Profiling the Autoantibody Repertoire of Acquired Thrombotic Thrombocytopenic Purpura (aTTP) Patients By Single Cell Sorting and Deep Sequencing of Splenic B-Cells

Introduction: Antibodies are the primary effector molecules in the humoral immune system. To create a diversity of receptors-antibodies B-cells undergo V(D)J gene recombination and somatic hypermutation. This allows the recognition of most of pathogens but also sometimes of self-antigen, which can lead to autoimmune diseases. Autoantibodies (Abs) neutralizing and/or accelerating the clearance of ADAMTS13 are present in nearly all acquired thrombotic thrombocytopenic purpura (aTTP) patients with severe ADAMTS13 activity (<5%). As increasing evidence points at the spleen as a major reservoir for antigen specific memory B-cells, we investigated the splenic B-cell repertoire. We sequenced heavy and corresponding light chains from single antigen-specific memory B-cells and deep sequenced global antibody repertoire among different B-cell populations. Combining both, the single antigen-specific repertoire and the general global repertoire from the same patients specimens allows an investigation of the V(D)J gene usage and abundance of specific V(D)J junction in the B-cell populations within one patient and also among different patients. Analysis of the repertoire enables better understanding of the humoral autoimmune response.

Methods: Specimens were analyzed from 8 aTTP patients who were splenectomized due to a refractory course of the autoimmune disease. Splenic mononuclear cells were sorted by flow-cytometry into naïve and transitional (CD27-, IgD+), unswitched memory (CD27+, IgD+) and switched memory (CD27+, IgD-) as well as (CD27+, CD20-/CD19-, CD38+) plasma cells. The frequencies of highly positive anti-ADAMTS13 B-cells in each B-cell population were calculated. Library consisting of antibody cDNA from each B-cell population was deep sequenced on MiSeq Ilumina. Prime analysis were performed in CLC Genomics Workbench. High quality sequences were submitted to IMGT/high V-QUEST web-based analysis tool to determine for V(D)J genes alignments, and V(D)J junction (antigen binding region) and mutations analysis. The IMGT output was parsed into MySQL database for further analysis. In addition, from 4 patients anti-ADAMTS13 specific IgG bearing cells were individually sorted. Gene transcripts of single cells were reverse-transcribed followed by nested PCR of IgG heavy/light chains. V(D)J pairings were visualized with Circos software. In MySQL we compared single cell antigen specific sequence with global repertoire.

Results: Anti-ADAMTS13 specific B-cells were detected in the spleen of all patients (average 0.01% of total B-cell population). Deep sequencing of the total B-cell repertoire revealed ~3 million productive sequences, from which ~2 million were unique. Splenic anti-ADAMTS13 specific B-cells of four aTTP patients revealed 80 antibodies with unique V(D)J junction. Among those most frequently used V-genes were IGHV1-69 and IGHV3-30 (15% and 12% respectively) in global repertoire 1.5% of antibodies were encoded with IGHV1-69 gene within respective population. The average identity to germline was lower in ADMTST13 specific B-cells (92% compare to 96%). Four V(D)J junctions were convergent in at least 2 patients (identical amino acid sequence).

Conclusion: Anti-ADAMTS13 specific B-cells were found in all aTTP patients in different B-cell populations. We observed enrichment of some variable gene segments when comparing the specific anti-ADAMTS13 to the total splenic repertoire (mainly IGHV1-69). Finding convergent V(D)J junction is very promising and also confirms previous finding of our group where similar V(D)J junction were found among two patients. Currently we clone selected single sorted monoclonal Abs (Schaller et al, Blood 2014;124(23):3469-79). Functional testing will allow the selection of the inhibitory Abs to be used as tools to develop anti-idiotypic specific therapies for aTTP patients.