Transcriptomic landscape of skin lesions in cutaneous leishmaniasis reveals a strong CD8⁺ T cell immunosenescence signature linked to immunopathology

Abstract

The severity of lesions that develop in patients infected by Leishmania braziliensis is mainly associated with a highly cytotoxic and inflammatory cutaneous environment. Recently, we demonstrated that senescent T and NK cells play a role in the establishment and maintenance of this tissue inflammation. Here, we extended those findings using transcriptomic analyses that demonstrate a strong co‐induction of senescence and pro‐inflammatory gene signatures in cutaneous leishmaniasis (CL) lesions. The senescence‐associated signature was characterized by marked expression of key genes such as ATM, Sestrin 2, p16, p21 and p38. The cell type identification from deconvolution of bulk sequencing data showed that the senescence signature was linked with CD8⁺ effector memory and T_EMRA subsets and also senescent NK cells. A key observation was that the senescence markers in the skin lesions are age‐independent of patients and were correlated with lesion size. Moreover, a striking expression of the senescence‐associated secretory phenotype (SASP), pro‐inflammatory cytokine and chemokines genes was found within lesions that were most strongly associated with the senescent CD8 T_EMRA subset. Collectively, our results confirm that there is a senescence transcriptomic signature in CL lesions and supports the hypothesis that lesional senescent cells have a major role in mediating immunopathology of the disease.

In this study, we analysed an immunosenescence signature in the transcriptomic environment of cutaneous leishmaniasis lesions caused by Leishmania braziliensis. We found a wide range of senescence‐associated genes overrepresented in lesions of patients, including key genes (ATM, Sestrin 2, p16, p21 and p38) as well as a substantial number of SASP and other pro‐inflammatory genes. Importantly, most of these genes were also correlated with estimated proportions of CD8+ TEMRA cells and senescent NK cells, supporting the hypothesis that lesional senescent cells have a major role in mediating immunopathology of this disease.

Abbreviations

CL

cutaneous leishmaniasis

DEG

differentially expressed gene

EM

effector memory T cells

EMRA

effector memory T cells that re‐expressing CD45RA

GEO

NCBI’s Gene Expression Omnibus

GO

Gene Ontology

HC

healthy controls group

LCL

localized cutaneous leishmaniasis group

MCL

mucocutaneous leishmaniasis

NKR

NK cell receptors

NKRL

NK cell receptor ligand

PCA

principal component analysis

SASP

senescence‐associated secretory phenotype

Tsen

senescent T cells

INTRODUCTION

Leishmaniasis is a group of severe neglected tropical diseases, caused by Leishmania parasites and is prevalent in over 98 countries in the world [1]. Leishmania braziliensis is the main causal agent of both cutaneous (CL) and mucocutaneous (MCL) leishmaniasis in Brazil, where mucosa and skin destructive lesions progressively develop [2].

The immune response against Leishmania is a complex process involving both innate and adaptive immunity, where T cells have a key role in promoting both protection and/or pathology [3, 4]. With regard to the pathology that occurs, substantial evidence has demonstrated the role of the overwhelming inflammatory processes, as well as non‐specific cytotoxic response in the disease [4, 5, 6, 7, 8, 9, 10]. This was observed in skin lesions of patients with MCL and CL that demonstrated the enrichment of cytotoxic and pro‐inflammatory cells with a conspicuous capacity to mediate tissue destruction [5, 6, 9]. Moreover, both cytotoxicity and inflammatory‐related genes are overexpressed in the lesional environment of CL patients that positively correlated with extensive necrosis and lesion size [11, 12, 13, 14, 15, 16].

Immunosenescence is a term that refers to the gradual change of the immune system resulting from ageing, persistent inflammatory states and chronic antigenic stimulation [17, 18]. Some of these changes are obvious within in the T cell compartment. These cells acquire a senescent phenotype and defective responses that are linked with the loss of proliferative capacity and decreased TCR‐related signalling and proliferation after activation [19, 20].

Senescent T cells (Tsen) undergo widespread changes in morphology, cell metabolism and secretory phenotype [21]. These cells are still metabolically active and secrete a wide array of pro‐inflammatory cytokines, chemokines, matrix metalloproteases and growth factors that is collectively known as the senescence‐associated secretory phenotype (SASP) that has considerable consequences for inducing the tissue inflammatory microenvironment [22, 23]. Moreover, both CD4⁺ and CD8⁺ Tsen cells express high levels of intracellular granules containing the cytotoxic proteins perforin and granzyme B, as well as express NK cell receptors (NKRs) [24, 25, 26]. Therefore, the increased inflammatory potential of T cells in CL may contribute to the pathology of this disease.

Recently, we have demonstrated that both CD4⁺ and CD8⁺ circulating Tsen cells with high inflammatory profile accumulate during infection with L. braziliensis. Moreover, they showed an increased propensity to migrate to the skin, where their potent inflammatory activity and accumulation were linked to the severity of skin lesions observed in cutaneous leishmaniasis patients [27, 28].

We have now analysed previously published RNA‐Seq data sets to evaluate the immunosenescence signature within lesional CL transcriptomes. Confirming previous observations, we found that lesions are transcriptionally distinct from healthy skin, exhibiting a strong co‐induction in the transcript levels of senescence‐associated genes and pro‐inflammatory immune responses. Moreover, transcriptional signatures of immunosenescence genes and pathways not only were found overrepresented in lesions, but these lesional signatures were linked to the immunopathology of CL. Therefore, this study sheds insight into the skin lesional environment and reveals gene signatures associated with the physiopathology of the disease, which may be explored and extrapolated to understand inflammation and immune‐mediated pathology in CL and several other related diseases.

MATERIALS & METHODS

Transcriptomic analyses of skin lesions from publicly available data sets

Raw counts matrix from a previous study on the lesional gene expression in CL was obtained from NCBI’s Gene Expression Omnibus (GEO) through accession code: GSE127831 [29]. All analyses of RNA‐Seq data were performed using R version 4·0.3 [30] in RStudio 1·4.1103 [31]. This resulted in count data from 21 skin samples from patients infected with L. braziliensis before treatment and 7 uninfected endemic controls. Patients from this data set consisted of 17 males and 4 females of age 38 ± 16,8 and lesions size ranging from 13 to 1237 mm². Genes that had less than 10 total counts in all samples were removed prior to subsequent analyses, resulting in a count table of 26,545 genes. DESeq2 [32] that was used to identify differentially expressed genes, considering those with Benjamini–Hochberg p‐adjusted value less than 0·05 as statistically significant. For visualization and clustering, count data were normalized, and variance stabilized through the vst function available in DESeq2 package. The Principal Component Analysis (PCA) was calculated using plotPCA function, also in the DESeq2 package. Gene Ontology (GO) overrepresentation was assessed using clusterProfiler package for Biological Processes in the GO database by using all DEGs and considering significant categories that had adjusted p‐value <0·05. The CellAge database [33] from Human Aging Genomic Resources [34] was adapted with added senescence‐associated genes and was used to investigate the senescence signature in the dataset. The STRING database v11·0 [35] was used to determine human protein‐protein interactions (minimum score of 700) within the senescence‐associated DEGs through use of the RITAN package. Immune cell deconvolution was performed through ImmQuant software [36] using the DMAP database, with scores calculated relative to the mean of all samples. Spearman's correlation was calculated between variance stabilized gene expression values and cell population scores predicted by ImmQuant. A correlation matrix was generated with the cor function in R, and the p‐values matrix was created using the cor_pmat function in ggcorrplot package. The results were then plotted as a heatmap. Several tools were used for visualizing the results of RNA‐Seq analyses. Heatmaps were constructed with relative expression vst transformed values, centred across gene means, through the ComplexHeatmap package. Volcano plot was made with EnhancedVolcano package. The package ggraph was employed to construct the network plots. All other plots were generated with ggplot2.

Parasite burden determination

Parasite burden data were obtained from Amorim and cols, 2019 [29], that it quantified by qPCR in the same RNA preparations used for RNA‐seq, as described by the authors. Briefly, a standard curve was prepared from total RNA extracted from 10⁷ L. braziliensis promastigotes recovered from axenic culture using the RNeasy Plus Mini Kit (Qiagen), and cDNA was generated with the High Capacity RNA to cDNA kit (Applied Biosciences). qPCR was performed using Power SYBR green master mix (Applied Biosciences) and primers targeting the L. braziliensis 18S ribosomal subunit (F: 5′‐TACTGGGGCGTCAGAG‐3′ and R: 5′‐GGGTGTCATCGTTTGC‐3′) and human GAPDH (housekeeping gene, F: 5′‐ GGTGTGAACCATGAGAAGTATGA‐3′ and R: 5′‐GAGTCCTTCCACGATACCAAAG‐3′).

Immunofluorescence study subjects

Skin biopsies from 10 untreated patients with cutaneous leishmaniasis seen at the University Hospital (HUCAM) of Universidade Federal do Espírito Santo, Brazil, were investigated in this study. Patient group (LCL) consisted of 5 males and 3 females of age 37 ± 10·3 years and lesion sizes ranging from 7 to 2734 mm² with durations ranging from 45 to 120 days. Diagnosis of CL was made with clinical and laboratory criteria, and all patients in this study were positive for the PCR/restriction fragment length polymorphism of L. braziliensis and reported no prior infections or treatments. The control group (HC) consisted of 6 healthy age (26,83 ± 5,64) and gender‐matched individuals with no history of leishmaniasis. All participants (patients and healthy volunteers) tested for HIV, HBV and HCV infections, and presented no history of chemotherapy, radiotherapy or treatment with immunosuppressive medications within the last 6 months. All samples in the study were obtained before the COVID‐19 outbreak. Patients provided written informed consent, and all study procedures observed the principles of the Declaration of Helsinki. This study was registered at HUCAM ethical committee reference number 735·274.

Skin Biopsies

Punch biopsies (8 mm in diameter) from the border of lesional skin from cutaneous leishmaniasis patients and from skin of healthy volunteers were obtained. Biopsy specimens were frozen in OCT compound (Sakura, Alphen aan den Rijin). Six‐micrometre sections were longitudinally sectioned to expose all skin layers and placed in poly‐L‐lysine‐coated slides (Star Frost^®). Tissues were then fixed in acetone and ethanol and stored in −80℃ until use.

Immunofluorescence

Six micrometres cryosections were obtained from skin biopsies of healthy individuals and patients with cutaneous leishmaniasis and prepared on poly‐L‐lysine‐coated glass slides (StarFrost, Knittel, Germany). Sections were stained for KLRG‐1 (Abcam, ab170959, 1:400) and ATF2 (Abcam, ab242381, 1:100) for 18 h at 4℃, followed by incubation with the secondary antibody goat anti‐rabbit IgG Alexa Fluor 568 (Thermo Fisher, A11036, 1:400) and mounting with Flouroshield Mounting Medium with DAPI (ABCAM, ab104139). Imaging consisted of obtaining five images of the dermis captured at 200x magnification using fluorescence microscopy (Leica DMi8) or slide scanner (Axio Scan). All positively stained cells and nuclei within each image were manually counted using computer‐assisted image analysis (National Institutes of Health Image Software ImageJ 1·52j; https://imagej.nih.gov/ij/). The data are shown as the average number of positive cells divided by the nuclei count x 100 to be expressed as percentage.

Statistical analysis

Analysis of ImmQuant results and gene correlation data was performed in R environment. For immunofluorescence data, analyses were performed in GraphPad Prism v8·0. Cell population scores were compared using unpaired Wilcoxon rank‐sum test. Immunofluorescence data normality was assessed with the Shapiro–Wilk test, and groups were compared using Student's t test with Welch's correction or Wilcoxon rank‐sum test. Correlations were assessed by Spearman's rank correlation for gene expression data, and Pearson's correlation for immunofluorescence data. Significance values were represented as follows: *p < 0·05; **p < 0·01; ***p < 0·001; ****p < 0·0001.

Code availability

A markdown report with all code used to produce the results in this study is provided as a supplementary file.

RESULTS

Transcriptional signature of CL lesions reveals a senescent profile

First, we analysed if a senescence signature was present in biopsies collected from the border of lesions from patients infected with Leishmania braziliensis when compared to skin biopsies from uninfected controls using RNA‐Seq data from a previous study [29]. We defined differentially expressed genes (DEGs) as those with an adjusted p‐value <0·05 and found a total of 12,306 DEGs, in which 6,076 were overrepresented and 6,230 were underrepresented (Table S1). The heatmap and hierarchical clustering of DEGs (Figure S1A), PCA (Figure S1B) and volcano plot (Figure S1C) of the total transcriptome data revealed distinct gene expression profiles between healthy skin and lesions from infected patients, supporting previous observations [13, 29, 37]. Gene Ontology (GO) overrepresentation analysis of all DEGs also revealed GO terms that are closely linked with both protective and pathological mechanisms of CL, as well as terms associated with senescence, namely the GO terms ‘ageing’ and ‘replicative senescence’ (Figure 1A).

FIGURE 1.

Differentially expressed genes demonstrate senescence profile. (a) Gene ontology overrepresentation analysis of all DEGs. The selected categories were all significantly enriched with p‐adjusted value <0·05. Size of the circles was used to distinguish the number of genes in each category. (b) Immune cell deconvolution using ImmQuant software. Values represent relative score based on mean of all samples in each cell type. (c) Dotplots and boxplots of selected cell populations from deconvolution analysis where groups are compared by Wilcoxon rank‐sum test. HC, Healthy controls; LCL, Localized Cutaneous Leishmaniasis patients

Specific senescence‐associated features have been mainly characterized in circulating T cells during cutaneous leishmaniasis [28]. To investigate the possible cell types involved on the immune response within CL lesions, we performed a deconvolution analysis of bulk RNA‐Seq data using ImmQuant [36]. This analysis revealed distinct immune cell type compositions between patients and healthy controls (Figure 1B). Notably, CD8 effector memory T (CD8 T_EM) and effector memory T cells that re‐express CD45RA (CD8 T_EMRA), NK cells and monocytes were significantly enriched in CL lesions (Figure 1C).

Senescence‐associated genes are found in CL lesions

We previously reported the skin lesion accumulation of senescent cells and its role in mediating pathology of lesions [28]. To confirm the presence of a lesional senescence signature, we next analysed a curated selection of genes from CellAge senescence database along with other known key genes associated with senescence. Using foldchange cut‐offs of 2 and −2, we found 107 differentially expressed genes in CL lesions, where 70 were overrepresented and 37 were underrepresented that were classified within oncogene‐, stress‐induced and replicative senescence categories (Figure S2A, Table S2). Interaction analysis using senescence DEGs revealed interactions among 48 overrepresented and 8 underrepresented genes (Table S3) that were mostly linked with the replicative senescence pathway (Figure S2B). Further analysis within key senescence DEGs [38] found a wide upregulation of key genes associated with immunosenescence that included cyclin‐dependent kinase inhibitors p21 (CDKN1A) and p16 (CDKN2A), DNA damage response protein ATM, p38 (MAPK11), Sestrin 2 (SESN2), ATF5, as well as key genes linked with cell differentiation such as EOMES and Tbet (TBX21) (Figure 2a). Moreover, they did not correlate with the age of patients, suggesting that in the lesional environment the expression of senescence key genes is governed by non‐age‐related factors (Figure S3A) Interestingly, most of the senescence genes correlated positively with the parasitic load observed in lesions of patients (Figure 2b), opposed to what was found regarding healing time, where only p21 (CDKN1A) showed a positive correlation (Figure S3B). Next, to evaluate the link between the expression of senescence genes and lesional cell populations, we next performed Spearman correlations between the ImmQuant scores [34] and senescence DEGs. We found that the expression of senescence genes positively correlates with the predicted relative proportions of several immune cell types, including CD8 T and NK cell subsets and monocytes. However, the strongest correlations were found within senescent CD8 (EMRA) and senescent NK (CD56⁺CD16⁺) subsets (Figure 2c). Senescence genes also had a positive correlation with monocytes (Figure 2c), to a lesser extent, that requires further analysis.

FIGURE 2.

Senescence‐associated genes are conspicuous in CL lesions. (a) Violin plot of selected key senescence‐associated genes. Gene expression is represented as log2 normalized and variance stabilized (vst) values, with corresponding adjusted p‐values as calculated by DESeq2, where healthy controls (HC, light blue) and lesions from LCL patients (LCL, red) are compared. (b) Spearman correlations between the selected senescence genes from A and the number of parasites detected by qPCR in lesions of patients. (c) Correlation between expression of selected senescence genes from A and predicted cell population scores calculated by ImmQuant. HC, healthy controls. LCL, Localized Cutaneous Leishmaniasis

Overall, these data support our findings on the accumulation of senescent T and NK cells in the lesional environment and suggest that they may be maintained or expanded by the presence of antigens or parasites. Moreover, we cannot exclude the possibility that tissue non‐lymphoid cells (such as keratinocytes, fibroblasts and/or endothelial cells) also exhibited a senescence profile as gene signatures for these cells were not part of the deconvolution signature.

Transcriptional landscape of CL lesions shows a profile of SASP, inflammatory cytokine and chemokine genes

Overwhelming inflammation is one of the main signatures of CL skin pathogenesis. Given the results indicating a senescent profile in lesions, we next investigated if senescent cells may instigate tissue damage through a senescence‐associated secretory phenotype (SASP). We performed a wide analysis of inflammatory cytokines, chemokines and their receptors that may induce skin pathology. Our analysis found 75 differentially expressed genes, of which 63 were increased by at least two‐fold (Table S4). Comparative analysis uncovered that among 68 SASP genes tested, 50 were DEGs and 25 were greatly increased, with fold changes ranging from 13‐fold (IL1A) to 535‐fold (MMP1) increases (Figure 3A and Table S4). The top 12 overrepresented SASP genes were associated with recruitment of inflammatory immune cells (CCL3, CCL8, CXCL1, CXCL11, CXCL13, CXCL8); metalloproteinases (MMP1, MMP10 and MMP3) and inflammatory cytokines (IL‐6, IL‐15 and IL‐1β) (Figure 3b, c and Figure S4). In addition, other non‐SASP inflammatory cytokines, chemokines and their receptors such as IFN‐γ (IFNG), TNF‐α (TNF), IL‐23 (IL23A), IL‐15 receptor (IL15RA), CXCL10, CXCL9, CCL4 and CCR5 were overrepresented in CL lesions (Figure 3C and Figures S5A,B), supporting previous findings [13, 14, 29, 37, 39]. As expected, most of the DEGs positively correlated with the predicted relative proportions of analysed cell subsets (Figure 3C). Interestingly, this correlation was most conspicuous within NK CD56⁺CD16⁺, CD8 T_EM and CD8 T_EMRA subsets, where the latter demonstrated the strongest association with all SASP, cytokines, chemokines and cytokine/chemokine receptors DEGs analysed (Figure 3C). Most of these genes also correlated with monocyte population, suggesting their presence in the lesional environment with an overlapping inflammatory profile with senescent cells (Figure 3C).

FIGURE 3.

Senescence‐associated secretory phenotype (SASP), inflammatory cytokine and chemokine genes show are overrepresented in CL lesions and are correlated senescent cells. (a) MA plot highlighting genes that were differentially expressed in the referred categories. Light grey represents non‐significant genes, medium grey represents underrepresented genes, and dark grey represents overrepresented genes. Indicated are the top 12 overrepresented genes from (b). B) Violin plots showing the marked increase of the top 12 overrepresented SASP genes. Gene expression is represented as normalized and variance stabilized (vst) values, with corresponding adjusted p‐values as calculated by DESeq2, where healthy controls (HC, light blue) and lesions from LCL patients (LCL, red) are compared. (c) Two heatmaps are shown. The heatmap on the left shows differentially expressed from the corresponding categories, comparing skin biopsies from healthy controls (HC, light blue) and lesions from LCL patients (LCL, red). Gene expression is represented in relative normalized and variance stabilized (vst) expression values (centred in row means), ranging from cyan (underrepresented) to yellow (overrepresented). Each row corresponds to a single gene and each column to a subject. The heatmap on the right represents the spearman correlation (ranging from dark blue to dark red) of expression of the same genes to predicted cell population scores calculated by ImmQuant. NS, non‐significant. HC, healthy controls. LCL, Localized Cutaneous Leishmaniasis

The expression of senescence markers is enhanced in CL lesions and correlates with lesion size

Senescent leukocytes may be defined by several phenotypic markers, as well as a state of cell cycle arrest [40]. We next performed in situ analysis by immunofluorescence for killer cell lectin‐like receptor 1 (KLRG‐1) that is a well‐known senescence marker of T and NK cells [41]. In addition, we also analysed the human ATF phosphorylated protein, an important transcription factor of the senescence pathway, involved in oxidative stress response and cellular growth arrest of both immune and non‐immune cells [42]. Compared to healthy controls, patients had an intense cell infiltration in the lesional skin associated with greater proportions of leukocytes and non‐immune senescent cells expressing both KLRG‐1 and ATF (Figure 4A,B,C and E). Gene expression analysis of KLRG‐1 and ATF genes also supports these findings, where an increase of sevenfold and fivefold over the HC group, respectively, was found (Figure 4D,F). Furthermore, we correlated the size of the skin lesions in patients with the extent of ATF or KLRG‐1 cells infiltration. Interestingly, the accumulation of both senescent markers strongly correlates with lesion size observed in patients (Figure 4G,H), supporting previous findings about senescent cells accumulation in the skin as well as suggesting that these cells have a greater impact on the immunopathology in the skin of patients with CL [27, 43].

FIGURE 4.

Senescent cells accumulate in CL lesions and correlate with the disease severity. Immunofluorescence staining and cumulative data of the expression of (a, c) KLRG‐1 and (b, e) ATF in the lesions of patients with CL (LCL) and in healthy skin (HC). (d, f) Fold changes of KLRG‐1 (7·18‐fold increase) gene and the ATF (5·67‐fold increase) gene with highest upregulation in CL lesions. (g, h) Pearson's correlation test between the expression of senescence markers and lesion size (mm²). The graphs show the mean ± SD. Scale bar: 100 µm

DISCUSSION

In this work, we used transcriptome profiling analysis to evaluate whether cell senescence is involved in the lesional signature and the pathology of cutaneous leishmaniasis. Thus, we now extend the knowledge of skin immunity in CL by identifying a key set of immunosenescence transcripts and pathways in the lesional environment during CL.

Infection of the skin by L. braziliensis promotes wide tissue remodelling and damage, with an intense leukocyte inflammatory infiltrate composed by immune cells [44]. We have recently demonstrated that circulating senescent T cells (Tsen) with high inflammatory profile and increased propensity to migrate to the lesional skin accumulate during infection with L. braziliensis [28]. Indeed, the accumulation of lesional senescent CD8 T cells positively correlated with lesion size, suggesting the involvement of these T cells in mediating the pathology of this disease [27, 28].

Senescent cells accumulate in many tissues during ageing, persistent antigenic stimulation and chronic infections, including cutaneous leishmaniasis [45, 46]. An interesting observation is that although they do not proliferate, these cells are still metabolically active and secrete a wide array of cytokines, chemokines and matrix metalloproteases, nominated as senescence‐associated secretory phenotype (SASP). This pronounced pro‐inflammatory secretome attract leucocytes, potentiates chronic inflammation and tissue damage, causing pathology in a similar way to those observed in cutaneous leishmaniasis [27, 47].

Activation of senescence pathways causes the alteration of immune and non‐immune cell secretory phenotypes [48, 49]. Our RNA‐seq analysis confirms this, where most of the well‐established senescence‐associated genes were found overrepresented in the lesions, including the serine‐protein kinase ATM, cell cycle regulator p16, p21 and p38 MAP kinase. We also observed an overrepresentation of other markers such as KLRG‐1, EOMES and Tbet that are associated with cell activation processes. Although this may be a limitation of this work, these markers are well‐established features to define immunosenescence when used in association with other markers, as demonstrated here [41, 48, 50].

In a deeper investigation of the host response, we used the ImmQuant software to infer the cell type composition of the bulk skin transcriptome [32]. We identified a key set of both senescence pathways and SASP‐related transcripts in CL lesions that positively correlated with predicted compositions of senescent cells and monocytes. These populations that includes monocytes/macrophages can share multiple phenotypic characteristics including the SASP secretome. Moreover, inflammation and senescence pathways and signs that share many components may corroborate to the increase of those genes in monocytes population, as found in our analysis [51, 52]. The local accumulation of cells with senescence features correlated positively with the lesion severity observed in patients. These observations support our previous results and sustain the hypothesis that lesional senescent cells are key for mediating immunopathology of CL through an exacerbated inflammatory immune response. Still related to pathology, although our data has mostly emphasized a strong transcriptional signature of immune cells as indicated by ImmQuant software analysis, we do not rule out the participation of non‐immune as well as non‐senescent cell populations at this process, which offers the basis for further complementary studies. Overwhelming inflammation is one of the main signatures of CL skin pathogenesis, where lesional transcriptomic analysis reveals a strong signature of inflammasome activation and release of IL‐1β, TNF‐α and IFN‐γ [5, 39, 53, 54]. Now we gather to these other cytokines, chemokines and growth factors, including a wide SASP secretome. Although non‐senescent cells can contribute to numerous inflammatory SASP‐related factors, most commonly reported SASP factors are expressed in response to replicative‐, oncogene‐ and DNA damage‐induced senescence that support the role of senescent cells to the lesional inflammation [55, 56]. In this scenario, SASP may contribute to induce senescence of adjacent cells in a paracrine manner by building a positive feedback loop, gradually exacerbating the inflammatory effects on tissues. In addition, several SASP components can also directly modulate the expression of stress ligands including those that bind to NK receptors by immune and non‐immune cells [57, 58]. The interaction between these ligands and NK receptors such as NKG2D on T/Tsen with NKRL‐expressing cells in the skin may lead to a cytotoxic killing mechanism with non‐specific tissue damage that amplifies the overwhelming inflammation, contributing to the progressive growth of CL lesions [59, 60].

In CL, the development of the classical ulcer occurs concomitantly with the local inflammation [54, 61]. Although there are few parasites, their presence in the skin tissue is not correlated to the size of the lesions observed in patients. Recent data show that the lesional parasite load directly correlates with the transcriptional signature of inflammatory, cytotoxic mediators and accumulation of cytotoxic CD8⁺ T cells [61], and these are the main responsible for the tissue pathology [14, 29]. The accumulation of senescent and non‐senescent cells in the cellular infiltrate as well as local inflammation may clarify the correlation observed between the parasites and senescence signature.

Moreover, the local accumulation of senescent cells may generate a destructive positive feedback loop with progressive and unspecific tissue damage and persistent inflammation, unable to eliminate parasites from the lesion but wieldy in promoting disease severity [61]. Therefore, controlling SASP/senescence pathways and/or the accumulation of senescent cells may also have protective effects, dampening local tissue damage.

Overall, the present study extends the understanding of the local senescence signature that occurs in the context of L. braziliensis infection. Moreover, our analyses provide information on how senescent cells may corroborate the progressive pathology in the skin lesions. Controlling senescence or the positive inflammatory feedback in this process may be the most effective strategy for targeting an efficient anti‐Leishmania immunity and laying the foundation for new therapies in leishmaniasis.

CONFLICT OF INTEREST

The authors state no conflict of interest.

AUTHOR CONTRIBUTIONS

CHF analysed RNA‐Seq the data. RGM and LPC carried out skin biopsies experiments and results analysis. CFA, PS and HLMG participated in analysis and discussion of the data and experiments. DCOG, LPC, AA, DM and CHF devised the study and discussed the data. AF selected the patients. DCOG, CHF, LPC, RG, AA and DM wrote the manuscript with the other co‐authors’ support. All authors contributed and approved the submitted version of the article.

Supporting information

Figure S1

Figure S2

Figure S3

Figure S4

Figure S5

Table S1

Table S2

Table S3

Table S4

Supplementary Material

Fantecelle CH, Covre LP, Garcia de Moura R, Guedes HLDM, Amorim CF, Scott P, et al. Transcriptomic landscape of skin lesions in cutaneous leishmaniasis reveals a strong CD8⁺ T cell immunosenescence signature linked to immunopathology. Immunology. 2021;164:754–765. 10.1111/imm.13410