Evidence for involvement of TNFR1 and TIMPs in pathogenesis of post-kala-azar dermal leishmaniasis

Abstract

Semi-quantitative RT–PCR was exploited to analyse the intralesional cytokine gene expression in 14 post-kala-azar dermal leishmaniasis (PKDL) and 10 kala-azar (KA) patients. The data provided evidence for both inflammatory and non-inflammatory responses, as reflected by elevated tumour necrosis factor (TNF)-α and interleukin (IL)-10 in PKDL lesions compared with normal skin tissue (n = 6). The ratio of TNF-α : IL-10 message was 2·66 in PKDL cases, substantially higher than in KA (1·18). Investigation of TNF-α receptors (TNFR1 and TNFR2) revealed a significant down-regulation of TNFR1 transcript in both PKDL and KA compared with control. In the presence of elevated levels of TNF-α transcript, interference with type 1 effector activity in PKDL may be due to minimal expression of the TNFR1 gene. Investigation of matrix metalloproteinases, known to be induced by TNF-α, and the tissue inhibitors of matrix metalloproteinases (TIMPs), provided evidence for the roles of TIMP-1 and TIMP-3 in the pathogenesis of PKDL.

Introduction

Leishmaniasis constitutes a globally widespread group of neglected diseases caused by an obligatory, intracellular, protozoan parasite of genus Leishmania. Visceral leishmaniasis (VL), also known as ‘black fever’ or kala-azar (KA) in Asia, results in the pentad of syndromes, comprised of irregular fever, loss of weight, splenomegaly, hepatomegaly and/or lymphadenopathy and anaemia, and has a high mortality if left untreated. The annual incidence of VL is 500 000, of which 90% of cases are found in five countries – Bangladesh, Brazil, India, Nepal and Sudan [1]. In India, the states of Bihar, Uttar Pradesh and West Bengal are highly endemic foci of VL, where periodic epidemics are common. The disease is spreading to other parts of India, as evidenced by reports of new foci in the hilly region of Uttarakhand, Himachal Pradesh and in the plains of Gujarat, due probably to increased human travel and environmental changes [2–4]. A small percentage (5–15%) of apparently cured VL patients and, occasionally, people from a VL-endemic area without a history of VL, develop a dermal manifestation termed post-kala-azar dermal leishmaniasis (PKDL). It is an unusual dermatosis that produces gross cutaneous lesions in the form of hypopigmented macules, erythema and nodules. The disease is seen mainly in the Indian subcontinent and parts of Africa, where the causative agent for VL is Leishmania donovani[5].

An effective vaccine against Leishmania infections is yet to be developed; current control measures rely upon parenteral administration of sodium antimony gluconate, which has been the cornerstone of treatment for the last six decades. However, it is fraught with problems, such as high toxicity or drug resistance. Furthermore, with the identification of new foci of leishmaniasis in India, the parasite is now considered an agent for diverse clinical manifestations, such as visceral (KA) and dermal (PKDL) caused primarily by L. donovani and cutaneous leishmaniasis caused primarily by L. tropica, but occasionally by L. donovani[3]. The same species of the parasite may lead to different disease manifestations as the host has the capacity to mount a differential immune response against the parasite, either by remaining an asymptomatic carrier clearing the infection or establishing disease, implying that the host immunodeterminants have a major role in determining the severity and progression of the disease [6].

The PKDL develops as a sequel of KA in apparently cured KA patients, but the precise immunological cause remains obscure. It is suggested that there may be immune suppression, allowing renewed multiplication of latent parasites from the viscera or reinfection in the skin [7]. Studies from our laboratory have shown that polymorphism in a specific genetic locus among the parasites isolated from KA and PKDL patients may account for the change in tissue tropism [8]. Further, we have a1so documented significant levels of interferon (IFN)-γ, tumour necrosis factor (TNF)-α, interleukin (IL)-10, transforming growth factor (TGF)-β and IL-6 in localized lesion tissues of PKDL patients and TNF-α as well as nitric oxide (NO) in serum samples [5,9,10]. In spite of the presence of effector molecules such as IFN-γ, TNF-α and NO during active disease, the parasite persists, implying that the biological processes involved in the disease pathogenesis are complex and cannot be interpreted as simple T helper 1 (Th1) or Th2-mediated processes, characteristics of the murine model of leishmaniasis.

The TNF-α is an inflammatory cytokine involved in immune regulation and resistance to various micro-organisms. It exerts a wide range of biological activities, including proliferation and differentiation, apoptosis, cytotoxicity, inflammation, immunomodulation and the production of chemokines and matrix metalloproteinases (MMPs) [11–13]. TNF-α provides signals to target cells through two different but structurally homologous homodimeric receptors: TNF receptor 1 (known also as TNFR1, CD120a, p55 or p60) and TNF receptor 2 (known also as TNFR2, p75, p80 or CD120b). Both TNF receptors (TNFRs) are type I transmembrane glycoproteins (gp) and members of the TNFR superfamily. The involvement of soluble TNFR1 and TNFR2 in Brazilian patients with KA is well documented [14]. Further, TNF-α is considered a potent inducer of MMP-9. MMPs form a family of zinc-containing proteases, implicated in tissue remodelling and chronic inflammation. They possess broad and overlapping specificities and, collectively, have the capacity to degrade all the components of the extracellular matrix [15,16]. MMPs are produced by many cell types, including lymphocytes and granulocytes, but in particular by activated macrophages [17]. MMPs are secreted as proenzymes, which are activated by proteolytic cleavage and regulated by a family of inhibitors called the tissue inhibitors of matrix metalloproteinases (TIMPs), which are produced constitutively by a variety of cells and recognized to have a potential role in parasitic diseases such as malaria [18]. Further, changes in actual MMPs activity are dependent upon the balance between production and activation of MMPs and the local levels of TIMPs.

Similarities exist between Leishmania gp-63 and members of the MMP, and Leishmania species engineered to express high levels of the surface metalloprotease gp63 have enhanced migration capacity through extracellular matrix [19]. Recently, an in vitro study has provided evidence for the participation of metallopeptidase in hepatocyte damage during L. chagasi infection [20].

Based on indicators of the involvement of MMPs and TNF receptors in Leishmania pathogenesis, the present study was designed to understand the modulation of these immunomodulatory molecules in localized infected tissue of KA bone marrow aspirates (BMA) and dermal lesion tissue of PKDL patients, which have provided evidence for their possible role in disease pathogenesis.

Material and methods

Study subjects

Dermal tissue lesions and BMA were collected from 14 PKDL and 10 KA patients with proven L. donovani infection, 1 week prior to the start of treatment, as described previously [5]. Of 14 PKDL cases included for the study, 12 had a history of KA, ranging from 1·5 years to 20 years (mean ± standard error, 9·39 ± 1·85 years), while two patients were not aware of a history of KA. The major characteristics of patients are given in Table 1. All patients were human immunodeficiency virus seronegative. Normal dermal tissue from six PKDL patients was used as control. The study was performed between July 2005 and December 2007. All patients originated from Bihar and reported to the Departments of Medicine or Dermatology, Safdarjung Hospital, New Delhi. The present study was approved by the Ethical Committee of Safdarjung Hospital on Human Subjects and informed consent was obtained from all patients.

Table 1.

Major characteristics of the study subjects, composed of kala-azar (KA), post-KA dermal leishmaniasis (PKDL) and controls.

Characteristic	Patients		Control*
	PKDL	KA
Subjects (n)	14	10	6
Age, years
Range	10–65	6–55	19–65
Mean ± standard error	31·29 ± 4·62	23·1 ± 5·18	29·17 ± 10·31
No. of males/no. of females	12/2	6/4	6/0

^*Control samples were collected from normal skin of patients with PKDL.

RNA isolation

Total RNA was isolated from normal and dermal lesion tissue samples, using Trizol reagent (Invitrogen, Carlsbad, CA, USA), followed by treatment of RNA with DNase (Sigma-Aldrich, St. Louis, MO, USA) as per the manufacture's instructions and stored at −70°C until use. The quantity and quality of the RNA was determined by spectrophotometer and agarose–formaldehyde gel electrophoresis.

cDNA synthesis and PCR

Two µg total RNA from each sample was used for cDNA synthesis using the SuperScript™ reverse transcriptase kit (Invitrogen); detailed methodology is given elsewhere [5]. Published sequences for hypoxanthine–guanine phosphoribosyl transferase (HPRT) and cytokine genes, metalloproteinases and tissue inhibitors of metalloproteinases were used in the study [5,21,22]. The primer sequences and estimated length of the PCR products are listed in Table 2.

Table 2.

Oligonucleotide sequence of primers used and PCR product sizes of transcripts amplified in this study.

Gene		Primer sequence (5′→3′)	Product size, base pairs
HPRT	Sense	CGAGATGTGATGAAGGAGATGGG	303
	Anti-sense	CCTGACCAAGGAAAGCAAAGTCTG
IL-10	Sense	TGAGAACCAAGACCCAGACATCAAG	309
	Anti-sense	CCAGATCCGATTTTGGAGACCTC
TNF-α	Sense	AGGCAGTCAGATCATCTTCTC	300
	Anti-sense	TCTTGATGGCAGAGAGGAGG
TNFR1	Sense	ACCAAGTGCCACAAAGGAAC	263
	Anti-sense	CTGCAATTGAAGCACTGGAA
TNF-αR2	Sense	GTT GGA CTG ATT GTG GGT GTG A	454
	Anti-sense	AGG GGC TGG AAT CTG TGT CTC
TIMP-1	Sense	CTTCCACAGGTCCCACAACC	303
	Anti-sense	CAGCCCTGGCTCCCGAGGC
TIMP-3	Sense	CGCTGGTCTACACCATCAAGC	638
	Anti-sense	CAGGAGGATAGTTCCCAATAAACC
MMP-9	Sense	GGCCCTTCTACGGCCACT	134
	Anti-sense	CAGAGAATCGCCAGTACTT

HPRT, hypoxanthine–guanine phosphoribosyl transferase; IL, interleukin; MMP, matrix metalloproteinase; TIMP, tissue inhibitor of matrix metalloproteinase; TNF, tumour necrosis factor; TNFR1, TNF receptor 1.

cDNAs were normalized based on the expression of HPRT products. The reaction mixture (20 µl) for PCR for HPRT, IL-10, TNF-α, TNFR1, TNFR2, MMP-9 and TIMPs (TIMP-1 and TIMP-3) contained normalized cDNA, 200 mmol/l of each dNTP, 1·5 mmol/l MgCl₂, 25 picomoles of each primer and 0·5 U Thermus aquaticus DNA polymerase in PCR buffer (Invitrogen). PCR was performed using the Personal Mastercycler thermal cycler (Eppendorf, Hamburg, Germany). Cycling parameters for TNFR1 and TNFR2 were as follows: initial denaturation 94°C/4 min, 94°C/30 s, 55°C/30 s (for TNFR2 60°C/30 s), 72°C/60 s and final extension 72°C/7 min. Conditions for TIMP-1 TIMP-3 and MMP-9 were as follows: initial denaturation 94°C/2 min, 94°C/30 s, 55°C/30 s, 72°C/30 s and final extension 72°C/5 min. PCR conditions for HPRT, IL-10 and TNF-α are given elsewhere [5]. The number of amplification cycles was 30 for HPRT, 32 for IL-10, 35 for TNF-α, TNFR1 and TNFR2, and 30 for MMP-9, TIMP-1 and TIMP-3.

After the PCR amplification, 2 µl tracking dye was added to the sample, and 10 µl sample products were run on 1% agarose gel containing ethidium bromide (1 µg/ml) in TAE buffer (0·04 M Tris acetate, 0·001 M ethylenediamine tetraacetic acid). A 100 base pairs DNA ladder was used as a molecular marker.

Quantitative analysis of PCR product

The PCR products were scanned using the gel doc system (Alpha Imager, San Leandro, CA, USA), and the intensity of PCR products present in each lane was measured densitometrically using AlphaEase software (Alpha Imager).

Statistical analysis

Data were analysed by Student's t-test and non-parametric tests including the Mann–Whitney U-test (GraphPad Prism 5 software). Differences were considered to be significant when the P-value was <0·05.

Results

Semiquantitative RT–PCR was used to evaluate transcripts of IL-10, TNF-α, TNFR1, TNFR2, MMP-9, TIMP-1 and TIMP-3 in lesion specimens collected from KA and PKDL patients (BMA and dermal lesion tissue respectively).

Cytokine message levels for IL-10 and TNF-α

Message levels for IL-10 and TNF-α were detectable, respectively, in 100% and 92·8% of PKDL patients and in 100% and 80% of KA patients. The expression level of IL-10 was found significantly elevated in PKDL or KA cases compared with control (P < 0·01), whereas expression of IL-10 was comparable between the PKDL and KA groups (P > 0·05), (Fig. 1). In the case of TNF-α, the level was elevated significantly in PKDL compared with KA or control (P < 0·01; Fig. 1). The ratio of TNF-α : IL-10 message was 2·66 in PKDL cases, substantially higher than in KA (1·18) (Fig. 1). In comparison with KA, the level of TNF-α transcript was 2·4-fold higher in PKDL lesions.

Fig. 1.

Transcripts of interleukin (IL)-10 (a) and tumour necrosis factor (TNF)-α (b) in lesion tissues from patients with post-kala-azar dermal leishmaniasis (PKDL; n = 14) and kala-azar (KA; n = 10) and control tissues (n = 6). Normalized cDNA was amplified with respective cytokine primers. PCR products were electrophoresed, and the intensity of signal was determined by densitometry. The graph shows results as an expression index, defined as the ratio of the intensity of cytokine with respect to the hypoxanthine–guanine phosphoribosyl transferase (HPRT) gene. The bars indicate standard errors. P < 0·05 was considered statistically significant.

Cytokine message levels for TNFR1 and TNFR2 in tissue lesions

Message levels for both TNFR1 and TNFR2 were detectable in 100% and 61·58% of PKDL patients and 100% of KA patients. The level of TNFR1 was found significantly down-regulated in both PKDL or KA compared with control (P < 0·01, P < 0·05; Fig. 2), whereas levels were comparable between PKDL and KA cases (P > 0·05). In the case of TNFR2, no significant difference was noted between the three groups, namely, PKDL, KA or control (P > 0·05; Fig. 2).

Fig. 2.

Transcripts of tumour necrosis factor receptor-1 (TNFR1) (a) and tumur necrosis factor receptor-2 (TNFR2) (b) in lesion tissues from patients with post-kala-azar dermal leishmaniasis (PKDL; n = 14) and kala-azar (KA; n = 10) and control tissues (n = 6). Normalized cDNA was amplified with respective primers. PCR products were electrophoresed, and the intensity of signal was determined by densitometry. The graph shows results as an expression index, defined as the ratio of the intensity of cytokine with respect to the hypoxanthine–guanine phosphoribosyl transferase (HPRT) gene. The bars indicate standard errors. P < 0·05 was considered to be statistically significant.

Cytokine message levels for TIMP-1, TIMP-3 and MMP-9 in tissue lesions

Transcripts of TIMP-1, TIMP-3 and MMP-9 were detected in 100%, 84·6% and 92·9% of PKDL cases, and in 100%, 10% and 100% KA cases respectively. The level of TIMP-1 was found significantly elevated in PKDL compared with KA or control (P < 0·05, Fig. 3), whereas the difference in expression of TIMP-1 was not significant between KA and control. Further, the level of TIMP-3 was found elevated significantly in PKDL compared with KA (P < 0·001). However, the levels were comparable between PKDL versus control or KA versus control (P > 0·05; Fig. 3). In the case of MMP-9, the levels were comparable between the three groups (PKDL, KA or control) (P > 0·05).

Fig. 3.

Transcripts of matrix metalloproteinase (MMP)-9 (a), tissue inhibitor of matrix metalloproteinase (TIMP)-1 (b) and TIMP-3 (c) in lesion tissues of patients with post-kala-azar dermal leishmaniasis (PKDL; n = 14) and kala-azar (KA; n = 10) and control tissues (n = 6). Normalized cDNA was amplified with respective primers. PCR products were electrophoresed, and the intensity of signal was determined by densitometry. The graph shows results as an expression index, defined as the ratio of the intensity of MMP or TIMPs with respect to the hypoxanthine–guanine phosphoribosyl transferase (HPRT) gene. The bars indicate standard errors. P < 0·05 was considered to be statistically significant.

Discussion

The treatment of PKDL cases is considered an important component towards control and eradication of KA, as these patients are considered to be the only known reservoir of L. donovani in India. The parasite survives and propagates in the dermis of PKDL patients and contributes towards drug resistance [23], in turn decreasing the efficacy of chemotherapy intervention, thereby making the scenario complex.

It has been shown previously that both IFN-γ and TNF-α mediate host protection against leishmaniasis, exerting synergistic action in killing Leishmania through induction of NO [24]. Further, it has been shown previously that unresponsiveness to Th1 stimuli may be due to the simultaneous presence of IL-10 and TGF-β, both of which have been associated with disease progression or in obstructing cure and counteracting IFN-γ activities [25,26]. In the present study, a similar situation could be predicted as the transcripts of both inflammatory (TNF-α) as well as anti-inflammatory (IL-10) were present simultaneously in lesions of PKDL cases. Interestingly, the ratio of TNF-α : IL-10 was 2·66, implying the preponderance of inflammatory cytokine over anti-inflammatory in PKDL. Inflammation is an essential host response to infectious challenge; however, when excessive, the inflammatory response becomes harmful. High plasma TNF-α has been associated with TNF promoter variants and predicts complications in malaria infection [27]. In Schistosoma infection, urinary tract morbidity was correlated with increased TNF-α and diminished IL-10 production [28]. Further, polymorphism at TNF-α locus has been associated with different clinical outcome of L. chagasi infection [29]. Such studies are warranted in the Indian population to elucidate the association of TNF-α with clinical outcome in Indian leishmaniasis, which may explain why only 10–15% of apparently cured KA patients develop PKDL. On the other hand, this situation is distinct in KA, where the level of inflammatory and anti-inflammatory cytokines was comparable, as evident from the present and our previous observations where no difference was evident in the level of TNF-α compared with controls at either message or protein levels [5,9]. There may be impairment in the production of TNF-α by monocytes (a major producer of TNF-α) in KA patients, as proposed in a recent study [30].

When the transcripts of TNF receptors were evaluated, significant down-regulation of TNFR1 was noted in both PKDL and KA cases, whereas no difference was observed for TNFR2 transcripts in either group. Recently, the down-regulation of TNFR1 on the surface of Chlamydia-infected cells has been demonstrated [31]. Interestingly, TNFR2 was not modulated during C. trachomatis infection, suggesting a selective mechanism underlying surface reduction of TNFR1 [31]. In tuberculosis, IL-6 has been shown to down-regulate the expression of TNF-α membrane receptors [32]; similarly, the high levels of IL-6 in PKDL and KA [5,9] may lead to a lower expression of TNF-α receptors in disease. Both receptors differ only by the presence of a conserved motif in the cytoplasmic tail, called the ‘death domain’[33]. TNFR1 mediates most of the biological properties of TNF-α, such as apoptotic and anti-apoptotic signalling, whereas TNFR2 is involved in anti-apoptotic signalling [34]. As the transcript of TNFR1 was decreased significantly in lesion tissue of both PKDL and KA patients, it would be of interest to investigate at protein level to unravel the concerted association in disease pathogenesis.

Parasites are deposited in the mammalian skin by infected sandflies and thereafter must interact with and overcome a variety of obstacles, including extracellular matrix to establish infection within macrophage phagolysosomes [35]. The Leishmania species of the L. donovani complex are viscerotropic, while how the same species are exclusively cutaneous in PKDL, free from viscera, is a key question that remains to be answered. A zymographic study of metalloprotease activities in extracts of five L. braziliensis strains have revealed distinct metalloprotease profiles and the differences were associated with distinct geographical origin of the strains with distinct clinical presentation [36]. Further, we have demonstrated differential expression of key surface proteins in PKDL isolates possessing the capacity of extracellular matrix degradation and of conferring resistance to complement-mediated lysis; such modifications may have a role in parasite persistence, of which PKDL is a typical example [37].

Among the MMPs, MMP-9 is of particular importance as it degrades type IV collagen, and the expression of Leishmania gp63 is correlated with degradation of type IV collagen [19]. Secondly, TNF-α is considered a potent inducer of MMP-9 [38]. Monocytes and macrophages release MMP-9 which degrades matrix proteins and sheds TNF-α from its circulating or cell-bound precursors [39]. Excessive inflammation following infection may cause tissue damage, and metallopeptidases are implicated in causing immunopathology [40]. In vitro and in vivo studies with Mycobacterium have demonstrated that infection induces MMP-9, a process which is regulated by TNF-α, IL-18 and IFN-γ[41]. Simultaneously, the role of MMPs in the pathogenesis of malaria and leprosy is well documented. [42,43], whereas little is known about MMP production during Leishmania infections and the contribution they make to immunity versus pathology, although there is evidence for the participation of metallopeptidase in hepatocyte damage during L. chagasi infection [20]. In the present study the expression of MMP-9 was found comparable between the three groups (lesions of KA, PKDL or control), possibly involving other MMPs, which remain to be investigated. Recently, significantly high levels of MMP-8 in patients with severe and uncomplicated malaria compared with control, as well as no differences in the level of MMP-9 between patients and control, have been demonstrated [18].

The MMPs play an important role in a variety of physiological and pathological conditions; however, the situation becomes pathological when the normal balance between the MMPs and their inhibitors, TIMP, is lost [44]. The transcript of TIMP1 was found elevated significantly in PKDL compared with KA or control, whereas TIMP3 was elevated significantly in PKDL compared with KA. Significant levels of TIMP1 and TIMP3 in PKDL in comparison with KA suggest that both molecules may have an important implication for distinct PKDL pathology.

High TIMP-1 level in PKDL raises the possibility of its counteracting activity on MMP-9, as it is known to be a major inhibitor of MMP-9. Simultaneously, it has also been demonstrated that IL-10 inhibits metalloproteinase and stimulates TIMP-1 production in human mononuclear phagocytes [45]. Further, it has also been demonstrated that TGF-β suppresses TNF-α-induced MMP-9 expression in monocytes [46]. The role of IL-10 and TGF-β in KA and PKDL is well documented [5,9], and a consistent correlation between IL-10 levels and the development of PKDL in Sudanese KA patients has been established [47]. The role of TIMP-3 in reduction of metastasis in a human breast cancer cell line, as well as the inhibition of invasion and induction of apoptotic cell death of cancer cell lines by over-expression, is well documented [48,49]. Based on the above finding, we could hypothesize the possible role of TIMP-3 in confinement of the parasite in the dermis of PKDL patients, contrary to the situation in KA, where TIMP-3 levels were low and there is abundant translocation of parasites to visceral organs such as spleen, bone marrow, liver, etc.

The present data provide evidence for the involvement of TNFR1, TIMPs and MMPs in the clinical outcome of PKDL. Further studies are necessary to clarify the role of metallopeptidases, their inhibitors and TNFR1 to unravel the concerted association of the pathogenesis of KA and PKDL.