MicroRNA profiles in colostrum and transition milk of non-dairy goats fed with or without resveratrol

Abstract

MicroRNAs are non-coding RNAs that regulate gene expression at the post-transcriptional level. To aid research on support the study of microRNAs in goat milk, we provide a dataset describing the microRNA profiles of small extracellular vesicles from colostrum and transition milk of dairy goats fed either a standard diet or a diet supplemented with resveratrol. In this study, we analyzed the microRNA profiles in small extracellular vesicles (sEVs) from colostrum and transition milk of dairy goats supplemented with or without resveratrol. A total of 30 milk samples were analyzed and 406 known microRNAs were identified, with let-7c-5p (10.48% ± 0.54%) being the most abundant, followed by let-7a-5p (9.18% ± 0.47%) and let-7f-5p (8.19% ± 0.43%). The microRNA profiles were clustered into two groups (0 d and 7 d) irrespective of resveratrol supplementation. This resource enables comparative studies of microRNA composition in goat colostrum and transition milk, facilitates research into the effects of diet on milk microRNAs, and supports studies on the transmission and function of microRNAs in neonatal development.

Background & Summary

Mammalian milk contains not only nutrients, but also biologically active molecules, including proteins, lipids, vitamins, minerals, oligosaccharides, hormones and nucleic acids such as microRNAs^1,2. MicroRNAs are small non-coding RNA with 19 to 26 nucleotides that can turn down the translation of mRNA by targeting the 3′ UTR of mRNAs to regulate gene expression at the post-transcriptional level^3–6. Previous research reported that microRNA derived from milk can be enclosed in degradation-resistant extracellular vesicles, such as exosomes^7,8, which can be ingested by the sucklers and may release microRNAs⁹ to play a potential role in improving the maturity of the newborn gut and immunological system^10,11. There are lots of studies focusing on the microRNA composition in dairy milk^7,8,12,13, but fewer in goat milk^2,14, especially the research about the effect of additives on maternal goats’ milk microRNA compositions are very limited. Ma et al. reported that the composition of microRNAs in goat milk didn’t show a significant difference in colostrum and 7 days post-parturition, but the number of samples was limited². Therefore, in the present study, the effect of colostrum and transition milk phases and resveratrol supplementation on the composition of milk-derived microRNAs was analyzed using 30 goat samples. An average of 36.94 ± 2.25 (standard error) million raw data was generated for all samples. After quality control, 7.68 ± 11.2 million reads were mapped to the Capra hircus genome database (Supplementary Table 1). From the 30 samples, a total of 406 known microRNAs were identified, mainly let-7c-5p (10.48% ± 0.54%), followed by let-7a-5p (9.18% ± 0.47%) and let-7f-5p (8.19% ± 0.43%) (Fig. 1; Supplementary Table 2). The Venn diagram showed that most microRNAs (307) were expressed in all samples, and some age- and treatment-specific microRNAs were also identified (Supplementary Fig. 1; Supplementary Table 3). The PCoA plot showed no clear separation of the expression profile of sEV-associated microRNAs among groups (Fig. 2A). The sPLS-DA plot (Fig. 2B) and heatmap (Fig. 3) showed that the samples were clustered into two groups (0 d and 7 d) regardless of resveratrol supplementation.

Fig. 1.

The relative abundance (proportion of each microRNA relative to the total number of reads) of top 10 microRNAs in the small extracellular vesicle of each milk sample collected from non-dairy goats supplemented with or without resveratrol. C0, colostrum sample collected from non-dairy goats supplemented without resveratrol; C7, transition milk samples collected from non-dairy goats supplemented without resveratrol; T0, colostrum sample collected from non-dairy goats supplemented with resveratrol; T7, transition milk samples collected from non-dairy goats supplemented with resveratrol.

Fig. 2.

(A) Principal coordinate analysis (PCoA) and (B) Sparse partial least squares discriminant analysis (sPLS-DA) of microRNA in the small extracellular vesicle of colostrum and transition milk samples collected from non-dairy goats supplemented with or without resveratrol. C0, colostrum sample collected from non-dairy goats supplemented without resveratrol; C7, transition milk samples collected from non-dairy goats supplemented without resveratrol; T0, colostrum sample collected from non-dairy goats supplemented with resveratrol; T7, transition milk samples collected from non-dairy goats supplemented with resveratrol.

Fig. 3.

Heatmap of the expression of microRNA in the small extracellular vesicles of colostrum and transition milk samples collected from non-dairy goats supplemented with or without resveratrol. C0, colostrum sample collected from non-dairy goats supplemented without resveratrol; C7, transition milk samples collected from non-dairy goats supplemented without resveratrol; T0, colostrum sample collected from non-dairy goats supplemented with resveratrol; T7, transition milk samples collected from non-dairy goats supplemented with resveratrol.

Methods

Animal and sample collection

The present study was conducted at the experimental station of Mianyang Academy of Agricultural Sciences (31.2 N, 104.5 E). Eighty multiparous Beichuan goats (first parity, 10 mo of age, and average 35 kg of body weight) were divided into two groups: The control group was fed a standard total mixture (TMR, n = 40), and the treatment group was fed a standard TMR supplemented with 2 g resveratrol¹⁵ per goat per day (n = 40). Colostrum and transition milk samples were only collected from goats with twin kids, at least one of which was male. Under these criteria, a total of 30 samples were collected, including 7 colostrum (C0) and 6 transition milk samples (C7) from the control group and 9 colostrum (T0) and 8 transition milk samples (T7) from the treatment group. Colostrum and transition milk were collected within 30 min and 7 d after birth, respectively.

Isolation of small extracellular vesicles from colostrum and transition milk

The sEVs were enriched from milk samples using a combination of size exclusion chromatography (SEC) and ultrafiltration. The enrichment procedure was adapted from a previously reported method^16,17, with appropriate modifications to ensure adherence to the MISEV2023 guideline¹⁸. Briefly, the melted samples were centrifuged at 2,000 g for 20 min at 4 °C. The supernatant layer and the lower precipitate were discarded. The liquid in the middle layer was separated into a new centrifuge tube. Then the product from the last step was centrifuged at 15,000 g for 20 min at 4 °C and the supernatant was transferred to another new centrifuge tube. Acetic acid was added to the product from the previous step at a ratio of 10:1, mixed well and allowed to stand for 15 min, then centrifuged at 4 °C for 5 min at 5,000 g, and the casein accumulated at the bottom. The supernatant was passed through a 0.22-µm filter membrane, then transfer to a 15-mL 100-kDa ultrafiltration tube (UFC910096,15 mL, Amicon® Ultra, Merck Millipore, Germany) and concentrated to 1 mL, followed by SEC purification on the Exosupur^® column system. A total of 2.5 mL eluate was collected and further concentrated to approximately 200 μL using a 100 kDa molecular weight cut-off ultrafiltration tube (UFC810096, 4 mL, Amicon® Ultra,. Merck Millipore, Germany) at 4,000 × g for 2 min.

Transmission electron microscopy (TEM) with uranyl acetate negative staining

The sEVs were characterized by TEM according to the guidelines of MISEV2023¹⁸. Ten µL of sEV solution was dropped onto a copper mesh, incubated at room temperature for 10 min, then washed with sterile distilled water, and the excess liquid was absorbed with absorbent paper. Ten µL of 2% uranium peroxide acetate was dripped onto a copper mesh for 1 min for negative staining. Floating liquid was absorbed with filter paper and dried under an incandescent lamp for 2 min. The copper mesh was placed under a transmission electron microscope for observation and imaging at 80 kV.

Nano-Flow Cytometry (Nano-FCM) for characterizing particle size, concentration and protein profiling of small extracellular vesicles

The Nano-FCM were performed according to the guidelines of MISEV2023¹⁸. The particle size distribution and particle concentration of the EVs were measured using a Flow NanoAnalyzer (NanoFCM Inc., Xiamen, China)^19,20. The particle concentration was estimated from the sample. The sample was diluted to a particle concentration of 108/ml for detection with PBS. Concentration detection was performed by measuring the number of particles in fluorescent nano-silica microspheres (250 nm) with known particle concentration within a certain time. The flow rate of the sample was calculated and the number of particles in the sample was combined under the same injection pressure conditions. Particle size distribution detection was used to create a standard operating curve of scattered light intensity and particle size by using standard silica spheres (68 nm, 91 nm, 113 nm and 155 nm), which were used to determine the scattered light intensity and particle size of the sample to be tested under the same conditions into particle size of sEV. In addition, analysis of sEV surface markers (CD81 and CD9) of a random sample was conducted according to the procedures described by Wang et al.²¹.

RNA isolation

RNA isolated using the Exosome RNA Purification Kit (5202050, Simgen, China) according to the manufacturer’s instructions. In brief, 700 μL buffer TL was added to 200 μL sEV suspension and mixed uniformly by vortex oscillation. Then 100 μL of buffer EX was added, shaken for 15 seconds to mix evenly, and centrifuged at 4°C, 12000 rpm for 15 minutes. The supernatant was transferred to a 2 mL centrifuge tube and anhydrous ethanol was added in double the amount of the supernatant. The centrifuge tubes were inverted several times to mix evenly. 700 μL of the mixture was transferred to a nucleic acid purification column and centrifuged for 30 s at 12,000 rpm. The filtrate was discarded and the nucleic acid purification column was returned to a 2-ml centrifuge tube. The remaining mixture from step 3 was transferred to the nucleic acid purification column and centrifuged for 30 s at 12,000 rpm. After the filtrate was discarded and the nucleic acid purification column was placed back into a 2 mL centrifuge tube, 700 µL of buffer WA was added and centrifuged for 30 s at 12,000 rpm. The filtrate was then discarded, the nucleic acid purification column was placed back into a 2 mL centrifuge tube, 600 µL of buffer WBR was added and centrifuged at 12,000 rpm for 30 s. The filtrate was discarded and the nucleic acid purification column was placed back into a 2 mL centrifuge tube. The filtrate was discarded and the nucleic acid purification column was placed back into a 2 mL centrifuge tube and centrifuged at 14,000 rpm for 1 min. Finally, the 2 mL centrifuge tubes were discarded. The nucleic acid purification column was placed in a clean 1.5 mL centrifuge tube, 35 µL of RNase-free water was added and incubated at room temperature for 1 min and centrifuged at 12,000 rpm for 30 s to elute the RNA.

Small RNA library preparation and Next Generation RNA sequencing

For a small RNA library, the input amount of each sample was 1 ng–500 ng, and an index sequence was added. Then QIAseq miRNA Library Kit (Qiagen, Frederick, MD) was used, and a specific UMI sequence was added to each small RNA. The quality of the library was assessed using Agilent Bioanalyzer 2100 and qPCR. The index-labeled sample generated clusters in the cBot Cluster Generation System by the TruSeq PE Cluster Kitv3-cBot-HS (Illumina, San Diego, CA). Double-sided sequencing was performed for all samples in one batch on the Illumina HiSeq platform.

MicroRNA sequencing data analysis

The quality of the microRNA library was assessed using the Agilent Bioanalyzer 2100 and qPCR. Raw data were quality controlled by Echobiotech (service@echobiotech.com) through in-house Perl scripts by discarding low-quality reads, 3′-adaptors, reads with unknown base content of more than 10% or without 3′-adaptor, and reads shorter than 15 or longer than 35 bp to obtain clean reads. Then, the clean reads were mapped by Bowtie (v1.3.1)²² to the Silva, GtRNAdb, Rfam, and Repbase databases with a mismatch less than 2 to filter the rRNA, tRNA, snRNA, snoRNA, and repeat sequences, generating unannotated reads with microRNAs. After mapping the unannotated reads to the goat reference genome (ARS1; assembly GCF_001704415.1) with Bowtie (v1.3.1), which allowed ≤ 2 mismatches, reads that aligned to the genome (with ≥ 5 matching bases) and were ≥ 15 bp long were retained. These reads were deduplicated using the mapper.pl (miRDeep2 v2.0.5²³). Subsequently, potential microRNAs were identified and quantified using Quantifier.pl (miRDeep2 v2.0.5) against miRBase v22.1²⁴ with the following parameters: -g 1 (≤1 mismatch against precursors), -e 5 (consider 5 bases upstream of mature microRNA), and -W (enable probabilistic weighting). Low-abundance microRNAs (TPM < 1 in any sample of any group) were removed before statistical analysis. The differential expression analysis of the abundance of microRNAs was conducted using DESeq. 2²⁵ package in R studio (R Foundation for Statistical Computing). Significant difference was declared at false discovery rate (FDR) corrected P value ≤ 0.05. The dimension reduction method, principal coordinate analysis (PCoA), was used for unsupervised sample classification, which uses spectral decomposition to approximate a matrix of distances by the distances among samples in a few dimensions²⁶. In addition, the supervised sample classification method, sparse partial least squares-discriminant analysis (sPLS-DA), is a versatile algorithm that can be used for model prediction and description, as well as for discriminative variable selection^27,28, and it was used to classify samples with the labels C0, C7, T0, and T7. The microRNAs with significantly different expressions and high relative expression abundance (TPM >1000) were classified by cluster heatmap, rows and columns were hierarchically clustered using Ward’s method (ward.D) with Euclidean distances²⁹.

Data Records

The microRNA sequences identified in this study were deposited in the NCBI sequence read archive at https://trace.ncbi.nlm.nih.gov/Traces/?study = SRP609998 (SRR35041470-SRR35041499)³⁰. It comprises raw sequence reads of 30 goat milk samples and a metadata was provided in Supplementary Table 4.

Technical Validation

Before the collection of colostrum or transition milk, the goats were tethered and the teats were sterilized with 15 ml of 0.5% iodine disinfectant. To avoid cross-contamination, the sampler was a fixed person wearing nitrile gloves, which were changed between animals. During sampling, the first 3 drops were discarded to avoid contamination with the disinfectant. Subsequently, 30 ml of the samples collected by hand from each goat were aseptically poured into three sterile 10 ml tubes and stored in a freezer at −80 °C until further analysis. During the sampling period, all kids were kept with their dams and sucked colostrum or transition milk freely.

The morphology of milk-derived sEVs isolated from a random sample was characterized by TEM and Nano-FCM (Supplementary Fig. 2). The distribution of milk-derived sEVs detected by the Flow Nano Analyzer was positively skewed for all sample groups, with a distinct peak and a narrow range from 60 nm to 80 nm (Supplementary Fig. 2). The mean size of milk-derived sEVs was 80.28 ± 5.89 nm for C0, 71.15 ± 2.17 nm for C7, 77.42 ± 4.30 nm for T0, and 71.80 ± 1.83 nm for T7.

Author contributions

T.M. and A.Z. conceived the study. T.S., S.W., M.H. and X.Z. performed the animal trial, sample collection, and the laboratory processing of samples. T.S. and S.W. completed the bioinformatic analyses and led the writing of the paper, with all authors reviewing draft versions of the manuscript. T.M. provided funding and resources to perform the analysis.

Code availability

No custom code was used in this study.

Competing interests

The authors declared no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41597-026-07403-z.