An investigation into autoimmune pathogenesis by tracking and profiling T cell clones

Abstract

Multiple immune tolerance checkpoints ensure that T cells carrying self-reactive T cell receptors (TCRs) are either purged during development or inactivated in the periphery. How these self-reactive T cell clones break tolerance mechanisms and drive disease in autoimmune disorders is poorly understood. When two tolerance checkpoints are crippled in mice by crossing Aire-deficient mice, which have a profound defect in thymic deletion, to Cblb-deficient mice, which display a defect in T cell anergy, the offspring develop an infant-lethal autoimmunity directed exclusively against the exocrine pancreas and salivary glands. Establishing a mechanistic basis behind the co-operation of Aire and Cblb deficiencies forms the primary goal of this thesis since it may reveal principles for understanding how specific autoimmune diseases are precipitated. We hypothesized that the underlying autoimmune pathogenesis in the Aire-/-Cblb-/- mouse model is dependent upon unique autoimmune "driver" T cell clones that are normally regulated by AIRE and CBLB-dependent tolerance mechanisms. To profile TCR diversity in organ-infiltrating T cells we first employed an enrichment strategy by adoptively transferring Aire-/-Cblb-/- lymphocytes into lymphopenic recipients. The paired TCRalpha and TCRbeta chains of pancreas-infiltrating T cells were then sequenced to identify potential disease driving clones. While the tissue pattern of autoimmunity following transfer from different Aire-/-Cblb-/- donors was remarkably homogenous, the TCRs used by the infiltrating clones were surprisingly heterogenous, implying that private clones with distinct receptors can cause destructive pancreatitis. To determine if the isolated T cell clones were capable of driving pancreas-specific autoimmunity, TCR retrogenic mice expressing the TCRs of the organ-infiltrating CD4 and CD8 clones were created. A dominant CD8 clone, termed 8.1, was able to cause pancreatitis and cachexia in the absence of other T cells with different specificities. The finding that this occurs in wild-type animals sufficient in both Aire and Cblb suggested that there is no direct regulation of this clone by these two tolerance checkpoints. A second CD8 clone isolated from the same donor mouse was found to display selective activation in the pancreatic lymph node. Results from retrogenic mice expressing TCRs isolated from CD4 cells, however, did not reveal any drivers of autoimmunity. To identify and profile T cell clones involved in autoimmunity at a greater depth, a high-throughput sequencing method that could profile either the TCRalpha or the TCRbeta repertoire of a large number of T cells was established. Important technical biases were identified while establishing this method such as multiplex PCR primer bias and sequencing errors. Deep sequencing of the TCRalpha repertoire of unmanipulated Aire-/-Cblb-/- mice revealed an unchanged frequency of the TCRalpha chain of clone 8.1 and an absence of clonal expansions. The deep sequencing method was extended to reveal distinct TCR repertoires of thymocytes undergoing clonal deletion in the thymus. Deep sequencing of the TCRbeta repertoire was additionally used to analyse the expression of an alternatively spliced TCR exon, Cbeta0, and its effect on a TCR surface density. Experiments investigating Cbeta0 led to the serendipitous finding that the quantity of surface TCR is biased towards particular TCRbeta Variable regions.