NOX4 is an enigmatic member of the NOX (NADPH oxidase) family of ROS (reactive oxygen species)-generating NADPH oxidases. NOX4 has a wide tissue distribution, but the physiological function and activation mechanisms are largely unknown, and its pharmacology is poorly understood. We have generated cell lines expressing NOX4 upon tetracycline induction. Tetracycline induced a rapid increase in NOX4 mRNA (1 h) followed closely (2 h) by a release of ROS. Upon tetracycline withdrawal, NOX4 mRNA levels and ROS release decreased rapidly (<24 h). In membrane preparations, NOX4 activity was selective for NADPH over NADH and did not require the addition of cytosol. The pharmacological profile of NOX4 was distinct from other NOX isoforms: DPI (diphenyleneiodonium chloride) and thioridazine inhibited the enzyme efficiently, whereas apocynin and gliotoxin did not (IC(50)>100 muM). The pattern of NOX4-dependent ROS generation was unique: (i) ROS release upon NOX4 induction was spontaneous without need for a stimulus, and (ii) the type of ROS released from NOX4-expressing cells was H(2)O(2), whereas superoxide (O(2)(-)) was almost undetectable. Probes that allow detection of intracellular O(2)(-) generation yielded differential results: DHE (dihydroethidium) fluorescence and ACP (1-acetoxy-3-carboxy-2,2,5,5-tetramethylpyrrolidine) ESR measurements did not detect any NOX4 signal, whereas a robust signal was observed with NBT. Thus NOX4 probably generates O(2)(-) within an intracellular compartment that is accessible to NBT (Nitro Blue Tetrazolium), but not to DHE or ACP. In conclusion, NOX4 has a distinct pharmacology and pattern of ROS generation. The close correlation between NOX4 mRNA and ROS generation might hint towards a function as an inducible NOX isoform.