Na⁺ and Cl⁻ Uptake Kinetics, Diffusive Effluxes and Acidic Equivalent Fluxes Across the Gills of Rainbow Trout II. Responses to Bicarbonate Infusion

Metabolic alkalosis during recovery from hyperoxia was associated with a greatly increased Cl⁻ influx, a reduced Na⁺ influx and a large uptake of acidic equivalents across the gills, indicative of dynamic modulation of Na⁺/acidic equivalent and CP/basic equivalent exchanges (Goss and Wood, 1990). Alterations in Na⁺ and Cl⁻ influx were achieved by changes in both the K_m (inverse of affinty) and J_max (maximal transport rate) of the respective transporters. However, this period is also characterized by marked changes in the perfusion and ventilation characteristics of the gills (Dejours, 1972, 1973; Wood and Jackson, 1980), raising the possibility that these ‘non-specific’ effects either contributed to, or detracted from, the observed responses. The primary goal of the present study was, therefore, to examine the same phenomena during a comparable metabolic alkalosis, but without the possible complicating effects of perfusion and ventilation changes.

Infusion of base in the form of NaHCO₃ results in an exogenously produced metabolic alkalosis which is, at least qualitatively, similar to that found during recovery from hyperoxia (Claiborne and Heisler, 1986; Heisler et al. 1988; McDonald and Prior, 1988; cf. Hõbe et al. 1984). However, all previous studies on this topic have used a bolus infusion method to induce the alkalosis. The present study, instead, has employed chronic infusion to create a relatively steady state of metabolic alkalosis within the fish, a balance between HCO₃⁻loading and excretion. The treatment was designed to produce an alkalosis of similar magnitude to that occurring during recovery from hyperoxia while avoiding the time course effects which.complicated the hyperoxia study.

Differential diffusive efflux of Na⁺ and Cl⁻ across the gills is also thought to play a significant role in acid-base correction in some circumstances (McDonald and Prior, 1988; McDonald et al. 1989). Some evidence for the involvement of this mechanism during post-hyperoxia alkalosis was obtained by Goss and Wood (1990) using a new method (efflux to NaCl-free water). Therefore, a second objective of the present study was to employ this technique to examine the contribution of differential diffusive efflux in detail at various times during the infusion period.

Blood acid-base status was examined during NaHCO₃ infusion to ensure that this treatment did, in fact, induce a similar alkalosis to that found during recovery from hyperoxia. The renal contribution to acid-base regulation during chronic infusion of NaHCO₃ was also assessed. Control infusions of neutral NaCl at the same concentration and rate were performed in all experiments to detect possible non-specific responses due to volume or salt loading.

Rainbow trout (Oncorhynchus mykiss (Walbaum, 1792); 250–400 g) were obtained and acclimated to 15±0.5°C in Hamilton tapwater as described in Goss and Wood (1990). To allow for repetitive blood sampling, infusion and urine collection, trout were anaesthetized (MS-222 1:10000; Sigma) and fitted with a dorsal aortic cannula filled with heparinized Cortland saline (Wolf, 1963; 50i.u.ml⁻¹ sodium heparin; Sigma), according to the method of Soivio et al. (1972), and a urinary catheter, according to the method of Wood and Randall (1973). The fish were then allowed to recover for 72 h before experimentation in the boxes described by Goss and Wood (1990).

Acclimation and experimental tapwater and the artificial NaCl-free water used in series II and III had the same composition as described in Goss and Wood (1990).

The term outflux (unidirectional outflux) refers to outflux in series I and III obtained by radioisotopic measurement of influx and net flux The term efflux (diffusive efflux) refers to effluxes measured in series II by a method that did not require radioisotopes.

Series I was designed to characterize the net and unidirectional ion and acid-base fluxes across the gills and kidney, and plasma ion and acid-base status associated with chronic infusion of either 140 mmol l^-1 NaCl (control, N=5) or 140 mmol 1⁻¹ NaHCO₃ (experimental, N=6). This concentration (140 mmol 1⁻¹) was chosen so as to minimize any changes in the osmolarity of the plasma. A further objective was to determine 4 h periods during which these fluxes might be relatively stable, thereby allowing the planned kinetic uptake measurements of series III.

Repetitive blood samples (300 μl) were withdrawn via the cannula at the following times: 0 (initial), 10, 15 and 19h after the start of infusion. Blood samples were analyzed for arterial pH (pHa) and plasma total [CO₂], [Na⁺] and [Cl⁻]. After the initial blood sample, the fish were then infused via the arterial catheter with either 140 mmol I⁻¹ NaCl or 140 mmol 1⁻¹ NaHCO₃ at a rate of 2.92±0.11mlkg⁻¹h⁻¹ (N=22) using a Gilson Minipuls peristaltic pump. The bicarbonate loading rate was therefore 410μequivkg⁻¹ h⁻¹. The goal was to induce an alkalosis of similar proportion to that found in recovery from hyperoxia by Hõbe et al. (1984), while minimizing the volume load. This rate of infusion and HCO₃⁻concentration were chosen on the basis of an initial trial experiment at a range of rates (not shown). After 10 h of infusion, the boxes were closed and both ²⁴Na⁺ (4.0μCi) and ³⁶Cl⁻ (1.0μCi) were added. An initial period of 10min was allowed for complete mixing. Thereafter, water samples (40 ml) were taken at 0.5 h intervals for 6h and analyzed for [Na⁺]_e, [Cl⁻]_e, total ammonia (Amm), titratable alkalinity (TAlk), and ²⁴Na⁺ and ³⁶Cl⁻ctsmin⁻¹. The boxes were flushed after 3 h of closure to ensure that the ambient ammonia (Amm) levels did not exceed 200μequivl⁻¹. Following the flushing, the appropriate amount of isotope was re-added to the box.

Series II was designed to measure the simple diffusive efflux of Na⁺ and CD from rainbow trout chronically infused with either 140 mmol I⁻¹ NaCl (N=11) or 140mmoll⁻¹ NaHCO₃ (N=12). This was performed concurrently with each of series I and III and therefore involved the same fish. The results from the two sets were not significantly different from each other and were therefore combined as series II. Control diffusive efflux measurements were taken immediately before control blood sampling; separate experimental diffusive efflux measurements were taken immediately prior to the 10 h and 19 h blood samples in each infusion group. At each measurement time the boxes were flushed with NaCl-free water of normal pH, Ca²⁺ and Mg²⁴⁺ content and TAlk, and diffusive effluxes were measured by monitoring the appearance of Na⁺ and Cl⁻ in the external water over 0-10 min, as described by Goss and Wood (1990).

Urine samples were collected in this series from an 8h pre-sampling control period, the first 10 h of infusion, 10–15 h of infusion and 15–19 h of infusion. These samples were analyzed for urine flow rate (UFR), total Amm and titratable acidity minus bicarbonate [TA-HCO₃⁻]. In a separate, identical series of experiments on different fish, urinary [Na⁺] and [Cl⁻] and UFR were measured to determine the ion excretion rates from the kidney during NaCl infusion (N=5) and NaHCO₃ infusion (N=5).

Series III was designed to measure the uptake kinetics of Na⁺ and Cl⁻ in fish chronically infused with either 140 mmol I⁻¹ NaCl (N=5) or 140 mmol I⁻¹ NaHCO₃ (N=6). The protocol for sampling of blood and infusion of salts was performed as outlined in series I. After 10 h of infusion, the flow of tapwater to the box was stopped and uptake kinetics determined over the following 4h. Methods were identical to those described in Goss and Wood (1990), except that Na⁺ and Cl⁻ kinetics were determined simultaneously in the same fish through the use of ²⁴Na⁺ and ³⁶Cl⁻. A common stock solution containing approximately 4.0μ Ci ²⁴Na⁺ and 1.0 μ Ci³⁶Cl⁻ was used for all additions to ensure the maintenance of a relatively constant specific activity over the entire range of [NaCl]_e.

Blood samples were drawn from the dorsal aortic cannula. Arterial pHa and plasma total CO₂ were immediately determined by standard Radiometer techniques (Wood and Jackson, 1980). The remainder of the sample was centrifuged and the plasma frozen for later ionic analysis. and plasma HCO₃⁻ levels were calculated by the rearrangement of the Henderson-Hasselbalch equation using the values of pK′ and αCO₂ tabulated in Boutilier et al. (1984).

Calculations of ion and acidic equivalent fluxes and transformation of the data by Michaelis-Menten kinetics and Eadie-Hofstee regression analysis were performed as outlined in Goss and Wood (1990), as were the statistical analyses. A significance level of P<0.05 was employed throughout.

Plasma [Na⁺] did not change in the NaHCO₃-infused group but in NaCl-infused fish there was a significant increase after 15 h of infusion (Fig. 1A). Plasma [Cl⁻] remained unchanged in the NaCl-infused group, despite the Cl⁻ load, but NaHCO₃ infusion resulted in a large loss of Cl⁻ from the plasma (133 to 108 mmol I⁻¹, Fig. 1B). As in fish recovering from exposure to hyperoxia, a pure metabolic alkalosis was present in fish infused with NaHCO₃, characterized by a large increase in arterial plasma [HCO₃⁻] (from 9.2 to 19.3 mmol 1⁻¹, Fig. 1C) with no elevation in (Fig. ID). As a result, pHa was greatly elevated from 7.89 to 8.25 (Fig. 1E). NaCl infusion did not cause any change in pHa, or [HCO₃⁻].

With a constant infusion rate of 2.92 ml kg⁻¹ h⁻¹, urine flow rate (UFR) increased by 86% (2.85±0.17 to 5.12±0.60mlkg⁻¹h⁻¹) and 80% (2.60±0.19 to 4.70±0.82ml kg⁻¹ h⁻¹) in the NaCl- and NaHCO₃-infused groups, respectively, at 10-16 h. These amounted to about 80% of the infusion rate. Volume loading of the fish was therefore relatively small, minimizing any complications that might be caused by changes in the cardiovascular system. Net basic equivalent excretion (=net acidic equivalent uptake) via the kidney increased dramatically in NaHCO₃- infused fish from a control value of 2 μequiv kg⁻¹ h⁻¹ to 85 μequiv kg⁻¹ h⁻¹, solely as a result of a large increase in the [TA-HCO₃⁻] component of renal acid-base flux (Fig. 2B). Net acidic equivalent uptake did not change in the NaCl-infused fish (Fig. 2A). In a separate group of fish, urinary Na⁺ and Cl⁻ excretion rate (data not shown) during NaCl infusion rose from control values of 12.5±1.4 and 10.2±1.5μequivkg⁻¹h⁻¹, respectively, to 49.9±9.9 and 58±11.8μequivkg⁻¹ h⁻¹ over the 19 h infusion period, whereas during NaHCO₃ infusion, urinary Na⁺ and Cl⁻ excretion rates rose from 16.9±3.0 and 15.2±5.3 μequiv kg⁻¹ h⁻¹ to 32.0±7.3 and 27.4±3.1 μequivkg⁻¹ h⁻¹, respectively, over the same time period. However, in each treatment, there was no differential loss of one ion over the other.

Infusion of NaCl or NaHCO₃ resulted in a highly negative (Fig. 3) of about −200μequivkg⁻¹ h⁻¹ in both cases, but there were large differences in the magnitude of the influx and efflux components of net ion balance. In particular, was reduced by about 50% from 400 to 200μequiv kg⁻¹ h⁻¹ in the NaHCO₃-infused group relative to the NaCl-infused group. was more variable in both groups, but overall was lower by about the same amount as a result of NaHCO₃ infusion. Responses of Cl⁻ fluxes were very different. In the NaCl-infused group, was again highly negative (−200μequivkg⁻¹h⁻¹; Fig. 4A) and approximately equal to (Fig. 3A). However, infusion of NaHCO₃ resulted in a positive of 150μequivkg⁻¹ h⁻¹, caused by a greatly stimulated (Fig. 4B), which was approximately doubled from 200 to 400μequivkg⁻¹ h⁻¹ compared to the NaCl-infused group. The differences in and were significant at almost every individual flux period and significant overall compared with the NaCl-infused group. tended to be smaller at most times in the NaHCO₃-infused fish, but none of these differences was significant.

There were profound differences in acidic equivalent fluxes between the two groups. While the NaCl-infused control group were in approximate H⁺ balance (Fig. 5A), the NaHCO₃-infused group exhibited a highly positive of about 400–500μequiv kg⁻¹ h⁻¹ (Fig. 5B), which was essentially equal to the rate of NaHCO₃⁻ infusion (mean of 410gequivkg⁻¹ h⁻¹). This difference was entirely due to an approximately threefold elevation of J^TA from 200 to 600μequivkg⁻¹h⁻¹ as a consequence of NaHCO₃ infusion. J^Amm remained unchanged at about −200μequivkg⁻¹h⁻¹ in both infusion groups (Fig. 5A,B). The differences in and J^TA between the two groups were significant at every flux period and significant overall, while for J^Amm there were no significant differences between the two groups.

Control diffusive efflux of Na⁺ in trout measured before the beginning of infusion was about —160 and −250; μ equiv kg⁻¹h⁻¹ in the NaCl- and NaHCO₃-infused groups, respectively (Fig. 7). These values were not significantly different from each other. In the NaCl-infused group, Na⁺ efflux increased significantly by 140% to −384^equivkg⁻¹h⁻¹ at 19 h while there was no change in the efflux of Na⁺ in the NaHCO₃-infused group. Control diffusive Cl⁻ efflux in the fish to be infused with NaCl was − 166μequivkg⁻¹h⁻¹ while the mean diffusive efflux for the group to be infused with NaHCO₃ was −306μequivkg⁻¹h⁻¹. The reason for this significant difference is unknown, for the two groups were treated identically up to this time. It may be a random effect of the small sample size. Infusion of NaCl resulted in an increase in diffusive efflux of Cl⁻ by 125 % to −376 μequiv kg⁻¹ h⁻¹ at 19 h, while infusion of NaHCO₃ resulted in a decrease in the diffusive efflux of Cl⁻ by 65% to −108μequivkg⁻¹h⁻¹. The net difference (i.e. Na⁺—Cl⁻) in the efflux rates of these ions was not significantly different from zero in either group at the control measurement or in the NaCl-infused group at any time during the experimental regime. However, infusion of NaHCO₃ resulted in significant changes in the diffusive efflux of Na⁺ and Cl⁻ such that Na⁺ efflux exceeded Cl⁻ efflux by −60μequivkg⁻¹h⁻¹ at 10 h and by −150 μequivkg⁻¹h⁻¹ at 19 h (Fig. 7).

The uptake kinetic curves obtained by observation of and over increasing [NaCl]_e in both NaCl- and NaHCO₃-infused groups are shown in Fig. 8. The lines fitted to the data are based on the Michaelis-Menten model using the means of the K_m and J_max values obtained for individual fish by Eadie-Hofstee regression analysis, as described by Goss and Wood (1990). in NaHCO₃-infused fish was lowered by 50–75 % compared to that of the NaCl-infused fish over the entire concentration range (Fig. 8A). In contrast, was approximately doubled in NaHCO₃-infused fish compared with NaCl-infused fish at every [Cl⁻]_e (Fig. 8B).

Estimates of the J_max and K_m of the transport system in each treatment group, based on the Eadie-Hofstee regression analyses are summarized in Fig. 9. During NaHCO₃ infusion, the affinity of the Na⁺ transporter decreased markedly, i.e. increased significantly by 70% from 276 to 463 μequivl⁻¹ (Fig. 9A). Concurrently, decreased by 60% (689 to 262μequivkg⁻¹h⁻¹) in NaHCO₃-infused fish compared to NaCl-infused fish (Fig. 9B). The Cl⁻ transporter manifested no change in as a result of NaHCO₃ infusion (Fig. 9A) when compared with NaCl-infused fish, but was almost doubled (360 to 674μequivkg⁻¹h⁻¹, Fig. 9B) in NaHCO₃-infused fish.

The effects of increasing [NaCl]_e on the unidirectional outfluxes of Na⁺ and Cl⁻ as well as on and its components were also examined in the kinetic experiments. At almost every level of [NaCl]_e, both and were significantly lower in the HCO₃⁻-loaded fish compared to the NaCl-infused controls (data not shown), in agreement with the steady-state fluxes of series I (cf. Figs 3 and 4). This difference was much more marked for (two- to fourfold difference) than for (1.5-fold to twofold difference). As [NaCl]_e was increased, both and tended to increase in each group, though the data were very variable, reflecting difficulties in making accurate J_out measurements at high substrate concentration over short flux periods, as discussed by Goss and Wood (1990). The only significant effect was an increase in at the two highest [NaCl]_e levels in NaHCO₃-loaded fish.

The flux of titratable acidity (J^TA) and the net flux of acidic equivalents were significantly higher in the NaHCO₃-infused group compared to the NaCl-infused group at every [NaCl]_e (Fig. 10). In contrast, J^Amm was not significantly different between the two groups at any [NaCl]_e. In NaCl-infused fish, ⁠, J^TA and were not significantly altered as [NaCl]_e increased (Fig. 10A). In marked contrast, there were significant elevations in both J^TA and as [NaCl]_e increased in the NaHCO₃-infused fish; J^Amm was unaffected. When plotted, these increases also seemed to follow typical Michaelis-Menten uptake kinetic curves. Thus, one could estimate the K_m and J_max of the acidic equivalent uptake (=basic equivalent excretion) in terms of [NaCl]_e. This was accomplished by plotting versus in an Eadie-Hofstee regression analysis (Michal, 1985) for each individual fish. was obtained from the intercept and from the slope. From this analysis it was found that the (expressed in terms of [NaCl]_e) of the transporter was 133±57_luequivl⁻¹ and the was 753±150μequivkg⁻¹h⁻¹. These may be compared with the very similar values for the Cl⁻ transporter (⁠⁠; ⁠) and the very different values for the Na⁺ transporter (⁠⁠; ⁠) The affinity and maximal transport rates of the Cl⁻ transporter were not significantly different from those of the acidic/basic equivalent transporter, while the Na⁺ transporter was significantly different in both terms. This similarity of kinetic parameters provides strong evidence for the linkage of Cl⁻ influx and net acidic equivalent uptake (=basic equivalent excretion) in this particular situation.

Chronic infusion of NaHCO₃ produced a number of significant changes in the blood and plasma composition of the rainbow trout, compared to few changes in fish infused with NaCl (Fig. 1). The increases in plasma [HCO₃⁻] and pHa were similar to those levels found during the first few hours of normoxic recovery from 72 h of environmental hyperoxia (Wood and Jackson, 1980; Hõbe et al. 1984). The elevation in pHa in response to NaHCO₃ infusion occurred without any significant changes in (i.e. a pure metabolic alkalosis; Fig. 1D). Despite the Na⁺ load, plasma [Na⁺] remained unchanged but plasma [Cl⁻] was partially replaced by HCO₃⁻ (Fig. 1A,B) in a similar manner to that found by Wheatly et al. (1984) in hyperoxic trout. This presumably reflects the constraint of electroneutrality, with Na⁺, Cl⁻ and HCO₃⁻ being the major ions involved in the present situation.

Continuous infusion of NaHCO₃ at a rate of 410μequivkg⁻¹ h⁻¹ did not result in a constantly increasing plasma [HCO₃⁻]. Instead, the fish reached a point where excretion rate matched infusion rate (Figs 1C and 5). Plasma [HCO₃⁻] in the present study was not elevated beyond the proposed plasma HCO₃⁻ threshold of 30 mmol 1⁻¹ (Claiborne and Heisler, 1984), although this concept remains the subject of controversy (Cameron and Iwama, 1987).

In both the NaCl-infused and the NaHCO₃-infused fish, urine flow rate (UFR) increased by 80 % of the infusion rate. If the remaining 20 % of the infusion rate stayed in the extracellular fluid (ECF), this would have acted to increase ECF volume in a 300g fish by an average of 3.3ml over the 19h infusion period. Assuming an ECF volume of 27% body weight (Milligan and Wood, 1982), this volume would amount to a 4 % increase in ECF volume and even less if some penetrated into the intracellular compartment. Therefore, it is likely that there was a minimal change in the blood volume over the entire experimental period. The implication of this result is that the chronic infusion of either NaCl or NaHCO₃ did not greatly alter the perfusion characteristics of the gills. While ventilation was not measured, it also seems likely that these treatments would not cause any substantial changes, based on current knowledge of ventilatory control in fish (Perry and Wood, 1989). Thus, any observed changes in ion and acidic equivalent fluxes should be much less subject to the confounding ‘non-specific’ effects (i.e. perfusion and/or ventilation changes) that are known to occur during recovery from exposure to hyperoxia.

The kidney is known to play a much smaller role than the gills in compensation of acid-base disturbances with the relative renal contribution varying from 7 to 32 % (McDonald and Wood, 1981; Cameron and Kormanick, 1982; Holeton et al. 1983; Wheatly et al. 1984; Perry et al. 1987b; Vermette and Perry, 1987b; Wood, 1988). This appears to have been the case in the present study, where the kidney contributed a maximum of 13 % to the total net base excretion in NaHCO₃-infused fish (Fig. 2B). This excretion rate was probably due to the increased UFR, and thus probably glomerular filtration rate, coupled with a large increase in the filtered HCO₃⁻ load. Infusion of NaHCO₃ may have surpassed the renal threshold for the reabsorption of HCO₃⁻, as discussed by Wheatly et al. (1984). Urine ion measurements obtained in the accompanying study do not indicate any differential efflux of Na⁺ over Cl⁻; however, it must be assumed that this excretion of HCO₃⁻ was accompanied by equimolar amounts of strong cations in order to maintain electroneutrality. Ca²⁺ and K⁺ excretion rates via the urine have been reported to change during acid-base disturbances (McDonald and Wood, 1981; Wheatly et al. 1984; Perry et al. 1987b) but, unfortunately, they were not measured in the present study. Therefore, it appears that the kidney played a significant but much smaller role in comparison to the gills in ion and acid-base regulation in NaHCO₃-infused trout, in agreement with previous studies.

Infusion of NaHCO₃ resulted in a greater compared to NaCl infusion (Fig. 4) while was greatly reduced (Fig. 3). In both groups there was a large branchial net loss of Na⁺ (approx. −200μequivkg⁻¹h⁻¹) due to infusion of Na⁺ salts. However, in NaHCO₃-infused fish, there was a net branchial Cl⁻ gain (approx. 200 μequiv kg⁻¹ h⁻¹) over the exposure period while in NaCl-infused fish there was a net Cl⁻ loss (approx. −200μequivkg⁻¹h⁻¹). These results confirm those of previous studies showing that the fish adjust their branchial Na⁺/acidic equivalent and Cl⁻/basic equivalent transporters in a manner consistent with the correction of acid-base status (Cameron, 1976; McDonald et al. 1983,1989; Wood et al. 1984; Perry et al. 1987a,tr, Wood, 1988; Goss and Wood, 1990). Furthermore, the fish were in approximate acid-base equilibrium, with the rate of (approx. +400μequivkg⁻¹h⁻¹) corresponding to the rate of HCO₃⁻ infusion (approx. 410μequivkg⁻¹h⁻¹), which in turn corresponded to the difference between and (approx. −400μequivkg⁻¹h⁻¹).

However, these fish were clearly not in Na⁺ and Cl⁻ equilibrium - i.e. acid-base homeostasis was achieved at the expense of ionic homeostasis. In the face of a loading rate of 410μequivkg⁻¹h⁻¹ for both Na⁺ and Cl⁻ in the NaCl-infused trout, and of 410 μequiv kg⁻¹ h⁻¹ for Na⁺ and 0 μequiv kg⁻¹ h⁻¹ for Cl⁻ in the NaHCO₃-infused trout, the gill net fluxes (see above) were clearly not in balance. The urinary fluxes (Fig. 2) at best could account for only a small part of this difference. The plasma ion data (Fig. 1) were also clearly not in accord with the net balance situation of the fish. For example, plasma [Cl⁻] declined and plasma [Na⁺] was stable in the NaHCO₃-infused trout (which were in overall positive balance for both Na⁺ and Cl⁻). From this we conclude that the ‘missing’ Na⁺ and Cl⁻ entered the intracellular compartment or some other sink within the fish. The homeostatic role of this compartment is a topic of interest in future research.

In both the NaHCO₃-loaded fish and fish undergoing the normoxia-hyperoxianormoxia regime (Goss and Wood, 1990), acidic equivalents were exchanged in approximately equimolar amounts for the difference in the net flux of Na⁺ and Cl⁻. In both studies, changes in were more important than changes in in contributing to the dynamic response. In turn, changes in were more important than those in Alterations in the net flux of acidic equivalents in these experiments were accomplished as a result of increases in the J^TA component, with no significant changes in J^Amm (Fig. 5). This provides additional evidence for the linkage of Cl⁻ influx and the flux of basic equivalents. In terms of influx manipulations, compensation from an alkalosis occurs mainly though stimulation of the Cl⁻/basic equivalent exchange with a minor, but significant, role for the inhibition of Na⁺/acidic equivalent exchange (Claiborne and Heisler, 1984; Wood et al. 1984). Modulation of differential diffusive efflux also plays a significant role (see below). The continued excretion of ammonia during alkalosis when Na⁺/aci-dic equivalent exchange is inhibited indicates that ammonia excretion probably follows the flexible model of Wright and Wood (1985), i.e. via either or both Na⁺/NH₄⁺ exchange and NH₃ diffusion across the gill epithelium. Cameron and Heisler (1983) found that the latter alone was sufficient to explain ammonia excretion under control conditions.

In the present study there was not a large stimulation of in NaHCO₃-infused fish (Fig. 4). This suggests that the simultaneous increases in both and occurring immediately after return to normoxia in the previous hyperoxic studies (Wood et al. 1984; Goss and Wood, 1990) were probably the result of changes in the perfusive and ventilatory characteristics noticed during this time (Dejours, 1972, 1973; Wood and Jackson, 1980). Thus, they would constitute a non-specific response, rather than a response to acid-base status..

While the relationship between the flux of acidic equivalents and the difference between the net fluxes of Na⁺ and Cl⁻ was linear and highly significant (Fig. 6), it did not yield a -1:1 stoichiometry, in contrast to the results during exposure to hyperoxia (Goss and Wood, 1990). The greater and more prolonged excursions of pHa may have been a factor here. Although the net fluxes of Na⁺ and Cl⁻ made up a large part of the charge balance (approx. 70%), other ions must certainly have been involved. Increased fluxes of other ions across the gills have been reported under certain conditions (e.g. K⁺, McDonald and Wood, 1981; Eddy, 1985; Ca²⁺, Perry and Wood, 1985; SO₄²⁻, Hõbe, 1987) and a similar lack of −1:1 stoichiometry has been found in several other studies (Claiborne and Heisler, 1984; Perry et al. 1987a; Vermette and Perry, 1987a; Wood, 1988).

Infusion of NaHCO₃ resulted in a differential diffusive efflux of Na⁺ over Cl⁻ (Fig. 7), confirming the result found during recovery from exposure to hyperoxia (Goss and Wood, 1990). This acts to decrease the strong ion difference (SID) and constrain a necessary net gain of acidic equivalents inside the fish, thereby aiding in reducing the alkalosis. Differential diffusive efflux (Fig. 7), when compared with (Fig. 5) at comparable times, may have accounted for as much as 35 % of the total branchial net acidic equivalent uptake (=basic equivalent excretion). Goss and Wood (1990) estimated a figure up to 50% during post-hyperoxic alkalosis. Thus, differential diffusive efflux constitutes a third mechanism of acid-base regulation during metabolic alkalosis (in addition to stimulations of and inhibition of ⁠).

An important contribution of the present study was the near simultaneous measurement of the differential diffusive efflux (to NaCl-free water) and the measured outflux of Na⁺ and Cl⁻ (using radio-tracer analysis) in the same fish during acid-base disturbance. Previous studies addressing the role of differential outflux/efflux in acid-base regulation have employed only the radioisotopic method (Wood et al. 1984; Wood, 1988; McDonald and Prior, 1988; McDonald et al. 1989). Comparison of the estimates in Table 1 indicates that there were significant differences in the estimated size of the differential outflux/efflux component of acid-base balance, depending on the technique applied. Radioiso-topic outflux measurements were probably higher because of the presence of exchange diffusion (Cl⁻/Cl⁻, Na⁺/Na⁺; see Goss and Wood, 1990). The exchange diffusion component of radioisotopically measured outflux, by definition, can play no role in acid-base correction. Therefore, the diffusive efflux method (to NaCl-free water) provides a more accurate technique for assessing the role of differential diffusive efflux in acid-base regulation.

Infusion of NaHCO₃ resulted in complex alterations in both a and of the Na⁺ transporter, while for the Cl⁻ transporter there was no significant change in but a large increase in (Figs 8,9). The observed inhibition of Na⁺ influx in NaHCO₃-infused fish (Fig. 3) was accomplished by decreasing both the affinity (increased ⁠) and the maximum transport rate of the Na⁺ transporter (Fig. 9). This is directly comparable with the response in Na⁺ kinetics noted during recovery from hyperoxia by Goss and Wood (1990). Changes in Cl⁻ influx during recovery from hyperoxia occurred as a result of both decreases in (but only relative to the elevated value of final hyperoxic acidosis) and increases in whereas, in the present study, only an increased J_max occurred in response to metabolic alkalosis (Fig. 9). The values of were the same during HCO₃⁻ infusion (132 μequivl⁻¹); NaCl-infusion (135 μequivl⁻¹; Fig. 9) and recovery from hyperoxia (135μequivl⁻¹; Goss and Wood, 1990).

Therefore, it appears that K_m of both the Na⁺ and Cl⁻ transporters under control normoxic conditions operate at or near the maximum affinity (low K_m). Reducing the influx can then be accomplished by increasing K_m or by decreasing /max-However, increases in influx at this time can be accomplished only by increasing because K_m under control conditions is already close to the minimum possible value. It is not surprising that there is a limit to which affinity can be increased, and the transporters normally operate close to this maximal affinity. No such restriction is apparent for J_max, which could be either increased or decreased with respect to control levels for the purpose of acid-base correction. Interestingly, the K_m of the Na⁺ transporter was increased (Fig. 9) during infusion with NaCl compared to the control normoxia value (Fig. 9). This might be the result of the Na⁺ load, even in the absence of an acid-base disturbance. Infusion of NaHCO₃ resulted in a further reduction in the affinity of the Na⁺ transporter.

The possible mechanisms for changes in K_m and J_max were dealt with at length in Goss and Wood (1990). However, the present experiment has indicated that changes in K_m and J_max observed during post-hyperoxic alkalosis were not the result of the ‘non-specific’ mechanisms (perfusive and convective changes, altered O₂ levels) that complicated interpretation of the post-hyperoxia responses.

The increased excretion of acidic equivalents as [NaCl]_e was increased in the kinetics experiment (Fig. 10) was similar to, but much more pronounced, than the corresponding relationship during post-hyperoxic recovery. This occurred as a direct result of alterations in the J^TA component while J^Amm remained unchanged. These data indicated that the external availability of NaCl was directly limiting on the rate of acidic equivalent excretion and support the hypothesis that recovery from an acid disturbance should occur faster in water with a higher [NaCl]_e (Perry, 1982; Tang et al. 1988; McDonald et al. 1989).

The uptake of acidic equivalents (=basic equivalent excretion) seemed to follow typical Michaelis-Menten kinetic curves (Fig. 10). The observation that kinetic parameters of basic equivalent excretion, when expressed in terms of [NaCl]_e, were very similar to those of Cl⁻, provides perhaps the strongest evidence to date that the major uptake mechanism involved in the excretion of basic equivalents across the gills is the Cl⁻ transporter.

This research was supported through an NSERC operating grant to CMW. Many thanks are extended to M. Kovacevic and R. Rhem for excellent technical support throughout the project.