Metabolic Heat Dissipation and Internal Solute Levels of Artemia Embryos During Changes in Cell-Associated Water

Post-diapause embryos (cysts) of the brine shrimp Artemia continue to be one of the most intensely studied anhydrobiotic systems (for reviews, see Crowe & Clegg, 1973; Clegg, 1981, 1984). In a previous communication, we examined mechanisms involved in the cessation of carbohydrate catabolism in dehydrated Artemia cysts from the Great Salt Lake (Glasheen & Hand, 1988). When hydrated, developing embryos were dehydrated by immersion in 5·0 mol 1⁻² NaCl, trehalose metabolism was inhibited at the series of reactions catalyzed by trehalase, hexokinase and phosphofructokinase. Intracellular pH (pHi) appeared to play no part in these metabolic transitions, in contrast to its pronounced regulatory role in Artemia embryos undergoing transitions into anaerobic dormancy (Busa et al. 1982; Busa & Crowe, 1983; Carpenter & Hand, 1986; Hand & Carpenter, 1986; Utterback & Hand, 1987; Hand & Gnaiger, 1988). Thus, entry into the quiescent state of anhydrobiosis in Artemia seems to be a strict function of cellular water content, as initially proposed by Clegg (1964, 1974, 1976a,b).

One deficiency currently preventing investigators from testing various mechanisms involved in dehydration-induced arrest of metabolism is the lack of a comprehensive analysis of the internal osmolyte system for Artemia embryos. Without such information, it is impossible to estimate the ionic and organic solute concentrations present in the embryo at the onset of metabolic arrest. Yet, in over two decades of research into the Artemia system, many of these important constituents of the intracellular milieu have never been quantified. The few values available (amino acid, trehalose, glycerol, sodium content; Emerson, 1967; Clegg, 1967 ; Conte et al. 1977) are taken from different populations of embryos and are often expressed in ways that make it difficult to estimate internal concentrations (e.g. per mg protein).

With the use of open-flow microcalorimetry, we characterize in this paper the dehydration-induced metabolic arrest in Artemia embryos from the Great Salt Lake population. Metabolic heat dissipation is a sensitive measure of total energy flow in the embryo (see Hand & Gnaiger, 1988) that can be applied in experiments where dehydration is accomplished with concentrated salt solutions. Use of calorimetry avoids the technical difficulties that arise when oxygen consumption is chosen as a means to estimate aerobic metabolism in solutions approaching saturation with NaCl; electrolyte solutions within the polarographic oxygen electrodes dehydrate under such conditions (Hale, 1983).

We relate the measured transitions in energy flow to the values for internal water content of embryos determined under identical experimental conditions. When the hydration values are coupled to the quantitative analyses of all major intracellular solutes, it is possible to estimate the composition of the internal milieu of these embryos at any point along the graded series of dehydration. Potential mechanisms for metabolic arrest are considered in the light of these new findings. Metabolic arrest may possibly be the result of disruption of macromolecular assemblies (e.g. enzyme-cytostructural interaction) owing to the elevated ionic strengths of intracellular contents in dehydrated embryos.

Dehydrated Artemia embryos (gastrula stage) from the Great Salt Lake were purchased from the Sander’s Brine Shrimp Co. in 1984 (Ogden, UT, USA), and stored at — 20°C. For all experiments, cysts were taken from the same commercial batch. The hatching percentage for this batch of cysts has been previously reported (Glasheen & Hand, 1988) and is virtually identical to the percentages determined in other studies using embryos from the Great Salt Lake population (Conte et al. 1977; Peterson et al. 1978; Warner et al. 1979). Dura Wipe filters were a product of Chicopee Mills, Inc. (New Brunswick, NJ, USA). Viton tubing was obtained from Cole Palmer (Chicago, IL, USA). Glycerokinase (from Candida mycoderma), glycerophosphate dehydrogenase (Type I, rabbit muscle) and highly polymerized DNA (calf thymus) were purchased from Sigma Chemical Co. (St Louis, MO, USA). All other chemicals and biochemicals were reagent grade.

In general, preparatory procedures followed those of Clegg (1974) and were always performed at 0°C. Dried cysts were initially washed five times in 0-25 mol 1⁻² NaCl. After each wash, floating cysts were decanted and discarded. Embryos were then suspended in 5·0 mol 1⁻² NaCl and collected from the surface onto Dura Wipe filters; sand and other debris that had sunk to the bottom of the salt solution were discarded. At this point, washing procedures varied according to the type of experiments for which embryos were intended. For studies requiring the removal of the cyst outer shell (dechorionation), embryos were soaked for 30–40 min in a modified antiformin solution (Nakanishi et al. 1962) containing 2·5% (w/v) active hypochlorite, 0·4mol1⁻² NaOH and 60mmol1⁻² Na₂CO₃. Following dechorionation, embryos were washed five more times in 0·25 mol 1⁻² NaCl and either used immediately in experiments or stored overnight at 0°C in the same medium.

If cysts were to be used for Na⁺ and Cl^- determinations, final washes were performed in deionized water (Milli-Q Reagent Water System; Millipore Corp.; resistance = 18 MΩ). Although it has been reported that soaking Great Salt Lake embryos in distilled water results in substantial lysis (Clegg, 1974), we found that if cysts were used immediately after brief washes in deionized water, the total percentage of ruptured cysts did not exceed that observed for cysts soaked in 0-25 mol 1⁻² NaCl (% ruptured cysts after deionized water wash = 9·5 ± 0·26; % ruptured after 0·25moll⁻² NaCl wash = 13·1 ± 0·70; mean ± S.E.M., N = 3 determinations).

Measurements of heat dissipation from Artemia embryos were made with an LKB2277 thermal activity monitor equipped with a 3·5 ml perfusion chamber adapted to ensure uniform flow of medium through the sample of embryos (Hand & Gnaiger, 1988). Perfusion media were equilibrated with a 60 % N₂:40 % O₂ gas mixture, calibrated with Tylan FC-260 mass flow meters. Flow lines for gases and for aqueous media were made of Viton and stainless steel, respectively. Flow rate of the perfusion media was set at 16·0 ml h⁻², and temperature was maintained at 23°C for all experiments. Heat dissipation in microwatts was time-corrected by the calibration unit of the calorimeter; typical time constants were t₁ = 600s and t₂ = 17100 s. Data were sampled every 60s and stored on an IBM XT personal computer, as well as on a strip chart recorder for immediate visual inspection. For all experiments, blanks (no embryos in perfusion chamber) were run and subtracted from the measurements obtained with embryos present. To ensure that oxygen was not limiting for embryos, the O₂ in the gas mixture was increased above 40 %, and no significant change in heat dissipation was recorded.

Embryos, prepared as described above (with intact chorions), were allowed to hydrate overnight at 0°C in 0·25 mol 1⁻² NaCl prior to the beginning of the calorimetry experiments. After calibration of the instrument, embryos were loaded into the perfusion chamber. Heat dissipation in 0·25 mol 1⁻² NaCl (control medium) was monitored for 4h. Perfusion medium was then switched to one of five different NaCl solutions (1·0, 2·0, 3·0, 4·0, and 5·0 mol 1⁻²) and heat dissipation recorded for another 16 h on average. Finally, incubation medium was switched back to 0·25 mol1⁻² NaCl and the recovery period monitored.

Hydration state in Artemia embryos can be precisely controlled by immersing cysts in solutions of varying ionic strength (Clegg, 1974), because the cyst wall (outer cuticular membrane) is impermeable to molecules other than low molecular weight dissolved gases and water (see Clegg & Conte, 1980, for a review). However, hydration values exist only for embryos from the San Francisco Bay population of Artemia (Clegg, 1974). Because significant interpopulational differences in hydration state may exist (Clegg, 1974), it was necessary for us to obtain these values for Great Salt Lake embryos to interpret more precisely the metabolic and intracellular solute data of this and previous studies (Glasheen & Hand, 1988).

First, to obtain a valid comparison with hydration values of San Francisco Bay embryos, experiments with the Great Salt Lake gastrulae were conducted in a virtually identical way to those in Clegg (1974), and the resulting data are presented as series I (see below). Second, hydration levels also had to be measured under conditions dictated by the various metabolic studies we have performed (series II, below).

Briefly, embryos were washed as described above and placed in desiccators over CaSO₄ (Drierite, WA Hammond Co., Xenia, OH, USA) until gravimetric equilibrium was reached. Cysts dehydrated in this manner were used directly for incubations in the experimental salt solutions. Approximately 1 g of dry cysts was added to triplicate 125 ml flasks containing 60 ml of one of the following solutions: 0·25, 1·0, 2·0, 3·0, 4·0 or 5·0 mol 1⁻² NaCl. Flasks were then shaken at 110 cycles min⁻² at 0°C for 24 h. Following incubation, 100-200 mg of cysts was filtered onto Dura Wipes and blotted between two sheets of Whatman no. 4 chromatography paper to remove interstitial water. Embryos were not washed in either deionized or dilute NaCl solutions prior to blotting. The amount of salt possibly adhering to the cysts after blotting was calculated to be negligible compared to the error introduced by even a 30s washing procedure. Results of a preliminary study show a significant increase in water content of fully desiccated cysts following a 30s rinse in 0·25moll⁻² NaCl (embryo H₂O = 11·2 ± 1·09g 100 g⁻² dry cyst; S.E.M., N = 3). Blotting was performed with a gentle circular motion until embryos attained the consistency of fine sand and flowed freely from the paper (Clegg, 1974). Cysts were then weighed on an analytical balance in tared aluminum weighing pans and placed over CaSO₄ in 1–1 glass desiccators. To maintain consistency between treatments, blotting time for each sample was 3·5 min, with a total of 5 min for filtration, blotting and weighing. Groups of three replicates obtained for each NaCl incubation were placed in the individual desiccators for 72 h, after which time dry masses were obtained.

Procedures employed for this series were identical to those for series I, except that the washed cysts were not dry, but rather pre-hydrated for 16 h in 0·25 mol 1⁻² NaCl, prior to starting the incubations in the experimental salt solutions. Following experimental incubations, the blotting, weighing and drying of cysts were performed exactly as they were in series I.

It is appropriate to note that the final hydration values reached after incubation in the experimental salt solutions are substantially different depending on whether one starts with dry cysts (series I) or cysts equilibrated with 0·25 mol 1⁻² NaCl (series II).

Metallic ion content was determined by atomic absorption spectroscopy. Cysts were washed and then dechorionated to eliminate artifacts due to ion entrapment in the porous, acellular chorion. Embryos were filtered onto Whatman no. 41 paper, samples placed in tared crucibles, and dried to constant mass (24 h in a 60°C oven). Cysts dried in a 60°C oven exhibited little additional mass loss over embryos dehydrated 72h over CaSO₄. Additional mass loss = 0·85 ± 0·17% (S.E.M., N = 18). Dry mass was measured, and embryos were placed in a Lindberg muffle furnace overnight at 550 °C. The resultant ash was suspended in 5 % HNO₃. After addition of lanthanum oxide and cesium chloride, solutions were brought to a final concentration of 7·5mmol1⁻² CsCl, 1% La₂O₃ and 0·5% HNO₃. The suspension was then centrifuged at 4000g for 10 min. Supernatants were decanted and analyzed for ions. An additional wash of the pellet was not necessary for complete ion extraction.

Sodium was measured in a Varian Techtron atomic absorption spectrometer at 589·0 nm using an air-acetylene flame. Potassium was determined at 766·5 nm using the air-acetylene mixture. Magnesium was quantified at 285·2 nm with acetylene-N₂O as fuel-oxidant, and calcium was measured at 422·7 nm, again with nitrous oxide as oxidant. Calibration curves were generated from standard solutions (Environmental Protection Agency, US).

For inorganic phosphate and chloride analyses, the dechorionated cysts were filtered onto Dura Wipes and excess moisture wicked away. Two samples of approximately equal mass were removed and wet masses obtained. One sample was used for ion analysis, while a similar portion was dried to constant mass. For phosphate analyses, samples were homogenized in 5 ml of 6 % perchloric acid and centrifuged at 10000g for 15 min. Supernatants were neutralized with 5 mol 1⁻² K2CO3 and centrifuged again to remove perchlorate salts. Inorganic phosphate was determined with a modified version of the Piper & Lovell (1981) method. The colorimetric reagent was made by adding 37·5 g of guanidine-HCl to 25 ml of a 7·5% ammonium molybdate/2·5mol1⁻² H₂SO₄ solution. This mixture was brought to a final volume of 100 ml with a 20 mg m1⁻² bovine serum albumin solution. The reagent was added 3:1 (v/v) to both samples and standards, and after 30 min the absorbance was measured at 370 nm. For chloride measurement, samples were homogenized in deionized water and centrifuged at 10000 g for 15 min. After addition of Ba(OH)₂ and ZnSO₄ to final concentrations of 0·025 mol 1⁻² and 1 %, respectively, supernatants were then left at room temperature for 48h to precipitate interfering materials. Following centrifugation, supernatants were analyzed with a Haake-Buchler digital chloridometer.

Dechorionated embryos were homogenized in two volumes of deionized water, and 175 μl of each homogenate was centrifuged at 140 000g in a Beckman Airfuge for 30 min. The osmotic pressure of the supernatants was measured with a vapor pressure osmometer. In order to equate the osmotic pressure measured in vitro to the in vivo situation, this approach assumes, of course, that all solutes in the embryo are in solution. It is likely that this assumption could result in an overestimate of the actual in vivo osmotic pressure.

Trehalose, glycogen and glycerol were measured in dechorionated cysts as described by Hand & Carpenter (1986). Ninhydrin-positive substances (NPS) were measured in neutralized perchoric acid extracts by the colorimetric method of Lee & Takahashi (1966), with aspartic acid as the standard. Total methylamines were analyzed with a modified version of the reineckate precipitation method of Barnes & Blackstock (1974). Samples were homogenized in 4 vols of 70 % ethanol, and then placed in a boiling water bath for 15 min. Homogenates were then centrifuged at 10 000g for 15 min, and the supernatants were collected and rotoevaporated. The residue was dissolved in 0·1mol1⁻² HC1 and clarified by centrifugation. A volume of this supernatant containing approximately 20μmol of methylamine (betaine standard) was diluted to 1·0 ml. Then 3·0 ml of ammonium reineckate reagent was added, the solution vigorously mixed, and precipitation allowed to occur overnight at 4 °C. The reineckate solution consisted of 5 g of NH₄-[Cr(SCN)₄(NH₃)₂]-H₂O brought to 100ml with 0·1mol1⁻¹ HC1. The solution was stirred for 45 min, titrated to pH 1·0, and filtered through Whatman no. 1 paper. The sample precipitate was collected by centrifugation at 3500g and pellets were washed in diethyl ether three times. Final pellets were brought to 6 ml with 70% acetone and absorbance read at 520 nm.

Total protein was determined in dechorionated embryos as described by Hand & Carpenter (1986) using the method of Peterson (1977). Embryo DNA content was quantified with the diphenylamine method of Richards (1974). Pellets from cysts homogenized in 6% perchloric acid (PCA) were resuspended in 10 ml of 1·5 mol 1⁻² PCA and heated at 70°C for 15 min with occasional mixing. Samples were then centrifuged at 10 000g for 15 min, and DNA measured in the supernatant. Total lipid content of dechorionated embryos was determined gravimetrically following the procedure of Folch et al. (1957), as modified by Ways & Hanahan (1964).

This set of experiments was designed to quantify the relationship between reductions in cellular hydration and the suppression of total energy flow in Artemia embryos. Heat dissipation from developing embryos rose steadily during the first 4h in control 0·25 mol 1⁻² NaCl. When the perfusion medium was switched to 1·0 mol 1⁻² NaCl, energy flow continued to rise, reaching 137 % of control (hour 4) values (Fig. 1; Table 1). However, when the control medium was switched to 2·0 moll⁻² NaCl, energy flow fell exponentially to 21% of control values after approximately 18h. In experiments with higher salt concentrations (3·5 mol 1⁻²), heat dissipation fell sharply to 6% or less of the rates at hour 4 (Fig. 2; Table 1). Thus, it appears that metabolism in developing embryos is disrupted at a critical hydration level promoted by an external NaCl concentration between 2·0 and 3·0 mol 1⁻². This range of external NaCl concentration needed for reduction of heat dissipation is comparable to that required for blockage of trehalose utilization (Clegg, 1976a; Glasheen & Hand, 1988) and the onset of respiration in Artemia (Clegg, 19766). It is interesting to note that dehydration of embryos in 3·0 mol 1⁻² NaCl or higher yields an energy flow (0·20 mWg⁻² dry mass) comparable to that observed in Artemia after several hours of anaerobic dormancy (Hand & Gnaiger, 1988). This low heat dissipation may reflect maintenance level metabolism in these embryos. In any case, intracellular hydration levels induced by these high-salt media appear incapable of supporting conventional metabolism.

When embryos were returned to 0·25 mol 1⁻² NaCl perfusion after dehydration bouts, recovery of energy flow was rapid (Figs 1, 2). In each case (2, 3, 4 and 5mol1⁻² NaCl), heat dissipation reached values that corresponded to approximately 135 % of the pre-dehydrated value (Table 1). Note that every time the salt concentration was changed, there was a 30 min gap in the heat dissipation profile. This disruption was due to heats of mixing, which artifactually forced the calorimetric readings off scale for those periods.

Results of hydration studies are presented in Table 2 (series I) and Table 3 (series II). For series I data, our Great Salt Lake cysts were prepared identically to the San Francisco Bay embryos in the studies by Clegg (1974); i.e. embryos were hydrated in the experimental solutions directly from the dry state. Differences in hydration behavior in response to external salinity exist between the two populations. The differences occurred primarily in the 1·0 and 2·0 mol 1⁻² NaCl incubations (Table 2); values were identical at the higher NaCl concentrations. These observations become particularly important when comparing the metabolic responses to dehydration between the two populations (e.g. relating microcalori-metric results with previous measurements of respiration).

Series II data are from embryos pre-hydrated in 0·25 mol1⁻² NaCl and then exposed to the increasing salt series (Table 3); this approach is the more practical for our metabolic studies. Note that in series II, embryos were effectively being dehydrated in order to reach the final hydration values, while in series I, the embryos were initially dry and gained water to reach the final hydration values. This directional difference in water movement apparently altered the final hydration levels obtained at 24 h; the water content was higher in embryos from series II. The reason for this difference has not been determined, but we suspect that numerous air spaces and channels in the chorion influence in some way the rate of embryo hydration versus dehydration. Preliminary studies indicated that hydration properties are markedly different with dechorionated embryos (data not shown).

Finally, the ‘corrected’ hydration levels presented in Table 3 represent estimates of the actual water content of the metabolically active embryonic tissues. These data, in combination with the microcalorimetry results, showed that energy flow was severely disrupted when embryo water was reduced to 70–85 g H₂O 100 g⁻² dry mass. This water content compares well with critical hydration levels required for the onset of carbohydrate catabolism and respiration, when corrected in a similar fashion (70 g H₂O 100 g⁻² dry cyst; Clegg, 1976a,b).

Ionic concentrations for dechorionated cysts are expressed as millimolal values in Table 4. Values calculated for each incubation condition were based on the corrected hydration values given in Table 3. The osmotic contribution of each ion reflect the embryo hydration levels in 0·25 mol 1⁻² NaCl and the indicated osmotic coefficient. Values for intracellular sodium, potassium and calcium were in good agreement with concentrations reported previously for barnacle muscle fibers (Balanus nubilis;Clark & Hinke, 1980). Inorganic phosphate levels were reasonably close to those reported for the decapod marine crustacean Nephrops (21 mmol 1⁻²; Robertson, 1961). Chloride was somewhat higher than the 30–35 mmol kg⁻² H₂O reported by Hinke et al. (1973) and Hinke & Gayton (1971) for Balanus muscle. Intracellular magnesium levels were unexpectedly high in Artemia embryos, more than three times the concentration found in some crustacean muscle (e.g. Balanus, Page et al. 1971). These ions are calculated to contribute an osmotic pressure of 275 mosmol kg⁻²H₂O when cysts are incubated in 0·25 mol 1⁻² NaCl. Since divalent ions were included in this value, their contribution to the osmotic pressure may be a slight overestimate, as the percentage of bound Mg²⁺, Ca²⁺ and Pi is unknown. If cysts are dehydrated to the critical range of 70–85 gH₂0100 g⁻² dry mass (corrected), the intracellular osmotic pressure contributed by inorganic ions increases to over 500 mos-mol kg⁻² H₂O. Inorganic ions account for 1·5 % of the dry embryo mass (Table 5).

Assuming that all the measured trehalose is soluble, this disaccharide accounted for 26 % of the osmotic pressure in Artemia embryos (Table 4), as well as 16 · 5 % of the dry mass (Table 5). This latter value is in close agreement with that reported by Clegg (1964). The polyalcohol glycerol also made a significant contribution to osmotic pressure (319 mosmol kg⁻²H₂O), again in agreement with values reported by Clegg (1964).

Nitrogenous solutes account for 12% of the osmotic activity in Artemia. Ninhydrin-positive substances contribute 62-3 mosmol kg⁻²H₂O (Table 4) and 1·11 % of the dry mass (Table 5). This value is significantly lower than for most marine invertebrates (Yancey et al. 1982), but it agrees well with data reported by Emerson (1967) and Clegg & Lovallo (1977) for Artemia embryos. Total methylamines (calculated using betaine as the standard) were present at 56·4 mosmol kg⁻²H₂O. This is the first report documenting the presence of these compounds in Artemia embryos. The beneficial effects of these osmolytes on enzyme structure and function have been documented (Clark & Zounes, 1977; Bowlus & Somero, 1979). In Artemia, trehalose and glycerol should also be considered as compatible solutes (Crowe, 1971; Crowe et al. 1987).

For the soluble components, osmotic coefficients have been assigned in the following manner: inorganic ions, 0·9 (Scatchard et al. 1938); ninhydrin-positive substances (NPS), 1·0 (Robertson, 1975); trimethylamine-N-oxide (TMA0) + betaine, 1·15 (Robertson, 1975); glycerol, 1·0 (Scatchard et al. 1938); and trehalose has been assumed to behave identically to sucrose (1·0). In summary, soluble organic compounds account for 778 mosmol kg⁻²H₂O, or 73% of the osmotic pressure observed in Artemia embryos. The measured organic solutes comprise 22·4% of cyst dry mass (Table 5).

Total osmotic pressure, measured by vapor pressure depression in dechorionated embryos, was 1300 mosmol kg⁻²H₂0 at the level of hydration induced in Artemia by immersion in 0·25moll⁻² NaCl (Table 4). This leaves an unidentified osmotic pressure component of 248 mosmol kg⁻²H₂O or 19% of the total. It is quite possible that much of this unknown component consists of osmotically active protein, oligopeptides, or polysaccharide moieties of glycoproteins, as was previously suggested for barnacle muscle (Clark & Hinke, 1980). When cyst hydration is decreased to 85·4 g H₂O 100 g⁻² dry mass (immersion in 2·0mol1⁻² NaCl), total osmotic pressure increases to 2048 mosmol kg⁻² H₂O. It is interesting to note that the internal osmotic pressure of Artemia embryos is substantially below that of the external medium for all incubations ⩾1·0 mol 1⁻² NaCl (see Discussion).

PCA-insoluble protein made up the majority of embryo dry mass in Artemia. The value of 54·4% agreed well with the protein component (58·5%) found in Balanus muscle (Clark & Hinke, 1980). Chloroform/methanol-extracted lipids also formed a large component of cyst dry mass (16·7%). These results were similar to those reported by Dutrieu (1960) for intact (chorionated) embryos. Glycogen values were low (1·6 %) in Artemia, as would be expected at this point in development based on previous studies of carbohydrate metabolism in dormant cysts (Glasheen & Hand, 1988; Carpenter & Hand, 1986; Clegg, 1976a). DNA content (0·81 %) was in line with values reported for most eukaryotes at resting stages in the cell cycle (e.g. blue crab, Wang & Stickle, 1986). Taken together, 97·4% of the dry mass in Artemia embryos has been identified and assigned to ionic, organic solute or macromolecular fractions (Table 5).

The results of the present study characterize more fully the metabolic arrest and the alterations in internal solute composition that occur as Artemia embryos enter and exit the state of anhydrobiosis. The use of microcalorimetry improves our understanding of these transitions by providing an estimate of total energy flow in the embryo. The comprehensive inventory of intracellular solutes, in conjunction with embryo hydration values, offers an opportunity to estimate more precisely the status of the internal milieu at the critical hydration point where metabolic arrest begins. Considering the numerous metabolic studies performed with Artemia embryos, it is surprising that a complete characterization of the internal osmolyte system has not been available until now. We will discuss below selected possibilities as to how changes in intracellular solution conditions might bring about the metabolic arrest observed during restricted water availability.

Our measurements of heat dissipation show that total energy flow is acutely suppressed when water content of the metabolically active tissue falls to 85 gH₂O 100 g⁻² dry mass. At 70gH₂0100g⁻² dry mass and below, heat dissipation has dropped to only a few percent of the control (0·25 mol1⁻² incubation) values or, in absolute terms, 0·14–0·26 mW g⁻² dry mass. During the quiescent state induced by anaerobiosis, Artemia embryos exhibit similar values during short-term (several hour) experiments (Hand & Gnaiger, 1988), although the mechanism of the arrest is apparently quite different. Substantial evidence supports a primary role for acidotic pHi in anaerobic dormancy; but during dehydration, metabolic arrest proceeds normally under conditions where pHi does not change (Glasheen & Hand, 1988). Both these forms of hypometabolism are fully reversible. Upon restoration of intracellular water, energy flow in embryos rises rapidly and dramatically to levels comparable to those in fully hydrated, developing embryos (Figs 1, 2). In previous studies, embryo respiration and trehalose usage displayed similar patterns (Clegg, 1976a,b;Glasheen & Hand, 1988).

Traditional views regarding this metabolic arrest in Artemia entering anhydrobiosis have centered upon the role of water in maintaining the structural and functional integrity of metabolic pathways. Minimum water content has been established for San Francisco Bay embryos that permit the onset of various metabolic functions (for a review, see Clegg, 1981). Hydration values for Great Salt Lake embryos have not been available heretofore, but now some interpopulation comparisons can be made. Critical hydration levels necessary for the onset of major metabolic events are similar for each population. A water content of 70 g H₂O 100 g⁻² dry mass is needed for respiration in the San Francisco Bay population (Clegg, 1974), while 70–85 g H₂0100 g⁻² dry mass is required for substantial heat dissipation in Great Salt Lake embryos (Table 2; Fig. 1). However, differences in hydration response are seen in solutions of low ionic strength. This metabolic shutdown observed during progressive dehydration/re-hydration has been explained by perturbation of ‘vicinal’ or protein-associated water (Clegg, 1984). Cellular regions of vicinal versus bulk water were described by Porter (1986) as ‘protein-rich’ and ‘water-rich’ phases, and this distinction was suggested by experiments performed as early as 1938 (Kopac, 1938; Chambers, 1940). In the present study, there is at least indirect evidence that the properties of intracyst water may be quite different from those of the external medium. For example, the difference between external and internal osmotic pressure for embryos incubated in 5·0 mol 1⁻² NaCl approaches 10000 mosmol kg⁻²H₂O. It is unlikely that a lack of osmotic equilibrium can explain this discrepancy, but rather that much of the intracellular water is restricted and not osmotically available.

While interpretations based on solvent status have proved highly useful, the important corollary of intracellular solute content has been relatively neglected vis à vis metabolic organization and control in Artemia. As a direct consequence of reductions in cell-associated water, intracellular solute concentrations rise dramatically. In Artemia, as water activity decreases to the critical level (as demonstrated by metabolic arrest), the concentration of inorganic ions approaches 500 mmol kg⁻² H₂O. In most marine species studied, levels above 100–200 mmol kg⁻² H₂O of univalent ions are severely deleterious to enzyme function, as judged by lowering of reaction velocities as well as elevation of K_m (Borowitzka & Brown, 1974; Bowlus & Somero, 1979; Yancey et al. 1982). In addition to these direct influences on enzyme function, evidence suggests that protein binding to subcellular components may be disrupted by elevated levels of inorganic ions. In mammalian muscle, the adsorption of the glycolytic enzymes aldolase, lactate dehydrogenase and pyruvate kinase to an F-actin-tropomyosin-troponin complex was significantly reduced in 150 mmol 1⁻² KC1 (Clarke & Masters, 1975). Likewise, 150 mmol 1⁻² NaCl resulted in desorption of glyceraldehyde-3-phosphate dehydrogenase from Band 3 protein in erythrocytes (Yu & Steck, 1975). Mammalian brain hexokinase, in its more active ‘bound’ form, is almost completely solubilized in 350 mmol 1⁻² KC1 or NaCl (Wilson, 1968). Partitioning of enzymes into soluble and particulate fractions (with differing kinetic properties) or altering the cohesiveness of metabolons (multienzyme associations) are attractive models for effective means of metabolic regulation (for reviews, see Welch, 1977; Kurganov, 1985; Welch & Keleti, 1981; Srere, 1987; Ovadi, 1988; G. N. Somero & S. C. Hand, in preparation).

Thus, in dehydrated Artemia, ionic perturbation of protein-protein or protein-cytoskeletal interactions may account for at least part of the observed metabolic arrest. Previous experiments have indicated simultaneous inhibition of three major enzymes during the glycolytic shutdown in dehydrated embryos, namely trehalase, hexokinase and phosphofructokinase (Glasheen & Hand, 1988). The catalytic properties of Artemia trehalase are significantly altered by modifications of macromolecular assembly state (Hand & Carpenter, 1986). While Artemia hexokinase does not reversibly associate with isolated mitochondria (Rees et al. 1989), as does the rat brain homologue, we have preliminary evidence that phosphofructokinase does bind to a particulate fraction in homogenates of these embryos (J. S. Glasheen & S. C. Hand, unpublished observations). The fact that the dehydration-induced metabolic shutdown in Artemia is fully reversible is consistent with the above view of metabolic organization. Such non-covalent assembly/disassembly transitions could alter enzymic function without compromising native protein structure. Indeed, it seems reasonable to suggest that conservation of macromolecular integrity is an important prerequisite for organisms capable of undergoing such drastic reductions in cell-associated water.

It is well established now that many organisms that frequently face water stress possess systems of ‘compatible’ solutes (Brown & Simpson, 1972), i.e. compounds either non-perturbing or actually stabilizing to macromolecular structure and function (Yancey et al. 1982; Somero, 1986). Generally, these organic solutes are either uncharged or zwitterionic, and quite often structurally resemble the stabilizing ions of the Hofmeister series (Clark & Zounes, 1977). Our analyses of intracellular solutes in Artemia reveal significant levels of methylamines (50-127 mmol kg⁻² H₂O, depending on hydration state), a family of compatible solutes known to exert stabilizing influences at concentrations as low as 50mmolkg⁻²H₂O (e.g. Hand & Somero, 1982). In addition, as dehydration reaches critical levels in the Artemia embryos, concentrations of the osmolytes trehalose and glycerol exceed 1000 mmol kg⁻² H₂O. In the case of biological membranes, Crowe et al. (1987) have suggested that trehalose serves a ‘water replacement’ function, stabilizing membrane structure through physical intercalation with phospholipids. In combination with transition metal ions, trehalose also protects proteins against denaturation incurred during freezing (Carpenter et al. 1986) and drying (Carpenter & Crowe, 1988).

Based on the comprehensive inventory of embryo solutes presented herein, we suggest that the role of intracellular osmolytes should be more fully integrated into hypotheses explaining the dehydration-induced arrest of Artemia metabolism. At low water activities, concentrations of inorganic ions rise appreciably, which could result in the disruption of protein-protein associations and the attendant loss of metabolic function. The substantial quantities of the organic solutes trehalose, glycerol and the newly detected methylamines may be involved in fostering reversibility of these anhydrobiotic bouts by preventing macromolecular denaturation during drying. Our current in vitro studies are incorporating these solute data to quantify the relative influences of inorganic and organic osmolytes on protein assembly state. The eventual goal is to clarify the importance of the internal milieu in regulating metabolic events, not only in terms of solvent interactions but also in terms of the nature and quantity of critical solutes.

This work was supported by National Science Foundation grants DCB-8702615 and BBS-8704421 to SCH. The authors thank Dr Dwayne W. Hamar (Department of Pathology, School of Veterinary Medicine, Colorado State University) for use of the chloridometer and Dr William D. Lewis, Jr (Department of EPO Biology) for use of the atomic absorption spectrophotometer. Appreciation is also extended to Mr Donald Morris for technical assistance.