Dynamics of Th1/Th17 responses and antimicrobial pathways in leprosy skin lesions

Abstract

BACKGROUNDReversal reactions (RRs) in leprosy are acute immune episodes marked by inflammation and bacterial clearance, offering a model to study the dynamics of host responses to Mycobacterium leprae. These episodes are often severe and difficult to treat, frequently progressing to permanent disabilities. We aimed to characterize the immune mechanisms and identify antimicrobial effectors during RRs.METHODSWe performed RNA-Seq on paired skin biopsy specimens collected from 9 patients with leprosy before and at RR diagnosis, followed by differential gene expression and functional analysis. A machine-learning classifier was applied to predict membrane-permeabilizing proteins. Antimicrobial activity was assessed in M. leprae-infected macrophages and axenic cultures.RESULTSIn the paired pre-RR and RR biopsy specimens, a 64-gene antimicrobial response signature was upregulated during RR and correlated with reduced M. leprae burden. Predicted upstream regulators included IL-1β, TNF, IFN-γ, and IL-17, indicating activation of both the Th1 and Th17 pathways. A machine-learning classifier identified 28 genes with predicted membrane-permeabilizing antimicrobial activity, including S100A8. Four proteins (S100A7, S100A8, CCL17, and CCL19) demonstrated antimicrobial activity against M. leprae in vitro. Scanning electron microscopy revealed membrane damage in bacteria exposed to these proteins.CONCLUSIONRR is associated with a robust antimicrobial gene program regulated by Th1 and Th17 cytokines. We identified potentially novel host antimicrobial effectors that showed activity against M. leprae, suggesting potential strategies to bolster Th1 and Th17 responses for combating intracellular mycobacterial infections.FUNDINGNIH grants R01 AI022553, R01 AR040312, R01 AR073252, R01 AI166313, R01 AI169526, P50 AR080594, and 4R37 AI052453-21 and National Science Foundation (NSF) grant DMR2325840.

Keywords: Adaptive immunity; Bacterial infections; Immunology; Infectious disease; Innate immunity.

Figure 1. Differential gene expression analysis of RR versus pre-RR groups.

(A) Heatmap displaying expression z scores for the 404 differentially expressed genes (Padj < 0.3) in RR versus pre-RR specimens, representing high (red) and low (light blue) expression levels. Samples were clustered using Euclidean distance and median linkage. (B) Volcano plot of the differential gene expression analysis showing RR-upregulated (red) and downregulated (blue) genes. The paired inverted β binomial test was used to perform differential gene expression analysis. The relevant genes are annotated in the plot.

Figure 2. Functional analysis of RR-upregulated genes.

(A) Dot plot of selected host defense functional pathways enriched (–log₁₀ Padj >1.3 = Padj < 0.05) in the RR-upregulated gene signature. (B) Venn diagram depicting overlap between the GeneCards database antimicrobial gene signature (n =1,693) and the RR-upregulated genes (n = 200). (C) Heatmap showing the expression of 64 antimicrobial genes in each patient before (pre-RR) and at RR clinical onset (RR). (D) Antimicrobial response signature z score for each patient before and at RR clinical onset. Statistical analyses were performed in GraphPad Prism 9.12 using a paired 2-tailed t test. **P < 0.01. (E) Dot plot showing the IPA UPR analysis of the 64 antimicrobial genes upregulated in RR skin lesions.

Figure 3. Different cell populations in RR skin lesions express the RR antimicrobial response signature.

Heatmap of average expression z scores for 53 of the 64 genes from the RR antimicrobial response signature (z score >2) detected in RR cell types defined by scRNA-Seq (GSE151528). The heatmap’s red-to-blue color scale indicates high to low expression. Cell type subclusters represent T cells (TC), myeloid cells (LC and ML), keratinocytes (KC), fibroblasts (FB), and endothelial cells (EC). The regulation of the antimicrobial genes (z score >2) by their respective UPRs is depicted as a heatmap at the bottom in light blue (IL17A), dark blue (IFNG), violet (TNF), and red (IL1B).

Figure 4. The RR antimicrobial response signature is more highly expressed in T-lep and RR pre-MDT patients and negatively correlates with bacillary load.

(A) Heatmap displaying expression z scores for the 64 RR antimicrobial genes in leprosy clinical forms, with red to light blue color scale indicating high to low expression. T-lep, RR pre-MDT, BL, and L-lep samples were grouped by hierarchical clustering using Canberra distance and the McQuitty linkage method. (B) Plot showing the antimicrobial response signature z scores for each patient from the T-lep (n = 10), RR pre-MDT (n = 12), BL (n = 6), and L-lep (n = 7) groups. Data represent the mean ± SEM. (C) Correlation analysis between RLEP expression and antimicrobial response gene signature z scores for each patient from the T-lep (red), RR pre-MDT (pink), BL (light blue), and L-lep (blue) groups. Statistical analyses were performed in GraphPad Prism 9.12 using ordinary 1-way ANOVA followed by Tukey’s multiple-comparison test (B) and Spearman’s correlation coefficient (C). *P < 0.05, **P < 0.01 and ****P < 0.0001.

Figure 5. Genes upregulated in RR skin lesions encode proteins with membrane-active AMP motifs.

(A) Graph displaying the amino acid position of the AMP-like motifs (yellow) identified along the protein sequence encoded by the RR-upregulated genes. (B) Venn diagram depicting the overlap between the 41 RR genes with AMP-like motifs and the human AMPs in the APD3 database (n = 117). (C) Evaluation of cationic and hydrophobic content of the AMP-like motifs detected in 41 RR antimicrobial molecules (colored circles and triangles) shown in a plot of lysine (K) to arginine (R) ratio = N_K/(N_K + N_R) versus the mean hydrophobicity, together with known α-helical AMPs from the APD3 database (black circles).

Figure 6. RNA-FISH shows antimicrobial gene expression in RR and pre-RR skin lesions by different cell populations.

(A) RNA-FISH of S100A7 (green) and staining for keratin 14 (KRT14) protein (red) in 1 representative pair of RR and pre-RR skin lesions (BL4/RR.BL4). S100A7 RNA dot quantification (number of dots) was performed on 4 pairs of RR and pre-RR skin lesions. (B) RNA-FISH of S100A8 (green) and protein staining of KRT14 (red) in 1 representative pair of RR and pre-RR skin lesions (BL5/RR.BL5). S100A8 RNA dot quantification (number of dots) was performed on 4 pairs of RR and pre-RR skin lesions. (C) RNA-FISH of CCL17 (red) and LYZ (green), a macrophage marker, in 1 representative pair of RR and pre-RR skin lesions (BL3/RR.BL3). CCL17 RNA dot quantification (number of dots) was performed on 4 pairs of RR and pre-RR skin lesions. (D) RNA-FISH of CCL19 (red) and COL1A1 (green), a fibroblast marker, in 1 representative pair of RR and pre-RR skin lesions (BL4/RR.BL4). CCL19 RNA dot quantification (number of dots) was performed on 4 pairs of RR and pre-RR skin lesions. Cell nuclei were stained with DAPI (blue). Images were acquired with a Leica TCS SP8 Digital Light Sheet microscope, and RNA dot quantification was performed using ImageJ. Scale bars: 10 μm; original magnification, ×630 (A–C) and ×630 with ×3 zoom (D). Statistical analyses were performed in GraphPad Prism 9.12 using the ratio paired t test (A and B) or paired 2-tailed t test (C and D). *P < 0.05 and **P < 0.01.

Figure 7. Protein expression of S100A7, S100A8, CCL17, and CCL19 in RR and pre-RR skin lesions.

(A) S100A7 and S100A8 protein expression in a representative pre-RR and RR skin lesion pair (LL1/RR.LL1) evaluated by IHC. (B) CCL17 and CCL19 protein expression in a representative pre-RR and RR skin lesion pair (BL4/RR.BL4) evaluated by IHC. CD68, a macrophage marker, was used as a positive control. Graphs show quantification of S100A7 (n = 6 pairs), S100A8 (n = 5 pairs), CCL17 (n = 4 pairs), and CCL19 (n = 5 pairs) staining 3-amino-9-ethylcarbazole (AEC)/nuclear area. Image J plugin ImmunoRatio was used for quantification, and staining was visualized and images were acquired using a Leica microscope (Leica 250). Scale bars: 25 μm; original magnification, ×200. Statistical analyses were performed in GraphPad Prism 9.12 using the paired t test (S100A7 and CCL19) or ratio paired, 2-tailed t test (S100A8 and CCL17). *P < 0.05.

Figure 8. S100A7, S100A8, CCL17, and CCL19 show antimicrobial activity against M. leprae in infected human macrophages.

(A–D) MDMs from healthy donors were infected overnight with M. leprae at a MOI of 5:1, followed by addition of 0.1 μM recombinant human S100A7, S100A8, CCL17, and CCL19 for 4 days. M. leprae viability was assessed by qPCR, and the percentage of antimicrobial activity was calculated by assigning 100% bacteria viability to the media control. Rifampin (10 μg/mL) (RIF) was added as a positive control. (E) Lysosome acidification was assessed by LysoTracker staining (green) in MDMs previously stimulated with 0.1 µM recombinant human S100A7, S100A8, CCL17, and CCL19 for 1 hour and then infected with M. leprae labeled with PKH26 (red) at a MOI of 5:1 over night. Leptin (0.1 μM) was used as a negative control. Images were captured using a Leica TCS SP8 Digital Light Sheet Microscope. DAPI (blue) was used to stain the nuclei. Scale bars: 10 μm; original magnification, ×630 with ×4 zoom. Statistical analyses were performed in GraphPad Prism 9.12 using the Friedman test followed by Dunn’s multiple-comparison test (A–D). Data represent the mean ± SEM (n = 6 for A and C) and (n =7, B and D). *P < 0.05 and **P < 0.01.

Figure 9. S100A7, S100A8, CCL17, and CCL19 show direct antimicrobial activity against M. leprae.

(A–D) Different concentrations of recombinant human S100A7, S100A8, CCL17, and CCL19 were added to M. leprae (2 × 10⁶ bacilli) in 7H9 broth with 10 mM sodium phosphate, pH 7.2, for 72 hours. Bacteria viability was assessed by qPCR, and rifampin (10 μg/mL) (RIF) was used as a positive control. (E) S100A7 (4.5 μM), S100A8 (9 μM), CCL17 (4.5 μM), and CCL19 (4.5 μM) were added to M. leprae (15 × 10⁶ bacilli) in 7H9 broth with 10 mM sodium phosphate, pH 7.2, for 6, 24, 48, and 96 hours, and bacteria morphology was evaluated by scanning electron microscopy. IL-26 (10 μM) was used as a positive control. Scale bar: 500 nm; original magnification, ×100,000. Statistical analyses were performed in GraphPad Prism 9.12 using repeated-measures of 1-way ANOVA with the Geisser-Greenhouse correction and Dunnett’s multiple-comparison test (A–D). Data represent the mean ± SEM (n = 4). **P < 0.01, ***P < 0.001, and ****P < 0.0001.