Introduction

Staphylococcal food poisoning (SFP) is one of the three major food poisonings that results from consuming foods containing one or more enterotoxins. These toxins are secreted by gram-positive Staphylococcus aureus coagulase bacteria and other enterotoxins-producing staphylococci and produce many extracellular proteins and toxins—the most important of which are staphylococcal enterotoxins (SEs) and similar staphylococcal enterotoxins (SELs). SEs are significant contributors to foodborne illnesses, such as rheumatoid arthritis, atopic eczema, and toxic shock syndrome (Yang et al. 2008). Additionally, SEs like trypsin and pepsin are resistant to proteolytic digestion. Therefore, they maintain their activity through the gastrointestinal tract and can consequently cause illness. SEA is the most common SEs connected to food poisoning; is relatively stable in heat, and reported as the cause of foodborne illnesses in many countries (Upadhyay and Nara 2018). It causes nausea, vomiting, abdominal cramping, and diarrhea; the infectious dose of SEA is 105–108 cfu g−1 or 1 ng toxin per gram of food consumed, 200 ng per kg of human body weight (Bhunia 2018). About 15% of foodborne illnesses related to staphylococcal in the developed countries are due to dairy products. Milk is an attractive medium for the growth of Staphylococcus aureus; it contaminates raw, sterilized, and pasteurized milk (Upadhyay and Nara 2018). Regular cooking and pasteurization temperatures cannot wholly disable enterotoxins. So, in many cases, the bacterial cell may die, while the food might still contain toxins and, therefore, be pathogenic. It seems necessary to detect bacterial poisons in food samples such as milk before consumption for prevention purposes.

For SEs to diagnose in foods, several immunological and molecular tests have been introduced. These include fluorescent immunoassays (Lian et al. 2010; Rubina et al. 2010; Vinayaka and Thakur 2012; Vinayaka and Thakur 2013), molecular biological methods (Ding et al. 2011; Jaulhac et al. 1992; Rodríguez et al. 2016; Suwanampai et al. 2011), chromatography methods (Muratovic et al. 2015; Sospedra et al. 2012), mass-based immunoassays (Karaseva and Ermolaeva 2015), LC-MS / MS (Muratovic et al. 2015; Sospedra et al. 2012), colorimetric immunoassays (Nouri et al. 2018; Zhang et al. 2019), chemiluminescence electrochemiluminescence immunoassays (Szkola et al. 2014; Zhang et al. 2013), surface plasmon resonance (SPR) immunoassays (Medina 2006; Soelberg et al. 2009), electrochemical immunoassays (Tang et al. 2010; Wu et al. 2013), aptamer-based bioassays (Wang et al. 2015), and molecularly imprinted polymers (MIPs)-based bioassay (Liu et al. 2014). Several kits are also available to simplify identification. Although these techniques are all sensitive or specific, there are shortcomings such as high cost, relatively long analysis time, use of sophisticated analytical instruments, and the need for trained personnel that limits their routine use. Therefore, we still require a quick, accurate, accessible, and low-cost diagnosis method to detect low SEs in routine clinical operations and food analysis. This paper introduces a simple and economical method based on fluorescent quenching of CQDs by SEAs conjugated magnetic antibody (Fe3O4@Ab@SEA). The use of highly sensitive biological sensors increases the tests’ accuracy, reduces the length of analysis time, and is economical.

CQDs are soluble luminescent carbon nanomaterial with sizes below 10 nm synthesize from inexpensive precursors with simple protocols (Wang and Hu 2014). Some features make CQDs ideal for various applications: bright fluorescence, low toxicity, stable photochemical properties, excellent chemical stability, excellent biocompatibility, and good water solubility as well as easy creation of functional groups at the surface (Wang and Hu 2014; Xu et al. 2015). Several natural and chemical carbon sources are available for CQDs preparation through various synthetic protocols regarding specific applications (Ramezani et al. 2018). Citric acid is a reliable chemical carbon source mostly due to its low cost and accessibility (Wang and Hu 2014; Xu et al. 2015) and produces CQDs with high and stable fluorescence and controllable surface functionality.

In this study, at first, a specific monoclonal antibody (SEA antibody) loaded Fe3O4 nanoparticles (Fe3O4@Ab) were prepared and used in a magnetic extraction of SEA from pasteurized and homogenized milk (Fe3O4@Ab@SEA); magnetic properties allow easy separation of the nanoparticles. Then, SEA was quantified by the fluorescent quenching of CQDs in the presence of Fe3O4@Ab@SEA; the method determines SEAs contents of the low-, semi-, and full-fat sterilized and homogenized milk.

Materials and Methods

Materials

Staphylococcal enterotoxin A (SEA) from Staphylococcus aureus and gout anti-mouse IgG antibodies, conjugated with HRP, were provided from Sigma. Monoclonal mouse anti-staphylococcal enterotoxin A antibody (Ab) was purchased from Abbexa (Milton, UK). One milligram Ab was transferred to a 10-mL volumetric flask; it was adjusted to the volume with phosphate saline buffer (PBS) containing sodium azide (NaN3). Then, the solution was divided into 1 mL portions—in 10 mL micro-tubes—and stored at − 20 °C before use. The stock solution of 100 μg mL−1 SEA was prepared by dissolving 0.1 mg SEA powder in 1 mL distilled water. Working solutions were made by serial dilution of an intermediate solution (10 μg mL−1 SEA).

All chemicals were of analytical grade (purity over 98%) and obtained from Merck (Darmstadt, Germany): Iron (II) chloride tetrahydrate (FeCl2·4H2O), iron (III) chloride hexahydrate (FeCl3·6H2O), sodium hydroxide, hydrochloric acid, citric acid, diethylenetriamine penta-acetic acid (DTPA), ethylenediamine, Tris (hydroxymethyl) aminomethane (Tris), dimethyl sulfoxide (DMSO), skim milk powder, glutaraldehyde, and polyethylene glycol 6000. N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDC) with purity over 98% (Fluka, Germany). N-Hydroxy succinimide with purity over 98% was prepared from Sigma-Aldrich. Deionized water was used throughout the study.

Apparatus

Morphology of Fe3O4 and Fe3O4@Ab nanoparticles was determined by a FEI TEC9G26 (USA) high-resolution transmission electron microscope (HRTEM) with an acceleration voltage of 200 kV and a TESCAN MIRA3 (Czech Republic) field-emission scanning electron microscope (FESEM). The samples were deposited on a copper grid (in case of HRTEM) and aluminum foil (and coated with gold for FESEM) imaging. MDK vibrating sample magnetometer (VSM, Kerman Kavir Daghigh magnetometer, Iran) determined Fe3O4 and Fe3O4@Ab nanoparticles’ magnetic properties. FTIR was recorded on Vortex 70 (Bruker, Germany) in the form of kBr pellet to identify the functional groups of CQDs, Fe3O4 nanoparticles, and Fe3O4@Ab. Atomic force microscope (Nanowizard, JPK, and Germany) was used to analyze the morphology and the thickness of as-prepared CQDs. For AFM evaluations, the samples were dispersed in acetone, and a few drops were placed on the Mica sheet and let dry while rotating the sheet. The fluorescence emission was measured by Lumina spectrofluorometer (Thermo Scientific, USA). UV spectra were recorded by an array UV-Vis spectrophotometer (photon X Ar 2015, Iran). A pH meter (Nick, Germany) was used to adjust the acidity whenever required.

Synthesis of Carbon-Quantum Dots (CQDs)

CQDs were prepared according to the method reported by Wang et al. using citric acid as a carbon source (Wang et al. 2015). In summary, 10 mL deionized water, 0.42 g of citric acid, and 530 μL of ethylenediamine were transferred into a single Teflon-lined stainless-steel autoclave, and the solution heated at 200 °C for 5 h. After cooling down to room temperature, the solution was filtered and dialyzed against distilled water; then, it was transferred into a 50-mL falcon tube and refrigerated for future use. The concentration of CQDs—determined after lyophilizing 3 mL portions—was 22 mg mL−1.

Synthesis of Fe3O4 Nanoparticles

Fe3O4 nanoparticles were prepared by the conventional co-precipitation method according to our previous protocols (Safdarian and Ramezani 2018; Safdarian et al. 2016). So, FeCl2·4H2O (5 mM) and FeCl3·6H2O (10 mM) dissolved in 50 mL HCl (1 M); then, this solution was added to an alkaline solution containing sodium hydroxide (100 mL of 2 M) under reflux conditions (at 70 °C; nitrogen atmosphere) with a stirring rate of 800 RPM; the process of stirring continued for 1 h; finally, the black precipitate of Fe3O4 separated by a 1.4 T permanent magnet and washed several times with deionized water to neutralize the supernatant. The particles were completely freeze-dried (Operon; Korea) under vacuum at − 150 °C. The solid nanoparticles were dispersed in 50 mL distilled water and kept for future use. Fe3O4 suspension was 6 mg mL−1.

Preparation of Antibody Loaded Fe3O4 (Fe3O4@Ab)

One milliliter of DMSO was transferred to a 5-mL volumetric flask; 200 mg DTPA was added and made to volume with 0.1 M Tris HCl buffer (pH 9.6). After complete mixing, 1 mL (6 mg mL−1) of Fe3O4 nanoparticles dispersion was added. The suspension was shaken well for 2 h at room temperature; Fe3O4 was separated and washed twice with Tris buffer at pH 8.2; the particles were transferred to a 10-mL volumetric flask and adjusted to the volume with Tris buffer; one milliliter of the antibody (100 μg mL−1) was added. The solution was shaken for 2 h (800 RPM) at room temperature and kept at 4 °C overnight. Next, skim milk (1 mL 4%) was added (to cap free OH group that remains on the Fe3O4 surface), and it was shaken at 800 RPM for 2.5 h; Fe3O4@Ab was separated by a permanent magnet, rinsed twice with 10 mM PBS (pH 7.2), and transferred to a 10-mL volumetric flask, and adjusted to the volume with PBS. Scheme 1 represents the whole protocols.

Scheme 1
scheme 1

Synthesis protocols for Fe3O4@Ab

Full size image

Enzyme Immune Assay (EIA)

EIA is performed according to Jia-hui Lei et al. (Jia-hui Lei et al. 2011). To three out of four 1.5 mL micro-tubes, 100 μL Fe3O4@Ab was added; activated and blocked Fe3O4 nanoparticle (100 μL) was transferred to the fourth one (as a negative control). The particles were washed twice with 900 μL 10 mM PBS (pH 7.2), stirred a few minutes, and separated by a 1.4 T magnet. Next, 900 μL PBS and 100 μL gout anti-mouse IgG antibodies conjugated with HRP-(dilutions: 1/2000, 1/4000, and 1/8000) were transferred to micro-tubes 1–3; to the fourth, 100 μL of dilution 1/2000 was added; the micro-tubes were shaken for 30 min at room temperature. The magnetic nanoparticles were washed twice with PBS, as described above, and dispersed in 100 μL PBS; 50 μL of tetra-methyl benzidine (TMB) was added to this uniform suspension and stirred for 5 min. A blue color appeared in micro-tubes no. 1 to 3, indicating that the monoclonal antibody of enterotoxin A binds to Fe3O4 (Fe3O4@Ab). No color changes are observed in the negative control (Fig. S1, ESI). The reaction was ceased by 50 μL of 2 M HCl. Then, 100 μL of the supernatant was transferred to the ELISA plate wells; the absorbance (OD) was recorded at 450 nm (two replicates); the results are brought in Table S1.

Quantification of SEA

Milk samples and SEA standard (concentrations ranging from 0.1 to 2000 ng mL−1; 200 μL) were mixed with 200 μL of 11 mg mL−1 Fe3O4@Ab suspension (2.2 mg Fe3O4@Ab) and 1.1 mL PBS (pH 7.2) and were shaken for 1 h (300 RPM); Fe3O4@Ab@SEA was separated by a 1.4 T magnet and washed twice with PBS (pH 7.2)—to remove any adsorbed impurity from the sorbent surface, it was transferred to a 5-mL volumetric flask and adjusted to the volume with PBS (pH 7.2). For fluorescence measurements, 1.5 mL of the extracted (loaded on Fe3O4@Ab) standards, milk, and blank were mixed with 100 μL of CQDs, and fluorescence intensities at 440 nm are recorded while excited at 350 nm. This extraction procedure was repeated for blank sample (200 μL distilled water instead of SEAs).

Preparation of Sterilized and Homogenized Milk Samples

Four sterilized and homogenized milks—two low fat (1.5%), a semi-fat (2.5%), and full fat (3%)—were collected from Ahvaz (Iran) local supermarkets. (Milks were spiked with SEA standard at two levels—100 and 400 pg mL−1—before dilution; so, 200 μL and 800 μL of 1 ng mL−1 SEA (two separate tubes) were mixed with 200 μL milk, and 1600 μL deionized water was added; SEA is replaced by deionized water. measurement). SEA contents of the samples were exposed to extraction and determination regarding the quantification of SEA procedure described above.

Results and Discussion

EDC-NHS and DTPA were used for the attachment of Ab to the surface of Fe3O4 (Fe3O4@Ab). Zil Nilson stain test and ELISA showed that Ab was attached to the Fe3O4 surface in both methods. At the same time, ELISA indicated that in EDC-NHS method, unbonded Ab is 1.17 times more than those obtained by DTPA (the attachment yield was 95% using DTPA, much higher than EDC-NHS). Therefore, we used DTPA for the attachment of Ab to the Fe3O4 surface.

Characterization of Fe3O4@Ab

An antibody is a huge Y-shaped protein that is made up of two areas: one variable and one constant area. The structure of the changing area depends on its corresponding antigen. EIA (Table S1, Fig. S1) and FTIR (Fig. 1A) identified that Ab has successfully attached to the Fe3O4 surface. Fe3O4@Ab conjugation with antigen was confirmed by the blue color developed in the presence of TMB; the color intensity has risen with an increase in antigen concentration (Fig. S1 and Table S1). Antibodies have amino acids in their chemical structures, and Fe3O4@Ab FTIR spectra show corresponding bands for both amino acids and Fe3O4. Peaks at 590 cm−1and 3400 cm−1 (Fig.1A(a)) are due to Fe-O and OH of Fe3O4 (Safdarian and Ramezani 2018; Safdarian et al. 2016; Ta et al. 2016). Characteristic FTIR bands of antibodies attached to the surface of Fe3O4 are: amide I band at 1680 cm−1; hydrogen-bonded C=O stretching at 1650 cm-1; asymmetric carboxylate C=O stretching and CH2 deformation at 1430–1470 cm−1; symmetric carboxylate stretching and amide II at 1390 cm−1 (Nie and Stutzman 2005); stretching vibrations of COOH, NH2, and NH3+ at 3427, 3216, and 3130 cm−1 and the bending vibration of C-O at 1400 cm−1 (Fig. 1A(b) relates to the amino acid groups of Ab. The sharp FTIR band at 2300 cm−1 shows SH of cysteine.

Fig. 1
figure 1

(A) FTIR spectra of (a) Fe3O4, (b) Fe3O4@Ab, and (B) VSM of (a) Fe3O4, (b) Fe3O4@Ab. The inset is a photo of Fe3O4@Ab suspension before and after the application of a magnetic field

Full size image

Magnetic hysteresis loops demonstrate the potential application of the Fe3O4@Ab as a magnetic sorbent for separation of SEA at room temperature (Figure 1B), the absence of hysteresis in two curves of Fe3O4 and Fe3O4@Ab. The saturation magnetizations are 41.68 and 10.79 emu g−1 for Fe3O4 and Fe3O4@Ab, respectively. Magnetization has decreased in Fe3O4@Ab compared with Fe3O4—the percent of the magnetic core is lower (Safdarian and Ramezani 2018; Safdarian et al. 2016). But, the Fe3O4@Ab still has enough magnetic properties and can be used as an excellent magnetic sorbent. The inset images in Fig. 1B shows the action of Fe3O4@Ab in the presence and absence of an external magnetic field. Fe3O4@Ab attracted to the magnet in less than 10 s, and a clear solution obtained.

The morphology of as-prepared particles has been investigated by HRTEM and FESEM (Fig. 2 a–d). The size distribution pattern was determined using ImageJ 1.46r image processing software. The HRTEM image (Fig. 2a) illustrated almost amorph Fe3O4 particles with mean sizes of about 6.9 nm (Fig. 2c); the size did not change extensively after the conjugation of Fe3O4 with Ab (Fig. 2b and d). Inset image of Fig. 2a and b show selected area electron diffraction patterns (SAED). Orientation and crystal structure affect SAED pattern in HRTEM images (Xiaodong et al. 2011). When two molecules are to react, they should contact in the right direction to facilitate the bond formation; therefore, the orientation of Fe3O4 nanoparticles may change while attaching Ab. The change in orientation of Fe3O4 can be seen in the SAED patterns in HRTEM (Fig. 2) and XRD (Fig. S2) of the Fe3O4 particles before and after Ab loading. SAED patterns of Fe3O4 nanoparticles consist of weak crystalline phases, while the rings are brilliant in Fe3O4@Ab particles. It means the crystalline phase of Fe3O4 improves extensively on antibody loading. XRD pattern also confirms the amorphous nature of Fe3O4 before loading the antibody (Fig. S2) and increases in crystallinity on Ab loading. The core-shell is also apparent in Fig. 2b; Fig. 2e indicates that the Fe3O4 are amorph compared with the antibody-loaded Fe3O4 (Fig. 2f), confirming XRD results (Fig. S2).

Fig. 2
figure 2

HRTEM image of (a) Fe3O4, and (b) Fe3O4@Ab (histogram of (c) Fe3O4 and (d) Fe3O4@Ab); SEM images of (e) Fe3O4, and (f) Fe3O4@Ab

Full size image

Scatteroscope I, Qudix (Korea), was used for particle size analysis (PSA) of as-prepared particles; Fig. S3 shows particle distribution for as-prepared particles. The results indicated that d(90) for Fe3O4 and Fe3O4@Ab were 12.9 and 17.9 nm, consequently, increasing the size while loading Ab. Polydispersity indexes were calculated for Fe3O4@Ab and CQDs by squaring the result of standard deviation of the mean particle size divided by the mean size, and were 0.29 and 0.0012, respectively. Therefore, particles are dispersed well, while they are mixed.

CQDs Characterization

CQDs were prepared according to the reported method (Wang et al. 2015) and characterized before their use as a fluorophore. UV/Vis, excitation/emission, and fluorescence spectra of CQDs are recorded (Fig. S4A–C). CQDS showed an excitation-independent emission with maximum emission at 440 nm while excited by 350 nm (Fig. S4B, C); the maximum excitation wavelength correlates well with UV/Vis spectra of CQDs (Fig. S4A; 350 nm). CQDs emit blue color under exposure to 360 nm (the inset image in Fig. S4B). The FTIR spectra of CQDs (Fig. S4D) have shown functional groups on the surface. FTIR bands between 2800 and 3400 cm−1 related to OH stretching, C=C of the polycyclic aromatic ring of CQDs, and CH2 and CH3 vibrations, respectively (Liu et al. 2018; Wang et al. 2015; Xu et al. 2015). Aromatic carbon skeletal (C=C) and C=O vibrations are observed at 1579.37 cm−1, and 1680 cm−1, consequently. A band at 1385.96 cm−1 was due to C-N and C-O of amine groups attached to the CQDs surface. The AFM image (Fig S4E, F) of as-prepared particles, on the mica sheet, indicates a height less than 2 nm and the particle size below 10 nm. Emission intensity at 440 nm (excited at 350 nm) is approximately constant in pH range 5 to 10 (Fig. S5)—a suitable range for biological applications. XRD pattern of CQDs indicated corresponding broad peaks around 2θ equal 28 (Fig. S4-1A); PSA results in Fig. S4-1B confirms the CQDs formation. At pH 7, CQDs were exposed to 350 nm for 60 min, and emission intensity at 440 nm was recorded every 10 min. Data illustrate high photostability (Fig. S6). The quantum dots are stable for 6 months if kept refrigerated.

CQDs Quenching Mechanism

An analyte molecule can change the fluorescent or phosphorescence emission of a sensor through the transfer of energy. In an energy transfer, one molecule acts as a donor, and the other acts as an acceptor. Forster resonance energy transfer (FRET)—one of the energy transfer methods—occurs when there is an energy match or the spectral overlap between the donor molecules’ fluorescence emission and the absorption of the acceptor molecule. When this condition is met, photons emitted by the donor are adsorbed by the acceptor; if the acceptor is emissive, it emits its characteristic wavelength. In cases when the acceptors are not emissive, quenching is observed (Mako et al. 2018). CQDs are luminescent carbon-based materials that captured the special attention of researchers because of their different optical and chemical properties (Devanaboyina et al. 2013; Mako et al. 2018; Wang and Hu 2014). Figure S7 illustrates that CQDs fluorescent emission quenched when Fe3O4@Ab conjugated with SEA. UV-Vis spectra of SEA in Fig. S4A (the blue line) identifies a considerable absorbance at 440 nm, the maximum emission wavelength of CQDs. Accordingly, the spectral overlap of fluorescence emission of CQDs with absorbance spectra of SEA causes the quenching of fluorescent emission of CQDs via FRET (CQDs act as donor, and the SEA as acceptor just absorb the light). The inner filter effect (IFE) may also affect fluorescent emission quenching of CQDs to a less extent since an absorbance overlap at 350 nm has been observed in the UV-Vis spectra of CQDs and SEA (Fig. S4A). Therefore, the mechanism of quenching can describe by both IFE and FRET.

Optimization

Different parameters were optimized to arrive at a sensitive assay: extraction pH and kind of buffer; the amount of CQDs; contact time; extraction time; the amount of Fe3O4@Ab; and ionic strength. Extraction pH has a strong effect on the method response. The effect of pH on the equilibrium constant of antibody-antigen results in the symmetric curve, with the highest equilibrium constant value at pH between 6.5 and 8.4; pH higher than 8.4 and lower than 6.5 may strongly inhibit the antigen and antibody response and thereby significantly reduce the equilibrium constant. For example, the equilibrium constant at pH 9.5 or pH 5 is 100 times lower than that of pH 6.5–7. Extreme pH causes considerable structural changes in the antibody molecule, which can destroy its binding to the antigen (Boenisch 1999; Devanaboyina et al. 2013; Hughes-Jones et al. 1964; Reverberi and Reverberi 2007). Maximum quenching has been observed at the extraction pH of 7 (Fig. 3a), so neutral or physiologic pH is considered optimum for SEA extraction. Different buffers, including phosphate, Britton-Robinson, and ammonia as well as phosphate buffer saline (PBS), had similar results; PBS was selected because it is the most used in physiological and immunological tests.

Fig. 3
figure 3

Effect of (a) pH, (b) CQDs amount, (c) contact time, (d) the amount of Fe3O4@Ab, (e) extraction time, and (f) KNO3 concentration on the response of the present biosensor. The results are mean of two replicates

Full size image

Fluorescence quenching of CQDs increased with an increase in its concentrations up to 5.5 μg mL−1, the optimum amount of CQDs (Fig. 3b), and when CQDs are in contact with Fe3O4@Ab@SEA for 10 min (Fig. 3c). Extraction time was selected concerning the time required for the conjugation of antibody-antigen; the conjugation is usually 1 h. Therefore, Fe3O4@Ab was dispersed in SEA solutions for 30, 45, and 60 min, consequently. The difference in fluorescent quenching is approximately the same for SEAs extracted after 45 and 60 min (Fig. 3e); 60 min was selected as the extraction time because the immunological test requires 60 min for the conjugation (Reverberi and Reverberi 2007). The amount of Fe3O4@Ab is also critical in the extraction of the SEA; conjugation is lower at a low amount of the sorbent. The results show that the fluorescence difference increased with an increase in the amount of Fe3O4@Ab up to 2.2 mg (Fig. 3d). Reduction in extraction efficiency at a higher amount of Fe3O4@Ab may be due to the improper dispersity of the particles or their aggregations. The effect of ionic strength on the device sensing ability has been evaluated by adding the different concentrations of potassium nitrate to Fe3O4@Ab@SEA suspension before contact with CQDs. Figure 3f indicates that the ionic strength does not have a considerable effect.

Method Evaluation

The figure of merit—accuracy, precision, the limit of quantitation (LOQ), and limit of detection (LOD), as well as linearity—was investigated to show that the method is reliable. The calibration graph was obtained under established experimental optimum conditions. Fe3O4@Ab@SEAfluorescence spectra—in contact with CQDs—has been recorded for SEA concentration in the range of 0 to 40000 pg mL−1 (Fig. 4). The Stern-Volmer equation can describe the fluorescence quenching performance (inset plot of Fig. 4):

Fig. 4
figure 4

Fluorescence spectra for SEA range from 4 to 40000 pg mL−1 (excited at 350 nm). The inset graph plots F0/F versus the concentration of SEA. The results are mean of three replicates

Full size image

\( {F}_{\raisebox{1ex}{$0$}\!\left/ \!\raisebox{-1ex}{$F$}\right.}=1+{k}_{sv}C \)F0 and F are fluorescence intensities of Fe3O4@Ab +CQDs and Fe3O4@Ab@SEA + CQDs, respectively; Ksv is the Stern-Volmer quenching constant; and C is the SEA concentration. Table 1 illustrates two linear ranges with an acceptable correlation coefficient. LOD and LOQ were calculated as the concentrations having a signal equal to three and ten times the standard deviation of blank divided by the slope of the calibration curve and were 2 and 4 pg mL−1, consequently. Repeatability and reproducibility of the measurements—expressed as coefficient of variation (CV)—are evaluated, by assay of three samples at different time intervals in a single day and six consecutive days (Table 1).

Table 1 Figure of merit of the proposed method for enterotoxin A determination
Full size table

SEA Analysis of Milk Samples

A suitable analytical technique should be reliable in the analysis of the target analyte in a complicated matrix. The complexity of the milk matrix makes the analysis difficult; herein, the specific sensor based on the magnetic SEA antibody can measure SEA in sterilized and homogenized milk with high reliability. The concentration of SEA in the sterilized and homogenized milk samples is determined before and after the addition of a certain amount of SEA standard (Table 2); the samples had different fat contents. The results in the table demonstrate that the method is reliable for the determination of SEA in sterilized and homogenized milk; recoveries are acceptable.

Table 2 Results for SEA determination in the different milk samples by the proposed method
Full size table

Comparison of the Previously Reported Methods

Table S3 lists some recently published articles on the quantitation of SEA. In some cases, sophisticated techniques such as HPLC equipped with a tandem mass detector required that limits their routine uses. LOD and LOQ and the linear range of the present method are in pg mL−1 range and improved extensively compared to the previously reported methods (Table S3). ELISA is a method of choice for SEA assay in the quality control labs. The ELISA kit is expensive; the detection limit is in ng mL−1, applied for a limited number of samples, and two-step incubation times required as well as interference from milk matrices such as fat and proteins lengthen the assay.

Last but not least, LOD, LOQ, and the linear determination range are higher than the current methodology. The present biosensor is simple, relatively fast, economical, and sensitive as well as specific. This technique can determine pg mL−1 of SEA in sterilized and homogenized milk samples with no need for fat removal. It also eliminates the elution steps usually used in the extraction techniques; so, no need for elution of SEA from Fe3O4@Ab@SEA before fluorometric assay.

Conclusion

It is the first time that a fluoro-sensor based on CQDs was introduced to detect and determine SEA in milk to the best of our knowledge. SEA antibodies successfully loaded on Fe3O4; the magnetic sorbent (Fe3O4@Ab) could extract SEA from sterilized and homogenized milk samples without fat removal, and there is no need for elution of SEA from the magnetic sorbent before sensing. The technique is simple, economical, reliable, sensitive, and relatively fast. It can detect the amount of SEA in different types of milk with excellent recovery. This biosensor application can be extended to SEA quantification in different food samples and biological matrixes such as plasma because Fe3O4@Ab is specific toward SEA.