Hybridization-driven fluorometric platform based on metal-organic frameworks for the identification of the highly homologous viruses

Graphical abstract

A MOF-modified fluorometric strategy was successfully established based on the hybridization-induced mechanism. It demonstrated the excellent analytical characteristics in the detection of SARS-CoV-2. Most importantly, SARS-CoV-2 and SARS-CoV can be efficiently distinguished only through a simple auxiliary test. Compared with the traditional means, this identification circumvents the trouble of pretreatment experiments, the primers optimization and the enzymatic mixtures. It is rapid, low-cost and easy-to-use tools.

Abstract

A novel fluorometric strategy for the simultaneous identification of SARS-CoV-2 and SARS-CoV was successfully established based on a hybridization-induced signal on–off-on mechanism. Here, one part of the probe (P1) of SARS-CoV-2 (P = P1/P2) is partially related to SARS-CoV, while the other part (P2) is completely irrelevant to SARS-CoV. They as smart gatekeepers were anchored on NH₂-MIL-88(Fe) (MOF@P1/P2) to turn off its catalytic performance. Only the specific SARS-CoV-2 genetic target can strongly restore the peroxidase-like activity of MOF@P1/P2. In the presence of o-phenylenediamine, SARS-CoV-2 can be efficiently detected with high sensitivity, accuracy, and reliability. This strategy demonstrated excellent analytical characteristics with a linear range (10^-9 M ∼ 10^-6 M) under the limit of detection of 0.11 nM not only in buffer but also in 10 % serum, which partly shows its practicability. Most importantly, with the help of the auxiliary test of MOF@P1 and MOF@P2, SARS-CoV-2 and SARS-CoV can be efficiently quantified and distinguished. This novel strategy has provided a breakthrough in the development of such identification. In the whole process, only a simple one-step experiment was involved. This circumvents the trouble of pretreatment experiments in traditional methods, including complex enzymatic mixtures, specialized experimental equipment, many primers optimization as well as reverse transcriptase. Additionally, this novel strategy is rapid, low-cost, and easy-to-use tools.

1. Introduction

Viral contamination of agricultural products can easily lead to various diseases. Some diseases have very similar biological characteristics [[1], [2]], but may be caused by different members of a given viral family, such as pneumonia [[3], [4]] caused by SARS-CoV or SARS-CoV-2 and hand, foot, and mouth disease caused by enterovirus ev71 or coxsackie virus [5]. If it isn’t distinguished which virus causes this disease, it is impossible to select appropriate therapies to personalize the treatments for patients. Even, huge economic losses will occur as the price of disease delay [[6], [7], [8]].

The bottleneck in the identification of these viruses is that the difference in their highly homologous genomic structures is extremely small. Although antibody detection is one of the current laboratory methods, in the face of these viruses, it suffers from the problem of false positives due to their cross antigenicity [[9], [10]]. Nucleic acid testing [[11], [12], [13]], such as PCR and RT-PCR, is the gold standard. Its positive rate is as high as 90 %. However, it is difficult to simultaneously distinguish every virus only though a simple one-step experiment. Because the nucleic acid signal readout depended on target nucleic acid transcribed by reverse transcriptase based on the gene sequence of one member of a given viral family and exponentially amplified by temperature cycle. For each member of a given viral family, we had to design its corresponding specific primer for its amplification [[12], [14]], for example, the experiment of dual PCR for the identification of goat pox and sheep pox viruses [15]. Obviously, the requirement of many primers optimization, reverse transcriptase, complex enzymatic mixtures, as well as specialized experimental equipment, make this approach time-consuming, expensive, susceptible to false-positive results, and unsuitable for on-site nucleic acid diagnostics. Therefore, it is important to develop rapid, low-cost, and easy-to-use tools for the simultaneous identification of these highly homologous viruses, especially in the point-of-care (POC) setting. Luckily, the continuous emergence of novel nanomaterials brought the unprecedented opportunities for the development of new strategies [[16], [17], [18], [19]].

Metal–organic frameworks (MOFs) are a class of porous inorganic–organic hybrid materials, exhibiting high surface areas, controlled pore sizes, open metal sites, and organic linkers [[20], [21], [22]]. The highly ordered pores can absorb and store enormous quantities of targeted molecules. The open metal sites in the framework provide intrinsic fluorescence-quenching properties or catalase-like catalytic activity. The tailored functionalization of the organic ligands endows this material with the ability to further chemical modifications. All of these promote its comprehensive applications in various fields [[23], [24], [25]], even in biosensors for bioanalysis [[26], [27], [28], [29], [30], [31]].

In this study, we only paid close attention to the construction of MOFs-based fluorescence sensing [[32], [33]]. The usual strategy mainly focuses on the use of MOFs as either signal probes or signal probe carriers, including the highly sensitive controllable release of fluorescence molecules or fluorescence-quenching-recovery detection [[34], [35]]. In these sensing areas, signal probes or signal probe carriers have to keep in constant contact with specimens. As we known, the real specimen is composed of proteins, enzymes, organic and inorganic compounds. Some components may affect not only the quenching effect of metal ions in MOFs but also the spectral position and intensity of fluorescent dyes released from MOFs. This will seriously affect the accuracy of the final results. Moreover, the cost of the strategy will significantly increase due to the addition of the fluorescent material. Benefiting from the unique intrinsic enzyme-mimic activity [[36], [37], [38]], high specific surface area, and good stability in water, the resultant NH₂-MIL-88(Fe) (MOF) nano octahedra were selected as the substrate [[39], [40], [41]]. In our sensing design, its nanozyme activity is used to design a turn off/on fluorometric platform.

Recently, SARS-CoV-2 and SARS-CoV viruses hidden in various agricultural products have brought irreversible harm to human health [42], therefore it is an urgent task to find a simple solution to the identification and treatment of such viruses [[43], [44], [45], [46]]. However, this pair of viruses belong to β-Coronavirus family with high genetic similarity. If we want to achieve its highly sensitive detection, the probe gene should be designed carefully. After a lot of screening, the specific SARS-CoV-2 gene (P = P1/P2) recommended by the WHO was finally anchored on MOF for the construction of a novel controllable platform (MOF@P1/P2). A partial sequence (P1) of SARS-CoV-2 gene is very close to that of SARS-CoV, only three fixed nucleotide bases were different. Another partial sequence (P2) of SARS-CoV-2 gene can detect only the SARS-CoV-2 virus. These probes as smart gatekeepers effectively blocked all inlets of the reaction tank on MOF. Only the target of SARS-CoV-2 (T = T1/T2) is the highly efficient key to powerfully 'turn-on' the nanozyme activity of MOF@P1/P2. In the presence of H₂O₂, o-phenylenediamine (OPD) is oxidized to 2,3-Phenazinediamine (DAP) fluorescence at 565 nm is displayed. This strategy can linearly detect the the 1μM ∼ 1nM target of SARS-CoV-2 with the detection limit of 0.11 nM, even in 10 % serum. Interestingly, the target of SARS-CoV (S = S1/S2), a partially similar but still not fully complement sequence doesn't have enough strength to strongly rejuvenate MOF@P1/P2. As a result, a weak fluorescence at 565 nm is shown.

However, only through this weak positive result, we define that it is SARS-CoV, which is a rather risky approach. After all, the other unrelated components in the test samples may also lead to the weak positive results. Under our careful design, T2, the partial sequence of target SARS-CoV-2, can completely hybridize with P2, but S2, the partial sequence of target SARS-CoV, cannot. According to this fact, we decided to carry out the auxiliary test depended on MOF@P1 and MOF@P2, while experiments based on MOF@P1/P2 are implemented. Judging the results from MOF@P1/P2, MOF@P1, and MOF@P2 comprehensively, we acquired this conclusion that three strong positives are demonstrated for SARS-CoV-2, only two weak positives for SARS-CoV, which directly help us to quantify and distinguish two highly homologous viruses without any requirement of complex enzymatic mixtures, specialized experimental equipment, many primers optimization as well as reverse transcriptase.

Additionally, during the whole experiment, this fluorometric platform can be immediately separated by centrifugation and water washing due to the water insolubility of MOF after sensing. This cleverly circumvents the various interferences from the real specimen, which ensures that the final result is more accurate. The platform may have several additional properties, such as cost-effectiveness, easy availability, environmentally friendly starting materials, and ease of fabrication. In a word, this strategy can be prospective new diagnostics method for the distinction of homologous viruses, which is sensitive and competitive with the traditional techniques but is quicker, easier and more user-friendly for on-site application.

2. Experimental

2.1. Chemicals and characterization

Iron (III) chloride hexahydrate (FeCl₃·6H₂O) was bought from Sigma-Aldrich, 2-amino benzene dicarboxylic acid (NH₂-BDC) from Sinopharm Chemical Reagent Co and N, N- Dimethyl formamide (DMF) from Aladdin Reagents (Shanghai, China) o-phenylenediamine (OPD) was bought from Aladdin. OMP31, BP26 were donated by college of animal science and technology, Shihezi University. Ethidium bromide stock solution (EB), 40 % acrylamide/bis-acrylamide gel solution (29: 1), 6 × glycerol gel loading buffer VII (with xylene cyanole, BPB), and 50 × Tris-Acetic acid-EDTA (TAE) buffer were purchased from Sangon Biotech (Shanghai) Co., ltd. (Shanghai, China) and tetra-methyl ethylenediamine (TEMED) from Beijing Woke Biotechnology Co., ltd. All oligonucleotide sequences listed in Table S1 were synthesized and purified by high performance liquid chromatography (HPLC) at Sangon Biotech (Shanghai) Co., ltd. (Shanghai, China). All reagents were of analytical grade quality and were used without further purification. All aqueous solutions were prepared with pure water produced by a Milli-Q water (resistance > 18 MΩ/cm) purifying system. NaAc-HAc buffer (pH 5.5, 0.2 M) was employed in the total experiment.

Scanning Electron Microscopy (SEM) was fulfilled with a JEOLJSM-7500F SEM instrument (Hitachi High-Technology Co., ltd. Japan), Transmission electron microscopy (TEM) with a JEM-2100 TEM instrument (Hitachi High-Technology Co., ltd. Japan) and Fourier transform infrared spectroscopy (FTIR) with the Bruker Tensor 70 spectrometer (Bruker Optics, Germany). All fluorescence measurements were performed on a FL-2700 fluorescence spectrophotometer (Shimadzu, Tokyo, Japan) under the following instrument parameters: Excitation and emission slit width 10 nm. The excitation and emission wavelengths were set at 435 nm and 565 nm, respectively.

2.2. The preparation of MOF@P1/P2

P1/P2 freeze-dried powder in the ratio of 1:1 was centrifuged at 4000 rpm for 1 min, then added into NaAc-HAc buffer (0.2 M). After the dilution, this solution was heated at 95 ℃ for 5 min, then rapidly cooled to room temperature. Next, 100 μL MOF (2 mg/mL) was mixed with 100 μL P1 and P2 for 20 h at 37 ℃. Finally, MOF@P1/P2 was centrifuged to remove loose and unabsorbed P1/P2, then dispersed in NaAc-HAc buffer (100 μL,0.2 M) for the further utilization (Scheme 1 A). For MOF@P1 and MOF@P2, the preparation method is the same as above, except that P1 or P2 replaces P1/P2.

Scheme 1.

(A) Schematic diagram of the synthesis of MOF@P1/P2. Fe₃-μ₃-clusters in MOF were blocked by P1 (blue single-stranded DNA) /P2 (purple single-stranded DNA) as smart gatekeeper (red cross). (B) Catalytic mechanism of MOF. OPD reacts in the presence of hydrogen peroxide to generate DAP (yellow fluorescence) under the catalysis of Fe₃-μ₃-clusters in MOF (red arrow). (C) Catalytic mechanism of MOF@P1/P2. The conversion efficiency of non-fluorescent OPD to fluorescent DAP is greatly reduced in hydrogen peroxide medium. (D) Catalytic mechanism of MOF@P1/P2 after the introduction of target not only in buffer but also in serum. (E) The catalytic results of MOF, MOF@P1/P2 and T1/T2-induced MOF@P1/P2. Here, smartphone image was converted to RGB intensity by ImageJ. (F) Identification of SARS-CoV-2 and SARS-CoV through MOF@P1, MOF@P2, MOF@P1/P2.

2.3. Optimization of experimental conditions for MOF@P1/P2

Optimization of pH and temperature: 20 μL MOF@P1/P2, 10 μL T1 and 10 μL T2 were mixed for 8 h, then centrifuged and diluted with 100 μL NaAc-HAc buffer with different pH values (0.2 M, pH 5 ∼ 6). Next, 20 μL H₂O₂ (10^-3 M) and 80 μL OPD (10^-3 M) were added into the above solution for 40 min. Finally, fluorescence signal of supernatant was measured according to the following parameter: the operating voltage is 700 V, the excitation and emission slits are set at 10 nm and the excitation wavelength is 435 nm. For the optimization of temperature and catalytic reaction time, the operation method is the same as above, only the temperature is changed from 4 °C to 45 °C or the catalytic time is varied from 20 to 60 min.

Optimization of adsorption time and desorption time: 100 μL MOF was added into 100 μL P1 and P2 (1:1, 10^-5 M) for 1 h ∼ 24 h incubation, then centrifuged and dispersed in NaAc-HAc buffer (100 μL,0.2 M). Subsequently, 20 μL MOF@P1/P2, 20 μL T1/T2 (1:1, 10^-6 M) were mixed for 8 h, then centrifuged and diluted with 100 μL NaAc-HAc buffer. Finally, 20 μL H₂O₂ (10^-3 M) and 80 μL OPD (10^-3 M) were added into the above solution for 40 min. Finally, fluorescence signal of supernatant was measured. The same experimental procedure was repeated, except that the desorption time between MOF@P1/P2 and T1/T2 varied from 1 h to 10 h.

2.4. The selectivity and stability of MOF@P1/P2

20 μL of MOF@P1/P2 was respectively dipped into several e-tubes, then mixed with 20 μL of targets one by one (T1/T2, 1:1, 10^-6 M; S1/S2 1:1 10^-6 M; WB1/WB2, 1:1 10^-6 M). After the incubation at 37 °C for 8 h, the mixture was centrifuged and washed with water. Next, 80 μL OPD (10^-3 M), 20 μL H₂O₂ (10^-3 M) and 100 μL NaAc-HAc buffer (0.2 M) were added for 40 min. Finally, the supernatant was transferred to a fluorescent quartz cell for the detection of fluorescence signal under the parameters below: the excitation wavelength of 435 nm, 700 V working voltage and 10 nm excitation slit and emission. Finally, 20 μL of MOF@P1 or MOF@P2 was respectively employed to mix with 20 μL of targets (T1, 10^-6 M. T2, 10^-6 M. S1, 10^-6 M. S2, 10^-6 M. WB1, 10^-6 M. WB2, 10^-6 M). After the same experimental procedure was repeated, fluorescence signals were collected.

Image software determination: The photos were taken in a well-lit room during the day without any external light sources using Honor 20. The brightness and distance between the material and the smartphone are fixed. After the catalysis for 40 min, photos of samples (MOF@P1/P2 + T1/T2, MOF@P1/P2 + S1/S2, MOF@P1/P2 + WB1/WB2) were taken, and the RGB intensity of the collected images was determined by image processing algorithm using Image J software. The image processing algorithm is to calculate the average intensity of three adjacent pixels.

For the stability of MOF@P1/P2, the operation method is the same as above, except that MOF@P1/P2 stored in the refrigerator for different times (0 d、2 d、4 d、6 d、8 d、10 d) replaces the freshly prepared MOF@P1/P2.

2.5. The biosensing of MOF@P1/P2 in buffer or in serum

20 μL of MOF@P1/P2 was respectively mixed with 20 μL target (T1/T2 in the ratio of 1:1) with different concentration ranged from 10^-6 M−10⁻⁹ M at 37 °C for 8 h. After the centrifugation and water washing, 80 μL OPD (10^-3 M), 20 μL H₂O₂ (10^-3 M) and 100 μL NaAc-HAc buffer (0.2 M) were added for 40 min. Then the fluorescence signal of the supernatant was measured. In addition, targets were respectively diluted with 10 % serum, then the above operation was repeated again.

2.6. Gel electrophoresis assay.

The experimental procedure of PAGE was implemented according to the following steps. First, 17.5 % PAGE was mixed with 40 % acrylamide/bis-acrylamide gel solution (3.5 mL), deionized water (4.25 mL), 1 × TAE buffer (160 μL), 10 % ammonium persulfate (APS) (80 μL) and TEMED (4 μL). After polymerization at 37 °C for 30 min, the freshly-prepared gel was soaked in 1 × TAE buffer. Then, samples (5 μL) were mixed with 6 × glycerol gel (2 μL). After loading, the gel electro phoresies were respectively performed at 180 V for 3 min and 135 V for 90 min. Next, the gel was stained with ethidium bromide (EB) dye for 30 min, and the stained gel plate was observed under a UV light with a wavelength of 254 nm and photographed for preservation.

3. Results and discussions

3.1. Characterization of MOF@P1/P2

Before performing detection experiments, MOF@P1/P2 was first characterized by scanning electron microscopy (SEM), transmission electron microscopy (TEM), Energy Dispersive Spectrometer (EDS) and flourier transform infrared (FTIR). As shown in Fig. 1 A, 1B, MOF@P1/P2 is the regular octahedral structure with a particle size of about 200 nm in a good dispersion, which is basically similar to that of MOF (Supplementary Material Fig. S1A, S1B). Only the surface of MOF@P1/P2 becomes slightly rough. This phenomenon also indirectly indicates that DNA film modified on the MOF is very thin. This is very beneficial to the recovery of nanozyme activity of MOF in the following studies. Additionally, in addition to C, O, N, and Fe, the P elements that cannot be captured in the EDS of the MOF (Fig. S1C) are clearly observed in that of MOF@P1/P2 (Fig. 1C). Moreover, in FTIR spectroscopy of MOF@P1/P2 (Fig. 1D), besides the characteristic absorption peak of MOF (Fig. S1D), 1250 cm⁻¹ ∼ 1000 cm⁻¹ related to the pentose phosphate backbone vibration on the DNA chain was showed. Among them, the symmetric stretching band of PO₂ ⁻ located at around 1083 cm⁻¹. This indicated that probe (P1/P2) has been successfully adsorbed on the MOF. Additionally, as we known, MOF itself contains a certain amount of positively charged amino groups. With the accumulation of electronegative phosphate of DNA, the electrical property of MOF changes from positive to negative. As shown in the Fig. 1D inset, the zeta potential of MOF with and without DNA gradually changes from positive to negative. This proves that the construction of MOF@P1/P2 is successful once again.

Fig. 1.

SEM image of MOF@P1/P2(A), TEM image of MOF@P1/P2 (B), EDS of MOF@P1/P2 (Illustration: p element map) (C) and FTIR of MOF@P1/P2 (Illustration: zeta potential maps of MOF and MOF@P1/P2) (D).

3.2. Gene sequence

As shown in Table S1, several gene fragments were employed in the study. Two parts, namely P1 and P2, are the specific gene fragment of SARS-CoV-2 (P) recommended by the WHO. T represents the specific SARS-CoV-2 genetic target. It is composed of T1 and T2. Among them, T1 is a perfect match with P1, T2 with P2. S contains two parts, namely S1 and S2, which is from the gene-related sequences of SARS-CoV (S). S2 is a completely mismatched sequence with P2. However, S1 has only three fixed nucleotide bases that are different from T1. Therefore, S1, a similar but not fully complemented sequence, was able to partially interact and hybridize with P1 at room temperature. But the hybridization intensity of S1 was clearly weaker than that of T1.W consists of WB1 and WB2, which neither match P1 nor P2.

3.3. The mechanism of strategy

The hydrogen peroxide-like enzyme activity of NH₂-MIL-88(Fe) plays a major role in our strategy (Scheme 1B). Fe₃-μ₃-clusters in MOF have high electron-accepting ability serving as electron-consuming pools to generate Fe²⁺ through electron transfer under H₂O₂ reaction. OPD is used as a substrate. Firstly, H₂O₂ is activated by Fenton-like reaction to generate OH radicals as shown in formulas (1), (2). Then OPD reacts in the presence of hydroxyl radicals to generate DAP as shown in formulas (3), (4). Interestingly, this can further speed up the production of OH radicals, greatly improve the conversion efficiency of non-fluorescent OPD to fluorescent DAP. Finally, a significant fluorescence at 565 nm is showed.

$${\text{H}}_2{\text{O}}_2+{\text{Fe}}^{3+}\to {\text{Fe}}^{2+}+{\text{O}}_2+2{\text{H}}^{+}$$ (1)

$${\text{Fe}}^{2+}+{\text{H}}_2{\text{O}}_2\to {\text{Fe}}^{3+}+·\text{OH}+{\text{HO}}^{-}$$ (2)

$$\text{RH}+·\text{OH}\to {\text{H}}_2\text{O}+\text{R}·\to {\text{CO}}_2+{\text{H}}_2\text{O}$$ (3)

(4)

Next, P1/P2 (P1:P2 = 1.09 ± 0.02:1) was modified on MOF only through the ionic bond or the weak electrostatic adsorption (Fig. S2). Interestingly, the conversion efficiency of non-fluorescent OPD to fluorescent DAP is greatly reduced in hydrogen peroxide medium, which indicated that the peroxidase activity of MOF is “turned off” in the presence of P1/P2 (Scheme 1C, Fig. S3). Based on the theory of hybridization induction, after the introduction of target (T1/T2), the catalytic activity of MOF@P1/P2 can be efficiently restarted. The strong fluorescence signal appears again (Scheme 1D). As indicated in Fig S3, 68 % improvement can be observed. In addition, we analyzed the relative intensities of post-catalysis colors by capturing images. It is consistent with the fluorescence results (Scheme 1E).

In the whole experiment, the electronegativity of the fluorescence platform also changed, which indirectly showed the adsorption–desorption dynamical process of probe DNA as smart gatekeeper (Fig S3 inset). Additionally, this dynamic behavior was also monitored by SDS-PAGE. As shown in Fig. 2 A, native PAGE was performed in 1 × TAE running buffer, and the bands in lane 1 ∼ 6 respectively represent P1, P2, T1, T2, the supernatant of MOF@P1/P2, the supernatant of MOF@P1/ P2 after the incubation with T1/T2. As expected, no new band appeared in lane 5 and two new bands in lane 6. This indicates that P1/P2 stably stays on the MOF and does not fall off without the induction of T1/T2. Only after the introduction of T1/T2, P1/P2 desorbs from MOF, then hybridizes with them. It is consistent with the expected result.

Fig. 2.

(A) SDS-PAGE of MOF@P1/P2 + T1/T2. The bands in lane 1, lane 2, lane 3, lane 4, lane 5, lane 6 respectively represent P1, P2, T1, T2, the supernatant of MOF@P1/P2, the supernatant of MOF@P1/ P2 after the incubation with T1/T2. (B) Fluorescence emission spectra of (a) MOF@P1/P2 + T1/T2 + OPD + H₂O₂, (b) MOF@P1/P2 + T1/T2 + H₂O₂, (c) MOF@P1/P2 + T1/T2 + OPD, (d) MOF@P1/P2 + OPD + H₂O₂, (e) MOF@P1/P2 + H₂O₂ and (f) MOF@P1/P2 + OPD.

PH value is an important factor for a MOF@P1/P2-OPD-H₂O₂ fluorometric platform. Higher pH may destabilize H₂O₂, resulting in reduced oxidative activity. Therefore, different pH values ​​(5.0–6.0) of NaAc-HAc buffer (0.2 M) were employed. According to the obtained results in Fig. S4, the fluorescence intensity gradually increased, then reaches maximum at pH 5.5, which indicated that pH 5.5 is the optimum pH value. Subsequently, the optimized temperature and optimized catalysis time were determined. As shown in Fig. S5 or S6, the strongest fluorescence signal appears at 37 °C and 40 min catalysis. Therefore, 37 °C and 40 min were selected in the subsequent experiments. As shown in Fig. S7, MOF@P1/P2 prepared for 20 h incubation demonstrated a strong fluorescence signal in the target assay. So 20 h was chosen for the incubation time between MOF and P1/P2. By the way, Fig. S8 shows that fluorescence intensity is the strongest for 8 h incubation between MOF@P1/P2 and T, so 8 h is selected for subsequent experiments.

3.4. The feasibility of strategy

In addition, to prove the feasibility of our scheme, we evaluated the fluorescence signal of H₂O₂ + OPD + MOF@P1/P2 + T1/T2, H₂O₂ + MOF@P1/P2 + T1/T2 and OPD + MOF@P1/P2 + T1/T2, H₂O₂ + OPD + MOF@P1/P2, H₂O₂ + MOF@P1/P2 and OPD + MOF@P1/P2. As shown in Fig. 2B, no obvious fluorescence signal in the wavelength range of 520–640 nm was observed for H₂O₂ + MOF@P1/P2 + T1/T2, OPD + MOF@P1/P2 + T1/T2, H₂O₂ + MOF@P1/P2 and OPD + MOF@P1/P2, only very weak fluorescence signal for H₂O₂ + OPD + MOF@P1/P2 and a strong fluorescence change for MOF@P1/P2 + H₂O₂ + OPD + T1/T2. This phenomenon indicated that this fluorometric platform didn't properly work in the presence of OPD or H₂O₂ alone, slowly worked in present of OPD and H₂O₂ but in absence of analyte and efficiently worked in present of OPD, H₂O₂ and analyte. It proves that this novel strategy is feasible.

3.5. Specificity of MOF@P1/P2

Before studying the specificity of MOF@P1/P2 in detail, we first study the characteristics of each part of each target molecule, using MOF@P1 and MOF@ P2 (Scheme 1F). According to the characteristic of DNA sequence we selected, theoretically, both T1 and S1 have the ability to activate the catalytic activity of MOF@P1, but the strength should be significantly different. As shown in SDS-PAGE (Fig. 3A, inset), no new band appeared in lane 5 and lane 8, one new band in lane 6 and lane 7. Moreover, the new band in lane 6 is brighter than that in lane 7. It indicated that P1 could desorb from MOF only under the induction of T1 or S1. Moreover, the inducement ability of T1 is stronger than that of S1. This is consistent with our conjecture. Here, the bands in lane 1 ∼ 15 respectively represent P1, T1, S1, WB1, the supernatant of MOF@P1, the supernatant of MOF@P1 after the incubation with T1, the supernatant of MOF@P1 after the incubation with S1, the supernatant of MOF@P1 after the incubation with WB1, P1, T2, S2, WB2, the supernatant of MOF@P1 after the incubation with T2, the supernatant of MOF@P1 after the incubation with S2, the supernatant of MOF@P1 after the incubation with WB2. Therefore, as shown in Fig. 3A, a strong fluorescence signal is generated through the catalysis of T1-induced MOF@P1, a weaker fluorescence signal through the catalysis of S1-induced MOF@P1. The ratio of the two values is about 2:1. This also indicated that although S1 has three fixed nucleotide bases which are different from T1, it still hybridizes with P1 to a certain extent.

For MOF@P2, only T2 can technically recovery the catalytic activity of MOF. As shown in SDS-PAGE (Fig. 3 B, inset), a new band appeared in lane 6 and no new band in lane 7 and lane 8. It indicated that P2 separate itself from MOF@P2 only after the introduction of T2. Therefore, a strong fluorescence signal appears through the catalysis of T2-induced MOF@P2 (Fig. 3B).

Fig. 3.

(A) Selectivity of MOF@P1 for different gene segments (T1, T2, S1, S2, WB1, WB2) (inset: gel electrophoresis images) (B) Selectivity of MOF@P2 for different gene segments(T1, T2, S1, S2, WB1, WB2)(inset: gel electrophoresis images) (C) Selectivity of MOF@P1/P2 for different gene segments (T1/T2, S1/S2 and WB1/WB2)(inset: ImageJ software analysis diagram of T1/T2-induced MOF@P1/P2-OPD-H₂O₂, S1/S2-induced MOF@P1/P2-OPD-H₂O₂ and WB1/WB2-induced MOF@P1/P2-OPD-H₂O₂). (D)Gel electrophoresis images of selectivity of MOF@P1/P2 for different gene segments (T1/T2, S1/S2 and WB1/WB2). The error bars represent the SD (n = 3).

Finally, the specificity of MOF@P1/P2 was investigated in detail. As shown in Fig. 3C, the catalytic activity of MOF can be powerfully turned on by T1/T2, weakly triggered by S1/S2, invalidly switched on by WB1/WB2. SDS-PAGE (Fig. 3D) indirectly proves this fact. Among all lanes, two new bands appeared in lane 10, one new band in lane 11 and no new band in lane 12. It demonstrated that P1 and P2 could desorb from MOF under the induction of T1/T2, only P1 under the induction of S1/S2 and neither P1 nor P2 under the induction of WB1/WB2. This further confirms our conjecture about the dynamic process of this strategy. As a result, the strongest fluorescence signal appears through the catalysis of T1/T2-induced MOF@P1/P2, half of fluorescence signal through the catalysis of S1/S2-induced MOF@P1/P2 and no obvious fluorescence signal through the catalysis of WB1/WB2-induced MOF@P1/P2. This confirms that MOF@P1/P2-OPD-H₂O₂ system has good specificity for the detection of SARS-CoV-2. In addition, we analyzed the relative intensities of post-catalysis colors by capturing images. Here, the optical image information was converted into hue parameters by ImageJ software. It is consistent with the fluorescence results (Fig. 3C, inset).

3.6. Biosensing of MOF@P1/P2 in buffer or in serum

In order to study the sensitivity of the MOF@P1/P2-OPD-H₂O₂ fluorometric platform, T1/T2 (1:1) with different concentrations was employed. As shown in Fig. 4 A, 4B, T1/T2 (1:1) with the higher concentration can wonderfully turn on this fluorometric platform. The change value of the fluorescence intensity and the logarithm of the target concentration have a linear relationship in the range of 10^-9 M to 10^-6 M with the detection limit of 0.075 nM and the corresponding regression equation is F = 2830 + 305.94lg [C_t1+t2] with the correlation coefficient R² = 0.995. LOD was calculated according to this formula (3.3 × S_b/a), where S_b is the standard error of the blank sample and a is the slope of the mean value of the constructed calibration curve.

Fig. 4.

(A) Fluorescence spectra of MOF@P1/P2-OPD-H₂O₂ induced by T1/T2 with the different concentrations (10^-9 M to 10^-6 M) in buffer. (B)The dynamic response range and calibration plot (inset) of MOF@P1/P2-OPD-H₂O₂ induced by T1/T2 with the different concentrations (10^-9 M to 10^-6 M) in buffer. (C) Fluorescence spectra MOF@P1/P2-OPD-H₂O₂ induced by T1/T2 with the different concentrations (10^-9 M to 10^-6 M) in 10 % serum. (D)The dynamic response range and calibration (inset) plot of MOF@P1/P2-OPD-H₂O₂ induced by T1/T2 with the different concentrations (10^-9 M to 10^-6 M) in 10 % serum. The error bars represent the SD (n = 3).

Although the fluorometric platform shows good analytical characteristics in buffer, based on the fact that the actual specimen is more complex, this platform has to be investigated in several possible coexisting interfering substances, including metal ions (K⁺, Na⁺, Ca²⁺, Zn⁺), BSA, OMP31, BP26, Urea and serum with the different concentrations (0 %,5%,10 %,20 %). As shown in Fig. S9, compared with the fluorescence signal of the MOF@P1/P2-OPD-H₂O₂ fluorometric platform in buffer, the fluctuation of the fluorescence signal of the MOF@P1/P2-OPD-H₂O₂ fluorometric platform in various media solutions is very small, except for in 20 % serum. This implies that the influence of these interfering substances on the accuracy of platform measurement is within a reasonable range and can be basically ignored.

According to the above conclusion, T1/T2 (1:1) was dissolved into 10 % serum for the study of the practical application ability of our platform. As shown in Fig. 4 C, D, the change of fluorescence signal in 10 % serum is basically consistent with that in buffer. A good linear relationship was demonstrated in the range of 10^-9 M to 10^-6 M with the detection limit of 0.110 nM. The corresponding linear regression equation was F = 2831 + 304.7lg [C_t1+t2] with the correlation coefficient R² = 0.992 (Fig. 4D). This proves the potential applicability of the MOF@P1/P2-OPD-H₂O₂ fluorometric platform in the actual specimen. Compared with other strategies, our method also has its own advantages (Table S2).

Additionally, the stability of MOF@P1/P2 was also investigated under optimized conditions. As shown in Fig. S10, after 6 days, the fluorescence signal intensity produced through the catalysis of T1/T2-induced MOF@P1/P2 remained basically unchanged, indicating that the method is relatively stable and reliable within a week. Although the following signal has obvious fluctuation, it is still within the allowable error range.

3.7. Analytical performance

In the design of the whole strategy, our ultimate goal is to distinguish two highly homologous viruses at the same time. The experiments based on MOF@P1/P2 only help us achieve the goal of the accurate detection of SARS-CoV-2. If we want to precisely define SARS-CoV, this weak positive result alone is not sufficient evidence. After all, the other unrelated components in the test samples may also lead to the weak positive results. For example, X is a simulated DNA fragment composed of X1 and X2. Among them, X1 has four bases matching P1, and X2 has five bases matching P2. As indicated in Fig. S11A, such sequence can also cause weak positive results in detection procedure.

Based on this fact, MOF@P1 and MOF@P2 were used as the auxiliary probe. Then, three long chains T' (T1 + T2), S' (S1 + S2), and W' (WB1 + WB2) were employed to induce MOF@P1, MOF@P2, and MOF@P1/P2 in turn. As shown in Fig. S11A, T' strongly restores the peroxidase-like activity of MOF@P1, MOF@P2, and MOF@P1/P2, while S' can only weakly turn on the activity of MOF@P1 and MOF@P1/P2. Ideally, W' cannot switch on any fluorescent platform. As shown in SDS-PAGE (Fig. S11B), the new bands appear in lanes 8, 9, 13, 18, and 19. This proves that the dynamic behavior of the binding-induced fluorescence sensor was consistent with our conjecture. Combining all the results, we found that the SARS-CoV-2 gene can be identified if three strong positive fluorescent signals are present. Obviously, two weak positives represent SARS-CoV-2 genetic. Finally, the near-blank signal is symbolic of other genes (Fig. S11B table).

4. Conclusion

A “on/off/on” MOF@P1/P2-H₂O₂-OPD system is established depending on DNA hybridization induction mechanism for the simultaneous identification of these highly homologous viruses not only in buffer but also in diluted serum. This strategy breaks through traditional thinking using multiple means and a large number of experiments, only a simple one-step experiment can distinguish our experimental model, namely SARS-CoV-2 and SARS-CoV, with high selectivity and sensitivity. This not only simplifies the detection steps and saves the cost, but also greatly improves the detection speed and strives effective time for efficient treatment. This strategy can also broaden its own application range by replacing relevant probes. Compared with other methods, this strategy is simple, efficient, and has irreplaceable advantages (Table S3).

CRediT authorship contribution statement

Wenjie Yang: Writing – original draft. Dong Li: Methodology, Formal analysis. Lili Chen: Data curation. Shuang You: Investigation, Software. Lihua Chen: Funding acquisition, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

No data was used for the research described in the article.