Transcript Document

Role of pathogen-driven selection in shaping
the predisposition to IBD:
identification of disease susceptibility alleles
Mario (Mago) Clerici, M.D.
Chair of Immunology
Head, PhD School in Molecular Medicine, Universita' di Milano
Scientific Director, IRCCS SM Nascente,
Fondazione Don C Gnocchi, Milano
Bardolino, January 16th, 2013
The evolutionary perspective for
Inflammatory Diseases
 Inflammatory/autoimmune diseases can have early onset (i.e.
before reproductive age)
 Inflammatory/autoimmune diseases have a strong genetic
component
 Inflammatory/autoimmune diseases have a relatively high
prevalence in human populations
Why has evolution failed to eliminate the risk alleles?
1) Susceptibility alleles increased in frequency by genetic drift
2) Susceptibility alleles increased in frequency as a result of natural
selection (they confer a selective advantage to the carriers; e.g.
protection from infection) [hygiene hypothesis]
3) Risk alleles were neutral under different environmental conditions
(e.g. high prevalence of infections/worms) [hygiene hypothesis]
Aims
Application of population genetic approaches to study
the evolutionary history of inflammatory disease risk
alleles in human populations
Study the role of past infections in shaping the presentday distribution of inflammatory disease risk alleles
Use evolutionary information to identify novel risk
variants for Crohn’s disease
An innovative approach
We developed a strategy to detect pathogen-driven selection
Pathogen-driven selection implies that allele frequencies at a
locus are shaped by selective pressure imposed by one or more
infectious agents
Strengths:
1) We test a specific hypothesis on the underlying selective
pressure (can distinguish among different pathogen groups)
2) High power to detect selection on standing variation
Weaknesses:
1) Use of a relatively low-density SNP panel
2) Use of a population panel with uneven geographic
representation
Pathogen-driven selection
Identifies correlations between genetic variability and
pathogen-driven selective pressure.
selected variant
Allele frequency
Allele frequency
neutral variant
pathogen-driven selective pressure
pathogen-driven selective pressure
We need a measure of selective-pressure that reflects historical
pressures (evolution acts over long time periods).
Pathogen diversity can be used as a measure of the selective pressure exerted
by infectious agents on human populations.
Pathogen diversity more closely reflects historical pressures than other estimates
such as the prevalence of specific infections
HLAA
HLAB
OUR APPROACH
Over 650 000 SNPs genotyped in 52 populations
(HGDP-CEPH panel).
We calculate Kendall's rank correlation
coefficient (tau) between allele frequencies in
HGDP-CEPH populations and pathogen
diversity.
Allele frequency
Pathogen diversity: number of different
pathogen species/families present in different
geographical areas of the world from the
Gideon database.
A SNP was considered to be significantly
associated with pathogen diversity if it
displayed a significant correlation and a rank
higher than 0.99
frequency
pathogen diversity
99th percentile
tau
Pathogen diversity:
Micro-pathogens: viruses, bacteria, fungi and protozoa
Macro-pathogens: insects, arthropods and helminths
 Among variants
subjected to pathogen
driven selection we
identified an IBDassociated SNP located
in IL18RAP.
The risk allele for IBD
correlates significantly
with pathogen richness
rs917997
Six out of 9 risk variants for CeD or IDB/Crohn's disease
(CD) were associated with micro- and macro-pathogen
richness
A specific measure
Quantifying selection
Estimate pathogen-driven selection (virus, protozoa, helminth, bacteria) for
single SNPs in the HGDP-CEPH panel
Retrieve all GWAS SNPs associated to any trait or disease from the NHGRI
Catalog of Published Genome-Wide Association Studies
Collapse SNPs in tight LD (r2 >0.8) into single loci and retain only variants
genotyped in the HGDP-CEPH panel (n=2773)
Total SNPs for CD: 43, UC: 42
Count SNPs that significantly correlate with the diversity of each pathogen
group (expected 5%; observed 18%)
Apply a re-sampling approach on the 2773 GWAS SNPs to assess significance
and calculate the empirical probability (on MAF-matched variants)
GWAS SNPs for CD, UC and CeD that correlate with
the diversity of different pathogens
Exploiting selection signatures to identify novel risk
variants that are not picked up by GWAS
Extract all SNPs with 0.05<p value <5x10-5 from CD meta-analysis (Bartett, 2008)
Identify those selected by protozoa
Rank them based on p value
Select genic SNPs
Discard SNPs close (less than 2 Mb) to previously associated CD loci
Analyse the top 5 SNPs [rs2364403 (ARHGEF2), rs3782567 (HEBP1), rs9636320
(ARID3B), rs199533 (NSF), rs1011312 (TPST2)] in an Italian cohort
Combine results with the partially independent 6-study meta-analysis
(Franke, 2010)
Association analysis
ARHGEF2: a central component of pathogen recognition by NOD1
NSF: involved in autophagy
HEBP1: promotes calcium mobilization and chemotaxis in monocytes and
dendritic cells
Interaction (Ingenuity model) network
known CD susceptibility
gene
Increased
activation in
mucosa of CD
patients;
pharamacological
inhibition of RhoA
patway reduces
colonic
inflammation in
rats
Disregulated in CD and IBD
mouse models; increased in
IBD-associated neoplastic
transformation;
underexpressed in Treg;
regulator of FOXP3 expression
Conclusions
Adaptation to pathogen exposure results in the selection for
alleles that confer increased protection against infections, but
predispose to CD
This information can be exploited to identify novel risk
variants for CD
These data suggest that infections (e.g. T. gondii) might
interact with genotype to determine CD susceptibility
These observations help building an evolutionary framework
for the development of novel treatment strategies
Lab 1
Chair of Immunology
University of Milano
MANUELA SIRONI
Rachele Cagliani
Uberto Pozzoli
Diego Forni
Stefania Riva
Matteo Fumagalli
Lab Lab 2
Don C Gnocchi Foundation
Milano