Impact of Diet Upon Intestinal Microbiota and Microbial Metabolites

Impact of diet upon intestinal microbiota and microbial metabolites

Harry J Flint

Microbiology Group Rowett Institute of Nutrition and Health University of Aberdeen, UK Impact of the gut microbiota on human nutrition and health metabolism of non-digestible modification of host dietary components secretions (mucin, bile, supply of additional gut receptors..) energy, vitamins immune function, inflammation release of phytochemicals, pathogenesis enterohepatic circulation of xenobiotics, drugs barrier function

Gut function, gut disorders: Colitis; Irritable bowel syndrome Colorectal cancer Infections Systemic effects: Energy supply, satiety Diabetes Heart disease Autoimmune disorders Modulation of the human gut microbiota by dietary fibres occurs at the species level [Chung WSF et al BMC Biology, 2016]

In vitro chemostat studies

[Separate experiments with 3 fecal microbiota]

Bacteroides spp. Modulation of the human gut microbiota by dietary fibres occurs at the species level [Chung WSF et al BMC Biology, 2016]

In vitro chemostat studies

[Separate experiments with 3 fecal microbiota]

Microbiota composition (family level) Modulation of the human gut microbiota by dietary fibres occurs at the species level [Chung WSF et al BMC Biology, 2016]

In vitro chemostat studies

[Separate experiments with 3 fecal microbiota]

Firmicutes DoesImpact the type of of dietary dietary noncarbohydrate-digestible alter carbohydrates the colonic in vivo

14 obese male volunteers with metabolic syndrome (mean age 54 years, mean BMI 39.4 kg/m2) Collection of faeces M: maintenance diet NSP: high non-starch polysaccharide, low RS M NSP RS WL RS: high resistant starch, low NSP 1 wk 3 wks 3 wks 3 wks WL: high protein, moderate carbohydrate M RS NSP WL

Mean dietary intake [g/d]:

Diet CHO starch RS NSP protein fat

M 427 230 5 28 103 126 Weight NSP 427 138 2 42 102 136 maintenance RS 434 275 26 13 109 127 WL 201 110 3 22 144 63 Weight loss

CHO: carbohydrate

added wheat bran

added type 3 resistant starch Walker AW et al. ISME J 2011 Some but not all species are diet-responsive Lower diversity with RS than WB NSP

[16S rRNA gene sequences]

350

300

250

200

150 Inverse Simpsonindex Inverse 100

Walker AW et al (2011) ISME J 5, 220-230 M NSP RS WL Salonen A et al (2014) ISME J 8, 2218-2230 [Hitchip] Increase with wheat bran NSP Increase with RS Some but not all species are diet-responsive Lower diversity with RS than WB NSP

[16S rRNA gene sequences]

350

300

250

200

150 Inverse Simpsonindex Inverse 100

Walker AW et al (2011) ISME J 5, 220-230 M NSP RS WL Salonen A et al (2014) ISME J 8, 2218-2230 [Hitchip] Increase with Increase with RS

wheat bran NSP Individual variation still dominates

Pearson

1 correlation

Volunteer No. Very extreme dietary changes can lead to wide-ranging shifts in the gut bacterial community in humans

‘animal based’ – 69.5 % (!) cals from fat, 30% from protein, <1% (!) from carbs/fibre

‘plant based’ - 32.5% from fat, 16.5% from protein, 50% from carbs/fibre

[David LA et al Nature 2014]

Rise in bile acids with high fat + lack of fibre considered key factors in ‘animal-based’ diet Responses occur rapidly and tend to be species-specific

Cluster IV Ruminococcus species - qPCR

Volunteer 19 Volunteer 16 R. bromii M NSP RS WL 50 50 M RS NSP WL R. albus

40 40 + R. bicirculans gene 30 30

R. flavefaciens rRNA 20 20 + R. champanellensis

10 10 + R. callidus % % 16S

0 0 R. champanellensis 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70

Volunteer 24 Volunteer 20

50 M NSP RS WL 50 M RS NSP WL

40 40 gene 30 30 Response on

rRNA 20 20 RS is due to

10 10 R. bromii % % 16S

0 0 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 time [days] time [days]

Walker AW et al. ISME J (2011); Salonen A et al. ISME J (2014) Nutritional specialization among Ruminococcus spp.

- selected glycoside hydrolase families encoded by genomes

xylanase/ amylase/ b-glucanase/ xylosidase/ a-glucosidase/ mannanase/ arabino- amylomaltase xyloglucanase cellulase furanosidase

12 R. bicirculans 80/3 Human colon

10 R. champanellensis 18P13 Human colon 8

6 R. flavefaciens FD1 Rumen (cellulolytic)

4 R. bromii L2-63 Human colon 2

0 GH2 GH3 GH5 GH26GH44GH74GH16GH48 GH9 GH10GH11GH43GH51GH13GH31GH77 “Ruminococcus bicirculans”

Able to utilize xyloglucan, beta-glucan for growth, but not cellulolytic

Complete genome of a new Firmicutes species belonging to the dominant human colonic microbiota (‘Ruminococcus bicirculans’) reveals two chromosomes and a selective capacity to utilize plant glucans. Wegmann U et al (2014) Environ Microbiol 16:2879-2890

Ruminococcus champanellensis

The only human colonic anaerobe able to degrade filter paper (crystalline) cellulose

Ruminococcal cellulosome systems from rumen to human. Ben David et al (2015) Environ Microbiol 17:3407-3426 Ruminococcus bromii

Specialist starch-degrader

Exceptional starch-degrading activity in the human colonic anaerobe Ruminococcus bromii coincides with unique organization of its extracellular enzymes into ‘amylosomes’. Ze X et al (2015) MBio 6 (5) e01058-15. Ruminococcus bromii

Specialist starch-degrader

Contrast - Bacteroides thetaiotaomicron Mainly periplasmic amylases starch utilization (sus) system (Salyers) susR susA susB susC susD susE susF susG

- soluble starches D G E F C OM B

TonB CM complex Ruminococcus bromii

Specialist starch-degrader

Degradation of corn starches by four amylolytic human gut bacteria [Ze X et al 2012 ISME J]

Digestible starch (S1) Resistant starches (S2, S3 (RS2) S4 (RS3))

100% Ruminococcus bromii 80% (Firmicutes)

60% Bifidobacterium adolescentis Gram- (Actinobacteria) positive 40% Eubacterium rectale 20% (Firmicutes)

0% Bacteroides thetaiotaomicron Gram- S1 S2 S3 S4 (Bacteroidetes) negative starches (boiled 10 mins) R. bromii as a ‘keystone’ species in degradation of [Walker AW et al. ISME J 2011 resistant starch Ze X et al. ISME J 2012]

qPCRqPCR - % analysis Ruminococcus – all timepoints(cl IV) v low R. bromii 16S rRNA gene copies 60 50

0 11 12 17 18 19 23 24 14 15 16 20 22 25 26 Rum % (RS diet) Rum % (NSP) Residual starch Incomplete starch fermentation in faeces 100

80 RS diet NSP diet 60

0 11 12 17 18 19 23 24 14 15 16 20 22 25 26

% residual faecal starch faecal residual % volunteer Microbial ecology of carbohydrate utilization in the gut: functional groups

(Flint HJ et al Env Micro 2007)

Insoluble complex carbohydrates ruminococci on faecal fibre

Primary degraders

Diet Soluble polysaccharides H-utilisers (acetogens, methano- Polysaccharide utilisers Metabolic gens, SRB); products lactate Oligosaccharides, sugars utilisers

Oligosaccharide/ sugar utilisers Enrichment of Firmicutes bacteria on wheat bran in vitro*

50 Otu039 Oscillospira guilliermondii (94% similarity) 40 Otu030 Coprococcus sp. ART55/1/eutactus (97% similarity) Otu029 Clostridium lactatifermentans 30 Otu028 Butyrate-producer A2-166 Otu024 Lachnospiraceae bacterium A4 (C. hathewayi - 96% similarity)

20 Otu017 Eubacterium siraeum (97% similarity) % % sequences Otu005 Roseburia faecis/hominis/intestinalis

10 Otu006 Butyrate-producer T2-145 (96% similarity) Otu007 Eubacterium xylanophilum (97% similarity) 0 2h 4h 8h24h48h 2h 4h 8h24h48h 2h 4h 8h24h48h 2h 4h 8h24h48h D1 D2 D3 D4

inflow 100% CO2 atmosphere pH- controlled, anaerobic in vitro fermentor system -

outflow

* Duncan SH et al. Environ Microbiol 2016 online medium % sequences acid” acid” microbiota of butyrate “Wheat bran promotes enrichment within the human colonic 10 20 30 40 50 60 0 Duncan Duncan

2h 4h SH D1 8h

24h etal 48h Environ 2h 4h D2 -

8h Microbiol 24h producing bacteria that release 48h

2h - 4h online 2016

D3 8h 24h 48h

2h 4h D4 8h 24h 48h Otu007 Otu006 Otu005 Otu017 Otu024 Otu028 Otu029 Otu030 Otu039 ferulic Action of isolated bacteria on wheat bran Otu017 acid ferulic phenolics weight loss (degradation) - relat ed Otu006 - related Otu005 - relat ed Impact of diet on plant-derived phenolic acids and [Russell WR et al microbial metabolites in human volunteers AJCN 2011]

4OH-3OMe-PPA 3,4OH-PPA 3OH-PPA

HO O HO O HO O HO O

ferulic acid hydrogenation demethylation dehydroxylation

O O HO HO OH OH OH 25

Faecal concentrations P = 0.007

) 3 - 20

(means, 8 overweight

g cm g  male volunteers) 15 P < 0.001 10 P < 0.001

Concentration Concentration ( 5

ferulic acid metabolite 1 metabolite 2 metabolite 3

Maintenance (carbs 400g, NSP fibre 28 g/day) diets High Protein WL (carbs 170g, NSP fibre 12 g/day) High Protein WL (carbs 23g, NSP fibre 6 g/day) The gut microbiota, bacterial metabolites and colorectal cancer [Louis P, Hold GL & Flint HJ 2014 NRM] Major fermentation pathways hexoses & pentoses hydrogen methane sulphide fucose, rhamnose Desulfovibrio spp. methanogenic sulfate archaea PEP DHAP + L-lactaldehyde H2 + CO2

Blautia formate hydrogenotrophica pyruvate oxaloacetate Bacteroidetes Eubacterium hallii some Negativicutes acetate acetyl-CoA Anaerostipes spp.

acetoacetyl-CoA Ruminococcus CO2 Roseburia inulinivorans bromii fumarate Ruminococcus obeum lactate Salmonella enterica Coprococcus catus succinate Megasphaera elsdenii butyryl-CoA ethanol acetate acetyl-CoA lactoyl-CoA propane-1,2-diol butyryl-P Eubacterium rectale Eubacterium hallii Roseburia spp. propionyl-CoA propionyl-CoA propionyl-CoA Anaerostipes spp. Coprococcus eutactus Coprococcus catus Coprococcus comes Faecalibacterium prausnitzii

butyrate propionate [propionate pathways -Reichardt N et al (2014) ISME J] Kettle H, Louis P, Duncan SH, Holtrop G, Flint HJ (2015) “Modelling the emergent dynamics of communities of human colonic microbiota: response to pH and peptide” Environ Microbiol 15: 1615-1630

NSP oligo- starch protein Bacterial functional groups saccharides, B1 = Bacteroidetes sugars B2 = Firmicutes (eg. R. bromii) B3 = Firmicutes (eg. E. eligens) B3, B6 B2, B4, B5 B1 B4 = Actinobacteria B1-B9 B5 = Roseburia group B6 = F. prausnitzii B7 = Negativutes B8 = E. hallii, Anaerostipes B9 = acetogens PEP B3, B1 B10 = methanogens pyruvate succinate formate B1-B9 H2 + CO2 B1 acetyl CoA B9 B10 lactate B5, B6 B8 acetate CH4 B7 butyrate propionate Changes in butyrate-producing bacteria within faecal microbiota in disease states/ extreme diets –

F. prausnitzii in ileal Crohn’s patients Sokol H et al PNAS (2008)

F. prausnitzii in frail elderly Claesson et al Nature (2012)

Roseburia, E. rectale, F. prausnitzii in type 2 diabetics Qin J et al Nature (2012)

Butyrate- producers in type 1 diabetics de Goffau MC et al Diabetes (2013)

Roseburia, butyrate-producers in CR cancer Wang T et al ISME J (2012)

Roseburia in constipation-type-IBS Chassard C et al Alim Ph Th (2012)

Roseburia in hepatic encephalopathy (mucosal flora) Bajaj JS et al Am J Clin Nutr (2012)

Butyrate-producers with improved insulin sensitivity Vrieze A et al Gastroenterol (2012)

Roseburia, E. rectale with low carb WL diets Duncan SH et al AEM (2007)

Roseburia, Faecalibacterium, Coprococcus in LGC individuals (increased metabolic syndrome) Le Chatelier et al Nature (2013) How do changes in gut microbiota composition impact on metabolite production?

Dietary intake, medication, inter-individual variation

GUT MICROBIOTA GUT metabolic pathways community structure ENVIRONMENT cross-feeding

METABOLIC PRODUCTS Relative abundance of Bacteroidetes correlates with % propionate

- in faecal samples - and in fermentor studies

Clone libraries from 1 volunteer 40 40 [14 male volunteers] mmdA 35 35 R² = 0,558 30 16S rRNA 25 30 20 25 15 10 M 20 NS

% total clones% of 5 P 15

0 RS % propionate % 10

Proportional ofabundance Proportional 5 propionate (% of total total ofSCFA) (% propionate

0 0 20 40 60 80 100 % Bacteroidetes (qPCR) Bacteroides spp. and Prevotella spp. % of total 16S rRNA genes

Salonen A et al. ISME J (2014) Chung WSF et al. BMC Biology (2016) • Total faecal SCFA , especially butyrate, decrease with lower carbohydrate intake study 1 study 2 Obese male subjects (UK) 120 mM Faecal SCFA: 100 1. Duncan SH et al AEM 2007 isobutyrate 80 isovalerate 2. Russell WR et al Am J Cl Nutr 2011 60 valerate butyrate 40 Diets: propionate M – maintenance (50-52% cals as CHO) 20 acetate

HPMC – Moderate carbohydrate (35%) 0 HPLC – Low carbohydrate (4-5%) M1 HPMC1 LC1 M2 HPMC2 HPLC2

• Total faecal SCFA, including butyrate, can be very high in populations that have high intakes of plant-derived foodstuffs

Ou J et al Am J Clin Nutr 2013

Native Africans - Ace: Prop: But 72: 27: 17 mM (total > 100 mM) African Americans - Ace: Prop: But 27: 7: 8 mM (total < 50 mM) Model describing the relationship between substrate abundance and gut microbiota product formation and its dependence on gut microbiota composition using the consumption of fermentable carbohydrates and the production of short chain fatty acids (SCFAs) as an example.

Do ‘restrictive’ communities lack certain keystone species?

Gary D Wu et al. Gut 2016;65:63-72

Changes in diet composition change the representation of certain ‘diet- responsive’ species within the microbiota

This reflects the fact that many species (especially Firmicutes?) are highly specialised in their substrate utilization

‘Some are more equal than others’ (Ze X et al Gut Microbes 2013) (after George Orwell) - certain ‘keystone species’ play primary roles in degradation of recalcitrant substrates

Changes in microbiota composition (due to diet or individual variation) impact on metabolite formation – especially loss of keystone species Acknowledgements - Institute of Food Research, UK Nathalie Juge, Emmanuelle Crost Rowett Institute of Nutrition and Health Rowett Institute (U Aberdeen, UK) Wellcome Trust Sanger Institute, UK Trevor Lawley, Julian Parkhill Alan Walker Sylvia Duncan U Groningen, Netherlands Petra Louis Hermie Harmsen Karen Scott Xiaolei Ze U Wageningen, Netherlands Faith Chung Jerry Wells Jenny Laverde Nicole Reichardt Weizmann Institute, Israel Alexandra Johnstone Ed Bayer, Sarah Morais Biomathematics and Statistics Scotland U Helsinki, Finland Anne Salonen, Willem de Vos Grietje Holtrop, Helen Kettle U. Girona, Spain Mireia Lopez-Siles, Jesus Garcia-Gill Support from :–

Commercial collaborators