Printer Friendly

Low-level environmental cadmium exposure is associated with DNA hypomethylation in argentinean women.

BACKGROUND: Cadmium, a common food pollutant, alters DNA methylation in vitro. Epigenetic effects might therefore partly explain cadmium's toxicity, including its carcinogenicity; however, human data on epigenetic effects are lacking.

OBJECTIVE: We evaluated the effects of dietary cadmium exposure on DNA methylation, considering other environmental exposures, genetic predisposition, and gene expression.

METHODS: Concentrations of cadmium, arsenic, selenium, and zinc in blood and urine of nonsmoking women (n = 202) from the northern Argentinean Andes were measured by inductively coupled mass spectrometry. Methylation in CpG islands of LINE-1 (long interspersed nuclear element-1; a proxy for global DNA methylation) and promoter regions of p16 [cyclin-dependent kinase inhibitor 2A (CDICN2A)] and MLH1 (mutL homolog 1) in peripheral blood were measured by bisulfite polymerase chain reaction pyrosequencing. Genotyping (n = 172) for the DNA (cytosine5-)-methyltransferase 1 gene (DNMT1 rs10854076 and rs2228611) and DNA (cytosine-5-)-methyltransferase 3 beta gene (DNMT3B rs2424913 and rs2424932) was performed with Sequenom iPLEX GOLD SNP genotyping; and gene expression (n = 90), with DirectHyb HumanHT-12 (version 3.0).

RESULTS: Cadmium exposure was low: median concentrations in blood and urine were 0.36 and 0.23 [micro]g/L, respectively. Urinary cadmium (natural log transformed) was inversely associated with LINE-1 methylation ([beta] = -0.50, p = 0.0070; [beta] = -0.44, p = 0.026, adjusted for age and coca chewing) but not with p16 or MLH1 methylation. Both DNMT1 rs10854076 and DNMT1 rs2228611 polymorphisms modified associations between urinary cadmium and LINE-1 (p-values for interaction in adjusted models were 0.045 and 0.064, respectively). The rare genotypes demonstrated stronger hypomethylation with increasing urinary cadmium concentrations. Cadmium was inversely associated with DNMT3B ([r.sub.s] = -0.28,p = 0.0086) but not with DNMT1 expression ([r.sub.s] = -0.075,p = 0.48).

CONCLUSION: Environmental cadmium exposure was associated with DNA hypomethylation in peripheral blood, and DNMT1 genotypes modified this association. The role of epigenetic modifications in cadmium-associated diseases needs clarification.

KEY WORDS: cadmium, DNMT1, DNMT3B, epigenetic, genotype, LINE-1, MLH1, p16, pyrosequencing. Environ Health Perspect 120:879-884 (2012). http://cbc.dolorg/10.1289/ehp.1104600 [Online 1 March 2012]

Cadmium is a widespread environmental pollutant associated with adverse health effects, including cancer, kidney damage, low bone-mineral density, bone fractures, and probably impaired early-life development (Engstrom A et al. 2011; Jarup and Akesson 2009; Kippler et al. 2011). The general population is exposed to cadmium mainly through the diet, in particular, through cereals and vegetables, and through smoking. Potential mechanisms of cadmium toxicity described so far include interactions with sulfhydryl groups and endocrine disruption, as well as oxidative stress (Hartwig 2010; Johnson a al. 2003; Monroe and Halvorsen 2006; Xu et al. 2011; Yu a al. 2010).

DNA methylation is essential for embryogenesis and for maintaining cell-lineage-specific gene expression throughout life. Dysregulation of epigenetic processes, such as DNA methylation, may lead to impaired childhood development or chronic diseases later in life, including cancer (Faulk and Dolinoy 2011; Feinberg 2007; Robins a al. 2011). In carcinogenesis, global DNA methylation is generally decreased, whereas specific regions of DNA that contain tumor suppressor genes often show hypermethylation (Chalitchagorn et al. 2004; Deng et al. 2006; Gaudet et al. 2003; Robertson and Jones 2000). In vitro studies have indicated that cadmium may interfere with epigenetic processes: Both genespecific DNA hypermethylation with gene silencing of pl6INK4a [CDKN2A (cyclin-dependent kinase inhibitor 2A)], RASSF1 [Ras association (RaIGDS/AF-6) domain family member 1], and MT1 (metallothionein 1) genes, and global DNA hypomethylation have been reported (Benbrahim-Tallaa et al. 2007; Huang et al. 2008; Jiang et al. 2008; Martinez-Zamudio and Ha 2011; Takiguchi et al. 2003). However, there are no reports on the epigenetic effects of cadmium exposure in humans.

The aim of this study was to evaluate whether environmental cadmium exposure, mainly through the normal diet, may affect DNA methylation, both of repetitive genetic elements (as a proxy for global methylation) and of specific genes.

Materials and Methods

Participants. We studied 202 women living in the Andean plateau (3,800 m above sea level) in northern Argentina, an area that has minimal industrial or vehicle-derived pollution. Most of the study participants (161) were from the village of San Antonio de los Cobres, which has about 5,000 inhabitants. The 41 remaining participants were from small surrounding villages. The study area is known to have elevated concentrations of arsenic (Vahter et al. 1995), but not cadmium (Concha et al. 2010), in the drinking water.

The study individuals were recruited with the assistance of local medical personnel, except in the small mining village of Tolar Grande, where we went from house to house to explain the study and invited the adults to participate. We included only women because men were often away from home for longer periods for work and were therefore likely to be exposed to a different variety of environmental hazards. Furthermore, women may be more susceptible to cadmium exposure than are men because of their increased gastrointestinal cadmium absorption when iron stores are low (Berglund et al. 1994), as is common in women of child-bearing age. Only four study participants smoked cigarettes, and one reported drinking alcohol, but almost half (46.5%) reported that they chewed coca leaves. As reported before, interviews with the study participants revealed that almost all women drank exclusively public water and that their diets consisted mainly of corn, beans, thicken, and pork (Broberg et al. 2011). Only three women reported taking any medication at the time of the study: one being treated for gastritis and two for hypertension. No women reported any history of or current malignancy.

We measured two biomarkers of cadmium exposure: blood cadmium, reflecting ongoing exposure, and urinary cadmium, reflecting long-term exposure (lamp and Akesson 2009). Both oral and written informed consent was provided by all the study participants. The study was approved by the Ministry of Health in Salta, Argentina, and the Regional Ethical committee at Karolinska Institute, Sweden.

Blood and urine collection. All biological samples were nonfasting spot samples collected during the daytime at local health clinics or the hospital in San Antonio de los Cobres. Peripheral blood for DNA extraction (n = 202) was collected in [K.sub.2]EDTA tubes (Vacuette; Vacuette Bio One, Frickenhausen, Greiner, Germany), blood for element analysis (fri = 202) in lithium heparin tubes (Vacuette Bio One), and blood for RNA extraction (n = 122) in PAXgene Blood RNA Tubes (Becton Dickinson, Franklin Lakes, NJ, USA). Spot urine samples were collected and processed as described previously (Concha et al. 1998). Blood, plasma, and urine samples were kept at -20[degrees]C before and after the transport (with cooling blocks) to Sweden for analysis.

Analysis of elements in blood and urine. We measured cadmium in blood and urine, and selenium and zinc in blood, because these elements may influence DNA methylation or one-carbon metabolism (Hong a al. 2000; Pilsner et al. 2011; Smeester et al. 2011). Cadmium, selenium, and zinc were measured using inductively coupled plasma mass spectrometry (ICP-MS; model 7500ce; Agilent Technologies, Waldbronn, Germany) with a collision/reaction cell system to remove potential interferences, particularly interference from molybdenum in the cadmium analysis (Concha et al. 2010; Kippler et al. 2009). The limit of detection (LOD; 3 x SD of blank values) was 0.001 [micro]g/L for urinary cadmium, and no sample had a cadmium concentration below the LOD. For blood measurements, the LOD was 0.011, 0.081, and 0.49 [micro]g/L for cadmium, selenium, and zinc, respectively, and no samples were below the LOD for any of the compounds measured. Analytical accuracy was controlled by analyzing commercially available reference materials (Concha et al. 2010; Kippler et al. 2009). Because the study area is known to have varying concentrations of arsenic in the drinking water, and because arsenic may alter DNA methylation (Chanda et al. 2006; Smeester et al. 2011), we also assessed exposure to arsenic by measuring arsenic metabolites (inorganic arsenic, methylarsonic acid, and dimethylarsinic acid) in urine using HPLC/ICP-MS (Engstrom et al. 2010). Additionally, the study area was recently found to have elevated concentrations of boron, lithium, and cesium in the drinking water, and we measured these elements in urine by ICP-MS (Concha et al. 2010). To control for variation in urine dilution, we adjusted the cadmium and arsenic concentrations in urine to the mean specific gravity of urine (1.020 g/mL) (Mermen et al. 2008). We measured the specific gravity of the urine samples with a digital refractometer (model RD 712 clinical refractometer; EUROMEX, Arnhem, Holland).

Polymerase chain reaction (PCR) and pyrosequencing. DNA was isolated using the QIAampe DNA Blood Midi kit (Qiagen, Hilden, Germany). We used the EpiTect4 kit (Qiagen) to treat 1 [micro]g of DNA (50 ng/[mu]L) with bisulfite. Bisulfite treatment of DNA converts unmethylated cytosine residues into uracil, leaving the methylated cytosine unchanged; these changes can be quantified by pyrosequencing. The bisufite-treated DNA was stored at -20[degrees]C until used for PCR.

We used the bisulfite-treated DNA (0.6-1 [micro]L) in a 15-to 25-[mu]L PCR reaction using the Pyromark PCR kit (Qiagen). One of the PCR primers was biotinylated, and the PCR product was purified using Streptavidin Sepharose High Performance beads (Amersham Biosciences, Uppsala, Sweden). The Sepharose beads with the immobilized PCR products were purified, washed, and denatured with 0.2 M NaOH and washed again using a vacuum prep tool (Pyrosequencing Inc., Westborough, MA, USA). After annealing 12 [micro]L of the 0.3-[mu]mol/L pyrosequencing primer to the purified single-stranded PCR product, we performed pyrosequencing using the PSQ HS96 Pyrosequencing System (Qiagen). The degree of methylation was expressed as the percentage of methylated cytosines over the sum of methylated and unmethylated cytosines. We verified bisulfite conversion using non-CpG (cytosine-guanine dinucleotide) cytosine residues as built-in controls; and complete conversion of cytosine at a non-CpG site indicated successful bisulfite conversion.

LINE-1, p16, and MLH1 methylation. As a surrogate measure of genome-wide methylation, we measured peripheral blood DNA methylation in the long interspersed nuclear element-1 (LINE-1) retrotransposable elements that make up about 17% of human DNA (Lander et al. 2001). We also measured methylation of CpG islands in p16 [cyclin-dependent kinase inhibitor 2A (CD1CN2A)], a tumor suppressor gene that is often hypermethylated in tumors and may be relevant in cadmium-induced tumorigenesis (Benbrahim-Tallaa et al. 2007), and in the MLH1 (mutL homolog 1) gene, a component of the DNA mismatch repair pathway.

Commercially available kits (Qiagen) were used to measure the methylation of four CpG sites in LINE-1 [positions 305-331 of GenBank accession no. X58075 (National Library of Medicine 2012a)], seven CpG sites in p16 (+148 to +174 in exon 1; GenBank accession no. L27211), and four CpG sites in MLH1 (-209 to--188; GenBank accession no. U07418) following the manufacturer's instructions. A single PCR fragment spanning a part of each genetic element was amplified, and the degree of methylation was analyzed in a single pyrosequencing reaction using 20 pL PCR product for LINE-1, 3 [micro]L for p16, and 4 [micro]L for MLH1. The samples were analyzed as singlets for LINE-1 and as triplicates for p16 and MLH1. We selected 10% of the samples at random and reassayed them. The repeatability of the method, expressed as the variation in coefficients, was 2.0%, 36.7%, and 26.4%, for LINE-1, p16, and MLH1 methylation, respectively. Previous reports have validated pyrosequencing as an effective technique for measuring LINE-1 methylation at the population level (Schernhammer et al. 2010; Tarantini et al. 2009).

Genotyping and gene expression. Genotyping for polymorphisms of DNA methyltransferase-1 [(cytosine-5-)-methyltransferase 1] (DNMT1 rs10854076 and rs2228611, one from each linkage disequilib-rium block in the gene) and DNA (cytosine-5-)-methyltransferase 3 beta (DNMT3B rs2424913 and rs2424932) was performed on DNA from peripheral blood with iPLEX GOLD SNP genotyping (Sequenom Inc., San Diego, CA, USA) the Swegene DNA facility at Malmo University Hospital (Malmo, Sweden), as described previously (Engstrom K et al. 2011). DNMT1 rs222861 I [minor allele frequency (MAF) = 40%] is a synonymous amino acid exchange, whereas rs10854076 (MAF = 35%) is situated in an intron. DNMT3B rs2424913 (MAF = 5%) and rs2424932 (MAF = 6%) are both situated in introns. The functional impact of these polymorphisms on DNA methylation, if any, is unknown.

RNA was extracted using the PAXgene Blood RNA kit (PreAnalytiX, Hombrechtikon, Switzerland) and stored at -80[degrees]C until isolation. We evaluated RNA concentration and purity using a NanoDrop spectrophotometer (NANODROP 1000; Thermo Scientific, Wilmington, DE, USA), and we evaluated the RNA integrity number (RIN) using a Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA). The results indicated good RNA quality (RIN > 7.5) for all of the samples Of 122 available PAX tubes for RNA extraction, 90 samples were selected for gene expression analysis based on their RNA quality and the amount of RNA left for analysis [for descriptive characteristics of this subset of the study population, see Supplemental Material, Table 1 (]. Of the 90 individuals selected, 75 were not first-degree relatives, and they were selected for the analysis of genetic effect modification on the association between cadmium exposure and epigenetic markers. DirectHyb HumanHT-12 (version 3.0; Illumina, San Diego, CA, USA) was used for gene expression analysis according to the manufacturer's instructions at the SCIBLU (Swegene Centre for Integrative Biology at Lund University) facility at Lund University. Background signals were filtered from the gene expression data by BioArray Software Environment (BASE) (Vallon-Christersson et at 2009). The DirectHyb HumanHT-12 (HT-12) array (Illumina) included one DNMTI transcript and five DNMT3B transcripts [the gene has alternatively spliced transcripts (National Library of Medicine 2012b)]. No LINE-1 transcripts were included on the array.

Statistical analyses. We used the averages of the relative amounts of methylated cytosine residues in the four, seven, and four CpG sites measured in LINE-1, p16, and MLH1, respectively to determine the overall methylation level for each genetic element analyzed in a given sample. We used Spearman's rank correlation analysis to estimate univariate correlations between cadmium in blood or urine and methylation and gene expression, as well as correlations with potential covariates such as age, body weight, body mass index (BMI), use of coca leaves, and zinc and selenium concentrations in whole blood.
Table 1. Descriptive information of the women and biomarkers
analyzed. (a)

Variable                      Median  5th-95th percentile

Age (years)                       34                18-64

Body weight (kg)                57.0            44.0-79.9

Height (cm)                      152              144-161

BMI (kg/[m.sup.2])              24.7            18.8-35.0

Blood hemoglobin (g/L)           155              130-180

Urinary creatinine (g/L) (b)    0.90             0.52-1.4

Coca chewing (%)                46.5

Methylation (%)

LINE1 (c)                       86.2            83.5-89.3

p16 (d)                          3.4              1.7-5.8

MLH1 (e)                         4.2              1.2-7.1

Blood measures

Cadmium ([mu]g/L)               0.36            0.20-0.76

Selenium ([mu]g/L)               176              150-221

Zinc (mg/L)                      6.8              5.5-8.2

Urinary measures

Cadmium ([mu]g/L) (b)           0.23           0.091-0.70

Arsenic ([mu]g/L) (b)            230               21-545

Boron (mg/L) (b)                14.6             2.6-27.9

Cesium ([mu]g/L) (b)             470             26.1-880

Lithium (mg/L) (b)               3.9            0.27-10.4

(a.) n = 202 subjects, except for creatinine (n = 194) and blood
selenium (n = 201) analyses. (b.) Adjusted for specific gravity of
urine (1.020 g/mL). (c.) Average of relative amounts of methylated
cytosines in four CpGs, positions 305-331, GenBank accession no.
X58075. (d.) Average of relative amounts of methylated cytosines in
seven CpGs, positions 148-174, accession no. L27211. (e.) Average
of relative amounts of methylated cytosines in four CpGs, positions
-209 to -188, accession no. U07418.

We first evaluated the gene methylation by quintiles of cadmium exposure [see Supplemental Material, Figures 1-3 (] and by scatter plots. Because visual inspection of the plots suggested linear relationships, we used linear regression models to estimate associations ([beta]-coefficients) of cadmium in blood and urine with the epigenetic markers. We performed a multivariable-adjusted analysis controlling first for age and coca chewing (yes/no) because age and coca chewing were correlated with at least one of the epigenetic markers (p-values < 0.2) and because cadmium exposure--outcome coefficients changed by > 5% when they were included in models. To assess the robustness of associations, we additionally controlled for urinary arsenic, boron, lithium, and cesium (modeled as untransformed continuous variables). In the linear regression analyses, urine and blood cadmium concentrations were natural log (In) transformed to obtain normal distributions of the residuals and a better linear fit with the outcomes.

The influence of DNMT1 genotypes (effect modification) on the association between urinary cadmium concentration and LINE-I methylation was analyzed in individuals that were not first-degree relatives (n = 172). This analysis was performed by modeling a cross-product term between genotype (for DNMT1 as an ordinal variable with three categories, and for DNMT3B dichotomized because few individuals were variant homozygotes) and lower-order terms for genotype and Inurinary cadmium concentrations. Analyses were then stratified by genotype to estimate associations between urinary cadmium and LINE-1 methylation according to each genotype. In addition, we used stratified models to estimate associations between urinary cadmium and DNMT1 expression according to genotype in the 75 participants that were not first-degree relatives. Associations between genotypes and the expression of DNMTs were estimated using analysis of variance with and without adjustment for age and coca chewing. We did not adjust p-values for multiple comparisons. A p-value < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS (version 18; SPSS, Chicago, IL, USA).
Table 2. Associations between In-transformed blood and urinary
cadmium concentration ([mu]g/L) and percent methylation of LINE1,
p16, and MLH1.

LINE1                p16                 MLH1

Predictor   [beta]   p-Value  [beta]   p-Value  [beta]
              (95%                  (95%                  (95%
              CI)                 CI)                 CI)


Unadjusted     0.16     0.61     0.11     0.65     0.14
(a)         (-0.46,           (-0.38,           (-0.52.
              0.78)             0.60)             0.80)

Adjusted 1     0.40     0.24     0.30     0.26     0.26
(b)         (-0.27,           (-0.23,           (-0.46,
              1.07)             0.83)             0.98)

Adjusted 2     0.45     0.19     0.24     0.37     0.19
(c)         (-0.23,           (-0.29,           (-0.53,
              1.12)             0.77)             0.91)


Unadjusted    -0.50   0.0070    -0.13     0.37   -0.062
(a)         (-0.86,           (-0.42,           (-0.45,
             -0.14)             0.15)             0.33)

Adjusted 1    -0.44    0.026   -0.048     0.76  -0.0054
(b)         (-0.83,           (-0.36,           (-0.43,
            -0.053)             0.26)             0.42)

Adjusted 2    -0.42    0.038    -0.11     0.51   -0.073
(c)         (-0.82,           (-0.42,           (-0.50.
            -0.025)             0.21)             0.36)

Predictor   p-Value


Unadjusted     0.69

Adjusted 1     0.47

Adjusted 2     0.61


Unadjusted     0.75

Adjusted 1     0.98

Adjusted 2     0.74

[beta] is the unstandardized regression coefficient.
(a.) Unjvariate analysis (n = 202): percent methylation =
[[alpha].sub.1] + [[beta].sub.1] cadmium in blood/urine. (b.)
Adjusted analysis (n = 202): percent methylation = [[alpha].sub.1]
+ [[beta].sub.1] cadmium in blood/urine + [[beta].sub.2] x age +
[[beta].sub.3] x coca chewing. (c.) Adjusted analysis (n = 201):
percent methylation = [[alpha].sub.1] + [[beta].sub.1] cadmium in
blood/urine + [[beta].sub.2] x age + [[beta].sub.3] x coca chewing
+ arsenic in urine.


Descriptive data for the study participants and element concentrations in blood and urine are presented in Table 1. In general, the women were relatively short, and, using the conventional cut-offs for overweight (BMI 25-29.9) and obesity (BMI > 30), about half of them were overweight and 18% obese. Blood concentrations of hemoglobin, selenium, and zinc indicated adequate nutritional status. Cadmium concentrations were 0.013-1.5 [micro]g/L in urine (median, 0.23 [micro]g/L) and 0.17-1.1 [micro]g/L in blood (median, 0.36 [micro]g/L). Four women who reported smoking had cadmium concentrations of 0.045-0.24 [micro]g/L in urine and 0.18-0.52 [micro]g/L in blood. Median LINE-1, p16, and MLH1 methylation levels were 86.2%, 3.4%, and 4.2%, respectively (Table 1).

The epigenetic markers were positively correlated with each other: the correlation between p16 and MM-11 methylation being the strongest ([r.sub.S] = 0.78,p < 0.01), followed by LINE-1 and MLH1 ([r.sub.S]= 0.51, p < 0.01), and LINE-1 and p16 ([r.sub.S]= 0.42, p < 0.01). Urinary cadmium concentrations were negatively correlated with LINE-1 methylation ([r.sub.S] = -0.19, p < 0.01). Blood cadmium concentrations were not correlated with LINE-1 ([r.sub.S] = 0.046) but were positively correlated with urinary cadmium concentration ([r.sub.S] = 0.46, p < 0.01). Neither urinary cadmium nor blood cadmium concentrations were correlated with p16 or AILH1 methylation ([r.sub.S] < 0.1). None of the nutritional markers (BMI, hemoglobin, selenium, and zinc) were significantly correlated with the epigenetic markers ([r.sub.S] < 0.11).

In the linear regression analysis, the In-transformed urinary cadmium concentrations of all women (n = 202) were inversely associated with LINE-1 methylation [for unadjusted model: [beta] = -0.50; 95% confidence interval (CI): -0.86, -0.14; p = 0.0070; Table 2]. Estimates did not change much after adjustments for age and coca chewing or after additional adjustment for urinary arsenic concentration (Table 2). Further adjustments for urinary boron, lithium, or cesium also had little influence on effect estimates (data not shown). When the analysis for LINE-1 included an interaction term between cadmium and dichotomized arsenic in urine (median split), no significant interaction was detected (p = 0.37). Excluding the four smokers (which left n = 198) did not change the estimated association between urinary cadmium and LINE methylation (in fully adjusted model: [beta] = -0.44; 95% CI: -0.85, -0.034). There was no association between Intransformed urinary or blood cadmium concentrations and p16 or MLHI methylation (Table 2).
Table 3. Effect modification of DNMT1 and DNMT3B genotypes on the
association between In-transformed cadmium concentrations in urine
and for LINE! methylation levels (n= 172 in all analyses).

            Unadjusted                    Adjusted
                analysis                    analysis

Genotype        [beta] (a)     p-Value  [beta] (c) (95%  p-Value
                (95% CI)       (b)          CI)          (d)



CC (n = 25)   -1.09 (-2.02,    0.041    -0.99 (-2.03.    0.045
                     -0.15)                    0.059)

CG (n = 69)   -0.96 (-1.68,             -1.04 (-1.91,
                     -0.24)                    -0.17)

GG (n = 78)          -0.052            -0.053 (-0.61,
                    (-0.58,                     0.51)


GG (n = 35)   -1.19 (-1.99,    0.050    -1.04 (-1.96,    0.064
                     -0.40)                    -0.12)

GA (n = 68)   -0.38 (-1.07,             -0.38 (-1.14,
                      0.31)                     0.38)

AA (n= 69)    -0.10 (-0.67,                     -0.13
                      0.47)              (-0.75,0.48)



CC + CT (n =  -1.22 (-3.20,     0.39    -1.04 (-3.36,     0.38
15)                   0.77)                     1.27)

IT (n = 155)  -0,45 (-0.84,             -0.39 (-0.81,
                    -0.064)                    0.024)


AA + GA (n =  -1.36 (-3.12,     0.28    -1.43 (-3.39,     0.25
18)                   0.40)                     0.53)

GG (n = 152)  -0.46 (-0.85,             -0.39 (-0.81,
                    -0.065)                    0.037)

(a.) Percent LINE1 methylation (stratified for each genotype) =
[[alpha].sub.1] + [[beta].sub.1] cadmium. (b.) p-Value interaction
from term [[beta].sub.3] in the model: percent LINEJ methylation =
[[alpha].sub.1] + [[beta].sub.1] genotype + [[beta].sub.2] cadmium
+ [[beta].sub.3] (genotype x cadmium). (c.) Percent LINE1
methylation (stratified for each genotype) = [[alpha].sub.1] +
[[beta].sub.1] cadmium + [[beta].sub.2] age + [[beta].sub.3] coca
chewing. (d.) p-Value interaction from term [[beta].sub.3] in the
model: percent LINE1 methylation = [[alpha].sub.1] + [[beta].sub.1]
genotype + [[beta].sub.2] cadmium + [[beta].sub.3] (genotype x
cadmium) + [[beta].sub.4] age + [[beta].sub.5] coca chewing.

Table 4. DNMT1 and DNMT3B gene expression according to DNMT1 and
DNMT3B genotypes. (a)

DNMT1                         DNMT3B

Genotype    Gene        p-Value  p-Value  Gene        p-Value  p-Value
            expression  (c)      (d)      expression  (c)      (d)
            (b)                           (b)



CC (n = 8)       654.9     0.14     0.18        92.2    0.027    0.040

CG (n =          640.3                          90.0

GG (n =          629.6                          87.3


GG (n =          653.6     0.18     0.18        91.0     0.23     0.26

GA (n =          639.6                          89.2

AA (n =          591.4                          87.8



CC + CT (n       583.0     0.83     0.67        85.1    0.068    0.036
= 7)

TT (n =          640.0                          89.2


AA + GA (n       559.0     0.68     0.38        86.4     0.19     0.11
= 8)

GG (n =          640.3                          89.1

(a.) DNMTT is represented by one transcript (ILMN_1760201), whereas
DNMT3B is represented by five transcripts (values from ILMN_1733929
presented here). DNMT genotypes were not associated with the
expression of other DNMT3B transcripts (p > 0.1). (b.) Median,
expressed as an arbitrary unit. (c.) p-Value from model gene
expression = [[alpha].sub.1] + [[beta].sub.1] x genotype. (d.) p-
Value from model gene expression = [[alpha].sub.1] + [[beta].sub.1]
x genotype + [[beta].sub.2] x age + [[beta].sub.3] x coca chewing.

The association of urinary cadmium with LINE-1 methylation showed statistically significant interactions in unadjusted models with the genotypes DNMTI rs10854076 and rs2228611 (Table 3). The association between urinary cadmium and LINE-1 methylation was significantly stronger in carriers of the CC and CG genotypes of DNMTI rs10854076 than in common GG carriers, and in carriers of the rare GG DNMT1 rs2228611 genotype than in GA and AA carriers. DNMT3B genotypes did not significantly modify the association between urinary cadmium and LINE-1 methylation, although associations were stronger among women with the variant genotypes of both DNMT3B polymorphisms.

Characteristics of the 90 women included in the analyses of DNMT gene expression [see Supplemental Material, Table 1 (] were similar to those of all participants (Table 1), except that the subset was slightly younger (median, 32 vs. 34 years; p = 0.019). DNMT1 expression in blood was much higher than DNMT3B expression (Table 4). DNMT3B _1733929 [lumina HT-12 transcripts are named ILMN(number)] expression was significantly higher in DNMT1 rs10854076 CC and CG carriers than in GG carriers, and in DNMT3B rs2424913 TT carriers than in rare C allele carriers (Table 4). Urinary cadmium concentration was inversely correlated with expression of one DNMT3B transcript (ILMN_1733929; [r.sub.S] = -0.28, p = 0.0086) but not with the other four transcripts ([r.sub.S] = -0.12 to 0.20; p = 0.062-0.78) or with DNMT1 expression ([r.sub.S] = -0.075, p = 0.48). Urinary cadmium also was not associated with DNMT1 expression when stratified by DNMT1 rs10854076 genotype (CC + CG carriers: ([beta] = -19; 95% CI: -88, 50; GG carriers: [beta] = 6.0; 95% CI: -63, 75) or by DNMT1 rs2228611 genotype (GG + GA carriers: ([beta] = -17.6; 95% CI: -80, 45; AA allele carriers: [beta] = -0.60; 95% CI: -75, 74). There were no correlations between DNMT1 or DNMT3B (ILMN_1733929) expression and LINE-1 methylation ([r.sub.S] = 0.050 and -0.16, respectively; p-values = 0.64 and 0.14).


We found that low-level urinary cadmium in women, which represents life-long cadmium exposure, was inversely associated with LINE-1 methylation, a marker of global DNA methylation. Urinary cadmium exposure was also inversely associated with the expression of DNMT3B, which is involved in de novo CpG methylation, but not with expression of DNMT1, which is primarily involved in maintaining CpG methylation patterns in the genome. The inverse association between urinary cadmium and LINE-1 methylation varied according to DNMT1 polymorphisms; the inverse association was stronger among carriers of the rare alleles of DNMT1 rs10854076 and rs2228611. The gene expression data from blood show associations between polymorphisms DNMT1 rsl 0854076 and DNMT3B rs2424913 and DNMT3B gene expression.

Most likely, diet was the main source of cadmium for the study subjects. Only a few of the women smoked cigarettes, another major source of cadmium exposure, and judging from their blood cadmium concentrations (0.18-0.52 [micro]g/L), it can be assumed they did not smoke much. Furthermore, the study area has minimal industrial pollution arid little vehicle traffic. The urine cadmium concentrations in our study population (median, 0.23 [micro]g/L; range, 0.013-1.5 [micro]g/L; or a median of 0.26 [micro]g/g creatinine) were similar to those reported for women of similar age in the United States (median, 0.19 [micro]g/g creatinine) (Gallagher et al. 2011). A slightly higher concentration has been reported for an Asian population (median, 0.59 [micro]g/L), which is possibly due to dietary exposure through rice, which easily takes up cadmium from soil (Kippler et al. 2007). Thus, if the observed association between cadmium and hypomethylation reflects a causal relation, effects may be common and stronger in groups with higher exposures, such as smokers and occupationally exposed workers. However, further studies in other populations, including occupationally exposed individuals, are warranted.

CpG sites in LINE-1 promoters are usually heavily methylated, and genome-wide methylation loss from these sites has been regarded as a common epigenetic event in malignancies that may play a role in carcinogenesis (Cho et al. 2010; Choi et al. 2009; Wilhelm et at 2010). /here is increasing concern that cadmium acts as a metalloestrogen (Fechner et al. 2011) and that it might increase the risk of hormone-related cancers in humans (Akesson et al. 2008; Gallagher et al. 2010; Siewit et al. 2010). The association found in the present study between cadmium and LINE-1 hypomethylation in women suggests that cadmium can alter DNA methylation, which could be another mechanism for cadmium carcinogenesis. in vitro data lend some support to this hypothesis: Takiguchi et al. (2003) reported that cadmium exposure of non transformed rat cells resulted in the inhibition of methyltransferase activity and, at higher exposures, hypomethylation. They also showed that after 10 weeks of exposure the cells became malignantly transformed and had increased methyltransferase activity. This phenomenon has also been shown in malignantly transformed prostate epithelial cells: cadmium exposure increased DNMT enzymatic activity and overexpression of DNMT3B accompanied by hypermethylation (Benbrahim-Tallaa et al. 2007).

Little is known about the potential effects of environmental pollutants on LINE-1 methylation. In a recent U.S. bladder cancer case--control study, higher toenail concentrations of arsenic (> 75th percentile) were associated with reduced LINE-1 methylation (Wilhelm et al. 2010). In contrast, urinary arsenic was not associated with LINE-1 in the present study. Possibly, this can be explained by specific genetic polymorphisms in the AS3MT [arsenic (+3 oxidation state) methyltransferase] gene that previously were found in the present study population to be associated with a very efficient methylation and detoxification of inorganic arsenic (Engstrom K et al. 2011). Other factors may also influence LINE-1 methylation. Exposure to coarse particulate matter (particulate matter with aerodynamic diameter [less than or equal to] 10 pm) (Tarantini et al. 2009) and black carbon (Baccarelli et al. 2009) were found to be inversely associated with LINE-I methylation in peripheral blood cells from humans. Maternal prenatal smoking was associated with lower LINE-I methylation in buccal cells of preschool children with glutathione S-transferase mu 1 (GSTM1)-null genotype, whereas there was higher methylation in the GSTM1-present children (Breton et al. 2009).

The maintenance of methylation in somatic cells is largely carried out by DNMTI (Chen and Li 2006), and DNMTI expression in normal peripheral blood has previously been reported to be higher than DNMT3B expression (Mizuno et al. 2001), consistent with our findings. DNMTI and DNMT3B colocalize in the nucleus (Kim et al. 2002), and it has been shown that in cancer cells DNMTI and DNMT3B work cooperatively to maintain DNA methylation (Rhee et al. 2002). Of note, the allelic frequencies of DNMT3B single-nucleotide polymorphisms were low in the present study, limiting the possibility to detect modest gene--environment interactions. Further, because the levels of DNMT3B expression were lower than those of DNMTI, these findings should be interpreted cautiously.

There was no association between blood cadmium concentrations, which represent a measure of short-term exposure, and our measure of global DNA methylation, although cadmium levels in blood and urine were moderately correlated. If a causal relation exists between cadmium exposure and DNA methylation, one may speculate that the cadmium effect is chronic in nature or occurred a long time before the analyses, even early in life. Furthermore, we found no effect of blood hemoglobin, selenium, or zinc status on the epigenetic markers. The lack of association between selenium concentrations and global DNA methylation in our study population is in contrast to the results of Pilsner et al. (2011), who measured plasma selenium in relation to global leukocyte DNA methylation. The difference in results may reflect differences in selenium status; the women in the present study all had adequate blood selenium concentrations (138-250 [micro]g/L), whereas the Bangladeshi individuals studied by Pilsner et al. (2011) showed plasma concentrations of 45-149 [micro]g/L (plasma concentrations are somewhat lower than selenium in whole blood), indicating lower selenium status. Other differences include the methods used to measure global DNA methyladon: Pilsner et al. (2011) used a [[.sup.3]H]-methyl incorporation assay assessing methylation at all genomic CpG sites, whereas we used bisulfite-PCR pyrosequencing to quantitate LINE-I methylation at four CpG sites, which served as a surrogate for global DNA methylation.

We conducted this study on a fairly homogeneous population in which subjects differed very little in terms of diet, air pollution, and other lifestyle factors, including smoking and alcohol consumption. This homogeneity may explain the small variation in degree of LINE-1 methylation we found compared with other studies (Choi et al. 2009; Tarantini et al. 2009). Also, we adjusted for other essential and toxic elements, such as arsenic, but these elements did not appear to modify the association between cadmium and LINE-1 methylation. A limitation of our study is that we measured methylation in the DNA isolated from peripheral blood only and not from individual cell populations or other tissues. DNA methylation patterns may differ among blood cell types (Wu et al. 2011).


We found that low-level environmental cadmium exposure was associated with a marker of global DNA hypomethylation in peripheral blood. 'Ibis association was modified in relation to DNMTI genotypes. Future studies are necessary to confirm our findings in other populations and determine whether cadmium-associated epigenetic modification is present in other tissues, and if epigenetic effects of cadmium may play a role in cadmium-associated diseases.


Akesson A, Julin B, Wolk A. 2008. Long-term dietary cadmium intake and postmenopausal endometrial cancer incidence: a population-based prospective cohort study. Cancer Res 68(15):6435-6441.

Baccarelli A, Wright RO, Bollati V, Tarantini L, Litonjua AA, Suh HH, et al. 2009. Rapid DNA methylation changes after exposure to traffic particles. Am J Respir Crit Care Med 179(7):572-578.

Benbrahim-Tallaa L, Waterland RA, Dill AL, Webber MM, Waalkes MP. 2007. Tumor suppressor gene inactivation during cadmium-induced malignant transformation of human prostate cells correlates with overexpression of de novo DNA methyltransferase. Environ Health Perspect 115:1454-1459.

Berglund M, Akesson A, Nermell B, Vahter M. 1994. Intestinal absorption of dietary cadmium in women depends on body iron stores and fiber intake. Environ Health Perspect 102:1058-1066.

Breton CV, Byun HM, Wenten M, Pan F, Yang A, Gilliland FD. 2009. Prenatal tobacco smoke exposure affects global and gene-specific DNA methylation. Am J Respir Crit Care Med 180(5):462-467.

Broberg K, Concha 6, Engstrom K, Lindvall M, Grander M, Vahter M. 2011. Lithium in drinking water and thyroid function. Environ Health Perspect 119:827-830.

Chalitchagorn K, Shuangshoti S, Hourpai N, Kongruttanachok N, Tangkijvanich P, Thong-ngam D, et al. 2004. Distinctive pattern of LINE-1 methylation level in normal tissues and the association with carcinogenesis. Oncogene 23(54):8841-8846.

Chanda S, Dasgupta UB, Guhamazumder D. Gupta M, Chaudhuri U, Lahiri S, et al. 2006. DNA hypermethylation of promoter of gene p53 and p16 in arsenic-exposed people with and without malignancy. Toxicol Sci 89(2):431-437.

Chen T, Li E. 2006. Establishment and maintenance of DNA methylation patterns in mammals. Curr Top Microbiol Immunol 301:179-201.

Cho YH, Yazici H, Wu H-C, Terry MB, Gonzalez K, Qu M, et al. 2010. Aberrant promoter hypermethylation and genomic hypomethylation in tumor, adjacent normal tissues and blood from breast cancer patients. Anticancer Res 30(71:2489-2496.

Choi SH, Worswick S, Byun H-M, Shear T, Soussa JC, Wolff EM, et al. 2009. Changes in DNA methylation of tandem DNA repeats are different from interspersed repeats in cancer. Int J Cancer 12513):723-729.

Concha G, Broberg K, Grander M, Cardozo A, Palm B, Vahter M. 2010. High-level exposure to lithium, boron, cesium, and arsenic via drinking water in the Andes of northern Argentina. Environ Sci Techno144(17):6875-6880.

Concha G, Nermell B, Vahter MV. 1998. Metabolism of inorganic arsenic in children with chronic high arsenic exposure in northern Argentina. Environ Health Perspect 106:355-359.

Deng G, Nguyen A, Tanaka H, Matsuzaki K, Bell I, Mehta KR, et al. 2006. Regional hypermethylation and global hypomethylation are associated with altered chromatin conformation and histone acetylation in colorectal cancer. Int J Cancer 118(121:2999-3465.

Engstrom A, Michaelsson K, Suwazono Y, Wolk A, Vahter M, Akesson A. 2011. Long-term cadmium exposure and the association with bone mineral density and fractures in a population-based study among women. J Bone Miner Res 26(3):486-495.

Engstrom KS, Vahter M, Lindh C, Teichert F, Singh R, Concha G, et al. 2010. Low 8-oxo-7,8-dihydro-2'-deoxyguanosine levels and influence of genetic background in an Andean population exposed to high levels of arsenic. Mutat Res 683(1-2):98-105.

Engstrom K, Vahter M, Mlakar SJ, Concha G, Nermell B, Raqib R, et al. 2011. Polymorphisms in arsenic(+111 oxidation state) methyltransferase (AS3M7) predict gene expression of AS3MT as well as arsenic metabolism. Environ 4-tealth Perspect 119:182-188.

Faulk C, Dolinoy DC. 2011. Timing is everything: the when and how of environmentally induced changes in the epigenome of animals. Epigenetics 6(7):791-797.

Fechner P, Damdimopoulou P, Gauglitz 6.2011. Biosensors paving the way to understanding the interaction between cadmium and the estrogen receptor alpha. PLoS One 6(8):e23048; doi:10.1371/journalpone.0023048 [Online 2 August 2011].

Feinberg AP. 2007. Phenotypic plasticity and the epigenetics of human disease. Nature 447(7143):433-440.

Gallagher CM, Chen JJ, Kovach JS. 2010. Environmental cadmium and breast cancer risk. Aging (Albany NY) 2(11):804-814.

Gallagher CM, Chen JJ, Kovach JS. 2011. The relationship between body iron stores and blood and urine cadmium concentrations in US never-smoking, non-pregnant women aged 20-49 years. Environ Res 111(51:702-707.

Gaudet F, Hodgson JO, Eden A, Jackson-Grusby L, Dausman J, Gray JW, et al. 2003. Induction of tumors in mice by genomic hypomethylation. Science 300(5618):489-492.

Hartwig A. 2010. Mechanisms in cadmium-induced carcinogenicity: recent insights. Biometals 23(5):951-960.

Hong KH, Keen CL, Mizuno Y, Johnston KE, Tamura T. 2000. Effects of dietary zinc deficiency on homocysteine and folate metabolism in rats. J Nutr Biochem 11(3):165-169.

Huang D, Zhang Y, Qi Y, Chen C, Ji W. 2008. Global DNA hypomethylation, rather than reactive oxygen species (ROS), a potential facilitator of cadmium-stimulated K562 cell proliferation. Toxicol Lett 179(1):43-47.

Jarup L, Akesson A. 2009. Current status of cadmium as an environmental health problem. Toxicol Appl Pharmacol 238(3):201-208.

Jiang G, Xu L, Song S, Zhu C, Wu Q, Zhang L, et al. 2008. Effects of long-term low-dose cadmium exposure on genomic DNA methylation in human embryo lung fibroblast cells. Toxicology 244(1):49-55.

Johnson MD, Kenney N, Stoica A, Hilakivi-Clarke L, Singh B, Chepko G, et al. 2003. Cadmium mimics the in vivo effects of estrogen in the uterus and mammary gland. Nat Med 9(8):1081-1084.

Kim GD, Ni J, Kelesoglu N, Roberts RJ, Pradhan S. 2002. Co-operation and communication between the human maintenance and de novo DNA (cytosine-5) methyltransferases. EMBO J 21(15):4183-4195.

Kippler M, Ekstrom EC, Lonnerdal B, Goessler W, Akesson A, El Arifeen S, et al. 2007. Influence of iron and zinc status on cadmium accumulation in Bangladeshi women. Toxicol Appl Pharmacol 222(2):221-226.

Kippler M, Goessler W, Nermell B, Ekstrom EC, Lonnerdal B, El Arifeen S, et al. 2009. Factors influencing intestinal cadmium uptake in pregnant Bangladeshi women-a prospective cohort study. Environ Res 109(7):914-921.

Kippler M, Tofail F, Gardner R, Rahman A, Hamadani JD, Bottai M, et al. 2011. Maternal cadmium exposure during pregnancy and size at birth: a prospective cohort study. Environ Health Perspect 120:284-289.

Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, et al. 2001. Initial sequencing and analysis of the human genome. Nature 409(6822):860-921.

Martinez-Zamudio R, Ha HC. 2011. Environmental epigenetics in metal exposure. Epigenetics 6(7):820-827.

Mizuno S, Chijiwa T, Okamura T, Akashi K, Fukumaki Y, Niho Y, et al. 2001. Expression of DNA methyltransferases DNMT1, 3A, and 36 in normal hematopoiesis and in acute and chronic myelogenous leukemia. Blood 97(5):1172-1179.

Monroe RK, Halvorsen SW. 2006. Cadmium blocks receptor-mediated Jak/STAT signaling in neurons by oxidative stress. Free Radic Biol Med 41(3):493-502.

National Library of Medicine. 2012a. GenBank. Available: http://wvwv.ncbi.nlanih.govigenbankl (accessed 29 March 2012).

National Library of Medicine. 2012b. Gene. Available: (accessed 29 March 2012).

Nermell B, Lindberg AL, Rahman M, Berglund M, Persson LA, El Arifeen S, et al. 2008. Urinary arsenic concentration adjustment factors and malnutrition. Environ Res 106(2)212-218.

Pilsner JR, Hall MN, Liu X, Ahsan H, llievski V, Slavkovich V, et al. 2011. Associations of plasma selenium with arsenic and genomic methylation of leukocyte DNA in Bangladesh. Environ Health Perspect 119:113-118.

Rhee I, Bachman KE, Park BH, Jair KW, Yen RW, Schuebel KE, et al. 2002. DNMT1 and DNMT3b cooperate to silence genes in human cancer cells. Nature 416(68801:552-556.

Robertson KD, Jones PA. 2000. DNA methylation: past, present and future directions. Carcinogenesis 21(3):461-467.

Robins JC, Marsit CJ, Padbury JF, Sharma SS. 2011. Endocrine disruptors, environmental oxygen, epigenetics and pregnancy. Front Biosci (Elite Ed) 3:690-700.

Schernhammer ES, Giovannucci E, Kawasaki T, Rosner B, Fuchs CS, Ogino S. 2010. Dietary folate, alcohol and B vitamins in relation to LINE-1 hypomethylation in colon cancer. Gut 59(6):794-799.

Siewit CL, Gengler B, Vegas E, Puckett R, Louie MC. 2010. Cadmium promotes breast cancer cell proliferation by potentiating the interaction between Elirt and c-Jun. Mal Endocrinol 24(51:981-992.

Smeester L, Rager JE, Bailey KA, Guan X, Smith N, Garcia-Vargas G, et al. 2011. Epigenetic changes in individuals with arsenizosis. Chem Res Toxicol 2412)165-187.

Takiguchi M, Achanzar WE, Du W, Li G, Waalkes MP. 2003. Effects of cadmium on DNA-(Cytosine-5) methyltransferase activity and DNA methylation status during cadmium-induced cellular transformation. Exp Cell Res 286(2):355-365.

Tarantini L, Bonzini M, Apostoli P, Pegoraro V, Bollati V, Marinelli B, et al. 2009. Effects of particulate matter on genomic DNA methylation content and iNOS promoter methylation. Environ Health Perspect 117:217-222.

Vahter M, Concha G, Nermell B, Nilsson R, Dulout F, Natarajan AT. 1995. A unique metabolism of inorganic arsenic in native Andean women. Eur J Pharmacol 293(4)*455-462.

Vallon-Christersson J, Nordborg N, Svensson M, I-Mkinen J. 2009. BASE-2nd generation software for microarray data management and analysis. BMC Bioinformatics 10:330; doll 0.1186/1471-2105-10-330 [Online 12 October 2009].

Wilhelm CS, Kelsey KT, Butler R, Plaza S, Gagne L, Zens MS, et al. 2010. Implications of LINE1 methylation for bladder cancer risk in women. Clin Cancer Res 16(5):1682-1689.

Wu H-C, Delgado-Cruzata L, Flom JD, Kappil M, Ferris JS, Liao Y, et al. 2011. Global methylation profiles in DNA from different blood cell types. Epigenetics 6(1):76-85.

Xu B, Chen S, Luo Y, Chen Z, Liu L, Zhou H, et al. 2011. Calcium signaling is involved in cadmium-induced neuronal apoptosis via induction of reactive oxygen species and activation of MAPK/mTOR network. PLoS One 6(4):e19052; doi:10.1371/journal.pone.0019052 [Online 22 April 2011].

Yu X, Robinson JF, Sidhu JS, Hong S, Faustman EM. 2010. A system-based comparison of gene expression reveals alterations in oxidative stress, disruption of ubiquitin-proteasome system and altered cell cycle regulation after exposure to cadmium and methylmercury in mouse embryonic fibroblast. Toxicol Sci 114(21:356-377.

Address correspondence to K. Broberg, Department of Laboratory Medicine, Division of Occupational and Environmental Medicine, Lund University, Klinikgatan 21, SE-22185 Lund, Sweden. Telephone: 46046173819. Fax: 46046143207. E-mail: karin.

Supplemental Material is available online (

We thank the Department of Oncology, Lund University, for technical assistance.

This research was supported by the Swedish Council for Working Life and Social Research, Karolinska Institutet, and Kungliga fysiografiska sallskapet, European Union (FP6; "PHIME" FOOD-CT-2006-016253).

The authors declare they have no actual or potential competing financial interests.

Received 10 October 2011; accepted 1 March 2012.

Mohammad Bakhtiar Hossain, (1), (2) Marie Vahter, (3) Gabriela Concha, (4) and Karin Broberg (l)

(1.) Department of Laboratory Medicine, Division of Occupational and Environmental Medicine, Lund University, Lund, Sweden; (2.) International Centre for Diarrhoeal Disease Research, Bangladesh, Dhaka, Bangladesh; (3.) Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden; (4.) Risk and Benefit Assessment Department, National Food Agency, Uppsala, Sweden
COPYRIGHT 2012 National Institute of Environmental Health Sciences
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Research
Author:Hossain, Mohammad Bakhtiar; Vahter, Marie; Concha, Gabriela; Broberg, Karin
Publication:Environmental Health Perspectives
Article Type:Report
Geographic Code:3ARGE
Date:Jun 1, 2012
Previous Article:Characterization of fine particulate matter and associations between particulate chemical constituents and mortality in Seoul, Korea.
Next Article:Potential for inhalation exposure to engineered nanoparticles from nanotechnology-based cosmetic powders.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters