부산대학교 도서관

Background: Human health is maintained by crosstalk between the gut microbiome and the gastrointestinal tract. The gut microbiome breaks down complex dietary compounds to produce metabolites which serve a multitude of functions and interact with host molecular pathways. The gut microbiome of an individual changes with host and environmental factors including location of residence and diet. Aims: To utilise open source software and apply tailored statistical approaches to investigate the role of the omics features of undernourishment, knee pain in osteoarthritis (OA), non-alcoholic fatty liver disease (NAFLD) and the effect of location of residence and dietary intakes. The objectives were: [1] To explore the relationship of the gut microbiome and its functional metabolic pathways with nutrition status and location of residence. [2] To explore the relationship of current knee pain with the serum metabolites and gut microbiome in people with OA. [3] To investigate the interplay of the gut microbiome, serum metabolome and dietary intakes in NAFLD compared to healthy participants. Methodology: This research incorporated analysis of 16S rRNA microbiome sequencing, serum metabolome, serum cytokines, and dietary recalls. Data analysis was performed using QIIME2 for microbiome sequence analysis, MicrobiomeAnalyst (Tax4Fun) for functional profiling from 16S rRNA, Metaboanalyst (Pathway Analysis) for investigating the pathways and FETA for calculating dietary intakes. The differential abundance (ANCOM), statistical significance inferences, correlation tests, regression modelling (linear, logistic and LASSO), classification modelling (OPLS-DA), meta-analysis (REML) and mediation analysis (causal mediation) were performed using statistical packages in R. Results: [1] Microbiome analysis showed an increased abundance of Prevotella 7 and uncultured members belonging to Prevotellaceae, Turicibacter, Megasphaera and decreased abundance of Marvinbryantia and Erysipelotrichaceae UCG-003 in participants residing in an urban location compared participants from rural Indian population. Also, the functional profiling showed the KEGG Orthologs involved in the carbohydrate metabolism, lipid metabolism and biosynthesis of secondary metabolites to be significantly altered based on the location regardless of the difference in the macronutrient intakes. This suggests that location of residence and not the nutritional status assessed based on BMI is the stronger determining factor modulating the gut microbiome and the functional metabolic pathways. [2] Meta-analysis of two independent UK studies showed levels of acyl ornithine, carnosine, cortisol, cystine, dihydroxyphenylalanine, glycolithocholic acid sulphate, phenylethylamine and succinic acid were significantly associated with knee pain. The pathway analysis further demonstrated amino acids and cholesterol metabolism pathways are involved in knee pain. These metabolites, irrespective of the association with the pain scores were also found to be significantly associated with the serum levels of cytokines including IL-10, IL-13, IL-1β, IL-2, IL-8 and TNF-α suggesting that they may have immunomodulatory properties. Significant associations of increased abundance of Parasutrella and decreased abundance of Streptococcus with an increase in pain and their significant associations with biogenic amines such as taurine, histamine and spermidine suggest their key role in secondary metabolism related to pain pathways. [3] Microbiome analysis revealed that an increased abundance of mainly Collinsella and Bifidobacterium is associated with an increase likelihood of developing NAFLD. Changes in levels of serum lipoproteins such as HDL and VLDL in NAFLD and their association with specific microbes suggests a potential role of microbes in secondary metabolism of lipids. The differences in the intakes of calcium, iron, selenium, total sugars and different food items classified under potatoes, meat and milk suggests that diet plays an important role in NAFLD. The significant mediation effect scores of the microbiome as mediator suggests its causal link between the diet and disease status. Conclusion: Taken together, this work has demonstrated the utility of integration and analysis of multi-omics data to establish the contribution of the gut microbiome, metabolome and diet towards understanding and improving human health. The omics features associated with the pathophysiological conditions or diseases identified in this research may serve as the molecular signatures or diagnostic biomarkers. This thesis paves a way for the discovery of new therapeutic targets or intervention strategies. This research makes an important contribution towards a deeper understanding and better management of nutrition, knee OA and NAFLD.