Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 24 (2), 809-22

Transcriptional and DNA Binding Activity of the Foxp1/2/4 Family Is Modulated by Heterotypic and Homotypic Protein Interactions

Affiliations

Transcriptional and DNA Binding Activity of the Foxp1/2/4 Family Is Modulated by Heterotypic and Homotypic Protein Interactions

Shanru Li et al. Mol Cell Biol.

Abstract

Foxp1, Foxp2, and Foxp4 are large multidomain transcriptional regulators belonging to the family of winged-helix DNA binding proteins known as the Fox family. Foxp1 and Foxp2 have been shown to act as transcriptional repressors, while regulatory activity of the recently identified Foxp4 has not been determined. Given the importance of this Fox gene subfamily in neural and lung development, we sought to elucidate the mechanisms by which Foxp1, Foxp2, and Foxp4 repress gene transcription. We show that like Foxp1 and Foxp2, Foxp4 represses transcription. Analysis of the N-terminal repression domain in Foxp1, Foxp2, and Foxp4 shows that this region contains two separate and distinct repression subdomains that are highly homologous termed subdomain 1 and subdomain 2. However, subdomain 2 is not functional in Foxp4. Screening for proteins that interact with subdomains 1 and 2 of Foxp2 using yeast two-hybrid analysis revealed that subdomain 2 binds to C-terminal binding protein 1, which can synergistically repress transcription with Foxp1 and Foxp2, but not Foxp4. Subdomain 1 contains a highly conserved leucine zipper similar to that found in N-myc and confers homo- and heterodimerization to the Foxp1/2/4 family members. These interactions are dependent on the conserved leucine zipper motif. Finally, we show that the integrity of this subdomain is essential for DNA binding, making Foxp1, Foxp2, and Foxp4 the first Fox proteins that require dimerization for DNA binding. These data reveal a complex regulatory mechanism underlying Foxp1, Foxp2, and Foxp4 activity, demonstrating that Foxp1, Foxp2, and Foxp4 are the first Fox proteins reported whose activity is regulated by homo- and heterodimerization.

Figures

FIG. 1.
FIG. 1.
Foxp4 represses lung-specific gene transcription. H441 cells were cotransfected with pCMVTag3B (pCMVTag), pCMV/Foxp1 (Foxp1), pCMV/Foxp2 (Foxp2), or pCMV/Foxp4 (Foxp4) expression plasmid and with the pCC10.luc (A) or pCMV.luc (B) reporter plasmid along with the pMSVβgal reference plasmid. Forty-eight hours after transfection, cells were harvested, and luciferase activity was measured and normalized to the activity obtained after transfection with the pCMVTag3B plasmid. All assays included the pMSVβgal control plasmid, and differences in transfection efficiencies were corrected using a commercial β-galactosidase assay. Assays were performed in triplicate, and the data are presented as the maximum percentage of relative luciferase activity obtained upon cotransfection of either reporter plasmid with the pCMVTag3B plasmid ± standard error of the mean. DBS, DNA binding site.
FIG. 2.
FIG. 2.
Identification of subdomains 1 and 2 in Foxp2 protein. NIH 3T3 cells were transfected with the pGAL4SV40.luc reporter plasmid along with expression plasmids encoding the GAL4 DNA binding domain or GAL4 chimeric expression vectors encoding the indicated regions of the mouse Foxp2 protein. Forty-eight hours after transfection, cells were harvested, and luciferase activity was measured and normalized to the activity obtained after transfection with the pGAL4 plasmid. (A) Definition of subdomains 1 and 2 as aa 260 to 419 and aa 418 to 500, respectively. (B) The zinc finger of subdomain 1 is not essential for transcriptional repression. All assays included the pMSVβgal control plasmid, and differences in transfection efficiencies were corrected using a commercial β-galactosidase assay. Assays were performed in triplicate, and the data are presented as the maximum percentage of relative luciferase activity obtained upon cotransfection of the pGAL4SV40.luc reporter plasmid with the pGAL4 plasmid ± standard error of the mean. GAL4 fusion proteins delineating subdomains 1 and 2 are indicated on the left in panel A. The zinc finger (black box) and leucine zipper (leuZip) (bracket) motifs are indicated. All GAL4 fusion proteins were expressed at approximately equal levels (data not shown).
FIG. 3.
FIG. 3.
Conservation of subdomain 1 but not subdomain 2 in Foxp1, Foxp2, and Foxp4 proteins. NIH 3T3 cells were transfected with the pGAL4SV40.luc reporter plasmid along with expression plasmids encoding the GAL4 DNA binding domain or GAL4 chimeric expression vectors encoding the indicated regions of the mouse Foxp1, Foxp2, and Foxp4 proteins. Forty-eight hours after transfection, cells were harvested, and luciferase activity was measured and normalized to the activity obtained after transfection with the pGAL4 plasmid. All assays included the pMSVβgal control plasmid, and differences in transfection efficiencies were corrected using a commercial β-galactosidase assay. Assays were performed in triplicate, and the data are presented as the maximum percentage of relative luciferase activity obtained upon cotransfection of the pGAL4SV40.luc reporter plasmid with the pGAL4 plasmid ± standard error of the mean. Note the conservation of repression in subdomain 1 (A) but not subdomain 2 (B) in Foxp1, Foxp2, and Foxp4. However, full-length Foxp1, Foxp2, and Foxp4 proteins display similar repression activity when fused to GAL4 (C). The zinc finger motif is indicated by a black box in panel A. All GAL4 fusion proteins were expressed at approximately equal levels (data not shown).
FIG. 4.
FIG. 4.
The corepressor CtBP-1 interacts with and represses transcription through Foxp1 and Foxp2 but not Foxp4. (A) The CtBP-1 (CtBP) binding motif is conserved in Foxp1 and Foxp2 but not Foxp4. Note the change from a conserved leucine to a proline. (B) CtBP-1 coimmunoprecipitates with Foxp1 and Foxp2. HEK-293 cells were cotransfected with either pCMV/Foxp1 (lanes 2 and 4), pCMV/Foxp2 (lanes 6 and 7), and pCMV/CtBP-1 (lanes 3, 4, 7, and 8) or with pCMVTag3B control vector (lanes 1 and 5). The Foxp1 and Foxp2 proteins were FLAG tagged, and CtBP-1 was myc tagged. Forty-eight hours after transfection, coimmunoprecipitations were performed on cell extracts using an anti-c-myc (α-myc) monoclonal antibody (9E10). Immunoprecipitates (IP) were probed with an anti-FLAG (α-FLAG) monoclonal antibody (M2). The presence of Foxp1 and Foxp2 coimmunoprecipitating with CtBP-1 is indicated by arrows. The cross-reactivity of the secondary antibody to the heavy chain of immunoglobulin G is indicated by asterisks. Expression of Foxp1 or Foxp2 (black arrowheads) or CtBP-1 (white arrowheads) in Western blots (WB) of cell extracts is indicated. (C) H441 cells were transfected with the pCC10.luc reporter plasmid along with expression plasmids encoding the full-length mouse Foxp1, Foxp2, and Foxp4 proteins and increasing amounts of the pCMV/CtBP-1 expression plasmid. Forty-eight hours after transfection, cells were harvested, and luciferase activity was measured and normalized to the activity obtained after transfection with the pCC10.luc plasmid and the pCMV/Foxp1, pCMV/Foxp2, and pCMV/Foxp4 plasmids without the pCMV/CtBP-1 expression plasmid. (D) Trun-cation of the CtBP-1 site in subdomain 2 of Foxp2 eliminates transcriptional repression by this domain and demonstrates that aa 418 to 429 are sufficient for repression by this domain. NIH 3T3 cells were transfected with expression vectors encoding the indicated Foxp2-GAL4 fusion proteins. Luciferase assays were performed, and luciferase activity was normalized as described in the legends to Fig. 1 to 3. All GAL4 fusion proteins were expressed at approximately equal levels (data not shown). (E) RT-PCR using H441 cell cDNA to detect CtBP-1 and CtBP-2 expression. Lanes 1 and 2 used oligonucleotides specific for CtBP-1, while lanes 2 and 3 used oligonucleotides specific for CtBP-2. Reaction mixtures lacking RT (lanes 1 and 3) and reaction mixtures containing RT (lanes 2 and 4) were used. (F) Mutation of the CtBP-1 binding site in Foxp1 and Foxp2 does not affect repression of the mouse CC10 promoter. Mutations were introduced into the CtBP-1 binding motif in Foxp1 and Foxp2 (Materials and Methods), and these plasmids (Foxp1-mut and Foxp2-mut) were transfected along with wild-type pCMV/Foxp1 (Foxp1-WT) and pCMV/Foxp2 (Foxp2-WT) plasmids and the pCC10.luc reporter plasmid into H441 cells. Forty-eight hours after transfection, cells were harvested, and luciferase activity was measured and normalized to the activity obtained after transfection with the pCC10.luc plasmid and the pCMVTag3B expression plasmid. All assays included the pMSVβgal control plasmid, and differences in transfection efficiencies were corrected using a commercial β-galactosidase assay. Assays were performed in triplicate, and the data are presented as the maximum percentage of relative luciferase activity obtained upon cotransfection of the pGAL4SV40.luc reporter plasmid with the pGAL4 plasmid ± standard error of the mean.
FIG. 5.
FIG. 5.
The leucine zipper motif is important for repression by subdomain 1. (A) Alignment of the leucine zipper (leuzip) motifs of Foxp1, Foxp2, Foxp3, Foxp4, and N-myc, showing the conservation of this region. Identical amino acids (white letters on black background) and similar amino acids (light gray shaded background) are indicated. Asterisks denote conserved positions of leucines, or in the case of Foxp1, Foxp2, Foxp3, and Foxp4, a valine substitution at the first position. The leucine mutated in these experiments is denoted with an arrow. (B) Mutation of the fourth leucine in Foxp2 eliminates repression of subdomain 1. NIH 3T3 cells were transfected with the pGAL4SV40.luc reporter plasmid along with expression plasmids encoding the GAL4 DNA binding domain or GAL4 chimeric expression vectors encoding the indicated regions of the Foxp2 protein. Forty-eight hours after transfection, cells were harvested, and luciferase activity was measured and normalized to the activity obtained after transfection with the pGAL4 plasmid. Note that mutation of either aa 407 and 408 (HL to AA) or just the leucine at aa 408 to alanine (L to A) results in a loss of repression by subdomain 1. (C) Leucine zipper mutations in subdomain 1 of Foxp1 and Foxp4 also eliminate repression. NIH 3T3 cells were transfected with the pGAL4SV40.luc reporter plasmid along with expression plasmids encoding the GAL4 DNA binding domain or GAL4 chimeric expression vectors encoding the indicated regions of the Foxp1 and Foxp4 proteins. Forty-eight hours after transfection, cells were harvested, and luciferase activity was measured and normalized to the activity obtained after transfection with the pGAL4 plasmid. Note loss of repression by mutation of the fourth conserved leucine in subdomain 1 of Foxp1 (aa 397) and Foxp4 (aa 376). All assays included the pMSVβgal control plasmid, and differences in transfection efficiencies were corrected using a commercial β-galactosidase assay. Assays were performed in triplicate, and the data are presented as the maximum percentage of relative luciferase activity obtained upon cotransfection of the pGAL4SV40.luc reporter plasmid with the pGAL4 plasmid ± standard error of the mean. All GAL4 fusion proteins were expressed at approximately equal levels (data not shown).
FIG. 6.
FIG. 6.
The leucine zipper motif in Foxp1, Foxp2, and Foxp4 mediates homo- and heterodimerization. HEK-293 cells were transfected with combinations of expression vectors encoding Foxp1, Foxp2, and Foxp4 proteins that had been tagged with FLAG or c-myc. The c-myc monoclonal antibody (9E10) was used to immunoprecipitate proteins from cell extracts. Immunoprecipitated proteins were resolved on SDS-polyacrylamide gels or Western blots, which were probed with the anti-FLAG M2 monoclonal antibody to reveal coimmunoprecipitated proteins. The first protein listed above each lane was tagged with c-myc, while the second protein was tagged with FLAG. The cells were transfected by plasmids encoding myc-tagged Foxp1 (Foxp1), myc-tagged Foxp2 (Foxp2), myc-tagged Foxp4 (Foxp4), myc-tagged Foxp1 and FLAG-tagged Foxp1 (p1/p1), myc-tagged Foxp2 and FLAG-tagged Foxp2 (p2/p2), myc-tagged Foxp4 and FLAG-tagged Foxp4 (p4/p4), myc-tagged Foxp1 and FLAG-tagged Foxp2 (p1/p2), myc-tagged Foxp1 and FLAG-tagged Foxp4 (p1/p4), myc-tagged Foxp2 and FLAG-tagged Foxp4 (p2/p4) proteins. (B) Western blots of sample cell lysate showing expression of myc- and FLAG-tagged Foxp1. Lysates from pCMVTag3B-transfected (pCMVTag), myc-tagged Foxp1 plasmid-transfected (p1/myc), FLAG-tagged Foxp1 plasmid-transfected (p1/FLAG), and both myc- and FLAG-tagged Foxp1 plasmid-transfected (p1/p1) HEK-293 cells were used. (C) Coimmunoprecipitation assays using Foxp1, Foxp2, and Foxp4 leucine zipper deletion (Δleu) mutants. Results are configured as in panel A, with the first protein being myc tagged and the second protein being FLAG tagged. The c-myc monoclonal antibody was used for immunoprecipitations, while the anti-FLAG monoclonal antibody was used on Western blots for immunodetection of coimmunoprecipitated proteins. The cells were transfected with myc-tagged Foxp2 and FLAG-tagged Foxp2 (p2/p2), myc-tagged Foxp1Δleu mutant and FLAG-tagged Foxp1 (p1Δleu/p1), myc-tagged Foxp2Δleu mutant and FLAG-tagged Foxp2 (p2Δleu/p2), myc-tagged Foxp4Δleu mutant and FLAG-tagged Foxp4 (p4Δleu/p4), myc-tagged Foxp4Δleu mutant and FLAG-tagged Foxp2 (p4Δleu/p2), myc-tagged Foxp4Δleu mutant and FLAG-tagged Foxp1 (p4Δleu/p1), myc-tagged Foxp2Δleu mutant and FLAG-tagged Foxp1 (p2Δleu/p1) expression plasmids. (D) Western blots of sample cell lysate showing expression of wild-type Foxp1 and Foxp1Δleu proteins. Lysates from pCMVTag3B-transfected (pCMVTag), myc-tagged Foxp1Δleu mutant plasmid-transfected (p1Δleu/myc), FLAG-tagged Foxp1Δleu mutant plasmid-transfected (p1Δleu/FLAG), and myc-tagged Foxp1Δleu mutant and FLAG-tagged Foxp1 plasmid-transfected (p1Δleu/p1) HEK-293 cells. The arrows in panels A and C indicate the 70- to 75-kDa bands of full-length Foxp1, Foxp2, and Foxp4 proteins. The antibodies used for each Western blot panel are indicated to the right in panels B and D.
FIG. 7.
FIG. 7.
Deletion of a single glutamic acid eliminates homo- and heterodimerization of Foxp1, Foxp2, and Foxp4. (A) Alignment of the leucine zipper (leuzip) motifs in Foxp1, Foxp2, Foxp3, Foxp4, and N-myc proteins, showing the conservation of the glutamic acid that is deleted in certain IPEX patients. Identical amino acids (white letters on black background) and similar amino acids (light gray shaded background) are indicated. (B) Coimmunoprecipitation assays using Foxp1, Foxp2, and Foxp4 ΔE mutants. Results are shown as described in the legend to Fig. 6, with the myc-tagged protein listed first and the FLAG- tagged protein listed second. The c-myc monoclonal antibody was used for immunoprecipitations, while the anti-FLAG monoclonal antibody was used on Western blots for immunodetection of coimmunoprecipitated proteins. Cells were transfected with myc-tagged Foxp2 and FLAG-tagged Foxp2 (p2/p2), myc-tagged Foxp1ΔE mutant and FLAG-tagged Foxp1 (p1ΔE/p1), myc-tagged Foxp2ΔE mutant and FLAG-tagged Foxp2 (p2ΔE/p2), myc-tagged Foxp4ΔE mutant and FLAG-tagged Foxp4 (p4ΔE/p4), myc-tagged Foxp4ΔE mutant and FLAG-tagged Foxp2 (p4ΔE/p2), myc-tagged Foxp4ΔE mutant and FLAG-tagged Foxp1 (p4ΔE/p1), myc-tagged Foxp2ΔE mutant and FLAG-tagged Foxp1 (p2ΔE/p1) plasmids. (C) Western blots of sample cell lysate showing expression of wild-type Foxp1, Foxp2, Foxp1ΔE, and Foxp2ΔE proteins. Lysates from myc-tagged Foxp1ΔE mutant plasmid-transfected (p1ΔE/myc), FLAG-tagged Foxp1 plasmid-transfected (p1/FLAG), myc-tagged Foxp1ΔE mutant and FLAG-tagged Foxp1 (p1Δleu/p1) plasmid-transfected, and myc-tagged Foxp2DE mutant and FLAG-tagged Foxp2 plasmid-transfected (p2ΔE/p2) HEK-293 cells. The arrow indicates the 70- to 75-kDa bands of full-length Foxp1, Foxp2, and Foxp4 proteins.
FIG. 8.
FIG. 8.
Disruption of the leucine zipper motif in Foxp1, Foxp2, and Foxp4 results in loss of transcriptional repression. (A) H441 cells were transfected with the pCC10.luc reporter plasmid along with the pCMVTag3B control plasmid (pCMVTag), wild-type Fopx1 (Foxp1), Foxp1Δleu mutant (Foxp1Δleuzip), wild-type Foxp2 (Foxp2), or Foxp2Δleu mutant (Foxp2Δleuzip). Forty-eight hours after transfection, cells were harvested, and luciferase activity was measured and normalized to the activity obtained after transfection with the pCMVTag3B plasmid. (B) H441 cells were transfected with the pCC10.luc reporter along with the pCMVTag3B control plasmid (pCMVTag) and with a plasmid encoding wild-type Foxp1 (Foxp1), Foxp1ΔE mutant (Foxp1Δ-E), wild-type Foxp2 (Foxp2), Foxp2ΔE mutant (Foxp2Δ-E), wild-type Foxp4 (Foxp4), or Fopx4ΔE mutant (Foxp4Δ-E) protein. All assays included the pMSVβgal control plasmid, and differences in transfection efficiencies were corrected using a commercial β-galactosidase assay. Assays were performed in triplicate, and the data are presented as the maximum percentage of relative luciferase activity obtained upon cotransfection of the pCC10.luc reporter plasmid with the pCMVTag3B plasmid ± standard error of the mean.
FIG. 9.
FIG. 9.
Dimerization is required for Foxp1, Foxp2, and Foxp4 proteins to bind DNA. The Fox DNA binding consensus site in the mouse CC10 promoter was used to perform EMSA on in vitro-translated Foxp1, Foxp2, and Foxp4 or GST-Foxp2 fusion proteins. (A) EMSA on unprogrammed cell lysate (lane 1) and cells treated with wild-type Foxp1, Foxp2, and Foxp4 alone (lanes 2 to 4) and with specific polyclonal antibodies (Ab) to each protein (lanes 5 to 7), and with mutant Foxp1, Foxp2, and Foxp4 ΔE proteins (lanes 8 to 10). Wild-type Foxp1, Foxp2, and Foxp4 DNA binding was eliminated with specific polyclonal antibodies to each protein (lanes 5 to 7). (B) Western blot of in vitro-translated wild-type Foxp1, Foxp2, and Foxp4 and Foxp1, Foxp2, and Foxp4 ΔE mutations. The same amount of in vitro-translated protein used in the EMSA (5 μl) was resolved on a SDS-polyacrylamide gel, blotted, and probed with the anti-myc monoclonal antibody. The position of the Foxp1, Foxp2, and Foxp4 proteins is indicated by the arrow. (C) EMSA using the mouse CC10 Fox oligonucleotide and GST protein (lane 1), untreated GST-Foxp2 protein (lane 2), GST-Foxp2 protein incubated overnight (ON) without thrombin (lane 3), and GST-Foxp2 protein incubated overnight with thrombin (lane 4). The shifted band is indicated by the arrow. (D) SDS-polyacrylamide gel showing the integrity of GST-Foxp2 (lane 1) and cleaved GST-Foxp2 (lane 2) proteins. The black arrow indicates the full-length GST-Foxp2 fusion protein, and the white arrow indicates the cleaved GST and Foxp2 portions of the fusion protein, which comigrate closely on the gel.

Similar articles

See all similar articles

Cited by 116 PubMed Central articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback