Microsomal Epoxide Hydrolase, Endotoxin, and Lung Function Decline in Cotton Textile Workers

Epidemiologic studies have shown that chronic occupational exposure to organic dust may induce chronic airway disease (1-3). Among many agents present in organic dusts, there is substantial evidence to suggest that bacterial endotoxin is a major causative agent that contributes to airway inflammation and airflow obstruction (4, 5). The concentration of inhaled endotoxin in bioaserosol is associated strongly with the development of airway disease in cotton and agricultural workers (4, 6, 7). Exposure to endotoxin may result in endothelial cell damage (8), influx of blood leukocytes into the airway (9), production of different cytokines (10, 11), as well as increased production of oxygen metabolites, such as reactive oxygen species (ROS) that may induce oxidative lung injury (12, 13).

The epoxide hydrolases are a family of enzymes that hydrate simple epoxide to vicinal diols and arene oxides to transdihydrodiols (14). Among this family, microsomal epoxide hydrolase (mEH) plays an important role in the detoxification of ROS and reactive epoxide intermediates (15). In the coding region of the mEH gene, two relatively common genetic polymorphisms are characterized within exon 3 and exon 4. In exon 3, the Tyr113His (C113T) polymorphism results in a 40 to 50% decrease in enzyme activity, and, thus, the His allele has been called the "slow allele". In exon 4, the His139Arg (G139A) polymorphism is associated with a 25% increase of enzyme activity, with the Arg allele termed the "fast allele" (16). Slow activity mEH genotypes have been associated with chronic obstructive pulmonary disease (COPD) in some studies (17-21), but not in others (22-26).

Our previous studies have shown that long-term occupational exposure to cotton dust may induce a faster decline of lung function, and this decline probably is more closely related to high concentrations of cndotoxin in cotton dust (6). mEHs metabolize the ROS generated by endotoxin exposure, therefore, polymorphisms of mEH with different enzyme activities may modify the associations between endotoxin exposure and lung function decline. Specifically, we hypothesized that the associations between endotoxin exposure and lung function decline will be stronger in genotypes with "slow" enzyme activities than in genotypes with "fast" enzyme activities. We tested these prior hypotheses in our 20-year longitudinal cohort of cotton textile workers.

METHODS

Additional detail is provided in the online supplement.

Subjects

The study population was derived from a population-based, 20-year longitudinal cohort for respiratory disease among cotton textile workers in Shanghai, China, which was established in 1981. Details of this cohort have been described previously (4). The first survey included 919 subjects (447 cotton and 472 silk workers), and 4 follow-up surveys were performed every 5 years in the following 20 years. The study was approved by the Institutional Review Boards of the Harvard School of Public Health, the Putuo District People's Hospital, and the Human Resources Administration of China. Written informed consent was obtained from each participant.

Exposure Assessment and Pulmonary Function Tests

Airborne cotton dust was measured with a Vertical Elutriator (General Mctalworks Corp., Mequon, WI), and gram-negative bacteria endotoxins were analyzed using the chromogenic Limulus amoebocyte lysate assay, as previously described (27, 28). Forced expiratory maneuver was performed at each survey (4). Annual decline in FEV^sub 1^ was calculated on the basis of the differences in preshift FEV^sub 1^ measurements between the last and first survey and was divided by 20 (years).

Genotyping

DNA was extracted from whole blood using a standard method. The mEH Tyr113His and His139Arg polymorphisms were genotyped by the 5' nuclease assay (TaqMan) using the ABI Prism 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA). The primers, probes, and reaction conditions are available on request.

Statistical Analysis

Population characteristics were tabulated, and significant differences in the distribution of the principal covariates were tested using the Chi-square, Fisher exact, and Student's t tests, where appropriate. Multivariate linear regression models were used to explore the association between mEH polymorphisms, alone or in combination with endotoxin exposure, and the annual decline rate of FEV^sub 1^ (as continuous variable), before and after adjusting for covariates, including age, sex, height, baseline FBV^sub 1^, smoking status, and packycars of smoking (29). For the His139Arg polymorphism, the His/Arg and Arg/Arg genotypes were combined because of low frequency for the Arg/Arg genotype. In all of the analyses, the Tyr/Tyr genotype of the Tyr113His polymorphism and His/Arg + Arg/Arg genotype of the His139Arg polymorphism were used as reference category, respectively, on the basis of the convention of faster enzyme activity for these genotypes. Three types of models were used to fit the data. Model 1 was fit to determine the overall association between mEH polymorphisms and lung function decline (main effect). At the same time, interactions among mEH polymorphisms, endotoxin exposure, smoking, as well as sex and age were determined, and they were retained in the model if reaching significance. Model II was to examine the modification effect of mEH polymorphisms on the association between endotoxin exposure and lung function decline, i.e., stratified analysis of endotoxin exposure and lung function decline in each genotype of mEH polymorphisms. Model III was used to investigate the joint effects of mEH polymorphisms and endotoxin exposure on the decline of lung function, using the combined mEH "fast" genotype (the two polymorphisms were analyzed in different models) and non-endotoxin-exposed as the common reference group. Statistical analyses were undertaken using the SAS personal computer software (version 8.2, 1999) (SAS Institute, Cary, NC).

RESULTS

Endotoxin Exposure Data

A total of 802 fullshift air samples were collected over the first four surveys in the yarn preparation areas of the two cotton mills and the silk mill. For the two cotton mills, geometric mean dust exposure of the work areas ranged from 0.2 to 1.6 mg/m^sup 3^ and geometric mean endotoxin of the work areas ranged from 27 to 12,038 EU/m^sup 3^. Because cotton dust and endotoxin levels in the silk mill were identified as nearly zero over the period of the follow-up study (4, 30), we stratified this population into two groups (non-endotoxin-exposed and endotoxin exposed) on the basis of occupation (silk or cotton workers).

Population Characteristics

Five hundred one of 919 subjects participated in the last survey and donated blood samples, including 234 (49.58%) in the non-endotoxin-exposcd group and 267 (59.73%) in the endotoxin-exposed group. There were no significant differences in age, sex, height, smoking status, and baseline FEV^sub 1^ between subjects with and without blood samples, in both nonexposed and exposed groups (p > 0.05).

A total 499 of 501 subjects were genotyped successfully for both Tyr113His and His139/Arg polymorphisms, including 234 in the non-endotoxin-exposed group and 265 in the endotoxin-exposcd group. Details of this population were presented in Table 1. There were no significant differences between non-endotoxin-exposed and endotoxin-exposed groups in age, sex, height, smoking status, and baseline FEV^sub 1^ (p > 0.05). The smoking amount among smokers was higher in the endotoxin-exposed group, with mean (SD) packyears of 29.04 (20.47) in endotoxin-exposed and 24.88 (15.87) in nonexposed groups, respectively (p < 0.05).

Genotype Distributions

Genotype frequencies of both Tyr113His and His139Arg polymorphisms were in Hardy-Weinberg equilibrium in both non-endotoxin-exposed and exposed groups (p > 0.05 by Chi-square test). There were no significant differences in the frequencies of mEH genotypes between non-endotoxin-exposed and exposed groups. For the Tyr113His polymorphism, the frequencies of Tyr/Yyr, Tyr/His, and His/His genotypes were 33.76%, 50.00%, and 16.24% in the non-endotoxin-exposed group, and 30.19%, 49.43%, and 20.38% in the exposed group, respectively (p = 0.44). For the His 139Arg polymorphism, genotype frequencies of Arg/Arg, His/Arg, and His/His were 0.85%, 23.08%, and 76.07% in the non-endotoxin-exposed group and 1.51%, 21.13%, and 77.36% in the exposed group, respectively (p = 0.71, Table 1).

mEH Polymorphisms and Lung Function Decline

Age, sex, height, smoking status, packyears of smoking, and baseline FEV^sub 1^ were all included in the multivariate linear regression model initially (Table 2). Sex and age were excluded in the final models because they were highly correlated with baseline FEV^sub 1^ (r = 0.59 between sex and baseline FEV^sub 1^ and r = 0.74 between height and baseline FEV^sub 1^, p < 0.001) and they turned out to be insignificant after baseline FEV^sub 1^ was included. In all of the analyses, age and packyears of smoking baseline FEV^sub 1^ were fitted as continuous variables.

The overall annual FEV^sub 1^ decline rate was -29.47 ml in the study population ("-" represents the direction of decline). Age, packyears of smoking, baseline FEV^sub 1^, and endotoxin exposure were all associated with an annual decline rate of FEV^sub 1^ (p < 0.01; Table 2). Among these, exposure to endotoxin caused a -3.23 ml (SE = 1.14) annual decline of FEV^sub 1^. There were no statistically significant associations between mEH polymorphisms and lung function decline (Model I). For the Tyr113His polymorphism, when compared with the Tyr/Tyr genotype, estimates of the annual decline rate of FEV^sub 1^ in Tyr/His and His/His genotypes were 0.03 ml (SE = 1.31) and -0.61 ml (SE = 1.69), respectively. For the His139Arg polymorphism, the estimate of the annual decline rate of FEV^sub 1^ for the His/His genotype was -1.87 ml (SE = 1.36) when compared with the combined Arg/Arg + His/Arg genotypes.

Second, we investigated the modification effect by mEH polymorphisms on the association between endotoxin exposure and the rate of lung function decline (Model II). Although endotoxin exposure was associated with a faster lung function decline in each mEH genotype group, the differences between exposed and nonexposed groups increased with the mEH enzyme activity decreased for both polymorphisms. For the Tyr113His polymorphism, the differences in estimates of the annual decline rate of FEV^sub 1^ between endotoxin-exposed and nonexposed groups were -2.33 ml (SE = 2.07, p = 0.26), -2.81 ml (SE = 1.66, p = 0.09), and -6.73 ml (SE = 2.83, p < 0.05) for the Tyr/Tyr, Tyr/His, and His/His genotype groups, respectively. For the His139Arg polymorphism, the difference between endotoxin-exposed and nonexposed groups was -1.82 ml (SE = 2.58, p = 0.48) and -4.27 ml (SE = 1.33, p < 0.01) for the Arg/Arg + His/Arg and His/His genotype groups, respectively (Table 3). When stratifying smoking status, we found that although cigarette smoking was related to additional lung function decline in all of the genotype groups, the modification effect of mEH polymorphisms on endotoxin exposure existed in both never and ever smokers (Table 3), and the three-way interactions between endotoxin exposure, smoking, and mEH polymorphisms were not statistically significant (p > 0.05).

Lastly, we investigated a potential joint effect of mEH polymorphisms and endotoxin exposure on the rate of lung function decline (Model III). In this model, five (Tyr113His polymorphism) or three (His139Arg polymorphism) dummy variables were created for the combination of mEH genotypes and endotoxin exposure. The results from this model were similar to the results from the gene-endotoxin exposure interaction model. For the Tyr113His polymorphism, the combined His/His genotype and endotoxin-exposed group was associated with the fastest annual FEV^sub 1^ decline (-4.91 ml, SE = 2.26, p = 0.03), when compared with the reference group of the combined Tyr/Tyr genotype and non-endotoxin-exposed group. A similar association was found for the His139Arg polymorphism: the combined His/His genotype and endotoxin-exposed group was associated with the fastest annual FEV^sub 1^ decline (-4.37 ml, SE = 1.91, p = 0.02), when compared with the reference group of combined Arg/Arg + His/Arg genotype and nonexposed group (Table 4).

DISCUSSION

Occupational exposure to cotton dust has been shown to cause airway disease characterized by acute changes in airflow and by chronic obstructive lung disease (30). Endotoxin is the principal component of cotton responsible for the development of inflammation of the airway and airflow obstruction (5). LPS intraperitoneal injection results in a significant increase of ROS production by peritoneal leukocytes, generating an oxidative stress situation (12). In vitro, after the human monocyte THP-1 cell line was treated with different doses of LPS, LPS elicited a dose-dependent oxidative burst (13). The oxidant/antioxidant theory postulates that an excess of oxidants in the lung promotes cellular and tissue damage, and is the major initiator of the disease process (31, 32). Our previous results have suggested that chronic occupational endotoxin exposure is associated with faster decline of lung function (4).

Although endotoxin exposure is associated with oxidative stress and airway inflammation, not all workers who are exposed to endotoxin develop airway disease. Healthy volunteers who were challenged with LPS had different responsiveness to inhaled LPS (33), suggesting that host factors may play an important role in the response to endotoxin exposure. In this study, we investigated the relationship between mEH polymorphisms and the annual decline rate of lung function in a population chronically exposed to endotoxin. Our results suggest that the mEH His/His genotypes of both Tyr113His and His139Arg polymorphisms associated with lower enzyme activities may increase lung function decline induced by endotoxin exposure. The modification effects of mEH polymorphisms on the association between endotoxin exposure and faster decline of lung function existed in both never and ever smokers.

The mEH enzyme is an important Phase II biotransformation enzyme, and it is highly expressed in several human tissues including bronchial epithelial cells (34), where it catalyzes the hydrolysis of various epoxides and reactive epoxide intermediates into less reactive and more water-soluble dihydrodiols, which then are excreted from the body (14, 35, 36). Hence, mEH is a protective enzyme involved in general oxidative defense against a number of environmental and endogenous stimulants including endotoxin exposure. Genotypes that are associated with "slow" enzyme activity will lead to inefficient metabolizing of ROS generated by endotoxin exposure, and may eventually induce faster lung function decline. Previous in vitro studies have suggested that the His/His genotypes of both Tyr113His and His139Arg polymorphisms are associated with decreased enzyme activity (16). Therefore, we observed the fastest annual decline rate of lung function associated with endotoxin exposure in the His/His genotypes of both mEH polymorphisms (Table 3).

Inconsistent associations between mEH polymorphisms and the development of COPD were suggested in different studies. The "slow" mEH genotypes were associated with higher susceptibility or seventy of COPD in five studies (17-21), but not in others (22-26). In one study with 591 white COPD patients who smoked, the "slow" mEH genotypes were associated with rapid decline of lung function (19). The inconsistent results may be due to the differences in the study population or environmental exposure levels. Unlike our 20-year prospective population, the majority of the previous studies were hospital-based case-control studies with the outcome of the development of COPD. Our results support the hypothesis that mEH genotype alone may only have a slight effect on lung function decline; however, mEH polymorphisms may modify the effects of endotoxin exposure on lung function decline. In the mEH polymorphisms-endotoxin joint effects analyses, the joint effects of endotoxin exposure and "slow" mEH genotypes are significantly related to faster decline in lung function, with a stronger effect than genotype or endotoxin exposure alone (Table 4), further suggesting the importance of investigating both genetic and environmental factors (endotoxin exposure and smoking) in lung function decline and the development of COPD.

In addition to endotoxin exposure, cigarette smoking is another important environmental factor for lung function decline. The mEH enzyme is also involved in the metabolism of smoking-induced highly reactive epoxide intermediates. However, unlike the metabolism of mEH on endotoxin-generated ROS, mEH has both Phase I and Phase II activity for tobacco chemical (37), which may be one reason that we did not find statistically significant interactions between cigarette smoking and mEH genotypes (data not shown). In this study, we found that although cigarette smoking induced additional lung function decline in different genotype groups, the modification effect of mEH polymorphisms on endotoxin exposure existed in both never smokers and ever smokers. The associations between mEH genotypes and endotoxin exposure were consistent in different models, including stratified analyses by mEH genotypes (Table 3) as well as gene-endotoxin exposure joint analyses (Table 4). We cannot exclude the possibility of gene-smoking interactions existing in nonoccupational populations, because our results were generated from the occupational cohort.

There are several limitations to our study. First, this is a moderate sample size study, especially in the stratified analyses by genotype groups. In this study, the study population was derived from a 20-year longitudinal cohort. In the initial enrollment, subjects were excluded if they had a history of respiratory diseases, and only healthy subjects were included in this cohort. We have complete demographic, smoking, and genotype data, and adjusted possible confounders in the multivariate liner regression models. Although a higher proportion of people in the endotoxin-exposed group donated blood samples (59.73%) than in the nonexposed group (49.58%), blood donations were random and not on the basis of genotypes, and the demographic distributions (age, sex, height, smoking, and baseline FEV^sub 1^) were similar between those with and without blood samples. Second, the use of area rather than personal endotoxin exposure data may have introduced some exposure misclassification. However, we collected complete environmental monitoring data, which allowed us to stratify the population into endotoxin-exposed and nonexposed groups, and to investigate the association between endotoxin exposure and lung function decline. Lastly, the enzyme activity associated with mEH genotypes was derived from in vitro studies, which might induce some misclassification for the function of mEH genotypes. We have observed consistent results for both of the mEH polymorphisms and in different multivariate regression models.

In conclusion, mEH polymorphisms play an important role in the association between endotoxin exposure and lung function decline. Our results suggest that mEH "slow" genotypes increase the annual decline rate of lung function for subjects who are occupationally exposed to airborne endotoxin.

Conflict of interest Statement: J.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this article; W.Z. does not have a financial relationship with a commercial entity that has an interest in the subject of this article; X.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this article; H.Z. does not have a financial relationship with a commercial entity that has an interest in the subject of this article; B.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this article; H.D. does not have a financial relationship with a commercial entity that has an interest in the subject of this article; L.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this article; D.C.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this article.

Acknowledgment: The authors thank the members of the Shanghai field team; Professors Gu Xue-qi, Lu Pei-lian, and Ye Ting-ting from Shanghai Medical University; the Shanghai Textile Bureau; and workers and members of the First and Second Textile Mills and the First Silk Mill of Shanghai, China. They also thank Janna Frelich for data management and Dr. Stephen Olenchock of the National Institute for Occupational Safety and Health for endotoxin analyses.

© 2005 American Thoracic Society Provided by ProQuest LLC. All Rights Reserved.