Downregulation of Endothelial Plexin A4 Under Inflammatory Conditions Impairs Vascular Integrity

Abstract

Objective: Besides hyperlipidemia, inflammation is an important determinant in the initiation and the progression of atherosclerosis. As Neuroimmune Guidance Cues (NGCs) are emerging as regulators of atherosclerosis, we set out to investigate the expression and function of inflammation-regulated NGCs. Methods and results: NGC expression in human monocytes and endothelial cells was assessed using a publicly available RNA dataset. Next, the mRNA levels of expressed NGCs were analyzed in primary human monocytes and endothelial cells after stimulation with IL1β or TNFα. Upon stimulation a total of 14 and 19 NGCs in monocytes and endothelial cells, respectively, were differentially expressed. Since plexin A4 (PLXNA4) was strongly downregulated in endothelial cells under inflammatory conditions, the role of PLXNA4 in endothelial function was investigated. Knockdown of PLXNA4 in endothelial cells markedly impaired the integrity of the monolayer leading to more elongated cells with an inflammatory phenotype. In addition, these cells showed an increase in actin stress fibers and decreased cell-cell junctions. Functional assays revealed decreased barrier function and capillary network formation of the endothelial cells, while vascular leakage and trans-endothelial migration of monocytes was increased. Conclusion: The current study demonstrates that pro-inflammatory conditions result in differential expression of NGCs in endothelial cells and monocytes, both culprit cell types in atherosclerosis. Specifically, endothelial PLXNA4 is reduced upon inflammation, while PLXNA4 maintains endothelial barrier function thereby preventing vascular leakage of fluids as well as cells. Taken together, PLXNA4 may well have a causal role in atherogenesis that deserves further investigation.

Keywords: atherosclerosis; endothelial function; inflammation; neuroimmune guidance cues; plexin; semaphorin.

Figure 1

Endothelial expression of NGCs under pro-atherogenic conditions. (A,B) Volcano plots depicting up and down regulation of NGCs after stimulation of primary human endothelial cells with IL1β (20 ng/mL) or TNFα (10 ng/mL) for (A) 5 or (B) 24 h. Results are depicted as mean of the natural logarithm transformed fold change in expression compared to unstimulated cells and plotted against significance. N = 4. (C) Quantitative PCR of mRNA expression of plexin receptors in endothelial cells after 5 h or 24 h of stimulation with IL1β (20 ng/mL) or TNFα (10 ng/mL). Data is depicted as mean of the natural logarithm transformed fold change in expression compared to unstimulated cells. Mean ± S.E.M. of N = 4, *P < 0.05. (D) Immunoblots and quantification of PLXNA4 and GAPDH protein in primary human endothelial cells stimulated with IL1β (20 ng/mL) or TNFα (10 ng/mL) for 24 h. Results are depicted as fold change in intensity. Mean ± S.E.M. of N = 4.

Figure 2

PLXNA4 downregulation alters endothelial cellular phenotype. (A) mRNA expression of PLXNA4, PLXNA1, PLXNA2, and PLXNA3 in endothelial cells treated with a lentiviral non-targeting shRNA (mock) or a shRNA against PLXNA4 (PLXNA4 KD). Results are expressed as copies corrected for GAPDH. Mean ± S.E.M. of N = 27, *P < 0.05. (B) Representative overview pictures of mock-treated and PLXNA4 KD endothelial cells. (C) mRNA expression of ICAM-1, IL-6, VE-cadherin and integrinβ1 (ITGB1) in mock or PLXNA4 KD endothelial cells. Results are expressed as copies corrected for GAPDH. Mean ± S.E.M. of N > 4, *P < 0.05. (D) Immunoblots and quantification of ICAM-1 and GAPDH protein in mock control cells and PLXNA4 KD cells. Results are depicted as fold change in intensity. Mean ± S.E.M. of N = 7. (E–I) Immunofluorescent staining of F-actin (red), VE-cadherin (green) and nuclei (blue) in mock or PLXNA4 KD endothelial cells. (E) Representative fluorescent overview photographs. (F–I) Quantification of (F) F-actin fluorescent signal, (G) F-actin cellular distribution, (H) VE-cadherin fluorescent signal or (I) VE-cadherin cellular distribution. Fluorescent signal is quantified as (mean fluorescent intensity*area)/nuclei and expressed as fold change relative to control cells. The cellular distribution is quantified as mean intensity per pixel of the border or interior area and expressed as the ratio between border and interior. Mean ± S.E.M. of N = 3, *P < 0.05. (J) Immunoblots and quantification of VE-cadherin and GAPDH protein in control cells and PLXNA4 knockdown cells. Results are depicted as fold change in intensity. Mean ± S.E.M. of N = 4.

Figure 3

Loss of endothelial PLXNA4 reduces cell proliferation and barrier function. (A) Proliferation of mock and PLXNA4 KD endothelial cells. Results are expressed as fold change to day 0 set at 1. Mean ± S.E.M. of N = 6, *P < 0.05. (B) mRNA expression of CDKN1A in mock or PLXNA4 KD endothelial cells. Results are expressed as copies corrected for GAPDH. Mean ± S.E.M. of N = 8. (C) Representative overview photographs and quantification of CDKN1A staining of mock and PLXNA4 KD endothelial cells. Results are expressed as fluorescent intensity in the nuclei of the cells. Mean ± S.E.M. of N = 4. (D) Immunoblots and quantification of CDKN1A and GAPDH protein in control cells and PLXNA4 knockdown cells. Results are depicted as fold change in intensity. Mean ± S.E.M. of N = 4. (E) Representative overview photographs and quantification of migration of mock and PLXNA4 KD endothelial cells over time. Results are presented as percentage of open area. Mean ± S.E.M. of N = 6. (F) Trans-endothelial electrical resistance of mock and PLXNA4 KD endothelial cells over time. Mean ± S.E.M. of N = 5, *P < 0.05. (G,H) Endothelial electrical resistance attributable to (G) cell-cell contacts (Rb) and (H) cell-matrix contacts (alpha) in a monolayer of mock and PLXNA4 KD endothelial cells. Mean ± S.E.M. of N = 5, *P < 0.05. (I, J) Effect of addition of 1 μg/mL SEMA3A to a stable monolayer of (I) mock-treated control or (J) PLXNA4 KD endothelial cells, on endothelial barrier function. Relative resistance is expressed as percentage of the average barrier before stimulation. Mean ± S.E.M. of N = 5, *P < 0.05.

Figure 4

Loss of endothelial PLXNA4 reduces tube-like structure formation. (A) Representative overview photographs and (B–E) quantification of tube formation of mock and PLXNA4 KD endothelial cells over time. Results are presented as (B) number of branches, (C) total length in pixels, (D) number of meshes and (E) number of nodes over time (left graph) and at 4 h (right graph). Mean ± S.E.M. of N = 3, *P < 0.05.

Figure 5

Loss of endothelial PLXNA4 induces vascular permeability. (A) Schematic overview of the leakage assay principle, where 3D cultured capillary-like vessels in the upper perfusion channel are separated from the collagen gel in the lower channel with a phaseguide. Leak tight capillary-like vessels will have no to limited leakage of red fluorescent-labeled albumin from the perfusion channel to the gel, while increased permeability of capillary-like vessels will result in increased fluorescent signal in the gel channel. (B) Representative overview photographs and (C) quantification of the apparent permeability of capillary-like vessels composed of endothelial cells with normal or decreased expression of PLXNA4 and with or without IL1β (20 ng/mL) or TNFα (10 ng/mL) stimulation. Mean ± S.E.M. of N = 4, *P < 0.05 for mock transduced cells vs. PLXNA4 KD cells ^#P < 0.05 for stimulated cells vs. untreated cells. (D) Representative overview pictures and quantification of migrated monocytes through a monolayer of mock or PLXNA4 KD endothelial cells toward MCP-1. Results are relative to mock control cells, set at 1. Mean ± S.E.M. of N = 8. (E) Quantification of adhered fluorescently labeled monocytes to mock and PLXNA4 KD endothelial cells. Results are relative to mock control cells, set at 1. Mean ± S.E.M. of N = 9.

Figure 6

Loss of endothelial PLXNA4 affects RhoA/ROCK pathway activity. (A,B) Activity of the small GTPases (A) RAC-1 or (B) RhoA in mock-treated or PLXNA4 KD endothelial cells. Results are expressed as fold change in absorbance compared to control cells, set at 1. Mean ± S.E.M. of N = 8, *P < 0.05. (C,D) Effect of 10 μM Y-27632 ROCK inhibitor on (C) barrier formation and (D) barrier function of mock control cells or PLXNA4 knockdown cells. Results are expressed as or as percentage of the average resistance of untreated control cells, respectively. Mean resistance ± S.E.M. of N = 3, *P < 0.05.

Figure 7

Downregulation of endothelial PLXNA4 under inflammatory conditions diminishes vascular integrity. Current hypothesis of the role of PLXNA4 in vascular integrity. Under homeostatic conditions PLXNA4 is expressed in endothelial cells resulting in a stable endothelial barrier. Inflammatory conditions reduce PLXNA4 expression resulting in rearranged cytoskeletal structures and cell-cell junctions. In turn this leads to decreased endothelial barrier function, impaired tube formation and increased vascular permeability to solutes and cells. This decrease in vascular integrity could contribute to atherogenesis by increasing deposition of (inflammatory) cells and lipids into the arterial wall.