Dysfunction of exhausted T cells is enforced by MCT11-mediated lactate metabolism

Tumor cell lines

We obtained B16-F10 (CRL-6475) and A549 (CCL-185) cells from the American Type Culture Collection. We obtained MC38 (available at Kerafast) cells from D.A.A. Vignali (University of Pittsburgh) and MEER cells from R. Ferris (University of North Carolina). The B16, A549 and MC38 cell lines were authenticated via sequencing by their supplier. MEER cells were generated by overexpressing E6/E7 and Ras in primary mouse tonsil epithelial cells50 and were probed by western blot to confirm E6/E7 and Ras overexpression. MC38 and MEER were confirmed mycoplasma free in 2016 and B16 in 2018. Primary and immortalized cell lines were maintained in laboratory-made R10 medium (RPMI 1640, 10% fetal bovine serum, 2 mM ʟ-glutamine, penicillin plus streptomycin, nonessential amino acids, 1 mM sodium pyruvate, 5 mM HEPES buffer and β-mercaptoethanol). The cultures were incubated in temperature-stable and partial-pressure-stable conditions at 37 °C and 5% CO2.

Mice

The work with mice was done in accordance with the Institutional Animal Care and Use Committee at the University of Pittsburgh. The mice were housed on a 12 h light–dark cycle in boxes (either four males or five females per box), in specific pathogen-free conditions. All experimental mice were fed with Purina Prolab Isopro RMH 3000 (5P75 and 5P76) chow ad libitum. Male and female mice were used on a C57BL/6 background between the ages of 8 and 12 weeks. Slc16a11f/f mice were generated by S. Gingras (University of Pittsburgh). NSG, C57BL/6, SJ/L (Thy1.1), CMVCre, Cd4cre and Tg(TcraTcrb)1100Mjb/J (OT-I) mice were obtained from the Jackson Laboratory.

Tumor growth curves and therapies

Slc16a11f/f or Slc16a11f/fxCD4Cre mice were injected with B16 100,000 cells in RPMI medium and treated with 200 μg per dose of αPD1 (Bio X Cell; clone RMP1-14; catalog number BE0146) for three total doses. For αMCT11-treated tumor growth curves, C57BL/6 or Rag2-knockout mice were injected with 250,000 cells of either B16, MC38 or MEER in serum-free RPMI. Starting when tumors were palpable (day 5–7) mice were treated with 200 μg per dose of mIgG2a isotype control (Bio X Cell), αPD1 (Bio X Cell; clone J43; catalog number BE0033-2), αMCT11 or LALAPG Fc mut αMCT11, which were administered every other day for a total of five doses. Anti-MCT11 was originally raised in mice against a human N-terminal peptide, and the monoclonal parental or LALAPG anti-MCT11 used in this study was generated recombinantly in CHO cells (Evitria). The tumors were measured three times per week using digital calipers. The maximal tumor size was reached when a tumor grew to 15 mm in size in any direction, at which point the tumor-bearing mouse was killed. The mice were excluded from analysis when tumors became ulcerated before the tumor reached maximal size.

Tumor and lymph node collection and mechanical disruption

For T cell single-cell suspension, LNs and spleens of mice were mechanically disrupted with the back end of a syringe plunger and filtered through 70 μm filters (Fisher brand). For tumor single-cell suspensions, whole tumors were injected with 2 mg ml−1 of collagenase type IV, 2 U ml−1 of dispase and 10 U ml−1 of DNAse I (Sigma) in buffered RPMI and incubated for 20 min at 37 °C. The tumors were then mechanically disrupted using the back end of a syringe plunger and filtered through 70 μm filters (Fisher brand).

LCMV infections

C57BL/6 mice were inoculated with LCMV clone 13 (2 × 106 plaque forming units (PFU) by retro-orbital injection). The mice were monitored for weight loss to determine whether they were infected. The mice were sacrificed on day 14 post infection and their LNs, spleens, liver, kidney, lung and BM were collected. The organs were mechanically disrupted as described above to reach a single-cell suspension. Antigen specific Tex cells were identified using a gp33+ tetramer gifted by Larry Kane (University of Pittsburgh).

CD8+ T cell negative selection

CD8+ T cells were purified by negative selection from LNs and tumor single-cell suspensions. These were performed using Mojosort magnetic-beads (BioLegend) and the following biotinylated antibodies (BioLegend): CD4 (RM4-5, catalog number 100508, lot number B286276, dilution 1:1,000), CD19 (6D5, catalog number 115504, lot number B353713, dilution 1:1,000), CD11c (N418, catalog number 117304, lot number B317309, dilution 1:1,000), CD11b (M1/70, catalog number 101204, lot number B307868, dilution 1:1,000), Ly6G/Ly6C (Gr-1, catalog number 108404, lot number B351067, dilution 1:1,000), TCRγδ (GL3, catalog number 118103, lot number B355058, dilution 1:1,000), B220 (RA3-6B2, catalog number 103204, lot number B352779, dilution 1:500), CD49 (DX5, catalog number 108904, lot number B285502, dilution 1:500), CD105 (MJ7/18, catalog number 120404, lot number B266720, dilution 1:500), CD24 (M1/69, catalog number 101803, lot number B360781, dilution 1:500) and CD16/32 (93, catalog number 101303, lot number B355428, dilution 1:500).

Hypoxia detection with pimonidazole

For experiments using Hypoxyprobe, the mice were retro-orbitally injected with pimonidazole (80 mg kg−1, Hypoxyprobe) in PBS 1 h before they were killed. Pimonidazole was detected using antipimonidazole antibodies (Hypoxyprobe) after 10 min of 4% paraformaldehyde (PFA) fixation, followed by Foxp3 Fix/Perm permeabilization for 20 min.

Flow sorting and cytometry

For extracellular stains, the samples were incubated on ice for 20 min in an antibody cocktail mix. For intracellular stains, the samples were incubated on ice for 20 min in an antibody cocktail mix after fixation. The antibodies were obtained from the following companies: BioLegend: anti-CD4 (GK1.5, catalog number 100412, lot number B184560, dilution 1:1,000), anti-CD8 (53-6.7, catalog number 100707, lot number B171971, dilution 1:1,000), anti-CD44 (IM7, catalog number 103032, lot number B267976, dilution 1:500), CD45 (I3/2.3, catalog number 147711, lot number B254856, dilution 1:1,000) anti-CD147 (OX-114, catalog number 123716, lot number B262975, dilution 1:500), anti-CD19 (6D5, catalog number 115530, lot number B276004, dilution 1:1,000), anti-Ly6G (1A8, catalog number 127616, lot number B248844, dilution 1:500), anti-Ly6C (HK1.4, catalog number 128017, lot number B213757, dilution 1:500), anti-MHCII (M5/114.15.2, catalog number 107612, lot number B251993, dilution 1:500), anti-F4 80 (BM8, catalog number 123149, lot number B326894, dilution 1:250), anti-CD279 (PD1, 29F.1A12, catalog number 135221, lot number B194160, dilution 1:250), anti-HAVcr-2 (TIM3, RMT3-23, catalog number 119705, lot number B224472, dilution 1:250), anti-IFNγ (XMG1.2, catalog number 505842, lot number B270630, dilution 1:250), anti-TNF (MP6-XT22, catalog number 506322, lot number B218553, dilution 1:500), anti-CD11b (M1/70, catalog number 101204, lot number B307868, dilution 1:250) and anti-CD11c (N418, catalog number 117320, lot number B286499, dilution 1:250); Invitrogen: anti-CD62L (MEL-14, catalog number 564109, lot number 7341887, dilution 1:500), and anti-TOX (TXRX10, catalog number 80-6502-82, lot number 2246902, dilution 1:250). The human samples were stained with the following antibodies (BioLegend): anti-PD1 (EH12.2Z7, catalog number 329904, dilution 1:200), anti-TIM3 (F382E2, catalog number 345006, dilution 1:200), anti-CD8 (HIT8a, catalog number 300918, dilution 1:200) and anti-CD3 (SK7, catalog number 344834, dilution 1:200). For panels targeting transcription factors, the cells were fixed with the FoxP3 fix/perm buffer set (BioLegend) according to the manufacturer’s protocol. For panels targeting cytokines, fixation was performed with Cytofix/Cytoperm (BD Biosciences) according to the manufacturers protocol. The data collection utilized BD FACSDiva v9.0 for flow cytometry and was analyzed via FlowjoV10.

CD8+ TIL cytokine production assay

Tumor and lymph node cell suspensions were stimulated in complete R10 medium with Golgi-Plug (BD Biosciences), 3 μg ml−1 plate-bound αCD3 and 2 μg ml−1 αCD28 in complete R10 medium with Golgi-Plug for 5 h at 37 °C. In addition, dLNs and tumor cells were cultured in complete R10 medium Golgi-Plug, as a no stimulus control to determine gating strategies.

CRISPR–Cas9 knockout

The CRISPR–Cas9-mediated knockout method was modified on the basis of a previous publication. A total of 10 µg of Alt-R S.p. Cas9 Nuclease V3 (IDT) was mixed with an LDHa-targeted sgRNA (AAGCTGGTCATTATCACCGC) to form an RNP complex. A total of 2 million MEER cells were mixed with the RNP and Lonza SF buffer along with the Alt-R Cas9 Electroporation Enhancer for electroporation. The DJ-110 program (Lonza 4D Nucleofector) was used for electroporation.

Tumor interstitial fluid lactate measurement

The tumors were collected from mice and measured for tumor weight. The tumors were then cut up and placed on 20 mm nylon filters (Spectrum labs) and placed in top of 50 ml conical tube, wedged between the cap and the filter. The conical tube was then spun for 5 min at 2,000g. Approximately 20 μl of tumor interstitial fluid was then collected from the bottom of the conical tube. To measure lactate concentrations, we utilized a lactate meter (Nova Biomdeicals). The lactate concentration was then divided by tumor weight to generate final concentration of lactate.

T cell transduction, retroviral overexpression of MCT11 and adoptive transfer

The Slc16a11 murine coding sequence was obtained from the National Institutes of Health (NIH) database and cloned via Gibson assembly into a murine stem cell virus retroviral expression vector, which also encodes for an internal ribosome entry site–mCherry cassette. The vector was transfected into the Plat-E retroviral packaging cell line. The CD8+ T cells were collected and negatively isolated with Mojo beads (as described above) and were stimulated for 24 h with 5 μg ml−1 of αCD3, 2 μg ml−1 of αCD28 and 50 U ml−1 of IL-2. After 48 h, Plat-E retroviral supernatant was collected and supplemented with 5 μg ml−1 polybrene. The OT-I T cells were spun down and transduced in the viral supernatant for 2 h at 2,000 rpm. The cells were expanded and sorted via mCherry fluorescence on day 3 post transduction. A total of 3 × 106 OT-I T cells were then retro-orbitally transferred into D7 B16OVA-bearing mice.

Human CAR-T cell production and adoptive transfer

Human CD8+ T cells were isolated from bulk PBMCs from the blood bank. The T cells were then stimulated with anti-CD3/CD28 Dynabeads in RPMI supplemented with 10% fetal bovine serum (v/v) and 200 U ml−1 human IL-2 at 37 °C with 5% CO2 for 48 h. Then, CD19-CAR-expressing retrovirus was added to the expanding cells. After 5 days of expansion, Dynabeads were magnetically removed, and the cells were expanded for another 5 days in the presence of IL-2. CD19-A549-bearing NSG mice were adoptively transferred with 3 × 106 human CD8+ T cells expressing CD19-CAR.

Transcriptomic analysis by RNA-seq

RNA-seq data of B16 melanoma CD8+ TIL populations21 was utilized to evaluate expression of SLCs (GSE175408). For sequencing of the MCT11 conditional knockout T cells, 350 Tex cells were sorted from B16 tumors into lysis buffer in a 96-well plate. For all samples, complementary DNA was generated using the SMARTer Ultra Low Input RNA Kit for sequencing. The libraries were generated using the Nextera XT kit (Illumina) with 1 ng of cDNA in a total of 5 μl. Sequencing was done using a P3 flow cell-NextSeq2000. Bulk RNA-seq analysis was performed on PartekFlow. The paired-end reads were concatenated into a single fastq file. The reads were trimmed for adapters using Cutadaptv1.12 before being aligned to Mus musculus reference genome (mm38) using the RNA-seq aligner STAR2.7. Using the raw counts, differential genes were found by DESeq2. Publicly available RNA-seq data21 from WT Tex cells were used as controls (GSE175408). A gene set enrichment analyses of selected immunologic signature and hallmark gene sets was performed with clusterProfiler (https://doi.org/10.1016/j.xinn.2021.100141).

Transcriptomic analysis by single-cell RNA-seq

Single-cell RNA-seq data from blood and tumor-infiltrating immune populations from a cohort of patients with head and neck cancer22,23 was utilized to evaluate the expression of MCT11 across subsets of CD8+ T cells. Feature/barcode expression matrices were downloaded from the Gene Expression Omnibus (GSE139324), and cell type annotations were inferred as previously described3,4. CD8+ T cells were then bioinformatically isolated from other immune populations, and the top 2,000 highly variable genes were used as input for dimensionality reduction with principal component analysis. The top principal components were identified heuristically by identifying the inflection point on an elbow plot and were subsequently used for generating uniform manifold approximation and projection embeddings and Louvian-based clustering. The coexpression of PDCD1, HAVCR2 and SLC16A11 was then evaluated across clusters in PBMC and TIL and was visualized with a heat map.

[14C]-lactic acid oxidation experiments

The protocol is modified from previous published research51. T cells were incubated for 6 h in 0.2 ml RPMI containing 3 mM uCi ml−1 [14C]-lactic acid, sodium salt (NEC599050UC, PerkinElmer) and 2.5 mM sodium lactate. Afterward, the medium was transferred to borosilicate glass tubes. Each glass tube contained a microcentrifuge tube filled with 1 N sodium hydroxide to absorb CO2. The glass tubes were then sealed, and 5 N hydrochloric acid was injected into each tube to stop cellular metabolism and release carbon dioxide. After overnight absorption of [14C]-CO2, the samples were analyzed by liquid scintillation counting. A total of 25,000 T cells were used for ex vivo TIL studies and 100,000 were used for MCT11 overexpressing cells.

Continuous stimulation under hypoxia assay

The protocol was previously published8. Briefly, CD8+ T cells were isolated from mouse LNs or human PBMCs. The T cells were then activated at a 1:1 ratio with CD3/CD28 Dynabeads for 24 h. After 24 h, the T cells were split into four groups: acute TCR stimulus (no beads for remainder of assay in 20% oxygen), acute TCR stimulus under hypoxia (no beads for remainder of assay in 1.5% oxygen), continuous stimulus (10:1 beads to T cells for remainder of assay in 20% oxygen) and continuous stimulus under hypoxia (10:1 beads to T cells for remainder of assay in 1.5% oxygen). The t cells were kept in those conditions for 6 days (mouse assay) or 8 days (human assay) at a 1 million per milliliter concentration in R10 media with 50 IU IL-2.

Extracellular flux analysis

In vitro cultured B16 and MEER tumor cell lines were plated on an Agilent Seahorse poly-d-lysine-coated cell culture plate at 25,000 cells per well in minimal RPMI supplemented with 2 mM glutamine. The basal glycolytic rates were measured for 30 min, following the injections of 10 mM glucose, 2 μM oligomycin and 10 mM 2-deoxyglucose, and the readings were continued for 3 h. The measurements were performed on the XFe96 Analyzer.

Western blots

The tumor cells were lysed in RIPA lysis buffer with sodium orthovanadate and protease inhibitor for 15 min on ice. The lysates were spun down via high-speed centrifugation to clear debris and samples were mixed with 4× lithium dodecyl sulfate (LDS) buffer and boiled for 10 min. The lysates were loaded onto gels and ran at 200 V for 60 min. The proteins were transferred onto nitrocellulose membranes with transfer buffer at 30 V for 90 min. Blocking of the membrane was performed with nonfat dry milk for 1 h and washed with Tris-buffered saline + 0.1% Tween-20 (TBST). The membranes were washed three times with TBST and probed with a primary antibody overnight and horseradish peroxidase (HRP)-conjugated secondary antibody for 1 h at 4 °C. The membrane was then washed three times with TBST and incubated for 1 min with chemiluminescent substrate. The western blots were detected via chemiluminescent exposure to film. The antibodies used were b actin (Cell Signaling Technology) and LDHA (Cell Signaling Technology).

Sample sizes, randomization and blinding

No statistical methods were used to predetermine sample sizes, but our sample sizes are similar to those reported in previous publications. The tumor growth curves were randomized on the basis of initial tumor size once tumors became palpable, ensuring even distribution across groups. All tumor growth curves were conducted in a blinded manner—one author administered treatments while another independently measured tumor size. For all other experiments, data collection and analysis were not performed blind to the conditions of the experiments.

Statistical analysis

The data distribution was assumed to be normal, but this was not formally tested. We used unpaired or paired Student’s t-tests, one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons or two-way ANOVA with Tukey’s, Dunnett’s or Šidák’s multiple comparison tests to calculate the P values in GraphPad Prism. For tumor growth curves and survival curves, the P values were calculated using two-way ANOVA with Tukey’s multiple comparison and a log rank Mendel–Cox test, respectively. Statistical tests for specific experiments can be found in the figure legends. Grubb’s test was utilized to determine whether a value was a significant statistical outlier.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

留言 (0)

沒有登入
gif