Fluorine is the thirteenth most abundant element in the earth's crust, but fluoride concentrations in surface water are low and fluorinated metabolites are extremely rare. The fluoride ion is a potent nucleophile in its desolvated state, but is tightly hydrated in water and effectively inert. Low availability and a lack of chemical reactivity have largely excluded fluoride from biochemistry: in particular, fluorine's high redox potential precludes the haloperoxidase-type mechanism used in the metabolic incorporation of chloride and bromide ions. But fluorinated chemicals are growing in industrial importance, with applications in pharmaceuticals, agrochemicals and materials products. Reactive fluorination reagents requiring specialist process technologies are needed in industry and, although biological catalysts for these processes are highly sought after, only one enzyme that can convert fluoride to organic fluorine has been described. Streptomyces cattleya can form carbon-fluorine bonds and must therefore have evolved an enzyme able to overcome the chemical challenges of using aqueous fluoride. Here we report the sequence and three-dimensional structure of the first native fluorination enzyme, 5'-fluoro-5'-deoxyadenosine synthase, from this organism. Both substrate and products have been observed bound to the enzyme, enabling us to propose a nucleophilic substitution mechanism for this biological fluorination reaction.

(4-Hydroxy)mandelate synthase (HmaS) and (4-hydroxyphenyl)pyruvate dioxygenase (HPPD) are two ␣-keto acid dependent mononuclear non-heme iron enzymes that use the same substrate, (4-hydroxyphenyl)pyruvate, but exhibit two different general reactivities. HmaS performs hydrogen-atom abstraction to yield benzylic hydroxylated product (S)-(4-hydroxy)mandelate, whereas HPPD utilizes an electrophilic attack mechanism that results in aromatic hydroxylated product homogentisate. These enzymes provide a unique opportunity to directly evaluate the similarities and differences in the reaction pathways used for these two reactivities. An Fe II methodology using CD, magnetic CD, and variable-temperature, variable-field magnetic CD spectroscopies was applied to HmaS and compared with that for HPPD to evaluate the factors that affect substrate interactions at the active site and to correlate these to the different reactivities exhibited by HmaS and HPPD to the same substrate. Combined with density functional theory calculations, we found that HmaS and HPPD have similar substrate-bound complexes and that the role of the protein pocket in determining the different reactivities exhibited by these enzymes (hydrogen-atom abstraction vs. aromatic electrophilic attack) is to properly orient the substrate, allowing for ligand field geometric changes along the reaction coordinate. Elongation of the Fe IV AO bond in the transition state leads to dominant Fe III OO •؊ character, which significantly contributes to the reactivity with either the aromatic -system or the COH -bond. magnetic circular dichroism ͉ density functional calculations ͉ reaction coordinates

The mechanism of Lewis acid catalysed hetero-Michael addition reactions of weakly basic nucleophiles to alpha,beta-unsaturated ketones was investigated. Protons, rather than metal ions, were identified as the active catalysts. Other mechanisms have been ruled out by analyses of side products and of stoichiometric enone-catalyst mixtures and by the use of radical inhibitors. No evidence for the involvement of pi-olefin-metal complexes or for carbonyl-metal-ion interactions was obtained. The reactions did not proceed in the presence of the non-coordinating base 2,6-di-tert-butylpyridine. An excellent correlation of catalytic activities with cation hydrolysis constants was obtained. Different reactivities of mono- and dicarbonyl substrates have been rationalised. A (1)H NMR probe for the assessment of proton generation was established and Lewis acids have been classified according to their propensity to hydrolyse in organic solvents. Brønsted acid-catalysed conjugate addition reactions of nitrogen, oxygen, sulfur and carbon nucleophiles are developed and implications for asymmetric Lewis acid catalysis are discussed.

A system is reported for the recombinant expression of individual ketoreductase (KR) domains from modular polyketide synthases (PKSs) and scrutiny of their intrinsic specificity and stereospecificity toward surrogate diketide substrates. The eryKR(1) and the tylKR(1) domains, derived from the first extension module of the erythromycin PKS and the tylosin PKS, respectively, both catalyzed reduction of (2R, S)-2-methyl-3-oxopentanoic acid N-acetylcysteamine thioester, with complete stereoselectivity and stereospecificity, even though the substrate is not tethered to an acyl carrier protein or an intact PKS multienzyme. In contrast, and to varying degrees, the isolated enzymes eryKR(2), eryKR(5), and eryKR(6) exercised poorer control over substrate selection and the stereochemical course of ketoreduction. These data, together with modeling of diketide binding to KR(1) and KR(2), demonstrate the fine energetic balance between alternative modes of presentation of ketoacylthioester substrates to KR active sites.

No abstract