Variations in Macromolecular Antifreeze Levels in Larvae of the Darkling Beetle, Meracantha contracta JOHN G. DUMAN Biology Department, University of Notre Dame, Notre Dame, Indiana 46556
ABSTMCT Overwintering larvae of the darkling beetle, Meracantha contracts, produce a macromolecular antifreeze that is similar in activity to the glycoproteinaceous and proteinaceous antifreezes found in some cold-water, marine teleost fishes. The antifreeze is not present in the hemolymph of the Meracantha larvae in summer, but its production begins by late September in the wild population. The antifreeze reaches a maximum concentration in February, decreases slowly through spring, and disappears by early June. The supercooling points of the larvae are lowest in February, when the antifreeze levels are highest, and increase as the antifreeze concentrations in the hemolymph decrease in the spring. Larvae collected in mid-February and warm-acclimated lost the antifreeze within 12 days. Larvae collected in early September and cold-acclimated required nearly two months to produce concentrations of antifreeze comparable to those of overwintering larvae. Temperature seems to be the major environmental factor responsible for the control of antifreeze levels in Meracantha;however, other environmental factors may also be involved. The overwintering phases of many polar and temperate-zone insects are well known for their abilities to survive subzero environmental temperatures (for reviews see Asahina, '66, '69; Salt, '61, '66). Many of these insects are capable of withstanding the actual freezing of their body fluids however, most seem to survive the winter by supercooling to temperatures below the equilibrium freezing point (the freezing point based on colligative properties) of their body fluids. The body fluids of most insects, even in summer, will supercool below the equilibrium freezing point. Therefore, i t is not the freezing point itself, but rather the temperature to which the body fluids can supercool before spontaneous nucleation occurs that is important to the survival of insects that are unable to survive freezing. Many physiological and biochemical adaptations are, no doubt, involved in the low temperature tolerance of these insects, especially in those that survive freezing (Baust, '73). However, the adaptation that has been best correlated with low-temperature tolerance in a number of overwintering insects is the production of high concentrations of polyhydric alcohols, such as glycerol, sorbitol, mannitol and threitol (Salt, '57, '59; Samme, '64, Chino, '57, '58; Baust and Miller, J. EXP. ZOOL., 201: 85-92.
'70, '72; Miller and Smith, '75; Frankos and Platt, '76). These polyols lower the freezing point of the hemolymph significantly and also depress the supercooling point allowing the insect to supercool to lower temperatures without spontaneous freezing. In addition the polyols may function as cryoprotectants to prevent damage if freezing should occur. Glycerol in particular is produced a t high levels in winter in several species of insects. For example, glycerol concentrations in the overwintering larvae of the wheat stem sawfly, Bracon cephi, are as high as 5 M (Salt, '59). Although numerous studies have shown the importance of low molecular weight polyols to insect low-temperature tolerance, only one has investigated the correlation between solutes of high molecular weight and overwintering capabilities (Duman, '77). In winter, larvae of the darkling beetle, ueracantha contracts (Coleoptera, Tenebrionidae) produce a macromolecular antifreeze which depresses the supercooling point of the larvae significantly below that of the summer larvae. This antifreeze does not function a s a cryoprotectant since the overwintering larvae are unable to tolerate ice formation in their tissues. The macromolecular antifreeze produces a thermal hysteresis in the hemolymph,
85
86
JOHN G . DUMAN
whereby the freezing point is depressed several degrees (as much as 6OC in some individuals) below the melting point. The thermal hysteresis in Meracantha hemolymph is similar to that produced by macromolecular antifreeze solutes that depress the freezing point of the body fluids of many cold-water, marine teleosts below the freezing point of seawater, thereby protecting the fish from freezing. These antifreeze solutes, first discovered in certain Antarctic nototheniid species (DeVries and Wohlschlag, '69),have since been demonstrated in many phylogenetically divergent species of cold-water fishes (Scholander and Maggert, '71; Hargens, '72; Duman and DeVries, '74, '75). Purification and characterization of the antifreeze substances from species of four different families of teleosts have shown that they are either proteins or glycoproteins, depending on the species (DeVries et al., '70; DeVries et al., '71; Komatsu et al., '70; DeVries, '71; Lin et al., '72; Shier et al., '72; Raymond et al., '75; Duman and DeVries, '76). The mechanism by which the fish antifreezes depress the freezing point of water 200 to 500 times more effectively than NaCl on a molal basis is unknown. However, evidence suggests that the antifreeze may adsorb to the surface of ice crystals, thereby preventing water molecules from joining the ice lattice and thus preventing growth of the crystal (Duman and DeVries, '72; Raymond and DeVries, '72). The Meracantha antifreeze has not been characterized chemically, but i t is proteinaceous. Dialysis has shown that the antifreeze has a molecular weight greater than 3,500 daltons (Duman, '77). In addition, the odd thermal hysteresis characteristic of both Meracantha and fish antifreezes permits speculation that the Meracantha antifreeze may be similar to that of the fishes. MATERIALS AND METHODS
Larvae of M. contractawere collected in and under partially decomposed, fallen logs in mature oak forests near South Bend, Indiana. Hemolymph samples were collected by puncturing the cuticle of the insect with a 23gauge needle in the dorsal midline approximately two-thirds of the way toward the posterior. Three to six microliter of hemolymph were then collected by capillary action in one end of a 10-pl glass tube. The opposite end of the capillary was sealed in a flame. (Care was
taken not to heat the hemolymph.) The sample was then centrifuged briefly to force the hemolymph into the sealed end of the capillary. Occasionally, cloudy samples were obtained. After centrifugation the cloudiness generally disappeared and a layer of material was left a t the top of the sample. This layer was removed with a drawn-out Pasteur pipet. Finally, the capillary was sealed with mineral oil, allowing a small air space to remain between the mineral oil and the hemolymph. If the freezing and melting points could not be determined immediately, the samples were frozen a t - 20°C for subsequent analysis. The freezing and melting points of the hemolymph samples were determined by a modification of the method of Scholander and Maggert ('71). A small seed crystal is formed in the sample by spraying with a spray-freeze (Cryokwik) and the capillary is placed in a refrigerated (alcohol) viewing chamber in which the temperature is finely controlled (* 0.01OC).The seed crystal is viewed through a microscope. The temparature of the bath is raised O.O2OC/5 minutes until the crystal disappears. The temperature a t which the crystal disappears is the melting point of the sample. Another crystal is then spray-frozen in the capillary and the temperature lowered O.l0C/5 minutes until the crystal begins to grow. This temperature is taken as the freezing point of the sample. With this technique, hemolymph from summer larvae, without a macromolecular antifreeze, or an aqueous solution that contains only "ideal" solutes, such as NaC1, exhibit freezing and melting points that are virtually identical (within 0.02"C), as theory predicts. However, in hemolymph that contains the macromolecular antifreeze, the freezing point may be depressed several degrees below the melting point. If Meracantha antifreeze behaves like fish antifreeze, measurement of the amount of thermal hysteresis present in a hemolymph sample should permit a n estimate of the levels of antifreeze in the sample since the amount of thermal hysteresis in a solution containing the fish antifreezes varies directly with antifreeze concentrations (DeVries, '71; Duman and DeVries, '74, '76). The supercooling point, the temperature a t which spontaneous nucleation occurs when the insect is supercooled below its freezing point, of larval Meracantha contracta was determined with a thermoelectric technique.
VARIATIONS IN MACROMOLECULAR ANTIFREEZE
A thermistor was attached with paraffin to the cuticle in the mid-dorsal region of the larvae. The termistor was connected to a YSI Model 42SC telethermometer and a recorder. The larva was placed in a freezing chamber and the temperature was lowered a t a rate of 0.4"CIminute. The supercooling point was easily detected as the temperature a t which a rapid increase in the body temperature of the insect occurred due to the release of the heat of fusion with the freezing of the body fluids (Salt, '66). Acclimation experiments were conducted in a Controlled Environments Inc. walk-in environmental chamber and in a Precision Scientific Model 805 incubator. Cold-acclimation of larvae collected in early September was conducted according to the following temperature schedule. The larvae were taken from room temperature to + 10°C in two days. Then the larvae were held a t 10°C for five days after which they were taken to 0°C over the course of the next ten days. The larvae were then held a t 0°C for the duration of the acclimation. The photoperiod was 12Ll12D. A control group of larvae was held a t 22"C, and a photoperiod of 12LI12D. Larvae collected in mid-February were warm-acclimated by allowing them to warm to room temperature and then placing them in an acclimation chamber a t 25°C. The photoperiod was 12LI 12D. Air temperature data for the South Bend area were received from the National Oceanic and Atmospheric Administration's Environmental Data Service; and they were collected a t the National Weather Service Office at the local airport. RESULTS
The freezing behavior of overwintering Meracanthu hemolymph is similar to that of the blood serum of teleost fishes that produce a macromolecular antifreeze. As the temperature of the sample is lowered slowly from the melting point temperature, no growth of the seed crystal occurs until the freezing point, which may be several degrees lower than the melting point, is reached, whereupon the crystal begins to grow rapidly by the production of monocline spears that grow rapidly down the length of the capillary tube. This is in contrast to the freezing behavior of a NaCl solution or of summer Meracanthu hemolymph that does not contain a macromolec-
87
ular antifreeze. In this case, as the temperature of the sample is slowly lowered from the melting temperature slow growth of the seed crystal is almost immediately observed, even if the decrease is only 0.02"C. Figure 1A shows the minimum monthly air temperatures recorded for the South Bend area during the period from August 1975 to July 1976. These data indicate the general trend in temperature during the year and show the minimum environmental temperatures which Meracantha might experience a t various times of the year. Temperatures inside the logs in which Meracantha were collected were not measured. However, because of the thermal buffering provided by the log and by snow cover in winter, the temperatures inside the log should generally have been higher than the minimum air temperatures recorded in figure 1A (Baust, '76). Figure 1B shows the amount of variation between the freezing and melting points (thermal hysteresis) of the hemolymph of larval Meracantha contracta during the year. As stated previously, the amount of thermal hysteresis indicates the amount of macromolecular antifreeze present in the hemolymph. There was no thermal hysteresis present in the larval hemolymph in August and early September. However, by late September small amounts of hysteresis appeared and by midNovember substantial levels of antifreeze were present. Antifreeze levels appear to become maximal in mid-February. However, overwintering larvae were not collected during the 2-month period from mid-December to mid-February. Therefore maximum levels of antifreeze may have occurred before midFebruary. A slow decline in the amount of thermal hysteresis in the hemolymph occurred throughout the early spring. In early June a more rapid decrease took place and by June 11there was no antifreeze detectable in the hemolymph. Figure 1C illustrates the variation in larval supercooling points over part of the year. They were lowest in mid-February when antifreeze levels were highest. As antifreeze levels decreased in the spring the supercooling points increased. Figure 2 shows the effects of cold-acclimation for 60 days on antifreeze production in Meracantha larvae collected in early September. During the first half of the acclimation only low levels of antifreeze were produced.
88
JOHN G. DUMAN I
A
B -5
-4
-3
-2
-1
c
I
C c
-5
-1 0 I
1
I
8
1
Dee Jan Feb Mar Apr May June July Time of Year Fig. 1 Variations in A: Minimum monthly air temperatures; B: freezing points ( 0 )and melting p i n t s ( 0 )showing Jg
Sept
Oct
Nov
the amount of thermal hysteresis (indicative of macromolecular antifreeze levels) in the hemolymph of Meracantha larvae over the time period from August 1975 to July 1976and C: supercoolingpoints of larvae over a period from December 1975 to July 1976.Values in B and C indicate the mean f standard deviation. Standard deviations less than 0.OB0C are not indicated.
89
VARIATIONS IN MACROMOLECULAR ANTIFREEZE
T
1 ,
I
lo
20 30 Time (Days)
40
$0
60
J
Fig. 2 Variations in the freezing points ( 0 )and melting points ( 0 )showing the amount of thermal hysteresis (indicative of macromolecularantifreeze levels) in the hemolymph of Mercantha larvae collected in early Septemberand cold-acclimated.Values indicate the mean It standard deviation. Standard deviations less than 0.08°C are not indicated.
This was followed by a rapid increase during the later half of the acclimation. The acclimation was terminated after 60 days. Control larvae held at 22°C for 60 days produced a significantly lesser amount of thermal hysteresis (1.22 It 0.34"C) than did the cold acclimated larvae after 60 days (2.32 It 0.24). Figure 3 illustrates the effect of warmacclimation (25°C) on larvae collected in midFebruary. The amount of thermal hysteresis decreased relatively quickly and within 12 days there was no antifreeze detectable in the hemolymph. DISCUSSION
The production of antifreeze by the wild population of Meracantha during the autumn began with the onset of cold weather. Similarly, cold-acclimation of larvae in late summer induced antifreeze production. However, environmental factors other than temperature may be involved in initiating antifreeze production. This is suggested by the finding that the control larvae held a t 22°C exhibited thermal hysteresis (122°C) in the hemolymph when they were assayed after 60 days. Although this control group did not produce as
much thermal hysteresis as did the cold-acclimated group (2.32"C) i t is interesting that the control larvae produced any antifreeze whatsoever. The photoperiod was identical for both groups (12L/12D). It may be that some, more subtle environmental factor(s) is important in initiating antifreeze production, especially in the absence of temperature variation. I t is also possible that a n internal clock is involved. Further studies into the initiation of antifreeze production in Meracantha are needed. Loss of antifreeze by the wild population of Meracantha in the spring, and by the warmacclimated larvae collected in mid-February coincided with the increasing temperatures. The loss of antifreeze in the wild population was a fairly slow process compared to the more rapid loss of antifreeze (12 days) exhibited by the warm-acclimated larvae. This is perhaps explainable by the sustained, high (25°C) temperature under which the larvae were acclimated. Actually in early June a rapid loss of antifreeze (over a 2-week period) also occurred in the wild population. In the spring of 1975 a similar rapid loss of antifreeze in the larvae occurred, but approx-
90
JOHN G . DUMAN
-6.0.
-5.0
-4.0
-
. O -3.0 v
?? 3
c
2 a,
Q
E, -2.0 I-
-1.0
1
2
3
4
5 6 Time (Days)
7
e
4
l 0 l i
ti
13'
Fig. 3 Variations in the freezing points ( 0 ) and melting points ( 0 ) showing the amount of thermal hysteresis (indicativeof macromolecular antifreeze levels) in the hemolymph of Meracuntha larvae collected in mid-February and warm-acclimated. Values indicate the mean 2 standard deviation. Standard deviations less than 0.08'C are not indicated.
imately two weeks earlier than in 1976 (Duman, unpublished data). The slow decline in antifreeze concentrations prior to the onset of sustained high temperatures is functional since, even though a general warming trend occurs in the spring, subzero temperatures can, and often do, occur in the South Bend area well into May (fig. 1A). Figure 1C shows that the supercooling points of Meracanthawere lowest during those periods when antifreeze levels were high. This depression of the supercooling point is important to the overwintering capabilities because the larvae are unable to survive ice formation in their body fluids (non-freeze tolerant). However, supercooling is a probabilistic event (Salt, '66) and because Merueantha may experience freezing temperatures for long pe-
riod of time, freezing and therefore death may occur in nature a t temperatures above the experimentally determined supercooling points listed in figure 1C. Therefore, the depression of the freezing point by the antifreeze may also be of considerable survival value. In addition, the antifreeze may hinder inoculative freezing (Salt, '63) which could be a problem in damp hibernaculae. The minimum monthly temperatures that occurred during mid-winter were several degrees below the supercooling points of the wild population (fig. 1). Therefore, since Meracanth is not freeze tolerant, it might be expected that these low environmental temperatures would have resulted in the death of many larvae. However, dead Meracanth were never discovered in the decomposing logs in which the larvae were
VARIATIONS IN MACROMOLECULAR ANTIFREEZE
hibernating. The answer most probably lies in the insulation from air temperatures provided by both the log and by any snow cover that may have been present (Baust, '76). Other than Meracantha the only known case among insects in which a factor that induces a thermal hysteresis in body fluids is produced is in the larvae of the mealworm, Tenebrio molitor (Ramsay, '64). Ramsay in his fine study of the cryptonephridial rectal complex of Tenebrio found a thermal hysteresis in fluid from the anterior and posterior perinephric space and occasionally in the hemolymph. The significance of this finding was not speculated upon by Ramsay although the possibility exists that the factor, later identified as a protein with a molecular weight of 10,000-12,000 daltons (Grimstone et al., '681, may be of some significance to the water reabsorption processes of the rectal complex (Wall, '71). Preliminary studies (Patterson and Duman, unpublished data) have confirmed that thermal hysteresis is sometimes present in the hemolymph of Tenebrio held at room temperature. However, cold-acclimation induces much greater thermal hysteresis (approximately 2.5"C). Associated with this increased thermal hysteresis is a decreased supercooling point in the cold-acclimated larvae. These preliminary data suggest that a function of the thermal hysteresis producing factor in Tenebrio larvae may be to enhance low temperature tolerance. Under natural conditions Tenebrio overwinter in the larval state (Metcalf and Flint, '62). The reliance on macromolecular antifreeze to achieve low-temperature tolerance does not seem to be widespread among insects in the South Bend area. Of 11overwintering species collected and checked for antifreeze during their hibernation in logs, under bark, etc., only Merueantha exhibited a thermal hysteresis in the hemolymph. It should also be noted, however, that of these 11 overwintering species only the overwintering adult queens of the bald-faced hornet, Vespula maculata (Hymenoptera, Vespidae) had high levels of glycerol (Duman, unpublished data). Actually, the means of low temperature tolerance of many insects is still unexplained (Asahina, '69; Semme, '64; Baust and Morrissey, '75). The macromolecular antifreeze of Meracantha has one obvious disadvantage when compared to the small molecular weight polyols, such as glycerol, used by some over-
91
wintering insects to achieve low temperature tolerance. Although both the polyols and the macromolecular antifreeze depress the supercooling point of the insect, the macromolecular antifreeze of Meracantha does not act as a cryoprotectant to protect the insect from freeze-damage, if freezing of body fluids should occur. The polyhydric alcohols are well known cryoprotectants. However, there may be certain advantages to the use of the macromolecular antifreeze. Production of the macromolecular antifreeze is not accompanied by a large increase in osmotic pressure of the body fluids, as evidenced by the occurrence of only a small decrease in the melting point of the hemolymph of overwintering Meracantha. However, the several molal concentrations of glycerol produced by some overwintering insects increases the osmotic pressure by several thousand milliosmols. The osmotic pressure of the body fluids of most animals, including insects, is normally closely regulated (Stobbant and Shaw, '74; Florkin and Jeuniaux, '74). The tremendous increases in osmotic pressure resulting from high glycerol concentrations may require considerable physiological and biochemical compensation. Macromolecular antifreeze has now been found in many species of cold-water marine teleosts and in two species of insects. In addition a n Antartic limpet, Patinigera polaris, secretes a n envelope of mucus around itself when i t is caught in the near-shore ice. An unknown factor in the mucus produces a thermal hysteresis, and ice propagation across the mucus is thereby retarded down to -10°C (Hargens, '73). In a recent study, a group of thermal hysteresis producing glycoproteins was isolated from the intertidal mussel, Mytilus edulis (Theede et al., '76). It was suggested that these glycoproteins contribute to the mechanism of freezing resistance in the mussel. The presence of a thermal hysteresis producing factor in three such phylogenetically divergent groups (teleost fishes, insects and molluscs) raises the question of the possible importance of macromolecular antifreezes to the low temperature tolerance of other poikilothermic animals. ACKNOWLEDGMENTS
This investigation was supported by Biomedical Sciences Support Grants FR/RR07033-09 and FR/RR-07033-10 and by the National Science Foundation.
92
JOHN G. DUMAN
LITERATURE CITED Asahina, E. 1966 Freezing and frost resistance in insects. In: Cryobiology. H. T. Meryman, ed. Academic Press, New York, pp. 451-494. 1969 Frost resistance in insects. In: Advances in insect Physiology. J . W.L. Beament, J . E. Treherme, and V. B. Wigglesworth, eds. Academic Press, New York, pp. 1-49. Baust, J. G. 1973 Mechanisms of Cryoprotection in freezing tolerant animal systems. Cryobiol., 10: 197-205. 1976 Temperature buffering in arctic microhabitats. Ann. Ent. SOC. Amer., 69: 117-119. Baust, J. G., and L. K. Miller 1970 Variations in glycerol content and its influence on cold hardiness in the Alaskan Carabid beetle, Pterostichus brevicornis. J. Insect. Physiol., 16: 979-990. 1972 Insect freezing protection in Pterostichus brevicornis (Carabidae). Nature, 236: 219-221. Baust, J . G., and R. E. Morrissey 1975 Supercooling phenomenon and water content independence in the overwintering beetle, Coleomgilla maculata. J. Insect Physiol., 21: 1751-1754. Chino, H. 1957 Conversion of glycogen to sorbitol and glycerol in the diapause egg of the Bombyx Silkworm. Nature, 180: 606-607. 1958 Carbohydrate metabolism in the diapause egg of the silkworm, Eombyz mori. 11. Conversion of glycogen into sorbitol and glycerol during diapause. J. Insect Physiol., 2: 1-12. DeVries, A. L. 1971 Glycoproteins as biological antifreeze agents in Antarctic fishes. Science, 172: 11521155. DeVries, A. L., S. K. Komatsu and R. E. Feeney 1970 Chemical and physical properties of freezingpointdepressingglycoproteins from Antarctic fishes. J. Biol. Chem., 245: 2901-2908. DeVries, A. L., J . Vandenheede and R. E. Feeney 1971 Primary structure of freezing point depressing glycoproteins. J. Biol. Chem., 246: 305-308. DeVries, A. L., and D. E. Wohlschlag 1969 Freezing resistance in some Antarctic fishes. Science, 163: 1073-1075. Duman, J. G. 1977 Macromolecular antifreeze: Low temperature tolerance in the darkling beetle, Meracantha contracta.Jour. Comp. Physiol., 115: 279-286. Duman, J. G., and A. L. DeVries 1972 Freezing behavior of aqueous solutions of glycoproteins from the blood of an Antarctic fish. Cryobiol., 9: 469-472. 1974 Freezing resistance in the winter flounder, Pseudopkuronectes americanus. Nature, 247: 237-238. 1975 The role of macromolecular antifreezes in cold water fishes. Comp. Biochem. Physiol., 52A: 193-199. 1976 Isolation characterization and physical properties of protein antifreezes from the winter flounder, Pseudopkuronectes americanus. Comp. Biochem. Physiol., 54E: 375-380. Florkin, M., and C. Jeuniaux 1975 Hemolymph: Composition. In: The Physiology of Insecta. Vol. 5.M. Rockstein, ed. Academic Press, New York, pp. 255-307. Frankos, V. H.,and A. P. Platt 1976 Glycerol accumulation and water content in larvae of Limenitis archippus:
Their importance to winter survival. J. Insect Physiol., 22: 623-628, Grimstone, A. V., A. M. Mullinger and J. A. Ramsay 1968 Further studies on the rectal complex of the mealworm, Tenebrio molitor. Phil Trans. Roy. Soc. London, 253: 343382. Hargens, A. R. 1972 Freezing resistance in polar fishes. Science, 176: 184-186. 1973 Protection against lethal freezing temperatures by mucus in an Antarctic limpet. Cryobiol.,lO: 331-337. Komatsu, S. K., A. L. DeVries and R. E. Feeney 1970 Studies on the structure of freezing point depressing glycoproteins from Antarctic fishes. J. Biol. Chem., 245: 2909-2913. Lin, Y., J. G. Duman and A. L. DeVries 1972 Studies on the structure and activity of low molecular weight glycoproteins from an Antarctic fish. Biochem. Biophys. Res. Comm., 46: 87-92. Metcaif, C. L., and W. P. Flint 1962 Destructive and Useful Insects. McGraw Hill Book Co., New York, pp. 922-923. Miller, L. K.,and J. S.Smith 1975 Production of threitol and sorbitol by an adult insect: Association with freezing tolerance. Nature, 258: 519-520. Ramsey, J. A. 1964 The rectal complex of the mealworm, Tenebrio molitor. Phil. Trans. Roy. SOC.London B., 248: 279-314. Raymond, J. A., and A.L. DeVries 1972 Freezing behavior of fish blood glycoproteins with antifreeze properties. Cryobiol., 9: 541-547. Raymond, J. A., Y. Lin and A. L. DeVries 1975 Glycoprotein and protein antifreezes in two Alaskan fishes. J. Ex@. Zool., 193: 125-130. Salt, R. W. 1957 Natural occurrence of glycerol in insects and its relation to their ability to survive freezing. Can. Ent., 89: 491-494. 1959 Role of glycerol in the cold hardiness of Eracon cephi Can. J . Zool., 37: 59-69. 1961 Principles of insect cold-hardiness. A. Rev. Ent., 6:55-74. 1963 Delayed inoculative freezing of insects. Can. Ent., 95: 1190-1202. 1966 Factors influencing nucleation in supercooled insects. Can. Ent., 95: 1190-1202. Scholander, P. F., and J. Maggert 1971 Supercooling and ice propogation in blood of Arctic fishes. Cryobiol., 8: 371374. Shier, W. T., Y. Lin and A. L. DeVries 1972 Structure and mode of action of glycoproteins from an Antarctic fish. Biochem. Biophys. Acta., 263: 406-413. S ~ m m eL. , 1964 Effects of glycerol and cold hardiness in insects. Can. J. Zool., 42: 87-101. Stobbart, R. H.,and J. Shaw 1974 Salt and water balance: Excretion. In: The Physiology of Insecta. Vol. 5. M. Rockstein, ed. Academic Press, New York, pp. 361-446. Theede, H., R. Schneppenheim and L. Beress 1976 Frostschutz-Glykoproteine bei Mytilus edulis ? Marine Biology, 36: 183-189. Wall, B. J. 1971 Local osmotic gradients in the rectal pads of an insect. Fed. Proc., 30: 42-48.
ABSTMCT Overwintering larvae of the darkling beetle, Meracantha contracts, produce a macromolecular antifreeze that is similar in activity to the glycoproteinaceous and proteinaceous antifreezes found in some cold-water, marine teleost fishes. The antifreeze is not present in the hemolymph of the Meracantha larvae in summer, but its production begins by late September in the wild population. The antifreeze reaches a maximum concentration in February, decreases slowly through spring, and disappears by early June. The supercooling points of the larvae are lowest in February, when the antifreeze levels are highest, and increase as the antifreeze concentrations in the hemolymph decrease in the spring. Larvae collected in mid-February and warm-acclimated lost the antifreeze within 12 days. Larvae collected in early September and cold-acclimated required nearly two months to produce concentrations of antifreeze comparable to those of overwintering larvae. Temperature seems to be the major environmental factor responsible for the control of antifreeze levels in Meracantha;however, other environmental factors may also be involved. The overwintering phases of many polar and temperate-zone insects are well known for their abilities to survive subzero environmental temperatures (for reviews see Asahina, '66, '69; Salt, '61, '66). Many of these insects are capable of withstanding the actual freezing of their body fluids however, most seem to survive the winter by supercooling to temperatures below the equilibrium freezing point (the freezing point based on colligative properties) of their body fluids. The body fluids of most insects, even in summer, will supercool below the equilibrium freezing point. Therefore, i t is not the freezing point itself, but rather the temperature to which the body fluids can supercool before spontaneous nucleation occurs that is important to the survival of insects that are unable to survive freezing. Many physiological and biochemical adaptations are, no doubt, involved in the low temperature tolerance of these insects, especially in those that survive freezing (Baust, '73). However, the adaptation that has been best correlated with low-temperature tolerance in a number of overwintering insects is the production of high concentrations of polyhydric alcohols, such as glycerol, sorbitol, mannitol and threitol (Salt, '57, '59; Samme, '64, Chino, '57, '58; Baust and Miller, J. EXP. ZOOL., 201: 85-92.
'70, '72; Miller and Smith, '75; Frankos and Platt, '76). These polyols lower the freezing point of the hemolymph significantly and also depress the supercooling point allowing the insect to supercool to lower temperatures without spontaneous freezing. In addition the polyols may function as cryoprotectants to prevent damage if freezing should occur. Glycerol in particular is produced a t high levels in winter in several species of insects. For example, glycerol concentrations in the overwintering larvae of the wheat stem sawfly, Bracon cephi, are as high as 5 M (Salt, '59). Although numerous studies have shown the importance of low molecular weight polyols to insect low-temperature tolerance, only one has investigated the correlation between solutes of high molecular weight and overwintering capabilities (Duman, '77). In winter, larvae of the darkling beetle, ueracantha contracts (Coleoptera, Tenebrionidae) produce a macromolecular antifreeze which depresses the supercooling point of the larvae significantly below that of the summer larvae. This antifreeze does not function a s a cryoprotectant since the overwintering larvae are unable to tolerate ice formation in their tissues. The macromolecular antifreeze produces a thermal hysteresis in the hemolymph,
85
86
JOHN G . DUMAN
whereby the freezing point is depressed several degrees (as much as 6OC in some individuals) below the melting point. The thermal hysteresis in Meracantha hemolymph is similar to that produced by macromolecular antifreeze solutes that depress the freezing point of the body fluids of many cold-water, marine teleosts below the freezing point of seawater, thereby protecting the fish from freezing. These antifreeze solutes, first discovered in certain Antarctic nototheniid species (DeVries and Wohlschlag, '69),have since been demonstrated in many phylogenetically divergent species of cold-water fishes (Scholander and Maggert, '71; Hargens, '72; Duman and DeVries, '74, '75). Purification and characterization of the antifreeze substances from species of four different families of teleosts have shown that they are either proteins or glycoproteins, depending on the species (DeVries et al., '70; DeVries et al., '71; Komatsu et al., '70; DeVries, '71; Lin et al., '72; Shier et al., '72; Raymond et al., '75; Duman and DeVries, '76). The mechanism by which the fish antifreezes depress the freezing point of water 200 to 500 times more effectively than NaCl on a molal basis is unknown. However, evidence suggests that the antifreeze may adsorb to the surface of ice crystals, thereby preventing water molecules from joining the ice lattice and thus preventing growth of the crystal (Duman and DeVries, '72; Raymond and DeVries, '72). The Meracantha antifreeze has not been characterized chemically, but i t is proteinaceous. Dialysis has shown that the antifreeze has a molecular weight greater than 3,500 daltons (Duman, '77). In addition, the odd thermal hysteresis characteristic of both Meracantha and fish antifreezes permits speculation that the Meracantha antifreeze may be similar to that of the fishes. MATERIALS AND METHODS
Larvae of M. contractawere collected in and under partially decomposed, fallen logs in mature oak forests near South Bend, Indiana. Hemolymph samples were collected by puncturing the cuticle of the insect with a 23gauge needle in the dorsal midline approximately two-thirds of the way toward the posterior. Three to six microliter of hemolymph were then collected by capillary action in one end of a 10-pl glass tube. The opposite end of the capillary was sealed in a flame. (Care was
taken not to heat the hemolymph.) The sample was then centrifuged briefly to force the hemolymph into the sealed end of the capillary. Occasionally, cloudy samples were obtained. After centrifugation the cloudiness generally disappeared and a layer of material was left a t the top of the sample. This layer was removed with a drawn-out Pasteur pipet. Finally, the capillary was sealed with mineral oil, allowing a small air space to remain between the mineral oil and the hemolymph. If the freezing and melting points could not be determined immediately, the samples were frozen a t - 20°C for subsequent analysis. The freezing and melting points of the hemolymph samples were determined by a modification of the method of Scholander and Maggert ('71). A small seed crystal is formed in the sample by spraying with a spray-freeze (Cryokwik) and the capillary is placed in a refrigerated (alcohol) viewing chamber in which the temperature is finely controlled (* 0.01OC).The seed crystal is viewed through a microscope. The temparature of the bath is raised O.O2OC/5 minutes until the crystal disappears. The temperature a t which the crystal disappears is the melting point of the sample. Another crystal is then spray-frozen in the capillary and the temperature lowered O.l0C/5 minutes until the crystal begins to grow. This temperature is taken as the freezing point of the sample. With this technique, hemolymph from summer larvae, without a macromolecular antifreeze, or an aqueous solution that contains only "ideal" solutes, such as NaC1, exhibit freezing and melting points that are virtually identical (within 0.02"C), as theory predicts. However, in hemolymph that contains the macromolecular antifreeze, the freezing point may be depressed several degrees below the melting point. If Meracantha antifreeze behaves like fish antifreeze, measurement of the amount of thermal hysteresis present in a hemolymph sample should permit a n estimate of the levels of antifreeze in the sample since the amount of thermal hysteresis in a solution containing the fish antifreezes varies directly with antifreeze concentrations (DeVries, '71; Duman and DeVries, '74, '76). The supercooling point, the temperature a t which spontaneous nucleation occurs when the insect is supercooled below its freezing point, of larval Meracantha contracta was determined with a thermoelectric technique.
VARIATIONS IN MACROMOLECULAR ANTIFREEZE
A thermistor was attached with paraffin to the cuticle in the mid-dorsal region of the larvae. The termistor was connected to a YSI Model 42SC telethermometer and a recorder. The larva was placed in a freezing chamber and the temperature was lowered a t a rate of 0.4"CIminute. The supercooling point was easily detected as the temperature a t which a rapid increase in the body temperature of the insect occurred due to the release of the heat of fusion with the freezing of the body fluids (Salt, '66). Acclimation experiments were conducted in a Controlled Environments Inc. walk-in environmental chamber and in a Precision Scientific Model 805 incubator. Cold-acclimation of larvae collected in early September was conducted according to the following temperature schedule. The larvae were taken from room temperature to + 10°C in two days. Then the larvae were held a t 10°C for five days after which they were taken to 0°C over the course of the next ten days. The larvae were then held a t 0°C for the duration of the acclimation. The photoperiod was 12Ll12D. A control group of larvae was held a t 22"C, and a photoperiod of 12LI12D. Larvae collected in mid-February were warm-acclimated by allowing them to warm to room temperature and then placing them in an acclimation chamber a t 25°C. The photoperiod was 12LI 12D. Air temperature data for the South Bend area were received from the National Oceanic and Atmospheric Administration's Environmental Data Service; and they were collected a t the National Weather Service Office at the local airport. RESULTS
The freezing behavior of overwintering Meracanthu hemolymph is similar to that of the blood serum of teleost fishes that produce a macromolecular antifreeze. As the temperature of the sample is lowered slowly from the melting point temperature, no growth of the seed crystal occurs until the freezing point, which may be several degrees lower than the melting point, is reached, whereupon the crystal begins to grow rapidly by the production of monocline spears that grow rapidly down the length of the capillary tube. This is in contrast to the freezing behavior of a NaCl solution or of summer Meracanthu hemolymph that does not contain a macromolec-
87
ular antifreeze. In this case, as the temperature of the sample is slowly lowered from the melting temperature slow growth of the seed crystal is almost immediately observed, even if the decrease is only 0.02"C. Figure 1A shows the minimum monthly air temperatures recorded for the South Bend area during the period from August 1975 to July 1976. These data indicate the general trend in temperature during the year and show the minimum environmental temperatures which Meracantha might experience a t various times of the year. Temperatures inside the logs in which Meracantha were collected were not measured. However, because of the thermal buffering provided by the log and by snow cover in winter, the temperatures inside the log should generally have been higher than the minimum air temperatures recorded in figure 1A (Baust, '76). Figure 1B shows the amount of variation between the freezing and melting points (thermal hysteresis) of the hemolymph of larval Meracantha contracta during the year. As stated previously, the amount of thermal hysteresis indicates the amount of macromolecular antifreeze present in the hemolymph. There was no thermal hysteresis present in the larval hemolymph in August and early September. However, by late September small amounts of hysteresis appeared and by midNovember substantial levels of antifreeze were present. Antifreeze levels appear to become maximal in mid-February. However, overwintering larvae were not collected during the 2-month period from mid-December to mid-February. Therefore maximum levels of antifreeze may have occurred before midFebruary. A slow decline in the amount of thermal hysteresis in the hemolymph occurred throughout the early spring. In early June a more rapid decrease took place and by June 11there was no antifreeze detectable in the hemolymph. Figure 1C illustrates the variation in larval supercooling points over part of the year. They were lowest in mid-February when antifreeze levels were highest. As antifreeze levels decreased in the spring the supercooling points increased. Figure 2 shows the effects of cold-acclimation for 60 days on antifreeze production in Meracantha larvae collected in early September. During the first half of the acclimation only low levels of antifreeze were produced.
88
JOHN G. DUMAN I
A
B -5
-4
-3
-2
-1
c
I
C c
-5
-1 0 I
1
I
8
1
Dee Jan Feb Mar Apr May June July Time of Year Fig. 1 Variations in A: Minimum monthly air temperatures; B: freezing points ( 0 )and melting p i n t s ( 0 )showing Jg
Sept
Oct
Nov
the amount of thermal hysteresis (indicative of macromolecular antifreeze levels) in the hemolymph of Meracantha larvae over the time period from August 1975 to July 1976and C: supercoolingpoints of larvae over a period from December 1975 to July 1976.Values in B and C indicate the mean f standard deviation. Standard deviations less than 0.OB0C are not indicated.
89
VARIATIONS IN MACROMOLECULAR ANTIFREEZE
T
1 ,
I
lo
20 30 Time (Days)
40
$0
60
J
Fig. 2 Variations in the freezing points ( 0 )and melting points ( 0 )showing the amount of thermal hysteresis (indicative of macromolecularantifreeze levels) in the hemolymph of Mercantha larvae collected in early Septemberand cold-acclimated.Values indicate the mean It standard deviation. Standard deviations less than 0.08°C are not indicated.
This was followed by a rapid increase during the later half of the acclimation. The acclimation was terminated after 60 days. Control larvae held at 22°C for 60 days produced a significantly lesser amount of thermal hysteresis (1.22 It 0.34"C) than did the cold acclimated larvae after 60 days (2.32 It 0.24). Figure 3 illustrates the effect of warmacclimation (25°C) on larvae collected in midFebruary. The amount of thermal hysteresis decreased relatively quickly and within 12 days there was no antifreeze detectable in the hemolymph. DISCUSSION
The production of antifreeze by the wild population of Meracantha during the autumn began with the onset of cold weather. Similarly, cold-acclimation of larvae in late summer induced antifreeze production. However, environmental factors other than temperature may be involved in initiating antifreeze production. This is suggested by the finding that the control larvae held a t 22°C exhibited thermal hysteresis (122°C) in the hemolymph when they were assayed after 60 days. Although this control group did not produce as
much thermal hysteresis as did the cold-acclimated group (2.32"C) i t is interesting that the control larvae produced any antifreeze whatsoever. The photoperiod was identical for both groups (12L/12D). It may be that some, more subtle environmental factor(s) is important in initiating antifreeze production, especially in the absence of temperature variation. I t is also possible that a n internal clock is involved. Further studies into the initiation of antifreeze production in Meracantha are needed. Loss of antifreeze by the wild population of Meracantha in the spring, and by the warmacclimated larvae collected in mid-February coincided with the increasing temperatures. The loss of antifreeze in the wild population was a fairly slow process compared to the more rapid loss of antifreeze (12 days) exhibited by the warm-acclimated larvae. This is perhaps explainable by the sustained, high (25°C) temperature under which the larvae were acclimated. Actually in early June a rapid loss of antifreeze (over a 2-week period) also occurred in the wild population. In the spring of 1975 a similar rapid loss of antifreeze in the larvae occurred, but approx-
90
JOHN G . DUMAN
-6.0.
-5.0
-4.0
-
. O -3.0 v
?? 3
c
2 a,
Q
E, -2.0 I-
-1.0
1
2
3
4
5 6 Time (Days)
7
e
4
l 0 l i
ti
13'
Fig. 3 Variations in the freezing points ( 0 ) and melting points ( 0 ) showing the amount of thermal hysteresis (indicativeof macromolecular antifreeze levels) in the hemolymph of Meracuntha larvae collected in mid-February and warm-acclimated. Values indicate the mean 2 standard deviation. Standard deviations less than 0.08'C are not indicated.
imately two weeks earlier than in 1976 (Duman, unpublished data). The slow decline in antifreeze concentrations prior to the onset of sustained high temperatures is functional since, even though a general warming trend occurs in the spring, subzero temperatures can, and often do, occur in the South Bend area well into May (fig. 1A). Figure 1C shows that the supercooling points of Meracanthawere lowest during those periods when antifreeze levels were high. This depression of the supercooling point is important to the overwintering capabilities because the larvae are unable to survive ice formation in their body fluids (non-freeze tolerant). However, supercooling is a probabilistic event (Salt, '66) and because Merueantha may experience freezing temperatures for long pe-
riod of time, freezing and therefore death may occur in nature a t temperatures above the experimentally determined supercooling points listed in figure 1C. Therefore, the depression of the freezing point by the antifreeze may also be of considerable survival value. In addition, the antifreeze may hinder inoculative freezing (Salt, '63) which could be a problem in damp hibernaculae. The minimum monthly temperatures that occurred during mid-winter were several degrees below the supercooling points of the wild population (fig. 1). Therefore, since Meracanth is not freeze tolerant, it might be expected that these low environmental temperatures would have resulted in the death of many larvae. However, dead Meracanth were never discovered in the decomposing logs in which the larvae were
VARIATIONS IN MACROMOLECULAR ANTIFREEZE
hibernating. The answer most probably lies in the insulation from air temperatures provided by both the log and by any snow cover that may have been present (Baust, '76). Other than Meracantha the only known case among insects in which a factor that induces a thermal hysteresis in body fluids is produced is in the larvae of the mealworm, Tenebrio molitor (Ramsay, '64). Ramsay in his fine study of the cryptonephridial rectal complex of Tenebrio found a thermal hysteresis in fluid from the anterior and posterior perinephric space and occasionally in the hemolymph. The significance of this finding was not speculated upon by Ramsay although the possibility exists that the factor, later identified as a protein with a molecular weight of 10,000-12,000 daltons (Grimstone et al., '681, may be of some significance to the water reabsorption processes of the rectal complex (Wall, '71). Preliminary studies (Patterson and Duman, unpublished data) have confirmed that thermal hysteresis is sometimes present in the hemolymph of Tenebrio held at room temperature. However, cold-acclimation induces much greater thermal hysteresis (approximately 2.5"C). Associated with this increased thermal hysteresis is a decreased supercooling point in the cold-acclimated larvae. These preliminary data suggest that a function of the thermal hysteresis producing factor in Tenebrio larvae may be to enhance low temperature tolerance. Under natural conditions Tenebrio overwinter in the larval state (Metcalf and Flint, '62). The reliance on macromolecular antifreeze to achieve low-temperature tolerance does not seem to be widespread among insects in the South Bend area. Of 11overwintering species collected and checked for antifreeze during their hibernation in logs, under bark, etc., only Merueantha exhibited a thermal hysteresis in the hemolymph. It should also be noted, however, that of these 11 overwintering species only the overwintering adult queens of the bald-faced hornet, Vespula maculata (Hymenoptera, Vespidae) had high levels of glycerol (Duman, unpublished data). Actually, the means of low temperature tolerance of many insects is still unexplained (Asahina, '69; Semme, '64; Baust and Morrissey, '75). The macromolecular antifreeze of Meracantha has one obvious disadvantage when compared to the small molecular weight polyols, such as glycerol, used by some over-
91
wintering insects to achieve low temperature tolerance. Although both the polyols and the macromolecular antifreeze depress the supercooling point of the insect, the macromolecular antifreeze of Meracantha does not act as a cryoprotectant to protect the insect from freeze-damage, if freezing of body fluids should occur. The polyhydric alcohols are well known cryoprotectants. However, there may be certain advantages to the use of the macromolecular antifreeze. Production of the macromolecular antifreeze is not accompanied by a large increase in osmotic pressure of the body fluids, as evidenced by the occurrence of only a small decrease in the melting point of the hemolymph of overwintering Meracantha. However, the several molal concentrations of glycerol produced by some overwintering insects increases the osmotic pressure by several thousand milliosmols. The osmotic pressure of the body fluids of most animals, including insects, is normally closely regulated (Stobbant and Shaw, '74; Florkin and Jeuniaux, '74). The tremendous increases in osmotic pressure resulting from high glycerol concentrations may require considerable physiological and biochemical compensation. Macromolecular antifreeze has now been found in many species of cold-water marine teleosts and in two species of insects. In addition a n Antartic limpet, Patinigera polaris, secretes a n envelope of mucus around itself when i t is caught in the near-shore ice. An unknown factor in the mucus produces a thermal hysteresis, and ice propagation across the mucus is thereby retarded down to -10°C (Hargens, '73). In a recent study, a group of thermal hysteresis producing glycoproteins was isolated from the intertidal mussel, Mytilus edulis (Theede et al., '76). It was suggested that these glycoproteins contribute to the mechanism of freezing resistance in the mussel. The presence of a thermal hysteresis producing factor in three such phylogenetically divergent groups (teleost fishes, insects and molluscs) raises the question of the possible importance of macromolecular antifreezes to the low temperature tolerance of other poikilothermic animals. ACKNOWLEDGMENTS
This investigation was supported by Biomedical Sciences Support Grants FR/RR07033-09 and FR/RR-07033-10 and by the National Science Foundation.
92
JOHN G. DUMAN
LITERATURE CITED Asahina, E. 1966 Freezing and frost resistance in insects. In: Cryobiology. H. T. Meryman, ed. Academic Press, New York, pp. 451-494. 1969 Frost resistance in insects. In: Advances in insect Physiology. J . W.L. Beament, J . E. Treherme, and V. B. Wigglesworth, eds. Academic Press, New York, pp. 1-49. Baust, J. G. 1973 Mechanisms of Cryoprotection in freezing tolerant animal systems. Cryobiol., 10: 197-205. 1976 Temperature buffering in arctic microhabitats. Ann. Ent. SOC. Amer., 69: 117-119. Baust, J. G., and L. K. Miller 1970 Variations in glycerol content and its influence on cold hardiness in the Alaskan Carabid beetle, Pterostichus brevicornis. J. Insect. Physiol., 16: 979-990. 1972 Insect freezing protection in Pterostichus brevicornis (Carabidae). Nature, 236: 219-221. Baust, J . G., and R. E. Morrissey 1975 Supercooling phenomenon and water content independence in the overwintering beetle, Coleomgilla maculata. J. Insect Physiol., 21: 1751-1754. Chino, H. 1957 Conversion of glycogen to sorbitol and glycerol in the diapause egg of the Bombyx Silkworm. Nature, 180: 606-607. 1958 Carbohydrate metabolism in the diapause egg of the silkworm, Eombyz mori. 11. Conversion of glycogen into sorbitol and glycerol during diapause. J. Insect Physiol., 2: 1-12. DeVries, A. L. 1971 Glycoproteins as biological antifreeze agents in Antarctic fishes. Science, 172: 11521155. DeVries, A. L., S. K. Komatsu and R. E. Feeney 1970 Chemical and physical properties of freezingpointdepressingglycoproteins from Antarctic fishes. J. Biol. Chem., 245: 2901-2908. DeVries, A. L., J . Vandenheede and R. E. Feeney 1971 Primary structure of freezing point depressing glycoproteins. J. Biol. Chem., 246: 305-308. DeVries, A. L., and D. E. Wohlschlag 1969 Freezing resistance in some Antarctic fishes. Science, 163: 1073-1075. Duman, J. G. 1977 Macromolecular antifreeze: Low temperature tolerance in the darkling beetle, Meracantha contracta.Jour. Comp. Physiol., 115: 279-286. Duman, J. G., and A. L. DeVries 1972 Freezing behavior of aqueous solutions of glycoproteins from the blood of an Antarctic fish. Cryobiol., 9: 469-472. 1974 Freezing resistance in the winter flounder, Pseudopkuronectes americanus. Nature, 247: 237-238. 1975 The role of macromolecular antifreezes in cold water fishes. Comp. Biochem. Physiol., 52A: 193-199. 1976 Isolation characterization and physical properties of protein antifreezes from the winter flounder, Pseudopkuronectes americanus. Comp. Biochem. Physiol., 54E: 375-380. Florkin, M., and C. Jeuniaux 1975 Hemolymph: Composition. In: The Physiology of Insecta. Vol. 5.M. Rockstein, ed. Academic Press, New York, pp. 255-307. Frankos, V. H.,and A. P. Platt 1976 Glycerol accumulation and water content in larvae of Limenitis archippus:
Their importance to winter survival. J. Insect Physiol., 22: 623-628, Grimstone, A. V., A. M. Mullinger and J. A. Ramsay 1968 Further studies on the rectal complex of the mealworm, Tenebrio molitor. Phil Trans. Roy. Soc. London, 253: 343382. Hargens, A. R. 1972 Freezing resistance in polar fishes. Science, 176: 184-186. 1973 Protection against lethal freezing temperatures by mucus in an Antarctic limpet. Cryobiol.,lO: 331-337. Komatsu, S. K., A. L. DeVries and R. E. Feeney 1970 Studies on the structure of freezing point depressing glycoproteins from Antarctic fishes. J. Biol. Chem., 245: 2909-2913. Lin, Y., J. G. Duman and A. L. DeVries 1972 Studies on the structure and activity of low molecular weight glycoproteins from an Antarctic fish. Biochem. Biophys. Res. Comm., 46: 87-92. Metcaif, C. L., and W. P. Flint 1962 Destructive and Useful Insects. McGraw Hill Book Co., New York, pp. 922-923. Miller, L. K.,and J. S.Smith 1975 Production of threitol and sorbitol by an adult insect: Association with freezing tolerance. Nature, 258: 519-520. Ramsey, J. A. 1964 The rectal complex of the mealworm, Tenebrio molitor. Phil. Trans. Roy. SOC.London B., 248: 279-314. Raymond, J. A., and A.L. DeVries 1972 Freezing behavior of fish blood glycoproteins with antifreeze properties. Cryobiol., 9: 541-547. Raymond, J. A., Y. Lin and A. L. DeVries 1975 Glycoprotein and protein antifreezes in two Alaskan fishes. J. Ex@. Zool., 193: 125-130. Salt, R. W. 1957 Natural occurrence of glycerol in insects and its relation to their ability to survive freezing. Can. Ent., 89: 491-494. 1959 Role of glycerol in the cold hardiness of Eracon cephi Can. J . Zool., 37: 59-69. 1961 Principles of insect cold-hardiness. A. Rev. Ent., 6:55-74. 1963 Delayed inoculative freezing of insects. Can. Ent., 95: 1190-1202. 1966 Factors influencing nucleation in supercooled insects. Can. Ent., 95: 1190-1202. Scholander, P. F., and J. Maggert 1971 Supercooling and ice propogation in blood of Arctic fishes. Cryobiol., 8: 371374. Shier, W. T., Y. Lin and A. L. DeVries 1972 Structure and mode of action of glycoproteins from an Antarctic fish. Biochem. Biophys. Acta., 263: 406-413. S ~ m m eL. , 1964 Effects of glycerol and cold hardiness in insects. Can. J. Zool., 42: 87-101. Stobbart, R. H.,and J. Shaw 1974 Salt and water balance: Excretion. In: The Physiology of Insecta. Vol. 5. M. Rockstein, ed. Academic Press, New York, pp. 361-446. Theede, H., R. Schneppenheim and L. Beress 1976 Frostschutz-Glykoproteine bei Mytilus edulis ? Marine Biology, 36: 183-189. Wall, B. J. 1971 Local osmotic gradients in the rectal pads of an insect. Fed. Proc., 30: 42-48.
