MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Other Versions of this Article:
MCB.01133-06v1
27/6/2155    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Minoo, P.
Right arrow Articles by Li, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Minoo, P.
Right arrow Articles by Li, C.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, March 2007, p. 2155-2165, Vol. 27, No. 6
0270-7306/07/$08.00+0     doi:10.1128/MCB.01133-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Physical and Functional Interactions between Homeodomain NKX2.1 and Winged Helix/Forkhead FOXA1 in Lung Epithelial Cells{triangledown}

Parviz Minoo,* Lingyan Hu, Yiming Xing, Nian Ling Zhu, Hongyan Chen, Min Li, Zea Borok, and Changgong Li

Department of Pediatrics, Division of Neonatology, and Department of Medicine, Will Rogers Institute Pulmonary Research Center, University of Southern California Keck School of Medicine, Los Angeles, California 90033

Received 23 June 2006/ Returned for modification 28 July 2006/ Accepted 25 December 2006


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
NKX2.1 is a homeodomain transcription factor that controls development of the brain, lung, and thyroid. In the lung, Nkx2.1 is expressed in a proximo-distal gradient and activates specific genes in phenotypically distinct epithelial cells located along this axis. The mechanisms by which NKX2.1 controls its target genes may involve interactions with other transcription factors. We examined whether NKX2.1 interacts with members of the winged-helix/forkhead family of FOXA transcription factors to regulate two spatially and cell type-specific genes, SpC and Ccsp. The results show that NKX2.1 interacts physically and functionally with FOXA1. The nature of the interaction is inhibitory and occurs through the NKX2.1 homeodomain in a DNA-independent manner. On SpC, which lacks a FOXA1 binding site, FOXA1 attenuates NKX2.1-dependent transcription. Inhibition of FOXA1 by small interfering RNA increased SpC mRNA, demonstrating the in vivo relevance of this finding. In contrast, FOXA1 and NKX2.1 additively activate transcription from Ccsp, which includes both NKX2.1 and FOXA1 binding sites. In electrophoretic mobility shift assays, the GST-FOXA1 fusion protein interferes with the formation of NKX2.1 transcriptional complexes by potentially masking the latter's homeodomain DNA binding function. These findings suggest a novel mode of selective gene regulation by proximo-distal gradient distribution of and functional interactions between forkhead and homeodomain transcription factors.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Members of the homeodomain family of transcription factors are distinguished by a highly conserved 60-amino-acid DNA binding motif (the homeodomain) and are critical to the process of embryonic development. Homeodomain proteins are found in multiple tissues, and yet they activate or repress distinct gene batteries in each organ. Even within the same tissue, homeodomain transcription factors may activate or repress distinct target genes, and their presence alone does not necessarily constitute target gene expression. In addition, it is difficult to envisage how transcriptional specificity can be accomplished by interactions of a highly conserved homeodomain DNA binding motif with variable target DNA. One simple and potential mechanism for transcriptional specificity may be via interactions with other proteins. Consistent with this view, homeodomain transcription factors are known to interact with a variety of other transcription factors, including SMAD (16), GATA (18), basic helix-loop-helix (bHLH) (26), and serum response factor (6).

The role of the homeodomain protein NKX2.1 (otherwise known as TTF-1 or T/EBP) is central to structural development and tissue-specific gene regulation in the brain, lung, and thyroid (14). In each tissue, NKX2.1 activates a distinct battery of genes. In vitro results using tissue-specific cell lines show that NKX2.1 can bind to and activate transcription from a number of lung-, thyroid- and brain-specific (or -enriched) genes. In the lungs of Nkx2.1–/– embryos, gene expression analysis has verified the role of NKX2.1 in regulation of lung-enriched genes, such as SpC and Ccsp, which encode the pulmonary Surfactant Protein C and Clara Cell-Specific Protein, respectively (20). In mouse embryonic lungs, Nkx2.1 is expressed in a proximo-distal (P-D) gradient, with the highest level found at the distal tips of endodermal airways, where it is known to activate SpC. In the proximal lung, where its levels are low, NKX2.1 activates Ccsp but not SpC. The mechanism for this selective gene activation remains unknown but likely involves specific interactions with other transcription factors. NKX2.1, like other homeodomain proteins, is known to interact with members of other transcription factor families, including RAR (31), GATA (18), bHLH (26), and SMAD (16).

In addition to NKX2.1, the promoter sequences of a number of lung genes, including Ccsp but not SpC, contain functional binding sites for members of the winged helix/forkhead family of FOXA transcription factors, including FOXA1 (HNF-3-alpha) and FOXA2 (HNF-3-beta). Homozygous disruption of Foxa1 leads to severe postnatal growth retardation, followed by death between postnatal days 2 and 12 (12). Whether there are lung or respiratory defects in these animals has not yet been investigated. Foxa2 is expressed in the node, notochord, floor plate, and gut in mouse embryos. Functional deletion of this gene leads to embryonic lethality before the formation of the lung primordium (2). Compound mutants lacking both Foxa1 and Foxa2 exhibit inhibition of lung cell proliferation, differentiation, and branching morphogenesis (27). Interactions between members of the FOXA family and other transcription factors have been reported to occur in regulation of both liver- and lung-specific genes (28).

We have examined the possibility that differential gene expression along the P-D axis of the lung may be due to specific interactions between NKX2.1 and members of the FOXA family of transcription factors. FOXA2 is known to interact with homeodomain proteins to bring about specific gene regulation. For example, characterization of null mice for Foxa2, Goosecoid, and Lim1 suggested that FOXA2 interacts with the other two transcription factors, although direct physical interactions between the individual proteins were not demonstrated (10, 24). Other homeodomain proteins, such as Otx2 (21), Pdx1 (19), and Engrailed (11), have been found to interact with FOXA2. Whether the homeodomain protein NKX2.1 interacts with the members of the FOXA protein family in regulating lung gene expression and whether these interactions result in stimulation or repression of target genes were questions that were addressed in the current study. The results demonstrate a unique mode of gene regulation involving the two classes of transcription factors whose spatial pattern of expression is reciprocal along the P-D axis of the lung. On NKX2.1 target promoters, such as SpC, which lack FOXA1 binding sites, FOXA1 attenuates the NKX2.1 activity by a mechanism involving DNA-independent protein-protein interactions. The affinity of FOXA1 for its cognate binding site, however, overrides its affinity for protein-protein interaction, and thus, on the Ccsp promoter, which includes both NKX2.1 and FOXA1 binding sites, the two transcription factors act independently to stimulate transcription. These last results provide a potential underlying mechanism by which fine-tuning of gene regulation along the P-D axis of the lung may be accomplished through interactions of homeodomain and forkhead transcription factor gradients.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cell culture and transient-transfection assays. The human pulmonary epithelial cell lines NCI H441 and A549 were maintained in RPMI medium 1640 and F-12K nutrient mixture (Amersham Biosciences, NJ), respectively, containing 10% fetal bovine serum and 1% penicillin-streptomycin. All plasmids used in transfection studies were purified on QIAGEN (CA) columns. Transient transfection of H441 cells and A549 cells was performed with SuperFect as described by the manufacturer (QIAGEN). In brief, cells in 35-mm dishes at 60 to 80% confluence were transfected with 2.25 µg of pSV-ß-gal (Promega, WI), 2.25 µg of test constructs, 2.25 µg of transactivator plasmids, and 15 µl of Superfect in 1 ml serum-free medium. Cells were cultured for 6 h and then changed to regular medium with 10% fetal bovine serum. After 42 h of culture, cells were lysed and the extracts were collected as described by the manufacturer (Promega, WI). Supernatants of the cell extracts were used for assay of ß-galactosidase and luciferase as described by the manufacturer (Promega, WI). MLE15 cells were cultured as described previously (5) and used for extracting nuclear protein for electrophoretic mobility shift assay (EMSA).

Plasmid construction and site-directed mutagenesis. The complete coding region of human Nkx2.1 was PCR amplified and cloned into EcoRI and HindIII sites, in frame with the GAL4 coding region of pM (Clontech, CA), and designated Gal4-Nkx2.1. A similar strategy was used for making GAL4 fusion constructs containing specific domain fragments of Nkx2.1. Gal4NN, Gal4HD, and Gal4NC were constructed by cloning the specific Nkx2.1 fragments that encode amino acids 1 to 141, amino acids 142 to 253, and amino acids 254 to 371 into EcoRI/BamHI, BamHI/MluI, and MluI/HindIII sites of pM, respectively. For glutathione S-transferase (GST) constructs, the same fragments of Nkx2.1 were PCR amplified and cloned into BamHI and EcoRI sites, in frame with the GST coding region, of pGEX-2T (Amersham Biosciences, NJ) and designated GST-Nkx2.1, GSTNN, GSTHD, and GSTNC.

The complete coding region of rat Foxa1 and Foxa2 was PCR amplified from cDNA clones (a gift from Robert Costa, University of Illinois at Chicago) and cloned into EcoRI and BamHI sites, in frame with the VP16 coding region of pVP16 (Clontech, CA), and designated VP16-Foxa1 and VP16-Foxa2. VP16FN, VP16WH, and VP16FC were constructed by cloning the specific Foxa1 fragments that encode amino acids 1 to 137, amino acids 138 to 317, and amino acids 318 to 466 into EcoRI/BamHI sites of pVP16 (Clontech). The Foxa1 and Nkx2.1 expression plasmids were as described previously (16). The SpC-luciferase reporter construct consisting of 320 nucleotides of the murine gene upstream of the transcriptional start site was a gift from Steve Glasser (Children's Hospital Medical Center, Cincinnati, OH) and has been described previously (13). The Ccsp-luciferase reporter construct contains 800 bp of the Ccsp promoter region in the pGL3 plasmid. SpA-Lux, a 1.5-kb SacI fragment of the human SpA gene including the 5'-flanking region, exon I, intron I, exon II, and part of intron II, was cloned into the SacI site of pGL2 basic. The orientation was checked by DNA sequencing.

Oligonucleotides. Synthetic oligonucleotides were annealed and diluted as described previously (4) and were used directly in EMSA as the cold competitor. For use as a probe in EMSA reactions, the annealed oligonucleotides were purified by gel electrophoresis on 3% low-melting-point agarose (Promega), excised, and then eluted using QIAEXII (QIAGEN). Two picomoles of the purified, annealed oligonucleotides were end labeled with T4 polynucleotide kinase and [{gamma}-33P]ATP as described previously (16). The labeled probes were purified from unincorporated [{gamma}-33P]ATP using a G-25 Sephadex column (Roche Applied Science). The DNA sequence of the oligonucleotide was TAGGCCAAGGGCCTTGGGGGCTCT. This sequence includes an NKX2.1 binding site found on the SpC promoter.

EMSA. Nuclear extracts were prepared using a mini-extraction procedure (4). Five micrograms of the nuclear extract was incubated with a 33P-end-labeled oligonucleotide probe with or without the cold competitor in 12.5 mM Tris-HCl (pH 7.5), 62.5 mM NaCl, 0.62 mM dithiothreitol, 10% glycerol, 0.05% NP-40, and 0.05 µg/µl poly(dI-dC) in a total volume of 20 µl at 4°C for 15 min. For experiments with antibody or GST fusion proteins, 4 µl antibody or 4.8 µg GST fusion protein was mixed with the nuclear extracts in a reaction mixture and incubated at 4°C for 15 min before addition of the 33P-end-labeled probe. Bound and free probes were separated by gel electrophoresis on a 4.5% nondenaturing polyacrylamide gel. The NKX2.1 antibodies were purchased from Lab Vision Corporation.

GST assay. GST interaction assays were performed as described previously (16). In brief, GST alone, GST-NKX2.1, or fragments thereof were expressed in E. coli and purified by adsorption to glutathione-Sepharose (Amersham Biosciences). 35S-labeled FOXA1 was prepared by in vitro transcription/translation in the presence of [35S]methionine using the TNT kit (Promega, WI). The Foxa1 template for TNT was PCR amplified with primers of the following sequences: forward, 5'-AATTAACCCTCACTAAAGGGAACAAAGAGCTCGGATGTTAGGGACTGTGAAG 3'; backward, 5'-CCATGCAGACAAACCCTCAGTTCTGGGAGCTAGGAAG-3'. After preincubation with glutathione-Sepharose, the translation mixture containing 35S-labeled FOXA1 was incubated with GST or GST-NKX2.1 proteins adsorbed to the Sepharose beads at 4°C for 1 h in the binding solution (1). The beads were then washed repeatedly, and the associated proteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The gels were then dried and exposed to X-ray films.

Coimmunoprecipitation. Coimmunoprecipitation was performed with protocol and reagents from Active Motif under highly stringent conditions. Briefly, nuclear proteins were extracted from H441 cells and incubated with an anti-NKX2.1 monoclonal antibody (Lab Vision Corporation) overnight. The protein G beads (Roche Applied Science) were washed (Active Motif) and added to the reaction mixture. After incubation for 1 h at 4°C on a rotator, the beads were washed thoroughly and resuspended in 2x reducing loading buffer (Active Motif) for Western blot analysis. As a negative control, a parallel reaction was carried out without the inclusion of the antibody. Western blotting was performed with either anti-NKX2.1 antibody (Lab Vision Corporation) or anti-FOXA1 antibody (CE Mines).

Mammalian two-hybrid assay. Expression constructs for GAL4-Nkx2.1 (1.125 µg) and VP16-Foxa1 or VP16-Foxa2 (5.625 µg) were cotransfected with the GAL4-luciferase reporter constructs pFR-Luc (3.375 µg; Stratagene) and pSV-ß-gal (Promega) into A549 or H441 cells. The interaction between GAL4-Nkx2.1 and VP16-Foxa1 or VP16-Foxa2 was measured as a function of transactivation of the heterologous GAL4 promoter as quantified by luciferase production. Vectors including pM and pVP16 were used in transfection experiments as a control for GAL4-Nkx2.1 and VP16-FoxA1 or VP16-FoxA2, respectively.

In situ hybridization. Antisense RNA probes were prepared with digoxigenin as described previously (15). Whole-mount in situ hybridization with E12 mouse embryonic lungs and in situ hybridization with embryonic lung sections were described previously (15, 17). The cDNA templates for in situ probes were as follows: Foxa1, 0.72-kb 3' untranslated region of Foxa1 amplified by reverse transcription (RT)-PCR from mouse embryonic lung; Foxa2, 0.5 kb of the Foxa2 coding region amplified by RT-PCR from embryonic lung; Ccsp, 0.4 kb of the Ccsp coding region amplified by RT-PCR from embryonic lung; SpC, 0.41-kb DNA fragment containing the 3' end of the SpC coding region and part of the 3' untranslated region amplified by RT-PCR from embryonic lung. All constructs were verified by DNA sequencing. The probe for Nkx2.1 was kindly provided by Shioko Kimora (NCI, NIH, Bethesda, MD).

siRNA transfection. Predesigned small interfering RNA (siRNA) against Foxa1 was synthesized and purified (Ambion). The sequences for targeting (GenBank accession number NM_008259) were as follows: sense, 5'-CGGGUUUCAUUAUUAUUCCtt-3'; antisense, 5'-GGAAUAAUAAUGAAACCCGtt-3'. Silencer Negative Control #1 siRNA (Ambion) was used as a negative control. Transfection of siRNA was performed in duplicate using the manufacturer's protocol for TransMessenger transfection reagent (QIAGEN). The cells were subsequently harvested and used either for Western blot analysis or RNA extraction for real-time PCR as described previously (15). The antibodies for FOXA1 and alpha-tubulin were purchased from CE Mines and Zymed.

Real-time PCR. The quantification of SpC mRNA by real-time PCR was performed using a LightCycler (Roche, Mannheim, Germany) as we have previously described (15). PCRs were recovered and verified by agarose gel electrophoresis. Two ratios were compared: the ratio of SpC to a reference gene (beta-actin) in untreated cells to the same ratio in siRNA-treated samples. The results were expressed as a normalized ratio (see Fig. 8). Sequences of the primers are as follows: beta-actin, 5'-CCAACCGTGAAAAGATGACC-3' (forward) and 5'-CCAGAGGCATACAGGGACAG-3' (backward). SpC, 5'-CAAAAACATACTGAGATGGTGAGTG-3' (forward) and 5'-TCTCTTCCTCCCGAACAGC-3' (backward).


Figure 8
View larger version (30K):
[in this window]
[in a new window]

  		FIG. 8. Inhibition of FOXA1 by siRNA increases SpC mRNA in MLE15 cells. MLE15 cells were treated with siRNA to Foxa1, and the cells were used for SpC mRNA quantification and Western blot analysis. Each experiment was carried out in duplicate (Materials and Methods). Upper panel, a representative SpC mRNA quantification by real-time PCR. Lower panel, Western blot analysis of FOXA1 and alpha-tubulin (control). The two asterisks denote the FOXA1 antibody cross-reactive peptide band.

 

   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
NKX2.1 and FOXA1 interact in a mammalian two-hybrid assay. To determine potential protein-protein interactions between NKX2.1 and FOXA1 and FOXA2, we used two approaches. First, a mammalian two-hybrid assay was used in which the Nkx2.1 cDNA was fused translationally to the DNA binding domain of GAL4 to generate the Gal4-Nkx2.1 plasmid. Two other constructs containing either Foxa1 or Foxa2 cDNAs, translationally fused to the transactivation domain of VP16, produced the plasmids designated Vp16-Foxa1 and Vp16-Foxa2, respectively. Cotransfection using two different lung epithelial A549 and H441 cell lines with the latter plasmids alone and in combination was conducted with a GAL4-responsive plasmid that consists of five tandem GAL4 binding sites cloned upstream of the luciferase reporter gene. Activation of transcription from the GAL4 promoter occurs when the DNA binding domain of GAL4 and the transactivation domain of VP16 are brought together to form a positive transcriptional complex. In turn, the latter occurs if the two chimeric fusion proteins interact. Therefore, luciferase production as measured spectrophotometrically is a reliable indicator of physical interactions between the two fusion proteins. As shown in Fig. 1A and B, the luciferase reading from the control experiments was assigned the value of unity and then used to normalize subsequent cotransfections. These control experiments consisted of the results of cotransfections of Vp16 with Gal4 alone (Fig. 1, bars a) and either Vp16-Foxa1 or Vp16-Foxa2 with Gal4 alone (Fig. 1, bars b). When results were normalized against these controls, cotransfection of the Vp16 plasmid with Gal4-Nkx2.1 showed a significant increase in the GAL4 promoter activity only in H441 cells (Fig. 1A and B, bars c, upper panels). In contrast, cotransfection of identical plasmids led to a decrease (below unity) in the luciferase readout in A549 cells (Fig. 1A and B, bars c, lower panels). Thus, in both A549 and H441 cells, there appear to be background interactions between the VP16 and GAL4-NKX2.1 proteins. Correcting for the latter background, significant activation of the GAL4 promoter was observed when Vp16-Foxa1 or Vp16-Foxa2 was cotransfected with Gal4-Nkx2.1 in either of the two cell lines (Fig. 1A and B, bars d). Even in A549 cells, cotransfection of Vp16-Foxa2 and Gal4-Nkx2.1 gives results onefold above the control level (Fig. 1, lower panels, compare bars c and d). The sum of these data indicates protein-protein interactions between NKX2.1 and both the FOXA1 and FOXA2 proteins.


Figure 1
View larger version (21K):
[in this window]
[in a new window]

  		FIG. 1. NKX2.1 interacts with FOXA1 and FOXA2 transcription factors. Gal4-Nkx2.1 and Vp16-Foxa1 (A) or Vp16-Foxa2 (B) were cotransfected with a Gal4-luciferase reporter construct (Gal4-Lux) into H441 (upper panels) or A549 (lower panels) cells according to the details described in Materials and Methods. A plasmid, pSV-ß-gal (Promega, WI), was used for normalization of transfection efficiency. In addition, Vp16 and Gal4 (pM) plasmids were used as a control. Luciferase values obtained for Vp16 and Gal4 alone or Vp16-Foxa1 in combination with Gal4 alone were adjusted to unity to normalize all other experimental values. Data represent results from at least three independent experiments.

 
FOXA1 and FOXA2 are nearly 93% identical in the winged/helix DNA binding domain (8), and hence in the remainder of our studies we focused on FOXA1 alone. To further establish the physical interactions between NKX2.1 and the FOXA1 protein, we investigated the precise domains on the respective proteins through which the interactions occurred. The human NKX2.1 protein can be divided into three specific functional domains (Fig. 2A). The N-terminal domain, designated NN, includes the Tinman homology domain and consists of amino acids 1 to 141. The DNA binding activity of NKX2.1 resides within the homeodomain that is internal to a fragment designated HD, which consists of amino acids 142 to 253. The C-terminal domain (NC) includes the NK2-specific domain and spans amino acids 254 to 371 (Fig. 2A). Specific fragments of the Nkx2.1 cDNA encoding each of the three domains were cloned downstream of the GAL4 coding sequence to generate three constructs, designated Gal4NN, Gal4HD, and Gal4NC (Fig. 2A). We also generated glutathione S-transferase fusion polypeptides of each fragment for subsequent studies (see below).


Figure 2
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 2. The homeodomain of NKX2.1 interacts physically with FOXA1. A. A simple map of the various functional domains of the human NKX2.1 protein. TN, Tinman domain; SD, NK2-specific domain. The hatched box represents the homeodomain or HD. Numbers indicate amino acids from the first ATG codon. Both GST- and GAL4-fused constructs are shown. B. Mammalian two-hybrid assay was conducted with A549 cells using VP16-fused FOXA1 and GAL4-fused peptide fragments of NKX2.1. pSV-ß-gal (Promega) was cotransfected in all experiments and used to normalize transfection efficiency. A Gal4-luciferase reporter construct (Gal4-Lux) was used for measurement of physical association between FOXA1 and various domains of NKX2.1. The first two reactions, "a" and "b," consisting of Vp16/Gal4 and Vp16-Foxa1/Gal4 combinations, were adjusted to unity and used for normalization of all other experimental values. Data represent results of at least three independent experiments.

 
Similarly, the human FOXA1 protein can be divided into three polypeptides representing distinct functional domains (Fig. 3A). The N-terminal end of FOXA1 includes the conserved transcriptional activation domain and consists of amino acids 1 to 137 (8). A second fragment includes amino acids 138 to 317 and contains the winged helix/forkhead DNA binding domain, conserved between FOXA1 and FOXA2 (8). The third fragment represents amino acids 318 to 466 and includes the C-terminal highly conserved regions II and III that are essential transactivation domains of FOXA1. Translational fusion of these last three fragments to VP16 generated VP16FN, VP16WH, and VP16FC, respectively (Fig. 3A).


Figure 3
View larger version (17K):
[in this window]
[in a new window]

  		FIG. 3. The Forkhead DNA binding domain of FOXA1 interacts with NKX2.1. A. A simple map of the various functional domains of the human FOXA1 protein. TAD, transactivation domain; IV/V, II, and III are various domains as described by Costa et al. (8). The hatched box represents the highly conserved Winged Helix-Forkhead DNA binding domain. Numbers indicate amino acids from the first ATG codon. The VP16-fused constructs are also shown. B. A mammalian two-hybrid assay was conducted with A549 cells using a Gal4-Nkx2.1 fragment in combination with a Vp16-Foxa1 fragment. pSV-ß-gal (Promega) was cotransfected in all experiments and used to normalize transfection efficiency. A Gal4-luciferase reporter construct (Gal4-Lux) was used for measurement of physical association between NKX2.1 and various FOXA1 fragments. The control reactions, consisting of Gal4/Vp16 (bar "a"), Gal4/Vp16-Foxa1, Gal4/Vp16FN, Gal4/Vp16WH, and Gal4/Vp16FC combinations (not shown), were adjusted to unity and used for normalization of all other experimental values. Data represent results of at least three independent experiments.

 
The mammalian two-hybrid assay with A549 cells was then used to test the ability of each NKX2.1 domain to interact with FOXA1. Conversely, FOXA1 domains were used to examine their potential interactions with NKX2.1. In the first set of experiments (Fig. 2B), only the GAL4HD peptide was found to interact with the VP16-FOXA1 protein, as evidenced by a nearly threefold increase in luciferase production from the GAL4 promoter (Fig. 2B, compare g and h). Neither the N-terminal (Fig. 2B, compare e and f) nor the C-terminal (Fig. 2B, compare i and j) peptides of NKX2.1 showed appreciable interactions with FOXA1. Similarly, the interaction of NKX2.1 with various functional domains of FOXA1 was examined with A549 cells (Fig. 3B). Here, we found interactions between NKX2.1 and the winged helix/forkhead DNA binding domain of the latter transcription factor (Fig. 3B, compare b and e). Interaction of NKX2.1 with functional domains of FOXA1was not as robust as its interaction with the intact FOXA1 protein (Fig. 3B, compare c and e), suggesting that other parts of FOXA1 may be required for optimal interaction. No interaction was observed between NKX2.1 and either the C-terminal or N-terminal domain of the FOXA1 protein.

Direct physical interaction between NKX2.1 and FOXA1. Direct physical interaction between the NKX2.1 and FOXA1 proteins was further examined by two approaches. First, we used a GST pull-down assay to examine the binding of GST-fused full-length NKX2.1 and its truncated domains (Fig. 2A) to in vitro-synthesized, [35S[methionine-labeled FOXA1. The results showed that FOXA1 forms protein-protein complexes with GST-NKX2.1 (Fig. 4, lane 4b). Also, there is clearly an interaction between the GST-fused HD of NKX2.1 and FOXA1 (Fig. 4, lane 6b). Consistent with the results from the two-hybrid assays shown in Fig. 2B, neither the N-terminal (Fig. 4, lane 5b) nor C-terminal (lane 7b) domain of NKX2.1 interacted with FOXA1. The control experiment showed that in vitro-synthesized FOXA1 does not associate with an excess amount of the GST peptide alone (no NKX2.1) (Fig. 4, lanes 3a and 3b).


Figure 4
View larger version (18K):
[in this window]
[in a new window]

  		FIG. 4. Physical interactions between NKX2.1 and FOXA1 in a GST pull-down assay. GST alone, GST-NKX2.1, or fragments thereof were expressed in E. coli as fusion proteins and purified by adsorption to glutathione-Sepharose (Amersham Biosciences). 35S-labeled FOXA1 was prepared by in vitro transcription/translation (TNT) in the presence of [35S]methionine. GST alone or GST-NKX2.1 fused proteins adsorbed to glutathione-Sepharose were incubated with TNT 35S-labeled FOXA1, and the Sepharose beads were washed repeatedly. The associated proteins were analyzed by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Panel A is the Coomassie blue-stained polyacrylamide gel. Panel B shows the autoradiogram of the same gel. Lane 1, molecular weight markers, given in thousands. Lanes 2 through 7 are the eluates of mixtures of 35S-labeled FOXA1 alone (lanes 2a and 2b) or in combination with either full-length NKX2.1 (4a and 4b) or fragments of the NKX2.1 protein (5a though 7a and the corresponding lanes 5b through 7b). Lanes 3a and 3b show the GST control. Please note the excess quantity of GST protein (lane 3a) and its absence of association with 35S-labeled FOXA1 (lane 3b). Arrows in panel A point to the sizes of the GST-fused polypeptides synthesized in E. coli. Lane 4a, GST-NKX2.1; lane 5a, GST-NN; lane 6a, GSTHD; lane 7a, GST-NC. The position of the radioactively labeled FOXA1 protein is indicated in panel B.

 
In a second approach, we used nuclear extracts from H441 cells to immunoprecipitate NKX2.1 and any associated proteins. The immunoprecipitated material was then analyzed by using duplicate Western blots, probed with either anti-NKX2.1 or anti-FOXA1 antibody. Figure 5 shows that FOXA1 is indeed a part of a complex that is immunoprecipitated by the anti-NKX2.1 antibody, demonstrating a close physical association between the two transcription factors in vivo. Thus, the results of these last two independent approaches clearly demonstrate physical interactions between NKX2.1 and FOXA1 proteins.


Figure 5
View larger version (11K):
[in this window]
[in a new window]

  		FIG. 5. Coimmunoprecipitation of NKX2.1 and FOXA1. Nuclear proteins from H441 cells were immunoprecipitated with (lane B) and without (lane A) anti-NKX2.1 antibody. The immunoprecipitation complexes were analyzed by Western blotting, using specific antibodies (see Materials and Methods) for the presence of NKX2.1 or FOXA1 (arrows).

 
Functional consequences of NKX2.1-FOXA1 interactions. As a first step in examining the functional significance of NKX2.1-FOXA1 interactions, we examined their spatial pattern of distribution in two progressive stages of murine embryonic lung development by in situ hybridization (Fig. 6). The mRNA for the two transcription factors showed a reciprocal pattern of spatial distribution along the P-D axis of the embryonic lung. Nkx2.1 mRNA is expressed along the P-D axis of the lung, with the highest level of expression at the tip of the branching airways in the distal lung (Fig. 6A and B). This spatial pattern of NKX2.1 is even more clearly demonstrated by immunohistochemistry using a commercial anti-NKX2.1 antibody (32). Low-level expression of Nkx2.1, can be readily discerned in the epithelium of the trachea, bronchi, and the proximal airways (Fig. 6B). In contrast, Foxa1 expression is highest in the tracheal, bronchial, and bronchiolar epithelium as well as in the proximal airways (Fig. 6C and D) and is reduced to negligible levels in the distal lung (Fig. 6D). Two other genes with a similar reciprocal P-D pattern of expression in the lung are SpC and Ccsp. Figure 6 also shows the reciprocal expression pattern that is observed in spatial distribution of the latter genes. A gradient of SpC expression occurs along the P-D axis of the lung (Fig. 6E), with the highest level found in the distal airways (Fig. 6E and F), overlapping with Nkx2.1. The onset of Ccsp expression occurs late; thus, there is an absence of signal in panel G. However, in E18 lungs, Ccsp is confined to the proximal lung (Fig. 6H). Whether NKX2.1-FOXA1 interaction bears functional significance for regulating the observed proximo-distally localized SpC and Ccsp gene expression was examined next.


Figure 6
View larger version (71K):
[in this window]
[in a new window]

  		FIG. 6. Spatial distribution of Nkx2.1, Foxa1, SpC, and Ccsp gene expression. Spatial distribution of Nkx2.1, Foxa1, SpC and Ccsp mRNA in embryonic day 12 (E12) (A to D) or E18 (E to H) lungs was determined by in situ hybridization. Note that Ccsp is expressed only in the E18 stage of lung development. Scale bar, 220 µm (A, C, E, and G) or 100 µm (B, D, F, and H). Arrowheads indicate branching airways in the distal lung. Arrows indicate trachea, bronchi, and proximal airways.

 
FOXA1 attenuates NKX2.1 activity on the SpC promoter both in vitro and in vivo. Based on the reciprocal pattern of gene expression observed in Fig. 6, we examined the functional impact of combinations of the transcription factors NKX2.1 and FOXA1 and each separately on activation of promoter/enhancer regions of SpC and Ccsp. The promoter/enhancer DNA of both target genes contains functional binding sites for NKX2.1. However, whereas the Ccsp promoter includes specific binding sites for FOXA proteins, no such site, based on DNA sequence, was found on the SpC gene. In A549 cells, which exhibit no endogenous Nkx2.1 gene expression, NKX2.1 increased the activities of the Ccsp and SpC promoters by 8.2-fold and 3-fold, respectively (Fig. 7). As expected with the absence of any known binding sites for FOXA1, the SpC promoter did not respond to the cotransfected cmv-Foxa1 plasmid (panel A, bar c), whereas there was a five-fold induction of Ccsp promoter activity by cmv-Foxa1 (panel B, bar c). Cotransfection with a combination of both transcription factors showed an additive effect on Ccsp (panel B, bar d). In contrast, and unexpectedly, cotransfection of cmv-Nkx2.1 and cmv-Foxa1 resulted in attenuation of NKX2.1-stimulated SpC transcription (panel A, bar d). This last finding was precisely reproducible by multiple repetitions of the experiment using the same or different batches of reagents. In addition, similar if not identical results were obtained in repeated experiments using H441 cells that exhibit endogenous Nkx2.1 gene expression (Fig. 7C and D).


Figure 7
View larger version (25K):
[in this window]
[in a new window]

  		FIG. 7. FOXA1 attenuates NKX2.1 stimulation of SpC promoter activity. Cotransfection of either the SpC promoter-luciferase or Ccsp promoter-luciferase was conducted with cmv-Nkx2.1, cmv-Foxa1, or a combination of the two in A549 (A and B) or H441 (C and D) lung carcinoma cell lines. Relative luciferase values were normalized to beta-galactosidase activity as described in Materials and Methods and plotted. Data represent results from at least three independent experiments.

 
To ascertain the in vivo validity of the above results, we examined the consequences of siRNA-mediated inhibition of Foxa1 on SpC mRNA in MLE15 cells. Treatment of MLE15 cells with siRNA effectively inhibited FOXA1 protein to a level nearly undetectable by Western analysis. As an internal control, an unknown protein that cross-reacts with the FOXA1 antibody (Fig. 8) remained intact, as did the level of alpha-tubulin. In association with FOXA1 inhibition, we found nearly a doubling of the steady-state level of SpC mRNA by real-time PCR (Fig. 8, upper panel). Since transcription of SpC in vivo is strictly dependent on NKX2.1 (20), the latter results are consistent with the model of FOXA1 attenuation of NKX2.1-driven SpC transcription.

Dose-dependent and DNA-independent attenuation of NKX2.1 activity by FOXA1. To further substantiate the latter findings, we examined whether FOXA1 can attenuate NKX2.1 activation of the SpC promoter in a dose-dependent manner. In these experiments, a constant amount of the cmv-Nkx2.1 expression plasmid was cotransfected with increasing molar ratios of the cmv-Foxa1 plasmid, using the SpC-luciferase construct as a reporter. All experiments included equal total quantities of cmv plasmid to avoid potential problems with transcription factor titration ("squelching") by cmv sequences. These studies showed a clear dose-dependent response in attenuation of NKX2.1 activity by FOXA1. At the highest dose used in these studies (5x), FOXA1 entirely repressed NKX2.1-dependent transcriptional stimulation of the SpC promoter activity (Fig. 9, compare bars c, b, and f). The sum of the results from these last studies show that FOXA1 can interact with NKX2.1 and attenuate its stimulatory role on transcriptional activation of the SpC promoter, which includes NKX2.1 but lacks FOXA1 binding sites. A corollary to this observation is that attenuation of NKX2.1 activity by FOXA1 may be DNA independent.


Figure 9
View larger version (17K):
[in this window]
[in a new window]

  		FIG. 9. Dose-dependent attenuation of NKX2.1 by FOXA1. Cotransfection of the SpC promoter-luciferase (SpC-Lux) was conducted with a fixed amount of cmv-Nkx2.1 plus increasing molar ratios of cmv-Foxa1 in A549 lung carcinoma cell lines. Each reaction mixture contains an equal total amount of cmv promoter DNA by addition of a vector plasmid of the cmv-Foxa1 construct. Relative luciferase values were normalized to beta-galactosidase activity as described in Materials and Methods and plotted. Data represent the results of at least three independent experiments.

 
A model of DNA-independent inhibition of NKX2.1 activity by FOXA1 predicts that the stimulatory impact of NKX2.1 would be abrogated on other promoters that lack FOXA binding sites. To examine the validity of this model, we tested the impact of excess FOXA proteins on NKX2.1-dependent activation of transcription from a 1.5-kb fragment of the lung-enriched Surfactant Protein A (SpA) gene promoter, which includes NKX2.1 binding sites but lacks those for FOXA. In cotransfection studies with A549 cells, NKX2.1 increased baseline expression of the SpA-luciferase construct by 6.5-fold (Fig. 10, compare bars a and b). The impact of either FOXA1 or FOXA2 alone on SpA transcription was inhibitory (bars c and d, respectively). Nevertheless, NKX2.1 stimulation of SpA transcription decreased to 4.2-fold when cmv-Foxa1 was added to the transfection at a ratio to Nkx2.1 of one to one (Fig. 10, compare c to e) and further to 3.5-fold in the presence of cotransfected cmv-Foxa2 plasmid (Fig. 10, compare bar d to bar f). These observations provide further support for a model in which FOXA proteins attenuate NKX2.1 activity on lung gene promoters that lack FOXA binding sites in a DNA-independent manner.


Figure 10
View larger version (13K):
[in this window]
[in a new window]

  		FIG. 10. FOXA1 and FOXA2 attenuate NKX2.1 stimulation of SpA promoter activity. Cotransfection of the human SpA promoter-luciferase (SpA-Lux) was conducted with cmv-Nkx2.1, cmv-Foxa1, cmv-Foxa2, or a combination in A549 lung carcinoma cell lines. Relative luciferase values were normalized to beta-galactosidase activity as described in Materials and Methods and plotted. Data represent results from at least three independent experiments.

 
GST-FOXA1 interferes with the binding of NKX2.1 to the SpC promoter. The results from the three independent mammalian two-hybrid, GST-pulldown, and coimmunoprecipitation assays showed that FOXA1 interacts with the homeodomain (HD) of NKX2.1 (Fig. 2B and 4). These findings suggest that attenuation of NKX2.1 activity by FOXA1 may occur through interference with the NKX2.1 DNA binding ability and/or destabilization of a transcriptional complex on the SpC promoter. To examine the latter, we conducted EMSA in the presence and absence of recombinant GST-FOXA1 protein. Increasing amounts of GST-FOXA1 fusion protein or GST alone (control) was used in EMSA to determine the impact of FOXA1 on the binding of NKX2.1 to its cognate site on the SpC promoter. Figure 11B shows a representative result of this analysis. Using nuclear extracts from MLE15 cells, which provide an enriched source of NKX2.1, one minor and two major nucleoprotein complexes were detected that bound the probe. These were designated a, b, and c in Fig. 11B, lane 1. Nonradioactive oligonucleotide probes included in the nuclear extract competed against all three complexes (panel B, lane 2), demonstrating the specificity of binding. In addition, inclusion of an anti-NKX2.1 antibody supershifted a nucleoprotein complex (Fig. 11B, lane 3). Of significance, inclusion of bacterially expressed GST-FOXA1 recombinant protein in an increasing gradient (0.08 to 8.0 µg/ml) (Fig. 11, lanes 4, 5, and 6) reduced the overall binding of all nuclear proteins. At 8.0 µg/ml, GST-FOXA1 robustly inhibited formation of all three complexes, including "b," but entirely blocked the formation of complex "c" (Fig. 11, lane 6). Although an identical increasing gradient of GST alone reduced the overall binding of nucleoproteins, this effect was minimal and nonspecific (Fig. 11, lanes 7, 8, and 9). These data indicate that DNA-independent interactions between FOXA1 and NKX2.1 result in disruption of nucleoprotein moieties that may be necessary for the formation of a functional and efficient transcriptional complex on the SpC promoter.


Figure 11
View larger version (98K):
[in this window]
[in a new window]

  		FIG. 11. FOXA1 interferes with the DNA binding ability of NKX2.1. A. Western blot analysis of purified recombinant GST-FOXA1 protein (arrow) synthesized in E. coli and used in panel B (lane 4). B. An electrophoretic mobility shift assay was performed with nuclear extract from MLE 15 cells. The probe was an oligonucleotide representing the NKX2.1 DNA-binding sequence found on the SpC promoter (please see Materials and Methods). The nucleoprotein complexes formed on the target oligonucletoide are designated "a," "b," or "c" (lane 1). The position of the supershifted band in the presence of anti-NKX2.1 antibody is shown by the arrowhead. The arrow points to nucleoprotein complex "c," which is specifically destabilized in the presence of the GST-FOXA1 protein (lane 4). Triangles represent increasing (from 0.08 µg/ml to 8 µg/ml) total recombinant protein amounts of GST-FOXA1 (lanes 4, 5, and 6) or GST alone (lanes 7, 8, and 9) added to individual reaction mixtures.

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The purpose of the current study was to examine potential interactions between the homeodomain, NKX2.1, and the forkhead/winged helix FOXA1 transcription factors. Using independent experimental approaches, it is clearly established that the NKX2.1 and FOXA1 proteins interact both physically and functionally in a DNA-independent manner to bring about what may represent a novel strategy for gene regulation along the P-D axis of the lung. On the proximally expressed Ccsp promoter, which includes binding sites for both, the impact of NKX2.1 and FOXA1 is additive. We propose that the latter is due to a preferential affinity of FOXA1 for its DNA binding site. In contrast, FOXA1 attenuates the stimulatory impact of NKX2.1 on the distally active SpC promoter, which includes NKX2.1 but not FOXA1 binding sites or response. Consistent with a model of transcriptional gradient gene regulation, we and others (3) found that the level and spatial distribution of FOXA1 and NKX2.1 are reciprocal along the P-D axis of the lung. These findings have functional implications regarding the mechanisms of selective gene regulation in specialized lung epithelial cell types in particular and a mechanism of gene regulation by gradients of interacting transcription factors in general.

Protein-protein interactions between transcription factors and accessory proteins are critical in the formation of a functional transcriptional assembly (29). For example, the DNA binding affinity of Msx2 is enhanced through direct protein-protein interactions with Miz1, a zinc finger-containing, DNA binding protein (30). While most interactions are thought to have a positive impact, some result in inhibition of transcription. Thus, GATA1 can interact through its N-zinc finger domain with STAT3 to inhibit its DNA binding activity in normal hematopoietic cells (9).

The homeodomain transcription factor NKX2.1 is expressed in thyroid, brain, and lung tissue. In these tissues, NKX2.1 not only directs architectural development of the organs but also controls tissue-specific gene regulation. Because the genes activated by NKX2.1 vary from organ to organ, it is rational to speculate that specificity may be accomplished through unique interactions with other nuclear and transcription factors. In the lung, NKX2.1 is known to interact physically and functionally with retinoic acid receptors (31) and SMAD3 (16). In each case these interactions modify and at times specify the precise function of NKX2.1 on its target genes. Whereas some may be stimulatory, other interactions attenuate the activity of NKX2.1 (16).

A number of Fox genes are known to be expressed in the lung (8). In embryonic lungs, temporal and spatial expression patterns of Foxa1 and Foxa2 are nearly identical (data not shown). The promoter/enhancer domains of a number of lung-specific or enriched genes, such as Ccsp (22), contain functional FOXA binding sites. In contrast, the SpC promoter is devoid of such sites, and hence its transcription is not stimulated by FOXA proteins.

Because their binding sites are present on the promoters of some lung-expressed genes, we examined the hypothesis that NKX2.1 and FOXA proteins may interact in regulating lung epithelial gene expression. In a mammalian two-hybrid assay, cotransfection of Vp16-Foxa1 with Gal4 alone had no effect on luciferase production (Fig. 1). However, only in H441 cells, a combination of Gal4-Nkx2.1 and Vp16 caused a 13.17-fold increase in luciferase production (Fig. 1A, c). This observation suggests one of two possibilities, either that NKX2.1 as part of the GAL4-NKX2.1 fusion protein provides some DNA binding activity and/or, alternatively, that it interacts with VP16. Despite this background in H441 cells, physical interaction between NKX2.1 and FOXA proteins was documented in both H441 and A549 cells, since it resulted in measurable activation of the GAL4-Luc construct (Fig. 1A and B). Further studies with domain-specific GAL4 fusion polypeptides of NKX2.1 showed that the interactions between NKX2.1 and FOXA1 occur through the DNA-binding homeodomain of NKX2.1. In previous work, the HD of NKX2.1 was shown to interact with calreticulin to activate thyroglobulin (25). NKX2.5 and NKX3.1, two other members of the NK family, also interact with nuclear proteins through their HD (6, 7).

To ascertain the functional implications of NKX2.1-FOXA1 interactions for lung genes, we selected promoter/enhancer constructs from two genes, SpC and Ccsp. This selection was based on two criteria. First, although both SpC and Ccsp promoters contain NKX2.1 binding sites, the SpC promoter used in our studies lacks a FOXA1 binding site, allowing it to serve the equivalent role of a FOXA1 deletion mutant in a natural promoter context. This circumvented any potential problems arising from introduction of mutations or large deletions that may otherwise alter promoter activity in an unforeseeable manner. Second, Ccsp and SpC expression domains are confined to two distinct cell types along the P-D axis of the lung. In situ hybridization showed a gradient pattern of distribution of Nkx2.1 and FoxA gene expression that was mirrored by the target genes, SpC and Ccsp (Fig. 6). Based on this information, we used simple cotransfection studies. In A549 cells, in which endogenous Foxa1 but not Nkx2.1 is expressed, the Ccsp and SpC promoter activities increased in response to NKX2.1 by 8.2-fold and 3-fold, respectively (Fig. 7). FOXA1 stimulated Ccsp promoter activity by 4.4-fold but had no effect on the SpC promoter. Significantly, FOXA1 attenuated the stimulatory effect of NKX2.1 on the SpC promoter. The SpC promoter/enhancer element in our study lacks FOXA1 binding sites, consistent with the absence of a transcriptional response to FOXA1 in either H441 or A549 cells. Thus, the attenuating impact of FOXA1 represents an indirect effect of this transcription factor on the SpC promoter. In the current work, we used multiple independent approaches, including two-hybrid and dose-response assays, to show that this indirect impact of FOXA1 may be explained by robust protein-protein interaction with NKX2.1. The latter interactions also suggested a model in which FOXA1 attenuates NKX2.1 activity in a DNA-independent manner. From a mechanistic viewpoint, the model predicted that FOXA1 should attenuate the effect of NKX2.1 on other promoters that do not contain a FOXA binding site(s). This is precisely what was observed in the studies on the SpA promoter response shown in Fig. 10.

The observation that a combination of NKX2.1 and FOXA1 showed an additive rather than synergistic impact on stimulation of Ccsp transcription (Fig. 6) suggests that NKX2.1 and FOXA1 do not interact on the Ccsp promoter. Instead, it appears that the two transcription factors act individually at their own binding sites. It is possible that the affinity of FOXA1 for its cognate DNA binding site on Ccsp may be greater than its affinity for NKX2.1. Thus, on a promoter that includes its binding site (e.g., Ccsp), FOXA1 preferentially binds to such a site and activates transcription. In contrast, on a promoter that lacks the FOXA1 binding site, NKX2.1-FOXA1 interactions predominate, leading in the case of SpC to attenuation of transcription, as we have observed in the current study. Furthermore, the finding that NKX2.1-FOXA1 interactions occur through the HD suggested that FOXA1 may interfere with the binding affinity of NKX2.1 for its cognate binding site on the SpC promoter. The validity of this hypothesis was examined by EMSA, which showed that a GST-FOXA1 protein specifically, reproducibly, and in a dose-dependent manner inhibited the formation of NKX2.1 nucleoprotein complexes on an SpC promoter oligonucleotide (Fig. 11).

The current study is the first to report DNA-independent physical interactions between NKX2.1 and FOXA1 transcription factors both in vitro and in vivo. These interactions are further shown to have functional consequences for regulation of SpC and Ccsp genes, whose expression occurs with precise P-D specificity in the lung, where the two transcription factors are also present in a reciprocal gradient. Although it is certain that establishment of a P-D gene expression pattern involves multiple factors with highly complex interactions, the findings described here regarding NKX2.1-FOXA1 interactions in a spatially selective manner may at least in part contribute to this process. Compared to the distribution of SpC and Ccsp, it is clear that the highest levels of FOXA1 are found in the proximal lung, where maximal Ccsp expression occurs. NKX2.1, a strong stimulator of both Ccsp and SpC, is expressed maximally in the distal epithelium and at low levels in the proximal lung. We propose that it is possible that a spatially selective SpC versus Ccsp pattern of gene expression is established, at least partly through NKX2.1-FOXA1 interactions in a P-D gradient. It is imperative to note that the proposed model is simplistic and may fall short of providing a mechanistic explanation for patterns of all gene expression along the P-D axis of the lung epithelium. For example, SpB is expressed in both proximal and distal lung epithelium, even though its promoter includes both FOXA1 and NKX2.1 binding sites (4). Although evidence for direct stimulation of SpB by FOXA1 is scant, the complexity of its promoter may necessitate invoking tertiary cis-acting or trans-acting factors beyond NKX2.1 and FOXA1 in its regulation. Despite this caveat, the gradient-dependent interaction between forkhead and homeodomain transcription factors may represent a novel strategy of gene regulation with implications pertaining to other organs, such as the brain and thyroid, where NKX2.1 and other members of the forkhead family of transcription factors are also expressed. The concept of spatially specific gene regulation through establishment of a transcription factor gradients applies to many biological processes, particularly the early events in development. In the Drosophila embryo, maternal transcription factors, such as Dorsal, are localized in a highly spatially predetermined manner. Gradients of Dorsal are known to be critical in activation of zygotic polarity genes in a spatially restricted manner (23). Regulation of proximal versus distal gene expression by DNA-independent protein-protein interactions of reciprocal gradients of homeodomain and forkhead transcription factors is a novel finding that may represent yet another variation on the overall mechanistic theme of spatially restricted gene regulation in metazoans.


   ACKNOWLEDGMENTS
 
We are grateful for the gift of plasmids provided by Robert Costa (University of Illinois, Chicago, IL), Shioko Kimura (NCI, NIH, Bethesda, MD), and Steve Glasser (Children's Hospital Medical Center, Cincinnati, OH).

This work was supported by NIH and The Hastings Foundation.


   FOOTNOTES
 
* Corresponding author. Mailing address: General Laboratories Building, 1801 E. Marengo Street, Room 1G1, Los Angeles, CA 90033. Phone: (323) 226-4340. Fax: (323) 226-5049. E-mail: minoo{at}usc.edu. Back

{triangledown} Published ahead of print on 12 January 2007. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 
1. Alliston, T., L. Choy, P. Ducy, G. Karsenty, and R. Derynck. 2001. TGF-beta-induced repression of CBFA1 by Smad3 decreases cbfa1 and osteocalcin expression and inhibits osteoblast differentiation. EMBO J. 20:2254-2272.[CrossRef][Medline]

2. Ang, S. L., and J. Rossant. 1994. HNF-3 beta is essential for node and notochord formation in mouse development. Cell 78:561-574.[CrossRef][Medline]

3. Besnard, V., S. E. Wert, W. M. Hull, and J. A. Whitsett. 2004. Immunohistochemical localization of Foxa1 and Foxa2 in mouse embryos and adult tissues. Gene Expr. Patterns 5:193-208.[CrossRef][Medline]

4. Bohinski, R. J., R. Di Lauro, and J. A. Whitsett. 1994. The lung-specific surfactant protein B gene promoter is a target for thyroid transcription factor 1 and hepatocyte nuclear factor 3, indicating common factors for organ-specific gene expression along the foregut axis. Mol. Cell. Biol. 14:5671-5681.[Abstract/Free Full Text]

5. Borok, Z., X. Li, V. J. Fernandes, B. Zhou, D. K. Ann, and E. D. Crandall. 2000. Differential regulation of rat aquaporin-5 promoter/enhancer activities in lung and salivary epithelial cells. J. Biol. Chem. 275:26507-26514.[Abstract/Free Full Text]

6. Chen, C. Y., and R. J. Schwartz. 1996. Recruitment of the tinman homolog Nkx-2.5 by serum response factor activates cardiac alpha-actin gene transcription. Mol. Cell. Biol. 16:6372-6384.[Abstract]

7. Chen, H., and C. J. Bieberich. 2005. Structural and functional analysis of domains mediating interaction between NKX-3.1 and PDEF. J. Cell Biochem. 94:168-177.[CrossRef][Medline]

8. Costa, R. H., V. V. Kalinichenko, and L. Lim. 2001. Transcription factors in mouse lung development and function. Am. J. Physiol. Lung Cell Mol. Physiol. 280:L823-L838.[Abstract/Free Full Text]

9. Ezoe, S., I. Matsumura, K. Gale, Y. Satoh, J. Ishikawa, M. Mizuki, S. Takahashi, N. Minegishi, K. Nakajima, M. Yamamoto, T. Enver, and Y. Kanakura. 2005. GATA transcription factors inhibit cytokine-dependent growth and survival of a hematopoietic cell line through the inhibition of STAT3 activity. J. Biol. Chem. 280:13163-13170.[Abstract/Free Full Text]

10. Filosa, S., J. A. Rivera-Perez, A. P. Gomez, A. Gansmuller, H. Sasaki, R. R. Behringer, and S. L. Ang. 1997. Goosecoid and HNF-3beta genetically interact to regulate neural tube patterning during mouse embryogenesis. Development 124:2843-2854.[Abstract]

11. Foucher, I., M. L. Montesinos, M. Volovitch, A. Prochiantz, and A. Trembleau. 2003. Joint regulation of the MAP1B promoter by HNF3beta/Foxa2 and Engrailed is the result of a highly conserved mechanism for direct interaction of homeoproteins and Fox transcription factors. Development 130:1867-1876.[Abstract/Free Full Text]

12. Kaestner, K. H., J. Katz, Y. Liu, D. J. Drucker, and G. Schutz. 1999. Inactivation of the winged helix transcription factor HNF3alpha affects glucose homeostasis and islet glucagon gene expression in vivo. Genes Dev. 13:495-504.[Abstract/Free Full Text]

13. Kelly, S. E., C. J. Bachurski, M. S. Burhans, and S. W. Glasser. 1996. Transcription of the lung-specific surfactant protein C gene is mediated by thyroid transcription factor 1. J. Biol. Chem. 271:6881-6888.[Abstract/Free Full Text]

14. Kimura, S., Y. Hara, T. Pineau, P. Fernandez-Salguero, C. H. Fox, J. M. Ward, and F. J. Gonzalez. 1996. The T/ebp null mouse: thyroid-specific enhancer-binding protein is essential for the organogenesis of the thyroid, lung, ventral forebrain, and pituitary. Genes Dev. 10:60-69.[Abstract/Free Full Text]

15. Li, C., L. Hu, J. Xiao, H. Chen, J. T. Li, S. Bellusci, S. Delanghe, and P. Minoo. 2005. Wnt5a regulates Shh and Fgf10 signaling during lung development. Dev. Biol. 287:86-97.[CrossRef][Medline]

16. Li, C., N. L. Zhu, R. C. Tan, P. L. Ballard, R. Derynck, and P. Minoo. 2002. TGF-beta inhibits pulmonary surfactant protein-B gene transcription through SMAD3 interactions with NKX2.1 and HNF-3 transcription factors. J. Biol. Chem. 277:38399-38408.[Abstract/Free Full Text]

17. Li, M., X. Wu, F. Zhuang, S. Jiang, M. Jiang, and Y. H. Liu. 2003. Expression of murine ELL-associated factor 2 (Eaf2) is developmentally regulated. Dev. Dyn. 228:273-280.[CrossRef][Medline]

18. Liu, C., S. W. Glasser, H. Wan, and J. A. Whitsett. 2002. GATA-6 and thyroid transcription factor-1 directly interact and regulate surfactant protein-C gene expression. J. Biol. Chem. 277:4519-4525.[Abstract/Free Full Text]

19. Marshak, S., E. Benshushan, M. Shoshkes, L. Havin, E. Cerasi, and D. Melloul. 2000. Functional conservation of regulatory elements in the pdx-1 gene: PDX-1 and hepatocyte nuclear factor 3ß transcription factors mediate ß-cell-specific expression. Mol. Cell. Biol. 20:7583-7590.[Abstract/Free Full Text]

20. Minoo, P., G. Su, H. Drum, P. Bringas, and S. Kimura. 1999. Defects in tracheo-esophageal and lung morphogenesis in Nkx2.1(–/–) mouse embryos. Dev. Biol. 209:60-71.[CrossRef][Medline]

21. Nakano, T., T. Murata, I. Matsuo, and S. Aizawa. 2000. OTX2 directly interacts with LIM1 and HNF-3beta. Biochem. Biophys. Res. Commun. 267:64-70.[CrossRef][Medline]

22. Nord, M., T. N. Cassel, H. Braun, and G. Suske. 2000. Regulation of the Clara cell secretory protein/uteroglobin promoter in lung. Ann. N. Y. Acad. Sci. 923:154-165.[Abstract/Free Full Text]

23. Papatsenko, D., and M. Levine. 2005. Quantitative analysis of binding motifs mediating diverse spatial readouts of the Dorsal gradient in the Drosophila embryo. Proc. Natl. Acad. Sci. USA 102:4966-4971.[Abstract/Free Full Text]

24. Perea-Gómez, A., W. Shawlot, H. Sasaki, R. R. Behringer, and S. Ang. 1999. HNF3beta and Lim1 interact in the visceral endoderm to regulate primitive streak formation and anterior-posterior polarity in the mouse embryo. Development 126:4499-4511.[Abstract]

25. Perrone, L., G. Tell, and R. Di Lauro. 1999. Calreticulin enhances the transcriptional activity of thyroid transcription factor-1 by binding to its homeodomain. J. Biol. Chem. 274:4640-4645.[Abstract/Free Full Text]

26. Sun, T., Y. Echelard, R. Lu, D. I. Yuk, S. Kaing, C. D. Stiles, and D. H. Rowitch. 2001. bHLH proteins interact with homeodomain proteins to regulate cell fate acquisition in progenitors of the ventral neural tube. Curr. Biol. 11:1413-1420.[CrossRef][Medline]

27. Wan, H., S. Dingle, Y. Xu, V. Besnard, K. H. Kaestner, S. L. Ang, S. Wert, M. T. Stahlman, and J. A. Whitsett. 2005. Compensatory roles of Foxa1 and Foxa2 during lung morphogenesis. J. Biol. Chem. 280:13809-13816.[Abstract/Free Full Text]

28. Waris, G., and A. Siddiqui. 2002. Interaction between STAT-3 and HNF-3 leads to the activation of liver-specific hepatitis B virus enhancer 1 function. J. Virol. 76:2721-2729.[Abstract/Free Full Text]

29. Westerman, B. A., C. Murre, and C. B. Oudejans. 2003. The cellular Pax-Hox-helix connection. Biochim. Biophys. Acta 1629:1-7.[Medline]

30. Wu, L., H. Wu, L. Ma, F. Sangiorgi, N. Wu, J. R. Bell, G. E. Lyons, and R. Maxson. 1997. Miz1, a novel zinc finger transcription factor that interacts with Msx2 and enhances its affinity for DNA. Mech Dev. 65:3-17. (Erratum, 69:219.)[CrossRef][Medline]

31. Yan, C., A. Naltner, J. Conkright, and M. Ghaffari. 2001. Protein-protein interaction of retinoic acid receptor alpha and thyroid transcription factor-1 in respiratory epithelial cells. J. Biol. Chem. 276:21686-21691.[Abstract/Free Full Text]

32. Yuan, B., C. Li, S. Kimura, R. T. Engelhardt, B. R. Smith, and P. Minoo. 2000. Inhibition of distal lung morphogenesis in Nkx2.1(–/–) embryos. Dev. Dyn. 217:180-190.[CrossRef][Medline]

Molecular and Cellular Biology, March 2007, p. 2155-2165, Vol. 27, No. 6
0270-7306/07/$08.00+0     doi:10.1128/MCB.01133-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Other Versions of this Article:
MCB.01133-06v1
27/6/2155    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Minoo, P.
Right arrow Articles by Li, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Minoo, P.
Right arrow Articles by Li, C.