Necroptosis blockade prevents lung injury in severe influenza

Abstract

Severe influenza A virus (IAV) infections can result in hyper-inflammation, lung injury and acute respiratory distress syndrome^1-5 (ARDS), for which there are no effective pharmacological therapies. Necroptosis is an attractive entry point for therapeutic intervention in ARDS and related inflammatory conditions because it drives pathogenic lung inflammation and lethality during severe IAV infection^6-8 and can potentially be targeted by receptor interacting protein kinase 3 (RIPK3) inhibitors. Here we show that a newly developed RIPK3 inhibitor, UH15-38, potently and selectively blocked IAV-triggered necroptosis in alveolar epithelial cells in vivo. UH15-38 ameliorated lung inflammation and prevented mortality following infection with laboratory-adapted and pandemic strains of IAV, without compromising antiviral adaptive immune responses or impeding viral clearance. UH15-38 displayed robust therapeutic efficacy even when administered late in the course of infection, suggesting that RIPK3 blockade may provide clinical benefit in patients with IAV-driven ARDS and other hyper-inflammatory pathologies.

Extended data Fig. 1. UH15-38 is a potent RIPK3 kinase inhibitor.

a, Immunoblot analysis of pMLKL and cleaved caspase 8 (CC8) in lysates obtained from Ripk3^−/− MEFs stably expressing 2xFv-RIPK3 and treated with dimerizer (AP20187, 100 nM) in the presence of the pan-caspase inhibitor (IDN-6556, 20 μM) and UH15-38 (500 nM) for 12 h. b, Table comparing UH15-38 to GSK’872 in the indicated recombinant kinase assays. Kinase assays were performed using ADPGlo (Promega) and on the DiscoverX platform (Eurofins DiscoverX). ND = Not determined. c, Overview of docking of UH15-38 into human RIPK3. d, Docking of GSK’872 into mouse RIPK3. e, Docking of GSK’843 into mouse RIPK3. f, MEFs infected with PR8 (MOI =2) were treated with indicated drugs in the presence of zVAD (50 μM) and cell viability was determined after 24 h. g, Cell survival kinetics of iBMDMs treated with LPS (10 ng/ml) and TAK1 inhibitor 5z7 (200 nM) following exposure to the indicated concentrations of the RIPK3 inhibitors UH15-38 and GSK’872 or the RIPK1 inhibitor GSK’547 for 6 h. h, Ripk3^+/+ and Ripk3^−/− MEFs were treated with TNF/5z7/IDN6556 and TNF/5z7, respectively, in the presence of the indicated concentrations of UH15-38, and viability was determined after 24 h. Ripk3^−/− MEFs treated with UH15-38 alone were used as controls. i, Viability of Ripk3^−/− MEFs treated with TNF/5z7 in the presence of DMSO or GSK’547 (10 μM). j, Immunoblot analysis of Gasdermin D (GSDMD) cleavage in primary BMDMs pre-treated with LPS (10 ng/ml) for 3 h followed by Nigericin (10 μM) or ATP (5 mM), along with DMSO or UH15-38 (500 nM), for an additional 1 h. k, Viability of primary BMDMs pre-treated with LPS (10 ng/ml for 3 h) followed by treatment with Nigericin (10 μM) and the indicated concentrations of UH15-38 for 1 h. Viability was determined by using CellTiter-Glo assay (as in g, h, k) or Trypan Blue exclusion assay (as in f). Error bars represent mean ± SD (n=3). Data are from one of the at least two independent experiments. Groups were compared using ordinary one-way ANOVA (as in f, k) or the unpaired two-sided Student’s t test (as in i).

Extended Data Fig. 2. Pharmacological profiling of UH15-38.

a, Half-life (T_1/2) and peak concentrations (C_max) of UH15-38 in tissue samples collected from mice treated with four once-daily) doses of UH15-38 (30 mg/kg/day, i.p.). b, c, Immunofluorescence staining for phosphorylated-MLKL (pMLKL) (b) and quantification of pMLKL signal in lung, heart kidney and liver sections (c) obtained from Casp8^−/−Mlkl^FLAG/FLAG mice following i.p administration of either vehicle (n = 3 mice) or UH15-38 (30 mg/kg, once-daily, n = 4 mice) for four days. Six fields of each tissue section per mice were quantified. d, InVEST Safety Panel profile of UH15-38 showing percent inhibition of each target in the presence of 1 μM compound. e, f, Immunohistochemistry for cleaved caspase 3 (CC3) (e) and quantification of cleaved caspase 3 signal (f) in lung, heart, kidney and liver sections (n = 4 mice) following i.p. administration of either vehicle or UH15-38 (30 mg/kg, once-daily) for seven days. IAV PR8-infected lung tissue (n = 4 mice) is shown as positive control in e and f (green). Groups were compared using the two-sided Mann-Whitney U test.

Extended Data Fig. 3. UH15-38 is a potent and specific inhibitor of necroptosis in murine and human cells.

a, Podoplanin (PDPN) staining demonstrates purity of primary Type I AECs. CD140a was used as control for fibroblastic (Fibs) contamination. b, Cell survival kinetics of Type I AECs treated with TCZ and exposed to the indicated concentrations of the RIPK3 kinase inhibitors UH15-38, GSK’843 or GSK’872). c, Cell survival kinetics of primary wild-type MEFs infected with PR8 (MOI = 2) and treated with the indicated RIPK3 kinase inhibitors in the presence or absence (DMSO) of zVAD. d, Immunoblot analysis of the indicated proteins in lysates prepared from primary MEFs infected with PR8 (MOI = 2) and treated with the indicated concentrations of UH15-38. Cell lysates were prepared 12 h after infection. e, Cell survival analysis following infection of MEFs with a panel of IAV and IBV strains and exposure to the indicated concentrations of UH15-38 in the presence or absence of zVAD (50 μM). f, Immunoblot analysis of the indicated proteins in cell lysates prepared from primary MEFs infected with a panel of IAV or IBV strains (MOI = 5) and treated with DMSO or UH15-38 (1 μM). Cell lysates were prepared 12 h after infection. g, Viability of primary Ripk1^−/− MEFs infected with PR8 (MOI = 2) and treated with DMSO or UH15-38 (1 μM). h, Viability of primary MEFs after infection with PR8 (MOI = 2) and treatment with UH15-38 or CSLP37 (a RIPK2 inhibitor) in the presence of either DMSO or zVAD (50 μM). (zVAD is included in the assays to block apoptosis so that effects of the test compounds on necroptosis can be evaluated). i, Viability of primary MEFs after infection with PR8 (MOI = 2) and treatment with UH15-38 or imatinib (an ABL inhibitor) in the presence of either DMSO or zVAD (50 μM). j, k, Zbp1 ^−/− MEFs stably expressing 2xFv-tagged murine ZBP1 were exposed to dimerizer (AP20187, 100 nM) in the presence of the indicated concentrations of UH15-38. Cell viability (j) and immunoblot analyses (k) of the indicated proteins (right) are shown. l, Viability of HeLa-RIPK3 cells treated with human TNFα (100 ng/mL) + cycloheximide (2.5 μg/mL) + zVAD (50 μM) (TCZ) in the presence or absence of UH15-38 (1 μM). m, HeLa-RIPK3 cells treated with TCZ and the indicated concentrations of UH15-38 were examined for pMLKL, total MLKL, and RIPK3 by immunoblot analysis 18 h after treatment. n, M29 cells infected with PR8 (MOI=10) were treated with increasing concentrations of UH15-38 for 24 h and examined for the indicated proteins by immunoblot analysis. H1N1 influenza strains: A/Puerto Rico/8/1934 (PR8), A/California/04/2009 (Cal/09); H3N2 influenza strains: A/Brisbane/10/2007 (Bri/07), A/Singapore/INFIMN-16-0019/2016 (Sin/16); Influenza B virus strains B/Colorado/06/2017 (Col/17) and B/Florida/04/2006 (Flo/06). zVAD (50 μM) was used to prevent apoptosis in this experiment. Cell viability in all panels was determined at 24 h after infection, using the Trypan Blue exclusion assay. Error bars represent mean ± SD of n =3 samples. All data are from one of at least two independent experiments with similar outcomes. Groups were compared using ordinary one-way ANOVA.

Extended Data Fig. 4. UH15-38 is a potent inhibitor of RIPK3 in cellulo.

a, Viability of MEFs following treatment with indicated concentrations of UH15-38. b, Immunoblot analysis of the lysates of MEFs treated with indicated concentrations of UH15-38 for the markers of apoptosis. CC8 = cleaved caspase 8, CC3 = cleaved caspase 3. c, Viability of wild-type MEFs treated with 100 x IC₅₀ (5 μM) of UH15-38 in the presence of DMSO or zVAD (50 μM), as compared to the viability of Ripk3 ^−/− MEFs treated with 100x IC₅₀ UH15-38. d, Viability of Type I AECs following treatment with indicated concentrations of UH15-38. e, Immunoblot analysis of the lysates of Type I AECs treated with indicated concentrations of UH15-38 for markers of apoptosis. Cell viability in all panels was determined using the Trypan Blue exclusion assay. Groups were compared using ordinary one-way ANOVA.Error bars are mean ± SD of n = 3 samples. Figures are representative of two (b, e) or three (a, c, d) independent experiments with similar outcomes.

Extended Data Fig. 5. UH15-38 prevents lethality in severe influenza.

a, Weight loss curves of mice infected with PR8 (6000 EID₅₀; ~LD₁₀₀) followed by i.p. administration of vehicle (n=20) or the following doses of UH15-38: 50 mg/kg (n = 20), 30 mg/kg (n = 21), 15 mg/kg (n = 15), 7.5 mg/kg (n = 15). Vehicle or UH15-38 was administered once-daily per the dosing schedule shown above the graph. b, c, Survival graphs (b) and weight loss curves (c) of mice (n=10) infected with PR8 (6000 EID₅₀) and treated with vehicle (n=12) or with UH15-38 (30 mg/kg, i.p.) for a shortened-, or delayed-dosing regimens, as indicated above panel b. d, e, Survival (d) and weight loss (e) curves of mice (n=7) infected with PR8 (4500 EID50) and treated i.p. with vehicle or GSK’872 (30mg/kg) as indicated above the graph. f, Weight loss analysis of mice infected with IAV H1N1 strain A/California/04/09 (600 EID₅₀; ~LD₆₀) and treated once-daily with either vehicle (n = 10) or UH15-38 (30 mg/kg, i.p., n =13) per the dosing schedule shown above the graph. g, Weight loss curves of mice infected with PR8 (4500 EID₅₀; ~LD₆₀) and treated i.p. with UH15-38 (30 mg/kg, i.p.), per dosing regimens shown above the graph (drug treatment beginning at day 3 after infection, n = 9; all other groups, n = 10/group). h, Wild type mice or Mlkl ^−/− mice (n = 10) were infected with PR8 (2500 EID₅₀). Wild type mice were treated with either vehicle (n = 10) or 30 mg/kg UH15-38 (n=12) once daily for four days starting 24 h after infection. i, Wild type mice or Mlkl ^−/− mice (n = 15) were infected with PR8 (4500 EID₅₀). Wild type mice were treated with either vehicle (n = 14) or 30 mg/kg UH15-38 (n=13) once daily for four days starting 24 h after infection. Mice were observed until 21 days and survival curves were plotted. Wild-type (Mlkl^+/+) mice in panels h and i were generated by intercrossing Mlkl^+/− mice. Groups were compared using the Log-rank (Mantel-Cox) test.

Extended Data Fig. 6. UH15-38 prevents necroptosis, inflammation, and injury in IAV-infected lungs.

a, b, Immunofluorescence staining of cleaved caspase 3 (CC3) (a) and quantification of CC3 signal (b) in lung sections harvested on the indicated days post-infection from mice infected with PR8 (6000 EID₅₀) and treated i.p. with either vehicle or UH15-38 (30 mg/kg once-daily), starting one day after infection, for up to four days. c, Primary Type I AECs were infected with PR8 (MOI = 2) and treated with DMSO or UH15-38 (1 μM). Supernatants were collected 12 h after infection and the indicated chemokines were analyzed on the Luminex platform. d, e, Morphometric images of Tenascin C stained lung sections showing Tenascin C positive area (red), inflamed tenascin C negative lung area (blue), larger airways (yellow) and normal lung area (green) (d) and proportion of Tenascin C positive area (e) in infected lungs (n = 5/group) nine days after infection, following i.p. administration of either vehicle or UH15-38 (30 mg/kg once-daily), starting one day after infection, for four days. f, Levels of IL-17 and CCL5 in BALF from infected mice (n=4/group) measured on the Luminex platform three days after infection, following i.p. treatment with either vehicle or UH15-38 (30 mg/kg once-daily), starting one day after infection, for two days. g, Histological scores of lung sections obtained from mice (n = 5/group) three days (alveolar inflammation and interstitial inflammation) or nine days (septal thickening and epithelial metaplasia) following infection with PR8 (6000 EID₅₀) and treated i.p. with either vehicle or UH15-38 (30 mg/kg once-daily), starting one day after infection, for up to four days. h, Primary BMDMs were infected with PR8 (MOI = 5) and treated with UH15-38 (1 μM) in the presence or absence of zVAD (50 μM). Supernatants were collected 24 h after infection and IL-1 β and IL-18 levels were measured by ELISA. i, Airway resistance (RI) was measured in mice uninfected (Mock / Vehicle, n = 5; Mock / UH15-38, n = 4) or infected (PR8, 2500 EID₅₀, n = 4/group) 10 days after infection following treatment with either vehicle or UH15-38 (30 mg/kg) intraperitoneally one dose daily for four days starting 24 h post infection. Data are presented as mean ± SD. (n = 6/group in b or n = 3/treatment condition in c, h). Comparison between groups was carried out by two sided Mann-Whitney U test (as in b, e, f, g) or by ordinary one-way ANOVA (as in c, h) or two-way ANOVA (as in i).

Extended Data Fig. 7. UH15-38 does not impede virus clearance or anti-IAV CD8⁺ T cell responses.

a,b, Lung morphometry showing virus spread (red areas) in lungs (a) and quantitation of percentage of viral antigen-positive lung area (b) at 6 days after infection. n = 5/group. c, Virus titers in lungs (day 3 and day 6, n = 10/group; day 9 and day 12, n = 5/group) at indicated time points as determined by plaque assay. d, Frequencies of IAV-specific CD8⁺ T cells nine days after infection with PR8 in BALF of vehicle- and UH15-38-treated mice using peptide:MHC tetramers incorporating PB1 residues 703–711 or PA residues 224–233 (n = 10/group). Comparison between groups was carried out by the two-sided Mann-Whitney U test (as in b, c, d), nd = not detected.

Extended Data Figure 8. Gating Strategy for flow cytometric analyses.

a,b, Gating strategy for the flow cytometric analyses presented in Fig. 4i and Extended Data Fig. 7d, respectively.

Extended Data Fig.9. MLKL binding requires the displacement of the RIPK3 αC helix.

a,b, Structures of monomeric (a) and MLKL-bound (b) mRIPK3. N-terminal lobe is depicted in gold and the C-lobe in light green. The αC helix is shown in light blue, and the beginning of the activation loop, including the DFG motif, in salmon. The side chains of key active site residues are shown as sticks and labeled. c. Overlaying the monomeric and MLKL-bound conformations of mRIPK3 shows that the αC helix swivels (blue arrow) outwards in the MLKL-bound conformation, with the E61 residue now facing away from the active site. The dashed red lines indicate axes of the αC helices in inactive and active conformations.

Figure 1.. UH15-38 is a potent RIPK3 kinase inhibitor.

a, Influenza A viruses activate ZBP1 and trigger RIPK3-driven parallel pathways of MLKL-dependent necroptosis and caspase 8-mediated apoptosis. Only necroptosis is reliant on RIPK3 kinase activity. b, Survival analysis of mice of indicated genotypes (Wild-type [WT], n=12; Mlkl^−/−, n=15; Ripk1^kinase-dead, n=13; Ripk3^−/−, n=12; Zbp1^−/−, n=11) following intranasal challenge with IAV H1N1 strain PR8 (6000 Egg Infectious Dose [EID]₅₀). c, Structure of UH15-38. d, Viability of primary wild-type MEFs exposed to the combination of murine TNFα (100 ng/ml), cycloheximide (250 ng/ml), and zVAD (50 μM) (TCZ) in the presence of UH15-38, GSK’872 or GSK’843. Viability was assessed 12 h after treatment. Data are mean ± SD, n = 3, examined over three biologically independent experiments. e, MEFs treated with TCZ in the presence of UH15-38 were examined for phosphorylated MLKL (pMLKL), total MLKL, ZBP1 and RIPK3 12 h after treatment. f, Comparative efficacy of UH15-38, GSK’872 and GSK’843 in human and mouse cells. g, Docking of UH15-38 into mouse RIPK3. N-terminal lobe (gold), αC helix (blue), activation loop (salmon), hinge region (grey), and C-terminal lobe (green) are shown. UH15-38 is shown in ball-and stick representation with a mesh surface, and atoms are color-coded by element. h, Detailed active- site view of UH15-38 docked into mouse RIPK3. Hydrogen bonds (2.0 – 3.1 Å) to the hinge residue M98 as well as the K51, E61 of the αC helix and D161 and F162 of the DFG motif are shown as red dashes. i, LigPlot showing interactions of UH15-38 with residues in mouse RIPK3. Hydrogen bonds are shown with red dashes and bond length in Angstroms. Cell viability in d was determined using Trypan Blue exclusion, and in f by the CellTiter-Glo assay. Groups were compared using the Log-rank (Mantel Cox) test (as in b).

Figure 2.. UH15-38 selectively blocks IAV-induced necroptosis in Type I AECs.

a, IAV mRNA expression in lung cell types at 6 days after infection with PR8 (2500 EID₅₀; i.n.). b, Distribution of IAV⁺ primary lung cell types at 6 days after infection c, IAV replication and ZBP1 expression in Type I AECs following IAV infection in vivo. d, Viability of primary Type I AECs from Zbp1^+/+ and Zbp1^−/− mice following infection with PR8 (MOI = 2). e, Expression of the indicated proteins in primary WT (Zbp1^+/+) Type I AECs following PR8 (MOI=2) infection. f, Viability of Type I AECs after infection with PR8 (MOI = 2) and treatment with UH15-38, GSK’872 and GSK’843 with or without zVAD (50 μM). g, Photomicrographs of primary Type I AECs infected with PR8 (MOI = 2) and treated with UH15-38 (1 μM) with or without zVAD (50 μM). Images are representative of three independent experiments. h, Immunoblots of lysates from primary Type I AECs infected with PR8 (MOI = 2, 12h) in the presence of UH15-38. Data are representative of at least two independent experiments. i, Viability of M29 cells infected with PR8 (MOI =10, 24h) and treated with or without zVAD (50 μM) or UH15-38 (1 μM). j-l, Live human lung sections (n=3/group) infected with PR8 (5 × 10⁶ pfu/ml) and treated with DMSO or UH15-38 (1 μM) for 24 h were stained for pMLKL (j, k) and IAV H1N1 antigen (j, l). Cell viability was determined by Trypan Blue exclusion in panels d, f, and i. Data are average of n=2 as in f and mean ± SD (n=3) as in d, i examined over three independent experiments with similar outcomes. Groups were compared using the unpaired two-sided Student’s t test (as in d), or by ordinary one-way ANOVA (as in i, k, l).

Figure 3.. UH15-38 prevents lethality in severe influenza.

a, Survival analysis of mice infected with PR8 (6000 EID₅₀) followed by i.p. administration of vehicle (n=20) or of UH15-38 at 50 mg/kg (n = 20), 30 mg/kg (n = 21), 15 mg/kg (n = 15), or 7.5 mg/kg (n = 15). Vehicle or UH15-38 was administered once-daily per the dosing schedule shown above the graph. b, Survival analysis of mice challenged with PR8 (4500 EID_50; ~LD₆₀) and treated i.p. with vehicle (n = 15) or with UH15-38 at 30 mg/kg, (n = 15); 15 mg/kg (n = 9), 7.5 mg/kg (n = 10), 3 mg/kg (n = 9), or 1 mg/kg (n = 9). Vehicle or UH15-38 was administered once-daily per the dosing schedule shown. c, Weight loss analysis of PR8 (4500 EID₅₀)-infected mice (n = 15/group) following four once-daily i.p. doses of vehicle or UH15-38 (30 mg/kg). d, Survival analysis of mice infected with IAV H1N1 strain A/California/04/09 (600 EID₅₀; ~LD₆₀) and treated once-daily with vehicle (n = 10) or UH15-38 (30 mg/kg, i.p., n =13) per the dosing schedule shown. e, Survival analysis of PR8 (4500 EID₅₀)-infected mice treated with UH15-38 (30 mg/kg, i.p.) starting either two (n = 10), three (n = 9), four (n = 10), or five (n = 10) days after infection. f, Survival analysis of PR8 (LD₆₀)-infected Mlkl^−/− (dashed lines) mice (vehicle group, n = 11; UH15-38 group, n = 10) or Ripk3^−/− (solid lines) mice (vehicle group, n = 10; UH15-38 group, n = 9) following once-daily treatment with vehicle (black lines) or UH15-38 (30 mg/kg, i.p, red lines) administered per the dosing schedule shown. The LD₆₀ for PR8 in Mlkl^−/− mice was 6500 EID₅₀, and for Ripk3^−/− mice was 2500 EID₅₀. Groups were compared using the Log-rank (Mantel-Cox) test.

Figure 4.. UH15-38 prevents necroptosis, inflammation, and injury in IAV-infected lungs.

a, b, Immunofluorescence staining of pMLKL (a) and quantification of pMLKL signal (b) in lung sections harvested from mice infected with PR8 (6000 EID₅₀) and treated i.p. with either vehicle or UH15-38 (30 mg/kg once-daily), starting one day after infection, for up to four days. c, Levels of inflammatory mediators in BALF (n = 4/group) from mice infected with PR8 (6000 EID₅₀) three days after infection. Mice were treated with either vehicle or UH15-38 (30 mg/kg once-daily) starting D1 post-infection. d, e, Lung images (d) and quantification (e) of neutrophil influx three days after infection with PR8 (6000 EID₅₀), following treatment with either vehicle or UH15-38 (30 mg/kg, i.p., once-daily) starting one day after infection. f, H & E stained lung sections of vehicle- and UH15-38-treated mice nine days after infection (PR8;4500 EID₅₀). Black arrow shows hyaline membranes and red arrow depicts denuded bronchioles. g, Histological scores of late lung injury nine days after infection. h. Arterial oxygen saturation (SpO₂) measured in uninfected (n = 4/group) or infected mice (PR8, 4500 EID₅₀; n = 6/group) 3 and 6 days after infection following treatment with vehicle or UH15-38 (30 mg/kg, i.p.) once daily for four days starting 24 h after infection. i, Frequencies of IFNγ-producing CD8⁺ T cells nine days after infection with PR8 in BALF of vehicle- and UH15-38-treated mice following in vitro stimulation with peptides corresponding to PB1 residues 703–711 or PA residues 224–233 (n = 10/group). Data are mean ± SD. n = 6 in b or n = 5 mice/group unless stated otherwise. Groups were compared using the two-sided Mann-Whitney U test (as in b, c, e, g, i, j), or by two-way ANOVA (as in h).