The 2D-structure of the T. brucei pre-edited RPS12 mRNA is not affected by macromolecular crowding

Mitochondrial transcript maturation in African trypanosomes requires RNA editing to convert nucleotide-deficient pre-mRNAs into translatable mRNAs. The different pre-mRNAs have been shown to adopt highly stable 2D-folds, however, it is not known whether these structures resemble the in vivo folds given the extreme “crowding” conditions within the mitochondrion. Here we analyze the effects of macromolecular crowding on the structure of the mitochondrial RPS12 pre-mRNA. We use polyethylene glycol as a macromolecular cosolute and monitor the structure of the RNA globally and with nucleotide resolution. We demonstrate that crowding has no impact on the 2D-fold and we conclude that the MFE-structure in dilute solvent conditions represents a good proxy for the folding of the pre-mRNA in its mitochondrial solvent context.


Introduction
The folding of RNA molecules into compact, native structures or ensembles of structures is dictated by a set of first principle physicochemical forces. 1 One of which is chargecompensation to overcome the electrostatic repulsion of the negatively charged phosphodiester backbone. [2][3][4] Mono and divalent metal-ions at low to high millimolar concentrations contribute to this task 5 and the effects of metal ion-radius and charge density have been studied in detail. [6][7][8] Next to metal-ions, metabolites, polyamines and osmolytes have been shown to modulate RNA structure 9-13 as well as high concentrations of macromolecules, which can occupy up to 30% of the total volume of a cellular compartment.
This generates so-called "crowded" solvent conditions, [14][15][16] which in general stabilize RNA 2D-and 3D-structures through the excluded volume effect and entropy perturbation of the folding landscape.
5, 17,18 RNA editing in African trypanosomes describes a post-transcriptional modification reaction of mitochondrial pre-mRNAs that is characterized by the site-specific insertion and deletion of exclusively U-nucleotides (for a review see ref. 19). The reaction takes place within the single mitochondrion of trypanosomes, which represents the most crowded intracellular environment of eukaryotic cells. Intra-mitochondrial macromolecular concentrations reach up to 560g/L. 20,21 Editing is catalyzed by a macromolecular protein complex, the 20S editosome, 19 which interacts with 18 mitochondrial pre-mRNAs as substrates in the processing reaction. The different transcripts encode subunits of the mitochondrial electron transport and oxidative phosphorylation chains and have been characterized by several unusual features: First, the majority of pre-mRNAs lacks substantial sequence information (on average 45%), hence they require RNA editing to be converted into translatable mRNAs. Second, the different pre-mRNAs are typified by an extraordinarily high G-content (34%), which in two thirds of the cases are clustered in tracts of G-nucleotides (2≤G≤8). Third, in vitro chemical probing studies revealed that the different pre-mRNAs adopt extraordinarily stable 2D-structures approaching the stability of structural RNAs. 22,23 Next to canonical base-pairing they contain pseudoknots and in many cases multiple G-quadruplex (GQ)-folds. However, it is not clear whether these 2D-structures resemble the in vivo folds given the extreme crowding conditions in the trypanosome mitochondrion. This is especially important since RNA editing in vitro has been shown to be sensitive to crowded solvent properties. 24 Here we analyze the effect(s) of macromolecular crowding on the structure of the mitochondrial RPS12 pre-mRNA as an archetypical example of a mitochondrial transcript in African trypanosomes. We use high molecular mass polyethylene glycol (PEG) as a neutral macromolecular cosolute to mimic intra-mitochondrial solvent conditions and we monitor the structure of the RPS12 transcript by temperature-dependent UV-spectroscopy and by selective 2'-hydroxyl acylation analyzed by primer extension (SHAPE).

Results and Discussion
To study the impact of macromolecular crowding on the structure of mitochondrial pre-mRNAs in Trypanosoma brucei we used the primary transcript of RPS12 as a representative model RNA. The pre-mRNA molecule is 325nt long. As a pan-edited transcript it is edited throughout its entire primary sequence with 132 U-nt inserted and 28 U's deleted. The RNA has a G-nt-content of 27% and a purine/pyrimidine (R/Y)-ratio of 1.3. Its experimentally determined minimal free energy (MFE)-2D-structure calculates to a Gibbs free energy (DG) of −152kcal/mol with a base-paired versus single-stranded nucleotide ratio (r bp/ss ) of 0.62. In addition, the RNA contains a pseudoknot. 22 Since in cell structure probing experiments have successfully been performed only with abundant, cytosolic RNAs (see ref. 5,25-28) we decided to mimic the crowded, intra-mitochondrial solvent environment by using a chemically inert, synthetic cosolute such as polyethylene glycol (PEG). 29, 30 We used PEG with a mean molecular mass of 4000g/mol (PEG 4000 ). The synthetic compound is characterized by a polymer crossover concentration (f*) of 4% (w/w), which marks the transition from a semidilute to a crowded solvent regime. [31][32][33] As a consequence, all experiments were performed at 6% (w/w) PEG 4000 .
As a first comparison, we measured UV-melting profiles of the RPS12 pre-mRNA in dilute and crowded solvent conditions. Representative normalized melting profiles and their    1A and Table 1A). This demonstrates that the crowding-driven stabilization of RPS12 RNA is by far weaker than the stabilization by divalent cations and that the impact on the overall structure of the transcript is minute.
As a follow up of these experiments we analyzed the effects of macromolecular crowding by probing the structure of the RPS12 RNA with nucleotide resolution. For that we used selective 2'-hydroxyl acylation analyzed by primer extension (SHAPE). 34   This behavior translates to the 2D-structure level when the normalized SHAPE-reactivities are used as pseudo free energy constraints to guide the structure prediction. Fig. 3 shows the pseudoknotted 2D-structure of the RPS12 transcript in dilute and crowded solvent conditions. Molecular crowding has no effect on the transcript if Mg 2+ -ions are present. However, since the average and median nucleotide flexibility are slightly decreased at crowded solvent conditions this results in a decrease of the Gibbs free energy (DG) of −3.9kcal/mol (Tab. 1B/C). By contrast, in the absence of Mg 2+ -ions the two structures are characterized by a roughly 15% higher DG and the absence of the pseudoknot. Sixty percent of the nucleotides retain their structural context at all conditions studied (Fig. S2).    were performed using CircleCompare. 45 MFE-structures were compared in terms of their "sensitivity" (sens) and their positive predictive values (ppv): Sens=fraction of bp in the reference structure also present in the non-reference structure; ppv= fraction of bp in the nonreference structure also occurring in the reference structure. RPS12 pre-mRNA probed in dilute buffer at physiological i.e. 10mM Mg 2+ -ion concentrations served as a reference state if not indicated otherwise.

Disclosure of Potential Conflicts of Interests
No potential conflicts of interest were disclosed.