© 2005 American Thoracic Society DOI: 10.1165/rcmb.F306
Thrombin Increases Lung Water by Decreasing Na,K-ATPase Activity
a Departments of Anesthesiology, Physiology and Biophysics, School of Medicine, University of Alabama at Birmingham, Birmingham, Alabama
In an article published in this issue of the AJRCMB (pp. 343–354), Vadász and coworkers (1) present evidence indicating that intravascular thrombin decreases the clearance of alveolar fluid across isolated-perfused rabbit lungs by promoting endocytosis of the Na,K-ATPase. In an elegant series of studies using isolated perfused lungs, as well as A549 and primary alveolar type II cells, the authors demonstrated convincingly that thrombin decreased levels of Na,K-ATPase at the basolateral membranes of A549 and ATII cells by activating NADPH, resulting in increased production of reactive oxygen species and activation of protein kinase C- This study is being highlighted in this issue of the Journal for a number of reasons: First, as the authors point out, there is considerable interest in identifying the mechanisms by which coagulation proteinases such as thrombin contribute to the pathogenesis of Acute Lung Injury (ALI) and Adult Respiratory Distress Syndrome (ARDS). Thrombin is a serine protease, a class of enzymes known to alter ion transport across intestinal, ocular, airway, and alveolar epithelial cells by activating protein activated receptors (PAR), phospholipases C (PLC), or tyrosine kinase pathways (2–5). However, presently the physiologic effects of thrombin on ion transport across the intact lung, as well as the mechanisms of its actions, have not been adequately elucidated.
Second, the investigators performed a number of complimentary physiologic, electrophysiologic, and biochemical measurements to identify both the physiologic consequences of thrombin administration to the development of lung injury and plausible mechanisms of its actions. To start with, they demonstrated that intravascular administration of thrombin damages both the pulmonary microvasculature, as shown by an increase of the capillary filtration coefficient, and the alveolar epithelium, as shown by the decrease of amiloride-sensitive alveolar sodium transport and the ability of the isolated lungs to clear aerosolized fluid. Then, by an ingenious set of experiments using both A549 and freshly isolated mouse alveolar type II (ATII) cells, they showed that the decrease in alveolar fluid clearance was not the result of damage to the apically located epithelial sodium channels (ENaC): protein levels of two of the ENaC subunits remained unchanged in biotinylated (membrane) fractions of thrombin-treated A549 and ATII cells. Also, in patch clamp studies, thrombin-treated mouse ATII and A549 cells exhibited unchanged whole cell currents as compared with untreated cells, indicating that apical pathways for entry of Na+ ions were unimpaired. On the other hand, thrombin decreased both the Na,K-ATPase activity of A549 and ATII cells, as measured by the ouabain-sensitive portion of 86Rb+-uptake, as well by levels of the This study also raises a number of basic physiologic questions. First, since the vectorial transport of Na+ ions across the alveolar epithelium requires the coordination of both apical and basolateral pathways (10), one would expect that a decrease of the Na,K-ATPase activity should lead to a downregulation of Na+ transport across apical pathways by decreasing either ENaC levels or the electrochemical gradient across them. When cells are patched in the whole cell mode, the movement of Na+ ions is governed by applied electrochemical gradients across their apical membranes. In this case, measurements of the amiloride-sensitive Na+ fluxes across isolated cells are needed to ascertain whether or not Na+ entry was compromised in vivo (11). Second, in contrast to other serine proteases, such as trypsin and elastase, which caused either a transient or steady increase of Na+ transport across confluent monolayers of airway cells, thrombin (which activates PAR1, PAR3, and PAR4) receptors had no effect on vectorial sodium transport across a variety of human airway epithelia (4, 5, 12). Clearly, Vadász and colleagues (1) showed that intravascular thrombin decreased active sodium transport and alveolar fluid clearance in isolated perfused rabbit lungs. However, the high values of epithelial lining fluid (ELF) indicate that the isolated perfused lungs used in their studies may have been damaged, possibly to the lack of lymph flow. Thus it is possible that thrombin decreases vectorial Na+ transport across the damaged, but not the normal alveolar epithelium. The investigators also attempted to measure in vivo fluid balance on the epithelial surface of the lung, and for this they are to be commended. To accomplish this, they performed bronchoalveolar lavage (BAL) and then estimated the volume of recovered ELF. Assuming that full mixing takes place, and knowing the fraction of instilled fluid recovered, allow the investigators to estimate the volume of ELF originally in the lung. All methods used to estimate ELF, however, have limitations. Exogenous markers added to the BAL fluid, which are then diluted by the recovered ELF, have been used, but overestimate recovery as the markers tend to be adsorbed to lung tissue (13). Several endogenous markers have been used. If the lavage is performed rapidly enough, urea can be used to estimate recovered ELF to within a factor of two or so (14). If the lavage is not performed rapidly, however, urea can diffuse into the BAL fluid from the circulation and the lung tissues resulting in an overestimate of recovered ELF (14). To assure full mixing, the investigators infused and aspirated a 30-ml volume three times. This differs slightly from the BAL method previously used by the same investigators, in which a 50-ml lavage was performed once (15). Presumably the choice of method represents a balance between complete mixing, which is required to estimate total lung ELF, and diffusion, which will increase the estimate of ELF. Sodium, the endogenous marker used by the investigators, may be a better choice, as it diffuses more slowly across epithelial surfaces than does urea (16). Even using Na+, the ELF estimate may be off by 40%. The use of Na+, moreover, assumes that its concentration in the ELF is the same as in plasma, or, in the case of the isolated perfused lungs, in the perfusate. This assumption is supported by direct estimates using micropuncture methods (17). The latter studies, which are appropriately widely cited, attempted to insert micropipets into the ELF on the surface of alveoli in intact lung. These demanding experiments raise the possibility that the fluid sampled may have been partially contaminated by interstitial fluid or plasma, particularly as the concentration measured in the ELF was the same as that in plasma. More recent studies have suggested that the ELF in the airways has a Na+ concentration less than that of plasma by 30–40% (18, 19). If the ELF Na+ is less than that of plasma (or perfusate), the estimated ELF volume would be too large. Thus, there are several methodologic considerations that could result in an overestimate of recovered ELF volume. For all these reasons, the volume of ELF estimated in the lungs of normal rabbits, 0.66 ± 0.08 ml is likely an overestimate. Strictly speaking, the investigators should probably refer to "apparent ELF." However, these methodologic issues are more of a problem for those who wish to know the absolute lung fluid volumes, than for those who wish to understand factors that regulate fluid flux. A more serious issue is the effect of the isolated perfused system used, which results in an increase in the ELF volume to 2.41 ± 0.31 ml. In fact, the change in ELF volume caused by the isolated-perfused lung system is considerably larger than that caused by the experimental interventions. This strongly suggests that the investigators are studying an injured lung, which may be considered a plus if they want to draw inferences of the effects of thrombin on the pathophysiology of ARDS. The key point, however, is that the investigators are able to quantify changes in fluid volumes and thus address the effects of thrombin on fluid balance. In summary, this article offers a complete picture of the physiologic effects of thrombin on lung fluid clearance: measurements across isolated rabbit lungs show that thrombin causes pulmonary edema by decreasing active sodium transport, while biochemical and physiologic studies establish a putative mechanism for the observed effect. As expected, these studies also raise a number of questions that will prompt additional investigations in this area. Footnotes Conflict of Interest Statement: Neither author has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. References
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
. 
(the catalytic) subunit of the Na,K-ATPAse in biotinylated (membrane) fractions. The authors should be commended for measuring not only Na,K-ATPase protein levels but also pump activity: a recent study showed that a 50% decrease of Na,K-ATPase resulted in no significant decrease of basal alveolar fluid clearance across the alveolar epithelium of mice (
1 but not