Introduction
Tenderness is a critical attribute in determining customer satisfaction of beef products. The complexity of the regulation of tenderness has long been recognized, and a great deal of resources have been dedicated to elucidating the biological basis for variation in tenderness with only moderate success. In these efforts, component traits such as sarcomere length and the extent of postmortem proteolysis have been measured to understand their contribution to tenderness variation (Huff Lonergan et al., 2010; Koohmaraie, 1996; Taylor et al., 1995; Wheeler et al., 2000). These component traits have provided insight into beef tenderness variation in many cases. However, in some instances, these traits failed to explain the tenderness differences observed across treatments (Cooper et al., 2025; Shackelford et al., 2012).
The advancement of molecular tools has enabled the profiling of the proteome and metabolome of beef differing in tenderness (Anderson et al., 2012b; King et al., 2019; Picard et al., 2018). These investigations have implicated a number of biomarkers in relation to tenderness variation, which highlight the importance of muscle metabolism in determining beef tenderness (Anderson et al., 2014; Antonelo et al., 2020; D’Alessandro et al., 2012). However, like the component traits, the effects of these biomarkers on tenderness are not always consistent across datasets. For example, King et al. (2019) reported increased malate to be associated with decreased tenderness of beef longissimus lumborum steaks. However, Cooper et al. (2025) reported a negative correlation between malate and slice shear force in beef longissimus lumborum steaks aged for 12 d and no correlation to tenderness in longissimus lumborum steaks aged for 26 d. Our laboratory has observed similar inconsistencies regarding other metabolites as well.
Questions facing meat scientists trying to understand biological variation in meat quality traits concern the nature of the relationships among biomarkers and the meat quality attribute of interest. We speculate that the inconsistencies in relationships across datasets are indicative of complex interactions among the various factors regulating the attribute. The present study was conducted to do the following: 1) validate the effectiveness of biomarkers previously associated with tenderness variation, 2) determine the marginal effects of metabolic traits on variation in beef tenderness when other factors are taken into account, and 3) explore interactions and interrelationships among the various metabolic traits in relation to beef tenderness.
Materials and Methods
This study used sampled carcasses at a US Department of Agriculture (USDA) Food Safety and Inspection Service inspected facility. As the work did not involve live animals, an animal care and use committee review was not required.
Sample handling and preparation
Carcass selection and sample processing were conducted as described by King et al. (2025). At approximately 24 h postmortem, commodity USDA Choice and Select grade beef carcasses (n = 100) were screened and selected for this study based on longissimus muscle pH less than 5.6, 5.60 to 5.74, 5.75 to 5.9, and greater than 5.9 (n = 25 from each group). Measurements were made with FC 200 B series electrode pH probes (Hanna Instruments, Woonsocket, RI) attached to REED SD-230 pH meters (Reed Instruments, Wilmington, NC). Strip loins, similar to institutional meat purchase specifications (IMPS) #180 (USDA, 2014), and top sirloin butt, similar to IMPS #184 (USDA, 2014), were transported to the US Meat Animal Research Center. A 50-g muscle sample was removed from the anterior aspect of each longissimus lumborum and gluteus medius muscle for 48-h laboratory muscle pH determination before cuts were vacuum packaged.
After being stored (4°C) until 14 d postmortem, subprimals were unpackaged and trimmed free of fat and accessory muscles before being cut into 28-mm thick steaks. Longissimus lumborum steaks 1 and 2 (most anterior) were assigned to trained sensory panel. Steak 3 was used for slice shear force determination. Steak 4 was used to measure oxygen consumption, nitric oxide metmyoglobin reducing ability, and the remainder was pulverized for use in other laboratory analyses. Steak 5 was used for a concurrent experiment. Gluteus medius steak 2 was used for laboratory experiments. Steak 3 was used for a trained sensory panel. Steak 4 was used for slice shear force determination. Steaks 1 and 5 were used for a concurrent experiment.
pH, myoglobin concentration, and 2-thiobarbituric acid reactive substances determination
The Bendall (1973) iodoacetate method was used to determine muscle pH on 48 h postmortem samples using a Reed SD-230 handheld pH meter with a pH probe (Omega PHE 2385 pH probe, Omega Engineering Inc., Stamford, CT). Myoglobin concentration was determined on samples aged for 14 d using the method reported by Hunt and Hedrick (1977) and modified by McKeith et al. (2016). Quantification of 2-thiobarbituric acid reactive substances (TBARS) was conducted on extracts of 14-d aged samples using a Quantichrom TBARS assay kit (DTBA-100, Fisher Scientific) as described by King et al. (2025).
Oxygen consumption and nitric oxide metmyoglobin reducing ability
As described by McKeith et al. (2016), oxygen consumption was measured as the proportion of surface myoglobin in the oxygenated formed after incubation in atmospheric conditions (bloom). Reducing capacity was estimated by measuring the initial levels of metmyoglobin formed after incubation with 0.03% sodium nitrite for 30 min at 20°C as described by McKeith et al. (2016). Surface myoglobin forms were calculated using the equations described by King et al. (2023).
Glycolytic potential and glycolytic metabolites
Metabolite concentrations were determined using coupled enzyme assay systems described by Bergmeyer (1974) and adapted for use with 96 well plates and using standard curves as described by Hammelman et al. (2003), with additional modifications described by King et al. (2025).
Mitochondrial DNA copy number
Relative mitochondrial abundance was determined as described by King et al. (2025). Mitochondrial and nuclear DNA was amplified by real-time polymerase chain reaction using primers for nicotinamide adenine dinucleotide + hydrogen (NADH) dehydrogenase 1 (ND1; forward primer ND1-3278F, ccactacgacccgctacatc, and reverse primer ND1-3438R, acggctaggcttgatatggc), and nuclear DNA was quantified using primers for vitamin D binding protein, group specific component (GC; forward primer GC-24746F, acctccttgatgcagtcctc, and reverse primer GC-24686R, caaatttgcccagaaagtgc) in separate reactions. The copy number of mitochondrial DNA was calculated as described by Schmittgen and Livak (2008).
Heat shock protein 70 and peroxiredoxin-2 abundance
Separate whole muscle extracts were prepared for peroxiredoxin-2 and heat shock protein 70 quantification as described by King et al. (2025). Supernatants were assayed for peroxiredoxin-2 and heat shock protein 70 using Bovine Peroxiredoxin-2 ELISA kit (catalog No. MBS2883044 Sandwich, MyBioSource.com, San Diego, CA) and bovine heat shock protein 70 ELISA kit (catalog No. MBS736940, My BioSource.com, San Diego, CA), respectively, using reagents and instructions provided by the manufacturer.
Carbonyls in soluble, insoluble, and mitochondrial fractions
Extraction and quantification of soluble and insoluble fractions were completed using methods provided in Rowe et al. (2004). The mitochondrial fraction was isolated, and mitochondrial protein concentration was determined according to Cawthon et al. (2001) and Iqbal et al. (2005) as modified by McKeith et al. (2016). Carbonyls on proteins from each fraction were determined by incubation with 2,4-dinitrophenylhydrazine to quantify the extent of oxidative damage in each muscle fraction, following the protocol of Reznick and Packer (1994) as modified by Rowe et al. (2004) and McKeith et al. (2016).
Myosin heavy chain isoforms, desmin degradation, and sarcomere length
A whole muscle extract prepared was used to quantify the proportion of myosin heavy chain isotypes using electrophoretic separation described by Picard et al. (2011) with modifications described by King et al. (2025). The percentage of desmin degraded was quantified on the tissue from the slices sheared in slice shear force determination that had been pulverized in liquid nitrogen by the Simple Western Automated Western Blot System (Bio-Techne/Protein Simple, San Jose, CA) as described by (Cooper et al., 2025). Sarcomere length was measured on pulverized cooked tissue from the slices sheared for slice shear force using laser diffraction method (Cross et al., 1981) as described by Cooper et al. (2025).
Slice shear force
Steaks designated for slice shear force determination were equilibrated to 5°C overnight before cooking for slice shear force determination. Steaks were cooked to a final internal temperature of 70°C on a TBG-60 Magigrill belt grill (MagiKitch’n Inc., Quakertown, PA) as described by (Wheeler et al., 1998). Longissimus lumborum steaks were sampled for slice shear force by removing a 5-cm section from the lateral end of the steak as described by Shackelford et al. (1999). Three 1-cm thick slices were removed, perpendicular to the steak surface, from the section. Gluteus medius steaks were sampled for slice shear force by removing three 5-cm sections from the lateral, center, and medial sections of the steak as described by King et al. (2009b). From each section, two 1-cm thick slices were taken perpendicular to the steak surface. All slices were sheared on a TMS-PRO Texture Measurement System (Food Technology Corp, Sterling, VA), and peak force was recorded and sheared as described by Shackelford et al. (1999).
Trained sensory panel analysis
The descriptive attribute panel was trained as prescribed by Cross et al. (1979) and the American Meat Science Association (2016). Panelists (n = 8) rated overall tenderness on an 8-point scale (1 = extremely tough, 8 = extremely tender). On each panel day, panelists were presented with samples from each of the muscle pH classes for which carcasses were selected. Presentation order of the pH classes was blocked across panel days. Steaks designated for trained sensory panels were thawed and allowed to equilibrate to 5°C for 24 h. Steaks were cooked to a peak internal temperature of 70°C as described above just in time to be sampled for sensory panel analysis. Exterior fat and connective tissue were removed, and the remaining steak was cut into 1.37-cm × 1.37-cm × steak thickness cubes. Panelists were served 3 random cubes from each sample. A warm-up sample was evaluated and discussed before formal sample evaluation began.
Statistical analysis
Simple statistics were generated for each of the biochemical and tenderness traits measured (Supplementary Tables 1 and 2) for longissimus lumborum and gluteus medius, respectively. Two measures of tenderness were determined for each muscle. A variable (composite tenderness) was derived via principal component analysis to represent both overall tenderness ratings and slice shear force values using the PCA() function of the FactoMineR (Lê et al., 2008) package in R (R Development Core Team, 2025). Composite tenderness (i.e., component 1) was used as the dependent variable for regression equations predicting tenderness.
Marginal effects of muscle pH, initial metmyoglobin formation, and carbonyls on sarcoplasmic proteins on composite tenderness values of longissimus lumborum steaks.
Marginal effects of glucose concentration. lactate concentration, glucose-6-phosphate concentration, and initial metmyoglobin formation on composite tenderness values of longissimus lumborum steaks. IMF, initial metmyoglobin formation.
We used a bidirectional stepwise selection approach (Efroymson 1960) to determine the best set of variables to predict composite tenderness for each muscle. This bidirectional approach ensures completeness of the model and prevents the elimination of important interactions (which may otherwise be removed due to lack of significance of first-order effects). We explored linear and generalized additive effects to ensure detection of both linear and nonlinear relationships. We began with an initial model, including all variables and interactions. Linear models were generated using the lm() function of base R and GAM models were generated with the gam() function from the mgcv package (Wood, 2011). The fit of models was compared using the AICc statistic, deviance, and analysis of variance and predicted values alongside raw data, and the best fit was used in further analysis. Variables with a P value less than .10 for prediction of composite tenderness were kept for further model development.
In the first round of selection (backward), nonsignificant interaction terms (P > .10) were removed from the model. Main-effect terms involved in an interaction were retained in the model. Initially, muscle pH and glycolytic variables (glycogen, glucose, glucose-6- phosphate, and lactate) were all included in model development. However, these variables displayed a high degree of collinearity and resulted in over-fitted models. Thus, a separate regression was done with each of these variables and their 2-way interactions, and the model was reduced. The result in longissimus lumborum was that muscle pH, glucose, glucose-6-phosphate, lactate, glucose × glucose-6-phosphate, and glucose × lactate were kept in the model. Any interactions involving muscle pH or glycolytic traits with nonglycolytic traits were kept in model development. Similarly, bloom and initial metmyoglobin formation were highly correlated. Initial metmyoglobin formation was more predictive of composite tenderness in longissimus lumborum muscles and, thus, was used for model development for that muscle. For the gluteus medius muscle pH, glycogen, glucose, glucose-6-phosphate, glucose × glucose-6-phosphate, and glucose × lactate were the glycolytic traits used in the initial model. Additionally, bloom was included in the initial model for gluteus medius steaks.
In the second round of selection (forward), each variable that was excluded from model development was plotted against each variable kept for model development. When the plot suggested an interaction, that interaction was tested, and if significant, the variable and the interaction were included in model development. Model fit was presented in terms of variance explained by the model (e.g., deviance), as this value is straightforward in its interpretation that the model describes a certain percentage of variance in the dependent variable, with the remainder being interpreted as error.
For both muscles, the final model was used to generate predictions for composite tenderness by holding all variables at the median for the dataset, except for the variables involved in the main effect or interactions being explored that were varied across the range observed within the dataset. The predicted values were plotted to demonstrate the effect of changing each variable on predicted composite tenderness values.
Results
Prediction of longissimus lumborum tenderness
The results of principal component analysis of tenderness traits in longissimus lumborum muscles were used to generate a composite tenderness value, which corresponded to the first principal component (Supplementary Figure 1A). Increased ratings for overall tenderness were associated with negative loadings for composite tenderness, while increased slice shear force values were associated with positive loadings for composite tenderness. The relationship of composite tenderness to overall tenderness ratings and slice shear force are presented in Supplementary Figures 1B and 1C, respectively. Composite tenderness explained 76% of the variation in both tenderness traits.
Initial analyses involved testing predictive relationships, including linear or curvilinear (spline) fits (Supplementary Table 3). Muscle pH and metabolites associated with glycolysis displayed curvilinear relationships to composite tenderness, which represented variation in tenderness. Similarly, the abundance of carbonyls on sarcoplasmic proteins, initial metmyoglobin formation, bloom (oxygen consumption), and the proportion of desmin degraded had a curvilinear relationship to composite tenderness. At the level of significance designated as the threshold for further analysis (P ≤ .10), mitochondrial copy number and peroxiredoxin-2 had linear relationships to composite tenderness. Sarcomere length had virtually no relationship to composite tenderness values but was included in later model development because sarcomere length interacted with carbonyls on proteins in the soluble fraction, glucose-6-phosphate, and lactate to affect composite tenderness. Several 2-way interactions existed (P < .10) among glycolytic traits, mitochondrial traits, and chaperones.
The reduced, final model is presented in Supplementary Table 4. This model explained 97% of the deviance in composite tenderness. As noted, sarcomere length was included in the initial model because of interactions with glycolytic metabolites, but these interactions were removed from the model due to lack of significance. However, the main effect of sarcomere length impacted composite tenderness in the final model (P = .003) and was the only main effect not involved in an interaction left in the model.
The reduced model for composite tenderness explained 73% and 74% of the variation in slice shear force values and overall tenderness ratings, respectively (Supplementary Figure 2). This is very close to the variance in each of these traits explained by composite tenderness. Thus, it is apparent that the reduced model sufficiently fits the observed tenderness traits in the dataset to enable visualization of marginal effects.
Marginal effects of each individual variable remaining in the final model were depicted by using the final model to predict composite tenderness. These predictions were made using the median value of each variable observed in the present experiment, except for the variable of interest. Values for that variable equally spanning the range observed in the present experiment (100 predictions) were plugged into the equation to estimate the effects of changes in that variable on composite tenderness. These predictions were plotted alongside the fit obtained by regressing that variable alone against composite tenderness and the raw data (Supplementary Figure 3). It must be noted that the predictions from the final model included main effects and multiple interactions. Thus, for some variables, the fit is quite different from the one observed by single variable regression. This is particularly evident for sarcomere length, which had virtually no relationship to composite tenderness individually, but when the other variables were considered, increased sarcomere length was associated with lower values for composite tenderness (increased tenderness). Similarly, when considered alone, increased peroxiredoxin-2 was associated with greater values for composite tenderness. However, when considered along with all other variables in the model, the inverse relationship was observed. These results indicate the interaction of multiple traits to affect tenderness.
Marginal effects of peroxiredoxin-2 abundance, extent of desmin degradation, and abundance of carbonyls on sarcoplasmic proteins on composite tenderness values of longissimus lumborum steaks.
Because most of the variables in the model were involved in interactions, plots were generated to illustrate the nature of these interactions. Figure 1 depicts the effects of altering muscle pH while also changing the level of initial metmyoglobin formation and the abundance of carbonyls on sarcoplasmic proteins. Increased muscle pH was associated with lower predicted values for composite tenderness. This effect interacted with initial metmyoglobin formation in relation to tenderness. The highest values of initial metmyoglobin depicted in Figure 1 were associated with lower values for composite tenderness relative to intermediate or low values of initial metmyoglobin formation. This value is similar to the mean for the glycolytic cluster reported in King et al. (2025). The mean of the high pH cluster reported in that investigation was in between the other 2 initial metmyoglobin levels depicted in Figure 1. Predicted composite tenderness values were slightly lower when very low values for carbonyls on sarcoplasmic proteins were relative to higher values for this variable.
Predicted composite tenderness values with different levels of glucose, glucose-6-phosphate, lactate, and initial metmyoglobin formation are presented in Figure 2. At very low levels of lactate, increasing glucose levels above approximately 6 μmol/g resulted in decreased predicted composite tenderness values, while increased glucose drastically increased predicted composite tenderness values at higher lactate concentrations. Intermediate lactate concentrations resulted in higher predicted composite tenderness values in association with greater glucose levels relative to higher lactate levels. In muscles with very low levels of initial metmyoglobin formation, changes in tenderness associated with increased glucose concentrations were less extensive than in muscles with greater initial metmyoglobin formation.
Figure 3 depicts the changes in composite tenderness associated with changes in peroxiredoxin-2 abundance as affected by the degradation of the cytoskeletal protein desmin and the abundance of carbonyls on sarcoplasmic protein. The abundance of carbonyls on sarcoplasmic proteins had a profound effect on the relationship of peroxiredoxin-2 concentration to composite tenderness. At 1.5 μmol/mg of carbonyls on sarcoplasmic proteins, increased peroxiredoxin-2 abundance was associated with lower composite tenderness (greater tenderness) values. At 2.5 μmol/mg sarcoplasmic carbonyls, increasing peroxiredoxin-2 abundance had little effect on composite tenderness values. At 3.5 μmol/mg carbonyls on sarcoplasmic proteins, increasing peroxiredoxin-2 content was associated with increased values of composite tenderness. Desmin degradation resulted in increased tenderness, but this effect was greater with greater abundance of peroxiredoxin-2.
Prediction of gluteus medius tenderness
A principal component analysis of gluteus medius slice shear force and overall tenderness ratings resulted in the first principal component, explaining much of the variation in both traits (Supplementary Figure 4A). Similar to the results from the longissimus lumborum, composite tenderness (component 1) explained 78% of the variance in each of these 2 traits (Supplementary Figure 4 B and C, respectively) and, thus, provides a variable that effectively combines the 2 tenderness measures.
Marginal effects of mitochondrial protein abundance, glycogen concentration, and muscle pH on composite tenderness values of gluteus medius steaks.
Individual metabolic traits were tested for their relationship to composite tenderness (Supplementary Table 5). Interestingly, within the gluteus medius most of the metabolic traits exhibited linear relationships to composite tenderness. Muscle pH, glycolytic potential, glycogen concentration, glucose-6-phosphate, bloom (oxygen consumption), and peroxiredoxin-2 displayed curvilinear relationships to composite tenderness in gluteus medius steaks so that spline models provided a better fit than linear models did. This contrasts with the longissimus lumborum results in which a greater proportion of the variables had a curvilinear relationship to composite tenderness. Similar to results observed for longissimus lumborum steaks, numerous interactions among glycolytic, mitochondrial, and chaperone traits indicate an interdependence among metabolic pathways in regulating beef tenderness.
Carcasses were selected for the present experiment based on muscle pH of the longissimus thoracis at grading. Thus, muscle pH was highly variable in the longissimus lumborum samples. However, muscle pH in the gluteus medius muscles from these same carcasses was much less variable (King et al., 2025). This may explain the more linear relationships between metabolic traits and tenderness in gluteus medius steaks compared to longissimus lumborum steaks.
The terms included in the final model predicting composite tenderness are presented in Supplementary Table 6. This model explained 88.8% of the variance in gluteus medius composite tenderness values. The relationships between the predicted composite tenderness and observed composite tenderness values, overall tenderness ratings, and slice shear force values are depicted in Supplementary Figure 5A, B, and C, respectively. The predicted composite tenderness values explained 63% and 76% of the variance observed in overall tenderness ratings and slice shear force values, respectively. The marginal effects of each individual variable on composite tenderness when all other variables are held constant are presented in Supplementary Figure 6. As noted for the longissimus lumborum tenderness, many of the variables included in the final model were involved in interactions regarding composite tenderness. As such, these individual plots do not fully explain the relationships of these traits to tenderness.
Marginal effects of glucose-6-phosphate concentration, peroxiredoxin-2 abundance, mitochondrial carbonyl abundance, and bloom on composite tenderness values of gluteus medius steaks.
Marginal effects of malate concentration, glucose-6-phosphate concentration, and glucose concentration on composite tenderness values of gluteus medius steaks.
Figure 4 depicts the interaction of mitochondrial protein abundance with muscle pH and glycogen concentration to affect predicted composite tenderness values. The impact of increasing mitochondrial protein abundance on composite tenderness predictions is highly dependent on muscle pH. At low ultimate muscle pH, value increases in mitochondrial protein abundance resulted in slightly lower composite tenderness predictions. However, at a muscle pH of 5.8, increased mitochondrial protein was associated with markedly higher predictions of composite tenderness. At an ultimate muscle pH of 6.2, very low mitochondrial abundance resulted in lower composite tenderness predictions than lower muscle pH; however, increases in composite tenderness predictions with increased mitochondrial protein abundance were greater than those at lower muscle pH values. At all muscle pH values, glycogen concentration had little effect on the predicted composite tenderness when mitochondrial protein abundance was low. However, at higher mitochondrial protein abundance, increasing glycogen concentration resulted in slightly greater predicted values for composite tenderness.
The effects of glucose-6-phosphate on composite tenderness values as impacted by peroxiredoxin-2 abundance, the concentration of carbonyls on mitochondrial proteins, and bloom are shown in Figure 5. Generally, as glucose-6-phosphate increased, composite tenderness values decreased, although an inflection point existed at approximately 7 μmol/g glucose-6-phosphate. At 5 μmol/mg mitochondrial carbonyls, the decrease in composite tenderness associated with increased glucose-6-phosphate was pronounced. This effect was less dramatic at 10 μmol/g mitochondrial carbonyl content, and composite tenderness values increased slightly when glucose-6-phosphate concentration increased beyond the inflection point. However, at very low glucose-6-phosphate levels, composite tenderness values were lower than those observed with lower mitochondrial carbonyl concentration. At 15 μmol/mg mitochondrial carbonyl concentration and very low glycose-6-phosphate, composite tenderness values were lower than at lower mitochondrial protein carbonyl concentration, and similar glucose-6-phosphate levels and increasing glucose-6-phosphate up to the inflection point resulted in slight deceases in composite tenderness values. However, at glucose-6-phosphate levels exceeding the inflection point, composite tenderness predicted values increased substantially.
Increasing bloom (i.e., reducing oxygen consumption) lowered predicted values for composite tenderness. At all levels of mitochondrial carbonyls, peroxiredoxin-2 abundance of 300 ng/g produced higher composite tenderness values greater than either 200 ng/g or 450 ng/g peroxiredoxin-2. Whereas either 200 ng/g or 450 ng/g peroxiredoxin-2 resulted in similar composite tenderness values at higher levels of mitochondrial carbonyls, when the abundance of mitochondrial carbonyls was low, 200 ng/g peroxiredoxin-2 produced lower predicted composite tenderness values than 450 ng/g peroxiredoxin-2.
Figure 6 depicts the effects of increasing malate concentration at differing levels of glucose-6-phosphate and glucose. At the lowest glucose-6-phosphate level, increasing malate did not affect the predicted composite tenderness value. However, at higher levels of glucose-6-phosphate, increased malate concentration was associated with slight increases in composite tenderness value. At the lowest value for glucose, the lowest value for glucose-6-phosphate produced greater composite tenderness value than higher concentrations of glucose-6-phosphate. As glucose concentration increased, the magnitude in difference in composite tenderness values associated with differing glucose-6-phosphate levels increased.
The impact of glucose, glucose-6-phosphate, and TBARS levels on composite tenderness predictions are presented in Figure 7. The impact of increased glucose concentration was dependent on the concentration of glucose-6-phosphate in the muscle. At the very lowest concentrations of glucose, composite tenderness values were similar across levels of glucose-6-phosphate. Under the lowest concentration of glucose-6-phsophate, increasing glucose concentration resulted in greater predicted composite tenderness values. At the intermediate glucose-6-phosphate level, increasing glucose-6-phosphate had minimal impact on predicted composite tenderness values. At the highest glucose-6-phosphate concentration, increasing glucose levels resulted in slightly lower predictions for composite tenderness. Increasing TBARS values resulted in lower composite tenderness values.
Marginal effects of glucose concentration, glucose-6-phosphate concentration, and 2-thiobarbituric acid reactive substances concentration on composite tenderness values of gluteus medius steaks. MDA, Malonaldehyde; TBARS, thiobarbituric acid reactive substances.
The concentration of mitochondrial proteins interacted with the abundance of carbonyls on mitochondrial proteins and sarcomere length to affect predicted composite tenderness values (Figure 8). At very low levels of mitochondrial protein, increasing the sarcomere length resulted in lower predicted values of composite tenderness. As sarcomere length increased up to approximately 1.7 μm, the differences in composite tenderness values associated with mitochondrial protein levels diminished. The impact of shorter sarcomere length was greater when the concentration of mitochondrial protein was lower, and at longer sarcomere lengths, mitochondrial protein had little impact on composite tenderness values. As the abundance of carbonyls on mitochondrial proteins increased, composite tenderness values were higher in steaks with shorter sarcomere lengths.
Marginal effects of mitochondrial protein abundance, sarcomere length, and abundance of carbonyls on mitochondrial proteins on composite tenderness values of gluteus medius steaks.
Discussion
A great deal of research has been conducted examining the relationships of metabolic characteristics to beef tenderness. Some of those reports examined some of the factors examined in the present experiment. However, very few investigations have attempted to examine interactions among such factors affecting tenderness. To our knowledge, no other studies have reported the effects of interactions among metabolic factors while accounting for factors not involved in that specific interaction. Thus, the present study is unique in examining interrelationships among metabolic factors and tenderness. The metabolic factors measured in the present experiment were effective in predicting tenderness of both longissimus lumborum and gluteus medius steaks. While these models are likely over-fit, they are useful in examining the interactions among metabolic traits. The 2 models reflect different relationships of metabolic traits to tenderness. Despite being obtained from the same carcasses, the range in muscle pH was much less in the gluteus medius muscles compared to the longissimus lumborum muscles (King et al., 2025). Moreover, the muscle specific nature of relationships among traits regulating beef tenderness is understood.
A great deal of research has been conducted using muscle pH as a proxy for metabolic status on beef tenderness (Anderson et al., 2012a; Eilers et al., 1996; Marsh et al., 1981; Watanabe et al., 1996). Generally, ultimate muscle pH exhibits a curvilinear relationship to tenderness, with the least tender beef having intermediate pH values (Grayson et al., 2016; Jeremiah et al., 1991; Wulf and Page, 2000). It is likely that the curvilinear relationship is the result of the direct effects of pH on the physical structure of the muscle in combination with the effects of metabolic factors associated with changes in muscle pH.
In the present experiment, the main effects of muscle pH indicated a linear relationship between muscle pH and tenderness in both muscles. In longissimus lumborum steaks, increases in muscle pH decreased predictions of tenderness. Moreover, muscle pH interacted with initial metmyoglobin formation such that high levels of initial metmyoglobin were associated with less tender beef. High levels of initial metmyoglobin formation indicate low levels of metmyoglobin reducing ability, which is an indicator of lesser mitochondrial function. Metmyoglobin reducing ability has been reported by our laboratory to be strongly, positively correlated to muscle pH (Cooper et al., 2025; King et al., 2024; King et al., 2025), and in those reports, we speculated that higher muscle pH is favorable to maintaining mitochondrial function in postmortem muscle. These results indicate that at a given muscle pH, lesser mitochondrial function resulted in lesser tenderness of longissimus lumborum steaks. In gluteus medius steaks, increased muscle pH increased the impact of greater abundance of mitochondrial protein. This further indicates an interaction between mitochondria and muscle pH (which is primarily driven by glycolysis) in regulating tenderness. Matarneh et al. (2017) previously reported that the presence of mitochondrial protein increases glycolytic flux, which would lower muscle pH.
Recent initiatives in beef tenderness research using molecular tools have implicated several metabolic pathways and mechanisms in impacting beef tenderness (Antonelo et al., 2020; Gagaoua et al., 2021; King et al., 2019). These findings provide evidence of metabolic impacts on meat tenderness that may be independent of muscle pH. Findings from our laboratory have indicated that considering a number of metabolic factors provides greater insight into the metabolic status of the muscle compared to muscle pH alone (King et al., 2024; King et al., 2025).
Glucose and glucose-6-phosphate have been reported to be increased in tender beef longissimus muscles compared to tough beef longissimus muscles at aging times ranging from 2 d to 28 d postmortem (King et al., 2019) Moreover, Cooper et al. (2025) reported negative correlations between glucose and glucose-6-phosphate and slice shear force of longissimus lumborum steaks after 12 d and 26 d of aging. Schulte et al. (2023) reported greater abundance of these metabolites in beef longissimus muscles with more rapid postmortem pH decline relative to muscles with slower pH decline. Those investigators also reported greater tenderness in the muscles with more rapid pH decline at 1 d postmortem, but not at 3, 7, or 14 d postmortem.
Phosphoglucomutase catalyzes the conversion of glucose-1-phosphate liberated from glycogen to glucose-6-phosphate. The equilibrium constant of this reaction favors the formation of glucose-6-phosphate (Colowick and Sutherland, 1942). Phosphoglucomutase abundance and phosphorylation have been reported to reflect tenderness differences in beef longissimus steaks (Anderson et al., 2014). The equilibrium of the reaction of glucose-6-phosphate to fructose-6-phosphate, catalyzed by phosphoglucoisomerase, favors glucose-6-phosphate (Stödeman and Schwarz, 2004). The next step in the glycolytic pathway is catalyzed by phosphofructokinase and is a control step in the pathway. Phosphofructokinase is inhibited and ultimately inactivated by decreasing pH and ceases glycolysis (Rhoades et al., 2005). Glucose-6-phosphate can also enter several alternative pathways (Rajas et al., 2019).
In the model for gluteus medius steaks, malate interacted with glucose-6-phosphate to affect tenderness. Although only 1 interaction involving malate remained in the final model, interactions among malate and peroxiredoxin-2, initial metmyoglobin formation, and glycogen were included in model development. Malate is an intermediate of the Kreb’s cycle, indicating the importance of mitochondrial metabolism in regulating tenderness. Additionally, malate can be involved in alternative pathways that may contribute to variation in tenderness. Previously, we reported relationships among malate, tenderness, and color stability and speculated that malate’s involvement in the malate-aspartate shuttle moving NADH to the mitochondria is important to regulating tenderness and color stability (Cooper et al., 2025). Additionally, Schulte et al. (2024) reported more abundant malate and malic enzyme in the early postmortem (1 h) proteome and metabolome of tender beef longissimus. Malic enzyme catalyzes the conversion of malate to pyruvate, producing nicotinamide adenine dinucleotide phosphate + hydrogen (NADPH) (Bottacchi and Di Donato, 1983). NADPH can then be used by NADPH oxidases (NOX1, 2, and 4), which produce reactive oxygen species (ROS) (Ferreira and Laitano, 2016). These ROS function in signaling and can induce ROS production by mitochondria and interact with chaperones such as heat shock protein 70 and peroxiredoxin-2 (Ferreira and Laitano, 2016; Stretton et al., 2020). In the present experiment, the interactions observed between malate and glycolytic metabolites, peroxiredoxin-2, and mitochondrial function (initial metmyoglobin formation) support both potential mechanisms and highlight the importance of relationships among glycolytic and mitochondrial metabolism and chaperones.
Desmin degradation and sarcomere length are component traits that have been used extensively to explain biological variation on beef tenderness. However these traits generally leave variation unexplained (King et al., 2003; King et al., 2009a; Rhee et al., 2004) and have failed to characterize tenderness differences across treatments (Cooper et al., 2025; Shackelford et al., 2012). The lack of consistency in the relationships between these traits and tenderness has been troubling researchers who try to characterize and optimize beef tenderness under differing production systems.
In the present experiment, desmin degradation was not predictive of gluteus medius steak tenderness, which had been subjected to just 1 aging time. In previous reports from our laboratory (King et al., 2009a), desmin degradation was not correlated to tenderness variation in gluteus medius steaks within aging times but was correlated to tenderness variation across aging times. In the present experiment, increased desmin degradation resulted in increased tenderness of longissimus lumborum steaks. However, the differences attributable to desmin degradation were less at low levels of peroxiredoxin-2 abundance than at higher levels of peroxiredoxin-2.
When considered alone, sarcomere length did not contribute to tenderness of longissimus lumborum steaks. However, when the effects of the other variables in the prediction model were accounted for, increased sarcomere length resulted in increased tenderness of longissimus lumborum steaks. Smulders et al. (1990) reported sarcomere length was not related to tenderness of beef loins with 3-h pH less than 6.3 but was highly correlated to tenderness of beef loins with a 3-h pH greater than 6.3. In gluteus medius steaks, the effect of sarcomere length on tenderness was highly dependent on the abundance of mitochondrial proteins as well as the abundance of carbonyls on mitochondrial proteins. The presence of mitochondrial protein has been reported to increase the rate of glycolysis (Matarneh et al., 2017). These results indicate that the degree to which tenderness variation can be explained by component traits is dependent on the redox status of early postmortem muscle.
Conclusion
Literature pertaining to the proteomic and metabolomic basis for variation in tenderness is growing and consistently implicates numerous metabolic pathways. However, the results regarding the importance of various biomarkers are mixed. Results of the present experiment confirm the importance of glycolytic flux, mitochondrial function, oxidation, and chaperone proteins in influencing beef tenderness. Moreover, these results indicate that the influence of any one of these pathways is interdependent with the status of other pathways. The impact of component traits on beef tenderness is mediated through the status of multiple metabolic processes. Thus, the complexity of the regulation of beef tenderness is highlighted by the numerous interactions among these pathways. Additional work elucidating the interactions among pathways and how these interactions can be influenced via management to affect tenderness is warranted.
Conflict of Interest
The authors declare no conflicts of interest regarding the content of this manuscript.
Acknowledgments
Research is funded by National Cattlemen’s Beef Association, a contractor to the Beef Checkoff. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture (USDA). The authors express appreciation for S. M. Lonergan and L. G. Johnson of Iowa State University for their insightful comments on these data. The authors are grateful to Megan Landes-Murphy, Peg Ekeren, Kristen Ostdiek, and Casey Trambly of the US Meat Animal Research Center for their assistance in the execution of this experiment and to Joanna VanDenBoom of the US Meat Animal Research Center for her secretarial assistance. USDA is an equal opportunity provider and employer.
Author Contribution
D. A. King contributed to conceptualization, data curation, methodology, data analysis, writing, and original draft preparation; S. D. Shackelford contributed to methodology, data curation, and editing; D. J. Nonneman contributed to methodology and editing; T. S. Katz contributed to data analysis and editing; and T. L. Wheeler contributed to methodology and editing.
Literature Cited
American Meat Science Association. 2016. Research guidelines for cookery, sensory evaluatoin, and instrumental tenderness measurements of meat. 2nd ed. AMSA; Champaign, IL.
Anderson, M. J., S. M. Lonergan, C. A. Fedler, K. J. Prusa, J. M. Binning, and E. Huff-Lonergan. 2012a. Profile of biochemical traits influencing tenderness of muscles from the beef round. Meat Sci. 91:247–254. doi: https://doi.org/10.1016/j.meatsci.2012.01.022
Anderson, M. J., S. M. Lonergan, and E. Huff-Lonergan. 2012b. Myosin light chain 1 release from myofibrillar fraction during postmortem aging is a potential indicator of proteolysis and tenderness of beef. Meat Sci. 90:345–351. doi: https://doi.org/10.1016/j.meatsci.2011.07.021
Anderson, M. J., S. M. Lonergan, and E. Huff-Lonergan. 2014. Differences in phosphorylation of phosphoglucomutase 1 in beef steaks from the longissimus dorsi with high or low star probe values. Meat Sci. 96:379–384. doi: https://doi.org/10.1016/j.meatsci.2013.07.017
Antonelo, D. S., N. R. B. Cônsolo, J. F. M. Gómez, M. Beline, R. S. Goulart, R. R. P. S. Corte, L. A. Colnago, M. W. Schilling, D. E. Gerrard, and S. L. Silva. 2020. Metabolite profile and consumer sensory acceptability of meat from lean Nellore and Angus×Nellore crossbreed cattle fed soybean oil. Food Res Int. 132:109056. doi: https://doi.org/10.1016/j.foodres.2020.109056
Bendall, J. R., editor. 1973. Postmortem changes in muscle. Academic Press, New York, NY
Bergmeyer, H. U., editor. 1974. Methods of enzymatic analysis. Academic Press, New York, NY.
Bottacchi, E., and S. Di Donato. 1983. Skeletal muscle NAD+ (P) and NADP+-dependent malic enzyme in Friedreich’s ataxia. Neurology. 33:712. doi: https://doi.org/10.1212/WNL.33.6.712
Cawthon, D., K. Beers, and W. G. Bottje. 2001. Electron transport chain defect and inefficient respiration may underlie pulmonary hypertension syndrome (ascites)-associated mitochondrial dysfunction in broilers. Poult. Sci. 80:474–484. doi: https://doi.org/10.1093/ps/80.4.474
Colowick, S. P., and E. W. Sutherland. 1942. Polysaccharide synthesis from glucose by means of purified enzymes. J. Biol. Chem. 144:423–437. doi: https://doi.org/10.1016/S0021-9258(18)72525-8
Cooper, J. V., D. A. King, S. D. Shackelford, P. Ekeren, C. L. Lorenzen, and T. L. Wheeler. 2025. Color stability and tenderness attributes of beef longissimus lumborum steaks at 2 different aging periods are impacted by metabolic characteristics. Meat and Muscle Biology. 9:18387,1–18. doi: https://doi.org/10.22175/mmb.18387
Cross, H. R., R. Moen, and M. S. Stanfield. 1979. Training and testing of judges for sensory analysis of meat quality. Food Technol. 32:48–54.
Cross, H. R., R. L. West, and T. R. Dutson. 1981. Comparison of methods for measuring sarcomere length in beef semitendinosus muscle. Meat Sci. 5:261–266. doi: https://doi.org/10.1016/0309-1740(81)90016-4
D’Alessandro, A., C. Marrocco, S. Rinalducci, C. Mirasole, S. Failla, and L. Zolla. 2012. Chianina beef tenderness investigated through integrated Omics. J. Proteomics. 75:4381–4398. doi: https://doi.org/10.1016/j.jprot.2012.03.052
Efroymson, M. A. 1960. Multiple Regression Analysis, Mathematical Methods for Digital Computers, ed. A. Ralston and H.S. Wilf, John Wiley & Sons Inc., New York.
Eilers, J. D., J. D. Tatum, J. B. Morgan, and G. C. Smith. 1996. Modification of early-postmortem muscle pH and use of postmortem aging to improve beef tenderness. J. Anim. Sci. 74:790–798. doi: https://doi.org/10.2527/1996.744790x
Ferreira, L. F., and O. Laitano. 2016. Regulation of NADPH oxidases in skeletal muscle. Free Radical Biol. Med. 98:18–28. doi: https://doi.org/10.1016/j.freeradbiomed.2016.05.011
Gagaoua, M., E. M. C. Terlouw, A. M. Mullen, D. Franco, R. D. Warner, J. M. Lorenzo, P. P. Purslow, D. Gerrard, D. L. Hopkins, D. Troy, and B. Picard. 2021. Molecular signatures of beef tenderness: underlying mechanisms based on integromics of protein biomarkers from multi-platform proteomics studies. Meat Sci. 172:108311. doi: https://doi.org/10.1016/j.meatsci.2020.108311
Grayson, A. L., S. D. Shackelford, D. A. King, R. O. McKeith, R. K. Miller, and T. L. Wheeler. 2016. The effects of degree of dark cutting on tenderness and sensory attributes of beef. J. Anim. Sci. 94:2583–2591. doi: https://doi.org/10.2527/jas.2016-0388
Hammelman, J. E., B. C. Bowker, A. L. Grant, J. C. Forrest, A. P. Schinckel, and D. E. Gerrard. 2003. Early postmortem electrical stimulation simulates PSE pork development. Meat Sci. 63:69–77. doi: https://doi.org/10.1016/s0309-1740(02)00057-8
Huff Lonergan, E., W. Zhang, and S. M. Lonergan. 2010. Biochemistry of postmortem muscle—lessons on mechanisms of meat tenderization. Meat Sci. 86:184–195. doi: https://doi.org/10.1016/j.meatsci.2010.05.004
Hunt, M. C., and H. B. Hedrick. 1977. Profile of fiber types and related properties of five bovine muscles. J. Food Sci. 42:513–517. doi: https://doi.org/10.1111/j.1365-2621.1977.tb01535.x
Iqbal, M., N. R. Pumford, Z. X. Tang, K. Lassiter, C. Ojano-Dirain, T. Wing, M. Cooper, and W. Bottje. 2005. Compromised liver mitochondrial function and complex activity in low feed efficient broilers are associated with higher oxidative stress and differential protein expression. Poult. Sci. 84:933–941. doi: https://doi.org/10.1093/ps/84.6.933
Jeremiah, L. E., A. K. W. Tong, and L. L. Gibson. 1991. The usefulness of muscle color and pH for segregating beef carcasses into tenderness groups. Meat Sci. 30:97–114. doi: https://doi.org/10.1016/0309-1740(91)90001-7
King, D. A., M. E. Dikeman, T. L. Wheeler, C. L. Kastner, and M. Koohmaraie. 2003. Chilling and cooking rate effects on some myofibrillar determinants of tenderness of beef. J. Anim. Sci. 81:1473–1481. doi: https://doi.org/10.2527/2003.8161473x
King, D. A., M. C. Hunt, S. Barbut, J. R. Claus, D. P. Cornforth, P. Joseph, Y. H. B. Kim, G. Lindahl, R. A. Mancini, M. N. Nair, K. J. Merok, A. Milkowski, A. Mohan, F. Pohlman, R. Ramanathan, C. R. Raines, M. Seyfert, O. Sørheim, S. P. Suman, and M. Weber. 2023. American Meat Science Association guidelines for meat color measurement. Meat and Muscle Biology. 6:12473,1–81. doi: https://doi.org/10.22175/mmb.12473
King, D. A., R. K. Miller, R. O. McKeith, A. L. Grayson, S. D. Shackelford, K. B. Gehring, J. W. Savell, and T. L. Wheeler. 2024. Multivariate examination of metabolic contributions to beef Longissimus lumborum flavor. Meat and Muscle Biology. 8:17055,1–14. doi: https://doi.org/10.22175/mmb.17055
King, D. A., S. D. Shackelford, C. D. Broeckling, J. E. Prenni, K. E. Belk, and T. L. Wheeler. 2019. Metabolomic investigation of tenderness and aging response in beef longissimus steaks. Meat and Muscle Biology. 3. doi: https://doi.org/10.22175/mmb2018.09.0027
King, D. A., S. D. Shackelford, D. Nonneman, T. S. Katz, and T. L. Wheeler. 2025. Categorization of beef Longissimus lumborum and Gluteus medius muscles based on metabolic attributes is more informative than muscle pH. Meat and Muscle Biology. 9:18388,1–20. doi: https://doi.org/10.22175/mmb.18388
King, D. A., T. L. Wheeler, S. D. Shackelford, and M. Koohmaraie. 2009a. Comparison of palatability characteristics of beef gluteus medius and triceps brachii muscles. J. Anim. Sci. 87:275–284. doi: https://doi.org/10.2527/jas.2007-0809
King, D. A., T. L. Wheeler, S. D. Shackelford, K. D. Pfeiffer, R. Nickelson, and M. Koohmaraie. 2009b. Effect of blade tenderization, aging time, and aging temperature on tenderness of beef longissimus lumborum and gluteus medius. J. Anim. Sci. 87:2952–2960. doi: https://doi.org/10.2527/jas.2009-1803
Koohmaraie, M. 1996. Biochemical factors regulating the toughening and tenderization processes of meat. Meat Sci. 43:193–201. doi: https://doi.org/10.1016/0309-1740(96)00065-4
Lê, S., J. Josse, and F. Husson. 2008. FactoMineR: an R package for multivariate analysis. Journal of Statistical Software. 25:1–18. doi: https://doi.org/10.18637/jss.v025.i01
Marsh, B. B., J. V. Lochner, G. Takahashi, and D. D. Kragness. 1981. Effects of early post-mortem pH and temperature on beef tenderness. Meat Sci. 5:479–483. doi: https://doi.org/10.1016/0309-1740(81)90046-2
Matarneh, S. K., E. M. England, T. L. Scheffler, C.-N. Yen, J. C. Wicks, H. Shi, and D. E. Gerrard. 2017. A mitochondrial protein increases glycolytic flux. Meat Sci. 133:119–125. doi: https://doi.org/10.1016/j.meatsci.2017.06.007
McKeith, R. O., D. A. King, A. L. Grayson, S. D. Shackelford, K. B. Gehring, J. W. Savell, and T. L. Wheeler. 2016. Mitochondrial abundance and efficiency contribute to lean color of dark cutting beef. Meat Sci. 116:165–173. doi: https://doi.org/10.1016/j.meatsci.2016.01.016
Picard, B., C. Barboiron, D. Chadeyron, and C. Jurie. 2011. Protocol for high-resolution electrophoresis separation of myosin heavy chain isoforms in bovine skeletal muscle. Electrophoresis. 32:1804–1806. doi: https://doi.org/10.1002/elps.201100118
Picard, B., M. Gagaoua, M. Al-Jammas, L. De Koning, A. Valais, and M. Bonnet. 2018. Beef tenderness and intramuscular fat proteomic biomarkers: muscle type effect. PeerJ. 6: e4891. doi: https://doi.org/10.7717/peerj.4891
R Development Core Team. 2025. R: a language and environment for statistical computing. https://R-project.org/https://R-project.org/
Rajas, F., A. Gautier-Stein, and G. Mithieux. 2019. Glucose-6 phosphate, a central hub for liver carbohydrate metabolism. Metabolites. 9:282. doi: https://doi.org/10.3390/metabo9120282
Reznick, A. Z., and L. Packer. 1994. Oxidative damage to proteins: spectrophotometric method for carbonyl assay. Methods Enzymol. 233:357–363. doi: https://doi.org/10.1016/s0076-6879(94)33041-7
Rhee, M. S., T. L. Wheeler, S. D. Shackelford, and M. Koohmaraie. 2004. Variation in palatability and biochemical traits within and among eleven beef muscles. J. Anim. Sci. 82:534–550. doi: https://doi.org/10.2527/2004.822534x
Rhoades, R. D., D. A. King, B. E. Jenschke, J. M. Behrends, T. S. Hively, and S. B. Smith. 2005. Postmortem regulation of glycolysis by 6-phosphofructokinase in bovine M. Sternocephalicus pars mandibularis. Meat Sci. 70:621–626. doi: https://doi.org/10.1016/j.meatsci.2005.01.024
Rowe, L. J., K. R. Maddock, S. M. Lonergan, and E. Huff-Lonergan. 2004. Influence of early postmortem protein oxidation on beef quality. J Anim. Sci. 82:785–793. doi: https://doi.org/10.2527/2004.823785x
Schmittgen, T. D., and K. J. Livak. 2008. Analyzing real-time PCR data by the comparative CT method. Nat. Protoc. 3:1101–1108. doi: https://doi.org/10.1038/nprot.2008.73
Schulte, M. D., K. G. Hochmuth, E. M. Steadham, S. M. Lonergan, S. L. Hansen, and E. Huff-Lonergan. 2024. Early postmortem beef muscle proteome and metabolome variations due to supranutritional zinc and ractopamine hydrochloride supplementation. J. Anim. Sci. 102:skae272. doi: https://doi.org/10.1093/jas/skae272
Schulte, M. D., K. G. Hochmuth, E. M. Steadham, S. M. Lonergan, S. L. Hansen, and E. J. Huff-Lonergan. 2023. Early postmortem muscle proteome and metabolome of beef longissimus thoracis muscle classified by pH at 6 hours postmortem. J. Proteomics. 271:104756. doi: https://doi.org/10.1016/j.jprot.2022.104756
Shackelford, S. D., D. A. King, and T. L. Wheeler. 2012. Chilling rate effects on pork loin tenderness in commercial processing plants. J. Anim. Sci. 90:2842–2849. doi: https://doi.org/10.2527/jas.2011-4855
Shackelford, S. D., T. L. Wheeler, and M. Koohmaraie. 1999. Evaluation of slice shear force as an objective method of assessing beef longissimus tenderness. J. Anim. Sci. 77:2693–2699.
Smulders, F. J. M., B. B. Marsh, D. R. Swartz, R. L. Russell, and M. E. Hoenecke. 1990. Beef tenderness and sarcomere length. Meat Sci. 28:349–363. doi: https://doi.org/10.1016/0309-1740(90)90048-B
Stödeman, M., and F. P. Schwarz. 2004. Importance of product/reactant equilibration in the kinetics of the phosphoglucose isomerization reaction by differential stopped flow microcalorimetry. Anal. Biochem. 329:307–315. doi: https://doi.org/10.1016/j.ab.2004.03.005
Stretton, C., J. N. Pugh, B. McDonagh, A. McArdle, G. L. Close, and M. J. Jackson. 2020. 2-Cys peroxiredoxin oxidation in response to hydrogen peroxide and contractile activity in skeletal muscle: a novel insight into exercise-induced redox signalling? Free Radical Biol. Med. 160:199–207. doi: https://doi.org/10.1016/j.freeradbiomed.2020.06.020
Taylor, R. C., G. H. Geesink, V. F. Thompson, M. Koohmaraie, and D. E. Goll. 1995. Is Z-disk degradation responsible for postmortem tenderization? J. Anim. Sci. 73:1351–1367.
US Department of Agriculture. 2014. Institutional Meat Purchase Specifications: Fresh Beef Series 100. https://ams.usda.gov/sites/default/files/media/IMPS_100_Fresh_Beef%5B1%5D.pdf. (Accessed 3 May 2018).https://ams.usda.gov/sites/default/files/media/IMPS_100_Fresh_Beef%5B1%5D.pdf
Watanabe, A., C. C. Daly, and C. E. Devine. 1996. The effects of the ultimate pH of meat on tenderness changes during ageing. Meat Sci. 42:67–78. doi: https://doi.org/10.1016/0309-1740(95)00012-7
Wheeler, T. L., S. D. Shackelford, and M. Koohmaraie. 1998. Cooking and palatability traits of beef longissimus steaks cooked with a belt grill or an open hearth electric broiler. J. Anim. Sci. 76:2805–2810. doi: https://doi.org/10.2527/1998.76112805x
Wheeler, T. L., S. D. Shackelford, and M. Koohmaraie. 2000. Variation in proteolysis, sarcomere length, collagen content, and tenderness among major pork muscles. J. Anim. Sci. 78:958–965. doi: https://doi.org/10.2527/2000.784958x
Wood, S. N. 2011. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J. R. Stat. Soc. B. 73:3–36. doi: https://doi.org/10.1111/j.1467-9868.2010.00749.x
Wulf, D. M., and J. K. Page. 2000. Using measurements of muscle color, pH, and electrical impedance to augment the current USDA beef quality grading standards and improve the accuracy and precision of sorting carcasses into palatability groups. J. Anim. Sci. 78:2595–2607. doi: https://doi.org/10.2527/2000.78102595x
Supplementary Materials
Principal component analysis of tenderness traits of longissimus lumborum steaks. Panel A depicts biplot of overall tenderness and slice shear force. Panel B depicts relationship of composite tenderness (Component 1 from principal component analysis) to Overall tenderness ratings. Panel C depicts relationship of composite tenderness (Component 1 from principal component analysis) to slice shear force values.
Fit of regression model using metabolic traits to predict composite tenderness values of beef longissimus lumborum tenderness traits (Panel A). Panel B depicts fit of regression model to overall tenderness ratings. Panel C depicts fit of regression model to slice shear force values.
Marginal effects of metabolic traits in predicting composite tenderness of longissimus lumborum steaks. Black lines represent relationship of raw data to composite tenderness. Blue lines depict marginal effects after holding all other variables in the model constant.
Principal component analysis of tenderness traits of gluteus medius steaks. Panel A depicts biplot of overall tenderness and slice shear force. Panel B depicts relationship of composite tenderness (Component 1 from principal component analysis) to Overall tenderness ratings. Panel C depicts relationship of composite tenderness (Component 1 from principal component analysis) to slice shear force values.
Fit of regression model using metabolic traits to predict composite tenderness values of beef gluteus medius tenderness traits (Panel A). Panel B depicts fit of regression model to overall tenderness ratings. Panel C depicts fit of regression model to slice shear force values.
Marginal effects of metabolic traits in predicting Composite tenderness of gluteus medius steaks. Black lines represent relationship of raw data to composite tenderness. Blue lines depict marginal effects after holding all other variables the model constant.
Simple statistics of metabolic and tenderness traits of longissimus lumborum muscles (n = 100)
| Variable | Mean | SD | Minimum | Maximum | Coefficient of Variation |
|---|---|---|---|---|---|
| Muscle pH | 5.74 | 0.33 | 5.35 | 6.79 | 5.83 |
| Glycolytic potential, μmol/g | 191.90 | 50.50 | 79.50 | 287.58 | 26.32 |
| Glycogen, μmol/g | 8.29 | 7.78 | 0.00 | 26.98 | 93.82 |
| Glucose, μmol/g | 4.92 | 3.06 | 0.05 | 10.84 | 62.23 |
| Glucose-6-phosphate, μmol/g | 5.30 | 4.22 | 0.00 | 14.48 | 79.65 |
| Lactate, μmol/g | 154.89 | 29.69 | 74.84 | 197.86 | 19.17 |
| Malate, μmol/g | 0.42 | 0.42 | 0.00 | 1.84 | 99.85 |
| Myoglobin, mg/g | 4.30 | 0.82 | 2.77 | 7.17 | 19.15 |
| Type I fibers % | 34.91 | 2.57 | 28.37 | 41.55 | 7.37 |
| Mitochondrial copy number | 1625.44 | 496.23 | 517.53 | 2835.58 | 30.53 |
| Mitochondrial protein, mg/g | 0.39 | 0.13 | 0.18 | 0.92 | 34.26 |
| Mitochondrial carbonyls, μmol/mg | 7.62 | 4.02 | 1.07 | 17.50 | 52.79 |
| Protein solubility, % | 11.19 | 2.14 | 3.85 | 15.87 | 19.13 |
| Sarcoplasmic carbonyls, μmol/mg | 1.70 | 0.97 | 0.00 | 5.57 | 56.86 |
| Insoluble carbonyls, μmol/mg | 5.93 | 3.65 | 0.47 | 19.71 | 61.56 |
| Bloom,1 % | 69.71 | 9.31 | 44.93 | 86.56 | 13.35 |
| Initial metmyoglobin formation,2 % | 59.34 | 5.66 | 44.75 | 66.74 | 9.55 |
| TBARS, μmol MDA/g | 11.30 | 7.24 | 1.82 | 37.11 | 64.04 |
| Peroxiredoxin 2, ng/g | 402.99 | 112.16 | 93.03 | 665.26 | 27.83 |
| Heat shock protein 70, ng/g | 4.79 | 1.42 | 2.17 | 9.78 | 29.69 |
| Desmin degraded, % | 74.72 | 23.76 | 0.00 | 99.59 | 31.80 |
| Sarcomere length, μm | 1.75 | 0.11 | 1.45 | 2.19 | 6.10 |
| Overall tenderness | 5.75 | 0.80 | 3.90 | 7.50 | 13.90 |
| Slice shear force, kg | 18.79 | 4.06 | 11.96 | 30.03 | 21.59 |
TBARS, thiobarbituric acid reactive substances.
Bloom = percentage of surface myoglobin in the oxymyoglobin form after incubation with atmospheric oxygen.
Initial metmyoglobin formation = percentage of surface myoglobin in the metmyoglobin form after incubation with 0.3% sodium nitrite.
Simple statistics of metabolic and tenderness traits of gluteus medius muscles (n = 100)
| Variable | Mean | SD | Minimum | Maximum | Coefficient of Variation |
|---|---|---|---|---|---|
| Muscle pH | 5.58 | 0.21 | 5.37 | 6.43 | 3.80 |
| Glycolytic potential, μmol/g | 207.85 | 42.47 | 102.06 | 340.46 | 20.43 |
| Glycogen, μmol/g | 10.66 | 10.55 | 0.00 | 49.86 | 98.95 |
| Glucose, μmol/g | 6.75 | 3.44 | 0.17 | 13.77 | 50.97 |
| Glucose-6-phosphate, μmol/g | 6.14 | 3.55 | 0.00 | 14.31 | 57.75 |
| Lactate, μmol/g | 160.75 | 23.12 | 93.00 | 194.55 | 14.38 |
| Malate, μmol/g | 0.37 | 0.28 | 0.00 | 1.27 | 75.69 |
| Myoglobin, mg/g | 5.62 | 0.94 | 3.96 | 8.21 | 16.67 |
| Type I fibers % | 34.39 | 2.74 | 27.76 | 40.70 | 7.97 |
| Mitochondrial copy number | 1917.07 | 1165.84 | 455.34 | 8857.56 | 60.81 |
| Mitochondrial protein, mg/g | 0.48 | 0.18 | 0.23 | 1.00 | 36.56 |
| Mitochondrial carbonyls, μmol/mg | 7.81 | 3.62 | 1.26 | 19.29 | 46.39 |
| Protein solubility, % | 11.09 | 2.20 | 7.09 | 15.11 | 19.82 |
| Sarcoplasmic carbonyls, μmol/mg | 1.41 | 1.05 | 0.04 | 4.88 | 74.27 |
| Insoluble carbonyls, μmol/mg | 4.61 | 3.11 | 0.30 | 16.12 | 67.39 |
| Bloom1, % | 78.20 | 8.80 | 47.71 | 90.55 | 11.26 |
| Initial metmyoglobin formation2, % | 62.78 | 4.20 | 48.44 | 69.25 | 6.69 |
| TBARS, μmol MDA/g | 12.20 | 6.69 | 4.48 | 29.92 | 54.84 |
| Peroxiredoxin 2, ng/g | 261.45 | 121.67 | 22.17 | 548.01 | 46.54 |
| Heat shock protein 70, ng/g | 6.42 | 2.98 | 2.20 | 20.88 | 46.48 |
| Desmin degraded, % | 87.35 | 10.06 | 42.68 | 99.57 | 11.52 |
| Sarcomere length, μm | 1.62 | 0.14 | 1.21 | 1.86 | 8.58 |
| Overall tenderness | 4.98 | 0.73 | 3.13 | 6.48 | 14.65 |
| Slice shear force, kg | 21.25 | 5.66 | 12.43 | 41.45 | 26.64 |
TBARS, thiobarbituric acid reactive substances.
Bloom = percentage of surface myoglobin in the oxymyoglobin form after incubation with atmospheric oxygen.
Initial metmyoglobin formation = percentage of surface myoglobin in the metmyoglobin form after incubation with 0.3% sodium nitrite.
Regression fit of metabolic traits main effects and significant interactions for predicting composite tenderness of longissimus lumborum steaks
| Predictor | Model Type | R2 | P > F |
|---|---|---|---|
| Muscle pH | Spline | 0.23 | <.001 |
| Glycolytic potential | Spline | 0.25 | <.001 |
| Glycogen | Spline | 0.10 | .07 |
| Glucose | Spline | 0.16 | .003 |
| Glucose-6-phosphate | Spline | 0.15 | .004 |
| Lactate | Spline | 0.21 | <.001 |
| Malate | Linear | 0.001 | .78 |
| Myoglobin | Linear | 0.004 | .51 |
| Myosin heavy chain type I | Spline | 0.01 | .16 |
| Mitochondrial copy number | Linear | 0.03 | .08 |
| Mitochondrial protein | Linear | 0.004 | .53 |
| Carbonyls on mitochondrial proteins | Linear | 0.003 | .56 |
| Protein solubility | Linear | 0.01 | .13 |
| Carbonyls on sarcoplasmic proteins | Spline | 0.12 | .04 |
| Carbonyls on myofibrillar proteins | Linear | 0.01 | .24 |
| Bloom | Spline | 0.16 | .001 |
| Initial metmyoglobin formation | Spline | 0.22 | <.001 |
| 2-thiobarbituric acid reactive substances | Linear | 0.01 | .25 |
| Peroxiredoxin-2 | Linear | 0.16 | <.001 |
| Heat shock protein 70 | Linear | 0.00 | .530 |
| Desmin degraded | Spline | 0.25 | <.001 |
| Sarcomere length | Linear | 0.00 | .97 |
| Muscle pH v sarcoplasmic carbonyls | Spline × spline | 0.27 | .14 |
| Muscle pH v initial metmyoglobin formation | Spline × spline | 0.30 | .02 |
| Muscle pH v peroxiredoxin-2 | Spline × linear | 0.27 | <.001 |
| Glycogen × peroxiredoxin-2 | Spline × linear | 0.16 | <.01 |
| Glucose v glucose-6-phosphate | Spline × spline | 0.21 | <.001 |
| Glucose × lactate | Spline × spline | 0.21 | <.001 |
| Glucose × sarcoplasmic carbonyls | Spline × spline | 0.28 | .03 |
| Glucose × initial metmyoglobin formation | Spline × spline | 0.30 | <.001 |
| Glucose-6-phosphate × sarcomere length | Spline × linear | 0.14 | <.01 |
| Glucose-6-phosphate × sarcoplasmic carbonyls | Spline × spline | 0.19 | .01 |
| Lactate × peroxiredoxin-2 | Spline × linear | 0.27 | <.01 |
| Lactate × sarcomere length | Spline × linear | 0.21 | <.001 |
| Sarcoplasmic carbonyls × sarcomere length | Spline × linear | 0.10 | .05 |
| Sarcoplasmic carbonyls × initial metmyoglobin formation | Spline × spline | 0.23 | <.001 |
| Sarcoplasmic carbonyls × desmin degradation | Spline × spline | 0.38 | <.001 |
| Sarcoplasmic carbonyls × peroxiredoxin-2 | Spline × linear | 0.19 | .01 |
| Initial metmyoglobin formation × peroxiredoxin-2 | Spline × linear | 0.28 | .01 |
| Desmin degradation × peroxiredoxin-2 | Spline × linear | 0.32 | .04 |
| Desmin degradation × mitochondrial copy number | Spline × linear | 0.36 | .03 |
Final model predicting composite tenderness of longissimus lumborum steaks
| Predictor | Type | P Value |
|---|---|---|
| Muscle pH | Spline | .05 |
| Glucose | Spline | .64 |
| Glucose-6-phosphate | Spline | .91 |
| Lactate | Spline | .99 |
| Sarcoplasmic carbonyls | Spline | .008 |
| Initial metmyoglobin formation | Spline | .32 |
| Desmin degradation | Spline | .36 |
| Peroxiredoxin-2 | Linear | .71 |
| Sarcomere length | Linear | .003 |
| Glucose × glucose-6-phosphate | Spline × spline | .02 |
| Glucose × lactate | Spline × spline | <.001 |
| Muscle pH × sarcoplasmic carbonyls | Spline × spline | .04 |
| Muscle pH × initial metmyoglobin formation | Spline × spline | <.001 |
| Glucose: × initial metmyoglobin formation | Spline × spline | <.001 |
| Glucose-6-phosphate × sarcoplasmic carbonyls | Spline × spline | .03 |
| Sarcoplasmic carbonyls: × initial metmyoglobin formation | Spline × spline | .01 |
| Sarcoplasmic carbonyls: × peroxiredoxin-2 | Spline × linear | .04 |
| Desmin degradation: × peroxiredoxin -2 | Spline × linear | .001 |
| R2 | .90 | |
| Deviance explained, % | 97.3 |
Regression fit of metabolic traits main effects and significant interactions for predicting composite tenderness of gluteus medius steaks
| Predictor | Model type | R2 | P > F |
|---|---|---|---|
| Muscle pH | Spline | 0.51 | <.001 |
| Glycolytic potential | Spline | 0.23 | <.001 |
| Glycogen | Spline | 0.26 | <.001 |
| Glucose | Linear | 0.14 | <.001 |
| Glucose-6-phosphate | Spline | 0.47 | <.001 |
| Lactate | Linear | 0.00 | .89 |
| Malate | Linear | 0.02 | .09 |
| Myoglobin | Linear | 0.02 | .18 |
| Myosin heavy chain type I | Linear | 0.02 | .17 |
| Mitochondrial copy number | Linear | 0.00 | .92 |
| Mitochondrial protein | Linear | 0.12 | <.001 |
| Carbonyls on mitochondrial proteins | Linear | 0.13 | <.001 |
| Protein solubility | Linear | 0.02 | .19 |
| Carbonyls on sarcoplasmic proteins | Linear | 0.02 | .19 |
| Carbonyls on myofibrillar proteins | Linear | 0.00 | .89 |
| Bloom | Spline | 0.26 | <.001 |
| Initial metmyoglobin formation | Linear | 0.07 | .006 |
| 2-thiobarbituric acid reactive substances | Linear | 0.11 | <.001 |
| Peroxiredoxin-2 | Spline | 0.13 | .02 |
| Heat shock protein 70 | Linear | 0.05 | .02 |
| Desmin degraded | Linear | 0.00 | .93 |
| Sarcomere length | Linear | 0.37 | <.001 |
| Muscle pH × mitochondrial protein | Spline × linear | 0.55 | .05 |
| Muscle pH × sarcomere length | Spline × linear | 0.58 | <.01 |
| Muscle pH × type I fibers | Spline × linear | 0.55 | <.001 |
| Glycogen × malate | Spline × linear | 0.30 | .07 |
| Glycogen × mitochondrial protein | Spline × linear | 0.29 | .01 |
| Glycogen × sarcomere length | Spline × linear | 0.47 | <.001 |
| Glucose-6-phosphate × bloom | Spline × spline | 0.52 | <.01 |
| Glucose-6-phosphate × glucose | Spline × linear | 0.38 | .10 |
| Glucose-6-phosphate × peroxiredoxin-2 | Spline × spline | 0.55 | .02 |
| Glucose-6-phosphate × malate | Spline × linear | 0.51 | .10 |
| Glucose-6-phosphate × mitochondrial carbonyls | Spline × linear | 0.50 | .06 |
| Glucose-6-phosphate × sarcomere length | Spline × linear | 0.55 | <.01 |
| Bloom × peroxiredoxin-2 | Spline × spline | 0.34 | .04 |
| Bloom × mitochondrial carbonyls | Spline × linear | 0.30 | <.01 |
| Bloom × sarcomere length | Spline × linear | 0.50 | <.001 |
| Glucose × mitochondrial protein | Linear × linear | 0.26 | .06 |
| Glucose × TBARS | Linear × linear | 0.23 | .04 |
| Peroxiredoxin-2 × glucose | Spline × linear | 0.22 | .03 |
| Peroxiredoxin-2 × malate | Spline × linear | 0.22 | .05 |
| Peroxiredoxin-2 × mitochondrial carbonyls | Spline × linear | 0.20 | .02 |
| Peroxiredoxin-2 × TBARS | Spline × linear | 0.29 | .02 |
| Peroxiredoxin-2 × sarcomere length | Spline × linear | 0.41 | <.001 |
| Mitochondrial protein × mitochondrial carbonyls | Linear × linear | 0.23 | .06 |
| Mitochondrial protein × sarcomere length | Linear × linear | 0.51 | .001 |
| Mitochondrial carbonyls × sarcomere length | Linear × linear | 0.49 | <.01 |
TBARS, thiobarbituric acid reactive substances.
Final model predicting composite tenderness of gluteus medius steak
| Predictor | Type | P > F |
|---|---|---|
| Muscle pH | Spline | .003 |
| Glycogen | Spline | .99 |
| Glucose | Linear | .99 |
| Glucose-6-phosphate | Spline | .07 |
| Malate | Linear | .62 |
| Mitochondrial protein | Linear | <.001 |
| Carbonyls on mitochondrial proteins | Linear | .002 |
| TBARS | Linear | .01 |
| Bloom | Spline | .003 |
| Peroxiredoxin-2 | Spline | .64 |
| Sarcomere length | Linear | .003 |
| Muscle pH × mitochondrial protein | Spline × linear | <.001 |
| Glycogen × mitochondrial protein | Spline × linear | <.001 |
| Glucose v TBARS | Linear × linear | .01 |
| Glucose-6-phosphate × glucose | Spline × linear | .002 |
| Glucose-6-phosphate × malate | Spline × linear | .09 |
| Glucose-6-phosphate × carbonyls on mitochondrial proteins | Spline × linear | .01 |
| Glucose-6-phoshate × peroxiredoxin-2 | Spline × spline | <.001 |
| Mitochondrial protein × sarcomere length | Linear × linear | <.001 |
| Carbonyls on mitochondrial proteins × sarcomere length | Linear × linear | .001 |
| Bloom × carbonyls on mitochondrial proteins | Spline × linear | .004 |
| Peroxiredoxn-2 × carbonyls on mitochondrial proteins | Spline × linear | .003 |
| R2 | R2 | .81 |
| Deviance explained | 88.80% |
TBARS, thiobarbituric acid reactive substances.