Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2001;9(6):265-81.
doi: 10.3727/000000001783992515.

Antagonistic Regulation of Dlx2 Expression by PITX2 and Msx2: Implications for Tooth Development

Affiliations
Free PMC article

Antagonistic Regulation of Dlx2 Expression by PITX2 and Msx2: Implications for Tooth Development

P D Green et al. Gene Expr. .
Free PMC article

Abstract

The transcriptional mechanisms underlying tooth development are only beginning to be understood. Pitx2, a bicoid-like homeodomain transcription factor, is the first transcriptional marker observed during tooth development. Because Pitx2, Msx2, and Dlx2 are expressed in the dental epithelium, we examined the transcriptional activity of PITX2 in concert with Msx2 and the Dlx2 promoter. PITX2 activated while Msx2 unexpectedly repressed transcription of a TK-Bicoid luciferase reporter in a tooth epithelial cell line (LS-8) and CHO cell line. Surprisingly, Msx2 binds to the bicoid element (5'-TAATCC-3') with a high specificity and competes with PITX2 for binding to this element. PITX2 binds to bicoid and bicoid-like elements in the Dlx2 promoter and activates this promoter 45-fold in CHO cells. However, it is only modestly activated in the LS-8 tooth epithelial cell line that endogenously expresses Msx2 and Pitx2. RT-PCR and Western blot assays reveal that two Pitx2 isoforms are expressed in the LS-8 cells. We further demonstrate that PITX2 dimerization can occur through the C-terminus of PITX2. Msx2 represses the Dlx2 promoter in CHO cells and coexpression of both PITX2 and Msx2 resulted in transcriptional antagonism of the Dlx2 promoter. Electrophoretic mobility shift assays demonstrate that factors in the LS-8 cell line specifically interact with PITX2. Thus, Dlx2 gene transcription is regulated by antagonistic effects between PITX2, Msx2, and factors expressed in the tooth epithelia.

Figures

Figure 1
Figure 1
Transcriptional control of the TK-bicoid promoter by Msx2 in CHO cells. CHO cells were transfected with either the TK-bicoid luciferase reporter gene containing four copies of the Pitx2 binding site (dashed boxes) or the parental TK-luciferase reporter without the bicoid sites. The cells were cotransfected with either the CMV-Pitx2 and/or -Msx2 expression plasmids (+) or the CMV plasmid without Pitx2 or Msx2 (–). To control for transfection efficiency, all transfections included the CMV β-galactosidase reporter. Cells were incubated for 24 h, then assayed for luciferase and β-galactosidase activities. The activities are shown as mean fold activation compared with TK-bicoid luciferase without Pitx2 expression and normalized to β-galactosidase activity (±SEM from four independent experiments). The mean TK-bicoid luciferase activity with Pitx2 expression was about 40,000 light units per 15 μg protein, and the β-galactosidase activity was about 70,000 light units per 15 μg protein.
Figure 2
Figure 2
Binding properties of Msx2. (A) Msx2 protein (∼130 ng) and PITX2 protein (∼150 ng) were incubated with the Msx or bicoid consensus sequence as the radioactive probe in the absence or presence of 50-fold molar excess unlabeled oligonucleotides as competitor DNAs. The EMSA experiments were analyzed in 8% native polyacrylamide gels. The free probe and bound complexes are indicated. (B) Scatchard plot of Msx2 protein binding to increasing amounts of bicoid probe. The free and bound forms of DNA were quantitated using the Molecular Dynamics STORM PhosphoImager.
Figure 3
Figure 3
PITX2 binds to DNA elements within the Dlx2 promoter. (A) Schematic of the Dlx2 promoter constructs used in transient transfection assays showing the location of bicoid and bicoid-like DNA elements, Bcd, bicoid, and bicoid-like sequences. (B) PITX2 protein (80 and 160 ng) was incubated with the Dlx2 bicoid consensus sequence (TAATCC), and the Dlx2 bicoid-like TATTCC sequence as the radioactive probe or our previously reported bicoid probe (4) in the absence or presence of 50-fold molar excess unlabeled bicoid element as competitor DNA. The EMSA experiments were analyzed in 8% native polyacrylamide gels. The bound forms of DNA were quantitated as described in Figure 2. The free probe and bound complexes are indicated.
Figure 4
Figure 4
Msx2 binds to DNA elements within the Dlx2 promoter. Msx2 protein (∼80 ng) was incubated with the radioactive Dlx2 probes described in Figure 3. The competitors used to demonstrate specific binding were our original bicoid oligonucleotide (4) and the two Dlx2 bicoid and bicoid-like oligonucleotides at 50-fold molar excess. The bound forms of DNA were quantitated as described in Figure 2. The free probe and bound complexes are indicated.
Figure 5
Figure 5
PITX2 and Msx2 proteins bind independently to the Dlx2 bicoid DNA element. PITX2 protein (either 40 or 80 ng) was mixed with varying amounts of Msx2 protein and incubated with the radioactive Dlx2 bicoid probe described in Figure 3. The bound forms of DNA were quantitated as described in Figure 2. The free probe and bound complexes are indicated.
Figure 6
Figure 6
Msx2 antagonizes PITX2 activation of the Dlx2 promoter. (A) CHO cells were transfected with either the Dlx2-3276 or Dlx2-200 luciferase reporter genes. The cells were cotransfected with either the CMV-PITX2 and/or -Msx2 expression plasmids (+) or the CMV plasmid without Pitx2 or Msx2 (−). (B) CHO cells were transfected with the Dlx2-3276 luciferase reporter and cotransfected with the indicated amounts of CMV-PITX2 and/or -Msx2 expression plasmids (+) or the CMV plasmid without PITX2 or Msx2 (−). To control for transfection efficiency, all transfections included the SV-40 β-galactosidase reporter. Cells were incubated for 24 h, then assayed for luciferase and β-galactosidase activities. The activities are shown as mean fold activation compared with the Dlx2 promoter plasmids without PITX2 expression and normalized to β-galactosidase activity [±SEM from four independent experiments for (A)]. Repression of the Dlx2 promoters by Msx2 is shown as being less than the control value set at 1 for fold activation. The mean Dlx2 promoter luciferase activity with PITX2 expression was about 100,000 light units per 15 μg protein, and the β-galactosidase activity was about 70,000 light units per 15 μg protein.
Figure 7
Figure 7
Identification of Pitx2 isoforms in LS-8 tooth epithelial cells. (A) Schematic of bacterial expressed PITX2 protein constructs showing the deleted portions of the proteins used in our assays (DPSKKKR, antibody recognition epitope; HD, homeodomain; OAR, 14 amino acid conserved region). (B) Western blot of purified bacterial proteins demonstrating specificity of the PITX2 P2R10 antibody. (C) RT-PCR of mRNA isolated from LS-8 cells showing the expression of Pitx2A and C isoforms. Primers specific to the N-terminal sequences, which differ in each isoform, were used in combination with an antisense homeodomain primer. Pitx2A and Pitx2C isoforms were detected using specific sense primers indicated by the arrows. All products were confirmed by sequencing the amplified bands. (D) Western blot of HeLa and LS-8 nuclear extracts (approximately 100 μg). No Pitx2 was detected in HeLa nuclear extracts while the two asterisks denote the Pitx2A (∼30 kDa) and the Pitx2C (∼36 kDa) isoforms in LS-8 nuclear extracts. (E) GST-PITX2 C173 pull-down assay with HeLa and LS-8 nuclear extracts. The endogenous Pitx2A and Pitx2C isoforms bind to the C-terminus of PITX2 and were detected using the PITX2 antibody by Western blot. This demonstrates that PITX2 can form homodimers through interaction within its C-terminal tail.
Figure 8
Figure 8
Msx2 attenuates PITX2 activation of the TK-bicoid luciferase reporter in LS-8 cells. (A) Western blot of LS-8 nuclear extract using a Msx2 antibody. LS-8 nuclear extract (10 and 20 μg) was tested for Msx2 protein expression. As a control, 400 ng of bacterial expressed Msx2 was used to show the correct migration of endogenous Msx2 and to demonstrate the purity of our Msx2 protein preparation. HeLa nuclear extract was used as a negative control. (B) LS-8 cells were transfected with either the TK-bicoid luciferase reporter gene or the parental TK-luciferase reporter without the bicoid sites. The cells were cotransfected with either the CMV-Pitx2 and/or -Msx2 expression plasmids (+) or the CMV plasmid without Pitx2 or Msx2 (−). To control for transfection efficiency, all transfections included the CMV β-galactosidase reporter. Cells were incubated for 24 h, then assayed for luciferase and β-galactosidase activities. The activities are shown as mean fold activation compared with TK-bicoid luciferase without Pitx2 expression and normalized to β-galactosidase activity (±SEM from five independent experiments). The mean TK-bicoid luciferase activity with Pitx2 expression was about 7000 light units per 15 μg protein, and the β-galactosidase activity was about 40,000 light units per 15 μg protein.
Figure 9
Figure 9
Transcriptional activity of the Dlx2 promoter is decreased in the LS-8 tooth epithelial cell line. LS-8 cells were transfected with either the Dlx2-3276 or Dlx2-200 luciferase reporter genes. The cells were cotransfected with either the CMV-Pitx2 and/or -Msx2 expression plasmids (+) or the CMV plasmid without Pitx2 (−). To control for transfection efficiency, all transfections included the CMV β-galactosidase reporter. Cells were incubated for 24 h, then assayed for luciferase and β-galactosidase activities. The activities are shown as mean fold activation compared with the Dlx2 promoters without Pitx2 expression and normalized to β-galactosidase activity (±SEM from three independent experiments). Repression of the Dlx2 promoters by Msx2 is shown as being less than the control value set at 1. The mean Dlx2-3276 luciferase activity with Pitx2 expression was about 5000 light units per 15 μg protein, and the β-galactosidase activity was about 40,000 light units per 15 μg protein.
Figure 10
Figure 10
LS-8 nuclear extract forms specific Pitx2-protein complexes. (A) LS-8 nuclear extract (NE) (∼3 μg) was incubated with the bicoid consensus sequence as the radioactive probe in the absence or presence of 50-fold molar excess unlabeled bicoid oligonucleotide (Bic) as competitor DNA. Approximately 80 ng of PITX2 CΔ39 was used in the EMSA to identify where the endogenous LS-8 Pitx2 homodimer migrates. We have previously shown that this protein readily forms homodimers (5). The EMSA experiments were analyzed on native 7% polyacrylamide gels. The free probe, dimer species, and bound complexes are indicated. (B) Purified bacterial expressed PITX2 was incubated with the bicoid probe in the absence and presence of PITX2 antibody (first two lanes on the left). The characteristic supershift caused by the PITX2–antibody complex is shown (supershift). LS-8 NE was incubated with the antibody prior to addition of the bicoid probe. HeLa NE was used to demonstrate the specificity of the LS-8 complexes. The last lane (far right) demonstrates that the PITX2 antibody does not bind to the probe. Asterisks denote the dimer and specific complexes.

Similar articles

See all similar articles

Cited by 27 articles

See all "Cited by" articles

Publication types

MeSH terms

LinkOut - more resources

Feedback