Tuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb), remains a major global health threat.
Challenges of compliance and toxicities of prolonged antimicrobial regimens and increasing drug resistance
highlight the need for novel treatment approaches. Host-directed therapy (HDT) harnesses host-intrinsic
mechanisms using small molecule drugs to accelerate sterilization and limit lung damage caused by host
immunity. We reported that metformin reduced lung bacterial load and immune pathology in Mtb-infected
mice. Retrospective data from over half a million individuals treated for diabetes with metformin support the
HDT potential of this drug. These studies reported lower rates of Mtb infection and progression from latent to
active TB, less cavitation and all-cause mortality, accelerated sputum conversion and less TB recurrence.
The goal of this project is to discern the mechanisms of metformin HDT efficacy, focusing on inflammation
and fibrosis. Our preliminary data suggest that metformin expands non-classical Ly6Clo monocytes in an
AMPK-dependent manner. These cells are essential for host protection in innate “trained” immunity and they
participate in tissue repair. Ly6Clo monocytes produce CXCL10 that recruits CD8+CXCR3+ memory (TM) cells,
which may restrict Mtb replication in an antigen-specific and non-specific manner. Our data also suggest that
metformin-mediated protection involves NOTCH pathway modulation, with suppression of NOTCH1 that
drives fibrosis and enhancement of NOTCH2 to regulates Ly6Clo monocyte and CD8+ TM cell differentiation.
Experiments in Aim 1 will establish the requirement for AMPK and NOTCH signaling in the protective
functions of metformin using the mouse aerosol TB model. Outcomes are lung bacterial load, immune
pathology, lung tissue damage with collagen remodeling and transcriptomic and proteomic markers of
fibrogenesis. Immunometabolic and epigenetic regulatory activities of metformin will be investigated in the
context of TB-HDT. Experiments in Aim 2 will establish the requirements for CD8+ TM cells and the CXCL10-
CXCR3 axis in host protection and evaluate potential similarities and differences between metformin-
educated CD8+ TM cells and virtual memory T cells. Both Aims include validation studies using alternative
mouse models (C3HeB/FeJ and Collaborative Cross) and de-identified human plasma, blood RNA and
PBMC samples from clinical TB studies having participants with or without metformin exposure. This project
will produce new knowledge about the mechanisms of metformin TB-HDT efficacy that will support the design
and interpretation of human clinical trials using metformin and might identify new targets for HDT agents
having greater specificity, efficacy and tolerability than metformin.