Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Acute and chronic gregarisation are associated with distinct DNA methylation fingerprints in desert locusts


Desert locusts (Schistocerca gregaria) show a dramatic form of socially induced phenotypic plasticity known as phase polyphenism. In the absence of conspecifics, locusts occur in a shy and cryptic solitarious phase. Crowding with conspecifics drives a behavioural transformation towards gregariousness that occurs within hours and is followed by changes in physiology, colouration and morphology, resulting in the full gregarious phase syndrome. We analysed methylation-sensitive amplified fragment length polymorphisms (MS-AFLP) to compare the effect of acute and chronic crowding on DNA methylation in the central nervous system. We find that crowd-reared and solitary-reared locusts show markedly different neural MS-AFLP fingerprints. However, crowding for a day resulted in neural MS-AFLP fingerprints that were clearly distinct from both crowd-reared and uncrowded solitary-reared locusts. Our results indicate that changes in DNA methylation associated with behavioural gregarisation proceed through intermediate states that are not simply partial realisations of the endpoint states.


Modification of neural DNA by cytosine methylation is emerging as an important mechanism in tailoring behavioural phenotypes to environmental conditions, including the social environment1,2,3,4. In most instances, however, the mechanistic role of DNA methylation in the chain of events from environmental signals to changes in behavioural phenotype is still poorly understood4. To what extent are changes in DNA methylation responsible for bringing about behavioural change, as opposed to serving to consolidate changes that first arose through other mechanisms? This is related to an even more fundamental question in understanding phenotypic transitions: are individuals in transition simply intermediates between more extreme endpoints, or are they better understood as ‘third states’? The answer may differ at different levels of analysis, as similar behavioural states may be underpinned by different mechanistic states.

Phenotypic plasticity is particularly common in insects, a fact implicated in their evolutionary success5. A striking example is provided by phase polyphenism in locusts. Locusts are grasshoppers (Acrididae) that can transform between two extreme phenotypes known as the solitarious and gregarious phase, which differ profoundly in morphology, physiology and behaviour6. Solitarious-phase locusts are cryptic and shy, and avoid conspecifics; gregarious-phase locusts are active and mobile and seek out conspecifics, causing them to aggregate in swarms. Several distantly related grasshopper species show phase polyphenism, with migratory locusts (Locusta migratoria) and desert locusts (Schistocerca gregaria) being amongst the most extreme and economically relevant. The sole direct environmental driver of phase change is the presence or absence of conspecifics. Solitarious desert locusts acquire gregarious behaviour within a few hours of forced crowding7,8. Behavioural solitarisation of long-term gregarious locusts is markedly slower, indicating a consolidation of the gregarious state with prolonged crowding. In desert locusts, phase state at hatching is additionally determined by trans-generational epigenetic inheritance9.

Phase change in locusts provides an attractive model for addressing fundamental questions about the role of DNA methylation in behavioural plasticity. Neural DNA methylation could conceivably contribute to several different aspects of behavioural phase polyphenism. It could be part of the effector cascade that initiates behavioural change; it could contribute to the consolidation of gregarious behaviour that occurs with prolonged crowding within a lifetime; and it could contribute to the inheritance of phase state across generations.

The DNA in locust central nervous systems is heavily methylated by insect standards: in S. gregaria, methylation occurs on 1.6–1.9% of all genomic cytosines and on over 3% of the cytosines in exons10,11. These values are over tenfold higher than in honeybees, where methylation is implicated in caste polyphenism12,13,14, suggesting that DNA methylation has important functions in locust behaviour. Two forms of (cytosine-5)-methyltransferases (DNMTs) are primarily responsible for DNA methylation in animals: DNMT3s, traditionally viewed as de novo methylation enzymes, and DNMT1s, which were thought to maintain the methylation state. The roles are not that clear cut, however15. In post-mitotic neurones, for example, DNMT1 and DNMT3 have overlapping roles in synaptic plasticity16. Unsuccessful attempts to identify Dnmt3 transcripts in EST libraries from either S. gregaria or L. migratoria led to the suggestion that Acridids may have secondarily lost this gene, as have several other insect clades10,17,18. Wang and colleagues19 reported the presence of a Dnmt3 gene in the L. migratoria genome but did not include their evidence, so that further studies are needed to resolve whether a functional gene is present. In species that have lost DNMT3, de novo methylation might be accomplished by DNMT118.

A practical difficulty is the lack of a sequenced reference genome for the desert locust and the huge genome sizes of Acrididae in general. In principle, several approaches are currently available for genome-wide studies of DNA methylation. Analysis of methylation-sensitive amplified fragment length polymorphisms (MS-AFLP), a gel electrophoresis-based technique, is still the most widely used approach in non-model organisms without a reference genome, despite its limitations20. MS-AFLP can only screen a limited number of loci simultaneously, and these remain anonymous, i.e., differences in banding pattern cannot be directly linked to known genomic loci. In model species, MS-AFLP has been largely superseded by more powerful techniques that include whole-genome bisulfite sequencing (WGBS) and reduced representation bisulfite sequencing (RRBS). However, these techniques are not readily transferrable to organisms for which a reference genome is not yet available21. WGBS would presently be prohibitively difficult and expensive in desert locusts given the huge genome size of 8.55 Gb22 and the expected abundance of repetitive elements (60% of the genome in L. migratoria19). RRBS is only sensible if the reads can be aligned against a reference genome, which presently precludes its use in desert locusts. The considerable technical complexities of RRBS23 also make it less suitable for an initial screening of the time course of changes. In L. migratoria, a species with a fully sequenced genome of about 6.5 Gb, RRBS identified about 90 differentially methylated genes in the brains of solitarious and gregarious nymphs19; this relatively moderate number probably represents only the tip of an iceberg and highlights the challenges of RRBS in locusts even when both a reference genome and substantial laboratory resources are available. New NGS-based methods are now appearing that are designed specifically to overcome the limitations encountered in non-model species (e.g., refs 21 and 24). At present, the primary utility of MS-AFLP is to inform the design of more detailed follow-up experiments using such emerging NGS approaches.

For this study, we chose MS-AFLP to compare the neural DNA methylation fingerprints of desert locusts with identical parental histories, but different individual social rearing histories. The reasons for this choice were (i) the fact that MS-AFLP does not require a sequenced reference genome; and (ii) the relatively low cost of MS-AFLP, which enabled us to individually profile a reasonable number of animals for statistical analysis. The study was designed to answer three questions. First, do long-term solitarious and gregarious desert locusts show differences in their global pattern of neural DNA methylation, as was recently reported in migratory locusts19. This question is of interest because the two species are only distantly related and have evolved phase polyphenism independently25. The primary focus of our study, however, was on whether the neural DNA methylation fingerprint changes over a timescale of crowding that is sufficient for behavioural gregarisation. We therefore asked whether a day of crowding is sufficient to cause detectable changes in the neural DNA methylation fingerprint of solitary-reared locusts; and if so, whether the methylation fingerprint of these acutely gregarised locusts already resembles that of long-term gregarious locusts.


Locust rearing and treatments

Desert locusts (Schistocerca gregaria Forskål, 1775) were obtained from an inbred gregarious colony at Leicester. Solitarious-phase locusts were produced from this stock by transferring them within a day of hatching into individual cages and rearing them in visual, tactile and olfactory isolation26. All locusts were maintained on a diet of fresh seedling wheat and dry wheat germ under a 12:12 photoperiod.

All locusts were virgin adults sacrificed 17–21 days after the final moult. Long-term gregarious (LTG) locusts were removed from the colony as final larval instars, sexed, and set up as one all-male and one all-female cohort of 40 each in separate tanks (40 × 30 × 25 cm3) in the controlled-environment room that also housed the solitarious locusts. Solitarious locusts were offspring from a single gregarious mother (first-generation solitarious, 1GS). There were three treatment groups of four males and four females each: (i) n = 8 1GS locusts that never experienced crowding; (ii) n = 8 LTG locusts; and (iii) n = 8 behaviourally gregarised 1GS locusts. These were produced by placing four male and four female 1GS locusts in the tanks that housed the 40 LTG virgins of the respective sex for 24 h before sacrifice. Locusts were sacrificed by decapitation and immediate dissection under ice-cold saline. The brain (excluding the retinae) and the thoracic ganglia were dissected out and snap-frozen on dry ice.

MS-AFLP analysis

Differences in DNA methylation patterns were detected by MS-AFLP analysis in n = 4 independent samples per treatment group, for a total of N = 12 samples. Each sample comprised the pooled brains and thoracic ganglia from one arbitrarily chosen male and female within the same treatment group. DNA was extracted with the QIAamp DNA Micro Kit (QIAGEN) following the manufacturer’s instructions.

Restriction digestion

The MS-AFLP protocol was based on27. For each sample of genomic DNA, one 500 ng aliquot was digested with EcoRI and MspI by combining 3 μl target DNA, 0.05 μl EcoRI (20,000 units), 0.25 μl MspI (20,000 units), 1 μl 10× NEBuffer 4 and 5.7 μl H2O; another 500 ng aliquot of genomic DNA was digested with EcoRI and HpaII by combining 3 μl target DNA, 0.5 μl EcoRI (20,000 units), 0.5 μl HpaII (10,000 units), 1 μl 10× NEBuffer 1 and 5.45 μl H2O at 37 °C for 3 h.

Adapter ligation

The EcoRI-MspI and EcoRI-HpaII restriction-digested products were ligated with EcoRI and HpaII-MspI adaptors (Table 1). The EcoRI adaptor was prepared from 5 μl EcoRI-F and 5 μl EcoRI-R, mixed in a final concentration of 5 pmol μl−1 each; the HpaII-MspI adaptor was prepared from 25 μl HpaII-MspI-F and 25 μl HpaII-MspI-R, mixed in a final concentration of 50 pmol μl−1 each. Both mixes were incubated at 65 °C for 10 min. For ligation, 3 μl digested product was combined with 7 μl of ligation reaction mixture (1 μl EcoRI adapter, 1 μl HpaII-MspI adapter, 0.25 μl T4 DNA ligase (400,000 units), 1 μl 10× T4 ligase buffer (New England Biolabs) and 3.75 μl H2O) at 37 °C for 3 h and then left overnight at room temperature. The ligation products were diluted with 100 μl of H2O and used as the template for pre-amplification.

Table 1 Sequences of ligation adapters, pre-amplification primers and selective amplification primers.


The pre-amplification PCR used 1 μl of ligation product with 1 μl each of EcoRIpre and HpaII-MspIpre primers (10 pmol ml−1; Table 1), and 7 μl of the reaction mix (0.8 μl 2.5 mM deoxynucleotide triphosphates (dNTPs), 1 μl 10× Paq5000 Hot Start Reaction Buffer, 0.3 μl Paq5000 Hot Start DNA Polymerase (500 units), 0.8 μl 25 mM MgCl2, 4.1 μl sterile H2O). The PCR conditions were 94 °C for 2 min, followed by 20 cycles of 94 °C for 30 s, 60 °C for 1 min and 72 °C for 1 min, followed by a final extension of 5 min at 72 °C. 3 μl of each PCR product was run on 3% agarose gel and appearance of a smear of DNA on the gel indicated that the pre-amplification PCR was successful.

Selective amplification

Seven μl of PCR products were diluted with 93 μl of H2O and used as the template for selective amplification. We used four different selective EcoRI primers and three different HpaII-MspI primers (Table 1), giving twelve unique EcoRI/HpaII-MspI primer pair combinations. In order to reduce the number of bands in the subsequent gel electrophoresis to a manageable number, each primer combination was used in a separate PCR, giving twelve PCR products per sample that were subsequently run on separate gels. The selective PCR reaction mixtures contained 1 μl pre-amplified product, 1 μl each of one of the HpaII-MspI primers and of one of the EcoRI primers (10 pmol ml−1) and 7 μl reaction mix (same as used for pre-amplification). PCR conditions were: (i) 94 °C for 2 min; (ii) 13 cycles of 30 s at 94 °C, 30 s at 65 °C (0.7 °C reduction per cycle) and 1 min at 72 °C; (iii) 23 cycles of 30 s at 94 °C, 30 s at 56 °C and 1 min at 72 °C; and (iv) a final extension at 72 °C for 5 min followed by a holding step at 4 °C.

Gel electrophoresis

PCR products were diluted with 100 μl H2O. 10 μl of diluted PCR product was mixed with 3 μl 1× loading buffer (Elchrom Scientific, Cham, Switzerland) and run on 9% poly(NAT) gels (Elchrom) on an Elchrom Origins electrophoresis system (120 V, 81 min at 55 °C). Gels were stained in the dark with SYBR Gold (Invitrogen; 1:10 000 in TAE buffer) followed by destaining in 100 ml TAE buffer alone.

Statistical analysis

Bands were scored automatically as either present or absent by the Java program GelJ28. Positions of loci are reported as molecular weights (base pairs). GelJ matches bands across different gel lanes based on a user-specified tolerance value (in base pairs) below which bands are considered identical. To ensure that our results were not sensitive to this arbitrary tolerance level, we generated matrices of band scores using tolerance values of 1–40 using custom R scripts. The resulting matrices were analysed for differentiation between groups by principal coordinates analysis (PCoA) and by analysis of molecular variance (AMOVA) in the R package msap29. We investigated the sensitivity of the ΦST value to the choice of tolerance value. This identified a broad range of tolerances which gave the same robust result (see Results and Discussion). Finally, to ensure that the ΦST values generated across this range of tolerance values were not due to chance, we compared ΦST from the real data with bootstrapped ΦST values from data generated by random sampling with replacement (N = 1000).

All R scripts used are available at All statistical analysis was carried out in R 3.2.330.

Results and Discussion

Our analysis of the MS-AFLP patterns across solitary-reared, 24 h crowded and crowd-reared locusts indicated clear differences between the three groups. This result was robust across a range of band scoring tolerances—the distance up to which bands in a given gel position are matched as identical between samples (Fig. 1). As expected, tolerance had some effect on our results. At low tolerances (<10 bp), a large fraction of the bands within each sample are treated as unique, leading to low calculated levels of differentiation between the groups (ΦST). Tolerances in the range of 10–25 bp produced robust ΦST values of approximately 0.2. The principal coordinate analysis for each of these tolerance levels is shown in Supplementary Figures S1–S3. Above this range, ΦST began to drop. This is to be expected, as bands that represent genuinely different fragments will now be grouped together, leading to less apparent differentiation between groups. This pattern of increasing and then decreasing ΦST is not due to chance as the bootstrapped data, generated by random sampling with replacement, do not show a similar pattern. Importantly, across the entire range of tolerances, the ΦST values obtained in the real data (red points in Fig. 1) are well outside the bootstrapped ΦST distributions (grey points and black boxplots in Fig. 1).

Figure 1

Sensitivity of our analysis to the tolerance value used for matching bands between samples.

ΦST represents the apparent degree of differentiation between the groups (solitary-reared, crowd-reared, and solitary-reared crowded for 24 h) and is plotted over the range of tolerance values (in base pairs). The red points represent ΦST values calculated from the real data, the grey points those calculated from N = 1000 bootstrapped data sets (generated by random sampling with replacement).

The following analysis is based on a tolerance value of 10 bp. This identified 294 unique AFLP bands (loci); of these, 282 were identified as methylation-susceptible based on different digestion patterns with HpaII and MspI, and 162 showed different banding patterns between individual samples (MS-polymorphic loci). Solitary-reared locusts had a slightly lower proportion of unmethylated loci than crowd-reared locusts (11.8% vs. 16.3%) and a slightly higher proportion of hypermethylated loci (65.6% vs. 55.9%). Acutely crowded solitary-reared locusts showed proportions of methylation that were intermediate (Table 2). The three treatment groups showed significant multi-locus differentiation in their methylation fingerprints (AMOVA, ΦST = 0.2264, p = 0.0002). Figure 2 gives a simplified representation of the multi-locus differentiation between the samples. The two axes represent the first two principal coordinates, which together explained 35.1% of the total variation.

Table 2 Proportion of methylation-sensitive restriction band patterns found in the CNS of locusts of different phase state, and their corresponding methylation status; methylated cytosines are indicated in bold type.
Figure 2

Principal Coordinate Analysis (PCoA) of epigenetic differentiation between uncrowded solitary-reared locusts (S), long-term gregarious locusts (G) and solitary-reared locusts crowded for 24 h (C), as identified by MS-AFLP (10 bp band matching tolerance).

The first two coordinates (C1, C2) are shown with the percentage of variance explained by them. Group labels show the centroid for each group, points correspond to individual MS-AFLP samples, ellipses represent their average dispersion around the group centroids.

A pair-wise comparison between crowd-reared and solitary-reared locusts identified significant epigenetic differentiation (ΦST = 0.2810, p = 0.0291), indicating that phase change in desert locusts entails modification of the neural DNA methylation pattern. This is maybe the least surprising of our results, considering that differences in brain DNA methylation between long-term phases have been previously reported in the distantly related migratory locust L. migratoria19. The differences observed in our experiment arose within a single generation, because we used solitary-reared locusts that were the direct offspring of long-term gregarious parents. It would now be interesting to see whether isolation over multiple generations further deepens the epigenetic differences between the two phases.

Our key finding, however, is that crowding solitary-reared locusts for 24 h resulted in a neural DNA methylation fingerprint that was distinctly different both from uncrowded solitary-reared locusts (ΦST = 0.238, p = 0.0283) and from crowd-reared locusts (ΦST = 0.166, p = 0.0288). This uncovers a disjunct between the global neural DNA methylation pattern and the behavioural phase state. Although one day of crowding is sufficient to establish gregarious behaviour7,8, we find that the neural MS-AFLP fingerprint is at this point still markedly different from that in long-term gregarious locusts. The time-course of behavioural gregarisation has been characterised primarily through work carried out in juveniles, where gregarious behaviour is established within 4 h of crowding (e.g., refs 7 and 31). Therefore, the difference that we observed between the methylation fingerprints of 24 h crowded and crowd-reared adult locusts could reflect an incomplete or slower behavioural gregarisation response in adults, e.g., due to age-related limits of plasticity. Bouaïchi and colleagues8 reported that very young adults, 2 days post-fledging, responded more strongly to crowding than older adults, 7–35 days post-fledging (‘Experiment 1’ in ref. 8). However, this effect was not robust (not evident in ‘Experiment 2’ of ref. 8), and after 8 h of crowding, adult locusts were found behaviourally indistinguishable from crowd-reared locusts. To the extent that they have been documented, age-related differences in the gregarisation response are minor and set against a clear behavioural transformation in less than 24 h, irrespective of age. We therefore expect that the result of this study generalises to juveniles.

Interestingly, the data points from 24 h crowded samples were set apart from the solitary-reared samples and the crowd-reared samples by shifts along both of the first two principal coordinate axes (Fig. 2). Although the exact position of the group centroids relative to the two axes depended on the tolerance value used in the band scoring algorithm, a clear triangular separation was maintained across the entire range of sensible tolerance values (Supplementary Figures S1–S3). In other words, the three groups never fell along a single line in the PCoA plots. We interpret this as evidence that the methylation patterns seen after 24 h of crowding are not simply intermediate between the two extremes, but reflect a distinct transitional epigenetic state.

The further changes in methylation that occur only some time after the first 24 h of crowding must then be mechanistically unrelated to the transition to, or expression of, gregarious behaviour. Previous behavioual studies have shown that the resilience of gregarious behaviour to re-isolation increases with time spent in crowded conditions26,32. When solitarious locusts are re-isolated after 24–48 h of crowding, they return to fully solitarious behaviour within 8 h. Long-term gregarious locusts, however, solitarise only partially when isolated for four days as final instar nymphs. Some of these late changes in methylation pattern may therefore represent consolidation mechanisms by which neurochemically mediated rapid changes in behaviour31 become more stable with time. However, differential DNA methylation may also underpin long-term phase differences in the CNS that are not directly responsible for generating phase-specific behaviour but represent adaptations to the respective life styles.

Our present results add to previous evidence for the existence of mechanistically distinct transitional phase states at different levels from the molecular to the behavioural. On a neurotransmitter level, serotonin concentrations in the CNS show a marked transient increase in the thoracic ganglia within the first few hours of crowding that has been causally linked to the transition to gregarious behaviour31. This is followed by an equally transient increase in the brain around 24 h33. On a neuronal level, acute and chronic crowding also have differential effects on serotonin-sythesising neurones, with one set of neurones responding to acute crowding with increased serotonin expression and a distinct set showing decreased serotonin expression in the long term34. Even on a behavioural level, when forming new associations between unfamiliar odours and toxic food, recently gregarised locusts differ from both long-term phases, in a way that matches their respective distinct ecological requirements35,36,37.

In conclusion, our results demonstrate that phase change in desert locusts is associated with distinct short- and longterm shifts in the neural DNA methylation fingerprint. A purely associative study like ours cannot prove causal connections between methylation and behaviour. However, an important consequence of our findings for future studies is that uncovering such causal connections will require analyses of transitional stages rather than only comparisons of endpoints19 because the transitional states are not simply partial realisations of the endpoints. Newly emerging NGS-based methods that are designed specifically to enable epigenome-wide association studies in non-model species (e.g., refs 21 and 24) hold great promise for following up the present results to identify specific changes at the level of identified methylation sites.

Additional Information

How to cite this article: Mallon, E. B. et al. Acute and chronic gregarisation are associated with distinct DNA methylation fingerprints in desert locusts. Sci. Rep. 6, 35608; doi: 10.1038/srep35608 (2016).


  1. Szyf, M., McGowan, P. & Meaney, M. J. The social environment and the epigenome. Environ. Mol. Mutagen. 49, 46–60, 10.1002/em.20357 (2008).

    Article  CAS  PubMed  Google Scholar 

  2. Day, J. J. et al. DNA methylation regulates associative reward learning. Nat. Neurosci. 16, 1445–1452, 10.1038/nn.3504 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Yan, H. et al. Eusocial insects as emerging models for behavioural epigenetics. Nat. Rev. Genet. 15, 677–688, 10.1038/nrg3787 (2014).

    Article  CAS  PubMed  Google Scholar 

  4. Isles, A. R. Neural and behavioral epigenetics; what it is, and what is hype. Genes, Brains Behav. 14, 64–72, 10.1111/gbb.12184 (2015).

    Article  CAS  Google Scholar 

  5. Simpson, S. J., Sword, G. A. & Lo, N. Polyphenism in insects. Curr. Biol. 21, R738–R749, 10.1016/j.cub.2011.06.006 (2011).

    Article  CAS  Google Scholar 

  6. Pener, M. P. & Simpson, S. J. Locust phase polyphenism: An update. Adv. Insect Physiol. 36, 1–272, 10.1016/S0065-2806(08)36001-9 (2009).

    Article  Google Scholar 

  7. Roessingh, P. & Simpson, S. J. The time-course of behavioural phase change in nymphs of the desert locust, Schistocerca gregaria. Physiol. Entomol. 19, 191–197, 10.1111/j.1365-3032.1994.tb01042.x (1994).

    Article  Google Scholar 

  8. Bouaïchi, A., Roessingh, P. & Simpson, S. J. An analysis of the behavioral-effects of crowding and re-isolation on solitary-reared adult desert locusts (Schistocerca gregaria) and their offspring. Physiol. Entomol. 20, 199–208, 10.1111/j.1365-3032.1995.tb00002.x (1995).

    Article  Google Scholar 

  9. Islam, M. S., Roessingh, P., Simpson, S. J. & McCaffery, A. Effects of population density experienced by parents during mating and oviposition on the phase of hatchling desert locusts, Schistocerca gregaria. Proc. Biol. Sci. 257, 93–98, 10.1098/rspb.1994.0099 (1994).

    Article  Google Scholar 

  10. Falckenhayn, C. et al. Characterization of genome methylation patterns in the desert locust Schistocerca gregaria. J. Exp. Biol. 216, 1423–1429, 10.1242/jeb.080754 (2013).

    Article  CAS  PubMed  Google Scholar 

  11. Boerjan, B. et al. Locust phase polyphenism: Does epigenetic precede endocrine regulation? Gen. Comp. Endocr. 173, 120–128, 10.1016/j.ygcen.2011.05.003 (2011).

    Article  CAS  PubMed  Google Scholar 

  12. Kucharski, R., Maleszka, J., Foret, S. & Maleszka, R. Nutritional control of reproductive status in honeybees via DNA methylation. Science 319, 1827–1830, 10.1126/science.1153069 (2008).

    Article  ADS  CAS  PubMed  Google Scholar 

  13. Lyko, F. et al. The honey bee epigenomes: differential methylation of brain DNA in queens and workers. PLoS Biol. 8, e1000506, 10.1371/journal.pbio.1000506 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Herb, B. R. et al. Reversible switching between epigenetic states in honeybee behavioral subcastes. Nat. Neurosci. 15, 1371–1373, 10.1038/nn.3218 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Bird, A. DNA methylation patterns and epigenetic memory. Genes & Dev. 16, 6–21, 10.1101/gad.947102 (2002).

    Article  CAS  Google Scholar 

  16. Feng, J. et al. Dnmt1 and Dnmt3a maintain DNA methylation and regulate synaptic function in adult forebrain neurons. Nat. Neurosci. 13, 423–430, 10.1038/nn.2514 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Robinson, K. L., Tohidi-Esfahani, D., Lo, N., Simpson, S. J. & Sword, G. A. Evidence for widespread genomic methylation in the migratory locust, Locusta migratoria (Orthoptera: Acrididae). PloS ONE 6, e28167, 10.1371/journal.pone.0028167 (2011).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  18. Glastad, K. M., Hunt, B. G. & Goodisman, M. A. D. Evolutionary insights into DNA methylation in insects. Curr. Opin. Insect Sci. 1, 25–30, 10.1016/j.cois.2014.04.001 (2014).

    Article  Google Scholar 

  19. Wang, X. et al. The locust genome provides insight into swarm formation and long-distance flight. Nat. Commun. 5, 2957, 10.1038/ncomms3957 (2014).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  20. Schrey, A. W. et al. Ecological epigenetics: beyond MS-AFLP. Integr. Comp. Biol. 53, 340–350, 10.1093/icb/ict012 (2013).

    Article  PubMed  Google Scholar 

  21. Klughammer, J. et al. Differential DNA methylation analysis without a reference genome. Cell Rep. 13, 2621–2633, 10.1016/j.celrep.2015.11.024 (2016).

    Article  CAS  Google Scholar 

  22. Camacho, J. P. M. et al. A step to the gigantic genome of the desert locust: chromosome sizes and repeated DNAs. Chromosoma 124, 263–275, 10.1007/s00412-014-0499-0 (2015).

    Article  CAS  PubMed  Google Scholar 

  23. Gu, H. et al. Preparation of reduced representation bisulfite sequencing libraries for genome-scale DNA methylation profiling. Nat. Protoc. 6, 468–481, 10.1038/nprot.2010.190 (2011).

    Article  ADS  CAS  Google Scholar 

  24. Schield, D. R. et al. EpiRADseq: scalable analysis of genomewide patterns of methylation using next-generation sequencing. Methods Ecol. Evol. 7, 60–69, 10.1111/2041-210X.12435 (2016).

    Article  Google Scholar 

  25. Song, H. & Wenzel, J. W. Phylogeny of bird-grasshopper subfamily cyrtacanthacridinae (orthoptera: Acrididae) and the evolution of locust phase polyphenism. Cladistics 24, 515–542, 10.1111/j.1096-0031.2007.00190.x (2008).

    Article  Google Scholar 

  26. Roessingh, P., Simpson, S. J. & James, S. Analysis of phase-related changes in behaviour of desert locust nymphs. Proc. Biol. Sci. 252, 43–49, 10.1098/rspb.1993.0044 (1993).

    Article  Google Scholar 

  27. Kronforst, M., Gilley, D., Strassmann, J. & Queller, D. C. DNA methylation is widespread across social Hymenoptera. Curr. Biol. 18, R287–R288, 10.1016/j.cub.2008.02.015 (2008).

    Article  CAS  PubMed  Google Scholar 

  28. Heras, J. et al. GelJ–a tool for analyzing DNA fingerprint gel images. BMC Bioinformatics 16, 270, 10.1186/s12859-015-0703-0 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Pérez-Figueroa, A. Msap: a tool for the statistical analysis of methylation-sensitive amplified polymorphism data. Mol. Ecol. Resour. 13, 522–527, 10.1111/1755-0998.12064 (2013).

    Article  PubMed  Google Scholar 

  30. R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, Vienna, Austria, 2015).

  31. Anstey, M. L., Rogers, S. M., Ott, S. R., Burrows, M. & Simpson, S. J. Serotonin mediates behavioral gregarization underlying swarm formation in desert locusts. Science 323, 627–630, 10.1126/science.1165939 (2009).

    Article  ADS  CAS  PubMed  Google Scholar 

  32. Simpson, S. J., McCaffery, A. R. & Hägele, B. F. A behavioural analysis of phase change in the desert locust. Biol. Rev. 74, 461–480, 10.1111/j.1469-185X.1999.tb00038.x (1999).

    Article  Google Scholar 

  33. Rogers, S. M. et al. Substantial changes in central nervous system neurotransmitters and neuromodulators accompany phase change in the locust. J. Exp. Biol. 207, 3603–3617, 10.1242/jeb.01183 (2004).

    Article  CAS  PubMed  Google Scholar 

  34. Rogers, S. M. & Ott, S. R. Differential activation of serotonergic neurons during short- and long-term gregarization of desert locusts. Proc. Biol. Sci. 282, 20142062, 10.1098/rspb.2014.2062 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  35. Despland, E. & Simpson, S. J. Food choices of solitarious and gregarious locusts reflect cryptic and aposematic antipredator strategies. Anim. Behav. 69, 471–479, 10.1016/j.anbehav.2004.04.018 (2005).

    Article  Google Scholar 

  36. Despland, E. & Simpson, S. J. Surviving the change to warning colouration: density-dependent polyphenism suggests a route for the evolution of aposematism. Chemoecology 15, 69–75, 10.1007/s00049-005-0296-6 (2005).

    Article  Google Scholar 

  37. Simões, P. M. V., Niven, J. E. & Ott, S. R. Phenotypic transformation affects associative learning in the desert locust. Curr. Biol. 23, 2407–2412, 10.1016/j.cub.2013.10.016 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references


Supported by research grants F/09 364/K from the Leverhulme Trust, UK (to S.R.O.), BB/L02389X/1 from the Biotechnology and Biological Sciences Research Council, UK (to S.R.O.) and NE/N010019/1 from the Natural Environment Research Council, UK (to E.B.M.). We thank Alice Round and Carl Breaker for managing our locust facilities.

Author information




S.R.O. and E.B.M. conceived the study and designed the experiments. S.R.O. carried out the animal treatments and dissections. H.E.A. carried out all MS-AFLP bench-work and initial gel analysis and prepared the first draft. E.B.M. and S.R.O. performed statistical analyses. S.R.O. wrote the final draft with input from H.E.A. and E.B.M. All authors gave final approval for publication.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Electronic supplementary material

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Mallon, E., Amarasinghe, H. & Ott, S. Acute and chronic gregarisation are associated with distinct DNA methylation fingerprints in desert locusts. Sci Rep 6, 35608 (2016).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing