Fasting alters histone methylation in paraventricular nucleus of chick through regulating of polycomb repressive complex 2

The developing brain is highly sensitive to environmental influences. Unfavorable nutrition is one kind of stress that can cause acute metabolic disorders during the neonatal period [1,2,3] and severe diseases in later life [4,5]. These early life experiences occurring during heightened periods of brain plasticity help determine the lifelong structural and functional aspects of brain and behavior. In humans, for example, weight gain during the first week of life increased the propensity for developing obesity several decades later [5]. This susceptibility is, if not all, related to the dynamic reversible epigenetic imprints left on the histones [6,7,8], especially during the prenatal and postpartum period [9].

Histones are highly dynamic and responsive towards environmental stress [10,11]. Through covalent modification of the histone tail, histones are able to direct DNA scaffolding and regulate gene expression [10,12]. Thus far, various types of post translational modifications have been identified on various histones tails [12]. Among them, the methylation and acetylation on lysine residue (K) 27 on histone 3 (H3) has been tightly linked to gene repression [13,14] and activation [15], respectively. EZh2 (enhancer of zeste 2) in the polycomb repressive complex 2 (PRC2) is the only methyltransferase that has been linked to catalyze this methylation reaction. In addition, SUZ (suppressor of zeste) and EED (embryonic ectoderm development) are two other key proteins in PRC2 function core that help EZH2. As previous reported, increased H3K27 methylation was monitored after fasting stress during neonatal period in chicks' paraventricular nucleus (PVN). In this study, we investigated the detailed mechanism behind changes in H3K27 methylation following fasting stress.

After 24 hours fasting on 3 days-of-age (D3), chicks exhibited elevated mRNA levels of PRC2 key components, including EZH2, SUZ and EED, in the PVN on D4. Western blots confirmed this finding by showing increased global methylation status at the H3K27 site in the PVN on D4. In addition, until 38 days post fasting, SUZ and EZH2 remained inhibited. A newly identified anorexigenic factor, Brain-derived neurotrophic factor (BDNF), was used as an example of multiple hormones expressed in PVN to verify this finding. Both BDNF protein and mRNA exhibited compatible changes to global changes of tri- (me3) and di-methylated (me2) H327. Furthermore, by using chromatin immunoprecipitation assays (ChIP), we were able to monitor the changes of H3K27me2/me3 deposition along the Bdnf gene. Fasting significantly increased H3K27me2/me3 as well as EZH2 at the Bdnf's promoter, transcription start site and 3'-untranslated region. These data show that fasting stress during the early life period could leave epigenetic imprinting in PVN for a long time. Next, we tried to understand the function of this epigenetic imprinting in the chicks' PVN. Thus, we compared naive chicks (never fasted) to chicks that received either a single 24 hour fast on D3 or two 24 hour fast on both D3 and 10 days-of-age (D10). We found that the D3 fasted group significantly increased the level of PRC2 key components and its product H3K27me2/me3 compared to the naive group. However, D3 fasting and D10 fasting together decreased the surges of H3K27me2/me3, SUZ and EED (not EZH2) compared to the naive group. We called this phenomenon "epigenetic memory". The Western blot, qPCR and CHIP assay results from BDNF all confirmed the existence of "epigenetic memory" for PRC2. These data suggested that fasting stress during the early period of brain development could leave long term epigenetic modifications in neurons. These changes could be beneficial to the body, which keeps homeostasis of inner environment and prevent massive response to future same stress.

The EZH2 protein was knocked down and the H3K27 methylation status changes were monitored after applying the same treatment. We first confirmed that EZH2 antisense oligonucleotides (5.5 ug), but not EZH2 siRNA and artificial cerebrospinal fluid (ACSF), inhibit EZH2 protein by 86 % in the PVN. Then, on D3, chicks were subjected to a 24 hour fasting stress (D3-fasting) post either EZH2 antisense or ACSF injection. The EZH2 antisense blocked the surge of both EZH2 mRNA and H3K27 methylation after D3-fasting. At the same time, BDNF exhibited elevated expression levels and less methylated H3K27 deposition along the Bdnf gene. In addition, we were also interested in the changes of "epigenetic memory" post EZH2 antisense injection. We found that after EZH2 antisense injection, chicks' PVN no longer exhibited any "epigenetic memory" to repetitive fasting stress. While EZH2 mRNA was constantly inhibited, SUZ, EED and H3K27me2/3 levels were unpredictable. These findings suggested that neurons in the PVN utilized PRC2 as a major H3K27 methylation tool. Knockdown of EZH2 in the PRC2 impaired the proper response in PVN to fasting stress and PVN's ability to acclimate to repetitive fasting stresses. Thus, EZH2 is an important H3K27 methyltransferase inside chicken hypothalamus to maintain homeostasis.

In conclusion, fasting stress during the early life period could leave epigenetic markers on chromosomes of neurons in the feeding regulation center. These epigenetic markers will be left on chromosomes for a long period of time and have a beneficial role in keeping homeostasis when individuals face future fasting stress again. H3K27 methylation is one of these epigenetic markers and inhibits expression of various genes inside neurons. EZH2 is so far the only detected methyltransferases for H3K27 that form the PRC2. Thus EZH2 plays a key function in the body's response to fasting.