MICHAEL P. HLASTALA. In vivo quantitation of carbonic anhy- drase and band 3 protein contributions to pulmonary gas ex- change. J. Appl. Physiol. 74(Z): ...
In vivo quantitation of carbonic anhydrase and band 3 protein contributions to pulmonary gas exchange ERIK
R. SWENSON,
JORGEN
GR0NLUND,
JAN
OHLSSON,
AND
MICHAEL
P. HLASTALA
Departments of Medicine and of Physiology and Biophysics, University of Washington, Seattle, Washington 98195 SWENSON,
ERIK
R., JORGEN
GRBNLUND,
JAN OHLSSON,
AND
MICHAEL P. HLASTALA. In vivo quantitation
of carbonic anhydrase and band 3 protein contributions to pulmonary gas exchange. J. Appl. Physiol. 74(Z): 838-848, 1993.-The contributions to pulmonary gas exchange of red blood cell (RBC) membrane band 3 protein HCO;-Clexchange and carbonic anhydrase- (CA) catalyzed HCO, dehydration have never been determined directly in the whole animal. We utilized an experimental and model approach to measure these by analysis of phase III exhaled CO, and 0, profiles in anesthetized dogs. In this method, we inhibit RBC membrane band 3 protein and cytoplasmic CA in RBCs passing the pulmonary capillaries and lung vascular luminal membrane-bound CA during a single ventilatory cycle. This is achieved with appropriately timed right atria1 infusions of 4,4’-dinitrostilbene-2,2’-disulfonate (DNDS) to inhibit band 3 protein, ethoxzolamide (a lipophilic CA inhibitor with rapid membrane penetrance) to inhibit RBC and lung tissue CA, and benzolamide (an extremely hydrophilic CA inhibitor with virtually no penetrance into RBC cytoplasm) to inhibit only lung vascular luminal membrane CA. DNDS caused a 15% reduction in CP, production (%ko& without any change in 0, consumption (VO,). The addition of.benzolamide to DNDS did not cause any further decrease in VCO~.Inhibition of RBC CA by ethoxzolamide caused a 67% reduction in ho2 and a 11.5% reduction in vo2. Inhibition of lung vascular CA by benzolamide alone caused no statistically significant changes in either ho, or VO,. These results are in general agreement with in vitro data and model calculations. The only exceptions are the higher than predicted effect of RBC CA inhibition on TO, (Bohr effect) and the lack of any contribution to CO, transfer in the dog by lung vascular CA with access to plasma as a possible consequence of an endogenous plasma CA inhibitor. chloride-bicarbonate exchange; dog; hypoxia; Bohr effect; carbon dioxide; oxygen; acetylene; ethoxzolamide; benzolamide; 4,4’-dinitrostilbene-2,2’-disulfonate
THE NORMAL UPTAKE of CO, by blood in peripheral
tissues and its excretion into alveolar gas in the lung is an effective process that occurs over very narrow PCO, differences. This is the consequence of the three major proteins in red blood cells: hemoglobin, cytoplasmic carbonic anhydrase (CA), and the membrane-associated electroneutral anion exchanger, band 3 protein. The large buffer capacity of blood, contained largely within red blood cells as hemoglobin, permits the convective transport of large amounts of metabolically produced CO, as bicarbonate and to a much smaller extent as carbamate. Although the equilibrium blood chemistry of 838
0161-7567/93
$2.00 Copyright
CO, was clearly established by the early 19OOs, it was not well appreciated that normal capillary transit times would permit very little CO, transfer because of the very slow rates of CO, hydration and H&O, dehydration. This biological dilemma was resolved with the discovery in red blood cells of the powerful enzyme catalyst, CA, by Meldrum and Roughton (28) in 1933. Despite the compartmentalization of hemoglobin and CA to red blood cells, plasma, in fact, transports the majority of HCO,. Thus effective utilization of the CO,-carrying capacity of plasma necessitates a mechanism for rapid movement of HCO, across the red blood cell membrane. This passive electroneutral exchange of HCO, for Cl- (chloride or Hamburger shift) is mediated by band 3 protein, an integral membrane protein discovered in 1976, the structure and kinetics of which are well established (4, 22). The quantitative contributions of red blood cell and lung CA and that of red blood cell anion (HCOi-Cl-) exchange to overall CO, elimination by the lung have been the subject of considerable study. Studies of CO, exchange in blood in vitro have yielded important kinetic data relevant to analysis of in vivo gas exchange (6, 11, l&23,29). Measurements of these kinetics and effects of inhibition in lung gas exchange have been performed utilizing inhibitors of CA and anion exchange (2,5,&l& 34) to quantitate contributions of CA and anion exchange. Taken together, these studies suggest that red blood cell CA mediates 60-80% and red blood cell anion exchange mediates l&--30% of normal CO, elimination by the lung. A number of sophisticated mathematical models based on these extensive kinetic and equilibrium blood chemistry data (1,7,19,35) arrive at similar conclusions. There are, however, possible deficiences of the in vitro lung studies that limit assessment of the true physiological contributions of red blood cell anion exchange and CA that cannot be adequately addressed by model simulations. Many isolated perfused lung studies did not use red blood cells, and the few that did may have lacked normal microcirculatory regulation. In whole animal experiments, only CA inhibitors have been studied, and their analyses are complicated by compensatory mechanisms that counter the loss of CA activity. To date, inhibitors of red blood cell anion exchange have not been studied in the whole animal. We have recently developed a method that permits measurement of CO, and 0, exchange in the course of single prolonged breaths in the intact dog (14). It accurately predicted the changes in CO, output caused by
0 1993 the American
Physiological
Society
CA
AND
RBC
BAND
3 PROTEIN
abrupt alterations in mixed venous Pco,, (PV,,,) of the blood arriving into the lung during a single breath. The strength of this method is that quantitative measurements may be made of various transport steps in CO, exchange in the intact normal lung before any compensatory adjustments occur. Thus, we sought in this study to determine the quantitative in vivo contributions of red blood cell CA and band 3 protein-mediated anion exchange by use of highly potent and selective inhibitors of these two fundamental red blood cell proteins. MATERIALS
Overview
AND
METHODS
of Experimental Approach
The principle is to measure breath-by-breath alveolar 0, and CO, exchange and, in the course of a single breath, to eliminate involvement of CA or red blood cell band 3 protein. After several normal breaths are analyzed to assure steady-state conditions, 0, and CO, flows are altered in the next respiratory cycle by an appropriately timed right atria1 infusion of drugs causing complete inhibition of these two proteins in the blood and/or lung during that breath. Thus any change in 0, uptake (VO,) and CO, production (VCOJ relative to the previous breath represents the quantitative contribution of CAcatalyzed HCO, dehydration and of red blood cell membrane HCO,-Clexchange. A pattern of artificial ventilation was chosen to ensure a duration of.exhalation long enough to permit estimates of VCO, and VO, during phase III but brief enough so that no recirculation of blood occurs within the experimental breath. Phase III is the second sloped portion of the exhaled breath (see Fig. 1) and is a mixed sample of alveolar gas having undergone exchange with the blood. Drug-induced changes in cardiac output and/or alveolar ventilation-flow relationships that might alter VCO, and VO, independent of their effects on CA or band 3 protein were monitored by simultaneous measurement of alveolar excretion of intravenously infused acetylene (C&H,). The excretion of this inert gas should not be effected by alterations in VCO, and Vo2 but only by the distribution of ventilation and perfusion and total pulmonary blood flow (10). In a steady state, C&H, is eliminated in the lungs at a constant rate and spontaneous breath-tobreath variations in the partial pressure profile of w2 in the expirate are very small. Thus any changes in the partial pressure vs. time curve of expired C,H, can be used to qualitatively detect changes in these parameters. Inhibition of band 3 protein and thus of red blood cell HCO&!lexchange was achieved with 4,4’-dinitrostilbene-2,2’-disulfonate (DNDS), a reversible competitive inhibitor of red blood cell band 3 protein (12). This agent binds directly to external- (plasma) facing anion binding sites of red blood cell band 3 protein with a dissociation constant (J&J of 2 PM at physiological chloride concentration (12). Inhibition of red blood cell CA was achieved with ethoxzolamide, a reversible highly permeant lipophilic inhibitor (25). It is a powerful inhibitor [inhibition constant (K,) = 7 X 10-l’ M], and its high lipid solubility [ether-water partition coefficient = 140 (26)] permits rapid red blood cell penetrance (21). Inhibition of only lung vascular endothelial luminal enzvme and possiblv
ROLES
IN ‘LUNG
GAS
EXCHANGE
839
extracellular-facing membrane-bound parenchymal lung cell CA with access to plasma was achieved with benzolamide, a reversible hydrophilic nonpermeant inhibitor (25). It is a powerful inhibitor (Kr = 9 X 10-l’ M), but its very low lipid solubility [ether-water partition coefficient = 0.001 (26)] greatly slows its penetrance into cells (21). All drugs were studied during normoxic [fractional concentration of inspired 0, (FI,~ ) = 0.2091 ventilation except in the case of ethoxzolamide, where hypoxic also was employed to study the PI o2 = 0.12) ventilation contribution of CA in the Bohr effect (27). It was our hypothesis that its role would be amplified in hypoxia when VO, occurs over a steeper portion of the 0, dissociation curve. With loss of leftward shift in the 0, dissociation curve during capillary transit through the lungs (dependent on CA-mediated H+ consumption), the effect of CA inhibition on VO, would therefore be greater in hypoxia than normoxia. Detailed Experimental Protocol
Twenty-three mixed-breed dogs weighing 21-29 kg were used. Anesthesia was induced with thiopental sodium (20-40 mg/kg) and mai ntaine Id with pentobarbital sodiu m (2-3 mg kg-l h-l). Dogs were intubated and ventilated with room air. To obtain prolonged expirations with a respiratory cycle lasting ~12 s (well below recirculation time), a specially designed piston ventilator was used. The ventilator is able to produce an inspiratory-to-expiratory ratio of 1:10 with a tidal volume up to 1 liter in addition to constant inspiratory and expiratory flows. We have shown previously that this form of ventilation yields normal arterial blood gas va lues in the dog (14). The ventilator is equipped with a potentiom .eter (indicating piston position), the signal of which was displayed on a strip chart recorder to obtain accurate values of expiration and inspiration times. The tidal volume of the ventilator was measured on a water-sealed spirometer (Warren E. Collins, Boston, MA) after the experiment. A Swan-Ganz double-lumen thermodilution catheter was inserted via the right external jugular vein and positioned with the distal port in the pulmonary artery and the proximal port in the right atrium. Cath .eters were also inserted in to th .e aorta and inferior ven .a cava via the femoral artery and vein. Systemic and PUlmonary arterial and airway (oral) pressures were recorded continuously in addition to intermittent measurements of the pulmonary wedge pressure. Cardiac output was measured with a thermodilution technique by use of a cardiac output computer (model 95lOA, Edwards Laboratories). A solution of C,H, was infused continuously into the inferior vena cavaata rate of 12 ml/min. The C,H, solution was prepared by mixing 600 ml of C,H, with 600 ml of a 5% dextrose solution in a l,OOO-ml evacuated bottle. The infusion was started at least 30 min before recording single-breath data to assure steady-state conditions. Hematocrits were measured in duplicate at the beginning and end of the experiment by a microcapillary technique. Pco~, PO,, and the partial pressure of C,H, (Pc,H,) in expired gas were measured with a mass specl
l
840
CA
AND
RBG
BAND
3 PROTEIN
trometer (MGA-1100, Perkin Elmer, Norwalk, CT) via a capillary inlet system sampling gas at the end of the endotracheal tube. The output signals of the mass spectrometer were transferred to a four-channel 12-bit analog-to-digital converter (resolution 5 mV) that was scanned by a computer (PDP 11/34, Digital Equipment Corporation) every 50 ms, and the recorded data were stored for later analysis. Before each experiment, the mass spectrometer was calibrated with known He, CO,, and 0, mixtures. Pc,H, was measured in arbitrary units. The respiratory frequency was initially set to 5 breaths/min with an expiration time of 10 s and a tidal volume of 900 ml. The settings of the ventilator were then adjusted to maintain an end-expiratory PCO, of ~40 Torr (adjusting frequency) and an end-inspiratory pressure ~25 cmH,O (adjusting tidal volume). The adjusted values of the inspiration time, expiration time, and tidal volume were 1.0 t 0.1 (SD) s, 9.0 t 1.4 s, and 891 t 63 ml. Duplicate measurements of the functional residual capacity (1,150 t 95 ml) were performed with a He dilution technique as follows. At the end of an expiration, the dog was switched to a bag containing a known volume and He concentration and manual rebreathing was maintained until a constant He concentration was measured (m 10 breaths in 15 s). Solutions of DNDS (100 mM), DNDS + benzolamide (100 mM/3 mM), benzolamide (3 mM), and ethoxzolamide (20 mM) in isotonic saline were prepared. These concentrations of DNDS and benzolamide are just soluble in saline after vigorous mixing and heating to 37OC. However, to dissolve the ethoxzolamide in saline, it was necessary to add an equimolar amount of NaOH, yielding a solution pH of -9.1. These solutions were infused into the right atrium (ascertained by pressure recordings) through the proximal port of the Swan-Ganz catheter (where the dead space was prefilled with the respective solution) at a rate of 46 ml/min with an infusion pump (Harvard Apparatus, Cambridge, MA). No attempt was made to make these solutions isosmolar, since dilution into the loo-fold greater cardiac output should quickly dissipate any high local osmolarity. The drugs were infused into the right atrium to obtain complete mixing and inhibitory effect before the blood arrived in the lung capillaries. The infusion was initiated at the end of the expiration in which the phase III portions of the 0, and CO, curves were used to calculate control values for VCO, and VO, and was continued throughout the duration of the succeeding breath. An outline of administration of the different drugs in the 23 dogs is given in Table 1. Eight dogs received an infusion of DNDS followed 3-4 h later by an infusion of DNDS + benzolamide. This interval was chosen to permit complete clearance of the first dose of DNDS, since it has a very rapid renal clearance (half time ~30 min; unpublished observations). In the interval, the dog was switched to a conventional ventilator. Seven dogs received an infusion of ethoxzolamide. Another seven dogs received an infusion of ethoxzolamide while being ventilated with a gas mixture containing 12% 0,. Two of these dogs received an infusion of benzolamide while being ventilated with room air -4 h before the infusion of eth-
ROLES
IN
LUNG
GAS
EXCHANGE
TABLE 1. Outline of experimental Dog No.
DNDS
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 16 19 20 21 22 23
* * * * * * * *
DNDS + Benzolamide
Benzolamide
series Ethoxzolamide Normoxia
Ethoxzolamide Hypoxia
* * * * * * * *
* * * * * * *
* * *
DNDS,
4,4’-dinitrostilbene-2,2’-disulfonate.
oxzolamide during hypoxia, whereas a third dog received only benzolamide. Expired CO,, 0,, and C&H, signals from the mass spectrometer were stored in the computer at 50-ms intervals starting three breaths before the drug infusions. Mixed venous and arterial blood samples were taken in the middle portion of the expiratory phase of the infusion period. PO,, Pcoz, and pH were measured by a blood gas analyzer (Instrument Laboratories 1302), which also calculated values of the hemoglobin 0, saturations and the base excess. Appropriate temperature corrections were made using the data of Siggaard-Andersen (32). Data Analysis
from phase III segVo2 and VCO, were determined ments of the PO, and Pcoz curves recorded during prolonged expirations. Only gas partial pressures appearing after expiration of the Fowler dead space plus 250 ml of gas (during the interval shown in Fig. 1) were used in the analysis to ensure no admixture of dead space gas. The principle used to calculate 00, and VCO, is based on the assumption that the phase III slope is entirely due to continuing gas exchange during the expiration. This assumption is discussed below. The conservation equations for 0,, CO,, and C,H, in the alveolar space in the time interval between to and tl during expiration can be expressed as ho 2avg (t 0 - t& = (FC02,V,,) (Fco2J4-J (1)
h
(~EFCO, ,dt) + & ,o,V;(Fco,
s
v”2
t1 - Fco, b)PB
t0 avp(to
-
tl) (2)
=
m&l)
-
PO,
toVto)
-
p
” to
G-F02
tw
CA AND RBC BAND
%%
a&
- t1) = (FC,H, J,,) m,H,
- J* (~EFc~H~ ,dt)
3 PROTEIN
SO
to&o)
(3)
+ PT CzH&(FC2H2
t1 -
Fw,
t*)PB
where ko,,,, VO,, , and h2H2, are the average fluxes of CO,, 0,, and C&H, across t gne alveolar membrane in the time interval. A general derivation of the gas flux equation is presented in the APPENDIX. Fco,, Fo,, and Fc,H, are the fractional concentrations of CO,, 0,, and C,H,; V is alveolar volume; i7E is expiration flow rate (constant during expiration); & coa and & c2H2are the effective solubilities of CO, and C,H, in lung tissue; VT is tissue volume; and PB is barometric pressure. Under - the assumption that net exchange of gas across the alveolar membrane approximates zero, lung volume at to is given bY V to = FRC + [VItinsp] - VD - 250 ml (4) where FRC is the functional residual capacity measured at the end of an expiration, VI is the inspiratory flow rate (constant), tinsp is the length of the inspiration, and VD is the Fowler dead space. The term 250 ml in EQ. 4 is the additional volume of expirate discarded before the phase III curve is considered to be representative of pure alveolar gas. Similarly, the lung volume at the end of the interval t, is given by V to= V to - [vE(tl - to)] (5) where VE is constant. The three gas fluxes, ho, avg, VO, avg, and \jC2H2 8v , were calculated by Eqs. 1-3 from the experimentalfy measured expired CO,, 0,, and C,H, curves between to and tl and V, and V,, calculated from Eqs. 4 and 5. The value used for VT was 180 ml (31). The value used for PT C2H2 was 0.13 ml 100 ml-l Torr-l (31). No value for lung tissue 0, solubility was used in Eq. 2, since it is virtually insoluble (31). The value used for & co2 was 0.38 ml 100 ml-l Torr-l (31) in all experiments, except those involving ethoxzolamide. In these a value of 0.08 ml 100 ml-l Torr-l was used to represent the physical solubility of dissolved CO, in lung tissue (31), with the assumption made that ethoxzolamide also inhibits all lung tissue CA. This is considered more fully in the discussion. Statistical analysis. Control and experimental values for single-breath ho,, TO,, and h,H, were compared by paired t test or Mann-Whitney U test. P < 0.05 was considered statistically significant. l
l
l
l
l
l
DNDSInfbsion
1
Ptl
to
841
ROLES IN LUNG GAS EXCHANGE
5
5
to’
40 Gl
X E
30 -
3 0"
20-
x lo-
I v 5
0 0
I
10 DNDSInfusion
l!wl *VW
I
I gl
n ~~
\” L
tn
t. n
L’
t.’
100 1
-fl E ~~ B. / t10
15
20
25
30
Time (sets) FIG. 1. Representative example of Pco~, PO,, and PC2H2 exhalation profiles before and during experimental inhibition of red blood cell band 3 protein-mediated HCO,-Cl- exchange by 4,4’-dinitrostilbene2,2’-disulfonate (DNDS) in dog 3 (see Table 2). Time intervals to-t, and t’,-t; are analyzed portions of phase III before and during DNDS infusion. Solid bar at top right denotes onset and duration of DNDS infusion. Effect of DNDS was to cause flattening of phase III slope of PCO~ profile and reduction in end-tidal Pco2 of that breath from -39 to 36 Torr. Neither O2 nor C,H, profile was altered.
The horizontal PC02 curve segment after t, and before the start of the secon .d inspi ration represents an expiratory pause required to flush the ventilator with atmospheric air before the next-inspiration. The PCO, tracing shows a marked effect (a flattening of the phase III slope) in the second expiration. With the use of Eq. 1 this change in phase III is equivalent to a change in VCO, from 139 to 125 ml/min caused by the inhibition of the HCO,-Clshift. Table 2 provides the individual data from each experiment. The average decrease in VCO, caused by DNDS in eight experiments was from 138 to 117 ml/min (15% reduction). This change was significant (P < 0.02, paired t test). vo2 and h,H, were unchanged.
RESULTS
DNDS: Inhibition of Red Blood Cell Band 3 Protein
Figure 1 shows a representative example of a PCO, vs. time curve recorded during two breathing cycles. The horizontal bar (top) shows the period of infusion of the DNDS solution into the right atrium. to and & are the times in the two breathing cycles after expiration of the Fowler dead space plus 250 ml of gas; t, and t; mark the ends of the first and the second expiration, respectively.
DNDS Plus Benzolamide: Inhibition and Vascular-Facing Lung CA
of Band 3 Protein
Table 3 provides individual data from each experiment. The average decrease in VCO, caused by DNDS plus benzolamide in eight experiments was from 121 to 101 ml/min (16.5% reduction). This change was significant (P < 0.02, paired t test #).However, the reduction. was not significantly different from the reduction in VCO,
CA AND RBC BAND
842 TABLE
3 PROTEIN
ROLES
IN LUNG
GAS EXCHANGE
2. Effect of DNDS on gas exchange during inhibition of red blood cell HCO,-ClbOz 9
P%,,
pvo,,
sao,9
svo,,
Dog No.
PH,
P&
Torr
Torr
Torr
Torr
%
%
1
7.372 7.305 7.300 7.367 7.361 7.340 7.290 7.370
7.380 7.269 7.279 7.345 7.345 7.286 7.250 7.350
31.4 38.1 39.8 34.8 39.0 32.6 38.6 28.8
41.1 44.8 44.2 37.8 43.0 42.3 47.3 32.6
89.0 82.0 95.2 104.0
43.0 48.0 49.2 52.5 55.5
96.7 95.0 96.3 98.2 97.5
75.1 77.1 78.7 84.6 86.5
110.0
51.3
97.7
81.0
88.2 97.9
46.8 55.8
95.3 97.2
7.338 to.034
7.313 to.047
35.5 k4.2
41.6 k4.6
97.9
50.4 24.2
96.7
k1.1
2 3 4 5 6 7 8
Mean t SD
% w, ml/min 1 2 3 4 5 6 7 8
Mean t SD
vco, @), ml/min
P&o,9
exchange in a single breath
117.1
t11.9
% CG, ml/min
Aho,, ml/min
G w, ml/min
Avo2, ml/min
156 18;
157 17;
-1
144 144 179
150 158 182
6 14 3
127 144 169 139 132 141 146 104
67 124 153 125 120 131 120 99
-60 -20 -16 -14 -12 -10 -26 -5
14;
15;
;
138
117 t25
-20 t17
159 k18
163 tl3
4
t18
-;
t7
1/min
Q9
Hct, %
BE, meq/l
74.5 86.9
3.1 4.5 5.1 4.5 5.1 3.6 3.7 4.6
44 36 28 34 39 36 30 41
-6.0 -6.4 -4.7 -3.1 -7.2 -7.5 -7.8
80.6 t5.0
4.3 kO.7
36 t5
-6.0 t1.6
kH, (21, au/min
A\~&H,,
%H2 (I), admin 643 560 813 155 543 835 545 448
675 563 785 145 543 825 520 441
568 t215
562 k215
-5.1
au/min 32 3 -27 -10 0 -10 -25 -7 -6 -+18
pH, and pH,, arterial and mixed venous pH, respectively; Pace, and PVcoz, arterial and mixed venous Pco~, respectively; Pao, and PVo,, arterial and mixed venous Pp,, respectively; Sao, and SV,,, arterial and mixed venous O2 saturation, respectively; Q, cardiac output; Hct, hematocrit; BE, base excess; VCO,, CO2 production; VO,, O2 consumption; iTCzHz, acetylene uptake; (l), control breath; (2), inhibited breath; A, difference between control and inhibited breaths; au, arbitrary units. * Data could not be evaluated.
observed with DNDS changed.
alone. vo2 and h,H,
Ethoxzolamide: Inhibition
were un-
of Red Blood Cell CA
Table 4 provides the individual data from each experiment in the dogs ventilated with room air (FI,, = 0.209). The average decrease in VCO~ caused by ethoxzolamide in seven experiments was from 174 to 60 ml/min (66% reduction). This change was highly significant (P < 0.001, paired t test). The decrease in vo2 was from 198 to 173 ml/min (12% reduction). This change was significant (P < 0.05, Mann-Whitney U test). $k,H, was unchanged. Table 5 provides the individual data from each experiment in the dogs ventilated with 12% 0, (FI,, = 0.12). The average decrease in VCO, caused by ethoxzolamide in seven experiments was from 123 to 40 ml/min (67% reduction). This change was significant (P < 0.005, paired t test) but was not significantly different from the reduction in VCO, found under normoxia with ethoxzolamide. The decrease in 60, was from 144 to 132 ml/min (9% reduction). This change was significant (P < 0.05, Mann-Whitney U test) but was not significantly different from the reduction in Vo2 found under normoxia. The average change in @Hz, from 509 to 486 arbitrary units/min, was not significant. Benzolamide: Inhibition of Vascular-Facing Lung CA
Table 6 provides the data from each individual experiment. The average change in VCO, caused by benzolamide was from 150 to 138 ml/min (8% reduction). The
changes in 60,) from 175 to 193 ml/min, and @H,, from 592 to 560 arbitrary units/min, were not significant. DISCUSSION
We hypothesized that it would be possible to achieve full inhibition of CA- and band 3 protein-mediated CO, and 0, exchange in vivo during single prolonged breaths with the goal of quantitatively determining their in vivo contributions. It was our purpose to assess critically the validity of previous calculations on the basis of I) model simulations of extensive in vitro blood and isolated lung data and 2) in vivo whole animal studies complicated by compensatory changes in the pattern of CO, transport that ensue with functional loss of these two proteins. The following discussion is organized into two major sections. The first deals with features of our experimental protocol and model that bear on data interpretation. The second section considers our data in relation to the large body of work that has attempted to define the quantitative physiological contributions of red blood cell band 3 protein and CA. Critique of Experimental Procedure and Model Effects of experimental ventilatory pattern. The experimental studies required a rather unphysiological ventilation profile, that of low frequency (5 breaths/min), high tidal volume (900 ml), and reduced inspiratory-to-expiratory ratio (1:5). This was necessary to maximize the time available for expiratory gas analysis in a single breath. Although this form of ventilation may cause pulmonary edema, the arterial blood gas data provided in
CA
AND
RBC
BAND
3 PROTEIN
ROLES
IN’LUNG
GAS
3. Effect of DNDS and benzolamide on gas exchange during inhibition and lung tissue CA with access to plasma in a single breath
TABLE
P%o*9
P%,9
PH,
PJ&
Torr
pvco29
-0, 9
Dog No. 1
7.346 7.309 7.360 7.368 7.405 7.318 7.370 7.430 7.363 kO.041
7.305 7.283 7.325 7.344 7.390 7.309 7.310 7.420 7.336 kO.047
34.0 35.6 35.7 33.7 31.5 34.6 30.1 25.2 32.6 t3.6
40.6 39.6 40.8 38.1 34.8 38.1 39.6 27.8 37.4 t4.3
89.0
44.0 50.0 49.9 49.9 52.9 51.4 45.2 47.8 48.9 k3.0
2 3 4 5 6 7 8 Mean t SD
t SD CA, carbonic
Torr
Torr
81.0 99.6 100.9 109.9 98.3 90.9 89.4 93.8
k9.1
vco2(l),
vco2(a, ml/min
Ako,, ml/min
% (l), mUmin
112 126 144 128 117 130 113 99 121 k14
58 121 136 114 113 112 77 78 101 t27
-54 -5 -8 -14 -4 -18 -36 -21 -20 t17
125 163 161 130 125 188 132 165 149 224
ml/min
Mean
Torr
vo2 G9, mUmin
125 148 159 134 137 184 126 160 147 t20
of red blood cell HCO;-Cl-
S%,>
svo, 9
%
%
96.4 94.9 97.2 97.4
75.0 80.2 81.7 82.5 86.5 82.2 76.0 84.3
98.1 95.8 96.6 96.9 96.7 tl.O Avo,, ml/min 0
-15 -2 4 12 -4 -6 -5 -2 t8
843
EXCHANGE
81.1 Ik3.9
exchange
l/mill
&
Hct, %
BE, meqll
3.1 4.6 4.3 3.1 4.3 3.8 2.9 3.6 3.7 to.6
46 37 29 35 39 40 30 44 38 t6
-5.4 -6.8 -4.7 -5.2 -3.9 -7.7 -7.0 -7.2 -6.0 1.4
%H, (21, au/min
v&H, (I), au/min
516 404 735 848 690 987 590 484 657 t197
545 405 731 819 766 956 580 414 652 t198
A~,H,,
au/min 29
1 -4 -29 76 -31 -10 -70 -5 t44
anhydrase.
Tables 2-6 show normal values in all dogs appropriate for the inspired 0, fraction. There were no significant changes from normal over 6-8 h in arterial oxygenation, total static respiratory system (lung plus chest wall) compliance, cardiac output, and systemic or pulmonary arterial wedge pressure. The mean pulmonary arterial pressure rose slightly by 5 mmHg. In our previous study (14) we determined that this ventilation profile results in a
cardiac output that varies -25% a respiratory cycle. Our error showed that single breath VCO, when compared with a constant equal to the mean.
about the mean within analysis in that paper would vary only by 3% pulmonary blood flow
Timing of inhibitor infusion. A precise timing of the drug infusions was essential in these experiments to achieve a constant and fully inhibitory concentration of
4. Effect of ethoxzolamide on gas exchange during inhibition of red blood cell and lung tissue CA in a single breath in normoxia
TABLE
Dog No.
pacoz 9
f&,2
pvo,9
saoz 3
PH,
Torr
=co, 7
PHll
9 10 11 12 13 14 15
7.340 7.380 7.360 7.400 7.370 7.350 7.308 7.358 to.030
7.320 7.300 7.330 7.310 7.330 7.310 7.253 7.307 kO.026
30.6 29.2 35.3 27.6 35.5 37.0 35.8 33.0 t3.8
33.4 38.4 41.7 35.0 40.8 42.4 46.5 39.7 t4.5
79.0 92.0
47.0 42.0 43.0 41.0 49.0 45.0 46.4 44.8 t2.9
95.0
79.5
97.1 96.5 95.0 96.7 97.0 94.3 95.9 21.1
72.5 75.3 71.7 81.9 76.1 67.7 75.0 24.8
Mean t SD
vco, w, 9 10 11 12 13 14 15
Mean t SD * Data
could
vco2m,
ml/min
ml/min
156 184 176 202 175 136 188 174 t22
66 65 56 61 54 51 67 60 t6
not be evaluated.
Torr
Torr
Torr
88.0 74.0 89.0 95.0 95.4 87.4 ,t8.1
Ako,, ml/min
a (0, ml/min
h (a, ml/min
-90
179 198 206 169 190 248 199 198 225
166 153 177 200 180 148 188 173 t19
-119 -120 -141 -121 -85 -121 -114 Ik20
%
A\jo2, ml/min
-13 -45 -29 31 -10 100 -10 -25 *40
=o, , %
Q
Hct, %
Vmin
4.6 3.9 3.4 3.6 4.6 3.5 4.0
39
38 45 40 47 46 37 42 +4
3.9
to.5
v&H, (11, au/min
%H, (2), au/min
188 575 826 642 711 466 864 610 5232
170 662 745 588 639 554 861 603 zk217
BE, meq/l
-7.4 -5.9 -3.9 -5.3 -3.2 -4.1 * -5.0 1.5 A~c~H~,
au/min
-18 87 -81 -54 -73 88 -3 8 *71
844
CA AND
RBC
BAND
3 PROTEIN
ROLES
IN
LUNG
GAS
EXCHANGE
5. Effect of ethoxzolamide on gas exchange during inhibition of red blood cell and lung tissue CA in a single breath in hypoxia TABLE
PaCO, 9
Torr
82.2 87.9 86.4 86.7 85.1 85.5 77.8
38.4 t4.7
51.3 t4.8
35.8 k3.9
84.5 t3.5
37.9
7.360 kO.044
7.329 to.034
33.0 k5.7
109 105 92 135 136 100 181
Mean t SD
123 -t-31
* Data
could
Torr
%
vo, (l),
00, c9,
Avo,,
ml/min
ml/min
ml/min
ml/min
3; 34 33 50 24 63
-7; -58 -102 -86 -76 -118
132 132 126 155 138 91 236
12; 106 122 132 102 211
-1; -20 -23 -10 11 -25
40 214
-85 k22
144 t45
132 t40
-13 t13
%
hH,
l/min
VW-b (3,
(I),
admin
admin
-6.1 t1.5 A\~c~H~,
au/min
459
-1
382 362 637 555 394 775
37; 303 598 557 370 709
-24 -66
509 t154
486 t159
-32 t28
-59 -39
2
not be evaluated.
Inhibitor
selection and calculated degree of inhibition.
Our analysis of the effects of red blood cell anion exchange and CA inhibition assumes 100% in vivo effect. Our measurements of average transit time from right atrium to small pulmonary artery clearly permit us to use, as a minimum, 4 s for transit to the capillaries. Thus it was necessary to select inhibitors with strong and rapid binding to their receptors and, in the case of CA, rapid diffusibility into red blood cells (ethoxzolamide) or very slow red blood cell uptake (benzolamide). DNDS. In the case of red blood cell band 3 protein (anion exchange inhibition), the reversible nontoxic stilbene disulfonate, DNDS, was selected. On average, the rate of DNDS infusion was 1% of cardiac output, resulting in an instantaneous 60-fold dilution of drug (100 mM to 1.66 mM) into the plasma of blood with a 40% hematocrit. If we assume 90% plasma protein binding at these very high concentrations and a short binding time (
