The majority of proteomic-based meat authentication studies have used either 2-dimensional gel electrophoresis (2-DE)/OFFGEL fractionation coupled with mass spectrometry (MS) (Montowska and Pospiech 2012; Sentandreu et al. 2010) or the multiple reaction monitoring (MRM) MS method (Watson et al. 2015). However, considering the unknown food fraud threats emerging globally, rapid extraction and fractionation protocol coupled with MS was developed in the current study for routine laboratory application. Given the enormous complexity of meat proteins, their separation prior to MS is needed to produce fractionated proteins in solution for comprehensive analysis of proteome. The following section describes the simple extraction and novel GELFrEE fractionation technique and MS for authentication of pork mixed with water buffalo meat. The current study also evaluated the robustness of 2-DE, a gold standard protocol used in proteomic studies for authentication of meat species.
Simple Extraction Protocol, GELFrEE Fractionation, and SDS-PAGE Analysis
Choice of muscle protein extraction method is critical for obtaining samples with high protein concentration and free of salt and lipids, which could interfere with further analysis. Researchers have extensively reported extracting muscle proteins with denaturing agents containing urea, thiourea, reducing agents, detergents, and salts (Anderson et al. 2011; Bowker et al. 2008) using in-house laboratory facilities like tissue homogenizers, high-speed refrigerated centrifuge, and sonicator followed by purification of extracted proteins for further characterization. Considering the demand for simple, cost-effective, and quick extraction methods for routine laboratory use, protein extraction from meat samples utilizing simple trituration with the standardized buffer containing a diverse proportion of phosphate-buffered saline (PBS), SDS, and EDTA (Tab. S1, Fig. S1) using pestle and mortar followed by filtration with Grade 1 filter paper was optimized in this study. The 0.01 M phosphate buffer (pH 7.4), 0.9% NaCl, and 1% SDS was found optimal for extraction of proteins from both raw and cooked meats for further authentication.
SDS-PAGE coupled to MS was reported to detect a minimum of 10% w/w adulteration of pork with chicken and in order to detect the lower adulteration level (up to 0.5%) an additional proteins/peptides enrichment step was recommended (Sentandreu et al. 2010). To achieve this we used novel Gel-eluted liquid fraction entrapment electrophoresis, a commercial Expedeon’s GELFrEE 8100 unit and their proprietary cartridges. The GELFrEE fractionates the broad mass range of proteins on the basis of molecular weight with high resolution and with >90% recovery of proteins (Tran and Doucette 2008). To assess the feasibility of GELFrEE as an efficient enrichment step, we fractionated 100% buffalo meat, 100% pork and buffalo meat: pork meat mixes at different levels. Fig. 1A provides an overview of the selected fractions obtained following GELFrEE separation of raw (Fig. 1A, a) and heat-processed (Fig. 1A, b) buffalo meat proteins. Similarly, GELFrEE fractionation of raw and heat-processed pork proteins were represented in Fig. 2A. Proteins were fractionated using 5% GELFrEE cartridge with a fractionation range of 3.5 to 500 kDa. This cartridge with an in-built 2.5 cm length gel, separated the meat proteins within 196 min. The rapid separations described were achieved by employing high electric fields (3-11 mA) on short gel columns. The protein fractions from GELFrEE were subjected to separation on 12% SDS-PAGE to visualize the protein bands and to validate the fractionation. The average MW of proteins obtained in the present study displayed a definitive increase as the separation progressed. In the GELFrEE technique, small molecular proteins are eluted first from the gel and subsequently confined in the collection chamber. After the initial run, a similar migration rate for larger proteins was observed. As evident from Figs. 1A and 2A, protein fractions were collected in an approximate “linear” molecular weight (MW) profile and while collecting, proteins of the entire mass range remained focused and recovered in a single fraction with a consistently high yield. The unique entrapment strategy of GELFrEE offers sample enrichment as opposed to dilution, and through automatic pauses at predefined entrapment intervals (longer trapping times at high MW proteins), the probability of trapping given proteins in single fractions is maximized. Our findings also revealed that GELFrEE fractionation did not require any sample purification or removal of salt and this ensured the retaining of even minor or lower abundance proteins. Tran and Doucette (2008) have demonstrated 100% recovery of standard proteins (BSA, cytochrome c, and ubiquitin) from a short GELFrEE column and observed a similar elution pattern and separation with a high degree of reproducibility. In comparison to protein separation using SDS-PAGE, GELFrEE separation trap the proteins and later elute them into a few relatively simple fractions that can be analyzed by MS with greater sensitivity over the whole proteome. These proteins were partitioned according to their size, resulting in 12 clear fractions encompassing total buffalo meat and pork proteins that span the entire mass range in a 2-3 h collection window. For the 5% cartridge used in the current study, an apparent increase in the size of proteins is visualized in SDS-PAGE as the GELFrEE run progresses. A similar logarithmic trend was observed by Orton et al. (2013) in their study on characterizing secreted proteome of a wild-type and mutant strain enteropathogenic Escherichia coli using multiplexed GELFrEE platform.
In the current study, GELFrEE fractions were further subjected to SDS-PAGE in order to validate the linear elution of proteins based on their MW and for the selection of fraction numbers in which the potential marker proteins may be eluted. Separation of GELFrEE fractions for raw and heat-processed (Fig. 1A) meat extracts on SDS-PAGE exhibited elution of low MW proteins in fractions 2-3 and high MW proteins from factions 4-9. As highlighted in Fig. 1A, similar protein band separation pattern/intensity was observed for low MW proteins even after heat-processing. On the contrary, the band intensity corresponding to high MW proteins was reduced to a greater extent and some bands were completely missing after cooking. These results indicated that heat processing or cooking of meat over 100 °C deteriorated the intensity (decrease in intensity and absence of protein bands) of high MW proteins, whereas low MW proteins were more resistant to heat-induced denaturation and suffered minimal damage (Montowska and Spychaj 2018). These low MW proteins might be used as markers for further species identification, as demonstrated by Sentandreu et al. (2010).
Mass Spectrometry Analysis and Identification of Species-specific Peptide Biomarkers
As a logical first step to ensure the authentication of meat species, we identified the species-specific proteins/peptides using 100% water buffalo (Fig. 1) meat and 100% pork (Fig. 2) separately. In-gel digestion and MALDI-TOF MS of low molecular weight protein bands from selected GELFrEE fractions revealed the presence of myoglobin (Mb) and carbonic anhydrase-3 (CA-3) derived peptides (Table 1) that are specific to water buffalo and pork. Myoglobin and CA-3 have been reported as the main proteins of heat-stable peptide markers with high degree of species-specificity (Stachniuk et al. 2019). In the current study, we identified two water buffalo-specific peptides derived from myoglobin ‘TEAEMKASEDLK’ (M+H+=1351.78), ‘HPSDFGADAQAAMSK’ (M+H+=1532.67), and corresponding peptides from pork myoglobin ‘SEDEMKASEDLK’ (M+H+=1381.61), ‘HPGDFGADAQGAMSK’ (M+H+=1488.65). Similarly, two specific peptides derived from water buffalo CA-3 ‘GGPLAAPYR’ (M+H+= 901.48), ‘GEFQLLLDALDK’ (M+H+= 1361.73) and corresponding pork CA-3 peptides ‘GGPLTAAYR’ (M+H+= 905.48), ‘GEFQLVLDALDK’ (M+H+= 1347.71) were identified. These peptides were detected in both raw and heat-processed meat samples indicating their heat stability. Myoglobin requires less aggressive extraction procedures and Mb derived peptides are heat stable suggesting their suitability as markers for authentication of both raw and heat-processed meat (Taylor et al. 1993). The good heat stability of the ‘HPSDFGADAQAAMSK’ peptide detected in our study was also confirmed by Claydon et al. (2015). The CA-3 is the most abundant cytosolic protein and was reported to discriminate meat species from fresh meat and their mixtures Kim et al. (2017).
The Mb and CA-3 peptides exhibited differences in sequence between buffalo meat and pork, which can be detected by mass spectrometry analysis. The differences in peptide sequence between Mb and CA-3 of buffalo meat and pork were compared using Clustal O (1.2.4) multiple sequence alignment (Figs. S2 and S3). The percent identity score for pairs of CA-3 and Mb sequences from buffalo meat and pork was calculated using Clustal 1.2. It was revealed that the amino acid sequences of CA-3 and Mb between buffalo meat and pork differed by 7% and 12%, respectively. The species-specific peptides derived from CA-3 and Mb were further subjected to sequence scoring and validated through NCBI BLAST. The representative MS/MS spectra of selected marker peptides from buffalo meat carbonic-anhydrase-3 (‘GGPLAAPYR’) (Fig. 1C) and pork myoglobin (‘HPGDFGADAQGAMSK’) (Fig. 2C) were presented. The spectra under each chromatographic peak were analyzed, and the ratio was determined between molecular weight to molecular ions. Sequence score and NCBI BLAST were used to identify the species-specific peptides as signature peptides. All of the peptide markers were found to be present in one of the two species under investigation, but not in the other, according to the peptide detection analysis. In each of the chosen peptides, there was a relatively low error mass (1.0073 Da) between theoretical and observed m/z, indicating a good match percentage. Previously, the myoglobin peptide HPGDFGADAQGAMTK was used to identify horse meat in fresh meat (Watson et al. 2015).
Establishing Limit of Detection from Binary Meat Mixture
Identification of buffalo meat and pork-specific peptides has prompted us to detect contaminating pork in water buffalo meat at a proportion as low as 0.5%. Different proportions of ground pork (0.5%, 1.0%, 5.0%, and 10%) were mixed with ground buffalo meat, cooked at 100 °C for 30 min followed by extraction of proteins from raw and heat-processed samples, as described earlier using optimized buffer, and finally, protein fractionation was performed using GELFrEE electrophoresis. All the 12 fractions from GELFrEE electrophoresis were collected separately and characterized on 12% polyacrylamide gel. Selected protein bands from the 1st GELFrEE fraction (Fig. 3) were subjected to trypsin digestion followed by protein identification using MALDI-TOF MS analysis. Buffalo meat and pork-specific peptides, that were identified as potential biomarkers in 100% raw and cooked buffalo meat and pork were also detected in 0.5% pork adulterated with buffalo meat under raw as well as in cooked conditions. Our repeated experiments (n = 6) with raw and cooked buffalo meat and pork mix revealed the elution of species-specific peptides in GELFrEE fraction 1 that can be used for meat authentication. We could successfully detect (Fig. 3) both water buffalo-specific peptides (‘TEAEMKASEDLK’, m/z 1352.31; ‘HPSDFGADAQAAMSK’, m/z 1532.63) as well as pork-specific peptides (‘SEDEMKASEDLK’, m/z 1381.24; ‘HPGDFGADAQGAMSK’, m/z 1488.66) derived from Mb in meat mixes (as low as 0.5% contaminating pork in buffalo meat mix). Similarly, water buffalo (‘GGPLAAPYR’ m/z 901.48; ‘EPITVSSDQIAK’, m/z 1287.20) and pork-specific (‘GGPLTAAYR’, m/z 905.48; ‘EPITVSSDQMAK’ m/z 1306.24) peptides derived from CA-3 in meat mixes were also detected. Detection of beef-specific Mb peptide ‘HPSDFGADAQAAMSK’ in the sample containing 5% (w/w) beef in the pork and horse meat matrix using label-free quantification combined with the infusion of high-resolution MS was also reported by Montowska and Spychaj (2018). A highly sensitive and specific method using liquid chromatography-tandem mass spectrometry was established by Li et al. (2018) for the simultaneous determination of heat-stable species-specific peptides from beef Mb (‘HPSDFGADAQAAMSK’), beef CA-3 (‘GEFQLLLDALDK’), pork myoglobin (‘HPGDFGADAQGAMSK’, ‘YLEFISEAIIQVLQSK’) and pork CA-3 (‘HDPSLLPWTASYDPGSAK, EPITVSSDQMAK’).
Application of GELFrEE for Authentication of Commercial Meat Products
We examined the applicability of the peptidomic approach adopted in this study (simple extraction, GELFrEE fractionation, and MALDI-TOF MS) using various commercially processed meat products including ham, sausage, frankfurter, bacon, etc. purchased from supermarkets, as well as in-house processed meat products as reference samples (Fig. S4). All the products were screened for the presence of species-specific Mb and CA-3 derived peptides, which were detected in 100% buffalo meat or pork and their mixes. The list of peptide markers identified in this study for the 10 different processed meat products is presented in Table 2. Most of the observed peptides identified as heat-stable markers belonging to Mb and CA-3 were unique to both species, confirming that the selected peptides are not only highly specific but that they have good thermal stability. In the commercial and in-house processed samples, we identified the most abundant Mb and CA-3 as previously mentioned in Table 1. The canned corned buffalo meat (sample no. 4), which was produced under high temperature and high pressure, retained at least two CA-3 peptides, thereby fully meeting qualitative requirements. Although the pork cocktail sausage (sample no. 3) was marked with chicken ingredients, the specific chicken meat content was not labelled and none of the specific peptides of chicken were detected. All the tested commercial products were in compliance with their label descriptions. The results confirmed that Mb and CA-3 were both heat-stable proteins and could be used as screening targets of species-specific markers. Li et al. (2021) also utilized pork-specific peptides, ‘EPITVSSDQMAK’, ‘GGPLTAAYR’, ‘HDPSLLPWTASYDPGSAK’ from CA-3 to accurately quantify pork content in different types of meat products. Pork-specific CA-3 derived processing resistant marker ‘HDPSLLPWTASYDPGSAK’ was reported by Kotecka-Majchrzak et al. (2021) in different meat and vegan products. Our results confirm the simple protein extraction coupled with GELFrEE fractionation as a robust and efficient enrichment step for low MW proteins, which can be successfully explored in the development of a peptidomic-based approach for meat authentication.
Two-Dimensional Gel Electrophoresis and MS Approach for Meat Authentication
Unlike simple extraction followed in GELFrEE, for 2-DE total proteins were extracted using the procedures of Montowska and Pospiech (2011) and subjected to TCA/acetone precipitation followed by fractionation of 100% buffalo meat, 100% pork and buffalo: pork meat mix (95.5:0.5). The 2-DE gels from raw and cooked meat mix containing buffalo: pork (95.5:0.5) are represented in Fig. 4. Based on the pioneering work reported by Montowska and Pospiech (2012, 2013) and our earlier research findings (Naveena et al. 2017), cluster of two to three protein spots as indicated in Fig. 4 was reported to provide species-specific information. From our laboratory, we have consistently obtained these spots and we anticipate that these spots will correspond to species-specific marker peptides derived from myosin light chain isoforms.
The MALDI-TOF MS analysis and species-specific peptides identified are listed in Tab. S3 with their accession number, mass, protein score, sequence coverage, and the number of matched peptides. The results showed the presence of both buffalo and pork-specific peptides in meat mix containing buffalo meat and pork. Buffalo-specific myosin light chain-3 (MLC-3) (spot No. R1) and pork-specific myosin light chain-4 (MLC-4) (spot No. R2) were detected in the raw meat mix. In cooked meat mix, buffalo-specific MLC-3 (spot No. C1) and pork-specific myosin light polypeptide 6 (spot no. C2) were detected. On 2-DE gels, MLC isoforms form a characteristic pattern, which is typical of each of the species, and on the basis of which it is possible to identify those species in meat mixtures. The MLC isoforms retain their species-specific electrophoretic mobility even after processing, including minced meat and various meat products (Montowska and Pospiech 2012).
The water buffalo-specific MLC-3 derived tryptic peptide ‘AAAAPAPAPAPPPAPEPSK’ detected in both raw and cooked meat mix and ‘EPEFDPSKIK’ in raw meat mix in the current study were also reported by Naveena et al. (2017, 2018) in protein extracts of meat mix identified by peptide mass fingerprinting using MALDI-TOF MS. The inter-species differences in MLC isoforms in raw meat of six species, namely cattle, pig, chicken, turkey, duck, and goose, have been observed by Montowska and Pospiech (2011). The study demonstrated the robustness of 2-DE for authentication of pork in buffalo meat mix up to 0.5% level in both raw and cooked conditions; however, 2-DE fractionation is laborious and consumes almost 48-72 h.