Transcription is the process of copying a fragment of DNA in the cell's nucleus into RNA. This copy is then used as a template to produce proteins, or it functions by itself as an enzyme, structural element or regulator. Transcription of protein-coding genes in eukaryotes is achieved by RNA polymerase II (Pol II), an enzyme that is tightly regulated to allow for the adaptation of transcript levels to both extracellular conditions as well as intracellular needs. My research has focused on understanding transcriptional regulation by two distinct factors: non-coding RNAs (ncRNAs) that are upregulated in response to cellular stress, and the DNA helicase RECQL5, a member of the highly conserved family of RecQ helicases involved in DNA repair.
Non-coding RNAs are an important transcriptional regulator when cells adapt to extreme conditions such as heat shock. In mouse and human cells, heat shock triggers an increase in levels of B2/B1 RNA and Alu RNAs, respectively, which regulate expression of protein-coding genes by Pol II. Although it had been shown that ncRNAs interact directly with Pol II to regulate transcription, many important questions remained unanswered: Where is the binding site for ncRNAs located? Does binding of ncRNAs interfere with the binding of DNA to Pol II? How are repressive and non-repressive ncRNAs, which are both upregulated in response to heat shock and which both bind to Pol II with high affinity, distinguished? To address these questions, I employed single-particle cryo-electron microscopy (cryo-EM) to determine the structures of human Pol II in complex with six different repressive and non-repressive ncRNAs from mouse and human. The structural data allowed me to identify a conserved docking site for ncRNAs in the active site cleft of Pol II; the location of this site was later confirmed independently by cross-linking studies in collaboration with the laboratory of James Goodrich. Collectively, my analysis of the cryo-EM reconstructions of ncRNA-Pol II complexes in conjunction with biochemical data from the Goodrich lab suggest that the distinction between repressive and non-repressive ncRNAs is made by the general transcription factor TFIIF based on certain flexible RNA elements that extend beyond the docking site.
RECQL5 is a DNA helicase implicated to function at the interface of the cellular DNA replication, DNA repair, and RNA transcription machineries. Although RECQL5 had previously been shown to interact directly with Pol II, its molecular mechanism of action remained elusive. My work aimed to answer the following questions: Where is the binding site for RECQL5 located on the surface of Pol II? Does binding of RECQL5 interfere with the binding of DNA or other transcription factors during transcription initiation or elongation? How is transcriptional repression by RECQL5 achieved at the molecular level? To answer these questions, we employed an integrative experimental approach, combining biochemical assays, X-ray crystallography, cryo-EM and small angle X-ray scattering. The crystal structure of a fragment of RECQL5's Pol II binding domain suggested that the topology of this domain is similar to a domain found in the transcription elongation factor TFIIS, which promotes continued transcription of arrested elongation complexes by stimulating the intrinsic RNA cleavage activity of Pol II. Using pull-down assays, I showed that RECQL5 and TFIIS compete for binding to Pol II, suggesting that the two proteins bind to overlapping sites. I corroborated these initial findings using an in vitro transcription assay, which confirmed that binding of RECQL5 to Pol II interferes with the function of TFIIS to promote read-through of intrinsic blocks to elongation. Using cryo-EM, I obtained a high-resolution reconstruction of an elongating Pol II complex repressed by RECQL5. By docking the known crystal structures of individual components into the EM map, I generated a pseudo-atomic model of the complex. This model confirmed the location of the binding site, and suggests a novel, dual mechanism for the regulation of transcription by RECQL5 that includes structural mimicry of the Pol II-TFIIS interaction.
Both ncRNAs and RECQL5 are important regulatory factors in human cells whose molecular mechanisms of transcriptional repression remained unknown. My research has provided important insights into their structure and function and, in the case of RECQL5, uncovered a novel mechanism of transcription regulation that might be employed by a number of other factors involved in transcriptional repression at the interface of the DNA recombination, replication and repair machineries.