Cyclophilin A as a novel biphasic mediator of endothelial activation and dysfunction

Abstract

Inflammation-mediated endothelial cell (EC) dysfunction likely contributes to the pathogenesis of several vascular diseases including atherosclerosis. We found that stimulation of human umbilical vein ECs with lipopolysaccharide induced secretion of cyclophilin (CyPA) an intracellular protein belonging to the immunophilin family. We then found that when added exogenously CyPA has direct effects on ECs in vitro. At low concentrations (10 to 100 ng/ml) CyPA increased EC proliferation, migration, invasive capacity, and tubulogenesis. Gelatin zymography indicated increased secretion of active matrix metalloproteinase-2, a mediator of cell migration and angiogenesis. At high concentrations (eg, 2 microg/ml) CyPA had opposite effects, decreasing EC migration and viability, possibly in relation to induction of Toll-like receptor-4 expression, detected by immunocytochemistry and flow cytometry. In vivo CyPA expression was not detectable in the luminal ECs of normal mouse carotid arteries but was rapidly induced after systemic lipopolysaccharide injection. In an experimental mouse model of atherosclerosis, CyPA expression was detected in the ECs of neocapillaries of carotid artery lesions, supporting its association with pathological angiogenesis suggested by our in vitro results. In conclusion, we found that CyPA has a biphasic activity on ECs in vitro and is up-regulated in vivo in ECs under pathological states. Our results suggest that CyPA is a novel paracrine and autocrine modulator of EC functions in immune-mediated vascular disease.

Figure 1

In vitro treatment with LPS induces secretion of CyPA by HUVECs. A: CyPA (18 kd) was detected in the culture medium of HUVECs by Western blotting. LPS is an effective stimulator of CyPA secretion by ECs (6 hours). B: Expression of CyPA was stimulated by LPS (1 μg/ml) in serum-free media in a time-dependent manner, with maximal stimulation detected at 12 hours. C: Detection of CyPA in HUVECs by immunofluorescence (red fluorescent signal, CyPA; blue fluorescent signal, nuclei counterstained with Hoechst). Addition of monensin, an inhibitor of protein secretion, increases intracellular accumulation of CyPA. Scale bar, 5 μm.

Figure 2

CyPA stimulates endothelial wound healing (left and middle rows) and proliferation in vitro. Scratch wound assay (dotted line indicates wound edge) showed that the low CyPA concentration tested (10 ng/ml) enhanced migration of ECs (detected with anti-CD31, green fluorescence). Nuclei are counterstained with Hoechst (blue). ECs migrating past the wound edge were quantified after 24 hours. Middle and right rows: EC proliferation, detected using incorporation of BrdU (red), in subconfluent wounded or intact monolayers under different conditions (immunofluorescence). Nuclei of proliferating cells appear as pink. CyPA (10 ng/ml) increased basal level (no treatment) of cell proliferation to a level comparable to that induced by VEGF, used as a positive control. A biphasic effect was observed, with high CyPA concentrations (2 mg/ml) decreasing cell proliferation.

Figure 3

A: CyPA increases HUVEC invasiveness through a gelatin matrix (Transwell assay, VEGF 10 ng/ml). B: CyPA increases secretion and activation of gelatinolytic activity associated with matrix metalloproteinase-2 (sodium dodecyl sulfate-polyacrylamide gel electrophoresis gelatin zymography). No effects on secretion of the related gelatinase matrix metalloproteinase-9 were detected, suggesting matrix metalloproteinase-2 as the likely enhancer of endothelial invasiveness.

Figure 4

In vitro endothelial tube formation assay. Top: ECs highlighted by immunofluorescence using anti-CD31 (red), nuclei counterstained with Hoechst (blue). LPS and CyPA treatments of HUVECs appear to have similar capacity to induce formation of EC tubes on Matrigel. Right bottom inset in no treatment section illustrates the negative control for immunocytochemistry (no Ab, no primary antibody). Left bottom inset in CyPA 1 mg/ml section illustrates at higher magnification the appearance of the lumen of a tube. Bottom (phase contrast): tube formation assay performed on growth factor-reduced Matrigel. The total number of tubes was increased ∼10.5-fold at 10 ng/ml of CyPA compared with the nontreated control (P = 0.0021), but decreased at high concentrations (2 μg/ml of CyPA). Insets (live-dead assay) illustrate effects on cell viability likely contributing to the differences—note the increased percentage of dead cells (red fluorescence) compared to live cells (green). Scale bars, 50 μm.

Figure 5

Induction of TLR-4 expression in HUVECs by CyPA and LPS treatments. Top: Immunocytochemical detection of TLR-4 expression in HUVEC monolayers. High concentrations of CyPA (2 μg/ml) induce expression similar to LPS treatment (red fluorescence, TLR-4; blue fluorescence, nuclei). Bottom right inset illustrates detection of picnotic nuclei at these concentrations. No Ab section illustrates the negative control for immunocytochemistry, cells processed in the absence of primary antibody. Bottom graphs: Flow cytometric analysis of TLR-4 expression on HUVECs before and after stimulation with LPS or CyPA demonstrates induction of TLR-4 by high concentrations of CyPA, as indicated by the right shift of the black curve.

Figure 6

In vivo vascular expression of CyPA is associated with acute and chronic pathological conditions. Immunofluorescence analysis of mouse carotid artery cross-sections from healthy, untreated mice (baseline), mice injected with LPS (+LPS), or from mice with experimental atherosclerotic lesions (+ athero). Blue fluorescence, cell nuclei counterstained with Hoechst. Top: LPS injection induces acute endothelial activation, as indicated by induction of ICAM (green fluorescence) and VCAM-1 (red fluorescence) expression. Compare to nondetectable baseline expression (insets) of these molecules in normal endothelium. Second row: Luminal endothelial layer is detected using anti-C31 (green fluorescence). Low baseline (left) level of CyPA expression (red fluorescence) is detectable within the normal arterial wall in smooth muscle cells (inset, arrows) but not in luminal ECs. Right: Injection of LPS (+LPS) increased expression of CyPA throughout the wall and induced expression associated with the luminal endothelium (yellow fluorescence). Bottom left inset illustrates the higher magnification of the boxed area. Note the appearance of the yellow signal (arrows), indicating co-localization of CyPA (red) and luminal ECs (green). Bottom: Experimental atherosclerotic lesions induced in the carotid artery of ApoE^−/− mice (+athero) a model for a chronic pathological condition of arteries. Top: CyPA (red) is detected diffusely within the intimal lesion and the medial layer. Left: CyPA did not appear associated with the luminal endothelium (thick arrow) but was co-localized (small arrows) with the ECs (green) of neocapillaries that develop within these lesions. Bottom left inset is a higher magnification of the boxed area illustrating the yellow signal in capillaries. Top inset illustrates the negative control for immunohistochemistry, consecutive section processed in the absence of primary antibodies (no Ab). Right: Simultaneous detection of CyPA (red) and macrophages (green), using anti-MAC3, suggesting little overlap. Note some obviously negative macrophages (arrowheads). Bottom: At higher magnification the staining patterns of consecutive sections (the asterisk indicates the same tissue area) for CyPA (red) and endothelium (CD31, green). Although CyPA expression is not restricted to the neocapillaries, these are clearly positive for CyPA. Scale bars, 50 μm.