CISH is induced during DC development and regulates DC-mediated CTL activation

Authors


Abstract

The cytokine inducible SH2-domain protein (CISH) is a well-known STAT5 target gene, but its role in the immune system remains uncertain. In this study, we found that CISH is predominantly induced during dendritic cell (DC) development from mouse bone marrow (BM) cells and plays a crucial role in type 1 DC development and DC-mediated CTL activation. CISH knockdown reduced the expression of MHC class I, co-stimulatory molecules and pro-inflammatory cytokines in BMDCs. Meanwhile, the DC yield was markedly enhanced by CISH knockdown via cell-cycle activation and reduction of cell apoptosis. Down-regulation of cell proliferation at the later stage of DC development was found to be associated with CISH-mediated negative feedback regulation of STAT5 activation. In T-cell immunity, OT-1 T-cell proliferation was significantly reduced by CISH knockdown in DCs, whereas OT-2 T-cell proliferation was not affected by CISH knockdown. CTLs generated by DC vaccination were also markedly reduced by CISH knockdown, followed by significant impairment of DC-based tumor immunotherapy. Taken together, our data suggest that CISH expression at the later stage of DC development triggers the shutdown of DC progenitor cell proliferation and facilitates DC differentiation into a potent stimulator of CTLs.

Introduction

DCs are the most well-known professional antigen-presenting cells, which play a key role in priming antigen-specific adaptive immunity and controlling immune homeostasis 1, 2. DCs have been studied widely because of their Th1-polarizing immunogenicity when inoculated in vivo, demonstrating their potential for use in the development of therapeutic cancer vaccines 3–5. Granulocyte-macrophage colony stimulating factor (GM-CSF) is a cytokine essential for DC development from hematopoietic progenitor cells 6. However, the molecular mechanisms underlying GM-CSF-mediated DC development in association with Th1-polarizing immunogenicity are poorly understood.

GM-CSF exerts its biological functions by phosphorylating at least two distinct domains in the β-chain of its receptor. One of these domains induces the activation of mitogen-activated protein kinases (MAPKs), which is followed by the activation of the PI3K/Akt/p21waf-1 pathway, and the other domain mediates activation of Janus kinase 2 (JAK2)/signal transducer and activator of transcription 5 (STAT5) signaling pathway 7. When activated, STAT5 migrates into the nucleus after mutual phosphorylation and dimerization, and then binds to specific DNA-binding sites, followed by the expression of STAT5 target genes 8. STAT5 is well established for its diverse activities in cell proliferation, renewal of hematopoietic stem cells (HSCs) and hematopoietic progenitor cells, and moderating cell differentiations 9.

Cytokine inducible SH2-domain protein (CISH) was first reported as an immediate early gene induced by GM-CSF, interleukin (IL)-2, IL-3 and erythropoietin (EPO) in hematopoietic cells 10. Later, CISH was identified as a STAT5 target gene 11 and a member of the suppressor of cytokine signaling (SOCS) family 12. The expression of CISH was found to induce feedback inhibition of STAT5 activation 11. CISH-transgenic mice displayed impaired responses to IL-2 and resemble STAT5-deficient mice 13, suggesting that a specific function of CISH is associated with the negative regulation of STAT5-mediated cytokine responses. However, a separate study reported that CISH enhances T-cell populations by activating TCR signaling through interactions with protein kinase C-θ in transgenic mice with CISH-expressing CD4+ T cells 14. These data suggest that the biological function of CISH likely varies depending on cell type and/or physiological conditions.

In a recent report, CISH gene mutations in humans were found to be strongly associated with susceptibility to diverse infectious diseases caused by pathogens 15. However, the detailed mechanisms underlying this susceptibility are still unknown. We found that CISH is significantly induced during DC development and plays an important role in DC-mediated CTL activation. A gradual increase in CISH expression during GM-CSF-mediated DC development was observed. Gene silencing experiments revealed that CISH inhibits the proliferation of DC precursor cells and triggers their differentiation at the later stage of DC development to become potent type 1 polarized DCs, which is essential for preventative or therapeutic immunity against exogenous pathogens or intrinsic tumors. Our findings provide a better understanding of the immunological mechanisms underlying the susceptibility to infectious diseases in patients with CISH mutations.

Results

CISH expression is induced during DC development

During DC development under GM-CSF conditions, BM progenitors underwent proliferation at the early stage up to 4 days and then mostly differentiated into DCs without further replication at the later stage (4–6 days; Fig. 1A). In the microarray analysis, the CISH mRNA level was significantly induced during DC development from mouse BM cells (Fig. 1B). In the quantitative real-time RT-PCR and kinetic western blot study, CISH expression, undetectable in BM cells, was strongly induced in the later stages of DC development from BM cells (Fig. 1C). Consistently, CISH expression increased during DC development in parallel with CD11c (Fig. 1D). CISH expression was also clearly detected in mouse splenic DCs (Supporting Information Fig. 1). These results suggest that CISH expression may be involved in GM-CSF-mediated DC development from BM cells.

Figure 1.

CISH expression during BM-derived DC (BMDC) development and effects of CISH knockdown on DC phenotype. Mouse BM cells (BALB/c) were cultured for 6 days in the presence of 10 ng/mL mGM-CSF. Cells were harvested at the indicated time points and analyzed. (A) To follow DC development from BM cells, mouse BM cells were labeled with CFSE (1 μM), cultured in the presence of mGM-CSF, harvested on days 0, 2, 4 and 6, and assessed by flow cytometry after staining with CD11c antibody. (B) Microarray data of the relative amounts of CISH mRNA during DC development from BM cells are represented by fold increases. (C) CISH mRNA and protein levels were assessed by real-time RT-PCR and Western blot analysis, respectively, during DC development. (D) Cells harvested on days 0, 2, 4 and 6 during DC development were assessed by flow cytometry after staining with PE-labeled CD11c mAb, CISH polyclonal antibody and FITC-labeled goat anti-rabbit IgG. Rabbit polyclonal IgG was used as an isotype control. (E) DC precursor cells were transfected with CISH siRNA (si-CISH) and control siRNA (si-con) on day 4 during DC development. Two days later, CISH knockdown was evaluated by Western blotting and flow cytometry analysis. (F) Mean fluorescence intensities (MFI) of selected surface phenotypic markers of CISH normal (si-con) and CISHKD (si-CISH) DCs were assessed by flow cytometry analysis. The data shown are from three independent experiments and are represented as mean±SD. **p<0.05 by unpaired Student's t-test. (A–E) Data shown are representative of more than three independent experiments, except the microarray data (B).

CISH knockdown affects the surface phenotype of DCs

When DC precursor cells were transfected with synthetic siRNA duplexes, more than 80% of CISH expression in DCs was efficiently silenced (Fig. 1E). CISH knockdown did not induce any changes in CD11c and MHC class II expression on the surface of DCs but caused tangible reduction (around 10%) in the surface expression of MHC class I and co-stimulatory molecules compared with control siRNA-transfected cells (Fig. 1F and Supporting Information Fig. 2). These data suggest that CISH expressed during DC development is likely to be involved in the type 1 polarization of DCs.

CISH expression down-regulates proliferation of DC precursor cells

CISH has been reported to have a negative effect on the proliferation of erythroid progenitor cells 16, and CISH knockout mice showed an increase in BM progenitor cells 17. Based on these previous findings, we investigated the effects of CISH expression on DC yield during DC development. CISH expression was highest on day 6 when DCs were fully differentiated (Fig. 1C and D). In the BrdU incorporation assay, DC precursor cells were highly proliferative during DC development for the first 5 days and proliferation markedly decreased on day 6 (Fig. 2A). CISH knockdown enhanced the proliferation of day 6 DC precursor cells by more than twice as compared with CISH normal cells (Fig. 2A) but did not much affect the development of neutropils and macrophages under the same culture conditions (Supporting Information Fig. 3). The populations of BrdU-positive proliferating cells were maintained at over 50% of the total cell population for the first 5 days and significantly decreased to <20% on day 6 (Fig. 2B). However, following CISH knockdown, the BrdU-positive cell populations on day 6 were maintained at about 47% (Fig. 2B), suggesting that CISH is likely involved in the inhibition of cell proliferation and/or the activation of cell apoptosis. In the MTT cell proliferation assay, the DC precursor population dramatically increased up to day 5 and slightly decreased on day 6 in control cells, while the CISHKD cell population continued increasing up to day 6 during DC development (Fig. 2C), suggesting that CISH plays a negative role at least in part in proliferation and/or cell viability during DC development.

Figure 2.

Effects of CISH knockdown on DC precursor cell proliferation during DC development. (A) Representative flow cytometry data showing BrdU-positive proliferating cells during BMDC development. Mouse BM cells (BALB/c) cultured in the presence of mGM-CSF were allowed to incorporate BrdU on days 2, 3, 4, 5 and 6 for 12 h (upper six panels). Cells on day 4 were transfected with si-CISH or si-con for 48 h, followed by BrdU incorporation for 12 h (bottom two panels). Cells were assessed by flow cytometry after staining with BrdU antibody. (B) BrdU-positive cells in (A) were assessed in triplicate and represented as mean±SD. (C) DC precursor cells were transfected with si-con or si-CISH on days 2, 3 and 4. Forty-eight hours after transfection, metabolically active (proliferating) cells were quantified using the MTT assay. Data from four independent experiments are represented as mean±SD. (D) Cell cycles of CISH normal (si-con) and CISHKD (si-CISH) DCs were assessed by flow cytometry after PI staining (upper panels), and the percentages of cells in each cell cycle are shown (lower panel). (E) Representative flow cytometry data showing cell apoptosis assessed by annexin V and PI staining of CISH normal (si-con) and CISHKD (si-CISH) DCs, respectively. Quantitative data of early apoptosis (annexin V+) and late apoptosis (annexinV+PI+) assessed from three independent experiments are represented as mean±SD. *p<0.01, **p<0.05 compared with control DCs by unpaired Student's t-test.

CISH is involved in cell-cycle arrest and cell apoptosis

To elucidate the mechanisms underlying CISH-mediated reduction of total cell population on day 6 of DC development, we examined the cell-cycle and cell apoptosis rate of CISH normal and CISHKD cells. In the cell-cycle analysis of CISH-normal and CISHKD cells on day 6, the number of cells in the S and G2/M phases was shown to be significantly higher while the number of cells in the G1 phase were decreased by CISH knockdown (Fig. 2D), suggesting that CISH is mainly involved in the G1 arrest of DC precursor cells during DC development. In addition, the reduction of the sub-G1 population (8.9–4.6%) by CISH knockdown (Fig. 2D) implies that CISH is also involved in cell apoptosis.

In the flow cytometric analysis of cell apoptosis, both annexin V+ and annexinV+PI+ cells were significantly reduced in CISHKD cells compared with CISH normal cells on day 6 of DC development (Fig. 2E), suggesting that CISH is also involved in early and late apoptosis, resulting in the reduction of DC yield.

Down-regulation of DC precursor cell proliferation by CISH-mediated feedback inhibition of STAT5

GM-CSF activates STAT5 via tyrosine phosphorylation, and the activated dimeric form of STAT5 translocates into the nucleus to induce the expression of STAT5 target genes, leading to induction of cell proliferation, differentiation and survival 18, 19. We examined the expression and activation patterns of STAT5 during DC development from BM cells. STAT5 mRNA levels were constant during DC development (Fig. 3A). In contrast, the STAT5 protein, barely detected in BM cells, was induced on day 2, reached a plateau on day 4 and gradually decreased on day 5 and day 6, which is consistent with our previous report 20 (Fig. 3B). In parallel with STAT5 activation, the well-known STAT5 target gene Bcl2 was similarly induced during DC development (Fig. 3B). DC yield was markedly reduced by treatment with STAT5 inhibitor during DC development along with a significant inhibition of Bcl2 and CISH expression (Fig. 3C). On the other hand, CISH knockdown enhanced STAT5 activation and accompanied by Bcl2 increase without affecting the levels of STAT5, leading to an increase in DC yield (Fig. 3D). Our data suggest that during GM-CSF-mediated DC development, GM-CSF signaling induces the expression and activation of STAT5, resulting in the induction of CISH expression at the later stage, which down-regulates STAT5 activation by feedback inhibition, leading to the inhibition of cell proliferation on day 6 and reduction in DC yield (Fig. 3E).

Figure 3.

CISH-mediated feedback inhibition of STAT5 activation during DC development. (A) Levels of STAT5 mRNA were assessed by RT-PCR, (B) STAT5 expression and its activation together with Bcl2 expression were assessed by western blot analysis during DC development. The data shown are representative of three independent experiments. (C and D) BMDCs on day 4 of development were transfected with si-con or si-CISH, and harvested 48 h after transfection. The effects of STAT5 inhibitor (100 μM) (C) or CISH-knockdown (D) on STAT5 activation and cell proliferation were assessed by western blotting and direct cell counting, respectively. *p<0.01 compared with control DCs by unpaired Student's t-test. The blots shown are representative of five independent experiments. Graphs are mean+SD of five independent experiments. (E) Schematic of CISH and STAT5 positioned downstream of the GM-CSF signaling pathway.

CISH expression in DCs plays an essential role in DC-mediated CD8+ T-cell immunity

We next examined the biological meaning of CISH expression during DC development apart from the aspect of cell proliferation. Down-regulation of MHC class I and other co-stimulatory molecules in CISHKD DCs (Fig. 1F) implies that CISH might be involved in CD8+ T-cell immunity. CISHKD DCs produced significantly lower amounts of the Th1-directing cytokines IL-12, IL-6 and TNF-α compared with control DCs (Fig. 4A). In a T-cell proliferation assay with OT-1 and OT-2 T cells, the ability of the CISHKD OVA-pulsed mature DCs to stimulate OT-1 T cells was significantly impaired when compared with control DCs while CISH knockdown did not affect the ability of DCs to stimulate OT-2 T cells (Fig. 4B and Supporting Information Fig. 4). These results were reproducible in three independent experiments (Fig. 4C), and further verified by repeated dose kinetic studies (Fig. 4D). CISH knockdown caused minor alterations in the levels of Th1/Th2/Th17 cytokines (Supporting Information Fig. 5) and in the population of Treg cells (Supporting Information Fig. 6) in co-culture experiments with CISHKD DCs, but CISH-knockdown did not affect DC-mediated entire proliferation of OT-2 T cells. Taken together, our results suggest that CISH expression during DC development is essential for type 1 polarization of DCs to activate CD8+ T cells.

Figure 4.

Effects of CISH knockdown on DC-mediated T-cell proliferation. (A) Th1 cytokines were assessed by ELISA from the cultures of CISHKD (si-CISH) and CISH normal (si-con) BALB/C-originated BMDCs matured by LPS (200ng/mL) stimulation. (B) Respective OVA peptide (10 μg/mL)-pulsed CISH-normal (si-con) and CISHKD (si-CISH) BMDCs of OT-1 or OT-2 mice were co-cultured with CFSE-labeled syngeneic T cells for 96 h. CFSE-positive proliferating T cells were gated and calculated (Supporting Information Fig. 4). (C) CFSE-labeled OT-1 and OT-2 T-cell proliferation at a DC:T-cell ratio of 1:1. (D) T-cell proliferation was assessed by the FACS beads count method from co-cultures of T/DC at the indicated ratios. *p<0.01, compared with control DCs by unpaired Student's t-test. Data shown are mean±SD and are representative of three independent experiments.

CISH knockdown impairs the CTL inducing capacity of DCs in vivo

To determine the effects of CISH on the effector functions of DCs in vivo, we analyzed CTLs induced by vaccination with CISHKD DC. The levels of INF-γ in the spleen and the draining lymph nodes (DLNs) of OT-1 mice vaccinated with OVA-pulsed CISHKD DCs were roughly 40–50% lower than that of OT-1 mice vaccinated with OVA-pulsed CISH normal DCs (Fig. 5A). OVA-specific CTL activity assessed with EG7 cells was clearly observed in a dose-dependent manner in the spleen and the DLNs of OT-1 mice vaccinated with OVA-pulsed CISH normal DCs, while CTL activity was not properly induced in mice vaccinated with CISHKD DCs (Supporting Information Fig. 7 and 8). Considering the CTL activity assessed with EL4 cells as control target cells, statistically significant OVA-specific CTL activities were only observed in mice vaccinated with OVA-pulsed CISH normal DCs but not with CISHKD DCs. These results were further verified by quantitative statistical analysis from three independent experiments with splenocytes (Fig. 5B) and DLNs (Fig. 5C) obtained from vaccinated mice. In addition, the population of perforin-secreting cells and the levels of Granzyme B were significantly reduced in OT-1 mice that were vaccinated with CISHKD DCs instead of CISH normal DCs (Supporting Information Fig. 9). However, the reduction of perforin and granzyme B expression in OT-1 mice vaccinated with CISHKD DCs was similar to the reduction of CTLs in these mice, suggesting that the reduction of these CTL-associated proteins in CISHKD DC-vaccinated OT-1 mice was not due to impaired expression of these genes, but rather due to a reduced number of CTLs in the vaccinated mice.

Figure 5.

CTL induction and tumor immunotherapy by CISH knockdown DCs. OT-1 mice were immunized twice at a 1 week interval with 1×106 OVA peptide-pulsed CISHKD (si-CISH) or CISH-normal (si-con) BMDCs from C57BL/6 mice. Cells in the spleen and draining lymph nodes (DLNs) of vaccinated mice were obtained and cultured in vitro in the presence of the OVA peptide for 7 days. (A) IFN-γ levels in culture supernatants were assessed in triplicates on day 2 of culture by ELISA and are represented as means±SD. (B and C) CTLs in spleens and DLNs of vaccinated OT-1 mice were assessed using CSFE-labeled EG7 and EL4 as target cells by flow cytometry (Supporting Information Fig. 7 and 8). Quantitative cytotoxic effects of CTLs are represented as mean±SD (n=3). (D and E) C57BL/6 mice were inoculated s.c. in the right flank area with EG7 or EL4 tumor cells (5×105), and immunized twice on days 3 and 10 with 1×106 cells of OVA peptide-pulsed CISHKD (si-CISH) or CISH-normal (si-con) BMDCs (derived from C57BL/6 BM cells). Upper panels show representative images of EG7 and EL4 tumors on day 20 in tumor-bearing mice. Tumor growth was monitored every 2 days and then represented as mean±SD (n=5) (bottom panels). *p<0.01, compared with the group of mice vaccinated with OVA-pulsed control (si-con) DCs by unpaired Student's t-test.

CISH is essential for DC-based tumor immunotherapy

To determine the effect of CISH on DC-based tumor immunotherapy, 3 day model tumor-bearing C57BL/6 mice were treated twice with CISH-normal and CISHKD OVA-pulsed DCs, respectively. EG7 cell-derived tumor-bearing mice showed partial or complete regression by vaccinations with OVA-pulsed CISH normal DCs but not by inoculation with CISHKD DCs (Fig. 5D). However, none of the EL4-derived tumor-bearing mice showed tumor regression by DC vaccination, even with OVA-pulsed CISH normal DCs (Fig. 5E). These data suggest that CISH is essential for DCs to induce anti-tumor immunity in DC-based tumor immunotherapy. In addition, anti-tumor immunity induced by DC vaccination was not enhanced by ectopic over-expression of CISH in DCs (Supporting Information Fig. 10A), implying that the level of endogeneous CISH expressed during DC development would be enough to potentiate the antitumor immunogenicity of the DC vaccine. We also examined the effect of STAT5 inhibitor on the antitumor immunity of the DC vaccine. STAT5 inhibitor treated at a later stage (D-4) of DC development did not much affect to the CISH expression and the antitumor immunogenicity of DC vaccine (Supporting Information Fig. 10B). However, we are unable to perform this experiment with cells treated with STAT5 inhibitor during the early stage of DC development because DC development was significantly inhibited when treated with STAT5 inhibitor during the early stage (D-2).

Discussion

SOCS family proteins regulate the responses of immune cells to cytokines 21, 22. Most SOCS family proteins are induced by cytokines and act in a classical negative-feedback loop to inhibit cytokine signal transduction 23. CISH was the first member of the SOCS family to be described 10. CISH is well known to be induced by cytokines in several hematopoietic cells 10, 18, 24, but CISH expression and its biological function have not been evaluated in association with DC development.

In the present study, we examined the expression and functional properties of CISH during DC development. We found that CISH, while not expressed in BM cells, is induced and significantly up-regulated during the later stages of DC development (Fig. 1). The level of MHC class I and co-stimulatory molecules was down-regulated on the surface of DCs by CISH knockdown during DC development (Fig. 1F). Induced CISH was shown to be involved in DC development by inhibiting the proliferation of precursor cells and leading to their differentiation into DCs (Figs 2 and 3). We next examined the role of CISH in DC-mediated immunogenicity in vitro and in vivo. We found that CISH induced and expressed during DC development plays a crucial role in the development of type 1 DCs, which is essential for CD8+ T-cell activation (Fig. 4). CISHKD DCs were not as effective as CISH normal DCs in the induction of antigen-specific CTLs in the immunized mice and DC-based immunotherapy in tumor-bearing mice (Fig. 5). These data indicate that the CISH abundantly expressed during DC development is essential for the differentiation of type 1 DC from mouse BM cells, leading to the activation of antigen-specific CTLs in DC-mediated immunotherapy.

CISH has been reported to be consistently up-regulated in T cells when activated by IL-2 in humans 25, and to be critical for T-cell proliferation and survival via activation of the T-cell receptor-mediated signaling pathway 14, 26. In this study, however, CISH was found to be a negative regulator of cell proliferation during DC development that acted by blocking STAT5 activation (Fig. 3). STAT5 and some of its target genes play a major role in cell proliferation and differentiation 9, 18, 19, 27, and aberrant STAT5 activation was reported to be associated with myeloid malignancies 28. Since CISH, a STAT5 target gene, is a negative feedback regulator of STAT5 activation 11, 13, we examined the expression and activation of STAT5 and CISH together with cell proliferation during DC development. We found that STAT5 expression and activation are gradually enhanced during the early stage of DC development but suppressed by CISH expression at the later stage of DC development, which results in decreased cell proliferation (Fig. 2). CISH knockdown was accompanied by STAT5 activation, resulting in increased cell proliferation (Fig. 3D). These results suggest that CISH may have at least two different functions – activation or cell proliferation – depending on the cell type or developmental stage.

CISH is known as a negative regulator of cytokine signaling 11–13. Recently, however, a positive role for CISH in immunity against infectious pathogens was reported, since the risk of infectious diseases was found to be higher in people carrying variant CISH alleles 15. This report implies that CISH plays a pivotal role in the immune response against foreign invaders, but the underlying immunological mechanisms have not yet been explored. It has been proposed that CISH mutations may be accompanied by enhanced activation of STAT5, which is a transcription factor of Treg cells, leading to an increase in the Treg-cell population, and may thus suppress effecter T cells, resulting in higher susceptibility to infectious agents 29. Our present findings suggest that the higher susceptibility to infectious diseases in patients with CISH mutations is at least in part due to impaired DC-mediated CTL induction, since CISH is essential for the differentiation of type 1 DCs and DC-mediated CTL activation. CISHKO mice will provide a better understanding of the role of CISH in immune responses in vivo. Cellular and molecular mechanisms underlying the prolonged CISH expression and CISH-mediated enhancement of MHC class I, co-stimulatory molecules and Th1 cytokines during DC development are under investigation.

Materials and methods

Mice, cell lines and reagents

Six- to eight weeks old female BALB/C (H-2d) or C57BL/6(H-2b) mice and C57BL/6-background (H-2b) OT-1 (OVA-specific CD8+ T-cell receptor transgenic) and OT-2 (OVA-specific CD4+ T-cell receptor transgenic) mice were used. All mice were maintained in the animal care facility of Sungkyunkwan University according to the University Animal Care and Use guidelines. EL4 (C57BL/6 mouse-derived thymoma cells) and EG7 (OVA-expressing EL4 cells) cells were obtained from American Type Culture Collection (ATCC). 5,6-Carboxyfluorescein succinimidyl ester (CFSE; Molecular Probes), bromodeoxyuridine (BrdU) cell proliferation kit (BD Pharmingen) and STAT5 inhibitor (Calbiochem) were used. CISH polyclonal antibody was purchased from Abcam. Anti-phospho-tyrosine-STAT5 and anti-pan-STAT5 antibodies were obtained from Cell Signaling, and anti-β-actin, HRP-conjugated anti-rabbit and anti-mouse IgGs were purchased from Sigma. Murine granulocyte-macrophage colony-stimulating factor (mGM-CSF) was obtained from Creagene. Lipopolysaccharide (O111:B4) was purchased from Sigma-Aldrich. Ovalbumin agonistic peptides (OVA257–264; SIINFEKL and OVA323–339; ISQAVHAAHAEINEAGR) synthesized based on sequence information 30 were provided by Peptron (Dajeon, Korea). The FITC-annexin V/propidium iodide (PI) kit and FITC- or PE-conjugated monoclonal antibodies for DC-specific surface markers were purchased from BD Biosciences Pharmingen (San Jose, CA, USA). FITC- or PE-conjugated CD4, CD8, Ly6G, F4/80, Foxp3 antibodies and cytokine ELISA kits for murine IL-6, IL-10, IL-12p70, IL-17A, TNF-α and IFN-γ were purchased from BioLegend. Mouse granzyme B ELISA kit and FITC-conjugated perforin antibody were obtained from eBioscience and mouse FasL ELISA kit from R&D System.

BMDCs and splenic DCs

BMDCs were generated from BM progenitor cells as described previously 20, 31. Briefly, BM cells were collected from the femurs and tibiae of 6 weeks old female BALB/c or C57BL/6 mice, and treated with ACK lysing buffer (Lonza) to remove erythrocytes. Cells were washed and cultured in complete RPMI 1640 media (RPMI 1640 supplemented with 10% FBS, penicillin/streptomycin) containing 10 ng/mL mGM-CSF (Creagene). After 2 days, the cultures were refreshed with 2 mL of fresh complete medium containing mGM-CSF. On day 4, 1 mL of fresh complete medium containing mGM-CSF was added to the culture. On day 6, non-adherent cells were collected as a source of immature DCs (imDCs). Splenic DCs were isolated from the spleens of 6- to 8 weeks old BALB/c mice using CD11c microbeads according to vender's instruction (Miltenyi Biotech).

Microarray analysis

cDNA microarray analysis was performed by Macrogene (www.macrogene.com). Briefly, BM cells obtained from BALB/c mice were cultured in the presence of mGM-CSF for 2, 4 or 6 days. BM cells were used as a 0 day sample; adherent cells harvested on day 2 by scraping cells using rubber policeman (SPL Life Science) were used as a 2-day sample, and the suspended cells harvested on days 4 and 6 were used as 4 and 6 day samples. Total RNAs from each culture were used to synthesize biotinylated cRNA using the Illumina TotalPrep RNA Amplification Kit (Ambion, Austin, TX, USA). Biotinylated cRNAs were hybridized to the Illumina MouseRef-8 Expression Beadchip (Illumina, San Diego, CA, USA).

CISH knockdown or CISH-overexpression in BMDCs

Two CISH-specific siRNAs were used to knockdown CISH expression. One was designed and synthesized by BLOCK-IT RNAi Designer (Invitrogen) (5′-cctttgcgtacagggatcttgtcct-3′) together with control siRNA (5′-ccttgcggacatagggttctttcct-3′), and the other was purchased from Santa Cruz Biotech (Cat. No. sc-61854). DC precursor cells were transfected with CISH or control siRNAs using the Gene Porter transfection kit as described 32, 33. For CISH overexpression in BMDC, full-length CISH cDNA, prepared by RT-PCR with primers (5′-gaattcatggtcctttgcgtacagg-3′, 5′-ctcgagtcagagttggaaggggtactgt-3′), was cloned into the pCDNA3-Flag plasmid. Forty-eight hours after transfection with pCDNA-Flag-CISH, exogenous CISH expression was examined by western blot with a Flag antibody.

Flow cytometry analysis

For phenotypic analysis, immunofluorescence staining was performed as described previously 31. Cells were stained in FACS buffer at 4°C for 20 min with the appropriate antibodies: FITC-labeled rat anti-mouse CD14 (rmC5-3), anti-mouse CD86 (GL1), anti-mouse I-A/I-E (2G9), anti-mouse MHCI (H-2Kd) and PE-labeled hamster anti-mouse CD11c (HL3), anti-mouse CD80 (16-10A1), rat anti-mouse-CD40 (3/23) (BD Pharmingen) together with PE- or FITC-labeled isotype control antibodies. After washing, cells were analyzed by FACS Calibur (BD) and CellQuest software. For intracellular CISH staining, cells prestained with PE-labeled hamster anti-mouse CD11c antibody were fixed and permeabilized by using the BD Cytofix /Cytoperm™ kit (BD Bioscience Pharmingens). Cells were then stained with rabbit anti-mouse CISH antibody (Abcam) or rabbit isotype control antibody in BD Perm/wash buffer for 1 h, followed by staining with second FITC-labeled goat anti-rabbit IgG antibody. After washing with BD Perm/wash buffer, cells were analyzed by flow cytometry. Intracellular staining for FoxP3 and perforin were conducted according to the manufacturer's protocol (BioLegend).

CFSE labeling

BM cells, DC precursor cells or T cells were labeled with CFSE according to the procedures described elsewhere 34, 35. Briefly, cells were incubated in PBS containing 1 μM CFSE and 5% FBS at 37°C for 10 min. Cells were then washed three times with ice-cold PBS containing 5% FBS. CFSE-labeled BM cells were cultured for 6 days as described in the DC generation sections. During the culture of CFSE-labeled BM cells were harvested on days 2, 4 and 6 for further analysis.

T-cell proliferation assay

CISH-normal and CISH-knockdown DCs were first prepared by transfecting DC precursor cells on day 4 with control (si-con) or CISH-specific siRNAs (si-CISH) for 2 days, matured by culturing with LPS (200 ng/mL) for 24 h, and then pulsed with OVA peptide (10 μg/mL). T cells were isolated from the spleen or DLNs of OT-1 and OT-2 mice as described 36. Briefly, mouse spleens or DLNs were homogenized in RPMI medium, passed through a 70 μM nylon cell strainer (BD Falcon) and treated with ACK lysing buffer (Lonza). T cells were purified by nylon wool columns (Poly Sciences, Warrington, PA, USA) or negatively selected using a mouse T-cell Isolation Kit (Miltenyi Biotech), and then CFSE labeled T cells were co-cultured with DCs at different ratios for 3–4 days. Before harvesting, a known number of calibrated beads (BD) and 10 μL of 1 μg/mL PI was added to each well according to a standard protocol 34, 37. Cells were harvested and analyzed by flow cytometry. Beads were gated as R1 and live cells gated as R2. Cell populations were assessed using the following formula: no. of cells=no. of live cells gated/no. of beads from ungated data×no. of beads added.

BrdU cell proliferation assay

BrdU incorporation was performed using the BrdU Flow Kit (BD Pharmingen). Briefly, 10 μL of BrdU solution (1 mM BrdU/mL stock) was directly added to each culture 12 h prior to cell harvest. After washing, cells were fixed with BD Cytofix/Cytoperm buffer according to vender's manual. Then, cells were treated with DNase I (30 μg/tube) for 1 h at 37°C, incubated with anti-BrdU antibody for 20 min at room temperature and were then analyzed by flow cytometry.

MTT assay

To examine the total live cell population during DC development, the MTT assay was performed as described 18. Briefly, DC precursor cells cultured in the presence of mGM-CSF were transfected with con-siRNA or CISH-siRNA on days 2, 3 and 4. Forty-eight hours after transfection, the MTT solution (20 μL/well) was added to each culture for 4 h followed by the addition of solubilizing buffer and incubation at 37°C. Absorbance was measured at 570 nm.

Cell-cycle analysis and apoptosis assay

Cell-cycle analysis was done according to a standard protocol as described previously 38, 39. CISH-normal and CISHKD BMDCs harvested on day 6 were washed and fixed with cold 70% ethanol for 1 h. The fixed cells were incubated in PBS containing 50 μg/mL RNase A, 0.25% Triton X-100 and 0.1 mM EDTA for 30 min at 37°C. PI was added to the cell suspension at a final concentration of 100 μg/mL and incubated for 15 min at room temperature in the dark. Cell cycle was analyzed by flow cytometry using the Cell Quest software (Becton Dickinson).

Cell apoptosis was assessed by flow cytometry after staining with 5 μL of annexin V-FITC and 5 μL of PI for 15 min using an annexin V-FITC Apoptosis Detection Kit (BD Biosciences Pharmingen).

Quantitative real-time RT-PCR

Total RNAs were purified from cells using TRIzol (Invitrogen). cDNA was synthesized, using RevertAid™ H Minus Reverse transcriptase and Oligo dT primers (Fermentas). Quantitative PCR was performed using the SYBR green PCR Mastermix (Qiagen), and RT-PCR was performed using the Maxime RT-PCR kit (iNtRON Biotechnology). The following primers were used: CISH sense (S) 5′-tgtgcatagccaagacgttc-3′, anti-sense (AS) 5′-tagtgctgcacaaggctgac-3′; STAT5A (S) 5′-gtgaagcgctcaacatgaaa-3′, (AS) 5′-cggtctgggaacacgtagat-3′; GAPDH (S) 5′-aatgtgtccgtcgtggatct-3′, (AS) 5′-tccaccaccctgttgctgta-3′; β-actin (S) 5′-gtatgcctcggtcgtacca-3′, (AS) 5′-cttctgcatcctgtcagcaa-3′.

Western blot analysis

Western blot analysis was performed as described 38, 39. Briefly, whole cell lysates were normalized, subjected to 12 % SDS-PAGE and then transferred to a PVDF membrane (Millipore, USA). Membranes were incubated with primary Abs at 4°C overnight, followed by staining with HRP-conjugated secondary Abs. Bound Abs were detected using a Chemiluminescent HRP substrate (Millipore) and autoradiographic films (Agfa Healthcare NV, Belgium).

CTL assay and ELISA for IFN-γ

OT-1 mice were injected twice at a 1 week interval with OVA-pulsed control (si-con) or CISHKD (si-CISH) DCs. Splenocytes were harvested from injected mice on day 14 and cultured for 7 days in the presence of 10 μg/mL OVA in six-well plates (2×106 per well), and then co-cultured with target cells (EL4 and EG7) labeled with 1 μM CFSE at different ratios for 4 h. After PI staining, CTL were assessed by flow cytometry as described 40, 41 with minor modifications, and the level of IFN-γ in the 2 day culture supernatant was assessed by ELISA.

DC-based tumor immunotherapy

C57BL/6 mice were injected with EL4 or EG7 cells (5×105 cells per mouse) subcutaneously (s.c.) into the right flank region 33. Tumor-bearing mice were vaccinated twice on days 3 and 10 after tumor inoculation s.c. with CISH normal or CISHKD DCs (1×106/mouse) pulsed with OVA (10 μg/mL). Tumor growth was monitored every 2 days using a caliper. Tumor mass was calculated as: V=(2A×B)/2, where A is the length of the short axis and B is that of the long axis.

Statistical analysis

All experiments were repeated at least three times with consistent results. Statistical data are presented as mean±SD. Comparisons of group means were assessed with Student's t-test. A p-value of <0.05 was considered statistically significant.

Acknowledgements

We thank Jun Chang (Ehwa Women's University, Korea) and Mi-Na Kwon (IVI) for providing OT-1 and OT-2 mice. We are grateful to Yoon Lee at Creagene for technical help, and Selim Ahmed and Su-Jin Park for assistance. This work was supported by the Specific Basement Grant from the Korea National Research Foundation (2009-0084683) and the Bio New Drug Grants (A102130 and A110054) from the Korean Ministry of Health and Welfare.

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

Ancillary