Experimental characterization of two archaeal inosine 5'-monophosphate cyclohydrolases.

There is variability as to how archaea catalyze the final step of de novo purine biosynthesis to form inosine 5'-monophosphate (IMP) from 5-formamidoimidazole-4-carboxamide ribonucleotide (FAICAR). Although non-archaea almost uniformly use the bifunctional PurH protein, which has an N-terminal IMP cyclohydrolase (PurH2) fused to a C-terminal folate-dependent aminoimidazole-4-carboxamide ribonucleotide (AICAR) formyltransferase (PurH1) domain, a survey of the genomes of archaea reveals use of PurH2 (with or without fusion to PurH1), the "euryarchaeal signature protein" PurO, or an unidentified crenarchaeal IMP cyclohydrolase. In this report, we present the cloning and functional characterization of two representatives of the known IMP cyclohydrolase families. The locus TK0430 in Thermococcus kodakarensis encodes a PurO-type IMP cyclohydrolase with demonstrated activity despite its position in a cluster of apparently redundant biosynthetic genes, the first functional characterization of a PurO from a non-methanogen. Kinetic characterization reveals a Km for FAICAR of 1.56 ± 0.39 µM and a kcat of 0.48 ± 0.04 s-1. The locus AF1811 from Archaeoglobus fulgidus encodes a PurH2-type IMP cyclohydrolase. This Archaeoglobus fulgidus PurH2 has a Km of 7.8 ± 1.8 µM and kcat of 1.32 ± 0.14 s-1, representing the first characterization of an archaeal PurH2 and the first characterization of PurH2 that naturally occurs unfused to an AICAR formyltransferase domain. Each of these two characterized IMP cyclohydrolases converts FAICAR to IMP in vitro, and each cloned gene allows the growth on purine-deficient media of an E. coli purine auxotroph lacking the purH2 gene.

Purine biosynthesis in archaea: variations on a theme.

BACKGROUND: The ability to perform de novo biosynthesis of purines is present in organisms in all three domains of life, reflecting the essentiality of these molecules to life. Although the pathway is quite similar in eukaryotes and bacteria, the archaeal pathway is more variable. A careful manual curation of genes in this pathway demonstrates the value of manual curation in archaea, even in pathways that have been well-studied in other domains. RESULTS: We searched the Integrated Microbial Genome system (IMG) for the 17 distinct genes involved in the 11 steps of de novo purine biosynthesis in 65 sequenced archaea, finding 738 predicted proteins with sequence similarity to known purine biosynthesis enzymes. Each sequence was manually inspected for the presence of active site residues and other residues known or suspected to be required for function.Many apparently purine-biosynthesizing archaea lack evidence for a single enzyme, either glycinamide ribonucleotide formyltransferase or inosine monophosphate cyclohydrolase, suggesting that there are at least two more gene variants in the purine biosynthetic pathway to discover. Variations in domain arrangement of formylglycinamidine ribonucleotide synthetase and substantial problems in aminoimidazole carboxamide ribonucleotide formyltransferase and inosine monophosphate cyclohydrolase assignments were also identified.Manual curation revealed some overly specific annotations in the IMG gene product name, with predicted proteins without essential active site residues assigned product names implying enzymatic activity (21 proteins, 2.8% of proteins inspected) or Enzyme Commission (E. C.) numbers (57 proteins, 7.7%). There were also 57 proteins (7.7%) assigned overly generic names and 78 proteins (10.6%) without E.C. numbers as part of the assigned name when a specific enzyme name and E. C. number were well-justified. CONCLUSIONS: The patchy distribution of purine biosynthetic genes in archaea is consistent with a pathway that has been shaped by horizontal gene transfer, duplication, and gene loss. Our results indicate that manual curation can improve upon automated annotation for a small number of automatically-annotated proteins and can reveal a need to identify further pathway components even in well-studied pathways.

Investigations of the roles of arginine 115 and lysine 120 in the active site of 5,10-methenyltetrahydrofolate synthetase from Mycoplasma pneumoniae.

5,10-Methenyltetrahydrofolate synthetase (MTHFS) catalyzes the conversion of 5-formyltetrahydrofolate to 5,10-methenyltetrahydrofolate coupled to the hydrolysis of ATP. A co-crystal structure of MTHFS bound to its substrates has been published (Chen et al., Proteins 56:839-843, 2005) that provides insights into the mechanism of this reaction. To further investigate this mechanism, we have replaced the arginine at position 115 and the lysine at position 120 with alanine (R115A and K120A, respectively). Circular dichroism spectra for both mutants are consistent with folded proteins. R115A shows no activity, suggesting that R115 plays a critical role in the activity of the enzyme. The K120A mutation increases the Michaelis constant (K(m)) for ATP from 76 to 1,200 microM and the K(m) for 5-formylTHF from 2.5 to 7.1 microM. The weaker binding of substrates by K120A may be due to movement of a loop consisting of residues 117 though 120, which makes several hydrogen bonds to ATP and may be held in position by K120.