The Time-Dependent Dynamics of Gut Microbiota and the Interaction With Host Metabolism in Mouse Pulmonary Fibrosis Models

23 Background: Pulmonary fibrosis (PF) is a chronic progressive disease whose 24 pathogenesis is thought to be associated with activation of the immune system and 25 consequent metabolic changes. Recent studies suggested that gut microbes are closely 26 related with host's immune response and metabolic changes in fibrotic hosts. However, 27 the dynamic changes of the gut microbiome and the interaction profiles with host 28 metabolism during the development of pulmonary fibrosis remain inconclusive. 29 Results: We collected serum and fecal samples from bleomycin - induced fibrotic mice 30 at 0, 7, 14, and 28 days and performed UPLC - MS analysis on serum metabolites and 31 metagenomic sequencing on fecal samples. It is found that the serum metabolic 32 profile and gut microbiome were significantly altered in mice during the progression 33 of fibrosis. Among the serum metabolites, the levels of three major types of lipids, i.e. , 34 35 time - dependent changes. correlation with that of serum metabolites. The taxa from Bacteroides , such as 45 Butyricimonas_synergistica and Muribaculaceae, show positive correlation with the 46 cluster of glycerophospholipids, while taxa from Firmicutes , such as Clostridioides 47 difficile and Enterococcus faecium exhibit negative correlation. Further functional 48 classification suggested that those taxa are involved in multiple functional modules, 49 such as Transporters, Secretion system, and Metabolism. 50 Conclusions: The results reveal the synergistic changes between the gut microbiome 51 and host metabolism and the dynamic responses of gut microbiome to host fibrosis 52 during the progression of fibrosis. 53


Abstract 23
Background: Pulmonary fibrosis (PF) is a chronic progressive disease whose 24 pathogenesis is thought to be associated with activation of the immune system and  profile and gut microbiome were significantly altered in mice during the progression 33 of fibrosis. Among the serum metabolites, the levels of three major types of lipids, i.e., 34 glycerolipids, glycerophospholipids, and fatty acids exhibit significant 35 time-dependent changes. 36 The glycerolipid TG and multiple glycerophospholipids (3 PG, 6 PE, and 1 PC) 37 decreased in the early stage of fibrosis and increased in the late stage. The other two 38 types of glycerolipids MG and DG and the fatty acids Cartinine and Punicic acid 39 decreased through the development of fibrosis. In the meantime, we detected 40 significantly elevated abundance of gut microbiome taxa, including Prevotella sp. 41 from Bacteroidetes, Lactobacillus from Firmicutes, and Bifidobacterium from INTRODUCTION 56 Pulmonary fibrosis (PF) is a chronic progressive disease with excessive 57 deposition of collagen tissue in the lung interstitium. It is a common and irreversible 58 interstitial lung disease and one of the serious diseases of the respiratory system [1]. 59 The patients would develop lung fibroblast proliferation, extracellular matrix collagen 60 deposition, airway inflammatory cell infiltration, etc., which eventually led to alveolar 61 and lung tissue fibrosis [2]. Due to the lack of reliable early diagnosis methods and 62 effective treatment methods to reverse the natural process and outcome of fibrosis, the 63 median survival time from diagnosis was only 2.5-3.5 years, and the 5-year survival 64 rate was less than 40% [3,4]. Pulmonary fibrosis may be even more malignant than 65 some malignant tumors because of progressive respiratory failure. 66 Pulmonary fibrosis is usually caused by different stimuli such as toxin, infection, 67 severe trauma, autoimmune reaction, adverse drug reaction, and idiopathic, unknown In the present study, we aimed to examine the interactions between the gut 111 microbiota and serum metabolome, and investigate their dynamic changes in mice 112 with pulmonary fibrosis using multi-omics approaches.  All mice were housed in a barrier facility and were fed normal-chow diet. They 121 were randomly assigned to the four groups (control, model_7Day, model_14Day, 122 model_28days) with 10 mice in each group. To induce pulmonary fibrosis, mice were 123 injected intraperitoneally and anesthetized suitably with 1% pentobarbital sodium, and 124 then vertically mounted on a slope. A laryngoscope was held at the throat of the mice 125 and the tongue was pulled lightly, a hermetic venous indwelling needle was plugged 126 in and the inner tube was taken out, the drug ( 2.0 mg/kg bleomycin for experimental 127 group, the same amount of normal saline for normal group) was dropped into the 128 trachea, the tube was unplugged and the mice laid in a mouse cage.

133
Mice were deeply anesthetized with sodium pentobarbital and died from blood

162
The Kruskal-Wallis H test was used to determine the differences between groups.

175
(Tianjin, China). FastQC is used to assess the quality of raw data comprehensively generated by sequencing, while we screened the data for quality, removing sequences 177 less than 50 bp in length or containing ambiguous bases. All metagenomic raw data 178 have been submitted to the NCBI database (accession: PRJNA722397).  (Table S1). 49 top metabolites significantly 229 changed in all three fibrotic models are presented in the hierarchical clustering 230 heatmap ( Figure 2B and Table S2). It is shown that those metabolites form four  The taxonomic abundance analysis showed that the gut microbiota of the mouse identified 79 taxa with differential abundance across the four groups (Table S3). Their 272 discriminating ability indicated as LDA (linear discriminant analysis) score in Figure   273 3C revealed that the control group is characterized by a high abundance of    Figure 5A).

335
Notably, the correlation matrix between gut microbiota KO categories and serum 336 metabolites is also stratified to two clusters, of which the hierarchical clustering 337 pattern regarding to metabolites is highly in parallel to that for the correlation matrix 338 between gut microbiota taxa and serum metabolites ( Figure 5B). Specifically, the first  were significantly reduced in patients with pulmonary fibrosis by plasma assays [31]. 371 We also found that serum concentrations of glycerophospholipids were significantly           of the main functional groups in the sample community from KEGG pathway level 1 to level 3( from inner to outer).

Figure 5
Correlation analysis of metabolites and fecal microbes. (A) Spearman's rank correlation between differential gut microbial species and metabolites, 40 samples were used for Spearman's rank correlation, (B) Spearman's rank correlation between differential gut microbial functional KO and metabolites. Red, represent positive correlations; blue, represent negative correlations.