Autoantigen-specific CD4+ T cells acquire an exhausted phenotype and persist in human antigen-specific autoimmune diseases

Pro-inflammatory autoantigen-specific CD4+ T helper (auto-Th) cells are central orchestrators of autoimmune diseases (AIDs). We aimed to characterize these cells in human AIDs with defined autoantigens by combining human leukocyte antigen (HLA)-tetramer-based and activation-based multidimensional ex vivo analyses. In aquaporin4-antibody-positive neuromyelitis optica spectrum disorder (AQP4-NMOSD) patients, auto-Th cells expressed CD154, but proliferative capacity and pro-inflammatory cytokines were strongly reduced. Instead, exhaustion-associated co-inhibitory receptors were expressed together with FOXP3, the canonical regulatory T cell (Treg) transcription factor. Auto-Th cells responded in vitro to checkpoint inhibition and provided potent B cell help. Cells with the same exhaustion-like (ThEx) phenotype were identified in soluble liver antigen (SLA)-antibody-autoimmune hepatitis and BP180-antibody-positive bullous pemphigoid, AIDs of the liver and skin, respectively. While originally described in cancer and chronic infection, our data point to T cell exhaustion as a common mechanism of adaptation to chronic (self-)stimulation across AID types and link exhausted CD4+ T cells to humoral autoimmune responses, with implications for therapeutic targeting. For single-cell RNA-seq, PBMCs were stimulated for 6 h (with AQP4, MP65, and Influenza HA peptide pools or PMA/Ionomycin), CD154+ memory cells were isolated by MACS and further purified by FACS sorting on a FACS Aria Fusion (BD Bioscience, San Jose, CA, USA) based on dual expression of CD154 and CD69 and expression of CD45RO. Cells were directly sorted into skirted 96-well PCR plates which were equipped with 2 l lysis buffer (nuclease-free water with 0.2% Triton-X 100 solution (Roche Diagnostics GmbH, Mannheim, Germany) with 4 U/µL RNase Inhibitor (New England Biolabs)) in advance. After FACS sorting plates were shortly centrifuged (30 sec, 200xg) and immediately frozen at -80°C until shipment. Library preparation and sequencing was performed at the DRESDEN-concept Genome Center (DcGC, TU Dresden, Germany). The workflow was based on the previously described SMARTseq2 protocol 81. After thawing the samples, 2 ul of a primer mix was added (5 mM dNTP (Invitrogen), 0.5 uM dT-primer*, 4 U RNase Inhibitor (NEB)). RNA was denatured for 3 minutes at 72°C and the reverse transcription was performed at 42°C for 90 min after filling up to 10 ul with RT buffer mix for a final concentration of 1X superscript II buffer (Invitrogen), 1 M betaine, 5 mM DTT, 6 mM MgCl2, 1 uM TSO-primer*, 9 U RNase Inhibitor and 90 U Superscript II. After synthesis, the reverse transcriptase was inactivated at 70°C for 15 min. The cDNA was amplified using 2X Kapa HiFi HotStart Readymix (Roche) at a final 1X concentration and 0.1 uM UP-primer* under following cycling conditions: initial denaturation at 98°C for 3 min, 22 cycles [98°C 20 sec, 67°C 15 sec, 72°C 6 min] and final elongation at 72°C for 5 min. The amplified cDNA is purified using 0.6X volume of hydrophobic Sera-Mag Beads (Cytvia: GE24152105050250,) and DNA was eluted in 12 ul nuclease free water. The cDNA quality and concentration of a few representative samples was determined with the Fragment Analyzer (Agilent). For library preparation, 2 µl amplified cDNA was tagmented in 1X Tagmentation Buffer using 0.8 µl bead-linked transposome (Illumina DNA Prep, (M) Tagmentation, Illumina) at 55°C for 15 min in a total volume of 4 µL. The reaction was stopped by adding 1 µl of 0.1 % SDS (37°C, 15 min). Magnetic beads were bound to a magnet, the supernatant was removed, beads were resuspended in 4 µl indexing PCR Mix containing 1x KAPA Hifi HotStart Ready Mix (Roche). After adding 700 nM unique dual indexing primers (i5 and i7) using Echo Acoustic Liquid Handler (Echo 525, Beckman Coulter), the reaction mixture was subjected to a PCR (72°C 3 min, 98°C 30 sec, 12 cycles [98°C 10 sec, 63°C 20 sec, 72°C 1 min], 72°C 5 min). After PCR, libraries are quantified with a Tecan plate reader Infinite 200 pro in 384 well black flat bottom low volume plates (Corning) using AccuBlue Broad range chemistry (Biotium) and equimolarly pooled using Echo Acoustic Liquid Handler (Echo 525, Beckman Coulter GmbH). The pooled libraries were then purified with 0.9x volume Sera-Mag SpeedBeads, followed by a double size selection with 0.6x and 0.9x volume of beads. Re-double size selection with 0.6x and 0.9x volume of beads was performed if needed. Sequencing was performed after quantification using a Fragment Analyzer on an Illumina Novaseq 6000 aiming at an average sequencing depth of 0.5 mio 100 bp paired-end fragments per cell. The method used for this single cell RNA seq is SMARTseq2 which is the full length transcriptome sequencing, thus the TCR sequences come from the same library as the expression data. However, always in some random cells, the full length of TCR alpha/beta or both are not sequenced, so it is not possible to recognize the TCR sequence, but still the expression data is available for these cells. This is the reason why fewer cells in the TCR file compare to the expression matrix (i.e. 928 samples vs. 924 columns in RawExpressionCountTable.csv, 829 data rows in TCR_data.csv)