Protein and Peptide Letters - Volume 19, Issue 4, 2012

Many proteins bear multi-locational characteristics, and this phenomenon is closely related to biological function. However, most of the existing methods can only deal with single-location proteins. Therefore, an automatic and reliable ensemble classifier for protein subcellular multi-localization is needed. We propose a new ensemble classifier combining the KNN (K-nearest neighbour) and SVM (support vector machine) algorithms to predict the subcellular localization of eukaryotic, Gram-negative bacterial and viral proteins based on the general form of Chou's pseudo amino acid composition, i.e., GO (gene ontology) annotations, dipeptide composition and AmPseAAC (Amphiphilic pseudo amino acid composition). This ensemble classifier was developed by fusing many basic individual classifiers through a voting system. The overall prediction accuracies obtained by the KNN-SVM ensemble classifier are 95.22, 93.47 and 80.72% for the eukaryotic, Gram-negative bacterial and viral proteins, respectively. Our prediction accuracies are significantly higher than those by previous methods and reveal that our strategy better predicts subcellular locations of multi-location proteins.

Computational prediction of protein structural class based on sequence data remains a challenging problem in current protein science. In this paper, a new feature extraction approach based on relative polypeptide composition is introduced. This approach could take into account the background distribution of a given k-mer under a Markov model of order k-2, and avoid the curse of dimensionality with the increase of k by using a T-statistic feature selection strategy. The selected features are then fed to a support vector machine to perform the prediction. To verify the performance of our method, jackknife cross-validation tests are performed on four widely used benchmark datasets. Comparison of our results with existing methods shows that our method provides satisfactory performance for structural class prediction.

DNA-binding proteins play an important role in most cellular processes, such as gene regulation, recombination, repair, replication, and DNA modification. In this article, an optimal Chou's pseudo amino acid composition (PseAAC) based on physicochemical characters of amino acid is proposed to represent proteins for identifying DNAbinding proteins. Six physicochemical characters of amino acids are utilized to generate the sequence features via the web server PseAAC. The optimal values of two important parameters (correlation factor δ and weighting factor w) about PseAAC are determined to get the appropriate representation of proteins, which ultimately result in better prediction performance. Experimental results on the benchmark datasets using random forest show that our method is really promising to predict DNA-binding proteins and may at least be a useful supplement tool to existing methods.

O-Acyl isodipeptides are prepared by coupling Boc-Ser/Thr-OBzl with Fmoc-Xaa-OH employing COMU, well known third generation peptide coupling agent. The reaction proceeds with high yield and the chemical homogeneity of the synthesized molecules were established via chiral HPLC analyses. The O-acyl isodipeptide units play crucial role in the success of ‘ click peptide’ protocol employed for assembling ‘ difficult sequence’ peptides.

Outer membrane proteins (OMPs) play important roles in cell biology. In addition, OMPs are targeted by multiple drugs. The identification of OMPs from genomic sequences and successful prediction of their secondary and tertiary structures is a challenging task due to short membrane-spanning regions with high variation in properties. Therefore, an effective and accurate silico method for discrimination of OMPs from their primary sequences is needed. In this paper, we have analyzed the performance of various machine learning mechanisms for discriminating OMPs such as: Genetic Programming, K-nearest Neighbor, and Fuzzy K-nearest Neighbor (Fuzzy K-NN) in conjunction with discrete methods such as: Amino acid composition, Amphiphilic Pseudo amino acid composition, Split amino acid composition (SAAC), and hybrid versions of these methods. The performance of the classifiers is evaluated by two datasets using 5-fold crossvalidation. After the simulation, we have observed that Fuzzy K-NN using SAAC based-features makes it quite effective in discriminating OMPs. Fuzzy K-NN achieves the highest success rates of 99.00% accuracy for discriminating OMPs from non-OMPs and 98.77% and 98.28% accuracies from α-helix membrane and globular proteins, respectively on dataset1. While on dataset2, Fuzzy K-NN achieves 99.55%, 99.90%, and 99.81% accuracies for discriminating OMPs from non- OMPs, α-helix membrane, and globular proteins, respectively. It is observed that the classification performance of our proposed method is satisfactory and is better than the existing methods. Thus, it might be an effective tool for high throughput innovation of OMPs.

A prior knowledge of protein structural class can provide useful information about its overall structure. So, it is vitally important to develop a computational prediction method for fast and accurately determining the protein structural class. In this paper, a dual-layer wavelet support vector machine (WSVM) is presented via the general form of Chou's pseudo amino acid composition, which is featured by introducing wavelet as a kernel and making decisions by the fusion from three individual classifiers. As a demonstration, the rigorous jackknife cross-validation tests were performed on two benchmark datasets, including the more challenging 25PDB dataset. Our success rates were reliable, and it has not escaped from our notice that the present method has specific ability to predict the most difficult case of α+β class. The program developed can be acquired freely on request from the authors.

Defensins are important components in host defense systems. The therapeutic use of β-defensins is limited by their innate toxicity and high cost due to the size and complex disulfide pairing. In this study, we used linear avian β- defensin-4 (RL38) without disulfide bonds as model peptide to derive two peptides by the truncation. GL23 is the Cterminal truncated sequence of RL38, and GLI23 is the derivative of GL23 by the replacement of cysteines with isoleucines. Results showed that these peptides exhibited strong antibacterial activity against gram-negative and gram-positive bacteria. An exception was that GL23 showed weak antimicrobial activity against gallinaceous pathogenic bacteria Salmonella Pullorum C79-13. Two truncated peptides GL23 and GLI23 displayed no or weak hemolysis, which was in accordance with little blue shifts of the peptides in the presence of synthetic eukaryotic membranes. CD spectroscopy demonstrated that these peptides appeared to be unfolded in aqueous solution but acquire structure in the presence of membrane- mimicking phospholipids. GLI23 kept the antibacterial activity at high concentrations of NaCl or low concentration of divalent cations (Mg2+ and Ca2+). The peptides preferentially bound to negatively charged phospholipids over zwitterionic phospholipids, which led to greater cell selectivity. The outer and inner membranes assay displayed that GLI23 killed bacteria by targeting the cell membrane. These results suggest the peptides derived by truncation of linear β-defensins may be a promising candidate for future antibacterial agent.

Identification on protein folding types is always based on the 27-class folds dataset, which was provided by Ding & Dubchak in 2001. But with the avalanche of protein sequences, fold data is also expanding, so it will be the inevitable trend to improve the existing dataset and expand more folding types. In this paper, we construct a multi-class protein fold dataset, which contains 3,457 protein chains with sequence identity below 35% and could be classified into 76 fold types. It was 4 times larger than Ding & Dubchak's dataset. Furthermore, our work proposes a novel approach of support vector machine based on optimal features. By combining motif frequency, low-frequency power spectral density, amino acid composition, the predicted secondary structure and the values of auto-correlation function as feature parameters set, the method adopts criterion of the maximum correlation and the minimum redundancy to filter these features and obtain a 95-dimensions optimal feature subset. Based on the ensemble classification strategy, with 95-dimensions optimal feature as input parameters of support vector machine, we identify the 76-class protein folds and overall accuracy measures up to 44.92% by independent test. In addition, this method has been further used to identify upgraded 27-class protein folds, overall accuracy achieves 66.56%. At last, we also test our method on Ding & Dubchak's 27-class folds dataset and obtained better identification results than most of the previous reported results.

As major chaperone of eye lens, alpha-crystallin (&agr:-Crs) is responsible for the transparency and refractive power of this region by preventing denaturation and precipitation of other proteins. As shown previously, cataract formation was positively associated with high salt intake and the elevation of blood sugar level. Here the effect of both temperature and ionic strength were studied on structure and chaperoning function of glycated and non-glycated &agr:-Crs. While chaperone activity of these proteins was increased as function of temperature elevation, in the presence of sodium salt (0- 160 mM), it was significantly decreased. As shown by fluorescence and circular dicroism (CD) instruments, the salt induced structural alteration of &agr:-Crs was accompanied with the exposure of hydrophobic surfaces and a transition from alpha- helical to beta-sheet structures. Moreover, the structural alterations induced by the salt were more pronounced in the case of glycated &agr:-Crs compared to that of non-glycated protein counterpart. Overall this study shows the structural changes accompanied with lose of the chaperone activity of &agr:-Crs induced by sodium chloride. Consequently, the obtained results may provide new evidences for the relationship between high salt intakes and cataract disease induced particularly by prolonged hyperglycemia.

Human antimicrobial peptides (AMPs), including defensins, have come under intense scrutiny owing to their key multiple roles as antimicrobial agents. Not only do they display direct action on microbes, but also recently they have been shown to interact with the immune system to increase antimicrobial activity. Unfortunately, since mechanisms involved in the binding of AMPs to mammalian cells are largely unknown, their potential as novel anti-infective agents cannot be exploited yet. Following the reported interaction of Human Neutrophil Peptide 1 dimer (HNP1) with a low density lipoprotein receptor (LDLR), a computational study was conducted to discover their putative mode of interaction. State-of-the-art docking software produced a set of LDLR-HNP1 complex 3D models. Creation of a 3D motif capturing atomic interactions of the LDLR binding interface allowed selection of the most plausible configurations. Eventually, only two models were in agreement with the literature. Binding energy estimations revealed that only one of them is particularly stable, but also interaction with LDLR weakens significantly bonds within the HNP1 dimer. This may be significant since it suggests a mechanism for internalisation of HNP1 in mammalian cells. In addition to a novel approach for complex structure prediction, this study proposes a 3D model of the LDLR-HNP1 complex which highlights the key residues which are involved in the interactions. The putative identification of the receptor binding mechanism should inform the future design of synthetic HNPs to afford maximum internalisation, which could lead to novel anti-infective drugs.

Measles virus (MV), one of the most contagious agents, infects immune cells using the signaling lymphocyte activation molecule (SLAM) on the cell surface. A complex of SLAM and the attachment protein, hemagglutinin (MVH), has remained elusive due to the intrinsic handling difficulty including glycosylation. Furthermore, crystals obtained of this complex are either nondiffracting or poorly-diffracting. To solve this problem, we designed a systematic approach using a combination of the following techniques; (1) a transient expression system in HEK293SGnTI(-) cells, (2) lysine methylation, (3) structure-guided mutagenesis directed at better crystal packing, (4) Endo H treatment, (5) single-chain formation for stable complex, and (6) floating-drop vapor diffusion. Using our approach, the receptor-binding head domain of MV-H covalently fused with SLAM was successfully crystallized and diffraction was improved from 4.5 Å to a final resolution of 3.15 Å. These combinational methods would be useful as crystallization strategies for complexes of glycoproteins and their receptors.

Bauninia forficata is trivially known as cow paw, and popularly used in Brazil for treatment of diabetes mellitus. Denominated baupain a cysteine proteinase was purified from B. forficata leaves. In this study, we investigated the baupain effect on aggregation of isolated human platelets in vitro and the results show that baupain hinders thrombin - but not ADP- and collagen- induced platelet aggregation. With synthetic quenched-fluorescent peptides, the kinetics of the cleavage site of human proteinase-activated receptor 1 / 2 / 3 and 4 [PAR-1 / 2 / 3 and 4] by baupain was determined. In conclusion, similar to bromelain and papain, baupain hinders human platelets aggregation, probably through an unspecific cleavage in the Phe-Leu bond of PAR1.