F10506 (tacrolimus), a macrolide compound with immunosuppressant activity, has been proven to have clinical importance and has been manufactured industrially since 1993 by using mutants with high FK506-production ability; these mutants have been developed from the wild strain Streptomyces tsukubaensis No. 9993. FR900525 is one of the by-products of FK506 production. However, there was no effective industrial method to separate FR900525 from FK506 due to the structural similarity between the two compounds. Therefore, reducing the level of FR900525 was a serious problem in the industrial strain A. In this study, we aimed to reduce the FR900525 production. We first determined that pipecolic acid level was a critical parameter for controlling FR900525 production in strain A. S-(2-Aminoethyl) L-cysteine (AEC)-resistant mutants has been reported to increase lysine productivity successfully in a variety of lysine-producing microorganisms. Therefore, next, we applied a selection of AEC -resistant mutants to enhance pipecolic acid biosynthesis. Finally, four AEC -resistant mutants were obtained from strain A using ultraviolet irradiation, and three of them showed less FR900525 productivity compared to the parental strain A. Our findings indicated that AEC resistance was effective phenotype marker for increasing pipecolic acid productivity and for reducing FR900525 production in S. tsukubaensis. Thus, our study provides an efficient method for reducing FR90025 level during FK506 biosynthesis. (C) 2017, The Society for Biotechnology, Japan. All rights reserved.