Type 1 conventional dendritic cell fate and function are controlled by DC-SCRIPT

See allHide authors and affiliations

Science Immunology  02 Apr 2021:
Vol. 6, Issue 58, eabf4432
DOI: 10.1126/sciimmunol.abf4432

DC-SCRIPT takes center stage

Dendritic cells (DCs) initiate diverse and context-specific adaptive immune responses. Zhang et al. define a role for the transcription factor DC-SCRIPT in the development and function of type 1 conventional DCs (cDC1s) by generating DC-SCRIPT reporter and knockout mouse strains. DC-SCRIPT was required for establishment of the cDC1 gene expression program, as well as cross-presentation of cell-associated antigen and production of IL-12, which are important for stimulating effective type 1 immune responses. The cell type–specific role of DC-SCRIPT was likely due at least in part to its ability to directly control expression of the transcription factor IRF8 in cDC1s, highlighting DC-SCRIPT as a key component of the gene regulatory networks coordinating cDC1 development and function.


The functional diversification of dendritic cells (DCs) is a key step in establishing protective immune responses. Despite the importance of DC lineage diversity, its genetic basis is not fully understood. The transcription factor DC-SCRIPT is expressed in conventional DCs (cDCs) and their committed bone marrow progenitors but not in plasmacytoid DCs (pDCs). We show that mice lacking DC-SCRIPT displayed substantially impaired development of IRF8 (interferon regulatory factor 8)–dependent cDC1, whereas cDC2 numbers increased marginally. The residual DC-SCRIPT–deficient cDC1s had impaired capacity to capture and present cell-associated antigens and to secrete IL-12p40, two functional hallmarks of this population. Genome-wide mapping of DC-SCRIPT binding and gene expression analyses revealed a key role for DC-SCRIPT in maintaining cDC1 identity via the direct regulation of cDC1 signature genes, including Irf8. Our study reveals DC-SCRIPT to be a critical component of the gene regulatory program shaping the functional attributes of cDC1s.


Dendritic cells (DCs) are a heterogeneous and functionally distinct group of immune cells that are pivotal to adaptive immune responses (1). In the steady state, DCs encompass conventional DC (cDC) and plasmacytoid DC (pDC) subsets. cDCs are highly potent at taking up, processing, and presenting antigens to T cells, orchestrating their expansion, functional polarization, and effector activity. In contrast, pDCs have a limited capacity to promote T cell expansion but have the characteristic ability to produce abundant type I interferon (IFN) early during viral infection (2). pDCs and cDCs can develop from a common DC progenitor (CDP) in the bone marrow (BM), although a large part of the pDC pool may come from another source, namely, CD135+ CD115 progenitors (37). cDCs are further divided into two subsets, cDC1 and cDC2 (8). Although both DC subsets are potent drivers of CD4+ T cell responses via major histocompatibility complex class II (MHCII), cDC1 are endowed with the ability to cross-present exogenous antigens to CD8+ T cells (9, 10), a property under intense investigation given its potential role in promoting antitumor responses (11, 12).

The initial specification of DC-committed progenitors and their subsequent lineage diversification is controlled by a group of transcription factors that act together in a gene regulatory network (13, 14). The transcription factors PU.1 (Spi1) and IFN regulatory factor 8 (IRF8) are essential for DC lineage priming in hematopoietic progenitors (1517). Recently, the analysis of cis-regulatory elements associated with Irf8 expression pointed to a critical role of the dynamic changes in enhancer usage as DC progenitors mature, enabling the recruitment of lineage-defining transcription factors at discrete developmental branching points (17, 18). Other determinants are also required for the development of either cDC1 or cDC2. The transcription factors BATF3, NFIL3, and ID2 are instrumental in controlling cDC1 development (19), whereas NOTCH2 and KLF4 have been implicated into the specification of functionally distinct cDC2 subsets (20, 21).

Previous work from our laboratory highlighted a transcriptional node controlling cDC identity regulated by PU.1 and DC-SCRIPT (22). DC-SCRIPT, encoded by Zfp366 in mouse and ZNF366 in human, is a relatively poorly understood Zn finger transcription factor whose expression is largely restricted to DCs in both human and mouse hematopoietic systems (22, 23). Gene knockdown studies in human monocyte-derived DCs suggested a critical role for DC-SCRIPT in controlling interleukin-10 (IL-10) release after Toll-like receptor (TLR) agonism (24, 25). DC-SCRIPT has also been proposed to direct the response of DCs and epithelial cells to steroid hormones through the control of nuclear receptor activity (23, 26, 27). Although these studies suggest that DC-SCRIPT is a critical protein in sensing environmental cues, the lack of genetic models for this transcription factor has hindered our capacity to decipher its importance in DC lineage specification and function.

In the present study, we characterized the expression pattern and the function of DC-SCRIPT in the cDC compartment using newly developed reporter and loss-of-function mouse models. DC-SCRIPT deficiency impaired both the development and functionality of cDC1s, whereas the number of cDC2s was slightly increased. Genome-wide transcriptome and DNA binding analysis identified a critical role for DC-SCRIPT in maintaining the gene signature characteristic of cDC1s. We also showed that DC-SCRIPT binds to the regulatory regions of many canonical cDC1 genes, including Il12b and the +32-kb Irf8 superenhancer essential for optimal IRF8 expression during cDC1 maturation. Together, this study demonstrates that DC-SCRIPT is a pivotal element of regulatory network controlling cDC lineage specification and effector functions.


DC-SCRIPT expression is restricted to cDCs, their committed progenitors, and selected tissue-resident macrophages

Earlier reports from our group, and others, have highlighted the restricted pattern of expression of DC-SCRIPT in the hematopoietic system (22, 23, 27). To accurately define its expression in vivo, we used CRISPR-Cas9 genome editing technology to insert an internal ribosome entry site (IRES)–tdTomato cassette followed by viral Thosea asigna virus 2A (T2A) cleavage site and Cre recombinase sequences into the 3′ untranslated region of Zfp366 (Fig. 1A). Zfp366-tdTomato/+ mice were indistinguishable in survival, hematopoietic cellularity, and lineage composition from C57BL/6 controls and had normal Mendelian inheritance.

Fig. 1 Expression of DC-SCRIPT defines the cDC lineage.

(A) Schematic representation of the Zfp366tdTomato reporter strain (not to scale). An IRES tandem dimeric (td)Tomato-T2A-Cre recombinase-targeting construct was inserted at the 3′ end of the Zfp366 gene. The Zfp366 exons were indicated as blue boxes and the introns as black lines. 3′UTR, 3′ untranslated region. (B) DC-SCRIPT expression levels in Zfp366tdTomato/+ splenocytes. Cell populations were gated as indicated. cDC1 (PILy6cLy6gSiglecHCD64MHCII+CD11c+XCR1+CD11b), cDC2 (PILy6cLy6gSiglecHCD64MHCII+CD11c+XCR1CD11b+), pDC (PISiglecH+BST2+CD11c+), RPMs (PIF4–80+CD64+CD11bMHCII+), monocyte (PILy6chiLy6glowSiglecHCD64CD11b+), granulocyte (PILy6clowLy6ghiSiglecHCD64CD11b+), NK cells (PINK1.1+CD19TCRβ), B cells (PICD19+MHCII+NK1.1TCRβ), CD8+ T cells (NK1.1CD19TCRβ+CD8+CD4), and CD4+ T cells (NK1.1CD19TCRβ+CD8CD4+). Gray histogram shows WT cDC1 as a background control. (C) Dot plot shows the gating strategy for tdTomato-expressing cells in the peripheral LNs from WT and Zfp366tdTomato/+ mice. SSC-A, side scatter area. (D) Quantitation of tdTomato expression in cDC1 and cDC2 from the indicated tissue of Zfp366tdTomato/+ mice. Gut (cDC1: PIMHCII+CD11c+CD64CD103+XCR1+CD11b, cDC2s: PIMHCII+CD11c+CD64CD103+XCR1CD11b+) and Peyer’s patch (PP) and mesenteric LN (MLN; cDC1: PIMHCII+CD11c+CD64XCR1+CD11b, cDC2: PIMHCII+CD11c+CD64XCR1CD11b+). Migratory (m) cDCs (PICD64MHCIIhiCD11clow). Resident (r) cDCs (PICD64MHCIIlowCD11chi). Thymus (cDC1: SiglecHCD64MHCII+CD11c+XCR1+CD11b, cDC2: SiglecHCD64MHCII+CD11c+XCR1CD11b+). Liver and lung (cDC1: PIMHCII+CD11c+CD64CD103+XCR1+CD11b, cDC2: PIMHCII+CD11c+CD64CD103XCR1CD11b+). MFI, mean fluorescence intensity. (E) Splenic B cells, cDC1, and cDC2 (tdTomatolow/high) were sorted from Zfp366tdTomato/+ mice. The expression level of the indicated was quantified by RT-qPCR. Gapdh was used as housekeeping gene. (F) Quantitation of tdTomato in epidermal LCs (CD45+MHCII+CD11c+CD11b+), lung alveolar macrophages (CD45+SiglecF+CD11c+CD64+F4/80+CD11b), and brain microglia (CD45+CD11b+MHCII+CD68+) of WT and Zfp366tdTomato/+ mice. Data in (D) and (E) are the means ± SEM, with each dot representing one mouse. Data represent at least three independent experiments.

Analysis of splenocytes isolated from Zfp366-tdTomato/+ mice highlighted the high tdTomato expression in cDC1 and its slightly lower and more varied expression in cDC2 (Fig. 1B). Similarly gating on tdTomato+ splenocytes revealed the selective expression of DC-SCRIPT in red pulp macrophages (RPMs; CD11blowF4/80+), cDC1s (F4/80SiglecHCD11c+MHCII+XCR1+CD11b), and cDC2s (F4/80SiglecHCD11c+MHCII+XCR1CD11b+) (Fig. 1B and fig. S1A), whereas in the lymph nodes (LNs), tdTomato+ cells were exclusively cDC1s or cDC2s (Fig. 1C). tdTomato fluorescence was also prominent in the migratory and resident cDC1 and cDC2 compartments from the skin-draining LN and several nonlymphoid organs (Fig. 1D and fig. S1B). To confirm that the allele accurately reported Zfp366 expression, we sorted cDC1, cDC2, and B cells according to the fluorescence level of tdTomato found in cDC1 (high), cDC2 (high and low), or B cells (negative). Consistent with the literature, cDC1 expressed higher level of Xcr1, Tlr3, Id2, and Batf3, whereas cDC2 expressed higher levels of Itgam or Csf1r. The expression of tdTomato correlated with the rate of Zfp366 transcription observed in the different DC compartments (Fig. 1E and fig. S1C).

To assess when DC-SCRIPT is expressed during DC development, we analyzed wild-type (WT) and Zfp366-tdTomato/+ BM progenitors by flow cytometry. CDPs (defined as LinIL-7RCD117intCD135+CD115+CD11c) and pre-cDC2 (LinCD117lowCD135+CD11c+MHCIIlow/-CD115CD127SiglecHLy6c+) were tdTomato negative, whereas tdTomato fluorescence was detected in about 40% of pre-DC1 (LinCD117lowCD135+CD11c+MHCIIlow/−CD115CD127SiglecHLy6c) (fig. S2, A and B). Zfp366 transcription in pre-DC1s, but not pre-cDC2s, was also observed in our reanalysis of publicly available single-cell RNA sequencing from the corresponding populations (fig. S2C) (6).

As noted above, examination of the non-cDC compartment of the spleen revealed tdTomato signal in RPMs, although it was undetectable in other myeloid or lymphoid cells including pDCs (Fig. 1B). Analysis of other tissue-resident myeloid cells revealed tdTomato fluorescence in epidermal Langerhans cells (LCs) but not brain microglia or alveolar macrophages (Fig. 1F). Thus, within the hematopoietic system, DC-SCRIPT has a very selective expression pattern, marking cDCs, their committed progenitors, and selected tissue-resident macrophages.

DC-SCRIPT ablation impairs DC development

To develop a mouse model of DC-SCRIPT deficiency, we used CRISPR-Cas9 technology to delete the entire Zfp366 coding region (exons 2 to 5) from mouse chromosome 13 (Fig. 2A). Western blot analysis confirmed the absence of DC-SCRIPT in both splenic cDC1 and cDC2 isolated from Zfp366−/− mouse (Fig. 2B). In concordance with its selective expression profile in immune cells, DC-SCRIPT ablation did not impede the development of B, T, or natural killer (NK) cells. The frequency of pDCs in the BM, but not the spleen, was increased in the absence of DC-SCRIPT (fig. S3, A and B). In contrast, the absence of DC-SCRIPT resulted in a substantial reduction in the numbers of cDC1s in the spleen, whereas cDC2s numbers were modestly increased (Fig. 2C).

Fig. 2 Decreased cDC1s in the absence of DC-SCRIPT.

(A) Schematic representation (not to scale) of the WT and deleted Zfp366 locus and the two sgRNAs. (B) Western blot for DC-SCRIPT in WT and Zfp366−/− splenic cDC1 and cDC2 [gated as in (C)]. ACTIN serves as a protein loading control. (C) Dot plot shows the proportion of pregated splenic cDCs (PICD45+SiglecHMHCII+CD11c+) that are cDC1 (CD172aXCR1+) or cDC2 (+CD172a+XCR1) from WT and Zfp366−/− mice. Bar graphs show the mean total cell number for the indicated cDC subset. (D) Bar graphs represent the mean relative percentage for the indicated cDC subset in the peripheral LN resident (r) and migratory (m) cDCs, liver, lung, and MLN resident (r) and migratory (m) cDCs (gated as in Fig. 1). (E) Histogram shows the expression of indicated cell surface marker in WT or Zfp366−/− splenic cDC1 and cDC2 [gated as in (C)]. (F) Dot plot shows the IRF8 and XCR1 expression in splenic cDC [gated as in (C)] from WT, Zfp366−/−, and ItgaxcreIrf8fl/fl mice. Histogram shows IRF8 expression in WT, Zfp366−/−, or ItgaxcreIrf8fl/fl splenic pDC, cDC2, and cDC1. Bar graph shows MFI ± SEM of IRF8 in pDC, cDC2, and cDC1; ItgaxcreIrf8fl/fl DCs were used as negative control. Data in (C), (D), and (F) are the means ± SEM, with each dot representing one mouse. P values are calculated using unpaired Student’s t test for cDC1 samples in (D) and (F) and one-way ANOVA for pDC and cDC2 samples in (F). *P < 0.05, **P < 0.01, and ***P < 0.001.

cDC2s can be separated in two distinct subsets according to their expression of the NOTCH2-dependent endothelial cell-selective adhesion molecule (ESAM). DC-SCRIPT deficiency did not influence the frequency of ESAM+ cDC2 compared with WT controls (fig. S3C). Recently, an alternative model for cDC2 diversity has been proposed on the basis of the expression of transcription factors TBET and RORγT to define the cDC2a and cDC2b subsets, respectively (28). Because cDC2b is defined by the coexpression of CLEC10a and CLEC12a, we analyzed expression of these cell surface markers on splenic cDC2s isolated from WT and Zfp366−/− mice by flow cytometry. Using these criteria, we observed a similar frequency of cDC2a and cDC2b in Zfp366−/− mice when compared with their WT counterparts (fig. S3D). Together, these observations suggest that DC-SCRIPT is dispensable for the development of both cDC2 subsets.

In contrast to the relatively normal cDC2 development, and consistent with the reduced number of cDC1s observed in the spleen in Zfp366−/− mice, proportion of cDC1 in total cDCs was decreased in peripheral tissues (including lung and liver) and also in cDCs resident in LNs (including peripheral and mesenteric LNs), whereas migratory cDC1s in these LNs were unaffected by the loss of DC-SCRIPT (Fig. 2D). The remaining resident cDC1s had an abnormal phenotype with reduced levels of MHCII and cDC lineage defining markers such as CLEC9a, CD103, or XCR1 and increased expression of cDC2 markers such as CD11b and CD172a (Fig. 2E).

These abnormalities led us to question the identity of the residual cDC1s in Zfp366−/− mice. cDC1s characteristically express high amounts of IRF8, which is required for their initial commitment and subsequent maturation (18, 29, 30). Consistent with the literature, pDCs and cDC1s isolated from WT animals expressed high amounts of IRF8, whereas cDC2s expressed minimal IRF8 (Fig. 2F). The concentration of IRF8 in Zfp366−/− pDCs was similar to their WT counterparts (Fig. 2F). Although there was a twofold reduction in the expression of IRF8 in Zfp366−/− cDC1, this was still much higher than for cDC2 (Fig. 2F). This suggests that the residual cDC1s observed in the absence of DC-SCRIPT represent bona fide cDC1, albeit with reduced expression of IRF8.

To confirm these observations in vitro, BM progenitors were isolated from WT and Zfp366−/− mice and stimulated with Flt3L for 7 days. Flow cytometric analysis highlighted a relative increase in pDC and cDC2 differentiation in culture originating from Zfp366−/− BM, whereas cDC1 development was substantially reduced compared with their WT counterparts (fig. S3E). Intracellular staining confirmed the reduced expression of IRF8 in cDC1 derived from Zfp366−/− BM progenitors (fig. S3F).

To determine the development stage at which cDC1s become dependent on DC-SCRIPT, we analyzed the pre-cDC1 compartment of the BM where CD226 expression on LinCD117CD135+ progenitors defines a specified pre-cDC1 progenitor (18). Consistent with this notion, we found that DC-SCRIPT expression in the BM was restricted to CD226+ pre-cDC1 progenitors (fig. S4A). Using this gating strategy, the number of pre-cDC1 cells in the BM was unaffected by DC-SCRIPT deficiency, although the number of pre-cDC1 seeding in the spleen was greatly reduced, suggesting that the initial requirement for DC-SCRIPT function is at this developmental stage (fig. S4, B and C).

Although DC-SCRIPT is also expressed in LCs and RPMs, analysis by flow cytometry revealed normal development of epidermal and skin-draining LN LCs (fig. S4, D to F) and splenic RPMs in Zfp366−/− mice when compared with their WT counterparts (fig. S4G). Together, these observations suggest that, despite being expressed in several other immune cells, in unchallenged mice, DC-SCRIPT deficiency mainly impairs the development of cDC1.

Cell-intrinsic requirement for DC-SCRIPT in cDC1

DC-SCRIPT expression has also been demonstrated in nonhematopoietic tissue (31, 32). Thus, to confirm that the reduction in cDC1s that we observed with DC-SCRIPT loss was cell intrinsic, we generated mixed BM chimeras, whereby lethally irradiated CD45.1+CD45.2+ hosts were reconstituted with a 1:1 BM mix of either WT CD45.1+:Zfp366−/− CD45.2+ or WT CD45.1+:WT CD45.2+. Analysis by flow cytometry of the spleen and LNs from these chimeric mice corroborated our findings that DC-SCRIPT is dispensable for the generation of B, T, and NK cells because WT and Zfp366−/− genotypes contributed equally to these populations (fig. S5, A and B). In contrast, loss of DC-SCRIPT was detrimental for the production of cDC1s in the spleen, lung, and gut, whereas the frequency of cDC2s increased in these organs (Fig. 3, A to C). Of note, the contribution of Zfp366−/− BM progenitors to the pDC compartment in the BM was also found to be increased when compared with the WT:WT chimeras (Fig. 3D). The cDC1 development defect observed in the Zfp366−/− compartment was accompanied by a reduced expression of IRF8 in the cDC1 (Fig. 3E), thus confirming our earliest observations (Fig. 2F). Together, these results highlight the cell-intrinsic role of DC-SCRIPT in controlling DC development at multiple anatomical sites.

Fig. 3 Cell-intrinsic control of cDC1 differentiation by DC-SCRIPT.

(A to D) Flow cytometric analysis of cDC1, cDC2, and/or pDC from the (A) spleen, (B) lung, (C) gut, and (D) BM of WT or Zfp366−/− mix chimeric mice. Chimeric mice were analyzed 6 weeks after BM reconstitution. CD45.1+CD45.2+ host cells were excluded. Donor cells were identified as CD45.1+ or CD45.2+ as appropriate and gated as in Fig. 1. Bar graphs represent the mean relative percentage for the indicated DC subset. Data are from six mice at least per group (A) and three mice per group (B to D). (E) Bar graphs represent the mean MFI of IRF8 in splenic cDC1, cDC2, and pDC in WT or Zfp366−/− mix chimeric mouse. IRF8 was detected via flow cytometry. Data are means ± SEM. P values compare the proportions of CD45.2+ WT and CD45.2+ Zfp366−/− cells in each indicated population using unpaired Student’s t test for (A) to (D) and paired Student’s t test for (E). *P < 0.05, **P < 0.01, and ***P < 0.001.

DC-SCRIPT controls cDC1 identity

To gain insights into the molecular function of DC-SCRIPT in DCs, we performed RNA-seq on splenic cDC1s and cDC2s isolated from WT or Zfp366−/− mice. Analysis of the Zpf366 locus confirmed the deletion of exons 2 to 5 (fig. S6A). Comparison of the data from WT and Zfp366−/− cDC1s identified 2584 genes activated by DC-SCRIPT, whereas 3034 genes were repressed by DC-SCRIPT in cDC1s (Fig. 4A and table S1). In contrast, deletion of DC-SCRIPT in cDC2 highlighted a milder effect on the transcriptome of cDC2, with 317 genes being activated by DC-SCRIPT and 308 genes being repressed (Fig. 4B and table S2). Thus, DC-SCRIPT deficiency has far more impact on the transcriptome of cDC1 than cDC2. Analysis of the top 100 differentially expressed genes between WT cDC1 and cDC2 revealed a gain of expression for transcripts associated with the cDC2 lineage in Zfp366−/− cDC1. These included the transcripts Itgam, Cd4, Irf4, Sirpb1a, Clec4b1, and Clec4a2 among others (Fig. 4C and table S3). In contrast, transcripts defining WT cDC1 such as Cd24a, Irf8, or Tlr3 were down-regulated in Zfp366−/− cDC1s. These results are in line with our earlier observations highlighting the reduced expression of several cell surface markers associated with the cDC1 lineage and the gain of cDC2 lineage-defining markers in Zfp366−/− cDC1s (Fig. 2E and table S1).

Fig. 4 Transcriptomic analysis of DC-SCRIPT–deficient cDCs.

(A and B) Multidimensional scaling (MDS) plot shows the differentially expressed (DE) genes (log2FC > 1.2, adjusted P < 0.05), which were up-regulated (red) and down-regulated (blue) in Zfp366−/− cDC1 (A) and cDC2 (B) compared with equivalent WT cells. Triplicate samples for each genotype and cell type were analyzed. FC, fold change. (C) Heatmap shows the top 100 DE genes between WT cDC1 and cDC2 and the corresponding gene expression in Zfp366−/− cDC1 and cDC2. The location and identity of genes of interest are indicated. (D and E) Gene set enrichment analysis for the cDC1 DE genes compared with the cDC1 (D) and cDC2 (E) signature genes. Genes are ordered right to left from most up-regulated (pink) to most down-regulated (blue) in Zfp366−/− versus WT cDC1 or cDC2. Signature genes are marked by vertical lines. Worm shows relative enrichment of the signature genes. (F and G) Expression in RPKM of the indicated cDC marker genes (F) and transcriptional regulators (G). Data are from (A) and (B). Data are means ± SEM with each dot representing one sample. P values are calculated using unpaired Student’s t test. **P < 0.01 and ***P < 0.001.

To more globally address the impact of DC-SCRIPT loss on the cDC1-specific gene expression program, we performed a gene set enrichment analysis using published gene signatures (33). This strategy revealed that cDC1 signature genes, such as Xcr1 or Itgae, were significantly down-regulated in the absence of DC-SCRIPT, whereas genes up-regulated in Zfp366−/− cDC1s were significantly enriched for genes belonging to the gene signatures of cDC2 (including Itgam and Sipr1a; Fig. 4, D to F) and pDCs (Tcf4 and Runx2; fig. S6, B and C) (3437).

Given the extent of the transcriptomic reprogramming occurring in Zfp366−/− cDC1s, we turned our attention to the expression of the cDC1 lineage–specific transcription factor Irf8, Batf3, Id2, Nfil3, Etv6, or Zeb2 (14). We observed a slight increase in Zeb2 expression in Zfp366−/− cDC1s; however, this did not correlate with a decrease in Id2, a known target of Zeb2 repression, which could have offered an explanation for the reduced number of Zfp366−/− cDC1s (Fig. 4G) (3840). Among the key transcription factors analyzed, only Irf8 transcripts were down-regulated in Zfp366−/− cDC1s (Fig. 4G and table S1). This is in line with the twofold decrease in IRF8 protein observed earlier (Fig. 2F and fig. S3F). Because cDC1 development is sensitive to Irf8 haploinsufficiency (41, 42), our results suggest that the down-regulation of IRF8 observed in DC-SCRIPT–deficient cDC1 potentially explains the perturbed development and identity of this compartment, a hypothesis that we subsequently pursue.

Genome-wide analysis of DC-SCRIPT DNA binding shows correspondence with BATF3 and IRF8 binding in cDC1

To gain insight into the molecular mechanisms underpinning DC-SCRIPT function in DCs, we adopted cleavage under targets and tagmentation (CUT&Tag) to identify the genes targeted by DC-SCRIPT in splenic cDC1s and cDC2s (43). We found that DC-SCRIPT bound to 31,473 sites in cDC1s, whereas 2776 regions were significantly enriched in cDC2. About 60% of the cDC2 peaks were shared with cDC1 peaks (Fig. 5A and table S4). DC-SCRIPT binding in the gene promoter regions and sequences surrounding a transcription starting site (−3 kb > TSS > +1 kb) accounted for 39 and 46% of the regions bound by DC-SCRIPT in cDC1s and cDC2s, respectively. Distal binding (>5 kb from a gene body) represented about 30% for both cDC1s and cDC2s, whereas intragenic binding represented about 32 and 23% of the binding sites detected in cDC1s and cDC2s, respectively (fig. S7A). Our analysis revealed that DC-SCRIPT binding was associated with genes such as Xcr1, Itgae, Itgam, or Sirp1a, whose expression was affected by DC-SCRIPT deficiency (Fig. 4F and fig. S7B).

Fig. 5 DC-SCRIPT controls the development of cDC1 via enhancing IRF8 expression.

(A) Venn diagram showing the overlap between DC-SCRIPT binding regions (CUT&Tag) in cDC1 (left) and cDC2 (right). The total number of bound regions is listed above the diagram. The number of binding peaks (×103) in each category is indicated. (B and C) Heatmap showing the cobinding of DC-SCRIPT with IRF8 and BATF3 in the open chromatin regions (OCR) and the RNA-seq DE genes in cDC1 (B) and cDC2 (C). (D) Venn diagram showing the overlap between DC-SCRIPT, IRF8, and BATF3 binding regions in cDC1 and (E) between DC-SCRIPT and BATF3 binding regions in cDC2. The number of binding peaks (×103) in each category is indicated. (F) Normalized sequencing tracks of CUT&Tag with anti–DC-SCRIPT antibody in WT or Zfp366−/− cDC1 and cDC2, and ChIP-seq with anti-IRF8 or anti-BATF3 antibodies in the indicated populations. The +32-kb Irf8 enhancer is shown as boxed. (G) Schematic representation of the DC-SCRIPT overexpression in mix chimeric mouse BM. Mice were analyzed 6 weeks after BM reconstitution. (H) Histogram shows the expression of intracellular IRF8 staining by flow cytometry from the indicated populations, manipulated as indicated in (G). (I) MFI of IRF8 in cDC1 and cDC2 in WT or Zfp366−/− mix chimeric BM cultures as in (H). Lines connect cDCs from the same chimeric BM culture. Each dot represents an independent experiment. (J) WT or Zfp366−/− BM progenitors were transduced with empty, IRF8, or DC-SCRIPT retroviruses (all expressing GFP). The frequency of GFP+ pDCs (SiglecH+CD11c+), cDC1 (SiglecHMHCII+CD11c+XCR1+CD11blow), and cDC2 (PIGFP+SiglecHMHCII+CD11c+XCR1CD11b+) was determined by flow cytometry. (K) Bar graph shows the means ± SEM percentage of pDC and cDC1 for the data from (J). Each dot represents one independent experiment. P values are calculated using two-way ANOVA with Tukey’s multiple comparison test. *P < 0.05, **P < 0.01, and ***P < 0.001.

Given the extent of the transcriptional changes in Zfp366−/− cDC1s and the magnitude of DC-SCRIPT binding in this subset, we next asked whether DC-SCRIPT binding was found in chromatin regions with increased accessibility and in close proximity of other transcription factors associated with cDC1 differentiation. To this end, we retrieved a publicly available chromatin immunoprecipitation dataset for the binding of IRF8 and BATF3 in cDC1s and BATF3 in cDC2 (41) and chromatin accessibility data from both subsets (44). DC-SCRIPT bound almost exclusively in open chromatin regions, with 47 and 40% of the regions containing either a BATF3 or an IRF8 peak, respectively (Fig. 5, B to D; fig. S8A, and table S5). In contrast, DC-SCRIPT was found in only 10% of the BATF3-bound regions in cDC2s (Fig. 5, C and E; fig. S8B; and table S5). Overall, 33% of DC-SCRIPT binding sites were cobound with BATF3 and IRF8 in cDC1s (Fig. 5, B and D).

We then sought evidence for the direct regulation of Irf8 by DC-SCRIPT. Recent studies have pointed to a critical role for the +32-kb distal region of Irf8 in regulating its own expression through the binding of IRF8 and BATF3 (18, 41). In cDC1s, DC-SCRIPT binding was detected in the promoter and also in the +32-kb region cobound by IRF8 and BATF3 (Fig. 5F). In contrast, these regions were devoid of DC-SCRIPT and BATF3 binding in cDC2s (Fig. 5F). Together, these results point to the critical role for DC-SCRIPT in coordinating the expression of Irf8 through its binding to the +32-kb enhancer.

DC-SCRIPT overexpression up-regulates IRF8 in cDC1s

The above findings suggest that DC-SCRIPT promotes cDC1 differentiation, at least in part through its regulation of IRF8 concentration. To test this hypothesis, we used a gain-of-function approach to overexpress DC-SCRIPT in BM progenitors isolated from 1:1 BM mix of either WT Ly5.1+:Zfp366−/− Ly5.2+ or WT Ly5.1+:WT Ly5.2+ (Fig. 5G). The retroviral vector coexpressed DC-SCRIPT and green fluorescent protein (GFP), allowing transduced cells to be defined as GFP+. Whereas DC-SCRIPT overexpression did not initiate IRF8 expression in cDC2s, there was a substantial increase of IRF8 abundance in Zfp366−/− cDC1s overexpressing DC-SCRIPT (Fig. 5, H and I). We then asked whether increasing IRF8 or DC-SCRIPT concentration in Zfp366−/− BM progenitors could rescue their potential to differentiate into cDC1s. To this end, WT or Zfp366−/− BM progenitors were transduced with retroviruses expressing GFP alone, or GFP and IRF8 or DC-SCRIPT. In agreement with our previous report, increased DC-SCRIPT expression in WT BM progenitors was detrimental for pDC differentiation while promoting cDC1 potential (Fig. 5, J and K). Complementation of Zfp366−/− BM progenitors with either DC-SCRIPT or, most notably, IRF8 restored their capacity to efficiently differentiate into cDC1s (Fig. 5, J and K). Together, these data show that DC-SCRIPT is instrumental in maintaining optimal levels of IRF8 in the cDC1 compartment, thereby controlling their full developmental potential.

Capture and presentation of cell-associated antigens is regulated by DC-SCRIPT

A cardinal feature of cDC1 is their capacity to present antigens via MHCI or MHCII (45). To directly compare the capacity of WT and DC-SCRIPT–deficient cDCs to capture and present antigens from dying cells, dye-labeled MHCI- and MHCII-restricted ovalbumin (OVA)–specific T cells (OT-I and OT-II, respectively) were transferred into WT and Zfp366−/− BM chimeric mice. To address cross-presentation capability, mice were immunized with OVA-coated irradiated splenocytes isolated from MHCI-deficient Bm1 mice (a model of cell-associated antigen), and the proliferation of OT-I cells was analyzed 3 days later. OT-I cell proliferation was substantially reduced in the Zfp366−/− group compared with their WT counterparts, thus suggesting a reduced capacity to capture and/or present cell-associated antigen onto MHCI (Fig. 6A). Because the reduced number of cDC1s present in Zfp366−/− group could account for the decreased OT-I cell proliferation, we turned to an in vitro antigen presentation assay where cDC1s numbers can be equilibrated between the genotypes and tested across a range of antigen doses. Similar to the in vivo findings, the lack of DC-SCRIPT substantially reduced the capacity of cDC1s to stimulate OT-I, whereas as expected, both WT and Zfp366−/− cDC2 lacked antigen cross-presentation capability (Fig. 6, B and C).

Fig. 6 Impaired antigen presentation in DC-SCRIPT–deficient mice.

(A) Representative flow cytometric analysis of in vivo cross-presentation of cell-associated antigen. Irradiated 2 × 107 OVA-loaded Bm1 splenocytes were injected intravenously into WT or Zfp366−/− chimeric mice, followed 1 day later by CTV-labeled CD45.1 OT-I cells. Mice were harvested 3 days after splenocyte injection. Data are pregated on CD45.1+CD8+TCRvα2+ cells, and the bar graph shows the number of CTVlow OT-I cells in the spleen for each genotype. Data are from five samples per condition. Nonimmunized mice received OT-I cells only and serve as controls. (B and C) Sorted splenic WT or Zfp366−/− cDC1s (B) and cDC2s (C) were assayed for cross-antigen presentation to OT-I in response to the indicated concentrations of OVA-loaded irradiated splenocytes from Bm1 mice. Data are representative of three independent experiments. (D) Representative flow cytometric analysis of in vivo presentation of cell-associated antigen. Irradiated 2 × 107 OVA-loaded MHCII-deficient splenocytes were injected intravenously into WT or Zfp366−/− chimeric mice, followed 1 day later by CD45.1 OT-II cells. Mice were harvested 3 days after splenocyte injection. Data are pregated on CD45.1+CD4+TCRvα2+ cells, and the bar graph show the number of CTVlow OT-II cells in spleen for each genotype. Data are from six samples per condition. Nonimmunized mice received OT-II cells only and serve as controls. (E and F) Sorted splenic WT or Zfp366−/− cDC1s (E) and cDC2s (F) were assayed for direct antigen presentation to OT-II in response to the indicated concentrations of OVA-loaded irradiated splenocytes from MHCII-deficient mice. Data are representative of three independent experiments. (G) Representative flow cytometric analysis of in vivo phagocytosis of cell-associated antigen. Irradiated 2 × 107 OVA-loaded BALB/c splenocytes were labeled with PKH26 and then injected intravenously into WT or Zfp366−/− chimeric mice. Spleen were harvested 16 hours after splenocyte injection. PKH26 signal was tested in B cells, pDC, cDC2, and cDC1. Data are from at least nine mice. Chimeric mice were analyzed 12 weeks after BM reconstitution. Data are shown as means ± SEM. Each dot represents one mouse. (A and D) P values are calculated using an unpaired Student’s t test. (B, C, E, and F) P values calculated are using two-way ANOVA with Tukey’s multiple comparison test. **P < 0.01 and ***P < 0.001.

We next examined the role of DC-SCRIPT in the direct presentation of cell-associated antigen to OT-II cells. CellTrace Violet (CTV)–labeled OT-II cells were adoptively transferred into either WT or Zfp366−/− BM chimeric mice. The following day, mice were immunized with lethally irradiated OVA-coated MHCII-deficient splenocytes, and proliferation of OT-II was measured by flow cytometry (Fig. 6D). OT-II cell proliferation was reduced in Zfp366−/− chimeric mice, suggesting that MHCII antigen presentation is also affected in the absence of DC-SCRIPT (Fig. 6D). When similar assays were performed in vitro using identical numbers of cDC1s, DC-SCRIPT–deficient cDC1s were also impaired in their capacity to stimulate OT-II cells (Fig. 6E). We did not detect efficient priming of OT-II cells by cDC2s in these assays (Fig. 6F).

In an attempt to gain insight into the mechanism associated with the impaired cross-priming of OT-I cells by model antigen associated to apoptotic cells, we assessed the capacity of DC-SCRIPT–deficient DCs to capture this form of antigen in vivo. To this end, OVA-loaded BALB/c splenocytes were labeled with the membrane dye PKH26, irradiated, and transferred into WT or Zfp366−/− mice. The extent of PKH26 uptake in B cells and DC subsets was measured by flow cytometry 16 hours later. Consistent with earlier reports (46), cDC1s had a greater capacity to efficiently phagocytose the labeled cells than cDC2s (Fig. 6G). However, the capture of apoptotic-labeled cells was substantially reduced in Zfp366−/− cDC1s (Fig. 6G). Together, these data show that DC-SCRIPT deficiency leads to reduced cross-priming and diminished uptake of cell-associated antigen from dying cells.

We next tested the capacity of WT and DC-SCRIPT–deficient cDCs to present peptide antigens, an activity that bypasses defects in antigen uptake and intracellular processing. To this end, CTV-labeled OT-I or OT-II cells were cocultured with cDC1s or cDC2s in the presence of OVA peptides 257 to 264 and 323 to 339, respectively. DC-SCRIPT–deficient cDCs promoted the robust proliferation of OT-I and OT-II T cells, indicating that DC-SCRIPT is dispensable for T cell stimulation in response to MHC-peptide complexes displayed on the cell surface (fig. S9, A and B). Together, these data demonstrate that the impaired T cell stimulation evoked by DC-SCRIPT–deficient cDC1s is caused by defects in phagocytosis and subsequent presentation of MHC-peptide complexes derived from cell-associated antigens.

DC-SCRIPT directly controls Il12b expression

To further characterize the functionality of DC-SCRIPT–deficient cDCs, we examined their capacity to respond to TLR agonists. Both BM-derived WT and Zfp366−/− cDC1s or cDC2s responded to the TLR ligand lipopolysaccharide (LPS) or CpG, by increasing their surface expression of MHCII, CD80, and CD86 (Fig. 7A and fig. S10A), canonical markers of DC maturation. CD40 was also up-regulated, although to a slightly reduced extent in Zfp366−/− cDC1s or cDC2s when compared with the levels observed in their WT counterparts (Fig. 7A). Despite this near-normal maturation response, we detected a substantial decrease in the capacity of DC-SCRIPT–deficient cDC1s (both in vivo and in vitro) to produce IL-12p40 in response to LPS or CpG (Fig. 7B and fig. S10B). The defect in IL-12p40 secretion was intrinsic to Zfp366−/− cDC1s, as it was also observed in BM cultures isolated from mixed chimeric mice (Fig. 7C). A broader analysis of the cytokines and chemokines present in the supernatants isolated from OVA 257 to 264 peptide pulsed cDC1 cocultured with OT-I cells revealed a decrease for only two cytokines: IL-12p40 and IFN-γ (fig. S10C). In contrast, culture supernatants isolated from OVA 323 to 339 pulsed cDC2s cocultured with OT-II T cells revealed no significant differences for the array of cytokines/chemokines tested (fig. S10D). These results suggest that DC-SCRIPT is a critical regulator of IL-12p40 in response to TLR agonism or antigen stimulation. This feature was specific to cDC1s because in vitro cultured monocyte-derived DCs and monocyte-derived macrophages generated from Zfp366−/− mice produced similar amounts of IL-12p40 in response to CpG stimulation than their WT counterparts (fig. S10, E and F).

Fig. 7 DC-SCRIPT controls IL-12p40 production in cDC1s.

(A) In vitro–generated WT or Zfp366−/− cDC1 and cDC2 were stimulated with CpG or LPS for 24 hours, and the expression of CD80, CD86, and CD40 was examined by flow cytometry. Heatmap shows the z score that was calculated on the basis of the MFI of CD80, CD86, and CD40 in cDC1 and cDC2. (B) Splenic CD11c+ cells were enriched by magnetic beads then stimulated with CpG for 4 hours. The expression of IL-12p40 was detected by intracellular staining. Bar graph shows the mean percentage of IL-12p40+ cDC1 (gated as PISiglecHMHCII+CD11c+XCR1+CD11b). (C) Mixed BM chimeric mice were generated and allowed to reconstitute for 12 weeks. BM cells were isolated from mixed chimeric mice, then cultured with FLT3L for 11 days, and stimulated with CpG for 6 hours. The expression of IL-12p40 was analyzed via intracellular staining. cDC1 were gated as in (B). Bar graph shows the mean relative contribution of IL-12p40+ cDC1s from each genotype. (D) In vitro–generated cDC1 and cDC2 were stimulated with CpG for 4 hours and then sorted. The expression level of IL12b was detected by RT-qPCR. Gapdh was used as housekeeping gene. (E) Sample tracks of DC-SCRIPT CUT&Tag from WT and Zfp366−/− cDC1 and cDC2 for the Il12b gene. (F) WT or Zfp366−/− BM progenitors were transduced with empty or DC-SCRIPT retrovirus at day 3 and then stimulated with CpG at day 9 for 6 hours. The IL-12p40 production in cDC1 [gated as in (B)] was assessed by intracellular staining. Transduced cells were marked as GFP+. (G) Bar graph shows the mean percentage of IL-12p40+ cDC1 for the data from (F). (H) Parasite burden of WT or Zfp366−/− chimeric mice injected with 10,000 Pru.tdTomato T. gondii tachyzoites 8 days earlier. Bar graph shows the mean percentage of tdTomato+ parasites. SSC-A, side scatter area. Data shown are the means ± SEM. Each dot represents one independent experiment (B, D, and G) or mouse (H). P values are calculated using (D) one-way and (G) three-way ANOVA with Tukey’s multiple comparison test. **P < 0.01 and ***P < 0.001.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR) analysis revealed that the CpG-induced regulation of Il12b (encoding IL-12p40) by DC-SCRIPT was transcriptional in cDC1s (Fig. 7D). The regulation appears to be direct because DC-SCRIPT bound in close proximity to the Il12b transcriptional start site, as well as 5′ and 3′ to the gene (Fig. 7E). To test this hypothesis, we performed a gain-of-function experiment, whereby WT and Zfp366−/− BM-derived DCs were transduced with retroviral vectors expressing either GFP alone or GFP and DC-SCRIPT, and the production of IL-12p40 after CpG stimulation was tested by intracellular staining. Transduction of either WT and Zfp366−/− BM DCs with DC-SCRIPT significantly increased the capacity of cDC1s to produce IL-12p40, compared with control cells (Fig. 7, F and G, and fig. S11, A and B). Together, these results highlight DC-SCRIPT as a critical regulator of IL-12p40 production and support the hypothesis that DC-SCRIPT activity is limiting for IL-12p40 production in cDC1s.

A critical role for cDC1-produced IL-12 has been highlighted in response to Toxoplasma gondii infection (47, 48). To test whether DC-SCRIPT deletion resulted in an increased susceptibility to T. gondii infection, we reconstituted lethally irradiated hosts with WT or Zfp366−/− BM. Mice were infected intraperitoneally with 10,000 tachyzoites of the Pru:tdTomato strain. In this chimera model, mice started to lose weight at day 5 after infection, and all mice had lost >15% of their initial body weight and had to be euthanized by day 8 (fig. S11C). Chimeric mice reconstituted with Zfp366−/− BM lost weight more rapidly and harbored a higher parasite burden than WT control chimeras (Fig. 7H and fig. S11C). Analysis of serum on day 4 after T. gondii infection revealed significantly lower IL-12p40 concentration in chimeric mice reconstituted with Zfp366−/− BM, when compared with their WT counterparts (fig. S11D). The decrease in IL-12p40 production was mirrored by a decreased serum concentration of IFN-γ in chimeric mice reconstituted with Zfp366−/− BM (fig. S11D). In summary, these studies show that DC-SCRIPT is a major regulator of cDC1 functionality, including promoting the production of IL-12p40 that is essential for an efficient response to parasite infection.


This study extends our understanding of DC development by characterizing the function of a previously poorly known transcription factor, DC-SCRIPT. By developing a mouse reporter strain, Zfp366tdTomato, we have defined DC-SCRIPT expression at the single-cell level. DC-SCRIPT expression could be detected in a fraction of clonogenic progenitors committed to the cDC1 lineage in the BM, whereas it was absent in cDC2 progenitors. These observations are reminiscent of findings made using the Zbtb46GFP reporter mice, where ZBTB46 expression in the BM was restricted to DC progenitors developing exclusively into cDC1s (41). Thus, our study supports the notion that some degree of commitment to distinct branches of cDCs occurs in the BM, as suggested earlier in both mice and humans (41, 49, 50). Upon seeding into peripheral organs, both cDC1s and cDC2s expressed substantial amounts of DC-SCRIPT, with the former having the highest level, yet again concordant with the ZBTB46 pattern of expression in mature cDCs. However, in contrast to the normal cDC development observed in the absence of ZBTB46 in the steady state, DC-SCRIPT ablation resulted in a reduced cDC1 potential, whereas cDC2 differentiation was slightly increased.

The remaining cDC1s observed in DC-SCRIPT–deficient mice displayed an aberrant cell surface phenotype with reduced expression of markers commonly used to identify this lineage (XCR1, CD103, or CLEC9a) and increased expression of canonical cDC2 markers (CD11b or CD172a). This altered identity was not limited to cell surface markers because transcriptional analysis showed that DC-SCRIPT positively controls nearly half (44%) of the cDC1 gene signature, whereas only 8% of the cDC1 signature was negatively regulated by DC-SCRIPT. We also noted that the DC-SCRIPT–deficient cDC1s gained pDC and cDC2 gene signatures, suggesting that DC-SCRIPT acts as a repressor of these programs in cDC1s. Curiously, despite its conspicuous expression, DC-SCRIPT played an unremarkable role in cDC2s, with <5% of genes belonging to the cDC2 signature being controlled by DC-SCRIPT. These observations suggest that DC-SCRIPT may coordinate the cDC1 program with a yet-to-be-defined partner(s) that is lacking in cDC2. In keeping with this hypothesis, DC-SCRIPT–bound chromatin regions in cDC1s greatly outnumber those in cDC2s, and these regions are often co-occupied with IRF8 and BATF3 in cDC1s, two transcriptional factors essential for their differentiation (19, 42). Collectively, these data demonstrate that DC-SCRIPT is an integral part of the transcriptional hub controlling the cDC1 lineage.

IRF8 is the lineage-defining transcription factor for cDC1s because it is highly expressed in all cDC1s and absolutely required for their development (15, 29, 5154). The induction of BATF3 along with ID2 in pre-cDC1 is thought to be a decisive step in defining the cDC1 lineage because it results in Irf8 autoactivation. The lack of BATF3 in pre-cDC2s has been proposed to be detrimental for the maintenance of IRF8 in these progenitors, preventing commitment into the cDC1 pathway and instead allowing cDC2 differentiation (41). Our study adds a layer of complexity to the molecular mechanisms controlling Irf8 expression in DC progenitors. As for BATF3, DC-SCRIPT binding was detected in the enhancer located at +32 kb relative to the Irf8 transcriptional start site and was critical in sustaining Irf8 expression in cDC1s. It is notable that the pre-cDC1 phenotype observed in this study mirrors the phenotype observed in Batf3−/− and Irf8+32kb−/− mice (18). DC-SCRIPT is dispensable for the specification of a pre-cDC1 progenitor in the BM but is essential to maintain adequate levels of IRF8 upon further maturation of this lineage, thus confirming the critical role of Irf8 +32-kb enhancer in controlling cDC1 maturation (18, 41).

How DC-SCRIPT regulates the activity of the Irf8 +32-kb enhancer is, at present, unclear but may potentially involve changes to the chromatin accessibility at the site because Batf3 expression per se was not altered in DC-SCRIPT–deficient cDC1s. The expression of DC-SCRIPT in pre-cDC1s, but not pre-cDC2s, and its binding in the +32-kb enhancer, supports the notion suggested earlier that Irf8 autoactivation constitutes an essential step in defining cDC1 fate and that the delay in the expression of Batf3 and Zfp366 in pre-cDC2s, resulting in decreased Irf8 expression, constitutes a key molecular switch in engaging the cDC2 developmental program (6, 41).

Although DC-SCRIPT deficiency resulted in a decreased number of resident cDC1s in all lymphoid and nonlymphoid tissue examined, DC-SCRIPT–deficient migratory cDC1 numbers were unaltered in the skin-draining LN. Similar observations were reported in migratory cDC1s isolated from mice lacking BATF3 (5557), suggesting that a putative compensatory mechanism is in place, beyond the function of DC-SCRIPT and BATF3, to support the development of migratory cDC1s in the skin-draining LN. The nature of this compensatory pathway warrants further investigation. Similarly, despite being abundantly expressed in cDC2, LCs, and RPMs, we have not observed any clear functions for DC-SCRIPT in any of these cell types. Whether this is due to limitations in the assays used to date or due to redundancy with a related transcription factor remains to be explored.

DC-SCRIPT is also expressed in human monocyte-derived DCs, where in vitro studies have suggested that it acts as a co-repressor of nuclear receptors and regulates IL-10 production (2426, 58). To date, however, there has not been any study addressing the role of DC-SCRIPT in the functionality of cDC lineages or its role in protective immunity. We provide evidence that DC-SCRIPT controls two of the cardinal functions of cDC1s, namely, capture and presentation of cell-associated antigens and IL-12 production. Lack of DC-SCRIPT in cDC1s resulted in impaired CD4+ and CD8+ T cell responses to cell-associated antigens, whereas presentation of soluble antigen was independent of DC-SCRIPT. Furthermore, combining genome-wide mapping of DC-SCRIPT binding in cDC1s, transcriptomics, and gain-of-function approaches, this study points to a critical role for DC-SCRIPT in controlling IL-12p40 production by cDC1s. IL-12 is a pivotal inducer of T helper 1 immune response, and consistent with an earlier report demonstrating its critical role in promoting IFN-γ–mediated control of T. gondii infection, we found that DC-SCRIPT–deficient mice were more susceptible to this parasite (47, 48).

Collectively, our work demonstrates that DC-SCRIPT is essential for the normal development of cDC1s. DC-SCRIPT is also required for the cell type–specific functions of cDC1s that are critical for protective immunity. Thus, DC-SCRIPT joins IRF8 and BATF3 as core components of the transcriptional circuitry governing the formation, maintenance, and functionality of cDC1 lineage.


Study design

The central aim of the study was to examine the importance of DC-SCRIPT for the development and function of mouse cDCs. We have generated DC-SCRIPT reporter and knockout mouse strains that were analyzed by flow cytometry. We made extensive use of RNA-seq to assess the changes in gene transcription in cDCs in the absence of DC-SCRIPT and developed CUT&Tag approaches to map the DNA binding by DC-SCRIPT in cDCs. Cell biological approaches were used to assess the impact of DC-SCRIPT loss on key cDC1 functions including antigen presentation, as well as the capacity of activated cDCs to express IL-12. In all assays, sample size was determined on the basis of pilot experiments in our laboratory to allow us to get reliable data to perform statistical analyses and ensure reproducibility. All experiments were replicated in at least two to three independent experiments. The number of individual replicates for each experiment is indicated in the figure legends. Animals were randomly assigned to study groups within each genotype. Mice were age- and sex-matched, and littermate controls were used where possible. The analyses were performed unblinded because mice were grouped according to the genotype and treatment.


Zfp366−/− mice were generated by deleting exons 2 to 5 via CRISPR-Cas9 technology in the Melbourne Advanced Genome Editing Centre (MAGEC) laboratory at Walter and Eliza Hall Institute. Two single-guide RNAs (sgRNAs) of the sequence AGGTGCTAGGAGCCGCCCAT and CGGTCTACCAGGACTAACTG and Cas9 nuclease were introduced into fertilized embryos to generate three founders lacking the entire coding domain of Zfp366. The correct deletion was confirmed by PCR, and sequencing and the founders were bred to C57BL/6 and heterozygous mice crossed to establish the Zfp366−/− strain. The combination of the forward and reverse primers gave an amplicon size of 300 base pairs (bp) for the deleted allele and 282 bp for the WT (see table S7 for primer sequences).

Zfp366-tdTomato mice were generated by the MAGEC laboratory by insertion of an IRES-tdTomato T2A Cre sequence after the endogenous Zfp366 STOP codon using CRISPR-Cas9 technology. Correct integration of the reporter gene was confirmed by conventional PCR (forward and reverse primers gave an amplicon size of 448 bp) and long-range PCR (5′ homology arm forward TGGGCATTGAGATTAGAGATCCTGGACACC and reverse GCTTCCTTCACGACATTCAACAGACCTTGC primers, amplicon size of 3159 bp; 3′ homology arm forward TACACCAAAATTTGCCTGCATTACCGGTCG and reverse CATAGGTAAATGATGGGGACTGGGTGAGCC primers, amplicon size of 3217 bp). The founders were bred to C57BL/6, and breeders were maintained as Zfp366-tdTomato/+ heterozygous mice.

Experiments involved male and female mice of 8 to 12 weeks of age. C57BL/6, C57BL/6-CD45.1, OVA-specific CD8 T cell receptor (TCR) transgenic OT-I mice, and OVA-specific CD4 TCR transgenic OT-II mice were maintained in the animal facility of the Walter and Eliza Hall Institute. CD45.1 OT-I mouse was donated by M. Jenkins (Walter and Eliza Hall Institute). Bm1−/− (MHCI-deficient) mice were from W. Heath (University of Melbourne). BALB/c, MHCII-deficient mice were from J. Villadangos (University of Melbourne). All mice were maintained on a C57BL/6 background and housed in specific pathogen–free conditions. Animal procedures were approved by the Walter and Eliza Hall Institute Animal Ethics Committee.

BM chimeras

To generate standard BM chimeras, 1.5 × 106 BM cells from the CD45.2 Zfp366+/+ or CD45.2 Zfp366−/− mice were intravenously injected into lethally γ-irradiated (2 × 550 rads) C57BL/6 × C57BL/6-CD45.1 F1 or C57BL/6-CD45.1 congenic recipient mice. To generate mixed BM chimeras, BM cells from WT or CD45.2Zfp366−/− mice (both CD45.2) were mixed with CD45.1 WT BM at a 1:1 ratio and then injected intravenously into lethally irradiated C57BL/6 × CD45.1 F1 recipients. The immune cell population from each compartment was assessed by flow cytometry after 6 to 12 weeks.

In vitro generation of DCs

Hips, femurs, and tibias from mice were crushed into fluorescence-activated cell sorting (FACS) buffer [phosphate-buffered saline (PBS) and 0.5% bovine serum albumin; Sigma-Aldrich], and erythrocytes were removed with ACK (ammonium-chloride-potassium) lysis buffer. BM cells were cultured at 1.5 × 106 cells/ml in RPMI 1640 supplemented with 10% heat-inactivated fetal calf serum (FCS), 2 mM l-glutamine (GIBCO), 50 μM 2-mercaptoethanol (Sigma-Aldrich), and penicillin/streptomycin (100 U/ml) (GIBCO) containing FLT3L (200 ng/ml; BioXCell) for 7 days to generate cDCs and pDCs.

Flow cytometry

Single-cell suspensions were resuspended in FACS buffer and stained at 4°C. Viable cells were defined as propidium iodide (PI) or fixable viability dye negative. Cytofix/Cytoperm kit (BD Biosciences) was used for intracellular cytokine staining according to the manufacturer’s protocol. For intracellular staining of transcription factors, cells were fixed and permabilized using FoxP3 staining kit (eBioscience) according to the manufacturer’s protocol. All analyses were performed on a BD FACSCanto or BD LSRFortessa (BD Biosciences), and data were processed using FlowJo version 10. To identify hematopoietic progenitors, BM was stained with a lineage (Lin) cocktail including monoclonal antibodies (mAbs) against CD3, TER119, and CD19. CD115+ CDPs were identified as PILinCD117midCD135+CD11cMHCII−/lowCD115+CD127; CD115 CDPs were identified as PILinCD117midCD135+CD11cMHCII−/lowCD115CD127; common lymphoid progenitors (CLPs) were identified as PILinCD117midCD135+CD11cMHCII−/lowCD115CD127+; macrophage/dendritic cell progenitors (MDPs) were identified as PILinCD117hiCD135+CD115+; pre-cDC1s were identified as PILinCD117midCD135+CD11c+MHCII−/lowCD115CD127Ly6cSiglecH or Lin (CD3, CD19, CD105, CD127, NK1.1, TER119, and Ly6G) CD135+CD117lowCD226+CD115; and pre-cDC2s were identified as PILinCD117midCD135+CD11c+MHCII−/lowCD115CD127Ly6c+SiglecH or Lin (CD3, CD19, CD105, CD127, NK1.1, TER119, and Ly6G) CD135+CD117lowCD226CD115+. Please refer to table S6 for antibody details.

Cell sorting and magnetic enrichment

Harvested spleens were digested by collagenase (Worthington) and deoxyribonuclease I (DNase I) (Sigma-Aldrich). Erythrocytes were removed with ACK lysis buffer. Light density cells were collected by a density centrifugation procedure. CD11c+ DCs were enriched by CD11c magnetic beads (Miltenyi). The purity of the enriched CD11c+ DC preparation was >90%. cDC1s were sorted as PINK1.1CD19TCRβSiglecHMHCII+CD11c+XCR1+CD172a. cDC2s were sorted as PINK1.1CD19TCRβSiglecHMHCII+CD11c+XCR1CD172a+. The in vitro–generated DCs were sorted as PISiglecHMHCII+CD11c+XCR1+CD172a (cDC1) and PISiglecHMHCII+CD11c+XCR1CD172a+ (cDC2). For OVA-specific OT-I or OT-II cell enrichment, LN was collected and passed through a 70-mm sieve. LN cells were negatively selected with a Lin cocktail including mAbs against MHCII, B220, MAC-1, Ly6G, CD44, and TER119 and positively selected with CD4 (for OT-II enrichment) or CD8 (for OT-I enrichment), using with BioMag Goat Anti-Rat IgG beads (Qiagen). Cells were checked for purity (>99%) by analytical flow cytometry using CD8 or CD4, and TCRvα2 antibody.

Cell isolation from nonlymphoid organs

For epidermal LC isolation, skin pieces (1 cm2) were incubated for 1 hour at 37°C in Dispase (2 U/ml; GIBCO). After this treatment, epidermal sheets were peeled from the dermis, cut into fine pieces, and digested with collagenase (50 U/ml) for 45 min at 37°C. For the isolation of cells from the gut, the fat tissue and Peyer’s patch were removed from the small intestine. The gut was cleaned from luminal contents, and tissues were cut into ~5-mm sections for digestion. Small fragments were then dissociated in 2% FCS Hank’s Ca+Mg+ Free Media with 5 mM EDTA for 40 min at 37°C under gentle agitation. The dissociated epithelial layer was then discarded, and remaining gut fragments were digested in Collagenase IV (1 mg/ml; Worthington Biochemical), DNase I (200 μg/ml; Roche), and Dispase (4 U/ml; Sigma-Aldrich) for 45 min at 37°C under gentle agitation. Lamina propria mononuclear immune cells were isolated by centrifugation on a 40 to 80% Percoll gradient. Lymphocytes were recovered from the interface and washed in cold FACS buffer. Brains were collected in RPMI 1640 and 10% FCS, passed through a 1-cm3 syringe to dissociate grossly the organ, and digested in Hanks’ balanced salt solution containing DNAse and collagenase for 40 min. The dissociated tissue was filtered, centrifuged, and resuspended in a 33% isotonic Percoll solution. After centrifugation, the pellet was resuspended in FACS buffer for further analysis.

In vitro TLR stimulation

FLT3L-cultured DCs (1 × 106) were stimulated with either LPS (10 μg/ml; Sigma-Aldrich) or CpG (200 nM; Invivogen) for 2 hours, treated with GolgiStop (BD Bioscience), and then incubated for another 2 or 4 hours. Cells were washed before being analyzed for cell surface and intracellular antigens. For analysis of CD80, CD86, or CD40 expression, cells were incubated with CpG or LPS for 24 hours before surface staining.

Antigen presentation and cross-presentation assays

For in vitro antigen presentation assay, purified OVA-specific CD8+ OT-I or CD4+ OT-II T cells were labeled with the CTV tracker (Invitrogen) according to the manufacturer’s recommendations. OT-I T cells (5 × 104) were cocultured with purified cDC1s or cDC2s (2.5 × 104) with or without OVA-loaded Bm1 or MhcII−/− splenocytes that were irradiated (1500 rads). Sixty to 72 hours after stimulation, numbers of proliferating T cells were measured by flow cytometry. For in vivo experiments, 1.5 × 106 purified CD45.1 CD8+ OT-I T cells were labeled with CTV and adoptively transferred intravenously into WT or Zfp366−/− full chimeric mouse. The next day, mice were injected intravenously with 2 × 107 OVA-loaded Bm1 splenocytes. On day 3 after immunization, spleen and LN were harvested, and proliferation pattern of OT-I cells were measured by flow cytometry.

Phagocytosis assay

BALB/c splenocytes were pulsed with OVA (10 mg/ml) and then washed and labeled with PKH26 according to the manufacturer’s instruction. The cells were irradiated (1500 rads) before intravenous injection (2 × 107 cells per mouse) into WT or Zfp366−/− chimeric mouse. The spleen was harvested after 16 hours, and detected PKH26+ cells were analyzed by flow cytometry.

Bio-Plex assay

WT or Zfp366−/− cDC1s or cDC2s were cocultured with OT-I or OT-II in the presence of OVA protein, OVA (257 to 264), or OVA (323 to 339) peptide. Supernatants were harvested after 60 to 72 hours. Bio-Plex Pro mouse cytokine chemokine and growth factor assays 23-plex panels (Bio-Rad) were used according to the manufacturer’s instruction. A minimum of 50 beads per cytokine were analyzed on a Luminex FLEXMAP 3D (Merck Millipore). Data acquisition used xPONENT 4.0 (Luminex) acquisition software, with data analyzed in Bio-Plex Manager 6.1.1 (Bio-Rad). Cytokine concentrations were calculated from the standard curve using a five-parameter logistic curve fit.

Western blot

Purified sorted cells were centrifuged and lysed for 30 min on ice with 20 mM tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, and 1 mM sodium pervanadate and protease inhibitors (Roche). Cell debris was removed by centrifugation (15,000 rpm/15 min). Extracts were subjected to SDS–polyacrylamide gel electrophoresis, transferred onto a nitrocellulose membrane, and blocked with 5% skimmed milk, and membranes were probed for anti–DC-SCRIPT and anti-ACTIN (see table S6 for antibody details).

RNA sequencing

DCs were enriched from the spleens of WT or Zfp366−/− full chimeric mice using CD11c magnetic beads according to the manufacturer’s recommendations. Selected cells were sorted as cDC1s (PICD45.1CD45.2+CD19NK1.1TCRβCD11c+MHCII+CD11bCD172aCD8+XCR1+) and cDC2s (PICD45.1CD45.2+CD19NK1.1TCRβCD11c+MHCII+CD11b+CD172a+CD8XCR1). RNA was isolated from three biological replicates with RNeasy Plus Mini kit (Qiagen). mRNA reverse transcription and complementary DNA (cDNA) libraries were prepared using the TruSeq RNA Sample preparation kit (Illumina) following the manufacturer’s instructions. Samples were sequenced with an Illumina NextSeq 500 sequencing system, producing between 14 × 106 and 20 × 106 single-end 85-bp reads per sample.

RNA-seq data analysis

RNA-seq reads were aligned to the mm10 genome using Rsubread 1.35.19 align and using Rsubread’s inbuilt mm10 RefSeq gene annotation to aid the identification of exon junctions (59). Read counts were obtained for Entrez Gene IDs using featureCounts and Rsubread’s inbuild annotation (60). Gene annotation was downloaded from (March 2020). Differential expression analyses were undertaken using the edgeR v3.28.0 and limma v3.42.0 software packages (61, 62). Unexpressed genes were filtered using edgeR’s filterByExpr function with min.count = 20 and = 25. Mitochondrial genes, ribosomal RNA, genes on the X or Y chromosomes, genes of type “other,” and obsolete gene IDs were also filtered. Library sizes were normalized by edgeR’s trimmed mean of M-values (TMM) method.

Differential expression was assessed using the limma-trend method with arrayWeights and a duplicate correlation to account for correlation between mice (63). Gene counts were transformed to log2 counts per million. Duplicate correlation was calculated with the individual mouse as the block using the duplicateCorrelation function from limma. The array weights per sample were estimated with the arrayWeights function from limma with method = “genebygene.” Robust empirical Bayes was used to protect against hypervariable genes (64). Differential expression was assessed using t tests relative to a threshold (TREAT)–moderated t tests (64) relative to a fold change threshold of 1.2. Genes with a TREAT false discovery rate below 0.05 were considered to be differentially expressed.

Gene set enrichment was tested using limma’s fry function, which provides fast approximation to rotation testing (66). Gene set enrichment was visualized using limma’s barcode enrichment plot. Heatmaps of the filtered and normalized logCPM value were plotted with the coolmap function from the limma package. Mean difference plots (MD plots) were plotted with the plotMD function.

Cleavage under targets and tagmentation

The CUT&Tag protocol was described previously (43). The Hyperactive In-Situ ChIP Library Prep Kit was purchased from Vazyme, and libraries were generated following the manufacturer’s recommendations. Briefly, 1 × 105 cDC1s or cDC2s sorted from the spleen of WT or Zfp366−/− full chimeric mouse were bound to concanavalin A–coated magnetic beads (Bags Laboratories) and were subjected for immunoprecipitation with 0.5 μg of primary antibody [anti–DC-SCRIPT or rabbit anti-mouse IgG (immunoglobulin G) control]. After primary incubation and washing using a magnetic stand, a secondary anti-rabbit antibody was added and incubated under gentle agitation for 1 hour (see table S6 for antibody details). Cells were washed and incubated for 1 hour in a mix of hyperactive pG-Tn5/pA-Tn5 transposon with Dig-300 buffer to a final concentration of 0.04 μM. Cells were washed and resuspended in 100 μl of tagmentation buffer (10 mM MgCl2 in Dig-300 buffer) and incubated at 37°C for 1 hour. Reaction was stopped by heat inactivation (55°C for 10 min) and DNA purified using AMPure XP beads (Beckman Coulter). For library amplification, 24 μl of DNA was mixed with 10 μl of 5× TruePrep Amplify Buffer (TAB), 1 μl of tris-acetate-EDTA, and 5 μl of uniquely barcoded i5 and i7 primers (67) and amplified for 14 cycles. PCR products were purified with AMPure XP beads and eluted in water. Libraries were sequenced on an Illumina NextSeq platform, and 150-bp paired-end reads were generated.

CUT&Tag data analysis

Paired-end reads were aligned to the mm10 genome using Rsubread align (59). Duplicate reads were removed, and the resultant bam files were sorted using Sambamba v0.6.6 (68). Reads in the bam files that were filtered on the basis of template length reads were required to have a template length of more than 10 and less than 300 bp. The HOMER pipeline was used to call peaks with the filtered bam files (69). Tag directories were created for each library with the genome specified and with checkGC. The tag directories for biological replicates were summed. The findPeaks function was used to find peaks with style as factor and size as auto in the WT samples with the corresponding Zfp366−/− sample as the input. BedGraph files were created with bamCoverage from Deeptools v2.5.3 (70).

Chromatin immunoprecipitation sequencing analysis

Fastq files for the chromatin immunoprecipitation sequencing (ChIP-seq) profiles of BATF3 and IRF8 in cDC1 cells and BATF3 in cDC2 cells were obtained from gene expression omnibus (GEO) GSE66899 along with an input sample (41). Reads were aligned to the mm10 genome using Rsubread align (59). Duplicate reads were removed, and the resultant bam files were sorted using Sambamba v0.6.6 (68). The HOMER pipeline was used to call peaks (69). Tag directories were created for each library with the genome specified and with checkGC. The findPeaks function was used to find peaks using the input sample with style as factor and size as auto. BedGraph files were created with bamCoverage.

Assay for transposase-accessible chromatin using sequencing analysis

Fastq files for the ATAC-seq (assay for transposase-accessible chromatin using sequencing) profiles of cDC1 and cDC2 were obtained from GEO GSE100738 (44). Read were aligned as for the ChIP-seq analysis. The tag directories for biological replicates were summed. The findPeaks function was used to find peaks with style as histone and size as auto. BedGraph files were created with bamCoverage.

Coverage plots

Bam files from biological replicates were summed, and bigWig files were created with bamCoverage from Deeptools using normalize using Reads Per Kilobase of transcript, per Million mapped reads (RPKM) and a bin size of 1 bp (70). Overlaps between peaks were identified with the overlaps. Any function of the IRanges package v2.20.2 with minimum overlap required to be 10 bp (71). Coverage was calculated over genomic regions using computeMatrix from Deeptools using reference-point mode, referencePoint center, 1 kbp upstream and downstream, and a bin size of 50 bp. The output was plotted in R with package pheatmap with no row or column clustering and scale = row.

Retroviral transduction

The DC-SCRIPT and IRF8 expression constructs and the retroviral transduction methodology were as described in (22, 72).

Enzyme-linked immunosorbent assay

Enzyme-linked immunosorbent assay for mouse IL-12p40 and IFN-γ (BioLegend) was performed according to the manufacturer’s instruction.

Toxoplasma growth and mice infections

Human foreskin fibroblasts were grown to confluence in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 1% (v/v) GlutaMAX (Invitrogen) and 10% (v/v) Cosmic Calf serum (HyClone, GE Healthcare) (D10). D10 media was replaced with D1 [DMEM supplemented with 1% (v/v) GlutaMAX and 1% (v/v) FCS] before parasite inoculation. T. gondii strain Prugniaud (Pru) expressing tdTomato under the constitutive tubulin promoter was used in this study (73). Cells were incubated at 37°C with 10% CO2 until the monolayer had lysed.

For mice infections, host cells were scraped and passed through a 27-G needle to release parasites. Parasites were resuspended in PBS at a concentration of 10,000 tachyzoites per 200 μl and injected intraperitoneally. Mice were weighed, and stage of infection was monitored daily. For some experiments, blood was collected on day 4 to measure cytokines.

Quantitative PCR

RNAs were isolated using the RNeasy Plus Mini Kit, and cDNAs were synthesized from total RNAs using the iScript reverse transcription supermix for RT-qPCR (Bio-Rad) according to the manufacturer’s instructions. Amplification was performed with SYBR green master mix (BioLabs) on a Bio-Rad CFX 384. Primers are listed on table S7.

Quantification and statistical analysis

The statistical significance of non–RNA-seq data was assessed with Prism 8 software (GraphPad). z score = (x − mean)/SD. Student’s t tests for all single comparisons were two-tailed and unpaired and assumed Gaussian distribution and equivalent SD unless indicated (table S8). One-way, two-way, and three-way analysis of variance (ANOVA) was performed with Tukey’s test for multiple comparisons. Bar graphs display the means ± SEM. P values of <0.05 were considered significant.


Fig. S1. Expression of DC-SCRIPT in the spleen and LN.

Fig. S2. Expression of DC-SCRIPT in BM progenitors.

Fig. S3. Increased pDC development potential in the absence of DC-SCRIPT.

Fig. S4. The influence of DC-SCRIPT deficiency on the development of DC-SCRIPT–expressing cell types.

Fig. S5. Normal lymphocyte development in the absence of DC-SCRIPT.

Fig. S6. Increased expression of pDC signature genes in DC-SCRIPT–deficient cDC1.

Fig. S7. Genome-wide analysis of DC-SCRIPT DNA binding.

Fig. S8. Genome-wide analysis of DC-SCRIPT DNA binding shows correlation with BATF3 and IRF8 binding.

Fig. S9. Normal antigen presentation in DC-SCRIPT–deficient cDC1s and cDC2s in response to peptide antigen.

Fig. S10. DC-SCRIPT specifically regulates IL-12p40 production in cDC1s.

Fig. S11. DC-SCRIPT–driven IL-12p40 production in cDC1s is required for an optimal immune response.

Table S1. Differentially expressed genes in cDC1s (.xls file).

Table S2. Differentially expressed genes in cDC2s (.xls file).

Table S3. Top 100 differentially expressed genes between WT cDC1 and cDC2 (.xls file).

Table S4. DC-SCRIPT peaks called in cDC1 and cDC2 (.xls file).

Table S5. Transcription factor–cobound regions in cDC1 and cDC2 (.xls file).

Table S6. Antibodies used in this study.

Table S7. Primer sequences.

Table S8. Raw data table (.xls file).


Acknowledgments: We thank our animal technicians and the institute flow cytometry facility for excellent technical assistance. Funding: This work was supported by the National Health and Medical Research Council of Australia [1155342 and 1054925 (S.L.N.); 1143976, 1150425, and 1080321 (A.M.L.); 1058892 (G.K.S.); and 1154502, 1113293, and 1163090 (J.A.V.)] and the Melbourne research scholarship (S.Z.). M.C. is supported by the Jenny Thatchell/Pauline Speedy and the Barbara McDonald Innovation grants, H.D.C. by the Marian and E.H. Flack Fellowship, and H.D.C. and G.K.S. by the Chan Zuckerberg Initiative EOSS program. This work was made possible through the Victorian State Government Operational Infrastructure Support and Australian Government NHMRC IRIIS. The generation of the Zfp366−/− and Zfp366-tdTomato mice used in this study was supported by the Australian Phenomics Network (APN) and the Australian Government through the National Collaborative Research Infrastructure Strategy (NCRIS) program. Author contributions: Conceptualization: M.C. and S.L.N. Methodology: M.C., S.Z., D.V.B., A.J.K., J.A.V., and G.K.S. Investigation: S.Z., Y.Z., W.C., M.A., S.S., C.J.T., A.M.L., A.D., N.J., M.C., and H.D.C. Writing (original draft): M.C. and S.Z. Writing (review and editing): M.C. and S.L.N. Resources: H.D.C., D.V.B., C.J.T., and J.A.V. Supervision: M.C. and S.L.N. Competing interests: The authors declare that they have no competing interests. Data and materials availability: RNA-seq and CUT&Tag raw data and read counts are available from the GEO repository as series GSE165361. Zfp366−/− and Zfp366tdTomato mice are available under a material transfer agreement from the corresponding authors. All data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials.

Stay Connected to Science Immunology

Navigate This Article