Pyruvate Kinase From the Land Snail Otala Lactea: Regulation by Reversible Phosphorylation During Estivation and Anoxia

Metabolic rate depression is a common strategy used by animals to survive periodic exposures to extreme environmental conditions that are unsuitable for normal life functions. An escape to a hypometabolic or dormant state is the frequent response to temperature extremes, aridity, lack of oxygen or lack of food and underlies such phenomena as estivation, hibernation, facultative anaerobiosis, anhydrobiosis and diapause (Hochachka and Guppy, 1987). The biochemical mechanisms that control metabolic arrest have been the subject of much recent work by our laboratory and we have found that the molecular controls serving anoxia-induced glycolytic rate depression (a reverse Pasteur effect) are conserved across a phylogenetically diverse group of facultative anaerobes including marine molluscs, goldfish and freshwater turtles (reviewed in Storey, 1988a). The same metabolic control principles could also apply to other forms of metabolic arrest.

Numerous species of terrestrial snails exploit seasonally arid environments by entering a state of estivation whenever water and food supplies disappear (Schmidt-Nielsen et al. 1971; Burky et al. 1972; Herreid, 1977; Horne, 1979; Umezurike and Iheanacho, 1983; Vorhaben et al. 1984). During estivation aerobic metabolic rate is reduced to 10 –30% of the resting rate when aroused, an adaptation that allows endogenous fuel reserves to be stretched to support months (or even years) of dormancy. The present study examines estivation in the land snail Otala lactea (Pulmonata, Helicidae). Physiological aspects of estivation have been well analyzed in this species. Estivation is characterized by hypoxia, hypercapnia, extracellular acidosis and a reduction of aerobic metabolic rate by as much as 85 % (Herreid, 1977; Barnhart, 1983, 1986a,b; Barnhart and McMahon, 1987, 1988). Carbohydrate reserves appear to be the primary fuel supporting metabolism in estivating snails (Umezurike and Iheanacho, 1983; Livingstone and de Zwaan, 1983; Cedeno-Leon, 1984) and, thus, control over glycolytic rate must play a central role in the transition to and from the dormant state.

One of the key mechanisms involved in metabolic depression in facultative anaerobes is the post-translational modification of regulatory enzymes via protein phosphorylation or dephosphorylation reactions to bring about rapid and readily reversible changes in the activity of regulatory enzymes (Storey, 1988a). This mechanism was first identified for pyruvate kinase (PK) in marine molluscs (Siebenaller, 1979; Holwerda et al. 1983; Plaxton and Storey, 1984a) but we now know that anoxia-induced covalent modification coordinates the activities of glycogen phosphorylase, 6-phosphofructo-l-kinase, 6-phosphofructo-2-kinase and PK to create a balanced depression of glycolytic rate (Storey, 1984, 1988a,b; L. Bosca and K. Storey, unpublished data). The present study examines the regulation of PK in O. lactea tissues. This species also has a substantial tolerance for anoxia [snails readily recovered after 4 days under a nitrogen gas atmosphere (Brooks and Storey, 1990a; this study)] and so we analyzed the effects of both estivation and anoxia on the properties of PK. The results show that in both states PK undergoes a stable modification that changes enzyme properties to those of a less active enzyme form. The changes in PK properties induced by either estivation or anoxia were virtually identical and could be mimicked by the actions of protein kinases on the enzyme in vitro. It is evident, then, that enzyme regulation via covalent modification may be a major control mechanism supporting metabolic rate depression during estivation.

All biochemicals and coupling enzymes were obtained from Sigma Chemical Co., Boehringer Mannheim Corp., or J. T. Baker Co. Snails, O. lactea, were provided by Dr M. C. Barnhart from an introduced population at Playa del Rey in Los Angeles County, California. In the laboratory, estivation is readily induced by the removal of water and food and, conversely, dormancy is broken by the reintroduction of these items. While being held in the laboratory, the snails were usually kept in their dormant state in a dry container and without food. Every 3 weeks they were aroused by the reintroduction of water (a light spraying plus damp paper towels) and the addition of food (cabbage sprinkled with chalk dust). Arousal usually lasted only a few days before the snails once again entered dormancy voluntarily.

All snails used for experiments were awakened from dormancy and given continuous access to food and regular spraying with water to ensure that they remained aroused for at least 2 days. Control snails (aroused and active) were sampled directly from this population. To impose anoxia, aroused snails were placed in ajar containing damp paper towels and small a piece of cabbage. The jars were then flushed with nitrogen gas (95 % N₂/5 % CO₂) and sealed with a screw-on lid and parafilm. A wire mesh barrier prevented the snails from climbing more than half-way up the jar. Under anoxic conditions the snails distended themselves out of their shells and pressed the foot out flat against the walls of the jar; they never ate under anoxia. To induce estivation, aroused snails were removed and placed in dry containers without food or paper toweling but with mesh lids to permit continuous aeration. In the dry environment the snails withdrew into their shells within about 2h. For the time course of dormancy and recovery, the length of dormancy was timed from the moment aroused snails (previously active for at least 2 days) were placed in the dry jar. After 22 days of constant dormancy, the length of recovery was timed from the moment that food and water were reintroduced, as described above. Because there was considerable variation in the rate at which individual snails egressed from their shells and resumed movement, we always chose from the experimental pool the four animals that were most active at each time point on the recovery time course.

To dissect snails the shell was broken open, the head end was severed from the remainder of the foot, and then foot, mantle and hepatopancreas were rapidly removed (within 1min). Tissues were immediately frozen in liquid nitrogen and stored at – 60°C until use. For the preparation of enzyme extracts, tissues were homogenized 1:5 (w/v) in homogenization buffer containing 100 mmol I⁻¹ NaF (to inhibit protein phosphatases), 5 mmol l⁻¹ EDTA, 5 mmol I⁻¹ EGTA (to inhibit protein kinases), 0.1 mmol I⁻¹ phenylmethylsulfonyl fluoride (a protease inhibitor), 30mmoll⁻¹ 2-mercaptoethanol, and 40% glycerol (v/v) (an enzyme stabilizer) in 50 mmol I⁻¹ imidazole-HCl (pH 7.0); a Polytron PT 10 or an Ultra-Turrax homogenizer was used. Homogenates were centrifuged for 20 min at 25 000g in a Sorvall RC-5B refrigerated centrifuge at 5 °C. The supernatant was removed and desalted by centrifugation (IEC benchtop centrifuge at top speed for 1min) through a 5 ml column of Sephadex G-25, as described by Helmerhorst and Stokes (1980). The column was equilibrated in 40 mmol E¹ imidazole-HCl (pH 7.0) with 5 mmol I⁻¹ EDTA, 15 mmol I⁻¹ 2-mercaptoethanol and 20% v/v glycerol. The filtrate was removed, stored on ice, and used for the analysis of enzyme kinetic properties.

PK activity was monitored by coupled enzyme assay following NADH utilization at 340 nm using a Gilford 240 recording spectrophotometer. Assay conditions for maximal PK activity in all three tissues were 50 mmol I⁻¹ imidazole-HCl (pH7.0) buffer, 3mmoll⁻¹ ADP, 3mmoll⁻¹ phosphoenolpyruvate (PEP), 50mmoH⁻¹ KC1, 5 mmol I⁻¹ MgCI₂, 0.2mmoll⁻¹ NADH and li.u.ml⁻¹ lactate dehydrogenase (previously dialysed when used for hepatopancreas PK assays) in a final cuvette volume of 1ml. All assays were initiated by the addition of enzyme preparation. I₅₀ values (the concentration of inhibitor producing a 50% decrease in enzyme rate) and K_a values (the activator concentration producing half-maximal activation) were determined at suboptimal PEP concentrations, 0.1 and 0.5 mmoll⁻¹, respectively. One unit of enzyme activity is defined as the amount of enzyme that catalyzes the utilization of 1 μ mol of PEP per min at 20°C.

Snails for these experiments were sampled from a second group of animals collected about 16 months after the animals that were used for all other experiments. The preparation of enzyme extracts from control versus estivating (4 weeks) animals followed the same protocol as that given above up to the point of centrifugation, except that tissues from freshly killed animals were used and tissues were homogenized in a 1:2.5 (w/v) ratio with buffer. After centrifugation, the supernatant was passed through a Sephadex G-25 spun-column (as described above) but with the column equilibrated in 40 mmol I⁻¹ imidazole-HCl (pH 7.0), 20% v/v glycerol, 10 mmol I⁻¹ potassium phosphate, 10 mmol I⁻¹ 2-mercaptoethanol and 0.1 mmol I⁻¹ EDTA. The filtrate was collected, divided into seven samples, and each sample was diluted 1:1 in a solution containing 40 mmol I⁻¹ imidazole-HCl (pH 7.0), 10 mmol I⁻¹ 2-mercaptoethanol and 20% v/v glycerol and one of the following to stimulate phosphorylation or déphosphorylation of the enzyme, (a) For cyclic-AMP-dependent phosphorylation; 20 mmol I⁻¹ MgCl₂, 40 mmol I⁻¹ NaF, 4 mmol I⁻¹ cyclic 3’5’-adenosine monophosphate (cyclic AMP), 2 mg ml⁻¹ cyclic-AMP-dependent protein kinase (from bovine heart, 2.0pmoll⁻¹ phosphate per μ g of protein phosphorylating activity) and 10 mmol I⁻¹ ATP. (b) For cyclic-GMP-dependent phosphorylation: 20 mmol I⁻¹ MgCl₂, 40 mmol I⁻¹ NaF, 8 mmol I⁻¹ cyclic 3’5’-guanosine monophosphate (cyclic GMP) and 10 mmol I⁻¹ ATP. (c) To stimulate phosphorylation by protein kinase C: 20 mmol I⁻¹ MgCl₂, 40 mmol I⁻¹ NaF, 2 mmol I⁻¹ CaCl₂, 20 μ gml⁻¹ phorbol 12-myristate 13-acetate, 80 μ gml⁻¹ brain extract (phospholipid source; Type 1, Sigma Chemical Co. no. 1502) and 10 mmol I⁻¹ ATP. (d) For protein dephosphorylation by alkaline phosphatases: 25 mmol I⁻¹ KC1, 5 mmol I⁻¹ EDTA, 15 mmol I⁻¹ MgCl₂ and 50i.u. alkaline phosphatase (grade 1, from calf intestine), (e) For protein dephosphorylation by alkaline and acid phosphatases: 25 mmol l ^{− 1} KC1, 5 mmol I⁻¹ EDTA, 15 mmol I⁻¹ MgCl₂, 50i.u. alkaline phosphatase (grade 1, from calf intestine), 2 mg ml⁻¹ acid phosphatase (type 1, from wheat germ, 0.46i.u.mg⁻¹ activity) and Imgml⁻¹ spermidine, (f) For stimulation of endogenous protein phosphatases: 25 mmol I⁻¹ KC1, 5 mmol I⁻¹ EDTA, 15 mmol I⁻¹ MgCl₂, and 1mgml⁻¹ spermidine (a protein phosphatase activator), (g) For the control: 40 mmol I⁻¹ NaF and 5 mmol I⁻¹ EDTA.

All samples were then incubated for 16 h at 30 °C for the foot muscle enzyme or 20 °C for the hepatopancreas enzyme (hepatopancreas PK was not stable at the higher incubation temperature). Samples were then passed through Sephadex G-25 spun-columns equilibrated in 45 mmol I⁻¹ imidazole-HCl (pH7.0), 0.5 mmol I⁻¹ EDTA, 10 mmol I⁻¹ mercaptoethanol and 20% v/v glycerol. The filtrates were used for subsequent kinetic analyses.

S_0.5 values (the substrate concentration producing half-maximal enzyme velocity), n_H (the Hill coefficient) and V_max (enzyme maximal velocity) values were determined by fitting the data to the Hill equation using a nonlinear least squares regression program for the computer. A modified Hill equation which included a Vo term (rate at zero activator) was used for determining K_a values. I₅₀ values were estimated from plots of reaction velocity vs [inhibitor]. Results are presented as means±s.E.M. Tests for significant differences between values used the Student’s t-test.

The kinetic properties of PK from O. lactea were analyzed in three tissues, foot muscle, mantle and hepatopancreas. In all tissues the enzyme underwent a stable modification (i.e. the modification was retained after gel filtration had removed low molecular weight metabolites) that changed enzyme properties during both estivation and anoxia.

The properties of PK from foot muscle of control (2 days aroused with food and water), estivating (22 days) and anoxic (45 h of nitrogen gas exposure) O. lactea are shown in Table 1. PK was modified during both estivation and anoxia; the changes induced by these states were very similar both qualitatively and quantitatively. In both situations, enzyme affinity for PEP decreased (S_0.5 increased by about 50 %) and inhibition by L-alanine and Mg-ATP increased (I₅₀ values decreased). PK from estivating snails also showed small, but significant, increases in S_0.5 ADP and in the activity of the enzyme in the tissue. Foot muscle PK was not activated by fructose-l,6-bisphosphate (F1,6P₂); levels up to 1 mmoll ⁻¹ F1,6P2 were tested, compared to physiological levels of less than 0.030 mmoll ⁻¹ (Churchill and Storey, 1989).

The properties of PK from O. lactea mantle are shown in Table 2. Overall, the kinetic properties of the mantle enzyme were very similar to those of foot muscle PK, particularly the properties of the enzyme from control (aroused) snails. Again, F1,6P₂ did not activate the enzyme. All the kinetic properties tested, with the exception of V_max (and n_H for S_0.5 ADP in estivation), were altered significantly during both estivation and anoxia. Affinity for both substrates decreased; S_0.5 for ADP increased about 25 % whereas S_0.5 PEP increased 1.8-fold in dormant and 2.4-fold in anoxic snails. I₅₀ for Mg-ATP decreased significantly under both experimental conditions but the change was again more pronounced in anoxia. The I₅₀ for L-alanine responded differently in the two experimental states, increasing by 74% in estivating animals but decreasing by 46% in anoxic snails.

The properties of PK from hepatopancreas of O. lactea are shown in Table 3. The properties of the enzyme in control (aroused) hepatopancreas were substantially different from those of the enzyme in foot or mantle. Enzyme affinity for PEP was much lower (S_0.5 about 10-fold higher) and inhibition by alanine was stronger (I₅₀ five-fold lower). In addition, the enzyme responded to F1,6P₂ activation with a K_a of only 0.004 mmol I⁻¹. All the properties of hepatopancreas PK were changed significantly in estivation and anoxia and the changes induced by either hypometabolic state were the same in both direction and magnitude.

Estivation or anoxia resulted in an increase in enzyme V_max, a decrease in S_0.5 PEP, increases in S_0.5 ADP, I₅₀ Mg-ATP and I₅₀ alanine, and a change from high sensitivity to F1,6P2 activation to no effect of F1,6P₂ at levels up to 5 mmol I⁻¹.

A number of other metabolites were tested for their effects on hepatopancreas PK (Table 3). Most had little effect on enzyme velocity, including phenylalanine, which is a well known inhibitor of mammalian liver PK. Proline (at 5 mmol I⁻¹) reduced the activity of PK from active snails by 45 % but had little effect on the enzyme from estivating or anoxic snails.

To analyze the progress of PK modification during estivation and subsequent arousal, selected properties of PK from foot and hepatopancreas were monitored over the course of 22 days estivation followed by 2 days of arousal. The parameters chosen were those that showed a large change between control and estivating states, namely the I₅₀ value for Mg-ATP for foot muscle PK and the S_0.5 for PEP for hepatopancreas PK (Tables 1 and 3). Fig. 1 shows the time course of changes in I₅₀ Mg-ATP of foot muscle PK. This parameter remained constant for the first 12 h after snails were moved to a dry container without food and water but dropped to the low value, characteristic of dormancy, within 48h. The I₅₀remained at this level for the subsequent 20 days of estivation. When food and water were reintroduced, however, the I₅₀ value for Mg-ATP of foot PK rebounded to control levels within 10 min. At about this time snails began to emerge from their shells and within 30 min they were actively crawling and feeding.

The time course of changes in S_0.5 PEP for hepatopancreas PK is similar during entry into dormancy. As Fig. 2 shows, S_0.5 remains high over the first 12 h but was reduced to a value characteristic of the estivating state within 48 h after placing snails into a dry environment. Changes in hepatopancreas PK during arousal, however, took a longer time than was seen for foot PK, occurring between 3 and 24 h after the reintroduction of food and water.

To determine whether the changes in PK kinetic properties occurring during estivation were the result of a post-translational modification of the enzyme via protein phosphorylation or dephosphorylation, PK from foot or hepatopancreas was incubated under various regimens to stimulate protein kinase or protein phosphatase actions on the enzyme. When PK from the foot muscle of control, active snails (2 days aroused) was incubated for 16 h in the presence of Mg-ATP, cyclic AMP and cyclic-AMP-dependent protein kinase, the I₅₀ value for alanine dropped to the value seen in either dormant or anoxic snails (Table 4). A similar significant decrease in I₅₀ was seen when the enzyme was incubated with cyclic GMP plus Mg-ATP to stimulate the activity of endogenous cyclic-GMP-dependent protein kinase or with Mg-ATP, Ca²⁺, phorbol 12-myristate 13-acetate and brain phospholipid fraction to stimulate protein kinase C. However, when the enzyme from active snails was incubated with alkaline phosphatase or with alkaline phosphatase plus acid phosphatase and the phosphatase stimulator spermidine, no change in the I₅₀ value was seen. PK from foot muscle of estivating (4 weeks) snails responded in the opposite direction. Phosphatase treatments significantly increased the I₅₀ value for alanine, restoring it to the value seen for PK from active snails. Stimulators of endogenous cyclic-GMP-dependent protein kinase or protein kinase C had no effect on foot muscle PK from dormant snails, but treatment with added cyclic-AMP-dependent protein kinase further lowered the 1₅₀ value of the enzyme.

Similar incubations were carried out with PK from hepatopancreas. When PK from the hepatopancreas of active snails was given treatments that promote protein kinase action, the enzyme showed a fourfold decrease in S_0.5 PEP, similar to the effect of estivation on the enzyme (Table 3). The effect was similar for PK treated with added cyclic-AMP-dependent protein kinase or with the stimulators of endogenous cyclic-GMP-dependent protein kinase or protein kinase C (Table 5). In addition, the activating effect of F1,6P2 was lost after protein kinase action; 0.05 mmol I⁻¹ F1,6P₂ produced a 3.8-fold activation of PK from control hepatopancreas but the ratio (V+F1,6P₂)/V was not significantly different from 1.0 after protein kinase treatments. However, treatments of the control enzyme with phosphatases or stimulation of endogenous phosphatases by spermidine had no effect on the properties of PK from control hepatopancreas. Equivalent treatments were carried out with hepatopancreas PK from estivating snails but neither S_0.5 PEP nor enzyme activation by F1,6P₂ was changed by any of the in vitro incubations. Parenthetically we note that the values for S_0.5 PEP for the enzyme in untreated incubations of extracts from active versus estivating snails differed somewhat from the comparable values reported in Table 3; we were unable to find a reason for this behavior except that a different population of snail was used for the in vitro incubations (see Materials and methods). However, in both cases the effect of estivation was the same, a strong decrease in S_0.5.

The properties of foot and mantle PK from O. lactea were very similar and not unlike the muscle form of PK from many species (Storey and Hochachka, 1975; Berglund and Humble, 1979; Storey, 1985). Thus, affinity for PEP was high (S_0.5 0.040 –0.090 mmol I⁻¹) and well within the range of PEP concentrations in O. lactea muscle (0.190 –0.330 mmol kg⁻¹ wet mass) (Churchill and Storey, 1989), the enzyme was unaffected by F1,6P₂, and Mg-ATP inhibition was weak. Like PK in most aerobic muscles, the enzyme in foot or mantle of O. lactea appears to be poised to be highly responsive to changes in glycolytic flux associated with muscular work. However, the enzyme displayed a significant sensitivity to inhibition by L-alanine with an I₅₀ of 5 mmol I⁻¹ and, as is the case for anoxia-tolerant marine molluscs (Hochachka and Mustafa, 1972; Plaxton and Storey, 1984a), such alanine sensitivity may be the key to enzyme inactivation in the hypometabolic state (see below).

The enzyme from hepatopancreas, however, showed very different kinetic properties, and probably represents a different isozymic form of PK, as also occurs in marine gastropods (Plaxton and Storey, 1985a). PK from O. lactea hepatopancreas showed a much lower affinity for PEP (S_0.5=0.570mmoll⁻¹) compared to foot and mantle PK and also compared to PEP levels in hepatopancreas in vivo (0.09 –0.15 mmol I⁻¹) (Churchill and Storey, 1989). The enzyme was also highly sensitive to F1,6P₂ activation and alanine inhibition. These properties are characteristic of PK in liver, or liver-like organs, from other species (Storey, 1985; Plaxton and Storey, 1985b; Engstrom et al. 1987) and they allow PK activity to be modulated in vivo in a multifunctional organ that includes both glycolytic and gluconeogenic carbon flow.

PK in all three tissues of O. lactea undergoes a stable modification of the enzyme protein upon entry into estivation or in response to anoxia. In all three organs this resulted in significant alterations to enzyme kinetic properties. The mechanism of this stable modification appears to be enzyme phosphorylation. Thus, the in vitro incubation studies revealed that kinetic changes characteristic of the control to estivating transition were mimicked by the actions of protein kinases on the control enzyme. Phosphatase treatments, however, did not affect the enzyme from control (aroused) snails but reconverted the kinetic parameters of PK from foot muscle of estivating animals to those characteristic of the control enzyme. These studies suggest that PK in tissues of control (aroused) O. lactea is a low-phosphate form of the enzyme whereas the enzyme in tissues from estivating/anoxic animals is a high-phosphate form. The effect of anoxia on PK in tissues of marine molluscs is the same. Anoxia induces a phosphorylation of PK and creates a much less active enzyme form (Plaxton and Storey, 1984a,b, 1985b). In the marine whelk, the effect is not mediated by cyclic-AMP-dependent protein kinase or the Ca²⁺-and phospholipid-dependent protein kinase C (Brooks and Storey, 1989a) but is due to the action of a cyclic-GMP-dependent protein kinase (Brooks and Storey, 1990b). The present results suggest that all three protein kinases could phosphorylate O. lactea PK in vitro but are also consistent with the presence of a second-messenger-independent protein kinase in O. lactea extracts. The protein kinase that mediates PK phosphorylation during the natural transition to an estivating state remains to be determined.

The changes in PK kinetic properties brought about by estivation or anoxia for the O. lactea foot and mantle enzymes are consistent with less active enzyme forms in the dormant or anoxic states. Thus, the affinity of foot muscle PK for its PEP substrate decreased in these states and inhibition by alanine or ATP increased. Similar results were seen for mantle (except for alanine inhibition in estivation), although here the changes in enzyme properties were more pronounced in anoxia than in estivation. These results are qualitatively the same as the effects of anoxia on PK from tissues of marine molluscs. In marine molluscs anoxia typically reduces enzyme V_max, lowers affinity for PEP, reduces F1,6P2 activation and enhances alanine inhibition of the enzyme (Holwerda et al. 1983; Plaxton and Storey, 1984,a; Hakim et al. 1985; de Vooys and Holwerda, 1986; Whitwam and Storey, 1990). Changes to PK that produced a less active enzyme form in foot muscle after long-term anoxia (45 h) also correlated well with changes in metabolic rate (measured as changes in rates of lactate accumulation in anoxia) and with changes in glycolytic rate inferred from measurements of the changes in concentrations of glycolytic intermediates. Thus, the rate of lactate accumulation declined rapidly under anoxia; the average was 1.90 μmolg⁻¹h ⁻¹ between 0 and 2h of anoxia and only 0.81 μmolg ⁻¹h⁻¹ between 2 and 14 h (Churchill and Storey, 1989). Furthermore, changes in the levels of PEP and pyruvate in foot muscle, when assessed via crossover analysis, indicated a relative activation at the PK locus during the initial stage of anoxia (2h), but subsequently showed strong metabolic inhibition at PK with longer exposure (14 h anoxia) (Churchill and Storey, 1989). During estivation, rates of oxygen uptake decline to about 70% of the rate in active snails within 6 –8h (Barnhart and McMahon, 1988) and average only 15 % of the active rate over the long term (2 –4 weeks of estivation) (Barnhart and McMahon, 1987). PK inactivation occurs between 12 and 48 h after estivation begins, suggesting that enzyme phosphorylation is not the first triggering event in the metabolic slow-down but is certainly one of the early events in establishing a stable hypometabolic state. Changes in metabolite levels in foot muscle were also consistent with metabolic inhibition at the PK locus between 3 and 22 days of estivation (Churchill and Storey, 1989). It appears, then, that enzyme phosphorylation, creating a less active enzyme form of PK, is an integral part of the transition from an active to a hypometabolic state in molluscs and that the response can be triggered by different stimuli (low water availability, lack of oxygen) in different species. The same use of post-translational modification to reduce the activity of glycolytic enzymes has also been identified in anaerobiosis in goldfish and turtles and hibernation in small mammals (Rahman and Storey, 1988; Brooks and Storey, 1988, 1989b; Storey, 1989).

The response of hepatopancreas PK to estivation or anoxia is more difficult to interpret. Changes in V_max, S_0.5 PEP and I₅₀ values for ATP and alanine would suggest a more active enzyme form in the dormant or anoxic state. This is opposite to the effect of anoxia on the kinetic properties of B. canaliculatum hepatopancreas PK (Plaxton and Storey, 1985,b). However, one dramatic change in the properties of O. lactea hepatopancreas PK occurred in anoxia or estivation and this was the loss of F1,6P₂ activation. Hepatopancreas PK changed from showing a high sensitivity to F1,6P₂ (K_a=0.004mmoll⁻¹) in control snails to being unaffected by the bisphosphate in estivation or anoxia. The action of F1,6P₂ on PK in liver-like tissues is frequently critical to enzyme function (Engstrom et al. 1987). PK in liver or hepatopancreas typically has a low affinity for PEP (S_0.5 is high relative to organ PEP levels) and feedforward activation of the enzyme by F1,6P₂ (the product of 6-phosphofructo-l-kinase), which lowers S_0.5 and overrides the effects of inhibitors, is necessary for substantial glycolytic flux. Thus, the loss of sensitivity to F1,6P₂ as the result of enzyme covalent modification may be all that is needed to produce the changes in hepatopancreas PK function that are required in the estivating or anoxic state.

Anoxia-induced phosphorylation of PK, when first described in marine molluscs, was proposed to be the mechanism responsible for regulating the PEP branchpoint and determining the aerobic (via PK and into mitochondrial oxidations) versus anoxic (via PEP carboxykinase and fermented to succinate or proprionate) fate of PEP in these animals. While this is obviously one very important consequence of PK phosphorylation in marine molluscs, the present study suggests that it may not be the primary role of PK phosphorylation. The fact that PK is phosphorylated in estivating or anoxic land snails suggests that the mechanism may have developed first as a method of glycolytic rate control. Thus, PK phosphorylation has no apparent function in redirecting glycolytic flux in O. lactea. The enzyme is covalently modified in estivation, even though this is an aerobic state where no fermentative end-products accumulate (Churchill and Storey, 1989) and glycolytic carbon continues to be fed via PK into the tricarboxylic acid cycle. PK is also modified in anoxia in O. lactea even though this modification does little to redirect carbon flow at the PEP branchpoint; thus, the ratio D-lactate:L-alanine:succinate was 8:2:1 over 14h of anoxia (Churchill and Storey, 1989). Furthermore, our recent studies with marine whelks have shown that at least three enzymes (PK, 6-phosphofructo-l-kinase and 6-phosphofructo-2-kinase) are phosphorylated and inactivated on parallel time courses during anoxia (Storey, 1988b; Whitwam and Storey, 1990; L. Bosca and K. Storey, unpublished data). Together these points suggest that PK phosphorylation functions primarily as a mechanism of overall glycolytic rate control and secondarily, in anoxia-tolerant marine molluscs, as a means of redirecting flux at the PEP branchpoint.

A time course for the changes in kinetic properties of foot muscle and hepatopancreas PK during estivation revealed that the transition of enzyme properties from those characteristic of the aroused state to those characteristic of estivation occurred between 12 and 48 h after the removal of food and water from the snails. As noted above, this indicates that PK modification is one of the early events associated with declining metabolic rate during entry into estivation. At 22 days of estivation the properties of PK were still the same as at 2 days, suggesting that the metabolic transitions associated with entry into estivation are made fairly quickly and abruptly and then maintained at a new steady state over the subsequent period of dormancy. The transition upon arousal from estivation was even faster. For foot muscle, the restoration of PK properties to those of the aroused (control) form, presumably via enzyme dephosphorylation, occurred within 10 min of the reintroduction of food and water. This corresponded with our observations that snails were beginning to emerge from their shells by this time. The restoration of PK to the control (aroused) form is apparently associated, therefore, with renewed muscular work. For hepatopancreas PK the time required for recovery was somewhat longer, between 3 and 24 h (although the high variation in the data points at 10 min and 1h resulted from one animal at each point that showed a high S₀.5 PEP of 0.55 mmol I⁻¹). Somewhat surprisingly, the time course studies revealed no intermediate values for the kinetic parameters during the transition times; PK from individual animals always showed kinetics characteristic of the active or the dormant states. This is unlike the situation seen in marine molluscs, where there is a gradation of change in kinetic parameters over the aerobic to anoxic transition (Holwerda et al. 1983; Whitwam and Storey, 1990). For whelk PK, however, this appears to be due to the presence of two phosphorylation sites per subunit of PK and differential rates of phosphorylation of these sites.

We are very grateful to Dr M. C. Barnhart for supplying the snails as well as information on their care and continued discussions on the physiology of estivation. Supported by an operating grant from the NSERC Canada to KBS.