Article pubs.acs.org/ac
Endonucleolytic Inhibition Assay of DNA/Fok I Transducer as a Sensitive Platform for Homogeneous Fluorescence Detection of Small Molecule−Protein Interactions Zhen Zhen,† Li-Juan Tang,† Jian Lin,† Jian-Hui Jiang,*,† Ru-Qin Yu,† Xiangling Xiong,‡ and Weihong Tan*,‡ †
State Key Laboratory of Chemo/Bio-Sensing and Chemometrics, College of Chemistry and Chemical Engineering, College of Biology, Hunan University, Changsha, 410082, P. R. China ‡ Center for Research at Bio/Nano Interface, Department of Chemistry and Department of Physiology and Functional Genomics, Shands Cancer Center, UF Genetics Institute and McKnight Brain Institute, University of Florida, Gainesville, Florida 32611-7200, United States S Supporting Information *
ABSTRACT: This paper reported a novel homogeneous fluorescence assay strategy for probing small molecule−protein interactions based on endonucleolytic inhibition of a DNA/ Fok I transducer. The transducer could cyclically cleave fluorescence-quenched probes to yield activated fluorescence signal, while protein binding to the small molecule label would prevent Fok I from approaching and cleaving the fluorescencequenched probes. Because of the efficient signal amplification from the cyclic cleavage operation, the developed strategy could offer high sensitivity for detecting small molecule− protein interactions. This strategy was demonstrated using folate and its high-affinity or low-affinity binding proteins. The results revealed that the developed strategy was highly sensitive for detecting either high- or low-affinity small molecule−protein interactions with improved selectivity against nonspecific protein adsorption. This strategy could also be extended for assays of candidate small-molecule ligands using a competitive assay format. Moreover, this strategy only required labeling the small molecule on a DNA heteroduplex, circumventing protein modifications that might be harmful for activity. In view of these advantages, this new method could have potential to become a universal, sensitive, and selective platform for quantitative assays of small molecule−protein interactions.
I
unique biochemical properties of DNA, which can be probed with DNA-modifying enzymes or DNA sensors,13−15 can facilitate signal transduction of binding events on the DNA moiety. Herein, we report the proof-of-principle of a novel fluorescence assay strategy using small molecule-labeled DNA for sensitive detection of small molecule−protein interactions, even those with low affinity. Because organic compounds generally feature much smaller molecular sizes than their protein targets, the binding of protein to a small-molecule moiety in the DNA module can dramatically increase steric hindrance at the small-molecule labeling site. Motivated by this hypothesis, we utilized small molecule-labeled DNA to construct an endonuclease Fok I-based DNA/Fok I transducer whose DNA-cleaving activity is inhibited by protein binding. In the absence of target protein, Fok I cyclically cleaves the fluorescence-quenched DNA probe, thereby activating a fluorescence signal. In contrast, binding of protein to the
nteractions between proteins and small molecules constitute a critical regulatory mechanism in many fundamental biological processes, and detection of these interactions represents a major avenue to biomarker analysis for clinical and public safety monitoring.1−4 Assay of small molecule− protein interactions commonly relies on immobilization or labeling of the organic ligands or the protein targets. Although technically endowed with the capacity of multiplexed assay of multiple ligands or targets, the surface-based methods5 may interfere with protein−ligand binding and frequently require cautious washing and blocking to combat nonspecific adsorption. The homogeneous assay such as fluorescence polarization,6 resonant energy transfer,7 and protein-fragment complementation,8 which can be implemented without any surface operations, then offers a preferable and robust technology for identification and validation of small molecule−protein interactions. However, these reported methods still show limited sensitivity, especially in the analysis of lowaffinity small molecule−protein interactions. Small-molecule labeled DNA represents an attractive option for investigating small molecule−protein interactions.9−12 The © 2012 American Chemical Society
Received: April 2, 2012 Accepted: June 6, 2012 Published: June 6, 2012 5708
dx.doi.org/10.1021/ac300889t | Anal. Chem. 2012, 84, 5708−5715
Analytical Chemistry
Article
monitored at 37 °C for 120 min by time-dependent fluorescence. DNA/Fok I Transducer-Based Detection of Small Molecule−Protein Interactions. In a typical assay, a 50 μL aliquot of the DNA heteroduplex solution (100 nM) was added to 40 μL of reaction buffer containing 50 mM KAc, 20 mM Tris-HAc (pH 7.9), 10 mM Mg(Ac)2, 2 mM DTT, 2 μM probe 3, and target protein, such as FR, DHFR, streptavidin, biotin antibody, or BSA, of a given concentration. The mixture was incubated at 37 °C for 20 min to allow complete interaction between the protein and the DNA heteroduplex. Then, 40 units of Fok I (10 μL) were added to the mixture to allow the formation of the DNA/Fok I transducer. DNA-cleaving behavior was monitored at 37 °C for 120 min by timedependent fluorescence. Time-dependent fluorescence responses of the DNA/Fok I transducer were performed with a time interval of 30 s in a 384well black microplate on a Tecan Infinite M-1000 microplate reader. The excitation wavelength was 490 nm, and the emission wavelength was 520 nm with both excitation and emission bandwidths of 10 nm. The reaction was terminated at 65 °C for 20 min with the addition of 20 mM EDTA. The resulting solution was immediately subjected to fluorescence spectral measurements. The fluorescence spectra were measured at room temperature in a 100 μL quartz cuvette on a Fluorolog-Tau-3 spectrofluorometer. The excitation wavelength was 490 nm, and the emission wavelength was in the range of 510 to 700 nm with both excitation and emission bandwidths at 2.5 nm.
small molecule-labeled DNA prevents Fok I from freely accessing the cleavage site, thus inhibiting the activity of Fok I and preventing the fluorescence enhancement. This allows the development of a universal strategy for detecting small molecule−protein interactions. Because cyclical cleavage of the input probe affords efficient amplification of the fluorescence signal, we are able to achieve very high sensitivity in detecting various small molecule−protein interactions, even those with low affinity. This strategy may create a new paradigm for efficient detection of small molecule biomarkers or their binding proteins in clinical diagnostics and public safety analysis.
■
EXPERIMENTAL SECTION Reagents and Materials. Folate receptor (FR), recombinant human dihydrofolate reductase (DHFR), folic acid or folate, 1-ethyl-3-(3-dimethyl-aminopropyl) carbodiimide hydrochloride (EDC), N-hydroxysulfosuccinimide (Sulfo-NHS), fluorescein, and monoclonal antibody (Ab) to biotin were purchased from Sigma. Streptavidin and Fok I (4000 U/mL) were obtained from New England Biolabs. Bovine serum albumin (BSA), human serum albumin (HSA), immunoglobulin G (IgG), transferrin (TF), α-fetoprotein (AFP), carcinoembryonic antigen (CEA), and carbohydrate antigen (CA) 19-9 were provided by the National Institute for the Control of Pharmaceutical and Biological Products. The oligonucleotides used in this work were synthesized by Takara Biotechnology Co. Ltd. The sequences of the synthesized oligonucleotides are given in Table S1, Supporting Information. Labeling and Purification of Folate to NH2-Modified Oligonucleotides. The folate label was conjugated to the NH2-modifier of oligonucleotide sequence 5 using the succinimide coupling (EDC-NHS) method.16 Briefly, 0.5 mL of 20 μM sequence 5 was mixed with 0.5 mL of 10 mM phosphate buffer (PB, pH 7.4), containing 10 mM folate, 1 mM EDC, and 5 mM Sulfo-NHS, and incubated for 2 h at 37 °C in the dark. The solution was then dialyzed against PB using a membrane with molecular weight cutoff of 1000 Da to remove excess folate. The dialysis was performed for 3 days with shielding from light and changes with fresh buffer every 4 h. DNA products were purified by semipreparative high-performance liquid chromatography (HPLC) on an Agilent 1200 Series Rapid Resolution HPLC analytical scale purification system (more details in Supporting Information). Construction of DNA/Fok I Transducer. The DNA/Fok I transducer was composed of a DNA heteroduplex to allow the assembly of the restriction endonuclease Fok I. The intermolecular DNA heteroduplexes were prepared by the annealing of two sequences, 1 and 2 (1 μM for each sequence), via heating for 5 min at 95 °C, followed by cooling at 50 °C for 1 h. The hybridization buffer contained 50 mM KAc, 20 mM Tris-HAc (pH 7.9), and 10 mM Mg(Ac)2. All the resulting DNA heteroduplexes, when stored at −18 °C, were found to be stable for months with no detectable variations for downstream assay. The DNA/Fok I transducer was prepared freshly before each assay. Typically, a 50 μL aliquot of a DNA heteroduplex solution (100 nM) was added to 40 μL of reaction buffer containing 50 mM KAc, 20 mM Tris-HAc (pH 7.9), 10 mM Mg(Ac)2, 2 mM DTT, and 2 μM probe 3. Then, 40 units of Fok I (10 μL) were added to the mixture to allow the formation of the DNA/Fok I transducer. DNA-cleaving behavior was
■
RESULTS AND DISCUSSION Design of DNA/Fok I Transducer for Small Molecule− Protein Interaction Assay. The DNA/Fok I transducer is constructed using a restriction endonuclease Fok I and a small molecule-labeled DNA heteroduplex (Figure 1A). To understand this design, it is important to recall the properties of Fok I. As one of the best-characterized type II−S endonucleases,17 Fok I recognizes the double-stranded 5′-GGATG-3′ site and precisely cleaves at the 9th and 13th nucleotides downstream from the recognition site. In our design of the DNA/Fok I transducer, the DNA heteroduplex is prepared via hybridization of DNA sequences 1 and 2 (Table S1 in Supporting Information) to incorporate a small-molecule label and two recognition sites for Fok I. Two Fok I molecules can assemble on the two recognition sites to form a dimer, which is known to be essential for its high endonucleolytic activity (Figure S1 in Supporting Information).18 The introduction of a smallmolecule label in the DNA heteroduplex is unique to the DNA/Fok I transducer, and it is a key component that can interact with the target protein and, in turn, inhibit the endonucleolytic activity of Fok I. The labeling site for the small molecule, the nucleotide three-bases downstream from a recognition site of Fok I, was determined via thorough optimization (Figure S2 in Supporting Information). The endonuclease activity of the DNA/Fok I transducer can be probed through fluorescence activation using an engineered DNA probe 3 (Figure 1B) labeled with a 5-carboxy fluorescein (FAM) fluorophore and a tetramethyl rhodamine (TMR) quencher in adjacent nucleotides. Thus, the FAM fluorescence is quenched in the intact probe. The probe’s sequence is complementary to the single-stranded tail downstream of the Fok I recognition site in the DNA heteroduplex. As a result, when the probe is added as the input, it hybridizes to the single5709
dx.doi.org/10.1021/ac300889t | Anal. Chem. 2012, 84, 5708−5715
Analytical Chemistry
Article
Figure 1. Design and working principle of DNA/Fok I transducer for small molecule−protein interaction assay. (A) Fok I assembled and dimerized on a small molecule-labeled DNA heteroduplex to form the DNA/Fok I transducer. The heteroduplex is shown as a hybrid of the two DNA sequences used in the experiments; the small molecule label site is denoted by a brown dot; and Fok I is symbolized by a blue recognition domain linked with a catalytic domain. (B) The DNA/Fok I transducer cyclically cleaving fluorescence-quenched probes and activating an amplified fluorescence signal. The DNA/Fok I transducer is schematically simplified into a cartoon complex of a DNA duplex and a protein domain. (C) The DNA/Fok I transducer with its smallmolecule label bound by target protein displaying inhibited activity in cleaving fluorescence-quenched probes.
stranded tail of the heteroduplex. Because Fok I is able to cleave DNA downstream of its recognition site, probe 3 will be split by Fok I into two short fragments that can be readily released from the single-stranded tail. This separates the fluorophore from the quencher, activates the fluorescence signal, and reexposes the single-stranded tail to the remaining intact input probes. In this way, the DNA/Fok I transducer can perform a cyclical DNA-cleaving operation via the repetition of annealing, cleaving, and releasing of the input probe. In contrast, when protein target is present, its binding to the small-molecule label will prevent Fok I from scissoring the input probe (Figure 1C), presumably because steric hindrance from the bound protein precludes Fok I from freely accessing the cleavage site. This prevents the DNA-cleaving operation, thus inhibiting fluorescence activation. As a result, endonucleolytic activity of the DNA/Fok I transducer, as indicated by the activated fluorescence signal, is dynamically inhibited by the small molecule−protein binding event. We have used this DNA/Fok I transducer as a new homogeneous fluorescence assay technology, termed endonucleolytic inhibition assay, which has several advantages for sensitive detection of small molecule−protein interactions. First, our design that the DNA/Fok I transducer can operate to
Figure 2. Time-dependent fluorescence responses of the DNA/Fok I transducer to small molecule−protein interaction. (A) Reactions of probe 3 (1 μM) plus 50 nM folate-labeled DNA heteroduplex of sequences 1 and 2 (brown), probe 3 (1 μM) plus 4 U/μL Fok I (cyan), and probe 3 (1 μM) plus 50 nM folate-labeled DNA/Fok I transducer (red). (B) Reactions of probe 3 (1 μM) plus 50 nM folatelabeled DNA/Fok I transducer in the presence of 50 nM FR (dark green), probe 3 (1 μM) plus 50 nM amino-labeled DNA/Fok I transducer (blue), and probe 3 (1 μM) plus amino-labeled DNA/Fok I transducer in the presence of 50 nM FR (dark yellow). (C) Reactions of probe 3 (1 μM) plus 50 nM folate-labeled DNA/Fok I transducer (pink) or amino-labeled DNA/Fok I transducer (black) in the presence of 5 μM DHFR.
cleave the input probe cyclically make the developed technology well suited for detecting low-affinity interactions. In our design, the DNA-cleaving operation is stopped when the protein target is bound to the small molecule label, and the DNA-cleaving operation reoccurs reversibly when the protein target is dissociated from the small molecule label. In the assay of low-affinity interactions with rapid dissociation and 5710
dx.doi.org/10.1021/ac300889t | Anal. Chem. 2012, 84, 5708−5715
Analytical Chemistry
Article
Figure 3. Fluorescence responses of folate-labeled DNA/Fok I transducer to folate−FA interaction. (A) Typical time-dependent fluorescence responses of 50 nM folate-labeled DNA/Fok I transducer to FR of varying concentrations. (B) Corresponding end-point fluorescence readouts versus FR concentrations. Error bars are standard deviations across four repetitive assays.
Figure 4. Fluorescence responses of folate-labeled Fok I DNA/Fok I transducer to folate−DHFR interaction. (A) Typical time-dependent fluorescence responses of 50 nM folate-labeled DNA/Fok I transducer to DHFR of varying concentrations. (B) Corresponding end-point fluorescence readouts versus DHFR concentrations. Error bars are standard deviations across four repetitive assays.
reassociation of protein targets from small-molecule labels, the DNA/Fok I transducer can still be inhibited during the association of target protein with the small molecule label, while continuing to cleave the input probe while the target protein is dissociated. On average, this makes the DNA/Fok I transducer show a decreased rate in cleaving the input probe over the entire reaction duration, thereby revealing the interaction of the small molecule label with its target protein. Second, the cyclical cleavage of the fluorescence-quenched input probe creates an efficient signal amplification mechanism for highly sensitive detection of the inhibition of Fok I. This provides high sensitivity for detecting protein receptors of low-abundance, as well as interactions with low affinity. Third, our strategy can exhibit improved selectivity as compared with existing methods, since nonspecific interactions are presumably too weak to preclude the access of Fok I to its cleavage site. Fourth, the fluorescence activation measurements in homogeneous format make this strategy easily automated for quantitative analysis. It can, therefore, be extended for multiplexed and highthroughput assays of hundreds of samples using multicolor fluorescence-quenched probes and microplate formats such as 96- or 384-well plates with the aid of robotic delivery systems. Compared with the FRET-based assay of small molecule− protein interactions, which requires preliminary labeling of the small molecules and target protein, respectively, with a fluorescence donor and the corresponding fluorescence acceptor, our strategy offers improved sensitivity and circumvents
protein modifications that may be harmful for activity and difficult for controllable labeling. It only requires that the small molecule be labeled on a DNA sequence, which can be accomplished using routine bioconjugation protocols. In this sense, our novel strategy holds great promise as a sensitive, specific, and robust platform for quantitative assay of small molecule−protein interactions. Note that the fluorescenced-quenched DNA probe 3 uses FAM as the fluorophore, which shows a pH-dependent fluorescent property and thus may limit the applications of our method. In situations where the small molecule−protein interactions are investigated under varying pH conditions, we can use some pH-insensitive fluorophores, such as Alexa Fluor 488 and BODIPY FL dyes, to resolve this limitation. Another limitation is that this method relies on the decrease of fluorescence signal for detecting small molecule−protein interactions, implying it is necessary to perform an interaction-free control experiment for every small molecule ligand to identify the specific interactions. Additionally, the conjugation of small molecule to DNA requires timeconsuming reactions and separation. However, this can be solved in the DNA synthesis step using small molecule-labeled phosphoramidites. Response Characteristics of DNA/Fok I Transducer for Small Molecule−Protein Interaction Assay. To investigate the response characteristics of the DNA/Fok I transducer regulated with small molecule−protein binding, we chose two 5711
dx.doi.org/10.1021/ac300889t | Anal. Chem. 2012, 84, 5708−5715
Analytical Chemistry
Article
Figure 5. Working principle of DNA/Fok I transducer for competitive assay of small−molecule binders. The inactivated DNA/Fok I transducer (small-molecule label bound to its target protein) interacting with competitive small-molecule binder, releasing the protein target and restoring the active DNA/Fok I transducer that is able to cyclically cleave fluorescence-quenched probes.
small-molecule compounds, folate and biotin, for a case study. Folate and biotin are both known to have two binding proteins each: folate receptor (FR) and dihydrofolate reductase (DHFR) for folate; streptavidin (SA) and antibiotin antibody (Ab) for biotin. The dissociation constants for FR−folate and DHFR−folate are ∼0.1 nM19 and ∼2.8 μM,20 respectively, while the dissociation constants for SA−biotin and Ab−biotin are ∼1 fM21 and ∼50 nM,22 respectively. These binding affinities cover the range of most interactions of interest in chemistry and biology. We modified these two small-molecule compounds separately on the labeling site in the DNA heteroduplex. The response characteristics of the DNA/Fok I transducer were then investigated using fluorescence measurements. The DNA-cleaving behavior of the DNA/Fok I transducer was studied using time-dependent fluorescence analysis (Figure 2). In the mixture containing only 50 nM folate-labeled DNA heteroduplex and 1 μM probe 3, a very weak fluorescence signal (∼210) was observed after an incubation of 2 h, and the signal did not show appreciable time-dependent changes (Figure 2A). A similar finding was obtained for the mixture containing only Fok I (4 U/μL) and probe 3. After adding Fok I (4 U/μL) to the mixture with 50 nM DNA heteroduplex and 1 μM probe 3, the fluorescence signal increased rapidly, yielding a saturation fluorescence readout of ∼3420 after 2 h. Such a desirable signal-to-background ratio (>16-fold) suggested that probe 3 was effectively cleaved by the DNA/ Fok I transducer. Because no substantial cleavage of probe 3 was observed when either Fok I or the DNA heteroduplex was absent in the analysis, it was clear that Fok I and DNA heteroduplex were both essential for the endonucleolytic activity. Furthermore, we found that the DNA/Fok I transducer was able to cleave probe 3 cyclically (Figure S3 in Supporting
Information), giving a clear route of signal amplification in determining the Fok I activity. It should be noted that probe 3 only had six bases complementary to the single-stranded tail in the DNA/Fok I transducer. The melting temperature was estimated to be ∼26 °C23 which was lower than the reaction temperature of 37 °C. However, efficient cleavage of probe 3 was obtained, presumably because endonucleolytic activity of the DNA/Fok I transducer required only the formation of a transient heteroduplex hybrid of the probe on the single-stranded tail. In the presence of 50 nM FR, the DNA/Fok I transducer showed a much slower rate of fluorescence activation (Figure 2B). After an incubation of 2 h, the end-point fluorescence readout showed only a ∼2-fold enhancement, compared with that for the noncleaved probe 3. This gave immediate evidence that binding of FR to the folate label could prevent the DNA/ Fok I transducer from cleaving probe 3. (Further evidence of the inhibition of the DNA/Fok I transducer was given in capillary electrophoresis analysis, as shown in Figure S4 in Supporting Information). As a control, we replaced the folate label by an amino modifier in the DNA/Fok I transducer. It was observed that, in the absence of FR, this amino-labeled DNA/ Fok I transducer gave the same fluorescence activation profile as the folate-labeled DNA/Fok I transducer. However, after addition of FR, the amino-labeled DNA/Fok I transducer displayed no significant variation in its fluorescence response curve. These results indicate that inhibition of endonucleolytic activity of the folate-labeled DNA/Fok I transducer specifically arises from the binding of FR to the folate label rather than nonspecific interaction between FR and the DNA/Fok I transducer. Having inspected the inhibitory role of the potent FR−folate interaction on the DNA/Fok I transducer, we then performed 5712
dx.doi.org/10.1021/ac300889t | Anal. Chem. 2012, 84, 5708−5715
Analytical Chemistry
Article
Thus far, we have demonstrated how small molecule− protein interactions can specifically inhibit the endonuclease activity of the DNA/Fok I transducer. This inhibitory role can be applied to the wide range of affinities (micromolar to subnanomolar dissociation constants) that span most of small molecule−protein interactions of interest in chemistry and biology. With this success, we then developed the small molecule-labeled DNA/Fok I transducer into a versatile and sensitive platform for quantitatively detecting protein targets or competitive binders for the small molecule. Typically, small molecule ligands to be investigated should be derivatized with a certain functionality such as amino, thiol, or carboxyl group, and conjugation of the small molecule through the functional group to DNA should not affect its binding affinity to target protein. Examples of these small molecule ligands include furosemide or 4-carboxybenzenesulfonamide (ligands of carbonic anhydrase II),24 lisinopril or captopril (ligands of angiotensin converting enzyme),25 aspirin (ligand of cyclooxygenase-1), and indomethacin (ligand of cyclooxygenase2).26 Additionally, our method can be used for antibody-based detection of toxin or drug small molecules, which have all been demonstrated in currently available immunoassays for the ability to be conjugated to some carrier protein without loss of affinity to antibodies. Quantitative Analysis of Small Molecule-Binding Protein Using DNA/Fok I Transducer. Figure 3 shows typical time-dependent fluorescence profiles of the folatelabeled DNA/Fok I transducer in response to FR of varying concentrations. The anticipated dose-dependent fluorescence activation profiles were observed, and the end-point fluorescence readouts were dynamically decreased (from ∼3420 to ∼420) with increasing concentrations of FR within the range of 0 pM to 50.0 nM. A plot of end-point fluorescence readout versus logarithmic FR concentrations revealed a quasi-linear correlation ranging from 10.0 pM to 50.0 nM with a detection limit of 5.0 pM (Figure S7 in Supporting Information). Consistent dose-dependent responses were also obtained with the end-point fluorescence spectra (Figure S8 in Supporting Information). Such a picomolar detection limit was ∼1000 times better than those obtained with SPR and FP methods in detecting small molecule−protein interactions with nanomolar dissociation constants,27,28 suggesting that the DNA/Fok I transducer is highly sensitive for detecting small molecule− protein interactions with dissociation constants in the nanomolar range. Also, the DNA/Fok I transducer showed very desirable reproducibility based on its homogeneous assay format. The relative standard deviations of end-point fluorescence readouts were 1.9%, 2.6%, 2.5%, 2.3%, and 1.1% in four repetitive assays of 10 pM, 75 pM, 500 pM, 5 nM, and 50 nM FR, respectively. Figure 4 displays the time-dependent fluorescence responses of the DNA/Fok I transducer to the low-affinity binder of folate, DHFR, with concentrations ranging from 0 nM to 27 μM. The fluorescence profiles showed that the activation rates dramatically decreased with increasing DHFR concentrations. The end-point fluorescence readouts also exhibited a quasilinear correlation to the logarithmic concentrations of DHFR in the range of 27 nM to 5.4 μM with a detection limit of 10 nM (Figure S9 in Supporting Information). This detection limit was ∼100 times better than those obtained with SPR and FP methods in detecting small molecule−protein interactions with micromolar dissociation constants.29,30 The relative standard deviations of end-point fluorescence readouts were 1.4%, 2.3%,
Figure 6. Fluorescence responses of folate-labeled DNA/Fok I transducer in competitive assay of aminopterin. (A) Typical timedependent fluorescence activation profiles obtained using mixtures of 50 nM folate-labeled DNA/Fok I transducer and 2.7 μM DHFR in response to aminopterin of varying concentrations. (B) Corresponding end-point fluorescence readouts versus aminopterin concentrations. Error bars are standard deviations across four repetitive assays.
further experiments using another folate-binding protein, DHFR, with very low affinity to folate (Figure 2C). As expected, in the presence of 5 μM DHFR (for low affinity interactions, high concentration of target was generally required in order to obtain a significant amount of binding complex), a slow rate of fluorescence activation was also observed, revealing a remarkable inhibition of the DNA/Fok I transducer by the low-affinity interaction between DHFR and folate. An additional control with the amino-labeled DNA/Fok I transducer also displayed insignificant inhibition of endonucleolytic activity in the presence of DHFR, further verifying the selectivity in inhibiting the DNA/Fok I transducer by the interaction between DHFR and the folate label. On the other hand, we performed end-point fluorescence spectral measurements to further characterize the DNA/Fok I transducer, which again verified that the folate-labeled DNA/Fok I transducer displayed an endonucleolytic activity specifically inhibited by interaction between folate and FR (Figure S5 in Supporting Information). To demonstrate the generality of the DNA/Fok I transducer for small molecule−protein interaction assays, we performed additional experiments using another small-molecule ligand and its binding proteins, biotin and either SA or biotin antibody. These results verified that binding of SA or Ab to the biotin label also potently inhibited the activity of this DNA/Fok I transducer (Figure S6 in Supporting Information). 5713
dx.doi.org/10.1021/ac300889t | Anal. Chem. 2012, 84, 5708−5715
Analytical Chemistry
Article
lated to the concentrations of aminopterin within the range of 0 nM to 50.0 μM. The relative standard deviations of end-point fluorescence readouts were 2.7%, 1.4%, 2.5%, and 1.3% in four repetitive assays of 27 nM, 270 nM, 1.75 μM, and 2.7 μM aminopterin. This suggested that aminopterin was also a DHFR ligand binding the same site as the folate label. In a generalized sense, this might imply that the DNA/Fok I transducer could be applied to the quantitative analysis of candidate smallmolecule ligands that bind to the protein target at the same site as the small-molecule label. Thus far, our method has been demonstrated to be directly used to quantitatively detect the concentration of protein targets or competitive binders for a small molecule. This method can also be adapted into a competitive immunoassay method for a small-molecule analyte. In this case, we only need to use an antibody against the small-molecule analyte in place of the target protein and add samples containing the smallmolecule analyte instead of the competitive small-molecule binders. This implies the utility of our strategy for sensitive detection of small-molecule biomarkers in clinical diagnostics and public safety analysis. However, because of nonlinear correlation between the fluorescence response and the concentration of Fok I/DNA transducer, this method cannot be used directly for measuring the affinity constant, and determination of the affinity constant may require complicated calculations with nonlinear fitting.
1.9%, and 2.4% in four repetitive assays of 27 nM, 270 nM, 1 μM, and 2.7 μM DHFR. These results, therefore, gave evidence that the DNA/Fok I transducer was adequately sensitive for detecting low-affinity small molecule−protein interactions with dissociation constants in the micromolar range. Additional assays were performed for the DNA/Fok I transducer using other serum proteins, such as human serum albumin, immunoglobulins G, transferrin, α-fetoprotein, carcinoembryonic antigen, and carbohydrate antigen 19-9 (Figure S10 in Supporting Information). These proteins also exerted little interference on fluorescence activation reactions mediated by the DNA/Fok I transducer. This further verified that endonucleolytic inhibition of the DNA/Fok I transducer was specific to the binding of FR to folate, implying high selectivity of the DNA/Fok I transducer for interrogating small molecule−protein interactions. Presumably, the improved selectivity was attributed to the fact that nonspecific interactions were too weak to preclude Fok I from accessing its cleavage site. Quantitative Analysis of Competitive Small-Molecule Binders Using DNA/Fok I Transducer. Besides the detection of binding proteins to the small-molecule label, the DNA/Fok I transducer can also be used to identify unkown small-molecule ligands that bind to the protein target at the same site as the small-molecule label on the DNA/Fok I transducer (Figure 5). Such assays might be used for identifying new ligands of a given protein target using one of its known small-molecule ligands as the label on the DNA/Fok I transducer. In this case, the known small-molecule ligand is labeled on the DNA/Fok I transducer, and the target protein of a limited concentration is incubated with the small-molecule labeled DNA/Fok I transducer, followed by the incubation with one of the candidate ligands. If the candidate ligand binds to the protein target at the same site as the small-molecule label, the equilibrium between the candidate ligand and the target protein will shift the equilibrium between the small molecular label and the target protein, resulting in the displacement of the small-molecule label by the candidate ligand. This can release the protein target from the DNA/Fok I transducer, thereby restoring the DNA/Fok I transducer to its free state and displaying high endonuclease activity (Figure 5). This, therefore, gives a quantitative indicator for the interactions between the candidate ligand and the protein target. Moreover, the concentration of free small molecule-labeled DNA/Fok I transducer will increase with the increasing concentration of the small-molecule ligand, allowing an indirect quantification of the ligand. For this study, we chose aminopterin, a potent smallmolecule ligand of DHFR,31 as the model system. In this assay, the folate-labeled DNA heteroduplex (50 nM) was first incubated with DHFR (2.7 μM) for 10 min. Then, aminopterin of varying concentrations was added and incubated for another 10 min. After the addition of Fok I, the mixtures were immediately subjected to time-dependent fluorescence measurements. If aminopterin binds to the same site as folate on DHFR, the equilibrium between aminopterin and DHFR (even though DHFR is present in excess) can displace the folate label and dissociate DHFR from the DNA/Fok I transducer, thereby recovering the endonuclease activity. Figure 6 shows the fluorescence activation curves corresponding to aminopterin of varying concentrations. It was observed that these fluorescence profiles exhibited clear dose-dependent activation kinetics, and the end-point fluorescence readouts were dynamically corre-
■
CONCLUSIONS We have developed a DNA/Fok I transducer as a novel strategy for homogeneous assays of small molecule−protein interactions. This strategy only requires labeling the small molecule on a DNA heteroduplex, circumventing protein modifications that may be harmful for activity. It is highly sensitive for detecting small molecule−protein interactions, either with high or low affinity, because of the efficient signal amplification from the cyclic cleavage operation. It also exhibits improved selectivity against nonspecific adsorption. This strategy can also be extended for identification of new small-molecule ligands and the immunoassay of small-molecule biomarkers using a competitive assay format. Finally, this strategy can be further adapted for multiplexed and high-throughput applications using multicolor fluorescence-quenched probes and microplate formats with the aid of automated robotic delivery systems. This developed strategy may hold great promise as a sensitive, specific, robust, and high-throughput platform for the identification and quantification of small molecules and their interactions with target proteins of varying affinities.
■
ASSOCIATED CONTENT
S Supporting Information *
HPLC purification of folate-labeled DNA, capillary electrophoresis experiments, additional table, and figures. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*Tel.: 86-731-88664085. E-mail:
[email protected] (J.H.J.);
[email protected]fl.edu (W.T.). Notes
The authors declare no competing financial interest. 5714
dx.doi.org/10.1021/ac300889t | Anal. Chem. 2012, 84, 5708−5715
Analytical Chemistry
■
Article
(31) Cocco, L.; Groof, J. P.; Temple, C., Jr.; Montgomery, J. A.; London, R. E.; Matwiyoff, N. A.; Blakley, R. L. Biochemistry 1981, 20, 3972−3978.
ACKNOWLEDGMENTS This work was supported by NSFC (21025521, 21035001, 91117006), National Key Basic Research Program (2011CB911000), European Commission FP7-HEALTH2010 Programme-GlycoHIT (260600), CSIRT Program, and NSF of Hunan Province (10JJ7002).
■
REFERENCES
(1) Griffin, J. L.; Shockcor, J. P. Nat. Rev. Cancer 2004, 4, 551−561. (2) Denkert, C.; Budczies, J.; Kind, T.; Weichert, W.; Tablack, P.; Sehouli, J.; Niesporek, S.; Könsgen, D.; Dietel, M.; Fiehn, O. Cancer Res. 2006, 66, 10795−10804. (3) Sreekumar, A.; Poisson, L. M.; Rajendiran, T. M.; Khan, A. P.; Cao, Q.; Yu, J.; Laxman, B.; Mehra, R.; Lonigro, R. J.; Li, Y.; Nyati, M. K.; Ahsan, A.; Kalyana-Sundaram, S.; Han, B.; Cao, X.; Byun, J.; Omenn, G. S.; Ghosh, D.; Pennathur, S.; Alexander, D. C.; Berger, A.; Shuster, J. R.; Wei, J. T.; Varambally, S.; Beecher, C.; Chinnaiyan, A. M. Nature 2009, 457, 910−915. (4) Preul, M. C.; Caramanos, Z.; Collins, D. L.; Villemure, J. G.; Leblanc, R.; Olivier, A.; Pokrupa, R.; Arnold, D. L. Nature Med. 1996, 2, 323−325. (5) Cooper, M. A. Nat. Rev. Drug Discovery 2002, 1, 515−528. (6) Bachovchin, D. A.; Brown, S. J.; Rosen, H.; Cravatt, B. F. Nat. Biotechnol. 2009, 27, 387−394. (7) Zhang, J.; Campbell, R. E.; Ting, A. Y.; Tsien, R. Y. Nat. Rev. Mol. Cell. Biol. 2002, 3, 906−918. (8) Kerppola, T. K. Nat. Rev. Mol. Cell Biol. 2006, 7, 449−456. (9) Melkko, S.; Scheuermann, J.; Dumelin, C. E.; Neri, D. Nat. Biotechnol. 2004, 22, 568−574. (10) Doyon, J. B.; Snyder, T. M.; Liu, D. R. J. Am. Chem. Soc. 2003, 125, 12372−12373. (11) Harris, D. C.; Chu, X. Z.; Jayawickramarajah, J. J. Am. Chem. Soc. 2008, 130, 14950−14951. (12) Wu, Z.; Zhen, Z.; Jiang, J. H.; Shen, G. L.; Ru, Q. Y. J. Am. Chem. Soc. 2009, 131, 12325−12332. (13) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Science 1997, 277, 1078−1081. (14) Liu, J. W.; Lu, Y. J. Am. Chem. Soc. 2007, 129, 9838−9839. (15) Ali, M. M.; Aguirre, S. D.; Lazim, H.; Li, Y. F. Angew. Chem., Int. Ed. 2011, 50, 3751−3754. (16) Zhang, Y. L; Huang, Y.; Jiang, J. H.; Shen, G. L.; Yu, R. Q. J. Am. Chem. Soc. 2007, 129, 15448−15449. (17) Kim, S. C.; Skowron, P. M.; Szybalski, W. J. Mol. Biol. 1996, 258, 638−649. (18) Bitinaite, J.; Wah, D. A.; Aggarwal, A. K.; Schildkraut, I. Proc. Natl. Acad .Sci. USA 1998, 95, 10570−10575. (19) Acharya, G.; Hartmann, L. C.; Low, P. S.; Savran, C. A. J. Am. Chem. Soc. 2007, 129, 15824−15829. (20) Greenfield, N. J.; Williams, M. N.; Poe, M.; Hoogsteen, K. Biochemistry 1972, 11, 4706−4711. (21) Bayer, E. A.; Wilchek, M. Methods Enzymol. 1990, 184, 49−51. (22) Berger, M. Biochemistry 1975, 14, 2338−2342. (23) Zuker, M. Nucleic Acids Res. 2003, 31, 3406−3415. (24) Papalia, G.; Leavitt, S.; Bynum, M. A.; Katsamba, P. S.; Wilton, R.; Qiu, H.; Stekers, M.; Wang, S.; Bindu, L.; Myszka, D. G. Anal. Biochem. 2006, 359, 94−105. (25) Dilsizian, V.; Eckelman, W. C.; Loredo, M. L.; Jagoda, E. M.; Shirani., J. J. Nucl. Med. 2007, 48, 182−187. (26) Blanco, F. J.; Guitian, R.; Moreno, J.; De Toro, F. J.; Galdo, F. J. Rheumatol. 1999, 26, 1366−1373. (27) Gillis, E. H.; Gosling, J. P.; Sreenan, J. M.; Kane, M. J. Immunol. Methods 2002, 267, 131−138. (28) Vilarino, N.; Fonfria, E. S.; Molgo, J.; Araoz, R.; Botana, L. M. Anal. Chem. 2009, 81, 2708−2714. (29) Yu, L. L.; Zhong, M.; Wei, Y. A. Anal. Chem. 2010, 82, 7044− 7048. (30) Hsieh, H. V.; Pfeiffer, Z. A.; Amiss, T. J.; Sherman, D. B.; Pitner, J. B. Biosens. Bioelectron. 2004, 19, 653−660. 5715
dx.doi.org/10.1021/ac300889t | Anal. Chem. 2012, 84, 5708−5715
