Evidence for Involvement of Th17 Type Responses in Post Kala Azar Dermal Leishmaniasis (PKDL)

Background Post kala-azar dermal leishmaniasis (PKDL), a dermal sequel of visceral leishmaniasis, caused by Leishmania donovani, constitutes an important reservoir for the parasite. Parallel functioning of counter acting immune responses (Th1/Th2) reflects a complex immunological scenario, suggesting the involvement of additional regulatory molecules in the disease pathogenesis. Methodology/Principal Findings In the present study, human cytokine/chemokine/receptor specific cDNA array technique was employed to identify modulations in gene expression of host immuno-determinants during PKDL, followed by evaluation of Th17 type responses by analyzing mRNA and protein expression of Th17 markers (IL-23, IL-17, RORγt) and performing functional assays using Leishmania antigen (TSLA) or recombinant (rec)IL-17. Array analysis identified key immuno-regulatory molecules including cytokines (TNF-α, IFN-γ, IL-10, IL-17), chemokines (MCP-1, MIP-1α), apoptotic molecules (FasL, TRAIL, IRF-1) and receptors (CD40, Fas). Up regulation in lesional expression of Th17 markers was observed during PKDL compared to control (IL-17 and IL-23, P = 0.0008; RORγt, P = 0.02). In follow-up samples, chemotherapy significantly down regulated expression of all markers. In addition, lesional expression of IL-17 was confirmed at protein level by Immuno-histochemistry. Further, systemic presence of Th17 responses (IL-17 and IL-23) was observed in plasma samples from PKDL patients. In functional assays, TSLA stimulated the secretion of IL-17 and IL-23 from PBMCs of PKDL patients, while recIL-17 enhanced the production of TNF-α as well as nitric oxide (NO) in PKDL compared to control (TNF-α, P = 0.0002; NO, P = 0.0013). Further, a positive correlation was evident between lesional mRNA expression of IL-17 and TNF-α during PKDL. Conclusion/Significance The results highlight key immune modulators in PKDL and provide evidence for the involvement of Th17 type responses in the disease pathogenesis.


Introduction
Visceral leishmaniasis (VL) or kala azar (KA), a potentially fatal protozoan disease caused by the members of Leishmania donovani complex, is endemic in 62 countries; with 200 million people at risk and an estimated 500,000 new cases worldwide annually [1,2]. A small percentage of apparently cured VL patients and occasionally, persons from endemic areas without a history of VL, develop a dermal manifestation known as post kala-azar dermal leishmaniasis (PKDL), characterized by macular, papular and/or nodular rash. The disease is relatively common in the Indian subcontinent (India, Nepal, and Bangladesh) and East Africa where the causative agent is L. donovani. However, the incidence of PKDL varies from 5-15% in India and to 50-60% in Sudan, the reasons for which are presently unknown [3]. PKDL assumes importance being the anthroponotic reservoir of VL in the Indian subcontinent, and contributing to increasing drug resistance [4,5].
While it is generally accepted that infection with different species of Leishmania leads to the establishment of different clinical forms, the same species of this parasite also leads to different disease manifestation in VL and PKDL demonstrating that the host's immune responses plays a vital role in the disease pathogenesis. Various factors implicated in the development of PKDL include nutrition, genetics, inadequate treatment of VL and immune suppression or reinfection [3]. Studies have suggested PKDL as a drug dependent manifestation since it is reported more frequently in SAG treated VL patients [6,7]. However, cases of PKDL develop even after treatment with other antileishmanial drugs [8].
A major role of immune responses in the development of PKDL is well recognized [9,10]. Antecedently, simultaneous presence of both Th1/Th2 responses with increased ratio of TNF-a/IL-10 and involvement of matrix metalloproteinases (MMPs) and tissue inhibitors of MMPs (TIMPs) was documented in tissue lesions of PKDL patients [11,12]. Further, recent reports documented the presence of T regulatory (Tregs) cells and demonstrated their role in parasite persistence by establishing positive correlation with parasite load in PKDL tissue lesions [13,14].
Th17 cells represent a newly described T-cell subset, characterized by production of IL-17 [15] and require IL-23 for differentiation and maintenance [16]. Th17 cells play a pivotal role in autoimmunity and chronic inflammatory diseases [17] and participate in defense mechanisms against certain pathogens including L. donovani [18,19]. Further, IL-17 was associated with intensity of infiltration and pathogenesis of cutaneous leishmaniasis (CL) [20].
In PKDL, majority of the studies in the past have typically focused on a small subset of genes as suspected immunodeterminants for understanding the immune regulation. In this study, cDNA array technology was employed to obtain a comprehensive picture of immunological scenario inside lesion tissue during active disease. Further, the elevated expression of IL-17 in gene array and clues from studies on CL incited us to investigate Th17 type responses and their role in PKDL. Here, we show that IL-17 production was enhanced during PKDL and contributed to elevated levels of TNF-a and nitric oxide (NO).

Materials and Methods
Patients PKDL patients originating from Bihar, reporting to the Department of Dermatology, Safdarjung Hospital, New Delhi were included in this study (Table 1). Slit-skin smears stained with Giemsa and histopathology of lesion biopsy for detection of LD bodies was performed for diagnosis of PKDL. In addition, QPCR was used to demonstrate Leishmania infection in lesions [21]. HIV positive patients were excluded from this study. Patients were treated with oral Miltefosine (150 mg/day) for 2 months which gave apparent clinical cure in all patients. The healthy individuals, all male, included in the study were from non-endemic area with age range of 18-33 years.

Sample collection
Skin biopsies (using 4 mm biopsy punch) from tissue lesions of PKDL patients were collected for RNA isolation in RNA later (Ambion, Austin, TX) and in neutralized formalin for IHC. Biopsies were collected from face or shoulder region. Follow-up samples were collected from the same site as at pre treatment stage, one month after completion of treatment. Heparinized blood was collected for plasma and PBMCs isolation. Normal skin tissues (n = 6, from the shoulder region) and blood (n = 10) were collected from healthy individuals.

Ethics statement
The study was approved by and carried out under the guidelines of the Ethical Committee of the Safdarjung Hospital, New Delhi, India. All patients and healthy individuals provided written informed consent for the collection of samples and subsequent analysis.

Analysis of mRNA expression using cDNA arrays
Total RNA was isolated from punch biopsy samples collected from PKDL (n = 6) patients and healthy individuals (n = 6) using Trizol (Invitrogen, Green island, NY) method. RNA samples were pooled in equal amount from each individual. Six micrograms of DNA-free RNA from each group was reverse transcribed, in the presence of 50 mCi of a-33 P dATP (specific activity $2000 Ci/ mmol; (Perkin Elmer, San Jose, CA) and gene specific primers for each gene represented on the array. The cDNA microarray (AtlasTM; CLONTECH, Palo Alto, CA) consisted of nylon membranes, spotted with 268 different human genes including those encoding cytokines, chemokines, growth factors, and cellular receptors (http://www.clontech.com/support/tools.asp). Briefly, [ 33 P] dATP-labelled cDNA was column purified and hybridized, at high stringency, to cDNA array overnight at 68uC. Membranes were washed at high stringency and exposed overnight to phosphor screens. Image was captured with phosphorImager Typhoon 9210 and analyzed by Imagequant TL software (Amersham Biosciences, Pittsburgh, PA). The intensity of each spot was corrected for background levels and normalized using the housekeeping genes. Array data was analyzed by summing the duplicated intensity signals for each gene in individual experiments and taking the average of 3 technical replicates. The data was expressed as the ratio of mRNA levels in PKDL and controls.

Author Summary
Post kala azar dermal leishamniasis (PKDL), an unusual dermatosis, develops in 5-15% of apparently cured visceral leishmaniasis cases in India and in about 60% of cases in Sudan. PKDL cases assume importance since they constitute an important human reservoir for the parasite. Host immunological responses, considered as major factors in PKDL development, are poorly understood. Limited studies have been performed to explore the host immune responses and that too, restricted to a few immune parameters. The present study employed cDNA array technique that identified various host immuno-determinants including cytokines, chemokines, apoptotic and signaling molecules which were not reported previously in PKDL. In addition, we showed for the first time that Th17 responses are present during L. donovani infection in PKDL which possibly contributes significantly to disease pathogenesis by inducing TNF-a and nitric oxide production.
Our findings lead to improved understanding of the host parasite interaction in terms of immune responses and pathology in tissue lesions of PKDL.

Cytokine measurement using ELISA
Cytokine levels of IL-17, IL-23 and TNF-a in plasma and supernatants were determined by ELISA (e-Biosciences, San Diego, CA) in accordance with manufacturer's instructions. The values were calculated as the concentrations in the stimulated cultures minus the unstimulated controls.

Immunohistochemistry
Skin punch biopsy was collected in neutralized formalin. The tissue was paraffin embedded and 5 mm sections were prepared on polylysine coated glass slides from all skin specimens. Immunohistochemical staining using anti-human IL-17 (Santa-Cruz, San Diego, CA) was performed as described earlier [14]. The immunohistochemical staining for IL-17 was evaluated semi quantitatively by counting the percentage distribution of lymphocytes in the dermal infiltrate showing cytoplasmic staining by counting at 4006 magnification. It was scored as 0: no staining or less ,10% cells labeled; 1: 11-25% of the cells labeled; 2: 25-50% of the cells labeled; 3: .50% cells positively labeled. Only cytoplasmic staining was considered positive.

PBMCs isolation, stimulation and cytokine production
PBMCs were isolated from PKDL patients at pre-treatment, post-treatment stage and healthy individuals by density gradient centrifugation with Ficoll-Hypaque (Sigma-Aldrich, St Louis, MO). The cells were cultured in RPMI 1640 (Sigma-Aldrich), 10% FCS (Life Technologies, NY), glutamine, HEPES and antibiotics. Briefly, 5610 5 cells were plated in 96 well flat bottom tissue culture plates (Axygen, Union city, CA, USA) and were kept only with media (unstimulated) or stimulated with 10 mg/mL of Phytohemagglutinin (PHA) (Sigma-Aldrich) as a positive control or with total soluble Leishmania antigen (TSLA) (10 mg/mL) or rec IL-17 (50 ng/ml) (Peprotech, Rocky Hill, NJ). After 72 hrs of incubation at 37uC and 5% CO 2 , supernatants were collected and stored at 270uC until further analysis.

Estimation of NO production
Nitrite accumulation, an indicator of NO production, was measured in cell culture supernatants using Griess reagent [22]. Briefly, 50 ml samples of culture supernatants were mixed with an equal volume of Griess reagent (Sigma-Aldrich) and incubated at room temperature for 15 min. The absorbance at 570 nm was measured in a microplate reader. The quantity of the respective nitrite was calculated in ng/ml using NaNO 2 standard curve.

Statistical analysis
Statistical analysis was performed with Mann-Whitney test/ paired t-test using Graph Pad Prism 5 (GraphPad Software, Inc., San Diego, CA). Correlation was evaluated using Spearman's rank correlation test. P values#0.05 were considered significant.

Study subjects
Of the 25 PKDL patients, Leishman-Donovan (LD) bodies were seen in 14 patients and histopathological features were compatible with PKDL in all patients revealing cellular infiltrate consisting of lymphocytes, plasma cells, and macrophages. All samples were positive in Leishmania specific quantitative real time PCR (QPCR). In post-treatment cases, subsidence of indurated lesions leaving normal or wrinkled skin, and histopathological absence of disease activity was considered as cured. Further no parasites were detectable by QPCR in any of post-treatment cases.

Gene expression profile in PKDL
Gene expression analysis using pooled RNA from dermal lesion tissues of PKDL and normal dermal controls was carried out using cDNA arrays. Sixty two genes out of 268 arrayed genes (23.1%) including cytokines, chemokines, receptors and other regulatory molecules, showed modulation 2 fold or more in tissue lesions compared to control. Table 2

IL-17 stimulation enhanced production of TNF-a and NO by PBMCs from PKDL patients
IL-17 is established as a strong inducer of TNF-a production by monocytes [25]. In order to find out the possible association of IL-17 and TNF-a in PKDL, levels of TNF-a (pg/ml) were measured in supernatants from PBMCs of PKDL (n = 8) and healthy individuals (n = 6) stimulated with TSLA or recombinant (rec)IL-17. Both TSLA and recIL-17 stimulated production of TNF-a (815.4666.01, 530.5673.68) which was significantly elevated in PKDL compared to control (81.8465.32, 65.3865.08, P = 0.0007 and 0.0002) (Figure 5b). In response to TSLA, a significant increase was observed in the level of TNF-a in post-treatment samples compared to control (P = 0.002, while the levels were comparable between pre and post-treatment stage (699.5645.39, P = 0.28). In response to recIL-17, a significant decrement was observed in the level of TNF-a in post-treatment samples compared to pre-treatment samples (70.16612.21, P = 0.0007). The levels were comparable between post-treatment samples and controls (P = 0.81) (Figure 5b). Since IL-17 is an inducer of NO production in various cell types [26], we also assessed the influence of IL-17 on NO production. In PKDL, both TSLA and IL-17

Transcripts of TNF-a were positively correlated with IL-17 transcripts
We measured mRNA expression of TNF-a in PKDL tissue lesions and investigated for any association between TNF-a and IL-17 expressions in the same set of patients. The message levels of TNF-a were found significantly elevated at pre-treatment stage compared to control (P = 0.0008) and decreased with chemotherapy at post-treatment stage (P,0.0001). Further analysis established a positive correlation between mRNA expression of IL-17 and TNF-a in tissue lesions of PKDL patients (r = 0.77) ( Figure 6). Such correlation was not observed for IL-23 or RORct.

Discussion
During the interaction with host cell, parasites attempt to take over cellular functions to their advantage while the host cell counteracts by mounting a variety of defensive responses that include induction of various cytokines, chemokines or other regulatory molecules, all of which are accompanied by changes in gene expression. In the present study, we document the modulation in gene expression profile of various cytokines, chemokines, receptors and apoptotic molecules during L. donovani infection in PKDL. The expression of cytokines IFN-c, TNF-a, IL-10 and TGF-b were found elevated, similar to previous reports [11], with TNF-a showing the highest expression (9 fold) in cytokine category. Further, chemokines including IL-8, CXCL-2, MIP-1a, MIP-1b and MCP-1, were found up regulated, all of which are known to primarily activate the microbicidal activity, by promoting chemotaxis of neutrophils and other target cells. Similarly, in Leishmania infection, MCP-1 and MIP-1a have been associated with lesional macrophage infiltration, enhancement of nitric oxide production and stimulation of leishmanicidal activity [27]. Higher expression of chemokines in PKDL implies that these may have a role in shifting the parasite from viscera to dermis, in keeping with a report demonstrating the presence of chemokine binding molecules on the surface of the kinetoplastid parasite [28]. Secondly, as MIP-1a and MIP-1b are potent chemo-attractants for monocyte/macrophages, these may be important in attracting the uninfected immature macrophages. Such cells would provide a safe shelter for the parasite as they get infected but do not kill the parasite, thus contributing towards persistence of parasites.
CD40, a member of the tumor necrosis factor receptor superfamily, interacts with CD40L on activated T cells and helps in immune activation during bacterial, fungal, parasitic and viral infections [29]. In our study, both CD40 and CD40L showed a moderate increment in gene expression in active PKDL. Studies in context with Leishmania have demonstrated that the interaction of CD40 with CD40L activates the macrophages resulting in killing of amastigotes [30] but at the same time, a weak CD40 signal promotes IL-10 and reduces IL-12 production [31]. Therefore, it remains a paradox whether CD40 induces host protection or disease promotion.
Another interesting molecule modulated in PKDL was interferon regulatory factor (IRF)-1. It was originally discovered as a protein that binds to DNA sequences termed IRF-Es [32]. It is required for in vivo generation of Th1 responses through IL-12 production [33]. Infection of human dendritic cells with L. major activated IRF-1 resulting in IL-12 production [34]. Furthermore, IRF-1 acted as a mediator of cytokine-induced apoptosis as hepatocytes from IRF-1-deficient mice were completely resistant to apoptosis induction by IFN-c [35]. A pronounced expression of IRF-1 pointed towards its involvement in PKDL pathogenesis; however its function during L.donovani infection in humans is not known.
In addition to prevalence of Th1/Th2 responses, elevated expression of IL-17 (5 fold) in array results pointed towards the presence of Th17 type responses in PKDL. IL-17 is involved in regulating tissue inflammation, development of autoimmune  [20,38]. IL-23, a heterodimeric cytokine, is associated with the induction of Th17 cells [39]. Products from microorganisms, including bacteria, intracellular parasites, viruses and fungi are strong inducers of IL-23 production in macrophages, monocytes, neutrophils, and DCs [40]. M. tuberculosis infection in human DCs induced IL-23 production which further contributed to IL-17 production from CD4 + cells [41]. In this study, IL-23 expression was found up regulated and PBMCs from PKDL patients were able to secrete IL-23 in response to Leishmania antigen. In addition to IL-23, generation of human Th17 cells is dependent on IL-1b, TGF-b and IL-6 [19] all of which were found up regulated (5.1, 2.9 and 2.1 fold respectively) in PKDL (Table 2). Thus, we suggest that abundance of IL-17 in PKDL could be due to enhanced IL-23 level which, in association with IL-6, TGF-b and IL-1b, induces differentiation to IL-17 secreting Th17 cells. In follow-up cases, after stimulation with TSLA, the Th17 cytokines levels were comparable to pre treatment stage which may be a memory T cell response. Similar to our results, antigen specific Th17 cytokine responses have been demonstrated in PBMCs from CL and ML patients [20]. In contrast to PKDL, a defective Th17 response in VL against whole parasite or soluble antigen is reported [18]. Possible discrepancy between VL and PKDL could be context dependent due to different niche and clinical manifestation.
To our knowledge, this is the first report demonstrating up regulation of IL-17 during PKDL, which further tended our interest to investigate the role of this cytokine. IL-17 is a known inducer of TNF-a production from endothelial cells, epithelial cells, monocytes and macrophages [25,42]. Increased production of TNF-a in PBMCs on addition of external IL-17 confirmed the impact of IL-17 towards enhanced TNF-a level. In contrast to PKDL, low abundance of IL-17 has been demonstrated recently in VL [43] that explains the impairment in the production of TNF-a by monocytes in VL [44]. TNF-a is functionally linked to the Th17 pathway by its ability to activate myeloid dendritic cells that synthesize IL-23 and other regulators of T-cell development [45,46]. Thus, TNF-a may serve as an indirect activator of Th17 responses and a significant positive correlation between these two molecules, as evident in our study, indicated their interrelationship. In addition, IL-17 stimulates NO production in mammalian cells [26] and an elevated level of NO has been shown in PKDL serum samples [47]. Stimulation with IL-17 enhanced the NO level in culture supernatants in PBMCs, establishing the association of IL-17 with elevated NO level. In CL, different studies have shown presence of Th17 responses, up regulation in iNOS expression and increase in NO serum levels [20,48,49], however, association between these molecules has not been reported. Here, we demonstrated a direct functional association of IL-17 and NO production during PKDL.
Recently, we have demonstrated accumulation of lesional Tregs and their correlation with parasite load in PKDL [14].
Here we provide evidence for enhanced Th17 responses in the same. Thus, we show a paradoxical gene expression signatures associated with inhibition of the immune response (Tregs) and the proinflammatory response (IL-17, TNF-a) in PKDL. However, no correlation was observed between parasite load and cytokine mRNA levels of Th17 markers. Previously, a number of studies have identified coexistence of Th17 and Treg cells in various diseases [50][51][52][53]. Considering the association of IL-17 with resistance to L. donovani infection [18], the data suggests that during PKDL, parasite induces production of Tregs that counteract inflammatory responses (IL-17, TNF-a and NO) by secreting IL-10 and promoting parasite persistence. It has been shown that adoptive transfer of Tregs inhibits ex vivo inflammatory Th17 responses in humans [54], although during active tuberculosis Tregs facilitated mycobacterial replication by inhibiting Th1 rather than Th17 responses [51]. In case of PKDL, further investigations are needed to explore cross talk between these counter acting responses.
Taken together, this study reveals that L. donovani infection has a remarkable effect on host cellular gene expression, and identified several new molecules whose implication remains to be defined in context with critical and non-overlapping functions of immunomodulators in regulating immunity in PKDL. Further, this study demonstrated enhanced expression and antigen specific stimulation of Th17 cytokines and the involvement of IL-17 in PKDL pathogenesis by demonstrating its association with TNF-a and NO production. Based on longitudinal follow-up studies on L.donovani infected humans [18] and present results, we suggest that enhanced Th17 responses may have a role in parasite clearance during PKDL. Considering the plasticity between Tregs and Th17 [55], it would be of great interest to investigate reciprocal regulation of these cells in PKDL. Improved mechanistic insight into the collaborative interaction between these immuno-determinants during PKDL would undoubtedly pave way for filling up gaps in our basic knowledge and contribute in developing strategies for disease control.