*2.4. Sensitivity and Specificity 2.4. Sensitivity and Specificity*

According to the optimal experimental conditions obtained from the above single-factor experiments, the sensitivity of the biosensor can be further analyzed by detecting the fluorescence intensity of different concentrations of OTA. As shown in Figure 7a, increasing OTA from 0.01 nM to 50 nM results in a gradual enhance in fluorescence signal. In addition, the fluorescence value of NMM at 608 nm is proportional to the logarithm of the OTA concentration from 0.01 nM to 0.5 nM (Figure 7b). The linear regression equation is y = 457.535lgx + 620.267 (x and y refer to OTA concentration and fluorescence intensity, respectively) with the correlation coefficient of 0.9948. Furthermore, the calculated detection limit (LOD) for OTA is 4.9 pM (according to 3σ/S rule). When the OTA concentration of the detected solution is above the linear range (i.e., 0.5 nM), by simply diluting the According to the optimal experimental conditions obtained from the above single-factor experiments, the sensitivity of the biosensor can be further analyzed by detecting the fluorescence intensity of different concentrations of OTA. As shown in Figure 7a, increasing OTA from 0.01 nM to 50 nM results in a gradual enhance in fluorescence signal. In addition, the fluorescence value of NMM at 608 nm is proportional to the logarithm of the OTA concentration from 0.01 nM to 0.5 nM (Figure 7b). The linear regression equation is y = 457.535lgx + 620.267 (x and y refer to OTA concentration and fluorescence intensity, respectively) with the correlation coefficient of 0.9948. Furthermore, the calculated detection limit (LOD) for OTA is 4.9 pM (according to 3σ/S rule). When the OTA concentration of the detected solution is above the linear range (i.e., 0.5 nM), by simply diluting the actual samples to a calculable concentration with buffer solution, the target concentration of OTA can be estimated quantitatively according to the multiple of dilution.

Compared with other proposed strategies for OTA detection (Table 1), although our method is not as sensitive as electrochemical and immunofluorescence assays, the platform is economical, convenient and fast, only takes one and a half hours from preparation to detection. In addition, this work uses non-enzyme and non-label strategies and has a lower detection limit compared with general colorimetry, fluorescence and chemiluminescence methods.

OTA detection.

*n* = 3.

wine samples.

Wheat flour

Red wine

**3. Conclusions** 

*2.5. Application in Practical Samples* 

actual samples to a calculable concentration with buffer solution, the target concentration of OTA can

**Figure 7.** (**a**) Fluorescence spectra of different concentrations of OTA. From a to j, the concentrations of OTA is 0, 0.01, 0.05, 0.1, 0.2, 0.5, 1.0, 5.0, 10, 50 nM, respectively; (**b**) Linear relationship between the fluorescence value of (**a**) at 608 nm versus logarithmic concentration of OTA. Error bars, SD, *n* = 3. **Figure 7.** (**a**) Fluorescence spectra of different concentrations of OTA. From a to j, the concentrations of OTA is 0, 0.01, 0.05, 0.1, 0.2, 0.5, 1.0, 5.0, 10, 50 nM, respectively; (**b**) Linear relationship between the fluorescence value of (**a**) at 608 nm versus logarithmic concentration of OTA. Error bars, SD, *n* = 3.


**Table 1.** Comparison of proposed OTA detection strategies and this work.

electrochemical soybean 5.2 fg mL−1 [26] immunofluorescence corn, rice, wheat 0.12 pM [50] fluorescence corn flour 30 pM [51] chemiluminescence wheat, rice, core 10.6 pM [52] chemiluminescence coffee 0.5 nM [53] electrochemical coffee 0.125 ng mL−1 [54] fluorescence wheat flour, red wine 4.9 pM this work In order to verify the specificity of the developed sensor, other mycotoxins (OTB and AFB1) were also tested under the same conditions. According to Figure 8, even if there were other mycotoxins with concentrations ten times higher than OTA, negligible changes in fluorescence intensity could be observed. Nevertheless, the presence of OTA caused significant increase in fluorescence intensity. In addition, the fluorescence signal of the mixture of OTA and other control mycotoxins was similar to that of the OTA group. The results suggest that this method possesses an outstanding specificity of OTA detection. *Toxins* **2020**, *12*, x FOR PEER REVIEW 9 of 13

fluorescence red wine 198.1 pM [24]

**Figure 8.** Specificity of the fluorescent biosensor to OTA other mycotoxins and the mixture of OTA, OTB and AFB1. The concentration of OTA is 10 nM and other mycotoxins are 100 nM. Error bars, SD, **Figure 8.** Specificity of the fluorescent biosensor to OTA other mycotoxins and the mixture of OTA, OTB and AFB1. The concentration of OTA is 10 nM and other mycotoxins are 100 nM. Error bars, SD, *n* = 3.

flour samples were between 97.9% and 105%, and the relative standard deviation (RSD) was lower than ±4.8%. In addition, the detected recovery of three wine samples was higher than 94% and the RSD in the range from 3.7% to 5.1%. The experimental results suggest that our approach may be an effective and convenient method for OTA detection in actual agricultural commodities, and may

**Table 2.** Application of fluorescent aptamer sensor for OTA determination in wheat flour and red

**Samples Added (pM) Found (pM) Recovery (%) a RSD (%) b**

a The mean of three measurements. b RSD = The relative standard deviation.

In short, combining with the target-triggered structure-switching signaling aptamer and HCR technology, we have proposed a non-enzyme and non-label fluorescence biosensing system for OTA detection, proved by computer simulations and biological experiments. The approach we developed exhibits several advantages. First, the biosensor model has specific recognition and awesome detection ability for OTA with no modification of fluorescent groups and quenching groups, making the experiments more economical and unsophisticated. Second, by conducting computer simulations to verify and optimize the experimental process, making the subsequent biological experiments more facile and effective. Third, using HCR for signal amplification rather than other proteases and complex thermal cycling processes makes the operation more convenient and controllable with an OTA detection down to 4.9 pM. Furthermore, the detection and analysis of other interfering mycotoxins and actual samples showed that the proposed sensor system possesses high specificity and practical application potential. Finally, the strategy has good universal adaptability for detecting other small molecules and proteins by skillfully designing aptamer sequence of the hairpin probe. In

1 10 10.5 105 ±4.1 2 50 48.9 97.9 ±3.3 3 100 104.8 104.8 ±4.8

1 10 9.4 94 ±3.7 2 50 48.4 96.8 ±5.1 3 100 95.7 95.7 ±4.9

provide promising strategies for improving food quality and safety.
