Specified pathogen free conventional C57BL/6OlaHsd males of 8 weeks old were purchased from a commercial supplier (Envigo, Horst, the Netherlands), and C57BL/6OlaHsd male 8 weeks old germ-free mice were obtained from a breeding colony at the animal facility of the Radboud University (Nijmegen, the Netherlands). Samples derived from germ-free mice were routinely screened once every 2 weeks on sterility. In addition, samples from germ-free mice the day before the end of the experiment were tested on sterility. All samples were screened on sterility by QM diagnostics (Nijmegen, the Netherlands) and confirmed to be sterile. At least 2 weeks before the start of β2→1-fructans treatment all animals were receiving a D12450B diet (10% fat, Research Diet Services, Wijk bij Duurstede, the Netherlands). The mice were kept on this diet throughout the experiment. All experiments were approved by the animal ethics committee of the University of Groningen (Dierenexperimentencommissie Rijksuniversiteit Groningen, DEC-RUG).

Frutalose^®OFP is a highly soluble powdered short-chain β2→1-fructan also called FOS of mainly DP2-10, produced by partial hydrolysis of chicory inulin. Frutafit^®TEX! (both obtained from Sensus, Roosendaal, the Netherlands), is a powdered food ingredient of DP10-60, based on native chicory inulin. DP profiles of supplement A and B are depicted in Figure S1 in Supplementary Material.

Endotoxin levels were tested by Toxikon (Leuven, Belgium) and found to be below 0.3 × 10⁻³ µg⁻¹. The fructans were dissolved in sterile water and given to the conventional mice by oral gavage. In addition, to ensure sterility of these solutions before administration to germ-free mice, short-chain β2→1-fructans were filtered with a 0.2 µm filter and long-chain β2→1-fructans were irradiated with 25 kGy (Synergy Health, Ede, the Netherlands), as it was to viscous for sterilization by filtration. Irridation induced some degradation of higher DP oligomers, but still significant amounts of oligomers up to DP30 were present. Sterility of the fructans after filtration or irradiation was tested by QM diagnostics (Nijmegen, the Netherlands) and confirmed to be sterile. Mice received 10 mg of these fibers or water only as a negative control by oral gavage once a day for 5 days in a row.

Colonic content from each mouse was first dissolved in PBS (100 mg/ml) and the supernatant was stored at −20°C after separation from the debris by centrifugation. These samples together with cecum contents derived from the same mice as prepared in the same way were used to analyze for short- and long-chain β2→1-fructans with HPAEC, which was performed on an ICS5000 system (Thermo Fisher Scientific, Waltham, MA, USA), equipped with a Dionex CarboPac PA-1 column (2 mm × 250 mm) in combination with a Carbopac PA-1 guard column (2 mm × 50 mm). Ten microliters of the samples were injected into the system using a Thermo Fischer Scientific ICS5000 autosampler. The system was equipped with pulsed amperometric detection (ED40 gold electrode). The flow rate was 0.3 ml/min, and the gradient used was 0–5 min 0.1 M NaOH, from 5 min in 40 min to 0.4 M NaOAc in 0.1 M NaOH; 5 min isocratically at 1 M NaOac in 0.1 M NaOH finally an equilibration step for 15 min at starting conditions. The software used was Chromeleon version 7 (Dionex). Monosugars and maltodextrin were used as standards.

For SCFA analysis of cecum content, HPLC was performed to quantify butyric acid, propionic acid, acetic acid, lactic acid, and succinic acid on a Ultimate 3000 HPLC (Thermo Fisher Scientific, Waltham, MA, USA) equipped with an autosampler, a RI-101 refractive index detector (Shodex, Kawasaki, Japan), and an ion-exclusion Aminex HPX 8H column (7.8 mm × 300 mm) with a guard column (Bio-Rad, Hercules, CA, USA). The mobile phase was 5 mM H₂SO₄, and the flow rate was 0.6 ml/min at 65°C. Samples were injected onto the column and SCFA production was quantified.

Spleen, PPs, and MLNs were isolated from each mouse. PPs were removed from the small intestine by tightening the patch with tweezers and cutting it with scissors, avoiding as much as possible inclusion of the surrounding intestinal epithelium. Fat tissue was carefully removed from the MLNs with a razor blade. After isolation, all organs were immediately placed in RPMI, 10% FBS, 1% Pen/Strep on ice. Single cell suspensions were obtained from all organs by crushing the organs with the plunger of a 2.5-ml syringe on 70 µm nylon cell strainers placed over a 50-ml tube. Red blood cells in spleen samples were lysed with a hypotonic lysis buffer. Cells were stained with Fixable Viability Dye eFluor 50 (eBioscience, Vienna, Austria) to exclude dead cells. Aspecific binding to FC receptors was prevented by incubating the cells with anti-CD16/32 (clone 93, Biolegend, Uithoorn, the Netherlands) for 15 min on ice. For extracellular staining, cells were incubated with the desired mixture of antibodies for 30 min on ice. After washing, cells were fixed with FACS lysing solution (BD Biosciences, Breda, the Netherlands). For intracellular staining, fixed cells were permeabilized with PERM (eBioscience, Vienna, Austria) and subsequently stained with the desired antibodies for 30 min on ice. For identification of the different T helper cell subsets, cells were stained with antibodies against: CD3e (clone 17A2), CD4 (clone GK1.5), T-bet (clone 4B10), RORyt (clone B2D), Gata-3 (clone TWAJ), CD25 (clone PC61), and Foxp3 (clone FJK-16S). Appropriate isotype controls were used to determine specificity of the staining. To identify naive or antigen-experienced T cells, the following antibodies were used: CD3e (clone 17A2), CD4 (clone GK1.5), CD8a (clone 53-6.7), CD69 (clone H1.2F3), CD44 (clone IM7), and CD62L (clone MEL-14). B-cells were stained with: CD19 (clone 6D5), B220 (clone RA-6B2), IgD (clone RTK2758), IgM (clone II/41), IgA (clone mA-6E1). For analysis of dendritic cells (DCs), other lineages were first excluded with CD3e (clone 17A2), CD19 (clone 6D5), B220 (clone RA-6B2), and NK1.1 (clone PK136). Next, DC subsets were identified with CD11c (clone HL3), MHC-II (clone M5/114.15.2), CD11b (clone M1/70), CD103 (clone 2E7), CD8a (clone 53-6.7), and CD80 (clone 16-10 A1). Samples were acquired with the FACSVerse (BD Biosciences, Breda, the Netherlands) and analyzed with Flow Jo software (Flow Jo LLC, OR, USA).

A part of the terminal ileum from each mouse was snap frozen in liquid nitrogen and stored at −80°C. RNA was isolated with the RNeasy kit (Qiagen, Valencia, CA, USA). Quantity of RNA was measured with the ND-1000 (NanoDrop Technologies, Thermo Fisher Scientific, Breda, the Netherlands) and quality of RNA was assessed with the Bioanalyzer 2100 (Agilent, Santa Clara, CA, USA). Total RNA (100 ng) was labeled utilizing the Ambion WT Expression kit (Life Technologies Ltd., Bleiswijk, the Netherlands) and the Affymetrix GeneChip WT Terminal Labeling kit (Affymetrix, Santa Clara, CA, USA). After labeling, samples were hybridized to Affymetrix GeneChip Mouse Gene 1.1 ST arrays. An Affymetrix GeneTitan Instrument was used for hybridization, washing, and scanning of the array plates. Bioconductor packages integrated in an online pipeline were used for quality control of the data (17, 18). Probe sets were redefined using current genome information (19). Probes were reorganized based on the Entrez Gene database (remapped CDF v14.1.1). Robust Multiarray Analysis preprocessing algorithm available in the Bioconductor library affyPLM (20) was used to obtain normalized expression estimates from the raw intensity values.

Another part of ileum and colon was fixed in Carnoy’s fixative to preserve the mucus layer. Samples were embedded in paraffin and afterward 4 µm sections of tissue were made with a microtome (Leica Biosystems, Nussloch, Germany). Paraffin was removed from the tissue slides by washing 2× with xylene for 10 min. Next, samples were washed in decreasing amounts of ethanol (99, 70, 60%) and finally in dH₂O. After washing, samples were incubated with 0.5 µg universal bacterial probe EUB338 (5′-GCTGCCTCCCGTAGGAGT-3′) or segmented filamentous bacteria (SFB)-specifc probe SFB1008 (5′-GCGAGCTTCCCTCATTACAAGG-3′) conjugated on the 5′ end with Alexa 488 (Eurogentec, Maastricht, the Netherlands) in hybridization solution (20 mM TRIS-HCl, pH 7.4, 0.9 M NaCl, 0.1% weight/volume SDS) at 50°C overnight in a humid environment. The next day, the samples were washed in FISH washing buffer (20 mM TRIS-HCl, pH 7.4, 0.9 M NaCl) for 20 min. Tissue slides were then washed 2× for 10 min with PBS and incubated with DAPI solution (30 nM) for 5 min to visualize nuclei of the epithelial cells. After washing 2× with PBS, samples were mounted with mounting medium. Data were acquired with a SP8 Leica confocal microscope (Leica Microsystems, Son, the Netherlands), and images were processed afterward with Imaris software (Bitplane, Zurich, Switzerland).

Fresh fecal samples obtained just after defecation were collected from all conventional mice before the start of fiber treatment. In addition, colonic content samples from these mice were collected at the end of the experiment. All samples were snap frozen in liquid nitrogen and stored at −80°C. These samples were used for 16S rRNA gene analysis for microbiota profiling with barcoded amplicons from the V1–V2 region of 16S rRNA genes generated using a 2-step PCR strategy that reduces the impact of barcoded primers on the outcome of microbial profiling (21). DNA extraction was performed using a combination of the bead-beating-plus column method and the Maxwell 16 Tissue LEV Total RNA purification kit (Promega, Leiden, the Netherlands). Beating of the fecal pellets took place as described before (22) but with STAR (Stool transport and recovery) buffer (Roche, Basel Switzerland). After centrifugation, 250 µl supernatant was taken for the Maxwell 16 Tissue LEV Total RNA Purification Kit, and the DNA was eluted in 50 µl DNAse free water. Twenty nanograms of DNA were used for the amplification of the 16S rRNA gene with primers 27F-DegS and 338R I + 338R II for 25 cycles as described before (23), only primers had a Universal Tag (UniTag) linkers attached; UniTag I (forward) and II (reverses) (I—GAGCCGTAGCCAGTCTGC; II—GCCGTGACCGTGACATCG). The first PCR was performed in a total volume of 50 µl containing 1× HF buffer (Finnzymes, Vantaa, Finland), 1 µl dNTP Mix (10 mM; Promega, Leiden, the Netherlands), 1 U of Phusion^® Hot Start II High-Fidelity DNA polymerase (Finnzymes Vantaa, Finland), 500 nM of the 27F-DegS primer (23, 24) that was appended with UniTag 1 at the 5′ end, 500 nM of an equimolar mix of two reverse primers, 338R I and II (24) based on three previously published probes EUB 338 I, II, and III (23), that were 5′-extended with UniTag 2, and 0.2–0.4 ng/µl of template DNA. The sequence of the UniTags were selected to have a GC content of ~66% and a minimal tendency to form secondary structures, including hairpin loops, heterodimers, and homodimers as assessed by the IDTDNA Oligoanalyzer 3.1 (Integrated DNA Technologies). Moreover, sequences were selected that had no matches in 16S rRNA gene databases [based on results of the “TestProbe” tool offered by the SILVA rRNA database project (25) using the SSU r117 database], and no prefect matches in genome databases with the Primer-BLAST tool (http://www.ncbi.nlm.nih.gov/tools/primer-blast/). The size of the PCR products (~375 bp) was confirmed by gel electrophoresis using 5 µl of the amplification reaction mixture on a 1% (w/v) agarose gel containing 1× SYBR^® Safe (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA). Five microliters of these PCR products were taken to add adaptors and a 8-nt sample-specific barcode in an additional five cycle PCR amplification. This second PCR was performed in a total volume of 100 µl containing 1× HF buffer, dNTP Mix 2 U of Phusion^® Hot Start II High-Fidelity DNA polymerase, 500 nM of a forward and reverse primer equivalent to the Unitag 1 and UniTag 2 sequences, respectively, that were each appended with an 8 nt sample-specific barcode (Hermes et al., in preparation) at the 5′ end. PCR products were purified with the magnetic beads (MagBio, London, UK) according to the HighPrepTM protocol of the manufactures instructions using 20 µl Nuclease Free Water (Promega Leiden, the Netherlands) and quantified using the Qubit (Life Technologies, Bleiswijk, the Netherlands). Purified PCR products were mixed in approximately equimolar amounts and concentrated by the magnetic beads as the purification before. Purified amplicon pools were 250 bp paired-end sequenced using Illumina Miseq (GATC-Biotech, Konstanz, Germany).

The Illumina Miseq data analysis was carried out with a workflow employing the Quantitative Insights Into Microbial Ecology (QIIME) pipeline (26) and a set of in-house scripts as described before for Illumina Hiseq 16S rRNA gene sequences (Hermes et al., in preparation). The set of in-house scripts processed the reads as follows: reads were filtered for not matching barcodes; otu picking and chimera removal was done via matching the sequences to the Silva 111 database, with only one mismatch allowed, and a biom and with clustalw a multiple alignment and phylogenetic tree file was generated. Further outputs were generated via QIIME, such as filtered reads per sample, PD whole tree diversity measurements and the level 1–6 taxonomic distributions with relative abundances.

Flow cytometry data are expressed as means, error bars represent SEM. To verify whether there were significant differences between the groups, one way ANOVA was performed followed by the Dunnett’s multiple comparisons test. All tests were performed with Graphpad software (Prism, La Jolla, CA, USA).

Differentially expressed probe sets were identified using linear models, applying moderated T-statistics that implemented empirical Bayes regularization of SEs (27). A Bayesian hierarchical model was used to define an intensity-based moderated T-statistic, which takes into account the degree of independence of variances relative to the degree of identity and the relationship between variance and signal intensity (28).

Statistical tests for gut microbiota composition were performed using R and Calypso (29), where the count data were not normally distributed and variances between groups were not equal, the Mann–Whitney U test was used.

We demonstrated previously that inulin β2→1-fructans stimulate immune cells directly and in a DP-dependent manner in vitro (15). In this in vivo study in mice, the immune effects of chemically characterized short (DP2-10) and long-chain chain β2→1-fructans (DP10-60) were compared. To this end, conventional mice received these fibers by oral gavage once a day for 5 days. Effects were studied in the PPs, MLNs, and spleen.

Th1, Th2, Th17, and Treg cells in the different organs were compared between mice treated with the two β2→1-fructans and controls (Figure 1A). No differences were found in the spleen (data not shown). However, in the PPs, both short and long-chain β2→1-fructan induced a higher number of Th1 cells (p < 0.05) compared to controls. In addition, in MLNs, short-chain β2→1-fructan but not long-chain β2→1-fructan treatment resulted in higher percentages of Treg cells (p < 0.05) and CD4⁺ T cells expressing the activation marker CD69 (p < 0.05). Next, we assessed the influence of β2→1-fructan treatment on B-cells in these tissues. No differences were observed in B cells in any organ tested (data not shown). Finally, considering the important role of DCs in initiating and modulating adaptive immune responses, we also evaluated the effect of β2→1-fructan treatment on DCs (Figure 1B). Strikingly, mice treated with short-chain but not the mice treated with long-chain β2→1-fructan had an increase in CD11b⁻CD103⁻ DCs in the MLNs (p < 0.01) and a decrease in CD11b⁺CD103⁺ DCs (p < 0.05). No differences were found in spleen or PPs (data not shown).

Thus, β2→1-fructans can modulate the immune system locally in the PP and MLN in vivo in a DP-dependent fashion. Only short-chain β2→1-fructan modulated DC and Treg numbers in MLNs.

To investigate the different impact in an unbiased manner, effects of the two β2→1-fructans on the host whole-genome expression of ileal tissue were studied with microarray. The ileum is assumed to be the location where β2→1-fructans interact with the mucosa (6). Treatment with both short and long-chain β2→1-fructan induced significant differential gene expression of approximately 100 genes in the ileum (p-value < 0.01, fold-change > 1.3) compared to controls (Figure 2A). Remarkably, the majority of affected genes were uniquely and differently regulated by either short- or long-chain β2→1-fructans (Figures 2B,C). There was only a minimal overlap in the induced responses by these β2→1-fructans. One of these genes was galactoside 2-alpha-l-fucosyltransferase 2 (Fut2), an enzyme involved in glycosylation. It was among the top-10 most upregulated genes by both short- and long-chain β2→1-fructans (Figure 2D). Recently, it has been demonstrated that Fut2 is upregulated by microbiota-dependent stimulation of IL-22 production by type 3 innate lymphoid cells, which in turn stimulates colonization resistance against pathogens (30). That IL-22 is regulated by both short- and long-chain β2→1-fructans was further confirmed by the observation that (i) genes known to be upregulated in epithelial cells by IL-22 (31) were strongly enriched in the top-10 most upregulated genes (Figure 2D) and (ii) that several of these genes clustered together after hierarchical cluster analysis (Figure 2C). However, short-chain β2→1-fructan seemed to induce a stronger response, reflected by higher expression of microbiota-dependent genes such as Fut2, SAA1, Retnlb, and Duox2 (Figures 2C,D). As upregulation of the IL-22-Fut2 axis is assumed to be gut microbiota dependent (30), these results suggest that the effects of β2→1-fructans on gene expression in the ileum are caused by differences in effects on gut microbiota. Therefore, we next studied and compared the microbiota composition in the mice treated with short- and long-chain β2→1-fructans.

As SFB promote production of IL-22 and upregulation of Fut2 (30), we first studied the ileum of short- and long-chain β2→1-fructan treated animals for possible differences in colonization by SFB (Figure 3A). Although we observed an overall higher density of microbiota after FISH staining (Figure 3B), especially in animals receiving short-chain β2→1-fructans, we never found SFB in the ileum in any sample. Also by applying 16S rDNA sequencing in fecal samples, we excluded promotion of SFB in the β2→1-fructans-treated mice (data not shown), which suggest that other mechanisms must be responsible for FUT2 upregulation.

Since β2→1-fructans have been shown to modulate gut microbiota composition (10), we analyzed gut microbiota composition with 16S rDNA sequencing in fecal samples before and after treatment with the dietary fibers (Figure 4A). No significant differences were found in any specific bacterial group after treatment (data not shown). Moreover, in agreement with the SFB-specific FISH staining, SFB was not detected in any of the samples (data not shown). Redundancy analysis suggested that gut microbiota composition was similar between the groups before treatment, as expected (Figure 4B). The samples from the three groups posttreatment separated into three different clusters, suggesting that β2→1-fructans had some effect on the composition of the microbial community, but this separation was not significant (Figure 4B).

Next, despite the absence of differences in microbiota composition between short- and long-chain β2→1-fructan-treated mice, we studied possible differences in immune active SCFA production as explanation for the observed immune differences. However, concentrations of butyrate, propionate, acetate, lactic acid, and succinic acid in cecal samples were not different, nor enhanced by the two β2→1-fructans (Figure 5B). As it might be suggested that this might be caused by lack of fermentation of the β2→1-fructans, we quantified the fructans in the colon and cecum of β2→1-fructan-treated mice by HPAEC. Both β2→1-fructans were undetectable, indicating completely utilization by the gut microbiota (Figure 5A, only plot of short-chain fructan is shown).

As we could not find pronounced difference in short and long-chain β2→1-fructans induced effects on microbiota or its degradation products, we determined and compared the impact of the two β2→1-fructans on the mucosal immunity in germ-free mice.

As expected, β2→1-fructans were not degraded in the intestine of germ-free mice (Figure 6A), and no SCFAs could be detected (data not shown). Analysis of immune cells population in the PPs, MLNs, and spleens was performed in the same fashion as in conventional mice. Th1, Th2, Th17, and Treg cells were not influenced by β2→1-fructans in the absence of microbiota (data not shown). In contrast, we did observe effects on B-cells and DCs in β2→1-fructan-treated germ-free mice, and the effects were DP dependent. Long-chain β2→1-fructans induced enhanced numbers of IgD^loIgM^hi B-cells in the MLNs (p < 0.05, Figure 6B) but not in the PPs or spleen (data not shown), while short-chains did not have such an effect. DCs in PPs were only modulated by short-chain β2→1-fructans as expression of the activation marker CD80 was lower in PP DCs that did not express CD11b or CD103 (p < 0.05) (Figure 6C). Only graphs of regulated cells are shown.

Also, the ileum of β2→1-fructan treated germ-free mice was studied by whole-genome expression with microarray. We found that both short- and long-chain β2→1-fructans modulate the gene expression in the ileum in the absence of the gut microbiota (Figure 7A). First, the gene profiles of short-chain β2→1-fructan-treated germ-free animals were compared to that of treated mice with a conventional microbiome to determine unique microbiota-dependent and microbiota-independent effects. The same was done for long-chain β2→1-fructan. After that, we compared the DP-dependent effects in the germ-free animals.

As shown in Figure 7B, there was only minor overlap between the genes affected in conventional and germ-free mice for both short and long-chain fructans, suggesting that β2→1-fructans modulate different pathways by direct interaction and by microbiota-dependent effects.

Strikingly, when studying DP-dependent effects in the mice there was also minimal overlap between differential gene expression induced by short and the long-chain treated germ-free mice (Figure 7C). This illustrates that similarly to what we observed in conventional mice, chain-length of β2→1-fructans is key in determining the type of induced response.

A heatmap of the 40 most differentially expressed genes in the ileum of germ-free mice after treatment with short or irradiated long-chain inulin further confirmed the presence of gene clusters specifically modulated by either short- and long-chain β2→1-fructan (Figure 8A). In particular, long-chain β2→1-fructan seemed to modulate B-cell responses, since several genes involved in antibody production were differentially expressed, including a strong upregulation of IgD and CD20 (Figure 8B). These data correspond to the flow cytometry data, where we observed an increase in IgD^loIgM^hi B-cells in the MLNs of germ-free mice after inulin treatment.