Function and Structure of RNA

BII Function and Structure of RNA


Our group focuses on identifying structural and functional motifs in RNA. We use a wide range of techniques to identify and characterize loci in different types of RNA comprising of sequence analysis focusing on covariation, annotation, analysis of chemical structure probing and crosslinking data, secondary and tertiary structure prediction methods, molecular modeling, and molecular dynamics simulations. Our primary targets of research are viral genomic and subgenomic RNAs including RNA-RNA and RNA-Protein interactions within infected host cells. Systems of interest span mosquito-borne infections, Enteroviruses, Influenza, as well as marine pathogens. Additionally, we are working to identify regulatory networks at the RNA level throughout human cell differentiation. Our lab was established in 2019 and has been able to get up to speed quickly.  

BII - Function and Structure of RNA Figure 1

Figure 1: Representation of Dengue Genome (orange) in complex with capsid proteins (purple) packaged into a mature virion (grey, blue). Mature virions contain an 11 kb RNA genome that exhibits complex local and long-range structure that is essential for viral function. We identified key RNA structural features even in the coding regions of the virus and showed conclusively that the RNA structure is necessary by synonymous mutagenesis which disrupted viral fitness and compensatory mutations that were able to restore viral function.

Functional Structural Elements in Dengue and Zika Genomes

Dengue and Zika viruses pose serious health risks to populations in tropical and subtropical regions. Climate change is likely to extend the habitats of the vector for these viruses, the Aedes mosquito. Unfortunately, neither vaccines nor antiviral therapeutics are available for the treatment of these serious and potentially lethal diseases. Our lab is working to elucidate the processes of genome packaging and regulation on the RNA level to identify key steps that could be targeted therapeutically. In collaboration with researchers from A*STAR GIS and Duke-NUS Medical School we were able to elucidate the structure of dengue and Zika genomes in mature virions as well as in infected cells while the genome is undergoing translation and replication.

We were able to closely analyze all four serotypes of Dengue virus alongside 4 strains of Zika virus in our study. This allowed us to characterize conserved structural features present in all related viruses, which in turn enabled our collaborators to design directed mutations to test the functional effect of disrupting these structures. As most of the structures we identified were present in coding regions of the viral genomes, special care was necessary to preserve protein-level conservation and avoid rare codons in our mutants, as these modifications would have an indeterminable additional effect on viral fitness beyond effects solely mediated by genome structure. In addition to restricting the possible mutations we could make, this constraint made it especially challenging to design compensatory mutations to validate our findings, but nevertheless we were able to do so and hence could conclusively prove that genomic RNA structures in the coding region are essential for viral fitness.

An interesting observation in our analysis of long-range interactions was that many of the sites involved engage in multiple interactions concurrently. We found that only approximately 60% of long-range interactions in virions comprise of two unique ends, whereas the other 40% of long-range interactions have at least one side that forms alternative structures. We have yet to fully investigate the meaning and effect of these alternative competitive interactions, as our current process only allows us to observe an ensemble average of structures present in the biological sample. We have developed some ideas about how to dereplicate these interactions that we are currently trying to implement for Enteroviruses. A key question in our opinion is whether these multiple interactions are evidence of a diversity in packaged states or reflect dynamics within single virions or at different stages of the intracellular stage of the viral life cycle.

A further curious finding we encountered during this study was that contrary to our expectations, the compact, packaged structures we observed in mature virions persisted even after lysis of the viral capsid with proteinase K. Our initial assumption was that removal of the spatial confinement in the capsid would be sufficient to effect the transition from the compact packaged form to the more accessible intracellular state, but this was inconsistent with our observations and indicates that a separate cellular process is responsible for opening up the packaged genome. We are trying to identify cellular components involved in this process as they might prove attractive targets for antiviral therapy.

Dengue and Zika RNA-Protein Interactions

In addition to intrinsic structures of the genomic RNA itself, viral RNA interacts with a variety of proteins. One of the most important interactions in the viral life cycle is the interaction of flaviviral RNA with their respective capsid proteins. These proteins are highly charged dimers with extended intrinsically disordered regions that have been shown to be crucial for their function. We investigated the structure and dynamics of these regions recently. Our current focus is on the interactions of capsid protein with genomic RNA which we are pursuing in two directions.

BII - Function and Structure of RNA Figure 2 

Figure 2: Domain organization of Dengue RNA. Identified long-range interactions allowed us to construct a computational model through Monte-Carlo simulation with constraints. This approach allowed us to sample a considerable number of states and settle in a minimum that fulfills most observed long-range interactions. Certain sites exhibit competing long-range interactions and thus several other states might be relevant in the viral life cycle as well.

The first avenue of inquiry is the distribution of capsid binding along the viral genome in the packaged state. It is generally assumed that the highly charged capsid protein fulfills a histonelike function in sequestering genomic RNA. We investigated the distribution of binding along the full genome and identified hotspots that preferentially bind RNA in vitro as well as in packaged mature virions. We were able to demonstrate that the structure inside virions differs from naïve binding to in vitro transcripts. Moreover, the binding pattern is relatively specific contrary to initial expectations, with considerable preference for structured regions and avoidance of interactions involved in long-range interactions. This leads us to conclude that local structural elements likely play a role in the early association of the viral genome with membrane-associated capsid protein at the time of packaging. Long-range interactions would then be formed throughout the budding process.

Building on that hypothesis, we are interested in characterizing the binding modes that occur between RNA segments and the capsid protein. We have reason to believe that structured and unstructured RNA regions interact with distinct regions on the capsid protein. We have also used molecular dynamics simulations to demonstrate that capsid binding likely involves significant structural distortion in target RNA. Our aim regarding the identification and characterization of binding modes is mainly to explain the observed structure and sequence specificity outlined above. Understanding the drivers for efficient packaging may on one hand allow us to target the underlying mechanism for therapy and on the other hand enable us to repurpose the viruses as delivery vehicles for drugs or gene therapeutics.

BII - Function and Structure of RNA Figure 3 

Figure 3: Identification of RNA-protein interactions by footprinting and cross-linking data. The genomic RNA is protected by or crosslinked with the protein segments bound at a specific location. Subsequent sequencing of the remaining products after digestion allows for the identification of these sites.

Miscellaneous Projects in 2019

Besides these major areas of focus, our group was able to contribute to a considerable number of research projects within BII and internationally. We managed to model specific polyanionic glycosaminoglycan mimetics that have shown promise as therapeutic agents among others in an antiviral context. We build a complex model of Mycobacterial ATP synthase by denovo modeling a fused segment in combination with EM maps. We identified conformational rearrangement in BRAF upon inhibition for specific cancer-associated BRAF mutants. We finalized a complex quantum mechanical and spectroscopic study on the conformational dynamics of carbonic acid.


Assistant Principal Investigator HUBER Roland G.   |    [View Bio]  
Senior Post-Doctoral Research Fellow DEFALCO Louis
Senior Post-Doctoral Research Fellow DELLI PONTI Riccardo  

Selected Publications