The Wing 2 Region of the FOXC1 Forkhead Domain Is Necessary for Normal DNA-Binding and Transactivation Functions

Abstract

purpose. To determine the biochemical defects that underlie Axenfeld-Rieger malformations, to determine a functional role for wing 2 in FOXC1, and to understand how mutations in this region disrupt FOXC1 function.

methods. Sequencing DNA from patients with Axenfeld-Rieger malformation resulted in the identification of two novel missense mutations (G165R and R169P) in wing 2 of FOXC1. Site-directed mutagenesis was used to introduce these mutations, as well as previously reported mutation (M161K), into the FOXC1 cDNA. These FOXC1 mutants were evaluated to determine their ability to localize to the nucleus, bind DNA and activate gene expression.

results. Two novel missense mutations were identified in unrelated patients, in wing 2 of the FOXC1 forkhead domain. Because there had been no previous biochemical analysis, the mutation M161K was also investigated. All three mutant proteins localized correctly to the nucleus. The G165R mutation maintained wild-type levels of DNA binding; however, both the M161K and R169P mutations displayed reduced DNA binding ability. Biochemical analysis showed that all three mutations disrupt FOXC1’s transactivation ability.

conclusions. Biochemical analysis of mutations G165R and R169P and of a previously reported mutation, M161K, demonstrate the functional significance of wing 2. M161K and R169P disrupt DNA binding of FOXC1, consistent with the hypothesis that wing 2 is necessary for DNA binding. The results also suggest that wing 2 plays a role in gene activation. These results provide the first insights into how mutations in wing 2 disrupt FOXC1 function.

Only a decade ago, the winged-helix-turn-helix (wHTH) DNA-binding domain was defined. ¹ The wHTH DNA-binding domain is a variant of the helix-turn-helix motif and consists of a three helix bundle and two large loops that form the “wing” structures that define the domain. ^{1 2 3 4} , ^{5 6 7} Characterized by a 110-amino-acid domain known as the forkhead domain (FHD), this DNA-binding motif was originally identified as a region of homology between the Drosophila melanogaster forkhead protein and rat hepatocyte nuclear factor 3 protein (known also as Foxa3). ⁸  

FOX (forkhead box) proteins are a unique subset of the wHTH proteins characterized by the presence of a loop C-terminal to β-strands called wing 2, that is essential for FOX/winged helix proteins to bind DNA as a monomer. ⁹ , ¹⁰ The three-dimensional solution structure of FOXA3/DNA has been determined, and the structure depicts wing 2 crossing the three-helix bundle in the vicinity of the N terminus of α-helix 1 and the C terminus of α-helix 3. ¹ The region spanning wings 1 and 2 appears flexible and may provide a way of connecting subdomains that produce contacts necessary for DNA binding. ^{3 11}  

Stevens et al. ¹¹ proposed an evolutionary pattern for the formation of wing 2 from an initial unstructured distal region in a basic wHTH protein. They further suggested that the acquisition of bulky aromatic and/or hydrophobic residues in this distal region allows for association with the hydrophobic and/or aromatic residues in the three-helix bundle. The ability to bind DNA as a monomer is a functional advantage that may have strengthened associations between residues in wing 2 and α-helix 1 that resulted in a functional dependence on wing 2 for stable DNA binding. ¹¹ However, the consequences of alteration of wing 2 on FOX protein function has not been studied. 

Mutations in FOXC1 lead to Axenfeld-Rieger (AR) malformations, a clinically heterogeneous condition with variable penetrance. ^{12 13 14 15 16 17 18 19 20} Patients manifest posterior embryotoxon, adhesions of the iris and the cornea, iris hypoplasia, corectopia, and/or an underdeveloped or aberrantly developed angle between the cornea and the iris. Nonocular features such as dental anomalies and redundant periumbilical skin may also be present. On rare occasions, patients have congenital cardiac anomalies. ^{12 15 21 22 23} Glaucoma is a progressive blinding condition that occurs in approximately half of patients with AR malformations. 

A molecular dissection of FOXC1 has shown that disease-causing missense mutations in the FHD disrupt FOXC1 function. ^{5 7} Mutations in α-helix 1, between α-helix 2 and 4, and those that reside in the recognition helix α-helix 3, have been characterized. ^{5 7} These mutations have been found to impede the ability of FOXC1 to bind DNA and activate transcription as well as to reduce levels of protein produced. ^{5 7} Molecular analysis of disease-causing missense mutations has enabled possible functional domains to be assigned to different regions in the FOXC1 FHD. ^{5 7}  

We report herein, the identification, patient phenotype, and molecular analysis of two novel and one previously reported ^{3 4} disease-causing missense mutations in wing 2 of the FHD of FOXC1. Molecular analyses reveal that all three mutations reduce the ability of FOXC1 to transactivate gene expression and that the R169P and M161K mutations also disrupt the ability of FOXC1 to bind DNA. Analysis of these mutations demonstrates the importance of wing 2, a poorly characterized region, to the functional integrity of FOXC1 DNA binding and FOX/wHTH proteins in general. 

Methods

Patient Report

This research adhered to the tenets of the Declaration of Helsinki. Patient 1 (G165R mutation) was a white woman who received the initial diagnosis at approximately 50 years of age. The patient presented with iridogoniodysgenesis, iris stroma hypoplasia, posterior embryotoxon, corectopia, and glaucoma, as well as some dental anomalies. Clinical photographs were not available for this patent. Patient 2 (R169P mutation) was also a white woman, in whom the disorder was initially diagnosed at 5 years of age. The patient had iris hypoplasia, hypertelorism, corneal opacity, and abnormal pupillary function (Fig. 1 ). Patient 2 did not exhibit dental anomalies; however, she had hearing loss. Mutations in both patients 1 and 2 were de novo. Two independent patients of Indian ancestry, both with the previously reported M161K mutation, were reported to have an anteriorly displaced Schwalbe’s line, abnormal angle tissue, corectopia, hypoplastic iris, and megacornea. ^{3 4}  

Mutation Detection

Conditions for amplification of FOXC1 are published elsewhere. ²⁰ PCR products were gel purified, extracted on separation columns (QIAquick; Qiagen, Valencia, CA), and sequenced directly by using ³³ P sequencing (Amersham, Arlington Heights, IL). 

Plasmid Construction

FOXC1 pcDNA4 His/Max B TM (Invitrogen, Carlsbad, CA) has been described previously. ⁵ Mutagenesis was performed with a kit (Quickchange; Stratagene, La Jolla, CA) with the addition of 10% dimethylsulfoxide and the appropriate primers. The mutagenic primer sequences were as follows: M161K, forward 5′-ccg gac tcc tac aac aag ttc gag aac ggc agc-3′ and reverse 5′gct gcc gtt ctc gaa ctt gtt gta gga gtc cgg-3′; G165R, forward 5′-aac atg ttc gag aac cgc agc ttc ctg cgg cgg-3′ and reverse 5′-ccg ccg cag gaa gct gcg gtt ctc gaa cat gtt-3′; and R169P, forward 5′-gag aac ggc agc ttc ctg ccg cgg cgg cgg cgc ttc aag-3′ and reverse 5′-ctt gaa gcg ccg ccg ccg cgg cag gaa gct gcc gtt ctc-3′. Potential mutant constructs were sequenced with an automated sequencer (LI-COR, Lincoln, NE) and confirmed mutants were subcloned into the FOXC1 pcDNA4 His/Max vector and resequenced. 

Cell Culture

COS-7 and HeLa cells were maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum. Transfections were performed with a reagent (FuGene6; Roche Diagnostics, Indianapolis, IN), according to the manufacturer’s protocol, with 2 μg DNA per 100-mm plate and 0.5 μg DNA per 25-mm well. 

Immunofluorescence

COS-7 cells were grown and transfected on coverslips with 1 × 10 ⁵ cell/mL placed into six-well tissue culture plates. After 24 hours, immunofluorescence was performed as described previously. ⁵  

Electrophoretic Mobility Shift Assays

COS-7 extracts containing recombinant FOXC1 were equalized for amounts of recombinant FOXC1 protein by Western analysis using the anti-Xpress antibody (Stratagene) against the vector encoded N-terminal Xpress tag. Protein extracts were incubated with 1.25 mM dithiothreitol (DTT); 0.3 μg sheared salmon sperm DNA; 0.125 μg poly dI^.dC (Sigma-Aldrich); and 80,000 cpm ³² P-dCTP–labeled, double-stranded DNA containing the in vitro preferred FOXC1-binding site (shown in italic; forward 5′-gat cca aag taa ata aa caa cag a-3′; reverse 5′-gat ctc tgt tgt tta ttt act ttg-3′). Reactions were incubated at room temperature for 30 minutes and resolved on a 6% polyacrylamide Tris/glycine/EDTA gel. 

Dual Luciferase Reporter Assays

Transactivation assays were performed as described previously (Dual Luciferase Reporter Assay System; Promega, Madison, WI). ⁵  

Modeling

The homology model of FOXC1 was created by aligning residues 76-177 of FOXC1 (GenBank Accession Q12948) with residues 117-218 of the crystal structure of FOXA3 ¹ in Swiss-PdbViewer. ²⁴ These sequences align with no gaps (Fig. 2B ). Although FOXC2, whose abbreviated FHD was solved by nuclear magnetic resonance (NMR) spectroscopy, ²⁵ shares very high sequence identity with FOXC1 (95.4%; Fig. 2B ), only FOXA3 (68.8% identity) was used as a template for homology modeling of FOXC1, because that purified peptide contained the most complete wing 2. The backbone atoms of FOXC1 were “fitted” against the template, and the modeling project was submitted to the SWISS-MODEL server (http://www.expasy.org/spdbv/; provided in the public domain by the Swiss Institute of Bioinformatics, Geneva, Switzerland). The resultant FOXC1 layer was subjected to in silico mutagenesis using the “mutate” tool of the Swiss-PdbViewer. For each amino acid, all pair-wise interatomic contacts of less than 0.7 nm (between residues that are greater than 11 positions apart), including a term for solvent accessibility, are converted to pseudoenergy terms (analogous to threading analysis). Low-scoring regions indicate structurally favorable environments and roughly correspond to the predicted α-helices and β-strands (not shown). The program ANOLEA ²⁶ was used to calculate pseudoenergy terms for individual residues in our wild-type and M161K models of FOXC1 through its knowledge-based atomic-level mean force potential. When M161 was changed to K161, the default (optimal) rotamer choice was selected to replace the wild-type side chain. 

Model structures were submitted to the ANOLEA server, http://www.swissmodel.unibas.ch/anolea/; provided in the public domain by the Laboratory of Structural Molecular Biology, The University of Namur, Namur, Belgium) with the window average set to one amino acid. 

Results

Mutation Identification

FOXC1 was screened for mutations by direct sequence analysis of PCR products of patient DNA. Two single nucleotide changes were identified within wing 2 of the FHD (Fig. 3 ) within two patients with AR malformations. A G-to-C transversion at codon position 165 (494G→C; G165R) was detected in patient 1. The transversion results in a glycine to arginine change. Another G to C transversion at codon position 169 (506G>C; R169P) was detected in patient 2. This transversion results in an arginine-to-proline change. Sequencing of the patients’ DNA and 100 normal control chromosomes confirmed that these mutations are not present in the normal population and are de novo mutations. A previously reported M161K mutation, resulting from a T-to-A mutation at codon position 161 (482T→A; M161K) in two different AR patients ^{3 4} but not previously biochemically analyzed, was also included in our analysis. 

Molecular Analysis

The mutations G165R, M161K, and R169P all occur in wing 2 of the FHD. To discern how these mutations alter FOXC1 function, site-directed mutagenesis was used to introduce the mutations into the FOXC1 cDNA. Immunofluorescent detection of Xpress tagged (Stratagene) recombinant FOXC1 showed that wild-type FOXC1, G165R, M161K, and R169P all localize predominantly to the nuclei of COS-7 cells (Fig. 4 ). The stability of FOXC1 was not compromised by these mutations as recombinant proteins of the correct molecular weight were detected by western analysis (Fig. 5A ). COS-7 cell extracts containing G165R, M161K, and R169P recombinant FOXC1 were used in electrophoretic mobility shift assays (EMSAs) to visually quantify if these mutations have an affect on the ability of FOXC1 to bind the previously described FOXC1-binding site. ^{5 27} The G165R mutation showed a capacity for DNA binding comparable to wild-type FOXC1, whereas the M161K and R169P mutations showed a reduction in DNA-binding by 20% to 30% and 50% to 60% of wild-type levels, respectively (Fig. 5B ). All mutations were tested against a panel of variants of the FOXC1-binding site, ⁵ to determine whether binding specificity was altered. None of the three wing 2 mutations altered the binding specificity of FOXC1 (data not shown). 

FOXC1 has been shown to transactivate gene expression of a reporter construct ^{5 6} ; thus, the effect of G165R, M161K, and R169P on the transactivation potential of FOXC1 was investigated. Using a luciferase reporter construct with FOXC1-binding sites positioned upstream of a thymidine kinase promoter, the transactivation ability of FOXC1 was found to be disrupted by all mutations (Fig. 5C ). FOXC1 in which the FHD was deleted (ΔBOX) activated luciferase expression at only 1% of wild-type levels. FOXC1 carrying G165R transactivated expression of the luciferase reporter at 36% of wild-type levels. M161K and R169P levels of transactivation were 20% and 23% of wild-type, respectively. These results indicate that although these mutant proteins are able to bind to the FOXC1-binding site in vitro and maintain nuclear localization, all three mutant proteins exhibit reduced levels of gene activation. Therefore, the overriding defect resulting from all three mutations is likely to be an impaired capability to activate transcription of target genes. 

Molecular Modeling

A theoretical structure of the FHD of FOXC1 was generated by homology modeling using FOXA3 (Fig. 2A ) as the template. Of the closely related FHDs with known structure, the FOXA3 model has the most complete wing 2 region and also contains identical residues at the positions equivalent to M161, G165, and R169 in FOXC1 (Fig. 2B ). The model structures were submitted to the ANOLEA server which computes a nonlocal energy profile for assessing protein structure. ²⁶ Methionine 161 in FOXC1 was mutated to lysine in silico to predict structural defects in the M161K molecule. Fig. 6 illustrates that the M161K mutation may disrupt structurally favorable interactions in an interior region of FOXC1 where T88, C135, F136, and M161 are predicted to converge. Assuming there are no changes in the structure of the protein backbone after mutagenesis, the optimal rotamer for K161’s side chain noticeably increased the ANOLEA scores of residues 88, 135, and 161 (Fig. 6C ). This pattern suggests that these residues are normally involved in a highly organized region of side chain packing (Fig. 6A ) such as would occur in a globular protein’s hydrophobic core. Disruption of this core-like module to accommodate the M161K amino acid substitution may lead to localized instability or deformation of wing 2. Similar results were obtained with in silico mutagenesis analysis of FOXA3 (Fig. 6C ) that has identical amino acids at the equivalent positions (Fig. 2B ). In silico mutagenesis to create G165R and R169P models did not result in any significant changes to ANOLEA scores of nonlocal residues (not shown) as expected for solvent-exposed positions. 

Discussion

Mutations in wing 2 of the FOXC1 FHD, or of any FOX proteins, have not been previously subject to molecular analyses. FOXC1 is a single-exon gene; therefore, the effects of amino acid substitution mutations are not predicted to involve alterations of exonic splicing, but rather at the protein function level. All the patients with mutations in wing 2 of FOXC1 present with AR phenotypes similar to those in patients with mutations in other regions of the FHD. Thus, there appears to be no phenotype–genotype correlation between the region of mutation in the FHD and patient phenotype. However, the identification of three missense mutations in wing 2 is helpful in refining our understanding of the FHD of FOXC1 and, more important, has shed light on the possible functional role of a previously poorly characterized region of the FHD. 

Nuclear Localization Disruption by Mutations G165R, M161K, or R169P

The nuclear localization signals (NLSs) of FOXC1 reside in the N and C terminus of the FHD. ⁶ Although G165, M161, and R169 all reside in the C-terminal region required for nuclear localization, substitutions at these sites are not sufficient to disrupt the signal and perturb nuclear localization (Fig. 4 ). Therefore, the wing 2 mutations are tolerated in the mutant FOXC1 proteins and do not disturb the NLSs. 

Effect of G165R on Wild-Type Levels of DNA Binding and Transactivation Ability

In the FOXA3/DNA crystal structure, position G165 faces outward and is not a predicted base, phosphate backbone, or water contact. ¹ Therefore, mutation at position 165 is not predicted to disrupt DNA binding. The fact that FOXC1 carrying G165R can maintain wild-type levels of DNA binding (Fig. 5B ) agrees with this prediction and indicates that the winged-helix structure necessary for DNA binding is not disrupted. The severe reduction in transactivation due to the G165R mutation implies a functional role for this position in transactivation. Previous characterization of two FOXC1 disease-causing missense mutations, F112S and I126M in the FHD of FOXC1 show similar disruptions in transactivation while maintaining normal DNA binding. ⁶ Because of the disruptions of protein–protein interactions, the G165R mutation may also disrupt the intramolecular interactions necessary for transcription activation. Previous reports suggest mutations F112S and I126M do not disrupt DNA binding because they are not involved in protein/DNA interactions. ⁵ Rather, they may disrupt transactivation due to disturbances in protein–protein interactions and/or intramolecular interactions. ⁵ Previous characterization of the FOXC1 FHD has implicated the N terminus of α-helix 1 in transactivation and DNA binding. ⁵ Wing 2 of FOXA3 meanders across the surface of the three-helix bundle in the vicinity of the N terminus of α-helix 1, ¹ establishing an environment receptive to interaction between regions of α-helix 1 and wing 2. Therefore we hypothesize that potential intramolecular interactions between wing 2 and α-helix 1 are disrupted by the G165R mutation, indirectly disrupting the transactivation ability while maintaining DNA-binding ability. 

Effect of M161 on DNA-Binding and Transactivation Abilities

Stable formation of wing 2 is essential for DNA binding of FOX proteins. ¹ In our predicted structure of FOXC1, M161 is spatially arranged such that its side chain resides in close proximity to the side chains of T88, C135, and F136 in a region of reduced solvent accessibility (Fig. 6A ), where the cusp of wing 1 converges with α-helix 1 and the middle of wing 2. The substitution of polar lysine for nonpolar methionine at position 161, as occurs in both patients with the reported mutation, places a larger positively charged residue in the vicinity of a hydrophobic-core–like module. In silico mutagenesis of M161 to M161K, in which only the atomic constituents and rotation of the amino acid side chain at position 161 were altered, suggests that the new side chain would no longer closely associate with the T88/C135/F136 cluster (Fig. 6B ). We hypothesize that M161K perturbs the stability of wing 2 enough to reduce DNA-binding ability by affecting the interatomic nonlocal interactions between residues T88, C135, and M161. 

Unlike G165R the loss of transactivation in M161K is most likely caused by the approximate 20% to 30% reduction in DNA binding of this mutation (Fig. 5B ). M161K is similar to a previous mutation, L86F, in α-helix 1 of the FHD of FOXC1. ⁷ L86F shows a minor reduction of DNA binding ability but shows marked reduction in transactivation from a luciferase reporter, ⁷ implicating L86F as having an important role in intramolecular interactions. ²² In FOXA3, position M202 is one of 35 known residues that contribute to the hydrophobic core of the FHD. ¹ This position is conserved in FOXC1 and corresponds to position M161. We predict therefore that M161 may also have a role in intramolecular interactions that the M161K mutation disrupts, impairing transactivation. 

Effect of Mutation at Position R169 on DNA Binding and Transactivation Ability

Based on the crystal structure of FOXA3, ¹ the R169P mutation occurs in a position that is predicted to be responsible for the sole side chain–base minor groove contact in wing. ¹ Our analysis predicts that R169P disrupts this base contact and potentially destabilizes the formation of wing 2, which is critical for DNA binding activity. ¹  

The severe reduction in DNA binding from mutation R169P is translated into a marked reduction in transactivation of a luciferase reporter (Fig. 5C ). The severe reduction in DNA binding seen in FOXC1 R169P, and the knowledge that R169 makes a DNA contact in the related FOX protein FOXA3, ¹ supports the hypothesis that this reduced transactivation is a result of reduced DNA-binding capability. 

The Role of Wing 2 in DNA-Binding and Transactivation of FOXC1

FOX proteins exhibit similar binding specificity to a core sequence (TAAAYA) due to the conservation of residues in the recognition helix ¹ ; however, it has been suggested that the DNA binding specificity may be determined by residues outside of α-helix 3. ²⁸ In sequence comparisons of FOX genes, residues of wings 1 and 2 have the most divergent sequences, suggesting their possible involvement in sequence recognition. ^{10 29} Although mutations M161K and R169P reduced the DNA binding ability of FOXC1, none of the three wing 2 mutations discussed herein affected the binding specificity of FOXC1 (data not shown). It cannot be ruled out, however, that other residues in wing 2 play a role in sequence specificity by interacting with other residues in the helix bundle. Nevertheless, mutations G165R, M161K, and R169P, although apparently disrupting the stable formation of wing 2 and subsequent DNA binding, do not appear to disturb internal interactions between wing 2 and other regions involved in DNA binding specificity. 

We propose that wing 2 plays an important role in DNA binding and the transactivation capability of FOXC1. Previous structural analyses of FOX proteins strongly support a role for wing 2 in DNA binding. ^{1 11 30} Disrupting R169, the sole base contact made in the minor groove by wing 2, is sufficient to reduce FOXC1 DNA binding to 20% to 30% of wild-type levels and reduce the capability of the mutated protein to transactivate genes nearly fourfold (Fig. 5C ). Disrupting residue M161, which is not predicted to be involved in any direct base contacts, ¹ is also sufficient to reduce the DNA binding and transactivation ability of FOXC1 (Fig. 5B ). Clearly, wing 2 of the FOXC1 FHD plays an important role in DNA binding, whether it is through a direct base contact or through intramolecular interactions that may be required for correct conformation of the FOXC1 protein. Our analysis of the G165R mutation indicates that wing 2 also plays a key role in transactivation, a role not initially predicted outside of wing 2 involvement in DNA binding. Future studies of the functional role of wing 2 will not only enhance our understanding of the FOX/wHTH family, but will also expand our knowledge of what our results indicate is a key region necessary for FOX protein DNA binding and transactivation. 

 Supported by a Graduate Assistance Award (TCM). MAW is an Alberta Heritage Fund for Medical Research Senior Scholar and a Canadian Institutes of Health Research (CIHR) investigator.

 Submitted for publication February 17, 2004; revised March 19, 2004; accepted March 29, 2004.

 Disclosure: T.C. Murphy, None; R.A. Saleem, None; T. Footz, None; R. Ritch, None; B. McGillivray, None; M.A. Walter, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Photographs of the eyes of the patient carrying the R169P mutation in FOXC1. (A) Right eye shows a small, eccentric pupil. (B) Left eye shows a thin, elongated horizontal pupil. This patient had iris hypoplasia, hypertelorism, corneal opacity, and abnormal pupillary function.

Figure 2.

 

Homology modeling of FOXC1. (A) The crystal structure of human FOXA3 ¹ is shown with the protein backbone represented as a white ribbon (wing 2 is blue) and the cocrystallized DNA binding target is shown as a space-filled yellow model. M, G, and R denote the residues equivalent to M161, G165, and R169 in FOXC1, which are mutated in AR. FOXA3 was used as the sole template structure for homology modeling of FOXC1. (B) Sequence alignment of FOXC1 with FHD-containing peptides whose structures have been determined by x-ray crystallography or NMR spectrometry. The secondary structure of FOXA3 depicted in the ideogram was predicted by Swiss-PdbViewer. The wings are suggested to comprise the non–B-strand regions after B-strand 1. Purple, identical residues; yellow, similar residues; black, conservation of M161, G165, and R169. *Peptides complexed with DNA.

Figure 3.

 

Identification of two novel missense mutations of FOXC1 in patients with Axenfeld-Rieger malformations. Individual lanes of patient and control DNA sequencing reactions were run side by side for comparison. The genomic DNA sequence of unaffected controls (unmarked) and patients with the FOXC1 mutations (*) are shown. The G-to-C transversion (GGC→CGC) in patient 1 resulted in an amino acid change of glycine to arginine, and the G-to-C transversion (CGC→CCG) in patient 2 resulted in an amino acid change of arginine to proline. Both transversions resulted in missense mutations in wing 2 of the FOXC1 FHD.

Figure 4.

 

G165R, M161K, and R169P mutations do not disrupt the nuclear localization of FOXC1. Immunofluorescence of recombinant FOXC1 constructs in COS-7 cells and the percentage of cells with nuclear, nuclear+cytoplasmic and cytoplasmic localization are indicated.

Figure 5.

 

G165R, M161K, and R169P mutations disrupt FOXC1 function. (A) Western blot of Xpress epitope-tagged FOXC1, G165R, M161K, and R169P detected with an anti-Xpress antibody. (B) G165R binds the FOXC1-binding site at near wild-type levels, whereas M161K shows a minor reduction (20%–30%) in binding, and R169P has a severe (50%–60%) DNA-binding deficit. 1×, 2×, and 5× corresponds to increased amounts of protein used in binding reactions. Arrowheads: protein/DNA complex. (C) Transactivation of a luciferase reporter by FOXC1 is impaired by G165R, M161K, and R169P mutations. At the top is a schematic of the reporter construct with six copies of in vitro derived FOXC1-binding site, driving expression of the luciferase gene. Error bars, SD.

Figure 6.

 

In silico mutagenesis of FOXC1. (A) A portion of the FOXA3-derived homology model of FOXC1 is shown with the protein backbone represented as a white ribbon (wing 2 is blue). Potential structurally important residues are depicted with space-filled heavy atoms: gray, carbon; red, oxygen; purple, nitrogen; yellow, sulfur. (B) Methionine 161 from FOXC1 (and the equivalent M202 residue in the FOXA3 model, see Fig. 4 ) was mutated to lysine using the Swiss-PdbViewer. (C) The wild-type and M161K-equivalent models for both FOXC1 and FOXA3 were submitted to the ANOLEA ²⁸ server. The differences between respective mutant and wild-type model scores (in E/kT units) are shown, highlighting which nonlocal residues are affected by changes at position 161.

The authors thank May Yu for technical assistance and tissue culture, Fred Berry for helpful discussions, and Alan Underhill and Susan Andrew for helpful comments concerning the manuscript.