Inducible DNA-loop formation blocks transcriptional activation by an SV40 enhancer

Abstract

It is well established that gene expression in eukaryotes is controlled by sequence-dependent binding of trans-acting proteins to regulatory elements like promoters, enhancers or silencers. A less well understood level of gene regulation is governed by the various structural and functional states of chromatin, which have been ascribed to changes in covalent modification of core histone proteins. And, much on how topological domains in the genome take part in establishing and maintaining distinct gene expression patterns is still unknown. Here we present a set of regulatory proteins that allow to reversibly alter the DNA structure in vivo and in vitro by adding low molecular weight effectors that control their oligomerization and DNA binding. Using this approach, we completely regulate the activity of an SV40 enhancer in HeLa cells by reversible loop formation to topologically separate it from the promoter. This result establishes a new mechanism for DNA-structure-dependent gene regulation in vivo and provides evidence supporting the structural model of insulator function.

Figure 1

Dimerization activity of tDs is necessary and sufficient for regulation of enhancer activity. (A) Outline of the reporter plasmid pWHE206 used for the analysis of tD-dependent looping out of an enhancer. An SV40 enhancer located 2.0 kb downstream, respectively 2.9 kb upstream of an SV40 promoter, activates the transcription of a luciferase reporter gene. The enhancer is flanked by a repeat of seven TetR-binding sides (tetO; 19 bp). The two (tetO)₇ boxes are separated by 327 bp including the 237 bp SV40 enhancer. (B) Tet-dimerizers (tD) consist of a GCN4- (tD_G), LexA- (tD_L) or a dimerization-deficient GCN4- (tD_Gmut) dimerization domain fused to the C-terminal end of sc TetR. Presence (C) and functionality (F) of the dimerization domain are necessary to regulate the activity of an enhancer. Transregulators are indicated schematically in column one. The predicted regulatory situations are shown in column two. Cells were cultured in the absence (white bars) or presence (gray bars) of dox. Repression factors and basal luciferase activity (thin line) are indicated. Values represent the means of triplicate samples with standard deviations given in corrected relative light units (corr. RLU) per μg of total cell protein. (D, E) Western blot analysis to detect transregulators in HeLa cell extracts.

Figure 2

Regulatory effect of sc TetR^* is nonspecific and depends on the amount of transregulator expressing DNA transfected. (A) tD_G and sc TetR^* exhibit different regulatory behaviors. As controls, pGL3-promoter (OFF) lacking the SV40 enhancer and tet operators or pWHE206 (ON) were transfected without transregulator (indicated by dotted lines). Cells were cultured in the absence (filled/open triangles) or in the presence (filled/open squares) of dox. Values represent the means of triplicate samples given in corrected relative ray light units (corr. RLU) per μg of total cell protein. (B) Factor of repression was determined for both tD_G and sc TetR^* by comparison of luciferase activity in the presence and the absence of dox and plotted against different amounts of transfected transregulator-expressing DNA. (C) Western blot analysis to detect the respective transregulators in HeLa cell extracts. Higher amounts of DNA transfected lead to higher amounts of transregulator protein.

Figure 3

tD_G interacts with tet operators on both sides of the SV40 enhancer in vitro. (A) Partial restriction map of the pWHE206 plasmid. If tD_L interacts with tet operators on both sides of the SV40 enhancer on supercoiled DNA, two DNA fragments (718 and 337 bp) generated after digestion with restriction enzymes AfeI and PstI are expected to be bound to tD_L and migrate as a single DNA–protein complex in a native gel. (B) Analysis of tD_L-induced DNA–protein complex formation on supercoiled DNA: supercoiled pWHE206 was pre-incubated with or without tD_L, then digested and analyzed in a native agarose gel. The expected products of digestion are shown on the left. Different bands in the gel are arbitrarily numbered. M, peqGOLD DNA ladder (PEQLAB, Germany). (C) Analysis of DNA composition of the tD_L-dependent DNA–protein complex. DNA was purified from different bands (I–III in (B)) and analyzed on a 1.2% agarose gel. M, peqGOLD DNA ladder (PEQLAB, Germany).

Figure 4

Regulation of enhancer activity is not restricted to the supercoiled state of plasmids. (A) Reporter plasmid pWHE206, encoding an SV40 enhancer flanked with seven tet operators and a luciferase gene under the control of an SV40-promoter, was linearized. (B) Transregulators are indicated schematically in column one. The predicted regulatory situations are shown in column two. Cells were cultured in the absence (white bars) or presence (gray bars) of dox. Repression factors and basal luciferase activity (dotted line) are indicated. Values represent the means of triplicate samples with standard deviations given in corrected relative light units (corr. RLU) per μg of total cell protein.

Figure 5

Different models can explain the operating mode of the homodimerization system. (A) Internal contact model; (B) bead string model; (C) loop model.

Figure 6

Exclusion of internal contact and bead string models for enhancer regulation. (A) tHD with different DNA-binding and heterodimerization domains. (B, D) Transregulators are indicated schematically in column one. The predicted regulatory situations are shown in column two. Cells were cultured in the absence (white bars), in the presence (dark gray bars) of dox, in the presence of AP21967 (light gray bars) or in the presence of dox and AP21967 (black bars). Repression factors and basal luciferase activity are indicated. Values represent the means of triplicate samples with standard deviations given in corrected relative light units (corr. RLU) per μg of total cell protein. (C, E) Western blot analysis to detect the respective transregulators in HeLa cell extracts.