The NLRP3 inflammasome – interleukin 1β axis in uveal melanoma

Uveal melanoma (UM) is the most common primary intraocular cancer in the adult population. Recent studies suggested that the NLRP3 inflammasome could be a therapeutic target for cutaneous melanoma (CM), but the role of NLRP3 in UM remains unknown. Here, we analyzed the NLRP3‐IL‐1β axis in 5 UM and 4 CM cell lines. Expression of NLRP3 mRNA in UM and CM was low, and expression in UM was lower than in CM (P < 0.001). NLRP3 protein levels were below detection limit for all cell lines. UM exhibited lower baseline IL‐1β secretion than CM, especially when compared to the Hs294t cell line (P < 0.05). Bioinformatic analysis of human tumor samples showed that UM has significantly lower expression of NLRP3 and IL‐1β compared with CM. In conclusion, our work shows evidence of extremely low NLRP3 expression and IL‐1β secretion by melanoma cells and highlight differences between CM and UM.

Uveal melanoma (UM) is the most common primary intraocular cancer in the adult population. Recent studies suggested that the NLRP3 inflammasome could be a therapeutic target for cutaneous melanoma (CM), but the role of NLRP3 in UM remains unknown. Here, we analyzed the NLRP3-IL-1b axis in 5 UM and 4 CM cell lines. Expression of NLRP3 mRNA in UM and CM was low, and expression in UM was lower than in CM (P < 0.001). NLRP3 protein levels were below detection limit for all cell lines. UM exhibited lower baseline IL-1b secretion than CM, especially when compared to the Hs294t cell line (P < 0.05). Bioinformatic analysis of human tumor samples showed that UM has significantly lower expression of NLRP3 and IL-1b compared with CM. In conclusion, our work shows evidence of extremely low NLRP3 expression and IL-1b secretion by melanoma cells and highlight differences between CM and UM.
Uveal melanoma (UM) is the most common primary intraocular cancer in the adult population, with an incidence of around 7 per million per year [1,2]. It can arise from melanocytes of any part of the uveal tract, but affects the posterior choroid in more than 90% of cases [3]. Uveal melanoma is slowly growing tumor, and less than 2% of patients have detectable systemic metastasis at presentation [4]. Despite effective primary treatment, more than half of the patients will develop metastatic disease [5,6], and once a metastasis is diagnosed, the median survival drops to 6-12 months [5,[7][8][9]. Great progress has been made in the past decade for advanced cutaneous melanoma (CM) with immunotherapy, such as immune checkpoint inhibitors [10,11], reducing the overall mortality rate by 18% and doubling the 5-year survival rate for metastatic CM from 15% to 30% between 2011 and 2017 [12,13]. On the contrary, such therapies have not been able to help patients with advanced UM [9,[14][15][16][17][18][19][20][21], for which the median survival after metastatic spread has remained largely unchanged over the past 40 years [7,22]. This highlights the intrinsic differences between these tumors and the need to better understand UM immune microenvironment and explore new treatment strategies for UM.
Recently, nucleotide-binding, leucine-rich repeat and pyrin domain-containing protein 3 (NLRP3) inflammasome has been suggested as a new therapeutic target for CM [23,24]. NLRP3 inflammasome is a multimeric intracellular protein complex involved in the innate immune response and is assembled and activated following damage-associated molecular patterns (DAMPs) or pathogen-associated molecular patterns (PAMPs) signaling, which promotes caspase-1 activation and the release of the pro-inflammatory interleukin-1b (IL-1b) and interleukin-18 (IL-18), and ultimately leads to cell death through pyroptosis [25]. Tengesdal et al. [23,24] showed that CM cell lines had active NLRP3-IL1b inflammasome axis aiding tumorinduced immunosuppression, and inhibition of NLRP3 could suppress melanoma growth in vivo in animal models. However, it is not known if this axis is present in UM as well. Here, we analyzed the baseline NLRP3-IL-1b axis in 5 different UM cell lines and contrasted them to four CM cell lines. We found that UM cells had extremely low levels of NLRP3 and IL1b. Furthermore this was confirmed in bioinformatic analysis of human samples, suggesting that in contrast to the work in CM, this may not be a promising target for future therapies of UM.

Cell culture
Human primary uveal melanoma cell lines (92.1, Mel270, and Mel285), human uveal melanoma liver metastasis cell lines (OMM2.3 and MM28), primary cutaneous melanoma cell lines (OCM3 and A375), and cutaneous melanoma lymph node metastasis cell lines (Hs294t and Sk-Mel28) were studied. Cell lines mutational profile and original references are summarized on Table S1

ELISA
Cells were seeded in 12-well plate at 10 5 cells per well and cultured for 48 h until confluency (#665180, Greiner Bio-One, Monroe, NC, USA). Media were then aspirated and 500 lL of fresh media was added, and cells were incubated for another 24 h. Supernatant was used for human IL-1b quantification using Quantikine ELISA kit (#DLB50, R&D Systems, Minneapolis, MN, USA), according to the manufacturer's protocol. Plate was read with SpectraMax 190 absorbance microplate reader (Molecular Devices, San Jose, CA, USA) using the SOFTMAX PRO software v5.4.1 (Molecular Devices). IL-1b quantification was measured five times per cell line.

Western blotting (WB)
Cells were seeded in 6-well plate (#657160, Greiner Bio-One) at 2.5 9 10 5 cells per well and cultured for 48 h until confluency with 2 mL of media per well. After reaching confluency, wells were washed with serum-free media and then stored overnight at À80°C. THP-1-derived macrophage was obtained by incubating THP-1 cells (5 9 10 5 cell per well in 6-well plate) overnight with 80 nM of phorbol 12-myristate 13-acetate (#P1585, Millipore Sigma) and

Quantitative polymerase chain reaction (qPCR)
Cells were seeded in 6-well plate (#657160, Greiner Bio-One) at 2.5 9 10 5 cells per well and cultured for 48 h until

Statistical analysis
Statistical analyses were performed by PRISM software 5.0 (GraphPad, San Diego, CA, USA). Results were described as mean AE SEM. ANOVA followed by post hoc Tukey HSD test was used to compare multiple groups. Statistical significance was achieved when P < 0.05.

Results
Uveal melanoma has lower NLRP3 expression compared with cutaneous melanoma qPCR analysis showed that NLRP3 gene expression by the five UM cell lines was significantly lower in comparison with CM cell lines (P < 0.001; Fig. 1A). Further, both uveal and cutaneous melanoma showed very low NLRP3 mRNA levels when compared to the monocyte cell line THP-1 (P < 0.001; Fig. 1B). Overall protein levels of NLRP3 by WB using a previously validated specific and very sensitive antibody [29], failed to detect any signal even after prolonged exposure (10 min) on any melanoma cell line, suggesting that the NLRP3 protein levels in melanoma, if present, are extremely low (Fig. 1C-G).

Uveal melanoma has minimal baseline secretion of IL-1b
We proceeded to investigate the expression levels of IL-1b by WB and ELISA using THP-1-derived macrophage as positive control (Fig. 2). WB with equal protein loading (Fig. 2F) showed no IL-1b signal on any cell line at 204 s (Fig. 2B,C), and after prolonged exposure, only a faint band on OCM3 cell line was revealed (Fig. 2D,E). ELISA detected low levels of IL-1b in all cell lines ( Fig. 2A) levels of NLRP3 and IL-1b compared with CM (Fig. 3A,B). Additionally, we explored if the NLRP3-IL-1b axis is activated on normal choroidal melanocytes, since these are the cells that originate UM. Transcriptomic analysis of RPE/choroid cells acquired from the Spectacle single-cell atlas [27] reveals that choroidal melanocytes do not express NLRP3 or IL-1b (Fig. 3C). Reclustering of choroidal melanocytes further demonstrates that the vast majority of cells do not express NLRP3 or IL1b (Fig. 3D,E).

Discussion
During the last decade, the role of inflammasomes in tumorigenesis, antitumor immunity, and response to cancer therapy has been investigated in many cancer models, with conflicting findings [25,[39][40][41]. While inflammasome-associated IL-1b secretion by tumorassociated macrophages and cancer-associated fibroblasts may promote tumor growth and metastasis by suppressing antitumor immunity [42][43][44], inflammasome activation also enhances antitumor activity of CD8 + T cells and NK cells through IL-18, promotes tumor infiltration of lymphocytes, and strengthens the response to checkpoint inhibitors [45][46][47][48]. NLRP3 inflammasome was investigated specifically in CM by other authors as well, with contrasting results. Lee et al. [49] demonstrated that NLRP3 inflammasome knockout or pharmacological inhibition reduced the ability of macrophages to promote migration and invasion of CM cells. Also, an increased expression of NLRP3 inflammasome-related genes in CM tumor samples was associated with low or no response to immune checkpoint therapy [50]. Conversely, a high expression of NLRP3 inflammasomerelated genes was associated with increased CM overall survival [50,51] and higher tumor immune infiltration of B cells, CD8 + T cells, and neutrophils [51]. In addition, NLRP3 inflammasome activation by nigericin induced pyroptotic cell death on BRAF V600 inhibitor-resistant CM tumor cells [52].
To date, there have been no reports of the NLRP3 inflammasome-IL-1b axis in UM. In this setting, we evaluated the baseline levels of IL-1b secreted by five different UM cell lines and four different CM cell lines. Nonstimulated secretion was overall low, on the picogram scale ( Fig. 2A). In addition, UM cells showed significantly lower levels of IL-1b secretion compared with the CM cell line Hs294t (P < 0.05), which might indicate an even lower potential of UM to recruit immune cells to the tumor site [23,24,53]. The levels reported in our study for the CM cell lines we used were comparable to other published literature (Table 1) and in accordance with Gehrke et al. [31] that showed no IL-1b secretion by CM cell lines. Most skin melanoma cell lines express low levels of IL-1b, with only few exceptions, particularly the cell line 1205Lu. It is worth mentioning that this specific cell line is potentially contaminated and might not reflect the original tumor characteristics [54,55]. These data are in accordance with a recently published paper that demonstrated through single-cell RNA-seq that myeloid cells, and not tumor cells, are the primary source of IL-1b in solid cancers [56].
Since we have investigated both primary and metastatic cell lines (Table S1), these data suggest that there is no meaningful endogenous IL-1b secretion by melanoma cell lines regardless of origin. Collectively, our findings indicate that melanoma-derived IL-1b might represent only a small fraction of the IL-1b within the tumor microenvironment and targeting NLRP3 would affect primarily immune rather than tumor cells. In addition, we demonstrate that UM cell lines have even lower NLRP3 gene expression and IL-1b secretion than CM cell lines, which might help explain, at least partially, the differences in UM response to immune checkpoint inhibitors [9,[14][15][16][17][18][19][20][21].
To further explore the topic, we conducted a bioinformatic analysis of human tumor samples from the TCGA database. The TCGA is a public available database of cancer multiomics data with over 20 000 cases of primary cancer and matched normal samples. UM samples had significatively lower expression levels of both NLRP3 and IL-1b compared with CM samples. Additionally, we investigated the gene expression of normal choroidal melanocytes through the Spectacle single-cell atlas [27] based on a dataset with 37 070 cells including 2939 melanocytes [28] and showed that these cells that originate UM also do not express NLRP3 or IL1B.
A limitation of our study is the sole use of in vitro and bioinformatic approaches. Although the use of animal models could provide data on the interaction between tumor and immune cells, the primary objective of this work was to determine the endogenous baseline NLRP3 inflammasome levels and IL-1b secretion of CM and UM. Another limitation of our study is that we only investigated established cell lines, which could differ from the original tumor cells. On the contrary, the use of nine different cell lines with concordant observations and the validation of our findings with human tumor samples data give us confidence in our results.
To conclude, our work shows evidence of very low NLRP3 expression and IL-1b secretion by melanoma cells and highlights the differences between cutaneous and uveal melanoma, providing data that may contribute to future melanoma translational research.

Conflict of interest
The authors declare no conflict of interest.
Author contributions VSMCC, NEE, and DGV conceived the paper. VSMCC and NEE designed the experiments. VSMCC, NEE, ZY, DPN, and TN conducted the experiments. VSMCC and NEE performed data analysis. VSMCC drafted the paper. EG and IKK contributed to the review and editing. DGV supervised the project. All authors contributed to the edits and approval of the final paper.

Data accessibility
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.