Slit2-Robo4 receptor responses inhibit ANDV directed permeability of human lung microvascular endothelial cells

Abstract

Hantaviruses nonlytically infect human endothelial cells (ECs) and cause edematous and hemorrhagic diseases. Andes virus (ANDV) causes hantavirus pulmonary syndrome (HPS), and Hantaan virus (HTNV) causes hemorrhagic fever with renal syndrome (HFRS). Hantaviruses enhance vascular endothelial growth factor directed EC permeability resulting in the disassembly of inter-endothelial cell adherens junctions (AJs). Recent studies demonstrate that Slit2 binding to Robo1/Robo4 receptors on ECs has opposing effects on AJ disassembly and vascular fluid barrier functions. Here we demonstrate that Slit2 inhibits ANDV and HTNV induced permeability and AJ disassembly of pulmonary microvascular ECs (PMECs) by interactions with Robo4. In contrast, Slit2 had no effect on the permeability of ANDV infected human umbilical vein ECs (HUVECs). Analysis of Robo1/Robo4 expression determined that PMECs express Robo4, but not Robo1, while HUVECs expressed both Robo4 and Robo1 receptors. SiRNA knockdown of Robo4 in PMECs prevented Slit2 inhibition of ANDV induced permeability demonstrating that Robo4 receptors determine PMEC responsiveness to Slit2. Collectively, this data demonstrates a selective role for Slit2/Robo4 responses within PMECs that inhibits ANDV induced permeability and AJ disassembly. These findings suggest Slit2s utility as a potential HPS therapeutic that stabilizes the pulmonary endothelium and antagonizes ANDV induced pulmonary edema.

Keywords: Hantavirus; Human endothelial cells; Inhibition; Permeability; Slit2.

Figure 1

Permeability and VE-cadherin internalization assays were performed 3 days post-infection as described (Gavrilovskaya et al., 2008). (A) Pulmonary Microvascular ECs (PMECs/Human microvascular ECs-lung-Cambrex) were mock, ANDV (CHI-7913), Hantaan virus (HTNV 76–118) or TULV (Tula/Moravia/MA 5302V/94) infected in BSL3 at an MOI of 0.5. In lower panel infected cells were immunoperoxidase stained for the hantavirus nucleocapsid protein (Geimonen et al., 2002). Three days p.i. monolayers were treated as indicated with VEGF (100 ng/ml) or Slit2 (N-terminal residues 1–1093 of human Slit2 homolog, Slit2-N; PeproTech, >98% pure) (10 nM) prior to assessing monolayer permeability to FITC-dextran (40,000; 0.5 mg/ml) as previously described. (B) PMECs were infected for 3 days and subsequently analysis of ANDV induced VE-cadherin internalization was performed as previously described on PMECs with or without Slit2 (Gavard and Gutkind, 2006; Gorbunova et al., 2010). Experiments were performed four times and error bars represent s.e.m.

Figure 2

PMECs or HYVECs were infected for three days at an MOI of 0.5 and subsequently assayed for the effect of Slit2 on ANDV permeability in (A) human umbilical vein ECs (HUVECs-Cambrex); or (B) PMECs 15, 30 and 60 minutes post-VEGF addition, was performed as in Fig. 1 and previously described (Gavrilovskaya et al., 2008).

Figure 3

Role of Robo1/4 in Slit2 regulation of PMECs. HUVECs (Pepini et al., 2010) (3A) or PMECs (3B) were infected as above and analyzed by qRT-PCR for Robo1 and Robo4 mRNAs relative to GAPDH mRNA levels (Pepini et al., 2010). (3C) PMECs were transfected with negative-control (si-NEG2) or Robo4-specific siRNA and assayed by qRT-PCR for Robo4 mRNA levels as in 3B (Pepini et al., 2010). (3D) PMECs were ANDV infected as in Fig 1, Robo4 or control siRNA transfected 1 day post-infection, as in 3C, and assayed for PMEC permeability in response to VEGF as in Fig 1.

Figure 4

PMECs were ANDV infected and assayed for changes in permeability as in Fig. 1 in the presence or absence of Slit2 (effective range 5–50nM), dasatinib or FTY720 (Gavrilovskaya et al., 2008; Gorbunova et al., 2010; Gorbunova et al., 2011).