Effect of Chinese Herbs on Serum Biochemical Parameters, Immunity Indices, Antioxidant Capacity and Metabolomics in Early Weaned Yak Calves

Simple Summary Herbs, as supplementary feed, have shown positive effects on livestock. The yak is a unique animal on the Qinghai-Tibetan plateau, and plays important roles in local livelihoods and ecology. Because of the harsh climate conditions, early weaned yak calves often face environmental and nutritional stress. The herbal active ingredients of some herbs improve the antioxidant capacity and immunity of animals. In the current study, early-weaned yak calves were supplemented with three widely used Chinese traditional herbal root extracts to examine whether they can improve the immune response and antioxidant capacity. The results demonstrated that the supplementary herbs increased the serum antioxidant capacity, and improved the energy and nitrogen metabolism of the yak calves. Abstract Chinese traditional herbs are used widely as feed supplements to improve the immune response and antioxidant capacity of livestock. Twenty early-weaned 4-month-old yak calves (72.3 ± 3.65 kg) were divided randomly into four groups (n = 5 per group); three groups were provided with supplementary 80 mL/kg DMI of the root water extracts of either Angelica sinensis, Codonopsis pilosula or Glycyrrhiza uralensis, and one group (control) was not provided with a supplement. Compared to control calves, calves consuming the three herbal extracts increased serum concentrations of albumin (ALB) and glutathione peroxidase (GSH-Px), but decreased serum concentrations of free fatty acids (FFAs) and malondialdehyde (MDA) (p < 0.05). Calves consuming A. sinensis decreased (p < 0.05) serum concentration of total cholesterol (TC), and increased (p < 0.05) serum concentration of total proteins (TP). Serum FFA concentrations increased (p = 0.004) linearly with time in the control group, but not in the groups consuming herbs. Serum metabolomic data demonstrated that A. sinensis and C. pilosula regulate mainly amino acid metabolism, while G. uralensis regulates mainly carbon and amino acid metabolism. It was concluded that the three herbal root extracts, as dietary supplements, improved energy and nitrogen metabolism, and enhanced the antioxidant capacity of yak calves.


Introduction
With the increasing global demand for safe and healthy animal products, natural botanical feed additives are being sought to reduce the heavy use of chemicals and an-

Preparation of Herbal Water Extracts
Sun-dried roots of two-year-old A. sinensis, C. pilosula, and G. uralensis were purchased in 2018 (Medicinal Trading Market, Longxi County, Gansu Province), and washed, dried, and ground to 1 mm in a hammer crusher (DF-50-A, Linda Machinery Co., LTD, Wenling, China). The leaching solution for roots of the herbs has been described earlier [16]. The preparation of the herbal extracts from the dry root powder followed Jiang et al. (2021) [15]. Thirteen batches of each herbal extract were prepared to span the needs of the entire study.

Experimental Design
Twenty, four-month-old weaned yak calves (72.3 ± 3.65 kg), 8 males and 12 females, were kept in one pen for one week with free access to concentrate, oat hay (2-3 cm length), and fresh water. Then, the calves were penned (2 m × 4 m) individually, divided into four groups (n = 5 for each group), matched for sex, and offered either no supplement (control group) or 80 mL/kg DMI of herbal root extract of either A. sinensis, C. pilosula, or G. uralensis. The root extract consumed by the calves was 3.12 g/kg DMI for A. sinensis, 3.19 g/kg DMI for C. pilosula, and 2.14 g/kg DMI for G. uralensis.
The study included a 15-day acclimation period for the yak calves and a 60-day measurement period. Each calf consumed approximately 170 mL of herbal root water extract per day, which was mixed with 300 g of concentrate. The actual daily dosage of root water extract was based on the average daily DMI in the previous two weeks, and was consumed completely by the calves at 07:00. Then the calves had free access to oat hay and concentrate, which were provided separately, and at 18:00, concentrate, oat hay, and fresh water were added.

Sampling Procedure
Herbal extracts were collected after each preparation and stored at −20 °C. Ten milliliters of jugular vein blood were collected in vacuum tubes on d 15, 30, 45, and 60 before morning feeding. The blood was centrifuged at 3000× g at 4 °C for 15 min, and the serum was collected and stored at −80 °C.

Analysis of the Main Active Ingredients of Herbal Root Extract
Water extracts of the three herbal roots were condensed from 500 to 50 mL by a rotary evaporator (Lab Tech EV312, Beijing LabTech Instruments Co., Ltd., Beijing, China) and

Preparation of Herbal Water Extracts
Sun-dried roots of two-year-old A. sinensis, C. pilosula, and G. uralensis were purchased in 2018 (Medicinal Trading Market, Longxi County, Gansu Province), and washed, dried, and ground to 1 mm in a hammer crusher (DF-50-A, Linda Machinery Co., LTD, Wenling, China). The leaching solution for roots of the herbs has been described earlier [16]. The preparation of the herbal extracts from the dry root powder followed Jiang et al. (2021) [15]. Thirteen batches of each herbal extract were prepared to span the needs of the entire study.

Experimental Design
Twenty, four-month-old weaned yak calves (72.3 ± 3.65 kg), 8 males and 12 females, were kept in one pen for one week with free access to concentrate, oat hay (2-3 cm length), and fresh water. Then, the calves were penned (2 m × 4 m) individually, divided into four groups (n = 5 for each group), matched for sex, and offered either no supplement (control group) or 80 mL/kg DMI of herbal root extract of either A. sinensis, C. pilosula, or G. uralensis. The root extract consumed by the calves was 3.12 g/kg DMI for A. sinensis, 3.19 g/kg DMI for C. pilosula, and 2.14 g/kg DMI for G. uralensis.
The study included a 15-day acclimation period for the yak calves and a 60-day measurement period. Each calf consumed approximately 170 mL of herbal root water extract per day, which was mixed with 300 g of concentrate. The actual daily dosage of root water extract was based on the average daily DMI in the previous two weeks, and was consumed completely by the calves at 07:00. Then the calves had free access to oat hay and concentrate, which were provided separately, and at 18:00, concentrate, oat hay, and fresh water were added.

Sampling Procedure
Herbal extracts were collected after each preparation and stored at −20 • C. Ten milliliters of jugular vein blood were collected in vacuum tubes on d 15, 30, 45, and 60 before morning feeding. The blood was centrifuged at 3000× g at 4 • C for 15 min, and the serum was collected and stored at −80 • C.

Analysis of the Main Active Ingredients of Herbal Root Extract
Water extracts of the three herbal roots were condensed from 500 to 50 mL by a rotary evaporator (Lab Tech EV312, Beijing LabTech Instruments Co., Ltd., Beijing, China) and then lyophilized (Labconco FreeZone 7.5, Kingston, NY, USA). Polysaccharides were detected using the phenol-sulfuric acid method as described by Jiang et al. (2021), and the content was determined colorimetrically at 485 nm (Molecular Devices, SpectraMax M5, Thermo Fisher Scientific, Waltham, MA, USA), with glucose as a standard. Total saponins were extracted from C. pilosula and G. uralensis root extracts following Jiang et al. (2021), and saponin content was determined colorimetrically at 560 nm, with ginsenoside as a standard for C. pilosula and at 589 nm, with mono-ammonium glycyrrhizinate as a standard for G. uralensis [15]. The main active ingredients of the three herbal root extracts were identified in lyophilized herbal extract by ultra-high-performance liquid chromatography (UHPLC, Column: BEH0C18, 1.7 µm × 2.1 mm × 100 mm; Agilent Technologies 1290, Santa Clara, CA, USA) and mass spectrometry (MS, Type: Q Exactive Focus, Thermo Fisher Scientific Waltham, MA, USA). The MS conditions were described by Jiang et al. (2021). Peak detection, extraction, alignment, and integration of the mass data were processed by R package-XCMS. An in-house MS2 database (BiotreeDB) was applied for compound annotation. The relative contents of the compounds in the root extracts were determined from the peak area ( Figure 2). then lyophilized (Labconco FreeZone 7.5, Kingston, NY, USA). Polysaccharides were detected using the phenol-sulfuric acid method as described by Jiang et al. (2021), and the content was determined colorimetrically at 485 nm (Molecular Devices, SpectraMax M5, Thermo Fisher Scientific, Waltham, MA, USA), with glucose as a standard. Total saponins were extracted from C. pilosula and G. uralensis root extracts following Jiang et al. (2021), and saponin content was determined colorimetrically at 560 nm, with ginsenoside as a standard for C. pilosula and at 589 nm, with mono-ammonium glycyrrhizinate as a standard for G. uralensis [15]. The main active ingredients of the three herbal root extracts were identified in lyophilized herbal extract by ultra-high-performance liquid chromatography (UHPLC, Column: BEH0C18, 1.7 μm × 2.1 mm × 100 mm; Agilent Technologies 1290, Santa Clara, CA, USA) and mass spectrometry (MS, Type: Q Exactive Focus, Thermo Fisher Scientific Waltham, MA, USA). The MS conditions were described by Jiang et al. (2021). Peak detection, extraction, alignment, and integration of the mass data were processed by R package-XCMS. An in-house MS2 database (BiotreeDB) was applied for compound annotation. The relative contents of the compounds in the root extracts were determined from the peak area ( Figure 2).

Serum Metabolome Analysis, Bioinformatics, and Statistical Analysis
Serum samples collected on d 60 were used for metabolome analysis. The prepara tion of the sample and analysis followed   [17].
One-way ANOVA (SPSS 25.0, SPSS Inc., Chicago, IL, USA) compared serum indice (including TP, ALB, BUN, IL-2, TNF-α, and MDA) among treatment groups, with the ya calf as the experimental unit. Where significance existed, the Tukey-adjusted p values sep arated means. Orthogonal polynomial contrasts determined whether responses to th herbal extracts were linear or quadratic with time when serial samples were collected Significance was accepted at p < 0.05 and as a tendency for significance at 0.05 < p < 0.10.
We combined the results in positive and negative ion modes for the main active in gredients of herbal root extract detection. ProteoWizard converted the raw data to th mzXML format, which was processed with an in-house program that was developed us ing R and based on XCMS (https://xcmsonline.scripps.edu/landing_page.php?pgcon tent=mainPage, accessed on 10 March 2021) for peak detection, extraction, alignment, an integration. An in-house MS2 database (BiotreeDB) was applied for metabolite annota tion. The cut-off for annotation was set at 0.3.
Multivariate analysis of normalized data used SIMCA software (V16.0.2, Umea, Swe den). Principal component analysis (PCoA) was used to visualize trends in the sample

Serum Metabolome Analysis, Bioinformatics, and Statistical Analysis
Serum samples collected on d 60 were used for metabolome analysis. The preparation of the sample and analysis followed   [17].
One-way ANOVA (SPSS 25.0, SPSS Inc., Chicago, IL, USA) compared serum indices (including TP, ALB, BUN, IL-2, TNF-α, and MDA) among treatment groups, with the yak calf as the experimental unit. Where significance existed, the Tukey-adjusted p values separated means. Orthogonal polynomial contrasts determined whether responses to the herbal extracts were linear or quadratic with time when serial samples were collected. Significance was accepted at p < 0.05 and as a tendency for significance at 0.05 < p < 0.10.
We combined the results in positive and negative ion modes for the main active ingredients of herbal root extract detection. ProteoWizard converted the raw data to the mzXML format, which was processed with an in-house program that was developed using R and based on XCMS (https://xcmsonline.scripps.edu/landing_page.php?pgcontent= mainPage, accessed on 10 March 2021) for peak detection, extraction, alignment, and integration. An in-house MS2 database (BiotreeDB) was applied for metabolite annotation. The cut-off for annotation was set at 0.3.
Multivariate analysis of normalized data used SIMCA software (V16.0.2, Umea, Sweden). Principal component analysis (PCoA) was used to visualize trends in the samples and orthogonal partial least-squares discriminant analysis (OPLS-DA) examined differences between treatment groups and control groups to remove noise. Metabolites were Animals 2022, 12, 2228 6 of 20 plotted according to their importance in separating the two groups based on the OPLS-DA results, and each metabolite received a variable importance in the projection (VIP) value. Metabolites with VIP values exceeding 1.0 were taken as changed variables, and these were assessed using a Student's t-test. A significantly differential metabolite between the groups was defined as a variable with a VIP > 1 and p < 0.05. Volcano plots were used to visualize the results. Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolic pathways were mapped and analyzed by MetaboAnalyst 3.0 (http://www.metaboanalyst.ca/, accessed on 20 October 2021) based on the differentiated metabolites. Differential pathways were identified based on p < 0.05, and the cutoff of impact value from the topology analysis pathway was set to 0.

Main Active Ingredients in the Herbal Root Extracts
Of the three herbal root extract groups, A. sinensis had the highest contents of phenylpropanoids, organic acids, and their derivatives; C. pilosula had the highest contents of organo-oxygen and miscellaneous compounds; and G. uralensis had the highest contents of terpenoids, alkaloids, phenols, and flavonoids ( Figure 2a). Ligustilide was the main flavonoid in A. sinensis ( Figure 2b); biochanin A and tectochrysin were the main flavonoids in C. pilosula; and maltol, liquiritigenin, baicalin, and wogonoside were the main flavonoids in G. uralensis. Herbal extracts of A. sinensis, C. pilosula, and G. uralensis contained 450, 366, and 476 mg polysaccharides per g DM, respectively (Figure 2c), and C. pilosula and G. uralensis contained 12.5 and 37.6 mg total saponins per g DM, respectively.

Effect of Herbal Extracts on Serum Biochemical Parameters
Serum TC concentration was lower in yaks consuming A. sinensis than in control yaks on d 30; serum TG concentration was lower in yaks consuming G. uralensis than control yaks on days 30 and 60; and serum TP concentration was higher in yaks consuming A. sinensis than in control yaks on d 15 (Table 1). There was no effect of dietary treatment on serum GH concentration, but there was an increasing trend (p = 0.065) of serum GH concentration with time in calves consuming G. uralensis. Calves consuming the three herbal root extracts had lower serum FFA concentrations on d 30 than control yaks, an effect that was maintained in calves consuming C. pilosula and G. uralensis on days 45 and 60. The serum ALB concentration of calves consuming G. uralensis was higher than control calves on days 15 and 60, and of calves consuming A. sinensis was higher (p < 0.05) than control calves on day 60. Serum BUN concentration increased quadratically in calves consuming C. pilosula, whereas, serum INS concentration decreased linearly with time in calves consuming G. uralensis.

Effect of Herbal Extracts on Serum Immunity Indices
Generally, there was no effect of herbal extract intake on concentrations of serum immunity indices, except for serum TNF-α, which was greater (p < 0.05) in calves consuming C. pilosula and G. uralensis than in control calves, and in calves consuming A. sinensis than in control calves on d 45 (Table 2). Serum IgA concentration decreased with time in calves consuming A. sinensis (p < 0.05) and decreased quadratically (p = 0.001) in calves consuming C. pilosula. In general, there were linear or quadratic decreases (p < 0.05) in serum concentrations of IL-2 and IL-6 with time in all calf groups. Control calves consuming A. sinensis decreased (p < 0.05) serum TNF-α concentration quadratically with time.

Effect of Herbal Extracts on Serum Antioxidant Capacity
Yak calves consuming root extract of A. sinensis, C. pilosula, and G. uralensis had lower (p < 0.05) serum MDA concentrations on d 15, consuming C. pilosula and G. uralensis had lower MDA concentrations on d 30, and consuming G. uralensis had lower MDA concentration on d 60 than control calves (Table 3). Serum GSH-Px concentrations were higher (p < 0.05) in calves consuming the three herbal root extracts on days 30, 45, and 60 than in control calves, with calves consuming G. uralensis having the highest concentration. Linear and quadratic increases in serum GSH-Px concentration with time (p < 0.05) emerged in control calves, and calves consuming C. pilosula and G. uralensis. Serum SOD concentration was not affected by diet (p = 0.40-0.80), but there were significant linear and quadratic increases in serum SOD concentration with time in all groups.

Identification and Quantification of LC-MS Compounds in Serum
The stability and repeatability of the system were tested using four quality control (QC) samples. The ionization source of LC-QTOF/MS was electrospray ionization, including positive (POS) and negative (NEG) ion modes. There was substantial overlap in peak retention time and peak area of total ion chromatograms (TIC) from all QC samples, indicating that the analytical system was stable (Figure 3a,b). A total of 6293 valid peaks were identified in POS modes and 6748 in NEG modes for the serum. These peaks were matched for 301 (POS) and 234 (NEG) plasma metabolites based on an in-house MS2 database and the KEGG compound metabolomics library. The metabolites were integrated from the two modes, and a total of 511 qualitative metabolites were obtained.

Metabolomic Profiles in Serum of Early Weaned Yak Calves
Score plots of the PCA (Figure 4a) and OPLS-DA (Figure 4b) modes illustrated separation between the control and A sinensis groups and all the samples fell with 95% confidence interval (Hotelling's T-squared ellipse). The permutation test was for verification (Figure 4c). The Q 2 was 0.00172, indicating that the predictive pow mode was not strong, which was in accord with the results that there were few signi plasma metabolites. For the C. pilosula and G. uralensis groups, the score plots of (Figure 4d,g) and OPLS-DA (Figure 4e,h) modes demonstrated satisfactory modelin predictive abilities between the control and treatment groups, and that all the sampl within the 95% confidence interval (Hotelling's T-squared ellipse). The permutatio was used for verification (Figure 4f,i), and the results indicated that the OPLS-DA m was suitable to test differences between the two groups, as there was no overfittin good stability. Following the criteria of a p < 0.05 for the t-test and the VIP > 1 for the O DA model, significant differential metabolites between the control and treatment g

Metabolomic Profiles in Serum of Early Weaned Yak Calves
Score plots of the PCA (Figure 4a) and OPLS-DA (Figure 4b) modes illustrated good separation between the control and A sinensis groups and all the samples fell within the 95% confidence interval (Hotelling's T-squared ellipse). The permutation test was used for verification ( Figure 4c). The Q 2 was 0.00172, indicating that the predictive power of mode was not strong, which was in accord with the results that there were few significant plasma metabolites. For the C. pilosula and G. uralensis groups, the score plots of PCA (Figure 4d,g) and OPLS-DA (Figure 4e,h) modes demonstrated satisfactory modeling and predictive abilities between the control and treatment groups, and that all the samples fell within the 95% confidence interval (Hotelling's T-squared ellipse). The permutation test was used for verification (Figure 4f,i), and the results indicated that the OPLS-DA model was suitable to test differences between the two groups, as there was no overfitting and good stability. Following the criteria of a p < 0.05 for the t-test and the VIP > 1 for the OPLS-DA model, significant differential metabolites between the control and treatment groups were screened from all identified metabolites. In the A. sinensis group, a total of 16 metabolites, mainly composed of amino acids, were identified and affirmed by combining MS/MS data and available biochemical databases. In the C. pilosula group, 29 metabolites, including amino acids, fatty acids, and glycerophospholipids, were identified. In the G. uralensis group, a total of 51 metabolites, including amino acids, glycerophospholipids, and fatty acids, were identified. were screened from all identified metabolites. In the A. sinensis group, a total of 16 metabolites, mainly composed of amino acids, were identified and affirmed by combining MS/MS data and available biochemical databases. In the C. pilosula group, 29 metabolites, including amino acids, fatty acids, and glycerophospholipids, were identified. In the G. uralensis group, a total of 51 metabolites, including amino acids, glycerophospholipids, and fatty acids, were identified. There were 16 potential biomarkers in the A. sinensis group identified by the Metabo-Analyst website, with six major metabolic pathways, including histidine, glyoxylate and dicarboxylate, β-alanine, glycerophospholipid, glycine, serine and threonine, and arginine and proline metabolism. (Figure 5a). Twenty-nine potential biomarkers were identified in the C. pilosula group, with seven major pathways, including pantothenate and CoA biosynthesis, phenylalanine, tyrosine and tryptophan biosynthesis, pyrimidine metabolism, phenylalanine, β-alanine, glyoxylate and dicarboxylate, and alanine, asparate, and glutamate. (Figure 5b). Fifty potential biomarkers were identified in the G. uralensis group, with six major metabolic pathways, including glyoxylate and dicarboxylate, ascorbate and aldarate, glutathione, pyrimidine, arginine and proline metabolism, and primary bile acid biosynthesis (Figure 5c). There were 16 potential biomarkers in the A. sinensis group identified by the Metabo-Analyst website, with six major metabolic pathways, including histidine, glyoxylate and dicarboxylate, β-alanine, glycerophospholipid, glycine, serine and threonine, and arginine and proline metabolism. (Figure 5a). Twenty-nine potential biomarkers were identified in the C. pilosula group, with seven major pathways, including pantothenate and CoA biosynthesis, phenylalanine, tyrosine and tryptophan biosynthesis, pyrimidine metabolism, phenylalanine, β-alanine, glyoxylate and dicarboxylate, and alanine, asparate, and glutamate. (Figure 5b). Fifty potential biomarkers were identified in the G. uralensis group, with six major metabolic pathways, including glyoxylate and dicarboxylate, ascorbate and aldarate, glutathione, pyrimidine, arginine and proline metabolism, and primary bile acid biosynthesis (Figure 5c).  Compared to control calves, calves consuming root extract of A. sinensis down-regulated the biomarkers of choline in the metabolite pathways of glycine, serine, threonine, and glycerophospholipid; and up-regulated methylimidazole in histidine metabolism, glycolate in glyoxylate and dicarboxylate metabolism, citrulline in arginine biosynthesis, and anserine in β-alanine metabolism (Figure 6a). Root extract of C. pilosula down-regulated the biomarkers of l-phenylalanine in phenylalanine, tyrosine and tryptophan biosynthesis, phenylalanine in phenylalanine metabolism, pantothenate and β-alanine in pantothenate and CoA biosynthesis, 4-acetamidobutanoic acid in arginine and proline metabolism, and deoxythymidylic acid (dTMP) and thymidine in pyrimidine metabolism; and up-regulated succinic acid semialdehyde in butanoate metabolism, and glycolic acid in glyoxylate and dicarboxylate metabolism (Figure 6b). Root extract of G. uralensis downregulated deoxyuridine monophosphate (dUMP) in pyrimidine metabolism and cholic acid and primary bile acids in primary bile acid biosynthesis; and up-regulated 5-oxoproline in glutathione metabolism, 4-hydroxyproline in arginine and proline metabolism, glycolic acid in glyoxylate and dicarboxylate metabolism, and L-gulonol-1,4-lactone in ascorbate and aldarate metabolism (Figure 6c). Compared to control calves, calves consuming root extract of A. sinensis downregulated the biomarkers of choline in the metabolite pathways of glycine, serine, threonine, and glycerophospholipid; and up-regulated methylimidazole in histidine metabolism, glycolate in glyoxylate and dicarboxylate metabolism, citrulline in arginine biosynthesis, and anserine in β-alanine metabolism (Figure 6a). Root extract of C. pilosula down-regulated the biomarkers of l-phenylalanine in phenylalanine, tyrosine and tryptophan biosynthesis, phenylalanine in phenylalanine metabolism, pantothenate and β-alanine in pantothenate and CoA biosynthesis, 4-acetamidobutanoic acid in arginine and proline metabolism, and deoxythymidylic acid (dTMP) and thymidine in pyrimidine metabolism; and up-regulated succinic acid semialdehyde in butanoate metabolism, and glycolic acid in glyoxylate and dicarboxylate metabolism (Figure 6b). Root extract of G. uralensis down-regulated deoxyuridine monophosphate (dUMP) in pyrimidine metabolism and cholic acid and primary bile acids in primary bile acid biosynthesis; and up-regulated 5-oxoproline in glutathione metabolism, 4-hydroxyproline in arginine and proline metabolism, glycolic acid in glyoxylate and dicarboxylate metabolism, and L-gulonol-1,4-lactone in ascorbate and aldarate metabolism (Figure 6c). s 2022, 12, x FOR PEER REVIEW 13 of 20

Effect of Herbs on Serum Biochemical Indices
After one month of herbal intake, serum FFA concentrations were generally lower in the three yak calf groups consuming the herbal extracts than in control yaks. High serum FFA concentrations are related to fat mobilization and negative energy balance [18]. Consumption of A. sinensis polysaccharides decreased the serum concentration of total cholesterol in diabetic mice and accelerated serum FFA oxidation in mice [19,20]. In addition, botanical essential oils, a main component of A. sinensis, lowered the serum concentration of total cholesterol in pigs and sheep [21,22]. In the current study, serum TG concentrations were generally lower in yak calves consuming the herbal extracts, especially G. uralensis, than in control yak calves. Previous studies reported that administration of G. uralensis reduced hepatic TG levels in mice [23], while consuming saponins reduced serum TG concentrations in sheep and chickens [24,25]. Furthermore, ursolic acid, the main component of saponins (terpene) in G. uralensis, alleviated lipid accumulation by activating the AMPK signaling pathway [26]. The high level of total saponins in the root extract of G. uralensis could explain the particularly low serum TG concentration in yak calves consuming this herbal extract. Further studies are needed to examine the effect of the different active ingredients of the three herbs on fat storage and mobilization.
The serum concentrations of total protein (TP) and albumin (ALB) reflect protein metabolism, mainly in the liver [27]. ALB is the most abundant serum protein, comprising 35 to 50% of the total, and is the major source of amino acids. It is important for the maintenance of homeostasis, contributing approximately 75% of the osmotic pressure of serum, for the transportation of substances, and for scavenging free radicals. The general increase in serum concentrations of TP and ALB by yak calves consuming herbal extracts could be related to polysaccharides. A previous study reported that polysaccharides of Mesona chinensis up-regulated the phosphorylation level of blood protein factors in mitogen-activated protein kinase signaling pathways [28].
Serum urea concentration did not differ among dietary treatments in the present study. The concentration of serum urea reflects protein and amino acid metabolism [29]. A decreased concentration of serum urea indicates increased nitrogen retention as a consequence of reduced protein catabolism in skeletal muscle [30]. The current results indicated that the three herbal water extracts did not affect body nitrogen metabolism.

Effect of Herbs on Serum Immunity Indices
The consumption of Astralagus polysaccharides improved immunity indices in lambs [31,32]. The concentrations of the serum immunoglobulin IgA, IgG, and IgM were all numerically higher at all time periods in the yak calves consuming the three herbal extracts than in the control calves in the present study, albeit the differences were not statistically different. An increase in these immunoglobulins enhances cell-mediated or systemic humoral immunity [33]. Herbal extracts had no effect on serum IL-2 and IL-6 concentrations; however, both decreased linearly with time in the four groups. Low air temperature can cause a decrease in inflammatory cytokines, including IL-2 and IL-6 [34]. The present study took place from autumn to winter, that is, with decreasing air temperature (Figure 1), which is, most likely, the reason for the decrease in serum IL-2 and IL-6 concentrations with time in the yak calves. IL-2 is produced by activated T-cells and stimulates immune responses and the release of IL-1 and IL-6 [35]. Consuming herbal extracts had no effect on serum TNF-α concentration in the yak calves in the present study. Serum TNF-α plays an important role in protection against invading microorganisms and in the regulation of the inflammatory functions of macrophages [31,36]. The immune response is related to the duration and amount of herbal consumption [37]. Generally, the doses of the three herbal extracts consumed by the early-weaned yak calves in the current study did not affect their immune responses.

Effect of Herbs on Serum Antioxidant Capacity
The serum concentration of MDA decreased in all groups consuming the herbal extracts when compared to control yak calves. Herbal compounds can prevent lipid peroxidation, and are used as synthetic antioxidants [38,39]. MDA is produced from the degradation of polyunsaturated fatty acids by lipid peroxidation, and is biologically active, possessing cytotoxic and genotoxic properties, which can be increased by reactive oxygen species (ROS) or other free radicals [40][41][42][43]. Consumption of an extract of dried Hibiscus sabdarrif flowers decreased MDA formation in rats, and flavonoids from natural plants inhibited MDA production [44,45], which could explain the reduction in serum MDA concentrations in yak calves consuming the three herbal root extracts in this study.
The three herbal root extracts increased serum GSH-Px concentrations in the yak calves, which is consistent with the report that a water extract of G. uralensis fed to lambs increased GSH-Px gene activities in muscle tissue [46]. The serum concentration of SOD was numerically higher in the yak calves consuming extracts from the three herbs than in the control yak calves, albeit the differences did not differ statistically. Serum GSH-Px and SOD scavenge ROS and reduce oxidative stress [47]. Administration of polysaccharides from A. sinensis increased the GSH-Px level in shrimp [48], while polysaccharides from C. pilosula increased serum GSH-Px activity in fish [49]. Therefore, the polysaccharides in the herbal water extracts could explain the increase in serum GSH-Px and SOD concentrations [50,51].

Serum Metabolome Responses
A chronic lack of glycine could retard normal growth and have adverse effects on health and immunity in mammals [52]. Glyoxylate synthesizes glycine through transamination, primarily utilizing alanine by alanine-glyoxylate aminotransferase (AGT) [53], and glyoxylate can be converted into oxalate by lactate dehydrogenase [54]. The present study indicated that the root extract of A. sinensis could reduce the activity of AGT and enhance the activity of lactate dehydrogenase. Glyoxylate can be produced through glycolate oxidation by peroxisomal glycolate oxidase, and is ultimately converted to oxalate [53,55]). C. pilosula and G. uralensis inhibited the activity of glycolate oxidase, which may be the reason for the increased glycolate content in yak calves in the present study. Flavonoids (quercetin and kaempferol) display anti-nephrolithic properties by selective interaction with glycolate oxidase, mainly through the inhibition of peroxisomal glycolate oxidase by isolated flavonoids [55]. The flavonoid, tectochrysin, in C. pilosula and G. uralensis (low content in A. sinensis) most likely inhibited glycolate oxidase [55], but this premise requires further study.
Ferulic acid was reported to increase antioxidant activity by increasing choline acetyltransferase and decreasing acetylcholinesterase activity [56], which is probably the reason for the down-regulating of choline in yak calves consuming A. sinensis in the present study. The up-regulation of citrulline in yak calves consuming A. sinensis is beneficial for the production of arginine. Nitrogen oxide is produced from arginine, and plays an important role in maintaining normal blood pressure and in scavenging pathogens in inflammatory diseases [57,58]. Anserine, the most abundant histidine-containing dipeptide in skeletal muscle in many vertebrates, reduces the risks of cardiovascular diseases [59,60]. The oxidation of N-methylhistamine by amine oxidase produces N-methylimidazole acetaldehyde, which is converted to methylimidazole acetic acid by aldehyde dehydrogenase [61]. Therefore, the components of A. sinensis regulate the two enzymes in the histidine metabolic pathway. Serum creatinine, a waste product from the breakdown of protein [62], is down-regulated with the consumption of C. pilosula. A previous study reported that polysaccharides from the stem of C. pilosula protected rats against renal injury, reduced blood creatinine levels, and improved renal glomerular function [63]. Active ingredients in the root extract of C. pilosula inhibit co-enzyme A synthesis, and affect the tricarboxylic acid cycle (TCA) cycle. The product of degraded thymidine, β-alanine, is metabolized into acetic acid [64,65], which can be diverted into pantothenic acid and co-enzyme A biosynthesis [66]. Pantothenate is a precursor of the fundamental enzyme co-factor co-enzyme A (CoA), which enters the TCA cycle [67]. The metabolome results indicated that consuming the root water extracts of A. sinensis and C. pilosula promoted energy and protein metabolism in yak calves, which is consistent with the findings of Ma et al. (2022) [68], who concluded that feeding Astragalus membranaceus root to pre-weaned dairy calves enhanced protein synthesis and gluconeogenesis.
The root extract of C. pilosula down-regulated the production of phenylalanine in the present study. Phenylalanine, via phenylalanine hydroxylase, produces tyrosine, which produces the neurotransmitter dopamine [69]. The ethanol eluent of Semen ziziphi and Radix polygalae down-regulated phenylalanine, mainly due to the sedative-hypnotic role of the flavonoid and saponin contents [70]. Glycosides of C. pilosula act as a sedative agent [71]. Thymidine, along with arachidonic acid and 2 -deoxycytidine, was identified as an indicator of cardiotoxicity, as an increase in thymidine was related to cardiotoxic drugs [72]. Thymidine can be converted into dUMP and dTMP by the catalysis of enzymes in pyrimidine metabolism. The current study down-regulated pyrimidine metabolism in yak calves consuming C. pilosula and G. uralensis root extract, which may be related to the protection of the heart and kidney.
In the current study, the up-regulating of 5-oxoproline, a product of glutathione metabolism, indicates oxidative stress, and is generally accompanied by the elevation of ascorbate [73,74], a typical antioxidant [75]. L-gulono-1,4-lactone is the intermediate in the synthesis of L-ascorbate by L-gulono-1,4-lactone oxidase [76]. Therefore, the root extract of G. uralensis enhances the antioxidant capacity of yak calves by promoting ascorbate production. It was reported that 4-hydroxyproline plays a key role in collagen stability, and in supporting muscle growth in animals [77,78]. The root extract of G. uralensis could potentially be beneficial for the growth rate of yak calves. G. uralensis root extract possesses detoxification properties by its regulatory effect on the plasma bile acid levels of mice, which up-regulates free and glycine bile acids and down-regulates taurine bile acids [79]. In the current study, yak calves consuming G. uralensis root extract down-regulated primary bile acids and cholic acid.

Conclusions
It was concluded that the root extracts of A. sinensis, C. pilosula, and G. uralensis consumed as dietary supplements by early-weaned yak calves enhanced antioxidant capacity, and increased energy and nitrogen metabolism. However, the immune-stimulating effect was not significant with the doses of the three herbal water extracts consumed by the yak calves in the current study. The three herbal root extracts up-regulated the metabolite, glycolate; A. sinensis down-regulated the metabolism of β-alanine, arginine, and histidine; and C. pilosula and G. uralensis had an inhibiting effect on pyrimidine metabolism. Further studies are needed to determine the optimal level of herb extract to be offered to the yak calves.