key: cord-0267477-w3y69v89
authors: Downey, Jeffrey; Randolph, Haley E.; Pernet, Erwan; Tran, Kim A.; Khader, Shabaana A.; King, Irah L.; Barreiro, Luis B.; Divangahi, Maziar
title: Mitochondrial cyclophilin D promotes disease tolerance by licensing NK cell development and IL-22 production against influenza virus
date: 2021-05-28
journal: bioRxiv
DOI: 10.1101/2021.05.28.445832
sha: 32a623aae0532e76a4a87915f68f2831be8e7613
doc_id: 267477
cord_uid: w3y69v89

Immunity to infectious disease involves a combination of host resistance, which eliminates the pathogen, and disease tolerance, which limits tissue damage. While the severity of most pulmonary viral infections, including influenza A virus (IAV), is linked to excessive inflammation, our mechanistic understanding of this observation remains largely unknown. Here we show that mitochondrial cyclophilin D (CypD) protects against IAV infection via disease tolerance. Mice deficient in CypD (CypD-/- mice) are significantly more susceptible to IAV infection despite comparable antiviral immunity. Instead, this susceptibility resulted from damage to the lung epithelial barrier caused by a significant reduction of IL-22 production by conventional NK cells in IAV-infected CypD-/- mice. Transcriptomic and functional data revealed that the compromised IL-22 production by NK cells resulted from dysregulated lymphopoiesis, stemming from increased cell death in NK cell progenitors, as well as the generation of immature NK cells that exhibited altered mitochondrial metabolism. Importantly, following IAV infection, administration of recombinant IL-22 abrogated pulmonary damage and enhanced survival of CypD-/- mice. Collectively, these results demonstrate a key role for CypD in NK cell-mediated disease tolerance.

In response to any given infection, host resistance mechanisms are involved in preventing 52 pathogen invasion or replication. These resistance mechanisms are a critical component of host 53 defense to infection, yet they come with a considerable inflammatory cost that paradoxically 54 threatens host fitness through excessive immunopathology. Thus, mechanisms of disease tolerance 55 are required to mitigate tissue damage, restore organ function and counter the cost of anti-microbial 56 inflammation (1-4). in the lungs of CypD -/mice ( Fig. 1M-N) . Taken together, these data show that CypD is required 140 in immunity to IAV by regulating disease tolerance, rather than host resistance. 12 weeks, reconstitution efficiency was greater than 92% (Fig. S2A) . For further analysis, we 148 chose day 7 post-IAV infection to match the peak lung damage and initial mortality of the CypD-149 deficient mice (Fig. 1) . Chimeric mice reconstituted with CypD -/-BM cells (CypD -/-→ WT) 150 showed a statistically significant increase in erythrocytes in the BAL and enhanced pulmonary 151 inflammation and collagen deposition when compared to WT ( Fig. 2A-D; Fig. S2B ). These results 152 indicate that the reduced disease tolerance in CypD-deficient mice is predominately mediated by 153 the hematopoietic compartment. 154 To elucidate which cell(s) of the hematopoietic compartment is/are responsible for the 155 increased lung damage, we extensively phenotyped innate immune cells by flow cytometry (full 157 and airways. We and others have previously shown that elevated levels of IMM and neutrophils 158 can compromise disease tolerance during IAV infection (11, 16) . However, the kinetics of IMM 159 and neutrophils in the BAL showed no significant differences in frequencies or numbers between 160 WT and CypD -/mice ( Fig. 2E-F) and even showed a trending decrease in CypD -/mice. 161 Interestingly, we found that the kinetics of NK cell accumulation in the airways differed between 162 CypD-deficient and WT mice, with a significant reduction in the frequency of NK cells at day 7 163 post-IAV infection in CypD -/mice, which aligned with the pulmonary damage (Fig. 2G) . No 164 differences in any of the cell populations assessed were observed in the lung parenchyma ( Fig.   165 S2C-E), hinting at the importance of spatial lung immunity (i.e. parenchyma versus airways) 166 during IAV infection. As NK cell kinetics and maturity differed between WT and CypD -/mice following IAV 189 infection, we next asked whether CypD was expressed in NK cells. To investigate this, we purified 190 NK cells from WT and CypD-deficient spleens and found significant expression of CypD in WT, 191 but not CypD -/cells, both at steady-state and upon infection (Fig. S3A) , which aligned with data 192 publicly available from ImmGen (44). We then performed bulk RNA-seq on deficient splenic NK cells isolated at day 5 post-IAV infection (Table S1) Table S3 ). In line with our observation that CypD -/mice display 208 significantly more immature (CD27 + CD11b -) and fewer fully mature (CD27 -CD11b + ) NK cells, 209 we found that genes previously associated with elevated expression in fully mature and Table S3 ), suggesting a distinct metabolic program in CypD-deficient 220 NK cells. Concordant with our enrichment analyses, the average expression of genes belonging to 221 the Wnt/β-catenin signaling pathway was significantly higher in WT compared to CypD -/mice 222 (Fig. 3H , t-test, p = 4.2x10 -3 ), while for the OXPHOS and mTORC1 signaling pathways, we 223 observed a trend toward higher expression in the CypD -/compared to WT mice ( Fig. S3D- steady-state, we detected no differences in OXPHOS in WT and CypD -/-NK cells (Fig. S3F-I) . 234 However, over the course of infection, we observed a significant increase in basal metabolic rate, Table S3) . Notably, among genes more highly-expressed 255 in WT NK cells, we also identified significant enrichments for multiple pathways involved in 256 cytokine secretion, including IL-2 production (FDR = 8.1x10 -3 ), IL-10 secretion (FDR = 1.2x10 -257 3 ) and production (FDR = 5.6x10 -3 ), negative regulation of IL-13 secretion (FDR = 6.8x10 -4 ), and 258 chemokine secretion (FDR = 1.8x10 -3 ), ( Fig. 4B; Table S3 ), suggesting that CypD -/-NK cells 259 display impaired anti-inflammatory cytokine production compared to WT NK cells. Finally, in 260 agreement with our previous observation that the WT NK cell transcriptome was enriched for 261 pathways involved in blood coagulation (Fig. 3B) and that CypD-deficient mice had enhanced 262 hemorrhaging ( Fig. 1H; Fig. S1D ), genes upregulated in WT compared to CypD -/cells were 263 enriched for hemostasis (FDR = 5.5x10 -4 ). Moreover, negative regulation of fibroblast growth 264 factor production (FDR = 1.5x10 -3 ) was enriched in WT NK cells, which might explain the reduced 265 fibrosis in WT lungs following IAV infection ( Fig. 1M-N) . Interestingly, one gene significantly between WT and CypD -/mice in the BAL, lung and blood ( Fig. 5C; Fig. S4A-B) , confirming 287 CCR2-mediated migration of NK cells into the airways is not dependent on CypD expression.

Although continual on-demand egress of NK cells from the BM into the blood to supply 289 peripheral tissues is well-described, it is now known that an additional population of long-lived Fig. 6B; Fig. S4F ), nor in the CLP ( Fig. 6C; Fig. S4G ). 317 Additionally, there were similar levels of the CMP and granulocyte-monocyte progenitor (GMP) 318 in WT and CypD-deficient mice ( Fig. S4H-I) . However, beginning at the NK cell-specific 319 progenitors, we observed a decrease in pre-NKP and NKP populations ( Fig. 6D-F) , as well as a 320 lower frequency of fully mature CD11b + CD27expressing NK cells in the BM upon infection 321 ( Fig. 6G; Fig. S4J ). Thus, CypD mediates NK cell hematopoiesis and an inability to progress 322 through the NK cell lineage correlates with a lack of mature NK cells in the periphery. Mice deficient in CypD are more susceptible to IAV due to a lack of IL-22 production by 343 conventional NK cells 344 Having established that CypD regulates NK cell hematopoiesis and that CypD -/-NK cells 345 are more phenotypically immature with altered transcriptomic and metabolic profiles, we finally 346 hypothesized that NK cells from these mice are functionally impaired and confer susceptibility to 347 IAV infection. As our RNA-seq data indicated impaired anti-inflammatory cytokine production 348 (e.g. IL-10) and wound healing/hemostasis pathways in CypD-deficient NK cells (Fig. 4B-C) , we 349 speculated that altered cytokine production was responsible for the heightened susceptibility and IFN-γ + (Fig. 7B; Fig. S5A ). However, in line with a previous study (66), using Ifngr -/mice, we 359 found no role for IFN-γ signaling in disease tolerance to IAV, with mice exhibiting similar amounts 360 of protein and number of erythrocytes in the BAL (Fig. S5B-C) , although a significant increase in 361 total leukocytes was noted (Fig. S5D) . To completely rule-out that the reduction in IFN-γ was 362 responsible for the damage in the CypD-deficient mice, we reconstituted the airways of WT and 363 CypD -/mice with 100ng of IFN-γ intranasally (i.n.), or vehicle control, at 5 days p.i., collected the 364 BAL at 7 days p.i and characterized pulmonary damage. As expected, we observed a significant 365 increase in protein and erythrocytes in the BAL of CypD-deficient mice that received PBS 366 compared to WT mice, but there was no amelioration in either group that received IFN-γ (Fig S5E-367 F). Collectively, these data confirm that, despite a reduction of NK cell-derived IFN-γ in CypD-368 deficient airways post-IAV infection, IFN-γ was not responsible for the break in disease tolerance. there was a specific reduction of IL-22 + NK cells in both the BAL and spleen of CypD-deficient 381 mice compared to WT ( Fig. 7D; Fig. S5J ). These data collectively confirm the importance of 382 CypD within CD49b + conventional NK cells in IL-22 production.

Because of the observation that IL-22 is involved in disease tolerance to IAV and that CypD -/mice with 100ng of IL-22 or PBS (Fig. 7E) and assessed mortality. As expected, CypD -/-388 mice that received PBS were significantly more susceptible than WT mice that received PBS. 389 Strikingly, after IAV infection, the survival of CypD-deficient mice that received exogenous IL-390 22 was significantly increased and comparable to both WT groups (Fig. 7F) . This enhanced 391 survival was independent of host resistance (Fig. 7G) , but dependent on disease tolerance as the cells, no variance in total leukocyte counts was observed in any group (Fig. 7M) . Furthermore, no 405 differences in viral load between any group could be delineated; thus, disease tolerance, rather 406 than host resistance, was responsible for the reduced airway damage (Fig. 7N) . Collectively, these 407 data indicate that NK cell-derived IL-22 is dependent upon CypD expression and is required for 408 regulating disease tolerance following IAV infection. unknown. Here, we found that CypD regulates IL-22 production by NK cells in the lung (Fig. 7C-482   D; Fig. S5K ), while expression of perforin and granzyme B by NK cells (Fig. 2H) and viral load 483 (Fig. 1C) were unaffected. Therefore, during the course of IAV infection, the functional capacity 484 of NK cells is dynamic and can be changed from promoting resistance to tolerance. tolerance is mediated by CypD. As our RNA-Seq and functional data ( Fig. 3G; Fig. 3I -P) 494 highlighted elevated reliance on OXPHOS in CypD -/-NK cells compared to WT, we propose that Lung tissues were perfused with 10 mL of PBS, harvested and minced before collagenase digestion 591 (150 U mL −1 ) for 1 h at 37 °C. Lungs were passed on a 40 µm nylon mesh, and red blood cells 592 were lysed. For bone marrow staining, cells were isolated following aseptic flushing of the tibiae 593 and femurs, and red blood cells were lysed. BAL were collected as previously described, spun 594 down and red blood cells lysed. Spleens were aseptically removed, crushed on a 40 µm nylon 595 mesh, and red blood cells were lysed. Then total cell counts were determined with a 596 hemocytometer. In some experiments BAL were counted prior to red blood cell lysis to enumerate 597 erythrocyte influx into the airways and then red blood cells were lysed. For peripheral blood 598 staining, the blood was collected by cardiac puncture in a BD Microtainer tube and stained 599 extracellularly; red blood cells were then lysed. The following nested linear model was used to identify genes for which expression levels 

Disease tolerance as a defense strategy

Disease Tolerance as an Inherent Component of Immunity

Two ways to survive infection: what resistance and tolerance can teach 811 us about treating infectious diseases

Unravelling the networks dictating host 813 resistance versus tolerance during pulmonary infections

Clinical aspects of pandemic 2009 influenza A (H1N1) virus infection

H5N1 and 1918 pandemic 817 influenza virus infection results in early and excessive infiltration of macrophages and neutrophils 818 in the lungs of mice

A systems analysis identifies a feedforward 820 inflammatory circuit leading to lethal influenza infection

Review Article: The Fraction of Influenza Virus Infections 822 That Are Asymptomatic: A Systematic Review and Meta-analysis

SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients

Surviving COVID-19: A disease tolerance perspective

Excessive neutrophils and neutrophil extracellular traps contribute to acute 828 lung injury of influenza pneumonitis

CCR2+ monocyte-derived dendritic cells 830 and exudate macrophages produce influenza-induced pulmonary immune pathology and 831 mortality

TNF/iNOS-producing dendritic cells are the necessary evil of lethal influenza 833 virus infection

Lung epithelial apoptosis in influenza virus pneumonia: the role of macrophage-836 expressed TNF-related apoptosis-inducing ligand

Contrasting effects of CCR5 and 838 CCR2 deficiency in the pulmonary inflammatory response to influenza A virus

Leukotriene B4-type I interferon axis 841 regulates macrophage-mediated disease tolerance to influenza infection

Extracellular Matrix Proteolysis by MT1-MMP Contributes to Influenza-844 Related Tissue Damage and Mortality

Natural Killer Cell Recruitment 846 to the Lung During Influenza A Virus Infection Is Dependent on CXCR3, CCR5, and Virus Exposure 847 Dose

Functions of natural killer cells

Elucidating the mechanisms of influenza virus recognition by Ncr1

Recognition of haemagglutinins on virus-infected cells by NKp46 activates 853 lysis by human NK cells

Respiratory Influenza Virus Infection Induces Memory-like Liver NK Cells in Mice

Lethal influenza infection in the absence of the natural killer cell receptor gene 857 Ncr1

In vivo treatment of mice and hamsters with antibodies to asialo GM1 859 increases morbidity and mortality to pulmonary influenza infection

Directing traffic: IL-17 and IL-22 coordinate pulmonary immune defense

Interleukin-22 Reduces Lung Inflammation during Influenza A Virus Infection and 864 Protects against Secondary Bacterial Infection

Interleukin-22 Is Produced by Invariant Natural Killer T Lymphocytes during 866 Influenza A Virus Infection POTENTIAL ROLE IN PROTECTION AGAINST LUNG EPITHELIAL 867 DAMAGES

Interleukin-22 (IL-22) Production by Pulmonary Natural Killer Cells and 869 the Potential Role of IL-22 during Primary Influenza Virus Infection

IL-22 from conventional NK cells is epithelial 872 regenerative and inflammation protective during influenza infection

IL-22 Is Essential for Lung Epithelial Repair following Influenza Infection

Loss of cyclophilin D reveals a critical role for mitochondrial permeability 877 transition in cell death

Cyclophilin D-dependent mitochondrial permeability transition regulates some 879 necrotic but not apoptotic cell death

Bcl-xL mediates RIPK3-dependent necrosis in M. tuberculosis-infected 881 macrophages

Dissecting host cell death programs in the 883 pathogenesis of influenza

NLRX1 prevents mitochondrial induced apoptosis and enhances macrophage 885 antiviral immunity by interacting with influenza virus PB1-F2 protein

Depletion of alveolar macrophages during influenza 888 infection facilitates bacterial superinfections

Maturation of mouse NK cells is a 4-stage developmental program

Immunometabolism and natural killer cell responses

Disease tolerance and immunity in host protection against 894 infection

Circadian control of lung inflammation in influenza infection

An Unbiased Linkage Approach Reveals That the p53 Pathway Is Coupled to NK Cell 898 Maturation

CD11b and CD27 reflect distinct population and functional specialization in human 900 natural killer cells

High-Dimensional Single-Cell Analysis Identifies Organ-Specific Signatures and 902

Gene-expression profiles and transcriptional regulatory pathways that underlie 904 the identity and diversity of mouse tissue macrophages

The Molecular Signatures Database (MSigDB) hallmark gene set collection

Fibronectin maintains survival of mouse natural killer (NK) cells via 908

CD11b/Src/beta-catenin pathway

The metabolic checkpoint kinase mTOR is essential for IL-15 signaling during the 910 development and activation of NK cells

mTORC1 and mTORC2 differentially promote natural killer cell development

Somehow a Master Regulator of Mitochondrial 914 Function

CD59a deficiency exacerbates 916 influenza-induced lung inflammation through complement-dependent and -independent 917 mechanisms

CCR2 defines a distinct population of NK cells and mediates 919 their migration during influenza virus infection in mice

Liver-resident NK cells confer adaptive immunity in skin-contact inflammation. The 921

Tissue-resident natural killer (NK) cells are cell lineages distinct from thymic and 923 conventional splenic NK cells

Natural Killer Cells: Development, Maturation, 925 and Clinical Utilization

Type I IFNs drive hematopoietic stem and progenitor cell collapse via impaired 927 proliferation and increased RIPK1-dependent cell death during shock-like ehrlichial infection

TNF-alpha Coordinates Hematopoietic Stem Cell Survival and Myeloid 930 Regeneration

at a glance

Cyclophilin D counteracts P53-mediated growth arrest and promotes Ras 935 tumorigenesis

p53 mediates loss of hematopoietic stem cell function and lymphopenia in Mysm1 937 deficiency

mdm2 Is critical for inhibition of p53 during lymphopoiesis and the response 939 to ionizing irradiation

Respiratory Influenza A Virus Infection Triggers Local and 941

Systemic Natural Killer Cell Activation via Toll-Like Receptor 7

Elevated Frequency of Gamma Interferon-Producing NK Cells in Healthy Adults 943 Vaccinated against Influenza Virus

IFN-gamma Treatment at Early Stages of Influenza Virus Infection Protects Mice 945 from Death in a NK Cell-Dependent Manner

IFN-gamma increases susceptibility to influenza A infection through suppression 947 of group II innate lymphoid cells

Response to influenza infection in mice with a targeted disruption in the 949 interferon gamma gene

Early local immune defences in the respiratory tract

Genetic variation for disease resistance and tolerance among Arabidopsis 954 thaliana accessions

Are tolerance and training required to end TB?

Host-Parasite Interactions Promote Disease Tolerance to Intestinal Helminth 959 Infection

Mitochondrial cyclophilin D regulates T cell metabolic responses and disease 961 tolerance to tuberculosis

Cyclophilin D Regulates Antiviral CD8(+) T Cell Survival in a Cell-Extrinsic 963

CypD: The Key to the Death Door

Cyclophilin D inactivation protects axons in experimental autoimmune 967 encephalomyelitis, an animal model of multiple sclerosis

Mitochondria in innate immune responses

Type 1 Interferons Induce Changes in Core Metabolism that Are Critical for Immune 972 Function

Skewing of the population balance of lymphoid and myeloid cells by secreted 974 and intracellular osteopontin

M. tuberculosis Reprograms Hematopoietic Stem Cells to Limit Myelopoiesis and 976 Impair Trained Immunity

Defining trained immunity and its role in health and disease

BCG Educates Hematopoietic Stem Cells to Generate Protective Innate 980 Immunity against Tuberculosis

BCG-induced trained immunity in NK cells: Role for non-specific protection 982 to infection

Targeted prostaglandin E2 inhibition enhances antiviral immunity through 984 induction of type I interferon and apoptosis in macrophages

The trinity of COVID-19: immunity, 986 inflammation and intervention

Anti-inflammatory effect of IL-10 988 mediated by metabolic reprogramming of macrophages

Assessment of cardiac fibrosis: a morphometric method comparison for collagen 990 quantification

Cutadapt removes adapter sequences from high-throughput sequencing reads

Near-optimal probabilistic RNA-seq quantification

Differential analyses for RNA-seq: transcript-level 996 estimates improve gene-level inferences

Bioconductor package for differential 998 expression analysis of digital gene expression data

limma powers differential expression analyses for RNA-sequencing and 1000 microarray studies

Statistical significance for genomewide studies

Genetic Ancestry and Natural Selection Drive Population Differences in Immune 1004 Responses to Pathogens

Fast gene set enrichment analysis. bioRxiv

ClueGO: a Cytoscape plug-in to decipher functionally grouped gene ontology and 1008 pathway annotation networks

Gene set enrichment analysis: a knowledge-based approach for 1010 interpreting genome-wide expression profiles

Cytoscape: a software environment for integrated models of biomolecular 1013 interaction networks

Tissue repair genes: the TiRe database and its implication for skin wound healing

Santé (FRQS) Award and the Strauss Chair in Respiratory Diseases. J.D. was supported by the 1042

Science Foundation Graduate Research Fellowship (DGE-1746045), E.P. was supported by a 1044

The funders had no role in study design, data 1045 collection and analysis, decision to publish, or preparation of the manuscript

Representative micrographs of hematoxylin and eosin-stained lungs (B; 1072 scale bar = 30 µM) or Masson's Trichrome (C; scale bar = 30 µM) as quantified in (D)

neutrophils (F) and NK cells (G) 1075 enumerated. Mean fluorescence intensities of perforin and granzyme B of NK cells at day 7 (H) 1076 post-IAV infection. (I) Representative FACS plot of CD27 and CD11b expression on NK cells in 1077 the BAL of WT versus CypD-deficient mice at 7 days post-infection. (J) Quantifications of the 1078 percentages (left panel) and total cells counts (right panel) of NK cell activation subsets in the 1079 BAL at day 7 post-infection

D, where each symbol is a randomly quantified 1081 micrograph (n=10 micrographs) and I, which is a representative plot compiled in J. All figures are 1082 a compilation of at least two experiments, except for H, which is one representative experiment of 1083 two. Statistical analyses were performed using the One-way ANOVA followed by Tukey's 1084 multiple comparisons (A, D), Two-way ANOVA followed by Sidak's multiple comparison test 1085 (E-G, J) or Two-tailed

CypD -/-alters the transcriptome of IAV-infected splenic NK cells and induces a 1088 distinct metabolic phenotype. (A) Volcano plot of genes significantly differentially expressed 1089 (DE) between IAV-infected splenic NK cells of WT (blue, log2 fold change (FC) < -0.5, FDR < 1090 0.10) and CypD -/-(green, log2 FC > 0.5, FDR < 0.10) mice. Genes with a negative log2 FC show 1091 higher expression in WT mice

Selected genes show the most extreme changes in expression between the two 1093 genotype groups. (B) Significant (FDR < 0.01) ClueGO pathway enrichments for genes showing 1094 higher expression (FDR < 0.10) in the WT mice versus (C) CypD -/-mice in IAV-infected splenic 1095 NK cells. (D) Heatmap showing expression levels (mean-centered and scaled) for a subset of 1096 maturity marker genes as identified in Chiossone et

= mature NK [CD27-CD11b+]). (E) Barcode enrichment plots for the 1099 hallmark Wnt/β-catenin signaling, (F) mTORC1 signaling and (G) oxidative phosphorylation 1100 pathways. A positive enrichment score (ES) corresponds to pathway enrichment among genes 1101 more highly expressed in CypD -/-mice (green), while a negative ES corresponds to pathway 1102 enrichment among genes more highly expressed in WT mice (blue). (H) Average, scaled logCPM 1103 expression estimates across genes in the hallmark Wnt/β-catenin signaling pathway. Oxygen 1104 consumption on purified splenic NK cells at day 3 (I-L)

Representative Seahorse curves from which the values in (J-L) and (N-P) were calculated, 1106 respectively. (Q) Representative flow cytometry plots of splenic NK cells differentially stained for 1107

Healthy respiring mitochondria 1108 were considered as Orange hi and Green hi (R) and disrupted mitochondria were considered as 1109

Frequency of MitoSox Red hi NK cells in the spleen. (I-P) All data 1110 are from one representative experiment of two independent experiments and each dot is indicative 1111 of a technical replicate of NK cells compiled from 5 different mice. (Q-T) Data are compiled from 1112 two independent experiments and each dot represents a unique mouse. Statistical differences were 1113 assessed by Student's T-Test (J-L and N-P) or Two-way ANOVA followed by Sidak's multiple 1114 comparisons test (R-T). See also Figure S3. 1115 1116 independent experiments

ANOVA followed by Sidak's multiple comparisons test was performed to determine significance

See also Figure S4

CypD -/-mice have reduced NK cell hematopoiesis in the bone marrow of infected 1144 mice due to cell death of progenitors. (A) Schematic of NK cell hematopoiesis starting at 1145 pluripotent LKS cells down to NK cells

Left panels are representative histograms taken at 5 days post-1151 infection as quantified in the right panels. (J) Differential expression of AnnexinV and NucSpot to 1152 determine the levels of dead cells within the NKP population at various times post-infection. FACS 1153 plots on the left are taken from day 5 post-IAV infection. Panels in B-F and I-J are a compilation 1154 of at least 2 experiments, while G-H are one representative experiment of three independent 1155 experiments, with each symbol representing an individual mouse. In all panels, significance was 1156 assessed by Two-way ANOVA followed by Sidak's multiple comparisons test. See also Figure 1157 S4. 1158 1159 representative of the quantification on the right and are gated against the FMO. (C) Levels of IL-1163 22 in the BAL of infected WT and CypD -/-mice as determined by ELISA. (D) Intracellular 1164 cytokine staining for IL-22 by NK cells in the BAL. FACS plots on the left are representative of 1165 the quantification on the right. An IL-22-deficient mouse was used as a staining control for 1166 specificity. (E) Model of recombinant IL-22 experiments as used in F-I. (F) Mice were infected 1167 with LD50 90 PFU

pulmonary inflammation and fibrosis were assessed by 1170 hematoxylin and eosin or Masson's Trichrome staining, respectively (H; scale bar = 30µM) or 1171 erythrocytes in the BAL were enumerated (I). (J) Schematic of NK cell transfer experiments 1172 performed in K-N

N data are combinations of two or three independent experiments. In B, D, G and M data are from 1175 one experiment that is representative of two or three independent experiments. Micrographs in H 1176 are representative of 3 or 4 individual mice

ANOVA followed by Tukey's multiple 1180 comparisons test; K-N One-way ANOVA followed by Dunnett's multiple comparisons test

See also Figure S5. 1182 1183 1184 1185 SUPPLEMENTARY FIGURE LEGENDS

A-D) Mice were infected with 50 PFU of IAV and levels of LDH 1188 in the BAL (A), IFN-β in the lung (B) or BAL (C) were quantified. (D) Representative pictures of 1189 the BAL of mice over the course of infection as quantified in Figure 1H. Gating strategy for WT 1190 (E) and CypD -/-(F) mice used to quantify innate cells in the study. In panels A-C, each symbol 1191 indicates a separate mouse. Panel A is a compilation of two independent experiments and B-C are 1192 from one experiment

Sidak's multiple comparisons test. Refers to Figures 1-2

Supplementary Figure 2: Kinetics and activation of NK cells. (A) Representative flow 1196 cytometry plots validating reconstitution of chimeric mice. (B-H) Mice were infected with 50 PFU 1197 of IAV. (B) Total cell counts in the BAL of chimeric mice at 7 days post-infection

neutrophils (D) and NK 1199 cells (E) in the lung. (F) Representative FACS plots of NK cell activation in the lung at day 7 and 1200 quantified in (G) with frequencies in the left panel and total cell counts in the right. (H) Frequencies 1201 (left panels) and total cell counts (right panels) of NK cell activation in the spleen at day 5 post-1202 IAV infection. In each panel, each symbol represents an individual mouse. Each panel is a 1203 compilation of two individual experiments, except H which is one representative experiment of 1204 two

Two-way ANOVA followed by Sidak's multiple comparisons test in all other panels

RNA-sequencing of purified, splenic NK cells in control and IAV-1209 infected conditions. (A) Expression of CypD in purified NK cells by qPCR. (B) PCA 1210 decomposition of the splenic NK cell expression data in WT (blue) and CypD -/-(green) mice in 1211 control and IAV-infected conditions. PC1 (percent variance explained = 56.3%) separates samples 1212 by infection status. (C) Barcode enrichment plots for maturity marker gene sets as

A negative enrichment score (ES) corresponds to pathway enrichment 1214 among genes more highly expressed in WT mice. (D) Average, scaled logCPM expression 1215 estimates across genes in the mTORC1 signaling pathway and (E) the oxidative phosphorylation 1216 pathway in IAV-infected splenic NK cells of WT and CypD -/-mice. (F-I) Oxygen consumption of 1217 naïve purified splenic NK cells as detailed in Figure 3I-P. Statistical differences were assessed in 1218 G-I by

A-E) Mice were infected with 50 PFU of IAV. The 1222 percentage of CCR2-expressing NK cells in the lung (A) and blood (B). (C) Differential expression 1223 of CD49b and CD49a in the lung, with a representative FACS plot on the left and quantified on 1224 the right. (D-E) Activation status of NK cells in the blood in uninfected (D; representative FACS 1225 plot on the left) and day 3 (E) post-infection. (F-O) Mice were infected with 50 PFU of IAV and 1226 phenotyped by flow cytometry. Frequencies of LKS (F) and CLPs (G) following infection were 1227 determined as well as frequencies

Relative activation states of naïve NK cells in the BM (J), as well as the percentage 1229 (left panels) and total cell counts (right panels) of Ki67-expressing pre-NKPs (K) and NKPs (L)

BAL (N) at day 5 post-1231 infection. (O) Representative FACS plot (left panel) and quantification (right panel) of cell death 1232 in pre-NKPs as determined by AnnexinV and NucSpot staining. Symbols indicate an individual 1233 mouse and

are one representative experiment of two. Statistical differences were assessed by Two-way 1235

ANOVA followed by Sidak's multiple comparisons test for all panels, except for N where a Two-1236 tailed Student's T-test was performed

The frequency of IFN-γ-1240 producing splenic NK cells at day 5 post-infection. (B-D) WT and Ifngr -/-mice were infected and 1241 the levels of protein (B), erythrocytes (C) and total cells (D) in the BAL at 7 days post-infection 1242 were determined. Following administration of recombinant IFN-γ, levels of protein (E) and 1243 erythrocytes (F) were assessed in the BAL. (G-I) WT and IL-22-deficient mice were infected and 1244 amount of protein (G) and number of cells (H) were enumerated, as well as pulmonary 1245 inflammation by hematoxylin and eosin staining (I; scale bar = 30µM). (J) Frequency of IL-22-1246 producing NK cells in the spleens of WT and CypD-deficient mice at 5 and 7 days post-IAV 1247 infection. In all panels except I and K, symbols represent an individual mouse. In I, micrographs 1248 are a representative image taken from one of four mice. In K quantification is done from 10 random 1249 micrographs and each symbol represents one micrograph. In A and J, data are taken from one 1250 experiment that is representative of three

Two-way ANOVA followed by Tukey's multiple comparisons test or Sidak's (J)

All raw data generated for this paper are available at GEO GSE163290, and processed data files