DEVELOPMENT AND EVALUATION OF ELISA KITS BASED ON PROTEOMICS TECHNOLOGY FOR DETECTION AND SEMI- QUANTIFICATION OF MEAT SPECIES ADULTERATION

Results: Evaluation of the developed kits in testing crooked reference sample mixtures revealed that the developed ELISA kits showed 100% specificity with no cross-reactivity detected for the non-target species. In terms of sensitivity, the developed ELISA kits were able to consistently detect pork in the sample mixtures and horses at levels down to 5% w/w. Although sheep and chicken were detected at levels as low as 1% w/w. The beef-specific ELISA test showed greater sensitivity.


INTRODUCTION
Food labeling legislation demands detailed and reliable information concerning the content of any food product; this information is likely to influence consumer choice based on moral, religious, ethical, dietary, or health factors. Enforcement of food labeling regulations requires robust and validated analytical methods to verify the authenticity of meat-based foods. At present, meat authentication is largely reliant on DNA-based detection technologies (O'Mahony, 2013). PCR methods offer great specificity and sensitivity, even in some heattreated foods and feed. However, both nuclear and mitochondrial DNA are sensitive to heavy processing methods (Sakalar et al., 2012) and therefore DNA-based technologies cannot always be robustly applied to highly processed foods, especially when quantification is necessary. ELISA methods are also used for meat speciation (Chen & Hsieh, 2000), but there are also a number of limitations with ELISA-based testing such as poor specificity, matrix interference, or the effect of different food processing procedures on antigen availability.
Proteins are substantially more diverse in terms of their physicochemical properties when compared with nucleic acids and some will survive the heating and other conditions associated with highly processed foods. Therefore, many proteins will suffer only partial degradation under such treatments, resulting in large fragments still present after processing. The primary structures (amino acid sequences) of these protein fragments can be analyzed using proteomics tools to determine the species of origin. Proteomics workflows have been developed in recent years using high-resolution MS to determine the species of origin of animal-derived ingredients in food and feed products (Montowska et al., 2014). This study aimed to use the proteomics approach to identify the unique peptide markers for five meat species to prepare specific and sensitive antisera against each to be used in developing ELISA kits to detect meat species adulteration and evaluate the developed kits in testing cooked reference sample mixtures.

Sample collection
Different meat species standard samples of known origin (beef, pork, horse, sheep and chicken) were supplied by the General Organization of Veterinary Services, Egypt. Other meat species commercial samples (raw meat, cooked meat, and thermally processed) were obtained from different butchers and supermarkets. All samples were stored at -20°C until used.

Preparation of thermally processed meat laboratory:
The different meat species were laboratory prepared to induce the industrial processing effects. Meat samples were minced firstly. Homogenized meat was cured in NaCl salt (20% w:w) for 4 days at 4°C. Following this curation, sufficient water was added to cover the meat samples and the samples were simmered for 20 min on a hot plate with stirring using a magnetic bar. The laboratory-cooked meat samples were then autoclaved with high temperature and pressure at 121°C for 15 min. These processed meat samples were stored at -20°C for further use.

Preparation of meat species mixture:
For the determination of LOD, different meat species mixtures were prepared by weighting respective amounts (10 mg) of meat to obtain beef samples spiked with various weight ratios of other meat species (10%, 5%, 1% and 0.5% w/w) in a 2 mL Eppendorf tube and then laboratory-processed (table 1).

Method of meat proteins extraction from both standard and commercial samples:
For each 1 gram of meat species sample, 10 ml of 50 mM Tris-HCl buffer, pH 8.0 were added then the sample was homogenized by using a Vortex for 2 minutes. The homogenate sample was centrifuged at 10,000 g for 20 minutes at 4 C. The supernatant that constituting the sarcoplasmic extract was discarded in which most of the sarcoplasmic proteins become denatured and insoluble with thermal processing. most of the sarcoplasmic proteins. The collected pellet was suspended in 10 ml of 50 mM Tris-HCl, pH 8.0, containing 6 M urea and 1 M thiourea, and then homogenized in a Vortex for 5 minutes for extracting the myofibrillar proteins. Most myofibrillar proteins remained in the extract in both fresh and thermally processed meats. The homogenate was centrifuged at 10,000 g for 10 minutes again. The supernatant containing the myofibrillar extract was collected and filtered through 0.45 m membrane filters (Millipore) before use. The protein concentration of the final extract was determined using the Bradford method.

Development of anti-species antibodies:
Construction of a database of species-specific peptides: All database searches were performed based on our selection criteria as the following; 1. Myofibrillar proteins: Myofibrillar proteins constitute a more reliable and robust source of peptide biomarkers because most of the sarcoplasmic proteins undergo denaturation after heat treatment making them insoluble. In contrast, most myofibrillar proteins remained in the same extract in both fresh and cooked meats, giving rise to the generation of the same peptide biomarkers in all cases. 2. Well distributed in muscles and other organs: myosin 3. Survive industrial processing to identify proteins present in heavily processed food products. 4. Solubility in salt: 5. Long amino acid chains: nearly 18-20 amino acids Using bioinformatics to compare amino acid sequences across a range of relevant meat animal species to find unique marker peptides: We created a database of unique peptides for beef, pork, horse, sheep, chicken, and rabbit using the UniProt database (www.uniprot.org). The full amino acid sequences of the proteins identified from each species were compared across all vertebrates to select a list of species-specific peptides for beef, pork, horse, sheep, chicken.
Initially, we searched against a single fasta file containing the UniProt reviewed sequences for species proteins to produce a conclusive list of identified abundant proteins. The amino acid sequences of selected proteins that were most confidently identified across the six species were then obtained for all species of interest, namely cow (Bos taurus), pig (Sus scrofa), horse (Equus caballus), sheep (Ovis aries), chicken (Gallus gallus), and rabbit (Oryctolagus cuniculus), from either UniProt or through BLAST (the Basic Local Alignment Search Tool) searching the NCBI non-redundant (NCBInr) database. This list of protein sequences was then used to construct a database containing only the unique tryptic peptides from each protein across the species. Firstly a peptide-picking algorithm was used to compare the in silico tryptic peptides of each protein across the different species entries. These unique tryptic peptides were then searched using BLAST to ensure that they were truly unique between all isoforms of all proteins currently in the databases for the six species (cow taxid:9913, pig taxid:9823, horse taxid:9796, sheep taxid:9940, rabbit taxid:9986, chicken taxid:9031).

Synthesis of multiple antigen peptides (MAP) to be used as an immunogen:
The immunogens used were synthetic unique marker peptides of the five species. They were constructed in multiple antigen peptide (MAP), tetra branched by: Hangzhou Dangang Biological Technology Co., Ltd., China.

Performing immunization protocol to develop anti-species antibodies:
This protocol is optimized to produce antiserum in the shortest possible time. The immunization schedule was performed using three NZW SPF rabbits (New Zealand White rabbits that are Specific Pathogen-Free) for each immunogen. Rabbits were injected subcutaneously (SC) with the immunogen as an emulsion in Complete Freund's Adjuvant (CFA) or Incomplete Freund's Adjuvant (IFA) as shown in table (2).

Table (2): Immunization protocol
Extraction and purification of the prepared antisera: After collecting blood, it was allowed to clot for 60 min at 37°C or over night at 4°C. We separated the clot from the sides of the tube (ringing) using a pasteur pipette and centrifuged at 10000 xg for 10 min at 4°C to separate the serum. Serum was stored at -20°C till used.

Development of ELISA kits and evaluation of its specificity and sensitivity:
The extracted and purified antisera were diluted in phosphate-buffered saline (PBS) (0.1 M, pH 7.4) and were coated (100 µL/well) and incubated overnight at 4 °C. Then, the 96 well plates was washed with PBS then blocked using BSA (5%) and incubated at room temperature for 1 h. Subsequently, the extracted meat protein from cooked reference sample mixtures was added (100 ul/well) and the plates were incubated at 37 °C for ½ h followed by a washing step. Diluted HRP-conjugated anti-pig IgG (1:1000) was added (100 µL/well) and the plate was incubated at 4 °C for 1 h followed by a washing step. As a final step, H2O2/TMB (200 µL/well) liquid substrate system for ELISA (1/20 diluted in citrate buffer pH 5.0) was added to each well and was incubated for 15 min in the dark. Sulfuric acid 0.1 M (50 µL/well) was used to stop the enzymatic reaction and the absorbance was measured at 450 nm using an ELISA plate reader (Biotek).

Identification of unique peptide markers:
We used bioinformatics to compare amino acid sequences across a range of relevant meat animal species to find heat-stable, species-specific peptides as a unique peptide marker for beef, pork, horse, sheep, and chicken. Our preliminary selection was shown in table (3). Based on our selection criteria previously mentioned, we construct a unique peptide marker for each species to develop antisera against each (table 4).

Specificity and Sensitivity:
As shown in Table ( 5 & 6), the developed ELISA kits showed 100% specificity during cooked reference sample mixtures testing, with no cross-reactivity detected for the non-target species in the beef/Pork, beef/horse, beef/sheep and beef/chicken mixtures.
In terms of sensitivity, the developed ELISA kits was able to consistently detect pork in the sample mixtures and horse at levels down to 5% w/w (    *Lowest consistent detection level (w/w) in a binary mixture of pork and beef. **Percentages are based on a total of 21 samples tested in duplicate with each assay.

Discussion:
Meat and meat-based products constitute a significant percentage of the human diet worldwide. Beef is the top red meats consumed. The contribution of beef to the global food trade id is 59.0 million tons of beef and veal that expected to be produced globally (USDA, 2016). Meat species may be easy identifiable when sold as whole cuts; however, processing conditions and techniques may change the texture, flavor, and color of meat, making it difficult to authenticate species in food products containing processed meats (Cawthorn et al., 2013). The inability to readily identify meat species in processed products gives rise to the potential for species mislabeling, in which one species is substituted either partially or completely for another species. In most cases, species mislabeling is a kind of economically motivated adulteration (EMA), in which a product is intentionally mislabeled for reasons of economic gain (FDA, 2009). Due to price differences among meat species, there are economic incentives accompanied with meat species mislabeling. For example, the average retail value of beef in 2015 was US$13.31/kg while the average retail value of pork for the same year was US$8.49 (USDA, 2016), resulting in an economic incentive for the substitution of pork for beef in a processed product. Substitution can occur at any point in the supply chain, from the slaughterhouse up until the point of sale (Premanandh, 2013). In addition to meat products may be mislabeled due to cross-contamination when processing different types of meat on the same equipment, as has been suggested in previous Mislabeling of meat species in processed foods has a number of potentially detrimental effects such as the exposure of consumers to meat allergen risks, infringement of religious practices, and economic deception (Ballin, 2010;Kane & Hellberg, 2016;Okuma & Hellberg, 2015). A number of studies have detected undeclared species in processed meat products such as deli meats, minced meats, burger patties, sausage and canned meats (Ayaz et al., 2006;Cawthorn et al., 2013;Di Pinto et al., 2015;Flores-Munguia et al., 2000;Okuma & Hellberg, 2015). Some of the most commonly undeclared species within these products were beef, pork, poultry, and sheep. These findings are concerning from a religious point of view, as some religions prohibit the consumption of beef and/or pork (Sattar et al., 2004). In addition to the studies mentioned above, horsemeat was detected as an undeclared ingredient in numerous beef products in the 2013 horse gate scandal in Europe (National Audit Office, 2013). Adulteration and misbranding of meat products is prohibited globally. Adulteration can occur when an ingredient has been completely or partially omitted, and/or if any ingredient has been substituted within a meat product. A misbranded product includes one whose labeling is false or misleading or if it is offered for sale under the name of another food.
To determine if meat species have been partially or completely substituted for cheaper alternatives in processed food products, DNA or protein-based methods are often used (Ballin, 2010). Two of the most commonly used methods for this purpose are enzyme-linked immunosorbent assay (ELISA), a protein-based method, and real-time polymerase chain reaction (PCR), a DNA based assay. Most of the regulatory bodies rely on ELISA method for identifying animal species in cooked and canned meat (USDA, 2005). ELISA technique is recognized as being sensitive and robust and it has been used in several studies for detecting low levels of a target species within mixtures (Yamamoto et al., 2015). Therefore, the objective of this study was to develop ELISA kits for detection of meat species adulteration and evaluate the developed kits in detection of meat species in cooked reference sample mixtures.
This work has identified muscle proteins that are able to persist in meat preparations that have undergone heavy processing. Bioinformatics analysis enabled us to obtain an initial list of peptides unique to beef, pork, horse, sheep and chicken. The main focus of this study was the identification of unique peptide markers for the five species to develop specific antisera for each species with no cross reactivity with others. Our preliminary selection has shown in table (3). Based on our selection criteria previously mentioned, we construct unique peptide marker for each species to develop antisera against each (table 4).
The developed ELISA kits showed 100% specificity during cooked reference sample mixtures testing, with no cross-reactivity detected for the non-target species in the beef/Pork, beef/horse, beef/sheep and beef/chicken mixtures. In terms of sensitivity, the developed ELISA kits were able to consistently detect pork in the sample mixtures and horse at levels down to 5% w/w (Table 5; Sample 10 & 11). Although sheep and chicken was detected at levels as low as 1% w/w (Sample 16 & 17), this result was found with all the duplicate samples.
The beef-specific ELISA test showed greater sensitivity. Our results are matched with previous studies using ELISA assay with monoclonal antibodies (mAbs) that have reported lower detection limits for different meat species. For example, studies using a sandwich ELISA with porcine-specific mAbs have reported the ability to detect pork at levels of 0.05 -0.5% w/w in various meat mixtures (Chen & Hsieh, 2000;Liu et al., 2006). Similarly, Yamamoto et al. (2015) were able to detect beef at levels of 0.1% w/w in a beef and pork meal mixture using a sandwich ELISA with two bovine-specific mAbs. Future studies are still required for validation of the developed kits and comparing its results with the commercially used kits before entry into the production phase.

CONCLUSION
Overall, the developed ELISA kits in this study was a specific and promising tool to detect meat species in the cooked reference sample mixtures and sensitive method for beef, followed by sheep and chicken, followed by pork and horse species. The results of this study also suggest that ELISA was found to be low time consuming and easy to perform. Further studies are still required for validation of the developed kits and comparing its results with the commercially used kits before entry into the production phase.