Biosensor-surface plasmon resonance: A strategy to help establish a new generation RNA-specific small molecules

Abstract

Biosensor surface plasmon resonance (SPR) is a highly sensitive technique and is most commonly used to decipher the interactions of biological systems including proteins and nucleic acids. Throughout the years, there have been significant efforts to develop SPR assays for studying protein-protein interactions, protein-DNA interactions, as well as small molecules to target DNAs that are of therapeutic interest. With the explosion of discovery of new RNA structures and functions, it is time to review the applications of SPR to RNA interaction studies, which have actually extended over a long time period. The primary advantage of SPR is its ability to measure affinities and kinetics in real time, along with being a label-free technique and utilizing relatively small quantities of materials. Recently, developments that use SPR to analyze the interactions of different RNA sequences with proteins and small molecules demonstrate the versatility of SPR as a powerful method in the analysis of the structure-function relationships, not only for biological macromolecules but also for potential drug candidates. This chapter will guide the reader through some background material followed by an extensive assay development to dissect the interactions of small molecules and RNA sequences using SPR as the critical method. The protocol includes (i) fundamental concepts of SPR, (ii) experimental design and execution, (iii) the immobilization of RNA using the streptavidin-biotin capturing method, and (iv) affinities and kinetics analyses of the interactions using specific example samples. The chapter also contains useful notes to address situations that might arise during the process. This assay demonstrates SPR as a valuable quantitative method used in the search for potential therapeutic agents that selectively target RNA.

Keywords: Biosensors biacore SPR; Drug discovery; Heterocyclic amidines; RNA- ligand interactions; Small molecules.

Fig. 1.

(A) Structure of some heterocyclic diamidine RNA binders and Neomycin; (B) 5′-biotin-labeled RNA example sequences used in this study; (C) Structure of the 5′-RNA biotin derivative nucleotide used during immobilization.

Fig. 2.

(A) The procedure for the immobilization of 5′-biotin labeled RNA on CM5 Chips; (B) Schematic representation of biomolecular interactions observed in a flow cell and SPR angle change with some of the critical components labeled; (C) SPR sensorgram and its components described in steps.

Fig. 3.

Example for immobilization of biotin-labeled RNA on a streptavidin chip. The triplet repeats RNA, r(CUG)_5, is immobilized in flow cell 3 (fc3).

Fig. 4.

Processing biosensor data. (A) Raw data from biosensor for DB182 injected over the r(CUG)₅ RNA sequence or surface; (B) Data sets from the reference surface were subtracted from the data from the reaction surface; (C) Axes calibration data obtained from Biaevaluation software by using kinetic/affinity commands; (D) Corrections of the response from the blank buffer injections; (E) Overlay of a series of DB182 injections.

Fig. 5.

DB75 interaction with r(CUG)₅ RNA sequence. (A) A representative set of sensorgrams with graded concentrations of DB75 are shown in the figure (red arrow) from 10 nM to 10 μM; (B) The corresponding RU at steady state is shown. Data were the best fit with an independent binding site size model (Eqs. (6) and (7)) with N = 8 ± 1, K_D = 3.4 ± 0.22 μM. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.

Fig. 6.

DB182 interaction with the r(GC)₄ RNA sequence. (A) A representative set of sensorgrams with graded concentrations of DB182 are shown in the figure (red arrow) from 10 nM to 10 μM; (B) The corresponding RU at steady state, as in Fig. 5, data were best fit with an independent binding site size model (Eqs. (6) and (7)) with N = 4 ± 1, K_D = 3.1 ± 0.22 μM. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 7.

DB1246 interaction with the TERRA G4-RNA sequence. (A) A representative set of sensorgrams with graded concentrations of DB1246 are shown in the figure (red arrow) from 10 nM to 250 nM; (B) The corresponding RU at steady state as in Fig. 5, data were best fit with two binding site model (Eq. (9)) with K_D1 = 3 ± 0.8 nM and K_D2 = 80 ± 12 nM. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 8.

Steady state result of Neomycin interaction with all 3 RNA sequences: with r(GC)₄ RNA (), with r(CUG)₅ RNA () and with TERRA G4-RNA (). Data were fitted with independent, identical binding site for r(GC)₄ DNA K_D = 16.2 ± 0.11 μm, N = 2 ± 1; r(CUG)₅ RNA K_D = 9.6 ± 1.2 μM, N = 3 ± 1; TERRA G4-DNA model K_D = 22 ± 7.2 μM, N = 1 ± 1.

Fig. 9.

DB182 interaction with the r(GC)₄ RNA sequence. (A) A steady state fit to obtain the K_D at 25 mM KCl. Steady state data were best fit with an independent binding site size model (Eqs. (6) and (7)) with K_D = 0.37 ± 0.09 μM, N = 6 ± 1; (B) Representative sensorgrams with graded concentrations of DB182:10, 50 and 100 nM (colored) and dissociation kinetic fitting curve (black), k_d = 0.03 ± 0.002 s⁻¹, k_a average (estimated) = (8.11 ± 2.04) × 10⁴ M⁻¹s⁻¹.