362 Journal o f Food Protection, Vol. 78, No. 2, 2015, Pages 362-369 doi: 10.4315/0362-028X.JFP-14-323 Copyright © , international Association for Food Protection
Development of an Indirect Competitive Enzyme-Linked Immunosorbent Assay Based on the Multiepitope Peptide for the Synchronous Detection of Staphylococcal Enterotoxin A and G Proteins in Milk MINGYAN LIANG,'f TINGTING ZHANG,'f XUELAN LIU,1 YANAN FAN,1 SHENGLIN XIA,2 YIQING XIANG,1 ZIQI LIU,1 AND JINNIAN LI1* 1College o f Animal Science and Technology, Anhui Agricultural University, 130 Changjiang Avenue, Hefei 230036, Anhui, People’s Republic o f China; and 2Center o f Animal Epidemic Disease Prevention and Control, 460 Liming Avenue, Wuhu 241000, Anhui, People’s Republic o f China MS 14-323: Received 10 July 2014/Accepted 25 September 2014
ABSTRACT Staphylococcal food poisoning (SFP), one of the most common foodbome diseases, results from ingestion of staphylococcal enterotoxins (SEs) in foods. In our previous studies, we found that SEA and SEG were two predominant SE proteins produced by milkacquired S. aureus isolates. Here, a tandemly arranged multiepitope peptide (named SEAGepis) was designed with six linear B-cell epitopes derived from SEA or SEG and was heterologously expressed. The SEAGepis-specific antibody was prepared by immunizing rabbit with rSEAGepis. Then, an indirect competitive enzyme-linked immunosorbent assay (ic-ELISA) based on rSEAGepis and the corresponding antibody was developed to simultaneously detect SEA and SEG. Under the optimized conditions, the ic-ELISA standard curve for rSEAGepis was constructed in the concentration range of 0.5 to 512 ng/ml, and the average coefficients of variation of intraand interassay were 4.28 and 5.61% during six standard concentrations. The average half-maximal inhibitory concentration was 5.07 ng/ml, and the limit of detection at a signal-to-noise ratio of 3 was 0.52 ng/ml. The anti-rSEAGepis antibody displayed over 90% cross-reactivity with SEA and SEG but less than 0.5% cross-reactivity with other enterotoxins. Artificially contaminated milk with different concentrations of rSEAGepis, SEA, and SEG was detected by the established ic-ELISA; the recoveries of rSEAGepis, SEA, and SEG were 91.1 to 157.5%, 90.3 to 134.5%, and 89.1 to 117.5%, respectively, with a coefficient of variation below 12%. These results demonstrated that the newly established ic-ELISA possessed high sensitivity, specificity, stability, and accuracy and could potentially be a useful analytical method for synchronous detection of SEA and SEG in milk.
Staphylococcus aureus, a major cause of clinical and subclinical mastitis in dairy cows, is responsible for staphy lococcal food poisoning (SFP) as a result of the heat-stable staphylococcal enterotoxins (SEs) produced by enterotoxigen ic strains (2). Within 1 to 6 h after ingestion of food containing 100 to 200 ng of SEs, symptoms of acute gastroenteritis can be observed (22). In recent years, the presence of enterotoxigenic S. aureus in food animals and livestock products has increased in some countries; in particular, cows with subclinical S. aureus mastitis can release large quantities of S. aureus into their milk (12, 15, 18). Although pasteurization kills S. aureus cells, thermostable SEs generally retain their biological activity. Thus, because of the importance of these toxins in the public health and food sectors, continuous monitoring of S. aureus and SEs in raw milk is required. SEs belong to a large pyrogenic toxin family. More than 20 serological types of SE have been identified to date, including classical SEs (SEA through SEE), newly reported SEs (SEG, SEH, SEI, SEJ, SER, SES, and SET), and SE-like * Author for correspondence. Tel: +86-551-65785935; Fax: + 86-55165785935; E-mail: lijinnian2000 @ 163.com. f Both authors contributed equally.
toxins (SELK, SELL, SELM, SELN, SELO, SELP, SELQ, and SELU) (1 ,3 ,6 , 16). Previously, reports identified SEA and SED as the predominant SEs in cases of SFP, followed by SEB, SEC, and SEE (8,19). However, recent reports show a higher frequency of the newly reported seg gene compared with the classical SE genes in S. aureus isolated from several sources (27). In our previous study, over 90% of 114 S. aureus strains isolated from mastitic milk were found to carry at least one type of SE gene. The most frequently identified SE gene was seg (89%), followed by sea (40%). Moreover, 39% of the S. aureus isolates analyzed carried, synchronous ly, seg and sea genes (data not shown). These differences may be a result of variations in the origin and geographical locations of the isolates. Enzyme-linked immunosorbent assay (ELISA) has been extensively used for diagnostic and residue detection purposes due to its sensitivity, specificity, rapidity, and simplicity. Saunders and Bartlett (20) first detected SEA in foods with ELISA. Since that time, a number of variations of ELISAbased methods have been introduced. There have been some commercial ELISA test kits to detect classical SEs (A, B, C, D, E) (4,21), but these were not available for detecting novel
J. Food Prot., Vol. 78, No. 2
SYNCHRONOUS DETECTION ASSAY FOR STAPHYLOCOCCAL ENTEROTOXINS A AND G
363
TABLE 1. Amino acid sequence o f linear B-cell epitope from SEA and SEG proteins Potential
epitope no. SEA
1 2
3 4 5 SEG
1 2
3 4 5 0
Amino acid position" 85-88 138-141 156-163 166-173 211-218 24-31 37-46 80-84 120-127 141-149
Peptide sequence HSWY HDNN DGKQNTVP TVKTNKKN HTSTEPSV NSQPDPKI VSDYKSNKGT EYKSY KSEPDINQ NSSENRDK
Avg antigenic index6 1.26 1 .1
1.58 2.05 1.54 1.92 2 .0 1 1 .8 8
2.38 2 .0 1
The corresponding position of each peptide in the native SEA and SEG proteins, respectively. Antigenicity index was computed by the Janeson-W olf algorithm. Average antigenicity index is the mean value o f all amino acid antigen indexes among epitope peptides.
SEs. In the present study, a tandemly arranged multiepitope peptide (SEAGepis) from both SEA and SEG protein was designed, prepared, and used to immunize rabbit to prepare anti-SEAGepis antibody. With these key fundamental materials, an indirect competitive ELISA (ic-ELISA) based on the multiepitope peptide was then developed to synchro nously detect the SEA and SEG in raw milk. M A TER IA LS A ND M E TH O D S Bacterial strains and plasmids. The Escherichia coli host strain Rosetta (DE3) was purchased from China Institute of Veterinary Drug Control and was cultured in Luria-Bertani medium (Beijing Solarbio Science & Technology Co., Ltd., Beijing, People’s Republic of China) supplemented with 100 pg/ml ampicillin as needed. The pE T -32a(+ ) vector (Novagen, Madison, WI) was used for gene expression. Prediction and validation of immunodominant linear Bcell epitopes of SE proteins. The amino acid sequences o f SEA and SEG proteins o f the S. aureus standard strain ATCC25923 were used for linear B-cell epitope prediction. The secondary structure, flexible regions, hydrophilicity, surface accessibility, and antigenic index of both proteins were analyzed using the Protean program from the Lasergene software package (DNASTAR, Inc., Madison, WI). The peptide sequences with high hydrophilicity, accessibi lity, flexibility, and antigenicity, which were located in the P-tum and random coil regions of the protein secondary structure, were considered as potential linear B-cell epitopes. The three-dimensional (3D) structure o f the SEA and SEG proteins was simulated using the Molecular Operating Environment Software (version 2008, Chemical Computing Group, Montreal, Canada), and the potential B-cell epitopes were represented using the synchronize visualization program (version 2005, Accrelrys, San Diego, CA). To verify the prediction, six potentially immunodominant epitope peptides that had higher antigenicity indices and were exposed on the surface of the 3D structure at varying depths, as well as an irrelevant control peptide (NTNYANSQ), were artificially synthesized (purity >98% ) by GenScript Biotechnology (Nanjing, People’s Republic o f China) and detected by peptide-ELISA with purified rabbit anti-m altose binding protein (M BP)-SEA and anti-MBP-SEG antibodies pre pared as described by Cen et al. (7); commercial rabbit anti-MBP tag antibody and normal rabbit serum were used as the negative control.
Design of a tandemly arranged multiepitope peptide of SE proteins. Based on the sequences o f the six immunodominant linear
B-cell epitopes of SEA or SEG proteins (SEA156_163aa, SEA166_173aa, SEA3i ]— 2 i 8 aa. SEG37^ f,all, SEG|2f>-i27m- and SEG| 4 i_i4 9 aa) (Table 1) and the tandem design principle, a tandemly arranged multiepitope peptide (named SEAGepis) was designed with six immunodominant linear B-cell epitopes linked with two alanines (AA) and one tyrosine (Y) as an AAY spacer between the epitopes. To maintain the independence and antigenicity o f each epitope and to avoid the generation of a junctional epitope, the connection order of the six epitopes was optimized by predicting the hydrophilicity, antigenicity, and surface probability o f the tandemly arranged multiepitope peptide using the Protean program from the Lasergene software package (DNASTAR, Inc.) (24). The multiepitope peptide gene (318 bp) was artificially synthesized and cloned into the pUC57 vector by GenScript Biotechnology (Nanjing, People’s Republic of China).
Expression and purification of the recombinant multi epitope peptide. The gene-encoding multiepitope peptide (SEA Gepis) was amplified from the recombinant plasmid pUC57SEAGepis by PCR using the following primers: 5'-CAG AT GGATCCC ATGATAATAATCGTTTG ACCGA AG AG-3' (BamhU site underlined) and 5'- GTGTC AAGCTTTTATTTATCGCGTTCATTTTCAGAACTATTAAA-3' (HindUl site underlined). Puri fied PCR product and plasmid pE T -32a(+) were digested with BamHl and Hind III, and then a ligation reaction was set up. The recombinant plasmid pET-32a-SEAGepis was verified by restriction enzyme digestion and sequencing. The positive recombinant plasmid was transformed into Rosetta (DE3) E. co//-competent cells upon induction with 1 mM isopropyl-fLo-thiogalactopyranoside (SigmaAldrich, St. Louis, MO) at 37°C for 2 h. The bacterial cells were then harvested by centrifugation at 10,000 x g for 5 min at 4°C. After analysis by sodium dodecyl sulfate-polyacrylamide gel electropho resis (SDS-PAGE), the soluble rSEAGepis was purified using a highaffinity Ni-NTA resin column (GE Healthcare, Piscataway, NJ) according to the manufacturer’s instructions. Protein purity was estimated by SDS-PAGE analysis using bioinformatics software BandScan (http://www.bbioo.com/Soft/2005/75.htm), and the con centration of purified protein was determined with a Bradford protein assay kit (Tiangen Biotech Co., Ltd., Beijing, People’s Republic of China). In addition, a Western blot was performed as described previously (5) with the mixture of anti-SEA and SEG polyclonal antibodies (prepared as described by Cen et al. (7)) to confirm the immunoreactivity of the rSEAGepis.
Preparation and purification of rabbit anti-rSEAGepis antibody. Adjuvant Montanide ISA763A oil (Seppic, Paris, France) was mixed with rSEAGepis at a 3:7 volume ratio and then was
364
LIANG ET AL.
stirred at room temperature for 1.5 h to obtain a stable emulsion. A New Zealand white rabbit (weight 2,000 g) was vaccinated with rSEAGepis (500 pg per rabbit in 500 pi) by multipoint subcutaneous injections in the back. After 2 weeks, a booster was given with the same amount of antigen without the adjuvant. On days 0 and 28 after the primajy immunization, blood samples were collected via the marginal ear vein, and then the serum antibody titer was assayed by indirect ELISA. Briefly, polystyrene microtiter plates (Coming Inc., Coming, NY) were coated with 100 pi of the rSEAGepis (5 pg/ml), and unbound antigens were removed by washing three times with phosphate-buffered saline (PBS) containing 0.25% Tween-20 (PBST). After blocking with 5% skim milk, serial dilutions of immune serum and preimmune serum samples were placed in wells and were incubated for 2 h at 37°C. Wells were then washed three times with PBST. A 1:3,000 dilution horseradish peroxidaseconjugated goat anti-rabbit IgG antibody (Tiangen Biotech, Co., Ltd., Beijing, People’s Republic of China) was added to the wells and incubated for 1 h at 37°C. Next, the color was developed by reaction with o-phenylenediamine (Ziyi Reagent Co., Shanghai, People’s Republic of China) substrate. The plates were read at 490 nm using an ELISA reader. The endpoint titer was defined as the highest dilution that gave an absorbance 2.1 times higher than the preimmune serum. The anti-rSEAGepis antibody in serum was purified by the combination of saturated ammonium sulfate (Ziyi Reagent Co.) precipitation and a HiTrap Protein GHP column (Amersham Pharmacia Biotech, Uppsala, Sweden) according to the manufactur er’s instructions. The purity, concentration, and titer of the purified antibody were detected by SDS-PAGE, the Bradford method, and the aforementioned indirect ELISA, respectively. Optimization of assay conditions. To develop a highly sensitive and specific ic-ELISA, assay conditions, such as the concentration of the coating antigen, dilutions of the primary antibody, selection of the buffer solution and blocking reagent, and incubation time and temperature, should be optimized. Criteria used to assess the conditions included maximum absorbance (B0, usually kept in the range of 0.5 to 1.5 according to the Lambert-Beer law for spectrophotometric analysis) and half-maximal inhibitory concen tration (IC50; the concentration of standard solution producing 50% of inhibition) (23). Various concentrations of coating antigen rSEAGepis (from 0.5 to 1.5 pg/ml) in combination with dilutions of rSEAGepis-specific antibody (from 1:8,000 to 1:1,024,000) were tested using a checkerboard procedure. Different buffer solutions (carbonate-buffered saline and PBS) and blocking reagents (sterile skimmed milk powder, bovine serum albumin, and casein) prepared with PBS at a concentration of 5% were tested to determine the effects of these factors. The effects of incubation temperature (37°C, room temperature) and time (2 h, 1 h, 30 min, and 15 min) during blocking and immunoreaction steps were also examined. ic-ELISA procedure. Polystyrene microtiter plates (Coming Inc.) were coated with rSEAGepis solution (0.75 pg/ml, 100 pi per weft) and incubated overnight at 4°C. Free sites were blocked with 5% sterile skimmed milk powder prepared with PBS for 30 min at room temperature. Fifty microliters of the purified rSEAGepisspecific antibody (at 1:256,000 dilution) was added to each well, followed by the addition of an equal volume of rSEAGepis standard solution or samples, both in triplicate. After an incubation period of 2 h, a 1:3,000 dilution of goat anti-rabbit IgG-horseradish peroxidase (Tiangen Biotech Co., Ltd., Beijing, People’s Republic of China) was added to each well and was incubated for 1 h. Next, the color was developed by reaction with o-phenylenediamine (Ziyi Reagent Co). Three washings were performed after each incubation
J. Food Prot., Vol. 78, No. 2
step with PBST, and the incubation temperature was 37°C. The absorbances were measured using an ELISA reader at 490 nm, and all measurements were performed in triplicate. Calibration curve generation and intra- and interassay variation determination. The rSEAGepis standard solution at six concentrations (0.5, 2, 8, 32, 128, and 512 ng/ml) was used as competitor, and ic-ELISA was performed as described above. The results are presented as inhibition (%) = [1 — (B/B0)] x 100, where B = absorbance of rSEAGepis diluted solution or test sample solution, and B0 = absorbance in the absence of rSEAGepis competitor. The calibration curve was plotted using the logarithm of concentrations of rSEAGepis as the lateral coordinates and corresponding inhibition as the longitudinal coordinates. The concentration of SEA and (or) SEG in a milk sample could be calculated by interpolating their mean absorbance values in the calibration curve run in the same plate. To assess the accuracy and stability of the ic-ELISA, intra-assay variation was tested by determining the coefficients of variation (CVs) of six rSEAGepis samples of various concentrations in replicates of three across the microtiter plate. Interassay variation was determined by evaluating the rSEAGepis samples in triplicate on three different microtiter plates for as long as 3 weeks. The CVs were then calculated as CV (%) = (sample standard deviation/sample mean) x 100. Sensitivity and specificity of the assay. IC50 value and limit of detection (LOD) were key criteria for evaluating the sensitivity of ELISA (9). IC50 value was estimated by using the equation of the inhibition standard curve of rSEAGepis. LOD (the lowest analyte concentration at a signal-to-noise ratio of 3) was calculated by subtracting three times the standard deviation of the obtained absorbance from the mean absorbance and then calculating the corresponding concentration with the established equation of the standard curve. The specificity of the anti-rSEAGepis antibody was assessed by detection of cross-reactivity against several enterotoxins, including SEA, SEG, SEB, SEC2, E. coli heat-stable enterotoxin, and Shiga toxin (all kindly provided by the Anhui Academy of Medical Sciences, Anhui, People’s Republic of China). The test enterotoxins (0.5 to 512 ng/ml) were used in the ic-ELISA procedure as described above. The cross-reactivity values were calculated as cross-reactivity (%) = (IC50 of rSEAGepis /lC5o of tested enterotoxin) x 100. Detection of SEAGepis, SEA, and SEG in artificially contaminated milk. To demonstrate the applicability of the icELISA, spiked milk samples were tested. Commercial ultra-hightemperature pasteurized milk purchased from a retail store was centrifuged at 3,500 x g for 10 min at 4°C, and the upper layer of cream was removed by absorption. Once the fat was removed, the milk was spiked with rSEAGepis, pure SEA, and SEG at various levels from 6.25 to 400 ng/ml. Spiked samples were vigorously homogenized by vortex oscillation for 5 min and then were detected using the established ic-ELISA. Unspiked sterilized milk was treated in the same way and set as a blank. A0 assays were carried out in triplicate, and recoveries were calculated as recovery (%) = (Css — CUJC S) x 100, where Css and Cus are concentrations of the enterotoxins measured in the spiked sample and unspiked samples, respectively, and Cs is the spiked concentration. Comparison of established ic-ELISA with commercial ELISA kit. A commercial ELISA kit for SEA (TECRA Diagnostics, Roseville, Australia) was tested along with the established ic-ELISA to compare the sensitivity of two methods. Pure SEA was added to defatted milk (treated as above) at 0.13, 0.26, 0.52, 1.04, and 2.08 ng /ml, and then samples were tested with the two methods. The
J. Food Prot., Vol. 78, No. 2
SYNCHRONOUS DETECTION ASSAY FOR STAPHYLOCOCCAL ENTEROTOXINS A AND G
365
(A) Sam Hl
Hindi
l
l
5' -G G A T C C ^ S E A 15^163-A A Y -S E A 166,173 -A A Y -S E A 211,218-A A Y -S E G 37.46-A A Y -S E G 120.127-A A Y -S E G 141,149 -T A A -A A G C T T - 3'
S top codon
(B) scale
l H ydrophtlid ty Ptot-Kyte -Doolittle
] Antigenic Index-Jameson-Wolf
D Surface Probability Plot-Emtm
FIGURE 1. Design and analysis o f the multiepitope peptide SEAGepis. (A) Design o f the tandemly arranged multiepitope peptide SEAGepis. The tandem contains six epitopes o f SEA or SEG, an AAY linker, a stop codon, a BamHI restriction site at the 5' end and a Hind ill restriction site at the 3' end. The optimized order o f epitopes was SEA, 56 - 163-SE A i 66- i 7j^SEA 2 n^ 2 i 8-S E G j 7^f6-S E G i 2o-i2 TSEG i 4 i _i 49. (B) Analysis ofhydrophilicity, antigenicity, and surface probability o f the tandem. The amino acid sequence o f the tandem was analyzed by the Kyte-Doolittle algorithm fo r hydrophilicity, by the Janeson-Wolf algorithm fo r antigenicity index, and by the Emini algorithm fo r surface probability. D ashed areas indicate six epitopes. Each epitope maintained its independence and antigenicity. commercial ELISA kit was used according to the manufacturer’s protocols. All samples were tested in duplicate, and all assays were repeated in triplicate.
RESULTS Immunodominant B-cell epitopes of SE proteins and design of a tandemly arranged multiepitope peptide. The
secondary structure, flexible regions, hydrophilicity, sur face accessibility, and antigenic index of the SEA and SEG proteins were analyzed using DNASTAR’s Protean program, and 10 potential epitopes were screened for their suitability (Table 1). Among potential epitope peptides, SEA156_]63aa, SEA166_i73aa, and SEA2U_218aa as well as SEG37^ 6aa, SEG j2 0 - 127 aa> a n d SEG i i_i 9aa were exposed on the surface 4
4
FIGURE 2. SDS-PAGE and Western blot analysis o f the SEAGepis expressed in E. coli Rosetta (DE3). (A) Lane 1, protein molecu lar standard; lane 2, uninduced pET-32aSEAGepisIRosetta (DE3); lane 3, induced pET-32a- SEAGepis!Rosetta (DE3); lane 4, 1 mglml bovine serum albumin (control); lane 5, purified rSEAGepis. (B) Lane I, prestained protein ladder; lane 2, Western blot o f purified rSEAGepis with the mixture o f anti-SEA and SEG polyclonal antibodies.
366
J. Food Prot., Vol. 78, No, 2
LIANG ET AL.
M
1
2
SDS-PAGE analysis showed that a 31.5-kDa soluble protein band appeared in the original and purified expression products (Fig. 2A, lanes 3 and 5), and its size was consistent with the theoretically expected relative molecular weight of the His-SEAGepis protein. No protein band was found at the same position in the uninduced recombinant E. coli Rosetta (DE3) (Fig. 2A, lane 2). After purification with a highaffmity Ni-NTA resin column, the purity of the recombinant protein was about 95%, as estimated by Coomassie blue staining on 12% (wt/vol) SDS-PAGE. The expression level of rSEAGepis was found to be 17 mg/litcr of medium. As shown by Western blot analysis, a single band at about 31.5 kDa was observed (Fig. 2B, lane 2). These results indicated that the purified rSEAGepis possessed sufficient immunoreactivity and should be satisfactory as an antigen for an immunoassay.
3
116kDa 66.2kDa 45kDa 35kDa
25kDa
18.8kDa 14.4kDa FIGURE 3. SDS-PAGE analysis of rabbit anti-rSEAGepis anti
body. M, protein molecular standard; lane 1, unpurified rabbit immune serum; lane 2, rabbit immune serum treated with saturated ammonium sulfate; lane 3, rabbit anti-rSEAGepis antibody, which was purified by protein G agarose affinity chromatography, under reducing condition. of the 3D structure model at varying depths. The results of peptide-ELISA showed that the six synthetic potential epitope peptides could be specifically recognized by the rabbit antiMBP-SEA or anti-MBP-SEG antibodies, but they did not bind with the commercial rabbit anti-MBP tag antibodies and normal rabbit serum. Any nonspecific reaction was also excluded because the irrelevant control peptide did not react with the antibodies against SEA, SEG, or the MBP tag peptide. Therefore, the six peptides represented immunodom inant linear B-cell epitopes of SEA or SEG. As shown in Figure 1, the tandemly arranged multiepitope peptide SEAGepis containing six immunodominant B-cell epitopes of SEA or SEG, five AAY linkers, a termination codon, a BamHl restriction site at the 5' end, and a HindUl restriction site at the 3' end was designed successfully. The opti mized connection order of these epitopes was SEA 156_i63 S E A i 6 6 _ 17 3 - S E A 2 i i _ 2 l 8 - S E G 3 7 ^ t 6 - S E G i 2 0 - 1 2 7 “ S E G i 4 i _ i 4 9 ,
in which each epitope maintained its independence and antigenicity as analyzed by the Lasergene Protean program. Purified rSEAGepis protein. To obtain the rSEAGepis, E. coli Rosetta (DE3) cells harboring pET-32a-SEAGepis were induced with isopropyl-(3-D-thiogalactopyranoside.
Antibody characterization and optimization of assay conditions. Antisera were collected and purified after 4 weeks of primary immunization. As shown in Figure 3, purified antibody exhibited two bands on SDS-PAGE; the molecular weight of the heavy chain and light chain of IgG were about 50 and 25 kDa under reducing condition with (3mercaptoethanol (Ziyi Reagent Co.). The titer of purified antibody was 1:819,200,1:409,600, and 1:409,600 as detected by ELISA using rSEAGepis, SEA, and SEG as coating antigens, respectively. According to the criteria with a higher Amax value and lower IC50 value, the immunoassay was optimized. Finally, carbonate-buffered saline and 5% skim milk were used for coating and blocking, respectively, in the assay. The optimal coating antigen concentration was 0.75 pg/ ml, and dilution of the primary antibody was 1:256,000, producing a maximum absorbance of approximately 1 in the absence of analyte (Table 2). The appropriate temperature of blocking and immunoreaction was 37°C, and incubation times of primary antibody, secondary antibody, and color develop ment were 2 h, 1 h, and 15 min, respectively. Specificity, sensitivity, and specificity of the assay. The stability of the assay was tested by running the ic-ELISA procedures three individual times. Figure 4 shows the average calibration curve of rSEAGepis generated from three assays: y = —38.286.x + 125.15, with a linear range from 0.5 to 512 ng/ml (R1 = 0.991, n = 3). As shown in Table 3, the average CVs of intra- and interassay were 4.28 and 5.61% during six standard concentrations. The average IC50 was 5.07 ng/ml, and the LOD was 0.52 ng/ml. Moreover, the established ic-ELISA was tested in comparison with a
TABLE 2. Checkerboard titration of rSEAGepis antigen versus anti-rSEAGepis primary antibody Fold dilution of primary antibody ( x 10-3) Concn of rSEAGepis coating antigen (pg/ml)
1:8
1:16
1:32
1:64
1:128
1:256
1:512
1:1024
1.5 1.25 1.0 0.75 0.5
2.84 2.84 2.75 2.84 2.70
2.32 2.47 2.42 2.49 2.34
1.95 2.11 2.09 2.04 1.90
1.39 1.51 1.57 1.73 1.47
1.13 1.13 1.22 1.31 1.17
0.86 0.95 0.88 1.00 0.93
0.72 0.72 0.64 0.57 0.68
0.40 0.41 0.44 0.42 0.44
J. Food Prot., Vol. 78, No. 2
367
SYNCHRONOUS DETECTION ASSAY FOR STAPHYLOCOCCAL ENTEROTOXINS A AND G
TABLE 4. Comparison o f sensitivity o f the established ic-ELISA with that o f the commercial ELISA kit fo r detection o f SEA in spiked milka
Test results Spiked concn (ng/ml) 0.13 0.26 0.52 1.04 2.08
ic-ELISA
Commercial ELISA kit
—
—
+ + +
+ +
a + , detection of SEA in spiked milk; —, no detection.
DISCUSSION FIGURE 4. The average standard calibration curve o f the established ic-ELISA fo r rSEAGepis run three times fo r as long as 3 weeks. Standard deviations (n = 3 ) are indicated as error bars.
commercial ELISA kit (Table 4). The ic-ELISA was able to detect 0.52 ng/ml of SEA in spiked milk. However, the ELISA kit was designed for a detection limit of 1 ng/ml, and the 0.52 ng/ml spiking concentration was below the LOD stipulated in the manufacturer’s instructions. These results suggest that the established ic-ELISA was highly sensitive and stable. The cross-reactivity values for SEA, SEG, SEB, SEC, E. coli heat-stable enterotoxin, and Shiga toxin were 90.9, 90.5, 0.2, 0.3,
Development of an Indirect Competitive Enzyme-Linked Immunosorbent Assay Based on the Multiepitope Peptide for the Synchronous Detection of Staphylococcal Enterotoxin A and G Proteins in Milk MINGYAN LIANG,'f TINGTING ZHANG,'f XUELAN LIU,1 YANAN FAN,1 SHENGLIN XIA,2 YIQING XIANG,1 ZIQI LIU,1 AND JINNIAN LI1* 1College o f Animal Science and Technology, Anhui Agricultural University, 130 Changjiang Avenue, Hefei 230036, Anhui, People’s Republic o f China; and 2Center o f Animal Epidemic Disease Prevention and Control, 460 Liming Avenue, Wuhu 241000, Anhui, People’s Republic o f China MS 14-323: Received 10 July 2014/Accepted 25 September 2014
ABSTRACT Staphylococcal food poisoning (SFP), one of the most common foodbome diseases, results from ingestion of staphylococcal enterotoxins (SEs) in foods. In our previous studies, we found that SEA and SEG were two predominant SE proteins produced by milkacquired S. aureus isolates. Here, a tandemly arranged multiepitope peptide (named SEAGepis) was designed with six linear B-cell epitopes derived from SEA or SEG and was heterologously expressed. The SEAGepis-specific antibody was prepared by immunizing rabbit with rSEAGepis. Then, an indirect competitive enzyme-linked immunosorbent assay (ic-ELISA) based on rSEAGepis and the corresponding antibody was developed to simultaneously detect SEA and SEG. Under the optimized conditions, the ic-ELISA standard curve for rSEAGepis was constructed in the concentration range of 0.5 to 512 ng/ml, and the average coefficients of variation of intraand interassay were 4.28 and 5.61% during six standard concentrations. The average half-maximal inhibitory concentration was 5.07 ng/ml, and the limit of detection at a signal-to-noise ratio of 3 was 0.52 ng/ml. The anti-rSEAGepis antibody displayed over 90% cross-reactivity with SEA and SEG but less than 0.5% cross-reactivity with other enterotoxins. Artificially contaminated milk with different concentrations of rSEAGepis, SEA, and SEG was detected by the established ic-ELISA; the recoveries of rSEAGepis, SEA, and SEG were 91.1 to 157.5%, 90.3 to 134.5%, and 89.1 to 117.5%, respectively, with a coefficient of variation below 12%. These results demonstrated that the newly established ic-ELISA possessed high sensitivity, specificity, stability, and accuracy and could potentially be a useful analytical method for synchronous detection of SEA and SEG in milk.
Staphylococcus aureus, a major cause of clinical and subclinical mastitis in dairy cows, is responsible for staphy lococcal food poisoning (SFP) as a result of the heat-stable staphylococcal enterotoxins (SEs) produced by enterotoxigen ic strains (2). Within 1 to 6 h after ingestion of food containing 100 to 200 ng of SEs, symptoms of acute gastroenteritis can be observed (22). In recent years, the presence of enterotoxigenic S. aureus in food animals and livestock products has increased in some countries; in particular, cows with subclinical S. aureus mastitis can release large quantities of S. aureus into their milk (12, 15, 18). Although pasteurization kills S. aureus cells, thermostable SEs generally retain their biological activity. Thus, because of the importance of these toxins in the public health and food sectors, continuous monitoring of S. aureus and SEs in raw milk is required. SEs belong to a large pyrogenic toxin family. More than 20 serological types of SE have been identified to date, including classical SEs (SEA through SEE), newly reported SEs (SEG, SEH, SEI, SEJ, SER, SES, and SET), and SE-like * Author for correspondence. Tel: +86-551-65785935; Fax: + 86-55165785935; E-mail: lijinnian2000 @ 163.com. f Both authors contributed equally.
toxins (SELK, SELL, SELM, SELN, SELO, SELP, SELQ, and SELU) (1 ,3 ,6 , 16). Previously, reports identified SEA and SED as the predominant SEs in cases of SFP, followed by SEB, SEC, and SEE (8,19). However, recent reports show a higher frequency of the newly reported seg gene compared with the classical SE genes in S. aureus isolated from several sources (27). In our previous study, over 90% of 114 S. aureus strains isolated from mastitic milk were found to carry at least one type of SE gene. The most frequently identified SE gene was seg (89%), followed by sea (40%). Moreover, 39% of the S. aureus isolates analyzed carried, synchronous ly, seg and sea genes (data not shown). These differences may be a result of variations in the origin and geographical locations of the isolates. Enzyme-linked immunosorbent assay (ELISA) has been extensively used for diagnostic and residue detection purposes due to its sensitivity, specificity, rapidity, and simplicity. Saunders and Bartlett (20) first detected SEA in foods with ELISA. Since that time, a number of variations of ELISAbased methods have been introduced. There have been some commercial ELISA test kits to detect classical SEs (A, B, C, D, E) (4,21), but these were not available for detecting novel
J. Food Prot., Vol. 78, No. 2
SYNCHRONOUS DETECTION ASSAY FOR STAPHYLOCOCCAL ENTEROTOXINS A AND G
363
TABLE 1. Amino acid sequence o f linear B-cell epitope from SEA and SEG proteins Potential
epitope no. SEA
1 2
3 4 5 SEG
1 2
3 4 5 0
Amino acid position" 85-88 138-141 156-163 166-173 211-218 24-31 37-46 80-84 120-127 141-149
Peptide sequence HSWY HDNN DGKQNTVP TVKTNKKN HTSTEPSV NSQPDPKI VSDYKSNKGT EYKSY KSEPDINQ NSSENRDK
Avg antigenic index6 1.26 1 .1
1.58 2.05 1.54 1.92 2 .0 1 1 .8 8
2.38 2 .0 1
The corresponding position of each peptide in the native SEA and SEG proteins, respectively. Antigenicity index was computed by the Janeson-W olf algorithm. Average antigenicity index is the mean value o f all amino acid antigen indexes among epitope peptides.
SEs. In the present study, a tandemly arranged multiepitope peptide (SEAGepis) from both SEA and SEG protein was designed, prepared, and used to immunize rabbit to prepare anti-SEAGepis antibody. With these key fundamental materials, an indirect competitive ELISA (ic-ELISA) based on the multiepitope peptide was then developed to synchro nously detect the SEA and SEG in raw milk. M A TER IA LS A ND M E TH O D S Bacterial strains and plasmids. The Escherichia coli host strain Rosetta (DE3) was purchased from China Institute of Veterinary Drug Control and was cultured in Luria-Bertani medium (Beijing Solarbio Science & Technology Co., Ltd., Beijing, People’s Republic of China) supplemented with 100 pg/ml ampicillin as needed. The pE T -32a(+ ) vector (Novagen, Madison, WI) was used for gene expression. Prediction and validation of immunodominant linear Bcell epitopes of SE proteins. The amino acid sequences o f SEA and SEG proteins o f the S. aureus standard strain ATCC25923 were used for linear B-cell epitope prediction. The secondary structure, flexible regions, hydrophilicity, surface accessibility, and antigenic index of both proteins were analyzed using the Protean program from the Lasergene software package (DNASTAR, Inc., Madison, WI). The peptide sequences with high hydrophilicity, accessibi lity, flexibility, and antigenicity, which were located in the P-tum and random coil regions of the protein secondary structure, were considered as potential linear B-cell epitopes. The three-dimensional (3D) structure o f the SEA and SEG proteins was simulated using the Molecular Operating Environment Software (version 2008, Chemical Computing Group, Montreal, Canada), and the potential B-cell epitopes were represented using the synchronize visualization program (version 2005, Accrelrys, San Diego, CA). To verify the prediction, six potentially immunodominant epitope peptides that had higher antigenicity indices and were exposed on the surface of the 3D structure at varying depths, as well as an irrelevant control peptide (NTNYANSQ), were artificially synthesized (purity >98% ) by GenScript Biotechnology (Nanjing, People’s Republic o f China) and detected by peptide-ELISA with purified rabbit anti-m altose binding protein (M BP)-SEA and anti-MBP-SEG antibodies pre pared as described by Cen et al. (7); commercial rabbit anti-MBP tag antibody and normal rabbit serum were used as the negative control.
Design of a tandemly arranged multiepitope peptide of SE proteins. Based on the sequences o f the six immunodominant linear
B-cell epitopes of SEA or SEG proteins (SEA156_163aa, SEA166_173aa, SEA3i ]— 2 i 8 aa. SEG37^ f,all, SEG|2f>-i27m- and SEG| 4 i_i4 9 aa) (Table 1) and the tandem design principle, a tandemly arranged multiepitope peptide (named SEAGepis) was designed with six immunodominant linear B-cell epitopes linked with two alanines (AA) and one tyrosine (Y) as an AAY spacer between the epitopes. To maintain the independence and antigenicity o f each epitope and to avoid the generation of a junctional epitope, the connection order of the six epitopes was optimized by predicting the hydrophilicity, antigenicity, and surface probability o f the tandemly arranged multiepitope peptide using the Protean program from the Lasergene software package (DNASTAR, Inc.) (24). The multiepitope peptide gene (318 bp) was artificially synthesized and cloned into the pUC57 vector by GenScript Biotechnology (Nanjing, People’s Republic of China).
Expression and purification of the recombinant multi epitope peptide. The gene-encoding multiepitope peptide (SEA Gepis) was amplified from the recombinant plasmid pUC57SEAGepis by PCR using the following primers: 5'-CAG AT GGATCCC ATGATAATAATCGTTTG ACCGA AG AG-3' (BamhU site underlined) and 5'- GTGTC AAGCTTTTATTTATCGCGTTCATTTTCAGAACTATTAAA-3' (HindUl site underlined). Puri fied PCR product and plasmid pE T -32a(+) were digested with BamHl and Hind III, and then a ligation reaction was set up. The recombinant plasmid pET-32a-SEAGepis was verified by restriction enzyme digestion and sequencing. The positive recombinant plasmid was transformed into Rosetta (DE3) E. co//-competent cells upon induction with 1 mM isopropyl-fLo-thiogalactopyranoside (SigmaAldrich, St. Louis, MO) at 37°C for 2 h. The bacterial cells were then harvested by centrifugation at 10,000 x g for 5 min at 4°C. After analysis by sodium dodecyl sulfate-polyacrylamide gel electropho resis (SDS-PAGE), the soluble rSEAGepis was purified using a highaffinity Ni-NTA resin column (GE Healthcare, Piscataway, NJ) according to the manufacturer’s instructions. Protein purity was estimated by SDS-PAGE analysis using bioinformatics software BandScan (http://www.bbioo.com/Soft/2005/75.htm), and the con centration of purified protein was determined with a Bradford protein assay kit (Tiangen Biotech Co., Ltd., Beijing, People’s Republic of China). In addition, a Western blot was performed as described previously (5) with the mixture of anti-SEA and SEG polyclonal antibodies (prepared as described by Cen et al. (7)) to confirm the immunoreactivity of the rSEAGepis.
Preparation and purification of rabbit anti-rSEAGepis antibody. Adjuvant Montanide ISA763A oil (Seppic, Paris, France) was mixed with rSEAGepis at a 3:7 volume ratio and then was
364
LIANG ET AL.
stirred at room temperature for 1.5 h to obtain a stable emulsion. A New Zealand white rabbit (weight 2,000 g) was vaccinated with rSEAGepis (500 pg per rabbit in 500 pi) by multipoint subcutaneous injections in the back. After 2 weeks, a booster was given with the same amount of antigen without the adjuvant. On days 0 and 28 after the primajy immunization, blood samples were collected via the marginal ear vein, and then the serum antibody titer was assayed by indirect ELISA. Briefly, polystyrene microtiter plates (Coming Inc., Coming, NY) were coated with 100 pi of the rSEAGepis (5 pg/ml), and unbound antigens were removed by washing three times with phosphate-buffered saline (PBS) containing 0.25% Tween-20 (PBST). After blocking with 5% skim milk, serial dilutions of immune serum and preimmune serum samples were placed in wells and were incubated for 2 h at 37°C. Wells were then washed three times with PBST. A 1:3,000 dilution horseradish peroxidaseconjugated goat anti-rabbit IgG antibody (Tiangen Biotech, Co., Ltd., Beijing, People’s Republic of China) was added to the wells and incubated for 1 h at 37°C. Next, the color was developed by reaction with o-phenylenediamine (Ziyi Reagent Co., Shanghai, People’s Republic of China) substrate. The plates were read at 490 nm using an ELISA reader. The endpoint titer was defined as the highest dilution that gave an absorbance 2.1 times higher than the preimmune serum. The anti-rSEAGepis antibody in serum was purified by the combination of saturated ammonium sulfate (Ziyi Reagent Co.) precipitation and a HiTrap Protein GHP column (Amersham Pharmacia Biotech, Uppsala, Sweden) according to the manufactur er’s instructions. The purity, concentration, and titer of the purified antibody were detected by SDS-PAGE, the Bradford method, and the aforementioned indirect ELISA, respectively. Optimization of assay conditions. To develop a highly sensitive and specific ic-ELISA, assay conditions, such as the concentration of the coating antigen, dilutions of the primary antibody, selection of the buffer solution and blocking reagent, and incubation time and temperature, should be optimized. Criteria used to assess the conditions included maximum absorbance (B0, usually kept in the range of 0.5 to 1.5 according to the Lambert-Beer law for spectrophotometric analysis) and half-maximal inhibitory concen tration (IC50; the concentration of standard solution producing 50% of inhibition) (23). Various concentrations of coating antigen rSEAGepis (from 0.5 to 1.5 pg/ml) in combination with dilutions of rSEAGepis-specific antibody (from 1:8,000 to 1:1,024,000) were tested using a checkerboard procedure. Different buffer solutions (carbonate-buffered saline and PBS) and blocking reagents (sterile skimmed milk powder, bovine serum albumin, and casein) prepared with PBS at a concentration of 5% were tested to determine the effects of these factors. The effects of incubation temperature (37°C, room temperature) and time (2 h, 1 h, 30 min, and 15 min) during blocking and immunoreaction steps were also examined. ic-ELISA procedure. Polystyrene microtiter plates (Coming Inc.) were coated with rSEAGepis solution (0.75 pg/ml, 100 pi per weft) and incubated overnight at 4°C. Free sites were blocked with 5% sterile skimmed milk powder prepared with PBS for 30 min at room temperature. Fifty microliters of the purified rSEAGepisspecific antibody (at 1:256,000 dilution) was added to each well, followed by the addition of an equal volume of rSEAGepis standard solution or samples, both in triplicate. After an incubation period of 2 h, a 1:3,000 dilution of goat anti-rabbit IgG-horseradish peroxidase (Tiangen Biotech Co., Ltd., Beijing, People’s Republic of China) was added to each well and was incubated for 1 h. Next, the color was developed by reaction with o-phenylenediamine (Ziyi Reagent Co). Three washings were performed after each incubation
J. Food Prot., Vol. 78, No. 2
step with PBST, and the incubation temperature was 37°C. The absorbances were measured using an ELISA reader at 490 nm, and all measurements were performed in triplicate. Calibration curve generation and intra- and interassay variation determination. The rSEAGepis standard solution at six concentrations (0.5, 2, 8, 32, 128, and 512 ng/ml) was used as competitor, and ic-ELISA was performed as described above. The results are presented as inhibition (%) = [1 — (B/B0)] x 100, where B = absorbance of rSEAGepis diluted solution or test sample solution, and B0 = absorbance in the absence of rSEAGepis competitor. The calibration curve was plotted using the logarithm of concentrations of rSEAGepis as the lateral coordinates and corresponding inhibition as the longitudinal coordinates. The concentration of SEA and (or) SEG in a milk sample could be calculated by interpolating their mean absorbance values in the calibration curve run in the same plate. To assess the accuracy and stability of the ic-ELISA, intra-assay variation was tested by determining the coefficients of variation (CVs) of six rSEAGepis samples of various concentrations in replicates of three across the microtiter plate. Interassay variation was determined by evaluating the rSEAGepis samples in triplicate on three different microtiter plates for as long as 3 weeks. The CVs were then calculated as CV (%) = (sample standard deviation/sample mean) x 100. Sensitivity and specificity of the assay. IC50 value and limit of detection (LOD) were key criteria for evaluating the sensitivity of ELISA (9). IC50 value was estimated by using the equation of the inhibition standard curve of rSEAGepis. LOD (the lowest analyte concentration at a signal-to-noise ratio of 3) was calculated by subtracting three times the standard deviation of the obtained absorbance from the mean absorbance and then calculating the corresponding concentration with the established equation of the standard curve. The specificity of the anti-rSEAGepis antibody was assessed by detection of cross-reactivity against several enterotoxins, including SEA, SEG, SEB, SEC2, E. coli heat-stable enterotoxin, and Shiga toxin (all kindly provided by the Anhui Academy of Medical Sciences, Anhui, People’s Republic of China). The test enterotoxins (0.5 to 512 ng/ml) were used in the ic-ELISA procedure as described above. The cross-reactivity values were calculated as cross-reactivity (%) = (IC50 of rSEAGepis /lC5o of tested enterotoxin) x 100. Detection of SEAGepis, SEA, and SEG in artificially contaminated milk. To demonstrate the applicability of the icELISA, spiked milk samples were tested. Commercial ultra-hightemperature pasteurized milk purchased from a retail store was centrifuged at 3,500 x g for 10 min at 4°C, and the upper layer of cream was removed by absorption. Once the fat was removed, the milk was spiked with rSEAGepis, pure SEA, and SEG at various levels from 6.25 to 400 ng/ml. Spiked samples were vigorously homogenized by vortex oscillation for 5 min and then were detected using the established ic-ELISA. Unspiked sterilized milk was treated in the same way and set as a blank. A0 assays were carried out in triplicate, and recoveries were calculated as recovery (%) = (Css — CUJC S) x 100, where Css and Cus are concentrations of the enterotoxins measured in the spiked sample and unspiked samples, respectively, and Cs is the spiked concentration. Comparison of established ic-ELISA with commercial ELISA kit. A commercial ELISA kit for SEA (TECRA Diagnostics, Roseville, Australia) was tested along with the established ic-ELISA to compare the sensitivity of two methods. Pure SEA was added to defatted milk (treated as above) at 0.13, 0.26, 0.52, 1.04, and 2.08 ng /ml, and then samples were tested with the two methods. The
J. Food Prot., Vol. 78, No. 2
SYNCHRONOUS DETECTION ASSAY FOR STAPHYLOCOCCAL ENTEROTOXINS A AND G
365
(A) Sam Hl
Hindi
l
l
5' -G G A T C C ^ S E A 15^163-A A Y -S E A 166,173 -A A Y -S E A 211,218-A A Y -S E G 37.46-A A Y -S E G 120.127-A A Y -S E G 141,149 -T A A -A A G C T T - 3'
S top codon
(B) scale
l H ydrophtlid ty Ptot-Kyte -Doolittle
] Antigenic Index-Jameson-Wolf
D Surface Probability Plot-Emtm
FIGURE 1. Design and analysis o f the multiepitope peptide SEAGepis. (A) Design o f the tandemly arranged multiepitope peptide SEAGepis. The tandem contains six epitopes o f SEA or SEG, an AAY linker, a stop codon, a BamHI restriction site at the 5' end and a Hind ill restriction site at the 3' end. The optimized order o f epitopes was SEA, 56 - 163-SE A i 66- i 7j^SEA 2 n^ 2 i 8-S E G j 7^f6-S E G i 2o-i2 TSEG i 4 i _i 49. (B) Analysis ofhydrophilicity, antigenicity, and surface probability o f the tandem. The amino acid sequence o f the tandem was analyzed by the Kyte-Doolittle algorithm fo r hydrophilicity, by the Janeson-Wolf algorithm fo r antigenicity index, and by the Emini algorithm fo r surface probability. D ashed areas indicate six epitopes. Each epitope maintained its independence and antigenicity. commercial ELISA kit was used according to the manufacturer’s protocols. All samples were tested in duplicate, and all assays were repeated in triplicate.
RESULTS Immunodominant B-cell epitopes of SE proteins and design of a tandemly arranged multiepitope peptide. The
secondary structure, flexible regions, hydrophilicity, sur face accessibility, and antigenic index of the SEA and SEG proteins were analyzed using DNASTAR’s Protean program, and 10 potential epitopes were screened for their suitability (Table 1). Among potential epitope peptides, SEA156_]63aa, SEA166_i73aa, and SEA2U_218aa as well as SEG37^ 6aa, SEG j2 0 - 127 aa> a n d SEG i i_i 9aa were exposed on the surface 4
4
FIGURE 2. SDS-PAGE and Western blot analysis o f the SEAGepis expressed in E. coli Rosetta (DE3). (A) Lane 1, protein molecu lar standard; lane 2, uninduced pET-32aSEAGepisIRosetta (DE3); lane 3, induced pET-32a- SEAGepis!Rosetta (DE3); lane 4, 1 mglml bovine serum albumin (control); lane 5, purified rSEAGepis. (B) Lane I, prestained protein ladder; lane 2, Western blot o f purified rSEAGepis with the mixture o f anti-SEA and SEG polyclonal antibodies.
366
J. Food Prot., Vol. 78, No, 2
LIANG ET AL.
M
1
2
SDS-PAGE analysis showed that a 31.5-kDa soluble protein band appeared in the original and purified expression products (Fig. 2A, lanes 3 and 5), and its size was consistent with the theoretically expected relative molecular weight of the His-SEAGepis protein. No protein band was found at the same position in the uninduced recombinant E. coli Rosetta (DE3) (Fig. 2A, lane 2). After purification with a highaffmity Ni-NTA resin column, the purity of the recombinant protein was about 95%, as estimated by Coomassie blue staining on 12% (wt/vol) SDS-PAGE. The expression level of rSEAGepis was found to be 17 mg/litcr of medium. As shown by Western blot analysis, a single band at about 31.5 kDa was observed (Fig. 2B, lane 2). These results indicated that the purified rSEAGepis possessed sufficient immunoreactivity and should be satisfactory as an antigen for an immunoassay.
3
116kDa 66.2kDa 45kDa 35kDa
25kDa
18.8kDa 14.4kDa FIGURE 3. SDS-PAGE analysis of rabbit anti-rSEAGepis anti
body. M, protein molecular standard; lane 1, unpurified rabbit immune serum; lane 2, rabbit immune serum treated with saturated ammonium sulfate; lane 3, rabbit anti-rSEAGepis antibody, which was purified by protein G agarose affinity chromatography, under reducing condition. of the 3D structure model at varying depths. The results of peptide-ELISA showed that the six synthetic potential epitope peptides could be specifically recognized by the rabbit antiMBP-SEA or anti-MBP-SEG antibodies, but they did not bind with the commercial rabbit anti-MBP tag antibodies and normal rabbit serum. Any nonspecific reaction was also excluded because the irrelevant control peptide did not react with the antibodies against SEA, SEG, or the MBP tag peptide. Therefore, the six peptides represented immunodom inant linear B-cell epitopes of SEA or SEG. As shown in Figure 1, the tandemly arranged multiepitope peptide SEAGepis containing six immunodominant B-cell epitopes of SEA or SEG, five AAY linkers, a termination codon, a BamHl restriction site at the 5' end, and a HindUl restriction site at the 3' end was designed successfully. The opti mized connection order of these epitopes was SEA 156_i63 S E A i 6 6 _ 17 3 - S E A 2 i i _ 2 l 8 - S E G 3 7 ^ t 6 - S E G i 2 0 - 1 2 7 “ S E G i 4 i _ i 4 9 ,
in which each epitope maintained its independence and antigenicity as analyzed by the Lasergene Protean program. Purified rSEAGepis protein. To obtain the rSEAGepis, E. coli Rosetta (DE3) cells harboring pET-32a-SEAGepis were induced with isopropyl-(3-D-thiogalactopyranoside.
Antibody characterization and optimization of assay conditions. Antisera were collected and purified after 4 weeks of primary immunization. As shown in Figure 3, purified antibody exhibited two bands on SDS-PAGE; the molecular weight of the heavy chain and light chain of IgG were about 50 and 25 kDa under reducing condition with (3mercaptoethanol (Ziyi Reagent Co.). The titer of purified antibody was 1:819,200,1:409,600, and 1:409,600 as detected by ELISA using rSEAGepis, SEA, and SEG as coating antigens, respectively. According to the criteria with a higher Amax value and lower IC50 value, the immunoassay was optimized. Finally, carbonate-buffered saline and 5% skim milk were used for coating and blocking, respectively, in the assay. The optimal coating antigen concentration was 0.75 pg/ ml, and dilution of the primary antibody was 1:256,000, producing a maximum absorbance of approximately 1 in the absence of analyte (Table 2). The appropriate temperature of blocking and immunoreaction was 37°C, and incubation times of primary antibody, secondary antibody, and color develop ment were 2 h, 1 h, and 15 min, respectively. Specificity, sensitivity, and specificity of the assay. The stability of the assay was tested by running the ic-ELISA procedures three individual times. Figure 4 shows the average calibration curve of rSEAGepis generated from three assays: y = —38.286.x + 125.15, with a linear range from 0.5 to 512 ng/ml (R1 = 0.991, n = 3). As shown in Table 3, the average CVs of intra- and interassay were 4.28 and 5.61% during six standard concentrations. The average IC50 was 5.07 ng/ml, and the LOD was 0.52 ng/ml. Moreover, the established ic-ELISA was tested in comparison with a
TABLE 2. Checkerboard titration of rSEAGepis antigen versus anti-rSEAGepis primary antibody Fold dilution of primary antibody ( x 10-3) Concn of rSEAGepis coating antigen (pg/ml)
1:8
1:16
1:32
1:64
1:128
1:256
1:512
1:1024
1.5 1.25 1.0 0.75 0.5
2.84 2.84 2.75 2.84 2.70
2.32 2.47 2.42 2.49 2.34
1.95 2.11 2.09 2.04 1.90
1.39 1.51 1.57 1.73 1.47
1.13 1.13 1.22 1.31 1.17
0.86 0.95 0.88 1.00 0.93
0.72 0.72 0.64 0.57 0.68
0.40 0.41 0.44 0.42 0.44
J. Food Prot., Vol. 78, No. 2
367
SYNCHRONOUS DETECTION ASSAY FOR STAPHYLOCOCCAL ENTEROTOXINS A AND G
TABLE 4. Comparison o f sensitivity o f the established ic-ELISA with that o f the commercial ELISA kit fo r detection o f SEA in spiked milka
Test results Spiked concn (ng/ml) 0.13 0.26 0.52 1.04 2.08
ic-ELISA
Commercial ELISA kit
—
—
+ + +
+ +
a + , detection of SEA in spiked milk; —, no detection.
DISCUSSION FIGURE 4. The average standard calibration curve o f the established ic-ELISA fo r rSEAGepis run three times fo r as long as 3 weeks. Standard deviations (n = 3 ) are indicated as error bars.
commercial ELISA kit (Table 4). The ic-ELISA was able to detect 0.52 ng/ml of SEA in spiked milk. However, the ELISA kit was designed for a detection limit of 1 ng/ml, and the 0.52 ng/ml spiking concentration was below the LOD stipulated in the manufacturer’s instructions. These results suggest that the established ic-ELISA was highly sensitive and stable. The cross-reactivity values for SEA, SEG, SEB, SEC, E. coli heat-stable enterotoxin, and Shiga toxin were 90.9, 90.5, 0.2, 0.3,
