Introduction

Magnetic beads have been demonstrated intensively as an ideal protein host since they have highly chemical stability and relatively environmental inertness [1, 2]. Nanometer-sized Fe3O4 particles have high-surface hydroxyl groups which are favorable for a wide variety of surface reactions and the binding of biomolecules including proteins, DNA and peptides [3, 4]. The bionanocomposites are readily separated from the mixtures with an aid of external magnetic field [5]. Thanks to these advantages, research has looked to explore innovative and powerful novel biofunctionalized nanometer-sized magnetic beads, controlling and tailoring their properties in a very predictable manner to meet the needs of specific applications because of high surface-to-volume ratio relative to bulk materials [6, 7].

Magneto-controlled biosensors and bioelectronics examine the effects of external magnetic fields on the electrical signals (e.g., current, voltage, resistance or capacitance) of functionalized magnetic beads associated with the electrodes [8]. Typically, there are two basic concerns on the successful development of high-efficiency magnetic biosensing systems. Undoubtedly, the first key point relies on the signal-transduction method and device. A pH meter, a device used for potentiometric pH measurements toward the hydrogen-ion activity in water-based solutions, works by measuring the potential difference with a pH (proton)-sensitive electrode (usually glass) and a reference electrode (usually silver chloride or calomel) [9]. Ideally, the electrode potential, E (mV), for the proton depends on the logarithm of hydronium ion in the activity followed by the Nernst equation [10]. Digital or handheld pH meters have been developed for the bioanalytical applications because of its simple operation, portability and easy quantification [11, 12]. Liang et al. designed pH meter-based electrochemical immunoassay for the detection of neuron-specific enolase using glucose oxidase-loaded liposome [13]. Wang et al. developed portable, self-powered, and light-addressable photoelectrochemical sensing platforms using pH meter readouts for high-throughput screening of thrombin inhibitor drugs [14]. Jakhar and Pundir prepared urease nanoparticles for the construction of an improved potentiometric pH urea biosensor [15]. Thus, our motivation of this study is to design a portable and magneto-controlled pH meter-based detection method for the analytical applications.

Another additional issue for the development of pH meter-based method depends on the signal amplification during the signal-generation process. Enzymes, e.g., glucose oxidase (GOX) and alkaline phosphate, are usually used for this purpose via enzymatic conversion of certain substrates or chemical decomposition of its soluble salts. However, the detectable signal is always limited thanks to the limitation of magnetic beads in surface coverage. Inspiringly, the emergence of DNA-based signal amplification strategies opens a new horizon by coupling with enzyme labels, e.g., hybridization chain reaction and rolling circle amplification (RCA) [16,17,18]. As a simplified version of natural rolling circle replication, RCA describes a process of unidirectional nucleic acid replication that can rapidly synthesize multiple copies of circular molecules of DNA or RNA [19]. Different from conventional DNA amplification techniques (e.g., polymerase chain reaction), RCA is an isothermal nucleic acid amplification technique where the polymerase continuously adds single nucleotide to a primer annealed to a circular template which results in a long concatamer single-stranded DNA (ssDNA) that contains tens to hundreds of tandem repeats (complementary to the circular template) [20, 21]. In the case, the formed repeats can hybridize with another ssDNA (similar to the circular template). When enzyme molecule is labeled to the ssDNA, the formed repeats hybridize to generate a long enzyme concatamer, thus causing the enzymatic conversion of numerous substrates [22]. Relative to magnetic bead alone, the conjugated enzyme molecules largely increase, thereby resulting in the signal amplification. To this end, RCA is expected to enhance the signal of magneto-controlled pH meter-based assay.

This report designs a sensitive and portable pH meter-based magnetic detection method for lead ion (Pb2+) by coupling with rolling circle amplification on magnetic beads (Scheme 1). This system involves in Pb2+-specific reaction with DNAzyme-coated magnetic bead, primer-triggered RCA reaction, hybridization of glucose oxidase-labeled ssDNA (GOx-ssDNA) and oxidization of GOx toward the substrate (glucose). Upon cleavage of Pb2+-specific DNAzyme, the formed GOx tandem repeats oxidizes glucose into gluconic acid and hydrogen peroxide. The produced gluconic acid changes the pH value of detection solution which is detected on a portable pH meter. The pH shift of the detection solution relative to background signal depends on the level of target Pb2+ in the sample. This study aims to explore a portable and sensitive magneto-controlled pH meter-based detection system for heavy metal ions in the environmental water samples.

Scheme 1
scheme 1

Schematic illustration of metal-ion-induced DNAzyme on magnetic bead (MB) for the detection of lead ion (Pb2+) with rolling circle amplification (RCA) on a handheld pH meter: a preparation process of Pb2+-specific DNAzyme on the magnetic bead, and b pH meter-based detection process by coupling with the RCA reaction and hybridization of glucose oxidase-labeled single-stranded DNA probe (GOx-ssDNA)

Full size image

Experimental

Reagent and chemical

Aqueous dispersion of magnetic nanoparticles with hydrodynamic diameter of 100 nm (25 mg mL−1; Number of particles: ~1.8 × 1015/g; Density: ~5.2 g cm−3; Storage: ddH2O) was purchased from Chemicell GmbH (Eresburgstrasse 22–23, Berlin, Germany www.chemicell.com). Bovine serum albumin (BSA; Vetec™ reagent grade, lyophilized powder, ≥98%), (3-glycidyloxypropyl) trimethoxysilane (GOPS; ≥98%), glucose oxidase (GOx; Type X-S; lyophilized powder; 100,000–250,000 units g−1 solid without added oxygen), D-glucose, sulfosuccinimidyl-4-(N-maleinidomethyl)-cyclohexane-1-carboxylat (sulfo-SMCC) and tris(2-carboxyethl)phosphine (TCEP) were gotten from Sigma-Aldrich (www.sigmaaldrich.com). Deoxyribonucleoside 5′-triphosphates mixture (dNTP), T4 DNA ligase (Activity: 5000 units mL−1), adenosine 5′-triphosphate (ATP) and phi29 DNA polymerase (Activity: 10,000 units mL−1) were purchased from Takara Biotechnol. Co., Ltd. (Dalian, China www.takara.com.cn). All other reagents were of analytical grade and used without further purification. Ultrapure water obtained from a Millipore water purification system was used in all runs (18.2 MΩ cm, Milli-Q, Millipore www.millipore.com). All buffers including phosphate-buffered saline (PBS) solutions with various pH values were the products of Sigma-Aldrich (www.sigmaaldrich.com).

All used oligonucleotides were obtained from Sangon Biotech. Co., Ltd. (Shanghai, China www.sangon.com), and their sequences were listed as follows:

  • Catalytic strand: 5’-NH2-TTT CAT CTC TTC TCC GAG CCG GTC GAA ATA GTG AGT-3’

  • Substrate strand: 5’-ACT CAC TAT rA GGA AGA GAT G-3’

  • Circular template: 5′-p-TTC GAC CGG AAC TGT CTT AGC AAA AAC TGT CTT AGC AAA CTC ACT AT-3′

  • Detection probe: 5’-SH-ACT CAC TAT TTC GAC CGG-3’

Pb2+-specific DNAzyme was designed according to our previous report [23]. The bold sequence at the catalytic strand was the primer for the RCA reaction. The italicized sequence at the circular template (p = 5′ phosphate) matched the bold bases of the catalytic strand. The detection DNA probe was the same as those in the italic on the circular template. GOx-labeled single-stranded DNA (i.e., detection probe) was prepared according to this literature [24], and described in the Electronic Supporting Material.

Preparation of Pb2+-specific DNAzyme-conjugated magnetic bead (DNAzyme-MB)

Before synthesis of DNAzyme-MB, the aminated DNA catalytic strands were covalently conjugated onto magnetic beads through the epoxy-amino reaction [25]. In detail, magnetic beads obtained by magnetic separation were initially dried in the vacuum at 80 °C for 60 min. Thereafter, magnetic beads (250 mg) were added to the GOPS dry toluene (5.0 mL, 5%, v/v) and stirred for 12 h (500 rpm, RT). Following that, the suspension was magnetically separated, and washed three times with toluene and ethanol solution, respectively. The GOPS-functionalized magnetic beads (GOPS-MBs) were dried and activated in an oven at 80 °C for 60 min under a nitrogen atmosphere. Afterword, the aminated DNA catalytic strands (100 μL, 1.0 mM) was injected into the GOPS-MB suspension at a concentration of 25 mg mL−1. The mixture was slightly stirred for 12 h at 4 °C for conjugation of amino groups on catalytic strands with epoxy group on the GOPS. Catalytic strand-conjugated MBs by magnetic separation were injected to substrate strand (200 μL, 1.0 mM) in ultrapure water, and incubated for 6 h at 37 °C to make the catalytic strand hybridize with the substrate strand for formation of Pb2+-specific DNAzyme. Finally, the magnetically separated DNAzyme-MBs were dispersed in PBS (10 mM, pH 7.4, 0.1 wt% sodium azide), and stored at 4 °C at a final level of 10 mg mL−1. The conjugation process of DNAzyme-MB is schematically illustrated in Scheme S1 of the Electronic Supporting Material.

Target-induced rolling circle amplification (RCA) and pH meter-based measurement

Target-induced RCA reaction on DNAzyme-MB was initially carried out as follows: (i) 50 μL of Pb2+ standard or sample was injected into DNAzyme-MB suspension (10 ng mL−1) in a 1.5-mL centrifugal tube, and reacted for 25 min at RT with slight shaking; (ii) 5.0 μL of circular template DNA (100 nM), 5.0 μL of T4 DNA ligase (4.0 U μL−1), 90 μL of phi29 DNA polymerase (2 unit mL−1) reaction buffer (50 mM, pH 7.5 Tris-HCl buffer, 10 mM magnesium acetate, 33 mM potassium acetate, 1.0 mM dithiothreitol, 10 mM dNTPs, and 0.1% Tween 20) and 1.0-mM ATP were added into the resulting magnetic particles, and incubated for 90 min at RT to conduct the padlock ligation reaction and the RCA reaction; (iii) 100 μL of the GOx-ssDNA was added to the centrifugal tube, and reacted for 120 min at RT; (iv) 200 μL of 2.0 mM (excess) glucose in PBS (10 mM, pH 7.4) was injected into the centrifugal tube and reacted for 10 min at RT; and (v) the pH of the resulting supernatant was determined on a handheld pH meter (0.00–14.00 pH; ± 0.1 pH, 20 °C; ATC® Aolilong Instr. Inc., Hangzhou, China). Note that the resulting suspension after each step was magnetically separated and washed with PBS (10 mM, pH 7.4). The pH value was collected and registered as the signal relative to target Pb2+ concentration. The signals in all experiments referred to average responses of reaction with the corresponding standard deviations in triplicate, unless otherwise indicated. All measurements were performed at room temperature (25 ± 1.0 °C).

Result and discussion

Choice of materials

Scheme 1 gives the schematic illustration of target-induced cleavage of Pb2+-specific DNAzyme on magnetic beads for the sensitive detection of metal ion by coupling rolling circle amplification with pH meter-based readout. In this system, magnetic beads are not only used as the affinity supports of the conjugation of Pb2+-specific DNAzyme, but are favorable for rapid separation and purification after each step. Importantly, this methodology pulls the biomolecules bound to magnetic beads from one laminar flow path to another applying a local magnetic field gradient and selectively removes them from the complex system. Aminated Pb2+-specific DNAzyme is conjugated to magnetic beads through the classical epoxy-amino reaction. Accompanying target-induced cleavage of Pb2+-specific DNAzyme on magnetic beads, the newly produced single-stranded DNA is used as the template for the RCA reaction in the presence of dNTP, thus resulting in a long concatamer ssDNA that contains many tandem repeats (complementary to the circular template). The concatamer repeats are utilized for the reaction with GOx-ssDNA conjugates via the principle of complementary base pairing. In this case, numerous GOx molecules are concatenated together by the RCA product, each of which oxidizes glucose into gluconic acid and hydrogen peroxide. Generation of gluconic acid changes the pH of the detection solution. By monitoring the pH change, the concentration of target Pb2+ is exactly evaluated in the sample.

Characterization of DNAzyme-MB and RCA reaction with GOx-ssDNA

For development of magneto-controlled detection platform, preparation of DNAzyme-MB is very crucial to induce target-triggered cleavage of DNAzyme. Figure 1a shows typical scanning electron microscope (SEM; Hitachi S4800, Japan www.hitachi.com) image of the DNAzyme-MB, and the average size was 110 nm in diameter (Fig. 1a, top inset). Most nanoparticles had the similar surface topological structures. The surface became rougher after magnetic beads were finally decorated with many concatenated GOx molecules through the RCA reaction (Fig. 1a, bottom inset). Since the conjugated DNAzyme molecules on the magnetic bead were easily carbonized at high applied voltage, they were difficultly observed in SEM images. To demonstrate this issue, dynamic light scattering (DLS; Zetasizer, Nano S90, Malvern, London, UK www.malvern.com) was employed to investigate the mean sizes and surface potentials of magnetic beads before and after modification with different components including GOPS, catalytic strand and RCA product, respectively (note: Substrate strand did not introduce in this case because the short single-stranded DNA did not almost change obviously the size and surface potential) (Fig. 1b). Clearly, the average sizes of magnetic beads gradually increased after modification with GOPS, catalytic strand, RCA product and GOx-ssDNA in sequence. The increasing size after the RCA reaction was relatively biggest than those of other conjugation components (column ‘d’ versus columns ‘b-c’), which derived from a long concatamer single-stranded DNA that contains many tandem repeats. Thanks to the presence of surface –O− groups, the unmodified magnetic beads exhibited a negative zeta potential (−12.2 mV, column ‘a’). Upon the epoxy group introduction, partial –O− groups were covalently conjugated with the GOPS to form epoxy group-functionalized surface, and the zeta potential decreased to −2.1 mV (column ‘b’). After the aminated catalytic strands reacted with the GOPS via the epoxy-amino reaction, however, the corresponding zeta potential increased to −13.1 mV (column ‘c’), which originated from the negatively charged oligonucleotides. Significantly, the zeta potential of the functional magnetic beads because more negative (−33.4 mV, column ‘d’) after RCA reaction, indicating formation of long-oligonucleotide strands. Since the isoelectrical point (Ip) of GOx is ~4.9, it is negatively charged in PBS (10 mM, pH 7.4). Therefore, the zeta potential of magnetic beads MBs after the final decoration with many concatenated GOx molecules was −78.3 mV (column ‘e’). These results preliminarily indicated that the aminated catalytic strands were conjugated to magnetic beads.

Fig. 1
figure 1

a SEM image of DNAzyme-MB (top inset: magnification image of DNAzyme-MB; bottom inset: SEM image of MBs after the final decoration with many concatenated GOx molecules); b DLS data and zeta potentials of a magnetic bead, b GOPS-MB, c catalytic strand-conjugated MB, d catalytic strand-conjugated MB after RCA reaction and e MBs after the final decoration with many concatenated GOx molecules; C UV-vis absorption spectra of a magnetic beads, b GOPS-MB and c catalytic strand-conjugated MB; and D agarose gel electrophoresis images for a 1.0-μM catalytic strand, b 1.0-μM substrate strand, c 1.0-μM catalytic strand +1.0-μM substrate strand, d DNAzyme +100-nM Pb2+, e DNAzyme +100-nM Pb2+ + RCA reaction and f DNAzyme +100-nM Pb2+ + RCA reaction + GOx-ssDNA

Full size image

Logically, one puzzling question arises as to whether Pb2+-specific DNAzyme was modified onto the magnetic beads. To clarify this concern, UV-vis absorption spectroscopy was used to monitor magnetic beads before and after formation of DNAzyme-MB (Fig. 1c). Curve ‘a’ represents UV-vis absorption spectrum of the unmodified magnetic beads, and almost no characteristic peaks were observed. Also, GOPS-modified magnetic beads gave the similar UV-vis absorption spectrum with magnetic beads alone (curve ‘b’). Favorably, an obvious characteristic absorption peak for the oligonucleotides was acquired at 260 nm (curve ‘c’), suggesting that the aminated catalytic strands were conjugated to magnetic beads. To further verify that Pb2+-specific DNAzyme was formed by the catalytic strand, gel electrophoresis was also utilized to characterize these processes including target-triggered HCR reaction (note: Because the big-sized magnetic beads were not migrated in gel electrophoresis, these experiments were carried out in the absence of magnetic beads) (Fig. 1d). Lanes ‘a’ and ‘b’ represent the gel electrophoresis images of catalytic strands and substrate strands, respectively. As seen from lane ‘c’, the DNA duplex was formed after mixture of catalytic strands with substrate strands. Inspiringly, the formed DNA duplex was cleaved in the presence of target Pb2+ (lane ‘d’), because the base number of the spot was almost the same as lane ‘a’, indicating the formation of Pb2+-specific DNAzyme. Lane ‘e’ shows the products of the cleaved DNAzyme by Pb2+ after the RCA reaction. Actually, the fuzzy spot lanes stemmed from the RCA products with different lengths of oligonucleotide strands. Note that the similar results were obtained when RCA products reacted with GOx-ssDNA (lane ‘f’). These results adequately revealed that target-induced DNAzyme on magnetic beads with the RCA reaction was used for the detection of Pb2+ through the GOx-ssDNA.

Control tests and analysis of feasibility

As mentioned above, the signal readout derived from the backfilling GOx-ssDNA after the RCA reaction. As control tests, two crucial points should be investigated prior to experiments: (i) whether target-induced RCA reaction was readily progressed step by step, and (ii) whether GOx-ssDNA was nonspecifically conjugated to DNAzyme-MB. Figure 2a displays the pH responses of this system after each step in the presence of glucose (200 μL, 2.0 mM) in PBS (10 mM, pH 7.4). Expectedly, almost no pH changes (versus background signal) were observed when DNAzyme-MB (column ‘b’) reacted with 5.0 nM Pb2+ (column ‘c’) and RCA reaction (column ‘d’) in turn, indicating that the conjugated DNAzyme and RCA products on magnetic beads were not change the pH value of the detection solution. When the RCA products formed on magnetic beads incubated with GOx-ssDNA, significantly, the pH value of the detection solution (column ‘d’) was obviously lower than that of background signal (column ‘a’), suggesting that the pH readout originated from the introduction of GOx-ssDNA. Naturally, one question to be produced was whether the pH change stemmed from the nonspecific adsorption of GOx-ssDNA in the absence of target analyte. As indicated from column ‘e’, the pH value of the detection solution was almost the same as the background signal. Therefore, the pH change was from target-triggered HCR reaction with GOx-ssDNA. Since the catalytic strand reacted with GOx-ssDNA, the hybridized products were also investigated by using the same assay mode. The pH value of the detection solution (column ‘f’) slightly decreased relative to background signal (column ‘a’). Compared with column ‘e’, pH response of this system without RCA reaction was largely lower than that with RCA reaction, revealing that RCA reaction amplified the detectable signal of this system. On the basis of these results, a conclusion was made that our strategy had the ability to determine target Pb2+ in the sample.

Fig. 2
figure 2

a The pH responses of different components in the presence of glucose (200 μL, 2.0 mM) in PBS (10 mM, pH 7.4): a DNAzyme-MB, b DNAzyme-MB + 5.0 nM Pb2+, c DNAzyme-MB + 5.0 nM Pb2+ + RCA reaction, d DNAzyme-MB + 5.0 nM Pb2+ + RCA reaction + GOx-ssDNA, e DNAzyme-MB + RCA reaction + GOx-ssDNA and f DNAzyme-MB + 5.0 nM Pb2+ + GOx-ssDNA; b pH responses (versus background signal) of RCA-based detection platform toward different-concentration Pb2+ standards; C the specificity of RCA-based Pb2+ detection platform; and D comparison of the results obtained from RCA-based Pb2+ detection method and ICP-MS for spiked water samples

Full size image

Analytical performance

Before investigating the analytical properties of RCA-based Pb2+ assay system with pH meter, the following parameters were optimized: (a) cleavage time for Pb2+-specific DNAzyme on magnetic beads; (b) RCA reaction time of primer; (c) hybridization time of RCA products for GOx-ssDNA; Respective data and figures are given in the Electronic Supporting Material (Fig. S1). The following experimental conditions were found to give best results: (a) cleavage time: 25 min; (b) RCA reaction time: 90 min; (c) hybridization time: 120 min. Under optimum conditions, the analytical performance of RCA-based pH assay platform was studied toward Pb2+ standards with different concentrations by coupling with rolling circle amplification and pH meter-based readout device. Figure 2b shows the pH changes of this system (relative to background signal) in the presence of target analyte. The signal increased with the increasing Pb2+ level, and a good linear relationship between the pH shift and the decimal logarithm of Pb2+ concentration was acquired within the dynamic range of 1.0–100 nM. The limit of detection (LOD) was evaluated to 0.91 nM on the basis of 3S/K (where K and S stand for the slope of the calibration plot and the standard deviation of blank solution, respectively). The regression equation was y = 0.07 + 1.59 × logC (nM, R2 = 0.9925, n = 15). Each data point represented the average value obtained from three different measurements. The maximum relative standard deviation (RSD) was 10.9%, suggesting a good reproducibility of this system. To further elucidate the signal-amplified advantage of our strategy by using RCA reaction, the analytical properties of this system including the linear range and LOD were compared with other detection schemes. As analyzed from Table 1, the LOD of our designed protocol were comparable with other analytical methods, although the linear range was relatively narrow. The reason might be ascribed to the limitation of the labeled amount of magnetic beads toward Pb2+-specific DNAzyme molecules.

Table 1 An overview on the reported nanomaterial-based methods for determination of lead ion
Full size table

Next, the specificity of our designed strategy was evaluated by challenging it against other metal ions including Hg2+, Ca2+, Mn2+, Co2+, Cu2+, Ag+, Ni2+, Zn2+, Fe3+, Cd2+ and Mg2+. Significantly, these non-targets did not almost cause the significant pH change of the detection solution (Fig. 2c). An obvious pH shift relative to background signal was achieved toward target Pb2+ ion. These results clearly revealed the high specificity of RCA-based Pb2+ sensing system. Further, the storage stability of DNAzyme-MB and GOx-ssDNA was measured during a six-month storage period. They were stored at 4 °C when not in use. By using 5.0 nM Pb2+ as an example, the pH signals preserved >90% of the initial signal after storage them for 5 months, suggesting a good stability.

Monitoring of environmental water samples

To further monitor the trueness and applicability of RCA-based Pb2+ detection system for testing environmental water samples, Pb2+ standards with various concentrations were initially spiked into drinking water. Following that, these water samples were determined by using our method and inductively coupled plasma mass spectrometry (ICP-MS; used as a reference method), respectively. The results obtained from two methods are summarized in Fig. 2d. Comparison of the experimental results obtained with RCA-based Pb2+ detection system with those of ICP-MS was performed via the use of a least-squares regression method. The regression line was fitted to y = (0.9944 ± 0.0266) x – (0.2967 ± 1.3961) (R2 = 0.9954, n = 45), where x stands for Pb2+ concentrations estimated with RCA-based Pb2+ detection system and y stands for those of the ICP-MS. The correlation between two methods was investigated by using t-test for comparison of experimental values of the intercept and slope to the ideal situation of zero intercept and slope of 1. As shown in Fig. 2d, the slope and intercept were close to the ideal values ‘1’ and ‘0’, respectively, thus revealing a good agreement between both analytical methods for the analysis of real water samples.

Conclusions

This contribution describes a portable potentiometric Pb2+ detection platform using rolling circle amplification on a handheld pH meter. The Pb2+ nanoprobes were prepared on the basis of Pb2+-specific DNAzyme conjugated onto the magnetic beads. Introduction of molecular biological technique (RCA reaction) and enzyme labels were utilized to amplify the detectable signal of this system. Nevertheless, one disadvantage of this method lies in the long reaction time owing to Pb2+-specific cleavage of DNAzyme and DNA-based RCA/hybridization reaction. Therefore, future work should be focusing on improvement of the whole assay time, e.g., by controlling the reaction temperature of DNA reaction.