Absence of the Transcriptional Repressor Blimp-1 in Hematopoietic Lineages Reveals Its Role in Dendritic Cell Homeostatic Development and Function

Mice and cell lines

Prdm1^f/f mice, generated as described (18), were bred with Tie2-Cre mice, B6, Tg(Tek-cre)12Flv/J (The Jackson Laboratory). Sex-matched Prdm1^f/f Tie2-Cre⁺ (conditional knockout, CKO) and littermate control (Ctrl) mice (2- to 4-mo-old) were used for all experiments in this study. All mice were maintained under specific pathogen free conditions in the animal facility of the Institute of Cellular and Organismic Biology at Academia Sinica. The Institute Animal Care and Utilization Committee of Academia Sinica approved all experiments. Mice were genotyped by genomic PCR using primers described in supplemental Table I.⁴ C57BL/6 and BALB/c mice were purchased from the Laboratory Animal Center in Taiwan and used at 6∼10 wk of age. For BrdU labeling in vivo, 1 μg BrdU (BD Biosciences) was injected into Ctrl or CKO mice by i.p. at 1 day before sacrifice for flow cytometric analysis of the percentage of BrdU-positive cells in splenic cDC subsets. JAWS II cells purchased from American Type Culture Collection were grown in MEMα medium (Invitrogen) supplemented with 20% FBS (Invitrogen), 100 U/ml penicillin, and 100 μg/ml streptomycin, 4 mM l-glutamine, 1 mM sodium pyruvate, and 5 ng/ml murine GM-CSF (eBioscience).

Preparation and culture of cells from lymphoid organs

For analysis of cell populations, lymphoid organs were treated with 0.02 mg/ml Liberase Blendzyme II (Roche Applied Science) for 20 min at room temperature. Single cell suspensions were washed with RPMI 1640 (Invitrogen) containing 50 μg/ml DNase I (Roche Applied Science) and 3% FCS (HyClone). RBCs were removed using RBC lysis buffer (Sigma-Aldrich). Splenic CD11c⁺ cells were isolated by positive selection using an autoMACS separator (Miltenyi Biotec) and stimulated with 1 μg/ml LPS (from Escherichia coli O26:B6, Sigma-Aldrich) or 100 μg/ml poly I:C (Amersham Biosciences) for 1 day in some experiments. Splenic B cells were cultured as described (22) and were stimulated with 2 μg/ml LPS for indicated days. Splenic CD4⁺ T cells were isolated by flow cytometry from total splenocytes cultured with 5 μg/ml anti-CD3 (eBioscience) and 10 ng/ml IL-2 (R&D Systems) for 3 days. The immediate splenic DC precursors (pre-cDCs) were defined as previously described (9). In brief, splenocytes were separated by density using Nycodenz (Axis-Shield), and cells of density between 1.076 and 1.084 were collected. B, T, and NK cells were excluded by staining with specific Abs to CD3, CD19, and CD49b. The lymphocyte-excluded medium-density fraction was further analyzed by the following parameters: low side scatter and expression of CD11c, CD45RA, CD43, SIRP-α, and CD24.

Bone marrow-derived DC (BMDC) cultures

GM-CSF culture of BMDCs was performed essentially as described (25). Cells (2 × 10⁵ cells/ml) were cultured in RPMI 1640 supplemented with 20 ng/ml murine GM-CSF (eBioscience), 10% FBS (BenchMark), 50 μM 2-ME, 100 U/ml penicillin, and 100 μg/ml streptomycin. After 8 days in culture, nonadherent cells were harvested as immature BMDCs. The remaining cells were replated at a density of 1 × 10⁶ cells/ml for stimulation with 10 ng/ml (or other indicated doses) LPS, 0.5 μg/ml poly I:C, 50 ng/ml CpG oligodeoxynucleotides 1826 (InvivoGen), or 10 ng/ml TNF-α (PeproTech) for 20 h. In some experiments, immature BMDCs harvested on day 8 were treated for 1 h with the NF-κB inhibitor BAY 11–0782 (Calbiochem), or the p38 MAPK inhibitor SB203580 (Calbiochem), followed by LPS stimulation for 20 h. Flt3L-cultured BMDCs were prepared as described (26). Cells were seeded at 1 × 10⁶ cells/ml and cultured in RPMI 1640, supplemented with 50 ng/ml murine Flt3L (R & D Systems), 10% FBS (BenchMark), 50 μM 2-ME, 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells were left untreated or stimulated with 0.5 μg/ml poly I:C or 0.3 μM CpG D19 (GGTGCATCGATGCAGGGGGG phosphorothioate-modified at sites underlined) on day 8 and harvested on day 9.

Abs and flow cytometry

Single cell suspensions were stained with appropriate Abs as described in supplemental Table II. For intracellular cytokine staining, T cells were washed and stimulated with 50 ng/ml PMA and 1 μg/ml ionomycin (Sigma-Aldrich) in the presence of brefeldin A (Biolegend) for 6 h. After fixation and permeabilization, cells were stained with anti-IFN-γ-APC (Biolegend) and anti-IL-4-PE (Biolegend). All flow cytometric analysis was performed using FACSCanto or FACSCalibur and results were analyzed by FACS Diva software (BD Biosciences) or Flow Jo software (Tree Star). In some experiments, CD11c^highCD8⁺, CD11c^highCD4⁺CD8⁻, or CD11c^highCD4⁻CD8⁻cells from purified splenic CD11c⁺ cells were sorted by FACSAria (BD Biosciences).

Western blotting and immunofluorescence staining

Western blotting with an anti-Blimp-1 mAb was performed as described (27). For detection of cytoplasmic IκBα degradation by rabbit anti-IκBα Ab (Santa Cruz Biotechnology) or nuclear translocation of p50 subunit by rabbit anti-p50 Ab (Santa Cruz Biotechnology), cytoplasmic and nuclear extracts were prepared as described (22). Goat anti-lamin B Ab was purchased from Santa Cruz Biotechnology. For NF-κB p65 subunit immunofluorescence staining, in brief, BMDCs were left unstimulated or stimulated with 100 ng/ml LPS for 2 h and prepared by cytospin at 500 rpm for 5 min. The cells were then fixed by 4% paraformaldehyde, permeabilized with 0.1% triton X-100, and blocked with 1% BSA, followed by staining with rabbit anti-p65 Ab (Santa Cruz Biotechnology) and Alexa Fluor 555-conjugated goat anti-rabbit IgG (Molecular Probes). The nuclei were counterstained by 4′-6-diamidino-2-phenylindole (Sigma-Aldrich). The fluorescent images were examined under a fluorescence microscope (Leica DM6000B).

Mixed lymphocyte reaction

Allogeneic T cell proliferation was measured using 10⁵ splenic CD3⁺CD4⁺ T cells from BALB/c as responders and various numbers of BMDCs (1000, 3000, or 10000 cells) irradiated with 30 Gy as stimulators. After 4 days in culture, cells were pulsed with 1 μCi [³H]thymidine for 16 h. [³H]thymidine incorporation was quantified with a microplate-scintillation counter (TopCount NXT, Packard). Cell proliferation was measured in cpm.

In vivo DC maturation

Ctrl and CKO mice were subjected to i.p. injection of LPS (3 μg/g body weight) solubilized in PBS. Control groups were injected with PBS alone. Ten hours after injection, mice were sacrificed. Single cell suspensions were prepared from mesenteric lymph nodes, stained, and subjected to flow cytometric analysis.

Lentiviral vector preparation and transduction

pFUGW, pFUW-GFPBlimp1, pFUGWCtrli, and pFUGWBlimp1-3952i have been described (27, 28). Generation of pseudotyped lentivirus was performed essentially as described (27, 29). Targeted BMDCs and JAWS II cells were infected at a multiplicity of infection of 5 in the presence of 8 μg/ml polybrene (Sigma-Aldrich).

Cytometric bead array (CBA), multiplex bead array assays, and ELISA

The Mouse Inflammation CBA Kit (BD Pharmingen) was used in accordance with the manufacturer’s protocol to measure select cytokines and chemokines (IL-6, IL-10, MCP-1, IFN-γ, TNF-α, and IL-12p70) in supernatants from BMDC cultures stimulated with 1 μg/ml LPS. Cytokines and chemokines in sera were determined by the MILLIPLEX Mouse Cytokine/Chemokine kit (Millipore) according to the manufacturer’s protocol and analyzed by Luminex 200. IgM production in culture supernatants was determined by ELISA essentially as described (22).

RNA isolation and reverse transcription-quantitative PCR (RT-QPCR)

These procedures have been described (22). The gene specific Taqman primer and probe sets, including Prdm1 (assay ID: Mm 00476128_ml) and L32 (used for normalization, assay ID: Mm 00777741_sH) were purchased from Applied Biosystems. The specific primer sets information was described in supplemental Table I.

Chromatin immunoprecipitation (ChIP) assay

Blimp-1 binding in vivo was detected essentially as reported (28, 30) using rabbit antiserum against Blimp-1 as described (31). In brief, 2∼3 × 10⁷ Ctrl or CKO BMDCs stimulated with LPS for 24 h were used per ChIP experiment. The immunoprecipitated DNA was analyzed by real-time PCR with SYBR green method on an ABI prism 7300 system. The primer sequences for ChIP assay were listed in the supplemental Table I.

DNA pull-down assay and EMSA

The DNA pull-down assay was performed essentially as described (32). In brief, biotinylated wild-type or mutated oligonucleotides coupled with streptavidin-conjugated Dynabeads (Invitrogen) were incubated with 150 μg precleared nuclear extract prepared from LPS-stimulated Ctrl and CKO BMDCs. After 1 h incubation at room temperature, the complexes were collected with a magnetic apparatus. Coprecipitated proteins were eluted and fractionated by 8% SDS-PAGE, followed by Western blotting analysis. In the competition assay, a 200-fold excess of unlabeled oligonucleotides was added. EMSA was performed using 2.5 μg nuclear extract prepared from 293T cells transfected with Blimp-1 expression vector and the LightShift chemiluminescent EMSA kit (Pierce) as described (22). For the supershift assay, 1 μg of mouse anti-Blimp-1 Ab or control Ab (Sigma-Aldrich) was added to the described reactions and incubated on ice for 30 min. The oligonucleotides used in this study were described in supplemental Table III.

Luciferase assay

Genomic fragments from +69 to +223 of murine Il-6, located in intron 1, and from −600 to +64 of murine Ccl2 were amplified by PCR from genomic DNA of primary mouse BMDCs derived from C57BL/6 mice and cloned into SmaI/BglII-digested pGL3 promoter vector (Promega) and KpnI/BglII-digested pGL3 basic vector (Promega), respectively. Detailed primer information for luciferase reporter constructions was described in supplemental Table I. JAWS II cells were cotransfected with 0.2∼1 μg luciferase reporter construct as mentioned above and various amounts (for Il-6 reporter assays: 50, 150, and 450 ng; for Ccl2 reporter assays: 30, 100 and 300 ng) of Blimp-1 expression vector (pCMV-Blimp1) or control vector (pCMV-5b) using TransIT-TKO reagent (Mirus) (33). Luciferase activity measurement and fold repression calculation was performed as previously described (30).

Statistical analysis

Statistical significance was measured using the two-tailed paired Student’s t test. Data represent means ± SD or means ± SEM as indicated.

Blimp-1 negatively regulates CD8⁻ cDC development

Prdm1^f/f mice harboring Prdm1 alleles containing two LoxP sites flanking exons 5 and 8 were generated previously (18) and have been used to demonstrate the role of Blimp-1 in B cell differentiation and T cell homeostasis in vivo (18, 19). To further understand the role of Blimp-1 in the immune system, such as in DC lineage development and function, we deleted Prdm1 in the hematopoietic system by crossing Prdm1^f/f mice with mice expressing the cre-recombinase (Cre) transgene under the control of the Tie2 regulatory element, Tie2-Cre (34). A three primer-based PCR assay using genomic DNA from various tissues isolated from either CKO mice (Prdm1^f/fTie2-Cre⁺) or littermate control (Ctrl) mice (Prdm1^f/fTie2-Cre⁻) demonstrated that the Prdm1 allele was efficiently deleted in CKO lymph nodes, spleen, thymus, and bone marrow (Fig. 1⇓A). Only a small proportion of Prdm1 was deleted in the brain, which is likely explained by endothelial and hematopoietic cells that are present in the brain. Prdm1 alleles were almost completely deleted in purified CD3⁺ T cells, B220⁺ B cells, and CD11c⁺ DCs isolated from CKO spleen (Fig. 1⇓A).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

The generation of CKO mice. A, PCR analysis of genomic DNA isolated from tissues (peripheral lymph node (LN), spleen (Sp), thymus (Thy), bone marrow (BM), and brain), from cells (splenic CD3⁺ T cells (T), B220⁺ B cells (B), and CD11c⁺ DCs (DC)) of Prdm1^f/fTie2-Cre⁻ and Prdm1^f/fTie2-Cre⁺ mice. B, Western blotting of Blimp-1 in Ctrl and CKO B cells treated with LPS for 3 days. α-tubulin was as a loading control. C, Flow cytometric analysis of plasma cells generated from Ctrl and CKO splenic B cells stimulated with LPS for 3 days. One representative experiment of three is shown. D, IgM levels in supernatants harvested from LPS-treated Ctrl and CKO B cells on days 3 and 5. Data represent means ± SD for one representative experiment of three. E, Flow cytometric analysis of splenic CD4⁺ T cells from Ctrl and CKO mice. Numbers in the quadrants indicate the percentage of naive (CD62L^highCD44^low) and effector (CD62L^lowCD44^high) T cells. F, Total splenocytes from Ctrl and CKO mice were cultured with PMA plus ionomycin in the presence of brefeldin A for 6 h for intracellular cytokine staining and subsequent flow cytometric analysis. Numbers in the quadrants indicate the percentage of IFN-γ- and IL-4-producing cells.

Splenic B cells from Prdm1^f/fCD19-Cre⁺ mice fail to differentiate when stimulated with mitogen-like LPS in vitro (18). Induction of Blimp-1 protein was largely abolished after LPS stimulation of CKO splenic B cells (Fig. 1⇑B). As anticipated, induction of the plasma cell surface marker CD138 and IgM production were defective when compared with Ctrl B cells (Fig. 1⇑, C and D), consistent with the notion that in vitro plasma cell differentiation is impaired in B cells lacking Blimp-1. The frequency of CD44^highCD62L^low effector T cells was increased in CKO spleen, and we isolated increased numbers of stimulated CKO splenic IFN-γ-producing T cells ex vivo (Fig. 1⇑, E and F), consistent with previous reports (19, 20).

We next examined the percentages and the total cell counts of various immune cell subsets in secondary lymphpoid organs. Fig. 2⇓A shows that the percentages of splenic CD3⁺ T cells, B220⁺ B cells, and F4/80^high macrophages were similar in Ctrl and CKO mice. Surprisingly, greater frequency of CD11c^high cells was found in CKO mice (Fig. 2⇓B, left) compared with Ctrl mice. The number of DCs found in CKO mice was even greater when we counted the total CD11c^high splenic CDCs (Fig. 2⇓B, middle). The CD11c^high populations in the peripheral lymph nodes of CKO mice were similarly augmented (Fig. 2⇓B, right). Thus, increased DC populations in CKO mice suggest a role for Blimp-1 in DC development.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Blimp-1 negatively regulates CD8⁻ DC homeostasis. A, Percentage of peripheral lymph node CD3⁺ T cells (n = 9), B220⁺ B cells (n = 9), and F4/80^high macrophages (n = 10) in Ctrl and CKO mice. B, Percentage (left) and total number (middle) of splenic CD11c^high cDCs (n = 12), and percentage of peripheral lymph node CD11c^high cDCs (n = 8, right) in Ctrl and CKO mice. C, Percentage of CD8⁺ and CD8⁻ cDCs in the CD11c^high gate in Ctrl and CKO spleens (n = 12). D, Total number of splenic CD11c^highCD8⁺ and CD11c^highCD8⁻ DCs in Ctrl and CKO mice (n = 12). E, Percentage of splenic CD11c^highCD4⁺CD8⁻ and CD11c^highCD4⁻CD8⁻ cells in the CD11c^highCD8⁻ gate from Ctrl and CKO mice (n = 12). F, Total number of splenic CD11c^highCD4⁺CD8⁻ and CD11c^highCD4⁻CD8⁻ cells in Ctrl and CKO mice (n = 12). Data represent means ± SEM. ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.005; ∗∗∗∗, p < 0.001 vs Ctrl.

Next, we found that the percentage of CD8⁻ cDCs was greater in CKO spleen than in Ctrl spleen (Fig. 2⇑C). Correspondingly, the percentage of CD8⁺ cDCs was less in CKO spleen. However, owing to the increased number of total cDCs in CKO mice, the total number of CD8⁺ cDCs in CKO spleen remained similar to that in Ctrl spleen, whereas the total number of CD8⁻ cDCs was greatly increased in CKO mice (Fig. 2⇑D). Among the CD8⁻ cDCs, the distribution and total number of CD4⁺CD8⁻ cDCs were significantly increased in CKO mice compared with Ctrl mice (Fig. 2⇑, E and F). Despite a decrease in the relative percentage of CD4⁻CD8⁻ cDCs in CKO spleen (Fig. 2⇑E), the total number of CD4⁻CD8⁻ cDCs was increased (Fig. 2⇑F). Therefore, both the CD4⁺ and CD4⁻ subpopulations of the CD8⁻ cDC population were expanded in CKO mice.

The expanded steady-state total CD8⁻ cDC population in CKO mice could result from increased numbers of bone marrow DC progenitors, changes in the frequency of immediate splenic precursors of the DC subset, and/or changes in cell proliferation/survival. We examined the distribution of DC precursors (defined as Lin⁻c-kit⁺Flt3⁺) (6) and the more restricted DC progenitor subpopulation (defined as Lin⁻c-kit^intFlt3⁺IL-7Rα⁻M-CSFR⁺) (7, 8) in Ctrl and CKO bone marrow cells (Fig. 3⇓A). The relative percentage of the Lin⁻c-kit⁺Flt3⁺ and Lin⁻c-kit^intFlt3⁺IL-7Rα⁻M-CSFR⁺ populations was similar in Ctrl and CKO bone marrow (Fig. 3⇓B), suggesting that the increase in total DC population in CKO mice may not be a consequence of increased generation of DC progenitors from bone marrow hematopoietic stem cells. We next examined splenic pre-cDCs, which are characterized as CD11c⁺CD45RA^lowCD43^intSIRP-α^int (9). There was no difference in the frequency of splenic pre-cDCs between Ctrl and CKO mice (Fig. 3⇓C). The immediate splenic CD8⁻ cDC precursors were examined by flow cytometric analysis of CD24^low/int cells from pre-cDCs (9, 35). The relative percentage of CD24^low/int cells was significantly increased in CKO mice (Fig. 3⇓, C and D), indicating that more CD8⁻cDC immediate precursors were present in CKO spleen than Ctrl spleen. We also found higher steady-state Blimp-1 mRNA levels in CD8⁻ cDCs, including CD4⁺CD8⁻and CD4⁻CD8⁻ cDCs (Fig. 3⇓E), consistent with these findings that CD8⁻ splenic cDCs were preferentially affected in CKO mice. In fact, CD4⁺CD8⁻and CD4⁻CD8⁻ cDCs have amounts of steady-state Blimp-1 mRNA comparable to those in CD44^highCD62L^low effector CD4⁺ cells and B220⁺ B cells (Fig. 3⇓E). The levels of Blimp-1 mRNA in the induced state were also assessed. Like stimulated splenic CD4⁺ T cells and B cells, Blimp-1 mRNA levels were induced by LPS in CD11c^high splenic cDCs (Fig. 3⇓F).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Increased percentage of CD8⁻ cDC immediate precursors in CKO mice. A, Analysis of bone marrow DC progenitors in Ctrl and CKO mice. Single cell suspensions of bone marrow cells were stained and analyzed by flow cytometry. The distribution of Lin⁻c-kit⁺Flt3⁺ progenitors (gate 1 plus gate 2, left contour plots) or Lin⁻c-kit^intFlt3⁺IL-7Rα⁻M-CSFR⁺ distinct DC progenitors (gate 3, right contour plots) are indicated in successive gates. One representative experiment of five is shown. B, Percentage of Lin⁻c-kit⁺Flt3⁺ (left, n = 6) in the Lin⁻ gate and Lin⁻c-kit^intFlt3⁺IL-7Rα⁻M-CSFR⁺ (right, n = 5) progenitors in the Lin⁻c-kit^intFlt3⁺ gate in Ctrl and CKO bone marrow cells. C, Distribution of splenic CD8⁻ DC progenitors in Ctrl and CKO mice. CD11c^intCD45RA^lowCD43^intSIRP-α^intCD24^low/int cells corresponding to CD8⁻ DC progenitors were analyzed by successive gates as indicated. D, The percentage of CD24^low/int cells that correspond to CD8⁻ DC progenitors in Ctrl and CKO mice (n = 3). E, Relative levels of Blimp-1 transcripts in various immune cell types. Purified C57BL/6 splenic B220⁺ B cells, naive (CD62L^highCD44^low) CD4⁺ T cells, effector (CD62L^lowCD44^high) T cells, and various indicated CD11c^high subsets were subjected to RT-QPCR analysis of Blimp-1 mRNA level. F, Relative levels of Blimp-1 transcripts in various immune cell types following stimulation. RT-QPCR was used to determine Blimp-1 mRNA levels in purified splenic B cells ± LPS stimulation for 3 days, CD11c^high DCs ± LPS stimulation for 1 day, and CD4⁺ T cells ± anti-CD3+IL-2 stimulation for 3 days. Results in E and F were calculated relative to nonstimulated B cells. Data represent means ± SEM. ∗, p < 0.05 vs Ctrl.

Loss of IRF-2, IRF-4, or IRF-8 alters CD8⁻ or CD8⁺ cDC distribution in mice (10, 11, 12). However, IRF-2, IRF-4, and IRF-8 mRNA levels in all splenic cDC subsets were similar in Ctrl and CKO mice (supplemental Fig. 1A). c-myc, which plays a role in cell proliferation, is a direct target of Blimp-1 in several cell lineages (28, 36). c-myc mRNA levels in each cell subpopulation were similar in Ctrl and CKO mice (supplemental Fig. 1B). Further investigation of proliferative CD11c^high splenic DC subsets in vivo revealed that Ctrl and CKO mice had similar frequencies of BrdU incorporation within CD8⁺ or CD8⁻ splenic DC subsets (supplemental Fig. 2A). Annexin V staining revealed similar survival rates for CD8⁺ and CD8⁻ splenic cDC subpopulations freshly isolated from Ctrl and CKO mice (supplemental Fig. 2B). Collectively, these results suggest that the increased number of steady-state CD8⁻ cDCs in CKO spleen likely results from increased generation of immediate CD8⁻ cDC precursors.

Stimulus-induced Blimp-1 expression during DC maturation is mediated by NF-κB and p38 MAPK pathways

DCs can be generated by addition of GM-CSF in the bone marrow progenitors in culture (37). We next measured the expression of Blimp-1 in bone marrow cultures supplemented with GM-CSF and evaluated whether its expression changed following stimulus-induced maturation. The Blimp-1 mRNA levels in sorted CD11c⁺ cells obtained from BMDCs were almost equivalent in cells cultured with GM-CSF for 4 and 8 days (Fig. 4⇓A), whereas Blimp-1 was greatly induced in BMDCs cultured with LPS for 24 h (Fig. 4⇓A). Similarly, maturation of GM-CSF-cultured BMDCs by TNF-α, CpG, or poly I:C induced Blimp-1 mRNA and protein expression (Fig. 4⇓, B and C). Using pharmacological inhibitors, we examined the signaling pathway(s) required for Blimp-1 induction in BMDCs. Induction of Blimp-1 by LPS was inhibited in a dose-dependent manner by the NF-κB inhibitor BAY 11-0782, and by the p38 MAPK inhibitor SB203580 (Fig. 4⇓D), but the JNK inhibitor SP600125, the ERK inhibitor PD98059, and the PI3K inhibitor LY294002 had no effect (data not shown). Thus, during BMDC maturation, Blimp-1 is induced by TLR ligands through the p38 MAPK and NF-κB pathways.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Blimp-1 expression is increased upon maturation mediated by NF-κB and p38 MAPK pathways. A, Relative Blimp-1 mRNA levels in CD11c⁺ cells sorted by flow cytometry from BMDCs cultured with GM-CSF for 4 or 8 days. Cells sorted after 8 days in culture were further stimulated with LPS for 1 day and subjected to RT-QPCR analysis of Blimp-1 mRNA. B, RT-QPCR analysis of Blimp-1 mRNA in immature (im) BMDCs and LPS-, TNF-α-, CpG-, or poly I:C-stimulated BMDCs cultured in the presence of GM-CSF. Data represent means ± SD of three experiments. C, Western blotting of Blimp-1 in immature BMDCs and LPS-, TNF-α-, CpG-, or poly I:C-stimulated BMDCs cultured with GM-CSF. D, After 8 days in culture with GM-CSF, BMDCs were treated with a NF-κB inhibitor, BAY 11-0782, or p38 MAPK inhibitor, SB203580, for 1 h, followed by LPS for 20 h. Cell lysates were analyzed for Blimp-1 expression by Western blotting.

Blimp-1 is required during DC maturation

We next examined whether BMDCs cultured with GM-CSF required Blimp-1. We verified impaired Blimp-1 induction by LPS treatment of CKO BMDCs cultured with GM-CSF (Fig. 5⇓A). The maturation status of DCs, which was analyzed by the expression of a panel of surface markers including CD25, CD40, CD86, and MHCII(I-A/E), did not differ significantly between Ctrl and CKO GM-CSF-cultured BMDCs after 8 days of culture (Fig. 5⇓B), suggesting that Blimp-1 is not essential for DC development under these conditions. Following stimulation with a variety of TLR ligands or by TNF-α for 24 h, induction of CD25, CD40, CD86, and MHCII was partially impaired in CKO GM-CSF-cultured BMDCs (Fig. 5⇓B). Likewise, up-regulation of CD86 and MHCII in BMDCs cultured with GM-CSF derived from C57BL/6 mice was blocked by lentiviral vectors expressing short hairpin RNAs against Blimp-1 (Fig. 5⇓C), suggesting that impaired maturation of CKO BMDCs is a specific effect. We excluded the possibility that defective BMDC maturation in CKO mice was caused by changes in the levels of various TLR mRNAs (supplemental Fig. 3A), the kinetics of IκBα degradation, or nuclear translocation of the NF-κB subunits p50 and p65 that is mediated by TLR4 receptor signaling (supplemental Fig. 3, B and C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Impaired DC maturation in CKO mice. A, Western blotting of Blimp-1 in GM-CSF-cultured immature (im) and LPS-stimulated BMDCs from Ctrl and CKO mice. B, Cell surface markers of BMDCs cultured with GM-CSF. Ctrl (shaded histogram) or CKO (open histogram) CD11c⁺ BMDCs and LPS-, poly I:C-, CpG-, or TNF-α-stimulated BMDCs were analyzed for surface marker expression by flow cytometry. One representative experiment of at least three is shown. C, Blimp-1 knockdown BMDCs are less mature. GM-CSF-cultured BMDCs were transduced with lentiviral vectors expressing short hairpin RNAs for Blimp-1 (Blimp1-3952i) or a control shRNA (Ctrli) after 8 days in culture. Surface expression of CD86 and I-A/E was examined in GFP⁺ cells 4 days after transduction. D, Allogeneic MLR elicited by stimulation of Ctrl or CKO BMDCs, as indicated. Data represent means ± SEM of three independent experiments. E, CKO DCs are refractory to maturation in vivo. The expression of indicated surface markers was analyzed from the CD11c^high gate of mesenteric lymph node cells. Results represent the mean fluorescence intensity (MFI) ± SEM for each marker (n = 5 in each group). ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.005 vs Ctrl.

Using a MLR assay, we demonstrated that LPS, poly I:C, CpG, and TNF-α stimulation of BMDCs from CKO mice resulted in a partially reduced ability to induce allogeneic T cell proliferation, when compared with Ctrl BMDCs (Fig. 5⇑D). The significance of Blimp-1 in DC maturation was further strengthened by in vivo experiments. Ten hours after LPS injection of Ctrl mice, CD40, CD54 and CD86 were up-regulated in mesenteric lymph node DCs (Fig. 5⇑E). By contrast, CKO DCs were resistant to up-regulation of several tested maturation markers following in vivo LPS injection (Fig. 5⇑E). Interestingly, it has been shown that mesenteric lymph node DCs which generally have higher levels of MHCII are more mature than splenic DCs (38, 39). This is in agreement with our data that lower basal MHCII levels were found in CKO mesenteric lymph node DCs that are less mature (Fig. 5⇑E).

Il-6 and Ccl2 were suppressed directly by Blimp-1

We also measured cytokine/chemokine production to assay the function of GM-CSF-cultured BMDCs from Ctrl and CKO mice (Fig. 6⇓A). After LPS stimulation for 1 day, CKO cultures produced more IL-6 and MCP-1 (encoded by Ccl2) than Ctrl cultures, as determined by CBA (Fig. 6⇓A), and this increase occurred at the transcription level (supplemental Fig. 4A). IL-6 and MCP-1 mRNA levels were also increased in LPS- and poly I:C-stimulated splenic CD11c⁺ DCs cultured from CKO mice (supplemental Fig. 4B). We thus hypothesized that Il-6 and Ccl2 expressed in the DC lineage are subject to Blimp-1 regulation. Il-6 contains three potential Blimp-1 binding sites: one in intron 1 and two in intron 3. The Ccl2 promoter region contains one potential Blimp-1 binding site, as revealed based on known Blimp-1 consensus sites (31) (Fig. 6⇓B). A ChIP assay using a Blimp-1-specific Ab showed that endogenous Blimp-1 bound to the Il-6 intron 1 and Ccl2 promoter sites in Ctrl BMDCs stimulated with LPS, and that Blimp-1 did not bind significantly to either site in LPS-stimulated CKO BMDCs (Fig. 6⇓C). In support of this finding, DNA pull-down assay demonstrated that Blimp-1 in nuclear extracts harvested from LPS-stimulated Ctrl BMDCs could bind to biotinylated oligonucleotides containing wild-type but not mutated Il-6 or Ccl2 sites (Fig. 6⇓D), and that competition assays using unlabeled oligonucleotides containing a known Blimp-1 binding site in the c-myc promoter (36) largely reduced the binding (Fig. 6⇓D). In contrast, nuclear proteins isolated from LPS-stimulated CKO BMDCs bound much less to wild-type Il-6 or Ccl2 biotinylated oligonucleotides (Fig. 6⇓D). Accordingly, EMSA showed that oligonucleotides containing these sites were shifted by nuclear protein extracts containing exogenously expressed Blimp-1 (supplemental Fig. 5, A and B). The shifted complex was disrupted by addition of excess unlabeled probe corresponding to the c-myc site (supplemental Fig. 5, A and B) and was supershifted by a Blimp-1-specific mAb (supplemental Fig. 5, A and B). By contrast, mutated oligonucleotides or control isotype Ab did not affect the mobility of the shifted complex (supplemental Fig. 5, A and B). Furthermore, ectopic Blimp-1 expression was sufficient to suppress the expression of endogenous IL-6 and MCP-1 in JAWS II cells, a murine immature DC line (33) (Fig. 6⇓E). Finally, we used a construct consisting of a minimal basal promoter fused with an intron 1 fragment of Il-6 containing the Blimp-1 binding site and the luciferase gene, and a similar construct containing the Ccl2 promoter fragment. We observed dose-dependent Blimp-1 suppression of luciferase expression with both constructs (Fig. 6⇓F). Taken together, these data indicate that Il-6 and Ccl2 are direct targets for Blimp-1 and that Blimp-1 negatively regulates these genes in DCs.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Blimp-1 regulates Il-6 and Ccl2 in DCs. A, CBA analysis of inflammatory cytokines in the supernatants of GM-CSF-cultured Ctrl or CKO BMDCs stimulated with LPS. One representative experiment of three is shown. Changes in the levels of IL-6 and MCP-1 were quantified in three independent experiments. Data are means ± SEM. B, Shown are the putative Blimp-1 binding sites in Il-6, between 3 kb upstream and 3 kb downstream of the transcription start site; also shown is the putative Blimp-1 binding site in the Ccl2 promoter. C, Blimp-1 binds to Il-6 and Ccl2 sites in vivo. Ctrl or CKO BMDCs were stimulated with LPS for 24 h, followed by ChIP analysis using primer sets designed to amplify the indicated gene loci. Snail1 was used as a negative control locus for Blimp-1 binding, as described (48 ). D, Western blotting analysis of Blimp-1 captured in DNA pull-down assays in which biotinylated probes coupled to streptavidin-coated magnetic beads were incubated with nuclear extracts from LPS-stimulated Ctrl and CKO BMDCs. E, Ectopic expression of Blimp-1 by pFUW-GFPBlimp-1 suppresses levels of IL-6 and MCP-1 transcripts as determined by RT-QPCR in JAWS II cells. F, Blimp-1 suppresses luciferase activity driven by an Il-6 intron 1 fragment containing the Blimp-1 binding site and a minimal promoter, as well as luciferase activity driven by a promoter fragment containing the Blimp-1 binding site in the Ccl2 promoter in JAWS II cells. Results in C, E, and F are means ± SEM (n = 3). ∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗∗, p < 0.001.