**Background:** Reactive oxygen species (ROS) were once considered mere metabolic byproducts, but the discovery of the NADPH oxidase (Nox) enzyme family—dedicated solely to ROS production—changed this view. Among the seven Nox isoforms, Nox3 was initially discovered in the inner ear and long thought to be restricted to that organ. This review by Marc Herb comprehensively examines Nox3, covering its discovery, structure, regulation, expression across tissues, physiological functions, and roles in disease. The author aims to dispel the "inner ear stigma" and highlight Nox3 as a versatile and underappreciated Nox isoform.
**Methods:** This is a narrative review that synthesizes findings from over 200 studies on Nox3. The author systematically organizes the literature into sections on Nox3 structure and subunits, tissue and cellular expression, activation and regulation, physiological functions, and disease involvement. The review includes data from in vitro co-expression systems, animal models (primarily mice and rats), and human genetic studies (GWAS). The author critically evaluates the quality of evidence, noting whether studies used genetic knock-down/knock-out or only correlative mRNA/protein expression.
**Key Results:**
- **Structure and Subunits:** Nox3 is a 568-amino-acid protein with 58% sequence similarity to Nox2. It forms a heterodimer with p22phox, which is essential for maturation and plasma membrane localization. Nox3 is uniquely flexible in its use of regulatory subunits: it can be activated by p47phox, NOXO1, NOXA1, or p67phox alone or in combination, and shows constitutive basal activity even without subunits. Rac is not essential but can enhance activity with certain subunit combinations.
- **Expression:** Contrary to the "inner ear only" dogma, Nox3 mRNA or protein has been detected in lung, liver, testes, white adipose tissue, tongue, skeletal muscle, brain, kidney, placenta, and various cancer cell lines. In the inner ear, Nox3 is expressed in non-sensory epithelial cells of the vestibular system (saccule, utricle, semicircular canals) and in cochlear supporting cells (Deiters' cells, Claudius' cells, inner sulcus cells), but not in hair cells.
- **Functions:** In the vestibular system, Nox3-derived ROS are crucial for otoconia formation via disulfide-linkage of otoconin-90/95. Nox3-deficient mice lack otoconia and show head-tilting and balance deficits. In the cochlea, Nox3 has no known physiological role; instead, its overproduction of ROS contributes to noise-, cisplatin-, and age-induced hearing loss. In the liver, Nox3 mediates insulin signaling (ERK1/2, VEGF-A) and TNF-induced insulin resistance via JNK1/2 and p38 pathways. In the lung, Nox3 contributes to emphysema and hyperoxia-induced injury, with TLR4 signaling inhibiting Nox3 via STAT3 and Hsp70.
- **Disease Associations:** GWAS have linked NOX3 SNPs to noise-induced hearing loss, pulmonary hypertension, primary graft dysfunction after lung transplant, abdominal aortic aneurysm, and anti-thyroid drug-induced agranulocytosis. In animal models, Nox3 is implicated in diabetic liver disease, stroke, heart failure, and renal disease.
- **Therapeutic Targeting:** siRNA against Nox3 has shown promise in reducing cisplatin-induced hearing loss in rats, lowering ABR threshold shifts from 35 dB to 23 dB. Other approaches include targeting TRPV1, STAT1, or using compounds like KR-22332, fenofibrate, and dexamethasone to reduce Nox3 expression and ROS production.
**Clinical Implications:** Nox3 is a potential therapeutic target for several conditions, particularly sensorineural hearing loss (noise- and drug-induced), diabetic complications, and lung diseases. However, the author emphasizes that most studies lack direct genetic evidence (knock-down/knock-out) and ROS measurements, relying instead on correlative mRNA expression. Future research should focus on ex vivo human tissues, confirm Nox3 as the ROS source, and explore its roles beyond the inner ear, especially in immunity and infection.