We used the well-established G. mellonella model system [6–14] to show that epigenetic reprogramming during insect metamorphosis, wound healing and infection is controlled by histone acetylation/deacetylation, which in turn is regulated by HATs and HDACs with opposing activities. We found that several genes encoding components of HATs and HDACs were upregulated during the transformation of prepupae into pupae (Figure 1). This concurs with previous reports discussing extensive transcriptional reprogramming at the onset of metamorphosis to induce genes involved in tissue and organ remodeling. For example, HAT activity is mediated by the histone H3 acetylase dGN5, which is a key modulator of chromatin structure and transcription during Drosophila melanogaster metamorphosis .
To provide experimental evidence for the epigenetic role of histone acetylation in G. mellonella metamorphosis, we postulated that imbalanced HDAC and HAT activities would either accelerate or postpone development. We therefore injected specific HDAC and HAT inhibitors into last-instar larvae and found that the HDAC inhibitors reduced histone deacetylation following precocious metamorphosis, whereas the HAT inhibitors postponed the formation of pupae (Figure 2A). This, to the best of our knowledge, is the first time that the inhibition of epigenetic regulators with opposing functions has been shown to induce correspondingly opposed developmental effects. Inspired by these results we sought to explore whether histone acetylation also regulates transcriptional reprogramming in response to wounding.
We pierced last-instar G. mellonella larvae with a needle to cause injuries accompanied by massive hemolymph loss and tissue damage that was ultimately lethal to the majority of the wounded larvae. We found that the injection of HDAC or HAT inhibitors prior to wounding resulted in opposite effects on mortality compared to control larvae injected with 1% DMSO (Figure 3). These data provide evidence that histone acetylation/deacetylation is also involved in wounding-mediated transcriptional reprogramming. The positive effect of HDAC inhibitors on G. mellonella survival after wounding agrees closely with the recently published results of a similar study using the mouse model, in which the HDAC inhibitor SAHA was shown to increase survival in mice suffering from septic shock without fluid resuscitation .
The contribution of histone acetylation/deacetylation to wounding-mediated transcriptional reprogramming was investigated in more detail by measuring the expression of genes encoding components of HDACs and HATs identified in our previous transcriptomic analysis of developmental and immunological gene repertoires in G. mellonella. Unfortunately, we have not found genes encoding further components of the histone acetylation/deacetylation complex such as SIN3A or SIR2.
As expected, wounding induced the expression of HDACs more than HATs (Figures 4 and 5) indicating that skewing the ratio of HDAC:HAT activity in favor of hypoacetylation ultimately increases the lethality associated with wounding. This is further supported by the observation that the injection of HDAC inhibitors prior to injury results in the accelerated and enhanced expression of both HDACs and HATs (Figures 4 and 5) and thus we hypothesize that larvae with a more balanced HDAC:HAT ratio were more likely to survive. We also studied the effect of HDAC inhibitors on the expression of HDACs using an independent method based on the photometric measurement of enzyme activity. We found that HDAC inhibitors attenuate HDAC level, thus reducing HDAC enzyme activity (Figure 6). Therefore, three lines of evidence indicate a role for histone acetylation in wounding-mediated reprogramming: the opposing effects of HAT and HDAC inhibitors on the survival of wounded larvae, the transcriptional activity of genes encoding HATs and HDACs, and the direct measurement of HDAC activity in the hemolymph.
Because the injection of HDAC inhibitors into last-instar larvae brings forward the onset of pupation, we postulated that the same treatment should enhance and/or bring forward the expression of genes contributing to both tissue remodeling during metamorphosis and wound healing. To test this hypothesis, we measured the expression of the first MMP discovered in Lepidoptera, since this is known to have pleiotropic roles in both development and immunity [18, 19]. Confirming our expectations, we found that the MMP gene was induced to higher levels following wounding in larvae treated with HDAC inhibitors (Figure 7A). The insect metalloproteinase inhibitor (IMPI), the only animal peptide that specifically inhibits microbial metalloproteases causing sepsis [16, 17], was induced 1 h post-injury, but the induction was earlier and stronger in larvae injected with HDAC inhibitors prior to injury (Figure 7B). This may explain the protective effect of HDAC inhibitors following injury accompanied by severe hemolymph loss. The observation that HDAC inhibitors suppress the expression of the antimicrobial peptide galiomycin in G. mellonella (Figure 7D) agrees with a recent study providing evidence that HDAC inhibitors impair innate immune responses to Toll-like receptor agonists and to infection in mice .
The observation that HAT and HDAC inhibitors have opposite effects on development and survival after wounding suggested that histone acetylation/deacetylation may also regulate transcriptional reprogramming during pathogen infections. It was recently reported that G. mellonella is an ideal model host for L. monocytogenes, a major food-borne human pathogen responsible for listeriosis, which in its severest form can cause fatal meningitis, meningoencephalitis and septicemia [22, 23]. The injection of G. mellonella larvae with bacterial LPS significantly enhanced survival rates upon subsequent infection with a nominally lethal dose of L. monocytogenes by inducing the expression of antimicrobial peptides that have now been characterized . We found that interactions between G. mellonella and virulent strains of L. monocytogenes caused a delay in the transformation of infected larvae into pupae (Figure 9). We therefore postulated that at least this pathogen can interfere with histone acetylation in the host because infection causes the same developmental delay that we observed when larvae were injected with HAT inhibitors (Figure 2). These findings agree with recent experiments showing that L. monocytogenes can also delay the induction of host cellular immunity in mammalian hosts to prolong the infection cycle . Furthermore, immune stimulation also accelerates the development of the red flour beetle Tribolium castaneum following a challenge with heat-killed bacteria, although links to epigenetic gene regulation have not been explored . Interestingly, we have also observed accelerated pupation in G. mellonella larvae following treatment with heat-killed L. monocytogenes (data not shown).
The ability of virulent L. monocytogenes and M. anisopliae strains to interfere with epigenetic gene regulation and exert codependent depressive effects on immunity and development led to a further hypothesis that non-virulent microbes lacking the ability to suppress the immune system might have a collateral accelerating effect on development. We therefore administered G. mellonella larvae with non-pathogenic E. coli or non-pathogenic M. anisopliae strain. As postulated, these infections had the opposite effect to infections with virulent pathogens, not only failing to delay metamorphosis but actually accelerating the process (Figure 8), as observed when larvae were injected with HDAC inhibitors (Figure 2).
The hemolysin known as listeriolysin O (LLO) has previously been shown to inhibit histone acetylation and phosphorylation in human epithelial cells infected with L. monocytogenes[26, 27], whereas prolonged exposure to LPS and infection with pathogenic Mycobacterium tuberculosis strains causes the transcriptional induction of multiple HDAC genes in mammals . Bacteria have therefore been shown to induce chromatin remodeling during infection, although their impact on host immunity and development is unclear. Our data provide the first conclusive evidence that pathogens can interfere with the epigenetic regulation of transcription to influence both immunity and development in insects. We found that infection with non-pathogenic E. coli induced HDAC gene expression after 1 h but that the levels were attenuated after 4 d, whereas infection with virulent L. monocytogenes resulted in the delayed but longer-lasting induction of HDACs, particularly HDAC8 (Figure 9). In contrast, HAT expression levels were not significantly affected at the later time point, indicating that the opposing developmental effects caused by virulent and non-virulent bacteria are predominantly mediated through the regulation of HDAC activity. Similarly, infection with entomopathogenic M. anisopliae strain 43 and 97 resulted in early mortality and late pupation than strain 79. We have recorded early induction of HDACs (2 and 4 days after infection) in larvae infected with M. anisopliae strain 43 and 97 but not with the strain 79. The skewing HDAC/HAT ratio following M. anisopliae strain 79 infection was apparently distinct only 9 days after infection, indicating the differences in the pathogenic potential of fungal strains in causing mortal infections. The proposed role of HDACs as a regulatory layer of immunity-related signaling in insects may be evolutionarily conserved  because our data are consistent with recent findings showing that the mouse metastatic tumor antigen 1 (MTA1) coregulator, a component of the nucleosome remodeling and deacetylase (NuRD) complex, modulates the NF-κB signaling network to maintain homeostasis during inflammatory responses .