Introduction
Development of adaptive immune response is regulated by immunosuppressive mechanisms that are involved in the maintenance of tolerance to self-antigens as well as in the control of tissue damage and homeostasis. A balance between activation and inhibition of immune response is regulated at many levels by life-essential mechanisms and various cell types. Among others, indoleamine 2,3-dioxygenase 1 (IDO) was shown to be involved in the formation of a tolerogenic environment [
1]. Moreover, in some types of cancer, IDO is considered to be engaged in the development of immunosuppressive microenvironment within the tumor and in the tumor-draining lymph nodes (TDLNs) [
2]. The effects of IDO activity such as local depletion of tryptophan and production of kynurenines, cause growth arrest of effector T cells, loss of cytotoxic function and polarization into T regulatory lymphocytes (Treg). Additionally, it was reported that IDO-secreting cells can mediate apoptosis of T cell clones [
3,
4]. IDO similarly to other amino acid degrading enzymes like arginase 1 (Arg1) can be induced during inflammation or anticancer therapy [
5]. Secretion of interferon γ (IFN-γ) and tumor necrosis factor α (TNF-α) was shown to increase IDO expression in various types of myeloid cells, including monocytes/macrophages, neutrophils, dendritic cells as well as tumor cells. In many types of tumors, elevated expression of IDO correlates with poor prognosis of patients [
6]. Therefore, IDO became a target for antitumor therapies and IDO inhibitors such as epacadostat, navoximod and indoximod are tested in clinical trials as mono- and combined therapies with other immunomodulatory drugs [
7].
Development of strong inflammation is well described as a first and decisive event after photodynamic therapy (PDT) of cancer. PDT is a clinically approved, noninvasive cancer treatment involving generation of cytotoxic reactive oxygen species (ROS) that result from photosensitizer activation by light of appropriate wavelength. PDT leads to direct tumor cell death, disruption of vasculature followed by induction of acute inflammation [
8,
9]. These events are associated with the release of various inflammatory mediators, recruitment and activation of innate immune cells and subsequent activation of a specific antitumor immune response. A great body of evidence indicates that the antitumor effects of PDT depend on the presence and activity of adaptive immunity [
10].
Various immunosuppressive processes are also activated in response to PDT, including an increase in the number of Treg and production of anti-inflammatory cytokines, such as IL-10 or transforming growth factor β (TGF-β) [
11]. Moreover, IL-10 and TGF-β mediate differentiation of CD4
+ T cells into Treg and cause anergy of CD8
+ T cells [
12]. Importantly, inactivation of immunosuppressive mechanisms leads to the development of efficient PDT-mediated antitumor adaptive immune response [
13].
An important role of immunomodulatory enzymes such as Arg1 or inducible nitric oxide synthases (iNOS) as well as myeloid cells in the shaping of PDT-treated tumor environment has been recently highlighted [
14,
15]. In this study, we analyzed the expression of enzymes: IDO, Arg1 and iNOS to elucidate the immunosuppressive mechanism induced by PDT. We confirmed that PDT-mediated inflammation is associated with Treg induction, and we found that PDT triggers expansion of myeloid cells with elevated expression of IDO. Finally, we showed that the combination of PDT with IDO inhibitor (epacadostat) augments the IL-6-dependent acute inflammation. The antitumor efficacy of the treatment combining PDT and IDO inhibitor is effective but accompanied by systemic toxicity.
Materials and methods
Cell culture and reagents
Mammary carcinoma 4T1 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) and E0771 in Roswell Park Memorial Institute (RPMI 1640) medium supplemented with heat-inactivated 10% fetal bovine serum (Invitrogen) and antibiotic/antimycotic solution (Sigma-Aldrich, A5955) under standard conditions (5% CO
2, humidified incubator at 37 °C). Epacadostat and its analogue—INCB024360-analog (Medkoo Bioscience Inc), were prepared for administration as it was described by Koblish et al. [
16]. Visudyne
® (Novartis), a liposomal formulation of verteporfin, was reconstituted as it was described before [
17].
Mice, tumor treatment and monitoring
Tumor cells were inoculated (5.0 × 10
4 of 4T1 or 1.5 × 10
5 of E0771 cells) into the second, left mammary fat pad of 8–12-week-old BALB/c or C57BL/6 female mice. The PDT was conducted on day 8th or 10th. Verteporfin was administered i.v., and PDT was performed as described previously [
17]. Epacadostat was administered: orally twice a day from day 9th to 13th or on day 9th and 10th and subsequently from day 13th to 20th at a dose of 50 mg/kg or intratumorally once a day from day 9th to 16th at a dose of 100 mg/kg. The anti-IL-6 mAbs or appropriate isotype control was used at a dose of 100 µg per mouse and injected every second day, in 6 doses, starting from 1 day post-PDT. Tumors were measured as described before [
17].
Real-time PCR
Total RNA was isolated, and real-time PCR was done as previously described in detail [
18]. The results were analyzed after amplification with LightCycler 480 Software 1.5 (Mannheim, Germany) and normalized for the content of the RPL32 as a housekeeping gene. The sequences of primers were as follows: IDO, forward-5′ GGTACATCACCATGGCGTATGTG-3′ and reverse-5′ TAAGACAGAATAGGAGGCA GGCC-3′; Arg1, forward-5′-GCAGTTGGAAGCATCTCTGG-3′ and reverse-5′-TCTACGTCTCGCAAGCCAAT; iNOS, forward-5′-GTCCTACACCA CACCAAACT-3′ and reverse-5′-CTCCAATCTCTGCCTATCCGT-3′; RPL32, forward-5′-TTAAGCGAAACTGGCGGAAAC and reverse-5′-TTGTTGCTCCCATAACCGATG-3′.
IDO activity assay
Tumors and TDLNs were harvested and lysed in 0.5% NP-40. The enzymatic assay has been performed according to the method of Takikawa et al. [
1] with some modifications. The reaction mixture (400 µl) contains: 1 M potassium phosphate buffer pH 6.5 (1 mM final concentration), 0.2 M sodium ascorbate (40 mM final concentration), 0.5 M methylene blue (20 µM final concentration), catalase (200 U/ml final concentration),
l-tryptophan (400 µM final concentration) and 500 µg of protein. All needed chemicals were purchased from Sigma-Aldrich. After 60 min of incubation at 37 °C, the reaction was terminated by the addition of 80 µl of 30% TCA. To convert
N-formylkynurenine to kynurenine, the reaction was carried out at 60 °C for 15 min, followed by centrifugation (10,000 ×
g, 20 min). Kynurenine was quantified by the addition of 200 µl of 2% Ehrlich’s reagent in a glacial acetic acid to an equal volume of sample supernatant. The activity of IDO was defined as the concentration of kynurenine that was generated during 60 min of enzymatic reaction. The kynurenine concentration was revealed as absorbance, measured at 480 nm (ASYS UVM 340, Biochrom) and shown on the graph as a percentage of controls.
Cell isolation
In order to obtain a single cell suspension, the tumors and TDLNs were incubated with collagenase IV (Sigma-Aldrich, C5138) and DNAse (Sigma-Aldrich, DN25) in IMDM medium (Invitrogen, 12440053) and subsequently forced through a 100-μm strainer. CD11b+ cells were isolated from tumor cell suspension by positive selection on magnetic beads according to the manufacturer’s instruction (EasySep #18000, Stemcell Technology). Spleens were mechanically disrupted through a 70-μm cell strainer.
White blood cells analysis
The smears of the blood were stained with the May–Grunwald–Giemsa method and properly air dried. Next, detailed cell morphology was assessed under the light microscopy with 100 × oil immersion objective. In the prepared smear, the percentages of the various population of white blood cells were counted.
Co-culture proliferation assay
Splenocytes were subjected to negative selection using magnetic beads (EasySep™ Mouse T Cell Enrichment Kit). Subsequently, CD3+ splenocytes were stained with CellTracker™ Violet BMQC Dye (Thermo Fisher) for 20 min in 37 °C and washed two times. Next, CD11b+ cells and T cells were seeded onto previously coated with anti-CD3 antibody (145-2C11, eBioscience) 96-well plate in 2:1 ratio. Then, cells were stimulated with anti-CD28 (145-2C11, eBioscience) for three consecutive days. Subsequently, T cells were stained and proliferation of CD8+ (53-6.7, eBioscience) and CD4+ (RM4-5, eBioscience) cells was evaluated with FACSCanto II using Diva software (Becton Dickinson, Franklin Lakes, New Jersey, USA).
Staining and flow cytometry
Cells were stained with Zombie NIR™ Fixable Viability kit (BioLegend, 423106) according to the manufacturer’s protocol and blocked with anti-CD16 mAbs. For surface markers, subsequent antibodies were used: anti-CD45.2-V500 (104, BD Bioscience), anti-CD11b-FITC (M1/70, eBioscience), anti-Ly6C-PerCp-Cy7 (AL27, BD Bioscience), anti-Ly6G-APC (1A8, BioLegend), anti-CD3-V450 (17A2, eBioscience), anti-IL4R-PE (552509, BD Bioscience). For intracellular staining, after the cells were fixed and permeabilized with Cytofix/Cytoperm (554722, BD Bioscience), the following antibodies were applied: anti-IDO-eF660 (Mido-48, eBioscience) and anti-Arg1-PE (IC5868P, R&D). In order to analyze the Treg population, the Mouse Phenotyping Kit (560767, BD Bioscience) was used. Cells, resuspended in FACS flow buffer, were analyzed on FACSCanto II using Diva software. The cytokine concentration was measured in mouse serum, separated from the blood collected from the cheek vein. Serum was stained with BD™ Mouse Inflammation Kit, Cytometric Bead Array (552364, BD Bioscience).
Statistical analyses
Data were analyzed with GraphPad Prism 6, and differences were calculated for significance by Mann–Whitney U test. The survival rate of animals was analyzed by log-rank survival test.
Discussion
PDT can be an effective and minimally invasive strategy to treat different types of superficial early-stage tumors without radiation and large incisions. In advanced metastatic tumors, it is rather a palliative treatment that can be used supplementary to surgery or optionally as an organ-sparing treatment [
25]. PDT leads to tumor cell death accompanied by extensive oxidative stress and induction of local inflammation [
26]. Both preclinical and clinical studies demonstrated that PDT through the induction of innate immune response is capable of activating adaptive immune response against tumors [
27,
28]. This feature makes PDT increasingly more attractive treatment option, which can be potentially used in combination with cancer immunotherapies. However, in experimental tumor models, the complete antitumor responses to PDT are limited to some particular conditions when the development of immune response is facilitated by the use of highly immunogenic, carcinogen-induced tumors or tumors that are derived from genetically modified cells that express strong tumor-associated antigens [
27‐
29]. Hence, it is widely discussed that PDT, apart from robust inflammation, induces certain compensatory mechanisms that limit the development of tumor-specific adaptive immunity [
12,
30]. Not many studies have focused on immune evasion-associated events during or after PDT. For example, PDT was shown to induce expansion of Treg and increased secretion of IL-10 and TGF-β [
11].
Here, we identify IDO as an immunoregulatory enzyme induced by PDT within tumors as well as in TDLNs. The obtained results suggest that the major source of IDO in the tumor microenvironment are granulocytic CD11b
+Ly6G
+ myeloid cells that strongly infiltrate the tumor after treatment and reveal the highest expression levels of this enzyme. Moreover, monocytic CD11b
+Ly6C
+ myeloid cells, although less abundant after PDT, upregulate IDO and might also play an important role in immunoregulation. The conclusion can be inferred from ex vivo studies, showing that CD11b
+ cells isolated from PDT-treated tumors suppress proliferation of T cells more effectively as compared with CD11b
+ cells isolated from control tumors. Although the Ly6G
+ myeloid cells revealed the immunosuppressive potential, they were also shown to play a crucial role in the induction of long-term antitumor immune response after PDT [
31]. Therefore, further mechanistic studies should be introduced to reveal the role of neutrophils subpopulations in PDT-mediated inflammation.
Increase in IDO activity after the PDT treatment is accompanied by the rise in the percentage of Treg in TDLN. It was previously shown that PDT increases the number of Treg in mice and the removal of Treg is associated with improved antitumor efficacy of PDT [
11]. Here we show that IDO inhibition with EPA brings back the number of Treg to control values indicating potential involvement of this enzyme in PDT-induced Treg expansion. Kynurenic acid, a direct product of tryptophan degradation catalyzed by IDO, was shown to induce Treg expansion, by activating aryl hydrocarbon receptors [
32].
Considering a number of immunoregulatory mechanisms associated with IDO activity, the selective inhibitors of this enzyme have been developed and progressed to clinical trials. IDO is an important immunoregulatory enzyme that evolved to control exuberant immune response mitigating tissue damage and immunopathology.
We observed exaggerated toxicity of orally applied EPA that evolved directly after PDT. Moreover, injection of EPA intratumorally prolonged the survival of PDT-treated mice what is in line with observation done by Lu et al. [
33]. Importantly, in our experimental settings the local administration of EPA did not protect completely from toxic effects. Acute inflammation was associated with IL-6 release and massive infiltration of granulocytic myeloid cells (CD11b
+Ly6G
+) to the tumor bed as well as to the TDLNs. The toxic reaction was not reported in the studies where depletion of Treg was combined with PDT [
11]. Nevertheless, increased concentrations of IL-6, as well as TNFα and IL-12, were previously described in response to simultaneous IDO inhibition and administration of apoptotic cells [
34]. Inhibition of this enzyme resulted in the loss of self-tolerance to apoptotic cell-associated antigens and susceptibility to lupus-like autoimmunity. These data indicate that IDO plays an important role in the regulation of immune tolerance to antigens released from dying cells and suggest that PDT-induced tumor cells death may be one of the triggers leading to increased immunopathology.
IL-6 was previously shown to be induced by PDT in vitro and in vivo but also in cancer patients [
31,
35]. The reports on its impact on the antitumor efficacy of PDT were discordant showing either negative [
36], positive [
37] or no effect [
38]. Systemic inflammation can be ameliorated by administration of IL-6-neutralizing antibodies that are clinically validated in the management of immune-mediated adverse events, developing after cancer immunotherapies with checkpoint inhibitors or adoptive treatment with chimeric antigen receptor (CAR) T cells [
39,
40]. Although IL-6 neutralization significantly potentiated antitumor efficacy of PDT, it eliminated the additional benefit of IDO inhibitor to the treatment. It is possible that IL-6 neutralization is associated with decreased IDO expression, as was shown in the previous study [
41], that would tuck away the target for EPA making the treatment superfluous. Intriguingly, and in contrast to our observations, several recent studies have shown that combined blockade of IL-6 and PD-1/PD-L1 checkpoint molecules promotes tumor infiltration of IFN-γ-producing CD4
+ T cells and exerts synergistic antitumor effects [
42,
43].
Although IDO is an attractive target for cancer immunotherapies, its role in the regulation of inflammation remains not completely understood. Importantly, the promising results obtained with IDO inhibitors in mouse tumor models have not been successfully translated into the clinical trials. Current results are rather disappointing as IDO inhibition did not improve the immunotherapy of cancer. These failed clinical trials with IDO inhibitors underlined the complexity of tryptophan metabolism. The tryptophan can be degraded not only by IDO but also by tryptophan-2,3-dioxygenase or can be introduced into the serotonergic pathway, which products have also immunosuppressive properties [
44]. On the other hand, in light of the success achieved by immunotherapies restoring the antitumor functions of T cells, studies on IDO inhibition provide a strong rationale for therapeutic targeting of this enzyme. Based on our results, we postulate to further elucidate the role of IDO in the systemic immune response to avoid unexpected acute reactions.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.