Disrupted Host-Microbiota Crosstalk Promotes Nonalcoholic Fatty Liver Disease Progression by Impaired Mitophagy

Disrupted Host-Microbiota Crosstalk Promotes Nonalcoholic Fatty Liver Disease Progression by Impaired Mitophagy | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Disrupted Host-Microbiota Crosstalk Promotes Nonalcoholic Fatty Liver Disease Progression by Impaired Mitophagy Wenjing Yin, Wenxing Gao, Yuwei Yang, Weili Lin, Wanning Chen, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4404936/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background : The intricate interplay between host genes and intrahepatic microbes is vital in shaping the hepatic microenvironment and contributes significantly to our understanding of nonalcoholic fatty liver disease (NAFLD). However, the underlying mechanisms of disease progression mediated by these interactions remain largely elusive. Methods : We conducted a comprehensive analysis of 570 liver biopsy transcriptomes from five cohorts, including 72 control, 124 nonalcoholic fatty liver (NAFL), 143 borderline and 231 nonalcoholic steatohepatitis (NASH) samples. Least Absolute Shrinkage and Selection Operator penalized regression and Sparse Canonical Correlation Analysis were utilized to identify host-microbiota interactions and their function. Results : We observed significant upregulations of key genes involved in mitochondrial organization across all disease stages, while genes related to antigen processing showed abnormal activations in advanced stages like NASH. Additionally, the abundances of intrahepatic microbes Methyloversatilis sp. RAC08 and Ralstonia insidiosa decreased significantly across all NAFLD stages. We identified 5537, 1937, 1485, and 2933 host-microbiota interactions in control, NAFL, borderline, and NASH samples, respectively. Notably, interaction strength showed a decreasing trend, especially during the transition from the borderline stage to NASH. In NAFL and borderline stages, bacteria like Bacillales, Ralstonia insidiosa , and Micromonosporaceae played pivotal roles in enhancing host mitophagy by interacting with genes including SQSTM1 , OPTN , and BNIP3L . However, such interaction functional clusters were absent in NASH samples. Conclusion : Disturbed host-microbiota interactions affecting the mitophagy process can lead to a pro-inflammatory hepatic microenvironment through activation of immune reactions, potentially driving disease progression to NASH. intrahepatic microbiome hepatic microenvironment host-microbe interaction nonalcoholic fatty liver disease mitophagy steatohepatitis mitochondria immune microenvironment disease progression bioinformatics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Nonalcoholic fatty liver disease (NAFLD) is a chronic liver disorder, affects an estimated 25% of the population and is increasingly recognized as a leading cause of liver-related mortality. NAFLD encompasses a spectrum of stages. Nonalcoholic fatty liver (NAFL), the initial period, is characterized by excessive hepatocyte triglyceride accumulation. Following a transitional stage termed “borderline”, the disease progresses into nonalcoholic steatohepatitis (NASH), marked by irreversible liver injury and severe hepatic inflammation. Alarmingly, some NASH patients are at risk of cirrhosis and hepatocellular carcinoma. In addition to local inflammation, systemic inflammation is also a typical characteristic in NAFLD patients, predisposing patients to various extrahepatic manifestations such as obesity, type 2 diabetes, and cardiovascular diseases. Multiple factors contribute to the pathogenesis of NAFLD, including genetic abnormalities, metabolic dysregulation, and gut microbiota dysbiosis. 1 – 3 Therefore, it is essential to consider the interplay between these elements to gain a more holistic understanding of the disease mechanisms. 4 , 5 For example, an integrated analysis that combined hepatic transcriptome, metabolome, and gut metagenome facilitated the depiction of coordinated disruption of gut-liver axis and causal mechanisms exploration in hepatic steatosis. 6 As gut-liver axis suggested, certain proinflammatory agents such as microbes were able to mitigated from gut to liver. 7 Using sequencing, qPCR, and FISH targeting 16S rRNA, studies have evidenced the presence of microbiota within the liver. 4 , 8 – 11 Albeit at a lower abundance than in the gut, these hepatic microbes provide crucial insights into the liver microenvironment since they reside directly within the liver. 4 For example, these hepatic bacteria, detecting by 16S rRNA sequencing, were shown a close association with liver phenomics including ballooning degeneration, inflammation and fibrosis. 4 Furthermore, these hepatic microbes were also active participators in various cellular process through interaction with host, 9 – 11 such as modulating host natural killer T cells programming to induce inflammation in both mice and humans. 9 Notably, the direct connection has been found in hepatic microbes with NAFLD/NASH risk alleles. For example, Enterobacter and Pseudoalteromonas were strongly associated with PNPLA3 rs738409 and TM6SF2 rs58542926 variants, suggesting the potential of hepatic microbes in modulating disease risk. 10 , 11 These exciting observations prompt us to delve into the hepatic microbes and their interaction with host, to gain a comprehensive understanding on the dynamics of tissue microenvironment. Given the complex, “multi-hit” nature of NAFLD pathogenesis, it is of significance to conduct a multi-dimensional analysis that integrates the entirety of all relevant genes, hepatic microbes, and their interactions. Moreover, a general pattern of NAFLD evolvement highlights the necessity to consider every progressive stage, as this approach not only facilitate the identification of features specific to each stage, but also aids in the exploration of potential mechanisms driving disease worsening. However, a comprehensive and systematic evaluation that simultaneously considers multi-dimensional perspectives across the whole progressive stages of NAFLD remains absent. Therefore, this study aims to fill this gap by evaluating the hepatic microenvironment at different stages of NAFLD. By understanding the distinct gene and microbial profiles, we aim to illuminate the complex pathogenesis of NAFLD progression and identify pivotal mechanisms driven by host-microbiota interactions. To achieve this goal, we conducted a comprehensive analysis of hepatic microenvironment in whole progressive stages including control, NAFL, borderline ang NASH. In addition, we integrated five cohorts from different regions to gain robust insights into the composition and interaction of host genes and hepatic microbes. 12 – 16 Our findings showed host-microbiota interactions in worse stages of NAFLD were disrupted, especially those related to mitophagy progress, which might contribute to disease progression. Collectively, focusing on holistic view of gene, microbes and host-microbiota interactions, this study introduced a novel perspective in understanding the dynamics of microenvironment during the progressive stages of NAFLD. Methods Sample collection We scoured public databases, including GEO, BioProject and ArrayExpress, for RNA sequencing data of liver biopsies from healthy or NAFLD stages (Online Resource 1). Written informed consent was obtained from all individuals. 12 – 16 Healthy liver biopsies were derived from people undergoing bariatric surgery. Seven studies with samples without a history of alcohol intake or viral infection were collected. To minimize the influence caused by different diagnostic standards, the disease stages of patients from three studies (PRJNA558102, PRJNA767535 and PRJNA682622) were reclassified uniformly according to the criteria of the NASH-Clinical Research Network system: samples with NAFLD Activity Score (NAS) scores < 3 were defined as NAFL, samples with NAS scores ≥ 5 were defined as NASH and the remaining were classified as borderline. 17 Samples from another two studies (PRJNA704861 and PRJNA523510) retained disease classification provided by the original publications (Online Resource 1). Hepatic gene expression preprocessing Raw sequencing data were downloaded and uniformly reprocessed to gene expression profiles with following steps. First, adapter removal and quality control were conducted by Trimmomatic (version 0.39) and the qualified reads were aligned to the newest human reference genome 18 (CHM13 v2.0) with STAR (version 2.7.10a). Focusing on genes encoding for proteins, we excluded other gene types such as micro-RNA using R package “biomaRt” (version 2.52.0). Subsequently, variance stabilizing transformation was performed using “DESeq2” (version 1.28.3), followed by filtering out less informative genes which variance below the 25th percentile cross all the samples. Hepatic microbial composition preprocessing Since microbes can also be captured by transcriptomics sequencing, we refined the workflow to detect the hepatic microbial sequences within liver biopsy RNA-seq data. 19 , 20 To mitigate false positives and potential contaminants, a rigorous decontamination process was conducted (Online Resource 2). First, to eliminate the distortion by a potentially high proportion of human sequences on microbial read counts, we employed Kneaddata (version 0.6.1) to remove the sequences aligned to CHM13v2 human genome (Module 1, Step 1 in Online Resource 2). The unaligned reads were regarded as candidate microbial sequences, subsequently annotated using KrakenUniq (version 1.0.4). This annotation enables precise differentiation between genuine microbial signals and false positives arising from highly repetitive or contaminated sequences 21 (Module 1, Step 2 in Online Resource 2). A customized database for KrakenUniq comprises the CHM13v2 human genome and complete microbial genomes downloaded from RefSeq (retrieved 5th September 2023), which includes 35,022 genomes from 8,907 bacterial species, 14,993 genomes from 13,986 viral species, 35 genomes from 35 fungal species, and 540 genomes from 423 archaeal species. Additionally, artificial sequences such as UniVec and EmVec are also included in the database. Bracken (version 2.5.0) was used subsequently to quantify the abundances. Since k-mers from a true organism will distribute evenly across the genome, the taxa with low unique k-mer counts were considered contaminations, using read length minus k-mer length + 1 as threshold (Unique k-mer filtering: Module 2, Step 3 in Online Resource 2). Then, taxa with total k-mer counts exceeding read counts by more than fivefold were retained (Total k-mer filtering: Module 2, Step 4 in Online Resource 2). For each taxon, the relationships between total read counts, unique k-mer counts and total k-mer counts were examined and those with significant Spearman correlations were considered true microbial sequences 22 (Module 2, Step 5 in Online Resource 2). Meanwhile, for those contaminations which are known commonly during experiments, we adopted a reference-based decontamination process. Taxa whose abundance resembled a distribution with the cell line data were considered experimental contaminations and were excluded from further analyses 22 (Module 2, Step 6 in Online Resource 2). The remaining taxa were compared with a published common reagent contaminations list 23 and all matches were removed (Module 2, Step 7 in Online Resource 2). As last, taxa with average relative abundance above 0.01% and present in at least 10% of the samples were retained for further analyses (Module 3, Step 8 & 9 in Online Resource 2). Interaction analysis between hepatic genes and microbes Least Absolute Shrinkage and Selection Operator (Lasso) penalized regression was used to identify the sparse interactions between host genes and hepatic microbes. To capture interactions at all taxonomic level, we combined abundance data at different taxonomic levels into a combined taxa matrix. For a given host gene, we detected the taxa that most accurately predict its expression pattern across samples within each stage. Subsequently, the combination of the host gene and the predicting taxa was considered as a host-microbiota interaction. R package “glmnet” was implemented for lasso analysis (version 4.1-7), with the optimal tuning parameter λ determined through leave-one-out cross-validation. Regression model significance was gauged using “hdi” (version 0.1-9), with interactions having a false discovery rate (FDR) less than 0.01 post Benjamini-Hochberg correction as significant. For stability assurance, the lasso model was iteratively refitted (100 times) to random sample subsets with perturbed λ. Interactions appearing in at least 60% iterations were considered stable. Interaction strength was quantified through Spearman correlation via “stats” (version 4.2.1). Interaction visualization and degree of the nodes was computed in Gephi. Sparse Canonical Correlation Analysis (SparseCCA) was utilized to identify the functional interaction clusters by projecting host genes and hepatic microbial data into a shared latent space, while maximizing the correlation between these two datasets at the same time. Combined taxa matrix at all taxonomic levels and the gene matrix were the input for SparseCCA in each stage. L1 penalty was incorporated to select the most associated host genes and microbes as a group. In each of the four stages, ten most associated groups were identified separately, and the significance of each group was computed using the leave-one-out cross-validation approach followed by post hoc Bonferroni correction. Such process was implemented using R package “PMA” (version 1.1). To explore the specific function of each group, we performed the KEGG pathway enrichment for the genes involved in each group using “enrichKEGG” function in R package “clusterProfiler”. Pathways were considered significant if they had a corrected FDR of less than 0.05. Statistical analysis Each profile of single dataset was combined into a unified one using “merge” function in R. Permutational Multivariate Analysis of Variance (PERMANOVA) was used to assess the general variance caused by batch effect variable “Study” and true biological variable “Stage”, whose specific influence on the main variance was analyzed using Kruskal-Wallis rank sum tests on the first two principal coordinates of Principal Coordinate Analyses. Batch effect adjustment was performed using “RemoveBatchEffect” function in limma R package (version 3.54.2) on the variance stabilized gene expression data and centered log ratio transformed microbial abundance data. Differentially expressed genes (DEG) and microbes were detected using limma R package. Multiple hypothesis correction was performed using the post hoc Bonferroni method. Differentially expressed genes were defined as those with absolute log fold change greater than 1.2 and FDR less than 0.05, and differentially abundant microbes refer to those with absolute log fold change more than 1 and FDR less than 0.05. Pathway enrichment for host genes was performed by gene set enrichment analysis (GSEA) using clusterProfiler R package (version 4.4.4). Significantly enriched pathways were defined as those with FDR less than 0.001. Results 1. Cohort collection and study design To investigate the hepatic microenvironment throughout the progression of NAFLD, we curated publicly available RNA sequencing samples from various data repositories. Five datasets, adhering to rigorous selection criteria, were integrated into this study. Samples involved were sourced from various countries and continents, including Japan, the United States, Denmark, and other countries in the European region. A total of 570 samples were enrolled, which encompassed 72 control samples, 124 NAFL patients, 143 borderline patients and 231 NASH patients (Online Resource 1). Here, the “borderline” stage was considered as a transitional state between NAFL and NASH, characterized by an intermediate NAS score of 3 and 4. 24 Detailed dataset information and sample demographic characteristics were summarized in Online Resource 9. After batch effect removal and rigorous decontamination processes (Online Resource 2–5), we obtained hepatic gene expression (14,414 genes) and microbial abundance (350 taxa) profiles for these samples. Subsequently, we delineated specific patterns in hepatic gene expression and microbial abundance. Underlying mechanisms of this dynamic niche were further elucidated by investigating the intricate interactions between the host and microbiota (Fig. 1 ). 2. Altered gene expression suggested disturbed energy metabolism and active immune response in NAFLD We firstly investigated hepatic genes associated with NAFLD with a DEG analysis. Taking control as the baseline, we identified 1639 DEGs in NAFL, 1802 DEGs in borderline and 2407 DEGs in NASH (Fig. 2 a & Online Resource 10–12). Among these, we found a consistent pattern in the most dramatically changed DEGs. For example, IL32 , AKR1B10 and FABP4 all increased progressively with the disease progressing, while EGR1 , CYP2C19 and VIL1 exhibited the opposite (Fig. 2 a). Besides, a total of 358 DEGs were consistently up-regulated and 536 DEGs were uniformly down-regulated in all three disease stages (Fig. 2 b-c). GSEA revealed pathway modulations in each disease stage compared to the control: in NAFL, 23 pathways were activated and 34 pathways were inhibited (Online Resource 6a & Online Resource 13); in the borderline stage, 25 pathways were activated and 5 pathways were inhibited (Online Resource 6d & Online Resource 14); in NASH, 61 pathways were activated and 16 pathways were inhibited (Online Resource 6g & Online Resource 15). Stage-specific and shared pathways revealed that both discrepancies and similarities existed along disease progression. In detail, earlier stage NAFL was specifically featured in the activation in ATP synthesis (Fig. 2 d) and oxidative phosphorylation (Online Resource 6b), and inhibition in organ development related pathway (Fig. 2 e & Online Resource 6c). Comparatively, borderline and NASH exhibited greater similarities, both showed the activation in extracellular matrix organization related pathway (Fig. 2 d & Online Resource 6h) and immune activities such as antigen processing and representation (Online Resource 6e & Online Resource 5g). In addition, consistent inhibition in xenobiotic catabolic process (Online Resource 6f & Online Resource 6g) and response too zinc ion (Fig. 2 e & Online Resource 6i) also suggested the shared abnormities between borderline and NASH. Notably, mitochondrion organization and ribosome biogenesis were activated across all disease stages compared to the control, suggesting that these cellular processes may play a persistent role during the whole process (Fig. 2 d). 3. Turbulent hepatic microbiota involved in the hepatic microenvironment in NAFLD The hepatic microbial landscape was comprehensively depicted by annotating multi-kingdom microbial reads of RNA-seq data, followed by a series of decontamination processes to remove human sequences, false positives during annotation, and common laboratory contaminations (Online Resource 3a-b). A total of 350 taxa ranging from domain to species were identified, covering bacteria (296 taxa), fungi (46 taxa), viruses (7 taxa) and archaea (1 taxa). 14 bacterial phyla were identified, with Proteobacteria, Actinobacteria and Firmicutes being predominant (Online Resource 7), indicating a distinct microbial profile compared to fecal microbiota which was dominated by Bacteroidetes. 25 Meanwhile, Basidiomycota and Ascomycota of fungi, and Artverviricota and Uroviricota of viruses were also observed presence (Online Resource 7). A turbulent microenvironment featured by dynamic hepatic microbiota composition was observed in NAFLD. First, the composition at the phylum level showed notable variations. Actinobacteria and Fusobacteriota were decrease while Firmicutes and Ascomycota were increase in NAFLD (Online Resource 7). In addition, a significantly higher alpha diversity distinguished NAFL with other stages (Fig. 3 a). Different composition among stages were further elucidated with a significant difference in beta diversity analysis (Fig. 3 b). An in-depth evaluation was further performed to assess the microbiota composition with higher resolution at species level. First, high-abundance species were analyzed, as their dominant abundance could potentially account for important effects. Here we defined high-abundance species as the taxa with abundance exceed the average of all samples were (Fig. 3 c). 37 high-abundance species were found totally, and 20 of them were showed in all four stages, accounting for 77.54% of the relative abundance (Fig. 3 c). Moreover, 15 of these high-abundance species exhibited differential abundances among stages (Fig. 3 d). Within these species, Cutibacterium acnes was the most abundant bacterium (Fig. 3 d). Comparatively, as the most abundant fungus, Malassezia restricta exhibited a stable abundance across the four stages (Fig. 3 d). Additionally, to understand the potential role of hepatic microbiota associated with NAFLD, we analyzed the differential abundant taxa at species level between each disease stage and the control stage. As a result, 18 species in NAFL, 12 in borderline, and 17 in NASH displayed significant differences in abundance (Online Resource 16). Among these differential species, eight bacterial species were observed significant differences in all of the disease stages compared to the control. Two species of them, Methyloversatilis sp. RAC08 and Ralstonia insidiosa , were previously identified as high abundance species and both showed decrease in all of the disease stages (Fig. 3 e). 4. Disrupted hepatic microenvironment characterized by impaired host-microbiota interactions in NAFLD patients One of the key objectives of this study was to explore the hepatic microenvironment and its relationship with the NAFLD progression, integrating both host genes and hepatic microbiota. To achieve this goal, we used LASSO regression to identify the interaction between a specific gene and a microbiota (Fig. 4 a). In the control stage, we identified 5,537 significant and stable host-microbiota interactions (Online Resource 17). Remarkably, these interactions were reduced to 1,937 in NAFL (Online Resource 18), further declined to 1,455 in the borderline stage (Online Resource 19), and 2,933 in the NASH stage (Online Resource 20). The number of involved genes (Fig. 4 b) and taxa (Fig. 4 c) exhibited a similar pattern, characterized by an overall decline with a marginal resurgence. Notably, we found no identical interactions appeared in four stages (Fig. 4 d), but similarities were exhibited among three disease stages. For example, an identical relationship between genus Micromonospora and FGF21 was observed in all of the three disease stages (Fig. 4 d). In addition, 487 identical interactions were present in both NAFL and NASH, with an additional 12 interactions appeared in both borderline and NASH, and 5 interactions appearing in both NAFL and borderline (Fig. 4 d). An overall decreasing strength was exhibited both positive and negative interactions as the disease progresses (Fig. 4 e). NAFL and NASH exhibited a higher degree of similarity, while the borderline stage demonstrated modest fluctuations. This may reflect the inherent instability of the transitional borderline stage. Despite the modest fluctuation, the decline was particularly dramatic when comparing the interaction strength of NASH to that of borderline (FDR < 0.0001 with Kruskal-Wallis rank sum tests for both positive and negative interactions), indicating an weak communication network between host and microbiota in advance stage NASH. To explore the interaction network more deeply, we examined the dominant interacted microbiota in every stage. Results showed that the host-microbiota interaction network were distinct in control but were analogous in three disease stages. Specifically, in the control liver microenvironment, Escherichia coli , Actinomyces naeslundii and Pseudomonas oleovorans were the most associated taxa (Fig. 4 f); however, interactions with these taxa were nearly absent in disease stages. On the contrary, the three disease stages exhibited a consistent pattern. The order Sphingomonadales and the family Sphingomonadaceae emerged as the most interactive taxa in NAFL and borderline stages, while Ralstonia insidiosa was identified as the taxon engaging the most extensive gene interactions in NAFL and NASH (Fig. 4 f). The reduced participation, weaker interplay and varied interaction network collectively suggested a significantly dysregulated and vulnerable hepatic microenvironment as the disease advances. 5. Hepatic microbiota was associated with host mitophagy and immune response in NAFLD To gain a deeper understanding of the potential pathogenetic mechanisms driving distinct patterns of host-microbiota interactions, we identified interaction clusters of host genes and microbiota. The biological processes indicated by the genes and taxa in each cluster shed light on the coordinate functions of these interactions. Our analysis revealed 145, 23 and 101 pathways in the control, NAFL, and borderline, respectively. However, no significant pathways were detected for NASH (Fig. 5 a & Online Resource 21), which is possibly attributed to the disturbed interactions precluding the execution of a consistent function in NASH. Five stage-shared pathways in NAFL and borderline were identified (Fig. 5 b). Subsequently, 89 stage-specific pathways were revealed, including 66, 3 and 20 pathways for control, NAFL and borderline, respectively (Fig. 5 a). Three of each stage-specific pathways were highlighted (Fig. 5 b) as they were of importance in fundamental cellular processes. Here, control-specific host pathways included autophagy and biosynthesis of unsaturated fatty acids, while NAFL-exclusive associations were enriched for cell signal pathways and ferroptosis (Fig. 5 b). Borderline-specific pathways included the mTOR signaling pathway, PI3K-Akt pathway and antigen processing and presentation pathway (Fig. 5 b). Mitophagy was especially focused on due to the essential role of mitochondria in energy production and close association with NAFLD. 8 Our results showed that Bacillales, Ralstonia insidiosa and Micromonosporaceae, two dominant taxa in the previous interaction network (Fig. 4 f), were involved in the host mitophagy process in both NAFL (Fig. 5 c) and borderline stages (Fig. 5 d). Relevant genes exhibited similarity as they both appeared in interaction clusters in both stages, including TBC1D17, HRAS. (Fig. 5 c & Fig. 5 d). To investigate the impact of these interactions on mitophagy, we examined the expression of related genes. As expected, with interactions absent in NASH, the up-regulated trend of related genes was relatively modest in NASH. For example, SQSTM1 was observed an insignificant variance in NASH compared to the borderline (Online Resource 8a). Similar pattern was recognized for OPTN and BNIP3L , encoding for other cargo receptors (Online Resource 8b-c). The mitochondrial kinase, PINK1 , was also up-regulated in NAFLD patients (Online Resource 8d). Intriguingly, STING , a cytosolic sensor involved in the xenophagy process — a process akin to mitophagy but targeting intracellular bacteria instead — showed an overall rising expression under disease liver stages (Online Resource 8e). In addition, T cell receptor signaling pathway was emphasized to explore the role of host-microbiota interactions on microenvironment inflammation during disease progression. As a result, genes including PTPN11 , PIK3CB and MAP2K7 were found to interact with hepatic microbiota in both NAFL and borderline (Online Resource 8f-h). These taxa included Bacilli, Sphingomonadales and Ralstonia insidiosa , were also the most gene-associated taxa in NAFL and borderline. These findings echoed the up-regulated antigen processing pathway in advanced stages such as borderline and NASH (Online Resource 6e). Genes in these interactions showed a similar up-regulation (Online Resource 8i-k), suggesting that these interactions could result in an immune-activated microenvironment. Altogether, these abnormities in host-microbiota interactions and related gene expressions inspired the hypothesis that hepatic microbiota may enhance the mitophagy process as a way of facilitating their survival in the liver. Should these interactions be disrupted, the resulting impaired clearance of dysfunctional mitochondria could release reactive oxygen species (ROS), which further leads to tissue damage and inflammation that drive disease progression (Fig. 6 ). Consequently, diminished mitophagy in NASH impair xenophagy inhibition, enabling more hepatic microbes to be degraded by host (Fig. 6 ). Discussion To explore potential mechanisms associated with NAFLD development and progression, this study encompassed the entire spectrum of NAFLD, from control, NAFL, the transient stage borderline, to the advanced stage NASH. In addition, multiple cohorts from various geographical regions were involved, allowing a more robust finding. Our study highlighted that, the previously overlooked hepatic microbiota was an indispensable constituent of the hepatic microenvironment. Most importantly, the intricate interplay between these microbiota and host provided a novel perspective to understand the role of microenvironment during NAFLD progression. By extracting microbial RNA fragments with rigorous decontamination, our study was the first to comprehensively investigate the active presence of hepatic microbes, which we found to be distinctly different from fecal microbes composition. 25 Bacteroidetes and Firmicutes were the most abundant phyla in fecal of NAFLD patients, 25 while Proteobacteria and Actinobacteria were predominant in liver of NAFLD. The disparities underscored the significance of exploring the tissue microbial signatures in the hepatic microenvironment. Interestingly, we also identified hepatic RNA fragments belonging to fungi. These fungi showed the underestimated diversity of liver microbiota, a field previously concentrated predominantly on bacterial components. 4 , 8 – 10 However, the potential origin of the hepatic microbiome necessitates further investigation. Specifically, we identified that Cutibacterium acnes , was predominant in the liver tissue and showed a decline in disease individuals. This species is known as a commensal bacterium, producing short-chain fatty acids to help maintain host homeostasis. We also found that a fungus species, Malassezia restricta , adapted well to the hepatic environment, possibly because a rich influx of lipids in liver satisfies its lipid-dependent growth. Overall, the diversity and dynamic abundance of hepatic taxa, not replicable by gut microbiota, highlight their previously underestimated but critical role in the hepatic microenvironment. A key contribution of this study is our approach to investigate about the hepatic microenvironment, which spans the whole spectrum of NAFLD stages including control, NAFL, borderline to NASH. This granularity holds clinical significance, as stage-specific features imply potential mechanisms for disease progression and may of therapeutic value. 16 Microenvironment landscape in each stage was first delineated with complementary results from host gene and hepatic microbiota compositions. Here, the turbulence in hepatic microbiota indicated a general dysbiosis of hepatic microenvironment, while abnormalities in host genes provide more specific insights in each disease stage. For example, NAFL was featured by abnormal energy metabolism while borderline and NASH exhibited activation of immune activities. These observations were also confirmed by other researches. 26 , 27 Most importantly, the analysis on interactions of host and microbes facilitated an in-depth examination of network and function within the hepatic microenvironment. The findings from this approach created a repository of potential mechanisms for disease progression, thereby extending the scope of conventional studies that typically focused on a single aspect. In general, our results showed the worse stages of NAFLD were featured by a sparser and weaker host-microbiota interaction network. This weaking was accompanied by a consistent disappearance of functional responses, which eventually loss in NASH. Take a closer examination, the hepatic microbiota Sphingomonadaceae family and the corresponding order Sphingomonadales maintained interactive status across NAFL and borderline. While these taxa have been experimentally detected in NAFLD livers by others studies, 8 , 9 , 11 their specific influence in NAFLD disease has not been extensively explored. Our studies suggested that these taxa might contribute to hepatic inflammation by affecting host T cell receptor pathway. These interactions were possibly mediated by glycosphingolipid, a metabolite secreted by Sphingomonadaceae and Sphingomonadales. This aligned with recent findings that intrahepatic Bacteroidetes phyla could modulate host T cells by glycosphingolipids. 9 Therefore, our study suggested that Sphingomonadaceae and Sphingomonadales may possess a more targeted therapeutic potential in NAFLD, surpassing the broader implications previously identified at the phylum level. 9 Furthermore, other microbial interactions that specifically appeared in the control stage may also prevent steatosis by aiding in the maintenance of tissue homeostasis. For instance, taxa such as Pseudomonas aeruginosa and Granulicatella were observed interacting with genes including HACD3 . These interactions could facilitate the host’s biosynthesis of unsaturated fatty acids, a process considered beneficial for liver health. Novel mechanisms was also revealed driving by host-microbiota interactions, highlighting the importance of this multifaceted approach assessing the hepatic microenvironment. Dysfunctional mitochondrial removal is evidenced in NAFLD and enhanced mitophagy rescues liver damage. 8 , 28 However, the exact internal biological dysfunctions that induce compromised mitophagy in NAFLD remains elusive. 28 Taking advantage of interaction analysis, we suggested that impaired mitophagy could attribute to the loss of certain host-microbiota interactions in NASH. The disruption of these interaction was possibly associated with the decline of core microbes including Bacillales, Ralstonia insidiosa and Micromonosporaceae, which were experimentally evidenced the ability to amplify host mitophagy. 29 Consequent accumulation of damaged mitochondria may produce reactive oxygen species (ROS), thereby exacerbate tissue inflammation. Such a pro-inflammatory microenvironment was also confirmed by the activation of immune response related genes and pathways. Taking together, investigating the host-microbiota interaction-driven microenvironment offers a broader and deeper view of the pathogenic mechanisms underlying NAFLD, surpassing the insights gained from transcriptomics analysis alone. A common challenge when analyze sequencing data is to obtain high repeatable results, with variabilities between different experimental batch. These variabilities if unrelated to the true biological variable, was considered batch effect. To gain a genuine and robust understanding of microenvironment components, our study was based on five cohorts with a rigorous consideration and correction of batch effect. 30 Our results showed a great concordance with previous findings, affirming their reliability and repeatability. For example, gene IL32 exhibited the most dramatically upregulated expression in all of the NAFLD stages, and CYP2C19 showed the opposite trend, these were both confirmed in existing findings. 26 , 27 With regard to hepatic microbiota, our results suggested Proteobacteria was the most abundant phyla, which was evidenced in previous experiments using 16S RNA and qPCR. 4 , 9 In summary, this study has comprehensively demonstrated the microenvironment across the whole progression of NAFLD, using an integrated analysis across multiple cohorts. Our results suggested disruption of host-microbiota interactions in NASH, especially those who are vital for mitophagy, implied novel mechanisms in disease stage transition. Collectively, our results shed light on a new perspective in the multifaceted mechanisms driving NAFLD progression, providing insights for potential therapeutic targets and intervention strategies. Abbreviations DEG , Differentially Expressed Genes; FC , Fold Change; FDR , False Discovery Rate; GSEA , Gene Set Enrichment Analysis; Lasso , Least Absolute Shrinkage and Selection Operator; NAFL , Nonalcoholic Fatty Liver; NAFLD , Nonalcoholic Fatty Liver Disease; NAS , NAFLD Activity Score; NASH , Nonalcoholic Steatohepatitis; PERMANOVA , Permutational Multivariate Analysis of Variance; ROS , Reactive Oxygen Species; SparseCCA, Sparse Canonical Correlation Analysis. Declarations Conflicts of interest disclosure: No potential conflict of interest was reported by the author(s). Data availability statement : All of the processed data in this study has been uploaded in the National Omics Data Encyclopedia under accession NO. OEP005066. The raw metagenomic data are available in the European Nucleotide Archive (https://www.ebi.ac.uk/ena/) under accession NOs. PRJNA558102, PRJNA767535, PRJNA682622, PRJNA704861 and PRJNA523510. The data relevant to the study are uploaded as supplementary information. The code and scripts are available on GitHub (https://github.com/tjcadd2020/NAFLD-Host-Microbe-Interaction). Funding statement: This work was supported by the National Natural Science Foundation of China (82000536 to NJ, 92251307 to RZ, 82170542 to RZ), the National Key Research and Development Program of China (2021YFF0703700/2021YFF0703702 to RZ), Guangdong Province “Pearl River Talent Plan” Innovation and Entrepreneurship Team Project (2019ZT08Y464 to LZ), the program of Guangdong Provincial Clinical Research Center for Digestive Diseases (2020B1111170004), and National Key Clinical Discipline. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Author contributions: NJ, LZ, and RZ conceived and designed the project. WY and WG drafted the manuscript. YY, WL, WC, XZ, RZ, LZ, and NJ revised the manuscript. All authors read and approved the final manuscript. References Wu D, Liu L, Jiao N, et al. Targeting keystone species helps restore the dysbiosis of butyrate-producing bacteria in nonalcoholic fatty liver disease. iMeta. 2022;1(4):e61. Jiao N, Loomba R, Yang Z-H, et al. Alterations in bile acid metabolizing gut microbiota and specific bile acid genes as a precision medicine to subclassify NAFLD. Physiol Genom. 2021;53(8):336–48. Jiao N, Baker SS, Chapa-Rodriguez A, et al. Suppressed hepatic bile acid signalling despite elevated production of primary and secondary bile acids in NAFLD. Gut. 2018;67(10):1881–91. Sookoian S, Salatino A, Castaño GO, et al. Intrahepatic bacterial metataxonomic signature in non-alcoholic fatty liver disease. Gut. 2020;69(8):1483–91. Xu T, Gao W, Zhu L et al. NAFLDkb: A Knowledge Base and Platform for Drug Development against Nonalcoholic Fatty Liver Disease. Journal of Chemical Information and Modeling . 2023;10.1021/acs.jcim.3c00395. Hoyles L, Fernández-Real J-M, Federici M, et al. Molecular phenomics and metagenomics of hepatic steatosis in non-diabetic obese women. Nat Med. 2018;24(7):1070–80. Jiao N, Baker SS, Nugent CA, et al. Gut microbiome may contribute to insulin resistance and systemic inflammation in obese rodents: a meta-analysis. Physiol Genom. 2018;50(4):244–54. Liu B, Zhou Z, Jin Y, et al. Hepatic stellate cell activation and senescence induced by intrahepatic microbiota disturbances drive progression of liver cirrhosis toward hepatocellular carcinoma. J Immunother Cancer. 2022;10(1):e003069. Leinwand JC, Paul B, Chen R, et al. Intrahepatic microbes govern liver immunity by programming NKT cells. J Clin Investig. 2022;132(8):e151725. Pirola CJ, Salatino A, Quintanilla MF, Castaño GO, Garaycoechea M, Sookoian S. The influence of host genetics on liver microbiome composition in patients with NAFLD. eBioMedicine. 2022;76:103858. Sookoian S, Pirola CJ. Liver tissue microbiota in nonalcoholic liver disease: a change in the paradigm of host-bacterial interactions. Hepatobiliary Surg Nutr. 2021;10(3):337–49. Gerhard GS, Legendre C, Still CD, Chu X, Petrick A, DiStefano JK. Transcriptomic Profiling of Obesity-Related Nonalcoholic Steatohepatitis Reveals a Core Set of Fibrosis-Specific Genes. J Endocr Soc. 2018;2(7):710–26. Govaere O, Cockell S, Tiniakos D, et al. Transcriptomic profiling across the nonalcoholic fatty liver disease spectrum reveals gene signatures for steatohepatitis and fibrosis. Sci Transl Med. 2020;12(572):eaba4448. Kozumi K, Kodama T, Murai H, et al. Transcriptomics Identify Thrombospondin-2 as a Biomarker for NASH and Advanced Liver Fibrosis. Hepatology. 2021;74(5):2452–66. Pantano L, Agyapong G, Shen Y, et al. Molecular characterization and cell type composition deconvolution of fibrosis in NAFLD. Sci Rep. 2021;11(1):18045. Yao K, Tarabra E, Sia D, et al. Transcriptomic profiling of a multiethnic pediatric NAFLD cohort reveals genes and pathways associated with disease. Hepatol Commun. 2022;6(7):1598–610. Brunt EM, Kleiner DE, Wilson LA, Belt P, Neuschwander-Tetri BA. for the NCRN. Nonalcoholic fatty liver disease (NAFLD) activity score and the histopathologic diagnosis in NAFLD: distinct clinicopathologic meanings. Hepatology. 2011;53(3):810–20. Nurk S, Koren S, Rhie A, et al. The complete sequence of a human genome. Science. 2022;376(6588):44–53. Poore GD, Kopylova E, Zhu Q, et al. Microbiome analyses of blood and tissues suggest cancer diagnostic approach. Nature. 2020;579(7800):567–74. Salzberg SL, Breitwieser FP, Kumar A, et al. Next-generation sequencing in neuropathologic diagnosis of infections of the nervous system. Neurol Neuroimmunol Neuroinflamm. 2016;3(4):e251. Breitwieser FP, Baker DN, Salzberg SL. KrakenUniq: confident and fast metagenomics classification using unique k-mer counts. Genome Biol. 2018;19(1):1–10. Ghaddar B, Biswas A, Harris C, et al. Tumor microbiome links cellular programs and immunity in pancreatic cancer. Cancer Cell. 2022;40(10):1240–53. Salter SJ, Cox MJ, Turek EM, et al. Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol. 2014;12(1):1–12. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41(6):1313–21. Zhu L, Baker SS, Gill C, et al. Characterization of gut microbiomes in nonalcoholic steatohepatitis (NASH) patients: A connection between endogenous alcohol and NASH. Hepatology. 2013;57(2):601–9. Powell NR, Liang T, Ipe J, et al. Clinically important alterations in pharmacogene expression in histologically severe nonalcoholic fatty liver disease. Nat Commun. 2023;14(1):1474. Moore MP, Cunningham RP, Meers GM, et al. Compromised hepatic mitochondrial fatty acid oxidation and reduced markers of mitochondrial turnover in human NAFLD. Hepatology. 2022;76(5):1452–65. Li R, Xin T, Li D, Wang C, Zhu H, Zhou H. Therapeutic effect of Sirtuin 3 on ameliorating nonalcoholic fatty liver disease: The role of the ERK-CREB pathway and Bnip3-mediated mitophagy. Redox Biol. 2018;18:229–43. Sanjuan MA, Dillon CP, Tait SWG, et al. Toll-like receptor signalling in macrophages links the autophagy pathway to phagocytosis. Nature. 2007;450(7173):1253–7. Gao W, Chen W, Yin W et al. Identification and validation of microbial biomarkers from cross-cohort datasets using xMarkerFinde. 2022. PROTOCOL (Version 1) available at Protocol Exchange [ https://doi.org/10.21203/rs.3.pex-1984/v1] . Online Resources Online Resources 9 and 10 are not available with this version. Supplementary Files SupplementaryMaterial.pdf OnlineResource11.pdf OnlineResource12.pdf OnlineResource13.pdf OnlineResource14.pdf OnlineResource15.pdf OnlineResource16.pdf OnlineResource17.pdf OnlineResource18.pdf OnlineResource19.pdf OnlineResource20.pdf OnlineResource21.pdf Graphicabstract.png Graphic Abstract This study encompassed 570 liver biopsy transcriptomes from five cohorts, including 72 control, 124 nonalcoholic fatty liver (NAFL), 143 borderline and 231 nonalcoholic steatohepatitis (NASH) samples. Hepatic microenvironment during disease progression was investigated from aspects including host transcriptome, hepatic microbiome and host-microbiota interactions. Notably, bacteria including Bacillales, Ralstonia insidiosa , and Micromonosporaceae were observed a vital role in enhancing host mitophagy by interacting with genes including SQSTM1 , OPTN , and BNIP3L . However, the absent of such interaction functional clusters might lead to a pro-inflammatory hepatic microenvironment through the activation of immune reactions, which potentially drove disease progression to NASH. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4404936","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":309377504,"identity":"9b9d54ac-3e1e-4d43-ad54-f654254be4c7","order_by":0,"name":"Wenjing Yin","email":"","orcid":"","institution":"Tongji University","correspondingAuthor":false,"prefix":"","firstName":"Wenjing","middleName":"","lastName":"Yin","suffix":""},{"id":309377505,"identity":"8d87104f-4a78-4ee4-88bc-093257aecd1e","order_by":1,"name":"Wenxing Gao","email":"","orcid":"","institution":"Tongji University","correspondingAuthor":false,"prefix":"","firstName":"Wenxing","middleName":"","lastName":"Gao","suffix":""},{"id":309377506,"identity":"330a87dc-f6bd-43e7-9253-3e01cf2a3282","order_by":2,"name":"Yuwei Yang","email":"","orcid":"","institution":"Tongji University","correspondingAuthor":false,"prefix":"","firstName":"Yuwei","middleName":"","lastName":"Yang","suffix":""},{"id":309377507,"identity":"5d4eba95-8d8b-4b56-b08d-542613094a5b","order_by":3,"name":"Weili Lin","email":"","orcid":"","institution":"Tongji University","correspondingAuthor":false,"prefix":"","firstName":"Weili","middleName":"","lastName":"Lin","suffix":""},{"id":309377508,"identity":"337566a7-29ed-4dd9-964f-ba0f2a8a4635","order_by":4,"name":"Wanning Chen","email":"","orcid":"","institution":"Tongji University","correspondingAuthor":false,"prefix":"","firstName":"Wanning","middleName":"","lastName":"Chen","suffix":""},{"id":309377509,"identity":"51ba9f1f-c11d-4463-92bb-ef19c2050626","order_by":5,"name":"Xinyue Zhu","email":"","orcid":"","institution":"Tongji University","correspondingAuthor":false,"prefix":"","firstName":"Xinyue","middleName":"","lastName":"Zhu","suffix":""},{"id":309377510,"identity":"8e881f81-470a-4a61-9674-b857b3d4a436","order_by":6,"name":"Ruixin Zhu","email":"","orcid":"","institution":"Tongji University","correspondingAuthor":false,"prefix":"","firstName":"Ruixin","middleName":"","lastName":"Zhu","suffix":""},{"id":309377511,"identity":"0dd277d7-2ae8-4ed9-b701-817a8e592885","order_by":7,"name":"Lixin Zhu","email":"","orcid":"","institution":"Sun Yat-Sen University","correspondingAuthor":false,"prefix":"","firstName":"Lixin","middleName":"","lastName":"Zhu","suffix":""},{"id":309377512,"identity":"62368d3d-a5ed-4402-97a3-5fb3b5f33888","order_by":8,"name":"Na Jiao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2ElEQVRIie3OsQrCMBCA4RTBLlXXE/QdAoJFKj7LlYJZiqN0EMmUSfBtnCMBu0RdC13q4uSgi7gotpNbWzfBfEsucD8JIYbxm1ASMu7kZ3Fp1E6mXf5NUlBfJK4tT/IRHYHGwTkjkedzey9Lk9EKcbvSKVB9dinRzOfODEsTKhFlS6RLmuAQLKF8Dg4tT44Zbp/iADRhd7BedZIEUbWEzJMwf4XXSjJUPR1AV1/mgDs2EE5Y9bEwuF2iCbRjtoHrwuuvbV2eEOLgZy7GZsV+zpbVO4ZhGP/tDfnVS+J4sypPAAAAAElFTkSuQmCC","orcid":"","institution":"Fudan University","correspondingAuthor":true,"prefix":"","firstName":"Na","middleName":"","lastName":"Jiao","suffix":""}],"badges":[],"createdAt":"2024-05-11 10:24:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4404936/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4404936/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":58378187,"identity":"46e8b24d-2477-4c5c-b64b-2bfd7d04065a","added_by":"auto","created_at":"2024-06-14 15:53:34","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":178705,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStudy Design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study encompassed samples covering the full spectrum of NAFLD, including control, NAFL, borderline and NASH. Host gene profiles and paired hepatic microbial profiles of each dataset were extracted and annotated from RNA sequencing data of liver biopsies. Integrated analyses were conducted to establish a unified representation of features at both host gene and hepatic microbiota level. Subsequently, host-microbiota interactions were delineated from both overall sparsity level and functional cluster level across stages, aiming to understand their contribution to the hepatic microenvironment.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/a67d6c029826c148a2a4ceab.png"},{"id":58378183,"identity":"255bbfbc-2ce4-4195-a3fb-114aab26372f","added_by":"auto","created_at":"2024-06-14 15:53:34","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":18845629,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHost Gene Characteristics Associated with NAFLD Progression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea) \u003c/strong\u003eScatter plots illustrating the log2 Fold Change (FC) values of genes (y-axis) between three disease stages vs. CTRL (control) (x-axis), red points represent differentially expressed genes (DEGs) whose FDR \u0026lt; 0.05 and absolute log2FC \u0026gt; 1.2. \u003cstrong\u003eb)\u003c/strong\u003e Venn plot displaying the overlap of up-regulated DEGs between three comparisons. \u003cstrong\u003ec)\u003c/strong\u003e Venn plot displaying the overlap of down-regulated DEGs between three comparisons. \u003cstrong\u003ed)\u003c/strong\u003eEnrichment pathway plot showing the up-regulated pathways in the disease stage in each comparison with control, enriched by gene set enrichment analysis (GSEA). Each dot represents a pathway; the dot size correlates with the number of genes in the pathway. Similar pathways were grouped into clusters, each shown in the same background color. The most significant pathway in a cluster was chosen as the representative of that cluster. \u003cstrong\u003ee)\u003c/strong\u003e Enrichment pathway plot showing the down-regulated pathways in the disease stage in each comparison with control, enriched by GSEA.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/9f1c97d865e25dcfe88d976d.png"},{"id":58377898,"identity":"820ed374-8f46-44f5-aadf-8d5851e14038","added_by":"auto","created_at":"2024-06-14 15:45:35","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":585864,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHepatic Microbiota Characteristics Associated with NAFLD Progression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea)\u003c/strong\u003e Shannon diversity of hepatic microbiota in different stages at species level. P-values were assessed via Kruskal-Wallis test, with FDR values evaluated using post hoc Bonferroni correction (*FDR \u0026lt; 0.05). \u003cstrong\u003eb)\u003c/strong\u003ePrincipal coordinate plot illustrating the differences across stages based on Aitchison distance of batch effect-adjusted hepatic microbiota abundances at species level (p = 0.001). The P-value was calculated using permutational multivariate analysis of variance. The colors of the dots correspond to the different sample stages. \u003cstrong\u003ec)\u003c/strong\u003e Upset plot depicting the intersects of high abundant taxa in different stages. High abundant taxa were defined as species whose average abundance exceeded the geometric mean of samples in the same stage. 20 high abundant taxa were shared in four stages and the corresponding bar was highlighted in orange. The pie chart shows the relative abundance coverage of these 20 shared high-abundant taxa. \u003cstrong\u003ed)\u003c/strong\u003e Heatmap illustrating the centered log-ratio (CLR)-transformed abundance of these 20 shared high abundant taxa in four stages. Annotation bar on the left indicates kingdom to which each taxon belongs. Taxa marked with asterisk means a significant difference was observed in abundance between stages (FDR \u0026lt; 0.05), as determined using Kruskal-Wallis test and corrected with post hoc Bonferroni method. The background color of the taxon name corresponds to the specific phylum it belongs to. \u003cstrong\u003ee)\u003c/strong\u003e Heatmap with phylogenetic tree demonstrating differentially abundant taxa in each disease stages comparing to the control stage. Differentially abundant taxa were fined as taxa whose FDR \u0026lt; 0.05 and absolute log2FC \u0026gt; 1. The background color of the taxon name corresponds to the belonging phylum. The left column of annotation bar categorizes each taxon by kingdom and the right column denotes whether the taxon was previously identified as highly abundant. Cells marked with an asterisk indicate that the taxon was identified as differentially abundant in the corresponding comparison. Upward arrows indicate a consistently increasing trend with disease progression, while downward arrows indicate the opposite trend.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/082e0f2c19fdb6dbeebd61bb.png"},{"id":58377892,"identity":"149071ce-c250-4d95-9043-e6075eeb8e89","added_by":"auto","created_at":"2024-06-14 15:45:34","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1790111,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistinct Characteristics of Host Gene-Microbiota interactions across the NAFLD Spectrum\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea) \u003c/strong\u003eBar plot displaying the total number of significant and stable host-microbiota interactions in each stage. \u003cstrong\u003eb)\u003c/strong\u003e Bar plot indicating the number of genes involved in the interactions. \u003cstrong\u003ec)\u003c/strong\u003e Bar plot indicating the number of taxa involved in the host-microbiota interactions. \u003cstrong\u003ed)\u003c/strong\u003eVenn plot showing the overlap of shared interactions, which refers to the relationship between the same gene and taxon, in four stages. \u003cstrong\u003ee\u003c/strong\u003e) Ridge plot showing interaction strength of the positive host-microbiota interactions (left) and negative host-microbiota interactions(right) in each stage, measured by Spearman correlation coefficients. The central line represents the median, while the outer lines represent the lower and upper quartiles. \u003cstrong\u003ef\u003c/strong\u003e) Network plots illustrating the general host-microbiota interaction communities in each stage. Orange dots and green dots represent host genes and hepatic microbes involved in host-microbiota interactions, respectively. Node sizes are determined by node degree. The most gene-associated taxa were annotated.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/28e7971d5c701faab2bebf77.png"},{"id":58377896,"identity":"abffd2a0-e0c7-45b6-bf2e-ccde200646c6","added_by":"auto","created_at":"2024-06-14 15:45:34","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":609609,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFunctional Annotations of Host-Microbiota Interactions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea) \u003c/strong\u003eVenn plot illustrating enriched pathways among genes correlated with specific hepatic microbes in control (green), NAFL (red), and Borderline (purple) (FDR \u0026lt; 0.05). No significant pathways were enriched for NASH. \u003cstrong\u003eb\u003c/strong\u003e) Five pathways showing enrichment in both NAFL and borderline are emphasized with a blue background color. Control-specific pathways are shown with a green background, NAFL-specific pathways with a red background, and borderline-specific pathways with a purple background. Dot sizes represent the number of genes in the pathways and dot colors denote their significance after post hoc Bonferroni correction. \u003cstrong\u003ec-d) \u003c/strong\u003eAssociation between hepatic microbiota and host genes within the mitophagy pathway, shared in both NAFL \u003cstrong\u003e(c)\u003c/strong\u003eand borderline cohorts \u003cstrong\u003e(d)\u003c/strong\u003e. The size of squares and triangles represents the absolute value of sparse CCA coefficients of genes and microbes, respectively. Genes and taxa that appeared in both NAFL and borderline are shown in blue. Overlapping triangles indicate microbial taxa from the same taxonomic clade.\u003cstrong\u003e e-f) \u003c/strong\u003eAssociation between hepatic microbiota and host genes within the shared T cell receptor signaling pathway in NAFL \u003cstrong\u003e(e)\u003c/strong\u003eand borderline cohorts \u003cstrong\u003e(f)\u003c/strong\u003e.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/2fc4823e17f287b489b0b0ff.png"},{"id":58377903,"identity":"534ae1da-eba2-4b8c-8492-7ff2ba42375f","added_by":"auto","created_at":"2024-06-14 15:45:35","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":162497,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePotential mechanisms for interaction between hepatic microbiota and host in affecting NAFLD disease progression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHepatic microbes assist in enhancing the host mitophagy process, which further inhibits the xenophagy process to ensure their survival in cells. Such interaction is diminished in NASH patients. Impaired clearance of damaged mitochondria could instigate inflammation and further immune response or tissue damage by releasing toxic chemicals like reactive oxygen species (ROS).\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/4b2d673e14792d1234dcd776.png"},{"id":58970188,"identity":"2dee2088-7b64-4f74-bed9-09c3b9f1c088","added_by":"auto","created_at":"2024-06-24 20:19:23","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":40200834,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/1dbc22ff-3f8a-4744-a1b1-0d6899b8605b.pdf"},{"id":58378774,"identity":"b76c8084-1813-4ba8-897d-f060434e0c65","added_by":"auto","created_at":"2024-06-14 16:01:34","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2989045,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/43eb180b17560eb5763b5c03.pdf"},{"id":58378182,"identity":"b3f44381-72cc-45fe-b659-62889ac35bcd","added_by":"auto","created_at":"2024-06-14 15:53:34","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":142509,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource11.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/5014a4ffc48d53dcfe895d57.pdf"},{"id":58377888,"identity":"48b6a8dd-7aaf-4ea5-82f3-ffb67306db3e","added_by":"auto","created_at":"2024-06-14 15:45:34","extension":"pdf","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":183235,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource12.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/b6978730bd760d46935bc624.pdf"},{"id":58377887,"identity":"c992e99b-27b0-4992-b975-444b0764c69c","added_by":"auto","created_at":"2024-06-14 15:45:34","extension":"pdf","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":37103,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource13.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/dcba65cb925f256eefd73ea3.pdf"},{"id":58378185,"identity":"23942029-f4a5-4311-98d0-a81d76173a8a","added_by":"auto","created_at":"2024-06-14 15:53:34","extension":"pdf","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":30377,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource14.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/a8b1b27cb42b7716e1556a1c.pdf"},{"id":58377890,"identity":"343aa28c-16c6-4ce3-987a-713143f29606","added_by":"auto","created_at":"2024-06-14 15:45:34","extension":"pdf","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":42059,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource15.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/0757bd4f8a3ec7541a767267.pdf"},{"id":58378184,"identity":"310e2e45-4b05-4a92-8f53-9c9c172a56ac","added_by":"auto","created_at":"2024-06-14 15:53:34","extension":"pdf","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":42641,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource16.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/23b384f4bdfd92980251d8d6.pdf"},{"id":58377905,"identity":"e1c20600-525b-4b8b-b4c5-a261320ce8c7","added_by":"auto","created_at":"2024-06-14 15:45:36","extension":"pdf","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":717657,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource17.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/a8b579966516bcd6aca13c28.pdf"},{"id":58377902,"identity":"24ae70fe-03ea-4911-843a-298080a9e80e","added_by":"auto","created_at":"2024-06-14 15:45:35","extension":"pdf","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":261247,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource18.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/e23ebe7ac56fe96695601228.pdf"},{"id":58377900,"identity":"3cd4b39c-adcf-4db3-a20c-38ee51bef17d","added_by":"auto","created_at":"2024-06-14 15:45:35","extension":"pdf","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":212878,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource19.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/0b732c792d6792dfdc7c4250.pdf"},{"id":58377899,"identity":"3bde91bd-7917-47da-b1df-ffbe31b71c37","added_by":"auto","created_at":"2024-06-14 15:45:35","extension":"pdf","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":388652,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource20.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/1c1344f555aba93fa336ea67.pdf"},{"id":58377901,"identity":"fdc319ed-eb2d-4f7b-a40c-a65e6531d0c8","added_by":"auto","created_at":"2024-06-14 15:45:35","extension":"pdf","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":129513,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineResource21.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/16822a909f8c24a0bf25e18c.pdf"},{"id":58377904,"identity":"95f4482f-ba6a-4dbd-aa0d-9ba785429e8e","added_by":"auto","created_at":"2024-06-14 15:45:35","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"supplement","size":888092,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGraphic Abstract\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study encompassed 570 liver biopsy transcriptomes from five cohorts, including 72 control, 124 nonalcoholic fatty liver (NAFL), 143 borderline and 231 nonalcoholic steatohepatitis (NASH) samples. Hepatic microenvironment during disease progression was investigated from aspects including host transcriptome, hepatic microbiome and host-microbiota interactions. Notably, bacteria including Bacillales, \u003cem\u003eRalstonia insidiosa\u003c/em\u003e, and Micromonosporaceae were observed a vital role in enhancing host mitophagy by interacting with genes including \u003cem\u003eSQSTM1\u003c/em\u003e, \u003cem\u003eOPTN\u003c/em\u003e, and \u003cem\u003eBNIP3L\u003c/em\u003e. However, the absent of such interaction functional clusters might lead to a pro-inflammatory hepatic microenvironment through the activation of immune reactions, which potentially drove disease progression to NASH.\u003c/p\u003e","description":"","filename":"Graphicabstract.png","url":"https://assets-eu.researchsquare.com/files/rs-4404936/v1/b8fa65b1d532be26942c144e.png"}],"financialInterests":"","formattedTitle":"Disrupted Host-Microbiota Crosstalk Promotes Nonalcoholic Fatty Liver Disease Progression by Impaired Mitophagy","fulltext":[{"header":"Introduction","content":"\u003cp\u003eNonalcoholic fatty liver disease (NAFLD) is a chronic liver disorder, affects an estimated 25% of the population and is increasingly recognized as a leading cause of liver-related mortality. NAFLD encompasses a spectrum of stages. Nonalcoholic fatty liver (NAFL), the initial period, is characterized by excessive hepatocyte triglyceride accumulation. Following a transitional stage termed \u0026ldquo;borderline\u0026rdquo;, the disease progresses into nonalcoholic steatohepatitis (NASH), marked by irreversible liver injury and severe hepatic inflammation. Alarmingly, some NASH patients are at risk of cirrhosis and hepatocellular carcinoma. In addition to local inflammation, systemic inflammation is also a typical characteristic in NAFLD patients, predisposing patients to various extrahepatic manifestations such as obesity, type 2 diabetes, and cardiovascular diseases.\u003c/p\u003e \u003cp\u003eMultiple factors contribute to the pathogenesis of NAFLD, including genetic abnormalities, metabolic dysregulation, and gut microbiota dysbiosis.\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e Therefore, it is essential to consider the interplay between these elements to gain a more holistic understanding of the disease mechanisms. \u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e For example, an integrated analysis that combined hepatic transcriptome, metabolome, and gut metagenome facilitated the depiction of coordinated disruption of gut-liver axis and causal mechanisms exploration in hepatic steatosis.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e As gut-liver axis suggested, certain proinflammatory agents such as microbes were able to mitigated from gut to liver.\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e Using sequencing, qPCR, and FISH targeting 16S rRNA, studies have evidenced the presence of microbiota within the liver.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan additionalcitationids=\"CR9 CR10\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e Albeit at a lower abundance than in the gut, these hepatic microbes provide crucial insights into the liver microenvironment since they reside directly within the liver.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e For example, these hepatic bacteria, detecting by 16S rRNA sequencing, were shown a close association with liver phenomics including ballooning degeneration, inflammation and fibrosis.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e Furthermore, these hepatic microbes were also active participators in various cellular process through interaction with host,\u003csup\u003e\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e such as modulating host natural killer T cells programming to induce inflammation in both mice and humans.\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e Notably, the direct connection has been found in hepatic microbes with NAFLD/NASH risk alleles. For example, Enterobacter and Pseudoalteromonas were strongly associated with \u003cem\u003ePNPLA3\u003c/em\u003e rs738409 and \u003cem\u003eTM6SF2\u003c/em\u003e rs58542926 variants, suggesting the potential of hepatic microbes in modulating disease risk.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e These exciting observations prompt us to delve into the hepatic microbes and their interaction with host, to gain a comprehensive understanding on the dynamics of tissue microenvironment.\u003c/p\u003e \u003cp\u003eGiven the complex, \u0026ldquo;multi-hit\u0026rdquo; nature of NAFLD pathogenesis, it is of significance to conduct a multi-dimensional analysis that integrates the entirety of all relevant genes, hepatic microbes, and their interactions. Moreover, a general pattern of NAFLD evolvement highlights the necessity to consider every progressive stage, as this approach not only facilitate the identification of features specific to each stage, but also aids in the exploration of potential mechanisms driving disease worsening. However, a comprehensive and systematic evaluation that simultaneously considers multi-dimensional perspectives across the whole progressive stages of NAFLD remains absent. Therefore, this study aims to fill this gap by evaluating the hepatic microenvironment at different stages of NAFLD. By understanding the distinct gene and microbial profiles, we aim to illuminate the complex pathogenesis of NAFLD progression and identify pivotal mechanisms driven by host-microbiota interactions.\u003c/p\u003e \u003cp\u003eTo achieve this goal, we conducted a comprehensive analysis of hepatic microenvironment in whole progressive stages including control, NAFL, borderline ang NASH. In addition, we integrated five cohorts from different regions to gain robust insights into the composition and interaction of host genes and hepatic microbes.\u003csup\u003e\u003cspan additionalcitationids=\"CR13 CR14 CR15\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e Our findings showed host-microbiota interactions in worse stages of NAFLD were disrupted, especially those related to mitophagy progress, which might contribute to disease progression. Collectively, focusing on holistic view of gene, microbes and host-microbiota interactions, this study introduced a novel perspective in understanding the dynamics of microenvironment during the progressive stages of NAFLD.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eSample collection\u003c/h2\u003e \u003cp\u003eWe scoured public databases, including GEO, BioProject and ArrayExpress, for RNA sequencing data of liver biopsies from healthy or NAFLD stages (Online Resource 1). Written informed consent was obtained from all individuals.\u003csup\u003e\u003cspan additionalcitationids=\"CR13 CR14 CR15\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e Healthy liver biopsies were derived from people undergoing bariatric surgery. Seven studies with samples without a history of alcohol intake or viral infection were collected. To minimize the influence caused by different diagnostic standards, the disease stages of patients from three studies (PRJNA558102, PRJNA767535 and PRJNA682622) were reclassified uniformly according to the criteria of the NASH-Clinical Research Network system: samples with NAFLD Activity Score (NAS) scores\u0026thinsp;\u0026lt;\u0026thinsp;3 were defined as NAFL, samples with NAS scores\u0026thinsp;\u0026ge;\u0026thinsp;5 were defined as NASH and the remaining were classified as borderline.\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e Samples from another two studies (PRJNA704861 and PRJNA523510) retained disease classification provided by the original publications (Online Resource 1).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eHepatic gene expression preprocessing\u003c/h2\u003e \u003cp\u003eRaw sequencing data were downloaded and uniformly reprocessed to gene expression profiles with following steps. First, adapter removal and quality control were conducted by Trimmomatic (version 0.39) and the qualified reads were aligned to the newest human reference genome\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e (CHM13 v2.0) with STAR (version 2.7.10a). Focusing on genes encoding for proteins, we excluded other gene types such as micro-RNA using R package \u0026ldquo;biomaRt\u0026rdquo; (version 2.52.0). Subsequently, variance stabilizing transformation was performed using \u0026ldquo;DESeq2\u0026rdquo; (version 1.28.3), followed by filtering out less informative genes which variance below the 25th percentile cross all the samples.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eHepatic microbial composition preprocessing\u003c/h2\u003e \u003cp\u003eSince microbes can also be captured by transcriptomics sequencing, we refined the workflow to detect the hepatic microbial sequences within liver biopsy RNA-seq data.\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e To mitigate false positives and potential contaminants, a rigorous decontamination process was conducted (Online Resource 2). First, to eliminate the distortion by a potentially high proportion of human sequences on microbial read counts, we employed Kneaddata (version 0.6.1) to remove the sequences aligned to CHM13v2 human genome (Module 1, Step 1 in Online Resource 2). The unaligned reads were regarded as candidate microbial sequences, subsequently annotated using KrakenUniq (version 1.0.4). This annotation enables precise differentiation between genuine microbial signals and false positives arising from highly repetitive or contaminated sequences\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e (Module 1, Step 2 in Online Resource 2). A customized database for KrakenUniq comprises the CHM13v2 human genome and complete microbial genomes downloaded from RefSeq (retrieved 5th September 2023), which includes 35,022 genomes from 8,907 bacterial species, 14,993 genomes from 13,986 viral species, 35 genomes from 35 fungal species, and 540 genomes from 423 archaeal species. Additionally, artificial sequences such as UniVec and EmVec are also included in the database. Bracken (version 2.5.0) was used subsequently to quantify the abundances. Since k-mers from a true organism will distribute evenly across the genome, the taxa with low unique k-mer counts were considered contaminations, using read length minus k-mer length\u0026thinsp;+\u0026thinsp;1 as threshold (Unique k-mer filtering: Module 2, Step 3 in Online Resource 2). Then, taxa with total k-mer counts exceeding read counts by more than fivefold were retained (Total k-mer filtering: Module 2, Step 4 in Online Resource 2). For each taxon, the relationships between total read counts, unique k-mer counts and total k-mer counts were examined and those with significant Spearman correlations were considered true microbial sequences\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e (Module 2, Step 5 in Online Resource 2). Meanwhile, for those contaminations which are known commonly during experiments, we adopted a reference-based decontamination process. Taxa whose abundance resembled a distribution with the cell line data were considered experimental contaminations and were excluded from further analyses \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e (Module 2, Step 6 in Online Resource 2). The remaining taxa were compared with a published common reagent contaminations list\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e and all matches were removed (Module 2, Step 7 in Online Resource 2). As last, taxa with average relative abundance above 0.01% and present in at least 10% of the samples were retained for further analyses (Module 3, Step 8 \u0026amp; 9 in Online Resource 2).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eInteraction analysis between hepatic genes and microbes\u003c/h2\u003e \u003cp\u003eLeast Absolute Shrinkage and Selection Operator (Lasso) penalized regression was used to identify the sparse interactions between host genes and hepatic microbes. To capture interactions at all taxonomic level, we combined abundance data at different taxonomic levels into a combined taxa matrix. For a given host gene, we detected the taxa that most accurately predict its expression pattern across samples within each stage. Subsequently, the combination of the host gene and the predicting taxa was considered as a host-microbiota interaction. R package \u0026ldquo;glmnet\u0026rdquo; was implemented for lasso analysis (version 4.1-7), with the optimal tuning parameter λ determined through leave-one-out cross-validation. Regression model significance was gauged using \u0026ldquo;hdi\u0026rdquo; (version 0.1-9), with interactions having a false discovery rate (FDR) less than 0.01 post Benjamini-Hochberg correction as significant. For stability assurance, the lasso model was iteratively refitted (100 times) to random sample subsets with perturbed λ. Interactions appearing in at least 60% iterations were considered stable. Interaction strength was quantified through Spearman correlation via \u0026ldquo;stats\u0026rdquo; (version 4.2.1). Interaction visualization and degree of the nodes was computed in Gephi.\u003c/p\u003e \u003cp\u003eSparse Canonical Correlation Analysis (SparseCCA) was utilized to identify the functional interaction clusters by projecting host genes and hepatic microbial data into a shared latent space, while maximizing the correlation between these two datasets at the same time. Combined taxa matrix at all taxonomic levels and the gene matrix were the input for SparseCCA in each stage. L1 penalty was incorporated to select the most associated host genes and microbes as a group. In each of the four stages, ten most associated groups were identified separately, and the significance of each group was computed using the leave-one-out cross-validation approach followed by post hoc Bonferroni correction. Such process was implemented using R package \u0026ldquo;PMA\u0026rdquo; (version 1.1). To explore the specific function of each group, we performed the KEGG pathway enrichment for the genes involved in each group using \u0026ldquo;enrichKEGG\u0026rdquo; function in R package \u0026ldquo;clusterProfiler\u0026rdquo;. Pathways were considered significant if they had a corrected FDR of less than 0.05.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eEach profile of single dataset was combined into a unified one using \u0026ldquo;merge\u0026rdquo; function in R. Permutational Multivariate Analysis of Variance (PERMANOVA) was used to assess the general variance caused by batch effect variable \u0026ldquo;Study\u0026rdquo; and true biological variable \u0026ldquo;Stage\u0026rdquo;, whose specific influence on the main variance was analyzed using Kruskal-Wallis rank sum tests on the first two principal coordinates of Principal Coordinate Analyses. Batch effect adjustment was performed using \u0026ldquo;RemoveBatchEffect\u0026rdquo; function in limma R package (version 3.54.2) on the variance stabilized gene expression data and centered log ratio transformed microbial abundance data.\u003c/p\u003e \u003cp\u003eDifferentially expressed genes (DEG) and microbes were detected using limma R package. Multiple hypothesis correction was performed using the post hoc Bonferroni method. Differentially expressed genes were defined as those with absolute log fold change greater than 1.2 and FDR less than 0.05, and differentially abundant microbes refer to those with absolute log fold change more than 1 and FDR less than 0.05. Pathway enrichment for host genes was performed by gene set enrichment analysis (GSEA) using clusterProfiler R package (version 4.4.4). Significantly enriched pathways were defined as those with FDR less than 0.001.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e1. Cohort collection and study design\u003c/h2\u003e \u003cp\u003eTo investigate the hepatic microenvironment throughout the progression of NAFLD, we curated publicly available RNA sequencing samples from various data repositories. Five datasets, adhering to rigorous selection criteria, were integrated into this study. Samples involved were sourced from various countries and continents, including Japan, the United States, Denmark, and other countries in the European region. A total of 570 samples were enrolled, which encompassed 72 control samples, 124 NAFL patients, 143 borderline patients and 231 NASH patients (Online Resource 1). Here, the \u0026ldquo;borderline\u0026rdquo; stage was considered as a transitional state between NAFL and NASH, characterized by an intermediate NAS score of 3 and 4.\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e Detailed dataset information and sample demographic characteristics were summarized in Online Resource 9.\u003c/p\u003e \u003cp\u003eAfter batch effect removal and rigorous decontamination processes (Online Resource 2\u0026ndash;5), we obtained hepatic gene expression (14,414 genes) and microbial abundance (350 taxa) profiles for these samples. Subsequently, we delineated specific patterns in hepatic gene expression and microbial abundance. Underlying mechanisms of this dynamic niche were further elucidated by investigating the intricate interactions between the host and microbiota (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2. Altered gene expression suggested disturbed energy metabolism and active immune response in NAFLD\u003c/h2\u003e \u003cp\u003eWe firstly investigated hepatic genes associated with NAFLD with a DEG analysis. Taking control as the baseline, we identified 1639 DEGs in NAFL, 1802 DEGs in borderline and 2407 DEGs in NASH (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea \u0026amp; Online Resource 10\u0026ndash;12). Among these, we found a consistent pattern in the most dramatically changed DEGs. For example, \u003cem\u003eIL32\u003c/em\u003e, \u003cem\u003eAKR1B10\u003c/em\u003e and \u003cem\u003eFABP4\u003c/em\u003e all increased progressively with the disease progressing, while \u003cem\u003eEGR1\u003c/em\u003e, \u003cem\u003eCYP2C19\u003c/em\u003e and \u003cem\u003eVIL1\u003c/em\u003e exhibited the opposite (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Besides, a total of 358 DEGs were consistently up-regulated and 536 DEGs were uniformly down-regulated in all three disease stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb-c).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eGSEA revealed pathway modulations in each disease stage compared to the control: in NAFL, 23 pathways were activated and 34 pathways were inhibited (Online Resource 6a \u0026amp; Online Resource 13); in the borderline stage, 25 pathways were activated and 5 pathways were inhibited (Online Resource 6d \u0026amp; Online Resource 14); in NASH, 61 pathways were activated and 16 pathways were inhibited (Online Resource 6g \u0026amp; Online Resource 15). Stage-specific and shared pathways revealed that both discrepancies and similarities existed along disease progression. In detail, earlier stage NAFL was specifically featured in the activation in ATP synthesis (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed) and oxidative phosphorylation (Online Resource 6b), and inhibition in organ development related pathway (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee \u0026amp; Online Resource 6c). Comparatively, borderline and NASH exhibited greater similarities, both showed the activation in extracellular matrix organization related pathway (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed \u0026amp; Online Resource 6h) and immune activities such as antigen processing and representation (Online Resource 6e \u0026amp; Online Resource 5g). In addition, consistent inhibition in xenobiotic catabolic process (Online Resource 6f \u0026amp; Online Resource 6g) and response too zinc ion (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee \u0026amp; Online Resource 6i) also suggested the shared abnormities between borderline and NASH. Notably, mitochondrion organization and ribosome biogenesis were activated across all disease stages compared to the control, suggesting that these cellular processes may play a persistent role during the whole process (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3. Turbulent hepatic microbiota involved in the hepatic microenvironment in NAFLD\u003c/h2\u003e \u003cp\u003eThe hepatic microbial landscape was comprehensively depicted by annotating multi-kingdom microbial reads of RNA-seq data, followed by a series of decontamination processes to remove human sequences, false positives during annotation, and common laboratory contaminations (Online Resource 3a-b). A total of 350 taxa ranging from domain to species were identified, covering bacteria (296 taxa), fungi (46 taxa), viruses (7 taxa) and archaea (1 taxa). 14 bacterial phyla were identified, with Proteobacteria, Actinobacteria and Firmicutes being predominant (Online Resource 7), indicating a distinct microbial profile compared to fecal microbiota which was dominated by Bacteroidetes.\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e Meanwhile, Basidiomycota and Ascomycota of fungi, and Artverviricota and Uroviricota of viruses were also observed presence (Online Resource 7).\u003c/p\u003e \u003cp\u003eA turbulent microenvironment featured by dynamic hepatic microbiota composition was observed in NAFLD. First, the composition at the phylum level showed notable variations. Actinobacteria and Fusobacteriota were decrease while Firmicutes and Ascomycota were increase in NAFLD (Online Resource 7). In addition, a significantly higher alpha diversity distinguished NAFL with other stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). Different composition among stages were further elucidated with a significant difference in beta diversity analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb). An in-depth evaluation was further performed to assess the microbiota composition with higher resolution at species level. First, high-abundance species were analyzed, as their dominant abundance could potentially account for important effects. Here we defined high-abundance species as the taxa with abundance exceed the average of all samples were (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). 37 high-abundance species were found totally, and 20 of them were showed in all four stages, accounting for 77.54% of the relative abundance (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). Moreover, 15 of these high-abundance species exhibited differential abundances among stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). Within these species, \u003cem\u003eCutibacterium acnes\u003c/em\u003e was the most abundant bacterium (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). Comparatively, as the most abundant fungus, \u003cem\u003eMalassezia restricta\u003c/em\u003e exhibited a stable abundance across the four stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAdditionally, to understand the potential role of hepatic microbiota associated with NAFLD, we analyzed the differential abundant taxa at species level between each disease stage and the control stage. As a result, 18 species in NAFL, 12 in borderline, and 17 in NASH displayed significant differences in abundance (Online Resource 16). Among these differential species, eight bacterial species were observed significant differences in all of the disease stages compared to the control. Two species of them, \u003cem\u003eMethyloversatilis sp. RAC08\u003c/em\u003e and \u003cem\u003eRalstonia insidiosa\u003c/em\u003e, were previously identified as high abundance species and both showed decrease in all of the disease stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4. Disrupted hepatic microenvironment characterized by impaired host-microbiota interactions in NAFLD patients\u003c/h2\u003e \u003cp\u003eOne of the key objectives of this study was to explore the hepatic microenvironment and its relationship with the NAFLD progression, integrating both host genes and hepatic microbiota. To achieve this goal, we used LASSO regression to identify the interaction between a specific gene and a microbiota (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). In the control stage, we identified 5,537 significant and stable host-microbiota interactions (Online Resource 17). Remarkably, these interactions were reduced to 1,937 in NAFL (Online Resource 18), further declined to 1,455 in the borderline stage (Online Resource 19), and 2,933 in the NASH stage (Online Resource 20). The number of involved genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb) and taxa (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec) exhibited a similar pattern, characterized by an overall decline with a marginal resurgence. Notably, we found no identical interactions appeared in four stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed), but similarities were exhibited among three disease stages. For example, an identical relationship between genus \u003cem\u003eMicromonospora\u003c/em\u003e and \u003cem\u003eFGF21\u003c/em\u003e was observed in all of the three disease stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed). In addition, 487 identical interactions were present in both NAFL and NASH, with an additional 12 interactions appeared in both borderline and NASH, and 5 interactions appearing in both NAFL and borderline (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAn overall decreasing strength was exhibited both positive and negative interactions as the disease progresses (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee). NAFL and NASH exhibited a higher degree of similarity, while the borderline stage demonstrated modest fluctuations. This may reflect the inherent instability of the transitional borderline stage. Despite the modest fluctuation, the decline was particularly dramatic when comparing the interaction strength of NASH to that of borderline (FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.0001 with Kruskal-Wallis rank sum tests for both positive and negative interactions), indicating an weak communication network between host and microbiota in advance stage NASH. To explore the interaction network more deeply, we examined the dominant interacted microbiota in every stage. Results showed that the host-microbiota interaction network were distinct in control but were analogous in three disease stages. Specifically, in the control liver microenvironment, \u003cem\u003eEscherichia coli\u003c/em\u003e, \u003cem\u003eActinomyces naeslundii\u003c/em\u003e and \u003cem\u003ePseudomonas oleovorans\u003c/em\u003e were the most associated taxa (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef); however, interactions with these taxa were nearly absent in disease stages. On the contrary, the three disease stages exhibited a consistent pattern. The order Sphingomonadales and the family Sphingomonadaceae emerged as the most interactive taxa in NAFL and borderline stages, while \u003cem\u003eRalstonia insidiosa\u003c/em\u003e was identified as the taxon engaging the most extensive gene interactions in NAFL and NASH (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef). The reduced participation, weaker interplay and varied interaction network collectively suggested a significantly dysregulated and vulnerable hepatic microenvironment as the disease advances.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e5. Hepatic microbiota was associated with host mitophagy and immune response in NAFLD\u003c/h2\u003e \u003cp\u003eTo gain a deeper understanding of the potential pathogenetic mechanisms driving distinct patterns of host-microbiota interactions, we identified interaction clusters of host genes and microbiota. The biological processes indicated by the genes and taxa in each cluster shed light on the coordinate functions of these interactions. Our analysis revealed 145, 23 and 101 pathways in the control, NAFL, and borderline, respectively. However, no significant pathways were detected for NASH (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea \u0026amp; Online Resource 21), which is possibly attributed to the disturbed interactions precluding the execution of a consistent function in NASH. Five stage-shared pathways in NAFL and borderline were identified (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). Subsequently, 89 stage-specific pathways were revealed, including 66, 3 and 20 pathways for control, NAFL and borderline, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). Three of each stage-specific pathways were highlighted (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb) as they were of importance in fundamental cellular processes. Here, control-specific host pathways included autophagy and biosynthesis of unsaturated fatty acids, while NAFL-exclusive associations were enriched for cell signal pathways and ferroptosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). Borderline-specific pathways included the mTOR signaling pathway, PI3K-Akt pathway and antigen processing and presentation pathway (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eMitophagy was especially focused on due to the essential role of mitochondria in energy production and close association with NAFLD.\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e Our results showed that Bacillales, \u003cem\u003eRalstonia insidiosa\u003c/em\u003e and Micromonosporaceae, two dominant taxa in the previous interaction network (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef), were involved in the host mitophagy process in both NAFL (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec) and borderline stages (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). Relevant genes exhibited similarity as they both appeared in interaction clusters in both stages, including \u003cem\u003eTBC1D17, HRAS.\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec \u0026amp; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). To investigate the impact of these interactions on mitophagy, we examined the expression of related genes. As expected, with interactions absent in NASH, the up-regulated trend of related genes was relatively modest in NASH. For example, \u003cem\u003eSQSTM1\u003c/em\u003e was observed an insignificant variance in NASH compared to the borderline (Online Resource 8a). Similar pattern was recognized for \u003cem\u003eOPTN\u003c/em\u003e and \u003cem\u003eBNIP3L\u003c/em\u003e, encoding for other cargo receptors (Online Resource 8b-c). The mitochondrial kinase, \u003cem\u003ePINK1\u003c/em\u003e, was also up-regulated in NAFLD patients (Online Resource 8d). Intriguingly, \u003cem\u003eSTING\u003c/em\u003e, a cytosolic sensor involved in the xenophagy process \u0026mdash; a process akin to mitophagy but targeting intracellular bacteria instead \u0026mdash; showed an overall rising expression under disease liver stages (Online Resource 8e).\u003c/p\u003e \u003cp\u003eIn addition, T cell receptor signaling pathway was emphasized to explore the role of host-microbiota interactions on microenvironment inflammation during disease progression. As a result, genes including \u003cem\u003ePTPN11\u003c/em\u003e, \u003cem\u003ePIK3CB\u003c/em\u003e and \u003cem\u003eMAP2K7\u003c/em\u003e were found to interact with hepatic microbiota in both NAFL and borderline (Online Resource 8f-h). These taxa included Bacilli, Sphingomonadales and \u003cem\u003eRalstonia insidiosa\u003c/em\u003e, were also the most gene-associated taxa in NAFL and borderline. These findings echoed the up-regulated antigen processing pathway in advanced stages such as borderline and NASH (Online Resource 6e). Genes in these interactions showed a similar up-regulation (Online Resource 8i-k), suggesting that these interactions could result in an immune-activated microenvironment.\u003c/p\u003e \u003cp\u003eAltogether, these abnormities in host-microbiota interactions and related gene expressions inspired the hypothesis that hepatic microbiota may enhance the mitophagy process as a way of facilitating their survival in the liver. Should these interactions be disrupted, the resulting impaired clearance of dysfunctional mitochondria could release reactive oxygen species (ROS), which further leads to tissue damage and inflammation that drive disease progression (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Consequently, diminished mitophagy in NASH impair xenophagy inhibition, enabling more hepatic microbes to be degraded by host (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eTo explore potential mechanisms associated with NAFLD development and progression, this study encompassed the entire spectrum of NAFLD, from control, NAFL, the transient stage borderline, to the advanced stage NASH. In addition, multiple cohorts from various geographical regions were involved, allowing a more robust finding. Our study highlighted that, the previously overlooked hepatic microbiota was an indispensable constituent of the hepatic microenvironment. Most importantly, the intricate interplay between these microbiota and host provided a novel perspective to understand the role of microenvironment during NAFLD progression.\u003c/p\u003e \u003cp\u003eBy extracting microbial RNA fragments with rigorous decontamination, our study was the first to comprehensively investigate the active presence of hepatic microbes, which we found to be distinctly different from fecal microbes composition.\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e Bacteroidetes and Firmicutes were the most abundant phyla in fecal of NAFLD patients,\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e while Proteobacteria and Actinobacteria were predominant in liver of NAFLD. The disparities underscored the significance of exploring the tissue microbial signatures in the hepatic microenvironment. Interestingly, we also identified hepatic RNA fragments belonging to fungi. These fungi showed the underestimated diversity of liver microbiota, a field previously concentrated predominantly on bacterial components.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e However, the potential origin of the hepatic microbiome necessitates further investigation. Specifically, we identified that \u003cem\u003eCutibacterium acnes\u003c/em\u003e, was predominant in the liver tissue and showed a decline in disease individuals. This species is known as a commensal bacterium, producing short-chain fatty acids to help maintain host homeostasis. We also found that a fungus species, \u003cem\u003eMalassezia restricta\u003c/em\u003e, adapted well to the hepatic environment, possibly because a rich influx of lipids in liver satisfies its lipid-dependent growth. Overall, the diversity and dynamic abundance of hepatic taxa, not replicable by gut microbiota, highlight their previously underestimated but critical role in the hepatic microenvironment.\u003c/p\u003e \u003cp\u003eA key contribution of this study is our approach to investigate about the hepatic microenvironment, which spans the whole spectrum of NAFLD stages including control, NAFL, borderline to NASH. This granularity holds clinical significance, as stage-specific features imply potential mechanisms for disease progression and may of therapeutic value.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e Microenvironment landscape in each stage was first delineated with complementary results from host gene and hepatic microbiota compositions. Here, the turbulence in hepatic microbiota indicated a general dysbiosis of hepatic microenvironment, while abnormalities in host genes provide more specific insights in each disease stage. For example, NAFL was featured by abnormal energy metabolism while borderline and NASH exhibited activation of immune activities. These observations were also confirmed by other researches.\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e Most importantly, the analysis on interactions of host and microbes facilitated an in-depth examination of network and function within the hepatic microenvironment. The findings from this approach created a repository of potential mechanisms for disease progression, thereby extending the scope of conventional studies that typically focused on a single aspect. In general, our results showed the worse stages of NAFLD were featured by a sparser and weaker host-microbiota interaction network. This weaking was accompanied by a consistent disappearance of functional responses, which eventually loss in NASH. Take a closer examination, the hepatic microbiota Sphingomonadaceae family and the corresponding order Sphingomonadales maintained interactive status across NAFL and borderline. While these taxa have been experimentally detected in NAFLD livers by others studies,\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e their specific influence in NAFLD disease has not been extensively explored. Our studies suggested that these taxa might contribute to hepatic inflammation by affecting host T cell receptor pathway. These interactions were possibly mediated by glycosphingolipid, a metabolite secreted by Sphingomonadaceae and Sphingomonadales. This aligned with recent findings that intrahepatic Bacteroidetes phyla could modulate host T cells by glycosphingolipids.\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e Therefore, our study suggested that Sphingomonadaceae and Sphingomonadales may possess a more targeted therapeutic potential in NAFLD, surpassing the broader implications previously identified at the phylum level.\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e Furthermore, other microbial interactions that specifically appeared in the control stage may also prevent steatosis by aiding in the maintenance of tissue homeostasis. For instance, taxa such as \u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e and Granulicatella were observed interacting with genes including \u003cem\u003eHACD3\u003c/em\u003e. These interactions could facilitate the host\u0026rsquo;s biosynthesis of unsaturated fatty acids, a process considered beneficial for liver health.\u003c/p\u003e \u003cp\u003eNovel mechanisms was also revealed driving by host-microbiota interactions, highlighting the importance of this multifaceted approach assessing the hepatic microenvironment. Dysfunctional mitochondrial removal is evidenced in NAFLD and enhanced mitophagy rescues liver damage.\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e However, the exact internal biological dysfunctions that induce compromised mitophagy in NAFLD remains elusive.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e Taking advantage of interaction analysis, we suggested that impaired mitophagy could attribute to the loss of certain host-microbiota interactions in NASH. The disruption of these interaction was possibly associated with the decline of core microbes including Bacillales, \u003cem\u003eRalstonia insidiosa\u003c/em\u003e and Micromonosporaceae, which were experimentally evidenced the ability to amplify host mitophagy.\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e Consequent accumulation of damaged mitochondria may produce reactive oxygen species (ROS), thereby exacerbate tissue inflammation. Such a pro-inflammatory microenvironment was also confirmed by the activation of immune response related genes and pathways. Taking together, investigating the host-microbiota interaction-driven microenvironment offers a broader and deeper view of the pathogenic mechanisms underlying NAFLD, surpassing the insights gained from transcriptomics analysis alone.\u003c/p\u003e \u003cp\u003eA common challenge when analyze sequencing data is to obtain high repeatable results, with variabilities between different experimental batch. These variabilities if unrelated to the true biological variable, was considered batch effect. To gain a genuine and robust understanding of microenvironment components, our study was based on five cohorts with a rigorous consideration and correction of batch effect.\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e Our results showed a great concordance with previous findings, affirming their reliability and repeatability. For example, gene \u003cem\u003eIL32\u003c/em\u003e exhibited the most dramatically upregulated expression in all of the NAFLD stages, and \u003cem\u003eCYP2C19\u003c/em\u003e showed the opposite trend, these were both confirmed in existing findings.\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e With regard to hepatic microbiota, our results suggested Proteobacteria was the most abundant phyla, which was evidenced in previous experiments using 16S RNA and qPCR.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eIn summary, this study has comprehensively demonstrated the microenvironment across the whole progression of NAFLD, using an integrated analysis across multiple cohorts. Our results suggested disruption of host-microbiota interactions in NASH, especially those who are vital for mitophagy, implied novel mechanisms in disease stage transition. Collectively, our results shed light on a new perspective in the multifaceted mechanisms driving NAFLD progression, providing insights for potential therapeutic targets and intervention strategies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cstrong\u003eDEG\u003c/strong\u003e, Differentially Expressed Genes;\u003cstrong\u003e\u0026nbsp;FC\u003c/strong\u003e, Fold Change;\u003cstrong\u003e\u0026nbsp;FDR\u003c/strong\u003e, False Discovery Rate; \u003cstrong\u003eGSEA\u003c/strong\u003e, Gene Set Enrichment Analysis; \u003cstrong\u003eLasso\u003c/strong\u003e, Least Absolute Shrinkage and Selection Operator; \u003cstrong\u003eNAFL\u003c/strong\u003e, Nonalcoholic Fatty Liver; \u003cstrong\u003eNAFLD\u003c/strong\u003e, Nonalcoholic Fatty Liver Disease; \u003cstrong\u003eNAS\u003c/strong\u003e, NAFLD Activity Score; \u003cstrong\u003eNASH\u003c/strong\u003e, Nonalcoholic Steatohepatitis; \u003cstrong\u003ePERMANOVA\u003c/strong\u003e, Permutational Multivariate Analysis of Variance; \u003cstrong\u003eROS\u003c/strong\u003e, Reactive Oxygen Species; \u003cstrong\u003eSparseCCA,\u0026nbsp;\u003c/strong\u003eSparse Canonical Correlation Analysis.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eConflicts of interest disclosure:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo potential conflict of interest was reported by the author(s).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003cstrong\u003e:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll of the processed data in this study has been uploaded in the National Omics Data Encyclopedia under accession NO. OEP005066. The raw metagenomic data are available in the European Nucleotide Archive (https://www.ebi.ac.uk/ena/) under accession NOs. PRJNA558102, PRJNA767535, PRJNA682622, PRJNA704861 and PRJNA523510. The data relevant to the study are uploaded as supplementary information. The code and scripts are available on GitHub (https://github.com/tjcadd2020/NAFLD-Host-Microbe-Interaction).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding statement:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Natural Science Foundation of China (82000536 to NJ, 92251307 to RZ, 82170542 to RZ), the National Key Research and Development Program of China (2021YFF0703700/2021YFF0703702 to RZ), Guangdong Province \u0026ldquo;Pearl River Talent Plan\u0026rdquo; Innovation and Entrepreneurship Team Project (2019ZT08Y464 to LZ), the program of Guangdong Provincial Clinical Research Center for Digestive Diseases (2020B1111170004), and National Key Clinical Discipline. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNJ, LZ, and RZ conceived and designed the project. WY and WG drafted the manuscript. YY, WL, WC, XZ, RZ, LZ, and NJ revised the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eWu D, Liu L, Jiao N, et al. Targeting keystone species helps restore the dysbiosis of butyrate-producing bacteria in nonalcoholic fatty liver disease. iMeta. 2022;1(4):e61.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiao N, Loomba R, Yang Z-H, et al. Alterations in bile acid metabolizing gut microbiota and specific bile acid genes as a precision medicine to subclassify NAFLD. Physiol Genom. 2021;53(8):336\u0026ndash;48.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiao N, Baker SS, Chapa-Rodriguez A, et al. Suppressed hepatic bile acid signalling despite elevated production of primary and secondary bile acids in NAFLD. Gut. 2018;67(10):1881\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSookoian S, Salatino A, Casta\u0026ntilde;o GO, et al. Intrahepatic bacterial metataxonomic signature in non-alcoholic fatty liver disease. Gut. 2020;69(8):1483\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu T, Gao W, Zhu L et al. NAFLDkb: A Knowledge Base and Platform for Drug Development against Nonalcoholic Fatty Liver Disease. \u003cem\u003eJournal of Chemical Information and Modeling\u003c/em\u003e. 2023;10.1021/acs.jcim.3c00395.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoyles L, Fern\u0026aacute;ndez-Real J-M, Federici M, et al. Molecular phenomics and metagenomics of hepatic steatosis in non-diabetic obese women. Nat Med. 2018;24(7):1070\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiao N, Baker SS, Nugent CA, et al. Gut microbiome may contribute to insulin resistance and systemic inflammation in obese rodents: a meta-analysis. Physiol Genom. 2018;50(4):244\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu B, Zhou Z, Jin Y, et al. Hepatic stellate cell activation and senescence induced by intrahepatic microbiota disturbances drive progression of liver cirrhosis toward hepatocellular carcinoma. J Immunother Cancer. 2022;10(1):e003069.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeinwand JC, Paul B, Chen R, et al. Intrahepatic microbes govern liver immunity by programming NKT cells. J Clin Investig. 2022;132(8):e151725.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePirola CJ, Salatino A, Quintanilla MF, Casta\u0026ntilde;o GO, Garaycoechea M, Sookoian S. The influence of host genetics on liver microbiome composition in patients with NAFLD. eBioMedicine. 2022;76:103858.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSookoian S, Pirola CJ. Liver tissue microbiota in nonalcoholic liver disease: a change in the paradigm of host-bacterial interactions. Hepatobiliary Surg Nutr. 2021;10(3):337\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerhard GS, Legendre C, Still CD, Chu X, Petrick A, DiStefano JK. Transcriptomic Profiling of Obesity-Related Nonalcoholic Steatohepatitis Reveals a Core Set of Fibrosis-Specific Genes. J Endocr Soc. 2018;2(7):710\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGovaere O, Cockell S, Tiniakos D, et al. Transcriptomic profiling across the nonalcoholic fatty liver disease spectrum reveals gene signatures for steatohepatitis and fibrosis. Sci Transl Med. 2020;12(572):eaba4448.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKozumi K, Kodama T, Murai H, et al. Transcriptomics Identify Thrombospondin-2 as a Biomarker for NASH and Advanced Liver Fibrosis. Hepatology. 2021;74(5):2452\u0026ndash;66.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePantano L, Agyapong G, Shen Y, et al. Molecular characterization and cell type composition deconvolution of fibrosis in NAFLD. Sci Rep. 2021;11(1):18045.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao K, Tarabra E, Sia D, et al. Transcriptomic profiling of a multiethnic pediatric NAFLD cohort reveals genes and pathways associated with disease. Hepatol Commun. 2022;6(7):1598\u0026ndash;610.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrunt EM, Kleiner DE, Wilson LA, Belt P, Neuschwander-Tetri BA. for the NCRN. Nonalcoholic fatty liver disease (NAFLD) activity score and the histopathologic diagnosis in NAFLD: distinct clinicopathologic meanings. Hepatology. 2011;53(3):810\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNurk S, Koren S, Rhie A, et al. The complete sequence of a human genome. Science. 2022;376(6588):44\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePoore GD, Kopylova E, Zhu Q, et al. Microbiome analyses of blood and tissues suggest cancer diagnostic approach. Nature. 2020;579(7800):567\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSalzberg SL, Breitwieser FP, Kumar A, et al. Next-generation sequencing in neuropathologic diagnosis of infections of the nervous system. Neurol Neuroimmunol Neuroinflamm. 2016;3(4):e251.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreitwieser FP, Baker DN, Salzberg SL. KrakenUniq: confident and fast metagenomics classification using unique k-mer counts. Genome Biol. 2018;19(1):1\u0026ndash;10.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhaddar B, Biswas A, Harris C, et al. Tumor microbiome links cellular programs and immunity in pancreatic cancer. Cancer Cell. 2022;40(10):1240\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSalter SJ, Cox MJ, Turek EM, et al. Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biol. 2014;12(1):1\u0026ndash;12.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology. 2005;41(6):1313\u0026ndash;21.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu L, Baker SS, Gill C, et al. Characterization of gut microbiomes in nonalcoholic steatohepatitis (NASH) patients: A connection between endogenous alcohol and NASH. Hepatology. 2013;57(2):601\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePowell NR, Liang T, Ipe J, et al. Clinically important alterations in pharmacogene expression in histologically severe nonalcoholic fatty liver disease. Nat Commun. 2023;14(1):1474.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoore MP, Cunningham RP, Meers GM, et al. Compromised hepatic mitochondrial fatty acid oxidation and reduced markers of mitochondrial turnover in human NAFLD. Hepatology. 2022;76(5):1452\u0026ndash;65.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi R, Xin T, Li D, Wang C, Zhu H, Zhou H. Therapeutic effect of Sirtuin 3 on ameliorating nonalcoholic fatty liver disease: The role of the ERK-CREB pathway and Bnip3-mediated mitophagy. Redox Biol. 2018;18:229\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSanjuan MA, Dillon CP, Tait SWG, et al. Toll-like receptor signalling in macrophages links the autophagy pathway to phagocytosis. Nature. 2007;450(7173):1253\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao W, Chen W, Yin W et al. Identification and validation of microbial biomarkers from cross-cohort datasets using xMarkerFinde. 2022. PROTOCOL (Version 1) available at Protocol Exchange [\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21203/rs.3.pex-1984/v1]\u003c/span\u003e\u003cspan address=\"10.21203/rs.3.pex-1984/v1]\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Online Resources","content":"\u003cp\u003eOnline Resources 9 and 10 are not available with this version. \u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"intrahepatic microbiome, hepatic microenvironment, host-microbe interaction, nonalcoholic fatty liver disease, mitophagy, steatohepatitis, mitochondria, immune microenvironment, disease progression, bioinformatics","lastPublishedDoi":"10.21203/rs.3.rs-4404936/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4404936/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e: The intricate interplay between host genes and intrahepatic microbes is vital in shaping the hepatic microenvironment and contributes significantly to our understanding of nonalcoholic fatty liver disease (NAFLD). However, the underlying mechanisms of disease progression mediated by these interactions remain largely elusive.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e: We conducted a comprehensive analysis of 570 liver biopsy transcriptomes from five cohorts, including 72 control, 124 nonalcoholic fatty liver (NAFL), 143 borderline and 231 nonalcoholic steatohepatitis (NASH) samples. Least Absolute Shrinkage and Selection Operator penalized regression and Sparse Canonical Correlation Analysis were utilized to identify host-microbiota interactions and their function.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e: We observed significant upregulations of key genes involved in mitochondrial organization across all disease stages, while genes related to antigen processing showed abnormal activations in advanced stages like NASH. Additionally, the abundances of intrahepatic microbes \u003cem\u003eMethyloversatilis sp. RAC08\u003c/em\u003e and \u003cem\u003eRalstonia insidiosa\u003c/em\u003edecreased significantly across all NAFLD stages. We identified 5537, 1937, 1485, and 2933 host-microbiota interactions in control, NAFL, borderline, and NASH samples, respectively. Notably, interaction strength showed a decreasing trend, especially during the transition from the borderline stage to NASH. In NAFL and borderline stages, bacteria like Bacillales, \u003cem\u003eRalstonia insidiosa\u003c/em\u003e, and Micromonosporaceae played pivotal roles in enhancing host mitophagy by interacting with genes including \u003cem\u003eSQSTM1\u003c/em\u003e, \u003cem\u003eOPTN\u003c/em\u003e, and \u003cem\u003eBNIP3L\u003c/em\u003e. However, such interaction functional clusters were absent in NASH samples.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e: Disturbed host-microbiota interactions affecting the mitophagy process can lead to a pro-inflammatory hepatic microenvironment through activation of immune reactions, potentially driving disease progression to NASH.\u003c/p\u003e","manuscriptTitle":"Disrupted Host-Microbiota Crosstalk Promotes Nonalcoholic Fatty Liver Disease Progression by Impaired Mitophagy","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-14 15:45:29","doi":"10.21203/rs.3.rs-4404936/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c11ef064-2f59-40dc-b1bc-1feebe48fbac","owner":[],"postedDate":"June 14th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-06-24T20:10:47+00:00","versionOfRecord":[],"versionCreatedAt":"2024-06-14 15:45:29","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4404936","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4404936","identity":"rs-4404936","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}