In prior studies in an estrogen receptor negative (ER^-) model of breast cancer, we demonstrated that CDDO-Me alters immune activation in the TME and redirects TAM activation from immunosuppressive to immunostimulatory (10, 11), To determine if CDDO-Me mediates similar effects in melanoma, human peripheral blood-derived monocytes were co-cultured for 7 days with SK-MEL-28 melanoma cells harboring the BRAF^V600E mutation using Transwells (Figure 1A). This culture system allows monocytes to differentiate in vitro in the presence of melanoma cell-secreted factors. Macrophage differentiation was verified by analysis of characteristic markers of activation (Supplemental Figures 1 and 2). The addition of CDDO-Me to cell cultures does not impair viability (Supplemental Figure 6).

To test if CDDO-Me alters immune activation of melanoma-conditioned macrophages, we first tested drug effects using a co-culture system. Cultures were pre-treated on Day 6 with 300 nM CDDO-Me or vehicle control for 16 hours (11), followed by activation with LPS for an additional 24 hours (Figure 1A). As demonstrated in Figure 1C, surface expression of CD16, a marker of non-classical monocytes (18), was inhibited by CDDO-Me, while expression of CD206 and CD163, which are upregulated on TAMs and correlate with poor patient prognosis (11, 19, 20), was not altered by drug treatment.

Because BRAF^V600E mutant tumors have been shown to regulate T cell recruitment during melanoma tumorigenesis (21), we next investigated the effect of CDDO-Me using a tri-culture that incorporate T cells to simulate a more complex tumor-like microenvironment (Figure 1A). Surprisingly, the addition of autologous T cells inhibited expression of CD206, CD163, and CD16 in tri-cultured melanoma-conditioned macrophages irrespective of CDDO-Me treatment (Figures 1B–D). Collectively, these results implicate a role for T cells in the regulation of pro-tumoral myeloid markers on macrophages in this human melanoma model system.

Given results obtained in Figure 1 and prior studies demonstrating CDDO-Me-mediated changes in TAM cytokine production, we next investigated potential drug effects on CCL2 and IL-6 in co- and tri-cultures. As shown in Figures 2A, B, CDDO-Me treatment in the co-cultures significantly hampered CCL2 both at the mRNA as well as the protein levels. A similar trend was observed with IL-6 protein levels post CDDO-Me treatment. However, IL-6 mRNA levels were not significantly altered.

Strikingly, addition of autologous CD3⁺ T cells attenuated CCL2 mRNA expression (Figure 2A) and protein production of both IL-6 and CCL2 in vehicle-treated cultures as compared to the co-cultures, and this decrease in CCL2 and IL-6 levels was augmented by CDDO-Me (Figures 2B, C). These results suggest CDDO-Me attenuates immunosuppressive macrophage activation in BRAF^V600E mutant melanoma, consistent with prior observations in ER^- breast cancer (11, 22–24).

Monoculture CCL2 and IL-6 levels were measured to assess SK-MEL-28 vs. melanoma-conditioned macrophage protein production (Supplemental Figure 3). As previously reported, SK-MEL 28 produce little to undetectable levels of IL-6 and CCL2 (25), suggesting macrophages account for most of the observed IL-6 and CCL2 production.

Because we showed that CDDO-Me inhibits TAM production of CCL2, which mediates myeloid cell recruitment to the TME (26–28), we hypothesized that CDDO-Me treatment would impair monocyte migration. We assessed the ability of CDDO-Me cultured media from co- and tri-cultures to impede migration. As expected, monocyte migration in co-cultured CDDO-Me media was significantly hindered (Figure 3). However, monocyte migration with tri-cultured media was hampered irrespective of CDDO-Me administration (Figure 3), consistent with results obtained in Figure 2. These results suggest CDDO-Me inhibits monocyte migration and this effect is enhanced by T cells.

To determine whether T cell effects on macrophage activation were mediated by soluble factors or if there were a requirement for direct contact between the cells, we performed Transwell co-culture studies. Monocytes were differentiated in conditioned media from SK-MEL-28 cells and incubated with autologous CD3⁺ T cells separated or not by Transwells (Figure 4A). As demonstrated in Figures 4B, D, production of CCL2 and VEGF was inhibited in a contact-independent manner in both vehicle and CDDO-Me-treated cultures. In contrast, direct T cell/macrophage contact was required for reduced IL-6 levels in vehicle-treated cultures, although this requirement was abrogated following CDDO-Me treatment (Figure 4C). These data suggest T cell-derived secreted factor(s) regulate macrophage cytokine production post-CDDO-Me treatment.

While TAMs are known to suppress T cell activation in the TME (29–31), the effect of activated T cells on TAM activation is poorly understood. Because the addition of T cells augmented CDDO-Me-mediated inhibition of macrophage production of IL-6 and CCL2, (Figure 2), we investigated the dependence of this synergy on T cell subset. We hypothesized that CD4⁺ T cells would be required for synergy with CDDO-Me, as CD4⁺ Th1 cells have been implicated in TAM redirection in B16 mouse melanoma models (7). Surprisingly, tri-cultures containing either CD4⁺ or CD8⁺ T cells were equally effective at suppressing production of CCL2 and IL-6 (Figures 4E, G) post-CDDO-Me treatment. Because TAMs sense hypoxia in avascular areas of tumors and react by producing pro-angiogenic factors such as VEGF-A (23, 24, 32), we interrogated the effect of CDDO-Me on VEGF-A production. As observed with CCL2 and IL-6, VEGF production was also hampered by both CD4⁺ and CD8⁺ T cell subsets post- CDDO-Me treatment (Figure 4F). Surface expression of CD206, CD163 and CD16 was not altered by either subset (Supplemental Figure 4). These results suggest CD4 and CD8 T cells inhibit immunosuppressive cytokine production in melanoma-conditioned macrophages, and that modulation of macrophage surface markers may require both T cell subsets, consistent with results in Figure 1.

CDDO-Me is a multifunctional drug that has been shown to inhibit activation of several signaling pathways important for cancer progression and metastasis, including MAPK, NF-κB, and STAT3 (10, 12, 33–37). Because STAT3 activation is associated with tumor growth, reduction of T cell infiltration, and maintenance of TAM immunosuppression in the TME, we hypothesized that one possible mechanism by which CDDO-Me attenuates TAM pro-tumoral activation is through inhibiting phosphorylation of STAT3 (11, 37–39). As a positive control for STAT3 phosphorylation, monocytes were differentiated with M-CSF alone in the absence of other cell types followed by stimulation with LPS. As demonstrated in Figures 5A, B, CDDO-Me reduced phosphorylation of STAT3 in all culture conditions, while total STAT3 levels were unaltered by CDDO-Me treatment. Full length immunoblots are provided in Supplemental Figures 5A–C. These results indicate that CDDO-Me inhibits STAT3 phosphorylation in human melanoma-conditioned macrophages and suggest CDDO-Me-mediated changes in myeloid activation may be due, at least in part, to suppression of STAT3 activation.

PBMCs were obtained by leukapheresis of healthy donors following written informed consent. Approval for this study was provided by the Institutional Review Board (IRB) of the Geisel School of Medicine, the Committee for the Protection of Human Subjects (protocol # 17011). This study was conducted in accordance with the human experimentation guidelines established by the Geisel School of Medicine Committee for the Protection of Human Subjects. Mononuclear cells were separated on Ficoll-Paque Premium (density: 1.077, GE Healthcare) and enriched for monocytes using cold aggregation. Monocyte purity was assessed at ≥95% using cytospin, Wright-Giemsa staining and flow cytometric analysis of CD14 (BioLegend, Cat.# 325618) expression.

Isolated CD14⁺ monocytes were differentiated with 20 ng/ml M-CSF in complete HEPES-buffered RPMI 1640 media supplemented with 10% FBS for 7 days. Macrophage polarization was verified using flow cytometry to measure expression of cell surface markers CD206 (clone: 15–2, Cat. #321103), CD163 (clone: GHI/61, Cat. #333606), and HLA-DR (clone: L243, Cat. #307618) and by qRT-PCR and ELISA analysis of secreted cytokines post-LPS stimulation.

Human monocytes were cultured in RPMI-1640 (Corning, Cat. #10-040-CV) supplemented with 10% FBS (HyClone, Cat. #SH30070.03), 0.25 M HEPES (Fisher Scientific, Cat. #BP299-100), and 1% penicillin streptomycin (Thermofisher, Cat. #15140122) and plated at 8 x 10⁵-1.2 x 10⁶/ml (tri- or co-culture, respectively) on day 1 in the top chamber of collagen-coated 0.4 μm PTFE membrane Transwells (Corning, Cat. #3491). Human malignant melanoma SK-MEL-28 cells carrying the BRAF^V600E mutation were cultured in E-MEM supplemented with 10% FBS and 1% penicillin streptomycin in the bottom chamber of Transwells at 2.5 x 10⁴/ml on day 0. For tri-culture studies, autologous T cells were activated for 2 days prior to addition to tri-cultures with anti-CD3/28/2 (Stemcell Technologies, Cat. #10990) from monocyte-depleted PBMCs in RPMI-1640 complete media supplemented with 10% FBS, 1% HEPES, 1% sodium pyruvate (HyClone, Cat. #SH30239.01), non-essential amino acids (HyClone, Cat. #SH30238.01), 0.055 mM BME (Gibco, Cat. #21985-023) and 1% penicillin streptomycin. T cells (2.5 x 10⁵/ml) were added to the top chamber of Transwells with monocytes. To compare the influence of CD4⁺ vs. CD8⁺ T cell subsets on tri-cultures, total CD3⁺ T cells were sorted using human CD4⁺ and CD8⁺ MicroBeads (Miltenyi Biotec, Cat. #130-045-101 and 130-045-201). For LPS activation studies, cells were pretreated with 300 nM CDDO-Me or DMSO vehicle control for 16 hours. After cultures were washed to remove drug, cells were stimulated with 10 ng/ml LPS (Sigma-Aldrich, Cat. #L4391-1MG) for an additional 24 hours. Graphic representations of culture assay systems are depicted in Figures 1A and 4A.

Total RNA was isolated using the quick RNA microprep kit (Zymo, Cat. #11-328M) per manufacturer’s instructions. Complementary DNA (cDNA) was synthesized from 100 ng total RNA and random hexamers using the qScript™ XLT cDNA SuperMix (QuantaBio, Cat. #955161-100). Quantitative real time PCR (qRT-PCR) was performed using TaqMan Probe single tube assays (Life Technologies, Cat. #4324018) for human CCL2 (Applied Biosystems, Cat. # Hs00234140), IL-6 (Cat. #Hs00985639_m1) and VEGF (Cat. # Hs00900055_m1) genes. The StepOnePlus Real-Time PCR System (Applied Biosystems) was used for amplification and detection. Threshold cycle number was determined using Opticon software. mRNA levels were normalized to β-actin, which control studies showed is not altered by CDDO-Me treatment (11), using the equation 2^-(Et-Rt), where Rt is the mean cycle threshold for the control gene and Et is the mean threshold for the experimental gene. Thermal cycling conditions for qRT-PCR consisted of an initial incubation at 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. Product accumulation was measured during the extension phase and all samples were run in triplicate.

Secreted protein expression in cell-culture supernatants was quantified by ELISA for CCL2 (Invitrogen, Cat. # 88-7391-88), IL-6 (Invitrogen, Cat. #88-7066-88), and VEGF (R&D Systems, Cat. #DY293B-05) according to manufacturers’ protocols.

All fluorophore-conjugated antibodies were obtained from BioLegend: anti-CD206-FITC (Cat. #321103), anti-CD163-PE(Cat. #333606), anti-CD64-PE/Cy7(Cat. # 305022), anti-CD80-APC (Cat. # 305220), anti-HLA-DR-APC/Cy7 (Cat. # 307618), anti-CD16-BV421 (Cat. # 360723), anti-CD1a-PerCP(Cat. # 300130), anti-CD45-APC/Cy7 (Cat. # 368516), anti-CD4-APC (Cat. # 300552), anti-CD8-BV510 (Cat. # 301048), anti-CD3-BV42 (Cat. # 344834). Cells were stained for 1 hour at 4°C with 2 mg/ml Globulins Cohn fraction II, III (Sigma) to inhibit non-specific antibody binding. In all conditions, doublets and multiplets were excluded by forward scatter pulse width (SSC-W) vs. forward scatter pulse area (SSC-A) gating. Gating of positively stained cells was determined by fluorescence-minus-one (FMO) controls. Cells were analyzed using an 8-color MACSQuant 10 (Miltenyi Biotec) with three laser sources (405 nm, 488 nm, 635 nm) and FlowJo 9.8.1 (Treestar).

Monocytes were differentiated with 20 ng/ml M-CSF alone or in co-culture with SK-MEL-28 cells or tri-culture with SK-MEL-28 cells and T cells as indicated, followed by treatment with 300 nM CDDO-Me or DMSO vehicle control for 16 hours as above. Cultures were then washed with media to remove drug and activated with 10 ng/ml LPS (Sigma-Aldrich, Cat. #L4391-1MG) for an additional 24 hours. Whole cell lysates were prepared using PRO-PREP™ Protein Extraction Solution (iNtRON Biotechnology, Cat. #17081). Lysates were analyzed for total protein concentration using a Micro BCA assay kit (Thermofisher Scientific, Cat. #23235). Four micrograms of each lysate were separated on Mini-PROTEAN TGS precast protein gels 4-15% (Bio-Rad, Cat. #45161084) and electrotransferred to nitrocellulose membrane (GE Healthcare, Cat. #10600012) in Tris-glycine buffer with 20% methanol. Membranes were blocked in 5% milk in 1xTBS and 0.05% Tween-20 (Santa Cruz, Cat. #SC-281695) for 1hr at RT. Blots were probed with primary detection antibodies for pSTAT3 (Cell Signaling, Cat. #4113S), STAT3 (Cell Signaling, Cat. #4904S) and GAPDH (Abclonal, Cat. #AC002). Primary antibodies for pSTAT3 and STAT3 were incubated at a 1:2000 dilution in Tris-buffered saline with 0.05% Tween 20 (Santa Cruz) and 5% BSA (Fischer BioReagents, Cat. #BP1600100) for 1hr at 4°C with rotation. Primary antibody for GAPDH was incubated at a 1:10000 dilution in Tris-buffered saline with 0.05% Tween 20 and 5% milk (Nestle Carnation) for 1hr at 4°C with rotation. Following incubation with primary antibodies, blots were incubated with horseradish peroxidase-conjugated secondary antibodies (goat anti-mouse or goat anti-rabbit) (Bio-Rad, Cat. #170-6516 or #170-6515) at a 1:2,000 dilution in Tris-buffered saline with Tween 20 (TBST) and 5% milk for 1hr at room temperature with gentle agitation on a rocker, followed by thorough washing in TBST. Blots for STAT3 and GAPDH were visualized using Pierce™ ECL Western Blotting Substrate (Thermofisher Scientific). Blots for pSTAT3 were visualized using SuperSignal™ West Femto Maximum Sensitivity Substrate (Thermo Scientific). To control for loading errors, total STAT3 and pSTAT3 proteins were normalized to their respective GAPDH signals. Using these normalized values, the ratio of pSTAT3 to total STAT3 was plotted for each co-culture and tri-culture condition.

Migration assays were performed using freshly-isolated monocytes in a modified Boyden chamber-based system. HTS Transwells, 96-well (5.0 μm pore, Corning, Cat. #3388), were used to assay chemotaxis. 150μl of tri-culture supernatant or media with and without CCL2 (Peprotech, Cat. #300-04-5ug) control (1 ng/ml) were transferred to the 96-well feeder tray (bottom). A cell migration chamber plate with 5.0 μm pores was placed over the 96-well feeder tray. A volume of 100μl containing 200,000 fresh monocytes cells/well resuspended in complete RPMI was transferred to each inner well of the cell migration chamber plate. The 96-well migration plate was placed in a CO₂ incubator at 37°C for 8 hours. At the conclusion of the incubation, migrated cells were lysed and dyed with CyQuant (Invitrogen, Cat. #C7026) GR, then read on a fluorescence plate reader at Excitation 480nm/Emission 520nm.

Figures are representative of at least three independent experiments as indicated in Figure Legends. All experiments were repeated at least three times, unless otherwise noted, and at least 2 technical replicates of each analyte were included in each assay. Results are described as mean ± SD and were analyzed by unpaired student’s t-Test or two-way ANOVA for multiple comparison (as indicated in Figure Legends) using GraphPad Prism 8. Significance was achieved at p < 0.05.