The first 2 authors contributed equally to this paper.
Expression and activity of IL-17 in cutaneous T-cell lymphomas (mycosis fungoides and sezary syndrome)
Article first published online: 2 JUN 2004
Copyright © 2004 Wiley-Liss, Inc.
International Journal of Cancer
Volume 112, Issue 1, pages 113–120, 20 October 2004
How to Cite
Cirée, A., Michel, L., Camilleri-Bröet, S., Jean Louis, F., Oster, M., Flageul, B., Senet, P., Fossiez, F., Fridman, W. H., Bachelez, H. and Tartour, E. (2004), Expression and activity of IL-17 in cutaneous T-cell lymphomas (mycosis fungoides and sezary syndrome). Int. J. Cancer, 112: 113–120. doi: 10.1002/ijc.20373
- Issue published online: 9 AUG 2004
- Article first published online: 2 JUN 2004
- Manuscript Accepted: 18 MAR 2004
- Manuscript Received: 14 NOV 2003
- Fondation de France
- Association pour la Recherche sur le Cancer. Grant Number: 970RC24
- Ligue Nationale contre le Cancer
- Université René Descartes
- tumor microenvironment;
- polymorphonuclear neutrophil
Interleukin-17 (IL-17) is a proinflammatory cytokine mainly produced by activated CD4+ CD45RO T cells. In mice, we have demonstrated that, depending on the model, IL-17 may act as a tumor growth-promoting or -inhibiting factor. In order to address the relevance of these models in human tumors, we look for the natural expression and activity of IL-17 in mycosis fungoides (MF) and Sezary syndrome (SS). These cutaneous T-cell lymphomas were selected because they are usually CD3+ CD4+ CD45RO+, a phenotype similar to nontransformed T cells producing IL-17. We show that in vitro activated malignant T cells derived from MF or SS patients express IL-17 mRNA and secrete this cytokine. However, IL-17 does not act in vitro as a growth factor for MF or SS cell lines. In addition, 5 out of 10 MF/SS biopsies expressed IL-17 mRNA, while this cytokine was not detected in normal skin. IL-17 was not observed in the biopsies derived from 2 patients initially identified as MF, whereas an upregulation of this cytokine was clearly demonstrated during progression of the disease in these patients. An association between IL-17 expression and polymorphonuclear neutrophil infiltration was also recorded in this group of MF/SS patients. A more detailed analysis of 2 patients with a pustular form of MF and SS revealed that IL-17 may participate in the recruitment of polymorphonuclear neutrophils via a paracrine mechanism involving keratinocyte-released IL-8. This study is the first report demonstrating that some human tumor cells could express IL-17, a cytokine that represents an early event in the development of the inflammatory reaction within the tumor microenvironment, a process that may influence tumor phenotype and growth. © 2004 Wiley-Liss, Inc.
Interleukin-17 (IL-17), originally identified by Rouvier et al.1 as cytolytic T-lymphocyte (CTL)-associated antigen 8, is a T-cell-derived cytokine with proinflammatory activity. It has a high homology with the open reading frame 13 (ORF13) of herpesvirus saimiri, also called vIL-17, and is now included in a novel emergent cytokine family with at least 6 members in the human genome (IL-17B, IL-17C, IL-17D, IL-17E, IL-17F).2
Members of the IL-17 family are all expressed as dimers and they exhibit the greatest similarity within a C-terminal stretch of 70 amino acids, where there are 4 highly conserved cysteines that participate in the formation of intrachain disulfide bonds.3 IL-17 is secreted mainly by activated CD4+ CD45RO+ memory helper T cells belonging to both type 1 or type 2 T cells regarding their secretion of cytokines.4, 5, 6 Microbial peptides, IL-2, IL-15 and IL-23 enhance the production of IL-17 by CD4+ T cells.7, 8, 9
Reduction of delayed-type hypersensitivity and T-dependent antibody production in IL-17-deficient mice strongly suggest that IL-17 plays an important role in T-cell responses.10 IL-17 is considered to be a proinflammatory cytokine because it increases IL-6 and IL-8 production by macrophages, fibroblasts, synovial cells and tumor cells.4, 5, 11, 12 It also induces nitric oxide production by human osteoarthritic cartilage13, 14 and stimulates the release of tumor necrosis factor (TNF)-α and IL-1β by macrophages and endothelial cells.15
In addition, IL-17 elicits the secretion of granulocyte colony-stimulating factor and CXC chemokines that stimulate granulopoiesis and recruitment of polymorphonuclear neutrophils into tissues.16, 17 Increased levels of IL-17 have been demonstrated in serum and in the diseased organs and tissues of patients with various autoimmune, inflammatory or allergic diseases, including rheumatoid arthritis, multiple sclerosis, systemic lupus erythematosus, asthma and Behcet's disease.18, 19, 20, 21, 22, 23
In previous studies, we have demonstrated that human tumors transfected with IL-17 showed increased growth in nude mice, whereas immunogenic tumors were inhibited in immunocompetent mice by means of a T-cell-dependent mechanism.12, 24 In all these studies, IL-17 was not naturally produced by the tumors but overexpressed secondary to the transfection of cells. In order to address the relevance of these models in tumors spontaneously occurring in humans, we look for the natural expression of IL-17 in mycosis fungoides (MF) and Sezary syndrome (SS), which are primary epidermotropic cutaneous T-cell lymphomas (CTCL). We have selected these tumors because MF and SS cells are usually CD3+ CD4+ CD45RO+, a phenotype similar to nontransformed activated T cells producing IL-17.25 We found that Sezary cells express IL-17, a feature associated with infiltration of polymorphonuclear neutrophils into these lesions. The IL-17-induced release of IL-8 may be partly responsible for neutrophil recruitment.
MATERIAL AND METHODS
Skin biopsy specimens were obtained from 4 patients with MF and 5 patients with SS. Tissues were divided into 2 equal parts: 1 portion was fixed in formalin for histologic examination, the other was snap-frozen in liquid nitrogen and stored at −70°C for RNA extraction and immunocytochemistry. The diagnosis of SS was established on the basis of clinical criteria (exfoliative erythroderma, pruritus, palmoplantar keratoderma, lymphnodes) and laboratory parameters including the presence of typical circulating Sezary cells (number > 1,000/mm3), histologic data (cutaneous epidermotropic T-cell lymphoma) and detection of a T-cell clone in the blood and skin by Vγ-Jγ PCR multiplex-based analysis. Essential characteristics of study patients with CTCL are provided in Table I. Approval was obtained from the institutional review board for these studies. Informed consent was provided according to the Declaration of Helsinki.
|TNM staging||Sezary cells (% of PBMCs)||TCR Vβ (% of CD4+ T-cells)|
|Patient 1 (SS)||T4N2M0||19%||Vβ6 (% N.A)|
|Patient 2 (MF)||T4N0M0||NA|
|Patient 2 (SS)||T4N0M0||9%||NA|
|Patient 3 (MF)||T4N1M0||NA|
|Patient 3 (SS)||T4N1M0||5%||NA|
|Patient 4 (SS)||T4N2M0||25%||Vβ2 (92%)|
|Patient 5 = SS1 (SS)||T4N1M0||7%||Vβ13.1 (91.5%)|
|Patient 6 (MF)||T3N0M0||NA|
|Patient 7 (MF)||T4N0M0||NA|
|SS2 (SS)||T4N1M0||40%||Vβ5.1 (94.8%)|
|SS3 (SS)||T4N2M0||49%||Vβ3 (94.5%)|
After rehydration and inhibition of endogenous peroxidases, 6 μm thin cryosections were incubated with primary-specific or isotype control-irrelevant antibodies for 30 min, then incubated with a multispecies biotinylated second antibody for 10 min before the addition of streptavidin-peroxidase reagent for 10 min (Ultratech HRP Streptavidin-Biotin Universal Detection System). The amino-ethyl-carbazol (AEC) substrate chromogen (Beckman-Immunotech, Marseilles, France) was finally used to reveal antibody fixation.
Sections were counterstained with Mayer's hematoxylin before glycerol mounting. A mouse monoclonal IgG1 antihuman neutrophil elastase (Dako, Trappes, France) diluted to 1/40, a mouse monoclonal IgG1 anti-CD4 (Dako) diluted to 1/70 and a mouse monoclonal anti-IL-8 (Genzyme Diagnostics, Cambridge, MA) diluted to 1/20 were used to detect neutrophils, CD4+ T cells and IL-8 expression, respectively. Isotype control antibodies were included in each experiment. To assess the CD4-expressing area on the sections, a LEITZ DMR (Leica, Deerfield, IL) microscope was used with a 10× objective, operating with a 3CCD Color Video Camera (DXC.93OP, Sony, Tokyo, Japan). The ratio of the stained CD4+ surface area over the total surface area was determined by using the Analysis software (SIS Soft Imaging Software, Reutlingen, Germany). The entire section was analyzed for each case.
Tumor cell lines
The SeAx Sezary cell line, established from the peripheral blood of a patient with SS, and the Myla cell line, established from a plaque lesion of a patient with MF, were both obtained from Reinhard Dummer (Zurich, Switzerland) with the permission of Keld Kaltoft (Institute of Human Genetics, University of Aarhus, Aarhus, Denmark). These cell lines were grown in RPMI-1640 medium (Gibco-Life Technologies, Paisley, U.K.) supplemented with 10% fetal calf serum (Dominique Dutscher, Brumath, France), 100 UI/ml penicillin, 100 UI/ml streptomycin, 5 mmol/l sodium pyruvate, 2 mmol/l glutamine (all from Gibco-BRL, Grand Island, NY). The NCTC 2544 human keratinocyte cell line was purchased from ICN (Orsay, France). These cells were cultured in Dulbecco's modified Eagle medium (DMEM; Gibco-Life Technologies) supplemented in a similar way to the RPMI medium.
Sezary T-cell lines derived from patients
Circulating Sezary cells were purified in 3 patients by using Ficoll-Hypaque density gradient centrifugation and magnetic beads coated with anti-CD4 or anti-TCR Vβ chain of TCR. The anti-TCR Vβ used was that previously identified as the dominant tumor clone. It was selected after TCR repertoire analysis performed by a semiquantitative PCR assay using the 24 different BV family-specific oligonucleotides.26 The monoclonality of these Sezary T-cell lines was assessed by flow cytometry with anti-CD4 and anti-TCR Vβ mAb (Beckman-Immunotech) as well as Vγ-Jγ multiplex-based analysis. The purity of recovered cells ranged from 97% to 99%.
Total cellular RNAs were extracted using the RNA Plus Kit (Quantum-Appligene, Ilkirch, France). Five micrograms of total cellular RNAs were reverse-transcribed with oligo-dT primers using the first-strand cDNA synthesis kit (Roche Molecular Biochmicals, Meylan, France). PCR was performed as previously described.27 The following oligonucleotides were used: β actin sense, TCGTCGACAACGGCTCCGGCATGTGC; β actin antisense, TTCTCCAGGGAGGAGCTGGAAGCAGC; IL-6 sense, ACGAATTGACAAACAAATTCGGTACA; IL-6 antisense, CATCTAGATTCTTTGCCTTTTTCTGC; IL-8 sense, TTCTGCAGCTCTGTGTGAAGGT; IL-8 antisense, GAAGAGGGCTGAGAATTCAT; IL-17 sense, ACTCCTGGGAAGACCTCATTG; IL-17 antisense, GGCCACATGGTGGACAATCG; IL-17R sense, CTAAACTGCACGGTCAAGAATAG; IL-17R antisense, ATGAACCAGTACACCCAC. Each RT-PCR includes a negative control without a cDNA template.
The cells were incubated for 24 hr at 2 × 106 cells per milliliter in medium alone or supplemented with 10 ng/ml phorbol 12-myristate 12-acetate (PMA) and 1 μM ionomycin (Sigma Chemical, St. Louis, MO) or IL-2 (1,000 U/ml) from Chiron (Emeryville, CA) and IL-15 (50 ng/ml) from R&D systems (Minneapolis, MN) or with anti-CD3 (1 μg/ml) from BD Pharmingen (San Diego, CA) and anti-CD28 (5 μg/ml) from BD Pharmingen. IL-17 in the cell-free supernatants was assayed using ELISA kits purchased from BioSource (Camarillo, CA). IL-6 and IL-8 measurements were performed using ELISA kits from Diaclone (Besançon, France). Cells were also stimulated with IL-17, a kind gift from F. Fossiez (Schering Plough, Dardilly, France).
SeAx and Myla cells (5 × 105 cells) were washed in Dulbecco phosphate-buffered saline (PBS; Gibco-Life Technologies) and resuspended in PBS containing 10% fetal calf serum for 30 min at 4°C. The cells were washed and stained with biotinylated anti-IL-17R goat antibody (10 μg/ml; R&D Systems) in PBS with 5% fetal calf serum or biotinylated irrelevant goat IgG (Jackson Immunoresearch, West Grove, PA) on ice. After washings, the cells were incubated with fluorescein isothiocyanate (FITC)-conjugated streptavidin (Dako) for 30 min at 4°C in the dark and then resuspended in 1% paraformaldehyde PBS. The cells were analyzed with FACScalibur (Becton Dickinson, Mountain View, CA).
The proliferative response of the tumor cells to varying concentrations of rhIL-17 was determined by measuring the [3H] thymidine incorporation (cpm) of 5 × 104 cells. This test was carried out in 96-well flat-bottomed plates in 0.2 ml of culture medium. After 4 days, culture wells were individually pulsed with 0.5 μCi of [3H] thymidine for 6 hr. [3H] thymidine incorporation was measured in a microplate scintillation counter (Topcount; Packard Instrument, Meriden, CT).
The chi-square test with Yates correction when necessary was used to analyze the relationship between IL-17 expression and polymorphonuclear neutrophil or CD4+ T-cell infiltration.
Sezary- and MF-derived cell lines express IL-17
SeAx and Myla cells, 2 long-term malignant T-cell lines, were shown to express IL-17 mRNA when activated by PMA-ionomycin and IL-2 plus IL-15. The induction of IL-17 mRNA was early (4 hr) and transient after stimulation with PMA-ionomycin and was observed later when IL-2 and IL-15 were added to the cells (Fig. 1a). In addition, highly purified circulating Sezary cells from 3 patients (SS1, SS2 and SS3) were stimulated with PMA-ionomycin or anti-CD3 plus anti-CD28. PMA-ionomycin treatment induced IL-17 mRNA expression in 2/3 patients tested, whereas anti-CD3 plus anti-CD28 stimulation resulted in an increase in IL-17 mRNA expression in 1/2 patients examined (Fig. 1b). We confirmed the expression of IL-17 at the protein levels as IL-17 was detected in the supernatant of all cell lines expressing IL-17 mRNA after PMA-ionomycin stimulation. (Fig. 1c and data not shown). IL-17 protein was also present at low concentrations in the supernatants of SS2 cells previously activated by anti-CD3 plus anti-CD28 but not in the supernatant of Myla or SeAx previously stimulated by IL-2 plus IL-15 (data not shown). Furthermore, we did not find any evidence of IL-17 transcripts by RT-PCR analysis in other human T-cell lines such as Jurkat or Molt 4 following stimulation with these various mitogens (data not shown).
Expression of IL-17 receptors by MF and Sezary cells
Nonstimulated Myla and SeAx cell lines as well as Sezary cells derived from patients expressed IL-17 receptors both at the mRNA and the protein levels (Fig. 2 and data not shown). No staining was observed when an isotype control antibody was used. These results are in accordance with the ubiquitous expression of the IL-17 receptor.28
Activity of recombinant IL-17 on Myla and SeAx cells
Since cell surface expression of IL-17R was found in Myla and SeAx cells, we investigated whether IL-17 could modulate the phenotype of these cell lines. IL-17 is considered to be a proinflammatory cytokine because it induces the secretion of IL-6 and IL-8 by many tumor cell lines and nontransformed endothelial, fibroblast and epithelial cells.28 When Myla and SeAx cells were incubated in the presence of various concentrations of recombinant human IL-17 for different times ranging from 4 to 48 hr, no increase in IL-6 and IL-8 was observed in the supernatants of the cell lines (Fig. 3a). These results were confirmed at the mRNA levels, when the presence of IL-6 and IL-8 transcripts were assayed by semiquantitative RT-PCR (data not shown).
The same recombinant human IL-17 was efficient to induce IL-6 and IL-8 by tumor cell lines derived from epithelial origin.12 As it has been reported that soluble IL-17 receptor inhibits the proliferation of T cells,28, 29 we looked for the potential effects of IL-17 on the growth rate of Myla and SeAx cells in vitro. As shown in Figure 3(b), IL-17 did not induce any significant change of the growth of SeAx and Myla cells.
Expression of IL-17 in biopsies derived from Sezary and MF patients
Expression of IL-17 mRNA was clearly detectable in 5 out of 9 biopsies derived from patients with MF (n = 4) or SS (n = 5; Fig. 4a). Three were derived from SS patients and 2 from MF patients, while expression was faintly detectable in an additional SS case (patient 4). As a control, IL-17 mRNA was not detected in 3 normal skin samples from healthy adults undergoing reductive plastic surgery (Fig. 4a and data not shown). Moreover, we analyzed IL-17 mRNA expression in sequential biopsies taken from CTCL lesions of 2 patients (patients 2 and 3) at different times during the progression of their disease. IL-17 mRNA expression was not detected when these patients were considered to have a localized disease (MF), whereas IL-17 mRNA was clearly expressed during the disseminated phase (SS) of the disease (Fig. 4a).
We have also analyzed the expression of IL-17 in other inflammatory skin diseases. We have shown a high expression of IL-17 mRNA in the 2 psoriatic skin lesions tested as already published (Fig. 4b).6, 11 We have also demonstrated the presence of IL-17 mRNA in 2/2 lepromatous leprosy skin lesions, whereas no IL-17 mRNA could be detected in the 2 tuberculoid leprosy skin lesions examined.
Since IL-17 is mainly expressed by normal and transformed CD4+ T cells, this heterogeneity of IL-17 mRNA expression in CTCL could be related to differences in the number of malignant CD4+ T cells infiltrating the tumor biopsies. However, we did not find any significant correlation between the level of infiltration of CD4+ T cells within the dermis and the detection of IL-17 mRNA in the biopsies of these patients (Table II).
|Patients||IL-17 expression by RT-PCR||CD4-expressing area|
|Patient 1 (SS)||−||8|
|Patient 2 (MF)||−||7|
|Patient 3 (MF)||−||3|
|Patient 4 (SS)||±||13|
|Patient 2 (SS)||+||1|
|Patient 3 (SS)||+||47|
|Patient 5 (SS)||+||3|
|Patient 6 (MF)||+||13|
|Patient 7 (MF)||+||18|
Relationships between IL-17 mRNA expression and polymorphonuclear neutrophilic infiltration
In this series of CTCL patients, a trend was observed between IL-17 expression and the degree of polymorphonuclear neutrophil infiltration (Table III and Fig. 5). Indeed, IL-17 mRNA expression was observed in skin lesions from all 4 patients who exhibited significant polymorphonuclear neutrophil infiltration (> 5%), whereas only 1 out of 6 patients with low or no polymorphonuclear neutrophil recruitment was shown to express IL-17 mRNA significantly (Table III). However, this relationship did not reach statistical significance possibly due to the limited number of patients analyzed (p > 0.05).
|Patients||IL-17 expression by RT-PCR||% polymorphonuclear neutrophils|
|Patient 1 (SS)||−||2|
|Patient 2 (MF)||−||1|
|Patient 3 (MF)||−||0|
|Patient 4 (SS)||±||1|
|Patient 2 (SS)||+||6|
|Patient 3 (SS)||+||0|
|Patient 5 (SS)||+||25|
|Patient 6 (MF)||+||16|
|Patient 7 (MF)||+||15|
IL-17-induced IL-8 release may contribute to polymorphonuclear neutrophil recruitment in skin lesions in vivo
Patients 5 and 7 exhibited a rarely observed pustular form of CTCL (Fig. 6a and data not shown). As shown in Table III, their skin lesions were heavily infiltrated by polymorphonuclear neutrophils throughout the dermis and epidermis and high levels of IL-17 mRNA were detected in these biopsies. In addition, purified circulating Sezary cells from patient 5 were found to secrete IL-17 (Fig. 1, patient SS1). Since IL-8 has been associated with polymorphonuclear neutrophil recruitment in cutaneous T-cell lymphoma, IL-8 expression was also analyzed by immunocytochemistry. The staining revealed an intense expression of IL-8 by keratinocytes in the 2 patients tested (Fig. 6b and c and data not shown). In the dermis, other cells such as fibroblasts and possibly tumor cells also appeared to be positively stained by anti-IL-8 antibodies. No staining was observed with an irrelevant isotype control antibody (data not shown).
In order to demonstrate a possible relationship between the high expression of IL-8 by keratinocytes with the expression of IL-17 previously reported in these 2 patients (Table III), we incubated the NCTC 2544 human keratinocyte cell line with recombinant IL-17. An increase in the production of IL-8 in the supernatant of IL-17-stimulated NCTC was clearly observed (Fig. 6d).
In this study, we have shown that in vitro activated cells derived from MF/SS patients as well as biopsies derived from this group of patients express IL-17 mRNA. As previously reported, IL-17 mRNA was not detected in nonlesional control biopsies.11 IL-17 protein was also measured in the supernatants of activated long-term established MF cells (Myla) and Sezary cells (SeAx cells) and also on fresh Sezary cells derived from patients.
To our knowledge, the present set of data provides the first evidence of IL-17 expression in CTCLs. Various studies have clearly demonstrated that the lesions in MF/SS preferentially contain IL-4, IL-5 and IL-10, indicating a bias toward a type 2 immune response.29, 30, 31 The preferential expression of IL-17 in lepromatous lesions (type 2-related lesions) also suggests that in some models IL-17 may be gathered with type 2 cytokines. However, as previously reported, IL-17 was also expressed in psoriatic lesions (Fig. 3), a disease associated with increased expression of type 1 cytokines. These results are in line with previous works indicating that IL-17 is secreted by both type 0, type 1 and type 2 T cells.5, 6
The expression of IL-17 in MF/SS biopsies is not directly related to the infiltration of malignant CD4+ T cells, which means that the detection of IL-17 cannot be considered as a marker of the extent of malignant cell infiltration, but it may help to identify a particular subgroup of patients. However, these results must be cautiously interpreted because IL-17 may also be derived from nontransformed CD4 T cells infiltrating the biopsies and certain MF/SS have also been reported to lose their CD4 expression.25 In addition, some infiltrating dendritic cells may express CD4.
Although MF/SS cells express the IL-17 receptor, we did not find any direct effect of IL-17 on the growth of Myla or SeAx cells in vitro. The role of IL-17 on the proliferation of normal T cells is debated. Indeed, a recombinant soluble IL-17 receptor inhibited T-cell proliferation and IL-2 production induced by PHA, concanavalin A (conA) and anti-TCR mAb,28, 29 but this effect of IL-17 on T cells was not found by other groups.4, 30 In other models, IL-17 also did not affect the in vitro proliferation of tumor cells.12, 24, 32 IL-17 therefore does not behave like other cytokines such as IL-2, IL-7 and IL-15, whose receptors share the same γ chain and are considered to be growth factors for normal T cells and Sezary cells.33, 34, 35 This difference may be explained by the fact that the IL-17 receptor does not belong to the cytokine receptor family associated with the γ chain.
Recombinant IL-17 did not increase the levels of IL-6 or IL-8 on Myla and SeAx cells. These results are in line with a previous report demonstrating that upregulation of proinflammatory cytokines by IL-17 was not observed in T cells and seemed to be restricted to epithelial, fibroblastic and endothelial cells together with their malignant counterparts.2 Although we cannot exclude an as yet unidentified autocrine activity of IL-17 on MF/SS cells, a paracrine pathway of IL-17-induced inflammatory changes is more likely in this tumor.
It is noteworthy that IL-17 expression was detectable in all CTCL lesions significantly infiltrated with polymorphonuclear neutrophils (Table III). In other clinical situations, elevated levels of IL-17 were recorded in association with high polymorphonuclear neutrophil numbers in bronchoalveolar lavage fluid during severe inflammation in human airways22 and in joint inflammation in rheumatoid arthritis.36 In mice, the absence of IL-17 receptors is associated with an inability to develop a full polymorphonuclear neutrophil response to Klebsiella pneumoniae.37 Recently, Chung et al.38 showed that abscess formation, mainly composed of polymorphonuclear neutrophils, associated with intraabdominal sepsis is strictly dependent on the T-cell-derived cytokine IL-17.
We have demonstrated an in vitro increase of IL-8 production by an IL-17-treated human keratinocyte cell line as previously reported.11 In addition, using immunohistochemistry techniques, we showed that keratinocytes derived from pustular forms of MF or SS lesions express high levels of IL-8. IL-17 could not be assessed by immunohistochemistry, as anti-IL-17 antibodies are not yet available for this purpose. However, IL-17 mRNA was clearly detected in these lesions and IL-17 protein was also detected in the supernatant of purified Sezary cells from patient 5. This study therefore suggests that IL-17 participates in the recruitment of polymorphonuclear neutrophils via a paracrine mechanism involving IL-8 released by keratinocytes. On the other hand, keratinocyte-derived IL-15, a cytokine known to induce IL-17, could also amplify the production and activity of IL-17.39
In rats, human IL-17 protein recruits polymorphonuclear neutrophils in airways in vivo partly via the induced release of a rat correlate to IL-8, macrophage inflammatory protein (MIP)-2.40 In another series of cutaneous T-cell lymphomas, the 2 biopsies that expressed IL-8 were highly infiltrated with polymorphonuclear neutrophils.41 A cutaneous T-cell lymphoma with disseminated pustulosis was associated with production of high levels of IL-8 by tumor cells.42
Of course, we cannot exclude that other polymorphonuclear neutrophil attracting CXC chemokines or IL-8 produced by other cells, such as fibroblasts, endothelial cells or tumor cells, possibly regulated by IL-17, may also contribute to polymorphonuclear neutrophil infiltration. The significance of polymorphonuclear neutrophil infiltration in tumors is a matter of controversy, as in mice polymorphonuclear neutrophils effectively inhibit tumor growth.43 In humans, high polymorphonuclear neutrophil infiltration in the tumor was associated with a poorer outcome possibly due to the release of proteases, angiogenic factors or growth factors.44, 45
The prognostic value of IL-17 expression by MF/SS tumors was not addressed in this study. In mice, the effect of IL-17 on in vivo tumor growth seems to depend largely on the tumor immunogenicity and cell types, as we and other groups have shown that IL-17 inhibits tumor growth in a T-cell-dependent manner.24, 46 In contrast, IL-17 has also been shown to promote tumor cell growth via proangiogenic and IL-6-dependent mechanisms.12, 35, 47
In the present study, 5 out of 10 MF/SS biopsies expressed IL-17, but in this small series of patients, no clear-cut difference in the expression of this cytokine was observed in MF or SS patients. Nevertheless, it is noteworthy that IL-17 was not detected in the biopsies derived from 2 patients initially identified as MF, whereas an upregulation of this cytokine was clearly demonstrated during progression of the disease in these 2 patients. A prospective study including a larger series of patients will be needed to assess the relevance of IL-17 as a clinical marker in cutaneous T-cell lymphomas.
In conclusion, we report for the first time that human tumor cells can express IL-17, a cytokine that represents an early event in the development of the inflammatory reaction within the tumor microenvironment, a process that may influence tumor phenotype and growth.48
The authors thank A. Janin and C. Danel for their support and help during this work.