Experimental design
The experiment was conducted from June 11 to September 15 in 2023 at Jinyuhaoxing Agricultural and Livestock Co. Ltd. in Ningxia province. There were 40 heads of Holstein cows selected for the experiment [parity (mean ± SD) = 2.35 ± 0.49, days in milk (DIM) = 92.35 ± 6.73 d, milk yield = 45.00 ± 3.35 kg/d]. Cows were allocated into one of the two treatment groups randomly: 1) high-protein group (HP, regular diet with 17.5% of crude protein), 2) low-protein group (LP, low-protein diet with 16.0% of crude protein) + 50g/cow/d rumen-protected lysine + 30g/cow/d rumen-protected methionine. There were 20 heads of cows in each group. The whole experimental period was 97 days, including 7 days of pre-treatment period for cows to get used to the diets and 90 days of treatment period for data collection.
Total mixed rations used in the experiment and feeding management
Cows were all fed three times per day in the form of a total mixed ration (TMR) and they had free access to feed and water. The two groups (HP and LP) were fed separately. The farm was disinfected following the routine management and procedures of large-scale farms. The ingredient composition of experimental TMR is shown in Table 1. The targeted nutrient composition of TMRs fed in the experiment is shown in Table 2.
Table 1
The ingredient composition (%, DM basis) of experimental TMR.
Feed ingredient | HP1 | LP2 |
Alfalfa hay | 8.72 | 8.75 |
Whole-plant corn silage | 28.54 | 28.65 |
Flaked maize | 28.36 | 28.47 |
Soybean meal | 12.37 | 10.35 |
Rumen-protected soybean meal | 2.06 | 2.07 |
Cotton meal | 2.12 | 2.13 |
Whole cottonseed | 2.54 | 4.24 |
Beet granules | 4.04 | 4.06 |
Molasses | 2.98 | 2.99 |
Premix3 | 5.35 | 5.37 |
Fat powder | 1.35 | 1.35 |
Fatty acid calcium | 0.45 | 0.46 |
Potassium bicarbonate | 0.37 | 0.37 |
Sodium bicarbonate | 0.23 | 0.23 |
Sustained release urea | 0.51 | 0.14 |
RPLys4 | 0.00 | 0.23 |
RPMet5 | 0.00 | 0.14 |
1High protein diet with 17.5% of crude protein. |
2Low protein diet with 16.0% of crude protein + 50g/cow/d rumen protected lysine + 30g/cow/d rumen protected methionine. |
3Contains (%, DM basis): 29.33% sodium bicarbonate, 21.37% calcium carbonate, 12.82% calcium hydrogen phosphate, 8.55% potassium bicarbonate, 5.98% 54% magnesium oxide, 5.15% salt, 4.27% calcium sulfate, 4.27% potassium chloride, 2.56% 17% magnesium sulfate, 1.71% NutriTek, 1.71% MT-Bond, 0.62% Availa-4, 0.43% Yeast-Sacc, 0.30% 30% manganese sulfate, 0.16% 34% zinc sulfate, 0.16% 50% VE, 0.13% 0.2% organic selenium, 0.13% DK Sarsaponin30, 0.09% chromium propionate, 0.09% 2% biotin, 0.09% Xtract, 0.05% 1% sodium selenite, 0.03% 25% copper sulfate, 0.03% VA (500IU/g), 0.01% VD3 (500IU/g), 0.002% EDDI. |
4Rumen protected lysine, LysiGEM, Kemin Industries Inc. |
5Rumen protected methionine, MetaSmart, Adisseo USA Inc. |
Note: Premix offers (per kg DM of TMR): VA 6960 IU, VD 1840 IU, VE 41.54 IU, Fe 214.96 mg, Cu 19.3 mg, Zn 80.19 mg, Mn 102.60 mg, Se 0.52 mg, I 0.96 mg, Co 0.84 mg, Cr 0.18 mg, Biotin 0.84 mg.
Table 2
The targeted nutrient composition of TMRs fed in the experiment.
Nutrient composition | HP1 | LP2 |
NEL, MJ/kg | 6.82 | 6.91 |
CP, % | 17.50 | 16.00 |
RDP, % | 10.84 | 9.88 |
RUP, % | 6.70 | 6.14 |
MP, g/d | 2374.80 | 2335.90 |
ST, % | 28.70 | 29.03 |
NDF, % | 17.53 | 18.54 |
ADF, % | 13.65 | 14.13 |
Ca, % | 0.95 | 0.97 |
P, % | 0.48 | 0.47 |
Lys3, %MP | 6.45 | 7.38 |
Met4, %MP | 2.10 | 2.41 |
1High protein diet with 17.5% of crude protein. |
2Low protein diet with 16.0% of crude protein + 50g/cow/d rumen protected lysine + 30g/cow/d rumen protected methionine. |
3Rumen protected lysine, LysiGEM, Kemin Industries Inc. |
4Rumen protected methionine, MetaSmart, Adisseo USA Inc. |
Note: The ration was designed based on Nutrient Requirement of Dairy Cattle: Eighth Revised Edition. The contents of CP and ST were measured and the contents of NEL, MP, RDP, and RUP were calculated by AMTS.
Feed samples
The TMR samples were collected for three consecutive days at 0d, 30d, 60d, and 90d for the analysis of dry matter (DM), crude protein (CP), and acid detergent lignin (ADL). The feed of three consecutive days was composited into one sample. All composited samples were dried at 103ºC for 4h to determine DM, and ground through a 1-mm sieve before further analysis. The concentration of CP was determined by multiplying 6.25 and total N, which was determined by the combustion of samples (43). The concentration of ADL was processed based on the method of (42). The amounts of ration fed to cows and orts left every day were collected for the calculation of the total feed intake of each group. The daily total feed amount and feed ort were recorded for the calculation of DMI of each group.
Fecal samples
All 20 cows from each group were used to collect fecal samples and urine samples. The fecal samples were collected consecutive three days at d0, d30, d60, and d90. The samples taken from each time point were fully mixed and divided into two parts. One part was used for the analysis of DM and ADL, and the other part was mixed with 10% sulfuric acid (20mL of sulfuric acid per 100g of fecal sample) for nitrogen preservation. Fecal samples were frozen at -80 ºC until further nitrogen analysis. The DM was determined by drying the fecal samples at 65 ºC for 48h. The concentrations of ADL and fecal nitrogen were done similarly to feed samples.
Urine samples
Three samples of urine were collected over a 3-day period at d0, d30, d60, and d90 of the experiment, and the samples from each timepoint were composited into one sample and equally divided into two parts. One part of the urine samples was used for the analysis of creatinine using the Creatinine (Cr) Assay Kit (sarcosine oxidase, Nanjing Jiancheng Bioengineering Institute, Jiangsu, China). Sulfuric acid (10% concentration, 50%/50% w/w) was added to the other part of the urine samples for nitrogen preservation. All urine samples were frozen at -80 ºC until nitrogen analysis. A urine sample collected at 90d was also used for the analysis of metabolites (described below).
Related calculations
Total nitrogen intake = DM intake (kg) * protein concentration (%) * 1000 / 6.25
Fecal excretion (kg/d) = ADL intake (kg/d) / ADL content in fecal (kg/kg) (16XXX)
Fecal nitrogen excretion (g/d) = fecal excretion (kg/d) * 1000 * fecal DM (%) * nitrogen content in fecal (%)
Urine excretion (L/d) = body weight (kg) × 29 / content of creatinine (mg/L) (17XXX)
Urine nitrogen excretion (g/d) = urine excretion (kg/d) * 1000 * nitrogen content in urine (%)
Total nitrogen excretion (g/d) = nitrogen in fecal (g/d) + nitrogen in urine (g/d)
Nitrogen digestibility (%) = (Total nitrogen intake – fecal nitrogen excretion) / total nitrogen intake * 100%
Nitrogen deposition (g/d) = total nitrogen intake (g/d) – total nitrogen excretion (g/d)
Nitrogen utilization efficiency (%) = nitrogen deposition (g/d) / total nitrogen intake (g/d) * 100%
Metabolite extraction
The urine samples collected at 90d were thawed at 4 ºC and vortexed to uniform. A subsample of 100 µL of urine samples was mixed with 400 µL of methanol that was refrigerated to 4 ºC, and the mixture was vortexed to uniform. The mixed sample was left in an ice bath ultrasound for 20 mins and frozen at -20 ºC for 1h before being centrifuged at 16,000g and 4 ºC for 20 mins. The precipitation was discarded. The supernatant was volatized in a high-speed vacuum concentration centrifuge. The sample was mixed with 100 µL of methanol solution (50%/50%, w/w) and vortexed for 1 min before being ultrasound for 3 min in ice bath. Lastly, the sample was centrifuged at 20,000g and 4 ºC for 15 mins, and the supernatant was used for metabonomic analysis.
Chromatographic separation
The separation process was maintained at 4 ºC, and a SHIMADZU-LC30 was used for ultra-high-performance liquid chromatography (UHPLC). The column was ACQUITY UPLC HSS T3 (2.1 × 100 mm, 1.8 µ) (Waters, Milford, MA, USA). Samples of 4 µL were taken automatically. The temperature of the column was set at 40 ºC, and speed was 0.3mL/min. Mobile phase A: 0.1% of formic acid. Mobile phase B: acetonitrile. The whole UHPLC process was about 15 mins per sample.
Mass spectrometry acquisition
Positive ions (+) and negative ions (-) were detected by electrospray ionization (ESI). The samples were separated by UPLC, subjected to mass spectrometry (QE plus mass spectrometer), and ionized with a HESI source.
Data analysis
Data analysis of nitrogen metabolism was performed by SPSS statistics (version 25, IBM, New York, USA). A t-test was conducted on the data of DMI, nitrogen intake, nitrogen content in milk, fecal and urine nitrogen excretion, total nitrogen excretion, nitrogen deposition, nitrogen digestibility, and nitrogen utilization efficiency to compare the means of the two groups. The least squares mean was compared using LSD, and statistical differences were declared significant when P < 0.05, and tendencies were considered when 0.05 ≤ P ≤ 0.10.
All multivariate data were analyzed and modeled using R software packages. Principal component analysis (PCA) and partial least square analysis and discrimination (PLS-DA) were used to build the model. Variable importance for the projection (VIP) obtained by the OPLS-DA model can be used to measure the influence intensity and explanatory ability of the expression pattern of each metabolite on the classification and discrimination of each group of samples. The VIP score represents the contribution of a variable to the difference between samples of all classes. T-test in SPSS was used to calculate P-values between groups and the fold change (FC) value of each metabolite. Metabolites with VIP > 1.0 and P < 0.05 were considered statistically different. The volcano plot was made by Ggplot 2.0 and Pheatmap 1.0, and the screened metabolites different between groups were analyzed and annotated by KEGG pathway analysis (http://www.kegg.jp).