Hunting down what makes good bacteria bad company : a study investigating the genetic and phenotypic differences between Limosilactobacillus fermentum AGR1485 and AGR1487 to elucidate the molecular basis underlying L. fermentum AGR1487's barrier-disruptive phenotype : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy, Massey University, Manawatū, New Zealand

Abstract

Background: Bacteria play crucial roles in human society, from food production to natural symbiosis. Among these, probiotic species such as Limosilactobacillus fermentum are widely recognised for their health benefits, although not all strains exhibit equivalent properties. Notably, strain AGR1487 of L. fermentum has been identified as an outlier, exhibiting a barrier-disruptive phenotype in an in vitro model of the intestinal epithelial barrier. Disruption of this barrier is clinically relevant, as it can increase intestinal permeability, facilitate translocation of luminal antigens or pathogens, and contribute to gastrointestinal disorders, including inflammatory bowel disease and irritable bowel syndrome. The in vitro model employs Caco-2 cells, derived from human colorectal adenocarcinoma, which differentiate into enterocyte-like cells, and measures transepithelial electrical resistance (TEER) to assess barrier function. In contrast, strain AGR1485, isolated from the same human population, does not display a barrier-disruptive phenotype. This study investigates the genetic and phenotypic differences between AGR1485 and AGR1487 to elucidate the molecular basis underlying the barrier-disruptive phenotype.

Methods: A variety of methods, including a multi-omics approach, were employed to analyse the genetic and phenotypic differences between AGR1485 and AGR1487. The barrier-disruptive phenotype was confirmed through TEER assays. Growth characteristics were assessed using spectrophotometry, and bacterial morphology was examined via electron microscopy. Genomic sequencing was performed using hybrid assembly approaches, combining Illumina short-read and PacBio long-read sequencing. Comparative genomic and transcriptomic analyses of both strains were conducted to identify non-identical genes and differential gene expression patterns.

Results: TEER assays demonstrated that AGR1487 significantly disrupts intestinal barrier integrity (-20.11 % TEER per hour), whereas AGR1485 had minimal effects. Electron microscopy revealed AGR1487’s tendency to autoaggregate, a feature absent in AGR1485. Both genomes were sequenced, and the assemblies were of exceptionally high quality, with BUSCO completeness scores of 99.5 % (98.8 % single-copy, 0.7 % duplicated, 0.5 % fragmented, 0 % missing), confirming their suitability for detailed comparative and functional analyses. Comparative genomic analysis showed that AGR1487 possesses 62 non-identical genes within regions of high genomic variability, suggesting horizontal gene transfer events and diversification or loss of these genes in other strains. Transcriptomic profiling indicated that AGR1487 expresses these non-identical genes before TEER assays, with 25 candidate genes potentially implicated in its barrier-disruptive properties. Among them, a phage-associated truncated metallopeptidase and mucin-binding protein were identified as strong candidates.

Conclusion: The findings of this PhD thesis highlight strain-specific variations within L. fermentum, emphasising the necessity of evaluating probiotics at the strain level. The identification of genetic elements responsible for the barrier-disruptive phenotype in AGR1487 underscores the complexity of bacterial-host interactions and the need for careful screening in probiotic applications. Future research will focus on the functional validation of candidate genes through knockout and complementation studies to confirm their roles in gut barrier integrity disruption