Comparison of cellular, surface and secreted proteins produced during starvation and acid stress conditions for Lactobacillus casei using proteomics

This thesis examines two aspects of stress physiology in Lb. casei strain GCRL163: the ability of this strain, and a broader range of lactobacilli isolates, to grow on Tween 80 as a carbon source, which is scavenged in the absence of other fermentable carbon sources (glucose and citrate); and adaptation to growth under acidic conditions. Members of the Lb. casei group include Lb. casei, Lb. paracasei and Lb. rhamnosus, which are difficult to differentiate on biochemical and physiological traits. The first part of the experimental work involved isolating a broader range of Lb. casei strains from fermented foods and milk sources, to address the question of whether Tween 80 use as a carbon source was a phenomenon common in this group or not. LC agar was used to isolate Lb. casei group strains selectively. Isolates from plates were speciated by 16S rDNA sequencing of PCR amplicons produced using universal primers (Turner's modification of Lane primers) while Lb. casei group-specific and species-specific (for Lb. rhamnosus) primers were used to speciate isolates. The major part of the study was on culturing Lb. casei GCRL163 in fermenters in two separate experiments (for 46h and 12 days, experiments P1 and P2 respectively) under starvation conditions in mMRS broths containing test and control growth conditions. The control culture had no acetate, citrate, Tween 80 or lactose/glucose, and the tests were the control medium with addition of Tween 80 only, Tween 80 plus citrate (TwCit) and citrate only, which allowed comparisons against the control and between these conditions. Another test condition in P1 involved dosing one fermenter with additional Tween 80 (TwD), to determine whether growth was further stimulated relative to the control. Proteomic and physiological analysis of nutritional stress was performed by application of a comprehensive statistical tool (Perseus version 1.5.031) to map proteomic differences between growth conditions ‚Äö- variables being added carbon source and growth phase. In both of these studies, proteomic analyses were conducted on cell-free extracts (CFE) and extracellular culture fluids (ECF), with LiCl extracts examined in the case of acid adaptation. This enabled evaluation of stress responses within cells and the secretomes and comparison between nutritional and acid stress responses. Thus the initial aim of this thesis was to determine whether Tween 80 could be used as a carbon source by Lb. casei GCRL163, and other strains in the Lb. casei/paracasei group, and to clarify the underlying biochemical pathways involved using proteomic and physiological analyses of cells. Acid stress studies were also conducted using pH 6.5, pH 4.5 and uncontrolled pH growth conditions in fermenters. Growth at low pH had been analysed previously in our laboratory so this set of conditions enabled comparisons across data sets, as the HPLC-MS and analytical platform had changed since these prior studies at UTAS. Furthermore, it emerged from the P1 and P2 studies that secreted proteins were an important element of the whole cell proteome, which may also be relevant in acid adaptation. Consequently, analyses were extended to examining the surface and secreted proteins under acidic growth conditions, relative to growth at neutral pH and when pH was not controlled in fermenters. In addition to identification of proteins at the cell surfaces, regulation of proteins during acid adaptation is described that has not been reported in prior studies. Fatty acid analyses were undertaken for cells from P1, P2 and the acid adaptation study. Results indicated that LC agar, a medium recommended for selective isolation of Lb. casei in yogurt, was not selective for the Lb. casei group members, as many of the isolates from milk were identified as Lactococcus species using 16S rRNA gene sequencing, demonstrating that LC agar failed to eliminate other LAB present in the complex microflora of milk and fermented food samples. The use of two sets of PCR primers (Kwon et al. 2004) enabled differentiation of Lb. casei group members from other lactobacilli and Lb. rhamnosus-specific primers clearly differentiated this species within the group. When tested for growth on Tween 80 in mMRS by Bioscreen there was no correlation observed between species or source of isolation and growth capability or patterns of growth, indicating considerable variability between the Lb. casei group isolates in capacity to grow on mMRS containing Tween 80. Co-metabolism of Tween 80 and citrate was observed when both were present in the medium resulting in faster growth, more rapid Tween 80 consumption (23% decline in concentration in 46 h in TwCit) and further upregulation of proteins already seen as upregulated in citrate-cultured cells (citrate metabolism, central glycolytic proteins), while lipid metabolism proteins were repressed. The expressed proteome of TwD cells indicated depletion in many cellular proteins and showed induction of stress proteins. Fatty acid analysis of cells indicates that 10-hydroxystearic acid was uniquely detected in P1 and P2 cells, likely the product of the Sph protein (oleate hydratase) which was detected in all cells and culture conditions at varying levels of expression. Cyclopropyl-C18:0 was also detected in P1 and for cells cultured at pH 4.5; this fatty acid has been linked to stress adaptation previously. In P2 cells, 10-hydroxystearic acid was the major fatty acid in Tween 80 cultured cells, while citrate and control cells demonstrated synthesis of C15 fatty acids from amino acid primers. The Tween 80 cultured cells lacked many of the Fab and other fatty acid synthesis proteins, whereas the control cells in P2 retained several Fab proteins. Analysis of CFEs and extracellular culture fluids (ECFs) for P2 failed to identify a lipase/esterase or hydrolase that may have been related to Tween 80 degradation, despite detecting several proteins that were unique to this experiment (against the acid adaptation study) in the ECFs. Many of the proteins unique to the ECFs (not detected in the CFEs) were transmembrane or secreted proteins (indicated from the presence of signal peptides) and transporters. Several of the proteins detected in the ECF were also detected in the CFEs as highly express proteins. However, the majority of proteins in the CFEs which were highly expressed were not detected in the ECF, so the presence of proteins in the ECF that are normally deemed as cytoplasmic could not be fully explained by cell lysis. Many of these ECF proteins have been previously described as moonlighting proteins that have also been detected at the cell surface. Analysis of surface, secreted and cellular proteins across the culture cycle for Tween 80 cells may provide some further insight into the degradation pathway of Tween 80. Analyses of CFEs showed that cells adapted to growth in acid conditions by rerouting carbon flow around pyruvate metabolism, with the noted upregulation of the malolactic enzyme MleA, as reported previously from transcriptomic studies on acid shock. Other upregulated lactate/malate dehydrogenases were also detected which may contribute to conversion of malate to pyruvate. However, when the pH was uncontrolled, and cells were moving into stationary phase at the point of cell harvest, it was clear that many proteins which were detected in the cells grown at pH 4.5 and 6.5 were not detected and that the pattern of expression was different to the pH 4.5-cultured cells. In several instances, proteins detected as upregulated at pH 4.5 were more highly expressed in cells with uncontrolled pH. Of the three identified pyruvate oxidases on the genome of GCRL163, one was more highly expressed at pH 4.5 but was a different protein to the Pox5 detected as regulated during growth on citrate. The latter protein (K0NBB6) was detected in cells grown with uncontrolled pH, suggesting cells were moving to utilisation of alternative substrates in mMRS, or recycling cellular components, at this stage of growth. In addition to impacting on central metabolism and the pentose phosphate pathway, modulation of proteins during adaptation to growth at low pH occurred principally across the following functionalities: amino acid/peptide degradation and acquisition; t-RNA ligase activity; cell wall biogenesis, cytokinesis and membrane bioenergetics; and lipid metabolism. Many of the upregulated proteins were in the uncharacterised or general prediction functional classes. One functionally uncharacterised aminotransferase was highly upregulated in all experiments for cells cultured at pH 4.5 and for the uncontrolled pH CFEs, indicating a key role in acid adaptation. Initial examination of the LiCl extracts indicated that there was lysis occurring during extraction, given that over 800 proteins were detected in extracts. By looking at proteins that were relatively enriched into LiCl, more than 200 proteins were deemed as potentially surface located and extractable into LiCl. These proteins were also checked for their presence or absence in the ECF. Highly upregulated proteins in LiCl extracts under acid adaptation (pH 4.5) found at the cell surface belonged to major functional classes such as cell wall biogenesis, ABC-type transporter systems, cytokinesis, amino acid-related metabolism, central glycolytic/intermediary pathways and uncharacterised proteins. Many of these proteins were transmembrane or had signal sequences, supporting surface or extracellular locations. BLAST analysis was used to assign presumptive identity to many of the proteins in the uncharacterised or general prediction categories. Many of the proteins in these categories were regulators, anchor proteins or surface/secreted hydrolases involved in cell wall biogenesis and cell division. Two notable proteins in this group were previously identified as major surface proteins (MSPs) in some lactobacilli species. Uncharacterised, acid regulated, LiCl-extracted protein K0...