Blood-gas equilibrium of carbon dioxide in lungs: a continuing controversy.

This review presents the experimental evidence that has been published in recent years both against and in support of the occurrence of negative blood-gas CO2 partial pressure differences (delta PCO2) in lungs in rebreathing equilibrium and during steady-state gas exchange in hypercapnia. Although some sources of potential experimental error can be pointed out, the reasons for the remarkably pronounced disagreement between the experimental data of the different studies cannot be definitely identified. Since a consistent and reproducible occurrence of negative blood-gas delta PCO2 in lungs in gas-blood equilibrium is not convincingly proved, it appears to be justified to continue accepting the validity of the conventional concept of equal PCO2 in blood and gas in equilibrium. Because the issue is of considerable importance in the analysis and understanding of alveolar gas exchange, pertinent evidence is expected from future work.

Model simulation of single-breath washout of insoluble gases from dog lungs.

Intrapulmonary gas mixing by convection and diffusion is analyzed on the basis of various analog lung models incorporating singly or in combination series inhomogeneity, parallel inhomogeneity, intercompartmental diffusion, sequential inspiration, and sequential expiration. The slope of the alveolar plateau of insoluble gases is used as an indicator for incomplete gas mixing. By use of the models, we have attempted to simulate experimental results obtained in artificially ventilated dog lungs with simultaneous single-breath washout of He and SF6 (Meyer et al., J. Appl. Physiol. 55: 1795-1802, 1983). Sufficient agreement with experimental data is obtained only when diffusion between serial-parallel compartments and sequential expiration are incorporated, the first being mainly responsible for separation of He and SF6, the latter for the slope of the alveolar plateau of these gases. A combined operational model comprising two parallel compartments connected to a mixing compartment, with both convective and diffusive gas transport between the compartments, could account for most of the experimental observations.

Kinetics of O2 uptake and release by human erythrocytes studied by a stopped-flow technique.

The kinetics of O2 uptake into and release from human erythrocytes was investigated at 37 degrees C by a stopped-flow technique. From the time course of O2 saturation (SO2) change a specific transfer conductance of erythrocytes for O2 (GO2) was calculated. The following results were obtained: 1) GO2 decreased in the course of O2 uptake, but initial GO2 was nearly independent of SO2 at which uptake started; 2) addition of albumin to the medium reduced GO2; 3) increasing dithionite concentration in the medium in O2-release experiments progressively enhanced GO2, which became virtually constant for nearly the entire course of release; and 4) O2 uptake and O2 release (without dithoite) in the same SO2 range yielded very similar GO2. These results suggested that O2 uptake and release were importantly limited by diffusion through the external medium and that in the SO2 range between 0.3 and 0.8, chemical reaction exerted little limiting effect. Since O2 release at the highest dithionite concentration (40 mmol/l) appeared to be virtually unlimited by external diffusion, GO2 measured under these conditions, averaging 8.7 ml X min-1 X Torr-1 X ml erythrocytes-1, was considered to mainly reflect intracellular diffusion limitation. The corresponding specific transfer conductance for O2 transfer in whole blood (hematocrit, 0.45) is 3.9 ml X min-1 X Torr-1 X ml blood-1.

Arterial-alveolar CO2 equilibration in exercising dogs during prolonged rebreathing.

Arterial-alveolar equilibration of CO2 during exercise was studied by normoxic CO2 rebreathing in six dogs prepared with a chronic tracheostomy and exteriorized carotid loop and trained to run on a treadmill. In 153 simultaneous measurements of PCO2 in arterial blood (PaCO2) and end-tidal gas (PE'CO2) obtained in 46 rebreathing periods at three levels of mild-to-moderate steady-state exercise, the mean PCO2 difference (PaCO2-PE'CO2) was -1.0 +/- 1.0 (SD) Torr and was not related to O2 uptake or to the level of PaCO2 (30-68 Torr). The small negative PaCO2-PE'CO2 is attributed to the lung-to-carotid artery transit time delay which must be taken into account when both PaCO2 and PE'CO2 are continuously rising during rebreathing (average rate 0.22 Torr/s). Assuming that blood-gas equilibrium for CO2 was complete, a lung-to-carotid artery circulation time of 4.6 s accounts for the observed uncorrected PaCO2-PE'CO2 of -1.0 Torr. The results are interpreted to indicate that in rebreathing equilibrium PCO2 in arterial blood and alveolar gas are essentially identical. This conclusion is at variance with previous studies in exercising humans during rebreathing but is in full agreement with our recent findings in resting dogs.

Pulmonary diffusing capacities for nitric oxide and carbon monoxide determined by rebreathing in dogs.

Pulmonary diffusing capacities (DL) of NO and CO were determined simultaneously from rebreathing equilibration kinetics in anesthetized paralyzed supine dogs (mean body wt 20 kg) after denitrogenation (replacement of N2 by Ar). During rebreathing the dogs were ventilated in closed circuit with a gas mixture containing 0.06% NO, 0.06% 13C18O, and 1% He in Ar for 15 s, with tidal volume of 0.5 liter and frequency of 60/min. The partial pressures of NO, 13C18O, 16O18O, N2, Ar, CO2, and He in the trachea were continuously analyzed by mass spectrometry. Measurements were performed at various O2 levels characterized by the mean end-expired PO2 during rebreathing (PE'O2). In control conditions ("normoxia," PE'O2 = 67 +/- 8 Torr) the following mean +/- SD values were obtained (in ml.min-1.Torr-1): DLNO = 52.4 +/- 11.0 and DLCO = 15.4 +/- 2.9. In hypoxia (PE'O2 = 24 +/- 7 Torr) DLNO increased by 11 +/- 8% and DLCO by 19 +/- 10%, and in hyperoxia (PE'O2 = 390 +/- 26 Torr) DLNO decreased to 87 +/- 3% and DLCO to 56 +/- 8% with respect to values in normoxia. DLNO/DLCO of 3.24 +/- 0.06 (hypoxia), 3.38 +/- 0.31 (normoxia), and 5.54 +/- 1.04 (hyperoxia) were significantly higher than the NO/CO Krogh diffusion constant ratio (1.92) predicted for simple diffusion through aqueous layers. With increasing O2 uptake elicited by 2,4-dinitrophenol, DLNO and DLCO increased and DLNO/DLCO remained close to unchanged. The results suggest that the combined effects of diffusion and chemical reaction with hemoglobin limit alveolar-capillary transport of CO. If it is assumed that reaction kinetics of NO with hemoglobin (known to be extremely fast) are not rate limiting for NO uptake, the contribution of the slow chemical reaction with hemoglobin to the total CO uptake resistance (= 1/DLCO) was estimated to be 38% in hypoxia, 41% in normoxia, and 64% in hyperoxia. The various factors expected to restrict the validity of this analysis are discussed, in particular the effects of functional inhomogeneity.

Alveolar slope and dead space of He and SF6 in dogs: comparison of airway and venous loading.

Series (Fowler) dead space (VD) and slope of the alveolar plateau of two inert gases (He and SF6) with similar blood-gas partition coefficients (approximately 0.01) but different diffusivities were analyzed in 10 anesthetized paralyzed mechanically ventilated dogs (mean body wt 20 kg). Single-breath constant-flow expirograms were simultaneously recorded in two conditions: 1) after equilibration of lung gas with the inert gases at tracer concentrations [airway loading (AL)] and 2) during steady-state elimination of the inert gases continuously introduced into venous blood by a membrane oxygenator and partial arteriovenous bypass [venous loading (VL)]. VD was consistently larger for SF6 than for He, but there was no difference between AL and VL. The relative alveolar slope, defined as increment of partial pressure per increment of expired volume and normalized to mixed expired-inspired partial pressure difference, was larger by a factor of two in VL than in AL for both He and SF6. The He-to-SF6 ratio of relative alveolar slope was generally smaller than unity in both VL and AL. Whereas unequal ventilation-volume distribution combined with sequential emptying of parallel lung regions appears to be responsible for the sloping alveolar plateau during AL, the steeper slope during VL is attributed to the combined effects of continuing gas exchange and ventilation-perfusion inequality coupled with sequential emptying. The differences between He and SF6 point at the contributing role of diffusion-dependent mechanisms in intrapulmonary gas mixing.

Pulmonary gas exchange in panting dogs.

Pulmonary gas exchange during panting was studied in seven conscious dogs (32 kg mean body wt) provided with a chronic tracheostomy and an exteriorized carotid artery loop. The animals were acutely exposed to moderately elevated ambient temperature (27.5 degrees C, 65% relative humidity) for 2 h. O2 and CO2 in the tracheostomy tube were continuously monitored by mass spectrometry using a special sample-hold phase-locked sampling technique. PO2 and PCO2 were determined in blood samples obtained from the carotid artery. During the exposure to heat, central body temperature remained unchanged (38.6 +/- 0.6 degrees C) while all animals rapidly switched to steady shallow panting at frequencies close to the resonant frequency of the respiratory system. During panting, the following values were measured (means +/- SD): breathing frequency, 313 +/- 19 breaths/min; tidal volume, 167 +/- 21 ml; total ventilation, 52 +/- 9 l/min; effective alveolar ventilation, 5.5 +/- 1.3 l/min; PaO2, 106.2 +/- 5.9 Torr; PaCO2, 27.2 +/- 3.9 Torr; end-tidal-arterial PO2 difference [(PE' - Pa)O2], 26.0 +/- 5.3 Torr; and arterial-end-tidal PCO2 difference, [(Pa - PE')CO2], 14.9 +/- 2.5 Torr. On the basis of the classical ideal alveolar air approach, parallel dead-space ventilation accounted for 54% of alveolar ventilation and 66% of the (PE' - Pa)O2 difference. But the steepness of the CO2 and O2 expirogram plotted against expired volume suggested a contribution of series in homogeneity due to incomplete gas mixing.

Regional blood flow in conscious resting rats determined by microsphere distribution.

To determine organ blood flow in the resting state, a box was designed to keep conscious untrained rats minimally disturbed. Blood pressure, heart rate, and organ blood flow, determined by the microsphere distribution and reference sampling technique, were measured in 11 Sprague-Dawley rats. After an acclimation period, 15-microns-diameter microspheres labeled with 113Sn were infused into the ascending aorta, a reference blood sample was withdrawn from the caudal artery, and organ blood flows were computed according to standard procedures. The average values of heart rate (365 beats/min) and blood flow to the brain (45 ml.min-1.100 g-1) and hindlimb muscles (15 ml.min-1.100 g-1) were significantly lower than most values reported earlier, whereas splanchnic blood flow was significantly higher (106 ml.min-1.100 g-1). Blood flow to the soleus muscle, which is considered the most active for postural maintenance, was relatively high (99 ml.min-1.100 g-1). The combination of low skeletal muscle and high visceral blood flows observed in these experiments suggests a low sympathetic tone, which is consistent with the low level of circulating catecholamines also observed in this study. It is hypothesized that the difference between our present and previous results is a lower level of stress, attributable to a more complete acclimation to the experimental environment.

Changes in regional blood flow distribution and oxygen supply during hypoxia in conscious rats.

The effects of acute hypoxia on central hemodynamics, regional blood flow, and regional oxygen supply (blood flow x arterial O2 concentration) were studied in conscious resting rats. Regional blood flow was determined by the radiolabeled microsphere technique. Blood pressure, heart rate; and aortic blood flow increased and total peripheral resistance decreased significantly during hypoxia. Blood flow to brain, respiratory muscles, and liver increased both in absolute value and as a fraction of the aortic blood flow. Fractional blood flow to the gastrointestinal tract, spleen, pancreas, skin, fat, and hindlimb bones decreased during hypoxia; blood flow decreased in absolute values only in stomach and fat. Oxygen supply to brain, respiratory muscles, and liver increased during hypoxia, whereas it decreased in the remaining organs investigated.

Cardiogenic oscillations in He and SF6 expirograms during airway and venous loading.

Cardiogenic oscillations in the expired partial pressure profiles of two inert gases (He and SF6) were monitored in seven anesthetized paralyzed mechanically ventilated dogs. He and SF6 were administered either intravenously by a membrane oxygenator and partial arteriovenous bypass [venous loading (VL)] or by washin into lung gas [airway loading (AL)]. The single-breath expirograms obtained during constant-flow expiration after inspiration of test gas-free air displayed distinct and regular cardiogenic oscillations. The relative oscillation amplitude (ROA), calculated as oscillation amplitude divided by mixed expired-inspired partial pressure difference, was in the range of 1-8%. The ROA for both He and SF6 was approximately 4.2 times higher in VL than in AL, which indicated that among lung units that emptied sequentially in the cardiac cycle, the effects of alveolar ventilation-perfusion (VA/Q) inequality were more pronounced than those of alveolar ventilation-alveolar volume (VA/VA) inequality. In AL, He and SF6 oscillations were 180 degrees out of phase compared with CO2 and O2 oscillations and with He and SF6 oscillations in VL, which suggests that regions with low VA/VA had high VA/Q and very low Q/VA. The ROA was practically unaffected by breath holding in both AL and VL, which indicates that there was little diffusive or convective (cardiogenic) mixing between the lung units that were responsible for cardiogenic oscillations. The ROA was consistently higher for He than for SF6, and the He-to-SF6 ratio was independent of route of test gas loading, averaging 1.6 in both AL and VL. This result may be explained by laminar Taylor dispersion, whereby oscillations generated in peripheral lung regions are dissipated in inverse proportion to diffusion coefficient during transit through the proximal (larger) airways.

Blood flow distribution in dog gastrocnemius muscle at rest and during stimulation.

The distribution of blood flow within the isolated perfused dog gastrocnemius muscle (weight 100-240 g) was studied by intra-arterial injection of radioactively labeled microspheres (diameter 15 micron) at rest and during supramaximal stimulation to rhythmic isotonic tetanic contractions of varied frequency against varied loads. After the experiment the muscle was cut into 180-250 pieces of approximately 0.75 g each, and the blood flow to each muscle piece was determined from its radioactivity. The inhomogeneity of blood flow was represented as the frequency distribution of the ratios of regional specific blood flow, i.e., blood flow per unit tissue weight of the piece, QR, to the overall specific blood flow of the muscle, Q. The QR/Q values for the individual pieces of a muscle were found to vary widely both at rest and during stimulation. With rising work load the frequency distribution had a tendency to broaden and flatten, indicating increasing perfusion inhomogeneity. On the average of the experiments, there was no significant difference in specific blood flow between the three anatomic components of the gastrocnemius (lateral and medial heads of gastrocnemius and flexor digitorum superficialis) nor between the superficial and deep portions within these anatomic components, only the distal third of the muscle was relatively less perfused compared with the proximal two-thirds. The considerable inhomogeneity of blood flow as revealed by microsphere embolization and by other methods is expected to exert important limiting effects on local O2 supply, particularly during exercise. Its neglect would lead to serious errors in the analysis of O2 supply to muscle tissue.

Convective and diffusive gas transport in canine intrapulmonary airways.

The significance of convective and diffusive gas transport in the respiratory system was assessed from the response of combined inert gas and particle boluses inhaled into the conducting airways. Particles, considered as "nondiffusing gas," served as tracers for convection and two inert gases with widely different diffusive characteristics (He and SF6) as tracers for convection and diffusion. Six-milliliter boluses labeled with monodisperse di-2-ethylhexyl sebacate droplets of 0.86-microns aerodynamic diameter, 2% He, and 2% SF6 were inspired by three anesthetized mechanically ventilated beagle dogs to volumetric lung depths up to 170 ml. Mixing between inspired and residual air caused dispersion of the inspired bolus, which was quantified in terms of the bolus half-width. Dispersion of particles increased with increasing lung depth to which the boluses were inhaled. The increase followed a power law with exponents less than 0.5 (mean 0.39), indicating that the effect of convective mixing per unit volume was reduced with depth. Within the pulmonary dead space, the behavior of the inert gases He and SF6 was similar to that of the particles, suggesting that gas transport was almost solely due to convection. Beyond the dead space, dispersion of He and SF6 increased more rapidly than dispersion of particles, indicating that diffusion became significant. The gas and particle bolus technique offers a suitable approach to differential analysis of gas transport in intrapulmonary airways of lungs.

Modeling of oxygen transport to skeletal muscle: blood flow distribution, shunt, and diffusion.

By injection of embolizing microspheres, by local radioactive xenon clearance and by inert gas washout in resting and stimulated gastrocnemius dog preparation, experimental evidence for unequal blood flow distribution and for shunt flow has been provided. Model calculations show that in some respect unequal blood flow and shunt produce effects predicted for a homogeneous model with diffusion limitation of O2 supply. This finding must be taken into account when the role of diffusion limitation to O2 supply is to be ascertained.

Alveolar-capillary gas transfer in lungs: development of concepts and current state.

Progress in research on pulmonary gas exchange, with special reference to the contribution of Gerhard Thews and associates, is reviewed. In particular, the following aspects are considered. (1) Oxygen transfer kinetics of red blood cells. Recent measurements, particularly on red blood cells in thin blood films, yield more rapid equilibration kinetics than previously recorded. A reevaluation of the roles of diffusion and chemical reaction in alveolar O2 uptake may become necessary. (2) Gas exchange in functionally inhomogeneous lungs. Besides the classical ventilation/perfusion (VA/Q) inequality, a variation of the diffusing capacity-to-perfusion ratio (DL/Q) appears to be of importance. The combination of VA/Q and DL 1Q inequalities may lead to a better understanding of alveolar gas exchange, particularly in diseased lungs. (3) Pulmonary diffusing capacity (DL) for oxygen. The rebreathing technique, which strongly reduces the effects of inequal VA/Q distribution effects, appears to be particularly suited for measurement of overall alveolar-capillary diffusion. But neither the factors determining DL, obtained by rebreathing or other methods, nor the relationships between DL for various gases are yet fully understood.

Role of diffusion shunt in transfer of inert gases and O2 in muscle.

Diffusion shunt is diffusive gas exchange between arterial and venous vessels. Evidence for diffusion shunt had been obtained in washout studies in the gastrocnemius muscle of the dog. According to models, diffusion shunt is expected to be enhanced at low blood flow, and for gases of high diffusivity. Shunting of O2 should be reduced in comparison to inert gases because of chemical binding in blood.

Diffusion-perfusion relationships in skeletal muscle: models and experimental evidence from inert gas washout.

In order to study the dependence of blood-tissue gas exchange upon diffusion, the simultaneous washout of two inert gases of differing diffusivity was investigated in isolated-perfused dog gastrocnemius preparations. The muscles were equilibrated with CH4 and SF6 via arterial blood. The washout kinetics were determined from venous blood samples analyzed by gas chromatography. The results revealed the following features: The washout of the test gases was pronouncedly multi-exponential, and could be described by three exponential components when analyzed to 5% of the initial value. The non-exponential washout was attributed to unequal distribution of capillary blood flow to tissue volume. The mean ratio of washout rate constants CH4/SF6 was within 1.10-1.25 and was even smaller than the ratio expected for pure perfusion limitation (1.46). Therefore, no evidence for effective tissue-blood diffusion limitation was obtained. The observed washout rate constant ratio could be explained by a model with veno-arterial back diffusion which more strongly retards washout kinetics of the better diffusible gas (CH4) as compared to the less diffusible gas (SF6).

Effects of temperature on oxygen transfer conductance of human red blood cells.

The influence of temperature (varied from 37 to 7 degrees C; average pH = 7.4) on the kinetics of O2 uptake and release by human red blood cells under stopped-flow conditions was investigated by double-beam spectrophotometry. The kinetics were characterized by the specific transfer conductance for O2, G. The temperature coefficient of G, Q10(G), for O2 uptake averaged 1.17, and activation energy, Ea(G) = 2.9 kcal/mol O2. The average values for O2 release were: Q10(G) = 1.30, and Ea(G) = 4.8 kcal/mol O2. The G values for release of O2 from oxyhaemoglobin solution, Gsol, yielded Q10(Gsol) = 2.06, Ea(Gsol) = 13.4 kcal/mol O2. Comparison of these Q10 and Ea values with those for diffusion of O2 and haemoglobin in aqueous media leads to the conclusion that the kinetics of O2 uptake and release by red blood cells in the stopped-flow condition is mainly limited by diffusion of O2 and haemoglobin in the red cell interior and by diffusion of O2 in the medium, and to a lesser degree by chemical reaction kinetics.

Elements for modeling inert gas washout from heterogeneous tissues.

Inert gas washout from tissue may be analyzed using Krogh's cylinder or a simplified model thereof, represented by a series arrangement of compartments. In particular with limited axial diffusion in tissue, washout from this model is multi-exponential, the rate of washout accelerating with time. To explain the experimentally observed multicomponent washout, in which washout retards with time, heterogeneity of flow and/or diffusing capacity to volume is required, either in series or in parallel compartments. While there appears to be a structural basis for both types of heterogeneity, parallel and series, a distinction between both on the basis of washout curves alone is impossible.

Dual role of diffusion in tissue gas exchange: blood-tissue equilibration and diffusion shunt.

The role of diffusion in inert gas washout from tissue is investigated using simple mathematical models incorporating diffusive blood-tissue equilibration and diffusion shunt due to diffusive gas transfer between precapillary and postcapillary vessels with counter-current blood flow. With increasing diffusivity blood-tissue equilibration is improved, but simultaneously the diffusion shunt by veno-arterial back diffusion is increased. Similarly, with decreasing blood flow, the extent of diffusion limitation in blood-tissue transfer is diminished, but at the same time veno-arterial diffusive shunting is enhanced. Diffusion shunt slows inert gas washout (local tissue clearance) and thus, if not taken into account, leads to an underestimation of capillary blood flow calculated from the washout rate constant. Diffusion shunting of O2 diminishes the efficacy of blood-tissue transfer, but its extent is predicted to be smaller compared to that for inert gases, because the chemical combination of O2 in blood decreases diffusive shunting.

Problems in determination of oxygen dissociation of avian blood.

No appreciable errors are expected in determination of blood gas values and pH using classical techniques provided time of anaerobic storage is kept small and is, if unavoidable performed on ice. In particular, dissociation curves may safely be analyzed with the Van Slyke technique which is in disagreement with the conclusions of Lutz et al. (1973). For measurement of PO2 and PCO2, delay time is mainly dictated by response time of the electrodes; measurements may have to be corrected for metabolism, particularly in high PO2 range.