No abstract

An acidic surface variant (ASV) of the "truncated" hemoglobin from Thermobifida fusca was designed with the aim of creating a versatile globin scaffold endowed with thermostability and a high level of recombinant expression in its soluble form while keeping the active site unmodified. This engineered protein was obtained by mutating the surface-exposed residues Phe107 and Arg91 to Glu. Molecular dynamics simulations showed that the mutated residues remain solvent-exposed, not affecting the overall protein structure. Thus, the ASV was used in a combinatorial mutagenesis of the distal heme pocket residues in which one, two, or three of the conserved polar residues [TyrB10(54), TyrCD1(67), and TrpG8(119)] were substituted with Phe. Mutants were characterized by infrared and resonance Raman spectroscopy and compared with the wild-type protein. Similar Fe-proximal His stretching frequencies suggest that none of the mutations alters the proximal side of the heme cavity. Two conformers were observed in the spectra of the CO complexes of both wild-type and ASV protein: form 1 with ν(FeC) and ν(CO) at 509 and 1938 cm(-1) and form 2 with ν(FeC) and ν(CO) at 518 and 1920 cm(-1), respectively. Molecular dynamics simulations were performed for the wild-type and ASV forms, as well as for the TyrB10 mutant. The spectroscopic and computational results demonstrate that CO interacts with TrpG8 in form 1 and interacts with both TrpG8 and TyrCD1 in form 2. TyrB10 does not directly interact with the bound CO.

Truncated hemoglobins (trHbs) are heme proteins present in bacteria, unicellular eukaryotes, and higher plants. Their tertiary structure consists in a 2-over-2 helical sandwich, which display typically an inner tunnel/cavity system for ligand migration and/or storage. The microorganism Bacillus subtilis contains a peculiar trHb, which does not show an evident tunnel/cavity system connecting the protein active site with the solvent, and exhibits anyway a very high oxygen association rate. Moreover, resonant Raman results of CO bound protein, showed that a complex hydrogen bond network exists in the distal cavity, making it difficult to assign unambiguously the residues involved in the stabilization of the bound ligand. To understand these experimental results with atomistic detail, we performed classical molecular dynamics simulations of the oxy, carboxy, and deoxy proteins. The free energy profiles for ligand migration suggest that there is a key residue, GlnE11, that presents an alternate conformation, in which a wide ligand migration tunnel is formed, consistently with the kinetic data. This tunnel is topologically related to the one found in group I trHbs. On the other hand, the results for the CO and O(2) bound protein show that GlnE11 is directly involved in the stabilization of the cordinated ligand, playing a similar role as TyrB10 and TrpG8 in other trHbs. Our results not only reconcile the structural data with the kinetic information, but also provide additional insight into the general behaviour of trHbs. Proteins 2010. (c) 2009 Wiley-Liss, Inc.