Dendritic Cells + Α and CD8 + Mouse CD4 Bcl6 Is Required for The

Bcl6 Is Required for the Development of Mouse CD4 + and CD8α+ Dendritic Cells Hiromi Ohtsuka, Akemi Sakamoto, Jing Pan, Sumina Inage, Satoshi Horigome, Hirohito Ichii, Masafumi Arima, This information is current as Masahiko Hatano, Seiji Okada and Takeshi Tokuhisa of September 26, 2021. J Immunol 2011; 186:255-263; Prepublished online 3 December 2010; doi: 10.4049/jimmunol.0903714

http://www.jimmunol.org/content/186/1/255 Downloaded from

References This article cites 50 articles, 26 of which you can access for free at: http://www.jimmunol.org/content/186/1/255.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 26, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Bcl6 Is Required for the Development of Mouse CD4+ and CD8a+ Dendritic Cells

Hiromi Ohtsuka,* Akemi Sakamoto,* Jing Pan,* Sumina Inage,* Satoshi Horigome,* Hirohito Ichii,* Masafumi Arima,* Masahiko Hatano,† Seiji Okada,‡ and Takeshi Tokuhisa*

Th2-type inﬂammation spontaneously shown in Bcl6-knockout (KO) mice is mainly caused by bone marrow (BM)-derived nonlymphoid cells. However, the function of dendritic cells (DCs) in Bcl6-KO mice has not been reported. We show in this article that the numbers of CD4+ conventional DCs (cDCs) and CD8a+ cDCs, but not of plasmacytoid DCs, were markedly reduced in the spleen of Bcl6-KO mice. Generation of cDCs from DC progenitors in BM cells was perturbed in the spleen of irradiated wild-type (WT) mice transferred with Bcl6-KO BM cells, indicating an intrinsic effect of Bcl6 in cDC precursors. Although cDC precursors were developed in a Bcl6-KO BM culture with Fms-like tyrosine kinase 3 ligand, the cDC precursors were more apoptotic than Downloaded from WT ones. Also p53, one of the molecular targets of Bcl6, was overexpressed in the precursors. The addition of a p53 inhibitor to Bcl6-KO BM culture protected apoptosis, suggesting that Bcl6 is required by cDC precursors for survival by controlling p53 expression. Furthermore, large numbers of T1/ST2+ Th2 cells were naturally developed in the spleen of Bcl6-KO mice. Th2 skewing was accelerated in the culture of WT CD4 T cells stimulated with Ags and LPS-activated Bcl6-KO BM-derived DCs, which produced more IL-6 and less IL-12 than did WT DCs; the addition of anti–IL-6 Abs to the culture partially abrogated the Th2 skewing. These results suggest that Bcl6 is required in cDC precursors for survival and in activated DCs for modulating the http://www.jimmunol.org/ cytokine proﬁle. The Journal of Immunology, 2011, 186: 255–263.

he human proto-oncogene Bcl6 has been identified from of splenic CD4 T cells in Bcl6-KO mice is due to CD4 T cells and chromosomal breakpoints involving 3q27 in diffuse large bone marrow (BM)-derived nonlymphoid cells, and one of the T B cell lymphomas (1–3). The Bcl6 gene encodes a 92–98- nonlymphoid cells is macrophages that present Ags to T cells kDa nuclear phosphoprotein (4, 5) that contains the BTB/POZ (19). Dendritic cells (DCs) are also the most effective APCs to domain in the NH2-terminal region and Kru¨ppel-type zinc finger stimulate naive T cells in vivo and influence Th1/Th2 differenti- motifs in the C-terminal region (1–3, 6). Because the NH2- ation. However, the functions of Bcl6-KO DCs and their role in by guest on September 26, 2021 terminal half of Bcl6 can bind to silencing mediator of retinoid this Th2 skewing have not been reported. and thyroid receptor protein and recruit the histone deacetylase Mouse splenic DCs consist of two populations; B220+ plasma- complex to silencer regions of target genes to repress expression cytoid DCs (pDCs) and CD11c+ B2202 conventional DCs (cDCs). of these genes (7, 8), Bcl6 acts as a sequence-specific transcrip- cDCs are subdivided into CD4+ cDCs, CD8a+ cDCs, and CD8a2 tional repressor (9–14). To observe physiological functions of CD42 cDCs (20) in a steady state. These DCs have different Bcl6, this gene was disrupted in the mouse germ line. Bcl6- immune functions. CD4+ cDCs have not been found to produce knockout (KO) mice showed growth retardation (15–17) and dis- cytokines, but they effectively present Ags to CD4 T cells (21). played Th2-type inflammatory responses in multiple organs, es- CD8a+ cDCs perform cross-presentation of foreign Ags to CD8 pecially in the heart and lungs, characterized by infiltration of T cells and are a major producer of IL-12 (21–24). CD8a2CD42 eosinophils at a young adult age (15, 16, 18). This Th2 skewing cDCs produce IFN-g (21). cDCs and pDCs can be generated from their progenitor cells in BM cells in vitro by Fms-like tyrosine kinase 3 ligand (FLT3L) stimulation (25). Transcription factors *Department of Developmental Genetics and †Department of Biomedical Science, ‡ required for generation of these subsets of DCs have been exten- Graduate School of Medicine, Chiba University, Chiba; and Division of Hemato- + poiesis, Center for AIDS Research, Kumamoto University, Kumamoto, Japan sively studied. IRF-2, IRF-4, and RelB are essential for CD4 cDC + Received for publication November 18, 2009. Accepted for publication October 27, generation (26–28), and IRF-8 and Id2 are critical for CD8a cDC 2010. generation (29, 30). Because the Bcl6 gene is one of the molecular This work was supported in part by grants-in-aid from the Ministry of Education, targets of IRF-8 (31) and IRF-4 (32), generation of DC subsets Culture, Sports, Science and Technology of Japan and the Global Center of Excellence may be perturbed in Bcl6-KO mice. This perturbation may cause Program (Global Center for Education and Research in Immune System Regulation and Treatment), Ministry of Education, Culture, Sports, Science and Technology, the Th2 skewing in Bcl6-KO mice. Japan. Furthermore, as a transcriptional repressor, Bcl6 regulates ex- Address correspondence and reprint requests to Dr. Akemi Sakamoto, Department of pression of genes, which are related to functions of Th2 cells Developmental Genetics (H2), Graduate School of Medicine, Chiba University, In- and their skewing, such as the IL-5 gene in CD4 T cells (33), ohana 1-8-1, Chuo-ku, Chiba 260-8670, Japan. E-mail address: sakamoto@faculty. chiba-u.jp the IL-18 gene (34), and the IL-6 gene (19) in macrophages. These Abbreviations used in this paper: BM, bone marrow; CBA, Cytometric Bead Array; results also suggested that abnormal functions of Bcl6-KO DCs cDC, conventional dendritic cell; DC, dendritic cell; FLT3L, Fms-like tyrosine kinase cause the Th2 skewing in Bcl6-KO mice. In this study, we exam- 3 ligand; KO, knockout; MFI, mean fluorescent intensity; ND, not detected; pDC, ined differentiation and functions of DCs in the spleen of Bcl6-KO plasmacytoid dendritic cell; PFTa, pifithrin-a; WT, wild-type. mice and in Bcl6-KO BM culture with FLT3L. We show that + + Copyright Ó 2010 by The American Association of Immunologists, Inc. 0022-1767/10/$16.00 numbers of CD4 and CD8a cDCs were dramatically reduced www.jimmunol.org/cgi/doi/10.4049/jimmunol.0903714 256 A ROLE FOR BCL6 IN DEVELOPMENT OF cDCs in the spleen of Bcl6-KO mice. BM cell-transfer experiments ratio of the calculated amount of candidate RNA in a given sample/calculated clearly showed that this reduction in cDCs was due to the intrinsic amount of the housekeeping control gene (b-actin) in the same sample. effect of Bcl6 in their precursor cells. Moreover, the profile of Flow-cytometric analysis of DCs cytokines produced by Bcl6-KO DCs activated with LPS was different from that produced by activated wild-type (WT) DCs. Spleen- or BM-cell suspension was treated with lysis buffer (0.155 M am- monium chloride, 0.1 M disodium EDTA, 0.01 M potassium bicarbonate) to We discuss the roles for Bcl6 in cDC generation and DC functions lyse erythrocytes before staining. A single-cell suspension was prepared in that are related to Th2 skewing. staining solution (PBS with 3% FCS and 0.1% sodium azide). Spleen or BM cells (106) were blocked with unconjugated anti-CD32/16 Abs (2.4G2; BD Materials and Methods Pharmingen), followed by incubation with biotinylated Abs and incubation with directly conjugated Abs and streptavidin-PerCP (BD Pharmingen). The Mice following Abs were used for staining: CD11c-FITC, CD11c-PE, CD11b- FITC, CD11b-PE, CD11b-allophycocyanin, CD8a-PE, CD24-PE, B220- C57BL/6 and BALB/c mice were purchased from Japan SLC (Hamamatsu, d Japan). Bcl6-KO (17), lck-Bcl6–transgenic (35), DO11.10-transgenic (36), PE, B220-allophycocyanin, CD45RA-PE, CD24-FITC, IA -PE, CD86-PE, and p53-KO (37) mice were described previously. Bcl6-KO mice were Ly5.2-FITC, Ly5.2-biotin, Ly5.1-biotin, Ly5.1-PE, avidin-allophycocyanin backcrossed to C57BL/6 or BALB/c mice for .10 generations. C57BL/6 (BD Pharmingen), CD8a-allophycocyanin, CD44-PE (Caltag Laboratories, (B6-Ly5.1) mice were purchased from Charles River Japan (Yokohama, Burlingame, CA), CD4-allophycocyanin, CD135-PE (eBioscience, San Japan). All mice were maintained under specific pathogen-free conditions Diego, CA), T1/ST2-FITC (MD Bioscience, St Paul, MN), mPDCA-1– in the animal center of the Graduate School of Medicine, Chiba University. allophycocyanin (Miltenyi Biotec, Bergish Gladbach, Germany), Annexin The care of all animals was in accordance with Chiba University Animal V-FITC, and Annexin V-biotin (Bender MedSystems, Vienna, Austria). The Care Guidelines. stained cells were analyzed using FACSCalibur. In some experiments, each subset of DCs was sorted by FACSVantage. The purity of those isolated Downloaded from BM-derived DCs cells was .95%. 2 2 2 2 2 2 Lineage-negative (B220 , Gr-1 , CD11b , CD8 , CD4 , CD5 , and Intracellular staining TER1192) BM cells were magnetically separated and then used for DC induction. Lineage-negative BM cells were cultured with RPMI 1640 DO11.10CD4TcellswereculturedwithLPS-activatedBM-derivedDCs and medium (Sigma-Aldrich, St Louis, MO) supplemented with 100 ng/ml OVA peptide (OVA323–329, 100 mg) for 7 d. These activated DO11.10 CD4 FLT3L(R&DSystems,Minneapolis,MN),10%FCS(Intergen,New T cells were stimulated with PMA (1 mg/ml) and ionomycin (1 mg/ml) for 6 h; monensin (2 mM) was added for the last 4 h at 37˚C. Restimulated CD4 York, NY), 50 mM 2-ME, 10 mM HEPES, 100 mg/ml streptomycin (Wako http://www.jimmunol.org/ Chemical, Osaka, Japan), and 100 U/ml penicillin G potassium (Banyu T cells and BM-derived DCs were stained with various Abs for surface Pharmaceutical, Tokyo, Japan) at 37˚C with 5% CO2. In the indicated staining and then were subjected to intracellular staining for various Abs experiment, lineage-negative BM cells were labeled with CFSE (Molec- using the CytoFix/Cytoperm kit (BD Pharmingen), according to the man- ular Probes, Eugene, OR), according to the manufacturer’s instructions, or ufacturer’s instructions. Abs used for intracellular staining included Bcl-2– cultured with 10 mM pifithrin-a (PFTa) (Calbiochem, San Diego, CA). FITC (BD Pharmingen), Bcl-xL–biotin (Chemicon International, Temecula, CA), p53-PE (BD Pharmingen), IFN-g–allophycocyanin (BD Pharmingen), Immunization of mice with Ag-pulsed BM-derived DCs and IL-4–PE (BD Pharmingen). 6 BALB/c mice were injected i.v. with DO11.10 CD4 T cells (5 3 10 )24h Assay for cytokines and chemokine before immunization. BM-derived DCs from Bcl6-KO and WT mice were stimulated with LPS (10 mg/ml; Sigma-Aldrich) for 12 h and incubated The amounts of cytokines and chemokine in culture supernatants were 5 by guest on September 26, 2021 with OVA peptide (OVA323–329). These OVA-pulsed DCs (5 3 10 ) were measured with the CBA. injected into the footpads of reconstituted mice. Popliteal lymph node cells were harvested at day 4 after immunization and cultured with WT splenic Statistical analysis APCs and OVA peptide for 72 h. The amount of cytokines in the culture Statistical analysis was performed using the unpaired t test; p values ,0.05 supernatants was measured using a Cytometric Bead Array (CBA) kit (BD were considered significant. Pharmingen, San Diego, CA), according to the manufacturer’s instructions. Stimulation of CD4 T cells with BM-derived DCs Results + a+ DO11.10 CD4 T cells (1 3 106) were cultured with LPS-activated BM- Numbers of CD4 and CD8 cDCs are reduced in the spleen derived DCs (5 3 105) and OVA peptides for 7 d. These cultured CD4 of Bcl6-KO mice T cells were restimulated with PMA and ionomycin for 6 or 24 h. The We examined subsets of DCs in the spleen of Bcl6-KO mice with amount of cytokines (IFN-g and IL-4) in activated CD4 T cells and in the culture supernatants was measured by FACS and CBA, respectively. a C57BL/6 background. As shown in Fig. 1A, the percentage and number of cDCs (CD11c+) in Bcl6-KO mice were less than those in BM-reconstitution assay WT mice. In contrast, the percentage and number of pDCs (B220+ + Half a million BM cells from Bcl6-KO (Ly5.2) and WT (Ly5.1) mice were mPDCA-1 ) were compatible between them. The percentages and mixed and injected i.v. into lethally irradiated (8.5 Gy) (Ly5.1 3 Ly5.2) F1 numbers of CD4+ and CD8a+ cDCs in Bcl6-KO mice were less than mice. DCs and CD4 T cells from the spleen of transferred mice were those in WT mice (Fig. 1B). However, the number of CD42CD8a2 analyzed by FACS 8 wk after transfer. cDCs in Bcl6-KO mice was similar to that in WT mice. Although Real-time quantitative RT-PCR there were a few CD4+ and CD8a+ cDCs in Bcl6-KO mice, the amount of CD11b on Bcl6-KO CD4+ cDCs was more than that on Total RNAwas isolated using TRIzol reagent (Invitrogen Life Technologies, + dull Carlsbad, CA). Total RNA was reverse transcribed with Superscript III WT CD4 cDCs (Fig. 1C). There were many CD8a cDCs in Reverse Transcriptase (Invitrogen Life Technologies) and oligo(dT) (Amer- Bcl6-KO mice, and those corresponded to the CD24+CD11bdull sham Biosciences, Piscataway, NJ), and the cDNAs were used for PCR. cDC population. A few CD8adull cDCs were also detected in WT Real-time RT-PCR was performed as described elsewhere (38). Real-time mice. CD135 (FLT3) expression was rather low on CD4+ cDCs, PCR with cDNAs was conducted using SYBR Green PCR Master Mix 2 2 + (Applied Biosystems, Foster City, CA) and run on an ABI Prism 7000 Se- CD4 CD8a cDCs, and pDCs, but it was high on CD8a cDCs in low + quence Detection System (Applied Biosystems). The following primers WT mice (Fig. 1D). CD135 CD4 cDCs were missing in Bcl6- were used: Bcl6: forward 59-CCGGCTCAATAATCTCGTGAA-39, reverse KO mice. The amount of CD135 on Bcl6-KO CD42CD8a2 cDCs 59-GGTGCATGTAGAGTGGTGAGTGA-39 and b-actin: forward 59-CC- was more than that on WT CD42CD8a2 cDCs. A similar impair- AGCCTTCCTTCTTGGGTAT-39, reverse 59-TGGCATAGAGGTCTTTA- CGGATGT-39. Data were captured and analyzed using Sequence Detector ment of cDC generation was observed in the spleen of Bcl6-KO software (Applied Biosystems) to determine the threshold fluorescence, mice with a different genetic (BALB/c) background (data not according to the manufacturer’s instructions. The data are reported as the shown). The Journal of Immunology 257 Downloaded from

FIGURE 1. DCs in the spleen of WT and Bcl6-KO mice were analyzed by FACS. A, A representative FACS profile. The number in each square rep- http://www.jimmunol.org/ resents the percentage. Open bars: WT mice. Filled bar: Bcl6-KO mice. Results represent mean 6 SD of six mice. B, CD4+ or CD8a+ cDCs within CD11chigh gated cells. The numbers in the FACS profiles indicate the percentage for each quadrant. Results represent mean 6 SD of six mice. C, CD11b and CD24 expression on CD4+ or CD8a+ cDCs. The numbers indicate the percentage for each square. Data are a representative of five independent experiments. D, CD135 (FLT3) expression on each DC subset. The numbers in the FACS profiles indicate the percentage. Data are a representative of five independent experiments. E, Bcl6 mRNA expression in each WT DC subset was analyzed by real-time RT-PCR. **p , 0.01; ***p , 0.005.

We analyzed the amount of Bcl6 mRNA in FACS-isolated DC was very low. The number of B2202 cDCs, but not B220+ pDCs, in subsets from the spleen of WT mice by real-time RT-PCR. The the Bcl6-KO BM culture was signiﬁcantly less than in the WT BM relative amounts in CD4+ and CD8a+ cDCs were 5.1- and 9.8-fold culture. There were fewer Bcl6-KO CD24+CD11bdull cDCs than by guest on September 26, 2021 greater, respectively, than in CD42CD8a2 cDCs (Fig. 1E). The amount of Bcl6 in pDCs was comparable to that in CD42CD8a2 cDCs. These results suggested that Bcl6 is required for the generation of CD4+ and CD8a+ cDCs, which express a larger amount of Bcl6 in the spleen. Bcl6 in cDC precursor cells is required for generation of cDCs We examined the presence of DC progenitor cells, which express CD135 or Ly49Q, in BM cells of Bcl6-KO mice. The percentages of CD135+ or Ly49Q+ cells in Bcl6-KO BM cells were comparable to those in WT BM cells (Fig. 2A). To examine whether the impaired cDC generation in the spleen of Bcl6-KO mice was due to a defect of Bcl6 in cDC precursor cells, we transferred lineage- negative Bcl6-KO (Ly5.2) and WT (Ly5.1) BM cells at a ratio of 1:1 into irradiated (Ly5.1 3 Ly5.2) F1 WT mice. As shown in Fig. 2B, the percentage of CD11c+ cDCs derived from Bcl6-KO BM cells was less than that from WT BM cells in the recipient spleen. However, the percentage of B220+mPDCA-1+ pDCs derived from Bcl6-KO BM cells was comparable to that from WT BM cells. The percentages and numbers (data not shown) of CD4+ and CD8a+ cDCs derived from Bcl6-KO BM cells were also less than those from WT BM cells. BM culture with FLT3L is a well-established system to examine FIGURE 2. Generation of DCs from DC progenitor cells in Bcl6-KO the in vitro generation of DCs, especially CD8a+ cDCs and B220+ BM cells in vivo. A, CD135 or Ly49Q expression on BM cells of Bcl6-KO and WT mice was examined by FACS. The numbers indicate the per- pDCs (25). To conﬁrm the role of Bcl6 in cDC generation, lineage- centage. Data are a representative of three independent experiments. B, negative Bcl6-KO and WT BM cells were cultured with FLT3L for + + Lineage-negative BM cells from WT (Ly5.1) and Bcl6-KO (Ly5.2) mice 7 d. WT BM cells differentiated into B220 pDCs, as well as CD24 were transferred into irradiated (Ly5.1 3 Ly5.2) F1 mice. Donor-derived dull low + 2 CD11b and CD24 CD11b DCs in the B220 (cDC) popula- DCs (Ly5.1+Ly5.22 or Ly5.12Ly5.2+) in the spleen were analyzed by + tion (Fig. 3A). In the Bcl6-KO BM culture, the percentage of CD24 FACS 8 wk after transfer. The numbers indicate the percentage. Data are CD11bdull cDCs, but not CD24lowCD11b+ cDCs and B220+ pDCs, a representative of three independent experiments. 258 A ROLE FOR BCL6 IN DEVELOPMENT OF cDCs

into CD24+CD11b+ cells in Bcl6-KO and WT BM culture at day 2 of culture (Fig. 4B). CD24+CD11bdull and CD24lowCD11b+ cDC precursors were generated in the WT BM culture after day 4. CD24lowCD11b+ cDC precursors, but not CD24+CD11bdull cDC precursors, were generated in the Bcl6-KO BM culture after day 4. B220+ pDC precursors were generated from DC progenitor cells in Bcl6-KO and WT BM cultures after day 4 (data not shown). The relationship between cell proliferation and generation of CD24+CD11bdull cDC precursors was examined using CFSE- labeled BM cells. Similar to WT BM cells, Bcl6-KO BM cells proliferated quickly after day 2 of culture (Fig. 4C). CD11bdull cells developed in the WT BM culture following cell-cycle progression. Bcl6-KO CD11b+ cells proliferated well, but differentiation of CD11bdull cells in the proliferating cells was impaired. Further- more, the percentage of CD11bdull cells in the nondividing cells (dotted square) increased in the Bcl6-KO BM culture at day 5. To detect apoptotic cells in the Bcl6-KO BM culture, CFSE- labeled or unlabeled BM-derived DCs were stained with Annexin

V at day 5 of culture. The percentage of Annexin V+ cells in the Downloaded from proliferating Bcl6-KO DC precursors was greater than in the WT DC precursors (Fig. 4D). The percentage of Annexin V+ cells in Bcl6-KO CD24lowCD11b+ cDC precursors and CD24+CD11bdull cDC precursors was greater than in WT ones. However, there was no difference in the percentages in B220+ pDCs between those

cultures. These results indicate that Bcl6 protects cell death of http://www.jimmunol.org/ FIGURE 3. Differentiation of DCs from DC progenitor cells in Bcl6-KO cDC precursor cells developed in BM culture with FLT3L. BM cells in vitro. Lineage-negative BM cells from Bcl6-KO and WT mice We analyzed the amounts of apoptosis-related proteins (Bcl-2, were cultured with FLT3L for 7 d. DCs in the culture were analyzed by Bcl-xL, and p53) in CD24lowCD11b+ cDC precursors, CD24+ FACS. A, A representative FACS profile. The numbers indicate the per- CD11bdull cDC precursors, and B220+ pDCs developed in the centage of each gate. Open bars: WT BM cells. Filled bars: Bcl6-KO BM cells. Data are a representative of five independent experiments. Results BM culture at day 5. The mean fluorescent intensity (MFI) of + dull are mean 6 SD of triplicate cultures. ***p , 0.005. B, BM cells from Bcl-2 in Bcl6-KO CD24 CD11b cDC precursors was slightly Bcl6-KO (Ly5.2) and WT (Ly5.1) mice were mixed and cultured with greater than that in WT ones (Fig. 4E). The amounts of Bcl-xL and FLT3L. DCs (Ly5.1+ or Ly5.2+) in the culture were analyzed by FACS. p53 in three subsets of Bcl6-KO DCs were greater than in WT DCs. The numbers indicate the percentage of each gate. Data are presented as The difference in the MFI of Bcl-2, Bcl-xL, and p53 between by guest on September 26, 2021 a representative of three independent experiments. Bcl6-KO and WT CD24+CD11bdull cDC precursors was the largest among three BM-derived DC subsets. Because the p53 gene is one WT ones. A similar impairment of cDC generation was observed in of the molecular targets of Bcl6 (39), the amount of p53 mRNA in the BM culture of Bcl6-KO mice with a BALB/c background (data Bcl6-KO BM-derived DCs was analyzed by real-time RT-PCR; it not shown). was 5-fold greater than in WT BM-derived DCs (data not shown). To examine the intrinsic effect of Bcl6 in cDC precursor cells PFTa (p53 inhibitor) was added to the Bcl6-KO BM culture + on cDC generation, lineage–negative Bcl6-KO (Ly5.2) and WT at the beginning of culture to examine the generation of CD24 dull (Ly5.1) BM cells were mixed at a 1:1 ratio and stimulated with CD11b cDC precursors at day 7. Although its addition did not FLT3L for 7 d. The Ly5.1+ WT subset accounted for 82% of the affect DC generation in the WT BM culture, it increased the per- + dull DCs (Fig. 3B). Bcl6-KO CD24+CD11bdull cDCs were fewer in centage and number of CD24 CD11b cDC precursors in the percentage and number compared with the WT cDCs (0.29 6 Bcl6-KO BM culture (Fig. 4F). Furthermore, the number of splenic 0.02 3 105/well versus 9.4 6 0.4 3 105/well, respectively; p = DCs in (p53-KO 3 Bcl6-KO) double-KO mice was compared with + 0.00048). Although the percentage of Bcl6-KO CD24lowCD11b+ that in p53-KO and Bcl6-KO mice. The numbers of CD11c cells + + cDCs was greater than WT cDCs, the number (4.1 6 0.3 3 105/ and CD4 and CD8a DCs in double-KO mice were similar to those well) of Bcl6-KO CD24lowCD11b+ cDCs was significantly less in p53-KO mice and were greater than those in Bcl6-KO mice (Fig. than that of WT ones (7.1 6 0.1 3 105/well; p = 0.0014), support- 4G). These results indicated that downregulation of p53 expression + ing the intrinsic effect of Bcl6 in cDC precursor cells on the by Bcl6 is important for normal CD11c DC generation. generation of cDCs. Th2 cells are dominantly induced from naive WT CD4 T cells Bcl6 is required for protecting apoptosis of cDC precursor stimulated with activated Bcl6-KO DCs and Ags cells Bcl6-KO mice spontaneously display Th2-type inflammation (18). Kinetics of Bcl6 mRNA expression in DCs generated in WT BM Indeed, T1/ST2, which is a marker for Th2 cells (40), was detected culture with FLT3L was analyzed by real-time RT-PCR. Bcl6 in significant percentages of CD4 T cells in the spleen of naive mRNA expression was induced in BM-derived DCs from day 2 Bcl6-KO mice with a C57BL/6 or BALB/c background (Fig. 5A). of culture, and the amount at day 2 of culture was the greatest until However, T1/ST2+ CD4 T cells were not detected in the spleen of day 5 of culture (Fig. 4A), suggesting the perturbation of DC gen- naive C57BL/6 or BALB/c mice. T1/ST2+ CD4 T cells in the spleen eration in the Bcl6-KO BM culture after day 2 of culture. The of Bcl6-KO mice were CD44+ (Fig. 5B), CD62L2, and CD45RB2 kinetics of DC generation in the Bcl6-KO BM culture with FLT3L (data not shown) memory phenotype CD4 T cells. When Bcl6-KO was compared with that in the WT BM culture. Lineage-negative mice were mated with lck-Bcl6–transgenic mice to generate Bcl6- 2 B220 CD24+CD11bdull BM cells at day 0 of culture differentiated KO-Rescure mice, in which the exogenous Bcl6 is expressed only in The Journal of Immunology 259 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 4. Kinetics of DC differentiation in Bcl6-KO BM culture. Lineage-negative BM cells from Bcl6-KO and WT mice were cultured with FLT3L for 7 d. A, Bcl6 mRNA expression in differentiating WT DCs was analyzed by real-time RT-PCR. B, Differentiating cDCs (B2202 gated) were analyzed by FACS. The numbers indicate the percentage of each gate. Data are a representative of five independent experiments. C, CFSE-labeled lineage-negative BM cells were cultured. CFSE and CD11b expression on DCs in the culture was analyzed by FACS. The numbers indicate the percentage of each square. Data are presented as a representative of three independent experiments. D, Annexin V expression on CFSE-labeled DCs; each subset of DCs in the culture was analyzed on day 5. Shaded graphs: WT mice. Open graphs: Bcl6-KO mice. The numbers indicate the percentage of each quadrant and Annexin V+ DCs. Data are a representative of three independent experiments. E, Apoptosis-related proteins in each subset of DCs were examined by FACS. Shaded graph: WT mice. Open graph: Bcl6-KO mice. The numbers indicate the MFI for each subset of DCs. Data are a representative of three independent experiments. F, PFTa was added to the BM culture at the beginning of culture. Representative FACS profile and the cell number of each subset of DCs. The numbers in the FACS profiles indicate the percentage of each square. Data are a representative of five independent experiments. Open bars: WT cells. Filled bars: Bcl6-KO cells. Results represent mean 6 SD of triplicate cultures. ***p , 0.005. G, DCs in the spleen of p53-KO, Bcl6-KO, and (p53-KO 3 Bcl6-KO) double-KO mice were analyzed by FACS. The numbers indicate the percentage of each square or quadrant.

T cells (35), the percentage of T1/ST2+ CD4 T cells in the spleen of supernatants of lymph node cells from mice injected with Bcl6- Bcl6-KO-Rescure mice was similar to that of Bcl6-KO mice, sug- KO DCs was greater than in mice injected with WT DCs (Fig. gesting that non-T cells are the main cells responsible for inducing 5D). These results indicate that BM-derived DCs are responsible Th2 skewing in Bcl6-KO mice. for inducing Th2 skewing in Bcl6-KO mice. Lineage-negative Bcl6-KO (Ly5.2) and WT (Ly5.1) BM cells To examine the function of BM-derived DCs, lineage-negative were transferred into irradiated (Ly5.1 3 Ly5.2) F1 WT mice. BM cells from Bcl6-KO mice with a BALB/c background were We examined the percentage of T1/ST2+ CD4 T cells in the spleen cultured with FLT3L for 7 d. These DCs were stimulated with of BM chimeric mice. When mice were simultaneously transferred LPS. Expression of IAd and CD86 was examined on activated with Bcl6-KO and WT BM cells, the percentages of T1/ST2+ CD4 Bcl6-KO and WT DCs by FACS. The MFI on activated Bcl6- T cells derived from Bcl6-KO and WT BM cells were greater than KO DCs was similar to those on activated WT DCs 24 h after in mice that received WT BM cells (Fig. 5C). Next, OVA-pulsed stimulation (Fig. 6A). Cytokine and chemokine production by BM-derived DCs were injected into the footpads of mice reconsti- activated Bcl6-KO DCs was analyzed using CBA. The amounts tuted with DO11.10 CD4 T cells. Popliteal lymph node cells from of IL-6 and MCP-1 secreted by activated Bcl6-KO DCs were the immunized mice were cultured with WT splenic APCs and OVA much greater than those secreted by activated WT DCs, and peptide for 72 h. The amount of IL-4, but not IFN-g, in culture TNF-a production by Bcl6-KO DCs was also greater than that 260 A ROLE FOR BCL6 IN DEVELOPMENT OF cDCs

(Fig. 7A). Because IL-6 (41) and IL-12 (42) affect Th2 and Th1 skewing, respectively, Abs against IL-6 or IL-12 were added to the CD4 T cell-activation culture at the beginning of culture. The addition of anti–IL-6 Abs to the culture partially abrogated the Th2 skewing with Bcl6-KO DCs but not that with WT DCs (Fig. 7B). The neutralization of IL-12 in the culture by anti–IL-12 Abs increased the percentage of T1/ST2+ CD4 T cells stimulated with WT DCs but not that stimulated with Bcl6-KO DCs. To conﬁrm Th2 skewing of CD4 T cells stimulated with BM-derived DCs, those cultured CD4 T cells were restimulated with PMA and ionomycin at day 7. IFN-g and IL-4 production in the activated CD4 T cells and in the culture supernatants was examined by FACS and CBA, respectively. CD4 T cells stimulated with Bcl6- KO DCs contained less IFN-g–producing cells and more IL-4 producing cells than did those stimulated with WT DCs (Fig. 7C). The amounts of IL-4 and IFN-g in the culture supernatants correlated with the percentages of IL-4– and IFN-g–producing cells. These results indicated that Bcl6-KO BM-derived DCs act

as effective Th2 inducers by producing greater amounts of IL-6 Downloaded from and lesser amounts of IL-12.

Discussion The number of cDCs, especially CD4+ and CD8a+ cDCs, was signiﬁcantly reduced in the spleen of Bcl6-KO mice. The numbers

of hematopoietic stem cells, common lymphoid precursors, com- http://www.jimmunol.org/ mon myeloid precursors (data not shown), and lineage-negative FLT3+ cells in BM cells were comparable between Bcl6-KO and WT mice. Recently, a single progenitor was reported for pDCs and cDCs (43, 44), and the number of pDCs in the spleen of Bcl6-KO mice was similar to that in WT mice. These results suggested that DC progenitors develop normally in the BM of Bcl6-KO mice. The abnormal development of cDCs in Bcl6-KO mice was conﬁrmed by BM culture with FLT3L. GM-CSF is also an important cytokine for DC development (45), and the number of cDCs that developed in by guest on September 26, 2021 FIGURE 5. Th2 skewing of CD4 T cells in the spleen of Bcl6-KO mice. the Bcl6-KO BM culture with GM-CSF instead of FLT3L was T1/ST2 expression on CD4 T cells in the spleen of naive Bcl6-KO mice similar to that in the WT BM culture (data not shown). Because was examined by FACS. A, CD4 T cells in the spleen of Bcl6-KO and WT the BM culture with FLT3L is more physiological for the develop- mice with C57BL/6 or BALB/c background were analyzed. The numbers indicate the percentage of each gate. Data are a representative of three ment of splenic DC precursor than is the BM culture with GM-CSF independent experiments. B, CD4 T cells in the spleen of various mice (46), Bcl6 may be required for generation of cDCs in vivo from DC with a C57BL/6 background were analyzed. The numbers indicate the progenitors with FLT3L stimulation. + dull percentage of each quadrant. Data are a representative of three independent There were few CD8a cDCs but detectable CD8a cDCs in experiments. C, Lineage-negative Bcl6-KO (Ly5.2) and WT (Ly5.1) BM the spleen of Bcl6-KO mice, and CD8adull cDCs belonged to the cells were transferred into irradiated (Ly5.1 3 Ly5.2) F1 WT mice. T1/ CD24+CD11bdull cDC population. A few CD8adull cDCs were also ST2+ CD4 T cells in the spleen of BM chimeric mice were examined by detected in the spleen of WT mice. CD8a on cDC precursors is FACS 8 wk after transfer. Open bars: WT BM cells. Filled bars: Bcl6-KO upregulated by the maturation process in the spleen of WT mice BM cells. Data are a representative of three independent experiments. D, (47), suggesting that CD8adull cDCs in the spleen are precursors of WT mice transferred with DO11.10 CD4 T cells were immunized with CD8a+ cDCs and that Bcl6 is required for the differentiation of OVA-pulsed BM-derived DCs into the footpads. Popliteal lymph node + dull cells were harvested and restimulated with APCs and OVA peptide for 72 CD8a cDCs from CD8a cDCs in the spleen. Kinetics of DC h. The amount of cytokines in the supernatant was analyzed by CBA. Open generation in the BM culture with FLT3L supports the differentia- + dull + + dull bars: WT BM-derived DC-immunized mice. Filled bars: Bcl6-KO BM- tion of CD24 CD11b CD8a cDCs from CD24 CD11b dull + 2 derived DC-immunized mice. Data are a representative of three in- CD8a cDC precursors. CD24 CD11b BM cells differentiated dependent experiments. Results are mean 6 SD of triplicate cultures. *p , into CD24+CD11b+ cDC precursors at day 2 of culture and into 0.05; **p , 0.01. CD24+CD11bdull cDC precursors at day 4 of culture. Furthermore, the cell-proliferation analysis using CFSE-labeled BM cells dem- of WT DCs (Fig. 6B). However, IL-12 secretion by Bcl6-KO DCs onstrated that CD11bdull cDCs developed from CD11b+ cDC pre- was less than that of WT DCs. These results were repeated using cursors following cell-cycle progression. These results suggested BM-derived DCs from Bcl6-KO mice with a C57BL/6 background that CD24+CD11b+CD8adull cDC precursors differentiate into (data not shown). Furthermore, these results, with the exception CD24+CD11bdullCD8adull cDC precursors and then into CD24+ of TNF-a production, were conﬁrmed using FACS-sorted splenic CD11bdullCD8a+ cDCs. Because the differentiation of CD24+ CD11c+ DCs stimulated with LPS for 24 h (Fig. 6C). CD11bdull cDC precursors from CD24+CD11b+ cDC precursors Naive CD4 T cells from DO11.10-transgenic mice were stimu- was perturbed in the Bcl6-KO BM culture, Bcl6 is also required lated with LPS-activated BM-derived DCs and OVA peptides for for the differentiation of CD24+CD11bdull cDC precursors in vitro. 14 d. The percentage of T1/ST2+ CD4 T cells in the CD4 T cells The largest percentage of Annexin V+ cells was observed within stimulated with Bcl6-KO DCs was greater than that with WT DCs CD24+CD11bdull cDC precursors in the Bcl6-KO BM culture with The Journal of Immunology 261

FIGURE 6. Function of activated Bcl6-KO DCs. Lineage-negative BM cells from Bcl6-KO and WT mice with a BALB/c background were cultured with FLT3L for 7 d. These cells were stimulated with LPS for 24 h. Open bars: WT BM-derived DCs. Filled bars: Bcl6-KO BM-derived DCs. A,IAd and CD86 expression (MFI) on those activated DCs were examined by FACS. Data are a representative of three independent experiments. B, Cytokine secretion from those activated DCs was examined in the culture supernatant. Results are mean 6 SD of triplicate cultures. C, CD11c+ DCs from spleens of Bcl6-KO and Downloaded from WT mice were stimulated with LPS for 24 h. IAd and CD86 expression (MFI) on those activated DCs were examined by FACS. The amount of cytokines in the culture supernatant was examined. Results are mean 6 SD of triplicate cultures. *p , 0.05; ***p , 0.005. ND, not detected.

FLT3L. Although a large percentage of CD24+CD11bdullCD8adull culture, suggesting that Bcl6 is required for differentiating CD8adull cDCs was detected in the spleen of Bcl6-KO mice, their number cDC precursors to protect apoptosis. Among the three DC subsets, 5 (0.7 6 0.02 3 10 /spleen) in the spleen of Bcl6-KO mice was the difference in the MFI of Bcl-2, Bcl-xL, and p53 expression http://www.jimmunol.org/ less than that (1.9 6 0.16 3 105/spleen) of CD24+CD11bdull between the Bcl6-KO and WT cDC population was greatest in the CD8a+∼dull cDCs in WT mice (p = 0.0047). These results sug- CD24+CD11bdull cDC precursors. The percentage of CD11bdull cDC gested that differentiating CD24+CD11bdullCD8adull cDC precur- precursors within the nondividing cells increased in the CFSE- sors in the spleen of Bcl6-KO mice die by apoptosis. Bcl6 mRNA labeled Bcl6-KO BM culture at day 5. Furthermore, the addition was strongly detected in CD8a+ cDCs in the spleen and induced of the p53 inhibitor (PFTa) to the Bcl6-KO BM culture increased in CD24+CD11b+ cDC precursors in the BM culture at day 2 of the percentage and number of CD24+CD11bdull cDC precursors. by guest on September 26, 2021

FIGURE 7. Effect of Bcl6-KO DCs on Th2 skewing of CD4 T cells. Naive WT CD4 T cells of DO11.10 mice were cultured with LPS-activated DCs derived from Bcl6-KO or WT BM culture with FLT3L and OVA peptides for 14 d. T1/ST2 expression on activated CD4 T cells was examined by FACS. A, Numbers indicate the percentage of each gate. Data are a representative of three independent experiments. B, Anti–IL-6 or anti–IL-12 Abs were added to the culture at the beginning of culture. Open bars: WT BM cells. Filled bars: Bcl6-KO BM cells. Results are mean 6 SD of triplicate cultures. C, At day 7 of culture, CD4 T cells were restimulated with PMA and ionomycin for 6 or 24 h. IFN-g and IL-4 production in activated CD4 T cells and in the culture supernatant were examined by FACS and CBA, respectively. Representative FACS proﬁle (left panels). Open graphs: restimulated CD4 T cells. Shaded graphs: nonrestimulated CD4 T cells. The numbers indicate the percentage of each gate. Percentage of cytokine-producing Th cells in the culture (middle panel). Open bars: WT DCs. Filled bars: Bcl6-KO DCs. Results are mean 6 SD of three independent experiments. Amount of cytokines in the culture supernatant (right panels). Open bars: WT DCs. Filled bars: Bcl6-KO DCs. Results are mean 6 SD of triplicate cultures. *p , 0.05; **p , 0.01; ***p , 0.005. 262 A ROLE FOR BCL6 IN DEVELOPMENT OF cDCs

The impairment of DC development in Bcl6-KO mice was rescued the Th2 skewing. Thus, Bcl6 may be important for maintaining the in the (p53-KO 3 Bcl6-KO) double-KO mice. Because p53 is one Th1/Th2 balance via the development of normal DC subsets and of the molecular targets of Bcl6 (39), p53 expression controlled by the regulation of DC functions, including IL-6 and IL-12 produc- Bcl6 in CD24+CD11bdull cDC precursors may be critical for cDC tion. development, and Bcl-2 and Bcl-xL may be produced in CD24+ dull CD11b cDC precursors to protect their apoptosis. Acknowledgments Differentiation of CD4+ cDCs can be traced by the expression + We thank H. Satake for skillful technical assistance and S. Nakamura for of CD24 and CD11b on their cell surface. The majority of CD4 secretarial services. cDCs in the spleen belong to the CD24lowCD11b+ cDC popula- + 2 tion. BM culture with FLT3L demonstrated that CD24 CD11b Disclosures BM cells differentiated into CD24+CD11b+ cDC precursors at The authors have no financial conflicts of interest. day 2 and into CD24lowCD11b+ cDC precursors at day 4, suggesting that CD24+CD11b+ cDC precursors differentiate into CD24low CD11b+CD4+ cDCs. CD24lowCD11b+CD4+ cDCs were barely References detected in the spleen of Bcl6-KO mice. However, the number 1. Kerckaert, J. P., C. Deweindt, H. Tilly, S. Quief, G. Lecocq, and C. Bastard. of CD24lowCD11b+ cDC precursors in the Bcl6-KO BM culture 1993. LAZ3, a novel zinc-finger encoding gene, is disrupted by recurring chromosome 3q27 translocations in human lymphomas. Nat. Genet. 5: 66–70. was slightly less than in the WT BM culture. Thus, Bcl6 is 2. Miki, T., N. Kawamata, S. Hirosawa, and N. Aoki. 1994. Gene involved in the + strongly required for generation of CD4 cDCs in vivo but not 3q27 translocation associated with B-cell lymphoma, BCL5, encodes a Kru¨ppel- for the generation in vitro with FLT3L stimulation. These results like zinc-finger protein. Blood 83: 26–32. Downloaded from low + 3. Ye, B. H., F. Lista, F. Lo Coco, D. M. Knowles, K. Offit, R. S. Chaganti, and suggested that CD24 CD11b cDCs that developed in the BM R. Dalla-Favera. 1993. Alterations of a zinc finger-encoding gene, BCL-6, in culture are still precursors of CD4+ cDCs in the spleen and that diffuse large-cell lymphoma. Science 262: 747–750. Bcl6 is required for the differentiation of CD24lowCD11b+ cDCs 4. Cattoretti, G., C. C. Chang, K. Cechova, J. Zhang, B. H. Ye, B. Falini, low + + D. C. Louie, K. Offit, R. S. Chaganti, and R. Dalla-Favera. 1995. BCL-6 protein into CD24 CD11b CD4 cDCs in vivo. is expressed in germinal-center B cells. Blood 86: 45–53. Bcl6 mRNA was induced in DC precursors developed in the BM 5. Onizuka, T., M. Moriyama, T. Yamochi, T. Kuroda, A. Kazama, N. Kanazawa, K. Sato, T. Kato, H. Ota, and S. Mori. 1995. BCL-6 gene product, a 92- to 98-kD culture with FLT3L, and the generation of DCs was perturbed in http://www.jimmunol.org/ nuclear phosphoprotein, is highly expressed in germinal center B cells and their the Bcl6-KO BM culture with FLT3L but not in that culture with neoplastic counterparts. Blood 86: 28–37. GM-CSF (data not shown), suggesting a role for Bcl6 in the signal 6. Fukuda, T., T. Miki, T. Yoshida, M. Hatano, K. Ohashi, S. Hirosawa, and transduction initiated by FLT3. FLT3L-KO mice showed reduced T. Tokuhisa. 1995. The murine BCL6 gene is induced in activated lymphocytes as an immediate early gene. Oncogene 11: 1657–1663. numbers of DCs and NK cells (48), and Bcl6-KO mice have 7. Dhordain, P., O. Albagli, R. J. Lin, S. Ansieau, S. Quief, A. Leutz, a normal number of NK cells (data not shown), suggesting that J. P. Kerckaert, R. M. Evans, and D. Leprince. 1997. Corepressor SMRT binds the BTB/POZ repressing domain of the LAZ3/BCL6 oncoprotein. Proc. Natl. Bcl6 is important for FLT3-mediated DC cell development but not Acad. Sci. USA 94: 10762–10767. for FLT3-mediated NK cell development. Several transcription 8. Wong, C. W., and M. L. Privalsky. 1998. Components of the SMRT corepressor factors were demonstrated to be critical for DC development. complex exhibit distinctive interactions with the POZ domain oncoproteins

PLZF, PLZF-RARalpha, and BCL-6. J. Biol. Chem. 273: 27695–27702. by guest on September 26, 2021 IRF-2–KO (26), IRF-4–KO (27), and RelB-KO (28) mice showed 9. Albagli, O., P. Dhordain, F. Bernardin, S. Quief, J. P. Kerkaert, and D. Leprince. + impaired generation of CD4 cDCs. IRF-8–KO (29) and Id2-KO 1996. Multiple domains participate in distance-independent LAZ3/BCL6- (30, 49) mice lacked CD8a+ cDCs. However, the type of impair- mediated transcriptional repression. Biochem. Biophys. Res. Commun. 220: 911–915. ment of DC development in the spleen of those KO mice was not 10. Chang, C. C., B. H. Ye, R. S. Chaganti, and R. Dalla-Favera. 1996. BCL-6, similar to that in Bcl6-KO mice, and the amounts of those tran- a POZ/zinc-finger protein, is a sequence-specific transcriptional repressor. Proc. scription factors (IRF-2, IRF-4, RelB, IRF-8, and Id2) in Bcl6-KO Natl. Acad. Sci. USA 93: 6947–6952. 11. Deweindt, C., O. Albagli, F. Bernardin, P. Dhordain, S. Quief, D. Lantoine, BM-derived DCs were similar to those in WT DCs (data not J. P. Kerckaert, and D. Leprince. 1995. The LAZ3/BCL6 oncogene encodes shown). Further study is required to elucidate the molecular tar- a sequence-specific transcriptional inhibitor: a novel function for the BTB/POZ gets of Bcl6 to generate CD4+ and CD8a+ cDCs in vivo. domain as an autonomous repressing domain. Cell Growth Differ. 6: 1495–1503. 12. Seyfert, V. L., D. Allman, Y. He, and L. M. Staudt. 1996. Transcriptional re- CD4 T cells in the spleen of Bcl6-KO mice showed Th2 skewing pression by the proto-oncogene BCL-6. Oncogene 12: 2331–2342. (41). BM-derived nonlymphoid cells, especially macrophages, are 13. Baron, B. W., R. R. Stanger, E. Hume, A. Sadhu, R. Mick, J. P. Kerckaert, the cells responsible for the Th2 skewing (19). We showed in this C. Deweindt, C. Bastard, G. Nucifora, N. Zeleznik-Le, and W. McKeithan. 1995. BCL6 encodes a sequence-specific DNA-binding protein. Genes Chromosomes study that Bcl6-KO DCs are critical for the Th2 skewing. DCs are Cancer 13: 221–224. important APCs that regulate Th1/Th2 polarization of CD4 T cells 14. Kawamata, N., T. Miki, K. Ohashi, K. Suzuki, T. Fukuda, S. Hirosawa, and by the amount of CD86, OX40L, and CD70 on their cell surface N. Aoki. 1994. Recognition DNA sequence of a novel putative transcription factor, BCL6. Biochem. Biophys. Res. Commun. 204: 366–374. (50). However, the expression of CD86, OX40L (data not shown), 15. Dent, A. L., A. L. Shaffer, X. Yu, D. Allman, and L. M. Staudt. 1997. Control of and CD70 (data not shown) on activated Bcl6-KO BM-derived inflammation, cytokine expression, and germinal center formation by BCL-6. DCs was similar to that on activated WT BM-derived DCs. Science 276: 589–592. + 16. Ye, B. H., G. Cattoretti, Q. Shen, J. Zhang, N. Hawe, R. de Waard, C. Leung, CD8a cDCs have Th1-skewing activity through their IL-12 pro- M. Nouri-Shirazi, A. Orazi, R. S. Chaganti, et al. 1997. The BCL-6 proto- duction (42), and Id2-KO mice showed Th2 dominance by the oncogene controls germinal-centre formation and Th2-type inflammation. Nat. defect in CD8a+ cDC generation (49), supporting the Th2 skew- Genet. 16: 161–170. + 17. Fukuda, T., T. Yoshida, S. Okada, M. Hatano, T. Miki, K. Ishibashi, S. Okabe, ing in Bcl6-KO mice with a reduced number of CD8a cDCs. IL- H. Koseki, S. Hirosawa, M. Taniguchi, et al. 1997. Disruption of the Bcl6 gene 6, IL-12, and IL-10 produced by activated DCs are also important results in an impaired germinal center formation. J. Exp. Med. 186: 439–448. 18. Yoshida, T., T. Fukuda, M. Hatano, H. Koseki, S. Okabe, K. Ishibashi, S. Kojima, in regulating the Th1/Th2 polarization of activated CD4 T cells. M. Arima, I. Komuro, G. Ishii, et al. 1999. The role of Bcl6 in mature cardiac IL-6 is one of the molecular targets of Bcl6 (19), and IL-6–based myocytes. Cardiovasc. Res. 42: 670–679. Th2 differentiation by Bcl6-KO macrophages was reported (41). 19. Toney, L. M., G. Cattoretti, J. A. Graf, T. Merghoub, P. P. Pandolfi, R. Dalla- Favera, B. H. Ye, and A. L. Dent. 2000. BCL-6 regulates chemokine gene We showed in this study that Bcl6-KO DCs produced greater transcription in macrophages. Nat. Immunol. 1: 214–220. amounts of IL-6 and lesser amounts of IL-12 than did WT DCs 20. Shortman, K., and Y. J. Liu. 2002. Mouse and human dendritic cell subtypes. after LPS stimulation. Because neutralization of IL-6 by anti–IL-6 Nat. Rev. Immunol. 2: 151–161. 21. Hochrein, H., K. Shortman, D. Vremec, B. Scott, P. Hertzog, and M. O’Keeffe. Abs partially abrogated the Th2 skewing induced by Bcl6-KO 2001. Differential production of IL-12, IFN-alpha, and IFN-gamma by mouse DCs, overproduction of IL-6 by Bcl6-KO DCs explains, in part, dendritic cell subsets. J. Immunol. 166: 5448–5455. The Journal of Immunology 263

22. den Haan, J. M., S. M. Lehar, and M. J. Bevan. 2000. CD8(+) but not CD8(2) 36. Ichii, H., A. Sakamoto, M. Arima, M. Hatano, Y. Kuroda, and T. Tokuhisa. 2007. dendritic cells cross-prime cytotoxic T cells in vivo. J. Exp. Med. 192: 1685– Bcl6 is essential for the generation of long-term memory CD4+ T cells. Int. 1696. Immunol. 19: 427–433. 23. Heath, W. R., G. T. Belz, G. M. Behrens, C. M. Smith, S. P. Forehan, I. A. Parish, 37. Donehower, L. A., M. Harvey, B. L. Slagle, M. J. McArthur, C. A. Montgomery, G. M. Davey, N. S. Wilson, F. R. Carbone, and J. A. Villadangos. 2004. Cross- Jr., J. S. Butel, and A. Bradley. 1992. Mice deficient for p53 are developmentally presentation, dendritic cell subsets, and the generation of immunity to cellular normal but susceptible to spontaneous tumours. Nature 356: 215–221. antigens. Immunol. Rev. 199: 9–26. 38. Yoshida, K., A. Sakamoto, K. Yamashita, E. Arguni, S. Horigome, M. Arima, 24. Reis e Sousa, C., S. Hieny, T. Scharton-Kersten, D. Jankovic, H. Charest, M. Hatano, N. Seki, T. Ichikawa, and T. Tokuhisa. 2006. Bcl6 controls granzyme R. N. Germain, and A. Sher. 1997. In vivo microbial stimulation induces rapid B expression in effector CD8+ T cells. Eur. J. Immunol. 36: 3146–3156. CD40 ligand-independent production of interleukin 12 by dendritic cells and 39. Phan, R. T., and R. Dalla-Favera. 2004. The BCL6 proto-oncogene suppresses their redistribution to T cell areas. J. Exp. Med. 186: 1819–1829. p53 expression in germinal-centre B cells. Nature 432: 635–639. 25. Naik, S. H., A. I. Proietto, N. S. Wilson, A. Dakic, P. Schnorrer, M. Fuchsberger, 40. Lo¨hning, M., A. Stroehmann, A. J. Coyle, J. L. Grogan, S. Lin, J. C. Gutierrez- M. H. Lahoud, M. O’Keeffe, Q. X. Shao, W. F. Chen, et al. 2005. Cutting edge: Ramos, D. Levinson, A. Radbruch, and T. Kamradt. 1998. T1/ST2 is preferen- generation of splenic CD8+ and CD82 dendritic cell equivalents in Fms-like tially expressed on murine Th2 cells, independent of interleukin 4, interleukin 5, tyrosine kinase 3 ligand bone marrow cultures. J. Immunol. 174: 6592–6597. and interleukin 10, and important for Th2 effector function. Proc. Natl. Acad. 26. Ichikawa, E., S. Hida, Y. Omatsu, S. Shimoyama, K. Takahara, S. Miyagawa, Sci. USA 95: 6930–6935. K. Inaba, and S. Taki. 2004. Defective development of splenic and epidermal 41. Yu, R. Y., X. Wang, F. J. Pixley, J. J. Yu, A. L. Dent, H. E. Broxmeyer, CD4+ dendritic cells in mice deficient for IFN regulatory factor-2. Proc. Natl. E. R. Stanley, and B. H. Ye. 2005. BCL-6 negatively regulates macrophage Acad. Sci. USA 101: 3909–3914. proliferation by suppressing autocrine IL-6 production. Blood 105: 1777–1784. 27. Suzuki, S., K. Honma, T. Matsuyama, K. Suzuki, K. Toriyama, I. Akitoyo, 42. Maldonado-Lo´pez, R., C. Maliszewski, J. Urbain, and M. Moser. 2001. Cyto- K. Yamamoto, T. Suematsu, M. Nakamura, K. Yui, and A. Kumatori. 2004. kines regulate the capacity of CD8alpha(+) and CD8alpha(2) dendritic cells to Critical roles of interferon regulatory factor 4 in CD11bhighCD8alpha2 den- prime Th1/Th2 cells in vivo. J. Immunol. 167: 4345–4350. dritic cell development. Proc. Natl. Acad. Sci. USA 101: 8981–8986. 43. Onai, N., A. Obata-Onai, M. A. Schmid, T. Ohteki, D. Jarrossay, and 28. Wu, L., A. D’Amico, K. D. Winkel, M. Suter, D. Lo, and K. Shortman. 1998. M. G. Manz. 2007. Identification of clonogenic common Flt3+M-CSFR+ plas- RelB is essential for the development of myeloid-related CD8alpha2 dendritic macytoid and conventional dendritic cell progenitors in mouse bone marrow. cells but not of lymphoid-related CD8alpha+ dendritic cells. Immunity 9: 839– Nat. Immunol. 8: 1207–1216. Downloaded from 847. 44. Naik, S. H., P. Sathe, H. Y. Park, D. Metcalf, A. I. Proietto, A. Dakic, S. Carotta, 29. Schiavoni, G., F. Mattei, P. Sestili, P. Borghi, M. Venditti, H. C. Morse, III, M. O’Keeffe, M. Bahlo, A. Papenfuss, et al. 2007. Development of plasmacytoid F. Belardelli, and L. Gabriele. 2002. ICSBP is essential for the development of and conventional dendritic cell subtypes from single precursor cells derived mouse type I interferon-producing cells and for the generation and activation of in vitro and in vivo. Nat. Immunol. 8: 1217–1226. CD8alpha(+) dendritic cells. J. Exp. Med. 196: 1415–1425. 45. Inaba, K., M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu, 30. Hacker, C., R. D. Kirsch, X. S. Ju, T. Hieronymus, T. C. Gust, C. Kuhl, T. Jorgas, and R. M. Steinman. 1992. Generation of large numbers of dendritic cells from S. M. Kurz, S. Rose-John, Y. Yokota, and M. Zenke. 2003. Transcriptional mouse bone marrow cultures supplemented with granulocyte/macrophage

profiling identifies Id2 function in dendritic cell development. Nat. Immunol. 4: colony-stimulating factor. J. Exp. Med. 176: 1693–1702. http://www.jimmunol.org/ 380–386. 46. Ardavıń, C. 2003. Origin, precursors and differentiation of mouse dendritic cells. 31. Lee, C. H., M. Melchers, H. Wang, T. A. Torrey, R. Slota, C. F. Qi, J. Y. Kim, Nat. Rev. Immunol. 3: 582–590. P. Lugar, H. J. Kong, L. Farrington, et al. 2006. Regulation of the germinal center 47. Martinez del Hoyo, G., P. Martıń, C. F. Arias, A. R. Marıń, and C. Ardavıń. gene program by interferon (IFN) regulatory factor 8/IFN consensus sequence- 2002. CD8alpha+ dendritic cells originate from the CD8alpha- dendritic cell binding protein. J. Exp. Med. 203: 63–72. subset by a maturation process involving CD8alpha, DEC-205, and CD24 up- 32. Lossos, I. S. 2007. The endless complexity of lymphocyte differentiation and regulation. Blood 99: 999–1004. lymphomagenesis: IRF-4 downregulates BCL6 expression. Cancer Cell 12: 48. McKenna, H. J., K. L. Stocking, R. E. Miller, K. Brasel, T. De Smedt, 189–191. E. Maraskovsky, C. R. Maliszewski, D. H. Lynch, J. Smith, B. Pulendran, et al. 2000. 33. Arima, M., H. Toyama, H. Ichii, S. Kojima, S. Okada, M. Hatano, G. Cheng, Mice lacking flt3 ligand have deficient hematopoiesis affecting hematopoietic M. Kubo, T. Fukuda, and T. Tokuhisa. 2002. A putative silencer element in the progenitor cells, dendritic cells, and natural killer cells. Blood 95: 3489–3497. IL-5 gene recognized by Bcl6. J. Immunol. 169: 829–836. 49. Kusunoki, T., M. Sugai, T. Katakai, Y. Omatsu, T. Iyoda, K. Inaba, T. Nakahata,

34. Takeda, N., M. Arima, N. Tsuruoka, S. Okada, M. Hatano, A. Sakamoto, A. Shimizu, and Y. Yokota. 2003. TH2 dominance and defective development of by guest on September 26, 2021 Y. Kohno, and T. Tokuhisa. 2003. Bcl6 is a transcriptional repressor for the IL-18 a CD8+ dendritic cell subset in Id2-deﬁcient mice. J. Allergy Clin. Immunol. gene. J. Immunol. 171: 426–431. 111: 136–142. 35. Ichii, H., A. Sakamoto, M. Hatano, S. Okada, H. Toyama, S. Taki, M. Arima, 50. Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. J. Liu, Y. Kuroda, and T. Tokuhisa. 2002. Role for Bcl-6 in the generation and main- B. Pulendran, and K. Palucka. 2000. Immunobiology of dendritic cells. Annu. tenance of memory CD8+ T cells. Nat. Immunol. 3: 558–563. Rev. Immunol. 18: 767–811.