Effects of Gene–Environmental Interaction on ... - ACS Publications

Jul 25, 2011 - Effects of GeneАEnvironmental Interaction on Noise-Induced Hearing ... on hearing.1 In addition, several other environmental factors, s...
0 downloads 0 Views 1MB Size
ARTICLE pubs.acs.org/est

Effects of GeneEnvironmental Interaction on Noise-Induced Hearing Threshold Levels for High Frequencies (HTLHF) Cheng-Yu Lin,†,‡,§ Tung-Sheng Shih,||,^,@ Yue-Liang Leon Guo,# Jiunn-Liang Wu,§ Yih-Min Sun,& and Perng-Jy Tsai*,‡,^,@ †

)

Department of Otolaryngology, Tainan Hospital, Department of Health, 125 Zhongshan Road, West Central District, Tainan City 700, Taiwan ‡ Department of Environmental and Occupational Health, Medical College, National Cheng Kung University, 138 Sheng-Li Road, North District, Tainan City 704, Taiwan § Department of Otolaryngology, Medical College and Hospital, National Cheng Kung University, 138 Sheng-Li Road, North District, Tainan City 704, Taiwan Institute of Occupational Safety and Health, Council of Labor Affairs, Executive Yuan, 99, Lane 407, Heng-Ke Road, Si-Jhih City, Taipei County 221, Taiwan ^ Department of Occupational Safety and Health, College of Public Health, China Medical University, 91 Hsueh-Shih Road, Taichung City 404, Taiwan # Department of Environmental and Occupational Medicine, Medical College, National Taiwan University, Room C339, 17 Syu-Jhou Road, Jhong-Jheng District, Taipei City 100, Taiwan & Department of Occupational Safety and Health, Chung Hwa University of Medical Technology, 89 Wun-Hua First Street, Ren-De Township, Tainan County 717, Taiwan ABSTRACT: In this study we assessed the interaction between glutathione S-transferase (GST) genetic polymorphisms and noise exposures, with regard to their effect on the hearing threshold levels for high frequencies (HTLHF). Research participants comprised 347 male workers, and each participant’s cumulative noise exposure was determined using a jobexposure matrix. Approximately 64.6% of the participants’ exposure in Leq-8 h was above 90 dBA. The mean HTLHF was 32.1 dB. A significant doseresponse relationship was found between noise exposure and HTLHF. We further converted the estimated total noise exposure level over each participant’s job history to a noise exposure level that corresponded to a 40-year exposure (Leq-40y). After we had adjusted the results for age, we found that workers carrying GSTM1 null, GSTT1 null, and GSTP1 Ile105/Ile105 genotypes were susceptible to the HTLHF when their Leq-40y were above 90 dBA. Therefore, GST genetic polymorphisms might affect HTLHF only when workers are exposed to high noise levels.

’ INTRODUCTION Excessive noise exposure is a pervasive occupational hazard with many adverse effects, including elevated blood pressure, reduced performance, sleeping difficulties, annoyance, tinnitus, and noise-induced hearing loss (NIHL). Long-term exposure to intense noise causes NIHL. In general, the longer the exposure lasts or the higher the noise level, or both, the greater the impact on hearing.1 In addition, several other environmental factors, such as exposure to organic solvents and heavy metals, can augment the effects of noise.28 Recently, animal model studies have suggested that genetic factors might influence individual susceptibility to noise.914 However, firm evidence associated with genetic factors in human NIHL is still limited, and sometimes contradictory results have r 2011 American Chemical Society

been found in different studies.1522 Rabinowitz et al. demonstrated that the glutathione S-transferase (GST) M1 wild genotype may have a protective effect on outer hair cell function in Americans working in noisy environments.15 Carlson et al. found no significant association between NIHL and GST genetic variants (i.e., GSTT1, GSTM1, and GSTP1) in a population of 215 noise-exposed workers in Sweden.17 Yang et al., in a study of 194 Chinese workers, showed that individuals with the GSTT1 null genotype might be more susceptible to NIHL.18 Our own Received: February 13, 2011 Accepted: July 25, 2011 Revised: July 18, 2011 Published: July 25, 2011 7128

dx.doi.org/10.1021/es200497v | Environ. Sci. Technol. 2011, 45, 7128–7134

Environmental Science & Technology prior research found that individuals carrying all genotypes of GSTT1 null, GSTM1 null and GSTP1 Ile105/Ile105 were susceptible to noise-induced temporary threshold shifts (TTS) for high frequencies.21 Researchers believe that the generation of reactive oxygen species (ROS) is an important factor associated with tissue destruction through metabolic stress.23 Because antioxidants can scavenge and eliminate the damaging ROS, antioxidants may interrupt the apoptotic biochemical cascade that results in the death of irreplaceable hair cells.24 One laboratory animal study has shown that glutathione (GSH), an important cellular antioxidant, limits noise-induced cochlear damage caused by ROS.25 The antioxidant enzymes involved in GSH metabolism include glutathione S-transferase (GST), glutathione peroxidase (GPX1), and glutathione reductase (GSR). Regarding the influence of genetic polymorphisms on hearing impairment for the above three GSH-related enzymes, GST is the one which has been studied most extensively in recent years.15,1719,21 GST enzymes catalyze the conjugation of GSH with xenobiotics and other compounds, and detoxify these chemicals. They play an important role in the antioxidant protection of the cochlea against noise-induced injuries.17 Individuals with GSTT1 null or GSTM1 null genotypes cannot conjugate metabolites specific for these enzymes.26,27 These two putative high-risk genotypes might render those individuals more prone to cellular damage induced by oxidative free radicals, which would possibly render such individuals more susceptible to NIHL. Additionally, GSTP1 is a member of the cytosolic GST superfamily. GSTP1 has two common nonsynonymous singlenucleotide polymorphisms (SNPs) that result in Ile105Val and Ala114Val alterations in the encoded amino acid sequence. Holley et al. showed that the GSTP1 Ile105 allele has reduced cellular proliferation and antiapoptotic activity through a c-Jun N-terminal kinase (JNK)-independent mechanism, as compared to the GSTP1 Val105 allele.28 Therefore, the GSTP1 Ile105 allele may be associated with an inability to repair cellular damage in the inner ear, thereby increasing the risk of NIHL. According to the findings of previous research, mutations in all three GST genes (i.e., GSTT1, GSTM1, and GSTP1) are known to result in oxidative stress, the retention of reactive quinone intermediates in cells, and the progression of cell death. In addition, the ability of cell restoration would be impaired. These pathogenetic pathways may render cells more sensitive to the toxic effects of oxidative stress, and may potentiate cell apoptosis. Rigorous research is needed to refine the assessment of exposure to noise, so that such assessments can be as accurate as possible and free from bias, thus avoiding the misclassification of noise exposure.29 To this end we wanted to investigate the geneenvironmental interaction and effect on NIHL. Although many studies have found a weak association between genetic influences and environmental exposure to noise,15,17 those results might have been contaminated by inaccuracies in noise exposure assessments. More recently, increasing attention has been paid to developing quantitative assessments of exposure, so as to provide better estimates and reduce the misclassification of various exposure groups.3033 In the current study we developed a jobexposure matrix (JEM), based on the measured noise levels of different work tasks and workers’ time/activity patterns. The JEM can accurately characterize each individual’s long-term cumulative noise exposure. Furthermore, we proposed that the risk of NIHL associated with cumulative noise exposure would be affected by the individual’s GST antioxidant capability.

ARTICLE

’ METHODOLOGY Subjects. We conducted our research in a steel manufacturing company in southern Taiwan. All workers in both the Steel Bar Manufacture and the Equipment Maintenance Departments were selected, apart from those who had potential exposure to organic solvents or polycyclic aromatic hydrocarbons. Subjects with a history of head injury, otological disease, ototoxic drug use, noise exposure during leisure time, congenital deafness, or preemployment hearing impairment were also excluded. All assessments were performed within one regular workday. The tests included a structured interview, an otoscopic examination, impedance tympanometry, pure tone audiometry (PTA), and a personal noise measurement. Our research was performed in accordance with the guidelines presented in the Declaration of Helsinki, and with approval from the Institutional Review Board (IRB) of the National Taiwan University Hospital, Taipei. Informed consent was obtained from all participants. Questionnaire. A structured questionnaire recorded the participants’ demographic characteristics, work tasks (including historically) and time/activity patterns, health habits (smoking or drinking alcohol), medical conditions, and the use of hearing protectors (HPs). Measurement of Daily Noise Exposure. Present Exposure. All participants worked for approximately eight hours per day; their 60-min lunch break was included in work time for the purpose of our study. Their tasks were carried out in a noisy environment which was located within 2 m of the production line. Personal noise exposure was assessed from 07:00 to 15:00 during the work shift, using TES-1355 Noise Dose Meters (TES Electrical Electronic Corporation, Taipei, Taiwan). The dosimeters measured the levels of different sound frequencies in every 10-min interval, then transformed the data into a continuous equivalent A-weighted sound pressure level (Leq) following the equal energy principle (3 dB rule). The timeweighted average (TWA) Leq during the 8-h shift (Leq-8 h) was calculated for each worker. The dosimeter used in this study conforms to the American National Standard Institute (ANSI) specifications.34,35 Past Exposure. According to the occupational health care database and our own questionnaire responses, most workers had worked in a single workplace during their entire employment history. Over the past few decades there had been very little change in their job descriptions or in the manufacturing processes and facilities (e.g., layout of the workplace, types of facilities, and work tasks). Therefore, we assumed that the daily noise levels for past exposure could be calculated by using the noise monitoring data collected from different areas of the building. For each individual area, we measured A-weighted noise levels in Leq (using the 3-dB rule) using a sound level meter (TES-52, TES Electrical Electronic Corporation, Taipei, Taiwan) placed at a height equivalent to that of workers’ ears (155 cm). If a worker used an HP in any given area, the noise reduction rating (NRR) specified for that particular HP was used to adjust the corresponding Leq-8 h. We measured the noise levels in all areas of the workplace as well as designated rest areas. Our sound level meter met the American National Standard Institute (ANSI) specifications.34,35 Total Noise Exposure Assessment over a Worker’s Career. Present and past occupational noise exposure was assessed by 7129

dx.doi.org/10.1021/es200497v |Environ. Sci. Technol. 2011, 45, 7128–7134

Environmental Science & Technology examining workers’ job histories according to our specific JEM. Work histories included successive job titles, dates, and departments. Job titles were grouped into 99 similar exposure groups, based on the workplace location and/or tasks of the worker. Total noise exposure level over the job history was estimated using the TWA exposure over the years (Leq-total, dBA-year). Leq-total was determined using logarithmic addition of the TWA noise levels as follows: h i Leq-total ¼ 10 3 log ΣTi 3 10ðLeq-8hi=10Þ

ARTICLE

Table 1. Demographic Variables and Distribution of the Subjects (n = 347) characteristics

mean (SD)

age (years)

48.9 (7.5)

employment (years) HTLHFa (dB)

24.2 (8.0) 32.1 (16.9)

tobacco smoking

8.0 (13.5)

(pack-year) smoking tobacco

where Ti is the duration of the work history i (years) and Leq-8hi is the predicted TWA noise exposure for the job held during the work period i. Assuming a worker’s lifetime working expectancy is forty years,36,37 we further converted Leq-total to a noise exposure level corresponding to a forty-year exposure (Leq-40y). For this purpose we used the equation suggested by Burn and Robinson,38 as follows:

yes

145 (41.8)

no

202 (58.2)

drinking alcohol yes no

65 (18.7) 282 (81.3)

habit of hearing protector usage

Leq-total ¼ 10 3 log 40 þ Leq-40y

yes

249 (71.8)

no

98 (28.2)

GSTM1

Leq-40y ¼ Leq-total  16:02 Evaluation of Hearing Status. Before having their audiological assessments, the participants were asked to avoid explosive noise exposure for at least 48 h. On the test day we performed hearing evaluations from 06:30 to 07:30. An otolaryngologist conducted otoscopic examinations to determine the status of the outer and middle ear. Middle ear function was assessed by a trained technician using impedance tympanometry (Grason Stadler GSI-37 Auto Tymp; Gordon N. Stowe and Associates, Inc., Wheeling, IL, USA). Pure-tone audiometry (PTA) was conducted by an audiologist using a Grason-Stadler GSI 68 audiometer (Gordon N. Stowe and Associates, Inc., Wheeling, IL, USA). The sound-attenuating chamber, with a background noise level of 25 dBA or less, complied with the testing conditions recommended by the International Organization for Standardization (ISO) 8253-1.39 The test frequencies for each ear included 500, 1000, 2000, 3000, 4000, 6000, and 8000 Hz. The hearing threshold level for high frequencies (HTLHF) in PTA was calculated as the average threshold level for 3000, 4000, and 6000 Hz respectively. Genotyping of GST Polymorphisms. We collected venous blood samples in heparinized tubes at the workplace, and placed them immediately into a refrigerated container. Genomic DNA was extracted using standard phenol/chloroform extraction techniques. The GSTM1 and GSTT1 genes were assessed simultaneously in a single assay using the multiplex polymerase chain reaction (PCR) approach described by Arand et al.40 The PCR produced three DNA fragments of 480 bp (GSTT1), 350 bp (albumin), and 215 bp (GSTM1). In both GSTT1 and GSTM1 polymorphisms, gene deletions are known to be responsible for the existence of null alleles. Individuals who are homozygous with respect to a given null allele lack the respective PCR-amplified DNA fragment. Albumin was used as an internal control for PCR efficiency. We used the method Watson et al. reported for determining the genotype at the GSTP1 locus, which combines PCR and restriction fragment length polymorphism (PCR/RFLP).41 The GSTP1 Ile105Val polymorphisms were classified as homozygous Ile105/Ile105 and carrier with either Val allele (Ile105/Val105 or Val105/Val105).

no. (%)

null-type

175 (50.4)

wild-type GSTT1

172 (49.6)

null-type

192 (55.3)

wild-type

155 (44.7)

GSTP1-105 Ile105/Ile105

230 (66.3)

Ile105/Val105 + Val105/Val105

117 (33.7)

GST antioxidant capabilityb poor-type normal-type

56 (16.1) 291 (83.9)

a

HTLHF: The average of hearing threshold levels at 3000, 4000, and 6000 Hz by pure-tone audiometry. b GST antioxidant capability: Individuals with the combination of GSTM1 null, GSTT1 null, and GSTP1 Ile105/Ile105 genotypes were classified as the poor group. Individuals with other combinations of the GSTT1, GSTM1, and GSTP1 alleles were classified as the normal group.

Statistical Analysis. We performed all analyses using Jump 5.0 software (SAS Institute Inc., Cary, NC, USA). We used the general linear regression model to analyze the doseresponse relationship between total noise exposure and HTLHF for different GST genetic variants. Those workers with GSTM1null, GSTT1-null, and GSTP1 Ile105/Ile105 were considered to have poor oxidative genotypes. A univariate regression model was used to compare the association between each possible risk factor and HTLHF. Our JEM, which assessed total noise exposure, took into account the time-weighted noise exposure level, use of HPs, and duration of employment. We therefore excluded those three variables from further statistical analysis. We used the multivariate regression model, with adjustments for the possible confounding variables, to estimate the HTLHF of different combinations of the three GST genetic variants. Statistical significance was set at a twotailed p-value of 0.05 or less.

’ RESULTS Demographic and Personal Characteristics. A total of 385 male workers were screened. Eleven were found to have eardrum 7130

dx.doi.org/10.1021/es200497v |Environ. Sci. Technol. 2011, 45, 7128–7134

Environmental Science & Technology

ARTICLE

Table 2. Characteristics of Work History Tasks and Time/Activity Patterns According to GST Antioxidant Capability (n = 347) GST antioxidant capability normal (n = 291) variables

no. (%)

poor (n = 56)

mean (SD)

no. (%)

mean (SD)

P value

age (years)

48.8 (7.6)

49.2 (6.8)

0.691

employment (years)

24.1 (8.2)

24.9 (6.8)

0.494

habit of hearing protector usage

0.701

yes

210 (72.2)

39 (69.6)

no

81 (27.8)

17 (30.4)

current time-weighted average noise level (Leq-8 h) (dBA) workplace (N = 99)a

82.1 (0.8)

85.1 (1.8)

0.127

rest place (N = 36) duration of hearing protector usage in workplace (hours/day)

62.4 (0.2) 4.1 (0.2)

62.9 (0.5) 4.8 (0.5)

0.356 0.165

duration of stay in workplace (hours/day)

5.9 (0.1)

6.0 (0.3)

0.816

duration of stay in rest place (hours/day)

2.2 (0.1)

2.1 (0.3)

0.824

total noise exposure (Leq-total) (dBA-year)

91.1 (0.7)

92.3 (1.5)

0.473

noise level for 40 years of exposure (Leq-40y) (dBA) < 70.0

a

0.597 91 (31.3)

15 (26.8)

70.079.9

113 (38.8)

23 (41.1)

80.089.9 > 90.0

47 (16.2) 40 (13.8)

7 (12.5) 11 (19.6)

N: Number of areas assessed.

Figure 1. Mean hearing threshold levels (HLs) (dB) at each hearing frequency, for groups with normal or poor GST antioxidant capability. (a) normal type, 2; (b) poor type, 3 3 3 0 3 3 3 . Error bars show the SD of the HLs.

perforation and/or purulent otorrhea, and seven reported a family history of congenital deafness. We excluded these workers as well as twenty individuals who had worked in other noisy environments before joining the company. Our final sample consisted of 347 workers, all male (347/385, 90.1%). Most participants were middle-aged with an average of 24.2 work years’ experience (Table 1). There were 41.8% workers with a habit of tobacco smoking. Two-thirds (71.8%) had the habit of wearing HPs in a noisy workplace, with earplugs being the preferred device.

The two genotypes GSTM1 null and GSTT1 null were each present in approximately half of the subjects, but GSTP1 Ile105/ Ile105 was found in two-thirds (Table 1). Fifty-six workers (16.1%) harboring all three genotypes (GSTM1 null, GSTT1 null, and GSTP1 Ile105/Ile105) were presumed to have poor antioxidant capability. Noise Exposure and Hearing Threshold Levels. Two-thirds of the participating workers (224 subjects, 64.6%) were exposed to a daily Leq-8 h above the permissible level of exposure (90 dBA). Among them, 49 subjects (21.9%) never used HPs. The reasons given included a feeling of discomfort, interference with job performance, and interference with communication. We found that mean current TWA noise level in the workplace for workers with normal GST antioxidant capabilities was 82.1 dBA, while for workers with poor antioxidant ability it was 85.1 dBA. The difference between these two means was not statistically significant (p = 0.127). As shown in Table 2, there was also no significant difference between the two groups (normal versus poor GST antioxidant ability) with regard to age, employment, habit of HP usage, and noise exposure parameters (p > 0.05). The mean hearing threshold levels at each frequency are presented in Figure 1. The average HTLHF was 32.1 ( 16.9 dB. For participants with normal and poor antioxidant ability, the average hearing threshold levels at 500, 1000, and 2000 Hz were 17.4 ( 7.8 and 18.4 ( 9.4 dB, respectively. These values were not significantly different from those found for the general population (17.2 ( 9.2 dB) (p > 0.05).42 Association between Risk Factors and Hearing Threshold Levels for High Frequencies. None of the following variables was found to be significantly associated with HTLHF: duration of employment, drinking alcohol, tobacco smoking, a habit of wearing hearing protectors, GSTM1-null genotype, GSTT1-null genotype, or GSTP1 Ile105/Ile105. However, through multivariate regression analysis we found that age and total noise exposure 7131

dx.doi.org/10.1021/es200497v |Environ. Sci. Technol. 2011, 45, 7128–7134

Environmental Science & Technology

ARTICLE

Table 3. Multivariate Regression Model of Hearing Threshold Levels by Different GST Antioxidant Capability (n = 347) HTLHFa normalb (n = 291)

poorb (n = 56)

βc

P value

β

P value

age (years)

0.523