Most of the chemical reactions of living cells are catalyzed by protein enzymes. These enzymes are very efficient and display a high degree of specificity with respect to the reaction catalyzed. Cellular activities depend critically on the precise three-dimensional structure and function of thousands of enzymes. Many enzymes require binding of metal ions or small organic molecules for their function. The organic molecules that are indispensible components of catalysis by proteins are called coenzymes. Pyridoxal 5ʹ-phosphate (PLP) is a versatile coenzyme found in all living cells. PLP-dependent enzymes play a key role in the function of most of the enzymes catalyzing reactions in the metabolic pathways of amino acid synthesis and degradation. The enzyme pyridoxal kinase serves to make available the co-enzyme PLP to apo-PLP dependent enzymes. Because of their key role in cellular function and their medical importance, the structure and function of PLP-dependent enzymes have been extensively investigated. In the past decade, detailed investigations on the structure and function of several PLP-dependent enzymes have been carried out in our laboratory. The enzymes studied are B. subtilis serinehydroxymethyl transferase (SHMT), S. typhimurium acetylornithine aminotransferase (AcOAT), S. typhimurium and E. coli diaminopropionate ammonia lyase (DAPAL), S. typhimurium D-serine dehydratase (DSD), S. typhimurium D-cysteine desulfhydrase (DCyD) and S. typhimurium arginine decarboxylase (ArgD). The extensive studies conducted on PLP-dependent enzymes in our laboratory during the past decade has not only resulted in deeper understanding of their structure and function but also raised several new questions regarding substrate recognition, reaction specificity, role of active site residues in the catalytic reaction, mechanism of catalysis and potential applications of these enzymes. This thesis is an attempt to answer some of these questions. The thesis also presents the structure and function of a new protein, Salmonella typhimurium pyridoxal kinase, the enzyme that provides PLP for PLP-dependent enzymes. Single crystal X-ray diffraction technique is the most powerful tool currently available for the elucidation of the three-dimensional structures of proteins and other biological macromolecules and for revealing the relationship between their structure and function. X-ray diffraction studies have provided in depth understanding of the topology of secondary structural elements in the three-dimensional structures of proteins, the hierarchical organization of protein domains, structural basis for the substrate specificity of enzymes, intricate details of mechanisms of enzyme catalyzed reactions, allosteric regulation of enzyme activity, mechanisms of feed-back inhibition, structural basis of protein stability, symmetry of oligomeric proteins and their possible biological implications and a myriad of other biochemical and biophysical properties of proteins. The work reported in this thesis is primarily based on X-ray diffraction studies. X-ray crystal structure investigations are complemented by spectral and biochemical studies on the catalyzed reactions. The thesis begins with an introduction to PLP-dependent enzymes and presentation of a brief summary of the earlier work carried out in our laboratory on PLP-dependent enzymes (Chapter 1). A brief description of earlier functional classification of PLP-dependent enzymes and the more recent classification of these enzymes into the four groups based on their three-dimensional structure is provided. Although enzymes belonging to these four structural classes have evolved from independent evolutionary lineages, they share some common features near their active sites and in the mode of PLP binding. Earlier work carried out elsewhere on pyridoxal kinase and its key role in maintaining PLP at a low concentration in the cytosol is presented. Different mechanisms that have been proposed for the transfer of PLP from pyridoxal kinase to other apo PLP-dependent enzymes are briefly described. The experimental procedures and computational methods used during the course of these investigations to obtain the results reported in chapters 3-6 are presented in Chapter 2. Most of these methods are applicable to the isolation of plasmids, cloning, over expression, protein purification, mutant construction, crystallization, X-ray diffraction data collection and processing, structure elucidation and refinement, validation and structural analysis presented in the next three chapters. Various programs and protocols used for data processing, structure determination, refinement, model building, structure validation and analysis are also briefly described. In chapter 3, the role of a number of active site residues in the reaction catalyzed by EcDAPAL, a fold type II PLP-dependent enzyme, the structure of which was determined earlier in the laboratory is explored by mutational, biochemical and structural analyses. Earlier studies had established the probable role of Asp120 and Lys77 in the reaction leading to the breakdown of D-DAP and L-DAP, respectively (Bisht et al., 2012). To further validate the earlier observations, a number of active site mutants were generated for Asp 120 (D120N, D120C, D120S and D120T), Asp 189 (D189N, D189C, D189S and D189T), Lys77 (K77T, K77H, K77R and K77A), His 123 (H123L) and Tyr 168 (Y168F). The structure of D120N mutant crystal obtained after soaking in crystallization cocktail containing D-DAP revealed the presence of an intact external aldimine complex at the active site supporting the earlier proposal that Asp120 is the base abstracting the Cα proton from the D-isomer of DAP. Biochemical and structural observations suggested that none of the Asp189 mutants may bind PLP and were catalytically inactive suggesting an essential role for Asp189 in catalysis. In contrast to type I PLP-dependent enzymes, none of the Lys 77 mutants of EcDAPAL could bind PLP either covalently or non-covalently and were inactive with both the isomers of DAP. Thus, Lys77 appears to be important for both PLP binding and catalysis. H123L mutant formed an external aldimine with D-DAP and a gem-diamine complex with L-DAP indicating that this residue is also crucial for catalysis. These studies have provided additional support to the catalytic mechanism of EcDAPAL proposed earlier. The next Chapter 4 explores the structure, function and catalytic mechanism of Salmonella typhimurium DAPAL (StDAPAL). The protein was purified from a construct carrying a hexa-histidine tag at the C-terminus by Ni-NTA chromatography. The purified protein was demonstrated to be homogeneous by SDS-PAGE and MALDI-TOF. Crystals of StDAPAL belonging to the C-centred monoclinic space group (C121) with four molecules in the asymmetric unit were obtained by the micro batch method and used for collecting X-ray diffracting data. The crystal structure was determined by molecular replacement using the homologous enzyme from E. coli (PDB code 4D9M, Bisht et al., 2012), which shares a sequence identity of 50% with the S. typhimurium enzyme as the phasing model in the program Phaser (McCoy et al., 2007) of the CCP4 suite. The model was refined with Refmac5 of CCP4 suite to R and Rfree values of 25.5% and 30.9%, respectively. A superposition of the structure so obtained over EcDAPAL revealed that the two structures are very similar. A sulfate molecule bound to the active site of StDAPAL could be located. The position of the sulfate corresponds to that of the carboxyl group of aminoacrylate intermediate of EcDAPAL (4D9M). The PLP was bound to Lys78 as an internal aldimine. Since the active sites of the two protomers in fold type II PLP-dependent enzymes are independent, it might be possible to obtain functional monomers of EcDAPAL. With this view, mutation of a conserved Trp (Trp399) present in the dimeric interface resulted in the destabilization of the dimeric interface and partial conversion of the dimeric protein to a monomeric protein. However, the monomeric species of EcDAPALW399R was unable to bind PLP and hence did not possess any catalytic activity. This highlights the importance of dimeric organization for efficient binding of PLP as well as for the activity of the enzyme. A remarkable difference between EcDAPAL and StDAPAL is the absence of a disulfide bond between residues Cys271 and Cys299 in StDAPAL equivalent to the bond formed between Cys265 and Cys291 in EcDAPAL. Mutation of Cys265 and Cys291 of EcDAPAL to Ser did not affect the activity of the enzyme towards either of the isomers of the substrate indicating that the disulfide bond is not crucial for enzyme activity. The stability of the loop corresponding residues 261-295 of EcDAPAL was believed to be promoted by the disulfide bond. However, the equivalent loop was found to be ordered in StDAPAL even though the disulfide bond is absent. In contrast to StDAPAL, EcDAPAL did not show any metal dependent activity. The previous two chapters dealt with fold type II PLP-dependent enzymes. In contrast, Chapter 5 deals with revisiting the structure and function of a fold type I PLP-dependent enzyme, Salmonella typhimurium arginine decarboxylase (StADC). ADC is a very large polypeptide in comparison with other fold type I enzymes. It is induced when the bacterium is subjected to low pH and plays a major role in protecting the cells from acid stress. The structure of StADC was determined but not satisfactorily refined by Dr. S. R. Bharat earlier. The X-ray diffraction data collected by Bharat needed to be improved and the structure needed to be further refined and compared with the homologous E. coli enzyme. Therefore, the entire process of data processing, structure solution and refinement was repeated. The refined structure of StADC was found to correspond to the apo form of the enzyme with only a phosphate molecule occupying the position equivalent to that of 5’ phosphate of PLP observed in EcADC holo enzyme structure. This allowed examination of structural changes that accompany PLP binding and formation of an internal aldimine. The apo to holo transition in StADC involves the movement and ordering of two loops consisting of residues 151-164 and 191-196 which are in the linker and PLP binding domains of the protein, respectively. Phosphate binding by itself appears to be insufficient for these structural changes. These two loops are close to the PLP binding site of the other protomer of the dimer. Hence, these movements are probably important for the catalytic function of the enzyme. Holo ADC has been found as a decamer in other studies. The decameric form of the apo-StADC suggests that PLP binding may not be essential for the oligomeric state of the protein. ADC appears to reduce proton concentration inside the cell in two ways; (i) by surface charge neutralization and (ii) by arginine decarboxylation by extracting a proton from the cytoplasm. The resulting product agmatine is exchanged for extra cellular arginine by arginine-agmatine antiporter. The low sequence identity and lack of structural similarity of the inducible and constitutive forms of ADC from S. typhimurium shows that these are unlikely to be products of divergent evolution. The final chapter 6 of the thesis presents the work carried out on S. typhimurium pyridoxal kinase (PLK). In the salvage pathway of pyridoxal 5’phosphate (PLP), PLP is produced as the product of the reaction catalyzed by PLK using PL, PN and PM as substrates. Thus, PLK plays the critical role of ensuring availability of PLP to the large number of PLP-dependent enzymes. S. typhimurium PLK was purified to homogeneity, crystallized in its native as well as ligand bound forms. It was necessary to circumvent an unusual problem caused by spots arising from a contaminant crystal to obtain the structure of the native crystals of PLK that belonged to the P212121 space group with two protomers in the crystal asymmetric unit. It was then straight forward to determine the ligand bound structures of StPLK (space group P43212) obtained by co-crystallization with ATP, PL and Mg2+ by molecular replacement using the wild type structure as the phasing model. The structures obtained by co-crystallization revealed the presence of ADP, Mg2+ and a PL bound to the active site Lys233 via a Schiff base (internal aldimine). This is the first structure in which the presence of an internal aldimine in the active site of PLK has been observed. Formation of the internal aldimine might be one way to prevent the release of excess PLP and protecting the cell from PLP induced toxicity. The enzyme was shown to be inhibited by the product which will also help in maintaining PLP concentration at low levels. It was also demonstrated that PLK interacts with apo-PLP-dependent enzymes. This observation supports possible direct transfer of PLP from PLK to PLP-dependent enzymes. The thesis ends with an appendix where the work carried out during the course of the thesis work but not as part of the thesis is briefly described.

The thesis describes structural and functional studies of two PLP-dependent enzymes, diaminopropionate (DAP) ammonia lyase (DAPAL) and N-acetylornithine aminotransferase (AcOAT). The main objective of this work was to understand the structural features that control and impart specificity for PLP-dependent catalysis. DAPAL is a prokaryotic enzyme that catalyzes the degradation of D and L forms of DAP to pyruvate and ammonia. The first crystal structure of DAPAL was determined from Escherichia coli (EcDAPAL) in holo and apo forms, and in complex with various ligands. The structure with a transient reaction intermediate (aminoacrylate-PLP azomethine) bound at the active site was obtained from crystals soaked with substrate, DL-DAP. Apo and holo structures revealed that the region around the active site undergoes transition from disordered to ordered state and assumes a conformation suitable for catalysis only upon PLP binding. A novel disulfide was found to occur near a channel that is likely to regulate entry of ligands to the active site. Based on the crystal structures and biochemical studies, as well as studies on active site mutant enzymes, a two base mechanism of catalysis involving Asp120 and Lys77 is suggested. AcOAT is an enzyme of arginine biosynthesis pathway that catalyses the reversible conversion of N-acetylglutamate semialdehyde and glutamate to N-acetyl ornithine and α-ketoglutarate. It belongs to subgroup III of fold type I PLP dependent enzymes. Many clinically important aminotransferases belong to the same subgroup and share many structural similarities. We have carried out extensive comparative analysis of these enzymes to identify the unique features important for substrate specificity. Crystal structures of AcOAT from Salmonella typhimurium were determined in presence of two ligands, canaline and gabaculine, which are known to act as general inhibitors for most of the enzymes of this class. There structures provided important insights into the mode of binding of the substrates. The structures illustrated the switching of conformation of an active site glutamate side chain on binding of the two substrates. In addition to that, structural transitions involving three loop regions near the active site were observed in different ligand bound structures. Kinetics of single turnover fast reactions and multiple turnover steady state reactions indicated that N-AcOAT dimer might follow a mechanism involving sequential half site reactivity for efficient catalysis. The changes observed in loop conformation that resulted in asymmetric forms of the dimer enzyme might form the structural basis for half site reactivity. Single site mutants were designed to understand the significance of these structural transitions and the specific role of active site residues in determining substrate specificity and catalysis. Biochemical characterization of wild type and mutant enzymes by steady state and fast kinetic studies, along with their crystal structures provided detailed insights into subtlety of active site features that manifest substrate specificity and catalytic activity. The thesis also describes the investigations on fold type II enzymes directed towards analyses of polypeptide folds of these enzymes, features of their active sites, nature of interactions between the cofactor and the polypeptide, oligomeric structure, catalytic activities with various ligands, origin of specificity and plausible regulation of activity. Analysis of the available crystal structures of fold type II enzymes revealed five different classes. The dimeric interfaces found in these enzymes vary across the classes and probably have functional significance. Contributions made towards structural and functional studies of three other PLP-dependent enzymes, serine hydoxymethyltransferase (SHMT), D-serine deaminase (DSD) and D-cysteine desulfhydrase (DCyD) are described in an appendix.

Diaminopropionate ammonia lyase (DAPAL) which belongs to the  class of PLP enzymes is reported only from prokaryotes. It is involved in the removal of two amino groups from its substrate, diaminopropionate, to form ammonia and pyruvate. DAPAL from Escherichia coli (eDAPAL) and Salmonella typhimurium (sDAPAL) was cloned, over expressed and purified using either affinity chromatography or conventional procedures. It was observed that eDAPAL (90 units / mg) was comparatively less active than sDAPAL (200 units / mg). Also the enzymes with the N-terminal His tag were found to be many fold less active than the enzymes without tag. DAPAL had a characteristic absorption maximum at 414nm due to the Schiff`s linkage between PLP and the € - amino group of the active site lysine residue. The apoenzyme was prepared by reaction with L-cysteine, and the resulting thiazolidine complex was easily dialyzed. On reconstitution with PLP, complete regain of absorption spectrum and 60% activity was seen. All the three enzymes (apo-, holo and reconstituted), when subjected to gel filtration chromatography were found to be homodimers of 88 kDa. The active site lysine 78 was mutated to glutamine, and the enzyme was purified to homogeneity. In the mutant enzyme PLP continued to be bound at the active site, but in a different orientation with an absorbance maximum at 406nm. The K78Q enzyme had negligible activity as compared to the wild type enzyme confirming the role of K78 in catalysis. Only a few of the enzymes of the  class have been investigated for their unfolding pathways. Urea induced unfolding studies on sDAPAL revealed that at lower concentrations of urea there was a loss in activity due to the disruption of Schiff's linkage. No gross conformational changes were observed at these concentrations of urea as seen from fluorescence and gel filtration experiments. Increase in concentration of urea led to unfolding of the protein thereby causing a shift in fluorescence maximum from 340nm to 357 nm due to the exposure of the buried tryptophans to the less hydrophobic environment. A considerable amount of aggregation was seen at intermediate urea concentrations, which was possibly the reason for the inability of the protein to refold completely. Based on the results, a concerted mechanism for dissociation and unfolding was proposed for sDAPAL. Aminooxy compounds, which are mechanism-based inhibitors for PLP enzymes have been used as drugs against various disorders for the last few decades. In order to probe the mechanism and efficiency with which these compounds inhibit sDAPAL, cycloserine (D and L), methoxyamine (MA) and aminooxyacetic acid (AAA) were chosen for the inhibition studies. The inhibition rates were measured by monitoring decrease in absorbance at 414nm, increase in the range of 320-330nm due to the product formation and loss of activity upon incubation with the inhibitor. It was seen that both the enantiomers of cycloserine were equally effective in disrupting the Schiff’s linkage with the second order rate constants of 15.8 and 36 M -1 sec –1 respectively. Spectral measurements showed two isosbestic points in the case of DCS and one in the case of LCS. Product of this inhibition reaction was identified to be a heat and acid stable compound namely a hydroxyisooxazole derivative of PMP. It was similar in nature to that reported from GABA aminotransferase. These results showed that unlike in the case of alanine racemase, sDAPAL could be inhibited equally well by both the enantiomers. The inhibition studies with the other two inhibitors namely AAA and MA, showed AAA to be more efficient at disrupting the Schiff’s linkage and causing inactivation of the enzyme. The visible absorbance spectrum showed a single isosbestic point in both the cases, indicative of a single step involved in the formation of the final product. The elution profile of the product of the enzymatic as well as non-enzymatic reactions on a C-18 HPLC column was similar and the product was identified to be an oxime. These inhibitors reacted with sDAPAL many fold better than the other PLP dependent enzymes and therefore these compounds can serve as potential drugs for sDAPAL.

Pyridoxal 5’-phosphate (PLP), the active form of vitamin B6, is a cofactor for many enzymes involved in the metabolism of amino acids, amino acid derived metabolites and some amino sugars. PLP is one of the most versatile cofactors and the PLP-dependent enzymes catalyze a variety of reactions including transamination, decarboxylation, inter-conversion of L-and D-amino acids and removal or replacement of chemical groups bound at β or γ carbon of amino acids. The thesis describes the structural studies carried out on three PLP-dependent enzymes; N-acetylornithine aminotransferase (AcOAT), serine hydroxymethyltransferase (SHMT) and diaminopropionate ammonia lyase (DAPAL). Chapter 1 of the thesis begins with a brief introduction to PLP-dependent enzymes and their classification. This is followed by a review of structures of enzymes belonging to the subgroup II aminotransferases. The last section of chapter I contains a detailed description of the structures available till date for SHMT from various sources and the mutational studies carried out on SHMT. All the common experimental procedures and computational methods used for the current investigations are described in chapter II, as most of these are applicable to all structure determinations and analyses. The experimental procedures described include cloning, overexpression, purification, crystallization, and X-ray diffraction data collection. Computational methods include details of various programs used during data processing, structure determination, refinement, model building, structure validation and analysis. AcOAT is one of the key enzymes in arginine and lysine metabolism. AcOAT belongs to the fold type I (αfamily) subgroup II family of PLP dependent enzymes. Both S. typhimurium and E. coli have two genes each, one involved in the biosynthesis of arginine and another in the biodegradation of arginine. Biosynthetic AcOAT catalyzes the conversion of N-acetylglutamate semialdehyde to N-acetylornithine (AcOrn) in the presence of L-glutamate and the conversion of N-succinyl-L-2-amino-6-oxopimelate to N-succinyl-L,L-diaminopimelate in lysine biosynthesis. Meso-DAP and lysine, the products of lysine biosynthesis pathway, are known to function as cross-linking moieties in the peptidoglycan component of bacterial cell wall. Therefore N-acetylornithine aminotransferase could serve as a target for designing antibacterials. Chapter III gives the details of the work carried out on AcOAT. Two genes each from S. typhimurium and E. coli coding for biosynthetic and biodegradative AcOAT were cloned in E. coli, overexpressed and purified by Ni-NTA affinity chromatography. Of the four enzymes, biosynthetic AcOAT from S. typhimurium (sArgD) crystallized in the unliganded form and in the presence of the inhibitor gabaculine or one of the substrates L-glutamate, diffracted to a maximum resolution of 1.90 Å and contained a dimer in the asymmetric unit. The structure was determined by the molecular replacement method using human ornithine aminotransferase (hOAT) as the starting model. The structure of unliganded sAcOAT showed significant electron density for PLP in only one of the subunits (subunit A). The asymmetry in PLP binding could be attributed to the ordering of the loop Lαk-βm in only one subunit. The Km and kcat/Km values determined with the purified sArgD suggested that the enzyme could accept both acetylornithine (AcOrn) and ornithine (Orn) as the substrates and had much higher affinity for AcOrn than for Orn. However, OAT accepts only Orn as the substrate. Comparison of the structurte of sArgD with T. thermophilus AcOAT and hOAT suggested that the higher specificity of sArgD towards AcOrn may not be due to specific differences in the active site residues but could result from minor conformational changes in some of them. sArgD was inhibited by gabaculine with an inhibition constant (Ki) of 7 µM and a second order rate constant (k2) of 0.16 mM-1s-1. The crystal structure of sArgD obtained in the presence of gabaculine and the spectral studies of sArgD with gabaculine suggested that the enzyme might have a low affinity for the PLP-gabaculine complex. Biosynthetic AcOAT from E. coli (eArgD) crystallized in the presence of gabaculine in hanging drop vapor diffusion method and diffracted X-rays only to a resolution of 3.5 Å. Two data sets were collected for the eArgD crystals. One of the data sets belonged to P1 (data 1) and the other to P321 space group (data 2) with a solvent content of ~70%. Data 1 was twinned and the unit cell was unusually large and could accommodate ~24 molecules in the asymmetric unit where as data 2 had four molecules in the asymmetric unit. Biodegradataive AcOAT from E. coli also crystallized in presence of gabaculine in hanging drop vapor diffusion method and suffered from low diffraction quality, where as that from S. typhimurium did not yield crystals. In chapter IV, X-ray crystallographic studies on various site specific mutants of SHMT from Bacillus stereotherophilus (bs) and a detailed comparison of structural data with the biochemical results in relation to mechanism of catalysis are presented. SHMT is a member of the α-class of PLP-dependent enzymes and catalyzes the reversible conversion of L-Ser and THF to glycine and 5,10-methylene THF. 5,10-methylene THF serves as a major source of one-carbon units in the biosynthesis of nucleotides and a few amino acids. SHMT also catalyses the cleavage of β-hydroxy amino acids like L-allo-threonine, transamination, racemization and decarboxylation reactions. SHMT shows increased activity along with enhanced nucleotide synthesis and therefore is a potential target for cancer chemotherapy. The availability of structural and biochemical data on SHMT from different sources ranging from human to E. coli enabled the identification of active site residues and a more critical examination of the role of these residues in the different steps of catalysis. The important mutants studied in the present investigation are E53Q, Y51F, Y61F, Y61A, Y60A, N341A and F351G of bsSHMT. The crystal structures of all these mutants are solved in the presence of various ligands, which gave many interesting results. E53, one of the active residues, interacts with the side chain hydroxyl group of serine bound to PLP in the wild type serine complex and N10 and formyl oxygen in the wild type glycine-FTHF complex. In E53Q glycine and serine complexes, glycine carboxyl and serine side chain were in two conformations, respectively, the new conformation being stabilized by their interaction with the mutated residue Q53. The structure of E53Q-Gly complex obtained in the presence and absence of 5-formyl THF(FTHF) showed an interesting case of enzyme memory in which the final conformational state depends on the way it was obtained and suggested that E53 is crucial for FTHF/THF binding. Though the spectrum showed that FTHF binds to the mutant initially, no density was observed for FTHF in the final structure. FTHF is believed to dissociate from the active site with prolonged incubation leaving behind a few significant conformational changes. Y51, one of the highly conserved tyrosines in SHMT, has hydrogen bonding interactions with the phosphate group of PLP and the active site lysine (K226) in bsSHMT. Mutation of Y51 to F resulted in significant changes at the active site. In all the structures of Y51F complexes, the phosphate group is in two conformations and F51 has moved away from the phosphate and in turn changed the position of Y61, another tyrosine in the active site. The residue Y61 is hydrogen bonded to R357 in the internal aldimine complex of bsSHMT. Addition of glycine/serine to bsSHMT resulted in the conformational change of Y61 away from R357 and towards E53, allowing the added glycine/serine to interact with R357. Mutation of Y61 to A did not bring significant structural changes. Structures of Y51F and Y61A mutants complexed with L-allo-Thr (cleaved to Gly by the wild type enzyme) showed that L-allo-Thr was not cleaved to glycine and acetaldehyde and confirmed the biochemical observation that these two residues are essential even for the THF-independent reaction. Residues Y60 and N341 are also highly conserved residues among SHMTs. Y60 stacks over PABA ring of FTHF in the wild type glycine-FTHF ternary complex. N341 has strong hydrogen bonding interactions with N1 and N8 atoms of the pteridine ring of FTHF. Mutation of either Y60 or N341 to A destroys the binding ability of FTHF/THF to the enzyme according to the biochemical and structural observations. The residue F351 exhibits different conformations in the two subunits of wild type glycine-FTHF ternary complex and is thought to be an important residue in determining the asymmetric binding of FTHF. Mutation of F351 to G did not affect the catalytic activity. Surprisingly, in the crystal structure obtained in the presence of L-allo-Thr, the ligand did not get cleaved to glycine, though in solution, the mutant is as active as the wild type enzyme. Chapter V describes the preliminary structural studies carried out on DAPAL from E. coli and S. typhimurium. DAPAL catalyzes the α, βelimination of both L-and D-diaminopropionate (DAP). DAP is the immediate precursor of two neurotoxins 3oxalyl and 2,3-dioxalyl DAP present in Lathyrus sativus, a grain legume rich in proteins and capable of growing well in drought conditions. The presence of these two neurotoxins precludes its use as a source of protein rich food. This enzyme is present only in bacteria and few species of actinomycetes. Unlike many other PLP-dependent enzymes, DAPAL does not catalyze any side reaction and is the only enzyme known to remove an amino group from the βcarbon of the substrate. The enzymes from E. coli (eDAPAL) and S. typhimurium (sDAPAL) produced diffraction quality crystals. However, crystals of sDAPAL did not survive heavy atom soaking and eDAPAL crystals suffered from poor reproducibility and severe non-isomorphism making it difficult to obtain suitable heavy atom derivatives for structure determination. Production of selenomethionine labelled proteins for these enzymes was initiated and thin crystals were obtained for eDAPAL. Improvement of the quality of these crystals is necessary in order to solve the structure of DAPAL by MAD method.