The effect of postmortem pH decline rate on mitochondrial apoptosis and tenderness of bovine skeletal muscle during aging

This study aimed to investigate the effect of postmortem pH decline rate on mitochondrial dysfunction-induced apoptosis and bovine muscle tenderness during aging. Protein denaturation, reactive oxygen species (ROS) levels, mitochondrial apoptosis factors, and shear force were assessed in bovine muscles with different pH decline rates. The results showed that compared to the slow group, fast pH decline in bovine muscle signicantly increased sarcoplasmic protein denaturation at 6–12 h and mitochondrial ROS levels at 6–24 h (P<0.05), and in turn signicantly accentuated mitochondrial dysfunction and cytochrome c oxidation levels (P<0.05), resulting in caspase-3 activation, apoptosis, and reduced muscle shear force (P<0.05). These results demonstrated that the fast postmortem pH decline enhanced mitochondrial apoptosis and bovine muscle tenderization by inducing ROS accumulation during aging. on the fast pH decline calpain-1 activity at lower microenvironments. apoptosis the regulation of postmortem pH decline rate on bovine muscle study a rst step towards understanding the role of postmortem pH in muscle tenderization during aging and this eld. apoptotic nucleus. TdT-mediated dUTP nick-end labeling (TUNEL) assay was used to detect the apoptotic nucleus. Muscle samples were cut into a block of 1.0 cm×0.5 cm×0.5 cm and in a 4% paraformaldehyde solution more than 48 Samples were and cut into ten µm-thickness tissue sections and transferred onto slides. and rehydration, the sections were immersed with formaldehyde solution containing 3% H 2 O 2 to remove endogenous peroxidase. The sections were permeabilized by immersing them in a solution containing 0.1% Na 2 NO 3 and 0.1% Triton X-100. The permeabilized tissue sections were rinsed and washed with buffer mmol/L NaCl, 4.3 mmol/L disodium phosphate, 2.7 mmol/L KCl, 1.4 mmol/L potassium dihydrogen phosphate, After washing, the sections were blocked by 5% goat serum for 30 min and then washed with the above buffer for 30 min. The labeling enzyme and the solution were prepared in a ratio of 1:9 to achieve the TUNEL reaction solution. The tissue sections were then covered with the solution prepared and incubated at 37°C for 60 min in a light protecting chamber. As for the negative controls, we used the lack of labeling enzyme during the TUNEL solution preparation. Positive controls were incubated with 5.0 U/mL DNase I for min at room temperature before TUNEL reaction was nuclei the tissue sections glycerol and placed under a uorescent for Regular cell nuclei were stained in blue uorescence by DAPI, and apoptotic cell nuclei were stained in green uorescence by TUNEL. At least three elds of view per section were chosen and quantied for each sample (400×), and the apoptotic rate was calculated as the ratio of apoptotic nuclei to regular nuclei. fragmentation index (MFI). MFI was determined by the method as reported previously 2 . In short, 5.0 g of meat was added with 20 mL of MFI buffer (100 mmol/L KCl, 20 mmol/L K 3 PO 4 , 1 mmol/L EDTA, 1 mmol/L MgCl 2 , 1 mmol/L NaN 3 , pH 7.1) and homogenized on ice. The homogenate was then centrifuged at 4°C at 1000×g for 15 min. After that, the supernatant was discarded. Next, 20 mL of MFI buffer was used to re-suspend the precipitate and the samples were homogenized again. The mixture was then centrifuged to discard the supernatant. Afterward, the precipitate was dissolved with 10 mL of MFI buffer, and was ltered through a 200-mesh nylon screen. Another 5 mL of MFI buffer was added to help the myobrillar protein pass through the lter. Filter myobrillar protein was quantied via the biuret method to determine its protein concentration, and its nal concentration was achieved at 0.5 mg/mL via additional MFI buffer. A UV spectrophotometer was used to measure the absorbance at 540 nm. The nal value was obtained by multiplying the reading 200 times to represent the MFI.


Introduction
Tenderness is one of the most important factors that in uence consumer satisfaction with meat and contribute to repeat purchases 1,2 . After slaughter, the occurring chemical and physical changes contribute to the development of muscle into the meat and promote muscle tenderization 3 . Among these factors, glycolysis is one of the most essential processes in the production of nal meat quality, and the rate of glycolysis in the early postmortem period may have a signi cant impact on muscle tenderization progression 4 . The inconsistency in meat tenderness could be attributed to variations in the rate of pH reduction in the early postmortem period 5,6 . Previous reports have suggested that the bovine muscle with a low pH before the onset of rigor mortis was transformed into tender beef 7,8 . These results suggest that the rate of postmortem pH decline had an important in uence on muscle tenderization.
Muscle tenderization is in uenced by the extensive interactions of biochemical processes in carcasses during postmortem aging, which includes the hydrolysis of key myo brillar proteins such as desmin, titin, and troponin-T 9 . The hydrolysis of myo brillar protein is the result of the activation of endogenous proteases, such as calpains, cathepsins, and caspases 10,11 . For decades, mitochondrial apoptosis has been shown to play an important part in the breakdown of myo brillar protein in muscle throughout aging 12,13 . In the mitochondrial apoptosis pathway, release of cytochrome c from mitochondria provided the stimulus for caspase-9 activation through the association of apoptotic protein activator-1 (Apaf-1) and procaspase-9 in the cytoplasm, and subsequently activated caspase-9 followed by caspase-3 activation thus enhancing muscle tenderness 14 . This type of cell death was different from necrosis, it was a process of programmed death caused by a lack of intracellular energy 15 . Furthermore, previous research has demonstrated that cytoplasmic acidi cation caused apoptosis and caspase-3 activation, suggesting that cytoplasmic pH changes occur early in the induction of mitochondrial apoptosis-related pathways 16 . After slaughter and bleeding of the cattle, the oxygen supply to the muscle is disrupted, and the aerobic respiration in the cells is switched to anaerobic glycolysis, resulting in lactate accumulation, ATP hydrolysis, and subsequently muscle acidi cation 17 . During postmortem aging, a fall in pH may further promote mitochondrial dysfunction and lead to apoptosis in bovine muscle 18,4 . In a low pH milieu, Reshkin et al. 19 discovered that various apoptosis-related enzymes, namely DNases and caspases, were activated. As a result, acidi cation causes mitochondrial apoptosis in muscle cells. However, there is still a lack of conclusive evidence connecting pH decline rate to mitochondrial apoptosis in postmortem muscle.
Thus, in the present work, the bovine carcasses of various pH in the early postmortem phase were selected to investigate the in uence of pH decline rate on mitochondrial apoptosis and muscle tenderness during aging. We aimed to clarify the mechanism of postmortem pH regulating muscle tenderness from the mitochondrial apoptosis pathway by analyzing protein denaturation temperature, ROS levels, mitochondrial apoptosis factors, myo brillar fragmentation index (MFI), and shear force in different pH decline rates groups of bovine muscles.

Results And Discussion
Changes in pH of bovine muscle during postmortem aging. The pH of bovine muscle with different pH decline rates during postmortem aging is shown in Fig. 1. The pH of the fast pH decline group decreased signi cantly at 24 h while the pH of the slow group decreased signi cantly at 48 h compared with 0.5 h (P<0.05). However, the pH in the fast group was lower than the slow group by 4.06%, 5.11%, 8.34%, 4.43%, and 2.67% at 0.5, 2, 6, 12, and 24 h (P<0.05), respectively. These ndings suggest a signi cant difference in pH decline rates between the two groups within 24 h postmortem.
After slaughter and bleeding, glycogen metabolism shifts from aerobic to anaerobic and the metabolite lactate accumulates, resulting in a decline in muscle pH 24,25 . Apaoblaza et al. 26 demonstrated that the ultimate pH of carcasses was correlated with glycogen content and a su cient amount of glycogen can supply the necessary fuel for the pH decrease. Previous studies have shown that the breakdown of ATP produced by postmortem muscle glycolysis led to the accumulation of protons (H + ), which was also caused a decrease in pH 27 . The rate of glycolysis is determined by the activity of glycolytic enzymes. Pyruvate kinase, enolase 2, phosphoglucomutase 1, fructose diphosphate aldolase, and enolase 3 are all phosphorylation-regulated enzymes that may be implicated in postmortem muscle glycolysis 28 . In addition, several other pre-and post-mortem factors might affect postmortem muscle glycolysis, such as genetics, pre-slaughter stress, carcass temperature, and S-nitrosylation of glycolytic enzymes, etc. 26,29,30 . In the present study, the fast group of muscles may have consumed glycogen more quickly, resulting in a low pH in the early postmortem period.
Effect of pH decline rate on protein denaturation and mitochondrial ROS accumulation in postmortem bovine muscle. The decrease in pH is generally thought to induce further protein denaturation in muscle. Therefore, we evaluated the in uence of pH decline rate on muscle protein denaturation temperature and ROS levels. A total of three heat absorption peaks were observed in the muscle tissue by DSC scanning, representing the denaturation temperature of the myosin head, sarcoplasmic protein, and actin 31 . The denaturation temperature of a protein increases when its structure is slightly altered. As shown in Table 1, both myosin and sarcoplasmic protein denaturation temperatures increased signi cantly in both groups during postmortem aging (P<0.05), however myosin denaturation temperature in the fast group was 2.60% higher than in the slow group at 12 h (P<0.05). Sarcoplasmic protein denaturation temperature was 1.79% and 1.39% higher than in the slow group at 6 and 12 h, respectively (P<0.05). However, actin did not show any differences in denaturation temperature during the postmortem aging in the two groups. In addition, the uorescence images of ROS in two groups are shown in Fig. 2A. The average uorescence values of ROS in muscle cells, re ecting the ROS levels, were signi cantly higher in the fast group than the slow group at 6-24 h postmortem (P<0.05; Fig. 2B), with a most signi cant difference at 12 h (P<0.01). The above results suggested that fast pH decline promoted myosin and sarcoplasmic protein denaturation and ROS accumulation in the postmortem muscle.
Insu cient ATP production and supply in the postmortem muscle in conjunction with the reduction of pH disrupt the normal redox system, leading to the accumulation of proapoptotic free radicals such as ROS 13,14 . There is a direct correlation between increased ROS levels and lower muscle pH 32 . Previous studies have shown that the antioxidant enzymes such as SOD, CAT, and GSH-Px showed decreased activity at the low pH in postmortem bovine muscle 33 . Sun et al. 34 showed that an ultimate pH of 5.3-5.6 during postmortem muscle aging did not indicate that muscle glycogen was depleted, but rather that the inactivation of enzymes in the glycolytic pathway caused the reaction to halt when the pH fell below a certain threshold. These ndings suggest that lower muscle pH induces a decrease in enzyme activity in the cytoplasm. The enzymes in the cytoplasm generally play an essential role in a neutral environment. Lower pH caused the active core group of the enzyme to deviate from its optimal dissociation state, reducing its capacity to bind to the substrate and catalytic activity 25,35 . In the present study, the fast pH decline group had greater ROS levels than the slow group, which could be owing to changes in the structure and activity of ROS scavenging enzymes by rapidly declining muscle pH. Taken together, it was clear that postmortem fast pH decline contributed to the accumulation of muscle ROS by inducing sarcoplasmic protein denaturation. Effect of pH decline rate on mitochondrial dysfunction and cytochrome c redox state. A mitochondrial dysfunction is a precursor to cell apoptosis. Therefore, the effect of pH decline rate on mitochondrial membrane permeability and cytochrome c redox state were assessed. Fig. 3A showed that the mitochondrial membrane permeability signi cantly increased during postmortem aging in both groups (P<0.05) and signi cantly higher in the fast group by 14.05%, 22.39%,18.34% and 25.28% than in the slow group at 6, 12, 24 and 72 h, respectively (P<0.05). The cytochrome c reduction levels are shown in Fig. 3B. Our results showed that cytochrome c reduction levels decreased signi cantly postmortem in both two groups (P<0.05).
Furthermore, the cytochrome c reduction in the fast group was signi cantly lower than the slow group by 16.71%, 23.39%, 17.05%, and 26.61% at 6, 12, 24, and 120 h, respectively (P<0.05), indicating that fast pH decline induced mitochondrial membrane dysfunction and cytochrome c oxidation during the postmortem aging.
ROS, produced in mitochondria, was an additional indicator for mitochondrial apoptosis 36 . The weakening of antioxidant enzymes and other defense systems in postmortem muscle inevitably led to the production of ROS 37 . Excessive accumulation of ROS induced swelling and destruction of mitochondria membrane and ultimately mediated mitochondrial apoptosis 14,38 . ROS could attack the mitochondrial membrane polyunsaturated fatty acids to produce lipid peroxides such as malondialdehyde, causing mitochondrial swelling and increasing mitochondrial membrane permeability 13,39 . Moreover, ROS induced cytoplasmic Ca 2+ in ux into mitochondria, leading to mitochondrial Ca 2+ overload, which further caused an increase in mitochondrial membrane pore opening and permeability 40 . Subsequently, the release of cytochrome c from damaged mitochondria into the cytoplasm was a universal event for mitochondrial apoptosis 41 , but its catalytic activity was largely dependent on its oxidative state 14 . Suto et al. 42 also showed that the oxidized cytochrome c was more effective in initiating mitochondrial apoptosis. In the present study, fast pH decline facilitated cytochrome c oxidation. Previous studies have showed that the oxidation of cytochrome c was associated with mitochondrial oxidants, such as H 2 O 2 37,43 . Together with the above ndings on the role of pH decline rate in ROS, it suggested that fast pH decline indirectly induced mitochondrial dysfunction and cytochrome c oxidation by promoting the accumulation of mitochondrial ROS.
Effect of pH decline rate on caspase-3, apoptosis and tenderness in bovine muscle. To further exam the downstream events of cytochrome c oxidation and mitochondrial dysfunction, the effect of pH decline rate on caspase-3 activity, apoptotic nucleus, MFI, and shear force were evaluated during postmortem aging. As depicted in Fig. 4, in both two groups, caspase-3 activity tended to increase initially and later decrease (P<0.05), with caspase-3 activity reaching the peak at 24 h. Besides, the caspase-3 activity in the fast group was signi cantly higher than the slow group at 12-168 h (P<0.05), except the 72 h time point. TUNEL assay of the apoptotic nucleus is shown in Fig. 5A. The normal nucleus was labeled with overall blue uorescence (DAPI), while the apoptotic nucleus was labeled with green uorescence (TUNEL positive). The proportion of apoptotic nuclei of the total nuclei was quanti ed and shown in Fig. 5B, with a signi cant increase in the total number of apoptotic nuclei in both groups (P<0.05). In the fast group, the total apoptotic nuclei were signi cantly higher than those in the slow group at 24-168 h (P<0.05).
The number of the MFI re ects the degree of fragmentation of myo brils and positively correlates with the degradation of myo brillar proteins. Table 2 showed a signi cant increase in MFI during aging for both groups (P<0.05), however, the fast group had a signi cantly higher MFI than the slow group at 12-168 h (P<0.05). Shear force was used to evaluate the tenderness of bovine muscle, and the smaller the value indicates the tenderer of the muscle. As shown in Table 2, the shear force of both fast and slow groups kept growing until 12 h (P<0.05). However, it plummeted at 24-168 h (P<0.05). In the fast group, the shear force was signi cantly higher than in the slow group at 6 h (P<0.05), but signi cantly lower than in the slow group at 12-168 h (P<0.05) and 10.49% lower than in the slow group at 168 h. The above results suggested that fast pH decline promoted caspase-3 activation, apoptosis, and bovine muscle tenderization during postmortem aging.
Caspase-3 is one of the most important effector caspases in the apoptotic pathway and plays a critical role in muscle tenderization 44 . The oxidized cytochrome c interacts rapidly with apoptotic protease activating factor (Apaf-1) and procaspase-9 in the presence of dATP to activate caspase-9, which further activates the downstream effector caspase-3 12,13,41 . The results of this study indicated that the caspase-3 showed the highest activity at 24 h time point. Our nding was con rmed by previous reports 45 , in which they suggested that caspase-3 activity in bovine skeletal muscle gradually increased to a maximum at approximately 24 h time point. Besides, our study also showed that the caspase-3 activity in the fast group was higher than that in the slow group. Combining the above effect of pH decline rate on ROS accumulation, mitochondrial dysfunction, and cytochrome c oxidation, we concluded that the fast pH decline promoted postmortem caspase-3 activation and apoptosis in the bovine muscle.
Postmortem muscle tenderization is a complex biochemical process involving the degradation of myo brillar proteins and is controlled by a variety of protein hydrolases, including caspase-3 and calpain-1 46 . According to Kemp and Parr 47 , the formation of 28-and 30-kDa troponin-T degradation products and the degradation of desmin improved the muscle tenderness. Although these proteins were more likely to be recognized by calpain-1, caspase-3 also played an important role in this process. Huang et al. 41 reported that caspase-3 also recognized and interacted with actomyosin, titin, nebulin, and tropomyosin in myo brillar protein and this interaction can trigger further hydrolysis. As a result, caspase-3 is vital in enhancing the tenderness of bovine muscle during aging. In the present study, the bovine muscle tenderness was higher in the fast pH decline group than in the slow group at 12-168 h postmortem and matched the caspase-3 activity. Hwang and Thompson 7 found that bovine longissimus dorsi with lower pH before the onset of rigour mortis produced more tender meat in the later stages of aging, which backed up the current ndings. Given the essential role of pH decline rate in caspase-3 activation, it is possible that fast pH decline increases bovine muscle tenderization via the mitochondrial apoptotic pathway. Our experiment also observed that the tenderness of the fast group was lower than that of the slow group at 6 h postmortem, which could be associated to calpain-1 activity. Previous research has revealed that calpain-1 activation occurred around 6 h postmortem in bovine muscle 48 , and that calpain-1 degradation activity on the myo brillar protein was strongest at pH 7.5 49 . Lomiwesa et al. 50 concluded that although calpain-1 autolysis was fast at a low pH, the hydrolysis of myo brillar proteins was diminished in pork muscle. Despite the fact that no experiments involving calpain-1 activity were conducted in this study, it can be hypothesized, based on previous studies, that the fast pH decline group may have decreased calpain-1 activity at 6 h due to the lower pH microenvironments. Therefore, various processes other than mitochondrial apoptosis may also be involved in the regulation of postmortem pH decline rate on bovine muscle tenderness during aging, and this needs to be investigated further in the future. The current study is simply a rst step towards understanding the role of postmortem pH in muscle tenderization during aging and we believe our ndings will encourage more research in this eld. Determination of the muscle pH. pH was measured using a portable pH meter (Testo® 230 meter, Testo GmbH & Co., Lenzkirch, Germany). The electrode was calibrated with standard buffer solutions at 4.0 and 7.0 pH values (Mallinckrodt Chemicals, Phillipsburg, NJ, USA). The electrode was placed into the muscle at random locations throughout the muscle for measurement. Before the next measurement, the electrode probe was cleaned with ultrapure water (pH 7.0) and wiped dry with lter paper. Each sample was measured three times.
Differential Scanning Calorimetry (DSC) analysis. A PerkinElmer differential scanning calorimeter (Waltham, USA) was used to analyze the denaturation temperature of bovine muscle. Minced muscle tissue (10-15 mg) was enclosed in an aluminum box and heated from 20°C to 100°C at a rate of 3°C/min, using the sealed empty box as a reference. Denaturation temperature (T, °C) was calculated by Pyris-12 software (Perkin Elmer-Instruments, USA).
Determination of reactive oxygen species (ROS). Muscle ROS levels were determined according to a previously described method 21 with some modi cations. Brie y, muscle samples were xed in 4% paraformaldehyde for 48 h and were trimmed, dehydrated in ethanol, followed by embedded in para n before being section into 4.0 µm thickness tissue slices. Tissue slices were dewaxed in xylene and anhydrous ethanol for 10 min, respectively. After being washed with PBS buffer (pH 7.2), the sections were incubated in 5 µmol/L DCFH-DA at 37°C for 60 min protected from light and were subsequently washed off with PBS. Finally, the slices were sealed with an anti-uorescence quenching sealer and placed under a uorescence microscope (CKX31, Olympus, Japan) for observation (Ex=490 nm and Em=520 nm). Sections were scanned (400×) by a panoramic scanner (CaseViewer 2.4, Hungary) and the average uorescence intensity of each cell cross-section was analyzed using the Halo v3.0.311.314 software (Indicalabs, USA).
Extraction of muscle mitochondria. The extraction of mitochondrial and cytoplasmic proteins from muscle tissue was followed as a previously described method 22 . In brief, approximately 5.0 g of muscle samples were mixed with 50 mL of mitochondrial separation medium (1 mmol/L EDTA, 300 mmol/L sucrose, 0.5% bovine serum albumin, 15 mmol/L Tris, pH 7.2) and homogenized in an ice bath. The homogenate was centrifuged at 4°C at 1,000 × g for 10 min. The obtained supernatant was centrifuged at 4°C at 8,000 × g for 20 min. The supernatant contained the cytoplasmic protein in solution while the precipitate was then extracted in mitochondria. The resulting precipitate was mixed with buffer (containing 300 mmol/L sucrose, 1 mmol/L EDTA, 15 mmol/L Tris, pH 7.2) to serve as the mitochondrial solution. The biuret method was employed to measure the nal protein concentration and the bovine serum albumin (Sigma Chemical Co., St. Louis, MO) was used as a standard.
Determination of mitochondrial membrane permeability. Mitochondrial membrane permeability was measured according to a previously described method 13 . The concentration of the mitochondrial solution was adjusted to 2 mg/mL and incubated with four times of volumes of cold test medium (including 250 mmol/L sucrose, 10 mmol/L Tris-HCl, pH 7.4) at room temperature for 3 min. Mitochondrial membrane permeability was determined by analyzing the light absorption value of the mixed solution with a UV spectrophotometer at 520 nm. The lower absorbance indicated the higher mitochondrial permeability.
Determination of cytochrome c redox state. The method to determine cytochrome c redox was based on a previously described method 24 with minor modi cations. Approximately 4.0 g of chopped muscle tissue was homogenized in 16.0 mL of cold separation buffer (0.3 mol/L sucrose, 0.5% bovine serum albumin, 1 mmol/L EDTA, 0.01 mol/L Tris-HCl, pH 7.4) and centrifuged at 4 ℃ at 1,000 × g for 10 min. After that, the supernatant was centrifuged at 4 ℃ at 8,000 × g for 15 min. The cytochrome c reduction levels were measured as the absorbance value of the obtained supernatant at 550 nm minus the absorbance value at 535 nm then divided by total protein concentration (ΔA 550−535 /mg protein). The biuret method was employed to measure the nal protein concentration and the bovine serum albumin (Sigma Chemical Co., St. Louis, MO) was used as a standard.
Determination of caspase activity. A caspase-3 activity assay kit (Sigma-Aldrich, USA) was employed to determine its activity. In short, samples (1.0 g) were lysed in a precooled lysis solution (1 mmol/L dithiothreitol, 100 mmol/L Hepes, 0.1% NP-40 and 10% sucrose, pH 7.5) Samples were then smashed by employing a glass pestle and mortar 30 times under ice, and then the mixtures were centrifuged at 12,000 × g for 10 min. Subsequently, 45 µL of 10 mmol/L dithiothreitols and 5 µL of 4 mmol/L DEVD-pNA substrate were added into 20 µL of sample aliquots. Then the mixed solution was incubated at 37°C for 1 h and a uorescence spectrophotometer (Shimadzu RF 5301, Kyoto, Japan) was used to measure the uorescent intensity of the sample (Ex=360 nm and Em=460 nm).
Determination of apoptotic nucleus. TdT-mediated dUTP nick-end labeling (TUNEL) assay was used to detect the apoptotic nucleus. Muscle samples were cut into a block of 1.0 cm×0.5 cm×0.5 cm and xed in a 4% paraformaldehyde solution for more than 48 h. Samples were para n-embedded and cut into ten µm-thickness tissue sections and transferred onto slides. After dewaxing and rehydration, the sections were immersed with formaldehyde solution containing 3% H 2 O 2 to remove endogenous peroxidase. The sections were permeabilized by immersing them in a solution containing 0.1% Na 2 NO 3 and 0.1% Triton X-100. The permeabilized tissue sections were rinsed and washed with buffer (137 mmol/L NaCl, 4.3 mmol/L disodium phosphate, 2.7 mmol/L KCl, 1.4 mmol/L potassium dihydrogen phosphate, pH 7.4). After washing, the sections were blocked by 5% goat serum for 30 min and then washed with the above buffer for 30 min. The labeling enzyme and the solution were prepared in a ratio of 1:9 to achieve the TUNEL reaction solution. The tissue sections were then covered with the solution prepared and incubated at 37°C for 60 min in a light protecting chamber. As for the negative controls, we used the lack of labeling enzyme during the TUNEL solution preparation. Positive controls were incubated with 5.0 U/mL DNase I for 10 min at room temperature before adding the TUNEL reaction solution. DAPI was added for nuclei staining, and the tissue sections were applied with glycerol and placed under a uorescent microscope (CKX31, Olympus, Japan) for observation. Regular cell nuclei were stained in blue uorescence by DAPI, and apoptotic cell nuclei were stained in green uorescence by TUNEL. At least three elds of view per section were chosen and quanti ed for each sample (400×), and the apoptotic rate was calculated as the ratio of apoptotic nuclei to regular nuclei.
Determination of the myo bril fragmentation index (MFI). MFI was determined by the method as reported previously 2 . In short, 5.0 g of meat was added with 20 mL of MFI buffer (100 mmol/L KCl, 20 mmol/L K 3 PO 4 , 1 mmol/L EDTA, 1 mmol/L MgCl 2 , 1 mmol/L NaN 3 , pH 7.1) and homogenized on ice. The homogenate was then centrifuged at 4°C at 1000×g for 15 min. After that, the supernatant was discarded. Next, 20 mL of MFI buffer was used to re-suspend the precipitate and the samples were homogenized again. The mixture was then centrifuged to discard the supernatant. Afterward, the precipitate was dissolved with 10 mL of MFI buffer, and was ltered through a 200-mesh nylon screen. Another 5 mL of MFI buffer was added to help the myo brillar protein pass through the lter. Filter myo brillar protein was quanti ed via the biuret method to determine its protein concentration, and its nal concentration was achieved at 0.5 mg/mL via additional MFI buffer. A UV spectrophotometer was used to measure the absorbance at 540 nm. The nal value was obtained by multiplying the reading 200 times to represent the MFI.
Determination of the shear force. The shear force of bovine muscle was measured according to a previously described method 14 . The samples were cooked in a bag and submerged in water at 80°C until the central temperature reached above 70°C for 30 min. Subsequently, the meat samples were cooled to room temperature, and three to ve meat columns (Φ 1.27) were removed along the direction of the muscle bers as representative samplers. Each sample was then cut by a blade perpendicular to the long axis of the muscle bers and the shear force was measured by using a tenderness meter (C-LM4, Harbin, China).
Statistical analysis. The data were analyzed using IBM SPSS software 21.0 (SPSS, Inc., Chicago, IL, USA). The results were presented as mean ± standard error (SE) from six biological replicates. The one-way analysis of variance (ANOVA) procedure followed by the Duncan's multiple range tests was used to separate the means. If P<0.05, the difference was signi cant. The student's t-test was employed to compare the differences in the means between the two groups (*P<0.05, **P<0.01). The results were presented as mean ± standard error (SE).

Conclusion
The present study investigated the effect of postmortem pH decline rates on mitochondrial apoptosis and tenderness in bovine muscle. The results demonstrate that the fast pH decline promoted sarcoplasmic protein denaturation and ROS accumulation, which further enhanced mitochondrial dysfunction and cytochrome c oxidation, thereby improving postmortem bovine skeletal muscle tenderness via caspase-3 activation. Therefore, the fast pH decline can increases bovine muscle tenderization by inducing mitochondrial apoptosis, and postmortem pH decline rates are important in ultimate muscle tenderness development throughout aging. This new observation will provide scienti c insight about the involvement of postmortem pH changes in apoptosis and beef tenderization.

Declarations Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Figure 1 The pH in bovine muscles of different glycolytic rates at 0.5, 2, 6, 12, 24, 48, and 72 h postmortem. A-D: Each letter denotes a statistically signi cant difference in the fast group at different postmortem times (P<0.05). a-e: Each letter denotes a statistically signi cant difference in the slow group at different postmortem times (P<0.05). ** and * indicate the signi cance compared the same time points between two groups (**P<0.01, *P<0.05). Bars represent the standard error of mean.   The caspase-3 activity in bovine muscles of different glycolytic rates at 0.5, 6, 12, 24, 72, 120, and 168 h postmortem. A-D: Each letter denotes a statistically signi cant difference in the fast group at different postmortem times (P<0.05). a-e: Each letter denotes a statistically signi cant difference in the slow group at different postmortem times (P<0.05). * indicates the signi cance compared the same time points between two groups (P<0.05). Bars represent the standard error of mean.