End healing mechanisms in DNA and RNA repair

The ability to repair breaks in the phosphodiester backbone of nucleic acids is important for genomic integrity. When breaks result in 5' phosphate and 3' hydroxyl termini, the ends can be sealed by classic polynucleotide ligases in an ATP dependent manner. However, when DNA and RNA damage generate 5' hydroxyl and 3' phosphate (or 2',3' cyclic phosphate) termini, the breaks are termed 'dirty ends' since they cannot be sealed by classic ligases. Mechanistically, the dirty end problem can be solved by two distinct ways -- end healing prior to ligation by classic ligases, and by covalently activating 3' phosphate ends by linkage to GMP. This thesis investigates reaction pathways that restore dirty ends in the context of DNA and RNA repair. Specifically, I have studied end healing of DNA 3' phosphate ends by an archaeal 3' phosphoesterase, the mechanism via which T4 phage opens cyclic phosphate ends in RNA, end healing of 5' hydroxyl RNA ends by a bacterial polynucleotide kinase-phosphatase, RNA 2' phosphate end healing by E. coli RtcA, and DNA 3' phosphate end activation by RtcB, an enzyme previously characterized solely as an RNA ligase.;LigD 3' phosphoesterase (PE) enzymes perform end healing reactions at DNA 3' phosphate breaks. Via structural and mechanistic investigations on a PE enzyme from the archaeon Candidatus Korarchaeum cryptofilum , I established how PE selectively binds to 'soft' metals in its active site, and that proper scissile phosphate and metal coordination geometry are imperative for PE catalysis.;During T4 phage infection, E. coli obstructs viral protein synthesis by inducing cleavage of host cell tRNALys leading to 2',3' cyclic phosphate and 5' hydroxyl RNA breaks. T4 phage repairs the broken RNA backbone via end healing by a polynucleotide kinase phosphatase (T4 Pnkp), and end sealing by an RNA ligase (T4 Rnl1). Using biochemical methods, I discovered that T4 Pnkp removes RNA cyclic phosphate ends by a multi-step processive reaction which involves formation of a 3' phosphoaspartyl-Pnkp followed by a RNA 3' phosphate intermediate, finally forming the RNA 3' hydroxyl end, thereby priming the RNA for ligation by T4 Rnl1. I also discovered a hitherto unknown 2' phosphatase activity of T4 Pnkp.;Clostridium thermocellum (Cth) Pnkp typifies a different flavor of RNA repair enzyme found in diverse bacterial taxa. I solved crystal structures of CthPnkp kinase at discrete states along the reaction pathway that helped reconstruct end healing by the kinase domain at atomic resolution. An absence of contacts to the ATP nucleoside moiety in the CthPnkp•ATP•Mg2+ structure hinted at non-specific nucleotide utilization by CthPnkp kinase (which I confirmed by biochemical assays). The structure of CthPnkp•GTP•Mg2+•DNA Michaelis complex illustrated a mechanism of general acid-base catalysis and identified the determinants of phosphoacceptor recognition.;A different solution to the dirty end problem lies in covalent activation of 3' phosphate ends by linkage to GMP. I discovered a novel DNA splicing activity in E. coli RtcB, which had been previously characterized solely as an RNA ligase. I demonstrated that RtcB catalyzes direct ligation of DNA 3' phosphate/5' hydroxyl ends via a unique 3' end activating chemical mechanism, distinct from known DNA ligases insofar as it does not require 'end healing' prior to ligation, and uses GTP instead of ATP as the energy source.