Dynabeads® M-450 Tosylactivated beads (Invitrogen by Life Technologies) were coupled with 50 μg of protein according the manufacturer’s guidelines and as previously shown [27]. Isotype beads were coupled with 10% Purified NA/LE Mouse IgG2a, k isotype control (BD Bioscience), 30% Ultra-LEAF Purified Mouse IgG1, k isotype control (Biolegend) and 60% Ultra-LEAF Purified Human IgG1 isotype control antibody (Biolegend). Stimulatory beads were designed using 5% Purified NA/LE Mouse anti-human CD3 (HIT3a; BD Biosciences), 30% Ultra-LEAF Purified Human IgG1 Isotype control antibody (Biolegend), 15% Ultra-LEAF Purified anti-human CD28 antibody (CD28.2; Biolegend) or Recombinant human B7–1/CD80-Fc Chimera Protein, CF (R&D Systems). Beads were stored in Ca²⁺ and Mg²⁺ free PBS with 0.1% BSA and 2 mM EDTA pH 7.4 at 4 °C.

Healthy donor leukapheresis packs were obtained from the NIH blood bank (protocol 99-CC-0168). T cells were negatively selected using an EasySep Human T cell isolation kit (Stemcell Technologies) and cryopreserved in 90% FBS and 10% DMSO until use. Cells were thawed in a 37°C water bath and cultured overnight in RPMI1640 medium containing 10% fetal bovine serum, 1% penicillin-streptomycin-glutamine, 1% MEM non-essential amino acids solution, 15 mM HEPES, 1 mM sodium pyruvate and 55 μM 2-mercaptoethanol. Cells were plated at 1*10⁵/200 μl in 96-well round-bottom plates with M-450 Tosylactivated beads. Dexamethasone was purchased from Sigma Aldrich (D4902) and dissolved in DMSO. Niviolumab and ipilimumab F(ab’)₂ were used to block PD-1 and CTLA-4, respectively. Ipilimumab F(ab’)₂ was created using a Pierce F(ab’)₂ Preparation Kit per the manufacturer’s instructions (ThermoFisher Scientific, MA, USA). Cells were incubated at 37 °C in 20% O_2, and 5% CO₂ for four days for proliferation analyses and two days for Western blot and qPCR analyses.

αCD152-PE (BNI), αCD8α-Pacific-Blue or Alexa 488 (RPA-T8), αCD4-APC (RPA-T4), αCD197-Brilliant Violet 605 (G043H7), and αCD45RO-Brilliant Violet 785(UCHL1) were purchased from BioLegend. Anti-CD4-Brilliant UV 496 (SK3) and anti-CD279-APC (clone EH12.2.2H7) were purchased from eBioscience. Cells were stained for 30 min at room temperature, rinsed with FACS buffer (1% BSA and 0.01% sodium azide in PBS), fixed with 4% paraformaldehyde (Alfa Aesar, WA, USA), and resuspended in FACS buffer for flow cytometry analysis. Cell Cycle analysis was conducted using a Click-iT® Edu Flow Cytometry Assay Kit (Invitrogen, Carlsbad, CA, USA) per the manufacturer’s instructions. Cell apoptosis was investigated using Annexin V-Pacific Blue (Biolegend, San Diego, CA, USA) and propidium iodide (Sigma Aldrich, MO, USA). Data were acquired on a BD LSR Fortessa SORP II or LSR Fortessa X50, analyzed with Flowjo version 9.9.4 and SPICE version 5.35.

Isolated human T cells were collected after 48 h of stimulation and lysed in RIPA buffer with EDTA- free protease inhibitor cocktail set III (EMD Millipore, Billerica, Massachusetts, USA). Pierce BCA protein assay kit (Thermo Fisher Scientific, Rockford, IL, USA) was used to determine protein concentration. Samples were separated by SDS-PAGE (Bio-Rad) and transferred onto 0.2 μm pore size polyvinylidene fluoride membranes (PVDF) (Invitrogen, Carlsbad, CA, USA). The following antibodies were purchased from Cell Signaling: cleaved caspase 3 (5A1E), p27kip (2552 s), cyclin D3 (DCS22), CDK4 (D9G3E). Anti-CTLA-4 (EPR1476) was purchased from Abcam. The bands were detected by Super Signal West Pico chemiluminescence reagent (Pierce, Rockford, IL, USA). Antibodies against β-actin (AG74) or GAPDH (FL-335) were used as internal standards.

Total RNA was extracted using Trizol reagent (Invitrogen). Reverse transcription was performed using the Superscript III First-Strand Synthesis System (Invitrogen) with oligo dT. Subsequent RT-PCR was performed using SYBR green reagent from ABI and run on an ABI Quant Studio7. β-actin was used as a control to normalize gene expression. Primers and conditions used for RT-PCR are listed in the table below.Taba

Gene	Primer Sequence	Tm(°C)
IFNγ	F- 5’ TCGGTAACTGACTTGAATGTCCA 3′R- 5’ TCGCTTCCCTGTTTTAGCTGC 3’	60
IL-2	F- 5’ AACTCCTGTCTTGCATTGCAC 3′R- 5’GCTCCAGTTGTAGCTGTGTTT 3’	60
CTLA-4	F- 5’ CATGATGGGGAATGAGTTGACC 3′R- 5’ TCAGTCCTTGGATAGTGAGGTTC 3’	60
β-actin	F-5’ CCACACTGTGCCCATCTAC 3′R-5’ CCATCTCTTGCTCGAAGTCC 3’	60

All experiments were approved by the NCI-Bethesda Animal Care and Use Committee. Six to 8 week-old female albino C57BL/6 mice were purchased from Jackson laboratories (Bar Harbor, ME). For intracranial tumor implantation, mice were injected with 1*10³ GL261 cells that were stably transduced with a firefly luciferase-mCherry lentiviral vector. Water-soluble dexamethasone (D2915, Sigma Aldrich) was administered at 1 mg/kg/day by oral gavage. The non-toxic solubilizer, 2-hydroxypropyl-beta-cyclodextrin (H-107, Sigma Aldrich) was dissolved with water and matched in concentration to water-soluble dexamethasone for vehicle control. InvivoMab anti-mouse CTLA-4 (BE0131, BioXCell) or InVivoMab polyclonal Syrian hamster IgG isotype antibody (BE0087, BioXCell) were administered by intraperitoneal injection. Tumor was detected by luminescence imaging and analyzed with LivingImage Software.

Statistical Analysis was performed using GraphPad Prism 7.0 software. Data are expressed as the mean +/− standard deviation and statistical significance evaluated by two-tailed Student’s t-test. P-values < 0.05 were considered significant.

T cells require two signals for an effective proliferative response; the first signal arises when a T cell receptor (TCR) binds its cognate peptide:MHC complex, resulting in an intracellular signaling cascade through CD3. The second, co-stimulatory signal, is received when CD28 on T cells binds either CD80 or CD86 on antigen-presenting cells (APCs) [28]. Here, T cells isolated from healthy donor peripheral blood samples were stimulated with an αCD3 antibody for signal 1 and either an αCD28 antibody or recombinant CD80 protein for co-stimulation [27]. Recombinant CD80 permitted us to dissect the role of extracellular molecular interactions between CD80 and its multiple binding partners on T cells. These partners include the positive co-stimulatory molecule, CD28, and negative co-stimulatory molecules, such as CTLA-4 and PD-1.

Human T cells incubated with αCD3 and CD80 underwent multiple rounds of division, demonstrating that recombinant CD80 provided an efficient co-stimulatory signal under normal culture conditions (Fig. 1a). However, dexamethasone exposure caused defects in cell division for both CD4 and CD8 T cells. Cell proliferation statistics, including precursor frequency, expansion index, and proliferation index, were performed on multiple heathy donor T cells isolated from peripheral blood mononuclear cells to assess population dynamics and identify the proliferation defect [29]. Precursor frequency, defined as the cell-intrinsic probability that a cell will undergo at least one division, was diminished when T cells were exposed to dexamethasone (Fig. 1b). Similarly, dexamethasone-exposed T cells also had a reduced expansion index, a cell-extrinsic statistic used to express the fold expansion of the final cell count compared to the initial cell count (including undivided cells). Finally, the proliferation index, which represents the number of divisions that cells in the post-mitotic population have undergone, was modestly decreased by dexamethasone. Thus, cell proliferation statistics demonstrate that dexamethasone impaired the ability of CD80 co-stimulated CD4 and CD8 T cells to divide and restricted the expansion potential of these populations. In contrast, dexamethasone imposed only subtle cell division defects under all conditions in which CD28 was exclusively ligated with an αCD28 antibody (Additional file 1: Figure S1), indicating that direct ligation of CD28 can confer resistance to dexamethasone.Fig1Fig. 1

T cell proliferation is impaired by dexamethasone. Healthy donor T cells were cultured for four days with αCD3/CD80 microbeads in the presence of vehicle or dexamethasone. a Representative flow cytometry plots of CellTrace violet dilution. Plots were derived from gated CD4 (top row) or CD8 (bottom row) T cells. b Negatively-selected healthy donor T cells were stained and proliferation analyses determined by flow cytometry following four days of culture under the indicated conditions. Precursor Frequency, Expansion Index, and Proliferation Index are shown. Each symbol is the average of duplicate wells, and each paired symbol represents a different donor (n = 5 donors). Statistical significance was determined with a paired two-tailed T test. c Cell cycle analysis was performed on healthy donor T cells cultured with vehicle or dexamethasone and stimulated with αCD3/CD80 microbeads. EdU uptake and DNA content were used to identify G₀/G₁, S, and G₂/M phases. Representative flow images (top) and quantification of duplicate wells are shown (bottom) from two independent experiments. d Lysates from healthy donor T cells incubated with the indicated microbeads and vehicle or dexamethasone were probed for the indicated proteins. GAPDH was used as a loading control and is shown for each individual blot. Data are representative of three independent experiments

Increased apoptosis and cell cycle blockade can each manifest as a proliferative defect in vitro. Dexamethasone did not increase the frequency of apoptotic T cells when cultured with αCD3 and either αCD28 or CD80 co-stimulation as demonstrated by flow cytometry and Western analysis of cleaved caspase 3 (Additional file 2: Figure S2), demonstrating that cell death was not responsible for the proliferative defect. Indeed, dexamethasone has been shown to protect T cells from activation-induced cell death [30].

Cell cycle progression of CD80-costimulated T cells was next evaulated in response to dexamethasone exposure. Dexamethasone induced, on average, a 13% increase in the G₀/G₁ phase along with a 67% reduction of cells in S-phase and a 49% reduction in the G₂/M phase (Fig. 1c). Dexamethasone-exposed T cells also contained increased levels of p27 protein, an inhibitor of cell cycle entry, and concurrent reductions in CDK4 and cyclin D3 protein. Together, these data demonstrate that dexamethasone blocks cell cycle entry in T cell stimulated with αCD3 and CD80.

Following stimulation-induced expansion T cells may advance to a more terminally differentiated state. However, T cells cultured with dexamethasone had a greater proportion of naïve T cells (T_N) and central memory T cells (T_CM) and fewer effector memory T cells (T_EM) than cells cultured with vehicle control after four days of stimulation (Additional file 3: Figure S3). Terminal effector T cells (T_TE) were unchanged. These data indicate that dexamethasone blocked the ability of T cells to differentiate in response to αCD3/CD80 stimulation. The loss in T_EM numbers was reversed by increasing concentrations of CD80 (Fig. 2 and Additional file 4: Figure S4), suggesting that strong co-stimulation may protect this subset from the inhibitory effects of dexamethasone.Fig2Fig. 2

Increased co-stimulation ameliorates the inhibitory effects of dexamethasone. Negatively-selected healthy donor T cells were cultured with 5 μg/mL αCD3 and increasing concentrations of CD80 in the presence of vehicle or dexamethasone. a-b CD4 T cells cultured with vehicle (a) or dexamethasone (b). Flow cytometry plots showing proliferation of cells cultured with the indicated concentration of CD80 (left) and total numbers of naïve (T_N), central memory (T_CM), effector memory (T_EM), and terminal effector (T_TE) T cells following four days of culture (right) are shown. Differentiation subsets were assessed by CD45RO and CCR7 staining. Each condition was plated in duplicate, and data are representative of three independent experiments. Data were analyzed with an unpaired, two-tailed T Test

CD80 provides a positive co-stimulatory signal to T cells when bound by CD28. However, following TCR signal transduction, CTLA-4 is translocated to the outer membrane of T cells where it can outcompete CD28 to bind CD80 and block co-stimulation [31–34]. Using flow cytometry, we confirmed that T cell stimulation led to increased extracellular CTLA-4 protein levels. In the presence of dexamethasone, however, stimulation caused a fourfold increase in surface CTLA-4 protein compared to vehicle treated (Fig. 4a) as well as an increase in CTLA-4 transcription (Fig. 4b). For these experiments, T cells were stimulated with αCD3 and αCD28 antibodies because surface CTLA-4 is internalized following ligation by CD80 [35], potentially impairing detection by flow cytometry antibodies. CTLA-4 expression was consistently found to be higher in CD4 T cells compared to CD8 T cells during dexamethasone treatment, likely rendering CD4 T cells more susceptible to cell cycle inhibition by checkpoint molecule.Fig4Fig. 4

CTLA-4 blockade partially restores T cell proliferation in the presence of dexamethasone. a Flow cytometry analysis of CTLA-4 surface expression on CD4 (left) or CD8 (right) T cells stimulated with αCD3/αCD28 microbeads. Unstimulated (dashed line), stimulated in presence of vehicle (solid line), and stimulated in presence of dexamethasone (filled red line) are shown (left) and median fluorescence intensity (MFI) of CTLA-4-expressing T cells is quantified (right). Data are representative of four independent experiments. b Expression of CTLA-4 by qPCR of T cells stimulated in the presence of vehicle or dexamethasone. Data are representative of four independent experiments. c Healthy donor T cells stimulated for four days in the presence of vehicle or dexamethasone and with or without ipilimumab F(ab’)₂ antibody. Proliferation analysis of CD4 T cells (top) and CD8 T cells (bottom) was performed, and the ratio of cells stimulated with dexamethasone relative to vehicle control are shown for Precursor Frequency (PF), Expansion Index (EI), and Proliferation Index (PI). All samples were plated in duplicate and the ratios were analyzed with an unpaired, two-tailed T test. Data are representative of 7 healthy donors. d Cells were cultured as in (c). The number of T cells in each differentiation group were quantified by flow cytometry and analyzed by SPICE. e Expression of the indicated cytokines was determined by qPCR. Five healthy donors were assayed for each condition. Each data point represents and average of triplicate wells. Data were analysed with a paired, two-tailed T test

To determine if these findings translated to in vivo models, the C57Bl/6 syngeneic glioma tumor line GL261 was implanted intracranially. Consistent with the human in vitro data, dexamethasone increased the percentage of CTLA-4-expressing CD4 T cells of tumor-bearing mice in a dose-dependent manner (Fig. 5a). The percent of CTLA-4-expressing CD8 T cells also significantly increased in mice treated with the highest concentration of dexamethasone (2.5 mg/kg/day). Because dexamethasone upregulated CTLA-4, we hypothesized that CTLA-4 blockade would provide a survival advantage to dexamethasone-treated mice. GL261 tumor cells were implanted one week before commencement of dexamethasone or vehicle treatment to allow for immune surveillance and development of differentiated anti-tumor T cells. CTLA-4 blockade or isotype antibody were administered on days 13, 16, and 19 to vehicle or dexamethasone-treated mice to emulate the clinical scenerios where patients may be off or on corticosteroids before receiving immunotherapy (Fig. 5b). To ensure comparable tumor burden, mice were randomized into cohorts with equivalent tumor luminescence prior to treatment, and luminescence was measured weekly thereafter (Fig. 5c).Fig5Fig. 5

CTLA-4 blockade enhances survival of dexamethasone-treated mice. a CTLA-4 was measured on circulating CD4 (left) and CD8 (right) T cells 1 h following oral gavage of vehicle or the indicated concentration of dexamethasone. Each cohort contained eight mice with intracranial GL261 tumors. Vehicle and dexamethasone-treated cohorts were statistically analyzed with an unpaired two-tailed student’s T-test. b Schema of treatment cohorts for (c-d). GL261 ffluc-mCherry glioma cells were orthotopically implanted into C57BL/6 mice one week before treatment initiation. Luminescence readings were acquired 6 days following tumor implantation and weekly thereafter. Mice were treated with vehicle or dexamethasone as indicated. CTLA-4 blocking antibody or isotype control were administered on days 13, 16, and 19 following tumor implantation. c Luminescence of tumor-bearing mice at days 13 and 20 following tumor implantation. d Kaplan Meier survival curves of mice receiving the indicated treatments. n = 7 to 8 mice per cohort. Data are representative of two independent experiments