{"http:\/\/dx.doi.org\/10.14288\/1.0379196":{"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool":[{"value":"Medicine, Faculty of","type":"literal","lang":"en"},{"value":"Non UBC","type":"literal","lang":"en"},{"value":"Medicine, Department of","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider":[{"value":"DSpace","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#identifierCitation":[{"value":"Nutrients 11 (2): 418 (2019)","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/creator":[{"value":"Abulizi, Nijiati","type":"literal","lang":"en"},{"value":"Quin, Candice","type":"literal","lang":"en"},{"value":"Brown, Kirsty","type":"literal","lang":"en"},{"value":"Chan, Yee Kwan","type":"literal","lang":"en"},{"value":"Gill, Sandeep K.","type":"literal","lang":"en"},{"value":"Gibson, Deanna L.","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/issued":[{"value":"2019-05-30T16:03:53Z","type":"literal","lang":"en"},{"value":"2019-02-16","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/description":[{"value":"The dynamics of the tripartite relationship between the host, gut bacteria and diet in the gut is relatively unknown. An imbalance between harmful and protective gut bacteria, termed dysbiosis, has been linked to many diseases and has most often been attributed to high-fat dietary intake. However, we recently clarified that the type of fat, not calories, were important in the development of murine colitis. To further understand the host-microbe dynamic in response to dietary lipids, we fed mice isocaloric high-fat diets containing either milk fat, corn oil or olive oil and performed 16S rRNA gene sequencing of the colon microbiome and mass spectrometry-based relative quantification of the colonic metaproteome. The corn oil diet, rich in omega-6 polyunsaturated fatty acids, increased the potential for pathobiont survival and invasion in an inflamed, oxidized and damaged gut while saturated fatty acids promoted compensatory inflammatory responses involved in tissue healing. We conclude that various lipids uniquely alter the host-microbe interaction in the gut. While high-fat consumption has a distinct impact on the gut microbiota, the type of fatty acids alters the relative microbial abundances and predicted functions. These results support that the type of fat are key to understanding the biological effects of high-fat diets on gut health.","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO":[{"value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/70425?expand=metadata","type":"literal","lang":"en"}],"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note":[{"value":"nutrientsArticleGut Mucosal Proteins and Bacteriome Are Shaped bythe Saturation Index of Dietary LipidsNijiati Abulizi 1,\u2020 , Candice Quin 1,\u2020 , Kirsty Brown 1, Yee Kwan Chan 1, Sandeep K. Gill 1 andDeanna L. Gibson 1,2,*1 Department of Biology, IKBSAS, University of British Columbia, Okanagan campus, Kelowna V1V 1V7,Canada; nijiati.abulizi@gmail.com (N.A.); candicequin@hotmail.com (C.Q.);kirsty.brown12@gmail.com (K.B.); cyk.carol@hotmail.com (Y.K.C.); sand.gill01@gmail.com (S.K.G.)2 Department of Medicine, Faculty of Medicine, University of British Columbia, Vancouver V6T 1Z3, Canada* Correspondence: deanna.gibson@ubc.ca; Tel.: +001-250-807-8790\u2020 These authors contributed equally.Received: 11 January 2019; Accepted: 13 February 2019; Published: 16 February 2019\u0001\u0002\u0003\u0001\u0004\u0005\u0006\u0007\b\u0001\u0001\u0002\u0003\u0004\u0005\u0006\u0007Abstract: The dynamics of the tripartite relationship between the host, gut bacteria and diet in the gutis relatively unknown. An imbalance between harmful and protective gut bacteria, termed dysbiosis,has been linked to many diseases and has most often been attributed to high-fat dietary intake.However, we recently clarified that the type of fat, not calories, were important in the development ofmurine colitis. To further understand the host-microbe dynamic in response to dietary lipids, we fedmice isocaloric high-fat diets containing either milk fat, corn oil or olive oil and performed 16S rRNAgene sequencing of the colon microbiome and mass spectrometry-based relative quantification ofthe colonic metaproteome. The corn oil diet, rich in omega-6 polyunsaturated fatty acids, increasedthe potential for pathobiont survival and invasion in an inflamed, oxidized and damaged gut whilesaturated fatty acids promoted compensatory inflammatory responses involved in tissue healing.We conclude that various lipids uniquely alter the host-microbe interaction in the gut. While high-fatconsumption has a distinct impact on the gut microbiota, the type of fatty acids alters the relativemicrobial abundances and predicted functions. These results support that the type of fat are key tounderstanding the biological effects of high-fat diets on gut health.Keywords: Host-microbe interactions; gut microbiome; dietary lipids; polyunsaturated fatty acids;monounsaturated fatty acids; saturated fatty acids; proteome; 16S rRNA gene amplicon sequencing;short-chain fatty acid metabolism1. IntroductionThe mammalian gut has co-evolved with thousands of bacterial species and together they forma complex dynamic relationship of which the physiological consequences are largely undiscovered.The gut microbial ecosystem has the potential to influence the overall health status of the mammalianhost by forming an interface between the gut mucosal surface and the luminal food ingested into thebody. Molecular crosstalk between the microbiome and the host epithelium influences intestinal barrierfunction, in part through the release of microbial metabolites like short-chain fatty acids (SCFA) [1].Increased intestinal permeability caused by a disruption of the microbiome, termed dysbiosis, has beenimplicated in diseases including inflammatory bowel disease (IBD), obesity and diabetes [2]. High-fatdiets have been shown to induce dysbiosis, primarily characterized by the escalation of Firmicutesaccompanied by a decrease of Bacteroidetes [3,4]. Changes in microbes induced through diet modulatemajor gene networks including signal transduction, inflammation, histamine, cell migration andNutrients 2019, 11, 418; doi:10.3390\/nu11020418 www.mdpi.com\/journal\/nutrientsNutrients 2019, 11, 418 2 of 24adhesion [5]. Therefore, identifying specific nutrients that prevent dysbiosis may be important inpreventing associated diseases.Lipids are essential for normal development and survival in mammals. Subtle differences in thechemistry of fatty acids effect mammalian physiology and inflammation. Oleic acid, a monounsaturatedfatty acid (MUFA), is the main component of olive oil, a major ingredient in the Mediterranean diet.In general, MUFA consumption is associated with health benefits including lower prevalence ofdigestive system cancers [6], decreased type 2 diabetes [7] and IBD [8]. Indeed, we have shown thatolive oil diets are effective at protecting against murine colitis [9]. In contrast, while North Americandietary guidelines recommend consuming omega-6 polyunsaturated fatty acids (n-6 PUFA), commonin vegetable seed oils, excessive consumption of n-6 PUFA is a risk factor for IBD in humans [10].In support of these findings, we have shown n-6 PUFA exacerbates murine colitis [9,11,12]. Conflictingdata exists for dietary intake of saturated fatty acids (SFA), which have no double bonds and are foundin dairy as well as coconut oil. SFA have been criticized as adversely affecting health over the past fewdecades, yet chronic inflammatory diseases are increasing while the global consumption of SFA havebeen in line with recommended low intakes [13]. Recently, a European prospective cohort study foundthat milk consumption is associated with decreased risk for IBD patients [14]. In contrast, SFA fed tomice resulted in increased spontaneous colitis in IL-10-\/- mice via conjugation of hepatic bile acidswhich promoted growth of Bilophila wadsworthia [15]. Yet, components of animal fat, such as butyricacid, suppress inflammation [16], protect against DSS-colitis [17] and stimulate colonic repair [18].In line with this, we have shown that milk fat promotes beneficial responses during colitis [9]. Whilethere is evidence that different dietary fatty acids have differential effects on host health, their effectson the gut bacterial ecosystem and their functional interaction with the host are not well explored.To understand the tripartite relationship between lipid diet, gut bacteria and the host, we fedmice a 40% (by energy) isocaloric and isonitrogenous diet composed of either corn oil, olive oil or milkfat for 5 weeks post-weaning. The gut tissues were collected for 16S rRNA gene amplicon sequencingand metaproteomic analysis. We show the corn oil diet, rich in n-6 PUFA, produces a microbiomepredicted to have enhanced virulence and pathogenicity potential. This was associated with a colonicproteome increased in proteins involved in inflammation, oxidative stress and barrier dysfunction.While the milk fat diet, rich in SFA, resulted in a host-microbe relationship indicative of inflammation,there was also a compensatory protective response evident by the increased host sirtuin signalingpathway and microbial production of SCFA. In marked contrast to both corn oil and milk fat, the oliveoil diet, rich in MUFA resulted in a microbiome most similar to a low-fat diet. These results supportthat not all high-fat diets promote similar host and microbial responses and that consideration of thetype of fat in high-fat diets is essential when investigating gut health. These results have the potentialto guide evidence-based nutrition recommendations for IBD patients who can suffer from nutrientdeficiencies from overly restrictive dietary regimes including low-fat diets.2. Materials and Methods2.1. Dietary Interventions and Tissue CollectionThree-week-old male and female C57BL\/6 mice (total n=32, n=8 each diet; 4 each sex) were fedirradiated isocaloric, isonitrogenous diets for 5 weeks. High-fat diets contained 40% energy fromolive oil, corn oil or anhydrous milk fat prepared by blending dietary oils to a basal diet mix aspreviously reported, whereas the chow control contained 9% energy from corn oil [11]. Mice wereraised in the same room and litter mates were separated into different diet groups post-natally andthen co-housed with four mice per cage. From these four, two mice per cage were used in thisstudy giving a total of 4 cages per group. Mice (Jackson Laboratories, Bar Harbor, Maine) weremaintained at the Center for Disease Modeling at the University of British Columbia (UBC), Vancouver,Canada. The animal room was temperature controlled (22+\/\u22122\u25e6C) with a 12-h light\/dark cycle andfed with respective diets ad libitum with free access to autoclaved pH neutral water under a specificNutrients 2019, 11, 418 3 of 24pathogen-free condition. Food intake and weight gain was monitored weekly. Mice were anaesthetizedwith isoflurane and euthanized by cervical dislocation. The distal region of the colon (with the luminalcontent and stool removed) was snap frozen in liquid nitrogen and stored at \u221280\u25e6C prior to ampliconsequencing\/proteomic experiments.2.2. Bacterial Genomic DNA ExtractionFrozen tissues were homogenized using stainless steel beads in Mixer Mill MM 400 (Retsch).Bacterial genomic DNA was extracted with QIAamp\u00ae DNA Stool Mini Kit according to themanufacturer\u2019s instructions. DNA concentration and purity were checked with Nanodrop 2000(Thermo Scientific). Primers were used to amplify the 16S rRNA gene as described previously [19].The PCR product was purified with QIAquick\u00ae Gel Extraction Kit (Qiagen) and PCR ampliconsconcentration was normalized with SequalPrepTM Normalization Plate Kit (Invitrogen). Librarypreparation, emPCR amplification and picotitre plate pyrosequencing using titanium chemistry wascarried out by Vancouver Prostate Centre, UBC and Vancouver General Hospital Centre of Excellencein accordance with Roche\/454 Life Sciences protocol on the 454 GS FLX+ System.2.3. Bioinformatics Sequencing and AnalysisSequencing was performed using Roche 454 technology. Sequences were analyzed using theQuantitative Insights Into Microbial Ecology (QIIME) pipeline [20] with default parameters. Sincereads in the 454 platform vary in length the two runs (male and female colons) were truncated to alength of 250 to retain at least 70% of the reads with the recommended 1% expected error threshold [21].Libraries were processed with a minimum quality score of 25 and a quality score window value of50. The quality filtered reads were then combined and chimeras were filtered using usearch61 [22].Sequences were aligned using PyNAST [23] and any sequences that failed to align were omittedfrom the subsequent tree and operational taxonomic unit (OTU) table. Both open-reference andclosed-reference OTU clustering was done at 97% similarity level against the most recent GreenGenesdatabase (gg_13_8_otus). An open-reference OTU table contains a combination of de novo OTUs(reads that do not match reference sequences) as well as reads that match sequences in the referencedatabase. Closed-reference OTU table discards any reads that do not match the sequence in thereference database. Prior to sequence processing, the individual sequencing statistics for the malecolon was 336,801 reads (max length 911, average 400.1) and 326,727 reads (max length 1190, average421.9) for the female colon. Following quality filtering, truncation and chimera removal a mean totalof 6586 high-quality bacterial 16S rRNA sequence reads from the 32 mice remained prior to rarefaction.Samples were rarefied to the same sequencing depth of 2069 (open-reference) for alpha diversity andbeta diversity, and 1458 (closed-reference), for phyla ratios using QIIME2. Alpha diversity rarefactioncurves were used to ensure appropriate sampling depth. Phylogenetic Investigation of Communitiesby Reconstruction of Unobserved States (PICRUSt) [24] and linear discriminant analysis effect size(LEfSe) [25] tools were used for further analyses.2.4. Alpha and Beta Diversity AnalysisAlpha diversity metrics included observed species richness, Chao1, Simpson\u2019s index (D) andShannon\u2019s diversity. A Kruskal\u2013Wallis analysis combined with Benjamini\u2013Hochberg adjustment formultiple comparisons was used to determine the gut microbiome differences between the dietarygroups. The structure of bacterial communities in each diet group were compared using weightedand unweighted UniFrac metrics [26]. Based on these distance matrices, a PERMANOVA [27]was used to analyze sample composition. Significance was assessed by 999 permutations for alldistance-based methods. An adjusted P value (Q-value) less than 0.05 was considered statisticallysignificant. To visualize microbial community composition, a principal coordinates analysis (PCoA)was performed on the distance data and the first two principal components were used to generate anordination plot in Primer 6.Nutrients 2019, 11, 418 4 of 242.5. Abundance AnalysisLEfSe was used to identify differences in taxa composition and Kyoto Encyclopedia of Genesand Genomes (KEGG) orthologs between the dietary groups. Differential abundance analysis wasperformed on the closed-reference OTU table with the logarithmic linear discriminant analysis(LDA) score of 2 as the cutoff and the less permissive \u2018all-against-all\u2019 strategy selected for pairwisecomparisons [25]. LEfSe first tests for statistical significance between dietary groups (non-parametricKruskal\u2013Wallis test) followed by quantitative tests for biological consistency (non-parametricWilcoxon-rank sum test). Multiple test corrections were performed by the Benjamini\u2013Hochbergprocedure-based false discovery rate (FDR) control (\u2018p.adjust\u2019 in R). An adjusted P value (Q-value) lessthan 0.05 was statistically significant.2.6. Amplicon Sequencing Prediction AnalysisPICRUSt was used to infer the relative abundance of gene families and biochemical pathwaysbased on the 16S rRNA data (version: 13_5) [24] The rarefied closed-reference OTU table was firstnormalized for the 16S copy numbers of each OTU and then linked to KEGG annotations of referencegenomes [28]. The generated KEGG pathways were submitted to HUMAnN (The HMP UnifiedMetabolic Analysis Network; version 0.99) for further analysis. The HUMAnN produced pathwaysummaries were analyzed by LEfSe to determine the differential abundance of KEGG pathways.BugBase [29], a microbiome analysis tool used to predict high-level phenotypes, was used to determinethe proportion of Gram-positive, Gram-negative, aerobic, anaerobic, facultative anaerobic, biofilmforming and mobile element containing bacteria.2.7. Short-Chain Fatty Acid AnalysisSCFAs were analyzed from cecal tissue samples using direct injection gas chromatography aspreviously described [19]. Tissue samples were homogenized in 700 \u00b5L of isopropyl alcohol, with0.01% 2-ethylbutryic acid as the internal standard, at 30 Hz for 13 min using stainless steel beads.Homogenate was centrifuged at 15,100 \u00d7 g for 10 min at 4 \u25e6C. Complete extraction was confirmed byabsence of SCFA in the supernatant after second re-extraction of the remaining tissue pellet. 0.9 \u00b5L ofcleared supernatant was directly injected to Trace 1300 Gas Chromatograph (D.I.A.B.E.T.E.S center,UBCO), equipped with flame-ionization detector, with AI1310 auto sampler in splitless mode. A fusedsilica FAMEWAX column 30 m \u00d7 0.32 mm i.d. coated with 0.25\u00b5m film thickness was used. Dataanalysis was carried out with Chromeleon 7 software. Peaks were analyzed on software and the areaunder peaks for acetic, propionic, and butyric acid data were represented as percent weight of totalwet cecal sample (mass %).2.8. Protein ExtractionFrozen colon pieces were scraped to separate the mucosa from the submucosa following a similarprotocol as previously described [30]. The submucosal and the mucosal samples were separatelyput into lysis buffer made up of 25 mM HEPES solution (pH = 7.5) with 1 tablet protease inhibitorcontaining bestatin, AEBSF, EDTA, pepstatin, and E-64 (Thermo Fisher Scientific), 7 M urea, 2 Mthiourea, and 4% CHAPS. The samples were homogenized via bead beating. Insoluble materials wereremoved by centrifugation and then soluble proteins were acetone precipitated from the supernatantand pelleted by centrifugation.2.9. Protein Digestion, Itraq Labeling and LC-MS\/MS AnalysisSamples were prepared for proteomic analysis at the University of Victoria, Genome BC ProteomeCenter located at the Vancouver Island Technology Park. Equal amounts of the extracted proteinfrom each mouse were pooled, group-wise, to generate pooled lysates for low fat (n = 6), milk fat(n = 6), olive oil (n = 6) and corn oil (n = 6) groups. Further, equal amounts of protein from all dietaryNutrients 2019, 11, 418 5 of 24groups were used to generate a total protein\u2019 pool. Sample pooling strategy has been used widely toreduce the effect of biological variation while dealing with clinical samples [31\u201334]. 100 \u00b5g of proteinfrom each dietary group was trypsin digested and then individually labeled using 8-plex iTRAQreagents (AB Sciex, ON, Canada). The labeled peptides were pooled and vacuum centrifuged until thefinal volume was approximately 100\u00b5L. An Agilent 1290 High-Performance Liquid Chromatography(HPLC) system (Agilent, CA, USA) was equipped with an XBridge C18 BEH300 (Waters, MA, USA)250 mm X 4.6 mm, 5 \u00b5m, 300 A HPLC column. The flow rate was set to 0.75 mL\/min, samples wereinjected onto the column and fractions were collected every minute for 96 min. The HPLC fractionswere then reduced in volume by lyophilization and concatenated into 24 fractions by combining every24th fraction. C18 StageTip concentrated samples were analyzed by reversed phase nanoflow HPLCwith nano-electrospray ionization using a LTQ-Orbitrap Velos Pro mass spectrometer operated inpositive ion mode with a 2 h reverse phase gradient per HPLC fraction. Each sample was rehydratedand samples were separated by on-line reversed phase liquid chromatography coupled on-line toan LTQ-Orbitrap Velos Pro mass spectrometer equipped with a Nanospray Flex source (ThermoFisher Scientific). Spectrum Selection was used to generate peak lists of the higher-energy collisionaldissociation (HCD) spectra (parameters: activation type: HCD; s\/n cut-off: 1.5; total intensity threshold:0; minimum peak count: 1; precursor mass: 350-5000 Da).2.10. Protein Data Processing and Sequence Database SearchingAll data was analyzed using Proteome Discoverer version 1.4. The peak lists were submitted to anin-house Mascot 2.4 (Matrix Science) server for database searching through the Proteome Discoverersoftware. All host data was searched against the mouse sequence database, Uniprot-Mouse database(43,908 sequences; 19,909,825 residues) using similar search parameters [35]. All bacterial data wassearched against Bacteroidetes (11363 entries) and Firmicutes (17039 entries). Scaffold (version 4.6.1,Proteome Software Inc., Portland OR), a software suite from Proteome Software was used for statisticalvalidation of MS\/MS based peptide and protein identifications. Scaffold software provides differentlevels of blocking in proteome analysis. Blocking is a statistical tool used to reduce biases and minimizevariances within a study. Scaffold provides four blocking levels, for example a single protein in theoriginal observation matrix can be summarized in terms of all spectra, unique spectra, unique peptidesand unique samples. Unique peptides is the preferred blocking level for analyzing the data [36],allowing users to compare measurements for each peptide. Since we pooled biological replicates tominimize biological variance, we do not have biological replicates in our study design. Therefore,we chose unique peptides as our statistical blocking method. Differential proteins, therefore, werepredicted using the differential peptides determined by Scaffold. Peptide identifications were acceptedif they could be established at greater than 95.0% probability by the Scaffold Local FDR algorithm andcontained at least two identified peptides for the host proteome and at least one identified peptidefor the bacterial proteome. Protein probabilities were assigned by the Protein Prophet algorithm [37].Spectra data were log-transformed, pruned of those matched to multiple proteins and those missing areference value, and weighted by an adaptive intensity weighting algorithm. Differentially expressedproteins were determined by applying Permutation Test with adjusted significance level P < 0.05corrected by Benjamini\u2013Hochberg.2.11. Ingenuity Pathways Analysis for Mucosal Host ProteinsIngenuity Pathways Analysis (IPA) was used to interpret the host proteome data in the contextof biological processes, pathways and networks. IPA infers hypothetical protein interaction clustersusing the Ingenuity Pathways Knowledge Base, a large database consisting of millions of individualrelationships between proteins. Given its proximity to the microbiome, the host mucosal proteomicsdata derived from the iTRAQ experiment was converted by IPA to \u2018fold change\u2019 and then uploadedinto the IPA program. No expression value cutoff was selected and both up- and down-regulatedidentifiers were defined as value parameters for the analysis. Heatmaps highlighting significantNutrients 2019, 11, 418 6 of 24downstream biological processes that are increased or decreased based on gene expression results aredisplayed as canonical pathways. To further explore connections between dietary intake and expressedgenes, hypothetical networks were generated followed by regulator effect analysis [38], using as manyproteins from the input expression profile as possible. Other proteins from the database were used tofill out a protein cluster when needed for a highly connected network as previously published [39].To identify mucosal phyla that correlate with the selected host proteins, a Spearman correlation matrixwas generated and plotted as a heatmap.2.12. Statistical AnalysisData is presented as mean \u00b1 standard deviation unless otherwise stated. The data was testedfor normality using Shapiro-Wilk test, and a Kruskal\u2013Wallis non-parametric test with a Benjamini\u2013Hochberg FDR-correction was used for comparing differences in the relative abundance ofGram-positive, Gram-negative, aerobic, anaerobic and facultatively anaerobic bacteria, Firmicutesto Bacteroidetes ratio, mobile elements and biofilm formers between dietary groups. SCFAs wereassessed using a Kruskal\u2013Wallis non-parametric test followed by a Dunn\u2019s multiple comparison test.2.13. Data Availability16S rRNA gene amplicon sequencing data is made available in the Genbank (SRA study ID:SRP082836). Metaproteome data is made available via ProteomeXchange for submucosal data(PXD008165) and mucosal data (PXD008152).2.14. Ethical ConsiderationsThe protocols used were approved by the Animal Care Committee of UBC under the protocolA15-0240 and in direct accordance with guidelines drafted by the Canadian Council on the Use ofLaboratory Animals.3. Results3.1. Dietary Lipid Type Affects Gut Microbial DiversityDysbiotic bacterial communities are often associated with low diversity [40], although a causalrelationship has not been established. Since the microbial composition of feces and mucosal tissuehave different microbiomes [41], we focused on the mucosal associated microbes as these microbes aremost likely respond to the dietary changes and have been suggested to be a reservoir for keystonespecies that contribute to disease activity [42]. Both the milk fat and corn oil diets resulted in increasedalpha diversity (Figure 1A). Specifically, observed species richness and Chao1 were increased withcorn oil and milk fat exposure whereas the olive oil diet resulted in similar richness to the low-fat chow.Similarly, Shannon\u2019s index revealed that milk fat had high richness and evenness compared to low-fatchow and olive oil groups. An increase in Simpson\u2019s index, indicating a decrease in evenness, wasobserved in all high-fat diets compared to the low-fat chow. These patterns of alpha diversity alignedwith comparisons between samples amongst dietary groups. The PCoA plot using the weightedUniFrac revealed three distinct clusters where the milk fat and corn oil groups clustered together andaway from olive oil and the low-fat groups (Figure 1B). A total of 80.8% of the overall variation intaxon composition was attributed to dietary exposure, of which the first and second axes explained71.6% and 9.2% of the total variation, respectively. While the permutational multivariate analysis ofvariance (PERMANOVA) based on the weighted UniFrac distance suggests that milk fat and cornoil groups are similar, the unweighted UniFrac showed separation between the milk fat and corn oilgroups suggesting that while the dominant species in the groups are similar, the rare species in themilk fat and corn oil groups are unique from each other (Figure 1C). Overall, the various dietary lipidseach uniquely predicted the microbial community composition that are present in the gut.Nutrients 2019, 11, 418 7 of 24Nutrients 2019, 11, x FOR PEER  7 of 25  Figure 1. The effect of lipid diets on the diversity of the gut microbiota. (A) Alpha diversity of colonic microbiota from mice fed high fats diets composed of low-fat chow (blue), milk fat (red), olive oil (purple) or corn oil (green). Observed species richness, Chao1, Shannon\u2019s, and Simpson\u2019s indexes are displayed. (B) PCoA plot of the weighted UniFrac distances of colonic microbial communities from mice fed high-fat diets composed of milk fat, olive oil, corn oil or low-fat chow. The first two principal components from the PCoA are plotted. (C) Statistical summary (p-values after Benjamini\u2013Hochberg adjustment for multiple comparisons) of all alpha and beta diversity measures. The differences in bacterial communities between the dietary cohorts were further evident when samples were ordered according to their Firmicutes to Bacteroidetes ratio. In agreement with previous literature [3,4], our results revealed that high-fat diets induce a microbiome with a high Firmicutes to Bacteroidetes ratio in the colon compared to the low-fat diet (Figure 2A), although we did not see a significant difference in body weight between any of the groups(Figure S1). These findings coincided with a predicted increase in the relative abundance of Gram-positive bacteria in the high-fat diets and a corresponding decrease in Gram-negative (Figure 2B). Additionally, our findings indicated there were changes to the relative abundances of facultative anaerobes, aerobes and anaerobes as a result of the type of fat feeding (Figure 2C). Olive oil diets associated with the least abundance of oxygen tolerating microbes, important given the hypothesis that oxygen tolerant microbes are abundant during gut stress [43]. Specifically, the predicted abundances of facultative anaerobic bacteria were higher in the low-fat dietary group compared to the olive oil and corn oil groups, and were higher in the milk fat group compared to the olive oil group. Finally, each dietary fat resulted in a unique set of taxa (Figure 2D). The low-fat diet had an increase in the abundance of Lachnospiraceae [Firmicutes (P = 0.01)], Aldercretzia spp. [Actinobacteria (P = 0.03)], family S24_7 [Bacteroidetes (P = 0.002)], and Ruminococcus spp. [Firmicutes (P = 0.005)]. Uniquely, olive oil resulted in an increased abundance of several Firmicutes including Clostridiaceae (P = 0.003), Peptostreptococcaceae (P = 0.01), Ruminococcaceae (P = 0.005), and Dorea spp (P = 0.003). In contrast, milk fat promoted different families of Firmicutes including Erysipelotrichales (P = 0.008) and several genera from Ruminicoccus (P = 0.003). Corn oil enhanced the abundance of Firmicutes family members Figure 1. The effect of lipid diets on the diversity of the gut microbiota. (A) Alpha diversity of colonicmicrobiota from mice fed high fats diets composed of low-fat chow (blue), milk fat (red), olive oil(purple) or corn oil (green). Observed species richness, Chao1, Shannon\u2019s, and Simpson\u2019s indexes aredisplayed. (B) PCoA plot of the weighted UniFrac distances of colonic microbial communities frommice fed high-fat diets composed of milk fat, olive oil, corn oil or low-fat chow. The first two principalcomponents from the PCoA are plotted. (C) Statistical summary (p-values after Benjamini\u2013Hochbergadjustment for multiple comparisons) of all alpha and beta diversity measures.The differences in bacterial communities between the dietary cohorts were further evident whensamples were ordered according to their Firmicutes to Bacteroidetes ratio. In agreement with previousliterature [3,4], our results revealed that high-fat diets induce a microbiome with a high Firmicutes toBacteroidetes ratio in the colon compared to the low-fat diet (Figure 2A), although we did not see asignificant difference in body weight between any of the groups (Figure S1). These findings coincidedwith a predicted increase in the relative abundance of Gram-positive bacteria in the high-fat dietsand a corresponding decrease in Gram-negative (Figure 2B). Additionally, our findings indicatedthere were changes to the relative abundances of facultative anaerobes, aerobes and anaerobes asa result of the type of fat feeding (Figure 2C). Olive oil diets associated with the least abundanceof oxygen tolerating microbes, important given the hypothesis that oxygen tolerant microbes areabundant during gut stress [43]. Specifically, the predicted abundances of facultative anaerobicbacteria were higher in the low-fat dietary group compared to the olive oil and corn oil groups, andwere higher in the milk fat group compared to the olive oil group. Finally, each dietary fat resulted in aunique set of taxa (Figure 2D). The low-fat diet had an increase in the abundance of Lachnospiraceae[Firmicutes (P = 0.01)], Aldercretzia spp. [Actinobacteria (P = 0.03)], family S24_7 [Bacteroidetes(P = 0.002)], and Ruminococcus spp. [Firmicutes (P = 0.005)]. Uniquely, olive oil resulted in an increasedabundance of several Firmicutes including Clostridiaceae (P = 0.003), Peptostreptococcaceae (P = 0.01),Ruminococcaceae (P = 0.005), and Dorea spp (P = 0.003). In contrast, milk fat promoted differentfamilies of Firmicutes including Erysipelotrichales (P = 0.008) and several genera from Ruminicoccus(P = 0.003). Corn oil enhanced the abundance of Firmicutes family members from Turicibacteraceae(P = 0.008) in addition to Coprococcus spp. (P = 0.002). Overall, high-fat diets resulted in analogousmodulation of the gut microbiota at higher taxonomic levels, but the type of fatty acid present in thedietary lipid uniquely altered the intestinal microbes at lower taxonomic levels.Nutrients 2019, 11, 418 8 of 24Nutrients 2019, 11, x FOR PEER  8 of 25  from Turicibacteraceae (P = 0.008) in addition to Coprococcus spp. (P = 0.002). Overall, high-fat diets resulted in analogous modulation of the gut microbiota at higher taxonomic levels, but the type of fatty acid present in the dietary lipid uniquely altered the intestinal microbes at lower taxonomic levels.  Figure 2. The effect of lipid diets on the gut microbial taxa. (A) Comparison of the log abundance of the Firmicutes to Bacteroidetes ratio among experimental diet groups in the colon. The y-axis of the box plot indicates the log of the abundance of the Firmicutes divided by the abundance of Bacteroidetes. The low-fat group had a significantly lower Firmicutes to Bacteroidetes ratio than all the high-fat diets. Within the high-fat diets, olive oil had a significantly lower ratio of Firmicutes to Bacteroidetes than the milk fat group. (B) Relative abundances of Gram-positive and Gram-negative Figure 2. The effect of lipid diets on the gut microbial taxa. (A) Comparison of the log abundance of theFir icutes to Bacteroidetes ratio a ong experi ental diet groups in the colon. The y-axis of the boxplot indicates the log of the abundance of the Firmicutes divided by the abundance of Bacteroidetes.The low-fat group had a significantly lower Firmicutes to Bacteroidetes ratio than all the high-fat diets.Within the high-fat diets, olive oil had a significantly lower ratio of Firmicutes to Bacteroidetes thanthe milk fat group. (B) Relative abundances of Gram-positive and Gram-negative bacteria in the dietgroups show a significantly lower abundance of gram positive bacteria and a corresponding higherabundance of gram negative bacteria in the low-fat dietary group. (C) Relative abundances of aerobic,Nutrients 2019, 11, 418 9 of 24anaerobic and facultatively anaerobic bacteria in the diet groups show significantly lower facultativeanaerobic bacteria in the olive oil group compared to the milk fat and low-fat group. An asterisk abovea single column indicates P < 0.05 for that group compared to every other group. An asterisk with aline connecting two groups indicates P < 0.05 between the two groups. (D) Differentially abundantmicrobial clades in the colon microbiota from mice fed high-fat diets composed of anhydrous milkfat, olive oil, corn oil or a low-fat normal chow. Cladogram represents taxonomic representation ofstatistically and biologically consistent differences among lipid diet groups. Significant differences arerepresented in the color of the most abundant class. Yellow circles represent non-significant microbialclades. The all-to-all version of LEfSe was used with Kruskal\u2013Wallis test (P < 0.05). LDA score thresholdwas set at default value 2.3.2. Dietary Lipid Type Confers Core Functionality to Each Microbial CommunitySince microbial compositions change according to type of lipid diets, we next investigatedhow lipids affect the functionality of the microbiota using amplicon sequencing predictions andcomparing SCFA metabolites. Amplicon sequencing functional content was predicted from markergenes (16S rRNA) and LDA was performed (Figure 3). The low-fat chow was predicted to enrichfunctions of the microbiome that included lipopolysaccharide biosynthesis, vitamin and cofactorbiosynthesis (including biotin metabolism, folate biosynthesis, pantothenate and CoA biosynthesis,lipoic acid metabolism, and riboflavin metabolism), protein export, and digestion and absorption.This suggests that all high-fat diets, despite the type of fatty acid, may have reduced capacity forvitamin biosynthesis and cofactor metabolism. The olive oil diet is predicted to result in a microbiomethat have an increased potential for pyruvate metabolism, enhanced synthesis and degradation ofketone bodies, butanoate metabolism and propanoate metabolism, enhanced lipid metabolism andabundant RIG-I-like receptor signaling important for viral immune recognition. The corn oil dietis predicted to result in a microbiome with functions characterized by increased flagellar assembly,ABC transporters, lipid metabolism (glycerolipid metabolism, sphingolipid metabolism, linoleic acidmetabolism), and carbohydrate metabolism for ATP production (pentose phosphate pathway, galactosemetabolism, starch and sucrose metabolism, fructose and mannose metabolism). Two componentsystems are also predicted be increased in the corn oil diet, which controls cellular processes suchas cell motility and virulence. This may suggest that corn oil may result in a microbiota that is moreinvasive. Predictions from the milk fat diets suggest the highest potential for bacterial chemotaxis andsimilar to the corn oil diet carbohydrate metabolism was also predicted to be higher. This includedglycolysis and gluconeogenesis, glyoxylate and dicarboxylate metabolism, C5-branched dibasic acidmetabolism, biosynthesis of unsaturated fatty acids and xenobiotic degradation (styrene, dioxin, andxylene). This suggests that milk fat results in a microbiota with increased capacity for energy harvest.Overall, the predicted functional analysis suggest that total calories from fat altered common functionalcharacteristics of the microbiota and that the type of lipid uniquely affected additional characteristics.Nutrients 2019, 11, 418 10 of 24Nutrients 2019, 11, x FOR PEER  10 of 25   Figure 3. The effect of lipid diets on predicted microbial functions. Statistically and biologicallydifferentially abundant pathways amongst the four dietary group shown as a histogram of the LDAscores. The length of the bars represents a log10 transformed LDA score set to a threshold value of 2.The one-to-all version of LEfSe was used with Kruskal\u2013Wallis test (P < 0.05).Nutrients 2019, 11, 418 11 of 24To determine if virulence attributes of the gut microbiome were modulated by high-fat diets,we used 16S OTUs to categorize functionality. We found that all high-fat diets predicted an increase inthe abundance of mobile genetic elements (Figure 4A); however, this finding should be interpretedcautiously as mobile elements are subject to microevolutionary processes and may vary over shortperiods of time [44]. The diets composed of corn oil, including the standard chow, predicted increasedlevels of biofilm formers (Figure 4B). Specifically, the abundance of biofilm formers in the low-fat dietwas significantly higher than the abundance of biofilm formers in the milk fat and olive oil groups butwere not statistically different from the corn oil dietary group. Since SCFA metabolism was predictedto be modulated based on the amplicon sequencing extrapolations, we examined the abundance ofcecal acetic, propionic and butyric acid to understand if the predicted changes in metabolic pathwaysaffected the bioavailability of SCFAs. We found that the milk fat group had similar levels of SCFA asthe low-fat chow groups whereas both olive oil and corn oil groups resulted in a decreased abundanceof acetic acid, important for lipid biosynthesis, and propionic acid, important for gluconeogenesis,compared to the low-fat chow, respectively (Figure 4C). A similar trend was observed with abundanceof butyric acid. Overall, these results suggest that the microbiome\u2019s ability to yield SCFAs resultingfrom high calories of fat can be compensated via exposure to milk fat.Nutrients 2019, 11, x FOR PEER  11 of 25  Figure 3. The effect of lipid diets on predicted microbial functions. Statistically and biologically differentially abundant pathways amongst the four dietary group shown as a histogram of the LDA scores. The length of the bars represents a log10 transformed LDA score set to a threshold value of 2. The one-to-all version of LEfSe was used with Kruskal\u2013Wallis test (P < 0.05). To determine if virulence attributes of the gut microbiome were modulated by high-fat diets, we used 16S OTUs to categorize functionality. We found that all high-fat diets predicted an increase in the abundance of mobile genetic elements (Figure 4A); however, this finding should be interpreted cautiously as mobile elements are subject to microevolutionary processes and may vary over short periods of time [44]. The diets composed of corn oil, including the standard chow, predicted increased levels of biofilm formers (Figure 4B). Specifically, the abundance of biofilm formers in the low-fat diet was significantly higher than the abundance of biofilm formers in the milk fat and olive oil groups but were not statistically different from the corn oil dietary group. Since SCFA metabolism was predicted to be modulated based on the amplicon sequencing extrapolations, we examined the abundance of cecal acetic, propionic and butyric acid to understand if the predicted changes in metabolic pathways affected the bioavailability of SCFAs. We found that the milk fat group had similar levels of SCFA as the low-fat chow groups whereas both olive oil and corn oil groups resulted in a decreased abundance of acetic acid, important for lipid biosynthesis, and propionic acid, important for gluconeogenesis, compared to the low-fat chow, respectively (Figure 4C). A similar trend was observed with abundance of butyric acid. Overall, these results suggest that the microbiome\u2019s ability to yield SCFAs resulting from high calories of fat can be compensated via exposure to milk fat. Figure 4. Predicted bacterial virulence traits and quantified secondary metabolites. Virulence traits such as (A) the relative abundance of bacteria which contain mobile elements and (B) the relative abundance of bacteria which are able to form biofilms are displayed for each diet group. (C) The effect of lipid diets on short-chain fatty acid production. Short-chain fatty acid analysis performed via gas chromatography on cecal samples from mice fed high-fat diets composed of milk fat, olive oil, corn oil or a low-fat chow. Acetic, propionic, and butyric acid are expressed as mass % of total cecal tissue sample. Values are expressed as mean +\/\u2212 SEM (n = 8\u201312). An asterisk above a single column indicates Figure 4. Pre icted bacterial virulence traits and quantified secondary metabolites. Virulence traitssuch as (A) the relative abundance of bacteria which contain mobile eleme ts and (B) the relativeabundance of b cteria which are able to for biofilms are isplayed for each diet group. (C) Theeffect of lipid diets on short-chain fatty acid production. Short-chain f tty acid analysis performedvi gas chromatography on cecal sa ples from mice fed high-fat diets composed of milk fat, oliveoil, corn oil or a low-fat chow. A etic, propionic, and buty ic acid are expressed s mass % of totalcecal tissue sample. Values are expressed as mean +\/\u2212 SEM (n = 8\u201312). An asterisk above a singlecolumn indicates P < 0.05 for that group compared to every other dietary group. An asterisk with aline connecting two groups indicates P < 0.05 between the two groups.Nutrients 2019, 11, 418 12 of 243.3. Dietary Lipids Alter Microbial and Host Proteins in the ColonTo understand the interactions between lipid diets, gut bacteria and the host, we performedmetaproteomics to examine microbial and host proteins associated with the colonic mucosa andsubmucosa. Over 300 bacterial proteins were identified in the mucosa (Table S1) and 112 were identifiedin the submucosa based on single peptide hits (Table S2). However, it is currently recommendedthat a minimum of four peptides are required to be matched for positive protein identification,to decrease the number of false positives [45]. Based on this recommendation the only bacterialprotein we could positively identify was the mucosal molecular chaperone dnaK protein which issignificantly upregulated in the corn oil group (0.4-fold increase) compared to the low fat and oliveoil groups (supplemental materials). In stark contrast to the bacterial proteome, 1956 host proteinswere identified in the submucosa and 1749 were identified in the mucosa based on a single peptidehit. Of these, 676 and 390 were confidently identified using four or more peptides with a P value\u22640.006 in the submucosa (Table S3) and mucosa (Table S4), respectively. Overall, there was lowhomogeneity between the submucosal and mucosal proteins with only 127 proteins overlappingthe two biological niches. Overlapping proteins (Table S5) largely included proteins important forhost fatty acid metabolism such as Apolipoprotein A-1, Apolipoprotein E, fatty acid-binding protein,fatty acid synthase and 2,-4-dienoyl-CoA reductase; proteins important for cellular function such asribosomal proteins, anion exchange proteins and endoplasmic reticulum resident proteins; proteinsinvolved in epithelial remodeling such as cadherin-17 and vinculin; and proteins involved in mucosaldefense and immunity including complement C3, and mucin-2.3.3.1. High-Fat Diets Associated with Decreased Death Receptor Signaling and Apoptosis andtRNA ChargingMolecular crosstalk between the commensal microbiota and the intestinal epithelial cells occursat the intestinal mucosal surface. As such, we focused our investigation on host proteins expressed inthe mucosa. To understand higher ranking response pathways due to different lipid diets, mucosalproteins were evaluated using IPA [46] which identifies the most significant canonical pathways,biological functions, and networks. After generating the pathway comparison heat map, we rankedthe effects of each diet and ordered the results in descending order based on the high-fat corn oil diet(Figure 5A). The IPA heatmap highlights that all high-fat diets have decreased predicted pathwaysassociated with cell death, indicated by the down-regulation of apoptosis signaling and death receptorsignaling pathways. These findings were based on the overall down-regulation of cell death proteinssuch as apoptotic chromatin condensation inducer 1, cytochrome c somatic, lamin A\/C, spectrin alphanon-erythrocytic 1, calpain 1, mitogen-activated protein kinase 1 and heat shock protein family B(small) member 1 (Table 1). While not included in the IPA pathway, increased interleukin-1 receptorantagonist in the corn oil and milk fat group, has also been shown to reduce apoptosis.Transfer RNA charging was similarly down-regulated in all high-fat diets. This was based onthe overall down-regulation of glutamyl-prolyl-tRNA synthetase, phenylalanyl-tRNA synthetasebeta subunit, lysyl-tRNA synthetase, asparaginyl-tRNA synthetase, arginyl-tRNA synthetase,threonyl-tRNA synthetase, valyl-tRNA synthetase and tyrosyl-tRNA synthetase. In contrast, all high-fat diets, had upregulated peroxisome proliferator activated receptor (PPAR)\u03b1\/ retinoid X receptor(RXR)\u03b1 pathways compared to the low-fat control. IPA selected proteins used to determine PPARactivation included: acyl-CoA oxidase 1, apolipoprotein A1, cytochrome P450 family 2 subfamily Cmember 18, fatty acid synthase, glycerol-3-phosphate dehydrogenase 1, heat shock protein 90 betafamily member 1, mitogen-activated protein kinase 1, and protein disulfide isomerase family A member3. Overall, increased consumption of fat regardless of the saturation index, results in decreased celldeath and tRNA charging signaling and increased PPAR\u03b1\/RXR\u03b1 activation signaling.Nutrients 2019, 11, 418 13 of 24Nutrients 2019, 11, x FOR PEER  18 of 25  Figure 5. Effects of lipid diets on the gut proteome shown by the Ingenuity pathway comparative analysis. (A) Heatmap visualization of metabolites detected in each dietary group. Orange color indicates a higher activation score, whereas blue color indicates a lower activation score. Ingenuity pathway analysis (IPA) identified many upstream regulators predicted to be active based on the gene expression profile including: (B) bleeding in the high-fat corn oil and (C) low-fat chow groups, (D) contractility of muscles in the corn oil group and (E) tumorigenesis of tissue in the olive oil group. Faded colors represent less of an effect. 3.1.3.3. Milk Fat Diet is Associated with Increased Inflammation and Compensating Restitution The milk fat dietary group had upregulated acute phase response (Figure 5A). Acute phase proteins are defined as proteins which are increased by at least 25 percent during inflammation and includes proteins such as apolipoprotein A1, ferritin, haptoglobin, interleukin-1 receptor antagonist protein, and serpin family A member 3, which were all upregulated in the milk fat group (Table 1). While not included in the IPA derived pathway, alpha-1 acid glycoprotein 1 was similarly upregulated and is involved in acute phase response. The milk fat group was the only high-fat diet to have upregulated fatty acid \u03b2-oxidation 1 and sirtuin signaling. Previous studies have shown that the sirtuin signaling pathways link inflammation and metabolism particularly protective restitutive responses helping to resolve inflammation [48]. Of the 22 proteins utilized in the IPA sirtuin signaling pathways, 6 were upregulated in the milk fat group including ATP synthase F1 subunit beta, H1 Figure 5. Effects of lipid diets on the gu proteome shown by th Ingenuity p thway comparativeanalysis. (A) Heatmap visualization of metabolites detected in each dietary group. Orange colorindicates a higher ctivati n score, whereas blue color indic tes a lower ctivati n score. Ingenuityp thw y analysis (IPA) identified many upstream regulators predicted to be active based on thegene expression profile including: (B) bleedin in the high-fat corn oi and (C) l -fat chow groups,(D) contractility of muscles in the corn oil group and (E) tumorigenesi of tissu in the live il group.Faded colors represent less of an effect.Nutrients 2019, 11, 418 14 of 24Table 1. Mucosal proteins contributing to IPA pathways and networks.Pathway Symbol Gene Name Low Fat Milk Fat Olive Oil Corn OilHigh fatDeathReceptor ACIN1apoptotic chromatin condensationinducer 1 0.4 \u22120.2 0 \u22120.3signaling CYCS cytochrome c, somatic 0.7 0.3 \u22120.2 \u22120.2HSPB1 heat shock protein family B (small)member 1 \u22120.5 0.5 0.1 \u22120.2LMNA lamin A\/C 0.6 \u22120.2 \u22120.1 \u22120.4SPTAN1 spectrin alpha, non-erythrocytic 1 0.6 \u22120.1 \u22120.1 \u22120.2Apoptosis ACIN1 apoptotic chromatin condensationinducer 1 0.4 \u22120.2 0 \u22120.3CAPN1 calpain 1 0.8 0 \u22120.1 \u22120.1CYCS cytochrome c, somatic 0.7 0.3 \u22120.2 \u22120.2LMNA lamin A\/C 0.6 \u22120.2 \u22120.1 \u22120.4MAPK1 mitogen-activated protein kinase 1 0.2 0 \u22120.2 \u22120.1SPTAN1 spectrin alpha, non-erythrocytic 1 0.6 \u22120.1 \u22120.1 \u22120.2IL1RN Interleukin-1 receptor antagonist protein \u22121 0.4 -0.1 0.7tRNAcharging EPRS glutamyl-prolyl-tRNA synthetase 0.8 \u22120.1 \u22120.2 0FARSB phenylalanyl-tRNA synthetase betasubunit 0.7 0 \u22120.2 \u22120.1KARS lysyl-tRNA synthetase 0.8 \u22120.2 \u22120.3 \u22120.2NARS asparaginyl-tRNA synthetase 0.9 \u22120.3 \u22120.5 \u22120.4RARS arginyl-tRNA synthetase 0.7 \u22120.1 \u22120.2 0TARS threonyl-tRNA synthetase 0.8 0 \u22120.2 \u22120.1VARS valyl-tRNA synthetase 0.4 0 \u22120.1 \u22120.2YARS tyrosyl-tRNA synthetase 0.5 \u22120.4 \u22120.3 \u22120.2PPARa\/RXRa ACOX1 acyl-CoA oxidase 1 0.5 0 \u22120.3 \u22120.4Activation APOA1 apolipoprotein A1 \u22120.7 0.2 \u22120.1 0.5CYP2C18 cytochrome P450 family 2 subfamilyC member 18 0.3 \u22120.4 0.4 \u22121.7FASN fatty acid synthase 0 0 0 0.4GPD1 glycerol-3-phosphate dehydrogenase1 1.3 \u22120.6 \u22120.4 \u22120.3HSP90B1 heat shock protein 90 beta familymember 1 0.2 \u22120.4 \u22120.3 \u22120.1MAPK1 mitogen-activated protein kinase 1 0.2 0 \u22120.2 \u22120.1PDIA3 protein disulfide isomerase family Amember 3 \u22120.5 0 0 0.1Corn oilGlycolysis I ALDOB aldolase, fructose-bisphosphate B 1.4 \u22120.5 \u22120.6 \u22120.5ENO1 enolase 1 \u22120.5 0.2 0.1 0.3FBP2 fructose-bisphosphatase 2 0.2 \u22120.2 \u22120.2 0.1TPI1 triosephosphate isomerase 1 \u22120.6 0 0 0.4Oxidative ATP5F1B ATP synthase F1 subunit beta \u22120.8 0.2 0 0.5phosphorylation ATP5PB ATP synthase peripheralstalk-membrane subunit b 0.6 \u22120.1 \u22120.1 \u22120.3ATP5PO ATP synthase peripheral stalksubunit OSCP \u22120.7 0.1 0.1 0.4COX5A cytochrome c oxidase subunit 5A \u22120.9 0.3 0.2 0.6CYCS cytochrome c, somatic 0.7 0.3 \u22120.2 \u22120.2NDUFA9 NADH:ubiquinone oxidoreductasesubunit A9 0.8 0.2 0.2 \u22120.1NDUFS1 NADH:ubiquinone oxidoreductasecore subunit S1 \u22120.4 \u22120.2 0 0.4NDUFS2 NADH:ubiquinone oxidoreductasecore subunit S2 0.6 0 0 \u22120.1NDUFS3 NADH:ubiquinone oxidoreductasecore subunit S3 \u22120.8 0.1 0.1 0.3NDUFV2 NADH:ubiquinone oxidoreductasecore subunit V2 \u22120.3 \u22120.1 \u22120.1 0.3UQCRB ubiquinol-cytochrome c reductasebinding protein 0.2 \u22120.3 \u22120.3 0Nutrients 2019, 11, 418 15 of 24Table 1. Cont.Pathway Symbol Gene Name Low Fat Milk Fat Olive Oil Corn OilCorn oilUQCRC2 ubiquinol-cytochrome c reductasecore protein 2 0.3 \u22120.1 \u22120.1 0.1NRF2-mediated CBR1 carbonyl reductase 1 \u22120.4 0.1 0.1 0.2oxidativestress CCT7chaperonin containing TCP1 subunit7 0.5 \u22120.2 \u22120.2 \u22120.3response DNAJB11 DnaJ heat shock protein family(Hsp40) member B11 0.7 \u22120.2 \u22120.4 \u22120.5FTH1 ferritin heavy chain 1 \u22120.4 0.3 0 0.1FTL ferritin light chain \u22120.2 0.2 0.2 0.3GSR glutathione-disulfide reductase 0.6 \u22120.1 \u22120.2 0.1GSTM3 glutathione S-transferase mu 3 1.2 0.3 0.4 0.3MAPK1 mitogen-activated protein kinase 1 0.2 0 \u22120.2 \u22120.1SOD1 superoxide dismutase 1 0.6 \u22120.1 \u22120.2 0.2USP14 ubiquitin specific peptidase 14 0.6 \u22120.2 \u22120.2 \u22120.1CA3 Carbonic anhydrase 3 0 \u22120.1 0 0.7ALDH2 Aldehyde dehydrogenase \u22120.6 0.2 0 0.6Glutathione-mediated ANPEP alanyl aminopeptidase, membrane 1.8 \u22121.1 \u22120.8 \u22120.8detoxification GGH gamma-glutamyl hydrolase 0.6 0.2 \u22120.1 0.6Gsta4 glutathione S-transferase, alpha 4 0.4 0 0 \u22120.5GSTM3 glutathione S-transferase mu 3 1.2 0.3 0.4 0.3GSTZ1 glutathione S-transferase zeta 1 \u22120.3 \u22120.1 0.1 0.5ILK signaling ACTN1 actinin alpha 1 0.3 0.1 0.1 \u22120.3ACTN4 actinin alpha 4 0.6 \u22120.3 \u22120.3 \u22120.2DSP desmoplakin 0.6 \u22120.1 0 \u22120.3FLNA filamin A 0.4 0.2 0.3 \u22120.4FLNC filamin C 0.7 0.1 0.1 \u22120.6FN1 fibronectin 1 0.7 \u22120.1 0.1 \u22120.8MAPK1 mitogen-activated protein kinase 1 0.2 0 \u22120.2 \u22120.1MYH9 myosin heavy chain 9 0.6 \u22120.2 \u22120.2 \u22120.2MYH11 myosin heavy chain 11 0.7 0.2 0.4 \u22120.6MYH14 myosin heavy chain 14 0.5 \u22120.1 \u22120.1 \u22120.2MYL9 myosin light chain 9 \u22120.5 0.4 0.6 \u22120.2PPP2R1A protein phosphatase 2 scaffoldsubunit Alpha \u22120.5 0.1 0.2 0.5VCL vinculin 0.6 \u22120.1 0.1 \u22120.4Epithelialintegrity Muc2 mucin-2 0.2 \u22120.1 \u22120.2 \u22120.6Cing cingulin 0.6 \u22120.3 \u22120.2 \u22120.3VEGFsignaling ACTN1 actinin alpha 1 0.3 0.1 0.1 \u22120.3ACTN4 actinin alpha 4 0.6 \u22120.3 \u22120.3 \u22120.2EIF2S3 eukaryotic translation initiationfactor 2 subunit \u03b3 0.4 \u22120.1 \u22120.1 \u22120.2ELAVL1 ELAV like RNA binding protein 1 0.7 \u22120.1 \u22120.1 \u22120.1MAPK1 mitogen-activated protein kinase 1 0.2 0 \u22120.2 \u22120.1VCL vinculin 0.6 \u22120.1 0.1 \u22120.4Bleedingnetwork APOE apolipoprotein E \u22120.7 0.2 \u22120.1 0.5CNN1 cluster of calponin-1 \u22120.1 0.5 0.5 \u22120.2FLNA filamin-a 0.4 0.2 0.3 \u22120.4MYH9 cluster of myosin-9 0.6 \u22120.2 \u22120.2 \u22120.2PLEC cluster of plectin 0.4 \u22120.1 0 \u22120.5IL1RN interleukin-1 receptor antagonistprotein \u22121 0.4 \u22120.1 0.7Contractilityof musclenetworkATP2A2 sarcoplasmic\/endoplasmicreticulum calcium ATPase 0.5 \u22120.2 \u22120.1 \u22120.4CKM cluster of creatine kinase M-type \u22120.3 0.2 0.3 \u22120.3DES cluster of desmin 0.6 0.4 0.3 \u22120.5MYH11 cluster of myosin-11 0.7 0.2 0.4 \u22120.6MYH14 myosin-14 0.5 \u22120.1 \u22120.1 \u22120.2VCL vinculin 0.6 \u22120.1 0.1 \u22120.4Nutrients 2019, 11, 418 16 of 24Table 1. Cont.Pathway Symbol Gene Name Low Fat Milk Fat Olive Oil Corn OilMilk fatAcute PhaseResponse APOA1 apolipoprotein A1 \u22120.7 0.2 \u22120.1 0.5C3 complement C3 0.4 \u22120.1 \u22120.8 \u22120.7FN1 fibronectin 1 0.7 \u22120.1 0.1 \u22120.8FTL ferritin light chain \u22120.2 0.2 0.2 0.3HP haptoglobin 0.8 0.2 \u22121.6 \u22121.3IL1RN interleukin 1 receptor antagonist \u22121 0.4 \u22120.1 0.7MAPK1 mitogen-activated protein kinase 1 0.2 0 \u22120.2 \u22120.1SERPINA3 serpin family A member 3 \u22120.6 0.6 \u22121.4 \u22121.1AAG1 alpha-1 acid glycoprotein 1 0.8 0.4 -0.8 -0.6Sirtuinsignaling ADAM10 ADAM metallopeptidase domain 10 0.4 \u22120.1 \u22120.2 0APEX1 apurinic\/apyrimidinicendodeoxyribonuclease 1 0.7 \u22120.1 \u22120.2 \u22120.3ATP5F1B ATP synthase F1 subunit beta \u22120.8 0.2 0 0.5ATP5PB ATP synthase peripheralstalk-membrane subunit b 0.6 \u22120.1 \u22120.1 \u22120.3CPS1 carbamoyl-phosphate synthase 1 2.3 \u22122 \u22121.3 \u22121.9H1F0 H1 histone family member 0 \u22120.5 1.2 0.4 \u22120.6Hist1h1e histone cluster 1, H1e 0.3 1.1 0.1 \u22120.6HMGCS2 3-hydroxy-3-methylglutaryl-CoAsynthase 2 \u22120.4 0 \u22120.2 \u22121.5MAPK1 mitogen-activated protein kinase 1 0.2 0 \u22120.2 \u22120.1NAMPT nicotinamidephosphoribosyltransferase 0.6 \u22120.1 \u22120.2 \u22120.3NDUFA9 NADH:ubiquinone oxidoreductasesubunit A9 0.8 0.2 0.2 \u22120.1NDUFS1 NADH:ubiquinone oxidoreductasecore subunit S1 \u22120.4 \u22120.2 0 0.4NDUFS2 NADH:ubiquinone oxidoreductasecore subunit S2 0.6 0 0 \u22120.1NDUFS3 NADH:ubiquinone oxidoreductasecore subunit S3 \u22120.8 0.1 0.1 0.3NDUFV2 NADH:ubiquinone oxidoreductasecore subunit V2 \u22120.3 \u22120.1 \u22120.1 0.3PDHA1 pyruvate dehydrogenase E1 alpha 1subunit \u22120.3 0 0 0.1SF3A1 splicing factor 3a subunit 1 0.7 0 0.1 \u22120.1SLC25A5 solute carrier family 25 member 5 0.9 0 0 \u22120.2SOD1 superoxide dismutase 1 0.6 \u22120.1 \u22120.2 0.2TIMM13 translocase of inner mitochondrialmembrane 13 \u22120.2 0 0 0.3UQCRC2 ubiquinol-cytochrome c reductasecore protein 2 0.3 \u22120.1 \u22120.1 0.1VDAC1 voltage dependent anion channel 1 0.1 0.3 \u22120.1 0.2Fatty acid Boxidation ACAA2 acetyl-CoA acyltransferase 2 \u22120.4 0.1 0.2 0HADHAhydroxyacyl-CoA dehydrogenasetrifunctional multienzyme complexsubunit alpha\u22120.1 0.2 0.1 0.2HADHBhydroxyacyl-CoA dehydrogenasetrifunctional multienzyme complexsubunit beta\u22120.4 0.2 0.1 0.1IVD isovaleryl-CoA dehydrogenase \u22120.6 \u22120.1 0 0.2Olive oilActincytoskeleton ACTN1 actinin alpha 1 0.3 0.1 0.1 \u22120.3signaling ACTN4 actinin alpha 4 0.6 \u22120.3 \u22120.3 \u22120.2ARPC5 actin related protein 2\/3 complexsubunit 5 \u22120.4 0.2 0 0.4FLNA filamin A 0.4 0.2 0.3 \u22120.4FN1 fibronectin 1 0.7 \u22120.1 0.1 \u22120.8Nutrients 2019, 11, 418 17 of 24Table 1. Cont.Pathway Symbol Gene Name Low Fat Milk Fat Olive Oil Corn OilOlive oilIQGAP2 IQ motif containing GTPaseactivating protein 2 0.7 0 \u22120.1 \u22120.2MAPK1 mitogen-activated protein kinase 1 0.2 0 \u22120.2 \u22120.1MYH9 myosin heavy chain 9 0.6 \u22120.2 \u22120.2 \u22120.2MYH11 myosin heavy chain 11 0.7 0.2 0.4 \u22120.6MYH14 myosin heavy chain 14 0.5 \u22120.1 \u22120.1 \u22120.2MYL9 myosin light chain 9 \u22120.5 0.4 0.6 \u22120.2VCL vinculin 0.6 \u22120.1 0.1 \u22120.4Col6a3 cluster of protein Col6a3 \u22120.1 -0.6 1.1 \u22121.1Tumorigenesisof ACOX1 acyl-coenzyme A oxidase 1 0.5 0 \u22120.3 \u22120.4tissue network APOA1 apolipoprotein a-1 \u22120.7 0.2 \u22120.1 0.5ATP2A2 sarcoplasmic\/endoplasmicreticulum calcium ATPase2 0.5 \u22120.2 \u22120.1 \u22120.4C3 complement C3 0.4 \u22120.1 \u22120.8 \u22120.7HP hippocalcin-like protein 1 0.8 \u22120.2 \u22120.2 0.2IL1RN interleukin-1 receptor antagonistprotein \u22121 0.4 \u22120.1 0.7MTTP microsomal triglyceride transferprotein large subunit 2.4 \u22122 \u22121.7 \u22122.1PC pyruvate carboxylase \u22120.2 0.2 0.1 0.4Displayed are the experimental log ratios. Compared to the pooled reference channel (value normalized tozero), positive values indicate an increased fold-change expression whereas negative values indicate a decreasedfold-change expression Italicized proteins were not included in IPA pathway. NA indicates the protein did notreach the 0.006 threshold. All groups are compared to the pool which was set to zero. Abbreviations used: OSCP,oligomycin sensitivity conferral protein; VEGF, vascular endothelial growth factor.3.3.2. Corn Oil Diets Show Responses Indicative of Increased Energy Requirements and OxidativeStress, and Decreased Barrier FunctionTwo of the most highly affected host pathways in the corn oil group were glycolysis I andoxidative phosphorylation, indicating increased energy demand in the mucosal epithelial cells of cornoil fed mice (Figure 5A). These pathways were determined through upregulated proteins involvedin glycolysis, such as enolase 1, fructose-bisphosphatase 2 and triosephosphate isomerase 1 (Table 1).Similarly, proteins involved in oxidative phosphorylation such as ATP synthase F1 subunit beta,ATP synthase peripheral stalk subunit OSCP (oligomycin sensitivity conferral protein), cytochromec oxidase subunit 5A, and nicotinamide adenine dinucleotide (NADH): ubiquinone oxidoreductasecore subunit S3 were upregulated in the high-fat corn oil and milk fat dietary groups. Other pathwaysheightened in corn oil diets are nuclear factor (erythroid-derived 2)-like 2 (NRF2) mediated oxidativestress response, and glutathione-mediated detoxification. IPA determined NRF2 mediated oxidativestress in the n-6 PUFA rich diets through the expression of glutathione-disulfide reductase, glutathioneS-transferase mu 3 and superoxide dismutase 1. In contrast to the low-fat diet, the high-fat diet also hadincreased expression of carbonyl reductase 1, ferritin heavy chain 1 and ferritin light chain. Carbonicanhydrase 3 (CA3) and aldehyde dehydrogenase (ALDH2) are similarly increased in the corn oilgroup but were not included in the NRF2-mediated oxidative stress pathway. Overall, increasedconsumption of n-6 PUFA diets show responses indicative of increased energy demands and oxidativestress. While all high-fat diets have down-regulated integrin-linked kinase (ILK) signaling, this wasespecially pronounced in the corn oil group which showed decreased expression of actinin alpha 1 and4, desmoplakin, filamin A and C, fibronectin 1, mitogen-activated protein kinase 1, myosin heavy chain9, 11, and 14, myosin light chain 9 and vinculin. This is important because ILK signaling has been foundto be indispensable for barrier function [47]. Other proteins important for epithelial integrity includemucins and proteins involved in junctional complexes. Here, we found that in addition to the priormentioned proteins, Mucin-2 (Muc2; fragments) and cingulin were down-regulated in the high-fatdietary groups, particularly in corn oil. This decreased barrier function is not limited to the epithelium.Nutrients 2019, 11, 418 18 of 24The pathways show decreased vascular endothelial growth factor (VEGF) signaling in the corn oiland olive oil group which corresponds with the network generated by IPA predicting that the corn oildiet would increase bleeding based on the down-regulation of cluster of calponin-1 (CCN1), filamin-a(FLNA), cluster of myosin-9 (MYH9), and cluster of plectin (PLEC) proteins and the upregulationof interleukin-1 receptor antagonist protein (IL1RN) and apolipoprotein E (Figure 5B). In contrast,low-fat diets had FLNA, MYH9, and PLEC and down-regulated IL1RN, CCN1 and APOE, and waspredicted to inhibit bleeding pathways (Figure 5C). Significant mucosal networks generated by IPAalso predicted that high-fat corn oil diets inhibit contractility of muscle based on the down-regulationof Sarcoplasmic\/endoplasmic reticulum calcium ATPase (ATP2A2), Cluster of Creatine kinase M-type(CKM), Cluster of Desmin (DES), Cluster of Myosin-11 (MYH11), myosin-14 (MYH14) and vinculinproteins (VCL) (Figure 5D). Taken together, diets rich in corn oil appear to have decreased barrierfunction, increased oxidative stress and require increased energy for maintenance.3.3.3. Milk Fat Diet is Associated with Increased Inflammation and Compensating RestitutionThe milk fat dietary group had upregulated acute phase response (Figure 5A). Acute phaseproteins are defined as proteins which are increased by at least 25 percent during inflammation andincludes proteins such as apolipoprotein A1, ferritin, haptoglobin, interleukin-1 receptor antagonistprotein, and serpin family A member 3, which were all upregulated in the milk fat group (Table 1).While not included in the IPA derived pathway, alpha-1 acid glycoprotein 1 was similarly upregulatedand is involved in acute phase response. The milk fat group was the only high-fat diet to haveupregulated fatty acid \u03b2-oxidation 1 and sirtuin signaling. Previous studies have shown that the sirtuinsignaling pathways link inflammation and metabolism particularly protective restitutive responseshelping to resolve inflammation [48]. Of the 22 proteins utilized in the IPA sirtuin signaling pathways,6 were upregulated in the milk fat group including ATP synthase F1 subunit beta, H1 histone familymember 0, NADH: ubiquinone oxidoreductase subunit A9, NADH: ubiquinone oxidoreductase coresubunit S3, and voltage dependent anion channel 1. In contrast to the corn oil and olive oil dietarygroups, IPA did not generate a hypothesis to explain how activation or inactivation of regulators leadsto an increase or decrease of function in the milk fat group. Given these data, milk fat appears to haveincreased expression of inflammatory pathways.3.3.4. Olive Oil Consumption Was Associated with Increased Cytoskeletal DynamicsSimilar to the low-fat group, the olive oil group had increased proteins involved in actincytoskeleton signaling and epithelial integrity (Figure 5A) due to increases in actinin alpha 1,filamin A, fibronectin 1, myosin heavy chain 11, myosin light chain 9 and vinculin. Additionally,the olive oil group had upregulated cluster of protein Col6a3 involved in microfibril formation.Predicted gene interaction networks show that olive oil was associated with inflammation ofliver (tumorigenesis of tissue), commonly caused by virial infections. This prediction was basedon the down-regulation of peroxisomal acyl-coenzyme A oxidase 1 (ACOX1), apolipoprotein a-1(APOA1), Sarcoplasmic\/endoplasmic reticulum calcium ATPase 2 (ATP2A2), complement C3 (C3),hippocalcin-like protein 1 (HP), interleukin-1 receptor antagonist protein (IL1RN), microsomaltriglyceride transfer protein large subunit (MTTP), and pyruvate carboxylase (PC) proteins (Figure 5E).3.4. Microbial Taxa Associate with Host ProteinsTo understand potential interactions between the bacteriome and host mucosal proteins,we evaluated Spearman correlations between mean phyla abundance and the selected mucosal proteinscontributing to IPA pathways and networks (Figure S2). The heatmap shows that low expressionof Proteobacteria and Tenericutes inversely correlates with apoptosis in high-fat diets. Increasedrelative abundance of TM7 in the corn oil group correlated with several host proteins involved inglycolysis, oxidative phosphorylation and NRF2-mediated oxidative stress response. With respect tothe olive oil diet, higher relative abundances of Bacteroidetes positively correlated with 42% of theNutrients 2019, 11, 418 19 of 24proteins involved in actin cytoskeleton signaling. Finally, 55% of the acute phase response proteins inthe milk fat group were associated with the relative abundances of Proteobacteria, Tenericutes andVerrucomicrobia. However, because the peptides were pooled, there were only four observationsavailable for the Spearman Rank correlation analysis, one for each diet group. As a result, we cannotrealistically draw conclusions from this correlative data, but rather advocate for well-controlled anddesigned experiments that ask specific questions based on the observations made here.4. DiscussionThe mammalian gut has co-evolved with trillions of microorganisms, the collection of which isreferred to as the gut microbiome. We have yet to understand how microbes succeed in the gut as aconsortium and then co-exist in a community and affect the host responses. It has been hypothesizedthat several external factors, including diet, play a role in the host-microbe interaction in the gut. Severalstudies over the past few years have shown that a high-fat diet can lead to different gut microbialprofiles, yet the effects on bacterial taxa and their functional responses caused by distinct types of fattyacids are not well understood. To define the specific changes in bacterial taxa as well as functionaloutputs, we analyzed the effect of commonly consumed dietary lipids on the colonic microbiome.While there were differences between the high-fat and low-fat diets suggesting calories may playa role, these diets are not directly comparable since the macronutrients and micronutrients are differentin the low-fat diet. By changing the amount of dietary fat, the proportion of carbohydrates andproteins automatically changes making it difficult to disentangle lipid driven changes. For example,the availability of dietary carbohydrates has been shown to modulate biofilm development [49]and acquisition of plasmids encoding relevant metabolic pathways (mobile genetic elements) [50]which could account for the predicted differences between the \u201clow\u201d and high-fat diets. Therefore,we focus on changes observed between the high-fat diets. We found that each type of dietary lipiddistinctly affected the clustering effects of the microbial communities indicating that different taxathrive with exposure to the types of fatty acids. While all high-fat diets caused an increase in theabundance of Firmicutes, each dietary lipid promoted specific taxa within the phyla with differingfunctions, indicating that different Firmicutes species thrive in the presence of different lipid substrates.Promoting the growth of certain bacterial species through diet, or prebiotics, has primarily beendocumented in carbohydrates [51], yet the potential for other macronutrients to act as prebiotics haslargely been unexplored. Correspondingly, SCFA production was also altered as a result of the type oflipid consumed. Propionic acid and acetic acid were suppressed in the corn oil and olive oil group,respectively, whereas the milk fat group had similar levels of SCFA production as the low-fat control.Our data also indicates that higher species richness is observed in mice fed corn oil and milk fat diets;however, the differential composition and predicted functions of the gut microbiota do not seem tobe associated with better health outcomes. This is apparent in the corn oil group which promoted amicrobiota with high invasive and infection potential. In support of this, previous studies from our labhave shown that the corn oil diet promotes exacerbated immune-driven damage when challenged withCitrobacter rodentium [11], whereas olive oil consumption is protective [9] despite the low microbialdiversity shown here. Therefore, diversity alone may not be a predictor for a better health.The data presented in this work show that diet induced changes to the microbiome mirrorsdiet induced responses from the host. For instance, the predicted increase in invasive potentialobserved in the corn oil bacteriome parallels the predicted decrease in pathogen resistance and barrierfunction observed in the host. Specifically, the corn oil diet increased the microbial diversity in thegut that was predicted to increase microbial virulence traits such as increased microbial motility andbacterial signal transduction by two component regulatory systems. The host proteome indicated theprotective barrier protein MUC2 was decreased alongside proteins important for tight and adherinsjunctions, and endothelial integrity (bleeding). Furthermore, decreased peristalsis (contractility ofmuscle), and increased oxidative stress response predicted in the corn oil group may be in responseto increased microbial invaders. This supports the phenotype observed in previous studies using aNutrients 2019, 11, 418 20 of 24similar diet which showed n-6 PUFA results in increased oxidative stress and tissue damage [11,52],increased inflammation and mortality during enteric infection [9,11], and metabolic insufficiencies [53].Our data, in combination with previous literature, indicates that increased n-6 PUFA in the dietmay be a risk factor for the development of a dysfunctional barrier in the gut. While descriptive,the data presented here provides a potential mechanism (bacterial-host interactions) by which cornoil, rich in n-6 PUFA, imparts toxicity in the gut. Indeed, an overabundance of dietary n-6 PUFApromotes chronic inflammation [54] and excessive consumption of n-6 PUFA is a risk factor for IBDin humans [10]. Prospective cohort studies conducted over a 5-year period demonstrated that PUFApositively associated with UC risk [55]. Similarly, retrospective case-control studies found increasedlevels of IBD in people consuming diets rich in n-6 PUFA [56]. Our research and others support theobservations made here that n-6 PUFA tends to increase gut inflammation and damage resultingin an exacerbated colitis in several animal models [11,57\u201361]. Currently, we do not understand themechanisms behind n-6 PUFA being detrimental during colitis but this study does reveal pathwaysthat need to be investigated further.Similar parallels between the microbiome and host responses were observed in the milk fatgroup. Specifically, the milk fat diet increased microbes in the gut whose functions are involved incarbohydrate and lipid metabolism. This was reflected in the host by increased proteins involvedin fatty acid \u03b2-oxidation. Mounting evidence supports that sirtuins link metabolism and hostinflammation. While inflammation is required to defend against invading organisms, compensatorymechanisms are required to prevent chronic inflammation. Host sirtuins, increased in the milkfat group, improve outcomes in chronic inflammatory diseases and sepsis by \u2018mending\u2019 the hostor promoting restitution through immune repression and restoring homeostasis following stressresponses [48]. Moreover, while the milk fat diet resulted in a host-microbe relationship thatpromoted host inflammation, there were no significant decreases in protective microbial SCFAresponses suggesting that both the host and the commensal microbes promote a homeostaticinflammation-resolution cycle. This supports previous studies showing increased pathology in themilk fat group during infection but also increased compensatory protective responses, unlike the cornoil group [9].This relationship between the microbiome, host and dietary lipids is not limited to the bacteriomeand there is evidence that the virome may similarly be involved. In support of this, sequencingdata predicted upregulated RIG-I-like receptor signaling pathways in the olive oil dietary group.The RIG-I-like family of pattern recognition receptors identify viral RNA [62], and are importantfor virus-host signaling crosstalk. Host mucosal proteins in the olive oil cohort also predicted liverinflammation (tumorigenesis of tissue) which is often caused by viral infections. Previous experimentshave shown viral infections can inhibit C3 complement production [63], and that the loss of IL1RN canenhance susceptibility to viral infections [64], which may suggest an interaction between dietary oliveoil, the host and the virome. However, 16S rRNA gene amplicon sequencing predictions face severallimitations, one of which is the inability to study viral microbiome community members. As such,future studies should test this potential relationship under controlled settings while specificallytargeting the virome.5. ConclusionsOverall, we conclude that the type of dietary lipids distinctly impacts the gut microbiome. Whilehigh fat consumption has a distinct impact on the gut microbiota as compared to a normal chow diet,the type of fatty acids alters the relative microbial abundances where olive oil was most distinct fromthe corn oil and milk fat. The corn oil and milk fat diets shared similarities in diversity but had differentfunctional characteristics. We show that the corn oil diet, rich in n-6 PUFA, resulted in a microbiomewith enhanced predicted virulence and pathogenicity associated with increased host inflammation,oxidative stress and increased barrier dysfunction. While the milk fat diet, rich in SFA, resulted in ahost-microbe relationship that promoted inflammation which could result in inflammatory inducedNutrients 2019, 11, 418 21 of 24intestinal damage, there was a compensatory protective response evident by the production of sirtuinsand SCFAs. In marked contrast to both corn oil and milk fat, the olive oil diet, rich in MUFA, resulted ina host-microbe dynamic suggesting the involvement of the less-explored virome. These results suggestthat fat type is an important consideration for gut health and not all high-fat diets are detrimental.However, given this study is descriptive in nature, future studies should focus on well-designedexperiments unraveling the mechanisms of each lipid on gut health. These results have the potentialto guide evidence-based nutrition recommendations for IBD patients who can suffer from nutrientdeficiencies from overly restrictive dietary regimes including low-fat diets.Supplementary Materials: The following are available online at http:\/\/www.mdpi.com\/2072-6643\/11\/2\/418\/s1,Figure S1: Mouse Weights, Figure S2: Spearman rank correlation depicting correlations between selected hostproteins and mucosal phyla, Table S1: All bacterial mucosal proteins identified by 1 or more peptides; Table S2: Allbacterial submucosal proteins identified by 1 or more peptides; Table S3: All host submucosal proteins identifiedby 4 or more peptides; TabS4: All host mucosal proteins identified by 4 or more peptides; TableS5: All matchinghost mucosal and submucosal proteins identified by 4 or more peptides.Author Contributions: N.A. performed QIIME, PICRUSt analysis and proteomic experiments and analysis,performed the statistical analysis, generated figures, wrote and revised the manuscript. C.Q. assisted withproteomic experiments, prepared the figures, critically analyzed the data and wrote and revised the manuscript.K.B. prepared the DNA amplicons, conducted QIIME analysis, helped generate figures and write and edit themanuscript. Y.K.C. performed the animal work, prepared the DNA library amplicons and helped write themanuscript. S.K.G. performed the short-chain fatty acid analysis and edited the manuscript. D.L.G. designedand supervised the project, funded the project and wrote the manuscript. All authors read and approved thefinal manuscript.Funding: This research was funded by Crohn\u2019s and Colitis Canada and the Natural Science and EngineeringResearch Council.Acknowledgments: We thank Chaubin Dai, Amy Botta and Sanjoy Ghosh for contributions to maintaining theanimal colonies and feeding experiments. We also thank the D.I.A.B.E.T.E.S. center for lipid analysis services.Conflicts of Interest: The authors declare no conflicts of interest.References1. Vernocchi, P.; Vannini, L.; Gottardi, D.; Del Chierico, F.; Serrazanetti, D.I.; Ndagijimana, M.; Guerzoni, M.E.Integration of datasets from different analytical techniques to assess the impact of nutrition on humanmetabolome. Front. Cell. Infect. Microbiol. 2012, 2, 156. [CrossRef] [PubMed]2. Chan, Y.K.; Estaki, M.; Gibson, D.L. Clinical consequences of diet-induced dysbiosis. Ann. Nutr. Metab. 2013,63, 28\u201340. [CrossRef] [PubMed]3. Ley, R.E.; Turnbaugh, P.J.; Klein, S.; Gordon, J.I. Microbial ecology: Human gut microbes associated withobesity. Nature 2006, 444, 1022\u20131023. [CrossRef] [PubMed]4. Turnbaugh, P.J.; Gordon, J.I. The core gut microbiome, energy balance and obesity. J. Physiol. 2009, 587,4153\u20134158. [CrossRef] [PubMed]5. Donovan, S.M.; Wang, M.; Li, M.; Friedberg, I.; Schwartz, S.L.; Chapkin, R.S. Host-microbe interactions in theneonatal intestine: Role of human milk oligosaccharides. Adv. Nutr. (Bethesda) 2012, 3, 450S\u2013455S. [CrossRef][PubMed]6. Psaltopoulou, T.; Kosti, R.I.; Haidopoulos, D.; Dimopoulos, M.; Panagiotakos, D.B. Olive oil intake isinversely related to cancer prevalence: A systematic review and a meta-analysis of 13,800 patients and 23,340controls in 19 observational studies. Lipids Health Dis. 2011, 10, 127. [CrossRef] [PubMed]7. Schwingshackl, L.; Hoffmann, G. Monounsaturated fatty acids and risk of cardiovascular disease: Synopsis ofthe evidence available from systematic reviews and meta-analyses. Nutrients 2012, 4, 1989\u20132007. [CrossRef][PubMed]8. Camuesco, D.; Galvez, J.; Nieto, A.; Comalada, M.; Rodriguez-Cabezas, M.E.; Concha, A.; Xaus, J.;Zarzuelo, A. Dietary olive oil supplemented with fish oil, rich in epa and dha (n-3) polyunsaturatedfatty acids, attenuates colonic inflammation in rats with dss-induced colitis. J. Nutr. 2005, 135, 687\u2013694.[CrossRef] [PubMed]Nutrients 2019, 11, 418 22 of 249. DeCoffe, D.; Quin, C.; Gill, S.K.; Tasnim, N.; Brown, K.; Godovannyi, A.; Dai, C.; Abulizi, N.; Chan, Y.K.;Ghosh, S.; et al. Dietary lipid type, rather than total number of calories, alters outcomes of enteric infectionin mice. J. Infect. Dis. 2016, 213, 1846\u20131856. [CrossRef]10. Tjonneland, A.; Overvad, K.; Bergmann, M.M.; Nagel, G.; Linseisen, J.; Hallmans, G.; Palmqvist, R.; Sjodin, H.;Hagglund, G.; Berglund, G.; et al. Linoleic acid, a dietary n-6 polyunsaturated fatty acid, and the aetiologyof ulcerative colitis: A nested case-control study within a european prospective cohort study. Gut 2009, 58,1606\u20131611.11. Ghosh, S.; Molcan, E.; DeCoffe, D.; Dai, C.; Gibson, D.L. Diets rich in n-6 pufa induce intestinal microbialdysbiosis in aged mice. Br. J. Nutr. 2013, 110, 515\u2013523. [CrossRef] [PubMed]12. Ghosh, S.; DeCoffe, D.; Brown, K.; Rajendiran, E.; Estaki, M.; Dai, C.; Yip, A.; Gibson, D.L. Fish oilattenuates omega-6 polyunsaturated fatty acid-induced dysbiosis and infectious colitis but impairs lpsdephosphorylation activity causing sepsis. PLoS ONE 2013, 8, e55468. [CrossRef] [PubMed]13. Micha, R.; Khatibzadeh, S.; Shi, P.; Fahimi, S.; Lim, S.; Andrews, K.G.; Engell, R.E.; Powles, J.; Ezzati, M.;Mozaffarian, D.; et al. Global, regional, and national consumption levels of dietary fats and oils in 1990and 2010: A systematic analysis including 266 country-specific nutrition surveys. BMJ 2014, 348, g2272.[CrossRef]14. Opstelten, J.L.; Leenders, M.; Dik, V.K.; Chan, S.S.; van Schaik, F.D.; Khaw, K.T.; Luben, R.; Hallmans, G.;Karling, P.; Lindgren, S.; et al. Dairy products, dietary calcium, and risk of inflammatory bowel disease:Results from a european prospective cohort investigation. Inflamm. Bowel Dis. 2016, 22, 1403\u20131411. [CrossRef][PubMed]15. Devkota, S.; Wang, Y.; Musch, M.W.; Leone, V.; Fehlner-Peach, H.; Nadimpalli, A.; Antonopoulos, D.a.;Jabri, B.; Chang, E.B. Dietary-fat-induced taurocholic acid promotes pathobiont expansion and colitis inil10-\/- mice. Nature 2012, 487, 104\u2013108. [CrossRef] [PubMed]16. Ohira, H.; Fujioka, Y.; Katagiri, C.; Mamoto, R.; Aoyama-Ishikawa, M.; Amako, K.; Izumi, Y.; Nishiumi, S.;Yoshida, M.; Usami, M.; et al. Butyrate attenuates inflammation and lipolysis generated by the interaction ofadipocytes and macrophages. J. Atheroscler. Thromb. 2013, 20, 425\u2013442. [CrossRef] [PubMed]17. Venkatraman, A.; Ramakrishna, B.S.; Pulimood, A.B.; Patra, S.; Murthy, S. Increased permeability in dextransulphate colitis in rats: Time course of development and effect of butyrate. Scand. J. Gastroenterol. 2000, 35,1053\u20131059. [CrossRef] [PubMed]18. Rabbani, G.H.; Albert, M.J.; Hamidur Rahman, A.S.; Moyenul Isalm, M.; Nasirul Islam, K.M.; Alam, K.Short-chain fatty acids improve clinical, pathologic, and microbiologic features of experimental shigellosis.J. Infect Dis. 1999, 179, 390\u2013397. [CrossRef] [PubMed]19. Brown, K.; Godovannyi, A.; Ma, C.; Zhang, Y.; Ahmadi-Vand, Z.; Dai, C.; Gorzelak, M.A.; Chan, Y.;Chan, J.M.; Lochner, A.; et al. Prolonged antibiotic treatment induces a diabetogenic intestinal microbiomethat accelerates diabetes in nod mice. ISME J. 2016, 10, 321. [CrossRef] [PubMed]20. Caporaso, J.G.; Kuczynski, J.; Stombaugh, J.; Bittinger, K.; Bushman, F.D.; Costello, E.K.; Fierer, N.; Pena, A.G.;Goodrich, J.K.; Gordon, J.I.; et al. Qiime allows analysis of high-throughput community sequencing data.Nat Methods 2010, 7, 335\u2013336. [CrossRef] [PubMed]21. Edgar, R.C. Expected Errors Predicted by Phred (q) Scores. Available online: https:\/\/www.drive5.com\/usearch\/manual\/exp_errs.html (accessed on 15 February 2019).22. Edgar, R.C. Search and clustering orders of magnitude faster than blast. Bioinformatics 2010, 26, 2460\u20132461.[CrossRef] [PubMed]23. Caporaso, J.G.; Bittinger, K.; Bushman, F.D.; DeSantis, T.Z.; Andersen, G.L.; Knight, R. Pynast: A flexible toolfor aligning sequences to a template alignment. Bioinformatics 2010, 26, 266\u2013267. [CrossRef] [PubMed]24. Langille, M.G.; Zaneveld, J.; Caporaso, J.G.; McDonald, D.; Knights, D.; Reyes, J.A.; Clemente, J.C.;Burkepile, D.E.; Vega Thurber, R.L.; Knight, R.; et al. Predictive functional profiling of microbial communitiesusing 16s rrna marker gene sequences. Nat Biotechnol. 2013, 31, 814\u2013821. [CrossRef] [PubMed]25. Segata, N.; Izard, J.; Waldron, L.; Gevers, D.; Miropolsky, L.; Garrett, W.S.; Huttenhower, C. Metagenomicbiomarker discovery and explanation. Genome Biol. 2011, 12, R60. [CrossRef] [PubMed]26. Lozupone, C.; Knight, R. Unifrac: A new phylogenetic method for comparing microbial communities.Appl. Environ. Microbiol. 2005, 71, 8228\u20138235. [CrossRef] [PubMed]27. Anderson, M.J. A new method for non-parametric multivariate analysis of variance. Austral Ecology 2008, 26,32\u201346. [CrossRef]Nutrients 2019, 11, 418 23 of 2428. Kanehisa, M.; Goto, S.; Sato, Y.; Furumichi, M.; Tanabe, M. Kegg for integration and interpretation oflarge-scale molecular data sets. Nucleic Acids Res. 2012, 40, D109\u2013D114. [CrossRef]29. Ward, T.; Larson, J.; Meulemanns, J.; Hillmann, B.; Lynch, J.; Sidiropoulos, D.; Spear, J.; Caporaso, G.;Blekman, R.; Knight, R.; et al. Bugbase predicts organism level microbiome phenotypes. bioRxiv 2017.[CrossRef]30. Luo, J.; Zheng, A.; Meng, K.; Chang, W.; Bai, Y.; Li, K.; Cai, H.; Liu, G.; Yao, B. Proteome changes in theintestinal mucosa of broiler (gallus gallus) activated by probiotic enterococcus faecium. J. Proteom. 2013, 91,226\u2013241. [CrossRef] [PubMed]31. Datta, A.; Chai, Y.L.; Tan, J.M.; Lee, J.H.; Francis, P.T.; Chen, C.P.; Sze, S.K.; Lai, M.K.P. An itraq-basedproteomic analysis reveals dysregulation of neocortical synaptopodin in lewy body dementias. Mol. Brain2017, 10, 36. [CrossRef] [PubMed]32. Datta, A.; Qian, J.; Chong, R.; Kalaria, R.N.; Francis, P.; Lai, M.K.; Chen, C.P.; Sze, S.K. Novelpathophysiological markers are revealed by itraq-based quantitative clinical proteomics approach in vasculardementia. J. Proteom. 2014, 99, 54\u201367. [CrossRef] [PubMed]33. Datta, A.; Sze, S.K. Data for itraq profiling of micro-vesicular plasma specimens: In search of potentialprognostic circulatory biomarkers for lacunar infarction. Data Brief 2015, 4, 510\u2013517. [CrossRef]34. Yang, W.; Woltjer, R.L.; Sokal, I.; Pan, C.; Wang, Y.; Brodey, M.; Peskind, E.R.; Leverenz, J.B.; Zhang, J.;Perl, D.P.; et al. Quantitative proteomics identifies surfactant-resistant alpha-synuclein in cerebral cortex ofparkinsonism-dementia complex of guam but not alzheimer\u2019s disease or progressive supranuclear palsy.Am. J. Pathol. 2007, 171, 993\u20131002. [CrossRef]35. Randall, D.R.; Park, P.S.; Chau, J.K. Identification of altered protein abundances in cholesteatoma matrixvia mass spectrometry-based proteomic analysis. J. Otolaryngol Head Neck Surg. 2015, 44, 50. [CrossRef][PubMed]36. Forbes, J.D.; Van Domselaar, G.; Bernstein, C.N. The gut microbiota in immune-mediated inflammatorydiseases. Front Microbiol. 2016, 7, 1081. [CrossRef]37. Nesvizhskii, A.; Keller, A.; Kolker, E.; Aebersold, R. A statistical model for identifying proteins by tandemmass spectrometry. Anal. Chem. 2003, 75, 4646\u20134658. [CrossRef] [PubMed]38. Chen, M.; Mistry, D. Ipa: Maximizing the Biological Interpretation of Gene Transcript & Protein ExpressionData with IPA, 2018. Available online: https:\/\/bioinformatics.rockefeller.edu\/wp-content\/uploads\/IPA-Training-Dev.pdf (accessed on 15 February 2019).39. Overgaard, A.J.; Thingholm, T.E.; Larsen, M.R.; Tarnow, L.; Rossing, P.; McGuire, J.N.; Pociot, F. Quantitativeitraq-based proteomic identification of candidate biomarkers for diabetic nephropathy in plasma of type 1diabetic patients. Clin. Proteom. 2010, 6, 105\u2013114. [CrossRef]40. Brown, K.; DeCoffe, D.; Molcan, E.; Gibson, D.L. Diet-induced dysbiosis of the intestinal microbiota and theeffects on immunity and disease. Nutrients 2012, 4, 1095\u20131119. [CrossRef] [PubMed]41. Zhang, S.; Cao, X.; Huang, H. Sampling strategies for three-dimensional spatial community structures in ibdmicrobiota research. Front. Cell. Infect. Microbiol. 2017, 7, 51. [CrossRef]42. Tang, M.S.; Poles, J.; Leung, J.M.; Wolff, M.J.; Davenport, M.; Lee, S.C.; Lim, Y.A.; Chua, K.H.; Loke, P.; Cho, I.Inferred metagenomic comparison of mucosal and fecal microbiota from individuals undergoing routinescreening colonoscopy reveals similar differences observed during active inflammation. Gut Microbes 2015,6, 48\u201356. [CrossRef]43. Baumler, A.J.; Sperandio, V. Interactions between the microbiota and pathogenic bacteria in the gut. Nature2016, 535, 85\u201393. [CrossRef] [PubMed]44. Brito, I.L.; Yilmaz, S.; Huang, K.; Xu, L.; Jupiter, S.D.; Jenkins, A.P.; Naisilisili, W.; Tamminen, M.; Smillie, C.S.;Wortman, J.R.; et al. Mobile genes in the human microbiome are structured from global to individual scales.Nature 2016, 535, 435\u2013439. [CrossRef] [PubMed]45. Damodaran, S.; Wood, T.D.; Nagarajan, P.; Rabin, R.A. Evaluating peptide mass fingerprinting-based proteinidentification. Genom. Proteom. Bioinform. 2007, 5, 152\u2013157. [CrossRef]46. IPA. Ingenuity Pathways Analysis Software. Available online: http:\/\/www.ingenuity.com (accessed on 15February 2019).47. Sayedyahossein, S.; Rudkouskaya, A.; Leclerc, V.; Dagnino, L. Integrin-linked kinase is indispensablefor keratinocyte differentiation and epidermal barrier function. J. Investig. Dermatol 2016, 136, 425\u2013435.[CrossRef]Nutrients 2019, 11, 418 24 of 2448. Vachharajani, V.T.; Liu, T.; Wang, X.; Hoth, J.J.; Yoza, B.K.; McCall, C.E. Sirtuins link inflammation andmetabolism. J. Immunol. Res. 2016, 2016, 8167273. [CrossRef] [PubMed]49. Santana, I.L.; Goncalves, L.M.; de Vasconcellos, A.A.; da Silva, W.J.; Cury, J.A.; Del Bel Cury, A.A. Dietarycarbohydrates modulate candida albicans biofilm development on the denture surface. PLoS ONE 2013, 8,e64645. [CrossRef]50. O\u2019Donnell, M.M.; O\u2019Toole, P.W.; Ross, R.P. Catabolic flexibility of mammalian-associated lactobacilli. Microb.Cell Fact. 2013, 12, 48. [CrossRef] [PubMed]51. Petschow, B.; Dore, J.; Hibberd, P.; Dinan, T.; Reid, G.; Blaser, M.; Cani, P.D.; Degnan, F.H.; Foster, J.;Gibson, G.; et al. Probiotics, prebiotics, and the host microbiome: The science of translation. Ann. N. Y.Acad. Sci. 2013, 1306, 1\u201317. [CrossRef]52. Ghosh, S.; Sulistyoningrum, D.C.; Glier, M.B.; Verchere, C.B.; Devlin, A.M. Altered glutathione homeostasisin heart augments cardiac lipotoxicity associated with diet-induced obesity in mice. J. Biol. Chem. 2011, 286,42483\u201342493. [CrossRef] [PubMed]53. Wong, C.K.; Botta, A.; Pither, J.; Dai, C.; Gibson, W.T.; Ghosh, S. A high-fat diet rich in corn oil reducesspontaneous locomotor activity and induces insulin resistance in mice. J. Nutr. Biochem. 2015, 26, 319\u2013326.[CrossRef]54. Simopoulos, A.; Cleland, L.G. Omega-6\/omega-3 essential fatty acid ratio: The scientific evidence. World Rev.Nutr. Diet. 2003, 92, 1\u2013194.55. Hart, A.R.; Luben, R.; Olsen, A.; Tjonneland, A.; Linseisen, J.; Nagel, G.; Berglund, G.; Lindgren, S.; Grip, O.;Key, T.; et al. Diet in the aetiology of ulcerative colitis: A european prospective cohort study. Digestion 2008,77, 57\u201364. [CrossRef] [PubMed]56. Sakamoto, N.; Kono, S.; Wakai, K.; Fukuda, Y.; Satomi, M.; Shimoyama, T.; Inaba, Y.; Miyake, Y.; Sasaki, S.;Okamoto, K.; et al. Dietary risk factors for inflammatory bowel disease: A multicenter case-control study injapan. Inflamm. Bowel Dis. 2005, 11, 154\u2013163. [CrossRef]57. Innis, S.M.; Dai, C.; Wu, X.; Buchan, A.M.; Jacobson, K. Perinatal lipid nutrition alters early intestinaldevelopment and programs the response to experimental colitis in young adult rats. Am. J. Physiol.Gastrointest. Liver Physiol. 2010, 299, G1376\u2013G1385. [CrossRef]58. Wan, J.M.; Grimble, R.F. Effect of dietary linoleate content on the metabolic response of rats to escherichiacoli endotoxin. Clin. Sci. (Lond) 1987, 72, 383\u2013385. [CrossRef]59. Andoh, A.; Tsujikawa, T.; Ishizuka, I.; Araki, Y.; Sasaki, M.; Koyama, S.; Fujiyama, Y. N-3 fatty acid-rich dietprevents early response of interleukin-6 elevation in trinitrobenzene sulfonic acid-induced enteritis. Int. J.Mol. Med. 2003, 12, 721\u2013725. [CrossRef]60. Hekmatdoost, A.; Wu, X.; Morampudi, V.; Innis, S.M.; Jacobson, K. Dietary oils modify the host immuneresponse and colonic tissue damage following citrobacter rodentium infection in mice. Am. J. Physiol.Gastrointest. Liver Physiol. 2013, 304, G917\u2013G928. [CrossRef] [PubMed]61. Kim, I.W.; Myung, S.J.; Do, M.Y.; Ryu, Y.M.; Kim, M.J.; Do, E.J.; Park, S.; Yoon, S.M.; Ye, B.D.; Byeon, J.S.; et al.Western-style diets induce macrophage infiltration and contribute to colitis-associated carcinogenesis.J. Gastroenterol. Hepatol. 2010, 25, 1785\u20131794. [CrossRef]62. Ramos, H.J.; Gale, M., Jr. Rig-i like receptors and their signaling crosstalk in the regulation of antiviralimmunity. Curr. Opin. Virol. 2011, 1, 167\u2013176. [CrossRef]63. Mazumdar, B.; Kim, H.; Meyer, K.; Bose, S.K.; Di Bisceglie, A.M.; Ray, R.B.; Ray, R. Hepatitis c virus proteinsinhibit c3 complement production. J. Virol. 2012, 86, 2221\u20132228. [CrossRef] [PubMed]64. Hill-Batorski, L.; Halfmann, P.; Marzi, A.; Lopes, T.J.; Neumann, G.; Feldmann, H.; Kawaoka, Y. Loss ofinterleukin 1 receptor antagonist enhances susceptibility to ebola virus infection. J. Infect. Dis. 2015, 212,S329\u2013S335. [CrossRef] [PubMed]\u00a9 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open accessarticle distributed under the terms and conditions of the Creative Commons Attribution(CC BY) license (http:\/\/creativecommons.org\/licenses\/by\/4.0\/).","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/hasType":[{"value":"Article","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt":[{"value":"10.14288\/1.0379196","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/language":[{"value":"eng","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#peerReviewStatus":[{"value":"Reviewed","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/provider":[{"value":"Vancouver : University of British Columbia Library","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/publisher":[{"value":"Multidisciplinary Digital Publishing Institute","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#publisherDOI":[{"value":"10.3390\/nu11020418","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/rights":[{"value":"CC BY 4.0","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#rightsURI":[{"value":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#scholarLevel":[{"value":"Faculty","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/subject":[{"value":"Host-microbe interactions","type":"literal","lang":"en"},{"value":"Gut microbiome","type":"literal","lang":"en"},{"value":"Dietary lipids","type":"literal","lang":"en"},{"value":"Polyunsaturated fatty acids","type":"literal","lang":"en"},{"value":"Monounsaturated fatty acids","type":"literal","lang":"en"},{"value":"Saturated fatty acids","type":"literal","lang":"en"},{"value":"Proteome","type":"literal","lang":"en"},{"value":"16S rRNA gene amplicon sequencing","type":"literal","lang":"en"},{"value":"Short-chain fatty acid metabolism","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/title":[{"value":"Gut Mucosal Proteins and Bacteriome Are Shaped by the Saturation Index of Dietary Lipids","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/type":[{"value":"Text","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#identifierURI":[{"value":"http:\/\/hdl.handle.net\/2429\/70425","type":"literal","lang":"en"}]}}