Chemistry and Biology of DNA-protein cross-links

Abstract

DNA-protein cross-links (DPCs) are ubiquitous DNA lesions formed when proteins become covalently trapped on DNA strands upon exposure to various endogenous, environmental and chemotherapeutic agents. Because of their considerable size, DPCs interfere with the progression of replication and transcription machineries, potentially contributing to mutagenesis and carcinogenesis. However, unlike small DNA lesions of which the biological consequences and repair mechanisms have been well characterized, biological effects and repair mechanism of DPC lesions remain to be established. A significant challenge in the field is the structural diversity of DPC lesions and the scarcity of experimental mythologies to create site-specific DNA-protein conjugates. The main objective of this thesis was to synthesize model DPC and to investigate their biological consequences and repair mechanisms. In Chapter II, we discovered, characterized and quantified 5-formylcytosine(5fC) mediated DNA-histone conjugates in human cells. 5-Formylcytosine (5fC) is an endogenous DNA modification enzymatically generated in the genome as an oxidation product of 5-methyl-dC (5mC). While 5mC is known to be an epigenetic mark that controls the levels of gene expression, the biological functions of 5fC are incompletely understood. In this chapter, we discovered that 5fC bases in DNA readily form Schiff base conjugates with Lys side chains of nuclear proteins such as histones, forming covalent DNA-protein conjugates. Isotope dilution nanoLC-ESI-MS/MS methodology was employed to detect and quantify 5fC-lys conjugate in human cells. We hypothesize that reversible 5fC-histone cross-linking contributes to epigenetic signaling, transcriptional regulations and chromatin remodeling. After the discovery of 5fC-mediated DNA-histone crosslinks in mammalian cells, we investigated their effects on DNA replication in Chapter III. DNA substrates containing site-specific DPCs were subjected to in vitro translesion synthesis (TLS) in the presence of TLS DNA polymerases. We found that DPCs containing various full-length proteins conjugated to DNA via the C-5 position of cytosine completely blocked human DNA polymerases, while the corresponding lesions containing shorter peptides were bypassed by translesion synthesis (TLS) polymerases. These results are consistent with the proposed DPC repair pathway in the literature, in which full-length DPCs are subjected to proteolytic degradation to generate short DNA-peptide cross-links, which may serve as substrates for translesion synthesis. In addition, our steady-state kinetics analysis and mass- spectrometry-based sequencing and quantification revealed that the bypass of DNA- peptide cross-links by human TLS polymerases was highly error-prone, introducing significant amounts of C to T and deletion mutations. In Chapter IV, we investigated the effects of DPCs on transcription using two model DPCs where the proteins are conjugated to the C5 position of cytosine or the C7 position of 7-deazaguanine. The latter serves as a hydrolytically stable model of the N7- guanine lesions, which commonly form upon exposure to bis-electrophiles such antitumor agents. We found that full-length proteins cross-linked to either 5fC or 7-deazaguanine completely blocked T7 RNA polymerase, while relatively short peptide cross-links were bypassed, although with low efficiency. Interestingly, the two model DPCs exhibited completely different mutagenic patterns are revealed by PCR and mass spectrometry based assays. While the bypass of peptide cross-linked to 7-deaza-G by T7 RNA polymerases induced very small numbers of mutations, transcription past peptide lesions conjugated to C-5 of C induced significant amounts of C to T transcriptional mutations. In Chapter V, we investigated the effects of 5fC-mediated DNA-peptide/protein cross-links on transcription and its potential repair by nucleotide excision repair (NER) in living cells. To accomplish this goal, structurally defined DPCs were site-specifically incorporated into plasmid molecules, which was then transfected into wild type cells or cells deficient in NER. RT-PCR and LC-MS/MS based strategy was then employed to quantitatively study the effects of DPC lesions on efficiency and fidelity of transcription in mammalian cells. We found that the presence of peptide cross-links conjugated to C-5 of cytosine significantly inhibited DNA transcription in human embryonic kidney cells. However, in contrast to our in vitro results, no transcriptional mutagenesis was observed. In addition, we compared the transcription bypass efficiencies of DpC lesions in wild-type and NER-deficient cell-lines, and also conducted the in vitro NER assays using cell-free extracts from human HeLa cells. Collectively, our data suggested that 5fC-mediated DNA- peptide cross-links are poor NER substrates, requiring a different pathway for their repair. Recent studies suggested that the bulky DPCs in cells are proteolytically processed to shorter DNA-peptide cross-links before they can be tolerated by translation synthesis mechanism or removed by nucleotide excision repair. DPCs can block DNA replication, signaling for recruitment of specialized metalloprotease (Spartan). However, the mechanisms of protease-mediated DPC digestion in the absence of DNA replication are incompletely understood. In Chapter VI, we employed an immunoprecipitation(IP)-PCR methodology to demonstrate that DPCs present on non-replicating plasmids are rapidly ubiquitinylated in mammalian cells, which likely serves as a signal for the proteasome- mediated DPC processing or other ubiquitin-mediated pathways to facilitate the DPC repair.