Summary The thienopyridine platelet antagonist ticlopidine is associated with development of thrombotic thrombocytopenic purpura (TTP) but the pathophysiology of this link is unclear. Severe deficiency of disintegrin and metalloproteinase with thrombospondin motif‐13 (ADAMTS13), described in familial cases and a significant fraction of idiopathic TTP, has been reported in only a few ticlopidine‐linked cases. As ticlopidine can disrupt production of extracellular matrix (ECM) components critical to microvascular endothelial cell (MVEC) integrity in vitro, we explored the hypotheses that ticlopidine and ticlopidine‐linked TTP plasmas induce MVEC apoptosis in a manner similar to that of idiopathic TTP plasmas, and that ECM components and related mitogen‐activated protein kinase (MAPK) signalling cascades may be involved in this process. Replicating the activity of plasmas from patients with idiopathic TTP, plasma from five ticlopidine‐linked TTP patients induced apoptosis of primary human dermal, glomerular and hepatic MVEC, but had no effect on pulmonary MVEC or large vessel endothelial cells (EC). Pharmacological levels of ticlopidine initiated apoptosis with similar EC lineage restriction. In parallel, ticlopidine and plasmas from idiopathic and ticlopidine‐TTP patients decreased transcripts for the ECM component thrombospondin‐1 in MVEC, but not in large vessel EC. These changes were accompanied by prolonged induction of MAPKs extracellular signal‐related kinase (ERK)‐1/2 and p38 only in TTP susceptible MVEC. Induction of apoptosis by ticlopidine and TTP plasma was abrogated by inhibitors of ERK‐1/2 and p38 phosphorylation. In conclusion, MVEC apoptosis related to altered ECM–MVEC interactions may be a key part of the pathology of ticlopidine‐linked and idiopathic TTP.

Summary We sought to model the mechanism by which leucocytes may be actively repulsed by a β-chemokine signal. This model is used to interpret an apparent paradox in chemokine biology, whereby high levels of a T-cell chemoattractant, stromal cell derived factor-1 (SDF-1), are present in bone marrow and thymic tissues despite a paucity of mature T lymphocytes in these areas. We postulate the differential involvement in cell migration of the two binding sites on SDF-1 for its sole receptor, CXCR4, depending on whether high or low concentrations of SDF-1 are encountered by the cell. Site choice would be mediated by divergent affinities of the two binding interactions. We also propose differential signalling following SDF-1/CXCR4 interactions on the plasma membrane versus ligand/receptor complexes in endocytic vesicles. Preliminary data showing divergent susceptibility to kinase inhibitors depending on whether a cell is attracted to or repulsed by SDF-1, are consistent with this model. In terms of physical movement toward or away from a chemokine gradient, we compare the cycling of surface receptors during migration to the caterpillar drive of a tractor, which can change direction simply by altering the direction of rotation of its threads. Finally, the potential clinical implications of concentration-dependent, chemokine-based cell attraction and repulsion are discussed.