Abstract

Introduction: Human parvovirus B19 (B19V) is small non-enveloped, single-stranded DNA virus belong to genus Erythrovirus. B19V can cause erythema infectiosum (fifth disease), oligoarthritis, hydrops fetalis and a plastic crisis in patients with sickle cell anemia. A variety of vaccine strategies have been employed targeting immune responses. However their results were controversy with a limiting in availability of viral antigen. Since B19V replicates predominantly in erythroid progenitor cells of human bone marrow, this makes a peptide-based vaccines a promising strategy for development of vaccine against B19V with less allergenic and reactogenic responses. The aim of the present study was to design an efficient multi-epitope vaccine for human B19 virus using VP1 glycoprotein. Material and method: Thirty six sequences of VP1 glycoprotein were retrieved from NCBI database in December 2017 and aligned to determine the conservancy between the retrieved strains. The IEDB different analysis resources were used to predict epitopes that could act as promising peptides vaccine against parvovirus B19. The predicted epitopes were further assessed for population coverage against the whole world population. Results: The epitopes _214-PEVP_-217, _675-GLHQPPP_-681 and _554-SLRPGPVSQPYH_-565 were found to be the most potential epitopes against B cells. For the T cell three epitopes namely _155-FRYSQLAKL_-163, _302-CTISPIMGY_-310 and _316-YLDFNALNL_-324 showed high affinity to MHC-I alleles. The epitopes (core) _155-FRYSQLAKL_-163, _438-FYVLEHSSF_-446 and _404-WVYFPPQYA_-412 showed high affinity to interact with MHC-II alleles. _155-FRYSQLAKL_-163 and _438-FYVLEHSSF_-446 showed high coverage for whole world population with percentage of 99.73% and 94.85% respectively. Conclusion: This study proposed eight epitopes for B and T cells that could be a powerful multi epitope vaccine against B19V. Particular concern directed towards the epitope _155-FRYSQLAKL_-163 which demonstrated merits by reacting efficiently with both MHC-I and MHC-II alleles. Clinical trial is required to proof the efficacy of these epitopes as promising candidate vaccine against parvovirus B19.

1. Introduction

Human parvovirus B19 (B19V) is a small, non-enveloped, single-stranded DNA virus, belongs to the genus Erythrovirus. The genus belongs to large family of Parvoviridae in which the word (parvum) means small. The family formed small capsids about 25 nm in diameter ^{1, 2}. The first discovery of B19V was in 1975 and around 1980s scientist linked the B19V infection to certain disease like a plastic crisis and fifth disease (erythema infectiosum) ^{1, 3, 4}. B19V causes aplastic crisis in patients with shortened red cell survival due to induction of apoptosis in infected erythroid progenitors. Erythema infectiosum also known as slapped cheek (fifth disease). It can also causes oligoarthritis and hydrops fetalis or intrauterine death in infected fetuses ^{3, 4}. The prevalence of B19V is highly common among children with sickle cell anemia and it leads to deaths due to acute and chronic anemia’s (crisis episodes). Moreover the prevalence of B19V among pregnant women remains as a major cause of high maternal mortality especially in Sudan. ^{5, 6, 7}

Parvovirus B19 is transmitted by respiratory aerosol spread from individuals with acute infection, blood products and vertical transmission from mother to fetus. The majority of infections occur during childhood and the risk being greatest in the first two trimesters of pregnancy ¹. Regional epidemics occur during late winter and spring due to absence of a lipid envelope in the virus and their genomic stability which makes virus notoriously resistant to heat inactivation and solvent detergents ^{1, 8, 9, 10}. This demonstrated seasonal variation in outbreak of the virus throughout the year and high prevalence among children and compromised and pregnant women ^{4, 11, 12}.

The parvovirus B19 genome consists of a single stranded linear molecule of 5596 nucleotides, composed of an internal coding sequence of 4830 nucleotides flanked by terminal repeat sequences of 383 nucleotides each. B19V contains two structural proteins, viral protein 1(VP1) and viral protein 2 (VP2). VP1differs from VP2 only in an N-terminal “unique region” composed of 227 additional amino acids which are mostly located outside the virion ^{1, 2, 13}. VP1 contains many linear epitopes recognized by neutralizing antibodies and accessible to antibody. This binding makes VP1 most fit capsid protein for designing a vaccine ^{1, 14}.

Till 1993 no human B19V vaccine is available. The development of such vaccine has been hampered by the limited availability of viral antigen since B19V replicates predominantly in Erythroid progenitor cells in human bone marrow. Moreover there is no convenient in vitro culture system available for routinely propagating large amounts of virus. However recently a recombinant design vaccine was established and their efficacy was controversial ^{15, 16, 17}.

In this report we attempted to design a vaccine for human B19 virus using VP1 glycoprotein as an immunogen to invigorate protective immune response using immunoinformatics tools. The insilico prediction of epitopes for appropriate protein residues would help in production of peptide vaccine with intense immunogenic and insignificant allergenic impact that opposed to antibody creation that relies on biochemical examinations. The traditional vaccines can be costly, time consuming and not generally work. Moreover formulation of attenuated or inactivated form of microorganism contains a hundred of unnecessary proteins that induced immunity, allergenic or reactogenic responses ^{18, 19, 20}.

2. Material and Methods

A total of 36 strains of VP1 glycoprotein were retrieved from NCBI (https://www.ncbi.nlm.nih.gov/protein?term=Human+parvovirus+B19+VP1) in Dec 2017 from different parts of the word. The retrieved strains and their accession numbers were depicted in Table 1.

The retrieved sequences were subjected to phylogenetic analysis to determine the common ancestor of each strain using different tools from (https://www.phylogeny.fr) ²¹.

The retrieved sequences of VP1 strains were aligned to obtain conserved regions using multiple sequence alignment (MSA). Sequences were aligned with aid of the ClustalW as implemented in the BioEdit program, version 7.2.5 ²².

To detect the candidate epitopes from VP1 of Parvo virus B19, for B-lymphocytes and T-lymphocytes, several analysis prediction tools from Immune Epitope Database (IEDB) (https://www.iedb.org/) ^{22, 23} were used.

B cell epitope is the portion of an immunogen which interacts with B lymphocytes. B-lymphocytes upon exposure differentiated into plasma cells and memory cells. Thus B cells epitopes are shown to being accessible and antigenic ²⁴. Accordingly the classical propensity scale methods and hidden Markov model programmed software were used for the following aspects:

BepiPred from immune epitope database (https://toolsiedb.ofg/bcell/) ²⁵ was used as a linear B-cell epitopes prediction from the conserved region of VP1 glycoprotein with a default threshold value of 0.5.

Emini surface accessibility prediction tool of the immune epitope database (IEDB) was used (https://tools.immuneepitope.org/tools/bcell/iedb) ²⁶. The surface accessible epitopes were predicted from the conserved region of VP1 glycoprotein with default threshold value 1.000.

The kolaskar and tongaonker antigenicity method was used to determine the antigenic sites with a default threshold value of 1.025 (https://tools.immuneepitope.org/bcell/) ²⁷.

Analysis of peptide binding to MHC class I molecules was assessed by the IEDB MHC-I prediction tool at (https://tools.iedb.org/mhci/n). MHC-I peptide complex presentation to T-lymphocytes underwent several steps. For instance the attachment of cleaved peptides to MHC-1 molecules was predicted by Artificial Neural Network (ANN) ^{28, 29, 30}. Also all epitopes lengths were set as 9 mers. Besides, all the conserved epitopes that bind to alleles at score equal to or less than 100 half-maximal inhibitory concentrations (IC-50) were selected for further analysis ³¹.

Analysis of peptide binding to MHC class II molecules was assessed by the IEDB MHC II prediction tool at (https://tools.immuneepitope.org/mhcii/) ^{32, 33}. For MHC-II binding prediction, human allele references set were used. MHC class II groove has the ability to bind peptides with different lengths. Therefore for the analysis, the NN-align as prediction method from IEDB MHC-II prediction tool were used. It allows for identification of the MHC class II binding core and epitopes binding affinity All conserved epitopes that bind to many alleles at score equal or less than 500 half-maximal inhibitory concentrations (IC-50) were selected for further analysis.

For the calculation of the population coverage for all potential MHC-I and II epitopes bindings, the IEDB tools (https://tools.iedb.org/tools/population/iedb_input) ³⁴ was used. The VP1 glycoprotein of B19V was assessed for population coverage against the whole world with selected MHC-I and MHC-II interacted alleles.

Raptor X protein structure prediction server (https://raptorx.uchicago.edu/StructurePrediction/predict/) was used for creation the 3D structure of the VP1 glycoprotein of B19 virus. The reference sequence [AAV35057.1] was used as an input and Chimera 1.8 was used as a tool to visualize the selected epitopes belonging to B cell and T cell (MHC-I and MHC-II). Homology modeling was used for visualization of the surface accessibility of the B lymphocytes predicted candidate epitopes as well as for visualization of all predicted T cell epitopes in the structural level.

3. Results

All retrieved strains were representing in phylogenic tree Figure 1. From the figure the Germany and Japan strains share common ancestor. Also Ghana, Brazil and United Kingdom shared common ancestor. Also in another site USA, Japan, China, Korea, German, United Kingdom and Belgium shared common ancestor.

Multiple sequence alignment was represented in Figure 2. Sequence alignment showed that some regions were mutated region, and dots show the conservancy between different retrieved sequences of the strains.

The reference sequence of the viral protein (VP1) was subjected to Bepipred linear epitope, Emini surface accessibility and Kolaskar and Tongaonkar antigenicity methods in IEDB to predict the likelihood of specific regions in the protein that bind to B cell receptor, being in the surface and immunogenic respectively. The thresholds of Bepipred linear epitope, Emini surface accessibility and Kolaskar and Tongaonkar antigenicity were shown in Figure 3. For Bepipred linear epitope prediction method, the average binding score of viral protein to B cell was 0.5. Fifty two epitopes were predicted as a linear epitopes and only 23 epitopes were conserved regions. Emini surface accessibility provided only thirteen epitopes that were potentially predicted on surface by passing the default threshold 1.000. Kolaskar and Tongaonkar antigenicity provided only eight epitopes that gave score above the default threshold 1.025. All the epitopes predicted by these different tools against B cell were provided in Table 2. Accordingly three conserved epitopes were successfully predicted to elicit the B cell lymphocytes since they were conserved among all retrieved strains, got higher score values in Emini surface accessibility and Kolaskar and Tongaonkar antigenicity prediction methods. These three epitopes were _214-PEVP_-217, _675-GLHQPPP_-681 and _554-SLRPGPVSQPYH_-565. The three dimension structural (3D) level of these epitopes was shown in Figure 4.

The reference structural protein (VP1) was analyzed using IEDB MHC-1 binding prediction tool to predict T cell epitopes interacting with different types of MHC-I alleles. Thirty three conserved peptides were predicted to interact with different MHC-1 alleles. The peptide _155-FRYSQLAKL_-163 had higher affinity to interact with 6 alleles, followed by _302-CTISPIMGY_-310 that interacted with four alleles and _316-YLDFNALNL_-324 that interacted with only two alleles as shown in Table 3. These three epitopes and their positions in structural level of VP1 were shown in Figure 5. The thirty three conserved peptides and their interaction with different MHC-1 alleles were supplemented in an extra sheet 1.

Seventy nine (79) conserved epitopes were predict from reference viral protein (VP1) which have ability to interact with MHC-II alleles. As shown in Table 4, three peptides (core) _155-FRYSQLAKL_-163, _438-FYVLEHSSF_-446 and _404-WVYFPPQYA_-412 demonstrated higher affinity to interact with MHC-П alleles. The three dimensional structural level (3D) of these epitopes within VP 1 protein was shown in Figure 6. The other core epitopes and their corresponding alleles that interacted with MHC-П were supplemented in an extra sheet- 2.

As shown in Table 5 two epitopes, _155-FRYSQLAKL_-163 and _316-YLDFNALNL_-324, interacted with most frequent MHC-I alleles and they demonstrated population coverage against the whole world 56.26% and 43.86% respectively. Two epitopes, _155-FRYSQLAKL_-163 and _438-FYVLEHSSF_-446, demonstrated population coverage against the whole world 99.37% and 94.85% respectively against MHC-II. Interestingly the epitope _155-FRYSQLAKL_-163 was shown to interact with both MHC-I and MHC-II alleles. As shown in Table 6 the overall epitope sets for the predicted epitopes against MHC-I and MHC-II alleles was 99.44%.

4. Discussion

Recently there have been considerable efforts toward developing a synthetic peptide vaccines which include B and T cell epitopes that could potentially lead to improved vaccination against parvovirus B19 ^{35, 36, 37}. Many recombinants B19 parvovirus capsids of various structural protein compositions like VP1 & VP2 were evaluated as vaccine. Other studies used virus-like particles (VLPs) by mix VP1 and VP2 with specific ratio which gives antibody levels similar to those elicited by infection but it caused significant reactogenicity like headache, fever, gastrointestinal related distress and fatigue ^{37, 38}.

To our knowledge, there is no peptide prediction has been conducted specifically for B19V with a good immune response and less immunological side effect. In this study we determine a 100% conserved regions which are then investigated to predict the highly potential immunogenic epitopes for both B and T cells using VP1 capsid protein. VP1 has been proposed from different study as a good candidate protein for vaccine against B19V ^{2, 38}. Our results revealed that all the proposed epitopes for B cell were above threshold scores in Bepipred linear epitope, Emini surface accessibility and Kolaskar and Tongaonkar antigenicity using prediction methods in IEDB. One of the promising predicted epitopes was _214-PEVP_-217. This tetra peptide probably activating humeral immune response as it is part of VP1 capsid protein. ³⁹

Since the immune response of T cell is long lasting response comparing with B cell, where the antigen can easily escape the antibody memory response and considering that CD8⁺T and CD4⁺T cell responses play a major role in antiviral immunity ⁴⁰. Many peptides were proposed by this study that binds to MHC-I molecules (Table 3). All the proposed epitopes were interacted with MHC-I exceeding the thresholds. Moreover the epitope _155-FRYSQLAKL_-163 binds to six alleles with worldwide population coverage 56.26% to MHC-I alleles. This finding is not consistent with another study which proposed a single HLAB 35-restricted peptide (QPTRVDQKM) to stimulate immunity ⁴¹ however this study recommended further mapping down to the epitopes restricted to MHC-I ⁴¹.

For MHC-II (CD4⁺) three peptides were proposed (_438-FYVLEHSSF_-446, _404-WVYFPPQYA_-412 and _155-FRYSQLAKL_-163) that they bound successfully to many HLA alleles of MHC-II (Table 4). These peptides were above proposed thresholds as well as they have higher population coverage. One of the most promising and interested finding of this study was that the epitope _155-FRYSQLAKL_-163 was shared in both MHC-I & II as well as it has population coverage of 99.73%.

5. Conclusion

Parvo virus (B19V) has high prevalence among children and due to it is vertical transmission from mother to fetus which lead to fetal loss, developing of a vaccine with less immunological side effect remains a challenging issue. In this study we proposed many peptides that could be a powerful multi epitopes vaccine against B19V especially the peptide _155-FRYSQLAKL_-163 which interacted against both MHC-I & II. To determine the efficacy of these proposed epitopes clinical trials is required.

Acknowledgments

We are grateful to Dr:Ahmed Hamdi Abu-haraz, Africa city of Technology, Khartoum, Sudan for his excellent editing of manuscript. Thanks to our colleagues in University of Bahri for their help and support.

Competing Interest

The authors declare that they have no competing interest.

Funding

This study received no specific grant from any funding agency in the public, commercial, or private sectors.

References