The LL and PM muscles from the left sides of 7 (n = 7) beef carcasses (USDA Choice; A maturity) were obtained from a local packing plant within 48 h of harvest. Seven 1.92-cm-thick steaks were fabricated from the anterior halves of each muscle. From each muscle, 3 steaks were allotted for surface color evaluation (on 0, 3, and 6 d), the fourth steak was used for pH measurements, and the remaining 3 steaks were used to create myoglobin standards for myoglobin redox calculations. Following surface color measurements, steaks designated to 0, 3, and 6 d were cut in half perpendicular to the oxygen-exposed surface, but in random axis. These halves were assigned randomly for biochemical and proteome analyses; while one half was used for mitochondrial proteome profiling (on 3 and 6 d), the other half was used for OC and MRA measurements.

The surface color was measured on 0, 3, and 6 d. Steaks on 0-d display were placed on Styrofoam trays and overwrapped in polyvinylchloride film (15,500–16,275 cm³ O₂/m²/24 h at 23°C; E-Z Wrap Crystal Clear Polyvinyl Chloride Wrapping Film, Koch Supplies, Kansas City, MO) and bloomed at 4°C for 2 h before the color measurement. The packages were displayed in a coffin-type retail case at 4°C (1,612–2,152 lx; Philips Delux Warm White Fluorescent lamps, Andover, MA; color rendering index = 86; color temperature = 3,000 K) and were rotated every 24 h to minimize temperature variations within the case. The polyvinylchloride film was removed, and L* (lightness), a* (redness), and b* (yellowness) values (CIE, 1976) were recorded at 3 random locations on the light-exposed steak surfaces using a HunterLab MiniScan XE Plus (HunterLab Associates, Reston, VA) with 2.54-cm diameter aperture, illuminant A, and 10° observer angle (AMSA, 2012). The colorimeter was calibrated with standard black and white plates each day before the readings were taken. Additionally, the CIE (1976) a* and b* values were used to calculate hue angle and chroma (AMSA, 2012).

Samples from steaks assigned to 0 d of display from both LL and PM, visually devoid of fat and connective tissue, were blended in an Omni tabletop mixer (Sorvall, Newton, CT). To determine pH, 10 g of pulverized steak that contained surface and interior was combined with 100 mL of deionized water and mixed for 30 s. The pH values were obtained by using an Accumet combination glass electrode connected to an Accumet 50 pH meter (Fisher Scientific, Fairlawn, NJ). Before measurements, the pH meter was standardized with pH 4 and 7 buffers.

The steak half assigned to MRA and OC was then bisected parallel to the oxygenated surface to expose the interior of the steak (resulting in 2 interior pieces). The first piece was used to measure MRA, and the second piece was used to measure muscle OC. The muscle OC was determined indirectly using the changes in oxymyoglobin level. A modified procedure of Madhavi and Carpenter (1993) was used to determine muscle OC on day 0, 3, and 6. Freshly cut steak interiors were allowed to oxygenate/bloom for 60 min at 4°C, vacuum packaged, and were scanned using a Hunter Lab MiniScan XE Plus spectrophotometer for reflectance from 400 nm to 700 nm to determine the initial percentage of oxymyoglobin (AMSA, 2012). OC (measured by conversion of oxymyoglobin to deoxymyoglobin) was induced by incubating samples at 30°C for 30 min. After incubation, the samples were rescanned immediately to determine remaining surface oxymyoglobin (AMSA, 2012). The reflectance (R) at 474, 525, 572, and 610 nm was converted to K/S values using K/S = (1 − R)²/2R, in which K and S indicate absorption and scattering coefficient, respectively. These K/S values were then utilized in the following equations to calculate the percentage of oxymyoglobin and metmyoglobin (AMSA, 2012). % oxymyoglobin =[K/S610÷K/S525 for 100% metmyoglobin−K/S610 ÷K/S525 for sample][K/S610÷K/S525 for 100% metmyoglobin−K/S610÷K/S525 for 100% oxymyoglobin]% metmyoglobin=[K/S572÷K/S525 for 100% oxymyoglobin−K/S572÷K/S525 for sample][K/S572÷K/S525 for 100% oxymyoglobin−K/S572÷K/S525 for 100% metmyoglobin]

To calculate muscle OC, the following equation was used: OC=%pre-incubation surface oxymyoglobin−%post-incubation surface oxymyoglobin

MRA was determined according to the protocol of Sammel et al. (2002). Samples (1.5 × 2.5 × 2.5 cm) were submerged in a 0.3% solution of sodium nitrite (Sigma, St. Louis, MO) for 20 min to facilitate metmyoglobin formation. The samples were then removed, blotted dry, vacuum packaged (Prime Source Vacuum Pouches, 4 mil, Koch Supplies Inc., Kansas City, MO), and scanned for reflectance spectra from 400 nm to 700 nm with a HunterLab MiniScan XE Plus spectrophotometer to determine pre-incubation metmyoglobin content (AMSA, 2012). The vacuum-packaged samples were incubated at 30°C for 2 h to induce metmyoglobin reduction. After incubation, the reflectance data were collected again to determine the post-incubation metmyoglobin content. The MRA was calculated using the following equation: MRA=100×[(% pre-incubation surface metmyoglobin−% post-incubation surface metmyoglobin)÷% pre-incubation surface metmyoglobin]

A split-plot experimental design was used to analyze the instrumental color, OC, and MRA data. The whole plot consisted of a randomized complete block design, in which carcass served as a block and muscles within a carcass were experimental units (LL or PM). In the subplot, steaks within a whole muscle served as experimental units assigned to day of display (0, 3, or 6). The random terms included carcass (Error A) and unspecified residual error (Error B). The data were analyzed using the Mixed procedure of SAS version 9.4 (SAS Institute Inc., Cary, NC), and the differences among means were detected using the least significant difference test at the 5% level.

Mitochondria were isolated from the beef muscles according to Lanari and Cassens (1991). On 3 and 6 d of display, LL or PM muscle tissue (35 g) was minced finely and suspended in 70 mL of mitochondrial suspension buffer (125 mM sucrose, 5 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid). The suspension was hydrolyzed for 15 min with proteinase K (protease/tissue, 1.0 mg/g for LL or 0.5 mg/g for PM), and the pH was maintained between 7.0 and 7.2. The suspension was diluted with mitochondrial isolation buffer (125 mM sucrose, 125 mM Tris-HCl, 10 mM KCl, 25 mM ethylenediaminetetraacetic acid, and 0.2% bovine serum albumin [pH 7.2]) to 350 mL and homogenized using a Kontes Duall grinder (Vineland, NJ), passed through a Wheaton Potter-Elvehjem grinder (Millville, NJ), and centrifuged for 20 min at 900 × g in a Sorvall refrigerated RC-5B centrifuge (Thermo Fisher Scientific, Waltham, MA). The resulting supernatant was passed through double-layered cheesecloth and centrifuged for 15 min at 14,000 × g. The pellet was washed twice and suspended in mitochondrial suspension buffer at pH 7.2. All steps were performed at 0°C to 4°C. Mitochondrial protein content was determined using a bicinchoninic acid protein assay, and the total mitochondrial yield was reported as milligrams per gram of tissue (Grubbs et al., 2013; Ke et al., 2017).

The aliquots of suspended mitochondria were centrifuged to pellet the mitochondria. The supernatant was discarded, and the crude mitochondrial pellet with 200 μL of the extraction buffer (8.3 M urea, 2 M thiourea, 2% 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate, and 1% dithiothreitol [DTT] [pH 8.5]) was shaken vigorously for 30 min at room temperature. Samples were centrifuged at 10,000 × g for 30 min. The supernatant was filtered, and the protein concentration was determined using a 2D Quant assay (GE Healthcare Life Sciences, Piscataway, NJ).

The mitochondrial proteome (900 μg) was mixed with DeStreak Rehydration Solution (GE Healthcare Life Sciences, Piscataway, NJ) with 2 mM DTT and 0.5% immobilized pH gradient (IPG) buffer (pH 3–10; GE Healthcare Life Sciences, Piscataway, NJ) to a total volume of 300 μL. The mix was loaded onto IPG strips (pH 3–10; 17 cm) and was subjected to passive rehydration for 16 h. First-dimension isoelectric focusing of the IPG strips was done in a Protean IEF cell system (Bio-Rad, Hercules, CA) by applying a linear increase in voltage initially and a final rapid voltage ramping to attain a total of 80 kVh. Subsequently, the IPG strips were equilibrated with equilibration buffer I (6 M urea, 0.375 M Tris-HCl [pH 8.8], 2% sodium dodecyl sulfate [SDS], 20% glycerol, 2% w/v DTT; Bio-Rad) followed by equilibration buffer II (6 M urea, 0.375 M Tris-HCl [pH 8.8], 2% SDS, 20% glycerol, 2.5% w/v iodoacetamide; Bio-Rad) for 15 min each. Second-dimension separation of proteins was done by 13.5% SDS polyacrylamide gel electrophoresis (38.5:1 ratio of acrylamide to bis-acrylamide; SDS-polyacrylamide gel electrophoresis) using a Protean XL system (Bio-Rad, Hercules, CA). The gels were stained with colloidal Coomassie Blue for 48 h and destained in deionized distilled water for 48 h or until sufficient background was cleared.

The digital images of the destained gels were captured using Versa Doc imager (Bio-Rad, Hercules, CA) and were analyzed using PDQuest software (Bio-Rad, Hercules, CA). Images were first subjected to automatic spot detection and matching optimized by the aid of landmark protein spots. The relative volume of each spot was normalized as a percentage of the total volume of all the spots detected on the gel (Joseph et al., 2012). For each spot in a given sample, spot quantity values in duplicate gels were averaged for statistical analysis (Nair et al., 2018a; Zhai et al., 2018). A spot was considered to be differentially abundant when it demonstrated 1.5-fold intensity difference between the treatments associated with P < 0.05 in a pairwise Student t test.

Protein spots differentially abundant between the treatments were excised from the gel using pipet tips. The spots were destained by two 30-min washes with 50 mM NH₄HCO₃/50% CH₃CN followed by vortexing for 10 min and drying using a vacuum centrifuge. Proteins in the gel were reduced by the addition of 50 mM NH₄HCO₃ containing 10 mM DTT and incubation at 57°C for 30 min. Further, the proteins were alkylated by the addition of 50 mM NH₄HCO₃ containing 50 mM iodoacetamide and incubation for 30 min in the dark at room temperature. The gel pieces were washed twice with 50 mM NH₄HCO₃ and once with CH₃CN and partially dried in a vacuum centrifuge. Dried gel pieces were rehydrated for 1 h (on ice) with a solution of 40 mM NH₄HCO₃/9% CH₃CN, containing proteomic grade trypsin (Sigma, St. Louis, MO) at a concentration of 20 ng/µL. An additional volume of 40 mM NH₄HCO₃/9% CH₃CN was added to cover the sample, and the gel was incubated for 18 h at 37°C. Peptide extraction from the gel was done by sonication in 0.1% trifluoroacetic acid for 10 min followed by vortexing for 10 min. Extraction was repeated with 50% CH₃CN/0.1% trifluoroacetic acid. The extracts were combined, and the volume was decreased to eliminate most of the acetonitrile. The peptide extracts were desalted and concentrated by solid phase extraction using a 0.1–10 µL pipet tip (Sarstedt, Newton, NC) packed with 1 mm of Empore C-18 (3M, St. Paul, MN), and peptides were eluted in 5 µL of 50% CH₃CN/0.1% trifluoroacetic acid. Desalted peptide extracts (0.3 µL) were spotted onto an Opti-TOF 384 well insert (Applied Biosystems, Foster City, CA) with 0.3 µL of 5 mg/mL α-cyano-4-hydroxycinnamic acid (Aldrich, St. Louis, MO) in 50% CH₃CN/50% 0.1% trifluoroacetic acid. Crystallized samples were washed with cold 0.1% trifluoroacetic acid and were analyzed using a 4800 MALDI TOF-TOF Proteomics Analyzer (Applied Biosystems). The initial matrix-assisted laser desorption ionization mass spectrometry (MS) spectrum was acquired for each spot with 400 laser shots per spectrum. From that, a maximum of 15 peaks with a signal-to-noise ratio of >20 were automatically selected for tandem mass spectrometry (MS-MS) analysis (1,000 shots per spectrum) by post-source decay. Peak lists from the MS-MS spectra were submitted for database similarity search using Protein Pilot version 2.0 (Applied Biosystems), and the search was performed in the Swiss Prot and National Center for Biotechnology Information database to identify the proteins using search parameters (search type: identification; enzyme: trypsin; database: bovine NCBInr; search effort: thorough; unused cutoff > 1.30, 95% confidence).

The differentially abundant proteins were matched against the STRING database (Szklarczyk et al., 2015) to determine the protein–protein interaction networks (on the influence of muscle source and retail display), in which the network nodes represented the proteins and the lines indicated functional associations.

The PM steaks had greater pH than LL steaks on day 0 of display (LL = 5.56, PM = 5.63; P < 0.05; standard error of the mean = 0.02). There was a muscle × display day interaction (P < 0.05) for instrumental color parameters (Table 1). The LL steaks had greater (P < 0.05) L*, a*, b*, and chroma values than the PM counterparts throughout display. Although the a*, b*, and chroma values decreased (P < 0.05) between 0 and 6 d of retail display in both muscles, the decline was less for LL steaks compared with PM ones. The L* value increased (P < 0.05) for both muscles from day 0 to day 3 and then decreased on day 6. On day 3 and 6, PM steaks exhibited greater (P < 0.05) hue angle than LL steaks. Whereas hue increased (P < 0.05) in PM steaks during display, it did not change (P > 0.05) in LL steaks.

Muscle source and display day influenced (P < 0.05; Table 2) MRA and OC. The LL steaks demonstrated greater (P < 0.05) MRA and lower (P < 0.05) OC than their PM counterparts during display. Both MRA and OC decreased (P < 0.05) in LL and PM steaks between day 0 and day 6. While MRA remained stable (P > 0.05) until day 3, it declined (P < 0.05) in both muscles between day 3 and day 6. On the other hand, OC decreased (P < 0.05) on day 3 and remained stable (P > 0.05) until day 6 in both LL and PM steaks.

The results of instrumental color (Table 1) and biochemical attributes (Table 2) reiterated that the LL is a color-stable beef muscle, whereas the PM is a color-labile muscle. These findings are in agreement with several previous studies (Hunt and Hedrick, 1977; O’Keeffe and Hood, 1982; Madhavi and Carpenter, 1993; McKenna et al., 2005; Seyfert et al., 2006; Kim et al., 2009; Mancini et al., 2009; Joseph et al., 2012; Canto et al., 2016; Nair et al., 2018a), which reported greater color stability in beef LL than in PM.

Mitochondria could not be isolated from PM steaks on day 6 because of extensive degradation. Previous research also reported negligible PM mitochondrial yield with storage time (Mancini et al., 2018). Therefore, proteome analyses were accomplished only on mitochondria isolated from LL steaks on day 3 and day 6 as well as PM steaks on day 3. This allowed for evaluating the effect of muscle (LL vs. PM on day 3) and retail display (LL on day 3 vs. LL on day 6) on mitochondrial proteome profile.

Image analyses of gels identified 7 proteins differentially abundant in the mitochondrial proteome isolated from beef LL and PM steaks after 3-d refrigerated retail display (Figure 1). The identified proteins are listed in Table 3 along with accession number, ProtScore, and sequence coverage. While 2 proteins were more abundant in day-3 LL steaks, 5 were overabundant in PM steaks. The differentially abundant proteins were enzymes, binding proteins, and proteins involved in biosynthesis. The network of interacting proteins generated using the STRING database (Figure 2) identified 7 proteins as key nodes in biological interactions on the influence of muscle source. These results indicated the contribution of mitochondrial proteome in muscle-specific beef color stability.

Gel image analyses indicated that 11 protein spots were differentially abundant between the LL steaks from day 3 and day 6 (Figure 3). MS-MS identified 7 proteins from the 11 spots because 2 proteins were present in multiple spots. These 7 proteins were more abundant in the mitochondria isolated from beef LL steaks on day 3 of refrigerated display (LL3) compared with those from LL steaks on day 6 (LL6), indicating that the duration of display impacted mitochondrial functionality and biochemistry. The differentially abundant proteins were enzymes and binding proteins (Table 4). In the protein–protein network generated using the STRING database, 7 proteins appeared as key nodes in biological interactions on the impact of retail display in LL steaks (Figure 4).