Structure of the Complex between Pyridoxal 5′-Phosphate and the Tyrosine 225 to Phenylalanine Mutant of Escherichia coli Aspartate Aminotransferase Determined by Isotope-Edited Classical Raman Difference Spectroscopy

The azomethine (Schiff base) linkage between the ϵ-amino group of active-site lysine 258 and the carbonyl moiety of enzyme-bound pyridoxal 5′-phosphate (PLP) normally exhibits absorbance maxima at ca. 360 (high-pH form) or ca. 430 nm (low-pH form). However, the absorbance maximum is shifted from 358 to 386 nm, a value which is similar to that of free PLP (λ_max = 388 nm), in a mutant form of Escherichia coli aspartate aminotransferase (AATase) in which tyrosine 225, which normally donates a hydrogen bond to the phenolate function of PLP, has been replaced with phenylalanine (Y225F). This spectral shift suggested that PLP binds to Y225F as the free aldehyde. The following evidence from isotope-edited classical Raman spectroscopy proves conclusively that the near-UV spectrum is anomalous and that PLP is bound to Y225F as a Schiff base: (1) A strong cofactor peak at 1630 cm⁻¹ in the holoenzyme-minus-apoenzyme difference spectrum of the unprotonated form of Y225F is red-shifted by 18 cm⁻¹ in enzyme labeled with ¹⁵N at lysine 258 and other positions. (2) This isotope-induced red shift is similar to that observed in the unprotonated form of the model Schiff base, PLP-valine. (3) The Raman spectrum of Y225F is unchanged in H₂¹⁸O, while peaks at ca. 1670 cm⁻¹ in the spectrum of free PLP or in that of a mutant of AATase in which Lys-258 is replaced with Ala, are red-shifted by ca. 30 cm⁻¹ in H₂¹⁸O. A molecular orbital explanation for the anomalous red shift in the near-UV spectrum of the Y225F-PLP complex is proposed. In addition, the O3′ atom of PLP is found to simultaneously accept hydrogen bonds from Tyr-225 and Lys-258 in the protonated internal aldimine form of AATase.