Depression of Aerobic Metabolism and Intracellular pH by Hypercapnia in Land Snails, Otala Lactea

Studies of the physiology and biochemistry of metabolic arrest suggest that hypoxia and changes of intracellular pH are possible factors governing reversible transitions between activity and dormancy in many animals (Hochachka, 1988; Busa & Nuccitelli, 1984; Hochachka & Guppy, 1987). Pulmonate land snails provide an interesting opportunity to investigate both these mechanisms. Dormancy in Otala lactea (Helicidae) is characterized by reduced aerobic metabolism and by simultaneous hypoxia, hypercapnia and respiratory acidosis brought about by hypoventilation of the lung. Mean declines to about 15% of that of resting, non-dormant individuals (Barnhart & McMahon, 1987). in the lung often falls below 10 mmHg, compared with 137 mmHg in active snails, whereas reaches 100 mmHg in dormancy, compared with 12 mmHg in active snails. Mean haemolymph pH falls 0·4 units as a result of hypercapnia and up to 0·5 units further during prolonged dormancy as [HCO₃⁻] decreases (Barnhart, 1986a,b).

Cycles of CO₂ accumulation and release coincide with fluctuations of in dormant Otala, and is inversely related to whole-body CO₂ content (Barnhart & McMahon, 1987). Both the broad correlation between hypercapnia and dormancy and the short-term correlation between body CO₂ content and suggest that CO₂ accumulation and/or resultant acidosis might be factors in the control of metabolic rate during dormancy. However, low values in the lung and haemolymph suggest that might be oxygen-limited during dormancy. To test these possibilities, the present study examines the effects of acute experimental hypercapnia and hypoxia on the oxygen consumption of active Otala. The effect of hypercapnia on whole-body intracellular pH (pHi) was also measured because intracellular acidification is a possible mechanism by which CO₂ could affect metabolism.

Adult Otala were collected at Playa del Rey in Los Angeles County, California, and maintained in the laboratory as described previously (Barnhart & McMahon, 1987). The individuals used had a mean whole mass of 8·44 ± 0·84g (S.D.).

O₂ uptake and CO₂ release of individuals were measured over successive 1-h intervals by closed-chamber respirometry, using 60 ml syringes as chambers (Barnhart, 1986b). For measurements in the absence of CO₂, an open vial holding a folded wick and 1 ml of 5% KOH was placed in the chamber to absorb respiratory CO₂. For measurements in the presence of CO₂ the KOH was replaced with distilled water. Temperature was 20–23°C.

After a brief period (2–7 days) of dormancy, snails were placed on wet paper towelling for 2h to induce arousal and permit hydration, and then placed in the respirometry chambers for 1h of acclimation. To measure gas exchange, the chambers were flushed with mixtures of air and CO₂ or N₂ prepared by Wösthoff gas-mixing pumps, closed for 1 h, and then placed in an infusion pump to expel the gas successively through a dessicant tube, an Applied Electrochemistry S3A electrochemical O₂ analyser and a CD3A infrared CO₂ analyser. Oxygen and CO₂ fractions were measured to the nearest 0·01% volume. The analysers were calibrated using nitrogen, dry CO₂-free air and a 5% CO₂ standard analysed with a Scholander device. Change in oxygen fraction in the respirometers was typically 1–2% of volume. Gas exchange was calculated according to Vleck (1987).

The effect of ambient on was examined in four experiments which ranged in duration from 5 to 14h. In each experiment, 24 individuals were distributed randomly among a control (zero CO₂) and several CO₂ treatment groups. of all individuals was first measured in the absence of CO₂. The treatment groups were subsequently exposed to ambient between 13 and 98 mmHg.

The relationship between and was measured in two experiments, one in the absence of CO₂ and the other at 66 mmHg ⁠. of each individual was determined initially in air and afterwards at one of six levels of between 14 and 160 mmHg.

The effects of hypercapnia on pH of venous haemolymph (pHe) and of intracellular fluids (pHi) were measured in individuals which had previously been catheterized to permit marker injection and serial sampling of haemolymph. pHi was estimated with the labelled weak acid distribution technique (Waddell & Butler, 1959), using 5,5-dimethyl[¹⁴C]oxazolidine-2,4-dione (DMO) as the pH indicator and polyethylene glycol[l,2³H] (PEG) as the extracellular fluid volume marker (isotopes from New England Nuclear). Catheters consisted of 21 gauge hypodermic infusion needles with attached tubing (Deseret Company). The needle was lightly coated with cyanocrylate glue just behind the tip, then inserted through a hole in the shell into the visceral sinus. If no leakage occurred around the insertion, the needle was sealed to the shell and supported with epoxy paste. A 1 cm length of tubing was left attached to the needle, and was reversibly closed by a crimp secured by a band of larger tubing. Animals were placed on moist paper towelling and used within 1–2 days of preparation.

The catheterized snails were placed in plastic bags for exposure to gas mixtures. The bags were flushed continuously with humidified mixtures of CO₂ and air to maintain 0, 46 or 98 mmHg ⁠. After waiting 3h to permit equilibration with ambient CO₂, each snail was injected with 0·15 μCi of DMO and 0·20μCi of PEG in 30μl of 200mmoll⁻¹ NaCl, followed by 100μl of snail haemolymph to flush the catheter. Haemolymph samples (about 60μl each) were collected at hourly intervals thereafter for 6h. The catheter dead-space volume was discarded. The next 40 μl was transferred anaerobically to a Radiometer capillary pH electrode for determination of pHe. Subsamples (10μl) were analysed for isotope activities (see below).

Haemolymph was estimated from pH using the in vitro relationship between pH and (Astrup method). Haemolymph from each individual was equilibrated at 46 and 98mmHg in a microtonometer and pH was measured. The results were plotted, assuming a linear relationship between pH and log (Barnhart, 1986a). In vivo pH was compared with these plots to estimate in vivo⁠.

Isotope activities in haemolymph (disintsmin⁻¹ ml⁻¹) were measured using a dual-channel liquid scintillation counter with external standard (Tracor Mark III). Haemolymph was digested in NCS tissue solubilizer and counted in OCS (Amersham), with correction for counting efficiencies in the calculation of activity. Extracellular volume was calculated from PEG dilution, extrapolating the exponential decline of haemolymph PEG activity to the time of injection. Extracellular water content (ECW, g) was calculated from extracellular volume assuming haemolymph water content to be 1gml⁻¹. Total body water (TBW, g) was measured as the difference between wet and dry body mass, and intracellular water (ICW, g) as the difference between TBW and ECW. pHi was calculated for the total intracellular space using the equation:

(Waddell & Butler, 1959), assuming pK_DMO = 6-27 (Boron & Roos, 1976). [DMO]_i was calculated as follows:

where DMO_inj(disintsmin⁻¹) is the total DMO injected.

decreased with time in snails respiring in the absence of CO₂ (see Fig. 2). Typically, declined gradually over 6–8 h to about 70% of the initial level, then remained stable for the remainder of the measurement periods (up to 14 h).

was rapidly depressed by ambient hypercapnia, and the reduction of was greater at higher ⁠. The relationship between and appears to be nonlinear, with maximum slope at roughly 40–60 mmHg (Fig. 1). The reduction of relative to the control group was maximal within the first hour of exposure (Fig. 2), although equilibration with altered ambient requires more than 2 h (Fig. 3, see below). When CO₂ was removed, increased within 1 h to a level similar to or somewhat higher than that of the control group (Fig. 2).

The behaviour of each animal was noted at hourly intervals. All individuals remained at least partly extended from the shell during respirometry. Crawling was infrequent in all groups (2% of observations). Over time, an increasing proportion of the snails partly retracted into the shell; this occurred sooner in low CO₂ than in high CO₂. In the experiment illustrated in Fig. 2, half or more of the snails in 0mmHg were partly retracted after 2h, whereas this occurred only after 7 and 11 h in 46 and 98 mmHg ⁠, respectively.

The time course of change of internal during ambient hypercapnia can be inferred from the respiratory exchange ratio, R (Fig. 3). R was low or negative immediately following an increase of ambient ⁠, which may be attributed to the accumulation of CO₂ in the body fluids as increased. R increased rapidly and became positive within 2h, which shows that of the body fluids rose rapidly and exceeded ambient within this period. R eventually rose beyond the control level (Fig. 3). The overshoot was greater at higher ⁠, and may have been due to an increase of haemolymph [HCO₃⁻] during hypercapnia (Burton, 1976; Barnhart, 1986a). Addition of HCO₃⁻ to the body fluids shifts the CO₂-carbonic acid-bicarbonate equilibrium and should therefore increase CO₂ loss and elevate R.

Mixing of the isotopes was judged to be complete within 3h of injection; after this period the mean [DMO]_e was stable, although PEG levels declined exponentially, as indicated by the linearity of plots of the logarithm of activity (disints min⁻¹) vs time from 3 to 6h after injection. These plots were extrapolated to time zero to estimate extracellular fluid volumes for each individual. Mean extracellular water as a percentage of total body water was 51·1 ± 6·74% (S.D., N = 15).

Between 3 and 6h after injection the mean values of [DMO]_e and pHe in the three groups of snails did not change significantly. Therefore, the [DMO]_e and pHe results were averaged over this period for each individual and used to calculate a single estimate of pHi (from equations 1, 2) and (from the Astrup line) for each (Fig. 4). Both pHe and pHi declined similarly with increase of haemolymph ⁠. Linear regressions of pHe and pHi on the logarithm of haemolymph yielded the following equations:

The regression coefficients for pHe and pHi did not differ significantly. Mean pHi was lower than pHe by 0·12 units; this difference was significant (analysis of covariance, critical P = 0·05).

was independent of ambient above about 90 mmHg in the absence of CO₂, and independent above about 50 mmHg when was lowered 45% by the presence of CO₂ (Fig. 5). Below these levels, varied with The relationship between ambient and -limited is approximately described by the following regression equation (see Fig. 5):

The y-intercept did not differ significantly from zero (5·1 mmHg; P = 0·2); consequently, the regression was forced through zero. This equation was used to predict the limit of as a function of (see Discussion).

Previous work has shown that CO₂ retention is correlated with lowered in dormant Otala (Barnhart & McMahon, 1987). The present results show that the oxygen consumption of active Otala is rapidly and reversibly depressed by imposed hypercapnia. Both results appear to be consistent with the hypothesis that hypercapnia resulting from hypoventilation reduces metabolic rate during dormancy. However, of hypercapnic active snails was not depressed to the same degree as during dormancy. The highest ambient tested lowered mean by only 63% (Fig. 1) compared with the 85% reduction during dormancy (Barnhart & McMahon, 1987). In absolute terms, the mean in 98mmHg ambient (37 μl g⁻¹ h⁻¹) was more than double the mean dormant and six times the minimum dormant observed previously (13·7 and 5·6μl g⁻¹h⁻¹, respectively; Barnhart & McMahon, 1987). Thus, it is still uncertain whether the effects of hypercapnia limit during true dormancy. Moreover, hypercapnia did not hasten withdrawal into the shell or induce epiphragm formation, events which normally accompany entry into dormancy.

A possible mechanism for depression of by CO₂ in Otala is through interference with O₂ delivery to the tissues, either by depressing circulation (heart rate decreases in hypercapnic individuals) or by the reversed Bohr effect of haemocyanin (Barnhart, 1986b) which might reduce at the tissues. Either effect might conceivably cause oxygen limitation of during hypercapnia. However, if hypercapnia inhibited oxygen transport, oxygen-limited should be lower at similar during hypercapnia due to reduced conductance for oxygen (Herreid, 1980). This is not the case (Fig. 5), and the results therefore argue that hypercapnia does not interfere with oxygen transport.

Change of intracellular pH is one of several possible mechanisms by which hypercapnia could initiate metabolic effects (Walsh et al. 1988). The present results show that intracellular acidosis must be considered as a potential mechanism for the observed effect of CO₂ on in Otala. The whole-body pHi measurement leaves open the question of the exact values of pHi in different tissues. These values were not determined in the present study because of uncertainty regarding the fate of the extracellular volume marker (PEG) as it left the haemolymph. The similar extracellular and intracellular acidosis during hypercapnia is curious because of the presumably higher buffering capacity of the intracellular space (Burton, 1976). Similar results were, however, obtained in the bivalve Mytilus (Lindinger et al. 1984).

Although intracellular pH has not been measured in dormant Otala, Rees & Hand (1987) report that pHe and whole-body pHi decline by similar increments during dormancy in the land pulmonate Oreohelix. It appears that pHi may not only fall but also fluctuate over a substantial range during dormancy due to changes in ⁠. CO₂ release of dormant Otala is periodic (Barnhart & McMahon, 1987) and haemolymph ranges between 25 and 100 mmHg (Barnhart, 1986b). Comparison of this range with Fig. 4 suggests that pHi fluctuates over about 0·4 units due to changes of during dormancy. These changes of pH may occur rapidly, particularly during CO₂ release. Measurement of pHi during dormancy using DMO thus presents special difficulties. It should be noted in this regard that dormant pulmonates are sensitive to disturbances associated with handling (Machin, 1975; Herreid & Rokitka, 1976; Herreid, 1977). The manipulations required by the present application of the DMO method for measuring pHi induce hyperventilation, release of CO₂ and increase of in dormant Otala (M. C. Barnhart, unpublished results).

Although the biochemical mechanisms that reduce the rate of energy utilization during dormancy and other hypometabolic states are not yet understood, considerable evidence indicates that glycolysis is suppressed (Hochachka & Guppy, 1987). Intracellular acidosis appears to suppress glycolysis in hibernating mammals (see review by Malan, 1986) and in the brine shrimp, Artemia, during metabolic arrest (Busa & Nuccitelli, 1984; Carpenter & Hand, 1986; Hand & Carpenter, 1986). Among molluscs, acid pH reduces phosphofructokinase activity in the adductor muscle of Mytilus (Ebberink, 1982), and acidosis has been implicated in reduction of phosphofructokinase activity and glycolysis during anoxia in the whelk Busy con (Ellington, 1983). Pyruvate kinase from Otala shows decreased affinity for phosphoenolpyruvate and ADP, increased inhibition by alanine and ATP, and decreased activation by fructose 1,6-diphosphate in vitro when pH is lowered from 7·0 to 6·5 (J. Fields, personal communication).

Other mechanisms which have been proposed to suppress glycolysis during hypometabolism involve neither CO₂ nor change in pH. These include phosphorylation of regulatory enzymes, dissociation of enzyme complexes to produce less active soluble forms, and decreased levels of fructose 2,6-bisphosphate, an activator of phosphofructokinase (Storey, 1988). Recent studies show that fructose 2,6-bisphosphate levels fall by 86–93% in Otala during 4 days of dormancy and subcellular binding of glycolytic enzymes decreases (K. Storey, personal communication). Thus, evidence suggests that both pH-dependent and pH-independent mechanisms may suppress glycolysis in Otala during dormancy.

in the lung and haemolymph of dormant Otala is on average only about one-third that of active snails, and is often less than 10 mmHg owing to hypoventilation (Barnhart, 1986b). Might be oxygen-limited during dormancy? One approach to this question is to compare V_O2 in dormancy with the limit of predicted at the low levels of typical of dormancy. As a first approximation, lung can be substituted for ambient in equation 5 [lung and ambient do not differ greatly in active Otala (Barnhart, 1986b)]. The mean of the lowest 10% of lung values observed in dormant Otala was 4·5 mmHg (data of Barnhart, 1986b). The predicted limit of at this is about 5 μl g⁻¹ h⁻¹ (from equation 5). This value is similar to the minimum sustained ⁠, observed during continuous respirometry (5·6 ± 0·57μmolg⁻¹h⁻¹, S.E.M.; Barnhart & McMahon, 1987). Thus it appears that lung sometimes approaches a critical level. However, the data do not indicate that normally limits during dormancy. Mean lung during dormancy is 38 mmHg (Barnhart, 1986a), which corresponds to a limit of 40μlg⁻¹h⁻¹, much higher than the mean during dormancy (14μlg⁻¹ h⁻¹; Barnhart & McMahon, 1987).

In conclusion, the effects of acute hypercapnia and hypoxia on oxygen consumption of active Otala lactea indicate that hypercapnia, but not hypoxia, explains much of the reduction of aerobic metabolism during dormancy. However, of acutely hypercapnic active snails is not depressed to the lowest levels typical of dormancy, suggesting that other factors also contribute to metabolic de-pression.

This work was supported by a postdoctoral fellowship from the Alberta Heritage Foundation for Medical Research, and by NSERC grant A5762.