Plasma CO2 reactions in Pacific spiny dogfish (Squalus acanthias) have access to plasma and gill membrane-associated carbonic anhydrase (CA). Acute severe experimental anemia and selective CA inhibitors were used to investigate the role of extracellular CA in CO2 excretion. Anemia was induced by blood withdrawal coupled to volume replacement with saline. Lowering hematocrit from 14.2 ± 0.4% (mean ± SE; N = 31) to 5.2 ± 0.1% (N = 31) had no significant impact on arterial or venous CO2 tensions (PaCO2 and PvCO2, respectively) over the subsequent 2 h. Pco2 was maintained despite the reduction in red cell number and a significant 32% increase in cardiac output (V̇b), both of which have been found to cause PaCO2 increases in teleost fish. By contrast, treatment of anemic dogfish with the CA inhibitors benzolamide (1.3 mg/kg) or F3500 (50 mg/kg), to selectively inhibit extracellular CA, elicited rapid and significant increases in PaCO2 of 0.68 ± 0.17 Torr (N = 6) and 0.53 ± 0.11 Torr (N = 7), respectively, by 30 min after treatment. These findings provide a functional context in which extracellular CA in dogfish contributes substantially to CO2 excretion. Additionally, the apparent lack of effect of V̇b changes on Pco2 suggests that, in contrast to teleost fish, CO2 excretion in dogfish does not behave as a diffusion-limited system.
- carbonic anhydrase
- 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid
severe anemia in teleost fish not only lowers blood O2 carrying capacity, activating a suite of physiological responses designed to maintain O2 delivery to the tissues, but also impairs CO2 excretion, causing blood CO2 tensions to rise. Significant increases in arterial and venous blood CO2 tensions (PaCO2 and PvCO2, respectively) were evident in brown bullhead (Ameiurus nebulosus) by 2 h after lowering hematocrit to 5% (23) and have been documented in starry flounder (Platichthys stellatus) and rainbow trout (Oncorhynchus mykiss) at 24 h to 5 days of severe anemia (24, 59). The anemia-induced hypercapnia, which is generally accompanied by an acidosis, appears to stem from two factors. First, the red blood cell (RBC), which contains high levels of cytoplasmic carbonic anhydrase (CA), plays a key role in CO2 excretion because it is the sole site of CA-catalyzed HCO3− dehydration in the gills of teleost fish (Refs. 26 and 42, see also Refs. 21 and 33 for reviews). Because >90% of CO2 is transported as HCO3− in the plasma (41, 57), and the rate of the uncatalyzed HCO3− dehydration reaction (rate constant >20 s; see Ref. 18) is too slow to support effective CO2 removal during the gill transit time (1–3 s; see Ref. 12), a reduction in the number of RBCs impairs CO2 excretion by reducing the available dehydration sites. Measurements of RBC HCO3− dehydration rates in vitro have indicated that RBC CA availability becomes limiting in rainbow trout at hematocrits of ∼5% (43).
To compensate for the reduced blood O2-carrying capacity that is a consequence of low hematocrit, several species elevate cardiac output (11, 23, 60), and this increase in blood flow may also contribute to the anemia-induced hypercapnia. CO2 excretion in teleost fish behaves as a diffusion-limited system because of chemical equilibrium limitations on the entry of HCO3− in the RBC to be dehydrated at the catalyzed rate (Refs. 16, 36, and 38; see Refs. 44 and 51 for reviews). Gas transfer efficiency is sensitive to transit time at the gas exchange surface in a diffusion-limited system. Because transfer efficiency falls as transit time decreases, increases in cardiac output (hence reduced gill transit time) cause PaCO2 to rise. In anemic bullhead, approximately one-half of the increase in PaCO2 that occurred over the initial 2 h of anemia was attributed to elevated blood flow (23). By contrast, where CO2 excretion is not diffusion limited, increases in cardiac output would be predicted to benefit CO2 excretion by enhancing the amount of CO2 transferred across the gas exchange surface for a given partial pressure gradient.
Although teleost fish lack plasma-accessible CA activity in the gills, chondrichthyans such as dogfish possess both plasma CA and branchial membrane-bound CA activities (22, 25, 27, 58, 63). CA activity circulating in the plasma is probably derived from endogenous (likely naturally occurring) hemolysis (32), whereas the branchial membrane-bound CA appears to be analogous to the mammalian type IV isozyme (25). Unlike mammalian pulmonary capillary CA IV (3, 13, 33), however, the extracellular CA activity of dogfish has been found to contribute significantly to CO2 excretion. Dogfish in which extracellular CA was selectively inhibited exhibited a reduced arterial-venous total CO2 concentration difference (∼44% reduction 15 min after inhibitor injection), resulting in an elevation of PaCO2 (by 0.22 Torr, or 19%, at 15 min) and a concomitant acidosis (a pH decrease of 0.11 units at 15 min; see Ref. 25).
Two predictions arise from the hypothesis that extracellular CA activity in dogfish contributes significantly to CO2 excretion. First, CO2 excretion should be less sensitive to a reduction in RBC number in dogfish than in teleost fish because of the possibility for extracellular CA activity to compensate for the lack of RBC dehydration sites. Second, the apparent diffusion limitations on CO2 excretion that exist in teleost fish (16, 23) should be relieved in dogfish by the presence of extracellular CA activity that contributes to CO2 excretion, and hence CO2 excretion will be perfusion, rather than diffusion, limited (44). Thus severely anemic dogfish should be capable of maintaining blood CO2 tension, unlike severely anemic teleost fish (23). In the present study, this prediction was tested. Severe anemia was induced by blood withdrawal coupled to volume replacement with saline, and the effects of severe experimental anemia on blood gas and acid-base status as well as cardiorespiratory parameters were assessed before and after treatment with selective CA inhibitors. Anemia was selected as a biologically relevant challenge to CO2 excretion; anemia is of significant natural occurrence in wild fish, presumably as a result of disease, injury, predation, or parasitism (60).
MATERIALS AND METHODS
Pacific spiny dogfish (Squalus acanthias Linnaeus) were collected by net during trawls by local fishermen or angled by hook and line and transported to holding facilities at Bamfield Marine Station where they were held in a 75,000-liter circular tank. The tank was supplied with full-strength seawater at 13°C and was kept under natural photoperiod. Dogfish were fed two times weekly with herring and used within 4 wk of capture. Two groups of dogfish were used; for experiments involving the recording of blood gases and cardiorespiratory variables, fish (N = 31) weighing 625–2,400 g [average mass = 1,545 ± 103 (SE) g] were used. Smaller fish (average mass = 472 ± 41 g, N = 6) were used for respirometry experiments.
Surgery was carried out on fish anesthetized by immersion in an aerated solution of benzocaine (ethyl-p-aminobenzoate; 0.1 g/l) then transferred to an operating table where the gills were irrigated continuously with the same anesthetic solution. In dogfish used for respirometry experiments, the caudal artery and caudal vein were cannulated. The hemal arch was exposed using a lateral incision at the level of the caudal peduncle, and flexible polyethylene tubing (Clay-Adams PE-50) was inserted in the vessels in the anterior direction (1). For all other experiments, the caudal artery and caudal vein were cannulated, and in addition a bidirectional cannulation (Clay-Adams PE-50) of the celiac artery was carried out (see Ref. 45). The bidirectional celiac cannulation permitted establishment of an extracorporeal blood circulation (see below; note that for venous blood measurements, blood flow to the extracorporeal loop was from the caudal vein), whereas the caudal vessel cannulas were used for blood withdrawal, saline infusion, and drug injection. Silk sutures were used to close incisions and secure cannulas to the skin, and all cannulas were filled with heparinized (100 IU/ml ammonium heparin) saline (500 mmol/l NaCl). Dogfish were, in addition, instrumented for measurements of blood flow and ventilation. To measure cardiac output, a 3S or 4S ultrasonic flow probe (Transonic Systems, Ithaca, NY) was placed around the conus arteriosus, which was exposed by means of a small (1.5-cm) midline ventral incision into the pericardial cavity. Lubricating jelly (K-Y Personal Lubricant; Johnson and Johnson) was used as an acoustic couplant. Again, silk sutures were used to close the incision and secure the flow probe lead to the skin. Ventilation was assessed by inserting catheters (Clay Adams PE-160) in the spiracular cavities and securing them to the head with silk sutures. After surgery, fish were revived and transferred to individual holding boxes of opaque acrylic or wood provided with flowing, aerated seawater for a 24-h recovery period before experimentation.
The objective of the experiments was to assess cardiorespiratory responses during the initial 2-h period after the induction of severe anemia (final hematocrit ∼5%) and to evaluate the contribution of extracellular CA to the regulation of PaCO2 under these conditions. After an initial 10-min period during which baseline conditions for arterial or venous blood gas and acid-base status, ventilation, and cardiovascular parameters were recorded, anemia was induced by blood withdrawal coupled with volume replacement with dogfish saline (in mmol/l: 250 NaCl, 7 Na2SO4, 3 MgSO4, 4 KCl, 2 CaCl2·2H2O, 5 NaHCO3, 0.1 Na2HPO4, 4 glucose, 325 urea, and 100 trimethylamine oxide, pH 7.8) containing 1% BSA. The initial 1 ml of blood withdrawn was used to determine the control hematocrit, O2 content (CaO2 or CvO2), and CO2 content (CaCO2 or CvCO2); 1 ml of blood was also withdrawn simultaneously for measurement of these parameters in venous blood if arterial blood was monitored via the extracorporeal loop, and vice versa. In some fish, arterial blood samples were also used to assess plasma lactate concentrations. Approximately 30% of the blood volume was removed and replaced with an equal volume of saline, and hematocrit was then remeasured. Up to three rounds of blood withdrawal and saline replacement over the course of 50 min (on average) were required to lower hematocrit to the desired value of 5%. Blood withdrawal averaged 17.2 ± 0.4 (N = 30), 16.3 ± 0.7 (N = 29), and 10.1 ± 0.7 (N = 20) ml/kg for the three rounds, with the decline in N reflecting the fact that multiple withdrawals were not required to make all fish anemic. Using a total blood volume of 5% body mass, these values translated to 34.4 ± 0.8% (N = 30), 32.6 ± 1.4% (N = 29), and 20.2 ± 1.3% (N = 20) of blood volume, respectively. A temporary hypotension likely accompanied blood withdrawal but would have been rapidly corrected by the subsequent saline infusion.
Once the desired hematocrit of 5% was achieved, cardiorespiratory parameters were monitored for 2 h, after which arterial and venous blood samples (1 ml each) were withdrawn for assessment of hematocrit, CaO2, CvO2, CaCO2, and CvCO2 and in some fish, plasma lactate levels. One of the following treatments was then administered, and the fish was monitored for 30 min: 1.3 mg/kg benzolamide (N = 7 for arterial blood, N = 6 for venous blood); 1 × 10−4 mol/l DIDS followed 40 min later by 1.3 mg/kg benzolamide (N = 6); 50 mg/kg polyoxyethylene-aminobenzolamide (F3500; N = 7); or 30 mg/kg acetazolamide or aminobenzolamide (N = 6). The objective of benzolamide and F3500 treatments was to selectively inhibit extracellular CA activity. Benzolamide is a relatively impermeant CA inhibitor (39) that, at low doses, enters the RBC only slowly and hence selectively inhibits extracellular CA activity (22, 25). F3500 consists of the CA inhibitor aminobenzolamide irreversibly linked to a nontoxic polymer and is restricted to the extracellular compartment by virtue of its high molecular weight (14, 40). Because the selectivity of benzolamide relies on the rate at which it permeates the RBC, trials using F3500 were also carried out. The effects of these treatments were compared with those of inhibiting RBC and extracellular CA activities using acetazolamide or aminobenzolamide, both of which are potent CA inhibitors that rapidly equilibrate across the RBC membrane. The effects of benzolamide were also examined under conditions of reduced HCO3− access to the RBC, achieved by treatment of the fish with the RBC anion exchange inhibitor DIDS (22, 25).
Drugs were administered as a bolus injection in the caudal artery or vein. Benzolamide and acetazolamide were prepared by dissolving the inhibitor in saline (500 mmol/l NaCl) with added NaOH and then slowly titrating the pH down to a level as close as possible to physiological. Aminobenzolamide and F3500 were synthesized according to the method of Conroy et al. (14) and tested for inhibitory activity against trout and dogfish RBC CA using either the electrometric ΔpH assay (31) or a radioisotopic [14C]HCO3− dehydration assay (62). Doses of these compounds for use in vivo in dogfish were selected on the basis of the results of Swenson et al. (54). DIDS was dissolved in dogfish saline containing 20% dimethyl sulfoxide (DMSO) and injected in the dogfish to achieve a nominal final concentration in the blood of 1 × 10−4 mol/l (0.05% DMSO), assuming even distribution throughout the extracellular fluid. Although the effect of 0.05% DMSO alone was not examined, previous studies have found 0.1% DMSO to be without effect on trout RBC CO2 excretion (43).
Flowthrough respirometry was used to assess oxygen consumption (Ṁo2) in a separate group of dogfish (N = 6) under control conditions and after 2 h of severe anemia (hematocrit ∼5%). Anemia was induced as described above. Respirometers (4.8-liter volume) were constructed from 100-mm-diameter PVC pipe. Ṁo2 was calculated as the difference in water Po2 between inflowing and outflowing water from the respirometer, taking into account the solubility coefficient of O2 in seawater (6), water flow rate, and mass of the fish.
An extracorporeal blood circulation (55) was used to continuously monitor blood respiratory variables. Blood was withdrawn at a rate of 0.55 ml/min from the arterial or venous vessel using a peristaltic pump and passed through an external circuit (of ∼1 ml volume) containing Po2, Pco2, and pH electrodes. To prevent clotting, the circuit was rinsed with heparinized (540 IU/ml) saline for 10–15 min before initiating blood flow. Blood pH, Pco2, and Po2 were measured using Metrohm (model 6.0204.100; pH) and Cameron Instruments (CO2, O2) electrodes housed in thermostatted cuvettes and connected to a blood gas analyzer (BGM 200; Cameron Instruments). Before each experiment, the pH electrode was calibrated by pumping precision buffer solutions through the circuit until stable readings were recorded. A similar procedure was used to calibrate the O2 and CO2 electrodes with a zero solution (2 g/l sodium sulfite; O2 electrode only) and/or water equilibrated with appropriate gas mixtures (supplied by a GF-3/MP gas mixing flowmeter; Cameron Instruments).
The pressure changes associated with breathing were assessed by connecting the spiracular catheter to a pressure transducer (Bell and Howell) linked to an amplifier (Harvard Biopac DA 100). The pressure transducer was calibrated daily against a static column of water. Blood flow was measured by attaching the factory-calibrated ultrasonic flow probe to a blood flowmeter (model T106; Transonic Systems).
A data acquisition system (Biopac Systems) with Acknowledge data acquisition software (sampling rate set at 10 Hz) and a Pentium personal computer were used to convert all analog signals (blood gases, ventilation pressures, and blood flow) to digital data. This system allowed continuous data recordings to be obtained for blood Pco2, Po2, and pH, mass-specific blood flow (V̇b), cardiac frequency (automatic rate calculation from the pulsatile V̇b trace), ventilation frequency (Vf; automatic rate calculation from the raw ventilation pressure trace), and ventilation amplitude (Vamp; the difference between maximum and minimum ventilation pressures). Cardiac stroke volume was calculated by dividing mass-specific blood flow by cardiac frequency.
Hematocrit was measured in triplicate using microcapillary tubes centrifuged at 6,000 g for 6 min. CaO2 and CvO2 were measured in triplicate on 50- or 100-μl samples with an Oxycon (Cameron Instruments). CaCO2 and CvCO2 were measured in duplicate on 50-μl samples with a total CO2 analyzer (Corning model 965). Plasma lactate concentrations were measured on deproteinized plasma samples using the NAD+/lactate dehydrogenase method outlined by Bergmeyer (2).
For respirometry experiments, water Po2 was measured with an O2 electrode (Cameron Instruments) and an O2/CO2 analyzer (Cameron Instruments) connected to a chart recorder (Goerz Metrawatt). A peristaltic pump was used to pass water at a flow rate of 4.15 ml/min from the inflowing or outflowing water to the O2 electrode, which was calibrated by pumping zero solution or air-equilibrated water across the electrode until stable readings were achieved. Water flow to the respirometer was adjusted to obtain a Po2 difference of ∼20 Torr between the inflowing and outflowing samples under control conditions.
Data are presented as means ± 1 SE. Mean blood gas, acid-base, ventilatory, and cardiovascular data were compiled for 2-min periods before and every 10 min after the induction of severe anemia, and every 2 min after drug treatments. Because of interindividual variation, data for individual fish were normalized by subtracting from each data point the value at the time of initiating recording after the induction of anemia (time 0) or drug administration (time = 150 or 190 min). To statistically analyze the effects of anemia, data for fish from all drug treatment groups were combined. Anemia effects within individual drug treatment groups were also examined and were in general similar to those of the combined data set; these data are presented in Figs. 4–6 and Table 3 for reference but are not otherwise discussed.
Baseline (pre-anemia) values were compared with values at time 0 using paired Student's t-tests. The effect of time on blood respiratory, cardiovascular, and ventilatory variables during the 2 h of anemia or after drug administration was analyzed statistically using a one-way repeated-measures ANOVA followed by the Bonferroni post hoc multiple-comparisons test, as appropriate. Where assumptions of normality or equal variance were violated, equivalent nonparametric analyses were utilized. The significance of differences in lactate concentration, Ṁo2, and percentage extraction of O2 from the blood [EavO2 = (CaO2 − CvO2)/CaO2 × 100%] between control and anemic conditions was assessed using paired Student's t-tests, whereas a two-way repeated-measures ANOVA was used to analyze the effects of sampling site (arterial or venous blood) and treatment (control or anemic) on blood O2 and CO2 contents. The fiducial limit of significance in all statistical analyses was 5%.
Blood withdrawal coupled to volume replacement with saline was effective in reducing hematocrit by 63%, from the control value of 14.2 ± 0.4% (N = 31) to an anemic value of 5.2 ± 0.1% (31). Despite the reductions in RBC number, and arterial and venous blood O2 contents, O2 uptake (measured by respirometry), and percentage extraction of O2 from the blood were maintained (Table 1). Plasma lactate was somewhat higher than levels typical of resting dogfish (∼1 μmol/l; see Ref. 47) but did not increase significantly with anemia (Table 1). Blood respiratory, ventilatory, and cardiovascular variables were monitored before blood removal and for 2 h from the point of reaching the final hematocrit. The process of removing blood caused arterial pH and Po2 to fall significantly, although venous values were not affected (Fig. 1 and Table 2). Over the initial 2 h of anemia, blood pH and Po2 then remained relatively constant, with PaO2 and PvO2 both declining slightly but significantly at the end of the monitoring period (Fig. 1). By contrast, neither arterial nor venous Pco2 was altered significantly during either the process of inducing anemia or the initial 2 h of severe anemia (Fig. 1A and Table 2), and arterial blood CO2 content was unaffected by anemia, whereas venous blood CO2 content decreased significantly (Table 1).
Cardiorespiratory variables were also affected by blood withdrawal. Ventilation amplitude decreased significantly during the lowering of hematocrit, with the new level being maintained over the initial 2 h of anemia (Fig. 2B and Table 2), whereas cardiac output was elevated by 32% during blood removal because of an increase in stroke volume (Fig. 3 and Table 2). Cardiac output then fell over the initial period of severe anemia, such that by 2 h it was elevated by only 9% over the baseline value (Fig. 3A).
Although blood Pco2 was unaffected by anemia, selective inhibition of extracellular CA activity using either low doses of benzolamide or the impermeant inhibitor F3500 resulted in rapid and significant increases in blood CO2 tension, with corresponding reductions in pH (Figs. 4 and 5 and Table 3). With benzolamide treatment, changes in venous Pco2 and pH appeared to lag behind those in the arterial blood, although there were no significant differences between arterial and venous blood in the overall magnitude of the changes by 30 min after benzolamide injection (Fig. 4, A and B). The elevation of blood Pco2 occurred despite significant increases in Vf with benzolamide treatment (Fig. 4C) and Vamp with either inhibitor (Figs. 4C and 5C). Blood flow was unaffected by benzolamide treatment (data not shown) and was not measured in F3500-treated fish because of equipment failure.
The effect of benzolamide on PaCO2 was enhanced by pretreatment of the dogfish with the anion exchange inhibitor DIDS (Fig. 6A and Table 3). The objective of this treatment was to further reduce RBC contributions to CO2 excretion; as in previous studies on dogfish of normal hematocrit, DIDS treatment was found to reduce HCO3− flux through the RBC by ∼40% (22). In anemic dogfish, DIDS treatment increased PaCO2 by 32% after 40 min, but this increase was dwarfed by the 152% increase in PaCO2 induced by subsequently administering benzolamide (Figs. 6A and 7). Indeed, the magnitude of the PaCO2 increase with DIDS plus benzolamide treatment was not significantly different from that observed after inhibition of both extracellular and RBC CA activities using acetazolamide or aminobenzolamide (Fig. 7 and Table 3). Corresponding to the increases in PaCO2, arterial pH decreased significantly after both DIDS and benzolamide injection (Fig. 6B). Ventilation was unaffected by either drug (data not shown). A significant, transient decrease in heart rate occurred after DIDS administration (Fig. 6C). This drop in heart rate tended to lower cardiac output, although not significantly. Over the 40-min monitoring period, heart rate and cardiac output returned to and exceeded, respectively, their initial levels; cardiac output then decreased toward the pre-DIDS value after benzolamide administration (Fig. 6C).
The key finding of the present study was that severely anemic dogfish were able to maintain normal blood CO2 tensions (Fig. 1), despite both the reduction in RBC number and a significant increase in cardiac output. This result is in sharp contrast to the situation in the teleost species that have been examined to date. In flounder, bullhead, and trout, decreasing hematocrit to 5% from starting values of 14, 23, and 20–25%, respectively, caused significant and rapid (2–24 h) increases in arterial and venous Pco2 (23, 59) that were sustained for several days (24). In particular, it is worth noting that the percentage reduction of hematocrit in flounder (59) was very similar to that of the dogfish of the present study. The elevation of blood CO2 tension in anemic teleost fish was attributed both to the loss of RBCs as sites for catalyzed HCO3− dehydration and to the impact of increased cardiac output on CO2 transfer efficiency in a system in which CO2 excretion behaves in a diffusion-limited fashion (23). The ability of dogfish to hold Pco2 constant during severe anemia did not stem from a reduction in metabolic rate, since O2 uptake was maintained (Table 1). Rather, extracellular CA appears to play a critical role because selective inhibition of extracellular CA activity using benzolamide or F3500 resulted in a rapid, significant elevation of blood CO2 tension (Figs. 4 and 5). Moreover, the use of DIDS treatment to reduce HCO3− dehydration via the RBCs to an even greater extent than that of anemia alone had only a small impact on PaCO2 relative to the subsequent effect of benzolamide administration (Figs. 6 and 7), again implicating extracellular CA as the key factor in maintaining CO2 excretion. The Pco2 rise with benzolamide or F3500 treatment was 20–33% of that elicited by inhibition of both RBC and extracellular CA activities using acetazolamide or aminobenzolamide, or by inhibition of extracellular CA with benzolamide, while blocking access to RBC CA through the use of DIDS (Fig. 7). These results suggest that ∼25% of total CO2 excretion in severely anemic dogfish was accomplished through extracellular CA. Similarly, extracellular CA was found to contribute 30–60% of total CO2 excretion in dogfish of normal hematocrit (25).
Extracellular CA in dogfish consists of gill membrane-bound CA that appears to be analogous to the mammalian type IV isozyme and CA activity that is present in the plasma (22, 25, 27, 58, 63). At present, the relative contributions of these two CA sources to HCO3− dehydration in the plasma during transit through the gills cannot be distinguished. Given that dilution of plasma CA by saline infusion probably occurred in the present study, apparently without effect, it seems likely that the physiological significance of plasma CA is outweighed by that of the branchial membrane-associated activity; this hypothesis remains to be tested. Unlike the plasma of many vertebrates (17, 20, 28, 32, 34, 46, 48, 49, 56), dogfish plasma does not contain an endogenous CA inhibitor (32), and its absence is likely a factor in the ability of extracellular CA to contribute to CO2 excretion. This would be particularly true for plasma CA activity, assuming that it is indeed derived from endogenous hemolysis (e.g., see Ref. 32), since cytoplasmic CA appears to be more susceptible to inhibition by plasma CA inhibitors than does membrane-associated CA activity (19, 30, 49, 56). Also, in dogfish and other cartilaginous fish, the plasma nonbicarbonate buffer capacity accounts for a relatively high proportion of whole blood buffering (27), and this factor may be important in supplying protons for HCO3− dehydration in the plasma (25, 27). In mammals, the contribution of pulmonary capillary CA IV to CO2 excretion under normal conditions is thought to be restricted to <10% because of limitations on proton availability in plasma (Refs. 5 and 29, see also Ref. 33 for a review).
Although at normal hematocrit CO2 excretion via pulmonary capillary CA IV in mammals is probably quantitatively unimportant (13, 52, 53), a reduction in hematocrit could enhance the role played by this extracellular CA activity. The mathematical model of Bidani and Crandall (4) suggests that, at 15% hematocrit (severe anemia), CO2 elimination would be increased by ∼30% by the presence of pulmonary capillary CA activity sufficient to increase the rate of the HCO3− dehydration reaction 100-fold over the uncatalyzed rate, whereas only a 12% increase would be achieved at 35% (normal) hematocrit. Perhaps surprisingly, however, the potential contribution of pulmonary capillary CA IV to CO2 excretion under conditions of reduced hematocrit has received little attention.
As in dogfish, blood Pco2 does not change in mammals subjected to isovolemic anemia (see Ref. 15 for discussion). The maintenance of normal blood Pco2 values has been attributed primarily to increases in cardiac output and an enhancement of the Haldane effect resulting from increased O2 extraction, with hyperventilation and improved efficiency of tissue and/or pulmonary gas exchange as additional factors that could potentially compensate for the reduction in RBC number (15). Whether gas transfer efficiency at the tissues or gas exchange surface is enhanced by anemia remains an open question in dogfish, as in mammals (15). Dogfish do not appear to utilize the strategy of increasing ventilation to compensate for reduced RBC number during severe anemia (Fig. 2). Indeed, ventilation amplitude was lower than the baseline value after blood withdrawal and saline replacement, and this reduction may have contributed to the parallel fall in arterial Po2 (Fig. 1). The appearance and persistence of ventilatory responses to severe anemia are variable among fish, with flounder and rainbow trout exhibiting at least a transient hyperventilation (11, 24, 50, 60), whereas dogfish (Fig. 2) and bullhead (23) do not hyperventilate.
In rabbits subjected to hemodilution, lowering hematocrit by 66% (from 36 to 12%) caused cardiac output to increase by 50% (15). Similar percentage reductions in hematocrit elicited a 32% increase in cardiac output in the dogfish of the present study (Fig. 3), and a 57% rise in cardiac output in bullhead (23); significant elevations of cardiac output have also been measured in severely anemic flounder and rainbow trout (11, 60). Increases in cardiac output will benefit CO2 excretion by permitting greater CO2 transfer across the gas exchange surface for the same partial pressure gradient (the driving force for CO2 diffusion), provided that CO2 excretion is not diffusion limited. In a diffusion-limited system, gas transfer efficiency is sensitive to changes in blood transit time at the gas exchange surface; therefore, increases in cardiac output, if translated into reductions in transit time at the gas exchange surface, cause PaCO2 to rise. This situation exists in the teleost fish that have been examined to date (16, 23, 36). The apparent diffusion limitations appear to arise from chemical equilibrium limitations on HCO3− entry in the RBC via the relatively slow anion exchanger, since they are relieved by addition of bovine CA to the circulation to provide an extracellular site for HCO3− dehydration (Refs. 16, 36, and 61, reviewed in Ref. 44). Thus, in teleost fish, the increase in cardiac output evoked by severe anemia benefits O2 delivery, but at the expense of increased blood CO2 tension and a concomitant acidosis (23). The lack of increase in PaCO2 in severely anemic dogfish, despite a significant increase in cardiac output, implies that CO2 excretion does not behave as a diffusion-limited system in this fish, presumably because the presence of extracellular CA circumvents chemical equilibrium limitations on HCO3− dehydration. In this case, the increase in blood CO2 tension after benzolamide or F3500 treatment in anemic dogfish could arise from the imposition of diffusion limitations on CO2 excretion as well as the loss of (extracellular) sites of catalyzed HCO3− dehydration. An alternative possibility is that CO2 excretion in dogfish is effectively diffusion limited but that a fall in gill transit time in anemia is avoided by lamellar recruitment. To distinguish between these possibilities, the experimental manipulation of cardiac output in dogfish of normal hematocrit should be undertaken in the presence/absence of selective CA inhibitors. A rise in PaCO2 with increasing cardiac output in dogfish subjected to selective inhibition of extracellular CA activity would support the hypothesis that apparent diffusion limitations on CO2 excretion do not exist because of the presence of extracellular CA.
Increased cardiac output in severely anemic rabbits was not sufficient by itself to maintain normal O2 uptake; hence, an increase in percentage utilization of O2 from the blood (i.e., a widening of the arteriovenous O2 content difference) occurred (15). The increased extraction of O2 from the blood lowers venous Hb saturation and is thought to benefit CO2 excretion by augmenting the role played by oxylabile Bohr protons (Haldane effect) in RBC HCO3− dehydration (15). In at least some teleost fish, elevated O2 extraction from the blood is also observed during severe anemia (60). However, this strategy is unlikely to be effective in enhancing oxylabile proton availability for HCO3− dehydration during anemia because many teleost fish possess nonlinear Haldane effects in which the majority of Bohr protons are released between 50 and 100% Hb O2 saturation (7, 10, 35, 37); hence, the contribution of the Haldane effect to CO2 excretion is maximal under conditions of normal O2 extraction (8, 9, 23). In dogfish, O2 extraction from the blood was not increased by severe experimental anemia (Table 1), nor would such a strategy benefit CO2 excretion, since the Haldane effect appears to be absent from dogfish blood (63).
Thus the chief compensatory response to severe anemia in dogfish appears to be an elevation of cardiac output. This response permits O2 uptake and delivery to the tissues to be maintained, as evidenced by the lack of significant changes in Ṁo2, O2 extraction from the blood, and plasma lactate concentration (Table 1). Given the presence of extracellular CA, increased cardiac output is also sufficient to maintain CO2 excretion during anemia in dogfish. Extracellular CA activity contributes to CO2 excretion by providing a plasma site for HCO3− dehydration at the catalyzed rate and by preventing the occurrence of chemical equilibrium limitations on HCO3− dehydration that might otherwise impose apparent diffusion limitations on CO2 excretion. Consequently, dogfish are able to withstand severe experimental anemia without a rise in blood Pco2. By contrast, because of the absence of extracellular CA, apparent diffusion limitations on CO2 excretion, and the possession of nonlinear Haldane effects, it appears that the only option available to severely anemic teleost fish to compensate for the loss of RBCs for CO2 excretion is to increase Pco2 gradients (through increased blood CO2 tensions) so as to enhance the HCO3− flux through the remaining RBCs.
This study was funded by Natural Sciences and Engineering Research Council (NSERC) of Canada research and equipment grants to K. M. Gilmour and S. F. Perry.
We are grateful to the director and staff of Bamfield Marine Station, and particularly to Nathan Webb (Research coordinator), for their help. Thanks are also extended to Dr. E. Swenson for the generous gift of benzolamide, Dr. J. Jaquith for the synthesis of F3500, and Dr. J. Richards for assaying plasma samples for lactate.
The animals used in this study were cared for in accordance with the principles of the Canadian Council for Animal Care, Guide to the Care and Use of Experimental Animals. Experimental protocols were approved by institutional animal care committees.
Present addresses: K. K. Gilmour, Department of Biology, Carleton University, Ottawa, ON, Canada K1S 5B6; S. F. Perry, Department of Biology, University of Ottawa, Ottawa, ON, Canada K1N 6N5.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society