Tobacco Documents Online

Eur Respir J 1997; 10:1380--1391 Pdnted in UK - all fights reserved REVIEW E9Z7 XE3Ze 1380 SAND EUR RESPIR J 97 (C]ffiJNKSGAARD INT PUBL LTD DE Copyright ©ERS Journals Lid 1997 European Respiratory Journal ISSN 0903 - 1936 Genetic risk factors for chronic obstructive pulmonary disease A.J. Sandford, T.D. Weir, P.D. Pard Genetic risk factors for chronic obstructive pulmonary disease. A.J. Sandford, T.D. Weir, P.D. Par~. ©ERS Journals Ltd 1997. ABSTRACT: Cigarette smoking is the major risk factor for chronic obstructive pulmonary disease (COPD). However, only a minority of cigarette smokers develop symptomatic disease. Studies of families and twins suggest that genetic factors also contribute to the development of COPD. We present a detailed literature review of the genes which have been investigated as potential risk factors for this disease. The only established genetic risk factor for COPD is homozygosity for the Z allele of the ¢z~-antitrypsin gene. Fieterozygotes for the Z allele may also be at in- creased risk. Other mutations affecting the structure of cq-antitrypsin or the regu- lation of gene expression have been identified as risk factors. Genes, including those for cq-antichymotrypsin, c~2-macroglobulin, vitamin D- binding protein and blood group antigens, have also been associated with the deve- lopment of COPD. Variants of the cystic fibrosis transmembrane regulator gene have been identified as risk factors for disseminated branchiectasis. The genetic basis to chronic obstructive pulmonary disease has begun to be elu- cidated and it is likely that several genes will be implicated in the pathogenesis of this disease. The knowledge gained from such studies may also prove relevant to other inflammatory diseases. Eur Respir J 1997; 10: 1380-1391. Respiratory Network of Centres of Ex- cellence, UBC Pulmonary Research Labo- ratory, St. Paul's Hospital, Vancouver, B.C., Canada. Correspondence: P.D. Par~ UBC Pulmonary Research Laboratory St. Paul's Hospital 1081 Burrard Street Vancouver B.C. Canada Keywords: Chronic obstructive pulmonary disease genetics risk factors Received: November 7 1996 Accepted for publication February 28 1997 Chronic obstructive pulmonary disease (COPD) is cha- racterized by decreased expiratory flow rates, increased pulmonary resistance and hyperinflation. The most impor- tant risk factor for the development of COPD is ciga- rette smoking [1]. Cigarette smoke, in combination with other factors, leads to two pathophysiological process- es in the lung. The first is proteolytic destruction of the lung parenchyma, which increases the size of the airspa- ces; these eventually coalesce to form emphysematous spaces. The development of emphysema is associated with a loss of lung elastic recoil. The second process is inflammatory narrowing of peripheral airways, which is characterized by oedema, mucus hypersecretion and fibrosis, scarfing, distortion and obliteration of periphe- ral airways. The loss of lung elastic recoil and the nar- rowing of the peripheral airways combine to decrease maximal expiratory flow from the lung and contribute to hyperinflation. In conjunction with gas exchange ab- normalities, hyperinflation produces the symptoms of COPD. Despite the clear association of smoking and airway obstruction, there remains marked interindividual varia- tion in the response to cigarette smoke. This indicates that there are additional genetic or environmental co- factors, which contribute to the development of COPD. It has been estimated that only 10--20% of chronic heavy smokers will ever develop symptomatic COPD [2, 3]. Co-factors, such as childhood viral respiratory infections and environmental and occupational pollution, undoub- tedly play a role in determining this susceptible subset. Furthermore, there is evidence that genetic susceptibility is of major importance. The epidemiological and clini- cal data that demonstrate a hereditary contribution to the development of COPD are summarized in table 1. Although the results of several of these studies show an aggregation of COPD in families, there is no clear Mendelian pattern of inheritance. Case-control studies have shown an increased prevalence of COPD in the relatives of cases as compared to the relatives of con- trois, which cannot be explained by differences in other known risk factors. There is also a higher prevalence Table 1. - Studies that demonstrate a genetic compo- nent to the development of COPD Study typc [Re~] Study showing clustering of COPD in families Family studies showing increased incidence of COPD or chronic bronchitis in relatives of cases compared to relatives of controls Studies showing significant correlations in lung function between parents and children and between siblings, and higher correlation between parents and children, or between siblings than between spouses Studies showing decreased prevalence of disease or less similarity in lung function with increased genetic distance Family studies showing a major gene effect or a genetic component to pulmonary function Studies of pulmonary function in monozygotic and dizygotic twins [41 [5-12] [4, 7, 13, 14] [5, 15, 16] [17, 181 [15, 19-241 COPD: chronic obstructive pulmonary disease. THIS ARTICLE IS FOR INDIVIDUAL USE ONLY AND MAY NOT BE FURTHER REPRODUCED OR STORED ELECTRONICALLY NZTHOUT NRZTTEN PERMISSIOH FROH THE COPYRIGHT HOLDER. UNAUTHORIZED REPRODUCTION MAY RESULT IN FINANCIAL AND OTHER PENALTIES.

GENETICS OF COPD 1381 of reduced lung function among the children of patients who have COPD than among their spouses. Cross-sec- tional studies have shown decreasing prevalence of dis- ease and less similarity in lung function with increasing genetic distance. Studies of twins support a large gene- tic contribution to the variability in lung function. Heri- tability estimates for forced expiratory volume in one second (FEV1) range 0.5-0.8. WEaST~R et al. [21] stud- ied the effects of smoking on lung function in mono- zygotic and dizygotic twins. They found that when one monozygotic twin was susceptible to the effects of ci- garette smoke, both twins developed reductions in lung function, whereas other monozygotic twin pairs appea- red to be nonsusceptible and, despite similar smoking intensity, maintained normal lung function. The same concordance of changes in the lung function with sim- ilar smoking intensity was not seen in dizygotic twins. Figure 1 presents the pathogenic mechanisms in COPD schematically. Our purpose in this article is to review the evidence that specific genes may contribute to genetic suscepti- bility to COPD. Identification of susceptibility genes Complex genetic diseases, such as COPD, are caused by the interaction of environmental factors and genetic susceptibility. Positional cloning has been used to iden- tify the genes for many Mendelian disorders, and has also proved successful in localizing multiple regions of interest in complex diseases, such as hypertension [25] and diabetes mellitus [26]. The positional cloning app- roach uses multiply-affected families, and compares the inheritance of the disease to the inheritance of genetic markers of known chromosomal location. If a genetic marker is consistently co-inherited with the disease, then it is inferred that the disease gene lies close to that mar- ker on the same chromosome. Additional markers from the region are used to progressively ref'me the localiza- tion, until the gene can be identified. The power of positional cloning studies is reduced by polygenic inheritance, genetic heterogeneity and inter- actions with environmental factors. Cigarette smoking is such an important risk factor for COPD that it is impossible to use family data in which the prevalence of cigarette smoking varies. Ideally, one would need multi- generation families, in which there were similar levels of exposure to cigarette smoke. However, this is extre- mely unlikely because of age- and gender-related dif- ferences in the prevalence of smoking. In addition, most patients with COPD do not come to medical attention until their fifth or sixth decade, by which time it is usu- ally impossible to obtain phenotypic data and deoxyri- bonucleic acid (DNA) from their parents, and their offspring are generally not old enough to have develop- ed significant symptoms of COPD. An alternative app- roach would be to use an intermediate phenotype: a trait which is known to predispose to the development of COPD in smokers, such as increased bronchial respon- siveness [27]. For these reasons, positional cloning is difficult to apply to genes involved in the pathogenesis of COPD. Therefore, an alternative strategy has been used; asso- ciation studies of candidate genes. The candidate gene approach involves identifying gene products that are clearly involved in the pathogenesis of a condition, and ' looking for genetic polymorphisms in the genes that code for these proteins. To determine if these variants contribute to the disease process, case-control studies are performed to test for the association of the polymor- phisms with the disease phenotype. The risk imparted by a particular phenotype can be calculated using the rela- tive risk (RR) or odds ratio (OR) equations. RR is given by: (a/[a+b])/(c/[c+d]); and the OR is: (a/b)/(c/d), where a and b are the number of patients with and without the risk allele, respectively, and c and d ate the number of controls with and without the risk allele, respectively. The calculation of OR and RR yields very similar val- ues when the prevalence of a condition is low; how- ever, the results diverge as the prevalence increases. This is illustrated in figure 2, in which the RR and OR for a genotype are calculated for different prevalences of the trait in the population. An increased OR or RR for a disease in individuals of a specific genotype may indicate that the genotype causes an abnormal gene prod- uct or gene regulation, which influences the disease pathogenesis. Alternatively, it is possible that the gene tested in the association study does not contribute to the disease process, but is in association with the true Environmental ~ , and occupationall ~ IChildhood respiratoryI pollution t ~ I. ,,, infections ~ Genetic /.~ susceptibility ~ .... " ~ ~Airway inflamm'ation~ I ~ Lung recoil 1]4 Expiratory flowll'_ and remodelling ~ hyperinflation Hg. 1. - Summary of the pathogenic mechanisms in chronic obstmc- five pulmonazy disease (COPD). Exposure to cigmeae smoke is the major factor in the pathogenesis of COPD but interacts with other risk factors, including genetic susceptibility, to produce airway obsm~c- tion by loss of elastic recoil and/or airway inflammation. • ~ 50" 40" 8.30- 20" 10" O0 011 012 013 014 015 '016 017 018 Prevalence of trait Hg. 2. - Dependence of estimates of relative risk (RR) on the pop- ulation prevalence of the trait under study. Values for RR and odds ratio (OR) diverge as the prevalence of the trait increases. ~: RR=5; - - o: RR=20.

1382 A.J. SANDFORD ET AL. disease-causing mutation. This is because the disease-causing muta- tion may have fast occurred on a chromosome containing the geno- type being tested in the study. If the two alleles are very close to each other, then they will remain in asso- ciation with each other for several generations and are said to be in linkage disequilibrium. The power of association studies has been clearly demonstrated [28]. Even genetic polymorphisms which impart only a slight increase in RR can be detected if sufficient numbers of patients and controls are obtain- ed. The weakness of the candidate gene approach is that only genes known to be involved in the patho- genic process can be examined. The other major difficulty is ensuring that the patient and control groups are adequately matched for every other variable that could influence the distribution of the genotype. Chief among these is ethnic origin. There is potential for false-positive or false-negative results if this factor is not carefully taken into account. Pulmonary ~ and bronchial ,~ capillaries • D-bi.nding pro.tein, / J:,:~., " ease cnemotaxic "~~ " ,. . Alpna2-macrogtooulin withC5a 7"~~''- !~ CFTR" ~:~!i,' iAlp.hal~antichy.mgtry.psin "-"""'~"~"~'-~ ~''~ Cytochrome P450 ~.. Elastase, _. ~;~~ Bronchial otlaerproteases t:~ara y' \ \ hyperresponsiveness andTNF-cz cell [ Int£r~titial genes ' SOD Fig. 3. - Schematic representation of an airway to illustrate how mutations in various genes may contribute to the development of chronic obstructive pulmonary disease (COPD). Alphal-antiprote- are (txl-antitrypsin), oq-antiehymotrypsin, and a2-macroglobulin are serum proteins that can inhibit inflammatory cell pretenses. Deficiencies in their function or level could enhance the proteolytic digestion of the lung parenehyma that characterizes emphysema. Cytoehrome P450 is an enzyme present in airway epithelial cells (primarily Clara cells) that converts inhaled toxic chemicals to their metabolites. A gene variant, which enhances the enzyme's activity, could increase the prevale; nee of lung cancer as well as accelerate the airway inflammation that characterizes COPD. There is an association between mutations in the CFTR gene and bronchiectasis. Variants of the vitamin D- binding protein may influence the susceptibility to COPD. This protein can be converted to a ma- emphage-aetivating factor and interact with complement factor 5a (C5a) and C5a cles-Arg to enhance ebemotaxis of inflammatory cells. Cb'TR: cystic fibrosis transmembrane regulator; SLPI: secreto- ry leueoeym proteinase inhibitor, TNF-ct: turnout necrosis factor-e; SOD: superoxide dismutase. For instance, an association of type 2 diabetes mellitus and an immunoglobulin G (IgG) haplotype was shown to be due to Caucasian admixture in a Native American population [29]. Caucasians have a lower incidence of diabetes and eoincidentally a higher prevalence of the IgG haplotype. Therefore, the haplotype appeared to be protective against diabetes, but in fact was only a mar- ker for Caucasian ancestry. The association was shown to be spurious because the protective effect was not seen in individuals with no Caucasian ancestry. Genetic factors in the pathogenesis of COPD Table 2 lists genes that have been tested as candidates for involvement in the pathogenesis of COPD. The table Table 2. - Genes implicated in the pathogenesis of COPD Genes for which association studies have shown a signifi- cant relationship between polymorphisms and COPD Alphal-antitrypsin Alphai-antichymotrypsin Cystic fibrosis transmembrane regulator Vitamin D-binding protein Alpha2-macroglobulin Cytochrome P450A1 ABH secretor, Lewis and ABe blood groups HLA Immunoglobulin deficiency Haptoglobin Candidate genes for which there are no significant asso- ciations at present Extracellular superoxide dismutase Secretory leucocyte proteinase inhibitor Cathepsin G COPD: chronic obstructive pulmonary disease; HLA: human leucocyte antigen. indicates those genes for which association studies have shown a significant relationship between specific poly- morphisms and COPD, and candidate genes that have the potential to be involved in the pathogenesis of COPD but for which there are no significant associations at present. Figure 3 is a schematic illustration of an air- way to depict how enhanced or deficient gene products could contribute to COPD. Alpha #-antitrypsin The recognition by LAURELL and ERIKSSON [30] that patients with extremely low levels of ~x-globulin had an increased prevalence of emphysema was the first study to show a genetic risk for COPD. Alphat-antitrypsin (cq- AT) is a powerful antiprotease and is one of the few enzymes that can inhibit leucocyte elastase. Alphal-AT is produced in large amounts by the liver, but is also produced by alveolar macrophages and peripheral blood monocytes [31]. It is a highly polymorphic protein and over 70 variants have so far been identified [32] using crossed electrophoresis [33] and isoelectric focusing [34]. The Z variant of ctt-AT has deficient antiproteolytic func- tion but, more importantly, it is improperly processed in the rough endoplasmic reticulum and aggregates with- in the cell. Large amounts of the Z variant of the AT protein accumulate in hepatocytes, where they can cause liver disease [35]. Individuals with homozygous Z mutations have extremely low levels of circulating at-AT (less than 15% of normal) and have a clearly accelerated rate of decline in lung function even in the absence of smoking [36, 37]. However, it is predomi- nantly among smokers who are homozygous that symp- tomatic airflow obstruction develops at a younger age

! i .i 1 GENETICS OF COPD 1383 [38, 39]. Although there is a clear association of homo- zygosity for this gene variant and the development of COPD, the homozygous state is rare in the population (1 in 1,670 [40] to 1 in 5,097 [41] live births in Caucasian populations) and, thus, can explain only a small percen- .tage of the genetic susceptibility to cigarette smoke. The discovery that homozygosity for the Z variant leads to increased risk for COPD led to numerous stud- ies in which an association of COPD and heterozygous genotypes was sought. The approximate allele frequen- cies of the most common gene variants M, S and Z are 0.93, 0.05 and 0.02, respectively. Patients with the MM genotype have the highest ~I-AT levels and are defined as normal. Patients who are heterozygous MS have mild reductions in ~I-AT levels to -80% of normal, whereas MZ heterozygotes have lower levels at -60% of normal. SZ compound heterozygotes are rare, but have even lower levels at --40% of normal. Two types of studies have attempted to identify an increased risk for COPD in the relatively common hetero- zygous MS and MZ genotypes. In ease-control studies, the prevalence of cq-AT genotypes in individuals with the clinical features of COPD is compared to control subjects without airflow obstruction, who are matched as closely as possible for other potential predictors of COPD. In general, the results of these case-control stud- ies have shown the OR to be significantly increased. As shown in table 3, the OR for COPD ranges 1.5-5.0. The prevalence of the MZ variant in the ease populations ranges 3.9-14.2%, whilst in the controls it ranges 1.0-- 5.3 . Investigators have also assessed the risk of the MZ genotype by studying lung function in the general popu- lation [49-56]. In these studies, a population sample is phenotyped for ~-AT variants and the prevalence of COPD in those with the MZ phenotype is compared with the prevalence in those with the MM phenotype. Many of these studies were based on small numbers of individuals and had insufficient power to detect an effect of the MZ or MS phenotype. However, even most of the larger studies showed no significant difference in respiratory symptoms or pulmonary function in the MZ individuals compared to MM subjects. In theory, popu- lation-based studies designed to examine the predictive value of a genotype are superior to case-control met- hods because there is less chance of a systematic bias. However, in COPD, where an environmental factor (i.e. cigarette smoking) plays an important role, population studies may have insufficient sensitivity to detect a fac- tor which only increases risk slightly. For example, in a collaborative study to assess risk of lung disease in MZ phenotype subjects, 143 MZ individuals did not have significantly lower lung function than 143 MM indi- viduals drawn from a population study of over 10,000 people [56]. However, only 37% of the subjects were current smokers, 35% had never smoked and 60% were less than 54 yrs of age. In contrast to these reports, the results of several pop- ulation studies have demonstrated differences between MZ and MM individuals. KLAY'ror~ et al. [57] found an increased prevalence of COPD in MZ heterozygotes who had smoked, but found no difference in the incidence of COPD between MM and MZ nonsmokers. CooP~.R et al. [58] found significantly decreased lung function in MZ individuals. However, both of these studies used relatives in the MZ study population and, therefore, the results may not be due to mutations in the Ctl-AT gene. TArr~RSALL et al. [59] found evidence for greater loss of elastic recoil in MZ versus MM smokers, but estima- tes of airway function were similar in both groups. HALL et aL [60] found that MZ heterozygotes had significantly lower expiratory flow rates, even in the absence of smok- ing. MholSO~q et al. [10] found more rapid decline in lung function in MZ individuals in a longitudinal study. Similarly, the results of a 10 ye~ longitudinal study of 28 MZ subjects demonstrated that deterioration in lung function was significantly greater than in a matched MM control group [61]. In addition to mutations that affect the basal serum levels of txt-AT, several mutations have been describ- Table 3. - Case-control studies of ~l-antitrypsin deficiency genotypes and chro- nic obstructive pulmonary disease (COPD) First [Ref.] Subjects Genotypes % OR author ' MZ MS ZZ SS SZ for MZ SmoF.or~ [42] 306 COPD patients 3.9 196 controls 1.0 B,~rMAr~ [43] 526 COPD patients 5.9 6.5 0.9 3.4 642 controls 1.2 6.5 0.3 0.2 Cox [44] 114 emphysema or 4.9 5.7 6.6 bronchitis patients 721 controls 1.9 7.9 0 J,,,r~s [45] 190 emphysema 14.2 5.3 2.6 1.1 patients - 1,303 controls 3.9 7 0.1 0.3 LmBF.gXt~ [46] 965 COPD patients 7.7 10.1 1.9 0.3 0.2 1,380 controls 2.5 8.0 0 0.1 0.4 MrrrMA~ [47] 350 COPD patients 10.0 6.3 3.4 0.9 0.9 2,830 controls 2.9 4.1 0.1 0.1 0.I Ktw~m~s [9] 114 COPD patients 7.9 4.4 2.6 0 114 controls 5.3 7.0 0 0.9 Bhm, ma'r [48] 107 COPD patients 9.3 5.6 1.9 91 controls 2.2 5.5 0 OR: odds ratio. ed that affect function [62], but these are relatively rare and can only explain a small percentage of the susceptible subgroup that develops COPD. Two separate groups have reported an associ- 4.0 ation between a mutation in the 3' region of the txl-AT gene and 5.0 COPD [63, 64]. KALSnF_XER et al. [63] found that this mutation was 2.6 associated with chronic lung dis- ease. Heterozygosity for the mu- tation in a group of patients with 3.9 pulmonary emphysema (I 8%), and in a group of patients with 3.3 bronchiectasis (19%), was signi- ficantly higher than in normal 3.8 controls (5%). However, the rea- son for the association of the 1.5 mutation with COPD was unclear, since it was not associated with 4.6 ~-AT deficiency or any partic- ular ct~-AT protein type. Subse- quently, these authors studied a r~

1384 A.J. SANDFORD ET AL. larger group of 140 patients with pulmonary emphyse- ma and bronchiectasis and found that 20% were hetero- zygous for the mutation (p=0.0015) [65]. The association has been independently confirmed by POLLER et al. [64] in a group of 137 COPD patients. The mutation was found in 15% of the patients and in only 5% of the healthy controls. In addition, a family was identified in which the mutation segregated with COPD, and, when homozygous, the mutation was associated with the onset of symptoms at a younger age. The 3' mutation could be associated with COPD as a result of linkage disequilibrium with the disease-causing allele. The at-antichymotrypsin gene has been mapped to within 220 kb of the cq-AT locus [66], and the mutant 3' allele could be in disequilibrium with an at-antichy- motrypsin deficiency allele. Alternatively, KALSrm~R and co-workers [65] have suggested that the 3' muta- tion may affect the regulation of at-AT gene expres- sion. Alphat-AT is an acute phase protein and its serum concentration increases two- to threefold during inflam- mation [67]. Presumably, the acute phase response has evolved to attenuate the proteolytic destruction that occurs at sites of acute tissue injury and, thus, prevents excessive tissue destruction. A deficient acute phase in- crease in cq-AT levels following viral or bacterial respi- ratory infections could exaggerate the proteolytie tissue destruction that accompanies the release of neutrophil elastase and other enzymes. It is possible that the 3' mu- tation could affect the acute phase response leading to reduced upregulation of ~zI-AT synthesis when inflam- mation is present. Alveolar and lung tissue macrophages are both capable of producing st-AT [31]. If the tzt-AT gene expression in tissue and alveolar macrophages is also affected by the mutation, then a disturbance of the proteolytic-antiproteolytic balance could develop with- in the microenvironment of the inflamed lung. MORGAN et al. [68] sequenced the 3' region of the tx~-AT gene, and showed that the mutation occurs in a region containing four consensus sequences for DNA- binding proteins, suggesting that it may affect a regu- latory element. Gel shift analysis and deoxyribonuclease (DNase) I footprinting experiments confirmed that all four potential regulatory regions bound nuclear factors [69]. However, the mutant sequence demonstrated poor binding, especially in the region of the mutation. To test for the functional significance of the mutation, both the wild type and mutant 3' regions were cloned into vectors, downstream of a reporter gene. These con- structs were used to transfect three different cell lines. In all of the cell types, the wild type sequence showed a 50--100% increase in gene expression compared to a control plasmid. Furthermore, the mutant sequence sho- wed two- to fourfold less activity than the wild type. The acute phase response is primarily mediated by interleukin 6 [70]. Recently, it has been proposed that transcription factors of the CCAAT box enhancer bind- ing protein (C/EBP) family play an important role in increasing acute-phase gene transcription [71]. The 3' region of the a~-AT gene contains a C/EBP binding site. Interestingly, the mutation in the 3' region appears to influence the binding to neighbouring regions, includ- ing the C/EBP site and, therefore, may influence acute phase gene expression. An additional polymorphism in the 3' region of the ~x~-AT gene has been shown to be associated with COPD [72]. The polymorphism was found in 3 out of 70 COPD patients but in none of 52 controls. The mutant allele showed loss of more than one restriction site, suggest- ing the presence of a deletion. Homozygosity for this mutation was associated with early onset COPD. This polymorphism was also associated with normal at-AT levels. 2063633577 Alpha l-antichymotrypsin Alphat-antichymotrypsin, like a~-antitrypsin, is a ser- ine protease inhibitor and acute phase reactant. Alphat- antichymotrypsin (txt-ACT) is known to inhibit pancreatic chymotrypsin, neutrophil cathepsin G, mast cell chy- mase and the production of neutrophil superoxide [73]. It is synthesized by hepatocytes and alveolar macropha- ges [74]. Alphal-ACT deficiency has a prevalence of approxi- mately 1% in the Swedish population. In cases where hereditary deficiency has been shown, transmission fol- lows an autosomal dominant inheritance pattern [75, 76]. No consistent clinical phenotype is associated with cq-ACT deficiency, although an increased prevalence has been reported in patients with childhood asthma [77] and COPD [78, 79]. In two other studies, deficient pati- ents had increased values of residual volume (RV) and of the RV/total lung capacity (TLC) ratio [75, 76]. Two point mutations in the ¢z~-ACT gene have been associated with decreased eft-ACT serum concentrations and COPD. POLLER and co-workers [78] described an amino acid substitution, Pro227--~Ala, which they found in four of 100 unrelated COPD patients and none of 100 controls in a German population (p=0.04). All four pa- tients with the mutant gene had serum cq-ACT concen- trations approximately 60% of normal and ~xt-AT levels within the normal range. However the prevalence of the pro227-->Ala mutation may vary in different populations, since it was not detected in 102 Russian COPD pati- ents [80]. A second amino acid substitution, Leu55-->Pro, was reported by POI.I.ER and co-workers [79] in three out of 200 unrelated COPD patients and none of 100 con- trols. Mean txt-ACT serum levels in the heterozygotes was 80% of normal, and the mutant protein had an altered pattern on isoelectric focusing and defective func- tion. One of the heterozygotes belonged to a family in which thre~ members were affected with severe early onset COPD. The mutant allele segregated with COPD in this three generation pedigree. Cystic fibrosis transmembrane regulator The cystic fibrosis transmembrane regulator (CFTR) gene product forms a chloride channel at the apical sur- face of airway epithelial cells and is intricately involved in the control of airway secretions. Homozygous defi- ciency or defective function of this protein results in cystic fibrosis (CF), characterized by elevated sweat chloride levels and early onset obstructive lung disease, secondary to chronic bacterial infection and bronchiec- tasis. The prevalence of CF is 1 in 2,000 to 1 in 3,000, with the carrier frequency estimated at 1 in 20 to 1 in

30 in populations of Northern European descent [81]. ~[i It has been hypothesized that this relatively high pre- valence arose from a selective advantage of carrying a CF allele. Resistance to pulmonary tuberculosis [82], influenza [83], and cholera [84] have each been suggest- |~ ed as a selective advantage. In an animal model, mice _.11 that were heterozygous for a mutant CFTR allele secre- ted 50% less intestinal fluid and chloride ion in response to cholera toxin [85]. CF heterozygotes could have altered airway water and ion regulation, altered mucoeiliary clearance and an in- creased susceptibility to challenges that are attenuated iI by these mechanisms. In the 1960s, several groups inves- _ tigated the hypothesis that CF heterozygotes may be pre- disposed to respiratory disease. Comparisons of parents of CF patients versus controls (mean age 34-36 yrs) did [ not reveal any significant differences in lung function _ or history of asthma or chronic bronchitis [86-89]. How- ever, obligate heterozygotes have been shown to have increased bronchial reactivity to methacholine [90], and I increased incidence of wheeze accompanied by decrea- _ sed FEVI and forced mid-expiratory fiow (FEF25-T5) [91]. More than 580 variants of the CFTR gene have been described; the most common mutation, AF508, is found on approximately 70% of all CF chromosomes [92]. Heterozygosity for the AF508 mutation was identified , in four of eight patients with disseminated bronchiecta- sis [93], and in five of 65 patients with bronchial hyper- - secretion [94]. In both studies, it is unclear whether the AF508 heterozygotes are predisposed to lung disease or whether they have mild, previously undiagnosed CF with unidentified CFTR mutations on their other chro- mosomes. In a study of patients with normal sweat chloride levels, G~.Rw,[s et al. [95] found the prevalence I of AF508 to be increased (four out of 47) in patients with bronchiectasis and not increased (seven out of 161) in patients with chronic bronchitis. The AF508 mutation was not found in any of 21 Japanese patients with diffuse panbronchiolitis, a disease with pathologi- cal~ and'. clinical characteristics similar to mild CF [96]. Recently, investigators have searched for associations between respiratory disease and other CFTR variants, GENETICS OF COPD 1385 sion. This variant was not found to be significantly in- creased in five of 33 COPD patients. Early work by the same authors did not support the involvement of CFTR in COPD.by linkage analysis with a CF locus marker [1011. In summary, heterozygosity for AF508 appears to pre- dispose for disseminated bronchiectasis, but the involve- ment of CFTR in other obstructive pulmonary diseases remains unproven. Studies of CPTR mutations in COPD patients who have documented lifelong airway challen- ges, such as cigarette smoking, have not been perform- ed. Vitamin D-binding protein (group-specific component) Vitamin D-binding protein (VDBP), also known as group-specific component, is a 55 kDa protein secreted by the liver, that is able to bind extracellular actin and endotoxin in addition to vitamin D. VDBP enhances the chemotactic activity of complement factor 5a (C5a) and C5a des-Arg for neutrophils by one to two orders of magnitude [102]. In addition, VDBP can act as a macro- phage-activating factor [103]. Thus, besides its vitamin D-binding function, VDBP can have important influen- ces on the intensity of the inflammatory reaction. Numer.ous isoforms of VDBP have been identified by isoelectrie focusing. Two common substitutions in exon 11 of the gene result in three possible isoforms, termed IF, 1S and 2. Figure 4 shows a partial gene map of VDBP and the substitutions responsible for protein iso- forms. Ku~PPEgS et al. [9] found a decreased frequency of the 2-2 genotype in COPD patients compared to con- trois. Subsequently, HoP~E et al. [ 104] performed a case- control study, in which they found that the prevalence of the 1F homozygote was significantly greater among patients with COPD than among controls, yielding a RR of 4.8. In addition, the genotypes that contained the 2 allele (2-1F, 2-1S and 2-2), had a protective effect. How- ever, this association remains controversial, since it was not replicated by KA~t¢~ et aL [105]. a) I in addition to AF508. A~rtaCH et al. [97] examined 100 Intron 10 Exon 11 Intron 11 patients with chronic bronchitis for the more common CFTR mutations (AF508, R553X, G551D, G542D, G542X, N 1303K and 621+1G--~T). The only mutation, /~~ o~. AF508, was found in one patient who also had bron- chicctasis, suggesting that none of these Cl:rrR muta- lfions predisposes to chronic bronchitis [97, 98]. Pmr~ATT~ and co-workers [99] performed detailed screening for approximately 70 CP-TR mutations. Although variants were found in two of 12 patients with COPD without bronchiectasis, and in two of 36 patients with non- obstructive pulmonary disease, the frequency of the mutations was not significantly different from that expect- eel However, CFTR mutations were found in five of 16 patients with disseminated bronchiectasis and normal sweat chloride levels (one each with mutations AF508, R75Q, Ml137V, 3667ins4, R1066C). In a subsequent study, five of the same 16 patients were also found to have the IVS8-5T variant (three of whom were previ- ously negative for other CPTR mutations) [100]. The IVS8-5T allele results in reduced CFTR gene expres- b) Isoform 1S 1F 2 Amino acid 416 Glu Asp Asp Amino acid 420 mhr Population frequency 0.57 0.18 0.25 Fig. 4. - Polymorphisms in the vitamin D-binding protein gene. a) Two point-mutations in exon 11 of the gene result in amino acid sub- stitutions at positions 416 and 420 of the protein, b) Amino acids pre- sent at position 416 and 420 in the three isoforms of the vitamin D-binding protein, and the frequencies of the isoforms in Caucasian populations. G: guanine; T: thymine; C: cytosine; A: adenine; Glu: glutamie acid; Asp: aspanic acid; Thr: threonine; Lys: lysine.

1386 Ad. SANDFORD ET AL. No studies have so far examined the influence of these genetic variants on the ability of the protein to act as a chemotactic enhancer of C5a or as a macrophage-activ- ating factor. However, the macrophage-activating factor is formed from VDBP by modification of an oligosac- chadde side chain. Less than 10% of the 2 isoform is glycosylated and able to form macrophage-activating factor [103], which is consistent with a protective effect for the 2 allele. Alpha~-macroglobulin Alpha2-macroglobulin is a broad spectrum scram pro- tease inhibitor. Normal serum levels of ~,2-macroglobulin are higher in females and decrease with age. Synthesized by bepatocytes, alveolar macrophages [106] and human lung flbroblasts [107], ot~-macroglobulin is thought to have a protective role in the lung. The large size of t~z- macroglobulin (725 kDa) prevents significant transport from blood to the lung interstitium or alveolar space, so that serum levels do not necessarily reflect its concen- tration in the lung. However, increased permeability of the vessel wall under inflammatory conditions could allow ~,~-maeroglobulin to enter the interstitial space [108]. An increase in ~.2-macroglobulin levels can be detected in the sputum of patients with acute chest in- fections [109]. Elevated ~-macroglobulin levels, up to two times control, have been reported in the serum of patients with ~t-AT deficiency, irrespective of the pre- sence or absence of COPD [110, 111]. Such an eleva- tion is not seen in patients with emphysema unrelated to ~1-AT deficiency. Alpha2-macroglobulin serum deficiency is rare and the cause is unknown. Two case studies described hereditary ~-macroglobulin deficiency with autosomal dominant transmission [112, 113]. Although symptoms suggestive of respiratory disease were not found in the deficient individuals, it is possible that the subjects were not old enough to develop COPD. Neither study included pul- monary function tests or smoking histories. Comple~ lack of ~-macroglobulin has not been described and may be incompatible with life. The ~,2-macroglobulin gene, located on chromosome 12, has been cloned and sequenced [114]. Whilst rest- rietion fragment length polymorphism (RFLP) vari- ants of the ~-macroglobulin gene have been described, only one variant has been reported to be associated with chronic lung disease in a single patient [115]. The patient had oa-macroglobulin serum levels 50% of normal and chronic pulmonary disease since childhood progressing to very severe COPD at the age of 42 yrs (smoking his- tory not reported). DNA from this patient was digested with 10 restriction enzymes and probed with an-~2- macroglobulin complementary deoxyribonucleic acid (eDNA) probe. All 10 restriction enzymes showed an alteration in the RFLP pattern suggesting a major alter- ation of the ~-macroglobulin gene. RFLP analysis with nine of the 10 restriction enzymes failed to demonstrate polymorphisms in 40 control and 39 COPD patients. The same author sequenced two functional domains of the ~2-macroglobulin gene in 30 COPD patients and 30 control subjects [114]. A common amino acid substitu- tion, Vall°°°--~lle, was detected equally in both groups. One COPD patient had an amino acid substitution, CysgV2---~Tyr, which was predicted to interfere with ot2- macroglobulin function. The serum level of ~t~-macro- globulin in this patient was within the normal range. Cytochrome P4501AI Cytochrome P4501A1 is an enzyme that metabolizes exogenous compounds to enable them to be excreted in the urine or bile. It is found throughout the lung, and may play a role in the activation of procarcinogens to their carcinogenic forms. The enzyme is produced by the CYPIA1 gene and mutations at this locus have been associated with lung cancer [116]. In a recent study, the prevalence of a mutation in exon 7 of CYPIA1 was assessed in lung cancer and COPD patients [117]. This mutation causes a substitution of isoleucine to valine at residue 462, and results in a pro- tein with almost twice the enzymatic activity of the isoleucine protein. The high-activity allele was found to be associated with susceptibility to centriacinar emphy- sema and lung cancer. The polymorphism was not link- ed to lung cancer in the absence of emphysema. Blood group antigens 2083833579 The association of COPD with the ABO, secretor and Lewis genes has been the focus of several studies. The ABO locus on chromosome 9 determines the activity of a glyeosyltransferase, which converts glyeoprotein H into the A or B antigens. An association between the ABO locus and COPD was found by Con~q et al. [118]. The results of this study suggested that impaired lung func- tion was associated with type A blood group. This was confirmed by the same authors in a 5 yr longitudinal study, in which there was a greater decline in lung func- tion in group A individuals compared with non-A sub- jects [119]. In direct contrast to these studies, I~,cz~ows~a et al. [120] found that subjects with blood group A had a smaller decline in lung function than individuals with other blood groups. The results of several other studies have failed to confirm any association of ABO alleles and pulmonary function [23, 121, 122]. ABO antigens are present on virtually all ceils of the body. Approximately 80% of the population secretes ABO antigens into saliva, plasma and respiratory tract secretions. The ability to do this is determined by the secretor locus on chromosome 19q and is a dominant trait. It has been reported that nonsecretors have impair- ed lung function compared to secretors [123, 124]. This suggests that the presence of ABO an[igens in the secre- tions of the respiratory tract may have a protective effect against lung impairment. This result was independent- ly confirmed by Kaor~a~rq et al. [105], who found sig- nificantly more nonsecretors of blood group O in subjects with low FEV1 measurements (OR=15.6). Secretor sta- tus was shown to have a protective effect against de- 'cline in peak expiratory flow rates [123], but, in this study, the effect was only observed in subjects over 40 yrs of age. These associations are controversial because they have not be~n replicated in other populations [121, 122, 125].

GENETICS OF COPD 1387 The Lewis blood group has also been investigated as a possible risk factor for airflow obstruction [126]. In Caucasian populations, ~90% of individuals have the dominant Le allele and produce Lewis a substance. In individuals who are secretors, this is converted to Lewis b substance, and therefore they have a and b substances in their serum. Lewis-positive nonsecretors only have a substance in their serum. HoP,~ et al. [126] found a significant increase in airflow obstruction in Lewis- negative subjects, with a RR of 7.2. The authors sug- gest that it is the presence of b substance rather than secretor status that protects against airflow obstruction. Since the blood group systems interact at the protein level, a recent study has considered all three gene loci together [127]. Blood group O individuals who were either Lewis-negative or nonsecretors, were found to have impaired lung function and higher prevalence of wheez- ing and asthma. Individuals who were both Lewis- negative and nonsecretors, had very low lung function. Lewis-positive secretors were found to have lower lung function if they had blood group A, compared with group O. The reason for the association of ABO, Lewis and secretor genes with COPD remains unclear, but it may be due to the role of these systems in the adhesion of infectious agents [128]. Recurrent respiratory infections, especially in childhood, are known to be a risk factor for COPD, and particular alleles of these blood groups may increase an individual's susceptibility to infection. Human leucocyte antigen locus Associations between the human leucocyte antigen (HLA) class I genes and COPD have been investigated in a study of heavy smokers with high FEV1 values and lifelong nonsmokers with low FEVI values [105]. The authors observed a significant decrease of the I-ILA- Bwl6 allele in those with low FEVI values (OR=0.2), and a significant increase of the HLA-B7 antigen in the same group (OR=3.8). HLA typing was also performed in a population of Japanese patients with diffuse panbronchiolitis [129]. The results demonstrated an increased prevalence of HLA-Bw54 in the patients compared to the control sub- jects (RR=13.3). This HLA type is only found in Japa- nese, Chinese and Korean individuals, and may explain why diffuse panbronchiolitis has not been reported in Caucasian or African populations. It is not yet clear whether these associations are due to variations in the HLA genes themselves or to suscep- tibility genes in linkage disequilibrium with the HLA alleles. of IgG2 and two had decreased levels of IgG3 [130]. All six of the IgG and IgA deficient subjects were found to have abnormal lung function. In addition, a signifi- cant correlation of IgG2 levels and FEV! values was found by O'K~ et al. [131]. Selective IgA deficiency has been found to segregate with COPD, in a large, three generation pedigree [133]. Haptoglobin The serum protein haptoglobin has several common polymorphisms. The prevalence of these variants was investigated in a population of subjects with low FEVI values [105]. Overall, no significant difference in the fre- quency of haptoglobin variants was observed between in- dividuals with low FEV1 values compared to those with high values. Among those with non-O blood groups, a possible protective affect of the HplS allele was detec- ted. However, a similar association was not found in an earlier study [9]. Other candidate genes for COPD Extracellular superoxide dismutase Extracellular superoxide dismutase (EC-SOD) is a sec- retory glycoprotein found mainly in the interstitial spa- ees, although -1% is found in the plasma, lymph and synovial fluids. It is the main extracellular antioxidant enzyme in the lung. EC-SOD has a high affinity for gly- cosaminoglycans, such as heparan sulphate, and there- fore >98% of the enzyme is found bound to beparan sulphate in the connective tissue matrix. EC-SOD is ideally localized to play an important role in attenuat- ing tissue damage secondary to oxygen radicals inhaled in cigarette smoke and generated by activated inflam- matory cells. A polymorphism in the EC-SOD gene results in the substitution of arginine to glycine at amino acid posi- tion 213 [134, 135]. Approximately 2% of a random population were found to be heterozygous for the sub- stitution [135]. This mutation (R213G) is located in the heparin-binding domain and results in a -30 fold incre- ase in the serum enzyme concentration, presumably due to a failure of EC-SOD to remain bound to interstitial glycosaminoglycans. A 10 fold increase in serum EC- SOD has been reported in a lung transplant patient with end-stage emphysema [136]. However, the R213G allele was not present in this patient, suggesting that further variants of this gene remain to be identified. Immunoglobulin deficiency Secretory leucocyte proteinase inhibitor The role of hereditary immunoglobulin A (IgA) and immunoglobulin G (IgG) deficiency in the aetiology of COPD has been examined in several studies [130, 13 I]. Patients with IgA deficiency, either alone or in combina- tion with IgG deficiency, are known to have recurrent respiratory infections [132]. In a study of IgA deficient individuals, four were found to have decreased levels Secretory leucocyte proteinase inhibitor (SLPI) is a 12 kDa serine antiprotease found in a variety of mucous secretions, including those of the respiratory tract. SLPI is produced loca!ly in the lung by airway epithelial cells and is able to inhibit neutrophil elastase [137]. Therefore, SLPI may play an important role in the prevention of tissue damage by neutrophils during inflammation. ABE

1388 A.J. SANDFORD ET AL. et al. [138] screened 114 individuals for polymorphisms in exons 2, 3 and 4 of the SLPI gene. The subjects in- eluded individuals with various cq-antitrypsin genotyp- es and 10 early onset COPD patients who did not have ct~-antitrypsin deficiency. However, no mutations were discovered, which suggests that structural alterations in the SLPI protein do not play a major role in the patho- genesis of COPD. Cathepsin G Cathepsin G is a sedne protease, and mutations in the gene for this enzyme may predispose individuals to COPD [139]. Therefore, exons 1-5 of the eathepsin G gene were screened for mutations in 180 individuals. A mutation was found in exon 4, which resulted in an amino acid substitution at position 125, but it was not associated with COPD. Summary In this manuscript, we have reviewed evidence for a genetic component to COPD and have described the genes that could contribute to the genetic risk. The diag- nosis of COPD is based on decreased expiratory air- flow, and it is possible that different pathophysiological processes contribute to this phenotype within and bet- ween patients. For example, bronchial smooth muscle cell hypertrophy, inflammatory narrowing of periphe- ral airways and loss of elastic recoil may contribute to a different extent in certain individuals. Susceptibility to these processes may have differing genetic bases. A search for genes that increase susceptibility to airflow obstruction among smokers may have implications be- yond the development of COPD. In epidemiological studies, a decrease in FEV1 has been shown to be a marker of premature mortality from other causes [140]. It is possible that an excessive pulmonary response to inhaled toxins and pollutants will serve as a marker of polymorphisms that increase susceptibility to other in- flammatory and degenerative diseases. The development of rapid, inexpensive molecular methods to screen for specific polymorphisms will allow an increased capacity to identify risk genotypes. This has profound relevance for the conduct of clinical investigations of environmen- tal risk, therapeutic interventions and clinical screening. References 1. Snider GL. Chronic obstructive pulmonary disease: risk factors, pathophysiology and pathogenesis. Ann Rev Med 1989; 40: 411--429. 2. Fletcher C, Peto R, Tinker C, Speizer bE. The natural history of chronic bronchitis: an eight year follow-up study of working men in London. Oxford, Oxford Uni- versity Press, 1976. 3. Bascom R. Differential susceptibility to tobacco smoke: possible mechanisms. Pharmacogenetics 1991; 1: 102- 106. 4. Tager IB, Rosner B, Tishler PV, Speizer bE, Kass EH. Household aggregation of pulmonary function and chro- nic bronchitis. Am Rev Respir Dis 1976; 114: 485--492. 5. Tager I, Tishler PV, Rosner B, Speizer bE, Litt M. Studies of the familial aggregation of chronic bronchi- tis and obstructive airways disease, lnt J Epidemiol 1978; 7: 55-62. 6. Khoury M J, Beaty TH, Newil CA, Bryant S, Cohen BH. Genetic-environmental interactions in chronic airways obstruction. Int J Epidemiol 1986; 15: 65-72. 7. Higgins M, Keller J. Familial occurrence of chronic respiratory disease and familial resemblance in ventila- tory capacity. J Chron Dis 1975; 28: 239-251. 8. Speizer FE, Rosner B, Tager I. Familial aggregation of chronic respiratory disease: use of national health inter- view survey data for specific hypothesis testing. Int J Epidemiol 1976; 5: 167-172. 9. Kueppers F, Miller R.D, Gordon H, Hepper NG, Offord K. Familial prevalence of chronic obstructive pulmonary disease in a matched pair study. Am J Med 1977; 63: 336-342. 10. Madison R, Zelman R, Mittman C. Inherited risk fac- tors for chronic lung disease. Chest 1980; 77 (2 Suppl.): 255-257. 11. Larson RK, Barman ML, Kueppers F, Fudenberg HH. Genetic and environmental determinants of chronic ob- structive pulmonary disease. Ann Intern Med 1970; 72: 627-632. 12. Cohen BH, Diamond EL, Graves CG, et al. A com- mon familial component in lung cancer and chronic obstructive pulmonary disease. Lancet 1977; ii: 523-526. 13. Kauffmann F, Tager IB, Mufioz A, Speizer FE. Familial factors related to lung function in children aged 6-10 years. Am J Epideraiol 1989; 129: 1289-1299. 14. Devor F_J, Crawford MH. Family resemblance for nor- mai pulmonary function. Ann Hum Biol 1984; 1 I: 439- 448. 15. Redline S, Tishler PV, Rosner B, et aL Genotypic and phenotypic similarities in pulmonary function among family members of adult monozygotic and dizygotic twins. Am J Epidemiol 1989; 129: 827-836. 16. Silverman E, Chapman H, Drazen J, et al. Early-onset chronic obstructive pulmonary disease (COPD): pre- liminary evidence for genetic factors other than Pi type. Am J Respir Crit Care Med 1996; 153: A48. 17. Rybicki BA, Beaty TH, Cohen BH. Major genetic mech- anisms in pulmonary function. J Clin Epidemiol 1990; 43: 667-675. 18. Astemborski JA, Beaty TH, Cohen BH. Variance com- ponents analysis of forced expiration in families. Am J Med Genet 1985; 21: 741-753. 19. Zamel N, Webster P, Lorimer E, Man S, Woolf C. En- vironment versus genetics in determining bronchial sus- " eeptibility to cigarette smoking. Chest 1981; 80: 57S. 20. Redline S, Tishler PV, Lewitter FI, Tager IB, Munoz A, Speizer bE. Assessment of genetic and nongenetic influences on pulmonary function: a twin study. Am Rev Respir Dis 1987; 135: 217-222. 21. Webster PM, Lorimer EG, Man SFP, Woolf CR, Zamel N. Pulmonary function in identical twins: comparison of nonsmokers and smokers. Am Rev Respir Dis 1979; 119: 223-228. 22. Hankins D, Drage C, Zamel N, Kronenberg R. Pulmo- nary function in identical twins raised apart. Am Rev Respir Dis 1982; 125: 119-121. 23. Hubert HB, Fabsitz RR, Feinleib M, Gwinn C. Genetic and environmental influences on pulmonary function in adult twins. Am Rev Respir Dis 1982; 125: 409--415. 24. Man SFP, Zamel N. Genetic influences on normal vari- ability of maximum expiratory flow-volume curves. J Appl Physiol 1976; 41: 874-877. 25. Soubrier F, Lathrop GM. The genetic basis of hyper- tension. Curr Opin Nephrol Hypertens 1995; 4:177-181. 2063633581

GENETICS OF COPD 1389 26. Davies JL, Kawaguchi Y, Bennett ST, et al. A genome- wide search for human type 1 diabetes susceptibility genes. Nature 1994; 371: 130-136. 27. Vollmer WM, Johnson LR, Buist AS. Relationship of response to a bronchodilator and decline in forced expi- ratory volume in one second in population studies. Am Rev RespirDis 1985; 132: 1186-1193. 28. Lander ES, Schork NJ. Genetic dissection of complex traits. Science 1994; 265: 2037-2048. 29. Knowler WC, Williams RC, Pettitt DL Steinberg AG. Gra3;5,t3,14 and type 2 diabetes mellitus: an associa- tion in American Indians with genetic admixture. Am J Hum Genet 1988; 43: 520-526. 30. Laurell CB, Eriksson S. The electrophoretic alphas- globulin pattern of serum in alpha~-antitrypsin deficiency. Scand J Clin Invest 1963; 15: 132-140. 31. Mornex J-F, Chytil-Weir A, Martinet Y, Courtney M, LeCocq J-P, Crystal RG. Expression of the alphas-anti- trypsin gene in mononuclear phagocytes of normal and alphal-anfitrypsin deficient individuals. J Clin Invest 1986; 77: 1952-1961. 32. Cox DW, Johnson AM, Fagerhol MK. Repor~ of nomen- clature meeting for alpha~-antitrypsin. Hum Genet 1980; 53: 429-433. 33. Fagerhol MK, Lanrell CB. The Pi system: inherited variants of serum ct~- antitrypsin. Prog Med Genet 1970; 7: 96-111. 34. Allen RC, Harley RA, Talamo RC. A new method for determination of alpha|-antitrypsin phenotypes using iso- electric focusing on polyacrylamide gel slabs. Am J Clin Pathol 1974; 62: 732-739. 35. Birrer P, McElvaney NG, Chang-Stroman LM, Crystal RG. Alpha~-antitrypsin deficiency and liver disease. J Inher Metab Dis 1991; 14: 512-525. 36. Black LF, Kueppers F. Alpha|-antitrypsin deficiency in nonsmokers. Am Rev Respir Dis 1978; 117: 421-428. 37. Janus ED, Phillips NT, Carrell RW. Smoking, lung function and alphal-antitrypsin deficiency. Lancet 1985; i: 152-154. 38. Brantley ML, Paul LD, Miller BH, Falk RT, Wu M, Crystal RG. Clinical features and history of the destruc- tive lung disease associated with alpha~-antitrypsin defi- ciency of adults with pulmonary symptoms. Am Rev Respir Dis 1988; 138: 327-336. 39. Tobin M J, Cook PJL, Hutchinson DCS. Alpha|-anti- trypsin deficiency: the clinical and physiological features of pulmonary emphysema in subjects homozygous for Pi type Z: a survey by the British Thoracic Association. Br J Dis Chest 1983; 77: 14-27. 40. Sveger T. Liver disease in alpha|-antitrypsin deficiency detected by screening of 200,000 infants. N Engl J Med 1976; 294: 1316-1321. 41. O'Brien ML, Buist NRM, Murpbey WH. Neonatal screen- ing for alpha~-antitrypsin deficiency. J Pediatr 1978; 92: 1006-1010. 42. Shigeoka JW, Hall WL Hyde RW, et aL The preva- lence of alpha~-antitrypsin heterozygotes (PiMZ) in • patients with obstructive pulmonary disease. Am Rev Respir Dis 1976; 114: 1077-1084. 43. Bartmann K, Fooke-Achterrath M, Koch G, et al. Heterozygosity in the Pi system as a pathogenetic co- factor in chronic obstructive pulmonary disease (COPD). Eur J Respir Dis 1985; 66: 284-296. 44. Cox DW, Hoeppner VH, Levison H. Protease inhibitors in patients with chronic obstructive pulmonary disease: the aipha~-anfitrypsin heterozygote controversy. Am Rev Respir Dis 1976; 113: 601-606. 45. Janus ED. Alpha|-antitrypsin Pi types in COPD patients. Chest 1988; 94: 446-447. 46. Lieberman J, Winter B, Sastre A. Alpha~-antittypsin Pi types in 965 COPD patients. Chest 1986; 89: 370-373. 47. Mittman C, Lieberman J, Rumsfeld J. Prevalence of abnormal protease inhibitor phenotypes in patients with chronic obstructive lung disease. Am Rev Respir Dis 1974; 109: 295-296. 48. Barnett TB, Gottovi D, Johnson AM. Protease inhibitors in chronic obstructive pulmonary disease. Am Rev Respir Dis 1975; 111: 587-593. 49. Webb DR, Hyde RW, Schwartz RH, Hall WJ, Condemi JJ, Townes TL. Serum a~-antitrypsin variants. Am Rev Respir Dis 1973; 108: 918-925. 50. Cole RB, Nevin NC, Blundell G, Merrett JD, McDonald JR, Johnston WP. Relation of alphal-antitrypsin pheno- type to the performance of pulmonary function tests and to the prevalence of respiratory illness in a working pop- ulation. Thorax 1976; 31: 149-157. 51. Morse JO, Lebowitz MD, Knudson ILl, Burrows B. Relation of protease inhibitor phenotypes to obstructive lung diseases in a community. N Engl J Med 1977; 296:1190-1194. 52. Lebowitz MD, Knudson R.l, Morse JO, Armet D. Closing volume and flow volume abnormalities in alphas-anti- trypsin phenotype groups in a community population. Am Rev Respir Dis 1978; 117: 179-181. 53. Chan-Yeung M, Ashley M J, Corey P, Maledy H. Pi phenotypes and the prevalence of chest symptoms and lung function abnormalities in workers employed in dusty industries. Am Rev Respir Dis 1978; 117: 239-245. 54. McDonagh DJ, Nathan SP, Knudson RJ, Lebowitz MD. Assessment of alpha~-antitrypsin deficiency heterozy- go~ity as a risk factor in the etiology of emphysema. J Clin Invest 1979; 63: 299-309. 55. Buist AS, Sexton GJ, Azzarn A-MH, Adams BE. Pulmo- nary function in heterozygotes for alpha~-antitrypsin deficiency: a ease-control study. Am Rev Respir Dis 1979; 120: 759-766. 56. Bruce RM, Cohen BH, Diamond EL, et al. Collaborative study to assess risk of lung disease in Pi MZ phenotype subjects. Am Rev Respir Dis 1984; 130: 386-390. 57. Klayton R, Fallat R, Cohen AB. Determinants of chro- nic obstructive pulmonary disease in patients with inter- mediate levels of alphal-antitrypsin. Am Rev Respir Dis 1975; 112: 71-75. 58. Cooper DM, Hoeppner V, Cox D, Zamel N, Bryan AC, Levison H. Lung function in alpha~-antitrypsin betero- zygotes (Pi type MZ). Am Rev Respir Dis 1974; 110: 708-715. 59. Tattersall SF, Pereira RP, Hunter D, Blundell G, Pride NB. Lung distensibility and airway function in inter- mediate alphal-antitrypsin deficiency. Thorax 1979; 34: "637-646. 60. Hall WJ, Hyde RW, Schwartz RH, et al. Pulmonary abnormalities in intermediate alpha~-antitrypsin defi- ciency. J Clin Invest 1976; 58: 1069-1077. 61. Tfi'jan E, Magyar P, V~iczi Z, Lantos /~, Vaszfir L. Longitudinal lung function study in heterozygous PiMZ phenotype subjects. Eur Respir J 1994; 7: 2199-2204. 62. Owen MC, Brennan SO, Lewis JH, Can'ell RW. Mutation "- of antitrypsin to antithrombin: oq-antitrypsin Pittsburgh (358 met---~arg), a fatal bleeding disorder. N Engl J Med 1983; 309: 694-698. 63. Kalsheker NA, Hodgson IJ, Watkins GL, White JP, Morrison HM, Stockiey RA. Deoxyribonucleic acid (DNA) polymorphism of the a|-antitrypsin gene in chro- nic lung disease. Br Med J 1987; 294: 1511-1514. 64. Poller W, Meison C, Olek K. DNA polymorphisms of the ~|-antitrypsin gene region in patients with chronic obstructive pulmonary disease. Eur J Clin Invest 1990; 20: 1-7. 65. Kalsheker NA, Watkins GL, Hill S, Morgan K, Stockley