Novel role of stathmin in microtubule-dependent control of endothelial permeability

Abstract

Microtubule (MT) dynamics in vascular endothelium are modulated by vasoactive mediators and are critically involved in the control of endothelial cell (EC) permeability via Rho GTPase-dependent crosstalk with the actin cytoskeleton. However, the role of regulators in MT stability in these mechanisms remains unclear. This study investigated the involvement of the MT-associated protein stathmin in the mediation of agonist-induced permeability in EC cultures and vascular leak in vivo. Thrombin treatment of human pulmonary ECs induced rapid dephosphorylation and activation of stathmin. Inhibition of stathmin activity by small interfering RNA-based knockdown or cAMP-mediated phosphorylation abrogated thrombin-induced F-actin remodeling and Rho-dependent EC hyperpermeability, while expression of a phosphorylation-deficient stathmin mutant exacerbated thrombin-induced EC barrier disruption. Stathmin suppression preserved the MT network against thrombin-induced MT disassembly and release of Rho-specific guanine nucleotide exchange factor, GEF-H1. The protective effects of stathmin knockdown were observed in vivo in the mouse 2-hit model of ventilator-induced lung injury and were linked to MT stabilization and down-regulation of Rho signaling in the lung. These results demonstrate the mechanism of stathmin-dependent control of MT dynamics, Rho signaling, and permeability and suggest novel potential pharmacological interventions in the prevention of increased vascular leak via modulation of stathmin activity.—Tian, X., Tian, Y., Sarich, N., Wu, T., Birukova, A. A. Novel role of stathmin in microtubule-dependent control of endothelial permeability.

The vascular endothelium functions as a semiselective barrier for mass transport through the vessel wall. Regulation of the endothelial cell (EC) barrier is achieved via the balance between actomyosin-driven contractile events and tethering forces implied by cell adhesive structures and the cortical actin cytoskeletal network (1, 2). Microtubules represent another key part of the cell cytoskeleton and regulate many processes, including mitosis and locomotion, as well as protein and organelle transport (3, 4). Cells respond to external stimuli by altering dynamics of microtubule assembly/disassembly and spatial rearrangements. Conversely, changes in microtubule dynamics modulate intracellular signal transduction.

We have previously shown that the EC barrier-disruptive effects of thrombin, TNF-α, and transforming growth factor-β (TGFβ) are associated with partial microtubule disassembly (5–7). In ECs, microtubule depolymerization leads to dissolution of the cortical actin cytoskeleton, increased myosin light chain (MLC) phosphorylation, stress fiber formation, contraction, and EC barrier dysfunction. MLC phosphorylation observed during microtubule depolymerization is a result of the activated small GTPase Rho and its effector Rho-kinase and involves microtubule-dependent mechanisms of Rho activation (5, 8, 9). Guanine nucleotide exchange factor H1 (GEF-H1) is a Rho-specific GEF, which localizes on microtubules. In the microtubule-bound state, the guanine-exchange activity of GEF-H1 is suppressed, whereas GEF-H1 release caused by microtubule disruption stimulates Rho-specific GEF activity (8). Indeed, microtubule disassembly caused by cell stimulation with thrombin, TNF-α, or nocodazole was linked to GEF-H1-dependent activation of Rho signaling leading to cell contraction (10–12).

Stathmin is an important regulator of microtubule polymerization and dynamics. In the unphosphorylated state, stathmin destabilizes microtubules in 2 ways: by reducing the microtubule polymer mass through sequestration of soluble tubulin into an assembly-incompetent T2S complex (2 α:β-tubulin dimers per molecule of stathmin) and by increasing the switching frequency (catastrophe frequency) from growth to shortening at plus and minus ends by binding directly to the microtubules. Phosphorylation of stathmin on one or more of its 4 serine residues (S¹⁶, S²⁵, S³⁸, and S⁶³) reduces its microtubule-destabilizing activity (13).

This study examined the role of stathmin in thrombin-mediated regulation of microtubule growth and stability and investigated microtubule-dependent mechanisms of vascular endothelial permeability regulation by stathmin. Using comprehensive molecular, biochemical, imaging, and functional approaches, we characterized a crosstalk between agonist-induced stathmin activation and Rho GTPase signaling via the Rho-specific guanine nucleotide exchange factor, GEF-H1.

MATERIALS AND METHODS

Reagents and cell culture

Unless specified, biochemical reagents were obtained from Sigma (St. Louis, MO, USA). Reagents for immunofluorescence were purchased from Molecular Probes (Eugene, OR, USA). Antibodies to phospho-myosin-associated phosphatase (MYPT) were purchased from Millipore (Billerica, MA, USA); GEF-H1 and diphospho-MLC from Cell Signaling (Beverly, MA, USA); stathmin, vascular endothelial cadherin (VE-cadherin), and end-binding protein-1 (EB1) from BD Transduction Laboratories (San Diego, CA, USA); phospho-stathmin-S¹⁶ and phospho-stathmin-S⁶³ from Abcam (Cambridge, MA, USA); and phospho-tau-S²⁶² and phospho-tau-S⁴⁰⁹ from Invitrogen (Carlsbad, CA, USA). 8-Bromo-adenosine-3′,5′-cyclic monophosphate (cAMP) was purchased from Calbiochem (La Jolla, CA, USA). Human pulmonary artery endothelial cells (HPAECs) were obtained from Lonza (Allendale, NJ, USA), maintained in a complete culture medium according to the manufacturer's recommendations, and used for experiments at passages 5–7.

Small interfering RNA (siRNA) and DNA transfections

To deplete endogenous stathmin, predesigned human or mouse Stealth Select siRNA sets of standard purity were ordered from Invitrogen. For GEF-H1 knockdown, predesigned standard purity human siRNA sets were ordered from Dharmacon (Lafayette, CO, USA). Transfection of ECs with siRNA was performed as described previously (14). Nontargeting, nonspecific siRNA (nsRNA) was used as a control treatment. After 72 h of transfection, cells were used for experiments or harvested for Western blot verification of specific protein depletion. Test experiments using control fluorescently labeled RNA oligonucleotides showed ∼90% efficiency of siRNA transfection of EC (data not shown). For in vivo experiments, we used polymer-based administration of nonspecific or specific siRNA conjugated with polycation polyethilenimine (PEI-22), described in our previous studies (14–16). This technique allows for lung-specific DNA and siRNA delivery (14, 17, 18) and preferential targeting of lung endothelium, although additional effects on other lung cells cannot be completely excluded. The optimal concentration of siRNA was determined in a series of preliminary experiments. siRNA at 2 mg/kg showed the most significant target gene inhibition after 72 h of transfection, determined by Western blot analysis. Treated mice showed no signs of nsRNA-induced inflammation. nsRNA (Dharmacon) was used as a control treatment for both in vitro and in vivo experiments. Plasmids encoding human GEF-H1 (aa 1-572) bearing an EGFP-tag and Stathmin-S63A bearing a His-tag were kindly provided by Dr. Gary M. Bokoch (Scripps Research Institute, La Jolla, CA, USA) and were used for transient transfections according to protocols described previously (5, 14). Control transfections were performed with empty vectors. In experiments with ectopic expression of GEF-H1 (aa 1-572) and Stathmin-S63A, significant decrease in basal levels of electrical resistance [693±187 Ω for GEF-H1 (aa 1-572) and 961±223 Ω for Stathmin-S63A vs. 1261±114 Ω for empty vector controls] was observed after 48-72 h of transfection, which is not surprising, given the critical role of Rho for the regulation of contractile mechanisms. Therefore, the following experiments were performed after 24 h of transfection. These conditions provided pronounced protein expression, verified by Western blot analysis (data not shown), but caused only a modest reduction in basal resistance [961±118 Ω for GEF-H1 (aa 1-572) and 1059±147 Ω for Stathmin-S63A vs. 1175±163 Ω for empty vector controls].

Measurement of transendothelial electrical resistance (TER)

The cellular barrier properties were analyzed by measurements of TER across confluent human pulmonary artery endothelial monolayers using an electrical cell-substrate impedance sensing system (Applied Biophysics, Troy, NY, USA), as described previously (5, 10). In the current studies, we did not observe significant effects of nsRNA, specific siRNA, or DNA constructs on cell viability and monolayer integrity, and initial testing of nontransfected, siRNA-transfected, and DNA-transfected EC monolayers did not reveal significant differences in basal TER levels.

Immunofluorescence and image analysis

Endothelial monolayers plated on glass coverslips were subjected to immunofluorescence staining with Texas Red phalloidin to visualize F-actin, as described previously (10, 14). For microtubule quantification, cells were fixed with −20°C methanol, and immunostaining was carried out with β-tubulin and VE-cadherin antibodies. The cells to be analyzed were outlined based on VE-cadherin staining and were divided at 70% radius. The integrated fluorescence density in the peripheral area was measured using MetaMorph software (Molecular Devices Corp., Sunnyvale, CA, USA) and was calculated as a percentage of the integrated fluorescence density in the total cell area. The results were normalized in each experiment. Similar methods were applied to EB1 quantification in fixed cells, except that EB1 immunoactivity was manually counted, and results were not normalized.

Live cell imaging and time-lapse microtubule plus-end tracking

Cells were plated on MatTek dishes (MatTek, Ashland, MA, USA) and transfected with GFP-EB1. Images were acquired with ×100 NA 1.45 oil objective in a 3I Marianas Yokogawa-type spinning-disk confocal system equipped with a CO₂ chamber and a heated stage (Intelligent Imaging Innovations, Inc., Denver, CO, USA). Time-lapse images were taken with 2-s intervals for 40 to 60 s. The 20 consecutive images in each condition were used for projection analysis using ImageJ software (U.S. National Institutes of Health, Bethesda, MD, USA). For tacking analysis, EB1 in the cell margin area (2–10 μm from cell border) was tracked with the Manual Tracking plug-in in ImageJ software. The median track length was calculated using Excel software (Microsoft, Redmond, WA, USA).

Isolation of microtubule-enriched fraction and immunoblotting

Confluent HPAECs were stimulated with thrombin, and microtubule-enriched fractions were isolated as described previously (5). For analysis of protein phosphorylation profile, cells were stimulated and then lysed, and protein extracts were separated by SDS-PAGE, transferred to PVDF membrane, and probed with specific antibodies as described previously (19).

Mechanical ventilation protocol

All experimental protocols involving the use of animals were approved by the University of Chicago Institutional Animal Care and Use Committee for the humane treatment of experimental animals. The 2-hit model of ventilator-induced lung injury was performed as described previously (14). In brief, C57BL/6J mice (8–10 wk old; males) with a weight of 20–25 g (Jackson Laboratories, Bar Harbor, ME, USA) were anesthetized with an intraperitoneal injection of ketamine (75 mg/kg) and acepromazine (1.5 mg/kg). A tracheotomy was performed, and the trachea was cannulated with a 20-gauge catheter (Exelint International, Los Angeles, CA, USA) that was tied into place to prevent air leak. The animals were then placed on a mechanical ventilator (Harvard Apparatus, Boston, MA, USA). Mice were given a single dose of thrombin-related activating peptide (TRAP6; 1.5×10⁻⁵ mol/kg, intratracheal instillation) followed by 4 h of mechanical ventilation at high tidal volume (HTV; 30 ml/kg) ventilation. After the experiment, animals were killed by exsanguination under anesthesia. Tracheotomy was performed, and the trachea was cannulated with a 20-gauge intravenous catheter, which was tied into place. Bronchoalveolar lavage (BAL) was performed using 1 ml of warmed sterile HBSS (30°C). The collected lavage fluid was centrifuged (2500 rpm, 20 min, 4°C); the supernatant was removed and frozen at −80°C for subsequent protein study. The BAL protein concentration was determined by BCA protein assay kit (Thermo Scientific, Rockford, IL, USA). BAL cell pellets were resuspended in 1 ml of red blood cell lysis buffer (ACK Lysing Buffer; BioSource International, Grand Island, NY, USA) and repelleted by centrifugation (2500 rpm, 20 min, 4°C). The resulting cell preparation, resuspended in PBS, was used for cell counting by a standard hemocytometer technique.

Statistical analysis

Results are expressed as means ± sd of 3–6 independent experiments. Experimental samples were compared to controls by unpaired Student's t test. For multiple-group comparisons, a 1-way ANOVA and post hoc multiple comparison tests were used. Values of P < 0.05 were considered statistically significant.

RESULTS

Stathmin knockdown prevents thrombin-induced EC barrier disruption

We used an siRNA-mediated knockdown approach to investigate stathmin involvement in thrombin-induced permeability. After 72 h of transfection, HPAECs were treated with thrombin or vehicle, and TER was monitored over the time. Knockdown of stathmin inhibited thrombin-induced TER decline, suggesting a strong barrier-protective effect (Fig. 1A).

Figure 1.

Stathmin knockdown prevents thrombin-induced EC barrier disruption and Rho signaling. A) Effect of stathmin siRNA transfection on thrombin-induced change in TER. HPAECs were transfected with stathmin siRNA or nsRNA, followed by thrombin treatment (0.3 U/ml) at the time indicated by arrow. B) Effect of stathmin siRNA transfection on thrombin-induced actin remodeling and paracellular gap formation (indicated by arrow). Stathmin-transfected cells were treated with thrombin for 15 min. F-actin was visualized by staining with Texas Red phalloidin. C) HPAECs were transfected with stathmin siRNA or nsRNA, followed by thrombin stimulation. Phosphorylation of VE-cadherin (p-VE-cadherin), MYPT (p-MYPT), and MLC (pp-MLC) were detected by phosphorylation-specific antibodies. Effective knockdown of stathmin is shown by immunoblotting with stathmin antibody. Equal protein loading was confirmed by determination of β-actin content in total cell lysates. Bar graphs depict quantitative analysis of Western blot densitometry data. *P < 0.05 vs. nsRNA; n = 5.

We further examined a role of stathmin in thrombin-induced F-actin remodeling and EC monolayer barrier disruption in immunofluorescence studies. Thrombin induced stress fiber formation and the appearance of paracellular gaps in control, nsRNA-treated ECs, whereas stathmin knockdown abolished disruptive effects of thrombin (Fig. 1B).

Several biochemical parameters reflecting EC permeability changes were also examined. Stathmin knockdown attenuated thrombin-induced VE-cadherin phosphorylation at Y⁷³¹ (Fig. 1C), the phosphorylation site known to promote disassembly of VE-cadherin-based cell-cell junctions (20). The Rho pathway of endothelial permeability involves phosphorylation of MYPT at the Rho kinase-specific site T⁸⁵⁰ (21), leading to increased MLC phosphorylation, EC actomyosin contraction, and disruption of EC monolayer integrity. Knockdown of stathmin suppressed thrombin-induced MYPT and MLC phosphorylation (Fig. 1C).

Stathmin activation is promoted by thrombin and critical for thrombin-induced microtubule disassembly

Phosphorylation at either S¹⁶ or S⁶³ reduces the ability of stathmin to bind to and sequester soluble tubulin and its ability to act as a catastrophe factor by directly binding to the microtubules (13). Our data show considerable levels of stathmin phosphorylation at S¹⁶ and S⁶³ in ECs under basal conditions and a pronounced decrease in phospho-S¹⁶ and -S⁶³ immunoreactivity on EC treatment with thrombin (Fig. 2A).

Figure 2.

Stathmin activation is promoted by thrombin and critical for thrombin-induced microtubule disassembly. A) HPAECs were treated with thrombin for different time durations. Site-specific phosphorylation of stathmin and tau, as well as tubulin acetylation, was detected by Western blot. Equal protein loading was confirmed by determination of stathmin or β-tubulin content in total cell lysates. Bar graphs depict quantitative analysis of Western blot densitometry data. *P < 0.05 vs. control; n = 3. B) HPAECs were transfected with stathmin siRNA or nsRNA, followed by thrombin treatment. Site-specific phosphorylation of tau and tubulin acetylation were detected by Western blot. Stathmin depletion and equal protein loading were confirmed by determination of stathmin and β-tubulin content, respectively, in total cell lysates. Bar graphs depict quantitative analysis of Western blot densitometry data. *P < 0.05 vs. nsRNA; n = 3. C) Cells grown on coverslips were transfected with stathmin siRNA and stimulated with thrombin (0.3 U/ml, 15 min), followed by immunofluorescent staining with an antibody against β-tubulin. Magnified images (insets) show details of microtubule structure. D) Quantification of peripheral microtubules in panel C was performed as described in Materials and Methods. *P < 0.05 vs. si-stathmin + thrombin; n = 5.

We have previously demonstrated that thrombin induces phosphorylation of the microtubule regulatory protein tau at Rho kinase-dependent sites S²⁶² and S⁴⁰⁹ (22), which leads to microtubule destabilization (5) and decreases the pool of acetylated tubulin representing stable microtubules (Fig. 2A). Notably, stathmin knockdown attenuated tau phosphorylation at S²⁶² and S⁴⁰⁹ and prevented thrombin-induced decreases in acetylated tubulin (Fig. 2B). Immunofluorescence analysis of microtubule structure showed preservation of the peripheral microtubule network in thrombin-treated ECs with depleted stathmin (Fig. 2C and insets). Quantitative analysis confirmed protective effects of stathmin depletion against thrombin-induced reduction of the peripheral microtubule pool (Fig. 2D). We noted a trend of an increased density of peripheral microtubules in nonstimulated cells with depleted stathmin, although these changes did not reach statistical significance.

Stathmin knockdown blocks thrombin-induced inhibition of microtubule growth

The effects of stathmin depletion on microtubule dynamics in control and thrombin-stimulated cells were further examined using a live imaging approach. For this purpose, ECs were transfected with GFP-tagged EB1, which tracks the growing plus end of microtubules. EB1 tracking in live cells was performed as recently described by Komarova et al. (23). Projection images were generated as described in Materials and Methods. Thrombin stimulation did not change the average rate of microtubule growth (data not shown) but caused reduction of the length of EB1 tracks, which represents episodes of uninterrupted microtubule growth (Fig. 3A, B, and Supplemental Videos S1 and S2). These thrombin effects were completely abolished by stathmin knockdown. Alternative analysis of thrombin effects on microtubule growth was performed by quantitative evaluation of the EB1 fluorescence at the peripheral areas of methanol-fixed ECs. The results show a decreased fraction of peripheral EB1 immunoreactivity in thrombin-treated ECs (Fig. 3C). This reduction of peripheral EB1 presence was abolished by stathmin knockdown. These results demonstrate that stathmin knockdown blocks thrombin-induced reduction in both the length of microtubule continuous growth and the number of growing ends at the cell margin.

Figure 3.

Stathmin knockdown blocks thrombin-induced inhibition of microtubule growth. A, B) Live cell imaging of HPAECs transfected with stathmin siRNA or nsRNA and labeled with GFP-EB1. A) Projection analysis of 20 consecutive images before and after thrombin treatment (0.3 U/ml, 15 min) shows changes in GFP-EB1 track length. B) Quantification of GFP-EB1 track length. Each pair of dots represents the median track length in a cell before and after thrombin treatment. C) Quantification of peripheral EB1 in methanol-fixed HPAEC. Cells were transfected with stathmin siRNA and treated with thrombin for 15 min. EB1 was visualized by immunostaining with EB1 antibody and quantified as described in Materials and Methods. *P < 0.05 vs. si-stathmin + thrombin; n = 5.

Phosphorylation-deficient stathmin delays EC recovery after thrombin challenge

To examine the role of stathmin phosphorylation in thrombin-induced EC permeability, cells were transfected with stathmin-S63A, a phosphorylation-deficient mutant. In comparison to ECs transfected with control vector, expression of stathmin-S63A delayed EC barrier recovery phase, characterized by a prolonged restoration of TER to basal levels (Fig. 4A). Delayed barrier recovery was also associated with higher phospho-MYPT and phospho-MLC levels in stathmin-S63A-transfected cells (Fig. 4B), which reflects more prolonged activation of Rho signaling induced by thrombin. Immunofluorescence analysis of thrombin-stimulated ECs with ectopic expression of nontransfected and stathmin-S63A-expressing ECs demonstrated similar patterns of stress fiber formation after 15 min of thrombin stimulation but significantly higher numbers of stress fibers in stathmin-S63A-expressing ECs after 30 min of thrombin treatment (Fig. 4C). These data are consistent with delayed barrier recovery (Fig. 4A) and more prolonged activation of Rho signaling (Fig. 4B) in ECs expressing phosphorylation-deficient stathmin mutant.

Figure 4.

Phosphorylation-deficient stathmin delays EC recovery after thrombin challenge. HPAECs were transfected with phosphorylation-deficient stathmin (STMN-S63A) or empty vehicle, followed by thrombin treatment (0.3 U/ml). A) TER measurements were performed over 4 h. B) Phosphorylation of MYPT and MLC was detected by Western blot after 15 and 30 min of thrombin treatment. Equal protein loading was confirmed by determination of β-tubulin content in total cell lysates. Bar graphs depict quantitative analysis of Western blot densitometry data. *P < 0.05 vs. empty vector; n = 4. C) Actin remodeling after 15 min and 30 min of thrombin treatment was examined by staining with Texas Red phalloidin. Stathmin was visualized by immunostaining with anti-stathmin antibody. Stathmin overexpressed cells are indicated by arrows.

cAMP-dependent stathmin phosphorylation is critical for protection against thrombin-induced microtubule disassembly and EC barrier disruption

We have previously described protective effects of a stable cAMP analog, Br-cAMP, against thrombin-induced EC barrier disruption (24). In the following studies, we examined whether modulation of stathmin activity is involved in cAMP effects. EC treatment with cAMP significantly increased stathmin phosphorylation at S¹⁶ and S⁶³, accompanied by an increase in acetylated tubulin (Fig. 5A). Furthermore, cAMP pretreatment suppressed thrombin-induced disassembly of peripheral microtubules (Fig. 5B, C).

Figure 5.

cAMP-dependent stathmin phosphorylation abolishes thrombin-induced microtubule disassembly A) Western blot analysis of lysates of HPAECs treated with cAMP analog (500 ng/ml) for indicated time duration. Site-specific phosphorylation of stathmin was detected by specific antibodies. Equal protein loading was confirmed by determination of stathmin or β-tubulin content in total cell lysates. B) Cells grown on coverslips were pretreated with cAMP analog for 15 min and stimulated with thrombin (0.3 U/ml, 15 min), followed by immunofluorescence staining with an antibody against β-tubulin. Insets: magnified images of boxed areas show details of microtubule structure. C) Quantification of peripheral microtubules in panel B. *P < 0.05 vs. cAMP + thrombin; n = 5.

Analysis of cAMP effects on thrombin-induced changes in the microtubule plus-end tracking assay with GFP-EB1 showed that cAMP completely abrogated thrombin-induced inhibition of microtubule continuous growth (Fig. 6A, B and Supplemental Videos S3 and S4). These results were consistent with inhibitory effects of cAMP pretreatment on thrombin-induced decreases in the fraction of peripheral EB-1 immunoreactivity determined in methanol-fixed EC preparations (Fig. 6C).

Figure 6.

cAMP-dependent stathmin phosphorylation blocks thrombin-induced inhibition of microtubule growth. Live cell imaging of HPAECs labeled with GFP-EB1 and treated with or without cAMP analog (500 ng/ml, 15 min). A) Projection analysis of 20 consecutive images before and after thrombin (Thr) treatment (0.3 U/ml, 15 min). B) Quantification of GFP-EB1 track length. Each pair of dots represents the median track length in a cell before and after thrombin treatment. Bar graphs depict quantitative analysis of Western blot densitometry data. *P < 0.05 vs. control; n = 3. C) Quantification of peripheral EB1 in methanol-fixed HPAEC monolayers. Cells were pretreated with cAMP, followed by thrombin stimulation for 15 min. EB1 was visualized by immunostaining with anti-EB1 antibody and quantified as described in Materials and Methods. *P < 0.05 vs. cAMP + thrombin; n = 5.

Next, the protective effects of cAMP against thrombin-induced permeability were tested in control cells and ECs expressing the stathmin-S63A mutant. Expression of stathmin-S63A abrogated barrier protective effects of cAMP (Fig. 7A). Furthermore, expression of this mutant suppressed inhibitory effects of cAMP on thrombin-induced activation of Rho signaling, as monitored by increased phosphorylation of the Rho-kinase targets MYPT and MLC (Fig. 7B). Finally, cAMP-induced attenuation of stress fiber formation caused by thrombin was abolished in cells expressing the stathmin-S63A mutant (Fig. 7C). These results strongly suggest the role of stathmin phosphorylation in the mechanism of barrier protection by cAMP.

Figure 7.

PKA-dependent stathmin phosphorylation at S⁶³ is critical for protection against thrombin-induced permeability and cytoskeletal recovery. HPAECs transfected with phosphorylation-deficient stathmin (STMN-S63A) or empty vehicle were pretreated with cAMP (500 ng/ml, 15 min), followed by thrombin treatment (0.3 U/ml). A) TER measurements were performed over 3 h. B) Phosphorylation of MYPT and MLC was detected by Western blot after 15 min of thrombin treatment. Equal protein loading was confirmed by determination of β-tubulin content in total cell lysates. Bar graphs depict quantitative analysis of Western blot densitometry data. *P < 0.05 vs. empty vector (Em Vec); n = 4. C) Actin remodeling after 15 and 30 min of thrombin treatment was examined by staining with Texas Red phalloidin. Stathmin was visualized by immunostaining with stathmin antibody. Arrows indicate stathmin-overexpressing cells.

Stathmin modulates GEF-H1 microtubule association and activity in thrombin-stimulated ECs

The results described above strongly suggest a link between stathmin-dependent regulation of microtubule dynamics and stability and Rho-dependent mechanisms of actin cytoskeletal remodeling and EC permeability. Reports by our group and others implicate microtubule-associated Rho-specific GEF-H1 in microtubule-dependent regulation of Rho activity (10–12). GEF-H1 activation is achieved via its release from microtubules. Indeed, thrombin stimulation decreased the total amount of microtubule-bound GEF-H1, as shown by analysis of GEF-H1 content in the microtubule-enriched fractions (Fig. 8A). This thrombin-induced decrease was abolished by depletion of stathmin.

Figure 8.

Stathmin modulates GEF-H1 microtubule association and activity in thrombin-stimulated ECs. A) HPAECs were transfected with stathmin siRNA or nsRNA, followed by thrombin stimulation (0.3 U/ml, 15 min). Microtubule-enriched faction was isolated, and microtubule-associated GEF-H1 was detected by Western blot. Bar graphs depict quantitative analysis of Western blot densitometry data. *P < 0.05 vs. nsRNA; n = 4. B) HPAECs were transfected with stathmin siRNA or nsRNA, followed by transfection of constitutively activated GEF-H1 (GEF-H1-CA) or empty vector. TER measurements were performed while cells were stimulated with thrombin. C) HPAECs were transfected with GEF-H1 siRNA or nsRNA, followed by transfection of phosphorylation deficient stathmin (STMN-S63A) or empty vector. TER measurement was performed while cells were stimulated with thrombin. Veh, vehicle.

Next, experiments were performed to define the hierarchy of stathmin-GEF-H1 relationships in control of EC barrier function. Thrombin-induced permeability responses were examined in stathmin-depleted ECs expressing constitutively active GEF-H1 (GEF-H1-CA; Fig. 8B). As described above, stathmin depletion alone abolished thrombin-induced EC barrier compromise. In turn, ectopic expression of GEF-H1-CA abrogated the protective effect of stathmin knockdown on thrombin-induced EC permeability.

In reciprocal experiments, the stathmin-S63A mutant, which favors microtubule disassembly and exacerbates thrombin-induced EC disruption (Fig. 4), was expressed in control ECd and cells with depleted GEF-H1 (Fig. 8C). Consistent with GEF-H1 role as an activator of the Rho pathway of EC permeability, its molecular inhibition by gene-specific siRNA attenuated thrombin-induced TER decline. Notably, ectopic expression of stathmin-S63A did not alter the protective effect of GEF-H1 knockdown against thrombin-induced hyperpermeability. Taken together, these results suggest that activation of GEF-H1 is downstream of stathmin-dependent regulation of microtubule dynamics in the pathway of thrombin-induced EC permeability.

Stathmin knockdown attenuates lung injury in the 2-hit model of ventilator-induced lung injury

Recent reports suggest Rho signaling as a critical mechanism for increased vascular leak associated with ventilator-induced lung injury (14, 25). The results described above suggest the role of stathmin in the microtubule-dependent mechanism of Rho activity regulation. We next tested the effect of stathmin molecular inhibition on parameters of lung injury in the 2-hit model of ventilator-induced lung injury. C57BL6J mice were treated with nonspecific and stathmin-specific siRNA and exposed to HTV (30 ml/kg, 4 h) concurrently with intratracheal administration of TRAP. Lung injury was assessed by analysis of protein content and cell count in BAL fluid. TRAP/HTV caused a prominent increase in protein concentration (Fig. 9A), total cell (Fig. 9B), and neutrophil counts (Fig. 9C) in mice transfected with nsRNA. These increases were significantly attenuated by depletion of stathmin. Western blot analysis of lung tissue samples showed that stathmin knockdown abolished TRAP/HTV-induced tau phosphorylation at a Rho-kinase specific site and preserved the pool of stable microtubules, as detected by increased levels of acetylated tubulin in stathmin-depleted lungs (Fig. 9D). These results demonstrate the critical role of stathmin in the control of lung barrier function in the 2-hit model of ventilator-induced lung injury.

Figure 9.

Stathmin knockdown attenuates lung injury in the 2-hit model of ventilator-induced lung injury. A–C) Mice were transfected with nonspecific or stathmin-specific siRNA (72 h), followed by intratracheal administration of TRAP6 (1.5×10⁻⁵ mol/kg) and HTV (30 ml/kg, 4 h). Protein concentration (A), cell count (B), and neutrophil count (C) in BAL samples were analyzed. *P < 0.05; n = 6. D) Lung samples were analyzed by Western blot. Site-specific phosphorylation of tau, tubulin acetylation, and stathmin depletion were examined. Equal protein loading was confirmed by determination of β-tubulin content in lung samples. Bar graphs depict quantitative analysis of Western blot densitometry data. *P < 0.05 vs. nsRNA; n = 6.

DISCUSSION

Published data describe a role of stathmin in the control of many cellular functions, such as regulation of cell cycle and cell surface receptor-coupled kinase systems (26) and sarcoma cell migration and invasion (27), as well as modulation of migratory properties of GN-11 neurons in vitro (28) or adult neurogenesis (29). The most significant finding of this study is a demonstration of a stathmin-dependent mechanism regulating agonist-induced endothelial barrier disruption and lung vascular leak.

Direct evidence of stathmin involvement in the EC barrier regulation was observed in experiments with stathmin knockdown. The results showed a striking inhibitory effect of stathmin knockdown on thrombin-induced peripheral microtubule disassembly, shortening of growing microtubules, activation of the Rho/Rho-kinase pathway manifested by increased MYPT and MLC phosphorylation, and EC permeability. Besides actin cytoskeletal targets, activated Rho-kinase phosphorylated another microtubule-associated protein, tau. Tau stabilizes microtubules and promotes microtubule assembly, whereas Rho-kinase-mediated tau phosphorylation at T²⁴⁵, T³⁷⁷, S²⁶², or S⁴⁰⁹ decreases tau binding to microtubules and may cause further microtubule disassembly (22). These signaling events may represent an amplification cascade of Rho-dependent microtubule disassembly caused by barrier-disruptive EC stimulation.

GEF-H1 is an important microtubule-associated activator of Rho. Inhibition of thrombin-induced GEF-H1 dissociation from a microtubule-enriched fraction achieved by stathmin knockdown strongly suggests a critical role of stathmin in the regulation of Rho signaling by control of the GEF-H1 microtubule-bound state. The mechanistic link between stathmin and GEF-H1 has been further defined in double transfection experiments. Expression of a constitutively activated GEF-H1 mutant unable to bind microtubules abrogated barrier-protective effects of stathmin knockdown on thrombin-induced EC permeability, while thrombin effects on permeability in cells expressing the stathmin-S63A mutant, shown to exacerbate the thrombin disruptive response (Fig. 4), were abolished by depletion of GEF-H1. These results strongly suggest GEF-H1 as a major regulator of EC permeability downstream of stathmin effects on microtubule dynamics.

Supporting the results of our studies in EC cultures, our data also show that stathmin knockdown significantly improved the parameters of lung barrier function in the setting of TRAP/HTV-induced lung injury in vivo. Of note, siRNA-induced GEF-H1 knockdown also significantly attenuated lung injury in mice exposed to mechanical ventilation at HTV (14). Taken together, these data strongly support the stathmin-GEF-H1 mechanism of microtubule-dependent endothelial barrier regulation in vitro and in vivo.

Physiological regulation of stathmin function in vivo is achieved by phosphorylation. Phosphorylation of stathmin on one or more of its 4 serine residues (S¹⁶, S²⁵, S³⁸, or S⁶³) reduces its microtubule-destabilizing activity (13). Considerable levels of stathmin phosphorylation at S¹⁶ and S⁶³ were observed in nonstimulated EC monolayers (indicative of basal inhibition of stathmin activity). In turn, stathmin phosphorylation rapidly decreased after thrombin challenge but returned to basal levels during EC barrier recovery. The mechanism of such dephosphorylation is not yet clear. One plausible explanation is thrombin-induced elevation of intracellular Ca²⁺, which activates Ca²⁺-dependent endothelial phosphatase 2B (30), leading to rapid stathmin dephosphorylation.

Another interesting observation in this study is an ability of cAMP to attenuate agonist-induced endothelial hyperpermeability via suppression of stathmin activity. Cell treatment with cAMP promoted stathmin phosphorylation at S¹⁶ and S⁶³, which was essential for increased microtubule stability, attenuation of thrombin-induced Rho pathway, and improved EC barrier recovery. In turn, the expression of stathmin phosphorylation-deficient mutant prolonged thrombin-induced MLC phosphorylation and actin stress fiber formation, exacerbated permeability response to thrombin, and abrogated protective effects of cAMP pretreatment against thrombin-induced EC permeability. Published data suggest that stathmin phosphorylation at S¹⁶ and S⁶³ has important biological significance for stathmin regulation in vivo, as these sites can be phosphorylated by different kinases including PKA and PAK1 (31–33). Of note, both PKA and PAK1 can be activated directly or indirectly by cAMP (24, 34, 35). The mechanisms regulating stathmin phosphorylation status and activity require further investigation.

Using EB-1 as a tracker of growing microtubule plus ends, we performed a direct analysis of agonist-induced changes in the microtubule dynamics in live cells and found that the microtubule growth rate was neither affected by thrombin nor by cAMP elevation (data not shown). In turn, direct measurements of mean length values of growing EB-1-positive ends in live cells showed a marked decrease by thrombin and restoration of microtubule length and peripheral structure in cells pretreated with cAMP prior to thrombin challenge. These data suggest that thrombin increases the rate of microtubule catastrophe events at the cell peripheral compartments. It is tempting to speculate that decreased peripheral microtubule density in thrombin-stimulated cells is linked to submembrane elevations in intracellular Ca²⁺ and activation of Ca²⁺-dependent phosphatase activities, which cause stathmin dephosphorylation and microtubule destabilization at the cell periphery.

To summarize our findings and published data, we propose a scheme of stathmin-induced regulation of endothelial permeability and lung injury (Fig. 10). Barrier-disruptive agonists, such as thrombin, decrease the pool of phosphorylated stathmin, leading to stathmin-dependent microtubule disassembly. This causes release from microtubules and activation of Rho-specific GEF-H1. In turn, GEF-H1 activation stimulates the Rho-dependent pathway of EC permeability, which involves Rho-kinase mediated phosphorylation and inhibition of MYPT phosphatase activity, accumulation of phosphorylated MLC, actomyosin contraction, cytoskeletal remodeling, and formation of paracellualr gaps, resulting in EC barrier disruption. In addition, activated Rho-kinase phosphorylates tau at S²⁶² and S⁴⁰⁹, causing tau dissociation from microtubule and microtubule destabilization, thus triggering a positive feedback loop of microtubule-dependent Rho activation. Stabilization of microtubules, reduction of Rho signaling, and attenuation of pulmonary EC permeability and acute lung injury may be achieved by protein knockdown or cAMP-induced phosphorylation.

Figure 10.

Role of stathmin in thrombin-induced endothelial barrier dysfunction and acute lung injury. Thrombin stimulation decreases the pool of phosphorylated stathmin, which leads to stathmin-dependent microtubule disassembly, activation of microtubule-released Rho-specific GEF-H1, and stimulation of Rho-dependent pathway of endothelial permeability. This pathway involves Rho-kinase mediated phosphorylation and inactivation of myosin phosphatase MYPT1, accumulation of phosphorylated MLC, actomyosin contraction, cytoskeletal remodeling, and endothelial barrier disruption. Activated Rho-kinase also phosphorylates tau at S²⁶² and S⁴⁰⁹, leading to tau dissociation from microtubules and microtubule destabilization. This mechanism triggers a positive feedback loop of microtubule-dependent Rho activation. Stabilization of microtubules, reduction of Rho signaling, and attenuation of pulmonary endothelial permeability and acute lung injury may be achieved by stathmin knockdown or stathmin inactivation via cAMP-induced phosphorylation on S¹⁶ and S⁶³.

In summary, using comprehensive molecular, biochemical, imaging, and functional approaches, we identified a crosstalk between agonist-induced stathmin activation and Rho GTPase-dependent signaling via Rho-specific GEF-H1. The results of this study suggest a novel approach for prevention of lung barrier dysfunction via stathmin-targeted stabilization of the microtubule network. Interference with stathmin microtubule-depolymerizing activity using small phospho-mimicking peptides, siRNA-induced knockdown, or structure-based small molecule design approaches may prove to be more efficient and target-specific strategies in comparison to microtubule stabilizing plant alkaloids to prevent severe vascular leak accompanying many pathological conditions including pulmonary edema and acute respiratory distress syndrome.