My microbes and me: A live-long relationship 1. The first come to us - - PowerPoint PPT Presentation

Peer Bork

Structural and Computational Biology EMBL, Heidelberg

Aiming at a functional understanding of biological systems

My gut microbes and me:

Exploring this live-long and fragile relationship

My microbes and me: A live-long relationship

• 2. Mother milk brings another boost (Bifidobacteria 20%), bottle-fed different
• 4. In parallel, the “inherited” part is being replaced by individually acquired

microbiota, thus environment plays a big role

• 1. The first come to us latest at birth (Caesarean delivery brings different bugs)
• 3. It takes 3-4 years until microbial composition is stabilizing, in interaction with
• ur immune system
• 5. The microbial composition still changes somewhat in childhood, seems

relatively stable in adulthood and diversity reduces in elderly, presumably associated with settled life style

Madan et al., JAMA Pediatr. (2016) 1

Müller et. al.,Trends Mol.Med 21(2015)109 Claesson et al., Nature 488(2012)178 Bork et al., in preparation

Gut microbiome: 2010 not even basics known How many microbial species and genes are in the gut?

Frequent genes saturate, but rare genes keep being added

Collaboration with BGI (China) and the EU MetaHit consortium Qin et al, Nature 464(2010)59

Human gut reference catalogue of 3.3 Mio microbial genes from 124 Europeans

2014: >10 Mio genes in >1200 individuals, with a stable core Li et al, Nat.Biotech. 32(2014)834

First illumina-based metagenomics: 10x cheaper, large scale, ca 250 species More bacterial than human cells, 1.5kg, >1000 species per person

3 distinct community types at genus level…

How different are our gut microbes?

Nature 473(2011)174

Enterotypes in the landscape of gut microbial community composition

Paul I. Costea1,*, Falk Hildebrand[1],[2],[3],*, Manimozhiyan Arumugam[4],[5], Fredrik Bäckhed[6],[7], Martin J. Blaser[8], Frederic D. Bushman[9], Willem M. de Vos[10],[11], S. Dusko Ehrlich[12],[13], Claire M. Fraser[14], Masahira Hattori[15], Curtis Huttenhower[16], Ian B. Jeffery[17], Dan Knights[18],[19], James D. Lewis[20], Ruth E. Ley[21], Howard Ochman[22], Paul

• W. O’Toole17, Christopher Quince[23], David A. Relman [24],[25],[26], Fergus Shanahan17,

Shinichi Sunagawa1, Jun Wang5,[28],[29],[30],[31], George M. Weinstock[32], Gary D. Wu[33], Liping Zhao[34], Jeroen Raes 2,3,[35],#, Rob Knight [36],[37],[38],[39],#, Peer Bork1,[40],[41],#

…now in context of a complex composition landscape

Paul Costea, Falk Hildebrand et al., in preparation

ET2 ET3 ET1

Enterotypen in the media ... and the phone kept ringing

www.bork.embl-heidelberg.de

my.microbes.eu

Donate or participate!

...

DNA extraction Sequencing Analysis Sampling

Operates since

• Sep. 2011

Goal is 5000 samples Currently 600 Euro 500k Euro so far

Enterotypes Version 2 approved by EMBL-ethics commission

Personalized my.microbes report

No approval for reporting antibiotics resistances, pathogen detection und some diseases

Our gut microbiome is linked to a multitude of different diseases

Crohn’s disease

• Gut 2006

Arthritis

• Nat. Rev.

Rheumatology 2011

Autism

• J. Med. Microbiol. 2005

Multiple Sclerosis

• Nature 2011

Parkinson Disease

• Eu. J. Neurosci. 2009

Obesity

• Nature 2006

Diabetes

• Nature (2012)

NASH

• Nature 2012

Athero- sclerosis

• Nature 2011

Colo- rectal cancer

• Genome Res. 2012

Neurological disorders Metabolic diseases Cardiovascular diseases Cancer Inflammatory diseases

Association of microbiota with colon cancer

Stool samples from 156 French individuals provided by Iradj Sobhani

Study design (cancer detection)

Per individual: on av. 9 Gb are currently sequenced, i.e. almost 3 human genome equivalents

Problem: Full cost model for metagenomics ca 1000 Euro, FOBT = 6 Euro

In combination with the “fecal occult blood test” (FOBT) >45% more cancers detected as FOBT alone

Zeller, Tap, Voigt et al, Mol.Sys.Biol. 10(2014)766

2014 Association of microbiota with colon cancer

22 Bacterial species in stool function as biomarkers in early cancer stages and can be utilized for non-invasive screening before colonoscopy

Cancer risk group: >50 years or >40 years, if obese etc. Stool sampling, preparation 384 well plate screening for marker genes of 22 species

Of each of the 22 species, specific primers from marker genes are used only Currently: ca 40 Euro total costs per sample (but can still be reduced)

qPCR based readout

Association of microbiota with colon cancer 2016

Some proteobacteria up after/during metformin treatment

Test for many confounding factors (diet, drug treatment etc.)

Metformin induces gut microbial composition changes in type II diabetes individuals

Chinese: Qin et al., Nature 2012; AUC 0.81 Swedish: Karlsson et al., Nature 2013; AUC 0.83 Danish: unpublished (54 T2D+75Ctrl); AUC 0.81

Metformin is strongest signal, distinct from T2D which alone is weak Combined cohort after omission of metformin-treated individuals: AUC = 0.53 Forslund, Hildebrand et al., Nature 228(2015)262 (Metahit)

Antibiotics (~15%) Antiparasitic, antifungal, antiviral (~8%) Human-targeted drugs (~77%)

Chemical library

96 and 384-well plates/species, all measures in triplicates

Screen of 40 prevalent gut bacterial species with ca 1200 FDA- approved drugs shows widespread fitness effects

Anaerobic incubation and

• ptical density readout

Bacterial growth patterns

Measuring growth differences after drug perturbation

Time (hours) OD (578 nm)

In progress: With Zeller, Typas and Patil groups at EMBL

Drug-bug interactions

>> 10% of human non-antibiotics drugs inhibit or reduce growth of gut bacteria, already at low dosages

High resolution microbiomics: From species to strains

Schloissnig et al., Nature 493(2013)45 Zhu et al., Genome Biol. 16(2015)82 Our microbial strains are individual, even in monozygotic twins Thus microbiome amplifies genetic individuality (e.g. in digestion) and could be target for personalized therapies Human – Human ca 0.3% SNPs Gene content Human – Human ca 0.2% Human – Chimpanzee ca 1% E.coli – E.coli ca 5% E.coli – E.coli ca 20% Human – Chimpanzee ca 2%

Each of us carries individual bacterial strains and at least healthy keep them for a long time

Faecal Microbiota Transplantation (FMT)

• Transfer of stool from a healthy donor to patient

– Usually following antibiotics treatment or bowel lavage

• Positive effects reported in GI and non-GI diseases

– Over 90% success in treating Clostridium difficile infection1

• Mechanism is currently unknown, e.g.

fate of native and introduced strains

– Specific bacteria introduced in patient2 – Replacement or ‘repair’ of ‘bad’ microbial species

1. van Nood, E. et al. (2013). N Engl J Med, 368, 407-15. 2. Lawley, TD., et al. (2012). PloS Pathog, 8, e1002995.

Analysis usually at species level, but most species are shared

Donor species and strain colonization after FMT for metabolic syndrome

Species Strains

5 time points up to 3 months after FMT, 164 metagenomes incl. donors

Measured using marker genes (mOTUs) Measured using discriminative SNVs, modified from Schloissnig et al., Nature 493(2013)45 Sunagawa et al., Nature Meth. 10(2013)1196

• Collab. with W. de Vos

and M. Nieuwdorp

Strain replacement after faecal microbiota transplantation (FMT) is easier than acquisition of new species

Donor species barely above random fluctuation Donors strains with durable colonization, often in coexistence

Li et al., Science 352(2016)586

• 2. Donor strains can colonise and persist
• ver at least 3 months
• 3. New donor strains colonise better

than new donor species, perhaps by being invisible to the host immune system

Strain replacement implies personalized treatment options, e.g. by replacing multidrug resistance.

Species Strains

• 1. There is no “superdonor” – 1 donor has

different outcomes

Screen for presence of helpful/harmful gut microbes

Basis for powerful health screenings in the future

www.bork.embl-heidelberg.de

Acknowledgements

Gut team

Shini Sunagawa Luis Coelho Georg Zeller

+ All other current and former group members as well as visitors….

Tools

Jens Kultima Ana Zhu Kristoffer Forslund Anita Voigt Simone Li Paul Costea Falk Hildebrand Matt Hayward Renato Alves Marja Driessen

Collaborators

IHMC IHMS, Metacardis, MetaHit (EU) D.Ehrlich

• W. De Vos (NL/FL)

S.Sunyaev (Harvard)

• I. Sobhani, (UPEC, F)
• M. von Knebel (HD)
• W. de Vos (Helsinki)

Genecore facil. (EMBL) TARA Oceans consort.

• R. Pepperkok (EMBL)

Ece Kartal

Thank you!

For details see: www.bork.embl.de

The Microbial Environment and its Influence on Allergy and Asthma in Early Life

Sabina Illi

On behalf of Erika von Mutius

• Dr. von Hauner Children‘s Hospital

Ludwig Maximilians University Munich, Germany

GABRIEL Study: Prevalences between farm and non farm children

%

5 10 15 20 25 30 35 40 45 asthma current asthma severe wheeze atopy hay fever atopic dermatitis Farm Non farm

European PASTURE Birth Cohort (N=1,133)

birth 2mo 1y 2y 3y 4y 5y 6y

Cord blood immune CrP asthma responses atopy atopy atopy

postnatal

Recruitment questionnaire in pregnancy diary yearly questionnaires until age 6 years

aOR from GEE model: 0.55 (95% CI: 0.43-0.71)

Contact to stable in 1. year of life and wheeze episodes

No contact to stable Contact to stable Age in weeks

Loss et al, AJRCCM 2016

Probability for wheeze in year 1

aOR from GEE model: 0.55 (95% CI: 0.43-0.71)

Contact to stable in 1. year of life and wheeze episodes

No contact to stable Contact to stable Age in weeks

Loss et al, AJRCCM 2016

Is protection mostly in those at risk of developing asthma with virus induced wheezing, i.e. carriers of the chromosome 17q21 asthma susceptibility locus ? These account for 75% of the total population.

aOR for contact to stable and wheeze in year 1

Loss et al, AJRCCM 2016

Only genotype GA/AA susceptible for protection from wheeze via environmental exposure

Chromosome 17q21 locus interaction with contact to stable

ORMDL3

asthma risk genotype

Chromosome 17q21 locus interaction with contact to stable

The same genotype that increases the risk of asthma in children with early virus induced wheeze also protects from asthma if children are exposed to stables during their first year of life.

Loss et al, AJRCCM 2016

Consistency of effects in GA/AAs across 5 PASTURE populations

aOR for contact to stable and wheeze in year 1

Loss et al, AJRCCM 2016

Dose response effect in GA/AAs

aOR for contact to stable and wheeze in year 1

Loss et al, AJRCCM 2016 No contact to stable Less than 2 hours per week More than 2 hours per week

WROCŁAW

Rural and urban life style in Poland

Sozanska et al, Allergy 2007

Sobotka: 1% live

• n farms

Villages: 55% live

• n farms

lowest prevalence in Europe

Rural and urban life style in Poland

Sozanska et al, Allergy 2007

UK-type prevalence Birth cohort (age) Prevalence of atopy (%)

lowest prevalence in Europe

Rural and urban life style in Poland

Birth cohort (age) Prevalence of atopy (%)

Sozanska et al, Allergy 2007

Difference appears to be explained by the cohort effect

• f a communal move

away from rural life.

Change in lifestyle in the villages after accession to the European Community (2004)

Age group

6-10 11-20 21-30 31-40 41-50 51-60 61+

Villages: participated in both surveys % atopic

Sozanska et al, JACI 2014

 Repetition of the study in 2012 in same population

Sozanska et al, JACI 2014 Villages 2003 2012 Regular or occasional contact with: Cows 24.3% 4.3% <0.001 Pigs 33.5% 14.0% <0.001 Poultry 46.8% 37.1% <0.001 Sheep or goats 3.2% 3.1% 0.900 Horses 0.8% 1.9% 0.040 Regular or occasional: Milking cows 12.0% 2.7% <0.001 Cleaning barns or stables 28.5% 15.6% <0.001 Collecting eggs 34.4% 28.4% 0.010 Drinking unpasteurized milk 34.9% 8.7% <0.001

Change in lifestyle in the villages after accession to the European Community (2004)

Increase in the prevalence of atopy after accession to the European Community (2004)

Only villagers that participated in both surveys Prevalence of atopy (%)

Sozanska et al, JACI 2014 2012 2003

 no change in prevalence in Sobotka

Increase in the prevalence of atopy after accession to the European Community (2004)

Sozanska et al, JACI 2014 2003 2012 N aOR 95%-CI Living on farm: No No 403 1.00 reference Yes No 72 1.07 0.51-2.50 No Yes 53 1.13 0.52-2.21 Yes Yes 243 0.38 0.20-0.72 Contact with cows: No No 620 1.00 reference Yes No 121 0.70 0.36-1.33 No Yes 10 − − Yes Yes 20 0.25 0.03-1.97 Contact with pigs: No No 537 1.00 reference Yes No 140 0.48 0.25-0.93 No Yes 22 0.59 0.17-2.12 Yes Yes 72 0.36 0.14-0.95 Contact with poultry: No No 430 1.00 reference Yes No 118 0.97 0.52-1.81 No Yes 49 1.27 0.57-2.84 Yes Yes 174 0.41 0.20-0.83 2003 2012 N aOR 95%-CI Living on farm: No No 403 1.00 reference Yes No 72 1.07 0.51-2.50 No Yes 53 1.13 0.52-2.21 Yes Yes 243 0.38 0.20-0.72 Contact with cows: No No 620 1.00 reference Yes No 121 0.70 0.36-1.33 No Yes 10 − − Yes Yes 20 0.25 0.03-1.97 Contact with pigs: No No 537 1.00 reference Yes No 140 0.48 0.25-0.93 No Yes 22 0.59 0.17-2.12 Yes Yes 72 0.36 0.14-0.95 Contact with poultry: No No 430 1.00 reference Yes No 118 0.97 0.52-1.81 No Yes 49 1.27 0.57-2.84 Yes Yes 174 0.41 0.20-0.83

Protective effect was maintained if farming was not given up

What are protective factors in stables ?

The diversity of microbial exposure and asthma

Ege et al, NEJM 2011

Cross-sectional study School aged children Dust from children‘s mattresses screened for bacterial DNA Cross-sectional study School aged children Settled dust from children‘s rooms analysed for bacteria and fungi by culture techniques

The diversity of microbial exposure is inversely related to asthma

Ege et al, NEJM 2011

Environmental microbial cocktail

Bacteria: Staphylococcus sciuri, Staphylococcus sp., Salinococcus sp., Macococcus sp, Bacillus sp., and Jeotgalicoccus sp., Listeria monocytogenes, Bacillus licheniformis, Bacillus sp., Corynebacterium sp., Methylobacterium sp., Xanthomonas sp., Enterobacter sp., Pantoea sp., Acinetobacter lwoffii and others. Fungi: Eurotium sp; Penicillium sp

Ege et al, NEJM 2011

Changes in the microbiome ?

Asthma Atopy

Changes in the microbiome ?

Nasal microbiome N=74 Throat microbiome N=327 More diversity in both

More diversity: less asthma More Moraxella: more asthma

Nose:

Throat: no signal with asthma

Bacterial indoor & outdoor exposures in the US

1100 homes with indoor and outdoor microbiome collections

Barberan et al, Proc R Soc B 2015

No clear regional similarity pattern

Bacteral richness higher indoors than outdoors: Home occupants matter

• utdoor,

skin, feces, vagina, insects

Barberan et al, Proc R Soc B 2015

Compos ition:

Protection from allergic asthma and RSV infection by dog in home

House dust from homes with and without a dog

• Protection from allergic

(OVA, cockroach) asthma.

• Protection from RSV

induced infection

Fujimura et al, PNAS 2014

Change in gut microbiome composition

by gavag e

Conclusions

• An environment rich in microbial exposures

protects from developing allergy, asthma and wheeze early in life.

• This protection is mediated by the diversity of

microbial exposures. Some microbes or microbial cocktails may be particularly beneficial.

• While infants may be particularly

responsive to microbial environments, effects may also be found in adulthood.

• Thus, continued microbial exposure may

matter rather than a single period in life such as the first year of life.

Conclusions

• In non-farm environments exposure

to dogs may be protective, potentially due to alterations in the microbiome.

• The pathway from environmental

microbiome to human microbiome to disease prevention is still not clear.

Conclusions

Munich: Erika von Mutius Bianca Schaub Markus Ege Martin Depner Antje Legatzki Juliane Weber Georg Loss PASTURE / EFRAIM / GABRIEL: Charlotte Braun-Fahrländer Roger Lauener Caroline Roduit Josef Riedler Jean-Charles Dalphin Dominique Vuiton Juha Pekkanen Jon Genuneit Elisabeth Horak Bill Cookson Miriam Moffat Michael Cox Dick Heederik Donata Vercelli Carole Ober Bart Lambrecht, Hamida Hamad