Alcohol alters DNA Methylation Patterns and Inhibits Neural Stem Cell

Alcohol alters DNA Methylation Patterns and Inhibits Neural Stem Cell Differentiation Feng C. Zhou1,2, Yokesh Balaraman1,7, MingXiang Teng3,4,5,8, Yunlong Liu3,4,5, Robindra Singh1, and Kenneth P. Nephew6 1Department of Anatomy and Cell Biology, Indiana University School of Medicine, Indianapolis, IN 46202, USA. 2Stark Neuroscience Research Institute, Indiana University School of Medicine, Indianapolis, IN 46202, USA. 3Department of Medical and Molecular Genetics, Indiana University School of Medicine, Indianapolis, IN 46202, USA. 4Center for Computational Biology and Bioinformatics, Indiana University School of Medicine, Indianapolis, IN 46202, USA. 5Center for Medical Genomics, Indianapolis, Indiana University School of Medicine, Indianapolis, IN 46202, USA. 6Medical Sciences Program and Department of Cellular Integrative Physiology, Indiana University School of Medicine, Indianapolis, IN 46202, USA. 7Institute of Psychiatric Research, Indiana University School of Medicine, Indianapolis, IN 46202, USA. 8School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Heilongjiang 150001, PR China. Abstract Background—Potential epigenetic mechanisms underlying fetal alcohol syndrome (FAS) include alcohol-induced alterations of methyl metabolism, resulting in aberrant patterns of DNA methylation and gene expression during development. Having previously demonstrated an essential role for epigenetics in neural stem cell (NSC) development and that inhibiting DNA methylation prevents NSC differentiation, here we investigated the effect of alcohol exposure on genome-wide DNA methylation patterns and NSC differentiation. Methods—NSCs in culture were treated with or without a 6-hr 88mM (“binge-like”) alcohol exposure and examined at 48 hrs, for migration, growth, and genome-wide DNA methylation. The DNA methylation was examined using DNA-methylation immunoprecipitation (MeDIP) followed by microarray analysis. Further validation was performed using Independent Sequenom analysis. Results—NSC differentiated in 24 to 48 hrs with migration, neuronal expression, and morphological transformation. Alcohol exposure retarded the migration, neuronal formation, and growth processes of NSC, similar to treatment with the methylation inhibitor 5-aza-cytidine. When NSC departed from the quiescent state, a genome-wide diversification of DNA methylation was observed—that is, many moderately methylated genes altered methylation levels and became hyper- and hypomethylated. Alcohol prevented many genes from such diversification, including Send all correspondence to: Feng C. Zhou, Ph.D. Indiana University School of Medicine Department of Anatomy & Cell Biology 635 Barnhill Drive, MS 508 Indianapolis, Indiana 46202 Telephone: 317-274-7359 Fax: 317-278-2040 imce100@iupui.edu. NIH Public Access Author Manuscript Alcohol Clin Exp Res. Author manuscript; available in PMC 2012 April 1. Published in final edited form as: Alcohol Clin Exp Res. 2011 April ; 35(4): 735–746. doi:10.1111/j.1530-0277.2010.01391.x. N IH -PA Author M anuscript N IH -PA Author M anuscript N IH -PA Author M anuscript genes related to neural development, neuronal receptors, and olfaction, while retarding differentiation. Validation of specific genes by Sequenom analysis demonstrated that alcohol exposure prevented methylation of specific genes associated with neural development [cutl2 (cut- like 2), Igf1 (insulin-like growth factor 1), Efemp1 (epidermal growth factor-containing fibulin- like extracellular matrix protein 1), and Sox 7 (SRY-box containing gene 7)]; eye development, Lim 2 (lens intrinsic membrane protein 2); the epigenetic mark Smarca2 (SWI/SNF related, matrix associated, actin dependent regulator of chromatin, subfamily a, member 2); and developmental disorder [Dgcr2 (DiGeorge syndrome critical region gene 2)]. Specific sites altered by DNA methylation also correlated with transcription factor binding sites known to be critical for regulating neural development. Conclusion—The data indicate that alcohol prevents normal DNA methylation programming of key neural stem cell genes and retards NSC differentiation. Thus, the role of DNA methylation in FAS warrants further investigation. Keywords Epigenetics; Epigenomics; MeDIP-Chip; Neural development; Fetal alcohol syndrome INTRODUCTION Multifactorial neurodevelopmental deficit, such as fetal alcohol spectrum disorder (FASD), is likely to occur through gene-environment interactions that alter the fate of early developing precursor cells. Neural stem cells (NSC) are capable of self-renewal and differentiate along neuronal and glial lineages. These processes are defined by the dynamic interplay between extracellular cues, and intracellular transcriptional signaling programs. Recently, epigenetic mechanisms—including DNA methylation, histone modifications, and non-coding RNA expression--have been shown to be closely associated with the fate specification of NSCs. These epigenetic alterations could provide coordinated systems for regulating gene expression at each step of neural cell differentiation. Epigenetic modifications regulate key developmental events, including germ cell imprinting (Bartolomei, 2003), stem cell maintenance (Cheng et al., 2005; Kondo, 2006; Zhang et al., 2006; Meshorer, 2007; Surani et al., 2007; Tang and Zhu, 2007), cell fate, and tissue patterning (Kiefer, 2007). Aberrant epigenetic alterations are known to disrupt key developmental events, particuarly in the nervous system, leading to conditions such as Rett's syndrome (Shahbazian and Zoghbi, 2002), immunodeficiency centromeric instability and facial syndrome (ICF) (Hansen et al., 1999, Tao et al., 2002, Ueda et al., 2006), and Prader- Willi/Angelman syndrome (Lalande et al., 1999; Xin et al., 2003; Lalande and Calciano, 2007). Environmental input also has a significant influence on development and can alter epigenetic programming. Recently, we and others have reported that alcohol exposure alters DNA methylation, resulting in genetic and phenotypic changes (Qiang et al., 2010; Ouko et al., 2009; Haycock, 2009; Pandey et al., 2008; Moonat et al., 2010; Shukla et al., 2008; Oberlander et al., 2008; Miranda et al., 2010; Liu et al., 2009; Kaminen-Ahola et al., 2010). Alcohol has multiple effects on methyl donors (Mason and Choi, 2005) and appears to interfere with the folate-methyl metabolic pathway for methyl donors by inhibiting methionine synthase and methionine adenosyltransferase. Inhibition of methionine synthase also creates a “methylfolate trap,” analogous to what occurs in vitamin B12 deficiency (Cravo and Camilo, 2000; Mason and Choi, 2005). Alcohol ingestion in animals has been shown to inhibit folate-mediated methionine synthesis, thereby interrupting critical methylation processes mediated through s-adenosyl methionine, the activated form of methionine and substrate for biologic methylation. In addition, some evidence indicates that alcohol may redirect the utilization of folate toward serine synthesis and thereby interfere Zhou et al. Page 2 Alcohol Clin Exp Res. Author manuscript; available in PMC 2012 April 1. N IH -PA Author M anuscript N IH -PA Author M anuscript N IH -PA Author M anuscript with a critical function of methylene tetrahydrofolate and thymidine synthesis (LaBaume et al., 1987; Cravo and Camilo, 2000). Thus, alcohol exposure, by resulting in downstream epigenetic modifications, could bridge the gene-environment interaction, providing a new causal paradigm for the neurodevelopmental deficit. Alcohol exposure, thus far, has been associated with the development of decreasing DNA methylation transferase (DNMT) mRNA levels in the sperm of male rats exposed to alcohol (Bonsch et al., 2006), and substantial evidence indicates that DNA methylation plays a significant role in neural cell lineage differentiation and early brain development (Takizawa et al., 2001; Setoguchi et al., 2006; Feng et al., 2007; MacDonald and Roskams, 2009). We recently demonstrated that alcohol exposure during early neurulation altered neural tube development and induced changes in DNA methylation, resulting in functional consequences on gene expression, cell cycle regulation, and neural development (Liu et al., 2009). The possibility of alcohol-induced alteration of stem cell DNA methylation changes is not clear, and evidence for alcohol causing aberrant stem cell development through altered DNA or histone methylation has not been examined. In the current study, we used an established dorsal root ganglial neural stem cells, in which multipotency and a differential phenotype profile has been characterized (Singh et al., 2009a). Epigenetic changes at the cellular level have also been demonstrated in this model system (Singh et al., 2009a). By using MeDIP-chip (anti-5-methylcytidine antibody immunoprecipitation followed by microarray analysis) a prominent effect of alcohol on DNA methylation profiles in neural stem cells was discovered. We report here that the normal DNA methylation program during differentiation is disrupted by alcohol, including those genes mediating early embryonic development. RESULTS Alcohol Effects on Neural Stem Cell Differentiation and Cellular DNA methylation As shown in Figure 1, undifferentiated NSCs in neurospheres were characteristically round and small with a distinct nucleus in the center of the small cytoplasm, as reported previously (Singh et al., 2009a). After a two-day culture on substrate-coated chamber slides, the neurospheres were flattened in appearance and cells in the outer layer of the neurosphere increased in size (Fig 1A). On the fourth day in culture (DIC4), noticeable zoning for a differential differentiating state was evident, as reported previously (Singh et al., 2009a). In brief, in the core of the neurosphere, the NSCs remained round and small (Fig 1A). However, in the periphery of the neurosphere, much like the undifferentiated groups, many NSCs remained round, but a subpopulation of the cells formed angular cytoplasmic expansions, some with fiber extensions. In the migrated zone (out of neurosphere), more differentiated cells grew processes, MAP2-immunostained (im), and acquired neuronal morphology with varicosities (Figure 1A); on the other hand, the OCT4-im was reduced in the peripheral area. The alcohol-treated NSC neurospheres, on the fourth day, displayed decreased migration, i.e. migrated cells traveled in smaller scales (Figure 1B, outside the core ) and fewer NSCs left the neurosphere (Figure 1C). The MAP2-im / OCT4-im cell ratio was reduced in comparison with that of Controls in the DIC4. In parallel with analysis of the genome-wide DNA methylation, it was next of interest to examine the effect of alcohol on the cellular distribution of DNA methylation marks, including DNA methylation mark 5-methyl cytidine (5-MeC) and DNA methyltransferase (DNMT). In undifferentiated NSC, 5-MeC immunostaining (im) was moderately intense and distributed throughout the nucleus. However, in the control-undifferentiated group, increased 5-MeC-im intensity was seen in the nucleus, but preferentially distributed near the nuclear membrane into differentiation. A redistribution of 5-MeC-im throughout most of the Zhou et al. Page 3 Alcohol Clin Exp Res. Author manuscript; available in PMC 2012 April 1. N IH -PA Author M anuscript N IH -PA Author M anuscript N IH -PA Author M anuscript nucleus was observed in alcohol-treated NSC during the differentiation stage (Figure 2A-C). A redistribution of DNMT1-im was also evident in the three treatment groups. The DNMT1 was distributed similarly to 5-MeC-im--low and throughout the nucleus--but translocated into cytoplasm in differentiated NSC. Alcohol exposure prevented this redistribution in a subpopulation of the NSC during differentiation (Figure 2D-E). The significance of these translocations may be pertinent to euchromatic vs heterochromatic distribution of epigenetic marks associated genes. A semi-quantitative measurement of staining indicated that 5-MeC and DNMT1 were dynamically changed from the undifferentiated control neurosphere to the differentiated Control and Alcohol groups in core and periphery of neurospheres and in migrated (Figure 2G). The distribution of multiple DNA methylation profiles changed during differentiation by presenting an increased diversity with more hypermethylated and hypomethylated genes, and a modified landscape of gene methylation. During the undifferentiated/renewal state, the vast majority of genes (93%) were moderately methylated (between 0.4 and -0.4 log mean methylation, see definition in Method) (Figure 3A). During differentiation, a total of 1143 genes changed the methylation status, and, out of those, alcohol prevented the change of 272 genes (p <0.05). Among the 344 genes that were changed from moderate methylation during differentiation, 222 became hypermethylated and 122 hypomethylated. When the DRG stem cells were treated with alcohol during differentiation, 97 genes out of those 344 changes were prevented or reversed; 78 genes were prevented from hypermethylation, and 19 genes changed from hypomethylation back to moderate methylation status (Figure 3B). The genome-wide MvA plot displayed difference of DNA methylation levels between two groups of all the genes according to their average DNA methylation levels e.g. control- undifferentiated vs. control-differentiated, control-differentiated vs. alcohol-differentiated. There was strong diversification of methylation during differentiation (comparing the undifferentiated with the differentiated). Such differentiation-related diversification was greatly reduced when treated with alcohol (comparing the undifferentiated with the differentiated + alcohol) (Figure 4), indicating alcohol prevented the diversification of DNA methylation Functional profile of Hyper- and Hypomethylated genes altered by Alcohol— The 78 genes that were prevented or reversed from hypermethylation by alcohol from moderate methylation category (Supplement Table 1) are involved in multiple key functions. The most noticeable group of genes are related to neuronal receptors, including glutamate transmitter receptors AMPA3 (Gria3) and glutamate receptor interacting protein 1 (Grip1), cholinergic receptor muscarinic 1 (Chrm1), Adrenergic receptor a1 (Adra1a), and water channels proteins Aqp8 and Aqp9. Other genes include neural development (Dlx3, Clcf1), Cell Cycle (Adra1a, Tnf, Pik3r1, Sh3bp2), path finding of retinal ganglion cells and extension of axons (Pou4f2, Pou4f3), synaptic transmission (Chrm1, Gria3, Tnf, Kcna1, Ptprc), and stem cell (Edg6, Pik3r1, Wnt16). An alcohol metabolism enzyme, alcohol dehydrogenase 4 (Adh4), was also affected by alcohol. Of the 19 genes in moderate methylation category prevented from hypomethylation by alcohol, many were noticeably associated with a group of sensory genes, including the olfactory receptor (Olr214, Olr304, Olr408, Olr611, Olr1646, Olr1139), and the taste receptor genes (Tas2r123, Tas2r7, T2r140, T2r18) (Supplement Table 2). Two-consecutive probe analysis—Since the effect of DNA methylation can be localized (e.g. within 200bp), probe-wise analyses were performed. We adopted a two- consecutive probe analysis (in which two adjacent probes detected the same direction of Zhou et al. Page 4 Alcohol Clin Exp Res. Author manuscript; available in PMC 2012 April 1. N IH -PA Author M anuscript N IH -PA Author M anuscript N IH -PA Author M anuscript DNA methylation changes), in parallel with the average DNA methylation analysis, to verify regional changes in DNA methylation. Based on the linear mixed model, 1054 (FDR <30%) two-consecutive probe regions showed a significant difference in DNA methylation between the control and alcohol-treated NSCs during differentiation. Among those genes, alcohol decreased methylation of 681 and increased methylation of 373. These regional changes were also used as the basis for validating regional DNA methylation changes of specific genes using Sequenom. Genes validated by Sequenom A set of genes with promoter DNA methylation changes during differentiation as well as by alcohol, i.e. genes altered during differentiation and affected by alcohol (changed both in control differentiated vs. control undifferentiated, and in control differentiated vs. alcohol differentiated) and related to the nervous system, were validated using Sequenom MASSarray technology. Overall, based on the two criteria in the Methods, we validated 12 genes out of 26 tested from the MeDIP-microarray results. Their methylation alteration site is shown in Table 1. Examples of comparison of MeDIP-ChIP and Sequenom is shown in Figure 5. DISCUSSION Mammalian development is associated with considerable changes in global DNA methylation levels during genomic reprogramming. The current genome-wide methylation study, showing widespread DNA methylation changes in specific genes of NSCs, supports our previous observation that DNA methylation is an active component and an intrinsic program during differentiation of NSCs (Singh et al., 2009a). In particular, we observed hyper- and hypo-methylation of moderately methylated genes during the reprogramming of quiescent NSCs into differentiation. There are two major questions addressed in this study: first, whether altered DNA methylation in the process of neural stem cell differentiation can be modified by the environmental input, i.e. whether alcohol, which is known to change methyl donor, can alter the intrinsic DNA methylation program during differentiation. Second, whether the altered DNA methylation (by alcohol) has a functional consequence on neural stem cell development. Two categories of alcohol effect are reported here. First, alcohol prevented programmed diversification of DNA methylation during differentiation, i.e. it prevented moderately methylated genes from increased or decreased methylation, or further, becoming significantly hyper- or hypo-methylated. As promoter DNA methylation represses transcription and recruits other repressive chromatin-modifying activities to the chromatin (Bird, 2002), the prevention of programmed hyper- and hypo-methylation of a set number of genes can disrupt chromatin remodeling, and, thus, differentiation reprogramming of quiescent NSCs might be incomplete. The second effect of alcohol is the methylation change of the genes that were not supposed to alter in the normal program at this stage of normal differentiation. The untimely methylation change of these genes also have an opportunity to affect the gene transcription and deviate differentiation. A number of genes verified by Sequenom MASSarray are known to closely participate in the neural stem cell differentiation program (Table 1). The Igf2 (insulin growth factor 2, an imprinting gene key in development) and Sox 7 (an activator of fibroblast growth factor 3 transcription) are involved in neural stem cell growth and patterning. The Lim2 and Cutl2 (Cux2, Cux2 cut-like homeobox 2), a homeodomain transcription factor, regulates neural stem cell proliferation and differentiation in the developing cortical ventricular zone and in the spinal cord (Cubelos et al., 2008,Iulianella et al., 2008). Cutl2 loss-of-function mouse mutants exhibit smaller spinal cords with deficits in neural progenitor development (Iulianella et al., 2008). Smarca2 (Brm; SWI/SNF related, matrix associated, actin dependent regulator of chromatin) with Brg1 are subunits of SWI/SNF complex essential for Zhou et al. Page 5 Alcohol Clin Exp Res. Author manuscript; available in PMC 2012 April 1. N IH -PA Author M anuscript N IH -PA Author M anuscript N IH -PA Author M anuscript the transition from neural stem/progenitors to postmitotic neurons (Lessard et al., 2007). The function of DNA methylation may regulate the recruitment of histone modification enzymes (e.g. histone deactylase