ETV3 and ETV6 enable monocyte differentiation into dendritic cells by repressing macrophage fate commitment

In inflamed tissues, monocytes differentiate into macrophages (mo-Macs) or dendritic cells (mo-DCs). In chronic nonresolving inflammation, mo-DCs are major drivers of pathogenic events. Manipulating monocyte differentiation would therefore be an attractive therapeutic strategy. However, how the balance of mo-DC versus mo-Mac fate commitment is regulated is not clear. In the present study, we show that the transcriptional repressors ETV3 and ETV6 control human monocyte differentiation into mo-DCs. ETV3 and ETV6 inhibit interferon (IFN)-stimulated genes; however, their action on monocyte differentiation is independent of IFN signaling. Instead, we find that ETV3 and ETV6 directly repress mo-Mac development by controlling MAFB expression. Mice deficient for Etv6 in monocytes have spontaneous expression of IFN-stimulated genes, confirming that Etv6 regulates IFN responses in vivo. Furthermore, these mice have impaired mo-DC differentiation during inflammation and reduced pathology in an experimental autoimmune encephalomyelitis model. These findings provide information about the molecular control of monocyte fate decision and identify ETV6 as a therapeutic target to redirect monocyte differentiation in inflammatory disorders.

Monocytes and monocyte-derived cells are central players in the initiation and resolution of inflammatory responses. In chronic inflammatory diseases, monocyte-derived antigen-presenting cells (APCs) become major drivers of the physiopathology by stimulating pathogenic T cells. Blocking monocyte differentiation therefore represents an attractive therapeutic strategy. A major hurdle is the paucity of molecular targets, due to limited knowledge of the molecular regulation of monocyte fate commitment.
Circulating monocytes infiltrate mucosal or inflamed tissues where they differentiate into mo-Macs or mo-DCs 1-3 . Mo-Macs are generally involved in homeostasis and tissue repair, whereas mo-DCs present antigens to T cells directly in tissues. However, in chronic nonresolving inflammation, T cell stimulation by mo-DCs becomes deleterious. In Crohn's disease, rheumatoid arthritis and psoriasis, mo-DCs secrete high amounts of interleukin (IL)-23 and stimulate helper type 17 T cells (T H 17) cells, two major drivers of the physiopathology [4][5][6][7] . In mouse models, mo-DCs induce pathogenic T cells that mediate tissue damage in experimental autoimmune encephalomyelitis (EAE) 8 and colitis 9,10 . Blocking monocyte differentiation has therefore emerged as a potential therapeutic strategy for inflammatory disorders. Pharmacological inhibition of monocyte recruitment suppresses the development of colitis 11 and the severity of EAE 12 . Inducing monocyte apoptosis with nanoparticles reduces inflammation and disease symptoms in colitis, EAE, peritonitis and virus-induced Article https://doi.org/10.1038/s41590-022-01374-0 To validate the differential expression of ETV3 and ETV6 in mo-DCs and mo-Macs generated in vitro, we measured their expression in sorted mo-DCs and mo-Macs after differentiation. Both transcription factors were more expressed in mo-DCs compared with mo-Macs at the messenger RNA (Fig. 1a) and protein levels (Fig. 1b).

ETV3 and ETV6 are essential for human mo-DC differentiation
To address the role of ETV3 or ETV6 in monocyte fate commitment, we silenced their expression using a lentivirus expressing a short hairpin (sh)RNA against ETV3, ETV6 or a scramble sequence. We assessed the effect of silencing on monocyte differentiation after 5 d by staining for phenotypic markers of mo-DCs (CD1a) and mo-Macs (CD16). We used three different shRNAs for each molecule and their efficiency was evaluated by measuring protein expression by immunoblotting ( Fig. 1c-f). These shRNAs all significantly decreased ETV3 or ETV6 expression with an efficiency of 40-90% (Fig. 1c,e). Silencing of ETV3 or ETV6 decreased mo-DC and increased mo-Mac differentiation (Fig. 1d,f, respectively). These results show that ETV3 and ETV6 play a key role in mo-DC differentiation. To characterize their expression kinetics during monocyte differentiation, we measured their expression by reverse transcription quantitative PCR (RT-qPCR) at different time points. ETV3 and ETV6 mRNA increased during the first hours in culture with a peak at 3 or 6 h for ETV3 and ETV6, respectively (Fig. 1g). To determine which signals increase their expression, we measured ETV3 and ETV6 mRNA in monocytes on exposure to M-CSF, in the presence or absence of IL-4 and TNF (Fig. 1h). ETV3 expression was induced by TNF and ETV6 expression by IL-4, with TNF sustaining its expression at later time points. These results show that ETV3 and ETV6 are expressed on exposure to inflammatory signals at an early stage of monocyte differentiation, suggesting that they could play a role in their lineage commitment toward mo-DCs.

ETV3 and ETV6 repress ISGs
To decipher the transcriptional network of ETV3 and ETV6, we first investigated the kinetics of their nuclear localization using imaging flow cytometry. To increase the resolution of our analysis, we sought to favor mo-DC differentiation in the culture system by using a modified cytokine cocktail (increased TNF concentration) (Extended Data Fig. 2a). We performed intracellular staining of ETV3 or ETV6 after 0, 1, 2, 3 or 6 d of culture. To quantify the expression of ETV3 and ETV6, we gated on ETV3-or ETV6-positive cells (Extended Data Fig. 2b,c). The percentage of ETV3 + and ETV6 + cells increased gradually, reaching a plateau at day 3 (Extended Data Fig. 2d). To quantify the nuclear localization of ETV3 or ETV6, we used the ImageStream software to calculate the 'similarity' of the ETV3 or ETV6 channel with the nuclear DAPI staining. High similarity (>1.8) indicates a nuclear localization of the transcription factor, whereas low similarity (<1.8) indicates a cytosolic localization (Extended Data Fig. 2e). We observed that ETV3 and ETV6 are located in the nucleus until day 3 in around 90% of the cells (Extended Data Fig. 2f,g). By contrast, at day 6, ETV3 and ETV6 are located in the cytosol in around 50% of the cells. As the transcriptional activity of ETV3 and ETV6 requires their nuclear localization, this observation suggests that ETV3 and ETV6 exert their function mainly during the first days of differentiation.
To identify the target genes of ETV3 and ETV6, we performed transcriptomic analysis by bulk RNA-seq on monocytes silenced or not silenced for ETV3 or ETV6, at day 3 of differentiation, with the modified cytokine cocktail to favor mo-DC development. Then, we performed a differential gene expression analysis using DESeq2 comparing control with silenced samples for ETV3 (Fig. 2a) or ETV6 (Fig. 2b) separately. We defined the differentially expressed genes (DEGs) by a log 2 (fold-change) (log 2 (FC)) > 0.5 and a P adj < 0.05. Comparison of the DEGs for each shRNA revealed unique transcriptional networks, because most of the genes are specific for ETV3 or ETV6 silencing ( Fig. 2c and Extended Data Fig. 3a; lists of DEGs can be found in Supplementary Tables 1 and 2). encephalitis 13 . Finally, impairing monocyte survival and differentiation via macrophage-colony-stimulating factor (M-CSF) receptor blockade reduces inflammation in arthritis 14,15 . However, a major caveat of these approaches is the potentially adverse effects due to disrupted differentiation of mo-Macs, which are involved in the resolution of inflammation. Blockade of the M-CSF receptor was reported to impair cardiac repair 16 and skeletal muscle regeneration 17 . Manipulating monocyte fate commitment toward mo-DCs versus mo-Macs would therefore provide an attractive alternative strategy. This would require a better understanding of the molecular regulators orchestrating the monocyte fate decision.
Monocyte fate is not transcriptionally imprinted 18,19 . Instead, monocytes respond to microenvironmental cues that can redirect their fate. Using in vitro models of human monocyte differentiation, we and others have shown that IL-4 signaling is essential to induce mo-DC differentiation 18,20 . Transcription factors involved in this process include IFN regulatory factor 4 (IRF4), aryl hydrocarbon receptor, B lymphocyte-induced maturation protein-1 (BLIMP-1) and the nuclear receptor corepressor 2 (NCOR2) 18,20 . What controls the balance of monocyte differentiation into mo-Macs versus mo-DCs remains unclear.
In the present study, we have identified ETV3 and ETV6 as important regulators of mo-DC differentiation and repressors of monocyte differentiation into macrophages. We provide evidence that ETV3 and ETV6 repress macrophage fate commitment independently of their action on IFN-stimulated gene (ISG) expression. Finally, we validate the role of ETV6 in monocytes for in vivo ISG repression, by analyzing single-cell transcriptomic data from ETV6-mutated patients and in mice deficient for Etv6 in monocytes. We further show that Etv6 in monocytes is required for mo-DC differentiation in a model of inflammatory peritonitis and modulates the severity of symptoms in a model of neuroinflammation. By enabling a better understanding of the molecular ontogeny of monocyte-derived cells, our results should provide opportunities for the therapeutic manipulation of monocyte differentiation.

ETV3 and ETV6 expression is greater in mo-DCs than in mo-Macs
We hypothesized that transcription factors differentially expressed between human mo-DCs and mo-Macs could be involved in their differentiation from monocytes. Our transcriptomic analysis of monocyte-derived cells from clinical samples identified ETV3 and ETV6 as potential candidates 18 . To assess ETV3 and ETV6 expression, we used our transcriptomics data from cells naturally occurring in vivo in peritoneal ascites 18 . ETV3 and ETV6 were more expressed in in vivo mo-DCs compared with mo-Macs (Fig. 1a). To address their potential role in monocyte differentiation, we used our previously published in vitro model allowing the simultaneous differentiation of mo-Macs and mo-DCs 18 . In this model, human CD14 + monocytes cultured for 5-6 d with M-CSF, IL-4 and tumor necrosis factor (TNF) differentiate into mo-Macs (CD16 + ) or mo-DCs (CD1a + ) or remain undifferentiated (double-negative cells). To verify monocyte purity, and in particular the absence of contaminating DC progenitors, we performed single-cell RNA-sequencing (scRNA-seq) on the initial population purified from two different donors (Extended Data Fig. 1a-d). We found two main populations of monocytes displaying high expression of S100A8 (clusters 0 and 1) or major histocompatibility complex class II (MHC-II) genes (cluster 2; Extended Data Fig. 1a), consistent with the 'neutrophil-like' and 'DC-like' monocyte populations previously reported 21 . In addition, we identified a small population of FCGR3A + monocytes (cluster 3, corresponding to CD14 + CD16 + intermediate monocytes), and a negligable proportion (2% each) of contaminating natural killer (NK) cells (cluster 4) and monocytes with high ISG expression (cluster 5) (Extended Data Fig. 1a,b). These results indicate that our culture model does not contain progenitor cells other than monocytes.
Article https://doi.org/10.1038/s41590-022-01374-0 To identify the molecular pathways controlled by ETV3 or ETV6, we performed network analysis. We calculated transcription factor activity using DoRoThEa regulons and VIPER 22 (Fig. 2d). STAT1 and STAT2 were the most active transcription factors in silenced samples. We then calculated the enrichment of gene ontology (GO) terms (biological process) for upregulated genes (Fig. 2e). Type I IFN response gene sets were enriched in silenced samples. This is consistent with the predicted activity of STAT1 and STAT2, which are known to control the expression of ISGs 23 . To confirm this, we filtered the DEG matrix for known ISGs. Most of the ISGs were expressed more in silenced compared with control samples (Fig. 2f,g). To determine the in vivo relevance of this finding, we reanalyzed scRNA-seq data from peripheral blood mononuclear cells (PBMCs) of patients carrying a germline mutation of ETV6 (P214L) resulting in loss of function 24 . We first filtered the data to retain only CD14 + and CD16 + monocytes from healthy and ETV6 P214L patients (Extended Data Fig. 3b,c). We then interrogated the single-cell data of ETV6 P214L and wild-type (WT) monocytes with different gene sets. Genes upregulated on ETV6 silencing in our in vitro system   . d,f, Mo-mac and mo-DC differentiation from monocytes after 5 d of ETV3 (d) or ETV6 (f) silencing . One representative donor is shown (n = 8) and the median (n = 8 in three independent experiments; paired one-way analysis of variance (ANOVA)). DN, double negative. g, ETV3 or ETV6 mRNA expression was analyzed by RT-qPCR. Each symbol represents an individual donor (n = 6 in three independent experiments). h, Monocytes were cultured for 3 and 6 h with medium only or combinations of M-CSF, IL-4 and TNF. Each symbol represents an individual donor (n = 5 in two independent experiments; paired one-way ANOVA). For all panels: * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001. All statistical tests were two sided. For immunoblots, paired samples were derived from the same experiment and processed in parallel.
Article https://doi.org/10.1038/s41590-022-01374-0 were enriched in ETV6 P214L monocytes compared with WT monocytes ( Fig. 2h and Extended Data Fig. 3d). Moreover, ETV6 P214L monocytes had a higher enrichment for ISGs than WT monocytes (Fig. 2i,j and Extended Data Fig. 3e), consistent with a previous report 24 . These results show that ETV3 and ETV6 repress ISG expression in monocytes in vitro and in vivo in humans and suggest that STAT1 signaling may be involved in the differentiation of monocytes.

The type I IFN pathway promotes mo-Mac differentiation
Given the impact of ETV3 or ETV6 silencing on mo-DC differentiation, our findings suggest that the type I IFN pathway may be activated in our model despite the absence of exogenous IFN in the culture system. However, we could not detect type I IFN secretion in the culture supernatant (not shown). To directly assess the effect of STAT1 activation on monocyte differentiation, we cultured monocytes in the presence of IFN-α or IFN-β. Type I IFN increased mo-Mac and decreased mo-DC differentiation in a dose-dependent manner ( Fig. 3a and Extended Data Fig. 4a). Neither IFN-α nor IFN-β affected monocyte-derived cell viability (Extended Data Fig. 4b). In addition, type I IFN increased the expression of CD163, an early mo-Mac marker, and decreased the expression of CD1b, an early mo-DC marker, on the double-negative cells ( Fig. 3b and Extended Data Fig. 4c). Collectively, these results show that activation of the type I IFN pathway promotes mo-Mac differentiation at the expense of mo-DCs. To understand how STAT1 signaling could affect monocyte fate decision, we exposed monocytes to IFN-α, in the presence or absence of cytokines, and measured the early expression of MAFB and IL-4-induced genes PPARG and ZNF366 (Fig. 3c). MAFB was rapidly increased on IFN-α exposure, along with the ISG MX1, whereas PPARG and ZNF366 expression on IL-4 treatment was inhibited by IFN-α. These results suggest that STAT1 activation modulates the mo-Mac:mo-DC balance by repressing the IL-4-induced mo-DC differentiation program.

Monocyte differentiation is controlled independently of ISGs
To directly test whether ISG expression plays a role in the control of monocyte differentiation by ETV3 or ETV6, we sought to inhibit type I IFN signaling in our culture model using recombinant viral B18R, a soluble receptor of type I IFN that prevents signaling through the IFN-α/β receptor (IFNAR) 25 . Exposure to B18R did not impact the proportions of mo-DCs and mo-Macs obtained with or without silencing of ETV3 or ETV6 ( Fig. 3d-f), even though B18R efficiently inhibited ISG expression, including MX1, CXCL10, OAS2 and IFIT3 (Fig. 3g,h). These results indicate that inhibition of the type I IFN pathway does not rescue mo-DC differentiation in the absence of ETV3 or ETV6 expression. We conclude that ETV3 and ETV6 regulate monocyte differentiation independently of their action on ISGs.

ETV3 and ETV6 repress mo-Mac program and differentiation
To identify the genes directly targeted by ETV3 and ETV6 during monocyte differentiation, we performed chromatin immunoprecipitation followed by sequencing (ChIP-seq) at day 3 of differentiation with the modified cytokine cocktail favoring mo-DC differentiation (a complete list of peaks can be found in Supplementary Table 3). We first analyzed the enrichment in known motifs using HOMER (Fig. 4a,b).
We found that the ETS motif was the most enriched in both ETV3 and ETV6 ChIP-seq datasets, with the ETS-IRF composite motif and IRF8 motif also significantly enriched, consistent with a previous report of interaction between ETV6 and IRF8 in macrophages 26 . Most identified genes were common between ETV3 and ETV6 (Fig. 4c), including at promoter regions. To assess the targets of ETV3 and ETV6 among DEGs, we intersected a list of genes from the RNA-seq and ChIP-seq datasets (Fig. 4d). We found that around 50% of regulated genes were directly bound by ETV3 or ETV6. Consistent with the transcriptomics analysis, ISGs were found among direct targets of ETV3 and ETV6 (Extended Data Fig. 4d-f).
To evaluate how ETV3 and ETV6 modulate monocyte fate commitment, we performed gene set enrichment analysis (GSEA) on DEGs and assessed the enrichment of monocyte, mo-Mac or mo-DC gene signatures. The mo-Mac signature was enriched in silenced samples, whereas mo-DC genes were enriched in the control condition (Fig. 4e). Among DEGs belonging to these signatures, we found a vast majority of direct targets identified in ChIP-seq (Fig. 4f). In particular, we found that MAFB expression, which is essential for mo-Mac differentiation 18 , was directly modulated by ETV3 and ETV6 (Fig. 4f,g). These results suggest that ETV3 and ETV6 repress the mo-Mac transcriptional program. To confirm this, we sought to overexpress ETV6 during monocyte differentiation. To avoid spontaneous expression of ETV6, we cultured monocytes with M-CSF only, a condition in which monocytes differentiate exclusively into mo-Macs. We validated the forced expression of ETV6 by immunoblotting (Fig. 4h). ETV6 overexpression decreased mo-Mac differentiation ( Fig. 4i,j). Taken together, these results indicate that ETV3 and ETV6 directly repress mo-Mac differentiation.

Etv6 represses ISG expression in mice
To validate the physiological relevance of our findings, we employed a mouse model that deletes Etv6 in Cx3cr1-expressing cells after induction with tamoxifen (Fig. 5a). Cx3cr1 is a canonical marker of patrolling monocytes, thus the deletion of Etv6 in Cx3cr1 + cells would be expected to delete Etv6 in monocytes and their progeny 27 . To confirm the cell types targeted by the deletion, we measured a yellow fluorescent protein (YFP) reporter mimicking the endogenous Cx3cr1 expression pattern. As expected, YFP was expressed at the highest level in Ly6c low patrolling monocytes, as well as in certain DC subsets including Esam − splenic cDC2 cells (Extended Data Fig. 5a). In addition, we measured Etv6 expression by RT-qPCR in cell-sorted populations (Extended Data Fig. 5b). Etv6 expression was significantly decreased in bone marrow (BM) and spleen monocytes of Cx3cr1-Etv6 Δ mice, as well as in spleen cDC1 and cDC2 cells but not pDCs. We have previously identified a population of peritoneal mo-DCs 18 . Etv6 was also significantly decreased in peritoneal mo-DCs of Cx3cr1-Etv6 Δ mice but not in peritoneal mo-Macs or resident macrophages (Extended Data Fig. 5b). To assess the impact of Etv6 deletion in Cx3cr1-expressing cells on ISG expression in vivo, we measured by flow cytometry the expression of Sca-1, an IFN-inducible protein 28,29 . We analyzed immune cells from WT and Cx3cr1-Etv6 Δ BM, blood and spleen (gating strategies in Extended Data Fig. 6a-c). Sca-1 expression was higher in Cx3cr1-Etv6 Δ than in WT BM monocytes (Fig. 5b). Sca-1 was also more expressed in Cx3cr1-Etv6 Δ mice in B cells, T cells and neutrophils in BM, blood and spleen, and in spleen cDC1 and cDC2 cells and pDCs  (b). c, Overlap of DEGs on ETV3 or ETV6 silencing. Up-or downregulation was calculated compared with the control condition. d, Inference of transcription factor activity with DoRoThEA. Activity (z-score) in silenced samples compared with the control is shown for top regulons (ETV3 in orange, ETV6 in purple). A, highest confidence; C, lowest confidence. e, Enrichment of the top GO terms (biological process) associated with genes upregulated in silenced compared with control samples. Fold enrichment is indicated by the size of the circle. The number of genes observed is indicated for each pathway. FDR, false discovery rate. f,g, Volcano plot of ISGs among DEGs for ETV3 (f) or ETV6 (g) silencing, obtained with DESeq2 (Wald's test). Colored dots indicate P adj < 0.05. h-j, ScRNA-seq data from blood monocytes of healthy (gray) and mutant ETV6 P214L (purple) patients downloaded from a public source. h, Enrichment score for the list of upregulated genes on ETV6 silencing in human monocyte culture. The median is shown. i,j, Enrichment score for the ISG signature in individuel cells (i) andmedian ( j) (Student's t-test: **** P < 0.0001). All statistical tests were two sided.   These results indicate that the increased ISG expression in Cx3cr1-Etv6 Δ mice is due to Etv6 deletion in monocytes rather than in DCs. To confirm our observations, we analyzed the expression of additional ISGs by RT-qPCR in BM monocytes and in peritoneal mo-DCs, mo-Macs and resident macrophages. Isg15, Mx1, Cxcl10 and Ly6a (encoding Sca-1) were expressed more in Etv6 Δ than in WT monocytes ( Fig. 5f) and in peritoneal Etv6 Δ mo-DCs compared with WT cells (Fig. 5g). Isg15 and Mx1 were also expressed more in Etv6 Δ peritoneal macrophages (Fig. 5g). This widespread spontaneous ISG expression suggests that Etv6 deletion induces type I IFN secretion by Cx3cr1 + cells. We were unable to detect circulating IFN-β (Extended Data Fig. 5d); however, we found that Ifnb1 spontaneously expressed in BM Etv6 Δ monocytes (Fig. 5h), but not in spleen DCs (Extended Data Fig. 5e) No IFN-α IFN-α 100 U ml -1 IFN-α 1,000 U ml -1 shControl shETV6

Etv6 controls mo-DC differentiation in mice
To determine whether Etv6 modulates monocyte differentiation in vivo, we first analyzed monocyte populations in steady-state blood, bone marrow (BM) and spleen of Cx3cr1-Etv6 Δ mice. The number of monocyte progenitors (cMoPs) in the BM was unchanged (Extended Data Fig. 5f). The differentiation of monocytes from the cMoPs of Cx3cr1-Etv6 Δ or WT mice was also similar in an in vitro assay (Extended Data Fig. 5g), excluding a role for Etv6 in the differentiation of cMoPs into monocytes. The numbers of B cells, T cells, neutrophils or Ly6C high monocytes were not affected by Etv6 deletion (Extended Data  Fig. 5h). The numbers of CD11b + CD115 + Ly6C int and CD11b + CD115 + Ly6C neg monocytes decreased in Cx3cr1-Etv6 Δ mice compared with WT mice. Moreover, the spleens of Cx3cr1-Etv6 Δ mice harbored decreased numbers of cDC2 cells, particularly of the Esam − subset that is transcriptionally and functionally related to monocytes (Extended Data Fig. 5h) 30 . To address the role of Etv6 in monocyte differentiation in vivo, we analyzed the peritoneal compartment in the steady state and during inflammation (Fig. 6a). In Cx3cr1-Etv6 Δ mice, mo-DCs and mo-Macs in the steady-state peritoneum were unaffected (Fig. 6b). By contrast, during thioglycolate-induced peritonitis, numbers of mo-DCs increased only in WT mice, whereas mo-Macs increased in Cx3cr1-Etv6 Δ mice (Fig. 6c). Monocyte recruitment to the inflamed peritoneum did not differ between WT and Cx3cr1-Etv6 Δ mice (Fig. 6c), suggesting that the mo-DC:mo-Mac balance was skewed in Cx3cr1-Etv6 Δ mice. To confirm that this phenomenon was a monocyte-intrinsic effect, we performed adoptive transfer of CD45.2 + WT or Etv6 Δ monocytes into the inflamed peritoneum of CD45.1 + recipient mice (Fig. 6d). Transferred monocytes differentiated in situ into mo-DCs and mo-Macs (Fig. 6e); however, the mo-DC output was significantly decreased in the progeny of Etv6 Δ monocytes compared with WT ones (Fig. 6f). These results show that Etv6 Δ monocytes are impaired in their differentiation into mo-DCs during inflammation in mice, as observed in human monocytes (Fig. 1g-i).
Finally, we sought to apply our findings to a physiopathological setting. Mo-DCs have a deleterious role in EAE 8 , an animal model for multiple sclerosis (MS). In addition, IFN-β treatment improves disease symptoms and was reported to act primarily on myeloid cells 31 . Therefore, we hypothesized that Etv6 deletion in monocytes would ameliorate EAE outcome. We induced EAE in WT and Cx3cr1-Etv6 Δ mice by injection of myelin oligodendrocyte glycoprotein (MOG; Fig. 6g). Cx3cr1-Etv6 Δ mice showed less severe symptoms during the course of EAE (Fig. 6h) and reduced incidence (Fig. 6i). Of note, Cx3cr1-Etv6 Δ mice also target microglia. Ets1, which encodes an Etv6 antagonist 32 , is highly expressed in microglia (Extended Data Fig. 7a), suggesting that Etv6 action in microglia is naturally inhibited in WT mice. To directly address the potential role of microglial Etv6 during EAE, we induced Etv6 deletion in Cx3cr1-Etv6 Δ mice by tamoxifen, then waited for 4 weeks to induce EAE (Fig. 6j). In this setting, only long-lived macrophages, including microglia, will remain deficient for Etv6, but short-lived monocytes will regain normal Etv6 expression 33 . We found no difference in the development of EAE symptoms between WT and Cx3cr1-Etv6 Δ mice (Fig. 6k) and even increased the peak score in deficient mice (Fig. 6i). In addition, to exclude a role for Etv6 in cDCs in the ameliorated symptoms, we induced EAE in Cd11c-Etv6 Δ mice. We found no significant difference in EAE severity between Cd11c-Etv6 Δ mice and WT littermates (Extended Data Fig. 7b). These results confirm that Etv6 deletion in monocytes, but not in microglia or cDCs, confers protection against severe EAE symptoms.
To understand the cellular mechanisms involved in ameliorated EAE outcome, we first analyzed DC populations in the lymph nodes draining the site of MOG injection during induction phase (7 d postimmunization) (Extended Data Fig. 7c). We found that mo-DCs were significantly decreased in the lymph nodes of Cx3cr1-Etv6 Δ mice (Fig. 6m,n), but not monocytes (Fig. 6n), neutrophils or other DC subsets (Extended Data Fig. 7d). It was recently shown that mo-DCs, but not cDC2 cells, are involved in the presentation of MOG antigen to CD4 + T cells in the lymph nodes 34 . MHC-II molecule expression by mo-DCs was unaffected by Etv6 deletion (Extended Data Fig. 7e). We hypothesized that decreased mo-DC numbers in Cx3cr1-Etv6 Δ mice would reduce the induction of pathogenic CD4 + T cells. To test this, we assessed the presence in lymph nodes of MOG-specific CD4 + T cells using tetramer staining (Fig. 6o). We found that MOG-specific CD4 + T cells were significantly decreased in Cx3cr1-Etv6 Δ mice compared with WT mice, which can explain the reduced EAE symptoms in the central nervous system.
Collectively, these results confirm that Etv6 controls monocyte differentiation in vivo in mice during inflammation. We also identified Etv6 in monocytes as a therapeutic target for chronic inflammatory disorders such as MS.

Discussion
In this work, we identified ETV3 and ETV6 as molecular regulators of the early stages of monocyte differentiation. We found that ETV3 and ETV6 act as repressors of mo-Mac fate commitment. We validated these observations in vivo, showing that mice deficient for Etv6 in monocytes display impaired mo-DC differentiation during inflammation. In addition, we found that Etv6 deletion in monocytes reduces the severity of EAE symptoms. Our findings allow a better understanding of the molecular control of monocyte fate decision and identify ETV6 as a therapeutic target in inflammatory disorders.
ETV3 and ETV6 are members of the Ets family of transcription factors. ETV6 is essential for hematopoietic stem cell survival 35 and ETV3 was shown to regulate cell-cycle arrest 36 . However, their role in immune cells remains poorly understood. We have previously shown that ETV6 is expressed in DCs and facilitates the functional differentiation of cDC1 cells 32 . ETV3 was proposed to be a potential anti-inflammatory mediator downstream of IL-10 (ref. 37 ). In the present study, we identify ETV3 and ETV6 as key transcriptional regulators of mo-DC differentiation. Additional transcriptional repressors are probably involved in this process, because ETV3 or ETV6 transcriptional activity requires their association with corepressors. In particular, ETV6 has been shown to associate with IRF8 in a murine macrophage-like cell line 26 and in mouse CD4 T cells 38 . Although IRF8 is essential for monocyte development from their progenitors 39,40 , whether it participates in mo-DC or mo-Mac differentiation is unknown. ETV6 has also been reported to associate in human PBMCs with NCOR2 (ref. 24 ), which regulates some of the IL-4-induced genes during human mo-DC differentiation 20 . In a human monocyte-like cell line, ETV3 was shown to associate with the repressor DP103, which interacts with the histone deacetylases HDAC2 and HDAC5 (ref. 36 ). Moreover, ETV6 recruits HDAC3 to the repressor complex in murine cell lines and human PBMCs 24,26,41 . Although a specific role for histone deacetylation in mo-DC fate commitment has not been described, it would be consistent with the fact that remodeling of histone acetylation occurs during monocyte differentiation 42 . Further work is needed to unravel the exact mechanism and molecular partners for the repression of ETV3 and ETV6 target genes in monocytes.
We have shown that ETV3 and ETV6 repress ISGs during monocyte differentiation and that ETV6 deletion in monocytes induces spontaneous ISG expression in vivo in mice. This is consistent with previous reports showing that ETV6 is involved in ISG repression in human PBMCs 24 and binds to an IFN-stimulated response element in a reporter assay 26 . We also found that genes targeted by ETV3 versus ETV6 were only partially overlapping. This is in line with the observation that ETV7, another member of the Ets transcription repressor family, represses a subset of ISGs, but not all ISGs, in virus-exposed cells 43 . These observations suggest the existence of a specific pattern of target ISGs for each member of the ETV family.
We find that activation of the type I IFN pathway promotes mo-Mac differentiation in our culture system, where human monocytes are exposed to M-CSF, IL-4 and TNF. This is consistent with the finding that monocytes differentiated with granulocyte-macrophage (GM)-CSF and IL-4 in the presence of IFN-β display altered phenotype and functional features, suggesting impaired mo-DC differentiation, although the reorientation of their fate was not investigated 44 . In addition, IFN-γ and IL-4 have been shown to mutually inhibit each other's programs in macrophages, via the crossrepression of STAT1 and STAT6 (ref. 45 ). In line with this idea, our results indicate that, when monocytes are exposed simultaneously to type I IFN and IL-4, the STAT1-induced program dominates that of STAT6. This suggests that STAT1 signaling For all panels: * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001. All statistical tests were two sided. Article https://doi.org/10.1038/s41590-022-01374-0 would need to be repressed in monocytes to enable expression of the IL-4-dependent mo-DC differentiation program.
Monocyte-derived cells have been shown to play a central role in neuroinflammation. Mice deficient for CCR2 or its ligand, in which monocytes cannot exit the BM, are resistant to EAE or develop milder disease depending on strains [46][47][48][49] . In addition, blocking monocyte recruitment using a pharmacological inhibitor diminishes the incidence and severity of EAE 12 . Monocyte depletion after EAE onset also reduces inflammation and disease symptoms 13,46,50 . Mo-DCs and mo-Macs appear to play different roles during EAE. Mo-DCs are responsible for the presentation of myelin antigens during the induction phase of EAE 34      Article https://doi.org/10.1038/s41590-022-01374-0 cells by secreting IL-23 (ref. 8 ). By contrast, mo-Macs display specific anti-inflammatory features during the resolution phase of EAE [51][52][53] . In patients with MS, monocyte recruitment is particularly increased in demyelinated areas 54 . Histological analysis also evidenced the presence around active MS lesions of myeloid cells that have a phenotype consistent with mo-DCs and are found to interact with numerous lymphocytes in situ 55 . Specific blocking of monocyte differentiation into mo-DCs, while preserving mo-Mac development, could therefore provide clinical benefits in neuroinflammation. Our results identify ETV6 as a candidate target to reorient monocyte fate decision for therapeutic strategies. Collectively, our findings suggest that active repression of mo-Mac differentiation is required to allow monocyte differentiation toward mo-DCs. We propose a model whereby the mo-DC fate commitment in response to external cues (such as IL-4 and TNF) would require both the activation of the mo-DC differentiation program by factors, including IRF4, and the transcriptional repression of mo-Mac development by factors, including ETV3 and ETV6. Given the central role of mo-DCs in fueling pathogenic inflammation in numerous chronic inflammatory diseases, our work should have important implications for the therapeutic manipulation of monocyte differentiation.

Online content
Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/s41590-022-01374-0.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/.

Human samples
Buffy coats from healthy donors (both male and female donors) were obtained from Etablissement Français du Sang (Paris) in accordance with INSERM ethical guidelines. According to French Public Health Law (art L 1121-1-1, art L 1121-1-2), written consent and institutional review board approval are not required for human noninterventional studies.

Mouse strains
Cx3Cr1-CreER were obtained from Jackson Laboratories (stock no. 021160). Cx3Cr1-CreER expresses the enhanced YFP from endogenous Cx3cr1 promoter/enhancer elements. Etv6 flox/flox mice were obtained from H. Hock 35 . Cx3cr1-Etv6 Δ mice were generated by crossing Cx3Cr1-CreER +/− mice with Etv6 flox/flox mice. Cd11c-Etv6 Δ mice have been described previously 32 . Cx3Cr1-CreER −/− Etv6 flox/flox or CD11c-CreER −/− Etv6 flox/flox littermates were used as WT controls, respectively. All mice were on a C57BL/6 background. Mice were maintained under specific pathogen-free conditions at the animal facility of Institut Curie or New York University School of Medicine, in accordance with institutional guidelines. Mice were housed in a 12 h light:12 h dark environment, with free access to water and food. Both male and female mice were used at age 7-9 weeks. All animal procedures were in accordance with the guidelines and regulations of the French Veterinary Department (authorization APAFIS no. 25217-2020042522586261 v.1) or the institutional animal care and use committee of New York University School of Medicine, and approved by the local ethics committee.

Monocyte isolation and culture
PBMCs were prepared by centrifugation on a Ficoll gradient (Lymphoprep, STEMCELL). Blood CD14 + monocytes were isolated from healthy donors' PBMCs by positive selection using magnetic beads (Miltenyi). Monocytes were 95-98% CD14 + CD16 − as assessed by flow cytometry. Monocytes (2 × 10 6 cells ml −1 ) were cultured for 5 d in RPMI-Glutamax medium (Gibco) supplemented with antibiotics (penicillin and streptomicin) and 10% fetal calf serum in the presence or absence of 100 ng ml −1 of M-CSF (Miltenyi), 5 ng ml −1 of IL-4 (Miltenyi) and 5 ng ml −1 of TNF-α (R&D Biotechne). Cytokines were added only at the start of the culture and the medium was not refreshed during the course of the culture. CD16 + or CD1a + cell populations were isolated by cell sorting on a FACSAria instrument (BD Biosciences). In some experiments, monocytes were cultured in the presence of 100 ng ml −1 of M-CSF (Miltenyi), 5 ng ml −1 of IL-4 (Miltenyi) and 20 ng ml −1 of TNF (R&D Biotechne), or in the presence of IFN-α (recombinant human (rh) interferon-alpha 1b, Immunotools, catalog no. 11343596) or IFN-β (generated in-house by the platform of recombinant proteins of Institut Curie).

Imaging flow cytometry
Cells were first stained with Live/Dead Aqua (Thermo Fisher Scientific) in PBS for 10 min at 4 °C. Then, cells were stained in PBS containing 0.5% human AB serum and 2 mM EDTA with anti-CD1a APC and anti-CD16 FITC for 30 min on ice. After washing, cells were fixed with paraformaldehyde 4% in PBS for 20 min at room temperature and permeabilized with Permeabilization Buffer (Fixation/Permeablization Kit, BD Biosciences) containing Fc block (Human TruStain FcX, BioLegend) and mouse serum (BioLegend) for 30 min on ice. Cells were then incubated with the primary antibody in permeabilization buffer, rabbit anti-ETV6/Tel (Novus Biologicals, NBP1-80695, dilution 1:1,000) or rabbit anti-ETV3 (Atlas Antibodies, HPA004794, dilution 1:1,000), at 4 µg ml −1 for 1 h at room temperature. Finally, cells were incubated with the secondary antibody anti-rabbit immunglobulin G (H + L) Alexa Fluor-594 (Molecular Probes, catalog no. A-11037, dilution 1:500) for 1 h at room temperature and then resuspended in staining buffer containing DAPI (Thermo Fisher Scientific, 50 ng ml −1 ). Samples were acquired in an Amnis ImageStream instrument (Luminex). Data were analyzed using the IDEAS software to obtain the similarity score between DAPI and Alexa Fluor-594 channels. Finally, data were exported to FlowJo for quantification and visualization.

Phenotypic analysis of mouse tissues
For phenotypic analysis, Cx3cr1-Etv6 Δ mice and WT (Etv6 flox/flox ) littermates were treated with 5 mg of tamoxifen (Sigma-Aldrich) resuspended in corn oil (Sigma-Aldrich) by oral gavage for 3 consecutive days and sacrificed 5 d after the last treatment.
BMs were flushed from one leg and filtered using 40-µm cell strainers; 50 µl of of blood was used and incubated twice with red blood cell (RBC) lysis buffer (Sigma-Aldrich) for 5 min at room temperature. Spleens were cut into small pieces and incubated for 30 min at 37 °C in a digestion mix (RPMI containing 0.4 mg ml −1 of DNAse I (Sigma-Aldrich) and 0.5 mg ml −1 of collagenase D (Roche)). Spleen suspensions were then incubated with RBC lysis buffer for 5 min and filtered using 40-µm cell strainers. Peritoneal lavage was recovered by intraperitoneal injection of 5 ml of PBS.  Supplementary Table 4. After washing, cells were resuspended in staining buffer containing DAPI (100 ng ml −1 ). Cells were acquired on a ZE5 flow cytometer (BioRad). Supervised analysis was performed using FlowJo software.

Quantitative PCR
For the analysis of BM monocytes, monocytes were purified with EasySep mouse monocyte isolation kit (STEMCELL) according to the manufacturer's recommendations. Cells from the peritoneal lavage or the spleen were sorted on a FACSAria Fusion instrument before lysis.
Cells were harvested and lysed in RLT buffer (QIAGEN). RNA extraction was carried out using the RNAeasy micro kit (QIAGEN) according to the manufacturer's instructions. Total RNA was retro-transcribed using the superscript II polymerase (Invitrogen), in combination with random hexamers, oligo-dT and dNTPs (Promega). Transcripts were quantified by RT-PCR on a 480 LightCycler instrument (Roche). Reactions were carried out in 10 µl, using a master mix (Eurogentec), with the following Taqman

ScRNA-seq
Monocytes were purified from two individual donors. Cells were barcoded per donor (donors A and B) using TotalSeq-anti-human Hashtag antibody (catalog nos. A0251 and A02052, respectively; BioLegend) according to the manufacturer's instruction. Barcoded cells were counted and mixed in a 1:1 ratio. Single-cell suspension was loaded into 10x Genomics Chromium. Libraries were prepared as per the manufacturer's protocol (Chromium Single Cell 3′ Reagent Kits v.3 protocol) and sequenced on an Illumina NovaSeq sequencer according to the 10x Genomics recommendations (paired-end reads) to a depth of approximately 50,000 reads per cell.
Initial processing was performed using Cell Ranger (v.3.1.0) and subsequent analysis with Seurat v.4.0 workflow 56 . Hashtag demultiplexing was performed using the function HTODemux() and positive.quantile = 0.99. Cells with >20% of mitochondrial genes or genes expressed in <3 cells were filtered out. Graph-based clustering, visualization and DEG analyses were performed using Seurat v.4.0. For clustering analysis, FindNeighbors() and FindClusters() functions of the Seurat package were used with the first 50 significant principal components (PCs) and a resolution of 1.3, respectively. For identification of DEGs, FindMarkers or FindAllMarkers function (test.use = 't', logfc.threshold = log[0.25]) were used based on normalized data. DEGs with P adj > 0.05 were filtered out. Data have been deposited in the National Center for Biotechnology Information's (NCBI's) Gene Expression Omnibus (GEO) and can be accessed at accession no. GSE211603.

RNA-seq library preparation
Monocytes were cultured for 3 d in the presence of 100 ng ml −1 of M-CSF, 5 ng ml −1 of IL-4 and 20 ng ml −1 of TNF. Total RNA was extracted using the RNAeasy minikit (QIAGEN) including on-column DNase digestion according to the manufacturer's protocol. The integrity of the RNA was confirmed in BioAnalyzer using RNA 6000 Pico kit (Agilent Technologies) (8.8 < RNA integrity no. (RIN) < 10). Libraries were prepared according to Illumina's instructions accompanying the TruSeq Stranded mRNA Library Prep Kit (Illumina). RNA, 500 ng, was used for Nature Immunology Article https://doi.org/10.1038/s41590-022-01374-0 each sample. Library length profiles were controlled using the LabChip GXTouchHT system (Perkin Elmer). Sequencing was performed in four sequencing units of NovaSeq 6000 (Illumina) (100-593 nt length reads, paired end) with an average depth of 40 × 10 6 clusters per sample. RNA-seq data discussed in this publication have been deposited in the NCBI's GEO and can be accessed at accession no. GSE188982.
GSEA GSEA 61 was performed using the GSEA software (v.4.0.3) with the default parameters, except for the number of permutations that we fixed at n = 1,000. The count matrix from RNA-seq studies was first normalized using DESeq2. Gene signatures of blood monocytes, mo-DCs and mo-Macs were designed from microarray data 18 .

Analysis of public RNA-seq data
ScRNA-seq data from healthy controls and patients with the ETV6 P214L mutation 24 were downloaded from BioProject, accession no. PRJNA657295 and processed using Cell Ranger (v.2.1.0) using the human reference genome (hg38). Count matrices were then integrated and analyzed using Seurat v.4. Cells with <200 genes detected or percentage of mitochondrial genes >15%, as well as genes detected in <3 cells, were excluded from the analysis. We used a reference-mapping approach to annotate cell labels in the query dataset using the MapQuery function of Seurat v.4. We first projected each query cell on to a previously computed Uniform Manifold Approximation and Projection (UMAP) visualization of the reference dataset 56 (Extended Data Fig. 2a,b) and then, we subset the CD14 + and CD16 + monocytes. The dimensionality reduction, PC analysis and UMAP projection were performed in this subset (Extended Data Fig. 2c). Differential signature enrichment was calculated using Student's t-test comparing the signature score in WT versus P214L patients.

ChIP coupled to sequencing
Monocytes from 2 individual donors were cultured for 3 d in the presence of 100 ng ml −1 of M-CSF, 5 ng ml −1 of IL-4 and 20 ng ml −1 of TNF. ChIP was performed from 80 × 10 6 monocytes as the starting material, using the ChIP-IT High Sensitivity kit (Active Motif) following the manufacturer's instructions. Monocytes were homogenized using a 7-ml Dounce glass homogenizer (Clinisciences). Sonication was performed using a Bioruptor Pico instrument (Diagenode) using 20 cycles (30 s on and 30 s off). Chromatin was immunoprecipitated using antibodies against ETV3 (ref. A303-737A, polyclonal, Bethyl) or ETV6 (clone R1092.1.1A9, CDI Laboratories Inc.). Libraries were prepared according to Illumina's instructions accompanying the TruSeq ChIPSeq Library Prep Kit (Illumina). DNA, 4-6 ng, was used for each sample. Sequencing was performed in one sequencing unit of NovaSeq 6000 (Illumina) (100-to 593-nt length reads, paired end).
The quality of the sequencing data was assessed using FastQC 57 . Reads were aligned to the genome using BWA-Mem 62 . Peak calling was performed with MACS2 (ref. 63 ) using default parameters for each donor and the corresponding input. Final peaks were obtained by intersecting both donors using BEDTools 64 with the parameter '-f 0.5'. Peak assignment was performed on the intersected bed file using the annotatePeaks function of HOMER 65 , with the genome 'hg38' as a reference. Known motif enrichment analysis was performed using HOMER with the findMotifsGenome function, and the parameters 'hg38 -size given --mask'. ChIP-seq data have been deposited in NCBI's GEO and can be accessed at accession no. GSE211604.

Experimental peritonitis
Cx3cr1-Etv6 Δ mice and WT (Etv6 flox/flox ) littermates were treated with 5 mg of tamoxifen resuspended in corn oil by oral gavage for 3 consecutive days (days 0-2). On day 5, mice received a fourth gavage of tamoxifen and were injected intraperitoneally with 1 ml of 3.8% brewer's thioglycolate medium (Sigma-Aldrich). Mice were analyzed 3 d after thioglycolate injection.

Monocyte adoptive transfer
Cx3cr1-Etv6 Δ mice and WT littermates were treated with 5 mg of tamoxifen resuspended in corn oil by oral gavage for 3 consecutive days (days 0-2). On day 5, monocytes were isolated from BM using the EasySep Mouse Monocyte Isolation Kit according to the manufacturer's instructions. Between 0.7 × 10 6 and 1 × 10 6 monocytes were injected intraperitoneally in CD45.1 C57BL/6 mice that had been injected 18 h before with 1 ml of 3.8% brewer's thioglycolate medium. Peritoneal lavage was analyzed by flow cytometry 3 d after monocyte injection.

Experimental autoimmune encephalomyelitis
Cx3cr1-Etv6 Δ mice and WT (Etv6 flox/flox ) littermates were treated with 5 mg of tamoxifen resuspended in corn oil by oral gavage twice a week, starting 1 week before immunization. In some experiments (deletion in microglia), Cx3cr1-Etv6 Δ mice and WT littermates were treated with tamoxifen twice, then rested for 4 weeks before immunization. In some experiments, Cd11c-Etv6 Δ mice and WT (Etv6 flox/flox ) littermates were used. Mice were immunized subcutaneously in the back with 100 µg of the MOG35-55 peptide (sb-PEPTIDE) emulsified in incomplete Freud's adjuvant (Invivogen) supplemented with 4 mg ml −1 of desiccated Mycobacterium tuberculosis (Sigma-Aldrich, catalog no. H37RA). Mice were injected intraperitoneally with 200 ng of pertussis toxin from Bordetella pertussis (Calbiochem) at days 0 and 2 after immunization. Mice were examined daily for clinical signs. In agreement with the local ethics committee, mice were scored as follows: 0, healthy; 0.5, tail weakness; 1, limp tail; 1.5, tail paralysis and hindlimb weakness; 2, tail paralysis and limping of one hindlimb; 2.5, tail paralysis and limping of both hindlimbs; 3, paralysis of tail and both hindlimbs; and 3.5, paralysis of tail and both hindlimbs and weakness in forelimbs. A score of 3 was predefined as the humane endpoint of the experiment.

Lymph node cell analysis during EAE
Inguinal lymph nodes were collected 7 d post-MOG immunization. For flow cytometry of DCs, lymph nodes were cut into small pieces and incubated for 30 min at 37 °C in digestion mix: RPMI containing 0.5 mg ml −1 of DNAse I and 0.5 mg ml −1 of collagenase D. Cell suspensions were then filtered using 40-µm cell strainers. An antibody panel can be found in Supplementary Table 1. For tetramer staining, lymph nodes were dissociated by forcing through a 40-µm cell strainer. Cells were incubated for 3 h at 37 °C in RPMI containing 10% FCS in the presence of phycoerythrin (PE)-conjugated MOG tetramer (I-A(b) GWYRSPFSRVVH, 2.7 mg ml −1 ) or control tetramer (I-A(b) PVSKMRMAT-PLLMQA, 2.7 mg ml −1 ) (both obtained from the National Institutes of Health (NIH) tetramer core facility). After washing, cells were stained with anti-CD8 BUV395 (BD Bioscience, clone H35-17.2), T cell receptor (TCR)-β BUV737 (BD Bioscience, clone H57-597), CD19 BV480 (BD

Statistical analysis
Sample-size calculations were performed using InVivoStat (v. 4.2). For experiments on human cells, based on our previous results, we estimated the interdonor variability to 100% coefficient of variation (CV) in these experiments. In these conditions, n = 5 is sufficient for 70% power and n = 6 for 80% power to detect biologically significant results with a 5% significance level. Therefore, we used n = 5-6 as a minimum group size for these experiments. For animal experiments, based on our previous results, we estimated the interanimal variability to 50% CV in these experiments. In these conditions, n = 9 is sufficient for 80% power to detect a doubling of response between groups. Therefore, we used a minimum of nine mice in these experiments. For RNA-seq analysis, we used five samples per group, based on the literature for optimal group size in RNA-seq experiments 66 . Wilcoxon's matched-paired test, Mann-Whitney U-test and unpaired Student's t-test were performed using Prism v.9 (GraphPad Software). All statistical tests were two sided. Data distribution was assumed to be normal but this was not formally tested. Statistical details for each experiment can be found in the corresponding figure legend; n corresponds to the number of biological replicates (individual human donors or individual mice). No animal or data points were excluded. Data collection was performed blinded and randomized. Blinding was not possible for EAE because measurements were performed longitudinally for each mouse. Blinding was not performed for immunoblots to allow silenced samples to be presented side by side.

Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability
Sequencing data have been deposited in the GEO under accession nos. GSE188982 (RNA-seq), GSE211603 (scRNA-seq) and GSE211604 (ChIP-seq). Publicly available data were obtained from the Immgen database (www.immgen.org), BioProject (accession no. PRJNA657295) and the reference genome hg38 from Ensembl (www.ensembl.org). All other data are in the article and supplementary files are available from the corresponding author upon reasonable request. Source data are provided with this paper.  (e) Top differentially expressed genes between healthy (WT) and ETV6 P214L monocytes (n = 8 for healthy and n = 5 for ETV6 P214L mutation).