Our studies on the resistance mechanisms and chemical modifications of aminoglycoside antibiotics led to the synthesis of arbekacin (ABK), which was refractory to most aminoglycoside-modifying enzymes in resistant bacteria. In 1990, ABK was launched into Japan as a chemotherapeutic agent for the treatment of infection caused by methicillin-resistant Staphylococcus aureus (MRSA). By 1992 only a few MRSA strains moderately resistant to ABK (6.25-12.5μg/ml) were clinically isolated. ABK was modified by reaction with excess of an enzyme preparation extracted from an ABK-resistant strain and three inactivated products, mainly ABK 2"-phosphate along with small amounts of 6'-N-acetyl-ABK and the doubly modified ABK were obtained. Based on these results, replacement of the 2"-OH by NH_2 in dibekacin (DKB) or in ABK was designed to obtain potent active derivatives against MRSA. Conversion of the 2"-OH by DMSO-DCC oxidation followed by reductive amination with NH_4OAc-NaBH_3CN gave 2"-amino-2"-deoxy-DKB (D1) and -ABK (A1). Their 5-deoxy, 5-epifluoro and 5-epiamino derivatives were also synthesized. All 2"-amino-ABK derivatives showed excellent activities against MRSA and Gram-negative bacteria, as expected. Among them, A1 having low toxicity was selected as a candidate for anti-MRSA agent.