- Open Access
PathLocdb: a comprehensive database for the subcellular localization of metabolic pathways and its application to multiple localization analysis
© Zhao and Qu; licensee BioMed Central Ltd. 2010
- Published: 2 December 2010
In eukaryotes, the cell is divided into several compartments enclosed by unitary membranes. Such compartmentalization is critical for cells to restrict different pathways to be carried out in different subcellular regions. The summary and classification of subcellular localizations of metabolic pathways are the first steps towards understanding their roles in spatial differentiation and the specialization of metabolic pathways in different organisms.
Integrating the subcellular localization of enzymes and their pathways from UniProt Knowledgebase and KEGG pathway databases, we present the first database for subcellular localization of 43014 pathways from 80676 UniProt entries and their pathway annotations from UniProt and KEGG pathway databases. To extract pathway localization across organisms, we defined 889 superpathways as clusters of basic pathways with the same pathway annotations from different organisms. Over eighty-eight percent of superpathways in the Swiss-Prot dataset occur in cytoplasm and mitochondria. And over seventy percent of UniProt superpathways have multiple localization annotations. We summarized four common reasons for the multiple localization of superpathways. Based on this database, we also discovered 88 potential transport systems between different steps of multiply localized pathways and 45 duplicated genes from 17 pathways, occurring in parallel in several locations in humans.
PathLocdb is a free web-accessible database that enables biochemical researchers to quickly access summarized subcellular localization of pathways from UniProt and KEGG pathway databases. As the first effort to systematically integrate pathway localization, this database is very useful in discovering the variation of localization of pathways between organisms and also cross-talk between different organelles within a pathway. The Pathlocdb database is available at http://pathloc.cbi.pku.edu.cn.
- Subcellular Localization
- Pathway Annotation
- Pathway Localization
- Localization Annotation
- KEGG Pathway Database
In eukaryotes, cells are subdivided into membrane-bound subcellular organelles . Subcellular localization of metabolic enzymes can give precise control over where these are synthesized and operate . Such compartmentalization of metabolic enzymes and metabolites of pathways provides a regulatory mechanism to control metabolic pathways. Moreover, many pathways such as the β-oxidation of fatty acids occur in several subcellular organelles. The differentiation of localization of these pathways often causes the differences in their efficiency in utilizing the metabolites [3, 4]. The regulatory mechanisms for coordinating the different metabolic environments in different organelles are more complex . Furthermore, incorrect localization of enzymes are often implicated in serious diseases . Thus, understanding the distribution patterns of pathway localization is essential in discovering potential regulatory mechanisms and the localization of metabolic pathways. In addition, the spatial distribution pattern information of pathways is also important in clarifying pathway boundaries and in discovering the mechanism of intermediate communication between different subcellular compartments .
However, the subcellular localization information of pathways and their related enzymes has not been systematically integrated. The popular pathway databases such as KEGG, BioCyc, MPW and aMAZE are constructed for specific research tasks such as the graphic representation or reconstruction of metabolic pathways [7–10]. Thus, they seldom provide integrated localization information.
Summary and classification of subcellular localization of metabolic pathways are the first steps towards understanding their roles in spatial differentiation and functional specialization. Here, we present the first effort to systematically collect pathway localization information from the UniProt  and KEGG Ligand databases . The strategy to summarize pathway localization is mainly based on the pathway and subcellular localization annotations of their participating enzymes. Using an automatic pipeline, the subcellular localization data of 43014 pathways were integrated from 80676 UniProt entries and their pathway annotations from UniProt and KEGG pathway databases. Furthermore, 889 superpathways involving 33953 organisms were summarized from the 43014 pathways. Our results indicate that over eighty-eight percent of superpathways in the UniProt dataset occur in the cytoplasm and the mitochondria. Also, over 70% of UniProt superpathways contain multiple locations. From our data, we have proposed four common reasons for multiple localization of superpathways. As the first system-wide collection for the subcellular localization of metabolic pathways, PathLocdb provides a valuable understanding of distribution patterns of metabolic pathways among cellular organelles in different organisms.
Integration of UniProt and KEGG Ligand databases provides many clues for studying pathways and pathway evolution. Here we present four types of applications for further studies.
Differences in data content and pathway localization annotations from UniProt and KEGG databases
Our strategy to summarize the pathway localization was mainly based on the pathway and subcellular localization annotations from their participating enzymes. Thus, only enzymes with both subcellular localization and pathway annotations satisfied our criterion. The enzyme localization data are from UniProtKB/Swiss-Prot (version 56.8) in UniProt (version 14.8, Feb. 2009) databases. The pathway annotations were fro UniProtKB/Swiss-Prot (version 56.8) and KEGG ligand databases (version 44.0). In total, the Pathlocdb database consists of subcellular localization information of 43014 pathways summarized from 80676 UniProt entries and their pathway annotations from UniProt and KEGG pathway databases. To extract pathway localization across organisms, we defined a superpathway as a cluster of basic pathways with the same name from different organisms. A total of 889 superpathways across 33953 organisms were extracted from 43014 pathways. Due to differences in data quality and in pathway annotations, three datasets were collected. The high quality dataset contains 3448 pathways with localization information as well as 337 annotated superpathways across 795 organisms summarized from 6630 UniProtKB/Swiss-Prot entries. The KEGG pathway localization dataset includes 2535 pathways and 215 superpathways with localization information of the participating enzymes summarized from 12281 UniProtKB/Swiss-Prot entries of 42 organisms. A comprehensive dataset including 37029 pathways with localization information was collected from 70566 UniProt entries involving 33953 organisms.
Discovery of conserved pathway locations across organisms and organism specific pathway localization
Nine superpathways with one single conserved location across organisms.
Amino-acid biosynthesis ; S-adenosyl-L-methionine biosynthesis ; S-adenosyl-L-methionine from L-methionine
Carbohydrate degradation ; glycolysis ; D-glyceraldehyde 3-phosphate from glycerone phosphate
Cofactor biosynthesis; ubiquinone biosynthesis
Energy metabolism; oxidative phosphorylation
Fermentation ; pyruvate fermentation to lactate ; (S) -lactate from pyruvate
Lipid metabolism; peroxisomal fatty acid beta-oxidation
Phenylpropanoid metabolism ; cinnamic acid biosynthesis ; trans-cinnamic acid from L-phenylalanine
Porphyrin metabolism ; protoporphyrin-IX biosynthesis ; 5-aminolevulinate from glycine
Purine metabolism ; GMP biosynthesis ; GMP from XMP (glutamine route)
Using our database, 63 organism specific superpathways are found (Fig. 1c), 35 of them being plant-specific such as “abscisic acid (ABA) biosynthesis” , and 21 being fungi-specific. Pathway localization specific to certain organisms could also be found by sorting out completely different patterns of localization between different organisms. For instance, the gene Cullin-1 in our database belonging to “protein modification; protein neddylation” in Arabidopsis thaliana mainly localizes to the nucleus during interphase and preprophase . This is obviously different from homologous genes in mammals with their cellular membrane and lipid raft localization . In total, 56 superpathways with such completely different localization patterns across organisms have been discovered (Additional file 1). Moreover, 184 superpathways with partially different localization patterns between organisms (Additional file 2) have also been found. For example, “glycan biosynthesis; starch biosynthesis” in all surveyed plants are carried out in the chloroplast and the plastid, but the superpathways from some plants could localize additionally to the amyloplast.
Classification of multiple subcellular locations of metabolic pathways and prediction of potential transport systems
Pathways with multiple locations are a prevalent phenomenon. Among the annotated 337 UniProtKB/Swiss-Prot superpathways, 235 of them occur in multiple locations. Here we propose four common possible reasons giving rise to these multiple localization annotation of 337 UniProtKB/Swiss-Prot superpathways at the subcellular level. Firstly, different organisms have different pathways, resulting in multiple annotations of subcellular localization of superpathways. The most common examples for this category are the pathways which generate ATP from mitochondria and chloroplasts in different organism groups. Secondly, many pathways are composed of a series of steps spanning several subcellular locations. Like a production line, many intermetabolites are produced from one organelle to another, stepwise in pathways. Thirdly, some pathways occur in parallel in several subcellular locations in a single organism. The obvious example is fatty acid β-oxidation occurring in the mitochondria as well as the peroxisome . The other most common reason for multiple annotation of pathways is the fact that they are summarized from multiple localizing enzymes. An extreme example here is the first enzyme of GPI biosysthesis in humans, with five subunits [16–18]. Four of its five subunits localize in the endoplasmic reticulum and the remaining one in the cytoplasm. However, the above four phenomena are often mixed together when we try to analyze the reasons for multiple localization of pathways (Fig. 1d).
Regardless of the complex reasons leading to multiple annotation of pathway localization, each possible reason is useful for researchers to further explore its potential significance. The transport system or cross-talk between organelles are made up of common processes for translocation of intermediates [6, 19]. Focusing on the 3448 pathways of the UniProtKB/Swiss-Prot dataset, 88 pathways with multiple locations are summarized from the multiple localizations of their translocation steps (Additional file 3). These results provide information on the potential existence of a transport system between these pathways with different locations and different steps. For instance, the pathway titled “amine and polyamine degradation; betaine degradation; sarcosine from betaine” consists of two steps in human, mouse and rat. The first step of this pathway occurs in cytoplasm in the three organisms, while the second step is in the mitochondria. Further experimental studies on the potential transport system between the first and second steps of the pathway in mammals may be interesting and useful.
Pathways occurring in parallel in multiple locations and duplicated pathways in human
Using our database, 28 pathways occurring in parallel in several locations in human are found. On combini ng these pathways with the gene duplication dataset in human [20, 21], 45 genes from 17 of the 28 pathways are detected to be duplicated during evolution (Additional file 4). Taking two duplicated genes 17-beta-HSD 1 and 17-beta-HSD 12 as examples, their protein products are localized to the cytoplasm and the endoplasmic reticulum respectively. Both proteins are involved in the “Steroid biosynthesis; estrogen biosynthesis” pathway. Similar phenomena also exist in the “Protein modification; protein sumoylation” and “Protein modification; protein ubiquitination” pathways in human. Occurrence of such a high proportion of enzyme duplication during pathway evolution also deserves further study.
To our knowledge, the present study represents the first of its kind, presenting a genome-wide collection of subcellular localization of metabolic pathways in cells across 33953 organisms. With high quality data from the UniProtKB/Swiss-Prot and KEGG databases, we have compared the pathway localization annotations of the two databases and found that the contributions from KEGG are larger than those from UniProtKB/Swiss-Prot.
A characteristic feature of this study is the summary of common reasons for the multiple annotations of pathway localization. Recently high throughput proteomics data suggest that proteins with multiple locations constitute 39 percent of the total 1404 proteins in mouse liver . Our database shows that over 70% of UniProtKB/Swiss-Prot superpathways have multiple localization annotations. The consistent phenomena at the pathway level confirm that differentiation of localization of biological function is prevalent. Intensive studies on the pathways which are carried out through a series of steps spanning several subcellular locations reveal 88 potential transport systems between different steps of multiple localizing pathways.
Based on the multiple localization pathways from different organisms, we discovered possibly for the first time, all the conserved pathway localizations across organisms and organism-specific pathway localizations. This systematic comparison of pathway localization between organisms reveals that numerous pathways occur at different locations, providing clues for the differentiation and specialization of the pathway localizations between organisms.
Focusing on the 28 pathways occurring in parallel in multiple locations in human, we identified 17 pathways containing gene duplication events to gain new locations. According to one of the most popular pathway evolution models, the “patchwork model”, enzymes with broad substrate specificities are more likely to be recruited to execute a new metabolic function . Although our analyses show that a high proportion of enzyme duplication occurs during pathway evolution, the mechanism of gaining new locations need to be elucidated. Phylogenetic distribution of N-terminal targeting signals of mitochondrial localizing proteins suggest that inter-compartmental duplication events could bring novel localization of gene products and expand the catalytic as well as the RNA processing repertoire . While multiple localizations of metabolic pathways provide more valuable clues for researchers to discover potential evolution routes, we hope further study will elucidate the possible mechanisms.
In conclusion, these analyses demonstrate that our database is valuable in the discovery of potential cross-talk between different steps and the variation of localization of pathways between organisms.
PathLocdb was constructed as a free web-accessible database and analysis server to enable biochemical researchers to quickly access to summarized subcellular localization of metabolic pathways of UniProt and KEGG pathway database. As the first effort to systematically collect pathway localization, this database is very useful in discovering the variation of localization of pathways between organisms and also cross-talk between different organelles within a pathway. The PathLocdb database can be accessed at http://pathloc.cbi.pku.edu.cn.
Data source and automation of data collection
An automatic computational pipeline was implemented to extract subcellular localization and pathway annotations from UniProtKB/Swiss-Prot and KEGG ligand databases using Perl scripts and the Swissknife module. Eukaryotic protein entries were screened by its taxonomy identification within the taxonomic groups including fungi, invertebrates, vertebrates, plants, mammals, rodents and human. As our investigation was focused on the cellular organelle level, all the subcellular localization descriptions were formatted at the organelle level. For instance, the “Mitochondrial intermembrane space” is assigned to Mitochondria. All the subcellular localization descriptions marked with “By similarity”, “Probable” and “Potential” were also accepted. In UniProtKB/Swiss-Prot, pathway annotations are in free text and step information was extracted using keywords, such as “step 2/2”.
To compare the subcellular localization annotations of pathways from different organisms, we defined a superpathway as a cluster of pathways having the same annotation from different organisms. The reference pathways in KEGG are combined multiple pathways from different organisms. Thus, the superpathway name of the KEGG dataset is the same as the name of reference pathways. As the pathway names from UniProt are also formatted to structured controlled vocabulary in UniPathway , the superpathway name of the UniProtKB/Swiss-Prot dataset is the same as that in UniPathway.
As an automatic pipeline for data collection, our database will be easy to update regularly when new versions of Swiss-Prot and KEGG Ligand are available. Besides, more data from other pathway databases and literatures will be compiled in the future.
Web interface construction
This work was supported by the China National High-tech 863 Programs (2006AA02Z334, 2006AA02Z314, 2006AA02A312, 2007AA02Z165), 973 Programs (2006CB910404, 2007CB946904). And we gratefully acknowledges the support of K. C. Wong Education Foundation, Hong Kong.
This article has been published as part of BMC Genomics Volume 11 Supplement 4, 2010: Ninth International Conference on Bioinformatics (InCoB2010): Computational Biology. The full contents of the supplement are available online at http://www.biomedcentral.com/1471-2164/11?issue=S4.
- Lehninger AL: Lehninger principles of biochemistry. Edited by: D.L. Nelson and M.M. Cox. 2008, NewYork :: W.H. Freeman, 5th ed.Google Scholar
- Ovadi J, Saks V: On the origin of intracellular compartmentation and organized metabolic systems. Mol Cell Biochem. 2004, 256-257 (1-2): 5-12. 10.1023/B:MCBI.0000009855.14648.2c.View ArticlePubMedGoogle Scholar
- Masterson C, Wood C: Mitochondrial and peroxisomal beta-oxidation capacities of organs from a non-oilseed plant. Proceedings of the Royal Society of London Series B-Biological Sciences. 2001, 268 (1479): 1949-1953. 10.1098/rspb.2001.1783.View ArticleGoogle Scholar
- Yu XX, Drackley JK, Odle J: Rates of mitochondrial and peroxisomal beta-oxidation of palmitate change during postnatal development and food deprivation in liver, kidney and heart of pigs. Journal of Nutrition. 1997, 127 (9): 1814-1821.PubMedGoogle Scholar
- Kim H, McGrath BM, Silverstone PH: A review of the possible relevance of inositol and the phosphatidylinositol second messenger system (Pi-cycle) to psychiatric disorders -focus on magnetic resonance spectroscopy (MRS) studies. Human Psychopharmacology-Clinical and Experimental. 2005, 20 (5): 309-326. 10.1002/hup.693.View ArticlePubMedGoogle Scholar
- Walter L, Hajnoczky G: Mitochondria and endoplasmic reticulum: the lethal interorganelle cross-talk. J Bioenerg Biomembr. 2005, 37 (3): 191-206. 10.1007/s10863-005-6600-x.View ArticlePubMedGoogle Scholar
- Aoki KF, Kanehisa M: Using the KEGG database resource. Curr Protoc Bioinformatics. 2005, Chapter 1: Unit 1-12.Google Scholar
- Caspi R, Karp PD: Using the MetaCyc pathway database and the BioCyc database collection. Curr Protoc Bioinformatics. 2007, Chapter 1: Unit1-17.PubMedGoogle Scholar
- Selkov E: MPW: the Metabolic Pathways Database. Nucleic Acids Res. 1998, 26 (1): 43-5. 10.1093/nar/26.1.43.PubMed CentralView ArticlePubMedGoogle Scholar
- Lemer C: The aMAZE LightBench: a web interface to a relational database of cellular processes. Nucleic Acids Res. 2004, 32 (Database issue): D443-8. 10.1093/nar/gkh139.PubMed CentralView ArticlePubMedGoogle Scholar
- Bairoch A, Apweiler R, Wu CH, Barker WC, Boeckmann B, Ferro S, Gasteiger E, Huang H, Lopez R: Universal Protein Resource (UniProt). Nucleic Acids Res. 2005, 33 (Database issue): 154-9. 10.1093/nar/gki070.View ArticleGoogle Scholar
- Calvo S: Systematic identification of human mitochondrial disease genes through integrative genomics. Nature Genetics. 2006, 38 (5): 576-582. 10.1038/ng1776.View ArticlePubMedGoogle Scholar
- Nambara E, Marion-Poll A: Abscisic acid biosynthesis and catabolism. Annual Review of Plant Biology. 2005, 56: 165-185. 10.1146/annurev.arplant.56.032604.144046.View ArticlePubMedGoogle Scholar
- Shen WH: Null mutation of AtCUL1 causes arrest in early embryogenesis in Arabidopsis. Molecular Biology of the Cell. 2002, 13 (6): 1916-1928. 10.1091/mbc.E02-02-0077.PubMed CentralView ArticlePubMedGoogle Scholar
- Chen YZ: APP-BP1 mediates APP-induced apoptosis and DNA synthesis and is increased in Alzheimer's disease brain. Journal of Cell Biology. 2003, 163 (1): 27-33. 10.1083/jcb.200304003.PubMed CentralView ArticlePubMedGoogle Scholar
- Inoue N: PIG-C, one of the three human genes involved in the first step of glycosylphosphatidylinositol biosynthesis is a homologue of Saccharomyces cerevisiae GPI2. Biochem Biophys Res Commun. 1996, 226 (1): 193-9. 10.1006/bbrc.1996.1332.View ArticlePubMedGoogle Scholar
- Miyata T: The Cloning of Pig-a, a Component in the Early Step of Gpi-Anchor Biosynthesis. Science. 1993, 259 (5099): 1318-1321. 10.1126/science.7680492.View ArticlePubMedGoogle Scholar
- Watanabe R: The first step of glycosylphosphatidylinositol biosynthesis is mediated by a complex of PIG-A, PIG-H, PIG-C and GPI1. Embo J. 1998, 17 (4): 877-85. 10.1093/emboj/17.4.877.PubMed CentralView ArticlePubMedGoogle Scholar
- Bick JA, Lange BM: Metabolic cross talk between cytosolic and plastidial pathways of isoprenoid biosynthesis: unidirectional transport of intermediates across the chloroplast envelope membrane. Archives of Biochemistry and Biophysics. 2003, 415 (2): 146-154. 10.1016/S0003-9861(03)00233-9.View ArticlePubMedGoogle Scholar
- Bailey JA: Recent segmental duplications in the human genome. Science. 2002, 297 (5583): 1003-7. 10.1126/science.1072047.View ArticlePubMedGoogle Scholar
- Bailey JA: Segmental duplications: organization and impact within the current human genome project assembly. Genome Res. 2001, 11 (6): 1005-17. 10.1101/gr.GR-1871R.PubMed CentralView ArticlePubMedGoogle Scholar
- Foster LJ: A mammalian organelle map by protein correlation profiling. Cell. 2006, 125 (1): 187-199. 10.1016/j.cell.2006.03.022.View ArticlePubMedGoogle Scholar
- Rison SC, Thornton JM: Pathway evolution, structurally speaking. Curr Opin Struct Biol. 2002, 12 (3): 374-82. 10.1016/S0959-440X(02)00331-7.View ArticlePubMedGoogle Scholar
- Szklarczyk R, Huynen MA: Expansion of the human mitochondrial proteome by intra- and inter-compartmental protein duplication. Genome Biology. 2009, 10 (11): --10.1186/gb-2009-10-11-r135.PubMed CentralView ArticlePubMedGoogle Scholar
- Hermjakob H, Fleischmann W, Apweiler R: Swissknife - 'lazy parsing' of SWISS-PROT entries. Bioinformatics. 1999, 15 (9): 771-2. 10.1093/bioinformatics/15.9.771.View ArticlePubMedGoogle Scholar
- consortium, T.S.-P.g.a.t.U.: UniPathway: a metabolic door to UniProtKB/Swiss-Prot. In-Silico Analysis of Proteins - Celebrating the 20th anniversary of Swiss-Prot. 2006, Fortaleza, BrazilGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.