Over the past 15 years, extensive research has been conducted around the world on the study of protein substances in raw meat and meat products, both native and those formed in the process of various technological treatments.

A classic quantification method in proteomics is the use of an isotopic tag, the modification of which has more than 40 species (Kopylov and Zgoda, 2007). There are also techniques that do not use isotopic labels (Kopylov, Zgoda and Archakov, 2009). The sensitivity of protein determination compared with gel electrophoresis increases by several orders of magnitude. More recently, the complexity of the study of phosphorylated proteins has been overcome. Various post-translational modifications of proteins with high sensitivity and specificity are studied by the Selected Reaction Monitoring (SRM) method (Zav'yalova, et al., 2014). Recently, a method of identifying species-specific molecular markers in the field of food analysis has gained strength, based on a combination of two methods, high-performance liquid chromatography (HPLC) and mass spectrometry (MS), used to detect peptides. Using this method, up to 0.5% (w/w) chicken meat was found in meat mixtures (Sentandreu et al., 2010). In more recent studies, in boiled meat products, up to 1.0% (w/w) impurities of beef, pork, chicken, duck and goose were detected (Montowska and Fornal, 2017). Heat treatment products were analysed using marker peptides derived from myosin 1 and 2 light chains. It is very important to determine the limit of detection (LOD) of the method. Using this criterion, one can compare various methods aimed at determining muscle tissue. Indicators of 0.5% and below were set for meat products. As an example, the established quantification limit for buffalo and sheep meat was up to 0.48% (w/w) meat (Naveena et al., 2017). The good thermal stability of the peptides was demonstrated by the authors to identify horse and pork markers a lower limit of 0.24% (Von Bargen, Brockmeyer and Humpf, 2014).

Model mixtures of minced muscle tissue were prepared in accordance with standard industrial procedures. A set of samples with a given recipe was prepared (Table 1). Beef muscle tissue content was 8% (w/w) and 16% (w/w). The calculation of muscle tissue content was carried out according to BEFFE (bindegewebseiweißfreies Fleischeiweiß − meat proteins that do not contain connective tissue) (Leitsätze für Fleisch und Fleischerzeugnisse, 2016). Samples of minced meat mixtures were placed in a collagen shell and cooked to a core temperature of 72 °C.

In this work, we used the Skyline program (2019), capable of theoretically cleaving proteins and listing the SRM for each peptide (Table 2). Protein analysis was performed by using biomodelling. If it is necessary to process complete protein sequences during analysis of LC-MS/MS data, it is possible to use software such as Spectrum Mill (Agilent Technologies, Santa Clara, CA, USA) (Sarah et al., 2016; Fornal and Montowska, 2019; Montowska and Fornal, 2017; Montowska and Fornal, 2019), Protein Lynx Global Server (Waters) (Naveena et al., 2017), Peaks Studio software (Bioinformatics Solutions, Waterloo, ON, Canada) (Prandi et al., 2017; Prandi et al., 2019) and MASCOT (Matrix Science, Boston, MA, USA) (Sentandreu et al., 2010; Naveena et al., 2017; Ruiz Orduna et al., 2015; Ruiz Orduna et al., 2017; Fornal and Montowska, 2019; Montowska and Fornal, 2017; Montowska and Fornal, 2019). In our work with the search for parameters for biomarkers on a mass spectrometer, the Skyline program proved to be the best. This is the best choice in the presence of a previously studied peptide sequence for develop of MRM methods. Most often, three transitions were selected. Only y-ions were used. The transition from parent ion (m/z) to product ions (m/z) occurred from a smaller to a larger one (m/z).

Peptides presented in a recent review (Stachniuk et al., 2019) were selected for comparison of potential biomarkers. Previously submitted peptides by us were analysed (Khvostov et al., 2019; Kulikovskii et al., 2019). One of the criteria for marker specificity is the presence of a sequence of more than six amino acids (Watson et al., 2015). This peptide length provides the species specificity of muscle protein. We decided to use the S/N indicator as the criterion for the comparison of heat-stable peptides.

Chromatograms of SRM peptide markers are shown in Figure 1a and Figure 1b. The four most intense peptides with a signal value of (50−250)*10³ cps are presented in Figure 1a. The remaining peptides in the intensity range of (10 - 50)*10³ cps are indicated in Figure 1b. The chromatogram data were obtained in a sample with a beef concentration of 16% (w/w), subjected to thermal treatment.

The S/N results for a sample of minced meat with 16% beef (w/w) after heat treatment are shown in Table 3.

The peptides are arranged in descending order of S/N. The data show that S/N is the highest for the peptide sequences NDMAAQYK (Kulikovskii et al., 2019) and YLEFISDAIIHVLHAK (Khvostov et al., 2019), which are myoglobin derivatives. Since beef contains a high level of myoglobin, we obtained the largest number of myoglobin peptide derivatives. The S/N ratio is above 10 for both raw and heat-treated samples. For the peptide HPSDFGADAQAAMSK (Claydon et al., 2015; Li et al., 2018; Khvostov et al., 2019), an additional MRM search was performed. Two parent ions, 766.8 (m\z) and 511.6 (m\z), were used. The most significant was ion 511.6 (m\z). The MRM intensity for this mass increased by 40% ±7.4 compared with ion 766.8 (m\z).

Samples were frozen and re-thawed. We evaluated the effect of one freeze/thaw cycle in digested samples on the intensity of the HPSDFGADAQAAMSK peptide in all mixtures. For samples subjected to and without heat treatment, S/N did not change. It was found that one freeze/thaw cycle did not affect the concentration of meat in mixture 1. If the beef content was less than 10% (w/w), the intensity decreased to 52.4 ±15.2. For peptides ALDLDSNC [+57.0] K (Wang et al., 2018), DLADEVALVDVMEDK, and GEFQLLLDALDK (Li et al., 2018), cross-contamination was recorded in a blank sample (no beef) (mixture 3) (Figure 2).

Many peptides did not meet the criterion of S/N >3.

Peptides representing from myosin proteins, such as MEIDDLASNVETISK (Montowska and Fornal, 2019) TLALLFSGPASGEAEGGPK (Claydon et al., 2015; Fornal and Montowska, 2019; Montowska and Fornal, 2019) were sensitive to heat-treated products with 16% muscle tissue (w/w). At lower concentrations, S/N approached 2 – 3. It was not possible to identify the DLADEVALVDVMEDK peptide (Li et al., 2018) in all types of samples. The S/N index for all samples was no greater than 1. A one-way analysis of variance (ANOVA) found an insignificant effect of temperature on the intensity of marker peptides at a concentration of 8% (w/w). In previous studies by Kulikovskii et al. (2019) and Khvostov et al. (2019), we established a limit of detection (LOD) of 0.29% for the NDMAAQYK peptides and 0.93% for the SNVSDAVAQSAR peptide. From the analysis of species-specific marker peptides, three peptides for determining muscle tissue in beef were selected, taking into account the following factors: high prevalence in muscle tissues (>50 сps), good S/N ratio at low concentrations (S\N >10), high specificity and the presence of trypsin-specific cleavage sites at both ends of the protein chain.

Two-way analysis of variance does not reveal differences in the assessment of the criterion for the influence of heat treatment of mixtures at a concentration of 8% beef, confirmed by statistical calculation of p (<0.71), which is higher than the significance level of alpha (0.05).

The developed methodology allowed us to simultaneously identify and compare up to 13 beef peptide biomarker. Using the S/N criterion, it was possible to compare peptide markers for the authenticity of raw meat and heat-treated meat. Considered successful candidates whose signal-to-noise ratio was higher than 3.

From the analysis of species-specific marker peptides, three peptides for determining muscle tissue in beef were finally determined: NDMAAQYK and YLEFISDAIIHVLHAK from myoglobin and SNVSDAVAQSAR from triosephosphate isomerase protein. For samples with two concentration levels and under cooking conditions at 100 °C, the S/N ratio was set above 10. This approach is universal. It is suitable for comparing meat biomarkers of other animal species. It will be able to identify the most suitable candidates. Selected peptide markers can be used to construct regression curves with good linearity, allowing a quantitative assessment of the types of meat present. The selected peptides can be used effectively to distinguish between accidental contamination (technologically unavoidable impurity) and deliberate falsification.

The developed methodology can aid in the study of the effect of meat protein on meat quality and functional characteristics, as well as the safety of finished meat products.