Phosphatidylinositol 3-Kinase–Independent Signaling Pathways Contribute to ICOS-Mediated T Cell Costimulation in Acute Graft-Versus-Host Disease in Mice

Jun Li, Jessica Heinrichs, Julien Leconte, Kelley Haarberg, Kenrick Semple, Chen Liu, Mathieu Gigoux, Mara Kornete, Ciriaco A. Piccirillo, Woong-Kyung Suh and Xue-Zhong Yu
J Immunol July 1, 2013, 191 (1) 200-207; DOI: https://doi.org/10.4049/jimmunol.1203485

Abstract

We and others have previously shown that ICOS plays an important role in inducing acute graft-versus-host disease (GVHD) in murine models of allogeneic bone marrow transplantation. ICOS potentiates TCR-mediated PI3K activation and intracellular calcium mobilization. However, ICOS signal transduction pathways involved in GVHD remain unknown. In this study, we examined the contribution of ICOS-PI3K signaling in the pathogenic potential of T cells using a knock-in mouse strain, ICOS-YF, which selectively lost the ability to activate PI3K. We found that when total T cells were used as alloreactive T cells, ICOS-YF T cells caused less severe GVHD compared with ICOS wild-type T cells, but they induced much more aggressive disease than ICOS knockout T cells. This intermediate level of pathogenic capacity of ICOS-YF T cells was correlated with similar levels of IFN-γ–producing CD8 T cells that developed in the recipients of ICOS-WT or ICOS-YF T cells. We further evaluated the role of ICOS-PI3K signaling in CD4 versus CD8 T cell compartment using GVHD models that are exclusively driven by CD4 or CD8 T cells. Remarkably, ICOS-YF CD8 T cells caused disease similar to ICOS wild-type CD8 T cells, whereas ICOS-YF CD4 T cells behaved very similarly to their ICOS knockout counterparts. Consistent with their in vivo pathogenic potential, CD8 T cells responded to ICOS ligation in vitro by PI3K-independent calcium flux, T cell activation, and proliferation. Thus, in acute GVHD in mice, CD4 T cells heavily rely on ICOS-PI3K signaling pathways; in contrast, CD8 T cells can use PI3K-independent ICOS signaling pathways, possibly through calcium.

Introduction

The ICOS is a member of the CD28 family of costimulatory receptors sharing structural similarity to CD28. Unlike CD28 that is expressed in naive T cells, ICOS is induced upon activation in CD4 and CD8 T cells and is a reliable marker of effector/memory populations. ICOS can enhance T cell proliferation, but to a lesser extent compared with CD28 mainly due to the limited amount of IL-2 production it can promote (1, 2). This proliferative effect of ICOS costimulation is not appreciable during in vivo immune responses presumably due to a dominant effect of CD28 (3). However, ICOS does play critical roles in generation of follicular Th cells (4), antitumor T cell responses (5), and graft-versus-host disease (GVHD) (6–9). These impacts are related to augmented expression of an array of Th1 and Th2 cytokines, including IFN-γ, IL-4, IL-10, and IL-21 by ICOS costimulation.

One of the reasons that ICOS costimulation serves a nonredundant role in T cell responses in vivo could be due to the broad, yet highly regulated, expression pattern of ICOS ligand (ICOSL). Unlike B7.1 and B7.2, whose expression is restricted to APCs, ICOSL is expressed in nonhematopoietic cells as well as APCs (10, 11). Furthermore, ICOSL is constitutively expressed in APCs (12, 13) and can be induced by inflammatory cytokines such as TNF-α (11). This may provide T cell costimulation in situations in which CD28 costimulation is limited.

In allogeneic bone marrow (BM) transplantation (BMT) settings, an initial study using a parent-into-F1 nonirradiated model showed that blocking ICOS interaction with an ICOS-specific Ab reduces chronic GVHD by selectively suppressing Th2 cytokines (14). Using myeloablative full MHC-mismatched BMT models, Taylor et al. (8) found that disruption of ICOS, achieved by using ICOS knockout (KO) mice or anti-ICOS mAb administration, resulted in significant inhibition of GVHD by reducing the number of alloantigen-specific effector cells. Results from another group indicated that ICOS blockade reduced GVHD while largely sparing graft-versus-leukemia activity by skewing toward Th2 differentiation, without affecting T cell activation, proliferation, cytotoxicity, and target organ infiltration (6). Studies by our own group indicate that ICOS deficiency or blockade results in significantly less GVHD morbidity and delayed mortality (7, 9). The effect of ICOS is predominately on CD4 T cells, and the deficiency of ICOS had no impact on their expansion, but significantly reduced their effector functions in term of expression in FasL and production of IFN-γ and TNF-α (9).

Analogous to the YMNM motif on CD28, ICOS also contains a YMFM motif in the cytoplasmic tail that is phosphorylated upon ligation and activates PI3K (4, 15). Multiple studies have shown that ICOS is much more potent in terms of PI3K activation compared with CD28 (4, 16, 17). Using a knock-in mouse strain, termed ICOS-YF, in which the cytoplasmic tail of ICOS cannot activate PI3K, we have shown that ICOS-PI3K signaling axis is critical for the generation of follicular Th cells (4). We also observed that, in preactivated CD4 T cells, ICOS could potentiate TCR-mediated calcium flux in a PI3K-independent manner. Although the ICOS-calcium signaling had a negligible impact on follicular Th cell generation, it may play bigger roles in other settings of immune responses. To address the potential role of PI3K-independent signaling pathways in GVHD, we examined the pathogenic function of ICOS-YF T cells in comparison with ICOS wild-type (WT) and knockout (KO) T cells in the induction of GVHD after BMT. We found that PI3K-independent ICOS signaling mechanisms contribute to T cell costimulation during GVHD. However, CD4 T cells depend more heavily on ICOS-PI3K signaling axis, whereas CD8 T cells are able to induce GVHD utilizing PI3K-independent ICOS signaling mechanisms.

Materials and Methods

Mice

ICOS-KO and ICOS-YF knock-in mice were generated, as previously described (4, 18), using 129 embryonic stem cells, and have been backcrossed 10 generations into C57BL/6 (B6) background. B6 (H-2b) and BALB/c (H-2d) mice were purchased from National Cancer Institute/National Institutes of Health. B6.C-H2bm1 (B6.bm1) and B6.SJL-Ly5a Ptprca Pep3b (B6.Ly5.1) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Mice were housed and bred at H. Lee Moffitt Cancer Center and Research Institute (Tampa, FL). Experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee.

T cell purification and stimulation

T cells were purified from spleen and lymph nodes. Our protocol for T cell purification by negative selection using a magnetic cell separation system (Miltenyi Biotec) has been described previously (9, 19), and the purity of T cells used for transplantation ranged from 91 to 97%. To stimulate T cells, 96-well flat-bottom plates were coated with anti-CD3 mAb (145-2C11; Bio Express) at 0.5 μg/ml in 4°C for overnight and subsequently with anti-ICOS mAb (7E.17; Bio Express) at the indicated concentrations at 37°C for 2 h. Purified T cells were labeled with CFSE and placed on precoated plates. T cell activation and proliferation were determined by measuring expression of CD25, IFN-γ, and CFSE dilution 2 d after stimulation.

Bone marrow transplantation

In nonmyeloablative transplantation models, recipient mice (B6.bm1) were exposed to 600 cGy total body irradiation at 120 cGy/min, a dose range that is immunosuppressive but not lethal for this strain of mice. Purified CD8 T cells from different donors on B6 background were suspended in PBS and injected via the tail vein into 7- to 8-wk-old irradiated recipients within 24 h after irradiation. In myeloablative models, BALB/c mice were irradiated with 800–900 cGy total body irradiation. T cell–depleted (TCD) BM cells from WT B6 donors or in combination with purified T cells from indicated donors were injected via the tail vein to recipients within 24 h after irradiation. Recipient mice were monitored every other day for clinical signs of GVHD, such as weight loss, ruffled fur, hunched back, lethargy, or diarrhea, and mortality. Mononuclear cells from liver and intestine were isolated, as has been described previously (9, 19).

Flow cytometric analysis

Two-, three-, or four-color flow cytometry was performed to measure the expression of surface molecules and intracellular cytokines, according to standard techniques. FITC-labeled anti-CD4, PE-labeled anti-CD4, anti–IFN-γ, anti–TNF-α, IgG isotype control, and Cy-chrome–labeled anti-CD4 were purchased from BD Pharmingen (San Diego, CA). PE-labeled IgG2a was purchased from Caltag (Burlingame, CA). Biotin-labeled anti-Ly5.1 mAb was prepared in our laboratory. Biotinylated Abs were detected with streptavidin Cy-chrome or streptavidin allophycocyanin.

Calcium flux and Western blot

CD8 T cells were purified using a negative selection kit (STEMCELL Technologies) and stimulated with plate-bound anti-CD3 (145-2C11) and anti-CD28 (37.51; eBioscience) for 2 d. Preactivated CD8 T cells were loaded with Indo-1-AM (Life Technologies) and then preincubated with biotinylated Abs against CD3 (0.2 μg/ml) and ICOS (2 μg/ml; C398.4A; eBioscience). Release of intracellular calcium was measured by flow cytometry using LSR (BD Biosciences) before and after addition of avidin (2 mg/ml; EMD Millipore). For Western blot analysis of ICOS-PI3K signaling pathways, preactivated CD8 T cells were expanded in the presence of IL-2 (20 ng/ ml; PeproTech) for additional 3 d. Cells were preincubated with anti-CD3 (1 μg/ml) plus hamster IgG (2 μg/ml; eBioscience) or anti-CD3 (1 μg/ml) plus anti-ICOS (2 μg/ml) for 1 min at room temperature. Subsequently, cells were treated with goat anti-hamster Ab (20 μg/ml) and immediately incubated at 37°C for up to 10 min. Cells were washed and lysed in lysis buffer (50 mM Tris [pH 7.5], 2 mM EDTA, 5 mM Na4P2O7, 100 mM Na3VO4, 5 mM NaF, 150 mM NaCl, and 1% Nonidet P-40), and protein extracts were prepared in Laemmli’s sample buffer. Proteins were separated in 10% SDS-PAGE gel and transferred to Hybond nitrocellulose membrane (GE Healthcare). After incubation with Abs against total Akt or phospho Ser473-Akt (Cell Signaling Technology), proteins were detected using HRP-conjugated goat anti-rabbit detection Ab (Bio-Rad Laboratories) and ECL reagents (GE Healthcare).

Cytokine and histopathological analysis

Blood samples were obtained from BMT recipients at the time specified, and cytokine analysis was performed using a cytometric bead array kit, as described previously (9). Histopathology on liver, lung, and small and large intestines was assessed by an expert pathologist using coded samples, as previously described (20).

Statistical analysis

For comparison of recipient survival among groups in GVHD experiments, the log-rank test was used to determine the statistical significance. To compare the engraftment and expansion of donor T cells, Student t test was used. The p values <0.05 were considered significant.

Results

ICOS-YF T cells induce substantially severe GVHD

We and others have shown that ICOS provides an important costimulation to T cell–mediated alloresponse and GVHD in vivo (6–9). Similar to CD28, ICOS also contains a YMFM motif in the cytoplasmic tail that is phosphorylated upon ligation and activates PI3K (15). Using a knock-in mouse strain, termed ICOS-YF, in which the cytoplasmic tail of ICOS cannot activate PI3K, we have shown that ICOS-PI3K signaling axis is critical for the generation of follicular Th cells (4). In this study, we examined the role of ICOS-medicated PI3K signal by testing the ability of ICOS-YF mutant T cells in the induction of GVHD using a MHC-mismatched BMT model: B6 (H2b)→BALB/c (H2d). T cells were purified from ICOS-WT, YF, or KO donors in B6 background, and their ability to induce acute GVHD was compared. WT T cells caused severe GVHD as all the recipients died within 40 d after BMT (Fig. 1). As expected, the recipients of ICOS-KO T cells had significantly reduced GVHD mortality (p < 0.0001) and prolonged survival than those of ICOS WT T cells. However, ICOS-YF T cells induced significantly more severe GVHD than KO T cells (p < 0.001), although less than WT T cells (p = 0.002). GVHD severity was also reflected by pathologic analysis, which showed that pathology scores were significantly lower in lung, liver, small intestine, and colon of the recipients transplanted with KO T cells than those with WT T cells (Fig. 1C). Furthermore, the recipients of YF T cells had significantly less injuries in lung, liver, and small intestine than those of WT T cells, but significantly more injuries in small bowel than those of KO T cells. Taken together, these data suggest that both PI3K-dependent and independent signals are required for optimal ability of T cells in the induction of acute GVHD.

FIGURE 1.
FIGURE 1.

ICOS-YF T cells retain the ability to induce acute GVHD. Lethally irradiated (800 cGy) BALB/c mice were transplanted with 5 × 106 B6 WT TCD-BM without or with 1 × 106 purified total T cells from ICOS-WT, YF, or KO mice in B6 background. Overall survival (A) and the percentage of original body weight over time (B) are depicted. The data are accumulated from three replicate experiments with 15–18 mice per group. BMT was set up as in (A) and (B), and liver, lung, small intestine, and colon samples were collected at 14 d after BMT for pathologic analysis (C). Pathological scores are presented individually using a semiquantitative scoring system, as described in Materials and Methods. The data are accumulated from two independent experiments with eight mice per group.

ICOS signaling affects T cell migration and infiltration into GVHD target organs

To elucidate potential mechanisms by which ICOS-WT, YF, and KO T cells induced GVHD of varying degrees, we evaluated T cell expansion and migration. The recipients of WT T cells had significantly fewer numbers of total spleen cells (23.3 ± 18.4 × 106/mouse, n = 8) than those of YF T cells (44.4 ± 15.6 × 106/mouse, n = 8) (p = 0.02), and the recipients of YF T cells had significantly fewer numbers of spleen cells than those of KO T cells (64.8 ± 12.2 × 106/mouse, n = 7) (p = 0.05). Because the total number of splenocytes reflects the efficiency of immune cell reconstitution by the donor BM cells in the recipient and it is known that this process is negatively correlated with the severity of GVHD, these data are consistent with the differential severity of GVHD induced by ICOS mutant T cells. However, the percentages of donor T cells were not different across ICOS genotypes (data not shown). In terms of absolute numbers, there were significantly fewer WT donor CD8 T cells than YF or KO counterparts in recipient spleen (p < 0.01), and the same trend was also observed for CD4 donor T cells, although the differences were not statistically significant (Fig. 2A, 2B). These data indicate that ICOS mutations in the donor T cells did not reduce their ability to expand in response to alloantigens in vivo.

FIGURE 2.
FIGURE 2.

ICOS-mediated signals affect T cell migration and infiltration. Lethally irradiated (800 cGy) BALB/c mice were transplanted with 5 × 106 WT TCD-BM from Ly5.1+ B6 donors without or with 1 × 106 purified total T cells from ICOS-WT, YF, or KO mice in B6 background (Ly5.1−). Recipient spleen, liver, and small intestine were collected at 14 d after BMT. Mononuclear cells isolated from these organs were counted and stained for expression of H-2Kb (donor marker), Ly5.1 (marker for donor BM-derived cells), CD4, CD8, and α4β7. Absolute numbers of donor CD4 (CD4+H-2KbLy5.1−) and CD8 (CD8+H-2KbLy5.1−) cells were calculated and displayed in spleen (A and B), liver (D and E), and small intestine (G and H). Percentages of α4β7+ cells gated on donor CD4 or CD8 cells are shown in spleen (C) and liver (F). The score of lymphocyte infiltration was shown in recipient small intestine (I). The data are obtained from one of two replicate experiments with four mice per group in each experiment (A–H), but from two pooled experiments with eight mice per group (I).

WT T cells made significantly more pathologic injuries than YF or KO T cells in all GVHD target organs tested. Although YF T cells were more pathogenic than KO T cells, the pathologic injuries caused by YF or KO T cells were not significantly different, with an exception in small intestine (Fig. 1C). Given the current work focuses on ICOS-mediated PI3K signal, we therefore were particularly interested in examining T cell infiltration in small intestine while looking at liver as a control site. There were significantly more WT donor CD8 T cells than YF or KO counterparts in recipient liver (p < 0.05), and the same trend was also observed for CD4 donor T cells, although the differences were not statistically significant (Fig. 2D, 2E). Numbers of YF and KO donor CD4 or CD8 T cells were also comparable in liver. However, there were significantly more WT donor CD4 and CD8 T cells than YF or KO counterparts in recipient small intestine (p < 0.05), and YF CD4 and CD8 T cells were slightly more than those of KO T cells, although the differences were not statistically significant (Fig. 2G, 2H). Furthermore, an independent semiqualitative measurement indicated that significantly more YF T cells infiltrated into the small intestine compared with KO T cells (Fig. 2I), but not into the liver (data not shown).

To further understand T cell infiltration, we measured chemokine receptors (CXCR3, CCR4, and CCR6) and integrin (α4β7, a molecule allowing T cell migration into gut) on donor T cells. Whereas expressions of chemokine receptors were comparable (data not shown), we found that WT T cells expressed significantly higher levels of integrin α4β7 than YF or KO T cells in recipient spleen and liver (Fig. 2C, 2F). The expression of integrin α4β7 was very low on the donor T cells infiltrated in gut regardless of cell type (data not shown), consistent with downregulation of this molecule in the gut.

ICOS signaling affects cytokine production in vivo

To further understand the potential mechanisms by which ICOS-WT, YF, and KO T cells induced GVHD of varying degrees, we compared cytokine expression levels within individual T cells as well as cytokine levels in the serum of recipient mice. Given their importance in GVHD, we chose to measure IFN-γ and TNF-α. Per cell basis, we found that the percentage of IFN-γ+ cells was significantly reduced in ICOS-KO CD8 T cells than in WT or YT T cells (Fig. 3B), but there was no difference among CD4 T cells (Fig. 3A). Comparable levels of TNF-α expression were found in different subsets of T cells (data not shown). In the serum of the recipient mice, we observed that the recipients of WT T cells had significantly higher levels of IFN-γ and TNF-α than those that received YF or KO T cells with no significant difference between YF and KO groups (Fig. 3C, 3D). There were no significant differences in the production of IL-4, IL-6, IL-10, or IL-17A in three groups (data not shown). The cytokine profiles do not provide clear-cut explanations to the intermediate severity of GVHD induced by ICOS-YF T cells. However, it is possible that the intact level of IFN-γ–producing ICOS-YF CD8 T cells contributed to the substantial pathogenicity above ICOS-KO T cells.

FIGURE 3.
FIGURE 3.

Cytokine profiles induced by ICOS mutant T cells. BMT was set up as in Fig. 1. Peripheral blood and spleens were collected from each recipient at 14 d after BMT. Spleen cells were stained for expression of H-2Kb (donor marker), CD4, CD8, and intracellular IFN-γ. (A) Percentages of IFN-γ+ cells are shown on gated donor CD4 or CD8 cells (B). The levels of IFN-γ (C) and TNF-α (D) in the serum of the recipient mice were measured using cytometric bead array kits. The data are accumulated from two independent experiments with four to eight mice per group.

CD4 T cells are heavily dependent on ICOS-PI3K signaling axis for induction of GVHD

Previous studies by us and others showed that ICOS contributes to the development of GVHD particularly by CD4 T cells (6, 8, 9). In this study, we asked the role of PI3K-dependent pathways through ICOS in GVHD induced by CD4 T cells alone. ICOS-WT T cells induced significantly more severe GVHD than YF or KO T cells (p < 0.05), but YF and KO T cells induced GVHD in a similar severity (p = 0.5) (Fig. 4). These data indicate that PI3K-dependent but not PI3K-independent signals primarily contribute to ICOS-mediated costimulation in GVHD induced by CD4 T cells.

FIGURE 4.
FIGURE 4.

ICOS-YF CD4 T cells are severely impaired in GVHD pathogenic potential. Lethally irradiated (800 cGy) BALB/c mice were transplanted with 5 × 106 WT B6 TCD-BM without or with 1 × 106 purified CD4 T cells from ICOS-WT, YF, or KO mice in B6 background. Overall survival (A) and the percentage of original body weight over time (B) are depicted. The data are accumulated from two replicate experiments with 10–12 mice per group.

ICOS-YF CD8 T cells can induce substantial GVHD

We next examined the capacity of ICOS-YF CD8 T cells to induce GVHD. Because CD8 T cells alone are unable to induce severe GVHD in B6→BALB/c BMT model, a small number of WT CD4 T cells was added into CD8 T cells isolated from ICOS-WT, YF, or KO B6 mice. Under this condition, WT and YF CD8 T cells were comparable to induce GVHD (44 and 55% survival, respectively), whereas CD8 KO T cells had reduced ability to cause GVHD (80% survival) (Fig. 5A). Although recipient survival rate did not reach statistical significance between ICOS-YF and ICOS-KO T cells, their body weight changes were distinctively different between genotypes with the recipients of ICOS-YF being intermediate (p < 0.001) (Fig. 5B). To exclusively test the ability of CD8 T cells in the induction of GVHD, we used an MHC I-mismatched B6→B6.bm1 transplant model in which only CD8 T cells can recognize alloantigens. Under this condition, we observed that WT CD8 T cells induced significantly more severe GVHD than KO T cells (p < 0.01). ICOS-YF T cells caused slightly lower rate of GVHD lethality than WT T cells, but this was not statistically significant (Fig. 5C, 5D). Taken together, these data indicate that, in acute GVHD models, ICOS-YF CD4 T cells behave like ICOS-KO CD4 T cells, whereas ICOS-YF CD8 T cells rather resemble WT CD8 T cells, suggesting a differential role of ICOS-PI3K signaling in different T cell compartments.

FIGURE 5.
FIGURE 5.

ICOS-YF CD8 T cells induce moderate GVHD. Lethally irradiated (800 cGy) BALB/c mice were transplanted with 5 × 106 TCD-BM alone or after supplementing with 0.2 × 106 WT CD4 T cells plus 1 × 106 CD8 T cells from ICOS-WT, YF, or KO mice. Overall survival (A) and the percentage of original body weight over time (B) are depicted. The data are accumulated from two independent experiments with 9–10 mice per group. For bm1 model, sublethally irradiated (600 cGy) B6.bm1 mice were transplanted with 1 × 106 CD8 T cells from ICOS-WT, YF, or KO mice. Overall survival (C) and the percentage of original body weight over time (D) are depicted. The data are accumulated from three independent experiments with 14–16 mice per group.

PI3K-independent signals contribute to ICOS-mediated costimulation in T cell activation and proliferation

PI3K is a common signal transducer used by CD28 and ICOS. However, CD28-mediated PI3K activation is dispensable for T cell proliferation and germinal center formation in vivo (21). Using ICOS-YF mice, we have shown that CD4 T cells lost the ability to potentiate TCR-mediated PI3K activation. However, CD4 T cells from ICOS-YF mice still retained the ability to augment TCR-mediated calcium mobilization (4). We hypothesized that ICOS-calcium signaling axis is also intact in CD8 T cells, and this may explain the substantial pathogenic potential of CD8 T cells in the settings of GVHD. Indeed, coligation of ICOS substantially augmented TCR-mediated AKT phosphorylation, a hallmark of PI3K activation in ICOS-WT CD8 T cells (Fig. 6A; WT). However, this PI3K potentiation effect was abrogated in ICOS-YF CD8 T cells (Fig. 6A; YF). Importantly, ICOS-YF CD8 T cells retained the ability to potentiate TCR-mediated calcium flux, although the magnitude of calcium signaling was partially reduced compared with that of ICOS-WT CD8 T cells (Fig. 6B). Thus, similar to CD4 T cells, CD8 T cells retain PI3K-independent calcium signaling capacity.

FIGURE 6.
FIGURE 6.

ICOS-YF can potentiate TCR-mediated calcium flux in CD8 T cells. Preactivated CD8 T cells were prepared from ICOS-WT, YF, and KO mice, and their ICOS-PI3K signaling (A) and ICOS-calcium signaling (B) capacities were examined, as described in Materials and Methods. Data shown are representative of three independent experiments.

We then tested the role of PI3K signaling in ICOS-mediated costimulation in T cell activation and proliferation in CD4 and CD8 T cells. To this end, purified T cells from ICOS-WT, YF, or KO mice were labeled with CFSE and stimulated with a suboptimal dose of anti-CD3 mAb in the presence of anti-ICOS mAb at various concentrations. Two days after stimulation, T cell activation and proliferation were measured based on CD25 expression and CFSE dilution (Fig. 7A). As expected, additional stimulation with anti-ICOS increased activation and proliferation of WT T cells in a dose-dependent manner for both CD4 and CD8 T cells (Fig. 7C, 7D). In contrast, additional stimulation with anti-ICOS had no effect on KO T cells. However, YF T cells also responded to anti-ICOS stimulation, although to lower extents relative to WT T cells (Fig. 7C, 7D). In addition to CD25 expression, intracellular IFN-γ expression was also measured along with CFSE dilution (Fig. 7B). We found that YF CD4 T cells failed to respond to ICOS stimulation similarly to KO cells (Fig. 7E), whereas YF CD8 T cells responded to ICOS stimulation robustly as WT cells (Fig. 7F). These in vitro results indicate that PI3K-independent signaling mechanisms are able to support ICOS-mediated costimulation, yet distinctly on CD4 versus CD8 T cells.

FIGURE 7.
FIGURE 7.

Contribution of PI3K signaling to ICOS costimulation in T cell activation and proliferation in vitro. Purified T cells from ICOS-WT, YF, or KO mice were labeled with CFSE and stimulated with immobilized anti-CD3 at 0.5 μg/ml in the presence of anti-ICOS at the concentrations indicated. Two days after stimulation, cells were harvested and stained for surface expression of CD4, CD8, and CD25, and intracellular expression of IFN-γ. (A) For WT T cells, CD4 and CD8 expression on gated live cells (left panels); CFSE profile and CD25 expression on gated live CD4 (middle) or CD8 cells (right). (B) The experiment was done as in (A); CFSE profile and CD25 expression are presented. Percentages of CD25+ and CFSE-diluted cells (upper left quadrant) are depicted on gated live CD4 (C) or CD8 (D) cells of ICOS-WT, YF, or KO genotype. Percentages of IFN-γ+ and CFSE-diluted cells are depicted on gated live CD4 (E) or CD8 (F) cells of ICOS-WT, YF, or KO genotype. Data shown are from a representative experiment of three independent experiments.

Discussion

In this study, by using ICOS-YF knock-in mice, we show that both PI3K-dependent and independent signals contribute to ICOS-mediated costimulation in T cell activation and proliferation in vitro as well as T cell alloresponse and GVHD in vivo. PI3K-dependent signals primarily contribute to overall production of inflammatory cytokines, but PI3K-independent signals can support IFN-γ production by CD8 T cells in vivo. Furthermore, we found that, in CD4 T cells, PI3K signaling is essentially for ICOS-mediated costimulation, whereas CD8 T cells can use PI3K-dependent and -independent pathways to exert their functions during GVHD progression in vivo. Because ICOS acts in concert with other costimulatory molecules (such as CD28 among others) to promote T cell activation and function, and others besides ICOS may also activate PI3K pathways, the current work highlights the contribution of ICOS-mediated PI3K-dependent and -independent signals in the context of other costimulations in T cell activation, migration, cytokine production, and pathogenicity in the induction of GVHD.

Similar to CD28, ICOS can activate PI3K through its Tyr-based motif (YMFM) in the cytoplasmic tail. However, unlike CD28, YMFM is the only motif known for ICOS, and PI3K and calcium are the only signal pathways known to be activated through ICOS (4, 22). Using T cells bearing YF mutation known to abrogate ICOS-mediated PI3K recruitment, we clearly demonstrated that PI3K signal pathway is not the only transducer of ICOS-mediated costimulation to T cell activation and proliferation in vitro (Fig. 7). Although the cytoplasmic motif or signal transducers remain to be identified, we showed that PI3K recruitment was not required for Ca2+ flux in either CD4 or CD8 T cells (Fig. 6), and both PI3K-dependent and independent signals mediated by ICOS contribute to CD25 expression and proliferation of both CD4 and CD8 T cells (Fig. 7C, 7D). However, ICOS-mediated PI3K activation is the main contributor to IFN-γ production by CD4 T cells, whereas ICOS-mediated PI3K-independent signal(s) is the primary contributor to IFN-γ production by CD8 T cells (Fig. 7E, 7F). The precise mechanisms by which ICOS-mediated PI3K activation plays distinct role in CD4 versus CD8 T cells are yet to be elucidated.

Similar to the studies in vitro, we found that substantial ICOS function remained when its ability to activate PI3K was selectively abrogated in vivo because ICOS-YF T cells caused significantly more severe GVHD than KO T cells (Fig. 1). Several groups, including ours, observed that ICOS KO T cells have reduced ability to cause acute GVHD. Although the underlying mechanisms have been somewhat controversial, reduction of Th1 cytokine production was a common finding by these groups (6–9). In current study, we compared the ability of donor T cell expansion and migration. Comparable (CD4) or even more (CD8) of either YF or KO T cells were found in recipient spleen as compared with WT T cells (Fig. 2A, 2B). Consistent with published results by us and others, the current data indicate that ICOS mutations did not affect the donor T cell expansion in response to alloantigens in vivo. In contrast, we found that less YF or KO T cells infiltrated into recipient liver (Fig. 2E) and small intestine (Fig. 2G, 2H). Given that WT T cells expressed significantly higher levels of integrin α4β7 than YF or KO T cells, we interpret that ICOS-mediated PI3K activation significantly contributes to upregulation of integrin α4β7, which facilitates T cell migration to liver and small intestine in particular.

We observed that Th1 cytokines, including IFN-γ and TNF-α, were significantly reduced in the recipients of ICOS-KO T cells compared with those of WT T cells (Fig. 3). Furthermore, the levels of IFN-γ and TNF-α in the recipients of ICOS-YF T cells were similarly reduced as those of KO T cells, indicating that PI3K signals were required for overall Th1 cytokine production potentiated by ICOS costimulation. However, IFN-γ production at single cell level was comparable on three types of CD4 T cells at the time tested (Fig. 3A), which seemed to be contradictory to in vitro results (Fig. 7E), in which both YF and KO CD4 T cells produced significantly lower levels of IFN-γ. Given ICOS was only costimulatory signal in the study in vitro, whereas other costimulatory signals were present in vivo, it is likely that other signals (e.g., CD28) compensate for the loss of ICOS signaling. IFN-γ production at single cell level was intact on ICOS-YF, but not in KO CD8 cells as compared with WT counterparts (Fig. 3B), which is consistent with the in vitro results (Fig. 7F). We thus interpret that ICOS is critical for IFN-γ production by CD8 T cells, but PI3K signal mediated by ICOS is dispensable. We further reason that IFN-γ production, and presumably other CTL effector functions as well (23), mediated by PI3K-independent signals, may contribute to the different capacities of YF versus KO T cells in the induction of GVHD particularly in small intestine.

Unlike intracellular cytokine expression, we found that the levels of IFN-γ and TNF-α in serum were similarly reduced in the recipients of YF or KO T cells compared with those of WT T cells (Fig. 3C, 3D). The discrepancy in cytokine levels in donor T cells versus recipient serum is most likely due to the fact that cytokines in serum result from multiple sources. Besides T cells from WT, YF, or KO donors, host hematopoietic (e.g., APCs) or even nonhematopoietic cells can also contribute to the serum cytokines in the recipients. Furthermore, altered ICOS signaling in donor T cells might indirectly promote IFN-γ and TNF-α production by host cells. Because ICOS-YF CD8, but not CD4, cells had an intact ability to produce IFN-γ as WT cells (Fig. 3A, 3B) and because Th1 cytokines are critical in the development of GVHD, we tested whether PI3K-mediated signals contributed to CD4 versus CD8 T cells differentially in the induction of GVHD. Indeed, we found that ICOS-YF CD4 T cells had a reduced ability to induce GVHD to a similar extent as KO CD4 T cells (Fig. 4), whereas ICOS-YF CD8 T cells had a substantial residual capacity to induce GVHD compared with ICOS-KO CD8 T cells (Fig. 5). These data again suggest that CD8 T cells were less dependent on ICOS-mediated PI3K signals for their optimal function to induce GVHD compared with CD4 T cells in vivo. Consistently, we have shown that generation of follicular Th cells, another subset of CD4 T cells, is severely impaired when ICOS-PI3K signaling is selectively abrogated (4). Interestingly, in the mixture of CD4 and CD8 T cells, ICOS-mediated PI3K-independent signals made a significant contribution to overall T cell responses both in vitro and in vivo (Figs. 1, 7). As ICOS-potentiated Ca2+ signaling was independent of PI3K activation (Fig. 6B), we attempted to evaluate the role of ICOS-potentiated Ca2+ signaling in T cell response and GVHD induction by blocking Ca2+ signaling using cyclosporine A. However, treatment with cyclosporine in vivo similarly alleviated GVHD induced by WT, YF, or KO T cells (data not shown), which presumably resulted from overall blockade of Ca2+ signaling mediated by collective T cell signals such as those induced by TCR, CD28, and ICOS. Thus, we interpret that Ca2+ influx may be responsible for residual function of ICOS in the absence of PI3K activation, but cannot exclude contributions by other unknown signal(s) as well.

In conclusion, our study shows that ICOS costimulation contributes to T cell responses and GVHD development beyond PI3K-dependent signaling pathways, possibly through Ca2+-dependent pathways. This finding reveals undefined downstream signal transduction mechanisms of ICOS that have potential implications in targeting ICOS for the control of T cell–mediated diseases in clinic.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank Dr. Claudio Anasetti for critical review of this manuscript and Dapeng Wang and Kane Kaosaard for technical assistance. We are grateful for the technical assistance provided by Flow Cytometry and Mouse Core Facilities at the Moffitt Cancer Center.

Footnotes

  • This work was supported in part by National Institutes of Health Grants R01s CA11816, CA143812, and AI 082685 (to X.-Z.Y.) and Canadian Institutes of Health Research Grant MOP-84544 (to W.-K.S.).

  • Abbreviations used in this article:

    BM
    bone marrow
    BMT
    BM transplantation
    GVHD
    graft-versus-host disease
    ICOSL
    ICOS ligand
    KO
    knockout
    TCD
    T cell–depleted
    WT
    wild-type.

  • Received December 20, 2012.
  • Accepted April 26, 2013.

References

  1. ↵
    1. Riley J. L.,
    2. P. J. Blair,
    3. J. T. Musser,
    4. R. Abe,
    5. K. Tezuka,
    6. T. Tsuji,
    7. C. H. June
    . 2001. ICOS costimulation requires IL-2 and can be prevented by CTLA-4 engagement. J. Immunol. 166: 4943–4948.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Yoshinaga S. K.,
    2. J. S. Whoriskey,
    3. S. D. Khare,
    4. U. Sarmiento,
    5. J. Guo,
    6. T. Horan,
    7. G. Shih,
    8. M. Zhang,
    9. M. A. Coccia,
    10. T. Kohno,
    11. et al
    . 1999. T-cell co-stimulation through B7RP-1 and ICOS. Nature 402: 827–832.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Suh W. K.,
    2. A. Tafuri,
    3. N. N. Berg-Brown,
    4. A. Shahinian,
    5. S. Plyte,
    6. G. S. Duncan,
    7. H. Okada,
    8. A. Wakeham,
    9. B. Odermatt,
    10. P. S. Ohashi,
    11. T. W. Mak
    . 2004. The inducible costimulator plays the major costimulatory role in humoral immune responses in the absence of CD28. J. Immunol. 172: 5917–5923.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Gigoux M.,
    2. J. Shang,
    3. Y. Pak,
    4. M. Xu,
    5. J. Choe,
    6. T. W. Mak,
    7. W. K. Suh
    . 2009. Inducible costimulator promotes helper T-cell differentiation through phosphoinositide 3-kinase. Proc. Natl. Acad. Sci. USA 106: 20371–20376.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Fu T.,
    2. Q. He,
    3. P. Sharma
    . 2011. The ICOS/ICOSL pathway is required for optimal antitumor responses mediated by anti-CTLA-4 therapy. Cancer Res. 71: 5445–5454.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Hubbard V. M.,
    2. J. M. Eng,
    3. T. Ramirez-Montagut,
    4. K. H. Tjoe,
    5. S. J. Muriglan,
    6. A. A. Kochman,
    7. T. H. Terwey,
    8. L. M. Willis,
    9. R. Schiro,
    10. G. Heller,
    11. et al
    . 2005. Absence of inducible costimulator on alloreactive T cells reduces graft versus host disease and induces Th2 deviation. Blood 106: 3285–3292.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Li, J., K. Semple, W. K. Suh, C. Liu, F. Chen, B. R. Blazar, and X. Z. Yu. 2011. Roles of CD28, CTLA4, and inducible costimulator in acute graft-versus-host disease in mice. Biol. Blood Marrow Transplant. 17: 962-969.
  8. ↵
    1. Taylor P. A.,
    2. A. Panoskaltsis-Mortari,
    3. G. J. Freeman,
    4. A. H. Sharpe,
    5. R. J. Noelle,
    6. A. Y. Rudensky,
    7. T. W. Mak,
    8. J. S. Serody,
    9. B. R. Blazar
    . 2005. Targeting of inducible costimulator (ICOS) expressed on alloreactive T cells down-regulates graft-versus-host disease (GVHD) and facilitates engraftment of allogeneic bone marrow (BM). Blood 105: 3372–3380.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Yu X. Z.,
    2. Y. Liang,
    3. R. I. Nurieva,
    4. F. Guo,
    5. C. Anasetti,
    6. C. Dong
    . 2006. Opposing effects of ICOS on graft-versus-host disease mediated by CD4 and CD8 T cells. J. Immunol. 176: 7394–7401.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Khayyamian S.,
    2. A. Hutloff,
    3. K. Büchner,
    4. M. Gräfe,
    5. V. Henn,
    6. R. A. Kroczek,
    7. H. W. Mages
    . 2002. ICOS-ligand, expressed on human endothelial cells, costimulates Th1 and Th2 cytokine secretion by memory CD4+ T cells. Proc. Natl. Acad. Sci. USA 99: 6198–6203.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Swallow M. M.,
    2. J. J. Wallin,
    3. W. C. Sha
    . 1999. B7h, a novel costimulatory homolog of B7.1 and B7.2, is induced by TNFalpha. Immunity 11: 423–432.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Liang L.,
    2. E. M. Porter,
    3. W. C. Sha
    . 2002. Constitutive expression of the B7h ligand for inducible costimulator on naive B cells is extinguished after activation by distinct B cell receptor and interleukin 4 receptor-mediated pathways and can be rescued by CD40 signaling. J. Exp. Med. 196: 97–108.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Watanabe M.,
    2. Y. Takagi,
    3. M. Kotani,
    4. Y. Hara,
    5. A. Inamine,
    6. K. Hayashi,
    7. S. Ogawa,
    8. K. Takeda,
    9. K. Tanabe,
    10. R. Abe
    . 2008. Down-regulation of ICOS ligand by interaction with ICOS functions as a regulatory mechanism for immune responses. J. Immunol. 180: 5222–5234.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Ogawa S.,
    2. G. Nagamatsu,
    3. M. Watanabe,
    4. S. Watanabe,
    5. T. Hayashi,
    6. S. Horita,
    7. K. Nitta,
    8. H. Nihei,
    9. K. Tezuka,
    10. R. Abe
    . 2001. Opposing effects of anti-activation-inducible lymphocyte-immunomodulatory molecule/inducible costimulator antibody on the development of acute versus chronic graft-versus-host disease. J. Immunol. 167: 5741–5748.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Coyle A. J.,
    2. S. Lehar,
    3. C. Lloyd,
    4. J. Tian,
    5. T. Delaney,
    6. S. Manning,
    7. T. Nguyen,
    8. T. Burwell,
    9. H. Schneider,
    10. J. A. Gonzalo,
    11. et al
    . 2000. The CD28-related molecule ICOS is required for effective T cell-dependent immune responses. Immunity 13: 95–105.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Arimura Y.,
    2. H. Kato,
    3. U. Dianzani,
    4. T. Okamoto,
    5. S. Kamekura,
    6. D. Buonfiglio,
    7. T. Miyoshi-Akiyama,
    8. T. Uchiyama,
    9. J. Yagi
    . 2002. A co-stimulatory molecule on activated T cells, H4/ICOS, delivers specific signals in T(h) cells and regulates their responses. Int. Immunol. 14: 555–566.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Parry R. V.,
    2. C. A. Rumbley,
    3. L. H. Vandenberghe,
    4. C. H. June,
    5. J. L. Riley
    . 2003. CD28 and inducible costimulatory protein Src homology 2 binding domains show distinct regulation of phosphatidylinositol 3-kinase, Bcl-xL, and IL-2 expression in primary human CD4 T lymphocytes. J. Immunol. 171: 166–174.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Tafuri A.,
    2. A. Shahinian,
    3. F. Bladt,
    4. S. K. Yoshinaga,
    5. M. Jordana,
    6. A. Wakeham,
    7. L. M. Boucher,
    8. D. Bouchard,
    9. V. S. Chan,
    10. G. Duncan,
    11. et al
    . 2001. ICOS is essential for effective T-helper-cell responses. [comment] Nature 409: 105–109.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Yu X. Z.,
    2. P. J. Martin,
    3. C. Anasetti
    . 1998. Role of CD28 in acute graft-versus-host disease. Blood 92: 2963–2970.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Liang Y.,
    2. C. Liu,
    3. J. Y. Djeu,
    4. B. Zhong,
    5. T. Peters,
    6. K. Scharffetter-Kochanek,
    7. C. Anasetti,
    8. X. Z. Yu
    . 2008. Beta2 integrins separate graft-versus-host disease and graft-versus-leukemia effects. Blood 111: 954–962.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    1. Dodson L. F.,
    2. J. S. Boomer,
    3. C. M. Deppong,
    4. D. D. Shah,
    5. J. Sim,
    6. T. L. Bricker,
    7. J. H. Russell,
    8. J. M. Green
    . 2009. Targeted knock-in mice expressing mutations of CD28 reveal an essential pathway for costimulation. Mol. Cell. Biol. 29: 3710–3721.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Rudd C. E.,
    2. A. Taylor,
    3. H. Schneider
    . 2009. CD28 and CTLA-4 coreceptor expression and signal transduction. Immunol. Rev. 229: 12–26.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Finlay D. K.,
    2. E. Rosenzweig,
    3. L. V. Sinclair,
    4. C. Feijoo-Carnero,
    5. J. L. Hukelmann,
    6. J. Rolf,
    7. A. A. Panteleyev,
    8. K. Okkenhaug,
    9. D. A. Cantrell
    . 2012. PDK1 regulation of mTOR and hypoxia-inducible factor 1 integrate metabolism and migration of CD8+ T cells. J. Exp. Med. 209: 2441–2453.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section