Hantaviruses (Bunyaviridas, Hantavirus) are rodent-borne emerging viruses [1] which cause over 100,000-200,000 cases per annually worldwide [2-5]. They are broadly classified into the New World Hantaviruses, which includes those causing hantavirus cardiopulmonary syndrome HCPS, and the Old World Hantaviruses, which are associated with another disease, hemorrhagic fever with renal syndrome (HFRS) [2-5]. This classification is determined by the geographic distribution of the rodent reservoirs [4].

HCPS is a severe respiratory disease, rare but often fatal, presents as a wide clinical spectrum ranging from brief febrile prodrome with headache, myalgia and thrombocytopenia, to rapidly progressive pulmonary edema characterized by increased vascular permeability due to dysregulation of the endothelial barrier function. The vascular leakage leads to non-cardiogenic pulmonary edema and in serious cases followed by respiratory failure, multi-organ failure, cardiogenic shock and death [5-10]. A different of pathogenic mechanisms have been suggested, including immune cell-mediated injury, cytokinemediated injury and enhanced VEGF responses from intercellular junctions resulting from highly specific virus–integrin interactions [8]. Since most deaths are caused by myocardial dysfunction and hypoperfusion rather than hypoxia, some investigators prefer using the term hantavirus cardiopulmonary syndrome rather than the previous term hantavirus pulmonary syndrome [3,10]. All of the etiologic agents associated with cardiopulmonary syndrome, at least 20 hantavirus genotypes, are found in the Western Hemisphere with the primary important etiologic agents include Sin Nombre Hantavirus (SNV) in the United States and Canada [11,12].

SNV is a Category A pathogen [13] that causes approximately 692 cases of HCPS according to the Centers for Disease Control and Prevention, from 1993-2016, [14] with a reported mortality rate ranging from 30% to 50%. [2,3,6]. SNV is an enveloped virus with a trisegmented, negative-sense RNA genome. The large (L) segment encodes an RNAdependent RNA polymerase; the medium (M) segment encodes a glycoprotein precursor, GPC, which is co-translationally cleaved into Gn and Gc transmembrane glycoproteins; the small (S) segment encodes a nucleocapsid protein (Np). Both glycoprotein and nucleocapsid have been shown to contribute to protective immunity [15-19]. SNV was first discovered within the 1993s outbreak, [2,3,15-17]. SNV-associated HPS is a severe and unpredictable public health threat, as affirmed by the 2012 outbreak at Yosemite National Park, California that resulted in 3 fatalities and more than 260,000 park visitors potentially exposed to a lethal virus [20]. SNV infections have been identified among persons with occupational exposure to North American deer mice (Peromyscus maniculatus) [21], which is the natural reservoir of SNV [10,16].

Similar to other viruses that cause HCPS, SNV has a serious potential applicability as a biological weapon because of its high lethality and known aerosol route of transmission [5]. Moreover, it causes a considerable economic harm and stigma to affected communities. Until now, Specific therapy is not available for the treatment of HCPS caused by SNV [2,3,5]. The main treatment of severe HCPS cases is completely supportive, often in intensive care unit surroundings [2]. Ribavirin, a broad-spectrum nucleoside analogue antiviral drug, is of some benefit as therapy in HCPS but it does not seem to have any clinical application in HCPS patients due to the lack of conclusive clinical data [3,4]. However, many efforts have been investigated in developing safe and effective vaccine against HCPS, included conventional approaches as well as molecular vaccine approaches using recombinant subunit vaccines utilizing vaccine delivery systems and plasmid DNA delivered by gene gun [2,3,12,13]. These approaches do not seem to have made much progress in this area, to date there is no vaccine proven to be highly efficacious against HCPS especially SNV [1-3,20]. Thus, epitopebased vaccines (EVs) have a priority for selection than other traditional vaccines. They offer many of advantages such as: biosafety, selecting of conserved or immunodominant epitopes for activating cellular or humoral responses and bio-processing of these epitopes are introduced economically and rapidly. In addition, the geographic distributions of the viruses are often well defined and the ethnic populations in need of vaccination can be determined [22] Indeed, many studies showed the immunological success of peptide-based vaccines against infectious diseases in experimental animals as well as in clinical trials, which demonstrated the responses to peptide vaccines against infectious diseases, for example malaria [23-27] hepatitis B, and HIV infections, but unfortunately, epitope mapping using experimental methods are costly and laborious.

More recently, immunoinformatics-based approaches in predicting most conserved and immunogenic epitopes from the viral genome sequence databases could give a significant amount of information for EVs developments. Several of recent publications explain in an excellently detail the values and benefits gained by the use of immunoinformatics and predictions in applied immunology and vaccinology [28]. Moreover, There are many studies used immunoinformatics-based approaches for designing EVs for different infectious and serious diseases [29,30]. In general, any active vaccine candidate must have to contain, at least, two antigenic epitopes; one to induce specific B cell or CTL responses while other induce specific Th cell response [19]. In this study, we aim to analyze envelope glycoprotein and nucleocapsid of SNV by using immunoinformatics tools housed in IEDB resources; in order to determine the most conserved and immunogenic epitopes for B cell and T cell which could trigger humoral as well as cell mediated immune response.

Retrieval of protein sequences

Sin Nombre virus’s glycoprotein and nucleocapsid sequences were obtained from National Centre for Biotechnology and Information (http://www.ncbi.nlm.nih.gov/protein/) databases in Nov 2016. Accession numbers of glycoprotein and nucleocapsid of retrieved strains and their date and area of collections are listed in Tables 1 and 2, respectively.

Table 1: Accession numbers of glycoprotein retrieved strains and their date and area of collection.

Table 2: Accession numbers of nucleocapsid retrieved strains and their date and area of collection

Determinations of conserve regions

The obtained sequences were aligned using multiple sequence aliment (MSA) with the aid of Clustal was implemented in the BioEdit program, version 7.0.9.0; in order to find the conserved regions among glycoprotein and neuclocapsid variants of the strains [31].

B-cell epitopes prediction

B cell epitope is part of an immunogen which interacts with B-lymphocytes. This interaction leads to differentiation of B-lymphocyte into an antibody-secreting plasma cell and memory cell [32].

The characteristic features of B cell epitopes are hydrophilicity and accessibility [19]. Therefore, the classical propensity scale methods and hidden Markov model programmed softwares were used from the IEDB (http://www.iedb.org/) analysis resource for the following aspects [33].

• linear B-cell Epitopes Prediction

Using Bepipred tool [34] from the conserved regions with a default threshold value of 0.35.

• Prediction of surface accessibility

Using Emini surface accessibility prediction tool [35] from the conserved region holding the default threshold value 1.000 or higher.

• Prediction of epitopes antigenicity sites

Using kolaskar and tongaonker antigenicity method [36] for detection of the antigenic sites with a default threshold value of 1.042.

T Cell epitope prediction

CD8+ T-cell epitopes prediction: For the prediction of CD8+ T-cell epitopes that bind to MHC-1, we used an artificial neural network (ANN) prediction method [37] from Immune Epitope Database (IEDB) and the predicted IC₅₀ value was less than 100 nm. Prior to the prediction, peptide length was set to 9 amino acids and only frequently occurring alleles were selected. Conserved epitopes having IC₅₀ value less than 100 nm were used for further analysis.

CD4+ T-cell epitopes prediction: The NN-align method under MHC-II binding prediction tool in IEDB was used for MHC-II binding [38]. Human allele references set were selected. The conserved predicted CD4+ T-cell epitopes having higher binding affinity to interact with alleles at IC₅₀ less than 1000 nm was chosen for further analysis.

Population coverage analysis

All CD8+ T-cell and CD4+ T-cell predicted epitopes were assessed for population coverage against the whole world population by using the IEDB population coverage calculation tool [39].

3D structure modeling

Sin Number virus’s glycoprotein and nucleocapsid 3D structures were obtained by RaptorX Property (http://raptorx2.uchicago.edu/ StructurePropertyPred/predict/), which is a web server predicting structure property of a protein sequence without using any template information [40]. For visualization of the 3D structure, we used UCSF Chimera (version 1.8). Chimera currently available within the Chimera package and available from the chimera web site (http://www.cgl.ucsf. edu/cimera) [41]. The 3D structure verifies the service accessibility and hydrophilicity of predicted B lymphocyte epitopes, and identifies all predicted T cell epitopes at the structural level [29].

Prediction of B-cell epitope

The glycoprotein and Nucleocapsid protein was exhibited to Bepipred linear epitope prediction, Kolaskar and Tongaonkar antigenicity and Emini surface accessibility prediction methods in IEDB, to predict the probability of specific regions in the protein to bind B cell receptor, being on the surface, being immunogenic in each. The results are illustrated in Figures 1-3 and also in Tables 3 and 4.

Figure 1: Prediction of B-cell epitopes for glyc oprotein by different scales.

Figure 2: Prediction of B-cell epitopes for nuclocapsid by different scales.

Figure 3: B-cell epitopes proposed for glycoprotein (visualization by UCSF chimera).

Table 3: List of predicted B cell epitopes for glycoprotein

Table 4: List of predicted B cell epitopes for nucleocapsid protein.

Prediction of cytotoxic T-lymphocyte epitopes and interaction with MHC Class I

The sequences of glycoprotein and nucleocapsid analyzed by IEDB MHC-1 binding prediction tool to predict T cell epitopes submit interacting with different types of MHC Class I alleles. Based on Artificial neural network (ANN) with half-maximal inhibitory concentration (IC₅₀) ≤ 100. The results of top epitopes are shown in Figures 3-5 and also in Tables 5 and 6.

Figure 4: B cell epitope proposed for nucleocapsid (visualization by UCSF chimera).

Figure 5: Glycoprotein epitope 208MTLPVTCFL216 suggested for cytotoxic T cell interaction (MHC 1) (visualization by UCSF chimera).

Table 5: The most potential 2 CD8+ T-cell epitopes of glycoprotein with interacting MHC-1 alleles.

Table 6: The most potential 2 CD8+ T-cell epitopes of nucleocapsid with interacting MHC-1 alleles.

Prediction of T helper cell epitopes and interaction with MHC Class II

The sequences of glycoprotein and nucleocapsid analyzed by IEDB MHC-II binding prediction tool based on NN-align with (IC₅₀) ≤ 1000. The results of two top epitopes are shown in Figure 6 and in Tables 7 and 8.

Figure 6: Glycoprotein epitope 458YTFTSLFSL466 suggested for cytotoxic T cell interaction (MHC 1) (visualization by UCSF chimera).

Table 7: The most potential 2 CD4+ T-cell epitopes of glycoprotein with interacting MHC-II alleles.

Table 8: The most potential 2 CD4+ T-cell epitopes of nucleocapsid with interacting MHC-II alleles.

Analysis of the population coverage

Epitopes with high affinity to interact with MHC-I and II international alleles were applied to population coverage analysis in IEDB. There were 22 and 70 epitopes in glycoprotein, and 10 and 44 in nucleocapsid in MHC-I and II have given a high percentage of population coverage. The results were shown in Tables 9 and 10 and proposed epitopes in both MHC class I and II were shown in Figures 7 and 8 and also in Tables 11 and 12.

Figure 7: Nucleocapsid protein epitopes suggested for both MHC-I and MHC-II interaction (visualization by UCSF chimera).

Figure 8: Glycoprotein epitopes suggested for T helper cell interaction (MHC-II) (visualized by USCF chimera).

Table 9: Glycoprotein epitopes with highest population coverage percentage.

Table 10: Nucleocapsid epitopes with high population coverage percentage.

Table 11: Proposed epitopes in both MHC class I and II of nucleocapsid.

Table 12: Proposed epitopes in both MHC class I and II of glycoprotein.

Vaccination, which is a form of preventative medicine, has an extremely powerful impact on the defense against infectious agents that cause disease and death, and sometimes vaccines have the ability to completely eradicate disease from the globe (e.g. smallpox) [42,43]. None the less, there are still many human diseases have no effective vaccines [42]. One of these diseases is HCPS especially caused by SNV. Despite SNV is not transmitted from person-to-person; SNV is highly pathogenic (case fatality rate 30 to 50%) [44] regardless of age, [45] health status, [46] or access to advanced medical care, and was responsible for the 2012 Yosemite National Park outbreak [21,47]. The absence of effective vaccines, post exposure prophylactics, or therapeutics to prevent HCPS caused by SNV participated in the worry experienced by about 270,000 Yosemite visitors who got notice of possible exposure [46]. This brings us to the urgent need for the nondevelopment of an effective vaccine against SNV. In our study, we analyzed envelope glycoprotein and nucleocapsid of SNV, and predicted most conserved and immunogenic B- and T-cell epitopes which could trigger humoral as well as cellular mediated immune responses in order to develop an effective epitope-based vaccine. This concept of epitopebased vaccine, in which a cocktail of B- and T-cell epitopes are used to enhance protection against viral infection, was applied by Steward et al. in a study on the Respiratory syncytial virus (RSV), where RSV-specific humoral and cellular immunities were induced after immunization with a cocktail of peptides. Following challenge infection, a 190-fold reduction in RSV titer was observed in the lungs of immunized mice [48]. Thus the combination of humoral and cellular immunity is more promising at clearing viral infection than humoral or cellular immunity alone. A number of studies have demonstrated the strong antigenicity of both the membrane glycoprotein and nucleocapsid protein. However, neutralizing antibodies have been shown to be sufficient to confer protection in passive immunization of experiments using glycoproteins-specific monoclonal and polyclonal antibodies [2,15,49]. The nucleocapsid is a multifunctional protein that plays several roles in virus replication and assembly and is the major antigen to which the host mounts the humoral immune response. It is highly conserved among various Hantavirus genotypes and is expressed early with huge amounts in the cell during infection. The antibody to the nucleocapsid protein is demonstrated early in the immune response [50,51].

The conserved predicted epitopes for B- and T-cell were retrieved from all declared strains of SNV glycoprotein and nucleocapsid in NCBI databases until Nov 2016. As obviously clear from the result, the conserved regions cross retrieved strains accompanied with a good level of confidence. To design an effective peptide antigen for B cell, epitopes were predicted from the conserved sequences by Bepipred prediction tools should be passed both Emini surface accessibility and Kolaskar and Tongaonkar antigenicity thresholds, and peptide sequences should best be within 8-22 amino acids in length [19]. However, 20 predicted epitopes of glycoprotein were satisfied both Emini surface accessibility and Kolaskar and Tongaonkar antigenicity tests in IEDB. The linear epitope 743CKKYAYPWQT752 showed high succeeded score in both tests as illustrated in Table 3. For nucleocapsid, 9 predicted epitopes for B cell were succeeded both Emini surface accessibility and Kolaskar and Tongaonkar antigenicity tests. The linear epitope 271QVDESKVSD279 was 9 amino acids residue long, when we decreased the epitope length to 8 amino acid 271QVDESKVS278 we got the highest score in accessibility with a succeeded score in antigenicity tests, as illustrated in Table 4. Sagadevan et al. predicted IMASKSVGS/TAEEKLKKKSAF from nucleocapsid protein of Hantaviruses causing HCPS, based on the mean percent prediction probability score, to be the best candidate B-cell epitope for developing immunoassays in the detection of antibodies to hantaviruses causing HCPS [51]. But this peptide has a low value of antigenicity and does not pass the threshold of Kolaskar and Tongaonkar antigenicity test in IEDB. Thus it may be useful for the development of immunodiagnostic tools towards pan-detection of Hantavirus antibodies causing HCPS but not effective vaccines.

T-cell mediated immunity plays a crucial role in the defense against viral infections [52]. The inclusion of T-cell epitopes in the vaccine development induces a strong and long lasting immune response and antigenic drift where antigen can easily escape the antibody memory response [19,22]. While CD8+ cytotoxic T-cells generally recognize intracellular peptides displayed by HLA class I molecules, CD4+ T-helper cells generally recognize peptides from the extracellular space, displayed by HLA class II molecules (CD4+ T-cell epitopes) [22]. A potential problem in the development of CTL epitope-based vaccines is the large degree of MHC polymorphism and the need for the understanding of HLA restrictions in the target population for vaccines [53]. In our study, we chose the most frequently occurring alleles in IEDB for MHC binding prediction. For glycoprotein, among 168 CD8+ T cell conserved predicted epitopes, we found the epitope 208MTLPVTCFL216 had higher affinity to interact with 8 alleles (HLA-A*02:06, HLA-A*02:01, HLA-A*30:01, HLA-A*68:02, HLA-B*58:01, HLA-B*57:01, HLA-C*03:03, HLA-C*14:02) and the epitope 458YTFTSLFSL466 had the affinity to interact with 7 alleles (HLA-A*02:06, HLA-A*02:01, HLA-A*68:02, HLA-B*39:01, HLA-C*03:03, HLA-C*12:03, HLA-C*14:02). These two epitopes had very good population coverage for class I alleles throughout the world, as shown in Table 5 and Table 12. Also we found 674 CD4+ T cell predicted conserved epitopes interacting with MHC-II alleles, as shown in Table 7. Epitopes 50ILLTQVADL58 and 165LLSSRIQVI173 had high affinity to interact with 22 and 21 alleles, respectively. They interacted with HLA-DPA1*01/2/3, -DPA1*02, -DQB1*02/3/4/5/6, -DRB1*03/4/7, -DRB3*01, -DRB4*01, -DRB5*01. Interestingly, these two epitopes covered 99.65% of the world population. For nucleocapsid, we found 66 CD8+ T cell and 228 CD4+ T cell conserved predicted epitopes, as shown in Tables 6 and 7. Most importantly, the epitopes 125YILSFALPI133 and 239FLAARCPFL247 had high affinity to interact with both MHC classes, I and II, and also they had excellent population coverage for Class I and II alleles throughout the world (Table 11). Epitope FLAARCPFL was the same epitope predicted by Sathish et al. in 2017 as a suitable epitope for vaccine specific to the SNV genotype [54].

The HLA-A*02 supertype is expressed in all main ethnicities in the 39-46% range. Since many peptides that bind A*02:01 also display degenerate binding (binding to multiple alleles), an A2 supertype multiepitope vaccine could be developed to allow broad, nondevelopment ethnically biased population coverage [53] Of greater significance is the finding that all CD8+ T cell predicted epitopes showed interactions with HLA-A*02 and most of HLA-A*30, -A*68, -B*08, -B*39, -B*57 and C*04:01 which are more predominant HLA alleles in the US and North America where infection and outbreak of SNV have occurred [54]. On the other hand, it was found that all the CD4+ T cell epitopes interacted with HLA-DRB1*03:02, -DRB1*03:01 and -DRB1*07:01 which are among most common haplotype observed within American populations, that reported by Martin et al. in 2007 [55].

Bioinformatics approaches are a fundamental component of a high throughput pipeline for in silico mapping of thousands of potential epitopes, minimizing the time and cost involved in the experimental testing of such peptides [53]. Target populations for the vaccine are inhabitants of SNV-endemic geographic areas and people with occupational risks of infection [1]. Analysis for the proteasomes and transporter associated with antigen processing (TAP) efficiency is required; in order to predict the candidate epitopes based on the processing of the peptides in vivo. Along with in silico study, further in vivo and in vitro experiments are needed to prove the effectiveness of triggering and mounting an immune response. Our study for the first time has predicted a cocktail of B- and T-cell epitopes for designing an effective vaccine against HCPS caused by SNV.