Investigation of Leish-111f Epitopes to Design a New Multi-epitope Vaccine Against Leishmania Major Using Immunoinformatics Approaches

Since the incidence of various types of leishmaniasis, no denitive treatment has been considered for the disease, and due to its high prevalence worldwide, this issue has caused many concerns. Cutaneous leishmaniasis is the most common form of the disease which, can cause malignant lesions on the skin. Vaccination for the prevention and treatment of leishmaniasis can be the most effective way to combat this disease. In this study, we designed a new multi-epitope vaccine using immunoinformatics tools, which conrmed its effectiveness in the in silico. Methods Sequences Leish-111f protein (TSA, Leif, and LMSTI1) of Leishmania major (L. major) were downloaded from GenBank and with the help of immunoinformatic tools, was designed a new multi-epitope vaccine antigen of L. major. Result Th and Tc epitopes of the leish-111f protein were predicted using bioinformatics tools. The nal multi epitope was consisted of 18 CTL epitopes that joined by AAY linker. There are also 9 HTL epitopes in the structure of the nal vaccine that were joined by GPGPG linker. The prolin adjuvant was also added into the construct by AAY Linker. There were 613 residues in the structure of the nal construct. The multi epitope was stable and non-allergic. the data obtained from the binding of nal multi-epitope vaccine-TLR11 residues (band lengths and weighted scores) showed that the ligand and the receptor have a high anity to bind to each other. Moreover, in silico cloning approach, was improved the expression of proposed vaccine in E. coli host. Codon adaptation index and GC percent were calculated 1.0 and 53.35, respectively Based on these results, we hope that the multi-epitope vaccine, which contains the most appropriate epitopes of a strong Leishmania major immunogen, along with an adjuvant of TLR11, further stimulate the the L.major. in the in vitro. One of these antigens is Leish-111f L. major Thiol-specic Antioxidant (TSA), L. inducible (LMSTI1), L. brazilensis elongation initiation factor (Leif) (12). In ve studies, immunogenicity evaluation performed using a vaccine containing Leish-111f antigen with MPL-SE adjuvant and all them in clinical trials were successful (13).


Background
The worldwide outbreak of leishmaniasis, which has endangered 98 countries and a population of about 350 million people over the years, has led to many efforts to solve this health crisis (1). Unfortunately, despite many studies about leishmaniasis, there is still no de nitive treatment for this zoonotic disease in human (2,3). Drugs such as pentavalent antimony (Sb 5+ ), pentamidine, amphotericin B and miltefosine are used to combat against types of the disease, but due to high cost and side effects, they cannot be prescribed with con dence (4,5). In addition, many vaccines were successful in vitro, but their effectiveness has not been con rmed in later stages of research and few vaccines such as LEISH-111f and ChAd63KH DNA vaccine have entered into clinical trials (6,7).
Understanding the host defense mechanism against the parasite is essential in the production of an effective vaccine. Studies on leishmaniasis have shown that stimulation of cellular immunity is essential for defense against the parasite (8). Once the parasite enters the host, macrophages M1 or M2 play a key role in the parasite's defense or survival. If the T helper 1 (Th1) response is activated, the cytokines IL-12, IL-2, IL-17, IFNγ and TNF-α are increased. The presence of these cytokines, especially IFNγ, leads to the differentiation of M1 macrophages, the production of NO by iNOS and nally the elimination of the parasite. On the other hand, by directing the immune response to T helper 2 (Th2), cytokines such as IL-4, IL-13, and IL-10 are produced, which inhibit IFN-γ production and differentiate M2 macrophages. This route causes the infection to remain in the body due to the ine ciency of the immune system in parasitic defense (9,10). Also, in some studies, an increase in humoral immune response has been demonstrated by IgG2a antibody production to defense against leishmaniasis (11). Therefore, vaccination aimed at stimulating Th1, T cytotoxic (Tc), and humoral immune responses can be highly effective in defense against the parasite.
Several Leishmania antigens have been used to synthesize a variety of vaccines against the disease, and some of these vaccines have shown good immunogenicity in the in vitro. One of these antigens is Leish-111f containing L. major Thiol-speci c Antioxidant (TSA), L. Major stress inducible 1 (LMSTI1), and L. brazilensis elongation initiation factor (Leif) (12). In ve studies, immunogenicity evaluation was performed using a vaccine containing Leish-111f antigen with MPL-SE adjuvant and all of them in clinical trials were successful (13).
These immunogens can be used to make a variety of vaccines, including protein, DNA and mRNA vaccines. As mentioned, the Leish-111f protein has been shown to be effective and can stimulate the immune system well, so it is a good candidate for synthesizing other types of vaccines. On the other hand, due to the large size of the genome of this polyprotein, it is better to use selected epitopes of its proteins in the synthesis of the vaccine, so that if it becomes a candidate for the synthesis of the DNA vaccine against Leishmania, it can easily enter the AAV (adeno-associated virus) vectors (14).
Nowadays, with the advancement of immunoinformatics approach, vaccines can be designed in the in silico that their effectiveness is con rmed before being tested in the in vitro and in vivo. According to the prediction methods of immunoinformatics, the most suitable epitopes of each immunogens are selected for vaccine design. Thus, it can decrease the cost of vaccine synthesis along with increase of safety, specify, and effectiveness of vaccines (15). Also, using a multi-epitope instead of a complete protein vaccine can reduce allergic reactions and is a good replacement to a protein vaccine. Moreover, the infectious agents present in the synthesis of attenuated live vaccines are not present in the synthesis of these vaccines (16). Therefore, we can design and synthesis a new multi-epitope vaccines against the disease using this therapeutic tool (17).
In this study, for the rst time, we designed a multi-epitope vaccine using Leish-111f immunogenic antigen of L. major to combat against cutaneous leishmaniasis with the help of immunoinformatic tools. To increase the immunogenicity of the vaccine, we also used a pro lin adjuvant that is a TLR11-agonist and can be attached at the level of macrophages and monocytes. For this purpose, the most suitable HTL and CTL epitopes of Leish-111f, were predicted and linked them by AYY and GPGPG linkers. Also, pro lin with the AAY linker was added to the selected epitopes. The designed vaccine was tested in several steps containing B-cell epitope prediction, tertiary structure prediction, tertiary structure validation and re nement, molecular docking of nal vaccine and TLR11, molecular dynamics simulation and codon optimization.

Study design
To design a new multi-epitope vaccine, several steps were performed: Prediction of HTL and CTL epitopes; Binding the selected T cell epitopes by using suitable linkers, and designing of nal vaccine construct; B cell epitope prediction; docking analysis between the nal vaccine and TLR11; molecular dynamics simulation. Finally, in silico cloning in E. coli was performed. The schematic diagram of the designed vaccine has been shown in Fig. 1.

T-cell epitope prediction
To predict the CTL and HTL epitopes of TSA, LMSTI1, and Leif proteins, the IEDB server was used. Prediction of 9-mer and 15-mer epitopes to bind to MHCI and MHCII were done by using IEDB (http://tools.iedb.org/mhcii/). The server predicts the CTL and HTL epitopes on the basis of SMM (stabilized matrix method), Arti cial Neural Networks (ANN) and, Allele speci c a nity cutoff (IC50) (18). In this study, we selected the mouse alleles to bind the MHCI and MHCII.

IFN-γ inducing epitope prediction
Using IFN epitope server (http://crdd.osdd.net/raghava/ifnepitope/), prediction of interferon-gamma inducing epitopes of HTL epitopes was performed. The server predicts the epitopes based on the methods such as SVM based, motif based, and hybrid approach. On this server, IFN-γ inducing and non-inducing MHC-II binders were determined.

Construction of nal multi-epitope vaccine
Using the in silico information, the nal multi-epitope vaccine was designed. First, AYY and GPGPS exible linkers were used to joined HTL and CTL epitopes, respectively (19). Then, to enhance the immunogenicity of nal multi-epitope vaccine, pro lin was added as an adjuvant to the N-terminal of the construct by using AYY linker.
Conformational and Linear B cell epitope prediction According to the 3-D structure of the construct, the ElliPro suite (http://tools.iedb.org/ellipro/) was used to predict of conformational and linear B cell epitopes of the nal vaccine construct (20).
The server assigns a score to each predicted epitope, named PI (Protrusion index). First, the three-dimensional shape of the protein is estimated by a number of spherical ellipses, which the higher the percentage, the fewer residues outside the ellipsoid. The PI value is de ned based on the center of mass remaining outside the largest possible ellipsoid and residues with higher scores have more access to the solvent. Discontinuous epitopes are characterized by PI values and clustered by R distance (in Å between remaining mass centers). The higher the R value, the larger the epitopes.
Tertiary structure prediction of nal multi-epitope vaccine I-TASSER Server (Iterative Threading ASSEmbly Re nement) available at (http://zhanglab.ccmb.med.umich.edu/ I-TASSER/) was used to predict of tertiary structure of the nal vaccine construct. I-TASSER predicts the best model with the highest C-score for further analysis. This server provides a PDB le that uses the multiple threading approach LMOTS and predicts the 3D structure using iterative templatebased pattern assembly simulations of full-length atomic models. Moreover, the server was the best in terms of function prediction in CASP9 (21).
Tertiary structure re nement of nal multi-epitope vaccine To re ne of 3D model of the nal vaccine construct, GalaxyRe ne server available at (http://galaxy.seoklab.org/) was applied (22). Mild and aggressive relaxation methods are used to reconstruct the side chain residues, repackaging and molecular dynamics simulation of proteins. The server can also enhance the local and global quality of the produced models by I-TASSER (22).
Validation of tertiary structure of nal multi-epitope vaccine PROCHECK (http://servicesn.mbi.ucla.edu/PROCHECK/), ProSA (http://prosa.services.came.sbg.ac.at/prosa.php), and ERRAT (http://services.mbi.ucla.edu/ERRAT) servers were used to validate the 3D structure of the nal vaccine construct. The PROCHECK gave a Ramachandran plot and assessed the stereochemical quality of the re ned model (23). Furthermore, the ProSA provided an overall quality score based on Z-score value. It should be noted that scores outside the range of native proteins are not acceptable (24).
Evaluation of physico-chemical properties of the nal vaccine construct ProtParam (http://web.expasy.org/protparam/), an online web server was used to calculate the physicochemical characteristics of nal multi-epitope vaccine containing aliphatic index, instability index, theoretical PI (isoelectric point) molecular weight, in vitro and in vivo halflife, and grand average of hydropathicity (GRAVY) (25). Then, the solubility of the nal vaccine construct over expression in E. coli was calculated by using SOLpro (http://scratch.proteomics.ics.uci.edu/) (26).
Allergenicity and antigenicity assessment of the nal vaccine construct AllerTOP v2.0 and AllergenFP1.0 were applied to determine the antigenicity of the nal vaccine construct. Using ve E-descriptors, including amino acid hydrophobicity, size, amino acid tendency to helix, amino acid abundance, and tendency of β-strand formation, the amino acids in the protein sequence are described in the data set of AllergenFP1.0 server. Also, based on auto-cross covariance (ACC) transformation, the strings were transformed into uniform vectors. The server categories the proteins in two allergen and non-allergen, according to Tanimoto coe cient (27). AllerTOP v2.0 also used the same ve E-descriptors that mention above. This server classi es the protein to allergen and non-allergen based on k-nearest neighbor algorithm (kNN, k=1) (28). Thereafter, the determination of antigenicity was done by using ANTIGENpro (http://scratch.proteomics.ics.uci.edu/) and VaxiJen v.2 (http://www.ddgpharmfac.net/) servers. ANTIGENpro is an alignment-free and sequence-based server. The accuracy of this pathogen independent server is 82% (29). Protective antigens were also predicted by VaxiJen, a rst alignment-independent prediction (30) .

Docking analysis of nal vaccine and TLR-11
ClusPro server available at (http://cluspro.bu.edu/login.php) was applied to determine the a nity of binding between ligand ( nal multiepitope vaccine) and receptor (TLR-11). Elimination of unstructured protein regions, use of attraction or repulsion, calculation of pairwise distance restraints, and homo-multimers construction, are some of the parameters that this server considers to provide docking results of two proteins. In addition, the selection of highly populated cluster centers of low-energy structures is another feature of this server (31).
Finally, visualization of docked complex was performed by PyMOL program.

Molecular dynamics simulation of nal vaccine and
The structural properties and interaction of ligand (the nal vaccine construct) and receptor (TLR-11) was evaluated in the in vivo conditions through molecular dynamics simulation. In this step, Using GROMOS 54a7 and 43a1 package, docked complex of nal vaccine and TLR-11 was simulated during 50000 picoseconds (Ps) at 310˚ K and 1 bar pressure. A Cubic box with periodic boundary conditions with water molecules of SPC/E was also used. Moreover, to minimize of system energy and relaxation of solvent molecule the steepestdescent algorithm was applied. Also, xation of the atom was performed by using A LINear Constraint Solver (LINCS) algorithm and SETTLE, an analytical algorithm, was used for solvent molecules. To calculate of total electrostatic energy and other non-bonded interactions, Particle Mesh Ewald (PME) summation method and L-J model (with 10 A˚ cutoff distances), were used, respectively. Moreover, the temperature and pressure of each system, at the time of coupling equal to 0.1 Ps were preserved using Berendsen weak coupling algorithm. Finally, the root mean square deviation (RMSD) and root mean square uctuation (FMSF) diagrams were plotted and the uctuations and stability of the docked complex were investigated.
In silico cloning of nal multi-epitope vaccine For this purpose, SnapGene tool was applied and cDNA was provided as an output of the server. The cDNA sequence was used to determine of CAI value and GC content. The CAI value higher than 0.8 and GC content between 30-70% are acceptable. Then, XhoI and NdeI restriction enzymes were added to the cDNA sequence. Finally, codon optimization in to pET28a (+) vector was done.

T-cell epitope prediction
Using IEDB server, CTL epitopes of leish111-f consisting three proteins was predicted. Six epitopes were predicted from each protein of leish111f. Therefore, a total of 18 epitopes with the lowest percentile rank was selected. Also, a total of 9 HTL epitopes of all proteins was selected that for each protein three epitopes were selected (  Construction of nal multi-epitope vaccine A total of 9 CTL epitopes was linked together by AAY linker. Also, GPGPG linker was utilized to join the HTL epitopes with them. Then, pro lin with the length of 163 amino acid was added at the N terminal of the construct. The construct was shown in Fig. 1. B cell epitope prediction of nal multi-epitope vaccine Prediction of linear and discontinuous epitopes of the nal vaccine construct was performed using ElliPro suite. According to the results obtained from the server, 12 linear and 7 discontinuous B cell epitopes were predicted in nal vaccine that can stimulate humoral immunity ( Fig. 2 and table 4, 5).  Prediction, re nement and validation of three dimensional structure of nal multi-epitope vaccine I-TASSER results showed ve predicted models for the 3D structure of the vaccine (3necA, 3tkpA, 3nec, 6tdaL, 3nec). Among them, the model 1(PDB Hit: 3nec) was recognized as the best model with the C-score of -2.23 (Fig. 3A). Then, the tertiary structure of the predicted model of the nal vaccine construct, was also re ned by using Galaxy Re ne server and the scores of model quality was calculated. Based on the GDT-HA ( 0.9545 ), RMSD ( 0.435 ), MolProbity (1.825), Clash score (13.1), Poor rotamers (0.7) and Rama favored (96.0),, model 1 was selected as the best model for the nal vaccine construct. Then, Ramachandran plot, ProSA and ERRAT tools were used to validate of the selected re ne model. For this purpose, veri cation of quality and potential errors of the crude 3D model was done. ERRAT tool showed the value of 86.42% as an overall quality factor of re ned model. The ProSA was calculated a Z-score of -4.55, that is a plausible value of native proteins as shown in Fig. 3B. In addition, the data obtained from the Ramachandran diagram showed 87% residues located in most favored regions, 9.9% residues in additional allowed regions, and 1.5% residues in allowed and 1.5 % residues in disallowed regions as shown in Fig. 3  nal vaccine were also indicated two hydrogen bands at 2.7 Å with Gly 922 (HH1) and Gly 924 (OH) from TLR11, respectively. (Fig. 4, Table   6).  In silico codon optimization of nal multi-epitope vaccine Based on the results of the Java Codon Adaptation tool (JCat), reverse translation and codon optimization in E.coli (stain K12) was carried out. After the design of cDNA, codon adaptation index and GC percent were calculated 1.0 and 53.35, respectively. The value of GC content was in acceptable rang of the results that is 30-70%. Also the CAI value showed the vaccine is capable of the highest expression in the pET28a (+) vector (Fig. 6).

Discussion
There is no de nitive treatment for leishmaniasis, and the cutaneous form of the disease, which affects 1.5 -2 million people annually, can cause several lesions on the patient's skin if it progresses (32,33). Researchers are still trying to make a vaccine for prevention and treatment of the disease (34). Identifying the host defense pathways against the pathogen and then strengthening these pathways can be the most important strategy in making an effective vaccine. To achieve this goal, immunoinformatics tools can be used to select the most appropriate epitope of the desired immunogens in less time with high accuracy (35), (36).
Leish-111f is an immunogen which a protein-based vaccine has already been approved and can be used as a candidate for another type of vaccine, such as a multi-epitope vaccine, which this type of vaccine is currently being considered in the treatment of various diseases including SARS-COV-2 (37).
The aim of this study was to design a new multi-epitope vaccine to be effective in combating leishmaniasis. In the rst step, HTL and CTL epitopes of Leish111-f polyprotein from L. major were predicted and ligated using AYY and GPGPG linkers. Also, due to the fact that in the vaccination against leishmaniasis, the immune system needs to be more activated, we used the pro lin as an adjuvant. The Pro lin can bind to TLR11 at the macrophage and dendritic surface, and leading to innate immune stimulation, increased IFN-γ, and consequently increased Th1. Therefore, pro lin can help increase the immunogenicity of this vaccine against Leishmania, which was our main goal. (38).
There were also 12 epitopes in the nal vaccine construct to induce interferon-gamma, which due to the role of interferon-gamma in inducing the immune system response against Leishmania, these epitopes can have a great effect on the defense against the parasite.
After the design of the nal vaccine construct, conformational and linear B cell epitopes were predicted. The number of twenty linear and conformational epitopes in the structure of the nal construct showed that this construct can stimulate the humoral immunity of the host well. Moreover, tertiary structure prediction of nal vaccine construct along with validation and re nement was done. Based on the results of Ramachandran diagram and ProSA, the nal vaccine construct was stable in nature. Also the quality and potential errors were predicted using a re nement of nal vaccine.
Thereafter, on the basis of physicochemical properties of nal vaccine, the values of PI and GRAVY were 5.17 and -0.157, respectively that shows the construct is acidic and hydrophilic in nature. Also, thermostability of nal vaccine was con rmed with the values of 5 and 10 for aliphatic index and instability index, respectively. Then, the solubility of the nal multi-epitope vaccine in E. coli was done.
In addition, the data obtained from the binding of nal multi-epitope vaccine-TLR11 residues (band lengths and weighted scores) showed that the ligand and the receptor have a high a nity to bind to each other. Also, the graphs of RMSD and RMSF indicated the docked complex was stable and has little variation. Therefore, the nal multi-epitope vaccine can have a stable binding to TLR11 in the body through pro lin.
In many vaccine design studies, docking and dynamic tests are performed to ensure the effectiveness of the construct. Because by con rming the a nity of binding between ligand-receptor by docking analysis in a vacuum condition and also ensuring the stability of this interaction in conditions similar to the body environment, the effectiveness of the vaccine in the in silico space is con rmed. Finally, the results of codon optimization (CAI value and GC content) showed that the multi-epitope vaccine could have the highest expression in the host.

Conclusion
In this study, immunoinformatic tools helped us design a new multi-epitope vaccine of Leish-111f of L. major. After identifying the most suitable epitopes of cellular immune stimulation of the host and adding an adjuvant to bind to TLR11 to increase the safety, the construct was designed. The B cell results showed that there are epitopes of humoral immune stimulation in the nal construct and since humoral immunity can also be effective in defense against leishmaniasis, this vaccine can be a good candidate. Other tests con rmed the effectiveness of the vaccine. Moreover, docking analysis of nal multi-epitope vaccine and TLR11 showed good interactions between them and molecular dynamics simulation of multi-epitope vaccine-TLR11 was also con rmed the stability of the docked complexes. Based on these results, we hope that the multi-epitope vaccine, which contains the most appropriate epitopes of a strong Leishmania major immunogen, along with an adjuvant capable of binding to TLR11, will further stimulate the immune system against the parasite. However, due to the effectiveness of the vaccine in the in silico, it is necessary to con rm the effectiveness of this vaccine in the later stages of the in vitro and in vivo. Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable. Figure 1 Schematic diagram of nal multi-epitope vaccine. The construct consists of 18 CTL epitopes with gray color that joined by AAY linker. There are also 9 HTL epitopes in the structure of the nal vaccine, which are shown in black color and were joined by GPGPG linker. The pro lin adjuvant was also added into the construct by AAY Linker. There are 613 residues in the structure of the nal construct.     In silico cloning of nal multi-epitope vaccine. The purple line indicates where the vaccine is attached to the vector and the black line indicates the vector.