The study of gene and protein interaction networks has improved our understanding of the multiple, systemic levels of regulation found in eukaryotic and prokaryotic organisms. Here we carry out a large-scale analysis of the protein-protein interaction (PPI) network of fission yeast (Schizosaccharomyces pombe) and establish a method to identify ‘linker’ proteins that bridge diverse cellular processes - integrating Gene Ontology and PPI data with network theory measures. We test the method on a highly characterized subset of the genome consisting of proteins controlling the cell cycle, cell polarity and cytokinesis and identify proteins likely to play a key role in controlling the temporal changes in the localization of the polarity machinery. Experimental inspection of one such factor, the polarity-regulating RNB protein Sts5, confirms the prediction that it has a cell cycle dependent regulation. Detailed bibliographic inspection of other predicted ‘linkers’ also confirms the predictive power of the method. As the method is robust to network perturbations and can successfully predict linker proteins, it provides a powerful tool to study the interplay between different cellular processes.
Analysis of protein interaction networks has been of use as a means to grapple with the complexity of the interactome of biological organisms. So far, network based approaches have only been used in a limited number of organisms due to the lack of high-throughput experiments. In this study, we investigate by graph theoretical network analysis approaches the protein-protein interaction network of fission yeast, and present a new network measure, linkerity, that predicts the ability of certain proteins to function as bridges between diverse cellular processes. We apply this linkerity measure to a highly conserved and coupled subset of the fission yeast network, consisting of the proteins that regulate cell cycle, polarized cell growth, and cell division. In depth literature analysis confirms that several proteins identified as linkers of cell polarity regulation are indeed also associated with cell cycle and/or cell division control. Similarly, experimental testing confirms that a mostly uncharacterized polarity regulator identified by the method as an important linker is regulated by the cell cycle, as predicted.
Citation: Vaggi F, Dodgson J, Bajpai A, Chessel A, Jordán F, et al. (2012) Linkers of Cell Polarity and Cell Cycle Regulation in the Fission Yeast Protein Interaction Network. PLoS Comput Biol 8(10): e1002732. doi:10.1371/journal.pcbi.1002732
Editor: Robert B. Russell, University of Heidelberg, Germany
Received: May 1, 2012; Accepted: August 21, 2012; Published: October 18, 2012
Copyright: © Vaggi et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors gratefully acknowledge support from a Human Frontier Science Program (HFSP) Young Investigator Grant (R.E.C.-S., A.C.-N., M.S.; HFSP RGY0066/2009-C), an European Research Council (ERC) Starting Researcher Investigator Grant (R.E.C.-S.; SYSGRO), a Grant-in-Aid for Young Scientists from the Japan Society for the Promotion of Science (JSPS) (M.S.) and A.C.-N., A.B., F.J. are funded by the Italian Research Fund FIRB (RBPR0523C3). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The eukaryotic cell cycle is one of the most important and evolutionary conserved processes of cells , . The cell cycle integrates signals from multiple pathways to control tissue growth and homeostasis in multicellular organisms, as well as reproduction and proliferation in single cell organisms . To ensure cell integrity, the cell cycle regulates and is regulated by other key processes such as DNA replication, cytokinesis and cell growth –. Disruption of the regulation between the cell cycle and other cellular processes can cause a myriad of cellular pathologies including defects in cell shape, abnormal cell growth and aneuploidy, potentially leading to cancer .
With the accumulation of data from high-throughput biology as well as the generalisation of manually curated online databases, we now can mine existing biological networks to make experimentally verifiable predictions about system-wide properties of genes and gene products. In this work, we present a new method to search for proteins that serve as linkers between distinct functional sub-networks. Because of the well-characterized interactions between the cell cycle and other processes in the fission yeast Schizosaccharomyces pombe, we focus our analysis on this organism, where these processes have not yet been investigated yet by protein interaction network analysis methods.
The fission yeast - a rod-shaped unicellular eukaryote - is ideally suited to study the relationship between cell cycle and cell polarity regulation, as its highly polarized growth pattern is tightly correlated with cell cycle progression , . After cytokinesis, newborn S. pombe cells resume growth in G1 in a monopolar fashion from their ‘old end’ - the cell end that existed prior to division - and later in early G2 activate growth at their ‘new end’ derived from the site of septation, an event termed new-end take-off or NETO . Bipolar growth then continues through G2 until cells reach a critical size, after which cells enter M phase again. At that point cells stop growing , mitosis takes place and each cell divides by growing a septum in its middle. Daughter cells resume their cyclic pattern of growth at the ends and division at the middle, a pattern that relies on the cytoskeleton of actin and microtubules and on diverse polarity-regulating proteins (‘polarity factors’). Cytokinesis, polarity, and the cell cycle have been extensively studied in fission yeast –using both experiments and mathematical modelling –. The insights gained from studies in fission yeast often carry over to higher eukaryotes, as the molecular machinery controlling those processes has been highly conserved throughout evolution , , .
Several proteins have been identified that play important roles connecting these processes in fission yeast. For example, the polarized growth-regulating DYRK kinase Pom1  was recently shown to form a spatial gradient that is used by the cell cycle machinery to sense the length of the cell , , . Another link was observed between the morphogenesis-related NDR kinase network (MOR) and the septation initiation network (SIN) . MOR is important for the localization of actin patches to sites of polarized growth, while SIN is responsible for triggering cytokinesis. It was discovered that SIN inhibits the MOR pathway, through inhibition of the Orb6 activator Nak1. MOR itself also interferes with SIN, and this antagonism is required for proper progression through the cell cycle , . Furthermore, a similar antagonism between the MOR and SIN pathways has also been observed in higher eukaryotes , . The NETO transition from monopolar to bipolar growth and the switch from polarized growth to actin ring-mediated cell septation are also controlled by the cell cycle , thus the cell cycle machinery enforces a major control on both polarized growth and cytokinesis. Although many polarity or cytokinesis regulators contain potential phosphorylation sites for the cell cycle-regulating Cyclin-Dependent Kinases  (CDK), the molecular details of these couplings are not well known. In the other direction, if either polarized cell growth or cytokinesis is inhibited, both can send signals to stop the cell cycle , , further underlining that these three functional modules are highly interlinked.
To tackle the interplay between different cellular processes, we utilized a network theory approach. Hitherto, network based approaches have only been used in a limited number of organisms, due to the paucity of genome-wide interaction data available for most species. Recently, however, improvements in automatic experimental annotation, literature mining , machine learning  and orthology annotations , are allowing the use of network approaches in a wider range of organisms. For example, ‘meta databases’ such as STRING , , benchmark information from multiple sources and provide for each possible interaction a confidence score that reflects the likelihood of a set of proteins of actually interacting. Here, we take advantage of such developments and build on the efforts of the fission yeast community in annotating protein functions –, to establish a new method to identify proteins linking diverse cellular processes, based on integrating Gene Ontology (GO) ,  and Protein-Protein Interaction (PPI) data together with network theory based measures. Network-based approaches in biology have been used in the past to identify community structures, study lethality, identify specific regulatory circuits and study hierarchical organization . In particular, the nature of large scale protein-protein interaction networks has recently been under considerable debate with different groups disagreeing about the modularity of networks, as well as the properties of the nodes responsible for bringing together different modules –. In this work, we sidestep the difficult problem of identifying hierarchical modules in a large, genome-wide network and focus instead on a method to identify proteins that link different cellular processes. To do this, we use the highly characterized sub-genomic network consisting of proteins regulating the cell cycle, cytokinesis, and polarized cell growth in fission yeast. We propose a new network measure, termed ‘linkerity’, and use it to predict a novel role for a number of proteins as key bridges between these biological processes.
Constructing and validating the fission yeast protein interaction network
We constructed the fission yeast protein-interaction network using data from STRING ,  and BioGRID . By applying a cutoff on the confidence score from STRING, we can reject interaction pairs for which there is a limited amount of evidence (see Materials and Methods for details on data in STRING) and use the remaining edges to construct a non-directed and non-weighted network.
We then examined the effects of increasing the cutoff in STRING confidence scores in both the genome-wide interaction dataset of fission yeast and that of the better characterized budding yeast Saccharomyces cerevisiae on the network topology. Increasing the cutoff decreased the amount of nodes (Figure 1A) and the edge density (Figure 1B) in the largest component (the connected component in the network containing the highest number of edges and nodes) of both the fission and budding yeast networks (Tables S1, S2). This decrease was less sharp in budding yeast compared to fission yeast due to the extensive amount of genome-wide interaction experiments carried out in the former, increasing the amount of high-confidence interactions. Interestingly, in the ‘core’ sub-network consisting of proteins involved in cell cycle regulation, polarity and cytokinesis (Figure 2 for fission yeast and Figure S1 for budding yeast), the drop off in the number of nodes and edges was far less significant in both yeasts, suggesting that interaction data for the core fission yeast network tends to be more reliable than interaction data for the rest of the network (Figure 1, red stars versus red dots, also Tables S1, S2, S3, S4). As a more stringent test, we constructed networks for both organisms using only data from BioGRID . BioGRID is a database that only contains data from manually annotated experiments (distinguishing between experiments that show direct physical interaction and genetic interactions). Networks built using the BioGRID physical interaction data also show that the core networks of fission yeast and budding yeast are relatively dense, while the fission yeast organism-wide network is rather sparse (Figure 1). Even with the relatively high coverage of the core (regulation of cell cycle, cytokinesis, polarity) network in fission yeast, it is important to note that fission yeast lacks any genome-wide protein-protein interaction experiments, and as such, several of the interactions predicted by STRING are based on indirect evidence such as genetic interactions, inference from homology, or literature mining , .
Figure 1. Dependence of network measures on protein-protein interaction data quality.
As we increase the minimal accepted confidence (cutoff) for the PPI data of the STRING database, the number of nodes in the largest connected component (A) and the network density (B) both decrease for all networks. This decrease is faster in fission yeast compared to budding yeast, and faster in the full organism network compared to the core network. Triangles overlaid on each curve show the same network measures for the PPI network based on the BioGRID database, the position on the x-axis of BioGRID data is calculated using linear interpolation to estimate the corresponding cutoff in STRING which would give a similarly-sized network, thus the overlay of the BioGRID data gives an indication how this relates to different cutoff STRING data. As can be seen from the figure panels the fission yeast core network is quite robust to cutoff changes and behaves similarly to the core network of budding yeast cells. This is also true for the core networks based on BioGRID data.doi:10.1371/journal.pcbi.1002732.g001
Figure 2. The cell cycle + cytokinesis + polarity = core interaction network of fission yeast proteins.
(A) Venn diagram showing the overlap among the different Gene Ontology functional groups for the proteins belonging to the core network. Proteins with multiple functional annotations have colours that are the sum of the colours of the individual functional annotations; proteins belonging to all three functional groups are in white. (B) Protein-protein interactions inside the fission yeast core network (from the STRING database at cutoff 0.7). Node colours are the same as in panel A. Node size is proportional to the degree of each protein, and node order within a category (clockwise) is also determined by degree. 165 black edges link proteins that do not share functional annotations, while 1869 grey edges link proteins that have at least one common GO annotation (thus white nodes have only grey links). White nodes (nodes belonging to all categories) are shown in the inner circle in the middle of the network.doi:10.1371/journal.pcbi.1002732.g002
As no analysis of the fission yeast network has been previously published, we performed a few checks to verify that our network construction procedure was giving sensible results, and that the data for fission yeast available in STRING was of sufficiently high quality. As a first check, we sought to replicate a number of analyses previously performed with budding yeast (Table 1). At a cutoff of 0.7 (defined by STRING as a ‘high confidence’ threshold), the genome-wide fission yeast network has 2770 nodes with at least one connection and 20432 edges compared to 5477 nodes and 105429 edges found in budding yeast, although they have approximately similar number of proteins. We calculated the degree distribution for the nodes in the network, and observed that, as previously described for numerous other complex networks , the fission yeast PPI network has a scale-free distribution (Figure S2). We also repeated analyses done in numerous other studies examining the relationship between network measures and gene deletion lethality –. As reported for budding yeast, we observed that degree (the number of interactions with other proteins) is the best predictor network measure of gene deletion-induced lethality in fission yeast, and that ratio of these essential genes among hubs (the top 20% of proteins by degree) is even higher in fission yeast than in budding yeast (see Text S1).
Table 1. Network statistics and gene essentiality comparison between the two yeasts.doi:10.1371/journal.pcbi.1002732.t001
Since there is no high-throughput genome-wide interaction data available for fission yeast, we tested the possibility that highly investigated proteins might have more interactions. To check this, we tested to see whether the number of abstracts in PubMed discussing a particular protein was correlated with the degree of that protein in the network. The Pearson correlation between the number of PubMed abstracts citing a protein and its degree in the network was 0.13 for budding yeast (p-value<10−19) and 0.14 for fission yeast (p-value<10−13) (details in Tables S1, S2), suggesting there is no fission yeast-specific bias for proteins with large amounts of publications in STRING networks. However, a large amount of evidence for the fission yeast interactions in STRING is obtained from homology, and specifically from interactions of homologues proteins in budding yeast. As essential genes are more likely to be conserved ,  and STRING is more likely to identify homology between highly conserved genes, it is possible that this might introduce a subtle bias making essential genes appear to be more highly connected in virtue of their higher conservation. This is consistent with the observation that a very high percentage of hubs in fission yeast appear to be essential (Text S1).
The core network of regulators of the cell cycle, cell polarity and cytokinesis
The sub-network of all proteins regulating cell cycle, cytokinesis and polarized growth, henceforth, the ‘core’ network (see Materials and Methods for definitions of exact GO terms used) in fission yeast contains 550 proteins: 384 of those are associated with regulation of cell cycle, 155 with cytokinesis and 139 with polarity. Using a cutoff of 0.7 in STRING, 429 of the total 550 proteins are connected to the largest connected component of the core network. Most of the proteins not in the network have no known interactions, and the second largest connected component contains only 4 proteins, thus we focus only on the interaction network of the largest connected component. There are a high number of proteins with multiple functions in the network (Figure 2A), 16 of them (Alp4, Cdc15, Gsk3, Lsk1, Mor2, Orb6, Pab1, Pmo25, Pom1, Ppb1, Ras1, Scd1, Shk1, Sid2, Tea1, Wsp1) are important for all three cellular processes and 77 have dual functions. The ratio of multifunctional proteins is quite similar to the ratio in the analogous core budding yeast network (Figure S1). Interestingly the budding yeast core network contains less nodes than the fission yeast core network (although it is more densely connected), this could be a consequence of the extensive studies of cytokinesis , cell cycle  and cell polarity  and their careful annotation in fission yeast –, but it also reflects the loss of some of the conserved eukaryotic cell cycle genes from budding yeast , .
The core interaction network contains several interactions between proteins that do not share a GO annotation; however the majority of links (91%) are between proteins which share at least one functional annotation among those under consideration (regulation of cell cycle, cytokinesis, and polarity) (Figure 2B). To probe this, we examined the relationship between the functional annotation of a node and that of its interaction partners. In fission yeast, any protein with a given functional annotation was 11 times (1.9 would be expected randomly, see Figure S3A) more likely to interact with another protein with the same functional annotation than with another protein with different functional annotations (for the budding yeast core network, this ratio was 4.5 vs. 1.06 expected, see Figure S3B). Since fission yeast has more proteins that belong to all three categories (16 in fission yeast versus 6 in budding yeast), we tested to see whether this observed functional modularity was due to their presence. We removed all proteins belonging to all three categories from both networks and repeated the analysis. This did not significantly alter the results as the ratios remained after the removal (10.38 times more likely for fission yeast and 4.16 for budding yeast) suggesting that the functional modularity observed in fission yeast is not caused by the presence of highly connected proteins with multiple annotations, but rather that the fission yeast network is characterized by strong connections between local communities that share functional annotations. It is however important to note that the GO categories ‘regulation of cell cycle’ and ‘cytokinesis’ are partially overlapping. In particular ‘regulation of cell cycle cytokinesis’ is a child term of both ‘regulation of cell cycle’ and ‘cytokinesis’. Even when taking this overlap into account in the analysis, we still observe a high degree of functional modularity in the core networks of both fission and budding yeast (not shown).
We further analyzed this effect using a community detection algorithm, which identifies local communities in a network and allows their overlap – as we have nodes with multiple annotations. We applied the k-clique propagation algorithm ,  and examined the communities generated by the method with k = 4. While the communities generated by the algorithm do not exactly match the functional annotations, we find that the cliques generated by the algorithm are primarily formed by proteins that share functional annotations (Figure 3A,B). Upon closer examination, the few proteins that do not share a functional annotation with the other members of a clique seem to have related roles: for example, in the 5th clique on Figure 3B, the lone ‘non-polarity’ protein is Rgf3, which was shown to play an important cell-wall remodeling role downstream of Rho1, one of the key regulators of polarity ,  (consult Table S5 for all clique members).
Figure 3. Segregation of functional communities in the core network.
A clique propagation algorithm was used to identify locally highly connected communities of the core network. The ten cliques generated by the algorithm segregate in the interaction network if laid out by a force-based algorithm that brings closer together the stronger interacting groups (A). Node colour determined by the functional annotation (same as Figure 2, inset on panel B here). Proteins belonging to the same clique share the same border colour. Proteins belonging to the same clique largely share functional annotations. Pie charts show the functional distribution of proteins found in each clique (B). Numbers report the number of proteins with the annotations corresponding to the given colour coded annotation (see inset for colours).doi:10.1371/journal.pcbi.1002732.g003
Identification of ‘linker’ proteins by network analysis
To systematically study proteins linking different cellular processes, we next used a network-based approach aiming to identify proteins that function as ‘linkers’ between different functional categories (Figure 4A). To do so, we constructed protein-protein interaction networks consisting only of proteins with one of the investigated functional annotations (cell cycle, cytokinesis or polarity regulation). We then calculated the betweenness centrality score for every node in each of these networks and in the merged core network. Betweenness Centrality (BC) measures how often a node is found in the shortest path between pairs of other nodes in the network; intuitively, it can be thought of as a measure of how central a node is in a network. If a node has a low centrality score it is localized at the fringe of a network, while if it has a high score it is localized near the centre. Next we ranked the proteins based on their BC score (in case of a tie, these proteins got their average rank). To ensure that this ranking method is robust even in the presence of imperfect interaction data certainly missing important links, we randomly added 10% extra edges to all the networks 1000 times, and recalculated the ranking of all proteins at each iteration (Figures S4). While the exact ranking of proteins is not very robust to addition of extra edges, if we examine all the proteins in the top 20%, we can observe that most fluctuate out of the top 20% only very rarely, and that we nearly never observe a protein in the top 10% drop out of the top 20%. It is also reassuring that the top of the rankings starts with expected key regulators of each function: the polarity landmark Tea1 –, the actin-regulating Rho GTPase Cdc42 ,  and actin (Act1) all came on the top of the polarity list. At the same time Cdc2, Wee1 and Cdc25  are on the top of the cell cycle list (and also on the top of the core list) and the SIN scaffold Cdc11  and the CDK counteracting, SIN activator phosphatase Clp1 ,  are leading the cytokinesis ranking (Figure S4 and Table S3).
Figure 4. Concepts of ‘linker’ protein detection and robustness of the method.
(A) ‘Linker’ proteins are found at the edge of a sub-network, but are central in the context of a larger network. Such proteins have low betweenness centrality (BC) score when considered in the context of their sub-network, but have a high BC score in the core network even though they do not have a functional annotation to the other category making up the core network. Black edges indicate edges between proteins that do not share functional annotations, while the other edges are gray. Table on right gives ranks and linkerity measures for all nodes in network ‘A’ in the same style as Table 2 does. (B) Analysis of the robustness of linkerity scores for the polarity network of fission yeast cells. We added 10% extra edges randomly to the network, and computed the linkerity score of all proteins after each iteration. Bars show mean ranking with standard deviation. Blue dashed line indicates cutoff for top 10% and red line marks the top 20% (results of other type of network perturbations are reported in Figure S5).doi:10.1371/journal.pcbi.1002732.g004
In the next step we compared the betweenness centrality rank of every protein in a sub-network to its relative rank in the core network. Only proteins that were originally in the sub-network were considered during this ranking based on scores they got for their position in the core network. We then calculated the ratio of the relative rank in the core network and the rank in the sub-network. We termed this calculated value ‘linkerity’, as this value is high for proteins that are found at the fringe of the network of proteins controlling a given cellular process, but central when considered in the context of a bigger network (Figure 4A):(1)
Proteins with high linkerity, we hypothesized, are likely to play a crucial role to function as linkers between different cellular processes. Specifically, we focused on the relationship of the polarity network to the rest of the core network to clarify how the cell cycle and the cytokinesis machinery control the temporal changes in the localization of polarized growth zones (top of Table 2, consult Table S3 for the rest of the list). Here, we show the top 10 proteins with the highest linkerity scores. These proteins became far more central when the polarity sub-network was embedded into the core network. Most of these proteins have GO annotations for multiple processes (among the annotations under consideration), thus their linking capacity is not that surprising. Novel linkers of polarity regulation could be those that were not associated with cytokinesis or cell cycle control but gained a high linkerity score in our analysis. The formin For3 , the AMP-activated, Snf1-like protein kinase Ssp2 , , the RNB-like protein Sts5  and the MRG family protein Alp13  are examples of proteins that match this. For3 is a well-characterized regulator of Tea1 to Cdc42 signalling , , the other three are less well characterized. The Rho GTPase Rho4  might be also an interesting linker candidate as it has established roles in polarity and cytokinesis regulation, but its exact function is not well characterized and it has no association to cell cycle regulation. Despite this, Rho4 has a central position in the core network that contains 75% cell cycle proteins (Figure 2A), furthermore its expression is cell cycle regulated . The highest linkerity proteins from the cytokinesis and cell cycle regulation networks also contain a number of proteins which are also associated with polarity regulation (Table 2). Scd1, Pom1 and Tea1 are on the top of the cell cycle linkerity list and Pmk1 , Shk1 and Tea1 lead the cytokinesis list after Bgs1, which is essential for cell wall synthesis , but has no polarity related GO annotation. These are on the edge of the cell cycle regulation or cytokinesis network but became central when they are merged with the polarity network, thus these can be also considered as linkers. As above for BC scores, we analysed the robustness of linkerity in the presence of imperfect network interaction data: we added or removed 10% of the edges from the core network at random or following a preferential attachment model and calculated linkerity scores for all proteins. Figure 4B reports the average and standard deviation from 500 random networks with 10% extra edge (other cases in Figure S5) for the top linkerity polarity proteins. Importantly the top 10 of the unperturbed list (Table 2) can be found in the top 16 of the list after 10% possible missing links were considered (Figure 4B).
Table 2. Top ten proteins with highest Linkerity measures from the three sub-networks.doi:10.1371/journal.pcbi.1002732.t002
As discussed above, in both fission yeast and budding yeast, we observe a high degree of functional modularity, i.e. proteins tend to interact with proteins that share their functional role. Since linker proteins play a special role in bringing together different cellular processes, we examined whether proteins with high linkerity interacted with proteins with different functional roles at a higher rate than low linkerity proteins. For all the proteins of the core network we calculated the number of its interactors (network neighbours) with cell cycle, cytokinesis and polarity annotations (Table S3). Then for every protein in each functional category (Figure 2) we calculated the ratio of the number of its interactions with proteins with the two other functional annotations to the number of its interactions with proteins with the same functional annotation. We observed that high linkerity is significantly correlated with having a high ratio of heterogeneously annotated neighbours across all functional categories in both yeasts, suggesting that linker proteins do play an important role in bridging proteins from different functional groups (see Text S2 for details).
Sts5 is a novel linker protein bridging cell polarity to cell cycle
Among predicted linker proteins we focused on Sts5, which is known to genetically interact with Ssp2 , which itself is likely to be linked with the cell cycle machinery as ssp2Δ cells cannot start mitosis when nutrient-starved . Sts5 is an orthologue of budding yeast SSD1  and therefore a candidate translational repressor. It is reported to control actin localisation in interphase and sts5Δ was shown to be compensated by mutations in Ssp2. Furthermore, Sts5 mRNA levels were shown to oscillate , . To examine the interplay between Sts5 and the cell cycle, we tagged the endogenous protein with a triple GFP tag and visualized its localization together with that of mCh-Atb2 (Alpha tubulin 2), which labels microtubules and hence served as a cell cycle stage marker. In interphase cells, Sts5 had a mostly diffuse cytoplasmic localization, however during mitosis it appeared to localize in dotted, cytoplasmic bodies (Figure 5A). The number of Sts5 dots increased throughout mitosis and peaked coinciding with the assembly of the Post Anaphase Array (PPA) of microtubules (Figure 5B). Time-lapse movies of mitotic cells also confirmed that the number of cytoplasmic dots increased until the formation of the PAA and sharply dropped to zero as cells entered interphase (Figure S6). Previous studies of Sts5  showed that it was required for correct cell growth and actin patch localization during interphase. Taken together with our results, this suggests that the cell cycle controls Sts5 activity by gradually sequestering it in cytoplasmic bodies during mitosis.
Figure 5. Localization in cells of Sts5 during the cell cycle.
(A) Imaging of fission yeast cells co-expressing Sts5-3GFP and mCh-atb2 (labelling the different microtubule structures seen through the cell cycle, and hence acting as cell cycle stage indicators). Interphase cells (I) have diffuse Sts5 localization (with a few cytoplasmic speckles) while cells in mitosis (either in anaphase (A) or during the time of the post anaphase array (PAA)) have several Sts5 cytoplasmic dots. Scalebar: 5 µm. (B) Population based analysis of cycling cells revealed that at metaphase the number of Sts5 speckles greatly increases and sharply drops during septum formation. Average and standard deviation of number of dots were automatically detected in multiple cells (see Materials and Methods for details).doi:10.1371/journal.pcbi.1002732.g005
In this work, we have carried out the first network analysis based, large-scale identification of proteins linking various cellular processes in the fission yeast protein-protein interaction network. Although data for fission yeast mostly comes from manually annotated experiments, literature mining and computational inference, the network displays features comparable to those observed in other organisms. We have shown that the relationship between lethality and different network measures holds in fission yeast, and that network based approaches can give meaningful and interesting results even in organisms lacking high-throughput interaction experiments.
Our analysis of the core network of all proteins regulating cell cycle, cytokinesis, and polarized growth revealed a striking degree of functional modularity, which we have found to be highly robust to the deletion of key nodes in the network. This functional modularity was also observed when examining the communities detected by a clique propagation algorithm. Detected communities had very low heterogeneity between the functional annotations of member proteins. We investigated this modularity further by using a network approach to identify linker proteins bridging different functional categories. We propose a new network measure, linkerity, which is the ratio of the ranking by betweennness centrality measures of all the nodes belonging to a given sub-network considered in the sub-network alone and considered in the context of a larger network (Figure 4A). This new network measure does not appear to show strong correlation with other existing network measures (Text S3). Due to the non-linear distribution of betweenness centrality measures in real systems , it might be necessary to normalize this linkerity measure in case linkers between large sub-networks are investigated.
We tested this concept on the connections of the polarized cell growth regulatory network to the cytokinesis and cell cycle networks of fission yeast cells. These are highly characterized and strongly interacting networks and the connection between these processes is of high importance in other organisms , , –. We confirmed that many of the highest linkerity scoring proteins in the polarity network were already known to play important roles in multiple processes. Among these the F-BAR protein Cdc15 provide good validation as it was already shown to play a role in switching from polarized growth to cytokinetic-actin ring formation in mitosis . Similarly Skb1  and Cdr1 ,  were shown to serve as links between cell cycle and cell polarity. All these proteins shifted from a low ranking in the polarity network to a high rank in the core network (Table 2), and thus their role in polarity regulation might come from the pleotropic behavior of these proteins or from their active role in connecting polarized growth regulation to cell cycle and cytokinesis. We also discovered that the proteins with high linkerity tend to interact with a more diverse set of proteins than those with low linkerity. This suggests that high linkerity proteins might play a pleiotropic role by linking together different functional processes , .
Sts5 had the second highest ranking in the polarity network among the top ten linkerity proteins (after actin, Act1 that is also essential for cytokinesis). Sts5 is known to play an important role in controlling the localization of the actin machinery to cell ends during interphase, although Sts5 is localized in the cytoplasm . We have shown that Sts5 is localized in cytoplasmic dots during mitosis, but diffuse during interphase, implying that its localization is cell cycle regulated. Growing tip localized polarity proteins change their localization when cells enter mitosis , , but it is not expected from a cytoplasmic protein to localize into clusters in a cell cycle dependent manner. The overall level of Sts5 protein slightly increases upon entry to mitosis (Figure S6), but its activity reaches its lowest level as its accumulation into cytoplasmic dots reaches a peak. This suggests that the cell cycle controls polarity by sequestering Sts5 in and out of cytoplasmic bodies, and the triggered release and sequestration function as switches between polarized cell growth and cytokinesis. The exact nature of those cytoplasmic bodies is still unclear, however the budding yeast Sts5 homologue SSD1 was shown to localize to P-bodies , the cytoplasmic centers of mRNA degradation. Interestingly, like Sts5, Ssp2 and the stress pathway kinase Wis4 are also localized into cytoplasmic dots  and it was proposed that the stress pathway and Sts5 might act in opposing manner on cell polarity . It will be important in the future to investigate if these proteins co-localize in the observed cytoplasmic dots and how these are exactly controlled by the cell cycle.
Sts5 was previously shown to genetically interact with members of the stress pathway . A number of other kinases associated with stress response (such as Sty1, Skb1, Orb6, Pmk1, Mkh1) have been shown to have defects in NETO ,  and many of these appear highly ranked in our linkerity lists (Table 2). Furthermore, the cell end-localized polarity factor Tea4 was also shown to interact with the stress pathway . These make the stress pathway a particularly intriguing target for further analysis in the search for proteins linking cell cycle and polarity, as it may play a special role as a pleiotropy integrator of both internal and external cellular signals in response to different stimuli in fission yeast and also in higher eukaryotes , . The linkerity analysis of cytokinesis and cell cycle regulatory proteins (bottom parts of Table 2) also give some interesting predictions. For instance the high linkerity of the transcription factor Cdc10  in the cell cycle network suggests its role controlling the transcription of important polarity and cytokinesis genes, especially with key regulators, such as Cdc15, Scd2, Sts5, Rho4 and Sid2 having periodic transcriptional profile , .
While we believe that the method presented here can be applied to other organisms and cellular processes to find linker proteins, different model organisms offer unique advantages and challenges. In this study, we took advantage of the extensive annotation of proteins by the fission yeast community to define discrete sub-networks, bypassing the very difficult problems involved in defining meaningful ‘communities’ using purely network based approaches , , . While this approach has its advantages, it is important to be aware of any partial overlaps between the used GO terms due to the presence of common child terms. The amount of overlap between child terms is also not consistent across multiple organisms, requiring special care when doing comparisons that involve multiple organisms (for example, the “regulation of cell cycle cytokinesis” is a child term of both “regulation of cell cycle” and “cytokinesis” and it contains 47 proteins in fission yeast, and only 4 proteins in budding yeast). Furthermore, while we have shown that the ranking of proteins within the communities is robust to noise, the actual communities detected by various algorithms as well as the structure of the network are strongly influenced by the granularity and quality of the interaction data used (Text S4 and ). In fission yeast, where interaction data is relatively sparse but there is extensive functional annotation, it makes sense to use GO annotations to define functional sub-networks . Very recent network predictions based on machine-learning methods  will enable us to perform more careful analysis in this organism as well. Other organisms with larger gene sets will often have a lower annotation coverage ; in these cases functional groups in the PPI network need to be identified by community detection algorithms or predefined by the authors . Once such functional groups are established, the described method provides a good means to identify proteins likely to have a role in connecting functional regulatory networks in any organism. Likewise, the defined linkerity measure can be used to identify key linker nodes of sub-networks in any complex network , –.
Materials and Methods
Bioinformatics data compilation
To obtain a list of proteins associated with specific cellular processes, we used the Gene Ontology (http://www.geneontology.org/) and downloaded all gene products associated with a given term. It is important to note that while ‘cytokinesis’ (GO:0000910) and ‘cell cycle regulation’ (GO:0051726) have specific terms that cover all proteins commonly associated with those processes, for polarity S. pombe proteins are split between ‘establishment or maintenance of cell polarity’ (GO:0007163) and ‘cell morphogenesis’ (GO:0000902). In the analysis, we thus used the umbrella term ‘polarity’ to include proteins in both of these categories. Data in STRING (http://string-db.org/) is present at different confidence scores. Confidence scores in STRING represent the likelihood of the two proteins actually interacting, and depend on the reliability of the source of the interaction. For example, an interaction that is reported in a single experiment will have a far higher confidence score than an interaction that is inferred through text mining or homology alone. We studied the effect of a cutoff in this confidence score on network size defined as the fraction of all proteins connected with at least one other protein; the main component size defined as the fraction of all proteins connected to the largest component in the network; and the edge fraction defined as the fraction of all edges found, compared to the theoretical maximum. To download the number of PubMed abstracts mentioning the name of a protein in the network, we relied on the Entrez module of the Biopython package (http://biopython.org/wiki/Biopython). Statistical analysis, including calculation of correlations, was carried out using the Statistics module of the SciPy package (http://www.scipy.org/). All network measures were calculated using pre-existing algorithms implemented in NetworkX (http://networkx.lanl.gov/). For community structure detection we used the k-clique propagation algorithm originally described in , and implemented in NetworkX . Packages were packaged in the Enthought Python Distribution courtesy of Enthought (http://www.enthought.com/).
Network analysis workflow
The pipeline used to create the networks was:
- We connected to the MySQL Gene Ontology database using custom python scripts, and downloaded all proteins associated with a given biological process.
- We took all proteins downloaded and used them to query STRING, downloading all the information about protein-protein interactions in PSI-MI-TAB format. It is important to note that STRING and Gene Ontology sometimes identify the same gene by a different name, therefore special care was taken to use consistent nomenclature.
- We parsed the PSI-MI-TAB file and transformed it into a NetworkX graph, which we could then study using both algorithms built into NetworkX as well as custom scripts.
We repeated the analysis described in the main text using networks obtained from BioGRID. In that case, instead of using STRING in step 2 we parsed the full network of a given organism from a PSI-MI-TAB file available for download on the BioGRID website, then extracted the sub-graph containing the nodes obtained in step 1 and edges of physical interactions stored in the database. The results presented are based on the state of all databases on 13 March 2012. The calculated network measures, PubMed citations and all presented numerical results are detailed in the Excel files of Tables S1, S2, S3, S4.
All Python scripts used to download data from databases as well as for analysis are available upon request.
Strains and strain construction
The S. pombe strain used in this study was MH123 (h- sts5-3GFP-L-nat Z2-mCh-atb2-hph leu1 ura4 ade6-M216 his7). Conventional PCR-based gene targeting methods for S. pombe were used for gene tagging –.
Live microscopy cell imaging
Prior to imaging, S. pombe strains were grown at 32°C in yeast extract with supplements (YES) (5) to exponential growth. Aliquots of 300 ml cells were mounted onto 1.5 coverslip glass-bottomed plastic dishes (MatTek; P35G-1.5-14-C) pre-coated with 10 ml 1 mg/ml lectin (Sigma; L1395 and Patricell Ltd; L-1301-25) that had been allowed to air dry. After a 30-minute incubation, cells unbound to the lectin-coated glass were removed by washing with minimal medium (EMM) – and the bound cells were kept in a final suspension of 1 ml EMM.
Imaging was performed with both: an OMX microscope (Applied Precision) in conventional resolution mode, with an Olympus UPlanSapo ×100 oil immersion lens (NA1.4) and 1.512 RI immersion oil (Applied Precision); and a DeltaVision microscope (Applied Precision), comprising an Olympus 1×71 widefield microscope, an Olympus UPlanSapo ×100 oil immersion lens (NA1.4) and an Photometrics CoolSNAP HQ2 camera. For analysis of Sts5-3GFP speckle number, stacks were taken at 0.4 um apart for 16 focal planes on the Deltavision microscope. Time lapses were taken for single focal planes at ten-minute intervals on the DeltaVision microscope.
Automated analysis of Sts5-3GFP speckle number
Cells within microscopy image fields were automatically segmented from the transmitted light channel using an algorithm developed in-house and coded in Matlab. For each cell, the cell-cycle stage was determined manually by looking at the mCh-Atb2 channel.Sts5-GFP speckles were detected using the spot detection module of the ICY software (http://icy.bioimageanalysis.org/; 
The cell cycle + cytokinesis + polarity = core interaction network of budding yeast proteins. (A) Venn diagram showing the overlap among the different Gene Ontology functional groups in the proteins present in the core network of budding yeast. Proteins with multiple functional annotations have colours that are the sum of the colours of the individual functional annotations, proteins belonging to all three functional groups are in white. (B) Protein-protein interaction in the budding yeast core network (from the STRING database at cutoff 0.7). Node colour same as in panel A. Node size is proportional to degree of the protein, and node order within a category (clockwise) is also determined by degree. 469 Black edges link proteins that do not share functional annotations, while 2146 grey edges link proteins that have at least one common GO annotation (thus white nodes have only grey links). White nodes (nodes belonging to all categories) are shown in the inner circle in the middle of the network.
Scale free distribution of networks. We calculated the degree of every node in the largest connected component of the genomwide network for both fission yeast (A) and budding yeast (B). We then calculated a histogram for frequency of degree (with number of bins equal to the maximum degree observed in the network) and plotted log(frequency) vs log(degree). Best fits to log(P(k))~log (ck−γ) were calculated using a least square minimization algorithm from scipy (http://www.scipy.org/).
Functional modularity in the core networks. To calculate how much the functional modularity (the ratio of interactions between nodes with a shared GO category versus interactions between nodes with no GO category in common) observed for the core network of budding and fission yeast deviated from a random network, we kept all the category labels for all the nodes, but rewired the network either completely at random (A, C), or using a method that preserves degree-distribution (B, D) . To rewire the networks at random, we removed every edge from the network then added an edge between any two nodes chosen at random until the total amount of edges in the network was equal to the original amount. To preserve degree distribution of the networks, we performed a double edge swap across the network. We picked two existing edges at random between nodes (u,v) and (x,y). We then added an edge between (u,x) and (y,v) and removed the original edge. Red arrows indicate the observed ratio for the core network, the distributions represent 1000 different random networks and their functional modularity.
Robustness analysis for the betweenness centrality ranking for polarity, cytokinesis and cell cycle networks in fission yeast. We analysed the robustness of ranking proteins by BC centrality in the presence of imperfect network interaction data. We added 10% extra edges at random to the network, calculated BC for every node after adding the edges, and ranked all the proteins. We calculated the mean and standard deviation for the rank of every protein in the network after repeating the procedure 1000 times. We normalized the rank of all proteins (Rank/number of nodes) and plotted the top 20% of nodes and their mean and standard deviation. The blue dotted line represents the cutoff for top 10% nodes, and the red dotted line represents the cutoff for top 20% of nodes. A, B, C are the top 20% proteins of regulation of cell cycle, cytokinesis and polarity of fission yeast.
Robustness analysis of linkerity of proteins in the fission yeast polarity network. We systematically analysed the robustness of linkerity in the presence of imperfect network interaction data. We added 10% edges preferentially to nodes with high degree (A) or removed 10% edges at random (B) to the core network. In the preferential attachment model, the probability P that a given node N had of gaining an edge was directly proportional to its degree P(N)~Degree(N). In the random model P(N)~k where k is a constant. Probabilities were normalized to increase or decrease the total edges of the network by 10%. We calculated the mean and standard deviation for the betweenness centrality of every protein belonging to the polarity sub-network after repeating the procedure 1000 times. We plotted the top 20% of nodes and their mean and standard deviation. The blue dotted line represents the cutoff for top 10% nodes, and the red dotted line represents the cutoff for top 20% of nodes.
Time-lapse analysis of Sts5 localization in fission yeast cells. Microtubules are visualized using mCherry labeled tubulin (Atb2) to identify cell cycle stage (A and B right column and Sts5-3GFP is visualized on the left). As the cell cycle progresses, Sts5 starts to accumulate into cytoplasmic dots, which then rapidly disappear upon septum formation. C is an automatic quantification of the amount of cytoplasmic dots in cells at different stages of the cell cycle.
Analysis of the genome-wide fission yeast network. See detailed description under Table S2.
Analysis of the genome-wide budding yeast network. Tabulated file (in .xls format) containing network measures for all protein in the largest connected component of the genome-wide network of fission (S1) and budding (S2) yeast. Columns include: Common name: Common name. Systematic name: Systematic name (for fission yeast), GO database ID (for budding yeast) Description: Brief description of known protein activity. PubMed count: Number of abstracts discussing that particular protein in fission yeast available in PubMed. Lethality: E (Essential) If deletion of the gene causes lethality, V (Viable) otherwise. Scores: Betweenness Centrality and Degree scores for the protein in the network using either STRING interaction data (at increasing cutoffs, 0.4, 0.7, 0.9) or data from BioGRID (only the physical protein-protein interaction data). Genes with no entry at a given cutoff have no other interactions with any proteins in the network.
Analysis of the core fission yeast network. See detailed description under Table S4.
Analysis of the core budding yeast network. Tabulated file (in .xls format) containing network measures for all protein in the core network of fission (S3) and budding (S4) yeast. Columns include: Common name: Common name. Systematic name: Systematic name (for fission yeast), GO database ID (for budding yeast) Description: Brief description of known protein activity. PubMed count: Number of abstracts discussing that particular protein in fission yeast available in PubMed. Lethality: E (Essential) If deletion of the gene causes lethality, V (Viable) otherwise. GO Categories: Which of the three categories (Cytokinesis (CY), Polarity (P), Cell Cycle (CC)) does the protein belong too? Scores: Betweenness Centrality and Degree scores and ranks for all the sub-networks the protein belongs to, as well as the core network. In the sub-network, betweenness rank (Ranksub-network) is calculated by ranking all the proteins from highest to lowest according to their betweenness. In the core network, the betweenness rank (Rankcore) is calculated only between proteins that are found in the original sub-network. To avoid artifacts due to the presence of multiple proteins with 0 betweenness, we assign consecutive proteins with the exact same score have the same rank, which is simply defined as the average of their ranking . For example: in a network of 6 proteins [A, B, C, D, E, F], with BC values of [10, 10, 7, 5, 5, 5], the ranking would be: [(A, 1.5), (B, 1.5), (C, 3), (D, 5), (E, 5), (F, 5)] Linkerity: Linkerity calculated for all the categories as given in Equation 1. Note that the linkerity for a protein that doesn't shift in rank is 1 by definition.
All members of the cliques identified on Figure 3B. List of all proteins belonging to the cliques described in Figure 3. Clique 1 corresponds to the top left clique in Figure 3B, with cliques increasing moving from left to right.
Predicting essentiality by network measures in fission and budding yeast.
Analysis of the neighbors of high linkerity proteins.
Correlation between linkerity and other network measures.
Linkerity at various network confidences.
The authors are thankful to L.J. Jensen, V. Wood, J. Bahler, H. Moriya, K. Nakano, M. Toya, J. Pines, M. Godinho Ferreira, T. Surrey, J. Howard, A. Ciliberto, C. Pal, A. Sveiczer, C. Bakal and the Carazo-Salas, Csikasz-Nagy and Sato groups for help and comments
Conceived and designed the experiments: FV RECS ACN. Performed the experiments: FV JD AC. Analyzed the data: FV AB AC ACN. Contributed reagents/materials/analysis tools: AB FJ MS. Wrote the paper: FV RECS ACN.
- 1. Nurse P (1990) Universal control mechanism regulating onset of M-phase. Nature 344: 503–508 doi:10.1038/344503a0.
- 2. Nurse P (2002) Cyclin dependent kinases and cell cycle control (nobel lecture). Chembiochem 3: 596–603 doi:10.1002/1439-7633(20020703)3:7<596::AID-CBIC596>3.0.CO;2-U.
- 3. Morgan DO (2006) The Cell Cycle: Principles of Control. New Science Press, Ltd.
- 4. Enoch T, Carr AM, Nurse P (1992) Fission yeast genes involved in coupling mitosis to completion of DNA replication. Genes Dev 6: 2035–2046 doi:10.1101/gad.6.11.2035.
- 5. Lacroix B, Maddox AS (2012) Cytokinesis, ploidy and aneuploidy. J Pathol 226: 338–351 doi:10.1002/path.3013.
- 6. Diffley JFX (2011) Quality control in the initiation of eukaryotic DNA replication. Phil Trans R Soc Lond B Biol Sci 366: 3545–3553 doi:10.1098/rstb.2011.0073.
- 7. Martin SG (2009) Geometric control of the cell cycle. Cell Cycle 8: 3643–3647. doi: 10.4161/cc.8.22.9891
- 8. Verde F, Mata J, Nurse P (1995) Fission yeast cell morphogenesis: identification of new genes and analysis of their role during the cell cycle. J Cell Biol 131: 1529–1538. doi: 10.1083/jcb.131.6.1529
- 9. Kellogg DR (2003) Wee1-dependent mechanisms required for coordination of cell growth and cell division. J Cell Sci 116: 4883–4890 doi:10.1242/jcs.00908.
- 10. Wilson PD (1997) Epithelial cell polarity and disease. Am J Physiol 272: F434–42.
- 11. Mitchison JM (1990) The fission yeast, Schizosaccharomyces pombe. Bioessays 12: 189–191 doi:10.1002/bies.950120409.
- 12. Mitchison JM, Nurse P (1985) Growth in cell length in the fission yeast Schizosaccharomyces pombe. J Cell Sci 75: 357–376.
- 13. Hayles J, Nurse P (2001) A journey into space. Nat Rev Mol Cell Biol 2: 647–656 doi:10.1038/35089520.
- 14. Vavylonis D, Wu J-Q, Hao S, Shaughnessy BO, Pollard TD, et al. (2007) Assembly mechanism of the contractile ring for cytokinesis by fission yeast. Science 319: 97–100 doi:10.1126/science.1151086.
- 15. Wu J-Q, Sirotkin V, Kovar DR, Lord M, Beltzner CC, et al. (2006) Assembly of the cytokinetic contractile ring from a broad band of nodes in fission yeast. J Cell Biol 174: 391–402 doi:10.1083/jcb.200602032.
- 16. Csikász-Nagy A, Gyorffy B, Alt W, Tyson JJ, Novák B (2008) Spatial controls for growth zone formation during the fission yeast cell cycle. Yeast 25: 59–69 doi:10.1002/yea.1571.
- 17. Martin SG, Berthelot-Grosjean M (2009) Polar gradients of the DYRK-family kinase Pom1 couple cell length with the cell cycle. Nature 459: 852–856 doi:10.1038/nature08054.
- 18. Csikász-Nagy A, Kapuy O, Gyorffy B, Tyson JJ, Novák B (2007) Modeling the septation initiation network (SIN) in fission yeast cells. Curr Genet 51: 245–255 doi:10.1007/s00294-007-0123-4.
- 19. Pollard TD, Wu J-Q (2010) Understanding cytokinesis: lessons from fission yeast. Nat Rev Mol Cell Biol 11: 149–155 doi:10.1038/nrm2834.
- 20. Csikász-Nagy A, Battogtokh D, Chen KC, Novák B, Tyson JJ (2006) Analysis of a generic model of eukaryotic cell-cycle regulation. Biophys J 90: 4361–4379 doi:10.1529/biophysj.106.081240.
- 21. Bähler J, Pringle JR (1998) Pom1p, a fission yeast protein kinase that provides positional information for both polarized growth and cytokinesis. Genes Dev 12: 1356–1370. doi: 10.1101/gad.12.9.1356
- 22. Hachet O, Berthelot-Grosjean M, Kokkoris K, Vincenzetti V, Moosbrugger J, et al. (2011) A Phosphorylation Cycle Shapes Gradients of the DYRK Family Kinase Pom1 at the Plasma Membrane. Cell 145: 1116–1128 doi:10.1016/j.cell.2011.05.014.
- 23. Moseley JB, Mayeux A, Paoletti A, Nurse P (2009) A spatial gradient coordinates cell size and mitotic entry in fission yeast. Nature 459: 857–860 doi:10.1038/nature08074.
- 24. Krapp A, Simanis V (2008) An overview of the fission yeast septation initiation network (SIN). Biochem Soc Trans 36: 411–415 doi:10.1042/BST0360411.
- 25. Ray S, Kume K, Gupta S, Ge W, Balasubramanian M, et al. (2010) The mitosis-to-interphase transition is coordinated by cross talk between the SIN and MOR pathways in Schizosaccharomyces pombe. J Cell Biol 190: 793–805 doi:10.1083/jcb.201002055.
- 26. Gupta S, McCollum D (2011) Crosstalk between NDR kinase pathways coordinates cell cycle dependent actin rearrangements. Cell Div 6: 19 doi:10.1186/1747-1028-6-19.
- 27. Cornils H, Kohler RS, Hergovich A, Hemmings BA (2011) Downstream of human NDR kinases: impacting on c-myc and p21 protein stability to control cell cycle progression. Cell Cycle 10: 1897–1904. doi: 10.4161/cc.10.12.15826
- 28. Hergovich A, Cornils H, Hemmings BA (2008) Mammalian NDR protein kinases: from regulation to a role in centrosome duplication. Biochim Biophys Acta 1784: 3–15 doi:10.1016/j.bbapap.2007.07.017.
- 29. Jensen LJ, Jensen TS, de Lichtenberg U, Brunak S, Bork P (2006) Co-evolution of transcriptional and post-translational cell-cycle regulation. Nature 443: 594–597 doi:10.1038/nature05186.
- 30. Rupes I, Webb BA, Mak A, Young PG (2001) G2/M Arrest Caused by Actin Disruption Is a Manifestation of the Cell Size Checkpoint in Fission Yeast. Mol Biol Cell 12: 3892–3903. doi: 10.1091/mbc.12.12.3892
- 31. Liu J, Wang H, Balasubramanian M (2000) A checkpoint that monitors cytokinesis in Schizosaccharomyces pombe. J Cell Sci 113: 1223–1230.
- 32. Jensen LJ, Saric J, Bork P (2006) Literature mining for the biologist: from information retrieval to biological discovery. Nat Rev Genet 7: 119–129 doi:10.1038/nrg1768.
- 33. Pancaldi V, Sarac OS, Rallis C, McLean JR, Prevorovsky M, et al. (2012) Predicting the Fission Yeast Protein Interaction Network. G3: Genes|Genomes|Genetics 2: 453–467 doi:10.1534/g3.111.001560.
- 34. Wood V (2006) Schizosaccharomyces pombe comparative genomics; from sequence to systems. In: Sunnerhagen P, Piskur J, editors. Comparative Genomics using fungi as models. Comparative Genomics using fungi as models. Springer. pp. 233–285.
- 35. Jensen LJ, Kuhn M, Stark M, Chaffron S, Creevey C, et al. (2009) STRING 8–a global view on proteins and their functional interactions in 630 organisms. Nucleic Acids Res 37: D412–6 doi:10.1093/nar/gkn760.
- 36. von Mering C, Jensen LJ, Kuhn M, Chaffron S, Doerks T, et al. (2007) STRING 7–recent developments in the integration and prediction of protein interactions. Nucleic Acids Res 35: D358–62 doi:10.1093/nar/gkl825.
- 37. Aslett M, Wood V (2006) Gene Ontology annotation status of the fission yeast genome: preliminary coverage approaches 100%. Yeast 23: 913–919 doi:10.1002/yea.1420.
- 38. Wood V, Harris MA, McDowall MD, Rutherford K, Vaughan BW, et al. (2012) PomBase: a comprehensive online resource for fission yeast. Nucleic Acids Res 40: D695–9 doi:10.1093/nar/gkr853.
- 39. Bitton DA, Wood V, Scutt PJ, Grallert A, Yates T, et al. (2011) Augmented annotation of the Schizosaccharomyces pombe genome reveals additional genes required for growth and viability. Genetics 187: 1207–1217 doi:10.1534/genetics.110.123497.
- 40. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, et al. (2000) Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25: 25–29 doi:10.1038/75556.
- 41. Gene Ontology Consortium (2012) The Gene Ontology: enhancements for 2011. Nucleic Acids Res 40: D559–64 doi:10.1093/nar/gkr1028.
- 42. Vaggi F, Jordan F, Csikász-Nagy A (2011) Keystone species of the molecular network of cellular functions. In: Jordan F, Jorgensen SE, editors. Models of the Ecological Hierarchy. Elsevier Press. pp. 71–86.
- 43. Han J-DJ, Bertin N, Hao T, Goldberg DS, Berriz GF, et al. (2004) Evidence for dynamically organized modularity in the yeast protein-protein interaction network. Nature 430: 88–93 doi:10.1038/nature02555.
- 44. Batada NN, Hurst LD, Tyers M (2006) Evolutionary and physiological importance of hub proteins. PLoS Comput Biol 2: e88 doi:10.1371/journal.pcbi.0020088.
- 45. Yu H, Kim PM, Sprecher E, Trifonov V, Gerstein M (2007) The importance of bottlenecks in protein networks: correlation with gene essentiality and expression dynamics. PLoS Comput Biol 3: e59 doi:10.1371/journal.pcbi.0030059.
- 46. Batada NN, Reguly T, Breitkreutz A, Boucher L, Breitkreutz B-J, et al. (2006) Stratus not altocumulus: a new view of the yeast protein interaction network. PLoS Biol 4: e317 doi:10.1371/journal.pbio.0040317.
- 47. Stark C, Breitkreutz B-J, Reguly T, Boucher L, Breitkreutz A, et al. (2006) BioGRID: a general repository for interaction datasets. Nucleic Acids Res 34: D535–9 doi:10.1093/nar/gkj109.
- 48. Barabási A-L (2009) Scale-free networks: a decade and beyond. Science 325: 412–413 doi:10.1126/science.1173299.
- 49. Kim D-U, Hayles J, Kim D, Wood V, Park H-O, et al. (2010) Analysis of a genome-wide set of gene deletions in the fission yeast Schizosaccharomyces pombe. Nat Biotechnol 28: 617–623 doi:10.1038/nbt.1628.
- 50. Decottignies A, Sanchez-Perez I, Nurse P (2003) Schizosaccharomyces pombe essential genes: a pilot study. Genome Res 13: 399–406 doi:10.1101/gr.636103.
- 51. Forsburg S, Nurse P (1991) CELL CYCLE REGULATION IN CEREVISIAE AND POMBE. Ann Rev Cell Biol 7: 227–56. doi: 10.1146/annurev.cb.07.110191.001303
- 52. Cross FR, Buchler NE, Skotheim JM (2011) Evolution of networks and sequences in eukaryotic cell cycle control. Phil Trans R Soc Lond B Biol Sci 366: 3532–3544 doi:10.1098/rstb.2011.0078.
- 53. Derényi I, Palla G, Vicsek T (2005) Clique percolation in random networks. Phys Rev Lett 94: 160202. doi: 10.1103/physrevlett.94.160202
- 54. Palla G, Derényi I, Farkas I, Vicsek T (2005) Uncovering the overlapping community structure of complex networks in nature and society. Nature 435: 814–818 doi:10.1038/nature03607.
- 55. Morrell-Falvey JL, Ren L, Feoktistova A, Haese GD, Gould KL (2005) Cell wall remodeling at the fission yeast cell division site requires the Rho-GEF Rgf3p. J Cell Sci 118: 5563–5573 doi:10.1242/jcs.02664.
- 56. Tajadura V, García B, García I, García P, Sánchez Y (2004) Schizosaccharomyces pombe Rgf3p is a specific Rho1 GEF that regulates cell wall beta-glucan biosynthesis through the GTPase Rho1p. J Cell Sci 117: 6163–6174 doi:10.1242/jcs.01530.
- 57. Sawin KE, Snaith Ha (2004) Role of microtubules and tea1p in establishment and maintenance of fission yeast cell polarity. J Cell Sci 117: 689–700 doi:10.1242/jcs.00925.
- 58. Niccoli T, Arellano M, Nurse P (2003) Role of Tea1p, Tea3p and Pom1p in the determination of cell ends in Schizosaccharomyces pombe. Yeast 20: 1349–1358 doi:10.1002/yea.1054.
- 59. Mata J, Nurse P (1997) tea1 and the microtubular cytoskeleton are important for generating global spatial order within the fission yeast cell. Cell 89: 939–949. doi: 10.1016/s0092-8674(00)80279-2
- 60. Kelly FD, Nurse P (2011) Spatial control of Cdc42 activation determines cell width in fission yeast. Mol Biol Cell 22: 3801–3811 doi:10.1091/mbc.E11-01-0057.
- 61. Chang F, Martin SG (2009) Shaping fission yeast with microtubules. Cold Spring Harb Perspect Biol 1: a001347 doi:10.1101/cshperspect.a001347.
- 62. Nurse P (2000) A long twentieth century of the cell cycle and beyond. Cell 100: 71–78. doi: 10.1016/s0092-8674(00)81684-0
- 63. Krapp A, Schmidt S, Cano E, Simanis V (2001) S. pombe cdc11p, together with sid4p, provides an anchor for septation initiation network proteins on the spindle pole body. Curr Biol 11: 1559–1568. doi: 10.1016/s0960-9822(01)00478-x
- 64. Cueille N, Salimova E, Esteban V, Blanco M, Moreno S, et al. (2001) Flp1, a fission yeast orthologue of the s. cerevisiae CDC14 gene, is not required for cyclin degradation or rum1p stabilisation at the end of mitosis. J Cell Sci 114: 2649–2664.
- 65. Trautmann S, Wolfe BA, Jorgensen P, Tyers M, Gould KL, et al. (2001) Fission yeast Clp1p phosphatase regulates G2/M transition and coordination of cytokinesis with cell cycle progression. Curr Biol 11: 931–940. doi: 10.1016/s0960-9822(01)00268-8
- 66. Feierbach B, Chang F (2001) Roles of the fission yeast formin for3p in cell polarity, actin cable formation and symmetric cell division. Curr Biol 11: 1656–1665. doi: 10.1016/s0960-9822(01)00525-5
- 67. Matsuzawa T, Fujita Y, Tohda H, Takegawa K (2012) Snf1-like protein kinase Ssp2 regulates glucose derepression in Schizosaccharomyces pombe. Eukaryot Cell 11: 159–167 doi:10.1128/EC.05268-11.
- 68. Matsusaka T, Hirata D, Yanagida M, Toda T (1995) A novel protein kinase gene ssp1+ is required for alteration of growth polarity and actin localization in fission yeast. EMBO J 14: 3325–3338.
- 69. Toda T, Niwa H, Nemoto T, Dhut S, Eddison M, et al. (1996) The fission yeast sts5+ gene is required for maintenance of growth polarity and functionally interacts with protein kinase C and an osmosensing MAP-kinase pathway. J Cell Sci 109 (Pt 9) 2331–2342.
- 70. Nakayama J-I, Xiao G, Noma K-I, Malikzay A, Bjerling P, et al. (2003) Alp13, an MRG family protein, is a component of fission yeast Clr6 histone deacetylase required for genomic integrity. EMBO J 22: 2776–2787 doi:10.1093/emboj/cdg248.
- 71. Wang H, Vavylonis D (2008) Model of For3p-mediated actin cable assembly in fission yeast. PLoS One 3: e4078 doi:10.1371/journal.pone.0004078.
- 72. Martin SG, McDonald WH, Yates JR, Chang F (2005) Tea4p links microtubule plus ends with the formin for3p in the establishment of cell polarity. Dev Cell 8: 479–491 doi:10.1016/j.devcel.2005.02.008.
- 73. Nakano K, Mutoh T, Arai R, Mabuchi I (2003) The small GTPase Rho4 is involved in controlling cell morphology and septation in fission yeast. Genes Cells 8: 357–370. doi: 10.1046/j.1365-2443.2003.00639.x
- 74. Gauthier NP, Jensen LJ, Wernersson R, Brunak S, Jensen TS (2010) Cyclebase.org: version 2.0, an updated comprehensive, multi-species repository of cell cycle experiments and derived analysis results. Nucleic Acids Res 38: D699–702. doi: 10.1093/nar/gkp1044
- 75. Madrid M, Soto T, Khong HK, Franco A, Vicente J, et al. (2006) Stress-induced response, localization, and regulation of the Pmk1 cell integrity pathway in Schizosaccharomyces pombe. J Biol Chem 281: 2033–2043. doi: 10.1074/jbc.m506467200
- 76. Cortés JCG, Konomi M, Martins IM, Muñoz J, Moreno MB, et al. (2007) The (1,3)beta-D-glucan synthase subunit Bgs1p is responsible for the fission yeast primary septum formation. Mol Microbiol 65: 201–217. doi: 10.1111/j.1365-2958.2007.05784.x
- 77. Valbuena N, Moreno S (2012) AMPK phosphorylation by Ssp1 is required for proper sexual differentiation in fission yeast. J Cell Sci 125: 2655–64. doi: 10.1242/jcs.098533
- 78. Kurischko C, Kim HK, Kuravi VK, Pratzka J, Luca FC (2011) The yeast Cbk1 kinase regulates mRNA localization via the mRNA-binding protein Ssd1. J Cell Biol 192: 583–598 doi:10.1083/jcb.201011061.
- 79. Rustici G, Mata J, Kivinen K, Lió P, Penkett CJ, et al. (2004) Periodic gene expression program of the fission yeast cell cycle. Nat Genet 36: 809–817 doi:10.1038/ng1377.
- 80. Alberghina L, Mavelli G, Drovandi G, Palumbo P, Pessina S, et al. (2012) Cell growth and cell cycle in Saccharomyces cerevisiae: basic regulatory design and protein-protein interaction network. Biotechnol Adv 30: 52–72 doi:10.1016/j.biotechadv.2011.07.010.
- 81. Kovács IA, Palotai R, Szalay MS, Csermely P (2010) Community Landscapes: An Integrative Approach to Determine Overlapping Network Module Hierarchy, Identify Key Nodes and Predict Network Dynamics. PLoS ONE 5: e12528 doi:10.1371/journal.pone.0012528.
- 82. Orr SJ, Boutz DR, Wang R, Chronis C, Lea NC, et al. (2012) Proteomic and protein interaction network analysis of human T lymphocytes during cell-cycle entry. Mol Sys Biol 8: 573 doi:10.1038/msb.2012.5.
- 83. Fankhauser C, Reymond A, Cerutti L, Utzig S, Hofmann K, et al. (1995) The S. pombe cdc15 gene is a key element in the reorganization of F-actin at mitosis. Cell 82: 435–444. doi: 10.1016/0092-8674(95)90432-8
- 84. Wiley DJ, Marcus S, D'urso G, Verde F (2003) Control of cell polarity in fission yeast by association of Orb6p kinase with the highly conserved protein methyltransferase Skb1p. J Biol Chem 278: 25256–25263 doi:10.1074/jbc.M209703200.
- 85. Szappanos B, Kovács K, Szamecz B, Honti F, Costanzo M, et al. (2011) An integrated approach to characterize genetic interaction networks in yeast metabolism. Nat Genet 43: 656–662 doi:10.1038/ng.846.
- 86. Nguyen T-P, Liu W, Jordán F (2011) Inferring pleiotropy by network analysis: linked diseases in the human PPI network. BMC Syst Biol 5: 179 doi:10.1186/1752-0509-5-179.
- 87. Martin SG (2009) Microtubule-dependent cell morphogenesis in the fission yeast. Trends Cell Biol 19: 447–454 doi:10.1016/j.tcb.2009.06.003.
- 88. Kurischko C, Kuravi VK, Herbert CJ, Luca FC (2011) Nucleocytoplasmic shuttling of Ssd1 defines the destiny of its bound mRNAs. Mol Microbiol 81: 831–849 doi:10.1111/j.1365-2958.2011.07731.x.
- 89. Koyano T, Kume K, Konishi M, Toda T, Hirata D (2010) Search for kinases related to transition of growth polarity in fission yeast. Biosci Biotechnol Biochem 74: 1129–1133. doi: 10.1271/bbb.100223
- 90. Tatebe H, Shimada K, Uzawa S, Morigasaki S, Shiozaki K (2005) Wsh3/Tea4 is a novel cell-end factor essential for bipolar distribution of Tea1 and protects cell polarity under environmental stress in S. pombe. Curr Biol 15: 1006–1015 doi:10.1016/j.cub.2005.04.061.
- 91. Williams T, Brenman JE (2008) LKB1 and AMPK in cell polarity and division. Trends Cell Biol 18: 193–198 doi:10.1016/j.tcb.2008.01.008.
- 92. Waskiewicz AJ, Cooper JA (1995) Mitogen and stress response pathways: MAP kinase cascades and phosphatase regulation in mammals and yeast. Cur Opin Cell Biol 7: 798–805 doi:10.1016/0955-0674(95)80063-8.
- 93. Aves SJ, Durkacz BW, Carr A, Nurse P (1985) Cloning, sequencing and transcriptional control of the Schizosaccharomyces pombe cdc10 “start” gene. EMBO J 4: 457–463.
- 94. Agarwal S, Deane CM, Porter M a, Jones NS (2010) Revisiting date and party hubs: novel approaches to role assignment in protein interaction networks. PLoS Comput Biol 6: e1000817 doi:10.1371/journal.pcbi.1000817.
- 95. Hakes L, Pinney JW, Robertson DL, Lovell SC (2008) Protein-protein interaction networks and biology–what's the connection? Nat Biotechnol 26: 69–72 doi:10.1038/nbt0108-69.
- 96. Rhee SY, Wood V, Dolinski K, Draghici S (2008) Use and misuse of the gene ontology annotations. Nat Rev Genet 9: 509–515 doi:10.1038/nrg2363.
- 97. Newman MEJ (2003) The Structure and Function of Complex Networks. SIAM Review 45: 58 doi:10.1137/S003614450342480.
- 98. Strogatz SH (2001) Exploring complex networks. Nature 410: 268–276 doi:10.1038/35065725.
- 99. Fortunato S (2010) Community detection in graphs. Physics Rep 486: 75–174 doi:10.1016/j.physrep.2009.11.002.
- 100. Hagberg A, Daniel S, Swart P (2008) Exploring Network Structure, Dynamics, and Function using NetworkX. In: Varoquaux G, Vaught T, Millman J, editors. Proceedings of the 7th Python in Science conference; 19–24 August 2008; Pasadena, California, United States. SciPy 2008. pp. 11–15.
- 101. Bähler J, Wu JQ, Longtine MS, Shah NG, McKenzie A, et al. (1998) Heterologous modules for efficient and versatile PCR-based gene targeting in Schizosaccharomyces pombe. Yeast 14: 943–951 doi:10.1002/(SICI)1097-0061(199807)14:10<943::AID-YEA292>3.0.CO;2-Y.
- 102. Sato M, Dhut S, Toda T (2005) New drug-resistant cassettes for gene disruption and epitope tagging in Schizosaccharomyces pombe. Yeast 22: 583–591 doi:10.1002/yea.1233.
- 103. Moreno S, Klar A, Nurse P (1991) Molecular genetic analysis of fission yeast Schizosaccharomyces pombe. Methods Enzymol 194: 795–823. doi: 10.1016/0076-6879(91)94059-l
- 104. De Chaumont F, Dallongeville S, Olivo-Marin J-C (2011) ICY: A new open-source community image processing software. IEEE International Symposium on Biomedical Imaging ISBI. IEEE. pp. 234–237. doi:10.1109/ISBI.2011.5872395.
- 105. Yook S-H, Oltvai ZN, Barabási A-L (2004) Functional and topological characterization of protein interaction networks. Proteomics 4: 928–942 doi:10.1002/pmic.200300636.
- 106. Joy MP, Brock A, Ingber DE, Huang S (2005) High-betweenness proteins in the yeast protein interaction network. J Biomed Biotechnol 2005: 96–103 doi:10.1155/JBB.2005.96.
- 107. Zotenko E, Mestre J, O'Leary DP, Przytycka TM (2008) Why do hubs in the yeast protein interaction network tend to be essential: reexamining the connection between the network topology and essentiality. PLoS Comput Biol 4: e1000140 doi:10.1371/journal.pcbi.1000140.
- 108. Gkantsidis C, Mihail M, Zegura E (2003) The Markov Chain Simulation Method for Generating Connected Power Law Random Graphs. In: Proceedings of the Fifth Workshop on Algorithm Engineering and Experiments; 11 January 2003; Baltimore, Maryland; United States. ALENEX 03. pp. 16–25.
- 109. Sokal RR, Rohlf FJ (1994) Biometry: The Principles and Practices of Statistics in Biological Research. W. H. Freeman.