The 2-week-old newborn rabbits (Oryctolagus cuniculus) were used to imitate NEC as the previous with a little change (Choi et al., 2010; Bozeman et al., 2013). Briefly, the newborn rabbits were fed with mother’s milk for 3 days before inducing NEC. The healthy controls were fed with mother’s milk all the time. The NEC rabbits were fed with homemade hypertonic formula milk (10 g protein powder dissolved in 75 ml Esbilac, 15 ml/kg/per, 3 times/day) and stimulated with hypothermia (10 min/4°C, 2 times/day for 3 days) and hypoxia (95% nitrogen and 5% oxygen. 10 L/min, 10 min/2 times/day for 3 days). The ingredients of sulperazone (cefoperazone sodium and sulbactam sodium for injection) are cefoperazone and sulbactam with a ratio of 2:1. Sulperazone was administrated to the rabbits by intravenous drip of 30 mg/kg/per 12 h. The rabbits of the groups reached 50% lethality as the end point of the experiment. The animal health condition was recorded for 2 weeks after model construction for 3 days. The animal study was reviewed and approved by the Ethics Committee of Southern Medical University (No. 2019R001-F05).

The 2-week-old newborn suckling rabbits were selected and divided into five groups randomly, 15 rabbits for each group [Group A: healthy rabbits taken at the end of the experiment; Group B: model control group (with jugular vein venous catheter for nutrition supply); Group C: conventional treatment group: after model building successfully treated with sulperazone; Group D: FMT group (fecal bacteria transplantation treatment group): after model building successfully, healthy rabbits, they were gavaged with feces from healthy rabbits; Group E: conventional treatment and FMT group: after model building successfully, the rabbits were treated with sulperazone and FMT.

The feces of healthy baby rabbits was mixed with physiological saline, centrifuged, filtered, and the stomach was gavaged. Rabeprazole was injected before gavage to inhibit gastric acid secretion and reduce the destruction of bacterial flora in the stomach. During the experiment, fluids and antibiotics were given through the central vein.

The intestinal tube was separated, and 2–3 cm of the ileocecum tissue was retrieved. After H&E staining, the ileocecum pathology score was performed. The pathological severity assessment was performed independently by two pathologists under double-blind conditions. The final pathology score calculation was half of the sum of the two scores. The scoring standard refers to the NADLER scoring standard: 0 points, intestinal villi and epithelium are intact, and the tissue structure is normal; 1 point, mild submucosal and/or lamina propria swelling and separation; 2 points, moderate submucosal and/or lamina propria swelling and separation, submucosal and/or muscle layer edema; 3 points, severe submucosal and/or lamina propria swelling and separation, submucosal and/or muscle layer edema, local villi loss; 4 points, intestinal villi disappearance with intestinal necrosis. The final pathology score ≥ 2 points is determined to be NEC.

The V3–4 hypervariable region of the bacterial 16S rRNA gene was amplified with the primers 338F (5′-ACTCCTACGGG AGGCAGCAG-3′) and 806R (5′-GGACTACNNGGGTATCT AAT-3′). For each feces sample, an eight-digit barcode sequence was added to the 5′ end of the forward and reverse primers (provided by Allwegene Company, Beijing). The PCR was carried out on a Mastercycler Gradient (Eppendorf, Germany) using 25-μl reaction volumes, containing 12.5 μl 2 × Taq PCR MasterMix, 3 μl of BSA (2 ng/μl), 1 μl of forward primer (5 μM), 1 μl of reverse primer (5 μM), 2 μl of template DNA, and 5.5 μl of ddH₂O. Cycling parameters were 95°C for 5 min, followed by 28 cycles of 95°C for 45 s, 55°C for 50 s, and 72°C for 45 s with a final extension at 72°C for 10min. The PCR products were purified using an Agencourt AMPure XP Kit.

Deep sequencing was performed on Miseq PE300 platform at Allwegene Company (Beijing). After the run, image analysis, base calling, and error estimation were performed using Illumina Analysis Pipeline Version 2.6.

The raw data were first screened, and sequences were removed from consideration if they were shorter than 230 bp, had a low-quality score (≤20), contained ambiguous bases or did not exactly match the primer sequences and barcode tags, and separated using the sample-specific barcode sequences. Qualified reads were clustered into operational taxonomic units (OTUs) at a similarity level of 97% using the Uparse algorithm of Vsearch (v2.7.1) software. The Ribosomal Database Project (RDP) Classifier tool was used to classify all sequences into different taxonomic groups against the SILVA128 database.

QIIME (v1.8.0) was used to calculate the richness and diversity indices based on the OTU information.

A total amount of 1.5 μg of RNA per sample was used as input material for the RNA sample preparations. Sequencing libraries were generated using NEBNext^® Ultra^TM RNA Library Prep Kit for Illumina^® (NEB, United States) following the recommendations of the manufacturer, and index codes were added to attribute sequences to each sample. Briefly, mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. Fragmentation was carried out using divalent cations under elevated temperature in NEBNext First Strand Synthesis Reaction Buffer (5X). First-strand cDNA was synthesized using random hexamer primer and M-MuLV Reverse Transcriptase (RNase H). Second-strand cDNA synthesis was subsequently performed using DNA Polymerase I and RNase H. The remaining overhangs were converted into blunt ends via exonuclease/polymerase activities. After adenylation of the 3′ ends of DNA fragments, the NEBNext Adaptor with hairpin loop structure was ligated to prepare for hybridization. In order to select cDNA fragments of preferentially 200–250 bp in length, the library fragments were purified with AMPure XP system (Beckman Coulter, Beverly, United States). Then 3 μl of USER Enzyme (NEB, United States) was used with size-selected, adaptor-ligated cDNA at 37°C for 15 min followed by 5 min at 95°C before PCR. Then PCR was performed with Phusion High-Fidelity DNA polymerase, Universal PCR primers, and Index (X) Primer. At last, PCR products were purified (AMPure XP system), and library quality was assessed on the Agilent Bioanalyzer 2100 system. The library preparations were sequenced on an Illumina Hiseq 4000 platform by the Beijing Allwegene Technology Company Limited (Beijing, China), and paired-end 150-bp reads were generated.

Gene Ontology (GO) enrichment analysis of the differentially expressed genes (DEGs) was implemented by the GOseq R packages based on the Wallenius non-central hyper-geometric distribution, which can adjust for gene length bias in DEGs.

Gut microbiota structure chaos is one of the leading causes of NEC. Recently, FMT is an emerging strategy to treat digestive tract disease caused by microbiome disorders. To verify whether FMT is beneficial for NEC newborn rabbits in our experiment, we constructed NEC models as described in “Materials and Methods” section, and NEC rabbits were treated with sulperazone, FMT, and its combination (Figure 1A). As shown in the figure, the NEC rabbits showed typical NEC pathology, such as intestinal mucosa destruction, villi shedding, inflammatory cell infiltration, etc. Two pathologists evaluated the sections independently, and the NEC model showed ≥ 2 score, which indicted the model had been constructed successfully. After different treatments, the pathology of NEC showed relief on different levels. However, the combined treatment of FMT and sulperazone could efficiently reverse the NEC symptom and prolong lives (Figure 1B).

It is worth noting that the sulperazone and FMT combination shows the most efficiency (Figures 2A,B), which indicates that this therapy strategy may be a better choice for NEC treatment. We performed 16s RNA-seq to investigate how the gut microbiome changes. First, we checked the diversities of the different groups by four diversity indexes (Figure 2A). The microbiome diversity was lowest in the NEC group compared with the other groups. The microbiome diversity was significantly increased after treatments, and the combination of the sulperazone and FMT group showed the highest efficiency and shared the most similarity with the healthy group (Figures 2B,C). This result suggests that microbiome homeostasis and diversity is very important for NEC treatment, and diversity recovery is helpful for NEC treatment (Guo et al., 2021).

Then, we checked the microbiome changes on a genius level. In healthy control, Bacteroides, Akkermansia, Ruminococcaceae NK4A214 group, Christensenellaceae R-7 group, Lachnospiraceae NK4A136 group, and Ruminococcus 1 were the top 10 abundant genera (Figure 2D). All the above microorganisms were probiotics, involved in food degradation, metabolism regulation, and inhibiting the pathogenic bacteria proliferation. Ruminococcus 1, Marvinbryantia, Parabacteroides, Kurthia, Flavonifractor, Lactonifactor, and Lachnospiaceae UCG 010 were the signatures of NEC (Figure 2D; Esber et al., 2020; Hiippala et al., 2020; Lordan et al., 2020). Parabacteroides was reported to be upregulated in NEC (Figure 2D). The others were defined as conditional pathogenic bacteria.

Next, we compared the top five abundant microbes among different groups. Christensenellaceae R-7 group and Akkermansia were not recovered after all treatments, which may indicate that both of them were not the key microorganisms for treatment used in our experiments. Enterococcus and Ruminococcaceae NK4A214 groups were significantly increased after all treatments. The Vagococcus and Myroides were the highest in the NEC group. Moreover, they did not exist in healthy control and were downregulated by all treatments, especially by sulperazone and FMT combination, and they were discovered first in our study. The hazard of Vagococcus was still not clarified. However, it is clear that Vagococcus is associated with purulent infection (Altintas et al., 2020). Myroides has strong resistance to bactericides such as antibiotics and leads to infection (LaVergne et al., 2019). Corynebacterium 1 and Aerococcus were the unique microorganisms of sulperazone treatment. A lot of studies reported that these two Gram-positive bacteria are conditional pathogenic bacteria (Rasmussen, 2016; Oliveira et al., 2017), and their benefits remain to be revealed. Escherichia–Shigella was a unique genus after FMT treatment. The abundance of Escherichia–Shigella was negatively correlated with NEC progress (Zhou et al., 2015). One of the mechanisms of FMT treatment to NEC was recovering Escherichia–Shigella. Clostridium sensu stricto 1 and Subdoligranulum were unique genera in sulperazone and FMT combination group. Clostridium sensu stricto 1 was helpful in promoting gut development and maintaining the gut microbiome diversity (Du et al., 2018). Subdoligranulum was one of the producers of butyrate, which is essential for gut health (Chassard et al., 2014). However, it could be detected in maternal feces and breast milk but does not exist in the gut of NEC babies. The combination treatment could fill the deficiency.

Interestingly, Enterococcus has a low proportion in NEC and healthy control but increased significantly in all treatment groups (Figures 2C,D). Enterococcus was classified as an infection source for inpatients and sometimes could be life threatening, but Enterococcus faecalis, one species of Enterococcus, was a common early colonizer in newborn gut and is associated with relieving NEC (Gao et al., 2018; Delaplain et al., 2019). Furthermore, a recent study showed that Enterococcus carries bacteriophage and interacts with human tumor antigen to enhance the T-cell immune response (Delaplain et al., 2019). Whether a similar function of Enterococcus, especially the faecalis species exists in NEC needs to be further investigated.

To investigate gene expression change, we performed transcriptome analysis. The transcript profile of sulperazone and FMT combination treatment showed the highest similarity with the healthy group and the sulperazone treatment showed the highest similarity with the NEC group, which reminds that the combination of sulperazone and FMT may show a better benefit for NEC (Figure 3A). Then, we compared the different groups to define the further changes. We found 3,330 upregulated genes and 2,332 downregulated genes in the NEC group (Figure 3B). These genes were enriched in alginic acid metabolism, thermogenesis, oxidative phosphorylation, ribosome, and viral process, suggesting that metabolism and microbiome dysregulation were the most significant molecular changes in NEC (Figure 4A). The 3,302 upregulated genes and 2,371 downregulated genes were defined after sulperazone treatment (Figure 3C). Function enrichment analysis showed that differentially expressed genes were involved in proteolysis, tryptophan metabolism, and indolakylamine metabolism, which showed that the sulperazone may help to inhibit the necrosis (Figure 4B; Cussotto et al., 2020; Yusufu et al., 2021). Interestingly, sulperazone may inhibit the inflammation because the downregulation of tryptophan and indolakylamine could be the signature of downregulation of inflammation. FMT treatment leads to 5,057 genes changed on the expression level (2,858 upregulated and 2,199 downregulated), and they were enriched in cytokine production, proteolysis, lymphocytes and monocytes behaviors, and thermogenesis of GO (Figures 3D, 4C). Finally, we checked the pathway changes of the combination of FMT and sulperazone. Response to lipid, especially cholesterol, was the most significant downregulated pathway in metabolism, and the pathways of cell locomotion were broadly downregulated (Figure 4D).