Human milk oligosaccharide 2’-fucosyllactose protects against high-fat diet-induced obesity by changing intestinal mucus production, composition and degradation linked to changes in gut microbiota and faecal proteome profiles in mice

Objective To decipher the mechanisms by which the major human milk oligosaccharide (HMO), 2’-fucosyllactose (2’FL), can affect body weight and fat mass gain on high-fat diet (HFD) feeding in mice. We wanted to elucidate whether 2’FL metabolic effects are linked with changes in intestinal mucus production and secretion, mucin glycosylation and degradation, as well as with the modulation of the gut microbiota, faecal proteome and endocannabinoid (eCB) system. Results 2’FL supplementation reduced HFD-induced obesity and glucose intolerance. These effects were accompanied by several changes in the intestinal mucus layer, including mucus production and composition, and gene expression of secreted and transmembrane mucins, glycosyltransferases and genes involved in mucus secretion. In addition, 2’FL increased bacterial glycosyl hydrolases involved in mucin glycan degradation. These changes were linked to a significant increase and predominance of bacterial genera Akkermansia and Bacteroides, different faecal proteome profile (with an upregulation of proteins involved in carbon, amino acids and fat metabolism and a downregulation of proteins involved in protein digestion and absorption) and, finally, to changes in the eCB system. We also investigated faecal proteomes from lean and obese humans and found similar changes observed comparing lean and obese mice. Conclusion Our results show that the HMO 2’FL influences host metabolism by modulating the mucus layer, gut microbiota and eCB system and propose the mucus layer as a new potential target for the prevention of obesity and related disorders.

portal and cava veins.Then, the mice were immediately euthanized by cervical dislocation.
One segment of colon from each mouse was opened, without flushing it before, for the collection of the mucus layer by gently scraping with a microscope glass slide and then weighed.

Biochemical Analysis
To determine the plasma insulin concentration, blood was harvested from the tip of the tail vein using capillaries prior to glucose load (−30 min) and 15 min after glucose load.Plasma insulin concentration was measured using an ELISA kit (Mercodia, Uppsala, Sweden), according to the manufacturer's instructions.Insulin resistance index was determined by multiplying the area under the curve of the blood glucose (−30 to 15 min) and plasma insulin (-30 min and 15 min).

Plasma Multiplex Analysis
Plasma levels of glucagon-like peptide 1 (GLP-1), peptide YY (PYY), ghrelin, leptin and glucagon were measured from the portal vein by multiplex assay kits based on chemiluminescence detection and following manufacturer's instructions (Meso Scale Discovery (MSD), Gaithersburg, MD, USA).Analyses were performed using a QuickPlex SQ 120 instrument (MSD) and DISCOVERY WORKBENCH® 4.0 software (MSD, Rockville, MD, USA).

RNA Preparation and gene expression analysis by real-time qPCR analysis
Total RNA was prepared from tissues using TriPure reagent (Roche).Quantification and integrity analysis of total RNA was performed by running 1 μl of each sample on an Agilent 2100 Bioanalyzer (Agilent RNA 6000 Nano Kit, Agilent).cDNA was prepared by reverse transcription of 1 μg total RNA using a Reverse Transcription System kit (Promega, Leiden, The Netherlands).Real-time PCRs were performed with the StepOnePlus real-time PCR system and software (Applied Biosystems, Den Ijssel, The Netherlands) using Mesa Fast qPCR sybr green mix (Eurogentec, Seraing, Belgium) and with the CFX Manager 3.1 software (Bio-BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) Rad, Hercules, CA) using Mesa Fast qPCR (GoTaq qPCR Master Mix, Promega, Madison, WI, USA) for detection, according to the manufacturer's instructions.RPL19 was chosen as housekeeping gene.All samples were run in duplicate in a single 96-well reaction plate, and data were analyzed according to the 2-ΔΔCt method.The identity and purity of the amplified product was checked through analysis of the melting curve carried out at the end of amplification.Primer sequences for the targeted mouse genes are available in Supplemental Table 9.

Analysis of the mucus layer thickness, goblet cells and immunohistochemistry
Colon segments were immediately removed and fixed in Carnoy's solution (ethanol 6: acid acetic 3: chloroform 1, vol/vol) for 2h at 4 °C.They were then immersed in ethanol 100% for 24 h.For the analysis of the mucus layer thickness and goblet cells, paraffin sections of 5 μm were stained with alcian blue.Images were captured at × 20 magnification and obtained using a SNC400 slide scanner and digital Image Hub software 561 (Leica Biosystems, Wetzlar, Germany).Analyses were performed using ImageJ (version 1.48r, National Institutes of Health, Bethesda, Maryland, USA) in a blinded manner.For the mucus layer thickness, two to six fields were used for each mouse and a minimum of 20 different measurements were made perpendicular to the inner mucus layer per field.For the goblet cells, the luminal side, muscularis mucosae, submucosa and muscle layer were removed and the blue area and the total area were measured separately in the remaining mucosal part of the colon.The proportion of the goblet cells was quantified based on the ratio between the blue area over the total area.

Histology and Fluorescent in situ hybridization
Segments of the distal colon from mice were fixed in water-free Methanol-Carnoy's fixative [60% methanol, 10% chloroform and 30% acetic acid] before paraffin embedding.Paraffin sections were dewaxed with Xylene substitute and hybridized with a general bacterial probe, EUB 338 conjugated to C3 (Merck, Ref: MBD0033).Immunostaining after hybridizations was performed with anti-MUC2C3 antiserum as described previously [4].Pictures were obtained with a LSM800 confocal microscope from Zeiss.

Bacterial distance and density
The bacterial penetration of the mucus was assessed using two parameters: the distance from the bacterial front to the epithelial cells and the density of bacterial cells within the inner mucus layer.The location of the bacterial front was easily delineated as the outermost border of the zone with high intensity for bacterial stain.The inner mucus was defined as the MUC2 positive layer between the bacterial front and the epithelial cells.To assess the first parameter, at least 10 pictures from different locations of at least 2 different distant sections were analyzed per mouse, with at least 10 measurements (distance between bacterial front and closest epithelial cell) taken per pictures to determine the average distance between the bacterial front and the apical side of the epithelial cells.For the second parameter, the bacterial density of the inner mucus, the area of the MUC2 positive layer between the bacterial front and the epithelial cells was measured and bacteria within this layer were counted manually by two independent investigators in a blinded manner.For this analysis, at least 5 pictures were analyzed per mouse.Analyses were performed using (Fiji Is Just) ImageJ 2.14.0/1.54fFor Mac OS and 2.14.0 for Windows.Measurements were first averaged per section, then per mouse, then per group.

Mucin glycan extraction and composition
Colonic mucus was suspended in 400 μl mucin extraction buffer (0.2 M Tris, pH 8, 1% SDS, 10 mM DTT).The samples were incubated at 60°C for 90 min.Iodoacetamide was added from a 1M stock solution to a final concentration of 100 mM.The samples were incubated at RT for 90 min in dark.The reduced samples were spin filtered through a 100k MWCO amicon 0.5 filter (merck) for 15 min at 14000 g.Lithium dodecyl sulphate (LDS) loading buffer (10 μl; Thermo Fischer) was added to the samples and loaded onto a 1% vertical agarose gel cast in Tris-Glycine-SDS (TGS) buffer (Biorad).Vertical agarose gel electrophoresis (VAGE) was carried out at 100 V for 45 min.The mucins/proteins were transferred onto Immobilon Psq(Merck) in Tris-glycine [5] buffer, using Trans-blot Turbo (25 V, 1 A, 60 min; Biorad).The region of the blot were mucins migrated was cut out and the blot was immersed into 500 μl 0.5M NaBH4 in 0.05 M NaOH.The β-elimination reactions were incubated at 45°C for 16 h and quenched by the stepwise addition of 1ml 5% aqueous acetic acid.The samples were desalted on inhouse prepared cation exchange columns using Amberlite 50Wx8 H+ 200-400mesh.The BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) samples were dried under vacuum and removal of borates was carried out with coevaporation with methanol under nitrogen.
For the base required for permethylation, 400 μl of 50% NaOH were mixed with 800 μl dry MeOH and 4 ml of anhydrous DMSO.The resulting gel was washed 5 times with 4 ml DMSO before resuspended in 4 ml DMSO.The dried samples were dissolved in 100 μl anhydrous DMSO, followed by the addition of 150 μl of the prepared base and 75μl of iodomethane.The samples were vortexed for 2 h at 2000 rpm and the reactions were quenched by the addition of 500 μl H2O.Excess of iodomethane was removed with a flow of nitrogen.
The permethylated glycans were loaded onto a Swift-HLB cartridge (Merck).Salts and other hydrophilic contaminants were removed with 4x1 ml washes with H2O and permethylated glycans were eluted with 4x1 ml of MeOH.The eluted glycans were dried under vacuum and redissolved in 10 μl of 30% acetonitrile in 0.1% aqueous trifluoroacetic acid (TA30).The sample (0.5 μl) was mixed with 0.5 μl of 2,5-dehydroxy-benzoic acid (DHB, 20 mg/ml in TA30) and spotted onto a MTP ground steel MALDI target plate.The samples were analysed by MALDI-ToF MS on a Bruker Autoflex in positive reflectron mode.Peak detection and integration in the mass spectra was done using flexAnalysis (v3.4,Bruker Daltonics) with the following settings: Peak detection algorithm was Snap2, signal to noise threshold = 2, relative intensity threshold = 0, minimum intensity threshold = 2, SNAP2 average composition was set to "sugar", baseline subtraction was set to TopHat.Relative abundance of each peak identified as glycan was calculated as the area of the peak over the sum of all peaks that were identified as glycans.Only glycans present in at least 3 mice and in at least one group were shown.

Endocannabinoid and lipid content
The endocannabinoid and lipid content in the cecal tissue was analyzed by UHPLC-MS.Briefly, lipids were extracted by ultraturax homogeneisation and internal standards (d4-AEA, d4-PEA, d4-OEA, d4-SEA and d5-1-2-AG) were added, followed by protein precipitation (acetone) and recover the supernatant.The samples were analyzed with Xevo-TQS mass spectrometer (from Waters).Absolute quantifications were obtained first by normalizing the area under the curve [6] of the lipid species with the AUC of the respective internal standard and second by extrapolation of the compound's ratio in his own calibration curve.The LC-MS methods was the following: BEH LC-18 column 50*2.1, 1.7µm (Waters) at 40°C.The mobile phase consisted BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) in a gradient between A: H2O 25% -MeOH 75%; B: MeOH 100%, all containing acetic acid (0.1%).ESI probe operated in positive mode was also used for sample ionization.The mass spectrometer parameters were the following: capillary voltage: 2.9kV ; cone voltage : 30V ; desolvation temperature : 550°C ; desolvation gas flow : 1100L/Hr : cone gas flow : 170L/Hr : nebuliser : 6bar.

DNA extraction and 16S rRNA gene amplicon sequencing
Analysis of gut microbiota composition was performed for fecal samples collected at the beginning (day 0) and at the end (day 45) of the study and for the caecal content collected and kept frozen at -80°C until use.Genomic DNA was extracted using a QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions, including a bead-beating step.The V4 region of the bacterial 16S rRNA gene was amplified with the primers 515F(GTGYCAGCMGCCGCGGTAA) and 806R (GGACTACNVGGGTWTCTAAT).Purified amplicons were sequenced using Illumina MiSeq technology following the manufacturer's guidelines.Sequencing was performed at MR DNA (www.mrdnalab.com;Shallowater, TX).Sequences were processed using the QIIME2 pipeline (version 2021.4).[7] Demultiplexed 225bp paired-end sequences were denoised using DADA2 to obtain an amplicon sequence variant (ASV) table.[8] Singletons (ASV present < 2 times) and ASVs present in less than 10% of the samples were discarded.Taxonomic classification was performed using a pre-trained naive Bayes classifier implemented in QIIME2 against the SILVA 132 reference database.[9]Taxa that could not be identified on genus-level are referred to the highest taxonomic rank identified.

Quantitative PCR for total bacteria
Quantification of total bacteria was carried out by qPCR with universal bacterial primers (338F: ACTCCTACGGGAGGCAGCAG, 518R: ATTACCGCGGCTGCTGG), with the StepOnePlus real-time PCR system and software (Applied Biosystems, Den Ijssel, The Netherlands) using GoTaq qPCR sybr green mix (Promega, Madison, Wisconsin, USA), according to the manufacturer's instructions.All samples were run in duplicate in a single 96-well reaction plate.The cycle threshold [1] of each sample was compared with a standard curve made by serially diluting genomic DNA isolated from a pure culture of the type strain of Lactobacillus acidophilus (DSM 20079 01-21) (BCCM/LMG, Ghent, Belgium; DSMZ, Braunshweig, Germany).
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) The absolute abundances of individual bacterial genera were estimated by multiplying their relative abundance by total bacterial density as described previously.[10]

Preparation of mouse fecal extracts
Fecal extracts were processed based on the protocol of Redinbo et al. [11], with modifications.
Briefly, 1-2 fecal pellets collected at the end of the experiment and stores at -80 °C were rehydrated with 350 μl cold extraction buffer (pH 6.5, 25 mM HEPES, 25 mM NaCl with Roche cOmpleteTM protease inhibitor cocktail).The mixture was then transferred in new tubes containing autoclaved 0.7 mm garnet beads and vortexed to break up dense and fibrous material.Bacterial cells were lysed using MP FastPrep-24 TM Classic high-speed benchtop homogenizer (MP Biomedicals, Santa Ana, CA, USA) for 2 minutes at 30 Hertz.The resulting homogenate was sonicated two times for 2 min, with an intermediate step of mixing by inversion.The resulting homogenate was centrifugated at 13,000xg for 10 min at 4 °C and the supernatant was decanted.The total protein concentration was calculated using Pierce TM BCA Protein Assay Kit (#23225, Thermo Fisher Scientific, Waltham, MA, USA).The mouse fecal extract was aliquoted and stored at -80°C until further use.

In-gel activity-based probes (ABP) fluorescent labelling of mouse fecal extracts
Mouse fecal extracts were diluted with buffer (pH 6.5, 125 mM HEPES, 125 mM NaCl, final) to have 1 µg of total protein in 9 µL of lysate working solution.1 µL of Cy5-ABP at a final concentration of 1 µM for alpha-L-fucosidase labeling (JJB381) [12] and 0.5 µM for alpha-Dgalactosidase (TB474) [13] was added to the lysate working solution (9 µL) on ice, and the resulting mixture was incubated at 37 °C for 1 h.The samples were denatured by adding 2.5 μL 5x Laemmli buffer (containing 0.3 M Tris-HCl pH 6.8, 50 % (v/v) 100 % glycerol, 8 % (w/v) dithiothreitol (DTT), 10 % (w/v) sodium dodecyl sulfate (SDS), 0.01 % (w/v) bromophenol blue) and boiled at 98 °C for 5 min.Samples were cooled on ice and run on 1.00 mm 10% polyacrylamide gel at 200 V. Wet-slab gels were scanned for ABP-emitted fluorescence using the Typhoon TM FLA 9500 scanner (Amersham Biosciences, Piscataway, NJ, USA), at 700 PMT and 50 μm resolution.Wet-slab gels were subsequently stained with Coomassie Brilliant Blue (CBB) staining agent to verify accurate protein loading.Full gel images and the relative CBB scanned images can be found in Supplemental Figure 7A,B.
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) normalization (Log transformation (base 10) and auto scaling) using MetaboAnalyst and PCA was computed from scaled data using PCA function in "FactoMineR" package.Differences between clusters were estimated by PERMANOVA test with 999 permutations on Euclidean distance using adonis2 function from "vegan" package.Proteins that were significantly up/down-regulated were used to create a gene list and execute Kyoto Encyclopedia of Genes and Genomics (KEGG) pathway and functional annotation clustering, giving which term/annotation groups were enriched (using DAVID 2021, https://david.ncifcrf.gov).[21,22] Following default settings, only clusters with P-values <0.05 (corresponding to enrichment scores ≥ 1.3) were shown in Supplemental figure 4 and 5.For human fecal proteomes, DAVID tool was also used to investigate the molecular function, biological process, KEGG pathway and diseases, as shown in figure 14.