Oxygen-Induced Conformational Changes in the PAS-Heme Domain of the Pseudomonas aeruginosa Aer2 Receptor.

The Aer2 receptor from Pseudomonas aeruginosa has an O2-binding PAS-heme domain that stabilizes O2 via a Trp residue in the distal heme pocket. Trp rotates ∼90° to bond with the ligand and initiate signaling. Although the isolated PAS domain is monomeric, both in solution and in a cyanide-bound crystal structure, an unliganded structure forms a dimer. An overlay of the two structures suggests possible signaling motions but also predicts implausible clashes at the dimer interface when the ligand is bound. Moreover, in a full-length Aer2 dimer, PAS is sandwiched between multiple N- and C-terminal HAMP domains, which would feasibly restrict PAS motions. To explore the PAS dimer interface and signal-induced motions in full-length Aer2, we introduced Cys substitutions and used thiol-reactive probes to examine in vivo accessibility and residue proximities under both aerobic and anaerobic conditions. In vivo, PAS dimers were retained in full-length Aer2 in the presence and absence of O2, and the dimer interface was consistent with the isolated PAS dimer structure. O2-mediated changes were also consistent with structural predictions in which the PAS N-terminal caps move apart and the C-terminal DxT region moves closer together. The DxT motif links PAS to the C-terminal HAMP domains and was critical for PAS-HAMP signaling. Removing the N-terminal HAMP domains altered the distal PAS dimer interface and prevented signaling, even after signal-on lesions were introduced into PAS. The N-terminal HAMP domains thus facilitate the O2-dependent shift of PAS to the signal-on conformation, clarifying their role upstream of the PAS-sensing domain.

A heme-binding protein produced by Haemophilus haemolyticus inhibits non-typeable Haemophilus influenzae.

Commensal bacteria serve as an important line of defense against colonisation by opportunisitic pathogens, but the underlying molecular mechanisms remain poorly explored. Here, we show that strains of a commensal bacterium, Haemophilus haemolyticus, make hemophilin, a heme-binding protein that inhibits growth of the opportunistic pathogen, non-typeable Haemophilus influenzae (NTHi) in culture. We purified the NTHi-inhibitory protein from H. haemolyticus and identified the hemophilin gene using proteomics and a gene knockout. An x-ray crystal structure of recombinant hemophilin shows that the protein does not belong to any of the known heme-binding protein folds, suggesting that it evolved independently. Biochemical characterisation shows that heme can be captured in the ferrous or ferric state, and with a variety of small heme-ligands bound, suggesting that hemophilin could function under a range of physiological conditions. Hemophilin knockout bacteria show a limited capacity to utilise free heme for growth. Our data suggest that hemophilin is a hemophore and that inhibition of NTHi occurs by heme starvation, raising the possibility that competition from hemophilin-producing H. haemolyticus could antagonise NTHi colonisation in the respiratory tract.

Coherence-assisted electron diffusion across the multi-heme protein-based bacterial nanowire.

Biological electron transfer (ET) is one of the most studied biochemical processes due to its immense importance in biology. For many years, biological ET was explained using the classical incoherent transport mechanism, i.e. sequential hopping. One of the relatively recent major observations in this field is long-range extracellular ET (EET), where some bacteria were shown to mediate electrons for extremely long distances on the micrometer length scales across individual nanowires. This fascinating finding has resulted in several suggested mechanisms that might explain this intriguing EET. More recently, the structure of a conductive G. sulfurreducens nanowire has been solved, which showed a highly ordered quasi-1D wire of a hexaheme cytochrome protein, named OmcS. Based on this new structure, we suggest here several electron diffusion models for EET, involving either purely hopping or several degrees of mixed hopping and coherent transport, in which the coherent part is due to a local rigidification of the protein structure, associated with a decrease in the local reorganization energy. The effect is demonstrated for two closely packed heme sites as well as for longer chains containing up to several dozens porphyrins. We show that the pure hopping model probably cannot explain the reported conductivity values of the G. sulfurreducens nanowire using conventional values of reorganization energy and electronic coupling. On the other hand, we show that for a wide range of the latter energy values, the mixed hopping-coherent model results in superior electron diffusion compared to the pure hopping model, and especially for long-range coherent transport, involving multiple porphyrin sites.

Structural and Functional Characterization of the Globin-Coupled Sensors of Azotobacter vinelandii and Bordetella pertussis.

Aims: Structural and functional characterization of the globin-coupled sensors (GCSs) from Azotobacter vinelandii (AvGReg) and Bordetella pertussis (BpeGReg). Results: Ultraviolet/visible and resonance Raman spectroscopies confirm the presence in AvGReg and BpeGReg of a globin domain capable of reversible gaseous ligand binding. In AvGReg, an influence of the transmitter domain on the heme proximal region of the globin domain can be seen, and k'CO is higher than for other GCSs. The O2 binding kinetics suggests the presence of an open and a closed conformation. As for BpeGReg, the fully oxygenated AvGReg show a very high diguanylate cyclase activity. The carbon monoxide rebinding to BpeGReg indicates that intra- and intermolecular interactions influence the ligand binding. The globin domains of both proteins (AvGReg globin domain and BpeGRegGb with cysteines (Cys16, 45, 114, 154) mutated to serines [BpeGReg-Gb*]) share the same GCS fold, a similar proximal but a different distal side structure. They homodimerize through a G-H helical bundle as in other GCSs. However, BpeGReg-Gb* shows also a second dimerization mode. Innovation: This article extends our knowledge on the GCS proteins and contributes to a better understanding of the GCSs role in the formation of bacterial biofilms. Conclusions:AvGReg and BpeGReg conform to the GCS family, share a similar overall structure, but they have different properties in terms of the ligand binding. In particular, AvGReg shows an open and a closed conformation that in the latter form will very tightly bind oxygen. BpeGReg has only one closed conformation. In both proteins, it is the fully oxygenated GCS form that catalyzes the production of the second messenger.

A Sec14-like phosphatidylinositol transfer protein paralog defines a novel class of heme-binding proteins.

Yeast Sfh5 is an unusual member of the Sec14-like phosphatidylinositol transfer protein (PITP) family. Whereas PITPs are defined by their abilities to transfer phosphatidylinositol between membranes in vitro, and to stimulate phosphoinositide signaling in vivo, Sfh5 does not exhibit these activities. Rather, Sfh5 is a redox-active penta-coordinate high spin FeIII hemoprotein with an unusual heme-binding arrangement that involves a co-axial tyrosine/histidine coordination strategy and a complex electronic structure connecting the open shell iron d-orbitals with three aromatic ring systems. That Sfh5 is not a PITP is supported by demonstrations that heme is not a readily exchangeable ligand, and that phosphatidylinositol-exchange activity is resuscitated in heme binding-deficient Sfh5 mutants. The collective data identify Sfh5 as the prototype of a new class of fungal hemoproteins, and emphasize the versatility of the Sec14-fold as scaffold for translating the binding of chemically distinct ligands to the control of diverse sets of cellular activities.