Recent progress in our understanding of IBD genetics

Report
State of the Art: Recent Progress in Understanding
of IBD Genetics
Judy H. Cho, M.D.
Ward-Coleman Professor of Translational Genetics and Medicine, Icahn
School of Medicine at Mount Sinai
December 5, 2014
Concepts
Rare vs. Common genetic variants
Rare variation
 Age of onset & population considerations
 When to sequence & sequence annotation
Common genetic variants & GWAS studies
 Population differences
 Non-coding functional variation
Rare vs. common human polymorphisms
Rare
Common
1. Often population-specific
1. Present across populations
2. Most missense alleles are rare
2. Large majority are non-coding
3. Selection: negative selection
3. Selection: balancing , positive
4. Association signals: greater
effect sizes (e.g. OR > 1.5) for
nominally associated variants
4. GWAS-identified significant loci
have modest effects (i.e., odds
ratios)—most less than 1.1
5. Less power—most samples
are homozygous wild-type
5. Greater power to detect
association
Monogenic Early-onset IBD
 Mendelian IBD: IL10
deficiency--neonatal
 Immunodeficiencies with
variable age of onset
 FOXP3: IPEX
 WAS: Wiscott-Aldrich
 LRBA: LPS-responsive, CVID8
 Chronic granulomatous disease
 XIAP: X-linked inhibitor of
apoptosis
10 years old
Uhlig et al., Gastroenterology 2014
XIAP, NOD2 and Crohn’s disease
 XIAP: X-linked inhibitor of apoptosis
 Many XIAP missense mutations—extremely rare
 XLP2: X-linked lymphoproliferative immunodeficiency
 CD-unique private mutations: (potentially) as high as 4% of
pediatric male IBD cases
 Ileocolonic plus perianal disease
 NOD2-signaling deficiency
 Variable disease course—majority responsive to
traditional therapy
 Numbers small, but XIAP-deficiency appears to have a
higher penetrance than NOD2-deficiency
Zeissig et al, Gut 2014
When to consider sequencing
Research? Clinical?
 Eventually all of us!!!
 Early onset IBD—especially males
 Consanguineous
 Immunodeficiency plus IBD
ATG16L1 & sequence annotation
 ATG16L1 T300A variant: alanine CD risk
 Caspase-3 & -7 cleavage sequence: -D-X-X-D-(T/A)Alanine variant  increased ATG16L1 degradation
 Clearly establishes that CD risk allele is correlated
with impaired autophagy
 Sequence annotation: prediction of new/altered
covalent modifications
Murthy et al., Nature 2014
Lassen et al., PNAS 2014
Rare variants are “less rare” in AJs
AJ
Flemish
 128 whole genome
sequenced in AJs
 Rare variants are
“less rare” in AJs
compared to Flemish
 Profound population
bottleneck in AJs
Derived allele frequency
Carmi et al., Nature Commun 2014
AJ-specific frameshift mutation in CSF2RB
associated to IBD
Gene
Function
IBD
Freq
CSF2RB
GMCSF
receptor
0.032
control
odds
P-value
type group
freq
ratios
AJ0.02 0.00104 1.6 IBD
specific
CSF2RB: common subunit to GMCSF, IL3, IL5
Truncation mutation: ~200 aa in cytoplasmic tail
Common variants in this region nominally
associated—also NCF4 (Rioux et al., Nature Genetics 2007)
GMCSF & IBD
 Anti-GMCSF antibodies
 CD patients with elevated anti-GMCSF Ab
 NOD2-deficient mice + neutralizing anti-GMCSF +
piroxicam ileitis
 Anti-GMCSF Ab levels correlate with increased disease
complications
 GMCSF cytokine
 Induces IL10, retinoic acid
 Potential therapeutic benefit
Han et al., Gastroenterology 2009, 1261
Gathungu et al., IBD 2013, 1671
Mortha et al., Science 2014
Korzenik et al., NEJM 2005
Concepts
Rare vs. Common genetic variants
Rare variation
 Age of onset & population considerations
 When to sequence & sequence annotation
Common genetic variants & GWAS studies
 Population differences
 Non-coding functional variation
Distinct genetic architecture for dominant loci:
European ancestry vs. Japanese Crohn’s disease GWAS
European ancestry
Arg381Gln
Japanese
MHC
TNFSF15
Yamazaki, et al,
Gastroenterology
2013; 781
GWAS/common alleles: vast majority demonstrate
similar direction & effect sizes across populations
Exceptions—biggest effect loci
European ancestry
Asian
Alleles
NOD2
R702W, G908R, fs1007
uncommon risk
monomorphic
IL23R
R381Q
uncommon protective monomorphic
non-coding IL23R
associated
associated
different alleles
ATG16L1
T300A
Risk is LOF
Smaller/no effect
TNFSF15
non-coding TNFSF15
Smaller effects
Risk is GOF
same alleles
Yang et al., IBD 2009
Kim et al., Gut 2014
Hedl et al., PNAS 2014
Expression quantitative trait loci (eQTL)
mapping
 mRNA expression as a continous trait
 Heritable
 Mappable to specific SNPs
 Cell lines, tissues and contextspecificity
 Presently defined eQTLs likely only a
subset of genuine eQTLs
 LPS- & IFNg stimulated monocytes
define more eQTLs
 80% of transcripts with eQTLs
Morley, et al., Nature 2004; 430: 743
Dixon, et al., Nature Genetics 3007; 39: 1202
Fairfax et al., Science 2014
 Diseases examined by GWAS
 21 different autoimmune diseases
 Comparative diseases
 Metabolic traits
 Neurologic diseases
 Most correlated: UC-CD-Ankylosing
spondylitis-psoriasis-Behcet’s disease
Navy blue: no correlation
Maroon red: highest correlation
Farh et al., Nature 2014
Fine-mapping in autoimmunity
 Fine-mapped autoimmune loci
 90% are non-coding
 60% map to immune enhancers
 Histone marks: greatest enrichment seen for
H3K27ac—active/stimulated enhancers
 Disease-associated SNPs in enhancers are near, but not
within consensus transcription factor binding sites
Farh et al., Nature 2014
Enrichment of GWAS signals in cell
type specific enhancers (H3K27ac)
T cells
B cells
Colonic
mucosa
 T cell enrichment: RA,
MS, IBD, celiac disease
 B cell enrichment: RA>
IBD
 Monocyte signal: largely
absent except CRP
 Tissue-macrophage
specific signals?
 Marked enrichment of
colonic mucosa in UC
RA
CD-UC
Y-axis: 33 cell type H3K27ac
X-axis: 39 diseases
Farh et al., Nature 2014
GWAS and eQTL SNPs show (apparently)
different molecular distributions
 eQTL dataset: peripheral blood RNA from twins
 BUT: eQTL distribution for stimulated (LPS, IFNg)
gene expression is likely quite different
 Value of eQTLs-risk allele alignment: implies
directionality—gain/loss of function
Wright et al., Nature Genetics 2014
Farh et al., Nature 2014
Conclusions
 Sequencing and rare variant-associated IBD
 Younger, severe, select populations
 Establishing genome-wide significance is
challenging—need for early functional analysis
 Common variant fine-mapping
 “Single” causal allele identifiable statistically in only a
minority of cases
 Greatest enrichment of GWAS signals in H3K27ac
active enhancer regions—near, but not in consensus
transcription factor binding sites
 Imperfect molecular matching of GWAS and eQTLs:
need to define cell and activation state specificities
 Power of genetics: earliest pathophysiologic
mechanisms, precise therapeutic targeting
Acknowledgements
 Cho lab
 Ken Hui

NIDDK IBD Genetics Consortium
 Steve Brant
 Monica Bowen
 Richard Duerr
 Kyle Gettler
 Dermot McGovern
 Nicole Villaverde
 John Rioux
 Nai-Yun Hsu
 Mark Silverberg
 Mark Daly
 Felix Chuang
 Phil Schumm
 Yashoda Sharma
 Mount Sinai collaborators

 Clara Abraham
 Inga Peter
 Eric Schadt
 Miriam Merad
 Bruce Sands
 Jean-Fred Colombel
Yale University
 Richard Flavell

The Ashkenazi Jewish Consortium
 Itsik Pe’er
 Todd Lencz

similar documents