a systems biology approach for the investigation of the heparin
TRANSCRIPT
1
A SYSTEMS BIOLOGY APPROACH FOR THE INVESTIGATION OF THE HEPARIN/HEPARAN SULFATE INTERACTOME
Alessandro Ori1,2, Mark C. Wilkinson1§ and David G. Fernig1§* From 1Institute of Integrative Biology and Centre for Glycobiology, University of Liverpool,
Liverpool L69 7ZB, United Kingdom 2Current address: Structural and Computational Biology Unit,
European Molecular Biology Laboratory, 69117 Heidelberg, Germany §Equally contributing authors
Running title: A system-level analysis of the heparin/HS interactome *Corresponding author: Professor David G. Fernig, Biosciences Building, Crown Street, Liverpool L69 7ZB, United Kingdom. Email: [email protected]; Tel. +44 (0)151 795 4471; Fax +44 (0)151 795 4406
A large body of evidence supports the involvement of heparan sulfate (HS) proteoglycans in physiological processes such as development, and diseases including cancer and neurodegenerative disorders. The role of HS emerges from its ability to interact and regulate the activity of a vast number of extracellular proteins including growth factors and extracellular matrix components. A global view on how protein-HS interactions influence the extracellular proteome and, consequently, cell function is currently lacking. Here, we systematically investigate the functional and structural properties that characterize HS-interacting proteins and the network they form. We collected 435 human proteins interacting with HS, or the structurally-related heparin, by integrating literature-derived and affinity proteomics data. We used this dataset to identify the topological features that distinguish the heparin/HS-interacting network from the rest of the extracellular proteome and to analyze the enrichment of gene ontology terms, pathways and domain families in heparin/HS-binding proteins. Our analysis revealed that heparin/HS-binding proteins form a highly interconnected network, which is functionally linked to physiological and pathological processes that are characteristic of higher organisms. Therefore, we then investigated the existence of a correlation between the expansion of domain families characteristic of the heparin/HS interactome and the increase in biological complexity in the metazoan lineage. A strong positive correlation between the expansion of the heparin/HS interactome and biosynthetic machinery and organism complexity emerged. The evolutionary role of HS was reinforced by the presence of a rudimentary HS biosynthetic machinery in a unicellular organism at the root of the metazoan lineage.
A major challenge for the post-genomic era is to establish functional and structural relationships between the components of biological systems. In the last decade, the development of high throughput methods for the study of genetic interactions (1) and protein-protein interactions (2-4) has enabled the collection of large datasets describing binary relationships between primary gene products. The accumulation of these large datasets required innovative ways to represent and analyse molecular networks, thus stimulating the development of a new discipline known as network biology (5-8). This new approach has been successfully used to integrate data from different experimental platforms (9), infer properties of interaction networks by applying statistical theories (6), assign protein function (10), identify network signatures characteristic of diseases such as cancer (8,10) and investigate the evolution of interaction networks (11,12). However, the chemical complexity of secondary gene products, such as glycans and lipids, and the technical challenges associated with the study of their interactions has generated a gap in our current models of interaction networks, and, as a consequence, the interactions of proteins with secondary gene products such as glycosaminoglycans (GAGs) have been excluded from the above systematic analyses.
The GAGs are linear polysaccharides whose synthesis is not template-driven. As the most complex of biological polymers, they provide access to a vast chemical information space. This has been exploited in eumetazoans to provide structural frameworks and active mediation of cell-cell communication, both absolute requirements for multicellularity. The sulfated GAGs such as heparin/heparan sulfate (HS) are synthesized serine-linked to the core proteins of proteoglycans (HSPGs) and are located on the plasma membrane and in the extracellular matrix (ECM). The chemical complexity of heparin/HS arises from the
http://www.jbc.org/cgi/doi/10.1074/jbc.M111.228114The latest version is at JBC Papers in Press. Published on March 30, 2011 as Manuscript M111.228114
Copyright 2011 by The American Society for Biochemistry and Molecular Biology, Inc.
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
2
initially synthesized monotonous polymer being extensively modified (epimerisation, sulfation at various positions in the sugar rings). These modifications are sub-stoichiometric and grouped to produce characteristic domains, which vary in size and number within each chain (13,14). Analysis of functional structures in vivo demonstrates that there is specific regulation of the structures of heparin/HS that are expressed at the cellular level (15-18) and thus it seems that biology exploits a substantial amount of the chemical information space of heparin/HS. The functions of heparin/HS are exerted through their capacity to engage protein ligands. The consequences of these interactions range from elaborating large-scale structures to regulating the gradient formation and signaling activities of growth factors, cytokines and morphogens, and the localization and activity of extracellular enzymes (16,19,20) and reviewed in (17). The scope of these functions is evidenced by the size of the human heparin/HS interactome: 216 proteins in a review published in 2008 (17). Many pathogens express proteins that interact with heparin/HS as part of their molecular adaptation to infection of mammals (21). Thus, HSPGs are key players in molecular networks driving biological phenomena such as development (22), inflammation and immune response (23,24), and disease (21,25).
The first aim of this paper is to integrate and rationalize available data on heparin/HS-protein interactions. The current coverage of heparin/HS-protein interactions in public databases is largely incomplete [Table 1]. Therefore, a literature mining effort (17) was combined with an affinity proteomic approach for the identification of heparin/HS-binding proteins (HBPs) [Supplementary Results and Supplementary Figures 1 and 2], and data retrieved from public databases to generate a comprehensive list of the interactions between heparin/HS and proteins described so far. The term “HBP” is used because heparin is commonly used as an experimental proxy for the sulphated domains of HS and many interactions have not been validated with HS. This dataset then enabled a new systematic way of analysing heparin/HS-protein interactions using tools widely applied in genomic and proteomic studies. The system-level analysis allowed the investigation of how HBPs interact with each other, by computing the topological properties of the network they form and to identify functional and structural features that are associated with
the heparin/HS-binding activity. Finally, in order to generate insights into the role of HSPGs in the evolution of multicellular organisms, the presence of orthologs of HS biosynthetic enzymes in the genome of the choanoflagellate Monosiga brevicollis was investigated. Choanoflagellates are unicellular and colony-forming organisms found in marine and freshwater environments. They use a single apical flagellum surrounded by a collar of actin-filled microvilli to swim and capture bacterial prey (26). Since choanoflagellates are not metazoans and did not evolve from sponges or more recently derived metazoan phyla, they are indicated as the last unicellular organisms that evolved before the origin and diversification of metazoans (27). Previous works indicate the presence in the genome of M. brevicollis of protein families that were thought to be exclusive to multicellular organisms (26-28). The presence of functional signaling cascades based on tyrosine phosphorylation has been also demonstrated (28). For these reasons, the study of M. brevicollis is considered to be crucial for the identification of the molecular networks that were present in the last common ancestor of choanoflagellates and metazoans and that likely contributed to the emergence of multicellularity and the development of animals.
EXPERIMENTAL PROCEDURES
Construction of the heparin/HS interactome. The heparin/HS interactome was built using a combination of literature curation, data retrieval from public databases and experimental data obtained by the affinity proteomic approach described in Supplementary information. An initial version of the literature curated dataset was first published in (17) and originally included 216 HBPs. The original dataset was expanded and in its current version (December 2009) includes 280 interactors. Additional data were retrieved from publicly available databases and GO classifications using the search criteria listed in [Table 1]. For GO (29) and UniProtKB (30), the search was restricted to human genes, while this was not necessary for MatrixDB, since all the interactions described in it are associated by default to human protein identifiers (31). Finally, the 147 HBPs identified by the heparin-affinity proteomic approach described in Supplementary information were included as a set of experimentally derived, non-literature-based data. The integration of these
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
3
datasets resulted in a non-redundant list of 435 human proteins. This list, referred to as the heparin/HS interactome, was used for the all the subsequent analysis presented and it is available in the Supplementary File 1. Construction and analysis of the heparin/HS-interacting network. A protein-protein interaction network based on the heparin/HS interactome was built using Cytoscape v.2.6.3 (32,33). The protein-protein interaction resource was the Cytoscape Human Interactome Dataset (2007), which was obtained by merging molecular interaction data from a variety of sources including IntAct (34), DIP (35) and HPRD (36). The NCBI Entrez gene ID for each HBP was obtained from their UniProt accession number using DAVID 6.7 (37) and used to extract HBPs and their interactions from the human interactome dataset. Topological parameters of the heparin/HS-interacting network, treated as undirected network, were computed using the Cytoscape plugin Network Analyzer v.2.6 (38). For a context relevant analysis, the topological parameters of the extracellular human interactome were compared to the one of the network formed by extracellular HBPs. Thus, the extracellular interactome was extracted from the human dataset by applying filters based on GO cellular component terms. The terms used were: GO:0005576 (extracellular region), GO:0005615 (extracellular space), GO:0031012 (ECM) and GO:0005604 (basement membrane). Extracellular HBPs and their interactions were then selected from the extracellular proteome using the NCBI Entrez gene IDs from the heparin/HS interactome. The properties and topological parameters of the networks are summarized in Figure 1B. Functional and structural analysis of the heparin/HS interactome. The over-representation (enrichment) of GO terms (29), KEGG pathways (39) and Pfam domain families (40) in the heparin/HS interactome was analyzed using the web-accessible program DAVID 6.7 (37). The list of HBPs’ UniProt accession numbers was used as the input list and the default human proteome was used as the background list. The significance of the enrichments was statistically evaluated with a modified Fisher’s exact test (EASE score), and a p-value for each term was calculated by applying a Benjamini-Hochberg false discovery rate correction (37). Cut-off values were 0.01 for GO biological process terms enrichment, 0.05 for KEGG pathways and Pfam domains.
Furthermore, Pfam domains significantly enriched in the heparin/HS interactome were associated with the corresponding SCOP superfamilies (41) in order to reduce redundancy and perform the analysis described in Figure 4. For GO terms enrichments, the GO FAT annotation available in DAVID was used. The GO FAT is a subset of the GO term set created by filtering out the broadest ontology terms in order to do not overshadow more specific ones. The enrichment of GO biological process terms was also analyzed using the Cytoscape plugin BiNGO v2.3 (42), using the complete GO term set and a hypergeometric statistical test with Benjamini-Hochberg false discovery rate correction. Identification of HS biosynthetic enzymes in M. brevicollis. The orthologs of human biosynthetic enzymes responsible for synthesis of the protein-GAG linker tetrasaccharide, and for the polymerization and modification of HS chains were identified using a reciprocal Blast best-hit approach (43). Thus, the protein sequences of human enzymes were searched against the non-redundant protein sequences databases of M. brevicollis (taxid: 81824), Fungi (taxid: 4751), D. discoideum (taxid: 44689) and Plants (taxid: 3193), using Blastp with default settings. The best hits for each group were then searched against the human non-redundant protein sequences database (taxid: 9606) and selected as orthologs only in the case where the reciprocal best-hit criterion was satisfied. In the cases where more than one paralog enzyme was present in the human genome, a paralog was arbitrarily chosen from the family and used for the Blast search. Only for these cases, the reciprocal best-hit criterion was considered satisfied if the best hit of the reciprocal Blast search was a member of the same enzyme family, but not necessarily the paralog used for the first search.
RESULTS
Topological analysis of the heparin/HS-interacting network. The first step towards a systematic analysis of the heparin/HS interactome was to build and analyze the topological parameters of the network of protein-protein interaction formed by HBPs. This analysis is based on the representation of protein-protein interaction networks as graphs, where proteins are represented as nodes connected by edges that indicate the presence of
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
4
an interaction between them [Figure 1A]. By applying statistical tools used for graph theory, it is then possible to compute topological parameters describing the properties of the network and use such parameters to compare different networks (for reviews see (6,38,44)) [Figure 1B, C]. Thus, HBPs and their interactions were extracted from a dataset of human protein interactions obtained by merging data retrieved from different repositories (see Experimental Procedures). Since the predominant location of HSPGs is extracellular, the extracellular heparin/HS-interacting network was compared with the network formed by extracellular non-heparin/HS-interacting proteins and the total extracellular interactome [Figure 1A]. These networks were extracted from the human interactome dataset using filters based on gene ontology (GO) cellular component terms associated with extracellular protein localization (see Experimental Procedures). The topological parameters of the networks analyzed are summarized in Figure 1B. By comparing the properties of these three extracellular networks, the main topological parameter associated with the heparin/HS-binding function was the high average clustering coefficient displayed by the heparin/HS-interacting network [Figure 1B, C]. The clustering coefficient is a measure of the modularity of a network (i.e. the tendency of nodes to form groups or clusters). Therefore, a high average clustering coefficient indicates the presence of highly interconnected groups of nodes (modules) within the network. Clustering coefficients of the heparin/HS-interacting network distribute at higher values when compared with the non-heparin/HS-interacting network and the total extracellular interactome [Figure 1C]. The high average clustering coefficient of the extracellular heparin/HS-interacting network indicates a stronger tendency of HBPs to form highly interconnected modules than other extracellular proteins. In Figure 2, selected examples of highly clustered modules extracted from the heparin/HS interactome are shown. These examples indicate how the tendency to form highly connected modules is independent of the nature of the HBP. The clusters shown are in fact formed by secreted growth factors such as VEGFB and transforming growth factor beta TGFβ2 and their transmembrane receptors [Figure 2A, D], as well as by structural components of the ECM such as fibrillins [Figure 2B], and plasma
proteins such as coagulation factors [Figure 2C]. Furthermore, the architecture of the heparin/HS-interacting network has also functional implications. These clusters represent examples of functional modules responsible for the regulation of complex biological processes such as angiogenesis (45), morphogenesis (46), ECM assembly (47) and regulation of the coagulation cascade (48). These data support the view of HSPGs as key mediators of the assembly of molecular complexes at the cell surface and in the extracellular space (49).
Functional analysis of the heparin/HS interactome. To gain insights into functional roles of HBPs, the over-representation (enrichment) of ontology terms and components of molecular pathways in the heparin/HS interactome was analyzed in comparison to their occurrence in the human proteome. The GO biological process terms describe biological objectives to which the gene or gene product contributes (29) and they can be either broad, generic terms such as “response to stimulus” or a more specific one such as “fibroblast growth factor receptor signaling pathway”. Therefore, the identification of biological process terms associated with a particular set of genes, in this case the HBPs, can be useful to highlight their functional roles at a network level. Ninety-four % of the HBPs were annotated with at least one biological process term [Table 2 and Supplementary File 2]. The remaining 6 % was not annotated due to the incompleteness of the current GO annotation. From this analysis, a strong correlation between the heparin/HS interactome and biological functions characteristic of multicellular and higher organisms emerged. Significantly enriched terms are associated with fundamental processes common to all multicellular organisms, such as cell-cell signaling, but also with more complex processes such as wound healing and the immune response that are characteristic of higher organisms [Table 2]. As with other ontologies, the biological process terms can be visualized as a graph where directed links describe the hierarchy and relationships between terms (29). This kind of visualization helps to group highly related/redundant terms typical of ontology classifications and identify relevant functional modules. The graph shown in Figure 3 highlights the existence of four main functional groups strongly enriched in the heparin/HS interactome. These include two clusters of biological processes involved in the
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
5
control of the immune system and of developmental processes and other two clusters related to the regulation of cellular processes such as cell proliferation, and cell-cell signalling [Figure 3]. It has to be highlighted that the graph in Figure 3 is based for clarity only on the subsets of terms with the highest p-value, therefore, other significant, and perhaps less investigated, biological functions (such as, for example, “cation homeostasis”, p-value 2.1 E-16) were not included in this analysis.
Next, a similar approach to identify those pathways that have a statistically significant over-representation of HBPs was applied by projecting the heparin/HS interactome on the Kyoto encyclopedia of genes and genomes (KEGG) collection of pathways. KEGG pathways are manually drawn graphs that are based on the current knowledge of molecular interactions and reaction networks (39). In Table 3 the pathways enriched in HBPs are summarized. HBPs are involved in pathways responsible for the control of key physiological and pathological processes characteristic of multicellular organisms. In particular, the enriched pathways highlight the role of the heparin/HS interactome in key mechanisms implicated in the regulation of the cellular response to external stimuli. These include interactions between soluble ligands and their cell surface receptors (“Cytokine-cytokine receptor interaction”), as well as cross-talk with components of the ECM (“ECM-receptor interaction”). These mechanisms are directly linked to the control of cell behavior via the regulation of processes such as cytoskeleton reorganization (“Regulation of actin cytoskeleton”, “Focal adhesion”) and activation of intracellular signaling cascades (e.g., “TGF-beta signaling pathway”). The deregulation of these pathways can lead to the establishment of pathological conditions such as cancer (e.g., “Melanoma”) and immunological disorders (e.g., “Systemic lupus erythematosus”). Furthermore, pathways linked to other pathologies caused by structural alteration of extracellular proteins and accumulation of amyloid plaques (e.g., “Prion diseases”) are significantly correlated with the heparin/HS-binding activity, since most of these proteins directly interact with GAGs.
In summary, the functional analysis of the heparin/HS interactome highlighted the following: (i) HBPs are functionally enriched in biological processes that are characteristic of
multicellular and higher organisms; (ii) HBPs are candidates for a potential key role in mediating the information flow between the extracellular space and intracellular signaling pathways; (iii) this role could imply a direct involvement of the heparin/HS interactome in complex physiological and pathological systems, such as organismal development, inflammation, cancer and neurodegenerative disorders.
Structural analysis of the heparin/HS interactome. The same strategy used for the functional analysis of the heparin/HS interactome was performed at a structural level by investigating the enrichment of protein domains in HBPs. Ninety-eight % of the HBPs were annotated to at least one Pfam domain and the domain families significantly enriched in the heparin/HS interactome are listed in Table 4. Pfam domain families associated with heparin/HS-binding activity are typically extracellular and the majority of them appear to be characteristic of the metazoan lineage (65% of the top-20 most enriched families [Table 4]). Highlighting the structural diversity of the heparin/HS interactome, the list includes domains that are characteristic of small soluble, single-domain, proteins such as cytokines and growth factors (e.g., “Small cytokines, interleukin-8 like”, “Fibroblast growth factor”, “Transforming growth factor beta like domain”), domains that assemble in large multi-domain proteins (e.g., “Thrombospondin type 1 domain”, “Laminin G domain”, “Fibronectin type III domain”) and domains associated with enzymatic activity (e.g., “Trypsin family”, associated with proteolytic activity). The enrichment of some domain families is the result of the presence of families of HBP paralogs that expanded during evolution (e.g., the chemokine and FGF family). While in other cases, non-homologous proteins contribute to one domain family by the arrangement of the same structural unit in various multi-domain architectures (e.g., Thrombospondin type 1 domain). Furthermore, some domain families are enriched in the heparin/HS interactome without being directly responsible for the interaction with the carbohydrate (e.g., EGF-like domain), thanks to their co-occurrence with heparin/HS-binding domains in multi-domain proteins. Next, the Pfam domain families enriched in the heparin/HS interactome were mapped to the corresponding structural classification of proteins (SCOP) superfamilies. The SCOP classification uses structural information derived
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
6
from the PDB database instead of sequence alignments to establish evolutionary relationships between proteins and domains (41). Seventy-one % of the Pfam domain families significantly enriched in the heparin/HS interactome were mapped to a SCOP superfamily [see Supplementary File 4]. This approach allowed the reduction of the redundancy of the Pfam classification by grouping domains of related structure into 25 superfamilies (for example the entries “EGF-like domain" and “Laminin EGF-like" of Pfam are grouped in the SCOP superfamily “EGF/Laminin") [Table 5]. As mentioned above, not all the structures are directly associated with heparin/HS-binding activity. Literature curation of the heparin/HS-binding sites (HBSs) identified so far revealed that 11 of the 25 superfamilies have never been described as mediating the interaction with the carbohydrate [Table 5]. Even though the lack of evidence in current literature does not necessarily exclude a role of these structures in heparin/HS binding, this analysis highlights those structures that are predominantly mediating the interaction with GAGs. For each of these structures, a referenced example and, when available, a three-dimensional structure of the domain in complex with heparin or a heparin-related molecule are reported in Table 5.
In summary, the structural analysis of the heparin/HS interactome revealed a high level of diversity between structures associated with heparin/HS-binding function, possibly with a direct link to the structural complexity of GAGs. These structures are typically found in extracellular proteins and the majority of them are specific to metazoans. They can either occur in single-domain proteins, that are in general part of large families of paralogs, or as units mediating the heparin/HS-binding activity that have been arranged during evolution in different multi-domain architectures. Finally, a fraction of the structures enriched in the heparin/HS interactome does not seem to be directly involved in the interaction with the carbohydrate and they might be over-represented because they are functionally and structurally linked with the heparin/HS-interacting modules.
Evolutionary aspects of the heparin/HS interactome. Since a strong association emerged between HBPs and biological functions and pathways characteristic of higher organisms, the evolution of the protein domains enriched in the heparin/HS interactome was analysed in
correlation with organism complexity across the tree of life. In a recent publication, Vogel and Chothia investigated the expansion of domain superfamilies across the genomes of 38 uni- and multicellular eukaryotes and correlated it to the increase in organism complexity, as measured by the number of different cell types of which the organisms are composed (50). They mapped the occurrence of different superfamilies using a database of 1219 hidden Markov models, based on the superfamily classification of domains (51). For each genome, they annotated single-domain proteins and the individual domains of multi-domains proteins to their respective superfamily and then calculated the abundance of each superfamily, as the number of proteins that contain at least one domain belonging to that particular superfamily. The normalized abundance profiles were then used to calculate a Pearson correlation coefficient (PCC), R, describing the correlation between superfamily abundance and the estimated number of cell types per genome (50). The PCCs for the 27 superfamilies enriched in the heparin/HS interactome were extracted from the dataset of Vogel and Chothia and they are plotted in Figure 4. The distribution of PCCs of the heparin/HS interactome superfamilies indicates a strong correlation between their abundance and organism complexity (mean R = 0.83, with R ≥ 0.8 indicating a strong positive correlation). In order to establish a link between domain functions and their correlation with organism complexity, the authors extended the domain annotations by manually assigning functional categories to each superfamily (50). They observed that only 15% of the superfamilies have a strong positive correlation (R ≥ 0.8) with organism complexity. Moreover, just two functional categories contributed nearly half of these positively correlated superfamilies. These two functional categories are superfamilies associated with extracellular processes (20%) and regulation (29%) (50). Therefore, since most of the superfamilies characteristic of the heparin/HS interactome are also associated with extracellular processes, their relative contribution to the extracellular functional category was investigated. Thus, the heparin/HS interactome superfamilies were annotated using the functional classification described in (50) and the distribution of the correlation coefficients of extracellular superfamilies enriched in the heparin/HS interactome was compared to the one of superfamilies annotated
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
7
as extracellular, but not enriched in the heparin/HS interactome, and of the whole extracellular category [Figure 4]. The mean correlation coefficient of the heparin/HS interactome superfamilies annotated as extracellular is even higher than the whole heparin/HS interactome set (mean R = 0.89), and, most importantly, the distribution of their PCCs is significantly different from the one of the whole extracellular category (p = 9.2 E-3, one-sided Wilcoxon rank sum test) and other extracellular superfamilies (p = 1.9 E-3, one-sided Wilcoxon rank sum test).
Next, the analysis of the correlation between HSPGs and organism evolution was extended to their biosynthetic enzymes. In this case the analysis was not performed using the SCOP superfamily classification, since functional specificity cannot be entirely recapitulated by the structural annotation (i.e. structurally related enzymes can be grouped in the same superfamily despite having different substrate specificity). Only the HS co-polymerases EXT1 and EXT2 are currently annotated to a superfamily (“Nucleotide-diphospho-sugar transferases”, 53448), which also includes other sugar transferases such as galactosyltransferases. However, this superfamily also shows a positive correlation with organism complexity (R = 0.77). The occurrence of HS biosynthetic enzymes across the tree of life was, therefore, investigated by multiple sequence alignment across 638 sequenced genomes using STRING (52). The results are schematically summarized in Figure 5. HS biosynthetic enzymes appear to be characteristic of the eumetazoan lineage and, therefore, strongly associated with the emergence of multicellularity. No homology is detectable in unicellular eukaryotes such as fungi and in plants [Figure 5]. Interestingly, some level of homology is detected in certain species of bacteria. The occurrence of enzymes able to synthesise GAG-related carbohydrates has been already described in some vertebrate pathogens such as Streptococci (53). The emergence of these microbial enzymes may have occurred by either convergent evolution or horizontal gene transfer under the selective pressure of the vertebrate host defense, as a mechanism to enhance the pathogen virulence (53). Interestingly, some HS post-synthesis editing mechanisms appear to have evolved at a later stage than the core biosynthetic machinery. No homology is detectable for the extracellular
HS cleaving heparanases (HPSEs) in the nematode Caenorhabditis elegans, though homology, albeit low, is observed for the extracellular HS sulfatases (SULFs) [Figure 5]. The association between increase in organism complexity and HSPGs is further corroborated by the expansion of the HS biosynthetic enzymes, in particular in the vertebrate lineage. Thus, while C. elegans possesses only one isoform for each of the key enzymes involved in HS biosynthesis (EXT: HS co-polymerase; NDST: N-deacetylase / N-sulfotransferase; GLCE: glucuronic acid C5 epimerase; and the sulfotransferases HS2ST, HS6ST and HS3ST), in humans all these enzymes have expanded by gene duplication with the exception of GLCE and HS2ST. Two EXT, four NDST, three HS6ST and six HS3ST isoforms together with one isoform each for GLCE and HS2ST form the HS biosynthetic machinery in humans. A similar pattern is observed for the HS core proteins. Only one syndecan and two glypican genes are found in C. elegans, while humans possess four syndecan and six glypican isoforms (54). Interestingly, a similar expansion is not observed for the ECM core proteins perlecan and agrin (54).
In summary, a strong correlation emerged between the abundance in a genome of domain superfamilies associated with the heparin/HS interactome and organism complexity. This correlation was already observed for extracellular domains (50), but it is statistically more pronounced for the heparin/HS interactome superfamilies than others. A similar correlation is also suggested for HS biosynthetic enzymes and core proteins based on the occurrence of these proteins across the tree of life and their expansion during evolution, especially in the vertebrate lineage.
The choanoflagellate Monosiga brevicollis possesses a rudimental HS biosynthetic machinery. Since GAGs are commonly considered a hallmark of eumetazoans and the expansion of HS biosynthetic enzymes and interacting domains is associated with an increase in organism complexity, the presence of orthologs of HS biosynthetic enzymes in the genome of M. brevicollis was investigated. Thus, orthologs of human HS biosynthetic enzymes were established using a reciprocal Blast best-hit procedure (43) against the non-redundant protein sequences database of M. brevicollis. The Blast search was also performed against the protein
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
8
sequences databases of the nematode C. elegans, known to produce GAGs, of the colony forming slime mold Dictyostelium discoideum, and of Fungi and Plantae. The biosynthesis of HS and heparin requires the formation of a linker tetrasaccharide by sequential transfer of a xylose (Xyl), two galactose (Gal) and a glucuronic acid (GlcUA) unit to serine residues located in a serine-glycine repeat consensus on the core protein. This linker region is common to other GAGs such as chondroitin sulfate (CS) and dermatan sulfate (DS). High score orthologs of the four enzymes required for the assembly of the protein-GAG linker tetrasaccharide are present in the genome of M. brevicollis, while clear orthology can not be established for the full set in lower eukaryotes and plants [Figure 6]. The addition to the linker region of a N-acetyl glucosamine (GlcNAc) or a N-acetyl galactosamine (GalNAc) residue commits the biosynthesis towards heparin/HS or CS/DS, respectively (55). The biosynthetic reaction then proceeds with the polymerization of the sugar backbone by alternate addition of GlcUA and GlcNAc units, in the case of heparin and HS. In human, the addition of the first GlcNAc residue is perfomed by the enzyme exostosin-like 3 (EXTL3) that possesses only GlcNAc-transferase activity, while the sugar polymerization is carried out by the enzymes exostosins (EXTs) that present both GlcUA- and GlcNAc-transferase activities (56). The two activities can be localized in the EXT enzymes in two separate domains with the N-terminal domain (EXT(N)) responsible for GlcUA-transferase activity and the C-terminal domain (EXT(C)) responsible for the GlcNAc-transferase activity (57). The EXT enzymes are likely to be the result of a gene fusion event between two functionally interacting enzymes carrying the GlcUA- and GlcNAc-transferase activities (55). The more rudimentary biosynthetic machinery of the nematode C. elegans comprises just a single enzyme (rib-2) with GlcNAc transferase activity, responsible for both chain initiation and polymerization, and a separate enzyme (rib-1) that presents high sequence homology with EXT(N) and possesses GlcUA-transferase activity, but only when in complex with rib-2 (58) [Figure 6]. Similarly, in M. brevicollis two genes are present that possess homology with the C-terminal portion of EXTL3 and EXT(N) [Figure 6]. These two genes present conserved features that have been associated with GlcUA- and GlcNAc-transferase activity in
human genes, such as a conserved DXD motif in the C-terminal portion of EXTL3 (55,57) [Supplementary Figure 4], and they indicate the possibility of the production of heparosan-like chains by choanoflagellates. The heparin/HS polymerizing chains undergo a series of modification by four classes of sulfotranferases (NDST, HS2ST, HS6ST and HS3ST) that introduce sulfate groups at different positions on both the GlcUA and GluNAc units, and by an epimerase (GLCE) that converts some GlcUA residues to iduronic acid (IdoA) (55,59). These modifications largely contribute to generate sequence diversity in HS chains. M. brevicollis possesses two genes homologous to members of two of the four families of HS sulfotransferases: HS2ST and HS3ST [Figure 6]. Also in this case, the choanoflagellate genes present conservation of key residues involved in sulfotransferase catalytic activity, such as binding sites for the sulfate donor PAPS (3′ -phosphoadenosine 5′ -phosphosulfate) [Supplementary Figure 4]. These data suggest that M. brevicollis potentially possesses a complete, albeit rudimentary, HS biosynthetic machinery able to synthesize sulfated forms of HS-like polysaccharides. There is no evidence supporting the presence of a GLCE enzyme in M. brevicollis, suggesting that this kind of post-synthesis modification might have been a more recent innovation in the metazoan lineage.
In summary, the identification of orthologs of key GAG biosynthetic enzymes in the genome of a unicellular choanoflagellate suggests for the first time the presence of HS-like sulfated polysaccharides in nonmetazoans. Intriguingly, the presence of a rudimentary HS biosynthetic machinery at the root of the metazoan lineage, but not in other eukaryotes could support the role of GAGs as key molecules for the emergence of metazoan multicellularity, as it has been suggested for other cell signaling, adhesion and ECM molecules exclusively shared by M. brevicollis and metazoans (27,28). Further studies including the biochemical characterization of the enzymes identified in this study and isolation of GAGs from M. brevicollis cells are required to validate this hypothesis.
DISCUSSION
The analysis undertaken of the heparin/HS interactome allowed the investigation of the properties of the heparin/HS-interacting network
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
9
of protein-protein interaction and the analysis, at a system level, of the functional and structural features characterizing HBPs. The heparin/HS interactome was built combining literature mining, data retrieval from public databases and proteomic experimental data. The affinity proteomic strategy described in Supplementary information led to the identification of 147 extracellular proteins of which 32 were previously described HBPs [Supplementary File 1]. The remaining 115 newly discovered HBPs were also included for the analysis of the heparin/HS interactome although these interactions will require independent experimental validation. The resulting dataset included 435 human proteins, mostly extracellular. The analysis of the heparin/HS-interacting network revealed that HBPs tend to form more highly clustered modules than other extracellular proteins. These clusters can assemble both at the cell surface and in the ECM, and they often also represent functional modules. From a functional point of view, the heparin/HS interactome is strongly associated with biological processes characteristic of multicellular organisms and with pathways that are crucial for the conversion of extracellular cues into intracellular signaling events and, finally, into a phenotypic response. These processes and pathways are central to complex biological phenomena particular to higher organisms such as development and the immune response, and they are consequently linked to pathological conditions such as cancer and neurodegenerative disorders. In this perspective, potential intracellular roles of HSPGs could provide additional mechanisms for GAG-mediated regulation of intracellular signaling (60). The structural analysis of the heparin/HS interactome revealed the existence of two main categories of domains associated with heparin/HS-binding activity: domains that are characteristic of soluble single-domain proteins and domains that occur mainly in multi-domain proteins and that have been assembled during evolution in different architectures. From an evolutionary perspective, the expansion of domains associated with the heparin/HS interactome strongly correlates with an increase in organism complexity, independently of their nature. A similar correlation was already described for domains and proteins generically associated with extracellular processes (50,61), however, also within this set, the heparin/HS interactome-associated domains display a
statistically significant higher correlation than other extracellular domains. It has been shown that the evolutionary rate of the extracellular proteome is faster then the intracellular one, probably due to the less chemically constrained environment faced by extracellular proteins (62). This evolutionary plasticity of the extracellular proteome could have been a driving force for the organization of more complex systems of intercellular communication and organization. The functional and structural link between the heparin/HS interactome and biological processes characteristic of complex organisms and the fact that HBPs are more correlated than other extracellular proteins with an increase in organism complexity, strongly suggests a pivotal role of HSPGs in driving the evolution of multicellular and higher organisms. In fact, the expansion of enzymes and core proteins responsible for the synthesis and localization of HS chains also correlates with an increase in organism complexity in the metazoan lineage. The core HSPGs biosynthetic machinery was thought to have evolved in early eumetazoans concomitantly with the emergence of multicellularity (63). Biochemical data describing the presence of heparin/HS-like GAGs in phyla at the root of the eumetazoan lineage, such as Cnidaria and Ctenophora (64,65) supported this view. However, the presence of orthologs of key HS biosynthetic enzymes in the choanoflagellate M.brevicollis suggests that GAGs are likely to be a molecular innovation that predates the origin of metazoans. HSPGs might have been a critical step for the assembly of an extracellular network of proteins required for the structural organization of the extracellular space and for the establishment of cell-cell communication and cell differentiation. Later on, gene duplication events, in particular the whole genome duplications that occurred at the root of the vertebrate lineage (66), contributed to expand the repertoire of both HSPGs biosynthetic enzymes and their interacting partners on which evolution could act. The fact that domains characteristic of the heparin/HS interactome are strongly correlated with organism complexity could indicate that their expansion has been one of the driving forces towards the organization of the more sophisticated and tunable extracellular networks that are required for the development of higher organisms and for the control of organism-level biological processes, such as the establishment of an immune system. In support of this, others
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
10
have already proposed HSPGs as key molecules for the emergence of neural connectivity (54,67). The authors collected a series of experimental evidences obtained in different model organisms describing the specific involvement of HSPGs in all the key processes required for the establishment of neural connectivity including axon guidance, neuron-target interaction and synapse development (54). Similarly to what has been proposed here, the authors suggested a role for HSPGs as versatile extracellular scaffolds that modulate extracellular cues influencing the response of neurons to their environment (54). In the future, heparin-affinity proteomic strategies similar to the one described in Supplementary information could be implemented in combination with quantitative mass spectrometry techniques, such as targeted proteomics (68) and staple-isotope labelling with amino acids in cell culture (69), to investigate
dynamic changes of the heparin/HS interactome associated, for example, with different developmental stages or pathological conditions. Such data, complemented by the characterization of the dynamic changes in HS structure, could be extremely valuable to elucidate HSPGs-mediated mechanisms involved in the control of physiological and pathological processes. This opens the door to the design of new, network based, therapeutic strategies targeting multiple protein-glycan interactions that are associated with multifactorial diseases. Finally, the structural and functional characterization of the proteome and glycome of organisms at the root of the metazoan lineage could illuminate the role of protein and glycan co-evolution in the assembly of the extracellular molecular networks necessary for the development of complex forms of life.
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
11
REFERENCES
1. Tong, A. H., Evangelista, M., Parsons, A. B., Xu, H., Bader, G. D., Page, N., Robinson, M., Raghibizadeh, S., Hogue, C. W., Bussey, H., Andrews, B., Tyers, M., and Boone, C. (2001) Science 294, 2364-2368
2. Uetz, P., Giot, L., Cagney, G., Mansfield, T. A., Judson, R. S., Knight, J. R., Lockshon, D., Narayan, V., Srinivasan, M., Pochart, P., Qureshi-Emili, A., Li, Y., Godwin, B., Conover, D., Kalbfleisch, T., Vijayadamodar, G., Yang, M., Johnston, M., Fields, S., and Rothberg, J. M. (2000) Nature 403, 623-627
3. Gavin, A. C., Bosche, M., Krause, R., Grandi, P., Marzioch, M., Bauer, A., Schultz, J., Rick, J. M., Michon, A. M., Cruciat, C. M., Remor, M., Hofert, C., Schelder, M., Brajenovic, M., Ruffner, H., Merino, A., Klein, K., Hudak, M., Dickson, D., Rudi, T., Gnau, V., Bauch, A., Bastuck, S., Huhse, B., Leutwein, C., Heurtier, M. A., Copley, R. R., Edelmann, A., Querfurth, E., Rybin, V., Drewes, G., Raida, M., Bouwmeester, T., Bork, P., Seraphin, B., Kuster, B., Neubauer, G., and Superti-Furga, G. (2002) Nature 415, 141-147
4. Krogan, N. J., Cagney, G., Yu, H., Zhong, G., Guo, X., Ignatchenko, A., Li, J., Pu, S., Datta, N., Tikuisis, A. P., Punna, T., Peregrin-Alvarez, J. M., Shales, M., Zhang, X., Davey, M., Robinson, M. D., Paccanaro, A., Bray, J. E., Sheung, A., Beattie, B., Richards, D. P., Canadien, V., Lalev, A., Mena, F., Wong, P., Starostine, A., Canete, M. M., Vlasblom, J., Wu, S., Orsi, C., Collins, S. R., Chandran, S., Haw, R., Rilstone, J. J., Gandi, K., Thompson, N. J., Musso, G., St Onge, P., Ghanny, S., Lam, M. H., Butland, G., Altaf-Ul, A. M., Kanaya, S., Shilatifard, A., O'Shea, E., Weissman, J. S., Ingles, C. J., Hughes, T. R., Parkinson, J., Gerstein, M., Wodak, S. J., Emili, A., and Greenblatt, J. F. (2006) Nature 440, 637-643
5. Jeong, H., Mason, S. P., Barabasi, A. L., and Oltvai, Z. N. (2001) Nature 411, 41-42 6. Barabasi, A. L., and Oltvai, Z. N. (2004) Nat. Rev. Genet. 5, 101-113 7. Grove, C. A., De Masi, F., Barrasa, M. I., Newburger, D. E., Alkema, M. J., Bulyk, M. L.,
and Walhout, A. J. (2009) Cell 138, 314-327 8. Taylor, I. W., Linding, R., Warde-Farley, D., Liu, Y., Pesquita, C., Faria, D., Bull, S.,
Pawson, T., Morris, Q., and Wrana, J. L. (2009) Nat. Biotechnol. 27, 199-204 9. Ideker, T., Thorsson, V., Ranish, J. A., Christmas, R., Buhler, J., Eng, J. K., Bumgarner, R.,
Goodlett, D. R., Aebersold, R., and Hood, L. (2001) Science 292, 929-934 10. Bandyopadhyay, S., Sharan, R., and Ideker, T. (2006) Genome Res. 16, 428-435 11. Fraser, H. B. (2006) Curr. Opin. Genet. Dev. 16, 637-644 12. Beltrao, P., and Serrano, L. (2007) PLoS Comp. Biol. 3, e25 13. Turnbull, J. E., and Gallagher, J. T. (1991) Biochem. J. 273 553-559 14. Murphy, K. J., Merry, C. L. R., Lyon, M., Thompson, J. E., Roberts, I. S., and Gallagher, J. T.
(2004) J. Biol. Chem. 279, 27239-27245 15. Allen, B. L., and Rapraeger, A. C. (2003) J. Cell Biol. 163, 637-648 16. Bornemann, D. J., Park, S., Phin, S., and Warrior, R. (2008) Development 135, 1039-1047 17. Ori, A., Wilkinson, M. C., and Fernig, D. G. (2008) Front. Biosci. 13, 4309-4338 18. Thompson, S. M., Jesudason, E. C., Turnbull, J. E., and Fernig, D. G. (2010) Birth Defects
Res. C. Embryo Today Rev. 90, 32-44 19. Vyas, N., Goswami, D., Manonmani, A., Sharma, P., Ranganath, H. A., VijayRaghavan, K.,
Shashidhara, L. S., Sowdhamini, R., and Mayor, S. (2008) Cell 133, 1214-1227 20. Yu, S. R., Burkhardt, M., Nowak, M., Ries, J., Petrasek, Z., Scholpp, S., Schwille, P., and
Brand, M. (2009) Nature 461, 533-536 21. Chen, Y., Gotte, M., Liu, J., and Park, P. W. (2008) Mol. Cells 26, 415-426 22. Lin, X. (2004) Development 131, 6009-6021 23. Parish, C. R. (2006) Nat. Rev. Immunol. 6, 633-643 24. Handel, T. M., Johnson, Z., Crown, S. E., Lau, E. K., and Proudfoot, A. E. (2005) Annu. Rev.
Biochem. 74, 385-410 25. Fuster, M. M., and Esko, J. D. (2005) Nat. Rev. Cancer 5, 526-542 26. Abedin, M., and King, N. (2008) Science 319, 946-948 27. King, N., Westbrook, M. J., Young, S. L., Kuo, A., Abedin, M., Chapman, J., Fairclough, S.,
Hellsten, U., Isogai, Y., Letunic, I., Marr, M., Pincus, D., Putnam, N., Rokas, A., Wright, K.
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
12
J., Zuzow, R., Dirks, W., Good, M., Goodstein, D., Lemons, D., Li, W., Lyons, J. B., Morris, A., Nichols, S., Richter, D. J., Salamov, A., Sequencing, J. G., Bork, P., Lim, W. A., Manning, G., Miller, W. T., McGinnis, W., Shapiro, H., Tjian, R., Grigoriev, I. V., and Rokhsar, D. (2008) Nature 451, 783-788
28. King, N., Hittinger, C. T., and Carroll, S. B. (2003) Science 301, 361-363 29. Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M., Davis, A. P.,
Dolinski, K., Dwight, S. S., Eppig, J. T., Harris, M. A., Hill, D. P., Issel-Tarver, L., Kasarskis, A., Lewis, S., Matese, J. C., Richardson, J. E., Ringwald, M., Rubin, G. M., and Sherlock, G. (2000) Nat. Genet. 25, 25-29
30. UniProt-Consortium. (2009) Nucleic Acids Res. 38, D142-148 31. Chautard, E., Ballut, L., Thierry-Mieg, N., and Ricard-Blum, S. (2009) Bioinformatics 25,
690-691 32. Shannon, P., Markiel, A., Ozier, O., Baliga, N. S., Wang, J. T., Ramage, D., Amin, N.,
Schwikowski, B., and Ideker, T. (2003) Genome Res. 13, 2498-2504 33. Cline, M. S., Smoot, M., Cerami, E., Kuchinsky, A., Landys, N., Workman, C., Christmas,
R., Avila-Campilo, I., Creech, M., Gross, B., Hanspers, K., Isserlin, R., Kelley, R., Killcoyne, S., Lotia, S., Maere, S., Morris, J., Ono, K., Pavlovic, V., Pico, A. R., Vailaya, A., Wang, P. L., Adler, A., Conklin, B. R., Hood, L., Kuiper, M., Sander, C., Schmulevich, I., Schwikowski, B., Warner, G. J., Ideker, T., and Bader, G. D. (2007) Nature Protocols 2, 2366-2382
34. Kerrien, S., Alam-Faruque, Y., Aranda, B., Bancarz, I., Bridge, A., Derow, C., Dimmer, E., Feuermann, M., Friedrichsen, A., Huntley, R., Kohler, C., Khadake, J., Leroy, C., Liban, A., Lieftink, C., Montecchi-Palazzi, L., Orchard, S., Risse, J., Robbe, K., Roechert, B., Thorneycroft, D., Zhang, Y., Apweiler, R., and Hermjakob, H. (2007) Nucleic Acids Res. 35, D561-565
35. Salwinski, L., Miller, C. S., Smith, A. J., Pettit, F. K., Bowie, J. U., and Eisenberg, D. (2004) Nucleic Acids Res. 32, D449-451
36. Keshava Prasad, T. S., Goel, R., Kandasamy, K., Keerthikumar, S., Kumar, S., Mathivanan, S., Telikicherla, D., Raju, R., Shafreen, B., Venugopal, A., Balakrishnan, L., Marimuthu, A., Banerjee, S., Somanathan, D. S., Sebastian, A., Rani, S., Ray, S., Harrys Kishore, C. J., Kanth, S., Ahmed, M., Kashyap, M. K., Mohmood, R., Ramachandra, Y. L., Krishna, V., Rahiman, B. A., Mohan, S., Ranganathan, P., Ramabadran, S., Chaerkady, R., and Pandey, A. (2009) Nucleic Acids Res. 37, D767-772
37. Huang, D. W., Sherman, B. T., and Lempicki, R. A. (2009) Nature Protocols 4, 44-57 38. Assenov, Y., Ramirez, F., Schelhorn, S. E., Lengauer, T., and Albrecht, M. (2008)
Bioinformatics 24, 282-284 39. Kanehisa, M., and Goto, S. (2000) Nucleic Acids Res. 28, 27-30 40. Finn, R. D., Tate, J., Mistry, J., Coggill, P. C., Sammut, S. J., Hotz, H. R., Ceric, G., Forslund,
K., Eddy, S. R., Sonnhammer, E. L., and Bateman, A. (2008) Nucleic Acids Res. 36, D281-288
41. Murzin, A. G., Brenner, S. E., Hubbard, T., and Chothia, C. (1995) J. Mol. Biol. 247, 536-540 42. Maere, S., Heymans, K., and Kuiper, M. (2005) Bioinformatics 21, 3448-3449 43. Koonin, E. V. (2005) Annu. Rev. Genet. 39, 309-338 44. Tucker, C. L., Gera, J. F., and Uetz, P. (2001) Trends Cell Biol 11, 102-106 45. Uniewicz, K. A., and Fernig, D. G. (2008) Front. Biosci. 13, 4339-4360 46. Belenkaya, T. Y., Han, C., Yan, D., Opoka, R. J., Khodoun, M., Liu, H., and Lin, X. (2004)
Cell 119, 231-244 47. Ritty, T. M., Broekelmann, T. J., Werneck, C. C., and Mecham, R. P. (2003) Biochem. J. 375,
425-432 48. Yu, H., Munoz, E. M., Edens, R. E., and Linhardt, R. J. (2005) Biochim. Biophys. Acta, Gen.
Subj. 1726, 168-176 49. Turnbull, J., Powell, A., and Guimond, S. (2001) Trends Cell Biol. 11, 75-82 50. Vogel, C., and Chothia, C. (2006) PLoS Comp. Biol. 2, e48 51. Gough, J., Karplus, K., Hughey, R., and Chothia, C. (2001) J. Mol. Biol. 313, 903-919
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
13
52. Jensen, L. J., Kuhn, M., Stark, M., Chaffron, S., Creevey, C., Muller, J., Doerks, T., Julien, P., Roth, A., Simonovic, M., Bork, P., and von Mering, C. (2009) Nucleic Acids Res. 37, D412-416
53. DeAngelis, P. L. (2002) Anat. Rec. 268, 317-326 54. Vactor, D. V., Wall, D. P., and Johnson, K. G. (2006) Curr. Opin. Neurobiol. 16, 40-51 55. Esko, J. D., and Selleck, S. B. (2002) Annu. Rev. Biochem. 71, 435-471 56. Kim, B. T., Kitagawa, H., Tamura, J., Saito, T., Kusche-Gullberg, M., Lindahl, U., and
Sugahara, K. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 7176-7181 57. Wei, G., Bai, X., Gabb, M. M., Bame, K. J., Koshy, T. I., Spear, P. G., and Esko, J. D. (2000)
J. Biol. Chem. 275, 27733-27740 58. Kitagawa, H., Izumikawa, T., Mizuguchi, S., Dejima, K., Nomura, K. H., Egusa, N.,
Taniguchi, F., Tamura, J., Gengyo-Ando, K., Mitani, S., Nomura, K., and Sugahara, K. (2007) J. Biol. Chem. 282, 8533-8544
59. Kusche-Gullberg, M., and Kjellén, L. (2003) Curr. Opin. Struct. Biol. 13, 605-611 60. Chen, L., and Sanderson, R. D. (2009) PLoS One 4, e4947 61. Huxley-Jones, J., Pinney, J. W., Archer, J., Robertson, D. L., and Boot-Handford, R. P.
(2009) Int. J. Exp. Pathol. 90, 95-100 62. Julenius, K., and Pedersen, A. G. (2006) Mol. Biol. Evol. 23, 2039-2048 63. Freilich, S., Goldovsky, L., Ouzounis, C. A., and Thornton, J. M. (2008) BMC Evol. Biol. 8,
247 64. Medeiros, G. F., Mendes, A., Castro, R. A., Bau, E. C., Nader, H. B., and Dietrich, C. P.
(2000) Biochim. Biophys. Acta, Gen. Subj. 1475, 287-294 65. Yamada, S., Morimoto, H., Fujisawa, T., and Sugahara, K. (2007) Glycobiology 17, 886-894 66. Panopoulou, G., and Poustka, A. J. (2005) Trends Genet. 21, 559-567 67. Lee, J. S., and Chien, C. B. (2004) Nat. Rev. Genet. 5, 923-935 68. Domon, B., and Aebersold, R. (2010) Nat. Biotechnol. 28, 710-721 69. Ong, S. E., Blagoev, B., Kratchmarova, I., Kristensen, D. B., Steen, H., Pandey, A., and
Mann, M. (2002) Mol. Cell. Proteomics 1, 376-386 70. Ori, A., Free, P., Courty, J., Wilkinson, M. C., and Fernig, D. G. (2009) Mol. Cell.
Proteomics 8, 2256-2265 71. Kuno, K., and Matsushima, K. (1998) J. Biol. Chem. 273, 13912-13917 72. Shen, L., Villoutreix, B. O., and Dahlback, B. (1999) Thromb. Haemostasis 82, 72-79 73. Ersdal-Badju, E., Lu, A., Zuo, Y., Picard, V., and Bock, S. C. (1997) J. Biol. Chem. 272,
19393-19400 74. Lyon, M., Rushton, G., and Gallagher, J. T. (1997) J. Biol. Chem. 272, 18000-18006 75. Ingham, K. C., Brew, S. A., and Atha, D. H. (1990) Biochem. J. 272, 605-611 76. Blom, A. M., Kask, L., and Dahlback, B. (2001) J. Biol. Chem. 276, 27136-27144 77. Kappler, J., Franken, S., Junghans, U., Hoffmann, R., Linke, T., Muller, H. W., and Koch, K.
W. (2000) Biochem. Biophys. Res. Commun. 271, 287-291 78. Rastegar-Lari, G., Villoutreix, B. O., Ribba, A. S., Legendre, P., Meyer, D., and Baruch, D.
(2002) Biochemistry 41, 6668-6678 79. Kuang, Z., Yao, S., Keizer, D. W., Wang, C. C., Bach, L. A., Forbes, B. E., Wallace, J. C.,
and Norton, R. S. (2006) J. Mol. Biol. 364, 690-704 80. Stephens, R. W., Bokman, A. M., Myohanen, H. T., Reisberg, T., Tapiovaara, H., Pedersen,
N., Grondahl-Hansen, J., Llinas, M., and Vaheri, A. (1992) Biochemistry 31, 7572-7579 81. Sachchidanand, Lequin, O., Staunton, D., Mulloy, B., Forster, M. J., Yoshida, K., and
Campbell, I. D. (2002) J. Biol. Chem. 277, 50629-50635 82. Shao, C., Zhang, F., Kemp, M. M., Linhardt, R. J., Waisman, D. M., Head, J. F., and Seaton,
B. A. (2006) J. Biol. Chem. 281, 31689-31695 83. Mine, S., Yamazaki, T., Miyata, T., Hara, S., and Kato, H. (2002) Biochemistry 41, 78-85 84. Clarris, H. J., Cappai, R., Heffernan, D., Beyreuther, K., Masters, C. L., and Small, D. H.
(1997) J. Neurochem. 68, 1164-1172
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
14
FOOTNOTES
We thank Dr. Martin Beck and the EMBL Proteomic Core Facility for their help with mass spectrometry data acquisition and analysis, Dr. Olga Vasieva and Dr. Krzysztof Wicher for many helpful discussions and bioinformatic support, and Hassanul Choudhury for technical assistance. The authors are supported by a European Commission Marie Curie Early Stage Training Programme (MolFun) (AO), the Cancer and Polio Research Fund Laboratories and the North West Cancer Research Fund (DGF).
The abbreviations used are: B3GA3: galactosylgalactosylxylosylprotein 3-beta-glucuronosyltransferase 3; B3GT6: beta-1,3-galactosyltransferase 6; B4GT7: xylosylprotein 4-beta-galactosyltransferase 7; CS: chondrotin sulfate; DS: dermatan sulfate; ECM: extracellular matrix; EXT: exostosin / heparan sulfate polymerase; EXTL3: exostosin-like 3 / glucosamine transferase; GAG: glycosaminoglycan; Gal: galactose; GalNAc: N-acetyl galactosamine; GlcUA: glucuronic acid; GlcNAc: N-acetyl glucosamine; GLCE: glucuronic acid C5 epimerase; GO: gene ontology; HBP: heparin-binding protein; HBS: heparin-binding site; HPSE: heparanase; HS: heparan sulfate; HS2ST: heparan sulfate 2-O sulfotranferase; HS3ST: heparan sulfate 3-O sulfotranferase; HS6ST: heparan sulfate 6-O sulfotranferase; HSPG: heparan sulfate proteoglycan; KEGG: Kyoto encyclopedia of genes and genomes; NDST: heparan sulfate N-deacetylase / N-sulfotransferase; PCC: Pearson correlation coefficient; SCOP: structural classification of proteins; SULF: extracellular heparan sulfate sulfatase; Xyl: xylose; XYLT1: xylosyltransferase 1.
FIGURE LEGENDS
Figure 1. Topological and functional analysis of the heparin/HS-interacting network. (A) The extracellular heparin/HS-interacting network (“Ec_hepint”, blue) was extracted from a dataset of extracellular protein-protein interactions, and it was compared with the non-heparin/HS-interacting network (“Ec_not-hepint”, red) and the whole extracellular interactome (“Ec”, green) (see Experimental Procedures). The position of the nodes in the networks as well as the length of edges are arbitrary and have only a graphical purpose. The properties and topological parameters of the network analyzed are summarized in (B). “Proteins” indicate the number of proteins (nodes) that form each network and “PPI” (protein-protein interactions) the number of interactions (edges) connecting them. The “Average degree” indicates the mean number of neighbors per node in the network. The “Characteristic path length” is the average over the shorter distances (number of links) separating all pairs of nodes in the network and offers a measure of a network's overall navigability. The clustering coefficient is defined as the number of links connecting the first neighbors of a given node divided by the total possible number of connections between them. It is a measure of the tendency of nodes to form highly interconnected modules. The “Average clustering coefficient” for a network is calculated as the mean of the clustering coefficients for each node having a degree ≥ 2. The “Ec_hepint-random” network was generated from the extracellular heparin/HS-interacting network by applying a degree preserving random shuffle of the edges (1320 shuffles). The “Ec_random” network was generated by randomly selecting a network of the same size as the extracellular heparin/HS-interacting network from the total extracellular interactome. The procedure was iterated 50 times and mean network parameters are shown with standard deviation in brackets. The average clustering coefficient of the extracellular heparin/HS-interacting network is six standard deviations higher than the average value calculated for randomly picked networks. In (C) nodes are binned according to their degree and the average clustering coefficient for each bin is plotted applying the same color-code used in A. The distribution of the clustering coefficients is characterized by the typical slope of protein-protein interaction networks, which indicates the presence of hierarchical modularity.
Figure 2. Examples of highly clustered modules of the heparin/HS-interacting network. HBPs with a high clustering coefficient were extracted together with their first neighbors from the extracellular heparin/HS-interacting network. Node color indicates the clustering coefficient of each node, according to the legend. Node label indicates the UniProt short name of the HBP. Protein-protein interactions are represented as green edges. The highest clustering coefficients represented were 1.0 for VEGFB and vascular endothelial growth factor receptor-1 (FLT1) (A), 0.67 for fibrillin-2 (FBN2)
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
15
(B), 0.60 for coagulation factor-IX (F9) (C) and 0.40 for TGFβ2 (D). The graphs were generated using Cytoscape v.2.6.3 (32).
Figure 3. The GO biological process terms enriched in the heparin/HS interactome are shown as nodes connected by directed edges that indicate hierarchies and relationships between terms. Node size is proportional to the number of HBPs belonging to the functional category. Node color indicates the corrected p-value (Benjamini-Hochberg false discovery rate correction) for the enrichment of the term according to the legend. For clarity, only highly significant terms are displayed (p < 1 E-21). The graphs were generated using Cytoscape v.2.6.3 (32) and its plugin BiNGO v2.3 (42).
Figure 4. Correlation between the abundance of superfamilies associated with the heparin/HS interactome and organism complexity. The Pearson Correlation Coefficients (PCCs) describing the association between superfamily abundance and organism complexity were extracted from (50). The PCCs for superfamilies: enriched in the heparin/HS interactome (“Hepint”), enriched in the heparin/HS interactome and annotated as extracellular in (50) (“Ec_hepint”), annotated as extracellular (“Ec”) and annotated as extracellular, but not enriched in the heparin/HS interactome (“Ec_not-hepint”) are plotted along the y axis. For each group a horizontal bar indicates the mean PCC. The dashed line indicates the threshold for strong correlation between superfamily abundance and organism complexity.
Figure 5. The occurrence of heparin/HS biosynthetic enzymes across the tree of life. The occurrence of heparin/HS biosynthetic enzymes across 630 organisms was analyzed using STRING 8.2 (52). The UniProt Accession numbers of the human HS biosynthetic enzymes were used as input. The conservation of each gene across different species is indicated by squares colored according to the sequence homology detected by STRING. For clarity, some species (e.g., Bacteria) were grouped in collapsed nodes, colored in grey, and the number indicated between brackets reports the number of species grouped in each node. In these cases, split squares report the highest and lowest score for the given gene within the grouped species.
Figure 6. The choanoflagellate Monosiga brevicollis possesses rudimentary HS biosynthetic machinery. Orthologs of the human HS biosynthetic enzymes were established by applying the reciprocal Blast best-hit criterion (43) against the non-redundant protein sequences databases of C.elegans, M. brevicollis, Fungi, D. discoideum and Plantae. The figure reports the Blastp score of the best hits only in the cases when the reciprocal best-hit criterion was satisfied. The “Linker” enzymes are responsible for the synthesis of the protein-GAG linker tetrasaccharide, which is shared between HS and other GAGs. The “HS” group includes enzymes that are specific for the HS-specific biosynthetic pathway that follow the formation of the linker tetrasaccharide. The full report of the Blastp search including M. brevicollis hits is provided in Supplementary File 6.
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
16
TABLES
Table 1. Current coverage of human HBPs in publicly available databases.
Source Search criteria Output Ref. GO consortium (April 2010) GO:0008201 heparin binding 109 human genes (29) GO:0043395 heparan sulfate binding 4 human genes MatrixDB (April 2010) heparin+heparan sulfate 90 human entries (31) UniProtKB (Release 2010_04) KW-0358 heparin-binding 66 human genes (30) Literature based review 216 human genes (17)
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
17
Table 2. GO biological process terms enriched in the heparin/HS interactome. “Count” indicates the number of HBPs and “%” the percentage of the mapped proteins associated to each term. For clarity only the twenty most significant terms are listed. A complete list can be found in Supplementary File 2.
Term Name Count % Corrected p-value
GO:0009611 Response to wounding 120 27.8 6.6E-69 GO:0042330 Taxis 55 12.8 2.6E-39 GO:0006935 Chemotaxis 55 12.8 2.6E-39 GO:0006954 Inflammatory response 73 16.9 3.4E-39 GO:0006952 Defense response 91 21.1 8.1E-35 GO:0007626 Locomotory behavior 62 14.4 3.3E-33 GO:0006955 Immune response 91 21.1 5.7E-31 GO:0042060 Wound healing 51 11.8 1.5E-30 GO:0016477 Cell migration 57 13.2 3.1E-28 GO:0007610 Behavior 71 16.5 3.7E-27 GO:0051674 Localization of cell 58 13.5 9.7E-27 GO:0048870 Cell motility 58 13.5 9.7E-27 GO:0042127 Regulation of cell proliferation 90 20.9 4.0E-26 GO:0006928 Cell motion 70 16.2 4.0E-26 GO:0032101 Regulation of response to external stimulus 43 10.0 8.7E-26 GO:0001568 Blood vessel development 51 11.8 2.4E-25 GO:0001944 Vasculature development 51 11.8 7.3E-25 GO:0051605 Protein maturation by peptide bond cleavage 33 7.7 1.2E-24 GO:0007267 Cell-cell signaling 76 17.6 1.8E-24 GO:0016485 Protein processing 36 8.4 3.9E-24
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
18
Table 3. KEGG pathways enriched in the heparin/HS interactome. “Count” indicates the number of HBPs and “%” the percentage of the mapped proteins associated to each term.
Term Name Count % Corrected p-value
hsa04610 Complement and coagulation cascades 42 9.7 1.4E-33 hsa04060 Cytokine-cytokine receptor interaction 63 14.6 7.3E-24 hsa04512 ECM-receptor interaction 35 8.1 7.6E-21 hsa04510 Focal adhesion 43 10.0 9.3E-14 hsa05200 Pathways in cancer 52 12.1 1.7E-11 hsa05218 Melanoma 22 5.1 7.8E-10 hsa04062 Chemokine signaling pathway 34 7.9 7.9E-09 hsa05020 Prion diseases 15 3.5 1.3E-08 hsa04810 Regulation of actin cytoskeleton 33 7.7 8.7E-07 hsa04350 TGF-beta signaling pathway 18 4.2 2.6E-05 hsa04672 Intestinal immune network for IgA production 13 3.0 6.8E-05 hsa05322 Systemic lupus erythematosus 18 4.2 1.4E-04 hsa04010 MAPK signaling pathway 30 7.0 1.4E-03 hsa04640 Hematopoietic cell lineage 14 3.2 4.5E-03 hsa04621 NOD-like receptor signaling pathway 11 2.6 1.0E-02 hsa05219 Bladder cancer 9 2.1 1.1E-02 hsa05310 Asthma 7 1.6 2.5E-02 hsa05222 Small cell lung cancer 12 2.8 3.0E-02
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
19
Table 4. Pfam domain families enriched in the heparin/HS interactome. “Count” indicates the number of HBPs associated to each domain family. a Domain family also present in the choanoflagellate M. brevicollis. b Domain family also present in bacteria, but not in non-metazoan eukaryotes. For clarity only the twenty most significant Pfam families are listed. A complete list can be found in Supplementary File 4.
Term Name Count Corrected p-value
Metazoan specific
PF00048 Small cytokines, interleukin-8 like 32 1.23E-38 √ PF00167 Fibroblast growth factor 19 5.08E-22 √ PF01391 Collagen triple helix repeat 24 1.09E-16 PF00090 Thrombospondin type 1 domain 21 1.17E-14 PF00008 EGF-like domain 22 1.03E-09 PF00089 Subtilase family 21 1.29E-09 PF01410 Fibrillar collagen C-terminal domain 8 5.98E-08 √a
PF00079 Serpin 12 2.14E-07 PF02210 Laminin G domain 12 3.49E-07 PF00019 Transforming growth factor beta like domain 11 1.15E-06 √ PF00093 von Willebrand factor type C domain 11 2.44E-06 √ PF00052 Laminin B (Domain IV) 6 2.27E-05 √b PF00053 Laminin EGF-like 10 2.88E-05 PF00084 Sushi domain 11 5.07E-05 PF02412 Thrombospondin type 3 repeat 5 7.16E-05 √b PF05735 Thrombospondin C-terminal region 5 7.16E-05 √b PF06008 Laminin Domain I 5 7.16E-05 √b PF06009 Laminin Domain II 5 7.16E-05 √b PF00219 Insulin-like growth factor binding protein 7 2.60E-04 √ PF00688 TGF-beta propeptide 7 3.27E-04 √b
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
20
Table 5. Superfamilies associated with the Pfam domain families enriched in the heparin/HS interactome. Not all the superfamilies are directly associated with heparin-binding activity. In the cases where an HBS was located in a domain belonging to a superfamily, the columns “HBS - UniProt AC” and “HBS - Ref.” report the UniProt accession number of the HBP and the publication describing the interaction, respectively. The column “PDB” reports, when available, the PDB number of a three-dimensional structure including a domain belonging to a superfamily in complex with heparin or a heparin derivative.
S.family Name HBS UniProt AC Ref. PDB
54117 Interleukin 8-like chemokines P02776 (70) 1U4L 50353 Cytokine P09038 (70) 1AXM 82895 TSP-1 type 1 repeat Q9UHI8 (71) 57196 EGF/Laminin 50494 Trypsin-like serine proteases P04070 (72) 1A7S 56574 Serpins P01008 (73) 1AZX 57501 Cystine-knot cytokines P01137 (74) 57603 FnI-like domain P02751 (75) 57535 Complement control module/SCR domain P04003 (76)
103647 TSP type-3 repeat 49899 Concanavalin A-like lectins/glucanases 57184 Growth factor receptor domain 50242 TIMP-like O95631 (77) 53300 vWA-like P04275 (78) 57610 Thyroglobulin type-1 domain P18065 (79) 57440 Kringle-like P00749 (80) 1GMN 49265 Fibronectin type III P02751 (81) 47874 Annexin P07355 (82) 2HYU, 2HYV 57424 LDL receptor-like module 49410 Alpha-macroglobulin receptor domain 48239 Terpenoid cyclases 57362 BPTI-like P10646 (83) 56491 A heparin-binding domain P05067 (84) 50755 Phosphotyrosine-binding domain 57630 GLA-domain 69179 Integrin domains 58010 Fibrinogen coiled-coil and central regions
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
21
FIGURES
Figure 1
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from
Alessandro Ori, Mark C. Wilkinson and David G. Ferniginteractome
A systems biology approach for the investigation of the heparin/heparan sulfate
published online March 30, 2011J. Biol. Chem.
10.1074/jbc.M111.228114Access the most updated version of this article at doi:
Alerts:
When a correction for this article is posted•
When this article is cited•
to choose from all of JBC's e-mail alertsClick here
Supplemental material:
http://www.jbc.org/content/suppl/2011/03/30/M111.228114.DC1
by guest on April 13, 2018
http://ww
w.jbc.org/
Dow
nloaded from