differences in necroptotic signaling between species, and suggest that, more broadly, divergent regulatory mechanisms may exist among orthologous pseudoenzymes.
PubMed ID: 32561735
PMC ID: PMC7305131
Article link: Nat Commun
Species referenced: Xenopus
Genes referenced: diablo mapk1 mlkl ripk3
Article Images: [+] show captions
|Fig. 1. Few MLKL orthologues reconstitute necroptotic signaling in mouse and human cells.Genes encoding human, mouse, rat, horse, pig, chicken, tuatara, frog and stickleback MLKL were stably introduced into Mlkl−/− mouse dermal fibroblast (MDF) (a) and MLKL−/− human U937 (b) cells and expressed upon doxycycline treatment (induced). Cells were either untreated (UT) or treated with a necroptotic stimulus (TNF, Smc mimetic, IDN-6556; TSI) to examine the capacity of each orthologue to reconstitute necroptotic signaling. Cell death was measured by propidium iodide (PI) uptake by flow cytometry. Data shown are mean ± SEM of independent experiments on one U937 cell line (n = 3 for mouse, rat, horse and chicken MLKL; n = 4 for human and frog MLKL; n = 5 for pig MLKL) or two biological replicate MDF lines (n = 6, except for n = 8 for mMLKL-FLAG). * represents statistical significance of p < 0.05 using a paired, two-tailed t-test: a mouse-FLAG p = 0.000015, horse p = 0.0017, pig p = 0.0290; and b human p = 0.0104, pig p = 0.0003. Source data are provided in a Source Data file. An example of the flow cytometry gating strategy used throughout this study is shown in Supplementary Fig. 2. Expression of introduced genes was verified by western blot (Supplementary Fig. 3).|
|Fig. 2. Rat and horse MLKL pseudokinase domain structures diverge from mouse MLKL.a The structure of the human MLKL pseudokinase domain (PDB, 4MWI39) shows a conventional active kinase-like conformation. The K230–E250 salt bridge (equivalent of the K72-E91 PKA interaction) between the β3 ATP-binding lysine and the Glu in the αC helix (shown in dark blue throughout) is shown in the zoomed inset. b The mouse MLKL pseudokinase domain (from PDB 4BTF18) shows a more open conformation owing to the activation loop adopting an unusual helical conformation (pale blue), which displaces the αC helix (dark blue). An unconventional interaction between the β3 K219 and the activation loop Q343 results (zoomed inset). c The rat MLKL pseudokinase domain adopts an active protein kinase-like conformation, resembling that of the human structure. The β3 K219 forms a conventional salt bridge with the αC helix E239 (zoomed inset), rather than a hydrogen bond with the activation loop Gln observed in the mouse structure. d The horse MLKL pseudokinase domain shows a similar active protein kinase-like conformation, with the K228:E248 salt bridge shown in the zoomed inset. The horse structure exhibits previously unobserved features, including burial of the activation loop in the pseudoactive site and an additional helix in the β3-αC loop (displayed in more detail in Fig. 4).|
|Fig. 3. Horse MLKL β3-αC loop and C-lobe residues facilitate mouse RIPK3 engagement.a The horse MLKL pseudokinase domain structure superimposed on the reported mouse MLKL pseudokinase:mouse RIPK3 kinase domain structure (PDB 4M69 (ref. 40)). A site in each of the N- and C-lobes of horse MLKL pseudokinase domain were identified as possible interfaces between horse MLKL and mouse RIPK3 (b, c). d Wild type or alanine substitutions of putative interface residues in horse MLKL were expressed in Mlkl−/− MDF cells and their capacity to induce death in the absence of stimulation (untreated, UT) or upon addition of a necroptotic stimulus (TSI) following doxycycline induction of protein expression was quantified by PI uptake. e Wild type or F373A mouse MLKL were expressed in Mlkl−/− MDF cells and their capacity to induce death evaluated as in panel d. Data in d, e are shown are mean ± SEM of ≥3 independent experiments for two biological replicate MDF lines (n = 6 for wild-type horse MLKL; n = 7 for S233A, R241A; n = 8 for R235A, F373A; and n = 9 for S237A, Y384A and mouse MLKL wild type). * represents statistical significance of p < 0.05 using a paired, two-tailed t-test: d horse MLKL wild type p = 0.0017, R236A p = 0.0162, S237A p = 0.0249, R242A UT uninduced vs UT induced p = 0.0236, R242A TSI uninduced vs TSI induced p = 0.0041; e wild-type mouse MLKL p = 0.0002. Source data are provided in a Source Data file.|
|Fig. 4. Horse MLKL activation loop phosphorylation induces conformational flexibility.a The activity of phosphomimetic mutations, S345D in rat MLKL, or T356E-S357E in horse MLKL, or alanine substitution (T356A-S357A) in horse MLKL, was tested in MLKL−/− U937 and/or Mlkl−/− MDF cells upon doxycycline induction of expression, in the presence (TNF, Smac mimetic, IDN-6556; TSI) and absence (untreated; UT) of necroptotic stimuli. These data are plotted alongside wild-type controls from Fig. 1. b, c The activation loop of the horse MLKL pseudokinase domain is buried in the pseudoactive site in a position occupied by ATP in conventional protein kinases, such as ERK (PDB 4GT3) (d). e Wild-type horse MLKL or alanine substitution mutants of activation loop and adjacent pseudoactive site residues were stably introduced into Mlkl−/− MDF cells and the capacity to kill cells in the presence (TSI) or absence (UT) of necroptotic stimuli in the presence (induced) or absence (uninduced) of doxycycline-induced exogene expression quantified by PI uptake using flow cytometry. Data in a and e are shown as mean ± SEM of ≥3 independent experiments for each of two biological replicate MDF lines (n = 6 for all in a, n = 7 for T208A and Q355A-T208A, n = 8 for Y282A and Q355A) or one U937 line (n = 3). * represents statistical significance of p < 0.05 using a paired, two-tailed t-test: a wild-type horse MLKL p = 0.0017, rat MLKL S345D in MDFs UT uninduced vs UT induced p = 0.0089, rat MLKL S345D in MDFs TSI uninduced vs TSI induced p = 0.0035, rat MLKL S345D in U937s UT uninduced vs UT induced p = 0.0108 and rat MLKL S345D in U937s TSI uninduced vs TSI induced p = 0.00114. For e T208A p = 0.0013, Q355A p = 0.0099 and Q355A-T208A p = 0.0056. f A comparison of molecular dynamics simulations on horse MLKL reveals increased activation loop flexibility in the phosphorylated MLKL model. The x-axis shows residue numbers and the y-axis shows root mean square fluctuation (RMSF) across the simulation. The phosphorylated residues, pT356 and pS357, are shown in red. g A series of snapshots of phosphorylated horse MLKL show the phosphorylated activation loop moving out of the pseudoactive site. Zoomed insets show hydrogen bonds at various stages of the transition.|
References [+] :
Adams, PHENIX: a comprehensive Python-based system for macromolecular structure solution. 2010, Pubmed