Fecal virus-like particles are sufficient to reduce necrotizing enterocolitis

Abstract

Fecal filtrate transfer (FFT) is emerging as a safer alternative to traditional fecal microbiota transplantation (FMT) - particularly in the context of necrotizing enterocolitis (NEC), a severe gastrointestinal condition affecting preterm infants. Using a preterm piglet model, FFT has demonstrated superiority over FMT in safety and NEC prevention. Since FFT is virtually devoid of bacteria, prokaryotic viruses (bacteriophages) are assumed to mediate the beneficial effects. However, this assumption remains unproven. To address this gap, we separated virus-like particles (30 kDa to 0.45 µm) of donor feces from the residual postbiotic fluid. We then compared clinical and gut microbiota responses to these fractions with the parent FFT solution after transferring them to NEC-susceptible preterm piglets. Virome transfer was equally effective as FFT in reducing the severity of NEC-like pathology. The bacterial compositional data corroborated clinical findings as virome transfer reduced the relative abundance of several NEC-associated pathogens e.g. Klebsiella pneumoniae and Clostridium perfringens. Virome transfer diversified gut viral communities with concomitant constraining effects on the bacterial composition. Unexpectedly, virome transfer, but not residual postbiotic fluid, led to earlier diarrhea. While diarrhea may be a minor concern in human infants, future work should identify ways of eliminating this side effect without losing treatment efficacy.

Keywords: Fecal virome transplantation; bacteriophages; diarrhea; fecal microbiota transplantation; necrotizing enterocolitis; preterm.

Figure 1.

Overview of study design. (a) fecal fractionating process. (b) preterm piglet experimental outline. Created with Biorender.com.

Figure 2.

Characteristics of fecal fractions. (a) fluorescent microscopy at 1000 × magnification showing SYBR™ gold-stained virus-like particles (VLPs) in each fecal solution; FFT = fecal filtrate transfer, VIR = isolated virus-like particles, RES = virus-depleted residual fluid, CON = autoclaved SM buffer. (b) estimation of average VLP concentration per mL in each fecal fraction from fluorescent microscopy of 10 representative pictures. Dotted line marks detection limit. (c) number of observed bacterial species as a measure of bacterial alpha diversity. Numbers above each bar represent the bacterial DNA concentration before amplification. (d) relative bacterial abundance summarized at species level. (e) number of observed viral species as a measure of viral alpha diversity. (f) relative viral abundance summarized at family level. (g) relative abundance of viral bacterial hosts summarized at family level. Taxonomic rank specifications: r = realm, c = class, o = order, f = family, g = genus, s = species.

Figure 3.

Efficacy parameters. (a) cumulative gross pathology score in small intestine and colon. (b) examples of gross pathology scores in colon. Left: NEC case from CON animal with arrow heads pointing at lesions with hemorrhage, pneumatosis intestinalis, and mucosal necrosis. Right: normal colon from VIR recipient. (c) microscopic lesion score in small intestine and colon. (d) examples of microscopic lesion scores in colons. Left: CON animal with transmural necrosis and pneumatosis intestinalis. Right: VIR recipient with normal histology. Scale bar = 0.3 mm. (e) measurement of interleukin (IL) 8 and IL-1β in lesion biopsies from small intestine and colon. FFT = fecal filtrate transfer, VIR = isolated virus-like particles, RES = virus-depleted residual fluid, CON = autoclaved SM buffer.

Figure 4.

Safety parameters. (a) survival. (b) time to first diarrhea episode. (c) body weight development showing percentage change from the median birth weight. (d) weights of intestinal segments relative to the body weight at euthanasia. Lines and boxplots not sharing the same letter are significantly different (p < 0.05). FFT = fecal filtrate transfer, VIR = isolated virus-like particles, RES = virus-depleted residual fluid, CON = autoclaved SM buffer.

Figure 5.

Colonic microbiota composition from 16S rRNA gene amplicon and viral metagenomics sequencing. (a) shannon diversity index as a measure of bacterial alpha diversity. (b) Principal coordinate analysis (PCoA) plot visualizing bacterial beta diversity based on bray-curtis dissimilarity metrics. FDR-adjusted P-values of pairwise comparisons are reported in adjacent table. (c) bacterial bray-curtis dissimilarity within each group as a measure of bacterial community heterogeneity. (d) shannon diversity index as a measure of viral alpha diversity. (e) PCoA analysis plot visualizing viral beta diversity based on bray-curtis metrics. Adjacent table show FDR-adjusted p-values of pairwise comparisons between the four groups of recipients. Filtrate = fecal fractions from donor pigs. (f) viral bray-curtis dissimilarity within each group as a measure of viral community heterogeneity. (g) mean relative bacterial abundance summarized at species level. (h) mean relative viral abundance summarized at family level. g) mean relative abundance of viral bacterial hosts summarized at family level. Taxonomic rank specifications: r = realm, c = class, g = genus. FFT = fecal filtrate transfer, VIR = isolated virus-like particles, RES = virus-depleted residual fluid, CON = autoclaved SM buffer. *p < 0.05, **p < 0.01, ***p < 0.001, ***p < 0.0001.

Figure 6.

Pairwise DESeq2 enrichment analyses with FDR correction visualized in volcano plots to detect changes in low-abundance microbes. (a) bacterial enrichment analyses based on 16S rRNA gene amplicon sequencing on species level. (b) viral enrichment analyses based on viral metagenomic sequencing on viral OTU level. 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. FFT = fecal filtrate transfer, VIR = isolated virus-like particles, RES = virus-depleted residual fluid, CON = autoclaved SM buffer.

Figure 7.

Spearman correlation analyses. (a) correlation between viral OTUs (blue dots) and key clinical parameters (green dots). Clinical parameters included; time of diarrhea onset, relative body weight loss, small intestinal weight, and gross pathology score. After FDR correction, no correlations appeared between viruses and pathology score or small intestinal weight, respectively. (b) correlation between viruses (blue dots, binned at genus level), bacteria (pink dots, binned at species level), and key clinical parameters (green dots). After FDR correction, no correlations appeared between viruses/bacteria and gross pathology score. Thicker lines indicate stronger correlations and color indicate direction; purple for positive and yellow for negative Spearman ρ correlation coefficient.