About this Journal Submit a Manuscript Table of Contents
International Journal of Genomics
Volume 2013 (2013), Article ID 456356, 7 pages
Research Article

Protein-Protein Interactions Inferred from Domain-Domain Interactions in Genogroup II Genotype 4 Norovirus Sequences

1Department of Computer Sciences, National Tsing Hua University, Hsinchu 30013, Taiwan
2Department of Computer Sciences and Information Engineering, Providence University, Taichung 43301, Taiwan

Received 28 January 2013; Accepted 1 April 2013

Academic Editor: Chun-Yuan Lin

Copyright © 2013 Chuan-Ching Huang and Chuan Yi Tang. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


Severe gastroenteritis and foodborne illness caused by Noroviruses (NoVs) during the winter are a worldwide phenomenon. Vulnerable populations including young children and elderly and immunocompromised people often require hospitalization and may die. However, no efficient vaccine for NoVs exists because of their variable genome sequences. This study investigates the infection processes in protein-protein interactions between hosts and NoVs. Protein-protein interactions were collected from related Pfam NoV domains. The related Pfam domains were accumulated incrementally from the protein domain interaction database. To examine the influence of domain intimacy, the 7 NoV domains were grouped by depth. The number of domain-domain interactions increased exponentially as the depth increased. Many protein-protein interactions were relevant; therefore, cloud techniques were used to analyze data because of their computational capacity. The infection relationship between hosts and NoVs should be used in clinical applications and drug design.

1. Introduction

Gastroenteritis and foodborne illness epidemics usually occur during the winter. In 1968, an acute gastroenteritis outbreak occurred at an elementary school in Norwalk, OH, USA [1]. Kapikian et al. used immune electron microscopy and an infectious stool filtrate to observe the virus particles and identified them as small round-structured viruses [2]. The nonbacterial pathogen was called the Norwalk virus after the outbreak region and renamed the Norovirus (NoV) at the International Congress of Virology in Paris in 2002. The NoV is the species of the genus of Norwalk-like viruses in the Caliciviridae family.

The NoV consists of a positive-sense, single-strand RNA (ssRNA) genome sequence of approximately 7.5 kb with 3 open reading frames (ORFs) [3]. The first open reading frame (ORF1) contains approximately 5 kb of sequences, which may enable 195 kDa polyproteins to support virus duplication. During the duplication phase, polyproteins are divided into at least 6 nonstructural proteins, such as p48, nucleotide triphosphatase (NTPase), p22, Vpg, proteinase, and RNA-dependent RNA polymerase [46]. ORF2 is approximately 1.8 kb long, which may transform into 60 kDa major structural protein VP1 with the following functions: self-assembly and capsid formation, receptor recognition, host specificity, strain diversity, and immunogenicity [7]. ORF3 consists of 0.6 kb nucleotides, which may translate into a 20 kDa minor structural protein (VP2), which facilitates the expression and stability of major structural protein VP1 [4, 8].

The NoV structure is formed by 180 capsid protein monomers organized into 90 dimeric capsomers [9]. Each capsid protein consists of an N-terminal shell (S) domain and a C-terminal protrusion (P) domain. The S domain forms a contiguous 8-stranded beta-barrel shell of the capsid and is commonly found in the capsid proteins of other viral families. The arch-like structure extending from the shell is formed by the P domain, which can be composed of P1 and P2 subdomains. The P2 subdomain is the critical segment because of its location on the capsid surface and because it exhibits the most variable sequence.

Because traditional cell cultures cannot replicate NoVs [10], the phylogenetic relationship is mainly investigated by using genome-sequencing techniques. Based on the phylogenetic analysis of genome sequences, 5 genogroups exist: GI, GII, GIII, GIV, and GV. GIII is found in cattle [11, 12], GV is found in mice [13], and human NoVs are divided into 2 major genogroups, GI and GII.

High-throughput sequencing techniques allow several genome sequences to be investigated simultaneously. This enables the mutation tendency of diverse NoV sequences to be examined. The sequence search tool on the Pfam website was used to collect approximately 30 NoV GII genotype 4 sequences to find their corresponding Pfam domains. The 7 main families are Cacli_PP_N (PF08405), RNA helicase (PF00910), Peptidase_C37 (PF05416), RdRP_1 (PF00680), Calici_coat (PF00915), Calici_coat_C (PF08435), and RNA_capsid (PF03035). Because the P2 subdomain is associated with antigenicity by binding to histoblood group antigen receptors, the infection relationship may be hidden by the interactions of Pfam domains extracted from NoV sequences. This study attempts to distinguish the protein-protein interactions (PPIs) from the domain-domain interactions (DDIs) and to use the PPIs to determine the biological regulation of NoVs. Many PPIs are expected, requiring high-capacity computation.

Cloud computing enables high-capacity computation. Cloud computing allows data to be collaboratively computed by different computer systems. Three main types of cloud computing services exist: infrastructure as a service (IaaS), software as a service (SaaS), and platform as a service (PaaS). User requirements determine the type of service used. When managing many PPIs [14], the distributed characteristic of cloud techniques enhances the efficiency of exploring the biological regulation caused by NoVs. This provides an opportunity to examine the relationship between hosts and NoVs and to investigate clinical applications and drug design.

2. Materials and Methods

Seven Pfam [15] domains were extracted from approximately 30 genome sequences from GII genotype 4 using the sequence search tool on the Pfam website. Four domains (PF08405, PF00910, PF05416, and PF00680) exist in ORF1 (Figure 1). Two domains (PF00915 and PF08435) exist in ORF2, and one domain (PF03035) exists in ORF3.

Figure 1: The domain composition of a NoV sequence which is aligned by the sequence search tool on the Pfam website.

The DOMINE database (version 2.0) [16] was used to investigate interactions between domains, including known and predicted protein domain interactions. For each of the 7 NoV domains, if a related DDI was found in DOMINE, the interaction and interactive domains were recorded. The interactive domains and the 7 NoV domains were grouped into a domain set of Depth 1. The interactive domain set with interactions in the Depth 1 domain set was denoted as the domain set of Depth 2. This method was used to collect domain sets of increasingly higher depth (Figure 2).

Figure 2: Illustration of DDIs from 7 NoV-related domains.

The number of domains and DDIs increases as the depth increases. The domains and interactions from Depths 1 to 9 are shown in Figure 3. Depth 2 has 17 times more domains and 27 times more interactions than Depth 1. Depth 3 has 10 times more domains and 24 times more interactions than Depth 2. The number of domains and the number of interactions plateau gradually after Depth 4.

Figure 3: Population of domains and interactions within each depth.

The set of DDIs within each depth was used to locate corresponding proteins for each domain and to construct the putative PPI network. Figure 4 shows the PPIs inferred from the DDIs within Depth 1. Corresponding proteins were collected by either using the NoV-related domains or by extending the domains within Depth 1 to build a PPI network to identify similar metabolic pathways.

Figure 4: Illustration of the PPI network extending from DDIs within Depth 1.

For each domain from a certain depth, all related proteins were extracted from the UniProt Knowledgebase (release 2011_07) [17]—a comprehensive database containing protein sequence and function information. Proteins were grouped by the Pfam domain (version 25.0) from the integrated source database, InterPro (version 33.0) [18].

An attempt was made to find connections between proteins based on the DDIs and to explore their accompanying protein functions. However, the protein interaction network was too complex to identify the proteins relevant to the NoVs. It would be beneficial to retain the most common PPIs that have been verified by the expert curators of the molecular interaction database, MINT [19]. Human proteins consisting of Pfam domains were selected to investigate the interaction relationship because the NoV uses people as hosts.

3. Results

To examine the DDI strength, the depth is defined as the distance of a domain from the previous domain. Depth 1 includes 13 domains and 12 interactions (Figure 3). However, the number of domains associated with interactions increases by almost 17 times in Depth 2. The number of domains and interactions only increase slightly in Depth 4, meaning that the number of DDIs reached a boundary value.

The DDIs in Depth 1 are illustrated using Cytoscape software [20] in Figure 5. Four Pfam domains (PF00910, PF00680, PF08435, and PF00915) were found in the interaction network of the 7 NoV-related domains. The first 2 domains are from ORF1, and the last 2 domains are from ORF2.

Figure 5: The domain-domain interaction network within Depth 1.

The DDIs in Depth 2 are more complex than those from Depth 1. The 4 NoV-related domains are shown in the top-right corner of Figure 6. The complex interactions are caused by 2 undirected domains (PF00270 and PF00271), which are centers for undirected domains and connect several related domains.

Figure 6: The domain-domain interaction network within Depth 2.

To show the complexity of the PPI network from each depth, PPI statistics and distinct proteins are listed in Table 1. From Depth 2 to Depth 3, the number of PPIs increases rapidly, and the number of proteins almost triples. Of the 1061 proteins in Depth 2, 107 proteins have higher links into other proteins, including ATP-dependent RNA helicase, DNA annealing helicase and endonuclease, putative DNA repair and recombination protein, and the transcription termination factor.

Table 1: The statistics of PPIs and proteins within each depth.

The PPI network from Depth 2 is shown in Figure 7 to show the interaction complexity between proteins. The 3 circles in the figure do not represent the proteins directly related to the 7 NoV domains and proteins in Depths 1 and 2. It is possible that proteins closer to the center have more putative linkages with other proteins. Infection may occur between NoVs and human hosts through some of these PPIs.

Figure 7: The protein-protein interaction network within Depth 2.

4. Discussion

High-throughput genome-sequencing technologies generate vast amounts of sequence data efficiently and economically. Personalized medication could be produced by analyzing individual genome sequences. Domains are the fundamental building blocks of proteins [21]. They play important roles in living organisms such as catalysis, signal transduction, and the transport of nutrients [22]. This study uses the Pfam domain database. The database contains protein domain families that share significant degrees of sequence similarity. This approach assumes that one type of domain combination represents a protein and that the interaction between proteins could be performed through DDIs. Exploring protein regulation by using DDIs is reasonable.

Although domains are smaller units than proteins, extending PPIs from the domain level is feasible. This causes 2 main problems: the vast number of protein connections from DDIs and the ability to examine available PPIs for NoVs. Data integration and relationship establishment require extremely complex computations. Cloud computing techniques are available to overcome these obstacles. Applying biological knowledge to organism specificity and infection pathways would assist in comprehending NoVs. This may help researchers develop medical therapies for NoVs.

5. Conclusions

Chemical reactions in living systems are performed by proteins. Understanding protein regulation may assist with understanding disease progression. Domains are basic protein folding and evolution units; therefore, investigating internal DDI relationships is valuable. For this study, we extracted NoV-related domains from the Pfam database to investigate the biological processes during NoV infection phases.

NoVs can cause acute gastroenteritis and foodborne illnesses, affecting lives and increasing health care costs. However, no efficient vaccine exists for NoVs. Exploring the regulation of PPIs extending from domains may help identify the relationship between hosts and NoVs. This information could be used in clinical applications and drug design.

Conflict of Interests

The authors of this paper have no competing interests to declare.


The authors are grateful to Professor Chi-Neu Tsai and to suggestions from anonymous reviewers. This work was supported by the National Science Council, China (Grant nos. NSC100-2221-E-126-011-MY3 and NSC100-2221-E-182-057-MY3).


  1. J. L. Adler and R. Zickl, “Winter vomiting disease,” Journal of Infectious Diseases, vol. 119, no. 6, pp. 668–673, 1969. View at Scopus
  2. A. Z. Kapikian, R. G. Wyatt, R. Dolin, T. S. Thornhill, A. R. Kalica, and R. M. Chanock, “Visualization by immune electron microscopy of a 27-nm particle associated with acute infectious nonbacterial gastroenteritis,” Journal of Virology, vol. 10, no. 5, pp. 1075–1081, 1972. View at Scopus
  3. X. Jiang, J. Wang, D. Y. Graham, and M. K. Estes, “Detection of Norwalk virus in stool by polymerase chain reaction,” Journal of Clinical Microbiology, vol. 30, no. 10, pp. 2529–2534, 1992. View at Scopus
  4. A. Scipioni, A. Mauroy, J. Vinjé, and E. Thiry, “Animal noroviruses,” Veterinary Journal, vol. 178, no. 1, pp. 32–45, 2008. View at Publisher · View at Google Scholar · View at Scopus
  5. G. Belliot, S. V. Sosnovtsev, T. Mitra, C. Hammer, M. Garfield, and K. Y. Green, “In vitro proteolytic processing of the MD145 Norovirus ORF1 nonstructural polyprotein yields stable precursors and products similar to those detected in calicivirus-infected cells,” Journal of Virology, vol. 77, no. 20, pp. 10957–10974, 2003. View at Publisher · View at Google Scholar · View at Scopus
  6. M. E. Hardy, “Norovirus protein structure and function,” FEMS Microbiology Letters, vol. 253, no. 1, pp. 1–8, 2005. View at Publisher · View at Google Scholar · View at Scopus
  7. R. Chen, J. D. Neill, J. S. Noel et al., “Inter- and intragenus structural variations in caliciviruses and their functional implications,” Journal of Virology, vol. 78, no. 12, pp. 6469–6479, 2004. View at Publisher · View at Google Scholar · View at Scopus
  8. A. Bertolotti-Ciarlet, S. E. Crawford, A. M. Hutson, and M. K. Estes, “The 3' end of norwalk virus mRNA contains determinants that regulate the expression and stability of the viral capsid Protein VP1: a novel function for the VP2 protein,” Journal of Virology, vol. 77, no. 21, pp. 11603–11615, 2003. View at Publisher · View at Google Scholar · View at Scopus
  9. B. V. V. Prasad, M. E. Hardy, T. Dokland, J. Bella, M. G. Rossmann, and M. K. Estes, “X-ray crystallographic structure of the Norwalk virus capsid,” Science, vol. 286, no. 5438, pp. 287–290, 1999. View at Publisher · View at Google Scholar · View at Scopus
  10. E. Duizer, K. J. Schwab, F. H. Neill, R. L. Atmar, M. P. G. Koopmans, and M. K. Estes, “Laboratory efforts to cultivate noroviruses,” Journal of General Virology, vol. 85, no. 1, pp. 79–87, 2004. View at Publisher · View at Google Scholar · View at Scopus
  11. T. Ando, J. S. Noel, and R. L. Fankhauser, “Genetic classification of 'Norwalk-like viruses',” Journal of Infectious Diseases, vol. 181, no. 2, pp. S336–S348, 2000. View at Publisher · View at Google Scholar · View at Scopus
  12. M. Koopmans, C. H. Von Bonsdorff, J. Vinjé, D. De Medici, and S. Monroe, “Foodborne viruses,” FEMS Microbiology Reviews, vol. 26, no. 2, pp. 187–205, 2002. View at Publisher · View at Google Scholar · View at Scopus
  13. S. M. Karst, C. E. Wobus, M. Lay, J. Davidson, and H. W. Virgin, “STAT1-dependent innate immunity to a norwalk-like virus,” Science, vol. 299, no. 5612, pp. 1575–1578, 2003. View at Publisher · View at Google Scholar · View at Scopus
  14. P. Suravajhala and V. S. Sundararajan, “A classification scoring schema to validate protein interactors,” Bioinformation, vol. 8, no. 1, pp. 34–39, 2012. View at Publisher · View at Google Scholar
  15. E. L. L. Sonnhammer, S. R. Eddy, and R. Durbin, “Pfam: a comprehensive database of protein domain families based on seed alignments,” Proteins-Structure Function and Genetics, vol. 28, no. 3, pp. 405–420, 1997.
  16. S. Yellaboina, A. Tasneem, D. V. Zaykin, B. Raghavachari, and R. Jothi, “DOMINE: A comprehensive collection of known and predicted domain-domain interactions,” Nucleic Acids Research, vol. 39, no. 1, pp. D730–D735, 2011. View at Publisher · View at Google Scholar · View at Scopus
  17. M. Magrane and U. Consortium, “UniProt Knowledgebase: a hub of integrated protein data,” The Journal of Biological Databases and Curation, vol. 2011, 2011. View at Publisher · View at Google Scholar · View at Scopus
  18. R. Apweiler, T. K. Attwood, A. Bairoch et al., “The InterPro database, an integrated documentation resource for protein families, domains and functional sites,” Nucleic Acids Research, vol. 29, no. 1, pp. 37–40, 2001. View at Scopus
  19. A. Ceol, A. Chatr Aryamontri, L. Licata et al., “MINT, the molecular interaction database: 2009 update,” Nucleic Acids Research, vol. 38, no. 1, pp. D532–D539, 2009. View at Publisher · View at Google Scholar · View at Scopus
  20. M. E. Smoot, K. Ono, J. Ruscheinski, P. L. Wang, and T. Ideker, “Cytoscape 2.8: new features for data integration and network visualization,” Bioinformatics, vol. 27, no. 3, pp. 431–432, 2011. View at Publisher · View at Google Scholar · View at Scopus
  21. C. Chothia, J. Gough, C. Vogel, and S. A. Teichmann, “Evolution of the protein repertoire,” Science, vol. 300, no. 5626, pp. 1701–1703, 2003. View at Publisher · View at Google Scholar · View at Scopus
  22. B. Rost, J. Liu, R. Nair, K. O. Wrzeszczynski, and Y. Ofran, “Automatic prediction of protein function,” Cellular and Molecular Life Sciences, vol. 60, no. 12, pp. 2637–2650, 2003. View at Publisher · View at Google Scholar · View at Scopus