|dc.description.abstracteng||Splicing is a crucial post-transcriptional processing event that entails the removal of non-coding intervening sequences (introns) from eukaryotic pre-mRNA and the ligation of the coding sequences (exons). It is carried out in a two-step reaction by the spliceosome, a giant and highly dynamic protein-rich ribonucleoprotein (RNP) enzyme. The spliceosome consists of five major subunits, U1, U2, U4/U6 and U5 snRNPs and multitude of non-snRNP proteins. The active center of the spliceosome only develops de novo on the pre-mRNA by a stepwise assembly of U snRNPs that is driven by several DExD/H-box ATPases/RNA helicases. Major structural and compositional rearrangements are required to render the spliceosome catalytically competent for promoting the two steps of splicing. The enzyme Brr2 plays a major role in this catalytic activation process. Brr2 is an exceptionally large DExH-box protein (ca. 250 kDa), and is a member of the Ski2-like RNA helicases in the spliceosome that stands out both structurally and functionally among other splicesosomal DExD/H-box proteins. It is composed of two putative helicase cassettes fused in tandem. Each helicase cassette contains conserved dual-RecA-like domains, flanked by a winged helix (WH) domain and a Sec63 homology unit of unknown function that may bestow specific properties upon the helicase. Brr2 is an integral component of the U5 snRNP and unlike other spliceosomal helicases it is preassembled with one of its substrates, the U4/U6 snRNPs, before recruitment to the pre-spliceosome. Furthermore, Brr2 remains stably associated with the splicesosome to function again during the disassembly step of the spliceosome. Thus, the RNPase activity of Brr2 needs to be reliably controlled to facilitate its multiple usages in the spliceosome. Indeed, Brr2 forms a stable complex with two U5 snRNP proteins, a large scaffolding protein Prp8 and the EF-2 like GTPase Snu114, both of which have been implicated in the regulation of Brr2 activity. In human, several mutations within Brr2 and the C-terminal tail of Prp8 cause a severe type of retinitis pigmentosa (RP), a progressive retinal dystrophy. It was hitherto unclear how Brr2 structurally and functionally adopts these capabilities and how the regulatory role of Prp8 on Brr2 can lead to a precise timing of the spliceosome activation and thus U4/U6 RNA unwinding by Brr2. In addition, the molecular basis of the way in which several RP-linked mutations in Brr2 and Prp8 may lead to the disease retinitis pigmentosa remained poorly understood. In this work, a crystal structure of the C-terminal Sec63 unit of Brr2 solved in collaboration with V. Pena and M. Wahl revealed the first insight into the structural similarity of the Brr2 helicase units with the DNA helicase Hel308. Guided by the Hel308 structure, the architecture of both Brr2 helicase cassettes could be modeled as a composite dual Hel308-like helicase. Functional roles for various predicted structural elements of Brr2 were then validated by mutational analysis in vitro and in living yeast cells. The results supported the idea that in analogy to Hel308 a conserved β-hairpin loop in the RecA-2 domain of the N-terminal helicase cassette may act as a strand separation device, during unwinding of U4/U6 RNAs.
More recently, the crystal structure of a larger fragment of human Brr2, encompassing both helicase cassettes, solved in collaboration with K. Santos and M. Wahl, revealed an extensive interaction surface between the C-terminal cassettes (respectively, Brr2NCand Brr2CC), and provided a framework for a detailed structure-based mutational analysis of Brr’s enzymatic activities. It could be shown that only the isolated Brr2NC harbors ATPase and helicase activities and that it threads single-stranded RNA through a central tunnel and across a helix-loop-helix domain during duplex unwinding. Although the Brr2CC is inactive on its own, it strongly stimulates the activity of the N-terminal cassette. Mutations of amino acid residues involved in the communication between the two cassettes, as well as mutations that interfere with the nucleotide-binding pocket of Brr2CC, strongly affected ATPase and/or helicase activities of the enzyme. Thus, while the Brr2CC does not seem to engage RNA, it binds ATP and acts as an intramolecular cofactor to stimulate Brr2NC helicase activity. Using various U4/U6 mutant constructs I was also able to show that Brr2 interacts with the single-stranded region of U4 preceding U4/U6 stem I (the U4 central domain), and translocates in a 3’ to 5’ direction along the U4 strand to unwind the U4/U6 stem I first.
In the second part of the work for this thesis I investigated the roles of the C-terminal RNase H-like (RH) and Jab1/MPN-like domains of Prp8 in the regulation of Brr’s enzymatic functions. Using UV-induced RNA-protein crosslinking and RNA structure probing methods I could show that the RNase H domain of Prp8 forms a specific complex with U4/U6 snRNAs in vitro, where it binds to a single-stranded region of U4 preceding U4/U6 stem I. Using mass spectrometry, RNA-protein crosslinks could be mapped at the base of a hairpin loop (β-finger) of the RH domain. Moreover, I was able to show that the Prp8 RNase H domain interferes with Brr2-mediated U4/U6 unwinding by sequestering Brr2’s targeting site, indicating that the RH domain negatively regulates Brr2 function and acts as a keeper to prevent premature activation of the spliceosome. These findings also support the idea that the Prp8 RH domain acts as a platform for the handover of the 5'-splice site from U1 to U6 snRNA prior to the activation step.
The Prp8 Jab1 domain is a ubiquitin-binding domain that comprises a globular domain followed by a protruding C-terminal tail, which is partly unstructured in the isolated Jab1 domain, and which represents a hotspot for mutations leading to retinitis pigmentosa. Using biochemical in vitro assays, I was able to show that the Jab1 domain binds only to the N-terminal helicase cassette and inhibits the helicase and ATPase activities of Brr2 by preventing Brr2 loading onto its RNA substrate U4/U6. Upon deletion of the unstructured C-terminal 16 amino acids, Jab1Δ16 now strongly stimulated Brr2’s ATPase and helicase activities, suggesting that the C-terminal tail of Jab1 may interfere with Brr2’s RNA binding capacity. The crystal structure of Brr2 in complex with the intact Jab1 domain, which was obtained in collaboration with M. Wahl, revealed the molecular basis for the biochemical observations. Jab1 rests with its globular part primarily on the IG-like domain of Brr2NC while the C-terminal tail interacts with the RNA binding motifs of the RecA domains, thus occluding the RNA binding tunnel of the N-terminal helicase cassette. I was also able to show that under conditions favoring RNA binding, the full-length Jab1 domain acts as a coactivator of Brr2, enhancing the coupling of ATP hydrolysis to duplex unwinding and the processivity of the helicase. This delicate regulation requires the dual-cassette organization of Brr2 and is not observed with the isolated N-terminal helicase cassette. Finally, I have investigated the effect of various RP-linked mutations in the Prp8 Jab1 domain on the regulation of Brr2’s activities in vitro and on the stability of tri-snRNP formation, cell viability and pre-mRNA splicing in vivo in yeast cells. Taken together, the results obtained uncover the mechanism underlying a unique dual-mode regulation of a superfamily 2 helicase by a protein cofactor and reveal that its disruption constitutes a disease principle underlying certain forms of retinitis pigmentosa.||de