Introduction and objectives Pulmonary artery hypertension (PAH) is associated with inappropriate vascular remodelling and inflammation. Recent studies have shown that vascular cells express the transmembrane roundabout (Robo) proteins, Robo1 and Robo4, and that interaction with a secreted glycoprotein ligand, Slit2, controls cell migratory and inflammatory response. We hypothesise that Robo1, Robo4 and Slit2 are expressed on pulmonary artery (PA) endothelial cells (EC) and smooth muscle cells (SMC). We also hypothesise that Slit2 will modulate PAEC and PASMC migration and inflammatory mediator release.

Methods Real-time-PCR determination of Robo1, Robo4 and Slit2 expression and the house-keeping gene, GAPDH, in PAEC, PASMC and for comparison, human umbilical vein endothelial cells (HUVEC); following incubation with TNFα (10ng/ml) or Slit2N (10nM) for 2h. Enzyme-linked immunosorbent assay measurements of granulocyte-macrophage-colony stimulating factor (GM-CSF) in supernatants of HUVEC, PAEC or PASMC pre-treated (1h) with Slit2N, followed by TNFα (17h). Migration assays (PAEC or PASMC) towards serum-containing medium (0.05 and 0.02%, respectively), for 4h with/ without Slit2N.

Results Basal mRNA expression of Robo1, Robo4 and Slit2 was detected in PAEC, PASMC and HUVEC (n = 3–4). Slit2N (2h) significantly (p < 0.05, n = 3) decreased Robo4 and Slit2 mRNA expression, but not Robo1, by 35% in PAEC; and had no effect on HUVEC or PASMC. TNFα had no significant effects on Robo1, Robo4 or Slit2, regardless of cell type. Despite a small (23%), but significant (p < 0.05) reduction of GM-CSF release from TNFα-activated HUVEC (n = 7), no similar effects were seen in PAEC or PASMC (n = 3). Moreover, whilst PAEC or PASMC migration to serum-containing medium increased (2.7- and 5.3-fold, respectively), co-incubation with Slit2 had no significant effect.

Conclusion The novel discovery of Robo1, Robo4 and Slit2 mRNA in PAEC and PASMC; and that Slit2 down-regulated Robo4 and Slit2 in PAEC, but not PASMC/HUVEC, might suggest negative feedback on the Robo4-Slit2 axis unique to PAEC. That neither PAEC nor PASMC responded to Slit2 in functional assays could reflect limitations in experimental assays. However, down-regulation of Robo4-Slit2 in PAEC might also explain lack of effect on GM-CSF release, when compared with HUVEC. Further studies to better delineate the role of the Robo-Slit2 pathway in PAH are required.