Search tips
Search criteria 


Logo of ijmsMDPIhomeThis articleThis journalInstructions for authorsSubscribeIJMS
Int J Mol Sci. 2017 February; 18(2): 371.
Published online 2017 February 10. doi:  10.3390/ijms18020371
PMCID: PMC5343906

Immunoinformatics Features Linked to Leishmania Vaccine Development: Data Integration of Experimental and In Silico Studies

Christopher Woelk, Academic Editor


Leishmaniasis is a wide-spectrum disease caused by parasites from Leishmania genus. There is no human vaccine available and it is considered by many studies as apotential effective tool for disease control. To discover novel antigens, computational programs have been used in reverse vaccinology strategies. In this work, we developed a validation antigen approach that integrates prediction of B and T cell epitopes, analysis of Protein-Protein Interaction (PPI) networks and metabolic pathways. We selected twenty candidate proteins from Leishmania tested in murine model, with experimental outcome published in the literature. The predictions for CD4+ and CD8+ T cell epitopes were correlated with protection in experimental outcomes. We also mapped immunogenic proteins on PPI networks in order to find Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways associated with them. Our results suggest that non-protective antigens have lowest frequency of predicted T CD4+ and T CD8+ epitopes, compared with protective ones. T CD4+ and T CD8+ cells are more related to leishmaniasis protection in experimental outcomes than B cell predicted epitopes. Considering KEGG analysis, the proteins considered protective are connected to nodes with few pathways, including those associated with ribosome biosynthesis and purine metabolism.

Keywords: immunoinformatics, epitope prediction, pathways, protein–protein interaction networks, reverse vaccinology, leishmaniasis

1. Introduction

Leishmaniasis is a wide-spectrum disease caused by parasites from Leishmania genus. It is prevalent in Americas, Europe, Africa and Asia. Overall, human infection is caused by at least 20 species whose vectors are phlebotomine sandflies [1]. Although being considered by many studies one of the best possible alternatives for this disease control, there is no human vaccine available [2].

In the advent of reverse vaccinology, in the latest years, a great effort has been made by bioinformaticians in order to provide epitopes predictors programs. Currently, it is possible to scan entire genomes searching for immunogenic epitopes and then select promising proteins for vaccine development. The bottleneck in this workflow analysis is the validation of predictions for protozoan parasites. Many predictors are available for B cells, T CD4+ cells and T CD8+ cells epitopes and subcellular localization. They are valuable in a pre-screening evaluation for vaccine targets and searching for diagnostic markers.

The building of protein-protein interaction (PPI) networks may give some insights to understand the biological role of these targets, and so might be a valuable asset in vaccine development. These networks are constituted by nodes that correspond to proteins, connected by edges, representing the interactions between two connected proteins. With PPI networks, we can have an overview of protein relationships and notice those with high connections (also referred as “hubs”). Hub proteins tend to have essential role in the parasite metabolism and might be good candidates to vaccinal and drug target [3,4].

To support Leishmania vaccine research, we developed an approach that integrates prediction of B and T cell epitopes, analysis of PPI networks and metabolic pathways. With the aim of validating this methodology, we selected Leishmania proteins tested as vaccine candidates in murine model, with experimental outcome (EO) published in the literature. After predicting epitopes in the selected proteins using specific computational programs, we correlated the predictions for T CD4+ and T CD8+ cells with protection in EO. Finally, we mapped the immunogenic proteins on PPI networks in order to find Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways associated with them.

2. Results

2.1. Leishmania Proteins Dataset Selection

Through the use of text mining technics from Pubmed website that included, but was not restricted to, categorization and entity extraction, we were able to identify and select 20 proteins from six different Leishmania species that were used in studies aiming the vaccine development against these parasites.

It is important to highlight that, for each one of those proteins, a specific MySQL ID was assigned to link GI accession number and TriTrypDB specific ID. Based on the results published, the EO was categorized into: (a) “no protection” (nine proteins); (b) “partial protection” (five proteins); and (c) “protection” (six proteins). The accession numbers of these proteins are depicted in Table 1.

Table 1
Selected candidate antigens from dermatotropic and visceratropic Leishmania species to leishmaniasis vaccine development. Proteins in literature tested in mice model were selected randomly.

2.2. Epitope and Subcellular Localization Predictions

With the purpose of selecting potential immunogenic epitopes in the selected experimental dataset, Structured Query Language (SQL) statements were used. The results obtained in terms of number of predicted binding Major Histocompatibility Complex (MHC) class I and II epitopes, B cell epitopes and subcellular locations are detailed in Table 2. Interestingly, the majority of the proteins within the “protection” group were predicted as extracellular and the proteins belonging to “no protection” group were predicted as located in nuclear and cytoplasmic compartments.

Table 2
Number of binding Major Histocompatibility Complex (MHC) epitopes, B cell epitopes and subcellular localization predicted by different computational programs.

Specifically regarding the epitopes capacity to bind MHC class I, MHC class II and epitopes for B cell activation, considering a path from “no protection” to “protection” groups, a gradual increase of the number of predicted epitopes for T cells and B cells was observed.

2.3. Predicted Epitopes and Experimental Outcome Correlation

To evaluate the possible association between the number of predicted epitopes (NPE) for B and T cells, and the EO of selected proteins, the following consensus predictions were produced: (a) prediction for T CD8+ epitopes obtained from NetMHC and NetCTL; and (b) prediction for B cell epitopes obtained from AAP12, BCPred12 and BepiPred. The consensus predictions were obtained overlapping identical predictions made by different methodologies.

To graphically depict the results, Box Plots and Correspondence Maps (CM) approaches were applied to visualize the potential associations determined through Spearman r and Chi-square distance, respectively. Firstly, the disease (leishmaniasis) was stratified into cutaneous leishmaniasis (CL) and visceral leishmaniasis (VL) and the EO of antigens from Leishmania that cause CL and VL were correlated with NPE (Figure 1a). Regarding VL analyses, significant correlation was observed only with EO and predicted epitopes for CD8+ T cells (p < 0.05) (data not shown). On the other hand, for CL analyses, it was observed significant correlation between EO and NPE for T CD4+, T CD8+ and B cells, as shown in Figure 1a. After that, analyses were performed concerning the disease without any stratification. As can be observed from Figure 1b (NPE and EO correlation for T CD4+, T CD8+ and B cells), a significant correlation exists between NPE specific to CD4+ and CD8+ T cells with r = 0.752/p < 0.05 and r = 0.793/p < 0.05. In addition, a weak association with B cell predicted epitopes (r = 0.515/p < 0.05) was observed. In other words, non-protective antigens have lowest frequency of predicted T CD4+ and T CD8+ epitopes, compared with protective ones.In regards to CM analysis (Figure 1c), considering the adopted variables (antigens EO versus NPE for T and B cells), the grouping outcome, which is related with data correlation, shows the same strong association above mentioned for leishmaniasis with no stratification.

Figure 1
Correlation analysis: (a) Box plots of the relationships between T CD4+, T CD8+ and B cell epitopes and experimental outcome of candidate antigens taking into account cutaneous leishmaniasis (CL); (b) box plots of the relationships between T CD4+, T CD8 ...

As the last analysis layer used to validate data correlation, the Chi-square results confirmed the significant association between EO and the predicted epitopes for T CD4+ and T CD8+, p < 0.05 (Tables S1 and S2) and the weak one between EO and predicted epitopes for B cells (p < 0.05, see Table S3).

2.4. Number of Alleles (NA) and Experimental Outcome (EO) Correlation

To hypothesize possible reasons linked with vaccine success or failure, an evaluation of allele-specific affinity (MHC I and II) was investigated. As illustrated in Figure 2a, the amounts of epitopes binding MHC haplotype d (BALB/c MHC alleles) and haplotype b binders (C57BL/6 MHC alleles) identified in the “protection” group were superior to those ones identified in the “no protection” group. In summary, our results indicate that epitopes for MHC class I and II haplotype b and d are more frequent in the success antigens tested for vaccine development. In addition, a detailed analysis in which MHC class I and II haplotypes were individually investigated revealed a strong association between NPE from MHC class I haplotype d and EO (p < 0.05 and r = 0.855) that is not observed for MHC class II haplotype d and b.

Figure 2
Evaluation of epitopes amount and relationship with experimental outcome: (a) bar graph showing number of epitopes for MHC class I and II (haplotype d and b) in the selected antigens classified in “protection” and “no protection” ...

2.5. Mapping Immunogenic Proteins on Protein-Protein Interaction Networks (PPI Networks)

We chose Cytoscape to model PPI networks using data from STRING v.10. Figure 3a presents these networks, annotated according to their related KEGG pathways. Analyses of enriched pathways are shown in Figure 3b, according to their False Discovery Rate. Considering the “no protection” group, the most common pathways found were: ribosome (58 genes involved), glutathione metabolism (17 genes), RNA degradation (15 genes), protein processing in endoplasmic reticulum (15 genes), peroxisome (sixgenes) and homologous recombination (sevengenes). On the other hand, for “protection” group, we identified pathways as ribosome (195 genes involved), metabolic pathways (12 genes), purine metabolism (sixgenes) and protein processing in endoplasmic reticulum (seven genes). Interestingly, target proteins from “no protection” group are connected with nodes from many different pathways. In contrast, we observed that proteins of the “protection” group are connected to nodes of few pathways. Thus, there is a strong negative correlation (r = −8.55) between the number of connected pathways and EO of the selected proteins (Figure 3c).

Figure 3
Immunogenic proteins mapped in Protein-Protein Interaction (PPI) networks: (a) PPI networks constructed starting with the target proteins (“no protection” and “partial protection/protection”) represented by triangles and ...

3. Discussion

Nowadays, many computational methodologies have been described for epitope predictions of bacterial, fungal and others microorganisms. For protozoan (specifically Leishmania species), there are not strong and validated platforms to identify promising antigens for Leishmania vaccines [23]. Herein, we developed a sturdy and complete platform with potential of identifying candidates for vaccines against leishmaniasis. This platform integrates prediction of B and T cell epitopes, analysis of PPI networks and signaling pathways.

The first step was to select the input antigens to validate the platform. In this concern, we did an extensive search in the literature regarding antigens tested in murine model. It was difficult to categorize the 20 selected antigens for this work because there is a lack of consensus concerning the model, challenge inoculum, mechanisms of protective immunity (response induction, parasite burden reduction), so there is no standardization to appoint if a vaccine indicates protection or not [24]. We tried to choose proteins with no interference of adjuvant since it can entirely modify the antigen response [25].

Regarding the platform validation, we performed the epitope mapping using ad-hoc algorithms. Our results suggest that antigens with more predicted epitopes for T CD4+ and T CD8+ cells could be associated with protection in EO. In this regard, our results revealed that there are strong correlation and association between predicted epitopes and the EO. The inertia values show the powerful association between the variables. The CM dimension 1 represents the highest inertia allowing the interpretation of the results in the first dimension. T CD4+ and T CD8+ predicted epitopes versus EO show higher inertia values when compared to B cells predicted epitopes emphasizing that epitopes for CD4+ and CD8+ T cells are crucial for Leishmania vaccines success. We used all available human and mouse alleles to restrict the epitopes allowing enhance the assertive prediction. This analysis is useful to identify conserved epitopes that can bind various alleles of MHC appointing for rare and promising epitopes. The in silico analyses and in vivo validation of epitopes demonstrates that some algorithms may be important tools for the identification of epitopes, and consequently of immunogenic proteins. The algorithm NetCTL version 1.2 makes prediction of peptide–MHC class I binding, proteasomal C terminal cleavage, both using artificial neural networks, and TAP transport efficiency using weight matrix. The tree predictions are then integrated [26]. Another predictor also used for MHC class I binding peptides was NetMHC version 3.0. It predicts binding of peptides to different HLA alleles using artificial neural networks and weight matrices. For peptide–MHC class II binding prediction NetMHCII, version 1.0, was used. It predicts binding of peptides to 14 different HLA-DR alleles, including human and mouse, using position specific weight matrices [27]. To perform B-cell epitopes predictions, we used only methods that predict continuous epitopes. We used first BepiPred, version 1.0, that predicts linear B-cell epitopes using a combination of Hidden Markov model and a propensity scale method [28]. Then we used BCPREDS server comprising the AAP12 and BCPred12 predictors. The first one is based on the finding that B-cell epitopes favor particular amino acid pair, and it was trained using support vector machine classifier. The second uses subsequence kernel trained using support vector machine classifiers with 701 linear B-cell epitopes, extracted from Bcipep database, and 701 non-epitopes, randomly extracted from SwissProt sequences [29,30,31]. Finally, we used WoLF PSORT predictor, which is an amino acid sequence predictor of subcellular localization sites of proteins. It uses known sorting signal motifs and some correlative sequence features [32]. The integration of these predictors could reveal proteins that are secreted or presented in parasite membrane, capable of eliciting B and T cells responses.

Herrera-Najera et al. [33] performed a large-scale prediction of T cell epitopes in the whole genome of L. major, obtaining 26 potential epitopes through prediction consensus. Fourteen of them revealed to be immunogenic epitopes that were capable to stimulate T cells to produce IFN-γ. Other studies employing computational predictions in specific Leishmania proteins have shown that it is quite possible to use combined algorithms in epitopes searching that could be validated by in vivo experiments [34]. Duarte et al. [35] developed a combined epitope prediction platform in order to investigate T CD8+ epitopes in 63 L. braziliensis proteins, demonstrating a cytotoxic activity of some predicted epitopes in Leishmania infected mice. Recently, in silico methods for linear epitope predictions (NetMHC, NetCTL, and NetMHCII) were combined with molecular modeling to identify potential epitopes in the whole L. braziliensis proteome. Therefore, the pipeline was validated based on stimulation of human peripheral blood mononuclear cells (PBMCs) proliferation. The results obtained after the in vitro assays showed that six of ten selected epitopes could be classified as potentially immunogenic [36].

The role of B cell epitopes still hasroom for discussion since there is no consensus if immunoglobulins could be associated with resistance or susceptibility in leishmaniasis [37,38]. On the other hand, our results demonstrate a correlation between protection and specific B cell epitopes. In this context, we suggest that an effective vaccine should have epitopes capable of eliciting a strong T cell response and B cells too. In addition to the epitopes for B cells and MHC class I and II, another important feature is the subcellular localization of the antigen. It is known that extracellular Leishmania proteins are more immunogenic and considered better targets for vaccine development [39,40]. This fact is indeed corroborated by our findings that show the majority of the antigens from “protection group” are linked with extracellular compartmentalization. In silico approaches have limitations regarding the proteome annotation (e.g., the data of L. amazonensis used in this work) and the large number of linear epitopes. Nevertheless, our results of epitope prediction indicate a higher assertive and successful prediction, so it can be a useful approach for vaccine development against leishmaniasis.

To better understand the biological importance of vaccine candidates, we proposed the use of PPI networks enriched with KEGG pathways information. It is well known that some proteins are essential for specific biological processes of Leishmania spp. [41,42]. In this context, we proposed to analyze antigen pathways through modeled PPI and its relation with protection and no protection of vaccine candidates after challenging with infective Leishmania parasites. Our analyses showed that many of the selected antigens do not have any KEGG pathway associated to them, but, instead, are connected to proteins that are part of some pathway. Pathways associated with ribosome biosynthesis, purine metabolism and metabolic processes are present in “protection” group networks. Ribosome related proteins were considered relevant molecules during infection, since in some circumstances they can modulate cell activities and cytokine release. Many works associated these pathways to immune response [43]. Cordeiro-Da-Silva and collaborators in 2001 characterized a Leishmania major gene considered to be homologous to the mammalian ribosomal protein S3a. This ribosomal protein can be found in many other Leishmania species such as L. infantum, L. amazonensis, and L. mexicana. The article authors suggested that this protein could participate in the Th1/Th2 immune response balance during leishmaniasis [44]. Soto and collaborators in 1993 using sera from dogs affected by visceral leishmaniasis identified high antigenic Leishmania acidic ribosomal proteins, also called P-type proteins [45]. Another work by Soto and collaborators in 2000 showed that intraperitoneal administration in BALB/c mice of the acidic ribosomal protein LiP2a, without adjuvants, elicited a strong humoral response and was capable of stimulating production of IFN-γ in cultured splenocytes from LiP2a-immunized mice [46]. Our findings also match with results obtained in the secretome of L. donovani, where the majority of virulent proteins (secreted proteins) belong mainly to metabolic and biosynthesis processes [47]. To check the importance of some proteins related to metabolic process, Leishmania knocked-out for protein kinases and phosphatases possible involved in parasite metabolism regulation were generated. After this process, in many cases highly attenuated or completely avirulent parasites could be observed [48]. Naderer and collaborators generated a Leishmania major mutant lacking the regulatory subunit of the Ca2+/calmodulin-dependent serine/threonine-specific phosphatase. This modified Leishmania grew normally at 27 °C. However, this parasite lost viability when exposed to 34 °C [49]. Target of rapamycin (TOR) kinases are involved in some regulatory pathways related to cell growth and structure in eukaryotes. Silva and collaborators generated TOR3 knocked-out Leishmania major parasites. These knocked-out parasites exhibited slower growth than wild-type parasites and were unable to survive or replicate in macrophages in vitro. These parasites were not capable of inducing disease or establish infection in mice in vivo [50]. In addition, McConville and Naderer [48] have shown that metabolic pathways are important to Leishmania virulence, since down regulation of metabolic genes causes latency of many Leishmania species. One possible application of these attenuated or avirulent parasites could be in whole parasites vaccines. Carter et al. [51] have noticed that purine metabolism is vital for Leishmania survival. Surprisingly, our findings show that proteins associated to protection are connected to few pathways when compared to proteins that are not protective.

In summary, in this work, we proposed and validated a computational approach regarding epitope prediction, topological structure and pathway analyses to drive a rational vaccine design against leishmaniasis.

4. Materials and Methods

4.1. Selection of the Leishmania Antigens

Leishmania proteins tested in murine model, with EO published in the literature, were selected. Studies describing vaccine effectiveness after challenge with Leishmania spp. were preferentially chosen. Bearing in mind that there is no standardization of the protection concept and that there is variation of results in the literature, it was necessary to create three categories as already described. According to our categorization, “no protection” group includes antigens that promote no or slight reduction of parasite burden or lesions after Leishmania challenge. The “protection” group includes antigens that promoted significantly strong reduction of the parasite burden or lesions showing strong immune response to Leishmania antigens. Thus, the term partial protection was used to classify antigens that are between a potential protection and no protection at all. The “partial protection” group comprises proteins that could slightly reduce the parasite burden and/or lesions more than “no protection” group. In addition, these proteins can elicit some immune response which results in ineffective protection. It is important to highlight that the influence of adjuvants was not taken into account for the categorization, thus only the response of antigens tested alone was chosen to categorize the groups. Twenty candidates from Old and New World Leishmania species were categorized according with their experimental results in the following groups: (a) “no protection”; (b) “partial protection”; and (c) “protection”. The selected experimentally validated data included information from proteins of L. amazonensis, L. braziliensis, L. major, L. mexicana, L. donovani and L. infantum that were subsequently used in this study to corroborate the in silico bioinformatics predictions. The selected antigens are described in Table 1.

4.2. Leishmania Proteome Data

The predicted proteome sequences of dermatotropic and viscerotropic Leishmania species were obtained from TriTrypDB (Kinetoplastid Genomics Resource) and the sequence of L. amazonensis was downloaded from [52]. Detailed information about the predicted proteomes versions used in this work can be found in Table 3.

Table 3
Leishmania predicted proteomes used in the study. The version and number of predicted proteins of each species are shown.

4.3. Epitope and Subcellular Localization Predictions

All proteomic data used in this work were screened in order to predict T CD4+ and T CD8+ epitopes, B cell epitopes and subcellular localization of proteins.

For T CD8+ epitope prediction (MHC class I binding epitopes), algorithms NetCTL [26,53,54] and NetMHC [55,56,57] were used. Regarding T CD4+ epitopes, NetMHCII [27,58] was used to predict MHC class II binding epitopes. For B cell epitopes, BepiPred [28] and BCPREDS (AAP12 and BCPred12 models) [29,30,31] were used to predict epitopes. Finally, the protein subcellular localization was predicted using WoLF PSORT [32,55,56,57]. The algorithms choice was made taking into account their viability for local stand-alone server installation, the number of citations found in literature, and the results previously published by Resende, Rezende, Oliveira, Batista, Correa-Oliveira, Reis and Ruiz [23] describing algorithms specificity, sensitivity and accuracy with parasite data obtained from UniProt ( and IEDB (Immune Epitope Database and Analysis Resourse) ( The analytical workflow used in this study is presented in Figure 4.

Figure 4
Workflow of analysis showing the steps followed along this work.

To carry out the predictions, the algorithms were parameterized for eukaryotic genomes. For MHC class I binding epitopes, predictions for 12 human supertypes and seven mice alleles were performed. The following alleles were used: A1, A2, A3, A24, A26, B7, B8, B27, B39, B44, B58, B62, H2-Db, Dk-H2, H2-Dd, H2-Kb, H2-Kd, Kk-H2, and H2-Ld. Concerning MHC class II binding epitopes, we used 14 human alleles and three mice alleles bringing the total of different alleles to 17, as follows: HLA-DRB1*01:01, HLA-DRB1*03:01, HLA-DRB1*04:01, HLA-DRB1*04:04, HLA-DRB1*04:05, HLA-DRB1*07:01, HLA-DRB1*08:02, HLA-DRB1*09:01, HLA-DRB1*11:01, HLA-DRB1*13:02, HLA-DRB1*15:01, HLA-DRB3*01:01, HLA-DRB4*01:01, HLA-DRB5*01:01, H2-IAs, H2-IAd and H2-IAb.

4.4. Development of Relational Database

Taking into account the great amount of data generated by the algorithms, we constructed a relational database using MySQL as Relational Database Management System (RDBMS) ( The use of a database system in this work represents a crucial step that allows the integration of the results from all predictors and a way of getting a data receptacle or conceptual repository from which is possible to extract data correlation, helping in the identification of target proteins. The MySQL GUI Tools ( were used as a graphical user interface for our MySQL database. The relational model was built in MySQL Workbench ( To extract, parse and load data into database, specific Perl scripts were developed using DBI (the Perl interface to databases) and BioPerl modules. The relational schema is presented in Figure 5.

Figure 5
MySQL relational database scheme developed to integrate data from predictions.

4.5. Mapping Immunogenic Proteins on Protein-Protein Interaction Networks (PPI Networks)

In this topic, two groups were created: P (protection), aggregating proteins classified as “partial protection” and “protection”, and NP (no protection) for proteins classified as “no protection”. We entered these proteins into STRING v.10 [59] looking for interaction networks associated with them, using Leishmania as reference organism. The active interaction sources used were text mining, experiments, database, co-expression, neighborhood, gene fusion and co-occurrence, with medium confidence score (0.400). The networks were built using the target proteins as central nodes and expanding them with the first neighbors, using 500 as maximum value of interactions. From STRING, we obtained the PPI network data and KEGG [60] pathway functional enrichments (p < 0.05) of proteins involved in these networks. Then we built the protein networks using Cytoscape [61], adding KEGG information over them.

4.6. Statistical Analysis

The analyses were performed using SPSS version 20 (SPSS Inc., Chicago, IL, USA). Association coefficients were determined using Spearman two tailed test and correspondence analysis were determined by Chi-squared, statistical significance was considered when p < 0.05.

5. Conclusions

In this work, we validated a computational approach regarding epitope prediction, topological structure and pathway analyses to drive a rational vaccine design against leishmaniasis, using antigens tested in murine model described in literature. Our results suggest that CD4+ and CD8+ T cells are more related to leishmaniasis protection in EO than B cells. For a deeper analysis, we also used PPI networks enriched with KEGG pathways information. According to our results, proteins associated to protection are connected to few pathways when compared to proteins classified as “no protection”. In addition, analysis of PPIs and KEGG pathways associated to proteins from “protection” group corroborate the idea already published in the literature that reverse vaccinology approaches are able to identify proteins related to pathogenicity of infectious agents, helping researchers to understand virulence mechanisms and how immune responses from hosts are able to fight them. Obtained results may be helpful in discovering new potential antigens using computational approaches.


The authors acknowledge the Brazilian agencies Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (MCTI/CNPq/CT-BIOTEC-GENOPROT Nº 21/2010 560943/2010-5–Alexandre B. Reis, MCTI/CNPq/CT-BIOTEC Nº 27/2013 402688/2013-9–Alexandre B. Reis, PQ 2015 301526/2015-0–Jeronimo C. Ruiz MCTI/CNPq Nº 14/2015 301652/2015-0–Jeronimo C. Ruiz and MCTI/CNPq Nº 14/2013 486618/2013-7–Jeronimo C. Ruiz), Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) (PRONEX APQ-01373-14–Alexandre B. Reis, FAPEMIG 01/2013 APQ-01661-13–Jeronimo C. Ruiz and PPM IX 02/2015 PPM-00710-15–Jeronimo C. Ruiz), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Universidade Federal de Ouro Preto (UFOP) and Fundação Oswaldo Cruz (FIOCRUZ) for financial support. Rory C. F. Brito, Daniela M. Resende and João P. L. Velloso are grateful for CAPES scholarships, Alexandre B. Reis, Rodrigo Corrêa-Oliveira and Jeronimo C. Ruiz are also grateful for CNPq fellowships.

Supplementary Materials

Supplementary materials can be found at

Author Contributions

Author Contributions

Rory C. F. Brito participated with ideas and suggestions linked with the development and implementation of the bioinformatics methodology; also involved in the selection of candidate proteins to vaccine development against leishamaniasis in the literature, epitopes and subcelullar localization predictions and formal analysis of the data. Frederico G. Guimarães was responsable for PPI networks construction and analysis, and in mapping KEGG pathways associated to immunogenic proteins in PPI networks. Rory C. F. Brito, Frederico G. Guimarães and João P. L. Velloso participated in drafting the article and/or revising it critically for important intellectual content, and also created the figures. Rodrigo Corrêa-Oliveira, Jeronimo C. Ruiz, Alexandre B. Reis and Daniela M. Resende contributed materials and analysis tools. Jeronimo C. Ruiz and Daniela M. Resende participated in the design of the bioinformatics work. Jeronimo C. Ruiz, Alexandre B. Reis and Daniela M. Resende participated in the study conception, project administration, critical revision of the article, supervision and funding acquisition.

Conflicts of Interest

Conflicts of Interest

The authors declare that there is no conflict of interest.


1. Peacock C.S., Seeger K., Harris D., Murphy L., Ruiz J.C., Quail M.A., Peters N., Adlem E., Tivey A., Aslett M., et al. Comparative genomic analysis of three Leishmania species that cause diverse human disease. Nat. Genet. 2007;39:839–847. doi: 10.1038/ng2053. [PMC free article] [PubMed] [Cross Ref]
2. Gillespie P.M., Beaumier C.M., Strych U., Hayward T., Hotez P.J., Bottazzi M.E. Status of vaccine research and development of vaccines for leishmaniasis. Vaccine. 2016;34:2992–2995. doi: 10.1016/j.vaccine.2015.12.071. [PubMed] [Cross Ref]
3. Rezende A.M., Folador E.L., Resende D.M., Ruiz J.C. Computational prediction of protein–protein interactions in Leishmania predicted proteomes. PLoS ONE. 2012;7:e51304 doi: 10.1371/journal.pone.0051304. [PMC free article] [PubMed] [Cross Ref]
4. Xu C., Ye B., Han Z., Huang M., Zhu Y. Comparison of transcriptional profiles between CD4+ and CD8+ T cells in HIV type 1-infected patients. AIDS Res. Hum. Retrovir. 2014;30:134–141. doi: 10.1089/aid.2013.0073. [PMC free article] [PubMed] [Cross Ref]
5. Salay G., Dorta M.L., Santos N.M., Mortara R.A., Brodskyn C., Oliveira C.I., Barbieri C.L., Rodrigues M.M. Testing of four Leishmania vaccine candidates in a mouse model of infection with Leishmania (Viannia) braziliensis, the main causative agent of cutaneous leishmaniasis in the New World. Clin. Vaccine Immunol. CVI. 2007;14:1173–1181. doi: 10.1128/CVI.00060-07. [PMC free article] [PubMed] [Cross Ref]
6. Campbell K., Diao H., Ji J., Soong L. DNA immunization with the gene encoding P4 nuclease of Leishmania amazonensis protects mice against cutaneous Leishmaniasis. Infect. Immun. 2003;71:6270–6278. doi: 10.1128/IAI.71.11.6270-6278.2003. [PMC free article] [PubMed] [Cross Ref]
7. Fedeli C.E., Ferreira J.H., Mussalem J.S., Longo-Maugeri I.M., Gentil L.G., dos Santos M.R., Katz S., Barbieri C.L. Partial protective responses induced by a recombinant cysteine proteinase from Leishmania (Leishmania) amazonensis in a murine model of cutaneous leishmaniasis. Exp. Parasitol. 2010;124:153–158. doi: 10.1016/j.exppara.2009.09.003. [PubMed] [Cross Ref]
8. Montalvo-Alvarez A.M., Folgueira C., Carrion J., Monzote-Fidalgo L., Canavate C., Requena J.M. The Leishmania HSP20 is antigenic during natural infections, but, as DNA vaccine, it does not protect BALB/c mice against experimental L. amazonensis infection. J. Biomed. Biotechnol. 2008;2008:695432. doi: 10.1155/2008/695432. [PMC free article] [PubMed] [Cross Ref]
9. Champsi J., McMahon-Pratt D. Membrane glycoprotein M-2 protects against Leishmania amazonensis infection. Infect. Immun. 1988;56:3272–3279. [PMC free article] [PubMed]
10. Gonzalez C.R., Noriega F.R., Huerta S., Santiago A., Vega M., Paniagua J., Ortiz-Navarrete V., Isibasi A., Levine M.M. Immunogenicity of a Salmonella typhi CVD 908 candidate vaccine strain expressing the major surface protein gp63 of Leishmania mexicana mexicana. Vaccine. 1998;16:1043–1052. doi: 10.1016/S0264-410X(97)00267-3. [PubMed] [Cross Ref]
11. Mendez S., Gurunathan S., Kamhawi S., Belkaid Y., Moga M.A., Skeiky Y.A., Campos-Neto A., Reed S., Seder R.A., Sacks D. The potency and durability of DNA- and protein-based vaccines against Leishmania major evaluated using low-dose, intradermal challenge. J. Immunol. 2001;166:5122–5128. doi: 10.4049/jimmunol.166.8.5122. [PubMed] [Cross Ref]
12. Rivier D., Bovay P., Shah R., Didisheim S., Mauel J. Vaccination against Leishmania major in a CBA mouse model of infection: Role of adjuvants and mechanism of protection. Parasite Immunol. 1999;21:461–473. doi: 10.1046/j.1365-3024.1999.00244.x. [PubMed] [Cross Ref]
13. Sjolander A., Baldwin T.M., Curtis J.M., Bengtsson K.L., Handman E. Vaccination with recombinant Parasite Surface Antigen 2 from Leishmania major induces a Th1 type of immune response but does not protect against infection. Vaccine. 1998;16:2077–2084. doi: 10.1016/S0264-410X(98)00075-9. [PubMed] [Cross Ref]
14. Webb J.R., Campos-Neto A., Ovendale P.J., Martin T.I., Stromberg E.J., Badaro R., Reed S.G. Human and murine immune responses to a novel Leishmania major recombinant protein encoded by members of a multicopy gene family. Infect. Immun. 1998;66:3279–3289. [PMC free article] [PubMed]
15. Solioz N., Blum-Tirouvanziam U., Jacquet R., Rafati S., Corradin G., Mauel J., Fasel N. The protective capacities of histone H1 against experimental murine cutaneous leishmaniasis. Vaccine. 1999;18:850–859. doi: 10.1016/S0264-410X(99)00340-0. [PubMed] [Cross Ref]
16. Soussi N., Milon G., Colle J.H., Mougneau E., Glaichenhaus N., Goossens P.L. Listeria monocytogenes as a short-lived delivery system for the induction of type 1 cell-mediated immunity against the p36/LACK antigen of Leishmania major. Infect. Immun. 2000;68:1498–1506. doi: 10.1128/IAI.68.3.1498-1506.2000. [PMC free article] [PubMed] [Cross Ref]
17. Carrion J., Folgueira C., Alonso C. Immunization strategies against visceral leishmaniosis with the nucleosomal histones of Leishmania infantum encoded in DNA vaccine or pulsed in dendritic cells. Vaccine. 2008;26:2537–2544. doi: 10.1016/j.vaccine.2008.03.003. [PubMed] [Cross Ref]
18. Santos-Gomes G.M., Rodrigues A., Teixeira F., Carreira J., Alexandre-Pires G., Carvalho S., Santos-Mateus D., Martins C., Vale-Gato I., Marques C., et al. Immunization with the Leishmania infantum recombinant cyclophilin protein 1 confers partial protection to subsequent parasite infection and generates specific memory T cells. Vaccine. 2014;32:1247–1253. doi: 10.1016/j.vaccine.2014.01.024. [PubMed] [Cross Ref]
19. Agallou M., Smirlis D., Soteriadou K.P., Karagouni E. Vaccination with Leishmania histone H1-pulsed dendritic cells confers protection in murine visceral leishmaniasis. Vaccine. 2012;30:5086–5093. doi: 10.1016/j.vaccine.2012.05.075. [PubMed] [Cross Ref]
20. Khoshgoo N., Zahedifard F., Azizi H., Taslimi Y., Alonso M.J., Rafati S. Cysteine proteinase type III is protective against Leishmania infantum infection in BALB/c mice and highly antigenic in visceral leishmaniasis individuals. Vaccine. 2008;26:5822–5829. doi: 10.1016/j.vaccine.2008.08.065. [PubMed] [Cross Ref]
21. Aguilar-Be I., da Silva Zardo R., Paraguai de Souza E., Borja-Cabrera G.P., Rosado-Vallado M., Mut-Martin M., Garcia-Miss Mdel R., Palatnik de Sousa C.B., Dumonteil E. Cross-protective efficacy of a prophylactic Leishmania donovani DNA vaccine against visceral and cutaneous murine leishmaniasis. Infect. Immun. 2005;73:812–819. doi: 10.1128/IAI.73.2.812-819.2005. [PMC free article] [PubMed] [Cross Ref]
22. Ghosh A., Zhang W.W., Matlashewski G. Immunization with A2 protein results in a mixed Th1/Th2 and a humoral response which protects mice against Leishmania donovani infections. Vaccine. 2001;20:59–66. doi: 10.1016/S0264-410X(01)00322-X. [PubMed] [Cross Ref]
23. Resende D.M., Rezende A.M., Oliveira N.J., Batista I.C., Correa-Oliveira R., Reis A.B., Ruiz J.C. An assessment on epitope prediction methods for protozoa genomes. BMC Bioinform. 2012;13:309 doi: 10.1186/1471-2105-13-309. [PMC free article] [PubMed] [Cross Ref]
24. Costa C.H., Peters N.C., Maruyama S.R., de Brito E.C., Jr., Santos I.K. Vaccines for the leishmaniases: Proposals for a research agenda. PLoS Negl. Trop. Dis. 2011;5:e943 [PMC free article] [PubMed]
25. Reed S.G., Coler R.N., Campos-Neto A. Development of a leishmaniasis vaccine: The importance of MPL. Expert Rev. Vaccines. 2003;2:239–252. doi: 10.1586/14760584.2.2.239. [PubMed] [Cross Ref]
26. Larsen M.V., Lundegaard C., Lamberth K., Buus S., Lund O., Nielsen M. Large-scale validation of methods for cytotoxic T-lymphocyte epitope prediction. BMC Bioinform. 2007;8:424 doi: 10.1186/1471-2105-8-424. [PMC free article] [PubMed] [Cross Ref]
27. Nielsen M., Lundegaard C., Lund O. Prediction of MHC class II binding affinity using SMM-align, a novel stabilization matrix alignment method. BMC Bioinform. 2007;8:238 doi: 10.1186/1471-2105-8-238. [PMC free article] [PubMed] [Cross Ref]
28. Larsen J.E., Lund O., Nielsen M. Improved method for predicting linear B-cell epitopes. Immun. Res. 2006;2:2. doi: 10.1186/1745-7580-2-2. [PMC free article] [PubMed] [Cross Ref]
29. Chen J., Liu H., Yang J., Chou K.C. Prediction of linear B-cell epitopes using amino acid pair antigenicity scale. Amino Acids. 2007;33:423–428. doi: 10.1007/s00726-006-0485-9. [PubMed] [Cross Ref]
30. El Manzalawy Y., Dobbs D., Honavar V. Predicting linear B-cell epitopes using string kernels. J. Mol. Recognit. 2008;21:243–255. doi: 10.1002/jmr.893. [PMC free article] [PubMed] [Cross Ref]
31. El Manzalawy Y., Dobbs D., Honavar V. Predicting flexible length linear B-cell epitopes. Comput. Syst. Bioinform. Conf. 2008;7:121–132. [PMC free article] [PubMed]
32. Horton P., Park K.J., Obayashi T., Fujita N., Harada H., Adams-Collier C.J., Nakai K. WoLF PSORT: Protein localization predictor. Nucleic Acids Res. 2007;35:W585–W587. doi: 10.1093/nar/gkm259. [PMC free article] [PubMed] [Cross Ref]
33. Herrera-Najera C., Pina-Aguilar R., Xacur-Garcia F., Ramirez-Sierra M.J., Dumonteil E. Mining the Leishmania genome for novel antigens and vaccine candidates. Proteomics. 2009;9:1293–1301. doi: 10.1002/pmic.200800533. [PubMed] [Cross Ref]
34. Agallou M., Athanasiou E., Koutsoni O., Dotsika E., Karagouni E. Experimental Validation of Multi-Epitope Peptides Including Promising MHC Class I- and II-Restricted Epitopes of Four Known Leishmania infantum Proteins. Front. Immunol. 2014;5:268. doi: 10.3389/fimmu.2014.00268. [PMC free article] [PubMed] [Cross Ref]
35. Duarte A., Queiroz A.T., Tosta R., Carvalho A.M., Barbosa C.H., Bellio M., de Oliveira C.I., Barral-Netto M. Prediction of CD8+ Epitopes in Leishmania braziliensis Proteins Using EPIBOT: In Silico Search and In Vivo Validation. PLoS ONE. 2015;10:e0124786 doi: 10.1371/journal.pone.0124786. [PMC free article] [PubMed] [Cross Ref]
36. Freitas e Silva R., Ferreira L.F., Hernandes M.Z., de Brito M.E., de Oliveira B.C., da Silva A.A., de-Melo-Neto O.P., Rezende A.M., Pereira V.R. Combination of In Silico Methods in the Search for Potential CD4(+) and CD8(+) T Cell Epitopes in the Proteome of Leishmania braziliensis. Front. Immunol. 2016;7:327. [PMC free article] [PubMed]
37. Day M.J. Immunoglobulin G subclass distribution in canine leishmaniosis: A review and analysis of pitfalls in interpretation. Vet. Parasitol. 2007;147:2–8. doi: 10.1016/j.vetpar.2007.03.037. [PubMed] [Cross Ref]
38. De Almeida Silva L., Romero H.D., Prata A., Costa R.T., Nascimento E., Carvalho S.F., Rodrigues V. Immunologic tests in patients after clinical cure of visceral leishmaniasis. Am. J. Trop. Med. Hyg. 2006;75:739–743. [PubMed]
39. Rosa R., Marques C., Rodrigues O.R., Santos-Gomes G.M. Immunization with Leishmania infantum released proteins confers partial protection against parasite infection with a predominant Th1 specific immune response. Vaccine. 2007;25:4525–4532. doi: 10.1016/j.vaccine.2007.04.003. [PubMed] [Cross Ref]
40. Lemesre J.L., Holzmuller P., Goncalves R.B., Bourdoiseau G., Hugnet C., Cavaleyra M., Papierok G. Long-lasting protection against canine visceral leishmaniasis using the LiESAp-MDP vaccine in endemic areas of France: Double-blind randomised efficacy field trial. Vaccine. 2007;25:4223–4234. doi: 10.1016/j.vaccine.2007.02.083. [PubMed] [Cross Ref]
41. Alcolea P.J., Alonso A., Sanchez-Gorostiaga A., Moreno-Paz M., Gomez M.J., Ramos I., Parro V., Larraga V. Genome-wide analysis reveals increased levels of transcripts related with infectivity in peanut lectin non-agglutinated promastigotes of Leishmania infantum. Genomics. 2009;93:551–564. doi: 10.1016/j.ygeno.2009.01.007. [PubMed] [Cross Ref]
42. Benitez D., Medeiros A., Fiestas L., Panozzo-Zenere E.A., Maiwald F., Prousis K.C., Roussaki M., Calogeropoulou T., Detsi A., Jaeger T., et al. Identification of Novel Chemical Scaffolds Inhibiting Trypanothione Synthetase from Pathogenic Trypanosomatids. PLoS Negl. Trop. Dis. 2016;10:e0004617 doi: 10.1371/journal.pntd.0004617. [PMC free article] [PubMed] [Cross Ref]
43. Iborra S., Parody N., Abanades D.R., Bonay P., Prates D., Novais F.O., Barral-Netto M., Alonso C., Soto M. Vaccination with the Leishmania major ribosomal proteins plus CpG oligodeoxynucleotides induces protection against experimental cutaneous leishmaniasis in mice. Microb. Inf. Inst. Pasteur. 2008;10:1133–1141. doi: 10.1016/j.micinf.2008.06.002. [PubMed] [Cross Ref]
44. Cordeiro-Da-Silva A., Borges M.C., Guilvard E., Ouaissi A. Dual role of the Leishmania major ribosomal protein S3a homologue in regulation of T- and B-cell activation. Infect. Immun. 2001;69:6588–6596. doi: 10.1128/IAI.69.11.6588-6596.2001. [PMC free article] [PubMed] [Cross Ref]
45. Soto M., Requena J.M., Garcia M., Gomez L.C., Navarrete I., Alonso C. Genomic organization and expression of two independent gene arrays coding for two antigenic acidic ribosomal proteins of Leishmania. J. Biol. Chem. 1993;268:21835–21843. [PubMed]
46. Soto M., Alonso C., Requena J.M. The Leishmania infantum acidic ribosomal protein LiP2a induces a prominent humoral response in vivo and stimulates cell proliferation in vitro and interferon-gamma (IFN-γ) production by murine splenocytes. Clin. Exp. Immunol. 2000;122:212–218. doi: 10.1046/j.1365-2249.2000.01372.x. [PubMed] [Cross Ref]
47. Silverman J.M., Chan S.K., Robinson D.P., Dwyer D.M., Nandan D., Foster L.J., Reiner N.E. Proteomic analysis of the secretome of Leishmania donovani. Genome Biol. 2008;9:R35. doi: 10.1186/gb-2008-9-2-r35. [PMC free article] [PubMed] [Cross Ref]
48. McConville M.J., Naderer T. Metabolic pathways required for the intracellular survival of Leishmania. Annu. Rev. Microbiol. 2011;65:543–561. doi: 10.1146/annurev-micro-090110-102913. [PubMed] [Cross Ref]
49. Naderer T., Dandash O., McConville M.J. Calcineurin is required for Leishmania major stress response pathways and for virulence in the mammalian host. Mol. Microbiol. 2011;80:471–480. doi: 10.1111/j.1365-2958.2011.07584.x. [PubMed] [Cross Ref]
50. Madeira da Silva L., Beverley S.M. Expansion of the target of rapamycin (TOR) kinase family and function in Leishmania shows that TOR3 is required for acidocalcisome biogenesis and animal infectivity. Proc. Natl. Acad. Sci. USA. 2010;107:11965–11970. doi: 10.1073/pnas.1004599107. [PubMed] [Cross Ref]
51. Carter N.S., Yates P., Arendt C.S., Boitz J.M., Ullman B. Purine and pyrimidine metabolism in Leishmania. Adv. Exp. Med. Biol. 2008;625:141–154. [PubMed]
52. Real F., Vidal R.O., Carazzolle M.F., Mondego J.M., Costa G.G., Herai R.H., Wurtele M., de Carvalho L.M., Carmona e Ferreira R., Mortara R.A., et al. The genome sequence of Leishmania (Leishmania) amazonensis: Functional annotation and extended analysis of gene models. DNA Res. 2013;20:567–581. doi: 10.1093/dnares/dst031. [PMC free article] [PubMed] [Cross Ref]
53. Peters B., Bulik S., Tampe R., Van Endert P.M., Holzhutter H.G. Identifying MHC class I epitopes by predicting the TAP transport efficiency of epitope precursors. J. Immunol. 2003;171:1741–1749. doi: 10.4049/jimmunol.171.4.1741. [PubMed] [Cross Ref]
54. Nielsen M., Lundegaard C., Lund O., Kesmir C. The role of the proteasome in generating cytotoxic T-cell epitopes: Insights obtained from improved predictions of proteasomal cleavage. Immunogenetics. 2005;57:33–41. doi: 10.1007/s00251-005-0781-7. [PubMed] [Cross Ref]
55. Buus S., Lauemoller S.L., Worning P., Kesmir C., Frimurer T., Corbet S., Fomsgaard A., Hilden J., Holm A., Brunak S. Sensitive quantitative predictions of peptide-MHC binding by a ‘Query by Committee’ artificial neural network approach. Tissue Antigens. 2003;62:378–384. doi: 10.1034/j.1399-0039.2003.00112.x. [PubMed] [Cross Ref]
56. Nielsen M., Lundegaard C., Worning P., Hvid C.S., Lamberth K., Buus S., Brunak S., Lund O. Improved prediction of MHC class I and class II epitopes using a novel Gibbs sampling approach. Bioinformatics. 2004;20:1388–1397. doi: 10.1093/bioinformatics/bth100. [PubMed] [Cross Ref]
57. Nielsen M., Lundegaard C., Worning P., Lauemoller S.L., Lamberth K., Buus S., Brunak S., Lund O. Reliable prediction of T-cell epitopes using neural networks with novel sequence representations. Protein Sci. 2003;12:1007–1017. doi: 10.1110/ps.0239403. [PubMed] [Cross Ref]
58. Nielsen M., Lund O. NN-align. An artificial neural network-based alignment algorithm for MHC class II peptide binding prediction. BMC Bioinform. 2009;10:296 doi: 10.1186/1471-2105-10-296. [PMC free article] [PubMed] [Cross Ref]
59. Szklarczyk D., Franceschini A., Wyder S., Forslund K., Heller D., Huerta-Cepas J., Simonovic M., Roth A., Santos A., Tsafou K.P., et al. STRING v10: Protein–protein interaction networks, integrated over the tree of life. Nucleic Acids Res. 2015;43:D447–D452. doi: 10.1093/nar/gku1003. [PMC free article] [PubMed] [Cross Ref]
60. Kanehisa M. KEGG Bioinformatics Resource for Plant Genomics and Metabolomics. Methods Mol. Biol. 2016;1374:55–70. [PubMed]
61. Shannon P., Markiel A., Ozier O., Baliga N.S., Wang J.T., Ramage D., Amin N., Schwikowski B., Ideker T. Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13:2498–2504. doi: 10.1101/gr.1239303. [PubMed] [Cross Ref]

Articles from International Journal of Molecular Sciences are provided here courtesy of Multidisciplinary Digital Publishing Institute (MDPI)