Intracellular pH in the Toad Bufo Marinus Following Hypercapnia

During hypercapnia, amphibians and reptiles regulate intracellular and extracellular pH relatively poorly compared to aquatic species (Boutilier et al. 1979; Toews and Heisler, 1982; Boutilier and Heisler, 1988). Furthermore, it has been suggested that pH regulation in these vertebrates may be limited by high resting levels of bicarbonate (Boutilier and Heisler, 1988), dictated by the physical characteristics of gas exchange in air. Thus, the transition from water to land, with an attendant increase in arterial $PCO2$⁠, may have placed constraints on the value of bicarbonate as a buffer: this raises questions about how lower terrestrial vertebrates regulate pH. Unfortunately, little information is available to enable us to formulate general patterns. For example, almost nothing is known about non-bicarbonate buffers in lower vertebrates. Information about pHi in some tissues, e.g. brain, is limited to aquatic vertebrates (Wood et al. 1990).

The present study examines pHi in brain and liver during short-term hypercapnia. In addition, we measured in vitro non-bicarbonate buffer curves to provide a better picture of the degree to which in vivo changes may represent compensation. Bufo marinus was selected for this initial study because of the wealth of information about acid-base regulation in this species (Boutilier et al. 1979; Toews and Heisler, 1982; Boutilier and Heisler, 1988).

Bufo marinus L., average body mass 210.9± 12.5 g (±S.E. range 167–306g, N=40), were obtained from Blue Spruce Biological Supply (Castle Rock, CO). Toads were housed in acrylic aquaria (2m×0.35 m×0.3m) containing 20% Holtfreter solution, prepared with deionized water. One end of each container was elevated to provide toads with wet and dry areas. Containers were washed daily with chlorinated tap water followed by deionized water. Toads were fed crickets daily. Temperature of air and water averaged 25±1.0°C.

Randomly selected individuals were anesthetized with a 0.75% aminobenzoic acid ethyl ester (MS-222) solution, pH adjusted to 7.7 with NaHCO₃. The right femoral artery was then cannulated (PE-50, Clay Adams, Parsippany NJ). Catheters were filled with heparinized amphibian Ringer’s solution, and secured with nylon sutures and tissue glue (Nexaband, CRX Medical, Raleigh NC).

After a minimum of 24h recovery, toads were placed in a 3.51 plastic chamber containing 200 ml of a 0.3 mmol I⁻¹ sodium bicarbonate solution. Gas lines and cannulae were passed through ports in the top of the chamber. The chamber was covered with opaque cloth to eliminate visual contact. In this way, blood samples were withdrawn from unanesthetized and undisturbed animals. Water and air temperatures averaged 25±0.5°C.

Intracellular pH was determined using the transmembrane distribution of [¹⁴C]DM0 (5,5-dimethyloxazolidine-2,4-dione-2-[¹⁴C]; New England Nuclear) with [³H]inulin (Amersham) to mark extracellular space (Waddell and Butler, 1959; Nestler, 1990). Inulin purity was 97% as checked with thin layer chromatography [solvent n-butanol: acetone: 95% ethanol: H₂O (5:5:2:1)]. Although DMO yields an average pHi, the values obtained with this method are similar to those measured with microelectrodes (Roos and Boron 1981). In addition, we (G. K. Snyder and J. I. Shapiro, unpublished observations) have recently determined pHi of limb skeletal muscle and brain using ³¹P-magnetic resonance spectroscopy with the techniques of Shapiro et al. (1989) and found that the values for both tissues were within ±0.02 pH units of, and not significantly different from, average values obtained with [¹⁴C]DMO. An equilibration time of 12 h for the radiotracers was established in a preliminary study.

Toads were injected with 300μ1 of a solution containing 15 μCi of DMO, 45 μCA of inulin, 6.9μmol of ‘cold’ DMO and 1.25μmol of ‘cold’ inulin into the femoral artery 12 h after they had been placed in the chamber. The chambers were flushed with room air (20.95% O₂: 0.04% CO₂) at 500ml min⁻¹ for 11 h. Air inflow for the ‘hypercapnic’ animals (N=9) was then switched to 20.95% O₂: 5% CO₂ and perfusion continued for an additional hour. Air inflow for ‘control’ animals (N=7) was continued as room air for the final hour.

After gas perfusion, blood (150–300μl) was drawn from the cannula to determine pH and $PCO2$ (Radiometer BMS3 Mk2 Blood Micro System, at 25°C). An additional 150μl sample was centrifuged, blood hematocrit recorded, and the plasma used for radioactive counts. Animals were then injected with pentobarbi-tol-KCl and immediately double pithed to prevent struggling. Heart, liver and skeletal muscle (left thigh muscles) were removed within 30–60 s. Removal of the brain required another 1–2 min. During this time, radiotracer washout was prevented by stopping the circulation prior to tissue sampling. All tissues were blotted, weighed and lyophilized to constant mass (24–36h). Perchloric acid extracts were prepared for liquid scintillation counting.

Tissue pHi was calculated using the equations of Wood and Cameron (1985). Liver accumulated inulin and, thus, artificially high (50–80%) extracellular water (ECW) values were obtained. Therefore, an ECW value of 15% was used for liver (Nestler, 1990). An opposite trend, low ECW (2–6%), was noted in brain, caused by the low permeability of the blood-brain barrier (Zubrod and Rail, 1959). Since values for toad brain ECW are not available, we measured ECW of brains from eight normocapnic toads using the methods of Bradbury et al. (1968). Brains were removed, the meninges teased away to allow free distribution of inulin, and the brain incubated for 1h in 5ml of a pH7.25 solution containing 2.0μCi of [³H]inulin. The incubation solution consisted of 90 mmol l⁻¹ NaCl, 25 mmol l⁻¹ NaHCO₃, 2.0 mmol l⁻¹ KC1, 3.0 mmol⁻¹ KH₂PO₄, 0.9 mmol l⁻¹ CaCl₂, 1.2mmoll⁻¹ MgCl₂, 1.8mmoll⁻¹ Na₂SO₄ and 13.8mmoll⁻¹ glucose. During incubation, the solution was bubbled with 95% O₂: 5% CO₂.

Bicarbonate concentrations were calculated with the Henderson-Hasselbalch equation. Values used for pK’ (6.05) and CO₂ solubility (0.248 mmol kPa⁻¹) were those reported for Bufo marinus at 25 °C (Boutilier et al. 1979).

Non-bicarbonate buffering capacity was determined using the methods of Castellini and Somero (1981). Toads (N=7) were cannulated and housed as described above for the control animals. After air-perfusion, a blood sample was taken and the animal was killed. Tissues were rapidly (30–60 s) removed, frozen in liquid nitrogen and stored at –85 °C. Frozen tissue samples were homogenized 1:20 (w/v) in 0.6% NaCl and titrated at 25°C with 0.2 or 0.02 mol l⁻¹ NaOH, between pH 6 and 7. Changes in pH, monitored with a Radiometer PHM64 pH meter and GK2401C combination electrode calibrated with precision buffers, were used to calculate intracellular non-bicarbonate buffer capacity with the equations of Heisler and Piiper (1971).

Statistical analyses were completed using SPSS-X. Differences in values between control and hypercapnic animals were examined using two-tailed t-tests and analysis of covariance (ANCOVA) with a significance level of 0.05. All data are expressed as mean±s.E.

Extracellular water volumes were statistically similar for all tissues. However, liver total water content was significantly lower than that of other tissues. One hour of hypercapnia did not significantly alter total tissue water content or extracellular water volume for any tissues (Table 1).

After 1h at 5% CO₂, blood $PCO2$ of Bufo marinus increased from 1.6±0.04 to 5.7±0.23kPa. Although the difference in $PCO2$ is highly significant, we found no evidence of a CO₂-stimulated hyperventilation, since the difference in $PCO2$ between arterial blood and inspired gas was roughly the same (1.5 kPa) for normocapnic and hypercapnic toads (e.g. at Boulder elevation $PCO2=84kPa×0.05=4.2kPa$⁠). After 1h of hypercapnia, plasma bicarbonate had increased significantly, from 23.9±1.0 to 27.0±1.8mequivl⁻¹, while arterial pH had decreased significantly, from 7.85±0.02 to 7.33±0.01. During this same time, brain pHi decreased from 7.16±0.02 to 6.72±0.05, similar to the decrease in pHi of skeletal muscle and heart (Fig. 1). Control pHi of liver, 7.44±0.14, in contrast, was significantly higher than in other tissues, and decreased the least following hypercapnia, to 7.25±0.05 (Fig. 1).

Tissue bicarbonate concentrations increased significantly following hypercapnia, to a greater extent than would be predicted from passive buffering, indicating compensation for the acidosis induced by hypercapnia. The increase in bicarbonate was greatest in liver (+14.4mequivl⁻¹) and least in brain (+2.0mequivl⁻¹).

When expressed as the difference between the observed pHi and the expected pHi at constant bicarbonate concentrations (Toews and Heisler, 1982), compensation in liver, 66%, is close to values for heart and skeletal muscle, while compensation in brain, 35%, is only slightly greater than in plasma, 25%.

Changes in pHi were uniformly less than changes in extracellular pH (pHe). In addition, pHi appeared to be best protected in the liver (pHi/pHe=0.37) and least protected in brain (relative change=0.83). Consequently, the in vivo non-bicarbonate buffer values were lowest in brain and highest in liver (Table 2). In addition, the in vivo non-bicarbonate buffer value for brain was significantly lower than the non-bicarbonate buffer curve measured in vitro (Table 2), indicating a net loss of base equivalents or a net gain of acid equivalents during hypercapnia. For liver, however, the apparent in vivo buffer value was significantly higher than the value measured in vitro, indicating a net gain of base equivalents or a net loss of acid equivalents in this tissue (Table 2).

Blood hematocrit after 1 h at 5% CO₂ increased significantly, from 23.9±1.5% (N=7) in control toads to 31.6±1.2% (N=9) in hypercapnic toads. Thus, although no corrections for plasma trapping were made, our results indicate that blood hematocrit may increase by as much as 32% during hypercapnia.

Exposing Bufo marinus to elevated ambient $PCO2$produces an immediate increase in lung ventilation followed by adaptation of the CO₂ ventilatory response (Boutilier et al. 1979). We did not observe a change in the difference between blood and ambient $PCO2$⁠. Thus, in our toads, hyperventilation was not a major factor following 1 h of acclimation; that is, adaptation of CO₂ sensitivity had occurred by the time blood samples were taken.

Control values of arterial pH, plasma bicarbonate and plasma buffers are all similar to those reported previously (Boutilier et al. 1979). In addition, control and hypercapnic pHi in skeletal muscle and heart are in the range of values reported for Bufo marinus (Toews and Heisler, 1982), other anuran amphibians (Malan et al. 1976) and fish (Wood et al. 1990). Brain and liver pHi values in amphibians have not been reported previously. However, our data for these tissues are consistent with pHi of the same tissues in mammals (Nestler, 1990).

We have examined compensation of hypercapnia by comparing in vivo buffer curves to those determined in vitro. Previous studies have shown that in vitro methods yield buffer values consistent with tissue protein concentrations (Zubrod and Rail, 1959; Wood et al. 1990) and similar to buffer values obtained using microelectrode techniques (Aickin and Thomas, 1977). Thus, although we cannot exclude the possibility that titration may overestimate buffer values in some tissues, it is likely that the lower in vivo than in vitro non-bicarbonate buffer values in all tissues except liver (Table 2) indicate a net accumulation of acid equivalents or a net loss of base equivalents during hypercapnia.

The relatively poor buffering of brain (Fig. 1, Table 2) is surprising, but similar to findings in fish (Wood et al. 1990) and mammals (Nestler, 1990). In contrast, the fact that brain pHi decreased to a greater extent than did pHi of other tissues is not consistent with findings from other vertebrates. In the skate, compensation of brain pHi was more rapid and more precise than in other tissues, in spite of the lower tissue buffer capacity (Wood et al. 1990).

The relatively poor buffering of brain cannot be attributed entirely to a lack of non-bicarbonate buffer capacity since the in vitro buffer value of brain is similar to those of heart and liver (Table 2). However, it is possible that secondary effects of elevated intracellular bicarbonate limit compensation in brain. Increased intracellular bicarbonate is followed by increased intracellular water. In the skate, this fluid shift appears to occur during rapid (<1 h) compensation, and persists for the first 24 h of hypercapnia (Wood et al. 1990). Further, intracellular fluid volume is eventually re-established by a reduction of [Cl⁻]₁ (Nishimura et al. 1989), which may interfere with electrophysiological mechanisms (Yoshizaki et al. 1989). Thus, it is possible that potential disruptions of normal brain function following loss of electrolytes may outweigh the benefit of rapid pH compensation.

The responses of liver to hypercapnia have not been studied previously, and the results presented here do not permit conclusions about the need for compensation specific to the liver. However, the presence of high concentrations of carbonic anhydrase in liver and its importance to normal hepatic tissue function are well established (Dodgson and Forster, 1986; Walsh and Milligan, 1989) and consistent with the relatively rapid accumulation of bicarbonate in this tissue. Furthermore, the toad bladder excretes H⁺ and NFL₄⁺ during hypercapnia (Tufts and Toews, 1986), which may be an important sink for the excess protons formed from bicarbonate production in liver.

This work was supported in part by National Science Foundation grant DCB881864 and Public Health Services grant RO1-HL32894 to G.K.S. The authors thank Kate Hallberg for assistance.