Lack of durable protection against cotton smoke-induced acute lung injury in sheep by nebulized single chain urokinase plasminogen activator or tissue plasminogen activator

All studies involving animals were approved by the Institutional Animal Care and Use Committees of The University of Texas Medical Branch at Galveston. The same studies and exchange of biological samples were also approved by the Institutional Animal Care and Use Committee of The University of Texas Health Science Center at Tyler. The ovine model of cotton smoke-induced ISIALI in awake animals was used in these studies, as we previously reported [5]. In the model, a total of 48 puffs of cotton bark smoke cooled to < 40 °C in 4 sets of 12 breaths was delivered via the tracheostomy tube by a modified bee smoker to induce ISIALI. A total of 46 sheep with ISIALI were utilized in this study, including those that were ventilated by either adaptive pressure ventilation/controlled mandatory ventilation (APVcmv; Group 1, n = 14) or synchronized controlled mandatory ventilation, (SCMV; Group 2, n = 32) using a Hamilton G5 ventilator (Hamilton Medical, Inc., Reno NV). Interventional agents were nebulized using an Aeroneb^® Pro vibrating mesh nebulizer (Aerogen, Mountain View, CA). In addition, another 3 naive (No any interventions i.e., no surgical procedure, no injury, no tracheostomy and ventilation) animals and 2 sham animals ventilated with the APVcmv for 48 h were used as non-ISIALI controls, so that a total of 51 sheep were used in this study. The ventilator parameters included the tidal volume = 12 mL/kg; frequency = 20 breaths/min; Positive end expiratory pressure; peep = 5 cmH₂O; time of inspiration to total respiratory cycle time; i/e = 1:2; heated/humidified inspired gas to 33 °C, with fractional concentration of oxygen; FiO₂ = 1, with all parameters constant to 3 h post injury, then titrated to maintain a PaO₂ ≥ 100 mmHg with frequency adjusted to maintain the PaCO₂ close to 30–40 mmHg.

Arterial blood samples were obtained at 3 and 4 h after injury and thereafter every 4 h in Group 1, and 3 h and 6 h after injury and thereafter every 6 h in Group 2 for measurement of arterial blood gas analyses (blood gas analyzer RAPIDPoint 500; Siemens Healthcare, Erlangen, Germany). The PaO₂/FiO₂ ratio was measured to assess pulmonary gas exchange. Peak and plateau airway pressures, lung resistance and static lung compliance were measured at these same intervals throughout the 48 h course of the experiments. All physiologic assessments from each of the animals in each of the treatment groups were included. Animals were euthanized if the animals reached euthanasia criteria i.e., PaO₂/FiO₂ < 50, PaCO₂ > 90 mmHg or mean arterial pressure < 50 mmHg sustained for 1 h. A schematic illustration of the sample collection format is depicted in Fig. 1, which illustrates that blood samples from sheep in Group 2 were collected at 3, 6, 12, 18, 24, 30, 36, 42 and 48 h after initiation of ISIALI. Blood was not collected from animals in Group 1. Nebulized interventional agents were delivered at 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, and 44 h after ISIALI was induced, after which the sheep were sacrificed at 48 h.

Lung preparation, histologic assessment and scoring of airway obstruction and bleeding were done as we previously reported [5]. Briefly, the right lung was removed, and a 1-cm-thick section was taken from the middle of the lower lobe, injected with and then immersed in 10% formalin. Samples (n = 4) were taken at predetermined sites for histological examination. Fixed samples were embedded in paraffin, sectioned at 4 μm, and stained with hematoxylin and eosin. A pathologist without knowledge of the group assignments evaluated the lung histology. The pathologist then evaluated all bronchi, bronchioles, and respiratory bronchioles in sections of the four tissue samples, and for each, estimated the percentage of the surface area of the lumen obstructed by cast material (0–100%), representing the obstruction score. Other scoring of histopathologic indicators of lung injury were done as we previously reported [5].

Activase [lyophilized alteplase or recombinant single chain (sc); referred to as tPA] was provided as a generous gift from Genentech through the Genentech Research Contracts and Reagents Program to SI. The preparation of scuPA used in all experiments was non-GMP (good manufacturing practices) bulk drug substance that was manufactured under NHLBI SMARTT Contract # HHSN268201100014C. This preparation of scuPA was initially assessed by Advanced Bioscience Laboratories (ABL, Inc). The specific enzymatic activity of the scuPA preparation was about 150,000 IU/mg, and initial purity was assessed by reverse phase high pressure liquid chromatography to be 99.5% scuPA. Bacterial endotoxin levels were low but detectable at 0.23 EU/mg. The aliquoted samples of scuPA were stored in sodium acetate buffer pH 4.5, frozen at − 80 °C and activity and purity of the stored scuPA were validated to be preserved by determination of the specific activity of the material every 6 months.

To assess stability of the fibrinolytic agents during nebulization, samples of 4 mg of single chain tissue plasminogen activator; sctPA or scuPA in 8 mL of PBS were nebulized using an Aeroneb^® Pro (Aerogen, Mountain View, CA) and captured for the further stability and activity measurements as described previously [14]. The specific enzymatic activity of the scuPA and sctPA preparations after nebulization were each 95 ± 8% of the measured stock preparation activity. Additionally, the quality of scuPA and tPA preparations used for in vivo experiments was confirmed independently by measuring the fraction of the active enzyme based on its ability to form an inhibitory complex with PAI-1 as described previously [14]. The molar fraction of active enzymes in all preparations was greater than 0.90.

BAL was accomplished by wedging a Foley urinary catheter within a left lower lobe segment and introducing 50 mL of PBS via the catheter. The harvested return ranged from 5 to 20 mL with gentle suctioning. Cells were removed via centrifugation at 2500g at 4 °C for 10 min and cell-free lavage was immediately frozen and stored at − 80 °C. Citrated plasma was collected from animals in Group 2 only as previously reported [8].

Measurements of the uPA and tPA activity and antigen in BAL were done using ELISAs (Molecular Innovations, Novi, MI) per the manufacturer’s protocol. Fibrinolytic activity in BAL was analyzed as we previously reported [10]. BAL fluids (0.1 mL) were added over a FITC-fibrin film, which was pre-formed in the 96-well plate, and supplemented with 100 nM of human Glu-plasminogen. Fibrinolytic activity was monitored by fluorescence emission at 510 nm (excitation 490 nm) with time. Plasminogen activation assays were performed as previously described [10]. Fibrin enzymography was done as we previously reported [9].

Levels of active PAI-1 in the BAL samples were determined using an active rat PAI-1 ELISA (Molecular Innovations, Novi, MI) following the manufacturer’s protocol and by titrating samples of BAL fluid with uPA of a known concentration, as previously described [15, 16].

A ventilator simulation circuit, described in detail and illustrated in Additional file 1: Figure S1, enabled us to assess losses/inactivation of the plasminogen activators.

αM complexes with either scuPA/uPA or tPA were assessed as we previously reported [14]. Briefly, PAI-1 resistant uPA amidolytic activity, which represents intrapleural uPA in complexes with αMs, was measured after samples of BAL fluids were supplemented with an excess (100–200 nM) of exogenous recombinant human PAI-1 to inhibit free enzyme.

The data was analyzed using SigmaPlot version 12.3 for Windows (Systat Software Inc., San Jose CA) and Graphpad Prism version 6 (Graphpad Software, La Jolla, CA). For each ventilation group and nebulized plasminogen activator dosing modality (at 2, 4 and 8 mg), changes in the physiologic parameters (PaO₂, PaO₂/FiO₂ ratio, peak and plateau airway pressures, resistance and static compliance) between treatment groups (controls, tPA and scuPA) over the 48 h study duration where compared using the Friedman’s 2-way ANOVA by rank, post hoc, where there is statistical significance, by pairwise Wilcoxon signed-rank tests. The impact of changes in the ventilation strategy (Group 1/Group 2) and an increase in dose (from 4 to 8 mg irrespective of treatment type) on the biochemical parameters (BAL and plasma uPA or tPA activity, antigen, active PAI-1) after 48 h of treatment were explored using the Kruskal–Wallis one-way ANOVA on ranks with post hoc using the Dunn’s test. Subgroup correlation analyses were also performed between physiologic and biochemical parameters at specific dosing levels. Statistical significance for each test was assessed at the 5% level of significance.

The dataset(s) supporting the conclusions of this article are included within the article and its Additional file 1.

Of the 14 sheep committed to APVcmv mechanical ventilation during the progression of ISIALI (Group 1), the first 2 animals were treated with nebulized 2 mg tPA delivered q4h. Two animals were euthanized because of respiratory and/or hemodynamic compromise at 29 h and 45 h, respectively. While we previously reported that this dose of nebulized tPA delivered every 4 h mitigated physiologic impairment in a similar model of inhalation/burn exposure [5], there was no improvement in the PaO₂/FiO₂ ratios, peak or plateau airway pressures, static compliance or airway resistance versus vehicle treated controls (n = 5) throughout the 48 h course of injury. The PaO₂/FiO₂ ratios of control saline-treated mechanically ventilated animals with evolving ISIALI generally declined more rapidly than controls included in the 2004 report [5], suggesting that the severity of injury could have contributed to the lack of salutary responses. Because we did not have enough material to provide continuous nebulization, we tested increased; 4 mg, doses of the PAs in the model and compared the effects of treatment with nebulized tPA (n = 4) versus scuPA (n = 3) using the same administration schedule (Fig. 1). One animal in the control group deteriorated and required euthanasia at 24 h. All other Group 1 animals survived over 48 h. In sham mechanically ventilated controls not exposed to cotton smoke, there was no physiologic change over 48 h, with the PaO₂/FiO₂ ratio maintained at ≥ 500 and the static lung compliance was maintained throughout 48 h at about 1.5 mL/cmH₂O/kg (not shown).

Following the administration of nebulized plasminogen activators (tPA and scuPA) at the 4 mg unit dose (Fig. 2a–f), there was a statistically significant increase in PaO₂/FiO₂ ratios over time to 48 h for animals treated with scuPA (p = 0.02) and tPA (p < 0.001), compared to controls. These were likewise significant increases than controls independent of time (p = 0.04 and < 0.001), respectively (Fig. 2b). Similar significant improvements occurred with decrements in the peak and plateau pressures (Fig. 2c, d, p < 0.001 in each case, independent of time) and the scuPA group demonstrated lower peak pressures than the tPA group (p = 0.03, independent of time, Fig. 2c). Lung compliance did not significantly differ between the sheep treated with nebulized plasminogen activators and saline vehicle-nebulized controls (Fig. 2e). Lung resistance was significantly reduced in the scuPA 4 mg group (p = 0.045, Fig. 2f) but not in the tPA group. At 48 h, the differences in PaO₂/FiO₂ ratios and all other physiologic parameters did not significantly differ between the treatment and control groups.

Given the potential for ventilator circuit losses (Additional file 1: Figure S1) and because pathophysiologic benefits of the nebulized PAs did not extend to 48 h in Group 1, we switched to the SCMV format and limited suction to augment airway delivery of the nebulized PAs (Group 2, n = 32 sheep). In our initial Group 2 experiments (Fig. 3a–f), we treated animals with 2 mg nebulized tPA (n = 2), which did not yield significant improvements in PaO₂/FiO₂ ratios, compliance or peak and plateau airway pressures versus vehicle-nebulized controls (n = 9, including two animals that were euthanized prior to 48 h). Interestingly, nebulized tPA at 4 mg (n = 6, including one animal that died before 48 h) was likewise ineffective in improving any of these same parameters versus controls (n = 9), so that the beneficial effects noted in Group 1 could not be corroborated in this larger comparison of sheep undergoing SCMV with reduced as needed, suctioning. Two animals were then treated with tPA 8 mg and each developed airway bleeding, including one animal that developed diffuse parenchymal bleeding involving all lobes and near occlusion of the airway with blood superimposed on uncleared airway casts. The second animal in this group developed airway hemorrhage that resulted in frank blood appearing in the ventilator tubing. This animal had grossly observed tracheal and large airway lesions (not shown). Both of these animals survived to 48 h, but there were no statistically significant improvements in oxygenation or other physiologic parameters.

The effects of scuPA nebulization in Group 2 sheep are shown in Fig. 4a–f. While there were no differences in arterial oxygenation (Fig. 4a), the PaO₂/FiO₂ ratios of the scuPA 8 mg (n = 8, including one animal that died prematurely) were increased significantly versus controls independent of time (n = 9, p = 0.02). The sheep treated with nebulized scuPA at 4 mg dosing did not differ from controls (Fig. 4b). Peak and plateau pressures were significantly reduced versus vehicle treated ISIALI controls by nebulized scuPA at 4 or 8 mg as a function of time, (p < 0.001) or independent of time (p < 0.001) in all cases except for the plateau pressure at scuPA 4 mg; p = 0.03, (Fig. 4c, d). Static lung compliance was increased in both the 4 and 8 mg scuPA groups versus vehicle nebulized controls independent of time during the 48 h course of progressive ISIALI (p = 0.02 in each case). Lung resistance was significantly reduced compared to control overtime (p < 0.001), with scuPA 4 mg (p = 0.01) and scuPA 8 mg (p = 0.02) treated groups, independent of time. By 48 h, only the plateau pressure of sheep nebulized with 8 mg scuPA remained significantly different from saline-treated controls and no overt bleeding into the airway or into the ventilator tubing was observed in sheep repeatedly treated at this nebulized unit dose.

Levels of PA activity in samples of BAL collected at 48 h after treatment with nebulized tPA and scuPA at the same 4 mg dose were generally higher in Group 2 using the SCMV/limited suction ventilation strategy (Fig. 5a). Further increments in BAL PA activity was observed with an increase in the dose of the plasminogen activator from 4 to 8 mg (Fig. 5a). Improved drug delivery was also observed at higher dosing levels based upon measurement of tPA and uPA antigen in BAL where total antigen increased significantly (p < 0.001) overall as ventilation strategy changes from Group1 at the dose 4 mg through Group 2 at 4 mg to Group 2 at 8 mg dose of plasminogen activators (Fig. 5b). No human uPA or tPA activity or antigen levels were detected in the BAL of control sheep treated with nebulized saline (n = 6), sham ventilated sheep without smoke exposure and no physiologic alterations over 48 h (n = 2) or naïve controls with no ISIALI nor mechanical ventilation (n = 3, data not shown).

In additional subgroup analyses, there was a significant correlation (r = 0.83, p < 0.02) between improvement of the PaO₂/FiO₂ ratios and uPA-related activity in the BAL of scuPA 8 mg treated animals (n = 7). When the uPA activities of BAL 4 mg scuPA group (n = 4) and the 8 mg scuPA group (n = 7) from Group 2 were pooled, there was also a significant correlation; r = 0.71, between improvement of the PaO₂/FiO₂ and BAL uPA activity in the pooled group (n = 11; p < 0.001). In this group, the PaO₂/FiO₂ and BAL uPA antigen were also significantly correlated (r = 0.88, p < 0.001). Interestingly, the correlation (r = 0.98) between PaO₂/FiO₂ ratios and total BAL tPA activity of the tPA 4 mg nebulized sheep (n = 4) was likewise statistically significant (p < 0.001). However, there was no significant correlation; r = 0.23, p = 0.70, between improvement of the PaO₂/FIO₂ ratio and total tPA antigen in the tPA 4 mg treated animals. Active PAI-1 was readily detectable in the BAL in control saline-treated animals with ISIALI (Fig. 6).

We next sought to assess fibrinolytic activity readouts in the BAL and initially used fibrin enzymography to detect zones of fibrinolysis associated with the nebulized PAs; tPA or scuPA. In the treatment groups in Group 2, a trend towards increased BAL fibrinolytic activity was observed in the tPA or scuPA groups at 4 h after the last administered dose (Fig. 7). Fibrinolytic activity was not detected in sham or naïve controls (n = 2 and 3 respectively, data not shown). By fibrin enzymography, zones of fibrinolysis associated with tPA or uPA were identified in all of the sheep treated with either nebulized tPA or scuPA 4 mg in Groups 1 and 2 and surviving to 48 h (Fig. 7, inset). No such bands were identified in the control saline-treated animals (Fig. 7, inset, lane 3). These data indicate that active plasminogen activator was identified in the BAL of sheep in which ISIALI was induced and that were then treated with nebulized tPA or scuPA at this dose. We next performed an independent analysis of fibrinolytic activity in BAL and detected low levels by fluorogenic analyses in some of the animals treated with either tPA or uPA in Group 2 (data not shown). The results are likely in part attributable to the lesser sensitivity of this assay versus that of fibrin enzymography in which the detector gels lyse over the course of 24 h or more. Plasminogen supplementation was next used to augment the fibrinolytic signal, since BAL represents a dilute form of extravascular lung lining fluid in which low plasminogen levels could limit detection of fibrinolytic activity. When supplemented by plasminogen, the differences in the fibrinolytic activities of the BAL samples of the tPA or scuPA-treated sheep became more apparent (Fig. 7, bar plots). Although there were no significant differences in fibrinolytic activity of the BAL from the nebulized tPA or scuPA groups, greater signal was detected in BAL of the animals from Group 2. There was a trend of increased levels of fibrinolytic activity in the scuPA or tPA-treated animals versus saline-treated controls in Group 2. d-Dimer levels in the BAL of sheep from these treatment groups exhibited the same trends (Fig. 8).

All cotton-smoke exposed animals had characteristic changes associated with ISIALI, with histologic changes consistent with our prior reported findings in the model, with scoring of the extent of injury done as we previously described [5]. Representative histologic images from control and treated animals in Group 2 are provided in the Additional file 1: Figure S2. There were no significant changes between the saline vehicle control-treated sheep versus those treated with nebulized tPA or scuPA with regard to alveolar edema, alveolar neutrophil, overall septal cell infiltration, atelectasis or major bronchial inflammation or bronchial or bronchiolar obstruction scores. In the tPA 8 mg group, one animal with extensive airway and parenchymal lung hemorrhage had focal necrotic areas in the trachea with a suggestion of bacterial invasion that can occur in the model. Similar lesions were identified in the major airways of other animals in Group 2 that did not experience hemorrhage. The second animal that had blood in the ventilator lines had grossly observed areas of focal large airway hemorrhage at post-mortem examination. Residual large airway casts of varying sizes were observed in most sheep with ISIALI at the time of gross post-mortem examination (not shown).

We lastly measured uPA and tPA antigens in the plasma of the sheep treated with nebulized Group 2 scuPA, tPA and saline controls (Fig. 9a–f). In saline-nebulized controls with ISIALI, no human tPA or scuPA was detectable in plasma, as anticipated (data shown). Levels of uPA antigen (Fig. 9a, b) in plasma of animals treated with 4 and 8 mg scuPA, respectively, increase significantly over time (p < 0.001). Difference were more profound for animals treated with scuPA at 8 mg compared with those treated with tPA or scuPA at 4 mg at each time point where samples were collected, while levels for the scuPA at 4 mg were not significantly different from those tPA at 4 mg (Fig. 9a–c). Detectable levels of αM/uPA complexes were found in the scuPA-nebulized animals (Fig. 9d, e) with the greatest levels observed in animals that received the highest; 8 mg dose of scuPA. These complexes were not found in the plasma of sheep treated with nebulized tPA (data not shown). Interestingly, a relatively high level of plasma tPA; 0.24 ng/mL at 48 h, was found in the tPA 8 mg animal that experienced major airway hemorrhage proximate to this determination. Plasma levels of d-dimer were not significantly different between controls and either the scuPA or tPA-nebulized sheep (Fig. 9f). Of particular note, d-dimers were not increased in conjunction with bioactive αM/uPA complex formation in plasma of sheep treated with nebulized scuPA.