Defects in lipid homeostasis reflect the function of TANGO2 in phospholipid and neutral lipid metabolism

**Background:** TANGO2 is a gene identified in a 2006 screen for genes required for Golgi transport in Drosophila. Unlike TANGO1, TANGO2 lacks a transmembrane domain and has been reported to be cytosolic and mitochondrial. Mutations in TANGO2 cause metabolic encephalopathy and arrhythmias, exacerbated by nutrient starvation and often fatal. The physiological role of TANGO2 and the mechanism by which its mutations lead to starvation-induced fatality were unknown. This study aimed to characterize TANGO2's intracellular localization and its function in lipid metabolism.

**Methods:** The intracellular localization of TANGO2 was studied in live HepG2 cells using C-terminally tagged TANGO2 (mScarlet or EGFP) co-expressed with fluorescent organelle markers (mitochondria, ER, peroxisomes, lipid droplets) and analyzed by confocal time-lapse microscopy. FRET microscopy assessed proximity between TANGO2-mScarlet and Tom20 (outer mitochondrial membrane marker). TANGO2 was depleted in HepG2 cells using two sequential rounds of siRNA transfection (~75% reduction). Lipid droplet morphology was assessed by confocal microscopy (HCS LipidTOX Deep Red, anti-ADRP) and quantified using a custom ImageJ macro. Total lipid droplet volume was measured by flow cytometry. Lipidomic analysis by mass spectrometry (Lipotype GmbH) quantified lipid classes in mock and TANGO2-depleted cells under control and starvation conditions (OptiMEM, 4 hr). mRNA expression of 13 lipid metabolism enzymes (GPATs, AGPATs, LPINs, PLA2G4A) was measured by RT-qPCR. ROS levels (CellROX Deep Red) and lipid peroxidation (Image-iT Lipid Peroxidation kit) were measured by flow cytometry. Experiments were also performed in primary fibroblasts from healthy donors and TANGO2 deficiency disorder (TDD) patients (TAN043, exon 3–9 deletion).

**Key Results:** TANGO2 localized predominantly to mitochondria (Pearson coefficient r = 0.88 with Mito-mTurquoise) and was enriched at sites where mitochondria juxtapose ER and lipid droplets. FRET confirmed proximity between TANGO2 and Tom20 (p = 0.0002). TANGO2-depleted HepG2 cells showed significantly larger lipid droplets: total LD volume per cell (p = 0.0005) and volume per LD (p = 0.002) were increased, but LD number per cell was unchanged. Starvation further enlarged LDs. Flow cytometry confirmed increased total LD volume in TANGO2-depleted HepG2 cells and TAN043 fibroblasts, with evidence of increased biogenesis and defective lipolysis. Lipidomics revealed that TANGO2-depleted cells had markedly elevated LPA and reduced PA and cardiolipin (CL) levels. Starvation exacerbated these changes: LPA increased further, PA decreased further, and PG levels dropped dramatically. RT-qPCR showed increased GPAT3 and AGPAT1 mRNA expression in TANGO2-depleted cells under both control and starvation conditions. TANGO2-depleted cells exhibited significantly increased ROS levels (CellROX) and lipid peroxidation, further elevated upon nutrient starvation. No significant difference in total mitochondrial mass was observed between mock and TANGO2-depleted HepG2 cells.

**Clinical Implications:** The study identifies TANGO2 as a key regulator of lipid homeostasis, specifically in the conversion of LPA to PA, likely through acyl-CoA metabolism. The resulting PA deficiency impairs cardiolipin synthesis, increases ROS and lipid peroxidation, and enlarges lipid droplets—all exacerbated by nutrient starvation. These defects mirror the clinical phenotype of TANGO2 deficiency: metabolic crises, rhabdomyolysis, cardiomyopathy, and arrhythmias triggered by fasting. The findings explain why vitamin B5 (a CoA precursor) and multivitamins may ameliorate metabolic crises in patients, as they support acyl-CoA production and provide antioxidant effects. The study provides a mechanistic basis for understanding TANGO2-related pathologies and suggests potential therapeutic strategies targeting lipid metabolism.