The Amidohydrolase Superfamily (AHS) of enzymes is one of the most structurally and functionally studied groups of biological catalysts, exquisitely designed to carry out an extensive number of reactions defined by a similar reaction mechanism. There are approximately 11,000 genes coding for AHS proteins from about 2,100 sequenced organisms. Sequence information for these genes has been catalogued in databases, the most instrumental being the National Center for Biotechnology Information (NCBI). Despite the accessible information organized in genomic databases, there is still an extensive problem of reliability in the functional annotation of gene products assigned to the AHS. Proteins in COG3964 of the AHS have been functionally identified as dihydroorotases and adenine deaminases. Eight proteins within three group families of COG3964 have been purified and fail to demonstrate the functionally annotated activity. A library of compounds developed by functional-group modifications was compiled and tested with these enzymes. A group of enzymes within COG3964 demonstrates the ability to hydrolyze stereospecific acetylated alpha-hydroxyl carboxylates. Substrate profiles were constructed for enzymes belonging to group 6 of COG3964. Atu3266, Oant2987 and RHE_PE00295 hydrolyze the R-isomers of a library of alpha-acetyl carboxylates of which acetyl-R-mandelate is the best substrate with catalytic efficiencies of 10^5 M^-1s^-1. This compound was identified after a series of modifications from a low-activity compound (V/K = 4 M^-1s^-1). Methylphosphonate analogs of acetyl-R-mandelate and N-acetyl-D-phenyl glycine are inhibitors of enzymes in group 6. The structure of Atu3266 was used in docking experiments to assess the selectivity of R- enantiomers over their S- counterparts. An additional group of orthologues share less than 40% sequence similarity to enzymes from group 6. EF0837, STM4445 and BCE_5003 from group 2 show significantly lower rates for the hydrolysis of alpha-acetyl carboxylates, including acetyl-R-mandelate, hydrolyzed at values of kcat/Km = 10^3 M^-1s^-1. This is also the only active compound for EF0837. Xaut_0650 and blr3349 from group 7 of COG3964 demonstrate less than 30% identity to enzymes in groups 2 and 6. These enzymes fail to hydrolyze any compound from an extended library of compounds. An annotated selenocysteine synthase gene (SelA) from COG1921 has been identified as a gene neighbor to almost every amidohydrolase from COG3964. Atu3263, Oant2990 and EF0838 are pyridoxal-5’-phosphate dependent enzymes that were purified and assayed with D- and L- amino acids. Initial thermal-shift fluorescence assays determined that in the presence of D-cysteine, the proteins were denatured at lower temperatures.

Philosophiae Doctor - PhD / In bacteria, nitrile hydratases and enzymes of nitrilase and signature amidase superfamilies hydrolyse nitriles and amides to their corresponding carboxylic acids releasing ammonia. Bacteria expressing these enzymes are typically isolated where a sole nitrogen and/or carbon source is used to support their growth. The majority of characterised enzymes of industrial potential have been identified for their stabilities at elevated temperatures. To date, no reports of such enzymes have been isolated from cold adapted bacteria.In this study, an extensive screening program of cold-active microbial isolates for enzymes of this group led to the selection and detailed characterisation of an aliphatic amidase from Nesterenkonia.Nesterenkonia AN1, a new psychrotrophic isolate of the genus, was isolated from soil samples collected from the Miers Valley, Antarctica. AN1 showed significant 16S rRNA sequence identity to known members of the genera, but this is the only strain that had optimal growth at approximately 21oC. AN1, similar to known members, is an obligately alkaliphilic (pH 9-10) and halotolerant (Na+ 0- 15% (w/v)) strain.The genome of Nesterenkonia AN1, sequenced in-house, revealed two ORFs encoding putative nitrilases, referred to as Nit1 and Nit2. Based on analysis of their deduced protein sequences, both belonged to the nitrilase superfamily. Both sequences showed conserved catalytic residues (EKEC), glycine residues and contained the characteristic áââá monomer fold. Homology modelling using known structures suggested that both genes could encode N-carbamoyl D-amino acid amidohydrolases, although neither showed conserved residues implicated in the hydrolysis of carbamoyls.Nit1 and Nit2 were expressed in Escherichia coli BL21 (DE3) pLysS as Cterminal and N-terminal hexahistidine tagged fusion proteins, and purified using Ni-chelation chromatography. Nit1 showed no activity towards nitrile, amide and carbamoyl substrates. This protein, unlike members of the multimeric enzymes of the nitrilase superfamily, was a monomer ~30 kDa protein. It is possible that the C-terminal hexahistidine tag might have prevented Nit1 from forming multimeric proteins.Nit2 showed substrate specificity similar to known aliphatic amidases with a preference for small amides. Nit2 had maximal activity at 30oC and between pH 6.5 and 7.5, properties compatible with its cold-adapted alkaliphilic origins. In addition, the enzyme was irreversibly inactivated at temperatures above 30oC and had a half-life of approximately 7 mins at 60oC. The crystal structure of Nit2 was solved to 1.66 Å. It revealed a ~45.5 kDa dimer, composed of two tightly bound ~30 kDa monomers. These monomers associated along the A surface forming a áââá-áââá sandwich architecture that is conserved in known structures of the nitrilase superfamily.Nit2 is distinct from known aliphatic amidases in both its structure and enzymic activity: the enzyme did not possess an extended C-terminal region; is active in dimeric form; has high affinity for 3C amides rather than 2C amides; and has a low overall catalytic rate. The short C-terminal region of Nit2 may have contributed to the low stability of the enzyme at elevated temperatures. A dendrogram composed of protein sequences of members of the nitrilase superfamily and Nit2 further supported evidence that this aliphatic amidase falls within a distinct group of enzymes.This is the first report of the enzymic characterisation and structural analysis of an aliphatic amidase from a psychrotolerant, alkaliphilic and halotolerant extremophile.

Tracking the evolution of function in enzyme superfamilies is key in understanding how important biological functions and mechanisms have evolved. New genes are being sequenced at a rate that far surpasses the ability of characterization by wet-lab techniques. Moreover, bioinformatics allows for the use of methods not amenable to wet lab experimentation. We now face a situation in which we are aware of the existence of many gene families but are ignorant of what they do and how they function. Even for families with many structurally and functionally characterized members, the prediction of function of ancestral sequences can be used to elucidate past patterns of evolution and highlight likely future trajectories. In this thesis, we apply in silico structure and function methods to predict the functions of protein sequences from two diverse superfamily case studies. In the first, the metallo-β-lactamase superfamily, many members have been structurally and functionally characterised. In this work, we asked how many times the same function has independently evolved in the same superfamily using ancestral sequence reconstruction, homology modelling and alignment to catalytic templates. We found that in only 5% of evolutionary scenarios assessed, was there evidence of a lactam hydrolysing ancestor. This could be taken as strong evidence that metallo-β-lactamase function has evolved independently on multiple occasions. This finding has important implications for predicting the evolution of antibiotic resistance in this protein fold. However, as discussed, the interpretation of this statistic is not clear-cut. In the second case study, we analysed protein sequences of the DUF-62 superfamily. In contrast to the metallo-β-lactmase superfamily, very few members of this superfamily have been structurally and functionally characterised. We used the analysis of alignment, gene context, species tree reconciliation and comparison of the rates of evolution to ask if other functions or cellular roles might exist in this family other than the ones already established. We find that multiple lines of evidence present a compelling case for the evolution of different functions within the Archaea, and propose possible cellular interactions and roles for members of this enzyme family.