Thymic epithelia amplify noise in chromatin accessibility via p53 repression to impose immune tolerance

Phenotypic plasticity of somatic cells is a principal feature of vertebrate adaptation as well as a hallmark of tumorigenesis. However, the determinants and mechanisms that regulate the stability of somatic cell identities remain unclear. Here, using the somatic plasticity of thymic epithelial cells – which facilitates the selection of a self-discriminating T cell repertoire – as a physiological model system, we show that stochastic fluctuations in background chromatin accessibility at nucleosome-dense regions are amplified by thymic epithelial cells to ectopically express thousands of genes highly restricted to other specialized cell types. We found broad regions of inaccessible chromatin flanking tissue-specific genes become 'destabilized' during thymic epithelial maturation independently of AIRE-induced transcription of these genes, but concurrently with the repression of the tumor suppressor p53. Augmenting p53 activity in thymic epithelial cells reduced noise in chromatin accessibility at nucleosome-dense regions, inhibited ectopic expression of tissue-specific genes and caused multi-organ autoimmunity. Furthermore, we found p53-regulated fluctuations in background chromatin accessibility in lung adenocarcinoma to be associated with high plasticity states that promote tumor progression. Taken together, our findings establish p53-dependent stabilization of nucleosomal barriers to cellular reprogramming as a fulcrum of cell fate integrity that underlies critical components of immune tolerance induction and oncogenic potential. Overall design: Mouse: Thymic epithelial cells were isolated using a previously published protocol (Kim & Serwold, Methods in Molecular Biology, 2019) with minor modifications. Briefly, thymi from 4-5 week-old mice were harvested and connective tissue was removed. Stromal tissue was perforated using scissors and incubated with rotation in DMEM-F12 (Sigma D6421) at room temperature for for 10min to liberate thymocytes. Remaining stromal tissue was enzymatically digested (0.5mg/mL Collagenase D (Sigma 11088858001), 0.2mg/mL DNaseI (Sigma 10104159001), 0.5mg/mL Papain (Worthington Biochemical LS003126)). Cells were stained with anti-EpCAM antibodies conjugated to APC-Cy7 and EpCAM+ cells were enriched via positive selection using magnetic anti-Cy7 beads (Miltenyi 130-091-652). Enriched fraction was stained with fluorescently conjugated antibodies for FACS or analysis. Human: Excised thymic tissue from one patient (3-month-old male) was obtained during the course of corrective cardiac surgery at the Advocate Children's Hospital, Chicago, IL, USA. Patient was de-identified upon receipt with written informed consent for release of genomic sequence data in accordance with IRB approved protocols by the Advocate Children's Hospital, University of Chicago, and the National Institute of Health. Connective tissue was removed, and stromal tissue was minced using scissors. Tissue was incubated with rotation in DMEM-F12 (Sigma) at 4C for 20min to liberate thymocytes. Further thymocyte liberation was achieved via enzymatic digestion using 0.5mg/mL Collagenase D (Millipore 11088858001) and 0.2mg/mL DNase I (Sigma 10104159001) at 37C for 20min. Remaining fragments were incubated with rotation in 0.5mg/mL Papain (Worthington), 0.25mg/mL Collagenase D and 0.1mg/mL DNase I at 37C for 20min. Digestion was quenched by addition of FACS buffer. Cells were stained with anti-EpCAM antibodies conjugated to APC-Cy7 (BioLegend 324245 clone 9C4 1:100) and EpCAM+ cells were enriched via positive selection with magnetic anti-Cy7 beads (Miltenyi). Enriched fraction was stained with DAPI (Invitrogen), CD45 (BioLegend 368508 clone 2D1 1:100), LY51/CD249 (Fisher Scientific BDB564533 clone 2D3/APA 1:100) and HLA-DRA (BioLegend 307622 clone L243 1:100) and sorted on a Symphony S6 (BD Biosciences). For all multiome experiments, we used the ATAC+GEX single-cell kit and protocol (10X Genomics 1000236 with protocol CG000338 RevE) with minor modifications to sample preparation. Briefly, 40,000 mTECs were FACS-sorted into 1X PBS supplemented with 2% BSA and centrifuged at 300rcf for 5 min. Cells were gently washed in 50uL Lysis Buffer (10mM Tris, 10mM NaCl, 3mM MgCl2 in nuclease-free water) and centrifuged at 300rcf for 5 min. Cells were resuspended in 50uL permeabilization buffer (10mM Tris, 10mM NaCl, 3mM MgCl2, 0.1% Tween20, 0.01% digitonin, and RNase inhibitor (Millipore) in nuclease-free water) and incubated for 5 minutes on ice. Nuclei were gently washed with wash buffer (10mM Tris, 10mM NaCl, 3mM MgCl2, 0.1% Tween20, and RNase inhibitor in nuclease-free water) and centrifuged at 500rcf for 5 min. Finally, nuclei wee resuspended in 5uL chilled diluted nuclei buffer (10X Genomics) and added to transposition mix. Paired-end, dual-index sequencing was performed on the Illumina NextSeq 500 or NovaSeq 6000 platform.