• Open Access

Parental phase status affects the cold hardiness of progeny eggs in locusts

Authors


Correspondence author. E-mail: lkang@ioz.ac.cn

Summary

1. The capacity to adapt to low temperature is critical for the persistence of insect populations in heterogeneous environments. Locusts show remarkable phenotypic plasticity, termed ‘phase change’, in response to local population density.

2. In this study, the hypothesis that population density, as a social factor, affects the cold hardiness of progeny is validated in the migratory locust, Locusta migratoria, which shows remarkable density-dependent phase changes between gregarious and solitarious phases.

3. We demonstrated that eggs of gregarious and solitarious migratory locusts significantly differ not only in size and weight, but also in cold hardiness. Eggs of solitarious locusts are more resistant to cold stress compared with those of gregarious locusts, with longer 50% lethal time at different low temperatures and under different acclimation treatments of cooling rates or cold exposure time, lower upper limit of chill injury zone, and lower sum of injurious temperature resulting from temperature–time regression. The lowest cooling rate (0·05 °C min−1) yields the highest survival of cold for eggs.

4. A custom-made microarray covering 9154 unigenes of the migratory locust demonstrated quite different gene expression profiles in the two phases in response to normal or low temperature. Under cold stress, the gregarious-phase eggs have higher transcriptional levels of heat-shock proteins, DOPA decarboxylases and tyrosine hydroxylase, whereas the solitarious-phase eggs exhibit stimulated lipid metabolism and carboxylic acid transport.

5. Hybridization between the two phases showed that the cold hardiness of eggs from the hybrid with solitarious females is significantly higher than that of the hybrid with gregarious females, and the cold hardiness of eggs from each reciprocal hybrid is close to their maternal origins. These results indicate that the cold hardiness of progeny is affected by the parental phase status.

Introduction

The population density-induced phase polymorphism is ubiquitous in animals, especially in insects, such as some species from Orthoptera, Lepidoptera, Hemiptera, Homoptera and Coleoptera (Applebaum & Heifetz 1999). The migratory locust Locusta migratoria is one of the most hazardous insect pests and exhibits typical polyphenism in morphology, behaviour and physiology in response to population density changes. The swarmed and gregarious locust populations often cause severe damage in agriculture throughout the world. Numerous studies have investigated the polyphenism of locusts, especially the differences between two extreme phases, the gregarious and solitarious phases (Pener & Simpson 2009). The transition between solitarious and gregarious phases is adaptive to temporal and/or spatial changes. The impact of parental phase status on progeny deserves particular attention because of the current research interest in the evolution of polyphenism.

It is found that the embryonic development and phase-dependent progeny characteristics are maternally controlled. Gregarious female locusts have lower fecundity and produce smaller eggs than solitarious locusts (Pener & Simpson 2009). In desert locusts Schistocerca gregaria, eggs from solitarious females produced green hatchlings, and eggs from gregarious females produced more heavily melanized hatchlings. It is assumed that a pheromonal factor is produced by foam plugs, which contain some active agents from female accessory glands (Miller et al. 2008; Pener & Simpson 2009). The close relationship between egg size and melanization degree in hatchlings is probably determined in the ovarioles (Tanaka & Maeno 2010). In the migratory locusts, transcriptomic data characterized the divergence of gene expression in eggs from the two phases of adults, indicating that the parental influence plays an important role in egg divergence, although the divergence of gene expression becomes more obvious from egg to adult (Chen et al. 2010). The phase differences associated with multicellular organismal process, intracellular structure, catalytic activity, lipid and glycan metabolism pathways, and cellular processes were observed at adult stage (Chen et al. 2010). Therefore, the maternal effect is related to the reproductive cycles and the genetic differences in eggs from solitarious and gregarious locusts. However, no research has yet compared the biological characteristics of eggs from the two typical phases of locusts.

The migratory locusts have a broad geographical distribution from tropical to temperate zones, and they can reproduce 1–4 generations per year depending on the local temperature. In temperate regions, eggs are laid at late September and overwinter in soil until the following April or May (Ma 1958). Our previous studies demonstrate significant differences in the cold tolerance of eggs from geographically and seasonally different populations (Jing & Kang 2003, 2004; Jing, Wang & Kang 2005; Wang & Kang 2005; Wang et al. 2006; Qi, Wang & Kang 2007; Wang, Qi & Kang 2010). Our recent work reveals different genome-wide transcriptome and small RNA transcription between the two phases from eggs to adults (Wei et al. 2009; Chen et al. 2010). Thus, we theoretically hypothesize that there are differences in the cold tolerance of eggs from the two phases of parents. Upon the outbreak of locust plagues, the locusts migrate long distances and hybridize with local populations. The consequence is changes in progeny cold tolerance, which affect the survival of eggs after overwintering. However, whether there are differences in the cold hardiness of locust eggs from gregarious and solitarious locusts remains unknown.

In this present study, we investigated the differences in cold hardiness of eggs from solitarious and gregarious locusts. First, we examined the size, mass and water content in eggs from gregarious and solitarious adults. To explore the genetic basis and differentially expressed genes of these eggs, their genome-wide transcriptional expression profiles before and after cold treatments were characterized using the oligo-nucleotide microarray covering 9154 unigenes based on the large-scale expressed sequence tags (ESTs) of the migratory locusts (Kang et al. 2004; Ma, Yu & Kang 2006). The locust microarray was successfully applied to study olfactory-related genes and dopamine pathway that regulates behavioural phase changes (Guo et al. 2011; Ma et al. 2011). Finally, we demonstrated that the cold hardiness of eggs from gregarious and solitarious locusts was significantly different and maternally inherited.

Materials and methods

Rearing of the Migratory Locust

The two phases of migratory locust stock colonies originated from adults obtained in April 2003 from Huanghua County (38′25°N, 117′20°E), Hebei Province, China, and were propagated for 10–11 generations in a laboratory. The gregarious locusts were reared in large, well-ventilated wooden cages (60 × 50 × 50 cm width × length × height) at a density of about 1000 insects per container until the hoppers grew to the adult stage. The solitarious locusts were kept in a separate room with each individual in a metal cage (28 × 12 × 12 cm) as described (Kang et al. 2004; Guo et al. 2011; Ma et al. 2011). One mature male and one female were paired in a wooden cage (32 × 32 × 30 cm) for mating and oviposition. Solitarious and gregarious locusts were carefully kept under similar conditions, a long-day photoperiod (14 h light/10 h darkness cycle) at 30 ± 1 °C. Both nymphs and adults were fed with wheat seedlings and wheat bran. Sterilized sand was used as the oviposition medium.

Egg Collection

Eggs from 50 pairs of gregarious locusts and 40 pairs of solitarious locusts were collected at 9:00 and 21:00 daily to ensure an even stage of development. Five gregarious egg pods were kept together in a plastic cup, and the solitarious ones were kept individually. The egg pods were incubated at 30 °C in sterilized sand with about 10% water content for 7 days to reach the mid stage of development (Wang & Kang 2005). Equal amounts of mid-stage eggs collected at 9:00 and 21:00 of the same day were removed from egg pods and used in all experiments except for examining the difference among various developmental stages.

Measurement of Egg Weight and Water Content

Fresh eggs were weighed on a Mettler (AE240) microbalance (±0·01 mg), and the results were recorded as fresh weight. Then they were placed in numbered vials, dried for 3 days at 60 °C and reweighed. Weight loss was considered as the corresponding water content. Ten individuals were used in each of the 20 replicates.

Supercooling Point Determination

One hundred and twenty healthy gregarious or solitarious eggs were fixed to a thermocouple, which was linked to an automatic recorder (uR100, Model 4152; Yologama Electrical Co, Seoul, Korea). The supercooling point was indicated on the recorder by a sudden spike in the thermocouple temperature. Details were described by Jing & Kang (2004).

Cold Hardiness of Eggs at Different Embryonic Developmental Stages

Locust embryonic development is usually divided into three stages: early stage (anatrepsis), mid stage (blastokinesis) and late stage (katatrepsis). Eggs that were incubated at 30 °C for 2, 7 and 12 days were regarded as early, mid and late stages, respectively (Wang & Kang 2005). To examine the cold hardiness difference in eggs at these three developmental stages between and within the two phases, survival rates were determined after 50 eggs at each developmental stage were exposed to −7·5 °C for 3, 6, 12, 24, 48 or 96 h. The experiment was replicated five times.

Analysis of the Relationship of Time and Temperature With Egg Survival

Eggs of gregarious or solitarious locusts were exposed to low temperatures (from 0 °C to −15 °C, in 2·5 °C increments) for different lengths of time (3, 6, 12, 24, 48, 96 h). The eggs were transferred into plastic tubes, whose temperature was controlled by a programmable refrigerated bath (Polysciences, Warrington, PA, USA). The temperature was decreased at a rate of 1 °C min−1 from 30 °C to the target temperature. After the cold treatment, the temperature was increased at a rate of 1 °C min−1 to 30 °C, and then the eggs were transferred to a 30 °C environmental chamber. The number of hatched eggs was counted everyday to calculate the survival rate. Forty individuals were used in each of five replicates.

Acclimation

Two types of acclimation were performed before egg survival rates were recorded. In one acclimation, eggs were first acclimated at 5 °C for 3, 10, 30 and 60 days, respectively, then exposed to −7·5 °C for 3, 6, 12, 24, 48 and 96 h, respectively. In the other acclimation, eggs were cooled from 30 to −10 °C with a cooling rate of 0·8, 0·2, 0·05 °C min−1 or to −10 °C directly (‘plunge’ treatment). Then the temperature was held at −10 °C for 10 h (this threshold was predetermined to result in c. 90% mortality), and the eggs were warmed to 30 °C at a rate of 0·1 °C min−1. Fifty eggs were used in each test point, and each test point was replicated five times.

Crossing Experiment and Cold Hardiness Analysis

To elucidate the parental effects on the cold tolerance of progeny, two parental crosses (gregarious male (♂) × gregarious female (♀), 30 pairs in one cage; solitarious ♂ × solitarious ♀, one pair in one cage) and two reciprocal crosses (gregarious ♂ × solitarious ♀ or solitarious ♂ × gregarious ♀, five pairs in one cage) were performed. Egg pods were collected every day and incubated at 30 °C to reach the mid stage of development. Eggs were removed from egg pods and cooled from 30 °C to −8 °C at a rate of 1 °C min−1, and the temperature was held at −8 °C for 2, 4, 8, 16 and 32 h, respectively. The eggs were then warmed to 30 °C by 1 °C min−1 after the cold treatments and incubated at 30 °C until hatching to calculate the survival rates. Forty eggs from each crossing group were used in each of six replicates.

Microarray Assay

We used the locust oligo-nucleotide microarray to explore the genetic basis of eggs from gregarious and solitarious locusts and the relationship of gene expression pattern and cold stress in the two phases. Mid-stage eggs were treated at 0 °C, −5 °C and −10 °C for 2 h, respectively, and then kept in liquid nitrogen. Another aliquot of eggs without cold treatment, i.e., maintained at ambient condition, was used for comparison. Three to five aliquots of 40 eggs were collected for the treatment and control groups.

The microarray was designed and hybridized as described by Ma et al. (2011). Total RNA (40 μg) was used to prepare cDNA probes, which were labelled with mono functional Cy3 and Cy5. Three to five independent hybridizations with biological replicates were performed using dye reversal strategy. Direct comparison method was chosen for microarray hybridization.

Quantitative PCR

Based on the microarray result, the differential expression of 15 genes was verified by quantitative real-time PCR as described previously (Wang et al. 2006; Guo et al. 2011; Ma et al. 2011). Primer sequences and annotations of the 15 genes are listed in Table S1 (Supporting information). Data were compared between the two phases at ambient condition and before and after the cold treatment at −5 °C for 2 h within each phase.

Data Analysis

The survival rates of eggs in all treatments were normalized by the percentage of hatched eggs in the control group, in which eggs were incubated at 30 °C. The 50% lethal time (Ltime50) was inferred from the survival rates, which were analysed as a function of temperature or time using probit of spss 11.0. Differences were evaluated statistically using spss 11.0 and statistica 5.0 software either by t-test to compare two means, or by one-way analysis of variance (anova) followed by a Tukey’s test for multiple comparisons. For the acclimation experiments, the general linear model univariate analysis in spss 11.0 was applied to evaluate the interaction effect of the two factors, phase and acclimation condition on eggs’ cold hardiness. Differences were considered significant at < 0·05. Values were reported as means ± SE.

The relationship of egg survival with time and temperature was regressed by the equation = 100/(1 + exp (bt (− c))), where S represents survival rate, t and T represent time and temperature, respectively, and a, b and c are three constant parameters (Nedvêd, Lavy & Verhoef 1998). c is an estimate of the upper limit of chill injury zone (ULCIZ). The ratio −a/b represents the sum of injurious temperature (SIT) with the unit of temperature in centigrade degree and time in hour.

The Limma package was used to correct microarray raw data background with minimum function in R (2.6.0). Intensity signals were normalized for bias elimination using rlowess function, and then data were fitted to fixed-effect anova model. All statistical analyses were conducted in R using the R/maanova software package with array as random effect. Fs-test P value was calculated for each gene. The smallest fold change of 1·5 within the cut-off value of < 0·001 was considered as significant. Hierarchical clustering (average linkage clustering) was performed using cluster software (Stanford University). Dendrograms and heat maps were generated by Java Treeview (Stanford University).

The unigenes were assigned to gene ontology (GO) categories using Blast2GO based on the sequence similarity to NR data base at National Center for Biotechnology Information (NCBI). GO categories were enriched for supplied gene list based on the algorithm presented by GOstat (Beissbarth & Speed 2004). For each GO term, the difference between tested gene group and reference gene group was represented by P value, which was approximated by chi-square test. Fisher’s exact test was used when any expected value of count was below 5, which causes inaccurate chi-square test. A Benjamini multiple-testing correction of the P value was performed by false discovery rate (FDR) (Benjamini & Hochberg 1995). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (Kanehisa et al. 2008) were assigned by searching the KEGG data base and finding the best hit for each sequence. The enrichment analysis of KEGG pathways was the same as GO enrichment analysis.

Results

Weight, Water Content and Supercooling Point of Eggs

The gregarious-phase eggs looked larger than the solitarious-phase eggs (Fig. 1a). This size difference was confirmed by the weight difference. Both fresh and dry weights of gregarious-phase eggs were markedly higher than solitarious-phase eggs (fresh: t = 12·7, < 0·001; dry: t = 9·4, < 0·001; Fig. 1b). However, there was no significant difference in water content or supercooling point of the eggs between the two phases (Fig. 1c, d).

Figure 1.

 The morphology (a), fresh and dry weight (b), water content (c) and supercooling point (d) of gregarious- and solitarious-phase eggs.

Cold Hardiness of Eggs

To explore the potential cold hardiness difference in eggs from gregarious and solitarious locusts, the 50% lethal time (Ltime50) of eggs at three embryonic developmental stages (early-, mid- and late-stage) was determined. More detailed cold hardiness properties, including the relationship of time and temperature with egg survival, Ltime50 or survival rate under various acclimation conditions, were compared only between the mid-stage eggs.

Cold Hardiness of Eggs at Three Embryonic Developmental Stages

The Ltime50 of eggs was estimated at three embryonic developmental stages at −7·5 °C. Mid- and late-stage eggs showed higher cold hardiness than early-stage eggs in both phases. The Ltime50 of solitarious-phase eggs was significantly higher than gregarious-phase ones at the early (18·2 h vs. 4·3 h, t = 41·0, < 0·001), mid (25·5 h vs. 14·5 h, t = 15·7, < 0·001) and late stages (38 h vs. 30·6 h, t = 3·9, = 0·017) (Fig. 2a). Because the locust eggs start the overwintering at the middle stage of embryonic development as eggs in the mid-stage development (Wang & Kang 2005), the cold hardiness of mid-stage eggs was tested in the following experiments.

Figure 2.

 Solitarious-phase eggs show higher cold hardiness than gregarious-phase eggs. (a) Ltime50 of eggs at −7·5 °C at three embryonic developmental stages. (b) Surface (3D) plots of egg survival depend on time and temperature of cold exposure for gregarious and solitarious phases, respectively. (c) Ltime50 of eggs at −7·5 °C after acclimated at 5 °C for various periods. NA represents nonacclimation treatments. (d) Survival of eggs after exposure to −10 °C for 10 h with different cooling rates.

Relationship of Duration of Acclimation and Temperature With Egg Survival

To systematically explore the cold hardiness differences in eggs between two phases, we measured the survival of two-phase eggs when they were exposed to low temperatures from 0 °C to −15 °C with 2·5 °C increments for various times (3, 6, 12, 24, 48, 96 h). The survival curves of eggs were successfully fitted to the extended logistic equation, = 100/(1 + exp (bt (T − c))), with the values = −0·86, = −0·022, = −2·34 for gregarious-phase eggs and = −1·07, = −0·014, = −3·73 for solitarious-phase eggs. This regression explained 79·3% and 83·7% of the variation under least squares loss function (squared correlation coefficient R2) in gregarious-phase and solitarious-phase eggs, respectively (Fig. 2b). The resulting graph represents the dependence of survival on duration of acclimation and temperature as well as the interaction between the two factors. The parameter analysis demonstrated that the solitarious-phase eggs had reduced ULCIZ (−3·7 °C) and decreased SIT (−76·1 h•degree) in comparison with the gregarious-phase ones (ULCIZ, −2·3 °C; SIT, −38·8 h•degree), indicating that the cold hardiness of solitarious-phase eggs was significantly higher than that of gregarious-phase ones.

Cold Hardiness of Eggs Under Various Acclimation Conditions

Two types of acclimation conditions were applied to evaluate the cold hardiness of eggs from gregarious and solitarious locusts. One was constant at 5 °C for various periods. This acclimation promoted the cold hardiness of eggs in both phases (gregarious, F4,14 = 16·3, < 0·001; solitarious, F4,14 = 10·7, = 0·001), and the acclimation for 60 days resulted in the longest Ltime50 (Fig. 2c). Comparison of eggs from gregarious and solitarious locusts revealed that the Ltime50 of solitarious-phase eggs was always longer than that of gregarious-phase counterpart in the same acclimation time. Thus, the solitarious-phase eggs exhibited higher cold hardiness than the gregarious-phase ones. However, as the acclimation time increased, the difference of Ltime50 decreased (nonacclimation: t = 15·7, < 0·001; acclimated 3 days: t = 5·7, = 0·005; acclimated 10 days: t = 20·9, < 0·001; acclimated 30 days: t = 5·3, = 0·006; acclimated 60 days: t = 3·5, = 0·024) (Fig. 2c). There was no interactive effect for the two factors, phase and acclimation time on eggs’ cold hardiness (F9,29 = 0·3, = 0·86).

The eggs were cooled from 30 to −10 °C at a series of rates, including a direct ‘plunge’, compared to which the cold hardiness of eggs from both phases increased at each cooling rate (gregarious, F3,19 = 330·6, < 0·001; solitarious, F3,19 = 185·9, < 0·001) (Fig. 2d). The lowest cooling rate, 0·05 °C min−1, led to the highest survival rate, and the solitarious-phase eggs had a significantly higher survival rate than the gregarious-phase eggs at all cooling rates. The biggest difference was observed at absolutely high cooling rates such as ‘plunge’ (t = 10·8, < 0·001) and 0·8 °C min−1 (t = 12·9, < 0·001) (Fig. 2d). The two factors, phase and cooling rate, had significant interactive effect on eggs’ cold hardiness (F7,39 = 11·4, < 0·001).

Gene Expression Profiles of Eggs at Ambient Condition

To explore the genetic basis of eggs from gregarious and solitarious locusts, we compared their genome-wide transcriptional profiles using the locust oligo-nucleotide microarray. The results showed that 459 genes (5·0%) displayed different expression levels between the two phases. 237 (52%) were up-regulated in gregarious-phase eggs and 222 (48%) were up-regulated in solitarious-phase eggs (Appendix S1, Supporting Information). The expression levels of 15 genes were verified by quantitative real-time PCR (qRT-PCR), which showed consistent results with microarray analysis (Fig. S1, Supporting information).

Among the 459 genes, 243 (53%) can be annotated with Blastx in NCBI. GO enrichment analysis revealed that the up-regulated genes in gregarious-phase eggs were enriched in chitin metabolism, cuticle structure, acid phosphatase activity and myosin complex components (Table 1). The KEGG pathways enriched in circadian rhythm and neuroactive ligand–receptor interaction (Benjamini <0·1) (Table 1). The up-regulated genes in solitarious-phase eggs did not show enrichment in GO categories (Benjamini >0·1), but when P value was not corrected by FDR, processes of DNA metabolism, biopolymer catabolism, RNA-dependent DNA replication and fructose metabolism were enriched (< 0·05) (Table 1). The enriched pathway of up-regulated genes in solitarious-phase eggs was renin–angiotensin system (Benjamini <0·1) (Table 1).

Table 1.   The differentially expressed genes enriched in GO function classes and KEGG pathways in eggs of gregarious and solitarious locusts at ambient condition
GO or KEGG IDGO or KEGG TermP value*Benjamini
  1. BP, biological process; MF, molecular function; CC, cellular component.

  2. *The P value represented the difference between tested gene group and reference gene group. It is approximated by chi-square test. Fisher’s exact test is used when any expected value of count is below 5.

  3. Benjamini is the multiple-testing correction of the P value by false discovery rate.

Up-regulated in gregaria
 GO:0042302 (MF)Structural constituent of cuticle4·42E-064·04E-04
 GO:0006030 (BP)Chitin metabolic process2·51E-054·04E-04
 GO:0016459 (CC)Myosin complex3·83E-043·47E-03
 GO:0003993 (MF)Acid phosphatase activity2·23E-030·0171
 KEGG:ko04710Circadian rhythm3·16E-051·93E-03
 KEGG:ko04080Neuroactive ligand–receptor interaction2·32E-047·06E-03
 KEGG:ko04530Tight junction5·01E-030·10
 KEGG:ko00061Fatty acid biosynthesis0·01930·29
 KEGG:ko04810Regulation of actin cytoskeleton0·02560·31
 KEGG:ko04612Antigen processing and presentation0·04410·38
Up-regulated in solitaria
 GO:0006259 (BP)DNA metabolic process0·02440·5
 GO:0043285 (BP)Biopolymer catabolic process0·03860·5
 GO:0006278 (BP)RNA-dependent DNA replication0·03130·5
 GO:0006000 (BP)Fructose metabolic process0·03940·5
 KEGG:ko04614Renin–angiotensin system8·14E-040·0611
 KEGG:ko00511Other glycan degradation0·0280·542
 KEGG:ko00150Androgen and oestrogen metabolism0·0330·542
 KEGG:ko04610Complement and coagulation cascades0·04150·542

Gene Expression Profiles of Eggs Under Cold Stress

To investigate the relationship of gene expression pattern and cold stress in the two phases, the locust eggs were exposed to three levels of cold stress, 0 °C, −5 °C and −10 °C, and the differentially expressed genes at each condition were identified by the fixed-effect anova model. In the 9154 unigenes, 274 (3%) and 477 (5%) cold-response genes were identified from gregarious-phase and solitarious-phase eggs, respectively. 132 (48%) genes in gregarious-phase eggs and 259 (54%) genes in solitarious-phase eggs can be annotated with Blastx in NCBI (Appendices S2 and S3, Supporting information). The two types of eggs shared only 87 cold-response genes, whereas 68% and 82% of the differentially expressed genes were unique to gregarious-phase and solitarious-phase eggs, respectively (Fig. 3a). The expression levels of 15 genes before and after −5 °C cold treatment in each phase were confirmed by quantitative real-time PCR (qRT-PCR) (Fig. S1, Supporting information).

Figure 3.

 The response genes to 0 °C, −5 °C and −10 °C treatments in gregarious- and solitarious-phase eggs. (a) Comparison of cold-response gene numbers between the two phases. Only 87 genes responded to cold stress in both phases. (b) and (c) Dendrograms and heat maps of gregarious-phase eggs (G) and solitary-phase eggs (S), respectively. Detailed expression pattern and annotation of genes that belong to specific functional groups from clusters of gregarious phase (A–C) and solitarious phase (A and B).

The gene expression patterns in eggs of gregarious and solitarious locusts under the three cold stress conditions were quite different. For gregarious-phase eggs, the gene expression level showed high variation at −5 °C and −10 °C treatments (Fig. 3b). Under −5 °C cold stress, the major change was the up-regulation of genes encoding endopeptidase inhibitors (Fig. 3b, cluster B), which depress p53 signalling pathway and complement and coagulation cascades (Table 2). At −10 °C, more genes were up-regulated and two groups of genes exhibited apparent induction (Fig. 3b, cluster C). One group is heat-shock proteins (hsps), including hsp90 (LM00716), hsp40 (LM00618), hsp70s (LM00690, LM00727), hsp20 (LM00711) and hsp105 (LM06070). The other group contains genes of DOPA decarboxylases (LM01398, LM02624) and tyrosine hydroxylase (LM01233), which are involved in the production of neurotransmitter dopamine. Moreover, genes of chitin metabolism and cuticle structure were down-regulated under −5 °C and −10 °C cold stress (Fig. 3b, cluster A). In solitarious-phase eggs, the most remarkable gene expression change occurred under 0 °C cold stress (Fig. 3c), when about half of the 477 cold-response genes were up-regulated and half were down-regulated. The up-regulated genes are involved in lipid metabolism and carboxylic acid transport (Fig. 3c, cluster B). Several pathways for catabolism of amino acids, including lysine, valine, leucine, isoleucine, tryptophan, and for metabolism of butanoate and fatty acid were stimulated (Table 2). Down-regulated genes are involved in microtubule-based movement and oxygen transport (Fig. 3c, cluster A).

Table 2.   The cold-response genes enriched in GO function classes and KEGG pathways in eggs of gregarious and solitarious locusts
GO or KEGG IDGO or KEGG TermP value*Benjamini
  1. BP, biological process; MF, molecular function; CC, cellular component.

  2. *The P value represented the difference between tested gene group and reference gene group. It is approximated by chi-square test. Fisher’s exact test is used when any expected value of count is below 5.

  3. Benjamini is the multiple-testing correction of the P value by false discovery rate.

Gregaria cluster A (down-regulated)
 GO:0006030Chitin metabolic process (BP)1·16E-062·14E-05
 GO:0042302Structural constituent of cuticle (MF)3·80E-065·11E-05
 KEGG:ko04710Circadian rhythm3·09E-085·26E-07
 KEGG:ko04080Neuroactive ligand–receptor interaction3·25E-042·77E-03
Gregaria cluster B (up-regulated)
 GO:0004866Endopeptidase inhibitor activity (MF)1·88E-049·44E-03
 KEGG:ko04115p53 signalling pathway0·01610·112
 KEGG:ko04610Complement and coagulation cascades0·01610·112
 KEGG:ko00310Lysine degradation0·03290·153
Solitaria cluster A (down-regulated)
 GO:0005921Gap junction (CC)2·79E-053·72E-03
 GO:0007018Microtubule-based movement (BP)7·94E-055·86E-03
 GO:0005344Oxygen transporter activity (MF)1·50E-046·65E-03
 KEGG:ko03320PPAR signalling pathway8·94E-056·44E-03
 KEGG:ko00625Tetrachloroethene degradation2·21E-030·0577
 KEGG:ko04540Gap junction2·41E-030·0577
 KEGG:ko00051Fructose and mannose metabolism5·34E-030·0777
 KEGG:ko00591Linoleic acid metabolism5·40E-030·0777
Solitaria cluster B (up-regulated)
 GO:0006629Lipid metabolic process (BP)0·01110·361
 GO:0046942Carboxylic acid transport (BP)0·01290·361
 GO:0008236Serine-type peptidase activity (MF)0·02320·361
 GO:0019211Phosphatase activator activity (MF)0·03080·361
 GO:0046914Transition metal ion binding (MF)0·04460·361
 KEGG:ko00310Lysine degradation2·35E-040·0104
 KEGG:ko00511Other glycan degradation2·77E-040·0104
 KEGG:ko00650Butanoate metabolism2·82E-030·0528
 KEGG:ko00071Fatty acid metabolism4·41E-030·0661
 KEGG:ko00072Synthesis and degradation of ketone bodies6·61E-030·0826
 KEGG:ko00280Valine, leucine and isoleucine degradation8·24E-030·0858
 KEGG:ko00380Tryptophan metabolism0·01030·0858

Maternally Heritable Effects Contribute to the Cold Hardiness of Eggs

To elucidate the parental effects on the cold tolerance of progeny, we performed two parental and two reciprocal crosses between solitarious and gregarious locusts, and analysed the cold hardiness of their eggs by measuring Ltime50 at −7·5 °C. There were significant differences among the four stocks (Fig. 4) (F3,31 = 59·2, < 0·01). The Ltime50 of two reciprocal hybrid stocks was intermediate of their parental stocks. The cold hardiness of the hybrid gregarious ♂ × solitarious ♀ was significantly higher than the hybrid solitarious ♂ × gregarious ♀. Furthermore, the cold hardiness of eggs from each reciprocal hybrid was close to their maternal origins, indicating that the cold hardiness of progeny is affected by the parental phase status, and reciprocal crosses between solitarious and gregarious locusts can change the cold tolerance of eggs.

Figure 4.

 Ltime50 at −7·5 °C for eggs from gregarious locusts (G♀ × G♂), solitary locusts (S♀ × S♂) and two reciprocal crosses (G♀ × S♂, S♀ × G♂).

Discussion

In our study, we demonstrated for the first time that there are phenotypic differences in the cold hardiness between gregarious-phase and solitarious-phase eggs of the migratory locusts, and these differences are maternally inherited. Eggs of gregarious and solitarious locusts have different gene expression profiles in response to low temperatures. Solitarious-phase eggs are more resistant to cold stress than gregarious-phase eggs as shown by the decreased ULCIZ and SIT values and the longer Ltime50 in the same acclimation treatments. For example, at −7·5 °C low temperature, 25·5 h were needed to kill 50% of solitarious-phase eggs, whereas 14·5 h resulted in 50% mortality in gregarious-phase eggs. However, we did not observe difference in another standard metric of insect cold hardiness, the supercooling point. In fact, different geographical and seasonal populations have similar supercooling points, although their cold tolerance is significantly different (Jing & Kang 2003, 2004). In general, the supercooling point is an indicator for the limits of geographical distribution of an insect population or species (Denlinger & Lee 2010), whereas ULCIZ and SIT values are better predictors of overwintering ability. Thus, solitarious-phase eggs that have higher cold hardiness should be able to overwinter more successfully than gregarious-phase eggs, but this difference may not influence their geographical distribution. Although numerous differences between solitarious and gregarious locusts have been extensively studied previously (Pener & Simpson 2009), the cold hardiness of locust eggs we report here provides novel insights into the phase-related characteristic of locusts.

Locust eggs adopt a freeze-avoiding strategy in response to coldness and can keep their body fluids below the ordinary melting point (Jing & Kang 2003, 2004). Several underlying mechanisms for the freeze avoidance in insects have been proposed, including removing the ice nucleators that initiate ice formation, synthesizing antifreeze proteins, accumulating sugars and polyols, and stabilizing membranes at low temperatures (Sinclair et al. 2003). Our previous studies showed that there are geographical and seasonal variations in the cold hardiness of eggs of the migratory locusts (Jing & Kang 2003, 2004). These variations are partially due to the different expression of heat-shock proteins and accumulation of sugars and polyols (Wang & Kang 2005; Wang, Qi & Kang 2010). One explanation for the phase-specific difference in cold hardiness of locust eggs is that some substances with anti-freezing activity exist in the yolk from maternal oocytes. A second possibility is that the embryos themselves can produce antifreezing substances in response to coldness. Although our microarray studies revealed the difference in egg metabolism between solitarious and gregarious phases, further investigations are required to depict the underlying mechanism.

In hybridization experiments, we demonstrated that the phase-specific cold hardiness of locust eggs is partially maternally controlled. The Ltime50 of each reciprocal hybrid of solitarious and gregarious locusts is closer to their maternal origins. In fact, several other phase-related traits of locusts, for example, hatching behaviour, coloration, mass, ovariole number, morphometry and development time, also display maternal effects (Pener & Simpson 2009). Such phase characteristics are thought to accumulate across generations through the maternal effect, and the trans-generational transmission of phase is an epigenetic phenomenon (Simpson & Miller 2007). Crowding of solitary-reared parents induces the development of gregarious characteristics in their hatchlings and vice versa. Although these phase-related traits are maternally inherited, they do not necessarily share the same underlying mechanism. Several studies reported that some compounds from the gregarious egg foam of desert locusts have behavioural gregarizing activity and cause dark coloration of isolated hatchling (Simpson & Miller 2007; Miller et al. 2008). Oocyte development regulated by juvenile hormone may contribute to the characteristics of eggs of two phases (Pener & Simpson 2009). The differential gene expression of juvenile hormone superfamily in the locusts could provide some cues (Kang et al. 2004). Other possible mechanisms include chemical modifications of histones, alternative splicing that produces different transcripts from a single gene and modulation of gene expression by small noncoding RNAs (Brennecke et al. 2008; Wei et al. 2009).

Microarray analysis has been proved to be a powerful technique to understand the processes and genes involved in the cold acclimation in several species of insects. One was performed in Drosophila melanogaster, where stress proteins, including Hsp23, Hsp26, Hsp83 and Frost, as well as membrane-associated proteins contribute to the cold-hardening response (Qin et al. 2005). Another was carried out in the goldenrod gall moth Epiblema scudderiana. A LIM protein, EsMlp, with a possible role in myogenesis, and six plasma membrane transporters were up-regulated in cold-exposed larvae (Storey & McMullen 2004). The physiological mechanisms of cryoprotective dehydration in the Arctic springtail Megaphorura arctica were recently reported (Clark et al. 2009). The production and mobilization of trehalose, protection of cellular systems via small heat-shock proteins, and tissue/cellular remodelling were induced during the dehydration process. Our present study revealed that under cold stress, the gregarious-phase eggs of migratory locusts exhibit higher expression of heat-shock proteins, whereas the solitarious-phase eggs show stimulated lipid metabolism and carboxylic acid (fatty acid and amino acid) transport. High concentration of heat-shock proteins could be toxic, directly interfering with ongoing processes in the cell (Feder & Hofmann 1999) or hindering reproduction ability of insects (Wang, Kazemi-Esfarjani & Benzer 2004). Considering the lower cold hardiness of gregarious-phase locust eggs, high expression of heat-shock proteins may rescue them from chilling injuries but with more fitness cost than solitarious counterparts. Lipid metabolism is an important physiological adaptation, probably involved in maintaining membrane lipid fluidity during cold stress, and some free amino acids have been shown to have cryoprotective properties (Clark & Worland 2008). Therefore, the active lipid metabolism and carboxylic acid transport could be responsible for the higher cold hardiness of solitarious-phase locust eggs. Interestingly, DOPA decarboxylases and tyrosine hydroxylase, which are highly expressed genes in dopamine pathway in gregarious nymphs (Ma et al. 2011), are also up-regulated in gregarious-phase eggs under cold stress, suggesting that the phase traits are closely associated with the cold tolerance characteristics of progeny. Although we have disclosed the gene transcript response towards cold stress here, the direct connection to physiology via protein synthesis should be addressed in future work.

Genes involved in structural constituent of cuticle and chitin metabolism in gregarious-phase eggs and cytoskeleton like microtubules in solitarious-phase eggs were down-regulated under cold stress. There is little evidence to support the relativity of cuticle modulation and cold tolerance in insects, apart from a study showing that seasonal cuticular modifications contributed to the inhibition of inoculative freezing in the fire-coloured beetle Dendroides canadensis (Olsen, Li & Duman 1998). Low-temperature alterations of cytoskeleton have been noted in several species of plants and animals. In some cases, microtubules depolymerize to enhance cold tolerance, as in Euplotes focardii, a cold-adapted Antarctic ciliate (Pucciarelli, Ballarini & Miceli 1997). A decrease in microtubule abundance was evoked by exposure to low temperature in nondiapausing Culex pipiens (Kim & Denlinger 2009). In other cases, microtubule assembly was induced and cytoskeleton components were up-regulated by low temperature (Clark & Worland 2008).

In our study, circadian rhythm genes seem to be differentially expressed between the two phases at ambient condition (Table 1) and in gregarious-phase eggs under cold stress (Table 2). We cannot exclude the possibility of developing asynchrony between the two phases even though we have carefully controlled the sampling procedure, sampling twice daily at fixed times. Whether the circadian rhythm genes take part in cold tolerance in locusts is not known, and specific experiments would be necessary to explore this topic further.

The phase-specific cold hardiness of eggs may have important adaptive significance for locust survival and reproduction. A high level of cold hardiness in solitarious-phase eggs could increase their success in overwintering to maintain the population density in the coming year. However, for gregarious locusts, more energy sources are reserved for migration rather than for reproduction (Rankin & Burchsted 1992). When locust plagues outbreak, long-distance migration often occurs. In the fields, the migratory gregarious populations probably hybridize with local solitarious population. Upon the reciprocal hybrid, the cold hardiness of their eggs will decrease, resulting in high mortality after overwinter. Therefore, elucidating the difference in cold hardiness between the two phases and the genetic characteristics will shed light on the locust population dynamics based on the variation of cold hardiness of progeny eggs at local and invasion area.

Acknowledgements

This work was supported by the grants of Chinese Academy of Sciences (KSCX2-YW-N-087), Natural Science Foundation of China (30830022) and Ministry of Agriculture of China (2009ZX08009-099B).

Ancillary