Abstract
Nicotinamide adenin dinucleotide (NAD+) is an important molecule as coenzyme in cellular redox reactions and signal-transduction pathways. The enzyme nictotinamide/nicotinate mononucleotide adenylyltransferase (NMNAT/NaMNAT) takes part in the biosynthesis of NAD+ by transferring the adenosine monophosphate moiety of ATP to nicotinamide/nicotinate mononucleotide.
The crystal structure of NMNAT from Homo sapiens in complex with NMN was solved to a maximum resolution of 2.9 Å with the SAD method. The structure of NMNAT consists of a six-stranded parallel b-sheet with helices on both sides, which in the core is the Rossmann fold. Electron density was observed for the ligand NMN but not for a loop of 37 amino acids, residues 109 to 146, that is positionally disordered. The structure of hNMNAT differs from the homologous proteins of M. thermoautotrophicum and M. jannaschii by this loop that contains a nuclear localization signal and by additional amino acids that form the helices H and I and the strand f in the human enzyme. Those secondary structure elements cause a different oligomerization compared to the archaeal proteins but all occur as hexamer as biological unit. Bacterial NMNATs occur as monomer (E. coli) or dimer (B. subtilis) and in the case of E. coli NMNAT has further secondary structure elements.
The ligand NMN is bound by the amino acids Ser16, Lys57, Trp92, Thr95, Leu168 and Trp169, all on one side of the b-sheet. Comparison of the NMN complex with human NMNAT structures solved by other groups (the apoenzyme and the NAD+ complex) and with the archaeal and procarial homologs yield a model for ligand binding and synthesis.
The homing endonuclease PI-SceI of S. cerevisiae is an intein, an internal protein, embedded into the extein sequences of the 69 kDa subunit of the vacuolar membrane H+-ATPase. In a protein splicing process it cuts itself out of the protein precursor. Responsible is domain I of PI-SceI that at the same time accounts for most of the binding energy to the specific DNA sequence of at least 31 bp. Domain II of PI-SceI contains the nucleolytic center, which cuts the specific recognition sequence and initiates the insertion of the vde-gene ("homing").
The crystal structure of domain I of PI-SceI was solved with the molecular replacement method to a maximal resolution of 1.35 Å. The core of PI-SceI domain I adopts the Hint fold of the Hedgehog/intein domain and consists mainly of b-strands. The active center of domain I, the protein splicing site, has cysteine 1 and two N-terminal extein residues but lacks the C-terminus of the intein, Asn454. The common model for the first step of the protein splicing reaction mechanism can be structurally confirmed.
Residues Gln55 to Glu66 were not observed in the electron density due to local disorder. They obviously form a flexible loop that is suspected to take part in the tight binding of DNA by domain I. Further residues that are supposed to participate in DNA binding because of biochemical data, lie at the front side of a tongs-like subdomain that is formed by strands i, j and k and helices B and C. Most likely, helices D and E as well as strand l are also part of this subdomain because they also contain potential DNA-binding residues. A geometry based docking model makes the possibility of a movement of about 60º relative to the core visible.
|
