Ionic immune suppression within the tumour microenvironment limits T cell effector function

Study Approval

Animal experiments were conducted with the approval of the NCI and NIAMS Animal Use and Care Committees. All NIH cancer patients providing human samples were enrolled in clinical trials approved by the NIH Clinical Center and NCI institutional review boards. Each patient signed an informed consent form and received a patient information form prior to participation

Mice and cell lines

Pmel-1 (B6.Cg-/Cy Tg [TcraTcrb] 8Rest/J), Rag2^−/−, OT-II (B6.Cg-Tg (TcraTcrb)425Cbn/J), and C57BL/6 mice were obtained from the Jackson Laboratory. C57BL/6 male mice of 6-8 weeks of age were used as recipient hosts for adoptive transfer unless otherwise indicated. We crossed Pmel-1 with Ly5.1 mice (B6.SJL-Ptprc^aPepc^b/BoyJ) to obtain Pmel-1 Ly5.1 mice. We crossed OT-II with Rag2^-/- to obtain OT-II Rag^-/- mice. All mice were maintained under specific pathogen–free conditions. B16 (H-2D^b), a murine melanoma, transduced as previously described³¹ to express gp100 with human residues at position 25-27; EGS -> KVP. Mel624 was obtained from ATCC and Platinum-E ecotropic packaging cells were obtained from Cell Biolabs. Cell lines and maintained in DMEM media with 10% FBS, 1% glutamine and 1% penicillin-streptomycin.

Cell line authentication

Mel624 and Platinum-E cells were obtained from ATCC following authentication and validation as being mycoplasma free. Authenticated B16 was obtained from the National Cancer Institute Tumour Repository and validated as being mycoplasma free via a PCR based assay.

Statistical analysis

Data were compared using either a 2-tailed Student’s t test corrected for multiple comparisons by a Bonferroni adjustment or repeated measures 2-way ANOVA, as indicated. Where necessary, the Shapiro–Wilk test was used to test for normality of the underlying sample distribution. Experimental sample sizes were chosen using power calculations with preliminary experiments or were based on previous experience of variability in similar experiments. Samples that had undergone technical failure during processing were excluded from analyses. The Kolmogorov–Smirnov test was used to evaluate the significance between different distributions. For adoptive transfer experiments, recipient mice were randomized prior to cell transfer. The products of perpendicular tumor diameters were plotted as the mean ± SEM for each data point, and tumor treatment graphs were compared by using the Wilcoxon rank sum test and analysis of animal survival assessed using a log-rank test. In all cases, P values of less than 0.05 were considered significant. Statistics were calculated using GraphPad Prism 7 software (GraphPad Software Inc.).

Electrolyte analysis of serum and interstitial fluid

We utilized a previously reported method to isolate tissue interstitial fluid via a centrifugation method^6,15,14. Briefly, en bloc tissue was harvested, placed on triple layered 10 µM nylon mesh, and spun at <50 g for 5 minutes to remove surface liquid. Next, samples were centrifuged at 400 g, a previously validated speed at which intracellular contents are not liberated^15,14, for an additional 10 minutes. Flow through from this step was retained as interstitial fluid and assayed for indicated electrolyte concentrations in a blinded fashion within the NIH Central Clinical Chemistry Laboratory using auto analyzer (AA) ion selective electrode (ISE) quantification (Cobas 6000; US Diagnostics)³².

External Solution Formulations

Unless otherwise indicated, re-activation of cells in elevated concentrations of potassium ([K⁺]e) was performed with an isotonic RMPI formulation with an additional 40 mM of potassium for mouse cells and 50 mM for human cells in comparison to the control condition media. In principal this media was produced by obtaining a custom formulation of RPI 1640 from Gibco that was devoid of NaCl. For control conditions, this media was reconstituted with NaCl to produce a solution equimolar to standard RPMI. Thus, the final inorganic salt concentrations for the control condition was identical to that in RPMI 1640 (Gibco) in mM: NaCl 103.4, NaHC0₃ 23.8, Na₂PO₄ 5.6, KCl 5.3, MgSO₄ 0.4, Ca(NO₃)₂) 0.4. For isotonic media containing an additional 40 mM KCl, this media was reconstituted with a combination of NaCl and KCl such that the final inorganic salt concentrations, again in mM, were: NaCl 63.4, NaHC0₃ 23.8, Na₂PO₄ 5.6, KCl 45.3, MgSO₄ 0.4, Ca(NO₃)₂) 0.4. For 40 mM hypertonic NaCl the final inorganic salt concentrations were, in mM: NaCl 143.4, NaHC0₃ 23.8, Na₂PO₄ 5.6, KCl 5.3, MgSO₄ 0.4, Ca(NO₃)₂) 0.4. For 40 mM hypertonic KCl the final inorganic salt concentrations were, in mM: KCl 143.4, NaHC0₃ 23.8, Na₂HPO4 5.6, KCl 5.3, MgSO₄ 0.4, Ca(NO₃)₂) 0.4. All other additives in RPMI 1640 (Vitamins, amino acids, glutathione, phenol red, etc) were unchanged in quality or quantity.

T cell TCR-induced Ca²⁺ was carried out in a combination of NaCl deplete RPMI 1640 reconstituted with NaCl or KCl and HBSS with Ca²⁺ and Mg²⁺ and without phenol red (Gibco). As such, the final inorganic salt concentrations in control conditions were, in mM: NaCl 120.6, NaHC0₃ 13.9, Na₂PO₄ 2.9, KCl 5.3, MgSO₄ 0.4, Ca(NO₃)₂) 0.2, CaCl₂ 0.63, KH₂PO₄ 0.22, MgCl₂ 0.25, and D-glucose 8.3 mM. While in the same experiment the concentration of inorganic salts, in mM, in the condition with elevated [K⁺]e were: NaCl 80.6, NaHC0₃ 13.9, Na₂PO₄ 2.9, KCl 45.3, MgSO₄ 0.4, Ca((NO₃)₂) 0.2, CaCl₂ 0.63, KH₂PO₄ 0.22, MgCl₂ 0.25, D-glucose 8.3 mM. As this final solution was a combination of HBSS and RPMI 1640 it contained all of the non-inorganic salt additives (vitamins, amino acids, glutathione, phenol red) normally within standard RPMI 1640, as reported in the publically available Gibco formulation, at one-half of the original concentration.

Generation and activation of effector T cells

In vitro activation of T cells was carried out either by negative enrichment (Miltenyi) of CD8⁺ T cells from C57BL/6 mice followed by activation using using immobilized anti-CD3 (145-2C11; eBioscience) and anti-CD28 (37-51; eBioscience) along with expansion in culture medium containing IL-2 for 4-5 days, or via isolation of Pmel-1 Ly5.1 murine whole splenocytes followed by stimulation in vitro with 1 μM hgp100_25–33 peptide and expansion in culture medium containing IL-2 for 4-5 days. For analysis of T cell effector function, these cells were then stimulated on day 4 or 5 of culture in the indicated conditions for 5 hours with anti-CD3 and CD28 without IL-2 in the presence of brefeldin-A and monesin (BD Biosciences). For assaying co-inhibitory signalling, PD-L1 based co-inhibition PD-L1-IgG2a (R&D) was conjugated along with anti-CD3 and anti-CD28 antibodies (eBioscience) and control IgG2a (R&D) onto M-450 Epoxy Dynabeads (ThermoFisher) at a ratio of (1 : 1 : 2 : 6) for (anti-CD3 : anti-CD28 : PD-L1-IgG2a : IgG2a) for a total of 5 µg mL^-1 protein per 2 x 10⁶ beads overnight at 4°C. For comparative control conditions IgG2a was replaced for PD-L1 to control for bead loading. These beads were then incubated with effector CD8⁺ T cells at a ratio of 1:1 in the indicated conditions for 5 hours. For CTLA-4 based co-inhibition, anti-CD3, anti-CD28, and CTLA-4-IgG2a (R&D) or IgG2a (R&D) were coated onto tissue culture treated 96-well plates at a concentration of 5 µg mL^-1 (in a fashion similar to standard immobilized antibody based stimulation) for each reagent overnight at 4°C. For assaying neoantigen specific reactivity, human TIL from NCI-14-C-0062, generated as described in section below, were re-activated for 5 hours with autologous pre-activated B-cells that were pulsed with the indicated wild-type or neo-antigen peptides as described below. Transduced human peripheral blood lymphocytes were co-cultured with the A375 patient derived melanoma tumour line, HLA-A*0201⁺ and positive for NY-ESO-1 expression³³, at a ratio of 1 T cell to 3 tumour cells in the presence of brefeldin-A and monesin (BD Biosciences) for 5 hours. For analysis of human CD8⁺ TIL of varied histologies in the presence or absence of elevated K⁺ or Ba²⁺, TILs were first grown from tumour fragments cultured in 6000 IU mL^-1 IL-2 in 1 to 1 mixture of RPMI 1640 and AIM-V, supplemented with 5% in-house human serum, penicillin and 100 µg mL^-1streptomycin, 2 mM L-glutamine, 10 µg mL^-1 gentamicin), antibody (Miltenyi Biotec) for approximately 14 days per standard operating GMP protocols practiced by the Surgery Branch³⁴. These T cells were then subjected to a rapid expansion protocol (REP) using irradiated PBMC at a ratio of 1 to 300 in in the same complete media with 30 ng mL^-1 OKT3 in preparation for subsequent patient transfer, these cells were activated via immobilized anti-CD3 and anti-CD28 in the presence of K⁺, Ba²⁺, and/ or OA as indicated. For experiments interrogating mouse CD4⁺ cells naïve were obtained by isolating splenocytes from 6-10 week old OT-II Rag^-/- mice and subjected to negative selection of naïve CD4⁺ T cells (Stemcell Technolgies). These naïve cells were activated with immobilized anti-CD3 and anti-CD28 (5 μg mL⁻¹ each) in media for 2 days followed by an additional 2 days on blank plates with relevant polarizing cytokines present during the entire duration of the 4 day culture: Th1 conditions (IL-12 10 ng mL^-1 or as indicated, R&D Systems) Th17 conditions (IL-6 (20 ng mL⁻¹, R&D Systems), human TGF-β1 (1 ng mL⁻¹, R&D Systems), anti-IFN-γ neutralizing antibodies (10 μg mL⁻¹), IL-1β (10 ng mL^-1), or iTreg conditions (human TGF-β1 200 pg mL⁻¹) as indicated.

For selected experiments as indicated cells were treated with Okadaic Acid (OA) at a concentration of 200 nM (mouse CD8⁺ T cells) and 125 nM (human CD8⁺ TIL), Gramicidin (0.75 or 1.5µM), Ouabain 125µM (pre-incubated for 30 min), 1-EBIO (50µM), or Valinomycin (2µM).

RNA purification, quantitative real-time RT-PCR, RNA-sequencing and bioinformatic analysis

RNA sequencing was performed and analysed as described previously³⁵. CD8⁺62L⁺ C57BL/6 splenocytes were FACS sorted from 6-8 week old mice in biological triplicate and activated with anti-CD3 and CD28 for 48 hours in IL-2 100 IU mL^-1 and cultured in RPMI complete media for an additional 72 hours. Cells were then subjected to ficoll density separation to isolate live cells, placed in complete media without IL-2 in the presence or absence of elevated [K⁺]e, and either re-stimulated with anti-CD3 and CD28 or kept in complete media for two hours with no stimulation (No stim). Cellular RNA was preserved with RNAlater (Qiagen) and purified with the RNeasy Plus Mini Kit (Qiagen, Valencia, CA). RNA was subsequently used to prepare RNA-seq libraries by using TruSeq SRRNA sample prep kit (FC-122-1001, Illumina) according to the manufacturer’s instructions. The libraries were sequenced for 150bp (paired-end) using a NextSeq500 sequencer (Illumina). Sequence reads from each cDNA library were mapped onto the mouse genome build mm9 by using Tophat, and the mapped data was then processed by Cufflinks³⁶. The obtained data was normalized based on RPKM (reads per kilobase exon model per million mapped reads). To define differentially regulated genes, we used a 1.5-fold change difference between treatment groups. Real-time RT-PCR was performed for genes following cDNA generation by reverse transcription (Applied Biosystems) with primers from Applied Biosystems by Prism 7900HT (Applied Biosystems). RNA-sequencing raw data files are deposited at GEO-GSE84996.

Extracellular Acidification Rate and Basal Oxygen Consumption Rate

OCR and ECAR were measured following re-stimulation of T cells using anti-CD3/28 Dynabeads (Invitrogen) in a 0.8:1 ratio at 37°C using an XF24 extracellular analyzer (Seahorse Bioscience) as previously described³⁷ in the indicated conditions. Oxygen consumption rates (OCR) and extracellular acidification rates (ECAR) were measured in XF media (nonbuffered RPMI 1640 containing 25 mM glucose, 2 mM L-glutamine, and 1 mM sodium pyruvate) in a 1:1 mixture with tonicity controlled additive free standard RPMI 1640 with normal or elevated [K⁺] under basal conditions and in response to 1 μM oligomycin, 2 μM fluoro-carbonyl cyanide phenylhydrazone (FCCP), or 100 nM rotenone with 1 μM antimycin A (Sigma).

Intracellular Cytokine Staining, PhosFlow, and flow cytometry

Suspensions containing T cells were stained with fixable live/dead (Invitrogen) in PBS followed by surface antibody staining in FACS buffer (PBS with 0.5% BSA and sodium azide). For intracellular cytokine staining, cells were stained for intracellular molecules following fixation and permeabilization. For phosphostaining BD PhosFlow reagents were utilized and fixation/permeabilization protocols were carried according the to manufacturer’s protocol. After washing, cells were stained with antibody-fluorochrome conjugates for the indicated phosphorylated proteins (pZap70^Y319 (BD Biosciences), all other phospho-antibodies were purchased from Cell Signaling). Antibodies for surface staining and intracellular cytokine staining were purchased from BD Biosciences and eBiosciences. For determination of cytoplasmic membrane potential (V_m) cells were incubated in 2uM DiSBAC₄3 (Invitrogen) in conditions as indicated for 60 minutes prior to evaluation. For determination of [K⁺]i, cells were loaded with the potassium sensitive dye Asante Green-4 (TEFLabs) with PowerLoad (Invitrogen) per the manufacturer’s protocols. All experiments were conducted on a BD Fortessa flow cytometer (Becton Dickinson) and analyzed with FlowJo software.

Identification and purification of mutation specific tumour infiltrating lymphocytes (TIL)

Cancer specific mutations and TIL targeted against those mutations from patients with metastatic melanoma were identified as previously described³⁸. Briefly, patients were enrolled on a clinical protocol (NCI-14-C-0062). Whole-exomic sequencing (WES) was performed on tumour tissue and normal peripheral blood cells by Personal Genome Diagnostics (PGDx, Baltimore, MD) and the data was aligned to genome build hg18. A viable metastatic tumour deposit was selected for resection as a source for TIL. The resected en-bloc tumour was subjected to mechanical disruption via the GenlteMACS Dissociator. The bulk tumour digest was cultured in complete media without IL-2 or other cytokines overnight. The following day CD3⁺PD-1⁺ T cells were FACS sorted using a BD Jazz Flow cytometer. These sorted T cells were then subjected to a rapid expansion protocol (REP) using irradiated PBMC at a ratio of 1 to 300 in 50/50 medium (1 to 1 mixture of RPMI 1640 and AIM-V, supplemented with 5% in-house human serum, 100 U mL^-1penicillin and 100 µg mL^-1streptomycin, 2 mM L-glutamine, 10 µg mL^-1gentamicin), 3000 IU mL^-1of IL-2, and 30 ng mL^-1of OKT3 antibody (Miltenyi Biotec) for approximately 14 days. At the conclusion of the first REP T cells were screened and FACS sort enriched with the BD Jazz sorter based on 41BB positivity³⁸, against tandem mini-gene (TMG) constructs encoding for tumour specific mutations identified by WES for each patient and tumour, as described above and previously^9,38,39. These enriched neo-antigen specific T cells were then expanded via a second REP. At the conclusion of the second REP T cells were activated with peptide pulsed (at 1 : 3; effector : target) autologous CD40L stimulated B-cells with either 10 µg mL^-1 of the full length 23-mer mutated or wild peptide (Patient A) or 1 µg mL^-1 of the minimal epitope (Patient B,C) in the indicated conditions.

Retroviral transduction

Platinum-E ecotropic packaging cells (Cell Biolabs) were plated one day prior to transfections on poly-D-lysine coated 10cm plates (Corning) at a concentration of 6x10⁶ cells per plate. Packaging cells were transfected with 20µg of retroviral plasmid DNA encoding MSGV-Thy1.1, MSGV-Kcna3-Thy1.1, MSGV-Kcna3-Thy1.1 (W389F) (referred to as Kcna3_PD) ^{25, 40}, MSGV-Ppp2r1a(K416E) -Thy1.1 (referred to as PP2A_DN^41,42), MSCV-IRES-Thy1.1 (pMIT), or pMIT Akt1-CA⁴³ where indicated along with 6µg pCL-Eco plasmid DNA using 60µL Lipofectamine 2000 in OptiMEM (Invitrogen) for 8 hours in antibiotic-free media. Media was replaced 8h after transfection and cells were incubated for a further 48 hours. Retroviral supernatants were then collected and spun at 2000g for 2 h at 32°C onto 24 well non-tissue culture treated plates coated overnight in Retronectin (Takara Bio). For in vivo experiments live cells were isolated via ficoll density separation (Cedarlane) and subjected to positive-selection via CD90.1-microbead column enrichment according to the manufacturer’s protocol (Miltenyi) prior to transfer, yielding a CD8⁺ T cell population 70-95% Thy1.1⁺ prior to transfer. For retroviral transduction of human PBL with the NY-ESO-1 TCR, patients with metastatic melanoma on the clinical protocol NCI-13-C-0214 were subjected to leukopheresis and their PBL were transduced with a MSGV1 backbone encoding the NY-ESO-1 murine derived TCR as previously described¹⁷. After primary transduction and culture (10 days), these cells were then further expanded via a REP as described above (days 10-23) and subsequently assayed for target-specific effector function in the indicated conditions.

Adoptive cell transfer (ACT) and tumour immunotherapy

For immunotherapy, C57BL/6 were implanted with subcutaneous B16 melanoma (5 × 10⁵ cells). At the time of ACT, 10 days after tumour implantation, mice (n ≥ 5 for all groups) were sub-lethally irradiated (600 cGy), randomized, and injected intravenously with 5x10 5 Pmel-1 Ly5.1 cells transduced with control or Kcna3 expressing retrovirus, and received intraperitoneal injections of IL-2 in PBS (6 × 10⁴ IU/0.5 ml) once daily for 3 days starting on the day of cell transfer. Tumours were blindly measured using digital callipers. Tumour size was measured in a blinded fashion approximately every two days after transfer and tumour area was calculated as length x width of the tumour. Mice with tumours greater than 400mm² were euthanized. The products of the perpendicular tumour diameters are presented as mean ± SEM at the indicated times after ACT. For functional analysis of transferred Pmel-1cells, B16 tumour-bearing mice received Pmel-1 cells as above, and six to eight days following cell transfer mice were injected with 500 uL of 0.5 mg mL^-1 brefeldin-A (Sigma) and six hours later tumours were harvested and processed for live/dead, surface, fixation, and intracellular staining for direct in vivo IFN-γ capture⁴⁴. For ex vivo restimulation, tumours were harvested, processed as above, red cells were lysed with ACK lysis buffer for 2 min at room temperature, then cell suspensions were subjected to live-cell isolation via ficoll density gradient separation (CedarLane) and stimulated in media containing Leukocyte activation cocktail with Golgiplug (BD biosciences) for 4 h at a final concentration of 2 µL mL^-1.

Viral infection and kinetic analysis

For assessing the response of CD8⁺ T cells to acute viral infection, 5 × 10⁵ transduced and Thy1.1⁺ enriched Pmel-1 Ly5.1 CD8⁺ T cells were transferred into recipient Thy1.2 Ly5.2 C57BL/6 mice. Immediately following transfer, mice were infected with rhgp100 1 × 10⁷ plaque-forming units (PFU). At the indicated time points following transfer recipient mouse blood was obtained via sub-mandibular venipuncture and analysis for phenotype and enumeration of the congenically identified transferred cells.

TCR crosslinking, western blots

As described previously⁴⁵, to induce TCR crosslinking in vitro generated T cells were rested overnight in the absence of IL-2, subjected to live cell isolation via ficoll density gradient separation, pre-incubated with soluble anti-CD3 and anti-CD28-biotin, appropriate surface antibodies, and pharmacologic inhibitors where indicated in additive free RPMI 1640 at 4°C. Cells were then washed, brought to 37°C, and stimulated via the addition of streptavidin in the indicated conditions. For western blots, at the indicated time points following cross-linking cells were lysed by the addition of 95°C 2x Laemmli sample buffer with 2-mercaptoethanol (Bio-Rad Laboratories) and sonicated for 40 seconds at 10% intensity prior to gel loading. For K_v1.3 protein analysis, Thy1.1⁺ enriched T cells were lysed in 1xRIPA buffer (Thermofischer) with protease and phosphatase inhibitors (Roche). Western blotting was performed using TGX reagents (Bio-Rad Laboratories) and protocols on PVDF paper. Following transfer blots were blocked with 5% BSA then incubated with antibodies against phospho-Tyrosine (4G10; EMD Millipore), pAkt-T308, Total Akt, β-actin, Akt-substrate, or phospho-Threonine with appropriate HRP-conjugated or Alexa 647-conjugated secondary antibodies (Cell Signaling Technology), anti-K_v1.3 (NeuroMab). For HRP-conjugated secondary antibodies blots were developed using chemiluminescence (Thermo Fisher Scientific), gel images were captured with the Gel Doc XRS (Bio-Rad Laboratories). For PI3K activity, cells were fixed at the indicated time points by the 1:1 addition of a mixed lysis solution containing 30 mM HCl, 65% methanol, and 30% chloroform, maintained at -80°C then processed and analysed as previously described⁴⁶.

TCR induced Calcium Influx

T cells were isolated and primed as above. Prior to analysis cells were ‘rested’ in IL-2 free complete RPMI 1640 for at least 8 hours. Cells were loaded with 1 μM Fluo3-AM and 1 μM Fura Red-AM (Invitrogen) for 30 min at 37°C in HBSS with Ca²⁺ and Mg²⁺ and 2% FCS, washed twice and then resuspended in HBSS with anti-CD3 and CD28-biotin conjugates (eBioscience) and live/dead (Invitrogen). For flow cytometry analysis, samples were resuspended in pre-warmed 37°C 1:1 mixtures of HBSS and isotonic normokalemic or hyperkalemic additive free RPMI (as described above in External Solution Formulations), a baseline measurement was recorded for 20 seconds, followed by the addition of Streptavidin (Invitrogen) to a final concentration of 20 µg mL^-1 to induce TCR cross-linking and Ca²⁺ influx. Kinetic analyses were performed with the FlowJo software package (TreeStar).

shRNA mediated Ppp2r2d knockdown

A pLKO-Thy1.1 construct targeting Ppp2r2d was a generous gift of Dr.’s Zhou and Wucherpfenning and lentiviral particles were generated per their protocol⁴. BM cells were collected from the femurs and tibiae of 6-8 week old donor mice. After red blood cell lysis, hematopoietic stem and progenitor cells were enriched by autoMACS depletion of lineage positive cells using the Lineage Cell Depletion Kit (Miltenyi) for BM cells. Negatively selected cells were cultured in chemically defined serum free medium X-vivo 10 with Gentamicin (Lonza) supplemented with L-glutamine (1x) (Gibco), beta-mercaptoethanol (50mM), mouse recombinant SCF (50 ng mL^-1), IL-6 (10 ng mL^-1), IL-3 (5ng mL^-1), FLT-3L (5ng mL^-1) and IL- 7 (5ng mL^-1) (Peprotech). The following day, these lineage depleted BM cells were transduced by spin-infection at 32°C degrees 2000RPM for 90 minutes in the presence of lentiviral supernatant and 5 µg mL^-1polybrene (Sigma-Aldrich). Cells were incubated for another 2-4 hours prior to tail vein injection into Rag2^-/- lethally irradiated (1000 cGy) recipient mice at 1-2 x 10⁶ cells per mouse in 500 uL sterile PBS. CD8⁺Thy1.1⁺CD44⁺CD62L⁺ cells were FACS sorted 6-8 weeks following adoptive transfer, activated, and assayed as above.

PP2A Phosphatase Assay

PP2A activity was evaluated after immunoprecipitation utilizing a malachite green phosphatase assay kit as per the manufacturer’s instructions (EMD Millipore).