Responses of trout gill ion transport systems to acute acidosis

The rainbow trout Salmo gairdneri, like many freshwater animals, supports body fluid homeostasis with outer boundary ion uptake systems. They are located on the gills and are probably represented by cells on the axis of the filaments between the respiratory lamellae. The sodium and chloride systems show no interdependence in the trout (Kerstetter, Kirschner & Rafuse, 1970; Kerstetter & Kirschner, 1972), although in the goldfish (Carassius) some depression in the rate of uptake of one ion species has been shown when a penetrating counter ion is absent (Garcia-Romeu & Maetz, 1964) as has also been shown in the crayfish (Astacus) (Shaw, 1960a, c). Coupled exchange mechanisms are thought to maintain electroneutrality. Krogh (1938) suggested NH₄⁺ and HCO₃^- as likely endogenous species for exchange with Na⁺ and Cl⁻. But it appears that H⁺ can exchange for Na⁺ in such systems. Hydrogen is used exclusively by adult frogs and mosquito larvae (Garcia-Romeu, Salibian & Pezzani-Hemandez, 1969; Kirschner, Greenwald & Kerstetter, 1973; Stobbart, 1971), while both H⁺ and NH₄⁺, in variable proportions, can serve as exchanging cations in teleosts (Maetz, 1973), the crayfish Astacus (Shaw, 1960b), and bullfrog (Rana catesbeiana) tadpoles (Dietz & Alvarado, 1974).

Two questions are considered in this paper. The first concerns the known stimulation of sodium influx in teleosts by the injection of ammonium salts (Maetz & Garcia-Romeu, 1964; Kerstetter et al. 1970). Given the observed increase in sodium uptake, is additional blood ammonia alone responsible, or is the resulting acidosis a contributing cause ? Secondly, might the Na⁺/H⁺ exchange system in the teleost gill respond to changes in blood pH and thus serve as an acid-base regulator ? The experiments described here were designed to show how ion transport responds to acute metabolic acidosis in the absence of ammonia loading.

Rainbow trout, 200–400 g body weight, were obtained from the Humboldt State University experimental fish hatchery and were maintained unfed at 6 °C in tap water, of which the NaCl content was 0·2 mm. At least 4 days before use they were transferred to a holding tank at 11 °C. The experiments were designed to measure the active transport of sodium and chloride by means of a small-volume system which recirculated external bathing solution (1·0 mm NaCl, 11° C) over the gills. Details of the method are described by Kerstetter et al. (1970). Light anaesthesia was induced with 0·1 % tricaine methane sulphonate and maintained with urethane (ethyl carbamate).

Each experiment was begun with a 1-h ion flux measurement to establish a control value for the animal. Following that, the shaft of a 23-gauge needle, connected to a syringe pump via PE 50 Intramedic tubing, was inserted into the ventral aorta. (An imaginary line connecting the anterior roots of the pectoral fins is an adequate landmark for guiding the insertion of the cannula.) Infusion of 0·05 M D, L-lactic acid, at 3·6 ml/h was then started, and 15 min later the second 1-h flux measurement began. At the end of the second hour, the infusion solution was changed to 0·145 mm NaCl, and the infusion was continued at the same rate during a third, hour-long flux determination. Control fish received 0·145 mm NaCl throughout. The lactic acid solution was made isotonic to fish body fluids by adding NaCl.

Blood was collected without exposure to air by withdrawing samples through the infusion cannula, and pH was measured at 13 °C in a capillary electrode (Instrumentation Laboratories Micro Blood Gas Analyser). Sodium analysis was by atomic absorption spectrophotometry; chloride was measured amperometrically with a Cotlove-type chloride titrator; and ²²Na was counted in a well-type gamma scintillation spectrometer. Cl-36 was determined by liquid scintillation after appropriate correction for ²²Na. Ammonia was measured by Nessler’s test after microdiffusion (Seligson & Seligson, 1951).

Sodium and chloride influxes were determined by calculating the rate of disappearance of isotope from the medium. To avoid influx rate changes due to variations in external salt concentration, an instantaneous rate was estimated near the beginning of each hour-long period, when the NaCl concentration was still very close to 1·0 mm. No correction for back flux of isotope was made since the specific activity of body fluid electrolytes was always less than 5 % of the external medium. Statistical significance was calculated with Student’s t test for difference between two means, using in most cases the one-tailed test.

Table 1 is a summary of results. Statistically significant differences are noted therein. It is readily apparent that lactate infusion did not stimulate sodium uptake, even though blood pH dropped sharply. The increase in Na⁺ influx during the second hour is clearly not significant, and is equalled by a similar response in the control group. Chloride uptake was also unaffected by the experimental treatment, the differences not being significant. Ammonia excretion was not significantly changed by acid infusion.

A previous report suggested that introduction of ammonium salts could stimulate sodium influx in the teleost gill by a secondary acidosis (Kerstetter et al. 1970). The work reported here contradicts that earlier speculation and clearly shows that acute metabolic acidosis does not affect the rate of either sodium or chloride influx ; it may be concluded that blood ammonium interacts directly with the sodium uptake system.

A commonly accepted model for Na⁺ and Cl⁻ uptake by the teleost gill is based upon unlimited supplies of H⁺ and HCO₃⁻ from the hydration of CO₂, catalysed by high levels of carbonic anhydrase in the gill (Maetz & Garcia-Romeu, 1964). In the model the HCO₃⁻ exchanges with Cl⁻, and the H⁺ can exchange directly with sodium or can combine with NH₃ to provide ammonium ion for exchange (Maetz, 1973). The model provides a possible mechanism for acid-base regulation via either unilateral changes in Cl⁻/HCO₃⁻ or Na/H⁺ (NH₄⁺) exchanges, or simultaneous changes in opposite directions. Shaw (1964) has postulated that ion uptake systems in Astacus also respond in a compensatory way to shifts in body fluid pH. Experimental evidence for the involvement of these ion transport systems in acid-base regulation was recently supplied by De Renzis & Maetz (1973). They induced shifts in blood pH by adapting goldfish to either sodium-free or chloride-free media; presumably the resulting decrease in either H⁺ or HCO₃⁻ excretion led to the observed pH shift. Fish kept in choline chloride became acidotic, and plasma levels of both Na⁺ and Cl⁻ were markedly reduced. Fish kept in Na₂SO₄ became alkalotic, plasma levels of both Na+ and Cl⁻ were similarly reduced, and when the fish were replaced in a NaCl medium, the Cl⁻ influx was increased threefold over a group adapted to de-ionized water. They reasoned that, since plasma sodium and chloride were reduced about equally, the signal for the unilateral increase in Cl^- influx was probably the low blood pH. The increase in Cl⁻ influx would tend to restore pH balance in the acidotic animals by the rapid excretion of HCO₃⁻. But the time course of the change in ion transport was not measured, and since the animals were kept in the depleted media for 3–4 weeks, transport rate could have changed slowly in response to the shift in blood pH. In addition, the depressed levels of plasma Na⁺ and Cl⁻ could well have been acting synergistically with the pH imbalance to bring about the observed effect. The experiments reported here make it clear that a rapid response to blood pH changes is not a characteristic of gill ion transport systems, since the sharp drop in blood pH associated with lactic acid infusion did not produce significant changes in either Na⁺ or Cl⁻ influx. A delayed response to blood pH shifts is not ruled out by these experiments, indeed the work of De Renzis & Maetz supports that possibility, but the role of the gill in short term, day-to-day pH regulation (with plasma electrolyte levels in the normal range) remains unclear.

This work was supported by NSF grant GB-35537.