Analyzing the interaction between RNA and RBPs (RNA-binding proteins) by NMR spectroscopy: the introduction of small molecules as third wheel

Background RNA-binding proteins (RBPs) are capable of regulating gene expression at the post-transcriptional level, controlling it by a complex network of RNA/protein interactions.1 HuR is an RBP belonging to Hu (or ELAV-L) family, and its target mRNAs encode proteins involved in tumorigenesis and metastasis. Experimental evidences correlate the overexpression of HuR to a variety of tumors 1, and for this reason it can be useful to develop new ligands that can inhibit it, acting as disrupters of the HuR-mRNA complex. Aims Previous studies2 led to the identification of small, drug-like molecules able to bind HuR by applying different approaches. Here, their ability to bind the protein in the RNA binding site (and so to interfere with HuR-RNA complex) is demonstrated, through competition experiments using Saturation Transfer Difference (STD) NMR analysis performed in presence of RNA and its PNA mimic (Peptide Nucleic Acid). Methods The intermolecular interactions between a protein and small molecules and their recognition mechanisms can be analyzed by NMR spectroscopy at the molecular or atomic level. 3 Among ligand-based techniques, Saturation Transfer Difference (STD) NMR spectroscopy is one of the most widespread NMR techniques for these interactions’ studies, allowing the identification of ligand epitope. Specifically, two scaffolds were selected and each synthesized compound was analyzed in solution in presence of HuR in a 400:1 ligand-protein ratio. The addition of RNA or PNA as displacing molecules allows to understand if they are able to compete for the same binding site. Results Starting with a structure-based procedure targeting the RNA binding site of HuR, new molecules able to bind it were designed and synthesized. (Figure 1) Figure 1. Image of the fundamental interactions that HuR protein establishes with mRNA. In silico studies combined with STD NMR, virtual screening, and fragment-based drug discovery allowed the selection of two lead compounds as promising binders for HuR. (Figure 2A) Moreover, through a protein-templated dynamic combinatorial chemistry (ptDCC) approach, a new library of molecules derived from compound 1 and 2 was developed in order to improve their affinity for the protein (Figure 2B). Figure 2. A) Chemical structure of the lead compounds 1 and 2. B) Chemical structure of RBA2, RBA3 and A-BOPC3. C) From above to below: 1H-NMR of RBA3 in PBS pH 7.4, STD spectrum of RBA3 in presence of HuR and STD spectrum of RBA3 in presence of HuR and PNA. The cartoon indicates that in a STD experiment only the protein is irradiated with a radiofrequency so that the ligand receives the magnetization only if it is in the binding site. Here the interaction studies, obtained by Saturation Transfer Difference (STD) NMR, between derivatives of compound 1 and 2 (Figure 2B) and HuR are presented. STD data allowed the validation of the ligands ability to bind HuR and the determination of their epitope map, illustrating which ligand’s moiety interacts with the protein. Since STD NMR allows to obtain information about the ligand and not about the protein binding site, to assess the capability of the ligands to interact in the same RNA binding pocket of the protein, competition experiments were performed adding a nucleic acid (c-Fos 3’ UTR mRNA (5’-AUUUUUAUUUU-3’)) to the ligand-HuR mixture. The results confirm that the selected RNA fragment can displace the ligands from the binding site. Since ribonucleases (RNase) enzymes are everywhere in the environment, RNase contamination is generally the leading cause of mRNA fast degradation in the experimental conditions. For this reason, to avoid the fast degradation of mRNA that can affect the NMR results, the use of Peptide Nucleic Acid (PNA) (H-ATTTTTATTT-NH2) as RNA mimic was exploited. The STD spectra (Figure 3C) showed that the addition of PNA to the ligand-protein complex reduces the intensities of the STD signal of the small molecule, indicating that the two are interacting with the same HuR site. Furthermore, when comparing RBA2 and RBA3, the quantitative analysis of STD data suggested that their affinities are different and that RBA3 is less displaced than RBA2 by the addition of PNA, suggesting its higher affinity for the protein. The approach proposed here demonstrated that PNA can be used instead as RNA mimic also in STD experiments. Conclusion These results suggested the ability of the selected molecules to bind the protein in the binding site for which they were designed acting as RNA-HuR complex disruptors and allowed the identification of the molecule with the higher affinity among the tested compounds. The proposed method is also a worth alternative for guiding medicinal chemistry efforts for identifying more potent molecules that can effectively modulate HuR-RNA complexes, as potential pharmacological agents and can be also expanded to other RNA binding proteins.