European corn borer eggs were obtained from the Maize Research Institute “Zemun Polje” in Belgrade, Serbia. Egg patches were sterilized with 4% Persteril solution (Overlack, Czech Republic), placed in translucent polycarbonate boxes with a semi-artificial diet (Salama, 1970) and were reared under different light:dark (LD) regimes at 22 ± 1°C. Non-diapausing larvae were reared under the photoperiod of 18L:6D, while diapause was induced with the photoperiod of 12L:12D. Larvae were observed on a daily basis and reared under the specified conditions until they reached the wandering stage of the final instar, when non-diapause (nd) and diapause-destined (dd) larvae were sampled (Figure 1).

Furthermore, diapause-destined larvae were either allowed to enter diapause at 22 ± 1°C (warm-acclimated) or were gradually chilled to 5 ± 1°C (−2°C/day) and then allowed to become dormant (cold-acclimated). Diapausing larvae were sampled every 2 months (Figure 1), except for warm-acclimated larvae that were sampled only once due to the high mortality (>95%) after 3 months in diapause. Overall, six experimental groups were established (Figure 1): non-diapause (nd), diapause-destined (dd); warm-acclimated, 2 months diapause (dwa2); cold-acclimated, 2 (dca2), 4 (dca4), and 6 (dca6) months diapause.

All chemicals were of analytical-reagent grade. Citric acid, γ-aminobutyric acid (GABA), cetyltrimethylammonium bromide (CTAB) and standards of nucleotides were obtained from Sigma (St. Louis, United States), methanol from Fluka (Buchs, Switzerland), and deionised water (18.2 Ωcm) from Elga Bucks, United Kingdom. Running electrolyte was prepared by dissolving citric acid to a concentration of 40 mM in water and titrating to pH 4.4 with solid GABA. Cationic surfactant CTAB was then added to a final concentration of 8 mM CTAB, and the solution was sonicated for 15 min at ambient temperature to thoroughly mix the surfactant with the running electrolyte. The standard solutions of each nucleotide were prepared separately by dissolving each nucleotide in deionized water (5 mg/mL). The standard mixture of nucleotides in the concentration of 1.0 mg/mL was prepared by mixing and diluting PAD standards with deionized water. Lower concentrations were prepared by diluting the mixture in deionized water.

Cytochrome c oxidase (COX) activity was assessed by monitoring the decrease in absorbance at 550 nm as described in Hofmann and Hand (1990). Reduced cytochrome c from the equine heart (RHCC) was prepared as described by Eddy et al. (2006) and stored at −20°C for up to 1 month. Before beginning the measurements, the stability of the reagent was validated by scanning the 450–600 nm spectrum. Ten frozen larvae per biological pool were homogenized on ice in 50 mM phosphate buffer pH 7 crushed (20% w:v) with a Teflon/glass homogenizer at 1,500 rpm with 10 strokes. The crude homogenate was transferred into a new tube, incubated on ice for 10 min and centrifuged at 1,000 g for 10 min to remove cellular debris. Supernatants were removed, placed into new tubes and centrifuged at 20,000 g for 20 min to pellet mitochondria. The pellet was washed with phosphate buffer, used for homogenization, and centrifuged again at 20,000 g for 20 min for further purification. The mitochondria precipitate was dissolved in the same phosphate buffer with added 0.5M sucrose, frozen in liquid nitrogen and stored at −80°C until analysis. Protein concentration was determined by Quick Start BioRad Assay (BioRad, Hercules, CA, United States) and bovine serum albumin was used as the standard. COX activity was measured at 550 nm using the Infinite M200 UV/VIS spectrophotometer/fluorimeter (Tecan Group ltd). One 250 μL reaction mixture contained phosphate buffer and 50 μM of RHCC. Reactions were initiated by adding up to 10 μL of mitochondrial fraction and the decrease in absorbance was recorded. Blanks omitted RHCC and conditions were optimized to ensure a linear reaction rate. One unit of enzyme activity was defined as the amount of enzyme that oxidized 1 nmol of RHCC per minute. All measurements were performed in technical duplicates for every biological pool. Data are expressed as units of enzyme activity per mg of total protein.

All analyses were performed in technical duplicates for each biological pool. Statistical analysis was performed on the means of the technical duplicates using the data analysis software system Statistica version 13 (StatSoft, Inc., Tulsa, OK, United States). The statistical significance of the differences between the larval experimental groups was tested using one-way ANOVA followed by post hoc Tukey’s test, with a level of significance of p < 0.05. Additionally, principal component analysis (PCA) was used to analyse the data.

Five distinct nucleotides (ATP, ADP, AMP, UMP, and GMP) and two coenzymes (NAD⁺ and NADP⁺) were detected and quantified in all experimental groups using CE (Figures 2–4). Moreover, the EC and ratios of the three distinct adenosine-phosphate nucleotides were determined.

The specific activity of COX significantly differed between experimental groups (Figure 6). The activity of COX was higher in larvae reared and acclimated to 22°C (nd, dd, dwa2) in comparison to groups acclimated to 5°C (dca2, dca4, dca6). It was the highest in non-diapausing larvae and the lowest in the middle of diapause (dca4), after which it increased in the latter part of diapause. COX activity differed significantly among diapausing larvae that have been acclimated under two distinct temperature regimes for 2 months (dwa2 and dca2) – it was higher in the group reared at 22°C (dwa2).

The relative gene expression of individual subunits of the electron transport chain in the fat body of larvae differed both between the experimental groups and between the analyzed genes (Figure 7). When compared to the non-diapausing group, the expression of CoQ-cytochrome c reductase and COX genes was significantly increased in diapausing groups (Figures 7B,C). On the other hand, the expression of NADH dehydrogenase and ATP synthase genes was reduced at the beginning of diapause, and then gradually increased, reaching the same level at the end of diapause as in non-diapausing larvae (Figures 7A,D).

Gene expression in diapause-destined larvae was different and specific for each gene. Expression of COX and ATP synthase genes was similar to that in the non-diapausing group, while the expression of NADH dehydrogenase and CoQ-cytochrome c reductase genes was increased in comparison to nd larvae (Figures 7A–D).

In diapausing larvae, the relative expression of all analyzed genes in the electron transport chain had a characteristic pattern – it was higher in the warm-acclimated group (dwa2) when compared to the cold-acclimated group (dca2). Going further, it usually followed a trend of gradual increase in gene expression during the time course of diapause (Figures 7A–D).

When it comes to ant and phb2, these two genes share a similar pattern of expression, with statistically significant differences between experimental groups. Compared to non-diapausing larvae, both of these genes are up-regulated in diapausing groups, except in cold-acclimated larvae in early diapause (dca2). In this group, gene expression is either unchanged (ant) or lowered (phb2) in comparison to non-diapausing larvae (Figures 7E,F).

In diapausing larvae, both genes exhibit a similar expression pattern – lowest expression in the dca2 group, with gradual increase until the highest expression is reached near the end of diapause (dca6). Also, as shown in Figures 7E,F, both of these genes have much higher expression in warm-acclimated larvae in early diapause (dwa2) compared to larvae of the same age, but cold-acclimated (dca2).

In order to determine where the variation in the dataset comes from, as well as to reveal what links the patterns of the performed measurements together, the gathered data was subjected to PCA and the results are presented in Figure 8. The PCA score plot grouped the samples into clusters according to their similarity (Figure 8A), while the PCA loading plot (Figure 8B) shows how each principal component is affected by the analyzed data. As can be seen in Figure 8, the PCA extracted two major principal components which cover 71.78% of the data variance. The first principal component (PC1) accounted for 43.90% of the total variance and it separated the nd, dd, and dca2 groups (negative scores) from the dwa2, dca4, and dca6 groups (positive scores). Based on this, it is highly likely that PC1 accounts for the variance caused by the time course of diapause (Figure 8A). Furthermore, PC1 was positively correlated with the expression of all analyzed genes, as well as UMP concentration and adenine nucleotide ratios. Negative correlations were observed for COX enzyme activity, EC content and the remaining nucleotide concentrations (Figure 8B). The second principal component (PC2) accounted for 27.88% of the total variance. It is likely that this component reflects the effect of metabolism rates on the analyzed characteristics, considering that it separated the metabolically less active dca2 and dca4 groups (negative scores) from the rest (Figure 8A). PC2 showed positive correlation with most of the analyzed characteristics, except for the concentrations of UMP, AMP, and NADP⁺, as well as AMP:ATP ratios (Figure 8B).

ATP is a universal energy tender in living systems and cells must maintain a high ratio of ADP:ATP and AMP:ATP nucleotides in order to preserve homeostasis. The AMP:ATP nucleotide ratio is particularly important, as these adenine nucleotides are known to be major allosteric enzyme modulators, but most often with opposing effects (Hardie and Hawley, 2001). In comparison to non-diapause, ATP/ADP concentrations and EC decrease in diapause-destined and all diapausing groups, regardless of temperature acclimation. Conversely, AMP concentration increases in diapause, especially in the cold-acclimated group that was dormant for 2 months (dca2), as also supported by the PCA. Thus, the ADP:ATP and AMP:ATP ratios are higher in diapause groups, compared to non-diapause groups. Since low levels of ATP/ADP and high levels of AMP are characteristic of early diapause at both high (dwa2) and low temperatures of acclimation (dca2), these results indicate endogenous regulation of the consumption of energy equivalents during diapause. However, it should be noted that adenine nucleotide concentrations were two times higher in the cold-acclimated (dca2) group compared to the warm-acclimated group (dwa2). Similarly, the activity of COX, concentrations of NAD⁺ and GMP, as well as the expression of all analyzed energy metabolism-related genes decrease during the first months of diapause in cold-acclimated larvae. Our findings are consistent with high levels of AMP and lower levels of ATP/ADP obtained in similar studies on dormant phases of the killifish Austrofundulus limnaeus (Podrabsky and Hand, 1999) and in the embryos of the shrimp Artemia franciscana (Zhu et al., 2009). Given the scarcity of literature data on the participation of GMP in diapause, the obtained results could be explained by assuming possible functions of a physiologically important form of GMP nucleotide – cyclic GMP (cGMP), which unfortunately could not be quantified separately from the non-cyclical form by CE in this study. Thus, we can assume that in non-diapausing larvae cGMP probably serves as a secondary messenger in various processes of active metabolism, while in diapausing larvae it is less important due to hypometabolism, until the termination of diapause and the forthcoming metamorphosis. Namely, in the flesh fly Sarcophaga crassipalpis, diapause was shown to be interrupted when cGMP or its derivatives were injected into diapausing pupae (Fujiwara and Denlinger, 2007), and in the moth Mamestra configurata, cAMP was found to participate in maintaining diapause, while cGMP brought on its termination (Bodnaryk, 1987). In the roundworm Caenorhabditis elegans, cGMP has been found to regulate nutrition and, based on nutrient availability, affect dormancy, allowing the body to survive adverse conditions (You et al., 2008). The concentration of NAD⁺ was expected to be reduced in cold-acclimated diapausing larvae with suppressed metabolism, in comparison to actively developing larvae with intensive respiration and metabolism. As NADH/NAD⁺ are the main carriers of reduced equivalents for the electron transport chain in mitochondria, a decrease in NAD⁺ concentrations in diapause was expected. Since O. nubilalis larvae do not feed during diapause and subsequent exposure to low temperatures the number of mitochondria and oxygen consumption decrease, and so does their metabolic rate (McLeod and Beck, 1963; Jovanović-Galović, 1997). Cold induced metabolic suppression in diapause of O. nubilalis was also supported by lower COX activity in diapausing compared to non-diapausing groups. Similar findings were obtained in the cotton bollworm Helicoverpa armigera (Yang et al., 2010), the shrimp A. franciscana (Reynolds and Hand, 2004), the moth Omphisa fuscidentalis (Singtripop et al., 2007), as well as in the moth species Epiblema scudderiana and the fly Eurosta solidaginis (McMullen and Storey, 2008). Moreover, in E. scudderiana and E. solidaginis, COX activity was found to decrease further with exposure of diapausing individuals to low temperatures (McMullen and Storey, 2008).

As opposed to GMP and NAD⁺, UMP and NADP⁺ concentration increased significantly in the first months of diapause (dca2, dca4), and then decreased ahead of diapause termination. Interestingly, the increase in NADP⁺ was observed already in dd larvae, which may signal preparation for diapause and the forthcoming metabolic depression. The concentration of UMP nucleotides in these groups was increased 10-fold and returned to diapause-destined levels at the end of diapause. The high concentration of NADP⁺ coenzymes in the initial months of diapause is in agreement with previous studies of diapause in the European corn borer, which showed that there is a close connection between glycerol synthesis, hexose-monophosphate shunt activity and the antioxidant protection system (Stanić et al., 2004). Hexose-monophosphate shunt activity and NADPH synthesis were found to increase in the initial months of diapause. During diapause, NADPH serves as a source of reducing equivalents in antioxidant protection, representing a substrate for the regeneration of antioxidant molecules such as GSH, thioredoxin, glutaredoxin, and others. Also, NADPH is necessary for biosynthetic processes, such as the production of glycerol which participates in the protection of cells from low temperatures (Stanić et al., 2004; Blagojević, 2007; Kojić et al., 2010, 2018; Purać et al., 2015; Uzelac et al., 2020). In addition to the concentration increase of the NADP⁺ in dca2 and dca4 larvae, a significant increase in the transcription of the thioredoxin (trx) gene was found in the same groups (Popović et al., 2015).

On the other hand, the importance of UMP nucleotides in insect development and dormancy is poorly described in literature. However, elevated uridine plasma levels have been shown after fasting in mice/rat models and humans, and the levels dropped rapidly after the fast was broken (Deng et al., 2017). Also, in these mammalian models, ATP depletion leads to increased levels of UMP and UDP, since ATP is a phosphate donor for UTP synthesis (Zhang et al., 2020). For this reason, UMP can be viewed either as a precursor for UDP nucleotide synthesis or, conversely, as a product of UDP hydrolysis in the ADP phosphorylation reaction and ATP regeneration. UDP nucleotides, as components of UDP-glucose, participate in the synthesis of trehalose and glycogen (Murphy and Wyatt, 1964). In O. nubilalis larvae, glycogen synthesis is intensive before entering diapause, after which glycogen serves as a source of glucose that is necessary for energy production, as well as for glycerol synthesis during diapause (Grubor-Lajšić et al., 1991). In addition to glycerol, trehalose is intensively synthesized in the initial months of diapause, as it is an important component of the development of resistance to low temperatures in the European corn borer (Kojić et al., 2010, 2018). In this regard, the increased concentration of UMP and NADP⁺ in the initial months of O. nubilalis diapause probably correlates with the processes of glycogen degradation, as well as the PPP and synthesis of glycerol, trehalose, and ATP. Our findings strengthen the established connection between the PPP pathway, glycogen metabolism and ADS in O. nubilalis diapause and cold hardiness development.

As respiration in individuals diapausing at low temperatures is known to be far lower than in non-diapausing ones (Wipking et al., 1995), low temperatures are likely to reduce the expression of examined genes for respiratory chain complexes, ATP synthesis and mitochondrial biogenesis, as evidenced in our study. Namely, the expression pattern of all analyzed energy-related genes – NADH dehydrogenase (nd1), CoQ-cytochrome c reductase (uqcr), cox, ATP synthase (atp), ADP/ATP translocase (ant), and prohibitin 2 (phb2), was the lowest in the first few months of diapause (dca2 group) and then gradually increased during the time course of diapause. In the larval diapause of the cotton bollworm H. armigera, cox expression was lower compared to the non-diapausing group (Yang et al., 2010), and in the fat body of the firebug Pyrrhocoris apterus, a gradual decrease in the number of cox transcripts was determined during 120 days of reproductive diapause (Koštál et al., 2008). Expression of the cox gene does not change during the E. scudderiana diapause when compared to the non-diapausing group, although a trend of lower expression during mid-diapause can be observed (McMullen and Storey, 2008). However, in the larval diapause of the moth O. fuscidentalis, treatment with juvenile hormone analog has been shown to lead to increased oxygen consumption, COX enzyme activity and cox1 gene expression, altogether resulting in the termination of diapause (Singtripop et al., 2007). Since prohibitin participates in mitochondrial biogenesis and proper respiratory chain formation, suppression of phb2 expression at the onset of diapause at low temperatures may be due to active mitochondrial depletion in O. nubilalis larvae, a common mechanism of adaptations to low temperatures in insects (McMullen and Storey, 2008). Also, suppression may reflect adaptations to dormancy and prolongation of life during starvation in the C. elegans diapause (Artal-Sanz and Tavernarakis, 2009). Namely, the authors believe that prohibitin promotes longevity by modulating the use of lipids and energy production through the mitochondrial respiratory chain, as they proved this by using RNA interference to suppress the expression of phb1 and phb2 in wild-type and mutants of the roundworm C. elegans. Reducing the expression of phb1/2 shortens the life of wild-type animals, but at the same time prolongs the life of individuals that are on caloric restriction, as well as mutants that cannot enter the so-called diapause or Dauer condition (Artal-Sanz and Tavernarakis, 2009). These findings on C. elegans are consistent with the results obtained in the present study, which showed that phb2 expression was reduced at the onset of diapause at lower temperatures, probably as a form of adaptation to the ongoing starvation and the subsequent metabolic depression. Also, in most organisms, it is known that metabolic depression is the most pronounced in the first part of diapause, while the metabolism recovers over time, ahead of diapause termination. For this reason, during the latter part of diapause, an organism prepares to terminate this hypometabolic state by intensifying the process of mitochondrial biogenesis, in which prohibitin plays an important role (Coates et al., 1997). Metabolic recovery is further evidenced by the increase in the expression of genes whose products are involved in the regulation of energy metabolism, as evidenced in our study.

This study has shown that energy metabolism in warm-acclimated diapausing larvae (dwa2) is distinct from cold acclimated diapausing groups due to higher COX activity, lower concentrations of NAD⁺/NADP⁺coenzymes, faster depletion of adenine nucleotides, and high expression of energy-related mitochondrial genes in the dwa2 group. Moreover, UMP and GMP levels in these larvae are similar to those in the cold-acclimated dca6 group whose metabolism is slowly recovering at low temperatures ahead of diapause termination. This is further supported by the PCA, which has clustered the dwa2 group with the dd and dca6 groups, with higher level of metabolic activity in comparison to dca2 and dca4. Moreover, a similar level of energy metabolism-related gene expression between the dwa2 and dd groups is most likely due to the development of larvae at high temperature (22 ± 1°C). In such conditions energy demands and metabolic rate remain at a high level, regardless of the fact that diapause-destined larvae are supposed to enter diapause due to a shorter photoperiod.

Given that warm-acclimated larvae (dwa2 and dd) had increased gene expression for ADP/ATP translocase, high COX activity, and decreased ATP concentrations, we can assume that larval metabolism is kept at a high level, despite the fact that these larvae are either in preparation for dormancy (dd) or are already dormant (dwa2). As the European corn borer larvae do not feed during diapause, ATP is synthesized in smaller quantities and intensively consumed in comparison to active development (nd). Such conditions lead to depletion of ATP in the cytoplasm, which is probably compensated by increased expression of the ADP/ATP transporter gene. In addition, this gene also has a pattern of increasing expression during the time course of diapause, which indicates its importance for the metabolic recovery prior to diapause termination and resumption of active development. The obtained results of relative gene expression indicate that transcription is in line with the energy needs of O. nubilalis larvae in both warm- and cold-acclimated experimental groups.

While at first glance diapause may seem like just a passive state of developmental arrest, the results of this and previous research on the European corn borer and other diapausing insects show that diapause is characterized by the maintenance of a dynamic equilibrium of energy metabolism. Furthermore, this balancing act is well synchronized with exogenous factors, such as the onset of low temperatures. This allows an organism to model its metabolism to respond to changes in the environment and direct it toward protective activities, such as synthesis of beneficial compounds (e.g., glycerol, trehalose, alanine, etc.) and removal of potentially harmful metabolites. Also, a gradual increase in the number of transcripts of all genes, and possibly their accumulation during diapause in cold conditions, may prepare the organism for termination of diapause and subsequent metamorphosis. After coming out of a long period of rest, during which metabolism was slowed down, mitochondrial biogenesis and reactivation of oxidative metabolism are necessary for rapid continuation of development, and numerous gene transcripts related to mitochondrial energy metabolism serve as a good source for protein synthesis.

The findings presented in this study evidence that diapause at high temperatures is shortened, stressful and leads to high mortality, while the exposure to low temperatures during diapause is substantial for long-term winter survival. Our results point to the possible adverse effects of the looming climate changes and global warming on species that are not accustomed to atypical temperature changes during specific stages of their life cycle, such as winter diapause.