Cross-linking methods to identify the mRNA targets of microRNAs

Abstract

microRNAs (miRNAs, miRs) constitute a class of small, endogenous, non-coding RNAs (ncRNAs) and have a great influence on various processes within the cell. By base‐pairing selectively to partially‐complementary sites located predominantly in the 3′-untranslated region (3′‐UTR) of target messenger RNAs (mRNAs), miRNAs participate in the post-transcriptional regulation of gene expression. miRNAs are responsible for controlling the expression of the majority of human protein‐coding genes and their dysregulation has been related to many pathological processes and diseases, including cancer. Thus, an in-depth understanding of the miRNA mechanisms of action is of great importance. One of the key challenges is the elucidation of the exact sites of the canonical and non-canonical interactions in a cell-specific context. An imperfect pairing between miRNAs and their target RNAs in animals, as well as high false-positive and false-negative rates for current prediction algorithms, generate a need for the experimental validation of the predicted binding sites. For this purpose, various cross-linking-based methods for target identification have been established. The microRNA cross‐linking and immunoprecipitation (miR‐CLIP) approach developed previously in our group allows capturing predicted and unpredicted miRNA targets in cells. The miR-CLIP technique employs the pre-miR-CLIP probes site-specifically modified with trioxsalen and biotin moieties introduced pre- or post-synthetically at the 2′-O-position of the nucleoside by the copper(I)-catalysed azide-alkyne cycloaddition (CuAAC/’CLICK’) reaction. It was shown that a miR‐106a-5p pre-miR-CLIP probe was able to cross‐link to the complementary regions present in the RNA targets. However, further tests suggested that the cross-linking with this probe design is more sequence-dependent than initially expected and therefore the outcome of the reaction is more difficult to predict. The overall objective of this research was to design miR-CLIP probes for all miRNAs of a cell, for which an understanding of the cross-linking chemistry with respect to sequence preferences, linker composition and sites of conjugation was required. The first project described in this thesis focused on exploring the nature of the cross-linking with the original miR‐106a-5p miR-CLIP probe. In vitro cross-linking assays performed with a set of trioxsalen-labelled miR-106a-5p analogues and probes for other miRNAs allowed us to establish a working hypothesis about the sequence preferences of cross-linking. The results suggested that inter-strand cross-linking most likely took place with a uridine positioned 2-3 base pairs upstream from the desired (juxtaposed) cross-linking site. This allowed proposing an alternative strategy for the positioning of the trioxsalen modification within these probes. In the second project, diazirine-based probes were synthesized and tested. The results showed that the attachment of the cross-linker at the 2′-O-position of the ribose or C5-position of the pyrimidine was ineffective and resulted in quenching of reactive species by the solvent molecules. Finally, a different approach aiming to achieve a versatile (sequence-independent) cross-linking was developed and tested. The new strategy was based on the utilization of probes in which the trioxsalen molecule was introduced at the N4-position of cytidine through short poly(ethylene glycol) (PEG)-based linkers. Probes for miR-124-3p were functionalized post-synthetically by the application of the convertible nucleoside approach and showed more efficient cross-linking than analogous probes prepared with the use of the CLICK chemistry.