Personal Use Only Not For Distribution

0 downloads 0 Views 2MB Size Report
A library of pro- teins or peptides is constructed and exposed to selected targets result- .... the understanding of PPI interfaces being undruggable has been.

5 Current Drug Metabolism, 2017, 18, 5-10

REVIEW ARTICLE ISSN: 1389-2002 eISSN: 1875-5453

Protein-Protein Interaction (PPI) Network: Recent Advances in Drug Discovery

Current Drug Metabolism

Impact Factor: 2.847

The international journal for timely in-depth reviews on Drug Metabolism

BENTHAM SCIENCE

Alexiou Athanasios1,2,*, Vairaktarakis Charalampos3, Tsiamis Vasileios4 and Ghulam Md. Ashraf 5,* Novel Global Community Educational Foundation, Australia; 2BiHELab, Department of Informatics, Ionian University, 49100 Corfu, Greece; 3Department of Computer Science and Biomedical Informatics, University of Thessaly, 35100 Lamia, Greece; 4 Department of Mathematics and Computer Science, University of Southern Denmark, Odense, Denmark; 5King Fahd Medical Research Center, King Abdulaziz University, Jeddah, Saudi Arabia

nl y

1

bu tio n

DOI: 10.2174/13892002176661611021506 02

Results: The Protein-Protein Interaction Network' alignment and mapping give the opportunity of further knowledge extraction concerning the evolutionary relationships between the species through conserved pathways and protein complexes. Additionally, Protein-Protein Interaction Network information has been demonstrated to be able to predict functionally orthologous proteins within sequence homology clusters. Our review analysis concluded that, while Protein-Protein Interaction was used to be characterized just by their large and plain interacting surfaces, they were considered inapplicable for drug discovery studies for a long time.

se

Received: February 15, 2016 Revised: February 15, 2016 Accepted: July 19, 2016

Methods: Our in depth review analysis, include Sixty-five peer-reviewed research and review studies from several bibliographic databases. The most significant components were fully described, filtered, combined and analyzed in order to provide documented proofs on the Protein-Protein Interaction Network' applications in biomedicine.

U

ARTICLE HISTORY

O

Abstract: Background: The investigation of the cellular components, their interactions and related functions constitute the major conditions in order to understand the cell as an integrated system. More specifically, the Protein-Protein Interactions and the obtained networks are very important in the majority of biological functions and processes, while most of the proteins appear to activate their functionalities through their interaction.

on al

Conclusion: The present review explores multiple technologies implicated in Protein-Protein Interaction Networks, implicating their potential role in drug discovery mechanisms.

tri

Keywords: Drug discovery, global network alignment, local network alignment, protein-protein interaction network, yeast two-hybrid system.



© 2017 Bentham Science Publishers

N

ot

Fo

rD

is

*Address correspondence to these author at the Novel Global Community Educational Foundation, School of Science, Australia; Tel: +306987467249; E-mail: [email protected] King Fahd Medical Research Center, King Abdulaziz University, P.O. Box 80216, Jeddah 21589, Saudi Arabia; Tel: +966593594931; E-mails: [email protected], [email protected]

or purification steps and can be easily expandable on a genomewide scale. These are the main reasons that this technique became ideal for large-scale PPI screening applications. In yeast two-hybrid system the first protein is expressed as a fusion to a specific DNAbinding domain called the bait, while the second protein is expressed as a fusion to an activation domain called the prey. Bait is linked through a promoter element to a reporter gene, although does not activate it while lacks an activation domain. Prey, on the other hand, is necessary to activate transcription. Transcription of the reporter gene is activated only if the two proteins interact, leading to a color reaction or growth to specific media. Through the years, yeast two-hybrid techniques have been used in many studies concerning PPI networks. Uetz et al. [6] through a comprehensive analysis of PPI in Saccharomyces Cerevisiae by yeast two-hybrid technique successfully expressed 6000 predicted yeast proteins as GAL4 DNA-binding domain fusion proteins detecting 957 putative interactions among 1004 involving Saccharomyces Cerevisiae proteins. Parrish et al. [7] reported a systematic identification of PPI of the bacterium Campylobacter Jejuni. They achieved the detection and reproduction of 11687 interactions and the resulted interaction map included 80% of the predicted Campylobacter Jejuni proteins. Fossum et al. [8] combined interactomes of different species by using yeast two-hybrid system to perform a comparative analysis of these species and extract interactions that cannot be reported in single species. In their study, they tested three herpes virus species (HSV-1, Murine Cytomegalovirus, and Epstein-Barr virus) and identified 735 interactions. Recently, Yu et al. [9] performed yeast two-hybrid screening to identify a group of proteins that interact with Neuroglobin, a globin protein that is expressed specifically in brain neurons. In their study they identified 36 potential Neuroglobin-interacting proteins, from a cDNA library of the mouse brain, which probably relate Neuroglobin with mitochondrial functions, energy metabolism, signaling pathways of multiple cell functions

INTRODUCTION The investigation of the cellular components, the interactions among them and the related functions, constitute the major conditions in order to understand the cell as an integrated system. More specifically, the PPI and the obtained networks are very important in the majority of biological functions and processes, while most of the proteins appear to activate their functionalities through their interaction. The PPI networks' alignment and mapping, even refer to the same or different species, give the opportunity for further knowledge extraction concerning the evolutionary relationships between the species through the conserved pathways and protein complexes [1-2]. Additionally, PPI network information has been demonstrated that is able to predict functionally orthologous proteins within sequence homology clusters [3]. While PPIs were used to be characterized just by their large and plain interacting surfaces, they were considered till the recent years as inapplicable for drug discovery studies, and only recently their attractive opportunities as drug targets have been appeared [4].

Pe rs

Current Drug Metabolism

Send Orders for Reprints to [email protected]

HIGH-THROUGHPUT EXPERIMENTAL TECHNIQUES FOR PPI MAPPING Fields and Song created the yeast two-hybrid system in 1989, a novel genetic system to study the PPI and demonstrate large-scale PPI for organisms [5]. Yeast two-hybrid system is easily automated for high-throughput PPI studies, with no demand for optimization

1875-5453/17 $58.00+.00

6 Current Drug Metabolism, 2017, Vol. 18, No. 1

Athanasios et al.

nl y

gene fragments insert into the phage genome. The resulted modified gene encodes a protein that is expressed on the capsid of the phage. If phagemid vector is used phage components will not be released from host cell before the infection of a helper phage [22]. A library of proteins or peptides is constructed and exposed to selected targets resulting bindings between some members of the library and the exposure target through interactions. Binding step is followed by purification during which phages that didn’t bind removed by washing the immobilized target. The remaining phages can be eluted and used to infect new host cells for amplification. Through the years, several studies have been used phage display technique for direct and indirect display on pIII [23, 24], cDNA display on pVI [25, 26], as well as cDNA display on T7 phage [17, 27]. Phage display technique gives the opportunity for the detection of proteins and peptides that can be used as a drug due to their ability to specifically bind with molecules related to diseases. COMPUTATIONAL METHODS AND ALGORITHMS FOR PPI NETWORK ALIGNMENT We must take into consideration that as the amount of PPI networks data increases, new and more efficient computational methods for their analysis and correlation are required. These processes mostly involve networks alignment, a term that is used to identify and quantify pairs of homologous proteins between two different species. During a network alignment the nodes and the edges of the imported networks are respectively aligned. The alignment’s objective is to find functional orthologous proteins within the PPI networks and between the testing species, to discover conserved sub-network motifs and to learn about the corresponding proteins topology [35]. Over the time, two different approaches for the network alignment have been proposed [2]: the Local Network Alignment (LNA) and the Global Network Alignment (GNA). The LNA aims to detect multiple, unrelated regions of the input networks and extract different local matching between a protein of one network and multiple proteins of a second network. In contrast, the GNA is used to detect a unique correspondence between all nodes and edges of the input networks.

N

ot

Fo

Phage display is a technique for high-throughput screening of protein interactions first reported by George Smith in 1985 [16]. According to this technique the DNA sequence that encodes a specific protein is ligated into the gene that encodes a coat protein of a phage, or in other words, a connection between genotype and phenotype occurs. In his study, Smith fused gene encoding antigenic regions of a protein to DNA encoding the gene 3 coat protein pIII of filamentous M13 phage. A similar approach has been used in several studies based on different phage systems; include T7 phage [17] and Lambda phage [18]. Filamentous M13 phage is composed of circular singlestranded DNA and capsid consisted of five different coat proteins, pIII, pVI, pVII, pVIII and pIX [16]. Minor capsid protein is pIII with up to 5 copies per phage and pVIII is the major one with up to 2700 copies, while circular single-stranded DNA allows the construction of varied libraries of up to 1011 unique members [19, 20]. On the other hand, T7 phage has linear double-stranded DNA, therefore, the construction of T7 phage libraries is more difficult compared to M13 phage libraries. Krumpe et al. [21] demonstrated that T7 system is able to produce more diverse peptide libraries than M13 system due to the differences in the processes of phage morphogenesis between them. The secretion process of the M13 system is taking place through the periplasmic space of the host cell resulting in possible disulfide bonds which is a disadvantage compared to the lytic T7 system. Phage display technique is described by the following steps [16]. Initially, before the phage system is inserted into the host cell,

bu tio n

tri

is

LNA METHODS AND ALGORITHMS PathBLAST [29] adjusts the basic ideas of the BLAST algorithm [30] for protein sequences local alignment, to the detection of local alignments between graphs in order to identify biological pathways. PathBLAST algorithm searches for alignments among pairs of protein interaction paths with a high score, for which the proteins of the first path are paired with orthologs that occur in the same order in the second path. Sharan et al. [31] performed multiple comparisons of PPI networks between 3 different species (Caenorhabditis Elegans, Drosophila Melanogaster, Saccharomyces Cerevisiae) and demonstrated a new method for simultaneous PPI network alignment of multiple species that search for similarities between the networks and identifies the conserved evolutionary pathways and clusters (Fig. 1). This method uses an algorithm that forms an alignment between the input PPI networks and search for conserved protein complexes. All the detected complexes are then assigned a score that depends on the similarities of each complex with a predefined protein complex model. Based on that method Sharan et al. [32] introduced the NetworkBLAST, an on-line platform for the identification of conserved protein complexes among PPI networks, that support both paired and single PPI network analysis. In the paired analysis, unlike single analysis that requires only PPI data for one species, the PPI network data of the two species and a sequence similarity file are required. The output data contain a set of discovered protein complexes and images of those complexes respectively. Koyoturk et al. [33] developed MaWISh, a tool for local PPI network alignment, which follows models of biological duplication and deletion to identify the evolutionary pathways. The framework is seeking for a group of proteins that contain conserved subnets based on a scoring function, -which accounts the

rD

Pe rs

on al

U

se

O

and also may give the opportunity to future researchers to develop therapeutic strategic against stroke and neurological disorders. Another technique that has been already proposed is based on the ability of mass spectrometry to identify and quantify numerous of proteins from proteinaceous complexes. A generic form of this technique has been described in five stages by Aebersold et al. [10]. According to this, initially, the proteins got isolated from the tissues by biochemical fractionation or affinity selection and then degraded to peptides for more sensitive mass spectrometry. The peptides get separated by high-pressure liquid chromatography and nebulized in charged droplets through an electrospray ion source before they enter to the mass spectrometer. Finally, a list of the peptides for a series of consecutively mass spectrometry experiments can be constructed computationally. Krogan et al. [11] focused on the identification of the protein complexes in the yeast Saccharomyces Cerevisiae, by using Tandem Affinity Purification for processing 4562 different tagged proteins and both chromatography-tandem mass spectrometry and matrix-assisted laser desorption/ionization-time of flight mass spectrometry were used for more accurate results. During Tandem Affinity Purification method, tagged bait proteins follow two tandem purification steps to reduce the background binders that are not specific [12]. These researchers [12] successfully identified 7123 PPI among 2708 tagged proteins organized in 547 protein complexes and successfully described a major part of the interactomes and protein complexes of Saccharomyces Cerevisiae proteome. Ewing et al. [13] reported the first large-scale study on PPI in human cells. For that purpose a method based on mass spectrometry has been used to identify the PPI networks for 338 bait proteins related to diseases, resulting 6486 PPI between 2371 proteins. Although, in both yeast twohybrid and mass spectrometry-based studies, non-detection of known PPI and a large amount of false positives have occurred. For example, Gavin et al. [14] in their study based on TAP mass spectrometry for the characterization of proteinaceous complexes in Saccharomyces cerevisiae reported the approximate possibility of 30% false positives among the detected PPI. Yeast two-hybrid was used for the analysis and characterization of binary PPI in contrast of mass spectrometry technique, which provided information about complexes. Thus, Deng et al. [15] suggested that complex datasets of mass spectrometry perform better in functional predictions than physical interactions datasets of yeast two-hybrid method. Both systems are biased towards stable interactions, however, yeast two-hybrid system is effective on detection of weak interactions.

Protein-Protein Interaction Network in Drug Discovery

Current Drug Metabolism, 2017, Vol. 18, No. 1

7

updated topological score. NETAL compared to IsoRank, GRAAL and MI-GRAAL found to has more efficient running time. Recently, Hu et al. [40] reported the development of a new global alignment algorithm for multiple (up to 6 species simultaneously) PPI networks called NetCoffee. The algorithm itself consists of four stages. Initially constructs the PPI network and a bipartite graph library and in the second stage applies a triplet extension approach to set the weight to every edge of these graphs. In the third stage, the algorithm determines the group of the networks' candidate edges and finally maximizes a scoring function as a result of a Metropolis Scheme process.

se

Fig. (1). Example of Drosophila PPI network with sticky gene mutant expression data. Red nodes represent genes that are up-regulated in sticky mutant and green represents down-regulation while yellow nodes represent genes with little to no change in expression. (Reprinted with permission from [28] - http://creativecommons.org/licenses/by/4.0/).

is

tri

Properties of PPI High-affinity bindings take place between small molecules and target proteins. Specifically, according to Wells et al. [42] a piece of the molecule's surface submerges into the target protein where via the connection with the binding pockets of the target protein, the bindings are formed. However, the lack of these pockets in PPI, disturb their applications in the therapeutic field. Although they were recognized as key elements of all the cellular processes, PPI has emerged as potential drug targets just recently [42-44]. Previously, PPI has been well-known not only for their lack of pockets but also for their flat, featureless and large interacting surfaces that made them inconvenient for drug discovery purposes. Over the past decade, investigations have been performed concerning the PPI surfaces with the discovery of the critical role of hotspots [45-47]; the understanding of PPI interfaces being undruggable has been eliminated. Even if the average length of a PPI surface is estimated to range between 1150Å2 and 4660Å2 [48] it was discovered that the center of the interaction between the proteins, is controlled by only a small number of amino acid residues that compose the hotspots and contribute to the required energy for the interactions to occur. These hotspots that dominate the energy of binding are buried on the surface by the other residues [49]. This forms the basis for a promising protein drug design strategy called hotspot grafting. The several hotspot residues can be distinguished by using alanine scanning mutagenesis experiments [46] where the results can be translated as binding with energy difference more than 2kcal/mol. Other studies [50] have shown that conversion of these hotspots, usually tyrosine, arginine and tryptophan into neutral alanine significantly decreases PPI, indicating that intervention of a single residue is sufficient to induce drastic structural changes at interacting protein surfaces. It is obvious, that the perspective of the scientific community concerning the PPI as the target for drug discovery field [51] has been completely changed in recent years.

rD

Pe rs

on al

U

evolutionary events. The scoring function idea extends the concept of the match, mismatch and gap in protein sequence alignment to PPI match and mismatch and duplication of proteins. Berg et al. [34] reported a solution of the local PPI network alignment problem using a Bayesian approach. They developed a method inspired by statistical models that determine the evolution of links and nodes between different species. According to their methods, network alignment is constructed based on a scoring function that determines the similarities between networks. The parameters of high scored alignments are regulated by a systematic Bayesian analysis. Generally, alignment between nodes achieved when their links follow adequately similar pattern, otherwise, no alignment occurs.

bu tio n

O

nl y

ADVANCES OF PPI NETWORKS TOWARDS DRUG DISCOVERY Unwrapping the Molecular Basis of Protein Function to Drug Discovery Protein-protein interactions are a significant tool, as far as the prediction of the result in various cellular processes is concerned. The accurate identification and determination of PPI and the networks correlation is a necessary step in order to understand the mechanisms that take place within the cell, unwrapping the molecular bases of diseases. In general, applications of PPI networks can be characterized efficient mostly if they cover the following four major areas [41]: (i) identification of new disease genes, (ii) study of their network properties, (iii) identification of disease-related sub-networks and (iv) network-based disease classification.

N

ot

Fo

GNA METHODS AND ALGORITHMS IsoRank [35] is the earliest global PPI network alignment algorithm approaching the solution to GNA problem, in correspondence to the way that Google’s PageRank algorithm is measuring the importance of website pages. According to this method, a protein in the first network should be matched with a protein in the second network only if their respective sequences and neighborhood topology can be also well matched. GRAph ALigner (GRAAL) [36] and H-GRAAL [37] algorithms depend exclusively on topological similarities of the networks based on graphlet degree vectors and are absolutely independent of other protein sequences or other information. In other words, a high similarity between two nodes of two different networks is expressed as a high topological similarity of their neighborhoods, which leads to alignment. The basic difference between those two algorithms is that GRALL is a greedy algorithm while H-GRAAL uses Hungarian algorithm. Matching-based Integrative GRAAL (MI-GRAAL) [38] in contrast to GRAAL and HGRAAL produces global alignments between networks based on any kind of similarity measurements between nodes including sequence similarities, topological network similarities as well as structural and functional similarities, resulting more stable alignments than the other GRAAL methods. NETAL [39] is a greedy algorithm that relies on the alignment scoring matrix that comes from both biological and topological information. The specificity of NETAL is the consecutive updating of topological information during the run time of the algorithm i.e. after the alignment of two nodes from different networks, the proteins remained to obtain an

THERAPEUTIC APPLICATIONS AND DRUG DISCOVERY PPI was long thought not to be druggable until recent new reports were published. First, the existence of the so-called hotspots and second, but equally important and innovative, was the scientific report that showed that even if the PPI interface appeared flat in Xray crystallography, in fact, there are small pockets seen in the in molecular dynamic simulation, that stay open for a short time, but

8 Current Drug Metabolism, 2017, Vol. 18, No. 1

Athanasios et al.

tamate receptor GluA2 and PICK1 previously known to intervene at the interaction of PICK1 with GluA2. CONCLUSION Protein to protein interactions network analysis plays a significant role in the identification of relationships between or within species and comprehension of biological properties that reveal functionalities. Several computational LNA and GNA methods presented in this review paper, have already used for PPI networks' alignment and mapping mainly for drug discovery and new treatment solutions. While algorithmic time and space complexity is a crucial factor for the efficiency of biological computational tools, the application of dynamic programming techniques in the addressing of alignment problem, seems to be a promising perspective for future software production.

nl y

LIST OF ABBREVIATIONS GNA = Global Network Alignment LNA = Local Network Alignment PPI = Protein to protein interactions

CONFLICT OF INTEREST The authors confirm that this article content has no conflict of interest. ACKNOWLEDGEMENTS Alexiou Athanasios acknowledges the facilities provided by the Novel Global Community Educational Foundation, Australia. Vairaktarakis Charalampos is grateful to the University of Thessaly, Greece for the facilities and Tsiamis Vasileios is grateful to the University of Southern Denmark for the facilities. Ghulam Md. Ashraf gratefully acknowledges the facilities provided by King Fahd Medical Research Center (KFMRC) and Deanship of Scientific Research (DSR), King Abdulaziz University, Jeddah, Saudi Arabia.

tri

se

REFERENCES [1]

rD

Fo

[2]

[3]

[4]

[5] [6]

N

ot

Kelley, B.P.; Sharan, R.; Karp, R.M.; Sittler, T.; Root, D.E.; Stockwell, B.R.; Ideker, T. Conserved pathways within bacteria and yeast as revealed by global protein network alignment. Proc. Natl. Acad. Sci. USA, 2003, 100, 11394-11399. Sharan, R.; Suthram, S.; Kelley, R.M.; Kuhn, T.; McCuine, S.; Uetz, P.; Sittler, T.; Karp, R.M.; Ideker, T. Conserved patterns of protein interaction in multiple species. Proc. Natl. Acad. Sci. USA, 2005, 102, 1974-1979. Bandyopadhyay, S.; Sharan, R.; Ideker, T. Systematic identification of functional orthologs based on protein network comparison. Genome Res., 2006, 16, 428-435. Arkin, M.R.; Well, J.A. Small-molecule inhibitors of proteinprotein interactions: progressing towards the dream. Nat. Rev. Drug Discov., 2004, 3, 301-317. Fields, S.; Song, O. A novel genetic system to detect proteinprotein interactions. Nature, 1989, 340, 245-246. 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.; Rothberg, J.M. A comprehensive analysis of proteinprotein interactions in Saccharomyces cerevisiae. Nature, 2000, 403, 623-627. Parrish, J.R.; Yu, J.; Liu, G; Hines, J.A.; Chan, J.E. Mangiola, B.A.; Zhang, H.; Pacifico, S.; Fotouhi, F.; DiRita, V.J.; Ideker, T.; Andrews, P.; Finley, R.L. Jr. A proteome-wide protein interaction map for Campylobacter jejuni. Genome Biol., 2007, 8(7), R130. Fossum, E.; Friedel. C.C.; Rajagopala, S.V.; Titz, B.; Baiker, A.; Schmidt, T.; Kraus, T.; Stellberger, T.; Rutenberg, C.; Suthram, S.; Bandyopadhyay, S.; Rose, D.; von Brunn, A.; Uhlmann, M.; Zeretzke, C.; Dong, Y.A.; Boulet, H.; Koegl, M.; Bailer, S.M.; Koszinowski, U.; Ideker, T.; Uetz, P.; Zimmer, R.; Haas, J. Evolutionarily conserved herpes viral protein interaction networks. PLoS Pathog., 2009, 5(9), e1000570.

is

U

on al

Pe rs

bu tio n

O

can be targeted for drug binding [52, 53]. Until now, many experimental techniques and methods have been developed to identify target proteins such as microbial genomics, differential proteomics, nuclear magnetic resonance, cell chip, gene transfection and gene knockout animal models. Unfortunately, these experiments are not only time-consuming but also expensive. Zhan-Chao Li et al. [54] based on graph theory, illustrated the topology information of proteins in the human PPI network with vertex-weighted and edgedweighted methods. With the use of minimal Redundancy Maximal Relevance algorithm in order to select the optimized feature subset and RF algorithm for the construction of the model, Zhan-Chao Li and his team composed a new, verified method which in comparison with the other experimental methods takes a few seconds only to identify whether the protein under study has its place or not in the target proteins. For a period of time, PPI has been targeted with peptides mimicking the interacting domains of proteins with the disadvantage of their poor biochemical and pharmacokinetic properties, like absorption, metabolism and excretion [55, 56], thus making these unsuitable as drugs or drug candidates. Usually, the conversion of a peptide into a drug-like molecule is achieved by creating non-peptidic scaffolds that mimic peptidic motifs by converting side chains into diverse chemical groups, a very lengthy and time-consuming process. Thus, faster and superior processes have been reported in the field of medical chemistry [57, 58] where the protein-binding interface is targeted by chemically conjugated peptide-based fragments or simple small molecules which are highly active in functional PPI. According to [59] the size of these small molecules must be around 750–1500Å2 in order to cover the standard interaction surface. This has resulted in the availability of, large libraries of small molecules that have become easily accessible providing a new base for drug discovery. A suitable example based on that scientific progress was the creation of H/Mdm2 ligands, a drug known for its effective attribute of binding to the tumor suppressor p53 [60]. Even though, the interest of the research investigations did not focus on linking disease-causing mutations with PPIs, Engin et al. [61] published a novel study where the interaction between ELANE and CSF3 have been modeled. Researchers claimed that the energy of PPI concerning CSF3-ELANE may be adjusted by the ELANE’s hotspot which is the variants of 101 amino acid residues, the same variants that are possibly linked with the metastasis advance in breast cancer patients. On the other hand [62] by using a comparative metabolic pathway approach, the same researchers managed to isolate 15 unique pathways out of 119 against Mycobacterium Tuberculosis and thus proposing a list of 18 drug targets for drug designing, thus, providing the possibility for the discovery of a broad spectrum of drugs. To justify this, eighteen out of eighteen targets occur in other pathogenic bacteria like Mycobacterium Leprae, Mycobacterium Bovis, and Mycobacterium Avium Paratuberculosis. It is a common knowledge that ion channels are a major controller of ions flux that enters and leave the cell. Ion channels are large transmembrane proteins, capable of regulating trafficking and biophysical properties of both voltage-gated and ligand-gated ion channels. These types of channels are fully implicated with a variety of human diseases, especially of the Central Neural System. One of the first successful campaigns had a primary target to identify the various molecules that disrupt the Kv1.1 channel [63]. The results from that study led to the identification of diverse compounds with favorable anti-seizure activity in vivo. A more recent generalized approach was centered on the idea of focusing directly on the interacting partner of a particular ion channel in order to locate ligands that could possibly act as competitive inhibitors in the Protein-Channel Interaction site. According to latest studies [64, 65] the discovery of a molecule named FSC231 is the result of the last method. FSC231 is a molecule inhibitor formed by the glu-

[7]

[8]

Protein-Protein Interaction Network in Drug Discovery

[10] [11]

[12]

[25]

[26]

[27]

[28]

[29]

[30]

[18] [19]

[20] [21]

[22] [23]

[24]

[33]

tri

[34]

bu tio n

[32]

is

[35]

[36]

rD

on al

Fo

[17]

ot

[16]

N

[15]

Pe rs

[14]

U

se

[31]

[37]

[38]

[39]

[40]

[41] [42]

[43]

[44] [45] [46] [47] [48]

9

tion of functional gene products linked to the genetic information responsible for their production. Gene, 1993, 137(1), 69-75. Somers, V.A.; Brandwijk, R.J.; Joosten, B.; Moerkerk, P.T.; Arends, J.W.; Menheere, P.; Pieterse, W.O.; Claessen, A.; Scheper, R.J.; Hoogenboom, H.R.; Hufton, S.E. A panel of candidate tumor antigens in colorectal cancer revealed by the serological selection of a phage displayed cDNA expression library. J. Immunol., 2002, 169(5), 2772-2780. Sioud, M.; Hansen, M.H. Profiling the immune response in patients with breast cancer by phage-displayed cDNA libraries. Eur. J. Immunol., 2001, 31(3), 716-725. Yamamoto, M.; Kominato, Y.; Yamamoto, F. Phage display cDNA cloning of protein with carbohybrate affinity. Biochem. Biophys. Res. Commun., 1999, 255(2), 194-199. Bauer, C.R. Epstein, A.M.; Sweeney, S.J.; Zarnescu, D.C.; Bosco, G. Genetic and systems level analysis of Drosophila sticky/citron kinase and dFmr1 mutants reveals common regulation of genetic networks. BMC Syst. Biol., 2008, 2, 101. Kelley, B.P.; Yuan, B.; Lewitter, F.; Sharan, R.; Stockwell, B.R.; Ideker, T. PathBLAST: a tool for alignment of protein interaction networks. Nucleic Acids Res., 2004, 32, W83-W88. Altschul, S.F.; Madden, T.L.; Schäffer, A.A.; Zhang, J.; Zhang, Z.; Miller, W.; Lipman, D.J. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res., 1997, 25(17), 3389-3402. Sharan, R.; Suthram, S.; Kelley, R.M.; Kuhn, T.; McCuine, S.; Uetz, P.; Sittler, T.; Karp, R.M.; Ideker, T. Conserved patterns of protein interaction in multiple species. Proc. Natl. Acad. Sci. USA, 2005, 102, 1974-1979. Kalaev, M.; Smoot, M.; Ideker, T.; Sharan, R. NetworkBLAST: comparative analysis of protein networks. Bioinformatics, 2008, 23(4), 594-596. Koyutürk. M.; Kim, Y.; Topkara, U.; Subramaniam, S.; Szpankowski, W.; Grama, A. Pairwise alignment of protein interaction networks. J. Comput. Biol., 2006, 13, 182-199. Berg J.; Lassig, M. Cross-species analysis of biological networks by Bayesian alignment. Proc. Natl. Acad. Sci. USA, 2006, 103(29), 10967-10972. Singh, R.; Xu, J.; Berger, B. Global alignment of multiple protein interaction networks with application to functional orthology detection. Proc. Natl. Acad. Sci. USA, 2008, 105, 12763-12768. Kuchaiev, O.; Milenkovi, T.; Memievi, V.; Hayes, W.; Prulj, N. Topological network alignment uncovers biological function and phylogeny. J. R. Soc. Interface, 2010, 7(50), 1341-1354. Milenkovi, T.; Leong, W.; Hayes, W.; Prulj, N. Optimal network alignment with graphlet degree vectors. Cancer Inform., 2010, 9, 121-137. Kuckaiev, O.; Przulj, N. Integrative network alignment reveals large regions of global network similarity in year and human. Bioinformatics, 2011, 15, 27(10), 1390-1396. Neyshabur, B.; Khadem, A.; Hashemifar, S.; Arab, S.S. NETAL: a new graph-based method for global alignment of protein-protein interaction networks. Bioinformatics, 2013, 29(13), 1654-1662. Hu, J.; Kehr, B.; Reinert, K. NetCoffee: a fast and accurate global alignment approach to identify functionally conserved proteins in multiple networks. Bioinformatics, 2014, 30(4), 540-548. Ideker, T.; Sharan, R. Protein networks in disease. Genome Res., 2008, 18, 644-652. Wells, J.A.; McClendon. C.L. Reaching for high-hanging fruit in drug discovery at protein-protein interfaces. Nature, 2007, 13, 459(7172), 1001-1009. Arkin, M.R.; Wells, J.A. Small-molecule inhibitors of proteinprotein interaction: progressing towards the dream. Nat. Rev. Drug Discov., 2004, 3(4), 301-317. Fry, D.C. Protein-protein interactions as targets for small molecule drug discovery. Biopolymers, 2006, 84(6), 535-552. Bogan, A.A.; Thorn, K.S. Anatomy of hot spots in protein interfaces. J. Mol. Biol., 1998, 280(1), 1-9. Clackson, T.; Wells, J.A. A hot spot of binding energy in a hormone-receptor interface. Science, 1995, 267(5196), 383-386. Delano, W.L. Unraveling hot spots in binding interfaces: process and challenges. Curr. Opin. Struct. Biol., 2002, 12(1), 14-20. Lo Conte, L.; Chothia, C.; Janin, J. The atomic structure of proteinprotein recognition sites. J. Mol. Biol., 1999, 285, 2177-2198.

O

[13]

Yu, Z.; Liu, N.; Wang, Y.; Li, X.; Wang, X. Identification of neuroglobin-interacting proteins using yeast two-hybrid screening. Neuroscience, 2012, 200, 99-105. Aebersold, R.; Mann, M. Mass spectrometry-based proteomics. Nature, 2003, 422, 198-207. Krogan, N.J.; Cagney, G.; Yu, H.; Zhong, G.; Guo, X.; Ignatchenko, A.; Li, J.; Pu, S.; Datta, N.; Tikuisis, A.P.; Punna, T.; Peregrín-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.; Greenblatt, J.F. Global landscape of protein complexes in the yeast Saccharomyces cerevisiae. Nature, 2006, 440, 637-643. Rigaut, G.; Shevchenko, A.; Rutz, B.; Wilm, M.; Mann, M.; Séraphin, B. A generic protein purification method for protein complex characterization and proteome exploration. Nat. Biotechnol., 1999, 17, 1030-1032. Ewing, R.M.; Chu, P.; Elisma, F.; Li, H.; Taylor, P.; Climie, S.; McBroom-Cerajewski, L.; Robinson, M.D.; O'Connor, L.; Li, M.; Taylor, R.; Dharsee, M.; Ho, Y.; Heilbut, A.; Moore, L.; Zhang, S.; Ornatsky, O.; Bukhman, Y.V.; Ethier, M.; Sheng, Y.; Vasilescu, J.; Abu-Farha, M.; Lambert, J.P.; Duewel, H.S.; Stewart, I.I.; Kuehl, B.; Hogue, K.; Colwill, K.; Gladwish, K.; Muskat, B.; Kinach, R.; Adams, S.L.; Moran, M.F.; Morin, G.B.; Topaloglou, T.; Figeys, D. Large-scale mapping of human protein-protein interactions by mass spectrometry. Mol. Syst. Biol., 2007, 3(89), Epub 2007 Mar 13. Gavin, A.C.; Bösche, M.; Krause, R.; Grandi, P.; Marzioch, M.; Bauer, A.; Schultz, J.; Rick, J.M.; Michon, A.M.; Cruciat, C.M.; Remor, M.; Höfert, 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.; Superti-Furga, G. Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature, 2002, 415, 141-147. Deng, M.; Sun, F. Chen, T. Assessment of the reliability of proteinprotein interactions and protein function prediction. Pac. Symp. Biocomput., 2003, 140-151. Smith, G.P. Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface. Science, 1985, 228(4705), 1315-1317. Danner, S.; Belasco, J.G. T7 phage display: A novel genetic selection system for cloning RNA-binding proteins from cDNA libraries. Proc. Natl. Acad. Sci. USA, 2001, 98(23), 12954-12959. Kuwabara, I.; Maruyama, H.; Mikawa, Y.G.; Zuberi, R.I.; Liu, F.T.; Maruyama, I.N. Efficient epitope mapping by bacteriophage lambda surface display. Nat. Biotechnol., 1997, 15(1), 74-78. Malik, P.; Terry, T.D.; Gowda, L.R.; Langara, A.; Petukhov, S.A.; Symmons, M.F.; Welsh, L.C.; Marvin, D.A.; Perham, R.N. Role of capsid structure and membrane protein processing in determining the size and copy number of peptides displayed on the major coat protein of filamentous bacteriophage. J. Mol. Biol., 1996, 260(1), 9-21. Weiss, G.A.; Sidhu, S.S. Design and evolution of artificial M13 coat proteins. J. Mol. Biol., 2000, 300, 213-219. Krumpe, L.R.; Atkinson, A.J.; Smythers, G.W.; Kandel, A.; Schumacher, K.M.; McMahon, J.B.; Makowski, L.; Mori T. T7 lytic phage-displayed peptide libraries exhibit less sequence bias than M13 filamentous phage-displayed peptide libraries. Proteomics, 2006, 6, 4210-4222. Mead, D.A.; Kemper, B. Climeric single-stranded DNA phageplasmid cloning vectors. Biotechnology, 1988, 10, 85-102. van Zonneveld, A.J.; van den Berg, B.M.; van Meijer, M.; Pannekoek, H. Identification of functional interaction sites on proteins using bacteriophage-displayed random epitome libraries. Gene, 1995, 29, 167(1-2), 49-52. Crameri, R.; Suter, M. Display of biologically active proteins on the surface of filamentous phage: a cDNA cloning system for selec-

nl y

[9]

Current Drug Metabolism, 2017, Vol. 18, No. 1

10 Current Drug Metabolism, 2017, Vol. 18, No. 1

nl y

[63]

[65]

bu tio n

O

[64]

tri

[58]

[62]

Miatkowski, K.; Nguyen, C.; Nguyen, T.B.; Park, S.; Pathan, N.; Penny, D.M.; Romanowski, M.J.; Scott, D.; Silvian, L.; Simmons, R.L.; Tangonan, B.T.; Yang, W.; Sun, L. Discovery of a potent and highly selective PDK1 inhibitor via fragment-based drug discovery. Bioorg. Med. Chem. Lett., 2011, 21, 3078-3083. Blazer, L.L.; Neubig, R.R. Small molecule protein-protein interaction inhibitors as CNS therapeutic agents: current progress and future hurdles. Neuropsychopharmacology, 2009, 34(1), 126-141. Acuner Ozbabacan, S.E.; Gursoy, A.; Keshin, O.; Nussinov, R. Conformational ensembles, signal transduction and residue hot spots: application to drug discovery. Curr. Opin. Drug Discov. Devel., 2010, 13(5), 527-534. Engin, H.B.; Gursoy, A.; Nussinov, R.; Keskin, O. Network-based strategies can help mono- and poly-pharmacology drug discovery: a systems biology view. Curr. Pharm. Des., 2014, 20, 1201-1207. Kushwaha, S.K.; Shakya, M. Protein interaction network analysisapproach for potential drug target identification in Mycobacterium tuberculosis. J. Theor. Biol., 2010, 262(2), 284-294. Lu, Q.; Peevey, J.; Jow, F.; Monaghan, M.M.; Mendoza, G.; Zhang, H.; Wu, J.; Kim, C.Y.; Bicksler, J.; Greenblatt, L.; Lin, S.S.; Childers, W.; Bowlby, M.R. Disruption of Kv1.1 N-type inactivation by novel small molecule inhibitors (disinactivators). Bioorg. Med. Chem., 2008, 16, 3067-3075. Stoilova-McPhie, S.; Ali, S.; Laezza, F. Protein-protein interactions as new targets for ion channel drug discovery. Austin J. Pharmacol. Ther., 2013, 1(2), 5. Thorsen, T.S.; Madsen, K.L.; Rebola, N.; Rathje, M.; Anggono, V.; Bach, A.; Moreira, I.S.; Stuhr-Hansen, N.; Dyhring, T.; Peters, D.; Beuming, T.; Huganir, R.; Weinstein, H.; Mulle, C.; Strømgaard, K.; Rønn, L.C.; Gether, U. Identification of a small-molecule inhibitor of the PICK1 PDZ domain that inhibits hippocampal LTP and LTD. Proc. Natl. Acad. Sci. USA, 2010, 107, 413-418.

is

[57]

[61]

se

[56]

[60]

rD

[55]

[59]

U

[54]

Fo

[53]

ot

[52]

on al

[51]

N

[50]

Moreina, P.I.; Santos, M.S.; Oliveira, C.R. Alzheimer’s disease: a lesson from mitochondrial dysfunction. Antioxid. Redox Signal., 2007b, 9, 1621-1630. Bogan, A.A.; Thorn, K.S. Anatomy of hot spots in protein interfaces. J. Mol. Biol., 1998, 280, 1-9. Jubb, H.; Higueruelo, A.P.; Winter, A.; Blundell, T.L. Structural biology and drug discovery for protein-protein interactions. Trends Pharmacol. Sci., 2012, 33, 241-248. Eyrisch, S.; Helms, V. Transient pockets on protein surfaces involved in protein-protein interaction. J. Med. Chem., 2007, 50(15), 3457-3464. Eyrisch, S.; Helms, V. What induces pocket openings on protein surface patches involved in protein–protein interactions? J. Comput. Aided Mol. Des., 2009, 23(2), 73-86. Li, Z.C.; Zhong, W.Q.; Liu, Z.Q.; Huang, M.H.; Xie, Y.; Dai, Z.; Zou, X.Y. Large-scale identification of potential drug targets based on the topological features of human protein–protein interaction network. Analytica Chimica Acta, 2015, 29(871), 18-27. Veber, D.F.; Saperstein, R.; Nutt, R.F.; Freidinger, R.M.; Brady, S.F.; Curley, P.; Perlow, D.S.; Paleveda, W.J.; Colton, C.D.; Zacchei, A.G.; et al. A super active cyclic hexapeptide analog of somatostatin. Life Sci., 1984, 34, 1371-1378. Diamond, J.M.; Wright, E.M. Biological membranes: the physical basis of ion and nonelectrolyte selectivity. Annu. Rev. Physiol., 1969, 31, 581-646. Hruby, V.J.; Cai, M.; Cain, J.; Nyberg. J.; Trivedi, D. Design of novel melanocortin receptor ligands: multiple receptors, complex pharmacology, the challenge. Eur. J. Pharmacol., 2011, 660, 8893. Erlanson, D.A.; Arndt, J.W.; Cancilla, M.T.; Cao, K.; Elling, R.A.; English, N.; Friedman, J.; Hansen, S.K.; Hession, C.; Joseph, I.; Kumaravel, G.; Lee, W.C.; Lind, K.E.; McDowell, R.S.;

Pe rs

[49]

Athanasios et al.