SiRNAs can cause unintended gene silencing due to miRNA-like effects because of the similarity in function of an siRNA guide strand and a miRNA. Here we evaluate the effect on miRNA-like off targeting of introducing the adenosine derivative 7-EAA and triazoles prepared from 7-EAA at different positions in an siRNA guide strand. We find that a sterically demanding triazole placed in the RNA duplex major groove at position six of the guide strand dramatically reduces miRNA-like off targeting potency. A high-resolution structure of an RNA duplex bearing a novel, major-groove localized triazole is reported, which suggests that modified triazoles could be disrupting the hAgo2-guide-target RNA ternary complex. Five different triazole modifications were tested at the guide strand 6-position for effects on on-target and miRNA-like off target knockdown potency. A 7-EAA triazole bearing a benzylamine substituent displayed on-target knockdown activity as potent as the native siRNA, while having an IC50 against a miRNA-like off target >100-fold higher. Melting temperature studies revealed no obvious correlation between potency in knockdown assays and a modification's effect on duplex stability. These results, along with known structures of hAgo2-guide-target ternary complexes, are used to rationalize the effect of 7-EAA triazoles on miRNA-like off target effects.

Short interfering RNAs (siRNAs) are promising therapeutics that make use of the RNA interference (RNAi) pathway, but liabilities arising from the native RNA structure necessitate chemical modification for drug development. Advances in the structural characterization of components of the human RNAi pathway have enabled structure-guided optimization of siRNA properties. Here we report the 2.3 Å resolution crystal structure of human Argonaute 2 (hAgo2), a key nuclease in the RNAi pathway, bound to an siRNA guide strand bearing an unnatural triazolyl nucleotide at position 1 (g1). Unlike natural nucleotides, this analogue inserts deeply into hAgo2’s central RNA binding cleft and thus is able to modulate pairing between guide and target RNAs. The affinity of the hAgo2–siRNA complex for a seed-only matched target was significantly reduced by the triazolyl modification, while the affinity for a fully matched target was unchanged. In addition, siRNA potency for off-target repression was reduced (4-fold increase in IC50) by the modification, while on-target knockdown was improved (2-fold reduction in IC50). Controlling siRNA on-target versus microRNA (miRNA)-like off-target potency by projection of substituent groups into the hAgo2 central cleft from g1 is a new approach to enhance siRNA selectivity with a strong structural rationale.

Alkyne-modified nucleoside analogs are useful for nucleic acid localization as well as functional and structural studies because of their ability to participate in copper-catalyzed azide/alkyne cycloaddition (CuAAC) reactions. Here we describe the synthesis of the triphosphate of 7-ethynyl-8-aza-7-deazaadenosine (7-EAATP) and the enzymatic incorporation of 7-EAA into RNA. The free nucleoside of 7-EAA is taken up by HeLa cells and incorporated into cellular RNA by endogenous RNA polymerases. In addition, 7-EAATP is a substrate for both T7 RNA polymerase and poly (A) polymerase from Escherichia coli in vitro, albeit at lower efficiencies than with ATP. This work adds to the toolbox of nucleoside analogs useful for RNA labeling.

Adenosine deaminases acting on RNA (ADARs) are RNA-editing enzymes responsible for the conversion of adenosine to inosine at specific locations in cellular RNAs. ADAR1 and ADAR2 are two members of the family that have been shown to be catalytically active. Earlier, we reported a phenotypic screen for the study of human ADAR2 using budding yeast S. cerevisiae as the host system. While this screen has been successfully applied to the study of ADAR2, it failed with ADAR1. Here, we report a new reporter that uses a novel editing substrate and is suitable for the study of ADAR1. We screened plasmid libraries with randomized codons for two important residues in human ADAR1 (G1007 and E1008). The screening results combined with in vitro deamination assays led to the identification of mutants that are more active than the wild type protein. Furthermore, a screen of the ADAR1 E1008X library with a reporter construct bearing an A•G mismatch at the editing site suggests one role for the residue at position 1008 is to sense the identity of the base pairing partner for the editing site adenosine. This work has provided a starting point for future in vitro evolution studies of ADAR1 and led to new insight into ADAR’s editing site selectivity.

Ribonucleoside analogues bearing terminal alkynes, including 7-ethynyl-8-aza-7-deazaadenosine (7-EAA), are useful for RNA modification applications. However, although alkyne- and triazole-bearing ribonucleosides are in widespread use, very little information is available on the impact of these modifications on RNA structure. By solving crystal structures for RNA duplexes containing these analogues, we show that, like adenosine, 7-EAA and a triazole derived from 7-EAA base pair with uridine and are well-accommodated within an A-form helix. We show that copper-catalyzed azide/alkyne cycloaddition (CuAAC) reactions with 7-EAA are sensitive to the RNA secondary structure context, with single-stranded sites reacting faster than duplex sites. 7-EAA and its triazole products are recognized in RNA template strands as adenosine by avian myoblastosis virus reverse transcriptase. In addition, 7-EAA in RNA is a substrate for an active site mutant of the RNA editing adenosine deaminase, ADAR2. These studies extend our understanding of the impact of these novel nucleobase analogues and set the stage for their use in probing RNA structure and metabolism.

Short interfering RNAs (siRNAs) are promising drug candidates for a wide range of targets including those previously considered “undruggable”. However, properties associated with the native RNA structure limit drug development, and chemical modifications are necessary. Here we describe the structure-guided discovery of functional modifications for the guide strand 5′-end using computational screening with the high-resolution structure of human Ago2, the key nuclease on the RNA interference pathway. Our results indicate the guide strand 5′-end nucleotide need not engage in Watson–Crick (W/C) H-bonding but must fit the general shape of the 5′-end binding site in MID/PIWI domains of hAgo2 for efficient knockdown. 1,2,3-Triazol-4-yl bases formed from the CuAAC reaction of azides and 1-ethynylribose, which is readily incorporated into RNA via the phosphoramidite, perform well at the guide strand 5′-end. In contrast, purine derivatives with modified Hoogsteen faces or N2 substituents are poor choices for 5′-end modifications. Finally, we identified a 1,2,3-triazol-4-yl base incapable of W/C H-bonding that performs well at guide strand position 12, where base pairing to target was expected to be important. This work expands the repertoire of functional nucleotide analogues for siRNAs.

ADARs (adenosine deaminases acting on RNA) are RNA editing enzymes that bind double helical RNAs and deaminate select adenosines (A). The product inosine (I) is read during translation as guanosine (G), so such changes can alter codon meaning. ADAR-catalyzed A to I changes occur in coding sequences for several proteins of importance to the nervous system. However, these sites constitute only a very small fraction of known A to I sites in the human transcriptome, and the significance of editing at the vast majority sites is unknown at this time. Site-selective inhibitors of RNA editing are needed to advance our understanding of the function of editing at specific sites. Here we show that 2′-O-methyl/locked nucleic acid (LNA) mixmer antisense oligonucleotides are potent and selective inhibitors of RNA editing on two different target RNAs. These reagents are capable of binding with high affinity to RNA editing substrates and remodeling the secondary structure by a strand-invasion mechanism. The potency observed here for 2′-O-methyl/LNA mixmers suggests this backbone structure is superior to the morpholino backbone structure for inhibition of RNA editing. Finally, we demonstrate antisense inhibition of editing of the mRNA for the DNA repair glycosylase NEIL1 in cultured human cells, providing a new approach to exploring the link between RNA editing and the cellular response to oxidative DNA damage.

ADAR2 is a member of a family of RNA editing enzymes found in metazoa that bind double helical RNAs and deaminate select adenosines. We find that when human ADAR2 is overexpressed in the budding yeast Saccharomyces cerevisiae it substantially reduces the rate of cell growth. This effect is dependent on the deaminase activity of the enzyme, suggesting yeast transcripts are edited by ADAR2. Characterization of this novel set of RNA substrates provided a unique opportunity to gain insight into ADAR2’s site selectivity. We used RNA-Seq. to identify transcripts present in S. cerevisiae subject to ADAR2-catalyzed editing. From this analysis, we identified 17 adenosines present in yeast RNAs that satisfied our criteria for candidate editing sites. Substrates identified include both coding and noncoding RNAs. Subsequent Sanger sequencing of RT-PCR products from yeast total RNA confirmed efficient editing at a subset of the candidate sites including BDF2 mRNA, RL28 intron RNA, HAC1 3′UTR RNA, 25S rRNA, U1 snRNA, and U2 snRNA. Two adenosines within the U1 snRNA sequence not identified as substrates during the original RNA-Seq. screen were shown to be deaminated by ADAR2 during the follow-up analysis. In addition, examination of the RNA sequence surrounding each edited adenosine in this novel group of ADAR2 sites revealed a previously unrecognized sequence preference. Remarkably, rapid deamination at one of these sites (BDF2 mRNA) does not require ADAR2’s dsRNA-binding domains (dsRBDs). Human glioma-associated oncogene 1 (GLI1) mRNA is a known ADAR2 substrate with similar flanking sequence and secondary structure to the yeast BDF2 site discovered here. As observed with the BDF2 site, rapid deamination at the GLI1 site does not require ADAR2’s dsRBDs.

Immune stimulation is a significant hurdle in the development of effective and safe RNA interference therapeutics. Here, we address this problem in the context of a mimic of microRNA-122 by employing novel nucleobase and known 2′-ribose modifications. The nucleobase modifications are analogues of adenosine and guanosine that contain cyclopentyl and propyl minor-groove projections. Via a site-by-site chemical modification analysis, we identify several immunostimulatory ‘hot spots’ within the miRNA guide strand at which single base modifications significantly reduce immune stimulation. A duplex containing one base modification on each strand proved to be most effective in preventing immune stimulation.

Considerable attention has focused on the use of alternatives to the native ribose and phosphate backbone of small interfering RNAs for therapeutic applications of the RNA interference pathway. In this synopsis, we highlight the less common chemical modifications, namely, those of the RNA nucleobases. Base modifications have the potential to lend insight into the mechanism of gene silencing and to lead to novel methods to overcome off-target effects that arise due to deleterious protein binding or mis-targeting of mRNA.