Aire mediates the expression of tissue-specific antigens in thymic epithelial cells to promote tolerance against self-reactive T lymphocytes. However, the mechanism that allows expression of tissue-specific genes at levels that prevent harm is unknown. Here we show that Brg1 generates accessibility at tissue-specific loci to impose central tolerance. We found that Aire has an intrinsic repressive function that restricts chromatin accessibility and opposes Brg1 across the genome. Aire exerted this repressive influence within minutes after recruitment to chromatin and restrained the amplitude of active transcription. Disease-causing mutations that impair Aire-induced activation also impair the protein’s repressive function, which indicates dual roles for Aire. Together, Brg1 and Aire fine-tune the expression of tissue-specific genes at levels that prevent toxicity yet promote immune tolerance.


Functional competence of the immune system requires the selection of T cells that bolster defense against pathogens and tumors while silencing T cells that react against self-constituents. T cell repertoire selection is limited by the range of self-antigens presented in the thymus1. The ectopic expression of thousands of tissue-specific self-antigens encompassing all parenchymal organs in the thymus induces immune tolerance to these self-antigens2,3. This ectopic transcription is driven largely by Aire, which operates in medullary thymic epithelial cells (mTECs), and mutations in this transcription factor cause autoimmune polyendocrine syndrome type 1 (APS-1)4,5.

Tissue-specific expression programs are generally defined by the developmentally coordinated actions of positive and negative regulators that influence the recruitment and release of RNA polymerase II (Pol II) at enhancers and promoters of lineage-specific genes6. Aire releases paused Pol II for productive elongation7,8,9,10, but the mechanisms that promote accessibility and Pol II binding are unclear. The determinants of this poising mechanism in mTECs are distinct from those in peripheral tissues, as the lineage-specific transcription factors essential in the latter context are unnecessary for thymic expression11.

While Aire has a unique ability to activate a large spectrum of peripheral-tissue-specific genes (PTGs), mTECs must limit the expression of these genes because many of them encode proteins that can perturb physiological processes. Indeed, the expression of Aire-induced PTGs in mTECs is orders of magnitude lower than that in the respective peripheral tissues10,12,13. The expression of PTG subsets within individual mTECs is transient and shuttles between distinct gene subsets, thereby reducing the number of mTECs necessary to present the entire PTG repertoire14. Elucidation of how the duration and amplitude of tissue-specific expression are controlled is essential for a full understanding of how Aire induces precisely quantified transcription.

Here we report that Aire represses chromatin accessibility through multimerization and histone-binding activities to restrain the amplitude of PTG transcription. We also found that Brg1 promotes accessibility at loci encoding tissue-specific antigens, thereby allowing Aire to function.


mTEChi differentiation promotes chromatin accessibility

To elucidate the Aire-independent mechanisms that poise tissue-specific genes in mTECs, we examined the developmental control of chromatin accessibility during mTEC differentiation to determine when tissue-specific loci become accessible. Using assay for transposase-accessible chromatin followed by deep sequencing (ATAC-seq)15 and gene expression profiling, we defined accessibility landscapes and transcriptomes for progenitor mTEC (mTEClo) and Aire+ differentiated mTEC (mTEChi) samples, which are distinguished by their surface expression of major histocompatibility complex class II (MHC II) (Supplementary Fig. 1a–e). We focused our ATAC-seq analyses on the main axis of variance from principal component analysis that separated mTEChi and mTEClo samples (Supplementary Fig. 1f), to reduce the contribution of batch effects and other technical noise16. Using MAVRIC (D.M.M. and W.J.G., unpublished data), a bioinformatic framework that attributes axes of variance to underlying sources (ATAC-seq peaks in this case), we identified 59,818 differentially accessible peaks that correlated with mTEChi maturation, accounting for approximately one-fourth of the total accessible genome detected (Fig. 1a). Nearly 70% of these peaks showed increases in accessibility during mTEChi differentiation and were enriched for their nearest gene being tissue-specific (Fig. 1a–d and Supplementary Fig. 1g). Cells in the mTEClo state had low levels of accessibility at these regions, similar to those in embryonic stem cells (ESCs) and naive CD8+ T cells (Fig. 1d). Polycomb repressive complexes were enriched at these sites in ESCs (Supplementary Fig. 1h), which suggested that these regions were likely to be within facultative heterochromatin, consistent with the reported enrichment of trimethylation of histone H3 at lysine 27 (H3K27me3) and depletion of Pol II binding and active histone modifications at Aire-regulated tissue-specific genes17. Accessibility changes at mTEChi-induced, but not mTEChi-repressed, peaks robustly correlated with the neighboring PTGs’ transcriptional changes (Fig. 1c). We observed that 90% of these peaks were >1 kb away from the nearest transcriptional start site (Fig. 1e). These results indicate a profound shift in the chromatin landscape during mTEChi differentiation that is driven by accessibility changes at distal cis-regulatory elements that control the transcription of tissue-specific genes.

Fig. 1: mTEChi differentiation promotes chromatin accessibility at tissue-specific loci.
Fig. 1

a, A heat map of normalized ATAC-seq fragment density at differential peaks (columns). WT, wild-type. b, Tissue restrictions of the indicated peak sets (filtered for peaks with accessibility fold-change >2) assessed on the basis of the tissue expression profile of the nearest gene. c, A cumulative distribution function plot of the transcriptional fold-change of tissue-specific gene expression between mTEChi and mTEClo cells near ATAC-seq peaks classified by differential accessibility changes. n = 9,857 and 4,096 for induced and repressed peaks, respectively. We used Mann–Whitney U-tests (two-tailed) for comparisons between classes of peaks. Confidence intervals (CI; 95%) are specified for each comparison. d, Density plots of ATAC-seq fragment dyads from mTEChi, mTEClo, ESC and CD8+ T cell groups at induced peaks near tissue-specific genes upregulated in mTEChi cells (left) or near housekeeping genes (right). e, Clustered correlation of chromatin accessibility at ATAC-seq peaks distal (left) and proximal (right) to tissue-specific genes with induced transcription in mTEChi versus mTEClo cells. The data shown are from n = 4 independent experiments (a,e) or are from pooled data representative of 4 independent experiments (c,d).

Aire and Brg1 have opposing influences on chromatin state in mTEChi samples

Subunits of the mSWI/SNF or BAF chromatin-remodeling complexes appeared in a hypothetical network for functional partners for Aire8, which suggested that BAF complexes specifically contribute to the poising of PTGs in mTECs. Thus, we profiled the accessibility landscapes in mTEChi and mTEClo samples from Foxn1ex9Cre;Brg1F/F mice (referred to as Brg1–conditional knockout (cKO) mice hereinafter), in which Smarca4, encoding the catalytic subunit Brg1, is deleted in thymic epithelial cells. The frequencies of mTEC compartments and the punctate localization of Aire were largely normal in Brg1-cKO mice (Supplementary Fig. 2a–c). To compare the influence of Brg1 on the mTEChi chromatin state with that of Aire, we profiled the accessibility landscapes in mTECs from Aire-knockout (Aire-KO) mice. Hierarchical clustering based on the correlations of accessibility at mTEChi-induced peaks near PTGs separated the Brg1-cKO and Aire-KO mTEChi samples (Fig. 2a), indicating distinct influences of Brg1 and Aire on the mTEChi chromatin landscape. Aire-KO and Brg1-cKO mTEChi samples distributed at opposite poles of the principal component (PC) of variability attributable to the genotypes of mTEChi samples (PC2) (Fig. 2b), which suggests divergent influences of Aire and Brg1 on the mTEChi-specific accessibility state.

Fig. 2: Aire and Brg1 are determinants of an mTEChi-specific chromatin state with opposing influences on accessibility.
Fig. 2

a, Correlations between the indicated ATAC-seq samples at tissue-specific regions exhibiting induced accessibility during mTEChi differentiation. b, Principal component (PC) analysis of ATAC-seq data. c, Comparison of the influences of Aire and Brg1 on mTEChi-specific accessibility states (ATAC-seq peaks correlated to PC2). Peaks whose nearest genes are upregulated (up) in mTEChi versus mTEClo cells are highlighted in orange and enumerated in each Cartesian quadrant (Quad) with the total for all peaks (blue). Frag., fragments. P values were determined by one-sided Fisher’s exact test with 95% CI (lower bound). d, Aire (top) and Pol II (bottom) ChIP-seq fragment dyad density at Aire-induced or Aire-repressed ATAC-seq peaks. e, Heat maps of Aire (top) and Pol II (bottom) ChIP-seq fragment dyad density at Aire-repressed ATAC-seq peaks. f, Aire (top) and Pol II (bottom) ChIP-seq fragment dyad density at Brg1-induced or Brg1-repressed ATAC-seq peaks. g, Genomic signal tracks of ATAC-seq fragments at two loci from the indicated mTEChi samples, and ChIP-seq signal tracks from wild-type mTEChi samples. Red arrowheads indicate differentially accessible regions. WT, wild-type; a.u., arbitrary units. The data shown are from n = 4 independent experiments (ac) or are from pooled data representative of 2 independent experiments (cg).

Using MAVRIC, we identified peaks correlated to PC2, and found a negative correlation (r = –0.43) between accessibility promoted by Aire and that promoted by Brg1 (Fig. 2c). Using arbitrary fold-change cutoffs to classify induced and repressed peaks (>2 and <0.5, respectively), we generated cumulative distribution function plots, which showed that Brg1-induced peaks were predictive for Aire repression and that Aire-induced peaks were predictive for Brg1 repression (Supplementary Fig. 2d). These data indicate that Aire and Brg1 exert opposing forces at the same sites. To assess the functional relevance of these regions for mTEChi differentiation, we assigned the nearest gene to each ATAC-seq peak. mTEChi-activated genes were far more frequently near peaks where accessibility was repressed by Aire and induced by Brg1 than near peaks induced by Aire and repressed by Brg1, or repressed by both Aire and Brg1 (Fig. 2c). To validate the negative influence of Aire on chromatin accessibility, we analyzed published ATAC-seq samples18, and found increases in fragment density in Aire-KO cells compared with wild-type mTEChi at peaks correlated to the primary axis of variance separating the genotypes (Supplementary Fig. 2h,i). Taken together, these data show that Aire has a repressive function at BAF-promoted accessible regions near genes upregulated during mTEChi differentiation.

To investigate whether Aire directly mediates these accessibility changes, we used published data18 from chromatin immunoprecipitation followed by sequencing (ChIP-seq) of mTEChi to assess Aire occupancy at differentially accessible peaks. Aire localization was robustly enriched at regions where chromatin accessibility was repressed by Aire (Fig. 2d–g and Supplementary Fig. 2g,j). Furthermore, Aire bound to regions where accessibility was induced by Brg1 (Fig. 2f,g and Supplementary Fig. 2g), which suggested a direct role for Aire in repressing accessibility at Brg1-promoted sites. These Aire-repressed, Brg1-promoted regions were mostly gene-distal, with only ~7% of peaks within promoters (Supplementary Fig. 2e,f). These regions were also enriched for Pol II, acetylation of histone H3 at lysine 27 (H3K27ac), topoisomerases I and IIa (Top1 and Top2a, respectively), and γ-H2AX (Fig. 2d–g and Supplementary Figs. 2g and 3), which further supported their enhancer potential. These data suggest that Aire directly opposes the accessibility that is promoted by BAF.

Accessibility at PTGs is promoted by Brg1 and repressed by Aire

To gain insight into the mechanism underlying Aire-induced ectopic expression of tissue-specific genes in mTECs, we focused our analyses on putative regulatory regions that promote the expression of PTGs during mTEChi differentiation. We defined these regions as those with ATAC-seq peaks where accessibility was specifically induced in mTEChi compared with mTEClo cells and was bound by Aire, and for which the nearest gene was tissue-restricted. In our data, 2,255 peaks met these criteria (Fig. 3a), which, considering the heterogeneity of individual mTECs expressing distinct subsets of the collective PTG repertoire10,11,13,14,19, is probably only a fraction of the true total. These qualifying regions presumably reflect PTGs with a higher probability of expression (up to 48% detected) in individual mTECs11. The accessibility changes at these peaks showed strong correlation with the transcriptional changes during mTEChi maturation (Fig. 3b), which indicates that these peaks may serve as cis-regulatory elements that control PTG expression.

Fig. 3: Accessibility at tissue-specific loci is promoted by Brg1 and repressed by Aire.
Fig. 3

a, MA plot of differential ATAC-seq peaks between mTEChi and mTEClo samples (blue), highlighting those for which the nearest gene is tissue-specific (light blue) and those that have Aire occupancy and neighboring tissue-specific genes (red). Frag., fragments. b, A cumulative distribution function plot of the transcriptional fold-change of tissue-specific genes between mTEChi and mTEClo samples near Aire-bound, ATAC-seq peaks classified by differential accessibility changes. n = 2,248 induced peaks. P values determined by two-tailed Mann–Whitney U-test with 95% confidence intervals (CI). c, Density plots of ATAC-seq fragment dyads from mTEChi samples of the indicated genotypes at Aire-bound, ATAC-seq peaks near tissue-specific genes (left) or housekeeping genes (right). a.u., arbitrary units. d, A heat map of normalized ATAC-seq fragment dyad density between samples of the indicated genotypes at mTEChi-induced, Aire-bound, tissue-specific loci. e, A heat map of Aire ChIP-seq fragment dyad density at the same regions as in d. The data shown are from n = 4 (a) or n = 2 (d) independent experiments or are pooled data representative of 4 (b) or 2 (c,e) independent experiments.

To determine whether BAF promotes the poising of these tissue-specific regions by facilitating their accessibility, we quantified the ATAC-seq fragment density at these sites, and found a substantial reduction of fragment density in Brg1-cKO mTEChi cells compared with that of wild-type cells (Fig. 3c,d). In contrast, the ATAC-seq fragment density at these sites in Aire-KO mTEChi samples was increased (Fig. 3c–e), which indicates that Aire repressed accessibility at these regions near transcriptionally upregulated PTGs. Furthermore, the increase in chromatin accessibility at these tissue-specific loci in mTEChi compared with mTEClo samples was also present in Aire-KO mTEChi samples (Supplementary Fig. 4a), which indicates that accessibility at these loci was induced in Aire-deficient mTECs. In contrast, the transcriptional activation of Aire-dependent tissue-specific genes was compromised in Aire-KO mTEChi cells compared with that in wild-type cells (Supplementary Fig. 4b). Similarly, the transcriptional activation of PTGs during the mTEClo-to-mTEChi transition in wild-type mice was largely Aire-dependent (Supplementary Fig. 4d). In contrast, Aire was broadly dispensable for the induction of accessibility at PTGs during normal mTEChi differentiation (Supplementary Fig. 4c). These data are consistent with reports that the amount of DNA methylation and promoter-proximal transcription at Aire-induced PTGs does not differ between Aire-KO and wild-type mTECs7,10. Taken together, these results suggest that Brg1 is essential for the accessibility near PTGs in mTEChi cells that Aire later targets for transcriptional activation. Moreover, the negative influence of Aire on accessibility may restrain the transcriptional amplitude at PTGs.

NF-κB target sites are hypersensitive to regulation by Aire and Brg1

Because the eight DNA-binding domains of the BAF subunits have only limited sequence specificity20 and Aire does not show sequence-specific DNA-binding21,22, we sought to identify transcription factors that potentially work in tandem with either complex to promote the mTEChi-specific chromatin state. Steric hindrance between transcription factors and the Tn5 transposase can be inferred from ATAC-seq data, which allows enrichment analyses for transcription factor footprinting15. Using an established framework to quantify the changes in accessibility across ATAC-seq peaks sharing known transcription factor motifs23, we identified a battery of transcription factors that were differentially enriched in wild-type mTEChi compared with mTEClo samples. Peaks with motifs for the transcription factors c-Rel, RelA and NF-κB1 showed the most prominent increases in accessibility (Fig. 4a and Supplementary Fig. 5a), consistent with the essential roles of NF-κB signaling in mTEChi maturation24,25. We also inferred a differential increase in binding of NF-κB complexes in mTEChi versus mTEClo samples from transcription factor footprinting (Supplementary Fig. 5e). In addition, we noticed substantial decreases in accessibility at sites that contained STAT, P53 and P63 motifs in mTEChi versus mTEClo samples (Fig. 4a and Supplementary Fig. 5a), along with putative loss of binding of these factors (Supplementary Fig. 5f,k,l,n–p). These data are consistent with the reported roles of Stat3 and p63 in the expansion and survival of the progenitor mTEClo compartment26,27,28.

Fig. 4: Regions that contain NF-κB motifs are highly sensitive to opposing regulation by Aire and Brg1.
Fig. 4

a, A heat map of deviation scores from expected ATAC-seq signal in mTEChi and mTEClo samples at ATAC-seq peaks containing known transcription-factor motifs. b, The change in deviation scores relative to the expected accessibility signal at ATAC-seq peaks containing known transcription-factor motifs. Diamonds indicate means; circles indicate replicates. c, Accessibility footprints at NF-κB motifs within ATAC-seq peaks from the indicated mTEChi samples. d, The top known motif enrichment within Aire ChIP-seq peaks. e, A heat map of Aire ChIP-seq fragment dyad density at ATAC-seq peaks with NF-κB motifs. WT, wild-type; a.u., arbitrary units. The data shown are from n = 4 (a) or 2 (b) independent experiments or are pooled data representative of 2 independent experiments (ce).

Because NF-κB depends on pre-existing accessibility for binding29, we next sought to determine whether NF-κB binding requires BAF chromatin remodeling to impose the mTEChi state. Indeed, Brg1 promoted chromatin accessibility most robustly at peaks that contained NF-κB motifs (Fig. 4b and Supplementary Fig. 5b,d). Furthermore, NF-κB footprinting and the flanking accessibility at these sites were substantially reduced in Brg1-cKO mTEChi cells compared with levels in wild-type cells (Fig. 4c). These data indicate that BAF poised the chromatin landscape to allow NF-κB binding at target loci and drive mTEChi differentiation.

Activation of NF-κB signaling in mTECs precedes the expression of Aire25. To determine the influence of Aire on NF-κB- and Brg1-induced chromatin accessibility, we quantified the changes in accessibility between Aire-KO mTEChi and wild-type mTEChi samples across peaks that contained each of the known motifs. We observed the greatest changes at peaks with NF-κB motifs, the accessibilities of which were repressed by Aire (Fig. 4b and Supplementary Fig. 5c,d). Moreover, the flanking accessibility and magnitude of the NF-κB footprinting were increased in Aire-KO mTEChi samples compared with those in wild-type controls (Fig. 4c). Robust Aire occupancy at NF-κB-motif-containing ATAC-seq peaks indicated direct repression by Aire at these Brg1-promoted sites (Fig. 4d,e). Considering the central role of NF-κB activation in mTEChi differentiation and the negative regulation of mTEChi cellularity by Aire5,13,30,31, these results suggest that Aire directly curtails fundamental components of the mTEChi differentiation program by limiting chromatin accessibility. Indeed, Aire bound to chromatin and diminished chromatin accessibility near genes that encode factors essential for mTEChi differentiation, such as Cd80, H2-Eb2 and Tvp23a (also known as Fam18a) (Fig. 2g and Supplementary Fig. 2g).

Brg1 imposes immunological tolerance

Because Aire acts on a pre-existing chromatin landscape, we asked whether the induced accessibility at tissue-specific loci was important for ectopic transcription of PTGs and central tolerance induction. Changes in accessibility at tissue-specific loci during the mTEClo-to-mTEChi transition were robustly predictive of changes in accessibility between Brg1-cKO and wild-type mTEChi samples (Fig. 5a). Gene expression profiling of Brg1-cKO mTEChi cells showed that the activation of hundreds of PTGs was diminished compared with that in wild-type cells, especially at Aire-regulated genes (Fig. 5b). In addition, the top decile of Brg1-upregulated genes showed high enrichment for PTGs expressed in a single peripheral tissue (Fig. 5c). Taken together, these results suggest that BAF bolstered PTG expression by promoting accessibility at distal regulatory elements during mTEC differentiation. To determine how the compromised activation of PTGs in Brg1-cKO mice affects the self-reactivity of the T cell repertoire, we assayed the frequency of activated T cells in the periphery. Although the frequency of thymocyte compartments was largely normal in Brg1-cKO mice (Supplementary Fig. 6a,b), the frequency of CD4+CD44hiCD62Llo effector T cells was doubled compared with that in wild-type mice, with nearly half of the splenic T cell compartment consisting of activated CD44hiCD62Llo effector T cells in 6-month-old mice (Fig. 5d–f). However, the total number of splenic T cells in Brg1-cKO mice was approximately half that in wild-type mice (Supplementary Fig. 6e), and it is likely that this was due to compromised function of Brg1-deficient cortical TECs (cTECs) that promote early thymocyte differentiation (Supplementary Fig. 6c,d). We next asked whether the activated effector T cells in Brg1-cKO mice could provoke autoreactive tissue damage. We found substantial lymphocytic infiltration in the kidney, liver, lung, and lacrimal and salivary glands in 3–6-month-old Brg1-cKO mice (Fig. 5g,h). Immunohistochemistry showed strong CD3 staining in these tissue infiltrates (Supplementary Fig. 6h).

Fig. 5: Brg1 imposes immunological tolerance.
Fig. 5

a, A cumulative distribution function plot of accessibility fold-changes at tissue-specific peaks classified as differentially accessible in wild-type mTEChi versus mTEClo cells. n = 10,135 induced peaks. P values determined by two-tailed Mann–Whitney U-test with 95% confidence intervals (CI). Frag., fragments. b, The role of Brg1 in the ectopic induction of tissue-specific genes (n = 6,811) during mTEChi differentiation. Dashed lines indicate fold-changes of 1.5 and 0.66 for each comparison. Numbers in corners indicate the total (black) and Aire-upregulated (orange) genes in each quadrant. P values correspond to genes in the upper right quadrant (one-tailed Fisher’s exact test with 95% confidence intervals). c, The fraction of genes from the top decile of Brg1-activated genes in mTEChi cells in each indicated tissue-restriction group (the number of peripheral tissues expressing the gene) compared with that in a random group of genes. d, Frequencies of activated splenic CD4+ T cells in 4-week-old mice (each symbol represents an individual mouse). Mean ± s.e.m. **P = 0.002 (two-tailed t-test). e, The frequencies of activated splenic T cells from mice of the indicated genotypes as a function of age. f, Representative cytometry plots of the frequency of activated CD4+ T cells in spleen from 6-month-old mice of the indicated genotypes. PE, phycoerythrin; APC, allophycocyanin. g, Histological analysis of the indicated tissues from 3–6-month-old wild-type or Brg1-cKO mice, stained with hematoxylin and eosin (H&E) for infiltrating lymphocytes. Each hexagon represents an individual mouse. h, Representative H&E stainings showing the histopathology in diseased tissues from Brg1-cKO mice. Scale bars, 200 μm. i, Histological analysis of tissues from nude mice via H&E staining for infiltrating lymphocytes 14 weeks after thymus transplant from the indicated donors. Scale bars, 100 μm. The data shown are pooled data representative of 2 independent experiments (a), or are from n = 2 (b,c,i), 8 (d,g,h), or 18 (e) independent experiments, or 1 experiment representative of 4 independent experiments (f).

We next addressed whether the autoimmunity in Brg1-cKO mice was due to failure in the negative selection of self-reactive T cells and/or the positive selection of Foxp3+CD4+CD25+ regulatory T cells (Treg cells), with either stemming from the diminished expression of PTGs in mTECs. Brg1-cKO mice had normal frequencies, numbers and function of splenic Foxp3+ Treg cells compared with those in wild-type mice (Supplementary Fig. 6f,g), suggesting a functional Treg cell compartment. The autoimmunity in Brg1-cKO mice could be driven by reduced thymic output, which might extend the neonatal window of physiologic lymphopenia, thus allowing aberrant homeostatic proliferation of self-reactive T cells32. To test for this possibility, we took thymic stroma depleted of hematopoietic cells by treatment with 2-deoxyguanosine from wild-type mice, Brg1-cKO mice or both, and grafted them under the kidney capsules of wild-type athymic nude mice. This experimental setting allowed us to test whether the normalized thymic output and the provision of Treg cells from wild-type cotransplanted thymus could rescue the autoimmune pathology driven by the self-reactive T cells that escaped negative selection in the Brg1-cKO thymus. Mice transplanted with both wild-type and Brg1-cKO thymic stroma showed lymphocytic infiltration in multiple organs (Fig. 5i), which indicated that the activity of self-reactive T cells from the Brg1-cKO thymus was dominant. However, the range of organs affected in these cotransplanted chimeric hosts was limited to the liver and salivary gland, whereas in mice that received transplants of only Brg1-cKO thymic stroma, infiltrations were also present in the kidney and lacrimal glands (Fig. 5i). These data suggest a potential role for lymphopenia-induced proliferation and/or defective positive selection of Treg cells in the autoimmunity of Brg1-cKO mice. Taken together, these results underscore how Brg1 plays an important part in promoting central tolerance induction by facilitating accessibility at tissue-specific loci, an essential component of PTG expression and selection of a functional T cell repertoire.

Aire rapidly represses accessibility after recruitment to chromatin

To elucidate the mechanism of Aire’s repressive function, we used the chromatin in vivo assay (CiA) system33,34,35, which allows precise temporal control of the recruitment of chromatin regulators to specific genetic loci via chemically induced proximity in living cells. The resulting minute-by-minute assessment of the effects of recruitment allows one to distinguish between rapid direct biochemical actions of Aire and slow secondary and indirect effects, such as influences on cell cycle or the actions of factors encoded by Aire-target genes. Mouse embryonic fibroblasts (MEFs) from CiA mice have two arrays of transcription factor binding sites and GFP inserted into one allele of Oct4 (Pou5f1)33, an Aire-regulated, tissue-specific locus in mTEChi cells31 (Fig. 6a and Supplementary Fig. 7a). This locus (CiA:Oct4) is faithfully integrated into the endogenous chromatin context with appropriate long-range, topological conformations34. Furthermore, this locus is transcriptionally silent and allows assessment of effects independent of transcription. We generated stable cell lines from these mice that expressed the ZFHD1 DNA-binding domain fused to FKBP12 (a component of the mTOR signaling pathway) and Aire fused to the FRB domain of mTOR and V5 epitope, serving as the anchor and recruitment partner, respectively (Fig. 6a). The addition of rapamycin elicited the rapid association of FKBP12 and FRB, which caused the robust recruitment of Aire (>200-fold increase as measured by V5 ChIP) to the CiA:Oct4 locus within minutes (Fig. 6b). We assessed the effects of Aire recruitment to the CiA:Oct4 locus and found an extensive loss of chromatin accessibility at the recruitment site (as measured by DNase I sensitivity) within 15 min of rapamycin addition compared with that observed after control ethanol treatment (Fig. 6b). The magnitude of this loss of accessibility at the modified Oct4 locus seemed to be similar to that observed after 1 h or 5 d of Aire recruitment (Supplementary Fig. 7b), which suggests that Aire induced its maximum repressive effect within minutes of recruitment. These observations indicate that Aire has direct repressive activity. This rapid repression was in contrast to the quick increase in accessibility after recruitment of BAF complex to the CiA:Oct4 locus34,35 (Supplementary Fig. 7c,d). Furthermore, the transcriptionally silent state of the CiA:Oct4 locus indicates that the repressive activity of Aire was uncoupled to transcription (Supplementary Fig. 7e). In a distinct approach, we used an orthogonal recruitment strategy with dCas9 and MS2-binding aptamers36 in a human cTEC line37 (Supplementary Fig. 7f,g). We chose an open tissue-specific locus upstream of PSMB11 (which encodes the β5t subunit of the thymoproteosome38) to test the influence of Aire recruitment on accessibility. Aire recruitment upstream of PSMB11 induced loss of accessibility compared with the recruitment of dCas9 alone (Supplementary Fig. 7h), confirming the results at the CiA:Oct4 locus in MEFs. Taken together, these results show that Aire exerted its repressive influence within minutes of recruitment to an Aire-regulated locus independently of transcription, and thus indicate a direct biochemical function.

Fig. 6: Aire rapidly represses accessibility after recruitment to chromatin.
Fig. 6

a A schematic of the inducible CiA system: recruitment to the modified Oct4 locus via rapamycin (Rap)-induced dimerization of Aire–FRB and FKBP–ZFHD1 fusion proteins. b, Changes in DNase I hypersensitivity (left) at the CiA:Oct4 locus after Rap-induced (15 min) Aire recruitment, as measured by V5 ChIP (right). Circles represent individual replicates. Mean ± s.e.m. TSS, transcription start site. c, A depiction of mutations in CARD, SAND and PHD1 domains found in subjects with APS-1 (murine orthologs). d, Western blot of Aire–FRB (and mutants) transgenic expression in the CiA system. e, Changes in DNase I hypersensitivity (left) at the CiA:Oct4 locus after Rap-induced (15 min) recruitment of wild-type Aire or mutant variants, as measured by V5 ChIP (right). Circles represent individual replicates. Mean ± s.e.m. n.s., not significant; *P < 0.05, **P < 0.01, ***P < 0.001 relative to the wild-type (two-tailed t-test). EtOH, ethanol. The data shown are from n = 4 (b, left; e, left) or 3 (b, right; e, right) independent experiments or are from 1 experiment representative of 2 independent experiments (d).

Repression by Aire requires the PHD1 and multimerization domains

Aire directly interacts with chromatin through PHD1, a histone-binding module whose recognition of unmodified histone H3 tail is necessary for tolerance induction22,39,40. Because the CiA system uses a ZFHD1–FKBP anchor, the recruitment of Aire to the CiA:Oct4 locus does not depend on Aire’s own domains, and thus can be used to uncouple the targeting and repressive functions of Aire. The Aire C311Y mutation in humans (C313Y in mice) (Fig. 6c) disrupts the zinc coordination of PHD141, thereby preventing Aire from interacting with chromatin22. This mutation gives rise to APS-14. When we recruited an Aire-C313Y variant to the CiA:Oct4 locus via rapamycin administration, the DNase sensitivity of the locus remained similar to that in ethanol-treated cells, thus indicating that the repressive activity of Aire was abolished. These results were observed despite the similar protein expression and locus recruitment in cells expressing Aire-C313Y versus wild-type Aire (Fig. 6d,e). Because the unfolding of the PHD1 finger in the Aire-C313Y mutant might affect other domains, we next recruited an Aire variant with a designed mutation (Aire-D299A) that mitigates the histone-binding activity of Aire, but maintains the integrity of the PHD finger22,41. Aire-D299A was unable to repress accessibility at the CiA:Oct4 locus compared with wild-type Aire (Fig. 6e), which indicates the importance of the histone-binding module of Aire for both repressive and activating functions of the protein. The APS-1-related G228W mutation in the SAND domain of Aire (G229W in mice) exerts dominant-negative activity to cripple Aire localization42. In contrast to the PHD1 mutants, Aire-G229W decreased chromatin accessibility after recruitment to the CiA:Oct4 locus to levels similar to those seen with wild-type Aire (Fig. 6e), which indicates that the targeting defects of Aire-G229W were rescued by the ZFHD1–FKBP anchor.

Aire multimerizes through its caspase-recruitment domain (CARD) to form a multiprotein complex of ~750 kDa18,43. The APS-1-associated L28P mutation (L29P in mice) disrupts the ability of Aire to multimerize, and thus causes defects in the subnuclear localization and activation of target genes43. We observed ~50-fold less recruitment of Aire-L29P to the CiA:Oct4 locus compared with recruitment of wild-type Aire, despite the similar protein expression in the respective CiA MEFs (Fig. 6d,e). Recruitment of Aire-L29P did not repress accessibility at the CiA:Oct4 locus compared with that of wild-type Aire (Fig. 6e), which suggests that the multivalent conformation of Aire within an oligomerized complex was essential to its repressive function. Taken together, these results show that the repressive function of Aire is keyed by the same multimerization and histone-binding activities that are necessary for the promotion of ectopic transcription, and thus indicate a unique duality in the functional domains of Aire.

Aire recruitment to an active locus restrains transcriptional amplitude

If the repressive influence of Aire on chromatin accessibility serves to constrain the transcription of tissue-specific genes, directed Aire recruitment to an active locus should reduce gene expression. To directly test the effect of Aire on active transcription, we constitutively recruited the transcriptional activator VP16 to the CiA:Oct4 locus and sorted for cells that activated the GFP reporter encoded in the CiA:Oct4 locus, before a 3-d treatment with rapamycin to corecruit Aire (Fig. 7a). Both the intensity of GFP expression and the frequency of GFP+ cells were reduced after corecruitment of Aire and VP16 compared with that in cells not treated with rapamycin (Fig. 7a–c). In contrast, corecruitment of VP16 and the BAF complex to the CiA:Oct4 locus increased the frequency of GFP+ cells (Fig. 7d,e). The repressive influence of Aire was dependent in part on the histone-binding module and multimerization potential of Aire (Fig. 7f). These observations suggest that multimerized Aire and the recognition of unmodified histone H3 tails through the PHD1 finger imposed a barrier for active transcription. Taken together, these results indicate that during mTEC development, the initial poising by Brg1 and the subsequent transcriptional triggering by Aire were restrained by the repressive influence of Aire on chromatin accessibility, akin to the action of a rheostat in an electrical circuit (Supplementary Fig. 7i).

Fig. 7: Aire recruitment to an active locus restrains transcriptional amplitude.
Fig. 7

(a) Top, sequential recruitment of Aire (Rap) to the CiA:Oct4 locus for 3 d after VP16 recruitment; the frequency of GFP+ cells compared with that in the ethanol-treated control (EtOH) was assessed via flow cytometry. Bottom, after VP16 recruitment, GFP+ cells were sorted and then treated with or without Rap for 3 d. Numbers adjacent to outlines indicate the percentage of cells in the gate. b, The frequency of GFP+ cells after VP16 recruitment with or without corecruitment by Aire. Circles represent individual replicates. Mean ± s.e.m. c, Representative histogram of GFP expression in Rap-treated versus EtOH-treated cells described in a. d, Sequential recruitment of BAF via Ss18 (Rap) for 3 d after VP16 recruitment. e, The frequency of GFP+ cells quantified from the samples in d. Circles represent individual replicates. Mean ± s.e.m. f, Quantification of the fold-change in frequency (left) and mean fluorescence intensity (MFI; right) of GFP+ cells between EtOH-treated and Rap-treated conditions after sorting as in a. Circles represent individual replicates. Mean ± s.e.m. Significant differences in fold-change (Rap/EtOH) between each mutant and the wild type are indicated. n.s., not significant; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (two-tailed t-test). The data shown are from 1 experiment representative of 6 (a, top), 5 (a, bottom; c), or 7 (d) independent experiments or are from n = 6 (b), 7 (e) or 3 (f) independent experiments.


Here we provide new insight into the mechanisms that promote the activation of thousands of tissue-specific genes in the thymic epithelium at levels that are sufficient for immune tolerance, but below those that could be harmful. We found that Aire, the transcriptional regulator necessary for much of the tissue-specific gene expression in mTECs, has an intrinsic repressive function that restricts chromatin accessibility. Our analyses indicate that Aire targeted Brg1-induced, mTEChi-specific regulatory sites and repressed their accessibility to curtail mTEChi differentiation. Taken together with the repressive actions of Aire at tissue-specific loci, these findings suggest that Aire limits the duration and amplitude of transcription of tissue-specific genes, as well as the number of cells that could become competent for this ectopic expression. The physiological necessity of this negative regulation is apparent when one considers the deleterious consequences of overexpression of tissue-specific factors such as insulin, calcitonin, and blood coagulation factors in mTECs44,45,46.

Our recruitment studies with chemically induced proximity at an Aire-regulated locus indicated that Aire exerts its repressive influence within minutes of recruitment to chromatin, which suggests a direct biochemical mechanism. Because this locus is transcriptionally silent, the repressive influence of Aire on accessibility precludes any dependence on active transcription. This repressive function required key residues that facilitate the multimerization and histone-binding activities of Aire—the same residues necessary for the expression of tissue-specific genes and the prevention of autoimmunity4,40. The histone-binding activity of Aire stabilizes its interaction with target loci, and thus contributes to its targeting function39,40. However, the CARD-mediated multivalent state allows the simultaneous recognition of multiple histone tails that are likely to favor nucleosomal occupancy and perhaps repression of transcription initiation. To our knowledge, this intrinsic functional duality has not been described for a single transcription regulatory factor. A recent study demonstrated Aire’s positive influence on chromatin accessibility at super-enhancers of mTECs18. Indeed, when we recruited Aire to the CiA:Oct4 locus in ESCs (the recruitment site was within the super-enhancer region35), we observed an increase in the intensity of GFP indicating an absence of repression (data not shown). In contrast, recruitment of HP1 to the same locus in ESCs reduced the frequency of GFP+ cells to ~17% within 5 d33. Collectively, these results suggest that the endogenous context of a given locus contributes to Aire’s ultimate impact on transcriptional amplitude.

We found that the Brg1 catalytic subunit of the mSWI/SNF remodeling complex is essential for promotion of the mTEChi-specific chromatin state through increased accessibility at tissue-specific loci prior to their expression. The mSWI/SNF or BAF complex represents one of about 30 nonredundant ATP-dependent chromatin-remodeling complexes with often instructive roles in developmental lineage specification and tumor suppression. BAF antagonizes Polycomb occupancy at target sites via direct physical interaction and ATP-dependent eviction34,47. Considering the enrichment of Polycomb subunits and H3K27me3 at Aire-regulated loci in peripheral tissues and ESCs17, it is likely that BAF uses similar mechanisms in mTECs to facilitate the poising of tissue-specific genes. In addition, BAF complexes associate with Top2 for the resolution of facultative heterochromatin35,48,49. These observations are in agreement with our results, as the mTEChi-specific regulatory regions where accessibility was promoted by Brg1 showed enrichment of Top2a. Aire inhibits Top218,50, which suggests a mechanism by which Aire could reduce accessibility to serve a repressive function.

Our studies indicate that Aire is unique in that the same domains mediate both activating and repressive functions. Although thousands of tissue-specific loci become accessible to the transcription factors present in mTECs, these putative regulatory regions seem to be held in check by Aire’s ability to oppose accessibility through these bifunctional domains. This opposition potentially guards tissue-specific genes from the hundreds of transcription factors (e.g., NF-κB) expressed in mTECs, yielding safe but immunologically effective levels of tissue-specific antigens in the thymus.



B6 Aire+/− mice5 from Jackson Laboratory were bred to generate Aire−/− and Aire+/+ littermates. Brg1F/F and Foxn1ex9Cre mice51,52 on a C57BL/6 background were bred to generate Brg1F/F and Foxn1ex9Cre;Brg1F/F littermates. We used female and male littermate controls in all experiments. All mice were maintained in accordance with Stanford University’s Animal Care and Use Committee guidelines. The animal protocol is annually approved by the Administrative Panel on Laboratory Animal Care of the Research Compliance Office of Stanford University.

Cell lines

Mouse embryonic fibroblasts (MEFs) were acquired from mice with the CiA:Oct4 allele at embryonic day 14.5 as described33. MEFs were transformed with simian virus 40 large T antigen and single-cell sorted after transfection with LGmCreER (self-deleting) plasmid53 (Addgene #33340) to enrich for cells with an excised neo cassette. Clones were screened for growth rate, VP16-mediated eGFP activation and DNase accessibility at the CiA:Oct4 locus. A single clone was used for all CiA recruitment experiments. MEFs were grown in high-glucose DMEM (Life Technologies; 11960) supplemented with 10% FBS (Omega Scientific; FB-11), 10 mM HEPES, pH 7.5, Minimal AA, glutaMAX, Na-pyruvate, penicillin–streptomycin, and 2-mercaptoethanol at 37 °C, 5% CO2. The 4D6 human thymic epithelial cell line37 was a gift from Maria Toribio and Diane Mathis. The 4D6 cells were grown in RPMI 1640 medium (Life Technologies; 21870092) supplemented with 10% FBS (Omega Scientific; FB-11) and penicillin–streptomycin at 37 °C, 5% CO2.

Flow cytometry

mTEC subsets were isolated as described54,55. Thymi were dissected, and capsules were incised and triturated with glass pipettes to release thymocytes from stromal fragments. Stroma were digested with Liberase TM (Roche) and DNase I (Roche), and TECs were enriched using anti-CD45 MACS microbeads (Miltenyi) or centrifugation on a Percoll PLUS gradient (GE Healthcare). Enriched TECs were stained with fluorochrome-conjugated antibodies (BioLegend) to CD45 (30-F11), Ly-51 (6C3), MHC-II I-A/I-E (M5/114.15.2) and EpCAM (G8.8), along with fluorescein-labeled UEA-I (Vector Labs) and DAPI (Life Tech). Thymocytes were stained with antibodies to TCR-β (H57-597), CD4 (RM4-5), CD8 (53-6.7), CD25 (PC61) and CD69 (H1.2F3). Spleen and lymph nodes were isolated, minced and stained with antibodies to TCR-β, CD4, CD8, CD25, CD44 (IM7) and CD62L (MEL-14). Intracellular staining for Aire (5H12) and Foxp3 (FJK-16s) was done with the eBiosciences Foxp3 staining kit. All antibody stainings were preceded by FcγR block (2.4G2). Cell sorting was done on a FACS Aria II (BD), and data were collected with an LSR II flow cytometer (BD) and analyzed with FACS Diva (BD) and FlowJo (Tree Star).

Gene expression profiling

mTECs were sorted into Trizol LS (Life Technologies), and RNA was extracted according to the manufacturer’s instructions and then purified with the RNeasy micro kit (Qiagen). Purified RNA was processed, amplified, labeled and hybridized to Affymetrix GeneChip MoGene 2.0 ST microarrays as described56. Expression signals were normalized with the Signal Space Transformation-Robust Multi-Chip Average (SST-RMA) algorithm on the Affymetrix Expression Console software. Normalized signals were analyzed with the Transcriptome Analysis Console software (Affymetrix). Tissue restriction for each gene was determined by assignments from the previously reported dynamic step method17. Expression values for 64 nonthymic physiological samples from the GNF Mouse GeneAtlas V357,58 were calculated and hierarchically clustered into 35 groups. Guided by thresholds typically chosen for microarray data analysis, we defined tissue-restricted probes as those with a minimum normalized expression value of 50 that showed a moderated exponential step-up in expression, such that expression was substantially higher in tissue groups 1–5 than in the sixth-highest tissue group. Only genes with unanimously tissue-restricted probe sets were designated as tissue restricted.

ATAC-seq sample preparation

We sorted 60,000 cells into a V-bottom 2-ml tube (E&K Scientific) and added 1 ml of RSB buffer without detergent (10 mM Tris, pH 7.4, 10 mM NaCl, 3 mM MgCl2) before pelleting in a swinging bucket centrifuge. Cells were resuspended in 200 μl of RSB buffer with 0.1% Tween-20 (Sigma-Aldrich) and incubated on ice for 5 min. Cells were pelleted, resuspended in 50 μl of TD buffer with 2.5 μl of Tn5 transposase (Illumina) and incubated at 37 °C for 30 min. Transposition reactions were cleaned up with MinElute columns (Qiagen), and libraries were constructed as described59. Libraries were sequenced by paired-end, dual-index sequencing on an Illumina HiSeq 2000 instrument with 51 × 8 × 8 × 51 cycle reads.

ATAC-seq data analysis

ATAC-seq data were processed as described23,60. Reads were trimmed with a custom script23,60, aligned with Bowtie2, and filtered for unique reads with alignment quality >q30. Reads mapping to mitochondrial, ChrY and unmapped contigs were removed. Peaks were called with MACS2 and filtered with a custom blacklist23. Peak summits were extended ±250 bp and filtered for nonoverlapping, maximally significant 500-bp peaks (n = 250,313). ATAC-seq fragment counts for each sample were calculated across all 250,313 peaks, quantile-normalized and GC-content-normalized as described23,60. The nearest gene to each peak was annotated with HOMER. Principal component analysis was carried out on normalized data with the MAVRIC R package (D.M.M. and W.J.G., unpublished data) for the top 100,000 peaks with the highest variance (ntop = 100000, alpha = 0.01, effsize = 0.05, corMethod = “pearson”). Peaks with variance significantly correlated to a given principal component (α < 0.05) were identified by MAVRIC (dimdesc($pcaobj)). Hierarchical clustering was done by Pearson correlation as the distance metric on normalized data. Normalized bedGraph files were created for each sample with bedtools genomecov to visualize insertion tracks. The sum of reads within peaks associated with ‘housekeeping’ genes (median microarray signal >75, coefficient of variation <0.15, fold change for wild-type mTEChi versus mTEClo > 0.9 and < 1.1) was used as the scaling factor for each sample as described61. Transcription factor (TF) deviation scores were calculated as described23,60. We assessed TF footprinting by plotting the normalized distribution of the 5′ ends of fragments (replicates pooled) spanning a 300-bp window relative to the motif center. Similarly, we plotted the normalized distribution of fragment dyads (midpoint) spanning a 4-kb window relative to the indicated peak center or transcription start site.

ChIP-seq data analysis

Fastq files for Aire, Pol II, Top1, Top2a, γ-H2AX and H3K27ac ChIP-seq18 from mTEChi, and for Lsd1, CoREST, Hdac262, Suz1263, Ring1b64, and Lamin B65 in ESCs were aligned to the mouse genome (mm9) with Bowtie, using default settings except that reads were filtered with multiple alignments (command line parameter --m 1). We processed BAM files by removing duplicates, unaligned reads and reads aligned to ChrM. Fragment sizes were estimated and reads were extended accordingly for each replicate. MACS2 was used to call peaks using the -callpeak function with IgG ChIP as the control. Signal tracks were generated with bedtools genomecov and were normalized by sequence depth. These signal tracks were cross-validated for significance with MACS2, with commands bdgcmp --m logLR --p 0.00001 to generate log10-likelihood bedGraph files. Enriched TF motifs for peak sets were identified with HOMER ( --size given --mask).

Chromatin in vivo assay

MEFs were acquired from mice with the CiA:Oct4 allele at embryonic day 14.5 as described33. MEFs were transformed with simian virus 40 large T antigen and single-cell sorted after transfection with LGmCreER (self-deleting) plasmid53 (Addgene #33340) to enrich for cells with an excised neo cassette. Clones were screened for growth rate, VP16-mediated eGFP activation and DNase accessibility at the CiA:Oct4 locus. A single clone was used for all CiA recruitment experiments. ESCs from mice with the CiA:Oct4 allele were established as described33. We cloned murine Aire with a C-terminal FRB tandem repeat (Aire-Frb2x-V5) or GAL4-binding domain (Aire-Gal4BD-V5) into a previously described lentiviral backbone33 for rapamycin-induced or constitutive recruitment, respectively. BAF was recruited via the Ss18 subunit, with previously described constructs (Frb2x-V5-Ss18, ZFHD1-Ss18)34,35. CiA anchor (ZFHD1–FKBP12) and VP16 (Gal4BD-VP16) constructs were described previously33. Lentivirus was generated as described66. MEFs and ESCs were infected and selected with puromycin (2 μg/ml), blasticidin (10 μg/ml) and/or hygromycin (200 μg/ml). For experiments using chemical-induced proximity (CIP), rapamycin (Selleckchem) was added at 3 nM. Media containing rapamycin was changed daily for experiments spanning more than 24 h. For early time points (<20 min), rapamycin at 12 nM was added to media as well as to the fresh trypsin used to dissociate cells.

CRISPR–Cas9-guided recruitment

Catalytic-dead Cas9 (dCas9)-mediated recruitment was done as described36. Three small guide RNAs (sgRNAs) targeting a DNase hypersensitivity site 200 bp upstream of the PSMB11 transcription start site were selected on the basis of scores for off-target matches as described67. These sgRNAs were cloned into the sgRNA(MS2)_zeo lentiviral backbone (Addgene #61427) containing MS2-specific hairpin aptamers. Murine Aire with C-terminal MS2 (Aire-MS2-V5) was cloned into a previously described lentiviral backbone33. The 4D6 human thymic epithelial cell line was sequentially transduced with lentiviruses encoding the sgRNAs and dCas9 or the sgRNAs, dCas9 and Aire-MS2.

Chromatin immunoprecipitation

ChIPs were essentially performed as described34,35,47. Cells were dissociated and washed, and 30 million cells were formaldehyde cross-linked (1%) at 37 °C for 12 min. For experiments using CIP (<20 min), rapamycin at 3 nM was added to media quenching trypsin, PBS wash and fix buffer. Nuclei were sonicated with a Covaris E220 ultrasonicator at 5% duty cycle, 4 intensity, 140-W peak incident power and 200 cycles per burst for 13 min. Insoluble chromatin was pelleted, and supernatant was diluted 1:1 in 2 × ChIP buffer (100 mM HEPES, pH 7.5, 600 mM NaCl, 2 mM EDTA, pH 8.0, 2% Triton X-100, 0.2% sodium deoxycholate, 0.1% SDS) and divided into four ChIP reactions (4–5 μg of antibody per reaction; anti-V5, clone R960-25, Life Technologies), which were incubated overnight at 4 °C. Then 25 μl of Protein G Dynabeads (Thermo Fisher) slurry washed in ChIP buffer was added to ChIP reactions and rotated at 4 °C for 1 h. Beads were washed three times in ChIP buffer, once in DOC buffer (10 mM Tris, pH 8.0, 250 mM LiCl, 0.5% NP-40, 0.5% sodium deoxycholate, 1 mM EDTA) and once in TE buffer (10 mM Tris, pH 8.0, 1 mM EDTA) before consecutive elutions (two times) in 150 μl of 0.1 M NaHCO3, 1% SDS. Eluates were subjected to RNase A and proteinase K before reverse cross-linking at 65 °C overnight. ChIP DNA was precipitated by phenol–chloroform extraction and reconstituted in TE buffer for qPCR reactions. Amplicon detection for each target region was normalized to that at respective control regions for all samples. Enrichment was calculated as the fold-change between the normalized bound/input values of rapamycin-treated versus ethanol-treated samples. Primers used for ChIP studies at the CiA:Oct4 locus are summarized in Supplementary Table 133,34, V5 ChIP at the CiA:Oct4 locus was normalized to that at the housekeeping Rps29 promoter.

DNase accessibility assay

DNase I sensitivity assays were carried out as described68. Cells were lysed in buffer A (15 mM Tris, pH 8.0, 15 mM NaCl, 60 mM KCl, 1 mM EDTA, 0.5 mM EGTA, 0.5 mM spermidine) and washed in buffer A, and 5 million nuclei were pelleted for each DNase I reaction. For experiments using CIP (<20 min), rapamycin at 3 nM was added to media quenching trypsin, PBS wash and buffer A. Nuclei were subjected to varying concentrations of DNase I (Sigma) for 3 min at 37 °C. Digestions were terminated with Stop buffer and exposed to proteinase K for 1 h at 55 °C. Samples were treated with RNase A, and DNA was purified over MinElute columns (Qiagen) for qPCR reactions. Amplicon detection at the CiA:Oct4 locus (Supplementary Table 1) for all DNase I conditions was normalized to that at a DNase-I-insensitive region at the Rho locus (Supplementary Table 1). Amplicon detection at the promoter of the PSMB11 locus for all DNase I conditions was normalized to that at the DNase-I-insensitive region at the RHO locus (Supplementary Table 1).

Histopathology and immunohistochemistry

Histopathology experiments were carried out as described40. Tissues were fixed in buffered 10% formalin and paraffin-embedded. Hematoxylin and eosin stainings were done via standard methods. Immunohistochemistry was carried out on 4-μm sections with ABC Vectastain kits (Vector Laboratories) using anti-CD3 (DakoCytomation; A045229), and developed with DAB.

Thymus transplants

Kidney capsule thymus transplants were done as described69. Thymi from newborn Brg1-cKO and wild-type littermates were cultured in 1.35 mM 2-deoxyguanosine for 7 d to deplete hematopoietic compartments. Thymic stroma were washed and transplanted under the kidney capsules of 6–8-week-old female nude mice. Thymopoiesis was monitored via cytofluorimetric analysis of blood at 5 and 10 weeks post-transplantation. Animals were examined 13–15 weeks after transplantation for T cell reconstitution, and peripheral organs were collected for histopathology.


Fisher’s exact tests for over-representation in fold-change versus fold-change plots were one-sided tests with 95% confidence intervals for the lower bound of odds ratios (Figs. 2c and 5b). For cumulative distribution function plots, we used Mann–Whitney U-tests to identify significant differences between comparisons with 95% confidence intervals at medians of pairwise comparisons (Figs. 1c, 3b and 5a, and Supplementary Figs. 2d and 4). Correlations between different peak sets were assessed by Pearson correlation test (Figs. 1e and 2a). Student’s t-tests were done as two-tailed tests; n values are specified for independent experiments in individual figures (Figs. 5, 6 and 7 and Supplementary Fig. 6).

Life Sciences Reporting Summary

Further information on experimental design is available in the Life Sciences Reporting Summary.

Data availability

ATAC-seq and microarray data sets reported in this article, including raw reads and fully processed count matrices, can be accessed in GEO with accession codes GSE102526 and GSE102525, respectively. Publicly available ChIP-seq and ATAC-seq data sets referenced in this study are as follows: GSE92597 (ChIP-seq) for Aire, Pol II, Top1, Top2a, γ-H2AX, and H3K27ac in mTECs; GSE39513 (ChIP-seq) for Suz12 in ESCs; GSE42466 (ChIP-seq) for Ring1b in ESCs; GSE28247 (ChIP-seq) for Lamin B in ESCs; GSE27841 (ChIP-seq) for Lsd1, CoREST, and Hdac2 in ESCs; and GSE94041 (ATAC-seq) for ESCs. Any custom code will be made available by the corresponding author upon request.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    Klein, L., Kyewski, B., Allen, P. M. & Hogquist, K. A. Positive and negative selection of the T cell repertoire: what thymocytes see (and don’t see). Nat. Rev. Immunol. 14, 377–391 (2014).

  2. 2.

    Takahama, Y., Ohigashi, I., Baik, S. & Anderson, G. Generation of diversity in thymic epithelial cells. Nat. Rev. Immunol. 17, 295–305 (2017).

  3. 3.

    Derbinski, J., Schulte, A., Kyewski, B. & Klein, L. Promiscuous gene expression in medullary thymic epithelial cells mirrors the peripheral self. Nat. Immunol. 2, 1032–1039 (2001).

  4. 4.

    Bruserud, Ø., Oftedal, B. E., Wolff, A. B. & Husebye, E. S. AIRE-mutations and autoimmune disease. Curr. Opin. Immunol. 43, 8–15 (2016).

  5. 5.

    Anderson, M. S. et al. Projection of an immunological self shadow within the thymus by the AIRE protein. Science 298, 1395–1401 (2002).

  6. 6.

    Long, H. K., Prescott, S. L. & Wysocka, J. Ever-changing landscapes: transcriptional enhancers in development and evolution. Cell 167, 1170–1187 (2016).

  7. 7.

    Giraud, M. et al. Aire unleashes stalled RNA polymerase to induce ectopic gene expression in thymic epithelial cells. Proc. Natl. Acad. Sci. USA 109, 535–540 (2012).

  8. 8.

    Giraud, M. et al. An RNAi screen for Aire cofactors reveals a role for Hnrnpl in polymerase release and Aire-activated ectopic transcription. Proc. Natl. Acad. Sci. USA 111, 1491–1496 (2014).

  9. 9.

    Yoshida, H. et al. Brd4 bridges the transcriptional regulators, Aire and P-TEFb, to promote elongation of peripheral-tissue antigen transcripts in thymic stromal cells. Proc. Natl. Acad. Sci. USA 112, E4448–E4457 (2015).

  10. 10.

    Meredith, M., Zemmour, D., Mathis, D. & Benoist, C. Aire controls gene expression in the thymic epithelium with ordered stochasticity. Nat. Immunol. 16, 942–949 (2015).

  11. 11.

    Villaseñor, J., Besse, W., Benoist, C. & Mathis, D. Ectopic expression of peripheral-tissue antigens in the thymic epithelium: probabilistic, monoallelic, misinitiated. Proc. Natl. Acad. Sci. USA 105, 15854–15859 (2008).

  12. 12.

    Jolicoeur, C., Hanahan, D. & Smith, K. M. T-cell tolerance toward a transgenic beta-cell antigen and transcription of endogenous pancreatic genes in thymus. Proc. Natl. Acad. Sci. USA 91, 6707–6711 (1994).

  13. 13.

    Derbinski, J., Pinto, S., Rösch, S., Hexel, K. & Kyewski, B. Promiscuous gene expression patterns in single medullary thymic epithelial cells argue for a stochastic mechanism. Proc. Natl. Acad. Sci. USA 105, 657–662 (2008).

  14. 14.

    Pinto, S. et al. Overlapping gene coexpression patterns in human medullary thymic epithelial cells generate self-antigen diversity. Proc. Natl. Acad. Sci. USA 110, E3497–E3505 (2013).

  15. 15.

    Buenrostro, J. D., Giresi, P. G., Zaba, L. C., Chang, H. Y. & Greenleaf, W. J. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nat. Methods 10, 1213–1218 (2013).

  16. 16.

    Wagner, A., Regev, A. & Yosef, N. Revealing the vectors of cellular identity with single-cell genomics. Nat. Biotechnol. 34, 1145–1160 (2016).

  17. 17.

    Sansom, S. N. et al. Population and single-cell genomics reveal the Aire dependency, relief from Polycomb silencing, and distribution of self-antigen expression in thymic epithelia. Genome Res. 24, 1918–1931 (2014).

  18. 18.

    Bansal, K., Yoshida, H., Benoist, C. & Mathis, D. The transcriptional regulator Aire binds to and activates super-enhancers. Nat. Immunol. 18, 263–273 (2017).

  19. 19.

    Brennecke, P. et al. Single-cell transcriptome analysis reveals coordinated ectopic gene-expression patterns in medullary thymic epithelial cells. Nat. Immunol. 16, 933–941 (2015).

  20. 20.

    Wang, W. et al. Architectural DNA binding by a high-mobility-group/kinesin-like subunit in mammalian SWI/SNF-related complexes. Proc. Natl. Acad. Sci. USA 95, 492–498 (1998).

  21. 21.

    Bottomley, M. J. et al. The SAND domain structure defines a novel DNA-binding fold in transcriptional regulation. Nat. Struct. Biol. 8, 626–633 (2001).

  22. 22.

    Koh, A. S. et al. Aire employs a histone-binding module to mediate immunological tolerance, linking chromatin regulation with organ-specific autoimmunity. Proc. Natl. Acad. Sci. USA 105, 15878–15883 (2008).

  23. 23.

    Buenrostro, J. D. et al. Single-cell chromatin accessibility reveals principles of regulatory variation. Nature 523, 486–490 (2015).

  24. 24.

    Manley, N. R. & Condie, B. G. Transcriptional regulation of thymus organogenesis and thymic epithelial cell differentiation. Prog. Mol. Biol. Transl. Sci. 92, 103–120 (2010).

  25. 25.

    Rossi, S. W. et al. RANK signals from CD4+3 inducer cells regulate development of Aire-expressing epithelial cells in the thymic medulla. J. Exp. Med. 204, 1267–1272 (2007).

  26. 26.

    Senoo, M., Pinto, F., Crum, C. P. & McKeon, F. p63 is essential for the proliferative potential of stem cells in stratified epithelia. Cell 129, 523–536 (2007).

  27. 27.

    Satoh, R. et al. Requirement of Stat3 signaling in the postnatal development of thymic medullary epithelial cells. PLoS Genet. 12, e1005776 (2016).

  28. 28.

    Lomada, D. et al. Stat3 signaling promotes survival and maintenance of medullary thymic epithelial cells. PLoS Genet. 12, e1005777 (2016).

  29. 29.

    Natoli, G. Control of NF-κB-dependent transcriptional responses by chromatin organization. Cold Spring Harb. Perspect. Biol. 1, a000224 (2009).

  30. 30.

    Lin, J. et al. Increased generation of Foxp3+ regulatory T cells by manipulating antigen presentation in the thymus. Nat. Commun. 7, 10562 (2016).

  31. 31.

    Gillard, G. O., Dooley, J., Erickson, M., Peltonen, L. & Farr, A. G. Aire-dependent alterations in medullary thymic epithelium indicate a role for Aire in thymic epithelial differentiation. J. Immunol. 178, 3007–3015 (2007).

  32. 32.

    Asano, M., Toda, M., Sakaguchi, N. & Sakaguchi, S. Autoimmune disease as a consequence of developmental abnormality of a T cell subpopulation. J. Exp. Med. 184, 387–396 (1996).

  33. 33.

    Hathaway, N. A. et al. Dynamics and memory of heterochromatin in living cells. Cell 149, 1447–1460 (2012).

  34. 34.

    Kadoch, C. et al. Dynamics of BAF-Polycomb complex opposition on heterochromatin in normal and oncogenic states. Nat. Genet. 49, 213–222 (2017).

  35. 35.

    Miller, E. L. et al. TOP2 synergizes with BAF chromatin remodeling for both resolution and formation of facultative heterochromatin. Nat. Struct. Mol. Biol. 24, 344–352 (2017).

  36. 36.

    Konermann, S. et al. Genome-scale transcriptional activation by an engineered CRISPR-Cas9 complex. Nature 517, 583–588 (2015).

  37. 37.

    Fernández, E. et al. Establishment and characterization of cloned human thymic epithelial cell lines. Analysis of adhesion molecule expression and cytokine production. Blood 83, 3245–3254 (1994).

  38. 38.

    Murata, S. et al. Regulation of CD8+ T cell development by thymus-specific proteasomes. Science 316, 1349–1353 (2007).

  39. 39.

    Org, T. et al. The autoimmune regulator PHD finger binds to non-methylated histone H3K4 to activate gene expression. EMBO Rep. 9, 370–376 (2008).

  40. 40.

    Koh, A. S., Kingston, R. E., Benoist, C. & Mathis, D. Global relevance of Aire binding to hypomethylated lysine-4 of histone-3. Proc. Natl. Acad. Sci. USA 107, 13016–13021 (2010).

  41. 41.

    Chakravarty, S., Zeng, L. & Zhou, M. M. Structure and site-specific recognition of histone H3 by the PHD finger of human autoimmune regulator. Structure 17, 670–679 (2009).

  42. 42.

    Su, M. A. et al. Mechanisms of an autoimmunity syndrome in mice caused by a dominant mutation in Aire. J. Clin. Invest. 118, 1712–1726 (2008).

  43. 43.

    Pitkänen, J. et al. The autoimmune regulator protein has transcriptional transactivating properties and interacts with the common coactivator CREB-binding protein. J. Biol. Chem. 275, 16802–16809 (2000).

  44. 44.

    Iglesias, P. & Díez, J. J. Management of endocrine disease: a clinical update on tumor-induced hypoglycemia. Eur. J. Endocrinol. 170, R147–R157 (2014).

  45. 45.

    Schneider, R., Heverhagen, A. E., Moll, R., Bartsch, D. K. & Schlosser, K. Differentiation between thyroidal and ectopic calcitonin secretion in patients with coincidental thyroid nodules and pancreatic tumors—a report of two cases. Exp. Clin. Endocrinol. Diabetes 118, 520–523 (2010).

  46. 46.

    Falanga, A., Schieppati, F. & Russo, D. Cancer tissue procoagulant mechanisms and the hypercoagulable state of patients with cancer. Semin. Thromb. Hemost. 41, 756–764 (2015).

  47. 47.

    Stanton, B. Z. et al. Smarca4 ATPase mutations disrupt direct eviction of PRC1 from chromatin. Nat. Genet. 49, 282–288 (2017).

  48. 48.

    Dykhuizen, E. C. et al. BAF complexes facilitate decatenation of DNA by topoisomerase IIα. Nature 497, 624–627 (2013).

  49. 49.

    Trotter, K. W., King, H. A. & Archer, T. K. Glucocorticoid receptor transcriptional activation via the BRG1-dependent recruitment of TOP2β and Ku70/86. Mol. Cell. Biol. 35, 2799–2817 (2015).

  50. 50.

    Abramson, J., Giraud, M., Benoist, C. & Mathis, D. Aire’s partners in the molecular control of immunological tolerance. Cell 140, 123–135 (2010).

  51. 51.

    Chi, T. H. et al. Sequential roles of Brg, the ATPase subunit of BAF chromatin remodeling complexes, in thymocyte development. Immunity 19, 169–182 (2003).

  52. 52.

    Gordon, J. et al. Specific expression of lacZ and cre recombinase in fetal thymic epithelial cells by multiplex gene targeting at the Foxn1 locus. BMC Dev. Biol. 7, 69 (2007).

  53. 53.

    Zhao, Z. et al. p53 loss promotes acute myeloid leukemia by enabling aberrant self-renewal. Genes Dev. 24, 1389–1402 (2010).

  54. 54.

    Gardner, J. M. et al. Deletional tolerance mediated by extrathymic Aire-expressing cells. Science 321, 843–847 (2008).

  55. 55.

    Jain, R. & Gray, D. H. Isolation of thymic epithelial cells and analysis by flow cytometry. Curr. Protoc. Immunol. 107, 3.26.1–3.26.15 (2014).

  56. 56.

    Cipolletta, D. et al. PPAR-γ is a major driver of the accumulation and phenotype of adipose tissue Treg cells. Nature 486, 549–553 (2012).

  57. 57.

    Su, A. I. et al. A gene atlas of the mouse and human protein-encoding transcriptomes. Proc. Natl. Acad. Sci. USA 101, 6062–6067 (2004).

  58. 58.

    Lattin, J. E. et al. Expression analysis of G protein-coupled receptors in mouse macrophages. Immunome Res. 4, 5 (2008).

  59. 59.

    Buenrostro, J. D., Wu, B., Chang, H. Y. & Greenleaf, W. J. ATAC-seq: a method for assaying chromatin accessibility genome-wide. Curr. Protoc. Mol. Biol. 109, 21.29.1–21.29.9 (2015).

  60. 60.

    Corces, M. R. et al. Lineage-specific and single-cell chromatin accessibility charts human hematopoiesis and leukemia evolution. Nat. Genet. 48, 1193–1203 (2016).

  61. 61.

    Denny, S. K. et al. Nfib promotes metastasis through a widespread increase in chromatin accessibility. Cell 166, 328–342 (2016).

  62. 62.

    Whyte, W. A. et al. Enhancer decommissioning by LSD1 during embryonic stem cell differentiation. Nature 482, 221–225 (2012).

  63. 63.

    Jia, J. et al. Regulation of pluripotency and self- renewal of ESCs through epigenetic-threshold modulation and mRNA pruning. Cell 151, 576–589 (2012).

  64. 64.

    Morey, L., Aloia, L., Cozzuto, L., Benitah, S. A. & Di Croce, L. RYBP and Cbx7 define specific biological functions of Polycomb complexes in mouse embryonic stem cells. Cell Rep. 3, 60–69 (2013).

  65. 65.

    Handoko, L. et al. CTCF-mediated functional chromatin interactome in pluripotent cells. Nat. Genet. 43, 630–638 (2011).

  66. 66.

    Tiscornia, G., Singer, O. & Verma, I. M. Production and purification of lentiviral vectors. Nat. Protoc. 1, 241–245 (2006).

  67. 67.

    Hsu, P. D. et al. DNA targeting specificity of RNA-guided Cas9 nucleases. Nat. Biotechnol. 31, 827–832 (2013).

  68. 68.

    John, S. et al. Genome-scale mapping of DNase I hypersensitivity. Curr. Protoc. Mol. Biol. 103, 21.27.1–21.27.20 (2013).

  69. 69.

    Anderson, M. S. et al. The cellular mechanism of Aire control of T cell tolerance. Immunity 23, 227–239 (2005).

Download references


We are grateful to N. Manley (University of Georgia, Athens, GA, USA) for Foxn1ex9Cre mice; N. Hathaway (University of North Carolina, Chapel Hill, NC, USA), C. Kadoch (Harvard Medical School, Boston, MA, USA), S. Braun (Stanford University, Stanford, CA, USA) and E. Chory (Stanford University, Stanford, CA, USA) for CiA constructs; M. Toribio (Universidad Autónoma de Madrid, Madrid, Spain) and D. Mathis (Harvard Medical School, Boston, MA, USA) for the 4D6 cTEC line; D. Mathis, M. Anderson and S. Denny for insightful comments; Y. Chien, C. Weber, J. Kirkland, L. Wagar and J. Ronan for critical reading of the manuscript; and J. Gardner, P. Chu and R. Li for technical assistance. We thank the Stanford Shared FACS facility and S. Kim for flow cytometry and cell sorting. The Stanford BioX3 cluster was used for computational analyses (NIH S10 grant 1S10RR02664701). This work was supported by the Howard Hughes Medical Institute (to G.R.C.), the NIH (grants CA163915 and NS046789 to G.R.C., P50-HG007735 to H.Y.C. and W.J.G., T32HG000044 to J.D.B., and T32 GM007790 to E.L.M.), the Lymphoma and Leukemia Society (A.S.K.) and the National Library of Medicine (Stanford Biomedical Informatics Training Grant LM-07033 to D.M.M.).

Author information


  1. Department of Pathology, Stanford University School of Medicine, Stanford, CA, USA

    • Andrew S. Koh
    • , Erik L. Miller
    •  & Gerald R. Crabtree
  2. Center for Personal Dynamic Regulomes, Stanford University School of Medicine, Stanford, CA, USA

    • Andrew S. Koh
    • , David M. Moskowitz
    • , William J. Greenleaf
    • , Howard Y. Chang
    •  & Gerald R. Crabtree
  3. Howard Hughes Medical Institute, Chevy Chase, MD, USA

    • Andrew S. Koh
    • , Erik L. Miller
    •  & Gerald R. Crabtree
  4. Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA

    • Erik L. Miller
    • , David M. Moskowitz
    •  & William J. Greenleaf
  5. Broad Institute of MIT and Harvard, Cambridge, MA, USA

    • Jason D. Buenrostro
  6. Harvard Society of Fellows, Harvard University, Cambridge, MA, USA

    • Jason D. Buenrostro
  7. Department of Developmental Biology, Stanford University School of Medicine, Stanford, CA, USA

    • Jing Wang
  8. Department of Applied Physics, Stanford University, Stanford, CA, USA

    • William J. Greenleaf
  9. Chan Zuckerburg Biohub, San Francisco, CA, USA

    • William J. Greenleaf


  1. Search for Andrew S. Koh in:

  2. Search for Erik L. Miller in:

  3. Search for Jason D. Buenrostro in:

  4. Search for David M. Moskowitz in:

  5. Search for Jing Wang in:

  6. Search for William J. Greenleaf in:

  7. Search for Howard Y. Chang in:

  8. Search for Gerald R. Crabtree in:


A.S.K. and G.R.C. conceived of the study and wrote the paper. A.S.K. planned and performed all experiments and data analysis. A.S.K., E.L.M., J.D.B. and D.M.M. performed ATAC-seq data analysis. J.W. performed kidney capsule transplants. W.J.G. and H.Y.C. provided conceptual insights and advised on data analysis and experimental design.

Competing interests

W.J.G. and H.Y.C. are cofounders of Epinomics, Inc. Stanford University has filed a patent on ATAC-seq on which W.J.G. and H.Y.C. are named as inventors.

Corresponding authors

Correspondence to Andrew S. Koh or Gerald R. Crabtree.

Integrated supplementary information

  1. Supplementary Figure 1 mTEChi differentiation promotes major shifts in the chromatin accessibility landscape.

    (a) Flow cytometry gating sequence for sorting mTECs. (b) Representative histogram of aggregate Tn5 insertions (blue) and smoothed signal (red) around transcriptional start sites (TSS). (c) Representative distribution of ATAC-seq fragment size exhibiting nucleosomal periodicity. (d) Representative comparison of ATAC-seq fragment density at peaks between biological replicates. (e) Signal enrichment for ATAC-seq libraries, defined by the fold-change between the maximum and minimum signals within the 4 kb region displayed in (b). (f) First principal component (PC) of PCA representing 30.36% of variance separates mTEChi and mTEClo samples. (g) Genomic signal tracks of ATAC-seq fragments at three loci from mTEChi and mTEClo samples. Red arrowheads indicate differentially accessible regions. (h) Representative density plots of ChIP-seq read dyads of indicated factors in embryonic stem cells at ATAC-seq peaks near tissue-specific genes whose accessibilities are induced during mTEChi differentiation. The data shown are from 1 experiment representative of > 20 (a), 16 (b,c), or 4 (d) independent experiments or are from n = 4 (f,g) independent experiments or from pooled data representative of 2 (h) independent experiments.

  2. Supplementary Figure 2 Aire and Brg1 have opposing influences on accessibility.

    (a) Representative immunofluorescence stainings of 4-week old thymic sections from indicated genotypes for medullary marker keratin-14 (red) and Aire (green). White scale bar = 10 um. (b) Representative frequencies of TECs in 4-week old thymi from indicated genotypes assessed by flow cytometry. (c) Representative frequencies of mTEChi and mTEClo compartments expressing Aire from 4-week old thymi of indicated mice. (d) CDF plots of indicated accessibility fold-changes at regions classified as differentially accessible in WT vs. Brg1cKO or AireKO mTEChi. P values were determined by Mann-Whitney U-test (two-tailed). (e) Distribution of distances of indicated peak sets to the nearest TSS. (f) Histogram of the distances between indicated peak sets (Fig. 2c) and the nearest TSS. (g) Genomic signal tracks of ATAC-seq fragments at six loci from indicated mTEChi samples (top). ChIP-seq signal tracks from WT mTEChi samples. Red arrowheads indicate differentially accessible regions. (h-j) ATAC-seq analyses on samples generated by Bansal et. al. 2016. (h) First principal component (PC) of PCA separating WT and Aire−/− mTEChi samples. (i) Heatmap of normalized ATAC-seq fragment density at differential peaks (rows). (j) Heatmap of Aire ChIP-seq fragment dyad density at Aire-repressed ATAC-seq peaks. The data shown are from 1 experiment representative of 3 (a), 10 (b,c) or 2 (g) independent experiments or from pooled data representative of 4 (c) or 2 (j) independent experiments or from n = 2 (h,i) independent experiments.

  3. Supplementary Figure 3 Regions in which accessibility is repressed by Aire and induced by Brg1 are enriched for H3K27ac and active topoisomerases.

    (a) ChIP-seq fragment dyad density of indicated factors/histones at Aire-induced or Aire-repressed ATAC-seq peaks. (b) Heatmap of ChIP-seq fragment dyad density of indicated factors/histones at Aire-repressed ATAC-seq peaks. (c) ChIP-seq fragment dyad density at Brg1-induced or Brg1-repressed ATAC-seq peaks.  The data shown are from pooled data representative of 2 independent experiments (a-c).

  4. Supplementary Figure 4 Aire is dispensable for induced accessibility at tissue-specific loci during mTEC differentiation.

    (a) CDF plot of accessibility fold-changes between AireKO mTEChi and AireKO mTEClo at indicated ATAC-seq peaks differentially accessible or unchanged in WT mTEChi vs. mTEClo. (b) CDF plot of transcriptional fold-changes between AireKO mTEChi and AireKO mTEClo at indicated genes upregulated or unchanged by Aire. (c) CDF plot of accessibility fold-changes between WT mTEChi and AireKO mTEChi at indicated ATAC-seq peaks differentially accessible or unchanged in WT mTEChi vs mTEClo. (d) CDF plot of transcriptional fold-changes between WT mTEChi and AireKO mTEChi at indicated genes upregulated or unchanged by Aire. The data shown are from pooled data representative of 2 independent experiments (a-d). P values determined from Mann-Whitney U-tests (two-tailed) (a-d).

  5. Supplementary Figure 5 Differential enrichment of transcription factor motifs and footprints during mTEC maturation.

    (a-c) Change in deviation from expected accessibility signal at ATAC-seq peaks containing known trans-factor motifs (key) between indicated samples. Diamonds represent means, circles represent replicates. (d) NF-kB motifs depicted from analyses in (a-c) and the changes in deviation scores from mTEChi samples of indicated genotypes. Mean +/− s.e.m. (e-p) Differential accessibility footprints in mTEChi vs. mTEClo samples at ATAC-seq peaks containing indicated motifs.  The data shown are n = 2 (a-c, d: Aire-KO or Brg1-cKO samples) or n = 4 (d: WT samples) independent experiments, or from pooled data representative of 4 (e-p) independent experiments. 

  6. Supplementary Figure 6 T cell compartments in Brg1-cKO mice.

    (a,b) Frequencies of thymocyte compartments in 4-week old mice from indicated genotypes assessed by flow cytometry Mean +/− s.e.m. n.s., not significant (two-tailed t-tests). (c) Frequency (left) and cellularity (right) of cTEC compartment in 4-week old mice from indicated genotypes. Mean +/− s.e.m. P values determined by two-tailed t-test < 0.001 (***). (d) Cellularity of thymocyte compartments in 4-week old mice from indicated genotypes. Mean +/− s.e.m. P values determined by two-tailed t-test < 0.05 (*), < 0.01 (**). (e) Cellularity of indicated splenic T cell compartments in 4-week old mice from indicated genotypes. Mean +/− s.e.m. P values determined by two-tailed t-test < 0.01 (**). (f) Frequency (left) and celluarity (right) of FoxP3+CD25+ regulatory T cells (Treg) in spleen from 4 week-old mice of indicated genotypes. Mean +/− s.e.m. P values determined by two-tailed t-test < 0.05 (*). (g) Purified CD4+CD25neg Tconv cells were mixed with CD4+CD25+ Treg from WT and Brg1-cKO mice in a criss-cross fashion as indicated, and assayed for proliferation in the presence of irradiated splenocytes and anti-CD3. Mean +/− s.e.m. (h) Histological analyses of indicated tissues from 6 month-old WT or Brg1-cKO mice via H&E and anti-CD3 immunohistochemistry stainings for infiltrating lymphocytes at indicated peripheral tissues. Scale bars for 10x, 60x images = 200 um, 50 um, respectively. The data shown are from 1 experiment representative of 8 (a) or 2 (h) independent experiments or from n = 8 (bf) or 3 (g) independent experiments.

  7. Supplementary Figure 7 Aire and BAF have divergent influences on accessibility after recruitment to chromatin.

    (a) The CiA:Oct4 locus exhibits DNase I sensitivity compared to the inaccessible rhodopsin (Rho) locus. Mean +/− s.e.m. (b) Rapid maximal reduction in DNase I sensitivity after recruitment of Aire to CiA:Oct4 locus via rapamycin. (c) Schematic of CiA recruitment system. (d) Changes in DNase I sensitivity at CiA:Oct4 locus post BAF vs. Aire recruitment. Mean +/− s.e.m. (e) Recruitment of BAF nor Aire to CiA:Oct4 locus activates transcription. (f) The PSMB11 locus exhibits DNase I sensitivity . Mean +/− s.e.m. (g) Schematic of dCas9 recruitment system: Aire fused to MS2 viral domain is targeted by guide RNA with MS2-binding aptamers. (h) Changes in DNase I sensitivity upon dCas9-induced Aire recruitment to PSMB11 locus, compared to changes at locus encoding ribosomal subunit (RPS29). Mean +/− s.e.m. (i) Model of rheostatic chromatin control of ectopic transcription of tissue-specific genes by Aire and Brg1: BAF chromatin remodeling complexes (orange) coordinate with transcription factors (purple) to poise and recruit transcriptional machinery (gold and gray) to tissue-specific loci during mTEC differentiation. Productive elongation of Pol II is inhibited by negative elongation factors, e.g. Nelf. Subsequent Aire expression and targeting brings interacting positive elongation factors (light blue) to release paused Pol II. Aire’s repressive function inhibits chromatin accessibility, reducing the occupancy of BAF, transcription factors and transcriptional machinery (indicated by fading opacity) and restraining amplitude of transcription. The data shown are from n = 6 (a), 1 (b), or 3 (d,f,h) independent experiments or from 1 experiment representative of 4 (e) independent experiments. 

Supplementary information

  1. Supplementary Text, Figures and Table

    Supplementary Figures 1–7 and Supplementary Table 1

  2. Life Sciences Reporting Summary

About this article

Publication history