Programmed cell death-1 receptor-mediated regulation of Tbet+NK1.1− innate lymphoid cells within the tumor microenvironment

Significance This work identifies an innate lymphoid cell subset (Tbet+NK1.1−) which possesses antitumor function in both experimental murine models of cancer and in human cancers. This immune arm can be harnessed for enhancing checkpoint inhibitor therapy in patients with solid cancers. The biggest challenge in solid cancers is to identify patients who benefit from checkpoint therapies, and we are still unable to define this population. Tbet+NK1.1− innate lymphoid cells (ILCs) within cancers is the missing link that can define checkpoint therapy success in cancers such as melanoma and cutaneous squamous cell carcinoma.


Metastatic melanoma
WT or B6.TbetZsGreen or B6.Pdcd1 -/mice were inoculated with either 0.5 or 2 x 10 5 B16F10 melanoma cells in 200µl of PBS via intravenous tail-vein injection. Mice were monitored for clinical signs of weight loss and euthanized at >25% weight loss. In certain experiments mice were treated with anti-PD-1 (200µg/mouse; clone: RPMI-40) or isotype control (IgG2a, 200µg/mouse; clone: 2A3) at day 7, 9, 11. Tissue was harvested at day 12 and analyzed for the presence of ILCs unless stated otherwise. None of these experiments included exogenous IL-33 administration.

AOM-DSS induced colorectal cancer
The AOM-DSS model was set up as previously reported 3 . Briefly, mice were treated with one dose of Azoxymethane (AOM; 12mg/kg) at day 1 and then 3% dextran sodium sulphate (DSS) was added as drinking water from day 3 for a week. Mice were allowed to recover for 3 weeks and then a 2 nd cycle of DSS was started. After the 3 rd cycle, animals were euthanized and immunobiology studied. None of these experiments included exogenous IL-33 administration.

TIL isolation
ILCs from the tumor tissue were isolated as previously described 4 . Briefly, tumor tissue was incubated at 37 o C for 30minutes in FBS free DMEM media containing liberase TM TL (0.25mg/ml; Roche) and DNAse I (0.5mg/ml; Roche). Single-cell suspensions were prepared by mechanically disrupting tissue through a 100µm nylon cell strainer into FBS. Lymphocytes were isolated using lymphocyte separation media (LSM; Promocell) and washed twice with complete media (DMEM supplemented with 10% FBS, glutamine (2mM), non-essential amino acids (0.1mM), sodium pyruvate (1mM), 2-mercaptoethanol (50µM) and penicillin and streptomycin (100 units/ml penicillin; 100µg/ml streptomycin; Gibco TM ) in order to remove traces of LSM. Cells were then analyzed by flow cytometry or stimulated to induce cytokine production. Ethical approval for experiments conducted on human tissue was provided by the South Central Hampshire B NRES Committee (reference number 07/H0504/187). Fresh tissue samples of cSCC and non-lesional skin were obtained from patients during surgery at the Dermatology Department, University Hospital Southampton NHS Foundation Trust. For lymphocyte isolation, samples of tumor and separately, non-lesional skin, were finely disaggregated with scalpels, incubated at 37 o C for 1.5 hours in RPMI media containing collagenase I-A (1mg/ml; Sigma-Aldrich) and DNAse I (10µg/ml; Sigma-Aldrich). The resulting suspension was then passed through a 70µm cell strainer and centrifuged over an Optiprep (Sigma-Aldrich) density gradient. Lymphocytes were then extracted and washed with PBS before use in experiments. In certain experiments, TILs were incubated for 60 mins with 2NBDG (200µg/ml) or PBS and then ILCs analyzed by flow cytometry.

ILC isolation from spleen
Single cell-suspension of splenocytes was generated by mechanically disrupting tissue through a 40µm filter into complete media. Cells were incubated with red blood cell (RBC) lysis buffer (Biolegend) for 3 minutes at room temperature and were then washed in complete media once before analysis by flow cytometry.

ILC isolation from tumor draining lymph nodes
Tumor draining lymph nodes (TDLN) were isolated from the inguinal draining lymph nodes.
Single cell-suspension was generated by mechanically disrupting TDLNs through a 40µm filter into complete media. Cells were then washed and analyzed by flow cytometry.

ILC isolation from lungs
Lungs were perfused with PBS in situ via the pulmonary artery prior to isolation. Tissue was incubated at 37 o C for 15 minutes in FBS free DMEM media containing Liberase TM TL (0.25mg/ml; Roche) and DNAse I (0.5mg/ml; Roche). Single cell suspension was generated by mechanically disrupting tissues through a 100µm filter into FBS followed by a Percoll gradient centrifugation (40% Percoll, GE Healthcare; FBS free media containing 0.5mg/ml DNase, Roche). Cells were then washed with complete media and analyzed by flow cytometry.

ILC isolation from small intestine
Small intestine tissue was harvested into complete media. Fecal matter was removed, and tissue was washed in buffer (PBS containing 5% FBS, Hepes (1mM; Sigma), 50µM 2mercaptoethanol (50µM; Sigma) and penicillin and streptomycin (100 units/ml penicillin; 100µg/ml streptomycin; Gibco TM )). This was followed by a wash with PBS to remove traces of FBS. Tissue was incubated at 37 o C for 30 minutes in FBS free DMEM media containing Liberase TM TL (0.25mg/ml; Roche) and DNAse I (0.5mg/ml; Roche). Single cell suspension was generated by mechanically disrupting tissues through a 100µm filter into FBS. This was followed by percoll gradient centrifugation (40% Percoll (GE Healthcare), containing 0.5mg/ml DNase (Roche)). Cells were washed with complete media and analyzed by flow cytometry.

ILC isolation from skin
Mouse normal dorsal skin was harvested, finely chopped and incubated for 2 hours in FBS free DMEM media containing Liberase TM TL (0.25mg/ml; Roche) and DNAse I (0.5mg/ml; Roche) at 37 o C. Single-cell suspensions were prepared by mechanically crushing tissue slurry through a 100µm nylon mesh into FBS. Cells were then subsequently filtered through 70-and 40-µm nylon filters. Cells were then analyzed via flow cytometry.

Antibodies
All the antibodies used to characterize murine and human ILCs were purchased from BioLegend or eBioscience unless otherwise stated. For analysis of murine cell surface markers, the following antibodies were used: Lineage consisted of CD3 (clone: 1452C11), CD5 (clone:

Transwell assays
Splenocytes were RBC lysed and then plated at a concentration of 0.5 x 10 6 cells per ml.
Transwell inserts (0.4 micron; ThermoFisher Scientific) were seeded with B16F10 melanoma cells at a 1:1 ratio with splenocytes (unless otherwise stated) and were incubated for indicated time points at 37 o C prior to flow cytometry analysis. For proliferation assays, transwell inserts were removed after 6 hours. For human experiments, PBMCs were acquired from healthy donors and were cultured at a concentration of 0.5 x 10 6 per ml. Transwell inserts were seeded with either C8161 human melanoma cell line or human cSCC cell line at a 1:1 ratio with PBMCs. For human experiments, transwell inserts were removed after 16 hours. Plates were incubated at 37 o C for indicated time points and were then analyzed by flow cytometry.

Lactate Assays
WT or Pdcd1 -/splenocytes were incubated with IL-2 (100 ng/ml) plus IL-7 (100 ng/ml) alone or in combination with lactic acid (20 mM) for 24 hrs and then PD-1 expression was measured by flow cytometry. For human studies, 1x10 6 PBMCs were incubated with IL-2 (100 ng/ml) plus IL-7 (100 ng/ml) alone or in combination with lactic acid (20 mM) for 24 hrs and then PD-1 expression was measured by flow cytometry within the Lineage neg CD45 + CD127 + CRTh2 -CD117 -Tbet + subset. B16F10 tumor cells were expanded and then supernatant was tested for lactic acid production as per the manufacturer's instructions (Abcam). Briefly, 2x10 6 cells were seeded in 24 well plates and then supernatant harvested after 4hrs or 24 hrs. The amount of lactic acid was determined using a lactic acid fluorometry kit

Phospho-P70S6Kinase Measurement
Tumors were resected when they reached >600mm 3 and then TILs were isolated. TILs were stimulated with IL-2 (80ng/ml) and IL-7 (40ng/ml) for 15 minutes. TILs were washed once with PBS and then stained for phosphoP70S6Kinase antibody and then analyzed by flow cytometry. In some experiments, TILs were enriched using CD90.2 microbeads as per manufacturer's instructions and then cultured for 3 days with IL-2 plus IL-7 alone or in combination with isotype (20µg/ml) or anti-PD-1 antibody (20µg/ml). At day 3 post cultures, TILs were washed with complete media once and then restimulated with IL-2 and IL-7 for 15mins. Following stimulation, phosphorylation of P70S6Kinase was measured.

Statistical Analysis
Statistical analysis was performed with GraphPad Prism using an unpaired student two-tailed T test for groups of two and a ONE-WAY ANOVA for multiple groups. Results are expressed as mean± standard error of the mean (SEM) and P-values ≤0.05 were considered significant.
Survival curve analysis was performed using a Kaplan-Meier survival curve and a log rank test. C57BL6 WT and pdcd1 -/mice were reconstituted with B16F10 melanoma cells via subcutaneous injection. Tumor volume was measured at day 10 and day 11 A. At day 12, tumors were resected and tumor infiltrating lymphocytes were isolated and then subjected to single cell analysis. WT and PD1 Ko mice liver, bone marrow and spleen were harvested.

Supplementary Figure Legends
Tbet + NK1.1-ILCs were characterized as shown in B. Frequency of Tbet + NK1.1 -ILCs in liver C, bone marrow F and Spleen I is shown. Frequency of Eomes + Tbet + ILCs in liver D, bonemarrow G and Spleen J is shown. Frequency of Eomes -Tbet + ILCs in liver E, bonemarrow H and Spleen K is shown. Data shown are Mean+SEM of n=4 mice, statistical significance was performed using an unpaired t test. Immunobiology experiments were repeated twice.