Acute lung injury (ALI), as occurs with prolonged mechanical ventilation, contributes to morbidity and mortality of critical illness, and studies on novel genetic or pharmacological targets are areas of intense investigation. Here, we systematically tested a murine model of ALI by using pressure-controlled ventilation to induce ventilator-induced lung injury. For this purpose, C57BL/6 or Sv129 mice were anesthetized and underwent tracheotomy followed by induction of ALI via mechanical ventilation. Mice were ventilated in a pressure-controlled setting at different inspiratory pressure levels (15–45 mbar) and over different times (0–90 min, 100% oxygen). As outcome parameters, we assessed pulmonary edema (wet-to-dry ratios), bronchoalveolar fluid albumin content, pulmonary myeloperoxidase activity, macrophage inflammatory protein-2, and pulmonary gas exchange. These studies revealed maximal differences in severity of lung injury between different mouse strains after 90 min of ventilation time at 45 mbar. Use of lower concentrations of inspired oxygen did not alter disease severity. Increases of CD73 transcript (5′-ectonucleotidase, pacemaker of extracellular adenosine production) or total pulmonary adenosine levels with mechanical ventilation were less pronounced in C57BL/6 mice, suggesting attenuated adenosine protection in C57BL/6 mice. Together, these studies demonstrate feasibility of this model to induce murine ALI.
- ventilator-induced lung injury
studies of tissue protection utilizing murine models of acute injury are areas of intense investigation (7–9, 11–16, 18–21, 24, 27, 31, 38, 39, 41). For example, previous studies have used intratracheal bleomycin (46), inhalation of cigarette smoke (43), hyperoxia (34), acute ischemia-reperfusion injury (47), oleic acid (6), endotoxin (32, 36, 41), or different murine pathogens (30, 41) to induce acute lung injury (ALI). In addition, several studies have used cyclic mechanical ventilation with increased tidal volumes (22, 33, 45) for the induction of ventilator-induced lung injury (VILI). As such, VILI closely resembles many aspects of ALI (acute setting, attenuated gas exchange, bilateral lung disease) and allows the study gene-targeted mice (2, 8, 26).
Previous studies have used volume-controlled mechanical ventilation to induce VILI. When using volume-controlled ventilation, relatively small tidal volumes specifically adjusted to the animals’ weight have to be delivered (between 5 and 20 ml/kg body wt). While murine ventilators can be used in a volume-controlled setting to ventilate mice over prolonged periods without injury (e.g., up to 8 h), an alternative ventilatory strategy to induce ALI resembles pressure-controlled ventilation. Here, we hypothesized that pressure-controlled ventilation can be utilized to induce a highly reproducible degree of injury. Based on previous studies showing differences in pulmonary mechanics between different inbred mouse strains (40) or different susceptibility to lung injury (25), we exposed two different inbred mouse strains (C57BL/6 or Sv129) to pressure-controlled mechanical ventilation. Different levels of inspiratory pressure and ventilation times revealed more differences between murine strains utilizing high inspiratory pressure levels (45 mbar) over a time period of 90 min in conjunction with a highly reproducible degree of ALI.
Murine mechanical ventilation.
All animal protocols were in accordance with the German guidelines for use of living animals and were approved by the Institutional Animal Care and Use Committee of the Tübingen University Hospital, the Regierungspräsidium Tübingen, or the University of Colorado Health Sciences Center, Denver. Mice on the C57BL/6 (C57BL/6NCrl) or the Sv129 (129S2/SvPasCrlf) strain were purchased from Charles River Laboratory. Mice were selected and allocated according to age, sex, and weight in a prospective manner. Animals were anesthetized with pentobarbital (70 mg/kg ip for induction; 20 mg·kg−1·h−1 for maintenance) and placed on a temperature-controlled heated table (RT; Effenberg, Munich, Germany) with a rectal thermometer probe attached to a thermal feedback controller to maintain body temperature at 37°C. In addition, all animals were monitored with an electrocardiogram (Hewlett Packard, Böblingen, Germany). Fluid replacement was performed with normal saline, 0.05 ml/h ip. Tracheotomy was performed in a supine position (8). The tracheal tube was connected to a mechanical ventilator (Servo 900C; Siemens, Germany, with pediatric tubing). Mice were ventilated in a pressure-controlled ventilation mode at different inspiratory pressure levels (15, 35, and 45 mbar) for different time periods (30–90 min). Respiratory rate and inspiratory to expiratory time (I:E) ratios were adjusted based on arterial blood gas sampling obtained by cardiac puncture in control animals to maintain a carbon dioxide partial pressure between 35 and 40 mmHg and a pH between 7.35 and 7.40 (Supplemental Table). All animals were ventilated with an inspired oxygen concentration of 100% (FiO2 of 1), except in studies on the influence of FiO2 on VILI. In a subset of experiments, animals were ventilated under deep anesthesia at 45 mbar until a cardiac standstill was observed in the surface electrocardiogram (8), and mice were subsequently euthanized.
To obtain bronchoalveolar lavage (BAL) fluid, the lungs were lavaged three times with 0.5 ml of PBS. (Supplemental data for this article is available online at the AJP-Lung web site.)
Measurement of albumin and MIP-2 in BAL fluid.
Albumin or MIP-2 concentrations were measured by ELISA. Values below the limit of the assay were considered “zero” (for details, see data supplement).
Pulmonary neutrophil sequestration was quantified using a myeloperoxidase (MPO) (for details, see data supplement).
Following ventilation with indicated settings, lungs were excised en bloc, and wet-to-dry ratios were determined (for details, see data supplement).
Transcriptional analysis and measurements of pulmonary adenosine levels.
Measurements of pulmonary CD73 transcript and adenosine levels are detailed in the supplementary data.
Blood gas analysis.
To assess pulmonary gas exchange, blood gas analyses were preformed in subsets of experiments by obtaining arterial blood via cardiac puncture(for details, see data supplement).
Histopathological evaluation of ALI.
Following ventilation at indicated settings, the mice were euthanized, and lungs were fixed by instillation of 10% formaldehyde solution, followed by histopathological elevation (for details, see data supplement).
Lung injury score and survival data are given as median (range); all other data are presented as means ± SD from four to six animals per condition. We performed statistical analysis using Student's t-test (2-sided, P < 0.05) or ANOVA to determine group differences. Lung injury score was analyzed with Kruskal-Wallis rank test. Kaplan Maier curves were compared using log rank test (Mantel-Haenszel) test. P < 0.05 was considered statistically significant.
Pressure-controlled ventilation to induce ALI.
C57BL/6 mice matched according to sex, age, and weight were ventilated at indicated settings using pressure-controlled ventilation (15, 35, and 45 mbar) over different time periods (0–90 min). In these studies, all animals were ventilated with an inspired oxygen concentration of 100% (FiO2 of 1) without positive end-expiratory pressure (PEEP). Outcome parameters of ALI included albumin leakage into the BAL, lung water content as assessed by wet-to-dry ratios, pulmonary neutrophil accumulation as measured by tissue MPO levels, and functional parameters of gas-exchange [ratio of arterial partial pressure of oxygen (PaO2) and fraction of inspired oxygen (FiO2)]. These studies revealed a highly reproducible degree of ALI in all examined parameters at 45 mbar inspiratory pressure (Fig. 1). Lung injury consistent with the definition of ALI (ratio of partial pressure of oxygen to the inspired oxygen concentration) (3) was achieved with ventilation times between 60 and 90 min at a pressure level of 45 mbar.
VILI is enhanced in C57BL/6 mice compared with Sv129 mice.
Previous studies had shown differences between inbred mouse strains with regard to their pulmonary mechanics (40). To test the usefulness of this model to determine such differences in the setting of ALI, we compared BAL albumin leakage and lung water content in C57BL/6 or Sv129 mice. As shown in Fig. 2A, Sv129 and C57BL/6 mice both showed increases in albumin leakage into the BAL fluid and in lung water content. However, these increases were more pronounced in C57BL/6 mice compared with Sv129 mice. Similarly, pressure-dependent increases in lung water were significantly enhanced in C57BL/6 mice (Fig. 2B). We next compared the levels of the pulmonary neutrophil accumulation by measuring MPO (Fig. 2C) or the proinflammatory cytokine MIP-2 (Fig. 2D) in C57BL/6 or Sv129 mice. In fact, PMN infiltration and MIP-2 levels were significantly more elevated in C57BL/6 compared with Sv129 mice. Similar findings were obtained when measuring pulmonary gas exchange (Fig. 2E). Based on previous studies showing a protective role of extracellular adenosine generation during VILI (8), we pursued the hypothesis that endogenous production of adenosine via the ecto-5′-nucleotidase (CD73) is attenuated in C57BL/6 mice compared with Sv129. As shown in Fig. 2, F and G, induction of CD73 transcript levels and increases of pulmonary adenosine levels were attenuated in C57BL/6 mice compared with Sv129, respectively. These studies suggest that ALI induced by mechanical ventilation is more severe in C57BL/6 compared with Sv129 mice and reveals differences in the generation of extracellular adenosine between both mouse strains. The latter could explain the differences in susceptibility to VILI.
Histological characteristics of pulmonary injury and gas-exchange in C57BL/6 or Sv129 mice.
As shown in Fig. 3, the differences between both mouse strains could be confirmed on a histological level. To investigate differences in pulmonary gas exchange during VILI as a surrogate parameter of function, blood gas analyses were performed. Together, these studies suggest that histological tissue injury is more severely impaired in C57BL/6 compared with Sv129 mice.
Influence of the inspired oxygen concentration on VILI in C57BL/6 or Sv129 mice.
Sv129 or C57BL/6 mice were ventilated with pressure-controlled ventilation at an inspiratory pressure of 45 mbar over indicated time periods (0 or 90 min) using different concentrations of inspired oxygen (20% or 100% oxygen). These studies confirmed differences between Sv129 and C57BL/6 mice with regard to pulmonary edema and lung inflammation during VILI (Fig. 4, A–D). However, different inspired oxygen concentrations did not alter these results, suggesting that differences between Sv129 and C57BL/6 mice during VILI occur independently of the inspired oxygen concentration in this model.
Different survival times of C57BL/6 or Sv129 mice during VILI.
Next, we exposed mice from both strains to VILI at an inspiratory pressure level of 35 (Fig. 4E) or 45 mbar (Fig. 4F). These studies confirmed the above differences between Sv129 and C57BL/6. To test the influence of PEEP on survival time in this model, we repeated these studies with the addition of 5 mbar of PEEP (Fig. 4G). Here we found optimal, better resolution of different susceptibility between both mouse strains. Together, these studies demonstrate suitability of pressure-controlled ventilation at high inspiratory pressure levels (45 mbar) to study survival times during VILI under deep anesthesia.
ALI significantly contributes to critical illness, as it occurs frequently (37) and carries a high mortality rate (44). Moreover, the only therapeutic interventions currently available are elimination of causative agents and supportive therapy (44). Therefore, murine studies of ALI to discover the contribution of specific gene products or novel pharmacological targets are areas of intense investigation. Such studies require models of ALI that are associated with highly reproducible outcome parameters, such as pulmonary edema, lung inflammation, histology, and functional parameters including gas-exchange or survival time. Here, we hypothesized that induction of ALI utilizing pressure-controlled ventilation may be associated with highly reproducible degrees of ALI in mice. Consistent with this hypothesis, we demonstrate here that pressure-controlled ventilation at high inspiratory pressure levels (45 mbar) over a time period of 60–90 min is associated with the induction of ALI, including bilateral pulmonary edema and attenuated gas exchange (3). Extension of these studies utilizing different inbred mouse strains (C57BL/6 or Sv129 mice) revealed that this model is reliable in resolving the influence of genetic factors on outcome parameters of ALI, including studies of survival. In addition, these studies revealed differences in CD73 induction and pulmonary adenosine generation as response to VILI. This could be an explanation for the increased susceptibility of C57BL/6 compared with Sv129 mice in this model.
Consistent with the present findings, previous studies have found different degrees of injury in mouse strains in models of acute injury. As such, previous studies described differences in pulmonary mechanics between different inbred mouse strains (40). Similarly, studies of murine myocardial resistance to ischemia revealed different degrees of susceptibility between inbred mouse strains to myocardial infarction induced by coronary artery occlusion. When exposed to similar ischemia time periods, myocardial infarction in CD1 mice was significantly larger than in C57BL/6 mice (9). Similarly, it is well established that differences in responses to renal ischemia have been observed when exposing mice with different genetic backgrounds to renal artery occlusion (1, 4). For example, a recent study searched for an in vivo model of intrinsic resistance to renal ischemia, as this would provide an invaluable tool to investigate putative protectors from renal injury. This study showed that Brown Norway rats are virtually resistant to renal ischemic injury, thus providing a novel model for studying mechanisms of renal protection from acute renal failure (1). Moreover, recent studies of hepatic ischemia revealed different susceptibility between inbred mouse strains to liver ischemia (23). Together, these studies reveal dramatic differences in different models of acute injury between inbred mouse strains.
Our studies of pulmonary adenosine production suggest an inverse correlation between the degree of lung injury and the capability to produce extracellular adenosine. These data suggest a protective role of extracellular adenosine production and signaling in lung protection during VILI. This is consistent with other studies using genetic and pharmacological approaches to study extracellular adenosine in lung protection. For example, studies in mice with genetic defects in extracellular adenosine production (cd39−/− or cd73−/− mice) show enhanced susceptibility to ALI induced by mechanical ventilation (8). Moreover, studies comparing gene-targeted mice for individual adenosine receptors (ARs) found a selective role of the A2BAR in attenuating injury during high-pressure mechanical ventilation (10). Consistent with other studies on adenosine in enhancing alveolar fluid clearance (17, 28), these studies revealed a role of A2BAR signaling in enhancing alveolar fluid clearance during VILI, suggesting that A2BAR signaling protects during VILI by drying out the lungs (10).
The present studies suggest the usefulness of pressure-controlled ventilation to induce ALI. In fact, the present data show rapid and reproducible lung injury within hours of mechanical ventilation. However, some limitations should be pointed out. First, previous studies have shown that ventilating animals with inspiratory pressures and zero PEEP results in very rapid massive injury, which is associated with mechanical disruption of lung structures (29). This is somewhat different from the modulation of innate immune responses without mechanical disruption seen at lower inspiratory pressures (5). However, the model used in the present studies allows the investigation of additional parameters, such as survival time, and more closely resembles the severe degree of impaired lung function seen in patients with ARDS. For studies to address more subtle innate immune responses (independently of disruption of lung structures), pressure-controlled ventilation at lower inspiratory pressure levels (e.g., 20 or 25 mbar) could be considered. Second, the ventilatory parameters required for the high-pressure model are only rarely applied in patients. However, particularly in European countries, pressure-controlled ventilation remains a common treatment form of ARDS in many patients (35) or is clinically applied during short-term recruitment maneuvers (42).
Together, the present studies demonstrate feasibility of pressure-controlled high-pressure ventilation to induce ALI as assessed by pulmonary edema, lung inflammation, gas exchange, and survival time. In addition, the present findings of significant differences between mouse strains during VILI underscore the role of genetic factors in susceptibility to ALI and suggest that it is critical for murine studies of ALI to carefully choose appropriate control groups (e.g., matched littermate control mice).
This work was supported by Fortune Grant 1416-0-0, IZKF Verbundprojekt 1597-0-0, DFG Grant EL274/2-2, a DFG Fellowship to A. Grenz, and a Foundation of Anesthesia Education and Research grant to H. K. Eltzschig and IZKF Nachwuchsgruppe 1605-0-0 to T. Eckle.
We acknowledge Marion Faigle and Stephanie Zug for technical assistance and Dr. Michel Mittelbron for histological analysis of tissue sections.
↵* T. Eckle and L. Füllbier contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society