Gene expression patterns associated with multidrug therapy in multibacillary leprosy

Abstract

Multidrug therapy (MDT) has been successfully used in the treatment of leprosy. However, although patients are cured after the completion of MDT, leprosy reactions, permanent disability, and occasional relapse/reinfection are frequently observed in patients. The immune system of multibacillary patients (MB) is not able to mount an effective cellular immune response against M. leprae. Consequently, clearance of bacilli from the body is a slow process and after 12 doses of MDT not all MB patients reduce bacillary index (BI). In this context, we recruited MB patients at the uptake and after 12-month of MDT. Patients were stratified according to the level of reduction of the BI after 12 doses MDT. A reduction of at least one log in BI was necessary to be considered a responder patient. We evaluated the pattern of host gene expression in skin samples with RNA sequencing before and after MDT and between samples from patients with or without one log reduction in BI. Our results demonstrated that after 12 doses of MDT there was a reduction in genes associated with lipid metabolism, inflammatory response, and cellular immune response among responders (APOBEC3A, LGALS17A, CXCL13, CXCL9, CALHM6, and IFNG). Also, by comparing MB patients with lower BI reduction versus responder patients, we identified high expression of CDH19, TMPRSS4, PAX3, FA2H, HLA-V, FABP7, and SERPINA11 before MDT. From the most differentially expressed genes, we observed that MDT modulates pathways related to immune response and lipid metabolism in skin cells from MB patients after MDT, with higher expression of genes like CYP11A1, that are associated with cholesterol metabolism in the group with the worst response to treatment. Altogether, the data presented contribute to elucidate gene signatures and identify differentially expressed genes associated with MDT outcomes in MB patients.

Keywords: bacillary load; gene signature; lipid metabolism; multibacillary leprosy; multidrug therapy.

Figure 1

Differentially expressed genes after multi-drug therapy of MB leprosy cases. (A) Heatmap depicting top 50 DEG (greatest |log₂FC| and FDR ≤ 0.1) after comparing after treatment (AT) vs. before treatment (BT). Z-score represents the number of standard deviations away from normalized log₂ gene expression. Samples (columns) were clustered using hierarchical clustering with Euclidean distance and complete agglomeration. Genes without an official HGNC symbol are shown as their ENSEMBL identifiers. (B) Volcano plot showing 121 DEG (blue points) with |log₂FC| ≥ 1 and false discovery rate (FDR) ≤ 0.1. Genes with the greatest mean differences are shown for emphasis. (C) Examples of regulated genes before (BT) after MDT (AT) in MB patients colored according to their responsive (R) or non-responsiveness (NR) to MDT.

Figure 2

Gene Ontology Biological Processes associated with DEG by GSEA. Gene set enrichment analysis (GSEA) highlighting the top 30 biological processes from Gene Ontology enriched in DEG between AT vs. BT. FDR, false discovery rate. BH, Benjamini-Hochberg. AT, after treatment; BT, before treatment. NES, normalized enrichment score. See also Supplementary Table 1.

Figure 3

Differentially expressed genes between patient responses to MDT. Top 50 DEG from responder (R) vs. non-responder (NR) LL patients before treatment (A) and after treatment (B). Z-score represents the number of standard deviations away from normalized log₂ gene expression. Top 50 genes with greatest |log₂FC| and FDR ≤ 0.1 are shown. Samples (columns) were clustered using hierarchical clustering with Euclidean distance and complete agglomeration. Genes without an official HGNC symbol are shown as their ENSEMBL identifiers. Plot showing the expression of a subset of DEG with distinct patterns between R and NR patients either before (C) or after treatment (D). In C-D, only the subset of samples from sequencing batch “CH” is shown as used in DESeq2’s generalized linear model (see Methods).