Pro-inflammatory endothelial cell dysfunction is associated with intersectin-1s down-regulation

Abstract

Background: The response of lung microvascular endothelial cells (ECs) to lipopolysaccharide (LPS) is central to the pathogenesis of lung injury. It is dual in nature, with one facet that is pro-inflammatory and another that is cyto-protective. In previous work, overexpression of the anti-apoptotic Bcl-XL rescued ECs from apoptosis triggered by siRNA knockdown of intersectin-1s (ITSN-1s), a pro-survival protein crucial for ECs function. Here we further characterized the cyto-protective EC response to LPS and pro-inflammatory dysfunction.

Methods and results: Electron microscopy (EM) analyses of LPS-exposed ECs revealed an activated/dysfunctional phenotype, while a biotin assay for caveolae internalization followed by biochemical quantification indicated that LPS causes a 40% inhibition in biotin uptake compared to controls. Quantitative PCR and Western blotting were used to evaluate the mRNA and protein expression, respectively, for several regulatory proteins of intrinsic apoptosis, including ITSN-1s. The decrease in ITSN-1s mRNA and protein expression were countered by Bcl-XL and survivin upregulation, as well as Bim downregulation, events thought to protect ECs from impending apoptosis. Absence of apoptosis was confirmed by TUNEL and lack of cytochrome c (cyt c) efflux from mitochondria. Moreover, LPS exposure caused induction and activation of inducible nitric oxide synthase (iNOS) and a mitochondrial variant (mtNOS), as well as augmented mitochondrial NO production as measured by an oxidation oxyhemoglobin (oxyHb) assay applied on mitochondrial-enriched fractions prepared from LPS-exposed ECs. Interestingly, expression of myc-ITSN-1s rescued caveolae endocytosis and reversed induction of iNOS expression.

Conclusion: Our results suggest that ITSN-1s deficiency is relevant for the pro-inflammatory ECs dysfunction induced by LPS.

Figure 1

LPS exposure causes an activated/dysfunctional ECs phenotype (I). ECs exposed to 1 μg/ml LPS for 48 h show open IEJs, (a, b), increased accumulation of actin filaments at cell periphery (c, arrows), significant increase in the number of Weibel-Palade (wp) bodies (d) and Golgi (f, circled areas). An enlarged and widespread endoplasmic reticulum network and the presence of a dilated tubular system are shown in e; frequently, increase number of lysosomal (lys) units is observed (a, d, f). Panel g shows a segment of cultured lung microvascular ECs comprising increased number of mitochondrial units (m), most of them with abnormal morphology. Note also the close association between ER and mitochondria. Frequently, mitochondria undergo a fission process (g1). Bars: 100 nm (a, b, d, e, f, g); 50 nm (c);

Figure 2

LPS exposure causes an activated/dysfunctional ECs phenotype (II). Fragments of control ECs show a normal endothelial phenotype: Golgi stacks and vesicles as well as two mitochondrial units with well-organized cristae (a), a wide endoplasmic reticulum network is located within the two mitochondrial units. Note also a Weibel-Palade (wp) body in close proximity of the basolateral plasma membrane. Caveolae are open apically or basolaterally, apparently charging or discharging their load, (a1, arrows). In panel b, a fragment of an ECs subjected to 1 μg/ml LPS shows frequently large vacuoles (v1, v2), membranous rings, (arrows) and tubules (t), within cytosol or open to the cell surface (c, d), suggestive of deficient endocytic transport. Bars: 50 nm (a); 100 nm (a1); 70 nm (b); 40 nm (c, d).

Figure 3

Effects of 1 μg/ml LPS exposure causes significant inhibition of caveolae internalization. A. Control and LPS-treated cells were subjected to biotinylation of cell surface proteins as described under Materials and Methods, followed by internalization of biotinylated proteins for 30 min, at 37°C. The number of biotin molecules present in lysates of control and LPS-treated cells was quantified by ELISA using streptavidin-HRP Ab, in 3 different experiments. The results were normalized per milligram total protein, per 30 min. Ordinate, amounts of biotin molecules/well detected per well; abscissa, μg total protein/well as result of serial dilution of control and LPS-treated ECs lysates. B. Degree of inhibition of caveolae-mediated uptake in LPS-treated ECs by reference to controls. Bars, ±SE.

Figure 4

NeutrAvidin Alexa Fluor 594 staining demonstrates inhibition of caveolae internalization by 1 μg/ml LPS exposure. NeutrAvidin Alexa Fluor 594 staining of control ECs subjected to biotinylation of cell surface proteins followed by internalization assay, indicates a strong punctate pattern throughout the cytosol (c), with some accumulation in the perinuclear area. An enlarged ECs (boxed area in c) subjected to the internalization assay is shown in c1. ECs exposed to LPS and subjected to the internalization assay followed by neutrAvidin Alexa Fluor 594 staining (b, b1) show large fluorescent puncta within the intracellular space with no perinuclear accumulation. Bars: a, a1, b, b1 - 20 μm.

Figure 5

1 μg/ml LPS exposure causes endothelial barrier dysfunction. A. At the time point indicated (arrow), ECs grown on gold microelectrodes, were exposed to 1 μg/ml LPS, and TER was monitored over time. Data are expressed as means ± SD of 3 independent experiments; n = 3 per condition for each experiment. DNP-BSA concentrations detected in growth media collected from the lower chamber in control vs. LPS-treated ECs. Tracer amounts were determined by ELISA in 3 independent experiments, n = 3 per condition for each experiment. Ordinate, amounts of DNP-BSA detected per well; abscissa, μl growth medium per well. Degree of DNP-BSA leakage in LPS-treated ECs monolayer by reference to controls. Values are means ± SD ng DNP-BSA/0.1 ml growth medium/1 hour.

Figure 6

Induction of iNOS and its mitochondrial variant expression by LPS exposure. Total cell lysates (70 μg/lane) of control and LPS-treated ECs (24 h and 48 h) were subjected to SDS-PAGE, electrotransfer and immunoblotting with NOS2 Ab, known to recognize the iNOS and mtNOS [31]. A representative blot, documenting the increased expression of iNOS caused by LPS is shown in (A). Blots obtained from 3 different experiments performed under identical experimental conditions (70 μg total protein/lane, 1:1000 NOS2 pAb dilution and 30 sec ECL exposure time) were subjected to densitometric analysis (B), Bars, ±SE C. Control and LPS-treated ECs were subjected to cell fractionation to obtain an enriched-mitochondrial fraction; mitochondria were then lysed and analyzed by SDS-PAGE and Western blot using NOS2 Ab. D. mtNOS activity was assessed by exposing freshly prepared, unlysed mitochondrial fractions of control and LPS-treated ECs to oxyHb assay. The rate of NO production was measured by the change in optical density of the reaction solution as generated NO quickly oxidized oxyHb to metHb. The bar graph shows the difference in rate of mitochondrial NO formation between control and LPS-treated ECs.

Figure 7

LPS-induces ITSN-1s deficiency without affecting the EC survival. Control and LPS-treated ECs lysates were analyzed by Western blot (A) and qPCR (B) for ITSN-1s protein and mRNA expression. The ITSN-1s mRNA levels, relative to the internal control, cyclophilin, were evaluated in three separate experiments. C. TUNEL staining of ECs treated with 1 μg/ml of LPS for 48 hours show no evidence for apoptotic activity (c2). A positive control - siRNA ITSN-1s - transfected ECs is shown in c1. D. Quantitative assessment was obtained by counting the number of TUNEL-positive and TUNEL-negative cells in 25 high-powered fields and calculating the percentage of apoptotic cells amongst the total cell population. No significant difference was noted in the percentage of apoptotic cells between LPS-treated and control ECs (0.89% vs. 0.98%, p = .45). E. Mitochondrial fractions and post-mitochondrial supernatants prepared for LPS-treated cells were analyzed for cyt c reactivity.

Figure 8

Control and LPS-treated cell lysates (24 h, 48 h) were also analyzed by qPCR for mRNA expression of Bim, Bcl-X_L and survivin, relative to the internal control, cyclophilin, in three separate experiments. Representative results of one of these experiments are shown. Total protein concentration in control and LPS-treated cell lysates was determined by BCA, and 80 μg of total protein per sample was subjected to SDS-PAGE. Gels transferred to nitrocellulose membranes were then probed with Bcl-XL, Bim and survivin Abs. These trends were confirmed in at least three separate experiments for each protein.

Figure 9

Expression of myc-ITSN-1s in LPS-treated ECs restores endocytosis and inhibits both iNOS and mtNOS protein expression. A. Myc-ITSN-1s-transfected ECs, 48 h post-transfection, were exposed to 1 μg/ml LPS cells for 48 h and then subjected to biotin internalization. NeutrAvidin Alexa Fluor 594 staining revealed the rescued ability of ECs to internalized biotin (a2). An enlarged region of interest (boxed area in a2) is shown in a3. Control ECs subjected to biotin internalization and NeutrAvidin Alexa Fluor 594 are shown for comparison in a1. Bars: a1, a2, a3 - 20 μm. B. Degree of inhibition of caveolae-mediated uptake in myc-ITSN-1s transfected and LPS-treated ECs, by comparison to controls was evaluated by ELISA in 3 different experiments. Bars, ±SD. C. Total ECs lysates (70 μg total protein/lane) prepared from control, LPS-treated and ITSN-1s transfected and LPS-treated cells were analyzed by SDS-PAGE, electrotransfer to nitrocellulose membranes and Western blotting for NOS2 protein expression. Actin was used as loading control. D. Enriched-mitochondrial fractions of control, LPS-treated and myc-ITSN-1s transfected/LPS-treated cells were analyzed as above for mtNOS expression.