C57BL/6J mice (6–8 weeks old, 20.0 ± 2 g) were supplied by Beijing Vital River Laboratory Animal Technology Co., Ltd (China, license No. SCXK (Beijing) 2016-0006). The mice were all housed in a specific pathogen-free environment with controlled temperature (22–26°C), humidity (40%−60%), and lighting (12/12-h light–dark cycle). All mice received water and standard food freely. All animal-related experiments were carried out with the evaluation and supervision of the Animal Experimentation Ethics Committee of Zhejiang Eyong Pharmaceutical Research and Development Center [certificate No. SYXK (Zhe) 2021-0033] and in compliance with the guidelines of the Institutional Animal Care and Use Committee.

All mice were allowed to acclimate to the conditions of the animal house for 7 days before the experiments. A mouse died during the adaptation phase, and subsequently, the surviving mice were randomized into five groups (n = 12/group): Group A: normal control group (NC); Group B: normal mice received FMT from HUC-MSC-treated mice group (FMT, the protocol of FMT is presented below); Group C: LPS-stimulated group (LPS); Group D: mice received antibiotics (ABX) and LPS stimulation group (ABX-LPS); Group E: mice treated with ABX, stimulated by LPS, and then received FMT from the HUC-MSC-treated mice group (ABX-LPS-FMT).

The flow of the experiment is shown in Figure 1. In brief, mice from the ABX-LPS and ABX-LPS-FMT groups received broad-spectrum antibiotics (0.1 mg/g neomycin, 0.1 mg/g ampicillin, 0.1 mg/g metronidazole, and 0.05 mg/g vancomycin) by gavage twice per day for 8 days; the step was utilized to remove the GM of the mice. The other mice were treated with an equal volume of sterile water intragastrically. Subsequently, the LPS, ABX-LPS, and ABX-LPS-FMT groups were administered with 100 μl LPS (10 mg/kg) intraperitoneally to construct the ALI model, and the remaining mice were perfused with the same amount of sterile saline (Yang et al., 2021). After 6 h, mice in the FMT and ABX-LPS-FMT groups received FMT treatment; fresh fecal samples were collected from normal mice that had received HUC-MSC treatment (in short, 6–8-week-old male C57BL/6 mice were treated with 100 μl 0.9% NaCl). After 6 h, the mice were injected intraperitoneally with 0.5 ml of phosphate-buffered saline (PBS) containing HUC-MSCs (2 × 10⁶ cells/ml, iCell Bioscience). After 3 days of HUC-MSC intervention, the fecal samples of the mice were collected (Wu et al., 2022). Then, the fecal samples were homogenized and diluted to a final concentration of 50 mg/ml using sterile saline. Afterward, the fecal samples were centrifuged, and the supernatant filtered by filters (70 μm) was applied for FMT. FMT was achieved by giving 200 μl fecal sample supernatant to the recipient mice intragastrically for four consecutive days (twice per day) (Wu et al., 2021). Correspondingly, mice in other groups were treated with an equal volume of sterile saline. The ALI model induced by LPS is deemed to be reliable and stable (Yang et al., 2021).

On the day of the last injection, fecal samples from the NC, FMT, LPS, and ABX-LPS-FMT groups were collected. Then, the mice were euthanized using 5% isoflurane, and the blood from the abdominal aorta was collected and centrifuged to obtain serum. After bronchoalveolar lavage fluid (BALF) was collected from the left lung, the right lung was harvested to evaluate the lung wet to dry (W/D) ratio and other indicators. Additionally, ileal tissues were removed from the mice and kept for further tests.

The trachea cannula of the left lung was washed with 0.7 ml of PBS repeatedly. The collected BALF was centrifuged, and the supernatant was utilized to determine BALF protein concentration and inflammatory cytokines. The cell pellets were resuspended by PBS, and then, the total cell number in BALF was counted by cell counting plate. Furthermore, the number of neutrophils in BALF was calculated by Wright–Giemsa staining.

The middle lobe of the right lung was excised and weighed immediately to obtain the wet mass of the lungs. Next, these lung tissues were dehydrated for 48 h at 80°C, to obtain the dry mass of the lungs. Finally, the lung W/D ratio was calculated to evaluate pulmonary edema.

Lung tissue samples were rinsed with PBS and then minced into pieces. The tissues were fully homogenized on ice with PBS. Then, the homogenates were centrifuged; the supernatant was obtained to measure the levels of detection of malondialdehyde (MDA), glutathione (GSH), superoxide dismutase (SOD), and myeloperoxidase (MPO) by commercial kits (A003-1; A006-2-1; A001-1; A044-1-1; Nanjing Jiancheng Bioengineering Institute, China).

The homogenates of ileal tissues were prepared as lung tissue. Then, commercial enzyme-linked immunosorbent assay (ELISA) kits were applied to test the contents of interleukin-1β (IL-1β, ml063132-1), interleukin-6 (IL-6, ml002293-1), and tumor necrosis factor-a (TNF-a, ml002095-1) in the BALF, serum, lung, and ileal tissues, following the manufacturer's instructions. All ELISA kits were bought from Shanghai Enzyme-linked Biotechnology Co., Ltd (China).

The whole blood of the mice was collected and anticoagulated with heparin sodium. Then, the blood was treated with corresponding antibodies for 0.5 h at room temperature. Specifically, Tc cells were evaluated by anti-CD3e (F2100301), anti-CD8a (F2100810), and anti-CD4 (F2100403) antibodies. DCs were measured using anti-CD11c (F21011C03) and anti-CD86 (F2100810) antibodies. NK cells were examined with anti-CD49b (F21049B03) and anti-CD3e antibodies. Tregs were tested using anti-CD4, anti-CD25 (F2102504), and anti-Foxp3 (F21FP302) antibodies. Finally, the mixture was centrifuged. After removing the supernatant, the pellet was resuspended in PBS and analyzed using flow cytometry. All antibodies used in the experiment were supplied by Liankebio (China).

The collected lung and ileal tissues were rinsed with PBS, followed by fixation and dehydration using 4% paraformaldehyde and a series of ethanol, respectively. Thereafter, the tissues were embedded in paraffin and cut into sections (5 μm thick) for staining with hematoxylin (H3136, Sigma, USA) and eosin (HE, E4009, Sigma, USA). The histopathological changes in the lung and ileal tissues were observed using a microscope (Eclipse Ci-L, Nikon, Japan).

Terminal deoxynucleotidyl transferase-mediated nick end labeling (TUNEL) assay was employed to assess the apoptosis in the lung tissues. In short, TUNEL staining was performed on the processed lung tissue sections by an in situ cell death detection kit (11684795910, Roche, France), as per the supplier's instructions. Finally, the slices were counterstained with 4′,6-diamidino-2-phenylindole for 10 min in the darkness to identify the nucleus. After completing staining, the non-apoptotic cell appeared blue, while the apoptotic cell was green. The slices were viewed using the microscope, and the apoptosis of the lung tissues was exhibited as the amount of TUNEL-positive cells.

The total RNA was isolated from lung tissue homogenate by TRIzol reagents. Then, the total RNA was used to synthesize cDNA using reverse transcription kits (CW2569, CWBIO). Thereafter, quantitative PCR (qPCR) was conducted by SYBR Green qPCR kits (11201ES08, YE SEM, China) and carried out on real-time PCR instruments (LightCycler 96, Roche, Germany). The relative mRNA expressions were calculated as the ratio of GAPDH and analyzed by the 2^−ΔΔCt method. The qPCR primer sequences designed in the experiment are presented in Table 1.

After fixation with paraformaldehyde, the lung and ileal tissues were embedded and cut as usual. Immunohistochemical staining was, then, performed to detect the expressions of TLR4, ZO-1, claudin-1, and occludin. In brief, the sections were deparaffinized and rehydrated prior to quenching in H₂O₂ (3%). Upon placing in 5% bovine serum albumin to block non-specific binding, the lung samples were probed with anti-TLR4 antibody (1:100, AF7017, Affinity, China), and the ileal samples were probed with anti-ZO-1 (1:500, ab221547, Abcam, UK), anti-claudin-1 (1:250, ab211737, Abcam, China), and anti-occludin (1:200, ab216327, Abcam, UK). The next day, the slices were washed and then reacted with secondary antibodies, stained with diaminobenzidine and hematoxylin, and finally examined under a microscope.

The total protein was isolated from lung and ileal tissue homogenate by Radioimmunoprecipitation assay buffer. Then, bicinchoninic acid kits were applied to detect protein concentration. Afterward, protein extracts were separated by SDS polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes. After shaking softly in blocking solution, the membranes were incubated with primary antibodies overnight at 4°C. The following day, the membranes were washed and incubated again with secondary antibodies. After the blots were visualized using enhanced chemiluminescence, data were analyzed by ImageJ software. All information on the primary antibodies is presented in Table 2.

DNA extraction and 16S rRNA gene sequencing of fecal samples were conducted with the help of BioNovoGene Co., Ltd. (China). In short, the total microbial DNA was extracted from the feces and then quantified by Nanodrop. Agarose gel electrophoresis (1.2%) was utilized to evaluate the quality of extracted DNA. After completing 16S rRNA amplification by PCR, purification, and recovery by magnetic beads, the products were subjected to fluorescence quantitative analysis. Subsequently, the TruSeq Nano DNA LT Library Prep Kit provided by Illumina was employed to construct sequencing libraries.

Raw sequence reads were trimmed and filtered into effective reads, which were further clustered into operational taxonomic units (OTUs) with a similarity >97% by UPARSE. The taxonomy of OTU was carried out by the SILVA database. In addition, the α-diversity of microbiota was analyzed by Shannon and Simpson indices, and the β-diversity of microbiota was assessed by weighted UniFrac distance and visualized with non-metric multidimensional scaling (NMDS) plot and analysis of similarities (ANOSIM) test. Additionally, linear discriminant analysis effect size was applied to taxonomic discovery analysis for microbial biomarker discovery.

The data were analyzed by SPSS 19.0 and presented as mean ± SD. Differences between multiple groups were compared by analysis of variance (ANOVA) and Tukey's tests. The Kruskal–Wallis H-test was applied when variances were not homogeneous. p < 0.05 was considered statistically significant. R software was used to perform amplicon sequencing statistical analysis. Furthermore, the Wilcoxon test was utilized to calculate the microbial α-diversity index, while the ANOSIM non-parametric test was conducted to compare microbial β-diversity between the two groups.

As shown in Figure 2A, LPS treatment leads to an obvious upregulation of the concentrations of total protein, the counts of total cells, and neutrophils in BALF of ALI rats (p < 0.01). Consistent with this, after mice were exposed to LPS, the lung W/D ratio increased clearly (p < 0.01), and the lung exhibited serious injury (Figures 2B, C); specifically, the remarkable thickening of the alveolar septum and wall was observed. In addition, from the results of HE staining, we also observed pulmonary capillary edema and congestion in LPS mice. When LPS mice were pre-treated with ABX, all the phenomena described were more pronounced (p < 0.01). However, FMT treatment effectively reversed the increase of total protein, total cells, and neutrophils in BALF, reduced the W/D ratio in the lung, and improved lung injury induced by LPS and ABX; the alveolar septa were thinning, and inflammatory cell infiltration was alleviated (p < 0.01).

Given that innate immune cells are crucial in ALI development, the percentage of immune cells was detected in the study. Flow cytometry results revealed that LPS exposure caused a significant increase in the percentages of CD4⁺ Th, CD8a⁺ Tc, CD86⁺ DC, and CD3e⁺ NK cells but blocked the percentages of CD25⁺ Tregs and CD4⁺ Tregs cells (p < 0.05, Figure 3). Interestingly, ABX pre-treatment worsened this situation (p < 0.05). As expected, it can be observed from Figure 3 that FMT treatment moderately attenuated ABX- and LPS-stimulated immune cell changes (p < 0.05).

The percentage of apoptotic cells was raised remarkably in the lung tissues of LPS-exposed mice (p < 0.01, Figure 4A). Accordingly, ABX pre-treatment deteriorated the apoptosis of the lung tissues in LPS mice (p < 0.01). Nevertheless, FMT treatment effectively counteracted the upregulation of apoptosis in the lung tissues after ABX and LPS stimulation (p < 0.01). The apoptosis-related proteins in the lung tissues were utilized to verify the extent of apoptosis. Unanimously, we found that LPS and ABX increase the Bax, caspase-3, and cleaved-caspase-3 protein expressions and decrease the Bcl-2 protein expression (p < 0.05, Figure 4B); however, those changes were offset by the FMT treatment (p < 0.01).

Next, the effect of FMT from HUC-MSC-treated mice on the inhibition of oxidative stress and inflammation response was determined. As shown in Figure 5A, the lung tissues of LPS-stimulated mice had higher MDA and MPO levels and lower GSH and SOD levels than normal control mice (p < 0.05). In addition, ELISA results also revealed that the LPS challenge led to serious inflammation in the BALF, serum, lung, and ileal tissues (Figure 5B). ABX pre-treatment further worsened those changes. However, ABX- and LPS-induced increases of MDA, MPO, IL-1β, IL-6, and TNF-α and reduction of GSH and SOD were effectively restored by FMT treatment (p < 0.01). Then, qPCR was conducted to further assess the antioxidant and anti-inflammatory effects of FMT treatment. It can be observed from Figure 5C that IL-1β, IL-6, TNF-α, iNOS, and Cox-2 mRNA expressions were significantly raised in the LPS and ABX-LPS groups (p < 0.05), but upon FMT treatment, the upregulation of these mRNA was effectively relieved (p < 0.05).

It is well-known that injured ileal tissues and impaired barrier function contribute to the development of ALI (Xu Y. et al., 2021). Thus, we checked whether FMT from HUC-MSC-treated mice could alleviate the injured degree, inhibit apoptosis, and improve the impaired barrier function for the ileal tissues. It can be observed that after stimulation by LPS, the ileal tissues appeared with obvious injury (Figure 6A). Accordingly, the situation was more serious in the ABX-LPS group. As expected, FMT treatment attenuated ABX- and LPS-induced ileal tissue injury. In addition, following LPS stimulation, Bax protein expression was increased, while Bcl-2 protein expression was decreased (p < 0.01), the situation further deteriorated in the ABX + LPS group (p < 0.01, Figure 6B). However, FMT from HUC-MSC-treated mice decreased Bax protein expression but increased Bcl-2 protein expression (p < 0.01). The expressions of ZO-1, claudin-1, and occludin are crucial for maintaining barrier function (Shi et al., 2020). The results of immunohistochemical staining revealed that ZO-1, claudin-1, and occludin protein expressions were obviously decreased upon ABX and LPS stimulation (p < 0.05, Figure 6C). Of note, FMT treatment elevated the decrease in ZO-1, claudin-1, and occludin protein expressions (p < 0.05). As expected, the Western blot results exhibited a similar trend as the results of the immunohistochemical staining (Figure 6D).

LPS has been found to play pro-oxidative and pro-inflammatory roles by the TLR4/NF-κB and Nrf2/HO-1 pathways (Xin et al., 2020). Therefore, the impact of FMT from HUC-MSC-treated mice on the TLR4/NF-κB and Nrf2/HO-1 pathways was tested. Immunohistochemical staining and Western blot analysis demonstrated that ABX and LPS treatments raised the protein expressions of TLR4, Myd88, COX-2, and iNOS, as well as the phosphorylation of NF-κB and IκBα in the lung tissues of mice, but decreased the protein expressions of Nrf2 and HO-1 (p < 0.05, Figure 7). Notably, FMT treatment restored the expressions and phosphorylation of TLR4/NF-κB and Nrf2/HO-1 pathway-related proteins in the lung tissues from ABX- and LPS-treated mice (p < 0.05). The results obtained from the ileal tissues were similar to those from the lung tissues (Figure 8). The mechanistic diagram is presented in Figure 9.

Further sequencing of the 16S rRNA gene was applied to detect the microbial profiles of the fecal samples from mice in the NC, FMT, LPS, and ABX-LPS + FMT groups. The top 20 bacterial genera are presented in Figure 10A in terms of the relative abundance. In addition, microbial α-diversity analyzed by Shannon and Simpson indices revealed that the community diversity of microbe in the LPS group was higher than those of the NC and FMT groups, despite the difference was not significant (Figure 10B). However, the community diversity of microbe from ABX-LPS + FMT mice was effectively decreased (p < 0.01). Additionally, microbial β-diversity was also analyzed by NMDS and ANOSIM, and the results revealed that the microbial community structure of ABX-LPS + FMT mice changed obviously (Figures 10C, D). Then, we further screen the potential GM change by FMT treatment in LPS mice. LEfSe analysis was conducted to detect the bacterial taxa that differed significantly among the groups (Figures 10E, F). Specifically, at the genus level, we found that relative to the NC group, the relative abundances of Bacteroides, Christensenella, Coprococcus, and Roseburia were significantly elevated in the LPS group. However, the relative abundances of Bacteroides, Christensenella, Coprococcus, and Roseburia were decreased obviously in the ABX-LPS + FMT group. Furthermore, the relative abundances of Xenorhabdus, Sutterella, and Acinetobacter obviously increased in the ABX-LPS + FMT group (Figure 10G).