Identifying promising peptide targets for leprosy serological tests: From prediction to ELISA

Abstract

Leprosy remains a significant health concern, particularly in India, Brazil, and Indonesia. Early diagnosis is essential to prevent complications, highlighting the need for improved diagnostic tools. This study aimed to identify novel Mycobacterium leprae antigens and assess their effectiveness against human sera through immunotools for antibody response evaluation. Using bioinformatics, we predicted B-cell epitopes in M. leprae, which were chemically synthesized and tested via dot blotting with sera from leprosy patients, tuberculosis patients, and healthy controls. Promising peptides underwent further analysis through ELISA using 465 serum samples from leprosy patients, household contacts, and healthy controls across Brazil. The samples were also tested against known antigens HSA-NDO, LID-1, and NDO-LID. A total of 102 epitope sequences were generated, of which eight (PEP1 to PEP8) demonstrated the ability to differentiate between individuals with and without exposure to M. leprae. The results of the ELISA test exhibited statistically significant differences in absorbance responses between the experimental groups for the novel synthetic peptides (p < 0.05). PEP3, PEP4, and PEP5 demonstrated the most favorable outcomes, with values of the area under the receiver operating characteristic curve (AUC) of 0.9759, 0.9796 and 0.9551 respectively in the comparison of healthy controls with household contacts, and 0.8257, 0.7945, and 0.7961 comparing the same controls with patients. Furthermore, the synthetic peptides demonstrated superior sensitivity, specificity, and AUC compared to HSA-NDO, LID-1, and NDO-LID. The identified peptides showed significant responses in samples from patients and household contacts (HHC), indicating their potential for tracing exposure to M. leprae bacilli. These novel synthetic peptides could enhance the sensitivity of rapid diagnostic tests for leprosy, facilitating early detection of the infection. This could help prevent disease progression and interrupt transmission.

1. Introduction

Leprosy is a chronic infectious disease caused by Mycobacterium leprae. In 2022, approximately 174,000 new cases were diagnosed worldwide, with a prevalence rate of 21.8 per million population. The highest incidences of the disease are still found in India, Brazil, and Indonesia.¹ Experts believe that the trend in the number of people with leprosy is, somewhat, higher than what specialists can diagnose.²

M. leprae infection is more prevalent than the disease but the distribution and transmission of the infection, and the factors that lead to illness are not fully understood, mainly due to the difficulty in cultivating the bacteria, which has limited research understanding leprosy. This disease is treatable, and curable, although, remains a neglected tropical disease, receiving no adequate attention from the public policies of developing countries.3, 4

The pathogen primarily affects the peripheral nervous system and skin. The disease path is determined by individual host immunity, and host genetic factors are also believed to influence susceptibility to infection and disease progression.5, 6 By the late 20th century, the World Health Organization (WHO) introduced a simplified classification system to enhance access to medical care in resource-limited settings. This system categorized leprosy cases into paucibacillary (PB) and multibacillary (MB) types, based on the bacilloscopic index (BI) and the number of skin lesions. PB cases were defined as having a BI lower than 2+ and up to five skin lesions and/or only one nerve trunk is involved, while MB cases were characterized by a BI of 2+ or higher and more than five skin lesions and/or more than one nerve trunk.⁷ Treatment protocols were also standardized: PB cases were managed with six monthly supervised doses of rifampicin and dapsone, supplemented by daily self-administered dapsone, whereas MB cases required 12 monthly supervised doses of rifampicin, clofazimine, and dapsone, along with daily self-administered doses of clofazimine and dapsone.⁸

Clinically, multibacillary (MB) lepromatous variants are distinguished from paucibacillary (PB) tuberculoid forms.⁹ The generally accepted concept is that MB patients are the main source of infection, and people living with these patients are more susceptible to being infected.10, 11 Although less contagious, PB patients are still infectious.¹² Thus, a control strategy based on diagnosed and treated cases should aim to reduce the transmission of the microorganism. Therewith, the chain of transmission would be broken, and then leprosy would cease.¹³ Hence, early diagnosis in both forms is crucial to avoid disease progression and sequelae. More accurate, rapid, and cost-effective screening tests are needed to improve case detection.

To date, no accurate serological test has been developed for the detection of patients with PB leprosy. To enhance serological sensitivity in PB forms and facilitate the early diagnosis of the disease, current studies have assessed the efficacy of rapid lateral flow (LF) tests, NDO-LID®, utilizing LID-1 antigen derived from the fusion of M. leprae recombinant proteins ML0405 and ML2331, conjugated to synthetic disaccharide-octyl of PGL-1 (ND-O).14, 15 These tests facilitate early diagnosis of leprosy in numerous cases and provide a device for fieldwork due to their relatively low cost and user-friendliness, which are advantageous in resource-limited settings. However, while high levels of antibodies against some of the aforementioned antigens were observed in patients with MB forms, low or medium levels were detected in patients with PB forms.16, 17 At present, there is no rapid diagnostic test that can distinguish between infection and exposure to M. leprae with high sensitivity, specificity, or accuracy.

Modern systems biology approaches are being employed as alternative diagnostic strategies and are currently under investigation as potential screening tools in research fields. Such methods include the use of antigens and peptides for antibody recognition in patients and household contacts, as well as sera for the detection of other neglected diseases.18, 19, 20, 21 The objective of our research was to identify and serologically validate previously unpublished molecules that could facilitate the screening of leprosy using a methodology for obtaining targets that combines bioinformatics computational tools and chemical synthesis.

2. Material and methods

2.1. Study population/Clinical samples

Serum samples from 475 individuals (6 – 88 years old) were selected for evaluation. Of these, 225 were patients with a previous diagnosis of leprosy under treatment, 166 were household contacts of those patients (contacts of MB = 94; contacts of PB = 72), and 74 were healthy control subjects from a medium endemicity area. Ten serum samples were from patients with Tuberculosis disease (TB). Samples were collected from two distinct geographical regions of Brazil: Southeast (Instituto René Rachou – Fiocruz Minas) and Northeast (Hospital Universitário de Sergipe/UFS). Leprosy patients were diagnosed by specialist physicians at their respective healthcare facilities using the operational classification. The study was conducted by observing the ethical principles of the National Health Council of Brazil and received the approval of the Research Ethics Committee of the Federal University of Minas Gerais (COEP: #11884919.4.0000.5149).

2.2. M. leprae B-cell epitopes prediction

The predicted Mycobacterium leprae TN strain proteome was retrieved from the NCBI (Biosample ID SAMEA1705921). The reference proteome was filtered to exclude sequences that did not start with ATG, did not end with a stop codon; had less than 100 amino acids, and those with “X” (unknown) as an amino acid. Proteins encoded by genes larger than 100 base pairs, with the presence of the initial methionine, and derived from non-pseudogenes were selected using PERL scripts. These proteins were then submitted to BepiPred (https://www.cbs.dtu.dk/services/BepiPred-2.0) for B-cell epitope predictions by using a cut-off of 0.6²² and IUPred 2.0 (https://iupred.elte.hu/) programs using default parameters for predictions of linear B-cell epitopes and structural disorder, respectively. In-house Perl scripts and Bio::Graphics were used to integrate the data generated from BepiPred and IUPred predictions, as previously described.²³ Peptides with 9-mer and 15-mer having the best prediction scores were selected for the synthesis.

To differentiate mycobacterial infections and select the best markers, a careful assessment of the involved microorganisms was crucial. Among the analyzed bacteria are Mycobacterium tuberculosis, M. bovis, M. africanum, M. microti, M. avium, M. intracellulare, M. kansasii, M. szulgai, M. xenopi, M. marinum, M. ulcerans, M. fortuitum, M. abscessus, M. chelonae, and M. scrofulaceum. Other agents causing skin infections, such as fungi (Tinea corporis − dermatophytosis), Sporothrix schenckii (Sporotrichosis), Paracoccidioides brasiliensis (Paracoccidioidomycosis), and Malassezia furfur (Pityriasis), were considered for differentiation. Additionally, protozoa responsible for skin infections, such as Leishmania sp. (cutaneous and visceral leishmaniasis) were also considered. Multiple alignment of selected targets was performed by the ClustalX 2.0. program,²⁴ using default parameters.²³ Any sequence that could present an overlap (at least 70% similarity along 70% of the length) with those microorganisms was excluded.

2.3. SPOT synthesis

Following the SPOT synthesis technique described in Frank, 1992,²⁵ a total of 102 peptides, including selected and control targets, were synthesized on a derivatized cellulose membrane (Intavis™) using an automated synthesizer ResPep SL (Intavis™). Briefly, peptides were constructed at specific spots using fluorenylmethyloxycarbonyl (Fmoc) protecting amino group for the N-terminus. Amino acid coupling was performed after Fmoc group removal (deprotection step) with 4-methylpiperidine (Sigma-Aldrich™) 25% v/v in dimethylformamide (DMF, Merck™) and carboxy activation with diisopropylcarbodiimide (DIC, Sigma-Aldrich™) and Oxyma Pure (Merck™), both at 1.1 M. Two coupling cycles were performed for each amino acid without intervening deprotection step to ensure coupling and good sequencing. Subsequently, an acetic anhydride solution (3% v/v in DMF) was used to promote the acetylation of any remaining free amine residues. The newly coupled amino acid Fmoc protector group was then removed with 25% v/v 4-methylpiperidine in DMF. These steps were repeated until all amino acids were added. At the end of the synthesis, the membrane was immersed in a cleavage solution of 95% (v/v) trifluoracetic acid (TFA, Sigma-Aldrich™)/2.5% (v/v) water/ 2.5% (v/v) triisopropylsilane (Merck™) for one hour to remove the last Fmoc and also the protecting groups from the amino acid side chains. Then, the membrane was washed 4 times with dichloromethane (DCM), 4 times with DMF, and ethanol twice. The membrane was dried, and the spots were examined under ultraviolet light.

2.4. Dot blotting screening

Pools of sera from patients infected with M. leprae in two different clinical forms were used in dot blotting assays: ten Paucibacillary (PB) – tuberculoid − and twenty Multibacillary (MB) – borderline-tuberculoid, borderline-borderline, borderline lepromatous and lepromatous; ten patients infected with M. tuberculosis presenting Tuberculosis (TB) disease; or the pool of sera from ten healthy individuals. Briefly, the membrane was blocked with 5% BSA and 4% sucrose in PBS for 16 h, washed 3 times for 10 min with 0.1% Tween-20 in PBS, and incubated with one of the above pools of serum diluted 1:100 in PBS-Tween 0.1% for 2 h. The membrane was then washed as described above, incubated with horseradish peroxidase-conjugated anti-human IgG (Sigma-Aldrich) diluted 1:10,000 in PBS/Tween 0.1% for 1 h, washed, and visualized by chemiluminescence with Luminata Forte Western HRP substrate (Merck) exposed for 1 or 10 min on an ImageQuant LAS 4000 digital imaging system (GE Healthcare). After data acquisition, the membrane was regenerated for use with another pool of sera. For regeneration, the membrane was washed three times with DMF for 10 min and then incubated with 8 M Urea, 1% SDS (Sodium Dodecyl Sulfate) solution for 16 h. Subsequently, the membrane was rinsed twice in the 8 M urea, 1% SDS solution for 30 min, once in deionized water for 2 min, and three times in 55% ethanol and 10% acetic acid solution (v/v) for 10 min. Finally, the membrane was washed in deionized water for 2 min.

2.5. Densitometric evaluation of the dot blotting assays

In this step, the reactivity of each membrane-synthesized peptide that was tested with pooled sera in the previous step was determined by calculating the densitometry value of each spot using the ImageJ software and the Protein Array Analyzer plug-in (https://image.bio.methods.free.fr/ImageJ/?Protein-Array-Analyzer-for-ImageJ.html).²⁶ The dot blotting membranes were evaluated individually based on color intensity, with higher color intensity corresponding to increased spot reactivity. The obtained densitometry values are proportional to the color intensity. To ensure optimal image quality, we assessed each membrane three times using the same pool, as some captured images may exhibit significant noise. For each group, we selected the image with the lowest noise interference. Subsequently, the membranes were normalized to one another using the brightest spot as a reference point, thereby adjusting the others. The aforementioned process was facilitated by the software itself through the “Masterize” function, which, after the procedure, generates a Microsoft Excel spreadsheet containing all values. The cut-off was determined based on the mean plus three times the standard deviation of the densitometric values of all spots evaluated on both the negative control and tuberculosis-tested membranes, which were considered here as control membranes. Only spots exhibiting a densitometric value equal to or greater than this threshold were considered reactive on membranes tested with sera from leprosy patients. The sequences of the selected peptides were also searched using BLASTp (Basic Local Alignment Search Tool − NCBI) to determine the characteristics of the proteins to which these peptides belong.

2.6. Synthesis of soluble peptides

A total of eight soluble peptides selected from the previous step were synthesized on a 10 μmol scale using ResPep SL (Intavis™) and named PEP1, PEP2, PEP3, PEP4, PEP5, PEP6, PEP7, and PEP8. The peptides were built on H-Rink Amide ChemMatrix resin (Sigma-Aldrich™). Briefly, Fmoc amino acids were activated with a 1:1 solution of Oxyma Pure (Merck) and DIC (Sigma-Aldrich), both at 1.1 M. Coupling was performed twice after deprotection with 4-methylpiperidine (25% v/v in DMF). A solution of acetic anhydride (3% v/v in DMF) was used for the acetylation of any free amine residues that may have remained. These steps were repeated until the synthesis of each of the peptides had been completed. A cysteine was coupled to the C-terminal to improve peptide ligation to cellulose membranes or immunoassay plates. The peptides were then treated with a solution of 92.5% trifluoroacetic acid, 2.5% water, 2.5% triisopropylsilane, and 2.5% beta-mercaptoethanol for 3 h with agitation to deprotect and release the peptides from the resin. The peptides were precipitated with cold methyl tert-butyl ether (Merck™) and lyophilized.

2.7. IgG detection by Enzyme-Linked immunosorbent assay (ELISA)

Following the synthesis of the peptides, an enzyme-linked immunosorbent assay (ELISA) was conducted to detect IgG antibodies specific to the selected peptides. The assay conditions were established to ensure the optimal antigen concentration and serum dilution. Subsequently, a pool of healthy controls, patients with paucibacillary (PB) leprosy, and patients with multibacillary (MB) leprosy were evaluated. Each of the eight peptides was tested individually to ascertain whether they exhibited the same differential reactivity as observed in the membrane. Subsequently, the ELISA was conducted on individual samples from the leprosy serologic panel, which had been previously classified by clinical and histopathologic examination and constituted the serologic library of the Laboratory of Immunology and Parasite Control/ICB/UFMG and the other partner institutions of this project, as previously described. In this study phase, the peptides were also tested in the sera of individuals with close contact to leprosy patients. Three of the eight peptides (PEP6, PEP7, and PEP8) were evaluated as a single peptide pool, as they are part of the same M. leprae protein. In this study, HSA-NDO, LID-1, and NDO-LID were also evaluated using the same serum bank. These antigens were kindly provided by Fundação Oswaldo Cruz (Fiocruz/Brazil) for research purposes only. The assay parameters were optimized using various antigen and antibody concentrations and comparing the results to select the concentration that yielded the optimal resolution between the negative control and positive samples. The ELISA plates were sensitized for 36 h at 4°C with 1.0 μg/well of soluble peptides/antigens diluted in 50 μL carbonate buffer (15 mM Na₂CO₃; 34 mM NaHCO₃; pH 9.6). After sensitization, the plates were blocked with 200 µL of 5% BSA in 1X PBS for 1 h at 37°C. After incubation, the plates were washed three times with 1X PBS solution containing 0.005% Tween20 (PBS/T). Subsequently, the sera were pipetted into a diluent solution (2.5% BSA in 1X PBS) at a 1:100 ratio and added to the plates in duplicate wells. After incubation for 1 h at 37°C, the plates were washed three times with PBS/T and a secondary anti-human IgG peroxidase-conjugated antibody (Sigma-Aldrich) diluted 1:15,000 in PBS/T was added (50 μL/well). After this period, the plates were washed three times with PBS/T, and 50 μL of reveal solution (0.1M Citrate; 0.2M Na₂PO₄; 0.05% OPD and 0.1% H₂O₂) was added. To stop the reaction, we added 25 μL of 2N H₂SO₄. The plates were then incubated for 15 min in the dark. An automated microplate reader (SpectraMax 340 PC, Molecular Devices) was used to measure the absorbance of the reaction at 492 nm.

2.8. Statistical Analyses

All serum samples were evaluated in duplicate, with the sample result being the mean OD value of these simultaneous determinations. GraphPad Prism 9 for Windows (Graphpad Software, Inc.) was used for all statistics. Shapiro-Wilk and Kolmogorov-Sminorff tests were performed to determine whether a variable had a normal distribution. P-values were determined using the Wilcoxon matched-pair test. All p-values < 0.05 were considered to be significant. A Receiver Operating Characteristic (ROC) curve was generated for each tested peptide/antigen considering the three groups evaluated, grouped as healthy control versus patients and healthy control versus household contacts. The optical density (OD) values were used to calculate the cut-off, area under the ROC curve (AUC), and 95% confidence interval (95% CI). Cut-off values for each peptide/antigen were defined based on the ROC curves, to identify the optimal compromise between sensitivity and specificity.²⁷

3. Results

3.1. Epitopes prediction and cellulose membrane synthesis

A total of 102 peptides were obtained through the prediction step, comprising 57 sequences of 9-mer and 45 sequences of 15-mer. These peptides were synthesized using the SPOT technique on a cellulose membrane. Fig. 1 illustrates the complete procedure.

Fig. 1.

Outline of B-Cell Epitope Prediction and Representative Results. The predicted proteome deposited in the NCBI database was subjected to computational tools, which yielded 102 peptide sequences of 9 and 15 amino acids, along with their respective molecular weights, as exemplified. These were then synthesized within the ResPep by Intavis. The figure was created using MS Office tools.

3.2. Dot blotting and densitometric evaluation

Following the analysis of all dot blotting membranes using ImageJ (Fig. 2), the mean densitometry value for the control membranes was found to be 6,174, with a standard deviation of 4,559. Subsequently, a minimum reactivity cut-off of 19,851 was established. In the membranes tested with sera from leprosy patients, eight peptides exhibited densitometric values above 19,851. Thus, these eight peptides, named PEP1, PEP2, PEP3, PEP4, PEP5, PEP6, PEP7, and PEP8, were selected for soluble synthesis to be used in ELISA tests with samples from the serum bank.

Fig. 2.

Membranes were tested with sera from different groups and analyzed using ImageJ. The membranes with the lowest noise interference were processed with ImageJ and normalized to each other. The color scale is located at the top of each membrane, with the least reactive (black) on the left and the most reactive (white) on the right. From left to right, the membranes correspond to those tested by a negative control pool (A), a pool of patients with tuberculosis disease (B), and pools of patients diagnosed with leprosy in the PB (C) and MB (D) forms, respectively. The spots in column 1 (C1), rows L1 to L15, and the spots in column 6 (C6), rows L19 and L20, do not contain peptides. Selected peptides and their respective best densitometric values in patients’ membranes are C2L20 (PEP1 – 38,643), C3L8 (PEP2 – 39,321), C3L18 (PEP3 – 57,352), C4L1 (PEP4 – 72,934), C4L10 (PEP5 – 37,079), C5L10 (PEP6 – 31,051), C6L14 (PEP7 – 33,633), C6L16 (PEP8 – 35,463).

PEP4 was chosen for further analysis despite its reactivity with negative sera in 2 out of 3 replicates. This decision was based on its strong performance in membranes tested with MB and PB sera. In the MB membrane, PEP4 showed a reactivity of 80,243, nearly 4 times higher than the cut-off (19,851) and 2.3 times higher than its value in the negative control membrane (34,189). Similarly, in the PB membrane, its reactivity was 72,934, approximately 3.7 times higher than the cut-off and 2.1 times higher than its value in the negative control membrane.

The M. leprae proteins to which the peptides belong were given by the prediction, as well as their molecular weights (Fig. 3). The BLASTp search confirmed the M. leprae proteins to which the peptides belong. PEP6, PEP7, and PEP8 belong to the same protein. The BLAST RID numbers can also be observed in Fig. 3.

Fig. 3.

Logo representation of the sequences of the eight selected peptides. The theoretical molecular weight (Da) and sequence of each peptide are displayed, with each colour representing a physicochemical property. The figure was created using the ggseqlogo package in R software and MS Office tools.

We then proceeded to the next step, as the soluble synthesized peptides showed a similar reactivity pattern to the cellulose membranes (data not shown).

3.3. Assessing IgG using ELISA

Fig. 4 presents the comparison between the groups for the new peptides. Statistical differences were observed between all groups, with the average OD being higher in the household contact individuals compared to those individuals previously diagnosed with leprosy. No sequence corresponding to the selected peptides was found in the previously described immunodominant proteins for M. leprae.

Fig. 4.

Graphs depicting the results of optical densitometry in ELISA assays conducted to test reactivity for immunoglobulin G (IgG) in sera from the following groups: healthy control subjects (Controls), household contacts of index patients (HHC), and patients previously diagnosed with leprosy (Patients). The mean values found are as follows: PEP1 – Controls: 0.1620; Contacts: 0.5711; Patients: 0.2678. PEP2 – Controls: 0.1353; Contacts: 0.4905; Patients: 0.2284. PEP3 – Controls: 0.1117; Contacts: 0.5645; Patients: 0.2399. PEP4 – Controls: 0.1311; Contacts: 0.7295; Patients: 0.2728. PEP5 – Controls: 0.1131; Contacts: 0.4548; Patients: 0.2288. POOL (PEP6, PEP7, PEP8) – Controls: 0.0945; Contacts: 0.2973; Patients: 0.1304. The distribution is not normal, and the statistical tests used were Kruskal-Wallis and ANOVA. Differences were considered statistically significant if the p-value was equal to or less than 0.05. * = 0.0199; **** < 0.0001.

The ROC curves showed good results in terms of area under the curve (AUC), especially for PEP3, PEP4, and PEP5, with PEP3 showing the best results, presenting values above 0.8 in both the comparison of healthy controls with patients (Table 1) and the comparison of the same controls with household contacts (Table 2). On the other hand, the new peptides showed better results when compared to the previously well-described antigens: HSA-NDO, LID-1, and NDO-LID (Table 1, Table 2).

Table 1.

Sensitivity, specificity, and accuracy percentages of immunoenzymatic assays for peptides (PEP1, PEP2, PEP3, PEP4, PEP5, PEP6, PEP7, PEP8) and previously described antigens (HSA-NDO, LID-1, NDO-LID) tested with human sera from patients who had been diagnosed with leprosy and were under treatment. AUC = area under the curve. CI = confidence interval. Values are related to the responses in the control sera.

	cut-off	AUC	Sensitivity (%)	95 CI (%)	Specificity (%)	95 CI (%)
PEP1	>0.1698	0.7121	70.33	63.64 to 76.44	70.59	58.29 to 81.02
PEP2	> 0.1433	0.7142	65.37	58.42 to 71.86	70.59	58.29 to 81.02
PEP3	> 0.1435	0.8257	74.38	67.80 to 80.24	75.36	63.51 to 84.95
PEP4	> 0.1436	0.7945	74.38	67.80 to 80.24	73.53	61.43 to 83.50
PEP5	> 0.1312	0.7961	75.48	69.05 to 81.17	75.00	63.02 to 84.71
PEP6,7,8	> 0.0925	0.6214	59.02	51.96 to 65.83	58.57	46.17 to 70.23
HSA-NDO	> 0.3464	0.6845	65.99	59.13 to 72.24	66.18	54.34 to 76.29
LID-1	> 0.5685	0.5930	56.89	50.36 to 63.19	55.88	44.08 to 67.05
NDO-LID	> 0.4103	0.6565	61.90	55.18 to 68.20	58.82	46.96 to 69.74

Table 2.

Sensitivity, specificity, and accuracy percentages of immunoenzymatic assays for peptides (PEP1, PEP2, PEP3, PEP4, PEP5, PEP6, PEP7, PEP8) and previously described antigens (HSA-NDO, LID-1, NDO-LID) tested with human sera from household contacts of patients who had been diagnosed with leprosy. AUC = area under the curve. CI = confidence interval. Values are related to the responses in the control sera.

	cut-off	AUC	Sensitivity (%)	95 CI (%)	Specificity (%)	95 CI (%)
PEP1	> 0.2780	0.9500	89.74	83.88  to 94.02	89.71	79.93 to 95.76
PEP2	> 0.2260	0.9500	88.34	82.4 to 92.83	88.24	78.13 to 94.78
PEP3	> 0.2225	0.9759	92.64	87.49  to 96.14	91.30	82.03 to 96.74
PEP4	> 0.2498	0.9796	92.55	87.34 to 96.09	92.65	83.67 to 97.57
PEP5	> 0.1965	0.9551	88.75	82.8 to 93.19	86.76	76.36 to 93.77
PEP6,7,8	> 0.1397	0.9366	86.45	80.04 to 91.41	84.29	73.62 to 91.89
HSA-NDO	> 0.3695	0.6238	67.11	59.30 to 74.07	63.24	51.36 to 73.70
LID-1	> 0.5017	0.6773	62.82	55.01 to 70.01	63.24	51.36 to 73.70
NDO-LID	> 0.4725	0.7254	67.90	60.37 to 74.60	66.18	54.34 to 76.29

4. Discussion

The prediction of B-cell epitopes represents a strategy that provides advantageous peptide sequences for incorporation into immunological constructions, including diagnostic and follow-up tools. Synthetic peptides have been identified as a potential means of identifying biomarkers for neglected diseases. Nevertheless, it has been demonstrated that the utilization of multiple peptides is essential to attain optimal specificity and sensitivity.²⁸

Our study predicted M. leprae B-cell epitopes that have not yet been described in the scientific literature, and some of the analyzed peptides have proven to be good candidates for indicating exposure to M. leprae in human sera (Fig. 2). While small peptides are typically investigated for their capacity to recognize T cells, there have been studies that have demonstrated their immunodominance for B cells as well.29, 30, 31 The peptides selected for analysis range in length from 9-mer to 15-mer, as illustrated in Fig. 3. This renders them more cost-effective and facilitates a more expeditious acquisition process. Nevertheless, the smaller peptides demonstrated significant reactivity with antibodies in a greater number of samples than the larger peptides (Fig. 4).

The patient and household contact groups exhibited significantly elevated reactivity to all peptides in comparison to the healthy control group (Fig. 4). This indicates that these peptides are specific for the detection of anti-M. leprae antibodies and may prove useful in the development of strategies for combating leprosy. The identification of individuals who are more susceptible to transmitting the disease can facilitate the disruption of the infection chain. It is important to note that individuals with high antibody levels also tend to have a high bacterial load.32, 33 This underscores the potential utility of antibody detection to M. leprae antigens as a means of identifying infections at an early stage and preventing bacterial dissemination.³⁴

The response of household contacts to our peptides was unexpected (Fig. 4). It is known that the host's metabolism at the time of infection, in conjunction with the immune response elicited by the host, which is shaped by genetic factors, plays a pivotal role in combating M. leprae.35, 36 Thus, the contacts in our study may exhibit, for instance, increased vitamin D levels and enhanced T lymphocyte activation, with a greater presence of M1 macrophages. This could lead to improved antigen presentation of M. leprae, particularly antigens that are not cell wall-associated, such as PEP1-PEP5, which do not belong to cell membrane proteins. Given that leprosy contacts are often considered to have a subclinical infection, which implies that they can be asymptomatic individuals, and that they can be considered to have a highly similar response to PB patients, the observation that IgG levels are lower in patients undergoing treatment compared to household contacts also can lends support to previous findings that patients receiving MDT have lower IgG antibody levels than individuals who are not receiving treatment.32, 36, 37, 38 Further research is needed to understand better the genetic factors and the type of immune response assembled by this group of infected but asymptomatic individuals, since it is a less studied group, although important to understand which factors can lead to disease establishment.

The PEP3, PEP4, and PEP5 peptides demonstrated the highest reactivity values in sera from groups exposed to M. leprae (Fig. 4), yielding superior results compared to previously described antigens in the literature (Table 1, Table 2). This finding lends further support to the hypothesis that small peptides can also exhibit robust immunodominance for B cells.³¹ Furthermore, these three peptides exhibit a higher proportion of hydrophobic amino acid residues (Fig. 3), which challenges the hypothesis proposed by some authors that a correlation exists between hydrophilicity and antigenicity.39, 40 PEP1 and PEP2 also demonstrated satisfactory reactivity, although the results were considerably lower than those of PEP3, PEP4, and PEP5.

The superior sensitivity, specificity, and area under the ROC curve demonstrated by the study's predicted peptides when compared to HSA-NDO, LID-1, and NDO-LID (Table 1, Table 2) reinforce the notion that these five synthetic peptides have the potential to be utilized in diagnostic applications and can facilitate the enhancement of existing tests, particularly for M. leprae, which has limited cultivation capabilities. Nevertheless, further testing of these peptides with recently diagnosed patients who are still not undergoing treatment is required to gain a more comprehensive understanding of the data. Although the patients were undergoing treatment, only HSA-NDO demonstrated a superior AUC value for patients (Table 1), confirming the accuracy of PGL-1 in detecting this group, as HSA-NDO is a synthetic form of PGL-1.⁴¹ The rapid serological commercial test adopted by the Brazilian Public Health System is the ML Flow test (Quibasa-Bioclin, Brazil), which utilizes PGL-1 for the detection of IgM. According to the manufacturer, the test demonstrates 90.8% sensitivity (95% CI: 86.11–95.57%) and 98.0% specificity (95% CI: 95.42–100%) for patients. While these data suggest superior performance compared to our peptides, a direct comparison cannot be made, as the commercial test targets IgM, whereas our focus is on IgG. Notably, regarding the HHC group and the IgG response, our findings represent a novel approach with highly promising results, particularly with the use of PEP3 and PEP4.

Despite the presence of a more significant number of amino acid residues, the majority of which possess hydrophilic properties, PEP6,7,8 did not yield favorable results in the comparison between patients and healthy controls (Fig. 4). Furthermore, it demonstrated reduced sensitivity and specificity in comparison to HSA-NDO and NDO-LID for this specific cohort of leprosy patients (Table 1). However, for the contacts group, it demonstrated superior reactivity, sensitivity, and specificity compared to HSA-NDO, LID-1, and NDO-LID, approaching the values of the other predicted peptides.

A systematic review revealed that researchers who utilized LID-1 identified sensitivity ranging from 0.43 to 0.64 for patients and employing HSA-NDO, sensitivity of 0.64.⁴² In another study, LID-1 and NDO-LID exhibited sensitivity levels of 6% and 33%, respectively, for active new cases, and 63% and 77%, respectively, for cases diagnosed in a reference center.⁴³ These findings are consistent with our results (Table 1), thereby confirming their diagnostic potential for patients. HHC are less extensively studied in terms of serological positivity. In a study involving 108 HHC in China, LID-1 exhibited approximately 23% positivity for this group, while NDO-LID demonstrated approximately 15%.¹⁶ In another study conducted in Brazil, LID-1, NDO-LID, and HSA-NDO exhibited approximately 3%, 6%, and 14% positivity, respectively.⁴⁴ These findings suggest that serological testing may exhibit discrepancies in this group, potentially attributable to genetic factors. Notably, our study's findings, derived from a more extensive sample size and a distinct population, exhibited enhanced outcomes compared to previous reports.

PEP6, PEP7 and PEP8 are peptides that differ by only one or two residues. As these peptides are linear B-cell epitopes and originate from the same protein, it can be hypothesized that this overlap occurred due to the limitations of our epitope prediction, which was restricted to 9-mer or 15-mer lengths. A prediction allowing for epitopes up to 20-mer in length could potentially identify a single 18-mer epitope with the following sequence: NTPTSSTSSSPTSSVPEG.

It is evident that all eight peptides exhibit minimal probability of constituting a shared epitope region within other pathogenic microorganisms described in the methods section, as any sequence demonstrating at least 70% similarity with these microorganisms was excluded. Moreover, the BLASTp search did not reveal any similarity with microorganisms known to be pathogenic to humans. The efficacy of this strategy is demonstrated in Fig. 2B, which illustrates that the 102 predicted peptides did not react with tuberculosis sera, the most prevalent human disease, which pathogenic agent belongs to the same family of M. leprae. Nevertheless, they may eventually share the region with some new pathogens, which would confer cross-reactivity to sera tests with these peptides, resulting in a false positive assay.35, 36, 37, 38 Our results are therefore promising, presenting a potential for developing novel immunological screening methods for contacts and contributing to enhancing epidemiological surveillance efforts in leprosy control.

5. Conclusion

In conclusion, the prediction of B-cell epitopes represents a promising strategy for identifying peptide sequences with beneficial properties for immunological applications, such as diagnostic tools. The present study successfully predicted M. leprae B-cell epitopes, with several peptides demonstrating robust reactivity and specificity in identifying anti-M. leprae IgG. This was demonstrated by dot blotting and ELISA techniques, particularly with the peptides PEP3, PEP4, and PEP5. The high reactivity observed in patient and household contact groups, in comparison to healthy controls, highlights the potential of these peptides in identifying individuals at risk of transmitting leprosy, thereby aiding in infection control. The findings also underscore the effectiveness of smaller peptides in improving specificity and sensitivity, while their cost-effectiveness and rapid production further enhance their applicability. The peptides showed no significant cross-reactivity with tuberculosis, reinforcing their potential as reliable candidates for serological testing. Further testing with other immunoglobulins and sera from recently diagnosed patients is required to validate these results and optimize their use in the early detection and prevention of infection.

Funding Information

This work was supported by Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG − PPSUS – Edital APQ – 4035/17; FAPEMIG BPD-00394–22). ACPJ is supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) and ALGO is supported by Conselho Nacional de Pesquisa (CNPq). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

CRediT authorship contribution statement

Augusto César Parreiras de Jesus: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Conceptualization. Vanêssa Gomes Fraga: Methodology, Investigation, Data curation. Samuel Alexandre Pimenta-Carvalho: Software, Formal analysis, Data curation. Tania Mara Pinto Dabés Guimarães: Investigation. Marcio Sobreira Silva Araújo: Investigation. Jairo Campos de Carvalho: Investigation. Marcio Bezerra Santos: Investigation. Marcelo Grossi Araújo: Investigation. Marcelo Antonio Pascoal-Xavier: Investigation. Sandra Lyon: Investigation. Sebastião Rodrigo Ferreira: Writing – review & editing, Investigation. Rocio Arreguin-Campos: Writing – review & editing. Kasper Eersels: Visualization. Bart van Grinsven: Writing – review & editing. Thomas Cleij: Visualization, Funding acquisition. Lilian Lacerda Bueno: Resources, Funding acquisition. Daniella Castanheira Bartholomeu: Software, Data curation. Cristiane Alves da Silva Menezes: Writing – review & editing, Validation, Supervision, Resources. Ana Laura Grossi de Oliveira: Writing – review & editing, Writing – original draft, Visualization, Validation, Project administration, Methodology, Investigation, Formal analysis, Conceptualization. Ricardo Toshio Fujiwara: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.