Discussion
Recent studies have used the MouseOx Plus for a variety of reasons including demonstration of hypoxaemia in transgenic mouse models,17 differential SaO2 following mechanical ventilation18 and monitoring of oxygen saturation post toxic gas inhalation.19 However, to date, no study has used this system to monitor arterial blood oxygen saturation following IT instillations of LPS, a well-characterised model of ALI/ARDS. This study demonstrated that infrared pulse oximetry can monitor the decline in SaO2 following IT-LPS and highlights for the first time the effect of cumulative neutrophil recruitment over several days, which results in oxygen saturation to decrease to 81.1%. A surprising result was that the lowest SaO2 levels were observed 96 h post IT-LPS instillation, a time point when markers of pulmonary injury and inflammation were returning to normal. This lag in lung function decline has not been appreciated before but implies that resolution of lung injury worsens oxygenation in mice. The mechanisms for this change are uncertain but may relate to restoration of blood flow to damaged areas of lung resulting in increased ventilation-perfusion mismatch.
An important finding of this study is that SaO2 correlates with well-characterised markers of pulmonary injury and inflammation used to assess the extent of ALI/ARDS in mice following IT-LPS. Indeed, SaO2 was the only cardiopulmonary parameter measured by the MouseOX Plus that correlated with the specific marker of alveolar epithelial cell damage, BAL RAGE expression. This may suggest that blood oxygen levels can be used in future experiments to verify whether quantitative lung injury markers directly affect lung function. However, a limitation of this study is that arterial blood samples were not analysed directly, in tandem to pulse oximetry to corroborate SaO2 readings. Changes in arterial PO2 have been shown recently to correlate to SaO2.18 In addition, our data suggest that SaO2 readings continue to drop from 48 to 96 h after IT-LPS even though heart and breath rate remain unchanged at these time points (figure 1C–E). Taken together these data suggest that SaO2 monitored during this study is reflective of the relative oxygen saturation within the artery.
Our data also suggest that RAGE expression increased 24 h post IT-LPS even though systemic hypoxaemia as determined by SaO2 was unchanged compared to PBS-treated controls. Previous in vitro experiments have suggested that RAGE expression is regulated by hypoxia by HIF1α,20 although to date type 1 lung epithelial cells have not been tested. Therefore, these data may reflect hypoxia-independent enzymatic cleavage and/or cytokine-induced RAGE release from epithelial cells or simply an effect of local tissue hypoxia prior to systemic hypoxemia.
Refractory expression of BAL VEGF was observed during the latter stages of lung repair following IT instillation of LPS. VEGF is predominantly expressed by alveolar type II cells in the lung,21 with contributions from macrophages and neutrophils during inflammatory responses.22 In this context, the role of VEGF in the lung is as a potent stimulus for endothelial and epithelial repair.23 ,24 The decrease in VEGF observed from 96 h in this model closely resembles the reduced VEGF levels observed in patients with ALI, which may be associated with impaired repair responses or reflect specific loss of alveolar type II cells following injury.25 Decline in lung function was also maximal at 96 h. This may suggest that future experiments using murine IT-LPS as a model of ALI should monitor this time point in particular. Although we did not extend our time points further than 9 days, it would be interesting in future studies to focus on this phase given that in addition to reduced VEGF levels, BAL neutrophilia and PPI remained elevated compared to resting levels at day 9 post IT-LPS. This was observed even though the levels of inflammatory chemokines associated with leucocyte chemotaxis such as CXCL1/KC and MIP-2 had returned to pre-IT-LPS levels. Taken together, these data are suggestive of permanent damage to alveolar-capillary barriers as a result of IT-LPS. A similar observation can be seen in other experimental models of inflammation when tissue function recovers but restoration of leucocyte populations do not return to predisease levels.26
Using pulse oximetry, we found the heart rate of our mice to average around 650–700 bpm (figure 1D). This mirrors recently published data which also used the MouseOx Plus system.27 However, resting mice have been previously shown to average 400–600 bpm.28 ,29 Heart rates of this magnitude were most similar to those of mice 24 h post IT-LPS. The MouseOx Plus system we used involved collar clips being placed directly on the animal and allowing them to wander freely in a cage. Therefore, we would suggest this to be a more sensitive reading of cardiac output than older technologies.
As a consequence of lung injury, direct readings obtained by the MouseOx Plus system became easier to monitor (more error-free data points) mainly due to the reduction in the mouse's activity. At extended time points, such as day 9, readings were more challenging not only due to the increased activity of the subject, but also due to regrowth of the hair around the neck and shoulders which can retard the infrared signals. These issues should be considered in future experiments using the MouseOx Plus system.
In conclusion, this study is the first to measure multiple quantitative markers of lung injury and inflammation alongside non-invasive monitoring of cardiopulmonary parameters during a mouse model of ALI. Our data revealed that lung function decline is maximal at 96 h post IT-LPS and that well-characterised indices of lung injury and inflammation correlate with SaO2. Pulse oximetry readings are easy to measure, and can be carried out with minimal stress to the animal, providing real-time data indicative of lung function as assessed by SaO2. Therefore this parameter may have the potential to predict outcome, help ensure humane endpoints are maintained and reduce animal usage by identifying points at which lung function differs from expected results.