EPH receptor B2 stimulates human monocyte adhesion and migration independently of its EphrinB ligands

Abstract The molecular basis of atherosclerosis is not fully understood and mice studies have shown that Ephrins and EPH receptors play a role in the atherosclerotic process. We set out to assess the role for monocytic EPHB2 and its Ephrin ligands in human atherosclerosis and show a role for EPHB2 in monocyte functions independently of its EphrinB ligands. Immunohistochemical staining of human aortic sections at different stages of atherosclerosis showed that EPHB2 and its ligand EphrinB are expressed in atherosclerotic plaques and that expression proportionally increases with plaque severity. Functionally, stimulation with EPHB2 did not affect endothelial barrier function, nor did stimulation with EphrinB1 or EphrinB2 affect monocyte‐endothelial interactions. In contrast, reduced expression of EPHB2 in monocytes resulted in decreased monocyte adhesion to endothelial cells and a decrease in monocyte transmigration, mediated by an altered morphology and a decreased ability to phosphorylate FAK. Our results suggest that EPHB2 expression in monocytes results in monocyte accumulation by virtue of an increase of transendothelial migration, which can subsequently contribute to atherosclerotic plaque progression.


INTRODUCTION
Cardiovascular disease (CVD), caused by atherosclerosis, remains the leading cause of death. 1 Despite the use of CVD risk lowering agents patients still suffer from CVD events, which suggest that additional, hitherto unaddressed, factors are involved. 2,3 Recently, a longassumed role for inflammation in the atherosclerotic process has been proven, [4][5][6] indicating that other pathophysiologic processes may play a role in CVD development. Unraveling novel players in the complex atherosclerotic process may ultimately result in novel targets for therapies to address the endemic burden of atherosclerosis.
Abbreviations: AKT, protein kinase B; ASO, antisense oligonucleotide; CCR2, C-C chemokine receptor 2; CD-, Cluster of differentiation; CVD, cardiovascular disease; ECIS, electric cell-substrate impedance sensing system; EPH, erythropoietin-producing hepatocellular receptor; Ephrin, EPH receptor interacting protein; FAK, focal adhesion kinases; HBSS++, HBSS with calcium and magnesium; HUVEC, human umbilical vein endothelial cell; ITGB1, integrin -1; MCP-1, monocyte chemotactic protein 1; NGC, neuroimmune guidance cue; PBS, phosphate buffered saline; Y, Tyr, Tyrosine. initial steps of atherogenesis. 10 Specifically, several studies have shown that members of the Ephrin family are involved in atherosclerosis related processes. [10][11][12][13][14][15] Erythropoietin-producing hepatocellular receptors (EPHs) and their EPH receptor interacting protein (Ephrin) ligands comprise a large family of receptor tyrosine kinases with 14 EPH receptors and 8 Ephrin ligands that are both membrane bound. A special feature of the Ephrins and their receptors is that they can induce bidirectional signaling. Not only does binding of the ligand to the receptor induce signaling (forward signaling), but also receptor-to-ligand binding induces signaling (reverse signaling). Both forward and reverse Ephrin signaling impacts on a variety of signaling pathways that mostly converge to regulation of the cytoskeleton and therewith can influence processes such as cellular adhesion, migration, and vascular stability. Due to its role in a variety of cellular processes, deregulation of the Ephrins have been associated with several diseases, including atherosclerosis. 16 Multiple Ephrins and EPH receptors have been found in human atherosclerotic plaques. 13,14,17 In addition, the EPH receptor genes EPHA2, EPHA8, and EPHB2 are located on chromosome 1 within region 1p34- 36, which has been identified as a locus for myocardial infarction by a genome wide search for susceptibility genes for myocardial infarction. 18 However, the functional role for Ephrins and EPH receptors in atherosclerosis is largely unexplored. We hypothesized that these molecules are expressed in human monocytes and endothelial cells, both culprit cell types in atherosclerosis, and contribute to atherogenesis. In this study, we identified the EPHB2 receptor and the ligands EphrinB1 and EphrinB2 as highly expressed Ephrin family members on monocytes and endothelial cells respectively. In addition, we showed increasing expression of both EPHB2 and EphrinB in progressing stages of atherosclerosis. Furthermore, we demonstrated an important EphrinB ligand-independent role for EPHB2 in the atherosclerotic process, by promoting monocyte adhesion through phosphorylation of focal adhesion kinases (FAK).

Database
We evaluated which NGCs are expressed by either monocytes or endothelial cells by means of the GENEVESTIGATOR 19

Transduction of THP1 cells
To achieve a knockdown of EPHB2 or EphrinB1, THP1 monocytes were transduced with lentiviral particles encoding shRNA against de coding region of EPHB2 or EphrinB1 (MISSION library Sigma-Aldrich, TRCN0000006424 or TRCN0000058656 respectively) or a mock. Selection of transduced cells was achieved using puromycin (3.33 µg/ml).

Barrier function assay
Endothelial barrier function analysis was performed with impedancebased cell monitoring using the electric cell-substrate impedance sens-

Migration assay
Chemotaxis of THP1 monocytes was measured using a 24-well Boyden chamber with a 5 µm pore size filter (734-1573, Corning, Kennebunk, ME, USA) coated with 10 µg/ml fibronectin (F4759, Sigma, Saint Louis, MO, USA). Cell migration toward 10 ng/ml recombinant human MCP-1 (279-MC, R&D Systems, Minneapolis, MN, USA) and/or 500 ng/ml EphrinB1 or EphrinB2 was measured after 3 h. Cells were resuspended and counted in randomly selected fields for each well to determine the number of cells that had migrated into the lower chamber. Each condition was performed in triplicate.

Statistical analyses
Data was analyzed by unpaired two-tailed t-tests for 2 groups or with ANOVA and post hoc t-tests by the Tukey method for multiple groups.
P-values of < 0.05 were considered statistically significant. All statistical analyses were performed with SPSS version 24 or Graphpad Prism 7.

Increased expression of EPHB2 and EphrinB in progressive human atherosclerotic lesions
Atherosclerosis is a systemic inflammatory disease, characterized by the accumulation of inflammatory cells in the vascular wall. 21  and an increase in endothelial mRNA expression of EphrinB1 was observed when cells were exposed to pro-atherosclerotic conditions The increase in EphrinB expression was mainly observed in the intima and primarily in the most severe disease state (Fig. 2E). In addition, when looking specifically to the plaque area, an increase in EphrinB expression was observed (Fig. 2F).

EPHB2-induced reverse signaling has no effect on endothelial barrier function
Based on the increased EphrinB expression in atherosclerotic plaques and in endothelial cells upon exposure to pro-inflammatory cytokines ( Fig. 2), we hypothesized that monocyte binding to the endothelium suring electrical resistance with ECIS (Fig. 3A). No difference in barrier function of the endothelial monolayer was observed when EphrinB reverse signaling was induced by addition of EPHB2 (Fig. 3B). Higher or lower concentrations of EPHB2 also did not alter barrier function ( Supplementary Fig. 3A), while induction of Semaphorin3A signaling did result in a decrease in barrier function (Fig. 3B).

EphrinB-induced forward signaling has no influence on monocyte trafficking
In addition to the role of reverse signaling on endothelial barrier function, we assessed the role of EphrinB1-and EphrinB2-induced forward signaling on monocyte adhesion and migration. Monocytes were stimulated with recombinant EphrinB1 or EphrinB2 for 30 min before adding them to a confluent monolayer of endothelial cells (Fig. 3C).
Both EphrinB1 and EphrinB2 stimulation did not change the adhesion ability of the monocytes, while stimulation with IL-1 did induce monocyte adhesion ( Fig. 3D and Supplementary Fig. 3B). Next, using the Boyden chamber assay (Fig. 3E), monocyte migration toward MCP-1 in the presence or absence of EphrinB1 or EphrinB2 was examined.

F I G U R E 3 Induced EPH-Ephrin signaling has no effect on endothelial barrier function, monocyte migration, and adhesion. (A) Schematic
We observed that both EphrinB1 and EphrinB2 in the absence of MCP-1 had no chemoattractant effect on the monocytes, nor did EphrinB1 or EphrinB2 in combination with MCP-1 had an antagonistic effect on monocyte chemotaxis (Fig. 3F).

EPHB2 on monocytes promotes monocyte adhesion and migration
As a ligand-dependent effect for EPHB2 could not be confirmed, we tested the ligand-independent potential of EPHB2 on monocyte adhesion and migration. For this, EPHB2 expression in THP1 cells was silenced using a lentiviral shRNA targeting EPHB2 mRNA. A non-EPHB2 targeting scrambled shRNA was used as a control. showing a decrease by ≈75% on mRNA level and ≈40% on protein level ( Fig. 4A-C). Adhesion of these THP1 cells to either fibronectin or to a confluent monolayer of HUVECs was diminished compared to control THP1 cells (Fig. 4D-F). Stimulation of THP1 cells with EphrinB1 or EphrinB2 before adhesion to HUVECs did not result in differences in adhesion capacity of the EPHB2 knockdown THP1 cells ( Fig. 4F and Supplementary Fig. 3B). In line with the reduced adhesion of EPHB2 knockdown THP1 cells, also migration toward MCP-1 was decreased in the EPHB2 knockdown THP1 cells compared to control cells, which was not influenced by the addition of EphrinB1 or EphrinB2 (Fig. 4G).

EPHB2 affects actin cytoskeleton via phosphorylation of FAK
To explain the reduced adhesion and migration of THP1 cells with reduced levels of EPHB2, we hypothesize that this could be mediated by changes in the expression levels of the MCP-1 receptor C-C chemokine receptor 2 (CCR2) or the main binding integrin of mono-cytes Integrin -1 (ITGB1). mRNA expression of CCR2 did not differ in EPHB2 knockdown cells compared to mock treated cells while mRNA levels of ITGB1 were slightly, but significantly decreased (Fig. 5A).
Despite the observed moderate decrease in ITGB1 on mRNA level, no change was observed in protein expression of ITGB1 (Fig. 5B). Visualization of the cells revealed a more rounded morphology and smaller cell area upon adhesion in monocytes with reduced expression of EPHB2 compared to control monocytes ( Fig. 5C and D). knockdown cells, but this difference did not reach statistical significance ( Fig. 5E and F). We next investigated whether phosphorylation of FAK via EPHB2 was entirely ligand-independent and not caused by a cis-interaction between the EPHB2 receptor and its EphrinB ligands on the same cell. From the EphrinB ligands, monocytes highly expressed EphrinB1 (Fig. 1A). We therefore transduced THP1 cells with a lentiviral shRNA targeting EphrinB1 mRNA. Gene and protein analysis showed a significant reduction in EphrinB1 expression and a slight reduction of EphrinB2 ( Supplementary Fig. 5A-C). In these cells, we observed an increase in phosphorylation of FAK upon knockdown of EphrinB ligands compared to control cells ( Fig. 5G and H). Together with the reduced FAK phosphorylation upon EPHB2 knockdown, this suggests that activation of FAK via EPHB2 most likely occurs via receptor dimerization and is independent of its ligand in both cis-and transinteractions (Fig. 5I).

DISCUSSION
It is acknowledged that Ephrin family members are involved in atherosclerotic related processes, like among others leukocyte chemotaxis, adhesion, and migration, and regulation of atherosclerotic inflammation. [12][13][14]24 This is not surprising since EPHA2, EPHA8, and EPHB2, are located within the murine Athsq1 atherosclerosis susceptibility locus, 25 which is highly homologous to the premature myocardial infarction susceptibility locus in human 18 that similarly contains EPHA2, EPHA8, and EPHB2. Using multiple functional assays, we now show pro-atherosclerotic functions of EPHB2, since reduced levels of this receptor resulted in less monocyte adhesion and migration via decreased phosphorylation of FAK, suggesting a role for EPHB2 in monocyte accumulation in atherogenesis.
In the current study, we have shown, for the first time to our knowledge, a plaque burden-dependent expression of EPHB2 and EphrinB ligands in atherosclerotic plaques. In accordance to a paper of Sakamoto and co-workers, we show that expression of EPHB2 and EphrinB is increased in advanced atherosclerotic plaques, 14   with a drug-conjugated Ab raised against EPHB2, which is expressed in melanomas, neuroblastomas, gastric, lung, and colon cancers, [43][44][45][46] which could induce cell death in EPHB2 expressing cells both in vitro and in vivo, 47 no clinical data has been reported yet. Whether this Ab will be a useful treatment option needs to be further investigated and discovering its therapeutic potential might even guide the way for its implication in other diseases like, for example, atherosclerosis.
Since we have shown in this study that lowering the expression of EPHB2 on monocytes inhibits monocyte adhesion and migration, cell specific targeting of EPHB2 remains a promising potential therapeutic target for atherosclerotic disease. The upcoming field of antisense oligonucleotides (ASOs) 48 might in time provide opportunities to specifically deliver ASOs raised against EPHB2 to inflamed regions and thereby reduce subendothelial monocyte accumulation. However, these options are still far from clinically relevant and further exploration of not only the ASOs but also EphrinB and EPHB2 is essential for discovering new therapeutic options.
In summary, the present study demonstrates an increased expression of EphrinB and EPHB2 in progressive human atherosclerotic tissue. Although the exact means by which Ephrins affect atherosclerosis development remains to be elucidated, we have shown that EPHB2 plays a role in atherosclerosis by mechanisms that are not related to the activation by trans nor cis interaction of the currently known EphrinB ligands. We show that the effect of EPHB2 is partially explained by its effect on FAK phosphorylation. The