A third of the yeast mitochondrial proteome is dual localized: a question of evolution

There are a growing number of examples of identical or almost identical proteins, which are localized to two (or more) separate compartments, a phenomenon that is termed protein dual localization, dual distribution or dual targeting. We previously divided a reference set of known yeast mitochondrial proteins into two groups, suggested to be dual localized or exclusive mitochondrial proteins. Here we examined this evaluation by screening 320 mitochondrial gene products for dual targeting, using the α-complementation assay. The analysis of the results of this experimentally independent screen supports our previous evaluation that dual localized mitochondrial proteins constitute a subgroup of mitochondrial proteins with distinctive properties. These proteins are characterized by a lower probability of mitochondrial localization (MitoProtII score), a lower net charge and are enriched for proteins with a weaker mitochondrial targeting sequence. Conversely, mRNAs of exclusive mitochondrial proteins are enriched in polysomes associated with mitochondria. Based on the discovery of more than 60 new gene products that are now assumed to be dual targeted, we have updated an annotation list of dual-targeted proteins. We currently estimate that more than a third of the mitochondrial proteome is dual targeted, and suggest that this abundant dual targeting presents an evolutionary advantage.     

Selection of relevant features from amino acids enables development of robust classifiers

Machine learning (ML) has been extensively applied to develop models and to understand high-throughput data of biological processes. However, new ML models, trained with novel experimental results, are required to build regularly for more precise predictions. ML methods can build models from numeric data, whereas biological data are generally textual (DNA, protein sequences) or images and needs feature calculation algorithms to generate quantitative features. Programming skills along with domain knowledge are required to develop these algorithms. Therefore, the process of knowledge discovery through ML is decelerated due to lack of generic tools to construct features and to build models directly from the data. Hence, we developed a schema that calculates about 5,000 features, selects relevant features and develops protein classifiers from the training data. To demonstrate the general applicability and robustness of our method, fungal adhesins and nuclear receptor proteins were used for building classifiers which outperformed existing classifiers when tested on independent data. Next, we built a classifier for mitochondrial proteins of Plasmodium falciparum which causes human malaria because the latest corresponding classifiers are not publically accessible. Our classifier attained 98.18 % accuracy and 0.95 Matthews correlation coefficient by fivefold cross-validation and outperformed existing classifiers on independent test set. We implemented this schema as user-friendly and open source application Pro-Gyan (http://code.google.com/p/pro-gyan/), to build and share executable classifiers without programming knowledge.     

Analysis of nuclear lamin isoprenylation in Xenopus oocytes: isoprenylation of lamin B3 precedes its uptake into the nucleus

Protein prenylation is a posttranslational modification involving the covalent attachment of a prenyl lipid to a cysteine at or near the COOH terminus of a protein. It is required for membrane localization and efficient function of a number of cytoplasmic as well as nuclear proteins including the proto-oncogenic and activated forms of Ras. Farnesylation in conjunction with a nuclear localization signal has been shown to be necessary to target newly synthesized nuclear lamins to the inner nuclear envelope membrane. It is, however, not clear where in the cell isoprenylation of nuclear lamins takes place. In this study we describe in vivo and in vitro experiments on the isoprenylation of the Xenopus oocyte nuclear lamin B3. We show by kinetic analysis that newly synthesized lamins are isoprenylated in the cytosol of oocytes before uptake into the nucleus. From our data it can be concluded that isoprenylation of lamins in the nucleus, as it is observed under certain conditions of isoprene starvation, represents a default pathway rather than the physiological situation. We further analyzed the capacity of isolated nuclei to carry out isoprenylation of B3. Our results are in line with a dual localization of a protein farnesyltransferase in the cytosol and nuclei of amphibian oocytes. Implications for the possible functions of a nuclear protein farnesyltransferase as well as possible mechanisms of the selective inhibition of farnesylation of cytoplasmic proteins by peptidomimetics are discussed.     

Ca(2+)-dependent association of S100A6 (Calcyclin) with the plasma membrane and the nuclear envelope.

Expression of S100A6 (Calcyclin), a member of the S100 family and of Zn(2+)-binding proteins is elevated in a number of malignant tumors. In vitro the protein associates with several actin-binding proteins and annexins in a Ca(2+)-dependent manner. We have now studied the subcellular localization of S100A6 using a new, specific monoclonal antibody. Immunofluorescence microscopy of unfixed, ultrathin, frozen sections demonstrated a dual localization of S100A6 at the nuclear envelope and the plasma membrane of porcine smooth muscle only in the presence of Ca(2+). The same localization was found by immunofluorescence and immunogold electron microscopy as well as by confocal laser scanning microscopy with cultured, fixed, human CaKi-2 and porcine ST interphase cells. Upon cell division, however, S100A6 was found exclusively in the cytoplasm. Cell fractionation studies showed that S100A6 was present in the microsomal fraction in the presence of Ca(2+) and was released from this fraction by the addition of EGTA/EDTA but not by Triton X-100. The data demonstrate that S100A6 is localized both at the plasma membrane and the nuclear envelope in vivo and suggest a Ca(2+)-dependent interaction with annexins or other components of the nuclear envelope.     

Quantitation of de novo localized (15)N-labeled lipoproteins and membrane proteins having one and two transmembrane segments in a Bacillus subtilis secA temperature-sensitive mutant using 2D-PAGE and MALDI-TOF MS

We developed a means of quantifying proteins that have just localized in the cytoplasmic membrane using 15N-whole cell labeling together with 2D-PAGE and MALDI-TOF MS. The localization of 18 among 20 proteins consisting of 8 lipoproteins, 11 integral membrane proteins having one or two transmembrane segments and one secretory protein in the membrane fractions of Bacillus subtilis, was inhibited by the absence of SecA in a temperature-sensitive mutant. The time course of inhibition indicated that SecA participates in the localization of those proteins through immediately dependent, delayed dependent, and independent ways.     

Nuclear localization of enzymatically active green fluorescent protein-CTP:phosphocholine cytidylyltransferase alpha fusion protein is independent of cell cycle conditions and cell types

To address the recent controversy about the subcellular localization of CTP:phosphocholine cytidylyltransferase alpha (CTalpha), this study was designed to visualize green fluorescent protein (GFP). CTalpha fusion proteins directly and continuously under different conditions of cell cycling and in various cell lines. The GFP. CTalpha fusion proteins were enzymatically active and capable of rescuing mutant cells with a temperature-sensitive CT. The expressed GFP.CTalpha fusion protein was localized to the nucleus in all cell lines and required the N-terminal nuclear targeting sequence. Serum depletion/replenishment did not cause shuttling of CTalpha between the nucleus and cytoplasm. Moreover, the subcellular localization of CTalpha was examined continuously through all stages of the cell cycle in synchronized cells. No shuttling of CTalpha between the nucleus and cytoplasm was observed at any stage of the cell cycle. Stimulation of cells with oleate had no effect on the localization of CTalpha. The GFP.CTalpha lacking the nuclear targeting sequence stayed exclusively in the cytoplasm. Regardless of their localization, the GFP.CTalpha fusion proteins were equally active for phosphatidylcholine synthesis and mutant rescue. We conclude that the nuclear localization of CTalpha is a biological event independent of cell cycle in most mammalian cells and is unrelated to activation of phosphatidylcholine synthesis.     

Identification of a novel nuclear localization signal common to 69- and 82-kDa human choline acetyltransferase

We demonstrated previously that 69- and 82-kDa human choline acetyltransferase are localized predominantly to the cytoplasm and the nucleus, respectively. We have now identified a nuclear localization signal common to both forms of enzyme using confocal microscopy to study the subcellular compartmentalization of choline acetyltransferase tagged with green fluorescent protein in living HEK 293 cells. To identify functional nuclear localization and export signals, portions of full-length 69-kDa choline acetyltransferase were cloned into the vector peGFP-N1 and the cellular distribution patterns of the fusion proteins observed. Of the nine constructs studied, one yielded a protein with nuclear localization and another produced a protein with cytoplasmic localization. Mutation of the critical amino acids in this novel putative nuclear localization signal in the 69- and 82-kDa enzymes demonstrated that it is functional in both proteins. Moreover, 69-kDa choline acetyltransferase but not the 82-kDa enzyme is transported out of the nucleus by the leptomycin B-sensitive Crm-1 export pathway. By using bikaryon cells expressing both 82-kDa choline acetyltransferase and the nuclear protein heterogeneous nuclear ribonucleoprotein with green and red fluorescent tags, respectively, we found that the 82-kDa enzyme does not shuttle out of the nucleus in measurable amounts. These data suggest that 69-kDa choline acetyltransferase is a nucleocytoplasmic shuttling protein with a predominantly cytoplasmic localization determined by a functional nuclear localization signal and unidentified putative nuclear export signal. For 82-kDa choline acetyltransferase, the presence of the unique amino-terminal nuclear localization signal plus the newly identified nuclear localization signal may be involved in a process leading to predominantly nuclear accumulation of this enzyme, or alternatively, the two nuclear localization signals may be sufficient to overcome the force(s) driving nuclear export.     

Predicting protein subnuclear localization using GO-amino-acid composition features

The nucleus guides life processes of cells. Many of the nuclear proteins participating in the life processes tend to concentrate on subnuclear compartments. The subnuclear localization of nuclear proteins is hence important for deeply understanding the construction and functions of the nucleus. Recently, Gene Ontology (GO) annotation has been used for prediction of subnuclear localization. However, the effective use of GO terms in solving sequence-based prediction problems remains challenging, especially when query protein sequences have no accession number or annotated GO term. This study obtains homologies of query proteins with known accession numbers using BLAST to retrieve GO terms for sequence-based subnuclear localization prediction. A prediction method PGAC, which involves mining informative GO terms associated with amino acid composition features, is proposed to design a support vector machine-based classifier. PGAC yields 55 informative GO terms with training and test accuracies of 85.7% and 76.3%, respectively, using a data set SNL_35 (561 proteins in 9 localizations) with 35% sequence identity. Upon comparison with Nuc-PLoc, which combines amphiphilic pseudo amino acid composition of a protein with its position-specific scoring matrix, PGAC using the data set SNL_80 yields a leave-one-out cross-validation accuracy of 81.1%, which is better than that of Nuc-PLoc, 67.4%. Experimental results show that the set of informative GO terms are effective features for protein subnuclear localization. The prediction server based on PGAC has been implemented at http://iclab.life.nctu.edu.tw/prolocgac.     

Subcellular localization prediction with new protein encoding schemes

Subcellular localization is one of the key properties in functional annotation of proteins. Support vector machines (SVMs) have been widely used for automated prediction of subcellular localizations. Existing methods differ in the protein encoding schemes used. In this study, we present two methods for protein encoding to be used for SVM-based subcellular localization prediction: n-peptide compositions with reduced amino acid alphabets for larger values of n and pairwise sequence similarity scores based on whole sequence and N-terminal sequence. We tested the methods on a common benchmarking data set that consists of 2,427 eukaryotic proteins with four localization sites. As a result of 5-fold cross-validation tests, the encoding with n-peptide compositions provided the accuracies of 84.5, 88.9, 66.3, and 94.3 percent for cytoplasmic, extracellular, mitochondrial, and nuclear proteins, where the overall accuracy was 87.1 percent. The second method provided 83.6, 87.7, 87.9, and 90.5 percent accuracies for individual locations and 87.8 percent overall accuracy. A hybrid system, which we called PredLOC, makes a final decision based on the results of the two presented methods which achieved an overall accuracy of 91.3 percent, which is better than the achievements of many of the existing methods. The new system also outperformed the recent methods in the experiments conducted on a new-unique SWISSPROT test set     

Nuclear transport of plant potyviral proteins.

We have used immunoblotting, immunocytochemical, and gene fusion methods to examine the differential subcellular partitioning of tobacco etch potyvirus proteins that are potentially associated with RNA replication. From the earliest timepoints at which viral proteins could be detected, proteins Nla (49-kilodalton proteinase) and Nlb (58-kilodalton polymerase) were localized primarily in the nucleus, whereas the 71-kilodalton cylindrical inclusion protein was identified in the cytoplasm. The Nla and Nlb coding regions were fused to the beta-glucuronidase (GUS) sequence in a plant expression vector, resulting in synthesis of chimeric proteins in transfected protoplasts and in transgenic plants. In situ localization of GUS activity revealed nuclear localization of the GUS-Nla and GUS-Nlb fusion proteins and cytoplasmic localization of nonfused GUS. These results indicate that both Nla and Nlb contain nuclear targeting signals, and that they may serve as useful models for studies of plant cell nuclear transport. A discussion of the general utility of the nuclear transport system described here, as well as the role of nuclear translocation of potyviral proteins, is presented.     

A mutant herpesvirus protein leads to a block in nuclear localization of other viral proteins.

The herpes simplex virus mutants KOS1.1 ts756 and HFEM tsLB2 express temperature-sensitive ICP4 proteins that are not localized properly to the cell nucleus at the nonpermissive temperature. In these infected cells at the nonpermissive temperature, nuclear localization of at least two other viral proteins, ICP0 and ICP8, is impaired. Replacement of the mutated sequences in the ICP4 gene of tsLB2 restored proper nuclear localization of all of the proteins. The ICP0 and ICP8 proteins expressed in cells transfected with their individual genes were localized to the cell nucleus. Therefore, in infected cells, the mutant ICP4 gene product appears to be the primary defect which leads to the block in nuclear localization of the other proteins. One viral protein, ICP27, was not inhibited for nuclear localization in these cells. These data indicate that there are at least two pathways for nuclear localization of HSV proteins, one of which is inhibited by the mutant ICP4 protein. The mutant ICP4 protein may define a probe for one of the pathways of nuclear localization of proteins.     

Yeast ubiquitin ligase Rsp5 contains nuclear localization and export signals

The Rsp5 ubiquitin ligase regulates numerous cellular processes. Rsp5 is mainly localized to the cytoplasm but nuclear localization was also reported. A potential nuclear export signal was tested for activity by using a GFP(2) reporter. The 687-LIGGIAEIDI-696 sequence located in the Hect domain was identified as a nuclear export signal active in a Crm1-dependent manner, and its importance for the localization of Rsp5 was documented by using fluorescence microscopy and a lacZ-based reporter system. Analysis of the cellular location of other Rsp5 fragments fused with GFP(2) indicated two independent potential nuclear localization signals, both located in the Hect domain. We also uncovered Rsp5 fragments that are important to targeting/tethering Rsp5 to various regions in the cytoplasm. The presented data indicate that Rsp5 ligase is a shuttling protein whose distribution within the cytoplasm and partitioning between cytoplasmic and nuclear locations is determined by a balance between the actions of several targeting sequences and domains.     

Using pseudo amino acid composition to predict protein subnuclear location with improved hybrid approach

Summary. The subnuclear localization of nuclear protein is very important for in-depth understanding of the construction and function of the nucleus. Based on the amino acid and pseudo amino acid composition (PseAA) as originally introduced by K. C. Chou can incorporate much more information of a protein sequence than the classical amino acid composition so as to significantly enhance the power of using a discrete model to predict various attributes of a protein, an algorithm of increment of diversity combined with the improved quadratic discriminant analysis is proposed to predict the protein subnuclear location. The overall predictive success rates and correlation coefficient are 75.4% and 0.629 for 504 single localization proteins in jackknife test, and 80.4% for an independent set of 92 multi-localization proteins, respectively. For 406 single localization nuclear proteins with ≤25% sequence identity, the results of jackknife test show that the overall accuracy of prediction is 77.1%. Authors’ address: Qian-Zhong Li, Laboratory of Theoretical Biophysics, Department of Physics, College of Sciences and Technology, Inner Mongolia University, Hohhot 010021, China