CONCENTRATIONS OF MACROMOLECULES IN THE AIR SPACE

TOP ABSTRACT CONCENTRATIONS OF... SMALL SOLUTE TRANSPORT GRAVIMETRIC MEASUREMENTS OF... IN VITRO STUDIES OF... SUMMARY REFERENCES

THE FIRST DIRECT EVIDENCE that mammalian lungs can absorb fluid from the airways was derived from observations made in fetal sheep. During most of gestation, the lungs secrete fluid, but, as birth approaches, a variety of hormones act to initiate fluid reabsorption. Experiments in 1971 by Normand et al. (40) reported that, although [³H]inulin (a polysaccharide) crosses capillary walls, it does not diffuse across the pulmonary epithelium into the plasma. In contrast, disaccharides, monosaccharides, thiourea, urea, and water diffuse from the air spaces into the plasma at rates inversely related to their molecular weight. Similar observations had been reported by Chinard in 1966 (9), and the studies of Taylor and Gaar (44) also indicated that the epithelium is much more impermeable than the pulmonary endothelium. However, Normand et al. (40) used decreases in the concentrations of [³H]inulin to calculate the rate at which fluid was secreted in the fetal lung. In 1978, Walters and Olver (45) used the same approach to study the effect of -adrenergic agents to induce fluid absorption from the air spaces. This represented the earliest indicator-dilution measurement of fluid absorption from mammalian air spaces. Walters and Olver (45) used changes in the concentration of ¹²⁵I-labeled albumin rather than [³H]inulin to measure fluid reabsorption. The equation used for this purpose was
where J_w is water flux; V₀ is the initial volume of fluid in the lungs; C_t and C₀ are the concentrations of ¹²⁵I-albumin at the beginning and end of the observation period, and t is the elapsed time. As indicated, absorption of fluid from the air spaces results in increases in the concentrations of the macromolecule in the air spaces and dilution of that in the perfusate. The observations of Walters and Olver (45) were in accord with the report of Enhorning et al. (21), who showed that an adrenergic agent given to pregnant rabbits reduced neonatal lung water, and that of Lawson et al. (34), who showed that epinephrine slowed secretion in fetal lambs. Reabsorption of fluid from fetal lungs appears to be initiated by increases in blood epinephrine and increases the activity of -adrenoreceptors on the pulmonary epithelium associated with increases in tissue cAMP (14, 46).

The macromolecular approach provides direct information concerning fluid absorption from the air spaces, both in vivo and in isolated, perfused lungs. Alterations in macromolecular concentrations in fluid instilled into the airways presumably reflect the transport activity of the epithelial cells lining the air spaces and airways. However, several caveats must be kept in mind. It has generally been assumed that a two-compartment model adequately describes these preparations (1, 2, 20, 27). However, fluid that is absorbed from the air spaces may not actually reach the vascular space and leave the lungs. This fluid may be retained within the cells or interstitium with swelling of these compartments. This is shown schematically in Fig. 1, in which the shift of fluid from the air spaces to the tissues ("A") would exceed that from the tissues to the vasculature ("B") when the pulmonary tissue became edematous. If an isolated lung is used in these studies, the volume of air space fluid that made its way into the vasculature can be followed by placing a macromolecular indicator in the perfusate. In the studies of Effros et al. (19), albumin labeled with Evans blue was placed into both the air spaces and perfusate. If it is assumed that there is no loss of this indicator from either the vascular or air space compartments, then increases in the airway concentrations indicate movement of fluid out of the air space, whereas decreases in perfusate concentrations indicate the movement of fluid into the perfusate. The volume of fluid entering the vasculature equaled that entering the perfusate, indicating that tissue volumes remained unchanged (19). Cellular and interstitial swelling would be expected if the lung were injured during the course of the experiments. Regardless of whether a vascular indicator is used, increases in air space concentrations can be used to quantify absorption of fluid from the air space compartment.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Schematic diagram of fluid reabsorption from the air spaces. As fluid containing relatively little of the macromolecular indicator (e.g., 125I-albumin) is absorbed from the air spaces (arrow A), concentrations of the indicator become concentrated in the air space compartment. Fluid is subsequently transferred from the tissues to the vasculature (arrow B), resulting in dilution of the perfusate. Shading is used to indicate concentration or dilution of the vascular indicator.

The osmolality of the air space and perfusate fluids should be monitored because drying can occur, resulting in an increase in the concentrations of both small and large molecules. Approximately one-half of the increase in air space concentrations in rabbits could be attributed to dehydration, which can lead to erroneously high estimates of fluid absorption (18). This is of particular concern in small lungs, such as those of mice, because the surface area-to-volume ratio is so great in these preparations. Evaporation could help explain the fact that reabsorption in mouse lungs appears to proceed much more rapidly than in rat lungs. Of course, rapid reabsorption of fluid from the air spaces could also be related to the high surface area-to-volume ratios associated with the smaller dimensions of alveoli in small animals.

If labeled albumin is used to track fluid reabsorption, care must be taken that the iodine label does not become dissociated from protein, something that might be promoted by proteases in the lungs. This should be checked at both the beginning and end of the experiments. If the lungs are injured, labeled albumin may leak across the pulmonary epithelium, resulting in underestimates of the rate of fluid absorption. Although dissociation of the label from the macromolecule must be monitored, this disadvantage is outweighed by the fact that none of the labeled indicator could have been derived from the tissues or plasma. Unlabeled protein concentrations (and surrogates such as refractive index) have been used as alveolar labels, but this approach necessitates additional studies to show that the airway protein concentrations are not influenced by epithelial secretion or transport from the vasculature. The air space proteins must be characterized, and an alternative label must be placed in the vasculature to show that there is no diffusion of protein from the vasculature into the air space (see below).

Matthay et al. (37) used the macromolecular approach to estimate fluid absorption from adult sheep lungs in 1982. Initially, these investigators instilled serum into the air spaces and measured increases in airway concentrations of proteins. Protein concentrations increased to levels that exceeded those in the vasculature, strongly suggesting fluid absorption. In subsequent studies, they used ¹²⁵I-albumin to measure fluid reabsorption. To determine whether protein leaked into the air spaces from the vasculature, ¹³¹I-labeled albumin was added to the perfusate. This represented a very small contribution to the final protein concentration. Furthermore, they did not detect the influx of other proteins into the air spaces by polyacrylamide electrophoresis.

	SMALL SOLUTE TRANSPORT

TOP ABSTRACT CONCENTRATIONS OF... SMALL SOLUTE TRANSPORT GRAVIMETRIC MEASUREMENTS OF... IN VITRO STUDIES OF... SUMMARY REFERENCES

The fact that fluid is absorbed from the air spaces does not in itself prove that an active, energy-consuming process is involved. The actions of inhibitors and promoters of transport lend strong support to this hypothesis, but their action in the lung may be more indirect than in experiments with epithelial cell monolayers. In the studies of Basset et al. (1, 2) and Effros et al. (19), sodium (Na) concentrations in the air space and perfusate compartment remained about the same, indicating isotonic absorption of fluid. Because absorption of fluid was virtually isosmotic, it could represent passive transport in response to hydrostatic flows from the air space to the interstitium. The argument for active transport would be more persuasive if it could be shown that Na can be transported against an electrochemical gradient. Effros et al. (19) showed that, under certain circumstances, Na transport from the air space to the perfusate persists, despite the fact that Na concentrations in the perfusate increase above those in the air space fluid. Provided that the experiment is not associated with the development of electrical potential differences, which could account for this transport, such observations would support the likelihood of active transport of Na. In the first set of experiments (20), the air spaces were filled with a solution in which the Na and chloride (Cl) in the perfusate were replaced with enough unlabeled mannitol to keep air space osmolality equal to that in the perfusate. Equal concentrations of ²²Na were then added to the perfusate and the air space fluid. Mannitol slowed absorption of fluid from the air spaces, but, over the course of 1 h, concentrations of air space ²²Na fell by an average of 35 ± 4 (SD)% of the initial concentration, whereas those in the perfusate increased by 16 ± 4%. This represented transport of ²²Na up a significant concentration gradient. During this interval, there was leakage of unlabeled Na into the air space fluid in the opposite direction (air space Na concentrations increased by 20 mM, and perfusate Na concentrations decreased by 10 mM). No movement of mannitol was observed.

In a second study designed to document movement of Na "uphill" against a concentration gradient, Effros et al. (19) added 150 mM mannitol to both the air space and perfusate compartments, and Na concentrations in these solutions were decreased from 135 to 90 mM to avoid hypertonicity. Na continued to move into the perfusate, despite the fact that the concentration of Na in the perfusate rose significantly above that in the air space solution.

Utilizing techniques developed by Goodman and Wangensteen (28), Goodman et al. (27) compared the transport of ²²Na and [¹⁴C]sucrose out of the air spaces. -Agonists accelerated the transport of ²²Na out of the air spaces but had no effect on the rate at which [¹⁴C]sucrose was lost from the air spaces, suggesting that intercellular transport was not influenced by these agents. Similar results were obtained by Basset et al. (1, 2), who used ¹²⁵I-albumin, ²²Na, and [³H]mannitol in their studies. Inclusion of an impermeant indicator such as ¹²⁵I-albumin is probably preferable, because passive leakage of ²²Na into the air spaces could conceivably occur through the epithelial cells and would not be reflected by a change in the diffusion of [¹⁴C]sucrose or [³H]mannitol.

	GRAVIMETRIC MEASUREMENTS OF FLUID ABSORPTION FROM THE LUNGS

TOP ABSTRACT CONCENTRATIONS OF... SMALL SOLUTE TRANSPORT GRAVIMETRIC MEASUREMENTS OF... IN VITRO STUDIES OF... SUMMARY REFERENCES

A gravimetric technique for measuring fluid reabsorption was introduced by Berthiaume et al. (3), who filled one lung with fluid containing 5% albumin and used the other lung as control. Wet (W) and dry weights (D) were measured for each lung. ¹²⁵I-albumin was added to the injected fluid to determine how much of the albumin was lost during the course of the study. In addition, Evans blue was incorporated into the air spaces to visualize the sites reached by the instilled fluid. The D of the albumin instilled into the experimental lung was multiplied by the fraction of ¹²⁵I-albumin remaining in the lungs, to yield the D of instilled albumin remaining in the lungs, which was then subtracted from the D of the experimental lung (D_e) to yield the initial D (D'_e). The following equation was used to calculate the weight of excess fluid remaining in the experimental lung at the end of the experiment (E)
where W_e is the W of the experimental lung, and W_c and D_c are the W and D of the control lung, respectively. E was then subtracted from the actual volume weight of fluid (W_inj), which was originally added to the air spaces to calculate the weight of fluid that was reabsorbed (W_inj  E).

One concern regarding the gravimetric approach is the possibility that protein might enter the lungs from the vasculature during the experimental period. As mentioned above, ¹³¹I-albumin was injected into the vasculature, and concentrations were measured in the experimental and control lungs at the end of the experiments. This proved to have a small effect on the data. Nevertheless, such studies must be included if both lungs are exposed to agents that would alter capillary and epithelial integrity. Once again, secretion of other proteins into the air spaces must be ruled out. Migration of inflammatory cells into injured lungs might contribute D to the lungs. Smedira et al. (43) estimated that this fraction of the lung weight was small in their studies but might become more troublesome if lung injury is severe and inflammatory cells are present in the perfusion fluid. W-to-D ratios (W/D) in the control lungs remained unchanged over the experimental period. Some change in the W/D of the control lungs would be expected if they were infused with agents that altered their permeability or transport properties.

Changes in the W of the lung may not accurately reflect changes in the volume of fluid in the air spaces. The onset of interstitial edema and cellular swelling could result in increases in the weights of the lungs, independent of any change in the volume of the instilled fluid remaining in the air spaces. Perfusate may also become trapped within the vasculature, adding to the weight of the lungs. The gravimetric approach is, therefore, not as specific as the indicator-dilution method for assessing alveolar function. Nevertheless, it can provide important information regarding fluid balance in the lung and could conceivably detect the onset of capillary leakage of fluid in lungs that continue to absorb fluid from the air spaces.

	IN VITRO STUDIES OF ALVEOLAR EPITHELIUM

TOP ABSTRACT CONCENTRATIONS OF... SMALL SOLUTE TRANSPORT GRAVIMETRIC MEASUREMENTS OF... IN VITRO STUDIES OF... SUMMARY REFERENCES

Much of our present understanding of active ion transport, barrier resistance, and fluid transport across isolated alveolar epithelium is derived from studies conducted over the last 20 years. These studies have provided insights into specific mechanisms of alveolar epithelial cell function, because isolated in vitro models allow more precise control of experimental variables. However, data obtained from isolated models must always be compared with and/or validated by complementary studies in intact and/or in vivo lungs. Such cross-confirmation between in vivo and in vitro models has permitted major progress in understanding alveolar pulmonary edema from bench to bedside since 1980.

A great deal of excellent work from several laboratories has appeared on these topics over the last 2 decades. In this brief overview, therefore, we have confined ourselves generally to studies of active Na transport across adult alveolar epithelium, relative to concepts of alveolar fluid homeostasis, primarily under normal conditions. Consequently, owing to space constraints, much outstanding work may not be specifically mentioned. We have tried, however, to highlight landmark and breakthrough developments, while summarizing many excellent contributions in this rapidly advancing field.

The earliest studies of an isolated "alveolar" epithelium took advantage of the hollow amphibian lung. Gatzy (23) utilized this preparation, whose cells morphologically resemble an amalgam of mammalian alveolar type I (AT1) and type II (AT2) cells, to demonstrate, in Ussing chamber experiments, spontaneous potential difference (SPD) of 10-20 mV (alveolar side negative), moderate transepithelial electric resistance (R_t) of 700-1,000  · cm², and short circuit current (SCC) of ~10 µA/cm². Interestingly, the entire SCC represents net Cl secretion, and no active Na flux was noted, suggesting that amphibian lung may be more akin to mammalian airway, as opposed to alveolar epithelium.

At about the same time, Misfeldt and coworkers (39, 41) improved methods for using mammalian epithelial cells in culture to study their barrier properties. On the basis of observations of fluid movement from apical to basolateral surfaces of primary mammary epithelial cell sheets to form "domes," they deduced that the cells must be performing active solute transport and forming occluding functions, thereby retaining those transport functions even under in vitro culture conditions (41). Misfeldt et al. (39) proceeded to grow a dog kidney epithelial cell line (Madin-Darby canine kidney) as monolayers on permeable supports. These monolayers, when studied in Ussing chambers, demonstrated SPD of 1.4 mV (apical side negative) and R_t of 84  · cm². These investigations provided an approach to the study of mammalian epithelial barriers in vitro that has led to numerous important insights into intact organ function.

A few years later, in a classic example of productive collaboration, Misfeldt and his group worked in alveolar epithelial cell research with lung scientists to produce a landmark paper relevant to alveolar fluid homeostasis (35). Mason et al. (35) harvested AT2 cells from rat lungs and plated them on both porous and nonporous substrata. On plastic culture dishes, numerous domes were formed, indicating that the cells were exhibiting active transport and forming occluding junctions. Pharmacological agents amiloride and ouabain inhibited dome formation, whereas terbutaline stimulated it, consistent with involvement of active vectorial Na transport from the apical to basolateral surface (24-26, 35).

When AT2 cells were plated on collagen-coated Millipore filters, grown to confluence, and studied in Ussing chambers, SPD of ~1 mV (apical side negative) and R_t of ~200  · cm² were noted (35). Apical amiloride and basolateral ouabain abolished SPD (although leakage of the agents through tight junctions could not be ruled out) and terbutaline-stimulated SPD. A series of papers utilizing this preparation was subsequently published in which pharmacological regulation of active transport was explored, the most important finding of which may have been the involvement of cAMP as a secondary intracellular messenger (10). Overall, these studies (10, 24-26, 35) allowed us to conclude that alveolar epithelium probably actively transports Na from the apical to basolateral surface, with Cl and water following passively, and that the ability to regulate this process could be therapeutically important. Of major significance is that these conclusions were tested and validated to the extent possible in numerous studies on intact mammalian lungs.

One concern that remained was the low R_t noted above, because the epithelium separating alveolar air and interstitial fluid-filled spaces would be expected to be much less leaky. Furthermore, direct measurement of ion fluxes was not possible (due to large passive leak), and pharmacological approaches always have their uncertainties. A breakthrough paper in this regard was made possible by the use of tissue culture-treated polycarbonate filters on which freshly isolated AT2 cells were plated (8). Under these conditions, SPD is 10 mV (apical side negative), SCC is 4 µA/cm², and R_t is >2,000  · cm². By measuring unidirectional ²²Na fluxes under short-circuit conditions, it was shown that the entire SCC was accounted for by active Na absorption. Active Na transport was decreased 70% by apical amiloride, decreased 100% by basolateral ouabain, and increased up to 100% by terbutaline (either side). The stage was thus set for in depth study in vitro of the properties in alveolar epithelial cells of their apical Na channels (both amiloride sensitive and insensitive), basolateral Na-K-ATPase (Na pumps), and occluding junctions.

Much has been learned from in vitro studies about Na channels and pumps in adult alveolar epithelial cells over the last decade (indeed, far too much to provide details here). Several groups made major contributions, including those of Berthiaume, Ingbar, Matalon, Sznajder, and others. Very briefly, highlights include short- and long-term regulation of adrenergic stimulation of active Na transport (38), possible requirement for activation of Cl channels by terbutaline to increase active Na flux (32), utilization of cutting-edge techniques (including patch-clamp and molecular biology) to define specific properties of amiloride-sensitive and -insensitive Na channels (31, 33, 36), and exploration of mechanisms for dopamine upregulation of active Na transport (29).

One major value of understanding mechanisms of alveolar fluid homeostasis lies in their potential therapeutic application. Very recently, Factor and co-workers (22) utilized gene transfer technology to insert additional -subunits of Na pumps into alveolar epithelial cells. This maneuver resulted in increased active Na transport across alveolar cell monolayers, a finding that has been shown as well in vivo (22). Although much development work remains, such observations are of major potential importance for providing new therapeutic approaches to management of pulmonary edema.

The in vitro studies described above lead immediately to the question of whether AT2 or AT1 cells, or both, are primarily responsible for alveolar transepithelial transport properties. It has been observed for some time (7, 35) that AT2 cells in culture morphologically come to resemble AT1 cells (their in vivo daughters), so determination of which phenotype is responsible for active transport becomes of considerable importance, especially with regard to lung injury. The development of AT1 cell-specific markers (13, 17, 42) allowed better determination of alveolar cell phenotype in vitro. It has been shown that, over time in culture, AT2 cell appearance and surfactant production are lost, whereas AT1 cell marker reactivity and cell morphology appear (4), consistent with recapitulation of in vivo cell lineage in vitro.

For some time, significant regulation of this apparent differentiation process remained elusive. However, several experimental developments have been reported that allow manipulation of phenotype in vitro. These include changes in cell shape (12), air vs. liquid interface (16), growth factors (6), and substratum properties (30). Regulation of differentiation in vitro utilizing these approaches not only provides us with the tools to investigate more precisely the transport properties of alveolar epithelium and its constituent cells, but may also allow development of innovative maneuvers to better manage alveolar edema in lung injury.

Of course, knowing phenotype in vitro with certainty is difficult without having an almost unlimited ability to probe cell-specific identity. Accordingly, very recently, two groups have made major new advances in successfully harvesting AT1 cells directly from rat lungs. Dobbs et al. (15) showed that AT1 cells have an extremely high cell membrane water permeability, a property of potential importance for alveolar fluid homeostasis. Borok et al. (5) demonstrated that Na transport proteins are expressed in freshly isolated AT1 cells. These studies are likely to represent only the beginning of direct exploration of AT1 cell biology.

	SUMMARY

TOP ABSTRACT CONCENTRATIONS OF... SMALL SOLUTE TRANSPORT GRAVIMETRIC MEASUREMENTS OF... IN VITRO STUDIES OF... SUMMARY REFERENCES

Many groups have contributed importantly to the rapidly advancing field of alveolar epithelial cell biology and physiology and the role of alveolar epithelium in fluid homeostasis. It is likely that investigations of fluid transport in intact lungs and in in vitro models will continue to be important tools in understanding pulmonary epithelial function. A recent National Institutes of Health Workshop Report (11) provides an excellent account of where we need to go now. As with all science, we have answered many questions, but the new questions that emerge from those answers are always the most exciting.