Introduction

Leishmaniasis comprises a group of parasitic diseases caused by the genus Leishmania, which are widely distributed in tropical and subtropical regions throughout the world (WHO 2010). Tegumentary leishmaniasis (TL) is endemic in more than 70 countries worldwide, in which 90% of the cases are registered in Brazil, Peru, Afghanistan, Algeria, Pakistan, Saudi Arabia, and Syria (Alvar et al. 2012). The disease exhibits distinct clinical manifestations, such as cutaneous leishmaniasis (CL), diffuse cutaneous leishmaniasis (DCL), and mucosal leishmaniasis (ML) (Grimaldi and Tesh 1993). Frequently, geographic distribution and clinical manifestations are applied to identify Leishmania spp. This measure is considered crucial to select the most appropriate therapeutic regimen to be administered to each patient, as well as to determine possible control measures in epidemiological studies and to predict the risk of dissemination in immunosuppressed patients (Foulet et al. 2006; Talmi-Frank et al. 2010).

The Montenegro skin test (MST) is used for the immunological diagnosis of TL; however, it does not make it possible to distinguish between an acute or chronic disease or if the patients have already been cured of disease (Gomes et al. 2008). Polymerase chain reaction (PCR)-based methods are a powerful tool to detect parasite DNA in TL patients, and they have proven to be highly specific when compared with traditional parasitological methods (Chargui et al. 2005). However, their sensitivity varies according to the time when the lesions occurred, since a low number of parasites can be present in later lesions, hampering the sensitivity of these tests (Goto and Lindoso 2010; Boggild et al. 2011). Additionally, these methods require invasive procedures for sample collection, which limit their use (Duarte et al. 2015).

A serological diagnosis is based on the presence of a specific antileishmanial humoral immunity. The sensitivity of the tests depends upon the assay and its methodology, but the specificity depends on the antigen rather than the methodological format used (Elmahallawy et al. 2014). ELISA has been employed as a diagnostic method for several diseases, including leishmaniasis. It requires a specific antigen as the starter reagent to capture a specific antibody; therefore, it is based on distinct antigenic molecules, such as surface antigens, ribosomal or nuclear proteins, histones, and kinesin-related proteins (Paiva-Cavalcanti et al. 2015). In this context, recombinant antigens have considerably improved the sensitivity and specificity of this technique. Between them, a recombinant kinesin-related protein, the recombinant K39 (rK39) protein, has been successfully employed for the serodiagnosis of visceral leishmaniasis (VL) (Sundar and Rai 2002; Maia et al. 2012). The introduction of the rK39 immunochromatographic strip rapid test has facilitated the field applicability of serological methods for VL control (Sakkas et al. 2016). On the other hand, for TL serodiagnosis, important drawbacks remain, since patients can present undetectable or low levels of anti-rK39 antibodies and, consequently, false-negative results in the serological exams (Ahluwalia et al. 2004; Romero et al. 2005). In addition, like VL, antibodies in TL can remain positive for months after the patient has been treated and do not differentiate between past and current infection. In this context, the search remains to identify new candidates to be used for a greater sensitivity and specific serodiagnosis of TL.

In a recent immunoproteomic approach performed in Leishmania braziliensis promastigotes and axenic amastigotes, the main etiological agents of TL in Brazil (Silveira et al. 2004), hypothetical proteins were recognized in immunoblotting assays by antibodies in TL patients’ sera (Duarte et al. 2015). These proteins are characterized as hypothetical due to their low amino acid sequence identity for known proteins (Lubec et al. 2005). Thus, in the present study, bioinformatics tools were used to functionally predict one of these proteins, namely LbHyM (XP_001566959.1), as well as to evaluate the antigenic properties of this recombinant antigen in an attempt to improve serological diagnoses of TL.

Materials and methods

Bioinformatics analyses

Sequence retrieval

The LbHyM protein FASTA sequence was collected from UniProt (http://www.uniprot.org/) using the NCBI reference sequence, as described in The UniProt Consortium (2015). In addition, a similar Trypanosoma cruzi protein (XP_806283.1) was used as a control sequence. The BepiPred 1.0 server was employed to evaluate the main B cell epitopes of these antigens. Further analyses for multiple sequence alignments were performed using the Clustal Omega server (http://www.ebi.ac.uk/Tools/msa/clustalo/), as described in Sievers et al. (2011).

Physicochemical characterization of LbHyM protein

The theoretical measurements of physiochemical properties of the LbHyM protein were performed using the ProtParam tool in the ExPASy server (Gasteiger et al. 2005). The parameters computed by the program include molecular weight, theoretical isoelectric point, extinction coefficient, instability index, aliphatic index, and grand average of hydropathicity (GRAVY).

Subcellular localization, function prediction, and performance assessment

The subcellular localization of LbHyM was analyzed using the PSORTb 3.0 (Yu et al. 2010), PSLpred (Bhasin et al. 2005), CELLO (Yu et al. 2004), and WolfPsort (Horton et al. 2007) programs. The SignalP 4.1 (Petersen et al. 2011) was used to predict signal peptide, while SecretomeP (Bendtsen et al. 2004) was employed to identify proteins involved in the non-classical secretory pathway. Transmembrane Hidden Markov Model (TMHMM) (Krogh et al. 2001) was used to predict the propensity of a protein to be a membrane protein. The functional domain of the LbHyM protein was predicted by using distinct available databases: SMART (Lee et al. 2016), ProtoNet 6.0 (Rappoport et al. 2012), SUPERFAMILY (Wilson et al. 2009), and CATH (Sillitoe et al. 2015). The receiver operator characteristic (ROC) curve was used to validate the predicted subcellular localization and the function of the LbHyM protein. The two binary numerals “0” and “1” were used to classify the prediction as true-negative (“0”) or true-positive (“1”); the integers 2, 3, 4, and 5 were used as confidence rating for each case (Sonego et al. 2008). The ROC curves were executed through 27 control sequences from Leishmania proteins with known function and subcellular localization (Ogungbe and Setzer 2013). The control sequences (supplementary table) were used to compare the values obtained with the LbHyM protein. The results were submitted to ROC curve analyses: Web-based Calculator for ROC curves (Eng 2014) in “format 1,” as required by the software. The online software calculates the ROC curves, and the results were expressed as sensitivity (Se), specificity (Sp), accuracy (Ac), and area under the curve (AUC).

Patients’ sera

The present study was approved by the Human Research Ethics Committee (COEP) from the Federal University of Minas Gerais (UFMG), Belo Horizonte, MG, Brazil (protocol number: CAAE–323431 14.9.0000.5149). Before collecting samples, patients received an individual copy of the study policy, which was reviewed by an independent person, and all participants signed an informed consent form in Portuguese before their samples were collected. All sera were collected by performing a venipuncture of the medial vein in tubes without an anticoagulant and were kept at 37 °C for 15 min, at which time they were centrifuged at 4000×g for 15 min. Samples were then separated and kept at −80 °C until use. Serum samples were obtained from patients with diagnoses of TL, which were classified as CL (n = 20, including 14 males and 6 females, with ages ranging from 30 to 56 years) or ML (n = 25, including 18 males and 7 females, with ages ranging from 28 to 57 years), which, in both cases, were collected from an endemic area of leishmaniasis (Belo Horizonte). The diagnosis was confirmed by means of clinical evaluation of the lesions, as well as by the direct demonstration of the parasites in Giemsa-stained smears of skin biopsies (CL) and/or mucosal fragments (ML), as well as by the PCR technique to identify the L. braziliensis kinetoplastid DNA. All patients also underwent a positive intradermal Montenegro test. In addition, none of the patients had been previously treated with antileishmanial drugs before their samples were collected. Serum samples were also obtained from healthy individuals living in endemic (n = 25, including 17 males and 8 females, with ages ranging from 24 to 55 years, Belo Horizonte, MG, Brazil) or non-endemic (n = 25, including 16 males and 9 females, with ages ranging from 21 to 45 years; Poços de Caldas, MG, Brazil) areas of leishmaniasis, which were used as controls. These subjects were clinically evaluated, showed no clinical symptoms of leishmaniasis, and presented negative results from antileishmanial serology. Serum samples from Chagas disease patients (n = 10, including 7 males and 3 females, with ages ranging from 24 to 58 years) were also used as cross-reactive samples.

Post-treatment evaluation of antileishmanial antibodies

For post-treatment serological follow-up, patients with CL (n = 6, including 4 males and 2 females, with ages ranging from 24 to 51 years) or ML (n = 15, including 9 males and 6 females, with ages ranging from 21 to 62 years) were serologically examined before the initiation to undergo treatment and 6 months after the end of the treatment sessions. Before treatment, CL patients had a disease evolution time varying from 5 months to 15 years, while ML patients presented a disease evolution time ranging from 6 months to 27 years. In addition, all patients were treated with pentavalent antimonials (Sanofi Aventis Farmacêutica Ltda., Suzano, São Paulo, Brazil).

Parasites

Leishmania (Viannia) braziliensis (MHOM/BR/1975/M2904) was used. Parasites were grown at 24 °C in Schneider’s medium (Sigma, St. Louis, MO, USA) supplemented with 20% inactivated fetal bovine serum (FBS, Sigma), 20 mM L-glutamine, 200 U/mL penicillin, and 100 μg/mL streptomycin, at pH 7.4. A soluble L. braziliensis antigenic preparation (SLA) was obtained from 2 × 108 stationary-phase promastigotes, as described in Coelho et al. (2003).

Cloning, expression, and purification of the LbHyM protein

The cloning, expression, and purification of the rLbHyM (XP_001566959.1) protein were performed as described (Duarte et al. 2015). After purification, the recombinant protein was passed through a polymyxin-agarose column (Sigma) to remove residual endotoxin content (<10 ng of LPS per 1 mg of recombinant protein, measured by the Quantitative Chromogenic Limulus Amebocyte Assay QCL-1000, BioWhittaker, MD, USA).

ELISA with patients’ sera

Titration curves were performed to determine the most appropriate concentration of serum samples and antigen dilutions to be used in the ELISA assays. Microtiter immunoassay plates (Jetbiofil®, Belo Horizonte) were coated with rLbHyM or L. braziliensis SLA (1.0 μg per well, in both cases), which were dissolved in 100 μL of coating buffer (50 mM carbonate buffer, pH 9.6) for 18 h at 4 °C. Next, free binding sites were blocked using 250 μL of PBS-T (phosphate buffer saline 1× plus Tween 20 at 0.05%), containing 5% albumin for 1 h and at 37 °C. After washing the plates five times with PBS-T, they were incubated with 100 μL of individual serum samples (1:100 diluted in PBS-T), for 1 h at 37 °C. Plates were subsequently washed five times in PBS-T and incubated with an antihuman IgG horseradish-peroxidase-conjugated antibody (1:6500 diluted in PBS-T), for 1 h at 37 °C. After washing the plates five times with PBS-T, the reactions were developed by incubation with 100 μL per well of a solution consisting of 2 μL H2O2, 2 mg orto-phenylenediamine and 10 mL citrate-phosphate buffer at pH 5.0, for 30 min, in the dark. Reaction was stopped by adding 25 μL 2 N H2SO4. The optical density (OD) was read in an ELISA microplate spectrophotometer (Molecular Devices, Spectra Max Plus, Canada), at 492 nm. All sera samples were evaluated on the same day using the same reagents (lots, dilutions, etc), aiming to reduce possible experimental variation and the interference in the interpretation of data, which could occur if the assays had been performed in different days. In addition, controls were included in all plates.

Statistical analysis

The results were entered into Microsoft Excel (version 10.0) spreadsheets and analyzed using GraphPad PrismTM (version 6.0 for Windows). The lower limits of positivity (cutoff) for the diagnostic antigens (rLbHyM and L. braziliensis SLA) were established for maximum likelihood ratio using the ROC curves. The curves for TL patients were plotted with the values from the cutaneous and mucosal leishmaniasis patients (TL, n = 45) versus those from the healthy control groups (HC, n = 50), which were composed by sera from healthy individuals living in endemic or non-endemic areas of leishmaniasis. The diagnostic capacity of each antigen was measured by assessing its sensitivity, specificity, and 95% confidence intervals (95%CI). To evaluate the post-treatment antileishmanial antibody levels, the reactivity index was computed by further correcting the mean OD value for each sample using the individual antigens by dividing it by their respective cutoff value previously selected. As a consequence, the reactivity index represents multiples of reactivity with respect to that of the cutoff value. The unpaired Student t test was used, and significant differences were considered with P < 0.05. The inter-run duplicates were developed by using the same samples tested in two independent ELISA experiment assays (sessions 1 and 2) for both antigens (rLbHyM and SLA), which were assessed by Spearman’s rank correlation coefficient and Bland-Altman plot analysis, showing excellent agreement between the OD results of both analyses. Also, a Bland-Altman plot analysis showed that differences in the OD values for both antigens between the duplicates did not increase with increases in their mean.

Results

Bioinformatics to evaluate the LbHyM protein

ProtParam tool was used to analyze physiochemical properties from the amino acid sequence of the LbHyM protein. The sequence was predicted to have 358 amino acids, with a molecular weight of 40.8 kDa and an isoelectric point of 5.43. The negative GRAVY index of −1.113 is indicative of a hydrophilic and soluble protein. A systematic workflow consisting of several bioinformatics tools and databases was defined and used with the goal of predicting the subcellular localization and functional annotation of LbHyM. Consensus predictions made by the web tools PSORTb, PSLpred, CELLO, WolfPsort, SignalP 4.1, SecretomeP, and TMHMM for subcellular localization, as well as SUPERFAMILY, CATH, SMART, and ProtoNet for conserved domains, were used to classify the sequence. In the results, LbHyM was suggested to be a cytoplasmic/extracellular protein with average accuracy, sensitivity, and specificity values of 68.75, 100, and 50%, respectively. Moreover, the protein was predicted to be involved in the non-classical secretory pathway by SecretomeP algorithm. Regarding the functional annotation of LbHyM, the results suggested that it is a kinesin-like protein, presenting average accuracy, sensitivity, and specificity values of 78.5, 78.5, and 100%, respectively (Fig. 1).

Fig. 1
figure 1

Computational framework used for predicting the subcellular localization and parasite function of the hypothetical LbHyM protein from Leishmania braziliensis species

Assessment of the rLbHyM protein for the TL serodiagnosis

In the present study, rLbHyM was evaluated as a diagnostic marker for TL serodiagnosis. More appropriately, this recombinant antigen was employed in serological assays by ELISA to detect CCL and ML patients, but not Chagas disease patients or healthy subjects living in an endemic area of disease. In addition, this protein was used to evaluate the antileishmanial serological follow-up in treated patients, since serum samples from these individuals were collected before and 6 months after the treatments were performed. As a comparative antigen, L. braziliensis SLA was employed in the assays. The results showed that 100% of TL patients (n = 45: CL (n = 20) and ML (n = 25)) presented mean reactivities above the cutoff value, when rLbHyM was used as an antigenic source, and their levels were significantly higher than those obtained when sera from healthy controls were evaluated (Fig. 2a). ROC curves were constructed, and the recombinant protein showed a maximum AUC value of 1.0 (Fig. 2b), whereas when using SLA, this value was of 0.96 (Fig. 2c). In addition, the sensitivity and specificity values of rLbHyM to differentiate TL patients from healthy subjects were 100 and 98.0%, respectively, whereas for the SLA, these values were 75.6 and 98.0%, respectively (Table 1). The cross-reactivity of the recombinant antigens against Chagas disease was also evaluated. Firstly, an amino acid sequence comparison between LbHyM and a T. cruzi hypothetical protein was performed (Fig. 3a). The results showed that the proteins present a nearly 78% similarity in their amino acid sequences. Moreover, when an evaluation of the specific B cell epitopes was performed, low identity was found between them in both protein sequences (data not shown). Secondly, the mean reactivities were evaluated using rLbHyM or SLA as antigens against serum samples collected from Chagas disease patients. In the results, a very low reactivity was found when rLbHyM was used in the assays; however, when SLA was used, mean reactivities were similar to those found in the TL group (Fig. 3b), demonstrating a high humoral cross-reactivity. Regarding the inter-run variability, Spearman’s rank correlation coefficient showed excellent agreement between the OD values in both ELISA assays performed, when both antigens (rLbHyM and L. braziliensis SLA) were employed in the experiments (Fig. 4). Two different ELISA sessions were developed by using the same samples tested in two independent experiments. A Bland-Altman plot analysis showed that differences in the OD values between the duplicates did not increase with the variation of their serological mean. When humoral immunity in patients undergoing follow-up was evaluated regarding anti-LbHyM antibodies levels before and after treatment, a significant decrease in the anti-protein antibody production was found after treatment, when compared to results obtained before treatment. On the other hand, when using SLA as an antigen source, similar antileishmanial antibodies levels were found before and after treatments performed in patients (Fig. 5).

Fig. 2
figure 2

Evaluation of ELISA reactivity against rLbHyM and L. braziliensis SLA. ELISA assays were performed using serum samples from patients with tegumentary leishmaniasis (TL) (n = 45, 20 serum samples from patients with cutaneous leishmaniasis and 25 serum samples from patients with mucosal leishmaniasis) and healthy controls (HC) (n = 50, consisting of sera from individuals living in endemic (n = 25) or non-endemic (n = 25) areas of leishmaniasis) (a). The ROC curves for TL patients were plotted with the values from cutaneous and mucosal leishmaniasis patients (TL groups) versus those from the healthy controls (composed by sera from healthy individuals living in endemic or non-endemic areas of leishmaniasis). The diagnostic capacity of each antigen (rLbHyM protein in b and L. braziliensis SLA in c) was measured by assessing its sensitivity (Se), specificity (Sp), area under the curve (AUC), standard deviation (SD), and 95% confidence interval (95%CI)

Table 1 Measures of diagnostic performance by ELISA using human sera. ELISA assays were performed with serum samples from patients with cutaneous (CL, n = 20) or mucosal (ML, n = 25) leishmaniasis, as well as with sera from healthy subjects living in endemic (n = 25) or non-endemic (n = 25) areas of leishmaniasis. ROC curves were used to determine ELISA sensitivity, specificity, 95% confidence interval (95%CI), and likelihood ratio (LR)
Fig. 3
figure 3

Comparative evaluation of the cross-reactivity of the diagnostic antigens. The rLbHyM and a Trypanosoma cruzi hypothetical protein (XP_845584.1) sequences were aligned using the Clustal Omega program. The identical residues are shown (black color), as well as the conservative (gray color) and semi-conservative (white color) substitutions (a). ELISA assays were performed using serum samples from patients with tegumentary leishmaniasis (TL) (n = 45, 20 serum samples from patients with cutaneous leishmaniasis and 25 serum samples from patients with mucosal leishmaniasis) and healthy controls (HC) (n = 50, consisting of sera from individuals living in endemic (n = 25) or non-endemic (n = 25) areas of leishmaniasis). In addition, sera from Chagas disease patients were also used (CD, n = 10) (b). The mean ± standard deviation of each group is shown

Fig. 4
figure 4

Evaluation of the inter-run variability of rLbHyM and L. braziliensis SLA antigens in ELISA experiments. Spearman’s correlation coefficient analysis for rLbHyM (a) and L. braziliensis SLA (b) were calculated when two different ELISA sessions (session 1 and session 2) were developed by using the same samples tested in two independent experiments. Also, a Bland-Altman plot of the differences between the OD values of the duplicates and their means for rLbHyM (c) and SLA (d) are shown. The curves for TL patients were plotted with the values from cutaneous and mucosal leishmaniasis patients (TL, n = 45) versus those from the healthy control groups (HC, n = 50), which were composed by sera from healthy individuals living in endemic or non-endemic areas of leishmaniasis. The dotted lines (panels b and d) represent the mean (heavy dotted line) and plus or minus two standard deviations (light dotted lines)

Fig. 5
figure 5

Post-treatment evaluation of antileishmanial antibody levels. ELISA assays were performed using serum samples from patients with cutaneous (CL) (n = 6, including 4 males and 2 females, with ages ranging from 24 to 51 years) or mucosal (ML) (n = 15, including 9 males and 6 females, with ages ranging from 21 to 62 years) leishmaniasis, which were collected before beginning treatment and 6 months after the onset of the treatments. The mean reactivities of each patient were normalized by cutoff values for each antigen. The data are shown as a reactivity index

Discussion

The diagnostic value of serology for TL depends on the causative Leishmania species, which differ in eliciting immunity, as well as in the immune response developed by the infected hosts and in the sensitivity of the assays used (Romero et al. 2005). Considering the fact that symptoms are not specific in disease diagnosis, TL is usually detected in a delayed manner (Gradoni 2013). In addition, parasitological diagnosis based on the direct demonstration of amastigote forms in tissue biopsy or mucosal fragments, considered the gold standard for TL, presents problems related to its sensitivity, especially in older lesions where parasites are scarce (Goto and Lindoso 2010). In this context, serological tests to detect anti-Leishmania antibodies, such as ELISA, IFAT (indirect immunofluorescent-antibody test), and DAT (direct agglutination test) can be considered but they also present problems related to their sensitivity and/or specificity, according the antigens employed (Gomes-Silva et al. 2009).

An ideal antigenic candidate for TL serodiagnosis should identify all patients, but not cured and/or treated patients, or those presenting cross-reactive diseases (Vexenat et al. 1996; Duarte et al. 2015; Coelho et al. 2016). In the present study, a hypothetical protein, which was recognized by antibodies in TL patients’ sera by a previous immunoproteomic study (Duarte et al. 2015), showed perfect sensitivity and specificity values and presented no cross-reactivity with Chagas disease patients’ sera.

Most serological studies on leishmaniasis have been developed for VL, since humoral immunity is usually exacerbated in patients infected with viscerotropic Leishmania species (Tavares et al. 2003). Serological tests employed to diagnose TL proved to be advantageous, as they are minimally invasive and easy to perform. However, these do lack specificity, given that many patients can present low antileishmanial antibody levels, leading to false-negative results and non-infected individuals. In addition, living in endemic areas of disease can also present false-positive results in the serological exams (Celeste et al. 2004; Souza et al. 2013; Duarte et al. 2015). In this context, the lack of an accurate diagnosis is one of the most important factors that has led to the expansion of TL cases not only affecting global health but also worsening poverty in low-income countries due to the greater morbidity (Costa et al. 2016). Therefore, improvements in TL serodiagnosis are of utmost importance. Furthermore, the identification of new antigens to be employed in more sensitive and specific serological assays should also be considered relevant.

Nowadays, the follow-up of antileishmanial antibodies after treatment is not considered to be a control measure for an effective therapy (Vallur et al. 2015). However, this procedure could reveal a decline of anti-Leishmania antibody levels in successfully treated patients (Gidwani et al. 2011), corroborating with improvements in their state of health. In the present study, rLbHyM was recognized in low levels when the sera from cured and treated patients were used. On the other hand, when using a L. braziliensis antigenic extract as an antigen source, a similar reactivity was found when the sera from patients were evaluated, making it difficult to serologically distinguish between treated and non-treated patients. Using an ELISA-Leishmania amazonensis SLA, Romero et al. (2005) showed a decrease in the antibody levels in CL patients after treatment but the results were unrelated to therapeutic responses. In contrast, Celeste et al. (2014) detected no decrease in the post-treatment antibody levels, when an ELISA assay using the recombinant HSP83 protein was employed. Although the cure control by serological tests is a controversial subject in the literature (Gomes-Silva et al. 2009; Souza et al. 2013; Menezes-Souza et al. 2014), the persistence of high antibody levels after treatment may well indicate the continued presence of viable parasites, raising the possibility of a future relapse of the disease (Fagundes-Silva et al. 2012; Duarte et al. 2015; Coelho et al. 2016).

Although nearly 130 serum samples were used in this study, new experiments are necessary to evaluate a larger serological panel, in order to further define the efficacy of our antigenic marker for TL serodiagnosis. For instance, our panel did not contain samples from VL patients, leprosy, malaria, and tuberculosis, among others. In addition, sera from treated patients were collected only 6 months after the end of the treatments. In this context, a longer follow-up should be performed to correlate the clinical cure of patients with lower anti-rLbHyM humoral reactivity. In this context, the data of the present study should be taken as a proof-of-concept of the capacity of this proposed antigen to achieve a proper TL serodiagnosis. These data should also be evaluated as a prospective antigen in the follow-up of cured and treated patients, rendering LbHyM a possible serological marker of clinical cure.

The hypothetical protein evaluated in this work was suggested to be a kinesin-like protein, due to the results obtained using bioinformatics tools. Some studies have shown that Leishmania kinesin is recommended for VL serodiagnosis. Burns et al. (1993) identified a kinesin-related gene product, LcKin, as a candidate diagnostic antigen by screening a Brazilian Leishmania infantum genomic library with serum of Leishmania donovani-infected patients. Among the most promising antigens derived from kinesin, rK39 has shown good results for the serodiagnosis of VL (Diro et al. 2007; Maia et al. 2012). A commercial kit (Kalazar Detect™ Test, InBios International, Inc., Seattle, WA, USA) was developed and is a non-invasive immunochromatographic strip assay for the qualitative detection of rK39-specific antibodies for the L. donovani complex in human serum samples, presenting good results in the diagnosis of VL patients (Srivastava et al. 2011; Costa et al. 2016). Although rK39 has been important for VL serodiagnosis, this antigen does not allow for the detection of antibodies in CL or ML patients (Singh and Sivakumar 2003), since these subjects remain serologically negative against this recombinant protein (Burns et al. 1993). However, to the best of our knowledge, the present study is the first of its kind in which a kinesin-like protein was successfully used to serologically identify TL patients, as well as to discriminate patients after the treatments had been performed, thus suggesting the possibility of using this recombinant protein as a serological marker for TL, and to correlate the presence of low levels of anti-rLbHyM antibodies with the clinic cure of patients.

Cross-reactivity of Leishmania antigens with serum samples from geographic areas where other protozoa diseases, particularly Chagas disease, overlap with leishmaniasis is common (Guimarães et al. 1991; Celeste et al. 2004). To overcome such non-specific reactions, recombinant proteins have been applied with varying success. The present study obtained maximum sensitivity and specificity indices for an ELISA-rLbHyM assay, to specifically differentiate the CL and ML patients from the other groups, whereas using SLA as an antigen, similar values were found when the sera from the TL and Chagas disease patients were tested.

Although the predictions performed using the LbHyM protein in relation to its subcellular localization and parasite function have not been experimentally validated yet, the in silico screening using distinct bioinformatics programs described here could be considered as a new technological strategy, which could be applied to predict other hypothetical proteins in distinct pathogens, aiming to identify new molecules to biotechnological applications against infections caused by different microorganisms.

In conclusion, the present study data showed that high levels of anti-rLbHyM antibodies could indicate active CL or ML. In addition, this recombinant antigen was not cross-reactive with Chagas disease patients’ sera or with those from healthy subjects living in endemic or non-endemic areas of leishmaniasis. Furthermore, rLbHyM showed promising results as a serological marker to evaluate post-treatment follow-up by monitoring the humoral immunity of these subjects.