Chemical and Biochemical Approaches for the Synthesis of Substituted Dihydroxybutanones and Di- and Tri-Hydroxypentanones.

Polyhydroxylated compounds are building blocks for the synthesis of carbohydrates and other natural products. Their synthesis is mainly achieved by different synthetic versions of aldol-coupling reactions, catalyzed either by organocatalysts, enzymes, or metal-organic catalysts. We have investigated the formation of 1,4-substituted 2,3-dihydroxybutan-1-one derivatives from para- and meta-substituted phenylacetaldehydes by three distinctly different strategies. The first involved a direct aldol reaction with hydroxyacetone, dihydroxyacetone, or 2-hydroxyacetophenone, catalyzed by the cinchona derivative cinchonine. The second was reductive cross-coupling with methyl- or phenylglyoxal promoted by SmI2, resulting in either 5-substituted 3,4-dihydroxypentan-2-ones or 1,4 bis-phenyl-substituted butanones, respectively. Finally, in the third case, aldolase catalysis was employed for synthesis of the corresponding 1,3,4-trihydroxylated pentan-2-one derivatives. The organocatalytic route with cinchonine generated distereomerically enriched syn-products (de = 60-99%), with moderate enantiomeric excesses (ee = 43-56%) but did not produce aldols with either hydroxyacetone or dihydroxyacetone as donor ketones. The SmI2-promoted reductive cross-coupling generated product mixtures with diastereomeric and enantiomeric ratios close to unity. This route allowed for the production of both 1-methyl- and 1-phenyl-substituted 2,3-dihydroxybutanones at yields between 40-60%. Finally, the biocatalytic approach resulted in enantiopure syn-(3 R,4 S) 1,3,4-trihydroxypentan-2-ones.

Long Time-Scale Atomistic Simulations of the Structure and Dynamics of Transcription Factor-DNA Recognition.

Recent years have witnessed an explosion of interest in computational studies of DNA binding proteins, including both coarse-grained and atomistic simulations of transcription factor-DNA recognition, to understand how these transcription factors recognize their binding sites on the DNA with such exquisite specificity. The present study performs microsecond time scale all-atom simulations of the dimeric form of the lactose repressor (LacI), both in the absence of any DNA and in the presence of both specific and nonspecific complexes, considering three different DNA sequences. We examine, specifically, the conformational differences between specific and nonspecific protein-DNA interactions, as well as the behavior of the helix-turn-helix motif of LacI when interacting with the DNA. Our simulations suggest that stable LacI binding occurs primarily to bent A-form DNA, with a loss of LacI conformational entropy and optimization of correlated conformational equilibria across the protein. In addition, binding to the specific operator sequence involves a slightly larger number of stabilizing DNA-protein hydrogen bonds (in comparison to nonspecific complexes), which may account for the experimentally observed specificity for this operator. In doing so, our simulations provide a detailed atomistic description of potential structural drivers for LacI selectivity.