IL-27 increases the proliferation and effector functions of human naïve CD8+ T lymphocytes and promotes their development into Tc1 cells

Authors

  • Raphael Schneider,

    1. Department of Medicine, Université de Montréal, CRCHUM-Notre-Dame, Pavilion J.A. De Sève, Montreal, QC, Canada
    Search for more papers by this author
  • Teodora Yaneva,

    1. Department of Medicine, Université de Montréal, CRCHUM-Notre-Dame, Pavilion J.A. De Sève, Montreal, QC, Canada
    Search for more papers by this author
  • Diane Beauseigle,

    1. Department of Medicine, Université de Montréal, CRCHUM-Notre-Dame, Pavilion J.A. De Sève, Montreal, QC, Canada
    Search for more papers by this author
  • Lama El-Khoury,

    1. Department of Medicine, Université de Montréal, CRCHUM-Notre-Dame, Pavilion J.A. De Sève, Montreal, QC, Canada
    Search for more papers by this author
  • Nathalie Arbour

    Corresponding author
    1. Department of Medicine, Université de Montréal, CRCHUM-Notre-Dame, Pavilion J.A. De Sève, Montreal, QC, Canada
    • Department of Medicine, Faculty of Medicine, Université de Montréal, Centre de Recherche-Centre Hospitalier de l'Université de Montréal Notre-Dame Hospital, Pavilion JA DeSève (Y-3609), 1560 Sherbrooke E, Montreal, QC, H2L 4M1, CanadaFax: +1-514-412-7602
    Search for more papers by this author

Abstract

IL-27 has been shown to exhibit both pro- and anti-inflammatory properties; it favors mouse naïve CD4+T-cell differentiation into Th1 cells to the detriment of Th17 and Th2 skewing and regulates IL-10 and IL-17 production by human CD4+ T cells. Moreover, IL-27 promotes proliferation and cytotoxic functions of mouse CD8+ T lymphocytes, but no data are available on human CD8+ T cells. We investigated the impact of IL-27 on human CD8+T cells. In contrast to mouse T cells, the IL-27 receptor (IL-27R), composed of T cell cytokine receptor (TCCR) and gp130, was detected on a greater percentage of human CD8+ than CD4+ T cells and these proportions increased upon polyclonal activation. IL-27 induced rapid STAT1 and STAT3 signaling, enhanced STAT1 protein levels, and induced SOCS1 and SOCS3 expression in a STAT1-dependent manner by human CD8+ T cells. Addition of IL-27 to α-CD3-activated naïve CD8+ T cells significantly increased T-box transcription factor expression levels, cell proliferation, and IFN-γ and granzyme B production leading to increased CD8+ T-cell-mediated cytotoxicity. These results demonstrate that IL-27, a rapidly produced cytokine by activated APC, has a profound impact on human naïve CD8+T cells, driving them to become highly efficient Tc1 cells.

Introduction

T cells partake in the adaptive immune responses, whereas CD4+ T cells play important roles in orchestrating multiple subsequent responses, CD8+ T cells are crucial in clearing intracellular pathogens. Activated CD8+ T lymphocytes can secrete cytokines and kill target cells via several mechanisms upon recognition of specific Ag-MHC class I complexes. Subsets of activated CD8+ T cells persist as memory cells providing long-term protection from Ag re-encounter 1. Cytokines present in the vicinity of T cells being activated profoundly influence the extent and quality of T-cell responses. Most studies addressing the influence of such mediators on T-cell responses have focused on CD4+ T cells. However, CD8+ T cells have been shown to respond differently to their environment than their CD4+counterparts 2–5.

IL-27 belongs to the IL-6/IL-12 family and is composed of two distinct subunits: Epstein–Barr virus-induced gene 3 (EBI3) and p28, each sharing similarity with the p40 and p35 subunits of IL-12, respectively 6. IL-27 is rapidly produced by APC upon stimulation through TLR signaling 6, 7 and thus can have a crucial impact on the initial T-cell responses 8. Both pro- and anti-inflammatory properties have been attributed to IL-27. Although it favors mouse naïve CD4+ T-cell differentiation into Th1 cells to the detriment of Th17 or Th2 differentiation, IL-27 negatively controls later Th1 responses 9–11. IL-27 suppresses IL-2 12 but enhances IL-10 production by CD4+ T cells 13–16. Furthermore, IL-27 promotes proliferation and cytotoxic functions of mouse CD8+ T lymphocytes 17–19, but similar data for their human counterparts have not been reported. IL-27 downstream signaling involves mainly STAT1 and STAT3 and to lesser extent STAT4 and STAT5 20, 21 depending on cell type and activation status.

The complete signaling receptor for IL-27 (IL-27R) consists of two chains: T-cell cytokine receptor (TCCR (also named WSX-1)) and gp130, and both subunits are required to transduce a signal 22. Murine CD4+ and CD8+ T lymphocytes enhance TCCR surface expression upon infection or TCR signaling regardless of polarizing conditions 23. Both IL-27R chains have been detected at the mRNA level in several human cells including NK cells, monocytes, DCs, T and B lymphocytes, mast cells and endothelial cells 22, 24, but the protein expression of this receptor has not been completely resolved.

Increased levels of IL-27 have been detected in inflammatory diseases both in humans and their animal models 8, 25, 26. However, as recently illustrated 27, IL-27 might have distinct effects on human cells compared to their mouse counterparts. Before IL-27 can become a therapeutic target in humans, pro- and anti-inflammatory effects of this cytokine need to be thoroughly elucidated. To expand the limited knowledge about the role of IL-27 in human CD8+ T-cell biology we investigated the presence of its receptor on human CD8+ and CD4+ T cells and its impact on CD8+ T-cell functions.

Results

A greater proportion of human CD8+ T cells expresses IL-27R compared to their CD4+ counterparts

As a variety of leukocytes has been shown to express IL-27R 22 we elected to investigate its presence on human ex vivo CD8+ and CD4+ T cells. We first confirmed the specificity of surface TCCR detection using both TCCR-transfected and IL-13R-transfected HEK-293 cells. Only TCCR expressing cells were positive using commercially available Abs (Fig. 1A and data not shown for rabbit α-TCCR).

Figure 1.

A greater proportion of human CD8+ T cells expresses IL-27R compared to CD4+counterparts. (A) Specificity of the anti-TCCR Ab was tested by FACS on HEK-293 cells expressing a control protein (left) or human TCCR (right); filled gray: isotype control; solid black line: anti-TCCR Ab. (B) PBMCs were stained ex vivo for surface TCCR, gp130, CD14, CD8, CD4, CD45RO, CD45RA and CCR7. Gating strategy is illustrated on first row: cell debris were excluded on FSC versus SSC, then cells either CD14+ (monocytes) or CD3+ (T cells) were selected; for T cells, further gating included CD4+or CD8+and for each of these subsets CD45RA+ or CD45RO+ cells were analyzed. Representative dot plots gated on either T cells (CD8+or CD4+) or monocytes (CD14+) are illustrated for TCCR and gp130 (upper row) or gated on CD8+T cells according to the memory (CD45RO+), naïve (CD45RA+CCR7+) and effector memory (CD45RA+CCR7-) subsets (bottom row). (C) For each cell subset, percentage of cells expressing both TCCR and gp130 obtained from ten different donors are shown; horizontal bar shows the mean. Paired Students' t-test comparing indicated subsets (CD8+ versus CD4+; CD8CD45RO+ versus CD8+CD45RA+) for ten donors. ***p<0.001.

A typical example of gp130 and TCCR FACS detection on PBMCs gated either on CD8, CD4 or CD14 is illustrated in Fig. 1B. We observed a significantly greater proportion of CD8+ T cells (6.6±1.0%, n=10,) expressing the complete IL-27R (TCCR and gp130), compared to CD4+ T cells (0.9±1.0%, n=10) (Fig. 1B and C). Moreover, most monocytes (CD14+ cells) (74.9±2.7%, n=10) in the same samples expressed high levels of TCCR (ΔMFI 2929±205) compared to T cells (ΔMFI 1144±72 for CD4; ΔMFI 983±66 for CD8). To further specify the phenotype of T cells expressing IL27R, we concomitantly looked for memory or naïve surface cell markers (CD45RO and CD45RA) and the homing receptor CCR7. A great proportion of CD8+ T cells expressing IL-27R was found to belong to the memory (RO) subset (mean 7.0±0.9%) compared to the naïve (CD45RA+) subset (mean 2.6±0.9%) (Fig. 1C). Amongst the CD45RA+ CD8+ T cells, similar proportions of CCR7+ cells (truly naïve) (3.5±0.8%) and CCR7- (effector memory) (2.4±0.8%) expressed the IL-27R. The proportion of CD4+ T cells expressing the complete IL-27R was similar in the naïve (2.3±0.2%) and memory (1.3±0.2%) subsets (Fig. 1C).

Expression of IL-27R by human T cells: Intracellular and surface TCCR detection and upregulation

We investigated whether TCCR is solely expressed on the surface of CD8+ T cells and observed that this protein is mainly located within permeabilized, untreated and activated CD8+ T cells, but also on the cell surface especially of activated cells (Fig. 2A). We also assessed TCCR intracellular and surface expression by FACS and detected significantly greater numbers of CD8+ T cells (48.2±18.1% n=5, Supporting Information Fig. 1B) expressing TCCR intracellularly than on their cell surface (Fig. 2B), thus both techniques confirmed the intracellular detection of TCCR in human CD8+ T cells. We also determined TCCR intracellular expression by FACS upon α-CD3 activation and detected similar numbers of CD8+ T cells expressing TCCR with and without activation (nil: 37.2±4.1% versus α-CD3: 34.4±3.2%, n=5, Fig. 2C) and similar ΔMFI. However, intracellular TCCR levels (ΔMFI=2150±75) were always greater than surface levels (ΔMFI=980±70) on activated cells. Similarly to what we observed for surface expression, a significantly greater proportion of CD8+ T cells expressed intracellular TCCR than their CD4+ counterparts (Supporting Information Fig. 1A and B).

Figure 2.

Expression of IL-27R by human CD8+ T cells: intracellular and surface TCCR detection and upregulation upon polyclonal activation. (A) Cellular localization of TCCR. PBMCs either treated with α-CD3 for 48 h or untreated were permeabilized and stained for CD8 (red) and TCCR (green) and nuclei (blue). Scale bar=2.5 μm, original magnification 1000×. (B) Representative FACS profiles of isotype controls, intracellular and surface stainings for TCCR detection gated on CD8+T cells. (C) Representative FACS profiles of isotype controls and intracellular TCCR detection gated on CD8+ T cells incubated in the presence or absence of α-CD3 for 48 h (left). Data (n=5 donors) gated on CD8+ cells showing percentage expressing TCCR incubated with or without α-CD3 for 48 h (right). Paired Students' t-test comparing nil versus α-CD3 for five donors. n.s., not statistically significant. (D) PBMCs were activated with α-CD3 48 h prior to TCCR, gp130 and CD8 surface staining. Data from five donors gated on CD8+ cells showing percentage expressing both TCCR and gp130 (left) or gp130 regardless of TCCR (right). Paired Students' t-test comparing nil versus α-CD3 for five donors; *p<0.05. (E) Western blot analysis for TCCR and actin from purified CD8+ T cells incubated in the presence or absence of α-CD3 for 48 h. One representative donor (left) and pooled data (n=4 donors) representing mean±SEM signal intensity relative of TCCR to actin (right) are shown; numbers indicate position of ladder markers. Paired Students' t-test comparing nil versus α-CD3 for four donors. n.s., not statistically significant.

To investigate whether the proportion of T cells expressing surface IL-27R varies upon activation, we evaluated IL-27R expression following in vitro α-CD3 T-cell activation. We observed a significant increase in the proportion of CD8+ T cells expressing the IL-27R after 2 days (18.3±3.3%, n=5, Fig. 2D) compared to nonactivated cells (2.9±1.1%, n=5). The proportion of CD4+ T cells expressing IL-27R also increased upon α-CD3 activation (nil: 0.6±0.1% versus α-CD3: 4.6±0.8%, n=5), albeit less than for CD8+ T cells (Supporting Information Fig. 1C). This elevated proportion was sustained after 4 and 6 days in culture (data not shown). Mouse CD4+ and CD8+ T cells have been shown to downregulate gp130 upon TCR cross-linking 28, yet an upregulation of gp130 on human CD4+ T cells upon similar activation (e.g. α-CD3) has been reported 29. Thus, we also assessed gp130 expression regardless of TCCR and did find that while α-CD3 activation led to a significantly enhanced proportion of gp130-expressing CD8+ T cells (Fig. 2D), the increase in gp130 expression on CD4+ T cells did not reach significance (Supporting Information Fig. 1D). We assessed total TCCR protein from CD8+ (Fig. 2E) and CD4+ T cells (data not shown) by western blot analysis. We found equivalent amounts of TCCR in both cell types and α-CD3 activation did not lead to an increase of total TCCR protein in either cell type; rather we observed a small nonsignificant decrease in TCCR upon activation in some donors (Fig. 2E).

IL-27 induces a rapid phosphorylation of STAT1 and STAT3 in human CD8+ T cells

We next examined the effect of IL-27 on the signaling transduction pathways downstream of TCCR/gp130; IL-27 has been shown to mediate its downstream effects mainly through STAT1 and STAT3 signaling 20, 22, 30, 31. We investigated the kinetics of STAT1 through STAT5 phosphorylation in human CD8+ T cells upon stimulation with IL-27. We observed a rapid and significant increase in STAT1 and STAT3 phosphorylation peaking at 15 min post IL-27 stimulation (Fig. 3A). The 100 ng/mL dose was almost or as efficient as the 250 ng/mL dose whereas 10 ng/mL of IL-27 had only a minor impact on STAT phosphorylation. Addition of IL-27 (100 ng/mL or more) to naïve CD8+ T cells (CD45RA+ cells) significantly increased STAT1 and STAT3 phosphorylation without significantly affecting the same signaling molecules in memory CD8+ T cells (CD45RO+ cells) (Fig. 3A). Stimulation with IL-27 did not lead to a significant increase in STAT2, STAT4 or STAT5 phosphorylation (data not shown). To assess the impact of IL-27 on total STAT1 and STAT3 protein levels, we stimulated purified naïve CD8+ T cells with IL-27 for 16 h and performed western blot analysis for STAT1 and STAT3 for four donors. We noticed that IL-27 significantly elevated STAT1 protein levels (Fig. 3B) relative to actin and compared to untreated cells. However, we did not observe any significant impact on total STAT3 protein in these experiments. Thus, IL-27 not only signals via STAT1 and STAT3, but also enhances STAT1 protein levels.

Figure 3.

IL-27 induces a rapid phosphorylation of STAT1 and STAT3 in naïve human CD8+ T cells. (A) Cells were incubated in the absence or presence of IL-27 for 15, 30, or 60 min, prior to staining with anti-CD8, anti-STAT1p and anti-STAT3p. Representative histograms of typical STAT1 and STAT3 phosphorylation detection in CD8+ T cells; tinted area: isotype; black line: cells incubated with IL-27 (100 ng/mL) for 15 min; dotted line: cells incubated with IL-27 for 60 min. Pooled data (n=4 donors) of STAT1 and STAT3 phosphorylation expression in absence or presence of IL-27 at different time points (mean±SEM) are illustrated for total CD8+ T cells. Levels of STAT1 and STAT3 phosphorylation upon IL-27 (100 ng/mL) addition specifically in naïve (CD45RA) and memory (CD45RO) CD8+ T cells (n=4 donors) are shown. Paired Students' t-test for four donors comparing 0 versus 15 min; *p<0.05. (B) Western blot detection of STAT1, STAT3 and actin in purified CD8+ T cells incubated in the presence or absence of IL-27 for 18 h. One representative western blot with molecular weight ladder and quantified pooled data (n=4 donors) are illustrated. Mean±SEM for four donors. Paired Students' t-test for four donors comparing nil versus 100 or 250 ng/mL;*p<0.05, **p<0.01.

IL-27 induces SOCS1 and SOCS3 expression in human CD8+ T cells

To determine whether IL-27 induces immune mediators via the STAT1 and STAT3 pathways in CD8+ T cells, we chose to examine the induction of SOCS1 and SOCS3 as readout of STAT1 and STAT3 signaling. SOCS1 is rapidly induced as an immediate early gene in response to cytokines activating STAT1. Similarly, SOCS3 is a target gene of STAT3 signaling, but STAT1 also contributes to its induction 32. We assessed SOCS1 and SOCS3 mRNA levels after 1 or 3 h of incubation in the absence or presence of IL-27 and observed their induction in a dose-dependent manner peaking at 1 h (data not shown) occurring without TCR activation. Treatment of cells with 100 ng/mL of IL-27 led to a significant induction of SOCS1 and SOCS3 in both total and naïve CD8+ T cells. To investigate whether signaling via STAT1 is essential for the IL-27-mediated effects, we elected to knockdown STAT1 levels in CD8+ T cells prior to IL-27 stimulation. We pretreated our cells with Fludarabine (fluda), a STAT1-specific inhibitor that does not affect other STATs 33, 34. We confirmed that fluda inhibited STAT1 and not STAT3 protein in CD8+ T cells by western blotting (Supporting Information Fig. 2). Overnight pretreatment of naïve CD8+ T cells with fluda (5 or 10 μM) significantly decreased the induction of SOCS1 and SOCS3 following 1-h stimulation with IL-27 (100 ng/mL) (Fig. 4A), demonstrating that STAT1 signaling is essential to induce these genes in naïve CD8+ T cells. IL-27 also increased SOCS1 (p=0.11) and SOCS3 (p<0.05) expression in memory CD8+ T cells (Fig. 4B). However, in contrast to naïve CD8+ T cells, pretreatment of memory CD8+ T cells with fluda did not significantly affect the IL-27-mediated induction of both SOCS (Fig. 4B).

Figure 4.

IL-27 induces the expression of SOCS1 and SOCS3 in CD8+ T cells. Purified CD45RA+CD8+ T cells (A) or CD45RO+CD8+ T cells (B) were cultured in the presence or absence of fluda (5 or 10 μM) for 18 h prior to being exposed to IL-27 (100 ng/mL) for 1 h. qPCR analysis (n=4 donors) of SOCS1 (left) and SOCS3 (right) mRNA expression; data are presented as relative expression compared to control (CTL). Mean±SEM of pooled data from four donors. Paired Students' t-test performed on data from four donors and comparing appropriate subsets as indicated on each graph. *p<0.05, **p<0.01, θp=0.11.

IL-27 increases T-box transcription factor (T-bet) expression in human naïve CD8+ T cells

IL-27 has been reported to induce T-bet, the lineage-specific transcription factor for Th1 development, in murine naïve CD4+ T cells 20, 31 whereas this cytokine has been shown to reduce GATA3 and RAR-related orphan receptor gamma (ROR-γ) without altering T-bet in human CD4+ T cells 16. Thus, to evaluate the impact of IL-27 on CD8+ T-cell lineage commitment, we performed qPCR for lineage specific transcription factors for Tc1 (T-bet), Tc2 (GATA3) and Tc17 (ROR-γ). For this purpose, purified naïve CD8+ T cells were cultured with α-CD3 in the absence or presence of IL-27 (100 ng/mL) for 16 h or 6 days. IL-27 led to a significant increase in T-bet expression (Fig. 5) after each time point; however, IL-27 did not have any significant impact on GATA3 or ROR-γ (data not shown). Our data suggest the presence of IL-27 during TCR activation of naïve CD8+ T cells favors Tc1 polarization.

Figure 5.

IL-27 enhances T-bet in naïve CD8+ T cells. Purified CD45RA+CD8+ T cells were cultured for 6 h or 6 days in the presence of α-CD3 with or without IL-27. qPCR analysis of T-bet mRNA expression (n=5 donors); data are presented as fold increase compared to treatment with α-CD3 alone at the corresponding time point (6 h or 6 days). Mean±SEM for pooled data obtained from five donors. Paired Students' t-test comparing data from five donors: 0 versus 100 ng/mL IL-27.*p<0.05.

IL-27 enhances proliferation and augments IFN-γ and granzyme B production by human CD8+ T cells

To evaluate the direct impact of IL-27 on CD8+ T cells, CFSE-labeled CD8+ T cells (total, naïve or memory) were stimulated with plate-bound α-CD3 in the absence or presence of IL-27. After a 6-day culture, cells were shortly activated with PMA, ionomycin and BFA and then stained for CD8, IFN-γ, and granzyme B. Activated CD8+ T cells underwent multiple divisions upon contact with α-CD3; addition of IL-27 significantly increased the number of cells entering divisions 3 and 4 (Fig. 6A) and the total proliferation in a dose-dependent manner. Naïve CD8+ T cells (CD45RA+ cells) put in culture in the presence of α-CD3 also increased their proliferation upon addition of IL-27 in a dose-dependent manner and a dose as low as 10 ng/mL was sufficient to significantly impact their proliferation (Fig. 6B). Although CD45RO+CD8+ T cells proliferated very well upon activation with plate-bound α-CD3, the addition of IL-27 had no impact even when additional time points (3–5 days) were tested (data not shown).

Figure 6.

IL-27 increases the TCR-induced proliferation, and IFN-γ and granzyme B production by naïve CD8+ T cells. CFSE-labeled CD8+ T-cell subsets were cultured in the absence (NIL) or presence of α-CD3 with or without recombinant human IL-27 (doses indicated in ng/mL) for 6 days prior to staining for IFN-γ and granzyme B (A). Representative FACS plots for proliferation and IFN-γ production. Percentages of CD8+ T cells that proliferated (B) and produced IFN-γ (C) or granzyme B (D) from pooled data (n=4–8 donors) are presented for indicated CD8+ T-cell subsets. (E) Ratio of the percentage of proliferated cells producing IFN-γ over all proliferated cells and ratio of the percentage of proliferated cells producing granzyme B over all proliferated cells are illustrated for CD45RA+CD8+ T cells. Mean±SEM for 4–8 donors. Paired Students' t-test for 4–8 donors comparing various doses of IL-27 to cells without IL-27. ***p<0.001, **p<0.01, *p<0.05.

Addition of IL-27 to total CD8+ T cells or naïve CD8+ T cells did also significantly increase the proportion of IFN-γ and granzyme B expressing cells (Fig. 6C and D) without affecting the same functions in memory CD8+ T cells (CD45RO+ cells). The ratios of IFN-γ+proliferation+ cells over total proliferation and of granzyme B+proliferation+ to total proliferation are illustrated in Fig. 6E for the CD8+CD45RA+cells. At higher doses (100 and 250 ng/mL for IFN-γ and 250 ng/mL for granzyme B), the addition of IL-27 not only increased the proliferation but also significantly enhanced the proportion of naïve CD8+ T cells that acquired these effector functions. To determine whether the CD8+CD45RA+ cells that were responding to IL-27 were truly naïve cells (CD8+CD45RA+CCR7+) we FACS sorted those cells with a very high purity (>99%) and put them in culture on plate-bound α-CD3 in the presence of IL-27 and assessed for proliferation, IFN-γ and granzyme B. The addition of IL-27 increased the proliferation, granzyme B and IFN-γ production of the truly naïve CD45RA+CCR7+ cells for both donors tested (Supporting Information Fig. 3), in a dose-dependent manner especially at 100 and 250 ng/mL with increases of 20–100% compared to α-CD3 alone.

IL-27 enhances CD8+ T cell-mediated cytotoxicity

To investigate whether IL-27 amplifies the capacity of CD8+ T cells to kill target cells, we performed an in vitro re-directed killing assay using a mast cell line (P815) as target cells, allowing us to detect killing activity regardless of TCR specificity. Naïve CD8+ T cells were cultured on plate-bound α-CD3 with or without IL-27 (100 or 250 ng/mL) and then collected, washed, and used as effector cells. Equal amounts of CFSEhi target cells loaded with α-CD3 and CFSElo target cells loaded with an isotype control were mixed with increasing numbers of effector cells in different E:T ratios (10:1; 20:1; 40:1). After 18-h incubation, the proportion of CFSEhi and CFSElo cells was analyzed by FACS. An overlay histogram of CD8+ T cells (white) and target cell populations (CFSEhi/CFSElo) (gray) prior to incubation shows (Fig. 7A) that these three cell populations could be easily distinguished and a representative example of CFSEhi cell killing which was enhanced in the presence of IL-27 is displayed in Fig. 7B (E:T ratio of 40:1). CD8+ T cells that had been exposed to IL-27 (100 or 250 ng/mL) were significantly more efficient at killing CFSEhi α-CD3-loaded target cells than CD8+ T cells that had been cultured with α-CD3 alone (pooled data n=5 donors, Fig. 7C) at E:T ratio from 20:1 and higher. Thus, our results suggest that addition of IL-27 to the TCR activation of naïve CD8+ T cells boosts their cytotoxicity.

Figure 7.

IL-27 enhances CD8+ T cell-mediated cytotoxicity. Naïve CD8+ T cells after a 6-day culture with α-CD3 in the absence or presence of IL-27 were exposed to α-CD3 and isotype-loaded target cells. (A) Histogram illustrating CFSE detection gated on CD8+ T cells (white) and mixed target cell populations: α-CD3 loaded CFSEhi and isotype-loaded CFSElo target cells (filled light gray). (B) Representative histogram of target cell killing. The filled light gray area represents target cells exposed to CD8+ T cells previously cultured with α-CD3 alone, the filled dark gray area represents target cells exposed to CD8+ T cells previously cultured with α-CD3 and IL-27 (100 ng/mL). E:T ratio of 40:1. (C) Pooled data of target cell killing (n=5 donors) mediated by CD8+ T cells previously cultured with α-CD3 alone (gray line), α-CD3 and IL-27 (100 ng/mL) (black line) or α-CD3 and IL-27 (250 ng/mL) (dotted line) are illustrated. E:T ratios are displayed on the x-axis. Mean±SEM for five donors. Paired Students' t-test for five donors comparing doses of 100 ng/mL or 250 ng/mL to effector cells cultured without IL-27. *p<0.05.

Discussion

Our study demonstrates that IL-27 significantly enhances the proliferation and effector functions of human naïve CD8+ T lymphocytes upon TCR activation. IL-27 induces a significant phosphorylation of STAT1 and STAT3, supporting the role of these signaling pathways downstream of IL-27R. Moreover, T-bet expression but not GATA3 nor ROR-γ, is upregulated in the presence of IL-27 implying that this cytokine favors the polarization of human naïve CD8+ T cells into Tc1 cells.

In contrast to a previous report documenting similar percentages of mouse CD8+ and CD4+ T cells expressing TCCR ex vivo 23, we observed that a greater proportion of human CD8+ compared to CD4+ T cells expresses surface and intracellular TCCR (Fig. 1). We are also the first ones to quantify TCCR expression on T cells obtained from multiple healthy donors, as a previous report showed only one example of FACS TCCR staining on human CD4+ T cells 16. Although we detected enhanced surface TCCR expression on T cells following TCR activation, western blot analysis showed similar levels with or without activation (Fig. 2). Therefore, TCR activation did specifically increase the TCCR expression at the surface leaving total protein levels unaffected. Our results underline the importance of assessing TCCR on cell surface rather than by other more global techniques (mRNA or western blot). The exact mechanisms controlling TCCR cell localization remain unresolved and are currently under investigation. It is plausible that the enhanced TCCR expression we observed following TCR-mediated activation (α-CD3 cross-linking) was due to surface expression of a fraction of the already existing intracellular pools of TCCR (Fig. 2). These pools did not significantly vary upon activation as illustrated by similar percentage of positive cells or ΔMFI and remained much higher than surface levels. In a physiological setting the upregulation of surface TCCR following CD8+ T-cell activation could render these CD8+ T cells more susceptible to IL-27 potentially leading to acquisition of physiological or pathological functions. Finally, although mouse CD4+and CD8+ T cells have been shown to downregulate gp130 upon TCR cross-linking 28, we detected an upregulation of this cytokine receptor on both human CD8+ and CD4+ T cells (Fig. 2 and Supporting Information Fig. 2) in line with a previous report on human CD4+ T cells 29. Thus, our results indicate that human T cells control the expression of these cytokine receptor chains using different mechanisms than their mouse counterparts. This observation could have a major impact not only on IL-27 effects but also for the numerous cytokines using gp130 as a receptor. Finally, a recent publication suggested that murine CD8+ T cells can be a source of IL-27 35. We evaluated whether purified human CD8+ T cells also transcribe IL-27 subunits (EBI3 and p28) upon α-CD3 and α-CD28 antibodies stimulation after 6 or 24 h. Although EBI3 was clearly upregulated after 24 h of stimulation as detected by qPCR, p28 was not detected even after 40 PCR cycles albeit we obtained a strong signal from our positive control, human activated dendritic cells. This suggests that human CD8+ T cells can not be a source of IL-27.

We performed our functional assays using isolated CD8+ T cells in the absence of APC to avoid confounding factors since APC also respond to IL-27 stimulation 27, 36. Moreover, we detected the effects of IL-27, at similar doses used by others 12, 27, 37. IL-27 has been shown to induce phosphorylation of STAT1, -2, -3, -4 and -5 in mouse naïve T cells and STAT1 and 3 in human monocytes, B cells, and naïve T cells 20, 27, 37. We observed that IL-27 quickly induced the phosphorylation of STAT1 and STAT3 but not STAT2, 4 or 5; and increased total STAT1 at the protein level in human naïve CD8+ T cells but these effects were not observed in memory CD8+ T cells (Fig. 3A). These observations could be due to a greater heterogeneity within the memory CD8+ T-cell compartment. Our results correlate with other reports demonstrating that several effects of IL-27 have been linked to STAT1 and STAT3 phosphorylation 8 and that IL-27 boosts STAT1 protein levels in human cells. STAT1 signaling is involved in IL-27-mediated induction of granzyme B and other mediators in mouse naïve CD8+ T cells. Similarly, ICAM-1, T-bet, IL-12Rβ2 and Th1 differentiation (including IFN-γ production) in mouse naïve CD4+ T cells have been linked to the activation of STAT1 18, 30. Overall, both STAT1- and STAT3-mediated signaling probably contribute to the impact of IL-27 on human CD4+ and CD8+ T cells. Recessive human STAT1 deficiencies have been shown to impair the IL-27 signaling pathway 38, whether it has a direct impact on CD8+ T cells has not been investigated.

It has been established that IL-27 induces SOCS3 in other cell types 39. However, although SOCS1 is rapidly induced in response to cytokines activating STAT1, there is no previous report demonstrating that IL-27 induces this protein. We observed that IL-27 induces a rapid and robust SOCS1 and SOCS3 expression that is partially mediated by STAT1-signaling in naïve CD8+ T cells (Fig. 4A). The partial effect of fluda could be due to incomplete knockout of the STAT1 pathway. However, it is also possible that IL-27 can induce SOCS1 and SOCS3 through additional STAT1 independent mechanisms especially in memory CD8+ T cells. Indeed, although IL-27 induced both SOCS1 and SOCS3 expression by memory CD8+ T cells, fluda pretreatment had only a minor impact (Fig. 4). Thus, our results suggest that IL-27 modulates early CD8+ T-cell responses by upregulating SOCS1 and SOCS3, which are important regulatory mediators 39.

IL-27 has been shown to influence the expression of crucial transcription factors associated with T-cell polarization. This cytokine favors murine naïve CD4+ T-cell differentiation into Th1 cells to the detriment of Th17 or Th2 differentiation 11, 40. A recent report showed that IL-27 reduced GATA3 and RORC without affecting T-bet and Foxp3 mRNA levels in human CD4+ T cells 16. We established that IL-27 increases T-bet expression in naïve CD8+ T cells (Fig. 5), leaving GATA3 and ROR-γ unmodified. Thus, our results imply that IL-27 distinctly influences the polarization of human CD4+ and CD8+ T cells. We determined that the activation of human naïve CD8+ T cells in the presence of IL-27-enhanced Tc1 effector functions as we detected significant increase in the proliferation and production of granzyme B and IFN-γ (Fig. 6) leading to amplified cytotoxicity (Fig. 7). Although a greater proportion of memory CD8+ T cells expresses IL-27R, the addition of IL-27 did not alter their proliferation or the effector functions we assessed. We can speculate that IL-27 could influence other functions of this subset. Similarly, human B cells distinctly respond to IL-27 according to their differentiation status and mode of activation 37. Consequently, IL-27 could influence several human immune cell subsets according to their maturation/differentiation stage.

Of clinical interest, elevated amounts of IL-27 have been reported in several human diseases such as Crohn's disease 41 and other Th1-associated granulomatous diseases 26. In addition, IL-27 has been implicated in host defense against viruses such as Hepatitis C 17. Therefore, the impact of IL-27 on human CD8+ T cells should be further investigated in human disease conditions especially since this cytokine and its receptor have distinct expression and effects on human cells compared to their mouse counterparts 27.

Materials and methods

Cell isolation procedure

A written informed consent was obtained from healthy donors and studies were approved by the Centre Hospitalier de l'Université de Montréal ethical boards. PBMCs were isolated as previously described 42 and either analyzed by FACS, cultured in complete RPMI (RPMI containing 10% FBS) or put in MACS buffer (PBS containing 2 mM EDTA and 0.5% FBS) prior to proceeding to magnetic beads T-cell isolation. CD8+ or CD4+ T-cell subsets were isolated using anti-human CD8 or CD4 beads respectively according to manufacturer's instructions (Miltenyi Biotech). CD45RA or CD45RO expressing cells were first removed prior to a CD8+ selection to isolate memory or naïve subsets respectively. Cell purity of each subset assessed by FACS was>95%.

FACS

Cells were stained as previously described 42 (see Abs Supporting information Table 1), and acquired on a LSR II flow cytometer (BD Biosciences). To assess STAT1p and STAT3p, PBMCs were first surface stained, then fixed with 1.5% paraformaldehyde, and finally permeabilized with 90% ice-cold methanol prior to labeling. Appropriate isotype controls were used in all steps. Staining specificity was confirmed using fluorescence minus one (all Abs minus one). FACS analyses were performed using FlowJo software (Treestar).

Confocal microscopy

PBMCs were cytospined with the StatSpin Cytofuge®2 (Iris sample processing), fixed with 2% paraformaldehyde and blocked for nonspecific binding with HBSS containing goat, horse and fetal bovine sera. Slides were incubated with mouse anti-CD8 and then permeabilized with PBS-Saponin 0.1% prior to being incubated with biotinylated rabbit anti-human TCCR Ab followed by addition of Cy3-conjugated goat anti-mouse Ab and Alexa Fluor® 488-conjugated streptavidin and finally with TO-PRO®-3 iodide (Invitrogen) for nucleus staining. Images were taken with a Leica SP5 confocal microscope.

In vitro culture of T cells

To assess gp130 and TCCR expression, PBMCs were put in culture for 2 days in complete RPMI in the absence or presence of α-CD3 (clone OKT3, grown and purified in-house) at 32.5 ng/mL (dose optimized as ED50 to provide 50% of maximal T-cell proliferation). Alternatively, CFSE-labeled cells 42 were cultured with or without plate-bound α-CD3 Ab using our determined ED50 doses for each cell subset in presence or absence of recombinant human IL-27 (R&D Systems) for 4–6 days and then stimulated with PMA, ionomycin and Brefeldin A as previously described 42, prior to FACS staining. To assess STAT1 and STAT3 phosphorylation, cells were cultured without serum for 1 h prior IL-27 addition.

Western blot

Cells were lysed in NP-40 buffer (10 mM Tris-HCl, 10 mM NaCl, 3 mM MgCl2, and 0.5% NP-40) supplemented with a protease inhibitor cocktail (BD Biosciences), sonicated and proteins stored at −20°C. Proteins were electrophoresed on a 10% SDS-polyacrylamide gel under reducing conditions, transferred onto PVDF membrane (BioRad), and membrane was then blocked in TBS-TM (20 mM Tris-HCl pH 7.6, 37 mM NaCl, 0.1% Tween 20®, 5% skim milk) prior to being incubated overnight with primary Ab (Supporting information Table 1) in TBS-TM followed by incubation with HRP-conjugated secondary reagent (Supporting information Table 1). Specific binding was visualized using ECL PLUS system (GE Healthcare). The observed band had the expected molecular weight for all proteins (Supporting information Table 1). Band intensity was quantified using the QuantityOne software (BioRad).

Quantitative Real-time PCR

Total RNA was extracted using RNeasy Mini kit (Qiagen) and then transcribed into cDNA using Quantitect RT kit according to the manufacturer' instruction (Qiagen). Relative gene expression levels were determined using primers and TaqMan FAM-labeled MGB probes for SOCS1, SOCS3, T-bet, GATA3 and ROR-γ and ribosomal 18S (VIC-labeled probe) (Applied Biosystems) and according to the manufacturer's instruction. qPCR cycling was performed according to the default temperature settings (2 min at 50°C, 10 min at 95°C, followed by 40 cycles of 15 s at 95°C, 1 min at 60°C) in a 7900HT Fast-Real-Time PCR System (Applied Biosystems). Gene-specific mRNA was normalized compared to endogenous control (18S) and relative expression quantified by extrapolating from an internal control using cDNA from cells having high expression levels.

Cytotoxicity assay

CD8+T cells were activated on plate-bound α-CD3 without or with IL-27 (100 or 250 ng/mL) for 6 days and then harvested for cytotoxicity assay. P815 cells were CFSE-stained with doses of either 5 μM (CFSEhi) or 0.5 μM (CFSElo). Fc receptors on P815 cells were saturated with 20 μg/mL of either α-CD3 (CFSEhi) or mouse IgG1 (CFSElo) (eBioscience). Activated CD8+ T cells were co-cultured for 18 h with CFSEhi-α-CD3 P815 cells (1×104) and CFSElo-IgG1 P815 cells (1×104) at different E:T ratios. Cells were then harvested and analyzed by FACS. Percentages of killing were calculated using the following formula:

equation image

Statistical analysis

Data were analyzed using GraphPad Prism software. Results are shown as mean±SEM and statistical analyses included paired Students' t-test.

Acknowledgements

The authors are grateful to Dr. Alexandre Prat for the scientific discussion and access to the flow cytometer LSR II and Dr. Jean-François Gauchat for providing HEK-293 transfected with human TCCR or another control molecule and for scientific discussion. This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada (♯355722-2008). R. S. obtained postdoctoral Fellowships from the German Academic Exchange Service (DAAD, ♯D0743793) and the Multiple Sclerosis Society of Canada. T. Y. and L. E. K. obtained Fellowships from the Neuroinflammation Training Program of the Canadian Institute of Health Research. N. A. holds a Donald Paty Career Development Award from the Multiple Sclerosis Society of Canada and a Chercheur-Boursier from the Fonds de la Recherche en Santé du Québec.

Conflict of interest: The authors declare no financial or commercial conflict of interest.

Ancillary