Evaluation of reciprocal interaction between host and pathogens in the development of diabetic foot infections using next-generation sequencing

Chronic and non-healing diabetic ulcers associated with high morbidity and mortality are serious complications in health care systems. Diabetic foot infections (DFIs) affecting fifty percent of people with diabetic ulcers have been identified as the most severe and costly complication in people with diabetes. However, due to the high heterogeneity of bacterial communities identified in DFIs, culture-based methods may not reflect DFIs complexity accurately. Over the last decade, new advances in molecular-based approaches have addressed the limitation of traditional approaches. DNA-based approaches (metagenomics), RNA-based approaches (metatranscriptomics), and analysis of 16S rRNA region (microbiome studies) have enabled the evaluation of DFIs beyond the limitation of culture-based methods. In this study, we have evaluated bacterial communities in DFIs taxonomically and transcriptionally using DNA- and RNA-based methods. Due to the high level of human cells contamination in human infected samples, extraction of high-quality genomic material from bacterial cells has always been challenging. In chapter three we have shown how the microbiome DNA enrichment method can impact downstream analysis in -omics approaches. According to our results, there was a significant difference (p < .001) in human contamination amount in samples extracted by the NEBNext method (high contamination) compared to samples extracted by QIAamp and HostZERO methods (low contamination). Taxonomic classification has shown that samples extracted by different microbiome DNA enrichment methods contained a total of five phyla dominating by Firmicutes and Actinobacteria. Also, samples extracted by EBNext had the most similar bacterial pattern to the control sample. Overall, all the microbiome DNA enrichment methods successfully identified clinically important isolated genera in DFIs with relatively similar abundance. In chapter four using an RNA sequencing approach, the composition, function, and pathogenicity of the active bacterial communities in DFIs have been investigated. Based on our findings, there was a significant difference in the abundance of transcripts assigned to Spiroplasma, Vibrio, and Mycoplasma in samples with different infection severity (P<0.05). Samples with mild and severe infections were dominated by Staphylococcus aureus and Porphyromonas asaccharolytica, respectively. Functional analysis showed the high abundance of pathways involved in ribosome and thiamine metabolism in DFIs samples. There was also a significant difference in pathways involved in lipoic acid metabolism, two-component systems, bacterial invasion to epithelial cells, glycerolipid metabolism, TCA cycle, and mismatch repair in samples with different infection severity (P< 0.05). Also, iron acquisition systems and pathways involved in the synthesis and regulation of cell-surface components were the most common virulence factors in DFIs. Multidrug efflux pumps/exporters were also the most predominant transcripts identified in DFIs samples compared to other antibiotic resistance mechanisms. We further investigated the reciprocal interaction between host and pathogens in DFIs in chapter five. According to human Gene Ontology analysis in DFIs samples, main clusters were involved in the recruitment of inflammatory cells and immune responses and skin cell development and wound healing processes such as extracellular structure organization and blood vessel development. Genes involved in adaptive/native immune responses and transport of mature mRNAs were also significantly different between samples with high and samples with a low number of virulence factors. Using a multi-omics approach in chapter six we observed minimum dissimilarities in the overall bacterial composition in samples with different infection severities. However, Staphylococcus aureus and Streptococcus were abundant in mild and moderate samples. These two genera showed the highest association to infection pathogenicity in DFIs. Moreover, samples with severe infections showed higher bacterial diversity compared to samples with mild and moderate infections. It may be concluded that molecular-based approaches are more reliable than traditional methods in the investigation of DFIs. However, due to the paucity of information in this regard, pre-clinical studies using animal models on a larger scale are needed to investigate DFIs pathogenicity in detail.  Since omics approaches are a relatively young but rapidly growing field, there is great growth potential. In the last chapter, the application of omics methods in pre-clinical studies using animal methods and final assessment of findings in this study has been discussed.