Material and methods Interaction Network Construction

Gallus gallus was used as a model organism in a chemical and biological interaction network built to better understand the effects of OCs com pounds on the physiological alteration related to the reproduction of birds at a molecular level. To achieve that, we performed a literature review and identified two OCs compounds to which predic- tion in G. gallus is available. Then, we searched for compound-protein interactions in the STITCH v.5.0 database (http://stitch.embl.de) (Szklarczyk et al. 2016). The STRING 10.5 search tool (http:// string-db.org) was used to find protein-protein interactions (Snel 2000; Kuhn et al. 2008; Szklar czyk et al. 2017) information on such interactions is widely dispersed across numerous databases and the literature. To facilitate access to this data, STITCH (‘search tool for interactions of chemi- cals’.) The interaction between chemical com pounds and proteins in STITCH is based on experimental evidence, databases, and published data (Kuhn et al. 2008). The following parameters were used to import sub-networks from STITCH: 50 interactions or less, medium confidence score (0.400), and network depth equal to 2; all prediction methods were activated except for text min- ing, gene fusion, and co-occurrence. This initial process generated two small compound-protein interaction sub-networks. Then, we searched for protein interactions in STRING 10.5. This web tool predicts protein interactions that can be directly (physically) or indirectly (functionally) associated (Snel 2000). The following parameters were used: no more than 100 interactions, medium confi dence score (0.400), and network depth equal to 2; all prediction methods were activated except for text mining, neighborhood, gene fusion, and cooccurrence. The distinct sub-networks generated in these two processes were joined in a single sub-network denominated Organochlorine-Protein Interaction Network (OPIN), using the Cytoscape 3.4.0 Advanced Merge Network software (Shan- non et al. 2003).

Topological Analysis

OPIN was analyzed by the Cytoscape 3.4.0 plugin MCODE (Molecular Complex Detection) (^{Bader and Hogue 2003}) phage display and mass spectrometry have enabled us to create a detailed map of biomolecular interaction networks. Initial mapping efforts have already produced a wealth of data. As the size of the interaction set increases, databases and computational methods will be required to store, visualize and analyze the information in order to effectively aid in knowledge discovery.This paper describes a novel graph theoretic clustering algorithm, \”Molecular Complex Detection\” (MCODE aimed at identifying modules/clusters (densely connected network regions) that suggest functional protein complexes. The following parameters were applied to MCODE to generate clusters visualization: loops included; degree cutoff 2; one connection node exclusion (haircut option activated); cluster expansion by one neighbor shell allowed (fluff option activat- ed); node density cutoff 0.1; node score cutoff 0.2; kcore 2; and maximum network depth 100. An MCODE score was calculated for each OPIN protein/compound.

Gene Ontology Analysis

The main processes associated with the clusters generated by the MCODE were analyzed by the Cytoscape 3.4.0 plugin BiNGO (Biological Network Gene Ontology) (Maere et al. 2005). The equation degree of a given cluster and category was quantitatively computed p-value) by hypergeo- metric distribution. Multiple correction tests were applied using the false discovery rate algorithm (Benjamini y Hochberg 1995) researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholar- ship. For more information about JSTOR, please contact support@jstor.org. SUMMARY The com- mon approach to the multiplicity problem calls for controlling the familywise error rate (FWER performed by BiNGO with a 5% significance level.

Centrality Analysis

Centrality analysis was performed using the Centiscape 2.1 plugin (Scardoni et al. 2009) and Cyto scape 3.4.0 to identify the nodes (proteins) with a central position within the network. Node degree and betweenness were the analyzed centralities; nodes with a relatively high degree are named ‘hubs’ and nodes with a relatively high betweenness are named ‘bottlenecks’. Hubs are highly connected nodes, while bottlenecks are more probable to connect different clusters (^{Barabasi and Oltvai 2004}; Yu et al. 2007). Therefore, a hub-bottleneck node (HB-N) can be considered a key regulator of biological processes, extremely important for successful information transferring within the network. Hence, disturbances in an HB-N can cause communication failures within the network.

Results and discussion

In recent years, the increasing number of “omic” databases has enabled the use of computational tools to understanding the interplay between the regulatory processes acting on a cell (Feltes et al. 2011). Based on previous studies that associated OCs exposure to reproductive failure in birds (Car- son and Darling 1962), we built an organochlorine- protein network interaction to investigate and to propose the molecular mechanism associated to the effect of organochlorines on birds.

Our OPIN network had 1,417 interactions and 137 nodes (Figure 1), of which 135 are G. gal- lus proteins and two are OCs compounds (DDT and 2,3,7,8-tetrachlorodibenzo-p-dioxin - TCDD) prospected from the literature review. Due to the scarce STITCH database about OCs compounds, only two compounds were prospected. In relation to network analysis had: clustering coefficient equal to 0.541, diameter to equal 4, radius equal to 3, centralization equal to 0.45, the shortest paths equal to 18632, characteristic path length equal to 2,172, avg. number of neighbors equal to 20,686, density equal to 0.152, heterogeneity equal to 0.669 and closeness equal to 0.0034. The MCODE analysis resulted in four clusters, a set of nodes connected to each other, very common in biological systems since cellular processes are ruled by biomolecules that form interaction modules (^{Pavlopoulos et al. 2011}. The BiNGO plugin indicated that the main metabolic processes associated to the cluster 1 (cluster with the highest score) are involved in the reproductive success and in the development of organs and tissues (Table 1). Hub-bottleneck proteins are nodes with the highest relevance inside a network and have a large number of connections capable of link- ing distinct clusters. The four main HB-N (UBB, ESR1, and HSP90AB1 proteins and organochlo- rine TCDD) found by the centrality analysis are shown in Figure 2.

Figure 1: Organochlorine-protein interaction network (OPIN) of the model Gallus gallus with 137 nodes and 1,417 edges. The blue nodes represent the chemical compounds DDT and TCDD, grey nodes represent the proteins, and the red nodes represent the two main hub-bottleneck proteins. The edges between nodes represent the interaction display of this biological system. 

Figure 2: Hubs-bottlenecks (HB) graph of the most important network nodes. The horizontal axis shows the betweenness values and the vertical axis shows the degree values. The mean of each centrality is represented by the dashed lines. The dots beyond the dashed lines are the HB nodes, highlighting the proteins ESR1, HSP90AB1, UBB, and the TCDD organochlorine. 

TCDD was found with one of the main hub-bottleneck proteins because it is able to connect a large number of clusters within a network, evidencing the great impact of this compound in a biological system. UBB is known as polyubiquitin b and plays an important role in proteins degradation, but also acts as a regulator of several biological mechanisms such as gene expression and stress responses. UBB is a highly conserved protein that plays critical roles in the functioning of eukaryotic cells (Hochstrasser 1996) and is frequently present as a HB-N in the system biology.

The other two main HB-N found in the analysis were the proteins ESR1 nd HSP90AB1. eSr1 (estrogen receptor 1) is responsible for the estro- gen reception in the cell. Once bound, the hormone and its receptor activate a series of signaling events involved in the genetic regulation that affect tissues and cellular proliferation (www. uniprot.org/uniprot/P06212) (Bateman et al. 2017).

Table 1: Gene Ontology (GO) classes derived from the cluster 1 network. GO identification number (ID), the significance value of each process, the bioprocess of each cluster, and the proteins involved are displayed. The main hub-bottleneck proteins are displayed in bold letters. 

Figure 3: A molecular model illustrating the effect of organochlorines on exposed birds. The organochlorine exposure probably leads to an increase in HSP90 proteins expression, inhibiting the hormone-binding conformation and consequently favoring the ESRI - organochlorine interaction. Thus, the disruption of the ESR1- estrogen interaction creates a calcium unavailability for eggshell formation, causing the thin-shell syndrome. 

(ESR1 UniProt code: P06212). HSP90AB1 is a chaperone protein that promotes the regulation of specific targets involved in the cell cycle and signal transduction, leading to a conformational change in target-proteins, causing their activation (HSP90AB1 UniProt code: Q04619). According to ^{Fliss et al. (2000}), HSP90 acts on maintaining the estrogen receptor in a high-affinity hormone- binding conformation, ensuring that the ESR1- hormone system performs the actions resulted from the hormonal activity. Animals exposed to negative environmental conditions, such as pes- ticide exposure, have been described to exhibit increased levels of HSP40, HSP70, and HSP90 (^{Brom et al. 2015}).

DDT and other OCs have mimicking action; they are capable of binding to the same receptors as estrogen, including ESR1. Once DDT compounds bind to estrogen receptors they can activate them,generating a signaling event cascade and the syn- thesis of certain cellular products (Jaga 2000). The best-documented case of the effects of environmental pollutants on birds is the thin-shell syndrome caused by dichlorodiphenyl (DDE) - a chemical derivative of DDT - particularly in spe- cies that feed on fishes, leading to a reproduc- tive failure (^{Gress et al. 1973}). The species more susceptible to the thin-shell syndrome appear to have a reduced ability to metabolize organochlo- rines (^{Schwarzbach et al. 1991}) p’ dicofol, p,p’ dichlorobenzophenone (p,p’ DCBP. Moreover, estrogen and thyroid hormones that control calcium availability are involved in the eggshell formation (Hoffmann and Volker 1969).

Thus, birds contaminated with OCs pesticides probable have an increased expression of HSP90 that inhibits the hormone-binding conformation and consequently favors ESRI-OCs interaction.

The result of the disrupted ESR1-estrogen inter- action is he calcium unavailability for eggshell formation, which leads to the thin-shell syndrome (Figure 3). As previously described, the syndrome is related to the calcium inhibition in the eggshell gland by DDE, an organochlorine very similar to DDT (Kolaja 1979). Derfoul et al. (2003), using hu

the estrogen-regulated calcium transport pathway. Furthermore, some studies provided novel insights into the roles of HSP90 and calcium transport. Li et al. (2014) observed that HSP90 affects the in- tracellular calcium homeostasis in human sperm. Organochlorine compounds were banned in the USA in the early 1970s. Even so, DDT and other organochlorine compounds are still found in eggs and carcasses of birds and in birds in- habiting urban and suburban areas (Espín 2010; Martínez 2015; Mora and col. 2016; Yohannes 2016; Schmitt and col. 2018). Organochlorine contamination has been recorded worldwide, in Europe (Espín 2010), Africa (Yohannes 2016), and South America (Martínez 2015). This may be occurring due to the bioaccumulative feature of these compounds; they remain in the food chain of ecosystems, including human’s. Since there are countries that continue to use these compounds on a large scale (^{Berg et al. 2017}), they are still a threat to the wildlife.