Identification of Target Genes Regulated by FOXC1 Using Nickel Agarose–Based Chromatin Enrichment

Abstract

purpose. To overcome the problem of antibody availability, often encountered during chromatin immunoprecipitation (ChIP) assays, nickel agarose–based chromatin enrichment (NACE) was developed. Based on the affinity of (His)-₆-tagged proteins for the nickel ion, this modified form of ChIP allows the isolation of chromatin in the absence of specific antibodies.

methods. Nonpigmented ciliary epithelium cells were transfected with (His)-₆-tagged FOXC1. FOXC1-enriched chromatin complexes were isolated by using the tight electrostatic interaction between histidine residues of the recombinant FOXC1 protein and nickel. One hundred fifty NACE-enriched clones were sequenced and subjected to in silico and biochemical analyses.

results. Twenty-six clones were detected near known genes: Eight were near predicted but uncharacterized genes, eight were within areas where neither known nor predicted genes have yet been mapped, four were chimeric, and the rest were either repetitive (n = 81) or poor-quality (n = 23) sequences. Twenty of the 26 known genes were expressed in the eye. Five of the NACE-enriched clones (BMP2K, DACH, FVT-1, SIX-1, and PGE-2 receptor), as well as nine clones selected from the literature, were validated by PCR amplification in two independent lots of NACE-enriched chromatin. All five NACE-selected genes were detected in two independent assays, as well as four (BMP7, SMAD2, TGF-B1, and WNT6) of the nine genes selected from the literature, consistent with these genes’ being regulated by FOXC1.

conclusions. NACE is a useful technique allowing specific chromatin enrichment in cases where antibodies are unavailable. Specific recovery of PTGER, DACH1, WNT6, and FVT-1 implicates FOXC1 in a variety of cellular events including modulation of intraocular pressure, cell cycle, ocular development, and oncogenesis.

Eukaryotic chromatin is an organized structural hierarchy in which the smallest unit is the nucleosome, a complex of ∼150 bp of DNA wrapped around two copies of each of four histones: H2A, -2B, -3, and -4. ¹ These histone-DNA complexes, associated with other components in the nucleus, are continuously interacting with one another to ensure the dynamic that characterizes living cells. Before transcription of target genes begins, transcription factors are gathered under complex interactions within the nucleus, as, simultaneously, the chromatin undergoes essential transformations (i.e., acetylation) to reach a relaxed form suitable for transcription initiation. ² However, the dynamic and structural complexity of chromatin usually hampers gene regulation studies. Chromatin immunoprecipitation (ChIP) is a powerful emerging technique designed to overcome this problem and to identify genes bound by transcription factors. ^{3 4 5} The ChIP technique involves formaldehyde treatment of cells to cross-link transcription factors to their target elements on DNA in vivo. After cell lysis, the protein-DNA complexes are recovered by immunoprecipitation, using an antibody to the transcription factor of interest. After reversing the cross-link, the target element DNA is analyzed by different methods including PCR amplification. The resultant PCR product minilibraries can be sequenced and compared to public DNA sequence databases to discover the identity and likely function of nearby genes. ⁶ Alternatively, the ChIP DNA pools can be labeled and hybridized to microarrays containing a large selection of expressed sequence tags (ESTs) or promoters. ^{7 8}  

ChIP analyses have been applied successfully to identify targets of yeast and mammalian transcription factors. ^{9 10} In the yeast Saccharomyces cerevisiae, ChIP was used to identify targets of the Gal4 and Ste12 transcription factors. Of the 10 Gal4 targets identified, 7 have been shown to be Gal4 targets, suggesting that ChIP analysis is robust and successful in obtaining bona fide target promoters. ¹⁰ The Gal4 DNA target sequence was found in these promoters. In mammals, ChIP analyses revealed that 127 promoters involved in the cell cycle were bound by the E2F4 repressor and that a subset of these were also bound by the E2F1 activator. ⁹  

FOXC1 (formerly FREAC3 or FKHL7) is a member of the Forkhead box (Fox) family of transcription factor genes named for a well-conserved 100-amino-acid DNA-binding domain homologous to Drosophila forkhead. FOXC1 is a monoexonic gene of 1.6 kb encoding a protein of 553 amino acids ^{11 12} located on chromosome 6 at p25 ^{13 14} and known to be expressed in the eye. ¹⁵ Mutations in FOXC1 are responsible for Axenfeld-Rieger (AR) malformations mapped to 6p25. ¹⁶ The most serious consequence for patients with AR is an approximately 50% risk of the development of glaucoma. How FOXC1 mutations lead to glaucoma is unknown. However, identification of potential candidate genes regulated by FOXC1 is a prerequisite for defining the role of FOXC1 in eye organogenesis and diseases. To address this question and to overcome the absence of an appropriate FOXC1 antibody, we modified the ChIP technique. Chromatin was affinity purified by using nickel agarose–based chromatin enrichment (NACE) to identify genes regulated by FOXC1 in the eye. 

Methods

Cell Culture and Transient Transfection Assay

Nonpigmented ciliary epithelial (NPCE) cells ¹⁷ were maintained and grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum (Gemini Bioproducts, Calabasas, CA), 50 U/mL penicillin, and 50 μg/mL streptomycin. Plates 100 mm in diameter containing 2 × 10⁶ cells were transiently transfected, at 70% to 80% confluence, by 4 μg plasmid DNA (pcDNA4 His/Max FOXC1) through use of a transfection reagent (FuGene6; Roche Diagnostics, Indianapolis, IN) according to the manufacturer’s protocol. The NPCE cells were also transfected with an empty vector (pcDNA4 His/Max), a FOXC1 ΔBox control (FOXC1 lacking the forkhead domain), or a non-FOXC1 binding site vector, as negative controls. After 48 hours of incubation at 37°C in a humidified atmosphere containing 5% CO₂, cells were subjected to NACE. 

Nickel Agarose–Based Chromatin Enrichment

Ni²⁺-NTA agarose beads (Qiagen; Valencia, CA) were used to isolate chromatin complexes containing the (His)-₆-FOXC1 recombinant fusion protein from nuclear lysates. Figure 1 illustrates schematically the different steps of NACE protocol. Briefly, transfected NPCE cells were brought to room temperature (RT) and cross-linked by adding formaldehyde directly to tissue culture medium to a final concentration of 1% and left at RT for 15 minutes with gentle shaking. The cross-linking was stopped by adding glycine to a final concentration of 0.125 M at RT for 5 minutes on a shaking platform. Cross-linked cells were washed twice with cold 1× phosphate-buffered saline (137 mM NaCl, 10 mM sodium phosphate buffer [pH 7.4], and 2.7 mM KCl), scraped, and lysed at a density of 5 × 10⁶ cells/mL for 10 minutes in ice-cold cell lysis buffer (5 mM piperazine-N-N′-bis(2-ethanesulfonic acid [PIPES; pH 8], 85 mM KCl, and 0.5% NP40 + protease inhibitors). Nuclei were pelleted by microcentrifugation at 6000 rpm, resuspended in 1× Ni²⁺-NTA resin binding buffer (50 mM NaH₂PO₄ [pH 8], 300 mM NaCl, and 10 mM imidazole) containing protease inhibitors (mammalian protease inhibitors; Sigma-Aldrich, St. Louis, MO) sonicated for 10 rounds of 20-second cycles, at 2-minute intervals, at 40% amplitude (Sonic Dismembrator, model 500; Fisher Scientific, Pittsburgh, PA). The sonicated chromatin (≤1.5 kb) was centrifuged at 4°C for 30 minutes at 12,000g and the supernatant transferred to a new tube containing 300 μL of charged nickel resin before overnight incubation in a cold room with gentle rocking. At this stage, a positive control (100% input, no Ni²⁺-NTA added to sample) and a negative control (mock, H₂O plus Ni²⁺-NTA) were also included. Samples were centrifuged at 3000 rpm in 4°C, supernatant was removed, and the bead pellet was washed in 1 mL washing buffer (50 mM NaH₂PO₄ [pH 8], 300 mM NaCl, and 10 mM imidazole) for 10 minutes with gentle shaking. The supernatant from the 100% input was set aside and used as a positive control for PCR amplification (described later). The washing procedure was repeated four times, to eliminate a maximum of nonspecific binding. The chromatin was eluted from the beads four times by adding 200 μL of elution buffer (50 mM NaH₂PO₄ [pH 8], 300 mM NaCl, and 250 mM imidazole), for 15 minutes at RT with rotation; centrifuging at 3000 rpm; and removing the supernatant without disturbing the nickel resin, by using a 26-gauge syringe. Eluates were treated for 5 hours at 70°C with 1 μL of 10 mg/mL RNase A and 50 μL of 5 M NaCl, to reverse the cross-linking, and precipitated overnight at −20°C in a 2.5-volume of ethanol. Samples were then microfuged at 14,000g for 20 minutes, and pellets were dissolved in 100 μL TE (50 mM Tris-Cl [pH 7.5], 25 mM EDTA), 25 μL of 5× proteinase K buffer, and 1.5 μL of 40 μg/mL proteinase K and incubated for 2 hours at 45°C to remove protein. Finally, the DNA was purified (QIAQuick kit; Qiagen). At this stage, target DNA was enriched and ready for cloning or for PCR amplification. 

Cloning, Construct Preparation, and Sequencing

Three hundred nanograms of enriched chromatin was added to 5 μL of 10× end-filling buffer (50 mM Tris-HCl [pH 8] and 10 mM MgCl₂), 10 μL of 0.25 mM dNTPs, and 3 U Klenow enzyme in a total volume of 50 μL. After an incubation of 1 hour at 37°C, the reaction was stopped at 75°C, and DNA was purified (Qiagen columns). One hundred fifty nanograms of the end-filled insert (enriched chromatin) was then added to 50 ng of EcoRV-digested DNA vector (pBluescript SKII; Stratagene, La Jolla, CA), 5 μL of 5× ligation buffer (60 mM Tris-HCl [pH 8], 20 mM MgCl₂, and 20 mM dithiothreitol [DTT]), 5 μL of 1 M polyethylene glycol, and 5 U of T4 ligase enzyme. The reaction was finally incubated at 5°C for at least 15 hours. The ligation product was subsequently used for the transformation of XL-Blue cells. Transformed cells were plated, and putative positive clones, based on white/blue selection, were cultured overnight at 37°C in 5 mL Luria-Bertani (LB) medium before DNA extraction (Miniprep kit; Qiagen). The extracted DNA was subsequently digested with the BamH1 and HindIII restriction enzymes, to identify the clones containing the insert. 

Fifty picomoles of forward and reverse M13 primers conjugated to chromophores emitting at different wavelengths (700 and 800 nm), were added to 10 μL of extracted DNA (1–3 μg) and subjected to a sequencing reaction with a fluorescence-labeled primer cycle sequencing kit with 7-deaza 2-dGTP (Thermo Sequenase; Amersham Pharmacia Biotech, Ltd., Amersham, UK). 

PCR Analysis

Sequences of oligonucleotide primer pairs used for PCR amplification are listed in Table 1 and were designed for each gene (Primer-3 program; http://www.broad.mit.edu/cgi-bin/primer/primer3_www.cgi; provided in the public domain by Massachusetts Institute of Technology, Cambridge, MA). The PCR reactions were performed in total volumes of 50 μL with a 50-second melting step at 95°C, an annealing step between 52 °C and 58°C (Tm − 5°C) for 50 seconds, and an elongation step at 72°C for 50 seconds. PCR reactions were preceded by a 5-minute melting period at 95°C and ended with an elongation period of 10 minutes at 72°C. The PCR products were resolved on 1% agarose gels, and bands were visualized by staining with ethidium bromide. 

In Silico Analysis

Clones were sequenced and analyzed in silico using BLAST against the Ensembl database (http://www.ensembl.org). ¹⁸ As well, expression analysis of candidate genes in different tissues was performed using the Source Web site (http://source.stanford.edu/cgi-bin/source/sourceSearch/ provided in the public domain by Stanford, University, Stanford, CA). 

Results

NACE Assay Specificity

To confirm the specificity of the enriched chromatin obtained by NACE, NPCE cells were transiently transfected with a FOXC1-pcDNA His/Max vector, ¹⁹ which expresses 6xHis-Xpress–tagged FOXC1, along with a reporter gene (luciferase) containing upstream FOXC1 binding sites (pGL3-TK-FOXC1). ¹⁹ The enriched chromatin was then analyzed by PCR amplification of the DNA flanking the FOXC1 binding site. The forward primer was designed in the pGL3-Basic vector upstream of the KpnI site, whereas the reverse primer was located in the TK promoter within the pGL3 vector. A product of 251 bp was obtained as expected in the positive control but not in samples without Ni²⁺-NTA or in the DNA mock samples (Fig. 2) , confirming the specificity of the NACE assays. Furthermore, sequence analysis confirmed the identity of the obtained bands. 

NPCE cells express FOXC1 endogenously (Fig. 3A) and therefore are an excellent reagent to use for FOXC1 target gene discovery. To isolate candidate target genes regulated by FOXC1, pcDNA FOXC1 was transfected into NPCE cells, as described earlier. One hundred fifty genomic DNA clones representing putative FOXC1 target sequences were randomly selected for sequence analysis using M13 forward and reverse primers. The size-range of the clones obtained varied from 500 to 1500 bp. 

In silico analysis of the sequences revealed 104 clones yielding sequences of poor quality or containing only repetitive elements, precluding further analyses. However, 26 clones were uniquely mapped in proximity and 5′ of known genes (Table 2) , and eight clones were mapped near predicted but uncharacterized genes. Four clones were chimeric, and eight were located within areas where no genes (known or predicted) have yet been mapped. According to published data and in silico analyses, 20 (76.9%) of the 26 known candidate genes are expressed in the eye and contain a potential FOX DNA binding site, consistent with their being targets of FOXC1 regulation of ocular gene pathways (Table 3) . 

Validation of Putative FOXC1 Target Genes Obtained from the NACE Microlibrary Screening

To determine whether the obtained potential targets would be consistently detected within the enriched chromatin, 5 of the 26 known candidate genes were subjected to PCR amplification to verify enrichment in the NACE eluate. Oligonucleotide primers were designed with the sequence of the 5′ upstream elements. Outputs from two independent NACE experiments were used as the DNA template. All the five candidates selected for validation, (BMP2K, DACH, FVT-1, SIX-1, and PGE-2 receptor) were enriched in both NACE assays (Fig. 4) . PCR failed to amplify these genes in a negative control consisting of enriched chromatin derived from cells transfected by an empty vector. These results confirm the reliability of the NACE technique and are consistent with the hypothesis that these genes are regulated by FOXC1 in the eye. 

To test if these selected genes were endogenously expressed in NPCE cells, primers were designed in the coding region of each gene and used for RT-PCR to amplify the NPCE cell mRNA template. All five NACE-selected genes were expressed by NPCE cells (Fig. 3) , consistent with these genes’ being potential FOXC1 targets in NPCE cells. 

Investigation of Selected Candidate FOXC1 Target Genes

The FOXC1-enriched chromatin was also used as a template for PCR amplification of nine candidate genes selected from the literature as potential ocular FOXC1 target genes, on the basis of their overlapping expression with FOXC1 and known role in the eye. Primers were designed within an ∼1-kb area 5′ upstream of the coding region of BMP4, BMP7, clathrin heavy chain, COX2, iNOS, SMAD2, TGF-B1, TYR, and WNT6 genes. Four (44.4%) genes (BMP7, SMAD2, TGF-B1, and WNT6) were detected in NACE-enriched chromatin consistently, in two independent lots. BMP4 displayed ambiguous expression, whereas TYR, clathrin heavy chain, COX2, and iNOS were not detected. Bands of the expected sizes were obtained exclusively in the enriched samples obtained from the nickel column, whereas no signal was detected in samples transfected with an empty vector or in the negative control samples (Fig. 5) . 

Discussion

NACE for Enrichment of Chromatin Regulated by Transcription Factors

Previous identification of 5′ upstream elements regulated by mammalian transcription factors has mainly relied on differential expression studies. ^{42 43} Recently, ChIP assays, either alone ⁶ or coupled to microarray analysis, ^{7 8} have been used to elucidate the target sequences regulated by transcription factors. Little progress has been made, however, in the study of gene regulation involved in eye diseases, including glaucoma, using this robust technique. This is principally due to an absence of suitable antibodies to relevant transcription factors. This prompted us to develop an antibody-free affinity-enrichment scheme based on Ni²⁺-NTA affinity chromatography. In the current study, we used NACE to isolate candidate target sequences bound by recombinant FOXC1. The use of NACE has several advantages, including an active conformation due to the immobilized (His)-₆-tagged biomolecules and optimal access of binding domains to potential interacting proteins. ⁴⁴ The tight-binding, 6xHis-tagged biomolecules-Ni²⁺ ions are not dependent on the three-dimensional conformation of the transcription factor and can be used under native or denaturing conditions, allowing stringent washing. ⁴⁴ Furthermore, the small size of the (His)-₆ tag does not interfere with protein structure or function in most cases and can be detected by several antibodies. ⁴⁵ In addition to the use of NACE in situations in which an antibody is not available, a major advantage is the utility of the NACE method for high-throughput screening of multiple transcription factors. 

However, like all sophisticated techniques, the NACE procedure has some limitations, mainly the potential of isolating, by means of transient transfection, false positives due to an overexpression of protein. Therefore, the choice of an appropriate cell line, in addition to strict controls, should be made carefully to avoid false results. To overcome these difficulties, we chose NPCE cells, a line that expresses FOXC1 endogenously. These cells were also used to validate our potential NACE-enriched targets. All the selected genes were expressed in NPCE cells, consistent with their being candidate targets of FOXC1 (Fig. 3) . Moreover, several controls were included to monitor the potential residual noise and to avoid harvesting nonspecific enriched chromatin. In addition to an empty-vector, negative control (Fig. 2) , we transfected cells with the FOXC1 vector lacking the forkhead box (ΔBOX) ¹⁹ as well as a reporter without an upstream FOXC1 binding site to estimate residual noises. ΔBOX FOXC1 did not detect the FOXC1 binding site located upstream of the luciferase reporter gene (data not shown). Last, it is important to remember that NACE, like all chromatin immunoprecipitation-based protocols, is an enrichment method. All recovered candidate targets must be validated, including demonstration of reproducible enrichment, expression in the starting cell material, and detection of potential transcription factor binding sites in the recovered chromatin. 

NACE was successful in isolating chromatin-containing recombinant (His)-₆-Xpress FOXC1 in NPCE cells. Of the known genes isolated, 76.9% were expressed in eye tissues (Table 3) . Furthermore, all five selected candidate genes were found to be enriched by NACE in two independent experiments (Fig. 4) , confirming the consistency of this technique. In parallel, nine genes were selected from the literature as being potential FOXC1 targets based on their known function and expression in the eye. NACE failed to detect four of the selected genes (clathrin, COX2, iNOS, and TYR), suggesting that these genes are not regulated by FOXC1. Alternatively, the failure to detect these genes may be simply because the chosen primers did not cover the location of the FOXC1 regulatory elements. BMP4 systematically showed an ambiguous pattern, despite stringent conditions. However, among the selected genes, four (BMP7, SMAD2, TGF-B1, and WNT6) were detected consistently in two different lots of NACE-enriched chromatin (Fig. 5) . Putative FOX binding sites (AAAT/CA) were found in the detected 5′ upstream regulatory elements. NACE recovery of these genes, together with their function, expression data, and putative FOX binding sites in their upstream regions, are consistent with the hypothesis that these genes are regulated by FOXC1 in the eye. 

Implication of Roles for FOXC1 in Regulation of the Cell Cycle and Cancer

We identified a genomic DNA element immediately 5′ of the follicular variant translocation protein gene (FVT-1) that was detected in two independent NACE experiments (Fig. 4C) . This represents the first evidence that FVT-1 may be subjected to regulation by FOXC1. The FVT-1 gene has been shown to be involved in cancer through chromosome translocation t(2;18) involving the 5′ region of the BCL-2 gene located 10 kbp downstream of FVT-1 on chromosome 18. ⁴⁶ In a recent report, FOXC1 was shown to be upregulated by TGF-β1 in several human cancer cell lines. ⁴⁷ In addition, homozygous deletion of FOXC1 was found in 6.7% of primary endometrial and ovarian cancers. ⁴⁷ As TGF-β1 was also enriched in NACE eluates, this pathway could form an important feedback loop involving FOXC1 in cell regulation. Consistent with this hypothesis, BMP7 and SMAD2, known to be involved in the TGF-β pathway, ^{48 49} were also enriched in the NACE eluates. Further studies may clarify the role of FOXC1 in human oncogenes. 

Role of FOXC1 in the Regulation of Intraocular Pressure

In two independent NACE experiments, we identified a NACE-enriched clone uniquely mapped to a region within the second intron of the prostaglandin E2 receptor (PTGER) gene (Fig. 4E) . PTGER is expressed in the ciliary body and plays a key role in modulating intraocular pressure (IOP) in humans. ⁵⁰ The corresponding ligand (PGE2) binds PTGER to initiate a cascade of events leading to IOP reduction in mice. Topical application of PGE2 in the mouse eye markedly reduces IOP and relaxes precontracted ciliary muscle. ²⁴ Prostaglandin has been approved for clinical use to increase drainage of intraocular fluid, probably through receptor E2 or E4, which through unknown mechanisms, promotes ciliary muscle relaxation and evacuation of the aqueous. ²⁸ It is tempting to speculate that failure of correct regulation of the PTGER gene by FOXC1 in patients with FOXC1 mutations may have a role in these patients’ elevated IOP and glaucoma. Additional experiments are needed to address this interesting question. 

Role of FOXC1 in the Regulation of Ocular Development and Maintenance of Cell Identity

The enriched 5′ upstream element of Dachshund homologue isoform A (DACH1) gene was approximately 18 kb upstream of the coding region and was consistently detected in NACE products (Fig. 4B) . DACH is expressed in several organs and malignant tissues, and particularly in the fetal eye, retinal fovea, macula, lens, eye anterior segment, and optic nerve. ⁵¹ Dach1 is found in the developing eye and ear of both chick and mouse ⁵² and has a role in retinal and pituitary precursor cell proliferation. It is also expressed in breast cancer cell lines, where it inhibits TGF-β signaling through binding Smad4. ³⁴ Moreover, Dach, Six, and Eya act synergistically through Creb-binding protein to activate downstream target promoters. ⁵³ Upstream elements of SIX1, SMAD4, and TGF-B1 were also detected in the enriched chromatin, implying that FOXC1 is upstream in this gene regulatory pathway. 

Consistent with this potential role of FOXC1 regulation of developing important genes, we also identified a 5′ upstream element approximately 2 kb upstream of the initiation codon of Sin oculis homeobox homolog 1 (SIX1), which was detected consistently in two NACE products (Fig. 4D) . Six1 is required for the early organogenesis of mammalian kidney ⁵⁴ and has a synergistic effect with Dach2 and Eya2 on the regulation of vertebrate muscle development as well as on Drosophila eye formation. ³⁵  

WNT6 was also reproducibly enriched by NACE from NPCE cells transfected with FOXC1, suggesting the potential regulation of WNT6 by FOXC1 in the eye. WNT6 is a member of a highly conserved family of developmental control genes, encoding secreted glycoproteins that are involved in signaling through Frizzled receptors. ⁵⁵ There is clear evidence that WNT genes are involved in the eye. ⁵⁶ Wnt5a, -5b, -7a, -7b, -8a, and -8b, as well as Frizzled receptors 1, 2, 3, 4, and 6, are detected in the lens, ⁵⁷ whereas Wnt2b controls retinal cell differentiation. ⁵⁸ Furthermore, it has been hypothesized that WNT6 also plays key roles in human carcinogenesis. ⁵⁹ The role that WNT6 has in the anterior chamber is not well understood at present. This is the first report showing a possible link of WNT6 to FOXC1, implying that FOXC1 has a role in regulating WNT6 expression to fulfill its developmental role in different organs, including the eye. Further experiments are in progress to investigate the interaction of these molecules in the regulation of target genes in the eye. 

NACE: A Useful Technique to Isolate Genes Regulated by Transcription Factors

The present findings reveal the ability of the NACE method to isolate recombinant (His)-₆-tagged FOXC1–enriched chromatin complexes. We propose that NACE can be an alternative to ChIP in cases in which suitable antibodies are unavailable. 

Further experiments are needed to investigate these potential targets of FOXC1 by the use of reporter gene assays, gel shift, and gene modeling of knockout mice. With these methods, it would be fruitful to explore the potential roles of FOXC1 in IOP regulation through regulation of PTGER and in the regulation of eye development through control of key genes, including DACH1, SIX1, and WNT6. Our results are consistent with the hypothesis that FOXC1 is directly involved in the regulation of a wide range of genes expressed in the eye, reinforcing the crucial role of FOXC1 in proper ocular development and function. ^{60 61}  

 Supported by the Alberta Heritage Foundation for Medical Research, the Canadian Institute for Health Research, and Novartis Ophthalmics.

 Submitted for publication May 31, 2004; revised July 22 and July 27, 2004; accepted July 30, 2004.

 Disclosure: Y. Tamimi, None; M. Lines, None; M. Coca-Prados, None; M.A. Walter, Novartis Ophthalmics (F)

 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.

 

Schematic representation of the NACE procedure. DNA-binding proteins are cross-linked to DNA with formaldehyde in vivo. Stars: recombinant (His)-₆-tagged FOXC1, hexagons represent other transcription factors present in the cells. After isolating the chromatin, the DNA along with bound proteins is sheared into small fragments by sonication. FOXC1 chromatin complexes are specifically isolated by nickel agarose (polygons). FOXC1-enriched chromatin is obtained after eluting and reversing the cross-linking to release the DNA and digest the proteins.

Table 1.

 

Sequences of the Oligonucleotide Used for PCR Amplification Assays

Gene	Forward Primer	Reverse Primer	Product Size (bp)
01 BMP-2K	5′-tgctggttggaccattgtta-3′	5′-gtgctaatgaaactaaggat-3′	527
02 BMP-2K	5′-tgaagaagttctctcggatgc-3′	5′-tacgcacgaggaaaactgtg-3′	207
03 BMP4	5′-cggggaagaggaggagga-3′	5′-cgtccctcagctcggatg-3′	150
04 BMP7	5′-cgtctgcagcaagtgacc-3′	5′-ctcgttgtccaggctgaagt-3′	216
05 Clathrin	5′-ttttgatgggcatgaatgag-3′	5′-ctgtgctgctggagtgacat-3′	207
06 Cox2	5′-atcacaggcttccattgacc-3′	5′-caggatacagctccacagca-3′	176
07 DACH	5′-aatatgaagcataatatgaa-3′	5′-ggaactccgcctgtagtgac-3′	530
08 DACH	5′-ggtgtgcaatgtggaacaag-3′	5′-cttaggaggccttccaggtc-3′	151
09 Ephrin	5′-cttgatctcagcctcccaag-3′	5′-ccagggatagtgtccatgctc-3′	310
10 FOXC1	5′-atcaagaccgagaacggtac-3′	5′-gtgaccggaggcagagagta-3′	635
11 FOXC1 binding site	5′-agtgcaggtgccagaacatt-3′	5′-caaaccctaaccaccgctta-3′	251
12 FVT-1	5′-ctcgcatctggcgctgtccg-3′	5′-gcctggagtgtttgggttt-3′	281
13 FVT-1	5′-gtgctgctgctgtacatggt-3′	5′-atggaaagcaccacctgttt-3′	233
14 iNOS	5′-aggaggagatgctggagatg-3′	5′-acatccccgcaaacatagag-3′	178
15 Prostaglandin	5′-cacacctgctgccagagtta-3′	5′-aaatcagagccttgcaggaa-3′	160
16 Prostaglandin	5′-cacacacggagaagcagaaa-3′	5′-atgtgatcctggcagaaagg-3′	175
17 SMAD2	5′-tgctgcgtttggtaagaaca-3′	5′-tttcactgctttctcacacca-3′	153
18 SIX-1	5′-gatgacttctggaatcaaga-3′	5′-gagaccagcctgagcaacat-3′	450
19 SIX-1	5′-aaggagaagtcgaggggtgt-3′	5′-tgcttgttggaggaggagtt-3′	206
20 TGF-β1	5′-gagctggtcgggagaagag-3′	5′-ctgagggacgccgtgtag-3′	161
21 WNT6	5′-ccggctctgatttcttctcc-3′	5′-gcatcgtgaccgccctac-3′	245

 Primers were either designed in the coding region (2, 8, 10, 13, 16, and 19) to test the expression of the corresponding gene in NPCE cells, or at the 5′ upstream of the coding region (1, 3, 4, 5, 6, 7, 9, 12, 14, 15, 17, 18, 20, and 21) to analyze their detection in NACE products. Primers 11 were designed in a region flanking the FOXC1 binding site.

Figure 2.

 

NACE specificity assay. NPCE cells were transfected with (His)-₆-tagged FOXC1 and enriched by NACE. The enriched chromatin was subjected to PCR amplification with primers flanking the FOXC1 binding site. Bands of the expected sizes (arrow) were obtained specifically in FOXC1-transfected samples enriched by NACE. Chromatin enriched from NPCE cells transfected by an empty vector was used as the negative control along with water and nickel agarose beads (mock).

Figure 3.

 

Expression of selected genes in the NPCE cell line. Confluent NPCE cells were harvested and subjected to RNA extraction. Reverse transcription was performed with oligo dT primers and amplified to monitor the expression of FOXC1 (A) as well as the selected genes (BMP2K, DACH, FVT-1, PGER2, and SIX-1) validated in NACE products (B).

Table 2.

 

NACE-Enriched Clones Matching Known Genes

Chromosome	Clone	Putative Identification [Function]
1	106∗	Ephrin receptor [cell–cell signaling]
	138∗	Prostaglandin E2 receptor (PTGER) [signal transduction]
2	147∗	Dynein [chromosome alignment and spindle organization during mitosis]
	54∗	Pleckstrin, (PLEK) [protein kinase C substrate, exact function unknown]
3	13∗	Roundabout, axon guidance receptor, homolog 1 (Drosophila), ROBOI [axon guidance]
	16	5 Azacytidine-induced gene 2 [unknown]
	85∗	HRAS-like suppressor (HRASLS) [cell signaling]
	123	CG158 [aminopeptidase, hydrolase activity)
4	132∗	Bone morphogenic proteins-2-inducible kinase (BMPK2) [may be involved in osteoblast differentiation]
5	23∗	RAS GTPase activating protein 2 [signal transduction]
9	11	SMC5 [ATP binding, chromosome segregation]
	51∗	Potassium channel KV11.1 [modulate channel activity]
	104∗	Contactin-associated protein-like 3 [belongs to the neuroxin family]
	142	DVS27-related protein [unknown]
10	86∗	TANKYRASE 2 (TNKS2) [regulation of vesicle trafficking]
11	6∗	Fibroblast growth factor 19 (FGF-19) [development]
13	67	Dachshund homolog (Drosophila) (DACH) [development]
14	63∗	Sine oculis homeobox homolog 1 (Drosophila) (SIX1) [development]
	80∗	Protein kinase C [signal transduction]
15	58∗	RAR-related orphan receptor A RORA (nuclear receptor ROR-α)/(Nuclear hormone receptor)
16	141∗	IRX-5 (iroquois homeobox protein 5) [transcription]
17	17∗	MAGUK P55 [signal transduction]
	22∗	Brain angiogenesis inhibitor (BAI-1) [proliferation regulation]
18	36∗	Follicular variant translocation protein (FVT1) [dehydrogenase, reductase]
	136∗	MYOSIN [muscle contraction]
22	107∗	Gamma-PARVIN [regulation of cell adhesion]

∗  Conserved between human and mouse.

Table 3.

 

Known Gene Expression Distribution

	Aort	Blood	Brain	Eye	Heart	Kidney	Liver	Lung	Muscle	Ovary	Pancreas	Placenta	Spinal Cord	Spleen	References
Ephrin receptor			•	•		•	•	•							20 21 22
PLK	•			•	•		•					•			23
PTGER3				•	•										24
Dynein				•											25 26
ROBOI		•	•	•	•		•	•							27
HRASLS			•	•				•							28
BMP2K		•	•	•						•					29
RAS GTPase AP				•											30
K channel KVII.1				•											31 32
Contactin AP-like 3				•											33
TNKS2			•	•						•		•	•	•	Source
FGF19				•											Tamimi et al, unpublished data
DACHI			•	•				•							34
SIXI			•	•				•	•						35 36
PKC				•					•						37
ROR-α				•											38
IRX				•											39
MAGUK				•											40
FVT-I				•				•		•	•	•			Source
Myosin				•					•						41

 Data were obtained from the Source Web site and/or published data.

 •, expression of the gene in the tissue.

Figure 4.

 

Validation of five NACE-enriched genes and the location of the 5′ upstream elements. Schematic representation of the 5′ upstream elements present in NACE-enriched target DNA for the following genes: (A) BMP2K, (B) DACH, (C) FVT-1, (D) SIX1, and (E) PTGER. Lane M: 100-bp DNA marker; lane 2: +Ni²⁺ enriched DNA lot1; lane 3: +Ni²⁺ enriched DNA lot2; and lane 4: +Ni²⁺ empty vector. Arrows, right: molecular sizes; arrows, left: locations of the NACE-enriched 5′ upstream elements.

Figure 5.

 

NACE analysis of candidate FOXC1-regulated genes selected from the literature. PCR amplification of genes thought to play a role in eye development pathways. Oligonucleotide primers were designed in the 5′ upstream region and used to amplify the two different lots of NACE-enriched chromatin. BMP7, SMAD2, TGF-B1, and WNT6 could be detected consistently in two different NACE-enriched samples.

The authors thank all members of Ocular Genetics Laboratory for help and discussions during the project, Fred Berry for help and discussion during the initiation of the project and for critical reading of the manuscript, Alan Underhill for comments during the preparation of the manuscript, Farideh Mirzayans and Tim Footz for technical help and support.