Edema formation is a cardinal sign of inflammation that contributes to the pathogenesis of inflammatory disorders. Leukocytes need to adhere to the endothelial cells in order to induce increased vascular permeability with concomitant tissue swelling and local edema, since anti-CD18 treatment inhibits both leukocyte adhesion and edema development 
. The current study clearly demonstrates that leukocyte adhesion per se is not responsible for the observed permeability increase, since despite equal number of adherent cells in the wild-type and Mac-1 deficient mice, permeability increased only in the wild-type group at early time points (30 min after MIP-2 addition to the superfusate, ). Instead down-stream events from adhesion, namely neutrophil emigration, seem to be responsible for increased vascular permeability. This observation is not dissimilar to studies that showed that chemoattractants given on the luminal side, caused adhesion but no emigration and no vascular permeability increase whereas application of chemoattractants to the ablumenal side increased adhesion, emigration and vascular permeability 
Leukocytes can transmigrate out of the vasculature by using either the paracellular or the transcellular route 
. The factors determining which pathway is preferred are unknown, but the pathway of choice might differ between organs, inflammatory stimuli and the type of leukocyte that is emigrating. Since emigration of neutrophils in vivo seems to occur mainly through the paracellular pathway 
, noting some exceptions including the use of high concentrations of bacterial products 
, one could speculate that transcellular migration is less regulated and might cause greater barrier damage. Recently, using intravital time-lapse video microscopy of muscle microcirculation exposed to MIP-2, we revealed that neutrophils upon adhering to the venular wall crawl to a junction where they then would emigrate 
. By using a confocal system, it was revealed that in this model 86% of the emigration occurred at junctional regions. We found that adhesion was dependent on the β2
integrin LFA-1 (CD11a/CD18), and crawling occurred through activation of the β2
-integrin Mac-1 (CD11b/CD18) 
. When the neutrophils were not permitted to crawl due to lack of Mac-1, adhesion was unaffected but 60% of the neutrophils emigrated transcellularly. This served as a model for paracellular (wt) and transcellular migration (Mac-1-/-) under otherwise similar conditions. Surprisingly, we found that the permeability increase of the vascular endothelium was not at all affected by the route of emigration, despite evidence of greater barrier disruption in the Mac-1-/- mice emigrating transcellularly. Electron microscopy however revealed a rate-limiting barrier or dome that sealed the neutrophil from the lumen thereby allowing extensive disruption of the endothelium under the neutrophil without physiologic impact on barrier function. The increase in vascular permeability would then only be a result of the plasma protein that got encapsulated along with the neutrophil.
Ligation of the endothelial adhesion molecules by the adherent leukocytes directly affect the endothelial cells, as demonstrated by transient changes in intracellular calcium, which leads to activation of myosin light chain kinase and endothelial cell shape changes 
. The induction of endothelial cells to change shape is believed to be important in order for the endothelial cells to retract. This localized dissociation of endothelial cell junctions occurs in order to enable emigration of leukocytes 
. Newer evidence shows that following firm neutrophil adhesion, the endothelial cells change their morphology so that transmigratory cups are formed, where microvilli-like projections rich in actin, ICAM-1 and VCAM-1 are docking the neutrophil to the endothelium 
. The endothelial cup-like structures partially engulfing the adherent leukocyte are formed regardless of the emigratory route taken 
. Electron micrographs also revealed these endothelial docking structures in vivo () but using both electron microscopy and spinning disk microscopy, we now for the first time show that these docking structures developed into endothelial dome-like encapsulations that completely surrounded the neutrophils ( and ). Interestingly, despite clear evidence that docking structures occur in vitro, dome structures have not previously been observed in vitro suggesting either that the domes do not form in vitro or are simply too difficult to see with presently available technology.
Recently, a new association between endothelial cells and leukocytes was described as transendothelial emigration of lymphocytes was found to be highly dependent on adherent lymphocytes forming podosome protrusions that initiated endothelial invaginations or pseudoprints 
. Extension of the lymphocyte podosome could eventually lead to pore formation through the endothelial cell, via which the lymphocyte would emigrate 
. Whether neutrophils could emigrate through transendothelial pores is still unknown, and neutrophil podosome formations have not yet been described in vivo although our electron micrographs did reveal small breaks (potentially pores) below neutrophils (). Contrary to lymphocytes and basophils, neutrophils emigrate in very large numbers during the early immune response. If this occurred through pores, the endothelial encapsulation might be an important phenomenon selectively for neutrophil emigration. Although domes were not described in in vitro studies, it could be that they form in microvascular endothelium under shear conditions when neutrophils are emigrating across an intact vessel wall, conditions difficult to mimic completely in in vitro systems.
Others have previously noted endothelial cytoskeletal rearrangements during neutrophil transmigration after exposure in vivo to either LTB4
or fMLP using electron microscopy 
. This resulted in endothelial domes surrounding transmigrated neutrophils that were suggested to reseal the endothelium after neutrophil diapedesis, but clear gaps were always identified in the endothelium either apically or basolaterally of the neutrophil 
. In this study, similar but complete encapsulations were observed in the wt and Mac-1 deficient mice, despite the fact that the endothelium on the basolateral side of the Mac-1-/- neutrophils often had multiple gaps directly below the neutrophils (see arrows ). However, somewhat unexpectedly the microvascular permeability as an index of barrier function was not higher in the Mac-1-/- than wt group, underscoring the potential importance of these domes. Since the Mac-1-/- neutrophils predominantly emigrate transcellularly, while the wild-type neutrophils exit through junctions, we can from our data conclude that transcellular and junctional emigration seems to cause permeability changes of the same degree. We would suggest that the dome-like endothelial structure completely covering the neutrophil would be the rate limiting structure for vascular permeability, minimizing additional plasma protein leakage regardless of route of emigration.
In conclusion, by comparing mice where emigration occurs predominantly paracellularly versus predominantly transcellularly, this study shows that vascular permeability changes seen during neutrophil recruitment may be limited by an encapsulation of the adherent neutrophil thereby forming an air lock type seal. The permeability would then be independent of the route of emigration. Moreover, the docking structures also referred to as endothelial emigratory cups formed by the endothelium to surround the base of emigrating neutrophils regardless of route of emigration 
were the initiators of the endothelial domes.