Predicting the subcellular localization of human proteins using machine learning and exploratory data analysis

Identifying the subcellular localization of proteins is particularly helpful in the functional annotation of gene products. In this study, we use Machine Learning and Exploratory Data Analysis （EDA） techniques to examine and characterize amino acid sequences of human proteins localized in nine cellular compartments. A dataset of 3,749 protein sequences representing human proteins was extracted from the SWISS-PROT database. Feature vectors were created to capture specific amino acid sequence characteristics. Relative to a Support Vector Machine, a Multi-layer Perceptron, and a Naive Bayes classifier, the C4.5 Decision Tree algorithm was the most consistent performer across all nine compartments in reliably predicting the subcellular localization of proteins based on their amino acid sequences （average Precision=0.88; average Sensitivity=0.86）. Furthermore, EDA graphics characterized essential features of proteins in each compartment. As examples, proteins localized to the plasma membrane had higher proportions of hydrophobic amino acids; cytoplasmic proteins had higher proportions of neutral amino acids; and mitochondrial proteins had higher proportions of neutral amino acids and lower proportions of polar amino acids. These data showed that the C4.5 classifier and EDA tools can be effective for characterizing and predicting the subcellular localization of human proteins based on their amino acid sequences.     

Control of nuclear-cytoplasmic shuttling of Ankrd54 by PKCδ

AIM To identify and characterize the effect of phosphorylation on the subcellular localization of Ankrd54.METHODS HEK293 T cells were treated with calyculin A, staurosporin or phorbol 12-myristate 13-acetate(PMA). Cells were transfected with eG FP-tagged Ankrd54 with or without Lyn tyrosine kinase(wild-type, Y397 F mutant, or Y508 F mutant). The subcellular localization was assessed by immunofluorescence imaging of cells, immunoblotting of subcellular fractionations. The phosphorylation of Ankrd54 was monitored using Phos-tagT M gel retardation. Phosphorylated peptides were analysed by multiplereaction-monitoring(MRM) proteomic analysis.RESULTS Activation of PKC kinases using PMA promoted nuclear export of Ankrd54 and correlated with increased Ankrd54 phosphorylation, assayed using Phos-tag TM gel retardation. Co-expression of an active form of the PKCδisoform specifically promoted both phosphorylation and cytoplasmic localization of Ankrd54, while PKCδ, Akt and PKA did not. Alanine mutation of several serine residues in the amino-terminal region of Ankrd54(Ser14, Ser17, Ser18, Ser19) reduced both PMA induced cytoplasmic localization and phosphorylation of Ankrd54. Using MRM proteomic analysis, phosphorylation of the Ser18 residue of Ankrd54 was readily detectable in response to PMA stimulation. PMA stimulation of cells co-expressing Ankrd54 and Lyn tyrosine kinase displayed increased coimmunoprecipitation and enhanced co-localization in the cytoplasm.CONCLUSION We identify phosphorylation by PKCδ as a major regulator of nuclear-cytoplasmic shuttling of Ankrd54, and its interaction with the tyrosine kinase Lyn.     

PredSL： A Tool for the N-terminal Sequence-based Prediction of Protein Subcellular Localization

The ability to predict the subcellular localization of a protein from its sequence is of great importance, as it provides information about the protein＇s function. We present a computational tool, PredSL, which utilizes neural networks, Markov chains, profile hidden Markov models, and scoring matrices for the prediction of the subcellular localization of proteins in eukaryotic cells from the N-terminal amino acid sequence. It aims to classify proteins into five groups： chloroplast, thylakoid, mitochondrion, secretory pathway, and ＂other＂. When tested in a fivefold cross-validation procedure, PredSL demonstrates 86.7% and 87.1% overall accuracy for the plant and non-plant datasets, respectively. Compared with TargetP, which is the most widely used method to date, and LumenP, the results of PredSL are comparable in most cases. When tested on the experimentally verified proteins of the Saccharomyces cerevisiae genome, PredSL performs comparably if not better than any available algorithm for the same task. Furthermore, PredSL is the only method capable for the prediction of these subcellular localizations that is available as a stand-alone application through the URL： http：//bioinformatics.biol.uoa.gr/PredSL/.     

Fast Fourier transform-based support vector machine for subcellular localization prediction using different substitution models

There are approximately 109 proteins in a cell. A hotspot in bioinformatics is how to identify a protein＇s subcellular localization, if its sequence is known. In this paper, a method using fast Fourier transform-based support vector machine is developed to predict the subcellular localization of proteins from their physicochemical properties and structural parameters. The prediction accuracies reached 83% in prokaryotic organisms and 84% in eukaryotic organisms with the substitution model of the c-p-v matrix （c, composition; p, polarity; and v, molecular volume）. The overall prediction accuracy was also evaluated using the ＂leave-one-out＂ jackknife procedure. The influence of the substitution model on prediction accuracy has also been discussed in the work. The source code of the new program is available on request from the authors.     

Subcellular Localization Analysis of Bovine Foamy Virus Borfl Protein

The Borfl protein is encoded by an immediate-early gene of the bovine foamy virus （BFV） and plays a key role in the viral life cycle. Borfl is a DNA binding protein which can transactivate both the long terminal repeat （LTR） and the internal promoter （IP） of BFV by specifically binding to the transactivation responsive element （TRE）. To analyze the subcellular localization of Borfl during the BFV life cycle, this gene was cloned into a prokaryotic expression vector and expressed in a soluble form. After the purification and immunization, we raised the mouse anti-Borfl serum with a high titer based on ELISA results. Western blot analysis showed that the antiserum could specifically recognize the Borfl protein that was expressed in 293T cells. With this specific serum, we revealed the nuclear and cytoplasmic localization of Borfl in HeLa cells that was transfected with Borfl. Moreover, the immuno-fluorescence assay also showed that the localization of Borfl during the infection and transfection of BFV was identical.     

Molecular properties of the Epstein-Barr virus BFRF3 gene

We report the results of cloning,expression,subcellular localization analysis,and molecular properties of the Epstein-Barr virus （EBV） BFRF3 gene,using several bioinformatics tools.Bioinformatics analysis indicated that the encoding protein of EBV BFRF3 gene （designated BFRF3） has a conserved Herpes_capsid domain,which was found to be closely related to the gammaherpesvirus capsid protein family,and is highly conserved among its counterparts.     

Mitochondrial pyruvate dehydrogenase E1 of Nosema bombycis： A marker in Microsporidian evolution

Microsporidia are a group of intracelluar eukaryotic parasites, which can infected almost all animals, including human beings. Till now, no mitochodria but mitosome, a remnant of mitochondria was discovered in this phylum. We present here the mitochondrial pyruvate dehydrogenase El （PDH, including PDHα and PDHβ） of the microsporidian Nosema bombycis, the pathogen of silkworm pebrine. Compared with PDH of microsporidian Encephalitozoon cuniculi and Antonospora locustae, both subunits are eonscrced. The phylogeny indicated that both subunits are mitochondrial. The syntenic maps revealed the subunits organization of NbPDH is distributed in different scaffolds, similar to that of EcPDH but different with AIPDH, and the relationship between phylogeny tree and organization of PDH suggest that the AlPDH subunits organization is the ancestral style of microsporidia, and through the genome evolution, the reshuffling of the chromosome of microsporidia occurred, the adjacent style of ALPDHE1 organization changed, and the two subunits separated and located to different chromosomes in E. cuniculi. For N. bombycis and N. ceranae, they locate to different scaffolds. In order to determine NbPDH subcellular localizations, we prepared the polyclonal antibodies against NbPDH prokaryotic fusion proteins, and adopted the colloidal gold immunological electron microscopy, the expression signals of NbPDH were observed in spores however, the subcellular localization were not definited. In general, through comparison of three mierosporidian PDH molecular phylogeny, subunits organization in chromosomes, localization indicated that PDH is an interesting marker in microsporidia evolution     

Identification of new centrosome proteins by autoimmune patient sera

Compared to other subcellular organelles, centrosome proteome can hardly be studied, due to the dif- ficulties in separation and purification of centrosome. Auto-antisera from 6 autoimmune patients, which recognized centrosome specifically in immunofluorescence, were used to identify the corresponding centrosomal proteins. The sera were first tested by Western blot on whole cell lysate, and all bound antibodies were then eluted from each single band in Western blot membrane to assure which antibody was responsible for the centrosome specific immunofluorescence staining. The corresponding pro- teins were obtained by immunoprecipitation and identified by mass spectrometry. Six centrosomal proteins, including 2 known centrosomal proteins and 4 proteins with unknown localization or report- edly non-centrosomal localization, were identified. These proteins apparently involve in cell cycle regulation, signal transduction pathways, molecular chaperons, and metabolism enzymes, which may reflect the expected functional diversity of centrosome.     

MORN motifs in plant PIPKs are involved in the regulation of subcellular localization and phospholipid binding

Multiple repeats of membrane occupation and recognition nexus （MORN） motifs were detected in plant phosphatidylinositl monophosphate kinase （PIPK）, a key enzyme in PI-signaling pathway. Structural analysis indicates that all the MORN motifs （with varied numbers at ranges of 7-9）, which shared high homologies to those of animal ones, were located at N-terminus and sequentially arranged, except those of OsPIPK1 and AtPIPK7, in which the last MORN motif was separated others by an -100 amino-acid ＂island＂ region, revealing the presence of two kinds of MORN arrangements in plant PIPKs. Through employing a yeast-based SMET （sequence of membrane-targeting） system, the MORN motifs were shown being able to target the fusion proteins to cell plasma membrane, which were further confirmed by expression of fused MORN-GFP proteins. Further detailed analysis via deletion studies indicated the MORN motifs in OsPIPK 1, together with the 104 amino-acid ＂island＂ region are involved in the regulation of differential subcellular localization, i.e. plasma membrane or nucleus, of the fused proteins. Fat Western blot analysis of the recombinant MORN polypeptide, expressed in Escherichia coli, showed that MORN motifs could strongly bind to PA and relatively slightly to PI4P and PI（4,5）P2. These results provide informative hints on mechanisms of subcellular localization, as well as regulation of substrate binding, of plant PIPKs.     

Identification of SUMO Targets by a Novel Proteomic Approach in Plants

Post-translational modifications (PTMs) chemically and physically alter the properties of proteins, including their folding, subcellular localization, stability, activity, and consequently their function. In spite of their relevance, studies on PTMs in plants are still limited. Small Ubiquitin-like Modifier (SUMO) modification regulates several biological processes by affecting protein-protein interactions, or changing the subcellular localizations of the target proteins. Here, we describe a novel proteomic approach to identify SUMO targets that combines 2-D liquid chromatography, immunodetection, and mass spectrometry (MS) analyses. We have applied this approach to identify nuclear SUMO targets in response to heat shock. Using a bacterial SUMOylation system, we validated that some of the targets identified here are, in fact, labeled with SUMO1. Interestingly, we found that GIGANTEA (GI), a photoperiodic-pathway protein, is modified with SUMO in response to heat shock both in vitro and in vivo.     

Depolymerization of actin cytoskeleton is involved in stomatal closure-induced by extracellular calmodulin in Arabidopsis

CaM ubiquitously presents inside eukaryotic cells. CaM抯 gene expression and its subcellular localization are regulated by light, osmotic stress, pathogens, plant hormones, etc.[1]. Intracellular CaM of plant displays important functions in pathogenesis and wounding reaction[2] and hypersensitive response[3]. CaM has been found extracellular spaces in many plant species, such as soluble extracts of oat coleoptile cell walls[4], the wheat coleoptile cell walls[5], maize root tips cell walls[6…     

Visualization of bHLH transcription factor interactions in living mammalian cell nuclei and developing chicken neural tube by FRET

Members of the basic helix-loop-helix （bHLH） gene family play important roles in vertebrate neurogenesis. In this study, confocal microscopy-based fluorescence resonance energy transfer （FRET） is used to monitor bHLH protein-protein interactions under various physiological conditions. Tissue-specific bHLH activators, NeuroD 1, Mash 1, Neurogenin 1 （Ngn 1）, Neurogenin2 （Ngn2）, and ubiquitous expressed E47 protein are tagged with enhanced yellow fluorescence protein （EYFP） and enhanced cyan fluorescence protein （ECFP）, respectively. The subcellular localization and mobility ofbHLH fusion proteins are examined in HEK293 cells. By transient transfection and in ovo electroporation, four pairs of tissue-specific bHLH activators and E47 protein are over-expressed in HEK293 cells and developing chick embryo neural tube. With the acceptor photobleaching method, FRET could be detected between these bHLH protein pairs in the nuclei of transfected cells and developing neural tubes. Mashl/E47 and Ngn2/E47 FRET pairs show higher FRET efficiencies in the medial and the lateral half of chick embryo neural tube, respectively. It suggests that these bHLH protein pairs formed functional DNA-protein complexes with regulatory elements of their downstream target genes in the specific regions. This work will help one understand the behaviours of bHLH factors in vivo.     

Phosphatidic Acid （PA） PP2A Activity and PIN1 Binds PP2AA1 to Regulate Polar Localization

Phospholipase D （PLD） exerts broad biological functions in eukaryotes through regulating downstream effectors by its product, phosphatidic acid （PA）. Protein kinases and phosphatases, such as mammalian target of rapa- mycin （mTOR）, Protein Phosphatase 1 （PP1） and Protein Phosphatase 2C （PP2C）, are PA-binding proteins that execute crucial regulatory functions in both animals and plants. PA participates in many signaling pathways by modulating the enzymatic activity and/or subcellular localization of bound proteins. In this study, we demonstrated that PLD-derived PA interacts with the scaffolding A1 subunit of Protein Phosphatase 2A （PP2A） and regulates PP2A-mediated PIN1 dephos- phorylation in Arabidopsis. Genetic and pharmacological studies showed that both PA and PP2A participate in the regu- lation of auxin distribution. In addition, both the phosphorylation status and polar localization of PIN1 protein were affected by PLD inhibitors, Exogenous PA triggered the membrane accumulation of PP2AA1 and enhanced the PP2A activity at membrane, while PLD inhibition resulted in the reduced endosomal localization and perinuclear aggregation of PP2AA1. These results demonstrate the important role of PLD-derived PA in normal PP2A-mediated PIN dephosphoryl- ation and reveal a novel mechanism, in which PA recruits PP2AA1 to the membrane system and regulates PP2A function on membrane-targeted proteins. As PA and PP2A are conserved among eukaryotes, other organisms might use similar mechanisms to mediate multiple biological processes.     

AtKP1, a kinesin-like protein, mainly localizes to mitochondria in Arabidopsis thaliana

Kinesins and kinesin-like proteins （KLPs） constitute a large family of microtubule-based motors that play important roles in many fundamental cellular and developmental processes. To date, a number of kinesins or KLPs have been identified in plants including Arabidopsis thaliana. Here, a polyclonal antibody against AtKP1 （kinesin-like protein 1 in A. thaliana） was raised by injection the expressed AtKP1 specific C-terminal polypeptides in rabbits, and immunoblot analysis was conducted with the affinity-purified anti-AtKP1 antibody. The results indicated that this antibody recognized the AtKP1 fusion proteins expressed in E. coli and proteins of -125 kDa in the soluble fractions of Arabidopsis extracts. The molecular weight was consistent with the calculated molecular weight based on deduced amino acids sequence of AtKP1. To acquire the subcellular localization of the protein, AtKP1 in Arabidopsis root cells was observed by indirect immunofluorescence microscopy. AtKP1 was localized to particle-like organelles in interphase or dividing cells, but not to mitotic microtubule arrays. Relatively more AtKP1 was found in isolated mitochondria fraction on immunoblot of the subcellular fractions. The AtKP1 protein could not be released following a 0.6 M KI washing, indicating that AtKP1 is tightly bind to mitochondria and might function associated with this kind of organelles.     

A PSTAIRE CDK-like protein localizes in nuclei and cytoplasm of Physarum polycephalum and functions in the mitosis

CDKs play key roles in controlling cell cycle progression in all eukaryotes. In plants, multiple CDKs are present,among which the best characterized CDKs are PSTAIRE CDKs. In this study, we carried out Western blot,immunoelectron microscopy and antibody treatment with an anti-PSTAIRE monoclonal antibody to explore the subcellular localization and functions of PSTAIRE CDKs in Physarum polycephalum. The results of Western blot and immunoelectron microscopy showed that in P. polycephalum, a PSTAIRE CDK-like protein was 34 kD in molecular weight and located in both nuclei and cytoplasm. In nuclei, the protein was mainly associated with chromosomes and nucleoli. The expression of the PSTAIRE CDK-like protein in both the plasmodia and nuclei showed little fluctuation through the whole cell cycle. When treated with an anti-PSTAIRE monoclonal antibody at early S phase, the cells were arrested in S phase, and the mitotic onset of P. polycephalum was blocked for about 1 h when treated at early G2 phase.Our data indicated that the PSTAIRE CDK-like protein has a direct bearing on the mitosis.     

Computational Identification of Protein-Protein Interactions in Rice Based on the Predicted Rice Interactome Network

Plant protein-protein interaction networks have not been identified by large-scale experiments. In order to better understand the protein interactions in rice, the Predicted Rice Interactome Network (PRIN; http://bis.zju.edu.cn/ prin/) presented 76,585 predicted interactions involving 5,049 rice proteins. After mapping genomic features of rice (GO annotation, subcellular localization prediction, and gene expression), we found that a well-annotated and biologically significant network is rich enough to capture many significant functional linkages within higher-order biological systems, such as pathways and biological processes. Furthermore, we took MADS-box do- main-containing proteins and circadian rhythm signaling pathways as examples to demonstrate that functional protein complexes and biological pathways could be effectively expanded in our predicted network. The expanded molecular network in PRIN has considerably improved the capability of these analyses to integrate existing knowledge and provide novel insights into the function and coordination of genes and gene networks.     

Cellular localization and biochemical characterization of a novel calcium-dependent protein kinase from tobacco

By screening tobacco cDNA library with MCK1 as a probe, we isolated a cDNA clone NtCPK5 （accession number AY971376）, which encodes a typical calcium-dependent protein kinase. Sequence analyses indicated that NtCPK5 is related to both CPKs and CRKs superfamilies and has all of the three conserved domains of CPKs. The biochemical activity of NtCPK5 was calcium-dependent. NtCPK5 had Vmax and Km of 526nmol/min/mg and 210μg/ml respectively with calf thymus histone （fraction Ⅲ, abbreviated to histone Ⅲs） as substrate. For substrate syntide-2, NtCPK5 showed a higher Vmax of 2008 nmol/min/mg and a lower Km of 30μM. The K0.5 of calcium activation was 0.04μM or 0.06μM for histone Ⅲs or syntide-2 respectively. The putative myristoylation and palmitoylation consensus sequence of NtCPK5 suggests that it could be a membrane-anchoring protein. Indeed, our transient expression experiments with wild type and mutant forms of NtCPK5/GFP fusion proteins showed that NtCPK5 was localized to the plasma membrane of onion epidermal cells and that the localization required the N-terminal acylation sites of NtCPK5/GFP. Taking together, our data have demonstrated the biochemical characteristics of a novel protein NtCPK5 and its subcellular localization as a membrane-anchoring protein.