Contents

Norwegian Research Council FUGE project

Secretomes of lactobacilli and enterococci and their impact on bacterial adaptation and human health

2008 - 2010

Summary

Lactic acid bacteria belonging to the genera Lactobacillus and Enterococcus are normal inhabitants of fermented food products and of the human gastrointestinal tract. The impact of these bacteria on human health varies from beneficial to harmful . We will conduct genome-wide studies of surface-located and secreted proteins (secretomes) of lactobacilli and enterococci, including comparative analyses (by experiment and bioinformatics) of two lactobacilli with different genome sizes (L. plantarum WCFS1, 3.31 Mb, and L. salivaricus, 2.13 Mb) and of enterococci displaying varying degrees of pathogenicity (E. faecalis and E. faecium). The bacteria will be grown under various conditions selected to resemble elements of the human ecosystem (bile salts, intestinal extracts, epithelial cells) and bacterial responses will be studied using transcriptomics and proteomics. In particular, we will analyze how the secretome is affected by growth conditions and how species differ in their adaptive behaviour. These studies will be complemented by phenotypic studies (e.g. immunomodulation, adhesion, survival), using e.g. knock-out mutants, over-expression of selected genes, and/or in vivo monitoring of labelled bacterial cells. In some cases, further insight will be obtained through detailed studies of target proteins emerging from the genome-wide work. The project involves two major European Research Centers (in Cork, Ireland and in Wageningen, The Netherlands) working on lactobacilli, and an Italian group that recently has developed new techniques for LC-MS/MS-based analysis of proteins displayed on the surfaces of Gram-positive bacteria. By conducting this systematic, partly comparative functional genomics analysis of bacterial secretomes, we hope to identify and understand mechanisms that are crucial for bacterial behaviour during interaction with the human ecosystem.

Background and general outline of the project

Lactic acid bacteria belonging to the genera Lactobacillus and Enterococcus are normal inhabitants of fermented food products and of the human gastrointestinal tract. The impact of these bacteria on human health varies from beneficial (e.g by displacing pathogens or being immunomodulators), to harmful (e.g. causing infectious diseases). Lactobacilli are regarded as harmless, whereas certain species in fact are considered beneficial for human heath. Currently, there is a considerable interest in these potentially beneficial harmless bacteria, which are referred to as “probiotics”, one goal being to unravel the molecular mechanisms behind their impact on human health (Marco et al. 2006). The status of enterococci is less univocal. On the one hand, enterococci are common, presumably beneficial components in fermented food. On the other hand, however, they are associated with a variety of diseases such as mucosal lesions, persistent root canal infections, endocarditis, and various hospital-acquired infections. In addition, some enterococcal isolates are known to be notoriously resistant towards antibiotics and there is a suspicion that presumably harmless enterococci may function as “reservoirs” and intermediate hosts in the spread of antibiotic resistance genes.

Available genome sequences of lactobacilli (e.g. Kleerebezem et al. 2003; Claesson et al. 2006; see also Eijsink & Axelsson, 2005) and enterococci (e.g. Paulsen et al. 2003) have revealed many genes that code for proteins putatively involved in the interaction between these bacteria and their human host (e.g. adhesion factors; Buck et al. 2005; Pretzer et al. 2005; Boekhorst et al. 2006a). DNA microarray studies of lactobacilli have shown that, as expected, the expression of many of these genes is modulated upon contact with human cells (M. Kleerebezem, personal communication). The contribution of some of these genes to the bacteria’s ability to adhere to human cells and/or to affect the physiology of these cells (e.g. cytokine production) is currently being studied (Morita et al. 2002; Valeur et al. 2004; O'Mahony et al. 2005; Pretzer et al. 2005; Vizoso Pinto et al. 2007). To understand bacterial behaviour and impact in the GI-tract, processes other than adhesion also need to be studied. Thus, the responses of these bacteria to a variety of conditions, including conditions that are relevant for behaviour in the GI-tract, are currently under investigation (e.g. Giard et al. 2001; Bron et al. 2004; Bron et al. 2006; Aakra et al. 2007). The cytoplasmic membrane in Gram-positive bacteria is surrounded by a rigid thick cell wall, which is composed of several distinct structures. The (small) space between the cell wall and the membrane bears some resemblance to the periplasmic space in Gram-negative bacteria (Merchante et al. 1995; Matias et al. 2006). The Gram–positive cell wall of lactic acid bacteria consists mainly of peptidoglucan, (lipo)teichoic acids, proteins and polysaccharides. The peptidoglucan network in the cell wall is called sacculus which is decorated with proteins, teichoic acids and polysaccharides. In general, proteins associated with the bacterial cell surface are divided into four subclasses: (1) proteins anchored to the membrane via a transmembrane domain; (2) lipoproteins which are covalently connected to a membrane lipid; (3) proteins which are covalently bound to the cell wall through a C-terminal cell wall anchor domain (LPxTG-motif); (4) proteins which are noncovalently attached to cell wall, usually via specific binding domains (e.g. LysM) (Desvaux et al. 2006). The impact and adaptability of these bacteria is obviously for a large part determined by proteins that reside on the cell surface (i.e. anchored to either the cell membrane or the cell wall) and proteins that are secreted into the environment. These proteins have numerous important functions such as the provision of nutrients, detoxification of the environment, killing of potential competitors, adhesion to surfaces including human epithelium, and immunomodulating effects in the GI-tract. Extracellular proteins of pathogenic bacteria play critical roles in virulence. The surface-exposed and secreted proteins, hereafter referred to as the secretome, can be recognized on the basis of genome sequences alone, because the primary gene products contain signals for secretion, anchoring and cognate post-translational modifications. Most papers describing the genomes of these bacteria present an analysis of putatively secreted or surface-associated proteins.

For example, the genome of Lactobacillus plantarum WCFS1 (Kleerebezem et al. 2003) is thought to encode totally 217 proteins with some kind of N-terminal signal sequence directing secretion or localization at the cell surface. 144 of these proteins contain putative signal peptidase cleavage sites (97 standard signal peptides and 47 N-terminal lipoprotein anchors), whereas others contain membrane anchors or cell wall anchoring domains (LPxTG or LysM domains). In their 2003 paper, Kleerebezem et al. concluded that “for the majority of the extracellular proteins, no definite function prediction can be made”. They also noted that many these of extracellular proteins may play a role in adhesion or binding to other cells or proteins, because they contain domains with homology to proteins in databases with predicted functions such as mucus-binding, fibronectin-binding, aggregation-promoting, intercellular adhesion, or cell clumping. Subsequent more extended in silico analyses of the secretome of L. plantarum have been conducted and are available in a database (http://www.cmbi.ru.nl/secretome) (Boekhorst et al. 2006b). Boekhorst et al (2006b) identified 69 extracellular proteins predicted to have enzymatic function. Despite, succesful efforts in bioinformatics, there is still a large need for experimental work to verify in silico predictions. The database for L. plantarum is an excellent basis for targeted experimental studies of the function of the extracellular proteins. Similar detailed in silico studies do not exist for enterococci but analyses are currently in progress, e.g. in the applicant’s laboratory.

Taking into account the predicted importance of the secretome for impact and adaptability, there are several reasons for taking a closer look at the secretome at the protein level: Primary gene products have to “travel” quite a long way (translocation, folding, post-translational modifications) before reaching their final location and functionality, casting a relatively large uncertainty over the quantitative links between transcriptional levels and levels of correctly localized functional proteins. Levels of correctly localized functional members of the secretome are highly interrelated since the secretome depends on the interplay between individual gene products of to-be-translocated proteins, the (partly translocated) secretion machinery, (translocated) enzymes involved in cell wall modification and anchoring, and (translocated) enzymes involved in extracellular post-translational modifications. This adds to the uncertainty concerning the quantitative links between transcriptional levels and levels of correctly localized functional proteins.

Most remarkably, while functional annotation of new bacterial genomes is reaching levels as high as about 80 % of the total number of predicted genes, the exact function of a relatively large number of secreted proteins in lactobacilli (Kleerebezem et al. 2003; Boekhorst et al. 2006b) and enterococci (Paulsen et al. 2003) remains unknown. Functional annotation of these proteins may be improved by disclosing their involvement in various bacterial responses. Alternatively, studies with knock-out mutants, including complementation studies with the deleted gene under control of a regulable promoter, may provide more insight. Alternatively, the function of identified key proteins may be assessed by overexpressing them and determining their structures. A structural approach may be seen as a last resort for proteins that are clearly important but for which the function remains unknown (the chances of successful functional identification on the basis of a structure alone are not big).

In this project we intend to study surface and secreted proteins of lactobacilli and enterococci, using a variety of techniques for fractionation, isolation and characterization of surface protein fractions, as outlined in detail below. The bacteria will be grown under various conditions selected to resemble elements of the human ecosystem (bile salts, intestinal extracts, epithelial cells). In addition, we will do comparative analyses with two lactobacilli with different genome sizes (L. plantarum WCFS1, 3.31 Mb, and L. salivarius, 2.13 Mb) and with enterococci displaying varying degrees of pathogenicity (various E. faecalis strains; E. faecium). Differences between the highly related species L. plantarum and E. faecalis will receive special attention. By conducting these studies, we wish to identify mechanisms and individual proteins/enzymes that are crucial for bacterial behaviour during interaction with the human ecosystem. The project will be conducted in close collaboration with two major European Research Centers (in Cork, Ireland and in Wageningen, The Netherlands) working on lactobacilli and with an Italian group that recently has developed new techniques for analysis of proteins displayed on the surfaces of Gram-positive bacteria. Whereas this project focuses on proteins, it will be closely linked to corresponding studies of the transcriptome both in lactobacilli (at the two foreign centers mentioned above) and in enterococci (through a close collaboration with the group of Prof. Ingolf F. Nes at the applying institute). We have already started experiments were we compare transcriptome and proteome data from enterococci responding to bile (Aakra, Vebø, Snipen, Mathiesen, Eijsink & Nes, unpublished data). The Ph.D. student working on the proteomics part of this ogoing activity will be incorporated in this new FUGE project.

Objectives

  • Major objective: To study how the secretome of lactobacilli and enterococci responds to environmental factors and to link this response to the impact of these bacteria on human health.
  • Sub-goals:
  1. Use transcriptomics and proteomics to study how the secretome of lactobacilli and enterococci responds in the intestinal ecosystem, exemplified by e.g. (a) bile salts, (b) human intestinal extracts (e.g. mucus), and (c) living human cells (Caco-2 and HT29 cells).
  2. In-depth study of the roles of selected proteins (knock-out mutations, effects of overexpression, enzymology, structure).
  3. To link observed changes in the secretome to phenotypic changes (e.g. adhesion, survival, immunomodulation, cell wall composition).