Roles of Buffering Capacity and Pentose Phosphate Pathway Activity in the Gas Gland of the Gulf Toadfish Opsanus Beta

In order to address these three facets of the metabolism of the gas gland, we have conducted in vitro experiments with the gas gland of the gulf toadfish, Opsanus beta. Our key findings are (1) that the tissue has an extremely high non-bicarbonate buffering capacity; (2) that the tissue has an active pentose phosphate pathway (as demonstrated by relative 14CO₂ production rates from C1-versus C6-labelled glucose); and (3) that flux through the pathway is enhanced under hyperoxic conditions.

Gulf toadfish, Opsanus beta (Goode and Bean), were collected by roller trawl from Biscayne Bay, Florida, in July and August 1990 for metabolic flux experiments and enzyme activity measurements and in November 1990 for buffering capacity measurements. Fish were held for no more than 1 week in the laboratory, without feeding, in flowing sea water, at 24–28°C and with a salinity of 33‰, under a natural photoperiod. Fish were anaesthetized with MS-222 for 5min at 0.5 gl⁻¹ in sea water buffered with NaHCO₃. The swim bladder was removed through a ventral incision and opened from the dorsal side, and the gas gland and surrounding layer of epithelial tissue was then gently peeled away from the ventral floor of the bladder. Detailed drawings of the toadfish swim bladder and gas gland appear in Fänge and Wittenberg (1958). For measurements of enzyme activities and buffering capacity, glands were placed in 1ml plastic cryotubes and frozen at −80°C. Enzyme activities and buffering capacities were measured within 30 days of freezing. The frozen glands for buffering capacity measurements were transported on dry ice to London, Ontario. The total non-bicarbonate buffer capacity of the gas gland was determined by acid titration as described by Cameron and Kormanik (1982). Individual gas glands were ground to a fine powder with a mortar and pestle under liquid nitrogen and then suspended in 3ml of 0.9% NaCl. The tissue homogenate was titrated to pH8.0 with 1mol l⁻¹ NaOH, allowed to stabilize and then back titrated to pH6.8 with 0.02mol l⁻¹ HCl at 25°C under a nitrogen atmosphere (to prevent absorption of atmospheric CO₂ by the sample, which would affect the buffering). The slope of the line relating pH with μmol HCl added was taken as an estimate of the total non-bicarbonate buffer capacity in μmolpHunit⁻¹ g⁻¹ (or slykes).

For radiotracer experiments, the gas gland was transferred to a glass slide and carefully sliced approximately in half with a scalpel such that each half had part of the rete and surrounding epithelial layer. Both halves were weighed (range was approximately 25–100mg per half) and placed in 0.95ml of toadfish Hanks salts solution, plus 3 mmol l⁻¹ NaHCO₃, 1mmol l⁻¹ CaCl₂ and 1mmol l⁻¹ glucose, in the bottom of a 20ml glass scintillation vial, as in prior studies of toadfish hepatocytes (Walsh, 1989). After a 15min preincubation period, 50 μl of saline containing 0.5 μCi of either C1-or C6-labelled [14C]glucose (Amersham, Arlington Heights, IL) was added to the flask.

Labelled CO₂ can be produced from C1-labelled glucose by both the pentose phosphate pathway and aerobic glycolysis, whereas C6-labelled glucose gives rise to 14CO₂ only via glycolysis. Therefore, the ratio of CO₂ production from C1-versus C6-labelled glucose gives the relative activity of pentose phosphate pathway activity. All experiments were paired in that half of the gland from each fish received C1 label and the other half received C6 label on a random basis, without regard for left or right or for slight differences in weights of the half glands. The flasks were immediately sealed with rubber stoppers containing centre wells with circles of 2.4cm diameter Whatman GF/C filters, and the tissues were incubated with gentle shaking for 2h. Chemical controls (i.e. as above without tissues) were run in parallel daily. For hyperoxic incubations, saline solutions were vigorously gassed with 100% O₂ for 1h prior to bicarbonate addition and pH adjustment and, after capping the vials, the gas space was purged with 100% oxygen for 30s. No loss of 14CO₂ by this purging was detected. Typical oxygen partial pressures were 73.3kPa (550mmHg) as determined with a Radiometer oxygen electrode and blood gas analyser. At the end of the incubation, 0.2ml of hyamine hydroxide solution (Sigma, St Louis, MO) was injected through the stopper onto the filter paper, and 0.1ml of 70% perchloric acid was injected into the saline. The samples were shaken for 1.5h, and the filters were removed and counted for 14C activity in 10ml of Ecolume (ICN) in a Tracor

Analytic Betatrac liquid scintillation counter. Counts were converted to total CO₂ production rates (μmol g glass scintillation vial, as in prior tissue h glass scintillation vial, as in prior) using the specific activity of glucose in the flask (i.e. disintsmin glass scintillation vial, as in prior of [14C]glucose added divided by 1 μmol total glucose).

For measurements of enzyme activities, glands were homogenized in their cryotubes in 0.5ml of 50mmol l glass scintillation vial, as in prior Hepes, pH7.5, using a Brinkman polytron. Samples were then centrifuged for 1min at 13000 g in a Fisher 235B microcentrifuge in a 4°C room. 10–50 μl of these supernatants (or 1:10 dilutions in the case of LDH) was used directly in enzyme assays. The following assays were performed, initially omitting the last listed item in each assay as a control. All assays were performed in a 1ml volume of Hepes, pH7.5, at 24±0.2°C and the change in absorbance was followed using an LKB Ultrospec (connected to a Kipp and Zonen chart recorder) at 340nm [micromolar extinction coefficient for NAD(P)H=6.22] unless noted.

Glucose-6-phosphate dehydrogenase (E.C. 1.1.1.49, G6PDH): 0.4mmol l⁻¹ NADP⁺, 7 mmol l⁻¹ MgCl₂, 1mmol l⁻¹ glucose 6-phosphate.

6-Phosphogluconate dehydrogenase (E.C. 1.1.1.44, 6PGDH): 0.8mmol l⁻¹ NADP⁺, 20mmol l⁻¹ MgCl₂, 3mmol l⁻¹ 6-phosphogluconate.

Transaldolase (E.C. 2.2.1.2, TA): 0.12mmol l⁻¹ NADH, 10mmol l⁻¹ EDTA, 0.8mmol l⁻¹ erythrose 4-phosphate, 0.6 units of alpha-glycerophosphate dehydrogenase, 9 units of triosephosphate isomerase, 2.7mmol l⁻¹ fructose 6-phosphate.

Transketolase (E.C. 2.2.1.1, TK): 20mmol l⁻¹ K₂HPO₄ (in addition to Hepes), 100mmol l⁻¹ KCl, 5mmol l⁻¹ MgCl₂, 0.12mmol l-1 NADH, 0.2mmol l⁻¹ thiaminepyrophosphate (=cocarboxylase), 0.6mmol l⁻¹ erythrose 4-phosphate, 0.6 units of alpha-glycerophosphate dehydrogenase, 9 units of triosephosphate isomerase, 1.0mmol l⁻¹ xylulose 5-phosphate.

Phosphoglucose isomerase (E.C. 5.3.1.9, PGI): 0.39mmol l⁻¹ NADP⁺, 6.8mmol l⁻¹ MgCl₂, 0.5 units of glucose-6-phosphate dehydrogenase, 1.4mmol l⁻¹ fructose 6-phosphate.

Phosphofructokinase (E.C. 2.7.1.11, PFK): 0.12mmol l⁻¹ NADH, 10mmol l⁻¹ MgCl₂, 50mmol l⁻¹ KCl, 2mmol l⁻¹ ATP, 1 unit of alpha-glycerophosphate dehydrogenase, 5 units of aldolase, 5 units of triosephosphate isomerase, 5mmol l⁻¹ fructose 6-phosphate.

Glyceraldehyde-3-phosphate dehydrogenase (E.C. 1.2.1.12, G3PDH): 0.12mmol l-1 NADH, 2mmol l⁻¹ MgSO₄, 0.9mmol l⁻¹ EDTA, 1mmol l-1 ATP, 10 units of phosphoglycerate kinase, 6mmol l⁻¹ glycerate 3-phosphate.

Lactate dehydrogenase (E.C. 1.1.1.2, LDH): 0.12mmol l⁻¹ NADH, 0.5mmol l⁻¹ pyruvate.

Malate dehydrogenase (E.C. 1.1.1.37, MDH): 0.12mmol l⁻¹ NADH, 0.5mmol l⁻¹ oxaloacetate.

Citrate synthase (E.C. 4.1.3.7, CS): pH8.0, 0.1mmol l⁻¹ 5,5 -dithiobis-(2-nitrobenzoic acid), 0.3mmol l⁻¹ acetylcoenzyme A, 0.5mmol l⁻¹ oxaloacetate. (Read at 412nm, micromolar extinction coefficient=13.6.)

All chemicals were reagent grade and biochemicals were purchased from Sigma (St Louis, MO).

Non-bicarbonate buffer capacity in toadfish gas gland was −105.4±9.4mmol pHunit⁻¹ kg⁻¹ wetmass (mean ± S.E., N=6). Compared to the buffer capacity of other tissues estimated in a similar fashion, that of the toadfish gas gland is 2–3 times that of most teleost skeletal muscle and is of comparable magnitude to that of tuna skeletal muscle (Castellini and Somero, 1981). It is of interest to note, however, that, despite the comparable buffer capacity, toadfish gas gland LDH activity is only one-tenth that reported for tuna muscle, indicating that the gas gland’s ability to produce copious amounts of lactate may be facilitated by an enhanced capacity to buffer the acidic end-products. This enhanced ability of the gas gland to buffer acid, and presumably to maintain intracellular pH despite continuous acid production, would explain the observation by Pelster and Scheid (1991) that the glycolytic enzymes of eel gas gland show no special adaptation to function in an acidic milieu.

Enzyme activities are listed by metabolic pathway in Table 1 and they generally agree well with prior studies (Boström et al. 1972; Ewart and Driedzic, 1990; Pelster and Scheid, 1991). There are substantial activities of two of the four enzymes of the pentose phosphate pathway (G6PDH and 6PGDH), but much lower activities of the remaining two enzymes (TA and TK). These low activities suggest that the shunt is not simply a bypass, but yields pentose sugars for general RNA synthesis, without rejoining glycolysis at the level of glyceraldehyde 3-phosphate to any appreciable extent. The gas gland has a substantial glycolytic activity, as indicated by activity of the enzymes PGI, PFK and G3PDH, but the activity appears to be largely anaerobic as LDH values are as high as those found in toadfish skeletal muscle (Walsh et al. 1989), and MDH and CS values are rather low. Interestingly, LDH in this tissue appears to be able to use NADPH as a cofactor at rates comparable to measured rates of shunt enzyme activities (Table 1). The NADPH used in this study was checked for NADH contamination by using purified rabbit M₄ LDH and no activity was noted with NADPH. However, given that the gas gland enzyme assays were performed with crude homogenates, the possibility of a novel form of LDH should be considered as preliminary.

Overall rates of glucose oxidation are comparable to those in prior studies with hepatocytes (Walsh et al. 1985), and the ratio of CO₂ production from C1-versus C6-labelled glucose is significantly greater than 1 (Table 2), indicating substantial flux through the pentose phosphate pathway. A typical value for this ratio in a teleost liver is 11.2 (Walsh et al. 1985), which is considered a relatively high pentose phosphate pathway activity. The high pentose phosphate pathway activity measured directly corroborates the previous suggestion by Pelster et al. (1989) of a highly active pentose phosphate pathway in teleost gas gland. Generation of CO₂ from C1-labelled glucose and the ratio of CO₂ production from C1-labelled versus C6-labelled glucose virtually double under hyperoxic conditions (Table 2), indicating an increase in flux through the pentose phosphate pathway. This observation strongly suggests that an additional role for this pathway in the teleost swim bladder is one of assisting in the protection of tissues from damage by oxygen radicals. Thus, further study of the gas secretion properties of the gas gland must continue to include as part of their experimental design careful control of oxygen partial pressures. Additionally, comparative studies which address both aspects of gas gland metabolism are strongly indicated.

1. Voltage and current-clamp steps were added to a sum of sine waves to measure the tetrodotoxin-insensitive membrane properties of neurons in the intact lamprey spinal cord. A systems analysis in the frequency domain was carried out on two types of cells that have very different morphologies in order to investigate the structural dependence of their electrophysiological properties. The method explicitly takes into account the geometrical shapes of (i) nearly spherical dorsal cells with one or two processes and (ii) motoneurons and interneurons that have branched dendritic structures. Impedance functions were analysed to obtain the cable properties of these in situ neurons. These measurements show that branched neurons are not isopotential and, therefore, a conventional voltage-clamp analysis is not valid.

2. The electrophysiological data from branched neurons were curve-fitted with a lumped soma-equivalent cylinder model consisting of eight equal compartments coupled to an isopotential cell body to obtain membrane parameters for both passive and active properties. The analysis provides a quantitative description of both the passive electrical properties imposed by the geometrical structure of neurons and the voltage-dependent ionic conductances determined by ion channel kinetics. The model fitting of dorsal cells was dominated by a one-compartment resistance and capacitance in parallel (RC) corresponding to the spherical, non-branched shape of these cells. Branched neurons required a model that contained both an RC compartment and a cable that reflected the structure of the cells. At rest, the electrotonic length of the cable was about two. Uniformly distributed voltage-dependent ionic conductance sites were adequate to describe the data at different membrane potentials.

3. The frequency domain admittance method in conjunction with a step voltage clamp was used to control and measure the oscillatory behavior induced by N-methyl-D-aspartate (NMDA) on lamprey spinal cord neurons. Voltage-clamp currents and impedance functions were measured at different membrane potentials. The impedance functions had a voltage-dependent resonance and phase shift characteristic of a negative conductance. These measurements provide a quantitative analysis of the conductances induced by NMDA in central neurons of the lamprey spinal cord and directly establish the basis of the non-linear oscillatory behavior previously observed in the presence of NMDA. NMDA was shown specifically to activate a negative and a positive conductance, both of which were markedly affected by the membrane potential. It is shown that the net current in the presence of NMDA must be considered as the algebraic sum of currents in opposite directions.(ABSTRACT TRUNCATED AT 400 WORDS)

We thank Jimbo Luznar for providing the toadfish and Dr T. P. Mommsen for advice on the TA and TK assays. This research was supported by an NSF grant to P.J.W. (IBN-9118819) and an NSERC grant to C.L.M.