In silico Analysis of CadF epitope- based vaccine design against Campylobacter jejuni

To eradicate infectious diseases caused by microorganisms, vaccination is a popular strategy against them and can be an effective approach. Among vaccines, subunit vaccines have been used to be more effective against diseases. CadF protein of Campylobacter jejuni is one of the important antigens in the pathogenic process of the bacterium. So, the aim of this work was to do a bioinformatics study for the identification of epitope-based CadF vaccine, as a subunit vaccine. CadF gene sequences were extracted from the NCBI database and suitable physic-chemical properties of CadF were evaluated by Protprom server. Some epitopes of the CadF protein with high affinity were detected by different servers, which were predicted based on MHC-peptide complex and B-cell epitopes. The results indicated that CadF is an antigenic and non-allergenic protein and provided a desirable structure for design vaccine. Among epitopes, LSDSLALRL was confirmed for the simulation of both of B and T cells. This 9-mers peptide was located in 135-143 sequences of CadF protein and interacted with HLA-A0101. The peptide isn’t allergen and has the ability to be an antigen for motivating immune system. Besides, the analysis implied that the epitope structure could allow designing a vaccine against C. jejuni.


Introduction
Campylobacter jejuni ( C. jejuni) is one of the significant pathogens belongs to the genus Campylobacter, which is a gram-negative, spiral, curved, and rod-shaped bacterium [1,2].
They can be transmitted to humans through direct contact with animals, consumption of contaminated food and water, unpasteurized milk or contact with patients [1,3]. Some gastrointestinal diseases are caused by this pathogen, which it has considered as one common bacterial cause of diarrhea, especially in children [1,[4][5][6]. According to researches, more than 229,000 cases of the campylobacteriosis have been reported worldwide [7]. Campylobacter jejuni can also be associated with Guillain Barre Syndrome (GBS), as a neurological disease in humans [5,8,9]. The disease is an autoimmune disorder which can be caused by cross-reactive immune response between LPS (Lipopolysaccharides) core of C. jejuni and gangliosides in the nerve cells of the human [10,11]. This bacterium has been also observed in pregnant women and can cause spontaneous abortion, stillbirth, prematurity, and neonatal sepsis in many cases.
Although, C. jejuni is self-limited but it is necessary to identify and confront with C. jejuni [12].
Virulence factors such as motility, adhesion, invasion, and toxin production contribute to the pathogenesis of C. jejuni [ 13]. Several adhesion factors e.g. CadF, CapA, JlpA, etc. are presented by C. jejuni to facilitate host-pathogen interactions [14]. Among these, CadF is one of the important proteins that bind to fibronectin (fibronectin-binding protein) and adheres the bacteria to host cell, resulting facilitate colonization. This is a 37kDa outer membrane protein has been described as a conserved and genus-specific protein. The reports have been showed that CadF can induce massive immune responses, including both humoral-and cell-immunities [7,15].
Today, vaccination has been a very successful strategy for the eradication and prevention of the infections. Different types of vaccines are currently available, including live attenuated, inactivated or killed, toxoid, and subunit vaccines [16,17]. Subunit vaccines include only parts of the microorganisms, which are known as safe and effective vaccines for human and animals. Investigations have also been shown that they are capable to induce both humoral-and cell-mediated immunities against the antigens of the microorganisms. To develop an effective subunit vaccine, the identification and prediction of the antigenic epitopes by bioinformatics tools are useful [18]. Bioinformatics methods provide new theoretical approaches for the design of the vaccines based on immunological databases such as the form of epitopes, MHC alleles, molecular interactions, and docking pathogens and host cells [19][20][21][22].
Although there are some studies on the evolution of the outer membrane proteins of C.
jejuni as vaccine candidates, CadF can be independently considered for the design of a protective vaccine [8].
Our aim was to analysis CadF protein for the identification of the epitope-based peptide candidates and to evaluate its proteomic database by In silico tools for developing a new vaccine candidate.

Protein analysis and identification of conserved regions
The sequences of the CadF protein were acquired from NCBI Protein Data Bank

Allergenicity and antigenicity assessment
The Allertop (www.ddg-pharmfac.net/AllerTOP) and AllergenFP (ddgpharmfac.net/AllergenFP) web servers were used to determine the allergenicity of the protein. The Allertop server was planned based on amino acids properties such as hydrophobicity, size, and helix forming which could classify a number of allergen and nonallergen targets. The AllergenFP data based to obtain a set of options for predicting allergens. After, protective antigens of the CadF protein were forecast by Vaxijen server.

Design and evaluation of molecular docking
To recognize the three-dimensional structures and biological functions, Phyre2 (www.sbg.bio.ic.ac.uk/phyre2), as an online protein fold recognition server, was used.
Secondary structure of the protein was also analyzed by Psipred

Result
CadF sequences were extracted from the NCBI database and FASTA format was used for analysis. The complete sequences of CadF protein were 319 amino acids and multiple sequence alignment showed that this protein is a highly conserved protein among Campylobacters and is belonging to the superfamily of outer member proteins (ompA) and in the 193-287 position of CadF protein an ompA-like domain can be detected. The result of the phylogenic tree also confirmed that CadF is classified in outer membrane protein superfamily (data not shown).

Antigenicity and Allergenicity protein analysis
The score of the antigenic prediction was calculated 0.79 by Vaxijen server. The results showed that this protein is probably antigen and can be used for the next analysis. The obtained data from the AllergenFP server was indicated a similarity 0.82 for the protein; hence it cannot be an allergen. Allertop server analysis also confirmed that CadF protein isn't an allergen.

The physico-chemical characterizes
Using ProtParam server, the MV (molecular weight) and PI (isoelectric point) parameters

Prediction of secondary and tertiary structures
To predict secondary and tertiary structures, we used different online servers. By PSIPRED server, graphical results of the secondary protein structures were obtained that indicated sheet, helix, and extracellular transmembrane structures (Additional file 1). In addition, The Phyre online servers were reported a three-dimensional structure of the modeled CadF with 97% confidence score and 192 known-domains aligns. The structural contents are included 16% alpha helix, 41% beta strands, and 16% disordered regions. Also, the prediction of the CadF protein shows a binding site at GIU-HIS-LYS residue and a lot of metallic heterogenic sections in its structure ( Fig. 1). Moreover, the three-dimensional (3D) structure belongs to selected epitopes was drawn by PyMOL software (Fig. 2).

Forecasted of antigenic T cell epitopes
To predict T cell epitopes, the best score of the epitopes were selected from SYFPEITH, IEDB, NetCTL, NHLAPred, NETMHC I, and MHCPred II online servers. Except by IEDB, which showed a high value for the lowest number, other servers have been specified a high value for the highest number. The epitopes of the MHC I (A-0101, A-0201, and B-2705) and the MHC II (DR1-0101 and DRB1-0401) were the most common epitopes in Iranian alleles that have been considered in this study. According to achieved data from abovementioned servers, the predicted epitopes of MHC I and MHC II are presented in Tables 1 and 2, respectively.
Among selected epitopes, LLCLGLASV, RRVDAKFIL, FSADNNVKF, and LSDSLALRL (belong to MHC class I) are shown in Table 1 and EGHFGFDKTTINPTF, QINFNHANH, LSDSLALRL, ASVLFSADNNVKFEI, and QINFNHANHNWVSTL (belong to MHC II) are showed in Table 2, as well. Due to achieving a high score in many servers and being antigen and loss of allergenicity, we estimated that they can act as a proper epitope.

B-cell epitopes prediction
After extracting the B-cell epitopes by IEDB, Bcepred, ABCpred, and SVMTrip servers, common and repetitive epitopes were selected. The results are showed in Table 3.
According to data, WVSTLGISFG, LETRDQINFN, VGEKFYFYGL, and NPRSSNDTKEGRADNRRVDA peptides were characterized that could be analyzed as predicted B-cell epitopes. Also, the graph was modeled by Bepipred server and in Fig. 3 shows that the yellow parts can be B-cell epitopes with a suitable threshold (0.5). Y-axe shows scores related to the amino acids and X-axe defines positions related to the protein regions.

Overall result of above-mentioned epitopes
The retrieved results of tables have identified the favored residues from T-and B-cells called LSDSLALRL, because it is common epitope with antigenicity and allergenicity properties and we suggest it as a candidate vaccine for next analysis.

Discussion
Campylobacter jejuni is one of the main reasons for gastroenteritis diseases throughout the world, specifically in developing countries [3,6]. It can be associated with diarrhea (from mild to severe), fever, neurological disorders, reactive arthritis, and weight loss. It is estimated about 14 cases for 100,000 people that were afflicted worldwide every year [4,23]. In order to handle the disease, the sciences are in urgent to raising their knowledge about C. jejuni. The pathogenicity of C. jejuni is up to the ability to linkage with other organisms and many proteins take part in the adhesion and colonization of and it had advanced in animal models but has failed at the human cases. Since the results were not able to gotten approved in term of safety for humans [5]. Also, different reports have demonstrated polysaccharide capsule (cps) of the bacteria that could be used for alleviating campylobacteriosis, but this approach was just tested at phase 1 clinical; however there are concerns about being the similarity between bacterial polysaccharides and human gangliosides [10].
Other antigens, e.g. ABC transporter (PEB1), which is known as an immunogenic and protective protein, could be a candidate for vaccination against C. jejuni. Due to the response of the immune system produced by lymphocytes inverse PEB1 protein, it could inhibit the development of the disease. However, this vaccine has been building in preclinical trials yet [5,27]. Beyond a PEB1-vaccine mentioned approach, other pathogenic proteins of C. jejuni have also been analyzed. Kobierecka et al. have found that CjaA, CjaD (binding proteins), FlaA (flagellin), and CmeC (outer-membrane protein) could protect the chickens against C. jejuni in many cases [9]. Neal-McKinney et al. also have asserted the vaccination of chickens with the CadF-FIaA-FIpA fusion protein that could reduce the rate of infection, but there is not any evidence whatever will be able to use in human models [8]. Moreover, some of the peptides related to invasive, virulent, and membrane of the bacteria, which contain FlaA, Cia, CadF, PEB1, PEB3, and MOMP have assessed which could be effective in design a vaccine against C.jejuni by immunoinformatics tools. So, despite many efforts to make the vaccine, there are not any approved the vaccine against C.
Our findings showed that CadF is a highly conserved protein among Campylobacter spp.
and is belonging to the superfamily of outer member proteins (ompA). In addition, our research also showed the correct topology model based on phyre2 server that predicts CadF is a stable target. This analysis was done with bioinformatics methods and helps to design novel vaccine according to sequence profile and spatial structure and dimension of protein.
In present study used from SYFPEITH, IEDB, NetCTL, NHLAPred, NETMHC I, MHCPred II, Bcepred, ABCpred, and SVMTrip online servers for searching of our study. Finally, common epitopes were identified and the LSDSLALRL epitope was selected as the best potential vaccine candidate. Moreover, the allergenicity of the LSDSLALRL epitope has also confirmed by Allertop server and confirmed that this sequence wasn't an allergen as well.
The epitope was located in 135-143 regions and can be interacted with HLA-A0101 according to collected results from many above-mentioned servers. In a partial contrary allergen, while no allergenicity was observed for our epitope "LSDSLALRL" in this study.
In addition, CadF is a significant protein for the colonization and binding the bacteria to host cells and maximum connection can be detected in the regions of the fibronectinbinding domain including phenylalanine-arginine-leucine-serine (FRLS) residues of CadF.
Although only fifty percent from the amino acids of our epitope was identified as the binding site to host cells, multiple servers confirmed that this reign has a high score for developing the vaccine.
According to aliphatic index, alanine, valine, isoleucine, and leucine amino acids were

Authors' Contributions
MMN, SS, MMN, and BB involved in the management of the project, the analysis of data, and writing up the paper. All authors read and approved the final manuscript.
corresponding author on reasonable request.

Consent for publication
Not applicable.

Ethics approval and consent to participate
Not applicable.

Funding
Not applicable. Tables Table 1. The list of high scored predicted T cell epitopes using online software and their Vaxijen and Allertop score. The purple highlighted epitopes were repeated in some servers but aren't selected because of being problems in their antigenicity and allergenicity. The blue-colored epitopes are suitable with regard to their traits and the green epitope was considering the most common peptide with correct abilities.   Table 3. The list of high scored predicted B cell epitopes using online software and their Vaxijen and Allertop scores. The purple highlighted epitopes were repeated in some servers but aren't selected because of being problems in their antigenicity and allergenicity. The blue-colored epitopes are suitable with regard to their traits and the green epitope was considering the most common peptide with correct abilities.

Additional File
Additional file 1: Figure 1. Secondary structure of CadF protein is displayed by psipred server. Figure 1 Three-dimension structure of CadF protein was predicted. The image colored by rainbow from N to C terminus. The Alpha helix (yellow and orange) and beta strand (green and blue arrows) and linker (thin strands).