Peripheral blood samples were obtained from healthy adult volunteers. Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation using Lymphoprep™ (Axis-Shield, Oslo, Norway). CD14⁺ monocytes and CD4⁺ T cells were isolated by magnetic-activated cell sorting (MACS) according to the manufacturer’s instructions (Miltenyi Biotec, Bergisch-Gladbach, Germany), and purity was confirmed by flow cytometry. Monocytes (average purity 98%) were isolated by positive selection using anti-CD14 microbeads. CD4⁺ T cells were isolated via negative depletion (average purity 95%), and in some experiments, CD45RO⁺ CD4⁺ T cells were subsequently enriched by positive selection using CD45RO microbeads (average purity 87%). In some experiments, CD4⁺ T cells were sorted to very high purity (> 99%) and part of the cells depleted of CD4⁺ CD25^highCD127^low Tregs by FACS-sorting after labeling cells with CD4 PerCP Cy5.5 (SK3), CD25 PE (M-A251), CD127 Alexa Fluor 488 (A019D5) mAbs (all from BioLegend, Cambridge, UK). The study was approved by the Bromley Research Ethics Committee (06/Q0705/20), and written informed consent was obtained from all participants.

Cells were cultured at 37°C with 5% CO₂ in culture medium [RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum (lot# 07F7435K, South American origin)] and 1% penicillin, streptomycin, and l-glutamine (all Life Technologies, Carlsbad, CA, USA). Freshly isolated bulk CD4⁺ or memory (CD45RO⁺)-enriched T cells (0.5 × 10⁶) and CD14⁺ monocytes (0.5 × 10⁶ unless otherwise indicated) were cocultured with 100 ng/ml anti-CD3 mAb (clone OKT3, Janssen-Cilag Ltd., High Wycombe, UK). Cocultures were incubated at 37°C with 5% CO₂ for 3 days. In some experiments, MACS-isolated CD4⁺ T cells were cultured alone with anti-CD3/CD28 stimulation. Anti-CD3 (OKT3) was coated onto culture plates at 1.25 µg/ml in PBS for a minimum of 2 h at 37°C. Wells were then washed three times with PBS before adding CD4⁺ T cells in culture medium. Anti-CD28 (clone CD8.2; BD Biosciences) was added to the well at a final concentration of 1 µg/ml. In some experiments, whole PBMCs were cultured with 100 ng/ml anti-CD3 mAb for up to 5 days. Where indicated, the following recombinant cytokines, neutralizing antibodies, or other reagents (from R&D Systems unless otherwise indicated) were added at the start of the culture period: recombinant human (rh) TNFα (10 ng/ml; Biosource, Camarillo, CA, USA), rhIL-27 (10–100 ng/ml), rhIL-10 (10 ng/ml; rhIL-10 used to generate data in Figure S5 in Supplementary Material from PeproTech, Rocky Hill, NJ, USA), neutralizing/blocking antibodies to IFNγ (clone 25718, 10 µg/ml), IL-17 (clone 41809, 10 µg/ml), IL-1R1 (polyclonal, 2 µg/ml), IL-10 (clone 23738, 10 µg/ml), IL-10R (clone 8516, 10 µg/ml), and IL-27 (polyclonal, 5 µg/ml). The following isotype control antibodies were used at an assay-appropriate concentration: human IgG1 (Abcam, Cambridge, UK), mouse IgG1, mouse IgG2a, mouse IgG2b, goat IgG (all R&D Systems). Clinical grade biologics were purchased via Guy’s Hospital Pharmacy. Adalimumab (Abbott Laboratories, Chicago, IL, USA) was frozen in aliquots at −20°C and used at final concentration of 1 µg/ml unless otherwise indicated, tocilizumab (Roche, Welwyn Garden City, UK) was used at 50 µg/ml, and abatacept (Bristol-Myers Squibb, New York, NY, USA) was used at 5 µg/ml. Pediacel [Diphtheria, tetanus, pertussis (acellular, component), poliomyelitis (inactivated), and Haemophilus type b conjugate vaccine] and Revaxis [Diphtheria, tetanus and poliomyelitis (inactivated) vaccine] (gifts of James Reading, Tim Tree lab, King’s College London) were added at a 1:1,000 dilution.

To assess intracellular cytokine expression after cell culture, cells were restimulated for 3 h in the presence of phorbol 12-myristate 13-acetate (PMA; 50 ng/ml, Sigma-Aldrich), ionomycin (750 ng/ml, Sigma-Aldrich), and GolgiStop (according to the manufacturer’s instructions; BD Biosciences). In some experiments, cells were labeled with a fixable viability dye (eFluor506 or eFluor780, eBioscience, San Diego, CA, USA; LIVE/DEAD fixable dead cell stains, ThermoFisher Scientific) according to manufacturer’s instructions. Cells from cocultures were surface stained using anti-CD14 APC-Cy7 (HCD14; BioLegend) or anti-CD14 APC Vio770 (TUK4; Miltenyi Biotec) and in some experiments, anti-CD8 Pacific Blue (RPA-T8), before fixing in 2% paraformaldehyde (Merck & Co., Inc., Kenilworth, NJ, USA). Cells were washed and then permeabilized with 0.5% saponin (Sigma-Aldrich) and intracellularly stained with various combinations of the following fluorescently conjugated antibodies (all from BioLegend): anti-CD3 PE Cy7 (UCHT1), anti-CD4 PerCP Cy5.5 or Pacific Blue (both SK3), anti-IL-10 Alexa Fluor 488 (JES3-9D7), anti-IL-17 PE (BL168), anti-IFNγ PerCP Cy5.5, Pacific Blue or APC (all 4 S.B3), anti-TNFα APC (MAb11), GM-CSF APC (BVD2-21C11). Antibodies to detect CD4 and/or CD3 were added intracellularly, to allow for staining of these markers following our T cell culture conditions. Stained cells were acquired using a FACSCantoII or LSRFortessa (BD Biosciences); in most experiments, 100,000 T cell events were recorded. The gating strategies used to analyze intracellular cytokine expression in CD4⁺ T cells are described in Figure S1 in Supplementary Material. All flow cytometry data were analyzed using FlowJo software (version 7.6.1 or 10, Tree Star, Inc., Ashland, OR, USA).

IL-10 was measured in cell culture supernatants using the IL-10 ELISA MAX kit (BioLegend), according to manufacturer’s instructions. IL-27 was measured in cell culture supernatants using the DuoSet Human IL-27 ELISA (eBioscience), according to manufacturer’s instructions. Microwell absorbance was read at 450 nm using a Wallac 1420 microplate reader (Perkin Elmer, Waltham, MA, USA). Concentrations of the analytes were determined based on the standard curve included on each plate.

Statistical testing was performed with GraphPad Prism 5.0 or 6.0 (GraphPad, San Diego, CA, USA). Data sets were tested for normality using the D’Agostino and Pearson omnibus normality test, followed by statistical significance testing using the appropriate tests as indicated in figure legends. Data sets with n values <8 were tested non-parametrically. P values < 0.05 were considered statistically significant.

To investigate whether TNF blockade regulates IL-10 expression in different pro-inflammatory cytokine-producing CD4⁺ T cells, we isolated CD4⁺ T cells from PB of healthy donors and cocultured the cells with CD14⁺ monocytes and anti-CD3 mAb (100 ng/ml) in the absence or presence of the anti-TNF mAb adalimumab (1 µg/ml), as previously described (20). After 3 days, cells were restimulated with PMA and ionomycin in the presence of GolgiStop for 3 h and stained for cytokine expression (gating strategy shown in Figure S1 in Supplementary Material). In agreement with our previous data (20), TNF blockade led to a significant increase in IL-10⁺ cells within total CD4⁺ T cells as well as in the IL-17⁺ CD4⁺ T cell subset. In addition, we found a strong increase in IL-10-expressing cells within IFNγ⁺, TNFα⁺, GM-CSF⁺, and IL-4⁺ CD4⁺ T cell populations (Figures 1A,B). TNF blockade did not alter the frequencies of IFNγ⁺, TNFα⁺, or GM-CSF⁺ CD4⁺ T cell populations but did induce a modest increase in IL-17⁺ CD4⁺ T cells, as previously shown (20), and in most donors in IL-4⁺ frequencies, although this did not reach statistical significance (Figures S2A, B in Supplementary Material). Addition of an isotype control human IgG1 mAb did not promote IL-10 expression in CD4⁺ T cells (Figure S3 in Supplementary Material).

Contrary to the effects of TNF blockade, addition of rhTNFα to cocultures of CD4⁺ CD45RO⁺ T cells, monocytes and anti-CD3 mAb led to a striking decrease in the percentage of IL-10⁺ cells within total CD4⁺ T cells and within IL-17⁺, IFNγ⁺, and TNFα⁺ CD4⁺ T cell subsets (GM-CSF⁺ and IL-4⁺ CD4⁺ T cells were not tested) (Figure 1C). TNFα addition did not significantly alter the frequencies of IL-17⁺, IFNγ⁺, or TNFα⁺ CD4⁺ T cell subsets (Figure S2C in Supplementary Material).

Since pro-inflammatory cytokine-expressing CD8⁺ T cells also contribute to immune-mediated inflammatory diseases (27–29), we investigated whether TNF blockade can regulate IL-10 expression in CD8⁺ T cells. We adapted the culture system to stimulate whole PBMC with anti-CD3 mAb (100 ng/ml) in the absence or presence of a dose range of adalimumab (0.01–10 µg/ml). After 3 days, the cultures were restimulated with PMA/ionomycin in the presence of Golgistop and cytokine expression was assessed within either the CD4⁺ or the CD8⁺ T cell populations. Significant increases in the percentages of IL-10⁺ cells within both CD4⁺ and CD8⁺ T cell populations were observed, including in the IL-17⁺ and IFNγ⁺ subpopulations (Figure 2). The regulation of IL-10 expression by TNF blockade was dose responsive in both CD4⁺ and CD8⁺ T cell subsets, with significantly increased IL-10⁺ frequencies following culture with either 1 or 5 µg/ml adalimumab (Figure 2C). These doses reflect the clinically effective trough levels for adalimumab in human serum following treatment (30, 31). In agreement with data from the CD4⁺ T cell/monocyte cocultures, addition of rhTNFα (100 ng/ml) to anti-CD3-stimulated PBMC led to a reduction in IL-10⁺ cell frequencies, in both CD4⁺ and CD8⁺ T cells (n = 3, data not shown).

Since these in vitro studies of T cell responses to TNF blockade all relied on monoclonal anti-CD3 stimulation, we tested whether TNF blockade could still elicit increased IL-10⁺ T cell responses in the context of a more physiological antigenic stimulation. Pediacel is a pentavalent vaccine consisting of purified diphtheria toxoid, purified tetanus toxoid, acellular pertussis vaccine, inactivated poliovirus, and Haemophilus influenzae type b polysaccharide. Revaxis is a booster vaccine containing purified diphtheria and tetanus toxoids and inactivated poliovirus. Each vaccine was added to CD4⁺ T cell/monocyte cocultures (1:1,000 dilution) in the presence or absence of TNF blockade. After 6 days, antigen-stimulated CD4⁺ T cells cultured in the presence of adalimumab demonstrated elevated IL-10⁺ frequencies as compared to cells that were not exposed to TNF blockade (Figure S4 in Supplementary Material).

Together these data demonstrate that in vitro TNF blockade provided at a physiological concentration and in a physiological setup can promote IL-10 expression in CD4⁺ and CD8⁺ T cells, including in subsets expressing pro-inflammatory cytokines.

Thus far, assessment of cytokine expression following in vitro TNF blockade was carried out after 3 days of culture. Time course experiments were performed to assess the kinetics of IL-10 expression in CD4⁺ and CD8⁺ T cells. PBMC were cultured with anti-CD3 mAb in the absence or presence of adalimumab, and the frequencies of IL-10 expressing cells were analyzed 4–120 h after stimulation. The data show that early after TCR stimulation (18–24 h) the frequencies of IL-10 expressing cells within CD4⁺ and CD8⁺ T cells increased rapidly irrespective of the presence of TNF blockade. However, at later time points (from 42 to 120 h), higher frequencies of IL-10⁺ cells were sustained within CD4⁺ and CD8⁺ T cells in the presence of TNF blockade (Figure 3). These experiments suggest that TNF blockade maintains, rather than directly induces, an IL-10⁺ phenotype in CD4⁺ and CD8⁺ T cells following TCR stimulation.

Our previous work demonstrated that in addition to adalimumab, other TNFα inhibitors (etanercept, infliximab, or certolizumab) as well as TNFR1/2 blocking mAbs were capable of increasing frequencies of IL-10-expressing IL-17⁺ CD4⁺ T cells (20). We investigated whether blockade of additional pro-inflammatory pathways could promote IL-10 expression in CD4⁺ T cells. Blockade of IL-17A did not enhance the frequencies of IL-10⁺ cells in any of the CD4⁺ T cell populations tested (Figure 4A). Blockade of IFNγ did not affect the percentage of IL-10⁺ cells within total CD4⁺ T cells, or within IFNγ or TNFα⁺ subpopulations, but led to modestly increased frequencies of IL-10⁺ expressing cells within the IL-17⁺ population, although this effect was much weaker than the effect of TNF blockade in parallel cultures (Figure 4A). Addition of tocilizumab (IL-6R blockade) or abatacept (CTLA4-Ig, which blocks CD80/CD86-mediated co-stimulation), both of which are biologic drugs routinely used in the clinic to treat rheumatoid arthritis, did not increase IL-10⁺ frequencies in CD4⁺ T cells, IL-17⁺, IFNγ⁺, or TNFα⁺ CD4⁺ T cell subpopulations (Figure 4B). To determine whether blockade of these pathways might regulate IL-10 expression with different kinetics to TNF blockade, IL-10⁺ frequencies were analyzed within both CD4⁺ and CD8⁺ T cells at different time points in anti-CD3-stimulated PBMC cultures exposed to these antibodies or drugs. IL-10⁺ CD4⁺ and IL-10⁺ CD8⁺ T cell frequencies were not regulated at any time point by blockade of IL-17, IFNγ, IL-6R, or CD80/CD86-mediated co-stimulation (Figure S5 in Supplementary Material). Blockade of IL-1R1 in CD4⁺ T cell/monocyte cocultures resulted in a significantly increased proportion of IL-10⁺ cells within total CD4⁺ T cells and within IL-17⁺, IFNγ⁺, or TNFα⁺ subpopulations (Figure 4B). However, this effect was not replicated in either CD4⁺ or CD8⁺ T cells in whole PBMC cultures (Figure S5 in Supplementary Material), indicating that the capacity of IL-1 blockade to regulate IL-10 expression may be dependent on the in vitro culture conditions. Together these data indicate that IL-10 expression in CD4⁺ T cells and CD8⁺ T cells can be regulated by blocking TNFα signaling, but not by blocking IFNγ, IL-17, IL-6R, or CD80/CD86-mediated co-stimulation, at least in vitro.

Since CD14⁺ monocytes are major producers of TNFα, we explored whether the presence of monocytes was required for the effects of TNF blockade on regulating IL-10. CD4⁺ T cells were sorted to a very high purity (>99%) and stimulated with plate-bound anti-CD3 and soluble anti-CD28 mAb in the absence of monocytes or placed in coculture with monocytes and anti-CD3 mAb, with or without addition of anti-TNF mAb. In the CD4⁺ T cell only cultures, TNF blockade still brought about increased IL-10⁺ cell frequencies within the total CD4⁺ population and also within IL-17⁺, IFNγ⁺, or TNFα⁺ subpopulations, similar to the increased IL-10⁺ frequencies observed in CD4 T cell/monocyte cocultures (Figures 5A,B). Analysis of supernatants from anti-CD3/CD28-stimulated CD4⁺ T cells by ELISA confirmed that levels of secreted IL-10 were increased during culture in the presence of TNF blockade (Figure 5C).

One potential interpretation of the observed increased frequencies of IL-10⁺ cells following TNF blockade could be that there is an expansion of a pre-existing population of Tregs. Our previous work indicated that in three donors, MACS-depletion of CD4⁺ CD25⁺ cells did not impair the anti-TNF-mediated increase in IL-10⁺ frequencies within IL-17⁺ CD4⁺ T cells and that no increase in FOXP3⁺ cell frequencies was apparent upon TNF blockade (20). To investigate this further, MACS-isolated CD4⁺ T cells were FACS sorted to very high purity (>99%) to yield either total CD4⁺ T cells or effector CD4⁺ CD25-CD127⁺ T cells (Teff; depleted of CD25^high CD127^low Treg). These cells were then cultured with anti-CD3/CD28 mAbs in the absence of monocytes, or placed in coculture with monocytes and anti-CD3 mAb, in the absence or presence of adalimumab. Our data show that TNF blockade still resulted in increased frequencies of IL-10-expressing cells in effector T cells in the absence of CD25^high CD127^low Tregs (Figure 6). Together these data indicate that TNF blockade regulates IL-10 expression directly in CD4⁺ effector T cells and does not require the presence of CD14⁺ monocytes or CD4⁺ CD25^highCD127^low Tregs.

In support of our findings that TNF blockade can exert its effects on CD4⁺ T cells in the absence of monocytes, we did not find evidence for a role of the monocyte-derived anti-inflammatory mediator IL-27 (32) in mediating IL-10 expression in the context of TNF blockade. This was demonstrated by the findings that TNF blockade did not induce IL-27 secretion, rhIL-27 enhanced IL-10 secretion but did not result in increased percentages of IL-10⁺ CD4⁺ T cells, and IL-27 blockade did not impair the increased IL-10⁺ CD4⁺ T cell frequencies brought about by TNF blockade (Figure S6 in Supplementary Material).

Finally, we explored whether modulation of IL-10 expression in CD4⁺ T cells by TNF blockade was dependent on IL-10 signaling itself. Neutralizing antibodies against IL-10 and IL-10R were added to CD4⁺ T cell/monocyte cocultures stimulated with anti-CD3 mAb, in the presence or absence of TNF blockade, and intracellular cytokine expression was assessed after 3 days. Blockade of IL-10 signaling in the absence of adalimumab led to significantly reduced IL-10⁺ frequencies within total CD4⁺ T cells and within IL-17⁺, IFNγ, or TNFα⁺ subpopulations (Figure 7A). When adalimumab and IL-10/IL-10R blocking mAbs were added in combination to the cocultures, a decrease in IL-10⁺ CD4⁺ T cell frequencies was observed as compared to the condition treated with adalimumab alone; however, an increase was still observed as compared to IL-10/IL-10R blockade alone. These data suggest that pathways additional to IL-10 signaling could play a role in the regulation of IL-10 expression by TNF blockade. Indeed, when we tested whether addition of exogenous IL-10 (10 ng/ml) could drive IL-10 expression in CD4⁺ T cell/monocyte cocultures, we found that this by itself was not sufficient to increase IL-10⁺ cell frequencies in total CD4⁺ T cells, or within IL-17⁺, IFNγ⁺, or TNFα⁺ subpopulations. Addition of rhIL-10 even decreased the percentage of IL-10⁺ cells in total CD4⁺ T cells and TNFα⁺ CD4⁺ T cells (Figure 7B). The biological activity of the recombinant IL-10 was validated by culturing freshly isolated monocytes with 0, 10, or 100 ng/ml rhIL-10 for 20 h and confirming that rhIL-10 addition resulted in increased CD14 and CD163 and reduced HLA-DR expression by flow cytometry, in accordance with previous studies (33, 34) (n = 2, data not shown). We next investigated the regulation of IL-10 in response to rhIL-10 in both CD4⁺ and CD8⁺ T cells using whole PBMC cultures assessed for IL-10 expression over a 5-day time course (Figure S5 in Supplementary Material). We found that at earlier time points (18–24 h), rhIL-10 addition reduced IL-10⁺ CD4⁺ T cell frequencies (n = 6, p = 0.03, Wilcoxon matched-pairs signed rank test), while at later time points (66–72 and 114–120 h), rhIL-10 slightly increased IL-10⁺ CD4⁺ T cell frequencies (n = 6, p = 0.03, Wilcoxon matched-pairs signed rank test). IL-10⁺ CD8⁺ T cell frequencies remained very low throughout, regardless of rhIL-10 addition. Together, these data indicate that IL-10 signaling contributes in part to TNF blockade-mediated regulation of IL-10 expression in CD4⁺ T cells, but that other factors also play a role. The effects of exogenous IL-10 on regulating IL-10 expression in CD4⁺ T cells appear to be dependent on both time and in vitro culture conditions.