Accepted Article
PROFESSOR JOCELYNE JUST (Orcid ID : 0000-0002-5646-2429)
DR MARIE-HÉLÈNE DIZIER (Orcid ID : 0000-0001-8460-7667)
Article type
: Original Article-Epidemiology of Allergic Disease
The COL5A3 and MMP9 genes interact in eczema susceptibility
Patricia Margaritte-Jeannin, PhD1,2*, Marie-Claude Babron, PhD1,2*, Catherine Laprise, PhD3,
Nolwenn Lavielle, MSc1,2, Chloé Sarnowski, PhD1,2, Myriam Brossard, PhD1,2, Miriam Moffatt, PhD4,
Valérie Gagné-Ouellet, MSc3, Adrien Etcheto, MSc5, Mark Lathrop, PhD6, Jocelyne Just, MD7,
William O. Cookson, MD4, Emmanuelle Bouzigon, MD, PhD1,2, Florence Demenais, MD,2, MarieHélène Dizier, PhD1,2
*: equal contribution
1. Inserm, UMR-946, Genetic Variation and Human Diseases unit, F-75010, Paris, France;
2. Univ Paris Diderot, Sorbonne Paris Cité, Institut Universitaire d’Hématologie, F-75010, Paris,
France;
3. Université du Québec, Chicoutimi, Canada ;
4. National Heart Lung Institute, Imperial College, London, UK,
5. Paris Descartes University, Rheumatology Department, Cochin Hospital, AP-HP; INSERM U1153,
Sorbonne Paris-Cité, Paris, France.
6. Mc Gill University and Genome Quebec's Innovation Centre, Montréal, Canada;
7. Service d’Allergologie Pédiatrique, Centre de l’Asthme et des Allergies. Hôpital d'Enfants ArmandTrousseau - UPMC Paris 06, F-75012, Paris, France ;
Corresponding author
Marie-Hélène DIZIER
UMR946, INSERM / Université Paris-Diderot
27 rue Juliette Dodu
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/cea.13064
This article is protected by copyright. All rights reserved.
Accepted Article
F-75010 Paris, France
Phone +33 1 72 63 93 25
Fax +33 1 72 63 93 49
Marie-Helene.Dizier@inserm.fr
Running Head:
Interaction of COL5A3 and MMP9 genes in eczema
Abstract
Background: Genetic studies of eczema have identified many genes, which explain only 14% of the
heritability. Missing heritability may be partly due to ignored Gene–Gene (G-G) interactions.
Objective: Our aim was to detect new interacting genes involved in eczema
Methods: The search for G-G interaction in eczema was conducted using a two-step approach, which
included as a first step, a biological selection of genes, which are involved either in the skin or
epidermis development or in the collagen metabolism, and as a second-step, an interaction analysis of
the selected genes. Analyses were carried out at both SNP and gene levels in three asthma-ascertained
family samples: the discovery dataset of 388 EGEA (Epidemiological study on the Genetics and
Environment of Asthma) families and the two replication datasets of 253 SLSJ (Saguenay-Lac-SaintJean) families and 207 MRCA (Medical Research Council) families.
Results: One pair of SNPs, rs2287807 in COL5A3 and rs17576 in MMP9, that were detected in EGEA
at P ≤ 10-5 showed significant interaction by meta-analysis of EGEA, SLSJ and MRCA samples
(P=1.1x10-8 under the significant threshold of 10-7). Gene-based analysis confirmed strong interaction
between COL5A3 and MMP9 (P=4x10-8 under the significant threshold of 4x10-6) by meta-analysis of
the three datasets. When stratifying the data on asthma, this interaction remained in both groups of
asthmatic and non-asthmatic subjects.
This article is protected by copyright. All rights reserved.
Accepted Article
Conclusion: This study identified significant interaction between two new genes, COL5A3 and
MMP9, which may be accounted for by a degradation of COL5A3 by MMP9 influencing eczema
susceptibility.
Further confirmation of this interaction as well as functional studies are needed to better understand
the role of these genes in eczema.
Key words (5 words in alphabetical order)
Asthma, Eczema, Gene-Gene interaction, Logistic Regression, meta-analysis
Introduction
Atopic dermatitis (AD), or eczema, is the most common chronic inflammatory skin disorder. Eczema
is also an allergic disease associated with asthma and allergic rhinitis. These co-morbidities may share
genetic determinants as suggested by their strong associations at both the individual and family levels
[1]. However, genes specific to eczema also exist, as first suggested by the observation that parental
history of eczema represents a more potent risk factor for eczema than parental history of asthma or
allergic rhinitis [1, 2].
A number of susceptibility genes for eczema has been found through genetic studies [3] based on
candidate genes, positional cloning approaches and GWAS. Many of these genes are involved in the
immune response, and are not specifically associated to eczema, but more generally to allergic
diseases including asthma and allergic rhinitis. Among the genes more specifically associated to
eczema are genes involved in the skin or epidermis development including SPRR3 [4], ACTL9 [5],
OVOL1 [5] and also FLG [6], SPINK5 [7], C11orf30 [8] which are related to epidermal barrier
dysfunction and genes involved in collagen metabolism like COL29A1 [9], IL6 [10] and IL6R [11].
However, all the genes associated with eczema explain only a small part of the heritability (1415%)[12, 13]. Mechanisms such as Gene-Gene (G-G) or Gene-Environment interactions may account
for this missing heritability. Indeed, most previous genetic studies on eczema focused on single marker
analysis. Five small scale G-G interaction studies were conducted for eczema, which involved only on
a very small number of genes [14-18].
This article is protected by copyright. All rights reserved.
Accepted Article
As it has been shown that ignoring interaction may hinder the detection of genetic factors [19], several
methods have been developed to identify G-G interactions in association studies. They can be applied
to the whole genome or to part of the genome, such as a linkage region. To alleviate the burden of
multiple testing, which already exists in single-marker analysis but is even more crucial in two-marker
analysis, two-step strategies have been proposed. They typically include a first step of marker
selection, followed by a second step of two-marker tests applied to the selected markers [20]. The
selection of SNPs at the first step can be based on single-marker association results or on biological
information [21]. We showed that, in case of strong interaction with a small marginal effect of each
marker, a first-step selection based on association tests was not powerful, as the markers would not be
selected at the first step [22]. Moreover, we can assume that genetic variants that do not have a too
small marginal effect would have already been detected by single marker association tests, through the
numerous GWAS conducted for eczema using large sample sizes [5, 8, 11, 23-26] which recently
reached more than 115,000 subjects [13].
The aim of the present study was to detect new interacting genes in eczema. To search for G-G
interaction, we applied a two-step approach, with a first step of gene selection based on biological
knowledge related to eczema. We selected genes which were, similarly to the genes previously found
specifically associated to eczema, involved either in the skin or epidermis development or in the
collagen metabolism. In a second step, we performed cross-gene SNP-SNP interaction analyses
followed by gene-based interaction analyses. These analyses were conducted in asthma-ascertained
family samples from three studies: a discovery dataset, the Epidemiological study on the Genetics and
Environment of Asthma (EGEA), and two independent replication datasets, the Saguenay-Lac-SaintJean (SLSJ) and the Medical Research Council (MRCA) studies.
This article is protected by copyright. All rights reserved.
Accepted Article
Material and method
Discovery sample
The EGEA study has been described in detail previously [27]. The EGEA family sample consisted of
388 French nuclear families, 253 families ascertained through offspring with asthma (one proband in
90% of families and two probands in the remainder) and 135 families ascertained through one parent
with asthma. Eczema status was based on the answer to the question: "Have you had eczema during
childhood?" (either at the time of data collection and/or at the 10-year follow-up of the EGEA study).
The same definition has already been used in previous linkage and association analyses of eczema
conducted in EGEA [28, 29]. Ethical approval was obtained from the relevant institutional review
board committees (Cochin Port-Royal Hospital and Necker-Enfants Malades Hospital, Paris). Written
informed consent was signed by all participants or by kin or guardians of the minors/children.
Replication samples
The Saguenay–Lac-Saint-Jean Familial Asthma Collection (SLSJ) comprised 253 French-Canadian
multigenerational families ascertained through two probands with asthma [30]. Eczema was defined
by a self-reported history of eczema on a questionnaire. Positive answers were validated by physicians
or through clinical records. The SLSJ local ethics committee approved the study, and all
subjects gave written informed consent.
The Medical Research Council (MRC) UK National family collection included 207 nuclear families,
recruited through at least one proband with asthma (MRCA sample)[31]. Eczema definition was based
on answers to questions on the presence of symptoms. The MRCA local ethics committee
approved the study, and all subjects gave written informed consent.
This article is protected by copyright. All rights reserved.
Accepted Article
Genetic data
The EGEA and the SLSJ samples were genotyped using the Illumina 610-Quad , as part of the
European Gabriel consortium asthma GWAS [32] while the MRCA sample was genotyped using the
smaller Illumina Sentrix HumanHap300 BeadChip [31]. To get a number of SNPs in the MRCA
dataset as large as in the EGEA and SLSJ datasets, SNP imputation was performed using MACH
v1.00 software (http://www.sph.umich.edu/csg/abecasis/MACH/) and HapMap2 CEU haplotypes as
reference panel. Imputed SNPs with imputation quality score greater than or equal to 0.9 were
retained. Stringent quality criteria (QC) [32] were used to select both individuals and SNPs.
After genotyping, an ancestry analysis was carried out in the three European descent samples (EGEA,
SLSJ and MRCA) [32] and putative non-European samples were eliminated from subsequent
analyses. Informative principal components for within-Europe diversity were included as covariates
in analyses.
Statistical analysis
Two step strategy in the EGEA discovery sample
The first step consisted in the selection of genes based on biological knowledge contained in the Gene
Ontology (GO) database (http://amigo.geneontology.org/). From the GO biological process categories,
we defined two groups. The first group included GO categories containing the term "collagen" i.e;
"collagen metabolic process" (GO:0032963), "collagen-activated signaling pathway" (GO:0038065),
"regulation of collagen binding" (GO:0033341). The second group included GO categories containing
the terms "epidermis" or "skin" i.e. "skin development” (GO:0043588) and "epidermis development”
(GO:0008544) after exclusion of sub-categories related to hair follicle, limb and nipple.
After
selection of the genes belonging to each group, all SNPs lying within the gene boundaries (Build 37.3)
and present in our three datasets were retained for analysis. The first group, called the "collagen"
group, comprised 121 genes (97 genes after excluding those having no SNP genotyped and 1,942
SNPs after QC checking), while the second group, the "epidermis and skin" group, totaled 251 genes
(176 genes after excluding those having no SNP genotyped and 2,189 SNPs after QC checking)
This article is protected by copyright. All rights reserved.
Accepted Article
(Supplementary Table I). Only SNPs located within the genes were considered to limit the number of
tests.
The second step consisted in conducting cross-gene SNP-SNP interaction analyses within each gene
group, i.e. between SNPs located in two different genes belonging to the same group. We performed
logistic regression using Stata® V14.1 (Stata Corporation, College Station, Texas) assuming an
additive model for SNP main effect and interaction and taking familial dependencies into account
through the cluster and robust options of the logit function. We used the same coding scheme as
usually proposed for SNP-SNP interaction modelling [21]. We modelled the additive effect of a
SNP by coding the genotypes of homozygotes for the minor allele, heterozygotes and
homozygotes for the major allele as 1, 0, and -1; the interaction term between two SNPs was
obtained by multiplication of these genotypic values for the two SNPs. Test of interaction was
performed by comparing the full model with SNP main effect and interaction to the restricted model
without interaction, using a likelihood-ratio test which follows a Chi-square distribution with one
degree of freedom. SNP pairs in which the expected number of joint genotypes in cases or in controls
was less than 5 were excluded. The total number of tested SNP pairs was 1.2x106.
To take into account the dependence between tests due to the Linkage Disequilibrium (LD) between
SNPs within the same gene, we calculated the effective number of independent SNP pairs. We first
calculated the LD matrix of each gene, i.e. the correlation between the SNPs of that gene. Then, as
proposed by Emily [33], the correlation matrix of a gene pair is set to the Kronecker product of the LD
matrices of the two genes. It is then possible to estimate the number of independent tests, Keff, for a
given gene pair, using the procedure of Moskvina and Schmidt [34]. Summing over all pairs of genes
provides a total Keff equal to 4.6 x105, thus giving a significance threshold of 10-7 after Bonferroni
correction for multiple testing.
Pairwise interactions at the gene level within each group of biological categories were also tested
using the AGGrEGATOr method [33]. Briefly, this method aggregates the SNP-SNP interaction Pvalues to estimate a gene-based interaction P-value, based on the minimum-P-value, accounting for
This article is protected by copyright. All rights reserved.
Accepted Article
the number of SNP pairs and dependency between these SNP pairs as proposed by Connely and
Boenhnke [35] . As 14,682 pairs of genes were tested, the Bonferroni significance threshold was equal
to 3.4x10-6.
Replication in SLSJ and MRCA
The SNP pairs exhibiting P SNP-SNP ≤ 10-5 in EGEA were tested for replication in SLSJ and MRCA. For
the MRCA imputed SNPs, logistic regression was carried out with allele dosage to take into account
the uncertainty of imputation. The threshold of 10-5 was chosen to obtain both a strong indication of
interaction and a reasonable number of SNP pairs (here 7, see results) to be tested in the replication
datasets. After Bonferroni correction for this number of SNP pairs), the threshold for significant
replication was thus equal to 7x10-3.
The gene-based interaction test was also applied to each replication sample for all pairs of genes with
PG-G ≤10-4 in EGEA. Three gene pairs were retained, leading to a Bonferroni threshold for significance
of 0.017.
We then conducted a fixed-effects meta-analysis of SNP-SNP interactions in EGEA, SLSJ and MRCA
for the SNP pairs detected in EGEA at P SNP-SNP ≤ 10-5 and for the SNP pairs belonging to the gene
pairs detected in EGEA at P G-G ≤ 10-4. Gene x gene analyses were performed using AGGrEGATOr
with P values of meta-analysis of SNPxSNP interaction test (Pmeta-SNP-SNP) in EGEA, SLSJ and MRCA.
The thresholds to declare significance were identical to those indicated above for analyses in the
discovery sample (10-7 for SNP-SNP interactions and 3.4x10-6 for G-G interactions respectively).
Finally, to investigate whether the evidence for interaction was related to the presence of asthma in
these asthma-ascertained samples, we performed a sensitivity analysis according to asthma status. In
this sake, the SNP-SNP interaction tests were performed in subjects with and without asthma
separately in the pooled EGEA, SLSJ and MRCA samples to have a sufficient sample size in each
group of subjects. Analyses were conducted in the pooled samples while adjusting for study and for
principal components within study. Homogeneity of interaction between the two groups was tested.
Gene-based interaction analyses were also repeated in subjects with and without asthma.
This article is protected by copyright. All rights reserved.
Accepted Article
Results
Descriptive statistics
Table I shows the number of genotyped subjects and their phenotypic distribution according to eczema
status in the EGEA, SLSJ and MRCA samples. The total number of subjects was 1,493 in EGEA,
1,153 in SLSJ and 254 in MRCA. The proportion of subjects with eczema was similar in the three
samples, ranging between 27 and 31%. The proportion of males was higher in subjects with eczema
than without eczema in MRCA, while it was smaller in SLSJ and similar in EGEA. The reason may be
that in MRCA, all subjects are children and most subjects with eczema are also asthmatic (89%) and it
is known that the risk of asthma in children is higher among boys than among girls. The mean age was
lower in subjects with eczema than in those without eczema in EGEA and SLSJ while it was similar in
these two groups in MRCA. The variation of age across studies, with younger subjects in MRCA and
older subjects in SLSJ, is at least partly due to the ascertainment of the samples, only children in
MRCA and the inclusion of fifty percent of three-generation pedigrees in SLSJ. As expected, the
proportion of subjects with asthma and atopy (defined by a positive skin prick test response to at least
one aeroallergen) was much higher among individuals having eczema than among unaffected subjects.
Note that in MRCA, whichever the eczema affection status, the proportion of asthmatics was higher
than in the other two samples, while the proportion of atopics was smaller. Analysis of offspring only
in MRCA and of subjects belonging to parental and offspring generation in the other two samples
might partly explain the higher proportion of asthmatics in MRCA families ascertained through
asthmatic offspring. The smaller proportion of atopics in MRCA may be due to the smaller number of
aeroallergens tested in MRCA (five allergens) than in the other two samples (ten or more allergens)
(see table I).
SNP-SNP and G-G interaction analyses
EGEA discovery dataset
In the SNP-SNP interaction analysis, seven SNP pairs, three within genes of the "collagen" group and
four within genes of the "epidermis and skin" group, had PSNP-SNP ≤ 10-5 in EGEA (Table II and more
details, supplementary Table II). No pair reached the significance threshold of 10-7. In the gene-based
This article is protected by copyright. All rights reserved.
Accepted Article
interaction analysis, three gene pairs (COL5A3 and MMP9, COL1A2 and COL6A1, CTSB and
COL5A3) had P
G-G
≤ 10-4 (Table III). None reached the significance threshold of 3.4 x10-6. As
expected, these three pairs of genes included the SNP pairs with the smallest P SNP-SNP value.
SLSJ and MRCA replication datasets
Only one SNP pair (rs2287807 in COL5A3 and rs17576 in MMP9) among the pairs detected in EGEA
was significantly replicated in SLSJ (PSNP-SNP =1.24x10-3, under the replication threshold of 7x10-3) but
not in MRCA (PSNP-SNP =0.13). However, meta-analysis of the EGEA, SLSJ and MRCA results for
this SNP pair showed strong evidence for interaction (Pmeta-SNP-SNP =1.11x10-8 , below the significance
threshold of 10-7) and strong improvement with respect to the interaction signal obtained in the EGEA
discovery dataset (PSNP-SNP =6.70x10-6) (Table II). This SNP pair was also the only one showing the
same direction of interaction term in the three datasets. This SNP pair showed a pattern of interaction
in which OR associated CC (or TT) genotype at rs2287807 (COL5A3) showed an opposite effect
depending on the genotype AA (or GG) at rs17576 (MMP9) (Figure 1 and supplementary Table III).
More precisely, for subjects homozygous for allele A at rs17576, T allele at rs2287807 will increase
the risk of eczema while C allele will be protective. For the other subjects, heterozygous or
homozygous for allele G, the effect of the alleles at rs2287807 will be inversed: the T allele will be
protective and the C allele at risk.
In the gene-based interaction analysis, only one pair of genes (COL5A3 and MMP9) among the gene
pairs detected in EGEA was significantly replicated in SLSJ (PG-G =8.79x10-3, below the replication
threshold of 1.5x10-2) but not in MRCA (PG-G =0.52). Moreover, this pair was also highly significant
by the meta-analysis of the three datasets (Pmeta-G-G = 4.04x10-8, well under the significance threshold
of 3.4x10-6). The other two gene pairs did not reach significance: Pmeta-G-G =7.18x10-2 for COL1A2 and
COL6A1 pair, and Pmeta-G-G =9.86x10-3 for CTSB and COL5A3 pair (Table III).
Finally, after stratification of the pooled sample according to asthma status, the SNP pair (rs2287807,
rs17576) showed similar evidence for interaction in 1423 asthmatic (P SNP-SNP =7.54x10-5) and in 1477
non-asthmatic subjects (PSNP-SNP =1.55x10-5). There was no heterogeneity between the two groups
This article is protected by copyright. All rights reserved.
Accepted Article
(Phomogeneity=0.16), clearly indicating this interaction is not related to the presence of asthma. Likewise,
in the gene-based interaction analyses, the COL5A3-MMP9 gene pair showed similar results in
asthmatic (P G-G =3.09x10-4) and non-asthmatic subjects (P G-G =6.28x10-5).
Discussion
This study identified the interactive effect of genetic variants in the COL5A3 and MMP9 genes on
eczema risk in families ascertained through asthmatic subjects. Evidence for this interaction rests on
the results obtained in the EGEA discovery sample and on the highly significant results obtained when
the EGEA dataset was combined with the two SLSJ and MRCA replication datasets. This result is
further strengthened by the fact that interaction was significant in both SNP-SNP and gene-based
interaction analyses. Moreover, the interaction between these two genes in eczema susceptibility
appears independently whichever the subject asthma status.
The EGEA results were significantly replicated in the French Canadian SLSJ dataset both at the SNP
and gene level, but not in the MRCA sample. This might be partly explained by the shared French
genetic background of EGEA and SLSJ families, but also by other phenotypic characteristics, shared
by EGEA and SLSJ, but different in MRCA. In particular, age at onset is known to be a factor of
genetic heterogeneity. Here, mean age was similar in EGA and SLSJ but much younger in MRCA,
indirectly suggesting earlier onset of eczema and asthma in MRCA. The proportion of asthmatic and
atopic subjects was similar in EGEA and SLSJ but different in MRCA. Other phenotypic differences
not measured here such as severity of eczema or asthma could also have influenced the results.
Another explanation might be the reduced power in MRCA because of smaller sample size.
Nevertheless, meta-analysis of EGEA, SLSJ and MRCA led to improvement of the SNP-SNP
interaction signal (P =1.11x10-8) compared to the one obtained in the meta-analysis of EGEA and
SLSJ (P =2.55x10-8).
Our study illustrates the advantage of SNP-SNP interaction analysis since the significant pair of SNPs
was not detectable by single SNP association test (P = 0.81 for rs2287807 and 0.03 for rs17576 in
EGEA). Similarly, COL5A3 and MMP9 would not have been detected by gene-based association
analysis (P=0.12 for COL5A3 and 0.03 for MMP9 in EGEA).
This article is protected by copyright. All rights reserved.
Accepted Article
Our findings also show the interest of a two-step strategy with selection of genes based on biological
knowledge, prior to performing interaction analysis. Indeed, if the first step had been based on singleSNP association analysis, neither the genes nor the SNP pairs showing significant interaction would
have been retained, because of the weak association signals observed in the discovery sample.
Likewise, the SNP pair would not have been detected by a one-step interaction strategy as the
correction for multiple testing would have been too stringent (threshold for significance estimated to
10-13 for a genome-wide interaction study of a 500K chip [36]).
We did not consider interaction across different GO classes because the hypothesis of a statistical
interaction between genes belonging to the same biological process is more plausible. Similarly, we
did not extend our search for interaction to other groups of genes, such as those involved in allergy
and immunity, which have already been extensively investigated in atopy and related phenotypes such
as asthma and eczema. Tested groups of genes included genes involved in regulatory T-cell function
[37], toll-like receptor (TLR)-related pathway [38] and TH2 differentiation pathway (IL13, IL4, IL33,
IL1RL1 and STAT6) [15, 39, 40]. Here, we focused on genes specifically related to eczema, a field
poorly investigated until now, in order to reduce the multiple testing burden and to gain power. This
strategy allowed detecting a significant interaction between the COL5A3 and MMP9 genes.
Note that the interaction detected here between two SNPs each with a very small or no marginal effect,
showed an inversion of the effect of genotypes at one variant depending on genotypes at the other
variant. Such situations correspond to flip-flop models. The biological plausibility of flip-flop models
may be discussed, however several human diseases, including Alzheimer’s disease [41], breast cancer
[42], melanoma [43] have been reported to be influenced by interaction of loci with no significant
main effect at either locus and showing a flip-flop type of interaction.
The present study is the first one to investigate G-G interaction in the genetic susceptibility to eczema
at a large-scale level.
Neither MMP9 nor COL5A3 were detected in the three previous G-G
interaction studies [14-18]. We also verified in the GWAS-Catalog of Published Genome-Wide
This article is protected by copyright. All rights reserved.
Accepted Article
Association Studies (http://www.ebi.ac.uk/gwas/), that no significant association of SNPs in COL5A3
nor in MMP9 had been reported with eczema and asthma and any other allergy related phenotypes,
such IgE sensitization, IgE levels. However, MMP9 has been shown to be associated with either atopic
[44] or non-atopic asthma [45] in candidate gene studies. To our knowledge, no association study of
this gene with eczema has been conducted. However, it is of note that MMP9 was shown to be overexpressed in the skin of patients with eczema[46] and differently expressed in lung tissue between pre
and post corticosteroids[47]. It was also found that allergen exposure induces the release of MMP9
into airway of allergic subjects [48].
SNP rs2287807 is located in intron 27 of the COL5A3 gene. Examination of functional annotations
using the HaploReg tool (http://compbio.mit.edu/HaploReg) shows that this SNP maps to promoter
and enhancer histone marks in skin (fibroblast, melanocyte and keratinocyte cells) and is also DNAse
hypersensitive in skin cells. SNP rs17576 is located in exon 6 of the MMP9 gene and corresponds to a
missense substitution (Gln>Arg) at position 279 of the protein. This SNP also maps to promoter and
enhancer histone marks in skin (fibroblast and keratinocyte cells).
The COL5A3 gene, located in 19p13.2 region, encodes an alpha chain for one of the low abundance
fibrillar collagens. Type V collagen is found in tissues containing type I collagen, the collagen the
most present in skin, and appears to regulate the assembly of heterotypic fibers composed of both type
I and type V collagen. This gene belongs to both "collagen" and "epidermis and skin" groups defined
here, and more precisely to the 3 GO biological process sub-categories, "collagen catabolic process",
"collagen fibril organization" and "skin development".
Interestingly, the gene MMP9 also belongs to the biological process sub-category "collagen catabolic
process". MMP9, located on chromosome 20q13.12, is involved in the breakdown of extracellular
matrix in normal physiological processes, such as tissue remodeling and collagen deposition. It also
plays a significant role in inflammation by facilitating cellular traffic, including neutrophils and
eosinophils. MMP9 also codes for an enzyme which degrades type IV and V collagens. Studies
showed experimental evidence of interaction (or binding) between MMP9 and type IV and V
collagens, e.g. COL4A1, COL4A2 and COL5A1, while COL5A3 and COL5A1 [49] have also been
This article is protected by copyright. All rights reserved.
Accepted Article
experimentally shown to interact (to be in the same protein complex). Moreover, MMP9 and COL5A3
were both showed experimentally to interact with the same protein COL1A2 [49, 50], which also
belongs to the category of collagen catabolic process.
All these observations support our finding of a statistical interaction effect between COL5A3 and
MMP9, which may be accounted for by a degradation of COL5A3 by MMP9 influencing eczema
susceptibility.
In conclusion, the present study detected two new genes having an interactive effect in eczema
susceptibility. It highlights that G-G interaction analyses, beyond single SNP association analysis, can
greatly contribute to the identification of new genes. Further confirmation of the interaction of MMP9
and COL5A3 as well as functional studies are needed to provide a full understanding of the role of
these genes in eczema.
Conflict of interest: none
Sources of support:
This work was supported by the GWIS-AM grant (ANR-11-BSV1-027, ANR-USPC-2013EDAGWAS). The Canada Research Chair in Environment and Genetics of Respiratory
Diseases and Allergy held by C. Laprise since 2005 allows the maintenance of the French
Canadian study.
Acknowledgments
We thank the EGEA cooperative group:
This article is protected by copyright. All rights reserved.
Accepted Article
EGEA cooperative group
Coordination: V Siroux (epidemiology, PI since 2013); F Demenais (genetics); I Pin (clinical aspects); R
Nadif (biology); F Kauffmann (PI 1992-2012).
Respiratory epidemiology: Inserm U 700, Paris: M Korobaeff (Egea1), F Neukirch (Egea1); Inserm U
707, Paris: I Annesi-Maesano (Egea1-2) ; Inserm CESP/U 1018, Villejuif: F Kauffmann, N Le Moual, R
Nadif, MP Oryszczyn (Egea1-2), R Varraso ; Inserm U 823, Grenoble: V Siroux. Genetics: Inserm U 393,
Paris: J Feingold ; Inserm U 946, Paris: E Bouzigon, F Demenais, MH Dizier ; CNG, Evry: I Gut (now
CNAG, Barcelona, Spain), M Lathrop (now Univ McGill, Montreal, Canada).
Clinical centers: Grenoble: I Pin, C Pison; Lyon: D Ecochard (Egea1), F Gormand, Y Pacheco ;
Marseille: D Charpin (Egea1), D Vervloet (Egea1-2) ; Montpellier: J Bousquet ; Paris Cochin: A
Lockhart (Egea1), R Matran (now in Lille) ; Paris Necker: E Paty (Egea1-2), P Scheinmann (Egea1-2) ;
Paris-Trousseau: A Grimfeld (Egea1-2), J Just.
Data and quality management: Inserm ex-U155 (Egea1): J Hochez ; Inserm CESP/U 1018, Villejuif: N
Le Moual ; Inserm ex-U780: C Ravault (Egea1-2) ; Inserm ex-U794: N Chateigner (Egea1-2) ; Grenoble:
J Quentin-Ferran (Egea1-2).
SLSJ: The Canada Research Chair in Environment and Genetics of Respiratory Diseases and Allergy
held by C Laprise and the funding supports from Canadian Institutes of Health Research (CIHR)
enabled the maintenance and continuation of the SLSJ asthma study. C. Laprise is the director of the
Asthma Strategic Group of the Respiratory Health Network of the Fonds de la recherche en santé du
Québec (FRSQ) and director of the SLSJ cohort. She is also member of Allergen network.
This article is protected by copyright. All rights reserved.
Accepted Article
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Dold S, Wjst M, von Mutius E, Reitmeir P, Stiepel E, Genetic risk for asthma, allergic rhinitis,
and atopic dermatitis. Archives of disease in childhood 1992;67: 1018-22.
Wadonda-Kabondo N, Sterne JA, Golding J, Kennedy CT, Archer CB, Dunnill MG, Association
of parental eczema, hayfever, and asthma with atopic dermatitis in infancy: birth cohort
study. Archives of disease in childhood 2004;89: 917-21.
Bin L, Leung DY, Genetic and epigenetic studies of atopic dermatitis. Allergy, asthma, and
clinical immunology : official journal of the Canadian Society of Allergy and Clinical
Immunology 2016;12: 52.
Marenholz I, Rivera VA, Esparza-Gordillo J, Bauerfeind A, Lee-Kirsch MA, Ciechanowicz A,
Kurek M, Piskackova T, Macek M, Lee YA, Association screening in the Epidermal
Differentiation Complex (EDC) identifies an SPRR3 repeat number variant as a risk factor for
eczema. The Journal of investigative dermatology 2011;131: 1644-9.
Paternoster L, Standl M, Chen CM, Ramasamy A, Bonnelykke K, Duijts L, Ferreira MA, Alves
AC, Thyssen JP, Albrecht E, Baurecht H, Feenstra B, Sleiman PM, Hysi P, Warrington NM,
Curjuric I, Myhre R, Curtin JA, Groen-Blokhuis MM, Kerkhof M, Saaf A, Franke A, Ellinghaus D,
Folster-Holst R, Dermitzakis E, Montgomery SB, Prokisch H, Heim K, Hartikainen AL, Pouta A,
Pekkanen J, Blakemore AI, Buxton JL, Kaakinen M, Duffy DL, Madden PA, Heath AC,
Montgomery GW, Thompson PJ, Matheson MC, Le Souef P, St Pourcain B, Smith GD,
Henderson J, Kemp JP, Timpson NJ, Deloukas P, Ring SM, Wichmann HE, Muller-Nurasyid M,
Novak N, Klopp N, Rodriguez E, McArdle W, Linneberg A, Menne T, Nohr EA, Hofman A,
Uitterlinden AG, van Duijn CM, Rivadeneira F, de Jongste JC, van der Valk RJ, Wjst M, Jogi R,
Geller F, Boyd HA, Murray JC, Kim C, Mentch F, March M, Mangino M, Spector TD, Bataille V,
Pennell CE, Holt PG, Sly P, Tiesler CM, Thiering E, Illig T, Imboden M, Nystad W, Simpson A,
Hottenga JJ, Postma D, Koppelman GH, Smit HA, Soderhall C, Chawes B, Kreiner-Moller E,
Bisgaard H, Melen E, Boomsma DI, Custovic A, Jacobsson B, Probst-Hensch NM, Palmer LJ,
Glass D, Hakonarson H, Melbye M, Jarvis DL, Jaddoe VW, Gieger C, Strachan DP, Martin NG,
Jarvelin MR, Heinrich J, Evans DM, Weidinger S, Meta-analysis of genome-wide association
studies identifies three new risk loci for atopic dermatitis. Nature genetics 2011;44: 187-92.
Palmer CN, Irvine AD, Terron-Kwiatkowski A, Zhao Y, Liao H, Lee SP, Goudie DR, Sandilands A,
Campbell LE, Smith FJ, O'Regan GM, Watson RM, Cecil JE, Bale SJ, Compton JG, DiGiovanna
JJ, Fleckman P, Lewis-Jones S, Arseculeratne G, Sergeant A, Munro CS, El Houate B,
McElreavey K, Halkjaer LB, Bisgaard H, Mukhopadhyay S, McLean WH, Common loss-offunction variants of the epidermal barrier protein filaggrin are a major predisposing factor for
atopic dermatitis. Nature genetics 2006;38: 441-6.
Kabesch M, Carr D, Weiland SK, von Mutius E, Association between polymorphisms in serine
protease inhibitor, kazal type 5 and asthma phenotypes in a large German population
sample. Clinical and experimental allergy : journal of the British Society for Allergy and
Clinical Immunology 2004;34: 340-5.
Esparza-Gordillo J, Weidinger S, Folster-Holst R, Bauerfeind A, Ruschendorf F, Patone G,
Rohde K, Marenholz I, Schulz F, Kerscher T, Hubner N, Wahn U, Schreiber S, Franke A, Vogler
R, Heath S, Baurecht H, Novak N, Rodriguez E, Illig T, Lee-Kirsch MA, Ciechanowicz A, Kurek
M, Piskackova T, Macek M, Lee YA, Ruether A, A common variant on chromosome 11q13 is
associated with atopic dermatitis. Nature genetics 2009;41: 596-601.
Soderhall C, Marenholz I, Kerscher T, Ruschendorf F, Esparza-Gordillo J, Worm M, Gruber C,
Mayr G, Albrecht M, Rohde K, Schulz H, Wahn U, Hubner N, Lee YA, Variants in a novel
epidermal collagen gene (COL29A1) are associated with atopic dermatitis. PLoS biology
2007;5: e242.
Gharagozlou M, Farhadi E, Khaledi M, Behniafard N, Sotoudeh S, Salari R, Darabi B, Fathi SM,
Mahmoudi M, Aghamohammadi A, Amirzargar AA, Rezaei N, Association between the
This article is protected by copyright. All rights reserved.
Accepted Article
11.
12.
13.
14.
15.
16.
17.
interleukin 6 genotype at position -174 and atopic dermatitis. Journal of investigational
allergology & clinical immunology 2013;23: 89-93.
Esparza-Gordillo J, Schaarschmidt H, Liang L, Cookson W, Bauerfeind A, Lee-Kirsch MA,
Nemat K, Henderson J, Paternoster L, Harper JI, Mangold E, Nothen MM, Ruschendorf F,
Kerscher T, Marenholz I, Matanovic A, Lau S, Keil T, Bauer CP, Kurek M, Ciechanowicz A,
Macek M, Franke A, Kabesch M, Hubner N, Abecasis G, Weidinger S, Moffatt M, Lee YA, A
functional IL-6 receptor (IL6R) variant is a risk factor for persistent atopic dermatitis. The
Journal of allergy and clinical immunology 2013;132: 371-7.
Portelli MA, Hodge E, Sayers I, Genetic risk factors for the development of allergic disease
identified by genome-wide association. Clinical and experimental allergy : journal of the
British Society for Allergy and Clinical Immunology 2015;45: 21-31.
Paternoster L, Standl M, Waage J, Baurecht H, Hotze M, Strachan DP, Curtin JA, Bonnelykke
K, Tian C, Takahashi A, Esparza-Gordillo J, Alves AC, Thyssen JP, den Dekker HT, Ferreira MA,
Altmaier E, Sleiman PM, Xiao FL, Gonzalez JR, Marenholz I, Kalb B, Pino-Yanes M, Xu CJ,
Carstensen L, Groen-Blokhuis MM, Venturini C, Pennell CE, Barton SJ, Levin AM, Curjuric I,
Bustamante M, Kreiner-Moller E, Lockett GA, Bacelis J, Bunyavanich S, Myers RA, Matanovic
A, Kumar A, Tung JY, Hirota T, Kubo M, McArdle WL, Henderson AJ, Kemp JP, Zheng J, Smith
GD, Ruschendorf F, Bauerfeind A, Lee-Kirsch MA, Arnold A, Homuth G, Schmidt CO, Mangold
E, Cichon S, Keil T, Rodriguez E, Peters A, Franke A, Lieb W, Novak N, Folster-Holst R,
Horikoshi M, Pekkanen J, Sebert S, Husemoen LL, Grarup N, de Jongste JC, Rivadeneira F,
Hofman A, Jaddoe VW, Pasmans SG, Elbert NJ, Uitterlinden AG, Marks GB, Thompson PJ,
Matheson MC, Robertson CF, Ried JS, Li J, Zuo XB, Zheng XD, Yin XY, Sun LD, McAleer MA,
O'Regan GM, Fahy CM, Campbell L, Macek M, Kurek M, Hu D, Eng C, Postma DS, Feenstra B,
Geller F, Hottenga JJ, Middeldorp CM, Hysi P, Bataille V, Spector T, Tiesler CM, Thiering E,
Pahukasahasram B, Yang JJ, Imboden M, Huntsman S, Vilor-Tejedor N, Relton CL, Myhre R,
Nystad W, Custovic A, Weiss ST, Meyers DA, Soderhall C, Melen E, Ober C, Raby BA, Simpson
A, Jacobsson B, Holloway JW, Bisgaard H, Sunyer J, Probst-Hensch NM, Williams LK, Godfrey
KM, Wang CA, Boomsma DI, Melbye M, Koppelman GH, Jarvis D, McLean WH, Irvine AD,
Zhang XJ, Hakonarson H, Gieger C, Burchard EG, Martin NG, Duijts L, Linneberg A, Jarvelin
MR, Nothen MM, Lau S, Hubner N, Lee YA, Tamari M, Hinds DA, Glass D, Brown SJ, Heinrich J,
Evans DM, Weidinger S, Multi-ancestry genome-wide association study of 21,000 cases and
95,000 controls identifies new risk loci for atopic dermatitis. Nature genetics 2015;47: 144956.
Namkung JH, Lee JE, Kim E, Kim S, Shin ES, Cho EY, Yang JM, Association of single nucleotide
polymorphisms in the IL-12 (IL-12A and B) and IL-12 receptor (IL-12Rbeta1 and beta2) genes
and gene-gene interactions with atopic dermatitis in Koreans. Journal of dermatological
science 2010;57: 199-206.
Ziyab AH, Davies GA, Ewart S, Hopkin JM, Schauberger EM, Wills-Karp M, Holloway JW,
Arshad SH, Zhang H, Karmaus W, Interactive effect of STAT6 and IL13 gene polymorphisms
on eczema status: results from a longitudinal and a cross-sectional study. BMC medical
genetics 2013;14: 67.
Shen C, Liu L, Jiang Z, Zheng X, Meng L, Yin X, Gao J, Sheng Y, Li Y, Zhou F, Xiao F, Sun L, Cui Y,
Yang S, Zuo X, Zhang X, Four genetic variants interact to confer susceptibility to atopic
dermatitis in Chinese Han population. Molecular genetics and genomics : MGG 2015;290:
1493-8.
Weidinger S, Baurecht H, Wagenpfeil S, Henderson J, Novak N, Sandilands A, Chen H,
Rodriguez E, O'Regan GM, Watson R, Liao H, Zhao Y, Barker JN, Allen M, Reynolds N, Meggitt
S, Northstone K, Smith GD, Strobl C, Stahl C, Kneib T, Klopp N, Bieber T, Behrendt H, Palmer
CN, Wichmann HE, Ring J, Illig T, McLean WH, Irvine AD, Analysis of the individual and
aggregate genetic contributions of previously identified serine peptidase inhibitor Kazal type
5 (SPINK5), kallikrein-related peptidase 7 (KLK7), and filaggrin (FLG) polymorphisms to
eczema risk. The Journal of allergy and clinical immunology 2008;122: 560-8 e4.
This article is protected by copyright. All rights reserved.
Accepted Article
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
O'Regan GM, Campbell LE, Cordell HJ, Irvine AD, McLean WH, Brown SJ, Chromosome
11q13.5 variant associated with childhood eczema: an effect supplementary to filaggrin
mutations. The Journal of allergy and clinical immunology 2010;125: 170-4 e1-2.
Culverhouse R, Suarez BK, Lin J, Reich T, A perspective on epistasis: limits of models
displaying no main effect. Am J Hum Genet 2002;70: 461-71.
Marchini J, Donnelly P, Cardon LR, Genome-wide strategies for detecting multiple loci that
influence complex diseases. Nat Genet 2005;37: 413-7.
Herold C, Steffens M, Brockschmidt FF, Baur MP, Becker T, INTERSNP: genome-wide
interaction analysis guided by a priori information. Bioinformatics 2009;25: 3275-81.
Babron MC, Guilloud-Bataille M, Sahbatou M, Demenais F, Genin E, Dizier MH, Comparative
power of family-based association strategies to detect disease-causing variants under twolocus models. Genet Epidemiol 2012;36: 848-55.
Sun LD, Xiao FL, Li Y, Zhou WM, Tang HY, Tang XF, Zhang H, Schaarschmidt H, Zuo XB,
Foelster-Holst R, He SM, Shi M, Liu Q, Lv YM, Chen XL, Zhu KJ, Guo YF, Hu DY, Li M, Zhang YH,
Zhang X, Tang JP, Guo BR, Wang H, Liu Y, Zou XY, Zhou FS, Liu XY, Chen G, Ma L, Zhang SM,
Jiang AP, Zheng XD, Gao XH, Li P, Tu CX, Yin XY, Han XP, Ren YQ, Song SP, Lu ZY, Zhang XL, Cui
Y, Chang J, Gao M, Luo XY, Wang PG, Dai X, Su W, Li H, Shen CP, Liu SX, Feng XB, Yang CJ, Lin
GS, Wang ZX, Huang JQ, Fan X, Wang Y, Bao YX, Yang S, Liu JJ, Franke A, Weidinger S, Yao ZR,
Zhang XJ, Genome-wide association study identifies two new susceptibility loci for atopic
dermatitis in the Chinese Han population. Nature genetics 2011;43: 690-4.
Hirota T, Takahashi A, Kubo M, Tsunoda T, Tomita K, Sakashita M, Yamada T, Fujieda S,
Tanaka S, Doi S, Miyatake A, Enomoto T, Nishiyama C, Nakano N, Maeda K, Okumura K,
Ogawa H, Ikeda S, Noguchi E, Sakamoto T, Hizawa N, Ebe K, Saeki H, Sasaki T, Ebihara T,
Amagai M, Takeuchi S, Furue M, Nakamura Y, Tamari M, Genome-wide association study
identifies eight new susceptibility loci for atopic dermatitis in the Japanese population.
Nature genetics 2012;44: 1222-6.
Ellinghaus D, Baurecht H, Esparza-Gordillo J, Rodriguez E, Matanovic A, Marenholz I, Hubner
N, Schaarschmidt H, Novak N, Michel S, Maintz L, Werfel T, Meyer-Hoffert U, Hotze M,
Prokisch H, Heim K, Herder C, Hirota T, Tamari M, Kubo M, Takahashi A, Nakamura Y, Tsoi LC,
Stuart P, Elder JT, Sun L, Zuo X, Yang S, Zhang X, Hoffmann P, Nothen MM, Folster-Holst R,
Winkelmann J, Illig T, Boehm BO, Duerr RH, Buning C, Brand S, Glas J, McAleer MA, Fahy CM,
Kabesch M, Brown S, McLean WH, Irvine AD, Schreiber S, Lee YA, Franke A, Weidinger S,
High-density genotyping study identifies four new susceptibility loci for atopic dermatitis.
Nature genetics 2013;45: 808-12.
Schaarschmidt H, Ellinghaus D, Rodriguez E, Kretschmer A, Baurecht H, Lipinski S, MeyerHoffert U, Harder J, Lieb W, Novak N, Folster-Holst R, Esparza-Gordillo J, Marenholz I,
Ruschendorf F, Hubner N, Reischl E, Waldenberger M, Gieger C, Illig T, Kabesch M, Zhang XJ,
Xiao FL, Lee YA, Franke A, Weidinger S, A genome-wide association study reveals 2 new
susceptibility loci for atopic dermatitis. The Journal of allergy and clinical immunology
2015;136: 802-6.
Kauffmann F, Dizier MH, Pin I, Paty E, Gormand F, Vervloet D, Bousquet J, Neukirch F, Annesi
I, Oryszczyn MP, Lathrop M, Demenais F, Lockhart A, Feingold J, Epidemiological study of the
genetics and environment of asthma, bronchial hyperresponsiveness, and atopy: phenotype
issues. Am J Respir Crit Care Med 1997;156: S123-9.
Guilloud-Bataille M, Bouzigon E, Annesi-Maesano I, Bousquet J, Charpin D, Gormand F,
Hochez J, Just J, Lemainque A, Le Moual N, Matran R, Neukirch F, Oryszczyn MP, Paty E, Pin I,
Vervloet D, Kauffmann F, Lathrop M, Demenais F, Dizier MH, Evidence for linkage of a new
region (11p14) to eczema and allergic diseases. Human genetics 2008;122: 605-14.
Dizier MH, Margaritte-Jeannin P, Madore AM, Esparza-Gordillo J, Moffatt M, Corda E, Monier
F, Guilloud-Bataille M, Franke A, Weidinger S, Annesi-Maesano I, Just J, Pin I, Kauffmann F,
Cookson W, Lee YA, Laprise C, Lathrop M, Bouzigon E, Demenais F, The ANO3/MUC15 locus is
This article is protected by copyright. All rights reserved.
Accepted Article
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
associated with eczema in families ascertained through asthma. J Allergy Clin Immunol
2012;129: 1547-53 e3.
Laprise C, The Saguenay-Lac-Saint-Jean asthma familial collection: the genetics of asthma in a
young founder population. Genes Immun 2014;15: 247-55.
Moffatt MF, Kabesch M, Liang L, Dixon AL, Strachan D, Heath S, Depner M, von Berg A, Bufe
A, Rietschel E, Heinzmann A, Simma B, Frischer T, Willis-Owen SA, Wong KC, Illig T, Vogelberg
C, Weiland SK, von Mutius E, Abecasis GR, Farrall M, Gut IG, Lathrop GM, Cookson WO,
Genetic variants regulating ORMDL3 expression contribute to the risk of childhood asthma.
Nature 2007;448: 470-3.
Moffatt MF, Gut IG, Demenais F, Strachan DP, Bouzigon E, Heath S, von Mutius E, Farrall M,
Lathrop M, Cookson WO, A large-scale, consortium-based genomewide association study of
asthma. N Engl J Med 2010;363: 1211-21.
Emily M, AGGrEGATOr: A Gene-based GEne-Gene interActTiOn test for case-control
association studies. Statistical applications in genetics and molecular biology 2016;15: 15171.
Moskvina V, Schmidt KM, On multiple-testing correction in genome-wide association studies.
Genetic epidemiology 2008;32: 567-73.
Conneely KN, Boehnke M, So many correlated tests, so little time! Rapid adjustment of P
values for multiple correlated tests. Am J Hum Genet 2007;81: 1158-68.
Becker T, Herold C, Meesters C, Mattheisen M, Baur MP, Significance levels in genome-wide
interaction analysis (GWIA). Ann Hum Genet 2011;75: 29-35.
Bottema RW, Kerkhof M, Reijmerink NE, Thijs C, Smit HA, van Schayck CP, Brunekreef B, van
Oosterhout AJ, Postma DS, Koppelman GH, Gene-gene interaction in regulatory T-cell
function in atopy and asthma development in childhood. J Allergy Clin Immunol 2010;126:
338-46, 46 e1-10.
Reijmerink NE, Bottema RW, Kerkhof M, Gerritsen J, Stelma FF, Thijs C, van Schayck CP, Smit
HA, Brunekreef B, Koppelman GH, Postma DS, TLR-related pathway analysis: novel gene-gene
interactions in the development of asthma and atopy. Allergy 2010;65: 199-207.
Savenije OE, Mahachie John JM, Granell R, Kerkhof M, Dijk FN, de Jongste JC, Smit HA,
Brunekreef B, Postma DS, Van Steen K, Henderson J, Koppelman GH, Association of IL33-IL-1
receptor-like 1 (IL1RL1) pathway polymorphisms with wheezing phenotypes and asthma in
childhood. J Allergy Clin Immunol 2014;134: 170-7.
Kabesch M, Schedel M, Carr D, Woitsch B, Fritzsch C, Weiland SK, von Mutius E, IL-4/IL-13
pathway genetics strongly influence serum IgE levels and childhood asthma. J Allergy Clin
Immunol 2006;117: 269-74.
Martin BK, Frangakis CE, Rosenberg PB, Mintzer JE, Katz IR, Porsteinsson AP, Schneider LS,
Rabins PV, Munro CA, Meinert CL, Niederehe G, Lyketsos CG, Design of Depression in
Alzheimer's Disease Study-2. The American journal of geriatric psychiatry : official journal of
the American Association for Geriatric Psychiatry 2006;14: 920-30.
Ritchie MD, Hahn LW, Roodi N, Bailey LR, Dupont WD, Parl FF, Moore JH, Multifactordimensionality reduction reveals high-order interactions among estrogen-metabolism genes
in sporadic breast cancer. American journal of human genetics 2001;69: 138-47.
Vaysse A, Fang S, Brossard M, Wei Q, Chen WV, Mohamdi H, Vincent-Fetita L, MargaritteJeannin P, Lavielle N, Maubec E, Lathrop M, Avril MF, Amos CI, Lee JE, Demenais F, A
comprehensive genome-wide analysis of melanoma Breslow thickness identifies interaction
between CDC42 and SCIN genetic variants. International journal of cancer 2016;139: 201220.
Nakashima K, Hirota T, Obara K, Shimizu M, Doi S, Fujita K, Shirakawa T, Enomoto T,
Yoshihara S, Ebisawa M, Matsumoto K, Saito H, Suzuki Y, Nakamura Y, Tamari M, A functional
polymorphism in MMP-9 is associated with childhood atopic asthma. Biochemical and
biophysical research communications 2006;344: 300-7.
This article is protected by copyright. All rights reserved.
Accepted Article
45.
Pinto LA, Depner M, Klopp N, Illig T, Vogelberg C, von Mutius E, Kabesch M, MMP-9 gene
variants increase the risk for non-atopic asthma in children. Respiratory research 2010;11:
23.
Harper JI, Godwin H, Green A, Wilkes LE, Holden NJ, Moffatt M, Cookson WO, Layton G,
Chandler S, A study of matrix metalloproteinase expression and activity in atopic dermatitis
using a novel skin wash sampling assay for functional biomarker analysis. The British journal
of dermatology 2010;162: 397-403.
Laprise C, Sladek R, Ponton A, Bernier MC, Hudson TJ, Laviolette M, Functional classes of
bronchial mucosa genes that are differentially expressed in asthma. BMC genomics 2004;5:
21.
Kelly EA, Busse WW, Jarjour NN, Increased matrix metalloproteinase-9 in the airway after
allergen challenge. American journal of respiratory and critical care medicine 2000;162:
1157-61.
Niyibizi C, Fietzek PP, van der Rest M, Human placenta type V collagens. Evidence for the
existence of an alpha 1(V) alpha 2(V) alpha 3(V) collagen molecule. The Journal of biological
chemistry 1984;259: 14170-4.
Makela M, Salo T, Larjava H, MMP-9 from TNF alpha-stimulated keratinocytes binds to cell
membranes and type I collagen: a cause for extended matrix degradation in inflammation?
Biochemical and biophysical research communications 1998;253: 325-35.
46.
47.
48.
49.
50.
Study
Phenotypes
No. (%)
Male sex
(%)
Mean age
(SD)
Asthma (%)
Atopy (%)
EGEA
All
1493 (100)
768 (51)
30.3 (0.44)
690 (46)
788 (53)
ECZ-
1082 (72)
553 (51)
32.9 (0.51)
419 (39)
510 (47)
ECZ+
411 (28)
215 (52)
23.3 (0.75)
271 (66)
278 (68)
All
1153 (100)
527 (46)
38.2 (0.64)
557 (48)
639 (56)
ECZ-
844 (73)
421 (50)
39.9 (0.76)
366 (43)
424 (51)
ECZ+
309 (27)
106 (34)
33.4 (1.17)
191 (62)
215 (70)
All
254 (100)
144 (57)
11.2 (0.30)
176 (69)
104 (41)
ECZ-
174 (69)
96 (55)
11.0 (0.37)
105 (60)
55 (32)
ECZ+
80 (31)
48 (60)
11.5 (0.49)
71 (89)
49 (61)
SLSJ
MRCA
ECZ+, with eczema; ECZ-, without eczema; SD, standard deviation;
Atopy: positive response of skin prick test to at least one allergen (house dust mite, cockroach,
cat, timothy grass, olive, birch, ragweed, pellitory, Alternaria, Aspergillus and Cladosporium
were tested in EGEA; house dust mite, dust, cat, herbs, trees, Alternaria, Aspergillus,
Cladosporium, Hormodendrum and Penicillium in SLSJ; dust mite, cat, timothy grass,
Aspergillus and Cladosporium in MRCA).
This article is protected by copyright. All rights reserved.
Accepted Article
Table I: Phenotypic characteristics of genotyped individuals in the three asthma-ascertained samples
Collagen
gene
EGEA
SLSJ
MRCA
All samples*
PSNP-SNP
PSNP-SNP
PSNP-SNP
Pmeta-SNP-SNP
SNP1
gene
SNP2
rs1293288
CTSB
rs3745596 COL5A3 3.56x10-6 8.65x10-1 5.65x10-1 1.20x10-3
rs2521205
COL1A2
rs1053312 COL6A1 6.70x10-6 4.36x10-1 8.61x10-3 2.88x10-2
rs2287807
COL5A3
rs17576
MMP9
6.70x10-6 1.24x10-3 1.33x10-1 1.11x10-8
SNP2
gene
PSNP-SNP
Epidermis or Skin
SNP1
gene
rs12854617 ATP8A2
rs6106524 TGM3
PSNP-SNP
PSNP-SNP
6.35x10-6 4.18x10-1 6.60x10-1
Pmeta-SNP-SNP
4.69x10-4
ITGA2
6.73x10-6 1.79x10-1 3.37x10-1 5.74x10-3
GAB1
rs6026561 GNAS
8.60x10-6 3.75x10-1 8.73x10-1 1.37x10-4
GAB1
rs6026561 GNAS
9.25x10-6 3.66x10-1 8.78x10-1 1.34x10-4
rs7586353
ABCA12 rs246406
rs6537155
rs3805251
*
: meta-analysis of interactive effects using fixed effect model
This article is protected by copyright. All rights reserved.
Accepted Article
Table II : SNPxSNP interaction test results in EGEA, SLSJ and MRCA for SNP pairs selected in
EGEA (PSNP-SNP≤10-5)
EGEA
SLSJ
MRCA
Meta-analysis
PG-G *
PG-G *
Pmeta-G-G **
gene1
gene2
No of PG-G *
SNP
pairs
COL1A2
COL6A1
7
2.14x10-5
2.87x10-1
2.90x10-2
7.18x10-2
COL5A3
MMP9
8
4.49x10-5
8.79x10-3
5.23x10-1
4.04x10-8
COL5A3
CTSB
47
1.48x10-4
2.61x10-2
7.11x10-1
9.86x10-3
* : gene x gene analysis using AGGrEGATOr
**: gene x gene meta-analysis using AGGrEGATOr with Pvalues of meta-analysis of SNPxSNP
interaction test in EGEA, SLSJ and MRCA
Table III: Gene x Gene interaction test results in EGEA, SLSJ and MRCA for SNP pairs selected in
EGEA (PG-G ≤ 10-4)
Figure legend
Figure 1. Two-locus Odds-ratio (ORs) and 95% confidence intervals for eczema estimated for the
genotypic combinations of the two SNPs, rs2287807 (COL5A3) and rs17576 (MMP9). The ORs were
computing using the estimates of SNP main effect and interaction obtained from the pooled analysis of
EGEA, SLSJ and MRCA.
This article is protected by copyright. All rights reserved.
Accepted Article
Figure 1 : Two locus Odds-ratio (ORs) and 95% confidence intervals for eczema for the genotypic
combinations of COL5A3 and MMP9 SNPS.
This article is protected by copyright. All rights reserved.