Q2

Q1

lable at ScienceDirect

Anaerobe xxx (2013) 1e4

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455

YANAE1196_proof ■ 1 August 2013 ■ 1/4

Contents lists avai

Anaerobe

journal homepage: www.elsevier .com/locate/anaerobe

Note

565758596061626364

A representative of the dominant human colonic Firmicutes, Roseburiafaecis M72/1, forms a novel bacteriocin-like substance

Diane Hatziioanou a, Melinda J. Mayer a,*, Sylvia H. Duncan b, Harry J. Flint b, Arjan Narbad a

aGut Health and Food Safety Programme, Institute of Food Research, Norwich Research Park, Colney, Norwich, NR4 7UA, UKbMicrobiology Group, Rowett Institute of Nutrition and Health, University of Aberdeen, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, UK

6566

67686970717273747576777879

a r t i c l e i n f o

Article history:Received 20 May 2013Received in revised form18 July 2013Accepted 22 July 2013Available online xxx

Keywords:Roseburia faecisBacteriocin-like inhibitory substance (BLIS)Firmicutes anaerobeGut microbiotaTrypsin

* Corresponding author. Tel.: þ44 1603 255284; faxE-mail address: Melinda.Mayer@ifr.ac.uk (M.J. May

1075-9964/$ e see front matter � 2013 Published byhttp://dx.doi.org/10.1016/j.anaerobe.2013.07.006

Please cite this article in press as: Hatziioanforms a novel bacteriocin-like substance, An

a b s t r a c t

During screening of human gut isolates in search of novel antimicrobials, the butyrate-producing strainRoseburia faecis M72/1 was found to produce an inhibitory substance active against Bacillus subtilis.Partial purification of the antimicrobial was achieved and activity found to be heat labile. Our findingssuggest that R. faecis M72/1 produces a proteinaceous inhibitor whose production may be triggered bytrypsin in the gastrointestinal tract.

� 2013 Published by Elsevier Ltd.

80

81 828384858687888990919293949596979899

100101102103104

Many species of bacteria have evolved and adapted to live in thehuman gut. Especially in the large intestine there is a complex anddynamic microbial ecosystem, with bacteria reaching concentra-tions of 1011e1012 cells/g [1]. The microbiota composition isthought to influence nutrition, metabolism, epithelial developmentand immune modulation [2,3]. In addition, some gut bacteriaproduce proteinaceous antimicrobials e bacteriocin-like sub-stances (BLIS) e that selectively inhibit other gut commensals andmay give a competitive advantage to the producers in the complexmicrobial environment [4]. Although numerous bacteriocins havebeen discovered from mammalian gut isolates [5e10] it is notknownwhat proportion of gut bacteria produce bacteriocins. Manystudies on lactic acid bacteria have shown the importance ofbacteriocin production in the dairy industry [11] and it has beensuggested that the role of bacteriocins in the gut may be similar tothose in other complex ecosystems [12]. Potential application ofthese BLIS ranges from use as antimicrobials in medicine and foodpreservation to cancer treatment [13e15]. Bacterial strains whichproduce BLIS could also find applications as probiotics [2,16,17].

Roseburia species are common gut bacteria belonging to theLachnospiraceae family, that comprises around one-third of the gut

: þ44 1603 507723.er).

Elsevier Ltd.

ou D, et al., A representativeaerobe (2013), http://dx.doi.

105106107108

microbiota, and Roseburia faecis is one of the 12 most dominantspecies present in the large intestine [18]. They contribute tocomplex polysaccharide breakdown in the colon and are one of thepredominant producers of short chain fatty acids (SCFA) and inparticular butyrate, which can influence a number of cellularfunctions associated with colonic health [18,19].

In this study the human gut isolate R. faecisM72/1 (DSM16840T)[20] was investigated for production of novel antimicrobials. Theisolate was grown on solid media and screened for antimicrobialactivity using overlay assays. Briefly, petri dishes with bacterialcolonies were exposed to chloroform in absorbent paper for 10 minthen aerated for 5 min before being overlaid with soft agar (0.7%)seeded with 2% (v/v) overnight grown culture of each indicatorbacteria and incubated at 37 �C. Indicator strains were from an in-house collection (Institute of Food Research, IFR) or from culturecollections NCIMB or ATCC andwere cultured in LB at 37 �C (Bacillussubtilis EC1280 and 94/50 (IFR), NCIMB3610), in MRS (Oxoid) at42 �C (Lactobacillus helveticus CH1 (IFR)) or 37 �C (Micrococcusluteus MC8166 (IFR)), in BHI (Oxoid) at 37 �C (Listeria innocuaATCC33090) or in M17 (Oxoid) with 0.5% glucose at 30 �C (Lacto-coccus lactis MG1614 (IFR)). Activity in supernatants was alsoassessed by drop tests, with 10 ml aliquots being spotted onto lawnsof indicator strains prior to growth.

In preliminary antimicrobial assays using BHI media withcomplements (50 mg/l vitamin K, 5 mg/l hemin, 1 mg/l resazurin

of the dominant human colonic Firmicutes, Roseburia faecis M72/1,org/10.1016/j.anaerobe.2013.07.006

109110

2 3 4 5 d

+trp

-trp

a

c

90°C

60°C

10 30 60 10 30 60 min

GM17R. faecis

b >30 kDa10-30

Fig. 2. Drop test of supernatant from R. faecis M72/1 spotted onto B. subtilis 94/50; a,R. faecis grown in GM17 with (þtrp) or without (�trp) trypsin (5 mg/ml) for 2e5 d; b,10e30 and >30 kDa filtered concentrated supernatant from R. faecis grown in GM17with 5 mg/ml trypsin; c, effect of temperature on concentrated supernatant (>10 kDa)of R. faecis (left) heated to 60 �C or 90 �C for 10, 30 or 60 min compared to a mediacontrol (right) which had received identical processing and treatments.

D. Hatziioanou et al. / Anaerobe xxx (2013) 1e42

111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175

176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212

YANAE1196_proof ■ 1 August 2013 ■ 2/4

and 0.5 g/l L-cysteine (BHIc)) there was evidence of antimicrobialactivity produced by R. faecis M72/1 against B. subtilis EC1280,B. subtilis NCIMB3610 and B. subtilis 94/50 while L. helveticus andL. innocua showed no inhibition. Initial antimicrobial productionwas lost after sub-culturing so a range of growth conditions andmedia compositions were tested to identify the necessary condi-tions for antimicrobial production. As noted with Ruminococcusgnavus [21], addition of 5e50 mg/ml trypsin to the culture mediainduced antimicrobial production from R. faecis M72/1 on solidmedia (Fig. 1), with 5 and 50 mg/ml trypsin giving similar levels ofinduction. Duplicate assays with bromocresol purple did notdemonstrate any alteration in pH, indicating that the inhibitionwasunlikely to be due to production of organic acids. Further testingdemonstrated that while trypsin was not always necessary forantimicrobial activity, its presence increased the zone of inhibition.

Longer incubation in liquid culture prior to growth on solidmedia gave increased activity in subsequent overlay assays (Fig. 1).To test for the presence of an inducing molecule produced by thebacterium, spent culture supernatant of 24 h cultures grownwithout trypsin was prepared by centrifugation (30 min, 10,000g,4 �C), concentrated 10-fold by freeze-drying and used to supple-ment fresh cultures (1/10, v/v). At 24 h intervals cultures wereplated in duplicate on solid BHI and overlay assays demonstratedthat spent culture supernatant was able to increase antimicrobialproduction without the presence of trypsin (Fig. 1f).

Production of antimicrobial activity by R. faecisM72/1 could notbe demonstrated when cultured in BHIc liquid media despiteaddition of trypsin or concentrated culture supernatant. Similardifficulties with extraction of bacteriocins from Gram-positivebacteria producing bacteriocin-like substances only on solid me-dia have been documented [22,23]. However, during culturing in arange of different media it was found that antimicrobial activitycould be detected in liquid media when M17 supplemented with0.5% glucose and 5 mg/ml trypsin was used, and this activityremained stable for 2e5 d (Fig. 2a). A range of other carbon sourceswere tested and the highest antimicrobial productionwas observedwith growth on glucose, xylose or cellobiose (data not shown).

Fig. 1. Overlay assay for antimicrobial activity of R. faecis M72/1 cultures using B. subtilis 94/50. R. faecis M72/1 was cultured in liquid BHIc for 24 h (aec) or 48 h (def) without(�sup, a, b, d, e) or with (þsup, c, f) concentrated supernatant of previously grown R. faecis M72/1 culture; cultures were then grown on plates containing no trypsin (�trp, a, c, d, f)or 5 mg/ml trypsin (þtrp, b, e).

Please cite this article in press as: Hatziioanou D, et al., A representative of the dominant human colonic Firmicutes, Roseburia faecis M72/1,forms a novel bacteriocin-like substance, Anaerobe (2013), http://dx.doi.org/10.1016/j.anaerobe.2013.07.006

213214215216217218219220221222223224225226227228229230231232233234235236237238239240

Table 1Effect of protein purification procedures on R. faecis M72/1 culture supernatant.

Step Conditions/samples tested Observation

(NH4)2SO4 precipitation 0e20%, 20e40%, 40e60%, 60e80% or 0e80% (NH4)2SO4

precipitates (antimicrobial activity tested after desalting).Majority of the antimicrobial activity observedin 60e80% saturated samples.

Desalting before andafter HIC

50 MWCO before and 10 kDa Amicon filter devices after,or dialysis with 14 kDa tubing at both stages.

Dialysis with 14 kDa dialysis tubing is able todesalt preparations, but filtration is preferredto avoid losses and antimicrobial degradation due to instability.

HIC purification 1 ml HiTrap columns: Butyl-s FF, Phenyl FF high sub,Phenyl FF low sub, Octyl FF, Butyl FF, Phenyl HP, Butyl HP.

Most concentrated antimicrobial elution observedfrom Phenyl HP resins (at 40e60% buffer B).

Concentration ofFPLC fractions

Phenyl HP eluate fractions. Using dialysed samples it was necessary to pool triplicateHIC preparations and concentrate by freeze drying.Using filtered samples the fractions were concentratedduring desalting by adjusting the retained volume to 250 ml.

M 1

188

62

49

98

38

28

17

M 1 2 3 4 5 6 7 8

71

55

41 29

kDa

b

a

d

c e

0

200

400

600

800

1000

1200

1400

1600

0 1 2 3 4 5 6 7 8 9 10 11121314 1516 171819 20

mA

u

1 2 3 4

5 6 7 8

U

B

ml

Fig. 3. Analysis of R. faecis M72/1 culture supernatant FPLC preparations; a, chro-matogram of FPLC purification, U, UV absorbance, B, & buffer gradient/ml; b, SDS-PAGE analysis of FPLC fractions from (a), 13 ml sample per lane. M, SeeBluePlus2marker, lanes 1, 0e7 ml at 0% buffer (fraction 1), 2, fraction 2 (0e20% buffer), 3, fraction3 (20e40%), 4, fraction 4 (40e60%), 5, fraction 5 (60e80%), 6, fraction 6 (80e100%), 7,fraction 7 (1 ml eluate after end of gradient), 8, Retained proteins eluted after end ofgradient; c, antimicrobial activity of FPLC fractions 1e8 from (b) on B. subtilis 94/50; d,SDS-PAGE analysis of pooled concentrated 40e60% buffer B fraction (13 ml, lane 1), M,SeeBluePlus2 marker. Arrow indicates band excised for analysis; e, antimicrobial ac-tivity of pooled concentrated fraction in (d) against B. subtilis 94/50.

D. Hatziioanou et al. / Anaerobe xxx (2013) 1e4 3

241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305

306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370

YANAE1196_proof ■ 1 August 2013 ■ 3/4

Other potential inducers tested (eg 5% or 0.5% faecal water, 0.31%volatile fatty acid mix) failed to trigger antimicrobial production.Interestingly, incubation of filter-sterilised culture supernatantwith trypsin did not produce an increase in antimicrobial activity,indicating that trypsin acts to induce production from the cellsrather than directly affecting a factor in the culture supernatant.

To attempt to concentrate the active component, sequentialultrafiltration was performed on filter sterilised culture superna-tants of R. faecisM72/1 initially through 1 kDa Amicon membranes(Fisher) followed by filtration of the retained proteins through10 kDa then 30 kDa MWCO regenerated cellulose filter devices(Amicon), with each fraction being freeze dried to 100� concen-tration. The antimicrobial content of each concentrated filtrate wastested, showing no activity in fractions below 30 kDa (Fig. 2b).Concentrated supernatant was also active against the Lactococcuslactis MG1614 and M. luteus Mc8166 (data not shown).

The heat stability of the concentrated antimicrobial substancewas examined using this concentrated fraction. Aliquots of 30 ml ora corresponding media control were heated to 60 �C for 10 min,30 min or 1 h, at 90 �C for 10 min or 30 min, or 121 �C for 15 min.Each sample was cooled on ice and tested for antimicrobial activityusing a drop test. Activity was found to be stable at 60 �C for up to1 h but was abolished after 10 min incubation at 90 �C (Fig. 2c).

Partial purification of the BLIS was performed on filter sterilisedculture supernatants from R. faecis grown in M17 supplementedwith 0.5% glucose and 5 mg/ml trypsin (Table 1). Initial purificationinvolved precipitation with 80% (NH4)2SO4 followed by desaltingusing a 50 MWCO Amicon filter device with three additions of 10volumes of 50 mM potassium phosphate pH 6.0 buffer. (NH4)2SO4was added to 2 M and hydrophobic interaction chromatography(HIC) was performed on an AKTA FPLC UPC-900 using a 1ml PhenylHP resin column (GE Healthcare). Elution was performed on a 10column volume gradient of 2 M (NH4)2SO4, 50 mM potassiumphosphate (pH 6.0) to 50 mM potassium phosphate (pH 6.0)(Fig. 3a). Each collected fraction was desalted using 10 kDa Amiconfilter devices by washing twice with an excess of phosphate buffer(pH 6.0) and the volume of each fraction adjusted to 250 ml.Desalted fractions were electrophoresed on a 3e8% TriseAcetategel (Invitrogen, Fig. 3b) and assayed for antimicrobial activity usingdrop tests (Fig. 3c). The majority of the activity was located infraction 4 (40e60% buffer B), which showed protein bands at c. 45,49 and 70 kDa.

Stepwise (NH4)2SO4 precipitation showed that the antimicrobialactivity was found at 60e80% saturation (Table 1). To improvepurification, 1 l supernatant was precipitated with 60e80%(NH4)2SO4 and dialysed in 2 M (NH4)2SO4, 50 mM potassiumphosphate buffer (pH 6.0). FPLC was performed as before and the40e60% buffer B fraction from triplicate FPLC runs was pooled,desalted by dialysis, concentrated 12-fold by freeze drying andanalysed by SDS-PAGE on 4e12% BiseTris gels (Invitrogen) in MES

Please cite this article in press as: Hatziioanou D, et al., A representative of the dominant human colonic Firmicutes, Roseburia faecis M72/1,forms a novel bacteriocin-like substance, Anaerobe (2013), http://dx.doi.org/10.1016/j.anaerobe.2013.07.006

D. Hatziioanou et al. / Anaerobe xxx (2013) 1e44

371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427

428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481

YANAE1196_proof ■ 1 August 2013 ■ 4/4

buffer and reducing conditions (Fig. 3d). This preparation hadantimicrobial activity against B. subtilis 94/50 (Fig. 3e). MALDI-ToFanalysis of tryptic digests from the major c. 45 kDa band providedpeptide mass fingerprint datawhichmay allow identification of theprotein when the R. faecis genome sequence becomes available.

To our knowledge this is the first evidence of BLIS production bya species of Roseburia, and one of a small number of high molecularmass BLIS produced by Gram positive bacteria, with the only othergut-associated BLIS being enterolysin A, produced by the ruminalisolate Enterococcus faecalis LMG 2333 [24]. Other large BLIS includehelveticin J, lactacin F andmillericin B, produced by L. helveticus 481[25] Lactobacillus acidophilus 11088 [26] and Streptococcus milleriNMSCC 061 [27] respectively. It has been shown that bacteriarelated to Roseburia and Eubacterium rectale comprise up to 10% oftotal bacterial diversity in the human colon and could thereforeplay an important role in the maintenance of gut health [28].Antimicrobial production may be one of the factors that allowR. faecis to compete successfully within the dense microbial pop-ulation in the colon. Interestingly, a recent study revealed a highdegree of inter-individual variation in the dominant Roseburiaspecies present among healthy individuals [29] and bacteriocinproduction would provide one possible explanation for this. Wefound that production of the antimicrobial could be improved bythe presence of trypsin, commonly found in the gut environment,and that the bacteria may produce a soluble inducer. The prefer-ence for production on solid media indicates that cellecell or cell-surface contact may be important and that the antimicrobial pro-duction was cell density dependent, possibly indicating a role forquorum sensing in R. faecis M72/1 antimicrobial production.Although the final antimicrobial yields were low in vitro it is hopedthat further characterization of this antimicrobial will give insightsinto the competence of R. faecis M72/1 in the gut.

Acknowledgement

This work was supported by a Biotechnology and BiologicalSciences Research Council studentship grant to DH. SHD and HJFacknowledge support from the Scottish Government RESASprogramme.

References

[1] Guarner F, Malagelada JR. Gut flora in health and disease. Lancet 2003;361:512e9.

[2] Howarth GS, Wang H. Role of endogenous microbiota, probiotics and theirbiological products in human health. Nutrients 2013;5:58e81.

[3] O’Shea EF, Cotter PD, Stanton C, Ross RP, Hill C. Production of bioactivesubstances by intestinal bacteria as a basis for explaining probiotic mecha-nisms: bacteriocins and conjugated linoleic acid. Int J Food Microbiol2012;152:189e205.

[4] Gillor O, Etzion A, Riley MA. The dual role of bacteriocins as anti- and pro-biotics. Appl Microbiol Biotechnol 2008;81:591e606.

[5] Booth SJ, Johnson JL, Wilkins TD. Bacteriocin production by strains of Bacter-oides isolated from human feces and the role of these strains in the bacterialecology of the colon. Antimicrob Agents Chemother 1977;11:718e24.

[6] Dabard J, Bridonneau C, Phillipe C, Anglade P, Molle D, Nardi M, et al. Rumi-nococcin A, a new lantibiotic produced by a Ruminococcus gnavus strain iso-lated from human feces. Appl Environ Microbiol 2001;67:4111e8.

Please cite this article in press as: Hatziioanou D, et al., A representativeforms a novel bacteriocin-like substance, Anaerobe (2013), http://dx.doi.

[7] Kalmokoff ML, Lu D, Whitford MF, Teather RM. Evidence for production of anew lantibiotic (butyrivibriocin OR79A) by the ruminal anaerobe Butyrivibriofibrisolvens OR79: characterization of the structural gene encoding butyrivi-briocin OR79A. Appl Environ Microbiol 1999;65:2128e35.

[8] Kawai Y, Saito T, Toba T, Samant SK, Itoh T. Isolation and characterization of ahighly hydrophobic new bacteriocin (gassericin A) from Lactobacillus gasseriLA39. Biosci Biotechnol Biochem 1994;58:1218e21.

[9] Laukova A, Marekova M. Antimicrobial spectrum of bacteriocin-like sub-stances produced by rumen staphylococci. Folia Microbiol (Praha) 1993;38:74e6.

[10] O’Shea EF, Gardiner GE, O’Connor PM, Mills S, Ross RP, Hill C. Characterizationof enterocin- and salivaricin-producing lactic acid bacteria from themammalian gastrointestinal tract. FEMS Microbiol Letts 2009;291:24e34.

[11] Pessione E. Lactic acid bacteria contribution to gut microbiota complexity:lights and shadows. Front Cell Infect Microbiol 2012;2:86.

[12] Riley MA, Gordon DM. The ecological role of bacteriocins in bacterialcompetition. Trends Microbiol 1999;7:129e33.

[13] Joo NE, Ritchie K, Kamarajan P, Miao D, Kapila YL. Nisin, an apoptogenicbacteriocin and food preservative, attenuates HNSCC tumorigenesis viaCHAC1. Cancer Med 2012;1:295e305.

[14] Shaikh F, Abhinand P, Ragunath P. Identification & characterization of Lacto-bacillus salavarius bacteriocins and its relevance in cancer therapeutics. Bio-information 2012;8:589e94.

[15] Balciunas EM, Castillo Martinez FA, Todorov SD, de Melo Franco BDG,Converti A, de Souza Oliveira RP. Novel biotechnological applications of bac-teriocins: a review. Food Control 2013;32:134e42.

[16] Di Pierro F, Adami T, Rapacioli G, Giardini N, Streitberger C. Clinical evaluationof the oral probiotic Streptococcus salivarius K12 in the prevention of recurrentpharyngitis and/or tonsillitis caused by Streptococcus pyogenes in adults.Expert Opin Biol Ther 2013;13:339e43.

[17] Malik DK, Bhatia D, Nimbriya A, Kumar S. Lactic acid bacteria and bacteriocin:a review. J Pharm Res 2012;5:2510e3.

[18] Flint HJ, Scott KP, Duncan SH, Louis P, Forano E. Microbial degradation ofcomplex carbohydrates in the gut. Gut Microbe 2012;3:289e306.

[19] Hamer HM, Jonkers D, Venema K, Vanhoutvin S, Troost FJ, Brummer RJ. Re-view article: the role of butyrate on colonic function. Aliment Pharmacol Ther2008;27:104e19.

[20] Duncan SH, Aminov RI, Scott KP, Louis P, Stanton TB, Flint HJ. Proposal ofRoseburia faecis sp. nov., Roseburia hominis sp. nov. and Roseburia inulinivoranssp. nov., based on isolates from human faeces. Int J Syst Evol Microbiol2006;56:2437e41.

[21] Gomez A, Ladire M, Marcille F, Fons M. Trypsin mediates growth phase-dependent transcriptional regulation of genes involved in biosynthesis ofruminococcin A, a lantibiotic produced by a Ruminococcus gnavus strain froma human intestinal microbiota. J Bacteriol 2002;184:18e28.

[22] Strompfova V, Laukova A. Enterococci from pigletseprobiotic properties andresponsiveness to natural antibacterial substances. Folia Microbiol (Praha)2009;54:538e44.

[23] Workman SN, Been FE, Crawford SR, Lavoie MC. Bacteriocin-like inhibitorysubstances from Campylobacter spp. Antonie Van Leeuwenhoek 2008;93:435e6.

[24] Nigutova K, Morovsky M, Pristas P, Teather RM, Holo H, Javorsky P. Productionof enterolysin A by rumen Enterococcus faecalis strain and occurrence of enlAhomologues among ruminal Gram-positive cocci. J Appl Microbiol 2007;102:563e9.

[25] Joerger MC, Klaenhammer TR. Characterization and purification of helveticin Jand evidence for a chromosomally determined bacteriocin produced byLactobacillus helveticus 481. J Bacteriol 1986;167:439e46.

[26] Muriana PM, Klaenhammer TR. Purification and partial characterization oflactacin F, a bacteriocin produced by Lactobacillus acidophilus 11088. ApplEnviron Microbiol 1991;57:114e21.

[27] Beukes M, Bierbaum G, Sahl HG, Hastings JW. Purification and partial char-acterization of a murein hydrolase, millericin B, produced by Streptococcusmilleri NMSCC 061. Appl Environ Microbiol 2000;66:23e8.

[28] Aminov RI, Walker AW, Duncan SH, Harmsen HJ, Welling GW, Flint HJ.Molecular diversity, cultivation, and improved detection by fluorescent in situhybridization of a dominant group of human gut bacteria related to Roseburiaspp. or Eubacterium rectale. Appl Environ Microbiol 2006;72:6371e6.

[29] Louis P, Young P, Holtrop G, Flint HJ. Diversity of human colonic butyrate-producing bacteria revealed by analysis of the butyryl-CoA:acetate CoA-transferase gene. Environ Microbiol 2010;12:304e14.

of the dominant human colonic Firmicutes, Roseburia faecis M72/1,org/10.1016/j.anaerobe.2013.07.006

482483484