Making central T-cell tolerance efficient: thymic stromal cells sample distinct self-antigen pools.

Abstract

Induction and maintenance of self-tolerance in the developing and mature T-cell repertoire is mediated by multiple mechanisms operating both in the thymus (“central tolerance”) and in peripheral lymphoid and non-lymphoid organs (“peripheral tolerance”) [1,2]. The thymus is viewed as the prime site of T-cell tolerance induction to ubiquitous (e.g. house keeping genes or MHC antigens) and abundant blood-borne antigens entering the thymus via the circulation (e.g. C5 complement protein) [3–5]. Though intensely investigated, tolerance to self-antigens which are confined to specific tissues and organs often expressed by only a minor set of cells (e.g. α-cells of the pancreas), has largely remained puzzling [68]. The prevailing view that tolerance induction via negative selection in the thymus applies only to a limited set of “abundant” proteins, however, has been challenged by recent fmdings that intrathymic expression of “tissue-specific” antigens is a common occurrence. This phenomenon is termed “ectopic” or “promiscuous” gene expression. Interestingly, this ectopic expression appears to be a unique property of thymic epithelial rather than bone marrow-derived antigen presenting cells [9–17]. Here we provide evidence that confinement of this interesting property to a certain stromal cell type, in conjunction with different efficiencies to present exogenous antigens, results in sampling by distinct thymic stromal cells of largely non-overlapping self-antigen pools. We will argue that this subtle and intricate functional complementation in tolerance induction by epithelial cells and hemopoietic cells extends the scope of central T-cell tolerance to a wide range of tissue-specific self-antigens.KeywordsTolerance InductionThymic Epithelial CellThymic MedullaThymic Stromal CellMedullary Epithelial CellThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.