Transcription elongation regulator 1 is a co-integrator of the cell fate determination factor Dachshund homolog 1

Abstract

DACH1 (Dachshund homolog 1) is a key component of the retinal determination gene network and regulates gene expression either indirectly as a co-integrator or through direct DNA binding. The current studies were conducted to understand, at a higher level of resolution, the mechanisms governing DACH1-mediated transcriptional repression via DNA sequence-specific binding. DACH1 repressed gene transcription driven by the DACH1-responsive element (DRE). Recent genome-wide ChIP-Seq analysis demonstrated DACH1 binding sites co-localized with Forkhead protein (FOX) binding sites. Herein, DACH1 repressed, whereas FOX proteins enhanced, both DRE and FOXA-responsive element-driven gene expression. Reduced DACH1 expression using a shRNA approach enhanced FOX protein activity. As DACH1 antagonized FOX target gene expression and attenuated FOX signaling, we sought to identify limiting co-integrator proteins governing DACH1 signaling. Proteomic analysis identified transcription elongation regulator 1 (TCERG1) as the transcriptional co-regulator of DACH1 activity. The FF2 domain of TCERG1 was required for DACH1 binding, and the deletion of FF2 abolished DACH1 trans-repression function. The carboxyl terminus of DACH1 was necessary and sufficient for TCERG1 binding. Thus, DACH1 represses gene transcription through direct DNA binding to the promoter region of target genes by recruiting the transcriptional co-regulator, TCERG1.

FIGURE 1.

DACH1 functions as trans-repressor through specific DNA sequence. A and B, HeLa cells were transfected with a vector encoding the single copy of DRE in either the forward or reverse orientation upstream of either SV40 promoter (A) or the promoter-less pGL3-basic reporter together with a DACH1 expression vector or control (B). The -fold repression by DACH1 on both reporter genes was calculated relative to vector control. The -fold change by DACH1 was further normalized to the reporter backbone (the activity from vector backbone was set as 1). C, cells were transfected with an expression vector encoding DACH1 and luciferase reporter construct containing six copies of DRE upstream of an E4 TATA box. D, DACH1 linked to the Gal4 DBD was assessed using the UAS-Gal4-DBD binding site linked to the minimal TATA box. Transfection with increasing amounts of Gal4-DACH1 repressed transcription. The data are shown as mean ± S.E. (p < 0.05).

FIGURE 2.

DACH1 inhibits transcription through the Forkhead protein response element. A, ribbon model representation of the FOXM1 protein with a DNA binding site is shown. The helix loop-helix (HLH) domain of DACH1 resembles the FOXM1 DBD. The DNA is shown as a stick model. The color coding for bases are: A, yellow; T, green; G, cyan; and C, red. The x-ray structure of the DNA binding motif of Dachshund is shown in ribbon representation. For both proteins, the HLH segments are shown by cyan-colored ribbon, and the rest of the ribbon structures are colored magenta. The picture was generated by the UCSF Chimera visualization program (23). B, comparison of DACH1 binding nucleotides to the consensus Forkhead binding site and FOXM1 binding site. C–E, (DRE)₆-Luc or (FoxA)₆-Luc transfected into HEK 293T cells, showing that DACH1 repressed reporter activity, whereas FOXM1 and FOXO induced the reporter activity. F, 293T cells transfected with (DRE)₆-E4-TATA-Luc, showing that FOXC2 induced the reporter activity. The data throughout are shown as mean ± S.E. (error bars) from at least two separate experiments with triplicate samples each (p < 0.01).

FIGURE 3.

DACH1 inhibits FOXM1 transactivation. A, FOXM1 expression vector was co-transfected with increasing amounts of DACH1 expression vector together with the (DRE)₆-Luc reporter into MCF-7 breast cancer cells. B, luciferase reporter assay was conducted in HEK 293T cells transduced with either shRNA for DACH1 or control. Data are mean ± S.E. (error bars). C, ChIP assays were conducted using anti-FOXM1 antibody in cells transfected with DACH1 shRNA or control.

FIGURE 4.

DACH1 represses FOXC2-dependent cellular migration. MCF-10A cells were transduced with retroviral expression vectors encoding FOXC2 and subjected to GFP-FACS with subsequent transient transfection with vector encoding FLAG-tagged DACH1. Cells were then analyzed for cellular migration by Transwell assay (A) and for invasion by three-dimensional invasiveness assay (B). Crystal violet dye staining of cells that migrated in the Transwell assays is shown. The data are shown as mean ± S.E. (error bars) of the number of cells migrated in three separate experiments.

FIGURE 5.

DACH1 binds the co-integrator TCERG1. A, experimental approach used for identification of DACH1-binding proteins. B, HEK 293T cells transiently transfected with FLAG-DACH1-expressing vector. 50 mg of whole cell lysates were subjected to an immune-affinity column preloaded with a 1-ml slurry of M2 agarose beads (Sigma). The proteins associated with agarose beads were eluted with buffer containing 100 μm FLAG peptide. Western blotting was conducted of the eluted DACH1 using the anti-FLAG antibody (Sigma). C, eluted proteins separated on 4–20% gradient SDS-PAGE using a silver-stained gel. D, peptide sequence aligning to TCERG1. E, IP-Western blotting to determine DACH1 and TCERG1 binding. Expression vector for DACH1 (FLAG-tagged) and a series of TCERG1 deletion mutants (T7-tagged) were used to transfect HEK 293T cells. IP was conducted with anti-FLAG antibody for DACH1 and Western blotting with T7 antibody for TCERG1. F, schematic representation of the TCERG1 expression vectors and observed DACH1 binding ability. G, co-immunoprecipitation assays performed by incubating GST fusion proteins of TCERG1 with FLAG-DACH1 protein expressed in HEK 293T cells. IP was conducted using anti-FLAG (M2) antibody followed by Western blotting using GST antibody to detect mutants of TCERG1.

FIGURE 6.

TCERG1 is required for DACH1 trans-repression. A, HEK 293T cells transfected with DACH1 and TCERG1 mutants together with (DRE)₆-Luc. The carboxyl-terminal deletion of TCERG1, which does not bind DACH1, is defective for repression. B, HEK 293T cells transfected with TCERG1 siRNA and control for 48 h followed by transfection with indicated luciferase reporter plasmids. C, HEK 293T cells transfected with TCERG1 and Gal4-DACH1 or control vector together with Gal4 reporter, showing that TCERG1 enhanced DACH1-mediated transcriptional repression.

FIGURE 7.

C terminus of DACH1 conveys trans-repression function. A, schematic representation of DACH1 expression vectors used for IP-Western blotting to determine the minimal region of DACH1 required for TCERG1 association. B, HEK 293T cells co-transfected with T7-tagged TCERG1 and FLAG-tagged full-length DACH1 or carboxyl-terminal truncated form (ΔC). IP was conducted with anti-FLAG antibody, followed by Western blotting with anti-T7 antibody. C, effect of ΔDBD and carboxyl terminus (C-ter) on wild-type DACH1 measured by reporter assay using (DRE)₆-Luc. D, HEK 293T cells co-transfected with increased amount of carboxyl-terminal deletion mutant (ΔC) plasmid DNA and vector expressing full-length DACH1 together with (DRE)₆-Luc reporter. The dominant effect is dose-dependent. -Fold repression was calculated as the ratio of activities measured from DACH1-transfected cells to those in vector control-transfected cells. The data are shown as mean ± S.E. (error bars) from at least two separate experiments with triplicate samples each. E, schematic representation of proposed model by which DACH1 and FOXM1 compete to regulate gene expression.