DNA Encountering Terbium(III): A Smart “Chemical ... - ACS Publications

Subscriber access provided by UNIV OF DURHAM

Article

DNA encountering terbium(III): A smart “chemical nose/tongue” for largescale time-gated luminescent and lifetime-based sensing Shi-Fan Xue, Zi-Han Chen, Xin-Yue Han, Zi-Yang Lin, Qi-Xian Wang, Min Zhang, and Guoyue Shi Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b05167 • Publication Date (Web): 13 Feb 2018 Downloaded from http://pubs.acs.org on February 13, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

DNA encountering terbium(III): A smart “chemical nose/tongue” for largescale time-gated luminescent and lifetime-based sensing Shi-Fan Xue, Zi-Han Chen, Xin-Yue Han, Zi-Yang Lin, Qi-Xian Wang, Min Zhang*, Guoyue Shi* Lab of Biochemical Sensing Technology, School of Chemistry and Molecular Engineering, Shanghai Key Laboratory for Urban Ecological Processes and Eco-Restoration, East China Normal University, 500 Dongchuan Road, Shanghai 200241, China. Phone & Tel: +86-21-54340042; Email: [email protected]；[email protected]. ABSTRACT: Recent years have witnessed the rapid development of pattern-based sensors due to their potential to detect and differentiate a wealth of analytes with only few probes. However, no one has found or used the combination of DNA and terbium(III) (Tb) as a pattern recognition system for largescale mix-and-measure assays. Here we report for the first time that DNA-sensitized Tb (DNA/Tb), as a label-free and versatile “chemical nose/tongue”, can be employed for widescale time-gated luminescent (TGL) monitoring metal ions covering nearly the entire periodic table in a cost-effective fashion. A series of guanine/thymine (G/T)-rich DNA ligands was screened to sensitize the luminescence of Tb (referring to antenna effect) as smart pattern responders to metal ions in solution, and metal ion-DNA interactions can differentially alter the antenna effect of DNA toward Tb as pattern signals. Our results show that as few as 3 DNA/Tb label-free sensors could successfully discriminate 49 analytes, including alkali metal ions, alkaline earth metal ions, transition/post-transition metal ions, and lanthanide ions. A blind test with 49 metals further confirmed the discriminating power of DNA/Tb sensors. Moreover, the lifetime-based pattern recognition application using DNA/Tb sensors was also demonstrated. This DNA/Tb pattern recognition strategy could be extended to construct a series of “chemical noses/tongues” for monitoring various biochemical species by using different responsive DNA ligands, thus promising a versatile and powerful tool for sensing application and investigating of DNA-involving molecular interactions.

Introduction The determination of biochemical species is the linchpin of bioanalytical chemistry, which is of essential importance for widespread applications, e.g. environmental monitoring, food safety, clinical diagnostics, etc.1-3 In this respect, luminescent sensors have acted as effective detection tools since they have many advantages including simple/rapid implementation, real-time monitoring specialty, easily interpreted outputs, and good reproducibility.4,5 While most luminescent sensors, to some extent, experience the problem of a background signal being observed in the absence of an analyte, which is unfavorable to the sensitivity and dynamic range that can be achieved.6,7 An ideal way to efficiently eliminate the interferences from background signal is time-gated luminescent (TGL) detection technique, which utilizes the temporal domain to distinguish their long-lifetime luminescence from short-lifetime autoluminescence.8,9 Typically, lanthanide-derived luminescent probes have attracted great attention owing to their unique spectroscopic properties, such as long luminescence lifetime, sharp emission bands, and large Stokes.10-12 These special characteristics make a boost to usefully address the issue of background signal and endow them with appropriateness for TGL detection. Since the traditional sensing method has utilized one sensor per analyte, it is often needed intensive labor to develop analyte-specific sensors. Considering this issue, much attention has been actively devoted to integrate with pattern recognition for exploiting array or differential sensors termed “chemical noses/tongues”, which can discriminate substantial analytes by investigating the combined pattern responses from a set of sensing elements. Remarkable examples have been reported for differentiating serum proteins with protein-nanoparticle conjugates,13 metal ions with oligodeoxyfluoroside chemosensors,14 and proteins with dye-labeled ensemble aptamers.15 Despite the progress, further

development of the pattern-based assay for largescale applications is hampered by limitations with respect to some sophisticated or expensive operation (e.g. high-cost complicated synthesis, and difficult design of pattern recognition elements). Thus, seeking available, cost-effective yet versatile pattern recognition elements is indispensable for the development of “chemical noses/tongues” with superior discrimination ability. DNA is fully suitable for this purpose. It is a negatively charged polymer of nucleotide bases (A, T, G or C) associated with phosphate groups, and possesses programmable composition and conformation thus providing high diversity, even a short 20-base DNA sequence contains as many as billions of combinations.16 Importantly, nearly any DNA sequence can readily be chemically synthesized and commercially available with robust stability, high purity, and low cost. Inspired by the above facts, we herein present a fundamentally new pattern-based detection platform via the combination of DNA with a common lanthanide ion (terbium, Tb), in which DNA-sensitized Tb (DNA/Tb) acts as a label-free and versatile “chemical nose/tongue” for widescale TGL sensing in a cost-effective manner (Scheme 1). Our presented “chemical nose/tongue” relies on the commercially available single-stranded DNA (ssDNA) as an antenna ligand for sensitizing the luminescence of Tb in aqueous solution (i.e. antenna effect). Subsequent addition of metal ions to interact with ssDNA can cause the alteration of antenna effect accompanying with substantial luminescence responses of Tb. We have previously described a series of guanine/thymine-rich DNA ligands can highly sensitize the luminescence of Tb.17 Using the same library, we unveiled 3 ligands that are effective in the regulation of antenna effect for yielding strong and varied luminescence responses of Tb to a broad range of metal ions, and successfully demonstrated as few as 3 DNA/Tb label-free and mix-and-measure sensors can be applied to discriminate 49 different metal species in

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

aqueous buffer. All of 49 metal ions have also been correctly verified in a blind test, confirming the discrimination power of our DNA/Tb pattern recognition system. Moreover, the long luminescence lifetime of DNA/Tb sensors enables background-free TGL and lifetime-based metal ion assays and investigations of metal ion-DNA interactions. This new DNA/Tb pattern recognition concept may be referentially expanded for detecting biochemical targets relating to different responsive DNA ligands and further assessing their molecular interactions

Scheme 1. Strategy of DNA/Tb-based pattern recognition. (a) Scheme illustration of the luminescence sensitization of Tb by an antenna effect from G3T oligo. (b) Mode of DNA/Tb sensor array ([G3T]3/Tb, [G3T]5/Tb, and [G3T]6/Tb) for metal ion differentiate and mapping of metal ion-DNA interactions by the pattern responses of luminescence in a time-gated and lifetime-based signal readout.

Experimental Section Chemicals. All DNA ligands were commercially available from Sangon Inc. (Shanghai, China). The following metal salts, CuCl2, PbCl2, CoCl2, ZnCl2, CdCl2, FeCl3, NiCl2, AgCl, AlCl3, MnCl2, FeCl2, CrCl3 and Hg(Ac)2 were reagent-grade and purchased from Sinopharm Chemical Reagent Co. Ltd (Shanghai, China). The following rare-earth-metal salts, Tb(NO3)3, Y(NO3)3, Eu(NO3)3, Pr(NO3)3, Ho(NO3)3, Nd(NO3)3, La(NO3)3, Er(NO3)3, Yb(NO3)3, Dy(NO3)3, Gd(NO3)3, Ce(NO3)3, Sm(NO3)3, Lu(NO3)3, Tm(NO3)3 were purchased from Diyang Chemical (Shanghai) Co. Ltd. 10×Tris-HCl buffer (100 mM, pH 7.4) was prepared using metal-free reagents in distilled water purified by a Milli-Q water purification system. All chemicals used in this work were analytical purity grade and obtained from commercial sources and directly used without further purification. Apparatus. Luminescence spectra were recorded by a microplate reader (infinite M200 pro, TECAN, Switzerland) using a black 384 well microplate (Corning, U.S.A.). The excitation wavelength used was 280 nm for the emission spectra. For the time-gated luminescence (TGL) spectra, a delay time of 50 µs and a gate time of 2 ms were used. The spectra of luminescence lifetime were measured by using FLS980 Fluorescence Spectrometer (Edinburgh Instruments Ltd, UK). Circular dichroism spectra were recorded by using Chirascan Circular Dichroism Spectrometer (Applied Photophysics, UK). Date analysis. Statistical Product and Service Solutions 22.0 software (IBM) was used to process principal component analysis (PCA) and hierarchical cluster analysis (HCA). Each

Page 2 of 7

sample was repeated in quintuplicate. The GraphPad Prism 7.0 software (GraphPad Software, San Diego, CA) was used to perform the data processing. Investigation of DNA to sensitize the fluorescence of Tb. DNA with varied sequences were prepared (Table S1). Different concentrations of DNA and Tb (10 µM) was prepared in 10 mM Tris-HCl buffer (pH 7.4). The mixture was incubated for 10 min at room temperature. Then, the time-gated luminescence excited at 280 nm was recorded; a delay time of 50 µs and a gate time of 2 ms were used. The optimal concentration of DNA was investigated. Luminescent assay of metal ions by DNA/Tb sensors. G3T oligos (8 µM [G3T]3, 4 µM [G3T]5, and 2 µM [G3T]6) were respectively mixed with 10 µM Tb3+ in 10 mM Tris-HCl buffer (pH 7.4) as DNA/Tb sensor array. After 10 min, 10 µL different kinds of metal ions were mixed with 90 µL DNA/Tb sensors, respectively. Then the mixtures were incubated for 4 min at room temperature. The resulting mixtures were placed in the black 384-well microplate and then shaked 10 s before being measured the time-gated luminescence (TGL) intensities. The relative fluorescence variation ((L0-L)/L0) was used as the TGL response, where L0 and L are the TGL intensities of DNA/Tb probes at 546 nm in the absence and presence of metal ions. Each metal ion has five replicates to react with DNA/Tb probes in the same condition. The raw data matrix (3 DNA/Tb probes×number of metal ions×5 replicates) was obtained. Principal component analysis (PCA) was used to process the multivariate pattern data. As a statistical analysis strategy, PCA could transform a series of data into some linearly unrelated components by using matrix conversion. To test the unknown metal ions, 10 µL of randomly selected metal ion was added into 90 µL DNA/Tb sensors, respectively. Then the TGL intensities were measured by the same way. Hierarchical cluster analysis (HCA), a statistical method to separate metal ions into different groups, was used to assess the similar responses to the metal ions. The Euclidean distance was chosen as the distance metric to describe the similarity among distinct analytes.

Results and Discussion Design of DNA/Tb-based pattern recognition assay. Tb alone is very poor at absorbing light directly due to its Laporte-forbidden f–f transitions, so direct excitation of Tb to luminesce is typically difficult.18 However, the luminescence of Tb can be sensitized by DNA as antenna ligand (the so called “antenna effect”) in a sequence and structure dependent manner.19 Especially, our previous studies have revealed that ssDNA with guanine (G)/thymine (T)-rich sequences can greatly enhance the Tb emission via efficient intermolecular energy transfer from GT-rich DNA to Tb.17 It is reported that nucleic acids (e.g. DNA) can interact with metal ions via their nucleobases and phosphate backbone,20,21 and functional nucleic acids with specific sequences (e.g. aptamer) have been demonstrated to undergo the conformational alteration for binding any targets of choice, such as metal ions, small molecules, proteins, etc.22 With the above facts in mind, the utilization of designed DNA would not only allow the tunable sensitization of the luminescence of Tb via structure-responsive antenna effect, but also the recognition and binding of analytes via target-induced allosteric effect, thus providing the opportunity to construct a novel DNA/Tb-based pattern recognition system. To find pattern recognition elements, we employed a DNA library composed

ACS Paragon Plus Environment

Page 3 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry of repeated G3T oligos (Table S1). Such DNA library shows large different sensitizing responses of the luminescence of Tb, and 3 G3T oligos ([G3T]3, [G3T]5, and [G3T]6) display a better quality for effective sensitizing of the luminescence of Tb under a 1:1 molar ratio (Figure S1). In this work, as a proof-of-concept, we selected these 3 G3T oligos ([G3T]3, [G3T]5, and [G3T]6) as antenna ligands of Tb to build a DNA/Tb-based label-free sensor array (Scheme 1). To further study the diversity of antenna effects between the chosen combination of 3 DNA/Tb sensors, the stoichiometry titration experiments were performed to explore the ratio between DNA and Tb (Figure S2). [G3T]3/Tb, [G3T]5/Tb, and [G3T]6/Tb sensors exhibit the peak antenna effects in various ratios (4:5, 2:5, and 1:5, respectively). Unless noted otherwise, the following assays were all carried out under this setting. Metal ions are vital to various biological, chemical, and environmental processes.23 Metal ion coordination to DNA is necessary for the biological action of DNA, and its different effect on the structural of DNA is also an open yet hot issue.24,25 For this reason, metal ions were chosen as model analytes to test the feasibility of DNA/Tb sensor array ([G3T]3/Tb, [G3T]5/Tb, and [G3T]6/Tb) for smart analyte differentiate and mapping of their molecular interactions by the pattern responses of luminescence (Scheme 1). Preliminary tests at two metal ions (Ag and Cr) (H2O as the Blank sample) demonstrated that luminescence-response pattern of the analytes was obtained due to their interactions with DNA/Tb sensors (Figure 1a-e). The luminescence responses of DNA/Tb sensors to these two metal ions are rapid, and can reach equilibrium within 4 min (Figure S3). DNA/Tb sensors challenged with Ag would display various luminescence enhancements (i.e., (L0-L)/L00) (Figure 1e). Principal component analysis (PCA), a powerful statistical technique,26 was then employed to quantitatively distinguish the resultant patterns of DNA/Tb sensors with Blank, Ag and Cr (Figure 1f). Five replicates of luminescence responses were collected for each metal ion against DNA/Tb sensors, and then subjected to PCA to gain three canonical factors (97.16, 2.79, and 0.05%), which stand for linear combinations of the luminescence response matrix (3 DNA/Tb sensors × number of analytes × 5 replicates). The two most significant factors were utilized to form a 2D plot (Figure 1f), in which each point denotes the response pattern for an individual sample against the DNA/Tb sensor array. The 15 canonical luminescence response patterns (3 analytes × 5 replicates) can be clustered into 3 distinct groups (Figure 1f). These three analytes (two metal ions, and one Blank) were obviously differentiated in this luminescent DNA/Tb-based pattern recognition system, confirming the feasibility of DNA/Tb sensor array as a novel label-free chemical nose/tongue. To decipher the possible mechanism involved in this system, luminescence lifetime and circular dichroism (CD) spectroscopy were used to characterized the interaction of DNA/Tb sensors with Blank, Ag and Cr, respectively. The DNA/Tb sensors showed long luminescence lifetimes of more than 0.4 ms (Figure 2d), promising the above label-free chemical nose/tongue suitable for background-free TGL metal ion assays. In a typical TGL assay, we can collect the TGL signal by setting an adequate delay time and gate time under pulsed excitation, which would remove the short-lived

background noise within the delay time and records only the long-lived luminescent signal from DNA/Tb sensors within the gate time. To demonstrate it, we investigated the responses of DNA/Tb sensors toward Blank, Ag and Cr in the presence of 0.4 µg/mL riboflavin (Ri) as a background additive (Figure S4). As shown in Figure S4a-c, there are almost background signals from Ri in the condition of steady-state luminescent measures (i.e., without delay time and with a 20 µs short gate time). In contrast, the signals of both DNA/Tb sensors and Ri can be observed in another measurement condition without delay time and with a 2000 µs long gate time (Figure S4d-f). While background-free signals of DNA/Tb sensors would be monitored with a TGL measurement setting of 50 µs delay time and 2000 µs gate time (Figure S4g-i), which reveals the advantage of TGL assay.

Figure 1. Preliminary identification of metal ions with DNA/Tb sensor array in a time-gated signal readout. TGL spectra of (a) [G3T]3/Tb, (b) [G3T]5/Tb and (c) [G3T]6/Tb against Ag and Cr (5 µM) (H2O as Blank), respectively. (d) TGL response patterns of DNA/Tb sensor array against Ag and Cr. (e) Heat map derived from the TGL response patterns of DNA/Tb sensor array against Ag and Cr. (f) 2D canonical score plot for the TGL response patterns as obtained from PCA against Ag and Cr (5 µM). Apart from TGL-based assay, the long lifetime of DNA/Tb would also allow to produce novel lifetime-based sensor array for metal ions discrimination. DNA/Tb sensors challenged with Ag and Cr show distinct changes in luminescence lifetimes in terms of molecular interactions within the samples (Figure 2a-e), which can also be interpreted from the results of CD spectroscopy (Figure S5). CD spectroscopy is widely used to characterize the structures of DNA. From Figure S5, we can notice that G3T oligos ([G3T]3, [G3T]5, and [G3T]6) presented a positive CD band around 260 nm and a negative band around 240 nm, and the binding of Tb to G3T oligos (DNA/Tb) caused some structural changes of G3T oligos (i.e. alterations in CD bands), resulting in certain spatial conformations to discriminately sensitize the luminescence of Tb with the different lifetimes indicated. Upon DNA/Tb sensors exposed to Ag and Cr, further changes in the CD spectra can be observed in such complex due to their molecular interactions, accompanying with significant pattern differences both in TGL intensities (Figure 1e) and lifetimes (Figure 2e). Besides, PCA was also wielded to differentiate these lifetime patterns of DNA/Tb sensors with Blank, Ag and Cr, a useful lifetime-based pattern recognition strategy can be accordingly realized using DNA/Tb sensors (Figure 2f). It is known that

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

intensity-based luminescent measurements are simple and accurate, but their performances largely depend on many factors, such as the intensity of exciting light, the extinction coefficient and concentration of the probe, the optical path length, the detector sensitivity, etc.27 Compared to intensity-based measurement, lifetime-based sensing is more robust because it is independent of intensity changes.28 Thus, it can be envisaged that the proposed lifetime-based pattern recognition concept would boost new opportunities for design of more novel pattern recognition strategies and expansion of wide application in different areas.

Figure 2. Preliminary identification of metal ions with DNA/Tb sensor array in a lifetime-based signal readout. Luminescence lifetime spectra of (a) [G3T]3/Tb, (b) [G3T]5/Tb and (c) [G3T]6/Tb against Ag and Cr (5 µM), respectively. (d) Luminescence lifetime (τ) response patterns of DNA/Tb sensor array against Ag and Cr. (e) Heat map derived from the τ response patterns of DNA/Tb sensor array against Ag and Cr. (f) 2D canonical score plot for the τ response patterns as obtained from PCA against Ag and Cr. Largescale detection and identification of metal ions. To broadly evaluate the differentiation power of these DNA/Tb sensors, we included essentially all the water-soluble metal ions (49 species in total) as analytes, including alkali metal ions, alkaline earth metal ions, transition/post-transition metal ions, and lanthanide ions. The most stable or common form of metals was chosen, in that metals may have multiple redox states. Alkali metal ions have low charge density,29 thus they are difficult to be detected and differentiated using luminescence sensors and relatively high concentrations are usually required.14,30 The full set of 49 metals was cross-screened with our 3 DNA/Tb sensors, using 500 µM for alkali metal ions and 5 µM for others to evaluate the largescale differentiation and detection of these metal species. Each luminescence response for a given metal was surveyed with 3 separate DNA/Tb sensors to examine reproducibility, and changes in TGL intensity were measured, generating pattern (L0-L)/L0 data and its corresponding heat map (Figure S6-S9). As a control experiment, the luminescence responses of the common counterions (fluoride, chloride, nitrate, sulfate, and ammonium) were also measured. Negligible signals were observed at 100 µM (Figure S10), verifying that the luminescence responses of DNA/Tb sensors indeed result from the metal ions. PCA and hierarchical cluster analysis (HCA) were then employed to quantitatively evaluate the metal ion-induced such pattern responses of DNA/Tb sensors. For the PCA, pattern responses from 5 alkali metal ions, 4

Page 4 of 7

alkaline earth metal ions, 24 transition/post-transition metal ions, and 16 nonradioactive lanthanide ions were clearly discriminated with high identification accuracy, respectively (Figure 3a, c, e, f). Similarly, the HCA using the Euclidean Distances unveiled the same results (Figure S11-S14), with all the experimental tests from each metal ion clustered and distinct from the other metal ions. The results of both PCA and HCA indicate that the luminescence patterns are closely connected with metal properties. For example, most alkali metal ions are grouped more closely except for K (Figure 3a, 3b and S11), which can be interpreted that K-binding G-rich DNA would form specific structure,31 resulting in unique pattern responses for understanding the K-DNA interactions. Likewise, the coordination of Ag to GT bases increases the rigidities of DNA ligands,22,32 thus enhancing the luminescence of Tb (Figure 1a-d) and Ag possesses more dissimilarity compared to other transition/post-transition metal ions (Figure 3e). Notably, this simple pattern recognition system can discriminate between highly similar lanthanide ions while allowing their analysis and grouping even with just 3 DNA/Tb sensors (Figure 3f), and we can clearly see that Tb locates more uniquely in the PCA plot than other grouped lanthanide ions due to its distinct interaction with DNA.12 Apart from the above TGL pattern recognition, lifetime-based pattern responses from two sets of metal ions (alkali and alkaline earth metals as examples) were demonstrated to be highly differentiated, respectively (Figure 3b, d), which have much similarity with the TGL results (Figure 3a, c). Thus, our proposed DNA/Tb-based pattern recognition system can be a smart “chemical nose/tongue” to identify largescale different metals with sufficient discriminating powers in both TGL and lifetime-based signal readout.

Figure 3. Largescale pattern differentiation of metal species. Canonical score plots of DNA/Tb sensors TGL responses to (a) alkali metals, (c) alkaline earth metals, (e) transition/post-transition metals, and (f) lanthanides. Canonical score plots of DNA/Tb sensors lifetime responses to (b) alkali metals, and (d) alkaline earth metals. Metal concentrations are 500 µM in (a, b) and 5 µM in (c−f). To test the sensitivity of DNA/Tb sensor array, four metal ions (K, Sr, Pb, and Tb) representing alkali, alkaline earth, transition/post-transition metal ions and lanthanide ions were respectively chosen to evaluate its analytical performances (Figure 4a-h, Figure S15-S18). Since the second discriminant factor (PC2) was smaller than 40%, it was possible simply to employ the first discriminant factor (PC1) to characterize the concentrations of metal ions (Figure 4a, c, e, g). Hence, this DNA/Tb sensor array was fully sensitive to determine these metal ions at subnanomolar concentrations (Figure 4b, d, f, h),

ACS Paragon Plus Environment

Page 5 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry the linear detection range of K is 0-0.1 mM and that of other metal ions (Sr, Pb, and Tb) are 0-5 µM, respectively. It is known that lanthanides aren’t easy to differentiate due to their similar chemical properties.33 It is noteworthy that, in this study, the discrimination of 16 lanthanide ions using DNA/Tb sensor array is effective (Figure 3f), moreover the sensitive detection of individual lanthanides (e.g. Tb) can be favorably realized (Figure 4h). More importantly, to the best of our knowledge, this is the first report about using lanthanide-based probe for the label-free detection and discrimination of lanthanide ions. The metal ion recognition ability of DNA/Tb sensors was not sacrificed in mixtures. We tested a series of mixtures of K and Rb, Ba and Ca, Ru and Ti, Lu and La with different molar ratios (Figure 4i-l, Figure S19-S22). These mixtures were clearly distinguished from each other in the PCA plot (Figure 4i-l). Additionally, we proved that DNA/Tb sensors were extremely sensitive to the valence states of metal ion (Figure S23). As depicted in Figure S23, a mixture of different valence states of Fe (FeII and FeIII) can be well differentiated from each other in the PCA plot. The two concentration levels of five metal ions were tested using DNA/Tb sensors to verify their extensible ability (Figure S24). The results revealed 10 groups of samples were effectively differentiated in 10 isolated clusters in the PCA plot, demonstrating that this platform would potentially allow analysis of complex composition. Furthermore, a list of unknown samples was used to probe the robustness of the developed smart DNA/Tb sensor array. In a blind test, all of 49 samples were identified correctly using DNA/Tb sensors, guaranteeing the identification accuracy of 100% (Table S2).

formed by spiking Eu and Ce at varying molar ratios (Eu/Ce = 0:1, 0.3:0.7, 0.4:0.6, 0.75:0.25, 1:0, all in µM) to lake water obtained from Cherry River, Shanghai. Six test samples were prepared for comparison: A) 0.8 µM Cu + 0.2 µM Pb, B) 0.5 µM Cu + 0.5 µM Pb), C) 0.3 µM Cu + 0.7 µM Pb), D) 0.8 µM Eu + 0.2 µM Ce, E) 0.5 µM Eu + 0.5 µM Ce, and F) 0.25 µM Eu + 0.75 µM Ce. PCA was used to data collected from the above standard mixtures and test samples (Figure 5). From the resultant PCA plot, the test samples can be noticed in the correct location according with their metal ratios. As an example, sample A was present between the two standards with Cu/Pb molar ratios at 1:0 and 0.75:0.25, respectively. This location illustrates that it contains 0.75-1 µM Cu and 0-0.25 µM Pb, which agrees with the results detected by ICP-AES. The table in Figure 5 summarizes the detailed data and emphasizes the ability of DNA/Tb sensor array for semi-quantitatively monitoring metal ions in complex environmental samples. The simple label-free luminescent array is a cost-effective, mix-and-measure method comparable to or better than ICP-AES analysis (Table S3).

Figure 5. Analysis of metal mixture in environmental sample. Canonical score plot for TGL response patterns obtained with DNA/Tb sensors against the standard mixtures and test samples of (a) Cu/Pb, and (b) Eu/Ce. Table shows the performance of analysis compared with ICP-AES.

Conclusion

Figure 4. Identification of metal ions at various concentrations using DNA/Tb sensor array. Canonical score plot for TGL response patterns obtained with DNA/Tb sensors against different concentrations of (a) K, (c) Sr, (e) Pb, and (g) Tb. Plot of the first discriminant factor (PC1) vs the concentrations of (b) K, (d) Sr, (f) Pb, and (h) Tb. Canonical score plot for DNA/Tb sensors against mixtures of (i) K and Rb, (j) Ba and Ca, (k) Ru and Ti, (l) Lu and La with different molar ratios. To elucidate the effectiveness of DNA/Tb sensors toward environmentally relevant mixtures, we prepared a series of complex samples. One set of samples was created by spiking Cu and Pb at various molar ratios (Cu/Pb = 1:0, 0.75:0.25, 0.4:0.6, 0.25:0.75, 0:1, all in µM), and the other set was

We have proven the first example of practical implementation of DNA/Tb sensor array as a label-free “chemical nose/tongue”, with the realization to the largescale detection of metal ions. Such DNA/Tb-based pattern sensing system depends on the synergetic usage of DNA-sensitized luminescence of Tb and metal/DNA interaction-induced alteration of antenna effect. Just only 3 DNA/Tb label-free sensors were used to sensitize the luminescence of Tb (referring to antenna effect) for constructing sensor array to strong pattern recognizing metal ions in solution, and metal ion-DNA interactions can diversely alter the preceding antenna effect as time-gated luminescent and lifetime-base pattern signals. The resulting DNA/Tb-based pattern sensing strategy offers multidimensional yet cost-effective powers for discriminating 49 metal species, including alkali metal ions, alkaline earth metal ions, transition/post-transition metal ions, and lanthanide ions. Unlike common pattern-based sensor

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

array, our label-free DNA/Tb platform avoids high-cost complicated synthesis (i.e. all regents used in the present study are commercially available with low price) and is largely free of difficulties in designing pattern recognition configurations. DNA/Tb-based “chemical noses/tongue” can be expanded to measure various biochemical species by integrating different functional DNA ligands. Owing to its intriguing features, we expect that the novel DNA/Tb-based differential sensors will lead to a host of versatile and powerful sensing applications, and play a vital role in the research of DNA-related molecular interactions.

Supporting Information This Supporting Information is available free of charge via the Internet at http://pubs.acs.org. Additional information as noted in text. Sequences of the Oligonucleotides; Luminescence spectroscopy of DNA/Tb; Luminescence titration plot of DNA/Tb; Steady-state luminescent and time-gated luminescent responses of DNA/Tb sensor array toward Blank, Cr and Ag upon the presence of riboflavin a background additive; Circular dichroism (CD) spectra of DNA, DNA/metal ions; Identification of metal ions (alkali, alkaline earth, transition/post-transition, lanthanide) with DNA/Tb sensor array; Investigating the effects of counterions on DNA/Tb sensor array; Dendrogram from agglomerative hierarchical clustering (AHC) of detecting alkali, alkaline earth, transition/post-transition and lanthanide ions; Identification of various concentrations of K, Sr, Pb, Tb with DNA/Tb sensor array; Identification of the mixture of K and Rb, Ba and Ca, Ru and Ti, Lu and La, FeII and FeIII with DNA/Tb sensor array; Identification of the two concentration levels of five metal ions (K, Mg, Ru, Ni, and Sc) with DNA/Tb sensor array; Identification of unknown metal ions using DNA/Tb sensor assay; Comparison of the performance between ICP-AES and our method.

AUTHOR INFORMATION Corresponding Author *Min Zhang, Email: [email protected] *Guoyue Shi, Email: [email protected] Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This work was supported by the National Natural Science Foundation of China (21775044, 21675053, 21635003, 21405047). M.Z. also thanks the support from the China Scholarship Council, Ministry of Education of China (201706145029).

REFERENCES (1) Hacia, J. G.; Brody, L. C.; Chee, M. S.; Fodor, S. P. A.; Collins, F. S. Nat. Genet. 1996, 14, 441-447. (2) Liu, Y. Q.; Zhang, M.; Yin, B. C.; Ye, B. C. Anal. Chem. 2012, 84, 5165-5169. (3) Zhang, M.; Le, H. N.; Wang, P.; Ye, B. C. Chem. Commun. 2012, 48, 10004-10006.

Page 6 of 7

(4) Anzenbacher, P.; Tyson, D. S.; Jursíková, K.; Castellano, F. N. J. Am. Chem. Soc. 2002, 124, 6232-6233. (5) Ma, D.; Li, B.; Zhou, X.; Zhou, Q.; Liu, K.; Zeng, G.; Li, G.; Shi, Z.; Feng, S. Chem. Commun. 2013, 49, 8964-8966. (6) Hagan, A. K.; Zuchner, T. Anal. Bioanal. Chem. 2011, 400, 2847-2864. (7) Wang, Q. Xue, S. F.; Chen, Z. H.; Ma, S. H.; Zhang, S.; Shi, G.; Zhang, M. Biosens. Bioelectron. 2017, 94, 388-393. (8) Zhu, D.; Chen, Y.; Jiang, L.; Geng, J.; Zhang, J.; Zhu, J. Anal. Chem. 2011, 83, 9076-9081. (9) Lu, Y.; Xi, P.; Piper, J. A.; Huo, Y.; Jin, D. Sci. Rep. 2012, 2, 837. (10) Bünzli, J. C. G.; Piguet, C. Chem. Soc. Rev. 2005, 34, 1048-1077. (11) Heffern, M. C.; Matosziuk, L. M.; Meade, T. J. Chem. Rev. 2014, 114, 4496-4539. (12) Zhang, M.; Qu, Z. B.; Han, C. M.; Lu, L. F.; Li, Y. Y.; Zhou, T.; Shi, G. Chem. Commun. 2014, 50, 12855-12858. (13) De, M.; Rana, S.; Akpinar, H.; Miranda, O. R.; Arvizo, R. R.; Bunz, U. H. F.; Rotello, V. M. Nat. Chem. 2009, 1, 461-465. (14) Yuen, L. H.; Franzini, R. M.; Tan, S. S.; Kool, E. T. J. Am. Chem. Soc. 2014, 136, 14576-14582. (15) Pei, H.; Li, Jiang.; Lv, M.; Wang, J.; Gao, J.; Lu, J.; Li, Y.; Huang, Q.; Hu, J.; Fan, C. J. Am. Chem. Soc. 2012, 134, 13843-13849. (16) Clelland, C. T., Risca, V.; Bancroft, C. Nature 1999, 399, 533-534. (17) Zhang, M.; Le, H. N.; Jiang, X. Q.; Yin, B. C.; Ye, B. C. Anal. Chem. 2013, 85, 11665-11674. (18) Akiba, H.; Sumaoka, J.; Komiyama, M. ChemBiochem 2009, 10, 1773-1776. (19) Fu, P. K. L.; Turro, C. J. Am. Chem. Soc. 1999, 121, 1-2. (20) Blackburn, G. M.; Gait, M. J. Nucleic Acids in Chemistry and Biology, 1990, Oxford University Press, Oxford, UK. (21) a) Lin, W. T. D.; Huang, P. J. J.; Pautler, R.; Liu, J. Chem. Commun. 2014, 50, 11859-11862; b) Li, W.; Zhang, Z.; Zhou, W.; Liu, J. ACS Sensors, 2017, 2, 663-669. (22) Liu, J.; Cao, Z.; Lu, Y. Chem. Rev. 2009, 109, 1948-1998. (23) Zhou, W.; Saran, R.; Liu, J. Chem. Rev. 2017, 117, 8272-8325. (24) Gao, Y. G.; Sriram, M.; Wang, A. H. J. Nucleic Acids Res. 1993, 4093-4101. (25) Anastassopoulou, J. J. Mol. Struct. 2003, 651–653, 19– 26. (26) Jurs, P. C.; Bakken, G. A.; McClelland, H. E. Chem. Rev. 2000, 100, 2649-2678. (27) Szmacinski H.; Lakowicz J. R. Topics in Fluorescence Spectroscopy, 2002, vol 4. Springer, Boston, MA. (28) Lu, Y.; Zhao, J.; Zhang, R.; Liu, Y.; Liu, D.; Goldys, E. M.; Yang, X.; Xi, P.; Sunna, A.; Lu, J.; Shi, Y.; Leif, R. C.; Huo, Y.; Shen, J.; Piper, J. A.; Robinson, J. P.; D. Jin. Nature Photon. 2014, 8, 32-36. (29) Mähler, J.; Persson, I. Inorg. Chem. 2012, 51, 425-438. (30) Meuwis, K.; Boens, N.; De Schryver, F. C.; Gallay, J.; Vincent, M. Biophys. J. 1995, 68, 2469-2473. (31) Mohanty, J.; Barooah, N.; Dhamodharan, V.; Harikrishna, S.; Pradeepkumar, P. I.; Bhasikuttan, A. C. J. Am. Chem. Soc. 2013, 135, 367-376. (32) Xu, L.; Zhou, L.; Chen, X.; Sheng, X.; Wang, J.; Zhang, J.; Pei, R. Spectrochim. Acta A 2017, 180, 85-90. (33) Huang, P. J. J.; Vazin, M.; Lin, J. J.; Pautler, R.; Liu, J. ACS Sens. 2016, 1, 732-738.

ACS Paragon Plus Environment

Page 7 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry for TOC only:

ACS Paragon Plus Environment