A hierarchical Bayesian model for predicting the functional consequences of amino-acid polymorphisms

Genetic polymorphisms in deoxyribonucleic acid coding regions may have a phenotypic effect on the carrier, e.g. by influencing susceptibility to disease. Detection of deleterious mutations via association studies is hampered by the large number of candidate sites; therefore methods are needed to narrow down the search to the most promising sites. For this, a possible approach is to use structural and sequence-based information of the encoded protein to predict whether a mutation at a particular site is likely to disrupt the functionality of the protein itself. We propose a hierarchical Bayesian multivariate adaptive regression spline (BMARS) model for supervised learning in this context and assess its predictive performance by using data from mutagenesis experiments on lac repressor and lysozyme proteins. In these experiments, about 12 amino-acid substitutions were performed at each native amino-acid position and the effect on protein functionality was assessed. The training data thus consist of repeated observations at each position, which the hierarchical framework is needed to account for. The model is trained on the lac repressor data and tested on the lysozyme mutations and vice versa. In particular, we show that the hierarchical BMARS model, by allowing for the clustered nature of the data, yields lower out-of-sample misclassification rates compared with both a BMARS and a frequen-tist MARS model, a support vector machine classifier and an optimally pruned classification tree.

X-ray scattering study of the effect of hydration on the cross-beta structure of amyloid fibrils

We have investigated the effect of sample hydration on the wide-angle X-ray scattering patterns of amyloid fibrils from two different sources, hen egg white lysozyme (HEWL) and an 11-residue peptide taken from the sequence of transthyretin (TTR105-115). Both samples show an inter-strand reflection at 4.7 Å and an inter-sheet reflection which occurs at 8.8 and 10 Å for TTR105-115 and HEWL fibrils, respectively. The positions, widths, and relative intensities of these reflections are conserved in patterns obtained from dried stalks and hydrated samples over a range of fibril concentrations. In 2D scattering patterns obtained from flow-aligned hydrated samples, the inter-strand and inter-sheet reflections showed, respectively, axial and equatorial alignment relative to the fibril axis, characteristic of the cross-β structure. Our results show that the cross-β structure of the fibrils is not a product of the dehydrating conditions typically employed to produce aligned samples, but is conserved in individual fibrils in hydrated samples under dilute conditions comparable to those associated with other biophysical and spectroscopic techniques. This suggests a structure consisting of a stack of two or more sheets whose interfaces are inaccessible to bulk water.

Interaction of sulphur-containing aroma compounds with proteins in both model systems and real food systems

Irreversible binding of key flavour disulphides to ovalbumin has been shown previously to occur in model systems. The extent of binding is determined by the availability of the sulphydryl groups to participate in disulphide exchange, influenced either by pH, or the state of the protein (native or heat-denatured). In this study, two further proteins, one with sulphydryl groups available in the native state (beta-lactoglobulin) and one with no sulphydryl groups in the native state (lysozyme) were used to confirm this hypothesis. When the investigation was extended to real food systems, a similar effect was shown when a commercial meat flavouring containing disulphides was added to heat-denatured ovalbumin. Furthermore, comparison of the volatiles generated from onions, cooked either alone, or in the presence of meat, showed a significant reduction of key onion-derived disulphides when cooked in the presence of meat, and an even greater reduction of trisulphides. These findings may have implications for consumer acceptance of food products; where these compounds are used as flavourings or where they occur naturally.

The adsorbed conformation of globular proteins at the air/water interface

External reflection FTIR spectroscopy and surface pressure measurements were used to compare conformational changes in the adsorbed structures of three globular proteins at the air/water interface. Of the three proteins studied, lysozyme, bovine serum albumin and P-lactoglobulin, lysozyme was unique in its behaviour. Lysozyme adsorption was slow, taking approximately 2.5 h to reach a surface pressure plateau (from a 0.07 mM solution), and led to significant structural change. The FTIR spectra revealed that lysozyme formed a highly networked adsorbed layer of unfolded protein with high antiparallel beta-sheet content and that these changes occurred rapidly (within 10 min). This non-native secondary structure is analogous to that of a 3D heat-set protein gel, suggesting that the adsorbed protein formed a highly networked interfacial layer. Albumin and P-lactoglobulin adsorbed rapidly (reaching a plateau within 10 min) and with little chance to their native secondary structure.

Protein-lipid interactions at the air/water interface

Surface pressure measurements and external reflection FTIR spectroscopy have been used to probe protein-lipid interactions at the air/water interface. Spread monomolecular layers of stearic acid and phosphocholine were prepared and held at different compressed phase states prior to the introduction of protein to the buffered subphase. Contrasting interfacial behaviour of the proteins, albumin and lysozyme, was observed and revealed the role of both electrostatic and hydrophobic interactions in protein adsorption. The rate of adsorption of lysozyme to the air/water interface increased dramatically in the presence of stearic acid, due to strong electrostatic interactions between the negatively charged stearic acid head group and lysozyme, whose net charge at pH 7 is positive. Introduction of albumin to the subphase resulted in solubilisation of the stearic acid via the formation of an albumin-stearic acid complex and subsequent adsorption of albumin. This observation held for both human and bovine serum albumin. Protein adsorption to a PC layer held at low surface pressure revealed adsorption rates similar to adsorption to the bare air/water interface and suggested very little interaction between the protein and the lipid. For PC layers in their compressed phase state some adsorption of protein occurred after long adsorption times. Structural changes of both lysozyme and albumin were observed during adsorption, but these were dramatically reduced in the presence of a lipid layer compared to that of adsorption to the pure air/water interface.

Infrared linear dichroism spectroscopy on amyloid fibrils aligned by molecular combing

We report the use of molecular combing as an alignment method to obtain macroscopically oriented amyloid fibrils on planar surfaces. The aligned fibrils are studied by polarized infrared spectroscopy. This gives structural information that cannot be definitively obtained from standard infrared experiments on isotropic samples, for example, confirmation of the characteristic cross-beta amyloid core structure, the side-chain orientation from specific amino acids, and the arrangement of the strands within the fibrils, as we demonstrate here. We employed amyloid fibrils from hen egg white lysozyme (HEWL) and from a model octapeptide. Our results demonstrate molecular combing as a straightforward method to align amyloid fibrils, producing highly anisotropic infrared linear dichroism (IRLD) spectra.

Measurement of intrinsic properties of amyloid fibrils by the peak force QNM method

We report the investigation of the mechanical properties of different types of amyloid fibrils by the peak force quantitative nanomechanical (PF-QNM) technique. We demonstrate that this technique correctly measures the Young’s modulus independent of the polymorphic state and the cross-sectional structural details of the fibrils, and we show that values for amyloid fibrils assembled from heptapeptides, a-synuclein, Ab(1–42), insulin, b-lactoglobulin,lysozyme, ovalbumin, Tau protein and bovine serum albumin all fall in the range of 2–4 GPa.

An optimized method for the extraction of bacterial mRNA from plant roots infected with Escherichia coli O157:H7

Analysis of microbial gene expression during host colonization provides valuable information on the nature of interaction, beneficial or pathogenic, and the adaptive processes involved. Isolation of bacterial mRNA for in planta analysis can be challenging where host nucleic acid may dominate the preparation, or inhibitory compounds affect downstream analysis, e.g., quantitative reverse transcriptase PCR (qPCR), microarray, or RNA-seq. The goal of this work was to optimize the isolation of bacterial mRNA of food-borne pathogens from living plants. Reported methods for recovery of phytopathogen-infected plant material, using hot phenol extraction and high concentration of bacterial inoculation or large amounts of infected tissues, were found to be inappropriate for plant roots inoculated with Escherichia coli O157:H7. The bacterial RNA yields were too low and increased plant material resulted in a dominance of plant RNA in the sample. To improve the yield of bacterial RNA and reduce the number of plants required, an optimized method was developed which combines bead beating with directed bacterial lysis using SDS and lysozyme. Inhibitory plant compounds, such as phenolics and polysaccharides, were counteracted with the addition of high-molecular-weight polyethylene glycol and hexadecyltrimethyl ammonium bromide. The new method increased the total yield of bacterial mRNA substantially and allowed assessment of gene expression by qPCR. This method can be applied to other bacterial species associated with plant roots, and also in the wider context of food safety.

Protein precipitation behaviour of condensed tannins from Lotus pedunculatus and Trifolium repens with different mean degrees of polymerization

The precipitation of bovine serum albumin (BSA), lysozyme (LYS) and alfalfa leaf protein (ALF) by two large- and two medium-sized condensed tannin (CT) fractions of similar flavan-3-ol subunit composition is described. CT fractions isolated from white clover flowers and big trefoil leaves exhibited high purity profiles by 1D/2D NMR and purities >90% (determined by thiolysis). At pH 6.5, large CTs with a mean degree of polymerization (mDP) of ~18 exhibited similar protein precipitation behaviors and were significantly more effective than medium CTs (mDP ~9). Medium CTs exhibited similar capacities to precipitate ALF or BSA, but showed small but significant differences in their capacity to precipitate LYS. All CTs precipitated ALF more effectively than BSA or LYS. Aggregation of CT-protein complexes likely aided precipitation of ALF and BSA, but not LYS. This study, one of the first to use CTs of confirmed high purity, demonstrates that mDP of CTs influences protein precipitation efficacy.

Sequence and function of lysosomal and digestive cathepsin D-like proteinases of Musca domestica midgut

Musca domestica larvae display in anterior and middle midgut contents, a proteolytic activity with pH optimum of 3.0-3.5 and kinetic properties like cathepsin D. Three cDNAs coding for preprocathepsin D-like proteinases (ppCAD 1, ppCAD 2, ppCAD 3) were cloned from a M. domestica midgut cDNA library. The coded protein sequences included the signal peptide, propeptide and mature enzyme that has all conserved catalytic and substrate binding residues found in bovine lysosomal cathepsin D. Nevertheless, ppCAD 2 and ppCAD 3 lack the characteristic proline loop and glycosylation sites. A comparison among the sequences of cathepsin D-like enzymes from some vertebrates and those found in M. domestica and in the genomes of Aedes aegypti, Drosophila melanogaster, Tribolium castaneum, and Bombyx mori showed that only flies have enzymes lacking the proline loop (as defined by the motif: DxPxPx(G/A)P), thus resembling vertebrate pepsin. ppCAD 3 should correspond to the digestive cathepsin D-like proteinase (CAD) found in enzyme assays because: (1) it seems to be the most expressed CAD, based on the frequency of ESTs found. (2) The mRNA for CAD 3 is expressed only in the anterior and proximal middle midgut. (3) Recombinant procathepsin D-like proteinase (pCAD 3), after auto-activation has a pH optimum of 2.5-3.0 that is close to the luminal pH of M. domestica midgut. (4) Immunoblots of proteins from different tissues revealed with anti-pCAD 3 serum were positive only in samples of anterior and middle midgut tissue and contents. (5) CAD 3 is localized with immunogold inside secretory vesicles and around microvilli in anterior and middle midguit cells. The data support the view that on adapting to deal with a bacteria-rich food in an acid midgut region, M. domestica digestive CAD resulted from the same archetypical gene as the intracellular cathepsin D, paralleling what happened with vertebrates. The lack of the proline loop may be somehow associated with the extracellular role of both pepsin and digestive CAD 3. (C) 2009 Elsevier Ltd. All rights reserved.

Role of the triad N46, S106 and T107 and the surface charges in the determination of the acidic pH optimum of digestive lysozymes from Musca domestica

Structures of digestive lysozymes 1 and 2 from housefly (MdL1 and MdL2) show that S106-T107 delimit a polar pocket around E32 (catalytic acid/base) and N46 contributes to the positioning of 050 (catalytic nucleophile), whereas those residues are replaced by V109-A110 and D48 in the non-digestive lysozyme from hen egg-white (HEWL). Further analyses revealed that MdL1 and MdL2 surfaces are less positively charged than HEWL surface. To verify the relevance of these differences to the acidic pH optimum of digestive lysozymes it was determined that pKas of the catalytic residues of the triple mutant MdL2 (N46D-S106V-T107A) are similar to HEWL pKas and higher than those for MdL2. In agreement, triple mutant MdL2 and HEWL exhibits the same pH optimum upon methylumbelliferylchitotrioside. In addition to that, the introduction of six basic residues on MdL1 surface increased by 1 unit the pH optimum for the activity upon bacterial walls. Thus, the acidic pH optimum for MdL2 and MdL1 activities upon methylumbelliferylchitotrioside is determined by the presence of N46, S106 and T107 in the environment of their catalytic residues, which favors pKas reduction. Conversely, acidic pH optimum upon bacterial walls is determined by a low concentration of positive charges on the MdL2 and MdL1 surfaces. (C) 2010 Elsevier Inc. All rights reserved.

Effect of solvent on the adsorption behavior and on the surface properties of Sterculia striata polysaccharide

Characterization of Sterculia striate polysaccharide (SSP) films adsorbed onto Si wafers from solutions prepared in ethyl methyl imidazolium acetate (EmimAc), water or NaOH 0.01 mol/L was systematically studied by means of ellipsometry, atomic force microscopy and contact angle measurements. SSP adsorbed from EmimAc onto Si wafer as homogeneous monolayers (similar to 0.5 nm thick), while from water or NaOH 0.01 mol/L SSP formed layers of similar to 4.0 nm and similar to 1.5 nm thick, respectively. Surface energy values found for SSP adsorbed from EmimAc or water were 68 +/- 2 mJ/m(2) and 65 +/- 2 mJ/m(2), respectively, whereas from NaOH it amounted to 57 +/- 3 mJ/m(2). The immobilization of lysozyme (LYS) onto SSP films was also investigated. The mean thickness of LYS (d(LYS)) immobilized onto SSP films adsorbed from each solvent tended to increase with the decrease of gamma(P)(S) and gamma(total)(S). However, the enzymatic activity of LYS molecules was higher when they were immobilized onto SSP films with higher gamma(P)(S) and gamma(total)(S) values. (C) 2010 Elsevier Ltd. All rights reserved.

Species-specific cell-matrix interactions are essential for differentiation of alveoli like structures and milk gene expression in primary mammary cells of the Cape fur seal (Arctocephalus pusillus pusillus)

Few models are in place for analysis of extreme lactation patterns such as that of the fur seals which are capable of extended down regulation of milk production in the absence of involution. During a 10–12 month lactation period, female fur seals suckle pups on shore for 2–3 days, and then undertake long foraging trips at sea for up to 28 days, resulting in the longest intersuckling bouts recorded. During this time the mammary gland down regulates milk production. We have induced Cape fur seal (Arctocephalus pusillus pusillus) mammary cells in vitro to form mammospheres up to 900 μm in diameter, larger than any of their mammalian counterparts. Mammosphere lumens were shown to form via apoptosis and cells comprising the cellular boundary stained vimentin positive. The Cape fur seal GAPDH gene was cloned and used in RT-PCR as a normalization tool to examine comparative expression of milk protein genes (αS2-casein, β-lactoglobulin and lysozyme C) which were prolactin responsive. Cape fur seal mammary cells were found to be unique; they did not require Matrigel for rapid mammosphere formation and instead deposited their own matrix within 2 days of culture. When grown on Matrigel, cells exhibited branching/stellate morphogenesis highlighting the species-specific nature of cell–matrix interactions during morphological differentiation. Matrix produced in vitro by cells did not support formation of human breast cancer cell line, PMC42 mammospheres. This novel model system will help define the molecular pathways controlling the regulation of milk protein expression and species specific requirements of the extracellular matrix in the cape fur seal.

In-situ confocal Raman monitoring of evaporation process during protein crystallization

We have introduced an in-situ Raman monitoring technique to investigate the crystallization process inside protein drops. In addition to a conventional vapour-diffusion process, a novel procedure which actively stimulates the evaporation from a protein drop during crystallization was also evaluated, with lysozyme as a model protein. In contrast to the conventional vapour-diffusion condition, the evaporation-stimulated growth of crystals was initiated in a simple dehydration scheme and completed within a significantly shorter time. To gain an understanding of crystallization behaviours under the conditions with and without such evaporation stimulation, confocal Raman spectroscopy combined with linear regression analysis was used to monitor both lysozyme and HEPES buffer concentrations in real time. The confocal measurements having a high spatial resolution and good linear response revealed areas of local inhomogeneity in protein concentration when the crystallization started. The acquired concentration profiles indicated that (1)ÿthe evaporation-stimulated crystallization proceeded with protein concentrations lower than those under conventional vapour diffusion, and (2)ÿcrystals under the evaporation-stimulated condition were noticeable within an early stage of crystallization before the protein concentration approached its maximum value. The HEPES concentration profiles, on the other hand, increased steadily towards the end of the process regardless of the conditions used for crystallization. In particular, the observed local inhomogeneities specific to protein distribution suggested an accumulation mechanism of protein molecules that initiates the nucleation of crystals.

The "refoldability" of selected proteins in ionic liquids as a stabilization criterion, leading to a conjecture on biogenesis

The folding of proteins is usually studied in dilute aqueous solutions of controlled pH, but it has recently been demonstrated that reversible unfolding can occur in other media. Particular stability is conferred on the protein (folded or unfolded) when the process occurs in ‘protic ionic liquids’ (pILs) of controlled proton activity. This activity (‘effective pH’) is determined by the acid and base components of the pIL and is characterized in the present study by the proton chemical shift of the N–H proton. Here we propose a ‘refoldability’ or ‘refolding index’ (RFI) metric for assessing the stability of folded biomolecules in different solvent media, and demarcate high RFI zones in hydrated pIL media using ribonuclease A and hen egg white lysozyme as examples. Then we show that, unexpectedly, the same high RFIs can be obtained in pIL media that are 90% inorganic in character (simple ammonium salts). This leads us to a conjecture related to the objections that have been raised to ‘primordial soup’ theories for biogenesis, objections that are based on the observation that all the bonds involved in biomacromolecule formation are hydrolyzed in ordinary aqueous solutions unless specifically protected. The ingredients for primitive ionic liquids (NH3, CO, HCN, CO2, and water) were abundant in the early earth atmosphere, and many experiments have shown how amino acids could form from them also. Cyclical concentration in evaporating inland seas could easily produce the type of ambient-temperature, non-hydrolyzing, media that we have demonstrated here may be hospitable to biomolecules, and that may be actually encouraging of biopolymer assembly. Thus a plausible variant of the conventional ‘primordial soup’ model of biogenesis is suggested.