The interface between microbiology and structural biology as viewed by nuclear magnetic resonance.

Nuclear magnetic resonance (NMR) spectroscopy is one of two principal experimental techniques used in structural biology. It can be used to determine structures at atomic resolution and to investigate the dynamics of macromolecules and intermolecular interactions. We aim to give an overview of the use of modern high resolution NMR methodology in microbiology.

Haemophore-mediated bacterial haem transport: evidence for a common or overlapping site for haem-free and haem-loaded haemophore on its specific outer membrane receptor.

Bacterial extracellular haemophores also named HasA for haem acquisition system form an independent family of haemoproteins that take up haem from host haeme carriers and shuttle it to specific receptors (HasR). Haemophore receptors are required for the haemophore-dependent haem acquisition pathway and alone allow free or haemoglobin-bound haem uptake, but the synergy between the haemophore and its receptor greatly facilitates this uptake. The three-dimensional structure of the Serratia marcescens holo-haemophore (HasASM) has been determined previously and revealed that the haem iron atom is ligated by tyrosine 75 and histidine 32. The phenolate of tyrosine 75 is also tightly hydrogen bonded to the Ndelta atom of histidine 83. Alanine mutagenesis of these three HasASM residues was performed, and haem-binding constants of the wild-type protein, the three single mutant proteins, the three double mutant proteins and the triple mutant protein were compared by absorption spectrometry to probe the roles of H32, Y75 and H83 in haem binding. We show that one axial iron ligand is sufficient to ligate haem efficiently and that H83 may become an alternative iron ligand in the absence of Y75 or both H32 and Y75. All the single mutant proteins retained the ability to stimulate haemophore-dependent haem uptake in vivo. Thus, the residues H32, Y75 and H83 are not individually necessary for haem delivery to the receptor. The binding of haem-free and haem-loaded HasASM proteins to HasRSM-producing strains was studied. Both proteins bind to HasRSM with similar apparent Kd. The double mutant H32A-Y75A competitively inhibits binding to the receptor of both holo-HasASM and apo-HasASM, showing that there is a unique or overlapping site on HasRSM for the apo- and holo-haemophores. Thus, we propose a new mechanism for haem uptake, in which haem is exchanged between haem-loaded haemophores and unloaded haemophores bound to the receptor without swapping of haemophores on the receptor.

Functional aspects of the heme bound hemophore HasA by structural analysis of various crystal forms.

The protein HasA from the Gram negative bacteria Serratia marcescens is the first hemophore to be described at the molecular level. It participates to the shuttling of heme from hemoglobin to the outer membrane receptor HasR, which in turn releases it into the bacterium. HasR alone is also able to take up heme from hemoglobin but synergy with HasA increases the efficiency of the system by a factor of about 100. This iron acquisition system allows the bacteria to survive with hemoglobin as the sole iron source. Here we report the structures of a new crystal form of HasA diffracting up to 1.77A resolution as well as the refined structure of the trigonal crystal form diffracting to 3.2A resolution. The crystal structure of HasA at high resolution shows two possible orientations of the heme within the heme-binding pocket, which probably are functionally involved in the heme-iron acquisition process. The detailed analysis of the three known structures reveals the molecular basis regulating the relative affinity of the heme/hemophore complex.

Evidence for a novel heme-binding protein, HasAh, in Alzheimer disease.

Multiple lines of evidence indicate that oxidative stress is an integral component of the pathogenesis of Alzheimer disease (AD). The precipitating cause of such oxidative stress may be misregulated iron homeostasis because there are profound alterations in heme oxygenase-1 (HO-1), redox-active iron, and iron regulatory proteins. In this regard, HasA, a recently characterized bacterial protein involved in heme acquisition and iron metabolism, may also be important in the generation of reactive oxygen species (ROS) given its ability to bind heme and render iron available for free radical generation through the Fenton reaction. To study further the role of heme binding and iron metabolism in AD, we show an abnormal localization of anti-HasA to the neurofibrillary pathology of AD, but not in normal-appearing neurons in the brains of cases of AD or in age-matched controls. These results suggest the increased presence in AD of a HasA homologue or protein sharing a common epitope with HasA, which we term HasAh. We conclude that heme binding of HasAh is a potential source of free soluble iron and therefore toxic free radicals in AD and in aging. This furthers the evidence that redox-active iron and subsequent Fenton reaction generating reactive oxygen are critical factors in the pathogenesis of AD.

The crystal structure of HasA, a hemophore secreted by Serratia marcescens.

Free iron availability is strongly limited in vertebrate hosts, making the iron acquisition by siderophores inappropriate. Pathogenic bacteria have developed various ways to use the host's iron from iron-containing proteins. Serratia marcescens can use the iron from hemoglobin through the secretion of a hemophore called HasA, which takes up the heme from hemoglobin and shuttles it to the receptor HasR, which in turn, releases heme into the bacterium. We report here the first crystal structure of such a hemophore, bound to a heme group at two different pH values and at a resolution of 1.9 A. The structure reveals a new original fold and suggests a hypothetical mechanism for both heme uptake and release.

NMR studies of the C-terminal secretion signal of the haem-binding protein, HasA.

HasA is a haem-binding protein which is secreted under iron-deficiency conditions by the gram-negative bacterium Serratia marcescens. It is a monomer of 19 kDa (187 residues) able to bind free haem as well as to capture it from haemoglobin. HasA delivers haem to a specific outer-membrane receptor HasR and allows the bacteria to grow in the absence of any other source of iron. It is secreted by a signal peptide-independent pathway which involves a C-terminal secretion signal and an ABC (ATP-binding cassette) transporter. The C-terminal region of the secretion signal containing the essential secretion motif is cleaved during or after the secretion process by proteases secreted by the bacteria. In this work, we study by 1H NMR the conformation of the C-terminal extremity of HasA in the whole protein and that of the isolated secretion signal peptide in a zwitterionic micelle complex that mimicks the membrane environment. We identify a helical region followed by a random-coil C-terminus in the peptide-micelle complex and we show that in both the whole protein and the complex, the last 15 residues containing the motif essential for secretion are highly flexible and unstructured. This flexibility may be a prerequisite to the recognition of HasA by its ABC transporter. We determine the cleavage site of the C-terminal extremity of the protein and analyse the effect of the cleavage on the haem acquisition process.

Conformational and functional properties of an undecapeptide epitope fused with the C-terminal end of the maltose binding protein.

Monoclonal antibody mAb164 is directed against the TrpB2 subunit of the Escherichia coli tryptophan synthase. It recognizes the synthetic peptide P11, constituted of residues 273-283 of TrpB, with high affinity. We constructed a hybrid protein in which the C-terminal end of protein MalE was linked with the N-terminal end of P11. Hybrid MalE-P11 was produced in E. coli from a plasmidic gene and purified in one step as MalE. MalE-P11 and the isolated P11 had identical conformational and functional properties according to the following criteria. The NMR spectra of MalE and MalE-P11 in TOCSY experiments showed that the P11 moiety of MalE-P11 moved independently from its MalE moiety. The chemical shifts of the protons for the P11 moiety of MalE-P11 and for the isolated P11 were very close and did not show significant deviations from random coil values. The equilibrium constant of dissociation (KD) from mAb164, measured by a competition ELISA, was identical for MalE-P11 and the isolated P11, around 6 nM. The change of the C-terminal residue of MalE-P11 from Lys into Ala increased 37-fold this dissociation constant. This increase showed that the P11 moiety of MalE-P11 was not degraded. The high molecular mass of MalE-P11 allowed us to follow its kinetics of interaction with immobilized mAb164 by surface plasmon resonance, using the BIAcore apparatus. The rates of association with mAb164 were similar for MalE-P11 and TrpB2, but the dissociation was faster for MalE-P11 than for TrpB2, as previously observed for the isolated P11 by a fluorometric method. Thus, the fusion of peptides with the C-terminal end of MalE could constitute an alternative to chemical synthesis for the study of their recognition by receptors, in vivo or in vitro.

Purification and characterization of an extracellular heme-binding protein, HasA, involved in heme iron acquisition.

Many bacterial hemoproteins involved in heme acquisition have been isolated recently, comprising outer membrane receptors and extracellular heme-binding protein. The mechanisms by which these proteins extract heme have not been described up to now. One such protein, HasA, which can bind free heme as well as capture it from hemoglobin, is secreted by the Gram-negative bacteria Serratia marcescens under iron deficiency conditions. The fact that HasA does not present sequence similarities with other known hemoproteins suggests that it possesses a new type of heme binding site. This work describes the main physicochemical properties of HasA, essential for understanding its function. HasA is a monomer of 19 kDa that binds one b heme per molecule with high affinity. The electron paramagnetic resonance spectra indicate that the heme iron is in a low-spin ferric state and that the two iron axial ligands are His and His-. The low oxidation-reduction potential value (-550 mV vs standard hydrogen electrode) of the heme bound to HasA suggests that heme could be exposed to the solvent. According to circular dichroism data, the binding of heme does not seem to modify the conformation of HasA.

NMR studies on the flexibility of nucleoside diphosphate kinase.

Human NDP kinase B, product of the nm23-H2 gene, binds DNA. It has been suggested that a helix hairpin on the protein surface, part of the nucleotide substrate binding site, could accommodate DNA binding by swinging away. The presence of flexible regions was therefore investigated by 1H NMR dynamic filtering. Although TOCSY peaks could be assigned to five residues at the N terminus of Dictyostelium NDP kinase, no flexible region was detected in the human enzyme. These data favor the idea that the protein offers different binding sites to mono- and polynucleotides.

NMR studies on the flexibility of the poliovirus C3 linear epitope inserted into different sites of the maltose-binding protein.

A set of permissive positions that tolerate insertions/deletions without major deleterious consequences for the binding activity of the protein was previously identified in the maltose-binding protein. The C3 epitope from poliovirus VP1 protein (93DNPASTTNKDK103) was inserted into eight of these positions and two nonpermissive control sites. NMR studies were performed on the MalE protein, the insertion/deletion mutants, and the C3MalE hybrids to selectively determine the flexible regions in these proteins. Comparison of the C3 epitope mobility in the different hybrid proteins indicates that, whatever its insertion site and independently from the specific sequences of its linkers, the epitope is mostly flexible. The vector protein was shown to unfold partially only in the two C3MalE hybrids that correspond to nonpermissive positions. For one of them (insertion at site 339), both sides of the insert are flexible, and at most one side for all the other hybrids. This result correlates with the antigenicity data on the inserted epitope (Martineau, P., Leclerc, C., and Hofnung, M. (1997) Mol. Immunol, in press.

1.8 A structure of Hypoderma lineatum collagenase: a member of the serine proteinase family.

Collagenase from the fly larvae Hypoderma lineatum cleaves triple-helical collagen in a single region. It was crystallized at neutral pH in the absence of inhibitor and 1.8 A data were collected using synchrotron radiation and a Mark II prototype detector. The structure was solved by combining multiple isomorphous replacement methods and rotation translation function in real space. Refinement between 7 and 1.8 A using the program X-PLOR led to a final R factor of 16.9%. The overall fold is similar to that of other trypsin-like enzymes but the structure differs mainly by the presence of a beta-sheet at position 31-44. The two embedded molecules of the asymmetric unit are related by a pseudo twofold axis. The beta-sheet 31-44 of one molecule is involved in hydrogen bonds with binding-pocket residues of the other molecule. It thus completely prevents access to the active site. The specificity of this enzyme probably results from the position of Phe192 and Tyr99 at the entrance of the active site.

Phosphorylation mechanism of nucleoside diphosphate kinase: 31P-nuclear magnetic resonance studies.

The phosphorylation mechanism of Dictyostelium discoideum nucleoside diphosphate (NDP) kinase was investigated by NMR. 31P chemical shifts were measured on both native and denatured enzyme. In the enzymatically phosphorylated enzyme denatured by 9 M urea or 7 M guanidine hydrochloride, the NDP kinase phosphohistidine signal appeared between the signals of N delta and N epsilon free monophosphohistidines used as reference compounds and added to the sample. A signal with the same intermediate position was also observed in the pronase digest of the alkaline-denatured phosphorylated enzyme. However, when phosphohistidines of the phosphorylated synthetic peptide pGlu-His-Gly were taken as references, the NDP kinase and the N delta peptide phosphohistidine signals were shown to be identical, providing evidence that phosphorylation occurs on the N delta of the active site histidine residue. Moreover, the rate of hydrolysis of the histidine-bound phosphate is in agreement with a modification at the N delta position. Phosphorylation of the NDP kinase by phosphoramidate provided a result similar to that of the enzymatic phosphorylation. In both cases, phosphorylation could not be detected on any amino acid other than histidine. Particularly, no phosphoserine residue was observed.

[Epitopes of the envelope of the hepatitis B virus: a structural approach]. / Les épitopes de l'enveloppe du virus de l'hépatite B: approche-structurale.

Hepatitis B is a major public health problem. More than 300 million people are chronically infected by the virus. During infection very large quantities of complete virions and empty envelopes, consisting of spherical or filamentous lipoprotein particles, are present in the blood. DNA genome coding for envelopes is divided into three domains, preS1, preS2 and S. All available data suggest that the preS1 and preS2 products are exposed at the surface of the virions. These proteins are more immunogenic than S in terms of in vivo antibody response and the number of epitopes identified. The three dimensional mapping of antigenic sites of the HBV will provide important strategic information for vaccine development and identification of targets for immunorecognition or immunoregulation of the disease.

Subgroup assignment of a human monoclonal anti-Rh(D) antibody.

Variable subgroups of both chains of a human monoclonal anti-RH(D) IgG1 (kappa), QA37C3G6, were determined from their N-terminal sequences. Sequence comparison with corresponding chains of other human antibodies indicated that the light chain belongs to the third subgroup of human kappa-light chains while the heavy chain belongs to the second subgroup of human heavy chains.

Partial amino acid sequence of the light chain of human anti-Rh(D) monoclonal antibody H2D5D2F5.

Antibody H2D5D2F5 is a human monoclonal anti-Rh(D) IgG1 (lambda) produced by Epstein-Barr virus-immortalized B lymphocytes from a healthy donor. The complete amino acid sequence of the light (L) chain, with the exception of positions 94-97, was determined by Edman degradation of the intact chain, containing 30 residues, and derived tryptic and thermolytic peptides. Sequences of the peptides were aligned by comparison with the sequences of previously reported L chains. H2D5D2F5 L chain belongs to the first variable subgroup of human chains. Its sequence does not reveal striking differences when compared to those of other human lambda chains issued from myeloma or hybridoma.

Purification and characterization of an extracellular 29-kilodalton phospholipase C from Listeria monocytogenes.

We purified and characterized an extracellular phospholipase produced by Listeria monocytogenes. This enzyme was separated as a homogeneous protein of 29 kDa by chromatography on DEAE-52 cellulose and Bio-Gel P100 columns. It is a zinc-dependent phospholipase C (PLC) that is mainly active at pH 6 to 7 and expresses lecithinase activity and a weaker sphingomyelinase activity. The exoenzyme also hydrolyzed phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and sphingomyelin but not phosphatidylinositol. It was distinct from the 36-kDa phosphatidylinositol PLC produced by L. monocytogenes and from the L. ivanovii sphingomyelinase. The pure protein expressed a weak, calcium-independent hemolytic activity and was not toxic in mice. Western immunoblot analysis using a rabbit immune serum raised against the enzyme showed that all virulent strains of L. monocytogenes tested produced in the culture supernatant a 29-kDa PLC. In contrast, no proteins antigenically related to the 29-kDa PLC were detected in supernatants of L. ivanovii, L. seeligeri, L. innocua, or L. welshimeri. The role in virulence of the 29-kDa PLC specifically produced by L. monocytogenes remains to be established.

[Model of protease I from the crab Uca pugilator]. / Modélisation de la protéase I extraite du crabe Uca pugilator.

Collagenolytic protease I from the fiddler crab Uca pugilator is one of the serine proteases of the trypsin family. A graphic molecular model was built on the basis of the sequences and crystalline structures of four homologous proteins which were superimposed in order to identify structurally conserved regions. The sequence of protease I was matched to sequences of the reference proteins, without allowing any deletions or insertions in these regions. For structurally variable regions, the most similar sequences of the four reference proteins were selected. Intramolecular steric clumping due to replacement of reference side-chains by protease I side-chains were corrected by adjusting side-chain conformations. The model was optimized by energy minimization. The conformation of the primary specificity pocket for protease I predicted by the model indicated a preference for P1 hydrophobic or positively charged substrates. This prediction is consistent with biochemical findings. Because soya bean trypsin inhibitor (STI) has been shown to inhibit protease I, a tentative model of the complex was constructed and possible protease I-STI interactions were analyzed.

Role of the gut proteinases from mosquito larvae in the mechanism of action and the specificity of the Bacillus sphaericus toxin.

Gut proteinases from larvae of mosquito species both susceptible and not susceptible to Bacillus sphaericus converted the 43-kDa toxin to a 40-kDa polypeptide exhibiting enhanced cytotoxicity to mosquito cell cultures. The toxin was also activated by gut proteinases from the nonsusceptible Lepidoptera Spodoptera littoralis in vitro and in vivo. Therefore, the specificity of Bacillus sphaericus toxin does not seem to be determined by gut proteinase action. However, susceptibility of mosquito cell cultures did not reflect the specificity of the toxin, which must now be investigated at the cellular level in the larvae.