The story begins with mitochondria. When fat tissue expands rapidly during weight gain, adipocytes come under intense oxidative stress, and their mitochondria become damaged and leaky. Fragments of mitochondrial DNA then escape into the cytoplasm of the cell. Because cytosolic DNA normally signals a viral or microbial invasion, the cell’s alarm system responds. An enzyme called cGAS detects the stray DNA and triggers STING, a protein on the endoplasmic reticulum that launches a cascade of inflammatory signaling, including the production of type I interferons and pro-inflammatory cytokines. In white adipose tissue, this chronic activation of the cGAS-STING pathway is increasingly recognized as a central driver of the tissue inflammation that underlies insulin resistance.

The researchers focused on NRF2, or nuclear factor erythroid 2-related factor 2, a transcription factor that functions as the cell’s principal antioxidant thermostat. Under normal conditions, NRF2 is held inactive by a protein called KEAP1, but oxidative stress liberates it, allowing it to enter the nucleus and switch on a battery of genes that detoxify reactive oxygen species, repair damaged proteins, and maintain mitochondrial integrity. The team first confirmed that NRF2 activity is markedly reduced in the white adipose tissue of mice fed a high-fat diet, and in mature adipocytes grown in the laboratory. That loss of antioxidant defense, they hypothesized, might be what allows mitochondrial DNA leakage and STING activation to proceed unchecked in obesity.

To test the idea, the researchers exposed differentiated adipocytes to palmitic acid, a saturated fatty acid abundant in high-fat diets that reliably induces lipotoxic stress. When they activated NRF2 in these cells, the results were striking. Oxidative stress markers fell, the amount of mitochondrial DNA leaking into the cytosol diminished, and the downstream STING cascade was throttled back: STING failed to translocate efficiently to its signaling compartments, and the production of inflammatory effectors dropped. Genetic experiments reinforced the picture. When the researchers knocked down NRF2, the anti-inflammatory protection vanished; when they overexpressed it, the suppression of STING signaling deepened. Transcriptomic profiling of the treated cells showed a broad quieting of inflammatory gene programs, consistent with NRF2 acting as an endogenous brake on the STING-driven inflammatory cascade rather than a mere general antioxidant.

The in vivo experiments were where the findings gained their translational weight. Obese mice treated with CDDO-Im, a potent synthetic NRF2 activator derived from oleanolic acid triterpenoids, showed significant improvement across the metabolic spectrum: reduced weight gain, better insulin sensitivity, and lower systemic inflammation. Crucially, the team used adeno-associated vectors carrying the adiponectin promoter—AAV-pAdipoq—to knock down NRF2 specifically within adipocytes of the fat tissue, or conversely to force overexpression of STING in the same cells. Either intervention largely abolished the metabolic benefits of the drug. That loss-of-effect experiment is the study’s strongest argument that the NRF2 activator works primarily by restraining STING signaling inside fat cells, not through some unrelated systemic action.

Multiplex imaging and cell-fractionation analyses added an important cellular detail: active STING signaling in obese adipose tissue operates predominantly in mature adipocytes, rather than in the immune cells or stromal cells that populate the tissue. This matters because much of the inflammation literature in obesity has centered on infiltrating macrophages. The new data suggest that the fat cells themselves are not passive victims of inflammation but active participants, running their own DNA-sensing alarm machinery when mitochondrial damage accumulates. Targeting the adipocyte-intrinsic STING pathway, therefore, may be a more direct route to calming inflamed fat tissue than previously appreciated.

Then came the twist. When the researchers compared the two major fat depots in mice—visceral epididymal white adipose tissue, or eWAT, and subcutaneous inguinal white adipose tissue, or iWAT—they found that NRF2 activation behaved very differently in each. Transcriptomic profiling showed that NRF2 robustly suppressed inflammatory pathways in visceral eWAT, the depot most closely linked to metabolic disease, but had minimal impact in subcutaneous iWAT. At the molecular level, NRF2 downregulated STING in eWAT while paradoxically upregulating it in iWAT. The same drug, acting on the same transcription factor, produced opposite effects on the inflammatory alarm system depending on where the fat was located.

The explanation the authors propose centers on adipogenesis—the process by which precursor cells mature into fat cells—and the differing adipogenic potential of the two depots. Because iWAT precursor cells have a higher capacity for differentiation, NRF2 activation preferentially inhibited adipogenesis there, enriching the tissue in immature adipocytes. These immature cells, the study found, express high levels of STING. In other words, the drug’s anti-adipogenic effect in subcutaneous fat inadvertently expanded the population of STING-rich immature cells, blunting the anti-inflammatory benefit and even reversing the direction of STING regulation. Adipocyte differentiation status, the authors conclude, is a key determinant of how the NRF2-STING axis responds, providing a mechanistic account of fat depot heterogeneity that goes deeper than the usual anatomical descriptions.

The implications for precision medicine are considerable. NRF2 activators, including CDDO-Im and related synthetic triterpenoids, have already been tested in humans for kidney disease and other conditions, so the pharmacological toolkit exists. But this study suggests that a single systemic dose may not deliver uniform benefits across all fat depots, and that therapies aimed at the NRF2-STING axis in obesity may need to be tailored to depot biology—or paired with strategies that account for the adipogenic state of the target tissue. Visceral fat, the primary culprit in metabolic disease, appears to be the depot where NRF2 activation delivers its cleanest anti-inflammatory payoff.

As with any mouse study, caveats remain. Epididymal and inguinal depots in rodents are not perfect analogues of human visceral and subcutaneous fat, and the AAV-based depot-specific manipulations, while elegant, do not fully replicate human pharmacology. The work was supported by the National Natural Science Foundation of China and published open access, with all animal procedures approved by the Institutional Animal Care and Use Committee of Xinjiang Medical University under ARRIVE guidelines. Still, the study delivers a compelling conceptual advance: it connects three of the hottest themes in metabolic research—mitochondrial DNA leakage, innate immune DNA sensing, and the antioxidant response—into a single regulatory circuit, and it demonstrates that the circuit’s behavior depends on the developmental identity of the fat cells involved. If those findings translate, the humble NRF2 protein may become a linchpin target in the effort to convert inflamed, insulin-resistant fat tissue into calmer, healthier fat.

Subject of Research: NRF2 regulation of STING signaling in obesity-induced adipose inflammation

Article Title: NRF2 activation attenuates obesity-induced adipose inflammation by suppressing STING signaling, with depot-specific regulation

Article References: Xia, Y., Zhai, X., Zhao, L., Song, Y., Wang, H., Lu, X., Qiu, Y., Tuersun, M., Yang, B., Lu, Y., & Jiao, Y. (2026). NRF2 activation attenuates obesity-induced adipose inflammation by suppressing STING signaling, with depot-specific regulation. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-08887-4

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08887-4

Keywords: obesity, NRF2, STING, cGAS-STING pathway, adipose inflammation, mitochondrial DNA, oxidative stress, insulin resistance, fat depot heterogeneity, adipogenesis, CDDO-Im, white adipose tissue