Type I interferon controls macrophage metabolism during mycobacterial infection
Gregory Olson [1,2], Tara Murray , Ana Jahn , Dat Mai , Alan H. Diercks , Elizabeth S. Gold , Alan Aderem 
 Center for Global Infections Disease Research, Seattle Children’s Research Institute, Seattle WA
 Medical Scientist Training Program, University of Washington, Seattle WA
Conflicting reports of the immunometabolic response of macrophages to Mycobacterium tuberculosis (Mtb) prevents targeting these fundamental pathways with host-directed therapies or vaccines. We were intrigued that Mtb viability could explain much of the inconsistent data and hypothesized that dissecting the responses to live and heat killed Mtb would reveal pathways controlling metabolism upon Mtb infection. We found that live—but not heat killed—H37Rv (lineage 4 Mtb) decreases murine bone marrow derived macrophage (BMDM) glycolysis and induces mitochondrial dysfunction and stress. RNAseq analyses highlighted the induction of type I interferon (IFN) and downstream signaling only upon live Mtb infection. Strikingly, BMDM lacking the type I IFN receptor, IFNAR, fail to discriminate between live and heat killed Mtb transcriptionally and maintain functional glycolysis and mitochondrial respiration upon live Mtb infection. Importantly, BMDM metabolism after treatment with type I IFN alone mirrors that after live Mtb infection: restrained glycolysis and mitochondrial dysfunction. To confirm type I IFN being upstream of mitochondrial damage, we show that the metabolic changes upon live Mtb infection require the signaling protein STING and are recapitulated with exogenous type I IFN in STING KO BMDM. Surprisingly, type I IFN alone or live Mtb infection increases rather than impairs mitochondrial respiration in primary alveolar macrophages cultured ex vivo. To investigate this in vivo, we sorted alveolar or interstitial macrophages from wild type or IFNAR KO mice 15 days after Mtb aerosol challenge and observed that IFNAR deficiency differentially alters the metabolism of these macrophage subsets. We suggest that macrophage type I IFN signaling, rather than any direct mycobacterial virulence mechanism, controls macrophage metabolism upon Mtb infection and might partially explain the different responses of macrophage subsets in vivo to Mtb.