The primary role of Tyr225 in the aspartate aminotransferase mechanism is to provide a hydrogen bond to stabilize the 3'O- functionality of bound pyridoxal phosphate. The strength of this hydrogen bond is perturbed by replacement of Tyr225 with 3-fluoro-L-tyrosine (FlTyr) by in vitro transcription/translation. This mutant enzyme exhibits kcat/values that are near to those of wild type enzyme; however, the kcat/vs pH profile is much sharper with similar pKas of approximately 7.5 for both the ascending and descending limbs. The pKas are assigned to the endocyclic proton of the internal aldimine and to the bridging hydrogen bond, respectively. The pKas in the kcat vs pH profile of 7.2 and 8.7 are assigned to the epsilon-NH3+ of lysine 258 and to the endocyclic protons of the ketimine complex, respectively. Arginine 292 forms a salt bridge with the beta-COOH of the substrate, aspartate. An improvement on the earlier attempt to invert the substrate charge specificity via R292D mutation-induced arginine transaminase activity [Cronin, C. N., & Kirsch, J. F. (1988) Biochemistry 27, 4572-4579] is described. Here Arg292 is replaced with homoglutamate (R292hoGlu). This construct exhibits 6.8 x 10(4)-fold greater activity for the cationic substrate D,L-[Calpha-3H]-alpha-amino-beta-guanidinopropionic acid (D,L-[Calpha-3H]AGPA) than does wild type enzyme. The gain in selectivity for this substrate is at least 4500-fold greater than that achieved in the 1988 experiment, i.e., [(kcat/KM)R292hoGlu/(kcat/KM)WT (D,L-[Calpha-3H]AGPA)] >/= 4500 x [(kcat/KM)R292D/(kcat/KM)WT (L-arginine)]. The value of (kcat/KM)R292D is 0.43 M-1 s-1 with L-Arg while (kcat/KM)R292hoGlu is 29 M-1 s-1 with D,L-[Calpha-3H]AGPA (it is assumed that the D-enantiomer is unreactive). The latter value is the lower limit because of the uncertain value of 3H kinetic isotope effect.