Search

Selection of oligonucleotide aptamers involves consecutive rounds of affinity isolation of target-binding oligonucleotides from a random-sequence oligonucleotide library. Every next round produces an aptamer-enriched library with progressively higher fitness for tight binding to the target. The progress of enrichment can only be accurately assessed with bulk affinity assays in which a library is mixed with the target and one of two quantitative parameters, the fraction of the unbound library (

Large molecules can be generically separated from small ones, though partially and temporarily, in a pressure-driven flow inside a capillary. This transient incomplete separation has been only applied to species with diffusion coefficients different by at least an order of magnitude. Here, we demonstrate, for the first time, the analytical utility of transient incomplete separation for species with close diffusion coefficients. First, we prove

Accurate Constant via Transient Incomplete Separation (ACTIS) is a new method for finding the equilibrium dissociation constant Kd of a protein-small molecule complex based on transient incomplete separation of the complex from the unbound small molecule in a capillary. This separation is caused by differential transverse diffusion of the complex and the small molecule in a pressure-driven flow. The advection-diffusion processes underlying ACTIS can be described by a system of partial differential equations allowing for a virtual ACTIS instrument to be built and ACTIS to be studied in silico. The previous in silico studies show that large variations in the fluidic system geometry do not affect the accuracy of Kd determination, thus, proving that ACTIS is conceptually accurate. The conceptual accuracy does not preclude, however, instrumental inaccuracy caused by run-to-run signal drifts. Here we report on assembling a physical ACTIS instrument with a fluidic system that mimics the virtual one and proving the absence of signal drifts. Furthermore, we confirmed method ruggedness by assembling a second ACTIS instrument and comparing the results of experiments performed with both instruments in parallel. Despite some unintentional differences between the instruments (caused by tolerances in sizes, positions, etc.) and noticeable differences in their respective separagrams, we found that the Kd values determined for identical samples with these instruments were equal. Conclusively, the fluidic system presented here can serve as a template for reliable ACTIS instrumentation.

Partitioning of protein–DNA complexes from protein-unbound DNA is a key step in selection of DNA aptamers. Conceptually, the partitioning step is characterized by two parameters: transmittance for protein-bound DNA (binders) and transmittance for unbound DNA (nonbinders). Here we present the first study to reveal how these transmittances depend on experimental conditions; such studies are pivotal to the effective planning and control of selection. Our focus was capillary electrophoresis (CE) which is a partitioning approach of high efficiency. By combining a theoretical model and experimental data, we evaluated the dependence of transmittances of binders and nonbinders on the molecular weight of protein target in two modes of CE-based partitioning: nonequilibrium capillary electrophoresis of equilibrium mixtures (NECEEM) and ideal-filter capillary electrophoresis (IFCE). Our data suggest that as the molecular weight of the protein target decreases: (i) the transmittance for binders remains close to unity in NECEEM but decreases drastically in IFCE and (ii) the transmittance for nonbinders increases orders of magnitude in NECEEM but remains relatively stable at a very low level in IFCE. To determine the optimal CE conditions for a given size of protein target, a balance between transmittances of binders and nonbinder must be reached; such a balance would ensure the collection of binders of sufficient purity and quantity. We conclude that, as a rule of thumb, IFCE is preferable for large-size protein targets while NECEEM should be the method of choice for small-size protein targets