Investigating the Merits of Microfluidic Capillary Zone Electrophoresis-Mass Spectrometry (CZE-MS) in the Bottom-Up Characterization of Larger RNAs.

Abstract

Established bottom-up approaches for the characterization of nucleic acids (NAs) rely on the strand-cleavage activity of nucleotide-specific endonucleases to generate smaller oligonucleotides amenable to gas-phase sequencing. The complexity of these hydrolytic mixtures calls for the utilization of a front-end separation to facilitate full mass spectrometric (MS) characterization. This report explored the merits of microfluidic capillary zone electrophoresis (CZE) as a possible alternative to common liquid chromatography techniques. An oligonucleotide ladder was initially employed to investigate the roles of fundamental analyte features and experimental parameters in determining the outcome of CZE-MS analyses. The results demonstrated the ability to fully resolve the various rungs into discrete electrophoretic peaks with full-width half-height (FWHH) resolution that was visibly affected by the overall amount of material injected into the system. Analogous results were obtained from a digestion mixture prepared by treating yeast tRNAPhe (75 nt) with RNase T1, which provided several well-resolved peaks in spite of the increasing sample heterogeneity. The regular shapes of such peaks, however, belied the fact that most of them contained sets of comigrating species, as shown by the corresponding MS spectra. Even though it was not possible to segregate each species into an individual electrophoretic peak, the analysis still proved capable of unambiguously identifying a total of 29 hydrolytic products, which were sufficient to cover 96% of the tRNAPhe's sequence. Their masses accurately reflected the presence of modified nucleotides characteristic of this type of substrate. The analysis of a digestion mixture obtained from the 364 nt HIV-1 5'-UTR proved to be more challenging. The electropherogram displayed fewer well-resolved peaks and significantly greater incidence of product comigration. In this case, fractionating the highly heterogeneous mixture into discrete bands helped reduce signal suppression and detection bias. As a result, the corresponding MS data enabled the assignment of 248 products out of the possible 513 predicted from the 5'-UTR sequence, which afforded 100% sequence coverage. This figure represented a significant improvement over the 36 total products identified earlier under suboptimal conditions, which afforded only 57% coverage, or the 83 observed by direct infusion nanospray-MS (72%). These results provided a measure of the excellent potential of the technique to support the bottom-up characterization of progressively larger NA samples, such as putative NA therapeutics and mRNA vaccines.