The immune profile of circulating autoreactive CD4 T cells is imprinted through tissue activation during autoimmune liver diseases (Exp 1)

Autoimmune liver diseases (AILD) are immune-mediated disorders in which CD4 T cells play a central role. However, the link between circulating self-antigen-specific CD4 T cells and the targeted tissue has not been extensively studied in AILD. We hypothesized that circulating autoreactive CD4 T cells were clonally and functionally related to dominant intra-hepatic pathogenic CD4 T cell clones. Single cell transcriptomic analysis of circulating self-antigen-specific CD4 T cells revealed a specific B-helper and immuno-exhausted transcriptional profile, which was conserved for different autoantigens, but distinct from several other types of foreign antigen specificities. In the blood, the dominant hepatic CD4 T cell clones had a similar transcriptomic signature and were enriched in the PD-1+ TIGIT+ HLA-DR+ CD4 T cell subset. In a mouse model, antigen-specific CD4 T cells acquired the immuno-exhausted transcriptional profile when they accumulated in the liver after local antigen reactivity. Locally, immune checkpoint molecules controlled the response of antigen-specific CD4 T cells responsible for liver damage. Our study reveals the origin and biology of liver-derived autoreactive CD4 T cells in the blood of AILD patients that are imprinted by the liver environment, and suggest a dysregulation of the immune checkpoint molecules pathways. Our study enables tracking and isolating circulating autoreactive CD4 T cells for future diagnostic and therapeutic purposes. Overall design: For scRNA-seq analysis of peptide-restimulated CD4 T cells, CD154+ memory CD4 T cells were first sorted on BD FACSAriaII, one cell per well, in 96-well plates containing specific lysis buffer at the CRTI, Nantes. Plates were immediately frozen for storage at -80°C, and sent on dry ice to the Genomics core facility of CIML, Marseille, for further generating scRNAseq libraries with the FB5P-seq protocol as described22,58. Briefly, mRNA reverse transcription (RT), cDNA 5'-end barcoding and PCR amplification were performed with a template switching (TS) approach. After amplification, barcoded full-length cDNA from each well were pooled for purification and library preparation. For each plate, an Illumina sequencing library targeting the 5'-end of barcoded cDNA was prepared by a modified transposase-based method incorporating a plate-associated i7 barcode. Resulting libraries had a broad size distribution, resulting in gene template reads covering the 5'-end of transcripts from the 3rd to the 60th percentile of gene body length on average. As a consequence, sequencing reads covered the whole variable and a significant portion of the constant region of the TCRa and TCRß expressed mRNAs, enabling assembly and reconstitution of TCR repertoire from scRNAseq data. Libraries prepared with the FB5P-seq protocol were sequenced on Illumina NextSeq2000 platform with P2 100-cycle kits, targeting 5×105-1×106 reads per cell in paired-end single-index mode with the following configuration: Read1 (gene template) 103 cycles, Read i7 (plate barcode) 8 cycles, Read2 (cell barcode and Unique Molecular Identifier) 16 cycles. We then used a custom bioinformatics pipeline to process fastq files and generate single-cell gene expression matrices and TCR sequence files, as described22,58.