The bacterial tryptophan synthase bienzyme complexes (with subunit composition alpha 2 beta 2) catalyze the last two steps in the biosynthesis of L-tryptophan. For L-tryptophan synthesis, indole, the common metabolite, must be transferred by some mechanism from the alpha-catalytic site to the beta-catalytic site. The X-ray structure of the Salmonella typhimurium tryptophan synthase shows the catalytic sites of each alpha-beta subunit pair are connected by a 25-30 A long tunnel [Hyde, C. C., Ahmed, S. A., Padlan, E. A., Miles, E. W., & Davies, D. R. (1988) J. Biol. Chem. 263, 17857-17871]. Since the S. typhimurium and Escherichia coli enzymes have nearly identical sequences, the E. coli enzyme must have a similar tunnel. Herein, rapid kinetic studies in combination with chemical probes that signal the bond formation step between indole (or nucleophilic indole analogues) and the alpha-aminoacrylate Schiff base intermediate, E(A-A), bound to the beta-site are used to investigate tunnel function in the E. coli enzyme. If the tunnel is the physical conduit for the transfer of indole from the alpha-site to the beta-site, then ligands that block the tunnel should also inhibit the rate at which indole and indole analogues from external solution react with E(A-A). We have found that when D,L-alpha-glycerol 3-phosphate (GP) is bound to the alpha-site, the rate of reaction of indole and nucleophilic indole analogues with E(A-A) is strongly inhibited. These compounds appear to gain access to the beta-site via the alpha-site and the tunnel, and this access is blocked by the binding of GP to the alpha-site. However, when small nucleophiles such as hydroxylamine, hydrazine, or N-methylhydroxylamine are substituted for indole, the rate of quinonoid formation is only slightly affected by the binding of GP. Furthermore, the reactions of L-serine and L-tryptophan with alpha 2 beta 2 show only small rate effects due to the binding of GP. From these experiments, we draw the following conclusions: (1) L-Serine and L-tryptophan gain access to the beta-site of alpha 2 beta 2 directly from solution. (2) The small effects of GP on the rates of the L-serine and L-tryptophan reactions are due to GP-mediated allosteric interactions between the alpha- and beta-sites.(ABSTRACT TRUNCATED AT 400 WORDS)

The allosteric behavior of the Co(II)-substituted insulin hexamer has been investigated using electronic spectroscopy to study the binding of different phenolic analogues and singly charged anions to effector sites on the protein. This work presents the first detailed, quantitative analysis of the ligand-induced T- to R-state allosteric transition of the insulin hexamer. Recent studies have established that there are two ligand binding processes which stabilize the R-state conformation of the Co(II)-substituted hexamer: the binding of cyclic organic molecules to the six protein pockets present in the Zn(II)-R6 insulin hexamer [Derewenda, U., Derewenda, Z., Dodson, E. J., Dodson, G. G., Reynolds, C. D., Smith, G. D., Sparks, C., & Swensen, D. (1989) Nature 338, 594-596] and the coordination of singly charged anions to the His(B10) metal sites [Brader, M.L., Kaarsholm, N.C., Lee, W.K., & Dunn, M.F. (1991) Biochemistry 30, 6636-6645]. The R6 insulin hexamer is stabilized by heterotropic interactions between the hydrophobic protein pockets and the coordination sites of the His(B10)-bound metal ions. The binding studies with 4-hydroxybenzamide, m-cresol, resorcinol, and phenol presented herein show that, in the absence of inorganic anions, the 4-hydroxybenzamide-induced transition, with a Hill number of 2.8, is the most cooperative, followed by m-cresol, phenol, and resorcinol with Hill numbers of 1.8, 1.4, and 1.2, respectively. The relative effectiveness of these ligands in shifting the allosteric equilibrium in favor of the Co(II)-R6 hexamer was found to be resorcinol > phenol > 4-hydroxybenzamide > m-cresol.(ABSTRACT TRUNCATED AT 250 WORDS)

No abstract