TUMOR-INFILTRATING NOCICEPTOR NEURONS PROMOTE IMMUNOSUPPRESSION [MDSC]

Nociceptor neurons impact tumor immunity. Removing nociceptor neurons reduced myeloid-derived suppressor cell (MDSCs) tumor infiltration in mouse models of head and neck carcinoma and melanoma. Carcinoma-released small extracellular vesicles (sEVs) attract nociceptive nerves to tumors. sEV-deficient tumors fail to develop in mice lacking nociceptor neurons. Exposure of dorsal root ganglia (DRG) neurons to cancer sEVs elevated expression of Substance P, IL-6 and injury-related neuronal markers while treatment with cancer sEVs and cytotoxic CD8 T-cells induced an immunosuppressive state (increased exhaustion ligands and cytokines). Cancer patient sEVs enhanced DRG responses to capsaicin, indicating increased nociceptor sensitivity. Conditioned media from DRG and cancer cell co-cultures promoted expression of MDSC markers in primary bone marrow cells while DRG conditioned media together with cancer sEVs induced checkpoint expression on T-cells. Our findings indicate that nociceptor neurons facilitate CD8+ T cell exhaustion and enhance MDSC infiltration. Targeting nociceptor-released IL-6 emerges as a novel strategy to disrupt harmful neuro-immune interactions in cancer and enhance anti-tumor immunity. Overall design: Eight-week-old female TRPV1Cre::DTAfl/wt mice, in which nociceptor neurons are genetically ablated (n = 4), and their TRPV1Cre::DTAwt/wt littermate controls (n = 3) each received 2 × 10^5 B16F10-OVA-mCherry melanoma cells suspended in 50 µL PBS, injected subcutaneously into the right flank. Fourteen days later the mice were euthanized and the tumors were harvested. The tumors were finely minced and digested for 40 min at 37 °C in DMEM supplemented with 5 % FBS, collagenase D (2 mg mL?¹), collagenase IV (1 mg mL?¹) and DNase I (40 µg mL?¹). The resulting suspension was purified on a 70 %/40 % Percoll gradient, passed through a 70-µm strainer and resuspended in PBS containing 2 % FBS and 2 mM EDTA, while lymph nodes were processed in parallel by mechanical dissociation and filtration. Single-cell suspensions were stained on ice and sorted on a BD FACSAria II. Non-viable cells were excluded with Zombie Aqua (BioLegend 423102), and Fc receptors were blocked with anti-CD16/32 (BD 553141, 0.5 mg mL?¹, 15 min, 4 °C). For lymphoid analyses, cells were labeled with CD45-BV421 (clone 30-F11, 1:100) and CD3e-FITC (clone 145-2C11, 1:100); for myeloid analyses, CD45-BV421 was combined with CD3e-PerCP-Cy5.5 (clone 145-2C11, 1:100). Intrinsic mCherry fluorescence identified melanoma cells. Tumor-infiltrating lymphocytes were further resolved with CD4-PerCP-Cy5.5 (clone GK1.5, 1:100), CD8a-Alexa Fluor 700 (clone 53-6.7, 1:100) and CD25-PE (clone PC61.5, 1:100), defining CD8? T cells as Zombie? CD45? CD3? CD8a? CD4? CD25? and regulatory T cells as Zombie? CD45? CD3? CD4? CD25^hi. Myeloid subsets were distinguished with CD11b-APC-Cy7 (clone M1/70, 1:100), Gr-1-APC (clone RB6-8C5, 1:100), CD64-PE-Cy7 (clone X54-5/7.1, 1:100), MHC II-FITC (clone M5/114.15.2, 1:100) and CD11c-PE (clone N418, 1:100), gating myeloid-derived suppressor cells as Zombie? CD45? CD3? CD11b? Gr-1? MHC II? and antigen-presenting cells as Zombie? CD45? CD3? CD11c? MHC II? CD11b? Gr-1?. Melanoma cells were sorted as Zombie? mCherry? CD45? and stromal cells as Zombie? mCherry? CD45?. Fluorescence-minus-one controls set all gates, and doublets were excluded by FSC-A/FSC-H and SSC-A/SSC-H discrimination. Six purified populations—CD8? T cells, regulatory T cells, myeloid-derived suppressor cells, antigen-presenting cells, mCherry? melanoma cells and mCherry?/CD45? stromal cells—yielded 1.5 × 10^3 to 1.8 × 10^5 cells per library. Sorted cells were collected directly into 500 µL TRIzol reagent (Invitrogen 15596026) and stored at -80 °C until RNA extraction according to established protocols. RNA quality assessed on an Agilent Bioanalyzer yielded RNA Integrity Numbers of 8.2–10. Poly-A-selected libraries were prepared with NEBNext Ultra II; a total of 48 libraries (eight per population) were constructed at Fulgent Genetics and sequenced on an Illumina HiSeq 4000 in 75-cycle single-end mode (project IEM-048-02, 10 May 2020), generating approximately 20 million reads per sample. Base calling used Illumina RTA 2.4.11, and demultiplexing was performed with bcl2fastq 2.20 (one index mismatch allowed). Adapter trimming and quality filtering were carried out with Trimmomatic, and high-quality reads were aligned to the GRCm38 mouse genome with STAR v2.5.1, which also generated gene-level counts. Differential expression analysis in DESeq2 used normalized counts; genes with an adjusted P value (false-discovery rate) < 0.05 were considered differentially expressed. Log2 fold changes and -log10 P values were calculated from the normalized data, and all downstream analyses and visualizations were performed in RStudio.