A ratiometric fluorescent aptamer homogeneous biosensor based on hairpin structure aptamer for AFB1 detection

On the basis of aptamer (Apt) with hairpin structure and fluorescence resonance energy transfer (FRET), a ratio fluorescent aptamer homogeneous sensor was prepared for the determination of Aflatoxin B1 (AFB1). Initially, the Apt labeled simultaneously with Cy5, BHQ2, and cDNA labeled with Cy3 were formed a double-stranded DNA through complementary base pairing. The fluorescence signal of Cy3 and Cy5 were restored and quenched respectively. Thus, the ratio change of FCy3 to FCy5 was used to realized the detection of AFB1 with wider detection range and lower limit of detection (LOD). The response of the optimized protocol for AFB1 detection was wider linear range from 0.05 ng/mL to 100 ng/mL and the LOD was 12.6 pg/mL. The sensor designed in this strategy has the advantages of simple preparation and fast signal response. It has been used for the detection of AFB1 in labeled corn and wine, and has good potential for application in real samples.


Introduction
According to the statistics, about 25% of the agricultural products are contaminated with mycotoxins during crop growth, storage or processing in the world [1]. Fungal contamination of food is a global concern, most of the toxic substances produced by pollution have strong carcinogenic effects on human beings [2]. Aflatoxin is the most important and toxic mycotoxin among many mycotoxins and AFB1 is the most toxic aflatoxin. Many crops such as peanuts, corn and grains are vulnerable to AFB1 contamination, which makes AFB1 a serious threat to food safety and human health [3]. At present, more than 100 countries and regions Min Wei wei_min80@163.com Zhiguang Suo zg_suo@163.com 1 College of Food Science and Technology, Henan Key Laboratory of Cereal and Oil Food Safety Inspection and Control, Henan University of Technology, 450001 Zhengzhou, PR China with hairpin structure labeled simultaneously with Cy5, BHQ2 and cDNA labeled with Cy3 to detect AFB1. The AFB1 aptamer with hairpin structure was used as recognition element, and the labeled BHQ2 which can quench the fluorescent dyes Cy3 and Cy5. The aptamer labeled with Cy5 and BHQ2 was paired with the cDNA labeled with Cy3 to form a double strand through complementary pairing. In the meantime, Cy3 was close to BHQ2, leading to fluorescence quenching. A high fluorescence emission of Cy5 emerged due to its distance from BHQ2. The hairpin structure would be opened in the presence of AFB1. The fluorescence intensity of Cy3 was increased due to it was far away from BHQ2, and the fluorescence intensity of Cy5 was decreased due to Cy5 approaching BHQ2. AFB1 was detected by comparing the ratio of F Cy3 to F Cy5 before and after adding AFB1.
Preparation of Cy5-Apt-BHQ2/cDNA-Cy3 sensor 20 µL cDNA (1 µM) was mixed with 18 µL Apt (1 µM) with stirring, the mixture was heated to 95℃ in PCR and kept for 5 min, then slowly cooled to 4℃, Cy5-apt-BHQ2 / cDNA-Cy3 sensor was prepared and stored at 4 ℃ until use. [8]. Li et al. reported a new fluorescence aptasensor, whcih can sensitively detect Pb [2+ [9]] . Yang et al. designed a novel fluorescence sensor based protein detection method using aptamer probe as recognition molecule and cationic conjugated polymer (CCP) as reporter molecule [10]. At present, the fluorescence aptasensors prepared by using aptamer instead of antibody has been widely used in the detection of mycotoxins, protein, cell and so on [11].
Most of the fluorescence aptasensors require a tedious and time-consuming process of embedding the probe or some recognition molecules on the surface of the material used to prepare the sensor, which will greatly increase the preparation time of the sensor. In addition, the separation and purification process involved in immobilizing probes and recognition molecules can lead to signal loss of the sensor, which in turn affects the performance of the sensor. To solve this problem, many researchers have chosen to use homogeneous sensors [[12, 13]] , which is kept in a solution from preparation to detection without any separation operation. Fluorescence resonance energy transfer (FRET) is a commonly used method for quenching fluorescent dyes in fluorescence sensors. The FRET process requires a distance between the quenching material and the fluorescence probe within 10 nm. When the quenching material and the fluorescence probe are dispersed in the system, the FRET process is weak. It almost has no effect on the signal of fluorescent probe, which is convenient for the application of homogeneous detection in fluorescent aptasensor. Lu [14] et al. provided a simple, homogeneous and sensitive strategy mediated FRET for the histone analysis. The homogeneous sensor has been widely used in the detection of mycotoxins, protein, cell [15].
Furthermore, the conventional fluorescent aptasensors were designed based on the changes of the "signal on" or "signal off" of single probe. The conventional fluorescent aptasensors were easily affected by the environmental and instrument factors in detection of the real sample, resulting in the degradation of the sensor performances [16]. Yang et al. [17] prepared a simple and rapid fluorescence homogeneous sensor for AFB1 detection by using the adaptation system of BHQ2-labeled complementary chain and Cy3 labeled fluorescent probe. However, the signal ratio is not sensitive to changes affected by environmental and instrumental. To improve the stability and accuracy of the sensor, the ratiometric fluorescent aptasensor was selects two signal probes with different excitation and emission wavelengths. The final signals are determined by the fluorescence signal ratio of the two probes [[18, 19]] . Therefore, the ratiometric fluorescent aptasensor using two signal probes has become a new research hotspot.
Herein, a simple, rapid and sensitive ratio homogeneous fluorescent aptasensor was constructed by using aptamer to F Cy5 before and after adding AFB1, the labeled samples were detected. All the test data were performed three times.
The contaminated feed (corn peanut meal) was ground to powder (1.0 g) and the extraction solvent was added after stirring for 30 min. The supernatant was treated to detect AFB1. All results were compared by HPLC. All the test data were performed three times.

Results and discussion
The design principle of the aptasensor for AFB1 detection Based on the hairpin structure of Apt and the characteristic that BHQ2 can simultaneously quench the fluorescence probe Cy3 and Cy5, a new fluorescence aptamer homogeneous sensor was designed for the detection of AFB1. The detection principle was illustrated in Fig. 1. The Apt labeled with Cy5 and BHQ2 and cDNA labeled with Cy3 form a double-stranded DNA through complementary base pairing. At this moment, Cy3 was close to BHQ2, resulting in fluorescence quenching. Due to the opening of the hairpin structure, Cy5 was far away from BHQ2 leading to the recovery of the fluorescence signal. The double-stranded DNA structure will be disintegrated in the presence of AFB1, resulting the removal of Cy3 and the close of Cy5 with BHQ2. At this time, a low fluorescence emission of Cy3 and a high fluorescence emission of Cy5 be generated due to the quenching effect of BHQ2. By comparing the ratio change of F Cy3 to F Cy5 before and after adding AFB1, the detection of AFB1 was realized. Therefore, the ratiometric fluorescent homogeneous aptasensor can be used for AFB1 detection theoretically.

Detection of AFB1
10 µL AFB1 (200 ng/mL) was mixed with 38 µL of prepared Cy5-Apt-BHQ2/cDNA-Cy3 sensor solution, and adding 20 mmol/L Tris-HCl (pH 7.4) buffer solution to 200 µL. Then the mixed solution was incubated at 37 ℃ with a water bath thermostatic oscillator for 40 min. In this paper, excitation wavelength scanning method was used to detect fluorescence intensity. The fluorescence intensity was measured at the excitation wavelength scanning mode by Hitachi F-7100 fluorescence spectrophotometer (Japan). All testing process were all carried out at room temperature. The fluorescence cell was a quartz fluorescence cell with a slit of 2 mm, and the emission spectrum of Cy3 probe was recorded in the excitation wavelength of 530 nm and in the wavelength range of 550-650 nm. The emission spectra of the Cy5 probe were recorded in the excitation wavelength of 635 nm and the wavelength range of 650-750 nm. The scanning voltage was 950 V, and the excitation and emission slit widths were set to 5 nm. All measurements were made at room temperature. Fluorescence spectrophotometer was used to record the fluorescence spectra of mixed solution before and after adding AFB1 at excitation wavelength of 530 nm (Cy3) and 635 nm (Cy5), the ratio value of fluorescence intensity of Cy3 (F Cy3 ) to fluorescence intensity of Cy5 (F Cy5 ) was compared before and after AFB1 added.All the test data were performed three times.

Application of the fluorescence sensor in real samples
To investigate the feasibility of the designed sensor, the spiked corn flour samples were tested. The processing method of corn flour referred to Seok's [18] processing method. 1 g corn meal sample was put into a 10 mL flask, and adding 1 mL analytically pure methanol was mixed with AFB1 in different known concentrations, eventually. The mixture was left in a fume hood for 12 h to completely evaporate the solvent, then extracted at room temperature with methanol/water (2:8, v/v) for 15 min. After centrifugation at 10,000 rpm for 15 min, the extract was filtered through a 0.45 mm injection filter. The filtrate was diluted with 50 mM Tris-HCl (pH 7.4) to the concentration of 0.1 ng/mL, 1 ng/mL, and 10 ng/mL. The prepared sensor was used to detect the spiked real samples, and the fluorescence spectra of Cy3 and Cy5 were recorded after the addition of the fluorescence sensor to the real samples.
After removing the foam from the wine samples by nitrogen blowing, the spiked wine samples were detected. The toxin concentration was diluted with Tris-HCl (pH 7.4) until 0.1 ng/mL, 1 ng/mL and 10 ng/mL with different known concentrations of AFB1. By comparing the ratio of F Cy3 Fig. 1 Preparation of Cy5-Apt-BHQ 2 /cDNA-Cy3 sensor and its detection principle diagram of AFB1 intensity of the sensor showed that the ratio of F Cy3 /F Cy5 was 1.1 in the absence of AFB1 (Red line). After adding 10 ng/mL AFB1 for incubation for 40 min, the value of the F Cy3 /F Cy5 was increased 6 times compared to without AFB1, indicating the feasibility of the prepared sensor for AFB1 detection (Blue line). In the presence of AFB1, the doublestranded DNA structure cleaved, leading to the separation of Cy3 from BHQ2 and the proximity of Cy5 from BHQ2, respectively. Due to the fluorescence quenching effect of BHQ2, the ratio of F Cy3 /F Cy5 increased to 6.6 when adding 10 ng/mL AFB1. The results were shown that the ratiometric fluorescent aptamer homogeneous sensor can be used to detect AFB1, sensitively.

Optimization of the experiment parameter
In order to get the optimal performance of the prepared sensor, the ratio value between Apt and cDNA should be optimized. The background signal will be affect by the change of the ratio value Apt to cDNA Because changing the ratio value of Apt to cDNA would affect the ratio of the initial The feasibility of the ratiometric fluorescent aptamer homogeneous sensor The feasibility of the prepared Cy5-Apt-BHQ2/cDNA-Cy3 sensor was analyzed in the absence and presence of AFB1. The result was shown in Fig. 2. The initial fluorescence Fig. 3 Effect of Apt and cDNA dosage ratio on Cy5-Apt-BHQ2 / cDNA-Cy3 sensor (AFB1 concentration is 10 ng/mL)

Fig. 2
The fluorescence spectra of Cy5-Apt-BHQ2 / cDNA-Cy3 sensor before and after AFB1 incubation (The concentration of AFB1 is 10 ng/mL) aptasensor with a single interfering toxin for 40 min, the ratio value of fluorescence intensity of the two dyes (F Cy3 /F Cy5 ), remained unchanged even though its concentration was 10 times higher than that of AFB1. However, as long as AFB1 was present, the value of F Cy3 /F Cy5 of fluorescence intensity could be significantly increased after incubation for detection, whether AFB1 was existed alone or AFB1 was mixed with interfering toxins, indicating that the designed aptasensor had good specificity.
was used to reflect the optimization results [18]. As shown in Fig. 3, (F cy3 /F cy5 )/(F cy3 /Fc y5 ) 0 was growing with the increasing the proportion between Apt and cDNA from 0.7 to 0.9. When the proportion between Apt and cDNA was greater than 0.9, (F cy3 /F cy5 )/(F cy3 /F cy5 ) 0 started to decrease. This was because the double-strand DNA formed by Apt and cDNA has reached a saturation state, and the redundant Apt reduces the sensitivity of the sensor and the signal change was no longer obvious. Therefore, 0.9 was selected as the optimal ratio value between Apt and cDNA.

Detection of AFB1
Under the optimal experimental parameters, F Cy3 /F Cy5 was measured by analyzing AFB1 solution with different concentrations. The detected spectrogram was shown in Fig. 4(A), the fluorescence intensity of the fluorescence sensor Cy3 (F cy3 ) increased and the fluorescence intensity of Cy5 (F cy5 ) decreased with the increasing concentration of AFB1 added (0, 0.05, 0.1, 0.5, 1, 10, 50, 100, 500 ng/ mL). A good linear relationship was obtained between 0.05 ng/ mL and 100 ng/ mL (Fig. 4B), the linear equation was F Cy3 /F Cy5 =2.34LgC AFB1 +4.60(R 2 =0.994). The LOD was calculated to be 12.6 pg/mL (3σ rule). The result indicated the prepared ratiometric homogeneous fluorescent aptasensor had a wider linear range and lower detection limit for AFB1 detection.
In order to study the specificity of the prepared fluorescence aptasensor to different toxins. Under the same experimental condition, 10 ng/mL of AFB1, 100 ng/mL of interference toxins such as OTA, FB1, DON, ZEN and mixture of AFB1 with all interference toxins were detected and results were shown in Fig. 5. After incubating the prepared  HPLC. The aptasensor designed in this paper is simple in preparation, fast in signal response, which has the application prospect of realizing fast detection of AFB1 on site.
The detection of the designed aptasensor was compared with other AFB1 aptasensors, and the results are shown in Table 1. It can be seen that the aptasensor designed in this experiment had a good linear range and a lower LOD compared with most sensors. This result indicated that the dualsignal change of the ratio sensor can effectively improve the sensitivity of the fluorescence aptasensor, and the homogeneous detection method improved the stability of the sensor.

Detection of the real samples
In order to further verify the accuracy and practicability of the ratio homogeneous fluorescent apatasensor, the spiked corn flour samples and spiked wine samples were tested, and the scalars were 0.1 ng/mL, 1 ng/mL, and 10 ng/mL, respectively. The detection results were shown in Table 2. The recovery of the detection results were 95-110% and 93-108%, respectively. The relative standard deviation (RSD) were not more than 10%. HPLC and the prepared fluorescence sensor were used to detect the contaminated feed. The results showed that the prepared sensor had certain accuracy and practicability which can be used as an alternative to traditional detection methods.

Conclusions
In summary, combining Apt with hairpin structure labeled with Cy5 and BHQ2 and cDNA labeled with Cy3 to detect AFB1, a simple, rapid and sensitive ratio homogenous fluorescent aptasensor was constructed. The ratio value of fluorescence intensity between Cy3 and Cy5 (F Cy3 /F Cy5 ), and LgC AFB1 showed a good linear relationship within the concentration range of AFB1 from 0.05 ng/ mL to 100 ng/mL, and the LOD was 12.6 pg/mL. The recovery rates of AFB1 in spiked corn meal samples and wine were 92-110% and 88-112%, respectively. Furthermore, the adequate determination of AFB1 in corn, wine samples designed, herein confirmed the applicability of this strategy compared with