IL-6 promotes T cell proliferation and expansion under inflammatory conditions in association with low-level RORγt expression

Abstract

IL-6 is a critical driver of acute and chronic inflammation, and has been reported to act as a T cell survival factor. The influence of IL-6 on T cell homeostasis is not well resolved. We demonstrate that IL-6 signaling drives T cell expansion under inflammatory conditions, but not during normal homeostasis. During inflammation, IL-6Rα-deficient T cells are unable to effectively compete with wild type T cells. IL-6 promotes T cell proliferation, and this is associated with low-level expression of the RORγt transcription factor. T cells upregulate Rorc mRNA at levels substantially diminished from that seen in Th17 cells. Blockade of RORγt through genetic knock-out or a small molecule inhibitor leads to T cell expansion defects comparable to those in IL-6Rα-deficient T cells. Our results indicate that IL-6 plays a key role in T cell expansion during inflammation, and implicates a role for the transient induction of low-level RORγt.

INTRODUCTION

Interleukin 6 (IL-6) is a pleiotropic cytokine that is a key mediator of the acute phase response and facilitates inflammatory responses (1–3). It is produced by T cells, B cells, macrophages, dendritic cells, and stromal and epithelial cells (4). While physiologic levels of serum IL-6 are normally low, IL-6 is rapidly induced in monocytes upon recognition of TLR agonists such as LPS (5, 6). Elevated IL-6 levels have been found in patients with rheumatoid arthritis, psoriasis, inflammatory bowel disease, Crohn’s disease, and other chronic inflammatory disorders (7–9).

The IL-6 receptor is a heterodimer formed by IL-6Rα (CD126) and the shared cytokine receptor gp130 (CD130) (10). Upon binding to IL-6, IL-6Rα forms a complex with gp130 that initiates signal transduction (11). IL-6Rα itself has no intrinsic signaling capacity, while gp130 activates JAK kinases and induces STAT1 and STAT3 phosphorylation (12, 13). Whereas gp130 is ubiquitously expressed, IL-6Rα expression is restricted to hepatocytes, megakaryocytes, and subsets of leukocytes, including dendritic cells, macrophages, and T cells (14, 15). IL-6Rα is highly expressed on naïve T cells (16) and IL-6 signaling drives the differentiation of Th17 T cells, a T helper subset characterized by the production of IL-17A (17–19). Retinoic acid (RA)–related orphan receptor gamma t (RORγt), encoded by Rorc, is the master transcriptional regulator defining the Th17 lineage (20), and IL-6 mediated STAT3 activation is important for RORγt-associated Th17 proliferation and survival (21).

IL-6 has been reported to promote T cell survival and limit apoptosis. IL-6 prevents activation induced cell death by downregulating Fas and FasL in an IL-2 independent manner (22, 23). It further supports T cell survival through the maintenance of anti-apoptotic factors (24–27). The effect of IL-6 on T cell proliferation is less clear. Early IL-6 signaling drives expansion and cytokine production of primed or memory CD4⁺ T cells (28, 29). Mice that overexpress IL-6 have increased peripheral T cells, although this is not due to activation induced proliferation, as the cells remain quiescent and do not exhibit any change in activation/memory phenotype (27). Blockade of IL-6 signaling in T cells has been shown to result in a proliferative defect in naïve T cells, particularly under lymphopenic conditions (30, 31). However, other studies have indicated that enhanced T cell survival in the presence of IL-6 is mediated by reduced apoptosis and not through increased proliferation (25). Thus, the direct effects of IL-6 on T cell proliferation, and any mechanism by which this may occur remain unclear.

We previously reported that, in contrast to wild type (WT) T cells, IL-6Rα-deficient T cells (IL-6Rα^Tdel) do not induce colitis upon transfer into lymphodeficient Rag1^−/− mice. This was associated with failure of the IL-6Rα^Tdel T cells to effectively expand (32). Our findings indicated that IL-6 is broadly necessary for T cell expansion during colitis, but did not identify the specific roles of lymphodeficiency and inflammation as drivers of this. We show here that IL-6Rα-deficient T cells are unable to effectively compete with WT T cells under lymphopenic and inflammatory conditions. This is due to inflammation induced IL-6 signaling into the cells, and not the lymphopenic environment. We correlated this defect with reduced expression of Ki67 and RORγt. IL-6Rα-deficient T cells isolated from inflammatory conditions expressed minimal levels of Rorc mRNA. In contrast, WT T cells upregulated Rorc, though to levels substantially lower than that observed in Th17 T cells. Blockade of RORγt, either by genetic knock-out or small molecule inhibitor, revealed a proliferative defect paralleling that seen in IL-6Rα-deficient T cells. Taken together our data indicate that while robust RORγt expression promotes and maintains Th17 differentiation, IL-6 signaling can also more broadly induce low level expression of RORγt that is correlated with T cell proliferation and expansion under inflammatory conditions.

MATERIALS AND METHODS

Mice:

B6.SJL-Ptprca Pepcb/BoyJ (CD45.1), B6.PL-Thy1a/CyJ (Thy1.1), B6.129S7-Rag1^tm1Mom/J (Rag1^−/−), and B6.SJL-Il6ra^tm1.1Drew/J (IL-6Rα^fl/fl) mice were purchased from The Jackson Laboratory. IL-6Rα^fl/fl mice possess loxP sites flanking exons 4–6 of the interleukin 6 receptor alpha (Il6ra) gene. IL-6Rα^fl/fl mice were bred with B6.Tg(Cd4-Cre)1Cwi/BfluJ (CD4 cre) mice (H. Chi, SJCRH). The resulting IL-6Rα^fl/fl CD4cre⁺ mice are referred to as IL-6Rα^Tdel mice. All mice were maintained under specific pathogen free conditions, including negative for detectable Helicobacter spp. Experimental protocols were approved by the St. Jude Children’s Research Hospital Animal Care and Use Committee.

T cell transfer model:

Naïve CD4⁺CD25⁻CD45RB^hi T cells were sorted from spleens and 1.0×10⁶ cells were transferred i.v. into Rag1^−/− mice. Body weight was monitored weekly. For IL-6Rα blockade, 1 mg anti-IL-6Rα mAb (clone 15A7, Bio-X-Cell) or control IgG was administered i.p. For RORγt inhibition, equal numbers (5×10⁵) of indicated T cells were transferred into Rag1^−/− mice. Three weeks later, mice were treated with the inhibitor, GSK805, or a vehicle control daily for one week. GSK805 was purchased from EMD Milipore with a purity ≥ 95% by HPLC. Mice received 10 mg/kg/day of GSK805 dissolved in DMSO by oral gavage in corn oil, or DMSO alone in corn oil as a vehicle control (33, 34). For co-transfers with Tregs, 1×10⁶ naïve CD4⁺CD25⁻CD45RB^hi T cells mixed with 0.2×10⁶ CD4⁺CD25⁺CD45RB^lo Treg cells from WT (CD45.1) and IL-6Rα^Tdel (CD45.2) mice were co-transferred together at a 1:1 ratio into Rag1^−/− mice. Organs were harvested 3 weeks later and cells were distinguished by CD45.1/CD45.2 expression.

LPS stimulation:

Naïve CD4⁺CD25⁻CD45RB^hi T cells were sorted from spleens of IL-6Rα^Tdel and WT mice. Cells were mixed at a ratio of 3:1.and 1×10⁶ total cells per mouse were transferred i.v. into Rag1^−/− mice. After 3 weeks, 1 mg/kg LPS or PBS was given i.v. Organs were harvested and analyzed 1 week after LPS treatment. For IL-6Rα blockade, 1 mg anti-IL-6Rα mAb (clone 15A7, Bio-X-Cell), which has been shown to neutralize IL-6 signaling in vivo (35) or control IgG was administered per mouse i.p. concurrent with LPS or PBS treatment. To test non-lymphopenic conditions, naïve CD4⁺ T cells from WT (CD45.1⁺Thy1.2⁺) and IL-6Rα^Tdel (CD45.2⁺Thy1.2⁺) mice were co-transferred into Thy1.1 mice at a 1:3 ratio (1×10⁷ total cells/mouse). LPS (1 mg/kg) or PBS was administered i.v. 4 days after transfer. Organs were harvested 1 week after LPS treatment.

IL-6 concentration:

Blood samples were obtained from retro-orbital puncture and immediately centrifuged at 4,000 rpm for 15 minutes. Serum samples were stored at −20°C until ready to be used. Serum levels of IL-6 were either measured by Luminex (Bio-Rad), or with a mouse IL-6 sandwich ELISA kit (eBioscience) according to manufacturer instructions.

Colon T cell isolation:

Colon segments were vigorously shaken in medium with 1 mM EDTA (Sigma-Aldrich) for 20 min at 37°C, and suspended cells collected and filtered through a cell strainer. Tissue was further minced and incubated at 37°C for 1 h in medium with 1 mM collagenase type IV (Sigma-Aldrich) and 40 U/ml DNase I (Roche) with agitation. Cells were filtered, washed, and isolated over a percoll step gradient (32).

Flow cytometry:

Cells were surface stained in FACS buffer (PBS with 2% FBS) as indicated with the following fluorescently labeled mAbs specific for mouse: TCRβ-FITC (clone H57–597) CD45.2-FITC (clone 104), CD44-FITC (clone IM7), CD8-Pacific Blue (clone 53–6.7), and CD62L-PE-Cy7 (clone MEL-14) from Biolegend; CD4-APC (clone RM4–5) from Tonbo; CD45.1 APC-C7 (clone A20), and CD69-PE (clone H1.2F3) from BD Biosciences; Thy1.2-PE (clone 30-H12), and CD45RB-PE (clone C363.16A) from ebioscience. For intracellular staining, cells were restimulated with cell stimulation cocktail containing Brefeldin A (eBioscience) for four hours at 37°C, washed with FACS buffer, and surface stained at room temperature for 30 minutes. Cells were fixed and permeabilized with the intracellular fixation and permeabilization buffer set (ebiosceince) and stained with IL-17A-FITC (clone TC11–18H10.1, biolegend), IFN-γ-APC (clone XMG1.2, biolegend), RORγt-PE (clone AFKJS-9, ebioscience), and Ki67-PerCP-Cy5.5 (clone B56, BD Biosceinces). For Foxp3 staining, cells were surfaced stained as above, followed by fixation, permeabilization, and staining with Foxp3-PE (clone FJK-16s, ebioscience) using the Foxp3 transcription factor buffer kit (ebioscience) according to manufacturer instructions. All samples were run on an LSRFortessa or LSRII flow cytometer (BD Biosciences) and data was analyzed using Cell Quest (BD Biosciences) or FlowJo (TreeStar) software.

Proliferation and cell death assays:

For BrdU, γH2AX, and cleaved-PARP staining, mice were injected i.p. with 100μl of 10 mg/ml BrdU in PBS 24 hours before analysis. Samples were prepared and stained using the Apoptosis, DNA Damage, and Cell Proliferation kit from BD Biosciences according to instructions. Briefly, samples were surface stained as described above for 30 minutes at room temperature. Cells were then fixed with BD Cytofix/Cytoperm solution for 30 minutes at room temperature, washed with BD Perm/Wash buffer, and treated with BD Cytofix/Cytoperm Plus Permeabilization buffer for 10 minutes on ice. Cells were washed with BD Perm/Wash buffer, re-fixed for 5 minutes at room temperature. Samples were then treated with 300 μg/ml DNAse for 1 hour at 37°C, to expose incorporated BrdU. Cells were then stained with anti-BrdU PerCP-Cy5.5 (clone 3D4), anti-H2AX(pS139)-Alexa Fluor 647 (clone N1–431), and anti-cleaved PARP(Asp214)-PE (clone F21–852). For Annexin-V and Propidium Iodide staining, cells were first surface stained as described above. Cells were washed once in PBS and once in Annexin binding buffer (BD biosciences). Cells were stained with Annexin-V-APC (BD biosciences) in binding buffer for 15 minutes at room temperature. Cells were washed with and resuspended in binding buffer. Propidium Iodide Staining Solution (BD biosceinces) was added and samples were assessed by flow cytometry within 2 hours. Active Caspase-3 levels were measured using the Active Caspase-3 Apoptosis Kit (BD Biosciences). Briefly, cells were surface stained as indicated above. Samples were washed twice in cold PBS and resuspended in BD Cytofix/Cytoperm solution and incubated for 20 minutes on ice. Cells were washed twice with BD Perm/Wash and stained with anti-active caspase-3-PE (clone C92–605) for 30 minutes at room temperature. Cells were washed with BD Perm/Wash, resuspended in FACS buffer, and analyzed by flow cytometry.

Intracellular Rorc mRNA detection:

Cells were harvested and mRNA expression of Rorc was analyzed on single cell level by flow cytometry in combination with CD4, TCRβ CD45.1 and RORγt protein staining using FlowRNA II Assay kit (PrimeFlow; Affymetrix/eBioscience) according to manufacturer instructions (36). Briefly, cells were washed with and resuspended in permeabilization buffer. Rorc mRNA was detected using an amplified signal FISH technique as previously described (37). Briefly, target probe hybridization was performed using type 1 (AlexaFluor647) probes for Rorc mRNA (VB1–17129). Splenocytes were incubated for 2 hours with the target probes at 40°C. All samples were then incubated with the PreAmplification (PreAmp) reagent for 2 hours and the Amplification (Amp) reagent for an additional 2 hours at 40°C. After signal amplification, cells were incubated with label probes at 40°C for 1 hour. Cells were washed and resuspended in staining buffer prior to acquisition.

In vitro Th17 culture:

Naïve (CD4⁺CD25⁻CD45RB^hi) cells were isolated by cell sorting to >99% purity. Cells were cultured at 5×10⁵ cells per well in 96-well plates pre-coated with 1 μg/ml anti-CD3 and 2 μg/ml anti-CD28 in complete RPMI 1640 media containing 20 ng/ml IL-6, 5 ng/ml TGF-β, 10 μg/ml neutralizing anti-IL-4 and 10 μg/ml neutralizing anti-IFN-γ. Negative control wells contained all reagents except IL-6. After 4 days, cells were collected for intracellular staining or Rorc mRNA detection (20, 38).

Statistical analysis:

Statistical significance was calculated using Prism5 software (GraphPad). Specific tests used for each experiment are detailed in the figure legends.

RESULTS

IL-6 promotes T cell expansion in a T cell transfer model

To assess the impact of IL-6Rα deficiency on T cell expansion, naïve CD4⁺CD25⁻CD45RB^hi T cells from WT or IL-6Rα^Tdel mice were separately transferred into Rag1^−/− hosts. Organs were analyzed 3 weeks later, prior to the development of the overt inflammatory disease seen with this T cell transfer model. Spleens from IL-6Rα^Tdel recipients were significantly smaller than those of recipients of WT T cells (Fig. 1A). Consistently, the number of CD4⁺ T cells in the spleen, MLN, and colon were dramatically reduced in recipients of IL-6Rα^Tdel T cells relative to WT T cells (Fig. 1B). Thus IL-6Rα is necessary for effective T cell expansion in these lymphopenic conditions.

Figure 1. IL-6 promotes T cell expansion in a T cell transfer model.

(A-B) Rag1^−/− mice received 1×10⁶ naïve CD4⁺CD25⁻CD45RB^hi T cells from WT or IL-6Rα^Tdel mice. Organs were harvested 3 weeks after transfer and analyzed. Significance was determined by unpaired t test. (A) Spleens and plotted ratio of spleen weight/body weight. (B) Absolute numbers of CD4⁺TCRβ⁺ T cells in the spleen, MLN, and colon. (C) Equal numbers (0.5×10⁶) of naïve CD4⁺CD25⁻CD45RB^hi WT (CD45.1⁺) and IL-6Rα^Tdel (CD45.2⁺) T cells were co-transferred i.v. into Rag1^−/− mice. Organs were harvested 3 weeks later. T cells were gated based on CD4 and TCRβ expression, and WT or IL-6Rα^Tdel cells were distinguished based on CD45.1/CD45.2 expression. Representative dot plots and summary data from 5 mice per group are shown. Significance was determined by paired t test. (D) Serum was collected from Rag1^−/− mice prior to T cell transfer, or at the indicated time after co-transfer of WT and IL-6Rα^Tdel T cells. Serum IL-6 concentrations were determined by ELISA. Significance was determined by one-way ANOVA with Tukey’s post-test for multiple comparisons. All data are representative of two or more independent experiments. *, p<0.05; **, p<0.01; ****, p<0.0001.

Transfer of naïve WT T cells causes severe colitis and IBD symptoms that are absent in mice that receive IL-6Rα^Tdel cells (32). To separate the role of T cell-intrinsic IL-6 signaling from broader inflammation-associated changes, we performed co-transfers of naïve WT and IL-6Rα^Tdel T cells, allowing comparison of their expansion in an identical environment. Three weeks after transfer at a 1:1 ratio, WT cells were three-fold more abundant than IL-6Rα^Tdel cells in the spleen and mesenteric lymph nodes (MLN) (Fig. 1C). A similar competitive disadvantage of IL-6Rα^Tdel T cells was seen the colon (Fig. 1C). A significant increase in serum IL-6 was observed in recipient mice after T cell transfer, and this was maintained over the course of the experiment (Fig. 1D). These cotransfers indicate that the reduced proliferative capacity of IL-6Rα-deficient T cells results at least in part from cell intrinsic effects of IL-6 signaling, and not from differences in the inflammatory environments present after alternatively transferring WT or IL-6Rα^Tdel T cells.

Naïve T cells acquire an activated or memory-like phenotype during homeostatic proliferation (39). By 3 weeks after co-transfer, both WT and IL-6Rα^Tdel T cells developed a CD44^hiCD62L^lo phenotype with similar expression of these markers (data not shown). CD69 expression was likewise comparable in WT and IL-6Rα^Tdel cells, indicating that the absence of IL-6 responsiveness did not inhibit T cell activation per se (Supplemental Fig. 1A).

As IL-6 is fundamental for Th17 differentiation, we also analyzed the proportions of Th17 cells in individually transferred WT and IL-6Rα^Tdel cell populations. Not surprisingly, significantly fewer IL-6Rα^Tdel than WT CD4⁺ cells expressed IL-17A (Supplemental Fig. 1B). However, only 3% of WT cells were IL-17A⁺, indicating that impaired Th17 cell differentiation in IL-6Rα^Tdel cells could not account for the observed expansion defect. A modest though significantly diminished percentage of IFN-γ⁺ cells was also observed, implying that IL-6 response deficiency also impacted Th1 differentiation. No difference in Foxp3⁺ regulatory T cell representation was seen (Supplemental Fig. 1B), consistent with prior reports (30). This suggests that differential activation and differentiation is not responsible for the broadly diminished competitiveness of IL-6Rα^Tdel T cells.

Reduced expansion of IL-6Rα^Tdel T cells is due to decreased proliferation and not increased apoptosis

To determine if receptivity to IL-6 led to differential proliferation of CD4⁺ T cells, we assessed WT and IL-6Rα^Tdel cells co-transferred in equal numbers into Rag1^−/− recipients for the expression of proliferation markers at both an early and later timepoint (Fig. 2A). Three days after transfer, the ratio of WT to IL-6Rα^Tdel cells remained approximately 1:1 in the spleen (Fig. 2B). Very few cells were recovered from the MLN and no cells were detected in the colon at this time point, preventing analysis at these sites. In vivo BrdU incorporation was significantly higher in WT cells than in IL-6Rα^Tdel cells recovered from the same recipients (Fig. 2C). Expression of the proliferation maker Ki67 was also significantly increased in WT T cells (Fig. 2D). The basal levels of γH2AX in WT and IL-6Rα^Tdel T cells were also assessed three days after co-transfer. While γH2AX is a marker of DNA damage, increased basal levels are detected as cells progress from G1 to G2/M, making it a useful marker of cell cycle progression (40–42). Basal γH2AX levels were significantly higher in recovered WT than IL-6Rα^Tdel cells (Fig. 2E). Overall, these results are consistent with reduced proliferation in IL-6Rα-deficient T cells at this early time point.

Figure 2. IL-6 signaling promotes proliferation in T cells.

Equal numbers (0.5×10⁶) of naïve CD4⁺CD25⁻CD45RB^hi WT (CD45.1) and IL-6Rα^Tdel (CD45.2) T cells were co-transferred i.v. into Rag1^−/− mice. Organs were harvested 3 days or 3 weeks later as indicated. T cells were gated based on CD4 and TCRβ expression and WT and IL-6Rα^Tdel T cells were distinguished based on CD45.1/CD45.2 expression. (A) Experimental design schematic. (B) Percent of CD4⁺TCRβ⁺ T cells in the spleen 3 days after transfer. (C) Percent BrdU⁺ T cells in the spleen 3 days after transfer. (D) Percent of Ki67⁺ T cells in the spleen 3 days after transfer. (E) Level of γH2AX in T cells 3 days after transfer, shown as MFI. (F) Percent of BrdU⁺ T cells in the spleen, MLN, and colon 3 weeks after transfer. (G) Percent of Ki67⁺ T cells in the spleen, MLN, and colon 3 weeks after transfer. (H) Level of γH2AX in T cells 3 weeks after transfer, shown as MFI. Representative plots and summary data from 4–5 mice per group are shown. Significance was determined by paired t test. *, p<0.05; **, p<0.01.

The same parameters were also assessed three weeks after T cell transfer, when cells had migrated to the MLN and colon, and WT cells were approximately 3-fold more abundant than IL-6Rα^Tdel cells. In vivo BrdU incorporation was significantly higher in WT cells than in IL-6Rα^Tdel cells isolated from the MLN, but not from the spleen or colon (Fig. 2F). More WT than IL-6Rα^Tdel cells were Ki67⁺ 3 weeks after transfer, and this was consistent across all organs (Fig. 2G). Finally, higher basal γH2AX levels were observed in WT cells in the spleen and MLN (Fig. 2H). Together, these data indicate that IL-6 responsiveness confers an increased proliferative capacity, and that although IL-6Rα^Tdel T cells become activated normally under lymphopenic and inflammatory conditions, they fail to proliferate at the same rate as WT T cells.

To determine if IL-6 deficiency led to increased cell death in CD4⁺ T cells, we also assessed WT and IL-6Rα^Tdel cells co-transferred into Rag1^−/− recipients for markers of cell death at early and late timepoints. Three days after transfer, there was no difference in levels of the apoptosis markers active caspase-3 or cleaved PARP between WT and IL-6Rα^Tdel T cells in the spleens of recipient mice (Supplemental Fig. 2A-B). There were also no differences in the percentage of cells undergoing early and late stage apoptosis, as determined by Annexin-V/PI staining, between WT and IL-6Rα^Tdel T cells (Supplemental Fig. 2C). Three weeks after co-transfer, no differences were also observed in any of these parameters in the spleen, MLN, or colon (Supplemental Fig. 2D-F). These results indicated that decreased proliferation, and not increased apoptosis of IL-6Rα^Tdel T cells is responsible for their reduced expansion.

IL-6 signaling is not required for T cell homeostasis under non-lymphopenic and non-inflammatory conditions

We next asked whether IL-6 signaling played a role in T cell expansion under non-lymphopenic and non-inflammatory conditions by comparing IL-6Rα^Tdel and IL-6Rα^fl/fl CD4cre⁻ littermate control mice at 6–8 weeks of age. The representation of CD4⁺ T cells among total T cells was not significantly different in the spleen or MLN of IL-6Rα^Tdel mice relative to littermate controls (Fig. 3A). Naïve and memory CD4⁺ T cell subpopulations were also equivalent in IL-6Rα^Tdel mice and littermate controls (Fig. 3B). The ratio of spleen to body weight in IL-6Rα^Tdel mice and littermate control mice was not significantly different, and the absolute number of splenocytes and total CD4⁺ T cells in the spleen and MLN of IL-6Rα^Tdel mice were comparable to that of control mice (Fig. 3C). This indicates that the absence of IL-6 signaling did not adversely affect the formation of peripheral CD4⁺ T cell populations under non-inflammatory conditions.

Figure 3. IL-6 signaling is not required for T cell homeostasis under non-lymphopenic and non-inflammatory conditions.

(A-C) Flow cytometric analysis of spleen and MLN from 8 week old IL-6Rα^Tdel mice or IL-6Rα^fl/fl CD4 cre^– littermates. T cells were gated based on CD4 and TCRβ expression. Representative dot plots and summary data from 5 mice per group are shown. Significance was determined by unpaired t test. (A) Percent of CD4⁺ T cells. (B) Percent of naïve (CD44^loCD45RB^hi) and memory (CD44^hiCD45RB^lo) CD4⁺ T cells. (C) From left to right, ratio of spleen weight/body weight, absolute number of splenocytes, and absolute CD4⁺ T cell numbers in the spleen and MLN. (D) Equal numbers (0.5×10⁶) of naïve CD4⁺CD25⁻CD45RB^hi WT (CD45.1⁺) and IL-6Rα^Tdel (CD45.2⁺) T cells were co-transferred i.v. into C57BL/6 mice congenic for Thy1.1. Spleens were harvested 2 weeks after transfer. Percent of WT or IL-6Rα^Tdel T cells among total CD4⁺TCRβ⁺Thy1.2⁺ T cells in the spleen were identified by CD45.1 and CD45.2 expression. Experimental schematic, representative dot plots and summary data from 5 mice are shown. Significance was determined by paired t test. All data are representative of at least two independent experiments.

To further investigate the role of IL-6 signaling in T cell persistence in lymphoid replete mice, we co-transferred equal numbers of naïve CD45.1 WT and CD45.2 IL-6Rα^Tdel T cells (Thy1.2⁺) into Thy1.1 congenic hosts. In contrast to results with lymphopenic Rag1^–/– hosts, the 1:1 ratio of transferred WT:IL-6Rα^Tdel T cells was maintained two weeks after transfer (Fig. 3D). Therefore IL-6 does not appear to have a detectable role in T cell expansion or persistence in lymphoid replete, disease-free animals.

LPS-induced IL-6 promotes T cells expansion

Although serum concentrations of IL-6 are normally low or undetectable, they rapidly rise with inflammation, and in particular the profound inflammation that occurs with conditions such as septic shock or hepatitis (2, 43, 44). Treatment of mice with LPS mimics sepsis and transiently induces IL-6, resulting in serum levels that peak after 3–6 hours and return to basal level within 24 hours (45). We observed a 3-fold greater expansion of WT T cells relative to IL-6Rα^Tdel T cells three weeks after co-transfer of equal numbers into Rag1^–/– mice (Fig. 1C). We hypothesized that LPS treatment, through the induction of elevated levels of IL-6, would further promote expansion of WT T cells.

Naïve T cells from WT or IL-6Rα^Tdel mice were mixed at a ratio of 1:3 and transferred into Rag1^−/− mice. Because of the accelerated expansion of the WT cells, an approximately 1:1 ratio of the two populations was present three weeks after transfer (Supplemental Fig. 3A). At this time, a non-lethal dose of LPS (1 mg/kg) or saline was administered (Fig. 4A). Serum concentrations of IL-6 increased from nearly undetectable to >3 ng/ml within 6 hours (Supplemental Fig. 3B). One week after this treatment, the ratio of WT:IL-6Rα^Tdel T cells further increased to 2:1 in mice that had received LPS, but was virtually unchanged in mice receiving saline (Fig. 4B and Supplemental Fig. 3C).

Figure 4. LPS-induced serum IL-6 further promotes T cell expansion.

Naïve CD4⁺CD25⁻CD45RB^hi T cells from WT (CD45.1) and IL-6Rα^Tdel (CD45.2) mice were co-transferred into Rag1^−/− mice at a 1:3 ratio (1×10⁶ total cells/mouse). A single dose of LPS (1 mg/kg) or PBS was administered 3 weeks post-transfer. Organs were collected 1 week after LPS treatment. T cells were gated based on CD4 and TCRβ expression, and WT and IL-6Rα^Tdel cells were distinguished by CD45.1/CD45.2 expression. (A) Schematic representation of experimental design. (B) Representative dot plots and calculated ratios of WT:IL-6Rα^Tdel cells from spleen, MLN, and colon after PBS or LPS treatment. Significance was determined by paired t test. (C) Anti-IL-6Rα or control IgG was administered concurrent with LPS or PBS treatment. The ratios of WT:IL-6Rα^Tdel cells in the spleen and MLN are shown, along with representative dot plots. Significance was determined by one-way ANOVA with Tukey’s post-test for multiple comparisons. (D) Percent of CD4⁺Ki67⁺ WT and IL-6Rα^Tdel T cells in spleens of mice treated with LPS or PBS and anti-IL-6Rα or control IgG. Representative dot plots are also shown. Significance was determined by one-way ANOVA with Tukey’s post-test for multiple comparisons. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. Data are representative of two independent experiments.

To confirm that IL-6 was the primary factor promoting T cell expansion after LPS treatment, a blocking anti-IL-6Rα antibody (46–48), or control IgG was administered concurrently with LPS or saline. As expected, the ratio of WT:IL-6Rα^Tdel T cells increased significantly with LPS treatment in the presence of control IgG, but this effect was completely abrogated by neutralizing IL-6 antibody (Fig. 4C). Consistent with prior data indicating that this antibody is not depleting, there was no difference in the ratio of IL-6Rα-expressing WT to IL-6Rα^Tdel cells in PBS treated mice that received control IgG or IL-6Rα blocking antibody. LPS treatment also increased the percent of proliferating WT Ki67⁺ cells, consistent with a role for IL-6 in supporting proliferation, and this effect was blocked by anti-IL-6Rα treatment (Fig. 4D). IL-6Rα^Tdel T cells did modestly increase Ki67 in response to LPS treatment, although to a lesser degree than WT T cells, indicating a lesser role for other LPS-induced mediators. Taken together, these results indicate that intrinsic IL-6 signaling plays a central role in T cell expansion in response to LPS treatment, and that elevated IL-6 levels are responsible.

IL-6-induced inflammation promotes T cell proliferation independent of lymphopenia

Our results indicated that IL-6 signaling promotes T cell proliferation and expansion in an inflammatory and lymphopenic environment, but not in a non-inflammatory and lymphoid sufficient one. We next asked if IL-6 signaling was sufficient to promote T cell expansion in the absence of lymphodepletion. Naïve T cells from WT or IL-6Rα^Tdel mice were mixed at a ratio of 1:3, transferred into lymphoid replete Thy1.1 hosts, and then treated with PBS or LPS as above. Similar to results in Fig. 3D, the PBS treated group maintained the WT:IL-6Rα^Tdel input ratio in the spleen (0.376+0.016) and MLN (0.347+0.020). LPS stimulation of these lymphoid replete mice, however, resulted in an increased ratio in the spleen (1.005+0.075) and MLN (1.084+0.073), indicating that IL-6 induced T cell expansion did not require prior lymphopenia (Fig. 5A).

Figure 5. Inflammation-induced IL-6 and not lymphopenia promotes T cell expansion.

(A) Naïve CD4⁺ T cells from WT (CD45.1⁺Thy1.2⁺) and IL-6Rα^Tdel (CD45.2⁺Thy1.2⁺) mice were co-transferred into Thy1.1 mice at a 1:3 ratio. LPS or PBS was administered 4 days after transfer. Organs were harvested 1 week later and the ratio of WT to IL-6Rα^Tdel cells was determined. Schematic of experimental design, representative dot plots, and summary data from 10 mice pooled from two independent experiments are shown. Significance was determined by paired t test. (B) 1×10⁶ naïve CD4⁺CD25⁻CD45RB^hi T cells and 0.2×10⁶ CD4⁺CD25⁺CD45RB^lo Treg cells were mixed. This mixture from WT (CD45.1⁺) and IL-6Rα^Tdel (CD45.2⁺) mice was co-transferred 1:1 into Rag1^−/− mice. Organs were harvested 3 weeks later and the percent of WT and IL-6Rα^Tdel CD4⁺ cells was determined. Schematic of experimental design, representative dot plots, and summary data from 10 mice pooled from two independent experiments are shown. Significance was determined by paired t test. ****, p<0.0001.

Inflammation in the Rag1^–/– T cell transfer colitis model can be abrogated by co-transfer of regulatory T cells (49). To further isolate the roles of inflammation and lymphopenia in the proliferative deficiency of IL-6Rα^Tdel T cells, naïve and regulatory T cells were mixed at a 5:1 ratio. Mixtures from WT and IL-6Rα^Tdel mice were then co-transferred at a 1:1 ratio into Rag1^–/– mice. Three weeks later, the presence of Treg and resulting curtailed inflammation led to preservation of the 1:1 ratio of WT: IL-6Rα^Tdel cells in the spleen and MLN of recipient mice (Fig. 5B). Taken together, these results indicate that T cell-intrinsic IL-6 signaling promotes T cell proliferation and expansion in the context of inflammation and IL-6 production. Lymphopenia itself does not foster this enhanced proliferation in the absence of inflammation, such as that occurring after naïve T cell transfer.

Low level expression of RORγt is correlated with IL-6 signaling and T cell proliferation

Though IL-6 enhances Th17 proliferation, reduced proliferation of IL-6Rα^Tdel T cells was observed in co-transfers despite the absence of a substantial Th17 presence (Supplemental Fig. 1B). Our results suggested that the effect of IL-6 on T cell proliferation and expansion was independent of its role in promoting Th17 differentiation. However, our findings did not exclude a role for IL-6-induced RORγt, the master transcriptional regulator of Th17 differentiation. We therefore asked whether RORγt expression correlated with expression of the Ki67 proliferation marker in our co-transfer model (Fig. 6A). Three weeks after co-transfer at a 1:1 ratio, fewer IL-6Rα^Tdel T cells were RORγt⁺ than WT T cells, as would be anticipated due the absence of IL-6 signaling (Fig. 6B and C). For either transferred population, the percent of Ki67⁺ T cells was significantly higher within the RORγt⁺ than RORγt^– population in all organs, indicating an association between RORγt expression and proliferation (Fig. 6B and C). This indicates that low-level RORγt expression in T cells is associated with T cell proliferation, and implies that the relative failure of RORγt upregulation in IL-6Rα^Tdel versus WT cells may account for the reduced IL-6Rα^Tdel T cell proliferation and expansion.

Figure 6. Low level expression of RORγt is associated with T cell proliferation.

Naïve CD4⁺CD25⁻CD45RB^hi T cells from WT (CD45.1⁺) and IL-6Rα^Tdel (CD45.2⁺) mice were transferred into Rag1^−/− mice at a 1:1 ratio. Spleens were collected 3 weeks after transfer and CD45.1/CD45.2 expression was used to differentiate WT and IL-6Rα^Tdel cells. (A) Schematic of experimental design. (B) Gating strategy and representative dot plots and (C) summary data showing RORγt and Ki67 expression in WT and IL-6Rα^Tdel T cells from spleen, MLN, and colon. Data from 10 mice pooled from two independent experiments are shown. Significance was determined by one-way ANOVA with Tukey’s post-test for multiple comparisons. (D) Rorc mRNA probe was used to flow cytometrically detect Rorc mRNA in WT and IL-6Rα^Tdel T cells isolated from co-transfers into three Rag1^−/− mice. In vitro differentiated WT Th17 cells and RORγt^−/− T cells were used as positive and negative controls, respectively. The MFI (mean fluorescence intensity) of each group is also plotted. Data are representative of two independent experiments with three mice per group. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.

On a more refined level, we also observed that the percent of Ki67⁺ cells was higher in WT relative to IL-6Rα^Tdel cells among either the RORγt^– or RORγt⁺ populations (Fig. 6C). This suggests either that RORγt cannot fully account for the proliferative discrepancies observed between WT and IL-6Rα^Tdel cells, or that differences in the level of RORγt expression between the cell populations that are not adequately resolved through our flow cytometric assay impacts T cell expansion.

Because intracellular staining of RORγt was weak in the transferred T cells, we also examined differences in Rorc mRNA expression. This was measured using a flow cytometry-based assay that significantly amplifies RNA signal (36). Co-transfers of WT and IL-6Rα^Tdel T cells were performed as described above, and transferred cells were isolated 3 weeks later and stained with a probe specific for Rorc mRNA. In vitro differentiated WT Th17 cells and T cells from RORγt^–/– mice were used as positive and negative controls, respectively. The Th17 cells expressed high Rorc mRNA levels, an order of magnitude greater than that observed in the transferred WT T cells, consistent with the role of RORγt as the primary mediator of Th17 differentiation. WT and IL-6Rα^Tdel T cells from co-transfers expressed intermediate and low levels of Rorc RNA respectively, relative to the RORγt-deficient cells (Fig. 6D). Taken together, these data suggest that IL-6 promotes a range of Rorc expression, with high expression associated with the Th17 lineage, low expression correlated with the enhanced cell proliferation among WT T cells, and further reduced expression in the relatively hypoproliferative IL-6Rα^Tdel T cells.

Inhibition of RORγt blocks T cells expansion

To further address the role of RORγt in T cell expansion, we performed 1:1 co-transfers of WT and RORγt^−/− T cells into Rag1^−/− mice (Fig. 7A). Similar to co-transfers with IL-6Rα^Tdel cells, WT T cells significantly outnumbered RORγt^−/− T cells 2:1 by 3 weeks after transfer (Fig. 7B). The percent of Ki67⁺ RORγt^−/− T cells was also reduced relative to WT T cells in all organs, indicating that RORγt status is associated with Ki67 expression and proliferation (Fig. 7C and Supplemental Fig. 4A). In contrast, there was no significant difference in the percent of activated caspase-3⁺ WT and RORγt^−/− T cells, suggesting that similar to IL-6 receptivity, RORγt expression does not influence apoptosis (Fig. 7D and Supplemental Fig. 4B). Taken together, these results indicate that IL-6 induces low levels of RORγt in transferred WT T cells, and that this IL-6-induced RORγt is associated with T cell proliferation in this setting. These findings do not exclude added effects of other IL-6 induced signaling pathways.

Figure 7. Blockade of RORγt inhibits T cell expansion.

(A-D) Naïve CD4⁺CD25⁻CD45RB^hi T cells from WT (CD45.1⁺) or RORγt^−/− (CD45.2⁺) mice were transferred 1:1 into Rag1^−/− mice. Organs were harvested 3 weeks post-transfer. CD4⁺ T cells were gated based on CD4, TCRβ, and WT and RORγt^−/− T cells were distinguished based on CD45.1/CD45.2 expression. (A) Schematic of experimental design. (B) Percent of WT and RORγt^−/− T cells among total CD4⁺ transferred cells, (C) Percent of Ki67⁺ and (D) percent active caspase-3⁺ WT or RORγt^−/− T cells in the spleen, MLN and colon. Significance was determined by paired t test. (E-G) Naïve CD4⁺CD25⁻CD45RB^hi T cells from WT or IL-6Rα^Tdel mice were transferred into Rag^−/− mice at a 1:1 ratio. After 3 weeks, mice were treated daily with an ROR inhibitor (GSK805). Organs were harvested one week later. (E) Schematic of experimental design. (F) Representative dot plots and summary data showing the ratio of WT to IL-6Rα^Tdel cells. Significance was determined by paired t test. (G) Percent Ki67⁺ WT or IL-6Rα^Tdel cells after treatment with RORγt inhibitor or DMSO (vehicle control). Significance was determined by one-way ANOVA with Tukey’s post-test for multiple comparisons. All data are representative of two independent experiments with 5 mice per group. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001.

As an additional approach to confirm the role of RORγt in T cell expansion, we performed co-transfer experiments using an RORγt inhibitor, GSK805. Naïve WT and IL-6Rα^Tdel T cells were co-transferred at a 1:1 ratio into Rag1^−/− mice. Three weeks later, mice were treated with inhibitor or a vehicle control daily for one week (Fig. 7E). As expected, mice that did not receive inhibitor had 2–3 fold more WT T cells than IL-6Rα^Tdel cells. After one week of inhibitor treatment, this effect was partially abrogated, resulting in ratios of approximately 1.5:1 of WT:IL-6Rα^Tdel cells (Fig. 7F). Importantly, RORγt inhibitor treatment reduced the percent of Ki67⁺ WT cells such that there was no difference between WT and IL-6Rα^Tdel cells (Fig. 7G and Supplemental Fig. 4C). Therefore inhibition of RORγt protein reduces the proliferative capacity of WT T cells. Cumulatively, these results identify low-level expression of IL-6-induced RORγt as a novel factor that is associated with T cell proliferation and expansion.

DISCUSSION

IL-6 can enhance T cell survival and proliferation (25), though the impact of these effects and the mechanisms by which they occur are unclear. In this study, we show that IL-6-induced RORγt is associated with T cell proliferation and expansion specifically under inflammatory conditions. T cell intrinsic IL-6 signaling induces low-level expression of RORγt which is associated with T cell proliferation, and this appears to be independent from the role of RORγt in Th17 programing. Blockade of RORγt, either by inhibition or genetic deletion, reduces proliferation and expansion toward levels observed in T cells unable to respond to IL-6, suggesting that IL-6 induced upregulation of RORγt is largely responsible for IL-6’s effects. Genetic deletion of RORγt was limited to T cells in our transfer model, ensuring that RORγt expression by innate lymphoid cells and their potential effects on T cell expansion were unaltered. Further, the use of co-transfers of WT and deficient T cells permitted the direct comparison the cell intrinsic effects of RORγt on T cells in identical environments.

Our prior observations demonstrating an obligate role for IL-6 responsiveness in T cell expansion and colitis induction in lymphoid deficient Rag1^–/– mice led us to further explore the role of IL-6 here. IL-6 induced RORγt promotes Th17 cell development and is necessary for colitis. Inflammation induced by the transfer of naïve T cells into lymphopenic mice is dependent on pathogenic Th17 cells and is RORγt dependent (50). The failure of transferred IL-6Rα^Tdel cells to induce colitis may have resulted from diminished Th17 development or defective T cell expansion (32). IL-17 producing cells constitute only 15–20% of T cells in the colon lamina propria and MLN in T cell transfer colitis models at 8 weeks after transfer when disease pathology is maximal (51). However, we observe here expansion defects in IL-6Rα^Tdel cells even 3 weeks after transfer when no overt disease symptoms are present and only 3% of T cells in the colon are IL-17⁺. More significantly, IL-6Rα^Tdel T cells fail to effectively compete with co-transferred WT T cells despite effective Th17 formation by the WT T cells and development of colitis at late time points in this co-transfer model (not shown). Therefore, IL-6 responsiveness is necessary for effective T cell expansion after transfer into lymphopenic animals, and the competitive disadvantage of IL-6 unresponsive T cells persists even when Th17 induced inflammation is restored by the cotransfer of WT T cells.

IL-6-induced RORγt also inhibits the conversion of T cells to Foxp3⁺ regulatory cells and increases Th17 differentiation (3, 52, 53). It is possible that differential expression of RORγt in WT and IL-6Rα^Tdel cells may influence their ability to convert to Foxp3⁺ Tregs and thus their ability to expand, as regulatory T cells suppress inflammation and proliferation in T cell transfer colitis. However, this appears unlikely as there are no differences in the percent of Foxp3⁺ cells between the WT or IL-6R$\mathrm{\alpha }$^Tdel populations. In addition, the use of a co-transfer system here ensures that expansion of both WT and IL-6R$\mathrm{\alpha }$^Tdel cells are equally suppressed by any Tregs in the environment, regardless of their ability to be converted.

Using low dose LPS to transiently boost serum IL-6 levels, we demonstrate that heightening IL-6 levels above baseline intensifies T cell expansion. The effect of LPS is predominantly dependent on IL-6 signaling and is observed both in lymphodeficient and lymphoreplete transfer models, demonstrating that IL-6’s actions are not exclusive to a lymphocyte-deficient setting. We did observe a slight increase in the number of IL-6Rα^Tdel T cells upon LPS treatment, indicating a possible influence of other cytokine pathways. LPS is a strong inducer of the TLR4-Myd88-NFκB and TRIF pathways (54, 55), which lead to production of an abundance of inflammatory cytokines (56) that may have IL-6-independent effects on T cell expansion. Regardless, the significant competitive disadvantage of IL-6Rα^Tdel cells in the presence of LPS stimulation supports a primary role for IL-6 in inflammation-induced T cell expansion.

IL-6 initiates signaling via two mechanisms. Classical cis-mediated signaling takes place when cells express both IL-6Rα and gp130, and binding of soluble IL-6 triggers signaling. IL-6Rα also exists as a soluble receptor that can bind to IL-6 in solution and initiate trans-signaling on cells that express only gp130 (57). Thus it is possible that IL-6 in the serum may bind to soluble IL-6Rα produced by other cell types and initiate trans-signaling in IL-6Rα^Tdel T cells, which retain expression of gp130. Classical IL-6 signaling has been implicated in T cell transfer colitis models (58), while trans-signaling is seen in Crohn’s disease and rheumatoid arthritis (24, 59). Effects of IL-6 trans-signaling would be controlled for by co-transfers of WT and IL-6Rα^Tdel, and we obtained similar results from transfers of individual populations and co-transfer of both populations. Our results were also similar using an anti-IL-6Rα antibody, which would eliminate both cis- and trans- mediated IL-6 signaling. Thus our data implies that the impact of trans signaling is relatively minor here, and that IL-6Rα acts largely cell intrinsically to promote T cell expansion in inflammatory environments.

Considering the minimal quantities of IL-6 present under normal physiological conditions and its rapid rise with inflammation, the low level expression of RORγt induced via IL-6 may prime T cells to begin proliferating in circumstances where rapid expansion is necessary for optimal immune function. IL-6 alone may initiate low level RORγt expression and T cell expansion, while IL-6 in concert with other cytokines including TGF-β, IL-1β, or IL-23 induces higher levels of RORγt that promote Th17 development (60, 61). One possibility is that STAT1 or STAT3 activation induced by IL-6 alone may promote low level RORγt expression, while additional signaling pathways stimulated by other cytokines are needed to more fully drive RORγt expression and Th17 conversion. Taken together, our results suggest a pro-proliferative role for IL-6-induced RORγt that promotes T cell expansion under inflammatory conditions. An FDA approved inhibitor of the IL-6 pathway has already proven efficacious in Cytokine Release Syndrome, Rheumatoid Arthritis, Juvenile Idiopathic Arthritis, and Castleman’s Disease, and is being evaluated for other conditions. An improved understanding of IL-6’s role in modifying inflammatory responses will be essential for the proper use of this and similar agents.