ABSTRACT
The effects of exposure to 0·94% (300mosmoll−1) sodium chloride on plasma electrolyte and acid-base status were examined in the freshwater stenohaline teleost Catostomus commersoni (Lacépède), the white sucker. Four days’ exposure to this maximum sublethal salinity resulted in an increase in plasma concentrations of both sodium and chloride but a decrease in the Na+/Cl− ratio. Since the plasma concentrations of free amino acids and other strong ions - Ca2+, Mg2+ and K+ - remained unchanged, plasma strong ion difference (SID) decreased. Additionally, plasma pH and bicarbonate concentration decreased at constant. The changes in electrolyte and acid-base status that occurred after the 96 h were not appreciably altered after a further 2–3 weeks of saline exposure. The ambient calcium concentration had no influence on these results.
Haemolymph non-bicarbonate buffer capacity (β) calculated as Δ[HCO3−]/ ΔpH, increased in saline-exposed fish. Consequently ΔH+, the apparent proton load, was zero despite the apparent change in acid-base status. Although β was directly proportional to the haemoglobin concentration in both control and experimental fish, this could not account for the increase in β since haemoglobin remained at control values. These results can be explained solely by the change in plasma SID and serve to illustrate the dependence of plasma acid-base status on the prevailing electrolyte characteristics, weak acid concentration and .
INTRODUCTION
The extracellular electrolyte and acid-base status (pH and bicarbonate concentration) of freshwater fish are currently considered to be regulated by Na+/H+ and C1−/HCO3− exchange mechanisms located in the gill epithelia. These exchange mechanisms were first proposed by August Krogh (1939) after he observed that the mechanisms of sodium and chloride uptake in the goldfish operate independently. He reasoned that the efflux of protons in exchange for sodium, and bicarbonate in exchange for chloride, was the most likely means of maintaining electroneutral charge movement across the gill epithelia. Currently it is believed that sodium uptake may utilize either protons or the ammonium ion, while chloride uptake may utilize either bicarbonate or hydroxyl ions (Kirschner, 1979; Evans, 1980a,b, 1982; Haswell, Randall & Perry, 1980). Thus, the active uptake of sodium and chloride serves not only to compensate for their diffusional loss, but is also thought to provide the means of removing acid-base constituents, protons and bicarbonate, and therefore to regulate extracellular pH.
The premise inherent in this model is that extracellular proton and bicarbonate concentrations are regulated in the same manner as sodium and chloride, i.e. the concentration of protons and bicarbonate is a function of the rate of input and rate of removal. Although intuitively appealing, this premise is simply incorrect, as can be seen from analyses from first principles (Stewart, 1978, 1981, 1983). Plasma proton and bicarbonate concentration are determined by (i) the difference between the sum of all strong cations and anions, i.e. the strong ion difference (SID); (ii) the total concentration of weak acids (ATOT), essentially plasma protein concentration; and (iii) the amount of dissolved carbon dioxide, Thus, an observed change in proton and bicarbonate concentrations can only reflect a change in one or more of these three independent variables. By extension, we propose that concepts such as non-bicarbonate buffer values (Δ [HCO3−]ΔpH), the base deficit (the proton load expressed as a loss of bicarbonate equivalents), and the buffering of plasma pH by intracellular proteins (i.e. haemoglobin) cannot be used meaningfully in terms of either quantitative or cause and effect analysis of the prevailing acid—base status, since they are all expressed as changes in dependent variables - proton and bicarbonate concentration.
The purpose of the present study was to demonstrate that the above concepts can be explained solely in terms of changes in the three independent variables, SID, and ATOT, which occur during acid-base disturbances. The approach adopted was to examine the changes in these independent variables and in the dependent variables, plasma pH and bicarbonate concentration, when the stenohaline freshwater teleost, Catostomus commersoni (the white sucker) is exposed to a hypersaline (0·94% sodium chloride) environment.
MATERIALS AND METHODS
White suckers, Catostomus commersoni (150–400 g) were obtained by live trapping from artificial lakes within the City of Calgary and were maintained at the University of Calgary for a minimum of 2 weeks prior to use.
Two control and two experimental regimes were used in the present study; control water contained high (> 1·0 mmol 1−1) or low (<0-1 mmol 1−1) calcium (hard and soft freshwater, respectively), while saline water contained high or low calcium in 0-94% (300mosmoll−1) sodium chloride (hard and soft saline water, respectively). These water conditions were obtained by supplementing dechlorinated tap water with calcium nitrate for hard control water, and by supplementing tap distilled water with sodium chloride for soft control water. Concentrations of sodium, calcium, potassium and magnesium in hard and soft freshwater are given in Table 1. Hard and soft saline waters were made in a similar manner except that the sodium chloride level was increased to 0·94% (300mosmoll−1). Preliminary studies (Wilkes, 1984) demonstrated that 300mosmoll−1 sodium chloride closely approximates to the maximum sublethal salinity for this species.
Experimental and control facilities consisted of constantly aerated, recirculating water on line with gravel filters. Each system embodied a small holding reservoir (approximately 1501) in which animals could be acclimated to the desired water conditions, and a wet table on which fish could be placed in individual fish boxes. Each fish box was 21 in volume, was darkened with opaque plastic, aerated and was supplied with a constant inflow of water from the reservoir at a rate between 0·5 and 1·01 min−1. The wet table was designed to return the overflow water from the fish boxes back through the gravel filter to the reservoir. The temperature of the systems was maintained at 15°C. Animals were fed daily on pellet trout food (Silver Cup) during holding and acclimation periods.
Experimental procedures
All fish were acclimated to either hard or soft freshwater conditions for 2 weeks prior to exposure to hard or soft saline water respectively. Some of the freshwater-acclimated fish were used as hard and soft freshwater controls. Three separate groups of hard freshwater-acclimated fish were exposed to hard saline water for 24 h, 96 h and several weeks (chronic), respectively, while a single group of soft-water-acclimated fish was exposed to soft saline water for 96 h.
Preliminary studies (Wilkes, 1984) demonstrated that caudal vessel cannulation clearly added an intolerable level of additional stress, at least to experimental groups. Blood samples were therefore taken by cardiac puncture. This factor necessarily influenced the experimental design.
Each fish was transferred, in its fish box, to the operating table where the input jack on the fish box was connected to the anaesthetic line (MS 222, 1:10000). The water used for anaesthetizing the fish was of the same temperature and electrolyte composition as that to which the fish had previously been exposed. This method permitted anaesthetization with a minimum of stress and no handling. As soon as the anaesthetic had taken effect (approximately 2Δ3 min) the fish was removed from the box, placed on the sling of the operating table, dorsal side up, and the gills ventilated through the mouth. The heart was exposed immediately using a mid-line incision through the cliethrum and a blood sample was drawn from the bulbus arteriosis into a heparinized syringe. Although the sample volume varied with the size of the fish it was generally possible to obtain 1–2 ml. The entire process, surgery and sampling, was completed within 2 min.
Acid-base analyses
Blood protein analyses
Total plasma protein concentration was measured with the Biuret reagent (Sigma). The haemoglobin concentration of whole blood was measured by the cyanomethaemoglobin method (Hycel). The intracellular haemoglobin concentration was estimated from ([Hb]/Ht) × 100.
Plasma and tissue electrolyte analyses
Plasma was analysed for the cations Na+, Ca2+, Mg2+, K+, and the anion Cl−, total free amino acids (ninhydrin-positive substances, NPS) and osmolality. Cations were measured either with a Jarrell Ash 850 or a Perkin Elmer 5000 atomic absorption spectrophotometer. Plasma samples were diluted in appropriate swamping agents (0·1% potassium chloride for sodium, 0·1% sodium chloride for potassium, and 1% lanthanum chloride for calcium and magnesium) prior to analyses. Plasma chloride was measured directly by coulometric titration (Buehler Chloridometer). Free amino acid concentration was calculated as the total ninhydrinpositive substances (NPS) following the method of Clark (1973). Although never more than 5% of the ninhydrin reaction, total ammonia concentration in the deproteinized sample was routinely measured (Solorzano, 1969) and subtracted from the total NPS. Plasma osmolality was measured with a vapour pressure osmometer (Wescor, 5100B).
Since the blood sampling procedure was terminal, the same animal could be used to obtain muscle samples for analyses of ions, amino acids and the percentage hydration. Muscle ions were measured by the method developed by McDonald, Hôbe & Wood (1980). Briefly, a 2- to 3-g muscle sample was removed from the epaxial muscle mass and analysed for wet weight/dry weight. The dried tissue was ground to a fine powder with pestle and mortar to facilitate elution of ions into 1·0equivl−1 nitric acid. After 48 h of constant rotation at room temperature the nitric acid supernatant was analysed for sodium, potassium, calcium, magnesium and chloride concentrations as described above. Muscle ion concentrations were reported in terms of total tissue water, intracellular plus interstitial. An additional wet tissue sample was homogenized in distilled water (1·0ml/0·25 g) using a Sorvall Omni mixer for measurement of free amino acids, also as described above for plasma. These muscle ion and amino acid analyses were performed only in a group of hard-water-acclimated control fish and in a group exposed to hard saline water for 96 h.
Calculations
RESULTS
Plasma and tissue electrolyte status
The only significant differences in plasma electrolyte status between control fish acclimated to hard and soft freshwater were reduced concentrations of sodium and calcium in the latter group (Figs 1, 2). A consequence of the lower plasma sodium concentration in soft-water-acclimated control fish was that the Na+/Cl− ratio of 1·16 ± 0·03 was significantly lower than the 1·30 ± 0·04 value found in hard-water control fish (Fig. 1).
Plasma sodium and chloride concentrations of fish initially acclimated to hard freshwater increased significantly after 24h exposure to hard saline water. By 96h, the concentrations of sodium and chloride were significantly greater than control values by 20% and 33%, respectively (Fig. 1). The greater increase in plasma chloride concentration over that of sodium resulted in a significant (17%) decrease in the Na+/Cl− ratio from 1·30 ± 0·04 to 1·08 ± 0·02. No further significant changes in sodium and chloride concentrations, or their ratio occurred in plasma of fish chronically exposed to hard saline conditions.
Changes in plasma sodium and chloride concentrations which occurred in fish exposed to soft saline water were qualitatively similar at 96 h to those observed in fish similarly exposed to hard saline water. The concentration of plasma chloride again increased substantially more than that of sodium, resulting in a significant (13%) decrease in the Na+/Cl− ratio from 1·16 ± 0·03 to 1·01 ± 0·02 (Fig. 1).
Exposure to either hard or soft saline conditions had no significant effect on the plasma concentrations of calcium, magnesium or potassium. Free amino acid concentration, measured only in the plasma of hard-water exposed fish, was also not significantly affected by saline exposure (Fig. 2).
The concentration of ions in epaxial muscle was measured only in a hard-water control group and in a group exposed to hard saline water for 96 h. Tissue calcium and magnesium concentration, and percentage tissue hydration in the saline-exposed animals were not significantly different from control values (Table 2). However, dramatic increases in sodium, chloride and free amino acid concentrations, of 85%, 175% and 66%, respectively, were measured in muscle samples from fish exposed to hard saline conditions for 96 h. Potassium concentration also increased significantly over control values, but only by 17% (Table 2).
Blood acid-base status
Both pH and bicarbonate concentration in true plasma of soft-water-acclimated control animals were significantly elevated over corresponding values measured in the hard-water control group (Fig. 3). Nevertheless, the acid-base status in blood of both groups was affected in a qualitatively similar manner by saline exposure. True plasma pH and bicarbonate concentration of fish exposed to hard saline conditions for 24 h were not significantly different from control values. However, by 96 h, pH had fallen significantly from a control value of 7·770 ± 0·053 to 7·408 ± 0·022, and fell even further to 7·241 ± 0·039 after chronic saline exposure. Initial true plasma bicarbonate concentration in hard-water control fish was 3·7 ± 0·03 mmol 1−1, and the 96 h value was significantly lower at 1·3 ± 0·1 mmol 1−1. No further statistically significant change occurred in the bicarbonate concentration during chronic saline exposure.
True plasma pH in the soft-water-exposed group fell significantly from a control value of 7·934 ± 0·030 to 7·485 ± 0·043 after 96 h exposure to soft saline water. True plasma bicarbonate concentration also decreased significantly during this period, from 7·6 ± 0·07 mmol 1−1 to 3·0 ± 0·2 mmol 1−1 (Fig. 3). The acid-base changes in fish exposed either to hard or soft saline water could not be attributed to a build-up of lactic acid which decreased progressively throughout saline exposure (Fig. 3).
The slopes (β) of whole blood buffer curves constructed from blood of fish exposed to hard and soft control conditions, to 96 h hard and soft saline water, and to 3 weeks hard saline conditions, were calculated from the change in true plasma bicarbonate concentration per unit change in pH, ΔHCO3−] ΔpH, which occurred after equilibration of whole blood to increasing levels of The mean and standard error for βat haemoglobin concentrations of 2, 4, 6 and 8g% were calculated by analysis of variance of the β-haemoglobin regressions according to Zar (1974). The results of these analyses demonstrated that the absolute magnitude of increased with increasing haemoglobin concentration in blood of both hard- and soft- water-acclimated control fish, and hard and soft saline-exposed fish. Additionally, βwas observed to increase independently of the prevailing haemoglobin concentration in saline-exposed fish (P<0·05 at mean haemoglobin concentrations). This latter effect was influenced more by the duration of saline exposure rather than water hardness (Fig. 4).
A consequence of the haemoglobin-independent increase in that occurred during saline exposure was that a negative value for ΔH+ was calculated for fish exposed to hard or soft saline conditions (Table 3) despite the decrease in pH and bicarbonate concentration. However, the ΔH+ values are in fact zero since there were no significant differences between H+c and H+8 (Table 3).
Blood protein characteristics
Plasma protein concentration, whole blood haemoglobin concentration, and haematocrit of soft-water-acclimated control fish were all significantly below values measured in hard-water-acclimated control fish (Table 4). Exposure to soft saline water for 96 h had no effect on plasma protein concentration. However, by 96 h and after chronic exposure to hard saline conditions, plasma protein concentration had fallen significantly below values found in hard-water-acclimated control fish. The only statistically significant effect of either hard or soft saline water on whole blood haemoglobin concentration was a lower value in the group chronically exposed to hard saline conditions. Haematocrit values were significantly higher than control values in the two groups exposed to hard and soft saline water for 96 h. Haematocrit values of fish chronically exposed to hard saline water were not significantly different from control values. The changes in haemoglobin concentration and haematocrit were such that the mean red cell haemoglobin concentration was significantly reduced in fish exposed to hard saline conditions for 96 h, and after chronic exposure to hard saline water. However, the decrease in mean red cell haemoglobin concentration could not be attributed to consistent changes either in whole blood haemoglobin concentration or in haematocrit.
DISCUSSION
Numerous studies have examined the effects of sublethal saline exposure on the plasma electrolyte levels of stenohaline freshwater fish (Davis & Simco, 1976; Norton & Davis, 1977; Kilambi & Zdinak, 1980; Maceina, Nordlie & Shireman, 1980; Hegab & Hanke, 1982). Although it is difficult to establish a specific maximum sublethal salinity, results from the above studies, and from Wilkes (1984), indicate that stenohaline freshwater fish cannot survive salinities in excess of 300–400 mosmol 1−1. A common finding in all the above studies, as well as the present results, is that stenohaline freshwater fish could survive a saline stress only so long as plasma osmolality was above ambient. Hegab & Hanke (1982) suggested that an inwardly directed osmotic gradient is required in order to obtain the water influx necessary to ensure continued renal function.
On initial exposure to a hypersaline environment some osmotic loss of water was expected, but this seemed to be minimal in C. commersoni since both plasma and tissue levels of calcium and magnesium remained constant (Fig. 2). Therefore, the observed increase in plasma osmolality was more a function of a net influx of sodium and chloride than of water loss. In either case, intracellular osmolality must match that of extracellular fluid in order to avoid cell shrinkage. Current theories on cell volume regulatory mechanisms state that both inorganic, essentially sodium and potassium, and organic electrolytes, principally non-essential amino acids and taurine (Gilles, 1979; Rorive & Gilles, 1979) are used to maintain the intracellular environment isosmotic with surrounding fluids. Although sodium, chloride and, to a lesser extent, potassium concentrations in tissue increased during saline exposure (Table 2), these concentrations are expressed in terms of total tissue water, intracellular plus interstitial. Therefore, some of the increase in sodium and chloride concentration was attributable to an increase in overall extracellular osmolality. However, the plasma (and therefore interstitial fluid) concentration of free amino acids did not change during 96 h of saline exposure, indicating that the measured 66% increase in tissue free amino acid concentration was predominantly intracellular.
Before analysing the effects of saline exposure on acid-base status, a brief discussion of the approach used in the present study is required. Wilkes, Walker, McDonald & Wood (1981) demonstrated that cannulation of a caudal vessel can result in decreased plasma protein concentration and haematocrit, which may in turn perturb the acid-base status. Cannulated fish rarely survived beyond 48 h in 300 mosmol 1−1 sodium chloride and experienced an acid-base disturbance which was often characterized by an increase in lactic acid concentration 24 h prior to death. However, before becoming lactacidotic and moribund, blood pH and bicarbonate concentration fell while (measured directly by a method adopted from Boutilier, Randall & Toews, 1978, and DeFur, Wilkes & McMahon, 1980) remained constant. These preliminary results indicate that the initial decreases in pH and bicarbonate concentration were consistent with a decrease in plasma strong ion differences (SID) brought about by a reduction in the plasma Na+/Cl− ratio. The in vitro methods described in the present study allow evaluation of the blood pH and bicarbonate concentration at constant without the complications brought on by surgical manipulation. The actual levels used, 1·3, 4·0, 9·1 and 13·0 Torr, were chosen so as to encompass the range of normal values but are not meant to be specifically representative of arterial or venous blood.
Therefore, the haemoglobin-independent increase in β observed in saline-exposed fish is a direct consequence of the lower plasma SID. The means by which the plasma SID decreases, i.e. whether through a decrease in the sodium chloride ratio as in the present study, or through a lactacidosis during exercise, is not important.
The consequences of the chloride shift to plasma proton and bicarbonate concentrations are greatly magnified by in vitro methodology. The greater the haematocrit of an isolated blood sample, the greater the chloride shift and the greater the increase in plasma SID for a given increase in In the whole animal, the effect of the chloride shift on plasma SID is diluted throughout the entire extracellular compartment, not just the plasma volume. Additionally, any compensatory effects of strong ion movement between the interstitial and intracellular compartments which would normally occur in vivo are also eliminated.
It is possible’to use equation 4 above from Stewart (1981) to demonstrate quantitatively the consequences of an increase in on the β values for two hypothetical blood samples. The first plasma sample has an initial SID of 35 mequivl−1 (represented by the stippled bars in Fig. 5). The second sample has an initial SID of 20mequiv 1−1 (represented by the clear bars in Fig. 5). These two hypothetical plasma samples are then subjected to an increase in (from 10 to 20 Torr) in the following three conditions. Condition I represents the case of plasma without red cells so that there is no chloride shift and SID does not change from its initial value when increases. Conditions II and III represent the same plasma samples but with red blood cells added to cause a 1 and 2mequivl−1 increase in plasma SID, respectively, to occur via the chloride shift as the is raised from 10 to 20 Torr. Note that by using these SID and values in equation 4 it is readily apparent (Fig. 5) that even these small, almost immeasurable, increases in plasma SID can have a pronounced effect on the proton and bicarbonate concentrations such that the lower the SID the greater the value of β. Additionally, the greater the red cell concentration in both of these samples, the greater the increase in. β
The above explanation for the haemoglobin-independent and haemoglobin dependent increases in ft observed in the present study points out the inability of the ΔH+ calculation to determine a ‘proton load’ (Table 3). Indeed, the term ‘proton load’ is meaningless since the proton concentration within a system is not a simple function of the number of protons added to it (Stewart, 1978, 1981, 1983). Since similar statements are equally true for bicarbonate, it follows that calculations based solely on the change in dependent variables (i.e. β and ΔH+) are of no functional use in determining the cause, or extent of a given change in acid-base status.
The initial differences in acid-base status in the two control groups of fish acclimated to hard and soft freshwater (Fig. 3) can probably be explained by the differences in plasma protein concentration (Table 4). Although the plasma SID was higher in fish acclimated to soft freshwater than in those acclimated to hard freshwater, the plasma protein concentration (ATOT) was much lower. Using equation 4 (from Stewart, 1981) it is possible to show that the influence of ATOT ON proton concentration is greatly enhanced at low levels. Thus, the reduced plasma protein concentration in soft-water-acclimated control fish has a more pronounced effect on proton concentration than the low SID. It is difficult to determine from these results whether the differences in protein concentration, and therefore acid-base status, observed in the two control groups is a function of water calcium concentration. Although Hóbe, Wood & McMahon (1984) reported that blood pH and bicarbonate concentration of C. commersoni reared in natural soft-water is higher than in C. commersoni reared in natural hard-water conditions, the plasma protein concentration in the former population was, in fact, higher than the levels reported in the present study for either control group. The lower plasma protein concentration observed in the present study may be attributable to a combination of the effects of several months’ captivity and of an artificial diet. Interestingly, the acid-base status of trout blood is not affected by water hardness of neutral pH (McDonald et al. 1980).
Relatively few studies have examined the effects of a change in external salinity on acid-base status of fish. DeRenzis & Maetz (1973) demonstrated that goldfish maintained in dilute choline chloride developed an acidosis, while dilute sodium sulphate exposure effected an alkalosis. Farrell & Lutz (1975) reported that when brackish water adapted eels were exposed to freshwater there was an increase in the SID in association with an increase in plasma bicarbonate concentration. Finally, Smatresk & Cameron (1982) reported the development of a metabolic acidosis, which they attributed to a decrease in SID, in the spotted gar after transfer from freshwater to 50% seawater. The results of these studies indicate that species which cannot readily tolerate or adapt to large changes in external salinity are subject to acid-base disturbances during such treatment. In all of the above studies the observed change in plasma pH is directly related to the change in plasma SID.
Apparently, euryhaline species are not subject to the same acid-base disturbances when external salinity is altered. While Perry & Heming (1981) showed a statistically significant alkalosis in trout transferred from freshwater to 25% seawater, the acid—base disturbance was, physiologically, quite small. Similarly, Milne & Randall (1976) and Bath & Eddy (1979a,b) showed no change in blood pH or bicarbonate concentration in rainbow trout after an abrupt change from freshwater to full strength seawater. Apparently, the ionoregulatory mechanisms of trout are capable of controlling both absolute and relative concentrations of plasma electrolytes during saline stress.
The results of the present study indicate that the Na+/H+-NH4+ and Cl−/HC03−-0H− of acid-base regulation is too simplistic. This model proposes that it is the rate of active uptake of sodium and chloride which is responsible for the removal, and therefore regulation, of acid-base constituents, protons and bicarbonate. However, as pointed out by Stewart (1978, 1981, 1983), the prevailing proton and bicarbonate concentrations have nothing to do with their rate of input to or removal from a given compartment. The proton and bicarbonate concentrations are dependent variables which are determined by the prevailing SID, ATOT and - If: (i) the plasma SID is determined predominantly by the difference between sodium and chloride concentration, and (ii) the sodium and chloride concentrations are functions of their rate of input (i.e. as determined by the magnitude of the rate constants for the active pumps and the external substrate concentration), and their rate of efflux (i.e. as described by the Nernst-Planck diffusion equation), then it follows that the only way in which the transepithelial movement of sodium and chloride can influence plasma acid-base status is if active and/or passive movement is such that the sodium to chloride ratio in the plasma is altered. A corollary of this is that any measured transepithelial flux of protons and/or bicarbonate (acidic equivalents) is a reflection of the change of SID in the external compartment brought about by the transepithelial movement of strong ions. The transepithelial flux of ‘acidic equivalents’ should not be interpreted as a factor responsible for causing or correcting the blood pH. Changes in transbranchial ion fluxes in C. commersoni during saline exposure are examined in the following paper.
ACKNOWLEDGEMENT
We would like to acknowledge our thanks and appreciation to Dr R. L. Walker for his assistance in the procurement of animals and the loan of equipment, and to Dr P. Stewart for his helpful comments and criticisms regarding the manuscript. This work was supported by NSERC Grant A5762 to BRM, and an AHFMR (Studentship) to PRHW.