microbe-host interactions in inflammatory bowel diseases · microbe-host interactions in...
TRANSCRIPT
Microbe-Host Interactions in Inflammatory Bowel Diseases
Hera Vlamakis
Oct 3, 2018
Mowat & Agace 2014 doi:10.1038/nri3738
Stool is proxy for gut microbes
Most of the bacteria in your body are in your gut
HEALTH BENEFITS
Breakdown of polysaccharides
Synthesis of vitamins
Colonization resistance
Maturation of our immune system
DISEASE
Inflammatory bowel diseases
Specific microbes enriched or
depleted
• Do infants acquire bacterial strains from their mother?
• Are different bacterial strains present in disease?
• Are microbial species associated with response to treatment or disease severity?
Today’s Questions:
Early life events shape the microbiome
Koenig et al, PNAS 2011
By age 3 years, the gut microbiome is adult-like.Yatsunenko et al. Nature 2012
Delivery mode affects the newborn’s gut microbiome.Dominguez-Bello et al. PNAS 2010Bäckhed et al., Cell Host & Microbe, 2015
Asnicar et al. mBio 2017
Large-scale paired longitudinal study of
both mother and child samples
Yassour et al., Science Trans. Med., 2016
Vertical transmission of microbes from mother to infants.
Longitudinal mother-infant cohort design
33 families in total = metagenomic sequencing for 264 samples.
Mother
Child
Age (months)
0 1 2 3-1-2
How similar are samples from the same family?
Yassour, Jason, Hogstrom et al. 2018 Cell Host and Microbe
Phylum-level microbial trajectories
ActinobacteriaBacteroidetesFirmicutesProteobacteriaVerrucomicrobia
Phylum:
Family
Birth
14 days
1
mo.
2
mo.
3
mo.
Infant profiles
Rela
tive a
bu
nd
an
ce
Gest.
week
27-28
Birth
3
mo.
Maternal profiles
Family
Rela
tive a
bu
nd
an
ce
Can we find distinctive species that were inherited by the child from the mother?
Do mother and child share their strains?
Yassour, Jason, Hogstrom et al. 2018 Cell Host and Microbe
ChildChildChild
Mother MotherMother
Dominant strain
transferred
Otherstrain
transferred
Secondarystrain
transferred
We are limited to species for which we have enough coverage
for both mother and child samples.
Do mother and child share their strains?
Dominant
strain
transferred
Secondary
strain
transferred
Neither strains
transferred
Z-sc
ore
fo
r se
con
dar
yst
rain
tra
nsf
er
Z-score for dominantstrain transfer
Yassour, Jason, Hogstrom et al. 2018 Cell Host and Microbe
mothersample
childsample
1 2 3 1 2 3
Mother dominant nucleotideMother secondary nucleotideOther nucleotide
z-sc
ore
fo
r s
ec
on
da
ryst
rain
tra
nsf
er
Bacteroides uniformis
Do mother and child share their strains?
Z-score for dominantstrain transfer
Yassour, Jason, Hogstrom et al. 2018 Cell Host and Microbe
mothersample
childsample
1 2 3 1 2 3
Mother dominant nucleotideMother secondary nucleotideOther nucleotide
Transmission pattern varies across species
z-score for dominant strain transfer
z-sc
ore
fo
r s
ec
on
da
ryst
rain
tra
nsf
er
Mostly secondary
strain transmissions
Mostly dominant
strain transmissions
Functional differences in maternal strains may
elicit inheritance mode
z-score for dominant
strain transfer
z-sc
ore
fo
r seco
nd
ary
stra
in t
ran
sfer
MISSING
FUNCTION?
Bacteroides uniformis
Yassour, Jason, Hogstrom et al. 2018 Cell Host and Microbe
z-score for dominant
strain transfer
z-sc
ore
fo
r seco
nd
ary
stra
in t
ran
sfer
MISSING
FUNCTION?
Bacteroides uniformis
Functional differences in maternal strains may
elicit inheritance mode
Yassour, Jason, Hogstrom et al. 2018 Cell Host and Microbe
Functional differences in maternal strains may
elicit inheritance mode
Maternal strains that can utilize human milk oligosaccharides
may be more likely to colonize the infant gut
Starch utilization system (sus) locus
• Do infants acquire bacterial strains from their mother?
• Are different bacterial strains present in disease?
• Are microbial species associated with response to treatment or disease severity?
Today’s Questions:
Intestinal epithelium
Environmental factors
DietHygiene
Microbial community
Genetic susceptibility
Immune systemInflammatory and non-inflammatory
organisms
PathogensAntibiotics
Dendritic cell
Paneth cell
IBD is caused by dysfunction in the composition of and interactionsbetween microbes, intestinal epithelium, and the immune system
Factors that influence development of
inflammatory bowel diseases
Differentially abundant species in IBD
Up in IBD species
Ruminococcus gnavus
Escherichia coli
Streptococcus salivarius
Streptococcus parasanguinis
Haemophilus parainfluenzae
Enterococcus faecium
Down in IBD species
Faecalibacterium prausnitzii
Eubacterium rectale
Ruminococcus bromii
Alistipes finegoldii
Clostridium leptum
Odoribacter splanchnicus
Obligate anaerobe
Facultative anaerobe
Hypothesis: Increased production of reactive oxygen species in IBD causes a decrease in obligate anaerobes and replacement by facultative anaerobes that can tolerate ROS Morgan et al. 2012 (Huttenhower)
Lewis et al. 2015 (Wu) Ballal et al. 2015 (Garrett)
Hall, Yassour et al. 2017 Genome Medicine
Transiently increased abundance of R. gnavus in IBD
Ind
ivid
ual
s
Ind
ivid
ual
s
Hall, Yassour et al. 2017 Genome Medicine
Ge
ne
cluste
rs
Metagenomic samples
Dark blue – presenceLight blue – absence
Clade 2
Pangenome analysis of R. gnavusClade 2 is only found in adults with IBD
Clade 1Metagenomic data from• IBD: LSS, Lewis et al. • Healthy: HMP, 500FG
R. gnavus clade 2 found in IBD patients from 3 clinical centers
Hall, Yassour et al. 2017 Genome Medicine
199 Clade 2 specific genes• Responses to oxidative stress• Mucus utilization• Adherence • Iron acquisition
• Do infants acquire bacterial strains from their mother?
• Are different bacterial strains present in disease?
• Are microbial species associated with response to treatment or disease severity?
Today’s Questions:
Study overview: PROTECT
0 4 12 52
Sample collection weeks:
Pediatric patients often exhibit a more severe clinical phenotype
compared to adults
2 treatment strategies
Strategy 1:
mesalamine
Strategy 2:
corticosteroids mesalamine
Evaluate efficacy
of standardized
initial therapy
405 pediatric, treatment-
naive cohort with new-
onset UC
263 rectal biopsies
949 stool
samples--------------------------
1,212 samples
Schirmer et al. 2018 Cell Host and Microbe
While disease severity improved for many patients, only 34% achieved week
12 remission and 37% achieved week 52 remission.
Severe
n=90 with
follow-up
Moderate
n=139 with
follow-up
Mild
n=75 with
follow-up
Disease progression and treatment response over time
Schirmer et al. 2018 Cell Host and Microbe
Samples cluster by disease severity
Changes in microbial community
composition are associated with disease
severity
Fecal calprotectin
significantly increases
with disease severity
* * *
100
1000
10000
inactive mild moderate severe
Fe
cal
ca
lpro
tec
tin
* * *
Gut microbial composition changes with increased disease severity
Schirmer et al. 2018 Cell Host and Microbe
• Heatmap shows 30/91 most significant OTUs (1,212 samples, FDR < 10-11)
• 75% of severity-decreased OTUs (54) were from the order Clostridiales
• Increase in oral taxa associated with severe disease
Oral
bugs
Do
wn
in
se
ve
reU
p in
se
ve
re
inactiv
e
mild
modera
te
severe
Bifidobacterium (OTU 997439)
Blautia (OTU 1078587)
Lachnospiraceae (OTU 975306)
Collinsella (OTU 147071)
Clostridiales (OTU 509416)
Blautia obeum (OTU 528606)
Blautia (OTU 368950)
Blautia (OTU 531675)
Erysipelotrichaceae (OTU 592616)
Blautia producta (OTU 696563)
Ruminococcus (OTU 199543)
Eggerthella lenta (OTU 682726)
Ruminococcaceae (OTU 309720)
Blautia producta (OTU 589313)
Clostridiales (OTU 562376)
Bifidobacterium (OTU 471180)
Blautia (OTU 580087)
Lachnospiraceae (OTU 195115)
Ruminococcus (OTU 591635)
Bifidobacterium (OTU 484304)
Coriobacteriaceae (OTU 175508)
Neisseriaceae (OTU 904345)
Eikenella (OTU 1026498)
Aggregatibacter segnis (OTU 92231)
Veillonella parvula (OTU 518743)
Pasteurellaceae (OTU 562825)
Fusobacterium (OTU 444857)
Veillonella dispar (OTU 342427)
Haemophilus parainfluenzae (OTU 865469)
Enterobacteriaceae (OTU 1111294)
−4.5
−4
−3.5
−3
−2.5
−2
−1.5
−1
Microbial factors are associated with disease severity
Schirmer et al. 2018 Cell Host and Microbe
• Severe patients achieving week 52 remission started with high levels and displayed a
continuous decrease of H. parainfluenzae
• Mild patients that failed to achieve week 52 remission started with lower levels but
displayed a substantial increase by week 52
●
●●
●●
●
● ●
●
●
●
●
●●
●●●
● ●
●
●
●●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●●
●●
●
●
●
●
●
●●
●●
●
●
●
●
●●
●
●●● ●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
● ●
● ●
●
●
●
●
●
●
● ●●
●●
●
●
●
●
● ●●
●
●
●
●●
●
●
●
● ●
●
●●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●1e−05
1e−03
1e−01
0 4 12 52
Collection week
OT
U 8
65
469
●
●
●
●
●
●
inactive
mild
moderate
severe
No remission
Remission
Haemophilus parainfluenzae
Longitudinal trends in H. parainfluenzae
implicated in remission
Schirmer et al. 2018 Cell Host and Microbe
Human Microbiome Project 2: Goals and Objectives of this Study
Pregnancy & PretermBirth
InflammatoryBowels Disease
Prediabetes(Type 2 diabetes)
Ramnik Xavier and Curtis Huttenhower
132 subjects; sampled every 2 weeks; 1 year
The IBD Multi’omics DataBase
http://ibdmdb.org
Thank you!
Ramnik Xavier Curtis Huttenhower
Timothy Arthur
Nadine Fornelos
Eric Franzosa
Ashley Garner
Brantley Hall
Larson Hogstrom
Xiaofang Jiang
Jason Lloyd-Price
Tiffany Poon
Melanie Schirmer
Alexandra Sirota-Madi
Moran Yassour
Human Microbiome Project 2
Lita Procter
Jon Braun
Dermot McGovern
Subra Kugathasan
Ted Denson
Janet Jansson
Owen White
Bruce Birren
Chad Nusbaum
Clary Clish
Joe Petrosino
Thad Stappenbeck
ibdmdb.org
…and the labs!
Eeva Jason
Mikael Knip Lab
Georgia Lagoudas
Paul Blainey Lab
PROTECT
Subra Kugathasan
Lee Denson
Jeff Hyams
We are now hiring postdocscontact us if interested