A rapid cold-hardening response protecting against cold shock injury in Drosophila melanogaster

Overwintering insects rely on a variety of physiological and biochemical adaptations to survive low temperature exposure (Baust and Rojas, 1985; Cannon and Block, 1988; Duman and Horwath, 1983; Lee, 1989; Ring, 1981; Storey and Storey, 1988; Zachariassen, 1985). These include the accumulation of glycerol and trehalose, and other low molecular weight polyols and sugars; the synthesis of antifreeze proteins and ice-nucleating agents; and changes in whole-body supercooling points (SCP), the temperature at which spontaneous nucleation of body fluids occurs (Salt, 1961). A species is described as freeze-tolerant if it survives below the SCP and freeze-intolerant if it does not survive extracellular ice formation. Although the SCP is sometimes used as a measure of the lower lethal temperature for freeze-intolerant species, lethal injury may occur at temperatures 10–15°C above the SCP (Ring, 1980; Lee and Denlinger, 1985; Knight et al. 1986; Bale, 1987).

Generally, studies of insect cold-hardiness have focused on the capacity for low-temperature survival for extended periods during the winter (Baust and Rojas, 1985; Somme, 1982; Turnock et al. 1983; Cannon and Block, 1988). Coldhardening in preparation for winter is triggered by exposure to low environmental temperatures (Baust and Lee, 1982) or, sometimes, photoperiodic cues (Horwath and Duman, 1982). In contrast, recent reports describe a novel rapid coldhardening response that occurs within minutes, even in non-overwintering insects, and protects against a form of non-freezing injury called cold shock (Chen et al. 1987; Lee et al. 1987). Cold shock, thermal shock or direct chilling injury is a form of cellular injury which occurs after rapid cooling, but in the absence of ice formation (Morris et al. 1983). Cold shock occurs in bacteria, protozoa, algae, fungi, higher plants, fish and mammalian somatic cells and spermatozoa (Morris and Watson, 1984).

The rapid cold-hardening response is distinctly different from winter coldhardening (Lee, 1989). Winter cold-hardening is a seasonal phenomenon that generally occurs in an inactive or diapausing stage, while the rapid cold-hardening response occurs even in feeding and reproductively active insects throughout the year. Studies of overwintering cold-hardiness typically classify insects as being freeze-tolerant or not, based on their ability to survive cooling to below the SCP. In contrast, the rapid cold-hardening response occurs at temperatures 10–15°C above the temperature at which body fluids freeze. In preparation for winter the cold-hardening process requires days or weeks, while the rapid cold-hardening process associated with cold shock involves only minutes or hours.

We report on the presence of the rapid cold-hardening response protecting against cold shock in the fruit fly D. melanogaster. Understanding the underlying mechanisms of survival of this dipteran at subzero temperatures is of particular importance since, at present, there is no system for the long-term cryopreservation of individuals of this species. The ability to cryopreserve Drosophila would save on the high cost of maintaining the many thousands of strains of this insect, as well as reduce the chance of stock loss due to either human error or change in genetic composition.

This study investigated the phenomenon of cold shock and rapid cold-hardening in larval, pupal and adult stages of D. melanogaster. Our specific objectives were: (1) to determine the lower lethal temperature and its relationship to the supercooling point, (2) to examine the effect of chilling at 5 °C on survival of subzero exposure, and (3) to investigate the effect of adult age on the capacity to cold-harden rapidly.

Drosophila melanogaster (Oregon R strain) (Diptera: Drosophilidae) were maintained at 23±1°C, 75 % relative humidity and L: D 12 h: 12 h in half-pint milk bottles. They were provided with Drosophila medium (Formula 4–24, Carolina Biological Supply) and yeast.

A 36 gauge copper-constantan thermocouple was positioned next to the insect to determine the SCP. Insects were cooled at a rate of approximately 0.1 °C min⁻¹ using a refrigerated bath. The SCP was recorded as the lowest temperature reached prior to the release of the latent heat of crystallization as body water freezes (Lee and Denlinger, 1985).

The SCPs were determined for actively crawling larvae (approximately 4–5 days old) and for pupae, 2–4 days post-pupariation. To load adults into the SCP apparatus it was necessary to anesthetize them lightly for 2–10 min using a cryolizer (Bioquip Products). This alternative approach to standard etherization methods was used because ether is reported to increase the sensitivity of D. melanogaster to low temperature (Novitski and Rush, 1949).

The lower lethal temperature (LLT) for D. melanogaster was determined by directly transferring flies from 23°C to zero or sub-zero test temperatures for 2h. The highest temperature at which no fly survived 2 h of exposure was defined as the lower lethal temperature (LLT). The flies were held in plastic test tubes (1.0cm×7.5cm) that were subsequently placed in dry test tubes immersed in a refrigerated bath. The exposure temperature was recorded using a digital thermometer (Sensortek, Inc. model BAT-12) inserted into the plastic test tubes. All flies were returned to 23°C and survival rates recorded 8 h later. Adult flies that were able to fly, walk or stand were scored as alive. Emergence was the survival criterion for pupae, while the incidence of pupariation was used for larvae.

The following protocol was based on the work of Chen et al. (1987) with Sarcophaga crassipalpis. Using the survival data from the LLT tests of D. melanogaster, we determined that −5°C was lethal for adults, and this temperature was chosen as the basis for the cold-hardening experiment. The control group was exposed to −5°C for 2h. A chilled group was subjected to 5°C for 2h, followed by a 2-h exposure to −5°C to determine if chilling prior to subzero exposure increased survival, compared with flies directly subjected to −5 °C. In separate experiments varying exposure periods at 5°C were used to determine the effect of the duration of pre-chilling on survival at −5 °C.

We investigated whether the age of the adult fruit fly affected its capacity to supercool or cold-harden rapidly. Although the adult fruit flies were the primary focus of this study, pupae and larvae were also tested for SCP, LLT and their potential to cold-harden rapidly.

Statistical analyses were made using analysis of variance (ANOVA) (SAS, 1982). All percentage survival data were arcsin-transformed before analysis. Mean separations were determined using Duncan’s multiple range test at a predetermined significance level of 0.05.

The SCPs for D. melanogaster larvae, pupae and adults were in the range −17 to −20°C (Fig. 1). The SCPs remained relatively constant for adults aged 1-28 days. Although the larvae had the highest SCP by rank order, no obvious trend was evident with regard to developmental stage. No individual at any stage of development survived cooling to below the SCP. The use of a cryolizer as a means of cold immobilization had no statistically significant effect on the supercooling point.

The lower lethal temperature (LLT) was defined as the temperature at which no individual survived a 2-h exposure. No adult survived exposure to −5 °C (Fig. 2). Surprisingly, however, raising the temperature to −4°C resulted in nearly 100% survival. The lower lethal temperature of pupae was −8°C, while no larvae survived 2 h at −5°C (Fig. 2).

Although 5-day-old adults did not survive exposure to −5 °C, if 2 h of chilling at 5°C preceded the −5 °C exposure, the rate of survival increased dramatically to approximately 70% (Fig. 3). One hour of chilling at 5°C prior to exposure to −5°C markedly increased survival to about 90%, while as little as 30 min of chilling increased the rate of survival to more than 50% (Fig. 4).

The capacity to cold-harden rapidly varied with the age of the adult (Fig. 5). Flies ranging in age from 1–30 days old were tested. Over 70 % of 3-to 5-day-old flies survived 2h at −5°C when they had previously been chilled at 5°C (Fig. 5). One-day-old flies were less able to cold-harden rapidly than 3-to 5-day-old adults. The capacity to cold-harden rapidly was variable and survival rates were lower among older individuals.

The effect of gender on the capacity to cold-harden rapidly was also examined. In general, males and females were equally capable of surviving 2 h at −5 °C when pre-chilled at 0°C (Student’s t-test, P>0.05). Although no quantitative studies were made to evaluate reproductive function in the rapidly cold-hardened flies, these individuals produced offspring following treatment with no evidence of impaired reproductive function.

Pupae and larvae were also able to cold-harden rapidly (Fig. 3). Chilling at 5°C for 2 h prior to exposure to subzero temperatures significantly increased survival in these developmental stages. Survival was defined as the ability to reach the subsequent stage of development. Larvae did not survive a 2-h exposure to −5°C, yet they were able to pupariate 63 % of the time if chilled at 5 °C (for 2 h) before the subzero exposure. This result was a substantial increase, considering that 83 % pupariated when held at 5 °C for only 2 h (Fig. 5). Similarly, all pupae died after 2 h at −8 °C, but 22% eclosed if pre-chilled at 5 °C.

In some insects the SCP is a measure of the lower lethal temperature, but in D. melanogaster we found significant mortality at temperatures far above the SCP (Fig. 2). The SCPs of larvae, pupae and adults of D. melanogaster were in the range −17 to −20°C (Fig. 1). These results are consistent with the data of previous studies of cold tolerance in drosophilid species (Tucic, 1979; Krunic et al. 1980; Enomoto, 1981). However, in adults and larvae a 2-h exposure to −5°C was lethal, and −8°C was the lower lethal temperature for pupae. These data indicate that mortality was due to cold shock, since the insects were cooled rapidly and spontaneous ice formation in extracellular fluids did not occur until approximately 15 °C lower (Chen et al. 1987; Lee et al. 1987).

The onset of mortality due to low temperature exposure in the adult fruit fly was particularly abrupt. Although more then 90 % of the flies survived 2 h of exposure to −4°C, all died when exposed to −5°C for the same interval. The rapid decline in survivorship within a 1°C increment demonstrates how suddenly pre-freeze mortality can occur. Previous studies describe similar ranges of lower lethal temperatures for Drosophila (Novitski and Rush, 1949; Chiang et al. 1962).

The rapid cold-hardening response observed in D. melanogaster occurred within minutes. As little as 30 min at 5 °C was sufficient to stimulate the rapid physiological adjustment that enabled adult flies to survive subzero exposure. After 90 min of chilling nearly all flies recovered from the −5 °C treatment (Fig. 4). Adults aged 3–5 days old exhibited the greatest capacity for rapid cold-hardening. Pupae and larvae, as well as adults, were able to cold-harden rapidly and completed the subsequent developmental stage. Two hours of chilling enabled approximately 63 % of larvae to pupariate following exposure to subzero temperatures. Similarly, pupae did not survive 2h at −8°C, but when pre-chilled (2h at 5°C) 22% were able to develop into adults (Fig. 4).

The capacity for rapid cold-hardening appears to be widespread among insects as a means of protection against cold shock. Chen et al. (1987) first described this response in larvae, pupae and pharate adults of both diapause and nondiapause flesh flies, Sarcophaga crassipalpis. This extremely rapid cold-hardening response occurs in non-overwintering stages of a beetle, Xanthogaleruca luteola, a true bug, Oncopeltus fasciatus, and another fly, Sarcophaga bullata (Lee et al. 1987). Meats (1973), in a study of a tephritid fruit fly, examined temperature acclimitization as it pertains to thresholds for torpor and flight. He reported that Dacus try one was capable of rapid acclimation, even at cooling rates as fast as 1 °C min⁻¹.

D. melanogaster is a cosmopolitan species which inhabits all biogeographic realms (David et al. 1983). This species is reported to be the most resistant to the environmental stresses of high-temperature desiccation and low temperature among six species of the melanogaster subgroup (Stanley et al. 1980). D. melanogaster, the most common fruit fly, is widely used as an experimental organism for the study of genetics: yet, it is not known how this organism overwinters under natural conditions in temperate regions (David et al. 1983).

Our project focused on a relatively short time in which D. melanogaster was exposed to subzero temperatures. Yet, the fact that all stages tested could cold-harden rapidly may offer insight into the overwintering biology of this species. From the data presented here, it appears that larvae, pupae and adult fruit flies are all likely candidates for the overwintering stage. Tucic (1979) found the adult fly to be most cold resistant, but added that the other stages (including the egg) are potentially capable of surviving moderately severe winter conditions, as all stages have the genetic capacity for cold resistance adaptation.

Lakovaara et al. (1972) report that three species of Drosophila in northern Scandinavia overwinter in an adult diapause. Reproductive diapause is characteristic of four strains of fruit flies in the D. auratia complex which inhabit the main islands of Japan (Kimura, 1984). D. melanogaster, in contrast, does not appear to overwinter in diapause. Rather, it is suggested that this fruit fly passes the winter months in a state of quiescence (McKenzie, 1975). The rapid cold-hardening response described here would enhance low-temperature tolerance in overwintering D. melanogaster and would also minimize mortality during unseasonably low temperatures in the autumn and spring.

In an ecological context the capacity to cold-harden rapidly, and thereby avoid injury due to cold shock, may be important for survival at any time of the year. In north temperate regions even non-overwintering developmental stages may be exposed to diurnal fluctuations in temperature of 20 to 30°C. The capacity to cold-harden rapidly may be particularly important to minimize mortality during periods of unseasonably low temperature during the spring and autumn. These data suggest that this extremely rapid acclimation response may allow insects to enhance cold tolerance as they track decreases in their environmental temperatures (Lee et al. 1987). It would be useful in future studies to determine whether slower cooling rates, which often occur under natural conditions, also enhance cold-hardening in Drosophila.

Freeze tolerance in diapausing larvae of another member of the Drosophilidae, Chymomyza costata, has recently been reported (Shimada and Riihimaa, 1988). When acclimated for a month at 0–10°C, freeze-inoculated and cooled slowly, this fly survives freezing to approximately − 80°C. Findings such as this, in addition to those reported here on D. melanogaster, make it necessary to re-evaluate the potential of ‘freeze-intolerant’ species to survive subzero exposure. In addition, information such as this may be applicable to the development of methods for the cryopreservation of this species.

We thank Edward J. DeVillez, Thomas G. Gregg, Kenneth G. Wilson and James T. Oris for their suggestions and comments on this study. This work was supported in part by National Science Foundation grants DCB-8517875 and 8811317 to REL.