Journal of Bioscience and BioengineeringVOL. 111 No. 2, 178–184, 2011

www.elsevier.com/locate/jbiosc

Brewer's yeast cell wall affects microbiota composition and decreases Bacteroidesfragilis populations in an anaerobic gut intestinal model

Yutaka Nakashimada,1,⁎ Atsuko Michinaka,2 Kentaro Watanabe,1 Naomichi Nishio,1 and Toshio Fujii2

⁎ CorrespondE-mail add

1389-1723/$doi:10.1016/j

Department of Molecular Biotechnology, Graduate School of Advanced Sciences of Matter, Hiroshima University, 1-3-1, Kagamiyama, Higashi-Hiroshima739-8530, Japan1 and Central Laboratories for Frontier Technology, Kirin Holdings co. ltd., Yokohama, 236-0004, Japan2

Received 9 February 2010; accepted 8 September 2010Available online 6 October 2010

Brewer's yeast cell wall (BYC) has been reported to have prebiotic activity that improves the microbiotal composition ofthe human gut. To understand the precise effect of BYC on gut microbiota and its metabolism, we used a three-stagecontinuous-flow reactor system that mimicked the environment of the large intestine. The reactor system was able tomaintain the bacterial community stably for a week. The Bacteroides fragilis population decreased drastically after theaddition of BYC into this system while the number of Lactobacillus was stably maintained. In addition, propionate and acetatelevels increased drastically. This metabolic change correlated with an increase in a number of specific operational taxonomicunits annotated to the genus Veillonella and Megasphaella. These results suggest that BYC affects the composition ofmicrobiota in an in vitro model system.

© 2010, The Society for Biotechnology, Japan. All rights reserved.

[Key words: Yeast cell wall; Prebiotics; In vitro culture system; Biomass; T-RFLP]

0

30

60

90

120

150Vessel 1

0

30

60

90

120

150

Con

cent

ratio

n (m

M)

Vessel 2

0

30

60

90

120

150

0 2 4 6 8 10 12 14

Culture time (d)

Vessel 3

Prebiotics are non-digestible food ingredients that affect the healthof the human host by selectively stimulating potentially beneficialmicrobes, thereby modifying the composition of the gut microbiota(1,2). Prebiotics can reduce the growth of pathogenic or virulentorganisms and induce health-promoting systemic effects (1). To bemost effective, prebiotics must be able to reach the large intestine andmust be utilized specifically by those microbes with proven health-promoting effects (3).

Saccharomyces sensu stricto strains, including S. cerevisiae, S.bayanus, and S. pastorianus, are GRAS (generally recognized as safe)organisms that have been used since ancient times in the production ofmany kinds of fermented foods and beverages. Saccharomyces sensustricto strains are also widely used as probiotics for humans andanimals (4–6). Yeast extracts, comprised of nucleic acids (e.g., inosinicacid) and amino acids, including glutamic acid, have also been used asfood seasonings. Intact dried yeast itself has gained increasing attentionas a healthful and nutritious food. The yeast cell wall, consisting mainlyof water-insoluble dietary fiber (β-glucan and α-mannan), and ofindigestible protein bound to the α-mannan, is also considered abeneficial component of foods. Administration of cell walls of brewer'syeast into experimentally constipated rats improved defecation (7), andyogurt supplemented with cell walls of brewer's yeast exhibitedbeneficial prebiotic effects in rats (8) and in healthy female adults.Several studies have suggested that dietary administration of yeast cellwalls can reduce cholesterol (9–11), improve hyperlipidemia (12), and

ing author. Tel./fax: +81 82 424 4443.ress: nyutaka@hiroshima-u.ac.jp (Y. Nakashimada).

- see front matter © 2010, The Society for Biotechnology, Japan. All.jbiosc.2010.09.005

increase anti-oxidant activity (13). However, many studies of theimpact of BYC on the gut microbiota may have underestimated effectsbecause non-cultivation-based methods of analysis were not applied.

FIG. 1. SCFA profiles of cultured microbiota from swine feces (sample A) using an invitro intestinal model. Symbols: closed circle, lactate; closed triangle, acetate; opentriangle, butyrate; open square, propionate; closed square, formate. Results wereshown as the average of duplicate experiments.

rights reserved.

TABLE 1. Mean concentrations of metabolites in continuous culture without BYC fromday 3 to day 9.

Vessel Concentration (mM)

Lactate Formate Acetate Propionate Butyrate

Experiment with sample A1 8.4 (4.9) 9.7 (4.1) 22.7 (3.8) 0.2 (0.2) 3.8 (2.4)2 2.9 (2.2) 6.0 (5.3) 55.8 (8.1) 11.2 (2.3) 12.0 (6.2)3 0.7 (0.7) 2.7 (4.3) 69.5 (8.0) 13.4 (2.5) 13.4 (10.1)

Experiment with sample B1 7.0 (4.8) 9.5 (1.3) 15.1 (4.3) 1.1 (0.7) 3.2 (3.6)2 ND 7.9 (4.5) 59.3 (10.4) 19.0 (4.3) 7.1 (2.7)3 ND 2.5 (3.1) 69.5 (11.5) 22.6 (6.3) 7.5 (2.0)

Values in parentheses are standard deviations.

BREWER'S YEAST CELL WALL AFFECTS MICROBIOTA COMPOSITION 179VOL. 111, 2011

In vitro models of the digestive tract are useful for investigatingmicrobial processes such as carbohydrate and protein fermentation,steroid and bile acid metabolism, hydrogen formation and disposal,mutagen formation and degradation, transformation of xenobioticsubstances, and the metabolism of lignans and phytoestrogens (14).They involve the use of pure cultures, defined mixed cultures, or fecalmaterial, and can range from simple batch fermentations performedin serum bottles, to sophisticated, pH-controlled, multistage, contin-uous culture systems designed to simulate environmental conditionsthat occur in different parts of the colon. These multistage continuouscultures are likely the most realistic in this regard because theyfacilitate long-term studies and allow perturbations to the microbiotato be investigated under steady-state conditions. The breakdown ofcomplex carbohydrates is one of the most important functions carriedout by the colonic microbiota and an important factor in maintaininggut health (1,15). Recently, several studies have used in vitro reactor

TABLE 2. The OTUs observed in the in

T-RFs (bp) Putative microorganisms

Genus (phylum) Vessel 1

89 Bacteroides Prevotella +91 Bacteroides Prevotella −100 Fusobacterium +107 Lachnospiraceae Incertae Sedis −165 Coprococcus Lactobacillus ++173 Eggerthella −178 Ruminococcus −184 Roseburia −190 Fusobacterium −195 (Veillonellaceae) −206 Enterococcus +++213 Clostridium, Streptococcus, Lactobacillus −216 Clostridium +++218 Clostridium ++220 Lactobacillus Clostridium −223 Lactobacillus Clostridium +234 (Ruminococcaceae) Lactobacillus −255 (Ruminococcaceae) Lactobacillus +351 Eggerthella −360 Bifidobacteirum +525 Fusobacterium −535 Clostridium butyricum Lactobacillus (Ruminococcaceae) +537 Ruminococcaceae Faecalibacterium (Ruminococcaceae) +551 Clostridium −571 Streptococcus Lactococcus −574 Streptococcus Lactococcus +576 Lactococcus ++581 Megashaera Veilonella −584 Megashaera Lactobacillus −588 Lactobacillus +688 Clostridium +

Putative microorganisms were identified based on RDP 9 database using 35F primers and H+++: N10%, ++: 10% ~5%, +: 5%~1%, −: b1%.

systems to evaluate prebiotics, including oligosaccharides (16),dextrose (17), and soy germ (18).

In the presentwork, amultistage continuous in vitromodel was usedto study the effect of BYC on the diversity of colon microbes at themolecular level. The impact on microbiota was determined using non-cultivation-basedmethods and included analysis of short chain fatty acid(SCFA) production. The results suggest that BYC, like other prebiotics,has a significant impact on gut microbiota and its metabolites.

MATERIALS AND METHODS

Microbial source Swine feces were used instead of human feces as a microbialsource in this study. The swine digestive system resembles that of humans with respectto neural systems, the process of fecal pellet formation, and the weight ratio of organsto large intestine microbiota; the resulting microbiota of the two large intestinalsystems are similar (19). Furthermore, it is much easier to verify an in vitro result bycomparison to an in vivo system, since a fecal sample can be taken directly from the piglarge intestine by equipping an apparatus for fecal sampling to the pig body.

Fresh swine feces were collected twice from grain-fed adult Landrace pigs raised ata hog factory in Hiroshima Prefecture, Japan. Samples were collected as sample A and Band were immediately frozen at −80°C for subsequent use as the seed for continuouscultures with and without BYC.

Continuous culture with in vitro intestinal model The in vitro intestinalmodel consisted of three vessels with operating volumes of 0.22, 0.32, and 0.32 liters, inaccordance with a previous report (14). The vessels were installed in an incubatormaintained at 37°C. A pH controller in the three vessels maintained the pH at 5.5, 6.2, and6.8, respectively. Each fermentor was magnetically stirred and maintained under anatmosphere of CO2. The growth medium was continuously sparged with O2-free N2 andfed by peristaltic pump into vessel 1. The effluent of vessel 1 was sequentially supplied tovessel 2 and vessel 3 via a series of tubes.

The culture medium consisted of the following constituents (g per liter) in distilledwater according to the previous study (14): starch, 5.0; citrus pectin, 2.0; guar gum, 1.0;mucin (porcine gastric type III), 4.0; xylan (oatspelt), 2.0; arabinogalactan (larchwood),2.0; inulin, 1.0; casein, 3.0; peptone water, 5.0; tryptone, 5.0; bile salts No. 3, 0.4; yeastextract, 4.5; FeSO4⋅7H2O, 0.005; NaCl, 4.5; KCl, 4.5; KH2PO4, 0.5; MgSO4⋅7H2O, 1.25;CaCl2⋅6H2O, 0.15; NaHCO3, 1.5; cysteine, 0.8; hemin, 0.05; Tween 80, 1.0. The system

vitro continuous reactor system.

No addition of BYC Addition of BYC

Vessel 2 Vessel 3 Vessel 1 Vessel 2 Vessel 3

+ + − +++ ++++++ + − + +− + + + −− − + + ++ + − − −− − + − −++ +++ − − −− − + − −+ + − − +− − + − −+++ +++ +++ +++ +++− − + + −+++ +++ − + +++ + +++ + +++ +++ + + +− + − − −− − + + ++ + − − −− − + + +− − − − −− + − − −− − − − −+ + − − −+ − − − −+ − − − −− − − − +++ + − − −− + − − −− − + + ++ + + − +− + − − −

haI. The average of relative ratio of each OTU during the day 3 to day 13 was calculated.

180 NAKASHIMADA ET AL. J. BIOSCI. BIOENG.,

was operated at a retention time of 7.1, 10 and 10 h for vessel 1, 2 and 3, respectively(14). The overall retention time of 27.1 h was calculated as the reciprocal of dilutionrate.

The fecal slurry was prepared by suspending 10 g of swine feces into 300 ml of salinesolution and then removing large solidswith a stainless steel sieve under anoxic conditionin an anaerobic chamber (Tabai, Osaka, Japan). Each vessel was inoculatedwith 100 ml offecal slurry. The fermentation systemwasallowed to equilibrate overnight, afterwhich thepump ran for at least 14 d per experiment, during which 5-ml samples were taken fromeach vessel daily.

Chemical analyses Unless otherwise stated, all chemicals were obtained fromSigma. Brewer's yeast cell wall (BYC), which was prepared as described previously (7),was obtained as a prebiotics reagent from Kirin Holdings Co. Ltd, Japan. Samples fromthe fermentation system were centrifuged at 19,000 ×g (20 min) to remove bacteriaand other particulate substances. Acetate, propionate, and butyrate were measured byhigh-performance liquid chromatograph (Shimadzu, Kyoto, Japan) equipped with a300×7.8 mm Aminex HPX-87H column, (Bio-Rad, Tokyo, Japan), maintained at atemperature of 65°C. The flow rate of the 0.005 M H2SO4 mobile phase was 0.8 ml permin.

DNA extraction A 5-ml portion of mixed liquor was collected at each vessel andcentrifuged. The supernatant was removed and washed three times by suspension in1.0 ml of extraction buffer (100 mM Tris–HCl, 50 mM EDTA; pH 9.0) and centrifugation

Log 10

cel

ls/D

NA

5 10 150

Culture time (d5 100

8

6

4

2

8

6

4

2

8

6

4

2

8

6

4

2

8

6

4

2

Vessel 1 Vessel 2

FIG. 2. Analysis of predominant bacterial groups of microbiota in the anaerobic intestinal mexperiment without BYC addition and closed square represents the experiments with additivessel 1 on the last day was out of the graph range. The value was 1.1×1011 cells/DNA. Primaverage of duplicate experiments.

at 14,000 x g in order to remove possible PCR inhibitors. Genomic DNA was extractedfrom the remaining pellet with the FastDNA SPIN Kit for soil (Qbiogene, Carlsbad, CA).The concentration of the extracted DNA was measured by a NanoDrop spectropho-tometer (NanoDrop Technologies, Wilmington, DE) and was adjusted to 10 ng/μl bydilution with milliQ water.

Terminal-labeled restriction fragment length polymorphism (T-RFLP)analysis 35F (5′-CCTGGCTCAGGATGAACG) and 1492R (5′-GGTTACCTTGTTAC-GACTT) were used for T-RFLP analysis (20). The forward primer 35F was labeledwith Beckman Dye3 at the 5′-end. PCR was carried out in a total volume of 50 μL with50 ng Template DNA, 0.2 μM of each primer, 0.2 mM dNTP (each), 1× standard PCRbuffer (Perkin-Elmer Corp., Norwalk, CT), 0.1% BSA (Takara Biomedicals, Ohtsu,Japan), and 1.25 U AmpliTaq Gold DNA Polymerase (Perkin-Elmer Corp.). PCRcycling consisted of an initial denaturation at 95°C for 10 min; followed by 30 cyclesof denaturation at 95°C for 30 s, annealing at 60°C for 30 s, and extension at 72°C for1 min; and a final extension at 72°C for 10 min. The fluorescently labeled PCRproducts were purified using Microcon YM-100 (Millipore Corporation, Bedford,MA) and eluted in a final volume of 30 μl of distilled water. Then, 7 μl of the purifiedPCR products were digested with 20 U of HhaI (Takara) at 37°C for 10 h. The lengthsof the fluorescently labeled terminal restriction fragments were determined for eachsample using the CEQ™ 8000 Genetic Analysis System (Beckman Coulter, Fullerton,CA). T-RFs were determined in the range of 60 to 600 bp, which was the range of the

Lactobacillus

B. fragilisgroup

C. coccoidesgroup

Atopobium

C. leptumsub group

15 15

)5 100

Vessel 3

odel with fecal sample A with and without BYC. Symbols of open diamond representson of BYC. Solid arrows indicate the initial day of BYC addition. The plot of Atopobium aters used for the Qt-PCR of each group are listed in Table S1. Results were shown as the

BREWER'S YEAST CELL WALL AFFECTS MICROBIOTA COMPOSITION 181VOL. 111, 2011

size marker that could be determined reliably under the applied electrophoresisconditions. Putative microorganisms were identified based on RDP 9 database(http://rdp.cme.msu.edu/).

Quantitative real-time PCR Quantitative PCR analysis (Qt-PCR) was performedaccording to previously reported methods (21). Briefly, the purified DNA samples weresubjected to Qt-PCR to determine the number of bacteria using Light Cycler (RocheDiagnostics, Basel, Switzerland). Amplifications of 16S rRNA gene were performed withthe SYBR Premix Ex Taq™ (TakaRa, Tokyo) containing 0.2 μM of each primer and 0.1 μg/μlBSA. The primers used for this study are listed in Table S1. Lactobacillus plantarum JCM1149T, Bacteroides vulgatus JCM 5826T, Collinsella aerofaciens JCM 10188T, Ruminococcusproductus JCM 1471T, and Faecalibacterium prausnitzii ATCC 27766T, were used asstandard strains of Lactobacillus, B. fragilis group, Atopobium cluster, C. coccoides group,and C. leptum group, respectively. Chromosomal DNA extracted from the dilution series ofthe standard strainswasused to quantify thenumberof cells in fecal samples. Resultswereindicated as the cell number of respective bacterial groups per 50 ng of purified DNAextracted from the solid fraction of culture broth, which reflected the proportion of theparticular bacterial group to the total amount of bacteria.

120

90

60

30

0120

90

60

30

0

Vessel 1

Con

cent

ratio

n (m

M)

Vessel 2

Addition of BYC

120

90

60

30

0

Vessel 3

0 2 4 6 8 10 12 14

Culture time (d)

FIG. 3. SCFA profiles of culturedmicrobiota from swine feces (sample A) with BYC in thein vitro intestinal model. BYC was added to the feeding medium on day 6 of the culture.Symbols: closed circle, lactate; closed triangle, acetate; open triangle, butyrate; opensquare, propionate; closed square, formate. Results were shown as the average ofduplicate experiments.

RESULTS

Carbohydrate metabolism of microbiota in an in vitrocontinuous culture system We constructed a three-stage contin-uous culture system for colon microbes, simulating colon fermentationproceeding from the proximal to the distal part of the colon. Continuousculture was started after an overnight batch culture. In experimentswithout BYC, using fecal sample A as microbial source, metaboliteproductionwas unstable during the first 2 days and then became nearlystable from day 3 to 9 (Fig. 1). After day 10, metabolites, especiallyacetate, began to increase, suggesting that the composition of themicrobiota was becoming unstable. The mean concentration ofmetabolites in continuous culture without BYC from days 3 to 9 oftwo experiments using different fecal sample A and B is summarized inTable 1. At steady state, acetate was the most abundant SCFAmeasured in all vessels; this is consistent with the organic acidcomposition in both the human and animal intestine. However, theconcentration of organic acids was higher than that reported inhumans due to a lack of absorption by intestinal epithelium cells(22). Acetate in vessels 2 and 3 simulating the distal colonwas higherthan in the proximal part (vessel 1). Butyrate and propionate alsoincreased in the distal vessels compared to the proximal vessels,whereas formate and lactate decreased. These observations sug-gested that lactate produced in the proximal vesselswas converted tobutyrate or propionate in the distal vessels as reported previously(23,24). It was presumed that formate was consumed as an electrondonor by the strictly anaerobic species. The decrease in lactate invessels 2 and 3 also suggested that the population of microorganism(s) that produce butyrate and/or propionate from lactate (25,26)was larger in these vessels than in vessel 1. This pattern was verysimilar even when a different microbial source (fecal sample B),indicating that the reactor system reproducibly maintained cross-feeding of the gut microbiota.

Microbial community in the in vitro continuous culturesystem In order to assess the microbial community in this reactorsystem, T-RFLP was performed using samples taken from the experi-ment shown in Fig. 1. The 31 terminally-labeled restriction fragments(T-RFs) detected are listed in Table 2. The dominant gram-negativespecies were from the phylum Bacteroidetes (Operation TaxonomyUnit(OUT) [89 bp], OTU[91 bp], OTU[100 bp]), and genus Fusobacterium(OTU [190 bp]). The dominant gram-positive bacteria were Firmicutesincluding Lactobacillus, Clostridium, Ruminococcus, Streptococcus, andEnterococcus. The composition of the microbiota was consistent withprevious reports on swine feces and manure (27,28). Hierarchicalclustering analysis of the T-RFLP profiles performed for daily samples ineach vessel suggested that the composition ofmicrobiota fromdays 3 to10was similar or did not change significantly in any of the three vessels.

Qt-PCR analysis performed for Lactobacillus and the previouslyreported 4 dominant groups of gutmicrobiota listed in Table S1 (29–31)

indicated that all the major groups were present in this system (Fig. 2).The Lactobacillus group was stably maintained in all three vessels. Onthe other hand, the population of the Bacteroides fragilis group wasunstable in vessel 1 and lower than in vessel 2 and 3. Clostridiumcoccoides group (Clostidium cluster XIVa) and Clostridium leptumsubgroup (Clostidium cluster IV) showed a tendency similar to that ofBacteroides. These two Clostridium groups and Bacteroides are likely tobe low-pH sensitive as reported (32). Atopobium was detected with aQt-PCRmethod but not represented in Table 2, which is based on the T-RFLP method, due to the limit of detection of the method. Although theOTU with 35F primer that represented Atopobium cluster was 27 bp, itcould not be measured using the T-RFLP method used in this study.

Impact of BYC on carbohydrate metabolism and compositionof microbiota in the in vitro continuous culture system Becausemeasurable parameters of the continuous culture systembetween day3and 9 were reproducible, the effect of BYC addition on production ofmetabolites was evaluated in duplicate experiments. In experimentsusing fecal sample A, the continuous culture was started with feedingmedium without BYC, followed by addition of 5% (w/v) BYC to thefeeding medium after day 6. The SCFA profile shown in Fig. 3 is quitedifferent from thatwithout BYCaddition as shown in Fig. 1. Immediatelyafter BYC addition, acetate decreased and lactate accumulated in vessel1, after which metabolite concentrations stabilized. On the other hand,in vessel 2, propionate accumulated significantly until day 7 and thendecreased, accompanied by a transient decrease in acetate. Lactateaccumulated fromdays 10 to 12, and butyrate accumulated until day 13.The metabolite profiles after BYC addition in vessel 3 were similar tovessel 2.

The impact of BYC on total microbiota composition was studied bycomparing the T-RFLP profiles in experiments without and with BYC(Fig. 4). A drastic decrease was observed in the peak corresponding tothe Bacteroides (OTU[89, 91 bp]) population after BYC addition. Thesize of the B. fagilis group was also confirmed by means of Qt-PCR(Fig. 2). Again, the decrease in size of the B. fagilis group was observedin vessels 2 and 3 after the addition of BYC. On the other hand, theLactobacillus group population was stably maintained or increasedin all three vessels after BYC addition. The addition of BYC was foundto have an impact on metabolite profiles (Fig. 3). The profile of the

Vessel 1 Vessel 2 Vessel 3

Vessel 1 Vessel 2 Vessel 3

Rel

ativ

e flu

ores

cenc

e un

itsR

elat

ive

fluor

esce

nce

units

A

B

0

Day 2

Day 3

Day 5

Day 7

Day 9

Day 11

Day 13

Day 2

Day 3

Day 5

Day 7

Day 9

Day 11

Day 13

100

Terminal restriction fragment length (bases)

Terminal restriction fragment length (bases)

200 300 400 500 600 0 100 200 300 400 500 600 0 100 200 300 400 500 600

0 100 200 300 400 500 600 0 100 200 300 400 500 600 0 100 200 300 400 500 600

FIG. 4. Impact of BYC onmicrobiota composition of in vitro intestinal model. Comparison of the T-RFLP profiles in experiment without BYC (A) andwith BYC (B). Solid arrow indicatesOTU[89, 91 bp] (B. fragilis group) and dotted arrow indicates OTU[255 bp] (Ruminococcus or Lactobacillus strains).

182 NAKASHIMADA ET AL. J. BIOSCI. BIOENG.,

levels of OTU[581 bp] and OTU[584 bp] corresponded significantlywith the profile of propionate (Fig. 5). These two OTUs correspond tothe genus Veillonella and/or Megasphaera.

In Qt-PCR analyses of Lactobacillus and the four previouslyreported dominant groups of gut microbiota listed in Table S1, thesize of the Atopobium cluster slightly decreased in vessels 2 and 3 ofthe experiment with BYC addition (Fig. 2). The physiological role ofthis bacterial group in the gut has not to be determined, althoughAtopobium vaginae has been strongly associated with bacterialvaginosis (33–35). Since an increase in Atopobium cluster counts hasbeen reported with fracto-oligosaccharide supplementation inhumans (36), a component in BYC might affect the growth of thiscluster. The size of the C. coccoides group decreased in vessel 2,whereas the size of the C. leptum group did not change in vessel 2 and3 after BYC addition. The populations of these two groups weresmaller and rather unstable in vessel 1 compared to vessels 2 and 3.

Leser et al. reported that the C. leptum subgroup was the dominantcolonizer in the colon but was rarely observed in the ileum of swine(27). These groups are likely to be sensitive to low pH (32). It shouldbe noted that some species of these groups prefer to use organic acidssuch as propionate and lactate, rather than carbohydrates, as carbonsources. These preferences may affect the population stability of thesegroups. A unique T-RF (OTU[255 bp]) annotated to Lactobacillus sp., orRuminococcus sp. was found to have increased in vessels 1 and 2 in theexperiment with BYC addition (Fig. 4), suggesting that the addition ofBYC affected microbiota composition not only at the group-level butalso at the species level.

DISCUSSION

Although yeast cells have been widely used as probiotics,prebiotics, and synbiotics in humans and animals, the mechanisms

Addition of BYC

Culture time (d)

Per

cent

age

of

OT

U[5

81 a

nd 5

84 b

p](%

)

00

5

10

15

20

2 4 6 8 10 12 14

FIG. 5. The relative amount ofM. elsdenii-related bacteria in the continuous culture withBYC addition in fecal samples A. The relative ratio of OTU[581 bp]+OTU[584 bp] wascalculated based on the height of the T-RFLP peaks. Symbols: open square, vessel 1;closed diamond, vessel 2; closed triangle, vessel 3. Results were shown as the average ofduplicate experiments.

BREWER'S YEAST CELL WALL AFFECTS MICROBIOTA COMPOSITION 183VOL. 111, 2011

by which yeast cell walls mediate beneficial effects on gut microbiotaare less well studied than those of plant-derived prebiotics. To addressthe effect of BYC on microbiota, we used an in vitro reactor modelmimicking intestinal conditions. The study of prebiotics in reactorsystems is gaining greater acceptance, due in part to facile samplingunder stable and regulated conditions that are difficult, if notimpossible, to obtain in vivo.

Several interesting features of BYC on swine microbiota wereobserved in the present in vitro study. The most intriguing effect ofBYC was the significant decrease in the population size of the B. fragilisgroup. This finding is consistent with a previous study that demon-strated a reduction, albeit a non-significant one, in the population size ofthe B. fragilis group inmice that received oral administration of BYC (7).

Although it is one of the most dominant groups of gram-negativebacteria inside the gut, several studies have suggested that a largepopulation of the B. fragilis groupmay not be beneficial to the host (37–39). Several strains of this group have genes that have been implicatedin causing diarrhea (40). The present study provides encouraging dataindicating that BYC has the potential to improve the gut environment.

The drastic reduction in the B. fagilis population that follows BYCaddition in the continuous reactor suggested that BYC might havegrowth-inhibitory effects on this group. However, because BYC is nottoxic, it was unlikely that BYC directly inhibited growth. Although thereason for the decrease due to BYC addition is unknown, therelationship between growth rate and dilution rate in the reactormight affect the population size of the B. fragilis group in vessels 2 and3 of the continuous culture system. Previous reports suggested that B.thetaiotaomicron and B. fragilis possess a variety of hydrolases and canassimilate mannose (41,42). For example, the population of the B.fragilis group in vessel 1 was much lower than in vessels 2 and 3 of oursystem (Fig. 2) because the growth rate of the B. fragilis group at low pHwas much lower than that of other bacteria. Interestingly, it has beenreported that mannose inhibits the uptake of galactose, glucose andarabinose in both batch and continuous cultures of B. thetaiotaomicron(15). The most marked inhibition was observed in nitrogen-limitedcontinuous cultures. It is possible that mannose released from BYCinhibits utilization of glucose in the B. fragilis group, and that theresulting growth rate is insufficient to maintain the population incontinuous culture.

BYC addition also altered the SCFA profile. In the experiment withfecal sample A, propionate, lactate and butyrate increased signifi-cantly after BYC addition (Fig. 3). Propionate increased within oneday, while lactate and butyrate increased later. The profiles suggestedthat lactate was mainly converted to butyrate. T-RFLP profile of theexperiment suggested that levels of OTU581[bp], which most likelywere derived from M. elsdenii, or Veillonella montpellierensis, were

responsible for propionate production. In TGY broth consisting oftrypticase, glucose and yeast extract, V. montpellierensis producesacetate and propionate, but not acid, from glucose. This suggests thatrelatives of V. montpellierensis transiently produce propionate fromBYC (43). Furthermore, Megasphaera, a relative of Veillonella belong-ing to the Clostridium cluster IX (28) and labeled as OTU584[bp],assimilates lactate and produces butyrate in the absence of cataboliterepression by carbohydrates (44,45) as observed at later phases ofexperiment with BYC addition. It is likely that whenMegasphaera andrelatives exist in the gut, they are activated by the addition of BYC, ashas been reported for chicory root and sweet lupine (46).

The level of OTU[255 bp] was clearly affected by BYC addition,because this OTU was significantly higher after the addition of BYC inFig. 4. The database annotation to RDP 9 (http://rdp.cme.msu.edu/)suggests that this OTU is froma species of Lactobacillus orRuminococcus.Lactobacillus produces lactate and some Ruminococcus species arebutyrate producers (47,48). The Qt-PCR analysis suggested that totalpopulation size of the genus Lactobacillus, the C leptum group, and the C.coccoides group did not increase significantly after the addition of BYC.Thus, amore directmeans of identification is necessary todetermine thespecies origin of OTU[255 bp]. Previous reports have suggested thatfeeding a mixture of prebiotics to piglets during weaning increases theLactobacillus amylovorus-like populations in the colon, and that thisLactobacillus amylovorus-like strain has a positive effect on thegrowth ofthehost (49–51). It is possible that BYCmayalso stimulate thegrowthofsome beneficial Lactobacilli in the colon of the pig.

In conclusion, our results reveal an interesting effect of BYC onreducing the population of the B. fagilis group and activatingMegasphaera- and Lactobacillus-like species. These findings werenot clearly observed in previous reports using cultivation methods.To our knowledge, ours is the first study to analyze the prebioticeffect of BYC by a non-cultivation-based method. We anticipate thatfuture oral administration studies using non-cultivation analysis willbe performed to confirm the beneficial effects of BYC on human oranimal gut microbiota composition.

Supplementary materials related to this article can be found onlineat doi: 10.1016/j.jbiosc.2010.09.005.

References

1. Duncan, S. H., Scott, K. P., Ramsay, A. G., Harmsen, H. J. M., Welling, G. W.,Stewart, C. S., and Flint, H. J.: Effects of alternative dietary substrates oncompetition between human colonic bacteria in an anaerobic fermentor system,Appl. Environ. Microbiol., 69, 1136–1142 (2003).

2. Roberfroid, M. B.: Prebiotics: preferential substrates for specific germs? Am. J. Clin.Nutr., 73, 406S–409S (2001).

3. Macfarlane, G. T. and Cummings, J. H.: Probiotics and prebiotics: can regulating theactivities of intestinal bacteria benefit health? West. J. Med., 171, 187–191 (1999).

4. Guslandi, M., Giollo, P., and Testoni, P. A.: A pilot trial of Saccharomyces boulardiiin ulcerative colitis, Eur. J. Gastroenterol. Hepatol., 15, 697–698 (2003).

5. Vilela, E. G., Ferrari, M. D. D., Torres, H. O. D., Pinto, A. G., Aguirre, A. C. C.,Martins, F. P., Goulart, E. M. A., and Da Cunha, A. S.: Influence of Saccharomycesboulardii on the intestinal permeability of patients with Crohn's disease inremission, Scand. J. Gastroenterol., 43, 842–848 (2008).

6. Zanello, G., Meurens, F., Berri, M., and Salmon, H.: Saccharomyces boulardii effectson gastrointestinal diseases, Curr. Issues Mol. Biol., 11, 47–58 (2008).

7. Nakamura, T., Agata, K., Mizutani, M., and Iino, H.: Effects of Brewer's yeast cellwall on constipation and defecation in experimentally constipated rats, Biosci.Biotechnol. Biochem., 65, 774–780 (2001).

8. Nakamura, T., Nishida, S., Mizutani, M., and Iino, H.: Effects of yogurtsupplemented with brewer's yeast cell wall on constipation and intestinalmicroflora in rats, J. Nutr. Sci. Vitaminol., 47, 367–372 (2001).

9. Hitomi, Y., Yoshida, M., Mizutani, M., Nakamura, T., and Shimasaki, H.: Effects ofbrewer's yeast cellwall on fecal steroid excretion in rats, J. Oleo Sci.,51, 335–346 (2002).

10. Nakamura, T., Hitomi, Y., Yoshida, M., Shirasu, Y., Tsukui, T., and Shimasaki, H.:Effect of yogurt supplemented with Brewer's Yeast Cell wall on levels of bloodlipids in normal and hypercholesterolemic adults, J. Oleo Sci., 51, 323–334 (2002).

11. Htwe, K., Yee, K. S., Tin, M., and Vandenplas, Y.: Effect of Saccharomyces boulardiiin the treatment of acute watery diarrhea in Myanmar children: a randomizedcontrolled study, Am. J. Trop. Med. Hyg., 78, 214–216 (2008).

184 NAKASHIMADA ET AL. J. BIOSCI. BIOENG.,

12. Hitomi, Y., Yoshida, M., Nakamura, T., and Shirasu, Y.: Effects of brewer's yeastcell wall on serum lipid levels in rats fed a high cholesterol and fat diet, J. Oleo Sci.,51, 141–144 (2002).

13. Jaehrig, S. C., Rohn, S., Kroh, L. W., Fleischer, L. G., and Kurz, T.: In vitro potentialantioxidant activity of (1–N3), (1–N6)-β -D-glucan and protein fractions fromSaccharomyces cerevisiae Cell Walls, J. Agric. Food Chem., 55, 4710–4716 (2007).

14. Macfarlane, G. T., Macfarlane, S., and Gibson, G. R.: Validation of a three-stagecompound continuous culture system for investigating the effect of retention timeon the ecology and metabolism of bacteria in the human colon, Microb. Ecol., 35,180–187 (1998).

15. Degnan, B. A. and Macfarlane, G. T.: Carbohydrate utilization patterns andsubstrate preferences in Bacteroides thetaiotaomicron, Anaerobe, 1, 25–33 (1995).

16. Tzortzis, G., Goulas, A. K., Gee, J. M., and Gibson, G. R.: A novel galactooligo-saccharide mixture increases the bifidobacterial population numbers in acontinuous in vitro fermentation system and in the proximal colonic contents ofpigs in vivo, J. Nutr., 135, 1726–1731 (2005).

17. Mäkeläinen, H. S., Mäkivuokko, H. A., Salminen, S. J., Rautonen, N. E., andOuwehand, A. C.: The effects of polydextrose and xylitol on microbial communityand activity in a 4-stage colon simulator, J. Food Sci., 72, M153–M159 (2007).

18. De Boever, P., Deplancke, B., and Verstraete, W.: Fermentation by gut microbiotacultured in a simulator of the human intestinal microbial ecosystem is improved bysupplementing a soygerm powder, J. Nutr., 130, 2599–2606 (2000).

19. Inoue, R., Tsukahara, T., Nakanishi, N., and Ushida, K.: Development of theintestinal microbiota in the piglet, J. Gen. Appl. Microbiol., 51, 257–265 (2005).

20. Hayashi, H., Sakamoto, M., and Benno, Y.: Evaluation of three different forwardprimers by terminal restriction fragment length polymorphism analysis fordetermination of fecal Bifidobacterium spp. in healthy subjects, Microbiol. Immunol.,48, 1–6 (2004).

21. Nishida, S., Michinaka, A., Nakashima, K., Iino, H., and Fujii, T.: Evaluation of theprobiotic potential of Lactobacillus paracasei KW3110 based on in vitro tests and oraladministration tests in healthy adults, J. Gen. Appl. Microbiol., 54, 267–276 (2008).

22. Cummings, J. H., Pomare, E. W., Branch, W. J., Naylor, C. P. E., and Macfarlane, G.T.: Short chain fatty acids in human large intestine, portal, hepatic and venousblood, Gut, 28, 1221–1227 (1987).

23. Bourriaud, C., Robins, R. J., Martin, L., Kozlowski, F., Tenailleau, E., Cherbut, C., andMichel, C.: Lactate is mainly fermented to butyrate by human intestinal microflorasbut inter-individual variation is evident, J. Appl. Microbiol., 99, 201–212 (2005).

24. Morrison, D. J., Mackay, W. G., Edwards, C. A., Preston, T., Dodson, B., andWeaver, L. T.: Butyrate production from oligofructose fermentation by the humanfaecal flora: what is the contribution of extracellular acetate and lactate? Br. J.Nutr., 96, 570–577 (2006).

25. Duncan, S. H., Louis, P., and Flint, H. J.: Lactate-utilizing bacteria, isolated fromhuman feces, that produce butyrate as a major fermentation product, Appl.Environ. Microbiol., 70, 5810–5817 (2004).

26. Belenguer, A., Duncan, S. H., Holtrop, G., Anderson, S. E., Lobley, G. E., and Flint,H. J.: Impact of pH on lactate formation and utilization by human fecal microbialcommunities, Appl. Environ. Microbiol., 73, 6526–6533 (2007).

27. Leser, T. D., Amenuvor, J. Z., Jensen, T. K., Lindecrona, R. H., Boye, M., andMoller,K.: Culture-independent analysis of gut bacteria: the pig gastrointestinal tractmicrobiota revisited, Appl. Environ. Microbiol., 68, 673–690 (2002).

28. Cotta, M. A., Whitehead, T. R., and Zeltwanger, R. L.: Isolation, characterizationand comparison of bacteria from swine faeces and manure storage pits, Environ.Microbiol., 5, 737–745 (2003).

29. Jackson, M. S., Bird, A. R., and McOrist, A. L.: Comparison of two selective mediafor the detection and enumeration of Lactobacilli in human faeces, J. Microbiol.Meth., 51, 313–321 (2002).

30. Matsuki, T., Watanabe, K., Fujimoto, J., Miyamoto, Y., Takada, T., Matsumoto, K.,Oyaizu, H., and Tanaka, R.: Development of 16S rRNA-gene-targeted group-specific primers for the detection and identification of predominant bacteria inhuman feces, Appl. Environ. Microbiol., 68, 5445–5451 (2002).

31. Matsuki, T., Watanabe, K., Fujimoto, J., Takada, T., and Tanaka, R.: Use of 16S rRNAgene-targeted group-specific primers for real-time PCR analysis of predominantbacteria in human feces, Appl. Environ. Microbiol., 70, 7220–7228 (2004).

32. Walker, A. W., Duncan, S. H., Leitch, E. C. M., Child, M. W., and Flint, H. J.: pH andpeptide supply can radically alter bacterial populations and short-chain fatty acidratioswithinmicrobial communities from thehuman colon, Appl. Environ.Microbiol.,71, 3692–3700 (2005).

33. Ferris, M. J., Masztal, A., Aldridge, K. E., Fortenberry, D., Fidel, P. L., and Martin,D. H.: Association of Atopobium vaginae, a recently described metronidazoleresistant anaerobe, with bacterial vaginosis, BMC Infect. Dis., 4, 5 (2004).

34. Verhelst, R., Verstraelen, H., Claeys, G., Verschraegen, G., Delanghe, J., VanSimaey, L., De Ganck, C., Temmerman, M., and Vaneechoutte, M.: Cloning of 16SrRNA genes amplified from normal and disturbed vaginal microflora suggests astrong association between Atopobiurn vaginae, Gardnerella vaginalis and bacterialvaginosis, BMC Microbiol., 4, 16 (2004).

35. Burton, J. P., Devillard, E., Cadieux, P. A., Hammond, J. A., and Reid, G.: Detectionof Atopobium vaginae in postmenopausal worrien by cultivation-independentmethods warrants further investigation, J. Clin. Microbiol., 42, 1829–1831 (2004).

36. Saulnier, D. M. A., Gibson, G. R., and Kolida, S.: In vitro effects of selectedsynbiotics on the human faecal microbiota composition, FEMS Microbiol. Ecol., 66,516–527 (2008).

37. Fukuda, S., Ishikawa, H., Koga, Y., Aiba, Y., Nakashima, K., Cheng, L., andShirakawa, T.: Allergic symptoms and microflora in schoolchildren, J. Adolesc.Health, 35, 156–158 (2004).

38. Odamaki, T., Xiao, J.-Z., Iwabuchi, N., Sakamoto, M., Takahashi, N., Kondo, S.,Miyaji, K., Iwatsuki, K., Togashi, H., Enomoto, T., and Benno, Y.: Influence ofBifidobacterium longum BB536 intake on faecal microbiota in individuals withJapanese cedar pollinosis during the pollen season, J. Med. Microbiol., 56,1301–1308 (2007).

39. Vael, C., Nelen, V., Verhulst, S., Goossens, H., and Desager, K.: Early intestinalBacteroides fragilis colonisation and development of asthma, BMC Pulm. Med., 8,19 (2008).

40. Pathela, P., Hasan, K. Z., Roy, E., Alam, K., Huq, F., Siddique, A. K., and Sack, R. B.:Enterotoxigenic Bacteroides fragilis associated diarrhea in children 0–2 years of agein rural Bangladesh, J. Infect. Dis., 191, 1245–1252 (2005).

41. Xu, J., Bjursell, M. K., Himrod, J., Deng, S., Carmichael, L. K., Chiang, H. C., Hooper,L. V., and Gordon, J. I.: A genomic view of the human-Bacteroides thetaiotaomicronsymbiosis, Science, 299, 2074–2076 (2003).

42. Kuwahara, T., Yamashita, A., Hirakawa, H., Nakayama, H., Toh, H., Okada, N.,Kuhara, S., Hattori, M., Hayashi, T., and Ohnishi, Y.: Genomic analysis of Bacter-oides fragilis reveals extensive DNA inversions regulating cell surface adaptation,Proc. Nat. Acad. Sci. U.S.A., 101, 14919–14924 (2004).

43. Jumas-Bilak, E., Carlier, J. P., Jean-Pierre, H., Teyssier, C., Gay, B., Campos, J., andMarchandin, H.: Veillonella montpellierensis sp nov., a novel, anaerobic, gram-negative coccus isolated from human clinical samples, Int. J. Sys. Evol. Microbiol.,54, 1311–1316 (2004).

44. Waldrip, H.M. andMartin, S. A.: Effects of an Aspergillus oryzae fermentation extractandother factors on lactate utilization by the ruminal bacteriumMegasphaera elsdenii,J. Anim. Sci., 71, 2770–2776 (1993).

45. Russell, J. B. and Baldwin, R. L.: Substrate preferences in rumen bacteria: evidence ofcatabolite regulatory mechanisms, Appl. Environ. Microbiol., 36, 319–329 (1978).

46. Mobak, L., Thomsen, L. E., Jensen, T. K., Knudsen, K. E. B., and Boye, M.: Increasedamount of Bifidobacterium thermacidophilum and Megasphaera elsdenii in thecolonic microbiota of pigs fed a swine dysentery preventive diet containing chicoryroots and sweet lupine, J. Appl. Microbiol., 103, 1853–1867 (2007).

47. Lawson, P. A., Song, Y. L., Liu, C. X., Molitoris, D. R., Vaisanen, M. L., Collins, M. D.,and Finegold, S. M.: Anaerotruncus colihominis gen. nov., sp nov., from humanfaeces, Int. J. Syst. Evol. Microbiol., 54, 413–417 (2004).

48. Duncan, S. H., Hold, G. L., Harmsen, H. J. M., Stewart, C. S., and Flint, H. J.: Growthrequirements and fermentation products of Fusobacterium prausnitzii, and aproposal to reclassify it as Faecalibacterium prausnitzii gen. nov., comb. nov, Int. J.Syst. Evol. Microbiol., 52, 2141–2146 (2002).

49. Konstantinov, S. R., Awati, A., Smidt, H., Williams, B. A., Akkermans, A. D. L., andde Vos, W. A.: Specific response of a novel and abundant Lactobacillus amylovorus-like phylotype to dietary prebiotics in the guts of weaning piglets, Appl. Environ.Microbiol., 70, 3821–3830 (2004).

50. Roselli, M., Finamore, A., Britti, M. S., Konstantinov, S. R., Smidt, H., de Vos, W. M.,and Mengheri, E.: The novel porcine Lactobacillus sobrius strain protects intestinalcells from enterotoxigenic Escherichia coli K88 infection and prevents membranebarrier damage, J. Nutr., 137, 2709–2716 (2007).

51. Konstantinov, S. R., Smidt, H., Akkermans, A. D. L., Casini, L., Trevisi, P., Mazzoni,M., De Filippi, S., Bosi, P., and de Vos, W. M.: Feeding of Lactobacillus sobriusreduces Escherichia coli F4 levels in the gut and promotes growth of infectedpiglets, FEMS Microbiol. Ecol., 66, 599–607 (2008).