Reverse transendothelial cell migration in inflammation: to help or to hinder?

Abstract

The endothelium provides a strong barrier separating circulating blood from tissue. It also provides a significant challenge for immune cells in the bloodstream to access potential sites of infection. To mount an effective immune response, leukocytes traverse the endothelial layer in a process known as transendothelial migration. Decades of work have allowed dissection of the mechanisms through which immune cells gain access into peripheral tissues, and subsequently to inflammatory foci. However, an often under-appreciated or potentially ignored question is whether transmigrated leukocytes can leave these inflammatory sites, and perhaps even return across the endothelium and re-enter circulation. Although evidence has existed to support "reverse" transendothelial migration for a number of years, it is only recently that mechanisms associated with this process have been described. Here we review the evidence that supports both reverse transendothelial migration and reverse interstitial migration within tissues, with particular emphasis on some of the more recent studies that finally hint at potential mechanisms. Additionally, we postulate the biological significance of retrograde migration, and whether it serves as an additional mechanism to limit pathology, or provides a basis for the dissemination of systemic inflammation.

Keywords: Endothelial cell; Intravasation; Monocyte; Neutrophil; Reverse interstitial migration; Reverse migration; T cell; Transmigration; rTEM.

Fig. 1

The contrasting mechanisms and downstream effects of neutrophil reverse transendothelial migration (rTEM). Top recent evidence shows that rTEM of neutrophils leads to the dissemination of inflammation following ischemia–reperfusion injury. Neutrophils are recruited by inflammatory mediators such as leukotriene-B4 (LTB4) made by neutrophils and other leukocytes. In addition to recruitment, LTB4 stimulates the production of neutrophil elastase, which proteolytically cleaves the junctional adhesion molecule (JAM-C). This results in a disruption to the integrity of the endothelial barrier, allowing transmigrated neutrophils to re-enter circulation and, therefore, disseminate inflammation. Interestingly, neutrophils that have undergone rTEM can be distinguished by high expression of ICAM-1 and low expression of CXCR1, thus providing useful markers for identifying rTEM neutrophils and their subsequent role in inflammation. Bottom based mainly on in vitro and zebrafish models, evidence suggests that rTEM of neutrophils can lead to resolution of inflammation. Neutrophils that perform phagocytosis appear to lose their ability to further migrate, and are possibly retained in tissue. Concordantly, overexpression of HIF-1α prevents tissue retention. Whether there is a link between phagocytosis and HIF-1α expression remains to be elucidated. Interestingly, the small molecule, tanshinone, promotes neutrophil rTEM and overcomes HIF-1α overexpression. In addition, rTEM of neutrophils is promoted by macrophage contact and this is regulated by redox-SFK signaling

Fig. 2

Evidence for rTEM by other immune cell types. a CCL5 is a strong chemoattractant molecule, promoting the movement of T cells across the endothelium. Conversely, recent in vitro data have suggested that endothelial cell-derived CXCL12 (acting through the receptor CXCR4) can drive T cells to undergo rTEM. b Monocytes have both macrophage and dendritic cell differentiation capabilities. The differentiation capability of monocytes recruited into tissue is dependent on their localization. Monocytes that remain associated with the endothelium (dependent on ICAM-1:LFA-1 interactions) may differentiate into dendritic cells (DCs). Endothelial-associated DCs (eDCs) through the concerted effect of endothelial-derived TGF and GM-CSF have been shown to have the ability to drive Th17 differentiation in some systems. Additionally, monocyte-derived dendritic cells (MD-DCs) in vitro are able to undergo rTEM and have T cell-activating potential, possibly at distal sites. c Systemic stimulation of monocytes with Chlamydia muridarum disseminates infection to the arterial intima. This bacterial infection is subsequently cleared from the intima without a local inflammatory response or the requirement for adaptive immunity. The infection induces CCL19/CCR7 signaling in DCs and this promotes their rTEM to clear the infection. MD-DC monocyte-derived dendritic cell, eDC endothelial-associated dendritic cell, TCR T cell receptor, MHCII major histocompatibility class II

Fig. 3

rTEM during tumor metastasis. Dissemination of tumor cells from tissues to the periphery relies on proteases such as ADAM12, produced by the endothelium, and MT4-MMP produced by tumor cells. These enzymes can pre-condition the endothelium to support rTEM by cleaving junctional molecules and extracellular matrix components. Likewise, macrophage-derived epidermal growth factor (EGF) acts on tumor cells to activate the PI3K signaling pathway, drive ‘invadopodia’ formation via NWASP/RhoA activation, and thus promote rTEM. ADAM12 a disintegrin and MMP domain 12, MT4-MMP4 membrane type 4-matrix metalloprotease, EGF epidermal growth factor, PI3K phosphoinositide 3-kinase, NWASP neural Wiskott–Aldrich syndrome protein, RhoA Ras homolog gene family, member A