batf3-dependent genes control tumor rejection …...research article batf3-dependent genes control...

Research Article

Batf3-Dependent Genes Control Tumor RejectionInduced by Dendritic Cells Independently ofCross-PresentationDerek J. Theisen1, Stephen T. Ferris1, Carlos G. Brise~no1, Nicole Kretzer1,Arifumi Iwata1, Kenneth M. Murphy1,2, and Theresa L. Murphy1

Abstract

The BATF3-dependent cDC1 lineage of conventional den-dritic cells (cDC) is required for rejection of immunogenicsarcomas and for rejection of progressive sarcomas duringcheckpoint blockade therapy. One unique function of thecDC1 lineage is the efficient cross-presentation of tumor-derived neoantigens to CD8þ T cells, but it is not clearthat this is the only unique function of cDC1 required fortumor rejection. We previously showed that BATF3 func-tions during cDC1 lineage commitment to maintain IRF8expression in the specified cDC1 progenitor. However, sincecDC1 progenitors do not develop into mature cDC1sin Batf3�/� mice, it is still unclear whether BATF3 hasadditional functions in mature cDC1 cells. A transgenicIrf8-Venus reporter allele increases IRF8 protein concentra-

tion sufficiently to allow autonomous cDC1 developmentin spleens of Batf3�/� mice. These restored Batf3�/� cDC1sare transcriptionally similar to control wild-type cDC1s buthave reduced expression of a restricted set of cDC1-specificgenes. Restored Batf3�/� cDC1s are able to cross-presentcell-associated antigens both in vitro and in vivo. However,Batf3�/� cDC1 exhibit altered characteristics in vivo andare unable to mediate tumor rejection. These results showthat BATF3, in addition to regulating Irf8 expression tostabilize cDC1 lineage commitment, also controls expres-sion of a small set of genes required for cDC1-mediatedtumor rejection. These BATF3-regulated genes may be usefultargets in immunotherapies aimed at promoting tumorrejection.

IntroductionConventional dendritic cells (cDC) develop as two major

lineages that are dependent on distinct transcriptional programsfor their development (1). The cDC1 lineage is dependent on thetranscription factors IRF8 and BATF3 for development andexpresses certain unique markers such as CD8a, CD103, andXCR1 in various tissues (1). BATF3-dependent cDC1s are special-ized for antigen cross-presentation and are required for antiviraland antitumor CD8þ T-cell responses (2–4). In particular, BATF3-dependent cDC1s are required for rejection of immunogenicsyngeneic fibrosarcomas (2), and type I interferon signalingsupports this capacity (5, 6). BATF3-depdendent cDC1s are alsorequired for T-cell priming in response to DNA vaccines (7) andtheir abundance in humans tumors correlated with improved

tumor regression (8, 9). The ability of checkpoint blockadeto mediate antitumor responses against progressively growingsarcomas was shown to also require BATF3-dependent cDC1s(10–12). However, despite the clearly important role of cDC1cells in antitumor immunity, much remains unclear about howthey survey tissues and initiate immune responses.

IRF8 and BATF3 play distinct roles in cDC1 development,which proceeds through distinct specification and commitmentstages (13). IRF8 is required for development of a cDC1-specifiedbone marrow (BM) progenitor, which is missing in Irf8�/� mice.In contrast, this specified progenitor develops in Batf3�/� mice,but fails to commit to mature cDC1s and instead diverts to thecDC2 lineage due to the inability to sustain the normal high Irf8expression (13). In this cDC1-specified progenitor, BATF3 andIRF8 cooperate in binding to an enhancer containing severalAP-1/IRF consensus elements (AICE; ref. 14) that functions tosustain Irf8 autoactivation initiated earlier in development (13).Thus, at least one function of BATF3 is exerted in the commitmentstage of cDC1 development.

Because there is currently no system for conditional deletionof Batf3, it has been difficult to determine whether BATF3 alsofunctions in cDC1 cells during immune responses. The normalrequirement for Batf3 for cDC1 development can be bypassedunder some conditions, although themechanism is not complete-ly understood. Splenic CD8aþ cDC1 were restored in Batf3�/�

mice during infection by Mycobacteria tuberculosis (15) or byadministration of IL12, and this restoration was blocked by invivo neutralization by IFNg (15). Molecularly, Batf can compen-sate for Batf3 in cDC1 development (15), and cDC1 developmentinduced by IL12 in Batf3�/�mice was reduced in Batf�/� Batf3�/�

mice, suggesting some role for Batf in the mechanism of

1Department of Pathology and Immunology, Washington University in St. Louis,School of Medicine, St. Louis, Missouri. 2Howard Hughes Medical Institute,Washington University in St. Louis, School of Medicine, St. Louis, Missouri.

Note: Supplementary data for this article are available at Cancer ImmunologyResearch Online (http://cancerimmunolres.aacrjournals.org/).

Current address for C.G. Brise~no: Amgen, South San Francisco, California;current address for N. Kretzer: University of Washington School of Medicine,Seattle, Washington; and current address for A. Iwata: Chiba University, Chiba,Japan.

Corresponding Author: Theresa L. Murphy, Washington University School ofMedicine, St. Louis, MO 63110. Phone: 314-747-3175; Fax: 314-747-4888; E-mail:[email protected]

doi: 10.1158/2326-6066.CIR-18-0138

�2018 American Association for Cancer Research.

CancerImmunologyResearch

www.aacrjournals.org 29

on June 15, 2020. © 2019 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

Published OnlineFirst November 27, 2018; DOI: 10.1158/2326-6066.CIR-18-0138

http://crossmark.crossref.org/dialog/?doi=10.1158/2326-6066.CIR-18-0138&domain=pdf&date_stamp=2018-12-19

http://cancerimmunolres.aacrjournals.org/

restoration (15). cDC1 development is transiently restored aftertransfer of Batf3�/� BM into irradiated recipients (16), althoughthe basis for this effect is unclear. In addition,Batf3�/�mice on theC57BL/6 genetic background frequently retain a population ofcells resembling cDC1 in skin draining lymph nodes, but not inthe spleen or other tissues (15). The basis for this strain- andtissue-specific phenomenon has not been established, butmay bespecific to microbiota, as it is not observed in all colonies ofC57BL/6 Batf3�/� mice.

Another system for BATF3 compensation is based on theparticular molecular mechanism of cDC1 development (13).Crossing a transgenic Irf8VENUS reporter strain (17) with Batf3�/�

mice also restores cDC1 development (13). In this case, cDC1development appears to result from the increased IRF8 expressionarising due to the three intact copies of IRF8 present in the BACreporter transgenes, leading to autonomous Irf8 autoactivation(13). In the present study, we used this Irf8VENUS reporter systemto allow an examination of cDC1 cells that develop and maintainIRF8 expression in the absence of BATF3 for their capacity tofunction in cross-presentation and tumor rejection, and to identifytranscriptional gene targets requiring BATF3.

Materials and MethodsMice

Wild-type and Batf3�/� mice on C57BL/6 background werecrossed to IRF8VENUS (17) as described (13). Experiments usingBatf3�/� Batf�/� mice used a mixed 129S6/SvEV/C57Bl/6 back-ground (18). C57BL/6-Tg(TcraTcrb)1100Mjb/J (OT-1) micewere purchased from The Jackson Laboratories. MHCI KO mice(Kb�/�Db�/�b2m�/�; TKO) were a gift from Herbert W. Virginand Ted Hansen, Washington University, St. Louis (19). Miceharboring a conditional allele of Itga8 (Itga8F/F) were obtainedfrom The Jackson Laboratories as B6.129S6-Itga8tm1.1Rdav/Jand crossed to CD11c-Cre (20). All mice were maintained in aspecific pathogen-free animal facility following institutionalguidelines and with protocols approved by the Animal StudiesCommittee atWashingtonUniversity in St. Louis, an InstitutionalAnimal Care and Use Committee. Experiments were performedwith sex-matchedmice 6–20weeks of agewithout randomizationor blinding.

Antibodies and flow cytometryCells were stained at 4�C in the presence of Fc Block (2.4G2; Bio

X Cell) in flow cytometry buffer (0.5% BSA/PBS). Antibodies tothe following proteins were used: from Becton Dickinson (BD):MHCII- IA/IE (M5/114 15.2, CD4 (RM4-5), andCD11b (M1/70);from BioLegend: XCR1 (ZET), CD24 (M1/65), CD326 (G8.8),TCR-Va2 (B20.1), and B220 (RA3-6B2); from Tonbo Biosciences:CD11c (N418) and CD45.1 (A20); from eBioscience: CD172a,CD45 (30F-11), and CD44 (IM7). Cells were analyzed on aFACSCanto II or FACSAria Fusion flow cytometer (BD) and datawere analyzed with FlowJo software (TreeStar). Unstained cells orgenetic controls that were missing populations of interest wereused as negative controls for antibody staining.

Tissue preparationMinced spleens, skin draining lymph nodes (pooled inguinal,

cervical, and brachial), and tumors, harvested 9–10 days aftertransplantation, were digested in collagenase B (0.25 mg/mL)and DNAse1 (30 u/mL) in complete IMDM (Iscove's modified

Dulbecco'smediumwith10%FCS, 2ME, penicillin/streptomycin,NEAA, and glutamine) for 40 minutes at 37�C with stirring andsubjected to ACK lysis. Prior to sorting, spleen cells were enrichedfor CD11cþ cells (Miltenyi). Earswere split into dorsal and ventralhalves prior tomincing anddigestion in Liberase (0.26 u/mL) andDNAse1 (30 u/mL) in complete IMDM for 40 minutes at 37�Cwith stirring. After addition of 5mmol/L EDTA, suspensions werefiltered through 70-mm nylon mesh, pelleted, subjected to ACKlysis, washed, and used for flow cytometry analysis.

Expression microarray analysiscDC1 (XCR1þ CD24þ CD172a� MHCIIþ CD11cþ B220�) or

(CD24þ CD172a� MHCIIþ CD11cþ B220�) were sorted fromspleens, and migratory (MHCIIhi CD11cint) and resident(MHCIIint CD11chi) cDC1s (CD326lowCD24þ CD172a�

B220�) were sorted from skin draining lymph nodes. Total RNAfrom spleen DCs was extracted using RNAqueous-Micro Kit(Ambion), amplified with the Ovation Pico WTA System(NuGEN) and hybridized to GeneChip Mouse Gene 2.0 STmicroarrays (Affymetrix). Total RNA from lymph node DCs wasextracted using Nucleo-spin RNA XS kit (Macherey-Nagel),amplified with GeneChip WT Pico Kit (Applied Biosystems)and hybridized to GeneChip Mouse Gene 1.0 ST microarrays(Affymetrix). Data were normalized by robust multiarray aver-age summarization and quartile normalization with ArrayStarsoftware (DNASTAR).

Data depositionGene-expression microarray data have been deposited in the

Gene-Expression Omnibus (accession no. GSE111034).

Tumor cell linesThe MCA-induced fibrosarcoma 1969 was a gift from Robert

Schreiber, Washington University School of Medicine, in 2014. Itwas generated in a female C57BL/6Rag2�/�mouse, was tested formycoplasma, and was banked at low passage as previouslydescribed (5, 21). Tumor cells derived from frozen stocks werepropagated for 1 week with one intervening passage in vitro inRPMI media supplemented with 10% FCS (HyClone), werewashed three times with PBS, resuspended at a density of6.67 � 106 cells/mL in endotoxin-free PBS, and then 150 mL wasinjected subcutaneously into the flanks of recipient mice. Cellswere not reauthenticated in the past year. Tumor growth wasmeasured with a caliper and expressed as the average of two per-pendicular diameters. An immunogenic fibrosarcoma expressingmembrane ovalbumin was generated from the MCA-inducedprogressor fibrosarcoma 1956, also a gift from Robert Schreiberin 2014. An mOVA fragment pCI-neo-mOVA (Addgene #25099)was ligated into MSCV-IRES-Thy1.1 vector (22) to generateMSCV-mOVA-IRES-Thy1.1. 1956 tumor cells retrovirally trans-duced with this vector were sorted for expression of Thy1.1, andsurface OVA expression was validated using flow cytometry(Millipore AB1225).

Cross-presentation assaysCross-presentation assays were as described (23, 24). Briefly,

MHCI-deficient splenocytes were ACK-lysed, loaded with oval-bumin (Worthington) in hypertonic medium (0.5 M sucrose,10% w/v polyethylene glycol 1,000, 10-mm Hepes, RPMI 1640pH 7.2) and irradiated (13.5Gy). OT-1 T cells (CD4� B220�

CD11c� CD8aþ Va2þ) were sorted from ACK-lysed, B220-macs

Theisen et al.

Cancer Immunol Res; 7(1) January 2019 Cancer Immunology Research30




depleted OT-1 splenocytes, and labeled with CFSE. For in vivocross-presentation, mice were injected intravenously (i.v.) with500,000 CFSE-labeled OT-1 T cells and on the next day withPBS or varying numbers of ovalbumin-loaded MHCI-deficientsplenocytes. Three days later, spleens and inguinal lymphnodes were harvested for analysis of CFSE dilution by flowcytometry. For in vitro cross-presentation, cDC1 (CD24þ

CD172a� CD11cþ MHCIIþ B220�) were sorted from spleno-cytes that had been positively enriched for CD11cþ cells(Miltenyi). Sorted cDC1 were placed in 96-well round bottomplates with ovalbumin-loaded MHCI-deficient splenocytes orheat killed Listeria monocytogenes expressing ovalbuminHKLM-ova (24) and CFSE-labeled OT-1 T cells in Iscove's MEMat 37�C in a CO2 incubator. Alternatively, migratory andresident DCs were sorted from mice 4 or 7 days after subcu-taneous implantation of 1956-ova tumor cells and were placedin 96-well round bottom plates with CFSE-labeled OT-1 T cells.CFSE dilution was analyzed on day 3.

Statistical analysisAll statistical analyses were performed using Prism (GraphPad

Software). One-way analysis of variance (ANOVA) was usedto compare means of population percentages between miceusing Sidak or Holm–Sidak multiple comparisons test with acutoff of 0.05.

ResultsTransgenic IRF8 overexpression eliminates dependence ofcDC1s on Batf3 for development

Transgenic Irf8VENUS reporter mice possess three cointegratedcopies of a phage artificial chromosome containing a 130-kb Irf8genomic regionwith an internal ribosome entry site and sequence

encoding the yellow fluorescent protein VENUS into the Irf8 30

untranslated region (17). Thus, mice with one transgenic reporterallele have a total of five functional Irf8 loci, resulting in increasedIRF8 expression that was approximately 2-fold higher inIrf8VENUSþ cDC1 cells compared with Irf8VENUS� cDC1 cells(13). Despite this slight IRF8 overexpression, expression of theVenus reporter is similar to endogenous Irf8 expression in severalways, such as being highly expressed in cDC1 and pDC lineages,and being reduced in cDC2 cells to the low expression typical ofIrf8VENUS� cDC2 (13).

We have previously shown that Batf3�/� Irf8VENUSþ mice hadnormal cDC1 development in contrast to Batf3�/� Irf8VENUS�

mice, which have impaired cDC1 development (13). Conceiv-ably, this effect of compensation could be dependent on endog-enous expression ofBatf inDCs, in themanner of IL12-dependentcompensationobservedpreviously (15). To address this,we askedwhether restoration of cDC1in Batf3�/� Irf8VENUSþ mice was dueto compensation by endogenous Batf (Fig. 1). In Batf�/� Batf3�/�

Irf8VENUS� mice that lack both Batf and Batf3, CD24þ CD172a�

cDC1s were reduced by 94% in spleen compared with wild-type(WT) Irf8VENUS� mice (Fig. 1A). This cDC1 population wasrestored by introduction of the Irf8VENUS transgene in Batf�/�

Batf3�/� Irf8VENUSþ mice to 60% of that in WT mice (Fig. 1A andB). Similarly, using XCR1 instead of CD24 to identify the cDC1population (25), cDC1s were restored in spleens of both Batf3�/�

Irf8VENUSþmice and doubly deficient Batf�/� Batf3�/� Irf8VENUSþ,and percentages of these were also reduced by 40% comparedwith WT (Fig. 1C and D). These results indicate that cDC1restoration in Batf3�/� Irf8VENUSþ mice bypasses dependence oneither Batf3 or Batf, unlike restoration induced by IL12 observedpreviously (15). Further, these results suggest that in spleencDC1s, the expression of XCR1 is not completely dependent onBatf3 or Batf.

11 2.6

1212

0 102 103 104 105

0

102

103

104

105

0.7

12

Irf8V

EN

US

–Irf

8VE

NU

S+

WT Batf3−/− Batf +/+

CD172a

CD

24

0

10

20

30**** *

CD

24+ (

% o

f cD

Cs)

XC

R1+ (

% o

f cD

Cs)

9.7

11 0.6

8.4

0.1

9.4

XC

R1

0102

103

104

105

0 102 103 104 105

***

0

5

10

15

20

A B

C D

Batf3−/− Batf −/−

CD172a

WT Batf3-/- Batf+/+ Batf3-/- Batf-/-

Irf8V

EN

US

–Irf

8VE

NU

S+

Irf8Venus − + − + − + Batf3 +/+ +/+ −/− −/− −/− −/− Batf +/+ +/+ +/+ +/+ −/− −/−

Irf8Venus − + − + − + Batf3 +/+ +/+ −/− −/− −/− −/− Batf +/+ +/+ +/+ +/+ −/− −/−

Figure 1.

Irf8VENUS bypasses the need for bothBatf3 and Batf in cDC1 development inspleen. A, Flow cytometry analysis ofsplenocytes of the indicated genotypesgated on live singlets that were B220�

CD11cþ MHCIIþ. Numbers indicate thepercentage of cells in the CD24þ

CD172a� gate. B, Compiled flow-cytometric data for samples analyzed asin A. Each symbol represents a singlemouse. One-wayANOVA, Sidakmultiplecomparisons test; adjusted P value:� , 0.0341; �� , <0.0001. C, Flowcytometry analysis of splenocytesgated on live singlets that were B220�

CD11cþ MHCIIþ. Numbers indicate thepercentage of cells in the XCR1þ

CD172a� gate. D, Compiled flow-cytometric data for samples analyzed asin C. Each symbol represents a singlemouse. One-way ANOVA, Holm–Sidakmultiple comparisons test, alpha 0.05;adjusted P value: � , 0.0386; �� , 0.0021.

Batf3-Dependent Genes in DC1 Control Tumor Rejection

www.aacrjournals.org Cancer Immunol Res; 7(1) January 2019 31




Batf3�/� cDC1 can cross-present cell-associated antigenWe analyzed the ability of Batf3�/� cDC1s restored by the

Irf8VENUS transgene to cross-present cell-associated antigen in vitro(Fig. 2). cDC1 purified from WT spleens, that were eitherIrf8VENUS� or Irf8VENUSþ, were able to cross-present to OT-1 Tcells (Fig. 2A and B), as expected. cDC1 purified from spleens ofBatf3�/� Irf8VENUSþ (Fig. 2A and B) and Batf�/� Batf3�/�

Irf8VENUSþ mice (Fig. 2C and D) cross-presented cell-associatedovalbumin as efficiently as WT cDC1 in vitro. As a control, cDC2cells purified from either WT or Batf�/� Batf3�/� Irf8VENUSþ micewere unable to cross-present cell-associated antigen (Fig. 2C andD), as expected. We next compared the efficacy of cross-presen-tation using a dose titration of HKLM-ova (Supplementary Fig.S1). cDC1s purified from spleens of Batf3�/� Irf8VENUSþcDC1swere able to cross-present HKLM-ova to OT-1 T cells as well asthose purified from WT spleens that were either Irf8VENUS� or

Irf8VENUSþ (Supplementary Fig. S1A). As a control, cDC2s did notcross-present HKLM-ova, as expected (Supplementary Fig. S1B).Thus, cDC1 cells that express IRF8 but lack Batf3 and Batf proteinsare capable of in vitro cross-presentation.

We have previously shown that cDC1s restored by IL12treatment of Batf3�/� mice were capable of cross-presentationin vivo (15). We next tested whether Batf3�/� Irf8VENUSþ micecould cross-present cell-associated antigen in vivo. TransferredOT-1 CD8 T cells proliferated in spleens and inguinal lymphnodes of WT mice, either with or without Irf8VENUS (Fig. 2Eand F). No OT-1 proliferation was observed in spleensor lymph nodes of Batf3�/� mice without Irf8VENUS, asexpected. However, OT-1 proliferation was observed inspleens and lymph nodes of Batf3�/� Irf8VENUSþ and wasequivalent to that in WT mice. These results show that cDC1sdo not require Batf3 for in vivo cross-presentation function. We

0.6

0 102 103 104 105

0

102

103

104

105

82

2.8

83

CFSE

CD

44

0 2.8

89

A B

10

0

2

103

104

105

70 2.4

5.6

90

1.5

4.4

CFSE

CD

44

100

DC1 DC2

2.74.3

0 102 103 104 105

C D

100

WTIrf8VENUS–

Irf8VENUS–

Irf8VENUS+WT

Irf8VENUS+

Irf8VENUS+Batf3 −/−

Irf8VENUS–WT

0

0

20

40

60

80

100

WT WT Batf3 −/−

0 50 100

DC1 DC2

050100

100

50

0Pro

lifer

atio

n (%

of W

T D

C1)

Pro

lifer

atio

n(%

CFS

E d

ilute

d in

CD

44+ )

Irf8VENUS+

WTIrf8VENUS–

Batf3−/− Batf −/−

Irf8VENUS+Batf3 −/− Batf −/−

Irf8VENUS+

Batf3 −/− Batf −/−

Irf8VENUS+

71

0 102 103 104 105

0

102

103

104

105

9.8

61

78

Batf3−/−

CFSE

CD

44 37

0 102 103 104 105

0

102

103

104

105

9.6

24

49

CFSE

CD

44

WT

E FBatf3−/− WT

Irf8V

EN

US

–Irf

8VE

NU

S+

Spleen SDLN

Irf8V

EN

US

–Irf

8VE

NU

S+

Figure 2.

Cross-presentation by IRF8VENUS

restored cDC1s does not require Batf3 orBatf.A andB,CFSEdilutionwithinOT-1 Tcells on day 3 after in vitro exposure tosorted splenic cDCs1 from mice of theindicated genotype and no antigen (0),or with 50,000 (50), or 100,000 (100)irradiated ovalbumin-loaded MHCI�/�

splenocytes. Numbers indicate thepercentage of CD44þ OT-1 cells withdiluted CFSE. OT-1 cells were identifiedas live singlet CD45.1þVa2þ CD3þ CD8þ.A, Representative flow cytometryanalysis. B, Each circle represents anindividual mouse. C andD, CFSE dilutionwithin OT-1 T cells on day 3 after in vitroexposure to sorted splenic cDC1s orcDC2s from mice of the indicatedgenotype and no antigen (0) or 100,000(100) irradiated ovalbumin-loadedMHCI�/� splenocytes. Numbers indicatethe percentage of CD44þOT-1 cells withdiluted CFSE. C, Representative flowcytometry analysis. D, Proliferation isexpressed as a percentage of OT-1 T-cellproliferation after exposure toWT cDC1sand 100,000 irradiated ovalbumin-loadedMHCI�/� splenocytes. Each circlewithin matching genotypes for cDC1sand cDC2s represents an individualmouse. E and F, Flow cytometry analysisof CFSE-labeled OT-1 T cells fromspleens (E) or inguinal lymph nodes(SDLN, F) after transfer into mice of theindicated genotype. CFSE dilution wasanalyzed on day 3 after injection ofCFSE-labeled OT-1 T cells on day �1followed by irradiated ovalbumin-loaded splenocytes on day 0. Numbersindicate the percentage of CD44þ OT-1cells with diluted CFSE. OT-1 cells wereidentified as live singlet CD45.1þ Va2þ

CD3þ CD8aþ. Batf3þ/þIRF8VENUS�,n¼ 2;Batf3þ/þIRF8VENUSþ,n¼ 1;Batf3�/�

IRF8VENUS�, n ¼ 2; Batf3�/� IRF8VENUSþ,n ¼ 4.

Theisen et al.





reported that cross-presentation of cell-associated antigens bymonocyte-derived DCs (mo-DC) is also independent of Batf3and Batf (26). Similarly, cross-presentation in cDC1s is notdependent on Batf3 itself, although the development of thecDC1 lineage is.

Batf3�/� cDC1 do not mediate fibrosarcoma rejection in vivoPreviously, we showed that transplanted immunogenic fibro-

sarcomas cannot be rejected by Batf3�/� mice, but that IL12-treated Batf3�/� mice reacquire a population of cDC1 cells andcan also mount antitumor responses (2, 15). However, usingIL12-induced cDC1 restoration, it is possible that IL12 mightindependently augment tumor rejection by actions on targets cellsother than cDC1s in Batf3�/� mice. Thus, there was a need toevaluate the intrinsic capacity of the Batf3�/� cDC1s to mediatetumor rejection in another system. Therefore, we next askedwhether Batf3�/� Irf8VENUSþ mice could also reject tumors, usinga C57BL/6 regressor fibrosarcoma, 1969, that is rejected by WTmice (refs. 5, 21; Fig. 3).

WT mice were able to reject the 1969 fibrosarcoma indepen-dently of being Irf8VENUS� or Irf8VENUSþ (Fig. 3A). By contrast,tumors grew progressively in Batf3�/� Irf8VENUS� mice, asexpected (Fig. 3A). However, tumors also grew progressively in

Batf3�/� Irf8VENUSþ mice (Fig. 3A), despite the presence of cDC1cells capable of cross-presentation (Fig. 1). Tumorswere of similarsizes inWT and Batf3�/�mice, for both Irf8VENUS� and Irf8VENUSþ

mice, 7 to 8 days after implantation. However, on day 18, tumorsin WT mice were rejected completely, whereas large tumorspersisted in all Batf3�/� mice, both Batf3�/� Irf8VENUS� andBatf3�/� Irf8VENUSþ genotypes (Fig. 3B).

We also examined the infiltration of 1969 tumors byDCs and Tcells in these mice. There was an infiltration of MHCIIþ CD11cþ

cells into tumors of all mice, and tumors in WT Irf8VENUSþ micecontained XCR1þ cDC1s and CD8aþ T cells (Fig 3C–E), asexpected. By contrast, progressively growing tumors in bothBatf3�/� Irf8VENUS� and Batf3�/� Irf8VENUSþ mice lacked XCR1þ

cDCs and CD8aþ T cells. The lack of CD8aþ T cells in thesetumors is consistent with the lack of tumor rejection in thesemice.Thus, despite restoration of cDC1 development and cross-pre-sentation in Batf3�/� Irf8VENUSþ mice, these Batf3�/� Irf8VENUSþ

cDC1s are insufficient for tumor rejection.

DCs in skin of Batf3�/� Irf8VENUSþ mice lack XCR1 expressionConceivably, cDC1 might mediate tumor rejection by priming

T cells, though cross-presentation of tumor antigens delivered tolymph nodes through lymphatics. Alternately, rejection might

WT Batf3−/−

0 5 10 15 20 250

5

10

15

20

DayM

ean

diam

eter

(mm

)

Mea

n di

amet

er (m

m)

0

3

6

9

12

15

18

Days 7−8

0

3

6

9

12

15

18

Days 18

A B

68

10

0

102

103

104

105

9.0

20

0 102 103 104 105

0

102

103

104

105

51

18 0.2

1.3

0 102 103 104 105

44

27 0.2

3.3

0 102 103 104 105

WT

Bat

f3−/

−

CD11c

MH

CII

Bat

f3−/

−

CD172a

XC

R1

CD8a

XC

R1

0

102

103

104

105

CCD45+ 7AAD− CD11c+ MHCII+ CD11c− MHCII−

Irf8V

EN

US

–Irf

8VE

NU

S+

Irf8V

EN

US

–Irf

8VE

NU

S+

Irf8V

EN

US

–

Irf8Venus − + − + Batf3 +/+ +/+ −/− −/−

0

5

10

15

20

XC

R1+ D

Cs

(% o

f CD

11c+ M

HC

II+ )

n.s.

*D

Irf8Venus − − + Batf3 +/+ −/− −/−

Irf8Venus − − + Batf3 +/+ −/− −/−

CD

8+ T c

ells

(% o

f CD

45+

7AA

D− )

0

5

10

15

20

25 ***

n.s.

E

Figure 3.

Batf3 is required for tumor rejectioneven in IRF8VENUSþ mice. A and B, Miceof the indicated genotype were injectedwith 1 � 106 1969 fibrosarcoma cellssubcutaneously. Data are combinedfrom three experiments. A, Each linerepresents mean tumor diameter for anindividual mouse. B, Mean tumordiameter compared between theindicated genotypes on days 7–8 andday 18. Each symbol represents anindividual tumor. C–E, Flow cytometryanalysis of fibrosarcomas from mice ofthe indicated genotype on days 9–10.C, Representative flow-cytometricanalysis. CD45þ 7AAD� cells, pregatedas B220�, were analyzed for thepercentage of CD11cþ MHCIIþ cells andthe percentage of CD11c� MHCII� cells(first column). CD11cþ MHCIIþ cells wereanalyzed for percentage of XCR1þ

CD172a� cDCs (second column). MHCII�

CD11c� cells were analyzed for CD8aþ

XCR1� T cells (third column). D and E,Cumulative flow-cytometric analysisfrom fibrosarcomas grown inmice of theindicated genotype. Each symbolrepresents an individual tumor.D,XCR1þ

cDCs. One-way ANOVA, Sidak multiplecomparisons test, alpha 0.05; adjusted Pvalue: � , 0.0241; ns, 0.8645. E, CD8aþ

T cells. One-way ANOVA, Sidak multiplecomparisons test, alpha 0.05; adjustedP value: �� , 0.0001; ns, 0.6075.






require cDC1 to acquire antigens directly from tumors and trafficto lymph nodes to prime T cells. We therefore analyzed the cDCpopulations indermis of non–tumor-bearingmice (Fig. 4A). First,dermis of both Irf8VENUS� Batf3þ/þ (WT) and Irf8VENUSþ Batf3þ/þ

mice contained the two major DC populations, the CD11bþ

cDC2 and XCR1þ CD24þ CD11b� cDC1s (Fig. 4B). In addition,dermal DCs from Batf3�/� Irf8VENUS� mice included CD11bþ

DC2, but not XCR1þ CD24þ CD11b� cDC1s. However, dermalDCs from Batf3�/� Irf8VENUSþ mice included the CD11bþ cDC2and CD24þ CD11b� cDC1s, but unexpectedly these CD24þ

CD11b� DCs did not express XCR1 (Fig. 4C).Next, we analyzed resident andmigratory cDC1 populations in

skin draining lymph nodes (SDLN; Fig. 5). The resident gate,identified as CD11chi MHCIIint, contained CD24þ CD172a�

cDC1s in WT mice, both Irf8VENUS�and Irf8VENUSþ. These cDC1salso highly expressed XCR1 (Fig. 5A and B). The migratory DCgate, identified as CD11cint MHCIIhi, also contained CD24þ

CD172a� cDC1s in WT mice, both Irf8VENUS�and Irf8VENUSþ.However, XCR1 expression on these migratory cDC1s was lowerthan XCR1 expression on resident cDC1 cells, consistent withincreased maturation in migratory compared with resident DCs(ref. 27; Fig. 5A and C).

We have previously observed that SDLN from Batf3�/�mice onthe C57BL/6 background can retain a CD24þ CD172a� popula-tion resembling cDC1s (28). Their development has been attrib-uted to compensation by Batf (15). As expected, Batf3�/�

Irf8VENUS� mice retain this residual cDC1 population in theresident DC gate of the SDLN, and have lower XCR1 expressionand increased CD172a expression compared with WT CD24þ

CD172a� resident cDCs (Fig. 5B). In contrast, in Batf3�/�

Irf8VENUSþ mice, the CD24þ CD172a� cDC1 population has

similar CD24 and CD172a expression compared with WT mice(Fig. 5B), but lower expression of XCR1, being expressed on 50%of the CD24þ CD172a� cDC1s.

The residual CD24þ CD172a� cDC1s are much less abun-dant in the migratory gate of SDLNs from Batf3�/� Irf8VENUS�

mice and here they do not express XCR1 at all (Fig. 5C).Migratory CD24þ CD172a� cDC1s were restored in Batf3�/�

Irf8VENUSþ mice, but these also did not express XCR1. Thus, incontrast to spleen, the Irf8-Venus transgene does not fullyrestore XCR1 expression by cDC1 in the dermis and SDLN inthe absence of Batf3. These results may suggest that there maybe parallel pathways for inducing XCR1 expression, one ofwhich is Batf3-dependent, and another that is dependent on asignal present in the spleen and LN, but lacking in the dermis.This conditional regulation of XCR1 could be similar to ourprevious demonstration of the expression of CD103, which canbe both Batf3-dependent and induced by GM-CSF in theabsence of Batf3 (28).

We wanted to test whether cDC1s from skin and from themigratory gate of SDLN of Batf3�/� Irf8VENUSþ mice are able tocross-present cell-associated antigen. As expected, migratory DCsfromna€�vemicewere alreadymatured and had lost the capacity tocross-present newly acquired antigen toOT-1 T cells (Supplemen-tary Fig. S2A; refs. 27, 29). Not surprisingly, our isolation proce-dure for skin dendritic cells resulted in their maturation, so wecould not directly assay their cross-presentation capacity (30).Therefore, to answer this question, we implanted a fibrosarcomaexpressingmembrane-anchored ovalbumin, 1956-ova, and askedwhether migratory cDCs purified from SDLN could stimulateproliferation of OT-1 T cells in vitro (Supplementary Fig. S2B).Because we have shown previously that cDC1 and not cDC2 areuniquely capable of cross-presentation (Supplementary Fig. S1)(24), we did not separate cDC1 from cDC2 in this experiment.Migratory, but not resident, cDCs from 1956 ova tumor-bearingWT Irf8VENUS� and WT Irf8VENUSþ mice stimulated robust OT-1proliferation. Migratory cDCs from Batf3�/� Irf8VENUSþmice alsostimulated OT-1 proliferation at approximately 40% of WT, butsignificantly more than their counterparts from the resident gate(Supplementary Fig. S2B). We conclude that migratory cDC1sfrom SDLN in Batf3�/� Irf8VENUSþ can cross-present antigen.However, we cannot be sure that these migratory cDC1s acquiredantigen while they resided in the skin, and it remains a possibilitythat migration of skin cDC1s or their antigen processing may bealtered.

Identification of Batf3-dependent target genes in cDC1 cellsTo determine Batf3-dependent genes that may contribute

to tumor rejection, we used microarray analysis to comparecDC1s purified from spleens of WT Irf8VENUS�, WT Irf8VENUSþ,and Batf3�/� Irf8VENUSþ mice (Fig. 6). First, increased IRF8 pro-vided by the Irf8VENUS transgene in WT cDC1s did not cause anoverall increase in gene expression for cDC1-associated genes(Fig. 6A). Irf8 expression was increased 1.6-fold in Irf8VENUSþ

compared with Irf8VENUS� cDC1s, presumably a result of the Irf8transgenes, but other cDC1-associated genes, such as Btla, Itgae,CD8a, Clec9a, and Xcr1, were not affected by Irf8VENUS. Batf wasnot induced by Irf8VENUS, which indicates that compensation byBatf is unlikely to be the mechanism by which Irf8VENUS restorescDC1 in Batf3�/� mice.

To determine which cDC1-associated genes were dependenton Batf3 for their expression, we compared cDC1s from

8.7

8.5

7.5

0 102 103 104 105

0

102

103

104

105

17

8.3

4.7

10

2.3

0.4

17

9.6

0.8

CD11c

MH

CII

CD11b

CD

24

XCR1

CD

24

WT Batf3−/−

Batf3−/−

Irf8VENUS− Irf8VENUS+Irf8VENUS−Irf8VENUS+

WT

Irf8VENUS− Irf8VENUS+Irf8VENUS−Irf8VENUS+

A

B

C

Figure 4.

Batf3�/� IRF8VENUSþ mice lack XCR1þ dermal cDC1s. A–C, Flow cytometryanalysis of ear skin of non–tumor-bearing mice of the indicated genotype. A,Percentage ofMHCIIþCD11cþ cells. Cellswere pregated as 7AAD�CD45þB220�

CD326INT/LOW. B, CD24þ CD11b� cells within MHCIIþ CD11cþ cells from A. C,CD24þ XCR1þ cDC1s within MHCIIþ CD11cþ cells from A. Data are representativeof at least three biological replicates for each genotype.

Theisen et al.





WT Irf8VENUSþ mice and Batf3�/� Irf8VENUSþ mice (Fig. 6; Sup-plementary Fig. S3). Approximately 10 genes had decreasedexpression in Batf3�/� Irf8VENUSþ cDC1 from spleen comparedwith WT Irf8VENUSþ cDC1 (Fig. 6C). Itga8 had the highest foldchange (10.4�) between WT Irf8VENUSþ and Batf3�/� Irf8VENUSþ

cDC1s. Other Batf3-dependent genes with 3-fold or greaterchanges between Batf3þ/þ Irf8VENUSþ and Batf3�/� Irf8VENUSþ

cells included the signaling adaptor molecule Clnk, the phospha-tase Ppef2 andGcsam, among others.We found no genes that wereincreased by the Irf8VENUS transgene in WT cells, with the excep-tionof Irf8 itself. Among a list of genes important for developmentor function inDCs, including Id2, Ciita, Nfil3, Zeb2, and Irf4, nonewere Batf3-dependent (Supplementary Fig. S3A). Compared withWT Irf8VENUSþ cDC1s, Batf3�/� Irf8VENUSþ cDC1 had higherexpression of Ccr7 and Ccr5, which are known to be involved inDC migration to lymph nodes (31, 32), but no difference in

expression of other chemokine receptors, including Xcr1 (Sup-plementary Fig. S3B). Itga8 was the only integrin for whichexpression was Batf3-dependent (Supplementary Fig. S3C).

Wehadobserved that XCR1 expressionwas reduced inBatf3�/�

Irf8VENUSþ migratory cDC1 from SDLN compared with WT mice(Fig. 5C), so we investigated whether other genes were alsoreduced in the lymph node cDC1s. We analyzed cDC1s fromSDLNs of WT Irf8VENUSþ mice and Batf3�/� Irf8VENUSþ mice bymicroarray (Fig. 6C; Supplementary Fig. S3D). All but one of thespleen Batf3-dependent genes, Ctla2b, were also Batf3-dependentin resident cDC1s from SDLN (Fig. 6C). These genes were alsoreduced inmigratory, i.e., mature, cDC1s compared with residentcDC1s fromWT Irf8VENUSþ mice and were not further reduced inBatf3�/� Irf8VENUSþ migratory cDC1s. A total of 34 genes, includ-ing 8 listed for spleen, were reduced in Batf3�/� Irf8VENUSþ

resident cDC1s compared with WT Irf8VENUSþ resident cDC1s(Supplementary Fig. S3D). Twenty-eight of these genes are amongthe approximately 900 genes that are 3-fold or greater reducedin migratory cDC1s compared with resident cDC1s from WTIrf8VENUSþ mice. Fifteen genes that maintained expression inWT Irf8VENUSþ migratory cDC1s were reduced in Batf3�/�

Irf8VENUSþ compared with WT Irf8VENUSþ migratory cDC1s.Xcr1 is 9.1-fold reduced in migratory, compared with resident,cDC1s from WT Irf8VENUSþmice, and stands out as being theonly gene that is Batf3-dependent in both resident and migra-tory cDC1s. In summary, the most Batf3-dependent genes arealso genes that are reduced upon the maturation of cDC1 andso are more expressed in resident cDC1 compared with migra-tory cDC1.

We analyzed IRF8 and BATF3 ChIP-seq binding in genomicloci of Batf3-dependent genes (Fig. 7). Shown are some exam-ples, Clnk, Gcsam, Itga8, and Xcr1, whose expression is specificto cDC1 compared with cDC2, is unaffected by the presence ofthe Irf8VENUS transgene in WT mice, and is Batf3-dependent(Fig. 7A). For each of these genes, we could find strong BATF3and IRF8 binding located in open chromatin as assessed bypeaks of H3K27Ac, either in the gene body (Gcsam) or nearby(Clnk, Itga8, and Xcr1; Fig. 7B), suggesting these genes may bedirect targets of Batf3 and Irf8. To begin to address which of theBatf3-dependent genes might be required for tumor rejection,we implanted 1969 fibrosarcoma cells into mice with condi-tional deletion of Itga8 (Supplementary Fig. S4). As controls,WT mice could reject 1969 tumor cells (Supplementary Fig.S4A) but Batf3�/� mice could not (Supplementary Fig. S4B).However, the conditional loss of Itga8 in CD11cþ cells was notsufficient to abrogate tumor rejection (Supplementary Fig.S4C). Additional analysis will be required to determine whichcomponents of the Batf3-dependent pathway are required fortumor rejection.

DiscussionOur aim was to determine which cDC1-specific genes rely on

Batf3 for their expression andwhether anyof these are required fortumor rejection mediated by cDC1. The requirement for Batf3 inmaintaining Irf8 expression during cDC1 development has pre-vented a direct determination of Batf3 target genes in cDC1,because cDC1 normally fail to develop in Batf3�/� mice. In thisstudy, we developed a system to allow cDC1 development in theabsence of Batf3 and identified a relatively small number of cDC1genes rely on Batf3 for their expression. This set of Batf3-target

1.7

1.1

0 103

104

0

-102

102

103

104

35

0 103

104

0

103

104

0 10 10

23

0 103

104

0

103

104

19

3 4

0

103

104

1.7

1.1

21

11

8.4

1.2

1.3

21

8.8

3.8

1.2

1.0

16

15

3.1

39

3 4

0

103

104

23 29 32

CD172a

XC

R1

CD172a

XC

R1

CD172a

CD

24

CD172a

CD

24

Irf8VENUS– Irf8VENUS+

CD11c

MH

CII

Irf8VENUS– Irf8VENUS+WT Batf3−/−

Resident DCs (MHCIIint CD11chi)

Migratory DCs (MHCIIhi CD11cint)

A

C

B

Figure 5.

Batf3�/� IRF8VENUSþ mice lack XCR1þ migratory cDC1s in SDLN. A–C, Flowcytometry analysis of SDLN (pooled inguinal, axial, brachial) of non–tumor-bearing mice of the indicated genotype. A, Gating for resident cDCs (MHCIIint

CD11chi), and migratory cDCs (MHCIIhi CD11cint), is shown. Cells were pregatedas B220�, CD326int/low. B, Resident cDCs (MHCIIint CD11chi) as gated fromA were analyzed for the percentage of CD24þ CD172a� or XCR1þ CD172a�

cDC1s. C,Migratory cDCs (MHCIIhi CD11cint) as gated fromAwere analyzed forthe percentage of CD24þ CD172a� or XCR1þ CD172a� cDC1s. Data arerepresentative of 6 biological replicates for each genotype.






genes includes genes required for cDC1 to mediate tumorrejection.

One Batf3 transcriptional target gene is Irf8. The high IRF8expression that is required for cDC1 development is achieved byIrf8 autoactivation that depends on the cooperative interactionbetween IRF8 and BATF3. This interaction is likely mediatedthrough an Irf8 þ32 kb enhancer that contains multiple AICEswhere BATF3 and IRF8 cooperatively bind (13). We took advan-tage of the phenomenon that the Irf8-Venus reporter canmaintaincDC1 development in the absence of BATF3 and BATF to discoverwhether the function of cDC1 is entirely a property of high IRF8expression or whether other genes in mature cDC1 that mayrequire BATF3 for their expression might contribute to cDC1function.

First, we found that the cDC1 that develop in Batf3�/�

Irf8VENUSþ mice were capable of cross-presentation of cell-asso-ciated antigen, a process that is specific to the cDC1 lineage (24).Thus, the transcriptional program for cross-presentation of cell-associated antigens may rely solely on high Irf8 expression,and not directly on Batf3-dependent gene expression. However,we found that Batf3�/� Irf8VENUSþ mice, like Batf3�/� mice,cannot reject immunogenic fibrosarcomas that are normallyrejected byWTmice, despite functional cross-presentation in vivo.This result is in contrast to the case in which cDC1 developmentwas restored in Batf3-deficient mice by IL12 treatment, in whichcDC1 restoration led to restored resistance to Toxoplasma gondii,restoration of cross-presenting cDC1 and to capacity for tumorrejection (15). Thus, IL12 treatment may provide additional ordifferent signals thatmaynot bepresent in theBatf3�/� Irf8VENUSþ

mice that compensates for the absence of Batf3 in cDC1 for thesefunctions.

All of the identified candidate genes for which expressionrequires Batf3 are expressed more highly in cDC1 compared withcDC2 cells. Such genes may be required for tumor rejection butwould not be predicted to be required for cDC1 development orcross-presentation. For example, Itga8 as a heterodimer withItgb1, a8b1, has several reported ligands including osteopontin,fibronectin, vitronectin, nephoronectin, tenascin C, and MFGE8(33–36). Although Itga8 expression is not limited to the immunesystem, its expression on cDC1s could facilitate interactions withthe tumor microenvironment (25, 37). However, a single genedefect in Itga8 expression did not block tumor rejection, but thismay be due to redundancy with other partners of Itgb1, whichwill require future studies.

In addition to cDC1s, Gcsam and its human counterpart HGALare also highly expressed in germinal center B cells. Gcsam is acytoplasmic protein that contains a putative immune tyrosineactivation motif. Gcsam's biological role is unknown, but it hasbeen shown that Gcsam is dispensable for the germinal centerreaction (38). Another Batf3-dependent candidate is Clnk, a mem-ber of the Blnk/SLP-76 adapter family. In addition to cDC1s,Clnk isexpressed in activated T cells, IL2-activated NK cells, and mast cells(39).No immune defects were observed inClnk-deficientmice, butthis was attributed to potential compensation by SLP-76. cDC1-specific functions for these and other BATF3-target genes have notbeen evaluated in the context of tumor rejection.

XCR1 has been recognized as a robust marker of cDC1 (40). Inour study, Xcr1 was not initially seen to be dependent on Batf3 in

Batf3 13.8 6.3 9.0Tgfbi 3.4 2.2 0.8

Gcsam 5.3 8.2 2.1Pdia5 4.0 2.7 1.4

Ocstamp 3.6 1.6 1.1Cxx1c 3.0 3.6 0.8Ppef2 4.3 1.8 1.1Clnk 4.0 4.0 1.5Itga8 10.4 7.7 1.4

Arhgap42 6.0 5.5 1.5Fgd6 3.2 3.7 0.9Fcrla 4.9 9.5 1.6

Ctla2b 5.8 1.1 1.7Tshz1 2.6 2.3 1.3Plce1 3.6 4.7 1.2

Spleen SDLN resident SDLN migratory

WTIrf8VENUS–

Batf3−/−Irf8VENUS+

WT Batf3−/− Batf3−/−Irf8VENUS+

WTIrf8VENUS+

WTFC FC FC

Expression(linear)

0 1.5××103

Expression(linear)

0 1×103

Expression(linear)

0 1×103

C

DC1/DC2(fold change) W

T Irf

8VE

NU

S+ /

Bat

f3−/

− Irf

8VE

NU

S+

(fold

cha

nge) Batf3 Itga8

0.01 0.1 10 100

0.01

0.1

10

100

0.01 0.1 10 100

0.01

0.1

10

100

Irf8

Xcr1

Clec9a

Clec4b2

Cd8aItgae

ClnkItga8

BtlaBatf3

Batf

WT

Irf8V

EN

US

+ / W

T Irf

8VE

NU

S–

(fold

cha

nge) Ctla2b

Gcsam

Arhgap42Fcrla

Clnk

Ppef2Fgd6

DC1/DC2(fold change)

A B

Figure 6.

A limited number of cDC1-specific genesare Batf3-dependent. A–C, Microarrayanalysis of cDCs from mice of theindicated genotype. A, Fold change inexpression of annotated probe setsbetween cDC1s and cDC2s (x-axis) isplotted against the fold change inexpression between WT Irf8VENUSþ andWT Irf8VENUS� cells (y-axis). B, Foldchange in expression of annotatedprobe sets between cDC1s and cDC2s(x-axis) is plotted against fold changebetween WT Irf8VENUSþ and Batf3�/�

Irf8VENUSþ cells (y-axis).C, Expression ofcDC1-specific genes that were�2.6-foldmore highly expressed in WT Irf8VENUSþ

compared with Batf3�/� Irf8VENUSþ

cDC1s from spleen, shown for spleencDC1s, and for resident and migratorycDC1s from SDLN (pooled inguinal,brachial, and cervical). FC indicates thefold change in gene expression betweenWT Irf8VENUSþ and Batf3�/� Irf8VENUSþ

cells. Each column represents anindependent microarray.

Theisen et al.





splenic cDC1, but was more Batf3-dependent in cDC1s fromSDLNs. Xcr1 expression was also diminished upon cDC1 matu-ration in SDLNs, being reduced by 9-fold in migratory cDC1srelative to resident cDC1s. These observations could imply thatthe Xcr1 gene is regulated differentially in different tissues. XCL1produced by CD8 T cells can act as a chemoattractant for XCR1þ

cDC1s (41). It is possible that lack of XCR1 on migratory cDC1scould help explain lack of tumor rejection in the Batf3�/�

Irf8VENUSþ mice. In a vaccinia infection model, Xcr1 was requiredfor the clustering of CD8þ T cells with cDC1s (42). Another studyshowed that NK cell release of XCL1 in tumors may recruit cDC1sinto tumors, but stated that loss ofXcr1was not sufficient to blockintratumor cDC1 accumulation (43). However, as far as we are

aware, neither these studies nor others (44, 45) reported whetherloss of XCR1 abrogated tumor rejection, so future studies will berequired to test this.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: D.J. Theisen, S.T. Ferris, N. Kretzer, T.L. MurphyDevelopment of methodology: D.J. Theisen, N. Kretzer, T.L. MurphyAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): D.J. Theisen, S.T. Ferris, C.G. Brise~no, A. Iwata,T.L. Murphy

Cln

k

0

100

200

300

400

100 kb mm10

H3K27ac25 _

0 _

BATF3150 _

0 _

IRF865 _

0 _

ClnkG

csam

Itga8

0.0025**

0.0082**

0.0044**

A B

0

500

1,000

0

100

200

300

cDC1 cDC2

Irf8Venus − + + − Batf3 +/+ +/+ −/− +/+

5 kb mm10

_H3K27ac25

0 _

BATF3160 _

0 _

IRF833 _

0 _

Gcsam

100 kb mm10

H3K27ac30 _

0 _

BATF3135 _

0 _

IRF890 _

0 _

Itga8Mindy3

H3K27ac

BATF3

IRF80

1,000

2,000

3,000

Xcr1

0.6587ns

10 kb mm10

Fyco1Xcr1

40 _

0 _

46 _

0 _

47 _

0 _

0

Figure 7.

Batf3-dependent genes harbor nearby enhancers binding BATF3/IRF8. A, Expression of selected Batf3-dependent genes in the indicated cell type andgenotype. Each symbol represents an independent expression array. Two-tailed unpairedStudent t tests; alpha¼0.05.B,ChIP-seq analysis for H3K27Ac, BATF3, andIRF8 in WT cDC1s for the indicated loci, Clnk, Gcsam, Itga8, and Xcr1 as indicated. Previously reported H3K27Ac, BATF3, and IRF8 ChIP-seq data (GSE66899)were remapped to mm10. Genomic scales are shown beneath each plot.






Analysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): D.J. Theisen, T.L. MurphyWriting, review, and/or revision of the manuscript: S.T. Ferris, K.M. Murphy,T.L. MurphyStudy supervision: T.L. Murphy

AcknowledgmentsThis work was supported by the Howard Hughes Medical Institute (K.M.

Murphy). We thank the Alvin J. Siteman Cancer Center at Washington Univer-sity School of Medicine for use of the Center for Biomedical Informatics and

Multiplex Gene Analysis Genechip Core Facility. We thank Robert Schreiber forthe 1969 and 1956 fibrosarcoma lines.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

ReceivedMarch 5, 2018; revised September 12, 2018; acceptedNovember 21,2018; published first November 27, 2018.

References1. Murphy TL, Grajales-Reyes GE, Wu X, Tussiwand R, Briseno CG, Iwata A,

et al. Transcriptional control of dendritic cell development. Annu RevImmunol 2016;34:93–119.

2. Hildner K, Edelson BT, Purtha WE, Diamond M, Matsushita H, KohyamaM, et al. Batf3 deficiency reveals a critical role for CD8alphaþdendritic cellsin cytotoxic T cell immunity. Science 2008;322:1097–100.

3. Torti N, Walton SM, Murphy KM, Oxenius A. Batf3 transcriptionfactor-dependent DC subsets in murine CMV infection: differentialimpact on T-cell priming and memory inflation. Eur J Immunol 2011;41:2612–8.

4. Nopora K, Bernhard CA, Ried C, Castello AA,Murphy KM,Marconi P, et al.MHC class I cross-presentation by dendritic cells counteracts viral immuneevasion. Front Immunol 2012;3:348.

5. Diamond MS, Kinder M, Matsushita H, Mashayekhi M, Dunn GP,Archambault JM, et al. Type I interferon is selectively required bydendritic cells for immune rejection of tumors. J Exp Med 2011;208:1989–2003.

6. Fuertes MB, Kacha AK, Kline J, Woo SR, Kranz DM, Murphy KM, et al.Host type I IFN signals are required for antitumor CD8þ T cellresponses through CD8{alpha}þ dendritic cells. J Exp Med 2011;208:2005–16.

7. Li L, KimS,Herndon JM,Goedegebuure P, Belt BA, SatpathyAT, et al. Cross-dressedCD8alphaþ/CD103þdendritic cells primeCD8þT cells followingvaccination. Proc Natl Acad Sci USA 2012;109:12716–21.

8. Broz ML, Binnewies M, Boldajipour B, Nelson AE, Pollack JL, Erle DJ,et al. Dissecting the tumor myeloid compartment reveals rare activatingantigen-presenting cells critical for T cell immunity. Cancer Cell 2014;26:638–52.

9. Roberts EW, Broz ML, Binnewies M, Headley MB, Nelson AE, Wolf DM,et al. Critical role for CD103þ/CD141þ dendritic cells bearing CCR7 fortumor antigen trafficking and priming of T cell immunity in melanoma.Cancer Cell 2016;30:324–36.

10. Gubin MM, Zhang X, Schuster H, Caron E, Ward JP, Noguchi T, et al.Checkpoint blockade cancer immunotherapy targets tumour-specificmutant antigens. Nature 2014;515:577–81.

11. Sanchez-Paulete AR, Cueto FJ, Martinez-Lopez M, Labiano S, Morales-Kastresana A, Rodriguez-Ruiz ME, et al. Cancer immunotherapy withimmunomodulatory anti-CD137 and anti-PD-1 monoclonal antibo-dies requires BATF3-dependent dendritic cells. Cancer Discov 2016;6:71–9.

12. Salmon H, Idoyaga J, Rahman A, Leboeuf M, Remark R, Jordan S, et al.Expansion and activation of CD103(þ) dendritic cell progenitors at thetumor site enhances tumor responses to therapeutic PD-L1 and BRAFinhibition. Immunity 2016;44:924–38.

13. Grajales-Reyes GE, Iwata A, Albring J,WuX, Tussiwand R, KCW, et al. Batf3maintains autoactivation of Irf8 for commitment of a CD8alpha(þ)conventional DC clonogenic progenitor. Nat Immunol 2015;16:708–17.

14. Murphy TL, Tussiwand R, Murphy KM. Specificity through cooperation:BATF-IRF interactions control immune-regulatory networks. Nat RevImmunol 2013;13:499–509.

15. Tussiwand R, LeeWL,Murphy TL,MashayekhiM,Wumesh KC, Albring JC,et al. Compensatory dendritic cell development mediated by BATF-IRFinteractions. Nature 2012;490:502–7.

16. Seillet C, Jackson JT, Markey KA, Hill GR, Macdonald KP, Nutt SL, et al.CD8alphaþDCs can be induced in the absence of transcription factors Id2,Nfil3 and Batf3. Blood 2013.

17. Schonheit J, Kuhl C, Gebhardt ML, Klett FF, Riemke P, Scheller M, et al.PU.1 level-directed chromatin structure remodeling at the Irf8 gene drivesdendritic cell commitment. Cell Rep 2013;3:1617–28.

18. Schraml BU,Hildner K, IseW, LeeWL, SmithWA, SolomonB, et al. The AP-1 transcription factor Batf controls T(H)17 differentiation. Nature2009;460:405–9.

19. Lybarger L,Wang X,HarrisMR, VirginHW,Hansen TH. Virus subversion ofthe MHC class I peptide-loading complex. Immunity 2003;18:121–30.

20. Caton ML, Smith-Raska MR, Reizis B. Notch-RBP-J signaling controls thehomeostasis of CD8- dendritic cells in the spleen. J Exp Med 2007;204:1653–64.

21. Shankaran V, Ikeda H, Bruce AT, White JM, Swanson PE, Old LJ, et al.IFNgamma and lymphocytes prevent primary tumour development andshape tumour immunogenicity. Nature 2001;410:1107–11.

22. Sedy JR, Gavrieli M, Potter KG,HurchlaMA, Lindsley RC,Hildner K, et al. Band T lymphocyte attenuator regulates T cell activation through interactionwith herpesvirus entry mediator. Nat Immunol 2005;6:90–8.

23. den Haan JM, Lehar SM, Bevan MJ. CD8(þ) but not CD8(-) dendritic cellscross-prime cytotoxic T cells in vivo. J Exp.Med. 2000;192:1685–96.

24. Kretzer NM, Theisen DJ, Tussiwand R, Briseno CG, Grajales-Reyes GE, WuX, et al. RAB43 facilitates cross-presentation of cell-associated antigens byCD8alphaþ dendritic cells. J Exp.Med 2016;213:2871–83.

25. Gurka S, Hartung E, Becker M, Kroczek RA. Mouse conventional dendriticcells can be universally classified based on the mutually exclusive expres-sion of XCR1 and SIRPalpha. Front Immunol 2015;6:35.

26. Briseno CG, Haldar M, Kretzer NM,Wu X, Theisen DJ, KCW, et al. Distincttranscriptional programs control cross-priming in classical andmonocyte-derived dendritic cells. Cell Rep 2016;15:2462–74.

27. Ardouin L, Luche H, Chelbi R, Carpentier S, Shawket A, Montanana SF,et al. Broad and largely concordant molecular changes characterize tolero-genic and immunogenic dendritic cell maturation in thymus and periph-ery. Immunity 2016;45:305–18.

28. Edelson BT, Bradstreet TR, KCW, Hildner K, Herzog JW, Sim J, et al. Batf3-dependent CD11b(low/-) peripheral dendritic cells are GM-CSF-indepen-dent and are not required for Th cell priming after subcutaneous immu-nization. PLoS One 2011;6:e25660.

29. Wilson NS, El Sukkari D, Belz GT, Smith CM, Steptoe RJ, Heath WR, et al.Most lymphoid organ dendritic cell types are phenotypically and func-tionally immature. Blood 2003;102:2187–94.

30. Malissen B, Tamoutounour S, Henri S. The origins and functions ofdendritic cells and macrophages in the skin. Nat Rev Immunol 2014;14:417–28.

31. Ohl L, Mohaupt M, Czeloth N, Hintzen G, Kiafard Z, Zwirner J, et al. CCR7governs skin dendritic cell migration under inflammatory and steady-stateconditions. Immunity 2004;21:279–88.

32. Aliberti J, Reis E Sousa, Schito M, Hieny S, Wells T, Huffnagle GB, et al.CCR5 provides a signal for microbial induced production of IL-12 by CD8alphaþ dendritic cells. Nat Immunol 2000;1:83–7.

33. Denda S, Reichardt LF, Muller U. Identification of osteopontin as a novelligand for the integrin alpha8 beta1 and potential roles for this integrin-ligand interaction in kidney morphogenesis. Mol Biol Cell 1998;9:1425–35.

34. Kiyozumi D, Takeichi M, Nakano I, Sato Y, Fukuda T, Sekiguchi K.Basementmembrane assembly of the integrin alpha8beta1 ligand nephro-nectin requires Fraser syndrome-associated proteins. J Cell Biol 2012;197:677–89.

Theisen et al.





35. Schnapp LM, Hatch N, Ramos DM, Klimanskaya IV, Sheppard D, Pytela R.The human integrin alpha 8 beta 1 functions as a receptor for tenascin,fibronectin, and vitronectin. J Biol Chem. 1995;270:23196–202.

36. Khalifeh-Soltani A,HaA, PodolskyMJ,McCarthyDA,McKleroyW,Azary S,et al. alpha8beta1 integrin regulates nutrient absorption through anMfge8-PTEN dependent mechanism. Elife 2016;5.

37. Heng TS, Painter MW, Immunological Genome Project Consortium. Theimmunological genome project: networks of gene expression in immunecells. Nat Immunol 2008;9:1091–4.

38. Schenten D, Egert A, Pasparakis M, Rajewsky K. M17, a gene specific forgerminal center (GC) B cells and a prognostic marker for GC B-celllymphomas, is dispensable for the GC reaction in mice. Blood 2006;107:4849–56.

39. Wu JN, Koretzky GA. The SLP-76 family of adapter proteins. SeminImmunol 2004;16:379–93.

40. Crozat K, Guiton R, Contreras V, Feuillet V, Dutertre CA, Ventre E, et al. TheXC chemokine receptor 1 is a conserved selective marker of mammaliancells homologous to mouse CD8alphaþ dendritic cells. J Exp Med2010;207:1283–92.

41. Dorner BG, Dorner MB, Zhou X, Opitz C, Mora A, Guttler S, et al. Selectiveexpression of the chemokine receptor XCR1 on cross-presenting dendriticcells determines cooperation with CD8þ T cells. Immunity 2009;31:823–33.

42. Brewitz A, Eickhoff S, Dahling S, Quast T, Bedoui S, Kroczek RA, et al.CD8þ T Cells Orchestrate pDC-XCR1þ dendritic cell spatial andfunctional cooperativity to optimize priming. Immunity 2017;46:205–19.

43. Bottcher JP, Bonavita E, Chakravarty P, Blees H, Cabeza-Cabrerizo M,Sammicheli S, et al. NK cells stimulate recruitment of cDC1 into the tumormicroenvironment promoting cancer immune control. Cell 2018;172:1022–37.

44. Yamazaki C, Sugiyama M, Ohta T, Hemmi H, Hamada E, Sasaki I, et al.Critical roles of a dendritic cell subset expressing a chemokine receptor,XCR1. J Immunol 2013;190:6071–82.

45. Kitano M, Yamazaki C, Takumi A, Ikeno T, Hemmi H, Takahashi N,et al. Imaging of the cross-presenting dendritic cell subsets in theskin-draining lymph node. Proc Natl Acad Sci USA 2016;113:1044–9.






2019;7:29-39. Published OnlineFirst November 27, 2018.Cancer Immunol Res Derek J. Theisen, Stephen T. Ferris, Carlos G. Briseño, et al. Dendritic Cells Independently of Cross-Presentation

-Dependent Genes Control Tumor Rejection Induced byBatf3

Updated version

10.1158/2326-6066.CIR-18-0138doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerimmunolres.aacrjournals.org/content/suppl/2018/11/27/2326-6066.CIR-18-0138.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerimmunolres.aacrjournals.org/content/7/1/29.full#ref-list-1

This article cites 43 articles, 16 of which you can access for free at:

Citing articles

http://cancerimmunolres.aacrjournals.org/content/7/1/29.full#related-urls

This article has been cited by 1 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerimmunolres.aacrjournals.org/content/7/1/29To request permission to re-use all or part of this article, use this link



http://cancerimmunolres.aacrjournals.org/lookup/doi/10.1158/2326-6066.CIR-18-0138

http://cancerimmunolres.aacrjournals.org/content/suppl/2018/11/27/2326-6066.CIR-18-0138.DC1

http://cancerimmunolres.aacrjournals.org/content/7/1/29.full#ref-list-1

http://cancerimmunolres.aacrjournals.org/content/7/1/29.full#related-urls

http://cancerimmunolres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerimmunolres.aacrjournals.org/content/7/1/29


batf3-dependent genes control tumor rejection …...research article batf3-dependent genes control...

Documents