and Endotoxin-Induced Lung Injury - ATS Journals

Role of Sphingomyelin Synthesis in Pulmonary Endothelial Cell Cytoskeletal Activation and Endotoxin-Induced Lung Injury Fatima Anjum1, Keval Joshi1, Natalia Grinkina1, Satish Gowda1, Michael Cutaia2, and Raj Wadgaonkar1,2 1

Department of Pulmonary and Critical Care Medicine, State University of New York Downstate Medical Center, Brooklyn, New York; and 2Molecular Medicine Laboratory, Veterans Administration Medical Center, Brooklyn, New York

Sphingomyelin (SM), a major sphingolipid in the lipid raft microdomains of the cell membrane, is synthesized by plasma membrane– bound sphingomyelin synthase 2 (SMS2). SMS2 is required for the maintenance of plasma membrane microdomain fluidity and receptormediated responses to inflammation in macrophages. However, the exact mechanism of SMS2 activation in endothelial barrier disruption and lung injury is not fully understood. To define the role of SMS activation in lung injury, we hypothesized that the inhibition of SM synthesis may provide protection against acute lung injury (ALI) by preserving endothelial barrier function. Using SMS2-silencing RNA (siRNA) treatment in human pulmonary endothelial cells (HPAECs) and tricyclodecan-9-yl-xanthogenate (D609), a competitive inhibitor of SMS, and phosphatidylcholine-specific phospholipase C in a murine model of bacterial LPS injury, we studied the role of sphingomyelin synthesis in ALI. Results show that pretreating mice with D609 significantly attenuated LPS-induced lung injury, as measured by a significant decrease in wet to dry ratio, bronchoalveolar lavage fluid cell and protein counts, and myeloperoxidase activity in lung tissue. Similarly, LPS-induced endothelial barrier disruption was significantly reduced in HPAECs pretreated with D609 or SMS2 siRNA, as demonstrated by an increase in paracellular integrity on an FITCdextran assay, by the inhibition of LPS-induced stress fibers, and by the formation of cortical actin rings and lamellipodia at the periphery. These results indicate that D609 attenuates LPS-mediated endothelial barrier dysfunction and lung injury in mice through inhibition of SMS, suggesting a novel and essential role of SMS inhibition in modulating endothelial barrier integrity via actin cytoskeletal activation, with a potential therapeutic role in ALI. Keywords: sphingomyelin synthase; endothelial barrier dysfunction; cytoskeletal rearrangement

Acute lung injury (ALI) is a diffuse inflammatory disorder characterized by the filling of alveoli with inflammatory cells and a protein-rich exudate that impairs pulmonary gas exchange, leading to hypoxemic respiratory failure. A range of local and systemic insults can result in ALI, including sepsis, pneumonia, aspiration, trauma, and other inflammatory disorders, such as acute pancreatitis. Recent estimates indicate approximately 190,000 cases per year of ALI in the United States, with an associated 74,500 deaths per year (1).

(Received in original form November 8, 2010 and in final form February 8, 2012) This work was supported by a Veterans Administration Merit Grant (R.W.) and a New York State ECRIP Grant (R.W.). Correspondence and requests for reprints should be addressed to Raj Wadgaonkar, Ph.D., Molecular Medicine Laboratory, Veterans Administration Medical Center and State University of New York Downstate Medical Center, Brooklyn, NY 11203. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org Am J Respir Cell Mol Biol Vol 47, Iss. 1, pp 94–103, Jul 2012 Copyright ª 2012 by the American Thoracic Society Originally Published in Press as DOI: 10.1165/rcmb.2010-0458OC on February 23, 2012 Internet address: www.atsjournals.org

Disruption of the endothelial barrier integrity plays a fundamental role in the pathogenesis of ALI. It is characterized by remarkable changes in the endothelial cell (EC) cytoskeleton, an increase in centripetal forces within cells causing cell contraction, disruption of intercellular adhesion junctions, and the formation of paracellular gaps, leading to an influx of protein-rich fluid and inflammatory cells into the alveolar spaces (2). A rapid decrease in lung compliance and impaired gas exchange warrant the use of mechanical ventilation in most patients. Unfortunately, mechanical ventilation further contributes to the disease process through the release of cytokines (3), whereas cyclic stretching in endothelial cells leads to the activation of cytoskeletal changes, cell contraction, and the disruption of cell junctions, promoting vascular leakage, as demonstrated in vitro and in isolated perfused rat lung models (4, 5). Bacterial endotoxins such as LPS cause further damage to the lungs via the release of cytokines and the activation of actin cytoskeletal remodeling, leading to endothelial barrier dysfunction and ALI (6, 7). Recent studies showed that sphingolipids, and specifically sphingosine-1–phosphate (S1P), attenuate the inflammatory lung injury induced by intratracheal LPS and mechanical ventilation in murine models of lung injury (8, 9). These findings, subsequently confirmed by the overexpression of sphingosine kinase 1 (SphK1) in SphK12/2 mice (10), demonstrated direct alterations in the endothelial cytoskeleton by increasing the cortical actin content while decreasing actin stress fibers. This results in an increase in endothelial resistance, as measured by using an electric cellsubstrate impedance sensing apparatus. In addition, intravenous S1P was shown to reduce inflammatory lung injury when administered as rescue therapy in a canine model of LPS-induced ALI, largely because of the reduction in endothelial permeability rather than cytokine modulation (11). These studies suggest that natural molecules such as bioactive lipids could provide effective therapy in attenuating vascular injury. Sphingomyelin synthase is an important enzyme involved in the synthesis of sphingomyelin, with two known mammalian isoforms, sphingomyelin synthase (SMS) 1 and 2. SMS1 is mainly located on cis-Golgi and medial-Golgi, whereas SMS2 is found in plasma membranes (12). They use ceramide and phosphatidylcholine as substrates to produce sphingomyelin (SM) and diacyl glycerol (DAG), thereby regulating SM, DAG, phosphatidylcholine (PC), and ceramide concentrations. Of the two isoforms, SMS2 is directly linked to cell membrane lipid messengers that play a role in cell survival and apoptosis (13, 14). Tricyclodecan-9-yl-xanthogenate (D609) has been accepted as a selective inhibitor of PC-specific phospholipase C (PLC) (15, 16). However, the mechanism of D609-dependent PC-PLC inhibition remains unclear. It has been used to establish the functional coupling of PC-PLC with sphingomyelinase, and to establish the coupling of PC hydrolysis to the activation of Raf-1 protein kinase. Some studies indicated that D609 can block sphingomyelin synthesis by inhibiting sphingomyelin synthase activity (13, 17). Recently, D609 was suggested to prevent platelet-activating factor (PAF)mediated lung edema by inhibiting the acid sphingomyelinase

Anjum, Joshi, Grinkina, et al.: Inhibition of Sphingomyelin Synthase

pathway (18). However, the exact mechanism by which pulmonary edema is prevented is not fully understood. In the present study, we sought to understand the pathway through which D609 can prevent ALI, and explore the role of SMS2 in endothelial barrier enhancement by performing experiments both in vivo and in vitro. Our results demonstrate that pretreatment with D609 in endothelial cells and a murine model of LPS-induced lung injury causes significant barrier enhancement, as demonstrated by the peripheral bundling of F-actin with augmented adherens junctions’ stability in human pulmonary artery endothelial cells (HPAECs), with a significant reduction in vascular leakage and degree of inflammation in mice. This result is analogous to the effect seen with other barrier-enhancing agonists such as S1P. Similar changes were observed in endothelial cells after the use of SMS2-silencing RNA (siRNA), suggesting the involvement of SMS2 and sphingomyelin synthesis inhibition in endothelial cell– barrier enhancement, with a potential therapeutic role in ALI.

MATERIALS AND METHODS

95

(1) Bronchoalveolar lavage (BAL) fluid protein quantification and BAL fluid cell count: BAL cell counts were determined as described previously (10, 19). (2) Myeloperoxidase activity (MPO) assay: the infiltration of polymorphonuclear granulocytes (PMNs) into lung tissue was assessed by measuring the activity of myeloperoxidase enzyme, a marker for PMNs, by kinetic reading for 30 minutes (10, 32). (3) Histologic examination: To characterize lung morphology, immediately after the mice were killed, the left lungs from mice in each experimental group were inflated to 20 cm H2O with 0.2% of low melting agarose through an intratracheal catheter (10, 19). Statistical Analysis Data analysis was performed using unpaired two-way ANOVA or the nonparametric Wilcoxon test when appropriate. Bonferroni’s multiple comparison of groups was used to test the effects of treatment. Data are presented as means 6 SD. P ¼ 0.05 was considered significant.

Cell Culture HPAECs were purchased from BRL Laboratories, and were grown in special medium containing growth supplements (M200 1 LSGSFBS), as described elsewhere (10).

Permeability Assay Measurements of FITC-dextran clearance to assess changes in endothelial permeability were performed with the In Vitro Vascular Permeability Assay Kit (ECM640; Millipore, Billerica, MA), and the methods were optimized for detecting barrier enhancement. ECs were plated on collagen-coated inserts at 100,000 cells/insert and grown to confluence. Cells were then treated with D609 in the luminal chamber only. After 1 hour of incubation, LPS was added with FITC-dextran (1:20 dilution, 40 kD; Sigma, St. Louis, MO) to the luminal chambers. After 2–24 hours of incubation, the inserts were removed, and the medium from the bottom plate was collected. Fluorescence detection was determined on a CytoFluor Multiwell Plate Reader Series 4000 (Applied Biosystems, Foster City, CA), with 485-nm and 530-nm filter settings.

SMS2 siRNA Treatment HPAECs were grown to 80% confluence in growth medium (M200, 10% FBS, and 2% LSGS) 100 nM (final concentration) SMS2 siRNA (P/N, SI00636636; Qiagen) or control siRNA (Qiagen) was complexed with oligofectamine in Opti-MEM, according to the manufacturer’s instructions. The resulting complex was added to HPAECs in Opti-MEM for 48 hours.

SMS Activity Assay Total SMS1 and SMS2 activity was measured by an SMS activity assay, as described elsewhere (13, 19). D609-treated HPAEC extracts or lung tissue was homogenized in a buffer containing 50 mM Tris $ HCl, 1 mM EDTA, 5% sucrose, and a cocktail of protease inhibitors (Sigma). Lipids were extracted in chloroform–methanol (2:1), dried overnight, and separated by thin-layer chromatography in a buffer containing methanol–chloroform–ammonium hydroxide (14:64:1).

Calcium Measurements HPAECs were grown to confluence on coverslips and preloaded with 5 mM Rhod-2–AM (Molecular Probes, Eugene, OR) and Fluo-4 for 30 minutes. Fluorescence was measured with an Aminco-Bowman Series 2 luminescence spectrometer at excitation wavelengths of 552 and 578 nm and emission wavelengths of 494 and 516 nm.

RESULTS D609 Attenuates LPS-Induced Murine Lung Inflammation and Injury

C57BL/6J mice were pretreated with intravenous D609 or vehicle for 10 minutes, and then exposed to LPS 2.5 mg/kg intratracheally for 8 hours. The animals were then killed, and various parameters were used to evaluate lung injury, as described in MATERIALS AND METHODS. As shown in Figure 1A, MPO was significantly decreased in mice pretreated with D609 (35%, P , 0.05), compared with the untreated group, indicating less neutrophilic infiltration and inflammation in this group. Moreover, the inflammatory cell count in BAL fluid, as shown in Figure 1B, was also reduced in the D609pretreated group versus the untreated group (25%, P , 0.05). These findings were confirmed via lung histology. As shown in Figure 1C, D609-pretreated mice demonstrated a dramatic decrease in inflammation after LPS exposure, showing less pulmonary edema, neutrophilic infiltration, and alveolar hemorrhage compared with LPS-pretreated mice (see also Figures E4 and E6 in the online supplement). These findings suggest that D609 significantly attenuates LPS-induced lung injury, achieving a level of inflammation comparable to that observed in uninjured lungs. Effect of D609 on LPS-Induced Lung Edema and Vascular Leakage

Because D609 pretreatment led to a significant reduction in lung inflammation, we looked for its effects on pulmonary edema and vascular leakage by measuring total lung weight, wet/dry ratio, BAL protein content, and Evans blue albumin (EBA) extravasation from the vascular space into surrounding lung tissue. As shown in Figure 2, D609-treated mice, when exposed to LPS, had significantly lower lung wet weights (Figure 2A) and wet/ dry (Figure 2B) ratios (25%, P , 0.05) compared with vehicletreated mice. Similarly, protein extravasation from the vascular space into surrounding lung tissue, as measured by BAL protein content and EBA leakage, was also decreased in D609pretreated mice, indicating less vascular leakage in this group (Figures 2C and 2D). These findings suggest that D609 could be enhancing endothelial barrier properties.

Assessment of Lung Injury

D609 Enhances Endothelial Cell Barrier Properties in an In Vitro Model

Lung injury was induced by exposure to 2.5 mg/kg LPS intratracheally. After 8 hours of LPS treatment, lung injury was assessed as described previously (19)

To validate our hypothesis that D609 enhances endothelial barrier function, we performed an in vitro vascular permeability experiment

96

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 47 2012 Figure 1. Effects of tricyclodecan-9-yl-xanthogenate (D609) on LPS-induced murine lung inflammation and injury. C57BL/6J mice were pretreated with intravenous D609 or vehicle for 10 minutes, and were then exposed to LPS (2.5 mg/kg) by intratracheal instillation for 8 hours. Lung injury was assessed using several parameters. (A) Myeloperoxidase (MPO) activity was measured as described in MATERIALS AND METHODS, and was compared in LPS-treated and D609-treated animals with control mice. MPO activity was significantly decreased in mice pretreated with D609 (35%, P , 0.05) compared with the untreated group, indicating less neutrophilic infiltration and inflammation in this group. (B) The inflammatory cell count in bronchoalveolar lavage (BAL) fluid was also reduced in the D609-pretreated group versus the untreated group (25%, P , 0.05). Statistical significance was calculated according to one-way ANOVA. For each set of conditions, six mice were tested. The standard deviation was determined for each set of experiments. Values that were significantly different between the LPS-treated groups at (P , 0.05) are indicated by asterisks, and values that were significantly different after D609 treatment are indicated by double asterisks. (C) Hematoxylin and eosin staining of lung tissue was performed after intratracheal LPS administration and D609 treatment for 8 hours. For each condition +/2 LPS (c and d) and +/2 D609 treatment (b and a) in two wild-type mice lungs, at least six fields were examined from four slices of the lungs per slide. LPS treatment produced prominent neutrophil infiltration and an occasional alveolar hemorrhage (arrow). (c) LPS-treated wild-type (WT) murine lung tissue at 320. (d) D609-treated mice further challenged with LPS.

to measure the effects of D609 on FITC-dextran clearance. ECs were grown on coculture inserts to form a confluent monolayer. Cells were then treated with control (media only) or D609 (in media). After 1 hour of incubation, cells were treated with LPS, and FITC tracking dye was added to the luminal chambers. Fluorescence

was measured at 4 hours and 24 hours after treatment. D609 reduced FITC-dextran clearance from 1,724.4 6 35.2 relative fluorescence units (RFUs) of control level to 1,499.5 6 20.6 RFUs at 4 hours after challenge, and from 1,713.8 6 42.9 RFUs of control level to 1,361.7 6 14.9 RFUs at 24 hours after challenge (Figure 3A), Figure 2. Effects of D609 on LPS-induced lung edema and vascular leakage. Pulmonary edema and vascular leakage were measured in terms of total lung weight, wet/dry ratio, BAL protein content, and Evans blue albumin (EBA) extravasations from the vascular space into surrounding lung tissue. (A and B) D609treated mice, when exposed to LPS, demonstrated a significantly lower lung wet weight/ total body (TB) weight ratio (25%, **P , 0.05) compared with LPS-treated mice. (C and D) EBA leakage and BAL protein content were also decreased in D609-pretreated mice, indicating less vascular leakage in this group (n ¼ 6 mice). Values that were significantly different between the LPS-treated and control groups at P , 0.05 are indicated by asterisks, and values that were significantly different after D609 treatment are indicated by double asterisks.


97

suggesting that D609 enhances endothelial barrier function by reducing permeability in a dose-dependent fashion (a more pronounced effect with 50 mm than with 25 mm), and the effect is sustained for at least 24 hours. In addition, D609 attenuated the LPS-induced increase in permeability at both 4 hours and 24 hours after challenge (from 2,058.4 6 23.8 RFUs of LPS to 1,346.0 6 16.9 RFUs of D609-pretreated cells with LPS exposure). To exclude D609-mediated interference with tracking dye fluorescence (a potential artifact), we used a cell-free system with the measurement of the fluorescence of control media and D609 in media with and without FITC-dextran. Without the addition of FITC, the fluorescence reading for control media was 2.8 6 0.7 RFUs, and for D609, it was 2.6 6 0.4 RFUs. In the presence of FITC, the fluorescence reading for control media was 4,108.3 6 89.2 RFUs, and for D609, it was 4,170.5 6 95.3 RFUs (Figure 3B), indicating that D609 does not cause autofluorescence or interfere with the fluorescence of the tracking dye. These findings suggest that D609 acts as potent endothelial barrier enhancing agent. D609 Induces the Stability of Cell Junctions and the Formation of Cortical Actin Rings

Actin cytoskeletal and cell-to-cell adherens junction stability plays a vital role in endothelial barrier regulation and integrity. Because most barrier-enhancing agonists are associated with changes in F-actin and myosin light chain (MLC) phosphorylation, we sought to explore the enhanced effect of D609 on endothelial barrier integrity by immune-fluorescence studies of the cell cytoskeleton. HPAECs were treated with D609, LPS, and combined D609 and LPS in various doses, and were fixed and stained at different time intervals to visualize F-actin and MLC phosphorylation (Figures 4 and 6). In contrast to control cells (Figure 4Aa), HPAECs treated with D609 showed significant alterations in their actin cytoskeleton (Figure 4Ab). Cells treated with LPS demonstrated significant actin stress fiber formation (Figure 4Ac), whereas cells pretreated with D609 showed an inhibition of LPSinduced central stress fiber formation (Figure 4Ad). D609 also led to the formation of well-defined cortical actin rings, with the bundling of F-actin and ruffling of lamellipodia at the periphery of cells upon exposure to LPS (Figures E1D–E1F), suggesting that less centripetal contractile force, as produced by LPS-induced stress fibers, and more cell-to-cell and cell-tomatrix adhesive forces led to endothelial mechanical stability. Similar effects were evident in another endothelial cell line, namely, EA hy926 cells. Inhibition of the effects of LPS by D609 at a 50-mM concentration was prominent in these cells (Figure E2). To understand further the effects of D609 on SMS activity, HPAECs were treated with two different concentrations of D609 (25 and 50 mM), and SMS activity was measured by quantitating SM synthesis. After 1 hour of D609 exposure to HPAECs, total SMS activity was significantly attenuated by D609 treatment, suggesting the direct correlation of inhibitory action of D609 on SMS activity (Figure 4B). Effect of SMS2 Depletion through Silencing RNA on Cytoskeletal Rearrangement

Previous data from our laboratory showed that SMS2 knockout (SMS22/2) mice are protected against LPS-induced ALI, indicating that SMS2 inhibition plays a protective role in ALI (19). To investigate further the mechanism of action for D609 with the similarity of SMS2 inhibition in cytoskeletal activation, we used SMS2-silencing RNA to inhibit its gene expression in HPAECs. Using 100 mM siRNA treatment, within 48 hours we observed almost 90% inhibition in SM synthesis, suggesting SMS2 as a major contributor for SM synthesis in HPAECs (Figure 5A). As shown in Figure 3, in addition to the effect of D609 treatment on barrier enhancement, we tested the effect of SMS2

Figure 3. D609 induces sustained and dose-dependent endothelial barrier enhancement. A confluent human pulmonary artery endothelial cell (HPAEC) monolayer was treated with D609 for 1 hour, followed by the addition of LPS (1 mg/ml) and FITC-dextran tracking dye to the media. Inserts were removed to stop the experiment after 4 hours and 24 hours after challenge, and the medium from the lower well was sampled for fluorescence. (A) The data show that D609 significantly decreased FITCdextran clearance at 4 hours and 24 hours after challenge in a dosedependent fashion. In addition, D609 attenuated the LPS-induced increase in endothelial permeability. Values that were significantly different between the LPS-treated and control groups at P , 0.05 are indicated by asterisks, and values that were significantly different after D609 treatment are indicated by double asterisks. (B) Fluorescence was detected neither in media nor in D609 itself, whereas the addition of FITC-dextran significantly enhanced measurements of fluorescence, an effect that was similar in both groups, suggesting that D609 did not interfere with fluorescence measurements in the endothelial permeability assay.

inhibition on barrier function. FITC dextran clearance was measured in silencing RNA-transfected cells at 4 hours and 24 hours of LPS treatment, as shown in Figure 5B. Compared with control siRNA-treated cells, SMS2 siRNA-transfected cells showed clear differences in FITC-dextran diffusion at both the 4-hour and 24-hour time periods. This result suggested an essential role of SM synthesis or SMS2 activation in the regulation of endothelial barrier function. Next, we tested the effect of SMS2 silencing on LPS-induced cytoskeletal rearrangement. At baseline, cells treated with SMS2 siRNA did not show any significant changes in actin cytoskeleton, as depicted in Figure 5C. Interestingly, treatment of the same cells with LPS exerted a remarkable effect on actin cytoskeletal rearrangement, compared with the control group (Figures 5Cb and 5Cc). In the control group, central stress fibers formed after exposure to LPS. However, SMS2 siRNAtreated cells demonstrated the bundling of F-actin at the periphery and enhanced cortical actin ring formation upon LPS exposure, an effect similar to that seen with D609 treatment. These changes became more pronounced after 30 and 60 minutes of LPS exposure (Figures 5Cd–5Cf). These findings suggest that SMS2 inhibition plays a regulatory role in endothelial barrier integrity, and is the probable pathway for the effect of D609 on cytoskeletal rearrangement.

98

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 47 2012

Figure 4. D609 induces actin cytoskeletal rearrangement and effects on the activity of sphingomyelin synthase–2 (SMS). HPAECs grown on gelatinized coverslips were incubated with LPS (200 ng/ml, 30 minutes), D609 (50 mm/ml, 1 hour), and combined D609 (50 mm/ml, 1 hour) and LPS for 10 minutes. (A) Cells were then fixed and stained to visualize F-actin. (a) Untreated cells show no stress fiber formation. (b) In D609-pretreated cells, the mild bundling of F-actin at the cell periphery was observed (arrow). (c) Significant increase in LPS-induced central actin stress fiber formation. (d) D609-treated cells were further exposed to LPS for 10 minutes. An inhibition of prominent central actin stress fibers and formation of cortical actin ring (arrow) at the cell periphery was observed. (B) Cells treated with D609 (50 and 25 mm/ml, 1 hour) were harvested for an SMS activity assay and extracted in assay buffer, as described in MATERIALS AND METHODS. Total lipids were extracted from cell lysates and separated on thinlayer chromatography plates. Fold increase in SMS activity, sphingomyelin (SM) synthesis. Cr, ceramide. Statistical significance was calculated as described in MATERIALS AND METHODS.

The effect of D609 on MLC phosphorylation is also shown in Figure 6. Treatment with D609 alone did not lead to a significant induction of MLC phosphorylation. Cells exposed to LPS alone demonstrated MLC phosphorylation prominently along stress fibers (Figure 6B). However, with SMS2 siRNA cell treatment, phosphorylated MLC redistributed from central stress fibers to the periphery in a prominent MLC-containing cortical ring (Figures 6E and 6F) with actin colocalization, suggesting a potential mechanism for structural stability for extending lamellipodia, thereby increasing cell–cell junctional interactions. These findings correlate well with the significant endothelial barrier enhancement by SMS2 inhibition or D609 treatment already observed with the FITC dextran assay.

culture plates, we loaded cells with the fluorescent calcium indicator dyes Rhod-2–AM and Fluo-4 (5 mM, 30 min) and then measured the fluorescence, using a luminescence spectrometer at excitation wavelengths of 552 and 578 nm and emission wavelengths of 494 and 516 nm. The addition of D609 did not cause an intracellular calcium spike (Figures 7A and 7B), whereas LPS, like thrombin, a known trigger of intracellular calcium signaling, produced a strong spike in cytosolic calcium. Interestingly, pretreatment with D609, followed by sequential stimulation with LPS, inhibited the LPS-induced intracellular calcium spike, an effect that correlates with the ability of D609 to inhibit LPS-induced stress fiber formation and barrier disruption.

D609 Inhibits the LPS-Induced Intracellular Calcium Spike

Failure of Sphingomyelinase Inhibition to Induce Cytoskeletal Activation

Variations in intracellular calcium concentrations play an important role in the regulation of endothelial barrier function. To explore the effect of D609 on calcium signaling in HPAECs grown on six-well

Earlier studies indicated that D609 interferes with sphingomyelin metabolism. However, it remains unclear whether D609 is a specific inhibitor for sphingomyelinases or SMS enzymes. Using imipramine,


99 Figure 5. Effects of SMS2 depletion on endothelial cell permeability and actin cytoskeletal rearrangement. HPAECs grown on coverslips were treated with SMS2 silencing RNA (siRNA) or control siRNA for 48 hours, as described in MATERIALS AND METHODS. (A) Inhibition of sphingomyelin (SM) synthesis after siRNA treatment. HPAECs transfected with SMS2 siRNA were harvested after 48 hours, and SM synthesis was measured from total lipids extracted by thin-layer chromatography. (B) SMS2 inhibition and barrier enhancement. Confluent HPAEC monolayer was transfected with control and SMS2 siRNA exposed to LPS (1 mg/ml) and FITCdextran tracking dye in the media. Inserts were removed to stop the experiment, 4 hours and 24 hours after the challenge, and medium from the lower well was sampled for fluorescence. *LPS induced significant increase in relative fluorescence and permeability. **SMS2 siRNA– transfected cells showed significant attenuation in LPS-induced permeability. (C) Cells were then treated with LPS for 15 and 30 minutes. Interestingly, the treatment of ECs with SMS2 siRNA prevented LPS-induced stress fiber formation (c), as opposed to control siRNA-treated cells (a and b). (d) Changes become more prominent after SMS2 siRNA treatment with formation of cortical actin rings (large arrow in e) and the ruffling of lamellipodia (small arrow in e) at the periphery after LPS exposure for 15 minutes (e) and (f) cortical actin ring was prominent up to 30 minutes with the attenuation of central stress fibers. These observations are representative of the entirety of siRNA-treated cells, and were reproduced in multiple independent experiments (at least n ¼ 3 for each condition).

a known pharmacological inhibitor of sphingomyelinase (45), to understand the specificity of D609, we treated the confluent HPAECs with imipramine, LPS, and combined imipramine and LPS. Cells were then fixed and stained to visualize for F-actin (Figure E3). Similar to the control sample (Figure E3A), cells treated with imipramine did not show any significant alterations in actin cytoskeleton. Cells treated with LPS demonstrated significant actin stress fiber formation (Figure E3B), whereas cells pretreated with imipramine failed to inhibit LPS-induced central stress fiber formation (Figure E3D). Similarly, pretreatment with imipramine did not lead to the formation of cortical actin rings (Figure E3C), suggesting that D609 specifically inhibits the SMS enzyme to induce cytoskeletal activation, causing endothelial barrier enhancement.

DISCUSSION Sphingomyelin is a major sphingolipid located in the lipid rafts of the cell membrane, which are considered microdomains of plasma membrane critical for signal transduction (20). The depletion of

sphingomyelin from lipid rafts causes an inhibition of TNF-a– mediated signaling pathways, preventing the activation of NF-kB (21), and such treatment also inhibits the proinflammatory signals mediated by LPS through Toll-like receptor (TLR)4 receptors (22). Sphingomyelin synthase is an important enzyme involved in SM synthesis. Two known isoforms of SMS are SMS1 and SMS2. Although both isoforms are involved in de novo SM biosynthesis and total cellular SM concentrations (23), SMS2 was implicated in the formation and maintenance of plasma membrane (24). The knockdown of SMS2 causes a depletion of SM in membrane lipid rafts (25), resulting into a stronger resistance to lysenin-mediated lysis. These results suggest that SMS2 plays a role in the remodeling or maintenance of SM-enriched lipid microdomains, and SMS2 deficiency could alter the signal transduction mediated by lipid raft–associated receptors (26). In the present study, we demonstrate a novel and essential role of SMS2 in modulating endothelial barrier integrity by actin cytoskeletal activation. D609 is a xanthate compound with antiviral, antitumoral (27) actions and antioxidant/glutathione mimetic properties, attributable

100


Figure 6. Effects of SMS2 inhibition and D609 treatment on myosin light chain (MLC) phosphorylation (p). HPAECs were grown on gelatinized coverslips. (A) Cells transfected with control siRNA were treated with D609 (50 mM) for 1 hour, and fixed and stained to determine actin stress fibers and MLC phosphorylation. (B) Cells transfected with control siRNA, exposed to LPS alone, demonstrated MLC phosphorylation prominently along stress fibers all over the cells. (C) However, in SMS2 siRNA-transfected cells, after LPS treatment, phosphorylated MLC redistributes from central stress fibers to the periphery in a prominent MLC-containing cortical ring with actin colocalization, suggesting a potential mechanism for structural stability and increased cell–cell junctional interaction. These observations are representative of the entire EC monolayer, and were reproduced in multiple independent experiments (n ¼ at least 3 for each condition).

to the presence of the thiol function (28, 29), and is also known to inhibit both SMS and PC-PLC enzymes (26, 30). Luberto and colleagues (31) found that D609 blocked the TNF-a–mediated and phorbol ester–mediated NF-kB activation that was concomitant with decreased concentrations of SM and DAG. This did not affect the generation of ceramide, suggesting that SM and DAG derived from SM synthesis via SMS pathway are involved in NFkB activation. Recently, D609 was shown to provide protection after stroke via SMS inhibition and elevated ceramide concentrations (32). Recent studies showed that D609 can reduce cytokine expression in LPS-stimulated macrophages (33, 34), and can prevent LPS-induced or TNF-induced lethal shock (35) as well as PAF-mediated lung edema (17). However, it remains unclear whether the specific inhibition of SMS activity was the major cause of the actions of D609, or if PC-PLC inactivation was involved (18). Our group recently showed (19) that SMS2 knockout mice are protected against endotoxin-mediated lung injury. Therefore, to substantiate our previous results, we compared D609dependent inhibition with SMS2 inhibition. Our results demonstrate that the severe lung injury produced by LPS was dramatically reduced by D609 pretreatment. This was demonstrated by significant reductions in BAL cell count, MPO activity, and protein content (Figure E4). Furthermore, the formation of edema in this model was shown to be virtually nonexistent. Neutrophil accumulation in the lungs of these mice was significantly reduced by the administration of D609. These findings indicate that D609 inhibits both vascular leakage and inflammation in LPS-induced injury. D609 was previously found to decrease interleukin-1b–induced and TNF-a–induced vascular cell adhesion molecule–1 expression in endothelial cells via PC-PLC inhibition (36, 37), which could account for the lower number of inflammatory cells in our model. However, the concomitant decrease in vascular leakage suggests an increase in

endothelial barrier integrity that could also lead to less inflammation by affecting the mobilization of leukocytes through the vessel wall. We observed reduced SMS activity in D609-treated lungs, further suggesting the specific inhibition of SMS by D609 treatment (Figure E5). To understand the effects of D609 on endothelial barrier integrity, we performed experiments in vitro involving HPAE and EA-hy926 endothelial cells (Figure E2). Our results show that D609 induces an immediate, dose-dependent augmentation of endothelial cell barrier function, with the full effect evolving at 40–60 minutes after stimulation, and this effect is sustained up to 24 hours after treatment. These observations were demonstrated via an FITC-dextran permeability assay that showed a significant reduction in FITC-dextran clearance after D609 treatment. In addition, D609 abolishes LPS-mediated increases in endothelial permeability, suggesting that D609 reinforces endothelial barrier integrity. Earlier studies showed that mediators such as thrombin, TNF-a, and LPS stimulate their respective receptors on endothelial cells to initiate signaling that increases cytosolic Ca21 and activates myosin light chain kinase and monomeric GTPases RhoA. This causes an increase in central stress fiber formation and MLC phosphorylation, which in concert results in cellular contraction, the disruption of cell junctions leading to the formation of paracellular gaps, and increased permeability (38–40). To understand the mechanism of D609-induced peripheral actin ring formation and permeability changes, MLC phosphorylation was further studied in D609treated cells via immunofluorescence and Western blotting. As shown in Figure 6 and in all other experiments, D609 alone did not induce MLC phosphorylation. However, LPS-induced MLC phosphorylation was notably localized to the periphery, and was quantitatively reduced in Western blots. Because D609 inhibits LPS-mediated increases in endothelial permeability in both mice and HPAECs, we hypothesized that D609 induces critical changes in cell cytoskeletons that are crucial in modulating the functional regulation of paracellular permeability. To confirm our hypothesis, we focused on immunofluorescence studies using Texas Red–phalloidin for F-actin staining. We demonstrated that D609 causes an inhibition of central stress fiber formation in response to LPS, thereby attenuating cell contraction and paracellular gap formation, and leading to the dose-dependent augmentation of endothelial cell barrier function and increased peripheral actin (Figure E2). D609 also markedly increased the redistribution of F-actin at the periphery of cells, with cortical actin ring formation and the ruffling of lamellipodia (Figures 3A and 3B). In addition, D609 caused the redistribution of MLC phosphorylation away from central stress fibers, to form a prominent cortical ring with actinstaining lamellipodia (Figure 6). The interaction of actin and myosin at these cortical rings provides cellular stability, whereas the lamellipodia extend into neighboring cells to interact and tighten cell–cell junctions (41), a feature observed with multiple barrier-enhancing stimuli, including S1P (42), polyethylene glycol (41), hepatocyte growth factor (43), and simvastatin (44). These findings suggest that D609 acts as an endothelial barrier– enhancing agonist through cytoskeletal rearrangement. Interestingly, D609 does not invoke a transient calcium spike, unlike S1P, a well known endothelial barrier–enhancing agonist and trigger of intracellular calcium (41, 42), indicating the involvement of calcium-independent signaling pathways for cytoskeletal activation (Figure 7). Furthermore, D609 also blocked LPS-induced calcium signaling. This is consistent with the ability of D609 to block LPS-induced barrier disruption, suggesting that the probarrier integrity effects of D609 may be modulating upstream signaling, before the calcium activation by LPS. Next, we explored the pathway through which D609 causes cytoskeletal reorganization. Interestingly, the inhibition of the SMS2


101

Figure 7. D609 inhibits LPSinduced calcium spike. HPAECs, grown on six-well culture plates, were loaded with fluorescent calcium indicator dyes Rhod2–AM and Fluo-4 (5 mM, 30 min), and then fluorescence was measured using a luminescence spectrometer at excitation wavelengths of 552 and 494 nm and emission wavelengths of 578 and 516 nm. High relative fluorescence units (RFUs) indicate increased calcium concentrations. (A and B) D609 reduced intracellular calcium concentrations compared with control samples. Sequential stimulation with LPS inhibited the LPS-induced intracellular calcium spike, which correlates with the ability of D609 to inhibit LPS-induced stress fiber formation and barrier disruption. (C) Mean calcium concentrations under different conditions. Similar results were reproduced with Fluo-4 (data not shown) (**P , 0.05). These observations were reproduced in multiple independent experiments (n ¼ at least 3 for each condition).

enzyme, using a silencing technique in vitro model, produced effects similar to those shown by D609. SMS2-silenced cells show some alterations in the actin cytoskeleton at basal level. However, on exposure to LPS, these changes become more pronounced, and the cells show cortical actin at the periphery instead of central stress fiber formation, and they also express ruffling lamellipodia. These changes are not as dramatic as those seen with D609, which may be explained by the fact that D609 inhibits both isoforms of

SMS enzyme, whereas in our study we used only SMS2-silencing RNA, showing that only a partial inhibition of SM synthesis is sufficient to block the effect of LPS-induced endothelial injury. This also suggests that SMS2 inhibition could exert more impact on the depletion of lipid-rich microdomains that are actively involved in signal transduction (26). To understand the specificity of D609 (SMS versus sphingomyelinase), we investigated the effects of sphingomyelinase inhibition on cytoskeletal activation by imipramine

102


(45). Surprisingly, imipramine did not induce the formation of cortical actin rings, and also failed to inhibit LPS-induced stress fiber formation, suggesting that D609 specifically inhibits the SMS enzyme to induce cytoskeletal activation, and that the inhibition of SMS is sufficient to enhance endothelial barrier integrity. In conclusion, we show that D609 acts as a potent endothelial barrier–enhancing agent by causing crucial changes in the EC actin cytoskeleton. In addition, D609 reduces LPS-induced endothelial lung injury through SMS inhibition, leading to a loss of sphingomyelin from lipid rafts, which in turn could affect downstream signaling, mediated by TNF-a and TLR4 receptors. The inhibition of SMS activity by SMS2 siRNA treatment and the subsequent increased barrier function clearly suggest a role for SM synthesis in reducing lung injury. Although D609 has a short half-life, its efficacy and rapidity in preventing barrier dysfunction and its sustained effects on endothelial barrier integrity make it an ideal therapeutic agent to reverse endothelial barrier dysfunction in ALI.

16.

17.

18.

19.

20. 21.

Author disclosures are available with the text of this article at www.atsjournals.org.

22.

References 1. Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med 2000;342:1334–1349. 2. Dudek SM, Garcia JGN. Cytoskeletal regulation of pulmonary vascular permeability. J Appl Physiol 2001;91:1487–1500. 3. Tremblay L, Valenza F, Ribeiro SP, Li J, Slutsky AS. Injurious ventilatory strategies increase cytokines and c-Fos m-RNA expression in an isolated rat lung model. J Clin Invest 1997;99:944–952. 4. Parker JC. Inhibitors of myosin light chain kinase and phosphodiesterase reduce ventilator-induced lung injury. J Appl Physiol 2000; 89:2241–2248. 5. Birukov KG, Jacobson JR, Flores AA, Ye SQ, Birukova AA, Verin AD, Garcia JGN. Magnitude-dependent regulation of pulmonary endothelial cell barrier function by cyclic stretch. Am J Physiol Lung Cell Mol Physiol 2003;285:L785–L797. 6. Bannerman DD, Goldblum SE. Endotoxin induces endothelial barrier dysfunction through protein tyrosine phosphorylation. Am J Physiol 1997;273:L217–L226. 7. Bannerman DD, Goldblum SE. Direct effects of endotoxin on the endothelium: barrier function and injury. Lab Invest 1999;79:1181–1199. 8. Peng X, Hassoun PM, Sammani S, McVerry BJ, Burne MJ, Rabb H, Pearse D, Tuder RM, Garcia JG. Protective effects of sphingosine 1– phosphate in murine endotoxin-induced inflammatory lung injury. Am J Respir Crit Care Med 2004;169:1245–1251. 9. McVerry BJ, Peng X, Hassoun PM, Sammani S, Simon BA, Garcia JG. Sphingosine 1–phosphate reduces vascular leak in murine and canine models of acute lung injury. Am J Respir Crit Care Med 2004;170:987–993. 10. Wadgaonkar R, Patel V, Grinkina N, Romano C, Liu J, Zhao Y, Sammani S, Garcia JG, Natarajan V. Differential regulation of sphingosine kinases 1 and 2 in lung injury. Am J Physiol Lung Cell Mol Physiol 2009; 296:L603–L613. 11. Szczepaniak WS, Zhang Y, Hagerty S, Crow MT, Kesari P, Garcia JG, Choi AM, Simon BA, McVerry BJ. Sphingosine 1–phosphate rescues canine LPS-induced acute lung injury and alters systemic inflammatory cytokine production in vivo. Transl Res 2008;152:213–224. 12. Huitema K, Dikkenberg JVD, Brouwers JF, Holthuis JC. Identification of a family of animal sphingomyelin synthases. EMBO J 2004;23:33–44. 13. Li Z, Hailemariam TK, Zhou H, Li Y, Duckworth DC, Peake DA, Zhang Y, Kuo MS, Cao G, Jiang XC. Inhibition of sphingomyelin synthase (SMS) affects intracellular sphingomyelin accumulation and plasma membrane lipid organization. Biochim Biophys Acta 2007; 1771:1186–1194. 14. Ding T, Li Z, Hailemariam TK, Mukherjee S, Maxfield FR, Wu MP, Jiang XC. SMS overexpression and knockdown: impact on cellular sphingomyelin and diacylglycerol metabolism, and cell apoptosis. J Lipid Res 2008;49:376–385. 15. Li F, Wu N, Su RB, Zheng JQ, Xu B, Lu XQ, Cong B, Li J. Involvement of phosphatidylcholine-selective phospholipase C in activation of

23.

24. 25.

26.

27.

28.

29.

30.

31.

32. 33.

34.

mitogen-activated protein kinase pathways in imidazoline receptor antisera-selected protein. J Cell Biochem 2006;98:1615–1628. Meng A, Luberto C, Meier P, Bai A, Yang X, Hannun YA, Zhou D. Sphingomyelin synthase as a potential target for D609-induced apoptosis in U937 human monocytic leukemia cells. Exp Cell Res 2004; 292:385–392. Yang Y, Yin J, Baumgartner W, Samapati R, Solymosi EA, Reppien E, Kuebler WM, Uhlig S. Platelet-activating factor reduces endothelial NO production: role of acid sphingomyelinase. Eur Respir J 2009;36: 417–427. Schu¨tze S, Potthoff K, Machleidt T, Berkovic D, Wiegmann K, Kro¨nke M. TNF activates NF kB by phosphatidylcholine specific phospholipase C induced “acidic” sphingomyelin breakdown. Cell 1992;71:765–776. Gowda S, Yeang C, Wadgaonkar S, Anjum F, Grinkina N, Cutaia M, Xian-Chen Jiang R, Wadgaonkar R. Sphingomyelin synthase 2 (SMS2) deficiency attenuates LPS induced lung injury. Am J Physiol Lung Cell Mol Physiol 2011;300:L430–L440. Simons K, Ikonen E. Functional rafts in cell membranes. Nature 1997; 387:569–572. Legler DF, Micheau O, Doucey MA, Tschopp J, Bron C. Recruitment of TNF receptor 1 to lipid rafts is essential for TNFalpha-mediated NF-kappaB activation. Immunity 2003;18:655–664. Triantafilou M, Miyake K, Golenbock T, Triantafilou K. Mediators of innate immune recognition of bacteria concentrate in lipid rafts and facilitate lipopolysaccharide-induced cell activation. J Cell Sci 2002; 115:2603–2611. Tafesse FG, Huitema K, Hermansson M, van der Poel S, van den Dikkenberg J, Uphoff A, Somerharju P, Holthuis JC. Both sphingomyelin synthase SMS1 and SMS2 are required for sphingomyelin homeostasis and growth in human HeLa cells. J Biol Chem 2007;282: 17537–17547. Tafesse FG, Ternes P, Holthuis JC. The multigenic sphingomyelin synthase family. J Biol Chem 2006;281:29421–29425. Li Z, Hailemariam TK, Zhou H, Li Y, Duckworth DC, Peake DA, Zhang Y, Kuo MS, Cao G, Jiang XC. Inhibition of sphingomyelin synthase (SMS) affects intracellular sphingomyelin accumulation and plasma membrane lipid organization. Biochim Biophys Acta 2007; 1771:1186–1194. Hailemariam TK, Huan C, Liu J, Li Z, Roman C, Kalbfeisch M, Bui HH, Peake DA, Kuo MS, Cao G, et al. Sphingomyelin synthase 2 deficiency attenuates NFkappaB activation. Arterioscler Thromb Vasc Biol 2008;28:1519–1526. Amtmann E. The antiviral, antitumoural xanthate D609 is a competitive inhibitor of phosphatidylcholine-specific phospholipase C. Drugs Exp Clin Res 1996;22:287–294. Sultana R, Newman SF, Abdul HM, Cai J, Pierce WM, Klein JB, Merchant M, Butterfield DA. Protective effect of D609 against amyloid-b 1–42 induced oxidative modification of neuronal proteins: redox proteomics study. J Neurosci Res 2006;84:409–417. Zhou DH, Lauderback CM, Yu T, Brown SA, Butterfield DA, Thompson JS. D609 inhibits ionizing radiation–induced oxidative damage by acting as a potent antioxidant. J Pharmacol Exp Ther 2001; 298:103–109. Luberto C, Hannun YA. SM synthase, a potential regulator of intracellular levels of ceramide and diacylglycerol during SV40 transformation: does SM synthase account for the putative PC-specific PLC? J Biol Chem 1998;273:14550–14559. Luberto C, Yoo DS, Suidan HS, Bartoli GM, Hannun YA. Differential effects of sphingomyelin hydrolysis and resynthesis on the activation of NF-kappa B in normal and SV40-transformed human fibroblasts. J Biol Chem 2000;275:14760–14766. Adibhatla RM, Hatcher JF. Protection by D609 through cell-cycle regulation after stroke. Mol Neurobiol 2010;41:206–217. Monick MM, Carter AB, Gudmundsson G, Mallampalli RK, Powers LS, Hunninghake GW. A phosphatidylcholine-specific phospholipase C regulates activation of p42/44 mitogen–activated protein kinases in lipopolysaccharide-stimulated human alveolar macrophages. J Immunol 1999;162:3005–3012. Zhang F, Zhao G, Dong Z. Phosphatidylcholine-specific phospholipase C and D in stimulation of RAW264.7 mouse macrophage-like cells by lipopolysaccharide. Int Immunopharmacol 2001;1:1375–1384.

Anjum, Joshi, Grinkina, et al.: Inhibition of Sphingomyelin Synthase 35. Machleidt T, Kramer B, Adam D, Neumann B, Schutze S, Wiegmann K, Kronke M. Function of the p55 TNF receptor “death domain” mediated by phosphatidylcholine-specific PLC. J Exp Med 1996;184: 725–733. 36. Cobb RR, Felts KA, Parry GC, Mackman N. D609, a phosphatidylcholinespecific phospholipase C inhibitor, blocks interleukin-1 beta–induced vascular cell adhesion molecule 1 gene expression in human endothelial cells. Mol Pharmacol 1996;49:998–1004. 37. Weber C, Erl W, Pietsch A, Danesch U, Weber PC. Docosahexaenoic acid selectively attenuates induction of vascular cell adhesion molecule–1 and subsequent monocytic cell adhesion to human endothelial cells stimulated by tumor necrosis factor–alpha. Arterioscler Thromb Vasc Biol 1995;15:622–628. 38. Birukova AA, Smurova K, Birukov KG, Kaibuchi K, Garcia JGN, Verin AD. Role of Rho GTPases in thrombin-induced lung vascular endothelial cells barrier dysfunction. Microvasc Res 2004;67:64–77. 39. Garcia JG, Verin AD, Schaphorst KL. Regulation of thrombin-mediated endothelial cell contraction and permeability. Semin Thromb Hemost 1996;22:309–315. 40. Vandenbroucke E, Mehta D, Minshall R, Malik AB. Regulation of endothelial junctional permeability. Ann N Y Acad Sci 2008;1123: 134–145.

103 41. Chiang ET, Camp SM, Dudek SM, Brown ME, Usatyuk PV, Zaborina O, Alverdy JC, Garcia JGN. Protective effects of high-molecular weight polyethylene glycol (PEG) in human lung endothelial cell barrier regulation: role of actin cytoskeletal rearrangement. Microvasc Res 2009;77:174–186. 42. Garcia JG, Liu F, Verin AD, Birukova A, Dechert MA, Gerthoffer WT, Bamburg JR, English D. Sphingosine 1–phosphate promotes endothelial cell barrier integrity by EDG-dependent cytoskeletal rearrangement. J Clin Invest 2001;108:689–701. 43. Liu F, Schaphorst KL, Verin AD, Jacobs K, Birukova A, Day RM, Bogatcheva N, Bottaro DP, Garcia JG. Hepatocyte growth factor enhances endothelial cell barrier function and cortical cytoskeletal rearrangement: potential role of glycogen synthase kinase–3beta. FASEB J 2002;16:950–962. 44. Jacobson JR, Dudek SM, Birukov KG, Ye SQ, Grigoryev DN, Girgis RE, Garcia JGN. Cytoskeletal activation and altered gene expression in endothelial barrier regulation by simvastatin. Am J Respir Cell Mol Biol 2004;30:662–670. 45. Cuschieri J, Bulger E, Billgrin J, Garcia I, Maier RV. Acid sphingomyelinase is required for lipid raft TLR4 complex formation. Surg Infect (Larchmt) 2007;8:91–106.