hpvPDB: An Online Proteome Reserve for Human Papillomavirus.

Human papillomavirus (HPV) infection is the leading cause of cancer mortality among women worldwide. The molecular understanding of HPV proteins has significant connotation for understanding their intrusion in the host and designing novel protein vaccines and anti-viral agents, etc. Genomic, proteomic, structural, and disease-related information on HPV is available on the web; yet, with trivial annotations and more so, it is not well customized for data analysis, host-pathogen interaction, strain-disease association, drug designing, and sequence analysis, etc. We attempted to design an online reserve with comprehensive information on HPV for the end users desiring the same. The Human Papillomavirus Proteome Database (hpvPDB) domiciles proteomic and genomic information on 150 HPV strains sequenced to date. Simultaneous easy expandability and retrieval of the strain-specific data, with a provision for sequence analysis and exploration potential of predicted structures, and easy access for curation and annotation through a range of search options at one platform are a few of its important features. Affluent information in this reserve could be of help for researchers involved in structural virology, cancer research, drug discovery, and vaccine design.

Introduction

Human papillomavirus (HPV), a virus from the papillomavirus family, is capable of infecting humans. About 200 different strains of HPV identified, based on DNA homology, have been found to be etiologically linked to cervical, vaginal, vulvar, penile, anal, oral, and plantar infectious lesions and cancers, as well [1, 2]. The HPV genome, a double-stranded DNA molecule, consists of 8 kilobase pairs (kbp) of nucleotides, which comprises 3 regions: 6 early open reading frames (ORFs)-E1, E2, E4, E5, E6, and E7; 2 late ORFs-L1 and L2; and an upstream regulatory region [3]. A considerable volume of HPV specific information pertaining to its genome, proteome, structure, and disease association is available scattered on the web with trivial annotations; however, it is not customized to explore for data analysis, host-pathogen interaction, strain-disease association, drug designing, and sequence analysis, etc. Therefore, we proposed to develop a comprehensive reserve on HPV with maximum possible inputs and outputs for the end users.

Methods and Results

Data retrieval and curation

Amongst the existing 200 strains of HPV, 150 have been sequenced as of now, and their data available at the National Center for Biotechnology Information (NCBI). Genome and proteome information of those viral strains was retrieved from NCBI. Besides PubMed, various other online resources and published literature were also explored for validating genomic, proteomic, as well as strain and disease-associated information on HPV strains. HPV strain-specific information, such as strain name, sequencing status, sequencing centre, NCBI accession ID, associated disease information with references, genome statistics (GC%, AT%, A, T, G, C count, genes, and proteins), etc., were curated from various online resources, and protein parameters (length, molecular weight, isoelectric point) were calculated using ExPASy ProtParam [4].

Protein structure prediction and validation

MODELLER9v10 [5] and the SWISS-MODEL [6] server were used for protein structure prediction. The stereochemistry of each protein was evaluated through PROCHECK [7] analysis, available at the RCSB validation server (http://deposit.rcsb.org/validate/), and validated using ProSA-web [8] (http://prosa.services.came.sbg.ac.at/prosa.php).

Reserve architecture and design

Human Papillomavirus Proteome Database (hpvPDB), the relational reserve, was developed using Microsoft SQL Server 2005 as the back end. The website is powered by XAMPP (Windows Version 1.7.3). HTML, JavaScript, and CGI-PERL-based web interfaces were employed to execute SQL queries. The curated data and related information were stored in tables. The application layer, the web interface, and the backend relational tables were integrated using CGI-PERL. The overall architecture of hpvPDB is shown in Fig. 1.

Fig. 1

System architecture of Human Papillomavirus Proteome Database (hpvPDB) showing data collection, analysis, union, and user interface. HPV, human papillomavirus; CDS, coding sequence.

Reserve features

hpvPDB interfaces are made to help the users for easy navigation and information retrieval. Home, About, Tools, Search, and Advanced Search interfaces can be explored to obtain strain- and protein-specific information. User can access the meta information about different strains using a search box. Reserve comprises the strain-specific detailed informationon its name, sequencing status, submission details, date of submission, NCBI IDs, disease types and subtypes, type of DNA, genome length, molecular weight, nucleotide composition (A, T, G, C, AT, GC content), number of genes and proteins, and protein list. A genome map of each strain obtained by Geneious 5.4.4 software (available from http://www.geneious.com/) is also integrated in this page. Users, through an advanced search option, can precisely access the Genome and Proteome information separately by selecting HPV genome or HPV proteome. Each protein entry comprises protein overview (name, locus, function, etc.), protein sequence information (amino acid sequences with NCBI accession number with provision for direct protein BLAST [9] against NCBI nr database), protein parameters (length, molecular weight, theoretical isoelectric point [pI], amino acid composition, etc.), protein structure (predicted 3D structure by homology modeling viewed by Jmol (available from http://www.jmol.org/) [10] with the JAVA platform, Ramachandran plot obtained by PROCHECK and Z-score and Energy plot obtained by ProSA-Web. hpvPDB platform also provides a phylogeny analysis tool to perform multiple sequence alignment and phylogenetic tree construction of selected HPV proteins using the Phylogeny. fr web service [11].

The original Human Papillomaviruses Database was developed and hosted by the Los Alamos National Laboratory (LANL) between 1994 and 1999 with funding from the National Institute of Allergy and Infectious Diseases (NIAID) [12]. 'Human Papillomaviruses: A Compilation and Analysis of Nucleic Acid and Amino Acid Sequences' contains four annual data books of papillomavirus information published in both paper and electronic form (1994, 1995, 1996, and 1997) but has not been updated since 1997 [12]. This contains nucleotide sequences of few HPV strains and other papillomaviruses, amino acid and nucleotide sequence alignments, analysis, related host sequences, and database communication. We did not find any structural information in that database. In hpvPDB, along with updated protein sequence information, genome and protein structure information is also provided.

Conclusion

hpvPDB brings together comprehensive information on a total of 1,036 protein sequences and 743 predicted structures. The outcome of this study might provide a platform for simultaneous structural comparative analysis of these proteins and help in finding out variations in their structures to explore why different strains of HPV have causative associations with different types of cancers. Further, this might also help in designing specific drugs or vaccines against specific strains of HPV. This reserve provides a resource to help virologists identify potential roles for viral protein. Currently the hpvPDB is updated manually through online resources and available scientific publication review; however, to sustain the quality, these data are analyzed and checked before incorporation into this reserve. Meanwhile, to provide regular updates, our team is committed to searching for newly sequenced HPV strains, updating this reserve, and serving the users.