The PD‐L1‐ and IL6‐mediated dampening of the IL27/STAT1 anticancer responses are prevented by α‐PD‐L1 or α‐IL6 antibodies

Interleukin‐27 (IL27) is a type‐I cytokine of the IL6/IL12 family and is predominantly secreted by activated macrophages and dendritic cells. We show that IL27 induces STAT factor phosphorylation in cancerous cell lines of different tissue origin. IL27 leads to STAT1 phosphorylation and recapitulates an IFN‐γ‐like response in the microarray analyses, with up‐regulation of genes involved in antiviral defense, antigen presentation, and immune suppression. Like IFN‐γ, IL27 leads to an up‐regulation of TAP2 and MHC‐I proteins, which mediate increased tumor immune clearance. However, both cytokines also upregulate proteins such as PD‐L1 (CD274) and IDO‐1, which are associated with immune escape of cancer. Interestingly, differential expression of these genes was observed within the different cell lines and when comparing IL27 to IFN‐γ. In coculture experiments of hepatocellular carcinoma (HCC) cells with peripheral blood mononuclear cells, pre‐treatment of the HCC cells with IL27 resulted in lowered IL2 production by anti‐CD3/‐CD28 activated T‐lymphocytes. Addition of anti‐PD‐L1 antibody, however, restored IL2 secretion. The levels of other TH1 cytokines were also enhanced or restored upon administration of anti‐PD‐L1. In addition, we show that the suppression of IL27 signaling by IL6‐type cytokine pre‐stimulation—mimicking a situation occurring, for example, in IL6‐secreting tumors or in tumor inflammation–induced cachexia—can be antagonized by antibodies against IL6‐type cytokines or their receptors. Therapeutically, the antitumor effects of IL27 (mediated, e.g., by increased antigen presentation) might thus be increased by combining IL27 with blocking antibodies against PD‐L1 or/and IL6‐type cytokines.

Cytokines such as interferons, IL2, IL12, IL7, IL15, IL27, and others are tested as potential anticancer immuno-stimulatory agents and some have progressed into clinical trials. [3][4][5] IFN-is one of those cytokines long known to be involved in antitumor defense, because IFN-R and STAT1 deficiency has been described to increase cancer susceptibility upon treatment with chemical pathogens in mice. 6,7 Interferon adjuvant therapy has unfortunately invariably been reported to have high toxicity, for example, in hepatocellular carcinoma (HCC) patients. 8 Also the antitumor effects of other cytokines such as IL12 have been attributed to expression of IFN-. 9,10 IL27 has been described to have pro-or anti-inflammatory functions in different systems [11][12][13][14] and it is capable of exerting antitumor effects. IL27 seems to be better tolerated than interferons 8 and IL12 15 in preclinical models, probably because less IFN-is released by IL27 compared to IL12. Combination of IL27 and IL2 has been shown to be strongly synergistic in a murine neuroblastoma model. 16 Treg levels were suppressed and IFN-levels elevated for the combination treatment compared to the single treatments. IL2 therapy in humans has been shown to have a relatively low risk of severe side effects, although it was only effective in 10-20% of the patients. [17][18][19] Therefore it still remains to be seen whether a combination of IL27 and IL2 will be more effectiveas in the murine model-and still well tolerated.
IL27 does, however, not only act on immune cells but can also elicit antitumor responses in cancer cells directly, such as antiproliferative, pro-apoptotic, anti-angiogenic, and anti-metastatic functions, which have recently been reviewed extensively. 13 Briefly, IL27 showed antiproliferative effects on various melanoma cell lines 46 and, like IFN-, inhibition of angiogenesis. 20,47 The growth of poorly immunogenic B16/F10 melanoma tumors in syngeneic mice was reduced when the cells had been transfected with scIL27. 15 We have previously reported IFN--like responses and antiviral activity of IL27 in hepatoma cells, primary hepatocytes, and hepatic stellate cells, 48,49 as has also been described for other cell types. 13,24,50 We have described before that IL27 signaling is susceptible to inhibition by SOCS3, which is induced, for example, upon phosphorylation of STAT3 by IL6-type cytokines. 50 Many tumors have been shown to secrete IL6-type cytokines and/or to have constitutive STAT3 activation (and thus elevated SOCS3 levels). [51][52][53] In addition, IL6 has been shown to be crucial for colorectal cancer and HCC development. [54][55][56] In cancer cachexia, IL6 and other pro-inflammatory cytokines are produced due to a tumor-induced systemic inflammation. 57,58 IL6 and TNF-blockade have been used successfully to suppress cancer cachexia. 59,60 Here we describe that IL27 can stimulate cancer cells of different lineages and further characterize IL27 signaling in these cell lines. Microarray analysis showed that IL27 induced an IFN--like, STAT1-dependent transcriptional response, mediating, for example, up-regulation of genes associated with increased tumor immune clearance and also mediating gene expression profiles associated with cancer immune escape, for example, PD-L1 and IDO-1 up-regulation.
We show that IL27-induced PD-L1 can suppress IL2 production by peripheral blood mononuclear cells (PBMCs) and that this effect can be rescued by anti-PD-L1 antibody administration. Additionally, we used an IL6-type cytokine pre-stimulation approach to mimic tumor inflammation-mediated STAT3-and SOCS3-dependent IL27 inhibition and we show that this inhibition can be prevented by blocking anti-IL6, -IL6R, or -OSM antibodies.

Cytokine expression analysis in HCC/PBMC cocultures
The experiment was performed as described 61 with minor changes.
The cryopreserved PBMCs were from Zenbio. Huh7 and Hep3B

Cell lysis and Western blot analysis
Cell lysis was performed as described before. 62   cates. The STAT3 specific luciferase reporter, pXP2d2-rat pancreatitisassociated protein 1 (rPAP1), was described earlier. 65

Flow cytometry
Cells were re-suspended in cold PBS supplemented with 5% FBS

Whole genome microarray analysis
The cell lines (PH5CH8, Hep3B, Huh7, NHEM, A375, MelJuso, NCM460, HCT116 and HT29) were left untreated or were stimulated with hyper-IL6 (hy-IL6) 66  Partek Genomics Suite software, v. 6.6. In order to remove uninformative features, only genes with a log 2 expression level above 5 in at least one sample were considered for further analysis. Statistical analysis was performed in R 68 and using linear models in an empirical Bayesian approach implemented in the R/Bioconductor package limma. 69 A global linear model accounting for cell lines and treatments was fit, and then P-values were estimated for various contrasts. P-values were adjusted for multiple comparison by Benjamin-Hochberg's FDR method 70 and genes with FDR < 0.05 were selected.

Principal component analysis
The microarrays of the cell lines responding to the 4 cytokine treatments (PH5CH8, Hep3B, Huh7, NHEM, A375, NCM460 and HT29) were subjected to principle component analysis (PCA). Computation and plotting were performed using R (v3.4.0) 68 and Rstudio (v1.1). 71 The PCA was computed using the R package FactoMineR (v1.36) 72 and plotted by the R package factoextra (v1.0.4). 73 Other plots were generated using the R package ggplot2 (v2.1) 74 and the collection of packages tidyverse (v1.1). 75 All experiments were scaled to the same unit, as the IFN-stimulation for the NHEM cells possessed a larger variance compared to the other experiments.

IL27 triggers phosphorylation of STAT1 and STAT3 transcription factors in many different cancer cell lines
IL27 has been described to use the ubiquitously expressed common signal transducing receptor of the IL6-type cytokines, IL6R , in combination with IL27R . We stimulated a number of cancer and noncancer cell lines from various tissues (see materials and methods) with IL27 to test their responsiveness (Fig. 1A). Phosphorylation of the STAT factors was investigated by Western blot analysis after one hour of stimulation. The cell lines all reacted to IL27 with a robust phosphorylation of STAT1 (pSTAT1) and a phosphorylation of STAT3 (pSTAT3), the latter being only slightly over background for some of the cell lines. STAT5 and STAT4 were not phosphorylated by IL27 in any of the cell lines (Data no shown).
Using flow cytometry we analyzed the surface expression of the IL27 signaling receptors, IL6R , and IL27R , on a selection of cell lines ( Fig. 1B) and found that both receptors were expressed on the surface of all investigated cell lines.

In contrast to IFN-, IL27 induces a stronger STAT3 phosphorylation, which is insufficient to activate STAT3-mediated reporter gene transcription
We compared the STAT1/3 phosphorylation pattern obtained for IL27 signaling to efficient activators of each type of STAT (OSM for STAT3 and IFN-for STAT1) by Western blot immuno-detection analyses in a number of cell lines ( Fig. 2A). IL27-mediated STAT1 phosphorylation was robust, as for IFN-, and STAT1 protein expression was induced.
As expected, OSM did not induce STAT1 protein up-regulation. Phosphorylation of STAT3 by IL27 was generally weaker than that induced by the strong STAT3 activator OSM. Thus, IL27 stimulates many cell types resulting in robust phosphorylation of STAT1 and a less efficient STAT3 phosphorylation (compared to IFN-and OSM, respectively).
To test for transcriptional activity of STAT3 downstream of IL27 we used a STAT3-specific reporter gene assay containing the rPAP1 promotor, 65 which is very sensitive and highly specific for STAT3 ( Fig. 2B). 50 IL27, as IFN-(which in contrast to IL27 only leads to STAT1 phosphorylation), did not activate the STAT3-specific reporter gene construct, whereas hy-IL6-and OSM-induced transcription of the reporter gene. Of note, Hep3B cells do not express the oncostatin M receptor (OSMR) and thus do not activate reporter gene expression upon OSM. 50 Thus, although IL27 led to detectable STAT3 phosphorylation, it did not induce STAT3-specific promotor construct transcription (similarly to IFN-, which does not trigger STAT3 phosphorylation in our system).

IL27 and IFN-stimulations lead to similar transcriptome expression profiles
The fact that we did not detect STAT3-dependent reporter gene  (Fig. 3B).
Interestingly, the microarray analyses also revealed that IFN-and IL27 regulate a set of genes associated with immune evasion of cancer (Fig. 3C). In a potential cancer treatment involving IL27, however, immune-inhibitory proteins such as CD274 (hence forward referred to as PD-L1), IDO-1, IL18BP, or LAP3 might prevent an efficient immune clearance and instead induce immune tolerance. These genes were described to limit inflammation and tissue destruction upon IFNand IL27 stimulation. In chronic cancer inflammation these genes are associated with a bad prognosis.  Fig. 2). Because HCC is our main focus, 50 we also performed quantitative Western blot analysis for three HCC cell lines and the PH5CH8 cells for several proteins (Fig. 4D). In general, proteins associated with increased immune clearance of cancer such as Tap2 and MHC-I were upregulated upon IL27, IFN-and IFN-treatment,  D) HepG2, Hep3B, Huh7, and PH5CH8 cell lines were stimulated for different time points with IL27, OSM, or IFN-or were left untreated. Western blots of the lysates were then immunodetected using antibodies against IDO-1, TAP2, and STAT1 and quantitated as described in materials and methods and represented as relative signal intensity compared to the strongest signal on the blots (mean and standard deviation of 3 biological replicates are shown). Tubulin served as control detection (see also Supporting Information Figure 2C) but not after OSM or hy-IL6 treatment in different cell types (Fig. 4A and B and Supporting Information Fig. 2A and C) (for HCC cells see also 50 ). Thus, IL27-treated cells are likely to present more antigen on the cell surface than untreated cells. FACS analysis showed that IL27, IFN-, and IFN-led to an efficient PD-L1 protein up-regulation in different cancer cell lines, and again, OSM or hy-IL6 did not ( Fig. 4C and Supporting Information Fig. 2B). Interestingly, quantitative Western blot analysis revealed that IDO-1 was inefficiently induced by IL27 in contrast to STAT1 or TAP2, whereas IFN-induced IDO-1 more efficiently (Fig. 4D and Supporting Information Fig. 2C). In addition, IDO-1 protein induction was especially strong in the IFN--stimulated nontransformed cell line PH5CH8.

In cocultures of HCC and PBMC, HCC cell-expressed PD-L1 can suppress IL2 production in anti-CD3/-CD28 activated T lymphocytes
To analyze if IL27-mediated PD-L1 expression in HCC cells can have an immune-suppressive effect, we performed coculture experiments in which we pre-stimulated HCC cells with IL27 for 24 hours, washed IL27 away and incubated the PD-L1-expressing HCC cells with anti-CD3/CD28-activated PBMCs for 3 days (Fig. 5A). Pre-treatment of the HCC cells with IL27 resulted in lower IL2 production in the coculture, which shows that PD-L1 expressed at the surface of the HCC cells can influence T-cell functions (Fig. 5B). Addition of a blocking PD-L1 antibody (concomitantly with the PBMCs), raised the IL2 secretion at least to the level observed before (or higher) ( Fig. 5B and Supporting Information Fig. 3A). The levels of other pro-inflammatory cytokines (e.g., IL1 , TNF-, and MCP1) were either restored or upregulated by anti-PD-L1 administration (Fig. 5C). However, IL6 levels, a cytokine described to mediate "pro-tumor" effects, were also decreased by IL27 treatment and efficiently increased upon anti-PD-L1 administration ( Fig. 5D and Supporting Information Fig. 3B).

The inhibition of IL27 responses by IL6-type-cytokines is prevented by administration of blocking antibodies against these cytokines or their receptors
IL27 signaling is susceptible to inhibition by SOCS3, which is induced, for example, upon phosphorylation of STAT3 by IL6-type cytokines. 50 IL6-type cytokine and non-IL6-type cytokine-mediated STAT3 phosphorylation is often observed in a cancer-associated inflammatory context and is thought to have "pro-tumor" effects 77 (see also Supporting Information Fig. 5). In addition, as shown in Figure 5, IL6 levels are also upregulated upon anti-PD-L1 administration.
To mimic cancer-associated constitutive activation of STAT3, we pre-stimulated the cells with OSM, hy-IL6 or left them untreated for 4 h before a 24 h stimulation with IL27. To monitor IL27-induced responses, we measured PD-L1 surface expression by flow cytometry and detected STAT1, TAP2, and IDO-1 by Western blot analysis in HCC cell lines. OSM or hy-IL6 pre-stimulation (Fig. 6A, solid line) reduced the IL27-dependent PD-L1 up-regulation (Fig. 6A, red peak) almost to the level of the non-IL27-stimulated sample (Fig. 6A, gray peak).
The identical observations were made for STAT1, TAP2, and IDO-1 as investigated by Western blot analysis (Fig. 6B). Thus, an anti-IL6-type cytokine approach might have beneficial effects in IL27 antitumor treatment, if constitutive STAT3 phosphorylation is observed. We have described before that IL27 can be inhibited by SOCS3 (which is a STAT3 target gene). We show that phosphorylation of STAT3 (Fig. 6B, lower panels) and up-regulation of SOCS3 mRNA (Fig. 6C) and protein (see Supporting Information Fig. 4) are suppressed by anti-IL6 treatment.

DISCUSSION
We show and discuss in the following that (i) IL27 induces STAT We investigated a number of cell lines from various tissue origins concerning their responsiveness to IL27. IL27 induced a robust phosphorylation of STAT1 and a phosphorylation of STAT3, the latter being only slightly over background for some of the different cell lines (Fig. 1A). The investigated cells also expressed the IL6R and IL27R , the canonical receptor subunits for IL27 signaling (Fig. 1B). 50 Since IL27 is a noncovalent dimer of the EBI3 and p28 subunits, a dissociation of IL27 in the presence of other competing cytokine subunits (p35, CLF1) might occur. p28 alone or in complex with CLF1 (cytokine-like factor 1) has been described to signal through IL6R / IL6R (although only at high cytokine concentrations) to activate STAT3, 78,79 which (i) is described to be pro-tumorigenic 80 and (ii) can inhibit IL27 signaling (by SOCS3 induction). 50 p28 (also called IL30) expression correlates with advanced grade of prostate cancer and promotes breast cancer growth. 81,82 In addition, local  IL2 (B), MCP1, IL1 , TNF-(C) and IL6 (D) levels are expressed as the mean ± standard deviation of 3 biological replicates. One-way ANOVA with Bonferroni posttest was performed using Graphpad InSTAT version 3.10 (www.graphpad.com). For statistical analysis error probabilities <0.05 were considered to be significant, ***: P < 0.001, **: P < 0.01, *: P < 0.05 F I G U R E 6 Blocking antibodies against IL6type cytokine or receptors prevent downregulation of IL27 signaling. (A) Expression of PD-L1 on the HCC cell lines HepG2, Huh7, and Hep3B was studied by flow-cytometry. Cells were left untreated (gray fill) or were stimulated with IL27 alone (red fill) or treated for 4 h with hy-IL6 or OSM prior to the IL27 treatment for 24 h (solid line). The dotted line indicates the situation in which blocking antibodies against IL6, IL6R, or OSM were added together with the IL6 or OSM pre-treatment. (B) HepG2, Hep3B, Huh7, and PH5CH8 cells were stimulated as described in (A) but using only hy-IL6 and anti-IL6. Western blots of the lysates were immuno-detected for the STAT1 target genes IDO-1, TAP2, STAT1. Vinculin and tubulin detections were used as loading controls. Phosphorylation of STAT3 was investigated on a separate blot using STAT3, vinculin and tubulin as controls. (C) HepG2, Hep3B, Huh7, and PH5CH8 cells were stimulated as described in (B). The cells were lysed after 6 h of treatment. SOCS3 mRNA expression was investigated by qPCR. Results show the mean of 3 biological replicates as a percentage of the highest SOCS3 induction with standard deviation expression of p35 (a subunit of IL12), could lead to competitive displacement of p28 from EBI3, to form IL35 (composed of EBI3 and p35), a cytokine that exerts its immunosuppressive functions by promoting iTreg differentiation and activity. 83,84 Thus, IL27 administration as a potential anticancer treatment should rather be performed with a covalently linked dimer of p28 and EBI3 to avoid the aforementioned pro-tumorigenic effects.
With respect to STAT factor activation, almost all tested cells reacted to IL27 with a robust phosphorylation of STAT1 (comparable to the IFN-response) and a phosphorylation of STAT3, the latter being weaker compared to the OSM-induced STAT3 phosphorylation ( Fig. 2A). Because studies in immune cells have discussed a role of STAT3 in transcriptional responses through IL27, [85][86][87] we tested for the presence of transcriptionally active STAT3 by using the STAT3-specific and very sensitive rPAP reporter gene construct (Fig. 2B). Only OSM, but not IL27 or IFN-, induced STAT3-dependent reporter gene transcription in the tested cells of different tissue types.
The possible reasons for such an inefficient STAT3 response have been discussed in detail before. 50 Whole genome microarray analyses of 7 cell lines (3 different tissue types) stimulated with IL27, IFN-, hy-IL6, or OSM showed that the IL27 response closely recapitulates an IFN-response ( Fig. 3 and Supporting Information Fig. 1A and B). IL27, like IFN-, regulates mRNAs involved in immune regulation and antigen presentation but is also related to other canonical functions of IFN-, for example, the antiviral defense (data not shown). In the antigen presentation pathway, genes encoding MHC-I subunits (HLA-A, -B, -E, -G, B2M), the immunoproteasome (PSMB-8 to -10), the S11 proteasome (PSME-1 and -2), and TAP transporters (TAP1 and 2) were upregulated (Fig. 3B). The up-regulation of the antigen presentation pathway is associated with increased immunogenicity of tumors. In addition, CXCL9 and 11 were upregulated, which promote anti-angiogenesis and recruit immune cells associated with TH1 responses. Interestingly, a subset of genes involved in immune escape of cancer (LAP3, IL18BP, IDO-1, and PD-L1 (CD274)) were also upregulated by IL27 and IFN-88 (Fig. 3C). These genes can promote immune evasion of cancer if overexpressed, as part of a dysregulated interferon response. 89 Very recently, a number of the proteins mentioned earlier was found to be upregulated by IL27 in ovarian cancer, neuroblastoma, and lung adenocarcinoma cells. 88,90 IL27 upregulates PD-L1 in leukocytes 37,38,91,92 and other cell types including cancer cells. 61,90,93,94 In the subsequent validations we showed that IL27, like IFN-, also upregulates the protein levels of a number of genes involved in increased cancer immune clearance (STAT1, TAP2, MHC-I) and proteins involved in immune escape of cancer (PD-L1, IDO-1) ( Fig. 4 and Supporting Information Fig. 2). Interestingly, PD-L1 is upregulated following IL27 and IFN-stimulation, but not upon OSM in all tested cell lines.
Because HCC was our main focus here and represents a tumor with still limited treatment options, we tested the induction of the aforementioned genes in more detail in these cells and in the nontransformed PH5CH8 liver cell line (Fig. 4D). The liver is characterized by an intrinsic immune suppressive microenvironment, which has been described to impede the effectiveness of anticancer vaccines. 95 Moreover, HCC is known to lack the antigen processing proteins necessary to present tumor antigens, which allows escape from CD8 + lymphocyte killing. 31  Thus, IL27 and IFN-can have immune-promoting functions but could also lead to tumor immune escape, a characteristic that might interfere with tumor clearance upon a potential IL27 therapy. Because the PD-L1 gene is efficiently induced by IL27, IL27 immune therapy might profit from anti-PD-L1 antibody co-treatment. Targeting PD-L1/PD-1 has shown impressive antitumor effects in solid tumors, [96][97][98][99] including HCC. 100 Clinical trials using immune checkpoint inhibitors, for example, anti-CTLA4 and anti-PD-1, demonstrated that the treatment was well tolerated. 100 In coculture experiments of HCC cell lines with PBMCs we could show that an IL27-dependent PD-L1 expression on HCC cells can decrease IL2 production by the PBMCs.
IL6, a "pro-tumor" cytokine, was also secreted into coculture supernatants upon anti-PD-L1 administration (Fig. 5D). IL6 and OSM were shown to be present in inflammatory tumors. Inflammation has been established to be involved in cancer development, 101 and constitutive activation of STAT3 found in many cancers is mediated by different mechanisms. 57,58,80,[102][103][104] In addition, IL6-type cytokines have been shown to be essential players in the development of at least colon and liver cancers. 55,77,105,106 We have shown before that IL6-type cytokine pre-stimulation, which induces SOCS3 expression, can suppress a subsequent IL27-response in a SOCS3-dependent manner. 50,107 Due to this cross-inhibition of "pro-tumor" IL6-type cytokines, IL27 responses and thus its antitumor efficacy might be considerably reduced in "inflammatory" tumors. We now show that anti-IL6-type cytokine antibodies can reconstitute efficient IL27 responses (Fig. 5).
In addition, IL6 can have other immunosuppressive effects in cancer.

Tumor-induced systemic IL6 can induce systemic metabolic changes.
It induces lipolysis of white adipose tissue, which can cause hyperlipidemia and results in insulin resistance of skeletal muscle (reviewed in 108 ). In the liver, IL6 inhibits PPAR , which reduces ketone generation under low energy/insulin resistance. The resulting hypoketonemia stimulates the adrenal gland to induce glucocorticoids, which suppress intratumoral immunity and even prevent anticancer immunotherapy with anti-PD-L1 antibodies. 109 In addition, IL6 can act locally in the tumor on immune cells to suppress anticancer immunity (reviewed in 33 ). IL6 favors the development of M2 macrophages, MSDCs and T H 17 cells, which is associated with secretion of immunosuppressive factors such as IL10 and IL4 (which again inhibit T H 1 cells, CD8 positive CTLs and NK cells). Cancer-associated fibroblasts and regulatory B cells have also been described to secrete IL6. Thus, blocking IL6 in patients with a prominent tumor-induced IL6-type cytokine-dependent inflammation might have more anticancer effects F I G U R E 7 Major IL27-mediated signaling pathways in cancer cells and possible treatment options to improve immune clearance of cancer cells. Pathways and treatments that promote antitumor actions are indicated by blue arrows. Pathways that promote tumor immune escape are indicated using red arrows. Black boxes indicate those blocking antibodies that have been used in this study. Already FDA-approved blocking antibodies inhibiting the same targets mentioned above are shown in dark green boxes. An IDO-1 inhibitor that entered clinical trials but is not yet FDA-approved is indicated in a grey box. In tumors with high IL6-type cytokine levels, anti-cytokine-receptor-or anti-cytokine-antibodies could be used in addition to IL27, to relieve the SOCS3-mediated cross-inhibition of IL6-type cytokines on IL27 signaling. PD-L1 is expressed in all investigated cancer cell types in response to IL27 and thus anti-PD-L1 antibodies could be envisaged as standard co-treatment with IL27 (or Interferon-) therapies to prevent immune cell deactivation (e.g., a CD8 + T-cell as shown in the scheme) by cancer cell-expressed PD-L1. IDO-1 is not induced by IL27 in all cells and thus IDO-1 inhibitor treatment could only be applicable to a subset of tumors than just preventing the inhibition of a potential IL27 treatment (see also Supporting Information Fig. 5).
In summary, we hypothesize that HCC (characterized by an intrinsic immune suppressive microenvironment) could benefit from IL27 treatment, which upregulates proteins involved in antigen presentation (Fig. 7). In addition, a co-treatment with anti-PD-L1 or anti-PD-1 antibodies might prevent immune suppressive effects also mediated by prolonged IL27 or interferon stimulation. A number of studies even suggested that blockade of the PD-1/PD-L1 pathway is a prerequisite for the necessary stimulatory functions of interferons in the antitumor response. [110][111][112][113][114] Due to the suppression of IL27 signaling, mediated by IL6-type cytokine-induced SOCS3 protein, patients with high circulating IL6 levels (e.g., by cancer inflammation-associated cachexia or IL6 producing tumors) will very likely only be responsive to IL27 Gemeinschaft (DFG). We thank Demetra Philippidou for excellent technical assistance and Sebastien Plançon for his expert help concerning flow cytometry. We also thank our collaborator Prof. Dr. Nobuyuki Kato (Department of Molecular Biology, Okayama University, Japan) for providing the PH5CH8 cells; Prof. Dr. Stefan Rose-John (University of Kiel, Germany) for providing hyper-IL6; and Prof. Jan Tavernier for providing the rPAP reporter gene construct.

DISCLOSURE
The authors declare no conflicts of interest.