Surface Plasmon Resonance Study of Cooperative Interactions of

Mar 30, 2009 - Institute of Materials Research and Engineering, Agency for Science, ... 3 Research Link, Singapore 117602, and Genome Institute of ...
4 downloads 0 Views 796KB Size
Download PDF
Anal. Chem. 2009, 81, 3344–3349

Surface Plasmon Resonance Study of Cooperative Interactions of Estrogen Receptor r and Transcriptional Factor Sp1 with Composite DNA Elements Siew Jun Neo,† Xiaodi Su,*,† and Jane S. Thomsen*,‡ Institute of Materials Research and Engineering, Agency for Science, Technology and Research (A*STAR), 3 Research Link, Singapore 117602, and Genome Institute of Singapore, Agency for Science, Technology and Research (A*STAR), 60 Biopolis Street, Singapore 138672 We have applied surface plasmon resonance (SPR) spectroscopy to study the cooperative interactions of estrogen receptor r (ERr) and transcription factor Sp1 with a composite DNA element, containing an estrogen response element (ERE) half-site upstream of two adjacent Sp1 sites (+571 ERE/Sp1 composite site in promoter A of the human PR gene). Using nuclear extracts of MCF-7 breast cancer cells as sample, we have shown that Sp1 is associated with Sp1-binding sites only, whereas ERr can be recruited to DNA both through direct binding to the ERE half-site and/or through protein-protein interactions with DNA-bound Sp1. The ERE half-site and the proximal Sp1 site are only 4 bp apart, and our data suggests that one transcription factor bound to DNA constitutes a sterical hindrance of the accessibility of the binding site for the other transcription factor. Our data confirms previous observations that ERr increases the amount of Sp1 recruited to the composite binding site in a dosedependent manner. Using recombinant proteins, we have unambiguously proved the formation of a ternary complex of ERr/Sp1-composite DNA, for which previously published electrophoretic mobility shift assay (EMSA) results are contradictive. With this study, we have demonstrated that the solid-liquid-phase SPR assay is a powerful alternative for studying multiprotein-DNA interactions and is superior to the EMSA experiments as it is capable of real-time measurements, can quantify the amount of protein bound, and can capture transient and weak binding interactions. The comprehensive characterization ofthesynergisticinteractionsbetweenERr-DNA,Sp1-DNA, and ERr-Sp1 contributes to the understanding of how ERr and Sp1 influence and activate gene transcription. Estrogen is a hormone of critical importance in the development and maintenance of reproductive tissues and in cardiovascular and bone physiology. Estrogen’s effects are mediated through its interaction with the intracellular estrogen receptors * To whom correspondence should be addressed. Phone: 65-68748420 (X.S.); 65-64788179 (J.S.T.). Fax: 65-68720785 (X.S.); 65-64789060 (J.S.T.). E-mail: [email protected] (X.S.); [email protected] (J.S.T.). † Institute of Materials Research and Engineering. ‡ Genome Institute of Singapore.

3344

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

(ERR and ERβ). Classical models of estrogen action have proposed that the binding of ERs to specific DNA sequences, estrogen response elements (EREs) (a palindromic repeat separated by a three-base spacer, 5′-GGTCAnnnTGACC-3′), is essential to initiate the cellular response to estrogen.1 Numerous studies have been conducted using electrophoretic mobility shift assay (EMSA),2-7 gel filtration chromatography assay,2,3 fluorescence anisotropy assay,8,9 and more recently surface plasmon resonance (SPR) spectroscopy,3,10-13 to document ERs’ DNA binding behaviors, including sequence specificity, ER-ERE binding stoichiometry, distinct binding behavior of different ER subtypes (R and β), and ligand effects. Parameters which are all important in determining how ERs regulate estrogen-responsive genes. In the context of complex regulatory regions, where imperfect EREs are encased among motifs for other transcription factors, estrogen regulation and gene transcription mechanisms are much more complicated, involving complex interactions between multiple proteins and DNA. Human progesterone receptor (PR),14-16 heat shock protein 27 (Hsp 27),17 cathepsin D,18 retinoic acid receptor,19 rat creatine kinase B (CKB),20 and c-fos protoonco(1) McDonnell, D. P.; Clemm, D. L.; Hermann, T.; Goldman, M. E.; Pike, J. W. Mol. Endocrinol. 1995, 9, 659–669. (2) Melamed, M.; Arnold, S. F.; Notidess, A. C.; Sasson, S. J. Steroid Biochem. Mol. Biol. 1996, 57, 153–159. (3) Cheskis, B. J.; Karathanasis, S.; Lyttle, C. R. J. Biol. Chem. 1997, 272, 11384–11391. (4) Cowley, S. M.; Hoare, S.; Mosselman, S.; Parker, M. G. J. Biol. Chem. 1997, 272, 19858–19862. (5) Hyder, S. M.; Chiappeta, C.; Stancel, G. M. Biochem. Pharmacol. 1999, 57, 597–601. (6) Loven, M. A.; Wood, J. R.; Nardulli, A. M. Mol. Cell. Endocrinol. 2001, 181, 151–163. (7) Yi, P.; Driscoll, M. D.; Huang, J.; Bhahat, S.; Hilf, R.; Bambara, R. A.; Muyan, M. Mol. Endocrinol. 2002, 16, 674–693. (8) Ozers, M. S.; Hill, J. J.; Erivin, K.; Wood, J. R.; Nardulli, A. M.; Royer, C. A.; Gorski, J. J. Biol. Chem. 1997, 272, 30405–30411. (9) Boyer, M.; Poujol, N.; Margeat, E.; Royer, C. A. Nucleic Acids Res. 2000, 28, 2494–2502. (10) Su, X. D.; Lin, C. Y.; O’Shea, S. J.; Teh, H. F.; Peh, Y. X.; Thomsen, J. S. Anal. Chem. 2006, 78, 5552–5558. (11) Su, X. D.; Neo, S. J.; Peh, Y. X. Anal. Biochem. 2008, 376, 137–143. (12) Teh, H. F.; Peh, Y. X.; Su, X. D.; Thomsen, J. S. Biochemistry 2007, 46, 2127–2135. (13) Peh, W. Y. X.; Reimhult, E.; Teh, H. F.; Thomsen, J.; Su, X. D. Biophys. J. 2007, 92, 4415–4427. (14) Shcultz, J. R.; Petz, L. N.; Nardulli, A. M. Mol. Cell. Endocrinol. 2003, 201, 165–175. 10.1021/ac802543x CCC: $40.75  2009 American Chemical Society Published on Web 03/30/2009

gene,21 for example, are all estrogen-inducible genes; however, the estrogen-responsive regions of these genes contain no classical palindromic EREs, instead they contain an ERE half-site (halfERE) adjacent to Sp1/GC-rich binding site(s). How ERs and Sp1 activate gene expression through these composite DNA elements and how ERs mediate estrogen responsiveness are issues requiring extensive studies of protein-DNA (i.e., ER-DNA and Sp1-DNA) and protein-protein (i.e., ER-Sp1) interactions. In this study, we have used SPR spectroscopy to determine the cooperative interactions of ERR and Sp1 with a half-ERE/Sp1 composite element in the promoter A (+571 to +595) of the human PR gene. The +571 half-ERE/Sp1 composite element of the PR-A gene is composed of an ERE half-site upstream of two Sp1 binding sites, and although this site has previously been thoroughly investigated using an EMSA approach,15,16 many of the complex regulatory mechanisms are still not fully understood, largely due to the limitations of the traditional EMSA technique.22 Using SPR measurement and with the assistance of specific antibodies, we have investigated the ability of physiological concentrations of ERR and Sp1 to bind to the composite DNA, we have determined the synergy effects between protein-DNA and protein-protein interactions, and have characterized the linear relationship between the amount of ERR recruitment to the composite site and the level of Sp1-DNA binding. We have also demonstrated unequivocally that ternary ERR/Sp1-DNA complexes are formed. These observations contribute to the understanding of PR-A gene regulation mechanisms and show how estrogen responsiveness can be mediated by ERR through a nonclassical pathway. Unlike the previous EMSA studies that used recombinant proteins and artificial protein mixtures to reveal the cooperative binding behaviors, our SPR study used nuclear extracts from the MCF-7 human breast cancer cell line, expressing endogenous ERR and Sp1. The use of biological material containing physiological levels of proteins (ERR and Sp1 in this case) give rise to results that are more closely correlated to the gene transcription processes in cells.23 The present study clearly demonstrates the strength of the SPR technology in the analysis of complex multiprotein-DNA interactions,24 and due to its closer resemblance to in vivo conditions, it could be the choice application for binding studies in the future. MATERIALS AND METHODS Materials. Purified recombinant human estrogen receptor (rhERR) and Sp1 (rhSp1) were purchased from PanVera Corp. (Madison, WI). Poly(dI-dC) and salmon sperm DNA, purchased from the Amersham, GE Health Sciences, were used to block (15) Petz, L. N.; Ziegler, Y. S.; Schultz, J. R.; Kim, H.; Kemper, J. K.; Nardulli, A. M. J. Steroid Biochem. Mol. Biol. 2004, 88, 113–122. (16) Petz, L. N.; Nardulli, A. M. Mol. Endocrinol. 2000, 14, 972–985. (17) Porter, W.; Saville, D.; Hoivik, D.; Safe, S. Mol. Endocrinol. 1997, 11, 1569– 1580. (18) Krishnan, V.; Wang, X.; Safe, S. J. Biol. Chem. 1994, 269, 15912–15917. (19) Rishi, A. K.; Shao, Z. M.; Baumann, R. G.; Li, X. S.; Sheikh, M. S.; Kimura, S.; Bashirelahi, N.; Fontana, J. A. Cancer Res. 1995, 55, 4999–5006. (20) Pentecost, B. T.; Mattheiss, L.; Dickerman, H. W.; Kumar, S. A. Mol. Endocrinol. 1990, 4, 1000–1010. (21) Duan, R.; Porter, W.; Safe, S. Endocrinology 1998, 139, 1981–1990. (22) Chorley, B. N.; Wang, X.; Campbell, M. R.; Pittman, G. S.; Noureddine, M. A.; Bell, D. A. Mutat. Res.: Rev. Mutat. Res. 2008, 659, 147–157. (23) Mao, G.; Brody, J. P. Biochem. Biophys. Res. Commun. 2007, 363, 153– 158. (24) Majka, J.; Speck, C. Adv. Biochem. Eng./Biotechnol. 2007, 104, 13–36.

nonspecific protein-DNA interactions. Rabbit anti-ERR (HC-20, directed against the C-terminus of human ERR), rabbit anti-Sp1 (H-225, directed against the N-terminus of human Sp1), and antirabbit IgG-HRP conjugated secondary antibody (SC-2004) were purchased from Santa Cruz Biotechnology (California). Streptavidin (SA) and phosphate-buffered saline (PBS), composed of 10 mM phosphate buffer, 137 mM NaCl, 2.7 mM KCl (pH 7.4), were purchased from Sigma-Aldrich Pte Ltd., Singapore. The oligonucleotide sequences and their descriptors are given below. Oligonucleotides were synthesized by Proligo Primer & Probes (Boulder, CO). The wild-type sequence, denoted as wt halfERE/Sp1, contains an ERE half-site (TGACC) upstream of two Sp1 binding sites. This sequence was mutated at either the Sp1 sites (mP/D), or the ERE half-site (mE), or both (scrambled sequence). The biotinylated strands and the antisense strands (sequences not shown) were annealed in a molar ratio of 1:10 in 10 mM PBS (pH 7.4), 137 mM NaCl, 10 mM MgCl2, 10 mM EDTA. • wt half-ERE/Sp1: 5′-biotin-TAGGAGCTGACCAGCGCCGCCCTCCCCCGCCCCCGACCA-3′ • mP/D half-ERE/Sp1: 5′-biotin-TAGGAGCTGACCAGCGTTGTACTCCCTTGTACCCGACCA-3′ • mE half-ERE/Sp1: 5′-biotin-TAGGAGCTGATTAGCGCCGCCCTCCCCCGCCCCCGACCA-3′ Scrambled sequence: 5′-biotin-TAGGAGCTGATTAGCGTTGTACTCCCTTGTACCCGACCA-3′ Cell Culture and Isolation of Nuclear Extract. MCF-7 human mammary carcinoma cells (ATCC, U.S.A.) were grown in Dulbecco’s modified Eagle’s medium (Invitrogen) supplemented with 5% heat-inactivated fetal bovine serum (FBS) (Invitrogen) at 37 °C in a humidified atmosphere of 5% CO2. For experimental conditions, stripped medium was used, i.e., cells were grown in phenol red-free medium (Invitrogen) supplemented with 5% dextran-coated charcoal-treated FBS (Invitrogen) for 72 h. Cells were then treated with 10 nM 17β-estradiol (E2) (Sigma) dissolved in dimethyl sulfoxide (ME2SO) for 45 min before isolation of nuclear extract. Nuclear extracts were prepared as described by Schreiber et al.25 Briefly, following a wash in ice-cold PBS, cells were resuspended in a hypotonic buffer containing 10 mM Hepes, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 1 mM DTT. After incubation on ice for 15 min, cells were lysed by addition of 0.25% Igepal CA630 and centrifuged at 10 000g for 5 min at 4 °C to collect the nuclei. The nuclear pellet was resuspended in a buffer containing 10 mM Hepes-KOH pH 7.9, 400 mM NaCl, 0.1 mM EDTA, 5% v/v glycerol, 1 mM DTT and incubated on ice for 15 min. Nuclear extract (supernatant) was recovered after centrifugation at 13 000g for 5 min at 4 °C, aliquoted, and stored at -80 °C. The protein concentration in the samples was investigated using Nanodrop. The presence of ERR and Sp1 in the nuclear extracts and their DNA binding activities were confirmed using Western blotting (25) Schreiber, E.; Matthias, P.; Muller, M. M.; Schaffner, W. Nucleic Acids Res. 1989, 17, 6419–6425.

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

3345

and transcription factor enzyme-linked immunosorbent assay (ELISA) kits (Panomics), respectively. In SPR experiments, the nuclear extract (4.7 mg/mL) in Hepes-KOH buffer containing 400 mM NaCl was diluted to 1.75 mg/mL using a NaCl-free Hepes buffer (the final buffer content was 10 mM Hepes-KOH, pH 7.9, containing 0.5 mM DTT, 0.05 mM EDTA, 1.2 mM MgCl2, 1.8% glycerol, and 150 mM NaCl). SPR Experiments. The SPR measurements were conducted using a two-channel, AutoLab ESPR (Eco Chemie, The Netherlands). The SPR disks form the base of a two-channel open cuvette, into which two DNA samples (same sequence or different sequence) can be immobilized for protein to bind. In a kinetic mode, SPR angle shift (∆θ) induced by protein and DNA adsorption was recorded over time. During the measurement, the liquid was mixed with an automatic aspiration/dispension pipet. The measurements were conducted at room temperature, and the noise level of SPR angle was 1 mdeg. According to the user manual of the AutoLab ESPR (version 2, 2002, Chapter 7, p 105) and the de Feijter formula,13,26 approximately 100 mdeg angle shift is equivalent to a change in surface concentrations of 83.3 ng/ cm2 of proteins (corresponding to dn/dc ) 0.18) and 78.9 ng/ cm2 of DNA (corresponding to dn/dc ) 0.19). Biotinylated DNA (200 nM) was immobilized on streptavidinmodified SPR disks using biotin-streptavidin-biotin bridge chemistry.10-13 The resulting DNA surface coverage was 5.7 ± 0.9 pmol/cm2 calculated from the angle shift of DNA immobilization (177 ± 12.5 mdeg) and the molecular weight of the 39-mer biotinylated DNA (24.428 kDa), using the conversion factor of 78.9 ng/(cm2 · 100 mdeg). Nuclear proteins were first incubated with nonspecific DNA, i.e., poly(dI-dC) and salmon sperm DNA (0.2 µg of poly(dI-dC) and 0.1 µg of salmon sperm for every 2 µg of protein) for 15 min at room temperature prior to application to DNA-immobilized surfaces. The binding of nuclear proteins to DNA was monitored for 15 min. To identify and detect nuclear ERR or nuclear Sp1 bindings, rabbit antiERR (10 µg/mL) or rabbit anti-Sp1 (50 µg/mL) was added, respectively, for 10 min, followed by addition of anti-rabbit IgG-HRP conjugate (10 µg/mL) for 10 min. To study recombinant protein-DNA binding, recombinant human ERR (rhERR), recombinant human Sp1 (rhSp1), or a mixture of these two proteins dissolved in protein binding buffer (40 mM Hepes, pH 7.4, containing 10 mM MgCl2, 0.2% Triton X-100, and 200 mM KCl) was applied, and the binding was monitored for 20 min at room temperature. Anti-ERR (10 µg/mL) and anti-Sp1 (50 µg/mL) were used to identify bound ERR and Sp1 from protein mixtures. To regenerate the immobilized DNA, 0.1% SDS (sodium dodecyl sulfate) was applied for 2-3 min after each cycle of protein binding. RESULTS AND DISCUSSION MCF-7 ERr and Sp1 Bind to Half-ERE/Sp1 Composite Sites. In the context of complex regulatory regions containing half-ERE and Sp1 sites, the interplay between the ER and Sp1 transcription factors are not fully understood. Using EMSA, Petz and co-workers15,16 have previously determined the binding of human recombinant ERR to the wild-type +571 half-ERE/Sp1 (26) de Feijter, J. A.; Benjamins, J.; Veer, F. A. Biopolymers 1978, 17, 1759– 1772.

3346

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

Figure 1. Detection of MCF-7 ERR-DNA binding. (A) Schematics of ERR binding to half-ERE/Sp1 composite sites with different mutations (wt, wild type; mE, ERE half-site is mutated; mP/D, Sp1 sites are mutated; scrambled DNA, both ERE half-site and Sp1 sites are mutated). (B) SPR curves depicting MCF-7 nuclear protein binding to the composite DNA, followed by identification and amplified detection of ERR using anti-ERR and a secondary antibody, respectively. The anti-ERR binding results in no detectable SPR angle increase (due to the low amount of bound Sp1) but a bulk refractive index change (caused by different buffer content in the anti-ERR solution) (ref 11). Bulk refractive changes are also observed upon addition of the secondary antibody. The angle shifts after rinsing are indicative of antibody binding.

composite site and the formation of an ERR-DNA complex in the absence or presence of Sp1. In this SPR study, we have used E2-treated MCF-7 nuclear extract as source to assess the amount of endogenous ERR and Sp1 proteins binding to either the wildtype DNA composite binding site (wt), the mutant half-ERE/Sp1 composite sites (mE and mP/D), or a scrambled sequence (with both the ERE half-site and the Sp1 sites mutated). Following the binding of nuclear proteins, anti-ERR antibody (or anti-Sp1) was added to identify bound ERR (or Sp1) among other protein components, and a secondary antibody was used successively to amplify the signal, ensuring detection of the low endogenous level of bound ERR (or Sp1). We have previously developed this “twostep antibody” SPR assay11 for detection of specific ERR-ERE interactions using endogenous levels of MCF-7 ERR nuclear protein. The angle shift induced by addition of secondary antibody (∆θAb) clearly showed DNA sequence-dependent differences in the amount of recruited ERR, allowing us to easily distinguish between a wt palindromic ERE sequence, a mutant ERE, and scrambled DNA. Results in Figure 1, specifically the angle shift caused by the secondary antibody (∆θsecondary Ab) following anti-ERR, show that ERR is present in substantial amounts in our estrogen-treated

MCF-7 nuclear extracts, and it binds effectively to both the wt DNA composite sequence (∆θsecondary Ab ) 37 ± 5 mdeg) and the sequence containing an intact ERE half-site next to mutated Sp1 sites (mP/D) (∆θsecondary Ab ) 105 ± 10 mdeg). Interestingly, ERR is also recruited to the DNA sequence with intact Sp1 sites but with a mutated ERE half-site (mE) (∆θsecondary Ab ) 72 ± 10 mdeg), indicating that ERR here is recruited by proteinprotein interactions rather than protein-DNA interactions. When we carried out the next permutation experiment, where the Sp1 sites were also mutated (in addition to the mutation of the half-ERE), no ERR binding was detected to this scrambled sequence, supporting the view that the recruitment of ERR to a non-ERE-containing DNA sequence is due to specific protein-protein (ERR-Sp1) interactions rather than unspecific DNA binding. The fact that ERR can bind to a non-ERE site through indirect tethering mechanisms has previously been reported for Sp1,14-17,27 AP1,28 and NFkB29 factors (i.e., ERR can regulate gene expression through DNA binding sites for these transcription factors). We have previously shown that approximately 4% of all the ERR binding sites identified in MCF-7 cells in a genome-wide manner are tethered sites (i.e., non-ERE sites).30 Currently the tethered sites are not well characterized, and in the few cases where they have been studied using, e.g., EMSA, it has been customary to use recombinant protein and artificially made ERR/Sp1 mixtures. The usage of MCF-7 nuclear extracts in this study provides a more realistic characterization because the biological samples express nuclear ERR and Sp1 protein at physiological levels. To determine the involvement of Sp1 in PR gene expression, we next carried out a series of experiments replacing anti-ERR antibody with anti-Sp1 antibody in our “two-step antibody” SPR approach to detect nuclear Sp1 binding to the half-ERE/Sp1 composite site and determine the contributions of the ERE halfsite and the Sp1 sites to Sp1-DNA complex formation. We could easily detect nuclear Sp1 binding to the mE sequence that contains Sp1 binding sites (∆θsecondary Ab ) 75 ± 10 mdeg), but not to the mP/D sequence, in which the Sp1 sites were mutated (Figure 2). That MCF-7 Sp1 was not recruited to the Sp1-mutated sequence, even in the presence of ERR (bound to half-ERE) is in concordance with the literature using recombinant proteins in EMSA experiments,16 showing that Sp1 cannot be recruited to DNA by protein-protein interactions, in clear contrast to the ERR scenario (shown in Figure 1). In summary, we confirm that ERR makes use of tethered sites, whereas Sp1 does not use this mechanism of action. Synergistic Function between ERE Half-Site and Sp1 Sites in Recruiting Proteins. Currently, close to a dozen half-ERE/ Sp1 composite sites have been characterized. At these sites the presence of Sp1 sites next to a half-ERE (typically 10-17 bp away) has been reported to increase the amount of ERR protein recruited (27) Li, C.; Briggs, M. R.; Ahlborn, T. E.; Kraemer, F. B.; Liu, J. Endocrinology 2001, 142, 1546–1552. (28) Jakacka, M.; Ito, M.; Weiss, J.; Chien, P. Y.; Gehm, B. D.; Jameson, J. L. J. Biol. Chem. 2001, 276, 13615–13621. (29) Shyamala, G.; Guiot, G. C. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 10628– 10632. (30) Lin, C. Y.; Vega, V. B.; Thomsen, J. S.; Zhang, T.; Kong, A. L.; Xie, M.; Chiu, K. P.; Lipovich, L.; Barnett, D. H.; Stossi, F.; Teo, A.; George, J.; Kuznetsov, A. A.; Lee, Y. K.; Charn, T. H.; Palanisamy, N.; Miller, L. D.; Cheung, E.; Katzenellenbogen, B. S.; Ruan, Y.; Bourque, G.; Wei, C. L.; Liu, E. T. PLoS Genet. 2007, 3, e87.

Figure 2. Detection of MCF-7 Sp1-DNA binding. (A) Schematics of Sp1 binding to mE and mP/D half-ERE/Sp1 composite sites. (B) The overlay of SPR curves depicting MCF-7 nuclear protein binding to the composite DNA, followed by identification and amplified detection of Sp1 using anti-Sp1 and a secondary antibody, respectively. The anti-Sp1 binding does not result in any detectable SPR angle increase (due to the low amount of bound Sp1) but a bulk refractive index change (caused by different buffer content in the antiSp1 solution) (ref 11). Bulk refractive changes are also observed upon addition of the secondary antibody. The angle shifts after rinsing are indicative of antibody binding.

to the sites compared to sites containing an ERE half-site only. In contrast, the current +571 half-ERE/Sp1 composite element of the PR-A gene, in which the ERE half-site and the most proximal of the two Sp1 sites are separated by only four nucleotides, has been reported to behave differently (using EMSA and recombinant proteins15,16). Not only do we confirm this observation, we have also obtained additional information allowing us to develop a more complex model of ERR/Sp1-DNA interaction as follows: From the ERR experiments (Figure 1), we observed that the amount of ERR bound to mP/D (DNA containing only an ERE half-site) is considerably higher (∆θsecondary Ab ) 105 ± 10 mdeg) than the amount of ERR (∆θsecondary Ab ) 37 ± 5 mdeg) bound to an ERE half-site with the presence of adjacent Sp1 sites (wt sequence). Similarly, when we investigated the binding of Sp1 protein, we found that Sp1 binding to the wt DNA was barely detectable (curve not shown), whereas the recruitment of Sp1 to the mE sequence (with the ERE half-site mutated, Figure 2) was readily detectable (∆θsecondary Ab ) 75 ± 10 mdeg). Since the distance between the ERE half-site and the most proximal of the two Sp1 sites in the PR-A composite site is only four nucleotides, we suggest a model of sterical hindrance to explain our results. Specifically, the binding of ERR to its ERE halfsite blocks the accessibility of the adjacent Sp1 site, resulting in lower recruitment of Sp1 protein to its binding element, and vice versa, binding of Sp1 to its respective binding sites constitutes a sterical hindrance for ERR resulting in reduced ERR-DNA interaction. That one transcription factor bound to DNA can block the accessibility of binding sites for other Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

3347

Figure 3. ERR enhances MCF-7 Sp1-DNA complex formation. SPR curves depicting the addition of anti-Sp1 and subsequently the secondary antibody to nuclear protein bound to wt half-ERE/Sp1 DNA (nuclear protein binding was not shown), in the presence (solid blue line) and absence (dashed pink line) of rhERR. The amount of rhERR added into the nuclear protein was 5 (left), 10 (middle), and 20 nM (right). Nuclear protein concentration was 1.75 mg/mL.

transcription factors has previously been proposed for other composite transcription factor binding motifs.31 Synergistic Function of the Two Transcription Factors. Sp1 protein is known to be involved in regulation of the PR gene, and ERR is known to increase the transcriptional rate of the PR gene in an estradiol-inducible manner. In addition to the observation that a mutated ERE half-site results in increased Sp1-DNA binding, the following crosstalk between the two transcription factors has been observed in this current study: When MCF-7 nuclear extract was incubated with the wt half-ERE/Sp1 composite site immobilized on SPR, a spike-in of human recombinant ERR protein (5, 10, and 20 nM) resulted in an increased proportion of Sp1 protein recruited to DNA (Figure 3), whereas, in the absence of human recombinant ERR, the nuclear Sp1 binding to the same DNA was barely detectable (the average ∆θsecondary Ab from three repeated measurements, dashed pink lines in the panels in Figure 3, was only 3.3 ± 0.3 mdeg). Using EMSA analysis of recombinant protein-DNA interactions, Petz et al.15 have made a similar observation, i.e., a faint, barely visible band representing the Sp1-DNA complex was enhanced gradually in the presence of increasing amount of ERR. Interestingly, the SPR data in this study further reveals that the magnitude of increased Sp1 binding (∆θsecondary Ab 20 ± 2.2, 46 ± 4.2, and 75 ± 5.3 mdeg) is linearly related to the amount of ERR (5, 10, and 20 nM) added (r2 ) 0.98). The quantification of the dose-dependent enhancement is an additional characterization complementary to the EMSA analysis. In order to investigate if it was possible to reciprocate the effect, i.e., would addition of Sp1 enhance ERR recruitment to the PR-A site, human recombinant Sp1 protein in concentrations of either 5 or 10 nM was added into nuclear extracts followed by measurement of ERR binding to the wt half-ERE/Sp1 site. No enhancement of ERR binding was observed; instead, the binding was reduced in an Sp1 concentration-dependent manner in concordance with previously published data obtained using EMSA.15,16 Thus, it appears that Sp1 reduces ERR recruitment due to increased (31) Safe, S.; Wormke, M.; Samudio, I. J. Mammary Gland Biol. Neoplasia 2000, 5, 295–306.

3348

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

sterical hindrance, whereas the stabilizing effect of ERR protein on the Sp1 complex more than compensates for the sterical hindrance taking place at the same time. The data indicate that the ratio of ERR/Sp1 protein is crucial for the amount of ERR/ Sp1 molecules being recruited to the site, which ultimately could affect the transcriptional rate of the PR gene. Formation of Ternary ERr/Sp1-DNA Complex. Although there is a large body of evidence from previous EMSA experiments14-18 and the current SPR assays that ERR and Sp1 have a synergy function in regulating gene expression and ERR enhances Sp1-DNA complex formation through protein-protein interactions, the question as for whether a ternary ERR/Sp1-DNA complex is formed has not been conclusive. In most of the EMSA experiments, no higher order of complexes were observed and could be assigned to a ternary ERR/Sp1-DNA complex,14,16-20 except for one example.15 The ambiguity may arise from the fact that EMSA experiments require the formation of stable complexes, which must be maintained during extended periods of electrophoresis. Taking the advantages of real-time binding studies for strong and weak complexes of the SPR assays, we proved unambiguously the presence of a ternary complex of ERR/ Sp1-DNA with the +571 half-ERE/Sp1 composite site. Figure 4 shows the real-time binding of a mixture of rhSp1 (500 nM) and rhERR (20 nM) to mE oligo immobilized surface, followed by an injection of anti-Sp1, and subsequently anti-ERR to identify Sp1 and ERR, respectively. The well-detectable anti-Sp1 (40 mdeg) and anti-ERR binding signals (29 mdeg), following the Sp1/ERR protein binding (76 mdeg), confirm the presence of bound Sp1 and ERR on the mE oligo. When 20 nM of ERR was added alone (in the control channel), no direct binding of ERR to the mE DNA was detectable. This further confirms that the bound ERR in the presence of Sp1 is through its interaction with Sp1 bound to the Sp1 sites. This result not only indicates the presence of a ternary ERR/Sp1-DNA complex but also proves the formation of ERR/ Sp1-DNA ternary complex in a tethered mechanism.14,15 Not only can we prove the formation of an ERR/Sp1-DNA ternary complex with SPR measurements, we can also quantify each protein component with the assistance of antibody signals.

of 0.65 and 0.25 pmol/cm2, respectively. In view of the much lower concentration of rhERR (20 nM) compared to rhSp1 (500 nM) in the protein mixture, the amount of ERR bound through the protein-protein interaction is quite significant. With this calculation we demonstrate that quantitative identification of protein components in multiprotein-DNA complexes is possible. With further experimentation it would be possible to determine the binding stoichiometry of protein-DNA and protein-protein interactions in the tethered complex.

Figure 4. Detection of formation of ERR/Sp1-DNA ternary complex through a tethered mechanism. SPR curve depicting the binding of a mixture of rhSp1 (500 nM) and rhERR (20 nM) (solid blue line) to mE DNA, followed by addition of 50 µg/mL anti-Sp1 and 10 µg/mL antiERR, sequentially. The dashed pink line is the SPR curve showing the addition of rhERR (20 nM) alone to the same mE DNA.

Our experiments with recombinant ERR, Sp1, and their corresponding antibodies indicate that anti-ERR and anti-Sp1 signals are linearly related to the amount of their specific proteins (see the Supporting Information). On the basis of the calibration curves depicting the relationship between SPR signal of anti-Sp1 versus bound Sp1 (∆θanti-Sp1 ) 0.6289∆θSp1, r2 ) 0.956) and anti-ERR versus bound ERR (∆θanti-ERR ) 1.4656∆θERR, r2 ) 0.957), the 40 mdeg of anti-Sp1 measured following the formation of a ternary complex correlates to a Sp1 amount of 63.6 mdeg and the 29 mdeg of anti-ERR to an ERR amount of 19.8 mdeg. The sum of ERR and Sp1 signals calculated from the antibody signals is 83 mdeg, being approximate to the measured total protein binding signals of 76 mdeg. Using the conversion factor of 83.3 ng/(cm2 · 100 mdeg) for protein and the molecular weight of Sp1 (81 kDa) and ERR (66 kDa), the bound Sp1 (63.6 mdeg) and ERR (19.8 mdeg) are equivalent to protein amounts

CONCLUSION We have extended the usage of SPR spectroscopy for studying protein binding to DNA containing multiple binding sites, using the +571 half-ERE/Sp1 composite site of the PR-A gene as a model. We have determined the ability of nuclear ERR and Sp1 to bind to the composite site and the synergy function of these two proteins and thereafter extended the understanding of how these two proteins are involved in PR gene expression. When compared to the gel mobility shift assay that detects protein-DNA complexes formed in homogeneous phase, the solid-liquid-phase SPR assay allows for real-time measurement of ternary ERR/ Sp1-DNA complex formation and provides quantification of the synergy effects between proteins and composite DNA. With this study, we have demonstrated the power of the SPR techniques for studying complex protein-DNA bindings. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review December 2, 2008. Accepted February 24, 2009. AC802543X

Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

3349