DNA nanotechnology uses the Watson-Crick base-pairing of DNA to self-assemble structures at the nanoscale. DNA nanomachines are active structures that take energy from the system to drive a programmed motion. In this thesis, a new design for a reversible DNA motor and an automatically regenerating track is presented. Ensemble fluorescence measurements observe motors walking along the same 42nm track three times. A second new motor was designed to allow motors on intersecting tracks to block each other, which can be used to perform logical computation. Multiple design approaches are discussed. The chosen approach showed limited success during ensemble fluorescence measurements. The 'burnt bridges' motor originally introduced by Bath et al. 2005 was also sent down tracks placed along the inside of stacked origami tubes that are able to polymerise to micrometre lengths. Preliminary optical microscopy experiments show promise in using such a system for observing micrometre motor movement. Scaffold-based DNA origami is the technique of folding a long single-stranded DNA strand into a specific shape by adding small staple strands that hold it in place. Extended staple strands can be modified to functionalise the origami surface. In this thesis, the threading of staple extensions through a freely-floating origami tile was observed using single-molecule Förster resonance energy transfer (smFRET). Threading was reduced by bracing the bottom of the extension or by using a multilayered origami. smFRET was also used to investigate the process of staple repair, whereby a missing staple is added to a pre-formed origami missing the staple. This was found to be successful when the staple is single-stranded, and imperfect when partially double-stranded. Finally the idea for a new "DNA cage", a dynamic octahedron called the "Holliday Octahedron", is presented. The octahedron is made of eight strands, one running around each face. Mobile Holliday junctions at each face allow the stands to rotate causing a conformational change.

付記する学位プログラム名: 充実した健康長寿社会を築く総合医療開発リーダー育成プログラム / 京都大学 / 新制・課程博士 / 博士(医科学) / 甲第23114号 / 医科博第125号 / 新制||医科||8(附属図書館) / 京都大学大学院医学研究科医科学専攻 / (主査)教授 斎藤 通紀, 教授 篠原 隆司, 教授 滝田 順子 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第20212号 / 理博第4297号 / 新制||理||1617(附属図書館) / 京都大学大学院理学研究科生物科学専攻 / (主査)教授 鹿内 利治, 准教授 小山 時隆, 教授 長谷 あきら / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

No description available.

Yes / Abstract The budding yeast genome contains regions where meiotic recombination initiates more frequently than in others. This pattern parallels enrichment for the meiotic chromosome axis proteins Hop1 and Red1. These proteins are important for Spo11-catalyzed double strand break formation; their contribution to crossover recombination remains undefined. Using the sequence-specific VMA1-derived endonuclease (VDE) to initiate recombination in meiosis, we show that chromosome structure influences the choice of proteins that resolve recombination intermediates to form crossovers. At a Hop1-enriched locus, most VDE-initiated crossovers, like most Spo11-initiated crossovers, required the meiosis-specific MutLγ resolvase. In contrast, at a locus with lower Hop1 occupancy, most VDE-initiated crossovers were MutLγ-independent. In pch2 mutants, the two loci displayed similar Hop1 occupancy levels, and VDE-induced crossovers were similarly MutLγ-dependent. We suggest that meiotic and mitotic recombination pathways coexist within meiotic cells, and that features of meiotic chromosome structure determine whether one or the other predominates in different regions.

Homologous recombination (HR) is essential for the repair of DNA doublestrand breaks (DSBs) and damaged replication forks. However, HR can also cause gross chromosomal rearrangements (GCRs) by producing crossovers (COs), resulting in the reciprocal exchange of sequences between non-sister chromatids. Therefore, HR-mediated GCRs are suppressed via the promotion of HR pathways that favour noncrossover (NCO) formation, such as the synthesis-dependent strand annealing (SDSA) and dissolution pathways, which are modulated by Mph1 and Sgs1 helicases, respectively. The mismatch repair (MMR) pathway is intricately associated with HR via its roles in repairing mismatches on heteroduplex DNA that can arise during HR and in preventing homeologous recombination. Using a plasmid break-repair assay, we have revealed a novel, MMR-independent role of MutSα in promoting the formation of a subset of COs that is specifically supressible by Mph1, during HR between two completely homologous sequences. In contrast, the MMR-dependent function of MutSα, together with Mph1 and Sgs1, was shown to be required for the suppression of CO formation during homeologous recombination. These data indicate that Mph1 can both antagonise and promote the functions of MutSα during DSB repair, depending on the levels of homology between the two recombining sequences. COs are generated by the resolution of Holliday junction (HJ) intermediates formed at the terminal stages of HR. Several S.cerevisiae proteins such as Yen1, Mus81, Slx1 and Rad1 have been implicated in HJ resolution. However, the in vivo roles of these proteins in HJ resolution remain to be confirmed. To directly and quantitatively monitor in vivo HJ resolution in S.cerevisiae, a transformation-based HJ resolution assay using a plasmid-borne HJ substrate has been developed. Using this system, we have demonstrated an in vivo HJ resolution function of Yen1, which acts redundantly with Mus81. Moreover, these redundant activities of Yen1 and Mus81 are essential for survival during replication stress, but are dispensable for DSB repair. An Slx4 and Rad1-dependent in vivo HJ resolution activity was also observed in the absence of Yen1 and Mus81 that was suppressed by presence of Slx1. Models describing how the nucleases interact to process HJs in vivo will be discussed.

Homologous recombination (HR) is a highly conserved cellular process involved in the maintenance of chromosomal integrity and generation of genetic diversity. Biochemical and genetic studies have suggested that HR is crucial for repair of damaged DNA arising from various endogenous or exogenous assaults on the genome of any organism. Further, HR is vital to repair fatal DNA damage during DNA replication. An instructive example of cross-talk between the processes of DNA recombination and replication can be construed in the processing of replication/recombination/repair intermediates. The impediment(s) to the progression of DNA replication fork is one of the underlying causes for increased genome instability and consequently this might compromise the survival of organism. Various processes manifest at stalled replication forks before they can be rendered competent for the replication-restart. One of the mechanisms of replication-restart involves replication fork reversal (RFR), which envisage unwinding of the blocked forks with simultaneous annealing of the parental and daughter strands o generate a Holliday junction intermediate adjacent to DNA double strand end. Genetic evidence shows that in E. coli dnaEts mutant, holD mutant and in helicase defective rep mutant, RFR is catalyzed by RuvAB complex. Classically, HJ intermediates are generated during the terminal stages of the HR pathway. In E. coli, branch migration and resolution of HJ intermediates is promoted by RuvA, RuvB and RuvC proteins, which participate at the late stages of HR. Structural, biochemical and mutational analysis suggest that E. coli RuvA binds Holliday junction DNA with high affinity and specificity. RuvB, a member of the AAA+ (ATPase associated with various cellular activities) family, is recruited to the RuvA-Holliday junction complex and functions as a motor protein. Together, RuvA and RuvB catalyze ATP dependent branch migration of HJ. The resolution of HJ is catalyzed by the RuvC endonuclease, which introduces coordinated cuts at two symmetrical sites across the junction. RuvAB complex, the Holliday junction branch migration apparatus, is ubiquitous in bacteria. Genetic, biochemical and structural studies have not only established the in vivo role of E. coli RuvAB, in context of HR pathway, but have also provided valuable insights into the mechanism of HJ processing by RuvAB complex. However, the paucity of extensive studies examining the biochemical properties of each member of the RuvABC protein complex restricts models in deciphering the functions of the individual components of this tripartite protein complex. Our current understanding of the biochemical function of E. coli RuvA is within the context of its interacting cellular partner, RuvB. Consequently, the inherent activities of RuvA in the context of DNA repair and HR are poorly understood. Moreover, it remains to be ascertained if RuvABC protein complex, its different sub-complexes, or the individual subunits can function differently in the processing of HJ intermediates generated during DNA repair and HR. The information from these studies would be helpful in understanding the mechanistic details of HR pathway in mycobacteria. Additionally, a number of important questions regarding the molecular basis of RuvAB catalyzed fork reversal remain unanswered. Therefore, exploration of biochemical details of the RuvAB mediated RFR would provide mechanistic insights into the dynamics of fork reversal process. Moreover, analysis of RuvAB catalyzed RFR might be helpful in validating the different assumptions of the RFR model that has been proposed on the basis of genetic analysis of certain E. coli replication mutants. Another interesting question that remains to be answered is, how under in vivo conditions, RuvABC protein complex or its individual subunits are regulated to function differently in the context of HR and DNA repair? Mycobacterium tuberculosis is an important intracellular pathogen which is likely to experience substantial DNA damage inside the host and thus may require an efficient DNA recombination and repair machinery for its survival. Our knowledge about the mechanistic aspects of genetic exchange in mycobacteria is rather limited. Therefore, understanding of the processes catalyzed by the components of HR pathway may help in molecular genetic analysis of mycobacteria. Sequence analysis of M. tuberculosis genome, followed by various comparative genomic studies, has revealed the presence of putative homologs of E. coli rec genes but it is not known whether these gene products are able to catalyze the reactions similar to their E. coli counterparts. In M. tuberculosis, the genes encoding for the enzymatic machinery required for branch migration and resolution of HJ intermediates are present. The ruvA, ruvB and ruvC genes form an operon, and are probably translationally coupled. Further, these ruv genes are DNA damage inducible. The transcript level of ruvC is regulated by both RecA dependent and independent mechanisms whereas ruvA and ruvB are induced only through RecA dependent SOS response. During M. tuberculosis infection of host cells, expression of ruvA and ruvB genes is upregulated. We therefore surmise that their gene product might be required for DNA replication, recombination or repair, and would be physiologically relevant under in vivo conditions. However, the details of reactions involved in the processing of HR intermediates and rescue of stalled replication forks in M. tuberculosis remains unknown. In the initial part of this study, we have investigated the function of M. tuberculosis RuvA protein using Holliday junctions containing either homologous or heterologous core. In the later part, we have explored the ability of M. tuberculosis RuvA and RuvB proteins to catalyze in vitro replication fork reversal. M. tuberculosis ruvA gene was isolated by PCR amplification and cloned in an expression vector to generate the pMTRA construct. Genetic complementation assays, using the pMTRA construct transformed into E. coli ΔruvA mutant, indicated that M. tuberculosis ruvA is functional in E. coli and suggested that it can substitute for E. coli RuvA in conferring resistance to MMS and survival following UV irradiation. Having established the functionality of M.tuberculosis ruvA, a method was developed for heterologous over-expression and purification of M. tuberculosis RuvA protein (MtRuvA). MtRuvA was purified to homogeneity and the identity of purified protein was verified using western blot analysis using the anti-MtRuvA antibodies. Purified MtRuvA was free of any contaminating endo- or exo-nuclease activity. Biochemical functions of MtRuvA were defined by performing detailed investigations of DNA-binding and Holliday junction processing activities. Substrate specificity of purified MtRuvA was examined,through DNA binding assays, by using oligonucleotide substrates mimicking differentintermediates involved in the pathway of recombinational DNA repair. Purified M. tuberculosis RuvA exhibited high affinity for HJ substrate but also formed stable complex with replication fork and flap substrate. DNase I footprinting of MtRuvA-homologous Holliday junction complex confirmed that MtRuvA bound at the junction center. The DNase I protection conferred by MtRuvA, on homologous HJ, was two-fold symmetric; the continuous footprint was 10 bp longon one pair of symmetrical arms and 7 bp on the opposite pair of arms. In parallel, DNase footprinting of MtRuvA-heterologous Holliday junction complex generated a footprint that encompassed 16 nucleotide residues on each strand of the Holliday junction. Different crystallographic studies have envisaged an important role for RuvA in base pair rearrangement atthe center of the junction. Also, in crystal structure of tetramer of EcRuvA-HJ complex twobases at the junction center were unpaired. To explore if RuvA binding leads to helical distortionof Holliday junction, MtRuvA-HJ complexes were subjected to chemical probing with KMnO4.In case of heterologous HJ, binding of MtRuvA resulted in appearance of sensitive T residues at the junction crossover. By contrast, binding of MtRuvA to homologous HJ rendered the T residues at the junction center and within the homologous core sensitive to oxidation by KMnO4.Taken together, these observations suggested that binding of MtRuvA distorts two base pairs at the junction crossover in heterologous HJ, whereas in case of homologous HJ base pairs distortion extends into the arms of the junction. These observations with KMnO4 probing were independently validated, in real time, by using sensitive to 2-aminopurine fluorescence spectroscopy measurements of MtRuvA-HJ complexes. To follow structural distortions upon interaction with MtRuvA, HJ variants carrying 2-AP substitution were generated for both homologous and heterologous HJ substrate. In each junction species, the 2-AP residue was uniquely present either at the junction center, adjacent to the center or away from the center. Incase of heterologous HJ, binding of MtRuvA resulted in increase of fluorescence emission of2-AP residues located at the junction crossover but not those of 2-AP residues that were present1-2 base pairs away from the junction center. Binding of MtRuvA to homologous HJ resulted in increase of fluorescence emission of 2-AP residues located at the junction crossover. Further, increase in fluorescence emission was also observed for 2-AP residues present within the homologous core or adjacent to the homologous core in a pair of symmetrically related arms. Thus, 2-AP fluorescence results suggested that binding of MtRuvA to homologous HJ causes base pair distortion within and adjacent to the homologous core whereas in case of heterologous HJ the base pair distortion is restricted to the junction center. Together, these results suggest thatMtRuvA causes two distinct types of base pair distortions between homologous and heterologous HJ substrates. To explore the relationship between binding of MtRuvA and alterations in global structure of the junction DNA, we employed the established technique of comparative gel electrophoresis. Analysis of data from comparative gel electrophoresis revealed that MtRuvA, upon binding to the Holliday junctions, converts the stacked-X structure of HJ to square-planar form and stabilizes the same for loading of RuvB rings and subsequent branch migration by RuvAB complex. Our results underline the possible existence of distinct pathways for RuvA function, which presumably depend on the structure and the nature of the DNA repair or HR intermediates. In summary, our results show that binding of MtRuvA to the HJ induced changes in the local conformation of junction, which might augment RuvB catalyzed branch migration. An unexpected finding is the observation that MtRuvA causes two distinct types of structural distortions, depending on whether the Holliday junction contains homologous or heterologous core. These observations support models wherein RuvA facilitates, in a manner independent of RuvB, base pair rearrangements at the crossover point of both homologous and heterologous Holliday junctions. Although the genetic basis of ruvA ruvB catalyzed RFR in E. coli has been understood in some detail but less is known about the genetic and molecular mechanism of fork reversal in mycobacteria or other organisms. Specifically, to examine if the E. coli paradigm can be generalized to other RuvAB orthologs, we explored the RFR activity of M. tuberculosis RuvAB using a series of oligonucleotides and plasmid-based substrates that mimic stalled replication fork intermediates. This approach might be useful in genetic analysis of factors involved in processing of stalled forks in M. tuberculosis wherein technical difficulties associated with the isolation and characterization of appropriate mutants have limited our understanding of DNA metabolism. Importantly, we have asked the questions as to how the structure at fork junction, extent of reversal and presence of sequence heterology might determine the outcome of RuvAB mediated RFR. The results from this study will be helpful in consolidating the proposed in vivo role for RuvAB complex in fork reversal. The open reading frame corresponding to M. tuberculosis ruvB gene was PCR amplified and cloned in an expression vector to generate the pMTRB construct. Genetic complementation assays were performed to assess the functionality of M. tuberculosis ruvB in E. coli ΔruvB mutant. The data from these assays suggested that M. tuberculosis ruvB is active in E. coli and it is able to make functional contacts with E. coli RuvA. Moreover, the efficient alleviation of MMS toxicity in E. coli ΔruvB mutant suggested that M. tuberculosis ruvB might have a role in relieving replication stress generated under specific in vivo conditions. For biochemical analysis, M. tuberculosis RuvB protein (MtRuvB) was over-expressed in a heterologous system and purified to homogeneity. The identity of purified MtRuvB was verified using western blot analysis using the anti-MtRuvB antibodies. Purified MtRuvB was free of any contaminating endo- or exo- nuclease activity. The DNA-binding properties of MtRuvB were analyzed, in conjunction with its cognate RuvA, by using different substrates that are most likely to occur as intermediates during the processes of DNA replication and/or recombination. MtRuvAB bound HJ, three-way junction and heterologous replication fork with high affinity but with relatively weaker affinity to flap and flayed duplex substrates. MtRuvB displayed very weak affinity for linear duplex and failed to bind linear single-stranded DNA. The high affinity of MtRuvB for HJ substrate, in presence of its cognate RuvA, is indicative of direct and functional interaction between RuvA and RuvB. To further test this idea, the catalytic activity of MtRuvB was assayed in the in vitro HJ branch migration assay. In this assay,MtRuvB, in association with its cognate RuvA, promoted efficient branch migration of homologous HJ over heterologous HJ. To decipher the role of MtRuvAB in processing of stalled replication fork we performed in vitro replication fork reversal (RFR) assay using both oligonucleotide and plasmid based model replication fork substrates. Initially, binding of MtRuvAB to different homologous fork (HomFork) substrates was analyzed using the electrophoretic mobility shift assays. MtRuvAB exhibited similar binding affinity towards different HomFork substrates bearing different spatial orientation of nascent leading and lagging strands. To gain insight into the role of MtRuvAB in processing of replication forks, in vitro RFR reactions were carried out using an array of synthetic homologous fork substrates. In all these reactions, MtRuvAB catalyzed efficient fork reversal leading to generation of both parental duplex and daughter duplex. In the kinetics of fork reversal reaction, for all the fork substrates,the accumulation of daughter duplex increased with time whereas the increase in parental or nascent strand DNA was negligible. Taken together, our results suggest that MtRuvAB can efficiently catalyze in vitro replication fork reversal reaction to generate a Holliday junction intermediate thus implicating that RuvAB mediated fork reversal involves concerted unwinding and annealing of nascent leading and lagging strands. Equally important, we demonstrate the reversal of forks carrying hemi-replicated DNA, thus indicating that MtRuvAB mediated fork reversal is independent of symmetry at the fork junction. For understanding the role of RuvAB mediated processing of stalled forks at chromosome level, the fork reversal assays were performed using plasmid derived model “RF” substrate. Fork reversal was monitored by restriction enzyme digestion mediated release of 5’ end labeled fragments of specific size from the fourth arm extruded at the branch point of fork junction. In these reactions MtRuvAB complex was proficient at generating the reversed arm de novo from the RF substrate. Further, MtRuvAB complex catalyzed extensive fork reversal as analyzed by release of linear duplex of2.9 kb from a JM substrate. Use of non hydrolysable analogs of ATP and analysis of restriction digestion mediated release of duplex fragments from the reversed arm suggested that MtRuvAB catalyzed RFR reaction is ATP hydrolysis dependent progressive and processive reaction. MtRuvAB complex catalyzed fork reversal on plasmid substrate that had been linearized thus indicating that MtRuvAB mediated RFR is uncoupled from DNA supercoils in the substrate. Notably, MtRuvAB promoted reversal of forks in a substrate containing short stretch of heterologous sequences, indicating that sequence heterology failed to impede fork reversal activity of MtRuvAB complex. These results are discussed in the context of recognition and processing of varied types of replication fork structures by RuvAB enzyme complex.

Meiosis is a specialized type of cell division where two rounds of chromosome segregation follow a single round of DNA duplication resulting in the formation of four haploid daughter cells. Once the DNA replication is complete, the homologous chromosomes pair and recombine during the meiotic prophase I, giving rise to genetic diversity in the gametes. The process of homology search during meiosis is broadly divided into recombination-dependent (involves the formation of double-strand breaks) and recombination-independent mechanisms. In most eukaryotic organisms, pairing of homologs, recombination and chromosome segregation occurs in the context of a meiosis-specific proteinaceous structure, known as the synaptonemal complex (SC). The electron microscopic visualization of SC has revealed that the structure is tripartite with an electron-dense central element and two lateral elements that run longitudinally along the entire length of paired chromosomes. Transverse filaments are protein structures that connect the central region to the lateral elements. Genetic analyses in budding yeast indicate that mutations in SC components or defects in SC formation are associated with chromosome missegregation, aneuploidy and spore inviability. In humans, defects in SC assembly are linked to miscarriages, birth defects such as Down syndrome and development of certain types of cancer. In Saccharomyces cerevisiae, genetic screens have identified several mutants that exhibit defects in SC formation culminate in a decrease in the frequency of meiotic recombination, spore viability and improper chromosome segregation. Ten meiosis-specific proteins, viz. Hop1, Red1, Mek1, Hop2, Pch2, Zip1, Zip2, Zip3, Zip4 and Rec8, have been shown to be the bona fide components of SC and/or associated with SC function. S. cerevisiae HOP1 (HOmolog Pairing) gene was isolated in a genetic screen for mutants that showed defects in homolog pairing and, consequently, reduced levels of interhomolog recombination (10% of wild-type). Amino acid sequence alignment together with genetic and biochemical analyses revealed that Hop1 is a 70 kDa protein with a centrally embedded essential zinc-finger motif (Cys2/Cys2) and functions in polymeric form. Previous biochemical studies have also shown that Hop1 is a structure-specific DNA binding protein, which exhibits high affinity for the Holliday junction (HJ) suggesting a role of this protein in branch migration of the HJ. Furthermore, Hop1 displays high affinity for G-quadruplex structures (herein after referred to as GQ) and also promotes the formation of GQ from unfolded G-rich oligonucleotides. Strikingly, Hop1 promotes pairing between two double-stranded DNA molecules via G/C-rich sequence as well as intra- and inter-molecular pairing of duplex DNA molecules. Structure-function analysis suggested that Hop1 has a modular organization consisting of a protease-sensitive N-terminal, HORMA domain (characterized in Hop1, Rev7, Mad2 proteins) and protease-resistant C-terminal domain, called Hop1CTD. Advances in the field of DNA quadruplex structures suggest a significant role for these structures in a variety of biological functions such as signal transduction, DNA replication, recombination, gene expression, sister chromatid alignment etc. GQs and i-motif structures that arise within the G/C-rich regions of the genome of different organisms have been extensively characterized using biophysical, biochemical and cell biological approaches. Emerging studies with guanine- and cytosine-rich sequences of several promoters, telomeres and centromeres have revealed the formation of GQs and i-motif, respectively. Although the presence of GQs within cells has been demonstrated using G4-specific antibodies, in general, the in vivo existence of DNA quadruplex structures is the subject of an ongoing debate. However, the identification and isolation of proteins that bind and process these structures support the idea of their in vivo existence. In S. cerevisiae, genome-wide survey to identify conserved GQs has revealed the presence of ~1400 GQ forming sequences. Additionally, these potential GQ forming motifs were found in close proximity to promoters, rDNA and mitosis- and meiosis-specific double-strand break sites (DSBs). Meiotic recombination in S. cerevisiae as well as humans occurs at meiosis-specific double-strand break (DSBs) sites that are embedded within the G/C-rich sequences. However, much less is known about the structural features and functional significance of DNA quadruplex motifs in sister chromatid alignment N during meiosis. Therefore, one of the aims of the studies described in this thesis was to investigate the relationship between the G/C-rich motif at a meiosis-specific DSB site in S. cerevisiae and its ability to form GQ and i-motif structures. To test this hypothesis, we chose a G/C-rich motif at a meiosis-specific DSB site located between co-ordinates 1242526 to 1242550 on chromosome IV of S. cerevisiae. Using multiple techniques such as native gel electrophoresis, circular dichroism spectroscopy, 2D NMR and chemical foot printing, we show that G-rich motif derived from the meiosis-specific DSB folds into an intramolecular GQ and the complementary C-rich sequence folds into an intramolecular i-motif, the latter under acidic conditions. Interestingly, we found that the C-rich strand folds into i-motif at near neutral pH in the presence of cell-mimicking molecular crowding agents. The NMR data, consistent with our biochemical and biophysical analyses, confirmed the formation of a stable i-motif structure. To further elucidate the impact of these quadruplex structures on DNA replication in vitro, we carried out DNA polymerase stop assay with a template DNA containing either the G-rich or the C-rich sequence. Primer extension assays carried out with Taq polymerase and G-rich template blocked the polymerase at a site that corresponded to the formation of an intramolecular GQ. Likewise, primer extension reactions carried out with KOD-Plus DNA polymerase and C-rich template led to the generation of a stop-product at the site of the formation of intramolecular I -motif under acidic conditions (pH 4.5 and pH 5.5). However, polymerase stop assay performed in the presence of single-walled carbon nanotubes (SWNTs) that stabilize I -motif at physiological pH blocked the polymerase at the site of intramolecular I -motif formation, indicating the possible existence of i-motif in the cellular context. Taken together, these results revealed that the G/C-rich motif at the meiosis-specific DSB site folds into GQ and i-motif structures in vitro. Our in vitro analyses were in line with our in vivo analysis that examined the ability of the G/C-rich motif to fold into quadruplex structures in S. cerevisiae cells. Qualitative microscopic analysis and quantitative analysis with plasmid constructs that harbour the GQ or i-motif forming sequence revealed a significant decrease in the GFP expression levels in comparison to the control. More importantly, all the assays performed with the corresponding mutant sequences under identical experimental conditions did not yield any quadruplex structures, suggesting the involvement of contagious guanine and cytosine residues in the structure formation. Prompted by our earlier results that revealed high binding affinity of Hop1 for GQ, we wished to understand the role of the GQ and i-motif structures during meiosis by analysing their interaction with Hop1 and its truncated variants (HORMA and Hop1CTD). In agreement with our previous observations, Hop1 and Hop1CTD associated preferentially with GQ DNA. Interestingly, whereas the full-length Hop1 showed much weaker binding affinity for i-motif DNA, Hop1 C-terminal fragment but not its N-terminal fragment exhibited robust i-motif DNA binding activity. We have previously demonstrated that Hop1 promotes intermolecular synapsis between synthetic duplex DNA molecules containing a G/C-rich sequence. Hence, to understand the functional role of the quadruplex structures formed at the meiosis-specific G/C-rich motif, we examined the ability of Hop1 to promote pairing between linear duplex DNA helices containing the G/C-rich motif. DNA pairing assay indicated that binding of Hop1 to the G/C-rich duplex DNA resulted in the formation of a side-by-side synapsis product. Under similar conditions, Hop1 was unable to pair mutant duplex DNA molecules suggesting the involvement of the G/C-rich motif in the formation of the synapsis product. Our results were substantiated by the observation that yeast Rad17 failed to promote pairing between duplex DNA molecules with a centrally embedded G/C-rich motif. Altogether, these results provide important structural and functional insights into the role of quadruplex structures in meiotic pairing of homologous chromosomes. The second part of the thesis focuses on the biochemical and functional properties of Red1 protein, a component of S. cerevisiae lateral element. RED1 was identified in a screen for meiotic lethal, sporulation proficient mutants. Genetic, biochemical and microscopic analyses have demonstrated the physical interaction between Hop1 and Red1. Given this, hop1 and red1 mutants display similar phenotypes such as chromosome missegregation and spore inviability and thus are placed under the same epistasis group. However, unlike hop1 mutants, red1 mutants show complete absence of SC. RED1 overexpression suppressed certain non-null hop1 phenotypes, indicating that these proteins may have partially overlapping functions. Further, although the functional significance is unknown, chromatin immunoprecipitation studies have revealed the localization of Red1 to the GC-rich regions (R-bands) in the genome, considered to be meiotic recombination hotspots. Although the aforementioned genetic studies suggest an important role for Red1 in meiosis, the exact molecular function of Red1 in meiotic recombination remains to be elucidated. To explore the biochemical properties of Red1, we isolated the S. cerevisiae RED1 gene, cloned, overexpressed, and purified the protein to near homogeneity. Immunoprecipitation assays using meiotic cells extracts suggested that Red1 exists as a Homodimer linked by disulphide-bonds under physiological conditions. We characterized the DNA binding properties of Red1 by analysing its interaction with recombination intermediates that are likely to form during meiotic recombination. Protein-DNA interaction assays revealed that Red1 exhibits binding preference for the Holliday junction over replication fork and other recombination intermediates. Notably, Red1 displayed ~40-fold higher binding affinity for GQ in comparison with HJ. The observation that Red1 binds robustly to GQs prompted us to examine if Red1 could promote pairing between duplex DNA helices with the G/C-rich sequences similar to Hop1. Interestingly, we found that Red1 failed to promote pairing between dsDNA molecules but potentiated Hop1 mediated pairing between duplex DNA molecules. Our AFM studies with linear and circular DNA molecules along with Red1 suggested a possible role of Red1 in DNA condensation, bridging and pairing of double-stranded DNA helices. Bioinformatics analysis of Red1 indicated the lack of sequence or structural similarity to any of the known proteins. To elucidate structure-function relationship of Red1, we generated several N- and C-terminal Red1 truncations and studied their DNA binding properties. Our results indicated that the N-terminal region comprising of 678 amino acid residues constitutes the DNA-binding region of Red1. The N-terminal region, called RNTF-II, displayed similar substrate specificity comparable to that of full-length Red1. Interestingly, site-directed mutagenesis studies with the Red1 C-terminal region revealed the involvement of two cysteine residues at position 704 and 707 in the disulfide bond mediated intermolecular dimer formation. Finally, to understand the functional significance of Red1 truncations we analyzed the subcellular localization of Red1 and its truncations. We made translation fusions of RED1 and its truncations by placing their corresponding nucleotide sequences downstream of GFP coding sequence in yeast expression vector. Confocal microscopy studies with S. cerevisiae cells transformed with the individual plasmid constructs indicated that the N-terminal variants localized to the nucleus, whereas the C-terminal variants did not localize to the nucleus. These results suggest that NLS-like motifs are embedded in the N-terminal region of the protein. Furthermore, other results indicated that the N-terminal region contains functions such as DNA-binding and intermolecular bridging of non-contiguous DNA segments. Altogether, these findings, on the one hand, provide insights into the molecular mechanism underlying the functions of Hop1 and Red1 proteins and, on the other, support a role for DNA quadruplex structures in meiotic chromosome synapsis and recombination.

Mon projet de recherche avait pour but de caractériser le rôle de deux protéines, ArgR et PepA, qui agissent en tant que facteurs accessoires de la recombinaison au niveau de deux sites cer du plasmide ColE1 présent dans la bactérie Escherichia coli. Ces deux protéines, couplées aux deux recombinases à tyrosine XerC et XerD, permettent la catalyse de la recombinaison site spécifique au niveau de la séquence cer, convertissant les multimères instables de ColE1 en monomères stables. Cette étude a principalement porté sur la région C-terminale de la protéine ArgR. Cette région de la protéine ArgR possède une séquence en acides-aminés et une structure similaire à celle de la protéine AhrC de Bacillus subtilis. De plus, AhrC, le répresseur de l’arginine de cette bactérie, est capable de complémenter des Escherichia coli mutantes déficientes en ArgR. Les régions C-terminales de ces protéines, montrent une forte similarité. De précédents travaux dans notre laboratoire ont démontré que des mutants d’ArgR comprenant des mutations dans cette région, en particulier les mutants ArgR149, une version tronquée d’ArgR de 149 acides-aminés, et ArgR5aa, une version comprenant une insertion de cinq acides-aminés dans la partie C-terminale, perdaient la capacité de permettre la recombinaison au niveau de deux sites cer présents dans le plasmide pCS210. Malgré cette incapacité à promouvoir la réaction de recombinaison en cer, ces deux mutants étaient toujours capables de se lier spécifiquement à l’ADN et de réprimer une fusion argA :: lacZ. Dans ce travail, les versions mutantes et sauvages d’ArgR furent surexprimées en tant que protéines de fusion 6-histidine. Des analyses crosslinking ont montré que la version sauvage et ArgR5aa pouvaient former des hexamères in-vitro de manière efficace, alors qu’ArgR149 formait des multimères de plus faible poids moléculaire. Des formes tronquées d’ArgR qui comportaient 150 acides-aminés ou plus, étaient encore capables de permettre la recombinaison en cer. Les mutants par substitution ArgRL149A et ArgRL151A ont tous montré que les substitutions d’un seul acide-aminé au sein de cette région avaient peu d’effets sur la recombinaison en cer. Les expériences de crosslinking protéine-à-protéine ont montré que le type sauvage et les formes mutantes d’ArgR étaient capables d’interagir avec la protéine accessoire PepA, également impliquée dans la recombinaison en cer. Les expériences de recombinaison in-vitro utilisant la forme sauvage et les formes mutantes d’ArgR combinées avec les protéines PepA, XerC et XerD purifiées, ont montré que le mutant ArgR149 ne soutenait pas la recombinaison, mais que le mutant ArgR5aa permettait la formation d’une jonction d’Holliday. Des expériences de topologie ont montré que PepA était capable de protéger l’ADN de la topoisomérase 1, et d’empêcher ArgRWt de se lier à l’ADN. Les deux mutants ArgR149 et ArgR5aa protègent aussi l’ADN avec plus de surenroulements. Quand on ajoute PepA, les profils de migration montrent un problème de liaison des deux mutants avec PepA. D’autres expériences impliquant le triplet LEL (leucine-acide glutamique-leucine) et les acides-aminés alentour devraient être réalisés dans le but de connaitre l’existence d’un site de liaison potentiel pour PepA. / My research project involved the role of two proteins, ArgR and PepA, which act as accessory factors in the ColE1 cer recombination system from the gram negative bacteria Escherichia coli. These two proteins, in addition to the tyrosine recombinases XerC and XerD, catalyze a site-specific recombination event at the cer sequence which converts unstable multimeric forms of ColE1 into more stable monomers. Our study mainly focused on the C-terminal end of the ArgR. This region of the ArgR protein possesses a structural and amino acid sequence similarity with the AhrC protein from Bacillus subtilis. Moreover, AhrC, the Arginine repressor of this bacterium, is able to complement Escherichia coli mutants deficient in ArgR. The C-terminal regions of these proteins, display a very high region of similarity. Previous work from our laboratory has shown that ArgR mutants with mutations in this region, especially the mutants ArgR149, a truncated 149 amino acids form of ArgR, and ArgR5aa, a form containing a five amino acid insertion in the C-terminal part, lost the ability to perform a recombination reaction at two cer sites in the plasmid pCS210. Despite this defect in promoting cer recombination, the mutants were still able to bind specifically to DNA, and to repress an argA :: lacZ genetic fusion. In this work, both wild type and mutant ArgR proteins were overexpressed as 6-histidine fusion proteins. Crosslinking analysis showed that both wild type and ArgR5aa efficiently formed hexamers in vitro, while ArgR149 formed lower molecular weight multimers. Truncated forms of ArgR that were 150 amino acids or longer, were able to support cer recombination. The substitution mutants between positions 149 to 151 all showed that single amino acid substitutions at this region had little effect on cer recombination. Protein-protein crosslinking experiments showed that wild type and mutant forms of ArgR, were able to interact with and the other accessory protein involved in cer recombination, PepA. In vitro recombination experiments using wild type and mutant forms of ArgR, combined with purified PepA, XerC and XerD showed that the ArgR149 mutant did not support recombination, but the ArgR5aa mutant did promote Holliday junction formation, raising the possibility that these two mutants interact differently with the Xer recombination machinery. Topology experiments showed that after adding topoisomerase 1, PepA is able to protect DNA from topoisomerase 1, and prevent ArgRWt binding to DNA. The two mutants ArgR149 and ArgR5aa are protecting DNA with more supercoiling. When PepA is added, migration profiles with the two mutants showed a binding problem with PepA. Other experiments involving the LEL triplet (leucine-glutamic acid-leucine) and amino-acids around it should be done in order to know the existence of a possible binding site for PepA.

Mon projet de recherche avait pour but de caractériser le rôle de deux protéines, ArgR et PepA, qui agissent en tant que facteurs accessoires de la recombinaison au niveau de deux sites cer du plasmide ColE1 présent dans la bactérie Escherichia coli. Ces deux protéines, couplées aux deux recombinases à tyrosine XerC et XerD, permettent la catalyse de la recombinaison site spécifique au niveau de la séquence cer, convertissant les multimères instables de ColE1 en monomères stables. Cette étude a principalement porté sur la région C-terminale de la protéine ArgR. Cette région de la protéine ArgR possède une séquence en acides-aminés et une structure similaire à celle de la protéine AhrC de Bacillus subtilis. De plus, AhrC, le répresseur de l’arginine de cette bactérie, est capable de complémenter des Escherichia coli mutantes déficientes en ArgR. Les régions C-terminales de ces protéines, montrent une forte similarité. De précédents travaux dans notre laboratoire ont démontré que des mutants d’ArgR comprenant des mutations dans cette région, en particulier les mutants ArgR149, une version tronquée d’ArgR de 149 acides-aminés, et ArgR5aa, une version comprenant une insertion de cinq acides-aminés dans la partie C-terminale, perdaient la capacité de permettre la recombinaison au niveau de deux sites cer présents dans le plasmide pCS210. Malgré cette incapacité à promouvoir la réaction de recombinaison en cer, ces deux mutants étaient toujours capables de se lier spécifiquement à l’ADN et de réprimer une fusion argA :: lacZ. Dans ce travail, les versions mutantes et sauvages d’ArgR furent surexprimées en tant que protéines de fusion 6-histidine. Des analyses crosslinking ont montré que la version sauvage et ArgR5aa pouvaient former des hexamères in-vitro de manière efficace, alors qu’ArgR149 formait des multimères de plus faible poids moléculaire. Des formes tronquées d’ArgR qui comportaient 150 acides-aminés ou plus, étaient encore capables de permettre la recombinaison en cer. Les mutants par substitution ArgRL149A et ArgRL151A ont tous montré que les substitutions d’un seul acide-aminé au sein de cette région avaient peu d’effets sur la recombinaison en cer. Les expériences de crosslinking protéine-à-protéine ont montré que le type sauvage et les formes mutantes d’ArgR étaient capables d’interagir avec la protéine accessoire PepA, également impliquée dans la recombinaison en cer. Les expériences de recombinaison in-vitro utilisant la forme sauvage et les formes mutantes d’ArgR combinées avec les protéines PepA, XerC et XerD purifiées, ont montré que le mutant ArgR149 ne soutenait pas la recombinaison, mais que le mutant ArgR5aa permettait la formation d’une jonction d’Holliday. Des expériences de topologie ont montré que PepA était capable de protéger l’ADN de la topoisomérase 1, et d’empêcher ArgRWt de se lier à l’ADN. Les deux mutants ArgR149 et ArgR5aa protègent aussi l’ADN avec plus de surenroulements. Quand on ajoute PepA, les profils de migration montrent un problème de liaison des deux mutants avec PepA. D’autres expériences impliquant le triplet LEL (leucine-acide glutamique-leucine) et les acides-aminés alentour devraient être réalisés dans le but de connaitre l’existence d’un site de liaison potentiel pour PepA. / My research project involved the role of two proteins, ArgR and PepA, which act as accessory factors in the ColE1 cer recombination system from the gram negative bacteria Escherichia coli. These two proteins, in addition to the tyrosine recombinases XerC and XerD, catalyze a site-specific recombination event at the cer sequence which converts unstable multimeric forms of ColE1 into more stable monomers. Our study mainly focused on the C-terminal end of the ArgR. This region of the ArgR protein possesses a structural and amino acid sequence similarity with the AhrC protein from Bacillus subtilis. Moreover, AhrC, the Arginine repressor of this bacterium, is able to complement Escherichia coli mutants deficient in ArgR. The C-terminal regions of these proteins, display a very high region of similarity. Previous work from our laboratory has shown that ArgR mutants with mutations in this region, especially the mutants ArgR149, a truncated 149 amino acids form of ArgR, and ArgR5aa, a form containing a five amino acid insertion in the C-terminal part, lost the ability to perform a recombination reaction at two cer sites in the plasmid pCS210. Despite this defect in promoting cer recombination, the mutants were still able to bind specifically to DNA, and to repress an argA :: lacZ genetic fusion. In this work, both wild type and mutant ArgR proteins were overexpressed as 6-histidine fusion proteins. Crosslinking analysis showed that both wild type and ArgR5aa efficiently formed hexamers in vitro, while ArgR149 formed lower molecular weight multimers. Truncated forms of ArgR that were 150 amino acids or longer, were able to support cer recombination. The substitution mutants between positions 149 to 151 all showed that single amino acid substitutions at this region had little effect on cer recombination. Protein-protein crosslinking experiments showed that wild type and mutant forms of ArgR, were able to interact with and the other accessory protein involved in cer recombination, PepA. In vitro recombination experiments using wild type and mutant forms of ArgR, combined with purified PepA, XerC and XerD showed that the ArgR149 mutant did not support recombination, but the ArgR5aa mutant did promote Holliday junction formation, raising the possibility that these two mutants interact differently with the Xer recombination machinery. Topology experiments showed that after adding topoisomerase 1, PepA is able to protect DNA from topoisomerase 1, and prevent ArgRWt binding to DNA. The two mutants ArgR149 and ArgR5aa are protecting DNA with more supercoiling. When PepA is added, migration profiles with the two mutants showed a binding problem with PepA. Other experiments involving the LEL triplet (leucine-glutamic acid-leucine) and amino-acids around it should be done in order to know the existence of a possible binding site for PepA.