Oxidative stress resulting from an increased amount of reactive oxygen species and an imbalance between oxidants and antioxidants plays an important role in the pathogenesis of asthma. The present study tested the hypothesis that genetic susceptibility to allergic and nonallergic variants of asthma is determined by complex interactions between genes encoding antioxidant defense enzymes (ADE). We carried out a comprehensive analysis of the associations between adult asthma and 46 single nucleotide polymorphisms of 34 ADE genes and 12 other candidate genes of asthma in Russian population using set association analysis and multifactor dimensionality reduction approaches. We found for the first time epistatic interactions between ADE genes underlying asthma susceptibility and the genetic heterogeneity between allergic and nonallergic variants of the disease. We identified
Bronchial asthma (BA) is a common chronic inflammatory disease of the airways characterized by variable and recurring symptoms, reversible airflow obstruction, and bronchospasm [
In the recent years, the relationships between common genetic variants and BA risk are being reported with rapidly increasing frequency. Large-scale genome-wide association studies (GWAS) have been recently done to look for asthma susceptibility genes in ethnically diverse populations of the world [
It is widely agreed that the expression of a disease phenotype may not accurately be predicted from the knowledge of the effects of individual genes because of complex nonlinear interactions between genes, including epistatic and additive interactions [
An important task for a genetic epidemiologist utilizing a candidate gene approach is the selection of appropriate genes and SNPs for testing the disease association. Compared with studying individual genes, the inferences derived from a hypothesis-driven candidate pathway study are enhanced by allowing global conclusions about the involvement of entire biochemical pathway to the pathogenesis of disease [
It is well known that oxidative stress resulting from an increased amount of ROS and an imbalance between oxidants and antioxidants plays a role in the molecular mechanisms underlying BA [
The study protocol was approved by the Ethical Review Committee of Kursk State Medical University, and written informed consent was obtained from each participant before the study. The participants comprised a total of 429 unrelated individuals (215 patients with asthma and 214 healthy controls); all are ethnically Russians from Central Russia (mainly from the Kursk region). All study subjects were recruited from the Division of Pulmonology at the Kursk Regional Clinical Hospital between 2003 and 2004. Asthma was diagnosed by qualified pulmonologists on the basis of the WHO criteria, as described previously [
The candidate genes for this study were selected according to the guidelines for genetic association studies proposed by Cooper and coauthors [
Genomic DNA of all study participants was isolated from 5–10 mL of peripheral blood samples, collected in K3-EDTA tubes by venipuncture, and maintained at −20°C until processed. Twenty-five of the selected gene polymorphisms had been genotyped in our previous studies [
Description of the polymorphisms included in this study.
Gene symbols |
Gene name | Polymorphism (SNP) | Location | SNP ID | |
---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 6 |
|
|||||
1 |
|
Glutathione peroxidase 1 | C>T (P198L) | exon 1 | rs1050450 |
2 |
|
Glutathione peroxidase 2 (gastrointestinal) | G>A (R146C) | exon 2 | rs17880492 |
3 |
|
Glutathione peroxidase 3 (plasma) | 249G>A | 3′ UTR | rs2070593 |
4 |
|
Glutathione peroxidase 4 (phospholipid hydroperoxidase) | C718T | 3′ UTR | rs713041 |
5 |
|
Glutathione reductase | T>C (30546636T>C) | intron 9 | rs2551715 |
6 |
|
Superoxide dismutase 2, mitochondrial | A16V | exon 2 | rs4880 |
7 |
|
Superoxide dismutase 3, extracellular | A40T (A58T) | exon 3 | rs2536512 |
8 |
|
Catalase |
−21A>T (−89A>T) | 5′ UTR | rs7943316 |
9 |
|
−262C>T (4760C>T) | 5′ UTR | rs1001179 | |
10 |
|
Glutamate-cysteine ligase, modifier subunit | −588C>T (4704C>T) | 5′ UTR | rs41303970 |
11 |
|
−23G>T | 5′UTR | rs743119 | |
12 |
|
NAD(P)H dehydrogenase, quinone 1 |
P187S | exon 6 | rs1800566 |
13 |
|
R139W | exon 4 | rs4986998 | |
14 |
|
Cytochrome b-245, alpha polypeptide |
242C>T (Y72H) | exon 4 | rs4673 |
15 |
|
640A>G (24G>A) | 3′ UTR | rs1049255 | |
16 |
|
−930A>G | 5′ UTR | rs9932581 | |
17 |
|
Myeloperoxidase | −463G>A (4535G>A) | 5′ UTR | rs2333227 |
18 |
|
Peroxiredoxin 1 | C>A | 5′ UTR | rs17522918 |
19 |
|
Thioredoxin reductase 1 | C>G | 5′ UTR | rs1128446 |
20 |
|
Flavin-containing monooxygenase 3 | E158K | exon 4 | rs2266782 |
21 |
|
Cytochrome P450, family 1, subfamily A, polypeptide 1 | I462V | exon 7 | rs1048943 |
22 |
|
T6235C | 3′ UTR | rs4646903 | |
23 |
|
Cytochrome P450, family 2, subfamily |
−1293G>C | 5′ UTR | rs3813867 |
24 |
|
−1053C>T | 5′ UTR | rs2031920 | |
25 |
|
7632T>A | intron 6 | rs6413432 | |
26 |
|
9896C>G | intron 7 | rs2070676 | |
27 |
|
Epoxide hydrolase 1, microsomal (xenobiotic) | Y113H (337T>C) | exon 3 | rs1051740 |
28 |
|
H139R (416A>G) | exon 4 | rs2234922 | |
29 |
|
Paraoxonase 1 | Q192R | exon 6 | rs662 |
30 |
|
Paraoxonase 2 | C311S | exon 9 | rs7493 |
31 |
|
Glutathione S-transferase mu 1 | Expressor/deletion | exons 6-7 | — |
32 |
|
Glutathione S-transferase theta 1 | Expressor/deletion | exon 4 | — |
33 |
|
Glutathione S-transferase pi 1 |
I105V | exon 5 | rs1695 |
34 |
|
A114V | exon 6 | rs1138272 | |
35 |
|
Tumor necrosis factor | −308G>A | 5′ UTR | rs1800629 |
36 |
|
Interleukin 1, beta | −511C>T | 5′ UTR | rs16944 |
37 |
|
Interleukin 3 (colony-stimulating factor, multiple) | S27P | exon 1 | rs40401 |
38 |
|
−15C>T | 5′ UTR | rs31480 | |
39 |
|
Interleukin 5 (colony-stimulating factor, eosinophil) | C-703T | 5′ UTR | rs2069812 |
40 |
|
Colony stimulating factor 2 receptor, beta, low-affinity (granulocyte-macrophage) | G1972A | exon 5 | rs131840 |
41 |
|
Interleukin 9 | T113M | exon 5 | rs2069885 |
42 |
|
Interleukin 13 | −1111C>T | 5′ UTR | rs1800925 |
43 |
|
Secretoglobin, family 1A, member 1 (uteroglobin) | A38G | exon 1 | rs11549442 |
44 |
|
Serpin peptidase inhibitor, clade A (alpha-1 antiproteinase, antitrypsin), member 1 | E288V | exon 3 | rs17580 |
45 |
|
D365N | exon 5 | rs143370956 | |
46 |
|
1331G>A | 3′ UTR | rs11568814 |
The concordance of genotypes prevalence in patients with asthma and healthy controls with values expected under Hardy-Weinberg equilibrium was assessed by Pearson’s chi-square test. The association between ADE gene polymorphisms and asthma was examined with binary logistic regression analysis with calculation of odds ratios (OR) and 95% confidence intervals (CI). The statistical calculations were done using Statistica for Windows (v8.0) software package (StatSoft; Tulsa, OK, USA). The statistical significance was established at the
Two bioinformatic approaches, SAA and MDR, were applied for the analysis of gene-gene interactions. The principle of SAA is described in detail elsewhere [
MDR is a flexible nonparametric and genetic model free method for analysis of high-order nonlinear or nonadditive gene-gene interactions [
Both the SAA and MDR are limited by the identification of a few number of high penetrance interacting genes, whereas a larger portion of genes of low-to-moderate effects remain out of the analysis. To address this issue, we performed post hoc comparisons of two-locus genotype combinations (only for those SNPs which were found in gene-gene interaction models obtained by SAA and/or MDR methods) between the case and control groups to look for the genotype combinations which determine the risk of asthma. The observed associations were adjusted for multiple tests using Bonferroni procedure.
Allele and genotype frequencies of the studied genes are shown in Tables
Allele frequencies of genes investigated in the present study.
Gene | Polymorphism | Alleles | Allele frequency | |||
---|---|---|---|---|---|---|
Controls |
Asthma, entire group |
Allergic asthma |
Nonallergic asthma | |||
1 | 2 | 3 | 4 | 5 | 6 | 7 |
|
||||||
|
G>A |
G | 0.991 | 0.981 | 0.987 | 0.964 |
A | 0.009 | 0.019 | 0.013 | 0.036 | ||
|
G>A |
G | 0.703 | 0.726 | 0.734 | 0.696 |
A | 0.297 | 0.274 | 0.266 | 0.304 | ||
|
C718T |
718T | 0.402 | 0.391 | 0.407 | 0.348 |
718C | 0.598 | 0.609 | 0.593 | 0.652 | ||
|
T>C |
T | 0.442 | 0.398 | 0.362 | 0.491 |
C | 0.558 | 0.602 | 0.638* | 0.509 | ||
|
A16V |
16A | 0.528 | 0.486 | 0.481 | 0.509 |
16V | 0.472 | 0.514 | 0.519 | 0.491 | ||
|
A40T |
40A | 0.322 | 0.321 | 0.324 | 0.295 |
40T | 0.678 | 0.679 | 0.676 | 0.705 | ||
|
C>A |
C | 0.923 | 0.937 | 0.942 | 0.920 |
A | 0.077 | 0.063 | 0.058 | 0.080 | ||
|
C>G |
C | 0.808 | 0.821 | 0.808 | 0.857 |
G | 0.192 | 0.179 | 0.192 | 0.143 | ||
|
E158K |
158E | 0.549 | 0.537 | 0.529 | 0.571 |
158K | 0.451 | 0.463 | 0.471 | 0.429 | ||
|
−308G>A |
−308G | 0.888 | 0.872 | 0.875 | 0.866 |
−308A | 0.112 | 0.128 | 0.125 | 0.134 | ||
|
−511C>T |
−511C | 0.710 | 0.664 | 0.670 | 0.652 |
−511T | 0.290 | 0.336 | 0.330 | 0.348 | ||
|
S27P |
27S | 0.738 | 0.685 | 0.686 | 0.688 |
27P | 0.262 | 0.315 | 0.314 | 0.313 | ||
|
−15C>T |
−15C | 0.741 | 0.683 | 0.686 | 0.688 |
−15T | 0.259 | 0.317 | 0.314 | 0.313 | ||
|
C-703T |
−703C | 0.673 | 0.778 | 0.788 | 0.732 |
−703T | 0.327 | 0.222* | 0.212* | 0.268 | ||
|
G1972A |
1972G | 0.831 | 0.866 | 0.872 | 0.848 |
1972A | 0.169 | 0.134 | 0.128 | 0.152 | ||
|
T113M |
113T | 0.820 | 0.863 | 0.856 | 0.893 |
113M | 0.180 | 0.137 | 0.144 | 0.107 | ||
|
−1111C>T |
−1111C | 0.729 | 0.693 | 0.692 | 0.714 |
−1111T | 0.271 | 0.307 | 0.308 | 0.286 | ||
|
A38G |
38A | 0.347 | 0.367 | 0.372 | 0.366 |
38G | 0.653 | 0.633 | 0.628 | 0.634 | ||
|
E288V |
288E | 0.993 | 0.991 | 0.990 | 0.991 |
288V | 0.007 | 0.009 | 0.010 | 0.009 | ||
|
D365N |
365D | 0.991 | 0.995 | 0.997 | 0.991 |
365N | 0.009 | 0.005 | 0.003 | 0.009 | ||
|
1331G>A |
1331G | 0.937 | 0.933 | 0.926 | 0.946 |
1331A | 0.063 | 0.067 | 0.074 | 0.054 |
Genotype frequencies of genes investigated in the present study.
Gene | Polymorphism | Genotypes | Genotype distributions, |
|||||||
---|---|---|---|---|---|---|---|---|---|---|
Controls |
Asthma, entire group |
Allergic asthma |
Nonallergic asthma | |||||||
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
|
||||||||||
|
G>A |
GG | 210 | 98.1 | 207 | 96.3 | 152 | 97.4 | 52 | 92.9 |
GA | 4 | 1.9 | 8 | 3.7 | 4 | 2.6 | 4 | 7.1 | ||
AA | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||
|
G>A |
GG | 105 | 49.1 | 113 | 52.6 | 83 | 53.2 | 28 | 50.0 |
GA | 91 | 42.5 | 86 | 40.0 | 63 | 40.4 | 22 | 39.3 | ||
AA | 18 | 8.4 | 16 | 7.4 | 10 | 6.4 | 6 | 10.7 | ||
|
C718T |
718TT | 31 | 14.5 | 33 | 15.3 | 25 | 16.0 | 8 | 14.3 |
718TC | 110 | 51.4 | 102 | 47.4 | 77 | 49.4 | 23 | 41.1 | ||
718CC | 73 | 34.1 | 80 | 37.2 | 54 | 34.6 | 25 | 44.6 | ||
|
T>C |
TT | 40 | 18.7 | 32 | 14.9 | 17 | 10.9* | 15 | 26.8 |
TC | 109 | 50.9 | 107 | 49.8 | 79 | 50.6 | 25 | 44.6 | ||
CC | 65 | 30.4 | 76 | 35.3 | 60 | 38.5 | 16 | 28.6 | ||
|
A16V |
16AA | 59 | 27.6 | 49 | 22.8 | 34 | 21.8 | 15 | 26.8 |
16AV | 108 | 50.5 | 111 | 51.6 | 82 | 52.6 | 27 | 48.2 | ||
16VV | 47 | 22.0 | 55 | 25.6 | 40 | 25.6 | 14 | 25.0 | ||
|
A40T |
40AA | 21 | 9.8 | 24 | 11.2 | 19 | 12.2 | 4 | 7.1 |
40AT | 96 | 44.9 | 90 | 41.9 | 63 | 40.4 | 25 | 44.6 | ||
40TT | 97 | 45.3 | 101 | 47.0 | 74 | 47.4 | 27 | 48.2 | ||
|
C>A |
CC | 182 | 85.0 | 188 | 87.4 | 138 | 88.5 | 47 | 83.9 |
CA | 31 | 14.5 | 27 | 12.6 | 18 | 11.5 | 9 | 16.1 | ||
AA | 1 | 0.5 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | ||
|
C>G |
CC | 140 | 65.4 | 145 | 67.4 | 101 | 64.7 | 42 | 75.0 |
CG | 66 | 30.8 | 63 | 29.3 | 50 | 32.1 | 12 | 21.4 | ||
GG | 8 | 3.7 | 7 | 3.3 | 5 | 3.2 | 2 | 3.6 | ||
|
E158K |
158EE | 57 | 26.6 | 59 | 27.4 | 39 | 25.0 | 20 | 35.7 |
158EK | 121 | 56.5 | 113 | 52.6 | 87 | 55.8 | 24 | 42.9 | ||
158KK | 36 | 16.8 | 43 | 20.0 | 30 | 19.2 | 12 | 21.4 | ||
|
−308G>A |
−308GG | 170 | 79.4 | 162 | 75.4 | 118 | 75.6 | 42 | 75.0 |
−308GA | 40 | 18.7 | 51 | 23.7 | 37 | 23.7 | 13 | 23.2 | ||
−308AA | 4 | 1.9 | 2 | 0.9 | 1 | 0.6 | 1 | 1.8 | ||
|
−511C>T |
−511CC | 114 | 53.3 | 91 | 42.1 | 67 | 43.0* | 23 | 41.1 |
−511CT | 76 | 35.5 | 105 | 48.6 | 75 | 48.1* | 27 | 48.2 | ||
−511TT | 24 | 11.2 | 20 | 9.3 | 14 | 9.0 | 6 | 10.7 | ||
|
S27P |
27SS | 120 | 56.1 | 104 | 48.2 | 77 | 49.4 | 25 | 44.6 |
27SP | 76 | 35.5 | 88 | 40.7 | 60 | 38.5 | 27 | 48.2 | ||
273P | 18 | 8.4 | 24 | 11.1 | 19 | 12.2 | 4 | 7.1 | ||
|
−15C>T |
−15CC | 120 | 56.1 | 103 | 47.7 | 77 | 49.4 | 25 | 44.6 |
−15CT | 77 | 36.0 | 89 | 41.2 | 60 | 38.5 | 27 | 48.2 | ||
−15TT | 17 | 7.9 | 24 | 11.1 | 19 | 12.2 | 4 | 7.1 | ||
|
C-703T |
−703CC | 90 | 42.1 | 132 | 61.1 | 97 | 62.2* | 31 | 55.4 |
−703CT | 108 | 50.5 | 72 | 33.3 | 52 | 33.3* | 20 | 35.7* | ||
−703TT | 16 | 7.5 | 12 | 5.6 | 7 | 4.5 | 5 | 8.9 | ||
|
G1972A |
1972GG | 136 | 67.7 | 160 | 74.1 | 118 | 75.6 | 39 | 69.6 |
1972GA | 62 | 30.9 | 54 | 25.0 | 36 | 23.1 | 17 | 30.4 | ||
1972AA | 3 | 1.5 | 2 | 0.9 | 2 | 1.3 | 0 | 0.0 | ||
|
T113M |
113TT | 146 | 68.2 | 159 | 73.6 | 113 | 72.4 | 44 | 78.6 |
113TM | 59 | 27.6 | 55 | 25.5 | 41 | 26.3 | 12 | 21.4 | ||
113MM | 9 | 4.2 | 2 | 0.9 | 2 | 1.3 | 0 | 0.0 | ||
|
−1111C>T |
−1111CC | 114 | 53.3 | 101 | 47.0 | 75 | 48.1 | 26 | 46.4 |
−1111CT | 84 | 39.3 | 96 | 44.7 | 66 | 42.3 | 28 | 50.0 | ||
−1111CT | 16 | 7.5 | 18 | 8.4 | 15 | 9.6 | 2 | 3.6 | ||
|
A38G |
38AA | 25 | 11.8 | 28 | 13.0 | 23 | 14.7 | 5 | 8.9 |
38AG | 97 | 45.8 | 102 | 47.4 | 70 | 44.9 | 31 | 55.4 | ||
38GG | 90 | 42.5 | 85 | 39.5 | 63 | 40.4 | 20 | 35.7 | ||
|
E288V |
288EE | 211 | 98.6 | 212 | 98.1 | 153 | 98.1 | 55 | 98.2 |
288EV | 3 | 1.4 | 4 | 1.9 | 3 | 1.9 | 1 | 1.8 | ||
288VV | — | — | — | — | — | — | — | — | ||
|
D365N |
365DD | 210 | 98.1 | 214 | 99.1 | 153 | 98.1 | 55 | 98.2 |
365DN | 4 | 1.9 | 2 | 0.9 | 3 | 1.9 | 1 | 1.8 | ||
365NN | — | — | — | — | — | — | — | — | ||
|
1331G>A |
1331GG | 188 | 87.9 | 187 | 86.6 | 133 | 85.3 | 50 | 89.3 |
1331GA | 25 | 11.7 | 29 | 13.4 | 23 | 14.7 | 6 | 10.7 | ||
1331AA | 1 | 0.5 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 |
Taking into account the polygenic basis of asthma, it was an important task to investigate high-order gene-gene interactions using specialized bioinformatics approach called set association analysis which captures the simultaneous effects of multiple genes and achieves a global view of gene action and interaction [
The results of statistical modeling of gene-gene interactions in allergic asthma using set association approach. Significance level of
Men
Women
The results of statistical modeling of gene-gene interactions in nonallergic asthma using set association approach. Significance level of
Men
Women
The MDR method was used for a purpose of modeling gene-gene interactions underlying allergic and nonallergic asthma in men and women. Firstly, we used an exhaustive search algorithm to evaluate all interactions among all possible subsets of the polymorphisms. Table
A summary of best 2-, 3-, and 4-locus models of gene-gene interactions obtained by MDR analysis in allergic and nonallergic asthma (exhaustive search algorithm).
Number of loci | Best |
Cross-validation consistency, % | Prediction error, % |
---|---|---|---|
Allergic asthma (men) | |||
2 |
|
40 | 52.5 |
3 |
|
50 | 50.3 |
4 |
|
30 | 53.6 |
|
|||
Allergic asthma (women) | |||
2 |
|
50 | 45.8 |
3 |
|
50 | 40.9 |
4 |
|
20 | 50.1 |
|
|||
Nonallergic asthma (men) | |||
2 |
|
30 | 52.3 |
3 |
|
20 | 57.0 |
4 |
|
30 | 49.2 |
|
|||
Nonallergic asthma (women) | |||
2 |
|
30 | 46.7 |
3 |
|
60 | 45.5 |
4 |
|
90 | 27.4 |
**A statistically significant (
Figure
Dendrograms of gene-gene interactions in the pathogenetic variants of asthma (MDR method). Dendrograms show both complexity and diversity of interactions between polymorphic genes of antioxidant defense enzymes in allergic and nonallergic asthma (dendrograms are stratified by gender). Each dendrogram comprises a spectrum of lines representing a continuum from synergy (black) to redundancy (gray) of gene-gene interactions. The lines range from bold black, representing a high degree of synergy (positive information gain), thin black, representing a lesser degree, and dotted line representing the midway point between synergy and redundancy. On the redundancy end of the spectrum, the highest degree is represented by bold gray (negative information gain) with a lesser degree represented by thin gray.
On the next step, a forced search algorithm was applied to analyze all possible
Then, we performed a post hoc comparison of genotype frequencies between the case and control groups with a focus on those ADE genes which were present in gene-gene interaction models obtained using SAA and MDR methods. Ten and nine two-locus combinations were found to be associated with allergic asthma in men and women, respectively (Table
Associations of genotype combinations with risk of allergic asthma (stratified by gender).
Combinations of genotypes | Allergic asthma | Controls | Chi-square |
OR (95% CI) | ||
---|---|---|---|---|---|---|
|
% |
|
% | |||
Men | ||||||
|
34 | 53.1 | 38 | 36.2 | 4.66 (0.03) | 2.00 (1.06–3.76) |
|
20 | 31.3 | 19 | 18.1 | 3.88 (0.05) | 2.06 (1.00–4.25) |
|
7 | 10.9 | 2 | 1.9 | 4.77 (0.03) | 5.40 (1.25–23.42) |
|
12 | 18.8 | 6 | 5.7 | 5.80 (0.02) | 3.64 (1.33–9.96) |
|
18 | 28.1 | 13 | 12.4 | 6.58 (0.01) | 2.77 (1.25–6.14) |
|
12 | 18.8 | 7 | 6.7 | 4.67 (0.03) | 3.13 (1.19–8.21) |
|
38 | 59.4 | 44 | 41.9 | 4.86 (0.03) | 2.03 (1.08–3.81) |
|
17 | 26.6 | 11 | 10.5 | 7.44 (0.01) | 3.09 (1.34–7.13) |
|
9 | 14.1 | 3 | 2.9 | 5.97 (0.01) | 5.01 (1.41–17.8) |
|
6 | 9.4 | 28 | 26.7 | 6.36 (0.01) | 0.30 (0.12–0.76) |
|
||||||
Women | ||||||
|
43 | 46.7 | 33 | 30.3 | 5.75 (0.02) | 2.02 (1.13–3.60) |
|
18 | 19.6 | 39 | 35.8 | 6.46 (0.01) | 0.44 (0.23–0.83) |
|
43 | 46.7 | 33 | 30.3 | 5.75 (0.02) | 2.02 (1.13–3.60) |
|
18 | 19.6 | 39 | 35.8 | 6.46 (0.01) | 0.44 (0.23–0.83) |
|
9 | 9.8 | 24 | 22.0 | 4.59 (0.03) | 0.40 (0.18–0.89) |
|
9 | 9.8 | 24 | 22.0 | 4.59 (0.03) | 0.40 (0.18–0.89) |
|
11 | 12.0 | 4 | 3.7 | 3.83 (0.05) | 3.31 (1.07–10.22) |
|
13 | 14.1 | 31 | 28.4 | 5.97 (0.01) | 0.41 (0.20–0.85) |
|
18 | 19.6 | 43 | 39.4 | 9.33 (0.002)* | 0.37 (0.20–0.71) |
Associations of genotype combinations with risk of nonallergic asthma (stratified by gender).
Combinations of genotypes | Nonallergic asthma | Controls | Chi-square ( |
OR (95% CI) | ||
---|---|---|---|---|---|---|
|
% |
|
% | |||
Men | ||||||
|
6 | 20.7 | 2 | 1.9 | 11.13 (0.001)* | 11.45 (2.49–52.66) |
|
4 | 13.8 | 3 | 2.9 | 3.50 (0.05) | 5.17 (1.20–22.31) |
|
5 | 17.2 | 4 | 3.8 | 4.58 (0.03) | 5.06 (1.37–18.99) |
|
6 | 20.7 | 5 | 4.8 | 5.68 (0.02) | 5.05 (1.49–17.14) |
|
5 | 17.2 | 4 | 3.8 | 4.58 (0.03) | 5.06 (1.35–18.99) |
|
11 | 37.9 | 15 | 14.3 | 8.12 (0.004) | 3.67 (1.47–9.28) |
|
8 | 27.6 | 3 | 2.9 | 15.31 (0.0001)* | 11.58 (3.07–43.72) |
|
4 | 13.8 | 1 | 1.0 | 7.16 (0.01) | 12.29 (1.84–82.03) |
|
5 | 17.2 | 1 | 1.0 | 10.55 (0.001)* | 15.64 (2.44–100.3) |
|
4 | 13.8 | 1 | 1.0 | 7.16 (0.01) | 12.29 (1.84–82.03) |
|
12 | 41.4 | 10 | 9.5 | 16.8 (0.00004)* | 6.71 (2.5–17.96) |
|
7 | 24.1 | 5 | 4.8 | 8.22 (0.004) | 6.09 (1.85–20.05) |
|
||||||
Women | ||||||
|
2 | 7.4 | 0 | 0.0 | 3.88 (0.05) | 21.47 (1.00–461.1) |
|
6 | 22.2 | 56 | 51.4 | 6.29 (0.01) | 0.29 (0.11–0.74) |
|
2 | 7.4 | 0 | 0.0 | 3.88 (0.05) | 21.5 (1.00–461.2) |
|
1 | 3.7 | 24 | 22.0 | 3.69 (0.05) | 0.20 (0.04–1.09) |
|
7 | 25.9 | 4 | 3.7 | 11.58 (0.001)* | 8.58 (2.43–30.26) |
The main purpose of our study was to investigate a comprehensive contribution of ADE genes to genetic susceptibility to allergic and nonallergic variants of BA. The single-locus analysis revealed that none of the ADE genes was associated with the risk of asthma. However, using two bioinformatics approaches, we found multilocus gene-gene interactions which are associated with the risk of allergic and nonallergic asthma in men and women in a gender-specific manner. Further, post hoc analysis allowed revealing two-locus combinations of genotypes which are significantly associated with allergic and nonallergic asthma in both sexes. A majority of the susceptibility genes identified in our study represented antioxidant defense enzymes. Moreover, interactions between ADE genes varied across the pathogenetic variants of asthma and were different in men and women suggesting both genetic heterogeneity and gender-specific genetic effects in the disease susceptibility.
The observed differences in gene-gene interactions between allergic and nonallergic variants of asthma demonstrate a genetic heterogeneity of the disease, a situation in which the same or similar phenotype of a complex disorder is caused by different susceptibility genes [
The results of gene-gene interactions analysis are consistent with observations of other genetic studies which demonstrated an importance of ADE genes for asthma pathogenesis. In particular, we confirmed a potential role in the pathogenesis of asthma for
Comparing the results obtained by the three statistical approaches to the analysis of gene-gene interactions, we can say that, despite gender-specific effects of genotypes on the pathogenetic variants of the disease, each of the methods showed own uniqueness and efficacy in the detecting genes associated with asthma risk. In our point of view, the advantage of SAA method is its capacity in the identification of “gene dosage effects” of different sets of ADE genes on asthmatic phenotype. Meanwhile, MDR method, especially its cluster technique, was found to be powerful in the detecting high-order epistatic interactions between ADE genes and their synergic and antagonistic effects on the asthma risk. The variability in the structure of gene-gene interactions models across the pathogenetic variants of asthma can be partially explained by differences in bioinformatic approaches to the analysis of multiple genes. Apparently, a similarity in gene-gene interactions between the models obtained by the two different bioinformatic tools may be explained by strong effects of particular genes on the asthmatic phenotype. This means that strong phenotypic effects of the
The dendrograms obtained by MDR technique (Figure
A post hoc comparative analysis of the frequencies of genotype combinations in the study groups revealed two-locus combinations of the ADE genotypes which increase the risk of the development of asthma. We found relatively rare combinations of genotypes which gave the highest asthma risk estimates but were limited to small subgroups of subjects. In particular, frequencies of these genotype combinations varied from 1 to 9% among healthy controls and from 17 to 41% among patients with nonallergic asthma, whereas odds ratios for disease risk varied from 6.7 to 15.6. Moreover, there was an obvious excess of combinations of variant genotypes among asthma patients compared with healthy subjects, and these differences reached statistical significance after adjusting for multiple tests.
Despite nonsignificant differences in the genotype distributions, the two-locus comparison of genotype frequencies between the study groups has shown that asthma patients more often than healthy subjects carry combinations of the genotypes which are known to determine a diminished activity of ADE towards ROS. This is supported by a number biochemical studies that observed massive generation of ROS and the insufficiency in antioxidant capacity in asthma [
Based on the literature data demonstrating biochemical abnormalities in redox homeostasis in asthma and the results of our study, we assumed possible relationships between these abnormalities and ADE genes showed the associations with asthma in our study (the data are shown in Table
Common biochemical abnormalities in redox homeostasis found in asthma and their possible relationship with genes for antioxidant defense enzymes which have been associated with risk of the disease in the present study.
Biochemical abnormalities in asthmatics |
ADE gene related with the abnormality | Allergic asthma | Nonallergic asthma | ||
---|---|---|---|---|---|
Men | Women | Men | Women | ||
Diminished capacity of glutathione peroxidases and catalase in detoxification of hydrogen peroxide [ |
|
+ + + | + + | + + | + + |
|
+ + | − | + + + | ||
|
− | − | + + | − | |
|
− | + | − | − | |
|
+ + + | − | + + | − | |
|
|||||
An enhanced production of ROS/hydrogen peroxide/superoxide anion radicals [ |
|
− | + | − | − |
|
− | − | + + + | − | |
|
|||||
Perturbations in glutathione (GSH) homeostasis [ |
|
+ + | − | + + + | − |
|
+ + | − | + + + | − | |
|
|||||
Increased EPHX1 activity, increased production of xenobiotics-generated epoxides, |
|
+ + | + + + | + + | + + + |
The number of pluses means a degree of the relationship between the gene and asthma risk. These measures reflect how many times a particular gene showed the link with asthma risk through the three methods used for evaluation of gene-gene interactions in the present study, namely, set association approach (SAA), multifactor dimensionality reduction (MDR) method, and post hoc association analysis of two-locus genotype combinations (AAGC): + + + means that the link was found thrice (i.e., using SAA, MDR, and AAGC methods); + + means that the link was found twice (i.e., using SAA or MDR and AAGC methods); + means that the link was found once by AAGC method. Associations are stratified by asthma type and gender.
Although we did not perform biochemical investigations of antioxidant status, taking the observed association of asthma with the ADE genotypes and their functional significance into account, it is likely that an imbalance between oxidants and antioxidants detected in asthma can be directly related to genetically diminished capacity of ADE. Such an imbalance results in oxidative stress caused by an excessive production of ROS and/or by inadequate antioxidant defense leading to damage of airway epithelial cells and inflammation due to upregulation of redox-sensitive transcription factors and proinflammatory genes [
An important finding of our study was that polymorphisms of many ADE genes showed sex-specific associations with the development of asthma. For instance, the
The study has limitations. Due to the relatively small sample sizes of the studied groups, the association analysis of two-locus genotype combinations was underpowered, especially after Bonferroni adjustment for multiple tests. Because of the limited sample size, we also cannot exclude the possibility that small effects of some ADE genes were not detected. Since BA is a multifactorial and genetically heterogeneous disease [
To the best of our knowledge, this is the first study investigating the associations between BA and 34 functionally significant polymorphic variants of ADE genes and 12 other candidate genes. So far, no genetic studies have reported a comprehensive evaluation of asthma susceptibility with a number of ADE genes at once. Methodological approaches used in this study were proved fruitful in uncovering the genetic architecture of complex interactions between genes involved in the regulation of redox homeostasis. This allowed finding for the first time that antioxidant defense enzymes genes are collectively involved in the molecular mechanisms of BA and can explain genetic heterogeneity between allergic and nonallergic variants of the disease. In particular, we found for the first time that the
Further studies focusing on the molecular mechanisms regulating redox homeostasis can provide more complete understanding of the role of the ADE genes in bronchial asthma and end up in the discovery of new drug targets for antioxidant treatment and prevention of the disease.
The authors declare that there is no conflict of interests regarding the publication of this paper.
The study was supported in part by the Federal Targeted Program “Scientific and Scientific-Pedagogical Personnel of the Innovative Russia in 2009–2013.” The authors thank Drs. Mikhail Kozhuhov and Valery Panfilov from Kursk Regional Clinical Hospital for their invaluable help in recruiting and examining asthma patients for the study.