Targeted gene suppression by inducing de novo DNA methylation in the gene promoter

Abstract

Background: Targeted gene silencing is an important approach in both drug development and basic research. However, the selection of a potent suppressor has become a significant hurdle to implementing maximal gene inhibition for this approach. We attempted to construct a 'super suppressor' by combining the activities of two suppressors that function through distinct epigenetic mechanisms.

Results: Gene targeting vectors were constructed by fusing a GAL4 DNA-binding domain with a epigenetic suppressor, including CpG DNA methylase Sss1, histone H3 lysine 27 methylase vSET domain, and Kruppel-associated suppression box (KRAB). We found that both Sss1 and KRAB suppressors significantly inhibited the expression of luciferase and copGFP reporter genes. However, the histone H3 lysine 27 methylase vSET did not show significant suppression in this system. Constructs containing both Sss1 and KRAB showed better inhibition than either one alone. In addition, we show that KRAB suppressed gene expression by altering the histone code, but not DNA methylation in the gene promoter. Sss1, on the other hand, not only induced de novo DNA methylation and recruited Heterochromatin Protein 1 (HP1a), but also increased H3K27 and H3K9 methylation in the promoter.

Conclusions: Epigenetic studies can provide useful data for the selection of suppressors in constructing therapeutic vectors for targeted gene silencing.

Keywords: DNA methylation; Epigenetics; Gene expression; Gene suppression; H3K27 methylation; Histone code.

Figure 1

Targeted suppression of the reporter gene by epigenetic suppressors. a. Schematic diagram of suppressor and reporter gene vectors. GAL4: the GAL4 DNA binding domain; GBS: GAL4-binding site; KRAB: kruppel-associated box domain; NSL: nuclear localization signal; pCMV: cytomegalovirus (CMV) promoter; PA: SV40 polyadenylation signal; Sss1: methyltransferase gene from Spiroplasma sp. strain MQ1; vSET: the histone H3 lysine 27 methyltransferase SET domain. Synthetic factors use the GAL4 domain to bind to the GBS site in the target gene vector, where the suppressor domain suppresses the activity of the downstream CMV promoter through epigenetic mechanisms. b. Relative expression of the reporter gene. 293 T cells were transiently co-transfected with 250 ng suppressor vectors, 250 ng luciferase target vector, and 25 ng thymidine kinase promoter-Renilla luciferase reporter (pRL-TK) control vector. The empty pcDNA3.1 vector was used as the study control. Forty-eight hours post-transfection, cells were harvested for luciferase assay. For comparison, the pcDNA3.1 control vector was adjusted to 100%. Each error bar represents the SEM of three independent experiments. a: P <0.05 as compared with the pcDNA3.1 control vector; b: P <0.05 as compared with the Sss1 group.

Figure 2

Suppression of the reporter gene by epigenetic ‘two-hit’ suppressors. a. Schematic diagram of the two-hit suppressor vectors. Two epigenetic suppressor domains are fused with the GAL4 domain. After binding to the target vector, the synthetic factors suppress the target gene using two distinct epigenetic pathways. b. Relative expression of the reporter gene. Forty-eight hours post-transfection, cells were harvested for luciferase assay as described in the Figure 1 legend. Each error bar represents the standard error of mean (SEM) of three independent experiments. a: P <0.05 as compared with the pcDNA3.1 control vector; b: P <0.05 as compared with the Sss1 group.

Figure 3

Epigenetic suppression of the copGFP reporter gene. a. Schematic diagram of the copGFP reporter gene system. b. Inhibition of the transiently-transfected copGFP gene. 293 T cells were transiently co-transfected with copGFP reporter and suppressor vectors. Forty-eight hours post-transfection, copGFP expression was analyzed by luminometer. Each error bar represents the standard error of mean (SEM) of three independent experiments. *Indicates P <0.05 versus pcDNA3.1 control vector. c. Inhibition of the copGFP gene that has been stably integrated in the genome of the target cell. Expression of the copGFP reporter gene was quantitated by real-time quantitative PCR. Each sample was analyzed in quadruplicate. a: P <0.05 as compared with 293 T cells transiently transfected with pcDNA3.1 empty vector; b: P <0.05 as compared with the Sss1 group.

Figure 4

DNA methylation of the CMV promoter. a. The schematic diagram of the CMV promoter and the location of cytosine-phosphate-guanine dinucleotide (CpG) islands. After treatment with sodium bisulfite, genomic DNA was amplified with PCR primer JH1351 and JH1370. Red bar: CpG islands that were sequenced. b. DNA methylation of the CMV promoter using methylation-specific PCR (MSP). Stable clone cells that have the genomically integrated GBS-pCMV-copGFP were transiently transfected with synthetic suppressor vectors. Genomic DNA was extracted and amplified with primers that specifically recognize the methylated CpGs (top panel). Total genomic DNAs were amplified with primers that recognize both unmethylated and methylated CpGs. c. Efficiency of de novo DNA methylation by synthetic suppressors. Stable GBS-pCMV-copGFP clone cells were transiently transfected with 1 μg suppressor vectors. Forty-eight hours post-transfection, cells were harvested for bisulfate sequencing. DNA methylation was calculated as the average percentage of methylated CpGs/(methylated CpGs + unmethylated CpGs) from five CpG islands (110, 122, 141, 165, and 174). *P <0.05 as compared with cells transiently transfected with pcDNA3.1 empty vector.

Figure 5

Promoter histone marks and heterochromatin factor HP1a binding. Chromatin immunoprecipitation (ChIP) assay was performed with anti-trimethyl H3K4 (A), anti-trimethyl H3K9 (B), anti-dimethyl H3K27 (C) and anti-HP1a (D) antibodies. Input DNA of each group was amplified by real-time qPCR as a positive control. Stable copGFP clone cells were transiently transfected with 1 μg blank pcDNA3.1 or suppressor vectors. Forty-eight hours post-transfection, cells were harvested for ChIP assay. a: P <0.05 versus pcDNA3.1 blank vector; b P <0.05 versus KRAB group; and c P <0.05 versus Sss1 group.