Discovery of serum biomarkers for diagnosis of tuberculosis by NMR metabolomics including cross-validation with a second cohort.

BACKGROUND: Tuberculosis (TB) is a disease with worldwide presence and a major cause of death in several developing countries. Current diagnostic methodologies often lack specificity and sensitivity, whereas a long time is needed to obtain a conclusive result. METHODS: In an effort to develop better diagnostic methods, this study aimed at the discovery of a biomarker signature for TB diagnosis using a Nuclear Magnetic Resonance based metabolomics approach. In this study, we acquired 1H NMR spectra of blood serum samples of groups of healthy subjects, individuals with latent TB and of patients with pulmonary and extra-pulmonary TB. The resulting data were treated with uni- and multivariate statistical analysis. RESULTS: Six metabolites (inosine, hypoxanthine, mannose, asparagine, aspartate and glutamate) were validated by an independent cohort, all of them related with metabolic processes described as associated with TB infection. CONCLUSION: The findings of the study are according with the WHO Target Product Profile recommendations for a triage test to rule-out active TB.

Active and prospective latent tuberculosis are associated with different metabolomic profiles: clinical potential for the identification of rapid and non-invasive biomarkers.

Although 23% of world population is infected with Mycobacterium tuberculosis (M. tb), only 5-10% manifest the disease. Individuals surely exposed to M. tb that remain asymptomatic are considered potential latent TB (LTB) cases. Such asymptomatic M. tb.-exposed individuals represent a reservoir for active TB cases. Although accurate discrimination and early treatment of patients with active TB and asymptomatic M. tb.-exposed individuals are necessary to control TB, identifying those individuals at risk of developing active TB still remains a tremendous clinical challenge. This study aimed to characterize the differences in the serum metabolic profile specifically associated to active TB infected individuals or to asymptomatic M. tb.-exposed population. Interestingly, significant changes in a specific set of metabolites were shared when comparing either asymptomatic house-hold contacts of active TB patients (HHC-TB) or active TB patients (A-TB) to clinically healthy controls (HC). Furthermore, this analysis revealed statistically significant lower serum levels of aminoacids such as alanine, lysine, glutamate and glutamine, and citrate and choline in patients with A-TB, when compared to HHC-TB. The predictive ability of these metabolic changes was also evaluated. Although further validation in independent cohorts and comparison with other pulmonary infectious diseases will be necessary to assess the clinical potential, this analysis enabled the discrimination between HHC-TB and A-TB patients with an AUC value of 0.904 (confidence interval 0.81-1.00, p-value < 0.0001). Overall, the strategy described in this work could provide a sensitive, specific, and minimally invasive method that could eventually be translated into a clinical tool for TB control.

Metal ions modulate the plastic nature of Mycobacterium tuberculosis chaperonin-10.

Chaperonin-10s possess a highly flexible segment of approximately 10 residues that covers their dome-like structure and closes the central cavity of the chaperonin assembly. The dome loop is believed to contribute to the plasticity of their oligomeric structure. We have exploited the presence of a single tryptophan residue occurring in the dome loop of Mycobacterium tuberculosis chaperonin-10 (cpn-10), and through intrinsic fluorescence measurements show that in the absence of metal ions, the tryptophan is almost fully solvent exposed at neutral pH. The dome loop, however, assumes a closed conformation in the presence of metal ions, or at low pH. These changes are fully reversed in the presence of chelating agents such as EDTA, confirming the role of cations in modulating the metastable states of cpn-10.

Characterization of the Mycobacterium tuberculosis H37Rv alkyl hydroperoxidase AhpC points to the importance of ionic interactions in oligomerization and activity.

An alkyl hydroperoxidase (AhpC) has been found frequently to be overexpressed in isoniazid-resistant strains of Mycobacterium tuberculosis. These strains have an inactivated katG gene encoding a catalase peroxidase, which might render mycobacteria susceptible to the toxic peroxide radicals, thus leading to the concomitant overexpression of the AhpC. Although the overexpressed AhpC in isoniazid-resistant strains of M. tuberculosis may not directly participate in isoniazid action, AhpC might still assist M. tuberculosis in combating oxidative damage in the absence of the catalase. Here we have attempted to characterize the AhpC protein biochemically and report its functional and oligomerization properties. The alkyl hydroperoxidase of M. tuberculosis is unique in many ways compared with its well-characterized homologues from enteric bacteria. We show that AhpC is a decameric protein, composed of five identical dimers held together by ionic interactions. Dimerization of individual subunits takes place through an intersubunit disulphide linkage. The ionic interactions play a significant role in enzymic activity of the AhpC protein. The UV absorption spectrum and three-dimensional model of AhpC suggest that interesting conformational changes may take place during oxidation and reduction of the intersubunit disulphide linkage. In the absence of the partner AhpF subunit in M. tuberculosis, the mycobacterial AhpC might use small-molecule reagents, such as mycothiol, for completing its enzymic cycle.

Homology model of a novel xylanase: molecular basis for high-thermostability and alkaline stability.

Xylanases form enzymes of considerable interest to a variety of biotechnological industries. Their industrial usage is especially attractive since they can replace some of the environmental pollutants, and are economically viable. Those with higher thermostability and optimal activity at alkaline pH are of particular importance to the paper and pulp industry due to the demands of conditions under which the enzymatic reactions are carried out. We have earlier isolated a xylanase from Bacillus sp. NG-27, which is active both at high temperature as well as at alkaline pH. In order to find out factors responsible for the adaptation of this enzyme to the extreme conditions, three dimensional structure of NG-27 xylanase has now been obtained by homology modelling. The tertiary structure shows TIM barrel fold consisting of 8 parallel beta-strands surrounded by alpha-helices. The active site is located at the carboxy terminal end of the TIM barrel. Factors which contribute to the thermostability of the enzyme are increased number of salt bridges. The salt bridges occur remarkably on one face of alpha-helices, with oppositely charged residues occupying i, i+4, i+7 positions. A solvent shielded salt bridge interaction is also observed, which is absent in the mesophilic homologous xylanases. Solvent shielding may enhance electrostatic interaction through lowering of the dielectric, and contribute to increased stability of the enzyme.

Structural characterization of protein-denaturant interactions: crystal structures of hen egg-white lysozyme in complex with DMSO and guanidinium chloride.

A variety of physico-chemical methods employ chemical denaturants to unfold proteins, and study different biophysical processes involved therein. Chemical denaturants are believed to induce unfolding by stabilizing the unfolded state of proteins over the folded state, either macroscopically or through specific interactions. In order to characterize the nature of specific interactions between proteins and denaturants, we have solved crystal structures of hen egg-white lysozyme complexed with denaturants, and report here dimethyl sulfoxide and guanidinium chloride complexes. The dimethyl sulfoxide molecules and guanidinium ions were seen to bind the protein at specific sites and were involved in characteristic interactions. They share a major binding site between them, the C site in the sugar binding cleft of the enzyme. Although the overall conformations of the complexes were very similar to the native structure, spectacular conformational changes were seen to occur locally. Temperature factors were also seen to drop dramatically in the local regions close to the denaturant binding sites. An interesting observation of the present study was the generation of a sodium ion binding site in hen egg-white lysozyme in the presence of denaturants, which was hitherto unknown in any of the other lysozyme structures solved so far. Loss of some of the crucial side chain-main chain interactions may form the initial events in lysozyme unfolding.

Conserved structural features and sequence patterns in the GroES fold family.

An irregular, all beta-class of proteins, comprising members of the chaperonin-10, quinone oxidoreductase, glucose dehydrogenase and alcohol dehydrogenase families has earlier been classified as the GroES fold. In this communication, we present an extensive analysis of sequences and three dimensional structures of proteins belonging to this family. The individual protein structures can be superposed within 1.6 A for more than 60 structurally equivalent residues. The comparisons show a highly conserved hydrophobic core and conservation of a few key residues. A glycyl-aspartate dipeptide is suggested as being critical for the maintenance of the GroES fold. One of the surprising findings of the study is the non-conservative nature of Ile to Leu mutations in the protein core, although Ile to Val mutations are found to occur frequently.

Identification of the INO1 gene of Mycobacterium tuberculosis H37Rv reveals a novel class of inositol-1-phosphate synthase enzyme.

1L-myo-inositol (inositol) is vital for the biogenesis of mycothiol, phosphatidylinositol and glycosylphosphatidylinositol anchors linked to complex carbohydrates in Mycobacterium tuberculosis. All these cellular components are thought to play important roles in host-pathogen interactions and in the survival of the pathogen within the host. However, the inositol biosynthetic pathway in M. tuberculosis is not known. To delineate the pathways for inositol formation, we employed a unique combination of tertiary structure prediction and yeast-based functional assays. Here, we describe the identification of the gene for mycobacterial INO1 that encodes inositol-1-phosphate synthase distinct in many respects from the eukaryotic analogues.

Function of the central domain of streptokinase in substrate plasminogen docking and processing revealed by site-directed mutagenesis.

The possible role of the central beta-domain (residues 151-287) of streptokinase (SK) was probed by site-specifically altering two charged residues at a time to alanines in a region (residues 230-290) previously identified by Peptide Walking to play a key role in plasminogen (PG) activation. These mutants were then screened for altered ability to activate equimolar "partner" human PG, or altered interaction with substrate PG resulting in an overall compromised capability for substrate PG processing. Of the eight initial alanine-linker mutants of SK, one mutant, viz. SK(KK256.257AA) (SK-D1), showed a roughly 20-fold reduction in PG activator activity in comparison to wild-type SK expressed in Escherichia coli (nSK). Five other mutants were as active as nSK, with two [SK(RE248.249AA) and SK(EK281.282AA), referred to as SK(C) and SK(H), respectively] showing specific activities approximately one-half and two-thirds, respectively, that of nSK. Unlike SK(C) and SK(H), however, SK(D1) showed an extended initial delay in the kinetics of PG activation. These features were drastically accentuated when the charges on the two Lys residues at positions 256 and 257 of nSK were reversed, to obtain SK(KK256.257EE) [SK(D2)]. This mutant showed a PG activator activity approximately 10-fold less than that of SK(D1). Remarkably, inclusion of small amounts of human plasmin (PN) in the PG activation reactions of SK(D2) resulted in a dramatic, PN dose-dependent rejuvenation of its PG activation capability, indicating that it required pre-existing PN to form a functional activator since it could not effect active site exposure in partner PG on its own, a conclusion further confirmed by its inability to show a "burst" of p-nitrophenol release in the presence of equimolar human PG and p-nitrophenyl guanidino benzoate. The steady-state kinetic parameters for HPG activation of its 1:1 complex with human PN revealed that although it could form a highly functional activator once "supplied" with a mature active site, the Km for PG was increased nearly eightfold in comparison to that of nSK-PN. SK mutants carrying simultaneous two- and three-site charge-cluster alterations, viz., SK(RE24249AA:EK281.282AA) [SK(CH)], SK(EK272.273AA;EK281.282AA) [SK(FH)], and SK(RE248.249AA;EK272.273AA:EK281.282AA+ ++) [SK(CFH)], showed additive/synergistic influence of multiple charge-cluster mutations on HPG activation when compared to the respective "single-site" mutants, with the "triple-site" mutant [SK(CFH)] showing absolutely no detectable HPG activation ability. Nevertheless, like the other constructs, the double- and triple-charge cluster mutants retained a native like affinity for complexation with partner PG. Their overall structure also, as judged by far-ultraviolet circular dichroism, was closely similar to that of nSK. These results provide the first experimental evidence for a direct assistance by the SK beta-domain in the docking and processing of substrate PG by the activator complex, a facet not readily evident probably because of the flexibility of this domain in the recent X-ray crystal structure of the SK-plasmin light chain complex.

Identification of an ABC transporter gene that exhibits mRNA level overexpression in fluoroquinolone-resistant Mycobacterium smegmatis.

We describe here the PCR amplification of a DNA fragment (mtp1) from Mycobacterium smegmatis using primers derived from consensus sequences of the ABC family of transporters. The fragment encodes amino acid sequences that exhibited significant homology with different ABC transporters. Amino acid sequence alignment of the full length gene with other transporters identified the ABC protein as the B-subunit of the phosphate specific transporter. Strikingly, a M. smegmatis colony which exhibited a high level of ciprofloxacin resistance showed mRNA level overexpression of mtp1. Thus this is the first report in any prokaryote indicating differential expression of an ABC transporter in a fluoroquinolone resistant colony.

Stabilization of human triosephosphate isomerase by improvement of the stability of individual alpha-helices in dimeric as well as monomeric forms of the protein.

Human triosephosphate isomerase (hTIM) is a dimeric enzyme of identical subunits, adopting the alpha/beta-barrel fold. In a previous work, a monomeric mutant of hTIM was engineered in which Met14 and Arg98, two interface residues, were changed to glutamine. Analysis of equilibrium denaturation of this monomeric mutant, named M14Q/R98Q, revealed that its conformational stability, 2.5kcal/mol, is low as compared to the stability of dimeric hTIM (19.3 kcal/mol). The fact that this value is also lower than the conformational stabilities usually found for monomeric proteins suggests that the hTIM monomers are thermodynamically unstable. In the present work, we attempted to stabilize the M14Q/R98Q mutant by introducing stabilizing mutations in alpha-helices of the protein. Five mutations were proposed, designed to increase alpha-helix propensity by introducing alanines at solvent-exposed sites (Q179A, K193A), to introduce favorable interactions with helix dipoles (Q179D, S105D), or to reduce the conformational entropy of unfolding by introducing proline residues at the "N-cap" position of alpha-helices (A215P). Three replacements (Q179D, K193A, and A215P) were found to increase the stability of the native dimeric hTIM and the monomeric M14Q/R98Q. These results suggest that the monomeric hTIM mutant can be stabilized to a considerable extent by following well-established rules for protein stabilization. A comparison of the stabilizing effect performed by the mutations on the dimeric hTIM and the monomeric M14Q/R98Q allowed us to reinforce a model of equilibrium denaturation proposed for both proteins.

Three hTIM mutants that provide new insights on why TIM is a dimer.

Human triosephosphate isomerase (hTIM), a dimeric enzyme, was altered by site-directed mutagenesis in order to determine whether it can be dissociated into monomers. Two hTIM mutants were produced, in which a glutamine residue was substituted for either Met14 or Arg98, both of which are interface residuces. These substitutions strongly interfere with TIM subunit association, since these mutant TIMs appear to exist as compact monomers in dynamic equilibrium with dimers. In kinetic studies, the M14Q mutant exhibits significant catalytic activity, while the R98Q enzyme is inactive. The M14Q enzyme is nevertheless much less active than unmutated hTIM. Moreover, its specific activity is concentration dependent, suggesting a dissociation process in which the monomers are inactive. In order to determine the conformational stability of the wild-type and mutant hTIMs, unfolding of all three enzymes was monitored by circular dichroism and tryptophan fluorescence spectroscopy. In each case, protein stability is concentration dependent, and the unfolding reaction is compatible with a two-state model involving the native dimer and unfolded monomers. The conformational stability of hTIM, as estimated according to this model, is 19.3 (+/-0.4) kcal/mol. The M14Q and R98Q replacements significantly reduce enzyme stability, since the free energies of unfolding are 13.8 and 13.5 (+/- 0.3) kcal/mol respectively, for the mutants, A third mutant, in which the M14Q and R98Q replacements are cumulated, behaves like a monomer. The stability of this mutant is not concentration-dependent, and the unfolding reaction is assigned to a transition from a folded monomer to an unfolded monomer. The conformational stability of this double mutant is estimated 2.5 (+/-0.1) kcal/mol. All these data combined suggest that TIM monomers are thermodynamically unstable. This might explain why TIM occurs only as a dimer.

Structure of the heat shock protein chaperonin-10 of Mycobacterium leprae.

Members of the chaperonin-10 (cpn10) protein family, also called heat shock protein 10 and in Escherichia coli GroES, play an important role in ensuring the proper folding of many proteins. The crystal structure of the Mycobacterium leprae cpn10 (Ml-cpn10) oligomer has been elucidated at a resolution of 3.5 angstroms. The architecture of the Ml-cpn10 heptamer resembles a dome with an oculus in its roof. The inner surface of the dome is hydrophilic and highly charged. A flexible region, known to interact with cpn60, extends from the lower rim of the dome. With the structure of a cpn10 heptamer now revealed and the structure of the E. coli GroEL previously known, models of cpn10:cpn60 and GroEL:GroES complexes are proposed.

Crystal structure of recombinant triosephosphate isomerase from Bacillus stearothermophilus. An analysis of potential thermostability factors in six isomerases with known three-dimensional structures points to the importance of hydrophobic interactions.

The structure of the thermostable triosephosphate isomerase (TIM) from Bacillus stearothermophilus complexed with the competitive inhibitor 2-phosphoglycolate was determined by X-ray crystallography to a resolution of 2.8 A. The structure was solved by molecular replacement using XPLOR. Twofold averaging and solvent flattening was applied to improve the quality of the map. Active sites in both the subunits are occupied by the inhibitor and the flexible loop adopts the "closed" conformation in either subunit. The crystallographic R-factor is 17.6% with good geometry. The two subunits have an RMS deviation of 0.29 A for 248 C alpha atoms and have average temperature factors of 18.9 and 15.9 A2, respectively. In both subunits, the active site Lys 10 adopts an unusual phi, psi combination. A comparison between the six known thermophilic and mesophilic TIM structures was conducted in order to understand the higher stability of B. stearothermophilus TIM. Although the ratio Arg/(Arg+Lys) is higher in B. stearothermophilus TIM, the structure comparisons do not directly correlate this higher ratio to the better stability of the B. stearothermophilus enzyme. A higher number of prolines contributes to the higher stability of B. stearothermophilus TIM. Analysis of the known TIM sequences points out that the replacement of a structurally crucial asparagine by a histidine at the interface of monomers, thus avoiding the risk of deamidation and thereby introducing a negative charge at the interface, may be one of the factors for adaptability at higher temperatures in the TIM family. Analysis of buried cavities and the areas lining these cavities also contributes to the greater thermal stability of the B. stearothermophilus enzyme. However, the most outstanding result of the structure comparisons appears to point to the hydrophobic stabilization of dimer formation by burying the largest amount of hydrophobic surface area in B. stearothermophilus TIM compared to all five other known TIM structures.

Protein crystallography and infectious diseases.

The current rapid growth in the number of known 3-dimensional protein structures is producing a database of structures that is increasingly useful as a starting point for the development of new medically relevant molecules such as drugs, therapeutic proteins, and vaccines. This development is beautifully illustrated in the recent book, Protein structure: New approaches to disease and therapy (Perutz, 1992). There is a great and growing promise for the design of molecules for the treatment or prevention of a wide variety of diseases, an endeavor made possible by the insights derived from the structure and function of crucial proteins from pathogenic organisms and from man. We present here 2 illustrations of structure-based drug design. The first is the prospect of developing antitrypanosomal drugs based on crystallographic, ligand-binding, and molecular modeling studies of glycolytic glycosomal enzymes from Trypanosomatidae. These unicellular organisms are responsible for several tropical diseases, including African and American trypanosomiases, as well as various forms of leishmaniasis. Because the target enzymes are also present in the human host, this project is a pioneering study in selective design. The second illustrative case is the prospect of designing anti-cholera drugs based on detailed analysis of the structure of cholera toxin and the closely related Escherichia coli heat-labile enterotoxin. Such potential drugs can be targeted either at inhibiting the toxin's receptor binding site or at blocking the toxin's intracellular catalytic activity. Study of the Vibrio cholerae and E. coli toxins serves at the same time as an example of a general approach to structure-based vaccine design. These toxins exhibit a remarkable ability to stimulate the mucosal immune system, and early results have suggested that this property can be maintained by engineered fusion proteins based on the native toxin structure. The challenge is thus to incorporate selected epitopes from foreign pathogens into the native framework of the toxin such that crucial features of both the epitope and the toxin are maintained. That is, the modified toxin must continue to evoke a strong mucosal immune response, and this response must be directed against an epitope conformation characteristic of the original pathogen.

Crystal structure of recombinant human triosephosphate isomerase at 2.8 A resolution. Triosephosphate isomerase-related human genetic disorders and comparison with the trypanosomal enzyme.

The crystal structure of recombinant human triosephosphate isomerase (hTIM) has been determined complexed with the transition-state analogue 2-phosphoglycolate at a resolution of 2.8 A. After refinement, the R-factor is 16.7% with good geometry. The asymmetric unit contains 1 complete dimer of 53,000 Da, with only 1 of the subunits binding the inhibitor. The so-called flexible loop, comprising residues 168-174, is in its "closed" conformation in the subunit that binds the inhibitor, and in the "open" conformation in the other subunit. The tips of the loop in these 2 conformations differ up to 7 A in position. The RMS difference between hTIM and the enzyme of Trypanosoma brucei, the causative agent of sleeping sickness, is 1.12 A for 487 C alpha positions with 53% sequence identity. Significant sequence differences between the human and parasite enzymes occur at about 13 A from the phosphate binding site. The chicken and human enzymes have an RMS difference of 0.69 A for 484 equivalent residues and about 90% sequence identity. Complementary mutations ensure a great similarity in the packing of side chains in the core of the beta-barrels of these 2 enzymes. Three point mutations in hTIM have been correlated with severe genetic disorders ranging from hemolytic disorder to neuromuscular impairment. Knowledge of the structure of the human enzyme provides insight into the probable effect of 2 of these mutations, Glu 104 to Asp and Phe 240 to Ile, on the enzyme. The third mutation reported to be responsible for a genetic disorder, Gly 122 to Arg, is however difficult to explain. This residue is far away from both catalytic centers in the dimer, as well as from the dimer interface, and seems unlikely to affect stability or activity. Inspection of the 3-dimensional structure of trypanosomal triosephosphate isomerase, which has a methionine at position 122, only increased the mystery of the effects of the Gly to Arg mutation in the human enzyme.

Crystal structure of peanut lectin, a protein with an unusual quaternary structure.

The x-ray crystal structure of the tetrameric T-antigen-binding lectin from peanut, M(r) 110,000, has been determined by using the multiple isomorphous replacement method and refined to an R value of 0.218 for 22,155 reflections within the 10- to 2.95-A resolution range. Each subunit has essentially the same characteristic tertiary fold that is found in other legume lectins. The structure, however, exhibits an unusual quaternary arrangement of subunits. Unlike other well-characterized tetrameric proteins with identical subunits, peanut lectin has neither 222 (D2) nor fourfold (C4) symmetry. A noncrystallographic twofold axis relates two halves of the molecule. The two monomers in each half are related by a local twofold axis. The mutual disposition of the axes is such that they do not lead to a closed point group. Furthermore, the structure of peanut lectin demonstrates that differences in subunit arrangement in legume lectins could be due to factors intrinsic to the protein molecule and, contrary to earlier suggestions, are not necessarily caused by interactions involving covalently linked sugar. The structure provides a useful framework for exploring the structural basis and the functional implications of the variability in the subunit arrangement in legume lectins despite all of them having nearly the same subunit structure, and also for investigating the general problem of "open" quaternary assembly in oligomeric proteins.

Cloning and overexpression of the triosephosphate isomerase genes from psychrophilic and thermophilic bacteria. Structural comparison of the predicted protein sequences.

We focused on the temperature adaptation of triosephosphate isomerase (TIM; E.C. 5.3.1.1.) by comparing the structure of TIMs isolated from bacterial organisms living in either cold or hot environments. The TIM gene from psychrophilic bacteria Moraxella sp. TA137 was cloned and its nucleotide sequence determined. Its deduced amino acid sequence revealed 34% identity with the thermophilic bacteria Bacillus stearothermophilus TIM. Expression vectors were constructed and recombinant Moraxella TA137 and Bacillus stearothermophilus TIMs were overproduced and purified to homogeneity. Recombinant TIM inactivation constants (Ki), measured at various temperatures, compared to those of the mesophilic Escherichia coli recombinant TIM clearly show that Moraxella TA137 and B. stearothermophilus TIMs have respectively psychrophilic and thermophilic characteristics. To try to elucidate the structure-thermolability and structure-thermostability relationship, factors affecting the overall stability of these two TIMs were examined, based on the alignment with the mesophilic chicken TIM, the three-dimensional structure of which is already known. From this comparison, it appears that the adaptability of TIM to high temperature is favored by better stabilizing residues for the helix dipole as well as better helix-forming residues whereas the adaptability of TIM to low temperature seems to reside in the nature of helix-capping residues.