RNA is considered by many to be the primordial molecule of life. The major component of the ribosome—the most ancient organelle, which is shared between all living things—is ribosomal (r)RNA; which also forms the catalytic center necessary for protein production. The instructions needed for producing proteins is encoded within messenger (m)RNA, which is decoded by the ribosome by way of transfer (t)RNA. Beyond the classical RNAs involved in protein synthesis, a wide array of noncoding (nc)RNAs exist that mediate important biological processes: e.g. regulation of gene expression, mRNA splicing, post-transcriptional modification, chromatin structure, and more. In the vast majority of cases, we know almost nothing about the function of ncRNA (the transcriptional “dark matter”); however, function is inferred from differential expression/processing of that RNA (e.g. in diseases such as cancer) or from its evolutionary conservation.    

An important feature of all functional RNAs, is the central role played by molecular structure. RNAs can fold back on themselves to form complex 2D (base paired) and 3D (atomic arrangement) shapes. These shapes govern how RNAs interact with other biomolecules (e.g. proteins, nucleic acids and small-molecules), form catalytic centers, determine molecular stability (e.g. lifetime) of RNA, and more. The major goal of the Moss Lab is to identify RNA sequences with a high propensity to form structure, deduce that structure and then determine the function of that RNA and the roles played by its structure. To do this we use tools from three different disciplines:

Bioinformatics: Extensive databases of 2D RNA structural motifs and their experimentally-measured free energies of folding are available. These data are incorporated into folding algorithms that can attempt to predict the native RNA 2D structure. Predicted structures and energies can be complemented by sequence analysis and other bioinformatic approaches to identify RNAs (or regions of RNA) that are likely to fold into functional structures. We are interested in adapting or improving approaches for structured RNA discovery; particularly in creating protocols that best make use of existing algorithms to exploit the unique features of our targets of interest.

Biochemistry: Prediction of an RNA structure based on the thermodynamics alone is expected to yield a structure that has roughly 70% of the base pairs predicted correctly. This is due to limitations in the model used and inaccuracies in the measured thermodynamics of small motifs. To overcome these limitations, it is possible to use biochemical structure probing, where small molecules that react with nucleotides in a structure-specific manner are used to constrain (or validate) predicted structures. We are particularly interested in coupling probing experiments to high-throughput sequencing readout of modifications sites to gain structural information for large transcripts or even transcriptomes.

Biology: RNA biology is an incredibly active field and one that is greatly facilitated by the presence of a wide array of modern tools (e.g. RNA-Seq, CRISPR/Cas9, etc.). We are applying these tools to understand the biological roles RNAs that we discover and the significance of their modeled structures. This can range from assessing their effects on gene expression, processing, identifying interactors within the cell (e.g. proteins) and more. Although a linear progression can be inferred from bioinformatics, to biochemistry and biology, each approach also informs the others and we tackle RNA holistically.


Currently we are focused on three major projects:

Viral ncRNAs: Much like the hosts they infect, viruses also generate their own ncRNAs. These viral ncRNAs are used to modulate host cells to facilitate infection and are often associated with disease. We are currently studying two viral ncRNAs generated by the Epstein–Barr virus (EBV) during a particularly oncogenic form of latent infection. These ncRNAs are derived from introns and accumulate to very high levels in infected human blood cells: they are thus names the stable intronic sequence (sis)RNAs -1 and -2. sisRNA -1 and -2 are 81 and 2971 nt long, respectively, and are highly conserved in sequence and structure within EBV strains and throughout related herpes viruses. We are working to validate their predicted structures, identify interacting molecules, determine their localization within the cell and determine their effects on host cells. A better understanding of the sisRNAs could help explain how EBV latent infection is connected to certain cancers.

Viral cis-regulatory elements: Many RNAs have regulatory elements encoded within their own sequences (in cis) that affect gene expression in a number of ways. These cis-regulatory elements are typically highly-structured (e.g. frameshift pseudoknots, riboswitches and internal ribosome entry sites). We have discovered many regions in EBV RNAs that have high propensity to form structure overlap with untranslated regions (UTRs), within intron, or at intron-exon junctions. These structured regions have a high likelihood of being cis-regulatory elements and we seek to understand their exact roles. For example, we are currently working on intronic and junctional structures found in the pre-mRNA for a critical viral latency protein: latent membrane protein (LMP)2. Specifically, we are investigating how structure is affecting the rate of splicing and isoform abundance: e.g. through modulating the accessibility or distance between splicing regulatory motifs (e.g. splice sites, branch points, enhancers, etc.) and the recruitment of regulatory proteins.     

Host-virus interactions: In order to maintain and propagate infections, viruses must affect host cells in a myriad of ways (e.g. altering host gene expression). One possible route by which viruses may be affecting host cells, is through altering host RNA structure. In addition to the most thermodynamically stable fold, RNAs have many near-energy “suboptimal” folds that can also be populated. The energy to alter this equilibrium can be small, and is easily provided by interactions with proteins (or other biomolecules), post-transcriptional RNA modifications, and other perturbations to the cellular environment: all of which can be induced by viral infection. We are investigating how different forms of EBV infection can be affecting folding in human RNAs and targeting these sites for further analyses to determine their roles in infection and disease.