Effects of dietary inclusion of yerba mate (Ilex paraguariensis) extract on lamb muscle metabolomics and physicochemical properties in meat

Abstract

This study aimed to evaluate the effect of dietary yerba mate (Ilex paraguariensis) extract (YME) on muscle metabolomics and physicochemical properties of lamb meat. Thirty-six uncastrated male lambs (90 d old) were fed experimental diets, which treatments consisted of 0%, 1%, 2%, and 4% inclusion of YME. Animals were fed for 50 d before slaughter. Muscle and meat samples were collected for metabolomics and meat quality analysis, respectively. The experiment was carried out in a randomized block design and analyzed using orthogonal contrasts. There was a quadratic effect of YME inclusion in tenderness (P < 0.05) and a positive linear effect on meat lightness (P < 0.05). No qualitative changes (P > 0.05) on individual metabolites were observed; however, changes in the quantitative metabolic profile were observed, showing that animals fed 1% and 2% of YME have a greater concentration of desirable endogenous muscle antioxidants, with direct impact on metabolic pathways related to beta-alanine metabolism and glutathione metabolism. Therefore, YME dietary supplementation up to 2% of the diet to lambs had little to no effects on the majority of meat quality traits evaluated; moreover, 4% of YME inclusion negatively affected feed intake and meat quality traits.

Introduction

Proteins and lipids that are the major components of meat are susceptible to oxidation leading to off-flavors and color changes causing consumer rejection of the product (Papuc et al., 2017). Additionally, the oxidation process may lead to loss in nutritional quality and generate compounds, such as metmyoglobin, lipid hydroperoxides, oxidized lipids, protein, and amino acids, with potential risk to consumers health (Papuc et al., 2017). Because of that, feed additives have been tested in animal diets as a source of biomolecules that delay the meat degradation process (Zawadzki et al., 2017; Ceribeli et al., 2018; Bellés et al., 2019).

It has been hypothesized that plant biomolecules, especially polyphenols from botanical extracts, such as flavonoids and phenolic acids, can improve antioxidant status in the muscle, by a direct effect, due to the intestinal permeability of some phenolic compounds (Scalbert et al., 2002). Furthermore, gastrointestinal tract microorganisms can metabolize plant polyphenolic compounds into bioactive components with more intestinal permeability (Déprez et al., 2000), which may be transported to the muscle and increase the antioxidant status of the tissue. Another benefit of plant bioactive molecules in ruminant nutrition is the modulatory effects on the rumen microbiome and consequently reduction of fatty acid biohydrogenation and increased availability of conjugated linoleic acid (CLA) and polyunsaturated fatty acids (PUFAs) for absorption in the small intestine (Vasta and Luciano, 2011; Kumar et al., 2015; Kalogianni et al., 2020). After absorption, PUFA and CLA can be incorporated into animals’ tissues, which may also benefit human health.

Yerba mate (Ilex paraguariensis) is a plant commonly used in human nutrition in Latin American Countries, whose leaves are traditionally used for tea making (Heck and de Mejia, 2007). The major components present in yerba mate tea according to Pomilio et al. (2002) and Zaporozhets et al. (2004) are polyphenols and purine bases, followed by alkaloids and amino acids. Berté et al. (2011) and Burris et al. (2012) reported that about 10% of dry tea or extract contains mainly caffeoyl derivatives and around 2% of caffeine. Those components are largely related to inhibition of the oxidation process and place yerba mate as an important source of natural antioxidants (Gullón et al., 2018; Kungel et al., 2018; Mateos et al., 2018).

The effect of dietary yerba mate on animal production is not well established; however, its effectiveness on increasing antioxidant activity of the products has been reported on dairy and beef cattle (Santos et al., 2017; de Zawadzki et al., 2017) and poultry (Ceribeli et al., 2018). Feeding yerba mate extract (YME) to dairy cows had no impact on milk production and yield of protein and fat; however, a greater oxidative stability of the milk lipids was observed (Santos et al., 2017). Ewes fed yerba mate during peripartum had an increase in milk fat, protein, and total solids yield (Po et al., 2012). Lambs fed up to 2% of YME had an increase in dry matter (DM) intake and body weight gain (Lobo et al., 2020); also, low thiobarbituric acid reactive substances and increase in glutathione peroxidase activity during meat aging improved the color stability of aged meat (Pena-Bermudez et al., 2020). Furthermore, YME increased tenderness of beefsteak and reduced the lipid oxidation in beef and poultry meat (de Zawadzki et al., 2017; Ceribeli et al., 2018).

Both reports tested dietary inclusion of YME and applied metabolomic approaches aiming to understand the metabolic pathway affected by the additive. In both cases, the antioxidant profile related to reduction of oxidative pathways was enhanced. To the best of our knowledge, there is no published study on the effect of YME on lamb’s diet that integrates muscle metabolome and meat quality; also, the metabolome analysis complements well the findings of Pena-Bermudez et al. (2020). Therefore, the objective of this study was to evaluate the effects of increasing levels of YME as a feed additive for lambs and its effect on metabolic profile of the muscle and instrumental meat quality traits. We hypothesized that dietary inclusion of YME may modulate the metabolic profile of the muscle and consequently the instrumental meat quality parameters (color, tenderness, weight losses, and composition) of lamb meat.

Materials and Methods

All the procedures using animals were approved by the Institutional Animal Care and Use Committee of the College of Animal Science and Food Engineering (Protocol number CEUA 3497040618) from the University of São Paulo, Brazil.

Location, animals, and experimental design

The animals were born and raised until weaning on a commercial farm and had access to pasture and feed supplementation on a creep-feeding system. After weaning, they were transported to the experimental unit of College of Animal Science and Food Engineering, University of São Paulo, where the experiment was carried out.

Thirty-six uncastrated male lambs from the same herd (crossbred [Texel × Santa Inês] × Dorper) weighing 23.77 ± 3.70 kg at 90 d of age were used in this trial. Upon arrival at the experimental facilities, all animals were dewormed with levamisole hydrochloride at 9.4 mg/kg of body weight (Ripercol, Zoetis) as an active ingredient, following the dose recommendations from the manufacturer. The animals were weighed and assigned into nine blocks. Each block had four animals with similar weight, and those four animals within each block were randomly assigned to one of the four experimental treatments (animal as experimental unit), according to a randomized block design arrangement. The animals had individual water and feeder access and were placed into an individual pen (width = 75 cm and length = 115 cm) with elevated plastic floor (50 cm from the ground) with holes to drain the urine and feces. Control diet (0% of YME as presented in Table 1) was fed for 3 d, and on the fourth day, the animals started to be fed with the assigned experimental diets for 50 d (until the time of slaughter).

Table 1.

Ingredients and composition of the experimental diets

	Yerba mate extract
	1%	2%	3%	4%
Ingredients
Corn silage, g/kg	400	400	400	400
Ground corn, g/kg	330	330	330	330
Soybean meal, g/kg	205	205	205	205
Salt, g/kg	1.5	1.5	1.5	1.5
Dicalcium phosphate, g/kg	1.5	1.5	1.5	1.5
Mineral1, g/kg	22	22	22	22
Yerba mate extract, g/kg	0	10	20	40
Kaolin, g/kg	40	30	20	0
Chemical composition
Dry matter, g/kg	688.9	688.3	687.7	686.4
Organic matter, g/kg	895.6	904.3	913.0	930.4
Crude protein, g/kg	214.6	215.6	216.6	218.7
Ether extract, g/kg	20.9	20.9	20.9	20.9
Non-nitrogen extract, g/kg	519.0	526.9	534.9	550.7
ADF2, g/kg	181.9	182.0	182.0	182.2
NDF3, g/kg	345.3	345.4	345.4	345.6
Lignin, g/kg	44.8	44.8	44.8	44.8
TDN4, g/kg	673.5	681.1	688.7	703.9
Gross energy, kcal/kg	4,188	4,229	4,271	4,354

¹Mineral composition per kilogram: calcium (maximum): 218 g; calcium (minimum): 190 g; cobalt (minimum): 148 mg; copper (minimum): 2,664 mg; sulfur (minimum): 64 g; fluoride (maximum): 1,600 mg; phosphorus (minimum): 160 g; iodine (minimum): 141 mg; manganese (minimum): 2,220 mg; selenium (minimum): 37 mg; and zinc (minimum): 7,992 mg.

²Acid detergent fiber.

³Neutral detergent fiber.

⁴Total digestible nutrients.

Diet and experimental treatment

Animals were fed ad libitum daily at 0800 and 1600 hours, and orts were weighed before morning feeding and allowed at least 10% of the offered, and diets were formulated to meet the nutritional requirements for finishing lambs (NRC, 2007). Basal diet was composed of 40% of corn silage and 60% of concentrate (on a DM basis) while the concentrate was a mixture of cornmeal, soybean meal, salt, dicalcium phosphate, mineral premix, and kaolin and/or YME. Kaolin (Al₂O₃(SiO₂)₂(H₂O)₂) was used as an inert ingredient to replace YME and maintain a constant nutrient concentration across experimental diets, as presented in Table 1.

Experimental treatments included one control diet (0% YME and 4% kaolin) and three diets with increasing concentrations of YME (1% YME and 3% kaolin, 2% YME and 2% kaolin, and 4% YME and 0% kaolin), all expressed on a percentage of DM basis. The doses of YME were selected based on the results from de Zawadzki et al. (2017), in which doses up to 1.5% had no negative effects on animal performance and greater impact on antioxidant profile were observed; however, greater doses were not tested. The ingredient and chemical composition of the diets are presented in Table 1.

The YME was provided by Centro Flora (Botucatu, SP, Brazil), from fresh leaves of I. paraguariensis by water:ethanol 75:25 v/v extraction at 90 °C, which were spray-dried until they reached 91.8% of DM. Total phenolic concentration was 21.7 g equivalents of gallic acid/100 g extract and contained 6% caffeine (w/w). Among the identified phenolic components, chlorogenic acid and 1,5-dicaffeoylquinic were the most abundant, representing 58.2% and 28.4% of the identified phenolic compounds, respectively. The YME used in this experiment was the same as the one used by de Zawadzki et al. (2017), who reported a complete characterization of the phenolic compounds using an ultra-performance liquid chromatography electrospray ionization mass spectrometry analysis. Bromatological characterization of YME was performed and YME had 10% ash (ASH), 10.4% crude protein (CP), 0.13% ether extract (EE), 0.7% neutral detergent fiber (NDF), 0.7% acid detergent fiber (ADF), 0.18% lignin (ADL), and 4.15 Mcal/kg of gross energy (GE). The YME was mixed with the concentrate, and at feeding time, the concentrate was mixed into the silage. The average daily feed intake was 1.12 kg as reported in our companion paper (Lobo et al., 2020).

Dietary ingredients were analyzed according to AOAC (1990) methods: DM (method 934.01), ASH (method 923.03), and EE (method 920.85), and nitrogen (method 920.87) was analyzed using the Kjeldahl method, and this value was multiplied by the factor 6.25 to obtain CP. NDF was analyzed according to Mertens (2002) using thermostable-α-amylase without sodium sulfate, and ADF and ADL (method 973.18) were measured according to Van Soest et al. (1991). GE was measured with a bomb calorimeter (C200 System, IKA, Staufen, Germany). Organic matter (OM) was calculated by subtracting the ash content from 100.

Slaughter and sample collection

The animals were raised for 50 d until they reached a final body weight of approximately 40 kg. Before slaughter, animals underwent 16 h of feed fasting. Slaughter was carried out by trained employees following standard procedures. Following captive bolt stunning and exsanguination, the skin, head, and hooves were removed from the carcasses, and eviscerations were made. A cubic muscle sample (3 × 3 × 3 cm) from each animal was collected from the right side of the carcass on the Longissimus lumborum (LL) region and subcutaneous fat of the sample was removed; a muscle subsample was collected and stored into screw cap tubes and snap-frozen in liquid nitrogen and then stored in a freezer (−80 °C) until metabolomics analysis. For the metabolomic analysis, muscle samples were collected after slaughter to represent the muscle biology of the live animal.

Carcasses were kept in cold storage (2 °C) for 24 h. One the day after slaughter, during the deboning process, the right side of the carcasses (the same side as muscle sampling) was used to collect chops samples from the LL region, and a total of 4 chops of 2.5 cm thickness were collected for instrumental meat quality analysis. Those samples were vacuum-packed and quickly stored in a freezer (−20 °C) for the instrumental meat quality analysis.

Instrumental meat quality analysis

The assessment of meat quality traits was done by following commercial standard procedures that involve freezing and thawing. Three frozen chops (as previously described) from each animal were thawed at 4 °C for 24 h. The first chop was unpacked and exposed for 30 min to oxygen. Subsequently, color (lightness—L*, redness—a*, and yellowness—b*) was measured in three different places from the same chop, according to the guidelines of the American Meat Science Association (AMSA, 2012), using a Minolta CR-400 colorimeter (Konica Minolta Sensing, Inc., Osaka, Japan) with diffuse illumination/0° angle of viewing, illuminant D65, and 8-mm opening. Other color parameters were calculated: ratio of a*:b* was calculated as (a*/b*), chroma was calculated as $\sqrt{a^2+b^2}$, and the hue angle was calculated as tan⁻¹ (b*/a*) × (180/π). The pH of thawed meat (same used for color analysis) was measured twice in different places of the same chop using a Testo pHmeter (205, Testo Inc., Sparta, NJ, USA) with a penetration electrode inserted into the chop sample.

Water holding capacity (WHC) was analyzed in duplicates. In accordance with Hamm (1960), 2 g of meat sample (same used for color and pH analysis) was placed in a folded paper filter (7 µm porosity), between two acrylic plates, subjected to 10 kg of pressure for 5 min, and weighed again. A second chop was used to measure cooking weight loss (CWL), and the entire chop of 2.5 cm of thickness was weighed and placed into a preheated grill (George Foreman—model GBZ4C, Orlando, USA) covered with aluminum foil. The internal temperature of the meat sample was measured by individual thermocouples (Flyever Ltda., São Carlos, SP, Brazil) placed into each sample until the internal temperature reached 71 °C and then removed from the grill. After that, the sample was left to cool and weighed again, according to de Mello et al. (2017). The results from WHC and CWL show the difference between the initial and final weights, which was reported as a percentage.

From the cooked chop in the CWL analysis, at least three cubic subsamples of each chop were collected and cut down at 110 mm of thickness. The cubic samples were placed with the fibers in a perpendicular way to the Warner-Bratzler shear device, connected to a TA-XT2i texture analyzer (Stable Micro Systems, LTD., Godalming, UK) to measure the shear force (SF) in Newton (N) in at least three cubic subsamples (triplicate) (de Mello et al., 2017).

On the third chop, subcutaneous fat was removed and the muscular material was lyophilized to obtain homogeneous and reduced-moisture samples. The proximate analysis was determined, according to AOAC (1990): DM (method 934.01), ASH (method 923.03), and total lipids (TL, method 920.85), and nitrogen (method 920.87) was analyzed using the Kjeldahl method, and this value was multiplied by the factor 6.25 to obtain protein. Moisture content (MC) of each sample was calculated subtracting DM from 100, as well as OM was calculated subtracting ASH from 100. The ratio between OM:protein was calculated as protein/OM, and the ratio between OM:TF was determined as TF/OM.

Metabolites extraction

Approximately, 0.3 g of lamb muscle from each of the 36 animals was homogenized with 1.2 mL of a cold solvent mixture (methanol:chloroform:water 2:2:1, v/v/v; Bligh and Dyer, 1959) using a commercial cell disruptor (FastPrep, MP Biomedicals) in homogenization tubes containing ceramic beads at a speed of 5.0 ms⁻¹ for 60 s. Homogenates were centrifuged for 10 min at 10,000 × g at 10 °C. The chloroform phase (lipid metabolites) was carefully separated from the hydroalcoholic phase (polar metabolites [PMs]) and reserved. The hydroalcoholic phase was mixed with 0.3 mL of chloroform, and the solution was agitated for 1 min (vortex). The mixture was centrifuged at the same previous condition, and the solvent phases were separated. Posteriorly, the chloroform phase was added together to the reserved phase mentioned above, and the hydroalcoholic phase (PMs) had the solvent evaporated in a vacuum centrifugal concentrator during 12 h (Speed-Vac, Thermo Savant). At the end of the centrifugation process, total PMs were quantified, and the ratio PM/OM was calculated.

Nuclear magnetic resonance analysis

The dried tissue extract containing the PMs was resuspended in 600 µL of deuterium oxide phosphate buffer—0.10 M, pD 7.7 (Glasoe and Long, 1960; Rubinson, 2017)—containing 0.050% w/w of sodium 3-trimethylsilyl-2,2,3,3-d4-propionate (TMSP-d₄, from Cambridge Isotopes, Leicestershire, UK), and the total volume was transferred to a 5-mm nuclear magnetic resonance (NMR) tube.

Metabolomics NMR analysis was conducted at 298 K on a 14-T Bruker Avance III spectrometer (Bruker BioSpin, Rheinstetten, Germany) equipped with a 5-mm PABBO probe head with gradients, automated tuning and matching accessory (ATMATM), BCU-I for the regulation of temperature, and a Sample-Xpress sample changer. The ¹H NMR spectra were acquired using 1D-NOESY-presaturation pulse sequence (Bruker 1D noesygppr1d pulse sequence) with irradiation at water frequency, 64 K data points, with a spectral width of 20 ppm, an acquisition time of 2.73 s, a recycle delay of 4 s, mixing time set to 10 ms, and no dummy scans, and the spectra are the accumulation of 256 scans. Free induction decays (FIDs) were multiplied with a 0.3-Hz exponential multiplication function before Fourier Transform was applied. Phase and baseline corrections were carried out within the instrument software, and the TMSP-d₄ signal was calibrated at δ 0.00 ppm.

The 1D NMR spectra were assigned using the databases software Chenomx NMR Suite (professional version 8.6), Human Metabolome Database (http://www.hmdb.ca/), and literature values (Zawadzki et al., 2017; Ceribeli et al., 2018), the 2D NMR experiments JRES, ¹H-¹³C HSQC, and ¹H-¹H COSY on selected samples.

Individual metabolite peaks were integrated and quantified relative to the “Electronic Reference to access In vivo Concentrations 2” (ERETIC2) signal experiment, which was performed using the Sucrose standard (Bruker) and the fixed receiving gain used for the samples. Identification of individual metabolites was carried out using absolute integrals and grouped according to their chemical classification and biological function in the muscle. Classification by chemical characteristic is as follows: amino acid (tyrosine, creatinine, anserine, glutathione, glycine, betaine, l-carnitine, β-alanine, methionine, l-glutamine, proline, l-arginine, l-alanine, l-threonine, l-valine, l-leucine, and l-isoleucine); organic acid (fumarate, l-lactate, citrate, succinate, acetate, and methylmalonate); sugar (glucose and glucose-6-phosphate); nucleotide and nucleoside (adenosine and inosine); and others (glycerol and choline). Metabolites groups by biological function or related to it are as follows: antioxidant (anserine, β-alanine, betaine, and glutathione); protein synthesis ( l-isoleucine, leucine, l-threonine, tyrosine, and l-valine); metabolism of energy ( l-alanine, adenosine triphosphate, l-carnitine, glucose, glycerol, methylmalonate, and glycine); maintenance, growth, reproduction, and immunity ( l-arginine and l-glutamine); and visual/flavor aspects of meat, or related (lactate and inosine).

Data processing and statistical analysis

The ¹H NMR data had the binning of 0.04 ppm applied, and they were transformed into a data matrix using MNova software. The NMR data were used from 0 to 7 ppm since the peaks signals in the aromatic region were discarded. This procedure was adopted because of variations of chemical shifts due to slight differences in pH and ionic strength. Data were analyzed in MetaboAnalyst 4.0 platform (http://www.metaboanalyst.ca/faces/home.xhtml), using principal component analysis (PCA). The parameters enrolled data filtering interquartile range, no sample normalization, no data transformation, and Pareto scaling (mean-centered and divided by the square root of the standard deviation of each variable). Five principal components were used for discrimination of the analyzed metabolites samples.

The experiment was structured following a random block design. Physicochemical parameters and individual/grouped metabolites were analyzed with SAS Studio (Cary, North Caroline, USA). Residual normality and influencer point analyses were estimated by MIXED procedures. Statistical analysis was carried out using GLM procedure, and the model was composed of the treatments (0%, 1%, 2%, and 4% of YME inclusion) and blocks (1 to 9) as fixed effects, as well as error was included into the model. The treatment effects were analyzed by orthogonal contrast (linear, quadratic, and cubic) using the coefficients from the IML procedure to adjust unequal spaced treatments and were considered significant if P ≤ 0.05 and considered a tendency if 0.05 < P ≤ 0.10. Power analysis was calculated, with data from previous experiments, using the POWER procedure of SAS (Statistical Analysis System, 9.4, Cary, NC, USA).

For pathway analysis, the differential metabolites were cross-listed with the pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG), using the previously identification at Bovine Metabolome Database (BMDB), PubChem Compound, Chemical Entities of Biological Interest (ChEBI), Japan Chemical Substance Dictionary Web (NIKKAJI), and Chemical Abstracts Service (CAS). The top altered pathways were identified and building according to the potential functional analysis. For this purpose, MetaboAnalyst 4.0 platform was also used.

Results

For most of the evaluated meat quality traits, there were no cubic effects of the orthogonal contrast (data not shown), unless otherwise stated.

Instrumental meat quality parameters

The results for the proximate analysis of the thawed lamb meat are presented in Table 2. Ash (P < 0.01), OM (P = 0.01), protein (P = 0.01), TL (P = 0.04), and the ratio PM:OM (P < 0.01) had a quadratic effect of the different feeding treatments. However, the other parameters showed no statistical difference (P > 0.05), such as MC (mean ranging from 60.46 to 65.51 g 100g⁻¹ meat), ratio protein:OM (mean ranging from 0.72 to 0.76 g of protein per g of OM), ratio TL:OM (mean ranging from 0.077 to 0.094 g of TL per g of OM), and PMs content (mean ranging from 2.36 to 2.62 g 100g⁻¹ meat).

Table 2.

Proximate analysis of the meat from lambs fed increased levels of yerba mate extract in the diet

Parameters1	Yerba mate extract					P-value3
(g 100g−1 meat)	0%	1%	2%	4%	MSE2	Lin	Quad
MC	65.51	60.46	60.55	65.48	1.75	0.64	0.11
ASH	1.56	1.70	1.96	1.60	0.09	0.80	<0.01
OM	32.92	37.71	37.50	32.92	1.66	0.62	0.01
Protein	24.25	27.13	28.09	24.79	1.16	0.99	0.01
TL	2.60	3.21	3.52	2.97	0.30	0.49	0.04
PM	2.36	2.62	2.38	2.55	0.07	0.24	0.76
Ratio protein:OM	0.74	0.72	0.75	0.76	0.01	0.16	0.63
Ratio TL:OM	0.077	0.082	0.094	0.079	0.007	0.79	0.12
Ratio PM:OM	0.075	0.071	0.064	0.078	0.003	0.46	<0.01

¹MC, moisture content; ASH, ash content; OM, organic matter content; TL, total lipids content; PM, polar metabolites; ratio protein:OM, ratio between protein and organic matter; ratio TL:OM, ratio between total lipids and organic matter; ratio PM:OM, ratio between polar metabolites and organic matter.

²MSE, mean square error.

³Values from orthogonal contrast (linear and quadratic) are significantly different if P ≤ 0.05 and tendency if 0.05 < P ≤ 0.10.

Instrumental meat quality parameters are presented in Table 3. Measurements of pH and WHC had no differences (P > 0.05). The same occurred with parameters of color such as a*, b*, ratio a*:b*, chroma, and hue angle. Despite that, CWL had a linear increase (P = 0.05) as the YME content increased in the diet. Shear force and L* parameter of the color tended to a quadratic effect (P = 0.07 and P = 0.09, respectively). Meat from lambs fed YME had an increase in tenderness (measured by the decrease of shear force), by 1%, 2%, and 4% of extract inclusion, respectively, comparing with control (0%).

Table 3.

Physicochemical parameters of the meat from lambs fed increased levels of yerba mate extract in the diet

Parameter1	Unit	Yerba mate extract					P-value3
		0%	1%	2%	4%	MSE2	Lin	Quad
pH		5.37	5.37	5.39	5.39	0.01	0.13	0.67
CWL	%	26.26	24.32	27.24	29.16	1.33	0.05	0.40
WHC	%	49.73	52.31	52.98	52.20	1.75	0.41	0.28
SF	N	32.21	24.28	28.28	29.63	2.12	0.90	0.07
Color
L*		38.18	38.19	37.96	41.00	0.72	< 0.01	0.09
a*		14.03	14.34	13.69	14.30	0.40	0.80	0.57
b*		3.63	3.99	4.49	4.27	0.37	0.22	0.28
Ratio a*:b*		3.69	3.77	3.19	3.48	0.25	0.40	0.41
Chroma		14.51	14.91	14.47	14.94	0.43	0.61	0.87
Hue angle		0.272	0.269	0.290	0.289	0.016	0.40	0.78

¹pH, potential of hydrogen; CWL, cooking weight loss; WHC, water holding capacity; SF, shear force; L*, lightness; a*, redness; b*, yellowness; ratio a*:b*, ratio between a* value and b* value.

²MSE, mean square error.

³Values from orthogonal contrast (linear and quadratic) are significantly different if P ≤ 0.05 and tendency if 0.05 < P ≤ 0.10.

Muscle metabolites

A total of 37 metabolites were identified in the NMR spectra of LL lambs’ muscle. A typical ¹H NMR spectrum from a sample of lamb meat fed 2% of YME is illustrated in Figure 1. Individual quantification of 29 of those metabolites was achieved (Supplementary Tables S1 and S2), and the other 8 (carnosine, malonate, taurine, creatine, adenosine, phenylalanine, niacinamide, and hypoxanthine) had low signal-to-noise ratio or overlapped signals.

Figure 1.

Proton nuclear magnetic resonance (¹H NMR) spectra obtained from the metabolites extracted from lamb muscle from animals fed 2% of yerba mate extract. Ile, l-isoleucine; Leu, l-leucine; Val, l-valine; Mma, methylmalonate; Lac, lactate; Thr, l-threonine; Ala, l-alanine; Arg, l-arginine; Ace, acetate; Pro, proline; Gln: l-glutamine; Glu, glutathione; Met, methionine; Suc, succinate; Crn, l-carnitine; Cit, citrate; Car, carnosine; Ans, anserine; Cre, creatine; Mal, malonate; BAla, β-alanine; Cho, choline; Tau, taurine; Bet, betaine; Glc, glucose-6-phosphate; Gly, glycine; GLY, glycerol; Ino, inosine; Crt, creatinine; ATP/ADP/AMP, adenosine tri-, di-, or mono-phosphate; GLC, glucose; Ado, adenosine; Fum, fumarate; Tyr, tyrosine; Phe, phenylalanine; Nia, niacinamide; Hyp, hypoxanthine.

l-Lactate was the metabolite in greater concentration on lamb muscle (Supplementary Tables S1 and S2), followed by creatinine, glucose-6-phosphate, adenosine triphosphate, anserine, and carnitine. Together those six metabolites represented around 80% of the total PMs extracted in lamb muscle. Regarding the individual metabolites expressed in milligrams per 100 g of muscle (Supplementary Table S1), a linear positive effect by the treatment was observed for methionine (P = 0.02), and a tendency for negative and positive effect was observed for l-lactate (P = 0.06) and glucose-6-phosphate (P = 0.07), respectively. The quadratic effect was observed for glucose (P = 0.04), acetate (P = 0.03), and l-alanine (P = 0.03), as well as a tendency effect for fumarate (P = 0.10).

Concerning the relative proportion of individual metabolites (Supplementary Table S2), a tendency of a negative linear effect was observed for β-alanine (P = 0.06) and l-alanine (P = 0.09). A quadratic effect was observed on the relative proportion of fumarate (P = 0.04) and glycine (P = 0.03), and a tendency was observed for glucose (P = 0.07) and acetate (P = 0.08).

Grouped metabolites by the chemical classification are presented in Table 4; there was a linear and quadratic tendency on organic acids (P = 0.06) and sugars (P = 0.07) expressed in milligrams per 100 g of muscle, respectively. A similar result was observed on the relative proportion of those metabolites, a linear and quadratic effect for organic acid (P = 0.02) and sugars (P = 0.05). Concerning grouped metabolites by their biological function and expressed in milligrams per 100 g of muscle, a negative linear tendency effect was observed for metabolites related to visual/flavor aspects of meat (P = 0.06). Also, a negative linear effect was observed for metabolites related to antioxidant function (P = 0.01) and visual/flavor aspects of meat (P = 0.02) expressed as relative proportion.

Table 4.

Quantification of metabolites (g/100g of muscle) from nuclear magnetic resonance (NMR) analysis of muscle longissimus thoracis from lambs fed increased levels of yerba mate extract in the diet

Metabolites	Yerba mate extract					P-value2
	0%	1%	2%	4%	MSE1	Lin	Quad
Quantification of metabolites
Based on chemical classification, mg 100 g−1 of muscle
Amino acids3	1,090.5	1,209.4	1,084.8	1,111.6	37.0	0.70	0.49
Organic acids4	709.7	736.7	660.7	662.1	24.1	0.06	0.88
Sugars5	321.1	323.4	290.1	358.2	18.1	0.18	0.07
Nucleotide and nucleoside6	245.4	300.5	242.2	269.8	20.5	0.84	0.78
Others7	41.4	46.2	39.7	42.9	1.7	0.90	0.93
Based on biological function, mg 100g−1 of muscle
Antioxidant function or related8	421.7	462.4	427.0	411.6	18.8	0.36	0.29
Protein synthesis or related9	31.3	37.9	29.6	32.5	3.2	0.78	0.88
Metabolism of energy or related10	461.2	528.1	487.5	474.6	31.9	0.89	0.32
Maintenance, growth, reproduction, and immunity, or related11	24.1	31.5	23.6	28.2	2.1	0.56	0.84
Visual/flavor aspects of meat, or related12	726.6	743.9	690.6	681.4	21.3	0.06	0.98
Relative proportion
Based on chemical classification, %
Amino acids3	46.47	46.32	45.62	44.57	0.94	0.13	0.86
Organic acids4	28.49	28.17	27.68	26.16	0.74	0.02	0.67
Sugars5	13.30	12.28	12.47	14.99	0.71	0.05	0.05
Nucleotide and nucleoside6	10.47	11.45	11.45	11.10	0.44	0.47	0.13
Others7	1.71	1.78	1.69	1.70	0.08	0.68	0.85
Based on biological function, %
Antioxidant function or related8	18.47	17.67	17.57	16.13	0.61	0.01	0.87
Protein synthesis or related9	1.28	1.44	1.12	1.28	0.11	0.62	0.64
Metabolism of energy or related10	18.69	20.15	20.24	18.93	0.84	0.95	0.12
Maintenance, growth, reproduction, and immunity, or related11	1.00	1.21	0.90	1.12	0.08	0.70	0.62
Visual/flavor aspects of meat, or related12	29.00	28.46	28.26	26.85	0.61	0.02	0.73

¹MSE, mean square error.

²Values from orthogonal contrast (linear and quadratic) are significantly different if P ≤ 0.05 and tendency if 0.05 < P ≤ 0.10.

³Sum of tyrosine, creatinine, anserine, glutathione, glycine, betaine, l-carnitine, β-alanine, methionine, l-glutamine, proline, l-arginine, l-alanine, l-threonine, l-valine, l-leucine, and l-isoleucine.

⁴Sum of fumarate, l-lactate, citrate, succinate, acetate, and methylmalonate.

⁵Sum of glucose and glucose-6-phosphate.

⁶Sum of adenosine and inosine.

⁷Sum of glycerol and choline.

⁸Sum of anserine, β-alanine, betaine, and glutathione

⁹Sum of l-isoleucine, leucine, l-threonine, tyrosine, and l-valine

¹⁰Sum of l-alanine, adenosine triphosphate, l-carnitine, glucose, glycerol, methylmalonate, and glycine.

¹¹Sum of l-arginine and l-glutamine.

¹²Sum of lactate and inosine.

Metabolomics

Results from principal component (PC) analysis from lamb muscle metabolomics of lambs fed increasing levels of YME are shown in Figure 2. The first graph in Figure 2A is the PCA score, in which 86.3% of the total variance were explained (82.2% in PC1 and 4.1% in PC2) in a 2D graphic regarding samples. Two contrasting groups of metabolomics profiles were clustered in component 1 (PC1). One group was represented by metabolomics profile corresponding to animals fed 0% and 4% YME vs. the second group, which includes animals fed 1% and 2% YME. PC3 was added; thus, a 3D figure is included (Figure 2B), that last added component contributed with 3% more, totaling 89.3% of the total variance. Furthermore, animals fed 2% YME were most different from animals fed 0% as seen in both PCA projections scores (Figure 2A and B).

Figure 2.

Principal components analysis scores of metabolites from lambs fed increased levels of yerba mate extract in the diet: (A) principal component (PC)1 × PC2 and (B) PC1 × PC2 × PC3.

The metabolites profile responsible for the differentiation between the degrees of YME, which the lambs have been fed, is presented by the PCA loadings charts, as shown in Figure 3, which had a threshold determined as 0.1 and 0.05 for PC1 × PC2 and PC1 × PC3; respectively. These thresholds separate the metabolites in the positive values since those were responsible for distinguishing animals fed with YME from the control animals.

Figure 3.

Principal components analysis (PCA) loadings of metabolites from lambs fed increased levels of yerba mate extract in the diet (A) PCA loading 1 vs. 2 and (B) PCA loading 1 vs. 3. Ile, l-isoleucine; Leu, l-leucine; Val, l-valine; Mma, methylmalonate; Lac, lactate; Thr, l-threonine; Ala, l-alanine; Arg, l-arginine; Ace, acetate; Pro, proline; Gln, l-glutamine; Glu, glutathione; Met, methionine; Suc, succinate; Crn, l-carnitine; Cit, citrate; Ans, anserine; Cre, creatine; BAla, β-alanine; Cho, choline; Bet, betaine; Glc, glucose-6-phosphate; Gly, glycine; GLY, glycerol; Ino, inosine; ATP, adenosine triphosphate; Ado, adenosine; Fum, fumarate; Tyr, tyrosine.

PC1 × PC2 and PC1 × PC3 loadings (Figure 3A and B) include the highlighted metabolites such as l-alanine, methionine, succinate, anserine, creatine, β-alanine, choline, l-carnitine, betaine, glucose-6-phosphate, glycine, glycerol, glutathione, and lactate. These were the most contributing species in the group with animals fed 1% and 2% YME against animals fed with 0% and 4% YME. In addition, PC1 × PC3 loadings have highlighted on the PC3 axis the metabolites l-leucine, l-valine, methylmalonate, l-alanine, methionine, β-alanine, choline, betaine, glucose-6-phosphate, and lactate (Figure 3B), which has shown a tendency of separation between the metabolome of muscle from animals fed control diet and 4% of YME.

Metabolic pathway analysis

In Figure 4A, the metabolic pathway of common metabolites as a view map shows 37 pathways, which are presented in Supplementary Table S3, concerning the 37 metabolites identified in all samples from the four different feeding regimes, 15 of those pathways had a pathway impact value greater than 0.1, which is the cutoff value for relevance. The 15 most important pathways are 1) phenylalanine, tyrosine, and tryptophan biosynthesis; 2) beta-alanine metabolism; 3) taurine and hypotaurine metabolism; 4) phenylalanine metabolism; 5) glutathione metabolism; 6) glycine, serine, and threonine metabolism; 7) glycerolipid metabolism; 8) nicotinate and nicotinamide metabolism; 9) tyrosine metabolism; 10) citrate cycle (tricarboxylic acid cycle); 11) arginine and proline metabolism; 12) histidine metabolism; 13) glyoxylate and dicarboxylate metabolism; 14) alanine, aspartate, and glutamate metabolism; and (15) cysteine and methionine metabolism.

Figure 4.

Pathway maps: (A) represents the pathway map of 37 common metabolites found in lamb muscle fed increased levels of yerba mate extract in the diet, named according to Supplementary Table S3; (B) represents the pathway maps of distinguish metabolites of cluster formed (Figure 2A) by treatment 1% and 2% of inclusion of yerba mate extract (l-alanine, methionine, succinate, anserine, creatine, β-alanine, choline, l-carnitine, betaine, glucose-6-phosphate, glycine, glycerol, glutathione, and lactate)—common metabolites named according to Supplementary Table S4. Greater diameter of the cycle in the figure A and B are correspondent to greater pathway impact and darker color if related to greater -Log(P) values.

Based on both P-value (P ≤ 0.05) and impact value (> 0.1), it is possible to point out from the 15 pathways cited above that 1), 2), 4), 6), 10), 13), and 14) were the most relevant metabolic pathways, which have their metabolites involved mainly in amino acid metabolism, energy, and buffer activity.

Regarding only the muscle metabolites highlighted in animals fed 1% and 2% of YME (Figure 3A and B) composed of l-alanine, methionine, succinate, anserine, creatine, β-alanine, choline, l-carnitine, betaine, glucose-6-phosphate, glycine, glycerol, glutathione, and lactate. Metabolic pathways shown in Figure 4B reported 24 metabolic pathways, which are presented in Supplementary Table S3, 7 of those had a pathway impact value greater than 0.1, which is the cutoff value for relevance. The seven most important pathways are 1) beta-alanine metabolism; 2) glutathione metabolism; 3) glycine, serine, and threonine metabolism; 4) glycerolipid metabolism; 5) starch and sucrose metabolism; 6) glyoxylate and dicarboxylate metabolism; and 7) cysteine and methionine metabolism.

Based on both P-value (P ≤ 0.05) and impact value (> 0.1), it is possible to point out from the seven pathways cited above that 1), 2), and 3) were the most relevant metabolic pathways, which have their metabolites involved in amino acid metabolism and antioxidant metabolism.

Discussion

Instrumental meat quality parameters

Concerning the proximate analysis of the meat (Table 2), no effect in MC was observed, even though, numerically, we can observe a discrepancy between control (0%) and 4% vs. 1% and 2% of mate extract. That numerical difference could have influenced the composition content of the meat and might help to explain the treatment effects on nutritional components due to the intrinsic characteristic of those variables. Proximate analysis is the nutritional composition of the food measured by chemical analysis and is usually expressed in percentage. The sum of all must be 100%, and high variation in one of the parameters analyzed can influence the other results, especially if the variation is concentrated in a variable which has a greater magnitude, such as water.

This difference between control and 4% vs. 1% and 2% may be explained by feed sorting (data not presented) and feed intake (Lobo et al., 2020). Animals fed the control diet, 1%, and 2% YME had a feed sort index close to 1, meaning that no sorting was detected. However, for 4% YME, a greater feed sort index was observed, indicating that animals potentially selected their feed. Furthermore, a linear reduction (P = 0.03) on feed intake in animals fed YME was observed; animals consumed 1.15, 1.15, 1.17, and 1.00 kg of DM/d when fed 0%, 1%, 2%, and 4% YME, respectively (Lobo et al., 2020). Therefore, animals fed 4% of YME had a lower feed intake, and lower YME intake, probably due to the bitter taste of YME.

Lamb meat pH post thawing had slightly lower values (Table 3) than the literature, which describes values of pH of meat after 24 h of slaughter between 5.5 and 5.8 (Jandasek et al., 2014; Bezerra et al., 2016; de Abreu et al., 2019). Our companion paper (Pena-Bermudez et al., 2020) reported that lambs LL muscle had no pH difference (P = 0.56) among treatments after 24 h of the slaughter, with an average of 5.5. Similar values were observed by Leygonie et al. (2011, 2012) in which the thawed meat had a lower pH compared with fresh meat. Since pH is the measurement of the quantity of the free hydrogen ions in the water environment as the solvent, the process of freezing meat could cause different effects in the meat’s pH, such as denaturation of muscle proteins, consequently changing water distribution, and pH in the tissue. In addition, the process of freezing and thawing could result in an exudate production, and it could change the concentration of solutes. Also, according to Leygonie et al. (2011, 2012), pH reduction could be caused by microbial or enzymatic activity, which releases hydrogen ion by deamination and hydrolysis of proteins.

Bouton et al. (1971) reported that small changes in pH may have major impacts on other parameters of meat quality, such as WHC and tenderness of meat. Hughes et al. (2014) pointed out that the kinetic of the pH reduction during the process of conversion of muscle to meat has important structural effects, such as myofibrillar lattice spacing, expulsion of fluid to the extracellular space, and development of drip channels. According to the authors, all three structural changes have major impacts on WHC, cook weight loss, tenderness, and color of the meat which may help to explain our results; however, more studies evaluating the drop kinetics of pH during the conversion of the muscle to meat are needed.

The results shown a quadratic response for tenderness, with the greater point of the curve on animals fed 1% of YME, and a positive linear effect on lightness of the meat from lambs fed YME (Table 3). That could be associated with tender meat with lower water losses during storage and cook, as well as a better visual appearance of the meat. According to Holman et al. (2017, 2020) and Khliji et al. (2010), the color and tenderness of the meat are important parameters of acceptability of the meat by consumers. For fresh lamb meat, redness (a*) and lightness (L*) values are considered acceptable for consumer when equal or greater than 9.5 and 34, respectively (Khliji et al., 2010).

In the study of YME feeding for beef cattle, by de Zawadzki et al. (2017), it was reported greater results for instrumental meat quality parameters. Regarding animals fed control diets and 1% of YME, the data showed a reduction of shear force from 58.06 to 49.72 N, respectively, a total of 8.34% reduction, and also, increasing of luminosity of the meat, ranging from 40.66 to 41.92, respectively, corroborating with our results (Table 3). However, more studies feeding YME and its effects on pH and meat quality should be done to give us a better understanding of the mechanism of action mechanism of plant biomolecules on early postmortem metabolism.

Metabolomics of lamb’s muscle

The metabolomics samples were collected after slaughter to assure that they represent the biology of the live animals. Although meat quality was assessed in thawed samples, to mimic what happens with the majority of lamb meat consumption, we acknowledge that the freeze-thawing process can increase membrane breakage and loss of metabolites in the purge, possibly modifying the metabolomics of the thawed samples; however, our goal was to assess the metabolomics of the live animal with subsequent meat quality. All 37 metabolites found in this study (Figure 1) corroborates with the literature (Osorio et al., 2012; de Zawadzki et al., 2017; Ceribeli et al., 2018). Although there was no statistical difference between the majority of the individual (Supplementary Table S1) and the grouped metabolites (Table 4), a global analysis of those metabolites indicates changes in the metabolic profile and consequently a clusterization of the results, as observed in Figure 2, which means that the profile of animals fed 1% and 2% YME was similar and it contrasted with the metabolome of animals fed 0% and 4% YME. Such clusterization was mostly affected by particular metabolites, which are shown in Figure 3 and discussed above.

Betaine and taurine, for example, induce and regulate the synthesis of muscle endogenous antioxidants. Taurine may act as a neurotransmitter in the brain, facilitator of ion transport, and stabilizer of cell membranes and mitochondrial pH gradient (Hansen et al., 2010; Oliveira et al., 2010; Beauclercq et al., 2016). This metabolite showed no or weak antioxidant activities, according to Wu et al. (2003), but it is related to protection against oxidative stress (Williams, 2007). Betaine can increase antioxidant defenses through the regulation of sulfur amino acid metabolism and comes from choline at glycine, serine, and threonine metabolism. Some studies with chicken muscle demonstrated that betaine may increase the activity of some enzymes, such as glutathione peroxidase, catalase, and superoxide dismutase. These enzymes enhance the protection against lipid peroxidation (Alirezaei et al., 2012; Jung et al., 2013; Beauclercq et al., 2016). Indeed, high levels of betaine can stimulate glutathione synthesis.

Beta-alanine is not an antioxidant by itself; however, this metabolite is relevant because it takes part in the biosynthesis of the histidine-derived metabolites carnosine and anserine (Harris et al., 2006; Culbertson et al., 2010; Derave et al., 2010). According to Ceribeli et al. (2018), carnosine and anserine may play an important role as endogenous antioxidants. Anserine together with carnosine is responsible for almost the entire amount of antioxidant compounds in the lamb meat, which corroborates with the scientific literature (Williams, 2007). Anserine is an efficient radical scavenger (Kohen et al., 1988) and may act as a chelator of redox-active metal ions, factors that contribute to the protection against oxidation (Wu et al., 2003). More specifically, anserine and carnosine can prevent myofibrillar protein oxidation leading to a tender meat.

The endogenous antioxidants cited below can have an impact on the color stability of the meat. Oxymyoglobin is the major pigment responsible for the red fresh meat color; however, when oxidized, it becomes metmyoglobin, which results in a brownish and less bright color. An improvement in the endogenous antioxidant profile, such as presented in Figures 3B and 4B, may help reducing meat oxidation and improve the lightness of the meat, as observed in our results. Also, our companion paper (Pena-Bermudez et al., 2020) demonstrated that dietary inclusion of YME is related to low thiobarbituric acid reactive substances and increased glutathione peroxidase activity at shelf life essay, which improved color stability. Besides the antioxidant effects in the color, tenderness can also be affected. According to Wang et al. (2018), oxidation could promote cell apoptosis in the early postmortem and initiate proteolytic degradation of fibers in the muscle which consequently could improve meat tenderness. However, we observed the opposite effect, an increase in endogenous antioxidant profile and a trend to reduce tenderness. However, the pathway changes by dietary YME in the early postmortem metabolism are still unknown.

Cysteine, histidine, proline, methionine, tyrosine, and tryptophan (the latest two as products of phenylalanine, tyrosine, and tryptophan biosynthesis) are the amino acids more susceptible to oxidation, and they have been proposed to be involved in the redox stabilization of proteins and antioxidant enzymes, when free in the biological medium. However, in meat, the oxidation of residues of tyrosine and cysteine can lead to the formation of cross linking of proteins (Lantto et al., 2007; Lund et al., 2008, 2011), which can promote change in the structure and loss of functionality of proteins.

l-Threonine, l-valine, l-leucine, and L-isoleucine are associated with in vivo protein synthesis, whereas arginine, proline, and glutamine are involved in maintenance, growth, reproduction, and immunity. However, some amino acid residues in the protein side chain such as arginine and proline are easily oxidized via metal-catalyzed reactions that generate carbonyl residues which are susceptible to cross-link reactions. These impact protein oxidation and compromise meat quality (Lund et al., 2011).

l-carnitine, l-alanine, adenosine triphosphate, glycerol, glucose, and glucose-6P are involved in energy processes. Glucose and glucose-6P are involved in the starch and sucrose metabolism and are the most important source of energy for the muscle cell. Glucose 6-P is the product of glucose phosphorylation, and in the muscle, it is used in the glucogenesis pathway (to produce glycogen) or in the glycolysis pathway, which will convert glucose 6-P into energy for the cell. The inclusion of YME had a quadratic effect on glucose, with lower values in the muscle of animals fed 1% and 2% of YME; however, when compared with the sum of sugar metabolites, animals fed 1% and 2% had an opposite response (greater values). Frylinck et al. (2013) reported that steers had greater muscle glucose than older animals, which could be related to a lighter meat. A similar result was observed; as described previously, animals fed 1% and 2% YME had a lower glucose content and a lighter meat. However, the relationship between glucose and meat lightness is not well established and further studies are needed.

Glycine is involved in the production of many important metabolites found in the lamb meat, one of them is creatine that releases creatinine. Creatine is a molecule found in meat and fish that can be synthesized by the liver and pancreas (Walker, 1979) from arginine, glycine, and methionine (Bloch and Schoenheimer, 1940). In the skeletal muscle, creatine also plays an important role in energy production. The use of creatine and creatinine can be identified in the metabolomics profile of animals fed 1% and 2% of YME. Some studies have reported that creatinine is involved in muscle hypertrophy by increasing free fat mass (Ingwall et al., 1974; Parise et al., 2001; Forbes et al., 2019). That idea corroborates with our companion article, in which animals fed 1% and 2% of YME achieved a greater body weight and a lower carcass fat thickness (Lobo et al., 2020). In addition, it was reported by Forbes et al. (2019) that, in the human body, increasing creatine is related to increasing free fat mass even for people who have a tendency to be obese.

Lactate is the most abundant metabolite on lamb muscle and it is related to meat color stability. According to Kim et al. (2006), lactate could be used in the lactate–lactic dehydrogenase system to help the reduction of NAD⁺ to NADH. The product of this reaction is required during metmyoglobin-reducing activity, which is related to great meat color stability (Mancini and Hunt, 2005). Ramanathan et al. (2011) reported that lactate enhances color stability and reduces darkness (L*) of meat. Also, Cônsolo et al. (2021) observed that darker meat had a lower concentration of lactate. However, lactate concentration in the muscle ongoing postmortem metabolism is variable, Frylinck et al. (2013) reported that lactate concentration of beef steers begins at 40 µmol/g and can reach around 75 µmol/g 20 h after slaughter. Our results showed that a lighter meat was produced by the inclusion of YME, suggesting that maybe the lactate–lactic dehydrogenase system was enhanced due to a greater concentration of lactate in the muscle at slaughter, a greater amount of NADH was produced and an increase in deoxymyoglobin by the metmyoglobin reduction could explain the reported effects. Another key metabolite, which has an important relationship with meat quality, is inosine, a metabolite present in fresh meat that is one of the components that give meat an “umami” taste (Maga, 1987) and is classified as a flavor precursor (Koutsidis et al., 2008).

Metabolic pathway

The most important metabolic pathway detected in the global analysis of pathways is the biosynthesis of aromatic amino acids such as phenylalanine, tyrosine, and tryptophan. Those amino acids are essential, which means the animal organism could not produce them in enough quantities to meet the animals’ requirements. Therefore, the supply of referred amino acids must come from the diet or the microbial community of the rumen and gut (Parthasarathy et al., 2018). Following the biosynthesis of aromatic amino acids, glycine, serine, and threonine metabolism is the second most relevant pathway for protein synthesis in the present study. Glycine, serine, and threonine metabolisms are related to protein synthesis in the muscle. More specifically, glycine itself is responsible for 11.5% of the total amino acid in body protein (Wu, 2010), and the collagen of the muscle requires it for every third position on its chain a glycine residue (Wang et al., 2013).

As we demonstrate, some protein synthesis metabolisms were highly activated in the lambs’ muscle; however, at the same time, an anabolic pathway may be taking place to provide a healthy protein turnover. The metabolism of phenylalanine is one of the pathways related to that anabolic process (Rasmussen et al., 2000). Despite building blocks of protein, some amino acids are related to other vital functions in the body, and alanine, aspartate, and glutamate metabolisms are related to the neural functions (Santos et al., 2012; Engskog et al., 2017).

Another main metabolism present in the muscle is the citrate cycle—this metabolic pathway consumes about 66% of the total oxygen obtained by the lungs and is responsible for generating the majority amount of energy (around 66%) in the body (Akram, 2014). This important metabolism is related to a key role in deamination, transamination, gluconeogenesis, and lipogenesis (Moffett and Namboodiri, 2003).

The beta-alanine metabolism is the second most important pathway in lamb muscle. This pathway is related to carnosine synthesis (Caruso et al., 2012; Boldyrev et al., 2013; Roveratti et al., 2019), which is reported to be stored in great amounts also in human muscle (Derave et al., 2010; Blancquaert et al., 2017). Carnosine is highly related to buffering capacity and antioxidant activity in muscle (Caruso et al., 2012; Perim et al., 2019), a key role in the homeostasis. Also, antioxidant metabolism was reported by the pathway analysis, the glutathione metabolism is responsible for glutathione production, this metabolite is considered to be an endogenous-reducing agent and antioxidant with key role in the cellular metabolism (Luo et al., 1996).

Glutathione metabolism is a very important pathway, highlighted in the muscle of animals fed 1% and 2% of YME. Its basic function in the muscle is detoxification of xenobiotics, cysteine reserve, and protein synthesis; however, this pathway is related to a defense against reactive oxygen species and the redox state of the cells (Sen, 1998). Glutathione is related to health status and immune response of the animal (Liu et al., 2005) and could be the reason for the observed increase in blood defense cells (segmented neutrophils and lymphocytes) in lambs fed up to 2% YME in the companion article (Lobo et al., 2020). Also, this metabolite is related to reduction of postmortem oxidation of components, such as lipids and proteins, enhancing the instrumental parameters of meat quality, such as tenderness and color (Mercier et al., 2004; Rowe et al., 2004).

The most important pathways are related to amino acid metabolism and protein synthesis, as we stated earlier; however, other crucial pathways were reported, such as energy metabolism and buffer activity metabolism. Those results were expected, because the animals from the experiment were young and growing, which require the building of protein and maintenance of muscle pH homeostasis. Also, energy metabolism is related to postmortem metabolism, in which muscle glycogen is converted to pyruvate and consequently into lactate, which can impact the pH drop and metmyoglobin-reducing activity during the postmortem metabolism. Compared with the metabolic pathway from the animals grouped in PC1 loadings (fed 1% and 2% YME), the pathway of endogenous antioxidant was reported with a key role, which indicates that animals fed those doses of YME had a better muscle antioxidant metabolism. These results are related to the differences reported in instrumental meat quality for animals fed 1% and 2% YME (Table 3).

In summary, YME dietary supplementation up to 2% of the diet to lambs had little to no effects on the majority of meat quality traits evaluated; moreover, 4% of YME inclusion negatively affected feed intake and meat quality traits. We were able to identify 37 metabolites from lamb muscle. Analysis of the majority of the individual metabolites was not qualitatively affected by treatments; however, the quantitative profile analysis showed an increased early postmortem of metabolites related to antioxidant activity in animals fed 1% and 2% of YME, such as l-alanine, methionine, succinate, anserine, creatine, β-alanine, choline, l-carnitine, betaine, glucose-6-phosphate, glycine, glycerol, glutathione, and lactate. Feed sort index linearly increased when animals were fed 4% of YME, which may limit YME intake and negatively affect meat quality traits. In general, there were no convincing evidences showing that YME improved meat quality at up to 2% of the diet, and at 4% inclusion, it reduced feed intake and meat quality.

Glossary

Abbreviations

¹H NMR

proton nuclear magnetic resonance

ADF

acid detergent fiber

ADL

acid detergent lignin

CLA

conjugated linoleic acid

CP

crude protein

DM

dry matter

EE

ether extract

GE

gross energy

LL

Longissimus lumborum

NDF

neutral detergent fiber

NMR

nuclear magnetic resonance

OM

organic matter

PC

principal component

PCA

principal components analysis

PM

polar metabolites

PUFA

polyunsaturated fatty acid

TL

total lipids

Conflict of interest statement

The authors declare no conflict of interest.