Beyond guanine quartets: cation-induced formation of homogenous and chimeric DNA tetraplexes incorporating iso-guanine and guanine

Background:iso-Guanine (iso-G) is the purine component of an isomeric Watson-Crick base pair that may have existed prebiotically. By comparing the abiotic molecular recognition properties of iso-G and its complement, iso-cytosine (iso-C), with those of genomic nucleotide bases, it may be possible to explain the exclusion of the iso-G-iso-C base pair from modern genomes. Whether a nucleobase forms quartets may have a key role in determining its functionality. Biotically, nucleic acid tetraplexes have been implicated in cellular functions; prebiotically, tetraplexes would probably interfere with replication. Recently, in vitro selection has yielded receptors and catalysts that incorporate G quartets. The versatility of these structures could be enhanced by expanding the range of bases that can form the quartet motif.Results: Native polyacrylamide gel electrophoresis of oligonucleotides bearing runs of iso-G provides strong support for tetraplex formation via cation-promoted DNA strand association. In particular, when strands of different lengths bearing the same iso-G tetrad recognition element were combined, five bands were observed after electrophoresis, corresponding to all possible heterotetraplexes with parallel strand alignment. An analogous experiment with a mixture of strands bearing iso-G or G tetrad recognition domains supports the existence of mixed iso-G/G tetraplexes with antiparallel strand alignment at chimeric junctions. iso-G tetraplex and quartet structure has also been probed by a photo-crosslinking experiment, ultra-violet spectroscopy and theoretical calculations.Conclusions: As iso-G and G both have a marked tendency to form tetraplexes, their tandem inclusion in genetic material may be problematic, leading to double-stranded DNA half composed of bases that have a tendency to auto-associate. The resulting density of ‘selfish’ bases could undermine Watson-Crick pair formation, especially in a prebiotic context devoid of enzymes. Nevertheless, the ability of iso-G to form mixed quartets with G may provide a basis for altering the properties of tetraplexes in the domain of artificial receptors or catalysts from in vitro selections.     

Kinetic analysis of an efficient DNA-dependent TNA polymerase

alpha-l-Threofuranosyl nucleoside triphosphates (tNTPs) are tetrafuranose nucleoside derivatives and potential progenitors of present-day beta-d-2'-deoxyribofuranosyl nucleoside triphosphates (dNTPs). Therminator DNA polymerase, a variant of the 9 degrees N DNA polymerase, is an efficient DNA-directed threosyl nucleic acid (TNA) polymerase. Here we report a detailed kinetic comparison of Therminator-catalyzed TNA and DNA syntheses. We examined the rate of single-nucleotide incorporation for all four tNTPs and dNTPs from a DNA primer-template complex and carried out parallel experiments with a chimeric DNA-TNA primer-DNA template containing five TNA residues at the primer 3'-terminus. Remarkably, no drop in the rate of TNA incorporation was observed in comparing the DNA-TNA primer to the all-DNA primer, suggesting that few primer-enzyme contacts are lost with a TNA primer. Moreover, comparison of the catalytic efficiency of TNA synthesis relative to DNA synthesis at the downstream positions reveals a difference of no greater than 5-fold in favor of the natural DNA substrate. This disparity becomes negligible when the TNA synthesis reaction mixture is supplemented with 1.25 mM MnCl(2). These results indicate that Therminator DNA polymerase can recognize both a TNA primer and tNTP substrates and is an effective catalyst of TNA polymerization despite changes in the geometry of the reactants.     

Biological activity of functionalized SiO2 thin films prepared by sol-gel method

In this work we investigated the biological properties of sol-gel films in aqueous medium. Functionalized silica films were prepared by the sol-gel process, from organically modified silicon alkoxides with amino or thiol groups. Covalent binding of proteins with different orientations according to the hydrophilic or hydrophobic character of the surface was studied. This binding occurred via a covalent coupling agent providing a very stable linkage. No denaturation was detected and a good detection of the antigen was observed. Immunoassays have demonstrated the biological activity of grafted antibodies.     

Enantiopure beta-hydroxysulfoxide derivatives as novel chiral auxiliaries in asymmetric biaryl Suzuki reactions

[reaction: see text] Highly diastereoselective biaryl Suzuki coupling reactions of (1R)-1-(2-iodo or bromophenyl)-2-(R)-(4-tolylsulfinyl)-1-ethanol derivatives with various aryl- or naphthylboronic acids (or esters) were performed with high yields (up to 99%) and an excellent control of the axial chirality (up to 98% de).     

Differential equations with singular fields

This paper investigates the well posedness of ordinary differential equations and more precisely the existence (or uniqueness) of a flow through explicit compactness estimates. Instead of assuming a bounded divergence condition on the vector field, a compressibility condition on the flow (bounded Jacobian) is considered. The main result provides existence under the condition that the vector field belongs to BV in dimension 2 and SBV in higher dimensions.     

Non-distortion of twin building lattices

We show that twin building lattices are undistorted in their ambient group; equivalently, the orbit map of the lattice to the product of the associated twin buildings is a quasi-isometric embedding. As a consequence, we provide an estimate of the quasi-flat rank of these lattices, which implies that there are infinitely many quasi-isometry classes of finitely presented simple groups. Finally, we describe how non-distortion of lattices is related to the integrability of the structural cocycle.     

On the convergence of formally diverging neural net-based classifiers

We present an analytical study of gradient descent algorithms applied to a classification problem in machine learning based on artificial neural networks. Our approach is based on entropy–entropy dissipation estimates that yield explicit rates. Specifically, as long as the neural nets remain within a set of “good classifiers”, we establish a striking feature of the algorithm: it mathematically diverges as the number of gradient descent iterations (“time”) goes to infinity but this divergence is only logarithmic, while the loss function vanishes polynomially. As a consequence, this algorithm still yields a classifier that exhibits good numerical performance and may even appear to converge.