Leukocyte rolling, adhesion and transmigration were all described by the pathologists of the nineteenth century 1,2 . With the discovery of integrins, selectins and their respective ligands, and of chemokines and chemokine receptors, the leukocyte adhesion cascade emerged as a concept that began to explain the recruit- ment of leukocyte subsets to specific sites. The original model of leukocyte adhesion proposed that the cascade achieved combinatorial specificity 3,4 through the three steps of selectin-mediated rolling, chemokine-triggered activation and integrin-dependent arrest. However, recent evidence suggests that additional steps occur during integrin-mediated leukocyte adhesion, which remain incompletely understood 5,6 . Transendothelial migration was first described almost 200 years ago 1 , but its molecular mechanisms were only discovered recently 7 and were not included in the classical adhe- sion cascade 3,4 .Integrin-mediated adhesion is char- acterized by at least two events — arrest from rolling, which is mediated by increased leukocyte avidity for the endothelium (BOX 1), and a post-binding phase of adhesion stabilization, the molecular basis of which is only now beginning to emerge, although it was cor- rectly predicted to be important more than 10 years ago 4 . In the past decade, new insights have been gained into the structures and signalling events that underlie integrin activation 5,6 , into the post-adhesion events that strengthen leukocyte attachment to the endothelium, and into the molecules that are involved in leukocyte transendothelial migration 7,8 . These insights have led to an expanded version of the original three-step leukocyte adhesion cascade, which now includes slow rolling, adhe- sion strengthening, intraluminal crawling, paracellular and transcellular migration, and migration through the basement membrane (FIG. 1). Leukocyte rolling The role of selectins. Rolling is mediated by L-selectin, P-selectin and E-selectin 9 , which interact with P-selectin glycoprotein ligand 1 (PSGL1) 10 and other glycosylated lig- ands. L-selectin is expressed by most leukocytes, whereas E-selectin and P-selectin are expressed by inflamed endothelial cells. P-selectin is also expressed by activated platelets. PSGL1 has a dominant role as a ligand for all three selectins, although it was originally described as a P-selectin ligand. The binding of PSGL1 to L-selectin nucleates leukocyte–leukocyte interactions, by which adherent leukocytes 11 and leukocyte-derived fragments 12 facilitate secondary leukocyte capture or tethering, terms that are used synonymously. Secondary tethering also ena- bles leukocytes that do not express ligands for E-selectin or P-selectin to reach sites of inflammation. Although PSGL1 is expressed on almost all leukocytes, it is functional only when glycosylated correctly (BOX 2). In addition to its expression by leukocytes, PSGL1 was recently found to be expressed by certain endothelial cells 13,14 . In addition to PSGL1, E-selectin also binds to glycosylated CD44 and E-selectin ligand 1 (ESL1) 15 . *Robert M. Berne Cardiovascular Research Center and Departments of Biomedical Engineering, Molecular Physiology and Biological Physics, University of Virginia,Charlottesville, Virginia 22908, USA. ‡ Department of Pathology and Center for Biomedical Computing, University of Verona, Verona 37134, Italy. § Toronto General Research Institute and Department of Laboratory Medicine and Pathobiology, University of Toronto,Toronto M5G 1L7, Canada. || Centre for Microvascular Research, William Harvey Research Institute, London, EC1M 68Q, UK. Correspondence to K.L. e-mail: klausley@virginia.edu or klaus@liai.org doi:10.1038/nri2156 Combinatorial specificity Specificity achieved in a sequential cascade. If there are 3 rolling molecules, 15 chemokines and 2 integrins, theoretically, 90 (3 × 15 × 2) specificities are possible. Getting to the site of inflammation: the leukocyte adhesion cascade updated Klaus Ley*, Carlo Laudanna ‡ , Myron I. Cybulsky § and Sussan Nourshargh || Abstract | Neutrophil recruitment, lymphocyte recirculation and monocyte trafficking all require adhesion and transmigration through blood-vessel walls. The traditional three steps of rolling, activation and firm adhesion have recently been augmented and refined. Slow rolling, adhesion strengthening, intraluminal crawling and paracellular and transcellular migration are now recognized as separate, additional steps. In neutrophils, a second activation pathway has been discovered that does not require signalling through G-protein- coupled receptors and the signalling steps leading to integrin activation are beginning to emerge. This Review focuses on new aspects of one of the central paradigms of inflammation and immunity — the leukocyte adhesion cascade. REVIEWS 678 | SEPTEMBER 2007 | VOLUME 7 www.nature.com/reviews/immunol © 2007 Nature Publishing Group