Endothelial PlexinB2 suppresses human neutrophil activation in a humanized xenotransplantation model

Background: Human neutrophils are early responders in pig-to-human xenotransplantation and can injure xenogeneic endothelium through direct cytotoxicity, oxidative burst, and neutrophil extracellular trap (NET) formation. Although endothelial expression of human CD47 has been proposed as an inhibitory “self” signal via SIRPα-dependent pathways, its capacity to suppress neutrophil activation programs remains uncertain. PlexinB2 was recently identified as an endothelial ligand for neutrophil Semaphorin 4D (SEMA4D) and has been shown to suppresses neutrophil activation. We compared PlexinB2 with CD47 for their ability to suppress human neutrophil effector responses to swine endothelial cells (SECs) in vitro and in a humanized in vivo model.
Methods: SECs were engineered to express PlexinB2 or CD47, and wild-type cells used as controls. Human neutrophils from healthy donors were co-cultured with endothelial cells and stimulated with phorbol 12-myristate 13-acetate (PMA). Neutrophil reactive oxygen species (ROS) production was quantified by flow cytometry, and NET formation was assessed using SYTOX Green staining by fluorescence microscopy and flow cytometry. For in vivo evaluation, genetically modified endothelial cells embedded in Matrigel were implanted subcutaneously into humanized NOG-hG-CSF mice reconstituted with human CD34+ hematopoietic stem cells. Human neutrophil infiltration and activation were assessed on days 3 and 7 by MPO staining and flow cytometric analysis of hCD45+CD66b+ cells (CD66b expression).
Results: PlexinB2-expressing SECs showed significantly reduced neutrophil-mediated cytotoxicity compared with wild-type cells under PMA stimulation (P = 0.006). CD47 expression also reduced the cytotoxicity (P = 0.008), although the magnitude of suppression was smaller than that of PlexinB2. Consistent with direct inhibition of neutrophil activation, PlexinB2 significantly attenuated ROS production (P < 0.001) and NET formation (P = 0.002) compared to wild-type cells.
In vivo, human neutrophils infiltrated all graft types. However, PlexinB2-expressing grafts exhibited significantly fewer MPO-positive neutrophils than wild-type controls (P = 0.006) and reduced neutrophil activation, as reflected by the lower proportion of CD66b-high cells (P = 0.02). CD47-expressing grafts showed intermediate effects, with a trend toward reduced MPO-positive neutrophil infiltration, but did not reach statistical significance (P = 0.051).
Conclusion: the expression of human PlexinB2 on SEC suppresses human neutrophil cytotoxicity, oxidative burst, and NETosis in an in vitro xenogeneic endothelial model. In addition, in vivo it reduces the activation of human neutrophil infiltration in the NOG-hG-CSF humanized xenograft platform, demonstrating stronger inhibition of neutrophil activation than CD47. PlexinB2 may be a promising genetic-modification strategy molecule for mitigating initial xenograft damege by neutrophil-driven xenograft injury.