Effects of Milk-Derived Extracellular Vesicles on the Colonic Transcriptome and Proteome in Murine Model

Evidence shows that effective nutritional intervention can prevent or mitigate the risk and morbidity of inflammatory bowel disease (IBD). Bovine milk extracellular vesicles (mEVs), a major bioactive constituent of milk, play an important role in maintaining intestinal health. The aims of this study were to assess the effects of mEV pre-supplementation on the colonic transcriptome and proteome in dextran sulphate sodium (DSS)-induced acute colitis, in order to understand the underlying molecular mechanisms of mEV protection against acute colitis. Our results revealed that dietary mEV supplementation alleviated the severity of acute colitis, as evidenced by the reduced disease activity index scores, histological damage, and infiltration of inflammatory cells. In addition, transcriptome profiling analysis found that oral mEVs significantly reduced the expression of pro-inflammatory cytokines (IL-1β, IL-6, IL-17A and IL-33), chemokine ligands (CXCL1, CXCL2, CXCL3, CXCL5, CCL3 and CCL11) and chemokine receptors (CXCR2 and CCR3). Moreover, oral mEVs up-regulated 109 proteins and down-regulated 150 proteins in the DSS-induced murine model, which were involved in modulating amino acid metabolism and lipid metabolism. Collectively, this study might provide new insights for identifying potential targets for the therapeutic effects of mEVs on colitis.


Introduction
Inflammatory bowel disease (IBD) is an incurable inflammatory disorder, which impairs the quality of life of patients and creates substantial costs for health care [1]. Moreover, IBD has evolved into a global disease with increasing incidence in every country [1]. The mechanisms of IBD pathogenesis are the result of an interaction between genetic predisposition, environmental factors, intestinal barrier and immune imbalance [2]. Genetic predisposition explains about 25% of the overall incidence of IBD, and environment factors play an increasingly considerable role in the development of IBD [3,4]. Therefore, manipulation of the susceptibility of gene expression via dietary intervention might help to reduce the risk of IBD [5,6].
Recently, dietary supplements or functional ingredients have gained wide attention, due to their nutritional and therapeutic values, as well as minimal side effects, even after long-term consumption. Milk extracellular vesicles (mEVs), with biomolecular lipid bilayer nanostructures and various cargos, are abundant bioactive ingredients in bovine milk [7]. Thousands of components of various types have been identified in mEVs, including predominantly proteins, nucleic acids, and lipids [8]. Among these cargos, miRNA, lncRNA and circRNA have been demonstrated to participate in regulating a wide variety of gene expression and multiple processes in hosts [9]. For example, Gao et al. found that mEV-derived bta-miR-34a could activate the apoptosis signaling pathway, thus improving intestinal epithelial cell survival [10]. Moreover, emerging evidence supports the claim that mEVs could enhance intestinal immunity [11], attenuate intestinal epithelial cell death [12], and exert protective effects against oxidative stress [13]. There is also a previous study of the effect of mEV supplementation on the gene expression of chronic experimental colitis [14]. However, the study of the protective effects of mEVs on acute colitis has focused mainly on phenotype changes [15], which may have restricted the knowledge and in-depth study of the roles of mEVs on the genes and protein expression mechanisms behind the attenuation of acute colitis.
The applications of multi-omics approaches to identify targets for therapeutic interventions are high on the research agenda for various diseases and help find low abundance targets that were underrepresented before [16]. To date, multi-omics sequencing has been widely employed in investigating the effects of functional foods supplementation on colitis [17,18]. Therefore, in this study, gut transcriptome and proteome were adopted to characterize the effect of mEVs on acute colitis. In addition, the relationship between colonic gene expression and protein levels were also identified. This study not only enriches the biological function of mEVs, but may also uncover the potential mechanisms of mEVs that alleviate acute colitis and the underlying potential therapeutic targets for the treatment of IBD.

Milk Collection
Bovine raw milk samples were collected from mid-lactation Holstein cows at the farm of the Institute of Animal Science, Chinese Academy of Agricultural Sciences, Beijing, China. Then, the fresh milk samples were centrifuged immediately at 3000× g for 30 min at 4 • C to obtain whey. The whey samples were stored at −80 • C for further analysis.

mEVs Isolation from Whey
The whey samples were then performed as described previously to obtain mEVs [13,19]. Briefly, the samples were centrifuged at 12,000× g at 4 • C for 60 min to remove fat globules, casein aggregates and residual chymosin. The supernatant was then centrifuged at 30,000× g, 75,000× g for 60 min each to remove large particles and residual debris. The resulting supernatant was ultracentrifuged at 120,000× g for 90 min using an SW32Ti rotor (Optima XPN-100; Beckman Coulter Instruments, Fullerton, CA, USA). Finally, the pelleted mEVs were resuspended in PBS, filtered through 0.22 µm filters (Millipore, Burlington, MA, USA), and stored at −80 • C for further processing.

Transmission Electron Microscopy (TEM)
The isolated mEVs were stained with uranyl acetate and identified using a HT7700 transmission electron microscope (Hitachi, Tokyo, Japan).

Nanoparticle Tracking Analysis (NTA)
Recent reports support the claim that particle concentration might be more suitable to determine the levels of mEVs [20]. Hence, the particle size and concentration of mEVs were measured by NTA at VivaCell Shanghai with ZetaView PMX 110 (Particle Metrix, Meerbusch, Germany). Data were analyzed by the equipped ZetaView v8.04.02 SP2 software (Particle Metrix, Meerbusch, Germany).

Animals and Treatments
Six-week-old male specific pathogen-free (SPF) C57BL/6 mice were provided by SPF Biotechnology Co., Ltd (Beijing, China). All mice were adapted to the laboratory conditions with free access to food and water and housed at 22 ± 2 • C with 12 h light-dark cycles for two weeks. Animal care was performed following the guidelines for Care and Use of Laboratory Animals of the Chinese Academy of Agricultural Sciences and was approved by the Animal Ethics Committee of Chinese Academy of Agricultural Sciences (approval number: IAS2021-235).
Acute colitis was induced by the addition of 3.5% DSS (molecular weight 36,000-50,000 kDa; MP Biomedicals, CA, USA) to autoclaved drinking water for 7 days. Thirty-six mice were randomly divided into normal control (NC), DSS+PBS, and DSS+mEVs (3.0× 10 9 particles per gram body weight) groups. Mice in the DSS+mEVs group received daily oral gavage with mEVs for 30 days, before addition of 3.5% DSS in the drinking water for 7 days. Body weight was monitored daily. Disease activity index (DAI) was determined by evaluating weight loss, stool consistency and presence of blood in the stool. All mice were sacrificed after anesthesia to measure the length of the colon. Part of the colon tissues was fixed in neutral buffered formalin and the rest of the samples were collected and stored at −80 • C for further analysis.

Histological Analysis
Fresh colon tissues were fixed in 10% neutral buffered formalin at 4 • C overnight. Paraffin was used to embed the tissues, which were then sliced to 10 µm thicknesses. Subsequently, the sections were stained with hematoxylin and eosin (H&E). Images of the sections were pictured using optical microscopy (Olympus, Tokyo, Japan).

Transcriptome Analysis
For RNA isolation, total RNA was extracted from the colon tissues using the mir-Vana™ miRNA Isolation Kit (Ambion, Austin, TX, USA) according to the manufacturer's instructions. RNA integrity was evaluated using 2% agarose gel electrophoresis and an Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA), and the purity of RNA was determined by NanoDrop 2000 (Thermo, Waltham, MA, USA). Only RNA integrity number (RIN) values above 7.0 were considered acceptable for sequencing.
RNA sequencing included the following steps: deplete Ribo-Zero and F=fragment RNA, synthesize first strand cDNA, synthesize second strand cDNA, purity, adenylate 3 Ends, ligate adapters, enrich DNA fragments, purity and validate library. TruSeq Stranded Total RNA with Ribo-Zero Gold (RS-122-2301, Illumina) was used in these processes, according to the manufacturer's instructions. The cDNA library was constructed and sequenced on the Illumina Hiseq X Ten, PE150 platform at Shanghai Ouyi Biomedical Technology co., Ltd (Shanghai, China). The sequencing depth was 12 G clean data.
For the bioinformatics analysis, SortMeRNA v4.2.0 (Cedex, Paris, France) and Trimmomatic v0.36 (Julich, Germany) were used to remove residual rRNA sequences and low-quality reads, respectively, resulting in high-quality clean reads. Hisat2 v2.0.4 (Baltimore, MA, USA) was used to align the high-quality clean reads to the reference genome of the mice. The samples were evaluated by genomic and gene alignment. Stringtie v1.3.1(Baltimore, MA, USA) was used to assemble the reads and splice new transcripts. Fragments per kilobase of transcript per million mapped reads (FPKM) were obtained by Cuffdiff software (Massachusetts, USA) to determine the level of transcript expression.
Expression abundances were log2-transformed to normalize the expression levels per transcript. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed with functional annotation of the differentially expressed genes.

Tandem Mass Tag (TMT) Mass Spectrometry Analysis
Frozen colon tissues were treated with a buffer (7 mM sucrose, 1 mM NaCl, 0.39 mM EDTA·2 Na, 0.086 mM DTT) and supplemented with 1 mM PMSF. Then, the samples were homogenized by sonication and added to Tris-phenol buffer and mixed for 30 min at 4 • C. The mixtures were centrifuged at 7100× g for 10 min at 4 • C to obtain the supernatants. Afterwards, the supernatants were precipitated for 5 times with the volumes of 0.1 M cold ammonium acetate-methanol buffer and kept at −20 • C overnight. After precipitation, the proteins were washed with cold methanol and acetone successively. Finally, the extracted protein concentrations were determined by BCA assay (cat. no. 23227, Pierce, Thermo Scientific, Landsmeer, Netherlands) and stored at −80 • C until use.
All raw data were analyzed by Proteome Discover 2.4 (Thermo, Waltham, MA, USA). Mass tolerances for precursor and fragment ions were 0.02 Da and 10 ppm, respectively. The false discovery rate (FDR) was less than 0.01 and the number of peptides had to be over 2 for further quantification. Protein-protein interaction (PPI) was used to analyze the differentially expressed proteins based on STRING database. Cytoscape v3.8.0 (Bethesda, Maryland, USA) software was used to visualize the PPI network.

Correlations Analysis of Proteome and Transcriptome
To evaluate the correlation between transcriptome and proteome levels in the colon, Pearson correlation coefficients were performed. A correlation heat map was presented by GraphPad Prism version 8.0 (GraphPad Software Inc., San Diego, CA, USA).

Real-Time PCR
To verify the results of the transcriptome, the extracted RNA from the colon was reverse transcribed into cDNA, using the PrimeScript RT Master Mix kit (cat. no. RR036a, TaKaRa, Shiga, Japan). Then, the cDNA was performed on the QuantstudioTM 7 flex system (ABI Q7 Flex 384 well, Life Technologies, San Diego, CA, USA) with the TB Green®Premix Ex Taq ™ II (cat. no. RR820a, TaKaRa, Tokyo, Japan) and gene-specific primers (Table S1). The relative expression was calculated by the 2-∆∆Ct formula, where ∆∆Ct = (CtTarget-Ct GAPDH) DSS+mEVs-(CtTarget-Ct GAPDH) DSS+PBS. GAPDH was used as the housekeeping gene for normalization.

Statistical Analysis
All data were presented as the mean ± SEM and drawn by GraphPad Prism 8.0 (GraphPad Software, Inc. San Diego, CA, USA). The comparation of body weights, DAI, colonic length and histological score of colons were performed by one-way ANOVA analysis, and relative mRNA expression obtained from real-time PCR was analyzed by unpaired Student's t-test of SPSS version 22.0 (SPSS, lnc., Chicago, IL, USA). A significant difference was declared at p < 0.05. The principal component analysis (PCA) was performed using R Script. For RNA sequencing data, the genes were filtered according to the value of counts, and only those with a value greater than 2 were analyzed further. DESeq2 v1.14 (New York, NY, USA) software was used to standardize the counts number of each sample gene, and calculate the foldchange (FC). Then, the NB (negative binomial distribution test) was used as the significance test. Only FC > 1.5/FC < 0.67 and FDR < 0.05 were considered statistically significant. For the proteomic data, the protein abundance was analyzed by Student's t-test with p < 0.05 and FC ≥ 1.2/FC ≤ 0.84 being considered statistically significant.

Characterization of Bovine mEVs
The mEVs isolated from bovine milk were characterized by TEM, NTA and WB. The shape of the mEVs showed a circular or elliptical structure ( Figure 1A). The peak diameter of mEVs was 154.78 ± 5.22 nm ( Figure 1C). The mEVs contained abundant positive surface markers HSP70, TSG101, CD63, and CD9 and only a marginal amount of the negative marker calnexin ( Figure 1B). The above results confirmed that the isolated particles were mEVs.

Protective Roles of mEVs on DSS-Induced Colitis
Compared with the DSS+PBS group, mEV intervention significantly attenuated DSSinduced colitis, as evidenced by the significantly reduced body weight loss, relieved DAI score, and decreased colon shortening (p < 0.05) (Figure 2A-C). Investigation of colonic morphology presented that oral mEVs significantly reduced the infiltration of inflammatory cells, mucosal damage and overall histology score, compared with the DSS + PBS group ( Figure 2D,E). These results suggested that pre-supplementation of mEVs was capable of suppressing DSS-treated colitis symptoms and colonic injury. The histology score. Data represent the mean ± SEM, n = 12 samples/group, * p < 0.05, ** p < 0.01, *** p < 0.001. * p < 0.05 relative to DSS+PBS group; # p < 0.05, ### p < 0.001 relative to NC group.

mEVs Changed Gut Gene Expression Profile and Signaling Pathways
To explore the mechanisms associated with the phenotype changes, we investigated the colonic transcriptome differences between the DSS+PBS and DSS+mEVs groups. A total of 1,289,007,754 raw reads (193.34 Gb) and 1,263,042,544 (185.56 Gb) clean reads were produced. The Q30 of all the reads ranged from 90.75% to 93.30% and the average GC content was 48.41% (Table S2). The average percent of reads that were consistently mapped to the transcriptome was 95.90%. The FPKM and gene expression density are shown in Figure S1, presenting comprehensive representation of the transcriptome ( Figure S1A,B).
mEVs significantly regulated the colonic gene expression in the DSS model, as evidenced by the PCA and hierarchical clustering based on Pearson's correlation coefficients in all samples ( Figure 3A,B). Compared with the DSS+PBS group, oral mEV intervention significantly up-regulated the expression of 246 genes (FC > 1.5 and p < 0.05) and downregulated the expression of 164 genes (FC < 0.67 and p < 0.05) ( Figure 3C, Table S3). In the presence of mEVs, the expression of proinflammatory cytokines (IL-1α, IL-1β, IL-33, IL-6 and IL-17A) and cytokine receptors (IL-1R1 and IL-1R2) were significantly decreased in DSS-treated mice (p < 0.05) ( Figure 3D, Table S3). Among these genes, IL-1β, IL-6 and IL-17A have critical roles in the onset and progression of IBD [21]. In addition, mEV administration significantly inhibited the expression of chemokine ligands (CXCL1, CXCL2, CXCL3, CXCL5, CCL3, CCL4 and CCL11) (p < 0.05), chemokine receptors (CXCR2 and CCR3), growth factors (CSF3), extracellular matrix remodeling enzymes (MMP3, MMP8, MMP9, MMP10, MMP12, MMP13, PLAUR and TIMP1), oxidative stress (PTGS2), proinflammatory mediators (S100a8) and other candidate biomarkers (PROK2) ( Figure 3D, Table S3), in comparison with the DSS+PBS group. To verify the results of the RNA-seq analysis, a total of six differentially expressed genes in this study were selected to further investigate their expression profiles ( Figure S2). Taken together, the reduction in inflammatory cytokines and chemokines levels in colon tissues suggested that mEVs exhibited preventive and therapeutic effects in colitis mice.
The changes in GO terms and KEGG pathways in response to mEV supplementation were also examined. According to the GO analysis of differentially expressed genes, extracellular region, extracellular space, inflammatory response, neutrophil chemotaxis and cytokine activity were the top five significantly down-regulated GO terms (p < 0.05) ( Figure S1D). In addition, circadian regulation of gene expression, rhythmic process, Z disc, circadian rhythm and membrane were the top five significantly up-regulated GO terms (p < 0.05) ( Figure S1E). The KEGG annotation analysis of differentially expressed genes revealed that 44 pathways were significantly down−regulated ( Figure 3E) and 13 pathways were significantly up-regulated (p < 0.05) ( Figure S1C). The down-regulated pathways were associated with inflammation, such as the IL-17 signaling pathway, TNF signaling pathway, NF-κB signaling pathway and NOD-like receptor signaling pathway ( Figure 3F-I). In addition, the IBD pathway was also down-regulated ( Figure 3E). Moreover, the up-regulated KEGG pathways were closely associated with the metabolism of xenobiotics by cytochrome P450, glycosaminoglycan biosynthesis-keratan sulfate and steroid hormone biosynthesis. These data demonstrated that mEVs ameliorated colitis through modulating multiple pathways, including inflammation−related signaling pathways and nutrition metabolism.
The 56 significantly enriched GO terms, based on the enrichment scores of the differentially expressed proteins in DSS-treated mice with and without receiving oral mEVs, in the cellular component, biological process and molecular function were listed in Table S5 (p < 0.05). The top three most enriched pathways were discoidal high-density lipoprotein particle, glutamine-fructose-6-phosphate transaminase (isomerizing) activity and CoA carboxylase activity. Further enrichment chord analysis showed that discoidal high-density lipoprotein particle, high-density lipoprotein particle assembly, response to oxidative stress, very low-density lipoprotein particle, positive regulation of cholesterol esterification and actin filament binding were involved in the attenuation of colitis by oral mEVs ( Figure 4D).  To further understand the cellular pathways through which the annotated unigenes were potentially related to ripening, the 259 differentially expressed proteins were mapped to the KEGG database. In total, 23 KEGG pathways were significantly up-regulated and 14 were significantly down-regulated. Oral mEVs significantly improved amino acid metabolism and lipid metabolism, including valine, leucine and isoleucine degradation, tryptophan metabolism, arginine and proline metabolism, histidine metabolism, pyruvate metabolism and fatty acid degradation ( Figure 4E). Moreover, Fc gamma R-mediated phagocytosis, leukocyte transendothelial migration and TNF signaling pathways were the main down-regulated KEGG pathways in response to oral mEVs ( Figure 4F). These results suggested that mEVs-induced metabolic reprogramming of amino acid metabolism and lipid metabolism in the colon was beneficial for protecting against colitis.

Correlation Analysis of mRNAs and Proteins
We explored the correlation between the transcriptome and proteome and found that the regulation of gene abundance resulted in a weaker tendency for changes in the corresponding protein level ( Figure 5A). A total of 60 proteins and genes with same changes were identified from the differentially expressed proteins and genes ( Figure 5B). Among them, 39 proteins−genes were co-upregulated (such as TGM3, CLCA1, SLC26A3, and LYPD8), and 21 were co-downregulated ( Figure 5B). GO and KEGG enrichment analyses were used to explore the potential functions of the 60 same changes in the proteins and genes. "Tissue remodeling" and "metabolism" were the main terms that were enriched in the biological process category ( Figure 5C). In addition, the KEGG pathway analysis showed that the genes were enriched in metabolic pathways and many signaling pathways, including histidine metabolism, arginine and proline metabolism, alanine, aspartate and glutamate metabolism, carbohydrate digestion and absorption and fatty acid degradation ( Figure 5D).

Discussion
mEVs have been proved to transfer from milk into the plasma through the intestinal tract of the consuming organism, and to affect gene expression once absorbed by target cells, thereby interfering with the physiological and pathological processes of the consumers [22][23][24]. Reif et al. reported that mEVs could be absorbed by intestine cells, exerting a therapeutic and anti-inflammatory role in DSS−induced colitis, which involves reducing the expression of IL-6 and TNF-α [25]. In another study, mEVs also had anti−inflammatory capacity by inhibiting the expression of the pro-inflammatory cytokines IL-6 [26]. Benmoussa et al. further determined the concentration of cytokines, chemokines and inflammation−related proteins in the colon of colitis mice, and demonstrated that mEVs decreased the production of the intestinal inflammation driver and pro-inflammatory cytokines [27]. Consistent with the above studies, we observed the protective and therapeutic roles of mEVs on DSS−induced murine colitis, as evidenced by a decrease in body weight loss, DAI score, colonic shortening and histology score. We also found that mEVs decreased the expression of pro-inflammatory factors (IL-1β, IL-6 and IL-17A) in colitis. Therefore, nucleic acids, proteins and lipids delivered by mEVs into recipient cells changed their biological responses, and effectively attenuated the symptoms of acute colitis.
Chemokines are a subset of cytokines, which could induce immune cell migration to sites of inflammation, ultimately resulting in tissue damage and destruction [28]. CCL11, also known as eotaxin-1, is considered as an eosinophil chemoattractant, and is involved in the pathogenesis of DSS−induced colitis [29]. CCL3 and CCL4, known as MIP-1α and MIP-1β, respectively, are indispensable mediators of polymorphonuclear leukocyte and macrophage recruitment. In addition, CCL3 can specifically move macrophages and granulocytes to the inflammatory area in acute inflammation [28]. Members of the CXCL chemokine families (e.g., CXCL1, CXCL2, CXCL3 and CXCL5) are also closely related with immune response. CXCL1 is also known as NAP3, which induces neutrophil migration [30]. CXCL2 is also known as MIP2-α, which is secreted by monocytes and macrophages [30]. Hence, the anti-inflammatory mechanisms of mEVs are mainly associated with its ability to inhibit chemokine expression.
Apart from the above chemokines' genes, mEVs also changed other genes' expression, when compared to the DSS+PBS group. CYP1A1 is reported to modulate immune responses in the intestine and exert protection against intestinal inflammation [31,32]. The SLC6A4 gene is significantly down-regulated during intestinal inflammation, which codes for a protein that is the only high affinity and competent transporter of serotonin [33]. SLC6A8, also known as CRT, is the only famous creatine transporter, and might regulate epithelial integrity and barrier function [34]. Hence, loss of SLC6A8 expression could induce metabolic disorders and can usually be found in IBD patients [34]. PHLPP2 might be a promising candidate therapeutic target for colitis, and the up-expression of PHLPP2 could inhibit the NF-κB signaling pathway to alleviate colonic inflammatory response [35,36]. ANO1 is vital for the protection of intestinal epithelium from colitis, whereas its expression was significantly decreased in the DSS-induced colitis mice [37,38]. NR1D1 knockout mice were more susceptible to experimental colitis, suggesting an essential role of NR1D1 in colitis development [39,40]. Moreover, NR1D1, as a transcriptional repressor, could repress the NF-κB/Nlrp3 axis to prevent colitis [39]. Although CPA6 is closely associated with the extracellular matrix or immune response, its expression was decreased in colitis [41]. GLP-2R has the ability to improve crypt cell proliferation, suppress epithelial cell apoptosis and maintain mucosal integrity. However, the level of GLP-2R was reduced in colitis in both mice and patients [42]. GSTT1 is found to alleviate colitis by the IL-22 dependent pathway [43]. SlCO2A1 deficiency is identified to deteriorate intestinal inflammation by activating the NLRP3 inflammasome in macrophages [44]. For the reasons above, the up-regulated CYP1A1, SLC6A4, SLC6A8, PHLPP2, ANO1, NR1D1, CPA6, GLP-2R, GSTT1 and SlCO2A1 in mEV−fed mice indicated that mEVs facilitated the expression of genes that had low expression in colitis, and hence ameliorated colitis.
Additionally, the results of the proteomic changes could provide further mechanistic understanding of how mEVs alleviate acute colitis. A previous study demonstrated that the expression of PBLD was significantly decreased in mice with DSS-induced colitis; however, its repressive action on NF-κB signaling could improve intestinal barrier function and alleviate intestinal inflammation [45]. FAM3D, as a protector of colonic homeostasis, is reported to maintain the normal differentiation of goblet cells, as well as the presence of acidic mucins and antimicrobial molecules [46]. The KRT8 protein plays an important role in anti-inflammatory processes and in sustaining cellular structural integrity [47]. SULT2B1 is a critical enzyme that catalyzes the synthesis of cholesterol sulfate from cholesterol, and its deficiency promotes pro-inflammatory macrophage polarization [48]. Protein GPA33 is a marker of Treg cells, and GPA33-deficient mice would increase the rate of colitis [49]. DEPTOR has the ability to inhibit the activity of mTOR, which could promote inflammation by secreting cytokines (TNF-α, IL-1β, and IL-6) and chemokines (CCL7 and CXCL16) [50]. ACSL4, an important enzyme of long-chain fatty acid degradation, is involved in inflammasome activation in neutrophils, and is significantly up-regulated in inflamed colon tissues [51]. The polymorphisms of IFITM1 increase the sensitivity to ulcerative colitis, and the expression of IFITM1 is also up-regulated in colitis [52]. HSD11B1 exerts an important role in the inflammatory response, and the high expression of HSD11B1 will increase the incidence of colitis [53]. Therefore, the increased levels of PBLD, FAM3D, KRT8, SULT2B1, GPA33, DEPTOR, and decreased expression of ACSL4, IFITM1 and HSD11B1 caused by mEVs has been demonstrated to attenuate colitis, implying that consumption of mEVs has the potential to improve intestinal health.
A previous study found that the correlation coefficients between mRNAs and protein levels are often low [54]. Our study also revealed the poor correlation between transcriptional activity and protein production. The poor correlation may be explained by the post-transcriptional regulation and different half-life of mRNAs and proteins [55]. Although our results showed a non-perfect correlation between the transcriptomic and proteomic surveys, there were some similar influences on several genes/proteins. In our study, the expression of TGM3, CLCA1, SLC26A3, and LYPD8 genes/proteins were significantly up-regulated in response to mEVs ( Figure 5B). TGM3 is expressed and synthesized in goblet cells, which exclusively determines the transamination activity of colonic mucus [56]. CLCA1 also maintains mucus homeostasis to decrease the risk of colitis [57]. SLC26A3 contributes to protecting the intestinal epithelial barrier and maintaining the enterocyte acid/base balance. Thus, the high expression of SLC26A3 could prevent DSS−induced acute colitis [58]. LYPD8 levels are significantly decreased in colitis patients, and supplementation of LYPD8 proteins might maintain colonic homeostasis in colitis patients by inhibiting the motility of pathogenic bacteria, such as γ-proteobacteria [59,60].
Overall, the altered gene and protein expressions derived from mEV intervention might play a central role in the attenuation of colitis, which could also be used as novel candidate biomarkers for the treatment of intestinal disorders. However, this study did not exclude the effects of slight impurities from the remaining milk ingredients. Furthermore, adding a mEV−depleted bovine milk group would offer more convincing evidence.

Conclusions
Bovine mEVs attenuated DSS−induced acute colitis models, as evidenced by increased body weight, reduced DAI and maintained integrity of the colon tissue. Transcriptome data revealed that mEVs inhibited the expression of pro-inflammatory cytokines (IL-1β and IL-6) and chemokines (CXCL1, CXCL2 and CXCL3) and promoted the expression of anti-inflammatory genes. Moreover, this study provided evidence of the proteomic variations in the colon in response to oral mEVs. By integrating the results of transcriptomics and proteomics, we found that amino acid metabolism and immune-related signaling pathways played a crucial role in the process of mEVs in alleviating colitis. Overall, mEV supplementation in the current study presented great potential and many desirable features for the prevention of acute colitis.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/nu14153057/s1, Table S1. The list of primer sequences. Table S2. The results of raw reads after quality control. Table S3. Quantitative analyses of transcriptomic data at the gene level. Table S4. Quantitative analyses of proteomic data. Table S5. GO analysis of differently expressed proteins in response to mEVs. Figure S1.