Introduction

Epigenetic modifications of chromatin play essential roles in the regulation of gene expression (1,2,3). One key modification is DNA methylation, which is typically associated with stable gene silencing in mammals. Over the past 30 years, a myriad of cellular processes have been examined that might be under the regulation of DNA methylation. The best-studied examples include genomic imprinting, X chromosome inactivation, and silencing of repetitive elements, all associated with hypermethylation at silenced loci. It is increasingly recognized that DNA methylation is also important for the silencing of genes during normal cellular differentiation and during cancer progression (4,5,6,7,8). As such, there is growing interest in identifying genes whose silencing is associated with DNA methylation. To this end, researchers have devised several approaches, such as interfering with DNA methyltransferase activity by conditional knockout or RNA interference (RNAi)-induced knockdown of the DNA methyltransferase I (Dnmt1) gene, and using drugs that inhibit DNA methylation. However, technical issues persist with these methods, including the lack of global demethylation, possibly due to the presence of other DNA methyltransferases, heterogeneity of demethylated cells in RNAi-induced Dnmt1 knockdown, and incomplete incorporation and toxicity of the methylation-inhibiting drugs (9,10,11,12).

One common approach to study DNA methylation is to treat cells with the demethylating drug 5-aza-2′-deoxycytidine (5-aza-2dC), a derivative of deoxycytidine, and identify genes activated by this treatment. 5-aza-2dC is incorporated into replicating DNA where it irreversibly binds to DNMT1, thus depleting maintenance methyltransferase activity in the cell. This results in global demethylation or hypomethylation. A problem with 5-aza-2dC treatment is that cells in culture do not respond equally to the drug. Given that only actively dividing cells incorporate 5-aza-2dC, nondi-viding or slow-dividing cells are not affected by the drug. Furthermore, cells in different stages of the cell cycle during drug treatment may undergo varying amounts of demethylation, and other factors, such as the rates of drug uptake and catabolism, may also differ among cells. Thus, when cultured cells are treated with 5-aza-2dC, the result is a heterogeneous population of cells with differing levels of demethylation.

Here, we report an approach that uses the reactivation of an X-inactivated green fluorescent protein (GFP) transgene in mice to enrich for cells that have undergone drug-induced demethylation. We show that this approach, when combined with microarray analysis, enhances the ability to identify genes activated by 5-aza-2dC treatment. This technique is widely applicable to dividing cells derived from the available X-linked GFP transgenic mouse line, including fibroblasts, osteoblasts, chondrocytes, myoblasts, hepatocytes, and many others. Additionally, the basic strategy can be extended to studies involving other chromatin-modifying drugs such as those that influence histones.

Materials and Methods

Cell Culture and Drug Treatment

Mice used have an X-linked GFP transgene driven by the chicken ß-actin promoter and cytomegalovirus (CMV) intermediate early enhancer in a 129/ICR mixed genetic background obtained from The Jackson Laboratory (Bar Harbor, ME, USA) (13). Male transgenic animals were bred to wild-type BL6 female mice to obtain females that were heterozygous for the transgene. Dermal fibroblasts were obtained from explant cultures of 1- to 3-day neonatal female mice from these crosses, using procedures as previously described (14). Cultures were maintained in Dulbecco's modified Eagle's medium (DMEM; Invitrogen; Carlsbad, CA, USA) with L-glutamine and high glucose plus 10% fetal bovine serum (FBS; Invitrogen) and 100 U/mL penicillin/streptomycin (Invitrogen) at 37°C and 5% CO₂. Cells were passaged at confluence by 0.25% trypsin EDTA (Invitrogen). To obtain sufficient cell number and fibroblast purity, cells were cultured for 12 days. Freshly prepared 5-aza-2dC (Sigma, St. Louis, MO, USA) stock solution was diluted in growth medium to concentrations of 1–10 M. A new preparation of the drug in growth medium was added after 24 h for experiments with a 48-h drug exposure.

Cell Sorting by Flow Cytometry

MoFloHTS™ (DakoCytomation, Glostrup, Denmark) was used for fluorescence-activated cell sorting (FACS) by the University of Chicago Immunology Application Core Facility, Chicago, IL. Cells were sorted in growth medium supplemented with 1 mM EDTA to prevent cell clumping. BD FACScan™ (BD Biosciences, San Jose, CA, USA) was used for cytometry with CellQuest(tm) software (BD Biosciences) for analysis.