This study was submitted and approved by the Ethics Committee from Universidade Federal de Juiz de Fora—UFJF (Research Protocol CAAE #56863016.6.1001.5147). All participants or their next of kin read and signed a free and informed consent form before inclusion in the study. Parents/guardians provided consent on behalf of minor participants.

The present study is an exploratory observational investigation carried out in Governador Valadares Municipality, Minas Gerais, Brazil, considered a hyperendemic area for leprosy (1.9 cases/10,000 inhabitants), based on the overall prevalence observed in Minas Gerais (0.5 cases/10,000 inhabitants) and Brazil (1.2 cases/10,000 inhabitants) (1). The participants were contacted at the Reference Center for Endemic Diseases and Special Programs (CREDEN-PES) outpatient unit, Department of Public Health, Governador Valadares, MG, Brazil.

This study comprises a cross-sectional investigation, named Leprosy – Brazil Project, that enrolled a total of 257 participants of both genders (52% males and 48% females), ages ranging from 5 to 85 years old, selected as a non-probabilistic convenience sampling, encompassing: 91 household contacts of leprosy patients (HHC, 38% males and 62% females, median age = 33 years old, ranging from 5 to 85 years old); 79 patients (L, 56% males and 44% females, median age = 47 years old, ranging from 6 to 80 years old) and 87 endemic healthy controls (EC, 52% males and 48% females, median age = 46 years old, ranging from 10 to 83 years old). According to the Brazilian Guidelines (16), the operational classification of leprosy cases is based on the number of skin lesions, considering the following criteria: Paucibacillary (PB) – cases presenting up to five lesions; Multibacillary (MB) – cases with more than five skin lesions. Although, positive baciloscopic index (BI≠0) classifies the cases as MB, negative baciloscopic index (BI=0) does not necessarily classify the patient as PB, as it depends on the number of lesions the case presents. Moreover, the number of lesions lower than five does not necessarily classify the case as PB, since cases with lesion number up to five, but presenting baciloscopic index (BI≠0) is classified as MB. Operational classification records were available for 88 out of 91 (96%) HHC. The bacilloscopic index values (BI=0) for index cases of HHC(PB) and number of lesions (>5) for index cases of HHC(MB) were obtained from medical records. Based on the operational classification records, the HHC subjects (n=88 out of 91) and all L patients (n=79) were further classified into subgroups, which reflect the degree of infection in patients and the level of exposure of their contacts. In our study these subgroups are referred as to: HHC(PB), (n=20) – household contacts of paucibacillary leprosy patients; HHC(MB), (n=78) – household contacts of multibacillary leprosy patients; L(PB), (n=23) – paucibacillary leprosy patients; L(MB), (n=56) – multibacillary leprosy patients. All patients were untreated and did not present reactions at the time of sample collection, ensuring that the immune response observed was not influenced by ongoing leprosy treatment. For operational purpose, household contacts (HHC) are defined as an individual who resides or has resided, or who lives or has lived with a leprosy patient within the household, during the five years prior to the diagnosis of the disease, whether a family member or not (16). Table 1 provides the study population’s major demographic features and clinical laboratory records. Figure 1 summarizes the compendium of study design, population & methods.

Heparinized whole blood samples (10mL) were obtained from each participant and used to isolate peripheral blood mononuclear cells (PBMC) for in vitro cultures and further quantification of chemokine and cytokine in culture supernatant. PBMC were separated by Ficoll Hypaque cushion gradient, washed twice, and resuspended in RPMI-1640, supplemented with 10% fetal bovine serum, 2mM L-glutamine, penicillin (100U/mL) and streptomycin (100mg/mL). Cell aliquots (2x10⁵/well) were cultured in duplicates on flat-bottom 96-well plates at 37°C, 5% of CO₂ in a humidified incubator. Parallel culture batches were carried out without exogenous stimuli (unstimulated culture) and in the presence of 10 irradiation-killed M. leprae bacilli/cell (M. leprae-stimulated culture). The M. leprae bacilli were prepared at Instituto Lauro de Souza Lima (São Paulo, Brazil), as previously described (17). Culture supernatants were harvested after five days, centrifugated to remove cells/debris, aliquoted, and stored at -80°C until processing for soluble mediator analysis. We selected day 5 for supernatant collection based on prior studies showing this time point best captures peak cytokine and chemokine production in PBMC cultures stimulated with antigen, as it is the most representative for detecting immune responses to M. leprae (14).

Quantitative analysis of chemokines and cytokines in cell culture supernatants was performed by Cytometric Bead Array (Human Chemokine Kit for CXCL8, CCL2, CXCL9, CCL5 and CXCL10 and Human Cytokine Flex Set kit for IL-6, TNF, IFN-γ, IL-17, IL-4, IL-10 and IL-2) purchased from BD Bioscience Pharmingen (San Diego, CA, USA) according to the manufacturer instructions. 400 events/chemokine or cytokine-specific beads were acquired on a FACSVerse™ Bioanalyzer (BD Bioscience, San Jose, CA, USA). Data was analyzed using the FCAP Array software (Soft Flow, Inc., St. Louis Park, MN, United States). The results were expressed in pg/mL based on the standard curves using the 5-parameter logistic regression model.

Statistical analysis was carried out using the Prism GraphPad software (version 10.2.0, San Diego, California, USA). Considering the non-parametric distribution of data sets, comparative analysis amongst groups was carried out by the Kruskal-Wallis test followed by Dunn’s post-test for multiple comparisons amongst EC vs. HHC vs. L as well as amongst EC vs. HHC(PB) vs. HHC(MB) vs. L(PB) vs. L(PB) subgroups. In all cases, significant differences were considered at p<0.05.

Additional descriptive analysis of soluble mediators, referred to as signatures, was carried out as previously reported by Cunha et al. (14) and modified as follows: the intrinsic median values for each cytokine and chemokine were derived based on their distribution under specific culture conditions (unstimulated culture or M. leprae-stimulated culture). Thus, two distinct medians were obtained for each biomarker, depending on the stimulus, providing a robust baseline for classifying the subjects into low and high expression groups. Continuous variables expressed in pg/mL were converted into categorical data using the intrinsic median values of unstimulated culture (CXCL8 = 6,194; CCL2 = 3,669; CXCL9 = 778; CCL5 = 707; CXCL10 = 1,351; IL-6 = 10,135; TNF = 8.1; IFN-γ = 560; IL-17 = 7.2; IL-4 = 0.8; IL-10 = 42 and IL-2 = 13 pg/mL) or M. leprae stimulated culture (CXCL8 = 5,460; CCL2 = 3,518; CXCL9 = 852; CCL5 = 641; CXCL10 = 1,248; IL-6 = 10,057; TNF = 8.2; IFN-γ = 456; IL-17 = 6.3; IL-4 = 1.0; IL-10 = 45 and IL-2 = 16 pg/mL) as the cut-off to identify volunteers with low and high levels of soluble mediators. Following, the proportion of subjects (%) with chemokine and cytokine levels above the intrinsic cut-off was calculated for each study group or subgroup. The chemokines and cytokines with the proportion of subjects above the 50^th percentile was underscored and considered as increased levels. The chemokines and cytokines signatures were assembled as ascendant signatures, and Venn diagrams were constructed to identify common and selective soluble mediators amongst groups.

Integrative networks of chemokines and cytokines were built based on correlation analysis (Spearman rank tests) between pairs of soluble mediators. Only significant correlations (p<0.05) were employed to construct networks (chemokines and cytokines) using the open-source Cytoscape software (available at https://cytoscape.org). The networks were assembled using cluster layouts with nodes representing each chemokine and cytokine and connecting lines identifying positive (“r” scores >0, continuous line) or negative (“r” scores <0, dashed line) correlations. The node sizes are proportional to the correlations between pairs of soluble mediators. Comparative analysis amongst groups was carried out considering the total number of correlations. Colormap analysis of integrative networks was used to illustrate the comparison amongst groups using a color key based on the percentile distribution (10^th/50^th/90^th) of correlation numbers calculated for soluble mediator, chemokine, and cytokine clusters or the total number of correlations.

An additional analysis was performed to summarize the significant changes observed in soluble mediators amongst groups. For this purpose, the fold changes in chemokines and cytokines concentrations in PBMC culture supernatants from HHC and L were calculated using the relative ratio to EC (Fold Change in concentration = HHC/EC or L/EC). To accomplish the comparative analysis of HHC and L subgroups according to operational classification, fold changes were calculated about PB subjects [Fold Change in concentration HHC(MB)/HHC(PB) or L(MB)/L(PB)]. Log10 of Fold Change values (magnitude) were combined with the -Log10 p values (significance) obtained by Dunn’s post-test for multiple comparisons.

Differences in chemokines and cytokines signatures were assessed for EC and PB subjects [Delta in % = %HHC – %EC or %L – % EC; %HHC(MB) – %HHC(PB) or %L(MB) – %L(PB)]. The Log10 of Delta values (magnitude) was combined with the -Log10 p values (significance) obtained by Chi-square comparisons.

Differences in chemokines and cytokines network correlations were assessed according to EC and PB subjects [Delta in correlation numbers = HHC – EC or L – EC; HHC(MB) – HHC(PB) or L(MB) – L(PB)]. Delta values (magnitude) were combined with the % of correlations from EC or PB (threshold = decrease or increase > 33%).

In order to characterize the soluble mediator profile of PBMC from EC, HHC, and L, the levels of chemokines (CXCL8, CCL2, CXCL9, CCL5, CXCL10) and cytokines (IL-6, TNF, IFN-γ, IL-17, IL-4, IL-10, and IL-2) were measured in the supernatant collected upon in vitro culture. Figure 2 illustrates the analysis of soluble mediators, expressed as the concentration of each biomarker in pg/mL, treated as continuous variables. Data from cultures without exogenous stimuli (unstimulated cultures) revealed that household contacts (HHC) exhibited lower levels of CXCL8 and CCL2 but higher production of IL-6, TNF, IL-17, and IL-10 when compared to endemic controls (EC). Additionally, the leprosy patients (L) group showed elevated TNF levels, but reduced secretion of CCL5, IFN-γ, and IL-4, compared to EC ( Figure 2A ).

In M. leprae-stimulated cultures, HHC maintained lower production of CXCL8 and CCL2 while displaying higher levels of TNF and IL-10, with a reduction in IL-4 production, compared to EC. Conversely, the L group showed consistently higher TNF levels and decreased production of IFN-γ and IL-4, along with lower levels of CCL5 and IL-2, and elevated IL-10 levels, when compared to EC ( Figure 2B ).

The soluble mediators measured in the supernatant from in vitro cultured PBMC were further assessed as chemokines and cytokines signatures. For this purpose, the levels of soluble mediators expressed in pg/mL were converted into categorical data as described in Methods, and the results are shown in Figure 3 . Data analysis of the proportion of subjects above the 50^th cut-off allowed the identification of clusters of soluble mediators featuring each study group ( Figure 3A ). A Venn Diagram analysis identified common and selective soluble mediators amongst EC, HHC, and L groups ( Figure 3B ). The results from Unstimulated cultures demonstrated that the production of CCL2 and IL-2 was the hallmark of EC. The hallmark of HHC was the increased production of IL-17 and IL-10, while TNF production represents discriminant attributes observed in L. The signature analysis of categorical data further identified CXCL9 as a selective attribute of the L group ( Figure 3B , Left panel).

The Venn diagram analysis of M. leprae-stimulated cultures reveals IL-17, IL-4, and IL-2 as selective attributes for EC, IFN-γ for HHC and CXCL9, CXCL10, and TNF for the L group ( Figure 3B ).

These findings underscore that upon M. leprae-stimuli, IFN-γ-upregulation is observed in household contacts, and CXCL10 is observed in leprosy patients.

Integrative networks were assembled based on the correlation profile between pairs of soluble mediators to investigate further the interplay amongst chemokines and cytokines in the PBMC culture microenvironment ( Figure 4 ). Data analysis showed that, regardless of the culture condition, the number of correlations displayed a decrease in HHC and L as compared to EC. Moreover, the M. leprae-stimuli led to a decrease in the number of correlations between soluble mediators as compared to the Unstimulated culture (EC = 68 → 62; HHC = 56 → 50 and L = 46 → 42) ( Figure 4A ). Color map analysis further illustrated that downregulation was universally observed in the total number of correlations, intra-cluster connectivity, and the analysis of single patterns of most soluble mediators ( Figure 4B ). Detailed description of correlations between pairs of chemokines and cytokines together with the Spearman rank “r” scores are provided in the Supplementary Figure 1 .

Figure 5 provides a snapshot of the significant changes observed in the chemokine and cytokine profile elicited by the in vitro culture of PBMCs. Distinct approaches were employed to scrutinize details of soluble mediator settings in HHC and L groups. For this purpose, the fold change in concentration magnitude, along with shifts in the soluble mediator signatures and integrative networks, were assessed according to EC.

The analysis of fold changes in soluble mediator concentrations demonstrated that regardless of the culture condition, decreased levels of CXCL8 and CCL2 and increased levels of TNF and IL-10 were observed for HHC. The M. leprae-stimuli downregulated the levels of IL-6 and IL-17. In the L group, consistent decrease of CCL5 and IL-4 was observed despite the culture condition. At the same time, IL-10 is observed in L patients only upon M. leprae stimuli ( Figure 5 , top panels).

Data from the soluble mediator signatures further corroborates the loss of IL-6 and IL-17 in HHC upon antigen stimulation. The analysis of L group brings about the insight that decreased levels of CCL5 and IL-4 are consistently observed in L regardless of the culture condition ( Figure 5 , middle panels).

The scrutiny of integrative networks supports the overall downregulation in the correlation numbers between soluble mediators in HHC and L as compared to EC ( Figure 5 , bottom panels).

An integrative analysis of changes in chemokines and cytokines highlights distinct immune response profiles that differentiate HHC and L from EC ( Supplementary Figure 2 ). Key biomarkers emerge, defining the common and selective immunological landscapes of each group. An increase in TNF stands out as common marker of inflammation in HHC group across all culture conditions. Conversely, a decrease of CCL5 and IL-4 appears prominently in the L group, regardless the culture condition ( Supplementary Figure 2 ).

To further characterize the chemokine and cytokine profiles of PBMCs from distinct subgroups of HHC and leprosy patients (L), subjects were stratified based on operational classification into the following groups: HHC(PB), HHC(MB), L(PB), and L(MB). The results are presented in Figure 6 .

In unstimulated cultures, comparative analysis revealed that HHC(MB) exhibited higher levels of IFN-γ and lower levels of CXCL8 compared to HHC(PB). Conversely, L(PB) showed elevated levels of IL-17 and IL-10, alongside reduced levels of CCL5, when compared to L(MB) ( Figure 6A ).

In M. leprae-stimulated cultures, HHC(PB) displayed higher levels of CXCL9, TNF, and IL-17, along with lower IFN-γ levels, while L(MB) demonstrated lower levels of CXCL8, CCL2, and IL-4 ( Figure 6B ).

The chemokine and cytokine signatures were assembled for HHC and L subgroups based on categorical data transformation using the specific cut-offs to calculate the proportion of subjects with high levels of soluble mediators as described in the methods. The results are presented in Figure 7 . Data analysis pointed out a set of soluble mediators with the proportion of subjects above the 50^th percentile ( Figure 7A , colored font). Further analysis identified intra-group common and selective soluble mediators for distinct study subgroups ( Figure 7B ). Data from unstimulated cultures revealed that IL-17 levels were elevated in HHC, independent of the operational classification, whereas CCL5 and IFN-γ were selectively increased in HHC(MB). Additionally, L subgroups showed higher levels of CXCL8 despite the operational classification. Comparative analyses of ascendant signatures from M. leprae-stimulated cultures revealed that CCL5 was a standard feature for HHC, while IL-17 and IL-4 were selectively increased in HHC(PB). Moreover, IFN-γ remained a selective HHC(MB) attribute. Data analysis also demonstrates that TNF was increased in both L subgroups ( Figure 7B ).

Integrative networks were assembled to evaluate the interactive profile between chemokines and cytokines in the PBMC culture supernatant based on the correlation interplay between pairs of soluble mediators. The results are presented in Figure 8 . Data analysis demonstrated that in Unstimulated cultures, the number of correlations is higher in Multibacillary subjects as compared to Paucibacillary, [HHC(MB) = 50; HHC(PB) = 36 and L(MB) = 40 and L(PB) = 18] ( Figure 8A ). In the M. leprae-stimulated cultures, the number of correlations remains higher in L(MB) as compared to L(BP) [L(MB) = 44 and L(PB) = 36] but is lower in HHC(MB) as compared to HHC(PB) [HHC(MB) = 36; HHC(PB) = 48] ( Figure 8A ). Color map analysis further corroborates these findings, as observed in the total number of correlations, the intra-cluster connectivity, and the analysis of single patterns of soluble mediators ( Figure 8B ). Detailed description of correlations between pairs of chemokines and cytokines together with the Spearman rank “r” scores are provided in the Supplementary Figure 1 .

The classification of household contacts (HHC) and leprosy patients (L) based on operational classification (paucibacillary [PB] and multibacillary [MB]) reveals distinct immune signatures that highlight key differences in the cytokine and chemokine profiles. These analysis consistently underscore the critical roles of specific biomarkers in shaping the immune response to M. leprae infection, with notable variations between PB and MB subgroups.

The Figure 9 summarizes the main changes identified in the chemokine and cytokine profile of HHC and L upon classification into subgroups according to operational classification. The analysis of fold changes and shifts of soluble mediator signatures and integrative networks was performed in Multibacillary HHC and L subgroups according to Paucibacillary subjects.

The analysis of fold changes indicated that, regardless of the culture condition, increased levels of IFN-γ were observed in HHC(MB). Conversely, the L(MB) subgroup exhibited elevated CCL5 levels, independent of the culture condition. ( Figure 9 – top panels).

Data from soluble mediator signatures further corroborated the increase of IFN-γ in HHC(MB) regardless of the culture condition. Data from the L(MB) subgroup confirmed the increased levels of CCL5 despite the culture condition ( Figure 9 , middle panels).

The analysis of correlation networks demonstrated that in unstimulated conditions, the correlation numbers between soluble mediators in MB subgroups increase compared to PB. However, upon M. leprae stimuli, while a reduction in the number of correlations was observed in HHC(MB), an increase in correlations was found for L(MB) ( Figure 9 , bottom panels).

An integrative analysis of changes in chemokines and cytokines highlights distinct immune response profiles that differentiate HHC(MB) and L(MB) from PB subgroups ( Supplementary Figure 2 ). Key biomarkers emerge, defining the common and selective immunological landscapes of each group. An increase in TNF stands out as common marker of inflammation in HHC group across all culture conditions. Conversely, a decrease of CCL5 and IL-4 appears prominently in the L group, regardless the culture condition ( Supplementary Figure 2 ).