Regulation of the mouse Treacher Collins syndrome homolog (Tcof1) promoter through differential repression of constitutive expression

Abstract

Treacher Collins syndrome is an autosomal-dominant mandibulofacial dysostosis caused by haploinsufficiency of the TCOF1 gene product treacle. Mouse Tcof1 protein is approximately 61% identical and 71% similar to treacle, and heterozygous knockout of Tcof1 causes craniofacial malformation. Tcof1 expression is high in developing neural crest, but much lower in other tissues. To investigate this dual regulation, highly conserved regions upstream of TCOF1 homologs were tested through deletion and mutation reporter assays, and conserved predicted transcription factor binding sites were assessed through chromatin binding studies. Assays were performed in mouse P19 embryonic carcinoma cells and in HEK293 cells to determine differential activation in cell types at different stages of differentiation. Binding of Cebpb, Zfp161, and Sp1 transcription factors was specific to the Tcof1 regulatory region in P19 cells. The Zfp161 binding site demonstrated P19 cell-specific repression, while the Sp1/Sp3 candidate site demonstrated HEK293 cell-specific activation. Moreover, presence of c-myb and Zfp161 transcripts was specific to P19 cells. A minimal promoter fragment from -253 to +43 bp directs constitutive expression in both cell types, and dual regulation of Tcof1 appears to be through differential repression of this minimal promoter. The CpG island at the transcription start site remains unmethylated in P19 cells, 11.5 dpc mouse embryonic tissue, and adult mouse ear, which supports constitutive activation of the Tcof1 promoter.

FIG. 1.

Deletion analysis of the mouse Tcof1 regulatory region. Deletions from the 5′ and 3′ ends of the regulatory region show different effects on expression of the reporter gene luciferase in vitro. (A) Schematic representation of the inserts used in deletion clones is shown. Six different 5′ ends are labeled 1–6, while three different 3′ ends are labeled a, b, and c, and position relative to the TSS (+1) is indicated below. Fragments were inserted into pGL3-basic, and are named using the combination of the numbered 5′ end and the lettered 3′ end. Corresponding restriction sites are 1, KpnI; 2, NdeI; 3, SspI; 4, HincII; 5, HaeII; 6, PvuII; a, NheI; b, SmaI; and c, NcoI. Vertical arrows indicate peaks of areas of conservation among vertebrates. A bent arrow indicates the initiator ATG codon, and the gray box indicates exon 1. (B) Transient transfection luciferase expression analysis of deletion clones in P19 mouse embryonic carcinoma cells. Transfections were performed in triplicate and are shown as a fold-of-control value with standard deviations. Comparisons of results are pairwise, with each construct compared to the immediate longer construct to assess specific effects of deletion. Bracketed groups assess significance of 3′ deletions by comparing pairwise 5′ deletions within the group and testing significance of the group. Significance of comparisons is determined using the appropriate Student's t-test with *p < 0.01, **p < 0.005, or ***p < 0.001.

FIG. 2.

Conserved putative transcription factor binding sites within the Tcof1 regulatory region. Consensus sites for conserved predicted transcription factor binding sites are given on the top row, with species names in the left column. Consensus binding site direction is shown as forward (+) or reverse (−) relative to the Tcof1 open reading frame. Nonconsensus nucleotides are in lower case and bold type, while consensus nucleotides are in upper case. The position indicated at the bottom is for the mouse Tcof1 sequence.

FIG. 3.

Specificity of conserved putative transcription factor binding sites within the mouse Tcof1 regulatory region. Site-directed mutation of candidate sites was performed, and constructs were transiently transfected into the P19 mouse embryonic carcinoma cell line. The site mutated is indicated with an asterisk followed by the name of the clone used as template. Transfections were performed in triplicate, and results are shown as a fold-of-control value with standard deviations. Comparisons of results are pairwise, with each mutated construct compared to the wildtype template. Significance of comparisons is determined using the appropriate Student's t-test with *p < 0.01, **p < 0.005, or ***p < 0.001.

FIG. 4.

Analysis of cell-specific regulation of Tcof1. (A) Transient transfection luciferase expression analysis of deletion clones in HEK293 human embryonic kidney transformed cells. Deletion clone constructs are the same as for P19 cells as in Figure 1A, and results are reported as in Figure 1B. (B) Transient transfection luciferase expression analysis of site-directed mutation clones in HEK293 cells. Clone constructs are the same as for P19 cells, and results are reported as in Figure 3.

FIG. 5.

Presence of candidate transcription factor transcripts in cell lines. Semi-quantitative reverse transcription PCR analysis of candidate transcription factors was performed in triplicate, and transcript level was determined as a ratio of GAPDH transcript level at 22 cycles of PCR. Values are given with standard deviations for (A) P19 cells and (B) HEK293 cells. Transcripts not detectable after 40 cycles of PCR are denoted as n.d.

FIG. 6.

Chromatin enrichment of Tcof1 regulatory region. PCR of enriched chromatin was performed using immunoprecipitated chromatin samples for (1) no antibody background control, (2) input chromatin positive control, (3) anti-c-myb, (4) anti-Cebpb, (5) anti-Zfp161, (6) anti-Sp1, (7) anti-Sp3, (8) anti-AP2α, and (9) water only. Marker (M) shows 200 and 300 bp (and 400 bp in middle row) bands. Amplification shown was performed using a primer set surrounding the (top) c-myb candidate binding site with a product size of 219 bp, (middle) CCAAT candidate binding site with a product size of 222 bp, and (bottom) Zfp161 candidate site with a product size of 252 bp.

FIG. 7.

MethylScreen analysis. PCR product is shown at 252 bp for six different genomic DNA samples: untreated P19 cell, HhaI methylated P19 cell, SssI methylated P19 cell, untreated E11.5 mouse whole embryo, and two untreated adult mouse ear punch samples. Genomic DNA was subjected to each of four restriction digest reactions: (A) HhaI methylation-sensitive restriction endonuclease, (B) McrBC methylation-dependent restriction endonuclease, (C) both HhaI and McrBC enzymes, and (D) glycerol only. Positive control template (+) for PCR is untreated P19 cell DNA not subjected to restriction digestion, and negative control template (−) for PCR is water only.