Bcl6-independent in vivo development of functional cDC1 supporting tumor rejection

Abstract

The transcriptional repressor Bcl6 has been reported as required for development of a subset of classical dendritic cell (cDCs), called cDC1, which is responsible for cross-presentation. However, mechanisms and in vivo functional analysis have been lacking. We generated a system for conditional deletion of Bcl6 in mouse cDCs. We confirmed the reported in vitro requirement for Bcl6 in cDC1 development and the general role for Bcl6 in cDC development in competitive settings. However, deletion of Bcl6 did not abrogate the in vivo development of cDC1. Instead, Bcl6 deficiency caused only a selective reduction in CD8α expression by cDC1, without affecting XCR1 or CD24 expression. Normal cDC1 development was confirmed in Bcl6^cKO mice by development of XCR1⁺ Zbtb46-GFP⁺ cDC1, by rejection of syngeneic tumors and by priming of tumor-specific CD8 T cells. In summary, Bcl6 regulates a subset of cDC1-specific markers, and is required in vitro, but not in vivo, for cDC1 development.

Introduction

The BTB-ZF (broad-complex, tramtrack and bric-à-brac- zinc finger) family of transcriptional repressors (1) includes Zbtb46, which is expressed in mature cDCs but is not required for their development (2), and Bcl6, which is expressed in and required for a wide range of immune lineages (3–5). Bcl6 was previously reported to be required for cDC development, but not for development of plasmacytoid DCs (pDCs), based on reduced numbers of CD11c⁺ cells in Bcl6^–/– mice (6). That study identified cDC1 and cDC2 subsets by expression of CD8α or CD4, respectively. A second study claimed only that Bcl6 was selectively required for cDC1 development (7), based again on reduced CD8α expression. However, CD4 and CD8α expression are induced by Notch2 signaling during late stages of cDC development and are controlled independently of cDC1 and cDC2 development (8,9). Thus, reduced expression of these markers may result from a limited requirement for Bcl6 by CD4 or CD8α, rather than a general requirement for Bcl6 by the cDC1 or cDC2 lineage. Indeed, similar claims that CIITA was required for cDC development turned out to be due only to reduced expression of major histocompatibility complex class II (MHC-II) genes in Ciita^–/– mice (10). The Zbtb46-GFP reporter mouse, in which GFP expression identifies all cDCs, provides the ability to identify cDCs independently of surface marker expression (2). Thus, we developed a mouse in which Bcl6 expression is specifically inactivated in myeloid cells (referred to here as Bcl6^cKO) and crossed this with the Zbtb46-GFP reporter strain. Here, while we can confirm previously reported in vitro effects of Bcl6 on CD11c⁺ cDC1 development, we find that loss of Bcl6 in vivo neither eliminates Zbtb46-GFP⁺ cDC1 development nor impairs their ability to cross-present tumor antigens to CD8 T cells. These findings indicate that Bcl6 is not required for cDC1 development in vivo, and that its role in vitro and in vivo requires additional study.

Materials and Methods

Mice

WT C57BL6/J, Bcl6^fl (B6.129S(FVB)-Bcl6^tm1.1Dent/J), Csf1r^Cre (C57BL/6-Tg(Csf1r-cre)1Mnz/J ), CMV^cre (B6.C-Tg(CMV-cre)1Cgn/J), and Rosa26^lSlYFP (B6.129X1-Gt(ROSA)26Sor^tm1(EYFP)Cos/J) mice were purchased from The Jackson Laboratory. Zbtb46^gfp/+ and Irf8 +32^–/– mice were previously described (2,11). B6.SJL (B6.SJL-Ptprc^a Pepc^b /BoyJ) mice were obtained from Charles River. Bcl6^−/− mice were obtained by crossing Bcl6^fl/fl mice to CMV^cre mice, generating germline deletion as previously reported(12). All mice were maintained on the C57BL/6 background in the Washington University in St. Louis School of Medicine specific pathogen-free animal facility. Experiments were performed with mice between 6 and 10 weeks of age.

Dendritic cell preparation

Lymphoid and nonlymphoid organ DCs were harvested and prepared as described previously(11). Briefly, spleens and inguinal skin-draining LNs were minced and digested in 5 mL of Iscove’s modified Dulbecco’s media (IMDM) + 10% FCS (cIMDM) with 250 μg/mL collagenase B (Roche) and 30 U/mL DNaseI (Sigma-Aldrich) for 45 min at 37ºC with stirring. After digestion was complete, single cell suspensions from all organs were passed through 70-μm strainers and red blood cells were lysed with ammonium chloride-potassium bicarbonate (ACK) lysis buffer. Cells were subsequently counted with a Vi-CELL analyzer (Beckman Coulter), and 3–5×10⁶ cells were used per antibody staining reaction.

Flow cytometry

Cells were kept at 4ºC while being stained in PBS supplemented with 0.5% BSA and 2mM EDTA in the presence of antibody blocking CD16/32 (clone 2.4G2; BD 553142). All antibodies were used at a 1:200 dilution vol/vol (v/v), unless otherwise indicated.

The following antibodies were used: Brilliant Ultraviolet 395–anti-CD117 (clone 2B8, 1:100 v/v), phycoerythrin(PE)-CF594–anti-CD135 (clone A2F10.1, 1:100 v/v), V500–anti-MHC-II (clone M5/114.15.2), Alexa Fluor 700 or peridinin chlorophyll protein (PerCP)-eFlour 710–anti-CD11b (clone M1/70), Alexa Fluor 700–anti-Ly6C (clone AL-21), Brilliant Violet 421–anti-CD127 (clone SB/199, 1:100 v/v), biotin–anti-CD19 (clone 1D3), Brilliant Violet 510 or PE/Dazzle 594–anti-CD45R (clone RA3–6B2), allophycocyanin(APC)–anti-CD317 (clone eBio927, 1:100 v/v), PE-Cy7–anti-CD24 (clone M1/69), PerCP–eFluor 710–anti-CD172a (clone P84), Brilliant Violet 711–anti-CD4 (clone RM4–5), v450–anti-CD8α (clone 53–6.7), Brilliant Violet 711–anti-CD115 (clone AFS98, 1:100 v/v), PE, APC, or Brilliant Violet 421–anti-XCR1 (clone ZET), Alexa Flour 700 or APC/Cy7–anti-F4/80 (clone BM8, 1:100 v/v), Brilliant Violet 605 or PE–anti-CD45.2 (clone 104), PerCP/Cy5.5 or biotin–anti-Ly6G (clone 1A8), biotin–anti-Ter119 (clone TER-119), biotin–anti-CD105 (clone MJ/718), Brilliant Violent 650–anti-CD45.1 (clone A20),APC–eFluor 780–anti-CD11c (clone N418), PE–anti-CD103 (clone 2E7) and PE-Cy7-anti-TCR β (clone H57–597). Cells were analyzed on a FACSCanto II or FACSAria Fusion flow cytometer (BD), or an Aurora spectral flow cytometer (Cytek), and data were analyzed with FlowJo v10 software (TreeStar).

Isolation and culture of BM progenitor cells and splenic DCs

Bone marrow progenitors and DCs were isolated as described(13,14). For BM sorting experiments, BM was isolated and depleted of CD3-, CD19-, CD105-, Ter119-, and Ly6G- expressing cells by staining with the corresponding biotinylated antibodies followed by depletion with MagniSort Streptavidin Negative Selection Beads (Thermo Fisher). All remaining BM cells were then stained with fluorescent antibodies prior to sorting. MDPs were identified as Lin^–CD117^hiCD135⁺CD115⁺ BM cells; CDPs were Lin^–CD117^intCD135⁺CD115⁺MHC-II^–CD11c^–; pre-cDC1s are Lin^–CD117^intCD135⁺ Zbtb46-GFP^pos, and pre-cDC2s as Lin^–CD117^loCD135⁺ Zbtb46-GFP^pos. For splenic sorting experiments, spleen was isolated and depleted of Ly6G-, Ter119-, CD19-, and CD105-expressing cells. cDC1 were identified as Lin^–CD45R^–CD317^– Zbtb46-GFP^posXCR1⁺ cells, cDC2 were identified as Lin^–CD45R^–CD317^– Zbtb46-GFP^posXCR1^-cells, and pDCs were identified as Lin^-CD45R⁺CD317⁺ cells. Cells were purified on a FACSAria Fusion into IMDM plus 10% FBS with 5% Flt3L conditioned media. Sort purity of >95% was confirmed by post-sort analysis before cells were used for further experiments. For experiments that included Flt3L cultures, whole BM was cultured for 8 d at 37 °C with 5% Flt3L conditioned media.

Expression microarray analysis

RNA was extracted with a RNAqueous-Micro Kit (Ambion) or a NucleoSpin RNA XS Kit (Machery-Nagel), then was amplified with Ovation Pico WTA System (NuGEN) or WT Pico System (Affymetrix) and hybridized to GeneChip Mouse Gene 1.0 ST microarrays (Affymetrix) for 18 h at 45 °C in a GeneChip Hybridization Oven 640. The data were analyzed with the Affymetrix GeneChip Command Console. Microarray expression data were processed using Command Console (Affymetrix, Inc) and the raw (.CEL) files generated were analyzed using Expression Console software with Affymetrix default RMA Gene analysis settings (Affymetrix, Inc). Probe summarization (Robust Multichip Analysis, RMA), quality control analysis, and probe annotation were performed according to recommended guidelines (Expression Console Software, Affymetrix, Inc.). Data were normalized by robust multiarray average summarization and underwent quartile normalization with ArrayStar software (DNASTAR). Unsupervised hierarchical clustering of differentially expressed genes was computed with ArrayStar (DNASTAR) with the Euclidean distance metric and centroid linkage method. Microarrays are available on the GEO database with the SuperSeries accession number https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE135904.

Tumor implantation and tetramer staining

The 1969 regressor fibrosarcoma has been previously described (15). Tumor cells were thawed and propagated in R10 medium (RPMI + 10% FBS + 0.1% 2-ME). On the day of injection, cells were harvested by incubation in 0.05% trypsin-EDTA, washed three times with endotoxin-free PBS, and then 2×10⁶ cells were injected subcutaneously in a total volume of 0.15 mL of PBS into the shaved flanks of mice. Tumor size was measured every three days beginning on day 4 and is presented as the surface area of the tumor (length X width).

Gdp2 H-2Kb biotinylated monomers were purchased from the immunomonitoring core lab at the Bursky Center for Human Immunology and Immunotherapy Programs. The peptide–MHC class I complexes refolded with an ultraviolet-cleavable conditional ligand were prepared as described with modifications(16). Briefly, recombinant the H-2Kb heavy chain and the human β2 microglobulin light chain were produced in Escherichia coli, isolated as inclusion bodies, and dissolved in 4 M urea, 20 mM Tris pH 8.0. MHC Class I refolding reactions were performed by dialyzing a molar ratio of heavy chain:light chain:peptide of 1:1:8 against 10 mM potassium phosphate, pH 7.4 for 48 h. Refolded peptide-MHC class I complexes were captured by ion exchange (HiTrap Q HP, GE), biotinylated, and purified by gel filtration FPLC. Ultraviolet-induced ligand exchange and combinatorial encoding of MHC Class I multimers was performed as described (17). Then, the peptide-MHC multimers were incubated with BV605 and BV710 conjugated streptavidin (SA) at a concentration of 1:5 for 30 minutes at 4°C protected from light in separate reactions. SA labeled tetramers were then incubated with 25 uM D-biotin for 20 min at 4°C protected from light to quench free fluorochrome labeled SA. 3 × 106 splenocytes were incubated with 10% BSA 2mM EDTA phosphate buffered saline (PBS) supplemented with 10% 2.4g2 supernatant for 5 min at 4°C. Fluorochrome-conjugated tetramers were added to the splenocytes at a concentration of 3:50 and incubated at 37°C for 30 minutes. Surface antibodies were added without washing and stained for another 30 minutes at 4°C.

Mixed bone marrow chimeras

Bone marrow cells from donor mice were collected as described above. Recipient CD45.1⁺ mice received a single dose of 1050 rads of whole-body irradiation. For mixed BM chimeras, donor marrow from mice of each genotype was mixed at a ratio of 1:1 before transplantation the next day. Mice were analyzed four to six weeks after transplantation for dendritic cell reconstitution.

Statistical analysis

Data were analyzed using Prism version 8 (GraphPad Software), using unpaired two-tailed Student’s t tests when comparing two groups.

Results

Bcl6 is required for development of cDC1 in vitro in a cell-intrinsic manner

Bcl6 is induced during the specification of pre-cDC1 and pre-cDC2 progenitors from the common dendritic cell progenitor (CDP) (Supplemental Fig. 1A) (13). This increase in Bcl6 in both pre-cDC1 and pre-cDC2 suggested it could be involved in overall cDC specification, as implied by a previous study (6). However, Bcl6^−/− mice suffer from perinatal mortality and do not survive to adulthood (18–20). To test whether Bcl6 is involved in the development of cDCs in adult mice, we therefore selectively inactivated Bcl6 in cDCs by crossing the Bcl6^fl/fl allele, which deletes the BCL6 DNA-binding domain (DBD) upon expression of Cre recombinase (21), to mice expressing the Cre recombinase under control of the Csf1 receptor gene (Csf1r), which targets gene expression in cDCs (22). This cross generated Bcl6 conditional knockout mice (Bcl6^cKO). We confirmed the activity of this Cre by examining activation of the Rosa26^lslYFP reporter (Supplemental Fig. 1B), finding ~90% YFP expression in CDPs, pre-cDC1, pre-cDC2, and splenic cDC1 and cDC2, as expected. We confirmed the deletion of the Bcl6 DBD using qPCR (Supplemental Fig. 1C). Bcl6 expression is controlled partially by autorepression at the Bcl6 locus(23,24); therefore, we confirmed a loss of BCL6 functional activity by upregulation of a Bcl6 sequence outside of the targeted DBD (Supplemental Fig. 1C).

We next found that loss of Bcl6 expression caused elimination of in vitro cDC1 development in Flt3L cultures (Fig. 1A-C), as previously reported (6). Zbtb46-GFP^pos cDCs were greatly decreased in cultures of Bcl6^cKO BM compared to WT BM (Fig. 1A, 1B). CD24⁺ CD172a^– cDC1s were eliminated entirely (Fig. 1C). We also generated mixed BM chimeras reconstituted with equal ratios of WT (CD45.1⁺) and Bcl6^cKO (CD45.2⁺) BM (Fig. 1D-F). We found a severe reduction in Bcl6^cKO CD45.2⁺ cDCs compared with WT CD45.1⁺ cDCs (Fig. 1D, Supplemental Fig. 1D), but normal development of CD11b⁺Ly6C⁺ monocytes and CD45R⁺ CD317⁺ pDCs (Fig. 1E, 1F). These results validate the effective deletion of Bcl6 in the Csf1r^Cre system by confirming previous in vitro and mixed bone marrow chimera results.

FIGURE 1. Bcl6 is required for in vitro cDC development and for cDC progenitor fitness.

(A) Representative FACS analysis of Zbtb46–GFP^posF4/80^– cells pre-gated on single cells from in vitro Flt3L cultures from BM from WT and Bcl6^cKO mice. Numbers indicate percentage of cells in the gate. Data representative of four independent experiments. (B) Analysis in (A) is presented for individual mice. (C) Representative FACS analysis of in vitro Flt3L-derived cDCs pregated on Zbtb46–GFP^pos cells. Numbers indicate percentage of cells in the gate. Data representative of four independent experiments. (D- E) FACS analysis of live splenocytes from chimeras generated with equal mixes of CD45.1⁺ B6.SJL and CD45.2⁺ Bcl6^f/f or Bcl6^cKO BM analyzed 4 wk after lethal radiation and transplant. Shown are representative two-color histograms for CD11c⁺ cells (D) and for monocytes (CD11b⁺Ly6C⁺), cDCs (MHC-II⁺CD11c⁺), cDC1 (CD24⁺CD172a^–), cDC2 (CD24^–CD172a⁺), and pDCs (CD45R⁺CD317⁺) (E). Numbers indicate percentage of cells in the gate. Data are representative of two independent experiments. (F) Contributions of Bcl6^cKO BM to the indicated lineages in chimeras generated in (E) and (F) shown as the ratio of monocyte contribution in the same mouse. Each dot represents a biological replicate from two independent experiments. ns, not significant. **P < 0.01, ****P<0.0001.

Bcl6 controls in vivo expression of CD11c, but not expression of Zbtb46

We next examined in vivo cDC development in Bcl6^cKO mice (Fig. 2). First, using CD11c⁺ expression to define cDC identity, we confirmed the previously described reduction in splenic cDCs in Bcl6^cKO mice compared with WT mice (Fig. 2A, 2B). pDC development was normal (Fig. 2C, 2D).The number of CD11c⁺ DCs was normal in mesenteric (mLN) and peripheral (pLN) lymph nodes, but there were fewer cells in the migratory and resident DC gates (Supplemental Figure 2A-D). We also confirmed previously described reductions in CD103⁺ cDC1s within the mesenteric lymph node and in CD8α⁺ cDC1s in the peripheral lymph nodes (Supplemental Figure 2E, 2F). However, using Zbtb46-GFP to define cDC identity (2,10), we did not observe any significant reduction in cDC numbers within Bcl6^cKO spleens and instead documented a reduction in CD11c expression on Zbtb46-GFP⁺ cDCs(Fig. 2E-H). We also did not observe any significant reduction in Zbtb46-GFP⁺ DCs in the mLN or pLN, though we did observe a specific reduction in Zbtb46-GFP⁺ resident DCs within the mLN (Supplemental Fig. 3A-D). These results suggested that Bcl6 might selectively regulate certain markers of cDCs, rather than controlling the development of the entire lineage, similar to how Ciita selectively regulates MHCII expression, rather than cDC development (10).

FIGURE 2. Bcl6^cKO mice show reduced expression of CD11c but retain expression of Zbtb46-GFP.

(A) Shown are histograms for CD11c expression of WT and Bcl6^cKO splenocytes. Numbers indicate percentage of cells. Data are representative of four independent experiments. (B) Analysis in (A) is presented for individual mice. (C) Representative FACS analysis of CD45R⁺CD317⁺ pDCs from WT and Bcl6^cKO mice, pre-gated on single cells. Numbers indicate percentage of cells. Data are representative of four independent experiments. (D) Analysis in (C) is presented for individual mice. (E) Representative FACS histograms for Zbtb46–GFP^pos splenocytes from WT and Bcl6^cKO mice. Numbers indicate percentage of cells. Data are representative of four independent experiments. (F) Analysis in (E) is presented for individual mice. (G) Representative FACS analysis for CD11c expression on Zbtb46-GFP⁺ cDCs in WT and Bcl6^cKO spleens. Data are representative of four independent experiments. (H) Analysis in (G) is presented for individual mice. Each dot represents a biological replicate from four independent experiments. ns, not significant. **P < 0.01, ****P<0.0001.

Therefore, we next directly compared CD8α and CD4 expression with the expression of other markers that define cDC subsets (Fig. 3). First, we confirmed that the number of CD4⁺ CD11c⁺ cells is reduced in Bcl6^cKO mice (Fig. 3A, 3B). Similarly, CD8α expression was reduced when cDC1s were identified using Zbtb46-GFP expression (Fig. 3C, 3D). However, cDC1s can also be identified independently of CD8α expression, as either XCR1⁺ cDCs (25) or as CD24⁺CD172a^- cDCs (26). Using these markers alongside Zbtb46-GFP to identify cDC1, we found that Bcl6^cKO mice contain a normal number of cDC1 compared with WT mice (Fig. 3C, 3D). Similarly, cDC2 can be identified as Zbtb46-GFP⁺ CD11b⁺ CD172a⁺. We found a reduction in Zbtb46-GFP⁺ CD172α⁺ cells, but normal numbers of Zbtb46-GFP⁺ CD4⁺ and Zbtb46-GFP⁺ CD11b⁺ cells (Fig. 3E, 3F). Further, we found that in Bcl6^cKO mice, there were fewer Zbtb46-GFP⁺ CD24⁺ DC1s which co-expressed CD8α, and fewer Zbtb46-GFP⁺ CD4⁺ DC2s which co-expressed CD172α (Fig. 3G, 3H).These results suggest that Bcl6 is not required for the development of the cDC lineage, and instead only alters certain surface marker expression.

FIGURE 3. Bcl6^cKO mice retain normal numbers of cDC1 and cDC2.

(A) Representative FACS analysis of CD8α⁺ and CD4⁺ cDCs pregated on CD11c⁺ splenocytes. Numbers indicate percentage of cells in the gate. Data are representative of four independent experiments. (B) Analysis in (A) is presented for individual mice. (C) Spleens from WT and Bcl6^cKO mice were analyzed for CD8α⁺Zbtb46–GFP^pos (left), CD24⁺Zbtb46–GFP^pos (middle) and XCR1⁺Zbtb46–GFP^pos (right) cells. Numbers indicate the percentage of cells in the gate. Data are representative of four independent experiments. (D) Analysis in (C) is presented for individual mice. (E) Spleens from WT and Bcl6^cKO mice were analyzed for CD4⁺Zbtb46–GFP^pos (left), CD11b⁺Zbtb46–GFP^pos (middle) and CD172a⁺Zbtb46–GFP^pos (right) cells. Numbers indicate the percentage of cells in the gate. Data are representative of four independent experiments. (F) Analysis in (E) is presented for individual mice. (G) CD24⁺ Zbtb46-GFP^pos cells were analyzed for expression of CD8α. Analysis is presented for individual mice. Data are representative of four independent experiments. (H) CD4⁺ Zbtb46-GFP^pos cells were analyzed for expression of CD172α. Analysis is presented for individual mice. Data are representative of four independent experiments. Each dot represents a biological replicate from four independent experiments. ns, not significant. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Bcl6 is not required for the in vivo development of functional cDC1 that support tumor rejection

To determine if Bcl6 was required for cDC specification or commitment in the BM, we analyzed all known progenitor populations that give rise to cDC subsets in Bcl6^cKO. We found that the macrophage/dendritic cell progenitor (MDP), CDP, pre-cDC1, and pre-cDC2 are found in normal numbers in Bcl6^cKO mice compared with WT mice (Fig. 4) (13,27,28). Thus, despite the induction of Bcl6 in the pre-cDC1 or pre-cDC2 from the CDP or the BCL6 present in the BM cells (5), our data indicate that on developmental grounds, Bcl6 is not required for early cDC specification or commitment.

FIGURE 4. Bcl6 is not required for cDC specification or commitment in the bone marrow.

(A) BM from WT and Bcl6^cKO was analyzed for Lin^–CD127^–CD117^hiCD135⁺CD115⁺ MDPs (green box). Numbers indicate percentage of cells in the gate. Data representative of three independent experiments. Lineage (Lin) includes CD3, CD19, Ter119, CD105, and Ly6G. (B) BM from WT and Bcl6^cKO was analyzed for Lin^–CD127^–CD117^intCD135⁺CD115⁺MHC-II^–CD11c^– CDPs (blue box). Numbers indicate percentage of cells in the gate. Data representative of three independent experiments. (C) BM from WT and Bcl6^cKO was analyzed for Lin^–CD127^–CD135⁺CD117^intZbtb46–GFP^pos pre-cDC1 (pink box) and Lin^–CD127^–CD135⁺CD117^lo Zbtb46–GFP^pos pre-cDC2 (purple box). Numbers indicate percentage of cells in the gate. Data representative of three independent experiments. (D) Data in (A), (B), and (C) presented as percentage of total Lin^– BM for individual mice.

We next performed microarray analysis on splenic CD45R⁺ CD317⁺ pDC, XCR1⁺ Zbtb46-GFP⁺ cDC1, and XCR1^- Zbtb46-GFP⁺ cDC2 from WT and Bcl6^cKO mice (Supplemental Fig. 4A, 4B). We found few changes in gene expression within Bcl6^cKO pDCs. We found a number of genes to be upregulated within Bcl6^cKO cDC1s and cDC2s, in keeping with the known role of Bcl6 as a transcriptional repressor (Supplemental Fig. 4C). We observed no significant change in signal for Bcl6, as loss of transcripts within the Bcl6 DBD was accompanied by an increase in transcripts in regions outside the Bcl6 DBD, driven by loss of BCL6 autorepression (Supplemental Figure 1C). We also found a reduction in expression of CD8a transcripts in Bcl6^cKO cDC1s, but no significant change in expression of Itgax (which codes for CD11c) suggesting at least some of the changes found in Bcl6^cKO cDC1s are owing to posttranscriptional effects. These results suggest that Bcl6 is not required for development of any DC lineage.

We next tested the functionality of Bcl6^cKO cDC1. We evaluated Bcl6^cKO mice for growth of fibrosarcomas whose rejection is known to be dependent on cDC1 (29,30). We find that the 1969 fibrosarcoma (15) is rapidly rejected by WT mice, but not by Irf8 +32^–/– mice that lack cDC1 (11) (Fig. 5A). However, Bcl6^cKO mice are fully able to reject the 1969 fibrosarcoma, consistent with their harboring cDC1. Further, CD8 T cells reactive to an antigen expressed by this tumor are selectively expanded in both WT and Bcl6^cKO mice, but not in naïve mice or Irf8 +32^–/– mice, again consistent with the presence of functional cDC1 in Bcl6^cKO mice (Fig. 5B, 5C).

FIGURE 5. Bcl6-deficient cDC1 are functional and support tumor rejection.

(A) Bcl6^f/f (left), Bcl6^cKO(middle), and Irf8 +32^–/– (right) mice were injected with 1969 regressor fibrosarcoma and tumors were measured for 16 d. Data representative of two independent experiments. (B) Representative FACS analysis depicting percentage of tetramer made from CD8α⁺TCRβ⁺ cells from tumor-laden Bcl6^f/f, Bcl6^cKO, and Irf8 +32^–/– mice at 10 d after injection. FACS analysis is pre-gated on single cells. Data representative of two independent experiments. (C) Data in (B) as presented as individual mice. ns, not significant.

Discussion

Our results indicate that the role of Bcl6 in cDC and cDC1 development has been poorly characterized. An initial study of Bcl6^–/– mice concluded that cDCs were reduced based on fewer CD11c⁺ cells (6). A second study of Bcl6^–/– mice concluded that cDC1s were reduced based on CD8α and CD103 expression, but not on other parameters nor on functional analysis (7). We find that CD11c expression is reduced in the absence of Bcl6, but that cDC lineages continue to develop and express other markers of cDCs, including Zbtb46-GFP. A possible explanation for these differences is that previous studies used Bcl6^KO mice, which experience pleiotropic defects and premature death that may affect cDC development through cell-extrinsic pathways(18). We have also restricted our studies to adult mice, leaving open the possibility that Bcl6 has an effect upon cDC differentiation in juvenile but not adult mice. While we confirm the in vivo reduction of CD8α and CD103 in Bcl6^cKO mice, we find normal cDC1 development based on other cDC1-specific markers, including XCR1 and CD24, and on functional analysis of a cDC1-dependent tumor rejection.

Finally, we have revealed that cDC1 development is Bcl6-dependent in vitro, but not in vivo. Discrepancies between in vitro and in vivo requirements of transcription factors are recognized within the area of cDC development. For example, deletion of the E protein E2A abrogates in vitro cDC1 development, but not in vivo cDC1 development (11). This difference may be due to selective in vivo compensation by the E protein E2–2, which may not take place in vitro. Similarly, the in vivo requirement for Nfil3 in cDC1 development is not exhibited in vitro (31). In summary, our results indicate that Bcl6 regulates a subset of cDC1-specific markers, such as CD8α, but is not required in vivo for cDC1 lineage development or cross-priming function.

Key points.

• Bcl6 affects expression of certain cDC markers such as CD11c and CD8α

• Zbtb46⁺ DC1s develop in Bcl6^cKO mice

• Bcl6^cKO DC1s are functional with respect to tumor rejection