A total of 76 newborn male and female C57BL/6 mice (8 weeks old, weight 2–4 g) were obtained from the Experimental Animal Center of Shanghai Jiao Tong University Affiliated Sixth People's Hospital. Intestinal epithelial specific SIRT1 knockout (KO) newborn mice (SIRT1^KO, villin-cre⁺, SIRT1^flox/flox) and their littermate Flox controls (villin-cre⁻, SIRT1^flox/flox) on a C57BL/6 background were generated as previously described (20). The mice were socially housed in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication, 8th Edition, 2011). All animal experiments were approved by the Ethics Committee on Animal Care of Shanghai Jiao Tong University Affiliated Sixth People's Hospital.

The neonatal mice were placed into an incubator with a constant temperature of 36°C and a humidity of 55% on the day of birth, and were artificially fed using a formula previously described (Table I) (21). The neonatal mice were intubated with a sterile 1.9F silicone tube (BD Biosciences) every 4 h. The skin around the mouth of the neonatal mice was cleaned with 0.9% sodium chloride and 75% ethanol solution before feeding. The first feeding volume was 0.1 ml, and 0.1–0.3 ml was added every 24 h.

Mice with NEC were established as previously described (22). Briefly, neonatal mice were fed with lipopolysaccharide (LPS; Sigma-Aldrich; Merck KGaA) at 2 mg/kg in addition to artificial feeding for 3 days. Furthermore, neonatal mice were exposed to hypoxia (99% N₂) for 1 min in a hypoxic chamber, followed by hypothermia (4°C) for 10 min twice daily for an additional 3 days (23). The NEC model was successfully established when the animals appeared to be severely ill with signs of cyanosis, lethargy and/or abdominal distension, which was similar to the human NEC in stage II and III. The mice continued to be artificially fed for 3 days and then sacrificed by cervical dislocation. During this period, if the mouse exhibited more serious symptoms, such as severe abdominal distention, disappearance of bowel sounds, serious decrease of heart rate, hypothermia, and/or shock, it was sacrificed by cervical dislocation. After establishing the NEC model, the newborn mice were orally administered with SB sachets (>1.3×10⁹ CFU/g; Laboratoires Biocodex) once a day at 800 mg/kg/day for 3 days. After the gavage, the neonatal mice were returned to the incubator, and the mice with NEC were sacrificed after 3 day SB treatment. Notably, our previous study indicated that 800 mg/kg SB significantly increased the body weight and survival cycle, and decreased the pathological score of neonatal mice with NEC compared with other SB doses (unpublished data). Therefore, 800 mg/kg SB was used in the present study.

Mice from the normal and control groups were sacrificed 6 days after breastfeeding or artificial feeding. Mice with NEC were sacrificed 6 days after the induction of NEC, while NEC + SB mice were sacrificed 3 days after SB treatment. These samples were all acquired at the same time among the experimental groups. After euthanasia, mice were incised from the median region, and the tissues and blood vessels were separated. Next, the ileocecum (~5 cm) was obtained, fixed in 4% paraformaldehyde for 24 h at room temperature, and embedded in paraffin. In addition, a section of ileocecal intestinal tissue was stored for subsequent use in immunofluorescence staining, western blotting and reverse transcription-quantitative PCR (RT-qPCR).

The embedded ileocecal intestinal tissue was sliced at a thickness of 4 µm, and stained with an H&E Staining kit (Beyotime Institute of Biotechnology) following the manufacturer's instructions. The diagnosis of NEC was performed based on the assessment of two independent pathologists, who were blinded in this study. NEC was scored based on the severity of the damage from the mucosa to the lamina propria as follows (24): i) 0, intact villi and epithelium with normal tissue structure; ii) 1, mild separation or edema in the submucosa and/or lamina propria; iii) 2, moderate separation or edema in the submucosa and/or lamina propria; iv) 3, severe separation or edema in the submucosa and/or lamina propria, and villi shedding; and v) 4, complete absence of epithelial structures and transmural necrosis. A pathological score ≥2 was classified as NEC.

The ileocecal tissue sections were fixed for 15 min with 4% paraformaldehyde at room temperature, blocked with 5% BSA (Beyotime Institute of Biotechnology) at room temperature for 30 min, and incubated with primary antibodies (SIRT1; 1:1,000; cat. no. ab220807; Abcam) at 4°C overnight. Following which, tissues were incubated at room temperature for 1 h with donkey anti-mouse lgG Alexa Fluor 350-conjugated secondary antibody (cat. no. A10035) and donkey anti-mouse lgG Alexa Fluor 594-conjugated secondary antibody (cat. no. A32744; 1:1,000; both purchased from Thermo Fisher Scientific, Inc.). Then, the sections were stained with DAPI (Sigma-Aldrich; Merck KGaA) at room temperature for 10 min. After final washing, the coverslips were mounted on slides using 50% glycerin. The sections were then observed using a fluorescence microscope (Olympus Corporation).

Cytoplasmic and nuclear proteins were extracted from ileocecal tissues using a Nuclear and Cytoplasmic Protein Extraction kit (Beyotime Institute of Biotechnology) according to the manufacturer's instructions. Protein concentrations were measured via bicinchoninic acid assay (Beyotime Institute of Biotechnology). Proteins (40 µg) were loaded onto each lane and separated by 8–12% SDS-PAGE and electro-transferred to nitrocellulose membranes (EMD Millipore). Then, the membranes were probed with antibodies against SIRT1 (cat. no. ab220807), NF-κB (cat. no. ab16502), β-actin (cat. no. ab8226) or lamin B1 (cat. no. ab16048; 1:1,000; all purchased from Abcam) at 4°C overnight, followed by incubation with a LICOR IRDye^® 800CW goat anti-mouse lgG secondary antibody (cat. no. 926-32210) and IRDye 680RD goat anti-rabbit IgG (cat. no. 926-32221; 1:10,000; both purchased from LI-COR Biosciences) at room temperature for 2 h. Finally, the signals were detected by Odyssey Infrared Imaging System (LI-COR Biosciences). Digitized images were determined by ImageJ software (v1.47; National Institutes of Health). Protein expression levels were calculated from the ratio of corresponding protein to β-actin (cytoplasmic loading control) or lamin B1 (nuclear loading control) multiplied by 100%.

Total RNA from ileocecal tissues was extracted with TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's protocol. The RNA was then subjected to RT to synthesize cDNA using the PrimeScript RT Reagent kit with gDNA Eraser (Takara Biotechnology Co., Ltd.) at 42°C for 15 min and 85°C for 5 min. Subsequently, 20-µl reactions with SIRT1, NF-κB or β-actin primers (Genewiz, Inc.) were detected using a PikoReal 96 Real-Time PCR system (Thermo Fisher Scientific, Inc.) with SYBR Green PCR Master mix (Applied Biosystems; Thermo Fisher Scientific, Inc.). Relative quantitative analysis in the mRNA expression was obtained using the 2^−ΔΔCq method (25) and normalized to β-actin.

Total DNA from the feces and intestinal contents was extracted using the QIAamp DNA Stool Mini kit (Qiagen GmbH). The abundance of gut microbial composition was assessed by qPCR using specific 16S ribosomal DNA (rDNA) primers (Genewiz, Inc.). The data were quantified by calculating the abundance of bacterial group-specific 16S rRNA genes and normalized to total bacterial rRNA genes.

The sequences of the primers were as follows: SIRT1 forward, 5′-CAGCTCTGCTACAATTCATCGCGTC-3′ and reverse, 5′-AATCTCTGTAGAGTCCAGCGCGTGTG-3′; NF-κB forward, 5′-ACGAGCAGATGGTCAAGGAG-3′ and reverse, 5′-CTTCCATGGTCAGTGCCTTT-3′; TNF-α forward, 5′-TACACCTCACCCACACAGTC-3′ and reverse, 5′-CGCACCTCACAGACTGTTTT-3′; IL-6 forward, 5′-GCCAGTTGCCTTCTTGGG-3′ and reverse, 5′-CGACTTGTGAAGTGGTATA-3′; β-actin forward, 5′-TAAAGACCTCTATGCCAACACAGT-3′ and reverse, 5′-CACGATGGAGGGCCGGACTCATC-3′; Eubacterium rectale/Clostridium coccoides (Erec) forward, 5′-ACTCCTACGGGAGGCAGC-3′ and reverse, 5′-GCTTCTTAGTCAGGTACCGTCAT-3′; Clostridium leptum (Clept) forward, 5′-GTTGACAAAACGGAGGAAGG-3′ and reverse, 5′-GACGGGCGGTGTGTACAA-3′; Lactobacillus sp. (Lact) forward, 5′-GACGGGCGGTGTGTACAA-3′ and reverse, 5′-CACCGCTACACATGGAG-3′; Bacteroides sp. (Bact) forward, 5′-CACCGCTACACATGGAG-3′ and reverse, 5′-GCTGCCTCCCGTAGGAGT-3′; mouse intestinal Bacteroides (MIB) forward, 5′-CCAGCAGCCGCGGTAATA-3′ and reverse, 5′-CCAGCAGCCGCGGTAATA-3′; segmented filamentous bacteria (SFB) forward, 5′-GACGCTGAGGCATGAGAGCAT-3′ and reverse, 5′-GACGCTGAGGCATGAGAGCAT-3′; and total bacteria forward, 5′-ACTCCTACGGGAGGCAGCAGT-3′ and reverse, 5′-ATTACCGCGGCTGCTGGC-3′.

Data were expressed as the means ± standard error of mean. Statistical analysis was performed by SPSS 23.0 (IBM Corp.). Significant differences between two groups were analyzed via Student's t-test, multiple comparisons of >two groups were performed using Tukey's honest significant difference test, and P<0.05 was considered to indicate a statistically significant difference.

Based on macroscopic observation of the intestinal specimens, the intestine of normal neonatal mice was bright in color, straight and smooth, without gas accumulation. The color of the intestine in the control (artificial feeding) group was slightly darkened, and the ileum was partially dilated. In the NEC group, the intestine was severely dilated, blackened and congested. In addition, the NEC mice had severe intestinal gas accumulation. The intestine of the NEC + SB neonatal mice was visually relieved, only showing mild edema and only slightly darkened in color, which was significantly improved compared with the NEC group. H&E staining revealed that the ileocecal intestinal tissue of the normal group was intact with continuous epithelium, regular glands and neat villi, and the mucosa, submucosa and lamina propria were free of congestion and edema (Fig. 1A). The pathological score was low in the normal group (Fig. 1B). The ileocecal mucosa, submucosa and lamina propria of the control group exhibited mild congestion and edema, and the glands were disorderly arranged, resulting in an increase in pathological score compared with the normal group (P<0.05). Furthermore, the ileocecum in the NEC group had obvious villi degeneration and edema, disordered gland arrangement, severe congestion, and edema of mucosa, submucosa and lamina propria. The pathological score was significantly higher than that of the control group (P<0.01). The NEC + SB group had a mild edema in the mucosa, submucosa and lamina propria, and the villi were uneven, although there were some arrangements, and the glands were arranged regularly. The pathological score in the NEC + SB group was significantly decreased compared with the NEC group (P<0.01).

The results from immunofluorescence staining indicated that a relatively high level of SIRT1 protein expression was primarily located in the intestinal epithelial cells of the normal group (Fig. 2). Artificial feeding (control) could lead to a decrease in the expression of SIRT1 in the ileocecum of neonatal mice, and there was a further decrease in the level of SIRT1 in the NEC group compared with the control group. Notably, NEC mice fed with SB could reverse the decrease in SIRT1 and markedly enhance its expression level. These results revealed that SB treatment enhanced SIRT1 protein expression in NEC neonatal mice.

To determine whether the SIRT1/NF-κB pathway is involved in the role of SB in NEC mice, western blotting was used to analyze the protein expression of SIRT1 and NF-κB. As revealed in Fig. 3A and B, the expression of cytoplasmic SIRT1 and nuclear NF-κB proteins in the control group was slightly decreased and increased, respectively, compared with the normal group (P<0.05). NEC insult led to a further decrease and increase of SIRT1 and NF-κB (P<0.01). SB treatment in NEC mice reversed the decrease in SIRT1 and the increase in NF-κB, and significantly promoted SIRT1 expression and reduced NF-κB expression (P<0.01).

Next, the effect of SB on SIRT1, NF-κB, TNF-α and IL-6 mRNA expression in NEC mice was investigated. Similar to the changes in SIRT1 and NF-κB at the protein level, a decrease in SIRT1 and an increase in NF-κB mRNA was observed in the control group compared with the normal group (P<0.05; Fig. 4A and B). NEC insult induced a significant decrease in SIRT1 mRNA and an increase in NF-κB mRNA levels compared with the control group (P<0.01). In addition, the mRNA levels of TNF-α and IL-6 in the NEC group were significantly increased compared with the control group (P<0.01; Fig. 4C and D). SIRT1 mRNA expression was significantly upregulated, and the mRNA levels of the members of the NF-κB pathway were all downregulated in the NEC + SB group compared with the NEC group (P<0.01).

To further testify the possible function of the SIRT1 pathway in NEC neonatal mice, an intestinal epithelium-specific SIRT1^KO mouse model was generated, in which the exon 4 of the mouse SIRT1 gene was deleted throughout the length of the intestinal epithelium (20). There was no obvious intestinal damage in SIRT1^KO mice with breastfeeding or artificial feeding (Fig. 5A). The ileocecal lesions caused by NEC in SIRT1^KO neonatal mice were significantly increased compared with the control (co-housed paired Flox) group, and SB treatment did not reverse this pathological damage. Furthermore, the pathological score of SIRT1^KO with artificial feeding was increased compared with SIRT1^KO with breastfeeding (P<0.05). The score of SIRT1^KO in the NEC + SB group was also significantly increased compared with the control group (P<0.01). In addition, both the SIRT1 protein (Fig. 5B and C) and mRNA (Fig. 5D) expression in the SIRT1^KO NEC + SB group were significantly decreased compared with the control group (P<0.01). Unexpectedly, the NF-κB protein (Fig. 5B and C) and mRNA (Fig. 5E) levels, and the mRNA levels of its downstream partners TNF-α and IL-6 (Fig. 5F and G) were all significantly increased in the SIRT1^KO group compared with the control group (P<0.01). All these results indicated that knockdown of SIRT1 suppressed the effect of SB in NEC mice.

To evaluate whether SB-induced alterations in NEC mice is associated with the regulation of the gut microbial ecosystem, total fecal microbiota profiles of neonatal mice were analyzed via 16S rRNA amplicon sequencing. Feces were collected, and the proportions of microbiota were evaluated by qPCR. As revealed in Fig. 6A, the abundance of all gut microbial composition except Lact, including Erec, Clept, Bact, MIB and SFB were decreased in the control group compared with the normal group (P<0.05). Furthermore, NEC insult induced a significant decrease in gut microbial abundance compared with the control group (P<0.01 in Erec, Clept and Lact; P<0.05 in Bact, MIB and SFB). SB treatment significantly upregulated the abundance of gut microbiota in NEC mice compared with the NEC group (P<0.01 in Erec, Clept, Lact and Bact; P<0.05 in MIB and SFB). Notably, the gut microbial abundance in the SIRT1^KO NEC + SB group was significantly decreased compared with the Flox NEC + SB group (P<0.01; P<0.05 in Lactl; Fig. 6B). These results revealed that the modulation of gut microbiota was also involved in the role of SB in NEC neonatal mice.