Methicillin-resistant Staphylococcus aureus (MRSA) is an important human pathogen, which is cross-resistant to virtually all beta-lactam antibiotics. MRSA strains are defined by the presence of mecA gene. The transcription of mecA can be regulated by a sensor-inducer (MecR1) and a repressor (MecI), involving a unique series of proteolytic steps. The induction of mecA by MecR1 has been described as very inefficient and, as such, it is believed that optimal expression of beta-lactam resistance by MRSA requires a non-functional MecR1-MecI system. However, in a recent study, no correlation was found between the presence of functional MecR1-MecI and the level of b-lactam resistance in a representative collection of epidemic MRSA strains. Here, we demonstrate that the mecA regulatory locus consists, in fact, of an unusual three-component arrangement containing, in addition to mecR1-mecI, the up to now unrecognized mecR2 gene coding for an anti-repressor. The MecR2 function is essential for the full induction of mecA expression, compensating for the inefficient induction of mecA by MecR1 and enabling optimal expression of beta-lactam resistance in MRSA strains with functional mecR1-mecI regulatory genes. Our data shows that MecR2 interacts directly with MecI, destabilizing its binding to the mecA promoter, which results in the repressor inactivation by proteolytic cleavage, presumably mediated by native cytoplasmatic proteases. These observations point to a revision of the current model for the transcriptional control of mecA and open new avenues for the design of alternative therapeutic strategies for the treatment of MRSA infections. Moreover, these findings also provide important insights into the complex evolutionary pathways of antibiotic resistance and molecular mechanisms of transcriptional regulation in bacteria.