Metformin reduces intracellular M. tuberculosis replication and inflammation in vitro: A potential host-directed therapy for ocular tuberculosis?

Ocular tuberculosis (TB) remains a leading cause of infectious uveitis in TB-endemic settings. Antitubercular therapy, with or without adjunctive corticosteroids or immunosuppressive agents, is the mainstay of treatment; however, more rapid resolution of intraocular inflammation is desirable to reduce visual morbidity. This preliminary study investigates metformin as a potential host-directed therapy (HDT) for ocular TB through in silico prediction and validation using in vitro disease models. In silico drug discovery was conducted using a bioinformatics platform. In vitro assays were performed using retinal pigment epithelial (RPE) cells and human retinal endothelial cells (RECs) infected with both the H37Rv and a hypervirulent clinical (Beijing) Mycobacterium tuberculosis (Mtb) strains, with and without metformin treatment. Colony-forming unit (CFU) assays and cytokine expression levels in culture supernatants were measured. Through in silico prediction, metformin was identified as a potential HDT for ocular TB. It exhibited significant “anti-match” characteristics when compared to inflammatory pathways derived from the transcriptomic profiles of Mtb-infected RPE cells. A reduction in intracellular Mtb replication (t = 48 h) was observed in Mtb-infected RPE cells treated with 4 mM metformin (median CFU: 2.4 × 10⁵ CFU/mL vs. 3.9 × 10⁵ CFU/mL in untreated cells; p = 0.019). Metformin treatment decreased the production of CCL2/MCP-1 and IL-6 in both Mtb-infected RPE cells and RECs. Additionally, in RECs, a significant reduction in the production of CXCL10/IP-10, CCL5/RANTES, and IL-8 was observed. Altogether, our data suggests that metformin demonstrates potential as a HDT for ocular TB by inhibiting intracellular Mtb replication and reducing inflammation in in vitro ocular TB models.