Enhancing the polyphenolic content and demand of Upper Midwestern, locally produced beef

Please note that all materials and methods were described in the published manuscript Krusinski et al. (2023) https://www.mdpi.com/2304-8158/12/19/3547 

Experimental design and diet composition:

• This two-year study (2019 and 2020) was conducted at the Upper Peninsula Research and Extension Center (UPREC) (Michigan State University) located in Chatham, MI (latitude: 46°20’ N, longitude: 86°55’ W; elevation: 271 m).
• Three treatments were tested: 1) a complex pasture mixture (GRASS), a total mixed feedlot ration (GRAIN), and a total mixed feedlot ration supplemented with 5% (dry matter - DM) GSE for the last 30 days of finishing (GRAPE). The experimental design for the GRASS and GRAIN groups was previously described (Krusinski et al. 2022).
• Forty (n = 40) and 32 Red Angus steers (n = 32) (14-20 months of age) were randomly allocated to the diets in 2019 and 2020, respectively. Thirty steers in 2019 and 22 steers in 2020 were randomly stratified and allocated to one of three pens for each of the GRASS and GRAIN treatments (for each diet and each year, n = 5 animals per pen, three pens). This design was followed in 2019, but due to lower male births in 2020, only two groups (one group with three steers and another group with four steers) were formed for the GRAIN diet. For GRAPE, ten Red Angus steers each year were kept in one same feedlot pen.
• For this project, for each year, nine samples per diet were randomly chosen. For GRASS and GRAIN in 2019, three samples per pen per year were randomly picked. The same design was followed for GRASS in 2020. Since only two pens for GRAIN were formed in 2020 due to lower male births, all seven samples were picked with an additional two random samples from 2019. Since GRAPE samples all came from the same pen, nine samples for each year were randomly chosen. The total number of beef samples analyzed for this study was 54 (n = 54).
• Plant species found in GRASS were meadow fescue (Schedonorus pratensis (Huds.) P. Beauv.), red clover (Trifolium pratense L.), timothy grass (Phleum pratense), alfalfa (Medicago sativa), white clover (Trifolium repens L.), birdsfoot trefoil (Lotus corniculatus), chicory (Cichorium intybus), orchardgrass (Dactylis glomerata L.), and dandelion (Taraxacum oficinale L.).
• The GRAIN diet was composed of orchardgrass hay, dry corn, high moisture corn, and pellets (36% crude protein).
• For GRAPE, 5% of GSE (DM basis) was added to GRAIN during the last 30 days of the finishing period. GSE was obtained from Pioneer Enterprises (Lewiston, ID, USA).

Sample collection:

• Feed samples were collected every two weeks (GRAPE samples were collected at the same frequency as GRAIN samples).
• For fatty acid analysis, the number of feed samples was 21 GRASS, 15 GRAIN, and 7 GRAPE in 2019 and 24 GRASS, 10 GRAIN, and 5 GRAPE in 2020. Since these feed samples were also analyzed as part of other projects, less samples were left for phytochemical analysis: 11 GRAIN samples and 4 GRAPE samples were analyzed. Unfortunately, no GRASS samples were left for phytochemical analysis. GRASS samples underwent multiple runs as part of a series of previous experiments, leaving little to no samples for phytochemical analysis.    
• For beef samples collection, each year animals were harvested on the same day (late September in 2019 and early October in 2020) in a United States Department of Agriculture (USDA) regulated facility (16-18 months old for GRAIN and GRAPE and 24-26 months old for GRASS), and ribeye samples were collected between the 11^th and 13^th rib on the left side of the carcass.
• Beef samples were then further processed at the Michigan State University Meat Laboratory; steaks were cut in 1 ´ 1 cm cubes and flash frozen with liquid nitrogen. Beef samples were stored at -80 °C until further analysis.

Fatty acids: 

Fatty acids in beef and feed samples were analyzed according to Krusinski et al. (2022) https://doi.org/10.3390/foods11213451

• Lyophilized and ground feed samples and minced beef samples were mixed with 8 mL 4:1 ethyl acetate:methanol (v/v) and underwent extraction in a CEM Mars 6 microwave digestion system equipped with a 24-vessel rotor and GlassChem tubes (CEM Corp., Matthews, NC, USA).
• Samples were then filtered using Whatman filter paper grade 597 into another set of test tubes containing 3.5 mL HPLC water and centrifuged to separate the top organic layer.
• The top layer was transferred to another set of tubes and evaporated under nitrogen gas. Samples were reconstituted with 4:1 dichloromethane:methanol (v/v) to reach a concentration of 20 mg of oil/mL.
• For the creation of FA methyl esters (FAMEs), 100 mL containing 2 mg of oil was resuspended in toluene with 20 mg of internal standard (methyl 12-tridecenoate, U-35M, Nu-Chek Prep, Elysian, MN, USA).
• After resuspending, 2 mL of 0.5 N anhydrous potassium methoxide was added before being heated at 50 °C for 10 min.
• Then, 3 mL of methanolic HCl (5%) was added before heating the samples at 80 °C for 10 min.
• After cooling down, 2 mL of HPLC water and 2 mL of hexane were added, and samples were centrifuged at 2,500 RPM for 5 min.
• The top layer was removed and dried under nitrogen gas.
• FAMEs were resuspended in 1 mL of isooctane (final concentration was 2 mg/mL).
• Samples were then transferred to gas chromatography-mass spectrometry (GC-MS) vials with glass inserts.
• For FAMEs quantification, a PerkinElmer (Waltham, MA, USA) 680/600S GC-MS in electron impact mode (70 eV) was used. The GC-MS was equipped with a HP-88 column (100 m, 0.25 mm ID, 0.2 mM film thickness) from Agilent Technologies (Santa Clara, CA, USA).
• One mL of feed samples was injected (GC temperature: 250 °C).
• One mL of beef samples was injected twice (20:1 split) (GC temperatures: 175 °C and 150 °C). For detailed temperature settings, please see Krusinski et al. (2022). A third injection in splitless mode followed for beef samples.
• The carrier gas was helium (1 mL/min), and MS data were recorded in full scan mode (m/z 70-400 amu).
• FAMEs identification was achieved using MassLynx V4.1 SCN 714 (Water Corp., Milford, MA, USA). Retention time and EI mass fragmentation were compared to a reference standard containing Supelco 37 Component FAME Mix with mead acid, docosatetraenoic acid, n-3 DPA, n-6 DPA, and palmitelaidic acid (Cayman Chemical, Ann Arbor, MI, USA). The CLA standard UC-59M from Nu-Chek Prep (Elysian, MN, USA) was used for the identification of CLA isomers.
• For FAs not included in the standards, elution order and EI mass fragmentation were used for identification. FAMEs were quantified by utilizing a standard curve containing reference and internal standards.

Vitamin E and mineral analysis of beef samples: 

• For vitamin E, 1 g of beef was thoroughly mixed with 5 mL of water.
• Proteins were precipitated with ethanol, and fat-soluble vitamins were extracted using hexane.
• Samples were centrifuged and the top layer was evaporated.
• Evaporated samples were resuspended in the chromatographic mobile phase and placed in vials.
• A six-point curve was made by serial diluting a vitamin E solution (Sigma-Aldrich, St. Louis, MO, USA) in ethanol (50 mg/mL to 0.2 mg/mL).
• A Waters Acquity system equipped with a Symmetry C18, 1.7 mm, 2.1 x 50 mm analytical column and Water Empower Pro Chromatography Manager software (Water Corp., Milford, MA, USA) were used for chromatography analysis.
• The mobile phase was acetonitrile:methylene chloride:methanol (70:20:10, v/v/v) and the flow rate was 0.5 mL/min. Detection was performed at UV absorption 292 nm.
• For mineral analyss, the Agilent 7900 Inductivity Coupled Plasma-Mass Spectrometer (ICP-MS) (Agilent Technologies, Inc., Santa Clara, CA, USA) was used, with a six-point calibration curve and bovine liver and mussel standards (National Institute of Standards and Technology, Gaithersburg, MD, USA).

Polyphenolic profiling: 

• Feed samples underwent freeze-drying for 18.5 h using a Harvest Right Home Freeze Dryer (Harvest Right, North Salt Lake, UT, USA) and were subsequently ground through a 1 mm screen in a Wiley mill (Arthur H. Thomas, Philadelphia, PA, USA) with dry ice.
• Beef samples were minced and pulverized on dry ice using a mortar and pestle (see picture).
• UHP-LC-MS-grade acetonitrile, methanol, DMSO, formic acid and water (Supelco LiChrosolv®) were ordered from Sigma-Aldrich (St. Louis, MO, USA).
• A QReSS™ internal standards kit, containing a mixture of isotopically-labeled metabolites, was purchased from Cambridge Isotope Laboratories (Tewksbury, MA, USA).
• Purified external standards of compounds were obtained from Sigma-Aldrich (St. Louis, MO, USA) and/or Cayman Chemical (Ann Arbor, MI, USA).
• Briefly, 200 mg of pulverized beef and 50 mg of feed samples were mixed with 1000 mL and 500 mL, respectively of methanol.
• At this time, 10 mL of QReSS™ internal standard (Cambridge Isotope Laboratories, Inc., Tewksbury, MA, USA) were added to the samples.
• Proteins were then precipitated under vigorous shaking for 10 min at 20 Hz using a QIAGEN TissueLyser II operated with two 5 mm glass beads (QIAGEN Sciences, Germantown, MD, USA).
• Samples were then centrifuged at 23,000 x g for 10 min at 4 °C after a 1 h protein freeze-out step at -20 °C.
• Supernatants were removed and transferred to a new set of tubes.
• Beef samples were diluted with 2 mL of water with 1% formic acid (v/v), while feed samples were diluted with 1 mL of the same mixture.
• Strata C18-E cartridges (Phenomenex, Torrance, CA, USA) were used for solid phase extraction (SPE).
• Cartridges were activated with 1 mL of methanol with 1% formic acid and washed with 1 mL of water with 1% formic acid.
• Samples were then loaded onto the cartridges. After passing the samples through the cartridges, beef samples were washed with 2 mL of water with 1% formic acid while feed samples were washed with 1.2 mL of water with 1% formic acid.
• Beef and feed samples were eluted with 1.2 mL of methanol in 0.1% formic acid.
• Beef and feed samples were then evaporated under a gentle stream of nitrogen gas before being reconstituted with 100 mL and 200 mL, respectively of methanol in 0.1% formic acid in 1.5 mL LC-MS amber vials (Agilent, Santa Clara, CA, USA) with 250 mL glass inserts.
• Compounds were detected concurrently as precursor ion/product ion pair through multiple reaction monitoring (MRM) using ultra-high performance liquid chromatography tandem mass spectrometry (UHPLC-MS/MS).
• The platform utilized a SCIEX Hybrid Triple Quad™ 7500 (Framingham, MA, USA) with a front-end Shimadzu Nexera LC-40 Series (Kyoto, Japan) liquid chromatography system.
• Sample extracts were held at 10 °C in an auto-sampler and compounds were separated at 30 °C using a reverse phase Kinetex F5 100Å column (2.1 mm x 150 mm, 2.6 μM) from Phenomenex (Torrance, CA, USA) with binary mobile phases of water (A) and acetonitrile (B), both containing 0.1% formic acid (v/v).
• Samples were run in both negative and positive electrospray ionization mode. The following source parameters were used in negative mode: 1,600 V for the ionspray voltage, 550 °C for the temperature, 40 psi for the curtain gas, 40 psi for the nebulizer gas (GS1), 60 psi for the heating gas (GS2).
• In the negative mode, the linear gradient consisted of an initial composition of 5% B for 2.1 min with a flow rate of 0.2 mL/min, which was ramped up gradually to 95% B and a maximum flow rate of 0.46 mL/min over 14 min to keep a constant pressure, prior to being switched to 5% B for the final 4 minutes with a minimum flow rate of 0.175 mL/min.
• The following source parameters were used in positive mode: 2,000 V for the ionspray voltage, 550 °C for the temperature, 40 psi for the curtain gas, 40 psi for the nebulizer gas (GS1), 60 psi for the heating gas (GS2).
• In the positive mode, the linear gradient consisted of an initial composition of 5% B for 2.1 min with a flow rate of 0.2 mL/min, which was ramped up gradually to 95% B and a maximum flow rate of 0.46 mL/min over 14 min to keep a constant pressure, prior to being switched to 5% B for the final 4 minutes with a minimum flow rate of 0.175 mL/min.
• For both modes, a pooled matrix sample (sample generated by taking a small volume from samples from different experimental conditions), a double blank (100% methanol), and mixture of purified standards of target compounds was injected using an unscheduled method to determine presence of compounds in the matrix sample and their retention times for the scheduled method.
• In both modes, the cycling time in the scheduled method was set to 1000 msec and the dwell time ranged from 3 to 250 msec depending on the number of MRMs triggered.
• Double blank (100% methanol) and blank internal standard samples (methanol spiked with QReSS™ isotopically-labeled internal standards) were ran every 15 samples for quality control purposes.
• Analyst 3.1 software (AB Sciex, Framingham, MA, USA) was used to acquire and analyze the chromatographic data.
• Peaks were integrated using area-under-the-curve and normalization was performed using QReSS™ isotopically-labeled internal standards to account for any loss of material during sample preparation.
• Unlabeled external standard mixes were run in parallel to the samples with known concentrations of the different metabolites to allow for quantitation (in mg/100 g) of various compounds with relevant nutritive value and for which a standard was available.
• For compounds with no relevant nutritive value or for which no standard was ran concurrently, the data is expressed as arbitrary units (AU).

Statistical analysis:

• RStudio (R Core Team, Vienna, Austria) was used for the statistical analysis.
• A linear regression model was used to test the effect of diet on fatty acids, micronutrients, and quantified phytochemicals in beef.
• Diet, year, and pen were considered fixed effects.
• The experimental unit was each animal.
• Tukey’s HSD was used for post hoc comparison.
• For feed samples, a t-test was performed to compare GRAIN vs. GRAPE samples.
• For all analyses, values that fell below the limit of detection were treated as zeroes. Results were considered significant at p < 0.05 and were reported as mean ± standard error from the mean (SEM).
• Principal Component Analysis (PCA) and Random Forest (RF) analysis were conducted using MetaboAnalyst 5.0 (metaboanalyst.ca).
• Phytochemical compounds were first normalized to mass and then log transformed. The goal was to visualize data sets and identify the top metabolites that discriminate between groups using mean decrease accuracy.