Quick viewing(Text Mode)

Recent Insight in Α-Glucan Metabolism in Probiotic Bacteria

Recent Insight in Α-Glucan Metabolism in Probiotic Bacteria

Biologia 69/6: 713—721, 2014 Section Cellular and Molecular Biology DOI: 10.2478/s11756-014-0367-7 Review

Recent insight in α-glucan in probiotic bacteria

Marie S. Møller1,YongJunGoh2, Alexander H. Viborg1,Joakim M. Andersen1, Todd R. Klaenhammer2,BirteSvensson1 &MaherAbou Hachem1*

1Enzyme and Protein Chemistry, Department of Systems Biology, Technical University of Denmark, Søltofts Plads, Building 224,DK-2800 Kgs. Lyngby, Denmark; e-mail: [email protected] 2Department of Food, Bioprocessing and Nutrition Sciences, North Carolina State University, Box 7624, Raleigh, NC 27695 USA

Abstract: α-Glucans from bacterial exo- or diet, e.g., resistant , legumes and honey are abundant in the human gut and fermentation of resistant fractions of these α-glucans by probiotic lactobacilli and bifidobacteria impacts human health positively. The ability to degrade polymeric α-glucans is confined to few strains encoding extra- cellular amylolytic activities of (GH) family 13. Debranching of the subfamily GH13 14 are the most common extracellular GH13 in lactobacilli, whereas corresponding enzymes are mainly α- and amylopullulanases in bifidobacteria. Extracellular GH13 enzymes from both genera are frequently modular and pos- sess starch binding domains, which are important for efficient and possibly to mediate attachment of cells to starch granules. α-1,6-Linked glucans, e.g., isomalto-oligosaccharides are potential prebiotics. The enzymes targeting these glucans are the most abundant intracellular GHs in bifidobacteria and lactobacilli. A phosphoenolpyruvate-dependent phos- photransferase system and a GH4 phospho-α-glucosidase are likely involved in metabolism of isomaltose and isomaltulose in probiotic lactobacilli based on transcriptional analysis. This specificity within GH4 is unique for lactobacilli, whereas canonical GH13 31 α-1,6- active on longer α-1,6-gluco-oligosaccharides are ubiquitous in bifidobacteria and lactobacilli. Malto-oligosaccharide utilization operons encode more complex, diverse, and less biochemically understood ac- tivities in bifidobacteria compared to lactobacilli, where important members have been recently described at the molecular level. This review presents some aspects of α-glucan metabolism in probiotic bacteria and highlights vague issues that merit experimental effort, especially oligosaccharide uptake and the functionally unassigned enzymes, featuring in this important facet of turnover by members of the gut microbiota. Key words: prebiotic; oligosaccharide uptake; ATP-binding cassette transport system; phosphotransferase system; tran- scriptional analysis; isomalto-oligosaccharide; malto-oligosaccharide. Abbreviations: ABC, ATP-binding cassette; CAZy, -Active enZymes database; CBM, carbohydrate binding module; GH, glycoside hydrolase; GI, ; GT, glycosyl family; IMO, isomaltooligosaccharide; PTS, phosphotransferase transport system; RS, resistant starch.

Introduction gut microbiota is dominated by mainly four bacterial phyla with Firmicutes being the most abundant fol- The intimate association between microbes and humans lowed by Bacteroidetes, Actinobacteria and Proteobac- (and other animals) has been increasingly recognized teria (Arumugam et al. 2011). This lower complexity at as a major factor affecting the co-evolution, physiol- phylum level belies the true diversity of the gut micro- ogy, and metabolic interplay between both classes of biota, that comprises an excess of 100 species (varies ac- organisms (Ley et al. 2008; McFall-Ngai et al. 2013). cording to study and protocol used) (Faith et al. 2013) Humans consist of their own somatic cells only until and which is unique to each individual. Common to birth, but in conjunction with delivery the newborn all humans, however, is that imbalance in the gut mi- are inoculated with microbes that proliferate to estab- crobiota (dysbiosis) as compared to the status-quo at- lish vast and complex communities in the gastrointesti- tained in the early years of life, is associated with seri- nal tracts (GI) and at other sites of the body. The ous disorders including obesity, diabetes, bowel cancer, gut microbial community develops rapidly in the new- and allergies ( Wallace et al. 2011; Kootte et al. 2012; born and stabilizes after two to three years (Morgan et Sommer & Baeckhed 2012). The precise composition al. 2013), forming one of the most densely populated of the gut microbiota is governed by a variety of fac- ecological niches in nature (Eckburg et al. 2005). The tors, e.g., age, genetics, method of delivery and lifestyle

* Corresponding author

c 2014 Institute of Molecular Biology, Slovak Academy of Sciences 714 M.S. Møller et al.

(Nicholson et al. 2012). Despite the apparent stability α-1,6-gluco-oligosaccharides possess prebiotic poten- of the gut microbiota through adulthood, it remains dy- tial as judged by preferential stimulation of probi- namic being able to respond to perturbations through otic bacteria and improved bowel function (Kaneko et changes in its composition reflected at the species level al. 1994, Yen et al. 2011). Additionally, a recent re- and in the ratio between different phyla, which has been port suggests that RS resulted in significant increase demonstrated to be of significance to health and dis- in lactic acid production and a change in the Bac- ease risks (Clemente et al. 2012). An obvious example teroidetes/Firmicutes ratio in fecal in vitro fermen- of such perturbations is changes in diet, which has been tations, supplemented with the probiotic strain Lac- shown to alter the composition of the microbiota (Scott tobacillus acidophilus NCFM (Knudsen et al. 2013). et al. 2013). These data are consistent with the importance of α- Humans possess remarkably limited saccharolytic glucan metabolism in probiotic bacteria, which is also machinery being only able to hydrolyze glycosidic reflected by the notably high representation of glyco- bonds in , and to a certain extent starch side hydrolase (GH) family 13 genes that are pivotal in by their digestive enzymes (Cantarel et al. 2012), ren- α-glucan metabolism by probiotic bacteria as compared dering most essentially non-digestible to hu- to other GH families according to the mechanism and mans. On the other hand, human diet contains large sequence-based Carbohydrate-Active enZymes (CAZy) quantities of glycans in cereals, fruits and vegetables, classification system (http://www.cazy.org/) (Cantarel most of which reach the distal gut (colon) intact and are et al. 2009). harvested by the gut microbiota to generate metabolic Considering the importance of probiotic bacteria fuel. This is in accord with the important role that gly- in the context of human health and the interest in non- can metabolism plays in modulating the composition of invasive manipulation of the gut microbiota compo- the gut microbiota, and the inherent impact of this on sition for therapeutic purposes, surprisingly little at- health (Lozupone et al. 2008; Muegge et al. 2011; Scott tention has been dedicated to α-glucan metabolism et al. 2013). in probiotic bacteria and many aspects of this part Starch, which is an exclusively α-glucan polymer of their glycan metabolism remain ill-understood. In organized in supramolecular insoluble semicrystalline the present review, we survey the content of GH13 granules (Tester et al. 2004), is the most abundant gly- in genomes of genera harbouring probiotic strains and can in human diet. Although humans possess the en- highlight recent insight into the routes of uptake and zymatic machinery for starch breakdown (mainly the metabolism of selected α-glucans. Recent work on α-1,4-glycosidic bonds), the degradability of this poly- metabolism in probiotic lactobacilli, which is mer varies considerably based on botanic origin and fine of relevance to fitness in the highly competitive and structural details, especially crystal-packing (Blazek & dynamic niche of the gut is also presented. Gilbert 2010). Thus, significant amounts of dietary re- sistant starch (RS) reach the distal colon where they are A genomic survey of GH13 α-glucan active fermented by members of the gut microbiota (Fuentes- CAZymes in Lactobacillus and Bifidobacterium: Zaragoza et al. 2011). Other dietary sources of α- a snapshot into α-glucan metabolism potential glucans likely to reach the colon intact are α-1,6- linked gluco-oligosaccharides (often designated as iso- At present, the family GH13 includes the vast major- maltooligosaccharides; IMOs) from legumes and honey ity of α-glucan-active enzymes in the CAZy database (Goffin et al. 2011), as these glucans are non-digestible (>15,800 sequences), making this family a good probe for humans except for the disaccharide isomaltose (α- for metabolic capabilities in relation to α-glucans. d-Glcp-(1,6)-d-Glcp). Beside dietary intake, capsular This section focuses on the distribution of the GH13 polysaccharides or exo-polysaccharides produced by CAZymes in the Lactobacillus and Bifidobacterium, members of the gut bacterial community (Duboc & which casts light on the commonalities and differences Mollet 2001) (or acquired via bacterial food contam- in the enzymology of α-glucan processing from a ge- inants) comprise another class of α-glucan substrates nomic perspective. Other selected CAZy families with that are available to support bacterial growth in the important activities relevant to α-glucan metabolism gut niche. will be mentioned later in other sections, but a com- Specific groups of gut microbiota, mainly belong- prehensive analysis of these is beyond the scope of the ing to the Lactobacillus and Bifidobacterium genera, review. have been shown to confer a variety of health benefits All GH13 entries from annotated genomes of lac- including pathogen exclusion and alleviation of inflam- tobacilli and bifidobacteria were retrieved from CAZy. matoryboweldisease,coloncancer, and allergies (Leyer At the time of the analysis, 315 and 326 entries were et al. 2009; Tang et al. 2010; Whelan 2011). These retrieved from 32 Bifidobacterium and 58 Lactobacil- health promoting probiotic bacteria have been shown to lus genomes, respectively. Thus, GH13 is the largest be selectively stimulated by a variety of non-digestible family based on the average GH13 number per genome β-linked oligo- and, to a less extent, poly-saccharides (9.8 and 11 in Bifidobacterium and Lactobacillus,re- acting as prebiotics, e.g., galacto-oligosaccharides, xylo- spectively) emphasizing the importance of α-glucan oligosaccharides and fructo-oligosaccharides (de Vrese metabolism in large subsets of lactobacilli and bifi- & Schrezenmeir 2008; Rastall 2010). Interestingly, dobacteria. The abundance of GH13, however, seems α-Glucan metabolism in probiotic bacteria 715 to be species and strain dependent, which is informa- tive from a niche adaptation point of view. For ex- ample, milk adapted Lactobacillus strains, e.g., Lacto- bacillus fermentum CECT 5716 and L. fermentum F-6 possess only one GH13 and the meat adapted Lactobacillus sakei subsp. sakei 23K possesses two, as compared to 7-9 GH13 members in the gut adapted acidophilus cluster lactobacilli, e.g., the characterized probiotic L. acidophilus NCFM (Sanders & Klaenham- mer 2001). The same applies in bifidobacteria, e.g., Bi- fidobacterium asteroides PRL2011 from insect (honey bee) gut has the lowest GH13 arsenal of all bifidobac- teria (4 sequences). Interestingly, there is a fairly large proportion of GH13 enzymes, which are not assigned into a subfam- ily (subfamilies in a GH are assigned based on primary structure similarities and members of a subfamily are often functionally related; Stam et al. 2006) and more so in Lactobacillus than in Bifidobacterium (Fig. 1A). Some of these enzyme sequences represent known speci- ficities, but are too distant from characterized counter- parts to enable reliable subfamily assignment. Other se- quences are expected to confer novel activities that re- main to be discovered. Further experimental work is re- quired to illuminate the biochemical properties of func- tionally unassigned enzyme sequences and their role in α-glucan turnover in these two important genera. Another notable observation is the high repre- sentation of subfamily 31 (GH13 31) in the GH13 repertoires of both Lactobacillus and Bifidobacterium, mounting almost to 30% of all GH13 members in lactobacilli (Fig. 1A). This subfamily contains α- 1,6-glucosidases that confer the utilization of α-1,6- containing gluco-oligosaccharides (IMOs and panose), consistent with the prebiotic effect of this type of glu- cans that endow probiotic bacteria with a competitive edge due to their preferential fermentation (Kaneko et al. 1994; Yen et al. 2011; Møller et al. 2012). Other subfamilies that are highly represented in both gen- era are GH13 18 sucrose and GH13 9 1,4-α-glucan branching enzymes (Fig. 1A). The former subfamily confers the phosphorolysis of sucrose moi- eties including those derived from soybean oligosaccha- rides of the raffinose family, which are also suggested to be prebiotics (Fredslund et al. 2011; Andersen et al. 2012), whereas the branching enzymes of the lat- ter subfamily GH13 9 are involved in glycogen synthe- sis (Goh & Klaenhammer 2013). Some subfamilies are uniquely encountered in Lactobacillus, e.g.,GH1320 maltogenic α-amylases that are frequently present in Fig. 1. (A) Distribution of GH13 subfamilies (% of total) as as- signed in the CAZy database (http://www.cazy.org/). At the the malto-oligosaccharide metabolism operons in lac- time of analysis 315 and 326 entries were retrieved from bifi- tobacilli (Nakai et al. 2009; Møller et al. 2012), while dobacteria (black bars) and lactobacilli (grey bars), respectively. others are unique for bifidobacteria, e.g., predicted ex- The GH13 statistics give insight into the representation of cer- tracellular α-amylases of GH13 32 and modular amy- tain subfamilies, and thereby indicate the genomic potential of these two genera to catabolize specific α-glucan bonds. (B) Sub- lopullulanases containing both GH13 14 and GH13 32 family distribution amongst putative extracellular GH13 entries catalytic modules (Fig. 1A, Fig. 2). The diversity of as predicted using SignalP (see text) from lactobacilli (9 entries). subfamilies is larger in bifidobacteria, as compared to (C) Predicted extracellular GH13 enzymes from bifidobacteria lactobacilli (Fig. 1B,C), which can be partially ex- (9 entries). Modular putative extracellular GH13 enzymes fre- quently contain an α- catalytic domain in bifidobacteria, plained by: (i) larger ratio of unassigned sequences whereas extracellular GH13 enzymes of lactobacilli are mainly in lactobacilli which will be eventually assigned into pullulanases. 716 M.S. Møller et al.

Fig. 2. Schematic overview of the domain organisation of putative extracellular GH13 proteins from lactobacilli and bifidobacteria. The domain assignments are based on output from the Conserved Domain Database (CDD; http://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml), together with information from CAZy and related literature. Members sharing the same domain organisation are listed once, and the producing organisms are enlisted in the figure together with the NCBI accession numbers. Signal peptides as predicted by SignalP are in yellow; carbohydrate binding modules (CBM family numbers are indicated) are in green tones; domains of unknown function (DUF) in grey; catalytic domain and associated C-domains are in blue tones; surface layer association protein abbreviated as SLAP in purple; α-amylase-like catalytic domain and associated C-domains are in red tones; Q-P-T and A/T-A-N-S-T repeat regions are in cyan; bacterial immunoglobulin-like domains is in pink. new subfamilies when biochemical data becomes avail- sary for the utilization of both granular starch, and able; and (ii) more extensive occurrence of α-glucan polymeric components thereof, as these large substrates enzymes from other GH families in lactobacilli, e.g., need to be degraded to smaller oligosaccharides to allow GH4 (contains both α-glucosidases and 6’-phospho-α- efficient uptake by glycan specific transporters. There- glucosidases) and GH65 (contains di-glucosyl phospho- fore, analysis of the extracellular GH13 content is in- rylases). Additional biochemical data is needed to pre- formative to assess the amylolytic potential of differ- cisely compare the metabolic reach of the gut adapted ent organisms. We analysed the putative extracellular species of lactobacilli and bifidobacteria. GH13 in lactobacilli and bifidobacteria by submitting all the retrieved GH13 sequences to the SignalP server Extracellular secreted GH13 and starch utilization by to predict the presence of secretion transit peptides probiotic bacteria (http://www.cbs.dtu.dk/services/SignalP/). Only 9 se- The best studied starch utilization system in the gut quences from either Lactobacillus or Bifidobacterium niche is from species of the dominant commensal (sym- were predicted to be extracellular, which corresponds to biont) genus Bacteroides (Cameron et al. 2012). By con- about 2.9% of the GH13 entries in each group. This low trast, the starch utilization capabilities of lactobacilli abundance of extracellular amylolytic enzymes suggests and bifidobacteria are less well documented. The pro- that the ability to degrade starch or polymeric starchy duction of amylolytic extracellular enzymes is neces- substrates is confined to a few strains within these taxa. α-Glucan metabolism in probiotic bacteria 717

Notably, most extracellular GH13 CAZymes from Bi- the probiotic L. acidophilus NCFM and in Bifidobac- fidobacterium contained putative α-amylase catalytic terium animalis subsp. lactis Bl-04 (Barrangou et al. domains from subfamily 28 or 32 (the role of sub- 2009), which represents a widely used probiotic sub- family 3 modules in bifidobacteria is currently un- species within the genus due to its technological and known), whereas the α-1,6-debranching pullulanases of health promoting properties ((Loquasto et al. 2013). GH13 14 were dominant in lactobacilli, especially those In L. acidophilus NCFM, the malto-oligosaccharide acidophilus group members that are adapted to the gene cluster encodes a transcriptional regulator (MalR), human gut (Fig. 1B,C, Fig. 2). The rareness of α- acetate kinase (AckA), a GH13 of unknown function amylases in lactobacilli is consistent with only 11 of (MalL), a GH13 20 maltogenic α-amylase (MalN), a 96 species reported to degrade starch (Petrova et al. GH65 maltose (MalP), a phosphogluco- 2013), in a highly strain dependent manner. The ma- mutase (PgmB) and an ATP-binding cassette (ABC) jority of the extracellular GHs were modular featuring transport system (Msmk, MalE, MalF, MalG) (Nakai domains of unknown function, cell attachment mod- et al. 2009). This arrangement is highly conserved ules, starch binding domains assigned into carbohy- across species in the genus with the main difference drate binding module (CBM) families 25, 26, 41, and 48 being that in other species, the gene cluster also in- (Christiansen et al. 2009) (Fig. 2) appended to catalytic cludes a GH13 31 glucan α-1,6-glucosidase that catal- GH13 modules. The wide occurrence of starch binding yses the of IMOs including panose. Re- domains in these enzymes suggests that binding to the cently, the glucan α-1,6-glucosidase from L. acidophilus is functionally important for catalysis, which NCFM, which is encoded by a separate genetic lo- has been shown in the α-amylases from L. amylovorus cus than the malto-oligosaccharide operon, has been and L. plantarum (Rodriguez-Sanoja et al. 2005). In a biochemically and structurally characterized providing competitive niche as the gut, another important func- the first insight into the enzymology α-1,6-glucosidases tion of CBMs is plausibly to mediate attachment be- from lactobacilli (Møller et al. 2012). The frequent co- tween the cells (via cell displayed modular enzymes) localization of α-1,6-glucosidase genes in the malto- and supramolecular substrates such as starch granules. oligosaccharide operon in most lactobacilli raises the We are currently investigating this hypothesis in the question of whether the same ABC transport system cell attached modular pullulanase from L. acidophilus is able to confer uptake of both IMOs and malto- NCFM (Fig. 2), which is essentially inactive on starch oligosaccharides, but experimental evidence is needed granules (M.S. Møller et al., unpublished data). This to verify this. organism is unable to grow on starch or other larger Another interesting enzyme encoded by the malto- glucans, thus the role of this enzyme is intriguing and oligosaccharide operon in lactobacilli is the GH65 mal- affords interrogation. Theoretically this enzyme could tose phosphorylase (MalP). We have provided exper- associate to starch granules via the CBM41 domain, imental evidence for the functionality of this enzyme but this needs experimental corroboration. Binding of from L. acidophilus NCFM (Nakai et al. 2009). The en- bacterial cells to starch granules (the site of primary zyme also proved very useful as a synthetic tool for degradation) may provide a competitive advantage by the synthesis of a variety of di-saccharides with an promoting cross-feeding on smaller α-glucans (di- and α-glucosyl residue at the non-reducing end including tri-saccharides) that escape capture by other organisms. some non-naturally occurring compounds, e.g., α-Glcp- Further work is required to investigate this possibility. (1,4)-Manp, α-Glcp-(1,4)-Xylp and α-Glcp-(1,4)-l-Fucp (Nakai et al. 2009). Surprisingly, some of these com- Metabolism of α-1,4- and α-1,6-gluco- pounds produced by this enzyme selectively stimulated oligosaccharides the growth of B. animalis subsp. lactis Bl-04 as com- pared to two other probiotic strains and a commen- Probiotic lactobacilli and bifidobacteria are able to sal strain in vitro suggesting that this approach may grow on different α-gluco-oligosaccharides (Sanz et al. be a viable one to engineer novel prebiotic compounds 2005; Sarbini et al. 2013). Probiotic bacteria can ac- (Vigsnæs et al. 2013). This also suggests that some the quire these substrates either directly from the host diet, uptake systems possess a degree of promiscuity, which by cross feeding from primary degraders (Flint et al. can be exploited in design of novel prebiotic oligosac- 2007) or more rarely by the action of their own ex- charides. tracellular GH13 enzymes. In all cases, the strategy Whole genome microarray transcriptional analy- adopted by probiotic bacteria is to rely on different sis suggested a novel route for the uptake of isoma- classes of transport systems for efficient uptake of di- ltose, isomaltulose (α-d-Glcp-(1,6)-d-Fruf)andpossi- and oligo-saccharides, which is followed by complete de- bly panose in L. acidophilus NCFM, as two loci encod- polymerisation by intracellular and acces- ing a phosphoenoylpyruvate-dependant phosphotrans- sory enzymes (Abou Hachem et al. 2013). Typically, ferase transport system (PTS) and a GH4, respectively, but not invariantly, the genes encoding these transport were highly up-regulated in response to growth on these systems, hydrolases and transcriptional regulator are saccharides (Andersen et al. 2012). Uptake through clustered together in functional operons. PTS systems is concomitant with phosphorylation of We have investigated the routes of utilization of the internalised glycan (mainly di-saccharides) at the both α-1,4-malto-oligosaccharides and α-1,6-IMOs in 6’-hydroxyl moiety. These phosphorylated substrates 718 M.S. Møller et al. are recognized by special GH4 glycosidases. The isoma- port systems and the functionally ambiguous hydrolases ltose responsive GH4 which resides on a separate locus to establish a more robust understanding of which and different from the PTS uptake system, is currently an- how α-glucans are internalised and processed by probi- notated as a maltose 6’-phospho-glucosidase. However, otic bifidobacteria. the transcriptomic data and sequence homology to a characterized enzyme from Lactobacillus casei ATCC Glycogen metabolism in Lactobacillus 334 active on phosphorylated sucrose isomers (66% identity) suggest that the physiologically relevant sub- Glycogen synthesis in bacteria involves -1- strate for this enzyme is not maltose, since the enzyme phosphate adenylyltransferase (GlgC or GlgCD), a from L. casei ATCC 334 was repressed upon growth on glycosyl transferase (GT) family 5 sucrose and maltose (Thompson et al. 2008). Nonethe- (GlgA) and glycogen-branching enzyme of GH13 9 less, the enzyme from L. casei was active on 6’-phospho- (GlgB). Glucose-1-phosphate serves as a substrate for maltose in vitro, albeit displaying five-fold lower spe- ADP-glucose synthesis catalyzed by GlgC or GlgCD. cific activity as compared to 6’-phospho-isomaltulose. Then, GlgA catalyzes the transfer of glucosyl units Phylogenetic analysis of all bacterial GH4 members from ADP-glucose to the elongating chain of lin- showed that the GH4 from L. acidophilus NCFM and ear α-1,4-glucan, where GlgB subsequently transfers the homologue from L. casei ATCC 334 reside on an a fragment of the glucan chains to form frequent α- exclusively Lactobacillus populated cluster of phospho- 1,6-branches characteristic to the glycogen structure. α-glucosidases (Abou Hachem et al. 2013). Interest- Glycogen catabolism involves ingly, GH4 sequences from other lactobacilli including of GT35 (GlgP) and glycogen-debranching enzyme of other L. casei strains cluster more closely to other char- GH13 (GlgX). The glycogen phosphorylase catalyzes acterized phospho-α-glucosidases showing high activ- the phosphorolytic breakdown of the glucan chain, ity on 6’-phospho-maltose as compared to 6’-phospho- whereas the debranching of limit dextrins produced isomaltulose or 6’-phospho-isomaltose (Fig. 3). Alto- by the action of the phosphorylase is carried out by gether, this suggests that GH4 phospho-α-glucosidases the debranching enzyme. Glycogen metabolism con- from gut adapted lactobacilli may have a preference for tributes to energy storage and other physiological func- phosphorylated isomaltose and other sucrose isomers tions in some , including colonization persis- rather than 6’-phospho-maltose, as shown for some or- tence (Jones et al. 2008; Busuioc et al. 2009), and serves ganisms from the soil niche and human pathogens. as a carbon capacitor that regulates downstream carbon The diversity of α-glucan metabolism capabilities fluxes (Belanger & Hatfull 1999). In addition, glyco- seems to be higher in bifidobacteria with regard to gen metabolism is involved in major cellular processes, both genetic organization and the type of transporters such as carbohydrate metabolism, energy production, and hydrolases. A notable difference from lactobacilli stress response and cell-cell communication (Eydallin is the co-localization and co-transcription of GH13 31 et al. 2007; Eydallin et al. 2010). Altogether, available and GH36 1 α-galactosidase genes responsible for the data suggest that glycogen metabolism potentially con- metabolism of IMOs and galactosides, respectively (An- tributes to the survival and probiotic activities of lac- dersen et al. 2013). The presence of genes encoding an tobacilliand possibly other probiotic microorganisms in ABC transport system within this operon is suggestive the GI environment. of a dual α-1,6-glycoside specificity of this transport The first example of a functional glycogen meta- system. Currently we are investigating the biochemical bolism pathway in probiotic lactic acid bacteria (LAB) and structural features of the solute binding protein me- was recently demonstrated in L. acidophilus NCFM diating the capture of these substrates for translocation (Goh & Klaenhammer 2013). The pathway is encoded via this ABC system. by an operon, designated as the glg operon, and consists With regard to α-1,4-maltodextrins, the picture of glgB, glgCD, glgA, glgP, amy (annotated as a putative is more complicated. Transcriptional analysis identi- α-amylase, but is more likely to have a debranching ac- fied a locus harbouring genes for two ABC transport tivity) and pgm (putative phosphoglucomutase) genes. systems, a GH13 30 α-glucosidase, a GH13 14 pullu- The co-transcription of these genes indicates that glyco- lanase, a putative 4-α-glucanotransferase of GH77 and gen synthesis and degradation occur in parallel and may a GH13 member of unknown function (Andersen et play a role in regulating the carbon flow inside the cells. al. 2013). Only one of the solute binding proteins as- Similar glg operons were found only in specific Lacto- sociated with one of the ABC transport systems (lo- bacillus species commonly associated with mammalian cus tag number Balac 1565) shows a distant relation- hosts or natural environments. The proposed niche- ship to the canonical maltose binding proteins, whereas specific function of glycogen metabolism was further the second solute binding protein is not homologous to highlighted by the inter-strain comparison among Lac- known malto-dextrin specific binding proteins. Interest- tobacillus helveticus and Lactobacillus delbrueckii ssp. ingly the presence of these two ABC import systems is bulgaricus, whereby the pathway was identified in a conserved in the B. animalis species and in other bifi- probiotic strain and a human isolate of L. helveticus dobacteria, e.g., Bifidobacterium thermophilum RBL67 and L. bulgaricus, respectively, but not in their dairy- and Bifidobacterium dentium Bd1. These observations associated counterparts (Goh & Klaenhammer 2013). merit the biochemical characterisation of these trans- In L. acidophilus,expressionoftheglg operon and α-Glucan metabolism in probiotic bacteria 719

Fig. 3. The phylogenetic tree calculated for all CAZy GH4 entries from Lactobacillus in addition to 6’-phospho-α-glucosidases with pref- erence for 6’-phospho-maltose (EC 3.2.1.122). The amino acid sequences were retrieved and aligned using the MAFFT program and the phylogenetic tree was calculated using the ClustalW, both interfaced at the European Bioinformatics Institute (http://www.ebi.ac.uk/). The tree was rendered in Dendroscope (http://ab.inf.uni-tuebingen.de/software/dendroscope/). For clarity, not all sequences were dis- played. The sequences forming the tree are labelled with their GenBank accession and some labels are omitted for clarity. The black subtree is populated by the GH4 sequence from L. acidophilus NCFM and L. casei ATCC334, which were implicated in the metabolism of isomaltose and sucrose isomers by transcriptomics (L. acidophilus NCFM) and proteomics (L. casei ATCC334; see Thompson et al. 2008). The grey subtree contains the rest of lactobacilli GH4 sequences together with characterized 6’-phospho-α-glucosidases with preference for 6’-phospho-maltose (labelled with full organism names for clarity). glycogen accumulation were carbon source- and growth Raffinose also induced an overall higher temporal ex- phase-dependent, and were repressed by glucose (Goh pression of the glg operon throughout growth compared & Klaenhammer 2013). The highest intracellular glyco- to . Overall, unlike cells grown on trehalose gen content was observed in early-logarithmic phase or glucose, the raffinose-grown cells showed remarkably cells grown on trehalose, which was followed by a dras- stable intracellular glycogen reserves during prolonged tic decrease of glycogen level prior to entering sta- growth periods. It was speculated that L. acidophilus tionary phase. When the soybean tri-saccharide raf- may maintain a higher level of intracellular glycogen to finose (6’-α-d-galactosyl-sucrose) was provided as the enhance sustainability when growing on a more com- carbon source, glycogen accumulation in the cells grad- plex carbon source, a scenario most likely encountered ually declined following early-log phase and was main- in the GI environment. These data prompted the pro- tained at stable levels throughout stationary phase. posal of a novel role conferred by raffinose and poten- 720 M.S. Møller et al. tially other prebiotic oligosaccharides, namely the en- Andersen J.M., Barrangou R., Abou Hachem M., Lahtinen S., hancement of the residence time of L. acidophilus in Goh Y.J., Svensson B. & Klaenhammer T.R. 2013. Transcrip- the GI tract by inducing the synthesis of glycogen stor- tional analysis of oligosaccharide utilization by Bifidobac- terium lactis Bl-04. BMC Genomics 14: 312. age reserves. Mutational analysis in L. acidophilus con- Arumugam M., Raes J., Pelletier E., Le Paslier D., Yamada T., firmed that glgA and glgB are responsible for glycogen Mende D.R., Fernandes G.R., Tap J., Bruls T., Batto J.M., and are required for efficient growth on raf- Bertalan M., Borruel N., Casellas F., Fernandez L., Gau- finose. Furthermore, deletion of the glgB or glgP gene, tierL.,HansenT.,HattoriM.,HayashiT.,KleerebezemM., Kurokawa K., Leclerc M., Levenez F., Manichanh C., Nielsen which are involved in glycogen synthesis or degrada- H.B., Nielsen T., Pons N., Poulain J., Qin J., Sicheritz-Ponten tion, respectively, resulted in defective growth and in- T., Tims S., Torrents D., Ugarte E., Zoetendal E.G., Wang creased bile sensitivity. The study indicates that glyco- J., Guarner F., Pedersen O., de Vos W.M., Brunak S., Dore gen metabolism contributes to the growth maintenance, J., Weissenbach J., Ehrlich S.D. & Bork P. 2011. Enterotypes 473: bile tolerance and efficient complex carbohydrate uti- of the human gut microbiome. Nature 174–180. Barrangou R., Briczinski E.P., Traeger L.L., Loquasto J.R., lization in L. acidophilus, and reflects the importance Richards M., Horvath P., Coute-Monvoisin A.C., Leyer G., of glycogen metabolism on the probiotic attributes of Rendulic S., Steele J.L., Broadbent J.R., Oberg T., Dudley lactobacilli. E.G., Schuster S., Romero D.A. & Roberts R.F. 2009. Com- parison of the complete genome sequences of Bifidobacterium animalis subsp lactis DSM 10140 and Bl-04. J. Bacteriol. Perspectives 191: 4144–4151. Belanger A.E. & Hatfull G.F. 1999. Exponential-phase glycogen The explosive growth of metagenomic data and the recycling is essential for growth of Mycobacterium smegmatis. myriad of studies exploring the interplay between the J. Bacteriol. 181: 6670–6678. gut microbiota, certain diet types and the human host Blazek J. & Gilbert E.P. 2010. Effect of enzymatic hydrolysis 11: have yielded fascinating insight into the importance of on native starch granule structure. Biomacromolecules 3275–3289. gut microbiota and diet to human health. Manipulat- Busuioc M., Mackiewicz K., Buttaro B.A. & Piggot P.J. 2009. ing the composition of the gut microbiota to promote Role of intracellular in persistence of Strepto- health and diminish the use of to combat en- coccus mutans. J. Bacteriol. 191: 7315–7322. teropathogens, comprises an important objective that Cameron E.A., Maynard M.A., Smith C.J., Smith T.J., Ko- would be promoted by a better molecular understand- ropatkin N.M. & Martens E.C. 2012. Multidomain carbohydrate- binding proteins involved in Bacteroides thetaiotaomicron ing of glycan metabolism in the gut niche. Despite the starch metabolism. J. Biol. Chem. 287: 34614–34625. importance of α-glucan metabolism, some fundamen- Cantarel B.L., Coutinho P.M., Rancurel C., Bernard T., Lombard tal aspects of it remain ill-understood. Future work ad- V. & Henrissat B. 2009. The Carbohydrate-Active EnZymes dressing α-glucan uptake by different microbiota mem- database (CAZy): an expert resource for Glycogenomics. Nu- 37: bers is important to understand niche adaptation re- cleic Acids Res. D233–D238. Cantarel B.L., Lombard V. & Henrissat B. 2012. Complex carbo- garding competition and cross-feeding amongst differ- hydrate utilization by the healthy human microbiome. PloS ent microbiota taxa. The enzymology and molecular One 7: e28742. recognition of several important players in α-glucan Christiansen C., Abou Hachem M., Janecek S., Viksø-Nielsen A., metabolism require more attention to establish a bet- Blennow A. & Svensson B. 2009. The carbohydrate-binding ter understanding of the dynamic metabolic fluxes that module family 20 – diversity, structure, and function. FEBS J. 276: 5006–5029. may contribute to the flexibility and fitness of microbial Clemente J.C., Ursell L.K., Parfrey L.W. & Knight R. 2012. The cells in the competitive gut niche. impact of the gut microbiota on human health: an integrative view. Cell 148: 1258–1270. de Vrese M. & Schrezenmeir J. 2008. Probiotics, prebiotics, and Acknowledgements synbiotics, pp. 1–66. In: Stahl U.D.U.E.B.N.E. (ed.) Food Biotechnology. This research was funded by a FøSu grant from the Dan- Duboc P. & Mollet B. 2001. Applications of exopolysaccharides ish Strategic Research Council to the project “Gene discov- in the dairy industry. Int. Dairy J. 11: 759–768. ery and molecular interactions in pre/probiotics systems”. Eckburg P.B., Bik E.M., Bernstein C.N., Purdom E., Dethlefsen DuPont Nutrition and Health, and North Carolina Dairy L., Sargent M., Gill S.R., Nelson K.E. & Relman D.A. 2005. Foundation are also acknowledged for financial support. Diversity of the human intestinal microbial flora. Science 308: PhD stipends from the Technical University of Denmark 1635–1638. for JMA and AHV. Eydallin G., Montero M., Almagro G., Sesma M.T., Viale A.M., Munoz F.J., Rahimpour M., Baroja-Fernandez E. & Pozueta- Romer J. 2010. Genome-wide screening of genes whose References enhanced expression affects glycogen accumulation in Es- cherichia coli.DNARes.17: 61–71. Abou Hachem M., Andersen J.M., Barrangou R., Møller M.S., Eydallin G., Viale A.M., Moran-Zorzano M.T., Munoz F.J., Fredslund F., Majumder A., Ejby M., Lahtinen S.J., Jacob- Montero M., Baroja-Fernandez E. & Pozueta-Romero J. sen S., Lo Leggio L., Goh Y.J., Klaenhammer T.R. & Svens- 2007. Genome-wide screening of genes affecting glycogen son B. 2013. Recent insight into oligosaccharide uptake and metabolism in K-12. FEBS Lett. 581: 2947– metabolism in probiotic bacteria. Biocatal. Biotransform. 31: 2953. 226–235. Faith J.J., Guruge J.L., Charbonneau M., Subramanian S., See- Andersen J.M., Barrangou R., Abou Hachem M., Lahtinen S., dorf H., Goodman A.L., Clemente J.C., Knight R., Heath Goh Y.J., Svensson B. & Klaenhammer T.R. 2012. Transcrip- A.C., Leibel R.L., Rosenbaum M. & Gordon J.I. 2013. The tional analysis of prebiotic uptake and catabolism by Lacto- long-term stability of the human gut microbiota. Science 341: bacillus acidophilus NCFM. PloS One 7: e44409. 44–53. α-Glucan metabolism in probiotic bacteria 721

Flint H.J., Duncan S.H., Scott K.P. & Louis P. 2007. Interactions Muegge B.D., Kuczynski J., Knights D., Clemente J.C., Gonzalez and competition within the microbial community of the hu- A., Fontana L., Henrissat B., Knight R. & Gordon J.I. 2011. man colon: links between diet and health. Environ. Microbiol. Diet drives convergence in gut microbiome functions across 9: 1101–1111. mammalian phylogeny and within humans. Science 332: 970– Fredslund F., Abou Hachem M., Larsen R.J., Sørensen P.G., 974. Coutinho P.M., Lo Leggio L. & Svensson B. 2011. Crystal Nakai H., Baumann M.J., Petersen B.O., Westphal Y., Schols H., structure of α-galactosidase from Lactobacillus acidophilus Dilokpimol A., Abou Hachem M., Lahtinen S.J., Duus J.O. NCFM: insight into tetramer formation and substrate bind- & Svensson B. 2009. The maltodextrin transport system and ing. J. Mol. Biol. 412: 466–480. metabolism in Lactobacillus acidophilus NCFM and produc- Fuentes-Zaragoza E., Sanchez-Zapata E., Sendra E., Sayas E., tion of novel α-glucosides through reverse phosphorolysis by Navarro C., Fernandez-Lopez J. & Perez-Alvarez J.A. 2011. maltose phosphorylase. FEBS J. 276: 7353–7365. Resistant starch as prebiotic: a review. Starch 63: 406–415. Nicholson J.K., Holmes E., Kinross J., Burcelin R., Gibson G., GoffinD.,DelzenneN.,BleckerC.,HanonE.,DeroanneC. Jia W. & Pettersson S. 2012. Host-gut microbiota metabolic & Paquot M. 2011. Will isomalto-oligosaccharides, a well- interactions. Science 336: 1262–1267. established functional food in Asia, break through the Eu- Petrova P., Petrov K. & Stoyancheva G. 2013. Starch-modifying ropean and American market? The status of knowledge on enzymes of lactic acid bacteria – structures, properties, and these prebiotics. Crit. Rev. Food Sci. Nutr. 51: 394–409. applications. Starch-St¨arke 65: 34–47. Goh Y.J. & Klaenhammer T.R. 2013. A functional glycogen Rastall R.A. 2010. Functional oligosaccharides: application and biosynthesis pathway in Lactobacillus acidophilus: expression manufacture.Annu.Rev.FoodScie.Technol.1: 305–339. and analysis of the glg operon. Mol. Microbiol. 89: 1187–1200. Rodriguez-Sanoja R., Ruiz B., Guyot J.P. & Sanchez S. 2005. Jones S.A., Jorgensen M., Chowdhury F.Z., Rodgers R., Hart- Starch-binding domain affects catalysis in two Lactobacillus line J., Leatham M.P., Struve C., Krogfelt K.A., Cohen P.S. α-amylases. Appl. Environ. Microbiol. 71: 297–302. & Conway T. 2008. Glycogen and maltose utilization by Es- Sanders M.E. & Klaenhammer T.R. 2001. Invited review: The cherichia coli O157 : H7 in the mouse intestine. Infect. Im- scientific basis of Lactobacillus acidophilus NCFM function- mun. 76: 2531–2540. ality as a probiotic. J. Dairy Sci. 84: 319–331. Kaneko T., Kohmoto T., Kikuchi H., Shiota M., Iino H. & Mit- Sanz M.L., Gibson G.R. & Rastall R.A. 2005. Influence of dis- suoka T. 1994. Effects of isomaltooligosaccharides with differ- accharide structure on prebiotic selectivity in vitro.J.Agric. ent degrees of polymerization on human fecal bifidobacteria. Food Chem. 53: 5192–5199. Biosci. Biotechnol. Biochem. 58: 2288–2290. Sarbini S.R., Kolida S., Gibson G.R. & Rastall R.A. 2013. In Knudsen A., van Zanten G.C., Jensen S.L., Forssten S.D., Saari- vitro fermentation of commercial α-gluco-oligosaccharide by nen M., Lahtinen S.J., Bandsholm O., Svensson B., Jespersen faecal microbiota from lean and obese human subjects. Br. J. L. & Blennow A. 2013. Comparative fermentation of insoluble Nutr. 109: 1980–1989. in an in vitro human feces model spiked with Scott K.P., Gratz S.W., Sheridan P.O., Flint H.J. & Duncan S.H. Lactobacillus acidophilus NCFM. Starch-St¨arke 65: 346–353. 2013. The influence of diet on the gut microbiota. Pharmacol. Kootte R.S., Vrieze A., Holleman F., Dallinga-Thie G.M., Zoe- Res. 69: 52–60. tendalE.G.,deVosW.M.,GroenA.K.,HoekstraJ.B.L., Sommer F. & Baeckhed F. 2012. The gut microbiota – masters of Stroes E.S. & Nieuwdorp M. 2012. The therapeutic poten- host development and physiology. Nat. Rev. Microbiol. 11: tial of manipulating gut microbiota in obesity and type 2 227–238. diabetes mellitus. Diabetes Obes. Metab. 14: 112–120. Stam M.R., Danchin E.G.J., Rancurel C., Coutinho P.M. & Hen- Ley R.E., Hamady M., Lozupone C., Turnbaugh P.J., Ramey rissat B. 2006. Dividing the large glycoside hydrolase family R.R., Bircher J.S., Schlegel M.L., Tucker T.A., Schrenzel 13 into subfamilies: towards improved functional annotations M.D., Knight R. & Gordon J.I. 2008. Evolution of mammals of α-amylase-related proteins. Protein Eng. Des. Sel. 19: 555– and their gut microbes. Science 320: 1647–1651. 562. Leyer G.J., Li S., Mubasher M.E., Reifer C. & Ouwehand A.C. Tang M.L.K., Lahtinen S.J. & Boyle R.J. 2010. Probiotics and 2009. Probiotic effects on cold and influenza-like symptom prebiotics: clinical effects in allergic disease. Curr. Opin. Pe- incidence and duration in children. Pediatrics 124: E172– diatr. 22: 626–634. E179. Tester R.F., Karkalas J. & Qi X. 2004. Starch – composition, fine Loquasto J.R., Barrangou R., Dudley E.G., Stahl B., Chen C. structure and architecture. J. Cereal Sci. 39: 151–165. & Roberts R.F. 2013. Bifidobacterium animalis subsp. lactis Thompson J., Jakubovics N., Abraham B., Hess S. & Pikis ATCC 27673 is a genomically unique strain within its con- A. 2008. The sim operon facilitates the transport and served subspecies. Appl. Environ. Microbiol. 79: 6903–6910. metabolism of sucrose isomers in Lactobacillus casei ATCC Lozupone C.A., Hamady M., Cantarel B.L., Coutinho P.M., Hen- 334. J. Bacteriol. 190: 3362–3373. rissat B., Gordon J.I. & Knight R. 2008. The convergence of Vigsnæs L.K., Nakai H., Hemmingsen L., Andersen J.M., Lahti- carbohydrate active gene repertoires in human gut microbes. nen S.J., Rasmussen L.E., Abou Hachem M., Petersen B.O., Proc. Natl. Acad. Sci. USA 105: 15076–15081. Duus J.O., Meyer A.S., Licht T.R. & Svensson B. 2013. McFall-Ngai M., Hadfield M.G., Bosch T.C.G., Carey H.V., In vitro growth of four individual human gut bacteria Domazet-Loso T., Douglas A.E., Dubilier N., Eberl G., on oligosaccharides produced by chemoenzymatic synthesis. Fukami T., Gilbert S.F., Hentschel U., King N., Kjelleberg S., Food Funct. 4: 784–793. Knoll A.H., Kremer N., Mazmanian S.K., Metcalf J.L., Neal- Wallace T.C., Guarner F., Madsen K., Cabana M.D., Gibson G., son K., Pierce N.E., Rawls J.F., Reid A., Ruby E.G., Rumpho Hentges E. & Sanders M.E. 2011. Human gut microbiota and M., Sanders J.G., Tautz D. & Wernegreen J.J. 2013. Animals its relationship to health and disease. Nutr. Rev. 69: 392–403. in a bacterial world, a new imperative for the life sciences. Whelan K. 2011. Probiotics and prebiotics in the management Proc. Natl. Acad. Sci. USA 110: 3229–3236. of irritable bowel syndrome: a review of recent clinical trials Møller M.S., Fredslund F., Majumder A., Nakai H., Poulsen and systematic reviews. Curr. Opin. Clin. Nutr. Metab. Care J.C.N., Lo Leggio L., Svensson B. & Abou Hachem M. 2012. 14: 581–587. Enzymology and structure of the GH13 31 glucan 1,6-α- Yen C.H., Tseng Y.H., Kuo Y.W., Lee M.C. & Chen H.L. 2011. glucosidase that confers isomaltooligosaccharide utilization in Long-term supplementation of isomalto-oligosaccharides im- the probiotic Lactobacillus acidophilus NCFM.J.Bacteriol. proved colonic microflora profile, bowel function, and blood 194: 4249–4259. cholesterol levels in constipated elderly people – a placebo- Morgan X.C., Segata N. & Huttenhower C. 2013. Biodiversity controlled, diet-controlled trial. Nutrition 27: 445–450. and functional genomics in the human microbiome. Trends Genet. 29: 51–58. Received January 22, 2014 Accepted March 13, 2014

Top View