Phosphoproteome Reveals the Lactation Mechanism in Caprine Mammary Gland Tissues in Peak and Late Lactation Stages

Background: Protein phosphorylation plays an important role in lactation. Differentially modied modication sites between peak lactation (PL, 90 days postpartum) and late lactation (LL, 280 days postpartum) were investigated using an integrated approach, namely, liquid chromatography with tandem mass spectrometry (LC-MS/MS) and tandem mass tag (TMT) labelling, to understand the molecular biological mechanisms in goat breast tissues. Results: A total of 1,938 (1,111 up-regulated, 827 down-regulated) differentially modied modication sites of 1,172 proteins were identied (P values < 0.05 and fold change of phosphorylation ratios > 1.5). In addition, the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis showed that ribosome (chx03010), RNA transport (chx03013), protein export (chx03060), calcium signalling pathway (chx04020), oxytocin signalling pathway (chx04921), RNA degradation (chx03018) and MAPK signalling pathway (chx04010) were enriched in relation to energy metabolism and protein translation. The results of western blot showed phosphorylation levels of ACACA, EIF4EBP1 and IRS1 increased and JUN decreased in PL compared with LL. The result was consistent with phosphoproteome. Conclusions: Overall, these data indicate that protein phosphorylation is closely related to lactation and differentially modied modication sites might have potential research value in the regulation of goat lactation. PLIN4: perilipin-4 isoform DFFA: IRS1: substrate 1; PRKAA: 5′-AMP-activated catalytic subunit alpha-2; RPS6KB: ribosomal protein S6 kinase beta-1 isoform X1; EIF4EBP1: eukaryotic translation initiation factor 4E-binding protein 1; TSC2: tuberin; JUN: transcription factor AP-1.


Background
Phosphorylation is one of the important post-translational modi cations of proteins; it is related to many activities of life, such as signal transduction, gene expression, cell cycle and cell apoptosis [1].
Esteri cation takes place as the side chain of protein amino acid is added to a phosphoric acid group with a strong negative charge, thereby changing the structure, activity and ability to interact with other molecules, affecting many biological processes. In eukaryotes, phosphorylation occurs primarily in serine, threonine and tyrosine [2].
Goat milk is very popular with the improvement of people's living conditions and the unique nutritional and health value of goat milk. Therefore, improving milk production and quality of dairy goat is the focus of our work. Dynamic changes in protein phosphorylation occur in goat mammary epithelial cells during PL period and LL period [3,4]. Studies have shown that protein phosphorylation regulates lactation in animals [5,6]. Although phosphoproteome analysis was conducted on the oocyte of goat [7] and the heart, liver of mice [8,9], phosphorylated proteomic analysis has not been performed in PL and LL stages of dairy goat mammary gland tissues. The regulation mechanism of phosphorylation protein to lactation synthesis is not well known.
Therefore, in this experiment, differentially modi ed modi cation sites of phosphorylated protein between PL and LL in goat breast tissues are identi ed by tandem mass tag (TMT)/isobaric tags for relative and absolute quantitation (iTRAQ) labelling, high-performance liquid chromatography (HPLC) fractionation, a nity enrichment and liquid chromatography with tandem mass spectrometry (LC-MS/MS) analysis. A total of 1,938 (1,111 up-regulated, 827 down-regulated) differentially modi ed modi cation sites of 1,172 proteins were identi ed and further expore the Gene Ontology (GO) and Kyoto Encyclopedia of Genes (KEGG) for differentially phosphorylated proteins. Phosphorylation levels of ACACA, EIF4EBP1, IRS1 increased and JUN decreased in PL compared with LL, the date showed that the results of phosphoproteome were reliable. The present study provided essential information to enhance the knowledge of expression dynamics of phosphorylated proteins during PL and LL and good reference data for further studying effect of protein phosphorylation to lactation in the mammary gland.

Results
Sample repeatability test and mass spectrometry quality control detection Differentially modi ed modi cation sites were analysed by LC-MS/MS and TMT between PL and LL of goat. The work ow of the present study is shown in Fig. 1a. The results of the sample repeatability test are shown in Fig. 1b. Figure 1b is a thermal chart drawn by calculating Pearson's Correlation Coe cient between two pairs of samples. Pearson's Correlation Coe cient showed that the reproducibility of the sample was good. The mass error was detected at 0 as the centre axis and concentrated in the range of less than 10 ppm. The length of most peptide segments was between 8 and 20 amino acid residues, which indicated that the preparation of the sample was standard in our study (Fig.S1).

Analysis of differentially phosphorylated proteins
A quantitative study of differentially modi ed modi cation sites was performed in this experiment by an integrated approach involving LC-MS/MS and TMT labelling. All samples were prepared in triplicate. A total of 6,979 phosphorylation sites were identi ed on 2,608 proteins, of which 5,901 phosphorylation sites of 2,454 proteins contained quantitative information. In the classi ed groups, the variation of the difference expressed was more than 1.5 times and less than 0.67 times the change criterion of signi cant up-regulation or signi cant down-regulation (P values < 0.05). A total of 1,938 (1,111 up-regulated and 827 down-regulated) differentially modi ed modi cation sites of 1,172 proteins were changed ( Fig. 2; Table S1).
To understand the phosphorylated proteins that were identi ed and quanti ed, such phosphorylated proteins and differentially modi ed modi cation sites were annotated in detail from the aspects of GO, protein domain, KEGG pathways and subcellular localisation (Table S2). To calculate the rule of amino acid sequence in the differentially modi ed modi cation sites, the amino acid sequence was detected before and after all phosphorylation sites in the samples. The results revealed that most quanti ed phosphorylated proteins had remarkable enrichment near the modi er sites (Fig.S2, Table S3).

Functional enrichment of differentially phosphorylated proteins
Proteins that contained differentially modi ed modi cation sites were analysed by GO enrichment analysis. From the analysis of the results, up-regulated phosphorylated site corresponding proteins were primarily enriched to biological processes related to protein translation and energy metabolism, such as 'translation', 'organic substance transport', 'protein localisation' and 'Golgi vesicle transport'. For molecular function, up-regulated phosphorylated sites corresponding proteins were primarily enriched to 'transporter activity' and 'structural constituent of ribosome', which were also related to translation and transportation. Similarly, up-regulated phosphorylated sites corresponding proteins were primarily enriched to organelles related to translation proteins in cellular component, such as 'ribosome', 'endoplasmic reticulum part' and 'intracellular ribonucleoprotein complex'. In biological processes, down-regulated phosphorylated proteins were primarily enriched in catabolic process of mRNA and protein translation, such as 'nucleartranscribed mRNA catabolic process', 'mRNA catabolic process' and 'nuclear-transcribed mRNA catabolic process, deadenylation-dependent decay'. In molecular function, 'enzyme inhibitor activity' and 'peptidase regulator activity' were enriched. In addition, in cellular components, down-regulated phosphorylated proteins that related to cytoskeleton were enriched, such as 'cytoskeleton', 'polymeric cytoskeletal bre' and 'intermediate lament cytoskeleton' (Fig. 5).
KEGG enrichment analysis revealed that proteins that contained up-regulated phosphorylated sites were associated with 'ribosome', 'RNA transport' and 'protein export' pathways. Proteins that contained downregulated phosphorylated sites were remarkably enriched in 12 pathways (Fig. 6), such as 'MAPK signaling pathway', 'RNA degradation', 'Oxytocin signaling pathway' and 'Calcium signaling pathway'.
Protein domain enrichment analysis revealed that 'Spen paralogue and orthologue SPOC, C-terminal' and 'Zinc nger, C2H2-like' were enriched in proteins that contained up-regulated phosphorylated sites. For proteins that contained down-regulated phosphorylated sites, 14 protein domains were enriched (Fig. 7).

Validation of changes in phosphorylated proteins level by western blot
Phosphorylation levels of ACACA, EIF4EBP1, IRS1 and JUN were tested by western blot in PL and LL. The result showed phosphorylation levels of ACACA, EIF4EBP1 and IRS1 increased and JUN decreased in PL compared with LL (Fig. 8).

Discussion
Protein phosphorylation regulated many cellular processes in eukaryotic cells by post-translational modi cation [10]. Protein phosphorylation revealed the effect of temperature on wheat leaf and spikelet [11] and the importance of protein phosphorylation for different chloroplast functions [12]. Protein phosphorylation had an important role in sperm quality [13] and an important connection with lactation [5,14]. The level of phosphorylation was constantly changing in breast tissues throughout the development of the gland or different stages of lactation [3]. In addition, the level of phosphorylated proteins was detected by LC-MS/MS and TMT in PL and LL. These results contribute to the elucidation of the mechanism underlying the regulation of lactation in goat breast tissues by phosphorylation.
The production of milk was increasing after parturition and subsequently reached the maximum [15]. These phenomena were dependent on a series of factors, including somatotropin, insulin and functional protein. Nevertheless, the effect of protein phosphorylation on goat breast tissues during lactation remained unclear. Therefore, the differentially phosphorylated proteins were explored in goat breast tissues. A total of 1,938 differentially modi ed modi cation sites of 1,172 proteins were found between PL and LL. The western blot result of phosphorylation levels of ACACA, EIF4EBP1, IRS1 and JUN showed that the results of phosphoproteome were reliable and these results were consistent with this studies [16,17]. Previous studies showed APOC3, FASN and ACACA proteins were highly correlated with the decomposition and utilisation of fat [18][19][20]. Moreover, the inactivation of PLIN4 could decrease fat accumulation [21]. The phosphorylation levels of mTOR, RPS6KB and STAT5B were closely related to the synthesis of milk fat and proteins [22][23][24][25]. SLC2A1 was the primary glucose transporter of glucose metabolism [26]. The effect of DFFA was closely related to apoptosis [27]. Insulin regulated glucose uptake and metabolism by binding to an insulin receptor [28], insulin receptors (IRS1 and IRS2) were necessary for insulin to function. The mTOR pathway was a key regulation pathway of lactation. PRKAA, IRS1, RPS6KB, EIF4EBP1, TSC2, FASN and ACACA were located in the mTOR pathway, previous studies showed that activation of mTOR pathway could enhance the synthesis of lactose and triglyceride [29,30]. As expected, the result of this study showed phosphorylation levels of these proteins (FASN, ACACA, APOC3, SLC2A1, FLNA, IRS1, PLIN4 and DFFA) were changed in different stages of lactation, so we speculate that the changes of these proteins phosphorylation levels may be related to lactation.
The energy metabolism of breast tissues increased remarkably during lactation [31]. GO enrichment showed several differentially phosphorylated proteins enriched in biological processes, which were related to energy metabolism, material transport and protein translation. The results provided an overview of dynamic changes in lactation. A series of signalling pathways was enriched in PL compared with LL, such as 'Ribosome' (chx03010), 'RNA transport' (chx03013), 'Protein export' (chx03060), 'Calcium signalling pathway' (chx04020), 'Oxytocin signalling pathway' (chx04921), 'RNA degradation' (chx03018) and 'MAPK signalling pathway' (chx04010). As such, the results showed that protein translation, energy metabolism and hormone regulation were essential for lactation and these KEGG pathways might play important roles during lactation. In summary, GO and KEGG analyses provided a better understanding of the cellular components, molecular functions and biological processes of differentially phosphorylated proteins in goat mammary gland epithelial tissues and provided a reference for future research.
Conclusions the present study was the rst to sequence phosphorylated proteomics in mammary tissues of dairy goats and this study markedly increased the current understanding of post-translational modi cations of proteins events that occur during the progression of lactation. The western blot result of phosphorylation levels of ACACA, EIF4EBP1, IRS1 and JUN showed that the results of phosphoproteome were reliable. FASN, ACACA, APOC3, SLC2A1, FLNA, IRS1, PLIN4 and DFFA proteins as part of the results might have potential research value in regulating goat lactation and we will study the function of these proteins in lactation in the future. So, the present study provided valuable information for the further study of the molecular mechanism of lactation in goats.

Mammary gland tissue collection and cell culture
Guanzhong dairy goats were obtained from Longxian Goat Breeding Center, in Longxian County of Shaanxi Province, China. All procedures in our animal study were approved by the Animal Care and Use Committee of the Northwest A&F University (Yangling, China) (permit number: 17-347, data:2017-10-13).
Breast tissues were collected at the LL (280 days postpartum) and PL (90 days postpartum) periods from 18 healthy Guanzhong dairy goats (3 year olds) by surgery (9 breast tissues in PL and 9 breast tissues in LL). First of all, one side of the mammary gland of the dairy goat was wiped clean with alcohol, and then local anesthesia was performed on the mammary tissue. A 5-6-cm incision was made and the skin pulled back using sterile forceps, exposing the mammary tissue. 1 cm 2 samples were taken using a sterile scalpel blade and forceps. In order to stop any external bleeding, pressure was applied with sterile gauze to stop any external bleeding. The incision was closed with 6 to 8 surgical staples (#89063337, Appose ULC Skin Stapler, 35 wide; Henry Schein Inc., Melville, NY). After that, Keep the goat in a clean and hygienic place, the skin around the site of wound was sterilized with iodine tincture and 75% alcohol, and assign special person to take care of it. All collected tissues were washed with Rnase-free PBS and then frozen in liquid nitrogen and stored at -80 °C.

Protein extraction
The mammary gland epithelial tissues of dairy goats in PL and LL were grinded into cell powder in the presence of liquid nitrogen. The cell powder was transferred to a 5 mL centrifuge tube. Subsequently, the cell powder was added to four volumes of lysis buffer (8 M urea, 1% Protease Inhibitor Cocktail, 2 nM EDTA), followed by thrice of ultrasonic treatment on ice with a high-intensity ultrasonic processor (Scientz, Ningbo, China). The remaining debris was removed by centrifugation at 12,000 g at 4 °C for 10 min. Finally, the supernatant was collected, and the protein concentration was quanti ed using the BCA kit (Scientz, Ningbo, China). Proteins were extracted from 18 breast tissues of dairy goats that included nine breast tissues in PL and nine breast tissues in LL. Afterwards, proteins extracted from nine breast tissues in PL were divided into three groups. Each group was made up of three parts of protein extracted from PL breast tissues. The same method was used to construct the LL protein library.

Trypsin digestion
The protein solution was reduced with 5 mM dithiothreitol for 30 min at 56 °C and alkylated with 11 mM iodoacetamide for 15 min at room temperature in the dark. Consequently, the protein solution was digested. The protein sample was diluted by adding 100 mM triethylammonium bicarbonate (TEAB). The concentration of urea in the solution should be less than 2 M. Finally, trypsin and protein were added at 1:50 mass ratio for the rst digestion overnight and at 1:100 mass ratio for the second digestion for 4 h.

TMT labelling and HPLC fractionation
The production was desalted by a Strata X C18 SPE column (Phenomenex, Torrance, USA) after trypsin digestion and vacuum drying. Afterwards, peptides were reconstituted in 0.5 M TEAB and processed using a TMT kit (Scientz, Ningbo, China). Peptides were labelled using TMT for quantitation; labelling was performed as previously described [32]. One unit of TMT reagent was thawed and reconstituted in acetonitrile. The peptide mixtures were incubated for 2 h at room temperature, pooled, desalted and dried by vacuum centrifugation. The tryptic peptides were divided into segments using a Thermo Betasil C18 column (5 μm particles, 10 mm ID, 250 mm length) by high pH reversed-phase HPLC. In summary, peptides were separated into 60 fractions after more than 60 min of gradient treatment with 8%-32% acetonitrile (pH 9.0). Subsequently, the peptides were combined into eight fractions and dried by vacuum centrifuging. Reactions were quenched using 8 μL of 11% lysine following the manufacturer's recommendations. Finally, 10% tri uoroacetic acid (TFA, Sigma-Aldrich) was added into the peptide solutions. The volume ratio of the peptide solution to TFA was 1:10, and the mixture was pooled, desalted and dried by vacuum centrifugation.

A nity enrichment
The mixtures were rst incubated with immobilised metal a nity chromatography (IMAC) microsphere suspensions with vibration in loading buffer (50% acetonitrile/6% TFA). Phosphopeptide-rich IMAC microspheres were collected by centrifugation. Afterwards, the supernatant was removed. The IMAC microspheres were washed with 50% acetonitrile/6% TFA and 30% acetonitrile/0.1% TFA to remove nonspeci cally adsorbed peptides. To elute the enriched phosphopeptides from the IMAC microspheres, an elution buffer containing 10% NH 4 OH was added, and the enriched phosphopeptides were eluted with vibration. The supernatant containing phosphopeptides was collected for LC-MS/MS analysis after lyophilisation.

LC-MS/MS analysis
The tryptic peptides were dissolved in 0.1% formic acid (solvent A), directly loaded onto a homemade reversed-phase analytical column (15 cm length, 75 μm i.d.). The gradient consisted of solvent B (0.1% formic acid in 98% acetonitrile), which increased from 6% to 23% over 26 min, 23% to 35% in 8 min and climbing to 80% in 3 min, followed by holding at 80% for the last 3 min, all at a constant ow rate of 400 nL/min on an EASY-Nlc 1,000 ultra-performance liquid chromatography (UPLC) system.
Peptides were characterised using tandem mass spectrometry (MS/MS) in Q Exactive™ Plus (Thermo) coupled online to the UPLC after the peptides were subjected to NSI source. The electrospray voltage applied was 2.0 Kv. The m/z scan was 350 to 1,800 for full scan, and intact peptides were detected in the Orbitrap at a resolution of 70,000. Using a normalization collision energy setting of 28, the fragments were detected in the Orbitrap at a resolution of 17,500 for the selected MS/MS of peptides. A datadependent procedure was alternated between one MS scan followed by 20 MS/MS scans with a 15.0 s dynamic exclusion. The automatic gain control was set at 5E4. The xed rst mass was set at 100 m/z.

Database search
Maxquant search engine (V.1.5.2.8) was used to process the resulting MS/MS data. The UniProt Ovis aries (27472 sequences) concatenated with a decoy database that was used for the tandem mass spectra search. The search parameters were as follows: 6-plex TMT, the minimum length of the peptides was set at 7 amino acid residues and the maximum modi cation number of the peptides was set at 5; xed modi cation (carbamidomethylation of cysteine residues); variable modi cation (protein N-term acetylation); and oxidation of methionine residues and phosphorylation of serine, threonine and tyrosine residues, thereby allowing two missing cleavages by cleavage enzyme trypsin/P. The mass error was set at 20 ppm in the rst search and 5 ppm in the primary search, and the fragment ions of mass tolerance were set at 0.02 Da. The minimum score for modi ed peptides was set at > 40, and a false discovery rate was set at < 1%.

Bioinformatics analysis
Gene Ontology (GO) annotation data were obtained from UniProt GoA database. Proteins were divided into three categories by GO annotation, namely, a biological process, cellular compartment and molecular function. The GO with a corrected P value < 0.05 was considered signi cant. InterProScan (http://www.ebi.ac.uk/interpro/) was used to detect the enrichment of functional domains of differentially expressed proteins compared with all identi ed proteins. Protein domains with a corrected P value < 0.05 were considered signi cant. The KEGG database (https://www.genome.jp/kegg/tool/map_pathway2.html) was used to identify enriched pathways. The pathway with a corrected P value < 0.05 was considered signi cant. Based on the KEGG website, these pathways were classi ed into hierarchical categories. Fisher's exact test was used to examine the enrichment of the differentially phosphorylation protein against all identi ed phosphorylation proteins.
Wolfpsort, a subcellular localisation predication wolf, was used to predict subcellular localisation. Soft motif-x was used to analyse the sequence model constituted with amino acids in speci c positions of modify-21-mers (10 amino acids upstream and downstream of the site) in all protein sequences. In addition, all database protein sequences were used as background database parameter, as well as other parameters with default.

Enrichment-based clustering
Hierarchical clustering was based on the functional classi cation of different proteins. Categories that were at least enriched in one of the clusters with P <0.05 were ltered after collating all the categories and their P values. This ltered P value matrix was transformed by the function x = − log10 (P value). Finally, these x values were z-transformed for each functional category. These z scores were then clustered by one-way hierarchical clustering (Euclidean distance and average linkage clustering) in Genesis. Cluster membership was visualised by a heat map using the "heatmap.2" function from the "gplots" R-package.

Western blot
Approximately 25 μg of protein which extracted from goat mammary gland epithelial tissues was transferred onto polyvinylidene di uoride membrane (PVDF, Merck Millipore, MA, USA). The polyvinylidene di uoride membrane was immersed in 10% skimmed milk powder and diluted with Trisbuffered saline inclusive of 0.1% Tween 20 (pH 7.6) for 2.5 h at room temperature. The membrane was immersed in corresponding primary antibodies overnight at 4 °C (Table 1). In the next step, the membrane was incubated in suitable HRP-conjugated secondary antibodies against mouse, rabbit at 4 °C for 2 h. Proteins were visualised using ECL prime western blotting detection reagent (Amersham, GE Healthcare Lifesciences, Swenden) by gel documentation system (Biospectrum 410, UVP). Proteins were quanti ed by the Quantity One program (Bio-Rad, California, USA).

Statistics
Each experiment was repeated at least three times. All the data were processed by SPSS 19.0 (Beijing, China). The results were shown as means ± SE (standard error), and the differences were compared by one-way ANOVA (**P < 0.01, *P < 0.05).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.