A Metal Organic Framework-Based Light Scattering ELISA for the Detection of Staphylococcal Enterotoxin B

: Enzyme-linked immunosorbent assay (ELISA) is one of the most commonly used method for the detection of staphylococcal enterotoxin B (SEB), the main protein toxin causing staphylococcal food poisoning. However, the traditional ELISA reaction needs to be stopped by sulfuric acid to obtain stable colorimetric signal, and it is easily inﬂuenced by a colored sample. In order to address this problem, a new ELISA method using zeolite imidazolate skeleton-8 metal-organic framework (ZIF-8 MOF) as a light scattering (LS) reporter for SEB detection was developed in this work. ZIF-8 MOF has the characteristics of high porosity, large speciﬁc surface area, clear pore structure, and adjustable size, which is one of the most representative MOFs constructed from Zn 2+ and 2-methylimidazole (2-mIM). The 2-mIM ligand of ZIF-8 exhibited antioxidant activity and can strongly react with H 2 O 2 , which could destroy the structure of ZIF-8, resulting in the obvious decrease in LS intensity. We combined this speciﬁc reaction with the sandwich immune reaction to construct the LS ELISA method for the successful detection of SEB. This method is more reliable than commercial tests kits for the detection of colored samples, and it is simple, sensitive, and selective, and has great potential in the detection of other toxins by simply changing the corresponding recognition units.


Introduction
Staphylococcal enterotoxin B (SEB) is one of the main protein toxins causing staphylococcal food poisoning [1]. It is resistant to high temperature, acid, and alkali, and easy to formulate as an aerosol. It has been not only linked to food poisoning but also classified as a potential Class B biological warfare agent by the US Centers for Disease Control and Prevention (CDC). Therefore, it is highly important to develop sensitive method to detect SEB. The traditional SEB detection methods include chromatography [2], mass spectrometry [3,4], spectral method [5], electrochemical method [6], Raman spectroscopy [7], enzyme-linked immunosorbent assay (ELISA) [8], and so on. At present, the detection limit of SEB reported in the literature has been as low as 4.29 fg/mL [9]. Among these methods, ELISA has attracted a lot of attention because of its high sensitivity, throughput, and specificity. Generally, the ELISA method for SEB detection depends on the double-antibody sandwich system and the colorimetric signal is produced by the specific reaction between 3,3',5,5'-Tetramethylbenzidine (TMB) and H 2 O 2 catalyzed by horseradish peroxidase (HRP) that is labelled to the secondary antibody. Although this reaction is highly efficient, it
The light scattering signals were measured with an F-2500 fluorescence spectrophotometer (Hitachi, Japan) by simultaneously scanning the excitation and emission monochromator of the spectrofluorometer with same starting excitation and emission wavelength (namely, λ em = λ ex ). Vacuum drying oven (DZF-6020, Keelrein, Shanghai, China) was used to dry ZIF-8. The scanning electron microscope (S-4800, Hitachi, Tokyo, Japan) was utilized to measure the morphology of ZIF-8. Dark field microscopy imaging was obtained using a BX51 light microscope (Olympus, Tokyo, Japan) equipped with a dark field con-denser (U-DCW, Olympus, Japan). The Fourier transform infrared spectrometer Prestige-21 (FTIR, Shimadzu, Kyoto, Japan) was used to test the infrared spectra of materials. X-ray diffractometer D8 ADVANCE (Brock, Germany) was used to test powder X-ray diffraction patterns of samples in a scanning range of 5-50 θ at a scanning speed of 1.5 • /min.

Synthesis of ZIF-8
The synthesis method of ZIF-8 was according to the reported literature with little modification [34]. First, 1.32 mol/L 2-mIM and 24 mmol/L Zn(NO 3 ) 2 ·6H 2 O were prepared using water as solvent using ultrasonic dissolution for 10 min. Next, 500 µL of 1.32 mol/L 2-mIM was added to the glass sample bottle and pre-stirred for 30 s, and then, 500 µL of 24 mmol/L Zn(NO 3 ) 2 ·6H 2 O was added. After being stirred for 5 min, the mixture was kept stagnant for 1 h. After washing twice with ethanol, the as-prepared ZIF-8 was stored at 4 • C for further use.

Dark Field Light Scattering Imaging of the Reaction between ZIF-8 and H 2 O 2
The reaction process of ZIF-8 and H 2 O 2 was monitored using a dark field light scattering microscope. The cationic slides were rinsed with ultrapure water and dried with nitrogen. Then, 100 µL of 50 µg/mL ZIF-8 ethanol solution was added to the surface of the cationic glass slide and was allowed to stand at room temperature for 15 min to facilitate ZIF-8 deposition on the surface of the cationic glass slide. The excess solution on the surface of the slides was then rinsed with ultrapure water and blow-dried with nitrogen. In order to monitor the reaction between ZIF-8 and H 2 O 2 in real-time, 200 µL ethanol solution was added between the slides and cover slides to obtain the original dark field image of ZIF-8. Next, ethanol solution was removed and 200 µL of 3% H 2 O 2 solution was added. Then, dark field microscopic images of the reaction process between ZIF-8 and H 2 O 2 were collected.

Detection of SEB
The light scattering ELISA method was developed based on the ELISA tests kits. A 50 µL volume of different concentrations of SEB were added to the 96-well plate premodified with Ab1 of SEB. Then, 100 µL of HPR-Ab2 was added and incubated at 37 • C for 1 h. Each well of the 96-well plate was washed with 350 µL washing buffer five times to remove the uncombined SEB and HRP-Ab. The above steps were consistent with the operation of the ELISA test kits. Then, 100 µL of 6% H 2 O 2 was added to each well and incubated for 15 min in a 37 • C incubator. The residual 100 µL H 2 O 2 of the reaction was transferred to the pre-ultrasonicated, dispersed 100 µL of 100 µg/mL ZIF-8 solution and reacted at room temperature. Finally, the LS spectrum was obtained using F-2500.

Principle of the SEB Detection
The SEB detection principle is shown in Figure 1. Firstly, ZIF-8 was synthesized using 2-mIM as the ligand and Zn 2+ as the central ion. As the ligand 2-mIM could be oxidized by H 2 O 2 , the structure of ZIF-8 could be destroyed by H 2 O 2 , and the LS intensity was reduced. Subsequently, we combined this reaction with SEB antibody immune sandwich assay to construct a new SEB detection method. When SEB was added, an Ab1/SEB/Ab2-HRP sandwich immune structure formed in the 96-well plate. After the addition of H 2 O 2 into the 96-well plate, the H 2 O 2 was consumed by the HRP on Ab2. Then, the residual H 2 O 2 after the reaction was transferred to the ZIF-8 solution. Since H 2 O 2 was consumed by HRP, the oxidation ability of H 2 O 2 toward 2-mIM was weak. And ZIF-8 maintained an intact structure, showing high LS intensity. When there was no SEB in the system, the sandwich immune structure cannot be formed, and the added H 2 O 2 will not be consumed by HRP. In this situation, the amount of H 2 O 2 was large, and the oxidation capacity of H 2 O 2 toward 2-mIM was strong, leading to the damage of the structure of ZIF-8 and resulting in the decreased LS intensity. Thus, quantitative detection of SEB can be achieved by the obviously changed LS intensity of ZIF-8 before and after adding SEB. consumed by HRP, the oxidation ability of H2O2 toward 2−mIM was weak. And ZIF−8 maintained an intact structure, showing high LS intensity. When there was no SEB in th system, the sandwich immune structure cannot be formed, and the added H2O2 will no be consumed by HRP. In this situation, the amount of H2O2 was large, and the oxidation capacity of H2O2 toward 2−mIM was strong, leading to the damage of the structure o ZIF−8 and resulting in the decreased LS intensity. Thus, quantitative detection of SEB can be achieved by the obviously changed LS intensity of ZIF-8 before and after adding SEB.

Characterization of ZIF-8 and Its Reaction with H 2 O 2
The as-prepared ZIF-8 was characterized using Fourier transform infrared (FTIR) and X-ray diffraction (XRD) measurements. In the absorption FTIR spectra of ZIF-8, peaks at 3138 cm −1 and 2933 cm −1 belong to the stretching vibration of C−H in methyl and imidazole rings, respectively ( Figure 2A). The XRD studies ( Figure 2B consumed by HRP, the oxidation ability of H2O2 toward 2−mIM was weak. And ZIF−8 maintained an intact structure, showing high LS intensity. When there was no SEB in the system, the sandwich immune structure cannot be formed, and the added H2O2 will no be consumed by HRP. In this situation, the amount of H2O2 was large, and the oxidation capacity of H2O2 toward 2−mIM was strong, leading to the damage of the structure o ZIF−8 and resulting in the decreased LS intensity. Thus, quantitative detection of SEB can be achieved by the obviously changed LS intensity of ZIF-8 before and after adding SEB.

Characterization of ZIF−8 and Its Reaction with H2O2
The as−prepared ZIF−8 was characterized using Fourier transform infrared (FTIR and X−ray diffraction (XRD) measurements. In the absorption FTIR spectra of ZIF−8 peaks at 3138 cm −1 and 2933 cm −1 belong to the stretching vibration of C−H in methyl and imidazole rings, respectively ( Figure 2A). The XRD studies ( Figure 2B) showed that ZIF−8 had multiple diffraction peaks, mainly consisting of 2θ = 7.3°, 10.2°, 12.5°, 14.5°, 16.2° 17.8°, 21.8°, etc. The above angles corresponded to planes (011), (002), (112), (022), (013) (222), and (114), respectively, with the sharpest peak at 2θ = 7.3 °, indicating that the syn thesized ZIF-8 has a higher crystallinity. These results confirmed that the ZIF−8 with a higher crystallinity was synthesized successfully.  The morphology of ZIF-8 before and after the reaction with different concentrations of H 2 O 2 was further characterized using scanning electron microscopy (SEM). As shown in Figure 2, ZIF-8 presented a regular dodecahedron structure ( Figure 3A). After the reaction with 0.1% H 2 O 2 , the regular morphology of ZIF-8 was destroyed ( Figure 3B). With the increase in H 2 O 2 concentration, the degree of destruction of ZIF-8 also increased ( Figure 3C,D). These results directly confirmed that H 2 O 2 can destroy the structure of ZIF-8.
Because H2O2 could destroy the structure of ZIF−8, so the LS intensity decreased significantly.
The morphology of ZIF−8 before and after the reaction with different concentrations of H2O2 was further characterized using scanning electron microscopy (SEM). As shown in Figure 2, ZIF−8 presented a regular dodecahedron structure ( Figure 3A). After the reaction with 0.1% H2O2, the regular morphology of ZIF−8 was destroyed ( Figure 3B). With the increase in H2O2 concentration, the degree of destruction of ZIF−8 also increased (Figure 3C, D). These results directly confirmed that H2O2 can destroy the structure of ZIF−8. We also investigated the dark field microscopic (DFM) images of ZIF−8 before and after the reaction with H2O2. As shown in Figure 4A, the DFM image of ZIF−8 showed an obvious doughnut-shaped image, and the LS intensity at the edge of the DFM image of ZIF−8 was very strong. However, after the reaction with 30% H2O2, the doughnut-shaped image changed into the solid scattering light spot, and the LS intensity decreased significantly ( Figure 4B). Then, in situ DFM imaging was conducted for real−time monitoring of the reaction between ZIF−8 and H2O2. As shown in Figure 4C, with the addition of H2O2 the size and intensity of dark field scattering spots of ZIF−8 gradually decreased with the extension of incubation time. These results further proved that H2O2 could destroy ZIF−8 resulting in the reduction of the LS intensity of ZIF−8. We also investigated the dark field microscopic (DFM) images of ZIF-8 before and after the reaction with H 2 O 2 . As shown in Figure 4A, the DFM image of ZIF-8 showed an obvious doughnut-shaped image, and the LS intensity at the edge of the DFM image of ZIF-8 was very strong. However, after the reaction with 30% H 2 O 2 , the doughnut-shaped image changed into the solid scattering light spot, and the LS intensity decreased significantly ( Figure 4B). Then, in situ DFM imaging was conducted for real-time monitoring of the reaction between ZIF-8 and H 2 O 2 . As shown in Figure 4C, with the addition of H 2 O 2 , the size and intensity of dark field scattering spots of ZIF-8 gradually decreased with the extension of incubation time. These results further proved that H 2 O 2 could destroy ZIF-8, resulting in the reduction of the LS intensity of ZIF-8.

LS Spectral Characteristics of ZIF−8 for SEB Detection
To investigate whether the proposed method could be applied for the detection o SEB, the LS spectral characteristics of ZIF−8 before and after the addition of SEB was meas ured. As shown in Figure 5A, in the presence of SEB, the LS intensity of ZIF−8 was signif icantly higher than that of the control group before adding SEB. The Ab1/SEB/Ab2−HRP sandwich immune complex structure was formed in the 96−well plate after the addition of SEB. Although HRP can react with H2O2 to produce OH•, only the remaining H2O2 ca react with ZIF−8 because of the short life of OH• [35]. Compared with the system withou SEB, the H2O2 content decreased and the oxidation capacity of ZIF−8 ligand 2−mIM de creased with the addition of SEB, so ZIF−8 maintained higher scattering intensity. There fore, SEB could be detected by the significantly enhanced LS intensity at 378 nm (Figur 5B).

Optimization of the SEB Detection Conditions
In order to obtain excellent analytical performance, some important experimenta conditions were optimized. We first optimized the reaction time of ZIF−8 with H2O2. A shown in Figure 6A, when the reaction time was 70 min, the light scattering signal differ ence (∆I), which was calculated by subtracting the scattering intensity of ZIF−8 before add ing SEB from that after the addition of SEB, reached the maximum. Thus, 70 min wa

LS Spectral Characteristics of ZIF-8 for SEB Detection
To investigate whether the proposed method could be applied for the detection of SEB, the LS spectral characteristics of ZIF-8 before and after the addition of SEB was measured. As shown in Figure 5A, in the presence of SEB, the LS intensity of ZIF-8 was significantly higher than that of the control group before adding SEB. The Ab1/SEB/Ab2-HRP sandwich immune complex structure was formed in the 96-well plate after the addition of SEB. Although HRP can react with H 2 O 2 to produce OH·, only the remaining H 2 O 2 can react with ZIF-8 because of the short life of OH· [35]. Compared with the system without SEB, the H 2 O 2 content decreased and the oxidation capacity of ZIF-8 ligand 2-mIM decreased with the addition of SEB, so ZIF-8 maintained higher scattering intensity. Therefore, SEB could be detected by the significantly enhanced LS intensity at 378 nm ( Figure 5B).

LS Spectral Characteristics of ZIF−8 for SEB Detection
To investigate whether the proposed method could be applied for the detection of SEB, the LS spectral characteristics of ZIF−8 before and after the addition of SEB was measured. As shown in Figure 5A, in the presence of SEB, the LS intensity of ZIF−8 was significantly higher than that of the control group before adding SEB. The Ab1/SEB/Ab2−HRP sandwich immune complex structure was formed in the 96−well plate after the addition of SEB. Although HRP can react with H2O2 to produce OH•, only the remaining H2O2 can react with ZIF−8 because of the short life of OH• [35]. Compared with the system without SEB, the H2O2 content decreased and the oxidation capacity of ZIF−8 ligand 2−mIM decreased with the addition of SEB, so ZIF−8 maintained higher scattering intensity. Therefore, SEB could be detected by the significantly enhanced LS intensity at 378 nm ( Figure  5B).

Optimization of the SEB Detection Conditions
In order to obtain excellent analytical performance, some important experimental conditions were optimized. We first optimized the reaction time of ZIF−8 with H2O2. As shown in Figure 6A, when the reaction time was 70 min, the light scattering signal difference (∆I), which was calculated by subtracting the scattering intensity of ZIF−8 before adding SEB from that after the addition of SEB, reached the maximum. Thus, 70 min was selected as the best reaction time. Subsequently, we optimized the concentration of H2O2. As shown in Figure 6B, when H2O2 concentration was 3%, ∆I was the largest, and 3% H2O2

Optimization of the SEB Detection Conditions
In order to obtain excellent analytical performance, some important experimental conditions were optimized. We first optimized the reaction time of ZIF-8 with H 2 O 2 . As shown in Figure 6A, when the reaction time was 70 min, the light scattering signal difference (∆I), which was calculated by subtracting the scattering intensity of ZIF-8 before adding SEB from that after the addition of SEB, reached the maximum. Thus, 70 min was selected as the best reaction time. Subsequently, we optimized the concentration of H 2 O 2 . As shown in Figure 6B, when H 2 O 2 concentration was 3%, ∆I was the largest, and 3% H 2 O 2 was selected as the best reaction condition. Finally, we optimized the concentration of ZIF-8. The low concentration of ZIF-8 will result in low LS intensity and small signal changes. However, high concentrations of ZIF-8 can also affect the detection sensitivity. As shown in Figure 6C, when ZIF-8 concentration was 10-50 µg/mL, ∆I showed an upward trend. When ZIF-8 concentration was 50-90 µg/mL, ∆I showed a downward trend. Therefore, 50 µg/mL was finally selected as the optimal concentration of ZIF-8.
Chemosensors 2023, 11, x FOR PEER REVIEW was selected as the best reaction condition. Finally, we optimized the concentr ZIF−8. The low concentration of ZIF−8 will result in low LS intensity and sma changes. However, high concentrations of ZIF−8 can also affect the detection sen As shown in Figure 6C, when ZIF−8 concentration was 10−50 µg/mL, ∆I showed ward trend. When ZIF−8 concentration was 50-90 µg/mL, ∆I showed a downwar Therefore, 50 µg/mL was finally selected as the optimal concentration of ZIF−8.

Analytical Performance of SEB Detection
Under optimal experimental conditions, the sensitivity of the method was gated. As shown in Figure 7, the LS intensity variation and SEB concentration sh good linear relationship within the range of 7−500 ng/mL. The linear regression e was ∆I = 2.982 cSEB + 33.575 (R 2 = 0.998), and the limit of detection (LOD, 3 σ/k) w ng/mL. This method is more sensitive to the response of SEB, as shown in Tabl detection limit of this method is still comparable to some reported methods.
Then, we examined the selectivity of this platform for the detection of SEB. Un same experimental conditions, we compared the response signals of this method toxin B1 (AFB1), hemoglobin (HGB), ochratoxin A (OTA), staphylococcal entero (SEA), carcinoembryonic antigen (CEA), bovine serum albumin (BSA), and SEB. A in Figure 8A, the LS intensity difference produced by SEA, AFB1, OTA, BSA, H CEA was very low, and only the presence of SEB could lead to an obvious valu These experimental results illustrated that this method had a good selectivity for tection of SEB.

Analytical Performance of SEB Detection
Under optimal experimental conditions, the sensitivity of the method was investigated. As shown in Figure 7, the LS intensity variation and SEB concentration showed a good linear relationship within the range of 7-500 ng/mL. The linear regression equation was ∆I = 2.982 c SEB + 33.575 (R 2 = 0.998), and the limit of detection (LOD, 3 σ/k) was 0.69 ng/mL. This method is more sensitive to the response of SEB, as shown in Table 1. The detection limit of this method is still comparable to some reported methods.  Since various potential substances in actual samples may affect the detection results, we investigated the influence of HCO3 − , I − , K + , Na + , Zn 2+ , Ca 2+ , Ba 2+ , Fe 2+ , Fe 3+ , Al 3+ , glucose, sucrose, Congo Red, and other substances on SEB detection. The concentrations of interferences (2500 ng/mL) were 10 times than that of SEB (250 ng/mL). The experimental results are shown in Figure 8B. The potential interfering substances had no significant influence on the detection of SEB.
In addition, we focus on exploring the influence of colored samples on colorimetric materials. As shown in the Figure S1, we washed the 96−well plate for 0 to 5 times, and then tested the SEB using both SEB ELISA test kits and light scattering ELISA. Under the linear range, the concentration of SEB used by the SEB ELISA test kits is 5 pg/mL, the concentration of Congo Red is 50 pg/mL, the concentration of SEB used by the light scattering ELISA is 250 ng/mL, and the concentration of Congo Red is 2500 ng/mL. After several plate washing operations, the detection results obtained using the SEB ELISA detection kits were all higher than the actual concentration ( Figure S2A); The detection results obtained by the light scattering ELISA method after the fourth plate washing operation Figure 7. Linearity of SEB detection. The inset is the linear relationship between ∆I and SEB concentration (7 ng/mL to 500 ng/mL). Experimental conditions: the concentration of SEB was from 7 ng/mL to 1000 ng/mL; reaction time, 70 min; ZIF-8, 50 µg/mL; H 2 O 2 , 3%. The error bars represent the standard deviation of three replicates. The black line represents that ∆I increases with the increase of SEB concentration until reaching the plateau period. The black line represents the linear range. Enzyme-linked immunosorbent assay 0.38 ng/mL [37] Raman spectroscopy and chemometric methods 0.2 ng/L [38] Light scattering ELISA 0.69 ng/mL This work Then, we examined the selectivity of this platform for the detection of SEB. Under the same experimental conditions, we compared the response signals of this method to aflatoxin B1 (AFB1), hemoglobin (HGB), ochratoxin A (OTA), staphylococcal enterotoxin A (SEA), carcinoembryonic antigen (CEA), bovine serum albumin (BSA), and SEB. As shown in Figure 8A, the LS intensity difference produced by SEA, AFB1, OTA, BSA, HGB, and CEA was very low, and only the presence of SEB could lead to an obvious value of ∆I. These experimental results illustrated that this method had a good selectivity for the detection of SEB.
Since various potential substances in actual samples may affect the detection results, we investigated the influence of HCO 3 − , I − , K + , Na + , Zn 2+ , Ca 2+ , Ba 2+ , Fe 2+ , Fe 3+ , Al 3+ , glucose, sucrose, Congo Red, and other substances on SEB detection. The concentrations of interferences (2500 ng/mL) were 10 times than that of SEB (250 ng/mL). The experimental results are shown in Figure 8B. The potential interfering substances had no significant influence on the detection of SEB. Figure 7. Linearity of SEB detection. The inset is the linear relationship between ∆I and SE tration (7 ng/mL to 500 ng/mL). Experimental conditions: the concentration of SEB was from to 1000 ng/mL; reaction time, 70 min; ZIF−8, 50 µg/mL; H2O2, 3%. The error bars represent t ard deviation of three replicates. The black line represents that ∆I increases with the increa concentration until reaching the plateau period. The black line represents the linear range Since various potential substances in actual samples may affect the detection we investigated the influence of HCO3 − , I − , K + , Na + , Zn 2+ , Ca 2+ , Ba 2+ , Fe 2+ , Fe 3+ , Al 3+ , sucrose, Congo Red, and other substances on SEB detection. The concentrations ferences (2500 ng/mL) were 10 times than that of SEB (250 ng/mL). The experim sults are shown in Figure 8B. The potential interfering substances had no signifi fluence on the detection of SEB.
In addition, we focus on exploring the influence of colored samples on colo materials. As shown in the Figure S1, we washed the 96−well plate for 0 to 5 tim then tested the SEB using both SEB ELISA test kits and light scattering ELISA. U linear range, the concentration of SEB used by the SEB ELISA test kits is 5 pg concentration of Congo Red is 50 pg/mL, the concentration of SEB used by the li tering ELISA is 250 ng/mL, and the concentration of Congo Red is 2500 ng/mL. A eral plate washing operations, the detection results obtained using the SEB ELIS tion kits were all higher than the actual concentration ( Figure S2A); The detectio obtained by the light scattering ELISA method after the fourth plate washing o In addition, we focus on exploring the influence of colored samples on colorimetric materials. As shown in the Figure S1, we washed the 96-well plate for 0 to 5 times, and then tested the SEB using both SEB ELISA test kits and light scattering ELISA. Under the linear range, the concentration of SEB used by the SEB ELISA test kits is 5 pg/mL, the concentration of Congo Red is 50 pg/mL, the concentration of SEB used by the light scattering ELISA is 250 ng/mL, and the concentration of Congo Red is 2500 ng/mL. After several plate washing operations, the detection results obtained using the SEB ELISA detection kits were all higher than the actual concentration ( Figure S2A); The detection results obtained by the light scattering ELISA method after the fourth plate washing operation were consistent with the actual concentration ( Figure S2B). Therefore, light scattering ELISA has greater advantages in the detection of colored samples.

SEB Detection in Complex Samples
We carried out the standard recovery experiment of SEB in orange juice, fresh milk, and skim milk powder to test whether this method could be applied in the real sample detection. Firstly, three different concentrations of SEB were added to the orange juice and detected by our method. The accuracy of this method for SEB detection was evaluated by Recovery. As shown in Table 2, the spiked recoveries of SEB in orange juice were 91.9-106.2%, and the relative standard deviations (RSD) were 2.10-8.21%, indicating that this method could be used for the detection of SEB in orange juice. Furthermore, the concentrations of SEB in fresh milk and skim milk powder were tested. Dissolve 200 mg skim milk powder in 1 mL sample diluent as its initial concentration. The recovery of SEB detection in fresh milk was 92.0-109.6% with RSD of 0.16-9.25%, and in skim milk powder, it was 90.2-107.0% with RSD of 1.45-6.85%. These results indicated that this method has good accuracy in the detection of SEB in different food substrates.

Staphylococcal Enterotoxin A (SEA) Detection
Finally, in order to investigate the universality of this method, we applied it for the detection of staphylococcal enterotoxin A (SEA) by changing the corresponding recognition units. Quantitative detection of SEA can be achieved by the obviously changed LS intensity of ZIF-8 before and after adding SEA. As can be seen from Figure 9A, ∆I gradually increases with the increase in concentration of SEA. The experimental results showed that there was a linear relationship between the ∆I with the concentration of SEA at the range of 30 ng/mL-70 ng/mL. The linear equation was ∆I = 29.042 c SEA − 442.509 (R 2 = 0.991), and the limit of detection (LOD, 3 σ/k) was 1.58 ng/mL.
Chemosensors 2023, 11, x FOR PEER REVIEW 10 of induced a high ΔI value ( Figure 9B), indicating that the method could be used for th specific detection of SEA. These results confirmed that this method is universal. Figure 9. Linearity of SEA detection (A). The inset is the linear relationship between ∆I and SE concentration (30 ng/mL to 70 ng/mL). Selectivity of the proposed strategy (B). Experimental cond tions: the concentration of SEB was from 30 ng/mL to 70 ng/mL; the concentration of SEB, AFB BSA, and HGB were 50 ng/mL; reaction time, 70 min; ZIF−8, 50 µg/mL; H2O2, 3%. The error ba represent the standard deviation of three replicates.

Conclusions
ZIF−8 MOF was synthesized and employed as a new LS reporter of the ELISA for th detection of SEB based on the unique reaction of ZIF−8 MOF and H2O2. This develope LS ELISA method has several advantages. Firstly, the synthesis of ZIF−8 is easy. Secondl the structure of the synthesized ZIF−8 was uniform, and its LS signal was strong and sta ble. Thirdly, compared with commercial SEB ELISA test kits, light scattering ELISA show more reliable properties for the detection of colored samples, and the operation steps ca be further simplified. Furthermore, this method does not need strong acid to stop the r action, and the experimental operation is safer. Therefore, this method has been succes fully applied for the direct detection of SEB in complex food samples. We believe th method could be extended to the detection of other biotoxins by changing the correspond ing antibodies.
Supplementary Materials: The following supporting information can be downloaded a www.mdpi.com/xxx/s1, Figure S1: The reaction between ZIF−8 and H2O2. (A) The LS spectrum ZIF−8 before and after the reaction with H2O2. (B) The LS intensity of ZIF−8 before and after th reaction with H2O2. Experimental conditions: the concentration of H2O2 in A and B was 0.1%, 3% 30%; reaction time, 70 min; ZIF−8, 50 µg/mL. The error bars represent the standard deviation three replicates. Figure S2: SEB ELISA test kits test results (A). The concentration of SEB is 5 pg/m Figure 9. Linearity of SEA detection (A). The inset is the linear relationship between ∆I and SEA concentration (30 ng/mL to 70 ng/mL). Selectivity of the proposed strategy (B). Experimental conditions: the concentration of SEB was from 30 ng/mL to 70 ng/mL; the concentration of SEB, AFB1, BSA, and HGB were 50 ng/mL; reaction time, 70 min; ZIF-8, 50 µg/mL; H 2 O 2 , 3%. The error bars represent the standard deviation of three replicates.
Then, the selectivity of the method was tested by comparing the signal generated by SEA, staphylococcal enterotoxin B (SEB), aflatoxin B1 (AFB1), bovine serum albumin (BSA), and hemoglobin (HGB) at the same concentration. It was found that only SEA induced a high ∆I value ( Figure 9B), indicating that the method could be used for the specific detection of SEA. These results confirmed that this method is universal.

Conclusions
ZIF-8 MOF was synthesized and employed as a new LS reporter of the ELISA for the detection of SEB based on the unique reaction of ZIF-8 MOF and H 2 O 2 . This developed LS ELISA method has several advantages. Firstly, the synthesis of ZIF-8 is easy. Secondly, the structure of the synthesized ZIF-8 was uniform, and its LS signal was strong and stable. Thirdly, compared with commercial SEB ELISA test kits, light scattering ELISA shows more reliable properties for the detection of colored samples, and the operation steps can be further simplified. Furthermore, this method does not need strong acid to stop the reaction, and the experimental operation is safer. Therefore, this method has been successfully applied for the direct detection of SEB in complex food samples. We believe this method could be extended to the detection of other biotoxins by changing the corresponding antibodies.