BackgroundSELEX is an iterative process in which highly diverse synthetic nucleic acid libraries are selected over many rounds to finally identify aptamers with desired properties. However, little is understood as how binders are enriched during the selection course. Next-generation sequencing offers the opportunity to open the black box and observe a large part of the population dynamics during the selection process.MethodologyWe have performed a semi-automated SELEX procedure on the model target streptavidin starting with a synthetic DNA oligonucleotide library and compared results obtained by the conventional analysis via cloning and Sanger sequencing with next-generation sequencing. In order to follow the population dynamics during the selection, pools from all selection rounds were barcoded and sequenced in parallel.ConclusionsHigh affinity aptamers can be readily identified simply by copy number enrichment in the first selection rounds. Based on our results, we suggest a new selection scheme that avoids a high number of iterative selection rounds while reducing time, PCR bias, and artifacts.

The dynamics of RNA hairpin tetraloops of the GNRA type [sequence G- any ribonucleotide (N)-purine (R)-A] was analyzed by fluorescence spectroscopy and by fluorescence-detected temperature-jump relaxation, using RNA oligomers with 2-aminopurine (2AP) substituted in two different positions of the loop sequence, Gp2APpApA (HP1) and GpAp2APpA (HP2), as indicator. The fluorescence of HP1 is much higher than that of HP2, indicating a lower degree of 2AP-stacking in HP1. Addition of Mg(2+) or Ca(2+) ions leads to an increase of fluorescence in HP1, whereas a decrease of fluorescence is observed in HP2. In both cases at least two ion-binding equilibria are required to fit titration data. T-jump experiments using fluorescence detection show a relaxation process with a time constant of 22 micros for HP1, whereas two relaxation processes with time constants 5 and 41 micros, are found for HP2. These results clearly demonstrate the existence of more than the single conformation state detected by NMR analysis. The T-jump amplitudes decrease with increasing bivalent ion concentration, indicating that one of the states is favored in the presence of bivalent ions. The loop relaxation processes are slower than standard stacking processes, probably because of activation barriers imposed by a restricted mobility of loop residues, and are assigned to a stacking rearrangement, probably between the 5' and the 3'-side. A similar process has been observed previously for the anticodon loop of tRNA(Phe). The rate constants of the transition are in the range of 10(4) s(-1) in the case of HP1. The data demonstrate the existence of structures that are not resolved by standard NMR because of fast exchange and are not found by X-ray analysis because of restrictions by crystal packing.

Guanine quadruplex (G-quadruplex) motifs in the 5′ untranslated region (5′-UTR) of mRNAs were recently shown to influence the efficiency of translation. In the present study, we investigate the interaction between cellular proteins and the G-quadruplexes located in two mRNAs (MMP16 and ARPC2). Formation of the G-quadruplexes was confirmed by biophysical characterization and the inhibitory activity on translation was shown by luciferase reporter assays. In experiments with whole cell extracts from different eukaryotic cell lines, G-quadruplex-binding proteins were isolated by pull-down assays and subsequently identified by matrix-assisted laser desorption ionization-time of flight mass spectrometry. The binding partners of the RNA G-quadruplexes we discovered included several heterogenous nuclear ribonucleoproteins, ribosomal proteins, and splicing factors, as well as other proteins that have previously not been described to interact with nucleic acids. While most of the proteins were specific for either of the investigated G-quadruplexes, some of them bound to both motifs. Selected candidate proteins were subsequently produced by recombinant expression and dissociation constants for the interaction between the proteins and RNA G-quadruplexes in the low nanomolar range were determined by surface plasmon resonance spectroscopy. The present study may thus help to increase our understanding of the mechanisms by which G-quadruplexes regulate translation.

Conformational changes of the hammerhead ribozyme were examined by fluorescence changes of 2-aminopurine riboside incorporated either in the substrate or in the ribozyme. Fluorescence changes could be observed for both the substituted substrate and ribozyme upon complex formation, indicating a different environment for the 2-aminopurine in the complex. Ribozyme-substrate constructs for ciscleavage containing 2-aminopurine at various sites were used for the determination of binding constants of Mg2+ and Ca2+. Depending upon the site of 2-aminopurine substitutions, the fluorescence intensity upon addition of Mg2+ or Ca2+ was reduced by 0-50%. The measurements were performed in high ionic strength buffers such that base pairing in the helical regions is expected to be complete. With three of the ribozymes, the dependence of the fluorescence emission as a function of Mg2+ concentration could be fitted by single binding processes, whereas for the two remaining ribozymes a second binding process needed to be included. The binding constants range from 7600 M-1 down to 12 M-1 in 75 mM Tris-HCl (pH 7.5) and indicate the presence of multiple binding sites in the ribozymes with varying degrees of affinity toward the metal ions. Mg2+ binding constants determined in the same buffer from the Mg2+ dependence of the cleavage rate are of the order of 100 M-1; thus, Mg2+ sites directly involved in catalysis are of intermediate affinity. The ribozyme containing 2-aminopurine in loop III demonstrated the highest binding constant whereas the ribozyme with a 2-aminopurine next to a 2'-deoxy-2'-aminocytosine at the cleavage site exhibited only low metal ion affinity. The data obtained for Ca2+ are very similar to those found for Mg2+. This approach provides a first set of data describing a Mg2+ binding topography to hammerhead RNA molecules and should be useful for the analysis of other RNA molecules.

No abstract