121 B-cells during preclinical phase of rheumatoid arthritis blood, stained, sorted as plasmablast/plasma cells or memory B cells and directly lysed to capture RNA. In the same patients, these analyses were done at different time points (every 6 months) to study potential phenotypic changes in B cell subsets as well as the stability of dominant B cell signatures over time. These analyses were done based on the sorting and sequencing strategy shown in supplementary figure 2B. The first comparison answers the question whether freshly isolated unsorted PBMCs show a similar clonal pattern as blood collected in PAXgene tubes (and thus immediately lysed). In the second comparison the question is answered to what extent the clonal pattern of plasmablast/plasma cell subset overlaps with the memory B cell subset. The last comparison shows to what extent the PAXgene repertoire is based on the repertoire within plasmablasts/plasma cells and/or the repertoire in memory B cells. RNA isolation and next-generation sequencing of the BCRh repertoire RNA isolation from peripheral blood was carried out using PAXgene blood miRNA isolation kit (Qiagen, Hilden, Germany). Complementary DNA (cDNA) of BCR heavy chain molecules was synthesized using a BCR heavy chain (BCRh) Joining reverse primer tagged with a 9 random nucleotide UMI and a consensus sequence, followed by Exonuclease I treatment (Thermo Fisher Scientific, Breda, The Netherlands) to remove left over primers. This was followed by a PCR with forward primers covering the BCRh Variable genes and a reverse primer binding to the consensus sequence previously introduced in the specific-cDNA and tagged with an 8 bp patient identifier (MID, Molecular Identifier). Obtained amplicons were purified using two rounds of AMPure XP beads clean-up (Beckman Coulter, Woerden, The Netherlands), quantified using Qubit dsDNA HS Assay Kit (Thermo Fisher Scientific), dual-indexed with i5 and i7 adapters (Nextera XT Index Kit v2) and sequenced using the Illumina Miseq Kit v3 2 x 300 bp technology according to the manufacturer’s manual (Illumina, San Diego, California, USA). Bioinformatics generation and analysis of BCRh repertoires The obtained sequencing reads were analyzed with an in-house developed pipeline for repertoire analysis “RESEDA” (REpertoire SEquencing Data Analysis, https://bitbucket.org/barbera/reseda), using the following steps: 1) pairwise assembly of the paired-end reads using PEAR, 2) identification of the 8 bp MID, 3) identification of the complement determining region 3 (CDR3), 4) alignment to the IMGT database to obtain the Variable and Joining gene assignment, 5) removal of reads with low quality bases (Q score < 30) in the CDR3, 6) clustering of reads in clones based on V gene name, J gene name and 100% amino acidic CDR3 identity, 7) UMI-based correction of clonal frequencies and 8) contamination check between samples from different 6
RkJQdWJsaXNoZXIy MTk4NDMw