Sphingolipids affect fibrinogen-induced caveolar transcytosis and cerebrovascular permeability

Abstract

Inflammation-induced vascular endothelial dysfunction can allow plasma proteins to cross the vascular wall, causing edema. Proteins may traverse the vascular wall through two main pathways, the paracellular and transcellular transport pathways. Paracellular transport involves changes in endothelial cell junction proteins, while transcellular transport involves caveolar transcytosis. Since both processes are associated with filamentous actin formation, the two pathways are interconnected. Therefore, it is difficult to differentiate the prevailing role of one or the other pathway during various pathologies causing an increase in vascular permeability. Using a newly developed dual-tracer probing method, we differentiated transcellular from paracellular transport during hyperfibrinogenemia (HFg), an increase in fibrinogen (Fg) content. Roles of cholesterol and sphingolipids in formation of functional caveolae were assessed using a cholesterol chelator, methyl-β-cyclodextrin, and the de novo sphingolipid synthesis inhibitor myriocin. Fg-induced formation of functional caveolae was defined by association and colocalization of Na+-K+-ATPase and plasmalemmal vesicle-associated protein-1 with use of Förster resonance energy transfer and total internal reflection fluorescence microscopy, respectively. HFg increased permeability of the endothelial cell layer mainly through the transcellular pathway. While MβCD blocked Fg-increased transcellular and paracellular transport, myriocin affected only transcellular transport. Less pial venular leakage of albumin was observed in myriocin-treated HFg mice. HFg induced greater formation of functional caveolae, as indicated by colocalization of Na+-K+-ATPase with plasmalemmal vesicle-associated protein-1 by Förster resonance energy transfer and total internal reflection fluorescence microscopy. Our results suggest that elevated blood levels of Fg alter cerebrovascular permeability mainly by affecting caveolae-mediated transcytosis through modulation of de novo sphingolipid synthesis.

Keywords: Förster resonance energy transfer microscopy; cholesterol; functional caveolae; protein leakage; total internal reflection fluorescence microscopy.

Fig. 1.

Fibrinogen (Fg)-induced permeability of mouse brain endothelial cells (MBECs). A and B: permeability of MBECs to Lucifer yellow (LY) and BSA tagged with Alexa Fluor 647 (BSA-647) in the presence of PBS in medium (control), 4 mg/ml fibrinogen (Fg4), 4 mg/ml Fg + 100 μM methyl-β-cyclodextrin (Fg4 + MβCD), or 100 μM MβCD. C and D: permeability of MBECs to LY and BSA-647 in the presence of PBS in medium (control), 4 mg/ml Fg (Fg4), 4 mg/ml Fg + 500 nM myriocin (Fg4 + myriocin), or 500 nM myriocin. Fluorescence intensity of each dye in samples collected from lower chambers of the Transwell system after 20, 40, 60, and 120 min was measured by a microplate reader (488-nm excitation and 520-nm emission for LY; 650-nm excitation and 668-nm emission for BSA-647). Results are expressed as ratio of fluorescence intensity of each dye in the lower chamber to fluorescence intensity of the respective dye in the original sample at the end of the experiment. Values are means ± SE; n = 4. *P < 0.05 vs. control. †P < 0.05 vs. Fg4 + MβCD or Fg4 + myriocin.

Fig. 2.

Cerebrovascular permeability to macromolecules in hyperfibrinogenic (HFg) mice. Pial venular permeability to FITC-BSA was assessed in HFg mice treated with myriocin (0.5 mg·kg⁻¹·day⁻¹) or PBS for 3 days. Fluorescence intensity changes in an area of interest adjacent to the venular segment were measured as described in methods. Venular permeability was assessed by changes in the ratio of fluorescence intensity measured in the interstitium adjacent to the venule to that inside the vessel. Values (means ± SE) are shown as percent change in fluorescence compared with PBS alone (control); n = 4. *P < 0.05 vs. HFg + PBS.

Fig. 3.

Fg-induced increase in sphingolipid synthesis in MBECs. MBECs were treated with PBS in medium (control) or 2 or 4 mg/ml Fg for 2 h (A) or 24 h (B). Levels of sphingolipids were assessed in cell culture medium by LC-MS/MS. Fg markedly increased levels of ceramide and sphingomyelin species in MBECs. Sphingolipids with different length and degree of saturation of fatty acids were measured. C18:2, fatty acid with 18 carbons and 2 double bonds; DH, dihydro; Sph, sphingosine; S1P, sphingosine-1-phosphate. Values are means ± SE; n = 3. *P < 0.05 vs. control. †P < 0.05 vs. 2 mg/ml Fg.

Fig. 4.

Comparison of sphingomyelin (SPM), ceramide (Cer), and glucosylceramide (GlcCer) content in plasma of hyperfibrinogenic (HFg) and wild-type (WT) mice. Plasma concentration of SPM, Cer, and GlcCer sphingolipids was measured by LC-MS/MS. Content of these lipids was higher in HFg than WT mice. Values are means ± SE; n = 4. *P < 0.05 vs. WT.

Fig. 5.

Fg-induced Na⁺-K⁺-ATPase activation in MBECs. Ouabain-sensitive Na⁺-K⁺-ATPase activity was measured as an indicator of active transport across the membrane. Cells treated with PBS in medium were used as a control group. Values are means ± SE; n = 3. *P < 0.05 vs. control.

Fig. 6.

Effect of Cer, GlcCer, SPM, and Fg on formation of functional caveolae in MBECs. Formation of caveolae was determined by Förster resonance energy transfer (FRET) and total internal reflection fluorescence (TIRF) microscopy. Cells were transfected with green fluorescence protein (GFP)-labeled plasmalemmal vesicle-associated protein-1 (PV-1) and/or mCherry-labeled Na⁺-K⁺-ATPase, and live cells were imaged as described in methods. A, B, and D: representative images from 3 individual experiments for epifluorescence (A), FRET (B), and TIRF (D). Merged images in B are in pseudocolor (gated to mCherry acceptor levels). Color scale shows reference spectrum: blue indicates no association (FRET and TIRF), and red indicates association (FRET and TIRF). C: data for 3-channel FRET efficiency after photobleaching from 3 individual experiments. In each experiment, FRET efficiency from 30–50 cells was calculated, averaged, and considered as 1 experimental value. E and F: expression of Na⁺-K⁺-ATPase and caveolin-1 in the plasma membrane was determined by TIRF, and the number of caveolae was counted as individual GFP (E) and mCherry (F) particles using ImageJ software. In each experiment, values from 30–50 cells were averaged and considered as 1 experimental value. Values are means ± SE; n = 3. *P < 0.05 vs. control.