From silencing to gene expression: real-time analysis in single cells

Abstract

We have developed an inducible system to visualize gene expression at the levels of DNA, RNA and protein in living cells. The system is composed of a 200 copy transgene array integrated into a euchromatic region of chromosome 1 in human U2OS cells. The condensed array is heterochromatic as it is associated with HP1, histone H3 methylated at lysine 9, and several histone methyltransferases. Upon transcriptional induction, HP1alpha is depleted from the locus and the histone variant H3.3 is deposited suggesting that histone exchange is a mechanism through which heterochromatin is transformed into a transcriptionally active state. RNA levels at the transcription site increase immediately after the induction of transcription and the rate of synthesis slows over time. Using this system, we are able to correlate changes in chromatin structure with the progression of transcriptional activation allowing us to obtain a real-time integrative view of gene expression.

Figure 1. Characterization of Human U2OS 2-6-3 Cells

(A) Schematic representation of the gene expression plasmid, p3216PECMS2β. The plasmid is composed of 256 copies of the lac operator, 96 tetracycline response elements, a minimal CMV promoter, CFP fused to the peroxisomal targeting signal SKL, 24 MS2 translational operators (MS2 repeats), a rabbit β-globin intron/exon module, and a cleavage/polyadenylation signal. Expression of CFP-lac repressor allows the DNA to be visualized and expression of pTet-On (rtTA) in the presence of doxycycline (dox) drives expression from the CMV minimal promoter. When MS2-YFP (YFP fused to the MS2 coat protein) dimerizes and interacts with the stem loop structure of the translational operator, it allows the transcribed RNA to be visualized. (B) Quantitative Southern blot of clone 2-6-3 genomic DNA. A 2.4 kb fragment is produced when clone 2-6-3 genomic DNA and p3216PECMS2β are digested with NcoI which cuts at the 5′ end of CFP and within the β-globin intron. Comparison of known quantities of plasmid DNA equal to 100, 200, 300, and 400 copies per cell showed that 2-6-3 cells contain ~200 stably integrated copies of p3216PECMS2β. (C) DNA fluorescence in situ hybridization (DNA FISH) of 2-6-3 cells shows that there is a single integration site in the euchromatic region of chromosome 1p36. (D) Northern blot time course analysis of RNA isolated 0, 5, 10, 15, and 30 min after the induction of transcription. The last lane shows a lighter exposure of the 30 min time point. Pre-mRNA transcripts run at 3.4 kb and spliced mRNA at 2.8 kb. The probe recognizes the MS2 repeats. Actin was probed as a loading control. (E) Immunoblot time course analysis of CFP-SKL expression 0, 1, 2, 3, 4, 5, and 6 hr after the addition of doxycycline.

Figure 2. Visualization of DNA, RNA, and Protein in Living Cells

(A–F) U2OS 2-6-3 cells were transiently transfected with pSV2-CFP-lac repressor, pTet-ON (rtTA) and MS2-YFP, and imaging was begun 2.5 hr posttransfection. (A–C) At 0 min (−) dox, CFP-lac repressor marks the locus (A) and MS2-YFP is diffusely distributed throughout the nucleus (B). (D–F) 2.5 hr after the addition of Dox, the locus is highly decondensed and CFP-SKL is seen in the cytoplasmic peroxisomes (D). MS2-YFP accumulates at the site of the decondensed locus and is present in a particulate pattern throughout the nucleoplasm (E). (G–L) Image stacks of cells expressing pSV2-CFP-lac repressor (pseudocolored red) and EYFP-rtTA-N1 (pseudocolored green) were collected and deconvolved in cells fixed 5 min (G–I) and 60 min (J–L) after the induction of transcription. Single sections from deconvolved stacks are shown. (M–U) Factors involved in gene expression colocalize with the decondensed locus. YFP-RNA polymerase II (M–O), YFP-SF2/ASF (P–R), and Cstf64 (S–U) are present at the active locus. Scale bar is equal to 5 µm. Scale bar in enlarged insets is equal to 1 µm.

Figure 3. Kinetics of RNA Synthesis

(A–O) Still images from a time series of U2OS 2-6-3 cells during transcriptional activation showing the relationship between the chromatin of the locus, marked by CFP-lac repressor, and the RNA, marked by MS2-YFP. Also see Supplemental Data and Supplemental Movie S1 available on Cell website. Scale bar is equal to 5 µm. (P) Quantitative analysis and modeling of RNA levels at the locus. The intensity of the MS2-YFP signal at the locus as it associates with the synthesized RNA was measured every 2.5 min. There is an initial linear increase after the induction of transcription which slows over time (following the form A*(1−exp(−B*t))). The fit shows a deviation from this increase at 140 min after which a decrease predominates. There is a delay in the increase at 50 min.

Figure 4. Characterization of the Condensed Heterochromatic Locus

YFP-HP1α (A–C), YFP-HP1β (D–F), and YFP-HP1γ (G–I) colocalize with the condensed locus, marked by CFP-lac repressor and the histone H3 is trimethylated on lysine 9 (H3 tri-meK9) (J–L). The H3 lysine 9 modification is not detected after the induction of transcription (M–O; 2.5 hr postdox). The histone H3 K9 methyltransferases YFP-Suv39h1 (P–R) and YFP-G9a-L (S–U) are present at the condensed locus. Scale bar is equal to 5 µm.

Figure 5. Chromatin Immunoprecipitation (ChIP) Analysis of Heterochromatin Proteins and Modifications on the Inactive Gene Expression Plasmid

Diagram (A) depicts the gene expression plasmid and the location of the primer pairs used: (a) promoter, (b) beginning of β-globin intron/exon module, (c) end of β-globin intron/exon module, (d) and (e) bacterial plasmid sequence. (B) Results of the ChIP analysis showing the localization of HP1α, HP1γ, histone H3 di- and tri-MeK9, and Eu-HMTase1 on the gene expression plasmid. (C) The levels of the associated factors and modifications are plotted.

Figure 6. Dynamics of YFP-HP1α Depletion from the Locus during Transcriptional Activation

YFP-HP1α colocalizes with the condensed locus (0 min, A–C) and the condensed regions during the early time points of transcriptional activation (D–F, 17.5 min). It is seen in punctate structures at the 30 min time point (G–I) and appears smooth and diffuse by 50 min (J–L). 180 min postinduction, a dark region (HP1α depleted) that colocalizes with the decondensed locus, is seen in the YFP-HP1α image. Also see Supplemental Data and Supplemental Movie S2 available on Cell website. Scale bar is equal to 5 µm.

Figure 7. Analysis of the Deposition of the Histone Variant, H3.3, at the Locus during Transcriptional Activation

H3.3-YFP is not enriched at the condensed locus (A–C). A small spot adjacent to the locus is seen immediately after induction (D–F; 7.5 min). Significant deposition begins around the periphery of the decondensing locus (G–I, 40 min; J–L, 75 min) and eventually H3.3-YFP appears in a concentrated region that does not completely colocalize with CFP-lac repressor (pseudocolored red) (M–O; 180 min). Also see Supplemental Data and Supplemental Movie S3 available on Cell website. Scale bar is equal to 5 µm.