1 Introduction

Marburg virus (MRV) is genus of Filoviridae family. MRV is negative stranded & non-segmented RNA virus, that is responsible for severe hemorrhagic fever, known as marburg hemorrhagic fever (MHF) in both humans and non-human primates. The MRV infection has approximately 23 to 100% fatality and lethality rates in humans and non-human primates (Mehedi et al., 2011). The systemic viral replication of MRV interferes with immune and inflammatory activities, the consequence of which are serious pathological features in patients, like hemorrhages, edema, coagulation imbalance, multiple-organ failure and shock, often resulting in death (Bente et al., 2009). Before the discovery of Ebola in 1967, the first MRV infection was observed in Germany and Serbia (Mehedi et al., 2011)followed by Zimbabwe/South Africa in 1975 (Conrad et al., 1978)Angola in 2004 (Towner et al., 2006). According to the report of CDC 2014, the recent outbreak was in Uganda from 2007 to 2014. Few studies demonstrated that MRV is highly infectious and very stable in experimental aerosol exposure (Alves et al., 2010), which raises the concern that MRV may be very suitable to be used as biological weapon (US Centres for Disease Control and Prevention/“Bioterrorism Agents/Diseases” report). Despite of all, in current scenario no treatment is available against MRV infection (Cross et al., 2018). The main enviable feature for any vaccine candidate is that the molecule should activate cell-mediated (T-Cell) and humoral (B-Cell) immune response followed by memory cell formation. CD4 + T-cells activation is mandatory for a competent humoral immune response for the induction of Immunoglobulin-G and memory B cells. CD4 + T-cells primarily recognize antigen peptides by CD4 co-receptor and only recognize the major histocompatibility complex (MHC) II protein on antigen-presenting cells; then memory B cells make a repository of infected virus for the farther prophylaxis arrangement (Clem, 2011). Filoviridae family virus consists of 7 structural proteins, among all, highly glycosylated (N- and O-linked glycans) envelope glycoproteins (Env GPs) are present over the cell surface. Host proteases like furin is responsible for the proteolysis of GP, resulting in two subunits, GP1 and GP2, linked by a disulfide bond (Volchkov et al., 2000). These GPs mediate and lead the viral entry into host cells (Takada et al., 1997)^,(Wool-Lewis and Bates, 1998), thus GPs are considered to be the ideal target for neutralizing antibodies against filoviruses. In current era as the electronic support increases in the life sciences, computational based approach provides access to researchers to deal with huge number of genome and proteome data of virus. Immunogenic, non-toxic and peptide-based vaccines would prove to be a good alternative treatment option for the management of MRV infection. In this article, we have focused on the special epitopes candidates those have non mutagenic tendency, which were thoroughly screened via protein variability server. We have identified various small fragments of Env GP those don’t have the mutation hot spot, which will lead the less chance of viral resistance. In current COVID scenario scientists have encountered with the toxicity issue of vaccines so here we have seriously focused on the non-toxic and highly immunogenic multi-epitopes from Env GP proteins of MRV virus. The recognition of world-wide HLA coverage analysis of CD4 + T-cell epitopes in Env GP protein was carried out by using artificial neural network algorithm (ANN) implemented in IEDB and NetMHCIIPan Server. This article, is clearly depicted that the focus of this article is to provide a rapid, cost effective and efficient vaccine candidate and process by using epitopes of viral proteins. This strategy was designed to keep in the mind of current COVID situation where we were looking safe and effective vaccine. Since a very recent statement of WHO, also claimed the inflammable problem of marburg virus as a “scary & deadly disease”. (https://www.express.co.uk/news/science/1474916/marburg-virus-news-disease-scary-deadly-world-health-organisation-africa-spt)

2 Materials and methods

2.6

2.6 Population coverage analysis

The population coverage rate of the final selected epitopes was calculated by using the IEDB population coverage tool (http://tools.immuneepitope.org/tools/population/iedb_input) (Bui et al., 2006). The predicted epitopes with their all-binding HLA alleles for the worldwide distribution were tabulated. IEDB server extract all allele genotypic frequencies related data from Allele Frequency database, which comprises with allele frequencies form huge population set of 115 countries and 21 different ethnicities grouped into 16 different geographical areas. The schematic representation of the entire methodology of in silico CD4 + T-cell epitope prediction and HLA distribution of MRV is mentioned in Fig. 1.

Close

Fig. 1

The schematic representation of the entire methodology.

Export to PPT

3 Results

3.1

3.1 Retrieval of Env glycoprotein sequences of MRV and multiple sequences alignment (MSA)

All 37 Env glycoprotein sequences of different strains of MRV (mentioned in Supplementary Data Set part 1) with more than 89% of similarity, were retrieved from uniprot database. CLUSTAL Omega was used for the identification of evolutionary relationship (as shown in Fig. 2) and percent of similarity between all 37 protein sequences of Env protein of MRV (mentioned in Supplementary Data Set part 2). Q6UY66|VGP_MABVO was considered as reference sequence and its comparative sequence similarity coverage with other proteins more than 89%. On the basis of similarity coverage, authors obtained the information about the variation in the sequences that may lead to hot points of the virus mutations (Fig. 3). Considering this, Protein Variability Server (PVS) was used to select the non-variable fragments of the viral Env proteins, which were used for the selection and identification of the most effective and immunogenic epitopes. Total ten fragments were obtained (Table 1) by PVS study. Often, only eight fragments were selected for the identification of epitopes because fragment no. 03 (PEIKPTSTPTDAT, 13 amino acids from 240 to 252) and 06 (NLSTLS, 06 amino acids from 350 to 354) have very lesser numbers of amino acids than the selection windows length for HLA-II peptide which was 15.

Close

Fig. 2

Phylogenetic relationship between all 37 Env proteins of all strains of Marburg virus.

Export to PPT

Close

Fig. 3

Protein Variability Plot of Env proteins of all 37 Sequences of Marburg Virus (MRV).

Export to PPT

Table 1 Protein variable fragments of EnvGP of MRV obtained via Protein Variability Server run.

Set No.	Start	End	Sequence	Vaxijen Score (0.4 Threshold)
1	3	27	TTCFFISLILIQGIKTLPILEIASN	0.6525(Probable ANTIGEN)
2	29	230	QPQNVDSVCSGTLQKTEDVHLMGFTLSGQKVADSPLEASKRWAFRTGVPPKNVEYTEGEEAKTCYNSVTDPSGKSLLLDPPTNVRDYPKCKTIHHIQGQNPHAQGIALHLWGAFFLYDRIASTTMYRGKVFTEGNIAAMIVNKTVHKMIFSRQGQGYRHMNLTSTNKYWTSSNGTQTNDTGCFGTLQEYNSTKNQTCAPSK	0.4850(Probable ANTIGEN).
3	240	252	PEIKPTSTPTDAT	1.3071(Probable ANTIGEN).
4	254	316	LNTTNPNSDDEDLTTSGSGSGEQEPYTTSDAVTKQGLSSTMPPTPSPQPGTPQQGGNNTNHSQ	0.4729(Probable ANTIGEN).
5	324	348	NTNTTAQPPMPSHNTTTISTNNTSK	0.4109(Probable ANTIGEN).
6	350	354	NLSTLS	0.8725(Probable ANTIGEN).
7	365	379	NTQSMATENEKTSAP	0.6994(Probable ANTIGEN)
8	381	405	KTTLPPTESPTTEKSTNNTKSPTTM	0.4557(Probable ANTIGEN).
9	415	496	SPSSTPNSTTQHLIYFRRKRSILWREGDMFPFLDGLINAPIDFDPVPNTKTIFDESSSSGASAEEDQHASSNISLTLSYLP	0.6654(Probable ANTIGEN).
10	498	680	SENTAYSGENENDCDAELRIWSVQEDDLAAGLSWIPFFGPGIEGLYTAGLIKNQNNLVCRLRRLANQTAKSLELLLRVTTEERTFSLINRHAIDFLLTRWGGTCKVLGPDCCIGIEDLSRNISEQIDQIKKDEQKEGTGWGLGGKWWTSDWGVLTNLGILLLLSIAVLIALSCICRIFTKYIG	0.4865(Probable ANTIGEN).

3.4

3.4 Population coverage analysis

IEDB population coverage server was used for the identification of population coverage of our screened non-mutagenic, antigenic and non-toxic peptides. This analysis indicates that 11 peptides (ILIQGIKTLPILEIA, LILIQGIKTLPILEI, VFTEGNIAAMIVNKT, AFFLYDRIASTTMYR, FTEGNIAAMIVNKTV, QHLIYFRRKRSILWR, LIQGIKTLPILEIAS, SKRWAFRTGVPPKNV, WGAFFLYDRIASTTM, FDESSSSGASAEEDQ and TAGLIKNQNNLVCRL) show highest HLA coverage. Each region has 100% population coverage except South Africa, which makes an average of 96.79% of world population coverage, as mentioned in Fig. 4.

Close

Fig. 4

Bar-diagram representation of population coverage of final 11 epitopes against different regions of globe.

Export to PPT

4 Discussion

Immunization is one of the major, successful and cost-effective preventive strategy for community health to combat against the fatal infectious diseases globally (Chabot et al., 2004). Although, there is continuous development in the area of vaccines, classical vaccination like whole pathogen immunization is still popular. These types of immunizations are known to produce long lasting and strong immunity, but the major concern is that it may induce strong allergic reactions (Skwarczynski and Toth, 2016). So, peptide-based vaccines (PBVs) or multi epitope vaccines have now become a better choice for safe vaccination. PBVs are a striking alternative approach that depends on selection and usage of short peptide fragments to engineer the stimulation of extremely targeted immuno-protective responses, avoiding allergenic sequences (Li et al., 2014). In this context multi epitope-based vaccines are competent of stimulating strong immunogenic responses and safer option than whole protein-based vaccines. Earlier various studies in public domain are also encouraging the efficacy and impact of multiple epitopes based in silico vaccinology (Chaitra et al., 2005; Parida et al., 2007; Wiwanitkit, 2007; Gupta et al., 2010; Shey et al., 2019).

As we all are aware of antimicrobial resistance, which is a severe problem of healthcare at present and affecting millions of people around the globe. Antiviral resistance on the other hand, has been considered as a lesser threat than antibiotic resistance because unlikely drugs, vaccines are used for the prophylactic roles (Kennedy and Read, 2017). However, lately vaccine resistance is also becoming an important and inflammable problem.

Viruses are known for the high mutational rate in very short replication time, which is led by the nucleotide sequence context on the template molecule as well as by external environmental factors. This kind of genetic variation is the guarantee of virus survival in extreme conditions, as the significance of high rate mutation escort to formation of quasi-species or new viral strains (Lauring and Andino, 2010). Single-stranded RNA virus like influenza and marburg often carries error prone polymerases, which habitually induce at least one (range 0.1–10) incorrect base selection during every round of replication and initiate rapid materialization for vaccine resistant (Domingo and Holland, 1997). In this article authors pinpoint three nodes of vaccine that are identification of non-mutagenic, highly antigenic and non-toxic peptides from Env gylcoprotein of marburg virus (MRV), which may have potential to point at a direction in designing of a new vaccine to combat marburg virus induced infections.

Env gylcoprotein of any virus is supposed to be a probable target for the vaccine construction, because only Env gylcoprotein is accountable for the docking and connection of virus with any kind of human protein or receptor which lead the entry of virus into the cell (Janahi et al., 2017). In current study, authors focused on selection of non-mutagenic, immunogenic and non-toxic epitopes of Env gylcoprotein of MRV. As we know, viruses are known for their characteristics of high genomic mutation rates which provides the protection coverage for the virus and this is one of the major causes of vaccine failure (Laughlin et al., 2015). Considering this problem, we have focused on 37 variable strains of Env glycoproteins with 89–90% homology. Selected amino acid sequences were considered as input data for Protein Variability Server (PVS) to identify the most probable hot spots of mutations in Env glycoprotein. After implementation of PVS techniques, 10 different non variable/mutagenic and antigenic fragments of Env glycoprotein were generated. Out of 10, 8 fragments were selected for further screening, those have at least equal to or more than 15 peptides length. In this study authors have only focused on the selection and identification of CD4 + T cell mediated immunity because, CD4 + T-cells activation is an initial and mandatory factor for a competent humoral immune response for the induction of immunoglobulin-G and memory B-cells. CD4 + T-cells primarily recognized by antigen peptides by CD4 co-receptor and only recognized by the major histocompatibility complex (MHC) II protein on antigen-presenting cells; then memory B cells construct a repository of infected virus for the further prophylaxis arrangement (Clem, 2011). Cytotoxic T-cells or TCD8 + have different roles in immunity, which is related to therapeutic understanding not a prophylaxis segment. That’s why for rapid and cost-effective development for prophylaxis vaccine development CD4 + T cell alone is capable and important to induce protective response. NetMHCPanII server was used for the identification of predicted putative CD4 + T-cell epitopic core sequences in each of fragments of Env proteins along with their respective binding HLAs. Only CD4 + T-cell epitopes were chosen because CD4 + T-cells are the only immune cells that initially identify antigenic proteins and forms major histocompatibility complex (MHC) II, followed by setting up the configuration of memory B cells for the further prophylaxis arrangement (Gupta et al., 2010). Total 5224 HLAs were listed in the NetMHCpan server (DQ: 2912, DP: 2247, DRB4: 06, DRB1: 15, DRB3: 29, DRB5:15). All eight fragments were screened with all 5224 HLAs. After generation of this huge data, we chose only strong peptide binder among all 8 sets of fragments with lowest affinity score. Best 29, strong binder epitopes were extracted, as mentioned in Table 2. Interferon -gamma (IFN-gamma) potential was also evaluated for all 11 epitopes, and we found in category 1 (IFN-gamma vs Non IFN-gamma) 5 out 11 epitopes were showing potential to induce the IFN-gamma response against viral infection and among 5 epitopes our best top two selected epitopes (ILIQGIKTLPILEIA: 1034 HLAs and LILIQGIKTLPILEI: 336 HLAs) were present those having highest HLA coverages. Second category (IFN-gamma vs other cytokinin) were showed all 11 epitopes were showing better potential to induce the IFN-gamma as compared to other cytokines (details as mentioned in Table 3). About the physiochemical characteristics we found that most of finally selected epitopes have broad range of HLA coverage, but they were not showing good solubility in water and thermodynamically unstable in biological system, since its very usual with all small peptides and nucleosides, they are very prone to degradation via circulating proteases and nucleases. These enzymes are abundantly found in the biological systems, that’s why in this case authors will suggest the nanoparticle (NP)-formulation coting of peptides to shield themselves from proteases enzymes. NPs have the ability to transport weak antigens or vaccines to the mature DCs within the secondary lymph organs. Nano formulation is able to protect peptides, from degradation by proteases. By using NPs as a delivery system, we can initiate stronger immune responses. Once NPs reaches DCs, then controlled release of epitopes can be achieved through chemical modification on their surface, thus activation of DCs can be achieved more efficiently (Jia et al., 2018). Based on all the previously mentioned information we strongly believe that, nano formation system can prove to be an effective and potentiating delivery system for multi-epitopes. Even recently developed COVID-19 vaccine by Pfizer–BioNTech and Moderna also encapsulating the mRNA in lipid nanoparticles (Chaudhary et al., 2021) to save them from nucleases enzymes in biological system. Biodegradable nanoparticles generally made up of poly(D,L-lactic acid-co-glycolic acid)/PLGA are approved for the use of human (Elmowafy et al., 2019). PLGA based peptide NPs are very popular and efficient option of vaccine delivery system for targeting DCs and the development of DCs based cellular vaccines (Athanasiou et al., 2017). So, in that case authors will suggest PLGA nano-formulation will be most suitable candidate for this encapsulation while it already approved by FDA in drug formulations and able to protect the non-water soluble and thermodynamically unstable epitopes. Furthermore, all 29 epitopes were screened based on antigenicity and final 11 epitopes were selected as probable multi epitopes vaccine candidates, with non-toxic properties and high world population coverage (as shown in Fig. 5). This result shows that the proposed epitopes would be significant vaccine contenders for large proportion of the human population which is around 96.78% globally. In short, this study generously focused on a strong prophylactic intervention against MRV with very low possibility of resistance, highly antigenic, non-toxic and high world population coverage. However, the T-cell stimulation potential of the predicted putative CD4 + T-cell epitopes are required to be validated by wet lab experiments for their efficient use as peptide vaccine candidates against marburg virus (MRV).

Close

Fig. 5

Bar-diagram representation of final 11 epitopes with their HLAs coverage analysis.

Export to PPT

5 Conclusion

In this study, immuno-informatics tools were employed to design a putative vaccine peptide coding for multiple T-cell CD4 + epitopes. Total 11 peptides were minutely screened based on high antigenicity, non-mutagenic, non-toxic and broad HLA coverage. This computer-based study uses strong technical and logical methodologies that help this study to become more precise and reproducible in real time models. The authors are very hopeful, however, the T-cell stimulation potential of these predicted peptides containing the core amino acid sequences are to be validated by using in vitro and in vivo experiments for their competent use as multiepitope vaccine candidates against MRV infection. This study can be highly useful for designing newer vaccine strategies to prevent and/or lower the death toll attributed to MRV infection in future.