Orthogonality of FH and FJ ribozymes: potential for simultaneous dual-color RNA

The ribozymes FJ1 and FJC9 demonstrated no reactivity with the FH14 ribozyme substrate N⁶-Biotin-ATP. FH14 and FH31 ribozymes were also tested for reactivity with N⁶ -Biotin-TenDP. FH14 showed no activity with the TenDP analog whereas FH31 managed to ligate this substrate to the target sequence, albeit at low efficiency. Thus, FJ1 and FJC9 can be considered as perfectly orthogonal to the FH14 ribozyme but not to FH31. The orthogonality of the ribozymes allowed labeling of a synthetic RNA transcript at two different positions simultaneously, using FJ1 and FH14. Cy5-TenDP and FAM-ATP were used as labeling substrates for FJ1 and FH14, respectively. Dual-color labeling of target RNA at two different positions is significant, as it allows the installation of FRET pairs in strategic positions.

Subsequently, the conformational dynamics of labeled RNA can be studied using the FRET signal. The main problem with large RNA molecules, in this case, is the lack of suitable techniques for separation of the doubly labeled RNA from the unlabeled or singly labeled products.

We have also managed to label large cellular RNA such as E. coli 16S or 23S rRNA, using FJ1 and FH14 ribozymes, both individually and at the same time. The results of these reactions once again asserted the orthogonality of these ribozymes as well as their high degree of specificity. Our method involved labeling of the target sequence RNA in total cellular RNA. The treated cellular RNA was then resolved on a gel and visualized via fluorescent imaging. This strategy can be compared to blotting techniques such as northern blotting. Our strategy, however, requires fewer steps, as in northern blotting the RNA is typically transferred and fixed onto a nylon membrane. Synthetic probes that are labeled either radioactively or fluorescently are then annealed to the target RNA followed by visualization on the membrane after washing off the unbound probe. In our strategy, RNA is visualized directly in the gel, without the need for additional labeled probes. For this strategy to replace northern blotting however, it is important to investigate the sensitivity of our technique. The lowest range of RNA that can be detected using this strategy needs to be determined. The smallest amount of RNA detected in northern blotting using near infra-red fluorescent dyes was determined to be around 0.05 fmol (Köhn et al., 2010; Miller et al., 2018). These systems however use multiple dyes per probe. The sensitivity of our system may also be improved by targeting the same RNA at several positions and attaching multiple labels, as demonstrated for 5S rRNA that was labeled at three sites.

4.3.5 FJ ribozymes and the path to RNA labeling in situ

The tenofovir-diphosphate analog used for selection of FJ ribozymes solves the orthogonality issue of the ATP based substrates used by FH ribozymes. The background signal produced as a result of FJ ribozyme labeling in live cell, may be lower as cellular polymerases do not incorporate tenofovir.

We have also tested resistance to debranching of the RNA samples branched using FJ1 or FH14. As expected, the phosphonomonoester linkage introduced by FJ1 was more resistant towards debranching by highly active recombinant Dbr1 compared to their phosphodiester counterparts. Although less enzymatically labile, the phosphonomonoester branch was not fully resistant towards debranching. The debranching rate of this product at Mn²⁺

concentrations above 10 µM was similar to the phosphodiester linkage. The results obtained in our debranching experiments by no means reflect the complex cellular conditions.

Debranching assay in environments such as cellular extracts should still be performed to provide us with a more realistic view.

There may be a way to make phosphonomonoester linkages even more resistant to debranching. Carrocci et al discovered that using Sp diastereomer of GTP-αS, in 10DM24 catalyzed branching reaction results in the formation of the Rp-thiophosphodiester (Figure 4-4 A). This branch type showed exceptional resistance to debranching even in 4 mM Mn²⁺

concentration (Carrocci et al., 2017). Similarly, thiophosphonate analogs can be synthesized and tested for FJ ribozyme mediated ligation (Figure 4-4 B). Ultimately, the thiophosphonate-branched RNA substrate can be examined for debranching resistance.

There is a likelihood that FJ ribozymes may not accept the thiophophonate analog as substrate. In that case it may be necessary to evolve new variants that utilize such tenofovir analogs. Rescue using thiophilic metal ions such as Cd²⁺ may also be another option (Basu and Strobel, 1999), however such ions are highly toxic for cells.

Figure 4-4 Debranching resistant NTP^s analogs. (A) GTPαS analog used by Carocci et al (B) Potentially debranching resistant substrate for FJ1 labeling.

Apart from substrate orthogonality and stability of the introduced label, FJ ribozymes share one major limitation with FH ribozymes. The dependency on high Mg²⁺ concentration is still a major issue that needs to be resolved prior to any cellular experiments. Folding in live cell and sensitivity to nuclease degradation is also an issue facing artificially developed functional nucleic acids (Filonov et al., 2015). This problem may be resolved by placing the FJ ribozymes within the context of stably folded RNA scaffold (Filonov et al., 2015) as mentioned in 1.4.1.3, for the case of fluorogenic aptamers.

Another limitation of both FH and FJ ribozymes, like most other covalent labeling methods, is the lack of fluorogenicity. It is important to design a fluorogenic substrate that only becomes fluorescent after it is attached to RNA. Otherwise, the background emission from the unligated substrate will interfere with the analysis. Absence of fluorogenicity limits the application of these ribozymes to fixed cells where the excess of unbound substrate can be washed out of the cell (Muthmann et al., 2020). The problem may be solved by using internally quenched nucleotide analogs, in which a fluorophore is conjugated to the nucleobase and a quencher to the γ-phosphate (Hacker et al., 2015). These analogs, however, are not naturally produced in the cell. The cellular entry of these types of substrates will also be challenging.