Clostridium butyricum and Clostridium tyrobutyricum: angel or devil for necrotizing enterocolitis?

ABSTRACT Necrotizing enterocolitis (NEC) is a challenging gastrointestinal disease that disproportionately affects premature neonates, with high incidence and case-fatality rates. Despite extensive research efforts, the pathogenesis and mechanisms of NEC remain unclear, making it difficult to effectively eradicate. However, it has been established that dysbiosis of gut microbes occurs before the onset of NEC, providing compelling evidence for the potential use of probiotic therapy. As such, we have focused our attention on two probiotics in particular: Clostridium butyricum and Clostridium tyrobutyricum, especially in light of recent breakthroughs that have linked several Clostridia species with NEC. To determine whether C. butyricum and C. tyrobutyricum are pathogenic or probiotic, we conducted a comparison of the phenotypic traits of NEC mice treated with each bacterium. Our results confirm that treatment with C. tyrobutyricum restores intestinal barrier integrity and alleviates inflammatory immune responses associated with NEC. In contrast, treatment with C. butyricum exacerbates intestinal barrier damage and promotes immune disorder, including increased numbers of macrophages, monocytes, and neutrophils in the intestinal lamina propria. Further analysis of the gut microbiome suggests that the positive effects of C. tyrobutyricum treatment are associated with an increase in Akkermansia muciniphila, while C. butyricum treatment decreases the level of A. muciniphila, which accounts for its negative effect on NEC. This study sheds light on the fact that treatment with C. tyrobutyricum, but not C. butyricum, has the potential to protect against NEC development. The opposite effects of these two probiotics on NEC may result from their different modulation of the level of A. muciniphila, a gut microbe that is closely associated with intestinal homeostasis. In summary, by improving the abundance of A. muciniphila to alleviate intestinal inflammation and enhance intestinal barrier integrity, supplementation with C. tyrobutyricum may become a promising therapy for NEC. IMPORTANCE This study sheds light on that treatment with Clostridium tyrobutyricum but not Clostridium butyricum is entitled to protect against necrotizing enterocolitis (NEC) development potentially. The mechanisms behind the opposite effect on NEC may result in different modulation on the level of Akkermansia muciniphila, which is deeply associated with intestinal homoeostasis. Briefly, through improving the abundance of A. muciniphila to alleviate intestinal inflammation and enhance intestinal barrier integrity, C. tyrobutyricum supplement may become a promising therapy for NEC.

developing NEC (2).The development of NEC is attributed not only to antenatal risk factors such as intrauterine inflammation, infection, and preeclampsia (3) but also to postnatal risk factors including prematurity, sepsis, formula feeding, and gut microbiota dysbiosis (4).However, due to inadequate understanding of the complex and multifactorial pathogenesis and mechanisms of NEC, the mortality of NEC neonates remains high, fluctuating between 20% and 30% even after surgery (5).Despite the complex pathogenesis of NEC, a wealth of literature indicates that an imbalance of gut microbiota plays a critical role in the development of the disease, often preced ing its onset (6).Early bacterial colonization is also crucial for the formation of intes tinal barrier integrity and systemic immune function in infants.When the chemical and physical barriers of the intestine become aberrant, bacterial invasion can occur, leading to perforation, necrosis, and systemic inflammatory reactions (7).Therefore, gut dysbiosis is a core risk factor to be reckoned with during the development of NEC, and probiotic therapy is emerging as an increasingly important approach to address this issue (8,9).Metagenomic analysis has revealed an increased relative abundance of the phylum Proteobacteria and a decreased relative abundance of the phyla Firmicutes and Bacteroidetes in infants with NEC (10).However, whether the genus Clostridium is positively or negatively associated with NEC remains unclear, posing a new challenge for the development of NEC therapeutic strategies.Some studies have suggested that specific strains of Clostridium, such as C. butyricum (11), Clostridium perfringens (12), and Clostridium neonatale (13), are positively associated with NEC.However, one study reported a decreased abundance of Clostridia with the development of NEC (14).Therefore, we focused our attention on two probiotics of Clostridium: C. butyricum and C. tyrobutyricum, both of which are butyric acid-producing bacteria and have been applied to treat gastroenteritis clinically or experimentally (15)(16)(17).To determine whether these Clostridia are pathogenic or probiotic for NEC, we compared the phenotypic traits, intestinal barrier integrity, and inflammatory immune response of NEC in a mouse model treated with the two Clostridia.Our results reveal that C. tyrobutyricum alleviates the symptoms of NEC, while C. butyricum promotes the development of NEC.Furthermore, 16S rDNA analysis revealed that C. tyrobutyricum treatment enhanced the abundance of A. muciniphila, while C. butyricum treatment weakened it, indicating the potential for interspecific competition and colonization resistance between these bacteria.Overall, our findings provide new insights into the screening of potential probiotics for clinicians to treat NEC with multiple strategies and highlight the significance of early bacterial colonization for infants, as well as the potential mechanisms of interspecific competition among these bacteria.

Human sample collection
Human samples, including colons and feces, were collected from the Jiangning Affiliated Hospital of Nanjing Medical University and Nanjing Children's Hospital, respectively, in sterile surgery rooms.Colon samples were transferred using germ-free 50-mL Eppendorf tubes containing sterile RPMI Medium 1640, and feces samples were transferred using germ-free collection tubes.All samples were transported under dry ice conditions for further study.The information for all human samples was provided in Table S1.

Mouse model of necrotizing enterocolitis with or without treatment
Pregnant C57/BL6 mice were purchased from Jiangsu Jicuiyaokang Biotechnology Co., LTD.When newborn mice were 1 wk old, mice were randomly divided into different groups.Control mice were fed with breast milk without separated from their dams.Necrotizing enterocolitis (NEC) mice were co-housed in a neonatal incubator at 32-35°C and 60-70% humidity.They were hand-fed with formula milk (40 µL/g), consisted of 15 g Similac Advance infant formula (Abbott Nutrition) and 10 g Esbilac canine milk replacer (PetAg) in 75 mL water with or without bacteria (12.5 µL stool slurry per milliliter or 10 9 colony-forming units per milliliter) six times a day.Meanwhile, they were stressed with hypoxia (5% O 2 , 95% N 2 ) for 10 min, followed by cold stress (placement in a refrigerator at 4°C) for 5 min later twice a day.Mice were humanely sacrificed after 1 wk and the intestines were collected immediately for further experiments.Briefly groups were as follows: control group, breast-fed only; NEC group, formula-fed, hypoxia, and cold stress; feces group, formula-fed with bacterial slurry, hypoxia, and cold stress; C. butyricum group, formula-fed with C. butyricum, hypoxia, and cold stress; C. tyrobutyricum group, formula-fed with C. tyrobutyricum, hypoxia, and cold stress.

Histological analysis
Mouse intestinal tissue sections were soaked in 4% Carnoy's fixative for 24 h and paraffinembedded.Then, hematoxylin-eosin staining (Leagene, DH0006) and Alcian blue staining (Leagene, 0041) were performed on paraffin sections according to manufacturer's protocol.All histological evaluation of the ileum and colon were conducted and scored in a blinded manner.The total score obtained was statistically analyzed.Alcian blue staining was used for quantifying the coverage of mucus layer.

Isolation of the lamina propria mononuclear cells (LPMCs) from ileum murine and colon
Fresh ileum and colon tissues were collected when mice were sacrificed.After being washed with cold PBS, tissues were cut into 1.5 cm and incubated in 1 mM dithiothreitol (DTT, Sigma-Aldrich, D9779) and 10 mM HEPES solution (Sigma-Aldrich, 375368) for 10 min at 37°C.After vigorous shaking and washing with PBS, we used 30 mM ethyl enediaminetetraacetic acid (EDTA, Sigma-Aldrich, 798681) and 10 mM HEPES solution to incubate tissues twice.Then tissues were incubated in RPMI Medium 1640 (Gibco, 31800-022) with 0.2 mg/mL collagenase I (Sigma-Aldrich, C2674) and 0.15 mg/mL DNase (Sigma-Aldrich, AMPD1) to digest tissues for 90 min at 37°C.After a vigorous stirrer, the cells were purified by using a 70-µm filter and were collected by centrifugation for 5 min at 500g.A discontinuous Percoll (Cytiva, 17089102) gradient (80%/40%) was then used to separate mononuclear cells through centrifugation for 30 min at 1,000g without break.The LPMCs were collected at the interface between two Percoll gradients for further flow cytometry.

Fluorescence in situ hybridization
Ileum tissues were fixed in 4% Carnoy's fixative and embedded in paraffin, as previously described.The tissues were then sectioned to a thickness of 4 µm and hybridized with a universal bacterial probe targeting the 16S rRNA gene (EUB338 probe: 5′-GCTGCC TCCCGTAGGAGT-3′).The samples were stained overnight at 4°C with mucin2 (Santa, sc-515032, 1:300) and Hoechst for 10 min at room temperature.Fluorescence analysis was performed using a Leica SP8 fluorescence microscope.The percentage of bacteria was quantified per square millimeter of ileum in each group.

RNA sequencing
Total RNAs of colon were extracted using Trizol Reagent and purified by Ribo-ZeroTM Magnetic Gold Kits (Illumina, MRZG126) to remove rRNA.We collected 3 µg RNAs of each biological sample to construct sequencing library using NEB Next Ultra Directional RNA LibraryPrep Kit for Illumina (NEB, Ispawich, USA).The mRNAs were measured by DEGseq software to find differential expressed genes (DEG) and DEGs were further annotated through NCBI, Uniport, GO, and KEGG database.

Cell viability assay
Cell viability was determined by the Cell Counting Kit-8 (CCK-8, APExBIO, K1018) assay.We seeded 1000 cells for 96-well plate in each well, respectively.After being treated with different bacteria directly with multiplicity of infection of 100 for 4 h under anaerobic condition, we removed the culture medium containing bacteria instead of 1640 medium with 10% FBS, 1% penicillin/streptomycin, and 100 mg/mL gentamycin for 2 h to kill the extracellular bacteria in wells.Then the medium was replaced by 1640 medium supplemented with 10% FBS and 10% CCK-8 for further analysis.

Bacterial competition in liquid cultures
The overnight cultures of C. butyricum, C. tyrobutyricum, and A. muciniphila were prepared for competition experiments beforehand.For direct competition, all bacte rial cultures were diluted to an initial OD 600 of 0.1 and bacteria were centrifuged at 2,000 rpm for 5 min.For hot-killed bacterial experiment, C. butyricum and C. tyrobutyri cum cultures were placed in an air oven at 70°C for 30 min to be killed completely.Then C. butyricum and C. tyrobutyricum were inoculated with A. muciniphila, respectively, resuspended in a mixed culture medium (half RCM and half BHI with 0.1% mucin), whereas co-culture of C. butyricum + A. muciniphila and C. tyrobutyricum + A. muciniphila were tested with an initial OD 600 of 0.05 of each strain.For bacterial indirect competition, after the C. butyricum and C. tyrobutyricum cultures grew overnight at an initial OD 600 of 0.1in a mixed culture medium, the conditioned medium supernatant was collected by centrifuging at 4,000 rpm for 10 min.Then conditioned mediums were used to culture A. muciniphila respectively with an initial OD 600 of 0.05 of A. muciniphila.Bacterial cultures were collected at 0 h, 24 h, 48 h and 72 h of each experiment simultaneously.Total bacterial DNA were extracted using Bacterial DNA Kit (OMEGA, D3350-01).Quantitative PCR on a 7500 Sequence Detector (Applied Biosystems, CA, USA) was used to calculate the number of C. butyricum, C. tyrobutyricum, and A. muciniphila 16s rDNA gene copies in the co-culture medium.The primers were as follows: A. muciniphila F, 5′-GTTCGGAATCAC TGGGCGTA-3′; A. muciniphila R, 5′-CGCATTTCACTGCTACACCG-3′.

Statistical analysis
Statistical significance was determined with t-test, one-way ANOVA test, or two-way ANOVA test, as indicated in the figure legends.Error bars indicate SEM on all graphs.Prism 9.0 (Graphpad, La Jolla, CA) was used for all statistical analyses.

Gut microbiota is involved in the development of NEC
To investigate whether dysbiosis of gut microbiota is associated with the development of NEC, we compared the gut microbiota composition of NEC infants with that of normal neonates using 16S rDNA sequencing.Although there was no significant difference in alpha diversity (Fig. 1A), NMSD analysis revealed a distinct change in beta diversity of microbiota composition between the NEC group and the control group (Fig. 1B).The differential taxonomic similarity of intestinal bacteria at the phylum level between each group was reflected in Actinobacteriota, Firmicutes, and Proteobacteria (Fig. 1C).Notably, bubble plot analysis also confirmed a significant alteration at the genus level (Fig. 1D).
Microbial network analysis also suggested potential interactions between specific groups of gut microbiota (Fig. S1A).The PICRUSt2 predicted results indicate that the distinct microbiome in NEC patients plays a crucial role in the metabolism and biosynthesis of amino acids, including arginine, proline, and histidine.These findings highlight the significant impact of the differentiated microbiome on the host's amino acid processes (Fig. S1B).
To further investigate the underlying mechanism, we used a mouse model to induce NEC, including formula feeding, hypoxia, and cold stress (20), and we treated the mice with bacterial stock (12.5 µL stool slurry in 1 mL formula) cultured from the stool of NEC patients for 1 wk (Fig. 1E).Considering that NEC infants are likely to present with pneumatosis cystoides intestinalis (21), we compared the severity of pathology in each group and calculated the number of pneumatosis cystoides intestinalis (Fig. 1F and G).The results showed that bacterial treatment facilitated the luminal gas trapping of NEC mice.Moreover, ileum and colon tissue damages were more severe after bacterial gavage (Fig. 1H), as judged from H&E staining.Overall, our data indicate that microbiome disorder promotes the development of NEC.

C. tyrobutyricum protected against NEC while C. butyricum worsened NEC
Recent studies, including epidemiological studies (22), clinical signs (23), and animal models (24), have demonstrated the involvement of Clostridia in the development of NEC.To investigate whether two different Clostridium probiotics, C. tyrobutyricum and C. butyricum, can be used for NEC treatment, we fed NEC mice with mixed formula contain ing both probiotics (at 10 9 CFUs/mL), as described previously (20) (Fig. 2A).We found that mice treated with C. tyrobutyricum had milder symptoms of pneumatosis cystoides intestinalis, while the symptoms of mice treated with C. butyricum became more severe (Fig. 2B and C).Furthermore, H&E staining results verified that C. tyrobutyricum alleviated ileum and colon damage, while C. butyricum aggravated the damage (Fig. 2D).
These findings suggest that each Clostridium strain has unknown mechanisms that impact the development of NEC, resulting in the opposite effects observed with C. tyrobutyricum and C. butyricum.Overall, the study indicates the potential therapeutic benefits of C. tyrobutyricum and the potential harm of C. butyricum in the treatment of NEC.

Intestinal inflammation was alleviated by C. tyrobutyricum but aggravated by C. butyricum
To investigate the mechanisms underlying the effects of both Clostridia species on the progression of NEC, the study focused on alterations at the transcriptional level of the disease itself.RNA sequencing was performed to assess messenger (m)RNA expression between NEC and normal neonate colon tissues, taking into consideration that the usual sites of NEC were the distal ileum and proximal colon (25).As expected, there was a significant difference in mRNA expression between the two groups (Fig. 3A).A total of 1970 upregulated mRNAs were identified, and differential inflammatoryrelated genes were labeled in the volcano plot (Fig. 3B).Furthermore, KEGG pathway enrichment revealed that Toll-like receptor signaling pathway and NF-κB signaling pathway were significantly upregulated (Fig. 3C), indicating the presence of an inflammatory state in neonates with NEC.
To further confirm the activation of intestinal inflammation, we extracted immune cells from the lamina propria layer of the ileum and colon of the mice (26).The number of CD11b + F4/80 + macrophages, CD11b + LY6c + monocytes, and CD11b + LY6g + neutrophils was detected using flow cytometry in each group.Notably, treatment with C. tyrobutyri cum decreased the number of macrophages, monocytes, and neutrophils increased by NEC, while treatment with C. butyricum failed to reverse it (Fig. 3D; Fig. S2A).We also detected Th17/Treg cells in four different groups.Our results showed that the number of CD4+RORγt+ (Th17) cells and CD25+FoxP3+ (Treg) cells was significantly altered in both the colon and ileum.Specifically, the NEC model disrupted the Th17/Treg cell balance in the ileum and colon.Treatment with C. tyrobutyricum reversed the decrease of Th17 cells in NEC model mice and decreased the number of Treg cells, while C. butyricum failed to produce the same effect (Fig. S2B and C).Consistent with the RNA-sequencing data, the expression of genes in intestinal tissues, such as Tlr4, Myd88, Nf-κb, and Il-1β, were significantly upregulated and reversed by C. tyrobutyricum, but not by C. butyricum (Fig. 3E).The protein level of intestinal tissues, including TLR4, MyD88, NF-κB, and phospho-NF-κB, was examined by western blot (WB).Treatment with C. butyricum elevated the expression of TLR4, MyD88, and phospho-NF-κB, while TLR4 and phospho-NF-κB were reduced when treated with C. tyrobutyricum (Fig. 3F).
Overall, the data suggest that NEC mice experienced severe intestinal inflamma tion, and the two different Clostridia probiotics had opposite effects.C. tyrobutyricum alleviated the inflammation through modulating immune cells and the TLR4/NF-κB signaling pathway, while C. butyricum aggravated the inflammation.

Intestinal barrier integrity was protected by C. tyrobutyricum but disrupted by C. butyricum
RNA-sequencing data were further analyzed to determine the alterations between NEC and normal neonates at the mRNA level in the colon.Gene set enrichment analysis (GSEA) of the differentially expressed genes between NEC and normal neonates detected enrichment of downregulated genes characteristic of tight junctions (Fig. 4A).We labeled the genes related to tight junctions among the 2073 downregulated genes in the volcano plot (Fig. 4B).Similarly, KEGG pathway enrichment emphasized that the tight junction signaling pathway was significantly inhibited, indicating damage to the intestinal barrier integrity in NEC neonates (Fig. 4C).
To investigate the link between intestinal barrier integrity and the opposite effects of two Clostridia on the development of NEC, we performed Alcian blue staining to observe the thickness of the mucus layer in the ileum (Fig. 4D).Treatment with C. tyrobutyricum significantly increased the thickness of the mucus layer, while treatment with C. butyricum significantly reduced it.Furthermore, more bacteria were translocated into the submucous layer in the ileum of NEC mice, and treatment with C. butyricum failed to alleviate the permeability of the intestinal barrier, in contrast to the dramatic protection provided by C. tyrobutyricum treatment (Fig. 4E).Meanwhile, the expression of the tight junction protein ZO-1 and claudin-1 and the cell adhesion protein E-cadherin in the ileum and colon were reduced in NEC mice through immunofluorescence and western blotting, and C. tyrobutyricum treatment remarkably improved the expression of ZO-1 and E-cadherin (Fig. 4F and G; Fig. S3A).We also examined the mRNA levels of Muc2, Muc5ac, Tff1, and Tff3, which are associated with mucus barrier integrity (27), in intestinal tissues through real-time PCR (Fig. 4H).The results further confirmed the protective effect of C. tyrobutyricum on the mucus barrier.All in all, these data suggest that intestinal barrier integrity is protected by C. tyrobutyricum but disrupted by C. butyricum.

The positive effect of C. tyrobutyricum and the negative effect of C. butyricum on NEC were associated with modulating the level of A. muciniphila
We performed bacterial 16s rDNA sequencing on mouse feces to investigate the potential association between gut microbiota dysbiosis and NEC development, as well as the effects of two Clostridia on microbiome composition.Although the bacteria Shannon index of the four groups did not show significant changes (Fig. S4A), three-dimensional principal component analysis revealed a distinct separation of microbiota for the control, model, C. butyricum, and C. tyrobutyricum groups (Fig. 5A).We further investigated the overall bacterial composition among each group by comparing the top 10 relative abundance of bacteria at the phylum level (Fig. 5B).Consistent with a previous study (10), the levels of Bacteroidetes, Proteobacteria, and Firmicutes were significantly altered.We observed a decrease in Verrucomicrobia in the NEC and C. butyricum groups and an increase in the C. tyrobutyricum group.The analysis of the genus confirmed the decline of Akkermansia in the NEC group and the improvement of Akkermansia in the C. tyrobutyri cum group (Fig. 5C; Fig. S4B).The result of the linear discriminant analysis effect size (LEfse) analysis also revealed this phenomenon in mice feces (Fig. 5D).Consistently, the abundance of Akkermansia significantly decreased in neonate samples (Fig. 5E).
To date, A. muciniphila is a well-known probiotic species in the genus Akkermansia that has been extensively studied and demonstrated to have beneficial effects on gut health.As a mucus-consuming bacterium in the intestine, the reduction of A. muciniphila is strongly associated with multiple diseases (28), and proper supplementation of A. muciniphila has been shown to improve intestinal barrier integrity and host immunity (29).Therefore, we measured the relative abundance of A. muciniphila using quantitative PCR in different mouse groups and found the same variation trend in mice feces (Fig. 5F).Overall, our findings suggest that, while C. butyricum treatment worsened the decline of A. muciniphila caused by the NEC model, C. tyrobutyricum treatment increased the abundance of A. muciniphila in mice feces.

Interspecific competition provided a fitness advantage to C. butyricum over A. muciniphila
To investigate the reason behind this observation, we examined whether interspecific colonization resistance between the two Clostridia and A. muciniphila could account for the variation of A. muciniphila.We prepared overnight liquid cultures of each strain for bacterial competition (30).We compared the growth rates of A. muciniphila directly with C. butyricum and C. tyrobutyricum under nutrient-limited conditions using absolute quantitative PCR (Fig. 6A).The results showed that A. muciniphila had lower growth rates when co-cultured with C. butyricum than when grown alone, but the yields did not change significantly when co-cultured with C. tyrobutyricum.We also found that hotkilled C. butyricum suppressed the growth of A. muciniphila (Fig. 6B), but the conditioned medium of C. butyricum had no effect on the growth of A. muciniphila (Fig. 6C).These findings suggest that the colonization resistance of C. butyricum to A. muciniphila is rooted in the bacterial structure itself and that the fermentation products or metabolites of C. butyricum and C. tyrobutyricum do not act as stimulatory molecules.Taken together, our data indicate that interspecific competition provided a fitness advantage to C. butyricum over A. muciniphila.

Conclusion
NEC has become a serious health threat in infants, particularly premature neonates.Despite efforts to understand its pathogenesis and underlying mechanisms, there is still an urgent need for more effective and safe therapeutic approaches (31).The gut microbiome plays a crucial role in the healthy growth of infants, including immune and intestinal barrier development, and recent insights suggest that gut microbiome dysbiosis occurs before the onset of NEC (32).Our 16S rDNA data and experimental design confirmed the involvement of gut microbiota in NEC development.Probiotic therapy has been shown to be beneficial for restoring intestinal homeostasis and reducing the incidence of NEC (33,34).However, not all probiotic products are suitable for NEC treatment, and there is still controversy regarding the effect of Clostridia on NEC development, with C. butyricum, a butyric acid-producing probiotic, being implicated in NEC (11).
To address this issue, we chose two Clostridium probiotics, C. butyricum and C. tyrobutyricum, to determine whether they are suitable for treating NEC.We established an NEC mouse model and fed the mice with exclusive formula mixed with either C. butyricum or C. tyrobutyricum.Surprisingly, C. tyrobutyricum treatment alleviated the severity of the NEC model, while C. butyricum treatment worsened the condition.Our RNA-sequencing results showed that NEC neonates had more intense inflammatory signaling in their intestines (35) and lower intestinal barrier integrity (36).C. tyrobutyricum treatment reduced inflammation and improved intestinal barrier integrity, while C. butyricum treatment exacerbated the condition, according to our data.We also investigated the bacterial diversity and homeostasis of human and mouse feces to identify potential pathogenic mechanisms.The 16S rDNA sequencing results indicated that the decline of A. muciniphila was associated with NEC development.Our findings supported the idea that C. tyrobutyricum treatment restored the level of A. muciniphila, while C. butyricum treatment exacerbated the loss of A. muciniphila.In addition, the fitness advantage of C. butyricum in interspecific competition may account for the reduction in A. muciniphila.
Although our study provided clear evidence of the roles of C. butyricum and C. tyrobutyricum in NEC progression, there are still some limitations to our study that need to be addressed.For example, we did not quantify the species level of C. butyricum and C. tyrobutyricum in mouse feces to provide a more detailed description of the microbiota composition in vivo.Additionally, the negative result of the competition between C. tyrobutyricum and A. muciniphila suggests another potential mechanism for exploring the fitness advantage of A. muciniphila under C. tyrobutyricum treatment.
In conclusion (Fig. 7), our study strongly supports the idea that C. tyrobutyricum treatment, but not C. butyricum treatment, is suitable for protecting against NEC by reducing intestinal inflammation and improving intestinal barrier integrity.Moreover, the

6 FIG 1
FIG 1 Gut microbiota is involved in the development of NEC.(A) Shannon index of bacteria in human feces (n = 8-10).(B) NMDS analysis of bacteria in human feces using Bray-Curtis metric distances of beta diversity (n = 8-10).(C and D) Relative abundance of bacteria composition at phylum and genus level (n = 8-10).(E) Experimental design for transplanting faeces from NEC patients to mice under NEC model (n = 10).(F and G) Representative images of enteric cavity and canal at 7 d.The number of pneumatosis cystoides intestinalis marked by arrows were counted and shown as scatter plots (n = 5).(H) H&E staining and histological scores of ileum and colon (n = 3-5).Quantified results were shown as mean ± SEM. p-Values were generated by one-way ANOVA with multiple comparisons.*P ＜ 0.05, **P ＜ 0.01,*** P ＜ 0.001.

12 FIG 5
FIG 5 The positive effect of C. tyrobutyricum and the negative effect of C. butyricum on NEC were associated with modulating the level of A. muciniphila.(A) Representative images of three-dimensional principal component analysis (PCA) of control, model, C. butyricum and C. tyrobutyricum mice (n = 5-7).(B and C) The relative abundances of bacteria at phylum and genus level were shown as a stacked bar plot.Each column corresponds to one group (n = 5-7).(D) Representative images of linear discriminant analysis effect size (LEfSe) of each two groups (n = 5-7).(E) Relative abundance of Akkermansia in neonate feces analyzed by Kruskal-Wallis test (n = 6-7).(F) Relative 16S rDNA expression of A. muciniphila measured by quantitative PCR analyzed by one-way ANOVA (n = 5).Quantified results were represented using box and whisker plots.p-Values were generated by one-way ANOVA with multiple comparisons.***P ＜ 0.001.

FIG 6
FIG 6 Interspecific competition provided a fitness advantage to C. butyricum over A. muciniphila.(A-C) Schematic diagram demonstrating the workflow of competition experiments in liquid cultures among different strains and the growth rates of A. muciniphila in each co-culture solution were represented by line charts.The growth rates of A. muciniphila were shown as the 16S rDNA copies measured by quantitative PCR (n = 3).Quantified results were shown as mean ± SEM.P-values were generated by two-way ANOVA with multiple comparisons.*P ＜ 0.05, **P ＜ 0.01,*** P ＜ 0.001.