To transmigrate or not to transmigrate

Neutrophil transmigration through the endothelial lining of blood vessels is regulated by signaling and adhesion molecules. Now, CD99L2 joins the club of molecules implicated in this process.

In a remarkable study published in this issue of Blood, Bixel and colleagues describe the important role of CD99L2, a molecule with a modest 32% homology to CD99 and no other close relatives, in neutrophil transendothelial migration. This increases the number of molecules known to be involved in neutrophil transmigration by one fifth, expanding on PECAM-1, JAM-A, JAM-C, ESAM and CD99, which have previously been implicated in this process.¹  In addition, ICAM-1 and ICAM-2 are involved in transendothelial migration, but also mediate neutrophil adhesion. Based on electron microscopy, CD99L2 seems to inhibit a distal step in transmigration, because in the presence of an antibody to CD99L2 neutrophils are found trapped between the endothelium and the basement membrane. A similar phenotype was previously seen in mice treated with an antibody to PECAM-1 and in PECAM-1 knockout mice.^2,3 

In an in vitro transmigration experiment through bEnd.5 endothelioma cells, transmigration of bone marrow neutrophils was inhibited by 40% by an intact antibody to CD99L2 and by 30% by F(ab′)₂ fragments. The effect of the full-length antibody was similar whether only endothelial cells were preincubated or the antibody remained present during the transmigration assay. Although no effect of exposing bone marrow neutrophils to CD99L2 antibody was seen, this remains to be confirmed with blood neutrophils, which express 100-fold more CD99L2. Incidentally, this regulation pattern makes CD99L2 one of the very few markers that distinguish mature (blood) from immature (marrow) neutrophils.

Although CD99L2 has a significant extracellular domain, it does not support homophilic adhesion. The calcium and magnesium sensitivity of the CHO-cell aggregation assay might suggest a possible integrin ligand for CD99L2, but this is pure speculation and awaits further study. Since all antibody reagents used were bivalent, ligation of CD99L2 may trigger a signaling event in endothelial cells, but the nature of this signaling and the domains in CD99L2 responsible remain unknown. Neutrophil transmigration requires an exquisitely regulated dance between the leukocyte and the vessel wall. When endothelial CD99L2 is ligated by antibody, the partners still find each other, since rolling and adhesion are unaffected, but their ability to dance is reduced. Future work will have to determine what the relevant ligands for CD99L2 might be and whether they are expressed on transmigrating neutrophils, adjacent endothelial cells, or both. The phenotype of CD99L2-deficient endothelial cells and mice remains to be explored. Based on the data presented by Bixel and colleagues, it seems unlikely that CD99L2 deficiency will produce a neutrophil transmigration defect.

Why are so many molecules involved in neutrophil transmigration? Possibly they form a chain of adhesion and signaling events, one handing off to the next. CD99L2 and CD99 are in the same general pathway, because blocking both has no additive effect on transmigration (as pointed out by Bixel and colleagues), but we don't know the details of their relationship. PECAM-1 has been described as functioning upstream (luminal) from CD99 in neutrophil transmigration⁴  and may engage in homophilic adhesion, which is clearly different from CD99L2. While many aspects of CD99L2 biology remain to be explored, the present paper takes the all-important first step in linking this molecule to transmigration.