Synthesis of point-modified mRNA.

Synthetic mRNA has recently moved into the focus of therapeutic and vaccination efforts. Incorporation of modified nucleotides during in vitro transcription can improve translation and attenuate immunogenicity, but is limited to triphosphate nucleotides which are accepted by RNA polymerases, and their incorporation is either random or complete. In contrast, site-specific modification, herein termed 'point modification' in analogy to point mutations, holds significant technical challenge. We developed fundamental techniques for isolation of long, translatable and internally point-modified mRNAs. Enabling concepts include three-way-one-pot splint ligations, and isolation of mRNA by real-time elution from agarose gels. The use of blue light permitted visualization of mRNA in pre-stained gels without the photochemical damage associated with the use of hard UV-radiation. This allowed visualization of the mRNA through its migration in the agarose gel, which in turn, was a prerequisite for its recovery by electroelution into precast troughs. Co-eluting agarose particles were quantified and found to not be detrimental to mRNA translation in vitro. Translation of EGFP-coding mRNA into functional protein was quantified by incorporation of 35S-labelled methionine and by in-gel EGFP fluorescence. This enabled the functional analysis of point modifications, specifically of ribose methylations in the middle of a 1371 nt long mRNA.

RNA marker modifications reveal the necessity for rigorous preparation protocols to avoid artifacts in epitranscriptomic analysis.

The accurate definition of an epitranscriptome is endangered by artefacts resulting from RNA degradation after cell death, a ubiquitous yet little investigated process. By tracing RNA marker modifications through tissue preparation protocols, we identified a major blind spot from daily lab routine, that has massive impact on modification analysis in small RNAs. In particular, m6,6A and Am as co-varying rRNA marker modifications, appeared in small RNA fractions following rRNA degradation in vitro and in cellulo. Analysing mouse tissue at different time points post mortem, we tracked the progress of intracellular RNA degradation after cell death, and found it reflected in RNA modification patterns. Differences were dramatic between liver, where RNA degradation commenced immediately after death, and brain, yielding essentially undamaged RNA. RNA integrity correlated with low amounts of co-varying rRNA markers. Thus validated RNA preparations featured differentially modified tRNA populations whose information content allowed a distinction even among the related brain tissues cortex, cerebellum and hippocampus. Inversely, advanced cell death correlated with high rRNA marker content, and correspondingly little with the naïve state of living tissue. Therefore, unless RNA and tissue preparations are executed with utmost care, interpretation of modification patterns in tRNA and small RNA are prone to artefacts.

Determination of enrichment factors for modified RNA in MeRIP experiments.

In the growing field of RNA modification, precipitation techniques using antibodies play an important role. However, little is known about their specificities and protocols are missing to assess their effectiveness. Here we present a method to assess enrichment factors after MeRIP-type pulldown experiments, here exemplified with a commercial antibody against N6-methyladenosine (m6A). Testing different pulldown and elution conditions, we measure enrichment factors of 4-5 using m6A-containing mRNAs against an unmodified control of identical sequence. Both types of mRNA carry 32P labels at different nucleotides, allowing their relative quantification in a mixture after digestion to nucleotides, separation by TLC and quantitative phosphorimaging of the labels.

AlkAniline-Seq: Profiling of m7 G and m3 C RNA Modifications at Single Nucleotide Resolution.

RNA modifications play essential roles in gene expression regulation. Only seven out of >150 known RNA modifications are detectable transcriptome-wide by deep sequencing. Here we describe a new principle of RNAseq library preparation, which relies on a chemistry based positive enrichment of reads in the resulting libraries, and therefore leads to unprecedented signal-to-noise ratios. The proposed approach eschews conventional RNA sequencing chemistry and rather exploits the generation of abasic sites and subsequent aniline cleavage. The newly generated 5'-phosphates are used as unique entry for ligation of an adapter in library preparation. This positive selection, embodied in the AlkAniline-Seq, enables a deep sequencing-based technology for the simultaneous detection of 7-methylguanosine (m7 G) and 3-methylcytidine (m3 C) in RNA at single nucleotide resolution. As a proof-of-concept, we used AlkAniline-Seq to comprehensively validate known m7 G and m3 C sites in bacterial, yeast, and human cytoplasmic and mitochondrial tRNAs and rRNAs, as well as for identifying previously unmapped positions.