Milk metabolomics analyses of lactating dairy cows with divergent residual feed intake reveals physiological underpinnings and novel biomarkers

The opportunity to select for feed efficient cows has been limited by inability to cost-effectively record individual feed efficiency on an appropriate scale. This study investigated the differences in milk metabolite profiles between high- and low residual feed intake (RFI) categories and identified biomarkers of residual feed intake and models that can be used to predict residual feed intake in lactating Holsteins. Milk metabolomics analyses were undertaken at early, mid and late lactation stages and residual feed intake was calculated in 72 lactating dairy cows. Cows were ranked and grouped into high residual feed intake (RFI >0.5 SD above the mean, n = 20) and low residual feed intake (RFI <0.5 SD below the mean, n = 20). Milk metabolite profiles were compared between high residual feed intake (least efficient) and low residual feed intake (most efficient) groups. Results indicated that early lactation was predominantly characterized by significantly elevated levels of medium chain acyl carnitines and glycerophospholipids in high residual feed intake cows. Citrate cycle and glycerophospholipid metabolism were the associated pathways enriched with the significantly different metabolites in early lactation. At mid lactation short and medium chain acyl carnitines, glycerophospholipids and amino acids were the main metabolite groups differing according to residual feed intake category. Late lactation was mainly characterized by increased levels of amino acids in high residual feed intake cows. Amino acid metabolism and biosynthesis pathways were enriched for metabolites that differed between residual feed intake groups at the mid and late lactation stages. Receiver operating characteristic curve analysis identified candidate biomarkers: decanoylcarnitine (area under the curve: AUC = 0.81), dodecenoylcarnitine (AUC = 0.81) and phenylalanine (AUC = 0.85) at early, mid and late stages of lactation, respectively. Furthermore, panels of metabolites predicted residual feed intake with validation coefficient of determination (R ²) of 0.65, 0.37 and 0.60 at early, mid and late lactation stages, respectively. The study sheds light on lactation stage specific metabolic differences between high-residual feed intake and low-residual feed intake lactating dairy cows. Candidate biomarkers that distinguished divergent residual feed intake groups and panels of metabolites that predict individual residual feed intake phenotypes were identified. This result supports the potential of milk metabolites to select for more efficient cows given that traditional residual feed intake phenotyping is costly and difficult to conduct in commercial farms.

Effects of dietary branched-chain amino acid supplementation on serum and milk metabolome profiles in dairy cows during early lactation

The 3 branched-chain AA (BCAA), Val, Leu, and Ile, are essential AA used by tissues as substrates for protein synthesis and energy generation. In addition, BCAA are also involved in modulating cell signaling pathways, such as nutrient sensing and insulin signaling. In our previous study, dietary BCAA supplementation was shown to improve protein synthesis and glucose homeostasis in transition cows. However, a more detailed understanding of the changes in metabolic pathways associated with an increased BCAA availability is desired to fine-tune nutritional supplementation strategies. Multiparous Holstein cows (n = 20) were enrolled 28 d before expected calving and assigned to either the BCAA treatment (n = 10) or the control group (n = 10). Cows assigned to BCAA were fed 550 g/d of rumen-protected BCAA mixed with 200 g/d of dry molasses from calving until 35 DIM, whereas the cows assigned to the control were fed only 200 g/d of dry molasses. Serum samples were collected on d 10 before expected calving, as well as on d 4 and d 21 postpartum. Milk samples were collected on d 14 postpartum. From a larger cohort, we selected 20 BCAA-supplemented cows with the greatest plasma urea nitrogen concentration, as an indicator for greater BCAA availability, for the metabolomics analysis herein. Serum and milk samples were subjected to a liquid chromatography-mass spectrometry-based assay, detecting and measuring the abundance of 241 serum and 211 milk metabolic features, respectively. Multivariable statistical analyses revealed that BCAA supplementation altered the metabolome profiles of both serum and milk samples. Increased abundance of serum phosphocholine and glutathione and of milk Val, Ile, and Leu, and decreased abundance of milk acyl-carnitines were associated with BCAA supplementation. Altered phosphocholine and glutathione abundances point to altered hepatic choline metabolism and antioxidant balance, respectively. Altered milk acyl-carnitine abundances suggest changes in mammary fatty acid metabolism. Dietary BCAA supplementation was associated with a range of alterations in serum and milk metabolome profiles, adding to our understanding of the role of BCAA availability in modulating dairy cow protein, lipid, and energy metabolism on a whole-body level and how it affects milk composition.

Maternal consumption of a fermented diet protects offspring against intestinal inflammation by regulating the gut microbiota

The neonatal intestinal tract is immature and can be easily infected by pathogens causing inflammation. Maternal diet manipulation is a promising nutritional strategy to enhance the gut health of offspring. A fermented diet is a gut microbiota targeting diet containing live probiotics and their metabolites, which benefit the gut and overall health host. However, it remains unclear how a maternal fermented diet (MFD) affects neonatal intestinal inflammation. Here, in vivo and in vitro models together with multi-omics analysis were applied to investigate the impacts and the underlying mechanism through which an MFD prevents from gut inflammation in neonates. An MFD remarkably improved the performance of both sows and piglets and significantly altered the gut microbiome and milk metabolome of sows. In addition, the MFD significantly accelerated the maturation of the gut microbiota of neonates and increased the abundance of gut Lactobacillus and the microbial functions of amino acid-related enzymes and glucose metabolism on the weaning day. Notably, the MFD reduced susceptibility to colonic inflammation in offspring. The fecal microbiota of sows was then transplanted into mouse dams and it was found that the mouse dams and pups in the MFD group alleviated the LPS-induced decrease in gut Lactobacillus abundance and barrier injury. Milk L-glutamine (GLN) and gut Lactobacillus reuteri (LR) were found as two of the main MFD-induced sow effectors that contributed to the gut health of piglets. The properties of LR and GLN in modulating gut microbiota and alleviating colonic inflammation by inhibiting the phosphorylation of p38 and JNK and activation of Caspase 3 were further verified. These findings provide the first data revealing that an MFD drives neonate gut microbiota development and ameliorates the colonic inflammation by regulating the gut microbiota. This fundamental evidence might provide references for modulating maternal nutrition to enhance early-life gut health and prevent gut inflammation.

The relationship between milk metabolome and methane emission of Holstein Friesian dairy cows: Metabolic interpretation and prediction potential

This study aimed to quantify the relationship between CH₄ emission and fatty acids, volatile metabolites, and nonvolatile metabolites in milk of dairy cows fed forage-based diets. Data from 6 studies were used, including 27 dietary treatments and 123 individual observations from lactating Holstein-Friesian cows. These dietary treatments covered a large range of forage-based diets, with different qualities and proportions of grass silage and corn silage. Methane emission was measured in climate respiration chambers and expressed as production (g per day), yield (g per kg of dry matter intake; DMI), and intensity (g per kg of fat- and protein-corrected milk; FPCM). Milk samples were analyzed for fatty acids by gas chromatography, for volatile metabolites by gas chromatography-mass spectrometry, and for nonvolatile metabolites by nuclear magnetic resonance. Dry matter intake was 15.9 ± 1.90 kg/d (mean ± SD), FPCM yield was 25.2 ± 4.57 kg/d, CH₄ production was 359 ± 51.1 g/d, CH₄ yield was 22.6 ± 2.31 g/kg of DMI, and CH₄ intensity was 14.5 ± 2.59 g/kg of FPCM. The results show that changes in individual milk metabolite concentrations can be related to the ruminal CH₄ production pathways. Several of these relationships were diet driven, whereas some were partly dependent on FPCM yield. Next, prediction models were developed and subsequently evaluated based on root mean square error of prediction (RMSEP), concordance correlation coefficient (CCC) analysis, and random 10-fold cross-validation. The best models with milk fatty acids (in g/100 g of fatty acids; MFA) alone predicted CH₄ production, yield, and intensity with a RMSEP of 34 g/d, 2.0 g/kg of DMI, and 1.7 g/kg of FPCM, and with a CCC of 0.67, 0.44, and 0.75, respectively. The CH₄ prediction potential of both volatile metabolites alone and nonvolatile metabolites alone was low, regardless of the unit of CH₄ emission, as evidenced by the low CCC values (<0.35). The best models combining the 3 types of metabolites as selection variables resulted in the inclusion of only MFA for CH₄ production and CH₄ yield. For CH₄ intensity, MFA, volatile metabolites, and nonvolatile metabolites were included in the prediction model. This resulted in a small improvement in prediction potential (CCC of 0.80; RMSEP of 1.5 g/kg of FPCM) relative to MFA alone. These results indicate that volatile and nonvolatile metabolites in milk contain some information to increase our understanding of enteric CH₄ production of dairy cows, but that it is not worthwhile to determine the volatile and nonvolatile metabolites in milk to estimate CH₄ emission of dairy cows. We conclude that MFA have moderate potential to predict CH₄ emission of dairy cattle fed forage-based diets, and that the models can aid in the effort to understand and mitigate CH₄ emissions of dairy cows.

Ultra-performance liquid chromatography-quadrupole-time of flight mass spectrometry MSE-based untargeted milk metabolomics in dairy cows with subclinical or clinical mastitis

In this study, a novel metabolomics technique based on ultra-performance liquid chromatography-quadrupole-time of flight mass spectrometry in the MS^E mode was used to investigate the milk metabolomics of healthy, subclinical, and clinical mastitis cows, which were classified based on somatic cell count and presentation of clinical symptoms. Meanwhile, univariate and multivariate statistical analyses were performed to identify the significant differences across the 3 groups. Compared with healthy milk samples, less glucose, d-glycerol-1-phosphate, 4-hydroxyphenyllactate, l-carnitine, sn-glycero-3-phosphocholine, citrate, and hippurate were detected in the clinical mastitic milk samples, whereas less d-glycerol-1-phosphate, benzoic acid, l-carnitine, and cis-aconitate were found in the subclinical mastitic milk samples. Meanwhile, the milk concentration of arginine and Leu-Leu increased in both the clinical and subclinical mastitis groups. Besides, less 4-hydroxyphenyllactate, cis-aconitate, lactose, and oxoglutarate were detected in the clinical than the subclinical mastitic milk samples, whereas the abundance of some oligopeptides (Leu-Ala, Phe-Pro-Ile, Asn-Arg-Ala-Ile, and Val-Phe-Val-Tyr) increased by over 7.95-fold. Our results suggest that significant variations exist across healthy and mastitis cows. The current metabolomics approach will help in better understanding the pathobiology of mastitis, although clinical validation will be required before field application.

Milk metabolome relates enteric methane emission to milk synthesis and energy metabolism pathways

Methane (CH4) emission of dairy cows contributes significantly to the carbon footprint of the dairy chain; therefore, a better understanding of CH4 formation is urgently needed. The present study explored the milk metabolome by gas chromatography-mass spectrometry (milk volatile metabolites) and nuclear magnetic resonance (milk nonvolatile metabolites) to better understand the biological pathways involved in CH4 emission in dairy cattle. Data were used from a randomized block design experiment with 32 multiparous Holstein-Friesian cows and 4 diets. All diets had a roughage:concentrate ratio of 80:20 (dry matter basis) and the roughage was grass silage (GS), corn silage (CS), or a mixture of both (67% GS, 33% CS; 33% GS, 67% CS). Methane emission was measured in climate respiration chambers and expressed as CH4 yield (per unit of dry matter intake) and CH4 intensity (per unit of fat- and protein-corrected milk; FPCM). No volatile or nonvolatile metabolite was positively related to CH4 yield, and acetone (measured as a volatile and as a nonvolatile metabolite) was negatively related to CH4 yield. The volatile metabolites 1-heptanol-decanol, 3-nonanone, ethanol, and tetrahydrofuran were positively related to CH4 intensity. None of the volatile metabolites was negatively related to CH4 intensity. The nonvolatile metabolites acetoacetate, creatinine, ethanol, formate, methylmalonate, and N-acetylsugar A were positively related to CH4 intensity, and uridine diphosphate (UDP)-hexose B and citrate were negatively related to CH4 intensity. Several volatile and nonvolatile metabolites that were correlated with CH4 intensity also were correlated with FPCM and not significantly related to CH4 intensity anymore when FPCM was included as covariate. This suggests that changes in these milk metabolites may be related to changes in milk yield or metabolic processes involved in milk synthesis. The UDP-hexose B was correlated with FPCM, whereas citrate was not. Both metabolites were still related to CH4 intensity when FPCM was included as covariate. The UDP-hexose B is an intermediate of lactose metabolism, and citrate is an important intermediate of Krebs cycle-related energy processes. Therefore, the negative correlation of UDP-hexose B and citrate with CH4 intensity may reflect a decrease in metabolic activity in the mammary gland. Our results suggest that an integrative approach including milk yield and composition, and dietary and animal traits will help to explain the biological metabolism of dairy cows in relation to methane CH4 emission.