Choline deficiency alters global histone methylation and epigenetic marking at the Re1 site of the calbindin 1 gene - PubMed (original) (raw)

Choline deficiency alters global histone methylation and epigenetic marking at the Re1 site of the calbindin 1 gene

Mihai G Mehedint et al. FASEB J. 2010 Jan.

Abstract

Maternal choline availability is essential for fetal neurogenesis. Choline deprivation (CD) causes hypomethylation of specific CpG islands in genes controlling cell cycling in fetal hippocampus. We now report that, in C57BL/6 mice, CD during gestational days 12-17 also altered methylation of the histone H3 in E17 fetal hippocampi. In the ventricular and subventricular zones, monomethyl-lysine 9 of H3 (H3K9me1) was decreased by 25% (P<0.01), and in the pyramidal layer, dimethyl-lysine 9 of H3 (H3K9me2) was decreased by 37% (P<0.05). These changes were region specific and were not observed in whole-brain preparations. Also, the same effects of CD on H3 methylation were observed in E14 neural progenitor cells (NPCs) in culture. Changes in G9a histone methyltransferase might mediate altered H3K9me2,1. Gene expression of G9a was decreased by 80% in CD NPCs (P<0.001). In CD, H3 was hypomethylated upstream of the RE1 binding site in the calbindin 1 promoter, and 1 CpG site within the calbindin1 promoter was hypermethylated. REST binding to RE1 (recruits G9a) was decreased by 45% (P<0.01) in CD. These changes resulted in increased expression of calbindin 1 in CD (260%; P<0.05). Thus, CD modulates histone methylation in NPCs, and this could underlie the observed changes in neurogenesis.

PubMed Disclaimer

Figures

Figure 1.

CD decreases H3K9me1 and H3K9me2 levels in specific areas of the fetal hippocampus. Pregnant mice were fed a CD diet or a CT diet from E12-E17. E17 fetal brains were used to assess the relative levels of H3K9me1 and H3K9me2 in the hippocampus and whole brains (_n_=7 independent samples/treatment; immunofluorescent labeling and Western blotting, respectively). A) Representative fluorescent images for both treatments (CD and CT); red staining indicates H3K9me1. Insets: nuclear staining. B) H3K9me1 levels are decreased in CD hippocampi as compared with CT (1.56±0.06 vs. 2.09±0.15), specifically in VZ and SVZ of hippocampus, which encompass the main proliferation area. CA, cornus ammoni; DG, dentate gyrus. C) Representative fluorescent images for both treatments; red staining indicates H3K9me2. Insets: nuclear staining. D) H3K9me2 levels were also decreased in pyramidal layer of CA (CD 1.03±0.15 vs. CT 1.61±0.27). Values are given as relative MOD levels (see Materials and Methods). Gray bars, CD; black bars, CT. Error bars =

. *P < 0.05, **P < 0.01; Student’s t test.

Figure 2.

CD changes the proliferation and apoptosis index in E14 mouse neural progenitors. Mouse neural progenitors were exposed to either a CD medium (5 μM choline, CD) or a control medium (70 μm choline, CT) for 0, 24, 48, 72, 96, and 120 h. TUNEL and BrdU immunolabeling was performed as described in Materials and Methods. Number of labeled nuclei was reported as percentage from total number of cells counted (_n_=5 independent samples/treatment). A) CD increases number of TUNEL-positive nuclei at 96 h (CD 29.12±1.57 vs. CT 12.55±0.59%) and 120 h (CD 38.33±0.94 vs. CT 14.74±2.75%) but not at 72 h. B) CD decreases proliferation at 48 h (CD 10.16±2.08 vs. CT 18.36±1.37%) and 72 h (CD 2.68±1.64 vs. CT 22.9±2.19%). Gray bars, CD; black bars, CT. Error bars =

se.

*P < 0.05, **P < 0.01; Student’s t test.

Figure 3.

CD decreases H3K9me1 and H3K9me2 but has no influence on H3K9me3, H3K9Ac, and H3K4me2 levels in the E14 NPC histone extracts. Mouse neural progenitors were exposed to either a CD medium (5 μM choline, CD) or a control medium (70 μm choline, CT) for 72 h. Immunolabeling was performed as described in Materials and Methods, and OD was measured (_n_=5 independent samples/treatment). A) Representative Western blot images of H3K9me1, H3K9me2, H3K9me3 (CD 0.81±0.11 vs. CT 0.77±0.043), H3K9Ac (CD 0.83±0.09 vs. CT 0.76±0.1), and H3K4me2 (CD 0.98±0.11 vs. CT 1.08±0.09). B) Values for each sample were normalized to the OD of total histone H3. CD decreased the H3K9me1 and H3K9me2 protein levels (H3K9me1: CD 0.59±0.19 vs. CT 1.43±0.15; H3K9me2: CD 0.71±0.08 vs. CT 1.42±0.17). Values are given as relative MOD levels (see Materieals and Methods). Gray bars, CD; black bars, CT. Error bars =

se.

*P < 0.05, **P < 0.01; Student’s t test.

Figure 4.

Regulation of chromatin compaction by histone methylation and deacetylation via REST mechanism. Chromatin repression is achieved following REST binding to the RE1 site, by recruiting histone modifying enzymes (G9a, SUV39, and HDAC). First HDAC performs deacetylation of H3K9 residues. After this event, G9a utilizes _S_-adenosylmethionine (SAM) as the methyl donor and methylates the available residue, decreasing chromatin availability. In a separate set of events, G9a and DNMT1 (which maintains the DNA methylation patterns established during development) are recruited by PCNA at the replication fork during DNA synthesis. They transmit the methylation status of both histones and DNA to the newly formed nucleosomes.

Figure 5.

CD alters G9a and Calb1 gene expression and protein levels but not SUV39. After 72 h exposure to a CD medium (CD) or a control medium (CT), gene expression of G9a histone methylase (in NPCs) and Calb1 (in NPCs and whole brain) was determined using quantitative real time RT-PCR (see Materials and Methods). A, B) No changes were detected in the SUV39 transcripts level in NPCs. G9a protein levels (A) were decreased by CD 0.53 ± 0.05 vs. CT 0.9 ± 0.04 in NPCs (B); increased immunoreactive bands of Calb1 (A) were detected under CD 1.18 ± 0.05 vs. CT 0.75 ± 0.04 in NPCs (B). Negative control blots were probed with reimmune serum (A). C) CD decreased G9a expression (CD/CT ratio 0.22±0.02, P<0.001; _n_=4/group), while Calb1 expression was increased (CD/CT ratio 2.62±0.05in NPCs and 2.84±0.8 in brain). Gray bars, CD; black bars, CT. Error bars =

. *P < 0.05, **P < 0.01; randomization test.

Figure 6.

CD changes the DNA-histone 3 interaction at the RE1 site of Calb1. After 72 h exposure to a CD medium (CD) or a control medium (CT), ChIP was used to assess changes in the levels of H3K9 and H3K4 methylation at the RE1 locus (B) and abundance of REST at the Calb1 site (C; Materials and Methods). A) Schematic representation of the RE1 site of Calb1 gene. RE1/NRSE (neuron-restrictive silencer element) site is within exon 1 of Calb1 and is in proximity to the transcription starting site. REST/NRSF (neuron-restrictive silencer factor) and coREST repressive complex (G9a, G9a HMTase; LSD1, lysine-specific demethylase 1) interacts with the RE1/NRSE site [adapted from Ooi et al. (49)]. B) CD decreased H3K9me1 and H3K9me2 levels upstream of RE1 in Calb1 decreased. Conversely, H3K4me2 was increased in proximity to the Re1 site. Inset indicates no amplification for negative controls (negative IgG and negative primer set) by qRT-PCR. *P < 0.05, **P < 0.01 vs. CT; randomization test. C) CD decreased the binding of REST to the RE1 site of Calb1 (CD 0.2±0.02 vs. CT 0.36±0.01, normalized to input DNA, P<0.01 by t test). Gray bars, CD; black bars, CT. Error bars =

. ipDNA, immunorecipitated DNA against REST antibody.

Figure 7.

CD induces hypermethylation of a specific CpG site within Calb1 gene. DNA pyrosequencing after bisulfite treatment indicates that CD induces specific hypermethylation of the CpG site 3 within the CpG island of Calb1 promoter (CD 16.9±0.2 vs. CT 14.2±0.2%). CpG sites 7–9 are located within the RE1 binding site (RE1 sequence underscored in inset). Gray bars, CD; black bars, CT. Error bars =

. *P < 0.00555. Each CpG site is marked with bold letters. Arrow indicates transcription start site for Calb1.

References

1. Fisher M C, Zeisel S H, Mar M H, Sadler T W. Inhibitors of choline uptake and metabolism cause developmental abnormalities in neurulating mouse embryos. Teratology. 2001;64:114–122. - PubMed
1. Fisher M C, Zeisel S H, Mar M H, Sadler T W. Perturbations in choline metabolism cause neural tube defects in mouse embryos in vitro. FASEB J. 2002;16:619–621. - PubMed
1. Blom H J, Shaw G M, den Heijer M, Finnell R H. Neural tube defects and folate: case far from closed. Nat Rev Neurosci. 2006;7:724–731. - PMC - PubMed
1. Shaw G M, Carmichael S L, Yang W, Selvin S, Schaffer D M. Periconceptional dietary intake of choline and betaine and neural tube defects in offspring. Am J Epidemiol. 2004;160:102–109. - PubMed
1. Craciunescu C N, Albright C D, Mar M H, Song J, Zeisel S H. Choline availability during embryonic development alters progenitor cell mitosis in developing mouse hippocampus. J Nutr. 2003;133:3614–3618. - PMC - PubMed

Choline deficiency alters global histone methylation and epigenetic marking at the Re1 site of the calbindin 1 gene - PubMed (original) (raw)