The mechanism of inactivation of pig brain γ-aminobutyric acid (GABA) aminotransferase by the antibiotic l-cycloserine was investigated. l-Cycloserine is a time-dependent inactivator of GABA aminotransferase; no enzyme activity returns upon gel filtration or dialysis. Treatment of GABA aminotransferase with [14C]-l-cycloserine, followed by rapid gel filtration, gives enzyme containing l.l equiv of radioactivity bound. Dialysis or denaturation by acid, base, or urea releases the radioactivity. Inactivation of [3H]pyridoxal 5‘-phosphate (PLP)-reconstituted GABA aminotransferase with l-cycloserine followed by dialysis or denaturation also leads to the release of radioactivity from the enzyme. Both the released [14C]- and [3H]-labeled adducts comigrate by HPLC, suggesting that the inactivation adduct is a condensation product of l-cycloserine with the PLP coenzyme. By HPLC comparison, it was shown that the radiolabeled adduct is not PLP, PMP, PLP oxime, or 4-[3-hydroxy-2-methyl-5-(phosphooxymethyl)-4-pyridinyl]-2-oxo-3-butenoic acid (20), the expected product of an enamine-type inactivation mechanism. On the basis of the stability of the released adduct to acid and base and its UV−visible spectrum, which has the appearance of a PMP analogue, a simple Schiff base between PLP and cycloserine also was excluded. HPLC of the cycloserine−coenzyme adduct had a retention time very similar to that of the gabaculine−coenzyme adduct. Electrospray ionization tandem mass spectrometry of the isolated cycloserine−coenzyme adduct is consistent with a structure that is one of the tautomeric forms of the Schiff base between PMP and oxidized cycloserine (21).

As a mechanism-based inactivator of PLP-enzymes, (S)-4-amino-4,5-dihydro-2-thiophenecarboxylic acid (SADTA) was cocrystallized with Escherichia coli aspartate aminotransferase (l-AspAT) at a series of pH values ranging from 6 to 8. Five structural models with high resolution (1.4-1.85 A) were obtained for l-AspAT-SADTA complexes at pH 6.0, 6.5, 7.0, 7.5, and 8.0. Electron densities of the models showed that two different adducts had formed in the active sites. One adduct was formed from SADTA covalently linked to pyridoxal 5'-phosphate (PLP) while the other adduct was formed with the inhibitor covalently linked to Lysine246,1 the active site lysine. Moreover, there is a strong indication based on the electron densities that the occurrence of the two adducts is pH dependent. We conclude that SADTA inactivates l-AspAT via two different mechanisms based on the binding direction of the inactivator. Additionally, the structural models also show pH dependence of the protein structure itself, which provided detailed mechanistic implications for l-AspAT.

(S)-4-Amino-4,5-dihydro-2-thiophenecarboxylic acid ((S)-6) was previously synthesized (Adams, J. L.; Chen, T. M.; Metcalf, B. W. J. Org. Chem. 1985, 50, 2730−2736.) as a heterocyclic mimic of the natural product gabaculine (5-amino-1,3-cyclohexadienylcarboxylic acid), a mechanism-based inactivator of γ-aminobutyric acid aminotransferase (GABA-AT) (Rando, R. R. Biochemistry 1977, 16, 4604). Inactivation of GABA-AT by (S)-6 is time-dependent and protected by substrate. Two methods were utilized to demonstrate that, in addition to inactivation, about 0.7 equiv per inactivation event undergoes transamination. Inactivation results from the reaction of (S)-6 with the pyridoxal 5‘-phosphate (PLP) cofactor. The adduct was isolated and characterized by ultraviolet−visible spectroscopy, electrospray mass spectrometry, and tandem mass spectrometry. All of the results support a structure (11) that derives from the predicted aromatization inactivation mechanism (Scheme ) originally proposed by Metcalf and co-workers for this compound. This is only the third example, besides gabaculine and l-cycloserine, of an inactivator of a PLP-dependent enzyme that acts via an aromatization mechanism.

No abstract

No abstract

No abstract

No abstract

Abstract— After dissolution of the membrane structure of chromatophores from Rhodospirillum rubrum, Rhodopseudomonas spheroides, and the R‐26 mutant of Rhodopseudomonas spheroides, active phototrap complexes from each have been purified by a column electrophoresis procedure. Phospholipids and transition metals were well separated from the phototrap complex in all three systems. The purified R. rubrum phototrap complex retained a full complement of antenna bacteriochlorophyll and carotenoid pigments which had nearly the same absorbance spectra as in the intact cell, and which delivered absorbed light energy to the phototrap with just as high efficiency as in the intact cell. Sodium dodecyl sulfate (SDS) disc gel electrophoresis using Tris buffer showed that these preparations often contained only two prominent polypeptides of 30,000 ± 2000 and 12,000 ± 4000 mol. wt., and a lesser amount of a third polypeptide of 21,000 ± 2000 mol. wt. The phototrap complexes prepared from the wild type and the R‐26 mutant of R. spheroides were similar, in that a partial separation from antenna pigments occurred during column electrophoresis. Both complexes had prominent polypeptides of 24,000 ± 2000 and 21,000 ± 2000 mol. wt., but no polypeptide of 30,000 mol. wt remained after electrophoresis. A third major polypeptide occurred with a mol. wt. of about 12,000 but seemed identifiable with an incompletely separated antenna pigment fraction. The phototrap complex prepared from the R‐26 mutant had a typical reaction center spectrum. In the case of wild type R. spheroides purification, two distinct protein‐pigment complexes separated. Although the absorbance of the bacteriochlorophyll and carotenoid pigments were little changed from those of the in vivo system, different polypeptides in the two fractions were observed by SDS disc gel electrophoresis; only one fraction seemed to be intimately related with the phototrap complex.