Abstract
Chronic obstructive pulmonary disease (COPD) is a prevalent and preventable disease associated with high morbidity and mortality. Severe and intermediate α1-antitrypsin (AAT) deficiency (serum levels <11 and 11–20 µmol·L−1, respectively) increase the risk of COPD in active smokers. However, little is known about the interaction of severe and intermediate AAT deficiency with modifiable COPD risk factors other than active smoking.
In this study, a MEDLINE search was carried out for studies investigating the combined effect of environmental inhalants (occupation and passive smoking) and AAT deficiency in the lung. A total of 18 studies using established methods for the assessment of AAT deficiency were included in this review.
Occupational exposures and passive smoking affected lung function decline or prevalence of respiratory symptoms in four out of five studies investigating subjects with severe AAT deficiency, and in eight out of 13 studies with a focus on intermediate AAT deficiency. While study designs mostly prohibited formal assessment of effect modification, an interaction between intermediate AAT deficiency and passive smoking was identified in two studies with children. Additional study limitations included small sample size, poor adjustment for confounding and misclassification of environmental exposure as well as AAT activity.
In conclusion, population-based epidemiological studies with associated biobanks are needed to identify gene–environment interactions and population subgroups susceptible to α1-antitrypsin deficiency.
- α1-Antitrypsin
- α1-antitrypsin deficiency
- gene–environment interaction
- occupational disorder
- occupational exposure
- passive smoking
α1-ANTITRYPSIN DEFICIENCY AND COPD
Aetiology and progression of chronic obstructive pulmonary disease (COPD) result from a complex interplay between genetic and environmental factors. More than 90% of COPD patients are current or ex-smokers. Thus, the environment is clearly of importance in disease development. However, the effect of cigarette smoking on pulmonary function is highly variable, suggesting a role for genetic susceptibility in the response to tobacco exposure 1, 2. Variations in the gene coding for α1-antitrypsin (AAT), the most abundant protease inhibitor circulating in the blood, is the only established genetic risk factor for COPD 3–5.
AAT is an acute-phase protein mainly produced by the liver. It protects the lung tissue from destruction by neutrophil elastase 6, 7. Absence or dysfunction of AAT leads to a shift in the protease–antiprotease balance in the lung and increases its susceptibility for the development of emphysema 8. The molecular basis for AAT deficiency is mostly genetic variation in the AAT gene, SERPINA1 9. AAT belongs to the superfamily of serine protease inhibitors (serpins) 10. The AAT gene and protein are highly polymorphic. As of May 2005, 186 single nucleotide polymorphisms (SNPs) in the AAT gene were listed in public databases (http://snpper.chip.org) 11. With isoelectric focusing (IEF), >90 protein variants, referred to as protease inhibitor (Pi) phenotypes, have been identified 12, 13.
Assessment of AAT deficiency in the clinical setting and in epidemiological studies has mostly been restricted to IEF. A common AAT classification divides the Pi variants into normal, deficient and null categories (no detectable AAT serum level), according to the serum concentration of the AAT protein 5. Protein variants associated with normal AAT concentrations are referred to as PiMM phenotype (table 1⇓). Several rare genetic variants mediate AAT deficiency. The most prevalent mutations in Caucasians are two SNPs occurring at a frequency of 0.6–11% (S-allele) and 0.3–4% (Z-allele) 14. In very rare instances, total AAT deficiency results from inheriting two null alleles 15. Thus, intermediate and severe AAT deficiency phenotypes in Caucasians mostly result from combinations of S-, Z- and null alleles. Severe AAT deficiency (i.e. AAT levels below a protective threshold of 11 µmol·L−1) includes subjects homozygous or heterozygous for the Z- or the null allele. Intermediate AAT deficiency includes subjects with PiMZ, PiSS and PiSZ phenotypes. Their serum levels range from 20–60% of normal (i.e. AAT levels between 11 and 20 µmol·L−1).
In patients with severe AAT deficiency, active smoking remains an important additional risk factor for the development of COPD 17–19. Active smoking was also associated with lung function deficits in individuals with intermediate AAT deficiency 20–22. In a study of PiSZ subjects, 29% of the never-smokers and 82% of the ex- and current smokers exhibited a predicted post-bronchodilator forced expiratory volume in one second (FEV1) <80% 20. Eriksson et al. 21 reported a significantly higher mean annual decrease in FEV1 in smoking PiMZ subjects as compared with nonsmoking PiMZ individuals. In a matched case–control study, significant differences in pulmonary function were observed between smoking, but not between nonsmoking, PiMZ and PiMM subjects 22.
Little is known about the interaction of severe and intermediate AAT deficiency with modifiable COPD risk factors other than active smoking in respiratory health, i.e. passive smoking, air pollution and occupational inhalants 23–25.
AAT DEFICIENCY AND MODIFICATION BY OCCUPATIONAL AND ENVIRONMENTAL INHALANTS: SUMMARY OF THE EVIDENCE
In this study, a MEDLINE search was conducted for studies investigating the combined effect of passive smoking/occupational exposures and AAT deficiency in the lung. A search for publications between January 1966 and October 2004 revealed 1,086 articles and included the terms “alpha1-antitrypsin”, “alpha1-antitrypsin deficiency” and “protease inhibitor”, in combination with “lung disease”. Inclusion of the search terms “environment”, “occupational exposure” and “occupational disorder”, as well as a search for additional references in pertinent reviews, resulted in 110 articles. Restriction to publications in English language and studies using established methods for the assessment of AAT deficiency revealed 18 articles, which assessed AAT status and environmental exposures other than active smoking (table 2⇓).
Five out of the 18 studies investigated children or adolescents 28, 39, 41–43. Health outcomes included continuous lung function parameters 27–33, 35–37, 41, 42, dichotomised lung health indicators including respiratory symptoms 29–32, 34–37, 41, 43, bronchial hyperresponsiveness 39 and FEV1 ≤65% predicted 26. Two publications investigated the impact of AAT deficiency on highly specific occupational disorders, such as byssinosis 38 and asbestosis 40. Investigated environmental and occupational modifiers of an AAT effect were primarily related to occupational exposures (gas, fumes, dusts, endotoxin, asbestos) 26, 27, 29–32, 34–40, with a few studies investigating exposure to passive smoking and industry and traffic-related air pollution 26, 28, 29, 33, 41–43. The combined effect of AAT deficiency and environmental exposures on lung health has been investigated predominantly in cross-sectional studies, except for one case–control study 40, and two studies reporting longitudinal data 37, 43. Five studies investigated occupational risks and passive smoking exposure in persons with severe AAT deficiency. As none of these studies included a comparison group of subjects without AAT deficiency, the modifying effect of environmental exposures on severe AAT deficiency could not be formally assessed (table 3⇓). The remaining 13 studies addressed environmental risks other than active smoking in intermediate AAT-deficient subjects (table 2⇑). Five of these studies allowed for formal assessment of effect modification; three of them actually tested for interaction between environmental exposures and AAT deficiency (table 4⇓).
Severe AAT deficiency and occupational exposure
In three out of four studies, subjects with severe AAT deficiency had significantly impaired lung function or reported more respiratory symptoms when exposed to occupational risks such as gas, fumes and dust, or to indoor air pollution caused by domiciliary kerosene heaters, when compared with nonexposed subjects 27, 29, 30. The observed combined effects of AAT deficiency and environmental exposures are likely to be independent of active smoking as two studies were restricted to nonsmoking PiZZ subjects 27, 29 and one study controlled for pack-years smoked 30. The study not confirming the impact of occupational exposure on lung health in subjects with severe AAT deficiency was hampered by an unadjusted statistically significant difference in the number of pack-years of cigarettes smoked and in age between the two comparison groups (FEV1 ≤65% pred versus >65% pred) 26. In one study, the effect was restricted to subjects >50 yrs 27.
Severe AAT deficiency and exposure to environmental tobacco smoke
In nonsmoking PiZZ individuals, Piitulainen et al. 29 did not find a difference in FEV1 and vital capacity (VC) between individuals with and without environmental tobacco smoke (ETS) exposure. However, the subgroup exposed to ETS for ≥10 yrs had a higher prevalence of self-reported chronic bronchitis. A cross-sectional analysis in 128 adolescents with severe AAT deficiency at the age of 18 yrs (range 17.7–19.9), drawn from the Swedish AAT newborn screening programme, revealed a lower FEV1/forced vital capacity (FVC) ratio in the subgroup exposed to parental smoking compared with the nonexposed group 28. In contrast to the previously discussed multivariate findings, the univariate analysis by Silverman et al. 26 found similar proportions of severe AAT-deficient subjects exposed to other smokers in the household in the low (FEV1 ≤65% pred) and high (FEV1 >65% pred) group.
Intermediate AAT deficiency and occupational exposure
Nine studies compared lung health parameters between subjects with and without intermediate AAT deficiency who were exposed to occupational dust and gas. Bronchial hyperresponsiveness was more common in subjects with intermediate AAT deficiency, and they exhibited lower FEV1 levels as opposed to those with normal AAT status 36, 39. In a longitudinal study that investigated 871 iron-ore workers over a period of 5 yrs 37, the rate of decline in the FEV1/FVC ratio over the follow-up period was significantly increased in intermediate deficiency phenotypes (PiMS, PiMZ) (−3.9±8.0) when compared with the PiMM group (−1.8±8.6). However, the investigators did not find a difference in the incidence of respiratory symptoms.
Intermediate AAT deficiency was also related to the risk of typical occupational diseases, namely endotoxin-related byssinosis and asbestosis 38, 40. After adjusting for potential confounders, the PiMZ phenotype or a serum AAT level <35 mmol·L−1 were related to a significant increased odds ratio (OR (95% confidence interval (CI))) for byssinosis (5.8 (1.1–30.3) for PiMZ and 5.0 (1.4–17.7) for AAT level ≤35 mmol·L−1) 38. PiZ heterozygosity or PiSS was significantly more prevalent in asbestosis cases compared with control workers with a similar asbestos exposure (8.0 (1.6–39.1)) 40.
Several cross-sectional surveys in working populations exposed to a variety of occupational inhalants failed to show an association between AAT status and lung function deficits or respiratory symptoms 31, 32, 35, 39. Small sample sizes in the PiMS (n = 18 and n = 25, respectively) and PiMZ (n = 5 and n = 2, respectively) groups 31, 32 and unadjusted confounding by active smoking 35 could have, in part, contributed to these inconsistencies. One study investigated an age, sex, and smoking stratified sample of inhabitants in a moderately polluted industrial community 33. Lung function measurements did not differ between the PiMZ, PiMS and PiMM groups. Again, not taking into account the observed differences in duration of residence in the community and smoking habits between the PiM and PiMS subjects might have biased these results.
Interaction between occupational exposure and AAT deficiency has not been formally assessed to date.
Intermediate AAT deficiency and exposure to environmental tobacco smoke
Two out of three studies that allowed for the formal investigation of interaction between ETS and AAT deficiency found evidence that passive smoking significantly modified the effect of intermediate AAT deficiency 41, 42. They were based on cross-sectional investigations on random samples of school children aged 9–13 yrs. Statistically significant interactions between parental smoking exposure and AAT status (PiMM versus PiM heterozygotes) have been reported for the FEV1/FVC ratio and maximum expiratory flow at 50% of VC (FEF50 L·s−1) 42. While no association between passive smoking with FEF50 and FEV1/FVC was present in PiM homozygote children, PiM heterozygotes exposed to ETS exhibited lung function deficits of −2.57% (−4.31– −0.83) for the FEV1/FVC ratio and −0.43 (−0.72– −0.12) for FEF50. These findings from Italy are consistent with a study by von Ehrenstein et al. 41 investigating the relationship between lung function, ETS exposure and plasma AAT levels in a random sample of German school children. Pronounced asymptomatic decrements in pulmonary function were observed in children with low AAT plasma levels (≤116 mg·dL−1, defined as the 5th percentile of the sample distribution) and exposure to ETS. Testing for interaction between low AAT plasma level and exposure to ETS revealed statistically significant results for FEV1 (% pred), FVC (% pred), FEF25 (% pred) and FEF25–75 (% pred). Similar decrements in lung function parameters in children with low AAT levels and exposure to truck traffic (>730 trucks a day in streets in a radius of 100 m around each child's home) were observed, but the interaction did not reach statistical significance.
In a national birth cohort, intermediate AAT deficiency due to PiS and PiZ variants was an independent risk factor for infant lower respiratory infection experienced by 2 yrs of age 43. The study did not find an association with a SNP in the enhancer 3′ region of the AAT gene (G1237A) thought to be associated with an impaired inflammatory response 44. The effects were not modified by parental smoking and atmospheric pollution. No associations of AAT status with adult respiratory outcomes (FEV1 decline between 43 and 53 yrs of age and respiratory symptoms, respectively) were found. The retrospective assessment of parental smoking at the age of 53 yrs and difficulties in the assignment of exposure to air pollutants are a potential source of misclassification of environmental exposure.
AAT DEFICIENCY AND MODIFICATION BY OCCUPATIONAL AND ENVIRONMENTAL INHALANTS: LIMITATION OF THE EVIDENCE
A formal meta-analysis of the available evidence was not feasible. A broad spectrum of respiratory health parameters, on the one hand, and of environmental exposures, on the other hand, was investigated. Few studies provided appropriately adjusted effect estimates; the type of effect estimates presented varied widely. Study designs varied greatly.
The interpretation of the available evidence with regard to the interaction between AAT deficiency and environmental inhalants is hampered by several factors associated with study design. Most studies did not allow for formal assessment of effect modification because they were either restricted to subjects with AAT deficiency or to subjects exposed to specific inhalants. The additional limitations of the presented studies are related to the assessment and assignment of an individual environmental exposure level, to differences in the ascertainment of the study population (i.e. index versus nonindex cases; occupational versus population-based study groups), and to the lack of power due to small sample sizes, as well as to unadjusted or residual confounding by active smoking and other factors.
A major limitation for investigating the interaction between AAT and environmental risks was the available methods for the evaluation of AAT deficiency. The assessment of AAT deficiency generally suffers from two major problems as follows.
The first problem relates to the fact that AAT activity in the lung epithelium, the target tissue of interest, cannot be directly inferred from serum AAT levels. In previous studies, assessment of AAT deficiency was restricted to capturing the most prevalent protein phenotypes known to be associated with serum AAT concentrations (i.e. IEF for the M, S and Z protein) or to determining serum AAT concentrations exclusively. Little is known about the correlation between AAT quantities in the interstitium of the lung and AAT serum concentrations. It has been demonstrated that AAT serum levels do not correlate with levels of AAT in bronchoalveolar lavage in patients with severe AAT deficiency after intravenous administration of human plasma-derived AAT 45. Furthermore, results on serum AAT concentrations may vary considerably depending on the commercially available standards and methods used for quantitative determination of serum AAT (i.e. radial immunodiffusion, nephelometry) 46.
The second problem is related to the measurement of functionality of a specific AAT protein. While genotypic information about an individual's AAT alleles may better reflect local AAT activity in the lung rather than serum AAT concentrations, additional problems arise for correlating genotyping results with expression and functionality of AAT in the target tissue. In the studies discussed, assessment of genetic variation was not exhaustive, since only the most prevalent AAT polymorphisms were captured through IEF phenotyping. Some common and rare AAT deficiency alleles have been reported to also alter its functional activity 47–49. For example, polymerised Z-AAT protein acts as a neutrophil chemoattractant in the lung 47, 49, and the rare PiMmineral springs allele has a reduced elastase inhibitor capacity 48. While these two alleles are associated with both low levels of AAT in the blood and modified functional activity, dysfunction of the AAT protein can be present even in the absence of AAT deficiency in the blood. Reactive oxygen species (ROS) and genetic variation in the regulatory site of the AAT gene locus has been shown to result in reduced AAT activity. Cigarette smoke, a potent source of ROS, has been found to reduce serum and alveolar AAT anti-elastase activity in healthy smokers in comparison with nonsmokers 50–52. Furthermore, a mutation in a 3′ enhancer region of the AAT gene is associated with normal basal protein expression, but has been reported to affect the acute-phase reaction, resulting in a diminished AAT response to inflammation 44, 53. Questions also remain about how to define activity of specific AAT genotypes and phenotypes. It is generally hypothesised that the role of AAT in the pathogenesis of emphysema acts through the inactivation of neutrophil elastase. However, additional molecular mechanisms, such as anti-inflammatory AAT effects 54, and antiproteolytic activity against other toxic metabolites and proteases 55 involved in lung inflammation may be of pathophysiological relevance, yet have not been taken into consideration for the classification of functionality of various genetic AAT polymorphisms. Finally, assessment of variation in the AAT gene itself may not adequately capture all of the individual variation in the production of AAT in the liver. It is likely that unrecognised genetic variants in genes modulating AAT expression are additional determinants of AAT activity.
CONCLUSION AND FUTURE RESEARCH NEEDS IN EPIDEMIOLOGICAL AAT RESEARCH
In summary, various occupational inhalants and passive smoking were found to impact on respiratory health in subjects with severe AAT deficiency. An association of intermediate AAT deficiency with respiratory health parameters in subjects exposed to occupational inhalants was reported by some, but not all, studies. The interaction between AAT deficiency and environmental inhalants was formally assessed in three studies only. In two of these studies, ETS was associated with lung function deficits in children with intermediate AAT deficiency and low serum AAT concentrations, respectively. Large, population-based epidemiological studies with associated biobanks are now needed to formally investigate the interaction between intermediate AAT deficiency and prevalent environmental inhalants, such as ETS and air pollutants, in both children and adults.
The public health relevance of the studies presented here is several-fold. First, COPD is a prevalent and preventable disease associated with high morbidity and mortality. AAT research that aims to identify susceptible population subgroups has the potential for targeted counselling and prevention. Secondly, intermediate AAT deficiency is prevalent as the report by de Serres 14 estimated at least 116 million carriers of at-risk alleles in the AAT gene worldwide. Thirdly, the study of the interplay between ambient air pollutants and genetic variation in SERPINA1 will be of special public health relevance given the high morbidity and mortality associated with current concentrations of air pollutants worldwide 56.
From a genetic perspective, the exhaustive identification of polymorphisms in and around the SERPINA1 gene and the investigation of haplotypes as opposed to single gene polymorphisms will be a great priority. While this is true for any future study on gene–environment interactions, it seems particularly important for AAT, given its highly polymorphic nature and its chromosomal proximity to additional genes with antiprotease activity in the serpin cluster (i.e. α1-antichymotrypsin, protein-c inhibitor and corticosteroid-binding protein). The ongoing HapMap Project 57 aims to identify “tagSNPs” representing the most frequent haplotypes. For the AAT gene locus, there is evidence about considerable allelic associations throughout the serpine cluster and a unique haplotype associated with the Z-allele has been reported 58.
In the development of a common complex disease such as chronic obstructive pulmonary disease additional genes may play an independent role and/or interact with α1-antitrypsin expression 23, 59, 60. Expanding the common single candidate gene approach to a “candidate pathways” (i.e. inflammation, oxidative stress) approach should provide further insights into risk factor patterns underlying chronic obstructive pulmonary disease.
- Received February 24, 2005.
- Accepted June 9, 2005.
- © ERS Journals Ltd