Tumor-derived microRNAs induce myeloid suppressor cells and predict immunotherapy resistance in melanoma

Abstract

The accrual of myeloid-derived suppressor cells (MDSCs) represents a major obstacle to effective immunotherapy in cancer patients, but the mechanisms underlying this process in the human setting remain elusive. Here, we describe a set of microRNAs (miR-146a, miR-155, miR-125b, miR-100, let-7e, miR-125a, miR-146b, miR-99b) that are associated with MDSCs and resistance to treatment with immune checkpoint inhibitors in melanoma patients. The miRs were identified by transcriptional analyses as being responsible for the conversion of monocytes into MDSCs (CD14+HLA-DRneg cells) mediated by melanoma extracellular vesicles (EVs) and were shown to recreate MDSC features upon transfection. In melanoma patients, these miRs were increased in circulating CD14+ monocytes, plasma, and tumor samples, where they correlated with the myeloid cell infiltrate. In plasma, their baseline levels clustered with the clinical efficacy of CTLA-4 or programmed cell death protein 1 (PD-1) blockade. Hence, MDSC-related miRs represent an indicator of MDSC activity in cancer patients and a potential blood marker of a poor immunotherapy outcome.

Keywords: Cancer immunotherapy; Immunology; Oncology.

Conflict of interest statement

Conflict of interest: The authors have declared that no conflict of interest exists.

Figures

Figure 1. Conversion of myeloid cells into MDSCs by melanoma EVs.

(A) HD-CD14+ cells (Mono) incubated 24 hours with melanoma EVs (Me EVs) downregulate HLA-DR (left, representative plot and summary of n = 7 HD), increase production of cytochemokines (middle), modulate HLA-DRA, IL6, and CCL2 gene transcription (right), and (B) suppress proliferation of activated CFSE-labeled T cells (percentage proliferation indicated). (C) CD14+HLA-DRneg cell frequency and HLA-DR expression on gated CD14+ cells in PBMCs of melanoma patients (Pts, n = 31) and HD (n = 15) by flow cytometry. (D) HLA-DRA downregulation of HD (n = 5) and patients’ (n = 4) monocytes cultured with melanoma EVs. (E) Induction of EV-MDSCs in CD14+ cells from a patient by autologous melanoma cell line EVs (left); suppressive activity on activated CD25+ T cells (percentages indicated, right). (F) NTA evaluation of EV size in plasma samples of patients and HD (n = 27/group) (top); correlation of EV mean size and frequency of CD14+HLA-DRneg in gated CD14+ cells of melanoma patients (bottom). (G) EV-MDSC converting potential of f1 and f2 plasma EVs from patients and HD (n = 5/group) shown as HLA-DRA downregulation in monocytes from 2 different HD; control: melanoma EVs. Data are presented as mean ± SEM. (H) Autologous (auto) plasma EVs f1 and f2 convert melanoma patient’s CD14+ cells, as shown by modulation of HLA-DRA, IL6, and CCL2 transcripts (top). EV-MDSCs generated with autologous plasma EVs f1 and f2 of melanoma patient inhibit T cell proliferation (percentages indicated, bottom). (I) Western blot of plasma EV fractions (f1, f2) of HD and patient. gMFI, geo mean fluorescence intensity; RE, relative expression. FC was by using as calibrator untreated monocytes. P < 0.001 (A, right; E); P < 0.01 (A, left); P < 0.05 (H, top), paired Student’s t test. P < 0.05, Mann-Whitney U test (D). *P < 0.05, ***P < 0.001, unpaired Student’s t test (C, F). Data are representative of 2 (E, H, I) and 3 (A, B) experiments.

Figure 2. Transcriptional regulation of EV-MDSCs at the gene level.

(A) Heatmaps of the genes regulated in EV-MDSCs (+EVs) compared with untreated monocytes (Mono, n = 5 HD) clustered according to the most representative functional categories. (B) EMT-related genes selected by a list of 839 genes from melanoma IPRES, EMT, and wound healing (5); from the literature-based EMT database (32); and from mesenchymal myeloid markers (33). (C) The FC of expression levels of selected genes in EV-MDSCs compared with untreated monocytes measured by qPCR. P < 0.01, paired Student’s t test. A representative HD of 5 tested from part A is shown. (D) GSEA plots obtained with the top 100 upregulated genes in EV-MDSCs showing a significant enrichment in data sets comparing the transcriptomic profiles of blood monocytes from colorectal, pancreatic, and breast cancer patients to those of monocytes obtained from HD. GSE47756, colon cancer (37); GSE60601, pancreatic cancer (38); GSE65517, metastatic breast cancer (39).

Figure 3. miR regulation in EV-MDSCs.

(A) Volcano plot of the miRs regulated in EV-MDSCs (n = 5 HD) compared with untreated monocytes based on microarray results, identification strategy of MDSC-miRs, and relative expression of selected miRs in EV-MDSCs compared with untreated monocytes assessed by qPCR in a representative HD. (B) Expression of MDSC-miRs in f1 and f2 plasma EVs of melanoma patients (n = 16). Box and whiskers (Tukey’s test). (C) Monocytes from HD transfected with MDSC-miR mimics (MmiRs) modulate HLA-DRA, IL6, and CCL2 gene expression compared with monocytes treated with scrambled control used as calibrator. (D) Immunosuppressive activity of mono+MmiRs on autologous activated CFSE-labeled T cells, as evaluated by CD25 expression and proliferation (percentage is indicated, left), and release of IFN-γ and TNF-α (right). (E) Loss of immunosuppressive activity of monocytes from a melanoma patient transfected with miR inhibitors (ImiRs) prior to coincubation with autologous activated CFSE-labeled T cells, as evaluated by flow cytometry (left) and cytokine release (right). (F) Expression of MDSC-miRs in CD14+ cells isolated from PBMCs of melanoma patients (n = 31) and HD (n = 15). AU, arbitrary units. P < 0.05, paired Student’s t test (A, C) and Mann-Whitney U test (F). *P < 0.05; **P < 0.01, paired Student’s t test (B, D, E). Results are representative of 5 (C), 4 (D), and 3 (E) experiments.

Figure 4. Inhibition of MDSC-miRs rescues monocytes from the acquisition of a suppressive phenotype.

(A) Effect of transfection with miR inhibitors or scrambled control (Scr) on MDSC-miR expression in INT12 melanoma cells and respective EVs (left panel). Reduced expression of miR-146a in silenced melanoma cells (Me) was confirmed by flow cytometry using an APC-fluorescent SmartFlare probe (right panel). (B) EVs derived from miR-silenced melanoma cells (+I-EVs) impaired the induction of the EV-MDSC phenotype compared with EVs derived from scrambled control cells (+S-EVs) (left panel); MDSC-miR expression was reduced in monocytes treated with I-EVs compared with those treated with S-EVs (right panel). (C) Effect of transfection with miR inhibitors or scrambled control of HD monocytes cultured in the presence of melanoma EVs, as evaluated by qPCR and flow cytometry, and (D) IL-6 and CCL2 release. (E) Loss of immunosuppressive activity of EV-MDSCs, generated from HD in the presence of miR inhibitors or scrambled control, on activated CFSE-labeled T cells, as evaluated by flow cytometry (left panel) and cytokine release (right panel), and (F) loss of immunosuppressive activity of EV-MDSCs generated from a patient with autologous melanoma–derived EVs. Percentages of CD25 and CFSE expression are indicated. Cells transfected with scrambled control were used as a calibrator (A–C). P < 0.05, paired Student’s t test (A–C); P < 0.001, 1-way ANOVA (D). *P < 0.05, **P < 0.01, paired Student’s t test (E, F). Experiments were repeated twice and performed in triplicate (A–D, F). Data are representative of 2 HD tested (E).

Figure 5. MDSC-miRs are expressed in tumor specimens and monocytes of melanoma patients.

(A) Correlation matrix of the expression levels of MDSC-miRs and CD163, CD14, CD209, ITGAM, CD33, CD68, CD3D, CD4, CD8A, PMEL, TYR, and MLANA in metastatic melanoma specimens (n = 58); PMEL, TYR, MLANA, miR-21, and miR-211 are used as unrelated controls. r values from univariate Spearman’s analysis in correlations with P < 0.05. (B) Immunostaining of CD163 infiltrate in melanoma (left panel) and correlation of the quantified CD163 signal with the expression levels of MDSC-miRs (right panel), as determined by qPCR (n = 20). (C) ISH images showing the expression of miR-146a, miR-100, miR-125b, and miR-155 in representative tumor sections. miR signals appear as brown dots localized both in CD163+ infiltrating cells and in melanoma cells. Scale bars: 100 μm. Representative images are shown in B and C.

Figure 6. MDSC-miRs are enriched in plasma from melanoma patients and associate to resistance to immunotherapy.

(A) MDSC-miR detection in the plasma of patients (Pts) and HD (n = 20/group). miR expression levels were normalized using ath-miR-159a as reference miR. RE, relative expression. *P < 0.05; **P < 0.01, Mann-Whitney U test. (B) OS of metastatic melanoma patients based on the expression levels of MDSC-miRs in plasma samples obtained at baseline of therapy, assessed by multivariable index score approach and AIM, in the global population (n = 87; left panel) and in the subsets of patients receiving immunotherapy (ICIs, n = 49) or targeted therapy (TKIs, n = 37). One patient was excluded from the latter because of receiving chemotherapy. Patients with low scores (0–1; showing 0 or only 1 increased miR) had a significantly better OS compared with patients with high scores (>1; having 2–5 increased miRs) only if receiving ICIs. Kaplan-Meier survival curves with log-rank P values are shown.