MiR-485 targets the DTX4 gene to regulate milk fat synthesis in bovine mammary epithelial cells

MicroRNAs (miRNAs) are mRNA suppressors that regulate a variety of cellular and physiological processes, including cell proliferation, apoptosis, triglyceride synthesis, fat formation, and lipolysis, by post-transcriptional processing. In previous studies, we isolated and sequenced miRNAs from mammary epithelial cells from Chinese Holstein cows with high and low milk fat percentages. MiR-485 was one of the significantly differentially expressed miRNAs that were identified. In the present study, the relationship between the candidate target gene DTX4 and miR-485 was validated by bioinformatics and real-time fluorescent quantitative PCR (qRT-PCR) and Western blot (WB) analyses in bovine mammary epithelial cells (bMECs). The results indicated that miR-485 negatively regulated the mRNA expression of the target gene DTX4. Furthermore, an shRNA interference vector for the target gene DTX4 was constructed successfully, and it increased the triglyceride content and reduced the cholesterol content of transfected cells. These results suggest that miR-485 may affect the contents of triglycerides (TGs) and cholesterol (CHOL) by targeting the DTX4 gene. This study indicates that miR-485 has a role in regulating milk fat synthesis and that miR-485 targets the DTX4 gene to regulate lipid metabolism in bMECs. These findings contribute to the understanding of the functional significance of miR-485 in milk fat synthesis.

Milk is secreted from the mammary glands of mammals and is a white-colored nutrient-rich liquid produced as the primary source of nutrition for infant mammals 1 . In 2011, dairy farms worldwide had approximately 730 million tons of milk from 260 million dairy cows (FAO, May 2012). According to the estimation of the FAO, in 2011, 85% of all the milk produced worldwide came from cows 2 . With the change in consumer demand, the likelihood of consuming milk fat, considered a good source of fat, has increased, and milk fat is used as an indicator of milk quality. The major constituents of milk fat are triglycerides (TGs), which comprise approximately 97-98% milk fat content. The other contents of milk fat include fatty acids (FAs) and cholesterol (CHOL) as minor components 3,4 . TGs are an important part of milk, and their synthesis in bovine mammary epithelial cells (bMECs) affects milk fat percentage 3,5 .
One of the molecular mechanisms regulating milk fat production involves the role of miRNAs in bMECs. MiRNAs are a class of non-coding RNAs that are approximately 22 nucleotides long, are fully or partially complementary to target mRNAs, and promote the degradation of target genes or the inhibition of translation, thereby enabling the regulation of target gene expression levels 6 . The expression of miRNAs has a developmental stage and tissue specificity; miRNAs are also expressed in the mammary gland and play essential roles in regulating mammary gland development, lipid metabolism, and lactation 7 . Studies have indicated that many miRNAs are involved in fat formation and lipolytic action 7 . For example, miR-224 overexpression decreased TGs synthesis and promoted apoptosis. Bta-miR-29b promoted TGs production and suppressed apoptosis 7 . Additionally, miR-33a, miR-21, miR-23a, and miR-877 have been reported to affect TGs production and apoptosis of bMECs 8 .
In a previous study, we performed miRNA sequencing on the bMECs of Chinese Holstein dairy cows with a high or low milk fat percentage 9 and selected miR-485 from the differentially expressed miRNAs to verify its role in milk fat metabolism. Previously, miR-485-5p was found to modulate mitochondrial fission by targeting mitochondrial-anchored protein ligase in cardiac hypertrophy 10,11 . MiR-485 mediated the response of human mesangial cells (HMCs) to high glucose (HG) concentrations. Overexpression of miR-485 suppressed HGinduced proliferation of HMCs 12,13 . Emerging evidence suggests that miR-485 plays a critical role in inflammation

Results
Expression trends of miR-485. Cell morphology and transfection efficiency were observed with a fluorescence microscope. The effect on fluorescence is shown in Fig. 1A-C. miR-485 mimics and inhibitor transfection was successful. After transfection of bMECs with miR-485 mimics, the expression of miR-485 detected by qPCR in the mimics group was significantly higher than that in the shNC group (*P < 0.05). Furthermore, compared with that in the shNC group, the expression of miR-485 in the inhibitor group was significantly reduced (Fig. 2).

Regulation of TGs, CHOL, and NEFA contents by miR-485 in bMECs.
Compared with that of the bMECs transfected with shNC, the TGs content of the bMECs transfected with miR-485 mimics was significantly upregulated. On the other hand, compared with that in the shNC group, the TGs content in the miR-485 inhibitor group was significantly downregulated. The above data suggest that miR-485 regulated TGs contents in bMECs (Fig. 3A). Additionally, miR-485 was found to regulate CHOL synthesis. Compared with that in the miR-485 shNC group, the CHOL content in the miR-485 mimics group was significantly downregulated (*P < 0.05). Compared with that in the miR-485 shNC group, the CHOL content in the miR-485 inhibitor group was significantly upregulated (Fig. 3B). Compared with that in the miR-485 shNC group, the NEFA content in the miR-485 mimics group was increased, and the NEFA content in the miR-485 inhibitor group was downregulated compared with that in the miR-485 shNC group (P > 0.05). (Fig. 3C).
Relative expression of DTX4. The relative mRNA expression of DTX4, a candidate target gene of miR-485, was significantly lower (**P < 0.01) in bMECs transfected with miR-485 mimics than in bMECs transfected with shNC ( Fig. 4). On the other hand, the relative mRNA expression of DTX4 in bMECs transfected with miR-485 inhibitor was significantly higher (**P < 0.01) than that in bMECs transfected with shNC ( Fig. 4). This indicated that miR-485 negatively regulated the mRNA expression level of DTX4.
Analysis of dual-luciferase activity. More than a dozen target genes related to lipid metabolism were identified by bioinformatics prediction. The TargetScan website predicted that the 3′UTR of the DTX4 mRNA contained binding sites for miR-485. Verified the miR-485 binding sites in the 3′UTR of DTX4 mRNA. The results of co-transfection showed that the luciferase activity of bMECs transfected with the DTX4-wt vector was significantly lower than that of bMECs transfected with the DTX4-si/mut vector, as shown in Fig. 5. These results confirmed that DTX4 was a target gene of miR-485.

Western blot analysis.
After analyzing the expression of DTX4 at the mRNA level, we also investigated the protein level of DTX4 to confirm the regulatory effect of miR-485 on the expression of the DTX4 gene (Fig. 6). The results suggested that the expression of DTX4 in the miR-485 inhibitor group was significantly increased compared to that of the miR-485 shNC group. The miR-485 mimics, and miR-485 shNC groups expression was lower than that of the miR-485 inhibitor group. This indicated that miR-485 mimics significantly decreased the expression of DTX4 at both the mRNA and protein levels. MiR-485 downregulated the expression of the DTX4 gene.
Expression of the DTX4 gene in bMECs after interference. By annealing the DTX4 interference sequence into the pb7SK vector, pb7SK-DTX4-shRNA interference vectors were successfully constructed (Fig. 7). The vectors were transfected into cells, and then the RNA of the cells was extracted 36 h later. The mRNA levels were analysed by reverse transcription and qPCR. qPCR detected that pb7SK-DTX4-shRNA3 had the highest interference efficiency among the three designed vectors (Fig. 8). www.nature.com/scientificreports/ and pb7SK-GFP-Neo. The TGs contents in bMECs of the pb7SK-DTX4-shRNA group were significantly increased (**P < 0.01) compared with those of the pb7SK-GFP-Neo group (Fig. 9A). However, the CHOL content in bMECs of the pb7SK-DTX4-shRNA group was significantly decreased (**P < 0.01) compared with that in bMECs of the pb7SK-GFP-Neo group (Fig. 9B).

Discussion
As the human population increases, the demand for milk has also increased. For meat supply, we have many sources, such as chicken, beef, mutton, etc., but milk for human consumption can only be obtained from female mammals, mainly from cows. Milk plays an essential role in the development of bone. Thus, scientists have put significant effort into improving the available cow breeds. Recently, the discovery of functional miRNAs has caused miRNAs to be recognized by an increasing number of scholars as important regulators of lipid metabolism in the body. MiRNAs participate in the regulation of fatty acids and metabolites in many different ways. Many miRNAs are essential to milk fat, lipid metabolism, and mammary gland development. Clusters of MECs are important components of mammary tissue that also have other biological functions. The growth and development of bMECs affect milk production capacity during lactation, and new bMECs formation is critical for maintaining tissue homeostasis 23 . Therefore, bMECs can be used as an important material for studying the mechanism of milk fat synthesis.
In our laboratory, microRNA sequencing was performed on mammary epithelial cells from Chinese Holstein cows with high or low milk fat percentages. MiR-485 was a predicted differentially expressed microRNA that may be involved in the regulation of milk fat metabolism. Bioinformatic analysis showed that the mRNA of the  www.nature.com/scientificreports/ DTX4 gene had the binding site for miR-485 in the 3′UTR. This was further verified through luciferase reporter assays. BMECs were cultured and transfected with miR-485 mimics, inhibitor, or shNC to determine the effect of modulating miR-485 in these groups. The corresponding assays were performed, including fluorescence imaging, RT-PCR, and Western blotting, to determine the cell number, the relative expression of miR-485 and its target mRNA, DTX4, and the protein expression of DTX4 in bMECs. Cell morphological analysis and fluorescence imaging showed that the transfection was successful. Furthermore, the expression of miR-485 was higher in the miR-485 mimics group than in the inhibitor and control groups, and its target mRNA, DTX4, showed the opposite expression pattern. Additionally, the levels of triglycerides and cholesterol were analysed. The results showed that the levels of TGs and CHOL were higher in the miR-485 mimics group than in the inhibitor and control groups, while the TGs and CHOL contents in the inhibitor group were lower than in the mimics and control groups. This delineates the role of miR-485 in regulating the expression of the DTX4 gene and ultimately regulating TGs and CHOL production in bMECs. DTX4 protein expression analysis via Western blot confirmed that the expression of the DTX4 protein in the miR-485 inhibitor group was significantly higher than that in the miR-485 mimics and miR-485-shNC groups. These results confirm the results of the bioinformatic analysis by revealing the expression pattern of DTX4 at the mRNA and protein levels and verify that it is a target gene of miR-485.  www.nature.com/scientificreports/ A previous study showed that miR-485-5p suppressed the invasion of cancer cells by targeting FLOT-1 in HPV-infected cervical carcinoma cells 24 . Another study showed that elevated expression of miR-485-3p suppressed ki-67 and cyclin D1 expression in NSCs. The same study also proved that miR-485-3p decreased the proliferation and induced the differentiation of NSCs by modulating TRIP6 expression. This information suggests that miR-485-3p has a vital role in the self-proliferation and differentiation of NSCs. MiR-485-3p and TRIP6 modulation may be promising therapy for treating neurodegenerative and neurogenic diseases 25 . One miRNA can regulate hundreds of genes, one gene may be affected by multiple miRNAs, and miRNAs can upregulate or downregulate gene expression by interacting with target genes. Previous studies have shown that the DTX4 gene, the adipocyte marker gene aP2, and the transcription factor PPAR-g all play important roles in the adipogenesis and differentiation of mouse 3T3-L1 preadipocytes and can promote adipogenesis and differentiation processes.  www.nature.com/scientificreports/ This hypothesis was proven by constructing stable DTX4 knockdown cells with lentivirus and non-targeted control cells. After stable knockdown of DTX4 was confirmed in the cells, a reduction in lipid droplet synthesis and a reduction in the adipogenic transcription factor PPAR-g and adipocyte marker gene aP2 were observed, thus proving that DTX4 is essential for adipogenesis in 3T3-L1 cells 26 . Peroxisome proliferator-activated receptors (PPARs) belong to the ligand-inducible nuclear hormone receptor superfamily and comprise three groups: PPARα, PPARβ/δ PPAR-g. PPAR-g improved adipogenic differentiation, lipid storage, and lipid metabolism 27 . This study provides a possible explanation of why DTX4 has a role in lipid metabolism through PPAR-g, which lays a theoretical foundation for further exploring the biological function of the DTX4 gene.  www.nature.com/scientificreports/ Our analysis includes evidence that the NEFA, TGs, and CHOL contents of bMECs are regulated by miR-485 through its effect on the DTX4 gene. This study also adds new insight into the regulatory pathway of lipid metabolism, providing simple but potentially critical future research directions.

Conclusion
The results show that miR-485 regulates the synthesis of TGs and CHOL in bMECs by targeting the DTX4 gene. This study lays the foundation for further research to clarify the molecular mechanism and provides evidence for the in-depth analysis of the function of DTX4 in milk fat synthesis.

Materials and methods
Samples and ethical approval. BMECs were provided by the Animal Science Laboratory of Jilin University, Jilin. According to previous work performed in our laboratory using the tissue nibble culture method, the bMECs in the study were purified and cultured from healthy Chinese Holstein dairy cows. Animal experiments were performed in strict accordance with the Guidelines for the Care and Use of Laboratory Animals of Jilin University (Animal Care and Use Committee permit number: SY201901007) 28,29 . This study only involved bMECs. The experimental design and procedures were performed the following Guangdong Ocean University biological experiment guidelines and regulations. Cow mammary gland tissues were collected under aseptic conditions and stored in 10× penicillin-streptomycin solution PBS buffer; the tissue was soaked in 75% alcohol for 30-60 s and then washed for 3-5 min each in 10×, 5×, and 2× PBS containing penicillin-streptomycin to remove blood and impurities from the tissue. The blood vessels and nerves visible to the naked eye in the tissue block were removed, clean tissue was collected, and the tissue was cut into pieces. bta Cell culture, total RNA extraction, and qPCR. The bMECs were cultured in a basal medium consisting of DMEM/F12 (HyClone, Grand Island, NY, USA) supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) penicillin and streptomycin (PAA, Pasching, Austria). The cells were incubated in 5% CO 2 at 37 °C. The plasmids were transfected into bMECs. Twelve hours before transfection, the bMECs were seeded in six-well cell culture plates, and transfection was performed when the cell confluence reached 80%. The cells were washed twice with PBS, and the PBS was replaced with a fresh medium. The transfection system consisted of 3 μg of plasmids, 7.5 μl of FuGENE HD Transfection Reagent (Promega Biotechnology Company, Beijing, China), and 150 μl of DMEM/F12. The transfected cells were incubated for 15 min at room temperature and then transferred to an incubator. After 36 h, the six-well plates were observed under a fluorescence microscope and photographed to evaluate the transfection efficiency.
RNA extraction was performed on the cells using TRIzol (15596-026, Invitrogen, USA). A spectrophotometer (Thermo, Nanodrop-2000) was used to determine the concentration and purity of total RNA. The concentration of the mimics was 293.1 ng/μl, and the OD was 1.80. The concentration of the inhibitor was 342.7 ng/μl, and the OD was 1.83. The RNA concentration in shNC was 189.0 ng/μl, and the OD was 1.71. After that, reverse transcription of the extracted RNA was performed using a Reverse Transcription Kit (RR047A, Takara Bio, Dalian, China), and then qRT-PCR SYBR Fluorescent Dye (RR420A, Takara Bio, Beijing, China) was used to examine the relative expression levels in the cells. All experiments were repeated three times.
Bioinformatics prediction of miR-485 target genes. The miRbase (www. mirba se. org/) software was used to select the mature sequence of miR-485 for the prediction of target genes. The target genes and binding sites of miR-485 potentially related to milk fat synthesis were predicted by TargetScan (http:// www. targe tscan. org/ vert_ 72/), and qPCR primer sequences were designed and synthesized according to miRNA primer design principles (synthesized by GENEWIZ, Suzhou, China) ( Table 1).
Luciferase reporter assay. The miR-485 binding site sequence of DTX4 was amplified by PCR for NotI (0551712, Biolabs, America) and XhoI (K2704AA, TAKARA, Beijing, China) restriction fragments. We cloned the sequences into the pmiR-RB-reporter vector to construct the pmiR-RB-report-DTX4 wt/mut/si vectors. All the sequence details of the DNA oligos are listed in Table 2. Plasmids were constructed by annealing, restriction digestion, ligation, transformation, and other methods (Endo Free Maxi Plasmid Kit, Tiangen Biotech Company, DP117, Beijing, China). The pmiR-BR-REPORT-DTX4-wt/mut/si vectors were cotransfected into the bMECs of the miR-485 mimics group. The firefly (hluc+) and Renilla (hRluc) luciferase activities were detected by the Dual-Glo Luciferase Assay System (Cat. #E1960, Promega, USA) according to the manufacturer's protocols; hluc+ luciferase was used as the reference to correct for variations in transfection efficiency, and the relative luciferase activity was calculated with the following equation: relative luciferase activity = hRluc/hluc+. Three Construction of DTX4 gene interference vectors. Three pairs of shRNA sequences were designed through the Thermo website (https:// rnaid esign er. therm ofish er. com/ rnaie xpress/), and the shRNAs were synthesized by Shenggong Biotechnology Company. The shRNA sequences are shown in Table 3. Annealed siRNA oligonucleotides were cloned into a modified pb7SK-GFP-Neo vector. Then, we selected the interference vector with the best interference efficiency.      www.nature.com/scientificreports/ Statistics and analysis. The data were analysed by GraphPad Prism 6 software. The mean and standard error (SE) of the mean were calculated from triplicate wells. P < 0.05 was considered statistically significant. For relative quantitation, the 2 −ΔΔCt method was used.