Portal de Periódicos da CAPES

Probiotics genomics

2009; Wiley; Volume: 3; Issue: 1 Linguagem: Inglês

10.1111/j.1751-7915.2009.00159.x

1751-7915

Roland J. Siezen, Greer Wilson,

Microbial Metabolites in Food Biotechnology

We were sitting in the Irish pub on quiz night, dumbfounded by trivia questions about ingredients of Mornay sauce and best-selling Boy Bands, when the following question came up: What are 'Live microorganisms which when administered in adequate amounts confer a health benefit on the host?' At long last, we had a correct answer: PROBIOTICS! The quizmaster was totally disinterested in our extensive elaboration on this topic, so we offer it to the readers of this Genomics Update. The Russian Noble Prize winner Elie Metchinkoff first suggested that certain bacteria could modify the composition of the gut flora (Metchnikoff, 1907). He suggested that the longevity of Bulgarians and Russians of the Steppes was due to their consumption of 'sour milk' containing beneficial microbes, which in fact probably were lactic acid bacteria (LAB) such as Lactobacillus bulgaricus. Henry Tissier of the Pasteur Institute isolated bacteria (now called Bifidobacterium bifidum) from the faeces of healthy breast-fed infants and recommended giving it to babies suffering from diarrhoea (Tissier, 1900). In 1935, Minoru Shirota in Japan developed the first commercial probiotic drink called Yakult, which contains Lactobacillus casei Shirota that can survive the passage through the stomach and colonize the intestine. The probiotic market is now estimated to be worth about $6 000 000 000 a year and is growing at around 10% annually (UBIC-Consulting, 2008). Since 1981 there have been over 2000 patent applications on probiotics filed (with 'probiotic' mentioned in patent somewhere) and some 524 granted (in the USA and Europe). The two most commonly used probiotics in commercial products are lactobacilli, members of the LAB, and bifidobacteria, but some yeasts and other bacteria have been claimed to have probiotic potential. See Table 1 and Ouwehand and colleagues (2002) for an overview of commercially used strains and their claimed probiotic effects. What do probiotics actually do? What is the meaning of 'confer a health benefit'? Probiotics are most commonly known as yoghurts or yoghurt-type drinks that people ingest. The consumption of probiotics by humans is intended to improve or maintain a healthy intestine. The claimed modes of action of probiotics include strengthening of the intestinal barrier function, modulation of immune responses, supply of vitamins, and antagonism of pathogens (or other commensals) either by producing antimicrobials or by binding to the mucosa (so called competitive exclusion). (For recent reviews see Marco et al., 2006; Ventura et al., 2007; Kalliomaki et al., 2008; Lebeer et al., 2008; Kleerebezem and Vaughan, 2009.) In general, desired attributes of probiotic strains include adequate survival of the stomach passage (i.e. low pH stability), and adaptation to the host gut environment, including stress response, active and synergistic metabolism, and adherence to the intestinal mucosa and mucus. Probiotics are presumed to have an ecological advantage owing to their capacity to metabolize complex sugars that are derived from the diet as well as from the host. Sugar metabolism enzymes include various glycosyl hydrolases (GHs) which can degrade plant-derived dietary fibres or complex host carbohydrate structures. Bacteriocin production may enhance their competitiveness in the gut. From an industrial perspective, crucial attributes of probiotic strains are good technological properties for production and storage and low health risk to consumers. Probiotics need not be restricted to food applications or oral delivery. Some can be applied to the skin as lotions or cream (Krutmann, 2009) and have been used to treat vaginal infections (Reid, 2008). Probiotics are also added to animal and fish feed to enhance growth, replacing the banned additive antibiotics or growth hormones (Gatesoupe, 2008; Higuchi et al., 2008; Wynn, 2009). They appear to work by inhibiting/reducing the pathogenic bacterial load that some animals or fish carry. There is evidence for all of these probiotic modes, but the exact mechanisms of action are still not very clear. Genome-scale analyses of health-promoting bacteria, also coined 'probiogenomics' (Ventura et al., 2009), should provide clues for probiotic mechanisms and potential. Here, we provide an update of recent genomics studies in this field. Table 2 and Fig. 1 give an overview of genome sequencing of putative probiotic bacteria that are publicly available, and Table 3 gives examples of proprietary sequences of commercial probiotics. By far the most used probiotics and the ones which have their genomes sequenced are those associated with gut health. Details of genomes sequenced before 2009 have been summarized by Mayo and colleagues (2008) and Ventura and colleagues (2009). Infants are born with a sterile gastrointestinal (GI) tract but in breast-fed babies colonization by bifidobacteria is rapidly seen. It is thought that these bacteria confer a health benefit to the infant. The first colonizer is Bifidobacterium longum ssp. infantis, which has the largest genome of any sequenced bifidobacteria at 2.83 Mb (Sela et al., 2008). The genome has complete pathways for the synthesis of some vitamins and a novel 43 kb gene cluster encoding a system for the import and degradation of human milk oligosaccharides (HMOs). After weaning, the numbers of this bifidobacterium decline but others become more dominant. Bifidobacterium animalis ssp. lactis, a resident of the GI tract and the most commonly used probiotic in Europe and North America, has a genome size of only 1.9 Mb. These bifidobacteria lack the HMO cluster as presumably post-weaned animals no longer require this functionality. They do, however, contain the fos gene cluster necessary to produce the enzymes to break down and utilize health-promoting fructo-oligosaccharides, a well-known prebiotic and bifidogenic factor. Evolutionary relationships between the main gastrointestinal tract commensal lactobacilli, based on a neighbour-joining tree of 16S rRNA gene sequences. Bootstrap values above 600 are indicated. Bacterial taxa for which whole genome sequences are available are shaded in green. The outgroup is shaded in grey. Lactobacilli for which genome sequencing is ongoing/incomplete are shaded in red. Reproduced and adapted from Ventura and colleagues (2009), with permission from Macmillan Publishers Limited, 2009. Several new genome sequences of probiotics have been released in 2009. Bifidobacterium animalis ssp. lactis AD011, isolated from a healthy breast-fed infant, has a high level of immunomodulatory activity (Kim et al., 2008b). Its genome encodes multiple glycosylases than can degrade plant- or milk-derived oligosaccharides, and the fos gene cluster for processing of fructo-oligosaccharides (Kim et al., 2009). Bifidobacterium animalis ssp. lactis strains B1-04 and DSM10140, both from commercial probiotic products, differ only by 47 single nucleotide polymorphisms and four small indels, of which one indel in a CRISPR (Barrangou et al., 2009). Lactobacillus johnsonii FI9785 is a competitive exclusion agent against pathogens in poultry (Wegmann et al., 2009). Lactobacillus plantarum JDM1 is a widely used Chinese commercial probiotic strain which appears to have lost 100 kb relative to the non-commercial strain WCFS1, encoding sugar transport and metabolism, possibly due to prolonged growth of this probiotic strain in rich medium (Zhang et al., 2009). Lactobacillus rhamnosus GG and Lactobacillus rhamnosus ATCC53103, probiotic strains used widely for nearly 20 years in a variety of functional foods, differ only by deletion of 5 kb in ATCC53103, and an inversion of 8.9 kb (Kankainen et al., 2009; Morita et al., 2009). Compared with other sequenced intestinal lactobacilli, both Lb. rhamnosus genomes have a relatively high number of proteins involved in carbohydrate and amino acid metabolism and transport, and defence mechanisms. In particular, 28 complete PTS-type transporters and 25 putative GHs are encoded, including the alpha-L-fucosidase (GH29; see Cazy database http://www.cazy.org) and alpha-mannosidase (GH38) families, which are not found in other sequenced lactobacilli. In addition, these Lb. rhamnosus genomes have three gene clusters encoding proteins with WxL domains which can attach to the peptidoglycan on cell surfaces (Siezen et al., 2006; Brinster et al., 2007); again, these gene clusters have not been found in other intestinal lactobacilli, but rather in plant-associated Gram-positive bacteria (Siezen et al., 2006). Most novel is the finding that Lb. rhamnosus GG has a gene cluster spaCBA, encoding three secreted pilin proteins with LPxTG-type peptidoglycan anchors, which is not present in the highly syntenous genome of Lb. rhamnosus LC705 (Kankainen et al., 2009). SpaA is the major scaffolding protein upon which the minor pili proteins SpaB and SpaC are attached. Using insertional inactivation of spaC, a truncated SpaC protein was produced which resulted in cells with a greatly reduced binding to human mucus (Kankainen et al., 2009). The authors suggest that the presence of SpaC-containing pili (Fig. 2) may possibly explain the longer persistence of this strain in the GI tract than strain LC705. Together with the high potential for sugar uptake and metabolism, this may explain probiotic effects of these Lb. rhamnosus strains. Identification of pili in L. rhamnosus GG by immunogold high-resolution electron micrography. Multiple pili are shown with gold-labelled SpaC proteins. Reproduced with permission from Kankainen and colleagues (2009). Ingested probiotic microbes themselves will react to the new environment of the intestine and change their gene expression accordingly. Transcriptional responses of bifidobacteria to human and formula milk have been described in in vitro and in vivo experiments, the latter from faecal samples of infants (Gonzalez et al., 2008; Klaassens et al., 2009). Carbohydrate metabolism genes are commonly upregulated, and include enzymes for degradation of complex plant carbohydrates, which are poorly digested by the host or other intestinal microbes (Klaassens et al., 2009), and for metabolism of mucin and HMOs (Gonzalez et al., 2008). In addition, putative genes were upregulated for cell-surface type 2 glycoprotein-binding fimbriae that are implicated in attachment and colonization in the intestine (Gonzalez et al., 2008). In vitro transcriptional response of Lactobacillus reuteri ATCC55730, a strain marketed for probiotic usage, to bile stress has been described (Whitehead et al., 2008). Upregulation was seen for genes involved in multidrug transport, membrane/cell wall stress, oxidative stress, DNA damage and protein denaturation. Transcription and comparative genomics analysis of Lb. johnsonii NCC533, an isolate characterized by long gut persistence, identified three genetic loci that were specifically expressed in the jejunum of mice mono-colonized with this strain, encoding a PTS-type sugar transporter, glycosyltransferases and an IgA-type protease (Denou et al., 2008). Several years ago, a very elegant resolvase-based in vivo expression technology was developed to study specific in vivo gene expression in L. plantarum WCFS1, using the mouse GI tract as a model system (Bron et al., 2004). This has now been followed up by whole genome transcriptome profiling of strain WCFS1 during colonization of the caeca of germ-free mice fed either standard low-fat rodent diet rich in complex plant polysaccharides or a Western diet rich in simple sugars and fats (Marco et al., 2009). Numerous carbon metabolism pathways of L. plantarum were upregulated on both diets, including uptake and utilization of raffinose, cellulose, maltose, lactose/galactose, sucrose, melibiose, sugar alcohols and sialic acid. Sialic acid is a common component of (human) gut glycoproteins. Host responses to potential probiotics have recently been described in intervention studies in healthy human volunteers. Duodenal mucosa was sampled after intraduodenal infusion (Troost et al., 2008) or oral ingestion (van Baarlen et al., 2009) of L. plantarum WCFS1. The continuous perfusion study showed that after prolonged exposure, mucosal cells switched to a more proliferative phase with upregulation of genes involved in lipid metabolism, cellular growth and development. Cell death and immune responses were triggered, but cell-death executing cells or inflammatory signals were not expressed. In the second study, consumption of live L. plantarum cells showed striking modulation of NF-κB-dependent pathways in mucosal cells, and identified cellular pathways that correlated with the establishment of immune tolerance in healthy adults (van Baarlen et al., 2009). Figure 3 summarizes some of the mechanistic events underlying probiotic effects that are beginning to be understood from these in vitro and in vivo studies. Bacterial and host effector molecules with potential probiotic effects. Lactobacillus strains are able to induce IL-10-producing, regulatory T cells (T reg) through DC-SIGN interaction (1). They can also induce hyporesponsive CD4+ T-cell populations after DC interaction (2). Lipoteichoic acid (LTA) composition is responsible for the differential modulation of cytokine production (3). Modulation of inflammatory responses by inactivation of the NF-κB signalling pathway is achieved through proteasome inhibition after IEC recognition of soluble probiotic components (4) or after recognition of bacterial motifs (e.g. CpG DNA by TLR9 receptors) (6). The induction of Hsps (either via 4 or 5) stabilizing the actin cytoskeleton would strengthen the mucosal barrier. Pathogen attachment and growth could be counteracted by strains possessing mannose adhesins (7) or by induction of hBD2 in IECs (8). M cell is an epithelial cell specialized in antigen uptake and transport. Reproduced and adapted with permission from Marco and colleagues (2006), Elsevier Ltd. Bifidobacterium longum DJO10A, an intestinal isolate, was shown to lose functionality by gene loss after prolonged pure culture (Lee et al., 2008). It would appear that when growing in the competitive environment of the colon the cells retained some important functionalities predicted to be involved in diverse traits pertinent to the human intestinal environment, specifically oligosaccharide and polyol utilization, arsenic resistance and bacteriocin production. The targeted loss of genomic regions was experimentally validated when growth of the intestinal B. longum in the laboratory for 1000 generations resulted in two large deletions, one in a bacteriocin-encoding region, analogous to a predicted deletion event in the commercial strain B. longum NCC2705 (O'Sullivan, 2008). This deletion strain showed a significantly reduced competitive ability against Clostridium difficile and Escherichia coli. The deleted region was between two IS30 elements which were experimentally demonstrated to be hyperactive within the genome. Hence, deletion of genomic regions, often facilitated by mobile elements, allows bifidobacteria to adapt to fermentation environments in a very rapid manner (two genome deletions per 1000 generations) and the concomitant loss of possible competitive abilities in the gut. This has implications for industry, because the claims for the use of a probiotic need to be fully substantiated. One of the most remarkable probiotic discoveries was made by the German Alfred Nissle in 1917 in World War I. Life in the trenches was dangerous and not just from the fighting. Disease was rife, especially enterocolitis (inflammation of the small and large intestine) caused by outbreaks of shigellosis. One soldier did not succumb to the disease and Nissle isolated from his faeces a bacterium with which he successfully treated other soldiers. Escherichia coli Nissle 1917 is still in use and is one of the few examples of a non-LAB probiotic (Mutaflor) (Table 1). At present, many of the commercial probiotic strains originate from the intestine of healthy infants and adults. Current research focuses on the determination of the characteristics these bacteria use to survive and compete successfully in the intestine, and with this knowledge more effective probiotic strains can be identified. To speed up this search, numerous gut metagenomic sequencing efforts are ongoing world-wide to identify potential new probiotic candidates (Gill et al., 2006; Kurokawa et al., 2007). See also the Human Gut Metagenome Initiative (http://www.international.inra.fr/press/mapping_the_human_intestinal_metagenome) and the Human Gut Microbiome Initiative (Gordon et al., 2006) (http://genomeold.wustl.edu/hgm/HGM_frontpage.cgi). Perhaps the future will bring us health-promoting drinks containing mixtures of many probiotic strains, much like the cocktails used these days for vaccination against infectious diseases. And what will be the next hype? Memory-enhancing drinks would definitely be a commercial success on quiz night in the pub! We thank Paul O'Toole (UC Cork) for providing the original version of Fig. 1, and Maria Marco (UC Davis) for the original of Fig. 3. This project was carried out within the research programmes of the Kluyver Centre for Genomics of Industrial Fermentation and the Netherlands Bioinformatics Centre, which are part of the Netherlands Genomics Initiative/Netherlands Organization for Scientific Research.