The expanding roles of 1-methyl-tryptophan (1-MT): in addition to inhibiting kynurenine production, 1-MT activates the synthesis of melatonin in skin cells

Authors

  • Ana C. R. Moreno,

    Corresponding author
    1. Department of Clinical Analysis and Toxicology, School of Pharmaceutical Sciences, University of São Paulo, Brazil
    • Correspondence

      A. C. R. Moreno, School of Pharmaceutical Sciences, University of São Paulo, São Paulo, SP, Brazil

      Fax: +55 11 3813 2197

      Tel: +55 11 3091 3741

      E-mail: carol@usp.br

    Search for more papers by this author
  • Renan O. Clara,

    1. Department of Clinical Analysis and Toxicology, School of Pharmaceutical Sciences, University of São Paulo, Brazil
    Search for more papers by this author
  • Janine B. Coimbra,

    1. Department of Clinical Analysis and Toxicology, School of Pharmaceutical Sciences, University of São Paulo, Brazil
    Search for more papers by this author
  • Ariane R. Júlio,

    1. Department of Clinical Analysis and Toxicology, School of Pharmaceutical Sciences, University of São Paulo, Brazil
    Search for more papers by this author
  • Renata C. Albuquerque,

    1. Department of Clinical Analysis and Toxicology, School of Pharmaceutical Sciences, University of São Paulo, Brazil
    Search for more papers by this author
  • Edson M. Oliveira,

    1. Department of Clinical Analysis and Toxicology, School of Pharmaceutical Sciences, University of São Paulo, Brazil
    Search for more papers by this author
  • Silvya S. Maria-Engler,

    1. Department of Clinical Analysis and Toxicology, School of Pharmaceutical Sciences, University of São Paulo, Brazil
    Search for more papers by this author
  • Ana Campa

    1. Department of Clinical Analysis and Toxicology, School of Pharmaceutical Sciences, University of São Paulo, Brazil
    Search for more papers by this author

Abstract

Indoleamine 2,3-dioxygenase 1 (IDO1), the rate-limiting enzyme of tryptophan catabolism, has been strongly associated with the progression of malignancy and poor survival in melanoma patients. As a result, IDO1 is a leading target for interventions aimed at restoring melanoma immune surveillance. Here, in a scenario involving the tryptophan catabolism, we report that melatonin biosynthesis is driven by 1-methyl-tryptophan (1-MT), a competitive inhibitor of IDO1, in human fibroblasts, melanocytes and melanoma cells. In addition to melatonin biosynthesis, 1-MT induced the expression of tryptophan hydroxylase, arylalkylamine-N-acetyltransferase and hydroxyindole O-methyltransferase mRNA in fibroblasts and melanocytes. We observed a great variability in the levels of IDO1 mRNA expression and kynurenine release between skin cells and melanoma cell lines in response to interferon-γ, a classical IDO1 inducer. In this setting, melatonin was shown to downregulate kynurenine production. Furthermore, in a condition of low basal activity of IDO1, it was observed that 1-MT, as well melatonin, inhibited the proliferation of human melanoma cells. Taken together, our results suggest that 1-MT may serve as more than just a tool to disrupt tumor immune escape (via the inhibition of IDO1) because it was shown to act directly on the proliferation of human melanoma cells and induce melatonin biosynthesis in the tumor milieu. Moreover, 1-MT-mediated inhibition of IDO occurs in normal skin and melanoma cells, which addresses the possibility that all cells in the skin microenvironment can be targeted by 1-MT. Our findings provide innovative approaches into understanding tumor therapy related to the control of tryptophan metabolism by 1-MT.

Abbreviations
1-MT

1-methyl-DL-tryptophan

AANAT

arylalkylamine-N-acetyltransferase

AHR

human aryl hydrocarbon receptor

HIOMT

hydroxyindole O-methyltransferase

IDO1

indoleamine 2,3-dioxygenase 1

IDO2

indoleamine 2,3-dioxygenase 2

IFN-γ

interferon γ

Kyn

kynurenine

TDO2

tryptophan 2,3-dioxygenase 2

TPH1

tryptophan hydroxylase

Trp

tryptophan

Introduction

Indoleamine 2,3-dioxygenase 1 (IDO1), an enzyme involved in tryptophan (Trp) catabolism that converts Trp to kynurenine (Kyn) (Fig. 1), is immunosuppressive and is strongly associated with the process by which malignant melanoma cells escape immune surveillance [1-3]. The main mechanism by which IDO drives immune escape in melanoma cells is mediated by the suppressive effects of Trp catabolism on effector T-cell function and the positive effects of this catabolism on regulatory T-cell differentiation [2-4].

Figure 1.

Trp metabolism. Trp is mainly metabolized by two pathways: the Kyn pathway and the serotoninergic/melatoninergic pathway. The rate-limiting step in the Kyn pathway is carried out by IDO1 to produce Kyn. The serotoninergic and melatoninergic pathways trigger the production of serotonin and melatonin, respectively. Their biosynthesis from Trp involves four well-defined intracellular steps catalyzed by TPH1, aromatic amino acid decarboxylase (AADC), AANAT and HIOMT.

IDO1 is overexpressed in various types of tumor cells and has been suggested to have a prognostic value in melanoma. Therefore, the upregulation of IDO1 in melanoma cells and tumor-draining lymph nodes has been shown to serve as an independent prognostic marker of less favorable prognosis and overall survival in patients with melanoma [2, 3, 5]. Moreover, an increased Kyn-to-Trp ratio in the serum of patients with malignant melanoma is closely related to tumor progression [6].

It is believed that IDO1 may be a leading target for interventions aimed at blocking melanoma immune escape [7], and this has also been demonstrated in experimental tumor mouse models [8]. 1-Methyl-tryptophan (1-MT), the gold standard for competitively inhibiting IDO1 [9], has been evaluated in clinical trials as a molecule directed at disrupting tumor tolerance [10]. In addition to blocking IDO1, the importance of 1-MT can be further highlighted due to its modulatory effects on dendritic cells, which could influence tumor growth [11].

With regard to tumor biochemistry, there is a lack of information concerning the impact of 1-MT on other routes of Trp metabolism. Since the serotoninergic and melatoninergic systems have been characterized in the main cellular population of humans [12-15] and rodent skin [16, 17], one possible role for endogenous melatonin could be its action modulating cell proliferation and viability [14]. Here, we have demonstrated that melatonin biosynthesis is driven by 1-MT in skin cells and melanoma cells. We also explored the effect of 1-MT on the expression of IDO1, tryptophan hydroxylase (TPH1), arylalkylamine-N-acetyltransferase (AANAT) and hydroxyindole O-methyltransferase (HIOMT) mRNA in human fibroblasts, keratinocytes, melanocytes and melanoma cell lines. Because we observed that 1-MT induced melatonin biosynthesis, we additionally studied the effect of 1-MT and melatonin on the clonogenicity of human melanoma cells.

Results

1-MT modulates the expression of genes associated with Trp metabolism in skin cells and melanoma cell lines

We evaluated the mRNA expression of Trp-metabolizing enzymes in cells that were treated with interferon- γ (IFN-γ) and/or 1-MT. IFN-γ, a classic IDO1 inducer [18], enabled the observation of 1-MT-mediated effects in cells with low basal IDO1 expression. We observed heterogeneity in gene expression between different cell types. IFN-γ produced a strong induction of IDO1 mRNA in all skin cell types and melanoma cell lines. The contribution of other Kyn-producing enzymes appears to be minimal, because cells did not express or expressed comparatively much lower indoleamine 2,3-dioxygenase 2 (IDO2) and tryptophan 2,3-dioxygenase 2 (TDO2) mRNA than IDO1 mRNA (Fig. 2A,B,C). The results outlined above indicated that IDO1 is mainly responsible for Kyn production in skin cells and melanoma cells upon IFN-γ treatment. Interestingly, among the skin cell types, fibroblasts demonstrated the greatest expression of IDO1 mRNA after IFN-γ stimulation, as this expression was ~ 2.5 and 1.5 times higher than that observed in keratinocytes and melanocytes, respectively (Fig. 2A). The induction of IDO1 mRNA in fibroblasts was similar to that observed in SK-Mel-147 cells under the same stimulation conditions. Among the melanoma cell lines, SK-Mel-147 cells were more responsive to IFN-γ-mediated upregulation of IDO1 mRNA than SK-Mel-19 cells. With regard to the transcripts associated with serotoninergic and melatoninergic pathways, there was a slight increase in the expression of TPH1, AANAT and HIOMT mRNA in keratinocytes treated with IFN-γ, and increases in TPH1 mRNA expression in treated SK-Mel-147 cells and AANAT and HIOMT mRNA expression in treated fibroblasts were also noted (Fig. 2D,E,F).

Figure 2.

Treatment of cells with 1-MT results in the expression of genes involved in Trp catabolism. The comparison of (A) IDO1, (B) IDO2, (C) TDO2 and (D) TPH1, (E) AANAT and (F) HIOMT mRNA expression in skin cells and melanoma cell lines under different treatment conditions. In response to 1-MT treatment, upregulation of IDO1 mRNA expression was observed in all of the cell types with the exception of keratinocytes. In addition, 1-MT upregulated the expression of TPH1 and HIOMT in the fibroblasts and melanocytes. The error bars correspond to the SEM from four independent experiments. *< 0.05; **< 0.01; ***< 0.001.

1-MT is a competitive inhibitor of IDO and is known to increase the expression of IDO in human cancer cells [19]. Upregulation of IDO1 mRNA induced by 1-MT treatment was also observed in fibroblast, melanocyte and melanoma cells but not in keratinocytes (Fig. 2A). These results indicate that there is no general mechanism responsible for the effect of 1-MT on the regulation of IDO1 mRNA expression. With the exception of keratinocytes, we also observed a slight increase in TDO2 mRNA after treatment with 1-MT (Fig. 2C).

In relation to the serotoninergic and melatoninergic systems, melanocytes were very active and demonstrated the highest expression of TPH1, AANAT and HIOMT mRNA in comparison with the other cell types. Surprisingly, 1-MT induced the expression of TPH1 and HIOMT mRNA in fibroblasts and melanocytes (Fig. 2D,E,F), indicating that due to the action of 1-MT these cells acquired the machinery necessary for the biosynthesis of melatonin.

1-MT reduces Kyn production and increases melatonin biosynthesis in fibroblasts, melanocytes and melanoma cells

In response to IFN-γ-mediated induction, an increase in IDO1 mRNA expression was noted, which correlated with an enhanced IDO1 enzymatic activity in all of the cell lines. Fibroblasts and SK-Mel-147 cells produced higher levels of Kyn than keratinocytes, melanocytes and SK-Mel-19, which demonstrated that these cells responded differently to IFN-γ stimulation. 1-MT treatment clearly inhibited Kyn production in all cell types, provided that they were stimulated with IFN-γ (Fig. 3A).

Figure 3.

1-MT activates melatonin biosynthesis. (A) Quantification of Kyn production in the supernatants of fibroblast, keratinocyte, melanocyte and SK-Mel-19 and SK-Mel-147 cell cultures in response to 1-MT treatment, as assessed by HLPC analysis. (B) Quantification of melatonin levels in the supernatants of fibroblast, keratinocyte, melanocyte and SK-Mel-19 and SK-Mel-147 cell cultures in response to 1-MT treatment, as assessed by MS analysis. The error bars correspond to the SEM from four independent experiments. *< 0.05; ***< 0.001.

Of the different types of skin cells studied, detectable levels of melatonin were observed only in melanocytes. Surprisingly, the basal concentration of melatonin was much higher in the two melanoma lines tested compared with that observed in melanocytes, which is consistent with previous studies [12]. Remarkably, in addition to inhibiting Kyn production, 1-MT treatment significantly stimulates melatonin biosynthesis in fibroblasts, melanocytes and the melanoma cell lines (Fig. 3B). Although Kyn and melatonin were present in micromolar and nanomolar concentrations, respectively, it was clear that 1-MT treatment resulted in melatonin biosynthesis, especially in the cultures of melanocytes and fibroblasts (Fig. 3B). In the latter cell type, we also observed a switch from no detectable to detectable production of melatonin. Although the melanoma lines also responded to 1-MT treatment, the induction of melatonin in these cells was comparatively modest. Keratinocytes did not synthesize melatonin (or it was not detectable) and were unaffected by 1-MT treatment in our experiments (Fig. 3B).

Melatonin modulates Kyn production in fibroblasts, keratinocytes, melanocytes and the SK-Mel-19 cell line

Furthermore, we investigated whether melatonin could influence IDO1 mRNA expression and Kyn release in skin cells and melanoma cell lines. There was heterogeneity with regard to IDO1 mRNA expression in response to melatonin or IFN-γ/melatonin treatment, indicating that melatonin treatment had different effects on the different cell types evaluated (Fig. 4A). However, melatonin treatment produced a slight decrease in the release of IFN-γ-induced Kyn in skin cells and SK-Mel-19 cells but not in SK-Mel-147 cells (Fig. 4B), and this effect did not correlate with the effect of treatment on IDO1 transcript levels. The inhibitory effect of melatonin on Kyn release was much lower than that observed in cells treated with 1-MT. These results indicate that the inhibition of Kyn synthesis driven by melatonin was modulated at the post-transcriptional level.

Figure 4.

Melatonin modulates Kyn production in skin cells and melanoma cell lines. (A) IDO1 mRNA levels in fibroblasts, keratinocytes, melanocytes and SK-Mel-19 and SK-Mel-147 cells treated with melatonin. (B) Quantification of Kyn levels in cells by HPLC analysis after treatment with melatonin. The error bars correspond to the SEM from four independent experiments. *< 0.05; ***< 0.001.

1-MT and melatonin treatment reduces the clonogenicity of melanoma cell lines

It has been demonstrated that 1-MT can inhibit IDO activity and act in concert with classical treatments to promote the regression of tumors in pre-clinical models [8]. However, in this context, the antitumor effects of 1-MT are thought to be immune mediated [8, 11, 20-22]. Here, we investigated the direct effect of 1-MT on the migratory and proliferative abilities of melanoma cell lines. Therefore, the directional migration of melanoma cell lines was evaluated by performing scratch wound healing assays, and the clonogenic assay was used to examine the effect of 1-MT on triggering the loss of reproductive integrity in melanoma cells. As shown in Fig. 5A,B, untreated wounds were used as controls for studying the progression of wound healing among melanoma cells. Cells treated with 1-MT demonstrated significant delays in wound closure, as 14% and 16% delays were observed for SK-Mel-19 and SKMel-147 cells, respectively. In the analysis of long-term colony formation, the proliferation rate of untreated SK-Mel-147 cells was 10-fold higher than that observed for SK-Mel-19 cells (Fig. 5C,D), and this phenotype may have been due to the different genetic background of the two cell lines [23]. Notably, after 15 days of treatment with 1-MT, there was a 17% drop in the reproductive ability of SK-Mel-19 to form large colonies (0.600 ± 0.016 versus 0.505 ± 0.027, colony area in mm2, = 0.02). The strongest effect was observed in the SK-Mel-147 cell line, where cells had lost 42% of their reproductive integrity (6.36 ± 0.36 versus 3.71 ± 0.28, colony area in mm2, < 0.0001) (Fig. 5C,D). It is worth noting that the antiproliferative effect of 1-MT in the long-term colony formation could be observed for nine different melanoma cell lines (data not shown). These data suggest that treatment with 1-MT alone could affect melanoma migration and proliferation.

Figure 5.

The effects of 1-MT on melanoma cell migration and proliferation. (A) The optical microscopy imaging of cellular migration in SK-Mel-19 and SK-Mel-147 melanoma cells treated for 24 h and 36 h, respectively, with 1 mm 1-MT. Cells cultured in medium alone were used as reference controls. (B) Panels on the right indicate the quantification of the percentage of cells that had migrated after treatment. The error bars correspond to the SEM from six independent experiments. (C) The optical microscopy imaging of cellular colony formation in SK-Mel-19 and SK-Mel-147 melanoma cells treated for 14 days with 1 mm 1-MT. Control cells were cultured in medium alone. (D) Panels on the right indicate the average quantification of the colony area after 1-MT treatment. The error bars correspond to the SEM from two independent experiments. *< 0.05; **< 0.01.

We next assessed the effects of melatonin on the proliferation of melanoma cells by performing the clonogenic assay. Melatonin treatment considerably reduced the size of the colonies that formed, demonstrating the existence of an inhibitory effect that was even greater than that induced by 1-MT. Remarkably, melatonin treatment decreased the reproductive integrity of SK-Mel-147 cells to form large colonies by 80% (6.36 ± 0.36 versus 1.25 ± 0.11, colony area in mm2, < 0.0001) (Fig. 6A,B). In contrast, the effect on SK-Mel-19 cells was modest (0.600 ± 0.016 versus 0.510 ± 0.020, colony area in mm2, = 0.01), with an observed reduction in reproductive ability of 15%.

Figure 6.

The effects of melatonin on melanoma cell proliferation. (A) Optical microscopy imaging of cell colony formation in SK-Mel-19 and SK-Mel-147 melanoma cells treated for 14 days with 1 mm melatonin. The control cells were cultured in medium alone. (B) The panels on the right depict the average quantification of colony area after melatonin treatment. The error bars correspond to the SEM from two independent experiments. *< 0.05; ***< 0.001.

Discussion

Despite the use of multiple new treatments in the form of either single agents or combinations, melanomas remain resistant to all therapies [24, 25]. Due to the demand for newer strategies, the use of 1-MT has been considered as a therapeutic approach to improve T-cell-mediated tumor control [8, 10]. This approach aims to disrupt tumor immune escape by inhibiting IDO, but 1-MT may also have off-target effects on various biological features that contribute to tumor progression, which could be additionally exploited. Alternatively, we have provided the first evidence of melatonin biosynthesis driven by 1-MT in human skin cells and melanoma cell lines. We found that treatment with 1-MT or melatonin produced effects on tumor cell migration and proliferation, and our findings suggest additional effects of 1-MT as an adjuvant in the treatment of melanoma.

We observed great variability in the levels of IDO1 mRNA expression and Kyn release between normal skin cells and melanoma cell lines. In particular, varying amounts of Kyn were released in fibroblasts, keratinocytes, melanocytes and melanoma cell lines (SK-Mel-19 and SK-Mel-147) in response to IFN-γ. Although TDO2 seems to be important for the production of Kyn by some types of tumors [26, 27], IDO1 was the enzyme mainly responsible for Kyn production in skin cells and melanoma cells upon IFN-γ treatment. Importantly, the amount of Kyn produced by fibroblasts was high and comparable with that produced by SK-Mel-147 metastatic melanoma cells (Fig. 3A). Sheipouri and collaborators recently emphasized that the role of the Kyn pathway in normal skin cells is not well understood [28]. Given the important role played by IDO in the engraftment of allogenic skin substitute in wound healing [29], the production of Kyn by fibroblasts observed in this study may contribute to the healing event. Our findings suggest that keratinocytes and melanocytes also have the capability to produce Kyn, which indicates that these cells could also contribute to the process of healing via Trp metabolism. Because IFN-γ is a good inducer of IDO in normal skin cells, this process may have a role in the maintenance or resolution of inflammatory processes in the skin.

Recently, Kyn was identified as an endogenous ligand for the human aryl hydrocarbon receptor (AHR) [30]. Immune and tumor cells respond differently to Kyn; the activation of AHR by Kyn modulates the function of dendritic cells, which leads to the generation of regulatory T cells [30-32] that promote tumor cell survival and motility [26]. These findings have elegantly linked the fields of immunology and cancer biology by providing a potential mechanism by which Kyn production helps tumor cells to overcome the immune response and progress to cause cancer [26, 30-32]. Although we did not evaluate AHR, differential expression of this receptor by SK-Mel-147 and SK-Mel-19 cells may explain the different clonogenic susceptibility of these cells to 1-MT and melatonin (Figs 5 and 6). This hypothesis is currently being evaluated in studies that are ongoing in our laboratory.

Our finding that 1-MT triggers melatonin biosynthesis may have important biological consequences. It has been proposed that melatonin can regulate skin functions [12, 15], and it is also known that melatonin can have an antiproliferative effect on tumor cells, including melanoma cells [33-41]. Melanomas represent a heterogeneous group of tumors [23] demonstrating diverse behaviors in response to various treatment strategies [24, 25]. This diversity was observed in the two melanoma cell lines used in this study, as these presented different susceptibilities to 1-MT and melatonin in the clonogenic assay. Indeed, melatonin may act on cells via a receptor-dependent or receptor-independent mechanism. Previous studies have shown that melatonin differentially suppressed proliferation in melanoma cell lines, and this behavior could be related to specific patterns of cellular receptors and/or cytosolic binding protein expression [33, 40-42]. Regardless, it is tempting to speculate that the antitumor effects of 1-MT may, at least in part, be related to increased melatonin biosynthesis in skin and melanoma cells. Locally produced melatonin may therefore contribute to the antitumor effects of 1-MT. Furthermore, one additional aspect worth noting is the fact that melatonin was shown to downregulate Kyn production (Fig. 4B).

With regard to the inhibition of Kyn production mediated by melatonin, it is possible that this effect was mediated by an inhibitory effect on IDO activity or the steps required for its activation. However, the first possibility was ruled out by previous studies of our research group (unpublished data) and others demonstrating that melatonin is not an IDO inhibitor [43]. The second possibility appears to be feasible given that melatonin is a powerful antioxidant and scavenges reactive oxygen species [44-46]. Moreover, IDO activation dependent on intracellular oxidants has not yet been identified [47-49].

Persistent chemoresistance and immunoresistance as well as secondary toxicities compromise the response to cancer treatment [50]. Future studies should investigate whether concurrent adjuvant therapy in combination with modern chemotherapy could have an impact on patient survival outcomes. In fact, the use of a combination of IDO inhibitors with other chemotherapeutic agents has been proposed [51]. Considering the findings that 1-MT induced melatonin biosynthesis, that 1-MT and melatonin treatments had antiproliferative effects, and that melatonin protected against the effects of chemotherapy [38], the response of patients to the combination of conventional drugs plus 1-MT or melatonin should be considered for the treatment of different types of tumors.

In conclusion, our findings demonstrated the following: (a) 1-MT-mediated inhibition of IDO occurs in normal skin and melanoma cells, which addresses the possibility that all cells in the skin microenvironment can be targeted by 1-MT; (b) in addition to the known effect of 1-MT on IDO inhibition, this molecule may directly act on tumor progression, as assessed in the clonogenic assay; (c) the mechanism of action of 1-MT may involve the induction of melatonin synthesis; and (d) melatonin can affect Kyn production. Because 1-MT has been proposed as an adjuvant to conventional antitumor therapy and melatonin is a well established oncostatic molecule, our results provide novel insights to understand tumor therapy regarding the control of Trp metabolism by 1-MT (Fig. 7). Other roles for 1-MT and melatonin may also broaden the combinations of treatments available against melanoma and other tumors.

Figure 7.

A schema describing how 1-MT and melatonin may act on skin cells and melanoma cells. 1-MT inhibits IDO in normal skin cells and melanoma cells, suggesting that (A) cells in the skin microenvironment could be targets for 1-MT; (B) some of the effects of 1-MT could be due to the induction of melatonin synthesis; and (C) melatonin affects Kyn production. Mlt, melatonin; dotted line, inhibition; continuous line, activation.

Materials and methods

Cell culture conditions and treatments

The human melanoma metastatic cell lines (SK-Mel-19 and SK-Mel-147) were donated by M. Soengas (Centro Nacional de Investigaciones Oncológicas CNIO/Spanish National Cancer Research Center, Madrid, Spain) [23, 52], and primary skin cells (keratinocytes, melanocytes and fibroblasts) were obtained from the foreskins of patients at the University Hospital (Hospital Universitário/HU-USP) and were donated by L. Maximiano (Ethics Committee ‘Comitê de Ética em Pesquisa do Hospital Universitário da Universidade de São Paulo – CEP-HU/USP’ no. CEP, Case 943/09). We confirm that written informed consent from the donor or the next of kin was obtained for use of the samples in our study. The melanoma cell lines and fibroblasts were cultured in RPMI 1640 medium (Gibco®, Life Technologies, Grand Island, NY, USA) and DMEM (Gibco®), respectively, supplemented with 10% fetal bovine serum (Gibco®), penicillin (50 U·mL−1) and streptomycin (50 μg·mL−1). Keratinocytes were maintained in Epilife medium (SKU # M-EPICF-500, Cascade Biologics, Portland, OR, USA) supplemented with human keratinocyte growth supplement (SKU # S-001, Cascade Biologics). The melanocytes were maintained in 254CF medium (SKU # M-500-254CF, Cascade Biologics) supplemented with human melanocyte growth supplement (SKU # S-002-5, Cascade Biologics). All of the cells were maintained at 37 °C in an atmosphere containing 5% CO2. To prepare the samples, cells were seeded in a six-well plate and cultured in the appropriate medium for 24 h until they were 60% confluent. Next, 1-MT (Sigma, St Louis, MO, USA) or melatonin (Sigma) was added to the cell culture for 24 h (1 mm final concentration) in the absence or presence of IFN-γ (1000 units·mL−1) (Peprotech, Ciudad de Mexico, DF, Mexico). The cellular supernatant was used for measuring the levels of Kyn and melatonin, and the adherent cells were harvested and used for RNA extraction. 1-MT and melatonin did not induce cell death under these conditions (data not shown).

Real-time PCR

The total RNA was extracted using the RNeasy Mini kit (Qiagen, Austin, TX, USA), according to the manufacturer's instructions. RNase-free DNase (Qiagen) was used to remove any potential contamination by genomic DNA. The samples were stored at −80 °C until needed. Up to 1 μg of total RNA was reverse transcribed using the High Capacity RNA-to-cDNA kit (Applied Biosystems, USA), according to the manufacturer's instructions. We performed real-time PCR using TaqMan fluorescence energy transfer assays with an ABI Prism 7500 Sequence Detection System (Applied Biosystems). The primers and fluorogenic probe used in this assay were obtained from Applied Biosystems, and they were ordered as Hs00266705_g1 (GAPDH), Hs00986554_m1 (IDO1), Hs01589373_m1 (IDO2), Hs00194611_m1 (TDO2), Hs01559141_m1 (TPH1), Hs01063209_g1 (AANAT) and Hs00946627_m1 (HIOMT). For each of the real-time reactions, GAPDH was used as an endogenous housekeeping gene, and its expression did not vary according to IFN-γ, 1-MT or melatonin treatment. The relative comparison method (2−ΔΔCT) was used to compare the expression levels of mRNA.

Kyn quantification

The supernatants from cultured cells were collected for the determination of Kyn levels. Briefly, 150 μL of cellular supernatant was added to 1500 μL of ice-cold acetonitrile, mixed for 1 min, and centrifuged at 14 000 g at 4 °C for 10 min. The organic layer was transferred to a glass tube, protected from light and evaporated at room temperature under a flow of nitrogen. The residue was dissolved in 150 μL MilliQ water and filtered using a spin-X® centrifuge tube filter (Corning® Costar®, Sigma). Next, 40 μL was injected into an HPLC Shimadzu SCL-10A vp system (Shimadzu Corporation, Kyoto, Japan). Kynurenine was separated using a Luna C18 Phenomenex column (250 mm × 4.6 mm inner diameter column; 5 μm) with 100% acetonitrile (A) and milli-Q® H2O (B) as the mobile phases. The linear gradient began at A/B = 0/100 and ended at 22 min with A/B = 10/90. Next, the mobile phase was held at A/B = 0/100 for 5 min. During the entire chromatography procedure, the flow rate was set at 1 mL·min−1. The detection was performed using a Diode Array SPD-M10A Shimadzu detector by selecting a wavelength of 365 nm. The sample concentrations were calculated using standard curves. The amount of Kyn found in untreated cells corresponded to the amount of Kyn found in the supplemented medium (data not shown).

Melatonin extraction and quantification

For melatonin extraction, the culture medium was separated and placed in a 15 mL Falcon tube containing 20 μL of 0.1 m NaOH and vortexed. Next, 100 μL of indole solution (100 μg·mL−1) was added as an internal standard, and the mixture was vortexed again. Then 2.5 mL of cold dichloromethane (stored at −20 °C) was added to the tube, which was vortexed continuously for 5 min and stored in an ice bath for another 5 min. Next, the tubes were centrifuged at 9.5 g for 10 min. The organic phase was transferred to a glass tube and dried under nitrogen gas flow. The samples were reconstituted by adding 150 μL of an acetonitrile-containing solution and 4 mm sodium formate (50 : 50 v/v), and the tubes were vortexed for 1 min. Next, 100 μL of this solution was transferred to a Spin-X centrifuge tube with a 22 μm filter (Costar, Corning) and centrifuged for 3 min at 1 g. Subsequently, 20 μL was injected into the equipment. Melatonin quantification was performed by liquid chromatography mass spectrometry using an ESI ion source, which was operated in a positive mode. The instrumentation consisted of LC 10Avp pumps, a SIL 10ADvp auto sampler, a SCL 10ADvp controller (all purchased from Shimadzu Co.) and a Quattro–Micro Triple Quadrupole (Micromass, Manchester, UK). The following source conditions were used for the equipment: 3.27 kV capillary, 15 V cone, 1 V extractor, 0.1 V rf lens, 100 °C source temperature, 300 °C desolvation temperature, 204 L·h−1 cone gas flow, and 561 L·h−1. The desolvation gas flow and the analyzer equipment conditions were as follows: for 10 LM1 and HM1, a resolution of 0.2 V, ion energy 1, 37 entrance2, 12 collision energy, and 37 exit; and for 10 LM2 and HM2, a resolution of 2.0 V, ion energy 2, 850 V multiplier and 2.52 × 10−3 mbar gas cell pressure. Each of these conditions was optimized by infusing a standard melatonin solution (1 μg·mL−1). The analysis was performed using a MassLynx V4.0 data analysis system (Micromass, Cary, NC, USA), and the data were collected in SIR or MRM mode by selecting a transition of 232.8(M + H) > m/z of 174 for melatonin and 118(M + H) > m/z of 89 for the internal standard, with −12 V as the collision energy. Chromatography was performed under a gradient of 4 mm ammonium formate (pH 4, in MilliQ water, ®Millipore) (referred to as buffer A) and 0.01% formic acid in acetonitrile (referred to as buffer B) at a flow rate of 0.35 mL·min−1. A Kinetex C18 100A (50 × 4.6 mm, 2.6 μm; Phenomenex) column coupled to a KrudKatcher ultra column in-line filter (Phenomenex) was used. The elution gradients for column re-equilibration were as follows: 0–1 min, 50 A : 50 B, v/v; 1–4.5 min, 0 A : 100 B, v/v; 4.5–6.5 min, 0 A : 100 B, v/v; 6.5–7 min, 50 A : 50 B, v/v; and 7–9 min, 50 A : 50 B, v/v.

Wound healing assay

To evaluate melanoma migration in response to the 1-MT treatment, cells were seeded in 24-well plates and cultured for 24 h until they reached 95% confluence. Next, the monolayers were carefully scratched with a 200 μL pipette tip followed by the addition of fresh culture medium containing 1-MT (1 mm). The cells were photographed after appropriate incubation times using a light microscope.

Clonogenic assay

We performed the clonogenic assay to determine the effect of 1-MT or melatonin treatment on the proliferation of tumor cells. Six hundred cells were seeded in 60 mm plates and cultured for 24 h. Next, the cells were treated with 1-MT or melatonin at a final concentration of 1 mm. Treated medium was replaced every 48 h. After 15 days, the cells were fixed in glutaraldehyde (6.0% v/v) and stained with crystal violet (0.5% w/v) to observe the formation of colonies. The size of the colonies was calculated using the carestream molecular imaging software from In-Vivo MS FX-PRO (Carestream Health Inc., Woodbridge, CT, USA). Colonies were defined as groups of neighboring cells that probably originated from a single parental cell.

Statistics

The statistical significance of differences in the mean values of all experimental groups was calculated using a one-way ANOVA. P values < 0.05 were considered to be statistically significant.

Acknowledgements

This work was supported by grants from the São Paulo Research Foundation – FAPESP (FAPESP 2009/14632-3, 2009/53800-9 and 2010/15919-1) and the National Council for Scientific and Technological Development – CNPq (CNPq 47151012010-6). Similarly, we acknowledge the scholarship from FAPESP to A. C. R. Moreno and R. Clara and from CNPq to J. B. Coimbra.

Ancillary