Elucidation of the evolutionary expansion of phosphorylation signaling networks using comparative phosphomotif analysis
© Yoshizaki and Okuda; licensee BioMed Central Ltd. 2014
Received: 11 February 2014
Accepted: 26 June 2014
Published: 1 July 2014
Protein phosphorylation is catalyzed by kinases and is involved in the regulation of a wide range of processes. The phosphosites in protein sequence motifs determine the types of kinases involved. The development of phosphoproteomics has allowed the identification of huge numbers of phosphosites, some of which are not involved in physiological functions.
We developed a method for extracting phosphosites with important roles in cellular functions and determined 178 phosphomotifs based on the analysis of 34,366 phosphosites. We compared the conservation of serine/threonine/tyrosine residues observed in humans and seven other species. Consequently, we identified 16 phosphomotifs, where the level of conservation increased among species. The highly conserved phosphomotifs in humans and the worm were kinase regulatory sites. The motifs present in the fly were novel phosphomotifs, including zinc finger motifs involved in the regulation of gene expression. Subsequently, we found that this zinc finger motif contributed to subcellular protein localization. The motifs identified in fish allowed us to detect the expansion of phosphorylation signals related to alternative splicing. We also showed that the motifs present in specific species functioned in an additional network that interacted directly with the core signaling network conserved from yeast to humans.
Our method may facilitate the efficient extraction of novel phosphomotifs with physiological functions, thereby contributing greatly to the analysis of complex phosphorylation signaling cascades. Our study suggests that the phosphorylation networks acquired during evolution have added signaling network modules to the core signaling networks.
Protein phosphorylation is a ubiquitous post-translational modification, which controls several cellular processes in cell signaling networks. Cellular protein phosphorylation is catalyzed by protein kinases, which are one of the largest gene families in eukaryotes. The disruption of cellular protein phosphorylation events causes several diseases, including cancer and diabetes, and kinase inhibitory molecules may have potential uses in targeted molecular therapy to combat certain diseases. The human genome encodes 518 protein kinases, which are divided into nine groups based on their sequence similarities, including one tyrosine kinase family and eight serine/threonine kinase families . Several protein kinases have local substrate specificities that are determined by the amino acid sequence surrounding the phosphosite as a kinase motif. The techniques used for protein phosphorylation analysis have advanced rapidly since the development of high-throughput phosphoproteomic analysis, which has yielded extremely large phosphopeptide datasets [2–4]. Techniques and algorithms have also been developed for bioinformatics analysis that combine large-scale phosphoproteomics datasets with other information from a wide variety of life science databases to predict signal transduction networks, thereby leading to a better understanding of cellular functions [5, 6]. However, it is difficult to extract useful information related to cellular functions based on phosphosites alone because it is necessary to understand both the kinase protein and the functional change in its substrate protein caused by phosphorylation. Thus, a large number of phosphosites have been reported, but only a small number of kinase–substrate relationships have been identified. Therefore, several studies have attempted to develop methods to predict these relationships [7–11].
Previous analyses of intracellular signaling networks used methods that subtract the subnetworks identified in various conditions, such as cell cycles and ligand stimulation, or in different species . Similarity methods such as evolutionary comparative analyses of phosphosite sequences have been used for phosphorylation signaling network analysis [13–16]. The position of a phosphosite in a protein is affected by both its primary structure and the corresponding kinase, which may be changed by evolutionary processes [1, 17]. In general, the amino acids in a protein have evolved at a constant rate from yeasts to humans. In the present study, we assume the evolutionary order of model organisms on the basis of the evolutionary distances from the viewpoint of humans, as described previously [18, 19]. Phosphorylated serine/threonine amino acids have an evolutionary rate similar to that of other amino acids and they exhibit rather low conservation [20, 21]. However, proteins with constrained specific amino acid sequences, such as R-X-X-S/T, rather than a phosphorylated serine alone, are likely to be evolutionarily conserved . These conserved motifs probably play important roles in cellular functions [23–25]. Thus, comparative evolutionary analyses based on sequence evolution are powerful tools for extracting phosphorylation signals that are directly linked to cellular functions. Previous comparative evolutionary analyses of phosphorylation have focused on conserved sites for exploring the important signaling networks that are considered to be essential for cellular functions [14, 16, 20] and for the rewiring of signaling networks [13, 21, 24, 25].
However, focusing on highly conserved sites would not allow us to extract the highly evolutionarily conserved signaling networks that are essential for cellular functions because additional signal transduction pathways have been acquired throughout evolution; therefore, low conservation sites are also important in cellular functions. However, it is also well known that several low conservation phosphosites do not have physiological functions. Previous comparative analyses could not distinguish these differences; therefore, we developed a method for investigating evolutionary patterns to determine the peripheral components that have been acquired throughout evolution. We analyzed a large number of phosphosites and simplified the complex phosphorylation signaling networks. The same motifs located in different proteins are likely to be phosphorylated by a kinase that belongs to a closely related protein family and these motifs may be useful for identifying candidate kinases responsible for phosphorylation . Thus, we identified motifs as putative signaling modules added during specific evolutionary processes. We demonstrate that our method is a useful tool for analyzing and understanding intracellular signaling networks, thereby determining the regulatory mechanisms of cellular processes.
Extraction of phosphomotifs
Previous studies attempted to construct phosphomotifs for each kinase . However, these studies focused on less than half of the known kinases, and the phosphomotifs are still not available to facilitate a comprehensive analysis of phosphorylation. By contrast, the information related to phosphosites has been updated and an analysis based on bioinformatics was published recently. To investigate the evolution of phosphomotifs, we extracted phosphomotifs from publicly available databases for phosphosites. Phosphomotifs comprise the downstream and upstream regions that together form a phosphosite . We extracted three types of phosphomotifs, i.e., X-X-X-X-X-pS/T/Y, X-X-X-pS/T/Y-X-X-X, and pS/T/Y-X-X-X-X-X, from 97,679 phosphosites in PhosphoSitePlus . We calculated the similarity scores between them by adding the score in the BLOSUM62 substitution matrix at each sequence position, as described previously . The sequence pairs with a similarity score >9 were clustered using the MCL  to generate motifs that corresponded to three motif sequence patterns. Thus, we obtained 474 clusters and manually extracted 178 phosphomotifs by identifying the conserved amino acids in each cluster (Additional files 1 and 2). These motifs covered 77% (75550/97679) of the phosphosites used in the clustering analysis. They also included 57%, 51%, and 68% of all the serine, threonine, and tyrosine residues, respectively, obtained from all human proteins.
Comparative evolutionary analysis of the phosphomotifs
We investigated the evolutionary conservation of the phosphosites in each motif to elucidate the physiological roles of the identified motifs. To analyze the evolutionary conservation, we selected model organisms with rich genome information, i.e., Saccharomyces cerevisiae (fission yeast), Schizosaccharomyces pombe (budding yeast), Caenorhabditis elegans (worm), Drosophila melanogaster (fly), Danio rerio (fish), Canis familiaris (dog), Mus musculus (mouse), Pan troglodytes (Chimpanzee), and Homo sapiens (human). We obtained orthologous gene sets from KEGG OC  for the nine species. We constructed multiple sequence alignments of phosphoproteins in each orthologous group and identified species where a known phosphosite was conserved in a motif. We counted the number of proteins with phosphomotifs (Additional files 3-A and 4). In addition, we counted the conservation of the serine/threonine/tyrosine residues observed in all human proteins as potential phosphosites. The number of conserved residues between evolutionarily neighboring species increased to a maximum of only approximately 30% between the fish and dog in the case of serine/threonine residues, and approximately 32% between the worm and fish in the case of tyrosine residues (Additional file 5). Using the conservation rates of known and potential phosphosites, i.e., serine/threonine/tyrosine residues observed in all human proteins, we defined a conservation index to estimate the evolutionary conservation of phosphomotifs. The conservation index was calculated as the sum of the differences between the conservation rate of a known phosphomotif and that of the corresponding known and potential phosphomotifs in all species (Additional file 3-B). This conservation index represented the conservation of a motif relative to the average amino acid conservation.
The conservation of known phosphosites decreased slightly from humans to yeasts in a similar manner to the conservation of serine/threonine residues as if the motifs had not been exposed to evolutionary pressure. However, the motifs subjected to evolutionary pressure had a sigmoid-type pattern with dramatic changes among species (Figure 2B and C). This suggests that the phosphorylation of motifs with sigmoid patterns is essential after speciation and that they originated from functional differences between species and their ancestral species (Additional file 3). We extracted sigmoid-type motifs on the basis of a threshold value using the average rates of transitions and their standard deviations (Figure 2B), which identified motifs 55, 56, and 68 as those that increased after worm speciation; motifs 82, 93, 121, and 129 for the fly; motifs 46, 135, 140, 159, and 160 for the fish; motifs 95 and 74 for the dog; and motif 84 for the mouse (Figure 2C). To visualize the conservation relationship among known phosphosites and their motifs, we generated sequence logos . The sequence logos confirmed that evolutionarily conserved phosphosites also contained highly conserved flanking regions (Figure 2D and Additional file 6). In addition, we investigated the conservation of phosphosites and the proteins containing these sites (Additional file 7), which detected the conservation of phosphosites related or unrelated to protein conservation.
Protein networks of highly conserved motifs in humans and the worm
C2H2 zinc finger proteins from highly conserved motifs in humans and the fly
Four phosphomotifs that increased after fly speciation were related to a C2H2 motif; therefore, we investigated whether the evolutionary conservation of the phosphomotifs and all C2H2 motifs were correlated (Figures 4C and 2C). We observed that the number of C2H2 motifs increased after fish speciation, whereas the conservation of phosphomotifs increased after fly speciation, which suggests two different hypotheses: proteins that possess C2H2 motifs without phosphomotifs have increased selectively since fish speciation, and proteins with a C2H2 motif related to phosphorylation have been duplicated since fish speciation. To test these hypotheses, we investigated the frequencies of all the phosphomotifs. In the human genome, 3749, 1423, and 2507 phosphomotifs were observed in motifs 82, 93, and 129, respectively. The human genome contained a total of 6743 phosphomotifs and 56% of them were related to motif 82, which was considerably higher than the number of C2H2 motifs that increased in the fly. This suggests that proteins in the human genome with C2H2 motifs have been duplicated from the proteins that increased since fly speciation. The increase in C2H2 motifs in fish may reflect an increase in retroelements because the C2H2 motifs in zinc finger proteins have been evolutionarily duplicated by retroelements . Thus, we investigated the sequence differences between all C2H2 and phosphorylated C2H2 motifs. We extracted a sequence that started two amino acids upstream of the first cysteine until the last histidine as the overall C2H2 motif. The C2H2 motif sequences were classified into three groups depending on their lengths, i.e., 23, 24, and 25 amino acids. We observed that 94% of the C2H2 motifs belonged to these three groups, where 90% of these were 23-amino acid C2H2 motifs (Additional file 9A). These three C2H2 motifs differed from the 23-amino acid-type motif in terms of the presence of only one and two amino acid insertions between the histidines in the 24- and 25-amino acid-type C2H2 motifs, respectively, and there were no other amino acid changes (Additional file 9-B). We investigated the differences between all C2H2 motifs and the phosphorylated 23-amino acid C2H2 motifs. The 1st tyrosine was present in 70.3% of all the 23-amino acid C2H2 motifs, and the 24th threonine was present in 68.9% of the C2H2 motifs. Therefore, these two amino acids could be universal in C2H2 motifs.
C2H2 zinc finger motifs are fundamental functional motifs in species ranging from yeasts to humans. We generated sequence logos for all C2H2 motifs coded by organisms from yeasts to humans (Additional file 9-C). In the first sites of motifs 82 and 163, a fraction of the tyrosines had increased throughout evolution, although phenylalanines dominated in yeasts. It has been reported that a phosphorylated threonine in motifs 121 and 129 affects the interactions between the C2H2 motif and DNA sequences, thereby regulating the cell cycle [41, 42]. These previous reports indicate that comparative motif analysis is a powerful tool for extracting biologically important phosphomotifs. We focused on motif 82, for which the regulation of phosphorylation has not been reported. To investigate whether the phosphorylation of motif 82 is related to the functions of C2H2 motifs, we generated oligo sequences that mimicked C2H2 motifs, i.e., 2×C2H2-YF, where we replaced Y with F in motif 82, and 2×C2H2-SN, where we replaced S with N in motif 163, which had the smallest evolutionary effect as a control sequence, based on the ZNF24 sequence containing a phosphorylated C2H2 motif for motifs 82 and 163 (Figure 4D). 2×C2H2 labeled with mCFP was transfected into Cos cells, and 2×C2H2 localized to the nucleus and cytosol. 2×C2H2 was not likely to have localized to the nucleus because of the loss of the SCAN domain and NLS . As a control, we observed localization when fluorescent proteins were co-expressed. m1Venus-2×C2H2 was highly localized to the nucleolus compared with mCFP. We compared the localization to the nucleolus of 2×C2H2 and 2×C2H2-YF (Figure 4E). We observed that the YF mutant was localized to the nucleolus at a higher level compared with the wild type. This suggests that the phosphorylation of motif 82 in the wild type inhibits localization to the nucleolus and that the tyrosine in motif 82 regulates the localization of C2H2 motifs because the YF mutant exhibited strong localization to the nucleolus.
Regulatory pathways of RNA splicing and insulin signaling based on highly conserved motifs in humans and the fish
Addition of the signaling network of phosphomotifs with sigmoid conservation patterns to the core signaling network led to the acquisition of new cellular functions
The phosphomotifs that appeared after fish speciation were components of the signaling networks related to fundamental functions such as the cytoskeletal regulation, signaling, and splicing. These networks are common signaling networks, but it is considered that these functions have probably been conserved since before fish speciation. These results suggest that the motifs that have increased since fish speciation may be related to the expansion of the signaling networks. Thus, we investigated the conservation of all known phosphosites in our motifs using human proteins from these signaling networks.
We investigated whether the motifs with the sigmoid pattern are related to the expansion of signaling networks. The sigmoid motifs appeared since the worm; thus, we defined the core signaling network as a network that comprised proteins with phosphosites conserved from the yeast to humans. In addition, the network of proteins that interacted with the core signaling network was defined as the additional network that expanded during evolution after the yeast (Additional file 11). Using these networks, we extracted a network that comprised proteins with the sigmoid-type motifs (Figure 6C). Most of the proteins with sigmoid-type motifs appeared to be in the additional network. To validate this bias toward the additional network among the sigmoid-type motifs, we compared it with the random networks. As a result, we found that the proteins with the sigmoid-type phosphomotifs were significantly more likely to be located in the additional network (Figure 6D). In the fly, a small number of proteins with sigmoid -type phosphomotifs were present in the network because most of these proteins are functionally unknown zinc finger proteins and the interactions among them are still unknown. However, both of the proteins related to the fly were present in the additional network, and we could not find them in the core signaling network. These results suggest that the sigmoid-type phosphomotifs could be related to the expansion of the core signaling network.
Furthermore, to investigate the relationship between phosphomotifs and their biological functions, we enriched the phosphomotifs based on their GO biological functions. In general, motifs that increased after speciation events distant from humans were more likely to have correlated functions, whereas motifs that increased after the emergence of mammals lacked correlations with any functions (Additional file 10). These results suggest that the conserved patterns of the phosphomotifs are closed in networks with similar functions.
In this study, we validated the conservation of motifs and reconstructed pathways based only on known phosphosites. However, there may be differences in the evolutionary conservation of the motifs from all proteins and those that are already known to be phosphorylated. We used 1,771,245 serine/threonine/tyrosine residues in this study. However, we only used 25,631 residues with known phosphorylations in the comparative analysis, which comprised only 1.5% of all residues. To isolate more reliable networks, we performed an analysis that was limited to the known phosphosites, although the number was considerably small. Recently, the volume of phosphosite information has increased dramatically because of the development of phosphoproteomic techniques based on mass spectrometry. These large-scale datasets should allow us to isolate larger numbers of more reliable signal transduction networks.
In this study, we identified 178 phosphomotifs from publicly available phosphosite information and developed a method for detecting conserved phosphomotifs among species. We found that the highly conserved phosphomotifs in a specific species were related to important cellular processes. Our study suggests that the phosphorylation networks acquired during evolution have added signaling networks to the core signaling networks. Our method can be helpful not only for screening functionally important phosphosites but also for the classification of molecular networks added throughout the evolutionary process.
Definition of phosphomotifs
We downloaded the phosphomotifs defined in PhosphoSitePlus  and PhosphoELM [Remark 1]  on May 8, 2012. We extracted phosphomotifs validated in high- and low-throughput experiments, and they were manually combined to yield unique motifs. We also extracted known motifs described in previous studies  and added them to our phosphomotif dataset. Finally, we obtained 178 known phosphomotifs.
Conservation of motifs in species ranging from yeasts to humans
We downloaded nine genomes from KEGG  (May 8, 2012): Homo sapiens (hsa), Pan troglodytes (ptr), Mus musculus (mmu), Canis familiaris (cfa), Danio rerio (dre), Drosophila melanogaster (dme), Caenorhabditis elegans (cel), Schizosaccharomyces pombe (spo), and Saccharomyces cerevisiae (sce). The three-letter code for each genome is the species identifier defined by KEGG. Orthologous genes among these genomes were defined by KEGG OC . For each ortholog cluster, multiple sequence alignments were constructed by MAFFT , which is a freely available, rapid, and reliable tool compared with other alignment tools. MAFFT was run with the “--auto” option, which automatically selected the optimal options. We explored the sequence regions that matched exactly with the known phosphomotifs from all species. We investigated the species conservation with respect to these known and potential phosphorylated sites. In this study, potential phosphosites were defined as all serine/threonine/tyrosine residues in human proteins derived from known phosphosites stored in the databases.
We assumed that the order of evolution for these species corresponded to the general and universal tree, depending on the evolutionary distances from the viewpoint of humans [18, 19]. Therefore, the most basal organisms were the two yeasts, followed by the remaining organisms in the following order: nematodes, fly, fish, and mammals. If an amino acid residue in a protein at the same position as a known phosphorylated site in the orthologous protein in humans did not correspond to the amino acid residues S, T, and Y, we considered that the site was not conserved in the species. However, if the amino acid residue was conserved in a species that was evolutionarily distant from humans, the site was regarded as conserved, even if it was not conserved in the intermediate species. We created sequence logos of 11 residues using the WebLogo application , which included known and potential phosphorylated sites in the central positions of these conserved phosphorylation motifs. We also created sequence logos of the motif regions observed in each genome.
To compare the conservation of phosphosites, we calculated the conservation rates for the motifs in each species. The conservation rate was defined as the number of motifs conserved in a species divided by the number of motifs observed in the human genome. To confirm that the evolutionary history dramatically affected the conservation rate of sequence motifs, we extracted known motifs where the conservation rate changed >50% between two evolutionarily adjacent species. The conservation rate patterns across all the species were clustered using R (http://www.r-project.org/) based on the Euclidean distance and Ward’s method.
where G denotes the set of genomes used in the present study, q is each genome selected from G, and Cq and Rq are the conservation rate and the reference conservation rate in q, respectively.
We extracted GO  annotation for the proteins with known phosphorylated sites in the human genome. The annotations at the known motif level were assigned on the basis of the GO biological processes for proteins with the motif. To clarify the functions of the motif, we performed an enrichment analysis using GoMiner . Significant GO annotations were extracted with cutoffs of FDR = 0.01 and P < 0.01.
To explore the associations between proteins related to a motif, we included their interaction information. We downloaded information related to intermolecular interactions from BioGRID (2.0.58)  and STRING (v8.2) . We extracted the interactions of proteins in a known motif from this interaction information. We also compared the interaction networks generated for the motifs with randomly generated networks. We selected the same number of human proteins as those with a motif and extracted the interaction networks of these proteins. This randomization procedure was repeated 100 times. The fold change in a motif was calculated as the number of interactions in a motif divided by the average number of interactions in a randomly generated network. Network visualizations for our data were created using Cytoscape .
Zinc finger analysis
The HMMER programs with the default parameters were used to extract Pfam motifs that corresponded to zinc finger motifs from the genomes . We counted the number of zinc finger motifs in human proteins.
We counted the conservation levels of the known phosphosites included in our motifs. We also manually extracted human proteins related to the spliceosome, insulin signaling, and the cytoskeleton based on the KO definitions. We counted the conservation levels of these proteins in addition to the level of conservation in all human proteins as a control. We also extracted proteins related to complexes A, B, C, and common components of the spliceosome using the KEGG BRITE functional category and determined their conservation levels.
Proportion of proteins shared between motifs
We calculated the proportion of proteins shared between two known motifs. We extracted all human proteins with the known motif. The proportion was defined as the number of common proteins in two known motifs divided by the total number of proteins in the two motifs.
Network expansion of sigmoid-type phosphomotifs
We isolated 585 proteins that possessed phosphosites conserved from yeast (spe and sce) to humans. We defined the interaction network of these proteins as the core signaling network. In addition, we extracted the interactions with the proteins in the core signaling network from BioGRID and STRING, and obtained the additional network for 996 proteins. We extracted the interaction network for the proteins with sigmoid-type phosphomotifs (motifs 55, 56, and 58 for worm; motifs 82, 93, and 121 for fly; and motifs 46, 135, 140, 159 and 165 for fish). We also constructed random interaction networks using the same number of proteins with sigmoid-type phosphomotifs in each genome. This randomization procedure was repeated 100 times. We compared the randomization results and the real counts of proteins with sigmoid-type phosphomotifs in the core signaling network and the additional network.
cDNAs of 2×C2H2WT, 2×C2H2SN, and 2×C2H2SN were subcloned into pCXN2-mCFP and/or pCXN2-mVenus. pCXN2-mCFP and/or pCXN2-mVenus are expression vectors, which encode monomeric CFP and monomeric Venus, a YFP variant, respectively . The cDNAs of 2×C2H2WT, 2×C2H2SN, and 2×C2H2SN were synthesized by Operon Biotechnology Inc. The pCXN2 vector, which carries a neomycin resistance gene, is derived from pCAGGS.
The Cos7 cells used in this study were Cos7/E3, a subclone of Cos7 cells established by Y. Fukui. Cos7 cells were maintained in Dulbecco’s modified Eagle’s medium (Sigma, St Louis, MO, USA) supplemented with 10% fetal calf serum. In the transient expression studies, the cells were transfected using Polyfect (Qiagen). The cells were analyzed at 24 h after transfection.
Imaging of the C2H2 zinc finger motif in living cells
Live cell imaging was performed essentially as previously described. In brief, cells plated on a collagen-coated 35-mm-diameter glass base dish (Asahi Techno Glass Co., Tokyo, Japan) were transfected with C2H2 zinc finger motif expression vectors and imaged every 2 min using an Olympus IX81 inverted microscope (Olympus Optical Co., Tokyo, Japan), which was equipped with a cooled CCD camera, (CoolSNAP HQ; Roper Scientific, Trenton, NJ) and controlled by MetaMorph software (Universal Imaging, West Chester, PA). For the dual-emission ratio imaging of the m1Venus-2×C2H2 and m1CFP-2×C2H2 mutants, we used an excitation filter, i.e., 440AF21 for CFP and S492/18X for YFP, with a dichroic mirror, i.e., 86006bs, and emission filters, i.e., 480AF30 for CFP and 535AF26 for YFP (Omega Optical Inc., Brattleboro, VT). The cells were illuminated with a 75-W xenon lamp through a 12% ND filter (Olympus Optical) and visualized using a 40× oil immersion objective lens. After background subtraction, the ratio of the intensity of the nuclear region relative to the whole cell region was calculated using MetaMorph, which was used to represent the efficiency of the retention of C2H2 motifs in nuclear regions.
Markov cluster algorithm
Kyoto encyclopedia of genes and genomes
Search tool for recurring instances of neighbouring genes
Cyan fluorescent protein
Yellow Fluorescent protein.
We would like to thank Dr. Toshiya Hayano and Dr. Etsuko Kiyokawa for providing experimental support. This research was supported by a Grant-in-Aid for Young Scientists (B) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT), by an Uehara Memorial Foundation Fellowship, and by the Program for Research of Young Scientists from Ritsumeikan University.
- Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S: The protein kinase complement of the human genome. Science. 2002, 298 (5600): 1912-1934.PubMedView ArticleGoogle Scholar
- Gnad F, Ren S, Cox J, Olsen JV, Macek B, Oroshi M, Mann M: PHOSIDA (phosphorylation site database): management, structural and evolutionary investigation, and prediction of phosphosites. Genome biology. 2007, 8 (11): R250-PubMed CentralPubMedView ArticleGoogle Scholar
- Linding R, Jensen LJ, Pasculescu A, Olhovsky M, Colwill K, Bork P, Yaffe MB, Pawson T: NetworKIN: a resource for exploring cellular phosphorylation networks. Nucleic acids research. 2008, 36 (Database issue): D695-D699.PubMed CentralPubMedGoogle Scholar
- Gould CM, Diella F, Via A, Puntervoll P, Gemund C, Chabanis-Davidson S, Michael S, Sayadi A, Bryne JC, Chica C, Seiler M, Davey NE, Haslam N, Weatheritt RJ, Budd A, Hughes T, Pas J, Rychlewski L, Trave G, Aasland R, Helmer-Citterich M, Linding R, Gibson TJ: ELM: the status of the 2010 eukaryotic linear motif resource. Nucleic acids research. 2010, 38 (Database issue): D167-D180.PubMed CentralPubMedView ArticleGoogle Scholar
- Huang PH, White FM: Phosphoproteomics: unraveling the signaling web. Molecular cell. 2008, 31 (6): 777-781.PubMed CentralPubMedView ArticleGoogle Scholar
- Cox J, Mann M: Quantitative, high-resolution proteomics for data-driven systems biology. Annual review of biochemistry. 2011, 80: 273-299.PubMedView ArticleGoogle Scholar
- Linding R, Jensen LJ, Ostheimer GJ, van Vugt MA, Jorgensen C, Miron IM, Diella F, Colwill K, Taylor L, Elder K, Metalnikov P, Nguyen V, Pasculescu A, Jin J, Park JG, Samson LD, Woodgett JR, Russell RB, Bork P, Yaffe MB, Pawson T: Systematic discovery of in vivo phosphorylation networks. Cell. 2007, 129 (7): 1415-1426.PubMed CentralPubMedView ArticleGoogle Scholar
- Moses AM, Heriche JK, Durbin R: Clustering of phosphorylation site recognition motifs can be exploited to predict the targets of cyclin-dependent kinase. Genome biology. 2007, 8 (2): R23-PubMed CentralPubMedView ArticleGoogle Scholar
- Miller ML, Jensen LJ, Diella F, Jorgensen C, Tinti M, Li L, Hsiung M, Parker SA, Bordeaux J, Sicheritz-Ponten T, Olhovsky M, Pasculescu A, Alexander J, Knapp S, Blom N, Bork P, Li S, Cesareni G, Pawson T, Turk BE, Yaffe MB, Brunak S, Linding R: Linear motif atlas for phosphorylation-dependent signaling. Science signaling. 2008, 1 (35): ra2-PubMedView ArticleGoogle Scholar
- Tan CS, Linding R: Experimental and computational tools useful for (re)construction of dynamic kinase-substrate networks. Proteomics. 2009, 9 (23): 5233-5242.PubMedView ArticleGoogle Scholar
- Hornbeck PV, Kornhauser JM, Tkachev S, Zhang B, Skrzypek E, Murray B, Latham V, Sullivan M: PhosphoSitePlus: a comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucleic acids research. 2012, 40 (Database issue): D261-D270.PubMed CentralPubMedView ArticleGoogle Scholar
- Ideker T, Krogan NJ: Differential network biology. Molecular systems biology. 2012, 8: 565-PubMed CentralPubMedView ArticleGoogle Scholar
- Emes RD, Pocklington AJ, Anderson CN, Bayes A, Collins MO, Vickers CA, Croning MD, Malik BR, Choudhary JS, Armstrong JD, Grant SG: Evolutionary expansion and anatomical specialization of synapse proteome complexity. Nature neuroscience. 2008, 11 (7): 799-806.PubMed CentralPubMedView ArticleGoogle Scholar
- Pincus D, Letunic I, Bork P, Lim WA: Evolution of the phospho-tyrosine signaling machinery in premetazoan lineages. Proceedings of the National Academy of Sciences of the United States of America. 2008, 105 (28): 9680-9684.PubMed CentralPubMedView ArticleGoogle Scholar
- Holt LJ, Tuch BB, Villen J, Johnson AD, Gygi SP, Morgan DO: Global analysis of Cdk1 substrate phosphorylation sites provides insights into evolution. Science. 2009, 325 (5948): 1682-1686.PubMed CentralPubMedView ArticleGoogle Scholar
- Tan CS, Pasculescu A, Lim WA, Pawson T, Bader GD, Linding R: Positive selection of tyrosine loss in metazoan evolution. Science. 2009, 325 (5948): 1686-1688.PubMed CentralPubMedView ArticleGoogle Scholar
- Hunter T: Tyrosine phosphorylation: thirty years and counting. Current opinion in cell biology. 2009, 21 (2): 140-146.PubMed CentralPubMedView ArticleGoogle Scholar
- Hedges SB: The origin and evolution of model organisms. Nature reviews Genetics. 2002, 3 (11): 838-849.PubMedView ArticleGoogle Scholar
- Ciccarelli FD, Doerks T, von Mering C, Creevey CJ, Snel B, Bork P: Toward automatic reconstruction of a highly resolved tree of life. Science. 2006, 311 (5765): 1283-1287.PubMedView ArticleGoogle Scholar
- Tan CS, Bodenmiller B, Pasculescu A, Jovanovic M, Hengartner MO, Jorgensen C, Bader GD, Aebersold R, Pawson T, Linding R: Comparative analysis reveals conserved protein phosphorylation networks implicated in multiple diseases. Science signaling. 2009, 2 (81): ra39-PubMedView ArticleGoogle Scholar
- Bashor CJ, Horwitz AA, Peisajovich SG, Lim WA: Rewiring cells: synthetic biology as a tool to interrogate the organizational principles of living systems. Annual review of biophysics. 2010, 39: 515-537.PubMed CentralPubMedView ArticleGoogle Scholar
- Landry CR, Levy ED, Michnick SW: Weak functional constraints on phosphoproteomes. Trends in genetics: TIG. 2009, 25 (5): 193-197.PubMedView ArticleGoogle Scholar
- Khalil IG, Hill C: Systems biology for cancer. Current opinion in oncology. 2005, 17 (1): 44-48.PubMedView ArticleGoogle Scholar
- Yeh BJ, Rutigliano RJ, Deb A, Bar-Sagi D, Lim WA: Rewiring cellular morphology pathways with synthetic guanine nucleotide exchange factors. Nature. 2007, 447 (7144): 596-600.PubMedView ArticleGoogle Scholar
- Shou C, Bhardwaj N, Lam HY, Yan KK, Kim PM, Snyder M, Gerstein MB: Measuring the evolutionary rewiring of biological networks. PLoS computational biology. 2011, 7 (1): e1001050-PubMed CentralPubMedView ArticleGoogle Scholar
- Ubersax JA, Ferrell JE: Mechanisms of specificity in protein phosphorylation. Nature reviews Molecular cell biology. 2007, 8 (7): 530-541.PubMedView ArticleGoogle Scholar
- Ptacek J, Devgan G, Michaud G, Zhu H, Zhu X, Fasolo J, Guo H, Jona G, Breitkreutz A, Sopko R, McCartney RR, Schmidt MC, Rachidi N, Lee SJ, Mah AS, Meng L, Stark MJ, Stern DF, De Virgilio C, Tyers M, Andrews B, Gerstein M, Schweitzer B, Predki PF, Snyder M: Global analysis of protein phosphorylation in yeast. Nature. 2005, 438 (7068): 679-684.PubMedView ArticleGoogle Scholar
- Dongen S: Graph Clustering by Flow Simulation. 2000, Amsterdam, Netherlands: PhD thesis, University of UtrechtGoogle Scholar
- Nakaya A, Katayama T, Itoh M, Hiranuka K, Kawashima S, Moriya Y, Okuda S, Tanaka M, Tokimatsu T, Yamanishi Y, Yoshizawa AC, Kanehisa M, Goto S: KEGG OC: a large-scale automatic construction of taxonomy-based ortholog clusters. Nucleic acids research. 2013, 41 (Database issue): D353-D357.PubMed CentralPubMedView ArticleGoogle Scholar
- Chatr-Aryamontri A, Breitkreutz BJ, Heinicke S, Boucher L, Winter A, Stark C, Nixon J, Ramage L, Kolas N, O'Donnell L, Reguly T, Breitkreutz A, Sellam A, Chen D, Chang C, Rust J, Livstone M, Oughtred R, Dolinski K, Tyers M: The BioGRID interaction database: 2013 update. Nucleic acids research. 2013, 41 (Database issue): D816-D823.PubMed CentralPubMedView ArticleGoogle Scholar
- Szklarczyk D, Franceschini A, Kuhn M, Simonovic M, Roth A, Minguez P, Doerks T, Stark M, Muller J, Bork P, Jensen LJ, von Mering C: The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic acids research. 2011, 39 (Database issue): D561-D568.PubMed CentralPubMedView ArticleGoogle Scholar
- Crooks GE, Hon G, Chandonia JM, Brenner SE: WebLogo: a sequence logo generator. Genome research. 2004, 14 (6): 1188-1190.PubMed CentralPubMedView ArticleGoogle Scholar
- Frodin M, Antal TL, Dummler BA, Jensen CJ, Deak M, Gammeltoft S, Biondi RM: A phosphoserine/threonine-binding pocket in AGC kinases and PDK1 mediates activation by hydrophobic motif phosphorylation. The EMBO journal. 2002, 21 (20): 5396-5407.PubMed CentralPubMedView ArticleGoogle Scholar
- Inagaki M, Schmelzle T, Yamaguchi K, Irie K, Hall MN, Matsumoto K: PDK1 homologs activate the Pkc1-mitogen-activated protein kinase pathway in yeast. Molecular and cellular biology. 1999, 19 (12): 8344-8352.PubMed CentralPubMedView ArticleGoogle Scholar
- Rintelen F, Stocker H, Thomas G, Hafen E: PDK1 regulates growth through Akt and S6K in Drosophila. Proceedings of the National Academy of Sciences of the United States of America. 2001, 98 (26): 15020-15025.PubMed CentralPubMedView ArticleGoogle Scholar
- Lawlor MA, Mora A, Ashby PR, Williams MR, Murray-Tait V, Malone L, Prescott AR, Lucocq JM, Alessi DR: Essential role of PDK1 in regulating cell size and development in mice. The EMBO journal. 2002, 21 (14): 3728-3738.PubMed CentralPubMedView ArticleGoogle Scholar
- Wolfe SA, Nekludova L, Pabo CO: DNA recognition by Cys2His2 zinc finger proteins. Annual review of biophysics and biomolecular structure. 2000, 29: 183-212.PubMedView ArticleGoogle Scholar
- Letunic I, Goodstadt L, Dickens NJ, Doerks T, Schultz J, Mott R, Ciccarelli F, Copley RR, Ponting CP, Bork P: Recent improvements to the SMART domain-based sequence annotation resource. Nucleic acids research. 2002, 30 (1): 242-244.PubMed CentralPubMedView ArticleGoogle Scholar
- Nieto MA: The snail superfamily of zinc-finger transcription factors. Nature reviews Molecular cell biology. 2002, 3 (3): 155-166.PubMedView ArticleGoogle Scholar
- Thomas JH, Schneider S: Coevolution of retroelements and tandem zinc finger genes. Genome research. 2011, 21 (11): 1800-1812.PubMed CentralPubMedView ArticleGoogle Scholar
- Jantz D, Berg JM: Reduction in DNA-binding affinity of Cys2His2 zinc finger proteins by linker phosphorylation. Proceedings of the National Academy of Sciences of the United States of America. 2004, 101 (20): 7589-7593.PubMed CentralPubMedView ArticleGoogle Scholar
- Rizkallah R, Alexander KE, Hurt MM: Global mitotic phosphorylation of C2H2 zinc finger protein linker peptides. Cell Cycle. 2011, 10 (19): 3327-3336.PubMed CentralPubMedView ArticleGoogle Scholar
- Bailey SL, Chang SC, Griffiths B, Graham AN, Saffery R, Earle E, Choo KH, Kalitsis P: ZNF397, a new class of interphase to early prophase-specific, SCAN-zinc-finger, mammalian centromere protein. Chromosoma. 2008, 117 (4): 367-380.PubMedView ArticleGoogle Scholar
- Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, Sherlock G: Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nature genetics. 2000, 25 (1): 25-29.View ArticleGoogle Scholar
- Chen S, Synowsky S, Tinti M, MacKintosh C: The capture of phosphoproteins by 14-3-3 proteins mediates actions of insulin. Trends in endocrinology and metabolism: TEM. 2011, 22 (11): 429-436.PubMedView ArticleGoogle Scholar
- Johnson C, Tinti M, Wood NT, Campbell DG, Toth R, Dubois F, Geraghty KM, Wong BH, Brown LJ, Tyler J, Gernez A, Chen S, Synowsky S, MacKintosh C: Visualization and biochemical analyses of the emerging mammalian 14-3-3-phosphoproteome. Molecular & cellular proteomics: MCP. 2011, 10 (10): M110 005751-PubMed CentralView ArticleGoogle Scholar
- Apostolatos A, Song S, Acosta S, Peart M, Watson JE, Bickford P, Cooper DR, Patel NA: Insulin promotes neuronal survival via the alternatively spliced protein kinase CdeltaII isoform. The Journal of biological chemistry. 2012, 287 (12): 9299-9310.PubMed CentralPubMedView ArticleGoogle Scholar
- Wahl MC, Will CL, Luhrmann R: The spliceosome: design principles of a dynamic RNP machine. Cell. 2009, 136 (4): 701-718.PubMedView ArticleGoogle Scholar
- Ast G: How did alternative splicing evolve?. Nature reviews Genetics. 2004, 5 (10): 773-782.PubMedView ArticleGoogle Scholar
- Irimia M, Penny D, Roy SW: Coevolution of genomic intron number and splice sites. Trends in genetics : TIG. 2007, 23 (7): 321-325.PubMedView ArticleGoogle Scholar
- Dinkel H, Chica C, Via A, Gould CM, Jensen LJ, Gibson TJ, Diella F: Phospho.ELM: a database of phosphorylation sites–update 2011. Nucleic acids research. 2011, 39 (Database issue): D261-D267.PubMed CentralPubMedView ArticleGoogle Scholar
- Kanehisa M, Goto S, Sato Y, Furumichi M, Tanabe M: KEGG for integration and interpretation of large-scale molecular data sets. Nucleic acids research. 2012, 40 (Database issue): D109-D114.PubMed CentralPubMedView ArticleGoogle Scholar
- Katoh K, Standley DM: MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Molecular biology and evolution. 2013, 30 (4): 772-780.PubMed CentralPubMedView ArticleGoogle Scholar
- Zeeberg BR, Feng W, Wang G, Wang MD, Fojo AT, Sunshine M, Narasimhan S, Kane DW, Reinhold WC, Lababidi S, Bussey KJ, Riss J, Barrett JC, Weinstein JN: GoMiner: a resource for biological interpretation of genomic and proteomic data. Genome biology. 2003, 4 (4): R28-PubMed CentralPubMedView ArticleGoogle Scholar
- Smoot ME, Ono K, Ruscheinski J, Wang PL, Ideker T: Cytoscape 2.8: new features for data integration and network visualization. Bioinformatics. 2011, 27 (3): 431-432.PubMed CentralPubMedView ArticleGoogle Scholar
- Eddy SR: Accelerated Profile HMM Searches. PLoS computational biology. 2011, 7 (10): e1002195-PubMed CentralPubMedView ArticleGoogle Scholar
- Nagai T, Ibata K, Park ES, Kubota M, Mikoshiba K, Miyawaki A: A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications. Nature biotechnology. 2002, 20 (1): 87-90.PubMedView ArticleGoogle 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/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.