Shedding light on class I phosphoinositide 3-kinase activity in endothelium

To the editor:

Class I phosphoinositide 3-kinases (PI3Ks) play a critical role in regulating chemoattractant-induced migration of neutrophils.¹  However, Puri et al have demonstrated that PI3K activity in vascular endothelium also contributes to the accumulation of these cells in tissues, as evidenced by the reduced ability of PI3K-deficient blood vessels to support rolling adhesion of WT neutrophils in response to the proinflammatory cytokine tissue necrosis factor α (TNFα; 20 ng).^2,3  By contrast, a recent publication by Liu et al suggests that endothelial PI3K activity is not essential for this process.⁴  Although both studies used chimeric animals in which PI3Kγ or PI3Kδ activity was present in neutrophils but absent in endothelium, it was speculated that the observed discrepancies were attributable to the amount of TNFα used to induce inflammation (500 ng vs 20 ng). We wish to point out that the use of transmitted light microscopy and electrocautery tissue dissection by Liu et al may account for these inconsistencies rather than differences in cytokine concentrations.

To confirm our hypothesis, we used WT or p110γ^−/− animals that were irradiated and reconstituted with WT fetal liver cells from mice expressing GFP in granulocytes, and evaluated neutrophil behavior in the microcirculation of cremaster muscle stimulated with TNFα (1000 ng). The absence of endothelial PI3Kγ activity was associated with a 5-fold increase in neutrophil rolling velocities, findings consistent with those of Puri et al (54.8 ± 4.2 μm/s vs 10.2 ± 0.6 μm/s in WT animals, P < .001, data represent means ± SEM, n = 80 cells per genotype). Moreover, the number of neutrophils extravasating into tissue was dramatically decreased in PI3Kγ chimeric animals as compared with WT counterparts (25 ± 2 cells vs 65 ± 2 cells, respectively, P < .001, Figure 1A,B,D). By contrast, the use of electrocautery rather than mechanical tissue dissection (microscissor) caused robust cell migration in PI3Kγ chimeric animals (79 ± 8 cells, P < .001, Figure 1B-D), results consistent with that of Liu et al.

Mechanistically, Puri et al speculated that class I PI3K activity in endothelium contributes to neutrophil trafficking into tissues by regulating the surface distribution but not expression of E-selectin, an adhesion molecule that mediates the rolling of these cells on the inflamed vessel wall. This is supported in the recent publication by Setiadi et al demonstrating that the physical clustering of E-selectin on cytokine-stimulated endothelium is essential for this process.⁵  Indeed, isoform-specific blockade of class I PI3K activity does alter E-selectin clustering of on the surface of TNFα-stimulated human umbilical cord vein endothelial cells (HUVECs; Cambrex Bio Science, Walkersville, MD; Figure 1E,F). By contrast, the surface distribution of PECAM-1, an adhesion molecule constitutively expressed on vascular endothelium, remained unchanged (Figure 1E,G), as did that of PSGL-1, a selectin ligand present on the surface of neutrophils (data not shown). These findings, in conjunction with the report by Sediati et al, provide a potential mechanism by which class I PI3Ks contribute to the proadhesive state of inflamed endothelium.

In conclusion, our data reaffirm the validity of the results by Puri et al and illustrate the importance of using appropriate imaging and surgical techniques when evaluating intracellular signaling pathways that play a critical role in neutrophil trafficking.