Fine-tuning of antigen-specific immune responses by regulatory T cells activated via antigen recognition-independent and humoral factor-dependent mechanisms.

CD4+ CD25+ Foxp3+ regulatory T cells (Tregs) are highly sensitive to IL-2, one of the many cytokines produced during immune responses, for their development, survival and functions. Although the effects of IL-2 administration on Tregs in vivo are well characterized, the effects on Tregs elicited by IL-2 produced during an immune response have not been elucidated. Hence, in this study, Treg behaviour during IL-2-producing immune responses was explored using in vivo and in vitro murine systems. The use of murine mixed lymphocyte culture (MLC) revealed that a large proportion of Tregs increased in size, accompanied by both cell death and proliferation status. Further, these large Tregs, which were found to not recognize specific antigens, were observed in MLCs as being functionally activated by various cytokines, including IL-2, produced by antigen-specific T cells. This 'bystander Treg activation' was also observed in mice with graft-versus-host reactions (GvHRs). Alternatively, effector cells from Treg-depleted MLCs exhibited lower antigen-specific responses or higher cross-reactivity as compared to control MLCs with Tregs. Taken together, these results suggest that Tregs are activated by cytokines, mainly IL-2, released from T cells that are activated by a specific antigen. Subsequently, these activated bystander Tregs contribute to the fine-tuning of highly antigen-specific immune responses.

IL-21 and IL-5 coordinately induce surface IgA+ cells.

Intestinal IgA is induced by microbes and food antigens. Peyer's patches (PPs) are known as one of the inductive sites for intestinal IgA production. However, the precise mechanism of IgA induction is as yet unknown. IgA secretion was induced from IgD+ B cells in vitro by stimulus with lipopolysaccharide in the presence of only retinoic acid (RA) and low doses of TGF-ß1. Surface IgA+ cells were effectively induced from IgD+ B cells in vitro by the mixture of RA and the cytokines TGF-ß1, APRIL, IL-5 and IL-21. rIL-21 upregulated surface IgA+ but impaired the proliferation of stimulated B cells in the presence of rTGF-ß1, RA and rAPRIL, in vitro. The addition of rIL-5 restored the impaired proliferation by rIL-21, resulting in the expansion of IgA+ cells. rIL-21 induced the expression of Aicda and Prdm1, and impaired Rel in IgD+ B cells. Blockade of IL-21R signaling by a neutralizing mAb in vivo led to lower frequencies of IgA+ and IgG2b+ cells and lower germinal center B cells in PPs in a homeostatic condition. Although amounts of small intestinal IgA and titers of anti-dsDNA, the major target of intestinal IgA, in these mice were not altered, anti-OVA IgA titers induced by OVA drinking in OVA-specific T-cell receptor (TCR) transgenic mice were decreased. PP-deficient TCR transgenic mice showed diminished anti-OVA IgA induction. Blockade of IL-5R signaling in vivo led to similar results with relatively weaker effects than that of IL-21R mAb administration. These results suggest that IL-21 and IL-5 play cooperative roles in surface expression of IgA in PPs.

Peyer's patches contain abundant isotype-switched B cells with activated phenotypes and are inductive sites for T-independent anti-DNA IgA.

Peyer's patches (PPs) are inductive sites for IgA production; however, the induction mechanism of IgA remains largely unknown. We found that the activated phenotypes of isotype-switched PP B cells were more abundant than those of splenic B cells. Immunoglobulins (Igs) from PP B cells reacted to several substances, including DNA and diet extract. Hybridomas established from PP B cells of untreated mice revealed that IgA mainly react with DNA. PP-deficient mice revealed that PPs were dispensable for a total intestinal IgA amount but were required for intestinal anti-diet extract and anti-DNA IgA. Antibiotic-treated mice and CD4+ T cell-depleted mice demonstrated that the intestinal anti-DNA IgA was induced by microbiota in a T-independent manner. Interestingly, the oral administration of IgA led to the expansion of intestinal bacteria in a reactivity-independent manner. Our findings suggest that PPs are unique and efficient inductive sites for IgA, particularly against T-independent antigens.

Tumor necrosis factor superfamily member (TNFSF) 13 (APRIL) and TNFSF13B (BAFF) downregulate homeostatic immunoglobulin production in the intestines.

Intestinal immunoglobulins (Igs) protect against microbes. However, the regulation of intestinal Ig production is poorly understood. In this study, we have investigated the roles of APRIL (tumor necrosis factor superfamily member [TNFSF] 13) and BAFF (TNFSF13B) in intestinal Ig induction. Peyer's patches (PPs) are, at least in part, an inductive site for Igs, including IgA. Introducing APRIL and BAFF in vivo lowered the frequency of IgG1+ or IgG2b+ B cells in PPs. Administration of TACI-Fc upregulated the frequency of IgG1+, IgG2b+, and IgA+ B cells in PPs, suggesting that APRIL and BAFF attenuate Ig production in these regions. TACI-Fc also upregulated intestinal IgA levels and expanded germinal center B cells in PPs. These results indicate that APRIL and BAFF paradoxically downregulate homeostatic Ig production in the intestines.

Germinal Center Hypoxia Potentiates Immunoglobulin Class Switch Recombination.

Germinal centers (GCs) are anatomic sites where B cells undergo secondary diversification to produce high-affinity, class-switched Abs. We hypothesized that proliferating B cells in GCs create a hypoxic microenvironment that governs their further differentiation. Using molecular markers, we found GCs to be predominantly hypoxic. Compared to normoxia (21% O2), hypoxic culture conditions (1% O2) in vitro accelerated class switching and plasma cell formation and enhanced expression of GL-7 on B and CD4+ T cells. Reversal of GC hypoxia in vivo by breathing 60% O2 during immunization resulted in reduced frequencies of GC B cells, T follicular helper cells, and plasmacytes, as well as lower expression of ICOS on T follicular helper cells. Importantly, this reversal of GC hypoxia decreased Ag-specific serum IgG1 and reduced the frequency of IgG1+ B cells within the Ag-specific GC. Taken together, these observations reveal a critical role for hypoxia in GC B cell differentiation.

Peyer's patch innate lymphoid cells regulate commensal bacteria expansion.

Anatomical containment of commensal bacteria in the intestinal mucosa is promoted by innate lymphoid cells (ILCs). However, the mechanism by which ILCs regulate bacterial localization to specific regions remains unknown. Here we show that Peyer's patch (PP) ILCs robustly produce IL-22 and IFN-γ in the absence of exogenous stimuli. Antibiotic treatment of mice decreased both IL-22+ and IFN-γ+ cells in PPs. Blockade of both IL-2 and IL-23 signaling in vitro lowered IL-22 and IFN-γ production. PP ILCs induced mRNA expression of the antibacterial proteins RegIIIß and RegIIIγ in intestinal epithelial cells. Furthermore, in vivo depletion of ILCs rather than T cells altered bacterial composition and allowed bacterial proliferation in PPs. Collectively, our results show that ILCs regulate the expansion of commensal bacteria in PPs.

IL-33 activates eosinophils of visceral adipose tissue both directly and via innate lymphoid cells.

Eosinophils are multifunctional leukocytes involved in allergic reactions as well as adipose tissue regulation. IL-5 is required for eosinophil survival; however, the in vivo mechanisms of eosinophil regulation are not fully understood. A tg mouse model with il5 promoter-driven EGFP expression was established for detecting the IL-5-producing cells in vivo. Il5-egfp tg mice expressed high levels of EGFP in gonadal adipose tissue (GAT) cells. EGFP(+) cells in GAT were mainly group 2 innate lymphoid cells (ILCs). IL-33 preferentially expanded EGFP(+) cells and eosinophils in GAT in vivo. EGFP(+) ILCs were found to upregulate prg2 mRNA expression in GAT eosinophils. These results demonstrate that ILCs activate eosinophils in GAT. The blockage of IL-33Rα, on the other hand, did not impair EGFP(+) ILC numbers but did impair eosinophil numbers in vivo. GAT eosinophils expressed IL-33Rα and IL-33 expanded eosinophil numbers in CD90(+) cell-depleted mice. IL-33 was further observed to induce the expression of retnla and epx mRNA in eosinophils. These findings demonstrate that IL-33 directly activates eosinophils in GAT, and together with our other findings described above, our findings show that IL-33 has dual pathways via which it activates eosinophils in vivo: a direct activation pathway and a group 2 ILC-mediated pathway.

NFκB attenuates IL-5 production and upregulates T-box transcription factors in Th2-like T cells.

IL-5 plays important roles in eosinophil differentiation, expansion, and recruitment. The regulation of IL-5 seems critical for the treatment of eosinophil-mediated allergic reactions. However, the precise mechanisms for IL-5 regulation remain unknown. In this study, we investigated how IL-5 production is regulated. The transduction of GATA-3 into a murine T cell hybridoma resulted in acquiring the ability to produce IL-5 in response to an antigenic stimulus like Th2 cells. This production was dependent on the cAMP-PKA pathway, but not on p38 activation. Transduction of NIK largely impaired IL-5 production. RelA and RelB similarly impaired IL-5 production. RelA decreased not only IL-5 protein amount but mRNA. RelA also inhibited Il5-luciferase reporter activity. The transduction of GATA-3 decreased the expression of Tbx21 and Eomes, but the additional transduction of RelA abrogated the decreased expression of GATA-3-induced Tbx21 and Eomes. Furthermore, the transduction of T-bet or Eomes into the GATA-3-transduced T cell hybridoma impaired IL-5 production. These results suggested that strong enhancement of the NFκB pathway downregulates IL-5 production and upregulates T-box protein expression to shift an immune response from Th2 to inflammatory Th1.

CD2-mediated regulation of peripheral CD4(+) CD25(+) regulatory T-cell apoptosis accompanied by down-regulation of Bim.

Extensive studies on CD4(+)  CD25(+) regulatory T (Treg) cells suggest that they are important in regulating immune responses. However, mechanisms of peripheral Treg cell homeostasis are unknown. We found that stromal cells isolated from secondary lymphoid organs such as spleen and lymph nodes could support the survival of Treg cells. This was dependent on CD2 engagement and a direct interaction between Treg cells and stromal cells. In the presence of stromal cells, Bim, a pro-apoptotic factor, was partially decreased in Treg cells. This effect could be inhibited by anti-CD2 blocking antibodies, indicating that stimulation through CD2 on Treg cells regulates Bim expression, which may be relevant to Treg cell apoptosis. Therefore, Treg cell interactions with stromal cells through CD2 may be essential for Treg cell survival. Surprisingly, the expression of CD2 ligands on stromal cells was not detected. Hence, it is not clear how CD2 on Treg cells contributes to a direct interaction with the stromal cells and participates in survival support for Treg cells. Taken together, CD2 stimuli were mandatory for Treg cell survival with reduced Bim expression, but CD2 may not function as a direct receptor for molecules on stromal cells.

Naïve CD4+ T cells of Peyer's patches produce more IL-6 than those of spleen in response to antigenic stimulation.

Peyer's patches (PPs) are potential sites where specific mucosal immune responses and oral tolerance are induced. The unique features of these immune responses are thought to occur in micromilieu and are largely affected by antigen-presenting cells (APCs) such as dendritic cells. In this study, we investigated the cytokine profiles induced by the activation of CD4(+) T cells of PPs. PP cells from TCR transgenic mice secreted greater amounts of IL-5 and IL-6 than spleen cells after antigenic stimulation. IL-5 was mainly produced by PP non-T cells, whereas IL-6 was secreted by PP CD4(+) cells. PPs contained two major populations including naïve and memory/activated CD4(+) cells; both populations secreted IL-6 upon activation. We also found that CD4(+)/CD62L(hi) naïve cells from PPs secreted a greater amount of IL-6 after stimulation than those from the spleen. Furthermore, subtraction and qPCR analyses revealed that PP CD4(+)/CD62L(hi) cells express a greater amount of transcripts of GA-binding protein ß subunit 1 than those of the spleen. These results suggest that naïve T cells as well as non-T cells and activated/memory T cells from PPs are distinct from their splenic counterparts and thus cause unique immune responses the in intestine.

Notch2 signaling is required for potent antitumor immunity in vivo.

CD8(+) T cells play a central role in cancer immunosurveillance, and the efficient induction of CTLs against tumor Ags is required for successful immunotherapy for cancer patients. Notch signaling directly regulates the transcription of effector molecules in CTLs. However, it remains unclear whether Notch signaling in CD8(+) T cells is required for antitumor CTL responses and whether modulation of Notch signaling can augment antitumor CTL responses. In this study, we demonstrate that signaling by Notch2 but not Notch1 in CD8(+) T cells is required for antitumor CTL responses. Notch2(flox/flox) mice crossed with E8I-cre transgenic (N2F/F-E8I) mice, in which the Notch2 gene is absent only in CD8(+) T cells, die earlier than control mice after inoculation with OVA-expressing EG7 thymoma cells. In contrast, Notch1(flox/flox) mice crossed with E8I-cre transgenic mice inoculated with EG7 cells die comparable to control mice, indicating that Notch2 is crucial for exerting antitumor CTL responses. Injection of anti-Notch2 agonistic Ab or delta-like 1-overexpressing dendritic cells augmented the antitumor response in C57BL/6 mice inoculated with EG7 cells. These findings indicate that Notch2 signaling in CD8(+) T cells is required for generating potent antitumor CTLs, thus providing a crucial target for augmenting tumor immune responses.

TACI induces cIAP1-mediated ubiquitination of NIK by TRAF2 and TANK to limit non-canonical NF-kappaB signaling.

B-cell-activating factor of the TNF family (BAFF) is a critical factor for B-cell survival and maturation through non-canonical nuclear factor kappaB (NF-kappaB) pathway, a NF-kappaB inducing kinase (NIK)-dependent pathway for the processing of NF-kappaB2 p100 to generate p52. While BAFF acts primarily through BAFF receptor (BAFF-R), the transmembrane activator and CAML interactor (TACI), the other receptor for BAFF, is thought to serve as a negative regulator for B-cell responses. However, how TACI regulates NF-kappaB2 activity is largely unknown. In this study, we showed that constitutive activation of TACI signaling suppressed BAFF-R-mediated NF-kappaB2 p100 processing with the up-regulation of cellular inhibitors of apoptosis 1 (cIAP1) and TNF receptor associated factor (TRAF)-associated NF-kappaB activator (TANK). The ubiquitination of NIK by cIAP1 was inhibited by the expression of TRAF2 with physical binding to cIAP1. TANK deficiency by small interfering RNA (siRNA) impaired TACI-dependent inhibition of NF-kappaB2 p100 processing. TANK also inhibited TRAF2-mediated cIAP1 inactivation. Moreover, the recruitment of TRAF2 to TACI induced the ubiquitination of NIK. Taken together, the regulation of NIK by TACI through the interaction of TANK/TRAF2/cIAP1 plays a pivotal role in the suppression of non-canonical NF-kappaB signaling.

Differentiation stage-specific requirement in hypoxia-inducible factor-1alpha-regulated glycolytic pathway during murine B cell development in bone marrow.

Hypoxia-inducible factor (HIF)-1alpha plays a central role in oxygen homeostasis and energy supply by glycolysis in many cell types. We previously reported that an HIF-1alpha gene deficiency caused abnormal B cell development and autoimmunity. In this study we show that HIF-1alpha-enabled glycolysis during B cell development is required in a developmental stage-specific manner. Supporting this conclusion are observations that the glycolytic pathway in HIF-1alpha-deficient B220(+) bone marrow cells is much less functionally effective than in wild-type control cells. The expression of genes encoding the glucose transporters and the key glycolytic enzyme, 6-phosphofructo-2-kinase/fructose-2,6-bishosphatase 3, was greatly reduced in HIF-1alpha-deficient cells. The compensatory adaptation to the defect of glycolysis was reflected in higher levels of expression of respiratory chain-related genes and TCA cycle-related genes in HIF-1alpha-deficient cells than in wild-type cells. In agreement with these findings, HIF-1alpha-deficient cells used pyruvate more efficiently than wild-type cells. The key role of HIF-1alpha-enabled glycolysis in bone marrow B cells was also demonstrated by glucose deprivation during in vitro bone marrow cell culture and by using a glycolysis inhibitor in the bone marrow cell culture. Taken together, these findings indicate that glucose dependency differs at different B cell developmental stages and that HIF-1alpha plays an important role in B cell development.

Coculture of osteoclast precursors with rheumatoid synovial fibroblasts induces osteoclastogenesis via transforming growth factor beta-mediated down-regulation of osteoprotegerin.

OBJECTIVE: The mechanisms of osteoclast maturation and the role of rheumatoid arthritis (RA) synovial fibroblasts in the control of osteoclastogenesis remain unclear. The purpose of this study was to determine the humoral factors that influence osteoclast differentiation resulting from mutual interactions between osteoclast progenitor cells and synovial fibroblasts. METHODS: The cloned mouse macrophage cell line RAW 264.7 or isolated human CD14+ monocytes were cocultured with RA or osteoarthritis (OA) synovial fibroblasts in the presence of RANKL. Osteoclasts were visualized by staining for tartrate-resistant acid phosphatase (TRAP), and their functions were evaluated by bone resorption assay. Transforming growth factor beta (TGFbeta) and osteoprotegerin (OPG) levels were measured by enzyme-linked immunosorbent assay. Expression of pSmad2 and Smad7 was analyzed by Western blotting. RESULTS: RANKL-mediated osteoclast formation was observed in cocultures of RAW cells with RA synovial cells, but not with OA synovial cells. This formation was inhibited by TGFbeta receptor kinase inhibitor or neutralizing TGFbeta antibody. Human CD14+ monocytes showed the same results with RAW 264.7, and bone resorption activity was consistent with osteoclast formation. RA synovial fibroblasts produced TGFbeta in response to cell-cell contact with RAW cells in a RANKL-dependent manner. TGFbeta reduced OPG production by RA synovial fibroblasts, but dose-dependently increased OPG secretion in OA synovial fibroblasts. TGFbeta decreased the expression of pSmad2 and increased the expression of Smad7 in RA synovial fibroblasts, but not OA synovial fibroblasts. CONCLUSION: Suppression of OPG production by down-regulation of TGFbeta/Smad2 signaling may contribute to RANKL-mediated osteoclastogenesis from RA synovial fibroblasts.

Notch2 integrates signaling by the transcription factors RBP-J and CREB1 to promote T cell cytotoxicity.

The acquisition of cytotoxic effector function by CD8(+) T cells is crucial for the control of intracellular infection and tumor invasion. However, it remains unclear which signaling pathways are required for the differentiation of CD8(+) cytotoxic T lymphocytes. We show here that Notch2-deficient T cells had impaired differentiation into cytotoxic T lymphocytes. In addition, dendritic cells with lower expression of the Notch ligand Delta-like 1 induced the differentiation of cytotoxic T lymphocytes less efficiently. We found that the intracellular domain of Notch2 interacted with a phosphorylated form of the transcription factor CREB1, and together these proteins bound the transcriptional coactivator p300 to form a complex on the promoter of the gene encoding granzyme B. Our results suggest that the highly regulated, dynamic control of T cell cytotoxicity depends on the integration of Notch2 and CREB1 signals.

TACI attenuates antibody production costimulated by BAFF-R and CD40.

B cell activating factor of the TNF family (BAFF), plays critical roles in B cell survival, activation, differentiation, and antibody (Ab) production. BAFF binds to three receptors: BAFF-R, transmembrane activator and calcium-modulator and cyclophilin ligand interactor (TACI) and B cell maturation antigen. While BAFF-R is the primary receptor for B cell costimulation by BAFF, TACI is reported to serve as a positive or negative regulator for B cell responses depending on conditions. To determine the real role of TACI in B cell responses, we examined the functional relationship between TACI and BAFF-R in Ab production from human peripheral blood B cells using agonistic mAb. BAFF-R and CD40 enhanced IgG secretion and B cell proliferation, which were inhibited by TACI. Although TACI induced mild B cell apoptosis, its extent did not correlate with that of TACI-mediated inhibition of IgG secretion. In addition, TACI inhibited B-lymphocyte-induced maturation protein-1 expression, IgG secretion from previously IgG-negative selected B cells, and activation-induced cytidine deaminase expression enhanced by BAFF-R and CD40. Importantly, BAFF-R and CD40 enhanced B cell responsiveness to TACI-mediated suppression. Thus, BAFF may attenuate T cell-independent and -dependent B cell responses by TACI.

TACI regulates IgA production by APRIL in collaboration with HSPG.

Transmembrane activator and calcium modulator and cyclophilin ligand interactor (TACI) is a member of the tumor necrosis factor (TNF) receptor family that serves as a receptor for B-cell activating factor of the TNF family (BAFF) and as a proliferation-inducing ligand (APRIL). Although TACI is reported to function as a positive or negative regulator for B-cell responses, its roles remain elusive. Experiments using TACI siRNA into B cells indicated that TACI positively regulated APRIL-induced IgA production in collaboration with heparan sulfate proteoglycans (HSPG). Furthermore, TACI negatively regulated BAFF-induced B-cell proliferation and production of IgA and IgG. In addition, B cells treated with heparitinase to denature HSPG showed that HSPG is essential for APRIL-induced B-cell responses such as B-cell proliferation, IgG and IgA production, induction of activation-induced cytidine deaminase (AID), and noncanonical NF-kappaB2. In contrast, phosphorylation of physiological AID kinase, protein kinase A (PKA), was dependent on TACI. Importantly, coligation of TACI and HSPG by specific antibodies, but not by TACI or HSPG ligation itself, could induce the phosphorylation of PKA and IgA production instead of APRIL. Our findings indicate that simultaneous binding of TACI and HSPG on B cells with APRIL is crucial for IgA production.

Cutting edge: hypoxia-inducible factor 1alpha and its activation-inducible short isoform I.1 negatively regulate functions of CD4+ and CD8+ T lymphocytes.

To evaluate the role of hypoxia-inducible factor 1alpha (HIF-1alpha) and its TCR activation-inducible short isoform I.1 in T cell functions, we genetically engineered unique mice with: 1) knockout of I.1 isoform of HIF-1alpha; 2) T cell-targeted HIF-1alpha knockdown; and 3) chimeric mice with HIF-1alpha gene deletion in T and B lymphocytes. In all three types of mice, the HIF-1alpha-deficient T lymphocytes, which were TCR-activated in vitro, produced more proinflammatory cytokines compared with HIF-1alpha-expressing control T cells. Surprisingly, deletion of the I.1 isoform, which represents < 30% of total HIF-1alpha mRNA in activated T cells, was sufficient to markedly enhance TCR-triggered cytokine secretion. These data suggest that HIF-1alpha not only plays a critical role in oxygen homeostasis but also may serve as a negative regulator of T cells.

Identification of a glioma antigen, GARC-1, using cytotoxic T lymphocytes induced by HSV cancer vaccine.

Despite several ongoing clinical trials of immunotherapies against glioma, few glioma-specific antigens recognized by cytotoxic T lymphocytes (CTLs) have been identified. We recently demonstrated that intratumoral inoculation with herpes simplex virus (HSV) as a cancer vaccine activates tumor-specific CTLs. To identify glioma antigens recognized by CTLs, we used the HSV cancer vaccine to vaccinate mice harboring a syngeneic mouse glioma cell line, GL261. From the splenocytes of the immunized mice, we generated an H-2Db-restricted CTL line, GCL-1, that was specific for GL261. Then, a cDNA expression library generated from GL261 was screened with GCL-1, and a new gene encoding glioma antigen, GARC-1, was isolated. Sequence analysis revealed that the GARC-1 gene isolated from GL261 had a point mutation causing an amino acid change (Asp to Asn at position 81). T-cell epitope analysis revealed that the mutated peptide GARC-1(77-85) (AALLNKLYA) but not the wild-type peptide (AALLDKLYA), was recognized by GCL-1. These results suggest that HSV cancer vaccination may be a useful method for inducing tumor-specific CTLs and identifying tumor antigens. Furthermore, this GL261/GARC-1 murine glioma model may be useful for the development of immunotherapy for brain tumors.

CD4+CD25+ regulatory T cells attenuate the phosphatidylinositol 3-kinase/Akt pathway in antigen-primed immature CD8+ CTLs during functional maturation.

This study was designed to determine the role of CD25(+)CD4(+) regulatory T (Tr) cells in CTL maturation and effector functions using a murine CTL line and in vitro MLC. Tr cells inhibited CTL functional maturation, but had no effect on CTL effector functions. In CD4(+) responder T cell-depleted MLC supplemented with IL-2, Tr cells suppressed mature CTL generation only when added within the first 2 days of culture. Tr cells down-regulated levels of active Akt, but not STAT5 or ZAP70 in Ag-primed immature CTLs. Down-regulation of active Akt was accompanied by a reduction in CTL cell size and IL-2Ralpha expression. In Tr cell-depleted MLC, CTLs were generated that exhibited high levels of nonspecific cytotoxicity. Our in vitro findings suggest that Tr cells regulate functional CTL maturation to generate optimal Ag-specific immune responses through the control of the PI3K/Akt pathway.