D‐Hydantoinases (E.C.3.5.2.2) are commercially valuable enzymes involved in the production of D‐amino acids. However, commercial exploitation of the biological process is rare, mainly because sufficient details are not available on the efficient production of these enzymes by microorganisms. In the present study, Agrobacterium radiobacter was used as the source of D‐hydantoinase and its production was optimized with inexpensive carbon and nitrogen sources. The four media components selected to study their effect on biomass and/or enzyme activities were molasses, ammonium nitrate, sodium di‐hydrogen orthophosphate, and manganese chloride. With the use of an empirical modeling technique (response surface method), we have optimized both biomass and enzyme production in this organism, with a minimal number of batches. Experiments were performed with optimized media components to validate the model. The maximum level of enzyme and biomass obtained was 35 U/mL and 1.69 mg/mL, respectively. © 1997 John Wiley & Sons, Inc. Biotechnol Bioeng 55: 148–154, 1997.

D-Hydantoinases (E.C.3.5.2.2) are commerciallycan subsequently be cleaved either enzymatically by valuable enzymes involved in the production of D-amino amidohydrolase (Olivieri et al., 1979; Runser et al., 1990; acids. However, commercial exploitation of the biological Syldatk et al., 1987) or by mild acid hydrolysis (Takaprocess is rare, mainly because sufficient details are not hashi et al., 1979) to the D-amino acids.available on the efficient production of these enzymes Several microorganisms have been identified which by microorganisms. In the present study, Agrobacterium radiobacter was used as the source of D-hydantoinase produce either or both of the enzymes involved in the and its production was optimized with inexpensive carconversion (Morin et al., 1987;Yamada et al., 1978). bon and nitrogen sources. The four media components Three different types of hydantoinases are known: Dselected to study their effect on biomass and/or enzyme specific (Vogels and Van der Drift 1976; Yokozeki et activities were molasses, ammonium nitrate, sodium dial., 1987); L-specific (Yamashiro et al., 1988); and nonhydrogen orthophosphate, and manganese chloride. With the use of an empirical modeling technique (re-specific (Mö ller et al., 1988; Vogels and Van der Drift sponse surface method), we have optimized both bio-1976; Watabe, 1992). The hydantoinase from Agrobacmass and enzyme production in this organism, with a terium radiobacter is D-specific (Vogels and Van der minimal number of batches. Experiments were per-Drift, 1976).formed with optimized media components to validate the D-Hydantoinase was first reported by Yamada et al. model. The maximum level of enzyme and biomass obtained was 35 U/mL and 1.69 mg/mL, respectively.organisms, D-hydantoinase and dihydropyimidinase acmodeling; factorial design; amino acids tivities are attributed to the same enzyme (Vogels and Van der Drift, 1976) and in others the activities are carried out by different enzymes. The dihydropyrimi-