SCREENING AND IDENTIFICATION OF DRUG TARGETS AND VACCINE CANDIDATES FOR HELICOBACTER PYLORI STRAIN Hp26695

Helicobacter pylori, a class 1 carcinogen colonizes stomach causing gastric carcinoma. Rising antibiotic resistance and reinfections are drawbacks of antibiotic therapies. Alternating drugs and vaccination may be the promising approach to prevent and treat reoccurring infections. Therefore, there is a need for discovery of drug targets, drugs and vaccine candidates for the treatment of H. pylori. An objective of this current study is to identify potential drug targets and suitable vaccine candidates for H. pylori strain Hp26695 by insilico genome and proteome analysis. Drug targets were identified initially by comparing the genomes between H. pylori and Homosapien sapiens using RAST. RAST identified a total of 569 unique genes. These unique genes later were subjected to non-homology and gene property analysis to identify the potential drug targets. BLASTpfollowed by gene property analysis of 569 unique genes identified seven potential drug targets. Vaccine candidates were identified initially by screening protein sequences for pathogenic factors. These pathogenic factors were screened to identify non-homologous molecules and secondary structure patterns (helices). The proteins ≤ 3 helices are subjected to screening of antigenic nature followed by allergenicity. The proteins qualifying the above criteria were screened for antigens, Bcells and T-cell epitopes. Proteins showing positive predictions for antigenic, B-cell, T-cell activities are thus shortlisted as vaccine candidates for vaccine designing. Analysis identified 16 immunogenic proteins contributing to immune-response. These methods have enabled rapid identification of potential drug targets and vaccine candidates for strain Hp26695 with possible therapeutic implications for gastric cancer. cancer.


INTRODUCTION
Gastric cancer is caused by infection of class 1 carcinogen Helicobacter pylori (Zhang, 1994). Treatment for H. pylori infection includes drugs to relieve from pain and acidity, but not for gastritis, peptic ulcers, and gastric cancer. Carcinogenic activity of H. pylori suggests the need for discovery of new drug targets and drugs for prevention of H. pylori. Laboratory techniques and bioinformatics approaches are used to identify drug targets which can influence growth, colonization and virulence of H. pylori (Neelapuet al., 2014). Availability of the complete H. pylori genome sequence of pathogens provides us the platform and opportunity to mine the genome and harness the potential drug targets. Comparative genomic analysis between host and pathogen would provide us with a tremendous amount of information that can be useful in drug target identification (Neelapuet al., 2013).Comparative genomics analysis of host with pathogens revealed potential drug targets in Staphylococcus aureus (Uddinet al., 2014), H. pylori (Neelapu and Pavani, 2013;Neelapuet al., 2015;Nammiet al., 2016)Listeria monocytogenes (Hossainet al., 2013), Leishmaniainfantum (Sutharet al., 2009), L. major (Florezet al., 2010), Mycobacterium leprae (Wiwanitkit, 2014), Pseudomonas aeruginosa (Sakharkaret al., 2004), Schistosomamansomi (Caffreyet al., 2009).Metabolic pathway analysis (Sarkaret al., 2012), reverse docking (Caiet al., 2006) and screening for essential genes (Duttaet al., 2006) are used to identify drug targets in H. pylori. However, there are no specific reports to date, on comparing genomes of H. pylori strain Hp26695 with host Homosapiens to identify drug targets in H. pylori. Therefore, comparing genome of host and pathogen may provide novel drug targets for H. pylori strainHp26695.
Rising antibiotic resistance due to efflux pumps (Rauws and Tytgat, 1990;Graham et al., 1991), potential reinfection of H. pylori even after successful eradication therapy (Arora and Czinn, 2005), inhibition of T-cell stimulation by vacuolatingcytotoxin (Vac A) (Molinari, 1998;Reyrat, 1999), and the highly inflammatory nature by the constituents of the cell wall suggests vaccination as an alternative for protection against H. pylori. For better immunization the half-life of the antimicrobial agents should be long enough to be effective and also penetrate mucosal barrier (Arora and Czinn, 2005). Hence, the use of vaccines with appropriate immunogens may provide immune protection against H. pylori. Therefore, mining the genomic sequences via bioinformatics approaches for immunological data would provide suitable vaccine candidates. The objective of the current paper is to screen and identify novel drug targets and vaccine candidates for therapeutic intervention of H. pylori.

Sample
Complete genomeof H. pylori strain Hp26695 with the geographical origin of Europe has the following accession number and genome length NC_018939 and 1,667,867 bp respectively (Manolovet al., 2014). In our study, identification of novel drug targets and vaccine candidates for H. pylori has been accomplished for the first time in H. pylori strain Hp26695. Novel drug targets were screened and identified using an integrated approach of genome, proteome and metabolic pathway analysis followed by primary property analysis of the genes/proteins using computational resources. Novel vaccine candidates were screened and identified by searching for pathogenic factors, followed by non-homology, secondary patterns and subsequent analysis for antigens, nonallergens and epitopes using computational resources.

Screening and Identification of Drug Targets for H. pylori
The following protocol was followed for screening and identification of novel drug targets in H. pylori ( Figure 1)

Drug target screening for identification of unique molecules in H. pylori
Comparative genome analysis was performed to screen the drug targets for pathogen H. pylori.Genomeof H. pylori strain Hp26695 was initially annotated and further reconstructed for metabolic pathways using Rapid Annotation Subsystem Technology (RAST) server (Aziz et al., 2008). Comparative genome analysis between pathogen H. pylori and host Homo sapiens sapienswas performed using RAST to screen unique genes that are only present in pathogen and not present in the host (Table 1). Genes which are unique to H. pylori in the above method were filtered and catalogued.

Drug target screening for confirmation of unique molecules in H. pylori
Bacterial genes which are non-homologous to humans are essential for pathogen. To identify the non-homologous molecules in H. pylori, homology at the level of sequence and structure of molecules were used as the parameters. BLASTp (Altschulet al., 1990) which is based on principle of homology was used to confirm the uniqueness of the catalogued genes in H. pylori by comparing genes against Homosapiens sapiens (Table 1).

Drug target identification
A set of computational resources were used to analyse the characteristic features of the genes, to identify the potential drug targets.BTXpred , SRTpred (Garg and Raghava, 2008), VGIchan  and VICMpred  are the potential targets servers (Table 1) to identify the potentiality of the drug targets. Catalogued genes were verified for their potentiality as drug targets using the above list of servers.

Screening and Identification of Vaccine Candidates for H. pylori
The following protocol was followed for screening and identification of vaccine candidates in H. pylori (Figure 2).

Screening of proteome for identification of pathogenic factors in H. pylori
The bacterial genome was retrieved and the translated protein sequences of the pathogen are screened for pathogenic factors. Virulence factors, secretory proteins, outer-membrane proteins, bacterial toxins are the pathogenic factors. VirulentPred, EffectiveDB, CELLO, BTXpred (Table 1) are used to screen virulence factors, secretory proteins, outer-membrane proteins and bacterial toxins respectively. Further, the pathogenic factors of the bacteria are screened for non-homologous proteins as per the procedure described above in "Drug target screening for confirmation of unique molecules in H. pylori".

Screening of non-homologous proteins for identification of secondary patterns in H. pylori
The non-homologous proteins of the bacteria are screened for secondary patterns -helices using Chou Fasman method by CFSSP: Chou and Fasman Secondary Structure Prediction Server (Ashok Kumar, 2013) (Table 1). Proteins with alphahelices ≤ 3 are selected for further analysis.

Screening of proteome for antigens
The proteins fulfilling the above criteria are screened for antigens using Antigenic Emboss Server (Kolaskar and Tongaonkar, 1990) (Table 1). These proteins are catalogued and subjected to further analysis.

Screening of proteome for non-allergenicity
The proteins which are antigenic in nature are screened for allergenicity using server Allergen Online (Maria et al., 2006) ( Table 1). The non-allergens are shortlisted and catalogued for further analysis.

Screening of non-allergenic proteome for identification of antigenic and epitope regions in H. pylori
The proteins fulfilling the above criteria are screened for promising epitopes which include both B cell & T cell epitopes. B-cells epitopes are screened and identified using ABCpred Server   (Table 1), whereas T-cell epitopes are screened and identified using HLApred (Table 1). Finally, proteins satisfying the above three criteria's i.e. proteins showing positive predictions for antigenic, B-cell, Tcell activities are short listed as vaccine candidates for vaccine designing.

Genome Wide InsilicoAnalysis for Screening of Drug Targets in H. pylori
Genome wide in silico analysis for screening drug targets identified 569 unique genes in H. pylori strain Hp26695 (Table  2). These molecules fall under 24 metabolic categories as shown in Table 2. Proteome analysis followed by gene property analysis of 569 unique genes identified seven potential drug targets in H. pylori (Table 3). These molecules fall under five metabolic categories as shown in Table 3.  Aziz et al., (2008) 2 BTXpred Server is for prediction of bacterial toxins and its function from primary amino acid sequence.  3 SRTpred Server classifies protein sequence as secretory or non-secretory proteins. Garg and Raghava, (2008) 4 VGIchan Server predicts voltage gated ion-channels and classifies them into sodium, potassium, calcium and chloride ion channels from primary amino acid sequences. Sahaet al., (2007) 5 VICM pred Server aids in broad functional classification of bacterial proteins into virulence factors, information molecule, cellular process and metabolism molecule.  6 VirulentPred VirulentPred predicts virulence proteins using reliable Support Vector Machine (SVM) algorithm. This server has a prediction accuracy of 65%. Garget al., (2008) 7 EffectiveDB EffectiveDB predicts putative effectors by identifying eukaryotic-like protein domains and by detecting the 2 known types of signal peptides. This server has a prediction accuracy of 80%.
Jehlet al., (2011) 8 CELLO-V CELLO (subcellular localization predictor) predicts protein present in outer membrane directly from protein sequences. The server uses two-level support vector machine (SVM) system: the first level contains SVM classifiers and the second level SVM classifier function to generate the probability distribution of decisions for possible localizations.
This server has a prediction accuracy of 90%. Yu et al., (2004) 9 CFSSP Server Chou &Fasman Secondary Structure Prediction (CFSSP) server predicts protein conformation like helices, beta sheets, random coils based on Chou &Fasman algorithm.
This server has a prediction accuracy of 88%.
Ashok Kumar, (2013) 10 Antigenic Antigenic server predicts potentially antigenic sites of a protein sequence. The server uses semi-empirical method consisting physicochemical properties of amino acids and their frequencies of occurrence in experimentally known epitopes. This server has a prediction accuracy of 75%. Kolaskar and Tongaonkar, (1990) 11 ABCpred ABCpred server predicts B-cell epitope using Recurrent Artificial Neural Network-(ANN-) based algorithm. This server has a prediction accuracy of 65.93%.  12 HLApred HLApred server identifies the experimentally proven binders taken from MHCBN database based on quantitative matrices HLA alleles which were obtained from literature. This server has a prediction accuracy of 80%.
http://www.imtech.res.in/raghava/hlapred/index.html 13 Allergen Online Cross reactive allergens are predicted using server Allergen Online based on BLOSUM50 scoring matrix algorithm. This Server has a prediction accuracy of 70%. Maria et al., (2006) Figure1 Screening and identification of novel drug targets for H. pylori
Insilicoscreening of peptides have helped examining the molecular properties, further in vivo-studies would be most helpful in bringing out potentially specific vaccine candidates.  (Arakawa et al., 2011). Therefore, designing an inhibitor for menaquinone via futalosine step 1 would affect the growth of H. pylori.

Drug targets influencing DNA metabolism of the pathogen
Type III restriction-modification system methylation subunit of restriction-modification (R-M) systems, is identified as the drug target in H. pylori. This drug target influences metabolic pathway DNA metabolism of the pathogen. H. pylori are naturally competent and prone to take DNA from the environment (Dorer et al., 2010) and bacteriophages also infect H. pylori (Heintschel et al., 1993). Missense and frameshift mutations can accumulate and inactivate genes when bacteriophages or free DNA or plasmids enter into other cells.
Evidence is there that sometimes even both endonuclease and methylase genes of R-M systems have to be turned off. However, H. pylori in a population have a very good defensive system, where R-M systems protect the genome of H. pylori from accumulated mutations when bacteriophages or free DNA or plasmids enter into other cells. Mutant strains lacking this display a pleiotropic phenotype, including increased mutability, hyper recombination, and increased sensitivity to DNAdamaging agents. Therefore, designing an inhibitor for type III restriction-modification system methylation subunit decreases the rate of survival of H. pylori due to gross changes occurring in the genetic material.

Drug targets influencing DNA metabolism of the pathogen
Dipeptide transport system permease protein DppB, dipeptide transport system permease protein DppC and ferric siderophore transport system, biopolymer transport protein ExbB are identified as drug targets in H. pylori. These drug targets influence membrane transport of the pathogen.
DipeptideDppABCDF and oligopeptideoppABCD genes are a class of ABC-type transporter in H. pylori. Dipeptide transport system permease protein -DppBC are responsible for transporting dipeptides. Dipeptide and oligopeptide system pylori lacked the ability to use certain dipeptides, hexapeptides, and nonapeptides due to compromisation of either substrate binding domain or permease domains (Weinberg, 2007). Therefore, designing an inhibitor to dipeptide transport system permease protein DppBC would affect the growth and survival of H. pylori.
Ferric transport system, biopolymer transport protein ExbB is a member of transporter proteins in H. pylori. All bacterial pathogens have developed highly sophisticated iron assimilation systems as a response to iron-limiting conditions encountered in environment and host's body fluids. Production of siderophores, small nonproteinaceous molecules with extremely high affinity for iron (III), is one of the most successful and widely utilized strategies of iron assimilation (Merrell et al., 2003). Common components of both siderophore-dependent and host iron-binding protein-dependent iron acquisition systems are receptor proteins involved in binding of siderophores and interacting with the host ironbinding proteins. These large outer membrane proteins are responsible for the transport of iron or iron-containing LGLILSLAAILIAFK responsible for the transport of iron or iron-containing compounds throughthe otherwise impermeable outer membrane (Ye et al., 2003). Ferric transport system Exb B biopolymer transport protein in H. pylori is responsible for the transport of iron or iron-containing compounds through the impermeable outer membrane. Sequence analysis in E. coli, Haemophilusinfluenzae, Neisseria meningitides and Pseudomonas putida provided information on existence mechanism that utilizes Ton-independent heme. Knockout mutant and complementation studies in Neisseria meningitides established this fact (Sarangi et al., 2009). Designing an effective inhibitor to the existing multiple proteins for the utilization of heme-containing compounds effects the survival of H. pylori in their natural habitat, human mucosal surfaces.utilization of heme-containing compounds effects the survival of H. pylori in their natural habitat, human mucosal surfaces.

Drug targets influencing RNA metabolism of the pathogen
Ribonuclease BN is identified as a drug target in H. pylori. This drug target influences RNA metabolism of the pathogen. Ribonuclease, BN, lacking RNase H and RNase D activity was identified in E. coli and it is different from other exoribonucleases known till date in E. coli. RNase BN is a substrate specific with specificity towards C-C-A sequence in tRNA than other types of tRNA and substrate specificity was proved both in vitro and in vivo. Mutants of these proteins affect the processing of tRNA's and ultimately synthesis of protein (Asha et al., 1983). Hence, an effective inhibitor for Ribonuclease BN can block the function of protein synthesis.  IVFCCFLRA  EPC467  LVIVFCCFL  EPC468  VIVFCCFLR  AG75  TQEFLYMKSSFVEFF  EPC470  YMKSSFVEF  AG76  KFYAYGISDV  EPC472  FYAYGISDV   AG81  LNSLSVTKVECSKGKHHAYVFVLSSDHKILSKL   EPC500  FVLSSDHKI  EPC501  YVFVLSSDH  EPC502  LSSDHKILS  EPC503  VLSSDHKIL  AG83  WFKCPKLSFVSDN  EPC505  WFKCPKLSF   AG86  IAFYFFAILTLSMALVVITTTNILYAITALASSMVFISAFFFLLDAEFLGVVQITVYVGAV  IVMYA   EPC528  FYFFAILTL  EPC529  ITVYVGAVI  EPC530  VVQITVYVG  EPC531  VYVGAVIVM  EPC532  LVVITTTNI  EPC533  VVITTTNIL  AG88  PKILCILSFGVALLLTLILSAPS  EPC540  LTLILSAPS  AG89  DAQIPNIKAIGYVLFTNYLIPFEAAALMLLVAMVGGI  EPC542  YVLFTNYLI  AG92 HIKVISI EPC572 HIKVISI

Drug targets influencing respiration of the pathogen
NADH-ubiquinone oxidoreductase chain J is identified as a drug target in H. pylori. This drug target influences metabolic pathway and effect respiration of the pathogen. The NADH ubiquinone oxidoreductase (Complex I), provides the input to the respiratory chain from the NAD-linked dehydrogenases of the citric acid cycle. The complex couples the oxidation of NADH and the reduction of ubiquinone, to the generation of a proton gradient which is then used for ATP synthesis. The complex occurs in the mitochondria of eukaryotes and in the plasma membranes of purple photosynthetic bacteria, and the closely related respiratory bacteria. All inhibitors affect the electron-transfer step from the high-potential iron-sulphur cluster to ubiquinone. Class I inhibitors appear to act directly at the ubiquinone-catalytic site which is related in complex I and glucose dehydrogenase (Friedrich.et al., 1994). Inhibitors designed to bind to NADH-ubiquinone oxidoreductase chain J competitively inhibit the protein from functioning which results in chemical asphyxiation of cells.

Vaccine Candidates for H. pylori
Constructive screening protocol was implemented to identify suitable vaccine candidates for H. pylori. Choosing such a conservative way to face vaccine design inevitably implies missing some pathogen antigens, but still this is a small price to reach a valuable compromise. Forwarding further bioinformatics analyses on selected ones, may prove successful (Sandroet al., 2006). Bioinformatics approach has helped us in shortlisting and in identifying pathogenic factors from the proteome of the pathogen. Screening of pathogen factors for non-homology would shortlist the proteins which have the potential to cross react when vaccine is administered. Usually a protein has high probability of failure to cloning and express in experiment when it is likely to have more helices. Hence, proteins with alpha-helices < 3 are selected for further analysis. (Sandroet al., 2006;Capecchiet al., 2004;Pizza et al., 2000).Screening of proteome for allergenicity would avoid the proteins which can elicit undesirable reaction during vaccination. Further, proteins showing positive predictions for antigenic, B-cell, T-cell activities are characterized as potential immunogens which are suitable for vaccine candidates.

CONCLUSION
Comparative genomics of H. pylori and Homosapien sapiens identified seven bacterial genes which are non-homologous to humans and are essential for pathogen. Four genes of the 7 predicted drug targets are already experimentally validated lending credence to our approach. These novel drug targets may have possible therapeutic implications for gastric cancer. Systematic insilico analysis approach identified 16 immunogenic proteins which are suitable vaccine candidates for H. pylori. Thus, bioinformatics approaches helped in rapid identification of novel drug targets and vaccine candidates for H. pylori strain Hp26695.