ELSEVIER Respiration Physiology 96 (1994) 259-272

R E S P I R A T I O N P H Y S I O L O G Y

Acid-base disequilibrium in the arterial blood of rainbow

trout

Kathleen M. Gilmour *, D.J. Randall I and Steve F. Perry

Department of Biology, University of Ottawa, 30 Marie Curie, Ottawa Ontario, KIN 6N5 Canada

Accepted 13 December 1993

Abstract

An extracorporeal blood circulation and a stopflow technique were used to examine the acid-

base status of arterial blood in the rainbow trout, Oneorhynchus mykiss. Arterial blood was routed

from the coeliac artery through an external circuit in which pH (pHa), partial pressure of oxy-

gen (Pao2) and partial pressure of carbon dioxide (Paco2) were monitored continuously. The stopflow condition was imposed by turning off the pump which drove the external loop. A

radioisotopic CO2 excretion assay was performed on blood samples collected periodically to

evaluate plasma carbonic anhydrase (CA) activity and hence red blood cell (rbc) lysis. An

acid-base disequilibrium was found in the post-branchial blood; pHa increased by 0.04-0.06

units, and Pac% by 0.03-0.10 Torr, during the stopflow period. The disequilibrium appeared to

arise primarily from the slow (uncatalyzed) rate of plasma HzCO 3 dehydration. This was con-

firmed by the intra-arterial injection of bovine CA (22 mg kg -1) prior to the stopflow; the dis-

equilibrium was abolished. When the CA inhibitor acetazolamide (30 mg kg -1) was injected, a

negative pH disequilibrium of 0.04 units, accompanied by a rise in Paco 2 of 0.57 Torr, was observed during the stopflow. These results can be explained by the acetazolamide-induced in-

hibition of rbc CA, which leads to continuing rbc CO 2 "excretion" in the post-branchial blood.

Key words: Acid-base equilibrium, arterial blood; Carbon dioxide; excretion, acid-base disequi-

librium; Carbonic anhydrase, fish blood; Fish; trout (Oncorhynchus mykiss)

1. Introduction

The process of carbon dioxide excretion in fish has been described (Fig. 1) (see re-

views by Perry, 1986; Perry and Wood, 1989; Perry and Laurent, 1990). Dehydratio n

of H2CO3 in the red blood cell (rbc) is rapid, owing to the presence of the enzyme

* Corresponding author. Tel.: (613) 564-6890; Fax: (613) 564-5014. Present address: Dept. of Zoology, University of British Columbia Vancouver, B.C. Canada.

0034-5687/94/$7.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0 0 3 4 - 5 6 8 7 ( 9 4 ) 0 0 0 0 6 - L

260 K.M. Gilmour et aLI Respiration Physiology 96 (1994) 259-272

RBC

H b <

+,,co;,"> c o , . , ..

plasma H+ 2 ~ " CO2 + H C O ~ --

gill epithelium

water

C O 2

Fig. 1. A diagrammatic representation of CO2 excretion in rainbow trout. CA = carbonic anhydrase, Hb = haemoglobin, HbH = protonated haemoglobin. Revised from Perry (1986).

carbonic anhydrase (CA), and produces molecular CO 2 which diffuses into the venti-

latory water. The H C O ~ for the dehydration reaction enters the rbc via the band 3

protein C 1 - / H C O 3 exchanger, while protons are provided by the dissociation of in-

tracellular buffers and the oxygenation of haemoglobin. HzCO 3 dehydration also takes

place in the plasma, but at the uncatalyzed rate, since CA activity is unavailable to

plasma flowing through the gills (Henry et al., 1988; reviewed by Perry and Laurent,

1990). In this respect, CO2 excretion in fish differs from that in mammals; lung endot-

helial CA in the latter provides an important secondary site of H2CO3 dehydration

(Perry and Laurent, 1990). The slow rate of the uncatalyzed HzCO 3 dehydration re-

action leads to the interesting possibility of an acid-base disequilibrium occurring in the

arterial (post-branchial) blood. Experimental results and computer simulations in mam-

mals indicate that, even in the presence of lung CA activity, a small postcapillary

K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272 261

acid-base disequilibrium may exist (Crandall and Bidani, 1981). When endothelial CA

activity is inhibited, extracellular HzCO 3 dehydration continues in the postcapillary

blood, giving rise to a significant alkaline pH 5mbalance (reviewed by Bidani and

Crandall, 1988). Therefore, the unavailability of CA activity to plasma in the gills in fish

might be expected to engender an alkaline acid-base disequilibrium.

The process ofrbc C1-/HCO3- exchange could also conceivably create an acid-base

disequilibrium in arterial blood. The transit time of blood in the gill is probably 0.5-

2.5 sec, while the reaction time (T67; Cameron, 1978) of rbc C1- /HCO£ exchange is

0.4 sec in rainbow trout (Cameron, 1978), so it is possible that, at least under some

conditions, rbc C1-/HCO3- exchange is not completed during the passage of the blood

through the gills. In post-branchial blood in which C1-/HCO3- exchange was not

finished, the continued loss of HCO3- to the rbc would be expected to result in an

acid-base disequilibrium tending towards acidification of the plasma (Crandall and

Bidani, 1981; Crandall et al., 1981; see reviews by Bidani and Crandall, 1988; Klocke,

1988). The goal of this study, then, was to determine whether an acid-base disequilibrium

is present in the arterial blood of rainbow trout. Secondly, we wished to establish the

underlying basis of any disequilibrium observed. An extracorporeal circulation (Thomas

and Le Ruz, 1982) in combination with a stopflow technique, a procedure well suited

to the detection of acid-base disequilibria, has been used. The experimental set-up is

similar to those which have been utilized for in vivo experiments on mammals (Bidani

and Crandall, 1988). Since rbc lysis has complicated the use of intact mammalian

preparations (e.g. Hill et al., 1977; Bidani and Crandall, 1978a), a recently developed

radioisotopic CO2 excretion assay (Wood and Perry, 1991) was used in this study to

provide an index of plasma CA activity, and hence haemolysis.

2. Materials and methods

Experimental animals The experiments were carried out on rainbow trout (Oncorhyn-

thus mykiss) weighing between 553 and 1346 g (experimental N= 12), obtained from

Linwood Acres Trout Farm (Campbellcroft, Ontario). An additional group of fish,

weighing between 253 and 363 g (experimental N= 8), was used to provide control

blood samples for the CO2 excretion assay. All fish were maintained on a 12L:12D

photoperiod in large fibreglass aquaria supplied with flowing, aerated, and dechlori-

nated City of Ottawa tap water ( [Na+]= 0.10 mmol L -1, [C1-] =0.15 mmol L-~;

[Ca 2+ ] = 0.35-0.40 mmol L 1, [K + ] = 0.03 mmol L -1, pH = 7.7-8.0). The water tem-

perature was 6-8 °C over the course of the experiments (December-January). Fish

were fed daily to satiation on a diet of commercial trout pellets; food was withheld

24 h prior to experimentation. Trout were acclimated to these conditions for at least

4 weeks before experiments were initiated.

Animalpreparation Fish were anaesthetized in a neutralized (pH 7.5-8.0) solution of

ethyl-m-aminobenzoate (0.1 g 1-1; MS 222), then transferred to an operating table that

permitted continuous irrigation of the gills with oxygenated anaesthetic solution. An

262 K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272

indwelling cannula was implanted into the dorsal aorta (Soivio et al., 1975) using

flexible polyethylene tubing (Clay-Adams PE 50; internal diameter = 0.580 mm, outer

diameter =0~965 mm); this permitted blood sampling from the control trout, and

measurement of dorsal aortic pressure (Pda) in the fish prepared for extracorporeal

experiments. Three additional catheters were placed in trout designated for extracor-

poreal trials. To measure ventilation amplitude, a catheter was inserted into the buc-

cal cavity using PE 160 tubing (internal diameter = 1.14 mm, outer diameter = 1.57 mm).

Two cannulae (PE 50) were implanted into the coeliac artery in the orthograde and

retrograde directions (Thomas and Le Ruz, 1982). Fish were revived on the operating

table by irrigation of the gills with water, and were then transferred to the experimen-

tal chamber where they were allowed to recover for 24 h.

Extracorporeal circulation and analytical procedures The extracorporeal circulation

(Fig. 2) was initiated by connecting the two coeliac artery cannulae in series with

cuvettes containing Po2, P¢o2 and pH electrodes. A peristaltic pump maintained blood

flow in the circuit at 1.2 ml min-1; at this flow rate, the transit time for blood flow from

the fish to the electrodes was approximately 30 sec. The volume of the external loop

was 1 ml, representing less than 4~o of the total blood volume of the fish. To prevent

clotting, the circuit was rinsed with 10 ml of heparinized (540 units m1-1) saline before

initiating blood flow. Using this extracorporeal loop, continuous recordings of arterial

PQ (Pao~), Pco~ (Pacts) and pH (pHa) could be acquired for extended periods of time.

The stopflow condition was imposed simply by turning off the peristaltic pump.

A Metrohm combination glass pH electrode (model 6.0204.100(OC)) in conjunction

with a Radiometer PHM 73 meter was used to measure pHa. The 50~o response time

of this electrode was approximately 25 sec for a 0.14 unit pH change in buffer solu-

tions. Pao~ and P a c t ~ were measured using Radiometer Po2 and Pco~ electrodes (E-

5046 and E-5036, respectively) and the PHM 73 analyzer. All three electrodes were

housed in thermostatted cuvettes maintained at ambient water temperature. The blood

gas electrodes were calibrated by pumping saline equilibrated with appropriate gas

mixtures (supplied by a W0sthoffgas-mixing pump) through the circuit. Similarly, buffer

solutions were used to calibrate the pH electrode. For the purpose of these experiments,

it was important to use a pH electrode which was relatively insensitive to changes in

flow. The sensitivity of the electrode was tested by performing stopflow trials on a blood

sample equilibrated with 0.5 or 0.1~o CO2 in air; the pH change associated with

stopping the flow was 0.007 _+ 0.003 units (mean + SEM; N= 6).

PQ of the inflowing water (Pwo~) was continuously monitored (Radiometer PHM

72 Mk2 meter) by siphoning a small volume of water from the experimental chamber

through a thermostatted cuvette containing a PQ electrode (Radiometer, E-5046). A

continuous measurement of dorsal aortic blood pressure was obtained by connecting

the dorsal aortic cannula to a pressure transducer (Bell&Howell, 4-327-I). Occasion-

ally, Pda decreased rapidly upon initiation of the extracorporeal circulation; in these

cases, the decrease in Pda was assumed to reflect internal bleeding and the fish was

rejected. Although haematocrit data were not collected in the present study, the hae-

matocrit was measured as 29 _+ 2~o (mean + SEM; N= 18) in a different series of ex-

periments performed by the same workers, using the same techniques and on the same

K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272 263

Data acquisition

II, Jl .... ' -I

meters pHa Pao2 Pacoz[

I TT .o. ............ | /

/

coeliac

ventilation

J 1'"aT

water

)

Fig. 2. A schematic representation of the extracorporeal preparation. Blood flow through the external cir- cuit is maintained by a peristaltic pump. The output of the blood and water electrodes and the pressure transducers is collected and stored using a data-acquisition program on a PC. pHa = arterial pH, Pao2 - partial pressure of 02 in arterial blood, Paco 2 = partial pressure of CO2 in arterial blood, ventilation = ventilation amplitude, Paa = dorsal aortic pressure, Pwo2 = water partial pressure of 02.

s tock o f fish. I t is unl ikely tha t the h a e m a t o c r i t o f the fish in the p re sen t s tudy differed

significantly f r o m this value. T h e bucca l cavi ty ca the te r was c o n n e c t e d to a s econd

p res su re t r ansduce r , and the a r i thmet ic dif ference be tween insp i ra to ry and exp i ra to ry

bucca l p ressures was used as a m e a s u r e o f ven t i l a t ion ampl i tude . In addi t ion , b rea th ing

f requency was m e a s u r e d per iod ica l ly by visual ly coun t ing bucca l m o v e m e n t s . T h e

p ressu re t r ansduce r s were ca l ib ra ted agains t a stat ic c o l u m n o f water .

264 K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272

The analog outputs from the pH, blood gas and water gas meters, as well as from

the pressure transducers were transformed into digital output using an analog-digital

interface (Data Translation Inc.). Customized software (written by P. Thoren; G0te-

borg, Sweden) was used to acquire data continuously; mean values for each variable

were recorded and stored at 5 sec intervals.

C02 excretion assay and analyticalprocedures The radioisotopic C O 2 excretion assay

has been described in detail by Wood and Perry (1991). The assay was used in this

study to quantify rbc lysis by providing an index of plasma CA activity, and therefore

was performed only on plasma samples. HzCO 3 dehydration rates for fish from extra-

corporeal experiments were measured on plasma samples obtained following each

stopflow period, and were compared to those for plasma samples from dorsal aortic-

cannulated (DA-cannulated) fish. Basal CO2 excretion rates were determined on plasma samples from DA-cannulated fish to which the CA inhibitor acetazolamide (Diamox;

Sigma; final concentration = 10-4 mol 1-1) had been added.

Blood samples obtained from the coeliac artery of fish prepared for extracorporeal

experiments, or by withdrawal from the dorsal aortic cannula of DA-cannulated fish,

were centrifuged (5900 xg for 2 min; 4 °C) to yield plasma. One ml aliquots of the

separated plasma were placed in glass scintillation vials (20 ml), 2 #Ci (10 /~1 of

200 #Ci ml-1) of sodium [ 14C]bicarbonate (in teleost Ringer; Wolf, 1963) were added,

and the vial was immediately sealed with a rubber septum. A plastic well, suspended

from the septum, contained a fluted filter paper impregnated with a CO2 trap (150 #1

hyamine hydroxide). The vial was placed in a shaking water bath for 3 min, after which

the filter was removed and assayed for 14C activity, as was a 50 #1 aliquot of plasma.

A second 50 #1 aliquot of plasma was assayed for total CO2 concentration (Cco~). The H2CO 3 dehydration rate for each vial was calculated by dividing filter paper 14C activity

by plasma specific activity (given by the plasma 14C activity and Cco~) and time.

Filter paper and plasma 14C activities were determined by liquid scintillation counting

(Packard TR 2500) with automatic quench correction. A commercial scintillation cock-

tall (ACS II; Amersham) was employed. Plasma Cco 2 was measured using a Corning

model 965 CO 2 analyzer.

Experimental protocol The experimental protocol consisted of imposing three con-

secutive 6 min stopflow periods on each fish (experimental N = 6). Once the measured

ventilatory, cardiovascular and blood respiratory variables had stabilized (generally

within 10-30 min of initiating the extracorporeal circulation), a control stopflow was

performed. The pump was then restarted, fish were given a bolus injection of bovine CA (200 Wilbur-Anderson units per rag; 22 mg kg -1 dissolved in 1 ml teleost Ringer)

and, when the measured variables had stabilized, a second stopflow was imposed. Finally, the fish was injected intra-arterially with acetazolamide (30 mg kg-1). This

resulted in a brief but pronounced alkalosis which was followed by a progressive aci-

dosis. The stopflow was initiated following the alkalosis, as pHa returned to pre-

acetazolamide levels (but was still decreasing). Ventilation frequency was measured during each stopftow period. Blood samples (1.5 ml) for each stopflow period were

collected 3 min after blood flow was re-initiated and were replaced by an equal volume

K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272 265

of teleost Ringer. In the associated control experiments, which were performed on a

second group of 6 fish, the CA and acetazolamide treatments were replaced by sham

injections of teleost Ringer.

Statistical analys& Data are presented as means + 1 s tandard error of the mean

(SEM). Statistical differences were determined by analysis of variance followed by

Fisher 's LSD multiple comparison test; 5~o was taken as the fiducial limit of signifi-

cance. The apparent half-time for the pH electrode response speed was estimated from

a plot of mean pH versus time by determining the time at which 50~o of the total pH

change had occurred.

3. Results

Plasma CA activity The H2CO 3 dehydration rate of p lasma from DA-cannula ted fish

to which the CA inhibitor acetazolamide had been added was significantly lower than

10

~ 6

~,~,~ 4

DA-cannulated Extracorporeal m

l

control ACTZ control CA CA+

ACTZ

Fig. 3. Plasma H2CO 3 dehydration rates, measured using an in vitro radioisotopic C O 2 excretion assay, for blood samples from DA-cannulated and extracorporeal fish. Acetazolamide was added to plasma samples from DA-cannulated fish (final concentration = 10 -4 mol 1-1). Dehydration rates for extracorporeal fish were measured on samples taken following each stopflow period: control, after the injection into the circulation of bovine CA (22 mg kg l) and following CA and acetazolamide (30 mg kg -I) infusion. * indicates a sig- nificant difference (P < 0.05) from the DA-cannulated control. Experimental N are shown in each bar of the histogram. CA = carbonic anhydrase, ACTZ = acetazolamide.

266 K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272

that of plasma alone (Fig. 3). However, there were no significant differences in the

H2CO 3 dehydration rates of control plasma samples from DA-cannulated and extra-

corporeal fish (Fig. 3), indicating that rbc lysis in the extracorporeal fish was no more

severe than is generally tolerated. Further, rbc lysis did not increase substantially with

the length of time the extracorporeal circulation was maintained; the plasma H2CO 3

dehydration rates of blood samples taken after successive control stopflow periods did

not differ significantly (data not shown). The injection of bovine CA into the circula-

tion of fish in extracorporeal experiments increased the plasma CA activity approxi-

mately three-fold (Fig. 3). This extra activity was inhibited by the infusion of aceta-

zolamide (Fig. 3).

Acid-base disequilibrium Mean ventilation frequencies and amplitudes during the three

stopflow periods are shown in Table 1 for both experimental and control fish. There

were no significant differences in ventilation parameters as a result of the drug treat-

ments, nor were the ventilation parameters of the drug-treated fish significantly different

from those of saline-injected fish.

Fig. 4 demonstrates that an acid-base disequilibrium is present in the arterial blood

of rainbow trout. Each line in Fig. 4 consists of continuous mean values for 6 fish;

standard errors are shown only every two min for clarity. Because the magnitude of

disequilibrium events was small relative to individual variability in pHa and Paco2, the

data for individual fish were normalized by subtracting from each value in a response

the value at the beginning of the stopflow period (Table 1). Data for Pao2 are not re-

ported since while the flow of blood was stopped they reflect the movement of 02 into

the rbc and the consumption of O2 by the electrode; Pao2 responses were not affected

by the drug treatments. The increase in pHa which occurred during the control stopflow

period was accompanied by a tendency for Paco 2 to rise (Fig. 4b, Table 1); due to the

small and variable nature of the response the increase was not significant. The

disequilibrium was reproducible, in that when three successive stopflow periods were

imposed, the extent of the pHa and Paco 2 increases did not change significantly

(Table 1).

Table 1

Ventilatory parameters during the stopflow periods, magnitudes of pHa (ApHa) and Paco2(APaco2) changes

during the stopflow period, and absolute values of pHa and Paco 2 at the beginning of the stopflow period.

Values are means _+ 1 SEM; N = 6 for each treatment. * indicates that APaco2 is significantly (one sample

t-test, P < 0.05) different from 0. fR = respiratory frequency, VA = ventilation amplitude.

fR VA pHa ApHa Paco 2 APaco 2 (min - 1) (cm H20 ) (Torr) (Torr)

Control 59_+3 5.37_+0.19 7.94_+0.03 0.04_+0.01 2.13_+0.15 0.03_+0.03

CA 62_+4 5.20_+0.45 7.99_+0.03 0.00 -+ 0.00 2.10_+0.20 0.00_+0.02

Acetazolamide 56 _+ 2 5.11 _+ 0.38 7.89 _+ 0.04 - 0.04 _+ 0.01 2.25 _+ 0.05 0.57 _+ 0.008*

Control 60 _+ 3 4.46 _+ 0.46 7.91 _+ 0.05 0.04 + 0.01 1.92 _+ 0.27 0.08 + 0.02*

Saline 59_+4 4.51_+0.50 7.90_+0.06 0.06+_0.02 1.71_+0.25 0.05_+0.02*

Saline 57_+4 4.62+0.53 7.88_+0.06 0.06_+0.02 1.65_+0.23 0.10_+0.02"

K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272 267

(a)

(b)

O N

t~

0.05

0.00

-0.05

-0.10

0.07

~ , 0.06

o 0.05

0.04 t.)

0.03

0.02 N

:~ 0.01

o.oo

0.00 0

-0.02

-0.03 0

:: ~.. • \

I

0 3 9

I I

6 12

t ime (min)

cont ro l

. . . . . . . C A

. . . . . . A C T Z

i i I

3 6 9

t ime (rain)

0.6

0.4

0.2

0.0

/

,z' J

I

0 3 6

/"

]./

9 12

I

12

Fig. 4. Mean normalized (see text) values for (a) pHa and (b) Paco2 (N = 6) during blood flow and the three consecutive stopflow periods (marked by the dotted lines): control (solid line), after CA injection (dashed

line) and after acetazolamide infusion (dot-dash line). Due to the difference in the magnitude of Paco 2 changes

under control and CA treatment versus acetazolamide treatment, the results for acetazolamide have been displayed separately (insert in b). Error bars represent SEMs and are shown only every 2 min for clarity.

* indicates a significant difference (P<0.05) at 9 min from the CA value, while + indicates a significant dif-

ference (P< 0.05) at 9 rain from the control value. CA = carbonic anhydrase, ACTZ = acetazolamide.

I n t h e p r e s e n c e o f e x o g e n o u s b o v i n e C A , t h e a c i d - b a s e d i s e q u i l i b r i u m w a s e s s e n t i a l l y

a b o l i s h e d a n d P a c o 2 d i d n o t i n c r e a s e (F ig . 4, T a b l e 1).

268 K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272

The injection of acetazolamide caused an increase in pHa which was followed by

a progressive acidosis and hypercapnia owing to the inhibition of rbc CA activity. Note

that the initial alkalosis was probably a result of the high pH (approximately 8.5) re-

quired to make acetazolamide soluble in saline at the desired concentration (30 mg

ml-I). The stopflow period was initiated as the arterial pH returned to the pre-injection

level (but was still declining); a negative pH disequilibrium accompanied by a large,

significant increase in PacQ was observed (Fig. 4a, b inset, Table 1).

4. Discussion

Critique o f experimental procedure The aim of this study was to clarify the acid-base

status of the arterial blood of rainbow trout. The method used, which was similar in

concept to established techniques for mammalian preparations, involved stopping the

flow of blood through an external circuit and monitoring continuing pHa and Paco2

changes. The existence of such changes supplies evidence that an acid-base disequi-

librium is present in the post-branchial blood.

A number of factors complicate the measurements made using the extracorporeal

stopflow technique. The electrodes must be insensitive to changes in flow; when the

electrodes used in this study were tested, artifacts associated with changes in flow were

found to be small relative to the magnitude of the disequilibrium. It is also likely that

unstirred layers exist within the stopflow apparatus. Although their influence on pH may

be ignored since the process under consideration is relatively slow, the effect of such

stagnant layers on diffusional CO2 movements, and hence Paco 2 data, may be signif-

icant (Klocke, 1988). To allow accurate estimation of reaction rates, the electrode re-

sponse time should be substantially less than the reaction half-time. The half-time of

the uncatalyzed H2CO 3 dehydration reaction, 25-90 sec at physiological temperature

(Perry, 1986), is greater than the measured 50~o response time of the pH electrode

(25 sec).

The transit time of the blood from the gills to the recording sites (approximately

30 sec) may lead to an underestimation of the extent of the disequilibrium because

acid-base changes begin as soon as the blood leaves the branchial vasculature. This

underestimation can be reduced by minimizing tubing lengths, but is to some degree

inevitable given the constraints of the experiment set-up. The magnitude of the pH

change observed also depends on the degree of haemolysis. Carbonic anhydrase re-

leased from lysed rbcs will decrease the extent of the disequilibrium (Fig. 4, Table 1).

Although haemolysis has been a severe problem in stopflow experiments on mamma-

lian preparations (Hill etal., 1977; Bidani and Crandall, 1978a), the results of

the in vitro CO2 excretion assay demonstrated that rbc lysis in extracorporeal fish was

not elevated in comparison to DA-cannulated fish (Fig. 3). Further, the addition of

acetazolamide to extracorporeal fish did not significantly reduce the rate of H2CO 3

dehydration in the plasma (Fig. 3). The assay also indicated that the infusion of bovine

CA was an effective means of elevating plasma CA activity, and that this additional

activity could be inhibited by the injection of acetazolamide (Fig. 3).

K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272 269

Basis o f the disequilibrium The issue of pH disequilibria in mammals has been ad-

dressed by a number of workers using computer modelling, isolated lung preparations,

and in vivo stopflow techniques (see reviews by Bidani and Crandall, 1988; Klocke,

1988). Early computer models predicted that, as a result of the slow (uncatalyzed) rate of plasma H2CO 3 dehydration, slow postcapillary increases in plasma pH would occur following gas exchange in the lungs (Hill et al., 1973; Forster and Crandall, 1975).

However, when these predictions were tested using isolated lung preparations and cell-free perfusates, postcapillary pH changes were absent or much smaller than pre- dicted except in the presence of CA inhibitors (Crandall and O'Brasky, 1978; Effros et al., 1978; Klocke, 1980). These results argued strongly for the accessibility of lung CA to perfusate reactions, and morphological data later provided confirmation of the existence of lung endothelial CA (see review by Bidani and CrandaU, 1988). Although

extrapolation to the in vivo situation implied that a state of equilibrium should exist in

blood leaving lung capillary beds, conflicting results have been obtained from in vivo

stopflow experiments on anaesthetized mammals. Both alkaline and acidic imbalances, in addition to equilibrium conditions, have been measured (Hill et aL, 1977; Bidani and

Crandall, 1978a; Rispens et aI., 1980; Chakrabarti et al., 1983). The consensus appears to be that postcapillary pH changes are small or absent in mammals trader normal conditions, owing to the presence of lung CA.

While the basic mechanism of CO2 excretion is similar in fish and mammals (Perry and Laurent, 1990), a major difference is the inaccessibility of CA activity to plasma in the gills in fish (Henry et al., 1988); in mammals, plasma HzCO 3 dehydration is catalyzed by both rbc and pulmonary endothelial CA (see reviews by Bidani and Crandall, 1988; Perry and Laurent, 1990). The absence of plasma CA activity in fish implies, by theory, that an alkaline pH disequilibrium should exist in the post-branchial

arterial blood. Both pHa and Pac t 2 increases were observed during the stopflow pe-

riod (Fig. 4, Table 1): the loss of CO2 to water in the gills drives plasma CO2- HCO3- -H + reactions towards HaCO 3 dehydration (Fig. 1) and because this process is uncatalyzed, it continues after blood leaves the branchial vasculature. The dehydra- tion reaction consumes protons (Fig. 4a) in forming CO 2 (Fig. 4b).

The alkaline imbalance essentially disappears in the presence of elevated plasma CA (Fig. 4a). This result is expected if the disequilibrium is, indeed, caused by the slow rate of plasma H2CO 3 dehydration. Similarly, in vivo experiments on mammalian prepara- tions, as well as computer models, indicate that CA released into the blood by inten- tional haemolysis abolishes postcapillary pH changes (Hill et al., 1977; Bidani and Crandall, 1978a).

By theory, an acidic pH disequilibrium would be anticipated if C1 -/HCO3- exchange

did not reach completion during the passage of blood through the gills because this exchange transfers HCO3- from the plasma to the rbc (Fig. 1). This prediction has been confirmed in experiments on mammalian preparations in which the rate of anion transport was reduced using C1-/HCO3 exchange inhibit0rs (Crandall et al., 1981). In the presence of CA the alkaline disequilibrium in the arterial blood of rainbow trout was abolished (Fig. 4a), but an acidic imbalance was not observed. This suggests that either a sufficient alkaline disequilibrium persisted despite the CA treatment to counter any acidic imbalance present, or, that C1-/HCO3- exchange reached completion during

270 K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272

the gill transit time. A potential role for anion transport to contribute to acid-base

imbalances under different conditions cannot, however, be ruled out. In particular,

circumstances which decrease the time the blood spends in the gill might lead to the

participation of C1-/HCO3 exchange in disequilibrium events, as the reaction time

(T6v = 0.4 sec; Cameron, 1978) for the exchange is close to the normal transit time of blood in the gill (0.5-2.5 sec).

The intention of the acetazolamide treatment was to inhibit the exogenous CA

previously added, thereby re-establishing the original alkaline disequilibrium. However,

in the presence of acetazolamide an acidic pH imbalance coupled with a large increase

in Pact2 was observed (Fig. 4, Table 1). Bidani and Crandall (1978b) detected a similar

pH decrease in stopftow experiments on anaesthetized mammals (dogs or cats) when

30 mg kg -1 acetazolamide was added to the blood. These results are explained by the

acetazolamide-induced inhibition of rbc CA in addition to exogenous plasma activity

(Bidani and Crandall, 1978b; Henry e ta l . , 1988). The disturbance in rbc CO 2-

HCO3- -H + reactions generated by the diffusional loss of COa to the water (Fig. 1)

could not, in this situation, be quickly balanced. That is, rbc CO 2 production, which

is normally completed before the blood leaves the gills, proceeded post-branchially,

driven forward by the production of Bohr protons from the oxygenation of haemoglobin

(Klocke, 1988; Perry and Gilmour, 1993). The CO2 formed in the rbc after blood leaves

the gills diffuses into the plasma. The ensuing rise in P a c t 2 drives plasma CO2-

HCO3- -H + reactions towards CO 2 hydration, generating a concomitant decrease in

pHa; these reactions continue during the stopflow period, giving rise to the acidic pH

imbalance (Fig. 4). In addition, anion exchange will persist in the post-branchial blood

in response to the continuing rbc CO2 "excretion", and this will contribute to the

negative disequilibrium.

The physiological significance of the disequilibrium in the arterial blood of rainbow

trout is uncertain. Indeed, it is possible that the disequilibrium is simply a consequence

of the lack of plasma CA activity in fish, and serves no physiological function, p e r se.

In this case, the question of why fish gills lack functional quantities of "endothelial"

CA must be addressed. Perry and Laurent (1990) suggested that an additional site of

H 2 C O 3 dehydration is not required in fish because water has a high CO2 capacitance, the countercurrent mechanism of gas transfer is very efficient, and the residence time

of blood in the gills is relatively long; additionally, fish exhibit a large Haldane effect.

The results of the present study indicate that previously reported measurements of

pHa are equilibrium values rather than those of blood leaving the gills. Whether or not

the CO2-HCO3- -H + system in fish blood ever reaches equilibrium in vivo is a matter

of speculation. Presumably, the excretion of CO2 into the blood from tissues also results

in disequilibria in venous blood. This has consequences for the extensive discussion of

the alpha-stat hypothesis, which is based entirely on pH measurements of blood at equilibrium, a situation that may never occur in vivo.

Based on the model of CO 2 excretion for trout (Fig. 1), the extent and/or direction

of the disequilibrium are likely to be affected by a number of factors (see also Bidani and Crandall, 1988). The CO 2 diffusion gradient may influence acid-base balance as

it determines the magnitude of the initial disturbance in C O 2 - H C O 3 - H + reactions.

Through their effect on pHa and Paco~ (Perry and Wood, 1989), ventilation param-

K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272 271

eters must also be considered. For this reason, ventilation amplitude and frequency

were monitored in this study (Table 1). Arterial pH is an important component of the

disequilibrium since it determines, in part, the availability of protons for the plasma

H2CO 3 dehydration reaction. Finally, factors which affect the gill transit time, such as

cardiac output, must be considered because of their critical role in determining the time

available for CO2 reactions to reach equilibrium. While this study has established the

presence of an acid-base disequilibrium in the arterial blood of rainbow trout, further

work is required to clarify the basis, as well as the physiological significance, of the

imbalance.

5. Acknowledgements

This study was supported by NSERC of Canada operating and equipment grants

to SFP. KMG was the recipient of an E.B. Eastburn Postdoctoral Fellowship.

6. References

Bidani, A. and E.D. Crandall (1978a). Slow postcapillary pH changes in blood in anesthetized animals. J.

AppL PhysioL 45: 674-680.

Bidani, A. and E.D. Crandall (1978b). Slow postcapillary changes in blood pH in vivo: titration with

aeetazolamide. J. AppL PhysioL 45: 565-573.

Bidani, A. and E.D. Crandall (1988). Velocity of CO 2 exchanges in the lungs. Annu. Rev. Physiol. 50: 639-652.

Cameron, J.N. (1978). Chloride shift in fish blood. J. Exp. Zool. 206: 289-295.

Chakrabarti, M.K., S.M. Cobbe, L. Loh and P.A. Poole-Wilson (1983). Measurement of pulmonary venous

and arterial pH oscillations in dogs using catheter tip pH electrodes. J. Physiol. (London) 336: 61-71.

Crandall, E.D. and J.E. O'Brasky (1978). Direct evidence for participation of rat lung carbonic anhydrase

in CO2 reactions. J. Clin. Invest. 62: 618-622.

Crandall, E.D. and A. Bidani (1981). Effects of red blood cell HCO£/C1- exchange kinetics on lung CO2

transfer: theory. J. Appl. Physiol. 50: 265-271.

Crandall, E.D., S.J. Mathew, R.S. Fleischer, H.I. Winter and A. Bidani (1981). Effects of inhibition ofrbc

HCO3/C1- exchange on CO 2 excretion and downstream pH disequilibrium in isolated rat lungs. J. Clin.

Invest. 68: 853-862.

Effros, R.M., R.S.Y. Chang and P. Silverman (1978). Acceleration of plasma bicarbonate conversion to

carbon dioxide by pulmonary carbonic anhydrase. Science 199: 427-429.

Forster, R.E. and E.D. Crandall (1975). Time course of exchanges between red cells and extraeellular fluid

during CO 2 uptake. J. Appl. Physiol. 38: 710-718.

Henry, R.P., N.J. Smatresk and J.N. Cameron (1988). The distribution ofbranehial carbonic anhydrase and

the effects of gill and erythrocyte carbonic anhydrase inhibition in the channel catfish Ietalurus punctatus.

J. Exp. Biol. 134: 201-218.

Hill, E.P., G.G. Power and L.D. Longo (1973). Mathematical simulation of pulmonary O 2 and CO 2

exchange. Am. J. Physiol. 224: 904-917.

Hill, E.P., G.G. Power and R.D. Gilbert (1977). Rate ofpH changes in blood plasma in vitro and in vivo.

or. Appl. Physiol. 42: 928-934.

Kloeke, R.A. (1980). Equilibrium of CO2 reactions in the pulmonary capillary. J. Appl. Physiol. 48: 972-976.

Klocke, R.A. (1988). Velocity of-CO2 exchange in blood. Annu. Rev. PhysioL 50: 625-637.

Perry, S.F. (1986). Carbon dioxide excretion in fishes. Can. J. ZooL 64: 565-572.

Perry, S.F. and C.M. Wood (1989). Control and coordination of gas transfer in fishes. Can. J. Zool. 67:

2961-2970.

272 K.M. Gilmour et al. / Respiration Physiology 96 (1994) 259-272

Perry, S.F. and P. Laurent (1990). The role of carbonic anhydrase in carbon dioxide excretion, acid-base

balance and ionic regulation in aquatic gill breathers. In: Transport, Respiration and Excretion: Com-

parative and Environmental Aspects, edited by J.P. Truchot and B. Lahlou. Basel: Karger, pp. 39-57.

Perry, S.F. and K.M. Gilmour (1993). An evaluation of factors limiting carbon dioxide excretion by trout

red blood cells in vitro. J. Exp. BioL 180: 39-54.

Rispens, P., B. Oesburg, J.P. Zock and W.G. Zijlstra (1980). Intra-aortic decrease in plasma pH. Pflagers

Arch. 386: 97-99.

Soivio, A., K. Nyholm and K. Westman (1975). A technique for repeated blood sampling of the blood of

individual resting fish. J. Exp. Biol. 62: 207-217.

Thomas, S. and H. Le Ruz (1982). A continuous study of rapid changes in blood acid-base status of trout

during variations of water Pco2. J. Comp. PhysioL 148: 123-130.

Wolf, K. (1963). Physiological salines for freshwater teleosts. Progr. Fish. Cult. 25: 135-140.

Wood, C.M. and S.F. Perry (1991). A new in vitro assay for carbon dioxide excretion by trout red blood

cells: Effects of catecholamines. G. Exp. BioL 157: 349-366.