A novel synergistic enhanced chemiluminescence achieved by a multiplex nanoprobe for biological applications combined with dual-amplification of magnetic nanoparticles

Abstract

A multiplex nanoprobe (CuS/DNA/Au/DNA/MNP) achieving a novel synergistic enhanced chemiluminescence (SECL) system is applied in a signal amplified bioassay strategy based on magnetic nanoparticles (MNPs) acting as both DNA molecular carriers and reporter labels, and further amplifying the signals by dual-amplification (DNA amplification and metal ions amplification). In this study, a luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL system is proposed for the first time, which is achieved by a multicomponent nanoparticle probe and significantly intensifies the CL signal compared to monometallic Cu²⁺ or Fe³⁺ catalyst. The combination of the remarkable sensitivity of the SECL method with the amplification of MNPs achieves the detection of specific DNA sequences and Ramos cells as low as 6.8 aM and 56 cells mL⁻¹, respectively, which is one of the most sensitive approaches for bioassays. In addition, the strategy is able to differentiate between Ramos target cells and CEM control cells based on the high specificity of the aptamer for Ramos cells, indicating the wide applicability of it for diseased cell detection.

---

Introduction

The emergence of nanotechnology is opening a new horizon for biological sensing, among which metal nanoparticles with distinctive photonic, magnetic and electronic properties offer an excellent prospect for highly sensitive assays of biological molecules.^1–3 So far, two main approaches have been developed for nanoparticle-based bioassays. (1) Nanoparticles are used as carriers to load a large amount of species, e.g., enzyme molecular, DNA, and luminescent labels.^4–6 As one of the most widely used nanomaterials in bioanalysis, magnetic nanoparticles (MNPs) are usually employed as special biomolecules immobilizing carriers because of their easy magnetic separation and large surface.^7–10 (2) Metal nanoparticles are directly used as reporters. One strategy is based on the aggregation of oligonucleotide-functionalized gold nanoparticles by colorimetric detection for DNA hybridization, metal ions or proteins.^11–13 Although this technique is easy to readout and rapid for detection, the low sensitivity limits its application. Another strategy of metal nanoparticle-based bioassay is to detect the metal ions released from the metal nanoparticle probes anchored on the reporters by oxidative metal dissolution.^14,15

As one of the most sensitive techniques for trace analysis of metal ions, chemiluminescence (CL) has become an attractive method for DNA and protein biosensing owing to its high sensitivity, inherent simplicity, and low cost.^16–18 In previous reports, the CL system was only studied by catalyzing with individual metal ions, such as Au³⁺,¹⁷ Fe³⁺,¹⁸ or Cu²⁺.¹⁹ However, so far there has been no report on two metal ions acting simultaneously as catalysts for the CL reaction system. Inspired by the above two application fields of metal nanoparticles in biological analysis, we hypothesized that the MNPs-based protocol could produce a universal, selective and sensitive method for the collection and subsequent detection of various target molecules by combining the easy separation of MNPs as carriers, high immobilization as labels, and sensitive detection of metal ions released from MNPs by CL detection.

Herein, we make the best use of MNPs for biological applications based on a novel luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL system by flow-injection chemiluminescence (FI-CL) detection, which is achieved from a CuS/DNA/Au/DNA/MNP multiplex nanoprobe and further accomplishes high sensitive and selective detection of specific DNA sequences. Meanwhile, MNPs are simultaneously acting as a dual-amplification intermediary for DNA amplification and metal ions amplification, allowing a detection limit of 6.8 aM for target DNA in DNA hybridization detection which is comparable with other most ultrasensitive DNA detection.^20,21 Furthermore, based on the structure–switching of aptamers specific for Ramos cells selected from cell-SELEX upon target binding,^22,23 the aptamer–ligand recognition event can be easily amplified following the nucleic acid sequence-based amplification protocols, which makes an extraordinarily low detection limit for Ramos cells possible. The remarkable sensitivity of the amplification strategy for the detection of Ramos cells can be as low as 56 cells mL⁻¹, which is much more sensitive than other optical cytosensors.^23,24 Given the excellent analytical properties and design flexibility, the developed strategy exhibits a promising analysis platform for diverse analytes assay. The preparations and analytical characteristics of the proposed DNA and whole-cell amplified assay schemes are detailed in this work.

Results and discussion

Principles of the proposed strategies for DNA and whole-cell amplified assay

In this study, MNPs were simultaneously acting as aptamer carriers and reporter labels, and further amplified the signals by employing MNPs as the dual-amplification intermediary based on the novel luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL system achieved with a CuS/DNA/Au/DNA/MNP multiplex nanoprobe. Scheme 1A depicts the principle of dual-amplification strategy for target DNA (t-DNA) detection in DNA hybridization. Briefly, carboxyl modified MNPs (carboxyl-MNPs, ∼500 nm) were firstly functionalized with 5′-NH₂ labeled bio-bar-code DNA (bbc-DNA) and first-probe DNA (p1-DNA), fabricating the bbc-p1-DNA/MNPs to avoid cross-linking reaction. Then, the streptavidin-coated substrate was modified with biotinylated capture DNA (biotin-c-DNA) (step I, Scheme 1A), followed by hybridizing with t-DNA on the substrate (step II, Scheme 1A). Subsequently, the bbc-p1-DNA/MNPs were conjugated onto the substrate by utilizing a hybridization reaction between p1-DNA and t-DNA (step III, Scheme 1A). In this case, only one MNP was bound to one t-DNA due to the effective avoidance of cross-reaction by employing bbc-p1-DNA/MNPs as labels, which was verified by hybridizing the CdTe quantum dots (CdTe QDs)-tagged probe DNA to bbc-p1-DNA/MNPs on the substrate. The centric distance between fluorescent spots indicated that one t-DNA could bring one MNP onto the substrate with a concentration of t-DNA of 1.0 × 10⁻¹⁴ M. Then, the bbc-p1-DNA/MNPs were released from the substrate via urea dehybridizaiton (step IV, Scheme 1A) and hybridized with the previously fabricated CuS NPs and Au NPs modified second-probe DNA (CuS/p2-DNA/Au) (step V, Scheme 1A). From Fig. 1B, the disappearance of fluorescence spots on the substrate indicated that the bbc-p1-DNA/MNPs were released from the substrate completely by a 50% (w/w) urea solution. It should be noted that in order to amplify the signals as large as possible, herein two kinds of second-probe DNA (p2-DNA), which were complementary to bbc-DNA and p1-DNA on the released MNPs surface, respectively, were employed. The nonspecifically bound nanoparticle probes could be automatically removed via magnetically controlled washing. Subsequently, once CuS NPs were dissolved from the hybrids by a certain concentration of nitric acid (0.1 M), a large amount of released cupric ions (Cu²⁺) could be sensitively determined by the luminol–H₂O₂ CL reaction system and generated a CL signal (step VI, Scheme 1A). The CL intensity was proportional to the concentration of t-DNA based on the amounts of dissolved Cu²⁺. Moreover, in order to amplify the CL signals at utmost, we dissolved ferric ions (Fe³⁺) of MNPs and Cu²⁺ of CuS NPs simultaneously via a higher concentration of nitric acid (0.4 M) (step VII, Scheme 1A) and surprisingly found the luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL reaction system, whose sensitivity was approximately one order of magnitude higher than that just employing monometallic Cu²⁺ as catalyst, achieving a low detection limit of target DNA at 6.8 aM.

Scheme 1 Schematic representation of SECL detection of DNA hybridization and Ramos target cells based on a dual-amplification strategy.

Fig. 1 Images of the microtiter plate with CdTe QDs/bbc-p1-DNA/MNPs hybridized with 1.0 × 10⁻¹⁴ M t-DNA (A) before and (B) after dehybridization using 50% (w/w) urea solution.

After investigating the mechanism of SECL system extensively (vide infra), a strategy for cancer cells detection was further developed based on the above proposed DNA-based amplification protocol. Scheme 1B displays the schematic diagram for Ramos cells detection. In this assay, amino-modified aptamers were attached on the surface of Fe₃O₄–Au core–shell magnetic nanoparticles (Fe₃O₄@Au) due to its large surface, and t-DNA, which was the same as that used in DNA hybridization detection with 12-mer complementary to the aptamer, was added to fabricate the t-aptamer-DNA/Fe₃O₄@Au complexes. The introduction of the Ramos target cells triggered the aptamers functionalized Fe₃O₄@Au assembling onto the cell surface through the conformational change of aptamers from a DNA/DNA duplex to a DNA/target complex. In this case, Fe₃O₄@Au-tagged target cells could be easily separated with a small magnet and again dispersed immediately after the magnet was removed. The released t-DNA was then detected according to the above amplified protocol for DNA hybridization detection by dissolving Cu²⁺ and Fe³⁺ simultaneously, whose signal was proportional to the concentration of Ramos cells.

Characterization of the DNA-NP conjugates

The synthesized Au NPs, CuS NPs, p2-DNA tagged with Au NPs and CuS NPs (CuS/p2-DNA/Au), and MNPs were characterized by TEM. From Fig. 2A–D, it can be seen clearly that in the absence of p2-DNA, only individual NPs were observed (A for Au NPs and B for CuS NPs, respectively), while in the presence of p2-DNA, aggregates of the Au NPs and CuS NPs were obtained (C). Fig. 2D shows the TEM image of the MNPs with an average diameter of 500 nm.

Fig. 2 The TEM images of (A) AuNPs (∼20 nm), (B) CuS NPs (∼5 nm), (C) p2-DNA tagged with Au NPs and CuS NPs (CuS/p2-DNA/Au), and (D) MNPs (∼500 nm). (E) UV-visible spectra of (a) p2-DNA, (b) CuS NPs, (c) Au NPs, and (d) p2-DNA difunctionalized with CuS and Au NPs on the 3′- and 5′-ends.

The UV-visible spectra of the CuS NPs, Au NPs, unmodified p2-DNA, and difunctionalized p2-DNA with CuS and Au NPs were recorded by the spectrophotometer as shown in Fig. 2E. Curve d exhibits both the characteristic absorbance of DNA (curve a) at 260 nm and Au NPs (curve c) at 520 nm. The results indicated that the CuS and Au NPs had been successfully labeled on the 3′- and 5′-end of p2-DNA, respectively.

DNA hybridization detection

In order to evaluate the feasibility of the demonstrated assay, the sensitivity of the DNA biosensor was first investigated. At step VIScheme 1A, once CuS NPs were dissolved from the hybrids, a large amount of released Cu²⁺ could be sensitively determined by the luminol–H₂O₂ CL reaction system, achieving a detection limit of 72 aM for t-DNA, which was comparable to that of 86 aM detected at step IVScheme 1A by the lunimol–H₂O₂–Fe³⁺ system. Here, t-DNA was 36-mer synthesized DNA sequence (sequence 3 in Table S1†).

As an expansion study, we further dissolved CuS NPs and MNPs simultaneously after dehybridizing by urea solution and hybridizing with CuS/p2-DNA/Au via certain concentrations of nitric acid and detection by the luminol–H₂O₂–Cu²⁺–Fe³⁺ system with a desire to achieve a higher sensitivity. Surprisingly, in this study we found that Cu²⁺ and Fe³⁺, both luminol–H₂O₂ CL system catalysts, could significantly intensify the CL signals compared to monometallic Cu²⁺ or Fe³⁺ catalyst. The results showed that the CL intensity of the luminol–H₂O₂–Cu²⁺–Fe³⁺ CL system increased with increasing concentrations of t-DNA ranging from 1.0 × 10⁻¹⁷ to 1.0 × 10⁻¹⁵ M with a detection limit of 6.8 aM by using 3σ (σ = S_b/m; S_b, standard deviation of blank sample, S_b = 3.4801 (n = 11) in this experiment; m = 15.3532, the slope of the calibration curve at low concentration, 1.0 × 10⁻¹⁷ to 1.0 × 10⁻¹⁶ M in this experiment). A relative standard deviation (RSD) of 4.2% for eleven successive measurements of 4.0 × 10⁻¹⁷ M target DNA exhibited a good reproducibility of the assay.²¹ Consequently, the sensitivity of the present work was significantly improved based on the dual-amplification of MNPs and further improved by detecting dissolved Cu²⁺ and Fe³⁺ at the same time (Table S2†). The detection limit was comparable with other most ultrasensitive DNA detection.^20,21 The mass concentration ratio of Cu²⁺ and Fe³⁺ dissolved from CuS/p2-DNA/Au/MNPs conjugates was 14/1. At this ratio, a synergistic effect between Cu²⁺ and Fe³⁺ for the luminol–H₂O₂ CL system was obtained, whose detection sensitivity was significantly improved compared to monometallic Cu²⁺ or Fe³⁺ as catalyst. In addition, the proposed amplification strategy was successfully applied in E. coli genomic DNA detection for 120- and 250-bp PCR products (sequences 11 and 12, respectively) by employing a new set of c-DNA (sequence 14), p1-DNA (sequence 15) and p2-DNA (sequence 16). The detailed results are shown in the Supporting Information.†

Investigation of luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL

In order to reason the mechanism of the luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL reaction system, the UV-visible absorption spectra and fluorescence spectra of different systems were investigated and nearly identical spectra were obtained (Fig. S8†). Thus, it could be proposed that no coordination complex was formed in the SECL system, and it was not the coordination complex mechanism. Subsequently, the effects of ·O₂⁻ quencher superoxide dimutase (SOD) and three HO· quenchers, mannitol, methanol and methyl benzoate, were investigated. As shown in Fig. 3A, the CL intensities were significantly quenched by these quenchers. Thus, it is reasonable that the possible mechanism of the luminol–H₂O₂–Cu²⁺–Fe³ SECL system was the “free radical mechanism” (Fig. 3B), in which a synergistic effect occurred between Cu²⁺ and Fe³⁺ in the generation of active oxygen radicals for the luminol–H₂O₂ CL system.

Fig. 3 (A) Effect of quenchers on the mechanism of the luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL system. 1.5 × 10⁻³ M luminol, 8.0 × 10⁻³ M H₂O₂, 1.0 × 10⁻⁸ g mL⁻¹ Fe³⁺ and 1.4 × 10⁻⁷ g mL⁻¹ Cu²⁺ were mixed and reacted with H₂O (black), 100 μg mL⁻¹ SOD (red), 0.1 M methanol (green), 0.01 M mannitol (blue), and 0.1 M methyl benzoate (cyan), respectively. (B) The presumptive mechanism of the luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL system.

In addition, for the general and wide application of the current SECL technique, the applicable ratios between Cu²⁺ and Fe³⁺ were also investigated. Considering the linear ranges of Cu²⁺ and Fe³⁺ in the luminol–H₂O₂ CL system were 1.0 × 10⁻⁹ to 1.0 × 10⁻⁸ g mL⁻¹ and 1.0 × 10⁻¹⁰ to 1.0 × 10⁻⁹ g mL⁻¹, respectively (Fig. S6 in Supporting Information†), in this study we set the concentration of Fe³⁺ as 1.0 × 10⁻¹¹ g mL⁻¹, at which the CL signal could not be detected in the luminol–H₂O₂–Fe³⁺ CL system, and investigated the ratios between Fe³⁺ and Cu²⁺ from 1 : 1 to 1 : 500. Among the ratios from 1 : 1 to 1 : 100, the corresponding concentrations of Cu²⁺ were from 1.0 × 10⁻¹¹ to 1.0 × 10⁻⁹ g mL⁻¹ which could also not be detected in the luminol–H₂O₂–Cu²⁺system. Thus, if a CL signal could be measured in a Cu²⁺ and Fe³⁺ mixture for luminol–H₂O₂ CL system in this case, the novel SECL system could be proved. As shown in Fig. 4, when the ratios were from 1 : 1 and 1 : 50, both CL signals of the luminol–H₂O₂–Cu²⁺ (purple) and luminol–H₂O₂–Fe³⁺ (blue) CL systems could not be detected. Only if Fe³⁺ and Cu²⁺ simultaneously existed in the luminol–H₂O₂ CL system from a ratio of 1 : 5, could the CL signals be detected and SECL began to make sense. Furthermore, when the ratios between Fe³⁺ and Cu²⁺ were 1 : 100 and 1 : 500, the corresponding concentrations of Cu²⁺ were 1.0 × 10⁻⁹ and 5.0 × 10⁻⁹ g mL⁻¹, the SECL still played a significant role in this case, although Cu²⁺ itself could be detected in the luminol–H₂O₂ CL system when the concentration of Cu²⁺ was higher than 1.0 × 10⁻⁹ g mL⁻¹. Consequently, it is well-grounded that the SECL technique has been successfully proposed and is applicable when the ratios between Fe³⁺ and Cu²⁺ are from 1 : 5.

Fig. 4 Investigation of the applicable ratios between Cu²⁺ and Fe³⁺ in the luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL system. (A) Kinetic curves of FI-CL: (a) luminol–H₂O₂–Fe³⁺, (b) luminol–H₂O₂–Cu²⁺, and (c) luminol–H₂O₂–Cu²⁺–Fe³⁺. (B) Corresponding histogram: CL signals of luminol–H₂O₂–Fe³⁺ (blue), luminol–H₂O₂–Cu²⁺ (purple) and luminol–H₂O₂–Cu²⁺–Fe³⁺ (yellow). The concentration of Fe³⁺ is fixed as 1.0 × 10⁻¹¹ g mL⁻¹, and the concentrations of Cu²⁺ are changed corresponding to the ratios. The ratio of 1 : 0 represents that there is only Fe³⁺ and not Cu²⁺ in the luminol–H₂O₂ CL system. The baseline of the FI-CL is ∼50.

Cancer cells detection

Sensitivity. The amplification strategy for cancer cells detection was proposed based on the above mentioned DNA amplified detection protocol in the luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL system using the high specificity and affinity of cell aptamers for the recognition of unique molecular signatures on the surface of cells. Scheme 1B displays the schematic diagram of Ramos cells detection. In this experiment, thiol-modified aptamers for Ramos cells were attached on the surface of Fe₃O₄@Au, and t-DNA, which was the same as that used in DNA hybridization detection (sequence 3) with 12-mer complementary to the aptamer, was added to fabricate the t-DNA/aptamer/Fe₃O₄@Au complexes. The introduction of target cells triggered aptamers functionalized Fe₃O₄@Au assembling onto the cell surface through the conformational change of aptamers from a DNA/DNA duplex to a DNA/target complex. The TEM images of Ramos target cells and CEM control cells mixed with the t-aptamer-Fe₃O₄@Au were taken respectively to verify the nanoparticles assembled on the surface of target cells (Fig. 6). Based on the TEM images, the Fe₃O₄@Au appeared to be assembled on the surface of Ramos cells, while not on CEM cells. This indicated that the aptamers on the Fe₃O₄@Au did cause the assembly of the Fe₃O₄@Au on the target cell surface leading to the release of t-DNA that have been described previously. The amounts of t-DNA/aptamer/Fe₃O₄@Au used (400 μL) was carefully selected to allow saturation of the highest number of cells while using the minimum amounts of nanoparticles to reduce the cost and background of the solution, which facilitated an easier detection of the lower end of the cell concentration (Fig. S9†). The released t-DNA was then detected as the following double amplification processes in the luminol–H₂O₂–Cu²⁺–Fe³⁺ SECL system whose signal was proportional to the concentration of Ramos cells (Fig. 5). The detection limit of 56 cells mL⁻¹ (estimated by 3σ) is much more sensitive than other cytosensors (Table S4†).

Fig. 5 (A) FI-CL signals for Cu²⁺ and Fe³⁺ dissolved simultaneously from a series of CuS/p2-DNA/Au/MNPs conjugates corresponding to different concentrations of Ramos cells: (a) 0; (b) 80; (c) 100; (d) 200; (e) 400; (f) 600; (g) 800; (h) 1000; (i) 2000; (j) 4000; (k) 6000; (l) 8000; (m) 10000 cells mL⁻¹. (B) The calibration curve of peak height versus the concentration of Ramos cells from 80 to 10000 cells mL⁻¹. Inset is the amplification of the linear range from 80 to 1000 cells mL⁻¹ for Ramos cells determination.

Fig. 6 CL signals of (a) t-DNA/aptamer/Fe₃O₄@Au incubating without cells (blank), (b) t-DNA/aptamer/Fe₃O₄@Au incubating with CEM control cells, (c) t-DNA/random DNA/Fe₃O₄@Au incubating with Ramos target cells, and (d) t-DNA/aptamer/Fe₃O₄@Au incubating with Ramos target cells. The TEM images of aptamer/Fe₃O₄@Au assembled on target cell (sample d) and control cell (sample b). The concentration of each cell sample is 4000 cells mL⁻¹.

Selectivity. To demonstrate the selectivity of the above design strategy, four samples were detected: t-DNA/aptamer/Fe₃O₄@Au (blank), t-DNA/aptamer/Fe₃O₄@Au with 4000 cells mL⁻¹ for Ramos target cells, t-DNA/aptamer/Fe₃O₄@Au with 4000 cells mL⁻¹ for CEM control cells, and t-random DNA-Fe₃O₄@Au with 4000 cells mL⁻¹ for Ramos target cells. As presented in Fig. 6, the SECL signals responded in a proportional manner to the amounts of released t-DNA from t-DNA/aptamer/Fe₃O₄@Au, which showed that the nonspecific binding did not occur and the selective binding resulted in the assembly of aptamer/Fe₃O₄@Au around the target cells and an increase in the SECL signals. The excellent selectivity of the assay stemed from the highly selective aptamer utilized in the experiment and also from the advantages of Fe₃O₄@Au themselves.

Mixed samples assay

In order to evaluate the potential application of the proposed method, determination of the detection capability of the assay was performed using artificial complex samples by mixing equal amounts of Ramos target cells and CEM control cells and imaged by confocal microscopy. In this experiment, CdTe QDs were used to label CEM cells prior to t-aptamer-Fe₃O₄@Au incubation in order to differentiate Ramos from CEM cells. CdTe-labeled CEM control cells were mixed with unlabeled Ramos cells with a ratio of 1 : 1 (Fig. 7A). After adding t-DNA/aptamer/Fe₃O₄@Au and incubating at 37 °C for 15 min with occasional gentle stirring, a magnetic field was applied to remove the control cells that were not attached to the aptamer-labeled Fe₃O₄@Au. The magnetically extracted and supernatant samples were then analyzed by confocal imaging as shown in Fig. 7B and C, respectively. The experiment was also performed by labeling Ramos target cells with CdTe QDs and mixing them (1 : 1) with unlabeled control cells as in Fig. 7D. The target cells separated by magnetic separation exhibited a CdTe fluorescence signal (Fig. 7E), while the supernatant control sample did not show fluorescence emission (Fig. 7F). The presence of the CdTe fluorescence proved that only the Ramos cells were collected and imaged. From the above results, it can be seen clearly that the assay was able to collect the Ramos cells in complex sample. The lack of the CdTe signal in Fig. 7B, along with the presence of the CdTe signal in Fig. 7E, proved that only target cells were binding collectedly using this method for extractions from 1 : 1 cell mixtures. The results were also confirmed by SECL strategy (Fig. S10†).

Fig. 7 Images of (A) 1 : 1 ratio of target cells with CdTe QDs-stained control cells. After magnetic extraction, (B) target cells extracted by aptamer/Fe₃O₄@Au, and (C) CdTe QDs-stained control cells in supernatant. (D) 1 : 1 ratio of CdTe QDs-stained target cells with control cells. After magnetic extraction, (E) CdTe QDs-stained target cells extracted by aptamer/Fe₃O₄@Au, and (F) control cells in supernatant. Here, Ramos cells are taken as target cells, and CEM cells as control cells.

Blood samples assay

Blood samples were also used to show the applicability of the method in a real assay of biological samples. In order to accomplish this end, Ramos cells were spiked into whole blood samples and compared with unspiked samples. After magnetic extraction, the stepwise double amplification SECL analysis process was performed as described above. Table S5† shows the SECL intensities of the blood samples with and without the defined added amounts of Ramos cells. The recovery was found to vary from 81.0 to 98.6%. Each sample was repeated three times and averaged to obtain the recovery values.

Conclusions

In summary, the cooperative action between Cu²⁺ and Fe³⁺ led unique SECL system for luminol–H₂O₂ is achieved from a CuS/DNAAu/DNA/MNPs multiplex nanoparticle probe, which notably improves the sensitivity compared to that employing a monometallic ion as catalyst. Coupled with dual-amplified function of MNPs by DNA amplification and synergistically enhanced metal ions amplification, the method allows the detection of DNA down to 6.8 aM, which is comparable with other most ultrasensitive DNA detection.^20,21 Moreover, for the detection of cancer cells, MNPs are also demonstrated to selectively extract Ramos target cells and significantly amplify FI-CL signals as reporter labels based on the cell-SELEX technique for aptamer selection. Whereas most previous studies for cells assay have to bind cells on supporters,^28,29 this strategy is simpler and more flexible due to the selectivity and affinity of aptamers selected for Ramos cells. The variability in fabrication procedures, automation and succession of the FI-CL detection step, and good reproducibility of the method opens up a new avenue for sensitive and selective detection of a wide spectrum of aptamer-based analytes and holds great promise for direct analytical applications in vivo or in vitro.

Experimental section

Synthesis of Au NPs

Au NPs were prepared according to the method of reduction of tetrachloroauric acid with trisodium citrate. Briefly, 100 mL of 0.01% HAuCl₄ solution was heated to boiling with vigorous stirring, and then 3.0 mL of 1% trisodium citrate solution was added dropwise rapidly. After the color of the solution changed from gray yellow to deep red, the heating source was removed and the resulting gold colloidal suspension with an average diameter of ∼20 nm (0.6 nM) was stirred for an additional 15 min to cool down to room temperature.

Preparation of water-soluble CuS NPs

In brief, 3.0 μL mercaptoacetic acid was added to 50 mL of 0.4 mM CuSO₄ solution, and the pH of the mixture was adjusted to 9.0 with 0.5 M NaOH solution. After being bubbled with N₂ for 30 min, 1.34 mM Na₂S solution was added dropwise to the mixture to keep the molar ratio of Na₂S to CuSO₄ at approximately 2.5. The reaction was carried out for 24 h under N₂, and a brown colloid was formed gradually.

Modification of p2-DNA with Au NPs and CuS NPs

First, the p2-DNA/Au NPs were synthesized by dispersing 3 mL of Au NPs (0.6 nM) in 2 mL of 1.0 × 10⁻⁷ M 3′-thiol-modified p2′-DNA and 1.0 × 10⁻⁷ M 3′-thiol-modified p2′′-DNA, respectively, which were previously deprotected with TCEP for 1 h,²⁵ followed by shaking gently for 16 h and “aged” in the solution (0.3 M NaCl, 10 mM Tris-acetate, pH 8.2) for another 48 h.^26,27 Excess DNA was removed by successive washing of the Au NPs conjugates with centrifugation at 15 000 rpm for 30 min. The red oily precipitate was washed with 5 mL of 0.1 M pH 7.0 PBS buffer containing 0.1 M NaCl, recentrifuged, and dispersed in 5 mL of 0.1 M pH 7.0 PBS buffer solution containing 0.3 M NaCl. Approximately 25–30% of the original nanoparticle concentration was lost during each centrifugation and workup procedure.²⁶ And then, 2.0 mL of 5′-amino group modified p2-DNA/Au solution was incubated with 200 μL of 0.1 M imidazole solution (pH 7.0) for 30 min, and dispersed in 100 μL of 0.1 M EDC solution and 2.0 mL of CuS colloid solution and stirred overnight. The resulting CuS/p2′-DNA/Au and CuS/p2′′-DNA/Au conjugates were centrifugated at 10 000 rpm for 30 min, followed by washing and resuspending in deionized and doubly distilled water. The two kinds of CuS/p2-DNA/Au (∼0.2 nM) were stored separately at 4 °C until use. The ratio of Au NPs, p2-DNA, and CuS NPs modified on the MNPs were ∼1:101:103.¹⁹

Preparations of t-DNA/aptamer/Fe₃O₄@Au and bbc-p1-DNA/MNPs

The aptamer/Fe₃O₄@Au conjugates were synthesized by adding ∼50 μL of aqueous Fe₃O₄@Au solution (5 mg mL⁻¹) to 500 μL of 2.0 × 10⁻⁸ M thiol-modified aptamer. Thiol-modified aptamer was deprotected and activated with TCEP (10 mM) for 1 h before attaching to Fe₃O₄@Au. After shaking gently for 16 h at room temperature, the aptamer-Fe₃O₄@Au conjugates were “aged” in the solution (0.3 M NaCl, 10 mM Tris-acetate, pH 8.2) for another 48 h. Excess reagents were removed by magnetic force, followed by adding 500 μL of 1.0 × 10⁻⁷ M t-DNA solution and incubating at 37 °C for 1 h to obtain the t-DNA/aptamer/Fe₃O₄@Au conjugates.

The fabrication of bbc-p1-DNA/MNPs was carried out based on a slightly modified literature protocol.¹⁸ Briefly, 100 μL suspension of MNPs (10 mg mL⁻¹) was transferred into a 1.5 mL Eppendorf (EP) tube. The MNPs were then washed three times with 400 μL of 0.1 M imidazol-HCl buffer (pH 7.0), and a 0.1 M imidazol-HCl buffer (300 μL) and a 0.8 M EDC (100 μL) were added to the EP tube. The mixture was incubated at 37 °C for 30 min to activate the carboxylate groups on the MNPs, followed by washing three times with 400 μL of 0.1 M PBS buffer. Then, 1.8 mL of 7.0 × 10⁻⁷ M bbc-DNA and 180 μL of 7.0 × 10⁻⁷ M p1-DNA were added and the mixture was incubated at 37 °C overnight with gentle mixing. Finally, the resulting MNPs, coated with bbc-DNA and p1-DNA, were washed three times with 400 μL of 0.1 M PBS buffer and resuspended in 200 μL PBS buffer for further use at 4 °C. Given the advantage of MNP in simple and complete magnetic separation, it could be considered that the MNP concentration was not lost during the modification procedures with the final concentration of ∼5 mg mL⁻¹ in 200 μL PBS buffer. The surface coverage of bbc-DNA and p1-DNA on MNPs was determined according to UV-visible absorbance and fluorescence. The ratios of MNPs, t-DNA and bbc-DNA were approximately 1/1.69 × 10⁵/1.22 × 10⁶.¹⁸

DNA assay

t-DNA sandwich hybridization assay. 96 Wells of streptavidin-coated microtiter plates were washed with 200 μL of PBS buffer solution twice before use. 100 μL of 1.0 × 10⁻⁷ M biotin-c-DNA in PBS was added into the wells and incubated for 30 min at room temperature. Unconjugated biotin-c-DNA was then removed by discarding the supernatants, and the wells were washed three times with 200 μL of PBS buffer solution, followed by 200 μL of 1.0 M NaCl for three times. And then, 10 μL of desired concentration of diluted t-DNA solution and 50 μL of 1.0 M NaCl were added in triplicate to the wells which has been immobilized c-DNA. After incubating for 4 h at 25 °C under gentle agitation to capture t-DNA onto the substrate, the solution was discarded and the substrate was washed three times with 200 μL of PBS buffer solution. Subsequently, an amount of 20 μL of the bbc-p1-DNA/MNPs suspension and 50 μL of 1.0 M NaCl was added into the well which contained captured t-DNA. After hybridization for 4 h to bind bbc-p1-DNA/MNPs to t-DNA sequences on the substrate of the well, unbound bbc-p1-DNA/MNPs was then removed, and the wells were washed three times with 200 μL of PBS buffer solution.

Release of bbc-p1-DNA/MNPs from substrate and hybridization with CuS/p2-DNA/Au conjugates. Double-stranded DNA can be dehybridized through strong alkali solutions, such as urea or NaOH. In this study, 200 μL of 50% (w/w) urea solution was introduced into the wells with sandwich-type hybridization complex to release bbc-p1-DNA/MNPs from the substrate through dehybridization. After reaction for 10 min at 25 °C, the suspension containing the released bbc-p1-DNA/MNPs in urea solution was transferred to a 1.5 mL EP tube. The bbc-p1-DNA/MNPs were magnetically seperated from the solution on a magnetic rack, and washed three times with 400 μL of 1.0 M NaCl. And then, the corresponding the amounts of 100 μL of 0.2 nM Au/p2′-DNA/CuS and 10 μL 0.2 nM Au/p2′′-DNA/CuS in 50 μL of 1.0 M NaCl were added to the EP tube containing the bbc-p1-DNA/MNPs, followed by hybridizing for 4 h at 25 °C. The resulting hybridization conjugates on MNPs were washed three times with 200 μL of PBS buffer solution.

FI-CL detection. Nitric acid was used for the dissolution step since it was observed that it can efficiently oxidize the CuS NPs but not the Au NPs and MNPs after optimizing concentration and reaction time. Preliminary experiments on dissolving CuS NPs with 0.1 M nitric acid and FI-CL detection of the released Cu²⁺ showed that cupric was fully oxidized after 5 min. In this assay, the above resulting MNPs modified with double-stranded hybrids labeled with Au and CuS NPs were transferred into a colorimetric tube. The dissolution of CuS NPs anchored on the hybrids was carried out by the addition of 200 μL of 0.1 M nitric acid solution for 5 min to ensure complete oxidation. Following a magnetic separation, the volume of the solution (containing the dissolved CuS NPs) was adjusted to 4 mL with deionized and doubly distilled water, and the pH was adjusted to 4.0 through 0.1 M Na₂B₄O₇-NaOH buffer (pH 10.0). The FI-CL detection of the dissolved CuS NPs was performed as described in our previous work.¹⁹ Briefly, 1.5 × 10⁻³ M luminol in pH 10.0, 0.1 M Na₂B₄O₇-NaOH buffer solution was first mixed with 8.0 × 10⁻³ M H₂O₂, and then reacted with metal ions in the flow cell to produce a CL signal. Furthermore, 0.4 M nitric acid was used to dissolve CuS NPs and MNPs simultaneously. The detection of luminol–H₂O₂–Cu²⁺–Fe³⁺ synergistic FI-CL procedures were the same as those of luminol–H₂O₂–Cu²⁺ FI-CL detection procedures.

Cell assay

Magnetic extraction was carried out by adding specified amounts of t-DNA/aptamer/Fe₃O₄@Au to each cell sample followed by incubating the mixture for 15 min at 37 °C. And then a magnetic field was introduced to the sample container. After the magnetic separation, 10 μL of the supernatant containing released t-DNA which was proportional to the amounts of target cells was decanted using a pipette followed by the above operations of the DNA assay.

Acknowledgements

The work was supported by the Excellent Young Scientists Foundation of Shandong Province (JQ200805), the National Natural Science Foundation of China (20827005), and the National Basic Research Program of China (2010CB732404).

Notes and references

1. X.-M. Qian and S. M. Nie, Chem. Soc. Rev., 2008, 5, 912–920.
2. N. L. Rosi and C. A. Mirkin, Chem. Rev., 2005, 105, 1547–1562.
3. R. Gill, M. Zayats and I. Willner, Angew. Chem., Int. Ed., 2008, 47, 7602–7625.
4. A. Bajaj, S. Rana, O. R. Miranda, J. C. Yawe, D. J. Jerry, U. H. F. Bunz and V. M. Rotello, Chem. Sci., 2010, 1, 134–138.
5. G.-F. Jie, P. Liu and S.-S. Zhang, Chem. Commun., 2010, 46, 1323–1325.
6. G. D. Liu and Y. H. Lin, J. Am. Chem. Soc., 2007, 129, 10394–10401.
7. S. Bi, Y. M. Yan, X. Y. Yang and S. S. Zhang, Chem.–Eur. J., 2009, 15, 4704–4709.
8. S. Bi, S. Hao, L. Li and S. S. Zhang, Chem. Commun., 2010, 46, 6093–6095.
9. J.-C. Liu, W.-J. Chen, C.-W. Li, K.-K. Tony Mong, P.-J. Tsai, T.-L. Tsai, Y. C. Lee and Y.-C. Chen, Analyst, 2009, 10, 2087–294.
10. Y.-M. Huh, E.-S. Lee, J.-H. Lee, Y.-W. Jun, P.-H. Kim, C.-O. Yun, J.-H. Kim, J.-S. Suh and J. Cheon, Adv. Mater., 2007, 19, 3109–3112.
11. R. Elghnian, J. J. Storhoff, R. C. Mucic, R. L. Letsinger and C. A. Mirkin, Science, 1997, 277, 1078–1081.
12. C.-W. Liu, Y.-T. Hsieh, C.-C. Huang, Z.-H. Lin and H.-T. Chang, Chem. Commun., 2008, 2242–2244.
13. J. W. Liu and Y. Lu, J. Am. Chem. Soc., 2007, 129, 8634–8643.
14. J. Wang, D. Xu, A.-N. Kawde and R. Polsky, Anal. Chem., 2001, 73, 5576–5581.
15. J. Wang, R. Polsky, A. Merkoci and K. L. Turner, Langmuir, 2003, 19, 989–991.
16. K. Aslan and C. D. Geddes, Chem. Soc. Rev., 2009, 38, 2556–2564.
17. A. P. Fan, C. Lau and J. Z. Lu, Anal. Chem., 2005, 77, 3238–3242.
18. S. Bi, H. Zhou and S. Zhang, Chem. Commun., 2009, 5567–5569.
19. S. Zhang, H. Zhong and C. Ding, Anal. Chem., 2008, 80, 7206–7212.
20. C. S. Thaxton, H. D. Hill, D. G. Georganopoulou, S. I. Stoeva and C. A. Mirkin, Anal. Chem., 2005, 77, 8174–8178.
21. B. Munge, G. Liu, G. Collins and J. Wang, Anal. Chem., 2005, 77, 4662–4666.
22. Z. Tang, D. Shangguan, K. Wang, H. Shi, K. Sefah, P. Mallikratchy, H. W. Chen, Y. Li and W. Tan, Anal. Chem., 2007, 79, 4900–4907.
23. C. D. Medley, J. E. Smith, Z. Tang, Y. Wu, S. Bamrungsap and W. Tan, Anal. Chem., 2008, 80, 1067–1072.
24. J. E. Smith, C. D. Medley, Z. Tang, D. Shangguan, C. Lofton and W. Tan, Anal. Chem., 2007, 79, 3075–3082.
25. J. Liu and Y. Lu, J. Am. Chem. Soc., 2005, 127, 12677–12683.
26. J. J. Storhoff, R. Elghanian, R. C. Mucic, C. A. Mirkin and R. L. Letsinger, J. Am. Chem. Soc., 1998, 120, 1959–1964.
27. J. Zhang, S. Song, L. Zhang, L. Wang, H. Wu, D. Pan and C. Fan, J. Am. Chem. Soc., 2006, 128, 8575–8580.
28. C. Hao, L. Ding, X. Zhang and H. Ju, Anal. Chem., 2007, 79, 4442–4447.
29. E. Han, L. Ding, H. Lian and H. Ju, Chem. Commun., 2010, 46, 5446–5448.

---

Footnote

† Electronic supplementary information (ESI) available: Additional experimental section, DNA hybridization assays, investigation of SECL system, cell assays of mixed samples and blood samples, respectively. See DOI: 10.1039/c0sc00341g

---