A novel aurone RNA CAG binder inhibits the huntingtin RNA-protein interaction

Abstract

Huntington's disease (HD) is a devastating, incurable condition whose pathophysiological mechanism relies on mutant RNA CAG repeat expansions. Aberrant recruitment of RNA-binding proteins by mutant CAG hairpins contributes to the progress of neurodegeneration. In this work, we identified a novel binder based on an aurone scaffold that reduces the level of binding of HTT mRNA to the MID1 protein in vitro. The obtained results introduce aurones as a novel platform for the design of functional ligands for disease-related RNA sequences.

Chart 1. Chemical structures of aurone derivatives 1a–1w and 2a–2c and aza-aurone (hemiindigo) 3 used for the screening and the structure of the known CAG RNA binder furamidine (4).

Fig. 1. Changes in the fluorescence intensity (in % of the initial intensity) of aurones 1a–1w and 2a–2c, aza-aurone 3 and furamidine (4) (c_lig = 5 μM) in the presence of 1 equiv. of the 5′-GCAGCAGCUUCGGCAGCAGC-3′ oligonucleotide in a buffer (pH = 7); positive values indicate fluorescence light-up in the presence of RNA and negative values indicate fluorescence quenching in the presence of RNA. The fluorescence output of each ligand was measured at three different excitation wavelengths: λ_ex = 350, 400 and 450 nm. The most reliable fluorescence output was obtained upon excitation at λ_ex = 350 nm for most compounds except for 1d (λ_ex = 450 nm) and 2b (λ_ex = 400 nm). To ensure the reproducibility, each measurement was repeated at least three times, and the repeat experiments gave values within 20%.

Fig. 2. Structures of the CUG RNA, HIV-1 TAR RNA and HIV-1 RRE-IIB RNA oligonucleotides used in this study and changes in the fluorescence intensity (in% of the initial intensity) of aurones (A) 2a and (B) 1d (c_lig = 5 μM) in the presence of 1 equiv. of each RNA oligonucleotide in a buffer (pH = 7); positive values indicate fluorescence light-up in the presence of RNA and negative values indicate fluorescence quenching in the presence of RNA. The fluorescence output was measured upon excitation at λ_ex = 350 nm for 2a and at λ_ex = 450 nm for 1d. To ensure the reproducibility, each measurement was repeated at least three times, and the repeat experiments gave values within 20%.

Fig. 3. (A) Spectrophotometric titration of 2a with the 5′-GCAGCAGCUUCGGCAGCAGC-3′ RNA oligonucleotide (c₂ = 5 μM and RNA/c₂ = 0–4) and (B) binding isotherm, i.e. a plot of the absorbance of 2aversus concentration of RNA (c_RNA), obtained from the photometric titration; black solid line: experimental data and orange dashed line: fit to the theoretical model.

Fig. 4. RNA–protein pull-down of MID1 with its target RNA HTT exon 1 in the absence (w/o compound 2a) or the presence of compound 2a. RNA-bound proteins were analyzed by western blot detecting MID1. (A) RNA–protein pull-down in the presence or the absence of compound 2a at a final concentration of 100 μM. A negative control that does not contain RNA was included (no RNA). The expected band of approx. 70 kDa was detected in the RNA pull-down without the compound in the cell lysate. (B) RNA–protein pull-down as described in (A) with different doses of compound 2a (final concentrations of 1 μM, 10 μM, and 100 μM).