| Nucleic Acids Research | Pages |
DNA binding properties of a chemically synthesized DNA binding domain of hRFX1
Introduction
Materials And Methods
Peptide synthesis
NMR experiments
Gel mobility shift assays
Surface plasmon resonance experiments
Determination of kinetic constants
DNA duplex melting temperature
Results
Peptide synthesis
NMR experiments
Gel mobility shift assays
Surface plasmon resonance experiments
DNA duplex melting temperature
Discussion
Acknowledgements
References
DNA binding properties of a chemically synthesized DNA binding domain of hRFX1
ABSTRACT
INTRODUCTION
The RFX family of proteins consists of highly homologous site-specific DNA binding proteins designated RFX1-5 in man and mouse (1,2) Sak1 in Shizosaccharomyces pombe (3), ScRFX in Saccharomyces cerevisiae and CeRFX in Caenorhabditis elegans (4). These proteins are characterized by several highly conserved sequences, among which the most prominent is a novel 76 amino acid DNA binding domain (DBD) designated the RFX DNA binding motif (1,2). The recent identification of several new RFX genes has permitted a precise definition of the consensus motif present in the DBD and has shed light on the evolutionary conservation and functional importance of RFX proteins (4). These proteins function as regulatory factors in a wide variety of unrelated systems, including regulation of the mitotic cell cycle in yeast fission, control of the immune response in mammals and promotion of infection by human hepatitis B virus. Thus RFX1 was shown to be a cellular transactivator used by the highly pathogenic hepatitis B virus (5,6), while RFX5, a 75 kDa subunit of a nuclear protein complex, is involved in transcription of MHC class II, Ii chain and DM genes (7-9). These genes play a key role in the immune system by controlling presentation of foreign antigenic peptides to CD4+ helper T lymphocytes. The RFX family of proteins appears, therefore, as widespread and functionally important group of proteins as those containing other well-known DNA binding motifs, such as the well-known zinc finger, homeodomain, basic leucine zipper and basic helix-loop-helix proteins.
The DNA binding motif shared by RFX1-5, Sak1, ScRFX and CeRFX proteins has no significant homology with any other known DNA binding motif (4; Fig.
Figure 1. (A) Alignment between the DBD sequences of all known human RFX proteins. Amino acids that are identical or similar are boxed. The secondary structure prediction (20) is indicated below the sequences. Positions of the predicted [alpha]-helices and [beta]-strand (extended) are indicated by boxes. (B) 23 bp DNA duplex containing the EF-C/MDBP target site of hRFX1/DBD. The presence of an inverted repeat in the sequence allows hRFX1 to bind as a homodimer (left). In order to simplify this model, our aim in this study was to obtain a 1:1 stoichiometric complex between the DBD of hRFX1 and one half-site of the EF-C/MDBP target site (right). The human hRFX1 protein binds DNA as homodimeric or heterodimeric (RFX1/RFX2 and RFX1/RFX3) complexes whose specific binding sites are inverted repeats, referred to as either EF-C or MDBP, or MHC class II X box motifs (Fig. In order to characterize more precisely the mode of DNA recognition by proteins of the RFX family, the DBD of hRFX1 (hRFX1/DBD) was chosen as a model. Synthesis of this domain (76 amino acids) was achieved by the solid phase method, which required optimization of several coupling steps. The resulting peptide was obtained in large quantities (90 mg), allowing us to perform preliminary 600 MHz 2D 1H NMR experiments, suggesting the occurrence of helical regions in the sequence. The synthetic peptide was also used to determine the minimum DNA binding site in order to simplify study of the interaction of hRFX1/DBD with DNA. The biotinylated 13mer DNA duplex characterized in this preliminary study was then immobilized on a streptavidin-coated biosensor chip in order to determine its kinetic binding parameters with hRFX1/DBD by surface plasmon resonance analysis and its selectivity for a double-stranded DNA. Finally, DNA duplex denaturation experiments were undertaken in order to investigate the effect of hRFX1/DBD on duplex stability.
MATERIALS AND METHODS
Peptide synthesis
Wang resin, piperidine, N-methylpyrrolidone, dichloromethane, dimethylaminopyridine, dicyclohexylcarbodiimide and 1-hydroxybenzotriazole were purchased from Perkin-Elmer (Saint Quentin en Yvelines, France). Trifluoroacetic acid was from SDS (Peypin, France) and triisopropylsilane and phenol were from Aldrich (Strasbourg, France).
Assembly of the protected peptide chain was carried out using the stepwise solid phase method of Merrifield (10) on an Applied Biosystems 431A automated peptide synthesizer. The different syntheses were run at 40 µmol scale on Wang resin (loading 1 mmol/g dry resin). Fmoc-protected amino acids were used with the following sidechain protection groups: t-butyl ether (Ser, Thr and Tyr); t-butyl ester (Glu and Asp); trityl (Cys, His, Asn and Gln); 2,2,5,7,8-pentamethylchroman-6-sulfonyl-5-(Arg), t-butoxycarbonyl (Lys and Trp). The hydroxy function on Wang resin was esterified by Fmoc alanine-activated as a symetrical anhydride with dimethylaminopyridine as the acylation catalyst. Deprotection of the Fmoc group during complete synthesis was obtained by three successive 3 min treatments with 20% piperidine in N-methylpyrrolidone. Successive couplings were performed with dicyclohexylcarbodiimide/hydroxybenzotriazole in N-methylpyrrolidone as coupling agent. Single couplings of 45 min were used (total cycle 1 h 15 min). N,O-bisFmoc derivatives of N-(2-hydroxy-4-methoxybenzyl)glycine was used for coupling of Gly53 in the second synthesis and for coupling of Gly44 and Gly53 in the third synthesis. In both cases coupling of the amino acid immediately following these glycines was performed using the symetric anhydride procedure with dicyclohexylcarbodiimide in dichloromethane as described in Johnson et al. (11). Fmoc deprotection in these syntheses was followed by spectrophotometric monitoring of the dibenzofulvene piperidine adduct at [lambda] = 301 nm (Fig.
Figure 2. Ultraviolet monitoring ([lambda]301 nm) of Fmoc deprotection during the three different syntheses of the DBD of hRFX1. The scheme is limited to the 38-55 region of hRFX1/DBD because the difficult couplings and deprotections were located only in this part of the synthesis. The sequence 38-55 of hRFX1 is represented in the reverse orientation, which corresponds to the order of incorporation of solid phase synthesis from the C- to the N-terminus. (A) Fmoc deprotection scheme of the first synthesis. Difficult couplings and deprotections are located from Ser49 to Gly44. (B) Incorporation of a 2-hydroxy-4-methoxy-benzyl as amino protecting group of Gly53 prevented any synthesis problem until Gly44, but new synthetic difficulties occurred at Phe43-Ala40. (C) The use of 2-hydroxy-4-methoxy-benzyl groups at Gly53 and Gly44 definitively solved the synthesis problems. The double intensity of peak heights corresponding to deprotection of Gly53, Met52, Gly44 and Phe43 is due to the presence of a N,O-bisFmoc on the glycine derivative, and the subsequent N,O double coupling of Fmoc methionine and Fmoc phenylalanine is due to the activation procedure of these residues as symmetrical anhydrides. Figure 3. HPLC chromatograms of the crude (A) and purified (B) products from the third synthesis of hRFX1/DBD. The analyses were obtained with an acetonitrile gradient of from 50 to 80% B over 15 min at a flow rate of 1 ml/min on a capcell C8 column (5 µm, 300 Å, 250 × 4.6 mm) with UV detection at [lambda] = 214 nm; A, 0.1% TFA in H2O; B, 70% CH3CN, 0.09% TFA in H2O. The retention time of the purified hRFX1 is 9.7 min under these conditions. (C) Ionisation electrospray mass spectrum of the purified hRFX1/DBD.
NMR experiments
hRFX1/DBD was dissolved in water containing 30% TFE-d2 (SDS, Peypin, France) to a final sample concentration of 1 mM. Two-dimensional phase-sensitive 1H Clean-TOCSY (70 ms spin lock) (12) and NOESY (mixing times 100 and 200 ms) (13) spectra utilizing time-proportional phase incrementation (14) were recorded at 293 K and pH 4.0.
NMR spectra were recorded on a Brüker AMX 600 spectrometer operating at 600.14 MHz without sample spinning with 2K real t2 points, with a spectral width of 7246 Hz, and 512 t1 increments. The transmitter frequency was set to the water signal, which was suppressed by irradiation during the relaxation delay of 1.6 s between scans, and a water gate sequence before acquisition. The temperature was externally controlled using a SK107 unit (Haake GmbH, Karlsruhe, Germany). The data were processed with FELIX (Biosym/MSI). Phase-shifted sine bell (shifted by [pi]/6) window functions were applied prior to Fourier transformation in both dimensions. The 1H chemical shifts were referenced to hexamethyldisilazane (HMDS; SDS, France), used as internal standard.
Gel mobility shift assays
Binding conditions and gel electrophoresis were carried out essentially as described (2,5). Briefly, 100 000 c.p.m. (50 pg) double-stranded oligonucleotide probes (end-labeled with 32P) were diluted with known amounts (see figure legends) of unlabeled oligonucleotide and incubated for 30 min on ice with 10 ng hRFX1/DBD in 20 µl binding buffer. Binding buffer was as described (2,5). Sequences of the duplex oligonucleotides were as follows: 23mer, 5[prime]-GGCCAGTTGCCTAGCAACTAATT-3[prime]; 16mer, 5[prime]-CCCCTAGCAACAGATG-3[prime]; 13mer, 5[prime]-CCCCTAGCAACAG-3[prime]; 12mer, 5[prime]-CCCTAGCAACAG-3[prime]. Only the upper strand is shown. Half-sites are underlined.
Surface plasmon resonance experiments
The BIAcore system used in these studies is from Pharmacia Biosensor AB (Uppsala, Sweden). Sensor chips and surfactant P20 were from Pharmacia Biosensor AB. N-Hydroxysuccinimide (NHS), N-ethyl-N[prime]-(3-diethylaminopropyl)carbodiimide (EDC), diaminopropane and streptavidin were purchased from Sigma. The complementary 13mers 5[prime]-CCCCTAGCAACAG-3 and 5[prime]-CTGTTGCTAGGGG-3[prime], biotinylated or not, were from Birsner & Grob-Biotech GmbH (Denzlingen, Germany).
The surface of the CM5 sensor chip was derivatized with streptavidin (2300 RU) using the EDC-NHS procedure according to the manufacturer's recommendations, followed by inactivation of the uncoupled dextran carboxylic groups with 1 M diaminopropane in H2O.
Hybridization of 2.5 µg 13mer 5[prime]-biotin(5[prime]-CTGTTGCTAGGGG-3[prime]) was performed using 3.75 µg complementary 13mer (5[prime]-CCCCTAGCAACAG-3[prime]) in 500 µl 50 mM HEPES, pH 7.4, 0.5 M NaCl, 5 mM MgCl2. After 1 h at 60°C, the temperature was allowed to decrease overnight to room temperature. Three different amounts of double-strand biotinylated oligonucleotide were immobilized on the sensor chip using 20 µl stock solution diluted 1:10 with prehybridized DNA which were injected at a flow rate of 5 µl/min over the three channels of the sensor chip. This led to immobilization of 120, 140 and 170 RU respectively.
Binding experiments were carried out at 25°C in 10 mM HEPES, pH 7.4, 0.15 M NaCl, 10 mM MgCl2, 0.3 mM DTT, 3 mM EDTA, 0.005% P20. For kinetic studies 40 µl solutions of HRFX1 at concentrations ranging from 50 to 800 nM were injected at 20 µl/min. Concentration and ratio between peptide and immobilized nucleic acids were determined using the equation:
| n = (RUpeptide/RUDNA) × (MWDNA/MWpeptide) | 1 |
After each injection of protein over the DNA, the sensor chip surface was regenerated by a 5 µl pulse (15 s) of 1 M NaCl in binding buffer.
Determination of kinetic constants
Apparent kinetic rates were calculated using the BIAcore 2.0 analytical software. To fit the dissociation curves, a single exponential dissociation rate equation
| y = Re-d |
2 |
For determination of the association rate constant Ka, the binding curve of the sensorgram was fitted to the exponential function
| R = Req(1 - e-Ks(t-t0) | 3 |
| Ks = KaC + Kd | 4 |
DNA duplex melting temperature
DNA denaturation experiments were achieved with a UVIKON 941 UV spectrophotometer piloted by KONTRON acquisition software for melting temperatures. Aliquots of 1.5 ml of a 1.5 µM solution of the 13mer oligodeoxynucleotide 5[prime]-CCCCTAGCAACAG-3[prime]/5[prime]-CTGTTGCTAGGGG-3[prime] solution and 2 µM hRFX1/DBD in 10 mM HEPES, pH 7.4, 1 mM [beta]-mercaptoethanol were heated over a gradient of temperature ranging from 4 to 65°C at 20°C/h. Two other cells containing the same concentration of oligodeoxynucleotide or hRFX1/DBD respectively were used as controls. The absorbance at [lambda]260 nm, which corresponds to maximum absorbance of nucleotide bases, was expressed in relative units and recorded. The absorbance of hRFX1/DBD alone at the same concentration (Do [sim] 0.01 RU) was recorded and substracted from the values obtained with DNA in the presence of hRFX1/DBD.
RESULTS
Peptide synthesis
Solid phase synthesis of the DBD of hRFX1 was carried out in a stepwise manner on Wang resin using the Fmoc/tBu strategy (15). Figure
NMR experiments
In a preliminary assay, the NMR spectrum of synthetic hRFX1/DBD was achieved in 20 mM acetate buffer, 0.3 mM DTT, pH 5.4, H2O/D2O 9:1 at 293 K. A large broadening of the signals, indicating aggregation, was observed under these conditions (data not shown). In 30% trifluoroethanol in water, 5 mM DTT, pH 4, at 293 K (data not shown) the linewidths of the signals were sharper. Figure
Figure 4. NH-NH region of a two-dimensional 600 MHz 1H NMR NOESY spectrum corresponding to a solution of the synthetic hRFX1/DBDin H2O/TFE 7:3, pH 4, 5 mM DTT at 293 K. The multiplicity of NH-NH NOE cross-peaks are indicative of helical structure.
Gel mobility shift assays
To determine whether the synthetic hRFX1/DBD polypeptide is biologically functional, i.e. capable of binding specifically to DNA, we performed gel mobility shift assays (Fig.
Figure 5. (A) Gel mobility shift assays were performed with 10 ng hRFX1/DBD and 100 000 c.p.m. (50 pg) 23 bp double-stranded oligonucleotide containing a palindromic binding site as shown in Figure 1B. The labeled probe was diluted with the quantities of unlabeled 23 bp oligonucleotide indicated above the lanes. In the left lane (-) no protein was added. Positions of free DNA (F) and protein-DNA complexes resulting from binding of a monomer (M) or a dimer (D) of hRFX1/DBD are indicated on the right. (B) Gel mobility shift assays were performed without (-) or with (+) hRFX1/DBD and double-stranded oligonucleotides of different lengths (16, 13 and 12 bp) containing a single half-site. Positions of free DNA (F) and the protein-DNA complex resulting from binding of an hRFX1/DBD monomer (M) are indicated on the right. Under the conditions used essentially all of the 16 and 13 bp oligonucleotides are complexed with hRFX1/DBD, while binding to the 12 bp oligonucleotide is severely impaired. Further gel shift experiments were performed to determine the minimum size of the DNA fragment required for binding of a monomer of hRFX1/DBD. A constant amount of hRFX1/DBD was incubated with double-stranded oligonucleotides of different sizes, each containing the same EF-C/MDBP half-site. Oligonucleotides of 20, 19, 16, 13, 12 and 10 bp were tested. Only the results for oligonucleotides of 16, 13 and 12 bp are shown in Figure 
Surface plasmon resonance experiments
In order to study the structural basis of the affinity and selectivity of the interaction between DNA and hRFX1 protein, surface plasmon resonance (SPR) measurements were used (18). The minimal sequence of DNA duplex able to interact with hRFX1 (5[prime]-CCCCTAGCAACAG-3[prime]/5[prime]-CTGTTGCTAGGGG-3[prime]) was biotinylated at the 5[prime]-end of the first strand and immobilized onto streptavidin-coated chips of a BIAcore biosensor apparatus. Figure
Figure 6. Determination of the kinetic parameters of the interaction of the DBD of hRFX1 to the DNA duplex using suface plasmon resonance (SPR). The response functions (sensorgrams) shown illustrate specific binding and dissociation profiles obtained after various concentrations of hRFX1 protein were added to the buffer flowing over a biosensor chip coated with the DNA duplex (CTGTTGCTAGGGG) immobilized via a streptavividin-biotin system. The data in these sensorgrams are expressed as corrected response units, since the contribution by the bulk refractive index (background) has been substracted. For clarity only the sensorgrams obtained using 80 (bottom), 140, 185, 250, 330, 445 and 600 nM (top) hRFX1 are shown. The association constant (Ka = 3.9 ± 0.3 × 105/M/s), dissociation constant (Kd = 6.0 ± 0.24 × 10-2/s) and the equilibrium dissociation constant (KD = 153 ± 18 nM) determined by this study are the average of three independent experiments at 25°C in 10 mM HEPES, pH 7.4, 0.15 M NaCl, 10 mM MgCl2, 0.3 mM DTT, 3 mM EDTA, 0.005% P20 with immobilized amounts of DNA duplex ranging from 120 to 170 RU. The association and dissociation rate constants were calculated for different protein concentrations (185, 250, 330, 445, 600 and 800 nM respectively) using the equations reported in Materials and Methods. Based on these measurements the data fitted a first order model well, yielding the kinetic parameters of binding. The values obtained for the association constant Ka (3.9 ± 0.3 × 105/M/s), the dissociation constant Kd (6.0 ± 0.24 × 10-2/s) and the equilibrium dissociation constant KD (153 ± 18 nM) are the average of three independent experiments (Fig. 
DNA duplex melting temperature
DNA denaturation experiments were performed on the 13mer duplex 5[prime]-CCCCTAGCAACAG-3[prime]/5[prime]-CTGTTGCTAGGGG-3[prime]. As shown in Figure
Figure 7. Melting temperature of a 1.5 µM solution of DNA duplex 13mer (5[prime]-CTGTTGCTAGGGG-3[prime]) in 10 mM HEPES, pH 7.4, 1 mM [beta]-mercaptoethanol obtained by UV monitoring at 260 nm with a gradient of temperature ranging from 6 to 70°C at 20°C/h. An elevation of 16°C in this melting temperature was obtained by adding 2 µM synthetic hRFX1/DBD to this solution, indicating increased stability of the DNA duplex.
DISCUSSION
With the aim of characterizing kinetic and physicochemical constants of a novel DNA binding motif present in the family of RFX proteins, solid phase peptide synthesis of a 76mer polypeptide, hRFX1/DBD, was performed using the Fmoc/tBu strategy. In the first synthesis difficulties in the deprotection steps and incomplete couplings were observed from Gly44 to Ser49 (Fig.
Careful characterization of 76 amino acid synthetic hRFX1/DBD by HPLC and electrospray mass spectrometry (Fig.
Gel mobility shift assays were performed with the synthetic peptide showing that it was able to bind to a double-stranded 23mer oligodeoxynucleotide containing the binding site of native hRFX1. Due to the presence of inverted repeats in the DNA, the peptide was shown to form a 2:1 complex in a cooperative manner with this oligonucleotide, as is observed with entire hRFX1 protein. Using oligonucleotides of different lengths, the optimum double-stranded oligonucleotide sequence leading to a 1:1 complex with hRFX1/DBD was determined to be the 13mer sequence 5[prime]-CCCCTAGCAACAG-3[prime]/5[prime]-CTGTTGCTAGGGG-3[prime] (Fig.
This 5[prime]-biotinylated DNA 13mer was used in SPR experiments to determine its association (4 × 105/M/s) and dissociation constants (6 × 10-2/s) for the hRFX1/DBD peptide, leading to an equilibrium dissociation constant KD of 153 nM. These values are in agreement with those found for other nucleic acid binding motifs, such as the zinc finger domain of HIV-1 NCp7 (21). However, a high selectivity and a higher affinity, by at least three or four orders of magnitude, of hRFX1 proteins for the EFC/MDBP binding site are expected to result from a cooperative effect between the two subunits of the hRFX1 dimer. This was observed with several DNA binding proteins, such as Escherichia coli catabolite activator protein, which binds DNA as a dimer of two identical protomers and exhibits 2-fold symmetrical interactions with a 22 bp 2-fold symmetrical DNA site (22). The equilibrium dissociation constant of catabolite activator protein is in the subnanomolar range, which is in agreement with the calculated increase in affinity, as compared with a monomer, resulting from interaction of a dimer with two symmetrical binding sites on DNA (23).
Finally, DNA duplex denaturation experiments performed on the DNA 13mer allowed us to demonstrate significant stabilization of the DNA duplex in the presence of synthetic hRFX1/DBD peptide.
In conclusion, this study has allowed us to determine the minimum double-strand DNA binding site and the binding constants for this domain. The results of these experiments are essential for subsequent investigation of the three-dimensional structure in solution of 13C/15N-labeled hRFX1/DBD by NMR and of its complex with the 13mer oligonucleotides by X-ray crystallography. These studies are now in progress.
ACKNOWLEDGEMENTS
We would like to acknowledge Dr J.Paoletti for providing UV spectrometry facilities and for helpful discussions and E.Barras for excellent technical assistance. We thank C.Dupuis for her invaluable help in drafting this manuscript.
REFERENCES
This article has been cited by other articles:
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 17 Apr 1998
Copyright©Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
L. D. Norquay, X. Yang, P. Sheppard, S. Gregoire, J. G. Dodd, W. Reith, and P. A. Cattini
RFX1 and NF-1 Associate with P Sequences of the Human Growth Hormone Locus in Pituitary Chromatin
Mol. Endocrinol.,
June 1, 2003;
17(6):
1027 - 1038.
[Abstract]
[Full Text]
[PDF]
S. Grkovic, M. H. Brown, M. A. Schumacher, R. G. Brennan, and R. A. Skurray
The Staphylococcal QacR Multidrug Regulator Binds a Correctly Spaced Operator as a Pair of Dimers
J. Bacteriol.,
December 15, 2001;
183(24):
7102 - 7109.
[Abstract]
[Full Text]
[PDF]
R. Tierney, H. Kirby, J. Nagra, A. Rickinson, and A. Bell
The Epstein-Barr Virus Promoter Initiating B-Cell Transformation Is Activated by RFX Proteins and the B-Cell-Specific Activator Protein BSAP/Pax5
J. Virol.,
November 15, 2000;
74(22):
10458 - 10467.
[Abstract]
[Full Text]
This Article 
Abstract
Print PDF (139K)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (10)
Request Permissions
Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Articles by Cornille, F.
Articles by Reith, W.
Search for Related Content
PubMed
PubMed Citation
Articles by Cornille, F.
Articles by Reith, W.
Social Bookmarking
What's this?