Quantitative lipidomic analysis of blood and milk in jennies with high and low milk fat content

Highlights

• •Composition of jenny blood lipids reflected mainly diet composition, but not milk fat content.
• •Jenny milk fat depends mainly on the synthesis by mammary tissue.
• •1,2-distearoyl-sn-glycerol may be a potential marker for low milk fat content in jennies.
• •Difference in milk fat content lies mainly in triglycerides, especially C8-C18 fatty acids.

Abstract

Jenny milk is characterized by low fat content and individual variability. To test the association between blood and milk lipid profiles between jennies with high-fat (HF) and low-fat (LF) content in milk, we compared blood and milk lipid compositions between HF and LF jennies on the same feeding and management regimes. Milk fat yield and milk fat content were greater in the HF than LF group, but milk yield, lactose and protein contents, density, pH, and somatic cell count did not differ between groups. Milk fat content was correlated negatively with density and positively with pH and SCC. In blood, glycerophospholipids accounted for more than 50% of total lipids (TL), while glycerolipids accounted for more than 50% of TL in milk. The composition or concentration of TL and lipid subclasses in the blood did not differ between the HF and LF groups, with only 11 lipid species differing. In milk, the composition and concentration of TL, 17 lipid subclasses, and 131 lipid species differed between groups. The main upregulated lipid compounds were triglycerides which consisted of C16 and C18 fatty acids, while the downregulated compound was 1,2-distearoyl-sn-glycerol (DSG). These differential lipids were involved in 8 metabolic pathways, including cholesterol metabolism, vitamins digestion and absorption, GL metabolism, regulation of lipolysis in adipocytes, fat digestion and absorption, and thermogenesis. We concluded that: 1) blood lipidomic profile cannot be used to predict milk fat content and composition in jenny; 2) the synthesis of jenny milk fat is influenced primarily by mammary tissue.

1. Introduction

Jenny milk is rich in lactose, whey protein, essential fatty acids(FA) (Baloš et al., 2023), and vitamins (Vincenzetti et al., 2021). It has low cholesterol content, high content of unsaturated FA, and a low somatic cell count (SCC) (Xu et al., 2025), thus gaining widespread attention and recognition as a healthy food. The milk contains immunocompetent substances required by the human body (Plotuna et al., 2025) and, compared to cow's milk, is more easily absorbed by the humans. In addition, the composition of jenny milk is similar to that of human milk (Zhang et al., 2022) and serves as a good substitute for human breast milk (Cimmino et al., 2023; Živkov Baloš et al., 2023).

However, the fat content in jenny milk fluctuates greatly, ranging between 0.16 % and 1.3 % (Baloš et al., 2023), which renders the quality of the milk unstable and is detrimental to the jenny milk industry. According to standards established in China, the requirement for fat content in raw jenny milk is 0.2 g/100 g (T/CAAA 057–2021) or 0.25 g/100 g (DBS65/017–2023), while the milk with fat content exceeding 0.5 % is regarded as high-quality donkey milk. Similarly, donkey milk powder is considered acceptable if its milk fat content reaches 0.25 % (GB 19,644–2024). The reasons behind the low milk fat content and the great individual variations among jennies are attracting much recent research.

Lipids are complex mixtures composed of lipid species and are classified into eight major categories (Fahy et al., 2005): fatty acyls (FA), glycerolipids (GL), glycerophospholipids (GP), sphingolipids (SP), sterol lipids (ST), prenol lipids (PR), saccharolipids (SL), and polyketides (PK). Dietary lipids consist mainly of GL and lipoids, with lipoids including GP, SL and ST; whereas blood lipids consist primarily of GL, GP, FFA, free cholesterol, cholesteryl esters, SP and ceramides (Vvedenskaya et al., 2019), and exhibit species specificity (Delavaud et al., 2022). Milk lipids consist of mainly into GP, GL, SP, and FA. These lipids comprise 98 % of the non-polar lipids that are primarily triacylglycerols (TGs) in milk fat globules, and 0.5 % to1 % of the polar lipids, including GP and SP in the milk fat globule membrane (Zheng et al., 2014). The biosynthesis of milk fat consists of three stages: accumulation of fatty acids in mammary cells via de novo synthesis or absorption from the blood, assembly of triacylglycerols, and formation and secretion of milk fat globules (Osorio et al., 2016). Generally, dietary lipids serve as precursors for lipid production (Singh et al., 2023), while blood lipids participate in transport and energy production. Milk lipids are produced by this process, forming a closed-loop metabolic pathway in lactating females (Osorio et al., 2016).

Milk fat in jenny milk is influenced by factors such as season (Martini et al., 2015, 2014), breed (Garhwal et al., 2023; Wu et al., 2023), age (Barbosa dos Santos et al., 2023), diet (Liang et al., 2025; Zhang et al., 2026), rearing environment, and lactation stage (Cosentino et al., 2012; Martini et al., 2015, 2014). Studies compared lipid differences between colostrum and milk in jennies (Li et al., 2020, 2024), as well as differences in milk lipids between jennies and other mammals, including humans (Zhang et al., 2021; Zou et al., 2013). Structural characterization and concentration were determined of non-polar lipids such as TG (Chiofalo et al., 2011), polar lipids such as phospholipids and glycosphingolipids (GSLs) (Calvano et al., 2018), and ceramide and cholesterol (Contarini et al., 2017) in jenny milk. However, little is known about the reasons contributing to the low milk fat content of jenny milk and the key factors affecting composition.

Dietary factors including type of roughage (Ren et al., 2023), addition of vegetable oil (Ren et al., 2025), and dietary proportions of lipids and carbohydrates affected jenny milk fat content. We hypothesized that, under the same dietary conditions, the differences between individuals with high and low milk fat percentages may be reflected in variations in blood lipid composition. To test this hypothesis, we compared the lipid composition between blood and milk of jennies with high and low milk fat contents under the same feeding conditions. The results could provide novel insights into lipid transport and synthesis in jenny milk.

2. Materials and methods

2.1. Experimental animals and design

All the experimental procedures in this research were approved by the Science and Technology Ethics Committee of Tarim University (No. 2022,003), and these procedures are in line with the principles and provisions of human and animal (biological) science and technology ethical protection of China.

To minimize differences in the physiological and health conditions among experimental animals, 12 non-pregnant lactating hybrid jennies from Xinjiang and Dezhou breeds, similar in age (7.0 ± 1.35 years, SE), parity (2nd or 3rd), days of lactation (20 to 30 days), body condition score (5 to 6 on Pearson and Ouassat nine-point scale scoring system) (Pearson & Owassat, 2000), and body weight (225 ± 38.4 kg) were selected. All jennies were fed the same diet, and the composition and nutrients of the diet are presented in Table 1. The ratio of roughage to concentrate was 6:4.

Table 1.

Ingredients and nutrient composition of the diets (DM basis).

Item	Amount
Ingredients (g/kg)
Sudan grass	250
Wheat straw	180
Alfalfa hay	170
Corn	195
Dephenolized cottonseed protein	110
Vegetable fat powder	50
NaCl	5
Premix1	40
Nutrient levels2(g/kg)
Digestible energy (DE, MJ/kg)	10.7
DM (Dry matter)	92
Crude protein (CP)	124
Ether extract (EE)	69
Calcium	3.7
Total phosphorus	2.3
Neutral detergent fiber (NDF)	435

¹The premix provided the following per kg of diet: Ca (as calcium carbonate) 8 g, Fe (as ferrous sulfate) 250–6 000 mg, Cu (as cupric oxide) 150–620 mg, Zn (as zinc sulfate) 900–3 000 mg, Mn (as manganese sulfate) 70–2 500 mg, Se (as sodium sulfate) 2.5–10 mg, I 2.5–10 mg, V_A 175–250 IU, V_D3 20–45 IU, V_E 500 IU, V_K3 12.5–75 mg, V_B1 25 mg, V_B2 25 mg, V_B6 12.5 mg, V_B12 0.12 mg, pantothenic acid 125 mg, nicotinic acid 125 mg, and folic acid 12.5 mg.

²DE was calculated according to the National Research Council (National Research Council of the national Academies, 2007), while the other parameters were measured.

After 14 days of feeding, the jennies were divided into two groups of six each based on the milk fat content and milk fat yield: high milk fat group (HF, with milk fat content > 0.4 % and milk fat yield > 5 g/day), and low milk fat group (LF, with milk fat content < 0.2 % and milk fat yield < 3 g/day).

2.2. Feeding and milking management

Each jenny and her foal were maintained in an individual stall throughout the experiment. The jennies were offered feed at approximately 3 % of their body weight per day (6.0 to 7.5 kg/d) in equal amounts at 10:00 and 17:00. The amount of feed was adjusted every three days to ensure that the remains were less than 5 %. Drinking water was freely available.

The jennies were milked in the morning from 9:00 to 10:00 using a pulsating mechanical milking machine with a vacuum pressure of 50 kPa and pulsation rate of 80 cycles/minute (Model HL-G1, Xinjiang Jufeng Animal Husbandry Machinery Equipment Co., Ltd., Wulumuqi, China). After milking, the jennies were penned with their foals and nursed. At 13:00, the jennies were separated from their foals and milked at 18:00. After this second milking, the jennies were penned and nursed their foals until 23:00 when they were separated until the next morning milking.

2.3. Sample collection and analysis

2.3.1. Measurement of body weight and feed intake

Feed remains were weighed daily before morning feeding, and feed samples and remains were dried at 105 °C for 48 h to calculate the dry matter intake (DMI). The jennies were weighed on d 1 and on d 28 before morning feeding to calculate average daily weight change.

2.3.2. Blood sample collection

On d 29 before morning feeding, jugular vein blood samples were collected into EDTA anticoagulant vacutainers. The samples were centrifuged at 1000 x g for 10 min, and the plasma was aliquoted into 1.5 mL centrifuge tubes, snap-frozen in liquid nitrogen, and stored at -20 °C. Two blood samples from each group were lost and, therefore, there were 4 samples for each jenny group for blood measurements.

2.3.3. Determination of milk yield and composition

On d 28, milk yields of the HF and LF jennies were measured during the morning and afternoon milkings, and 50 mL from each milking were combined. One aliquot of the milk sample was used for determining milk composition and somatic cell count (SCC), while the other aliquot was snap-frozen in liquid nitrogen and stored at -20 °C for later analysis.

Milk fat, lactose, protein, pH, and non-fat solids were measured using a calibrated milk composition analyzer (Model UL 40BC, Hangzhou Zhejiang University Youchuang Technology Co., Ltd., Hangzhou, China). The pH was measured using a calibrated pH meter (Model FE28-standard, Mettler Toledo, China), and SCC was determined by a somatic cell counter (Lactoscan SCC, Milkotronic Ltd., Nova Zagora, Bulgaria). Based on the milk yield and time separated from the foal, the 24-hour corrected milk yield was calculated using the following formula:

$$\text{Corrected}\text{milk}\text{yield}\left[\text{kg}/\left(\mathrm{d}·\text{head}\right)\right]=\text{milking}\text{yield}/\text{separation}\text{time}\times 24\mathrm{h}.$$

2.3.4. Quantitative lipidomics analysis

The blood and milk samples were stored at -20 °C for quantitative lipidomic analysis (MetWare Biotechnology Co., Ltd, Wuhan, China). The blood samples from the HF and LF groups were designated as HMF and LMF, and the milk samples were designated as MHMF and MLMF, respectively.

2.3.4.1. Sample pretreatment

After thawing, samples were vortexed for 10 s, and 20 μL of milk or 50 μL of blood were transferred into a tube with 1 mL of extraction solvent (methyl tert‑butyl ether: methanol = 3:1, v/v) containing internal standard lipids (see Supplementary Table 1). After shaking the mixture for 15 min, 200 μL of ultrapure water were added, vortexed for 1 min, centrifuged at 10,600 × g for 10 min at 4 °C, and 200 μL of the upper organic layer were collected and evaporated using a vacuum concentrator. The dry extract was dissolved in 200 μL reconstituted solution (acetonitrile: isopropanol =1:1, v/v).

2.3.4.2. Chromatography and mass spectrometry

The sample extracts were analyzed using an LC-ESI-MS/MS system (UPLC, ExionLC AD, https://sciex.com.cn/; MS, QTRAP® 6500+ System, https://sciex.com/). The analytical conditions were as follows, UPLC: column, Thermo Accucore™C30 (2.6 μm, 2.1 mm×100 mm i.d.); solvent system, A: acetonitrile/water (60/40,V/V, 0.1 % formic acid, 10 mmol/L ammonium formate), B: acetonitrile/isopropanol (10/90 VV/V, 0.1 % formic acid, 10 mmol/L ammonium formate); gradient program, A/B(80:20, V/V)at 0 min, 70:30 V/V at 2.0 min, 40:60 V/V at 4 min, 15:85 V/V at 9 min, 10:90 V/V at 14 min, 5:95 V/V at 15.5 min, 5:95 V/V at 17.3 min, 80:20 V/V at 17.3 min, 80:20 V/V at 20 min; flow rate, 0.35 ml/min; temperature, 45 °C; injection volume: 2 μL. The effluent was connected alternatively to an ESI-triple quadrupole-linear ion trap (QTRAP)-MS.

Linear ion trap (LIT) and triple quadrupole (QQQ) scans were acquired on a triple quadrupole-linear ion trap mass spectrometer (QTRAP), QTRAP® 6500+ LC-MS/MS System, equipped with an ESI turbo ion-spray interface, operating in positive and negative ion modes and controlled by Analyst 1.6.3 software (Sciex). The ESI source operation parameters were as follows: ion source, turbo spray; source temperature 500 °C; ion spray voltage (IS) 5500 V (positive), -4500 V (negative); ion source gas 1 (GS1), gas 2 (GS2), curtain gas (CUR) were set at 45, 55, and 35 psi, respectively. Instrument tuning and mass calibration were performed with 10 and 100 μmol/L polypropylene glycol solutions in QQQ and LIT modes, respectively. QQQ scans were acquired as multiple reaction monitoring (MRM) experiments with collision gas (nitrogen) set to 5 psi. DP and CE for individual MRM transitions were done with further DP and CE optimization. A specific set of MRM transitions were monitored for each period according to the metabolites eluted within the period.

2.3.4.3. Lipid qualification and quantification

Qualitative analysis was based on a self-built database (MetWare Database), using the retention time (RT) of the detected substances and parent-daughter ion pair information. Lipid quantification was accomplished via the MRM mode of triple quadrupole mass spectrometry. After obtaining the lipid mass spectrometric data of different samples, peak area integration was determined for all chromatographic peaks of the substances, and quantitative analysis used the internal standard method. Mass spectrometric data were processed using Analyst 1.6.3 software. The integrated peak area ratios of all detected samples were substituted into the following formula to calculate the actual concentration of the target substance in the sample:

$$\text{Concentration}\text{of}\text{target}\text{lipid}\text{in}\text{sample}\left(\text{nmol}/\text{mL}\right):X=0.001\times R\times c\times F\times V/\mathrm{m}$$

where X = concentration of target lipid in sample (nmol/mL); R = peak area ratio of the target substance to the internal standard; c = concentration of internal standard (μmol/L); F = internal standard correction factor for different types of substances; V = volume of sample extraction solution (μL); m = volume of sample taken (mL).

2.4. Data processing and statistical analysis

The lipidome data were unit variance scaled before unsupervised principal component analysis (PCA) using statistics function prcomp within R (www.r-project.org). Differential metabolites were determined by variable importance in the projection (VIP, > 1) and P-value (< 0.05, Student’s t-test). VIP values were extracted from Orthogonal partial least squares-discriminant analysis (OPLS-DA), which also provided score plots and permutation plots, and generated by R package (MetaboAnalystR). The data were log transformed (log2) and mean centered before OPLS-DA. To avoid overfitting, a permutation test (200 permutations) was performed. Identified metabolites were annotated using KEGG compound database (http://www.kegg.jp/kegg/compound/), and annotated metabolites were mapped to KEGG pathway database (http://www.kegg.jp/kegg/pathway.html). Pathways with significantly regulated metabolites were mapped and fed into metabolite sets enrichment analysis (MSEA) Significance was determined by hypergeometric test’s P-values.

Data on milk yield and milk composition were analyzed using SPSS 19.0 software (SPSS, Chicago, IL, USA). An independent samples t-test was applied, and the partial correlation analysis was conducted with body weight and age as control factors. A two-tailed P-value < 0.05 was accepted as significant. Data are presented as mean ± standard error (SE), unless specified otherwise.

3. Results

3.1. Body weight and feed intake

Body weight, body weight change, and DMI did not differ between HF and LF jennies (Table 2).

Table 2.

Body weights and dry matter intake over 28 days of jennies with high and low milk fat content.

Index	Groups1		P-value2
	HF	LF
Body weight (kg)
Initial weight	240.0 ± 18.33	209.2 ± 10.49	0.176
Final weight	241.0 ± 17.21	213.6 ± 12.17	0.223
Weight gain (g/d)	35.7 ± 100.3	157.1 ± 75.0	0.365
Dry matter intake (kg/d)	7.0 ± 0.50	6. 3 ± 0.37	0.301

¹HF=high milk fat jennies, LF= low milk fat jennies.

3.2. Milk yield and composition

As designed, the milk fat yield and milk fat content were greater (P < 0.05) in the HF than LF group (Table 3). However, corrected milk yield, lactose content, density, milk protein content, pH, and SCC did not differ between groups.

Table 3.

Milk yield and composition of jennies with high and low milk fat content.

Index	Group1	Mean	SD	CV	P-value2
24 h milk yield(g/d)	HF	1958	301.8	37.75	0.180
	LF	1475	105.2	17.47
Milk fat content ( %)	HF	0.42	0.08	44.04	0.004
	LF	0.13	0.01	22.42
Milk fat yield (g/d)	HF	7.28	0.70	23.64	<0.001
	LF	1.97	0.24	29.59
Non-fat solid ( %)	HF	9.10	0.14	3.78	0.745
	LF	9.16	0.11	2.86
Milk lactose content ( %)	HF	6.66	0.08	2.89	0.720
	LF	6.70	0.06	2.13
Density (g/mL)	HF	1.032	0.001	0.17	0.505
	LF	1.033	0.001	0.10
Milk protein content ( %)	HF	1.76	0.05	7.18	0.790
	LF	1.78	0.04	5.46
pH	HF	7.29	0.08	2.71	0.150
	LF	7.14	0.05	1.74
Somatic cell count(×103 /mL)	HF	57.3	28.34	121.08	0.398
	LF	28.0	17.40	152.25

¹HF=high milk fat jennies, LF= low milk fat jennies.

Milk fat content was correlated negatively (P = 0.013) with density and positively (P < 0.05) with pH and SCC (Table 4). Non-fat solids were correlated positively with lactose content, density, and milk protein content (P < 0.01), and negatively with pH (P = 0.013). Lactose, milk protein, and density were correlated positively (P < 0.01) with each other, while pH was correlated positively (P = 0.01) with SCC. Milk fat content and milk yield were not correlated.

Table 4.

Correlation coefficients between milk production and milk composition in jennies.

	24 h corrected milk yield	Fat content	Non-fat solid	Lactose content	Density	Protein content	pH	Somatic cell count
24 h corrected milk yield	1.000	-0.192	0.114	0.150	0.322	0.092	-0.309	-0.375
Milk fat content		1.000	-0.423	-0.474	-0.749*	-0.355	0.745*	0.696*
Non-fat solid			1.000	0.998**	0.896**	0.997**	-0.745*	-0.530
Milk lactose content				1.000	0.922**	0.990**	-0.780**	-0.568&
Density					1.000	0.860**	-0.914**	-0.762**
Milk protein content						1.000	-0.701*	-0.480
pH							1.000	0.895**
Somatic cell count								1.000

^&0.05< P < 0.10, *P < 0.05, **P < 0.01.

3.3. Quality control of lipidomics analysis

The total ion chromatogram (TIC) (Fig. 1A) displayed high overlap, indicating good instrument stability. High correlation of quality control (QC) samples (Fig. 1B), over 95 % of substances with coefficient of variation (CV) values < 0.3 in QC samples (Fig. 1C), and PC1 scores of QC samples within ±2 standard deviations (Fig. 1D) collectively indicated the reliability of the data. The tight clustering of QC samples (Fig. 1E) further confirmed the repeatability of the experiment.

Fig. 1.

Quality control of samples.

HMF: blood samples from high-fat jennies; LMF: blood samples from low-fat jennies; MHMF: milk from high-fat jennies; MLMF: milk from low-fat jennies. QC: quality control. A: total ion flow chart; B: QC sample correlation analysis; C: coefficient of variation profile; D: PCA scoring figure; E: PC1 scores control charts.

3.4. Quantitative analysis of different lipid classes

A total of 935 lipid species were detected in the studies (Table 5), classified into 5 classes: FA, GL, GP, SP, and ST. GP and GL were the most abundant classes, accounting for over 80 % of total lipids (TL). The number of lipid compounds varied across sample types: approximately 790 lipid species in blood and 500 species in milk were identified. In blood, GP accounted for over 50 %, and GL for 27 % of TL; in milk, GL accounted for greater 50 % and GP for approximately 30 % of TL.

Table 5.

Number and proportion of lipid classes or subclasses in jenny blood and milk.

Category1	Class	all samples	HMF2	LMF	MHMF	MLMF
Class
Total		935	791	790	511	488
GL		298	214	215	269	259
GP		465	427	428	156	153
SP		82	72	72	32	27
FA		69	57	55	47	43
ST		21	21	20	7	6
GL %		31.9	27.1	27.2	52.6	53.1
GP %		49.7	54.0	54.2	30.5	31.4
SP %		8.77	9.10	9.11	6.26	5.53
FA %		7.38	7.21	6.96	9.20	8.81
ST %		2.25	2.65	2.53	1.37	1.23
Subclass
TG	GL	268	195	196	248	240
PE	GP	78	69	68	36	36
PC	GP	72	69	69	17	17
PI	GP	60	57	58	22	22
PG	GP	46	38	38	18	17
PE-O	GP	39	39	39	12	12
CAR	FA	34	32	31	16	16
LPC	GP	32	30	30	12	12
FFA	FA	31	21	20	28	23
PS	GP	31	24	24	15	15
PA	GP	25	23	24	4	4
PE-P	GP	25	25	25	4	4
DG	GL	22	13	13	16	14
Cer-NS	SP	18	17	17	6	6
SM	SP	16	15	15	5	5
HexCer-AP	SP	14	13	13	8	7
LPE	GP	11	9	9	5	3

¹CAR: Acylcarnitines; Cer-NS: Ceramides - Non-Hydroxy Fatty Acid & SphingoBase; DG: Diacylglycerols; FA: Fatty Acyls; FFA: Free Fatty Acids; GL: Glycerolipids; GP: Glycerophospholipids; HexCer-AP: Hexosylceramides - Amino Phosphate; LPC: Lysophosphatidylcholines; LPE: Lysophosphatidylethanolamines; PA: Phosphatidic Acids; PC: Phosphatidylcholines; PE: Phosphatidylethanolamines; PE-O: Phosphatidylethanolamines with Ether Linkage; PE-P: Phosphatidylethanolamines with Vinyl Ether Linkage; PG: Phosphatidylglycerols; PI: Phosphatidylinositols; PS: Phosphatidylserines; SM: Sphingomyelins; SP: Sphingolipids; ST: Sterol Lipids; TG: Triacylglycerols.

²HMF: blood samples from high-fat jennies; LMF: blood samples from low-fat jennies; MHMF: milk from high-fat jennies; MLMF: milk from low-fat jennies.

At the subclass level (Table 5), TG was the most abundant, with 195 and 196 species detected in the HMF and LMF groups, respectively, and 248 and 240 species in MHMF and MLMF groups. TG accounted for 24.6 % of TL in HMF (Fig. 2A), 24.8 % in LMF (Fig. 2B), 48.5 % in MHMF (Fig. 2C), and 49.2 % in MLMF (Fig. 2D). Phosphatidylethanolamines (PE) were the second most abundant class (7.0 % to 8.7 %).

Fig. 2.

The number proportion of subclass lipids in jenny blood and milk.

A(HMF): blood samples from high-fat jennies; B(LMF): blood samples from low-fat jennies; C(MHMF): milk from high-fat jennies; D(MLMF): milk from low-fat jennies.

BA: Bile Acids; CAR: Acylcarnitines; CASE: Sitosterol Esters; CE: Cholesteryl Esters; Cer-AP: Ceramides - Amino Phosphate; Cer-AS: Ceramides - Amino Sugar; Cer-NDS: Ceramides - Non-Dihydroxy SphingoBase; Cer-NP: Ceramides - Non-Phosphorylated; Cer-NS: Ceramides - Non-Hydroxy Fatty Acid & SphingoBase; CerP: Ceramide-1-Phosphate; DG: Diacylglycerols; DG-O: Diacylglycerols with Ether Linkage; DGDG: Digalactosyldiacylglycerols; FFA: Free Fatty Acids; Hex2Cer: Dihexosylceramides; HexCer-AP: Hexosylceramides - Amino Phosphate; HexCer-NS: Hexosylceramides - Non-Hydroxy Fatty Acid & SphingoBase; LNAPE: N-Acyl-Lysophosphatidylethanolamines; LPA: Lysophosphatidic Acids; LPC: Lysophosphatidylcholines; LPC-O: Lysophosphatidylcholines with Ether Linkage; LPE: Lysophosphatidylethanolamines; LPE-P: Lysophosphatidylethanolamines with Vinyl Ether Linkage; LPG: Lysophosphatidylglycerols; LPI: Lysophosphatidylinositols; MG: Monoacylglycerols; PA: Phosphatidic Acids; PC: Phosphatidylcholines; PC-O: Phosphatidylcholines with Ether Linkage; PE: Phosphatidylethanolamines; PE-O: Phosphatidylethanolamines with Ether Linkage; PE-P: Phosphatidylethanolamines with Vinyl Ether Linkage; PG: Phosphatidylglycerols; PI: Phosphatidylinositols; PMeOH: Phosphatidylmethanols; PS: Phosphatidylserines; SHexCer: Sulfated Hexosylceramides; SM: Sphingomyelins; SPH: Sphingosines; TG: Triacylglycerols; TG-O: Triacylglycerols with Ether Linkage.

3.5. Concentration analysis of different lipid classes

The TL concentration was calculated by adding the concentrations of all quantified lipid compounds in one sample. There was no difference in TL concentration between HMF and LMF groups (Table 6), while the TL concentration of MHMF was greater (P = 0.05) than that of MLMF group. There was no difference in TL concentration between blood and milk samples.

Table 6.

Concentrations of class and subclass lipids in jenny blood and milk.

			Treatment1 (nmol/mL)						P-value
Category2	HMF	LMF	MHMF	MLMF	Blood	Milk	SEM	HMF vs LMF	MHMF vs MLMF	Blood vs milk
TL	4991	5220	10,800	3823	5106	7024	871.0	0.624	0.050	0.215
Class
FA	548	488	1068	1239	518	1153	71.1	0.516	0.249	< 0.01
GL	705	686	9216	2252	696	5734	757.6	0.887	0.044	0.010
GP	2684	2832	411	279	2758	345	163.1	0.741	0.217	< 0.01
SP	139	153	32.7	23.0	146.1	27.8	9.58	0.622	0.048	< 0.01
ST	916	1060	72.6	28.8	988	50.7	68.47	0.530	0.035	< 0.01
Subclass
BA	8.36	9.40	17.36	14.56	8.88	16.49	1.997	0.648	0.419	0.005
CAR	0.50	0.48	8.38	8.84	0.49	8.47	0.493	0.798	0.743	< 0.01
CASE	18.3	24.1	ND	ND	21.2	ND	1.896	0.342	-	< 0.01
CE	684	784	5.90	3.11	734	4.45	56.950	0.602	0.008	< 0.01
Cer-AP	0.063	0.062	0.258	0.119	0.062	0.187	0.0165	0.948	0.019	0.002
Cer-AS	0.098	0.081	0.383	0.369	0.089	0.378	0.0161	0.153	0.752	< 0.01
Cer-NDS	0.015	0.015	ND0	ND0	0.015	ND0	0.0009	0.899	-	< 0.01
Cer-NP	0.024	0.026	0.175	0.172	0.025	0.178	0.0143	0.681	0.930	< 0.01
Cer-NS	2.33	1.96	0.63	0.46	2.15	0.55	0.201	0.582	0.054	0.001
CerP	89.6	101.3	ND	ND	95.4	ND	9.76	0.693	-	< 0.01
Cholesterol	205	243	49.4	11.1	224	28.1	16.56	0.313	0.074	< 0.01
DG	95.7	99.1	903	530	97.4	681	46.88	0.722	0.017	< 0.01
DG-O	106	130	3.28	1.88	118	2.41	7.05	0.260	0.010	< 0.01
DGDG	ND	ND	1.40	0.58	ND	1.03	0.260	-	0.430	0.123
Eicosanoid	2.92	2.73	12.4	12.7	2.83	12.6	1.329	0.788	0.969	< 0.01
FFA	545	485	1047	1218	515	1121	79.7	0.530	0.254	< 0.01
Hex2Cer	0.042	0.029	ND	ND	0.036	ND	0.0030	0.207	-	< 0.01
HexCer-AP	1.82	1.59	13.5	14.5	1.70	13.8	0.509	0.128	0.425	< 0.01
HexCer-NS	1.82	1.61	0.86	0.24	1.71	0.54	0.231	0.673	0.076	< 0.01
LNAPE	7.90	8.54	27.6	8.42	8.22	17.9	2.644	0.646	0.065	0.070
LPA	2.33	2.01	2.01	5.28	2.17	3.70	0.900	0.584	0.218	0.344
LPC	51.5	47.4	4.73	2.33	49.4	2.45	3.895	0.688	0.311	< 0.01
LPC-O	0.367	0.365	0.030	0.031	0.366	0.031	0.0155	0.964	0.846	< 0.01
LPE	7.27	6.78	4.90	0.80	7.03	1.63	1.059	0.731	0.110	0.021
LPE-P	0.375	0.355	ND	ND	0.365	ND	0.0257	0.791	-	< 0.01
LPG	0.060	0.060	0.027	0.029	0.060	0.028	0.0028	0.976	0.460	< 0.01
LPI	0.197	0.230	0.128	0.052	0.213	0.076	0.0167	0.233	0.026	< 0.01
MG	211	210	543	605	211	577	16.3	0.884	0.160	< 0.01
PA	81.1	67.7	11.52	8.45	74.4	9.84	10.488	0.610	0.690	0.001
PC	2185	2344	87.0	37.8	2264	60.7	153	0.720	0.067	< 0.01
PC-O	12.6	17.7	ND	ND	15.1	ND	1.24	0.214	-	< 0.01
PE	27.6	25.8	22.4	7.81	26.7	14.9	3.37	0.722	0.059	0.027
PE-O	23.0	24.4	2.53	0.72	23.7	1.54	1.514	0.741	0.066	< 0.01
PE-P	44.4	47.6	30.6	28.5	46.0	29.6	2.01	0.550	0.016	< 0.01
PG	21.3	23.7	7.98	6.65	22.5	7.32	1.789	0.561	0.321	< 0.01
PI	85.9	108	30.2	14.4	96.8	20.8	6.37	0.179	0.029	< 0.01
PMeOH	10.5	10.5	0.08	0.08	10.52	0.08	0.054	0.923	0.888	< 0.01
PS	122	97.2	179	158	110	166	26.95	0.087	0.740	0.076
SHexCer	0.197	0.244	ND	ND	0.220	ND	0.0265	0.585	-	0.001
SM	43.1	46.1	16.9	7.12	44.6	10.8	2.71	0.549	0.035	< 0.01
SPH	0.025	0.027	ND	ND	0.026	ND	0.0008	0.453	-	< 0.01
TG	292	247	7765	1115	269	4208	700.6	0.692	0.046	0.022
TG-O	0.064	0.087	0.013	0.002	0.075	0.008	0.0116	0.348	0.433	< 0.01

¹HMF: blood samples from high-fat jennies; LMF: blood samples from low-fat jennies; MHMF: milk from high-fat jennies; MLMF: milk from low-fat jennies.

²BA: Bile Acids; CAR: Acylcarnitines; CASE: Sitosterol Esters; CE: Cholesteryl Esters; Cer-AP: Ceramides - Amino Phosphate; Cer-AS: Ceramides - Amino Sugar; Cer-NDS: Ceramides - Non-Dihydroxy SphingoBase; Cer-NP: Ceramides - Non-Phosphorylated; Cer-NS: Ceramides - Non-Hydroxy Fatty Acid & SphingoBase; CerP: Ceramide-1-Phosphate; DG: Diacylglycerols; DG-O: Diacylglycerols with Ether Linkage; DGDG: Digalactosyldiacylglycerols; FA: Fatty Acyls; FFA: Free Fatty Acids; GL: Glycerolipids; GP: Glycerophospholipids; Hex2Cer: Dihexosylceramides; HexCer-AP: Hexosylceramides - Amino Phosphate; HexCer-NS: Hexosylceramides - Non-Hydroxy Fatty Acid & SphingoBase; LNAPE: N-Acyl-Lysophosphatidylethanolamines; LPA: Lysophosphatidic Acids; LPC: Lysophosphatidylcholines; LPC-O: Lysophosphatidylcholines with Ether Linkage; LPE: Lysophosphatidylethanolamines; LPE-P: Lysophosphatidylethanolamines with Vinyl Ether Linkage; LPG: Lysophosphatidylglycerols; LPI: Lysophosphatidylinositols; MG: Monoacylglycerols; PA: Phosphatidic Acids; PC: Phosphatidylcholines; PC-O: Phosphatidylcholines with Ether Linkage; PE: Phosphatidylethanolamines; PE-O: Phosphatidylethanolamines with Ether Linkage; PE-P: Phosphatidylethanolamines with Vinyl Ether Linkage; PG: Phosphatidylglycerols; PI: Phosphatidylinositols; PMeOH: Phosphatidylmethanols; PS: Phosphatidylserines; SHexCer: Sulfated Hexosylceramides; SM: Sphingomyelins; SP: Sphingolipids; SPH: Sphingosines; ST: Sterol Lipids; TG: Triacylglycerols; TG-O: Triacylglycerols with Ether Linkage; TL: total lipids.

The GP accounted for the greatest proportion in blood samples, whereas GL accounted for the greatest proportion of TL in milk samples, followed by FA. The concentrations of FA, GL, GP, SP, and ST did not differ between the HMF and LMF groups, while the concentrations of GL, SP, and ST were greater (P < 0.05) in MHMF than in MLMF.

Of the 44 lipid subclasses, the five most abundant in blood were phosphatidylcholines (PC), cholesteryl esters (CE), free fatty acids (FFA), TG, and monoacylglycerols (MG). In milk, the top five lipid classes were TG, FFA, diacylglycerols (DG), MG, and PC. There was no difference in the concentrations of any lipid subclass between HMF and LMF groups. However, the concentrations of CE, ceramide-AP (Cer-AP), DG, oxidized diacylglycerols (DG-O), lysophosphatidylinositols (LPI), phosphatidylethanolamines-P (PE-P), phosphatidylinositols (PI), sphingomyelins (SM), and TG differed (P < 0.05) between MHMF and MLMF groups. The concentrations of 41 out of the 44 subclasses differed (P < 0.01) between MHMF and MLMF groups.

3.6. Analysis of differential lipid compounds

Based on PCA (Fig. 3A), there were differences in lipid composition between blood and milk. The OPLS-DA results (Fig. 3B–D) further demonstrated distinct differences between HMF and LMF groups, and between MHMF and MLMF groups. Among the 935 detected lipid species, 11 differential lipid compounds (DLC) were identified between HMF and LMF groups, with 10 belonging to the GP class (6 downregulated and 5 upregulated) (Fig. 3E, Supplementary Table 2). PS (18:2_20:2) was the most abundant upregulated compounds, while PI (16:0_18:1) was the most abundant downregulated lipid. These DLC featured carbon chain lengths ranging from 18 to 40 and degree of unsaturation of 1 to 5 double bonds.

Fig. 3.

Differential lipid analysis.

HMF: blood samples from high-fat jennies; LMF: blood samples from low-fat jennies; MHMF: milk from high-fat jennies; MLMF: milk from low-fat jennies. A: PCA plot of 4 groups; B: OPLS-DA plot of 4 groups; C: OPLS-DA plot of HMF_vs_LMF; D: OPLS-DA plot of MHMF_vs_MLMF; E:S-plot of HMF_vs_LMF differential lipids); F: S-plot of differential lipid between MHMF_vs_MLMF; G: Carbon chain length distribution of differential lipids between MHMF_vs_MLMF; H: unsaturated double bonds of differential lipids between MHMF_vs_MLMF.

Between MHMF and MLMF groups, 131 DLC were identified (5 downregulated and 126 upregulated) (Fig. 3F, Supplementary Table 3). Among these DLC, 78 species (59.5 %) belonged to the TG subclass in the GL class, followed by 32 species (24.4 %) from the GP class (including PI, PC, PE). The most abundant downregulated compound was DG (18:0_18:0). Upregulated DLC predominantly had carbon chain lengths of 30 to 40 and 50 to 60 (Fig. 3G) and degree of unsaturation of 0–3 double bonds (Fig. 3H).

Further screening of DLC (VIP > 1.2, fold change > 4, P〈 0.05, and average concentration of DLC in MHMF group 〉 10 nmol/mL), 26 key DLC were identified, with 25 TG (Table 7) and 1 DG (21:0_16:0). These DLC of TG subclass were characterized by even-carbon FA ranging from C8 to C18.

Table 7.

Concentrations of differential lipid compounds in jenny milk.

						Treatment1 (nmol/mL)
Compounds	Class2	Subclass	Molecular Weight	Ionization model	Formula	MHMF	MLMF	SEM	VIP	Fold Change	P-value3
TG (16:0_18:1_18:1)	GL	TG	858.8	Positive	C55H102O6	240.8	37.2	27.20	1.49	6.47	0.007
TG (16:0_16:0_18:1)	GL	TG	832.8	Positive	C53H100O6	167.5	22.9	19.23	1.48	7.33	0.008
TG (16:0_16:0_16:0)	GL	TG	806.7	Positive	C51H98O6	38.5	8.3	4.06	1.47	4.64	0.007
TG (16:0_16:0_16:1)	GL	TG	804.7	Positive	C51H96O6	30.9	3.4	4.00	1.41	9.08	0.014
TG (16:1_18:1_18:1)	GL	TG	856.8	Positive	C55H100O6	76.1	5.7	10.49	1.39	13.31	0.016
TG (16:0_18:0_18:2)	GL	TG	858.8	Positive	C55H102O6	24.9	3.0	3.37	1.35	8.23	0.017
TG (16:0_16:1_18:1)	GL	TG	830.7	Positive	C53H98O6	371.6	44.1	51.86	1.35	8.42	0.019
TG (16:0_18:1_18:2)	GL	TG	856.8	Positive	C55H100O6	296.4	41.7	41.95	1.33	7.11	0.020
TG (18:1_18:1_18:1)	GL	TG	884.8	Positive	C57H104O6	98.3	9.6	15.10	1.30	10.23	0.025
TG (14:0_16:0_16:0)	GL	TG	778.7	Positive	C49H94O6	26.7	6.4	3.53	1.30	4.18	0.023
TG (18:0_18:1_18:2)	GL	TG	884.8	Positive	C57H104O6	13.5	1.5	2.04	1.29	8.91	0.024
DG (21:0_16:0)	GL	DG	638.6	Positive	C40H78O5	548.5	92.4	85.28	1.29	5.94	0.022
TG (16:0_18:0_18:1)	GL	TG	860.8	Positive	C55H104O6	48.2	4.6	7.32	1.29	10.50	0.025
TG (16:1_18:1_18:2)	GL	TG	854.7	Positive	C55H98O6	149.8	15.2	23.39	1.27	9.85	0.029
TG (18:1_18:1_18:2)	GL	TG	882.8	Positive	C57H102O6	140.0	16.2	22.09	1.26	8.66	0.030
TG (16:1_16:1_18:1)	GL	TG	828.7	Positive	C53H96O6	101.7	10.8	15.97	1.26	9.43	0.031
TG (14:0_16:0_18:1)	GL	TG	804.7	Positive	C51H96O6	126.6	19.5	18.57	1.26	6.51	0.030
TG (16:0_16:1_16:1)	GL	TG	802.7	Positive	C51H94O6	69.4	8.5	10.70	1.25	8.12	0.032
TG (16:0_18:2_18:2)	GL	TG	854.7	Positive	C55H98O6	109.7	8.9	17.88	1.25	12.29	0.033
TG (12:0_14:0_18:0)	GL	TG	750.7	Positive	C47H90O6	15.6	3.4	2.26	1.25	4.65	0.029
TG (14:0_18:0_18:2)	GL	TG	830.7	Positive	C53H98O6	57.1	6.3	9.30	1.22	9.09	0.036
TG (16:0_16:1_18:2)	GL	TG	828.7	Positive	C53H96O6	147.7	16.0	24.57	1.22	9.26	0.037
TG (14:0_14:1_16:1)	GL	TG	746.6	Positive	C47H86O6	52.0	5.4	8.85	1.21	9.64	0.041
TG (10:0_16:0_18:1)	GL	TG	748.7	Positive	C47H88O6	266.3	46.4	42.65	1.20	5.74	0.039
TG (8:0_14:0_18:0)	GL	TG	694.6	Positive	C43H82O6	61.0	9.9	9.85	1.20	6.17	0.040
TG (12:0_12:0_18:0)	GL	TG	722.6	Positive	C45H86O6	15.3	3.3	2.43	1.20	4.56	0.038

¹MHMF: milk from high-fat jennies; MLMF: milk from low-fat jennies.

²GL: Glycerolipids; TG: Triacylglycerols.

3.7. KEGG and enrichment analysis of differential lipids

After KEGG and enrichment analysis of the DLC, GP metabolism (Fig. 4A) was the primary metabolic pathway affected in the blood. In contrast, between MHMF and MLMF groups, 8 pathways, including insulin resistance, cholesterol metabolism, vitamin digestion and absorption, GL metabolism, regulation of lipolysis, fat digestion and absorption, and thermogenesis were identified and upregulated (Fig. 4B).

Fig. 4.

KEGG pathway enrichment analysis of differential lipids.

The vertical axis represents the Rich Factor. The size of the dots indicates the number of differential lipids enriched in each pathway. The larger the dot, the greater the number of lipids. The color of dots reflects the magnitude of the P-value. The closer they are to red, the smaller the P-value. HMF: blood samples from high-fat jennies; LMF: blood samples from low-fat jennies; MHMF: milk from high-fat jennies; MLMF: milk from low-fat jennies. A: HMF_vs_LMF; B: MHMF_vs_MLMF.

4. Discussion

The present study is the first to compare blood and milk lipids based on quantitative lipidomic analysis between lactating jennies with low and high milk fat content. The relationship between blood and milk lipids of HF and LF jennies is important for revealing the key factors and potential biomarkers of jenny milk lipid synthesis, and for seeking strategies to increase the milk fat content and yield. The relatively small sample size was a limitation of this study, however, selecting jennies with similar age, body weight, parity, lactation stage, and body condition and controlling diets and management ensured that the genetic and metabolic differences were reflected.

4.1. Milk yield and composition

The milk fat content in this study exhibited a high variability among all components, whereas the contents of non-fat solids, lactose, and milk protein were relatively stable, which is consistent with reports from other donkey breeds in Brazil (Barbosa dos Santos et al., 2023) and Italy (Martemucci & D'Alessandro, 2012). The body weight, body weight gain and DMI were did not differ between the HF and LF jennies, which indicated that the feed allowance (3 % of body weight) and nutritional level (124 g/kg crude protein and 10.7 MJ/kg digestible energy) met their energy and nutritional needs during the first 2 months of lactation. These results indicated that the higher percentage of fat content in the milk of HF jennies could be attributed to genetic potential and the regulation of gene expression related to lipid synthesis and not due to body fat mobilization.

In the present study, milk yield varied considerably, but was not correlated with milk composition (e.g., milk fat, lactose), pH, density, or SCC. In dairy cows, the heritability of milk yield is 0.25, with 75 % of its variation attributed to feeding management, environment, and physiological status (Collier et al., 2017). It is theorized that the factors affecting milk yield in jennies are similar to those in dairy cows. Population surveys and genetic selection for high milk yield and fat content could possibly reduce the large individual variations in milk production and composition among jennies.

Milk pH values from ruminants are slightly acidic (6.5 to 6.7), whereas from humans and equids (e.g., horses, donkeys) are slightly alkaline (7.0–7.5) (Aspri et al., 2017). In the present study, the pH value of the jenny milk ranged from 7.1 to 7.3, which is consistent with previous reports for jennies (Barbosa dos Santos et al., 2023). The SCC serves as an indication of the health of dairy animals, hygiene conditions of farms and the quality and safety of dairy products (Alhussien & Dang, 2018). SCC is affected by factors including the productivity of the dairy animal, health status, parity, lactation period, and feeding management (Mondini et al., 2025). Although the SCC of the jennies in the present study varied considerably, it remained within the range reported in previous studies (Barbosa dos Santos et al., 2023). The SCC of raw cow milk should be ≤ 400,000 cells/mL (Ministry of Agriculture and Rural Affairs of China, 2021), whereas, of raw donkey milk should be SCC ≤ 100,000 cells/mL, indicating that the SCC of jenny milk is lesser than cow milk. In the present study, the average SCC of jenny milk was 42,000 cells/mL, which met the requirements of the standard (China Animal Husbandry Association, 2021). In addition, there were positive correlations between jenny milk SCC and milk fat content, density, and pH value. Future studies should examine the underlying mechanisms balancing SCC and milk fat content in donkey breeding programs.

In the present study, non-fat solids was correlated positively with lactose, milk protein, and density, which was attributed primarily to the relatively high contents of lactose and protein in jenny milk (Baloš et al., 2023), as these two components are key factors influencing milk density. Additionally, pH was correlated positively with density, non-fat solids, fat, lactose, and protein, which suggests that pH may be used to predict the solid content of jenny milk.

4.2. Blood lipidome

A total of 935 lipid species belonging to five classes, namely, FA, GL, GP, SP and ST, were detected in the present study. Approximately 790 lipid species were identified in blood, while 500 were detected in milk, indicating that the number of lipid compounds in blood is greater than in milk. GP and GL were the two most abundant lipids in the blood and milk, exceeding 80 % of the total lipids. In the jenny blood, the percentage of the GP class accounted for over 50 % of the total lipids, and TG was most abundant (24.6 % to 24.8 %), followed by PE and PC. For lipid concentration in blood, PC was highest, followed by CE, FFA, TG, and MG. In comparison, lipid concentrations in human blood consists primarily of CE, SM, lysophosphatidylcholines (LPC), FFA, and PC (Hornburg et al., 2023), while in dairy cow blood consists primarily of CE, PC, and SM, with moderate concentrations of TG and FFA (Liu et al., 2025). This variation in blood lipid composition may be attributed to such factors as species, nutrition, and physiology.

The concentrations of 5 lipid classes and 47 subclasses did not differ between blood from high-fat and low-fat jennies. For specific lipid compounds, only 11 DLC were identified, and these were associated mainly with GP metabolism. Studies on ruminants (Delavaud et al., 2022) and rats (Pomar et al., 2020) concluded that blood lipids are influenced strongly by diet. The uniform dietary consumed by the two jenny groups could explain the results in the present study. Therefore, the concentrations and composition of blood lipids did not reflect jenny milk fat content.

4.3. Milk lipidome

The precursors for milk fat synthesis are derived from blood, while the biosynthesis of milk fat consists of three stages: accumulation of fatty acids in mammary cells via de novo synthesis or absorption from the blood, assembly of triacylglycerols, and formation and secretion of milk fat globules (Osorio et al., 2016). Unlike the blood lipids, TG percentage in jenny milk accounted for 48.5 % to 49.2 %, nearly twice the proportion of TG in blood (24.6 to 24.8 %), which indicates that the jenny mammary gland can selectively uptake TG from circulation. For lipid concentrations, jenny milk was rich in TG, FFA, DG, MG, and PC, with moderate levels of SM and LPC. Ren et al. (2025) detected 725 lipid species belonging to 21 subclasses in jenny milk, among which TG, DG, and PC were the three most dominant, with TG accounting for 56.7 % of the total. The dietary composition in the present study included feed ingredients such as fresh alfalfa, Sudan grass, wheat straw, corn, and cottonseed protein. In contrast, Ren et al. (2025) fed the jennies mainly wheat straw, corn, wheat bran, and soybean meal, and supplemented different vegetable oils, which were reported to affect the levels of DG, SP, and acylhexosylceramides (AcHex) in milk. This difference in dietary composition may explain the difference in lipid species between the present study and Ren et al. (2025). These results suggest that the milk fat composition of jenny milk partly reflects their feed intake. In addition, there were 23 more lipid compounds in the milk of high-fat than low-fat jennies, which warrants further study.

The concentrations of 3 lipid classes, 9 subclasses, and 131 DLC differed between milk from high-fat and low-fat jennies. Among these, the GL class, and TG and DG subclasses were influential, which aligned with the differences in milk fat content. In the 25 key upregulated TG compounds in the milk of the high-fat jennies, the FA compositions consisted entirely of even-carbon FAs ranging from C8 to C18, with C16 FA and C18 FA being the most abundant. Moreover, the primarily upregulated DG (21:0_16:0) also contained C16 FA. In contrast, there were minimal differences in odd-carbon FA and FA with carbon chain lengths > 18, which are derived primarily from the diet. In mammals, C16 FA and C18 FA are the end products of de novo lipogenesis and FA chain elongation, respectively (Yu et al., 2024).

Milk fat percentage of livestock is influenced by the genetic make-up (Singh et al., 2023), and the regulation of gene expression during lactation in the process of FA uptake and transport, TG synthesis, and secretion of lipid droplets (Yu et al., 2024). The origin of FA in TG is from de novo synthesis in mammary glands and absorption from circulating blood (Tian et al., 2022), and these FA may further undergo desaturation or chain elongation. Under uniform diet and similar blood lipidome, we believe that the difference in fat content and composition of jenny milk in the present study was due primarily to de novo lipogenesis, desaturation and elongation by the mammary gland, which depends on the expression of key enzyme genes, such as acetyl-CoA carboxylase (ACC), fatty acid synthase (FAS), elongase and stearoyl-CoA dehydrogenase (Osorio et al., 2016). We surmise that the uptake and transport of blood lipids, de novo lipogenesis, desaturation and elongation determine the final composition and content of milk fat in jennies.

The key downregulated DLC in the jenny milk was DG (18:0_18:0), that is, 1,2-distearoyl-sn-glycerol (DSG). As a type of 1,2-diacylglycerols (DAGs) localized on the plasma membrane, DSG acts as a second messenger to activate protein kinase C (PKC) (Lyu et al., 2020), thereby regulating cellular functions such as signal transduction, metabolism, and secretion (Kolczynska et al., 2020). DAG inhibits fat accumulation and promotes oxidation (Hibi et al., 2009), which may explain the lesser milk fat content in these jennies. Thus, DSG has the potential to serve as a marker for predicting milk fat content. In addition, DAG combines with fatty acyl groups to form triacylglycerols under the action of diacylglycerol acyltransferases 1 (DGAT1) (Yu et al., 2024). The accumulation of DSG may be associated with individual differences in DGAT1 gene expression among donkeys, which requires further verification.

4.4. Main differential pathways in blood and milk

There is a lack of studies that directly examines the connection between blood and milk lipidome in mammals. Blood lipidome encompasses lipids from organs and tissues, and is commonly used to indicate the health status of the animal (Hornburg et al., 2023). Milk lipids reflect the energy metabolism, nutrient allocation, and endocrine status in the mammary gland (Li et al., 2020). Milk fat synthesis involves a complex metabolic network, and is regulated by such factors as genetics, physiology, nutrients, and hormones, including prolactin and insulin (Osorio et al., 2016; Yu et al., 2024). In the present study, the differential metabolic pathways between milk from HF and LF jennies included insulin resistance, cholesterol metabolism, vitamin digestion and absorption, GL metabolism, regulation of lipolysis in adipocytes, fat digestion and absorption, and thermogenesis. These differences collectively were due to differences in energy utilization efficiency, lipid synthesis ability, and related metabolic regulatory capacity in mammary glands between HF and LF jennies during lactation.

The metabolic pathway of jennies in the present study differed from jennies supplemented with vegetable oil (Ren et al., 2025) or fed with roughages (Ren et al., 2023). Vegetable oil supplementation influenced mainly GP and GL metabolism, while the roughages affected mainly of GL, SP and linoleic acid. As this study used a uniform diet, we concluded that genetic differences were the primary reason for the differences in milk fat content. In line with this reasoning, the differential pathway in blood between HF and LF jennies was only for GP metabolism, which are the main lipid components of cellular membranes to maintain cellular structure and function (Osawa et al., 2024). This indicated that the physiological and health statuses of the jennies were similar when fed the same diet, and that the differences in milk fat could be attributed mainly to the metabolism in mammary gland tissue. Genetic diversity and expression regulation of lipogenic genes in mammary gland tissue should be the focus of future research to improve milk quality.

5. Conclusions

This is the first study that compared milk yield, milk composition, and blood/milk lipidomic profiles between jennies with different milk fat contents under identical dietary and environmental conditions. Milk fat content, yield, and SCC exhibited high variability among jennies, and a positive correlation emerged between milk fat content and SCC. The composition and concentration of blood lipidome in high-fat and low-fat jenny milk differed slightly. However, 131 DLC were identified between high-fat and low-fat jenny milk, consisting mainly of C16 FA and C18 FA. The fasting blood lipid profile cannot be used to predict the content and composition of milk fat, but DSG has the potential to act as a marker for milk fat synthesis capacity. These differences in milk lipidome were regulated by multiple pathways, including cholesterol metabolism, vitamin digestion and absorption, GL metabolism, regulation of lipolysis in adipocytes, fat digestion and absorption, and thermogenesis. For developing specialized dairy donkey breeds, the genetic polymorphism of key genes such as DGAT1, FAS and ACC, and the metabolic regulation for milk lipogenesis should be studied. The DSG could potentially be used as a key biomarker to characterize milk fat synthesis capacity in jenny, but this should be verified.

Data availability

Data will be made available on request.

Ethical approval

All the experimental procedures in this research were approved by the Science and Technology Ethics Committee of Tarim University (No. 2022003), and these procedures are in line with the principles and provisions of human and animal (biological) science and technology ethical protection of China.

CRediT authorship contribution statement

Le-tian Zhang: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. Ke Zhang: Methodology, Investigation, Formal analysis. Qiong-xian Yan: Supervision, Project administration, Methodology, Conceptualization. Liang Deng: Writing – review & editing, Resources. Allan Degen: Writing – review & editing, Conceptualization. Li-lin Liu: Methodology, Data curation. Xiao-ling Zhou: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.