C57BL/6 mice were obtained from The Jackson Laboratory (#000664). OT-I mice (28) on the Rag2^−/− background were purchased from Taconic (#2334-F). All experiments in this study comply with the institutional guidelines approved by the Wake Forest Animal Care and Usage Committee.

Splenocytes, isolated from 6 to 12 week old female C57BL/6 mice, were incubated in the presence of 10⁻⁷M or 10⁻¹¹M OVA_257−264 peptide (SIINFKEL) (Chi Scientific) or no peptide for 3 h at 37°C in a 5% CO₂ humidified incubator. Stimulators were washed, irradiated (2,000 rads) and stained with CFSE (CFSE cell division tracker kit) (Biolegend) to allow their identification by subsequent flow cytometric analyses. OT-I Rag2^−/− splenocytes, isolated from 6 to 12 week old female mice were co-cultured with the irradiated C57BL/6 peptide pulsed splenocytes (1:16 ratio) in the presence of 10 U/ml recombinant human IL-2 (rhIL-2, carrier free, Biolegend) for 7 days at 37°C in a 5% CO₂ humidified incubator. Cultures were maintained in 24-well plates in RPMI-1640 medium supplemented with 2 mM L-glutamine, 0.1 mM sodium pyruvate, 1X non-essential amino acids, 100 U/ml penicillin, 100 μg/ml streptomycin, 10 mM HEPES (Gibco), 0.05 mM 2-mercaptoethanol (Sigma Aldrich) and 10% fetal bovine serum (Atlanta Biologics). OT-I CD8⁺ T cells were restimulated weekly with irradiated peptide-pulsed C57BL/6 splenocytes and rhIL-2 (10 U/ml). For neutralization of IL-4, 5 μg/ml of anti-mouse IL-4 (clone 11B11) (Leaf Purified anti-mouse IL-4, Biolegend) or 5 μg/ml of Rat IgG1, k isotype control (BD Pharmingen Purified NA/LE Rat IgG1, K isotype control, BD Biosciences) was added to the cultures weekly during the culture set up. For calcineurin inhibition assays 100 ng/ml of Cyclosporine A (Sigma Aldrich) or 25 μM of INCA-6 (CAS 3519-82-2) (Millipore Sigma) were added to cultures. Ionomycin (Sigma Aldrich) was added at 500 ng/ml to cultures to induce intracellular Ca²⁺ increases in the cells. DMSO (Sigma Aldrich) was used as a vehicle control.

OT-I Rag2^−/− cells from cultures stimulated with 10⁻⁷M or 10⁻¹¹M OVA pulsed stimulators were harvested on day 5 following weekly stimulation and plated in a 96-well round bottom plate (1 × 10⁵ cells/well). Cells were restimulated with graded concentrations of OVA_257−264 peptide (10⁻¹³M−10⁻⁶M) in the presence of brefeldin A (1:1,000 dilution) (BD GolgiPlug protein transport inhibitor, BD Biosciences) for 5 h at 37°C in a 5% CO₂ humidified incubator. Cells were pelleted and stained with Zombie Violet fixable viability kit (Biolegend) followed by staining with PerCP/Cyanine5.5 anti-mouse CD8α antibody (clone 53.6-7) (Biolegend). For intracellular cytokine staining, cells were fixed and permeabilized with Cytofix/Cytoperm (BD Biosciences) and subsequently stained with either APC anti-mouse IFNγ antibody (clone XMG1.2) (Biolegend) or APC Rat IgG1, k isotype control antibody (clone RTK2071) (Biolegend). Samples were acquired using the Fortessa X20 (BD Biosciences) and data analyzed using DIVA software (BD Biosciences).

Supernatant was harvested from OT-I cell cultures at 5, 12, 24, 48, and 72 h following stimulation with OVA_257−264 peptide-pulsed stimulators in the presence of rhIL-2 (10 U/ml). IFNγ secretion was measured using the mouse IFN-γ ELISA MAX set (Biolegend). IL-4 secretion was measured using the mouse IL-4 ELISA MAX set (Biolegend). The concentration of each cytokine was calculated per manufacturer's instructions using the kit's standard.

Murine P815 cells were pulsed with titrated (60 μg−0.73 μg/ml) amounts of purified hamster anti-mouse CD3ε Ab (clone 145-2C11) (BD Pharmingen) for 1 h on ice. Cells were irradiated with 10030 rads and washed. Irradiated P815 cells (3 × 10⁵) were cultured with naïve OT-I splenocytes (3 × 10⁵) (1:1 ratio) with 10 U/ml of rhIL-2 in a humidified incubator at 37°C under 5% CO₂ in 24-well plates. Supernatant was collected at 24, 48, and 72 h post stimulation.

CD8α⁺ cells were positively selected from naïve OT-I splenocytes using CD8α⁺ microbeads (MACS, Miltenyi Biotec, #130-117-044) as per the manufacturer's protocol. Isolated populations were routinely >96% CD8⁺. CD8α⁺ enriched cells from the OT-I mice (1 × 10⁵ cells/well) were cultured with high (10⁻⁷M), low (10⁻¹¹M) or no (NS) OVA_257−264 pulsed stimulators (1 × 10⁶ cells/well) in the presence of 10 U/ml of rhIL-2 in 96-well flat bottom plates for 48 h at 37°C in a 5% CO₂ humidified incubator. Supernatants were harvested at 48 h post stimulation and IL-4 production assessed via ELISA.

OT-I Rag2^−/− splenocytes were co-cultured with irradiated C57BL/6 OVA_257−264 peptide-pulsed splenocytes (1:10 ratio) with 10 U/ml of rhIL-2 for the IL-4Rα assay and in the absence of rhIL-2 for the pSTAT6 assay. CD8⁺ T cells were harvested at 5, 24, and 48 h post primary and secondary stimulation. For surface analysis of IL-4Rα, cells were pelleted and stained with Zombie Aqua fixable viability kit (Biolegend) followed by AF488 anti-mouse CD8α (clone-53-6.7) (Biolegend) and PeCy7 anti-mouse IL-4Rα (clone-I015F8) (Biolegend) or PeCy7 rat IgG2b, k isotype control (clone RTK4530) (Biolegend). For analysis of phosphorylation of pSTAT6, CD8⁺ T cells were fixed with 2% paraformaldehyde (PFA) for 10 min at 37°C to halt phosphorylation events. Cells were washed and then permeabilized with chilled TruePhos perm buffer (Phospho-protein Fixation/Permeabilization Buffer Set) (Biolegend) at −20°C for 60 min. Permeabilized cells were then washed with PBS containing 1% fetal bovine serum and stained with PE rat anti-mouse CD8α (clone 53-6.7)(BD Biosciences) and Alexa Fluor 647 anti-mouse STAT6 (pY641) (BD Phosflow, BD Biosciences) or Alexa Fluor 647 mouse IgG1, k isotype control (BD Phosflow, BD Biosciences). Cells were acquired using the Fortessa X20 (BD Biosciences) and data analyzed using DIVA (BD Biosciences) and FlowJo (BD Biosciences) software.

OT-I Rag2^−/− splenocytes (3 × 10⁵ cells) were co-cultured with 5 × 10⁶ irradiated CFSE-labeled C57BL/6 splenocytes that were either peptide-pulsed or left untreated. Cells were cultured with 10 U/ml of rhIL-2 at 37°C in a 5% CO₂ humidified incubator in 24-well plates containing complete media. OT-I cells cultured in the absence of peptide (non-stimulated) were harvested at 24 h and peptide stimulated OT-I cells were harvested at 48 h post stimulation. Cells (7 × 10⁵) were transferred to poly-L-lysine coverslips and fixed with 2% paraformaldehyde for 20 min. Cells were permeabilized using 0.1% Triton X100 (Fisher Scientific) for 5 min and non-specific binding was blocked using PBS with 10% goat serum (Lampire biological) for 20 min. Cells were stained with purified mouse anti-NFATC1 (Clone-7A6) (BD Pharmingen) for 30 min and then Rhodamine Red-X-conjugated AffiniPure Fab Fragment goat anti-mouse IgG (H⁺L) (Jackson Immuno Research laboratories Inc) for 20 min at room temperature. Cells were then incubated for 20 min with a high concentration (40 μg/ml) of AffiniPure Fab Fragment goat anti-mouse IgG (H⁺L) (Jackson Immuno Research laboratories Inc) to mask mouse IgG epitopes, followed by staining with a second mouse monoclonal antibody to NFATC2 (clone-25A10.D6.D2) (Thermo Fisher Scientific) for 30 min. Antibody was detected with Alexa Fluor 647-conjugated AffiniPure Fab Fragment goat anti-mouse IgG (H⁺L) (Jackson Immuno Research laboratories Inc.). Cells were mounted and nuclei stained with ProLong Gold Antifade reagent with DAPI (Invitrogen). Fluorescent images were collected with a 12-bit Retiga EX 1350 digital camera using a Nikon TE300 microscope equipped with a high-numerical aperture 60X/1.4 oil-immersion objective.

OT-I Rag2^−/− splenocytes (1 × 10⁵ cells) were co-cultured with irradiated OVA_257−264 peptide-pulsed splenocytes (1 × 10⁶) in the presence of 10 U/ml of rhIL-2 at 37°C in a 5% CO₂ humidified incubator_. Cultures were cultured for 8 h in 96-well plates containing complete media. Cells were then plated in XF culture plates coated with CellTak (Corning) in Seahorse Media (DMEM based media with 10 mM glucose, 2 mM L-Glutamine and 1 mM sodium pyruvate). Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured under basal conditions and after sequential treatments with 1 μM oligomycin, 1 μM fluoro-carbonyl cyanide phenylhydrazone (FCCP), and 0.5 μM rotenone/antimycin A using a Seahorse XF 96 analyzer (Agilent). Data was analyzed using Wave software according to manufacturer's instructions.

OT-I Rag2^−/− splenocytes (1 × 10⁵ cells/well) were co-cultured with irradiated 10⁻⁷M or 10⁻¹¹M OVA_257−264 peptide-pulsed C57BL/6 splenocytes or non-peptide pulsed splenocytes (1 × 10⁶ cells/well) in the presence of 10 U/ml of rhIL-2 in 96-well flat bottom plates at 37°C in a 5% CO₂ humidified incubator. Cells were treated with either 1 μM Oligomycin (Agilent), 10 mM 2-Deoxy-D-Glucose (Sigma Aldrich) or left untreated. Cell supernatants were harvested at 48 h for analysis of IL-4 and TNFα by ELISA. Cell viability was assessed by trypan blue exclusion.

Statistical analysis was performed using Prism 7 (GraphPad) and the results are represented as the mean ± SEM.

Fluorescent images were processed with the EBImage package (29) in the open source programming environment R (30). Micrographs were acquired with appropriate bandpass filters for DAPI, CFSE, Rhodamine Red-X and Alexa Fluor 647 as 12-bit, 676 × 516-pixel grayscale images. The fluorescent illumination and amplifier settings were adjusted to ensure that the signal for each fluorochrome was acquired in the linear range of the instrument. Raw images were scaled by subtracting the modal value of background pixels in the respective image. Binary nuclear masks defining individual nuclei were generated by applying a local thresholding and watershed algorithm to the low-pass-filtered DAPI image. Nuclear masks along the edge of the image and nuclear masks with an apparent area less than the 5th percentile or greater than the 95th percentile were omitted. Each nuclear mask was enlarged by 15 pixels (3.2 μm) using Voronoi tessellation to create a binary mask defining each cell. The nuclear masks were reduced by 3 pixels (1.2 μm) to define an inner nuclear mask. A perinuclear mask was defined as an annulus of 6 pixels (2.4 μm) surrounding the nuclear border. Registration between the nuclear mask and each fluorescent image was measured and the fluorescent image translated as needed to correct any misalignment. The fluorescent signal within each mask was determined and used for further analysis. CFSE-stained cells were identified and excluded from analysis. Cells out of the plane of focus also were excluded from analysis. These cells were identified as having a perinuclear fraction of DAPI staining that exceeded the median by 1.5-times the interquartile range of this parameter. In three experiments, between 50 and 200 cells could be evaluated for each unique sample. Staining for NFATC1 and NFATC2 was ~2–10-fold greater than staining due to the secondary antibodies alone. The total NFATC1 and NFATC2 signal was normalized to the DAPI signal in each cell and analyzed as log-transformed values. The signal for NFATC1 and NFATC2 in the various masks was measured and expressed as a fraction of the total signal to provide a measure of the cytoplasmic, nuclear, and perinuclear fraction. Because the ratios and log-transformed values were not normally distributed, imaging data were evaluated by non-parametric methods. The Wilcoxon rank sum test was used to detect differences between pairs of samples. Multiple samples were analyzed in two steps. The potential for differences among groups was identified by a significant (p < 0.05) outcome with the Kruskal-Wallis omnibus test. Pairwise non-parametric comparisons were then performed by the Dunn test, which adjusts for multiple comparisons (Dunn, 1964). P < 0.05 were considered to identify significantly different pairs.

CD8⁺ effector T cells were generated by weekly restimulation of cells from OT-I Rag2^−/− mice with C57BL/6 splenocytes that had been pulsed with a high (10⁻⁷M) or low (10⁻¹¹M) concentration of OVA_257−264 peptide. The functional avidity was assessed weekly by quantifying IFNγ production following stimulation with graded concentrations of peptide antigen as we have previously reported (5, 16–20, 31–33). Both of the stimulation conditions resulted in robust recruitment of OT-I cells into the response as demonstrated by expression of CD69 and CD44 as well as recovery and acquisition of effector function (Supplementary Figure 1).

Following primary stimulation, effector cells generated by stimulation with the high peptide condition required significantly more peptide to achieve 50% maximal IFNγ production compared to OT-I cells maintained with low peptide (Figure 1A). The increased requirement of cells maintained on 10⁻⁷M peptide was apparent throughout the first four stimulation cycles assessed here (Figure 1B and Supplementary Figure 2). Our previous studies suggest that the functional avidity established in these lines becomes a fixed property of the cells that is maintained long-term (our unpublished data). The modulation of functional avidity by peptide in the cultures shown here is in keeping with our previously published lines generated from OT-I or P14 TCR transgenic mice (16–19, 32) as well as WT mice (5).

While the ability of high peptide/MHC to program lower avidity is well-established, the mechanism responsible for avidity modulation is unknown. We assessed cytokine production following stimulation with high vs. low peptide/MHC with the hypothesis that encounter with a high vs. low level of peptide/MHC would differentially induce cytokine production by the OT-I cells.

Splenocytes isolated from naïve OT-I Rag2^−/− mice were stimulated with splenocytes previously pulsed with either high (10⁻⁷M), low (10⁻¹¹M) or no (NS) OVA_257−264 peptide. IL-4 and IFNγ production were measured at 5, 12, 24, 48, and 72 h post primary stimulation. We also assessed secondary stimulation as an indicator of whether the pattern of cytokine production was maintained. As would be expected, OT-I cells produced IFNγ under both conditions, although the level detected was significantly higher following stimulation with high peptide (Figure 2A). Intracellular cytokine staining (ICCS) analysis at 48 h following stimulation demonstrated that the higher amount of IFNγ in the supernatant was the result of both an increase in the percentage of cells producing IFNγ and the amount produced on a per cell basis (Supplementary Figure 3). IFNγ continued to accumulate in the supernatant through 72 h post primary stimulation in both high and low peptide stimulated OT-I cells. Following secondary stimulation, IFNγ was produced earlier (5–12 h) as would be expected. Similar to the initial stimulation, high peptide resulted in increased IFNγ levels compared to low peptide (Figure 2A).

We next assessed the potential for responding T cells to produce IL-4 in a peptide dose dependent fashion. Significant amounts of IL-4 were detected in cultures stimulated with 10⁻⁷M peptide-pulsed splenocytes following both primary and secondary stimulation (Figure 2B). IL-4 was initially detected at 24 h and peaked at 48 h. Only minimal IL-4 was present in the cultures maintained on 10⁻¹¹M pulsed stimulators (Figure 2B).

To strengthen our interpretation that the IL-4 was produced by CD8⁺ T cells, we isolated CD8⁺ cells from OT-I splenocytes by magnetic bead selection. Enriched populations were >96% CD8⁺. CD8⁺ OT-I cells were co-cultured with OVA_257−264 peptide pulsed stimulators. IL-4 was readily detected following stimulation with 10⁻⁷M, but not 10⁻¹¹M or non-peptide pulsed irradiated splenocytes, supporting the ability of OT-I T cells to produce this cytokine (Figure 2C). These results reveal the production of IL-4 as a previously unknown response of CD8⁺ T cells to stimulation with high amounts of peptide/MHC.

These data suggest that high TCR engagement is the mechanism responsible for the IL-4 production following stimulation with splenocytes pulsed with high peptide. If so, we predicted that exposure to increasing amounts of anti-CD3 antibody would also drive increases in IL-4 production. To test this, naïve OT-I cells were stimulated with titrated amounts of anti-CD3ϵ Ab presented by P815 cells via binding to the FcR, an approach that allows for optimal crosslinking of CD3. IL-4 was measured in the supernatant at 24, 48, or 72 h of culture. We observed a dose dependent increase in IL-4 production, consistent with the high IL-4 produced following stimulation with APC presenting high levels of peptide/MHC (Figure 2D). As with peptide stimulation, IL-4 in the cultures peaked at 48 h. These data establish IL-4 production as a consequence of high TCR engagement on CD8⁺ T cells.

Given the differential production of IL-4 by OT-I cells stimulated with high vs. low peptide/MHC, we considered the possibility that the expression of the IL-4 receptor (IL-4R) may also be regulated by antigen dose in a fashion that impacts the potential for IL-4 to serve in a regulatory capacity. IL-4Rα expression was measured at baseline and following stimulation with high (10⁻⁷M) or low (10⁻¹¹M) peptide-pulsed stimulators. IL-4Rα was readily detected ex vivo in naïve cells (Figure 3A). Stimulation resulted in an increase in IL-4Rα expression regardless of peptide concentration. We did note that while the two stimulation conditions resulted in similar expression of IL-4Rα at 48 h, the kinetics of upregulation differed, with high peptide/MHC stimulation resulting in a more rapid increase in IL-4Rα (Figure 3A). These data suggest the IL-4 pathway is primarily constrained by IL-4 production as opposed to responsiveness that depends on receptor expression.

Our data thus far show that high peptide increases autocrine IL-4 production and that either high or low peptide activated OT-I cells robustly express IL-4Rα. If the IL-4 produced as a result of high peptide stimulation was contributing to the differentiation of OT-I effectors we would expect to see robust phosphorylation of the IL-4R signaling molecule STAT6. To test this possibility, phosphorylation of STAT6 (pSTAT6) was assessed at 5, 24, and 48 h following stimulation with high (10⁻⁷M) or low (10⁻¹¹M) peptide-pulsed splenocytes. At 24 h post primary stimulation, on average, nearly 65% of OT-I cells stimulated with high peptide were positive for pSTAT6 and this increased to 77% at 48 h (Figure 3B, right panel and Figure 3C). pSTAT6 was also observed following secondary stimulation (Figure 3D). In contrast, minimal STAT6 phosphorylation was observed in cells stimulated with low peptide (Figure 3B, left panel and Figure 3C). Together, these data show that OT-I cells respond to the IL-4 that they produce following high TCR engagement.

We next tested the potential for high peptide/MHC-induced autocrine IL-4 to modulate functional avidity. While exogenous addition of IL-4 has been reported to decrease avidity in CD8⁺ T cells, the amount utilized in these studies (25 ng/ml) was much greater than that produced by the OT-I cells (34). To determine the modulatory potential of the autocrine IL-4 produced in our cultures, OT-I effectors were generated by weekly stimulation with high (10⁻⁷M) or low (10⁻¹¹M) OVA_257−264 pulsed splenocytes in the presence of neutralizing IL-4 antibody. The effectiveness of neutralization was assessed by an ELISA that relied on the same antibody used in the stimulation cultures. We reasoned cytokine binding by the antibody in the culture would prevent IL-4 detection in the ELISA due to competition for the detection antibody in the ELISA. This approach showed an inability to detect IL-4 in the ELISA, consistent with highly efficient blocking of cytokine activity (Supplementary Figure 4A).

Addition of neutralizing IL-4 antibody in the cultures stimulated with high peptide/MHC resulted in a significant increase in peptide sensitivity (Figure 4). In fact, the amount of peptide required to activate the cultures generated by stimulation with high peptide/MHC in the presence of neutralizing IL-4 antibody was similar to that required by effectors generated with 10⁻¹¹M peptide-pulsed splenocytes (Figure 4). In contrast, there was no significant change in functional avidity of cells maintained on 10⁻¹¹M peptide when neutralizing antibody to IL-4 was added to the cultures (Figure 4), as would be expected given the lack of IL-4 production. We also assessed the phosphorylation of STAT6 in cultures that were stimulated with 10⁻⁷M peptide-pulsed stimulators in the presence of neutralizing IL-4 antibody. In comparison to the isotype control, the average percentage of CD8⁺ T cells positive for pSTAT6 decreased by 4.27 fold in cultures where IL-4 was neutralized (Supplementary Figure 4B). The MFI of the remaining pSTAT6⁺ cells decreased on average by 5.05 fold. Overall these data demonstrate a previously unknown mechanism for self-tuning of the activation setpoint of CD8⁺ T cells that is determined by production of and signaling in response to autocrine IL-4.

We next sought to understand the effects of high peptide/MHC encounter that resulted in the avidity modulating production of IL-4 in CD8⁺ T cells. In CD4⁺ T cells, IL-4 production is controlled by nuclear factor of activated T cells (NFAT) family members present in the nucleus (35, 36). NFAT proteins are inactive in a phosphorylated form in the cytoplasm; however, when intracellular Ca²⁺ levels rise calcineurin dephosphorylates NFAT and these proteins are translocated to the nucleus where they bind the AP-1 promoter and activate transcription (37). IL-4 transcription in CD4⁺ T cells has been reported to be associated with nuclear NFATC1 (36), while IFNγ appears dependent on NFATC2 (35). To our knowledge, the role of NFAT as a driver of IL-4 production by CD8⁺ T cell has not been previously evaluated.

OT-I cells were stimulated with high (10⁻⁷M) or low (10⁻¹¹M) peptide-pulsed splenocytes. OT-I cells cultured in the absence of peptide served as a baseline (Naive, N). The amount of NFATC1 and NFATC2 was evaluated by quantitative fluorescent microscopy 48 h after primary stimulation. Non-stimulated cells were evaluated at 24 h of culture as cell viability was becoming compromised at 48 h. The fluorescent intensity in cells stained with secondary antibody alone served as the background. Cells stimulated with low peptide/MHC (Figure 5A) contained low levels of nuclear NFATC1 and NFATC2. The brightness of the images in Figure 5A was increased 3-fold over those in Figure 5B in order to better visualize the cellular distribution of NFAT. Cells stimulated with 10⁻¹¹M peptide-pulsed splenocytes displayed a perinuclear enrichment of NFAT, which is characteristic of a cytoplasmic protein that is excluded from the nucleus. Further, ~40% of the cells displayed a prominent perinuclear aggregate of NFAT protein as seen in the representative cells in panels 1, 2, 5, and 6 of Figure 5A and panels 2, 3, and 4 of Figure 5C. Stimulation with high peptide/MHC increased the levels of NFATC1 and NFATC2 present at 48 h (Figure 5B). Prominent perinuclear aggregates of NFAT protein were observed in <10% of these cells. In the absence of stimulation, low levels of nuclear NFATC1 and NFATC2 were observed (Figure 5C). The brightness of the images for both non-stimulated and 10⁻¹¹M stimulated OT-I cells was increased 3-fold to facilitate visualization of NFAT. The localization of NFAT protein was determined with an annulus extending from 1.2 μm within the nuclear border to the periphery of the cell. A non-parametric Dunn's test for multiple comparisons established that the increased nuclear levels of NFATC1 and NFATC2 for cells pulsed with 10⁻⁷M was significantly greater than levels in 10⁻¹¹M cells (5.2–5.5-fold) or naïve cells (4.8–4.3-fold) with an adjusted p < 0.001 (Figure 5D).

The increased nuclear NFATC1 and perhaps NFATC2 in cells stimulated with high peptide/MHC is consistent with a model wherein IL-4 production is a downstream consequence of the increased presence of these transcription factors in the nucleus. We hypothesized that IL-4 production would be altered by modulating signaling through the calcium/calcineurin pathway. Naïve OT-I cells were stimulated with high (10⁻⁷M) or low (10⁻¹¹M) peptide-pulsed splenocytes in the presence of cyclosporine A (CsA) to inhibit calcineurin or ionomycin to increase intracellular Ca²⁺. The effect on IL-4 was assessed at 48 h following stimulation. Cells that were stimulated with high peptide/MHC in the presence of CsA exhibited a significant decrease in IL-4 production when compared to the vehicle control (Figure 6A, right panel), while cells that were stimulated with low peptide/MHC in the presence of ionomycin had significantly increased production of IL-4 (Figure 6A, left panel).

To confirm that IL-4 production was a downstream consequence of increased NFAT activation we used a small molecule inhibitor know as INCA-6 which selectively binds with high affinity to calcineurin, blocking calcineurin dependent NFAT dephosphorylation (38). Unlike CsA, INCA-6 does not interrupt any other downstream calcineurin signaling [e.g., IκBα kinase (IKIC) and NFKβ (39, 40) other than NFAT (38). Autocrine IL-4 was measured in the supernatants of cells at 48 h following stimulation with or without INCA-6. Cells that were stimulated with high peptide in the presence of INCA-6 exhibited a decrease in IL-4 production compared to high peptide stimulated cells with a vehicle control (Figure 6B). These data confirm the effects of CsA.

Our data show that stimulation with high peptide increases NFAT activation and IL-4 production that is mediated by regulated cytosolic calcium flux. The coordination of these two processes is highly dependent on mitochondrial and glycolytic metabolism. Previous studies show that mitochondria localize to the immunological synapse during T cell activation, where these organelles regulate calcium flux, resulting in NFATC1 activation (41). Further, T cell stimulation is known to enhance nutrient uptake and breakdown of glucose as well as enzymatic activity of key enzymes in the TCA cycle, thereby increasing mitochondrial respiration (42). Based on this understanding, we predicted that stimulation with high peptide/MHC would result in an altered metabolic profile compared to stimulation with low peptide/MHC.

Oxidative phosphorylation and glycolysis were measured 8 h after stimulation with splenocytes pulsed with high (10⁻⁷M) or low (10⁻¹¹M) peptide or splenocytes without peptide (NS). High peptide stimulation resulted in a 40% (^**p < 0.007) increase in basal oxygen consumption rate (OCR) when compared to low peptide and an over 30% (^*p < 0.03) increase when compared to non-stimulated cells. In stark contrast, cells stimulated with the low peptide condition did not increase oxidative phosphorylation compared to the non-stimulated cells. This suggests that high peptide exposed OT-I cells have selectively elevated levels of oxidative phosphorylation compared to low peptide exposed cells (Figures 7A,B). Interestingly, the spare respiratory capacity was higher in low peptide stimulated OT-I cells when compared to the rest of the groups (Figure 7C). Cells stimulated with the low concentration of peptide also exhibited a significant increase in the percent maximal respiration (normalized to baseline OCR) (Supplementary Figure 5A).

To examine the potential for a shift in glycolysis we evaluated glycolytic flux by measurement of the extracellular acidification rate (ECAR). Our data indicated that high peptide stimulated cells had an over 40% (^**p < 0.001) increase in ECAR when compared to low peptide stimulated cells and an over 60% (^***p < 0.0001) increase when compared to non-stimulated cells (Figure 7D). Low peptide stimulation of OT-I cells also increased glycolysis, but this was significantly decreased compared to cells exposed to high peptide. While there was an overall greater increase in metabolic activity in the 10⁻⁷M stimulated cells compared to 10⁻¹¹M stimulated cells, the OCR/ECAR ratio was not significantly different at this time point, suggesting there was no peptide concentration dependent preferential upregulation of the two metabolic pathways (Supplementary Figure 6).

The altered metabolic profile observed with increased peptide stimulation led us to investigate the effects of OXPHOS and glycolysis on the autocrine production of IL-4. OT-I cells were cultured with irradiated splenocytes previously pulsed with high (10⁻⁷M) or low (10⁻¹¹M) peptide or no peptide (NS). Cells were treated with either 1 μM of oligomycin, a known inhibitor of OXPHOS through the inhibition of ATP synthase, or 10 mM of 2-Deoxy-D-glucose (2-DG), a glucose analog that inhibits glycolysis. Production of autocrine IL-4 was measured at 48 h. The inhibition of OXPHOS in cells stimulated with high peptide resulted in a near complete loss of IL-4 production when compared to the untreated control (Figure 7E). A similar effect was observed when glycolysis was inhibited (Figure 7F). Treatment with 2-DG or oligomycin did not result in a reduction in the cell counts, showing that the decrease in IL-4 production was not due to decreased cell viability (Supplementary Figure 5B).

We also measured the effect of these treatments on the production of TNFα. Cells stimulated with high peptide in the presence of oligomycin, while trending toward a reduction, showed no significant decrease in cytokine production compared to cells without treatment (Figure 7G, right panel). In contrast, cells stimulated with low peptide had a near complete loss of TNFα production (Figure 7G, left panel). Inhibition of glycolysis also resulted in a near complete inhibition of TNFα production in cells stimulated with low peptide (Figure 7H, left panel), while having no significant effect on those stimulated with high peptide (Figure 7H, right panel). Together, these data show IL-4 production is highly dependent on both OXPHOS and glycolysis as blocking either pathway results in the near complete loss of IL-4 production. While TNFα production following stimulation with the low amount of peptide was also dependent on both the glycolytic and OXPHOS pathways, inhibiting either of these pathways individually in OT-I cells stimulated with high peptide did not result in loss of the ability to produce TNFα. These data suggest the metabolic regulation of cytokine production diverges in CD8⁺ T cells depending on the strength of stimulation.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.