Post-transcriptional modifications are very important after all

In the beginning there was RNA and our lives as molecular biologists and biochemists proceeded simply. All one needed to know, to explain all things nucleic acid, were 4 canonical nucleotides: G, A, C and U. Then, magic happened! Pseudouridine, a 5th nucleotide resembling uridine but not quite chemically the same, was described in the early ‘50s.^1-3 This marked the birth of the nucleotide post-transcriptional modification field. In the years that followed, there were more examples of these oddities that were chemically distinct from the 4 canonical nucleotides and the excitement grew. Notably, the cloverleaf structure of tRNA presented by Robert Holley in the ‘60s revealed the presence of inosine, pseudouridine, dihydrouridine and several methylated nucleotides.⁴ Specifically, the discovery of inosine at the first position of the anticodon of tRNA^Ala was a crucial lead-in for the formulation of the wobble hypothesis.⁵ Despite these early reports, and due in part to the intractability of most modifications, many scientists did not know what to make of these chemical eccentricities. Save for a few aficionados, for many years nucleotide modifications where brushed under the proverbial cellular rug. This was not totally unwarranted, given that many biochemical assays (for example amino acylation) did not seem grossly affected by the lack of modifications in a given substrate. Naturally, people's attention to modifications, which was originally piqued by the potential new chemistries, slowly weaned. Lack of interest in modifications was further compounded by 2 important hurdles in the field: 1) analyzing and characterizing new modifications was, and still is, technically challenging; 2) genetic approaches led to the conclusion that modifications were not all that important. It turns out that in organisms amenable to genetic manipulation, such as S. cerevisiae or E. coli, deletion of genes known to encode modification enzymes, had little phenotypic effects on the well being of cells. Of course the mutants analyzed depended on what modifications were identified and in what substrate. One should mention that although many modifications had been described by thin-layer chromatography (TLC) methods, generally it was still not clear as to what specific nucleotide position in RNA they belonged; thus no substrate, no enzyme. It was perhaps no coincidence that the few mutants analyzed all encoded enzymes whose products were easily detectable and unfortunately these also happened to be those causing only subtle phenotypes. Again the prevailing thought that modifications were of no importance remained.

Despite the challenges encountered in the early days, there was a lingering question: if modified nucleotides were not important, why did nature maintain so many in a given RNA at evolutionary conserved positions? For example, pseudouridine 55, various modified purines at position 37 and modified uridines at the first position of the anticodon in tRNAs. The degree of conservation is in fact sometimes so astonishing that in many cases the same (or at least chemically very similar) modifications are found at a given position in the same tRNA in organisms spanning the 3 domains of life, Eukarya, Archaea and Bacteria.⁶ Still, answering the nucleotide modification evolutionary questions, let alone questions of their significance, remained challenging. Yet science does not stand still and little by little a cadre of geneticists, chemists and biochemists chipped away at the difficulties and many technical hurdles were overcome. The major leap in our thinking especially came with the advent of genomics, which coupled with major developments in the detection of modifications by mass spectrometry and the establishment of modification-specific chemistries have led to what is now a blossoming field.

To date, there are over 100 different modifications described and the list will certainly not end there. Most importantly, modifications are popping up everywhere.^6,7 It is now safe to say that every nucleic acid in a cell undergoes some form of post-transcriptional modification. At the nucleotide level, every position in the purine and pyrimidine bases has now been seen naturally modified; often the sugar itself is also modified. Although many modifications play roles in modulating nucleic acid structure, it is becoming increasingly clear that many, if not all, modifications affect cells in sometimes unpredictable ways. In fact, there is strength in numbers and it is now recognized that organisms dedicate a significant portion of their genome to encode modification enzymes. This is best documented in the yeast S. cerevisiae where approximately 1% of the genome encodes proteins dedicated to the biosynthesis of modifications in various nucleic acids.⁶ This exceeds what is dedicated to tRNA genes themselves or even the portion of the genome dedicated to canonical nucleotide biosynthesis.

As to their importance, there are several growing themes. First, many modifications do not act alone and they are part of a larger cascade, which combinatorially may change the information content of nucleic acid polymers beyond what is intended from the genome.⁸ Second, some modifications are so chemically intricate that their synthesis draws important cofactors from many products and byproducts of central metabolism, raising the question of higher order relationships between these pathways in cells.⁸ This has now led to explorations of the role of modifications as an integral part of sensing mechanisms, signal transduction cascades, if you will, to connect environmental cues to a cells metabolic state.^9,10 These have been recently documented in terms of a cell's response to environmental stress, but it can be argued that the ever-changing modification landscape may be part of the cells normal program of growth. Moreover, the function of modifications, in this sense, will be deeply rooted in how a cell deals with nutrient availability.

This special issue of RNA Biology on nucleotide post-transcriptional modifications gathers current knowledge and covers the state of this growing field with the hopes that the importance of post-transcriptional modifications, which many of us have appreciated for many years, now becomes commonplace in the daily thinking about cells and organisms. The articles include new discoveries that at various levels touch on the growing themes discussed above. Some articles highlight the complex chemistry involved in the biosynthesis of most modifications, while stressing the critical mechanistic aspects of enzymology and substrate recognition by modification enzymes. Other articles expand on modern approaches to detect and map modifications, where the challenges have rested on our ability to describe modifications in a position-specific manner in a particular RNA sequence. Here, one must appreciate the importance of such mapping experiments, given that the same modification may impart different effects on a particular RNA depending on the sequence context. These effects may then range from the subtle to the extreme. For example, still one of the most difficult questions is the function of modifications in organelles. In mitochondria a growing number of diseases are correlated with point mutations in mitochondrial tRNAs ¹¹; many of these affect a site that is normally modified and a few connections have now been made between mitochondrial modification defects and disease states.¹² However, most of these questions are only tractable genetically, and still pinning the defect to the lack of a specific modification in a specific RNA at a particular position is challenging. Referring to organelles the biggest difficulty is to obtain sufficient sample amounts to place the modifications within the limit of detection by mass spectrometry, the golden standard in modification identification and assignment.

Fortunately, we are entering an exciting time and the modification field has benefited greatly from the latest technological developments described. Still plenty of questions and challenges remain. In the coming years, it will be most interesting to see how modifications are integrated into various cellular systems and more specifically in eukaryotes how intracellular compartmentalization affects modification sets. What makes all this especially difficult is the fact that it now appears that RNAs are not fully modified at all times and at all sites. Rather, there is a modification continuum whereby the level of modifications may be constantly changing as RNA's way to match coding capacity to substrate availability. Thus given the varied effects on RNA and implicitly on gene expression, modifications are truly epigenetics gone wild; always hard to predict what one would get. But given the growing list of malfunctions associated with defects in post-transcriptional modifications, the field is ripe for the picking and modern approaches shall prove instrumental in peeling all the layers of the complex business that is the modification field. What should no longer be questioned is the importance of modifications. We should all reconcile with the fact that nucleotide modifications are part and parcel of cellular metabolism and homeostasis at more levels than meet the eye. The following set of articles should further cement this dictum while serving as a good forecast for greater things to come.