The aryl hydrocarbon receptor promotes inflammation‐induced dedifferentiation and systemic metastatic spread of melanoma cells

The aryl hydrocarbon receptor (AHR) is a ligand binding‐transcription factor of the basic helix‐loop‐helix family regulating multiple cellular functions such as differentiation, cell cycle, apoptosis, and inflammatory reactions. In neoplastic diseases, the AHR has been described to modulate proliferation and differentiation in dichotomous ways, either inhibiting or augmenting the growth of tumors. The precise role of AHR in melanoma is mostly unknown. Here, we report a functional effect of AHR activation on inflammation‐induced melanoma cell dedifferentiation and the development of lung metastases in a mouse model. Via in silico analyses of “The Cancer Genome Atlas” human melanoma cohort, we detected a correlation between AHR expression levels and a dedifferentiated melanoma cell phenotype with an invasive gene signature, which we were able to functionally recapitulate in a panel of human melanoma cell lines. Both human and mouse melanoma cell lines upregulated AHR expression after inflammatory stimulation with tumor necrosis factor‐α (TNF‐α). Activation of AHR in human and mouse melanoma cell lines with the endogenous ligand formylindolo(3,2‐b)carbazole (FICZ) promoted inflammation‐induced dedifferentiation in vitro. Importantly, mouse melanoma cells with CRISPR/Cas9‐mediated disruption of the AHR gene showed impaired in vivo tumor growth after transplantation in the skin as well as decreased numbers of spontaneous lung metastases. Taken together, our results demonstrate a functional role for AHR expression in melanoma development and metastatic progression. This provides a scientific basis for future experiments that further dissect the underlying molecular mechanisms and assess the potential for AHR inhibition as part of multimodal melanoma treatment strategies.


| INTRODUCTION
Despite increasing efforts for preventive and therapeutic options, the incidence of cutaneous melanoma continues to increase in European countries and the United States. This trend is projected to uphold in the near future. 1 The major environmental risk factor for melanoma development is UV light exposure. UV irradiation can induce oncogenic mutations and at the same time can promote an inflammatory microenvironment that supports the survival, proliferation, and migration of genetically altered melanocytes. [2][3][4][5] Proinflammatory mediators such as tumor necrosis factor-α (TNF-α ) can promote the emergence of dedifferentiated melanoma cell phenotypes that are thought to reactivate the migratory potential of their embryonal precursors in the neural crest. 6 UV irradiation also generates the tryptophan derivative formylindolo (3,2-b)carbazole (FICZ) in the skin which is a high-affinity ligand for the aryl hydrocarbon receptor (AHR). The AHR is a transcription factor of the basic helix-loop-helix Per-Arnt-Sim family that was initially described in xenobiotic metabolism and is highly expressed in barrier organs like the skin. [7][8][9][10][11] First observations linking AHR activation with pigmentation and melanocyte biology derive from accidental mass-poisoning incidents, for example, in Taiwan or Japan. 12,13 In support of this connection, it was found that mice lacking the AHR gene show impaired skin tanning. 14 Large-scale gene expression analyses in the Cancer Cell Line Encyclopedia (CCLE) revealed that melanoma cell lines can also express AHR. 15 Furthermore, shRNA-mediated knockdown of AHR led to growth inhibition, providing first evidence for a functional role of AHR. 16 The link between AHR and melanoma was further strengthened by recent genome-wide association studies that implicate the AHR gene locus with increased risk to develop cutaneous melanoma. 17,18 In our work, we further explored a potential function of AHR signaling in melanoma pathogenesis. We show that the expression of AHR is enhanced by inflammatory mediators and augments inflammation-induced dedifferentiation in human and mouse melanoma cells. Moreover, we observe that the AHR expression promotes metastatic tumor progression in a syngeneic mouse model in vivo. were authenticated using STR profiling in 2020 (Microsynth, Balgach, Switzerland). HCmel cell lines were derived from Hgf-Cdk4 R24C mice as described previously. 5
The aryl hydrocarbon receptor (AHR) has important functions in mediating xenobiotic metabolism and in regulating numerous enzymes and transcriptional programs in cells.
AHR also is expressed in melanoma and its knockdown has been shown to inhibit tumor cell growth. In our study, the AHR was found to promote inflammation-induced dedifferentiation of both human and mouse melanoma cells in vitro.
In syngeneic immunocompetent mice, AHR expression pro-

| Cell growth assay
Cell growth was quantified by counting viable cells using Trypan blue.
In short, genetically manipulated HCmel12 cells and HCmel12 control cells (transfected with an "empty" CRISPR vector) were seeded at low density in biological triplicates and incubated at standard conditions for 72 hours. After this time, all cells were harvested and collected following standard protocols. Cell number was determined using a hemocytometer with a DME microscope (Leica, Wetzlar, Germany).

| Immunoblot analysis
Whole cell lysates were extracted from cultured cells using the M-

| Quantitative real-time PCR
Total RNA from cultured cells was isolated and purified using the NucleoSpin RNA XS kit (Macherey-Nagel, Düren, Germany). RNA con-

| Mice
Wild-type C57BL/6J mice were purchased from Janvier (Le Genest-Saint-Isle, France) or taken from own breeding. Age-and sex-matched cohorts of mice were randomly allocated to the different experimental groups at the start of each experiment.

| Tumor transplantation
Cohorts of syngeneic C57BL/6J mice were injected intracutaneously with 2 × 10 5 HCmel12 CRISPR ctrl cells or a mixture of three HCmel12 AHR CRISPR KO monoclones in equal proportions resuspended in 100 μL PBS (Life Technologies) into the right flank.
Tumor growth was monitored by inspection and palpation. Tumor size was measured at least twice times weekly with a vernier caliper and recorded as the mean diameter of two perpendicular measurements.
Mice were sacrificed when tumors exceeded 20 mm in diameter or when signs of illness were observed. Lung metastases were counted by macroscopic inspection. All experiments were performed in groups of five or more mice and repeated independently three times.

| Statistical analyses
We considered tumor growth, cell proliferation and migration as well as qRT-PCR expression as normally distributed with similar variances.
For these data, we performed parametric testing with unpaired, two-  Figure 1A). Differential gene expression analysis between melanoma samples with high AHR levels and those with low AHR levels revealed 2055 genes with a log2-fold change >j2j (Table S2). Grouping of genes downregulated in AHR high samples by Gene Ontology Biological Process (GO-BP) terms showed strong enrichment of genes involved in the cellular pigmentation machinery ( Figure 1B). Further analyses of the melanoma cohort revealed that the expression of AHR was inversely correlated with the expression of pigmentation genes such as PMEL and TYRP1 ( Figure 1C). Additionally, the expression of AHR was inversely correlated with a gene signature that has previously been associated with a proliferative phenotype and directly correlated with a gene signature associated with an invasive phenotype ( Figure 1D). 20 Incidentally, the proliferative gene set largely contains melanocytic differentiation antigens, whereas the invasive gene set includes markers of a dedifferentiated cell state.
Taken together, this supports the notion that AHR may functionally participate in melanoma pathogenesis.

| Activation of AHR enhances inflammationinduced dedifferentiation in differentiated human melanoma cell lines
Next, we sought to recapitulate the association between AHR and a dedifferentiated, invasive phenotype also in human cell lines. For this, we performed bioinformatic analyses on gene expression data from the CCLE. Again, AHR was negatively correlated with the proliferative gene signature and putatively correlated with the invasive gene signature ( Figure S1A,B, left panel). These findings were also reproducible in a set of melanoma cell lines established in our laboratory that was selected to reflect the mutation spectrum and differentiation phenotypes in human melanoma ( Figure 2A; Figure S1A,B, right panel). From these, we selected three differentiated and three dedifferentiated cell lines based on their expression of a set of bona fide pigmentation genes for further functional analyses. We previously reported that TNF-α can shift melanoma cells towards a more dedifferentiated cell state. 5 We therefore

| Inflammation induces the expression of AHR and modulates inflammatory responses in HCmel12 cells
To

| AHR-deficient HCmel12 cells maintain a more differentiated phenotype upon inflammatory stimulation
In order to further investigate the function of AHR, we generated genetic knockouts via CRISPR/Cas9 genome editing in HCmel12 melanoma cells ( Figure S2A). Successful disruption of the gene was validated via next generation sequencing and Western blot analyses ( Figure S2B; Figure 5A). AHR knockout cells no longer responded to FICZ with the induction of the prototypic AHR target gene Cyp1a1, confirming the disruption of functional AHR signaling ( Figure 5B). AHR knockout HCmel12 cells exhibited a decreased shift towards a dedifferentiated state compared to wild-type controls and failed to further dedifferentiate in response to FICZ ( Figure 5C,D). This further supports a potential role for the UV-induced metabolite FICZ and the AHR signaling pathway in melanoma pathogenesis.

| Loss of AHR reduces HCmel12 melanoma growth and metastasis
Since we found that AHR expression was correlated with an expression profile associated with invasiveness in human melanomas and because AHR enhanced inflammation-induced dedifferentiation in human and mouse melanoma cell lines, we were interested to evaluate whether AHR would influence melanoma growth and metastasis in vivo. Therefore, we transplanted AHR-competent and -deficient HCmel12 melanoma cells into groups of syngeneic, immunocompetent C57BL/6J mice ( Figure 6A). The median survival of mice inoculated with AHR knockout melanoma cells was slightly but significantly prolonged compared to mice inoculated with control melanoma cells (median of 30.5 vs 27 days, log-rank test P < .05, Figure 6B). Strikingly, AHR knockout melanomas showed an intense pigmentation in vivo ( Figure 6C). Importantly, mice transplanted with AHR knockout melanoma cells revealed a significant reduction in the number and frequency of pulmonary metastases (median of 8 vs 0, Mann-Whitney U test P < .005, Figure 6D). Subsequent experiments revealed a slightly reduced proliferation rate of AHR-deficient HCmel12 cells compared to wild-type controls as well as an impaired TNF-induced migration in a transwell migration assay in vitro ( Figure S3). Taken together, these results provide functional evidence that AHR expression can promote melanoma growth and metastatic disease progression.

| DISCUSSION
It is well established that melanoma cells can exist in a dynamic equilibrium between more differentiated cell states associated with proliferative potential and more dedifferentiated cell states associated with invasive potential and properties of their embryonal precursors from the neural crest. [20][21][22][23] Inflammatory stimuli can shift melanoma cells towards the dedifferentiated cell state. 6,19,23 In the present study, we found that AHR expression directly correlated with the degree of dedifferentiation in human melanoma samples from the TCGA cohort as well as in a panel of human and mouse melanoma cell lines. Interestingly, we also observed a direct induction of AHR after stimulation with TNF in human and mouse melanoma cell lines. A similar induction of AHR by the inflammatory mediator lipopolysaccharide (LPS) has been previously described in a mouse sepsis model. 24 As an underlying mechanism, a direct interaction of NF-kB with the AHR promoter and following upregulation after LPS as well as a direct AHR knockout clones treated with FICZ for 24 hours (mean ± SD). C, qRT-PCR comparing the expression of selected melanocytic differentiation antigens. Treatments as indicated for 72 hours. Bars represents the mean fold change compared to the housekeeping gene Ubc ± SD (unpaired, two-sided Student's t test; ns, nonsignificant; ***P < .005). D, Immunoblots comparing the melanoma cell differentiation states between AHRcompetent, CRISPR ctrl HCmel12 cells and AHR CRISPR KO cells after 72 hours of treatment with TNF-α and/or FICZ. HCmel12 AHR CRISPR KO cells are a composite of three independent monoclones mixed in equal numbers. Data in A-D are representative for one out of three independently performed experiments. B + C were performed in three biological and three technical replicates ability of AHR to promote a phenotypic switch of cancer cells toward more dedifferentiated and stem-like phenotypes has also been reported for breast cancer and oral squamous cell carcinoma. 27,28 Mechanistically, both a direct interaction of AHR with SOX2 and an activation of JNK have been described. 29 Moreover, AHR has been shown to promote the expression of stemness-associated genes such as MYC or KLF4 in adenocarcinoma and squamous cell carcinoma cell lines. 30 We hypothesize that in our model, the AHR signaling pathway directly influences differentiation pathways and promotes cellular plasticity of melanoma cells, allowing them to dynamically alter their phenotype and thereby enhance their malignant potential.
We observed decreased local growth and metastatic dissemina- has also reported to increase tumor cell proliferation in some experimental settings including melanoma. 41 The reason for this discrepancy is currently unclear and may be explained by cell type-and cell statedependent differences in AHR signaling.
AHR has also been implicated in the resistance of melanoma cells to signal transduction inhibitors. 42 Mechanistically, a direct binding of to AHR leading to its inhibition have been described. 42,43 In addition to its direct effects on tumor cells, AHR signaling can also promote an immunosuppressive microenvironment. 44 This can be mediated by the tryptophan metabolizing enzyme indoleamine 2,3-dioxygenase 1 (IDO1) which converts the amino acid tryptophan to the metabolite kynurenine that in turn can bind to and activate the AHR. 45 Kynurenine was also shown to modulate dendritic cell functions and favor the induction of regulatory T cells via AHR signaling. 46