Differentiation of Inflammatory Dendritic Cells Is Mediated by NF-κB1–Dependent GM-CSF Production in CD4 T Cells

Ian K. Campbell, Annemarie van Nieuwenhuijze, Elodie Segura, Kristy O’Donnell, Elise Coghill, Mirja Hommel, Steve Gerondakis, José A. Villadangos and Ian P. Wicks
J Immunol May 1, 2011, 186 (9) 5468-5477; DOI: https://doi.org/10.4049/jimmunol.1002923

Abstract

Rel/NF-κB transcription factors regulate inflammatory and immune responses. Despite possible subunit redundancy, NF-κB1–deficient (Nfkb1−/−) mice were profoundly protected from sterile CD4 T cell-dependent acute inflammatory arthritis and peritonitis. We evaluated CD4 T cell function in Nfkb1−/− mice and found increased apoptosis and selectively reduced GM-CSF production. Apoptosis was blocked by expression of a Bcl-2 transgene without restoring a disease response. In contrast with wild-type cells, transfer of Nfkb1−/− or GM-CSF–deficient CD4 T cells into RAG-1–deficient (Rag1−/−) mice failed to support arthritis induction. Injection of GM-CSF into Nfkb1−/− mice fully restored the disease response, suggesting that T cells are an important source of GM-CSF during acute inflammation. In Ag-induced peritonitis, NF-κB1–dependent GM-CSF production in CD4 T cells was required for disease and for generation of inflammatory monocyte-derived dendritic cells (MoDC), but not conventional dendritic cells. MoDC were identified in inflamed synovium and draining lymph nodes during arthritis. These MoDC produced high levels of MCP-1, a potent chemoattractant for monocytes. This study revealed two important findings: NF-κB1 serves a critical role in the production of GM-CSF by activated CD4 T cells during inflammatory responses, and GM-CSF derived from these cells drives the generation of MoDC during inflammatory disease.

Dendritic cells (DC) are crucial for the generation of immune responses and comprise several subsets in mice. In the absence of overt infection or inflammation (steady-state conditions), these include the resident and migratory conventional DC (cDC), plasmacytoid DC, and Langerhans DC (1, 2). After exposure to inflammatory stimuli, monocytes are recruited and differentiate into inflammatory monocyte-derived DC (MoDC), in a GM-CSF–dependent manner (3–5). Although numerous lymphoid and myeloid cell types can produce GM-CSF in vitro, it is unclear how the GM-CSF required for MoDC differentiation is provided in vivo.

The Rel/NF-κB family of transcription factors has been shown to contribute to the development and differentiation of multiple cell types (6), and regulate the expression of multiple gene products that are likely to contribute to human rheumatoid arthritis (RA) (7). The REL subunit has recently been identified in genome-wide association studies as a genetic risk factor for RA (8). However, few in vivo studies have examined individual Rel/NF-κB subunits in relevant experimental models of inflammatory disease.

There are five mammalian Rel/NF-κB proteins: NF-κB1 (p50 and its precursor, p105), NF-κB2 (p52 and its precursor, p100), RelA (p65), RelB, and c-Rel (6). Rel/NF-κB proteins form homodimers and heterodimers, with the most common transcription factor comprising a dimer of NF-κB1 and RelA. In most cells, the majority of Rel/NF-κB is maintained as inactive cytoplasmic complexes by a family of inhibitor proteins (IκB). In response to multiple immune stimuli that include cytokines, TCR stimulation, and TLR ligands, IκB proteins are phosphorylated and targeted for proteasome-mediated degradation. In the case of NF-κB1, proteolytic processing of p105 produces p50, which homodimerizes or heterodimerizes before translocating to the nucleus and binding κB elements found in the transcriptional regulatory regions of target genes (6). However, control of gene transcription by NF-κB family members is complex, and a number of modulating cofactors have been identified (9). NF-κB1 lacks an intrinsic transactivation domain, and NF-κB1 homodimers generally inhibit transcription via the histone deacetylase, HDAC1 (10, 11). NF-κB1 activates gene transcription through interaction with its major transactivating dimer partners RelA and c-Rel, or by interacting with the coactivators, Bcl-3 (12) and CREB-binding protein (13).

We previously reported that NF-κB1–deficient (Nfkb1−/−), but not c-Rel–deficient (Rel−/−), mice were protected from CD4 T cell-dependent acute inflammatory arthritis (14). We also previously reported that MoDC differentiated from monocyte precursors in the spleen during acute inflammatory peritonitis (3). In this study, we show that Nfkb1−/− mice are also protected from acute inflammatory peritonitis, and that protection in these models of acute inflammation was due to a critical requirement for the NF-κB1 subunit in CD4 T cells. Protection from disease was not due to enhanced T cell apoptosis or to the failure to upregulate cellular activation markers or cytokine receptors, but to a relatively selective defect in the production of GM-CSF. We also show that the differentiation of MoDC, which could be identified in inflamed synovial tissue, draining lymph nodes (LN), and the spleen during both inflammatory arthritis and peritonitis, was dependent on the expression of NF-κB1 in CD4 T cells, and T cell-derived GM-CSF. Our data provide direct experimental evidence that NF-κB–mediated production of GM-CSF by activated CD4 T cells is pivotal for MoDC differentiation in inflammatory conditions.

Materials and Methods

Mice

Mice were backcrossed onto C57BL/6 (B6), as indicated. Nfkb1−/− mice (15) and IL-1R1–deficient (Il1r-1−/−) mice (16) had eight and six crosses, respectively. RAG-1–deficient (Rag1−/−) mice (17) (15 crosses), GM-CSF–deficient (GM-CSF−/−) mice (18) (>8 crosses), B6.Ly5.1, and CD11c-Diphtheria toxin receptor (DTR) transgenic (Tg) mice (19) were obtained from the Walter and Eliza Hall Institute (WEHI) animal services (Kew, VIC, Australia). Nfkb1−/−/OT-II mice and Nfkb1−/−/vav-Bcl2 mice were generated by crossing Nfkb1−/− with OT-II (20) and vav-Bcl2 Tg B6 mice (21), respectively. B6 mice were used as wild-type (WT) controls in all experiments. All mice were aged >7 wk. The WEHI animal ethics committee approved all mouse procedures.

Acute monoarticular arthritis

As previously described (14), on day 0, mice were injected intra-articularly into the knee joint with 200 μg methylated BSA (mBSA; Sigma). Human rIL-1β (250 ng; PeproTech) was injected s.c. into the footpad on days 0, 1, and 2. In some experiments, mice were injected i.p. three times daily for 7 d with 120,000 U murine rGM-CSF (sp. act. 1 × 108 U/mg; WEHI), beginning on day 0. Mice were sacrificed at day 7 and the knees processed for histological assessment. Sections were graded from 0 (normal) to 5 for the severity of 5 histological features of arthritis (i.e., total out of 25) by a blinded assessor.

Induction of acute peritonitis

Acute peritonitis was induced as described previously (3, 22). On days 0 and 14, mice were injected intradermally at the base of the tail with 100 μl mBSA (1 mg/ml in CFA). On day 21, mice were injected i.p. with 100 μg mBSA to induce inflammation. Peritoneal exudate and spleen were analyzed by flow cytometry on day 23.

Isolation of CD4 T cells

Single-spleen cell suspensions were incubated with FITC-conjugated anti-CD4 mAb (Caltag) followed by anti-FITC microbeads (Miltenyi Biotec). Magnetically labeled cells were positively selected using an Automacs Magnetic Cell Sorter (Miltenyi Biotec). Sorted cell purity was >95% by flow cytometry.

CD4 T cell transfer to Rag1−/− and irradiated Nfkb1−/− mice

CD4 T cells (5 × 106) were injected i.v. into either nonirradiated Rag1−/− mice or sublethally gamma-irradiated (4.0 Gy) WT or Nfkb1−/− mice. Thirty minutes before cell transfer, mice were injected i.p. with anti-CD8 mAb (0.3 mg each of YTS169 and 53-6.7) to deplete any cotransferred CD8 cells. After 4–5 wk, donor T cells were assessed in the peripheral blood of recipients by flow cytometry and acute arthritis was induced. In experiments with WT or Nfkb1−/− mice (each Ly5.2), evaluation of CD4 reconstitution was facilitated by the use of Ly5.1 donor CD4 T cells.

In vitro stimulation of CD4 T cells

CD4 T cells were cultured in 96-well plates at 37°C, 5% CO2 in RPMI containing 10% FCS, 20 μM 2-ME, 2 mM glutamine, and 1 mM sodium pyruvate. One or more of the following stimuli were added: human rIL-2 (1000 U/ml; Cetus), 1–5 μg/ml anti-CD3 (clone 145-2C11; WEHI; adsorbed to the wells for 2 h at 37°C), and 1–10 μg/ml anti-CD28 (clone 37.51; WEHI) mAb. For Th17 polarizing conditions, cultures included anti-CD3, anti-CD28, human rTGF-β1 (5 ng/ml; R&D Systems), human rIL-6 (20 ng/ml; R&D Systems), anti–IFN-γ (10 μg/ml; clone R4-6A2; WEHI), and anti–IL-4 (10 μg/ml; clone 11B11; WEHI). Cell surface activation marker expression was determined by flow cytometry using PE-conjugated mAb to CD25 (clone PC61; Pharmingen) or CD69 (clone H1.2F3; Pharmingen). Dead cells were excluded by propidium iodide staining (Calbiochem).

RNase protection analysis

Total RNA from 1 × 107 cells was analyzed using the RiboQuant RNase Protection System (Pharmingen) with the murine cytokine receptor template set mCR-1. Dried polyacrylamide gels were visualized after an overnight exposure by PhosphorImager (Molecular Dynamics) using Image Quant software.

Intracellular cytokine staining

CD4 T cells (1 × 105) were cultured with 10 μg/ml anti-CD3 and 2 μg/ml anti-CD28. For Th17 conditions, 5 μg/ml anti-CD3 and 1 μg/ml anti-CD28 were used, and TGF-β1, IL-6, anti–IFN-γ, and anti–IL-4 were added, as described earlier. Monensin A (1 μM; Sigma) was added during the last 4 h of culture. Cells were restained for CD4, then fixed, permeabilized, and stained for intracellular cytokines using the Cytofix/Cytoperm kit (BD Biosciences) and PE-labeled Ab (Pharmingen).

Cytokine production by naive CD4 T cells

Splenic CD4 T cells were prepared from naive mice, further stained with anti–CD25-PE (clone PC61), anti–CD44-PE (clone 1M7), anti–CD62L-Biotin (clone MEL-14), and streptavidin-allophycocyanin (all from Pharmingen), and sorted by flow cytometry (MoFlo Cytomation) to obtain naive CD4+CD44−CD62L+CD25− T cells. Cells (1 × 105) were cultured in a 96-well plate in the presence of 10 μg/ml anti-CD3 + 2 μg/ml anti-CD28. Culture supernatants were removed after 72 h and evaluated for cytokines and chemokines using the Bio-Plex mouse cytokine 23-plex panel (Bio-Rad Laboratories) according to the manufacturer’s instructions. The assay was read on a Bio-Plex 200 instrument and analyzed using the Bio-Plex Manager V5.0 software.

Nuclear and cytoplasmic extracts, EMSA, and Western blot

Nuclear and cytoplasmic extracts were prepared and EMSA performed as previously described (23). In brief, 1.5-μg nuclear extracts were incubated with [32P]-labeled double-stranded κB3 probe (5′-CGTAAGCAGCGGGAAATCCCCC-3′) (23) on ice for 20 min. Ab (2 μg; Santa Cruz Biotechnology) specific to NF-κB1 (sc-1192X), RelA (sc-372X), c-Rel (sc-71X), NF-κB2 (sc-298X), and RelB (sc-226X) were added for a further 20 min on ice. Reactions underwent electrophoresis at 250 V for 2 h on a 5% nondenaturing polyacrylamide gel. Gels were dried onto 3MM paper and exposed to film.

For Western blots, the NuPage system (Invitrogen) was used with 10 μg extract per lane. Bands were transferred to a PVDF membrane using iBlot (Invitrogen) and the membrane was incubated with rabbit anti–c-Rel Ab (sc-71X), followed by HRP-conjugated sheep anti-rabbit IgG (Chemicon). Bands were visualized using the ECL Western blotting detection system (Amersham Biosciences).

Depletion of CD11c+ DC during acute inflammatory arthritis

Lethally irradiated (2 × 500 rad) WT mice were injected i.v. with CD11c-DTR Tg (19) or WT bone marrow cells. Mice were allowed to reconstitute for 5 wk. Conventional CD11c+ DC were depleted by i.p. injection of 100 ng Diphtheria toxin on days −1 to 4 of the acute arthritis model. Control mice received saline. DC depletion was >60%.

Detection and sorting of DC in lymphoid and synovial tissues

Light density cells were purified as previously described (24). In brief, spleens or LN were digested with DNase I (0.1%; Roche Molecular Biochemicals) and collagenase (1 mg/ml, type II; Worthington Biochemical); synovial tissue was digested as described previously (25) and centrifuged in Nycodenz medium (density 1.082 g/cm3; Axis-Shield, Norway). Light density cells were then stained for flow cytometric analysis using Ab against CD11b, CD11c, Ly6C, and MHC II (all from Pharmingen). Dead cells were excluded by propidium iodide staining. For experiments with sorted DC, cDC (CD11chiCD11b+/−Ly6C−MHCIIhi) and MoDC (CD11cintCD11bhiLy6C+/hiMHCIIhi) were sorted using a DIVA automated cell sorter (BD Biosciences).

Cytokine production by cDC and MoDC purified from inflamed synovial tissue

DC were sorted from inflamed synovial tissue, obtained at the peak of acute inflammatory arthritis (day 7) as described earlier. Cells (5 × 104) were cultured in 200 μl RPMI containing 10% FCS, 20 μM 2-ME, 2 mM glutamine, and 1 mM sodium pyruvate for 16 h at 37°C, 5% CO2, in triplicate. Culture supernatants were removed after 16 h and evaluated for cytokines and chemokines using the Bio-Plex mouse cytokine 23-plex panel (Bio-Rad Laboratories). The assay was read on a Bio-Plex 200 instrument and analyzed using the Bio-Plex Manager V5.0 software.

Measurement of apoptosis

For analysis of apoptosis, cells were washed in PBS and resuspended in hypotonic buffer containing 50 μg/ml propidium iodide, 0.1% trisodium citrate (Merck Chemicals, Kilsyth, VIC, Australia) and 0.1% Triton X-100 (Calbiochem), and incubated at 4°C for >2 h. Flow cytometry was used to determine the percentage of DNA fragmentation as described previously (26).

MLR and presentation of OVA by DC

For MLR, splenic cDC from WT or Nfkb1−/− mice were mixed with CD4 T cells (1 × 105 cells/well) from BALB/c mice. For the reciprocal MLR, CD4 T cells from WT or Nfkb1−/− mice were mixed with BALB/c irradiated (20 Gy) stimulator splenocytes. For Ag-specific presentation, WT and Nfkb1−/− mice were injected i.v. with 3 mg OVA (Sigma) or PBS. After 16 h, spleens were harvested and cDC were purified as described previously (27). OVA-specific CD4 T cells were purified from spleens of OT-II mice and cocultured (1 × 104 cells/well) with increasing numbers of DC. In reciprocal experiments, WT cDC from OVA-injected mice were cultured with WT or Nfkb1−/− OT-II CD4 T cells. For each assay, 1 μCi/well [3H]thymidine (Amersham Biosciences) was added for the final 8 h of culture, cells were harvested onto glass fiber filters, and radioactive incorporation was counted on a scintillation counter as a measure of CD4 T cell proliferation.

Statistics

For histological scores, the Mann–Whitney two-sample rank test was used to determine the level of significance between means of groups. For other comparisons, two-way ANOVA and Student t test were used, where indicated. Tests were two-tailed, and p < 0.05 was considered statistically significant.

Results

A critical role for NF-κB1 in CD4 T cells during acute inflammatory arthritis

To investigate the selective protection of Nfkb1−/− mice from mBSA/IL-1 acute inflammatory arthritis, a CD4 T cell-dependent model (25), we transferred purified WT or Nfkb1−/− splenic CD4 T cells into Rag1−/− mice, which have normal APC function (28). There were similar percentages of WT and Nfkb1−/− CD4 T cells in the peripheral blood of recipient Rag1−/− mice immediately before arthritis induction (% of total leukocytes: 18.4 ± 2.7 versus 17.3 ± 2.2, respectively; mean ± SEM; n = 4). As expected, arthritis could not be induced in naive Rag1−/− mice (Fig. 1A); however, Rag1−/− mice that received WT CD4 T cells (Fig. 1B) developed acute arthritis, with features and severity similar to those of B6 mice (Fig. 1D). In contrast, Rag1−/− mice receiving Nfkb1−/− CD4 T cells had markedly reduced disease compared with recipients of WT CD4 T cells, with very mild histological features (Fig. 1C, 1E) that resembled the minor pathology seen in Nfkb1−/− mice (14).

FIGURE 1.
FIGURE 1.

CD4 T cells require NF-κB1, but not responsiveness to IL-1, for acute inflammatory arthritis. WT, Nfkb1−/−, and Il1r-1−/− splenic CD4 T cells were transferred to Rag1−/− mice as described in Materials and Methods. Acute inflammatory arthritis was induced after 5 wk and the joints examined 7 d later. Representative H&E-stained sections (original magnification ×200) of knee joints from Rag1−/− mice that had received (A) no cells, (B) WT cells, or (C) Nfkb1−/− cells. D, Disease in a WT B6 mouse is shown for comparison. E, Total histological scores (mean ± SEM; n = 7 joints) for Rag1−/− mice receiving either WT or Nfkb1−/− cells. Mice receiving no cells had similar scores to those transferred with Nfkb1−/− cells (not shown). *p < 0.05. Results are representative of three experiments. F, Total histological scores (mean ± SEM; n = 20 joints, pool of 3 experiments) are shown for Rag1−/− mice receiving either WT or Il1r-1−/− cells. G, Restoration of disease susceptibility in Nfkb1−/− mice by transfer of WT CD4 T cells. Total histological scores (mean ± SEM; n = 20 joints) are shown for sublethally irradiated WT or Nfkb1−/− mice receiving WT cells.

A link between IL-1R signaling and NF-κB, coupled with the ability of IL-1 to stimulate T cells, raised the possibility that the reduced disease response in Rag1−/− mice injected with Nfkb1−/− CD4 T cells might be because of the inability of these cells to respond to IL-1. Il1r-1−/− CD4 T cells were therefore transferred into Rag1−/− mice and arthritis was induced. Comparable disease between mice receiving Il1r-1−/− and WT CD4 T cells (Fig. 1F) indicated direct CD4 T cell responsiveness to IL-1 was not required.

To confirm the exclusive requirement for NF-κB1 expression in CD4 T cells, we injected sublethally irradiated Nfkb1−/− mice with WT CD4 T cells, which restored a normal disease response in the mice (Fig. 1G). Collectively, these data demonstrate that NF-κB1 function in CD4 T cells is essential for the induction of mBSA/IL-1 acute arthritis, via a pathway that does not involve IL-1 acting directly on CD4 T cells.

Arthritis is not restored in Nfkb1−/− mice by the prosurvival protein Bcl2

The Rel/NF-κB family is important for cell survival through the inhibition of apoptosis, and indeed, we found that cell death was greater in activated Nfkb1−/− compared with WT CD4 T cells (Fig. 2Ai, 2Aii, 2B). To address whether increased apoptosis could account for the failure of Nfkb1−/− mice to succumb to acute arthritis, we intercrossed Nfkb1−/− mice with mice that expressed a Bcl2 Tg in all leukocytes (21). Although Bcl2 Tg expression prevented the enhanced apoptosis seen in activated Nfkb1−/− CD4 T cells in vitro (Fig. 2Aiv, 2B), it was unable to restore disease in vivo (Fig. 2C), thereby excluding reduced CD4 T cell survival as the explanation for the protection of Nfkb1−/− mice.

FIGURE 2.
FIGURE 2.

Increased apoptosis in TCR-stimulated Nfkb1−/− CD4 T cells is prevented by vav-Bcl2 Tg but has no effect on acute arthritis. A, CD4 T cells from (i) WT, (ii) Nfkb1−/−, (iii) vav-Bcl2 Tg, and (iv) Nfkb1−/−/vav-Bcl2 Tg mice were cultured for 72 h with anti-CD3 + anti-CD28. Propidium iodide-stained cells were evaluated by flow cytometry. Representative histogram plots are shown with the percentage of cells exhibiting DNA fragmentation (apoptosis) indicated. B, Mean percentages of apoptotic cells (± SEM; n = 8 mice). Representative of three experiments. *p < 0.0001, Student t test. C, Acute inflammatory arthritis was induced in WT and Nfkb1−/−/vav-Bcl2 Tg mice and evaluated 7 d later. Results show total histological scores (mean ± SEM; n = 10 joints) and are representative of two experiments. †p < 0.0001.

Nuclear induction of NF-κB1 after CD4 T cell activation

To understand how NF-κB1 signaling might contribute to CD4 T cell function after cellular activation, we performed EMSA on nuclear extracts of stimulated WT and Nfkb1−/− CD4 T cells. Because of the low frequency of mBSA-specific T cells in the acute arthritis model (25), these studies were performed using nuclear extracts isolated from CD4 T cells after activation with anti-CD3 + anti-CD28 mAb in culture. Three low-abundance complexes that bound specifically to a canonical κB site were detected in unstimulated WT CD4 T cells (Fig. 3A). Based on Ab supershifts, these complexes were NF-κB1/RelA heterodimers and NF-κB1 homodimers (Fig. 3A, upper panel). After activation, nuclear levels of NF-κB1 homodimers were markedly increased, as were NF-κB1/c-Rel heterodimers, albeit to a lesser extent (Fig. 3A, lower panel). Consistent with previous findings (29), NF-κB2 and RelB were undetectable in CD4 T cell nuclei. Nuclear extracts from unstimulated Nfkb1−/− CD4 T cells only bound weakly to the κB site (Fig. 3A, upper panel). However, a c-Rel complex was upregulated in stimulated Nfkb1−/− CD4 T cells (Fig. 3A, lower panel), which may represent c-Rel homodimers. The presence of enhanced levels of c-Rel in the nuclei of stimulated WT and Nfkb1−/− CD4 T cells was confirmed by Western blot (Fig. 3B). Together, these data suggest that NF-κB1/c-Rel heterodimers and NF-κB1 homodimers are the major NF-κB subunits that are normally induced in activated CD4 T cells, and that in the absence of NF-κB1, c-Rel homodimers predominate.

FIGURE 3.
FIGURE 3.

NF-κB1 in activated CD4 T cells. A, Subunit composition of Rel/NF-κB complexes in WT and Nfkb1−/− splenic CD4 T cells. Nuclear extracts of CD4 T cells that were unstimulated (top panel), or stimulated for 72 h with anti-CD3 + anti-CD28 (bottom panel) mAb, were examined by EMSA in the presence and absence of specific Ab to Rel/NF-κB subunits. Arrows indicate supershifts (SS). B, Western blot of c-Rel expression in the nuclear extracts and cytoplasm of WT (+/+) and Nfkb1−/− (−/−) CD4 T cells that were unstimulated or stimulated for 72 h with anti-CD3 + anti-CD28 mAb. All results are representative of two or more experiments.

Reduced production of the arthritogenic cytokine GM-CSF by Nfkb1−/− CD4 T cells

Involvement of NF-κB in the regulation of proinflammatory cytokine gene transcription is well documented. In addition to IL-1, we and others have shown that the acute arthritis model is critically dependent on the T cell-derived cytokines IL-2, GM-CSF (30, 31), and IL-17 (32). We therefore evaluated the effect of NF-κB1 deficiency on expression of early activation markers and cytokine receptors, and production of the key arthritogenic cytokines IL-2, GM-CSF, and IL-17 by CD4 T cells. Flow cytometric analysis of Nfkb1−/− CD4 T cells after stimulation with anti-CD3 + anti-CD28 mAb showed normal expression of CD25 (IL-2Rα) and CD69 (Fig. 4A), and RNase protection analysis showed normal levels of mRNA for the cytokine receptors IL-2R α- and β-chains, the IL-4Rα-chain, and the γ common chain (Fig. 4B). Intracellular cytokine staining of Nfkb1−/− CD4 T cells after stimulation with anti-CD3 + anti-CD28 mAb or under Th17 polarizing conditions showed normal or enhanced expression of IL-17 and IL-2, respectively, but a marked reduction in GM-CSF (Fig. 4C).

FIGURE 4.
FIGURE 4.

Activation marker expression and cytokine production by Nfkb1−/− CD4 T cells. A and B, Normal CD4 T cell activation markers and cytokine receptor expression in the absence of NF-κB1. A, WT or Nfkb1−/− CD4 T cells were cultured for 24 h in the absence (−) or presence (+) of anti-CD3 + anti-CD28 mAb, then stained with Ab to CD25 or CD69 and analyzed by flow cytometry. B, RNase protection analysis for cytokine receptor mRNA expression by WT (+/+) and Nfkb1−/− (−/−) CD4 T cells stimulated for 48 h with anti-CD3 + anti-CD28. L32 and GAPDH were RNA loading controls. C and D, Reduced GM-CSF production by CD4 T cells in the absence of NF-κB1. C, WT and Nfkb1−/− CD4 T cells were cultured for 72 h with anti-CD3 + anti-CD28, either without additives or under Th17 conditions (TGF-β + IL-6 + anti–IFN-γ + anti–IL-4). The production of cytokines critical for arthritis development (IL-2, IL-17, and GM-CSF) was evaluated by intracellular cytokine staining. Representative flow cytometry plots are shown; values are the percentage of cytokine-producing CD4 T cells. D, Naive WT and Nfkb1−/− CD4 T cells (CD4+CD44−CD62L+CD25−) were cultured for 72 h with anti-CD3 + anti-CD28, and cytokine production was evaluated in culture supernatants by Bio-Plex assay. Values show the mean (± SEM; n = 3 mice) concentrations in pg/ml. *p < 0.05, †p < 0.01, Student t test. All results are representative of two or more experiments.

Cytokine production was further evaluated in the culture supernatants of stimulated naive CD4 T cells (CD4+CD44−CD62L+CD25−) by Bio-Plex assay. In accordance with the intracellular cytokine staining, the supernatants of Nfkb1−/− CD4 T cells had a 5.8-fold reduction in GM-CSF production (Fig. 4D). These data also confirmed the relatively specific effect of NF-κB1 deficiency on naive CD4 T cell cytokine production, with normal or enhanced production of IL-2, IFN-γ, and TNF. There was also 1.7-fold less IL-3 production by the Nfkb1−/− CD4 T cells, and the Th2 cytokines IL-5 and IL-13 were reduced (2.6- and 1.8-fold, respectively), as previously reported for Nfkb1−/− CD4 T cells (33), but the latter were minor products compared with GM-CSF (log scale shown in Supplemental Fig. 1).

Critical dependence of acute joint inflammation on NF-κB1–dependent GM-CSF production by CD4 T cells

Collectively, our findings suggested that Nfkb1−/− mice were protected from acute arthritis through the inability of CD4 T cells to produce the critical arthritogenic cytokine, GM-CSF (34). To further evaluate the relation between CD4 T cells and GM-CSF in acute arthritis, we transferred WT or GM-CSF−/− CD4 T cells to Rag1−/− mice and arthritis was induced. A substantial reduction in the disease response was observed in mice that received GM-CSF−/− CD4 T cells, indicating CD4 T cell-derived GM-CSF is indeed essential for full disease expression (Fig. 5A). Because the effect was not as profound as NF-κB1 loss in CD4 T cells, and IL-3 could also be involved, we sought to confirm the importance of GM-CSF in disease development in Nfkb1−/− mice. Injection of GM-CSF throughout the 7 d of the model fully restored a normal response in the Nfkb1−/− mice, for all features of disease (Fig. 5B).

FIGURE 5.
FIGURE 5.

NF-κB1 and GM-CSF production by CD4 T cells are critical for acute inflammatory arthritis. A, WT and GM-CSF−/− CD4 T cells were transferred to Rag1−/− mice (see Materials and Methods). Arthritis was induced after 5 wk and the joints examined 7 d later. Total histological scores (mean ± SEM; n = 24–25 joints) are shown for Rag1−/− mice receiving either WT or GM-CSF−/− cells. Data are pooled from two experiments. *p < 0.001. B, Arthritis was induced in WT and Nfkb1−/− mice that were also injected thrice daily with either GM-CSF (Nfkb1−/− mice only) or vehicle. Results show the mean histological scores (± SEM; n = 10 joints) for arthritis features at day 7. GM-CSF injections fully restored the disease response in Nfkb1−/− mice (p < 0.001, compared with vehicle-treated Nfkb1−/− mice).

Ag presentation by DC is required for the induction of acute joint inflammation

The dependence of acute arthritis on CD4 T cells infers that APC are essential to disease. To establish the importance of DC in this model, we used DTR Tg mice that were depleted (by >60%) of CD11chi cells, as previously described (19). Arthritis was markedly reduced in the CD11c-depleted mice, but some residual disease was present (Fig. 6A). In this system, Diphtheria toxin preferentially depletes CD11chi cells (i.e., cDC) over CD11cint cells (i.e., plasmacytoid DC and MoDC) (35), suggesting that cDC are necessary but not sufficient for a complete disease response and that CD11cint inflammatory MoDC could be responsible for the residual disease. To establish the role of NF-κB1 itself in cDC, we evaluated Nfkb1−/− cDC in vitro for allogeneic (MLR) and Ag-specific (OVA) T cell stimulation (Fig. 6B–D). In agreement with previous studies (36, 37), we found that Nfkb1−/− cDC were similar to WT cDC in both assays. In contrast, when we crossed OT-II mice to Nfkb1−/− mice, OVA-specific, Nfkb1−/− CD4 T cells were unresponsive to stimulation in vitro with splenic WT cDC that had been loaded with OVA in vivo (Fig. 6E).

FIGURE 6.
FIGURE 6.

DC are essential for acute inflammatory arthritis, but DC function is not dependent on NF-κB1. A, Dependence of acute arthritis on CD11c+ DC. Arthritis was induced in WT mice reconstituted with DTR-CD11c or WT bone marrow. DC were depleted by i.p. injection of 100 ng Diphtheria toxin on days −1 to 4. Control mice received saline. Data show the total histological scores (mean ± SEM; n = 14 joints, pool of 2 experiments). *p < 0.001. B and C, MLR. B, BALB/c CD4 T cells cocultured with purified cDC isolated from WT or Nfkb1−/− mice, or (C) CD4 T cells from WT or Nfkb1−/− mice cocultured with irradiated splenocytes from BALB/c mice. D and E, Purified splenic cDC were isolated from (D) OVA- or vehicle-injected WT and Nfkb1−/− mice and cultured with WT OT-II CD4 T cells, or from (E) OVA-injected WT mice and cultured with WT or Nfkb1−/− OT-II CD4 T cells (n = 3 mice). For all experiments, T cell proliferation was assessed by [3H]thymidine (TdR) incorporation. Results show the mean cpm (± SEM; n = 6) and are representative of three experiments. p < 0.0001 (C) and p < 0.001 (E), by two-way ANOVA, for Nfkb1−/− versus WT cells.

To address the involvement of GM-CSF-dependent MoDC (defined as CD11cintCD11bhiLy6C+/hiMHCIIhi) in acute inflammatory arthritis, we dissociated inflamed synovial tissue on day 7 of disease. Flow cytometry revealed that MoDC comprise 85% of the CD11c+ cell population in inflamed synovial tissue, and these cells were also found in the draining popliteal LN at the peak of disease in WT mice (Fig. 7A, 7C). In contrast, MoDC could not be detected in popliteal LN from naive mice. The number of these cells was markedly reduced in popliteal LN from GM-CSF−/− mice on day 7 of inflammatory arthritis, demonstrating that differentiation of these cells during inflammation in vivo is dependent on GM-CSF (Fig. 7A, 7B). To establish that the MoDC in synovial tissue can drive arthritis, we purified these cells, as well as cDC, from digested synovial tissue of 10 WT mice on day 7 of arthritis. Analysis of cytokine production after a short period of in vitro culture by Bio-Plex revealed that compared with cDC, MoDC produce more of the arthritogenic cytokine IL-6 and high levels of the chemokine MCP-1. Production of MCP-1 by these cells suggests a positive amplification loop to recruit more monocytes and MoDC to the site of inflammation (Fig. 7D).

FIGURE 7.
FIGURE 7.

MoDC are present in draining LN and inflamed synovial tissue from WT mice but not GM-CSF−/− mice in acute inflammatory arthritis. Single-cell suspensions of draining popliteal LN from WT mice, or GM-CSF−/− mice (A), or light density cell suspensions of inflamed synovial tissue (C) at the peak of acute inflammatory arthritis (day 7) were stained with Ab for CD11b, CD11c, Ly6C, and MHC II, and analyzed by flow cytometry. Naive WT mice were used as a control. CD11cintCD11b+Ly6C+MHCII+ MoDC could be detected in inflamed tissue, but not naive tissue or GM-CSF−/− tissue. B, Numbers of MoDC detected in popliteal LN from WT or GM-CSF−/− mice at the peak of acute inflammatory arthritis (day 7). Mean ± SEM, n = 5–6/group. *p < 0.05, compared with mBSA-treated WT mice, Student t test. Experiment was performed twice. D, Cytokine production by purified cDC and MoDC from inflamed synovial tissue of WT mice (pg/ml).

NF-κB1 in CD4 T cells is required for the induction of acute peritonitis

To determine whether these findings in experimental arthritis apply to other models of T cell-dependent inflammation, and to access tissue-infiltrating cells more readily, we examined Ag-induced peritonitis. This model is also dependent on GM-CSF (22), and we established in an earlier study that MoDC differentiate in the spleen from monocyte precursors during this response (3). Rag1−/− mice were reconstituted with purified CD4 T cells derived from WT or Nfkb1−/− mice, and acute peritonitis was induced. Mice were primed and boosted with mBSA in CFA, challenged by i.p. injection of mBSA 7 d later, and evaluated after 48 h. Mice that received Nfkb1−/− CD4 T cells developed markedly reduced peritoneal exudate (total white blood cells and neutrophils) compared with mice receiving WT CD4 T cells, indicating that NF-κB1 was required in CD4 T cells for disease (Fig. 8A). Evaluation of cytokines in the peritoneal exudate showed reduced levels of proinflammatory cytokines (IL-6, IL-1β, TNF) in mice that received Nfkb1−/− CD4 T cells, in particular, GM-CSF (Fig. 8B), which was one of the most prevalent cytokines detected. Therefore, NF-κB1 in CD4 T cells, as well as GM-CSF, is required in both models of acute tissue inflammation (i.e., arthritis and peritonitis).

FIGURE 8.
FIGURE 8.

NF-κB1 in CD4 T cells is critical for acute inflammatory peritonitis and the differentiation of MoDC, but not cDC, in the spleen. WT or Nfkb1−/− CD4 T cells were transferred into Rag1−/− mice, acute peritonitis was induced (control mice received saline), and peritoneal exudate was evaluated after 48 h. A, Number of white blood cells (WBC; left panel) and neutrophils (right panel) in peritoneal washes (mean ± SEM; n = 10–13 mice). *p < 0.05, Student t test, compared with mBSA-treated WT mice. Data pooled from four experiments. B, Analysis of cytokines in peritoneal exudates. Results show the mean (± SEM; n = 4 mice/group) concentration of cytokines as measured by Bio-Plex Array. C, Splenic DC were examined by flow cytometry 48 h after i.p. challenge. Light density spleen cells were gated for CD11b+CD11cint/+. MoDC were defined as Ly-6C+MHCII+ and cDC as Ly-6C−MHCII+. Ungated cells are monocytes. D, Numbers of splenic MoDC (left panel) and cDC (right panel). Results show the mean ± SEM (n = 15–16 mice/group, pool of 4 experiments). *p < 0.05.

NF-κB1 is required for generation of inflammatory DC

We showed previously that CD11cintCD11bhiLy6C+/hiMHCIIhi MoDC, which we detected in inflamed joints during arthritis, also appear in the spleens of mice with mBSA-induced acute peritonitis (3). To evaluate the importance of Nfkb1−/− in CD4 T cells in the generation of MoDC in this model, we reconstituted Rag1−/− mice with purified CD4 T cells derived from WT or Nfkb1−/− mice, and acute peritonitis was induced. Splenic MoDC numbers were consistently reduced in mice reconstituted with Nfkb1−/− CD4 T cells. In contrast, the number of cDC, although more variable over multiple experiments, was not statistically different (Fig. 8C, 8D).

Discussion

The NF-κB family is of central importance to immune responses and inflammation, but there has been relatively little in vivo evaluation of the contributions of individual subunits. This study of NF-κB1 in inflammatory arthritis and peritonitis revealed two important findings: NF-κB1 serves a critical role in the production of GM-CSF by activated CD4 T cells during inflammatory responses, and GM-CSF derived from these cells preferentially drives MoDC development during inflammatory disease.

Altered Nfkb1−/− CD4 T cell function was shown to be responsible for resistance to two models of acute inflammation: arthritis and peritonitis. Rag1−/− mice, which were rendered susceptible to disease by transfer of WT CD4 T cells, remained refractory when engrafted with Nfkb1−/− CD4 T cells. Arthritis susceptibility was restored in Nfkb1−/− mice by injection of WT CD4 T cells, indicating the defect was T cell intrinsic and that trans-endothelial migration and leukocyte recruitment to joints was not impaired in Nfkb1−/− mice. Although both CD4 T cells and IL-1 are essential components of this arthritis model, they are not directly interdependent. We found that CD4 T cells do not require a response to IL-1 in acute inflammatory arthritis, because Il1r-1−/− CD4 T cells could transfer disease susceptibility when injected into Rag1−/− mice.

One potential explanation for the reduced function of Nfkb1−/− CD4 T cells would be globally impaired activation. However, after strong polyclonal stimulation, Nfkb1−/− CD4 T cells displayed a normal activation profile (CD69, CD25) and cytokine receptor (IL-2R, IL-4R, γc) expression. It has been proposed that persistent Rel/NF-κB activation and inhibition of apoptosis contributes to the chronic inflammatory synovitis of human RA (38). The diminished [3H]thymidine incorporation we observed for both alloantigen and Ag-specific stimulation of Nfkb1−/− CD4 T cells could be explained by enhanced T cell apoptosis. However, we found that enforced expression of Bcl-2 in the hemopoietic compartment did not influence acute arthritis susceptibility in Nfkb1−/− mice, although it did improve the viability of activated T cells in vitro. Although reduced CD4 T cell survival involving the Bcl-2–dependent pathway does not therefore account for the reduced disease response in Nfkb1−/− mice, we cannot exclude enhanced Nfkb1−/− CD4 T cell apoptosis via the extrinsic, death receptor pathway contributing to disease resistance in this model.

A notable further finding was reduced cytokine production by activated Nfkb1−/− CD4 T cells, particularly GM-CSF. Remarkably, this was relatively selective, and activated Nfkb1−/− CD4 T cells actually showed enhanced production of TNF and IFN-γ compared with WT cells. Rag1−/− mice reconstituted with GM-CSF−/− CD4 T cells developed markedly less inflammatory arthritis than mice given WT cells, and injection of GM-CSF fully restored the disease response in Nfkb1−/− mice. Although NF-κB1 has a role in GATA3 expression and Th2 differentiation (33), there is only a minor contribution to mBSA/IL-1–induced arthritis from the Th2 cytokine, IL-4 (31). In contrast, disease is markedly dependent on endogenous GM-CSF, IL-2 (30, 31), and IL-17 (32). Given that IL-2 production by Nfkb1−/− CD4 T cells was enhanced and IL-17 expression was unchanged, the reduced GM-CSF production by Nfkb1−/− CD4 T cells raised the possibility that this source of GM-CSF could be pivotal in the arthritis model. Although IL-3 production was reduced in Nfkb1−/− CD4 T cells, and thus could also account for the mild disease response in Nfkb1−/− mice, injection of GM-CSF alone was sufficient to restore a normal disease response. It is interesting to note that the T cells of Rel−/− mice also have reduced GM-CSF production (39), whereas these mice exhibited a normal disease response in the acute arthritis model (14). A possible explanation lies in the markedly enhanced GM-CSF production by activated Rel−/− peritoneal macrophages (39), which may provide sufficient GM-CSF for a disease response.

Interestingly, each of the genes for the cytokines that were deficient in Nfkb1−/− CD4 T cells (GM-CSF, IL-3, and the Th2 cytokines IL-4, IL-5, and IL-13) map within a 0.7-Mb region on mouse chromosome 11 (and an orthologous region on human chromosome 5). In contrast, under the same conditions, the production of the Th2 cytokine IL-10, which maps to mouse chromosome 1, was normal in Nfkb1−/− CD4 T cells. It is therefore tempting to speculate that in activated CD4 T cells, NF-κB1 may directly or indirectly interact with noncoding elements responsible for the coordinate regulation of these genes on chromosome 11. There is evidence for such control regions in human chromosome 5, which are phylogenetically conserved (40, 41).

Our data show that cytoplasmic c-Rel was able to translocate to the nucleus in stimulated Nfkb1−/− CD4 T cells, suggesting that NF-κB1 is not required for nuclear shuttling of this NF-κB subunit. Although c-Rel was detected in the nuclei of stimulated Nfkb1−/− CD4 T cells and could bind to a consensus κB site, it was unable to effectively activate gene transcription in the absence of NF-κB1. GM-CSF was one of the most abundantly produced cytokines in WT CD4 T cells and was proportionally the most severely reduced in Nfkb1−/− CD4 T cells. In T cells, GM-CSF production is mainly controlled at the transcriptional level, is usually transient, and depends on activation through the TCR and costimulatory molecules, such as CD28 (42, 43). These findings are consistent with the demonstration of NF-κB1 and c-Rel binding to the GM-CSF promoter (44, 45). We conclude that the most likely reason for the Nfkb1−/− phenotype in the inflammation models is lack of NF-κB1/c-Rel heterodimer transcriptional activity in activated CD4 T cells. However, p50 homodimers have been reported to promote gene transcription in some circumstances (13), so we cannot exclude NF-κB1 homodimers as transcriptional activators in CD4 T cells.

Although originally described as a hemopoietic cell growth factor, GM-CSF can have numerous proinflammatory effects on mature myeloid cells (reviewed in Ref. 34), and GM-CSF is a potent inducer of DC differentiation. In a peritonitis model that is dependent on CD4 T cells and GM-CSF, we found reduced splenic MoDC numbers in Rag1−/− mice injected with Nfkb1−/− CD4 T cells. These mice had normal splenic cDC levels, suggesting that this cell type is unable to drive inflammation in the absence of sufficient MoDC. This possibility is supported by our observation that the disease response in acute inflammatory arthritis was markedly reduced, but not abolished, when CD11chi cells were depleted using the CD11c-DTR system. GM-CSF is also critical for the development of experimental autoimmune encephalitis, and in this model, GM-CSF production by effector T cells was shown to activate microglial cells (46). Interestingly, IL-23 stimulated the production of GM-CSF, as well as IL-17 by murine T cells (47), which may help explain earlier findings with GM-CSF−/− T cells (48). GM-CSF promoted the induction and survival of autoimmune Th17 effector cells in a CD4 T cell-mediated model of myocarditis (49). T cells appear to lack GM-CSF receptors; thus, this effect was most likely via enhancement of IL-6 and IL-23 production from APC. Although it was previously thought that GM-CSF−/− mice had normal DC (50), GM-CSF was recently found to be required for the accumulation of Langerin+ CD103+ DC (51), and GM-CSF production by CD8+ T cells, in combination with microbial stimuli, was postulated as a “licensing factor” for DC (52).

GM-CSF can be produced by macrophages but, interestingly, apparently not by DC (49). In vivo, ongoing DC recruitment and differentiation in the setting of tissue inflammation may be highly dependent on other cell types (28) and the supply of critical cytokines such as GM-CSF (51, 52). Transfer of WT CD4 T cells into Nfkb1−/− mice was sufficient to restore a normal disease response to these mice, indicating that APC function was not limiting. DC function in Nfkb1−/− mice was normal for Ag-specific (OVA) and allogeneic (MLR) T cell stimulation, indicating that Ag uptake, processing, and presentation by cDC was not impaired in Nfkb1−/− mice. These data confirm and extend the findings of previous reports (36, 37) that show normal cDC development and function in Nfkb1−/− mice, with a marginal reduction in numbers being the main difference to WT mice. Importantly, we found MoDC in the directly involved joint tissue of WT mice with inflammatory arthritis, and MoDC were identified as a likely target for CD4 T cell-derived GM-CSF. MoDC may be especially relevant during persistent tissue inflammation (53). It has been suggested that, in such conditions, these cells may even be superior APC, as steady-state cDC have been exposed to the immunosuppressive tissue environment (4). In a recent study, MoDC arising in response to an infectious challenge were found to localize to the T cell areas in LN, favoring functional interaction. These MoDC were found to produce a strong MLR and were actually superior to cDC in Ag presentation, including cross-presentation (54). Our finding that MoDC derived from inflamed synovial tissue produce high levels of MCP-1, a potent chemoattractant for monocytes, also suggests that MoDC might provide a positive amplification loop through which monocytes are recruited to sites of inflammation and then differentiate into DC.

Collectively, these data raise the intriguing possibility of a positive feedback loop between CD4 T cells and DC, whereby NF-κB signaling in CD4 T cells that initially encounter Ag presented by cDC causes GM-CSF production that specifically promotes the recruitment or development of MoDC from monocytes (Supplemental Fig. 2).

The important role the Rel/NF-κB transcription factors serve in regulating immune and inflammatory responses makes this pathway an excellent potential target for therapeutic intervention in inflammatory diseases such as RA. Indeed, inhibition of NF-κB in T cells has been shown to alleviate collagen-induced arthritis (55). However, global inhibition of Rel/NF-κB, even within a single cell type, may severely compromise the immune system. Anti–GM-CSF clinical trials are planned or under way in several candidate diseases, including RA; thus, it is vital to understand where, when, and how GM-CSF production is controlled. Our findings show that NF-κB1 deletion alone is sufficient to markedly reduce GM-CSF production by activated CD4 T cells, and also demonstrate the importance of T cell-derived GM-CSF in promoting MoDC development and inflammatory disease.

Disclosures

I.K.C. previously received a royalty payment related to a patent covering GM-CSF antagonism. The work in this study was not supported by and was not related to the patent.

Acknowledgments

We thank D. Baltimore (Caltech) and W. Sha (University of California, Berkeley, Berkeley, CA) for the provision of Nfkb1−/− mice, J. Adams (WEHI) for the vav-Bcl2 Tg mice, T. Kay (St. Vincent’s Institute, Melbourne, VIC, Australia) for the Il1r-1−/− mice, and P. Morgan (WEHI) for the murine rGM-CSF. We are grateful to F. Rodda for animal care, S. Mihajlovic for histology, and A. Strasser and M. Blewitt for helpful discussions.

Footnotes

  • This work was supported by the Reid Charitable Trusts, the Arthritis Foundation of Australia, the National Health and Medical Research Council of Australia, the Anti-Cancer Council of Victoria, and CellCept Australia Research Grant 971274 (to S.G.). I.P.W. and I.K.C. are supported by fellowships from the Australian National Health and Medical Research Council. E.S. is supported by Marie Curie fellowships from the European Commission.

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    B6
    C57BL/6
    cDC
    conventional dendritic cells
    DC
    dendritic cells
    DTR
    Diphtheria toxin receptor
    GM-CSF−/−
    GM-CSF–deficient
    Il1r-1−/−
    IL-1R1–deficient
    LN
    lymph nodes
    mBSA
    methylated BSA
    MHC II
    MHC class II
    MoDC
    monocyte-derived DCs
    Nfkb1−/−
    NF-κB1–deficient
    RA
    rheumatoid arthritis
    Rag1−/−
    RAG-1–deficient
    Rel−/−
    c-Rel–deficient
    Tg
    transgenic
    WEHI
    Walter and Eliza Hall Institute
    WT
    wild-type.

  • Received September 1, 2010.
  • Accepted February 21, 2011.

References

  1. ↵
    1. Segura E.,
    2. J. A. Villadangos
    . 2009. Antigen presentation by dendritic cells in vivo. Curr. Opin. Immunol. 21: 105–110.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Heath W. R.,
    2. F. R. Carbone
    . 2009. Dendritic cell subsets in primary and secondary T cell responses at body surfaces. Nat. Immunol. 10: 1237–1244.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Naik S. H.,
    2. D. Metcalf,
    3. A. van Nieuwenhuijze,
    4. I. Wicks,
    5. L. Wu,
    6. M. O’Keeffe,
    7. K. Shortman
    . 2006. Intrasplenic steady-state dendritic cell precursors that are distinct from monocytes. Nat. Immunol. 7: 663–671.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Le Borgne M.,
    2. N. Etchart,
    3. A. Goubier,
    4. S. A. Lira,
    5. J. C. Sirard,
    6. N. van Rooijen,
    7. C. Caux,
    8. S. Aït-Yahia,
    9. A. Vicari,
    10. D. Kaiserlian,
    11. B. Dubois
    . 2006. Dendritic cells rapidly recruited into epithelial tissues via CCR6/CCL20 are responsible for CD8+ T cell crosspriming in vivo. Immunity 24: 191–201.
    OpenUrlCrossRefPubMed
  5. ↵
    1. León B.,
    2. M. López-Bravo,
    3. C. Ardavín
    . 2007. Monocyte-derived dendritic cells formed at the infection site control the induction of protective T helper 1 responses against Leishmania. Immunity 26: 519–531.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Sun S. C.,
    2. S. C. Ley
    . 2008. New insights into NF-κB regulation and function. Trends Immunol. 29: 469–478.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Ghosh S.,
    2. M. S. Hayden
    . 2008. New regulators of NF-κB in inflammation. Nat. Rev. Immunol. 8: 837–848.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Gregersen P. K.,
    2. C. I. Amos,
    3. A. T. Lee,
    4. Y. Lu,
    5. E. F. Remmers,
    6. D. L. Kastner,
    7. M. F. Seldin,
    8. L. A. Criswell,
    9. R. M. Plenge,
    10. V. M. Holers,
    11. et al
    . 2009. REL, encoding a member of the NF-κB family of transcription factors, is a newly defined risk locus for rheumatoid arthritis. Nat. Genet. 41: 820–823.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Wan F.,
    2. M. J. Lenardo
    . 2010. The nuclear signaling of NF-κB: current knowledge, new insights, and future perspectives. Cell Res. 20: 24–33.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Zhong H.,
    2. M. J. May,
    3. E. Jimi,
    4. S. Ghosh
    . 2002. The phosphorylation status of nuclear NF-κB determines its association with CBP/p300 or HDAC-1. Mol. Cell 9: 625–636.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Grundström S.,
    2. P. Anderson,
    3. P. Scheipers,
    4. A. Sundstedt
    . 2004. Bcl-3 and NF-κB p50-p50 homodimers act as transcriptional repressors in tolerant CD4+ T cells. J. Biol. Chem. 279: 8460–8468.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Fujita T.,
    2. G. P. Nolan,
    3. H. C. Liou,
    4. M. L. Scott,
    5. D. Baltimore
    . 1993. The candidate proto-oncogene bcl-3 encodes a transcriptional coactivator that activates through NF-κ B p50 homodimers. Genes Dev. 7(7B): 1354–1363.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Cao S.,
    2. X. Zhang,
    3. J. P. Edwards,
    4. D. M. Mosser
    . 2006. NF-κB1 (p50) homodimers differentially regulate pro- and anti-inflammatory cytokines in macrophages. J. Biol. Chem. 281: 26041–26050.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Campbell I. K.,
    2. S. Gerondakis,
    3. K. O’Donnell,
    4. I. P. Wicks
    . 2000. Distinct roles for the NF-κB1 (p50) and c-Rel transcription factors in inflammatory arthritis. J. Clin. Invest. 105: 1799–1806.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Sha W. C.,
    2. H.-C. Liou,
    3. E. I. Tuomanen,
    4. D. Baltimore
    . 1995. Targeted disruption of the p50 subunit of NF-κB leads to multifocal defects in immune responses. Cell 80: 321–330.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Labow M.,
    2. D. Shuster,
    3. M. Zetterstrom,
    4. P. Nunes,
    5. R. Terry,
    6. E. B. Cullinan,
    7. T. Bartfai,
    8. C. Solorzano,
    9. L. L. Moldawer,
    10. R. Chizzonite,
    11. K. W. McIntyre
    . 1997. Absence of IL-1 signaling and reduced inflammatory response in IL-1 type I receptor-deficient mice. J. Immunol. 159: 2452–2461.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Mombaerts P.,
    2. J. Iacomini,
    3. R. S. Johnson,
    4. K. Herrup,
    5. S. Tonegawa,
    6. V. E. Papaioannou
    . 1992. RAG-1-deficient mice have no mature B and T lymphocytes. Cell 68: 869–877.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Stanley E.,
    2. G. J. Lieschke,
    3. D. Grail,
    4. D. Metcalf,
    5. G. Hodgson,
    6. J. A. M. Gall,
    7. D. W. Maher,
    8. J. Cebon,
    9. V. Sinickas,
    10. A. R. Dunn
    . 1994. Granulocyte/macrophage colony-stimulating factor-deficient mice show no major perturbation of hematopoiesis but develop a characteristic pulmonary pathology. Proc. Natl. Acad. Sci. USA 91: 5592–5596.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Jung S.,
    2. D. Unutmaz,
    3. P. Wong,
    4. G.-I. Sano,
    5. K. De los Santos,
    6. T. Sparwasser,
    7. S. Wu,
    8. S. Vuthoori,
    9. K. Ko,
    10. F. Zavala,
    11. et al
    . 2002. In vivo depletion of CD11c+ dendritic cells abrogates priming of CD8+ T cells by exogenous cell-associated antigens. Immunity 17: 211–220.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Barnden M. J.,
    2. J. Allison,
    3. W. R. Heath,
    4. F. R. Carbone
    . 1998. Defective TCR expression in transgenic mice constructed using cDNA-based α- and β-chain genes under the control of heterologous regulatory elements. Immunol. Cell Biol. 76: 34–40.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Ogilvy S.,
    2. D. Metcalf,
    3. C. G. Print,
    4. M. L. Bath,
    5. A. W. Harris,
    6. J. M. Adams
    . 1999. Constitutive Bcl-2 expression throughout the hematopoietic compartment affects multiple lineages and enhances progenitor cell survival. Proc. Natl. Acad. Sci. USA 96: 14943–14948.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Cook A. D.,
    2. E. L. Braine,
    3. J. A. Hamilton
    . 2004. Stimulus-dependent requirement for granulocyte-macrophage colony-stimulating factor in inflammation. J. Immunol. 173: 4643–4651.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Grumont R. J.,
    2. S. Gerondakis
    . 1994. The subunit composition of NF-κB complexes changes during B-cell development. Cell Growth Differ. 5: 1321–1331.
    OpenUrlAbstract
  24. ↵
    1. Segura E.,
    2. A. L. Albiston,
    3. I. P. Wicks,
    4. S. Y. Chai,
    5. J. A. Villadangos
    . 2009. Different cross-presentation pathways in steady-state and inflammatory dendritic cells. Proc. Natl. Acad. Sci. USA 106: 20377–20381.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Lawlor K. E.,
    2. I. K. Campbell,
    3. K. O’Donnell,
    4. L. Wu,
    5. I. P. Wicks
    . 2001. Molecular and cellular mediators of interleukin-1-dependent acute inflammatory arthritis. Arthritis Rheum. 44: 442–450.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Nicoletti I.,
    2. G. Migliorati,
    3. M. C. Pagliacci,
    4. F. Grignani,
    5. C. Riccardi
    . 1991. A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry. J. Immunol. Methods 139: 271–279.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Vremec D.,
    2. K. Shortman
    . 1997. Dendritic cell subtypes in mouse lymphoid organs: cross-correlation of surface markers, changes with incubation, and differences among thymus, spleen, and lymph nodes. J. Immunol. 159: 565–573.
    OpenUrlAbstract
  28. ↵
    1. Shreedhar V.,
    2. A. M. Moodycliffe,
    3. S. E. Ullrich,
    4. C. Bucana,
    5. M. L. Kripke,
    6. L. Flores-Romo
    . 1999. Dendritic cells require T cells for functional maturation in vivo. Immunity 11: 625–636.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Zheng Y.,
    2. F. Ouaaz,
    3. P. Bruzzo,
    4. V. Singh,
    5. S. Gerondakis,
    6. A. A. Beg
    . 2001. NF-κB RelA (p65) is essential for TNF-α-induced fas expression but dispensable for both TCR-induced expression and activation-induced cell death. J. Immunol. 166: 4949–4957.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Yang Y. H.,
    2. J. A. Hamilton
    . 2001. Dependence of interleukin-1-induced arthritis on granulocyte-macrophage colony-stimulating factor. Arthritis Rheum. 44: 111–119.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Lawlor K. E.,
    2. P. K. K. Wong,
    3. I. K. Campbell,
    4. N. van Rooijen,
    5. I. P. Wicks
    . 2005. Acute CD4+ T lymphocyte-dependent interleukin-1-driven arthritis selectively requires interleukin-2 and interleukin-4, joint macrophages, granulocyte-macrophage colony-stimulating factor, interleukin-6, and leukemia inhibitory factor. Arthritis Rheum. 52: 3749–3754.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Egan P. J.,
    2. A. van Nieuwenhuijze,
    3. I. K. Campbell,
    4. I. P. Wicks
    . 2008. Promotion of the local differentiation of murine Th17 cells by synovial macrophages during acute inflammatory arthritis. Arthritis Rheum. 58: 3720–3729.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Das J.,
    2. C. H. Chen,
    3. L. Yang,
    4. L. Cohn,
    5. P. Ray,
    6. A. Ray
    . 2001. A critical role for NF-κ B in GATA3 expression and TH2 differentiation in allergic airway inflammation. Nat. Immunol. 2: 45–50.
    OpenUrlCrossRefPubMed
  34. ↵
    1. Hamilton J. A.
    2008. Colony-stimulating factors in inflammation and autoimmunity. Nat. Rev. Immunol. 8: 533–544.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Sapoznikov A.,
    2. J. A. Fischer,
    3. T. Zaft,
    4. R. Krauthgamer,
    5. A. Dzionek,
    6. S. Jung
    . 2007. Organ-dependent in vivo priming of naive CD4+, but not CD8+, T cells by plasmacytoid dendritic cells. J. Exp. Med. 204: 1923–1933.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Ouaaz F.,
    2. J. Arron,
    3. Y. Zheng,
    4. Y. Choi,
    5. A. A. Beg
    . 2002. Dendritic cell development and survival require distinct NF-κB subunits. Immunity 16: 257–270.
    OpenUrlCrossRefPubMed
  37. ↵
    1. O’Keeffe M.,
    2. R. J. Grumont,
    3. H. Hochrein,
    4. M. Fuchsberger,
    5. R. Gugasyan,
    6. D. Vremec,
    7. K. Shortman,
    8. S. Gerondakis
    . 2005. Distinct roles for the NF-κB1 and c-Rel transcription factors in the differentiation and survival of plasmacytoid and conventional dendritic cells activated by TLR-9 signals. Blood 106: 3457–3464.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Pope R. M.
    2002. Apoptosis as a therapeutic tool in rheumatoid arthritis. Nat. Rev. Immunol. 2: 527–535.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Gerondakis S.,
    2. A. Strasser,
    3. D. Metcalf,
    4. G. Grigoriadis,
    5. J. Y. Scheerlinck,
    6. R. J. Grumont
    . 1996. Rel-deficient T cells exhibit defects in production of interleukin 3 and granulocyte-macrophage colony-stimulating factor. Proc. Natl. Acad. Sci. USA 93: 3405–3409.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Loots G. G.,
    2. R. M. Locksley,
    3. C. M. Blankespoor,
    4. Z. E. Wang,
    5. W. Miller,
    6. E. M. Rubin,
    7. K. A. Frazer
    . 2000. Identification of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 288: 136–140.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Osborne C. S.,
    2. M. A. Vadas,
    3. P. N. Cockerill
    . 1995. Transcriptional regulation of mouse granulocyte-macrophage colony-stimulating factor/IL-3 locus. J. Immunol. 155: 226–235.
    OpenUrlAbstract
  42. ↵
    1. Holloway A. F.,
    2. S. Rao,
    3. X. Chen,
    4. M. F. Shannon
    . 2003. Changes in chromatin accessibility across the GM-CSF promoter upon T cell activation are dependent on nuclear factor κB proteins. J. Exp. Med. 197: 413–423.
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Fraser J. D.,
    2. A. Weiss
    . 1992. Regulation of T-cell lymphokine gene transcription by the accessory molecule CD28. Mol. Cell. Biol. 12: 4357–4363.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Himes S. R.,
    2. L. S. Coles,
    3. R. Reeves,
    4. M. F. Shannon
    . 1996. High mobility group protein I(Y) is required for function and for c-Rel binding to CD28 response elements within the GM-CSF and IL-2 promoters. Immunity 5: 479–489.
    OpenUrlCrossRefPubMed
  45. ↵
    1. Schreck R.,
    2. P. A. Baeuerle
    . 1990. NF-κB as inducible transcriptional activator of the granulocyte-macrophage colony-stimulating factor gene. Mol. Cell. Biol. 10: 1281–1286.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    1. Ponomarev E. D.,
    2. L. P. Shriver,
    3. K. Maresz,
    4. J. Pedras-Vasconcelos,
    5. D. Verthelyi,
    6. B. N. Dittel
    . 2007. GM-CSF production by autoreactive T cells is required for the activation of microglial cells and the onset of experimental autoimmune encephalomyelitis. J. Immunol. 178: 39–48.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    1. Aggarwal S.,
    2. N. Ghilardi,
    3. M. H. Xie,
    4. F. J. de Sauvage,
    5. A. L. Gurney
    . 2003. Interleukin-23 promotes a distinct CD4 T cell activation state characterized by the production of interleukin-17. J. Biol. Chem. 278: 1910–1914.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Wada H.,
    2. Y. Noguchi,
    3. M. W. Marino,
    4. A. R. Dunn,
    5. L. J. Old
    . 1997. T cell functions in granulocyte/macrophage colony-stimulating factor deficient mice. Proc. Natl. Acad. Sci. USA 94: 12557–12561.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    1. Sonderegger I.,
    2. G. Iezzi,
    3. R. Maier,
    4. N. Schmitz,
    5. M. Kurrer,
    6. M. Kopf
    . 2008. GM-CSF mediates autoimmunity by enhancing IL-6-dependent Th17 cell development and survival. J. Exp. Med. 205: 2281–2294.
    OpenUrlAbstract/FREE Full Text
  50. ↵
    1. Vremec D.,
    2. G. J. Lieschke,
    3. A. R. Dunn,
    4. L. Robb,
    5. D. Metcalf,
    6. K. Shortman
    . 1997. The influence of granulocyte/macrophage colony-stimulating factor on dendritic cell levels in mouse lymphoid organs. Eur. J. Immunol. 27: 40–44.
    OpenUrlCrossRefPubMed
  51. ↵
    1. King I. L.,
    2. M. A. Kroenke,
    3. B. M. Segal
    . 2010. GM-CSF-dependent, CD103+ dermal dendritic cells play a critical role in Th effector cell differentiation after subcutaneous immunization. J. Exp. Med. 207: 953–961.
    OpenUrlAbstract/FREE Full Text
  52. ↵
    1. Min L.,
    2. S. A. B. Mohammad Isa,
    3. W. Shuai,
    4. C. B. Piang,
    5. F. W. Nih,
    6. M. Kotaka,
    7. C. Ruedl
    . 2010. Cutting edge: granulocyte-macrophage colony-stimulating factor is the major CD8+ T cell-derived licensing factor for dendritic cell activation. J. Immunol. 184: 4625–4629.
    OpenUrlAbstract/FREE Full Text
  53. ↵
    1. Auffray C.,
    2. M. H. Sieweke,
    3. F. Geissmann
    . 2009. Blood monocytes: development, heterogeneity, and relationship with dendritic cells. Annu. Rev. Immunol. 27: 669–692.
    OpenUrlCrossRefPubMed
  54. ↵
    1. Cheong C.,
    2. I. Matos,
    3. J. H. Choi,
    4. D. B. Dandamudi,
    5. E. Shrestha,
    6. M. P. Longhi,
    7. K. L. Jeffrey,
    8. R. M. Anthony,
    9. C. Kluger,
    10. G. Nchinda,
    11. et al
    . 2010. Microbial stimulation fully differentiates monocytes to DC-SIGN/CD209(+) dendritic cells for immune T cell areas. Cell 143: 416–429.
    OpenUrlCrossRefPubMed
  55. ↵
    1. Seetharaman R.,
    2. A. L. Mora,
    3. G. Nabozny,
    4. M. Boothby,
    5. J. Chen
    . 1999. Essential role of T cell NF-κB activation in collagen-induced arthritis. J. Immunol. 163: 1577–1583.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section