Lacticaseibacillus rhamnosus HN001 alters the microbiota composition in the cecum but not the feces in a piglet model

Abstract

The probiotic Lacticaseibacillus rhamnosus strain HN001 has been shown to have several beneficial health effects for both pediatric and maternal groups, including reduced risk of eczema in infants and gestational diabetes and postnatal depression in mothers. While L. rhamnosus HN001 appears to modify immune and gut barrier biomarkers, its mode of action remains to be fully elucidated. To gain insights into the role of HN001 on the infant microbiome, the impacts of L. rhamnosus HN001 supplementation was studied in 10-day old male piglets that were fed either infant formula, or infant formula with L. rhamnosus HN001 at a low (1.3 × 10⁵ CFU/ml) or high dose (7.9 × 10⁶ CFU/ml) daily for 24 days. The cecal and fecal microbial communities were assessed by shotgun metagenome sequencing and host gene expression in the cecum and colon tissue was assessed by RNA-seq. Piglet fecal samples showed only modest differences between controls and those receiving dietary L. rhamnosus HN001. However, striking differences between the three groups were observed for cecal samples. While total lactobacilli were significantly increased only in the high dose L. rhamnosus HN001 group, both high and low dose groups showed an up to twofold reduction across the Firmicutes phylum and up to fourfold increase in Prevotella compared to controls. Methanobrevibacter was also decreased in HN001 fed piglets. Microbial genes involved in carbohydrate and vitamin metabolism were among those that differed in relative abundance between those with and without L. rhamnosus HN001. Changes in the cecal microbiome were accompanied by increased expression of tight junction pathway genes and decreased autophagy pathway genes in the cecal tissue of piglets fed the higher dose of L. rhamnosus HN001. Our findings showed supplementation with L. rhamnosus HN001 caused substantial changes in the cecal microbiome with likely consequences for key microbial metabolic pathways. Host gene expression changes in the cecum support previous research showing L. rhamnosus HN001 beneficially impacts intestinal barrier function. We show that fecal samples may not adequately reflect microbiome composition higher in the gastrointestinal tract, with the implication that effects of probiotic consumption may be missed by examining only the fecal microbiome.

Introduction

Probiotics are defined as living microorganisms that confer health benefits on the host when consumed in adequate amounts (1). They are typically lactic acid bacteria belonging to the Lactobacillus or Bifidobacterium genera. Due to their demonstrated health benefits, they are commonly added to infant formula (2).

Several beneficial health effects have been reported for the probiotic Lacticaseibacillus rhamnosus (formerly Lactobacillus rhamnosus) strain HN001 for both pediatric and maternal groups, including reduced risk of eczema in infants (3) and reduced risk of gestational diabetes and postnatal depression in mothers (4). Effects may also be long lasting, with evidence showing early childhood supplementation with L. rhamnosus HN001 can lead to reduced incidence of eczema at 11 years of age (5). While L. rhamnosus HN001 appears to modify immune and gut barrier biomarkers (6), its mode of action remains to be fully elucidated. Consumption of L. rhamnosus HN001 does not tend to result in colonization (7) and a recent study of infant microbiomes from fecal samples collected as part of an anti-eczema study showed few significant changes in the microbiome, both in terms of taxa and metabolic pathways (8).

The effects of probiotics are likely to occur through several mechanisms, individually or in combination, depending on the specific strain of bacterium. One possible mechanism is through the improvement of intestinal barrier integrity (9) to reduce the translocation of antigens across the gut epithelium. Probiotics may also modulate the host immune system via specific receptors that sense microbial metabolites such as short chain fatty acids (SCFAs) and polysaccharides (10–12). A further putative mechanism is the influence of the probiotic on the microbiota present in the gut. However, the extent to which probiotics are able to modulate the resident microbiota is unclear, with some studies showing little effect (13) and others showing that effects are dependent on the starting microbiome (14). A contributing factor to the variable demonstrated impacts of probiotics on the gut microbiome may be that most studies measure changes in the fecal microbiome, which may not fully reflect the situation in the gastrointestinal tract itself (15).

During early life, a probiotic added to infant formula may shape the large intestinal environment into one that promotes a more beneficial microbiota, and thus enhance its development and maturation. In this study, we aimed to better understand the impact of L. rhamnosus HN001 in early life on the gastrointestinal microbiome using piglets fed an infant formula supplemented with two different doses of L. rhamnosus HN001. Compared with rodents, the neonatal pig shares greater similarities with neonatal humans in terms of anatomical and physiological features (16, 17). The cecal and fecal microbial communities were compared using shotgun metagenome sequencing. Associated changes in the cecal and colonic tissue gene expression profiles were compared using RNA-seq.

Materials and methods

Animals

The study was carried out in strict accordance with the NZ Animal Welfare Act 1999, and was approved by the AgResearch Limited (Grasslands) Animal Ethics Committee (Ethics Approval No.: 13982).

Twenty-four male large white cross 10-day old piglets were obtained from a commercial farm in the Manawatu-Wanganui region of New Zealand. All piglets were housed in custom cages constructed to allow animals to see, hear, and smell adjacent piglets while minimizing physical contact (18). On arrival at the animal facility (day 1), the piglets were pair-housed for two nights. The piglets were exclusively fed reconstituted infant formula [control formula; Fonterra Nutritional Base (protein 13.63%, fat 26.2%, ash 2.8%, carbohydrate 55.07%) + DCL100 at 5 g/L of formula] 2 h post-arrival and every 4 h thereafter. From day 3 to day 24 the piglets were individually housed. During this period, the piglets were let out into a shared pen and allowed to physically interact for an hour of social time each day.

Lactocaseibacillus rhamnosus strain HN001 (formerly known as Lactobacillus rhamnosus HN001, also known as LactoB HN001™ or DR20™) was supplied by Fonterra (Palmerston North, New Zealand).

From day 3 to day 24, the piglets were assigned to one of three treatment groups; 8 receiving control formula; 8 receiving control formula supplemented with 1.3 × 10⁵ CFU/ml of L. rhamnosus HN001 (HN001 Low); and 8 receiving control formula supplemented with 7.9 × 10⁶ CFU/ml of L. rhamnosus HN001 (HN001 High). The doses used were within the range of commonly used amounts for infant formulae (19). All formulas were dispensed using an automated system programmed to offer the formula every 2 h (starting at 2 L/day on day 3, increasing to 5.8 L/day by day 24) with automatic measurement of refusals. The piglet caging and automated feeding systems were designed and manufactured by ShapeMaster (Ogden, IL, USA).

Fecal samples were collected each day and stored at −80°C for later analyses. At the end of the study, the piglets were euthanized by sedation using a cocktail of tiletamine hydrochloride, zolazepam hydrochloride, xylazine, and ketamine hydrochloride (final solution 50 mg/ml of each drug, administered at a dose rate of 0.3 mL of the mixed solution/10 kg body weight), followed by intracardiac puncture with sodium pentobarbitone (300 mg/ml; 0.27 ml/kg body weight). Cecal contents were collected and stored at −80°C for microbiome analyses. Tissue samples from the cecum and ascending colon were collected into RNAlater (Qiagen, Hilden, Germany) and stored at −80°C for gene expression analysis.

Microbiome

Metagenomic DNA was extracted from cecal contents and fecal samples from the final day using Macherey Nagel NucleoSpin Soil kits (Düren, Germany) following the manufacturer’s instructions, with the addition of bead beating on a BioSpec Mini-Beadbeater 96 (Bartlesville, OK, USA) set to 4 min at 40 oscillations/s.

A total amount of 1 μg of metagenomic DNA per sample was used as input material for the DNA sample preparations. Sequencing libraries were generated using NEBNext^® Ultra™ DNA Library Prep Kit for Illumina (New England Biolab, Ipswich, MA, USA) following the manufacturer’s recommendations and index codes were added to attribute sequences to each sample. Briefly, the DNA sample was fragmented by sonication to a size of 300 bp, then DNA fragments were end-polished, A-tailed, and ligated with the full-length adaptor for Illumina sequencing with further PCR amplification. The PCR products were purified (AMPure XP system) and libraries were analyzed for size distribution by the Agilent 2100 Bioanalyzer and quantified using real-time PCR. Clustering of index-coded samples was performed on a cBot Cluster Generation System according to the manufacturer’s instructions. After cluster generation, the library preparations were sequenced on an Illumina HiSeq X platform and 150-bp paired-end reads were generated.

The software Trimmomatic v.0.36 (20) was used for removal of adapters, low quality (Phred scores < 30), and short (<50 bp) sequencing reads. Read pairs were joined using PEAR version 0.9.6 (21) with default settings. Read pairs that did not join were concatenated with a spacer consisting of a string of N’s using the “fuse” function from the BBMAP package version 38.22-0 (22). Evaluation and detection of host reads were done using the bbduk.sh function from the BBMAP package version 38.22-0 (22), a k-mer based filter, with the pig genome (Sscrofa 11.1 release 96) as reference. The “blastx” function of DIAMOND version 0.9.22 (23) was used to map the reads against the “nr” NCBI database. MEGAN6 Ultimate Edition (24) was used to assign putative functions to the DIAMOND alignment files against the SEED Subsystems database (25).

Comparisons of overall community compositions were performed using permutational multivariate analysis of variance (PERMANOVA) of distance matrices, implemented through the adonis function from the vegan package (26) for R. Differences in the relative abundances of individual taxa and gene functions were analyzed using ANCOM-BC (27) package in R, with Q-values < 0.05 considered significant. ANCOM-BC is a method for detecting differences in abundance while accounting for the compositional nature of microbiome datasets.

Piglet gut tissue transcriptome

Gene expression profiles from cecal and colon tissue samples were analyzed by RNA-seq. Total RNA was extracted from samples using RNeasy Mini Kits (Qiagen, Germantown, MD, USA). Total RNA quality and quantity were determined using an Agilent 2100 Bioanalyzer Instrument (Agilent, Santa Clara, CA, USA) and Nanodrop (Thermo Fisher Scientific, Waltham, MA, USA), and sample quality was also assessed using agarose gel electrophoresis. Samples that passed the RNA integrity number (RIN) threshold of 6.5 were submitted for sequencing. Strand-specific cDNA libraries were prepared using NEBNext^® Ultra Directional RNA Library Prep Kit for Illumina^® (New England Biolab, Ipswich, MA, USA) according to the manufacturer’s guidelines. Libraries were size selected for 250–300 bp fragments and sequenced using the Novaseq 6000 platform (Illumina) to produce 150 bp paired-end sequences. The reads were quality trimmed using Trimmomatic 0.36 (20) in paired-end mode using the following parameters; LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36. Read pairs that passed quality trimming were mapped against the genome (Sscrofa 11.1 release 96) using STAR (28). Uniquely mapped read pairs were summed for each gene and analyzed using the EdgeR package (29) in R. The resulting counts were analyzed using a likelihood ratio generalized linear model, with genes that had > 1.5-fold difference (i.e., log fold change > | 0.58|) and FDR < 0.05 considered differentially expressed. Gene expression profiles were also analyzed by gene set enrichment analysis (GSEA) using the mroast function from limma (30) and KEGG pathways (31, 32) as gene sets. GSEA involves the analysis of the collective expression of groups of genes treated as a unit rather than as individual genes.

Results

There were no significant differences in body weight between the groups at the start of the study (P = 0.96) or at the conclusion of the study (P = 0.94). Similarly, no differences in the percentage weight gain were observed (P = 0.95). The amount of infant formula consumed over the course of the study did not differ between groups (P = 0.31) and no differences in animal health or adverse effects were apparent.

Sequence reads are available from the NCBI Sequence Read Archive (SRA), accession PRJNA823879.

Microbiome

Following quality trimming the median number of paired-end reads per sample was 10.7 M, with a standard deviation, minimum, and maximum number of reads of 1.8, 6.8, and 15.6 M, respectively.

Lacticaseibacillus rhamnosus relative abundances

The taxonomic data summarized at the species level from the MEGAN analysis of NCBI nr hits, showed supplementation with L. rhamnosus HN001 resulted in a higher relative abundance of L. rhamnosus in the cecal and fecal community (P < 0.001 and P = 0.007, respectively) at the higher dose, compared to the Control and HN001 Low groups (Figures 1A,B). Because of the uncertainty with classification to the species level (33, 34), we also considered the relative abundance of taxa classified as uncultured and unclassified Lactobacillus (which in this instance includes Lacticaseibacillus and other former members of the recently reclassified Lactobacillus genus). When these were combined with sequences identified as L. rhamnosus, we saw a similar pattern in the cecum where the relative abundance of these combined sequences was higher in the HN001 High group (P = 0.018; Figure 1C). However, no differences were observed in the fecal community (P = 0.232; Figure 1D).

FIGURE 1

Cecal microbiome

The supplementation of infant formula with L. rhamnosus HN001 also led to dramatic changes in the overall cecal microbiome composition (Figure 2, PERMANOVA P = 0.001, Table 1). This included a large increase in Gram-negative phyla. Piglets in both HN001 High and HN001 Low groups had increased relative abundance of Bacteroidetes compared to the Controls (global Q < 0.001). Similarly, both doses of HN001 increased relative abundances of Proteobacteria (global Q = 0.03). At the genus level Prevotella were relatively more abundant in piglets fed either dose of L. rhamnosus HN001 compared to Control (Q < 0.05). The change in Prevotella was particularly notable as they underwent a major fold change while also comprising a substantial proportion of the community (Control 8.9% ± 1.7; HN001 Low 23.9% ± 3.1; HN001 High 17.6% ± 3.3; mean% ± SEM).

FIGURE 2

The increase in Gram-negative bacteria was accompanied by a concomitant decrease in Gram-positive bacteria including the Firmicutes (global Q = 0.05) and Actinobacteria, although the latter was not considered a significant difference (Q = 0.11). Methanobrevibacter was also lower in piglets receiving HN001 (Q = 0.004; Control 0.29 ± 0.06; HN001 Low 0.11 ± 0.04; HN001 High 0.09 ± 0.04; mean% ± SEM). An exception to the overall decrease in Gram-positive bacteria was Lactobacillus, which was significantly higher in piglets fed the high dose of HN001 compared to the Control group (Q < 0.01).

Changes to the cecal microbiome were also apparent in the differences in the relative abundances of genes related to a wide range of metabolic processes and pathways. Analysis of genes mapped to SEED Subsystems level 2 functions showed 158 out of 597 functions had significantly different abundances (global ANCOM-BC Q < 0.05). The 50 most relatively abundant significantly different level 2 functions are shown in Table 2. These included genes involved in carbohydrate metabolism; genes annotated to the SEED function Lactate utilization were more relative abundant in the HN001 Low and HN001 High group compared to Controls (Q < 0.001), while genes categorized to Pyruvate:ferredoxin oxidoreductase and Methanogenesis were lower in relative abundance for both HN001 groups compared to Controls (Q < 0.001). Other SEED categories related to methanogens and methane metabolism also differed in relative abundance; Carbon monoxide induced hydrogenase, H₂:CoM-S-S-HTP oxidoreductase, and Aromatic amino acid interconversions with aryl acids were also lowered in both HN001 groups compared to Controls (Q < 0.05), while genes assigned to Formate dehydrogenase function were increased in piglets receiving HN001 (Q = 0.009).

Other differences included genes involved in amino acid metabolism, nitrogen metabolism, cell wall and capsule related genes, cell motility, and virulence, all of which in general were more abundant in the cecal microbiome of piglets fed HN001. Further differences included genes related to sulfur metabolism, vitamin K metabolism (menaquinone and phylloquinone biosynthesis), and choline and betaine metabolism, which were also significantly (Q < 0.05) more abundant in HN001 fed piglets. Finally, other gene categories that differed included phage related genes, which were less abundant in HN001 groups.

Fecal microbiome

While feeding L. rhamnosus HN001 clearly impacted the cecal microbiome, the effects were much less noticeable in the feces. Fecal communities did not display conspicuous separation between groups by PCA (Figure 3) and permutation MANOVA only trended toward significance (P = 0.07). Furthermore, at the genus level, no significant differences were found between groups (FDR > 0.4). Although fecal communities did not appear to differentiate by treatment unlike in the cecum, fecal community composition did correlate with the cecal community from the same piglet. Procrustes rotation analysis showed a significant correlation (r = 0.7, P < 0.001) between PCA projections of the cecal and fecal microbial communities (Figure 4).

FIGURE 3

FIGURE 4

Gut tissue transcriptome

The median number of paired-end reads per sample following quality trimming was 13.4 M, with a standard deviation, minimum, and maximum number of reads of 1.9, 10.4, and 18.3 M reads, respectively.

RNA-seq analysis of the gut tissue transcriptome showed changes in expression of 3 transcripts at the highest HN001 dose in the cecum (Figure 5). Of these genes, 2 had known functions; FKBP2 (Ensembl ID ENSSSCG00000033757) and CYP4 × 1 (Ensembl ID ENSSSCG00000031778). FKBP2, which encodes FK506 Binding Protein 2, had a 2.6-fold lower expression in the HN001 High group compared to Controls (FDR = 0.001). Similarly, expression of CYP4 × 1, which encodes one of the Cytochrome P450 family of enzymes, was also downregulated in the HN001 High group compared to Controls (6.6-fold lower, FDR = 0.001).

FIGURE 5

While at the per gene level L. rhamnosus HN001 only modified the expression of three genes and only in the HN001 High group, GSEA identified 10 KEGG pathways that were differentially expressed (P < 0.05) in the cecum of piglets in the HN001 High group compared to the Control group, and these two pathways were also differentially expressed in the HN001 Low group (Figure 6).

FIGURE 6

Analysis of the collective gene expression profiles in each of these KEGG pathways by permutation ANOVA also confirmed the overall significant changes in expression patterns in four pathways between the HN001 High and the Control groups; Tight junction (ssc04530), Basal transcription factors (ssc03022), and RNA transport (ssc03013) pathways were more highly expressed in the HN001 High group (P = 0.005, 0.034, and 0.008, respectively), whereas the Regulation of autophagy (ssc04140) pathway showed higher expression in the Control group (P = 0.008). Although expression of KEGG pathways were not significantly altered in the HN001 Low group compared to Controls, with the exception of the GnRH signaling pathway (ssc04912) and Regulation of autophagy (ssc04140) pathway, expression profile patterns were generally intermediate between Controls and the HN001 High group piglets (Figure 7). Consistent with the microbiome results where L. rhamnosus HN001 significantly altered the cecal microbiome but only had minor effects on the fecal microbiome, feeding L. rhamnosus HN001 did not lead to any genes passing the thresholds for differential expression in the colon, nor did GSEA show any differentially expressed KEGG pathways.

FIGURE 7

Discussion

In this study we show that supplementation with L. rhamnosus HN001 can have dramatic effects on the microbiome of the cecum in a piglet model, with broad changes in relative abundance of the dominant taxa. At the same time, significant effects on the fecal microbiome were not observed. Associated with the change in cecal microbiota composition, we showed L. rhamnosus HN001 also altered the cecal tissue transcriptome compared to controls. However, this effect was not observed in the colon tissue transcriptome, which is consistent with our results from the fecal microbiome.

Supplementation with L. rhamnosus HN001 at the highest dose led to increased proportions of L. rhamnosus in the cecum and in the feces, as expected. However, L. rhamnosus HN001 at the lower dose did not lead to a significant increase. Similarly, the relative abundance of the wider Lactobacillus genus overall was increased in the HN001 High group. Interestingly, the relative abundance of the Lactobacillus genus was substantially higher than that of L. rhamnosus itself (over 40-fold higher). Because taxonomic identification to the species level is not always reliable from either 16S rRNA sequences or shotgun metagenome data (33, 34), it is uncertain whether the increased Lactobacillus was due to increased sequences originating from L. rhamnosus that could not be classified to the species level due to factors such as read length and alignment position, or whether it was from increases in Lactobacillus of different species. Lactobacillus was present in the cecum of control piglets, indicating the presence of this genus as an autochthonous member of the gut microbial community.

While cecal Lactobacillus proportions increased from around 3% in control piglets to around 6% in the HN001 high group, greater changes, both in terms of fold change and relative abundance, were observed in a wide range of taxa. Bacteroidetes in the cecum were increased by both low and high doses of HN001 and within the Bacteroidetes, the most prominent change occurred in Prevotella, which increased over threefold on supplementation with HN001. Despite being a prominent member of the human gastrointestinal microbiome, the role of Prevotella is still relatively unclear. There is evidence for Prevotella as a biomarker for diets high in plant fiber (35, 36), and increased Prevotella has been associated with improved glucose response (37) and reduction in body fat (38). Conversely, a reduction in Prevotella has been associated with increased behavioral problems in infants at 2 years of age (39). In contrast, there are also studies that show increased Prevotella can exacerbate intestinal inflammation and autoimmune disease (40–42). However, in the absence of an underlying health issue, the balance of the evidence points to Prevotella being an important member of the gut community that contributes to polysaccharide breakdown and SCFA production (43). Indeed, the increase in abundance of genes related to utilization of polysaccharides such as xylose, trehalose, and fucose in HN001 groups is probably reflective of the increase in Prevotella (44–46). The most likely polysaccharide substrates relevant to the piglets in our study are likely to be host-derived glycans, such as mucin and lactose, which can be metabolized by different species of Prevotella (47–49).

While it is unclear why supplementation with HN001, a Lacticaseibacillus, would stimulate an increase in Prevotella, a previous study has shown that bacteriocin-producing Ligilactobacillus salivarius increased intestinal Prevotella in mice, an effect which was not observed with L. salivarius lacking the specific bacteriocin gene (50). Bacteriocins produced by L. rhamnosus HN001 appear to have a limited spectrum and are only active against some Gram positive bacteria (51), which could be providing a competitive advantage for Prevotella, leading to their increase in relative abundance.

Other changes in the cecal microbiome included a significant decrease in Methanobrevibacter in the HN001 groups. In the gut, Methanobrevibacter primarily converts CO₂ and H₂ to form methane, although formate can also be used as substrate for methanogenesis (52). For methanogens, an important limiting factor is the concentration of H₂, which in the gut is mainly produced by microbial fermentation of carbohydrates (53). Interestingly, the use of lactic acid bacteria has been proposed for reducing methane production in ruminants, with one of the hypothesized mechanisms of action being the promotion of lactate-utilizing microbes and shifting microbial fermentation toward pathways that reduce the formation of hydrogen (54). Supporting this theory, we found genes involved in lactate utilization were increased in the cecum community of piglets supplemented with HN001. Furthermore, free hydrogen appears to be produced primarily by the Firmicutes (55), which were significantly reduced in piglets supplemented with HN001. Other differences that indicate pathways involved in carbohydrate fermentation and hydrogen utilization are altered by HN001 include changes in gene abundances related to methanogen metabolism, such as the SEED functions Carbon monoxide induced hydrogenase, H2:CoM-S-S-HTP oxidoreductase, and Aromatic amino acid interconversions with aryl acids. Carbon monoxide induced hydrogenase genes can be used by methanogens to use methyl groups as a substrate (56), while reduction of heterodisulfide (CoM-S-S-HTP) is a key energy metabolism pathway in methanogenic Archaea (57). Interestingly, the SEED function Aromatic amino acid interconversions with aryl acids, an important pathway used by methanogens to transform aromatic amino acids to aryl amino acids (58), was also reduced in both HN001 groups compared to Controls. Another factor which may impact the relationship between hydrogen availability and methane production are formate dehydrogenase genes, which convert formate to CO₂ and H₂. Genes assigned to this function were more abundant in the HN001 groups.

The cecal microbiomes of HN001 fed piglets were also characterized by a noticeable increase in the abundance of genes related to amino acid, nitrogen, and protein metabolism. A wide variety of gut microbes are able to decarboxylate amino acids to produce an amine plus carbon dioxide, and of these, lactobacilli and Proteobacteria are prominent representatives (59, 60), both of which were also increased following HN001 supplementation. Further conversion steps following decarboxylation of amino acids can lead to a variety of end products, but the most abundant are actually SCFAs (60). This is an important point as fermentation of amino acids and proteins is often assumed to lead to toxic compounds, and therefore negative impacts on the host. However, there is no evidence of any such toxicity effects arising from HN001. The metagenomes from piglets fed L. rhamnosus HN001 were also characterized by a prominent increase in the relative abundance of genes relating to K group vitamins. Lactobacilli have been shown to produce vitamin K (61, 62), which is a plausible explanation for the differences observed in these pathways. Changes in vitamin biosynthesis in the gut may also have wider impacts on the microbiome ecosystem as many commensal bacteria depend on the supply of vitamins through cross-feeding, including prominent butyrate producing members of the Firmicutes, such as Faecalibacterium prausnitzii (63). Together, these observations may explain how the dietary supplementation of a single lactic acid-producing bacterium, L. rhamnosus HN001, can lead to such dramatic changes in the overall microbiome composition, by influencing microbiome carbohydrate fermentation, amino acid metabolism, and vitamin availability.

In addition to the extensive alterations in the microbial community composition of the cecum, supplementation with L. rhamnosus HN001 also led to changes in gene expression profiles in the cecal tissue. Although analysis at the gene level only revealed two differentially expressed genes with known functions, GSEA highlighted several differentially expressed pathways. Overall expression of genes involved in tight junction formation and activity were more highly expressed in the cecum of piglets receiving the higher dose of L. rhamnosus HN001, compared to Controls. Tight junction proteins are important molecules that strongly influence intestinal barrier integrity, and perturbations in this barrier can lead to inflammation and serious disorders in the gastrointestinal tract and other parts of the body (64). Previous studies have shown that L. rhamnosus HN001 can beneficially enhance intestinal barrier integrity, both in cell (9) and animal models (65). Treatment with L. rhamnosus HN001 in animal models has also been shown to decrease the severity of necrotizing enterocolitis, which is characterized by extensive destruction of the intestinal epithelial cell layer (6). In this instance, the protective effects of HN001 were modulated by the activation of Toll-like receptor 9 (6), a receptor for microbial DNA expressed in immune system cells including dendritic cells, and other antigen presenting cells (66).

Related to the effects on tight junction gene expression, L. rhamnosus HN001 also altered expression in pathways related to neutrophin signaling and autophagy. The neurotrophic factor glial cell-derived neurotrophic factor (GDNF) has been shown to attenuate inflammation by increasing expression of tight junction proteins in a mouse model (67). Increased expression of another neurotrophic factor, brain-derived neurotrophic factor (BDNF), has also been shown to suppress autophagy in mice (68). In our study, expression of neutrophin signaling pathways in the cecal tissue was increased by L. rhamnosus HN001 at the high dose, while expression of autophagy pathways was decreased. Autophagy is a cellular process controlling the ordered removal of dysfunctional cells, which plays a major role in the regulation of inflammation (69). Direct links between autophagy proteins and tight junction integrity has also been demonstrated in previous studies. For example, autophagy-related protein-6 (ATG6) has been shown to disrupt tight junction integrity in a cell model by promoting the endocytosis of the tight junction protein occludin (70), while in a rat model of intestinal inflammation, decreased autophagy was associated with upregulation of claudin-2 (71), another tight junction protein. These studies and our results highlight how inflammation, intestinal barrier function and the intestinal microbes, whether resident or introduced, are interconnected.

Our study also highlights the importance for considering the specific location in the gastrointestinal tract when studying the microbiome; almost all microbiome studies in humans use fecal samples as a proxy for the microbial communities within the gut. While using fecal samples is usually the only option due to the difficulty of accessing the communities within the gut of a healthy human subject, our study shows that care must be taken when interpreting the results because they may not adequately reflect any differences seen within the entire gastrointestinal tract.

Conclusion

Supplementation of infant formula with L. rhamnosus HN001 can have dramatic effects on the developing cecal microbiome in a piglet model. The changes observed could not be explained simply by the expansion of the Lactobacillus genus as the magnitude of differences in other taxa was greater than the magnitude of Lactobacillus increase in the HN001 supplemented groups. The differences in taxonomic composition and relative abundances of gene functions in the cecum suggests L. rhamnosus HN001-induced changes in the microbiome included alterations in carbohydrate, hydrogen, methane, and amino acid metabolism. Concomitant with alterations in the cecal microbiome, host gene expression profiles in the cecal tissue were also impacted by L. rhamnosus HN001 supplementation. Changes in host gene expression induced by L. rhamnosus HN001 is consistent with improved intestinal barrier integrity and decreased inflammation. While L. rhamnosus HN001-induced changes in the cecal microbiome and cecal tissue gene expression were apparent, differences in the colonic tissue gene expression and fecal microbiome were less distinct. This work shows L. rhamnosus HN001 can influence the microbiome and host physiology, with changes that are likely to be beneficial to the host. While our study has shown L. rhamnosus HN001 impacts the cecal microbiome composition and metagenome, future studies could gain further insights by investigating the microbiome meta-transcriptome and metabolome to better understand the activity of the microbiome. Finally, the translation of animal model results, such as ours, to humans requires careful attention and confirmation in follow up studies.

References

• 1.

  Joint FAO/WHO Working Group.Probiotics in Food: Health and Nutritional Properties and Guidelines for Evaluation.Geneva: World Health Organization (2006).

  • Google Scholar

• 2.

  BraeggerCChmielewskaADecsiTKolacekSMihatschWMorenoLet alSupplementation of infant formula with probiotics and/or prebiotics: a systematic review and comment by the ESPGHAN committee on nutrition.J Pediatr Gastroenterol Nutr. (2011) 52:238–50.

  • Google Scholar

• 3.

  WickensKBarthowCMitchellEAStanleyTVPurdieGRowdenJet alMaternal supplementation alone with Lactobacillus rhamnosus HN001 during pregnancy and breastfeeding does not reduce infant eczema.Pediatr Allergy Immunol. (2018) 29:296–302. 10.1111/pai.12874

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4.

  SlykermanRFHoodFWickensKThompsonJMDBarthowCMurphyRet alEffect of Lactobacillus rhamnosus HN001 in pregnancy on postpartum symptoms of depression and anxiety: a randomised double-blind placebo-controlled trial.EBioMedicine. (2017) 24:159–65. 10.1016/j.ebiom.2017.09.013

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5.

  WickensKBarthowCMitchellEAKangJvan ZylNPurdieGet alEffects of Lactobacillus rhamnosus HN001 in early life on the cumulative prevalence of allergic disease to 11 years.Pediatr Allergy Immunol. (2018) 29:808–14. 10.1111/pai.12982

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6.

  GoodMSodhiCPOzolekJABuckRHGoehringKCThomasDLet alLactobacillus rhamnosus HN001 decreases the severity of necrotizing enterocolitis in neonatal mice and preterm piglets: evidence in mice for a role of TLR9.Am J Physiol Gastrointest Liver Physiol. (2014) 306:G1021–32.

  • Google Scholar

• 7.

  TannockGWMunroKHarmsenHJWellingGWSmartJGopalPK. Analysis of the fecal microflora of human subjects consuming a probiotic product containing Lactobacillus rhamnosus DR20.Appl Environ Microbiol. (2000) 66:2578–88. 10.1128/AEM.66.6.2578-2588.2000

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8.

  MurphyRMorganXCWangXYWickensKPurdieGFitzharrisPet alEczema-protective probiotic alters infant gut microbiome functional capacity but not composition: sub-sample analysis from a RCT.Benef Microb. (2019) 10:5–17. 10.3920/BM2017.0191

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 9.

  AndersonRCCooksonALMcNabbWCKellyWJRoyNC. Lactobacillus plantarum DSM 2648 is a potential probiotic that enhances intestinal barrier function.FEMS Microbiol Lett. (2010) 309:184–92. 10.1111/j.1574-6968.2010.02038.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10.

  UlluwishewaDAndersonRCYoungWMcNabbWCvan BaarlenPMoughanPJet alLive faecalibacterium prausnitzii in an apical anaerobic model of the intestinal epithelial barrier.Cell Microbiol. (2015) 17:226–40. 10.1111/cmi.12360

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 11.

  VillenaJAsoHKitazawaH. Regulation of toll-like receptors-mediated inflammation by immunobiotics in bovine intestinal epitheliocytes: role of signaling pathways and negative Regulators.Front Immunol. (2014) 5:421. 10.3389/fimmu.2014.00421

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12.

  MoensFVan den AbbeelePBasitAWDodooCChatterjeeRSmithBet alA four-strain probiotic exerts positive immunomodulatory effects by enhancing colonic butyrate production in vitro.Int J Pharm. (2019) 555:1–10. 10.1016/j.ijpharm.2018.11.020

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 13.

  ZmoraNZilberman-SchapiraGSuezJMorUDori-BachashMBashiardesSet alPersonalized gut mucosal colonization resistance to empiric probiotics is associated with unique host and microbiome features.Cell. (2018) 174:1388–405.e21. 10.1016/j.cell.2018.08.041

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14.

  HouQZhaoFLiuWLvRKhineWWTHanJet alProbiotic-directed modulation of gut microbiota is basal microbiome dependent.Gut Microbes. (2020) 12:1736974. 10.1080/19490976.2020.1736974

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15.

  LavelleALennonGO’SullivanODochertyNBalfeAMaguireAet alSpatial variation of the colonic microbiota in patients with ulcerative colitis and control volunteers.Gut. (2015) 64:1553–61. 10.1136/gutjnl-2014-307873

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16.

  PuimanPStollB. Animal models to study neonatal nutrition in humans.Curr Opin Clin Nutr Metab Care. (2008) 11:601–6. 10.1097/MCO.0b013e32830b5b15

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17.

  CalderPCKrauss-EtschmannSde JongECDupontCFrickJSFrokiaerHet alEarly nutrition and immunity - progress and perspectives.Br J Nutr. (2006) 96:774–90.

  • Google Scholar

• 18.

  FilJEJoungSHayesCADilgerRN. Influence of rearing environment on longitudinal brain development, object recognition memory, and exploratory behaviors in the domestic pig (Sus scrofa).Front Neurosci. (2021) 15:649536. 10.3389/fnins.2021.649536

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 19.

  Food Standards Australia New Zealand.Microbiology risk assessment: L-lactic acid producing microorganisms.Proceedings of the Safety & Food Technology Consultation Paper Proposal p1028–Review of the Regulation of Infant Formula Products.Wellington: Food Standards Australia New Zealand (FSANZ) (2021). P. 156–21.

  • Google Scholar

• 20.

  BolgerAMLohseMUsadelB. Trimmomatic: a flexible trimmer for Illumina sequence data.Bioinformatics. (2014) 30:2114–20. 10.1093/bioinformatics/btu170

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 21.

  ZhangJKobertKFlouriTStamatakisA. PEAR: a fast and accurate illumina paired-end reAd mergeR.Bioinformatics. (2014) 30:614–20. 10.1093/bioinformatics/btt593

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 22.

  BushnellB.(Sponsor Org.: USDOE Office of Science (SC)).Berkeley, CA: U.S. Department of Energy (2021).

  • Google Scholar

• 23.

  BuchfinkBXieCHusonDH. Fast and sensitive protein alignment using diamond.Nat Methods. (2015) 12:59–60. 10.1038/nmeth.3176

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24.

  HusonDHBeierSFladeIGórskaAEl-HadidiMMitraSet alMegan community edition - interactive exploration and analysis of large-scale microbiome sequencing data.PLoS Comput Biol. (2016) 12:e1004957. 10.1371/journal.pcbi.1004957

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 25.

  OverbeekRBegleyTButlerRMChoudhuriJVChuangHYCohoonMet alThe subsystems approach to genome annotation and its use in the project to annotate 1000 genomes.Nucleic Acids Res. (2005) 33:5691–702. 10.1093/nar/gki866

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26.

  DixonP. Vegan, a package of R functions for community ecology.J Veg Sci. (2003) 14:927–30. 10.1111/j.1654-1103.2003.tb02228.x

  • CrossRef
  • Google Scholar

• 27.

  LinHPeddadaSD. Analysis of compositions of microbiomes with bias correction.Nat Commun. (2020) 11:3514. 10.1038/s41467-020-17041-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28.

  DobinADavisCASchlesingerFDrenkowJZaleskiCJhaSet alStar: ultrafast universal RNA-seq aligner.Bioinformatics. (2013) 29:15–21. 10.1093/bioinformatics/bts635

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29.

  RobinsonMDMcCarthyDJSmythGK. edgeR: a bioconductor package for differential expression analysis of digital gene expression data.Bioinformatics. (2010) 26:139–40. 10.1093/bioinformatics/btp616

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 30.

  SmythGK. Linear models and empirical bayes methods for assessing differential expression in microarray experiments.Stat Appl Genet Mol Biol. (2004) 3:3. 10.2202/1544-6115.1027

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31.

  KanehisaMGotoS. KEGG: kyoto encyclopedia of genes and genomes.Nucleic Acids Res. (2000) 28:27–30. 10.1093/nar/28.1.27

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32.

  CarlsonM.KEGG.db: A Set of Annotation Maps for KEGG. R Package Version 3.2.3. (2016).

  • Google Scholar

• 33.

  PeabodyMAVan RossumTLoRBrinkmanFSL. Evaluation of shotgun metagenomics sequence classification methods using in silico and in vitro simulated communities.BMC Bioinform. (2015) 16:362. 10.1186/s12859-015-0788-5

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34.

  JohnsonJSSpakowiczDJHongBYPetersenLMDemkowiczPChenLet alEvaluation of 16S rRNA gene sequencing for species and strain-level microbiome analysis.Nat Commun. (2019) 10:5029. 10.1038/s41467-019-13036-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 35.

  De FilippoCCavalieriDDi PaolaMRamazzottiMPoulletJBMassartSet alImpact of diet in shaping gut microbiota revealed by a comparative study in children from Europe and rural Africa.Proc Natl Acad Sci U S A. (2010) 107:14691–6. 10.1073/pnas.1005963107

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36.

  De FilippisFPasolliETettATaralloSNaccaratiADe AngelisMet alDistinct genetic and functional traits of human intestinal prevotella copri strains are associated with different habitual diets.Cell Host Microbe. (2019) 25:444–53.e3. 10.1016/j.chom.2019.01.004

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37.

  Kovatcheva-DatcharyPNilssonAAkramiRLeeYSDe VadderFAroraTet alDietary fiber-induced improvement in glucose metabolism is associated with increased abundance of prevotella.Cell Metab. (2015) 22:971–82. 10.1016/j.cmet.2015.10.001

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 38.

  HjorthMFBlædelTBendtsenLQLorenzenJKHolmJBKiilerichPet alPrevotella-to-Bacteroides ratio predicts body weight and fat loss success on 24-week diets varying in macronutrient composition and dietary fiber: results from a post-hoc analysis.Int J Obes. (2018) 43:149–57. 10.1038/s41366-018-0093-2

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39.

  LoughmanAPonsonbyALO’HelyMSymeonidesCCollierFTangMLKet alGut microbiota composition during infancy and subsequent behavioural outcomes.EBioMedicine. (2020) 52:102640. 10.1016/j.ebiom.2020.102640

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40.

  ScherJUSczesnakALongmanRSSegataNUbedaCBielskiCet alExpansion of intestinal Prevotella copri correlates with enhanced susceptibility to arthritis.eLife. (2013) 2:e01202. 10.7554/eLife.01202

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 41.

  IljazovicARoyUGálvezEJCLeskerTRZhaoBGronowAet alPerturbation of the gut microbiome by Prevotella spp. enhances host susceptibility to mucosal inflammation.Mucosal Immunol. (2020) 14:1–12. 10.1038/s41385-020-0296-4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42.

  LeiteAZRodriguesNCGonzagaMIPaioloJCCde SouzaCAStefanuttoNAVet alDetection of increased plasma interleukin-6 levels and prevalence of prevotella copri and Bacteroides vulgatus in the feces of type 2 diabetes patients.Front Immunol. (2017) 8:1107. 10.3389/fimmu.2017.01107

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 43.

  PrecupGVodnarD-C. Gut Prevotella as a possible biomarker of diet and its eubiotic versus dysbiotic roles: a comprehensive literature review.Br J Nutr. (2019) 122:131–40. 10.1017/S0007114519000680

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 44.

  Linares-PasténJAHeroJSPisaJHTeixeiraCNymanMAdlercreutzPet alNovel xylan-degrading enzymes from polysaccharide utilizing loci of Prevotella copri DSM18205.Glycobiology. (2021) 31:1330–49. 10.1093/glycob/cwab056

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 45.

  GálvezEJCIljazovicAAmendLLeskerTRRenaultTThiemannSet alDistinct polysaccharide utilization determines interspecies competition between intestinal Prevotella spp.Cell Host Microbe. (2020) 28:838–52.e6. 10.1016/j.chom.2020.09.012

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 46.

  Fehlner-PeachHMagnaboscoCRaghavanVScherJUTettACoxLMet alDistinct polysaccharide utilization profiles of human intestinal Prevotella copri isolates.Cell Host Microbe. (2019) 26:680–90.e5. 10.1016/j.chom.2019.10.013

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 47.

  EfimovBAChaplinAVShcherbakovaVASuzinaNEPodoprigoraIVShkoporovAN. Prevotella rara sp. nov., isolated from human faeces.Int J Syst Evol Microbiol. (2018) 68:3818–25. 10.1099/ijsem.0.003066

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48.

  RouxVRobertCRaoultD. Non-contiguous finished genome sequence of Prevotella timonensis type strain 4401737(T.).Stand Genomic Sci. (2014) 9:1344–51. 10.4056/sigs.5098948

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49.

  WrightDPRosendaleDIRobertsonAM. Prevotella enzymes involved in mucin oligosaccharide degradation and evidence for a small operon of genes expressed during growth on mucin.FEMS Microbiol Lett. (2000) 190:73–9. 10.1111/j.1574-6968.2000.tb09265.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 50.

  Riboulet-BissonESturmeMHJefferyIBO’DonnellMMNevilleBAFordeBMet alEffect of lactobacillus salivarius bacteriocin abp118 on the mouse and pig intestinal microbiota.PLoS One. (2012) 7:e31113. 10.1371/journal.pone.0031113

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 51.

  Aguilar-UscangaBRSolís-PachecoJRPlascenciaLAguilar-UscangaMGGarcíaHSLacroixM. Effect of culture medium on bacteriocin production by Lactobacillus rhamnosus HN001 and Lactobacillus reuteri ATCC 53608.J Microbiol Biotechnol Food Sci. (2013) 2:2462–8.

  • Google Scholar

• 52.

  KellyWJMackieRIAttwoodGTJanssenPHMcAllisterTALeahySC. Hydrogen and formate production and utilisation in the rumen and the human colon.Anim Microbiome. (2022) 4:22. 10.1186/s42523-022-00174-z

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 53.

  FlintHJScottKPDuncanSHLouisPForanoE. Microbial degradation of complex carbohydrates in the gut.Gut Microb. (2012) 3:289–306. 10.4161/gmic.19897

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 54.

  DoyleNMbandlwaPKellyWJAttwoodGLiYRossRPet alUse of lactic acid bacteria to reduce methane production in ruminants, a critical review.Front Microbiol. (2019) 10:2207. 10.3389/fmicb.2019.02207

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 55.

  CarboneroFBenefielACGaskinsHR. Contributions of the microbial hydrogen economy to colonic homeostasis.Nat Rev Gastroenterol Hepatol. (2012) 9:504–18. 10.1038/nrgastro.2012.85

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 56.

  DanielsLFuchsGThauerRKZeikusJG. Carbon monoxide oxidation by methanogenic bacteria.J Bacteriol. (1977) 132:118–26.

  • Google Scholar

• 57.

  SetzkeEHedderichRHeidenSThauerRK. H2: heterodisulfide oxidoreductase complex from Methanobacterium thermoautotrophicum. Composition and properties.Eur J Biochem. (1994) 220:139–48. 10.1111/j.1432-1033.1994.tb18608.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 58.

  PoratIWatersBWTengQWhitmanWB. Two biosynthetic pathways for aromatic amino acids in the archaeon Methanococcus maripaludis.J Bacteriol. (2004) 186:4940–50. 10.1128/jb.186.15.4940-4950.2004

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 59.

  PuginBBarcikWWestermannPHeiderAWawrzyniakMHellingsPet alA wide diversity of bacteria from the human gut produces and degrades biogenic amines.Microb Ecol Health Dis. (2017) 28:1353881. 10.1080/16512235.2017.1353881

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 60.

  OliphantKAllen-VercoeE. Macronutrient metabolism by the human gut microbiome: major fermentation by-products and their impact on host health.Microbiome. (2019) 7:91. 10.1186/s40168-019-0704-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 61.

  MorishitaTTamuraNMakinoTKudoS. Production of menaquinones by lactic acid bacteria.J Dairy Sci. (1999) 82:1897–903. 10.3168/jds.S0022-0302(99)75424-X

  • CrossRef
  • Google Scholar

• 62.

  LeBlancJGLaiñoJEdel ValleMJVanniniVvan SinderenDTarantoMPet alB-Group vitamin production by lactic acid bacteria – current knowledge and potential applications.J Appl Microbiol. (2011) 111:1297–309. 10.1111/j.1365-2672.2011.05157.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 63.

  Soto-MartinECWarnkeIFarquharsonFMChristodoulouMHorganGDerrienMet alVitamin biosynthesis by human gut butyrate-producing bacteria and cross-feeding in synthetic microbial communities.mBio. (2020) 11:e886–820. 10.1128/mBio.00886-20

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64.

  GroschwitzKRHoganSP. Intestinal barrier function: molecular regulation and disease pathogenesis.J Allergy Clin Immunol. (2009) 124:3–20; quiz 21–2. 10.1016/j.jaci.2009.05.038

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 65.

  TannockGWTaylorCLawleyBLoachDGouldMDunnACet alAltered transcription of murine genes induced in the small bowel by administration of probiotic strain Lactobacillus rhamnosus HN001.Appl Environ Microbiol. (2014) 80:2851–9. 10.1128/AEM.00336-14

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66.

  KawasakiTKawaiT. Toll-like receptor signaling pathways.Front Immunol. (2014) 5:461. 10.3389/fimmu.2014.00461

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 67.

  ReinshagenMRohmHSteinkampMLiebKGeerlingIVon HerbayAet alProtective role of neurotrophins in experimental inflammation of the rat gut.Gastroenterology. (2000) 119:368–76. 10.1053/gast.2000.9307

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 68.

  NikoletopoulouVSidiropoulouKKallergiEDaleziosYTavernarakisN. Modulation of autophagy by bdnf underlies synaptic plasticity.Cell Metab. (2017) 26:230–42.e5. 10.1016/j.cmet.2017.06.005

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 69.

  Matsuzawa-IshimotoYHwangSCadwellK. Autophagy and inflammation.Ann Rev Immunol. (2018) 36:73–101. 10.1146/annurev-immunol-042617-053253

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70.

  WongMGanapathyASSuchanecELaidlerLMaTNighotPet alIntestinal epithelial tight junction barrier regulation by autophagy-related protein ATG6/beclin 1.Am J Physiol Cell Physiol. (2019) 316:C753–65. 10.1152/ajpcell.00246.2018

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 71.

  HuangLZhangDHanWGuoC. High-mobility group box-1 inhibition stabilizes intestinal permeability through tight junctions in experimental acute necrotizing pancreatitis.Inflam Res. (2019) 68:677–89. 10.1007/s00011-019-01251-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar