Homologous recombination is a fundamental cellular process evolved to maintain genomic integrity and to generate genetic diversity. It plays a crucial role in DNA repair, correct segregation of meiotic chromosomes and resumption of the stalled replication forks. In vitro, the homologous recombination pathway is kinetically separable into a four step process involving initiation, homologous pairing, branch migration and junction resolution. The process of pairing and strand exchange between two homologous double-stranded DNA molecules leads to the formation of an intermediate structure called the Holliday junction (HJ). The crucial enzyme involved in this step in bacteria is RecA. In eubacteria, the junction is processed by three proteins, collectively referred to as the RuvABC protein complex. RuvA binds to the HJ, while RuvB, a helicase, binds to the RuvA-HJ complex and pumps the duplex DNA thus facilitating branch migration. The work reported here is concerned with structural studies on mycobacterial RecA and RuvA. X-ray crystallography was used to solve the protein crystal structures. The hanging drop vapour diffusion method was used for crystallization in all cases. X-ray intensity data were collected on a MAR Research imaging plate mounted on a Rigaku RU200 X-ray generator except for two data sets collected using synchrotron radiation. The data were processed mostly using Mosflm and Scala and few data sets were processed using the HKL program suite. The molecular replacement method using programs Phaser and AMoRe was used for structure solution. Structure refinements were carried out using programs CNS and PHENIX. Model building was performed using COOT and O. PROCHECK, MOLPROBITY, ALIGN and NACCESS were used for structure validation and analysis of the refined structures. Mycobacterium smegmatis RecA (MsRecA) and its nucleotide complexes crystallize in three different, but closely related, forms characterized by specific ranges of unit cell dimensions. The six crystals discussed in the earlier part of the thesis and the five reported earlier, all grown under the same or very similar conditions, belong to these three forms, all in space group P61. They include one obtained by reducing the relative humidity around the crystal. In all crystals, RecA monomers form filaments around a 61 screw axis. Thus, the c-dimension of the crystal corresponds to the pitch of the RecA filament. As reported in the case of E.coli RecA, the variation in the pitch among the three forms correlate well with the motion of the C-terminal domain of the RecA monomers with respect to the main domain. The domain motion is compatible with formation of inactive as well as active RecA filaments involving monomers with a fully ordered C-domain. It does not appear to influence the movement upon nucleotide-binding of the switch residue Gln 196, which is believed to provide the trigger for transmitting the effect of nucleotide-binding to the DNA-binding region. Interestingly, partial dehydration of the crystal results in the movement of the residue, in a way similar to that caused by nucleotide-binding. The ordering of the DNA-binding loops L1 and L2, which present an ensemble of conformations, is also unaffected by domain motion. The conformation of loop L2 appears to depend upon nucleotide-binding presumably on account of the movement of the switch residue which forms part of the loop. The conformations of loops L1 and L2 are correlated and have implications to intermolecular communications within the RecA filament. The structures resulting from different orientations of the C-domain and different conformations of the DNA-binding loops appear to represent snapshots of the RecA molecule at different phases of activity and provide insights into the mechanism of action of RecA. Crystal structures of mutants of MsRecA involving changes of Gln 196 from glutamine to alanine, asparagine and glutamic acid, wild type MsRecA and several of their nucleotide complexes were subsequently determined using mostly low temperature and partly room temperature X-ray data. At both the temperatures, nucleotide binding results in a movement of Gln 196 towards the bound nucleotide in the wild type protein. This movement is abolished in the mutants, thus establishing the structural basis for the triggering action of the residue in terms of the size, shape and the chemical nature of the side chain. The 25 crystal structures reported in this thesis, along with the 5 MsRecA structures reported earlier, provide further elaboration of the relation among the pitch of the `inactive´ RecA filament, the orientation of the C-terminal domain with respect to the main domain and the location of the switch residue. The low temperature structures define one extreme of the range of positions the C-domain can occupy. The movement of the C-domain is correlated to those of the LexA binding loop and the loop that connects the main and the N-terminal domains. These elements of molecular plasticity are made use of in the transition to the `active´ filament, as evidenced by the recently reported structures of RecA-DNA complexes. The available structures of RecA resulting from X-ray and electron microscopic studies appear to represent different stages in the trajectory of the allosteric transformations of the RecA filament. This work contributes to the description of the early stages of this trajectory and provides insights into structures relevant to the later stages. The interesting results observed in the case of MsRecA prompted similar studies on the RecA from Mycobacterium tuberculosis (MtRecA). In this study, the crystals were grown at slightly different conditions and examined at different relative humidities and temperatures. Surprisingly, in spite of the 92% sequence identity between the two proteins, the structures indicated MtRecA to be substantially less plastic than MsRecA. The crystal structures do not provide an obvious explanation for this difference. Further studies are warranted to explain the molecular basis of the difference. RuvA, along with RuvB, is involved in branch migration of heteroduplex DNA in homologous recombination. The structures of four crystal forms of RuvA from Mycobacterium tuberculosis (MtRuvA) have been determined. The RuvB-binding domain is cleaved off in one of them. Detailed models of the complexes of octameric RuvA from different species with the Holliday junction have also been constructed. A thorough examination of the structures determined as part of the doctoral programme and those reported earlier bring to light the hitherto unappreciated role of the RuvB-binding domain in determining inter-domain orientation and oligomerization. These structures also permit an exploration of the interspecies variability of structural features such as oligomerization and the conformation of the loop that carries the acidic pin, in terms of amino acid substitutions. These models emphasize the additional role of the RuvB-binding domain in HJ binding. This role along with its role in oligomerization could have important biological implications. In addition to the work on RecA and RuvA, which forms the body of the thesis, the author was also involved in a structural bioinformatics study in which several carbohydrate binding proteins were probed to identify common minimum principles required for binding mannose, glucose and galactose. The study, presented in an Appendix, identified interactions that were specific to particular sugars, leading to individual fingerprints. These fingerprints were then used for exploring lead compounds, using a fragment based approach. This investigation helped the author to familiarize himself with the analysis of protein structures and ligand design based on them.

Genetic material of bacteria is subject to damage due to multitudinous factors, both extrinsic and intrinsic in origin. Mechanisms for the maintenance of genomic integrity are thus essential for a bacterium to survive. Bacterium also requires appropriate minor changes in the genetic material so as to adapt to the changing environments. Structural and related studies of two proteins from mycobacteria, one involved in recombinational DNA repair (RecA) and the other involved in SOS response which helps in adaptation to stress (LexA) form the subject matter of the thesis. The available literature on structural and related studies on RecA and LexA are reviewed in the introductory chapter. The action of RecA involves transition to an active filament formed in association with DNA and ATP, from an inactive filament in the absence of DNA. The structure of the inactive filament was first established in E. coli RecA (EcRecA). The interaction of RecA with non-hydrolysable ATP analogues and ADP were thoroughly characterised and the DNA binding loops were visualised in this laboratory using the crystal structures involving the proteins from Mycobacterium tuberculosis (MtRecA) and Mycobacterium smegmatis (MsRecA). A switch residue, which triggers the transformation of the information on ATP binding to the DNA binding regions, was identified. The 20 residue C-terminal stretch of RecA, which is disordered in all other relevant crystal structures, was defined in an MsRecA-dATP complex. The ordering of the stretch is accompanied by the generation of a new nucleotide binding site which can communicate with the original nucleotide binding site of an adjacent molecule in the filament. The plasticity of MsRecA and its mutants involving the switch residue was explored by studying crystals grown under different conditions at two different temperatures and, in one instance, at low humidity. The structures of these crystals and those of EcRecA and Deinococcus radiodurans RecA (DrRecA) provide information on correlated movements involving different regions of the molecule. MtRecA has an additional importance as an adjuvant drug target in Mycobacterium tuberculosis. Apart from recombination, another important property of RecA is its coprotease activity whereby it stimulates the inherent cleavage of a certain class of proteins. One of the substrates for the coprotease activity of RecA is LexA. LexA is a transcriptional repressor involved in SOS response in bacteria. LexA performs its function through an autoproteolysis stimulated by RecA, resulting in the derepression of the genes under its control. Structural studies on LexA from E. coli have shown that it has an N-terminal domain involved in binding to DNA and a C-terminal domain involved in catalysis and dimerisation. LexA mediated SOS response in bacteria has been shown in many cases to be responsible for the resistance gained by bacteria on treatment with antibiotics. In that respect, LexA is considered to be a potential drug target in Mycobacterium tuberculosis. Structures of crystals of Mycobacterium tuberculosis RecA, grown and analysed under different conditions and reported in the thesis, provide insights into hitherto underappreciated details of molecular structure and plasticity (Chapter 2). In particular, they yield information on the invariant and variable features of the geometry of the P-loop, whose binding to ATP is central for all the biochemical activities of RecA. The strengths of interaction of the ligands with the P-loop reveal significant differences. This in turn affects the magnitude of the motion of the ‘switch’ residue, Gln195 in M. tuberculosis RecA, which triggers the transmission of ATP-mediated allosteric information to the DNA binding region. M. tuberculosis RecA is substantially rigid compared with its counterparts from M. smegmatis and E. coli, which exhibit concerted internal molecular mobility. Details of the interactions of ligands with the protein, characterised in the structures, could be useful for design of inhibitors against M. tuberculosis RecA. Eleven independent simulations, each involving three consecutive molecules in the RecA filament, carried out on the protein from M. tuberculosis, M. smegmatis and E. coli and their ATP complexes, provide valuable information which is complementary to that obtained from crystal structures, in addition to confirming the robust common structural frame work within which RecA molecules from different eubacteria function (Chapter 3). Functionally important loops, which are largely disordered in crystal structures, appear to adopt in each simulation subsets of conformations from larger ensembles. The simulations indicate the possibility of additional interactions involving the P-loop which remains largely invariant. The phosphate tail of the ATP is firmly anchored on the loop while the nucleoside moiety exhibits substantial structural variability. The most important consequence of ATP binding is the movement of the ‘switch’ residue. The relevant simulations indicate the feasibility of a second nucleotide binding site, but the pathway between adjacent molecules in the filament involving the two nucleotide binding sites appears to be possible only in the mycobacterial proteins. As described in Chapter 4, full length LexA, the N-terminal and C-terminal segments defined by the cleavage site, two point mutants involving changes in active site residues (S160A and K197A) and another involving change at the cleavage site (G126D) were cloned, expressed and purified. The wild type protein cleaves at basic pH. The mutants do not autocleave at basic pH even after incubation for 12 hours. The wild type and the mutant protein dimerise and bind DNA with equal facility. The C-terminal segment also dimerises, but has a tendency to form tetramer as well. The full length proteins including the mutants and the C-terminal segment crystallised. The structure of the crystals obtained for mutant G126D could not be solved. Each of the other crystals, four in number, contained only the catalytic core and a few residues preceding it, indicating that the full length proteins underwent cleavage, at the canonical cleavage site or elsewhere, during the long period involved in the formation of the crystals. Crystals obtained from the solutions of the wild type protein and the C-terminal segment contains dimers of the catalytic core. Crystals obtained using the active site mutants appear to contain different type of tetramers. One of them involves the swapping of the peptide segment preceding the catalytic core. Models of tetramerisation of the full length protein similar to those observed for the catalytic core are feasible. A model of a complex of MtLexA with M. tuberculosis SOS box could be readily built. In this complex, the mutual orientation of the two N-domains of the dimer is different from that in the EcLexA-DNA complex.

Mobile genetic elements are DNA sequences that move around to different positions within one genome or between different genomes. Mobile DNA elements were initially considered as selfish DNA sequences parasitizing the organism’s genome. However, this view has changed with the discovery of several mobile genetic elements which play important evolutionary and functional roles. Such understanding has led to a new connotation for these genetic elements such as drivers or natural molecular tools of genome evolution. Extensive research over the past several years has also led to the identification of several new mobile genetic elements including transposons, segregation distorters, heritable organisms, introns and inteins. Homing endonucleases (HEnases) are a group of rare cutting site-specific doublestranded DNA endonucleases encoded by open reading frames within introns, inteins or free standing genes in all the three forms of life including viruses. These enzymes confer mobility to themselves and their encoding sequences by a gene conversion event termed “homing”. During the homing process, the endonuclease inflicts a double-strand break at or near the homing site of the intein-/intron-less allele, which is subsequently repaired by the host DNA repair machinery resulting in the inheritance of intein/intron. The first homing endonuclease identified was the Saccharomyces cerevisiae mitochondrial genetic marker ‘ω’, which affects the polarity of recombination. This genetic marker, which was later shown to be a mobile group I intron, was present in the mitochondrial 21S rRNA gene and encodes a homing endonuclease. HEnases are distinguished for being able to recognise long DNA sequences (14-40 bp), and display disparate cleavage mechanisms. Unlike restriction endonucleases, these enzymes tolerate sequence polymorphism in their recognition region which provides a mechanism for increasing their genetic diversity. Substantial efforts are underway to explore the possibility of utilizing HEnases as tools for genome mapping, cloning of megabase DNA fragments and gene targeting. HEnases are divided into five sub-families on the basis of their conserved sequence and structural motifs: LAGLIDADG, GIY-YIG, H-N-H, His-Cys box and PD-(D/E)-XK families. Among these, LAGLIDADG family is the largest, most prevalent and well-studied class of HEnases. Homing enzymes that contain a single copy of LAGLIDADG motif per polypeptide chain, such as ICreI, I-MsoI and I-CeuI function as homodimers and recognize and cleave palindromic and pseudo-palindromic DNA sequences. On the other hand, HEnases that harbour two copies of LAGLIDADG motifs including I-AniI, PI-SceI and I-SceI act as monomers and recognize and cleave their DNA target sites with considerable asymmetry. Eubacterial RecA proteins are important for a number of cellular processes such as homologous recombination, DNA repair, restoration of stalled replication forks and SOS response. RecA protein and the process of homologous recombination, which is the main mechanism of genetic exchange, are evolutionarily conserved among a range of organisms. However, few mycobacterial species such as Mycobacterium tuberculosis and Mycobacterium leprae were found to be an exception as they harboured in-frame insertion of an intein-coding sequence in their recA genes. In these organisms, RecA is synthesized as a large precursor, which undergoes protein splicing resulting in the formation of an intein and functionally active RecA protein. The milieu in which RecA precursor undergoes splicing differs substantially between M. tuberculosis and M. leprae. M. leprae RecA precursor (79 kDa) undergoes splicing only in mycobacterial species, whereas M. tuberculosis RecA precursor (85 kDa) is spliced efficiently in Escherichia coli as well. Intriguingly, M. tuberculosis and M. leprae RecA inteins differ greatly in their size, primary sequence and location within the recA gene, thereby suggesting two independent origins during evolution. The occurrence of inteins in the obligate mycobacterial pathogens M. tuberculosis, M. leprae and M. microti, initially suggested that RecA inteins might play a role in pathogenesis or virulence, however this was found to be not the case due to the subsequent identification of these intervening sequences in several non pathogenic mycobacterial strains. Sequence comparison of RecA inteins suggested that they belong to the LAGLIDADG class of homing endonucleases. Accordingly, we have shown earlier that M. tuberculosis RecA intein (PI-MtuI), is a novel LAGLIDADG homing endonuclease, which displays dual target specificity in the presence of alternative cofactors in an ATP-dependent manner. The genome of M. leprae, a gram positive bacillus reveals that in contrast to the genomes of other mycobacterial species, it has undergone extensive deletions and decay and thereby represents an extreme case of reductive evolution. In such a scenario of massive gene decay and function loss in the leprosy bacillus, and dissimilarities in size and primary structures among mycobacterial RecA inteins, it was of interest to examine whether M. leprae recA intervening sequence can encode a catalytically active homing endonuclease. To this end, the intervening sequence corresponding to M. leprae recA intein was PCR amplified, cloned, overexpressed and purified to homogeneity using IMPACT protocol. The identity of the purified RecA intein was ascertained by sequencing 9 amino acid residues at the N-terminal end and Western blot analysis using anti-PI-MleI antibodies. Purified enzyme was found to be devoid of any contaminating exonuclease. Protein crosslinking experiments using glutaraldehyde suggested that PI-MleI exists in solution as a monomer, consistent with double-motif LAGLIDADG enzymes. To test whether the purified PI-MleI can bind to the DNA and display any DNA-binding specificity, we carried out electrophoretic mobility shift assays with both single-stranded and double-stranded cognate DNA. The enzyme displayed robust binding to cognate doublestranded DNA, compared to the cognate single-stranded DNA. DNA binding was further found to be sequence independent though the presence of the cognate sequence was required for maximal binding. The stability and specificity of PI-MleI-cognate DNA complexes were further examined by salt titration and competition experiments, which indicated high stability and specificity. After establishing the stable binding of recombinant PI-MleI to the cognate duplex DNA, we next investigated its endonuclease activity on the cognate plasmid pMLR containing the intein-less recA allele, in the absence or presence of divalent cations. The cleavage was monitored by the conversion of supercoiled pMLR to nicked circular as well as linear duplex DNA. PI-MleI exhibited both single-stranded nicking and double-stranded DNA cleavage activity. PI- MleI exhibits endonuclease activity both in the presence of Mg2+ or Mn2+ through a two step reaction. PI-MleI mediated cleavage though was found to be divalent cation dependent however was nucleotide cofactor independent, unlike PI-MtuI, which cleaves the cognate DNA substrate in the presence of ATP. PI-MleI endonuclease activity was assayed under different conditions and found to display a broad divalent cation, pH and temperature dependence. The kinetic experiments revealed slow turnover rate of PI-MleI suggesting its weak endonuclease activity in contrast to robust cleavage activity displayed by several other known LAGLIDADG homing endonucleases. An intriguing observation emerged from the cleavage site mapping of PI-MleI at singlenucleotide resolution. PI-MleI displayed a staggered double- strand break in the homing site by nicking in the left flanking sequence 44 to 47 bp and in the right flanking sequence 16 to 25 bp, away from the intein insertion site. Similar cleavage patterns have been earlier observed for few GIY-YIG homing endonucleases. To gain further mechanistic insights into the PI-MleI mediated catalysis, we examined the binding of PI-MleI to the cognate DNA by DNase I and (OP)2 Cu footprinting experiments. Both the footprinting approaches revealed interaction of PI-MleI with a region upstream and downstream of its own insertion site, conferring protection to 16 nucleotide residues on the upper and 12 nucleotide residues on the lower strand, respectively. The asymmetric footprints have been earlier observed for some members of LAGLIDADG-type homing endonucleases wherein protection on the complementary strands was found to be out of register by 2 to 3 nucleotides, respectively. In case of PI-MleI, however the footprint formed on the complementary strands of the homing site is non-overlapping, indicating the asymmetric mode of interaction of the enzyme. Surprisingly, PI-MleI footprint was not evident at the cleavage sites and this could be due to the unstable binding of the intein at these regions. To decipher the interaction of PI-MleI at the cleavage sites and to ascertain if these interactions have any functional implications in terms of alterations in base-pairing positioning or strand separation to mediate DNA catalysis, we probed the structure of PI-MleI-DNA complexes with KMnO4. KMnO4 treatment of PI-MleI-cognate DNA complexes revealed the presence of hypersensitive T residues on both the strands at the cleavage sites, but showed no such reactive T residues within the PI-MleI-binding regions. Also, hyper-sensitive T residues were not seen at or near the intein-insertion site or in the region between binding and cleavage sites suggesting that PI-MleI upon binding its cognate DNA induces distortions selectively at the cleavage region. To validate these findings and to test whether such alterations occurred on all substrate DNA molecules or on a small sub-population of target molecules, we used a more sensitive 2-aminopurine fluorescence approach. To this end, six cognate duplex DNA molecules each containing 2-aminopurine (2-AP) at different positions such as at the insertion site, in the DNAbinding region, at or near to the cleavage sites were synthesized to monitor helical distortions in the target DNA. The 2-AP containing cognate DNA duplexes were incubated with increasing concentrations of PI-MleI in the assay buffer and monitored the changes in 2-AP fluorescence intensity in the spectral region from 330 to 450 nm. Out of the 2-AP placed at several positions within the cognate substrate, only the 2-aminopurines at the cleavage site showed enhanced fluorescence with PI-MleI addition, consistent with the hyper-sensitivity of T residues during KMnO4 probing. The findings suggest that DNA distortion might assist PI-MleI in widening the minor groove at the cleavage site and make the scissile phosphates accessible to the enzyme active site similar to what has been seen with other LAGLIDADG homing enzymes. These observations suggest that PI-MleI binds to cognate DNA flanking its insertion site, induces helical distortion at the cleavage sites and generates two staggered double-strand breaks. Together, these finding indicate the modular structure of PI-MleI having separate domains for DNA target recognition and cleavage and a bipartite structure of its homing site. After demonstrating the endonuclease activity of PI-MleI, we next examined the active site residues of PI-MleI involved in double-stranded DNA cleavage, which would further provide insights into its catalytic mechanism. Previously, sequence alignment analyses of LAGLIDADG enzymes carried out using different alignment programs identified the presence of 115VLGSLMGDGP123 sequence as DOD motif I (Block C) and 185LQRAVYLGDG194 or 210VLAIWYMDDG219C sequences as catalytic DOD motif II (Block E) in M. leprae RecA intein (PI-MleI). The bioinformatics analyses though on one hand identified the catalytic motifs in PI-MleI, on the other hand led to conflicting data in regard to the identity and the specific position of the catalytic DOD motif II within the PI-MleI polypeptide. We therefore, performed site-directed mutagenesis of key residues in these catalytic motifs and examined their effect on PI-MleI mediated catalysis. A wealth of mutagenesis and structural data, which exists concerning HEnases, suggests that catalytic centers carry essential aspartate residues, one in each of the LAGLIDADG motifs Accordingly, we chose to mutate conserved aspartates that have been previously implicated in catalysis. By site-directed mutagenesis, we constructed five mutant proteins, in which Asp122 was mutated to alanine, cysteine and threonine; whereas Asp193 and Asp218 were mutated to alanine. The identity of each mutant was ascertained by determining the complete nucleotide sequence of the mutant gene. Mutant proteins were further purified to >95% homogeneity using the purification strategy developed for wild type PI-MleI and were found to be devoid of any contaminating exonuclease. To study the effect of mutations in PI-MleI active site residues on its DNA-binding affinity, we examined the binding characteristics of the wild type PI-MleI and its aspartate variants with the intein-less recA substrate and the stability of protein-DNA complexes. All the mutants displayed similar binding affinity to the cognate DNA as that of the wild type PI-MleI, as judged by the comparison of their binding constants (Kd) which were found to be of the same order. Comparison of salt titration isotherms of wild type PI-MleI and its aspartate variants further revealed the similar salt titration midpoint for most of the mutants as that of wild type enzyme suggesting similar protein-DNA complexes stability. Although these results indicate the occurrence of stable complexes between PI-MleI variants and target DNA, to further define the DNA-binding properties of each mutant protein, wild-type PI-MleI and its variants were assayed by DNase I footprinting. All the mutants (D122A, D122C, D122T, D193A and D218A) showed an asymmetric footprint and protection of ~16 nucleotide residues on the upper and 12 nucleotide residues on the lower strand, respectively, near the intein-insertion site similar to the wild type PI-MleI. Together, these observations suggest that the aspartate substitutions in the catalytic motifs do not alter DNA recognition specificity of PI-MleI or its variants, and may not play a direct role in protein-DNA interactions, again implicating the existence of a modular structure of PI-MleI with distinct DNA-binding and catalytic domains. Wild-type PI-MleI although binds near the intein insertion site, but however was found to induce helical distortions only at the cleavage sites. To explore, if aspartate substitutions have any effect on the structural modifications in target DNA sequence, we carried out 2-aminopurine fluorescence with wild type PI-MleI and its variants. In agreement with the wild type enzyme, all the mutants showed increase in fluorescence with target DNA containing 2-AP only at the cleavage sites, but not at the binding sites. However, quantitative measurements of fluorescence change suggested that D122A and D193A mutants show nearly two-fold decrease in the magnitudes of spectral change at the cleavage site compared to wild type and other variants suggesting their involvement in the helical distortion process. To study the effect of Asp substitutions on the catalytic activity of PI-MleI, we performed cleavage assays using cognate plasmid pMLR DNA, with increasing concentrations of wild-type PI-MleI, or its variants and measured the double-stranded cleavage activity. Whereas, D122A and D193A mutants were completely inactive in double-stranded DNA cleavage under the conditions of the cleavage assay, D218A showed DNA cleavage activity comparable to that of the wild type PI-MleI. Similarly, D122T showed decrease in doublestranded DNA cleavage activity. Interestingly, D122C variant showed ~2-fold enhanced DNA cleavage, compared to the wild-type enzyme.Together, these findings provide compelling evidence to conclude that 115VLGSLMGDGP123 and 185LQRAVYLGDG194 motifs (Blocks C and E, respectively), but not 210VLAIWYMDDG219 motif (Block E), and that residues Asp122 and Asp193 play a direct role with respect to the catalytic mechanism of PI-MleI. In summary, these results suggest that the structural and mechanistic aspects of PI-MleI catalysis are distinct from other well-characterized LAGLIDADG-type homing endonucleases and thus provide further insights into understanding the function and evolution of LAGLIDADG homing enzymes.