Meatball model of porcine DNA detection by TaqMan probe real-time PCR

*Sajali, N., Ting, S.M.L., Koh, C.C., Desa, M.N.M., Wong, S.C. and 3 Abu Bakar, S. Centre for Research of Innovation and Sustainable Development, School of Engineering and Technology, University of Technology Sarawak, 96000 Sibu, Sarawak, Malaysia Halal Products Research Institute, Universiti Putra Malaysia, 43400 Serdang, Selangor Darul Ehsan, Malaysia Department of Biomedical Science, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor Darul Ehsan, Malaysia Department of Science and Technology, Faculty of Humanities, Management and Science, Universiti Putra Malaysia Bintulu Sarawak Campus, Nyabau Road, 97008, Bintulu, Sarawak, Malaysia


Introduction
Food fraud is a global phenomenon that affects consumers worldwide. Food fraud, or adulteration, is defined as the practice of deliberately diminishing the quality of food products, either by substitution, the addition of inferior, and cheaper ingredients or by removing some valuable ingredients (Jha, 2016). Fish and meat products were among the top two product categories with reported cases of food fraud in 2018. In addition, the main type of food fraud related to these products is mislabelling which accounts for 41.89% of instances, followed by adulterations involving the replacement of high-priced meat with lower-priced meat, which accounts for 19.25% (European Commission, 2018). At present, the intentional or unintentional adulteration of meat-based products containing pork DNA has attracted attention among Muslims and Jews due to religious concerns. Furthermore, this practice often leads to unfair trade competition (Raharjo, Chudori and Agustina et al., 2019).
In this regard, the authentication of meat-based food products is critical, due to legislation, public health and religious reasons (Omran et al., 2019). Products' labelling which indicates their origin and approved formulation (Fajardo et al., 2010) is also taken into account for authentication. Thus, several analytical methods have been developed to identify the species and analyse the authenticity of food products. Commonly used analytical methods, such as DNA-based techniques, are more likely to be used for fraud detection in raw meat species and heat-processed meat products because DNA has better thermal stability during food processing, and has distinct DNA sequences of specific animal species (Rashid et al., 2014). Several DNA-based techniques have been used in meat species eISSN: 2550eISSN: -2166 © 2022 The Authors. Published by Rynnye Lyan Resources FULL PAPER authentication, including randomly amplified polymorphic DNA (Huang et al., 2003), restriction fragment length polymorphisms (Doosti et al., 2014), species-specific PCR (Kitpipit et al., 2013) and real-time PCR (Druml et al., 2016). Hamzah et al., (2014) reported that conventional PCR can be used to detect porcine DNA in meat products. However, conventional PCR presents some drawbacks, such as it requires end-point analysis. Further, it yields a qualitative outcome, therefore it is less sensitive. In this study, the TaqManbased real-time PCR assay is used to detect porcine DNA as it is more sensitive, provides a quantitative outcome, while eliminating the need for post-PCR processing (Yusop et al., 2012).
Known as bebola in Malaysia, meatballs are one of the most preferred meat-based products (Rohman et al., 2011) among consumers due to their rich source of animal protein and carbohydrates (Huda et al., 2010). Meatballs typically comprise 90% chicken or beef (Aina et al., 2019). Some manufacturers seek to maximise their profits by substituting halal beef with cheaper pork when producing meatballs (Montowska and Pospiech, 2011). Previous studies have documented the use of porcine DNA detection in meatball models targeting various mitochondrial genes, such as NADH dehydrogenase subunit 1 (ND-1) , and D-loop (Orbayinah et al., 2020). The primers and probe sequences designed from the cytochrome B gene sequence have higher interspecies variability compared to 12S rRNA, ND2, ATPase6 and 16S rRNA (Mohamad et al., 2013). New species-specific primers and probe sequences were designed in this study, amplifying 121 bp at different regions of the gene, from regions 516 to 637 in the nucleotide of the cytochrome B gene. This model was different from previous studies, in terms of the porcine meat percentage incorporated in the meatball and its exploration of whether the application of a prolonged heat treatment leads to a significant difference in detection capabilities. Furthermore, the meatball model developed in this study was designed to reflect the complex composition of commercial meatballs by enabling the accurate quantification of targets in comparison with binary mixtures of two different species. The specificity and sensitivity of the designed primers and probe sequences on porcine meat DNA were indicated through validation tests with raw meat species and spiked beef and chicken meatball models, respectively.

Samples
Raw meats (pork, chicken, mutton, duck, quail, buffalo and beef) were purchased from a local butcher shop in Sibu, Sarawak, Malaysia. Rabbit and deer were obtained from Universiti Putra Malaysia (UPM), Bintulu Campus, Sarawak, Malaysia. Spiked meatball models were prepared according to the method detailed by Razzak et al. (2015). Raw meats were used for the determination of specificity, and the spiked meatball model was used in the sensitivity of the designed primers and probe.

In-silico design of primers and probe sequences
The nucleotide sequence of the mitochondrial cytochrome B (cytb) gene for pork, chicken, mutton, cow, duck, buffalo, deer, rabbit and quail was obtained from the National Centre of Biotechnology Information (NCBI) at https://www.ncbi.nlm.nih.gov/. The sequences of the mitochondrial cytb for various animals were compared and aligned using ClustalW Multiple Alignment tool from Bioedit software (version 7.2.5). The pork-specific primer and probe designated as Pork-F: CAAAGCAACCCTCACACGAT; pork probe: 5HEX -TTACCGCCCTCGCAGCCGTA-3IABkFQ and Pork-R: AGATTCCGGTAGGGTTGTTG were used in this study. These sequences were designed to amplify 121 bp of the Sus scrofa domesticus (pork) mitochondrial cytb. The designed primers and probes were verified using the Basic Local Alignment Search Tool (BLAST) in NCBI to ensure the specificity of the designed primers and probe. Mitochondrial 18S rRNA primers and probe sequences that amplified 140 bp designed by Rojas et al. (2011) were included to serve as an internal control for assay validation. The designed primers and probe were synthesised by Integrated DNA Technologies, Singapore.

DNA extraction and quantitation
DNA from raw meats and meatballs were extracted using DNeasy Mericon Food Kit (Qiagen, German). The extraction process was performed according to a standard protocol for raw meat samples and small fragments protocol for meatball models, with initial samples of 200 mg. The protocols were modified slightly to optimize the quality and quantity of extracted DNA. The purity and concentration of the DNA were determined using Genesys TM 10S UV-Vis Spectrophotometer (Thermo Fisher Scientific, USA). The absorbance of the diluted DNA sample at 260 nm and 280 nm were recorded. DNA integrity was verified by using 1% agarose gel electrophoresis. The gel electrophoresis was performed at 120 V for 30 mins.

Construction of real-time PCR standard curve
Porcine DNA was diluted into five DNA 10-fold serial dilutions (10 -1 until 10 -5 ) for the construction of the standard curve. qPCR assay was carried out based on the eISSN: 2550-2166 © 2022 The Authors. Published by Rynnye Lyan Resources FULL PAPER protocol for GoTaq Probe qPCR Master Mix product information (Promega, USA), with a slight modification in terms of total volume. Master Mix containing 5 μL of 5× GoTaq Probe qPCR Master Mix (Promega, USA), 0.25 μL of 20 μM of each forward and reverse primer, 0.125 μL of 20 μM of hydrolysis TaqMan probe and 3.375 μL nuclease-free water were prepared. Then, 1 μL of DNA template was added to each respective PCR tube which made up a final volume of 10 μL. Five tubes containing DNA templates with 10-fold serial dilutions were run simultaneously in triplicates. The reaction mixtures without template or NTC were included as the negative control. The PCR tubes were centrifuged briefly to ensure the reagent components remain at the bottom of the tube prior to loading into DTprime Real-time thermal cycler (DNA-Technology, Russia). Real-time PCR was operated according to the recommended thermal cycling condition as such, GoTaq activation at 95°C, 2 mins for 1 cycle; denaturation at 95°C, 15 s; annealing/extension at 60°C, 1 min, both for 40 cycles. The linearity, efficiency and sensitivity of the qPCR were determined by plotting the standard curve. The standard curve was plotted as threshold cycle (C t ) values against the logarithm of the template concentration.

Determination of specificity of the designed primers and probe sequences
The specificity of the designed primers and probe for TaqMan qPCR assays were determined using the DNA template of the raw meat samples. The reaction and thermal cycling conditions were carried out according to the method described in Section 2.4. A no-template control (NTC) containing nuclease-free water was added to the NTC tube which acted as a negative control. Internal control or positive control was prepared by adding 18S rRNA primer and probe to the master mix. qPCR reaction was done in triplicates for two independent studies.

Preparation of spiked meatball model
The chicken and beef meatball model was prepared according to Razzak et al. (2015), with a slight modification in the amount of spiked pork meat. Table 1 indicates the formulation for chicken and beef meatballs. The prepared chicken and beef meatballs were spiked with 10% (w/w) of pork meat individually and collectively to make ≥35g of meatballs per piece. The amount of minced chicken and beef used to make each meatball was similar. Another set of prepared meatballs was boiled at 100°C for 2 hrs and 30 mins. The DNA extraction and quantitation were done based on the method described in Section 2.3.

Determination of detection limit of designed primers and probe sequences
The detection limit was determined using DNA extracted from raw and heat-treated beef and chicken spiked meatball models. DNA was extracted from beef and chicken spiked meatballs containing 10% (w/w) of pork meat. Then, 10 ng/µL DNA extracted from each beef and chicken spiked meatball was serially diluted 10fold from 10 ng/µL to 0.0001 ng/µL. The step was repeated with a heat-treated spiked meatball to determine the effect of heat treatment on this assay. qPCR reaction was done in triplicates for two independent studies. The reaction and thermal cycling conditions were carried out according to the method described in Section 2.4. The detection rate of the designed primers and probe were determined using Equation 1 (Cai et al., 2017). The detection rate was calculated for each meatball spiked model and pork meatball model.
Where the number of tubes having positive detection for porcine DNA in sample replicates was calculated; PP is positive detection for porcine DNA in pork sample replicates; BP is positive detection for porcine DNA in beef sample replicates; CP is positive detection for porcine DNA in chicken sample replicates; N is the total number of tubes of sample replicates; DR is the detection rate.

Statistical analysis
Statistical analysis was conducted using Student's ttest (IBM SPSS Statistics Version 20) to determine the differences between the C t values of the raw spiked meatball models and heat-treated spiked meatball models.
(Positive detection) PP, BP, CP / (N) = DR (1)  Figure 1 shows multiple alignments of the sequences of the mitochondrial cytb for various animals using ClustalW Multiple Alignment tool (Bioedit software version 7.2.5). The non-conserved region of cytb sequences of Sus scrofa domesticus among the others was selected as primers and probes to ensure the primers and probes were specific for pork. In this study, Taqman probe qPCR assay was designed to amplify a short amplicon length of 121 bp. This is in agreement with Lim et al. (2011) who reported that amplification of less than 150 bp increases qPCR efficiency. Thus, a shorter amplicon provides higher chances of detection, reduces the chances of secondary structure formation (Toouli et al., 2000), and ensures the applicability of sequence detection in processed food products, particularly those that have been subjected to thermal treatment. The incorporation of Taqman probe with species-specific primer sequences in food matrices improves the consistency of the assay (Köppel et al., 2011).

Quantification of DNA quantity and quality
The A 260 /A 280 ratio provides an insight into the purity of the extracted DNA. In this study, the results were reported based on the data from two independent studies. The possible contaminants of DNA comprise protein, polyphenols, polysaccharides and other PCR inhibitors (Piskata et al., 2019). Four samples namely lamb, duck, rabbit and deer were reported with A 260 /A 280 absorbance ratio that is lower than 1.7. Meanwhile, no DNA sample was reported for an absorbance ratio of more than 2.0. A 260 /A 280 higher than 2.0 usually indicates RNA contamination (Piskata et al., 2019) whereas a ratio below 1.7 indicates protein contamination. The residual impurities carried over from the DNA extraction process such as ethanol or phenol can also lower the A 260 /A 280 of the extracted DNA. Laube et al. (2007) reported that the DNA yield of the extracted DNA can be affected by the sources of tissues. The DNA extracted from fatty tissue contains a lower concentration of DNA compared to DNA extracted from the kidney, liver, heart and tendon tissues. In addition, butter used in producing the meatball model may lower the DNA yield as high lipid content affects DNA extraction (Costa et al., 2010). Thus, optimisation has been done in this study to obtain a high yield of DNA concentration sufficient for qPCR assay.
In this study, the quality of the extracted DNA can be verified by electrophoresis analysis through a 1% agarose gel. All the DNA extracted from the raw meat was of good quality as the DNA band remained intact. For mutton, duck, rabbit, buffalo and quail sample, the DNA band appeared thicker as the concentration of the DNA was higher (data not shown). However, all the DNA extracted from heat-treated meatball models and raw spiked chicken meatball models were degraded as it appeared as an expanded smear with some of the fragmented DNA band. This is in agreement with Piskata et al. (2017) who stated that heat, physical or chemical treatment may negatively affect the quality and quantity of the extracted DNA as it will result in smearing and fragmentation of the extracted DNA. Furthermore, Malentacchi et al. (2014) reported that the sample quality, sample age, repeated freezing-thawing, retention to the tubes and storage condition can significantly affect the DNA integrity. Instead, newly short amplicon primer and probe sequences designed in this study are workable to detect porcine DNA in the spiked meatball model. This could be due to the availability of certain DNA fragments which may contain the target sequences, ensuring successful detection by the short amplicon. Hence, these sequences provide an alternative by which the chances of detection could be increased through amplification of the DNA fragment from heat-treated food products.

Construction of standard curve for primers and probe
The prerequisite step for conducting qPCR assays is to construct a standard curve for the target gene. The performance of the subsequent qPCR assays can be FULL PAPER estimated through a standard curve and it is used to determine the efficiency of the designed primers and probe. The optimal correlation coefficient (R 2 ) of a standard curve should be more than 0.99 as it measures how well the data fit on the standard curve. The values of the slope between -3.1 and -3.58 are acceptable, which correspond to the qPCR amplification efficiency ranging between 90 to 110% (Bio-Rad Laboratories, 2006;Science Squared, 2016). Based on Figure 2, C t was plotted against log 10 of five different 10-folds DNA dilutions (10, 1, 0.1, 0.01 and 0.001 ng/μL) without detection for the non-template control (NTC), which indicates the absence of contamination and primerdimer. The amplification of five DNA 10-fold serial dilutions showed a good linear regression and correlation coefficient (R 2 ) of 0.997. The assay indicates an acceptable value of 95% efficiency which corresponds to -3.431 slope value. Thus, the reliability of the designed cytb primers and probe in detecting porcine DNA at different concentrations was established. In addition, Figure 3 shows the C t plotted against log 10 of five different 10-folds DNA dilutions (10, 1, 0.1, 0.01 and 0.001 ng/μL), without amplification observed for NTC. The amplification of the five DNA 10-fold serial dilutions showed a very good linear regression and correlation coefficient (R 2 ) of 0.9909. The assay indicates an acceptable value of 97% efficiency which corresponds to -3.381 slope value. Hence, the result indicates that the internal control used in this study has high efficiency and reproducibility.

Determination of the specificity of the primers and probe
Specificity of the designed cytb primers and probe against porcine DNA was validated with eight meat species namely chicken, beef, duck, mutton, buffalo, deer, rabbit and quail through qPCR assay. The qPCR assays for the validation of specificity were conducted in triplicates for two independent studies. The designed cytb primers and probe is porcine-specific as indicated by the absence of amplification against DNA extracted from the other eight species. The specificity of the mitochondrial 18S rRNA primers and probe designed by Rojas et al. (2011) was determined to validate the presence of eukaryotic DNA that cannot be detected and amplified by porcine specific PCR system (Kim et al., 2016). The 18S rRNA is a structural ribosome for the small element of eukaryotic ribosomes, responsible for synthesizing protein and serves as the basic components of all eukaryotic (Uddin and Cheng, 2015). These universal sequences designed within 18S rRNA positively detect eukaryotic cells and preclude false negative outcomes. It also helps to confirm the functionality of the reaction mix, factors that might affect the real-time PCR amplification process and the possibility of false-positive detection in the qPCR (Hossain et al., 2017). Table 2 shows the mean C t value of the amplified DNA for specificity assay using the designed cytb primers and probe and 18S rRNA. Figures  4 and 5 show the representative qPCR amplification curve for the cytb and 18S rRNA specificity assay, respectively. DNA amplification was observed for all species which indicates the 18S rRNA primers and probe as highly specific to eukaryotic DNA.

Determination of sensitivity for the newly designed cytb primers and probe
The sensitivity of primers and probes in qPCR assay is vital as it provides an insight into the suitability and capacity of a detection system. In this study, the limit of detection (LOD) was determined to establish the lowest concentration of porcine DNA that can be detected by the designed primers and probe. LOD is defined as the lowest concentration that provides a positive result in all sample replicates corresponding to at least a 95% detection rate (Cai et al., 2017). A spiked meatball model was prepared to represent the complex matrices of the commercially processed meat products. The sensitivity of the assay was determined using 10-fold serially diluted raw and heat-treated beef and chicken spiked with pork meatball. The experiment was repeated for two independent studies, each in triplicates. Raw meatballs and heat-treated pork meatballs served as a control for comparison purposes. The LOD of the qPCR assay for  Table 3. In this study, the LOD for raw spiked beef and chicken meatballs were 0.01 ng/μL and 0.1 ng/μL, respectively. In comparison, the LOD of raw pork meatball is 0.001 ng/μL as it consists of solely pork meat and serves as a control.
LOD of the heat-treated meatball model was determined using the meatball model subjected to boiling at 100°C for 2 hrs and 30 mins. The LOD of heat-treated spiked chicken and the beef meatball was 0.1 ng/μL. Furthermore, the LOD of the heat-treated spiked meatball model is lower than the LOD of the raw meatball model. According to Bhat et al. (2016) although DNA has a stable structure compared to protein, DNA present in the food matrices may be damaged and disintegrated into smaller pieces when subjected to heat treatment. Processing and intensive cooking may lower the concentration of the DNA as, during meat processing or heating, the cellular membrane of the meat is disrupted. A previous study by Ali et al. (2012) reported a detection limit of 0.01% (w/w) of pork in a beef meatball spiked model following boiling in water for 15 mins. In comparison with the present study, prolonged boiling of adulterated meatballs in water does not significantly affect the detection limit. The C t values were the mean of replicate assays (n = 6). c SD: Standard deviation d 40±0.00 : Porcine DNA not detected after cycle 40 Table 2. Mean C t value of two independent studies for cytb and 18S rRNA specificity assay

Statistical analysis
The statistical analysis between raw and heat-treated spiked meatball model was determined using Student's ttest. As shown in Table 4, no significant difference (p>0.05) was observed in the porcine DNA detection between raw and heat-treated spiked chicken and beef meatballs. This result is in agreement with López-Andreo et al. (2012) who concluded that heat treatment does not significantly affect the result of qPCR against the different concentrations of DNA. However, there was a significant difference in mean C t values between raw and heat-treated pork meatballs at 10 to 0.01 ng/μL as the p values are less than 0.05. According to Hird et al. (2006), although heat-treatment may reduce the amount of detectable DNA, the DNA extracted from a heattreated sample can still be detected as the primers used in qPCR targeting only a short DNA fragment.

Conclusion
Consumption of food products that are adulterated with pork or its derivatives is forbidden for Muslims and Jews. In addition, adulterated foods can also cause adverse health effects to the individual who is prone to an allergic reaction against certain meat spp. Therefore, knowledge of food science and technology is critical to counter food adulteration issues. In this study, the presence of porcine DNA was detected through qPCR assay using the designed cytb primers and probe sequences. Having an amplification efficiency of 95%, the designed primers and probe are specific to porcine DNA as no other species were detected by the assay. Moreover, it can detect as low as 0.01 ng/μL and 0.1 ng/ μL of pork DNA in raw spiked beef and chicken meatball model, respectively. Pork DNA was detected at a concentration of 0.1 ng/μL in the heat-treated spiked chicken and beef meatball model. This study shows that the extracted DNA following prolonged meatballs heat treatment does not affect qPCR, which was comparable with other meatball models that were only subjected to boiling for 15 mins. Other thermal treatment processing parameters such as using a microwave oven could be incorporated in the future study to observe the effect of different conditions on detection limits. Furthermore, the designed primers and probe can be further applied for detection of porcine DNA in commercial beef and poultry canned food products with bogus Halal logo and without halal logo. Finally, as the designed primers and probe are sensitive and specific in detecting the presence of porcine DNA in raw and heat-treated products, it has the potential to be used in the screening for food adulteration and mislabelling which will protect consumers from food fraud.