Molecular signatures of anthroponotic cutaneous leishmaniasis in the lesions of patients infected with Leishmania tropica

Abstract

Anthroponotic cutaneous leishmaniasis (CL) caused by Leishmania tropica (L. tropica) represents a public health challenge in several resource poor settings. We herein employed a systems analysis approach to study molecular signatures of CL caused by L. tropica in the skin lesions of ulcerative CL (UCL) and non-ulcerative CL (NUCL) patients. Results from RNA-seq analysis determined shared and unique functional transcriptional pathways in the lesions of the UCL and NUCL patients. Several transcriptional pathways involved in inflammatory response were positively enriched in the CL lesions. A multiplexed inflammatory protein analysis showed differential profiles of inflammatory cytokines and chemokines in the UCL and NUCL lesions. Transcriptional pathways for Fcγ receptor dependent phagocytosis were among shared enriched pathways. Using L. tropica specific antibody (Ab)-mediated phagocytosis assays, we could substantiate Ab-dependent cellular phagocytosis (ADCP) and Ab-dependent neutrophil phagocytosis (ADNP) activities in the lesions of the UCL and NUCL patients, which correlated with L. tropica specific IgG Abs. Interestingly, a negative correlation was observed between parasite load and L. tropica specific IgG/ADCP/ADNP in the skin lesions of CL patients. These results enhance our understanding of human skin response to CL caused by L. tropica.

Figure 1

Schematic representation of the study work flow and methodology. Assays used in the study included. (a) The ITS1 PCR–RFLP assay for diagnosis of L. tropica patients. (b) Quantitative polymerase chain reaction (qPCR) analysis for quantification of L. tropica. (c) NextSeq 500 platform based RNA-seq method for investigation of whole genome transcription in biopsy samples. (d) Proximity extension assay (PEA) in biopsy supernatants for inflammatory protein assessment. (e, g) Enzyme-linked immunosorbent assay (ELISA) in biopsy supernatants and serum samples for L. tropica specific IgG antibody assessment. (f, h) Antibody-dependent neutrophil phagocytosis (ADNP) and Antibody-dependent cellular phagocytosis by human monocytes (ADCP) in biopsy supernatants and serum samples.

Figure 2

Transcriptomics analysis of skin lesions in ulcerative cutaneous leishmaniasis (UCL) and non UCL (NUCL) of L. tropica-infected patients, and healthy skins. (a) Partial least squares-discriminant analysis (PLS-DA) modeling was applied for classification of UCL and NUCL lesions and skin biopsies from healthy individuals. Volcano plots of UCL (b) and NUCL (c) based on DESeq2 results. Dots in red (Log FC > 1) and blue (Log FC < 1) represent differentially expressed genes with adjusted p-value < 0.05. All three figures were created using R version 3.5.1, PLS-DA with R package mixOmics 6.6.2 and volcano plots using R package ggplot2 3.2.1.

Figure 3

Common enriched GSEA pathways in the lesions of UCL and NUCL patients. Bar plot (a) and enrichment map plot (b,c) of GSEA revealed enrichment of 45 common pathways in the skin lesions of L. tropica-infected UCL and NUCL patients. In (b) and (c), dash circles denote five main clusters of enriched pathways. The color and length of the boxes represent normalized enrichment scores (NES), and the number of genes mapped to the indicated pathways, respectively. Positive and negative NES show up- and down-regulated pathways in the skin lesions of the CL patients compared with skin biopsies of healthy individuals, respectively. For the pathway analysis ReactomePA R package version 1.28.0 was used. This package version uses Reactome version 70 (https://reactome.org/about/news/142-version-70-released). All figures were created using R version 3.5.1.

Figure 4

Gene networks combined by the common GSEA pathways. The GSEA results revealed significant transcriptional changes of several pathways involved in inflammatory responses in the L. tropica infected lesions, including “Chemokine receptors bind chemokines” , “Signaling by Interleukins”, “IL-10 signaling”, “IL-4 and IL-13 signaling” “IFN-γ signaling”, “IFN-α/β signaling”, and “IFN Signaling”, for UCL (a) and NUCL (b). This figure was created using Cytoscape version 3.7.2 and the Reactome FI plugin version 7.2.0.

Figure 5

Inflammatory proteins expression in the lesions of the UCL and NUCL patients compared with healthy skins. The supernatants from UCL, NUCL and healthy biopsies cultured in RPMI, were analysed using Olink Inflammation panel. (a) NPX values for proteins with significant expression (p < 0.05) in the all lesions (full squares = UCL, open squares = NUCL) relative to healthy skin (Full circles). The inflammatory proteins were categorized into: CC and CXC chemokines family, Surface Molecules and Receptors and other cytokine and inflammatory proteins. (b) Circos plot displaying the inflammatory proteins in the UCL, NUCL and healthy samples. Results were analyzed with a t test/Mann–Whitney test, and Šidák adjusted p-values < 0.05 were considered statistically significant. The Circos plot in (b), generated using CIRCOS v 0.69-6, visualizes the similarities and differences between UCL and NUCL patients. Red lines (p < 0.01), orange lines (p < 0.05) and yellow lines (p < 0.1).

Figure 5

Inflammatory proteins expression in the lesions of the UCL and NUCL patients compared with healthy skins. The supernatants from UCL, NUCL and healthy biopsies cultured in RPMI, were analysed using Olink Inflammation panel. (a) NPX values for proteins with significant expression (p < 0.05) in the all lesions (full squares = UCL, open squares = NUCL) relative to healthy skin (Full circles). The inflammatory proteins were categorized into: CC and CXC chemokines family, Surface Molecules and Receptors and other cytokine and inflammatory proteins. (b) Circos plot displaying the inflammatory proteins in the UCL, NUCL and healthy samples. Results were analyzed with a t test/Mann–Whitney test, and Šidák adjusted p-values < 0.05 were considered statistically significant. The Circos plot in (b), generated using CIRCOS v 0.69-6, visualizes the similarities and differences between UCL and NUCL patients. Red lines (p < 0.01), orange lines (p < 0.05) and yellow lines (p < 0.1).

Figure 6

The enriched “FcγR dependent phagocytosis” gene pathway and antibody-dependent phagocytosis assays reveal antibody-dependent cellular phagocytosis (ADCP) and antibody-dependent neutrophil phagocytosis (ADNP) in the lesions of CL patients. Gene networks related to “FcγR dependent phagocytosis” pathway was enriched in the GSEA in skin lesions from UCL (a) and NUCL (b) patients. The dot colors indicate fold change in the gene expression as stated in the map. Violin plots show L. tropica-specific IgG antibody levels (ng/mL) in skin biopsy supernatants (c) and serum samples (d), ADCP activity in skin biopsy supernatants (e) and serum samples (f), as well as ADNP activity in skin biopsy supernatants (g) and serum samples (h). The phagocytic activity (AUC) was measured by the uptake of L. tropica-specific antibody-bead complexes by either human peripheral neutrophils or THP-1 cells, using flow cytometry. Spearman correlation between IgG antibody levels and ADCP, in skin biopsies (i) and serum samples (j), as well as between IgG antibody levels and ADNP, in skin biopsies (k) and serum samples (l). All figures were created using R version 3.5.1.