Updates on the COPD gene list.

A genetic contribution to develop chronic obstructive pulmonary disease (COPD) is well established. However, the specific genes responsible for enhanced risk or host differences in susceptibility to smoke exposure remain poorly understood. The goal of this review is to provide a comprehensive literature overview on the genetics of COPD, highlight the most promising findings during the last few years, and ultimately provide an updated COPD gene list. Candidate gene studies on COPD and related phenotypes indexed in PubMed before January 5, 2012 are tabulated. An exhaustive list of publications for any given gene was looked for. This well-documented COPD candidate-gene list is expected to serve many purposes for future replication studies and meta-analyses as well as for reanalyzing collected genomic data in the field. In addition, this review summarizes recent genetic loci identified by genome-wide association studies on COPD, lung function, and related complications. Assembling resources, integrative genomic approaches, and large sample sizes of well-phenotyped subjects is part of the path forward to elucidate the genetic basis of this debilitating disease.

Introduction

Chronic obstructive pulmonary disease (COPD) is the third-leading cause of worldwide mortality and is predicted to remain a major public health problem in the near future.1,2 It is characterized by airflow limitations that occur in approximately 10% of adults aged ≥ 40 years.3 Cigarette smoking is the primary risk factor. However, only a fraction of smokers (~20%) develop the disease, and host differences in susceptibility are thus persuasive. The author has previously reviewed the genetics of COPD and COPD-related phenotypes.4 The current review aims to: (1) update this publication, (2) provide a comprehensive literature overview on the genetics of COPD, (3) highlight the most promising findings during the last few years, and ultimately (4) provide an updated COPD gene list.

Chronic obstructive pulmonary disease candidate-gene studies

A systematic review of the literature was conducted in order to provide a comprehensive overview of genes associated with COPD and related phenotypes. PubMed was searched using the string “genetics and COPD” on January 5, 2012. All titles and abstracts were reviewed for inclusion. The goal was to obtain all publications testing genetic variants in humans for association with COPD and related phenotypes (ie, spirometric measurements, emphysema, chronic bronchitis, lung-function decline, etc). Population-based, case-control, and family studies were included. The author attempted to include all reported articles without quality assessment or exclusion criteria based on sample size or other criteria. The search for relevant publications was complemented using the list of references in relevant manuscripts and the COPD genetic association compendium.5 Readers are welcome to contact the author for any articles missed in the current review.

A large number of candidate gene–association studies were conducted to identify the COPD-susceptibility genes. Table 1 provides a comprehensive overview of the genes associated with COPD and related phenotypes using this genetic approach. Supplementary Table 1 presents additional genes tested but showing lack of association with COPD and related phenotypes. Most genes in these tables were studied because of their potential role in the pathobiology of COPD, but some also represent follow-up genes originally identified from genome-wide linkage and association studies. Genes are presented in alphabetical order. Single studies and metaanalyses testing each gene are indicated. An attempt was made to classify each article as supportive or not of a given gene based on the conclusions provided by the authors. Single genetic markers, haplotypes, or combinations of variants associated with COPD, COPD severity, COPD-related phenotypes, or complications were considered as positives. Table 1 aims to provide an exhaustive list of publications for any given gene.

Table 1

List of genes associated with chronic obstructive pulmonary disease

Symbol	Name	Chromosome	References
			Single studies		Meta-analyses
			Positive	Negative	Positive	Negative
A2M	Alpha-2-macroglobulin	12	51
ABCC1	ATP-binding cassette, sub-family C (CFTR/MRP), member 1	16	52–54
ACE	Angiotensin I converting enzyme (peptidyl-dipeptidase A) 1	17	55–60	61,62		5
ADAM33	ADAM metallopeptidase domain 33	20	63–68	69,70
ADRB2	Adrenergic, beta-2-, receptor, surface	5	71–82	83		5,83
ALOX5AP	Arachidonate 5-lipoxygenase-activating protein	13	84
AQP5	Aquaporin 5	12	85,86
BCL2	B-cell CLL/lymphoma 2	18	87
BDKRB2	Bradykinin receptor B2	14	88
CASP10	Caspase 10, apoptosis-related cysteine peptidase	2	89
CAT	Catalase	11	90	91,92
CCL5 (RANTES)	Chemokine (C-C motif) ligand 5	17	93	79
CCR2	Chemokine (C-C motif) receptor 2	3	94
CD14	CD14 molecule	5	95,96
CD40	CD40 molecule, TNF receptor superfamily member 5	20	97
CD63	CD63 molecule	12	98
CD86	CD86 molecule	3	99
CDC6	Cell division cycle 6 homolog (S cerevisiae)	17	100
CDKN1A (p21)	Cyclin-dependent kinase inhibitor 1A (p21, Cip1)	6	101
CFTR	Cystic fibrosis transmembrane conductance regulator (ATP-binding cassette sub-family C, member 7)	7	102–108	109,110
CHI3L1	Chitinase 3-like 1 (cartilage glycoprotein-39)	1	111
CHRNA3	Cholinergic receptor, nicotinic, alpha 3 (neuronal)	15	26,30,31, 112,113
CHRNA5	Cholinergic receptor, nicotinic, alpha 5 (neuronal)	15	26,30,31, 112,113
CLCA1	Chloride channel accessory 1	1	114
COL4A3	Collagen, type IV, alpha 3 (Goodpasture antigen)	2	115
CRP	C-reactive protein, pentraxin-related	1	116	117–119
CSF2	Colony stimulating factor 2 (granulocyte-macrophage)	5	120	121
CSF3	Colony stimulating factor 3 (granulocyte)	17	121
CTLA4	Cytotoxic T-lymphocyte-associated protein 4	2	99,122,123
CTSS	Cathepsin S	1	124
CYBA	Cytochrome b-245, alpha polypeptide	16	125
CYP1A1	Cytochrome P450, family 1, subfamily A, polypeptide 1	15	125–128	129,130
CYP1A2	Cytochrome P450, family 1, subfamily A, polypeptide 2	15	129,131	125,128
CYP2E1	Cytochrome P450, family 2, subfamily E, polypeptide 1	10	127,132	130
CYP2F1	Cytochrome P450, family 2, subfamily F, polypeptide 1	19	133
CYP3A5	Cytochrome P450, family 3, subfamily A, polypeptide 5	7	134
DEFB1	Defensin, beta 1	8	135,136	137
DEFB4A	Defensin, beta 4A	8	138
EDN1	Endothelin 1	6	139–141	142,143
EDNRB	Endothelin receptor type B	13	143
ELN	Elastin (supravalvular aortic stenosis, Williams–Beuren syndrome)	7	144,145	146,147
EPHX1	Epoxide hydrolase 1, microsomal (xenobiotic)	1	77,83,130, 146–167	127,168–174	175	5,8,176
ESR1	Estrogen receptor 1	6	177
FAM13A	Family with sequence similarity 13, member A	4	26
FGF10	Fibroblast growth factor 10	5	178
GC	Group-specific component (vitamin D binding protein)	4	179–186	146,147,151, 155,187
GCLC	Glutamate-cysteine ligase, catalytic subunit	6	188	172,189
GCLM	Glutamate-cysteine ligase, modifier subunit	1	190	172,188
GSTCD	Glutathione S-transferase, C-terminal domain containing	4	191
GSTM1	Glutathione S-transferase M1	1	127,148,161, 164,165, 192–202	90,130,146,147, 151,169,203–206	5,7,8
GSTO1	Glutathione S-transferase omega 1	10	207
GSTO2	Glutathione S-transferase omega 2	10	207
GSTP1	Glutathione S-transferase pi 1	11	77,90,146, 148,151,152, 157,164,165, 193,194,196, 204,208–210	69,127,130,147, 149,159,171,185, 197,203,211,212	8,213	5,214
GSTT1	Glutathione S-transferase theta 1	22	127,165,193, 196–198, 204–206	90,130,148,161, 164,169,194, 199–201,203		5,7,8
HCK	Hemopoietic cell kinase	20	215
HHIP	Hedgehog interacting protein	4	26,191,216
HLA	Classical class 11 subregion of the MHC	6	217,218	219,220
HMOX1	Heme oxygenase (decycling) 1	22	130,151,166, 221–224	69,147,185, 196,225
HTR4	5-hydroxytryptamine (serotonin) receptor 4	5	191
IFNG	Interferon, gamma	12	226–228
IL1A	Interleukin 1, alpha	2	227
IL1B	Interleukin 1, beta	2	227,229–233	120,228, 234–238		5,8
IL1RN	Interleukin 1 receptor antagonist	2	231,232, 234,235	228,230, 236–238	8
IL2	Interleukin 2	4	227
IL27	Interleukin 27	16	239
IL4	Interleukin 4	5	71,227,240	120,241,242		5
IL4R	Interleukin 4 receptor	16	227,243	79,241
IL5	Interleukin 5 (colony-stimulating factor, eosinophil)	5	244
IL6	Interleukin 6	7	118,228,234, 245–247	116,233, 236,248		5,8
IL8	Interleukin 8	4	120	234,235,238, 249,250
IL8RA	Interleukin 8 receptor, alpha	2	251	120,146,147
IL8RB (CXCR2)	Interleukin 8 receptor, beta	2	250	120,146,147
IL10	Interleukin 10	1	149,227,235, 248,252–254	120,234,255
IL12B	Interleukin 12B (natural killer cell stimulatory factor 2, cytotoxic lymphocyte maturation factor 2, p40)	5	227	239
IL13	Interleukin 13	5	79,241,242, 256–261	71,120,238, 243,262		5
IL13RA1	Interleukin 13 receptor, alpha 1	X	241
IL17F	Interleukin 17F	6	263
IREB2	Iron-responsive element binding protein 2	15	26,30,47
KCNMB1	Potassium large conductance calcium-activated channel, subfamily M, beta member 1	5	264
KEAP1	Kelch-like ECH-associated protein 1	19	265
LEP	Leptin	7	266
LEPR	Leptin receptor	1	267
LTA	Lymphotoxin alpha (TNF superfamily, member 1)	6	234,268–272	120,233,248, 273–275		5
LTA4H	Leukotriene A4 hydrolase	12	84
LTBP4	Latent transforming growth factor beta binding protein 4	19	146,147
MBL2	Mannose-binding lectin (protein C) 2, soluble	10	276,277
MICB	MHC class I polypeptide-related sequence B	6	278
MIR196A2	MicroRNA 196a-2	12	279
MIR499A	MicroRNA 499a	20	279
MMP1	Matrix metallopeptidase 1 (interstitial collagenase)	11	146,280,281	69,128,147, 151,282–285
MMP2	Matrix metallopeptidase 2 (gelatinase A, 72 kDa gelatinase, 72 kDa type IV collagenase)	16	285	69,281
MMP3	Matrix metallopeptidase 3 (stromelysin 1, progelatinase)	11	286	128,287
MMP9	Matrix metallopeptidase 9 (gelatinase B, 92 kDa gelatinase, 92 kDa type IV collagenase)	20	128,202,281, 282,284, 288–290	69,147,151, 280,283,285,287		5,8
MMP12	Matrix metallopeptidase 12 (macrophage elastase)	11	280,283, 291,292	69,146,147, 282,284, 285,287
MMP14	Matrix metallopeptidase 14 (membrane-inserted)	14	293
MSR1	Macrophage scavenger receptor 1	8	137,294
NAT2	N-acetyltransferase 2 (arylamine N-acetyltransferase)	8	132
NFE2L2	Nuclear factor (erythroid-derived 2)-like 2	2	265,295
NFKBIB	Nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, beta	19	185
NOS3	Nitric oxide synthase 3 (endothelial cell)	7	57,62, 296,297	149
NQO1	NAD(P)H dehydrogenase, quinone 1	16	90
NR3C1	Nuclear receptor subfamily 3, group C, member 1 (glucocorticoid receptor)	5	298	299
OGG1	8-oxoguanine DNA glycosylase	3	300	189
OR4X1	Olfactory receptor, family 4, subfamily X, member 1	11	301
PDE4D	Phosphodiesterase 4D, cAMP-specific (phosphodiesterase E3 dunce homolog, drosophila)	5	302
PLAUR	Plasminogen activator, urokinase receptor	19	303,304
PPARG	Peroxisome proliferator-activated receptor gamma	3	163
PTEN	Phosphatase and tensin homolog	10	14
PTGDR	Prostaglandin D2 receptor (DP)	14	305
PTGS2 (COX2)	Prostaglandin-endoperoxide synthase 2 (prostaglandin G/H synthase and cyclooxygenase)	1	306,307
SERPINA1	Serpin peptidase inhibitor, clade A (alpha-1 antiproteinase, antitrypsin), member 1	14	76,308–325	326–336
SERPINA3	Serpin peptidase inhibitor, clade A (alpha-1 antiproteinase, antitrypsin), member 3	14	337–343	146,147,149, 151,310,314, 326,332,344		5
SERPINE2	Serpin peptidase inhibitor, clade E (nexin, plasminogen activator inhibitor type 1), member 2	2	77,146,149, 326,345–348	147,152,171, 349,350
SFTPA1	Surfactant protein A1	10	69,351
SFTPA2	Surfactant protein A2	10	69
SFTPB	Surfactant protein B	2	147,151,171, 351–354	69,77,146, 149,152,355
SFTPC	Surfactant protein C	8	356	357
SFTPD	Surfactant protein D	10	69,358,359	151,351
SIRT2	Sirtuin 2	19	185
SLC6A4	Solute carrier family 6 (neurotransmitter transporter, serotonin), member 4	17	360
SLC11A1	Solute carrier family 11 (proton-coupled divalent metal ion transporters), member 1	2	361
SMAD3	SMAD family member 3	15	362
SMOC2	SPARC related modular calcium binding 2	6	363
SOD2	Superoxide dismutase 2, mitochondrial	6	364–366	91,92,271		8
SOD3	Superoxide dismutase 3, extracellular	4	90,91,364, 367–370		5	8
SOX5	SRY (sex determining region Y)-box 5	12	371
STAT1	Signal transducer and activator of transcription 1, 91 kDa	2	185
STAT3	Signal transducer and activator of transcription 3 (acute-phase response factor)	17	372
STAT6	Signal transducer and activator of transcription 6, interleukin-4 induced	12	79	241
STIP1	Stress-induced-phosphoprotein 1	11	373
TBXA2R	Thromboxane A2 receptor	19	244,374
TGFB1	Transforming growth factor, beta 1	19	69,77,146, 147,238, 375–382	30,149,151, 171,383	5,8	384
TGFBR3	Transforming growth factor, beta receptor III	1	190
TIMP1	TIMP metallopeptidase inhibitor 1	X	285	69
TIMP2	TIMP metallopeptidase inhibitor 2	17	146,385,386	147,151,387		5
TLR4	Toll-like receptor 4	9	388,389	96,271
TNF	Tumor necrosis factor (TNF superfamily, member 2)	6	11,149,151, 234,238,250, 262,268, 270–272, 390–398	83,120,146, 147,155,230, 233,235–237, 248,269, 273–275, 399–403	5,8–11
TNS1	Tensin 1	2	191
TP53 (p53)	Tumor protein p53	17	101,307
TRPV4	Transient receptor potential cation channel, subfamily V, member 4	12	404
TSLP	Thymic stromal lymphopoietin	5	405
VDR	Vitamin D (1,25-dihydroxyvitamin D3) receptor	12	406–408	409
VEGFA	Vascular endothelial growth factor A	6	410	411
XRCC1	X-ray repair complementing defective repair in Chinese hamster cells 1	19	300
XRCC5	X-ray repair complementing defective repair in Chinese hamster cells 5 (double-strand-break rejoining)	2	412

A total of 192 genes are summarized in Table 1 and Supplementary Table 1. Figure 1 illustrates these genes based on the number of publications supporting the association with COPD phenotypes. Briefly, 86 genes are supported by one study, 36 genes by two to five studies, 15 genes by six to ten studies, and seven genes by more than ten studies. The latter seven genes include ADRB2, TGFB1, TNF, GSTM1, GSTP1, SERPINA1, and EPHX1. Note that Figure 1 must be interpreted with caution. Replication of genotype–phenotype associations is the gold standard to identify genes conferring susceptibility.6 However, the number of supportive studies is not necessarily an indication that a gene is consistently replicated. Figure 2 illustrates the relationship between the number of studies supporting and not supporting the list of COPD genes. It seems that genes replicated many times in COPD are simply the most popular genes studied. For example, the author found 20 studies supporting TNF as a COPD-susceptibility gene. However, lack of association between this gene and COPD phenotypes was found in 20 other studies (Table 1). Considering publication bias, candidate genes associated with COPD are not consistently replicated and the overall results are rather inconclusive. In fact, excluding SERPINA1 (encoding the alpha-1 antitrypsin protein), none of the other genes are well-proven susceptibility genes for COPD. Perhaps the most convincing candidate COPD genes up to now are those less studied but consistently replicated, such as SOD3. Many of the most studied COPD genes have now been investigated in meta-analyses.

Figure 1

Candidate genes associated with chronic obstructive pulmonary disease (COPD) or related phenotypes.

Notes: The upper part shows a histogram of the number of COPD susceptibility genes based on the number of publications supporting a significant genetic association. The lower part shows the corresponding genes in each bar. Official gene symbols are indicated. The number of publications that are supportive is indicated in parentheses. References are provided in Table 1 for genes supported by at least one publication and in Supplementary Table 1 for genes tested but not supported.

Figure 2

Scatter plot showing the number of studies supporting and not supporting candidate genes for chronic obstructive pulmonary disease.

Notes: A total of 192 genes are illustrated. Note that many genes overlap in the lower-left corner and the 192 dots cannot be visualized on this display. The gray and red lines are the regression and identity lines, respectively. Genes studied many times or more consistently replicated are illustrated.

Meta-analyses

A number of meta-analyses have been conducted to identify genes robustly associated with COPD and lung function. So far, meta-analyses have been conducted for genes involved in the following pathways: inflammation (IL4, IL6, IL13, IL1B, IL1RN, LTA, TNF, and TGFB1), protease/antiprotease (MMP9, TIMP2, and SERPINA3), oxidative stress (GSTM1, GSTP1, GSTT1, EPHX1, SOD2, and SOD3), and others (ACE and ADRB2). These studies and their main outcomes are summarized by gene in Table 1. Among these genes, GSTM1 was consistently associated with COPD in more than one meta-analysis.5,7,8 This is also true for TNF, but only in Asian populations.5,8–11 In contrast, other genes have not been supported in meta-analyses conducted so far, including GSTT1,5,7,8 IL1B,5,8 IL6,5,8 and MMP9.5,8 The other genes considered in meta-analyses were either reported in only one study or showed conflicting results across studies (Table 1).

As genetic data accumulates, more genes and polymorphisms will be considered in meta-analyses. Combining the findings of an increasing number of studies will allow pooled analyses in more homogenous subgroups based on ethnicity, smoking history, emphysema vs airway type of COPD, and others. These subgroup analyses are likely to be important in finding susceptibility genes for COPD. Ongoing activities gathering genetic data in the field of COPD are important. For example, a web application summarizing candidate-gene studies was recently established.5 At the time of publication, this database included 108 genetic-association studies, including population-based and case-control studies but excluding family-based studies. Seventy-two genes were studied, focusing strictly on single-marker biallelic polymorphisms. A total of 27 genetic variants were found to be reported in three or more independent study populations and summarized into a meta-analysis. Four genes were found to carry a single genetic variant significantly associated with COPD, being GSTM1, TGFB1, TNF, and SOD3. It should be noted that this COPD genetic-association compendium has not been updated since April 2010 and does not included more recent genetic studies on COPD. Updating this type of resource is important to draw reliable conclusions about the contribution of genes. The number of studies for most COPD-susceptibility genes is currently insufficient to reach firm conclusions.

Multi-gene-association studies

A systematic replication study of genes associated with lung function was recently conducted in the SpiroMeta Consortium.12 A literature search identified 104 publications reporting a positive association with lung-function traits in the general populations of diverse origins or in cohorts of patients with respiratory diseases. A total of 130 genes and 48 intergenic regions were studied in 20,288 individuals. Among the 16,936 genotyped or imputed single-nucleotide polymorphisms (SNPs) in these loci, none was significantly associated with forced expiratory volume in one second (FEV1) or FEV1/forced vital capacity (FVC) ratio after correction for multiple testing. The strongest genetic association signals with FEV1 were observed in ever-smokers in the SERPINA1 and PDE4D genes.

Smaller-scale studies testing multiple genes were also conducted in China. First, 170 asthmatic cases and 347 controls were evaluated for 119 SNPs in 98 genes for association with lung function.13 After correction for multiple testing, none of the SNPs was significantly associated with lung function (ie, FEV1, FVC, or FEV1/FVC). The strongest association was observed between rs320995 (Phe309Phe) in CYSLTR1 and FEV1/FVC (P = 0.0004). Second, 1,261 SNPs in 380 candidate genes for cancer or other human diseases were tested for association with COPD in 53 cases and 107 controls with in-home coal exposure.14 A total of 22 genes were associated with COPD risk, but only PTEN was significant after correction for multiple testing. Considering the small sample sizes, the results of these studies must be replicated before reaching firm conclusions.

Genome-wide association studies on COPD

Table 2 summarizes COPD susceptibility loci identified by genome-wide association (GWA) studies. The results of the first GWA study on COPD were published in 2009.15 The GWA study was conducted in a case-control cohort of Norway (823 COPD cases and 810 controls), and the top 100 SNPs were followed up in the family-based International COPD Genetics Network (ICGN). Two susceptibility loci were identified. The most definitive evidence of association was found with two SNPs at the α-nicotinic acetylcholine receptor locus on chromosome 15q25, the same locus implicated in the risk of lung cancer.16–18 Two SNPs at the hedgehog interacting protein (HHIP) locus on chromosome 4q31 also showed strong associations.

Table 2

Susceptibility loci for chronic obstructive pulmonary disease (COPD) and related phenotypes identified by genome-wide association studies

Reference	Study*	Sample size (cases/controls)	Disease/trait	Platform (# SNPs)	Region (size)	Gene	Key SNPs
Pillai et al15	Norway	823/810	COPD	Illumina (Human Hap550)	15q25	CHRNA3	rs8034191
	ICGN	1891				CHRNA5	rs1051730
	NETT-NAS	389/472
	EOCOPD	949
					4q31	HHIP	rs1828591
							rs13118928
Cho et al19	Norway	2940/1380	COPD	Illumina (Human Hap550 or Quad610)	4q22	FAM13A	rs7671167
	NETT-NAS						rs1903003
	ECLIPSE
	COPDGene	502/504
	EOCOPD	949
	ICGN	2859
					15q25	CHRNA3	rs1062980
						CHRNA5
						IREB2
					4q31	HHIP	rs1828591
Cho et al20	ECLIPSE	1764/178	COPD	Illumina (Human Hap550, Quad610, or Omni1 Quad)	19q13	RAB4B	rs7937
	NAS-NETT	373/435				EGLN2	rs2604894
	GenKOLS	863/808				MIA
	COPDGene	499/501				CYP2A6
	ICGN	983 probands/ 1876 siblings
					4q22	FAM13A	rs1964516
							rs7671167
					4q31	HHIP	rs13141641
							rs13118928
					15q25	CHRNA3	rs11858836
						CHRNA5	rs13180
						IREB2
Wilk et al37	FHS	1059–1222	Ten spirometry phenotypes	Affymetrix (70,987)	10q25	GSTO2	rs156697
Wilk et al38	FHS	7691	FEV1/FVC	Affymetrix (500 K + 50 K)	4q31	HHIP	rs13147758
	Family heart study	835
Repapi et al40	SpiroMeta Consortium	20,288	FEV1 and FEV1/FVC	Affymetrix and Illumina (2.5 million)	4q31	HHIP	rs12504628
	CHARGE consortium	32,184
		21,209
	Health 2000 survey	883
			FEV1		2q35	TNS1	rs2571445
					4q24	GSTCD	rs10516526
					5q33	HTR4	rs3995090
			FEV1/FVC		6p21	AGER	rs2070600
					15q23	THSD4	rs12899618
Hancock et al39	CHARGE Consortium	20,890	FEV1/FVC	Affymetrix and Illumina (2,515,866)	2q36	PID1	rs1435867
	SpiroMeta consortium	16,178
					4q22	FAM13A	rs2869967
					4q31	HHIP	rs1980057
					5q33	HTR4	rs11168048
					5q33	ADAM19	rs2277027
					6p21	AGER-PPT2	rs2070600
					6q24	GPR126	rs3817928
					9q22	PTCH1	rs16909898
			FEV1		4q24	INTS12	rs17331332
						GSTCD
						NPNT
Soler Artigas et al41,**	23 studies	48,201	FEV1	Illumina and Affymetrix (2,706,349)	3q26	MECOM	rs134555
	17 studies	46,411
					6p22	ZKSCAN3	rs6903823
					10q22	C10orf11	rs11001819
			FEV1/FVC		1p36	MFAP2	rs2284746
					1q41	TGFB2- LYPLAL1	rs993925
					2q37	HDAC4- FLJ43879	rs12477314
					3p24	RARB	rs1529672
					5q15	SPATA9- RHOBTB3	rs153916
					6q21	ARMC2	rs2798641
					6p21	NCR3-AIF1	rs2857595
					12q13	LRP1	rs11172113
					12q22	CCDC38	rs1036429
					16q13	MMP15	rs12447804
					16q23	CFDP1	rs2865531
					21q22	KCNE2- LINC00310	rs9978142
			FEV1 and FEV1/FVC		10p23	CDC123	rs7068966
Imboden et al42	SAPALDIA	2677 nonasthmatic, 1441 asthmatic	FEV1 decline in nonasthmatic	Illumina	13q14	DLEU7	rs9316500
	ECRHS			Human
	EGEA			610quad
	FHS	10,858 nonasthmatic, 1138 asthmatic
	ARIC
	B58C
	Dutch asthma study
			FEV1/FVC decline in asthmatic		8p22	TUSC3	rs4831760
Kong et al43	ECLIPSE	1557	Emphysema (qualitative)	Illumina	12q11	BICD1	rs10844154
	Norway	432		Human			rs161976
				Hap550 (499,578)
Wan et al44	ECLIPSE	1734	Cachexia-related phenotypes (BMI and fat-free mass index)	Illumina	16q12	FTO	rs8050136
	Norway	851
	NETT	365
	COPDGene	502

Notes:

Bold entries indicates replication cohorts;

only the new loci are identified for this study, but ten loci previously reported by Hancock et al39 and Repapi et al40 were also detected.

Abbreviations: ARIC, Atherosclerosis Risk in Communities; B58C, British 1958 Birth Cohort; EOCOPD, Boston Early-Onset COPD Study; BMI, body mass index; COPDGene, COPDGene study; ECLIPSE, Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints; ECRHS, European Community Respiratory Health Survey; EGEA, Epidemiological study on the Genetics and Environment of Asthma; FEV1, forced expiratory volume in 1 second; FHS, Framingham Heart Study; FVC, forced vital capacity; GenKOLS, Bergen, Norway COPD Cohort; ICGN, International COPD Genetics Network study; NAS-NETT, Normative Aging Study and National Emphysema Treatment Trial; SAPALDIA, Swiss Cohort Study on Air Pollution and Lung and Heart Disease in Adults; SNPs, single-nucleotide polymorphisms.

The case-control cohort of Norway was then combined with the COPD cases from the National Emphysema Treatment Trial (NETT) and unaffected individuals from the Normative Aging Study (NAS), as well as cases and controls from the multicenter Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) Study.19 A total of 2940 cases and 1380 controls were considered. Loci 15q25-CHRNA3/CHRNA5/IREB2 and 4q31-HHIP were replicated in this study. A third locus was also identified at 4q22.1 harboring the FAM13A gene. The latter was followed up and validated in the COPDGene study and the ICGN. A trend was also observed in the Boston Early-Onset COPD Study (EOCOPD). The latest GWA study on COPD was performed using 3499 cases and 1922 controls regrouping the ECLIPSE, NETT-NAS, Norway, and COPDGene studies. 20 The three GWA-nominated COPD-susceptibility loci (ie, CHRNA3/CHRNA5/IREB2, HHIP, and FAM13A) were confirmed in this extended GWA study. In addition, a new COPD locus was identified on chromosome 19q13, which harbored the RAB4B, EGLN2, MIA, and CYP2A6 genes. It was estimated that the four GWA-nominated COPD loci accounted for ~5% of the total variance of the sibling relative risk of COPD.20

Two of the four genome-wide associated loci found in COPD – 15q25 and 19q13 – were previously associated with cigarettes smoked per day and cotinine levels,21–25 suggesting that the risk alleles are acting through smoking behavior. Further studies support this hypothesis on 15q25. In fact, previous studies suggested that sequence variants on chromosome 15q25 confer risk of smoking-related lung diseases (ie, COPD and lung cancer) through its effect on tobacco addiction.17,26 This is consistent with the lack of association between the 15q25 locus and lung cancer among never-smokers.27–29 In contrast, other evidence argues against this hypothesis, showing weak or no evidence that the 15q25 locus directly influences smoking behavior,15,16 no appreciable variation in the risk of lung cancer across smoking categories,18 and significant effect of the 15q25 locus on smoking-related diseases after adjustment for smoking exposure.30,31 Multiple distinct loci affecting both smoking behavior24,31 and lung cancer32 were reported on 15q25. It is still unknown whether genes located at any of these loci are causally involved in the pathogenesis of COPD and lung cancer or the effect is mediated by changing smoking behavior. Risk alleles on chromosome 15q25 were shown to modulate the mRNA expression levels of the CHRNA5 gene in the brain33,34 and lung35 tissues as well as the expression of CHRNA5 and IREB2 genes in sputum.36 The regulation of genes in primary disease tissues, such as lung and sputum, suggests a direct effect of 15q25 genes on COPD susceptibility. More functional studies are needed to find the causal alleles and genes on 15q25 as well as to disentangle their impact on correlated traits associated with this chromosomal region.

GWA studies on lung function

In 2007, Wilk et al37 reported the first GWA study on lung function in approximately 1200 individuals. The study was conducted as part of the Framingham Heart Study. Association tests were performed on 70,987 autosomal SNPs and for ten spirometry phenotypes. No SNP was associated with lung-function phenotypes using stringent criteria for genome-wide significance, but suggestive evidence of association was provided for a nonsynonymous coding SNP in exon 5 of the GSTO2 gene. In 2009, a larger GWA study from the Framingham Heart Study was performed in 7691 participants.38 Interestingly, the 4q31-HHIP COPD locus was associated with percent predicted FEV1/FVC ratio. This locus was confirmed in a second set of participants from the Family Heart Study (n = 835).

In January 2010, two articles reported GWA studies for lung function.39,40 First, Repapi et al40 performed a GWA study on FEV1 and FEV1/FVC ratio in the SpiroMeta consortium (20,288 individuals of European ancestry). They have also followed up the best associated SNPs in 32,184 additional individuals. Overall, they have identified five novel genome-wide significant loci for pulmonary function, being 2q35 (TNS1), 4q24 (GSTCD), and 5q33 (HTR4) for FEV1, and 6p21 (AGER) and 15q23 (THSD4) for FEV1/FVC. Second, Hancock et al39 conducted a GWA study on the same two clinically important pulmonary function measures in the CHARGE consortium consisting of 20,890 participants of European ancestry. They identified significant associations with FEV1/FVC ratio for SNPs located in seven previously unrecognized loci: 6q24 (GPR126), 5q33 (ADAM19), 6p21 (AGER and PPT2), 4q22 (FAM13A), 9q22 (PTCH1), 2q36 (PID1), and 5q33 (HTR4). For FEV1, one new locus annotated by three genes (INTS12, GSTCD, and NPNT) on 4q24 was identified. 4q24 (GSTCD), 5q33 (HTR4) and 6p21 (AGER) were common in both consortia, ie, SpiroMeta and CHARGE. The previously reported 4q31 locus located upstream of the HHIP gene associated with FEV1 and FEV1/ FVC ratio was also confirmed in these consortia.

More recently, a larger GWA study of FEV1 and FEV1/FVC ratio was reported, comprising more than 48,000 individuals of European ancestry and followed up for replication in more than 46,000 individuals.41 Ten out of eleven loci previously reported by the SpiroMeta and CHARGE consortia were replicated in this extended GWA study. Only PID1 on 2q36 was not replicated. More interestingly, 16 new loci were identified, including twelve loci for FEV1/FVC, three for FEV1, and one for both traits. Thus, 26 loci were associated with lung function in this GWA study. Together, these loci explain 3.2% of the additive polygenic variance for FEV1/ FVC and 1.5% of the variance for FEV1.

The first GWA study on lung-function decline was recently reported.42 Briefly, genome-wide analyses on FEV1 and FEV1/FVC decline were conducted in 2677 nonasthmatics and 1441 asthmatics separately. The top hits were then replicated in 10,858 nonasthmatic and 1138 asthmatic participants. Decline of FEV1 and FEV1/FVC ratio was evaluated during a follow-up examination period of roughly 10 years in these participants. No SNP reached genome-wide significance in the discovery set. However, one locus on chromosome 13q14.3 containing the DLEU7 gene was strongly associated with FEV1 decline in nonasthmatics from the discovery set and confirmed in the replication set. A strong association signal was also reported on 8p22 harboring the TUSC3 gene for FEV1/FVC decrease in asthmatics, but not validated in the replication set. Many loci previously associated with cross-sectional lung function in GWA studies described above were replicated with baseline lung function in either asthmatic or nonasthmatic subjects. However, few GWAS-nominated lung-function loci were associated with lung-function decline, suggesting different genetic mechanisms governing baseline lung function and decline with age. In addition, this study showed the genetic heterogeneity of lung-function decline between subjects with and without asthma. Table 2 summarizes lung-function susceptibility loci identified by GWA studies.

GWA studies on COPD-related phenotypes

Other GWA studies were reported on COPD-related phenotypes. Emphysema is an important feature of COPD and varies considerably between patients. A recent GWA study was performed on emphysema measures by computed tomography scan and defined by radiologist qualitative scores and quantitative assessments of low-attenuation areas.43 The qualitative scores obtained in 1557 patients from the ECLIPSE study and 432 subjects from the Norway cohort led to the identification of an emphysema locus on chromosome 12p11.2. The most strongly associated SNP is located in the BICD1 gene, known to be involved in regulating telomere length. The ECLIPSE, Norway, and NETT studies were also used to perform a GWA study on COPD-related cachexia phenotypes, including body mass index and fat-free mass index.44 Cachexia occurs in approximately 10% of patients with COPD and is associated with increased mortality. The GWA study on body mass index and fat-free mass index in patients with COPD identified a single susceptibility locus that harbored the FTO gene, the most robust gene associated with obesity. Whether FTO acts through obesity or directly affects lung function remains to be elucidated.

GWA studies on COPD, lung function, and related phenotypes provided strong and consistent evidence of genetic susceptibility loci. These studies also highlight the large number of participants required to identify reproducible genetic loci. So far, GWA studies have identified only a small fraction of the genetic variants contributing to COPD risk, related complications, and lung-function variability. GWA studies on larger sample sizes, especially for COPD, will be required to identify the genetic factors underpinning COPD and related phenotypes. Large international efforts are under way to increase sample sizes and use more comprehensive molecular phenotyping (eg, gene expression in the lung) to elucidate the genetic component of COPD.45,46 It should be emphasized that the causal genes and genetic variants of all these newly discovered loci by GWA studies remain to be identified. More integrative genomic approaches will be required for these purposes. Different study designs testing rare and copy-number variants as well as gene-smoking interaction are also needed.

Integrative genomic approaches

More studies are being conducted using integrative genomic approaches in order to identify COPD susceptibility genes. For example, the IREB2 gene was identified by combining gene expression in human lungs and genetic association in COPD cohorts.47 In this study, lung specimens were obtained from patients undergoing lung nodule resection, and gene expression was compared between 15 COPD and 18 non-COPD patients using whole-genome gene-expression arrays. A total of 889 SNPs found in the 62 genomic regions containing genes differentially expressed between patients with or without COPD were tested for association with COPD and lung function. Seventy-one SNPs nominally associated (P ≤ 0.05) with COPD in the NETT-NAS study were followed up for replication in the EOCOPD study. A gene-based replication was then completed to confirm genetic association between genetic variants in the IREB2 gene and lung function. Overall, the IREB2 gene was shown to be upregulated in lung specimens of COPD patients and to contain genetic variants associated with COPD. Gene expression in a larger number of lung specimens will be required to test whether COPD-associated SNPs in the IREB2 gene influence the expression of its gene product.

Although Table 2 shows the major susceptibility loci identified by GWA studies, many additional loci were borderline significant in these studies. Many true positives are likely to be missed by this approach owing to the stringent threshold used to control for false-discovery rates. Different weighting methods and SNP-prioritization strategies are currently used to find true-positive signals from previous GWA studies. For example, the FGF7 gene was recently identified as a COPD susceptibility locus by weighting GWA analysis on regions of conserved homozygosity haplotype in subjects affected with COPD compared to unaffected subjects.48 As mentioned previously,49 further studies reanalyzing genome-wide SNP datasets with weighting methods based on function annotations (eg, coding variants or regions) or prior knowledge (eg, candidate genes or genome-wide linkage studies) will be required. Similarly, ongoing lung expression quantitative trait loci (eQTLs) mapping data36,46 are likely to leverage the impact of previous GWA studies on COPD by providing a list of SNPs that regulate gene expression in relevant tissues. SNPs associated with gene expression will provide crucial functional information to understand the molecular changes introduced by the susceptibility DNA variants. The identification of SNPs associated with both disease traits and quantitative transcript levels of one or more genes in relevant tissues will highlight the most likely causal gene within the susceptibility loci and the functional SNPs that are prime candidates to be directly involved in the pathogenesis of COPD.

Conclusion

Elucidating the genetic component of COPD and lung function turned out to be a challenging task. Major resources and collaborative efforts will be required to achieve our goal. In this review, the author provides an updated list of COPD genes and a summary of GWAS results conducted during the last few years. It is hoped that the gene list can be used by investigators to replicate or refute susceptibility genes of COPD. As eluded above, this gene list can also be used to reanalyze GWA data by prioritizing genes previously associated with COPD or related phenotypes or enter into more global gene network and causality analyses. Owing to the challenge faced by the genetic community, large collections of patients well characterized for COPD phenotypes are ongoing to identify the genuine COPD genes. A lumping and splitting strategy is an old idea in the field of genetics of complex traits50 that will certainly be essential in the field of COPD. Pooling resources (ie, lumping) is required to obtain proper sample sizes, but is likely to increase heterogeneity. These larger sample sizes, however, provide the opportunity to subdivide (ie, splitting) the pooled data into more homogeneous subgroups where the molecular defects are more likely to be similar. Accordingly, not only the genetic community but the entire spectrum of experts managing and treating patients with COPD will be required to provide samples, precise phenotypes, and expertise to search for the underlying genetic mechanisms. In parallel, complementary multidimensional genomic data in relevant tissues (eg, lung eQTLs) will be crucial to uncover causal genes and genetic variants that contribute to COPD and to discover new molecular targets for prevention, diagnosis, and treatment.

Table S1

Genes tested but showing lack of association with chronic obstructive pulmonary disease

Symbol	Name	Chromosome	References
			Single studies		Meta-analyses
			Positive	Negative	Positive	Negative
AGER	Advanced glycosylation end product-specific receptor	6	1,2
CASP8	Caspase 8, apoptosis-related cysteine peptidase	2	3
CCL17 (TARC)	Chemokine (C-C motif) ligand 17	16	4
CCL2	Chemokine (C-C motif) ligand 2	17	5
CFLAR	CASP8 and FADD-like apoptosis regulator	2	3
COL6A5	Collagen, type VI, alpha 5	3	6
CXADR	Coxsackie virus and adenovirus receptor	21	7
CYP1B1	Cytochrome P450, family 1, subfamily B, polypeptide 1	2	8,9
CYP2D6	Cytochrome P450, family 2, subfamily D, polypeptide 6	22	10
DCN	Decorin	12	11
DNAJB1	DnaJ (Hsp40) homolog, subfamily B, member 1	19	12
EDNRA	Endothelin receptor type A	4	13
FGA	Fibrinogen alpha chain	4	14
FGB	Fibrinogen beta chain	4	14,15
FGG	Fibrinogen gamma chain	4	14
FKBP4	FK506 binding protein 4, 59 kDa	12	12
FKBP5	FK506 binding protein 5	6	12
FLCN	Folliculin	17	16
GABPA	GA binding protein transcription factor, alpha subunit 60 kDa	21	17
GPX1	Glutathione peroxidase 1	3	18,19
GSTM3	Glutathione S-transferase mu 3 (brain)	1	20
HDAC2	Histone deacetylase 2	6	1
HDAC5	Histone deacetylase 5	17	1
HSP90AA1 (HSPCA)	Heat shock protein 90 kDa alpha (cytosolic), class A member 1	14	12
HSP90AB1 (HSPCB)	Heat shock protein 90 kDa alpha (cytosolic), class B member 1	6	12
HSPA1A	Heat shock 70 kDa protein 1A	6	21
HSPA1B	Heat shock 70 kDa protein 1B	6	21
HSPA1L	Heat shock 70 kDa protein 1-like	6	21
HSPA8	Heat shock 70 kDa protein 8	11	12
IL11	Interleukin 11	19	1
IL13RA2	Interleukin 13 receptor, alpha 2	X	22
ITGB5	Integrin, beta 5	3	7
JAK3	Janus kinase 3	19	1
KCND2	Potassium voltage-gated channel, Shal-related subfamily, member 2	7	1
MAP3K5	Mitogen-activated protein kinase kinase kinase 5	6	1
MIR146a	MicroRNA 146a	5	23
MRPL44	Mitochondrial ribosomal protein L44	2	24
ORMDL3	ORM1-like 3 (S cerevisiae)	17	25
PTGES3	Prostaglandin E synthase 3 (cytosolic)	12	12
RARRES2	Retinoic acid receptor responder (tazarotene induced) 2	7	1
SCGB1A1 (CC16)	Secretoglobin, family 1A, member 1 (uteroglobin)	11	26
SOD1	Superoxide dismutase 1, soluble	21	18,27
TBX21	T-box 21	17	28
THSD4	Thrombospondin, type I, domain containing 4	15	2
TLR2	Toll-like receptor 2	4	29,30
TLR6	Toll-like receptor 6	4	31
TNFRSF1A	Tumor necrosis factor receptor superfamily, member 1A	12	32
TNFRSF1B	Tumor necrosis factor receptor superfamily, member 1B	1	32