Figures and data

Chemostat setup and bacterial changes over time with lactose medium (LAC) or lactose and milk oligosaccharide medium (LAC-MO).
A) The experimental setup of chemostat cultivations of donor feces (n=2 for LAC and n=4 for LAC-MO). Donor feces was inoculated into a fermenter followed by batch growth for 24 h. The chemostat culture was then run until 10 volume exchanges. The sampling points for metabolite and microbiome analyses are indicated. B) Processing of chemostat culture to separate virus-like particles from bacteria and the culture residuals. C) Number of observed bacterial species. D) Relative bacterial abundance per replicate summarized at species level. Taxonomic rank specifications: p = phylum, f = family, g = genus. E) Volcano plot visualizing bacterial DESeq2 enrichment analysis with FDR correction based on 16S rRNA gene amplicon sequencing on species level. Horizontal dotted lines show a significance level of P < 0.05 and vertical lines show ± 0.6 log2 fold change (FC) cut-off. NS = non-significant. F) Principal coordinate analysis (PCoA) plot visualizing bacterial beta diversity over time, based on Bray-Curtis dissimilarity metrics.

Metabolite production in cultures of piglet fecal microbiota during batch phase (0-24 h) and chemostat phase (24-75 h) between replicas (R) of cultivations with lactose (LAC) and lactose-milk oligo-saccharide (LAC-MO) medium.
Statistical significance is calculated between LAC-MO and LAC experiments.

Virome changes over time in chemostat cultures with lactose medium (LAC) or lactose-milk oligosaccharide medium (LAC-MO).
A) Number of observed viruses. B) Principal component analysis (PCoA) plot visualizing viral beta diversity over time, based on Bray-Curtis dissimilarity metrics. C) Relative abundance of viral bacterial hosts per replicate summarized at genus level. D) Predicted lifestyle of bacteriophages based on presence (temperate) or absence (virulent) of integrase and/or recombinase genes. E) Vulcano plot visualizing viral OTU differences between substrates at 75 hours. The DESeq2 enrichment analysis with FDR correction was based on viral metagenomics sequencing. Horizontal dotted lines show significance level of P < 0.05 and vertical lines show ± 0.6 log2 fold change (FC) cut-off. NS = non-significant. F) Depletion of eukaryotic viral families presented as heatmaps. The upper panel shows a raw analysis of culture samples. The lower panel shows the analysis, where the baseline viral content in the medium was deducted from culture samples. Color intensity indicates the relative abundance of specific eukaryotic viral contigs from total viral eukaryotic contigs.

Isolation and enumeration of viruses.
A) Inocula were produced by isolating virus-like particles (VLPs) from 1) donor feces to produce fecal virome transfer (FVT) solution, from 2) chemostat culture with lactose (LAC) medium to produce chemostat virome transfer (CVT) solution, and from 3) chemostat culture with lactose-milk oligosaccharide (LAC-MO) medium to produce CVT-MO solution. B) Estimation of average VLP level in donor feces (VLP/g), the chemostat cultures (VLP/mL), and the dose-adjusted inocula (VLP/mL). C) Number of observed viral species as a measure of viral alpha diversity in each inoculum and control (CON) solution.

Study design and clinical data from Piglet Experiment 1.
A) Cesarean section-delivered piglets received either control solution (CON) or one of the VLP-rich inocula (FVT, CVT, CVT-MO) over four oral treatments. The pigs were fed increasing volumes of enteral formula 1 until euthanasia. B) Survival curve. C) Gross pathology score in stomachs. D) Cumulative gross pathology score in small intestines and colons. E) Cumulative microscopic lesion score in small intestines and colons. F) Time to diarrhea onset. G) Body weight development showing percentage change from the median birth weight (±SD). Lines not sharing the same letter are significantly different (P < 0.05). PGroup = P-value for group effect, PInt = P-value for group and time interaction. FVT = fecal virome transfer, CVT = chemostat virome transfer, and CVT-MO = CVT propagated with milk oligosaccharides (n = 14-15/group).

Gut virus engraftment in Piglet Experiment 1.
A-B) Number of observed viral species as a measure of viral alpha diversity between groups (A) and between NEC and healthy appearing colons (B). C) Principal coordinate analysis (PCoA) plot visualizing viral beta diversity based on Bray-Curtis dissimilarity metrics. False discovery rate (FDR) adjusted P-values for pairwise comparisons are reported in the adjacent table. D) Mean relative abundance of viral bacterial hosts summarized at genus level. E) Predicted lifestyle of bacteriophages based on presence (temperate) or absence (virulent) of integrase and/or recombinase genes. F) The percentage of shared viral OTUs between recipients of each group and the inoculum material of the group. CON = control, FVT = fecal virome transfer, CVT = chemostat virome transfer, and CVT-MO = CVT propagated with milk oligosaccharides (n = 10-13/group). *P < 0.05.

Bacterial composition in the colon and transcriptome profile in mesenteric lymph nodes of preterm piglets (Piglet Experiment 1).
A-B) Number of observed bacterial species between NEC and healthy appearing colons (A) and between groups (B). C) Principal coordinate analysis (PCoA) plot visualizing bacterial compositional differences as determined by Bray-Curtis dissimilarity metrics. False discovery rate (FDR) adjusted P-values for pairwise comparisons are reported in the adjacent table. D) Mean relative bacterial abundance summarized at species or genus (g) level. E) Principal component analysis (PCA) score plot illustrating the global gene expression differences in mesenteric lymph nodes. F-G) Gene set enrichment analysis (GSEA) based on gene ontology classification, showing all significantly enriched pathways (false discovery rate adjusted P-value < 0.05) between CVT and FVT (F) and between CVT-MO and FVT (G). The size of the dots indicates the gene ratio, while the yellow color indicates a lower adjusted P-value. NS = non-significant. CON = control, FVT = fecal virome transfer, CVT = chemostat virome transfer, and CVT-MO = CVT propagated with milk oligosaccharides (n = 10-13/group). *P < 0.05.

Study design and clinical data from Piglet Experiment 2.
A) Overview of study design. Cesarean section-delivered piglets were allocated to one control group (CON) receiving SM buffer and one treatment group receiving chemostat virome transfer propagated with milk oligosaccharides (CVT-MO). The pigs were fed increasing volumes of enteral formula 2 until euthanasia after 96 hours. B) Survival curve. C) Gross pathology score in stomachs. D) Cumulative gross pathology score in small intestines and colons. E) Cumulative microscopic lesion score in small intestines and colons. F) Time to diarrhea onset. G) Body weight development showing percentage change from the median birth weight (±SD). PGroup = P-value for group effect, PInt = P-value for group and time interaction.