INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

RELATIVELY LITTLE IS KNOWN about the distribution of solute and water exchange across the arterial, capillary, and venous segments of the pulmonary vasculature. Differences in the morphology of these regions suggest that important differences might exist regarding the manner in which transport occurs in these vessels (9). Stop-flow studies have been used to characterize the longitudinal distribution of solute absorption and secretion in the nephron (4), but this approach has not been widely applied to the pulmonary vasculature. Stop-flow studies were used by Piiper (7) to monitor gas exchange in the lungs. Chinard (1) also referred to its use in a review published in 1980 to study ²²Na⁺ and ³⁶Cl exchange in the pulmonary circulation. However, interpretation of the latter data was complicated by uncertainties concerning the volume of the collected perfusate in the vasculature during the stop-flow interval.

In a recent report we used anterograde and retrograde stop-flow experiments to compare the sites of filtration and diffusion in the pulmonary vasculature. In these studies, isolated lungs were perfused with an intravascular indicator (FITC-dextran with a molecular weight of 2 × 10⁶) (3). Flow was then discontinued, and intravascular pressures were increased to 20 cmH₂O for 10 min. At the end of this time, a small volume of an isotonic solution containing ³HOH was instilled into the air spaces to effectively label the fluid that was in the exchange sites of the lungs during the stop-flow interval. The fluid remaining in the vasculature was then flushed from the lungs into serial sample tubes, and concentrations of FITC-dextran were used to determine how much fluid had been lost from the vasculature by filtration. Comparison of net filtration of unlabeled water from and diffusion of ³HOH into the collection samples suggested that significant filtration occurs from venous vessels in the lungs but not from arterial vessels. This approach represents an important advantage of using stop-flow studies in the lungs, since retrograde flow is not possible in the nephron.

There are a number of potential disadvantages to using instilled ³HOH to label the exchange vessels in the lungs: it is difficult to ensure uniform delivery to the pulmonary capillaries, ³HOH remaining in the air spaces may continue to equilibrate with the fluid used to flush the vasculature after the perfusate that was in the exchange volume has been collected, and instillation of fluid into the air spaces could alter regional lung perfusion. We proposed that an alternative method for labeling the pulmonary exchange volume would be to perfuse the lungs with a solute that is taken up and retained within the tissues during the stop interval (3). If virtually all the indicator were removed from the perfusate in the exchange vessels of the lung, then decreases in indicator concentration could be used to determine how much of each collection sample was in the exchange vessels during the stop interval. This, in turn, could help document the site of filtration and the sites of uptake or production of metabolites in the pulmonary circulation. We previously obtained evidence for efficient uptake of ²⁰¹Tl⁺ in perfused lungs (2), suggesting that this K⁺ analog could be used for this purpose. In the present study, ²⁰¹Tl⁺ and ⁸⁶Rb⁺ were evaluated as potential indicators for labeling the pulmonary vascular exchange volume in stop-flow studies, and we have compared their uptake with that of other indicators that diffuse into the pulmonary tissues.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Experimental approach. Rat lungs were initially perfused at a constant rate with an unlabeled solution. After several minutes the perfusate was changed to a solution that contained ¹²⁵I-albumin, as a reference intravascular indicator, and two or more of the following low-molecular-weight indicators: [³H]mannitol, [¹⁴C]urea, ³HOH, ⁸⁶Rb⁺, or ²⁰¹Tl⁺ (Table 1). Stop-flow experiments were divided into three stages: 1) during the prestop (loading) interval the lungs were perfused at various flow rates with a physiological solution containing a mixture of the indicators, 2) flow was arrested and left atrial pressures were maintained at 0 or 10 cmH₂O during the stop interval, and 3) the fluid that was in the lungs during the stop interval was flushed out during the poststop (flush) interval by resuming perfusion with the labeled solution.

Surgical procedure. Lungs were harvested from 52 Sprague-Dawley rats [average weight 441 ± 61 and 476 ± 71 (SD) g in low- and high-flow studies, respectively, P = NS]. The rats were anesthetized to deep pain with an intraperitoneal injection of 0.6-0.75 ml of 65 mg/ml pentobarbital sodium. After anesthetization the chest was opened and the rat was anticoagulated with an intracardiac injection of 0.4 ml of 1,000 U/ml heparin sulfate. The trachea was intubated, and catheters were placed in the pulmonary artery and left atrium before excision of the heart and lungs from the chest. The lungs were kept inflated with 95% O₂-5% CO₂ at a pressure of 5 cmH₂O.

Perfusate solutions and lung weights. The initial, unlabeled perfusate contained 135 mM Na⁺, 120 mM Cl, 25 mM HCO₃, 4 mM K⁺, 2.5 mM Ca²⁺, 0.8 mM Mg²⁺, 5 g/dl BSA, 150 mg/dl glucose, 10 mg/dl urea, and 5 mg/dl mannitol. The labeled solution was made by adding to each milliliter of the unlabeled perfusate three or more of the following radionuclides: 0.05 µCi ¹²⁵I-albumin (used in all experiments, human; Mallinckrodt, St. Louis, MO), 0.2 µCi ²⁰¹Tl⁺ (used in all experiments), 0.2 µCi [³H]mannitol, 0.2 µCi [¹⁴C]urea, 0.2 µCi ⁸⁶Rb⁺, and 0.8 µCi ³HOH (New England Nuclear, Wilmington, DE). More than 98% of the ¹²⁵I-albumin radioactivity was retained by a 30,000-mol-wt-cutoff Amicon filter, indicating that the ¹²⁵I remained associated with the protein molecule. The pH of the fluids were adjusted to 7.4 with small volumes of 1 N NaOH in the presence of 94% O₂-6% CO₂, and PCO₂, PO₂, and pH were measured at 37°C. Gas concentrations were measured in two preparations (not included in Table 1) before and after the stop period. The pH remained constant, and PCO₂ remained at 36-40 Torr. PO₂ remained above 120 Torr. Increases in K⁺ concentrations were achieved by replacing NaCl with KCl in the perfusate solution.

Analysis and calculation. Aliquots (50 µl) of the outflow fluid were placed into 3 ml of scintillation fluid (LKB Optiphase Hisafe 3, Pharmacia) and counted first in a gamma counter (for ¹²⁵I-albumin and ²⁰¹Tl⁺ radioactivity) and then in a beta counter (for [³H]mannitol, ³HOH, ⁸⁶Rb⁺, and [¹⁴C]urea radioactivity). These counts were corrected for crossover and ²⁰¹Tl⁺ decay, and the outflow counts per minute were divided by those in the arterial perfusion solution to yield outflow-to-inflow concentration ratios.

(1)

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Schematic diagram of volume (V) calculations. Vev,mannitol was calculated as shaded area between 125I-albumin and [3H]mannitol curves with Eq. 1. Similarly, Vev,urea and Vev,water were calculated from areas between 125I-albumin and urea and water curves (not shown). Fraction of each venous sample that was in exchange vessels of lungs during stop-flow interval (vexch/vven) was calculated from observed and interpolated 201Tl+ curves with Eq. 3. Total exchangeable volume in vasculature (Vexch) was calculated with Eq. 4. Values calculated for Vexch can equal true volume of exchange vessels only if indicator is completely removed from vasculature during stop-flow interval. In an anterograde flow study (flushed from artery to vein), 1st portions of exchangeable volume of pulmonary vessels recovered are from more venous vessels than later samples. Volume of perfusate collected is plotted on abscissa and is equal to product of flow and time. Vev, extravascular volume; a, time at which infusion solution is switched to labeled solution; b, time at which concentrations of outflow and inflow have become the same; d, time when flush begins; e, time when fven becomes equal to fp; f, concentration ratio; i, indicator; fp, concentration ratio interpolated between values just before and after stop interval; fven, concentration ratio of indicator in venous sample.

(2)

(3)

(4)

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Slow perfusion: effect of duration of stop flow on ²⁰¹Tl⁺ and ⁸⁶Rb⁺ uptake. As indicated in Fig. 2, there was uptake of ²⁰¹Tl⁺ before and after the stop-flow interval in the slow-perfusion studies. Outflow concentration ratios of ²⁰¹Tl⁺ were below those of the vascular indicator ¹²⁵I-albumin and below those of [³H]mannitol and [¹⁴C]urea, which each diffuses out of the vasculature into the pulmonary tissues. The decrease in ²⁰¹Tl⁺ concentrations after the stop flow was attributed to further uptake during the stop-flow interval. Increasing the duration of the stop-flow interval increased the calculated volume of perfusate that was cleared of ²⁰¹Tl⁺ (Fig. 3). However, prolonging the duration of the stop-flow interval from 90 to 300 s resulted in a relatively small increase in V_exch of ²⁰¹Tl⁺ in the slow- and the fast-flow experiments (see below).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Slow-perfusion studies (flow stopped for 90 s). Concentration ratios of [3H]mannitol are just below those of 125I-albumin, and concentration ratios of [14C]urea are just below those of [3H]mannitol. Concentration ratios of 201Tl+ are below those of each of other indicators, and concentrations fall significantly after stop-flow interval, in a manner consistent with additional uptake by tissues during stop-flow interval. Outflow concentrations of 201Tl+ remain well below those of 125I-albumin for a prolonged interval (inset), indicating that tissues had not become saturated with 201Tl+.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Effect of duration of stop-flow interval on Vexch for 201Tl+ (Vexch,thallium) in slow-perfusion experiments. Vexch,thallium appears to approach maximal values when stop-flow interval is 90 s.

Fast perfusion: effects of K⁺ concentration, duration of stop flow, and left atrial pressures during stop interval. Outflow concentration ratios of ⁸⁶Rb⁺ were slightly greater than those of ²⁰¹Tl⁺ before and after the stop-flow period (Fig. 4). However, the relative decrease in concentrations after the stop-flow interval was similar when K⁺ concentrations in the perfusate were between 4 and 10 meq/l. This was reflected by similar clearances of ²⁰¹Tl⁺ and ⁸⁶Rb⁺ from the perfusate in each sample during the stop interval (indicated by equivalent values of v_exch/v_ven in Fig. 5) and calculated V_exch (2nd, 5th, and 6th sets of bars in Fig. 6; the apparent differences at 1 and 40 meq/l K⁺ were not significant). Increasing the concentrations of K⁺ to 40 meq/l decreased V_exch of both indicators (Figs. 6 and 7). Concentration ratios of ²⁰¹Tl⁺ were closer to those of ¹²⁵I-albumin before and after the stop-flow interval when lungs were perfused at rapid rather than slow rates (cf. Figs. 2 and 4).

View larger version (35K): [in this window] [in a new window]   Fig. 4.   Fast-perfusion studies (flow stopped for 90 s, K+ = 4.0 meq/l). Concentration ratios of 201Tl+ are significantly closer to those of 125I-albumin before and after stop interval than in slow-perfusion studies (Fig. 2). This is consistent with effect of rapid flow on 201Tl+ extraction observed in earlier studies (2). Unlike results obtained at slow flow, it was possible to document a decline in concentrations of [3H]mannitol after stop interval, indicating some additional uptake of this indicator by lung tissue during stop interval. Concentration ratios of 86Rb+ tended to be greater than those of 201Tl+ before and after stop interval, but decreases in concentration ratios that occurred after stop interval were similar for 86Rb+ and 201Tl+.

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Decreasing left atrial pressure (LAP) during stop interval reduced apparent volume of exchangeable vessels (vexch) in lung, reflected by lower values of vexch/vven in collected samples. Nearly all fluid in exchange volume of lung during stop interval was recovered during first 5 s after stop interval. vven, Volume of venous sample.

View larger version (41K): [in this window] [in a new window]   Fig. 6.   Values of Vexch in fast-perfusion experiments. Increasing perfusate K+ concentrations to 40 meq/l significantly decreased Vexch of 201Tl+ and 86Rb+. Increasing duration of stop interval to 300 s did not have significant effects on Vexch for these indicators. Vexch of 86Rb+ was reduced when left atrial pressures were zero rather than 10 cmH2O during stop intervals. Vexch,i, indicator Vexch.

View larger version (26K): [in this window] [in a new window]   Fig. 7.   Increasing concentration of K+ in perfusate resulted in a decrease in fall of 201Tl+ in samples collected after a 90-s stop-flow period. Decrease in uptake is attributed in part to a decrease in Nernst potential of pulmonary cells but may also reflect a decrease in exchangeable vascular volume (see DISCUSSION). fi, Concentration ratio for indicator i.

Uptake of other small indicator molecules before and during stop flow. Concentration ratios of [³H]mannitol and [¹⁴C]urea were less than those of ¹²⁵I-albumin during loading and flushing in the slow-flow experiments, indicating loss of these indicators from the vasculature during perfusion (Fig. 2). Losses of [¹⁴C] urea exceeded those of [³H]mannitol: the calculated values of V_exch of [³H]mannitol averaged 0.06 ± 0.01 ml (n = 16), whereas V_exch of [¹⁴C]urea averaged 0.13 ± 0.01 ml in these slow-flow experiments (P < 0.01, by paired t-test).

View larger version (32K): [in this window] [in a new window]   Fig. 8.   Concentration ratios of 3HOH were significantly below those of 125I-albumin before stop flow in each of 4 experiments conducted at fast perfusion rates, indicating diffusion out of vasculature before stop interval. After a 90-s stop interval, a small decrease in 3HOH concentration was observed, indicating that additional 3HOH had equilibrated with pulmonary tissue during stop interval.

Pulmonary arterial pressures and lung water content. Pulmonary arterial pressures were kept at 20 cmH₂O in the slow-perfusion studies. Much higher pulmonary arterial pressures were observed in the fast-perfusion experiments: 35.0 ± 9.5 (SD) cmH₂O during the 10-s loading period in 19 experiments in which it was successfully measured and 32.1 ± 11 cmH₂O during the 10-s flush period in 18 experiments. Pulmonary arterial pressures were >50 cmH₂O during the loading and flush periods in two lungs perfused with 40 meq/l K⁺ and two lungs perfused with 1 meq/l K⁺. Lung weights were slightly but significantly greater at the end of the fast-perfusion experiments than in the slow-perfusion experiments: 1.90 ± 0.23 vs. 1.67 ± 0.18 (SE) g (P < 0.01). Similarly, wet-to-dry weight ratios were slightly greater in the fast- than in the slow-perfusion studies: 5.84 ± 0.38 vs. 5.40 ± 0.76 (P < 0.01).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

²⁰¹Tl⁺ and ⁸⁶Rb⁺ are efficiently removed from the circulation of a wide variety of organs (2, 5, 6, 8). Because uptake of these indicators is similar to that of K⁺, they become concentrated within the cellular compartment. In a recent study we showed that ²⁰¹Tl⁺ is progressively removed from the pulmonary circulation and that extravascular concentrations reach levels that are ~18 times intravascular concentrations (2). Three observations in the present group of experiments are consistent with the conclusion that nearly all the ²⁰¹Tl⁺ is removed from the perfusate remaining in the exchange vessels during stop-flow experiments: 1) Although concentrations of ²⁰¹Tl⁺ in perfusate flushed from the lungs decrease more when the stop-flow period is extended from 10 to 300 s, increasing the stop interval from 90 to 300 s had a relatively minor effect on uptake (Figs. 3 and 6). This was not because the tissues had become saturated with this indicator but was presumably because there was very little left in the exchange portions of the vascular bed. As soon as flow was resumed, extraction of ²⁰¹Tl⁺ from the circulation returned to levels close to those that were observed before flow was discontinued. 2) Increasing K⁺ concentrations in the perfusate should reduce the Nernst potential between the cellular and intracellular compartments, and uptake of ²⁰¹Tl⁺ and ⁸⁶Rb⁺ did decline when K⁺ concentrations were increased to 40 meq/l. However, at 10 meq/l, changes in K⁺ concentrations did not have a significant effect on the loss of ²⁰¹Tl⁺ from the venous samples, suggesting that over this range of K⁺ levels there was very little ²⁰¹Tl⁺ left in the plasma. 3) Clearance of ²⁰¹Tl⁺ and ⁸⁶Rb⁺ from the perfusate during stop-flow studies was similar at physiological K⁺ concentrations. As in the cardiac circulation (5), uptake of ²⁰¹Tl⁺ was slightly greater than that of ⁸⁶Rb⁺ during the loading and flush intervals, but values calculated with Eq. 4 for V_exch with each of these indicators were not significantly different.

Some of the decrease in uptake of ²⁰¹Tl⁺ and ⁸⁶Rb⁺ at very high concentrations of K⁺ could also be related to vasoconstriction and derecruitment, which would be expected with hyperkalemia (10). Pulmonary arterial pressures did tend to be high at high and low K⁺ concentrations.

If it can be assumed that concentrations of ²⁰¹Tl⁺, ⁸⁶Rb⁺, or comparable solutes fall to negligible values during stop-flow periods of sufficient duration, then the fraction (v_exch/v_ven) of each of the venous collection samples can be calculated with Eq. 3 and the total exchangeable volume in the pulmonary vasculature can be calculated with Eq. 4. As indicated in our earlier study (3), the first samples of the perfusate that were in exchange vessels during stop flow are derived from more venous portions of the vasculature than were the later samples when the lungs are perfused in an anterograde direction from artery to vein. During retrograde flushes the initial samples of the fluid would be derived from more arterial portions of the vasculature than were later samples. Retrograde experiments were not conducted in the present study but were reported previously, with ³HOH instilled into the air spaces as a marker for exchange vessels rather than ²⁰¹Tl⁺ or ⁸⁶Rb⁺ (3). Although the anatomic site of indicator exchange is not defined in experiments of this sort, it is possible to determine the relative locations of such exchange processes as diffusional exchange and edema formation and reabsorption utilizing stop-flow experiments (3), and it may also be possible to follow release and uptake or metabolism of mediators along the vasculature with this approach, much as solute uptake has been characterized in the nephron.

An ideal indicator of the exchange volume of the lung would be removed from the exchange vessels minimally during loading and flushing and maximally during the stop interval. Three types of indicators were tried in this experiment: K⁺ analogs (²⁰¹Tl⁺ and ⁸⁶Rb⁺), small solutes, which diffuse primarily into the pulmonary interstitium ([³H]mannitol and [¹⁴C]urea), and labeled water, which rapidly enters most lung compartments. Much of the loss of [³H]mannitol and [¹⁴C]urea from the circulation actually occurred during the loading period, and at low perfusion rates it was difficult to detect any additional decrease in the concentrations of these indicators during the stop interval. ³HOH rapidly entered the extravascular compartment, and again, it was difficult to detect any additional losses from the circulation during the slow stop-flow experiments. In contrast, much more of the uptake of ²⁰¹Tl⁺ and ⁸⁶Rb⁺ occurred during the stop interval, because the cells represent a large reservoir for these indicators. Volatile agents represent a fourth type of indicator that could be tried to mark the exchange vessels of the lungs during stop-flow studies. For example, preliminary studies in our laboratory indicate that a large fraction of labeled acetone is lost from the vasculature during stop-flow intervals, presumably into the air spaces. Alternatively, metabolites that would be efficiently removed or degraded during a stop interval could be utilized.

The duration and speed of loading and flushing the perfusate from the lungs have an important bearing on the outcome of stop-flow experiments. In the slow-perfusion studies, loading and subsequent flushing occurred over a longer interval, and there was considerable uptake of indicators from the circulation before and after the stop-flow interval. This was particularly true for ²⁰¹Tl⁺: concentration ratios of this indicator before and after the stop interval were considerably lower when the lungs were perfused at 8 than at 34 ml/min (cf. Figs. 2 and 4). However, despite greater extraction of ²⁰¹Tl⁺ and ⁸⁶Rb⁺ before and after the stop interval in the slow-perfusion studies, calculated values of V_exch of these indicators were not significantly influenced by the rates of loading and flushing.

V_exch of ⁸⁶Rb⁺ appeared to be decreased when left atrial pressures were decreased to 0 from 10 cmH₂O (Figs. 5 and 6). This would be expected if the amount of perfusate in the exchange volume of the lung was reduced at lower left atrial pressures.

Decreases in concentrations of [³H]mannitol, [¹⁴C]urea, and ³HOH after the stop interval could only be observed when loading was rapid and the duration of loading was short. This suggests that rapid loading and flushing are advantageous if indicators of this nature are being investigated. However, excessive perfusion rates should be avoided during these intervals, since increases in pulmonary arterial pressures could be associated with pulmonary vascular injury. We found evidence for a small amount of edema in lungs briefly perfused at rapid rates.

It is likely that much of the ²⁰¹Tl⁺ and ⁸⁶Rb⁺ lost from the vasculature crosses the luminal surface of the pulmonary capillaries and enters the endothelial cells. Additional quantities of these indicators may diffuse through the intercellular junctions and enter endothelial, interstitial, and epithelial cells of the lungs. It is generally assumed that much of the [¹⁴C]urea and [³H]mannitol that is lost from the vasculature diffuses through the interendothelial junctions into the pulmonary interstitium, and comparable quantities of ²⁰¹Tl⁺ and ⁸⁶Rb⁺ may diffuse through these structures into the interstitium. Pulmonary edema caused by increased intravascular pressures or injury to the capillary wall may alter the manner in which ²⁰¹Tl⁺ and ⁸⁶Rb⁺ are removed from the pulmonary circulation and the quantity of fluid in the exchange vessels of the lungs. However, these circumstances would not invalidate the calculation of V_exch if cellular uptake continues to remove all the ²⁰¹Tl⁺ and ⁸⁶Rb⁺ remaining in the exchange sites of the lungs during the stop interval.

³HOH rapidly crosses cellular membranes and equilibrates with water in the cellular and interstitial compartments. It has been generally assumed that the distribution of ³HOH between the vasculature and tissues of the lungs is limited by flow rather than by the permeability of the pulmonary vasculature to water. The small decreases in concentrations that were observed after the stop interval indicated that additional ³HOH had diffused out of the vessels after the 10-s loading period. This could represent some diffusion from perfused regions of the lungs to those that remained unperfused rather than a barrier at the capillary wall. In other words, the rate of diffusion of ³HOH across the capillary wall could in theory be virtually the same as that observed between tissue compartments that are not separated by barriers, but equilibration between compartments may be delayed by the time required for free diffusion to distant sites. Nevertheless, this observation does suggest that if tissue perfusion is not uniform, delivery of ³HOH to the lung tissues may be limited in part by diffusion rather than flow.