ABSTRACT
The great complexity of extracellular freezing stress, involving mechanical, osmotic, dehydration and chemical perturbations of the cellular milieu, hampers progress in understanding the nature of freezing injury and the mechanisms to cope with it in naturally freeze-tolerant insects. Here, we show that nuclear DNA fragmentation begins to occur in larval haemocytes of two fly species, Chymomyza costata and Drosophila melanogaster, before or at the same time as the sub-zero temperature is reached that causes irreparable freezing injury and mortality in freeze-sensitive larval phenotypes. However, when larvae of the freeze-tolerant phenotype (diapausing–cold acclimated–hyperprolinemic) of C. costata were subjected to severe freezing stress in liquid nitrogen, no DNA damage was observed. Artificially increasing the proline concentration in freeze-sensitive larvae of both species by feeding them a proline-enriched diet resulted in a decrease in the proportion of nuclei with fragmented DNA during freezing stress. Our results suggest that proline accumulated in diapausing C. costata larvae during cold acclimation may contribute to the protection of nuclear DNA against fragmentation associated with freezing stress.
INTRODUCTION
Freeze tolerance, the ability to survive ice formation in extracellular spaces, has evolved in many insects (Sinclair, 1999; Sinclair and Renault, 2010), plants (Pearce, 2001) and exceptionally also in ectothermic animals such as frogs, salamanders, turtles and lizards (Storey and Storey, 1988). The formation of ice crystals outside cells increases the concentration of extracellular solutes, which drives osmotic outflow of water from cells, which in turn causes a reduction of cell volume, tight packing of cytosolic components and an increase in the concentration of electrolytes, metal ions, protons and other chemical perturbants inside the cell (Mazur, 1984; Muldrew et al., 2004). The great complexity of freezing stress hampers progress in understanding the nature and hierarchy of freezing damage, which operates at different levels of biological organisation, from macromolecules and their complexes, to organelles, cells, tissues and finally to systemic function, survival and fitness (Rozsypal, 2022; Toxopeus and Sinclair, 2018). Similarly, mechanisms to prevent or repair the freezing damage are thought to form a broad adaptive complex composed of many interacting components (Storey and Storey, 2017; Toxopeus and Sinclair, 2018).
Drastic changes in extracellular and intracellular conditions associated with freezing stress can lead to loss of conformational stability of macromolecules and integrity of their assemblies. Therefore, destabilisation of macromolecules, such as irreversible denaturation of proteins and loss of lipid bilayer integrity, are considered to be the primary molecular mechanisms of freezing damage in cells or freeze-sensitive organisms (Mazur, 1984; Muldrew et al., 2004; Steponkus, 1984). Conversely, stabilisation of proteins and lipid bilayers is a general mechanism considered to be crucial for survival during freezing stress in diverse systems such as cryopreserved mammalian cells (Mazur, 2010), or naturally freeze-tolerant plants (Steponkus, 1984) and ectothermic animals (Storey and Storey, 1988), including insects (Lee, 2010; Rozsypal, 2022; Teets and Denlinger, 2014; Toxopeus and Sinclair, 2018). One class of macromolecules, DNA, has so far received limited attention from the research communities focusing on plant (Jaikumar et al., 2020) or animal freeze tolerance (Lubawy et al., 2019). However, a number of studies on the cryopreservation of mammalian sperm or blood cells have confirmed a close link between freezing stress and loss of DNA integrity (Donnelly et al., 2001; Restrepo et al., 2019; Ross et al., 1990; Valcarce et al., 2013). Nuclear DNA fragmentation is thought to be caused primarily by oxidative stress due to the increased concentration of reactive oxygen species (ROS) associated with the various steps of the cryopreservation protocol (Aitken and De Iuliis, 2009; Hughes et al., 1998). We are aware of only one study to date that has experimentally assessed the sensitivity of nuclear DNA to freezing stress in insects: Lubawy et al. (2019) showed that haemocytes of the tropical cockroach Gromphadorina coquereliana exhibit similar levels of DNA damage when exposed to either freezing stress at −6°C or heat stress at 44°C. While all cockroaches survived well after heat stress, the majority succumbed to freezing stress. Some important questions were left open: (i) how widespread is this response among insect species; (ii) is there a causal link between DNA damage (or the inability to repair it) and mortality after freezing stress; and (iii) can naturally freeze-tolerant insects avoid or repair DNA damage? Here, we asked similar questions and report on experiments expanding our understanding of the association between freezing stress, loss of DNA integrity and mortality in two fly species: the malt fly Chymomyza costata and the vinegar fly Drosophila melanogaster (both Diptera: Drosophilidae).
Fully grown third instar larvae were used for all experiments. The larvae of the two species are similar in size and general morphology. During the favourable summer period, they grow and develop rapidly on decaying plant material. The two species are from different geographical areas: while C. costata inhabits cold temperate and subarctic regions (Hackman et al., 1970) and its larvae spend the cold winter period in diapause (Riihimaa and Kimura, 1989), D. melanogaster is native to the Central African tropics, it has only spread to temperate regions during the last century (Throckmorton, 1975) and its larvae are found only in the warm season as adults overwinter (Schmidt and Paaby, 2008). As a consequence, D. melanogaster larvae are sensitive to even mild chilling (Koštál et al., 2011a; Strachan et al., 2010), whereas C. costata larvae are hardy: summer-active larvae are still relatively sensitive to freezing stress and only a small proportion of them can survive after freezing to −10°C; winter diapausing larvae become extremely freeze tolerant (Shimada and Riihimaa, 1988) and survive freezing at −100°C (Shimada, 1990) or even cryopreservation in liquid nitrogen at −196°C (Moon et al., 1996). In addition, we have found that larval tolerance to freezing can be modulated by artificially increasing the concentration of the free amino acid proline in their bodies by feeding the larvae a proline-enriched diet. Drosophila melanogaster larvae fed proline survive freezing to −5°C with 75% of body water converted to ice crystals (Koštál et al., 2016a, 2012). Proline-fed summer-active larvae of C. costata develop extreme hardiness, including the ability to survive cryopreservation in liquid nitrogen (Koštál et al., 2011b; Rozsypal et al., 2018).
Here, we subjected differently acclimated larvae of the two fly species to progressive freezing stress (decreasing freezing stress temperature) and asked: (i) whether there is any relationship between the temperature causing freeze mortality (taken from our previous studies) and the temperature at which nuclear DNA damage occurs (as assessed by the DNA comet assay); (ii) whether diapausing, cold-acclimated (extremely hardy) larvae of C. costata exhibit nuclear DNA fragmentation following exposure to liquid nitrogen using two different cryopreservation protocols, one of which results in survival of most larvae, the other causing death of all larvae; and (iii) whether artificially increasing the proline concentration in the larval body (achieved by feeding larvae a proline-enriched diet) has any effect on the incidence of nuclear DNA damage during freezing stress.
MATERIALS AND METHODS
Rearing and acclimation of flies
The two fly species were cultured in MIR 154 incubators (Sanyo Electric, Osaka, Japan) and fully grown third instar larvae were used for all experiments. Briefly, the larvae of the malt fly Chymomyza costata (Zetterstedt 1838) (Sapporo strain) are photoperiodically sensitive. We reared them either under long day length (LD, 16 h/8 h light/dark cycle; 18°C), which simulates the development of summer-active larvae and promotes direct development to pupae and adults, or under short day length (SD, 12 h/12 h light/dark cycle; 18°C), which induces winter dormancy of diapause type – a hormonally regulated developmental arrest (Koštál et al., 2016c, 2017). Next, we subjected the diapausing (SD) larvae to gradual cold acclimation (SDA, transfer to 11°C and constant darkness at 6 weeks of age for 1 week, then transfer to 4°C for another 4 weeks), which leads to a dramatic increase in freeze tolerance and acquisition of the ability to survive after cryopreservation in liquid nitrogen (Koštál et al., 2011b) (note: we keep the abbreviations LD, SD and SDA in this paper as they were used in our previous publications). Drosophila (Sophophora) melanogaster Meigen 1830 (Oregon-R strain) larvae are photoperiodically insensitive and were reared under constant 15°C and 12 h/12 h light/dark cycles. Prior to wandering stage (migration out of diet to seek sites suitable for pupariation), larvae were transferred to constant darkness and a fluctuating temperature regime (FTR, 6°C 20 h/11°C 4 h) for 3 days. Under FTR conditions, larvae enter the quiescence type of dormancy – a developmental arrest directly induced by low temperature, and increase their cold tolerance (Koštál et al., 2011a, 2016b). Vinegar flies were reared on a standard cornmeal–yeast–agar diet, and malt flies were reared on a similar cornmeal–yeast–agar diet supplemented with ground malted barley (Lakovaara, 1969). To modulate the level of larval freeze tolerance, the standard diets were supplemented with l-proline (Sigma-Aldrich, St Louis, MO, USA) (50 mg proline g−1 standard diet), which further increases freeze tolerance in larvae of both species (Koštál et al., 2016a, 2012, 2011b; Rozsypal et al., 2018).
Freezing of fly larvae
The larvae were frozen according to our previously developed optimised freezing protocols (Koštál et al., 2016a; Rozsypal et al., 2018). Briefly, larvae were frozen using a programmable Ministat 240 cold circulator (Huber, Offenburg, Germany). Groups of approximately 20 larvae were wrapped between two layers of cellulose moistened with distilled water. The moist cellulose ‘ball’ with the larvae inside was placed in the plastic tube (diameter 1 cm, length 5 cm). At the start of the freezing assay, a small ice crystal was added to the surface of the moist cellulose. The rate of cooling to a target temperature was 0.1°C min−1 and 0.01°C min−1 for C. costata and D. melanogaster, respectively. The ice crystals caused early freezing of water in the cellulose and stimulated ice penetration and inoculative freezing of larval body fluids at temperatures close to −1°C in D. melanogaster larvae (Koštál et al., 2016a), or between −1 and −4°C in C. costata larvae (Rozsypal et al., 2018). At the end of the freezing assay, immediately after thawing, larval haemolymph was collected to assess nuclear DNA integrity in haemocytes. Data on survival after slow extracellular freezing inoculated by external ice crystals were taken directly from our previous studies (Koštál et al., 2016a, 2012, 2011b; Rozsypal et al., 2018). In addition, haemolymph was collected from C. costata larvae cryopreserved in liquid nitrogen using the optimal and pesimal protocols. In the optimal protocol, larvae were slowly pre-frozen to −30°C (see above), then immersed in liquid nitrogen for 1 h and then warmed to 5°C at a rate of 0.6°C min−1. The LD larvae do not survive this treatment, whereas virtually all SDA larvae survive and almost 40% of them will metamorphose into fertile adults (Rozsypal et al., 2018). In the pesimal protocol, the SDA larvae are immersed in liquid nitrogen after insufficient pre-freezing, e.g. only to −5°C, resulting in none of them surviving the cryopreservation protocol (Rozsypal et al., 2018).
Assessment of nuclear DNA integrity in larval haemocytes
Circulating haemocytes were isolated by collecting 5 µl of haemolymph from a pool of 15–20 larvae into a calibrated microcapillary tube (Drummond Sci., Broomall, PA, USA). Contaminating fat body cells occurred sporadically in D. melanogaster haemolymph samples, but were excluded from analysis because of the size of the nuclei (haemocytes have 2n or 4n DNA content and nuclear diameter <10 µm, while fat body cell nuclei have very high ploidy and diameter >20 µm; Lee et al., 2009). The OxiSelect Comet Assay Kit (Cell Biolabs, San Diego, CA, USA) was used to assess nuclear DNA integrity in isolated haemocytes according to the manufacturer's instructions. Briefly, the haemocytes in 5 µl of haemolymph sample were added to 70 µl of agarose (provided in the kit), melted at 42°C and spread over the surface of a microscope slide (provided in the kit). After gelation of the agarose (at 4°C for 15 min), the slides were placed in 20 ml of ice-cold lysis buffer and incubated at 4°C for 60 min. The lysis buffer was then replaced with 20 ml of ice-cold alkaline solution and the slides incubated at 4°C for 30 min. The slides were then washed with alkaline electrophoresis solution and placed in the gel tray of the horizontal electrophoresis tank. The voltage was set at 1 V cm−1 of distance between the electrodes and applied for 15 min. The comet assay detects damage associated with the formation of single- or double-strand DNA breaks leading to chromosome fragmentation and mobilisation of DNA fragments in the electric field (Collins et al., 2023). After electrophoresis, the slides were immersed in 25 ml of ice-cold deionised water for 3×2 min and finally in 20 ml of ice-cold 70% ethanol for 5 min. Slides were then dried overnight at room temperature. DNA nuclei and comets were stained with Vista Green DNA dye (provided in the kit) for 15 min and photographed (using 480/535 nm excitation/emission filters) under a Zeiss Axioplan 2 fluorescence microscope (Zeiss, Oberkochen, Germany) equipped with an Olympus DP73 CCD camera (Olympus, Tokyo, Japan). To provide a positive control for DNA comets, C. costata (LD acclimation group) haemocytes were treated with hydrogen peroxide (1 mmol l−1 H2O2, Sigma-Aldrich) at 4°C for 30 min and then assayed for DNA integrity as described above (see Fig. S1).
RESULTS
DNA damage occurs at temperatures that cause freeze mortality
In Fig. 2, mortality data are presented as: (i) survival to larval stage (live/dead larvae scored 12 h after the end of the freezing assay); and (ii) survival to adult stage, where all live larvae were maintained at 18°C for the following 14 days (D. melanogaster) or 2 months (C. costata) and successful metamorphosis to adult stage was taken as the final criterion of survival. The shaded areas delimit the interval of sub-zero temperatures that cause 50% acute larval mortality and 50% delayed mortality, expressed as the inability of larvae to progress to the adult stage. In other words, either acute (causing instant mortality) or irreparable freezing injury (causing delayed mortality) occurred in larvae exposed to temperatures within this interval. The key message from Fig. 2 is that in both fly species, nuclear DNA fragmentation in larval haemocytes begins before or after reaching the sub-zero temperatures that cause irreparable freezing injury and mortality in freeze-sensitive larval phenotypes.
Nuclear DNA of C. costata SDA larvae is well protected against freezing damage
Fig. 3 shows that the nuclear DNA in haemocytes suffered drastic damage when the optimal cryopreservation protocol was applied to LD larvae of C. costata: the proportion of intact nuclei (white parts of the bars) decreased from 91.0% to 7.4%. In contrast, the nuclear DNA remained intact during the same treatment applied to SDA larvae: the proportion of intact nuclei remained practically constant at 67.4% and 70.0% in control and cryopreserved SDA larvae, respectively. Importantly, the application of a pesimal cryopreservation protocol (acutely killing all larvae) also resulted in no significant damage to nuclear DNA in SDA larvae: the proportion of intact nuclei was 79.4%.
Feeding larvae with proline-enriched diet partially protects nuclear DNA from freezing damage
Fig. 4 shows that frozen larvae fed a proline-enriched diet exhibited less damage to nuclear DNA in haemocytes than larvae fed a standard diet. The effect was statistically significant for both fly species. In C. costata frozen at −30°C, the proportion of intact nuclei increased from 16.7% to 64.0% when fed a proline-enriched diet. In D. melanogaster, the increases were from 56.1% to 92.3% and from 9.4% to 26.9% in larvae exposed to −10°C and −30°C, respectively.
DISCUSSION
Here, we show that nuclear DNA damage occurs in association with extracellular freezing in larvae of two fly species. We also show that the relative proportion of haemocytes with fragmented DNA increases with decreasing temperature of freezing stress and that the temperature at which DNA damage begins to occur corresponds relatively well with the temperature at which irreparable freezing injury occurs in the insect organism and mortality is induced.
These results, together with primary observations in the cockroach G. coquereliana (Lubawy et al., 2019), open avenues for further investigation into the potential importance of DNA stabilisation for survival during freezing stress in naturally freeze-tolerant insects. Although nuclear DNA damage occurred at freezing temperatures that cause mortality in both the present study and that by Lubawy et al. (2019), further research is needed to decide whether the relationship between DNA damage and mortality is causal or just a coincidence. So far, freeze-induced DNA damage has only been studied in insect haemocytes. However, the partial loss of the haemocyte population could be easily tolerated given their relatively high capacity to proliferate from stem cells (Stanley et al., 2023). Future efforts should be directed towards analysing freeze-induced DNA damage in other insect tissues. In addition, the relationship between DNA damage and mortality in the two insect studies was not entirely straightforward. First, although heat-stressed and freeze-stressed cockroaches experienced similar levels of DNA damage, the former survived while the latter died (Lubawy et al., 2019). This observation suggests that the difference between heat-stressed and freeze-stressed animals may lie in their different ability to stimulate DNA repair mechanisms after stress (Aleksandrov et al., 2020; Caldecott, 2008; Friedberg, 2003). It is also highly likely that the freeze-stressed animals suffered additional (irreparable) freezing injuries, given the great complexity of freezing stress and its detrimental influence on different biological structures and processes (Rozsypal, 2022; Toxopeus and Sinclair, 2018). Second, acute and absolute larval mortality was observed in selected treatments without any detectable (increase in) damage to haemocyte nuclear DNA, i.e. in the C. costata SDA larvae exposed to liquid nitrogen using the pesimal cryopreservation protocol and in the D. melanogaster larvae fed proline-enriched diet and exposed to −10°C. These results again confirm that freezing damage is a complex phenomenon, of which DNA damage may not be a necessary component.
The insensitivity of haemocyte nuclear DNA to severe or even lethal freezing stress in SDA C. costata larvae exposed to liquid nitrogen deserves special attention. Of course, we cannot exclude the possibility that nuclear DNA was damaged in a more subtle way that was not detectable by the comet assay. However, it is also reasonable to assume that the DNA of dormant insects is truly insensitive to, or protected from, freezing injury. Diapausing insects have their developmental pathways arrested and show relatively low behavioural and metabolic activity, corresponding to low gene transcriptional activity (Reynolds, 2017; Teets et al., 2023). Similarly, C. costata larvae that maintain deep diapause at a warm temperature of 18°C show only minimal changes in transcriptional profiles over time (Koštál et al., 2017). Thus, the hypothesis that DNA stability in dormant insects is associated with the formation of highly condensed, transcriptionally silent heterochromatin, which could better withstand the mechanical stresses associated with cytosolic freeze dehydration, cell volume shrinkage and organelle translocation, is tempting. However, we used cold-acclimated diapause larvae (SDA) of C. costata for our experiments. Although these larvae are transferred to a low temperature of 4°C during cold acclimation, we have previously shown that the transfer is accompanied by a significant change in the transcriptional profile (including both down- and up-regulation of numerous gene transcripts), so that the DNA of SDA larvae can no longer be considered transcriptionally silent (Koštál et al., 2017). Another possibility is that the nuclear DNA of SDA larvae is protected by an external stabilising agent.
During cold acclimation, C. costata SDA larvae accumulate a mixture of several metabolites in their haemolymph and tissues, dominated by proline and trehalose (Koštál et al., 2011b; Kučera et al., 2022; Moos et al., 2022; Rozsypal et al., 2018). Using a combination of in vivo and in vitro experiments, we have recently shown that these compounds contribute to the stabilisation of larval fat body cell membranes during freezing stress, protecting lipid bilayers from loss of structural integrity and barrier function (Grgac et al., 2022). The results presented in this study suggest that proline may also contribute to the stabilisation of nuclear DNA during freezing stress. Larvae of both species fed a proline-enriched diet showed less DNA damage than their counterparts fed standard diet. The molecular mechanism by which proline exerts its protective effect on nuclear DNA remains unclear. One possible explanation is related to the chemical potential of proline to act as a scavenger of ROS (Kaul et al., 2008; Rejeb et al., 2014; Signorelli et al., 2014). There is strong evidence that ROS cause single- and double-strand breaks in DNA (Driessens et al., 2009; Mahaseth and Kuzminov, 2016), which is also documented in our ‘positive control’ experiment, where exposure of larval haemocytes to hydrogen peroxide resulted in the appearance of DNA comets similar to those observed after freezing stress (Fig. S1). ROS are also thought to cause DNA damage in frozen and cryopreserved mammalian cells (Aitken and De Iuliis, 2009; Hughes et al., 1998). Therefore, it may be interesting to pursue the possibility that proline accumulated in C. costata larvae during cold acclimation acts as a ROS scavenger, protecting nuclear DNA (and possibly other macromolecules) from oxidative damage associated with freezing stress.
Conclusions
We have shown that the occurrence of nuclear DNA damage (fragmentation) in haemocytes is closely related to temperatures that cause irreparable freezing injury and mortality in freeze-sensitive larvae of C. costata and D. melanogaster. In contrast, almost no DNA damage was observed in freeze-tolerant C. costata larvae (diapausing, cold-acclimated phenotype) cryopreserved in liquid nitrogen. Feeding proline-enriched diet to freeze-sensitive larvae of both species resulted in less DNA damage during freezing stress. Our results suggest that DNA stabilisation may be an important facet of cold hardiness in naturally freeze-tolerant insects.
Acknowledgements
We thank Irena Vacková for maintenance of insect colonies.
Footnotes
Author contributions
Conceptualization: T.Š., V.K.; Methodology: T.Š., V.K.; Formal analysis: V.K.; Investigation: T.Š.; Writing - original draft: V.K.; Writing - review & editing: T.Š.; Visualization: V.K.; Supervision: V.K.; Project administration: V.K.; Funding acquisition: V.K.
Funding
This study was supported by Grantová Agentura České Republiky (GAČR; 19-13381S to V.K.).
Data availability
All data are presented in the paper and the supplementary information.
References
Competing interests
The authors declare no competing or financial interests.