Gut microbiota dysbiosis after spinal cord injury (SCI) can affect microglial activation and exacerbate secondary neuroinflammation via the gut-brain axis. Although every-other-day fasting (EODF) reshapes gut microbiota, promotes microglial polarization toward the M2 phenotype, and ameliorates neuroinflammation after SCI, whether it exerts these effects depending on gut microbiota remains unclear. Using a pseudo-germ-free (gut microbiota depletion, GMD) model and fecal microbiota transplantation (FMT), this study aims to investigate whether EODF promotes microglial M2 polarization after SCI by modulating gut microbiota, thereby alleviating secondary inflammation.

A total of 89 healthy male SPF grade SD rats (age 5 to 6 weeks, weighing 160 to 180 g) were randomly divided into groups. The GMD group (drinking antibiotic cocktail, n=6) and the ad libitum (AL) group (drinking sterilized water, n=6) were treated for 14 d before rat feces were collected for 16S rDNA high-throughput sequencing to validate the GMD model. Normal diet fecal microbiota transplantation (AL-FMT) and EODF-FMT were prepared after 4 weeks of intervention with normal diet (AL, n=12) and EODF (n=20), respectively. 45 rats were subjected to C5 right-sided hemiclamp SCI modeling on the basis of the GMD model, and then divided into a Sham group, an AL-SCI group, an EODF-SCI group, an AL-FMT-SCI group, and an EODF-FMT-SCI group, with 9 rats in each group. All of the above interventions lasted for 28 d. Gut microbiota was analyzed by 16S rDNA high-throughput sequencing. Food intake and body weight of rats in each group were recorded weekly. Upper limb motor function was evaluated using grooming test. Serum levels of brain-derived neurotrophic factor (BDNF), IL-4, and lipopolysaccharide (LPS) were measured by ELISA. The protein levels of the M1 microglial marker iNOS and the M2 microglial marker Arg-1, as well as the levels of the spinal cord tissue inflammation-related proteins IL-1β and TNF-α, were detected by Western blotting.

Compared with the AL group, both α-diversity and β-diversity were significantly decreased in the GMD group (P<0.05). At the family level, the relative abundances of Bacteroidaceae and other microbiota were markedly reduced in the GMD group (P<0.05), indicating successful establishment of the GMD model. In the SCI model, compared with the EODF-SCI and AL-SCI groups, the relative abundances of short-chain fatty acids (SCFAs)-producing microbiota were increased in the EODF-FMT-SCI group, whereas compared with the AL-FMT-SCI group, the EODF-FMT-SCI group exhibited significantly increased relative abundances of beneficial bacteria such as Bacteroidaceae and Lactobacillus gasseri. A significant difference in body weight was observed in the EODF-FMT-SCI group between 14 and 7 d postoperatively (P<0.05). In the grooming test, the scores at 7 d postoperatively were significantly higher in the EODF-FMT-SCI group than those at 3 d after surgery (P<0.05), and the scores at 7, 14 and 28 d after surgery in the EODF-FMT-SCI group were obviously higher than those in the AL-SCI group at corresponding time points (P<0.05). Compared with the EODF-SCI and AL-SCI groups, the EODF-FMT-SCI group showed significantly increased serum levels of IL-4 and BDNF, decreased LPS content, and reduced protein levels of IL-1β and TNF-α in the spinal cord tissues (P<0.05), downregulated expression of iNOS and upregulated expression of Arg-1 in the spinal cord tissues (P<0.05). Furthermore, the expression level of Arg-1 protein in the spinal cord tissue was significantly higher of the EODF-FMT-SCI group than that of the AL-FMT-SCI group (P<0.05).

EODF may alleviate secondary inflammation and promote motor functional recovery by reshaping gut microbiota and promoting microglial M2 polarization after SCI.

To explore the regulatory role and molecular mechanism of sodium butyrate（NaB）in lipopolysaccharide（LPS）induced inflammatory phenotype in BV2 microglia cells and the molecular mechanism.

BV2 cells were treated with different concentrations of NaB alone or combined with LPS. CCK8 assay was used to detect the change of cell viability and to screen the optimal concentration of the drug. Then the BV2 cells were divided into control, LPS group, low- and high-dose NaB groups（0.125 and 0.25 mmol/L）. An inflammatory model of BV2 cells were established by pretreatment with NaB for 17 h followed by LPS treatment for 24 h. Griess assay was used to measure the NO content in the supernatant, and ELISA was employed to determine the contents of IL-1β, TNF-α and IL-10 in culture medium. The mRNA levels of iNOS and CD206 were detected by RT-qPCR. Immunofluorescence staining was used to observe the expression of iNOS and CD206. The protein levels of TLR4, NF-κB p65m, IKB-α and p-IKB-α were detected by Western blotting. Cytoplasmic and nuclear proteins were isolated to detect the nuclear translocation of NF-κB p65.

Griess assay showed that the secretion of NO in the high- and low-dose NaB groups were lower than that in the LPS group（P<0.05）. ELISA results indicated that the high- and low-dose NaB groups had lower secretion levels of IL-1β and TNF-α, but high secretion level of IL-10 than the LPS model group（all P<0.05）. RT-qPCR revealed that high- and low-dose NaB treatment significantly decreased the mRNA expression of iNOS（M1 phenotypic marker）and promoted that of CD206（M2 phenotypic marker）（both P<0.05）. Immunofluorescence staining presented similar results as RT-qPCR（P<0.05）. Western blot results displayed that high- and low-dose NaB resulted in significantly inhibited protein expression of TLR4, NF-κB p65 and p-IKB-α as well as nuclear translocation of NF-κB p65 when compared with the LPS group（P<0.05）.

NaB significantly reduces LPS-induced neuroinflammatory responses in BV2 cells, which might be associated with its inhibition of TLR4/NF-κB signaling pathway and promotion of BV2 cells shifting from M1 to M2 phenotype.