Associations of Arsenic Exposure With Telomere Length and Naïve T Cells in Childhood—A Birth Cohort Study

Associations of Arsenic Exposure With Telomere Length and Naïve T Cells in Childhood—A Birth... Abstract There is limited knowledge of association between arsenic exposure and telomere length (TL) and signal joint T-cell receptor excision circle (sjTREC) that are potential biomarkers of immune senescence and disease susceptibility. We aimed to clarify whether long-term inorganic arsenic exposure influences TL and sjTRECs in childhood. Children born in a longitudinal mother-child cohort were followed-up at 4.5 (n = 275) and 9 years (n = 351) of age. Arsenic exposure was assessed by metabolite concentrations in urine (U-As) from mothers at gestational week 8 (prenatal) and their children at 4.5 and 9 years. TL and sjTRECs were determined in blood cells using quantitative PCR. The oxidative DNA damage marker 8-hydroxy-2ʹ-deoxyguanosine (8-OHdG) in plasma was measured by ELISA. In multivariable-adjusted spline regression analyses, both prenatal and childhood arsenic exposure above U-As of 45 µg/l were significantly inversely associated with TL and sjTRECs at 9 years. Fraction of monomethylarsonic acid (MMA) above spline knot 7% were significantly inversely associated with both TL and sjTRECs reflecting increased toxicity due to less-efficient arsenic metabolism in 9--year-old children. Prenatal and childhood arsenic exposure were positively associated with 8-OHdG at 9 years which in turn was inversely associated with sjTRECs at 9 years. However, adjustment with 8-OHdG did not change the estimates of the association of U-As with sjTRECs reflecting little contribution from 8-OHdG-induced oxidative stress. Our findings suggest that chronic arsenic exposure from early life can result in TL attrition and lower production of naïve T cells potentially leading to immunosenescence and immunodeficiency. telomere length, immunosenescence, sjTRECs, arsenic, birth cohort, school children Chronic exposure to inorganic arsenic (iAs), well-known environmental pollutant and carcinogen has been associated with various forms of cancers and numerous other pathological effects in humans (Gamboa-Loira et al., 2017). Growing evidence suggests that chronic arsenic exposure during pregnancy and early life is associated with increased risk of infectious diseases, in particular respiratory and gastrointestinal infections (Farzan et al., 2016; Rahman et al., 2011; Raqib et al., 2009; Smith et al., 2013). Several cross-sectional studies have indicated that arsenic adversely affects the immune system in both children and adults (Dangleben et al., 2013). However, very few longitudinal studies in human have further indicated that the adverse effects of the toxicant start from in utero (Ahmed et al., 2011, 2012, 2014, 2017; Raqib et al., 2009, 2017). A telomere is a region of nucleotide repeats (TTAGGG) present at each end of the eukaryotic chromosome which is highly regulated to protect chromosomes from recombination or degradation, ensuring integrity during replication. Telomeres shorten after each replication cycle, hence telomere length (TL) is generally considered to reflect cellular senescence (Blackburn et al., 2015). Many genetic and environmental factors can affect TL such as oxidative stress, diet, smoking, infections and different types of cancer and noncommunicable diseases (Barrett et al., 2015; Gianesin et al., 2016; Zhou et al., 2016). Shorter and longer TL both are associated with risk of cancer (Li et al., 2015; Zhang et al., 2017). Limited studies have addressed the association of arsenic with TL of human immune cells (Ameer et al., 2016; Gao et al., 2015; Li et al., 2012). The critical role of the thymus in the generation of a diversified population of peripheral T lymphocytes is well-established. The quantification of signal joint T-cell receptor excision circle (sjTREC) within peripheral T cell populations provides insight into the frequency of recent thymic emigrants or naïve T cells (Ye and Kirschner, 2002). Arsenic is known to produce reactive oxygen species leading to oxidative stress and DNA damage (Yamauchi et al., 2004). This can be assessed by determining 8-hydroxy-2ʹ-deoxyguanosine (8-OHdG) concentration, a known biomarker of oxidative DNA damage. Our earlier studies have shown that arsenic exposure during pregnancy increased 8-OHdG expression in cord blood and placenta, with increased inflammatory responses and reduced T cell counts in the placenta (Ahmed et al., 2011) and reduced sjTRECs in cord blood (Ahmed et al., 2012). Therefore, in the present longitudinal study, we aimed to investigate whether chronic arsenic exposure from in utero to pre-adolescent age influences TL and whether the negative impact of chronic arsenic exposure on sjTRECs at birth persists in 9-year-old children. Additionally, we measured 8-OHdG in plasma to assess its association with arsenic exposure, sjTRECs, and TL. MATERIALS AND METHODS Study design, area, and participants Matlab, a rural area, is situated about 53 km southeast of Dhaka. icddr,b established a health research and training center with a central hospital and 4 smaller subcenters in Matlab and operates a health and demographic surveillance system (HDSS) covering a population of about 220,000. Elevated arsenic concentrations in groundwater are common in Matlab where approximately 95% of the population uses hand-pumped tubewell water as their main source of drinking water. About 13,000 tubewells in this area were tested for arsenic concentration during 2002–2003, about 70% of which exceeded 10 µg/l of arsenic, the WHO guideline value (WHO, 2011), and >60% exceeded 50 μg/l, the national standard (Rahman et al., 2006). Installation of deep tubewells has resulted in somewhat decreased exposure from the time of pregnancy till 9 years but still remained elevated with 43% of tubewells containing arsenic >10 µg/l (Nahian, 2016). However, substantial arsenic exposure through food probably comes from the use of arsenic-rich irrigation water from shallow pumps (Kippler et al., 2016). This study is part of our ongoing research activities on effects of early-life arsenic exposure and immune function outcomes (Ahmed et al., 2012; Moore et al., 2009). It is nested in a large, randomized, food and multiple micronutrient supplementation trial (MINIMat trial) conducted in 2001–2003 in 4436 women recruited in early pregnancy evaluating nutritional impacts on pregnancy outcomes and child health (Persson et al., 2012). The MINIMat children were followed up at 4.5 years of age (n = 2735) to address different objectives. To reduce the burden of examinations and repeated biological sampling from each child, they were divided into 2 groups based on their year of birth (Group A, April 2002–June 2003; Group B, June 2003–June 2004) (Figure 1). For this study, we selected children from Group B, and followed them at 4.5 years (n = 640) and 9 years (n = 564) (Ahmed et al., 2013, 2014, 2017; Raqib et al., 2017; Mannan et al., 2016). Adequate blood cells for studying sjTRECs and TL were available from 275 children at 4.5 years and 351 children at 9 years of age, and among them 213 children were common in both time points (Figure 1). Figure 1. View largeDownload slide Flow diagram depicting the recruitment of children in the current study nested in the MINIMat birth cohort study in rural Bangladesh carried out in 2012–2013. Figure 1. View largeDownload slide Flow diagram depicting the recruitment of children in the current study nested in the MINIMat birth cohort study in rural Bangladesh carried out in 2012–2013. Maternal anthropometry and family socioeconomic status (SES) were collected in early pregnancy in the MINIMat trial (Persson et al., 2012). The family SES score was estimated via an asset index, generated through principal component analysis of household assets and was updated during the follow-up studies. Anthropometry data of children were collected at 4.5 and 9 years (Raqib et al., 2017). The measured height and weight were converted to height-for-age, weight-for-age, and body mass index-for-age z-scores (SD scores), using the WHO growth reference for school-aged children and adolescents (de Onis et al., 2007). Data on illness including cold, cough, fever, diarrhea, and dysentery in the past 6 months were collected based on structured questionnaire. Comparison of baseline features between the children included in this study and those who were not included did not show any significant differences in general characteristics (Supplementary Table 1). Specimen collection Heparinized venous blood was divided into 2 parts, 1 was used to determine specific cell types in blood as complete blood count with differentials using automated hematology analyzer (Sysmex XT-1800i, Kobe, Japan). The other part was centrifuged to separate plasma from buffy coat. Peripheral blood mononuclear cells (PBMCs) was separated from buffy coat by density-gradient centrifugation and stored in RNALater (Qiagen GmbH, Hilden, Germany) in −80°C. Urine samples collected as described earlier (Ahmed et al., 2017) from pregnant mothers during gestation week 8 (GW8) and their children at 4.5 and 9 years of age were stored at −80°C. Assessment of arsenic exposure iAs is metabolized by a series of reduction and methylation reactions producing monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA), the relative fractions of which reflect efficiency of arsenic metabolism. Increased toxicity of arsenic was shown with the excretion of higher MMA% and a lower DMA% in the urine (Vahter, 2002). Arsenic exposure was assessed based on the concentration of the sum of iAs, MMA, and DMA in urine, hereinafter referred to as urinary arsenic (U-As) reflecting exposure to iAs from both water and food (Diaz et al., 2015; Kippler et al., 2016). The arsenic metabolite concentrations in urine were measured using high-performance liquid chromatography online with hydride generation and inductively coupled plasma mass spectrometry. The intra- and interassay coefficients of variation (CV%) were approximately 4%, based on the measurements of a reference urine sample (CRM No.18, National Institute for Environmental Studies, Tsukuba City, Japan (Ahmed et al., 2014). All U-As concentrations were adjusted to the average specific gravity (1.012 g/ml for all urine samples), measured by a digital refractometer (RD712 Clinical Refractometer; EUROMEX, Arnhem, the Netherlands) (Nermell et al., 2008) using the formula U-As (1.012-1)/(measured specific gravity of each urine sample-1). Estimation of sjTRECs sjTRECs are small circles of DNA created in T-lymphocyte during maturation and rearrangement of T-cell receptor genes in the thymus. sjTRECs contents in blood are used as surrogates of recent thymic emigrant T cells (RTEs) and is a well-known biomarker for thymic output of RTEs or thymic function (Douek et al., 1998; Somech, 2011). From the frozen PBMC, DNA was isolated by using the QIAamp DNA Mini Kit (Qiagen GmbH) according to the manufacturer’s instructions. From 200 ng of DNA/sample, quantification of sjTREC was performed by using SYBR Green real-time quantitative PCR and CFX96TM real time system (C1000 Thermal cycler, Bio-Rad Life Science Research, Hercules, California) as described earlier in Ahmed et al. (2012) and Raqib et al. (2007). The following primers were used: forward primer 5ʹAAAGAGGGCAGCCCTCTCCAAGGCAAA3ʹ and reverse primer 5ʹAGGCT-GATCTTGTCTGACATTTGCTCCG3ʹ. A standard (kindly donated by PT Ngom) was prepared by using serial dilutions of a known number of copies of a fragment of the sjTREC gene sequence and included in each run to generate a standard curve (Ngom et al., 2004). The copy number of sjTRECs in the DNA samples was determined automatically using standard curves and expressed as sjTRECs content (copy numbers/100 ng of DNA). The CV% of intra- and interrun variation of determination of sjTREC level were 8.5% and 10.2% respectively. Estimation of TL Absolute TL in PBMC DNA was measured as described earlier (O’Callaghan and Fenech, 2011) by performing 2 RT-PCR reactions using 70 ng of DNA samples for each reaction in a single plate using 2 different oligomer standards. Telomere oligomer standard was used to establish telomere standard curve in telomere PCR reaction whereas single copy gene 36B4 oligomer standard was used for creating 36B4 standard curve in 36B4 PCR reaction. The primer sequences were: telomere forward primer 5’CGGTTTGTTTGGGTTTGGGTTT-GGGTTTGGG-TTTGGGTT3’; and reverse primer 5’GGCTTGCCTTACCCTTACCCTTACCC-TTACCC-TTACCCT3’; 36B4 forward primer 5’CAGCAAGTGGGAA-GGTGTAATCC3’ and reverse primer 5’CCCATTCTATCATC-AACGGGTACAA3’. Plasmid DNA (pBR322) was used with each standard to maintain a constant 20 ng of total DNA per reaction tube. The data obtained from RT-PCR method to measure absolute TL were analyzed as kb/reaction and genome copies/reaction for telomere and 36B4, respectively. The telomere kb per reaction value is divided by diploid genome copy number of 36B4 to give a total TL in kb per human diploid genome. The CT values within the linear range of the standard curves and between 15 and 35 threshold cycles were used for analysis. The CV% of intra- and interrun variation of TL assessment were <5% and <7.5%, respectively. Oxidative stress marker Plasma concentration of 8-OHdG, a biomarker of oxidative DNA damage was analyzed by a competitive ELISA kit (Highly Sensitive 8-OHdG Check ELISA, Japan Institute for the Control of Aging, Fukuroi, Shizuoka, Japan). Samples were measured in duplicates and analysis was repeated if results of duplicate samples differed >10%. The intra- and interassay CV% were 2.85% and 4.6%, respectively. Statistical analysis Statistical analyses were conducted using the software PASW 20.0 (SPSS Inc., Chicago, Illinois) and Stata/IC 13.0 (StataCorp, College Station, Texas). Associations between exposures, outcomes, and covariates were initially evaluated using Spearman’s rank correlation, Mann-Whitney U-test, analysis of variance, or Kruskal-Wallis test, as appropriate. Associations between different exposure biomarkers (U-As at 4.5 and 9 years; MMA%, DMA%, iAs%) and outcomes (sjTRECs and TL) were examined graphically using lowess moving average curves searching for linear or non-linear associations. When linear associations were obtained, multivariable-adjusted linear regression analyses were carried out. When deviation from linearity was found, we applied splines at apparent thresholds. In scatter plots with Lowess curves (Figs. 2A–F), both TL and sjTRECs slightly increased up to 45 µg/l of U-As and then started to decline showing deviation from a linear pattern. Thus, due to the nature of non-linear curve the associations of U-As at all-time points (GW8, 4.5 and 9 years of age) with TL and sjTRECs were assessed by multivariable-adjusted spline regression analyses with a spline knot introduced at 45 µg/l (log2-transformed: equivalent to 5.5). To obtain normally distributed residuals with a homogeneous variance, all exposure variables (U-As at GW8, 4.5 and 9 years of age) were log2-transformed. We chose log2-transformation of TL and sjTRECs values to simplify the interpretation of the beta-coefficients in the regression analyses (average changes in outcome associated with each doubling of exposure). To analyze the influence of arsenic metabolism efficiency on the associations between U-As and sjTRECs or TL, initially we analyzed by multivariable-adjusted linear regression. Again, we observed nonlinear association between % of arsenic metabolites and TL or sjTRECs, thus spline regression was applied. The spline knot values of the metabolites were introduced at 7, 80 and 15 for MMA%, DMA%, iAs%, respectively (Supplementary Figure 2). Linear mixed effects models were applied to evaluate the effect of exposure over time (4.5 and 9 years of age) on outcome changes (TL and sjTRECs) with a spline knot at 45 µg/l. The association between U-As and 8-OHdG was evaluated by linear regression analysis. The models were adjusted for covariates that were significantly associated with both exposure and outcome or changed the effect estimates by 5% or more. The covariates were child sex, age, HAZ, SES, mother’s education and plasma CRP. The p values < .05 were considered statistically significant. Figure 2. View largeDownload slide Associations between U-As concentration (log2-transformed) at GW8, 4.5 and 9 years and TL (A–C) and sjTRECs (log2-transformed) (D–F) in 9-year-old children. In the scatter plots, the solid line represents a Lowess (locally weighted scatter plot smoothing) moving-average curve for the raw data. Figure 2. View largeDownload slide Associations between U-As concentration (log2-transformed) at GW8, 4.5 and 9 years and TL (A–C) and sjTRECs (log2-transformed) (D–F) in 9-year-old children. In the scatter plots, the solid line represents a Lowess (locally weighted scatter plot smoothing) moving-average curve for the raw data. RESULTS Demographic Data and Arsenic Exposure The median age of the children at the different follow-ups were 4.6 years (range: 4.5–5.2) and 8.8 years (range: 8.6–9.6), respectively (Table 1) and were referred to as 4.5 and 9 years of age. At each follow-up, girls (49%) were shorter (p for all ≤ .003) and weighed less (p for all ≤ .049) than the boys (51%). Median U-As in the children at 9 years was lower than that of their mothers during pregnancy (median 88, range 1.9–1576 µg/l) and at 4.5 years of age indicating gradual decline in arsenic exposure over time. However, the mother’s U-As concentrations were strongly correlated with children’s exposures (rs = 0.48 for 4.5 years and rs = 0.40 for 9 years, both p < .001). Family SES did not change significantly over time from GW8 to 9 years of age. The median TL at 9 years decreased significantly from that at 4.5 years, though there was a strong correlation (Supplementary Figure 1). The sjTRECs content did change significantly over time. Mean plasma concentration of 8-OHdG was significantly higher in 9-years old compared with 4.5-years old (p < .001) indicating accumulating oxidative stress among older children (Table 1). Table 1. General Characteristics of the Study Participants Variablesa  4.5 years (n = 213)  9 years (n = 351)  p-Value  Age, months  55.9 ± 1.4  104.4 ± 1.2  <.001  Gender, male, n (%)  99 (46.5%)  178 (50.7%)    Weight (kg)  13.7 ±1.5  22.0± 3.1  <.001  Height (cm)  100.0 ±4.0  123.0 ± 5.2  <.001  Stuntedb  61 (28.6%)  76 (21.7%)  .912  Underweightb  97 (45.5%)  147 (41.9%)  .452  U-As (µg/l)c  57.1 (12.9, 1125.0)  53.9 (8.9, 1268.0)  .007  TL (kb/dg)  188.9 (68.6–464.0)  164.8 (53.3–455.0)  <.001  sjTRECs (copy number/100 ng DNA)  4.8 (0.23–684.0)×104  5.7 (0.18–84.5)×104  .910  8-OHdG (ng/ml)  2.5 ± 1.4  3.3 ± 1.7  <.001  Family SESd         First tertile, n (%)  71 (33.3%)  117 (33.3%)     second tertile, n (%)  65 (30.5%)  117 (33.3%)  .910   Third tertile, n (%)  77 (36.2%)  117 (33.3%)    Variablesa  4.5 years (n = 213)  9 years (n = 351)  p-Value  Age, months  55.9 ± 1.4  104.4 ± 1.2  <.001  Gender, male, n (%)  99 (46.5%)  178 (50.7%)    Weight (kg)  13.7 ±1.5  22.0± 3.1  <.001  Height (cm)  100.0 ±4.0  123.0 ± 5.2  <.001  Stuntedb  61 (28.6%)  76 (21.7%)  .912  Underweightb  97 (45.5%)  147 (41.9%)  .452  U-As (µg/l)c  57.1 (12.9, 1125.0)  53.9 (8.9, 1268.0)  .007  TL (kb/dg)  188.9 (68.6–464.0)  164.8 (53.3–455.0)  <.001  sjTRECs (copy number/100 ng DNA)  4.8 (0.23–684.0)×104  5.7 (0.18–84.5)×104  .910  8-OHdG (ng/ml)  2.5 ± 1.4  3.3 ± 1.7  <.001  Family SESd         First tertile, n (%)  71 (33.3%)  117 (33.3%)     second tertile, n (%)  65 (30.5%)  117 (33.3%)  .910   Third tertile, n (%)  77 (36.2%)  117 (33.3%)    Abbreviations: U-As, Urinary arsenic; kb/dg, kilo base pair/diploid genome; sjTRECs, signal-joint T cell receptor excision circles; 8-OHdG, 8-hydroxy-2'-deoxyguanosine; SES, Socioeconomic status. a Values are given either as mean ± SD or median with range within brackets. b Defined as children with HAZ, height for age or WAZ, weight for age<−2 SDs from the median value of height or weight for age of reference population (according to WHO). c Adjusted to average specific gravity of 1.012. d SES score was estimated via an asset index, generated through principal component analysis of household assets. Table 1. General Characteristics of the Study Participants Variablesa  4.5 years (n = 213)  9 years (n = 351)  p-Value  Age, months  55.9 ± 1.4  104.4 ± 1.2  <.001  Gender, male, n (%)  99 (46.5%)  178 (50.7%)    Weight (kg)  13.7 ±1.5  22.0± 3.1  <.001  Height (cm)  100.0 ±4.0  123.0 ± 5.2  <.001  Stuntedb  61 (28.6%)  76 (21.7%)  .912  Underweightb  97 (45.5%)  147 (41.9%)  .452  U-As (µg/l)c  57.1 (12.9, 1125.0)  53.9 (8.9, 1268.0)  .007  TL (kb/dg)  188.9 (68.6–464.0)  164.8 (53.3–455.0)  <.001  sjTRECs (copy number/100 ng DNA)  4.8 (0.23–684.0)×104  5.7 (0.18–84.5)×104  .910  8-OHdG (ng/ml)  2.5 ± 1.4  3.3 ± 1.7  <.001  Family SESd         First tertile, n (%)  71 (33.3%)  117 (33.3%)     second tertile, n (%)  65 (30.5%)  117 (33.3%)  .910   Third tertile, n (%)  77 (36.2%)  117 (33.3%)    Variablesa  4.5 years (n = 213)  9 years (n = 351)  p-Value  Age, months  55.9 ± 1.4  104.4 ± 1.2  <.001  Gender, male, n (%)  99 (46.5%)  178 (50.7%)    Weight (kg)  13.7 ±1.5  22.0± 3.1  <.001  Height (cm)  100.0 ±4.0  123.0 ± 5.2  <.001  Stuntedb  61 (28.6%)  76 (21.7%)  .912  Underweightb  97 (45.5%)  147 (41.9%)  .452  U-As (µg/l)c  57.1 (12.9, 1125.0)  53.9 (8.9, 1268.0)  .007  TL (kb/dg)  188.9 (68.6–464.0)  164.8 (53.3–455.0)  <.001  sjTRECs (copy number/100 ng DNA)  4.8 (0.23–684.0)×104  5.7 (0.18–84.5)×104  .910  8-OHdG (ng/ml)  2.5 ± 1.4  3.3 ± 1.7  <.001  Family SESd         First tertile, n (%)  71 (33.3%)  117 (33.3%)     second tertile, n (%)  65 (30.5%)  117 (33.3%)  .910   Third tertile, n (%)  77 (36.2%)  117 (33.3%)    Abbreviations: U-As, Urinary arsenic; kb/dg, kilo base pair/diploid genome; sjTRECs, signal-joint T cell receptor excision circles; 8-OHdG, 8-hydroxy-2'-deoxyguanosine; SES, Socioeconomic status. a Values are given either as mean ± SD or median with range within brackets. b Defined as children with HAZ, height for age or WAZ, weight for age<−2 SDs from the median value of height or weight for age of reference population (according to WHO). c Adjusted to average specific gravity of 1.012. d SES score was estimated via an asset index, generated through principal component analysis of household assets. Chronic Arsenic Exposure and TL Adjusted regression models showed significant inverse associations between maternal (GW8) and childhood U-As (4.5 and 9 years of age) and TL at 9 years of age above the spline knot 5.5 (Table 2). In contrast, concurrent U-As below the spline knot was significantly positively associated with TL at 9 years of age. In 4.5-year-old children, U-As at GW8 and concurrent U-As above the spline knot were inversely associated with TL, although the associations were not statistically significant. These associations remained unchanged when total available children (n = 275) at 4.5 years of age were considered (Figure 1, Supplementary Table 2). The estimates of the associations between U-As at GW8 with TL at 4.5 (β = −9.7) and 9 years (β = −10.4) were similar (Table 2). When association between U-As at 9 years and TL were further adjusted with blood cell types, we did not find any significant influence of the various cell types (data not shown). Table 2. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years and Arsenic Metabolites at 9 Years With TL at 4.5 and 9 Years of Age TL  Children at 4.5 Years (n = 213)  Children at 9 Years (n = 351)  β (95% CI)a  β (95% CI)a  U-As at GW8       <5.5b  7.1 (–26.9, 41.0)  10.4 (–8.4, 29.1)   >5.5  –9.7 (–19.4, 0.05)*  –10.4 (–16.9, –3.9)**  U-As at 4.5 years       <5.5b  –8.7 (–42.8, 25.4)  –3.3 (–25.7, 19.2)   >5.5  –4.7 (–16.0, 6.6)  –15.6 (–23.6, –7.6)**  U-As at 9 years       <5.5b  —  23.2 (5.9, 41.6)**   >5.5  —  –33.9 (–41.9, –25.9) **  MMA in urine (%)       <7.0b  0.01 (–0.08, 0.10)  0.02 (–0.05, 0.08)   >7.0  –0.02 (–0.05, 0.01)  –0.05 (–0.07, –0.02)**  DMA in urine (%)       <80.0b  0.01 (–0.02, 0.03)  0.02 (–0.004, 0.04)   >80.0  0.003 (–0.02, 0.03)  0.02 (–0.01, 0.04)  iAs in urine (%)       <15.0b  0.02 (–0.01, 0.04)  –0.001 (–0.02, 0.02)   >15.0  –0.07 (–0.13, –0.003)**  –0.03 (–0.09, 0.02)  TL  Children at 4.5 Years (n = 213)  Children at 9 Years (n = 351)  β (95% CI)a  β (95% CI)a  U-As at GW8       <5.5b  7.1 (–26.9, 41.0)  10.4 (–8.4, 29.1)   >5.5  –9.7 (–19.4, 0.05)*  –10.4 (–16.9, –3.9)**  U-As at 4.5 years       <5.5b  –8.7 (–42.8, 25.4)  –3.3 (–25.7, 19.2)   >5.5  –4.7 (–16.0, 6.6)  –15.6 (–23.6, –7.6)**  U-As at 9 years       <5.5b  —  23.2 (5.9, 41.6)**   >5.5  —  –33.9 (–41.9, –25.9) **  MMA in urine (%)       <7.0b  0.01 (–0.08, 0.10)  0.02 (–0.05, 0.08)   >7.0  –0.02 (–0.05, 0.01)  –0.05 (–0.07, –0.02)**  DMA in urine (%)       <80.0b  0.01 (–0.02, 0.03)  0.02 (–0.004, 0.04)   >80.0  0.003 (–0.02, 0.03)  0.02 (–0.01, 0.04)  iAs in urine (%)       <15.0b  0.02 (–0.01, 0.04)  –0.001 (–0.02, 0.02)   >15.0  –0.07 (–0.13, –0.003)**  –0.03 (–0.09, 0.02)  Abbreviations: TL, Telomere length; β, regression coefficient; CI, confidence interval; U-As, sum of urinary arsenic metabolites; GW, gestation week; MMA, monomethylarsonic acid; DMA, dimethylarsinic acid; iAs, inorganic arsenic. a Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein, and total leukocytes at 9 years, family socioeconomic status at 9 years, mothers’ education. b Spline regression model using spline knot at log2 U-As 5.5 (corresponding to 45 µg/l), fraction of MMA at 7%, fraction of DMA at 80% and fraction of iAs at 15%. * indicates p < .10 and **indicates p < .05. Table 2. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years and Arsenic Metabolites at 9 Years With TL at 4.5 and 9 Years of Age TL  Children at 4.5 Years (n = 213)  Children at 9 Years (n = 351)  β (95% CI)a  β (95% CI)a  U-As at GW8       <5.5b  7.1 (–26.9, 41.0)  10.4 (–8.4, 29.1)   >5.5  –9.7 (–19.4, 0.05)*  –10.4 (–16.9, –3.9)**  U-As at 4.5 years       <5.5b  –8.7 (–42.8, 25.4)  –3.3 (–25.7, 19.2)   >5.5  –4.7 (–16.0, 6.6)  –15.6 (–23.6, –7.6)**  U-As at 9 years       <5.5b  —  23.2 (5.9, 41.6)**   >5.5  —  –33.9 (–41.9, –25.9) **  MMA in urine (%)       <7.0b  0.01 (–0.08, 0.10)  0.02 (–0.05, 0.08)   >7.0  –0.02 (–0.05, 0.01)  –0.05 (–0.07, –0.02)**  DMA in urine (%)       <80.0b  0.01 (–0.02, 0.03)  0.02 (–0.004, 0.04)   >80.0  0.003 (–0.02, 0.03)  0.02 (–0.01, 0.04)  iAs in urine (%)       <15.0b  0.02 (–0.01, 0.04)  –0.001 (–0.02, 0.02)   >15.0  –0.07 (–0.13, –0.003)**  –0.03 (–0.09, 0.02)  TL  Children at 4.5 Years (n = 213)  Children at 9 Years (n = 351)  β (95% CI)a  β (95% CI)a  U-As at GW8       <5.5b  7.1 (–26.9, 41.0)  10.4 (–8.4, 29.1)   >5.5  –9.7 (–19.4, 0.05)*  –10.4 (–16.9, –3.9)**  U-As at 4.5 years       <5.5b  –8.7 (–42.8, 25.4)  –3.3 (–25.7, 19.2)   >5.5  –4.7 (–16.0, 6.6)  –15.6 (–23.6, –7.6)**  U-As at 9 years       <5.5b  —  23.2 (5.9, 41.6)**   >5.5  —  –33.9 (–41.9, –25.9) **  MMA in urine (%)       <7.0b  0.01 (–0.08, 0.10)  0.02 (–0.05, 0.08)   >7.0  –0.02 (–0.05, 0.01)  –0.05 (–0.07, –0.02)**  DMA in urine (%)       <80.0b  0.01 (–0.02, 0.03)  0.02 (–0.004, 0.04)   >80.0  0.003 (–0.02, 0.03)  0.02 (–0.01, 0.04)  iAs in urine (%)       <15.0b  0.02 (–0.01, 0.04)  –0.001 (–0.02, 0.02)   >15.0  –0.07 (–0.13, –0.003)**  –0.03 (–0.09, 0.02)  Abbreviations: TL, Telomere length; β, regression coefficient; CI, confidence interval; U-As, sum of urinary arsenic metabolites; GW, gestation week; MMA, monomethylarsonic acid; DMA, dimethylarsinic acid; iAs, inorganic arsenic. a Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein, and total leukocytes at 9 years, family socioeconomic status at 9 years, mothers’ education. b Spline regression model using spline knot at log2 U-As 5.5 (corresponding to 45 µg/l), fraction of MMA at 7%, fraction of DMA at 80% and fraction of iAs at 15%. * indicates p < .10 and **indicates p < .05. To evaluate long-term changes of arsenic on TL, we evaluated the association of childhood exposure to arsenic in relation to TL at 4.5 and 9 years of age by mixed effect models. A significant inverse association was found between childhood arsenic exposure above the spline knot and childhood TL (β = −15.4, 95% CI = −22.19, −8.64; p < .001). The TL was significantly inversely associated with the fraction of MMA above the spline knot of 7% in 9 years old and above 15% for iAs in 4.5-year-old children (Table 2) likely reflecting increased toxicity due to poor methylation of arsenic in these children. Chronic Arsenic Exposure, sjTRECs, and Immune Cells Significant inverse associations were found between U-As at GW8, 4.5 and 9 years of age and sjTRECs at 9 years of age above the spline knot 5.5 (Table 3). Only in 9-year-old children a significant positive association of sjTRECs was obtained with U-As below the spline knot. To examine whether elevated levels of sjTRECs were linked with illness in these children, we evaluated association between sjTREC concentrations and morbidity outcomes and found no association (data not shown). Again, when association between U-As and sjTRECs in 9-years old were further adjusted with cell types, we did not find any significant influence of the various cell types (data not shown). Table 3. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years and Arsenic Metabolites at 9 Years With sjTRECs at 4.5 and 9 Years of Age sjTRECs  4.5-Year-Old Children (n = 213)  9-Year-Old Children (n = 349)   Model 1 β (95% CI)a  Model 1 β (95% CI)a  Model 2 β (95% CI)b  U-As at GW8   <5.5c  0.72 (–0.02, 1.5)  0.29 (–0.23, 0.81)  0.32 (–0.20, 0.85)   >5.5  –0.12 (–0.33, 0.09)  –0.43 (–0.61, –0.25)*  –0.42 (–0.61, –0.24)*  U-As at 4.5 years   <5.5c  0.36 (–0.39, 1.1)  0.03 (–0.58, 0.65)  0.09 (–0.55, 0.73)   >5.5  –0.11 (–0.35, 0.14)  –0.62 (–0.84, –0.40)*  –0.59 (–0.82, –0.36)*  U-As at 9 years   <5.5c  —  1.3 (0.87, 1.8)*  1.3 (0.80, 1.8)*   >5.5  —  –1.3 (–1.5, –1.1)*  –1.3 (–1.5, –1.1)*  MMA in urine (%)   <7.0c  –0.03 (–0.32, 0.26)  0.19 (–0.04, 0.42)  0.04 (–0.08, 0.16)   >7.0  –0.05 (–0.14, 0.04)  –0.16 (–0.24, –0.08)*  –0.26 (–0.38, –0.14)*  DMA in urine (%)   <80.0c  –0.03 (–0.11, 0.06)  0.03 (–0.04, 0.11)  0.05 (–0.03, 0.14)   >80.0  0.04 (–0.04, 0.13)  0.03 (–0.04, 0.10)  0.02 (–0.05, 0.10)  iAs in urine (%)   <15.0c  –0.02 (–0.10, 0.07)  0.01 (–0.07, 0.08)  –0.01 (–0.09, 0.07)   >15.0  0.15 (–0.06, 0.36)  0.04 (–0.16, 0.23)  0.14 (–0.10, 0.38)  sjTRECs  4.5-Year-Old Children (n = 213)  9-Year-Old Children (n = 349)   Model 1 β (95% CI)a  Model 1 β (95% CI)a  Model 2 β (95% CI)b  U-As at GW8   <5.5c  0.72 (–0.02, 1.5)  0.29 (–0.23, 0.81)  0.32 (–0.20, 0.85)   >5.5  –0.12 (–0.33, 0.09)  –0.43 (–0.61, –0.25)*  –0.42 (–0.61, –0.24)*  U-As at 4.5 years   <5.5c  0.36 (–0.39, 1.1)  0.03 (–0.58, 0.65)  0.09 (–0.55, 0.73)   >5.5  –0.11 (–0.35, 0.14)  –0.62 (–0.84, –0.40)*  –0.59 (–0.82, –0.36)*  U-As at 9 years   <5.5c  —  1.3 (0.87, 1.8)*  1.3 (0.80, 1.8)*   >5.5  —  –1.3 (–1.5, –1.1)*  –1.3 (–1.5, –1.1)*  MMA in urine (%)   <7.0c  –0.03 (–0.32, 0.26)  0.19 (–0.04, 0.42)  0.04 (–0.08, 0.16)   >7.0  –0.05 (–0.14, 0.04)  –0.16 (–0.24, –0.08)*  –0.26 (–0.38, –0.14)*  DMA in urine (%)   <80.0c  –0.03 (–0.11, 0.06)  0.03 (–0.04, 0.11)  0.05 (–0.03, 0.14)   >80.0  0.04 (–0.04, 0.13)  0.03 (–0.04, 0.10)  0.02 (–0.05, 0.10)  iAs in urine (%)   <15.0c  –0.02 (–0.10, 0.07)  0.01 (–0.07, 0.08)  –0.01 (–0.09, 0.07)   >15.0  0.15 (–0.06, 0.36)  0.04 (–0.16, 0.23)  0.14 (–0.10, 0.38)  Abbreviations: sjTRECs, signal-joint T cell receptor excision circles; β, regression coefficient; CI, confidence interval; U-As, sum of urinary arsenic metabolites; GW, gestation week; MMA, monomethylarsonic acid; DMA, dimethylarsinic acid; iAs, inorganic arsenic. sjTRECs were log2-transformed. a Model 1. Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein and lymphocytes at 9 years, socioeconomic status of the family at 9 years of age, mothers’ education. b Model 2. Variables additionally adjusted by 8-OHdG at 9 years. c Spline regression model using spline knot at log2 U-As 5.5 (corresponding to 45 µg/l), fraction of MMA at 7%, fraction of DMA at 80% and fraction of iAs at 15%. * indicates p < .05. Table 3. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years and Arsenic Metabolites at 9 Years With sjTRECs at 4.5 and 9 Years of Age sjTRECs  4.5-Year-Old Children (n = 213)  9-Year-Old Children (n = 349)   Model 1 β (95% CI)a  Model 1 β (95% CI)a  Model 2 β (95% CI)b  U-As at GW8   <5.5c  0.72 (–0.02, 1.5)  0.29 (–0.23, 0.81)  0.32 (–0.20, 0.85)   >5.5  –0.12 (–0.33, 0.09)  –0.43 (–0.61, –0.25)*  –0.42 (–0.61, –0.24)*  U-As at 4.5 years   <5.5c  0.36 (–0.39, 1.1)  0.03 (–0.58, 0.65)  0.09 (–0.55, 0.73)   >5.5  –0.11 (–0.35, 0.14)  –0.62 (–0.84, –0.40)*  –0.59 (–0.82, –0.36)*  U-As at 9 years   <5.5c  —  1.3 (0.87, 1.8)*  1.3 (0.80, 1.8)*   >5.5  —  –1.3 (–1.5, –1.1)*  –1.3 (–1.5, –1.1)*  MMA in urine (%)   <7.0c  –0.03 (–0.32, 0.26)  0.19 (–0.04, 0.42)  0.04 (–0.08, 0.16)   >7.0  –0.05 (–0.14, 0.04)  –0.16 (–0.24, –0.08)*  –0.26 (–0.38, –0.14)*  DMA in urine (%)   <80.0c  –0.03 (–0.11, 0.06)  0.03 (–0.04, 0.11)  0.05 (–0.03, 0.14)   >80.0  0.04 (–0.04, 0.13)  0.03 (–0.04, 0.10)  0.02 (–0.05, 0.10)  iAs in urine (%)   <15.0c  –0.02 (–0.10, 0.07)  0.01 (–0.07, 0.08)  –0.01 (–0.09, 0.07)   >15.0  0.15 (–0.06, 0.36)  0.04 (–0.16, 0.23)  0.14 (–0.10, 0.38)  sjTRECs  4.5-Year-Old Children (n = 213)  9-Year-Old Children (n = 349)   Model 1 β (95% CI)a  Model 1 β (95% CI)a  Model 2 β (95% CI)b  U-As at GW8   <5.5c  0.72 (–0.02, 1.5)  0.29 (–0.23, 0.81)  0.32 (–0.20, 0.85)   >5.5  –0.12 (–0.33, 0.09)  –0.43 (–0.61, –0.25)*  –0.42 (–0.61, –0.24)*  U-As at 4.5 years   <5.5c  0.36 (–0.39, 1.1)  0.03 (–0.58, 0.65)  0.09 (–0.55, 0.73)   >5.5  –0.11 (–0.35, 0.14)  –0.62 (–0.84, –0.40)*  –0.59 (–0.82, –0.36)*  U-As at 9 years   <5.5c  —  1.3 (0.87, 1.8)*  1.3 (0.80, 1.8)*   >5.5  —  –1.3 (–1.5, –1.1)*  –1.3 (–1.5, –1.1)*  MMA in urine (%)   <7.0c  –0.03 (–0.32, 0.26)  0.19 (–0.04, 0.42)  0.04 (–0.08, 0.16)   >7.0  –0.05 (–0.14, 0.04)  –0.16 (–0.24, –0.08)*  –0.26 (–0.38, –0.14)*  DMA in urine (%)   <80.0c  –0.03 (–0.11, 0.06)  0.03 (–0.04, 0.11)  0.05 (–0.03, 0.14)   >80.0  0.04 (–0.04, 0.13)  0.03 (–0.04, 0.10)  0.02 (–0.05, 0.10)  iAs in urine (%)   <15.0c  –0.02 (–0.10, 0.07)  0.01 (–0.07, 0.08)  –0.01 (–0.09, 0.07)   >15.0  0.15 (–0.06, 0.36)  0.04 (–0.16, 0.23)  0.14 (–0.10, 0.38)  Abbreviations: sjTRECs, signal-joint T cell receptor excision circles; β, regression coefficient; CI, confidence interval; U-As, sum of urinary arsenic metabolites; GW, gestation week; MMA, monomethylarsonic acid; DMA, dimethylarsinic acid; iAs, inorganic arsenic. sjTRECs were log2-transformed. a Model 1. Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein and lymphocytes at 9 years, socioeconomic status of the family at 9 years of age, mothers’ education. b Model 2. Variables additionally adjusted by 8-OHdG at 9 years. c Spline regression model using spline knot at log2 U-As 5.5 (corresponding to 45 µg/l), fraction of MMA at 7%, fraction of DMA at 80% and fraction of iAs at 15%. * indicates p < .05. The mixed effect model with spline function showed an inverse association between childhood arsenic exposure and sjTRECs (β = −0.68, 95% CI = −0.85, −0.49; p < .001) above the spline knot at 5.5. However, a positive association was found between childhood arsenic exposure and sjTRECs in all children (β = 0.70, 95% CI = 0.29, 1.10; p < .001) below the spline knot. Further adjustment with U-As at GW8 did not change the estimates. Significant negative correlation of MMA% with sjTRECs levels was found above spline knot of 7% at 9 years of age. No other significant association of sjTRECs was found with other arsenic metabolites at either time points (Table 3). Arsenic Exposure, Oxidative Stress, TL, and sjTRECs Prenatal and childhood U-As were significantly positively associated with plasma 8-OHdG at 9 years of age, but not at 4.5 years of age (Table 4). Again, linear regression analysis showed a significant inverse association between 8-OHdG and sjTRECs at 9 years of age (β = −0.17; 95% CI, −0.30, −0.04; p = .01), but not at 4.5 years of age (β = −0.15; 95% CI, −0.37, 0.08; p = .195). However, we did not observe similar associations between TL and 8-OHdG either at 4.5 or 9 years of age. Table 4. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years With 8-OHdG at 4.5 and 9 Years of Age   8-OHdG at 4.5 Years (n = 213)   8-OHdG at 9 Years (n = 335)     Unadjusted  Adjusteda  Unadjusted  Adjusteda  U-As at GW8  –0.06 (–0.20, 0.07)  –0.06 (–0.19, 0.08)  0.13 (0.02, 0.24)*  0.13 (0.01, 0.24)*  U-As at 4.5 years  –0.07 (–0.24, 0.09)  –0.08 (–0.24, 0.09)  0.23 (0.09, 0.36)*  0.22 (0.08, 0.36)*  U-As at 9 years  —  —  0.30 (0.17, 0.43)*  0.32 (0.19, 0.45)*    8-OHdG at 4.5 Years (n = 213)   8-OHdG at 9 Years (n = 335)     Unadjusted  Adjusteda  Unadjusted  Adjusteda  U-As at GW8  –0.06 (–0.20, 0.07)  –0.06 (–0.19, 0.08)  0.13 (0.02, 0.24)*  0.13 (0.01, 0.24)*  U-As at 4.5 years  –0.07 (–0.24, 0.09)  –0.08 (–0.24, 0.09)  0.23 (0.09, 0.36)*  0.22 (0.08, 0.36)*  U-As at 9 years  —  —  0.30 (0.17, 0.43)*  0.32 (0.19, 0.45)*  Abbreviations: 8-OHdG, 8-hydroxy-2’-deoxyguanosine; U-As, sum of urinary arsenic metabolites; GW, gestation week. Data were expressed as regression coefficient (β) and 95% confidence intervals (CI). a Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein at 9 years, family socioeconomic status at 9 years of age, and mothers’ education. * indicates p < .05. Table 4. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years With 8-OHdG at 4.5 and 9 Years of Age   8-OHdG at 4.5 Years (n = 213)   8-OHdG at 9 Years (n = 335)     Unadjusted  Adjusteda  Unadjusted  Adjusteda  U-As at GW8  –0.06 (–0.20, 0.07)  –0.06 (–0.19, 0.08)  0.13 (0.02, 0.24)*  0.13 (0.01, 0.24)*  U-As at 4.5 years  –0.07 (–0.24, 0.09)  –0.08 (–0.24, 0.09)  0.23 (0.09, 0.36)*  0.22 (0.08, 0.36)*  U-As at 9 years  —  —  0.30 (0.17, 0.43)*  0.32 (0.19, 0.45)*    8-OHdG at 4.5 Years (n = 213)   8-OHdG at 9 Years (n = 335)     Unadjusted  Adjusteda  Unadjusted  Adjusteda  U-As at GW8  –0.06 (–0.20, 0.07)  –0.06 (–0.19, 0.08)  0.13 (0.02, 0.24)*  0.13 (0.01, 0.24)*  U-As at 4.5 years  –0.07 (–0.24, 0.09)  –0.08 (–0.24, 0.09)  0.23 (0.09, 0.36)*  0.22 (0.08, 0.36)*  U-As at 9 years  —  —  0.30 (0.17, 0.43)*  0.32 (0.19, 0.45)*  Abbreviations: 8-OHdG, 8-hydroxy-2’-deoxyguanosine; U-As, sum of urinary arsenic metabolites; GW, gestation week. Data were expressed as regression coefficient (β) and 95% confidence intervals (CI). a Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein at 9 years, family socioeconomic status at 9 years of age, and mothers’ education. * indicates p < .05. To investigate whether the estimated effects of arsenic on sjTRECs may be mediated via arsenic-induced oxidative stress, we adjusted the associations of prenatal and childhood U-As with sjTRECs by 8-OHdG in children at 9 years. The estimates of the associations of U-As with sjTRECs did not change after being additionally adjusted for 8-OHdG above or below the spline knot of 5.5 (Table 3). At 9 years, concentration of plasma 8-OHdG above the spline knot was significantly higher (mean 3.6 ng/ml) than below (mean 3.0 ng/ml) the spline knot (p = .007), but this was not the case at 4.5 years of age. DISCUSSION The findings suggest that prenatal and childhood (4.5 and 9 years) arsenic exposures negatively influence TL and production of naïve T cells when U-As is higher than 45 μg/l and the strongest associations were observed with the children’s concurrent exposure. On the other hand, with low level of arsenic exposure (<45 μg/l), increases in TL and sjTRECs concentrations were found in 9-year-old children. Furthermore, arsenic-induced 8-OHdG did not seem to impact the associations of arsenic exposure and sjTRECs levels suggesting nonoverlapping pathways of action. There is mounting interest in studies on environmental exposures and telomere attrition which have mostly been carried out in adults in cross-sectional studies. However, longitudinal birth cohort studies are lacking. Here we found that prenatal and persistent childhood exposure to arsenic above 45 μg/l of U-As reduced TL in children. For every doubling of arsenic exposure, there was a decrease in 34 kb/dg of TL in 9-year-old children. The deleterious impact of arsenic exposure on TL seemed to increase with duration of exposure, with increasingly stronger association being evident at 9 years compared with in utero and 4.5 years of age (Table 2). One study showed that adults with chronic exposure to arsenic (average total U-As ≥19.3 μg/l) have shortened TL in leukocytes in presence of hOGG1 Cys polymorphism indicating that arsenic-mediated telomere shortening was influenced by defects in DNA excision repair (Borghini et al., 2016). However, other epidemiological studies in adults have shown longer TL in PBMC in relation to high arsenic exposure. One study (Chatterjee et al., 2015) reported that people with arsenic-induced skin lesions (mean U-As 290 μg/l) exhibited telomerase-independent elongation of TL compared with subjects with lower exposure (mean U-As 30.5 μg/l). Studies conducted in people chronically exposed to high arsenic showed significantly longer telomeres being associated with higher U-As (median U-As 230 µg/l (Li et al., 2012); 80–196 µg/l (Ameer et al., 2016); mean U-As 856.0 μg/g of creatinine (Gao et al., 2015). In the latter 3 studies, poor arsenic methylation efficiency and elongated TL appeared to play an important role in arsenic-related carcinogenesis. We found poor methylation efficiency associated with shortening of TL among these children. The seemingly counterintuitive findings of shortening or elongation of TL by arsenic exposure could depend on various environmental and other factors (such as arsenic dose, duration of exposure, age, DNA repair mechanisms) that can manipulate the TL maintenance machinery and have opposing directions of effects (Romano et al., 2013; Zhang et al., 2013). For example, median U-As concentrations in our study (88, 57, and 54 µg/l at GW8, 4.5 and 9 years, respectively) were much lower than the above studies. Ours was a birth cohort study and the negative association of arsenic exposure (above U-As of 45 µg/l) with TL remained evident at each time point as opposed to the above cross-sectional, case-control studies in adults. It is plausible that far longer duration of arsenic exposure in adults compared with children influenced the outcome. In this study we found that U-As concentrations below 45 µg/l appeared to be associated with longer telomeres in the children. An ex vivo study demonstrated that treatment with lower concentration (0.0001 μM) of iAs markedly increased TL in cord blood leukocytes but at higher concentration (1 μM) significantly decreased the TL; this occurred in parallel to decreased telomerase expression (Ferrario et al., 2009). Similarly, another study (Zhang et al., 2003) indicated that arsenite at low concentrations (<1 μM) promoted telomerase activity and maintained TL in cell lines, while at high concentrations (>1 μM) there was drastic reduction in TL and increased apoptosis. One in vitro study reported that treatment with arsenic (0.75 µM) inhibited telomerase transcription and resulted in TL shortening and chromosomal end lesions with a dominance of chromosomal end-to-end fusions (Chou et al., 2001). Thus, shorter telomeres may promote genomic instability and initiation of carcinogenesis or reduce cell survival through enhanced apoptosis (Hackett and Greider, 2002). There may be a critical threshold of arsenic exposure beyond which exposed cells either undergo attrition or elongation of TL. However, it is difficult to directly compare the in vitro conditions with far more complex in vivo microenvironment. We have earlier shown in the MINIMat cohort that arsenic exposure reduced thymic size in infants (Raqib et al., 2009) and decreased thymic output in neonates reflected by reduced sjTRECs levels (Ahmed et al., 2012). We extend those findings here by demonstrating that both prenatal and childhood arsenic exposure decreased sjTRECs levels at 9 years of age with progressively stronger association being evident with concurrent exposure. Thymic involution begins from an early age of 1 year and with progression of age a shift occurs from efficient thymic lymphopoiesis to T-cell generation through peripheral replication which becomes the dominant mechanism of replenishing the T-cell pool (Mackall and Gress, 1997). Decrease in sjTREC concentrations reflects immunodeficiency, a well-known phenomenon in clinical conditions including HIV-infection (Douek et al., 2000; Zhang et al., 1999), chemotherapy, bone marrow transplantation, severe respiratory syncytial virus infections in neonates (Gul et al., 2017). Several trials of highly active antiretroviral therapy treatment of HIV/AIDS patients, both adult and pediatric, have demonstrated regeneration of sjTRECs containing T lymphocytes and recovery from immunodeficiency (Sandgaard et al., 2014; Ye et al., 2004). Thus, reduction of sjTRECs due to arsenic exposure suggests depletion of T cell pool in the children eventually leading to immunodeficiency. Our findings are in keeping with previous studies demonstrating reduced frequency of T cells in children and adults with chronic arsenic exposure (Hernandez-Castro et al., 2009; Rocha-Amador et al., 2011; Soto-Pena et al., 2006). Immunosuppressive effects of arsenic are mediated through, among other mechanisms, induction of cell apoptosis (Rocha-Amador et al., 2011). We have previously shown that arsenic exposure during pregnancy suppressed T cells in the placenta and cord blood and upregulated apoptosis related genes in neonatal/cord blood (Ahmed et al., 2011,, 2012). Childhood arsenic exposure reduced T cell-mediated function in these MINIMat children at 4.5 years (Ahmed et al., 2014) and impaired mumps-vaccine specific responses at 9 years of age (Raqib et al., 2017). T cells are required for an effective adaptive immune response. Depletion of T cells is likely to hamper adaptive immunity while aging T lymphocytes (with shorter TL) are hyporesponsive to infection or vaccination (Torrao et al., 2014). Thus, our findings of inadequate production of sjTRECs or naïve T cells and cellular exhaustion due to persistent arsenic exposure in childhood are closely linked to accelerated aging of immune cells (immunosenescence) that may subsequently result in mounting of suboptimal immune responses and increased disease susceptibility. Recent reports suggest that having a short or long TL may be largely established early in life and serve as a marker of susceptibility to chronic diseases and cancer in later life (Benetos et al., 2013; Daniali et al., 2013). In vitro and experimental studies have shown that oxidative stress causes telomere attrition (Liu et al., 2003). In support of our earlier studies (Ahmed et al., 2011, 2012), we found that prenatal and childhood arsenic exposure increased oxidative stress at 9 years, and again oxidative stress appeared to reduce sjTRECs levels in children. However, the associations between U-As and sjTRECs remained unaffected after adjustment with plasma 8-OHdG, suggesting that the mechanisms of 8-OHdG-mediated oxidative damage of naïve T cells may be distinct from arsenic-induced oxidative damage with little overlap. The strengths of our study include the longitudinal design, relatively large sample size, availability of temporal arsenic exposure data (from pregnancy to early and late childhood) and biomarkers for immunosenescence and naïve T cells at multiple time points. One limitation was that we measured TL in undifferentiated blood leukocytes that only reflect total PBMC not specific immune cells such as B and T lymphocytes. Another limitation was that the effect of arsenic on telomerase enzyme that compensates telomere shortening was not assessed. In conclusion, chronic high levels of arsenic exposure from fetal life throughout childhood can result in TL attrition and lower production of naïve T cells that could contribute to immunosenescence and immunodeficiency in later life. The mechanisms of these adverse effects of arsenic and 8-OHdG-induced oxidative damage were nonoverlapping. Environmental exposure to arsenic during early life may result in lifelong changes in health trajectories. SUPPLEMENTARY DATA Supplementary data are available at Toxicological Sciences online. FUNDING This work was supported by the Swedish Research Council, and the Swedish International Development Cooperation Agency (Sida/SAREC Agreement support with icddr,b; grant GR00599; GR00933). ACKNOWLEDGMENTS icddr,b acknowledges with gratitude the commitment of Swedish Research Council and Sida to its research efforts. icddr,b is also grateful to the Governments of Bangladesh, Canada, Sweden and the UK for providing core/unrestricted support. We gratefully acknowledge the participation of the women and their children. ETHICAL APPROVAL The study was approved by the Ethical Review Committee at the icddr,b Dhaka, Bangladesh. Written informed consent was obtained from patients before enrollment into the study that included permission to store specimens for future use. REFERENCES Ahmed S., Ahsan K. B., Kippler M., Mily A., Wagatsuma Y., Hoque A. M., Ngom P. T., El Arifeen S., Raqib R., Vahter M. ( 2012). In utero arsenic exposure is associated with impaired thymic function in newborns possibly via oxidative stress and apoptosis. Toxicol Sci.  129, 305– 314. Google Scholar CrossRef Search ADS PubMed  Ahmed S., Akhtar E., Roy A., von Ehrenstein O. S., Vahter M., Wagatsuma Y., Raqib R. ( 2017). Arsenic exposure alters lung function and airway inflammation in children: A cohort study in rural Bangladesh. Environ. Int.  101, 108– 116. Google Scholar CrossRef Search ADS PubMed  Ahmed S., Mahabbat-e Khoda S., Rekha R. S., Gardner R. M., Ameer S. S., Moore S., Ekström E. C., Vahter M., Raqib R. ( 2011). Arsenic-associated oxidative stress, inflammation, and immune disruption in human placenta and cord blood. Environ. Health Perspect.  119, 258– 264. Google Scholar CrossRef Search ADS PubMed  Ahmed S., Moore S. E., Kippler M., Gardner R., Hawlader M. D. H., Wagatsuma Y., Raqib R., Vahter M. ( 2014). Arsenic exposure and cell-mediated immunity in pre-school children in rural Bangladesh. Toxicol. Sci.  141, 166– 175. Google Scholar CrossRef Search ADS PubMed  Ahmed S., Rekha R. S., Ahsan K. B., Doi M., Grandér M., Roy A. K., Ekström E.-C., Wagatsuma Y., Vahter M., Raqib R.et al.  , ., ( 2013). Arsenic exposure affects plasma insulin-like growth factor 1 (IGF-1) in children in rural Bangladesh. PLoS One  8, e81530. Google Scholar CrossRef Search ADS PubMed  Ameer S. S., Xu Y., Engstrom K., Li H., Tallving P., Nermell B., Boemo A., Parada L. A., Peñaloza L. G., Concha Get al.  , ., ( 2016). Exposure to inorganic arsenic is associated with increased mitochondrial DNA copy number and longer telomere length in peripheral blood. Front. Cell Dev. Biol . 4, 87. Google Scholar CrossRef Search ADS PubMed  Barrett J. H., Iles M. M., Dunning A. M., Pooley K. A. ( 2015). Telomere length and common disease: Study design and analytical challenges. Hum. Genet.  134, 679– 689. Google Scholar CrossRef Search ADS PubMed  Benetos A., Kark J. D., Susser E., Kimura M., Sinnreich R., Chen W., Steenstrup T., Christensen K., Herbig U., von Bornemann Hjelmborg J.et al.  , ., ( 2013). Tracking and fixed ranking of leukocyte telomere length across the adult life course. Aging Cell  12, 615– 621. Google Scholar CrossRef Search ADS PubMed  Blackburn E. H., Epel E. S., Lin J. ( 2015). Human telomere biology: A contributory and interactive factor in aging, disease risks, and protection. Science  350, 1193– 1198. Google Scholar CrossRef Search ADS PubMed  Borghini A., Faita F., Mercuri A., Minichilli F., Bustaffa E., Bianchi F., Andreassi M. G. ( 2016). Arsenic exposure, genetic susceptibility and leukocyte telomere length in an Italian young adult population. Mutagenesis  31, 539– 546. Google Scholar CrossRef Search ADS PubMed  Chatterjee D., Bhattacharjee P., Sau T. J., Das J. K., Sarma N., Bandyopadhyay A. K., Roy S. S., Giri A. K. ( 2015). Arsenic exposure through drinking water leads to senescence and alteration of telomere length in humans: A case-control study in West Bengal, India. Mol. Carcinog.  54, 800– 809. Google Scholar CrossRef Search ADS PubMed  Chou W. C., Hawkins A. L., Barrett J. F., Griffin C. A., Dang C. V. ( 2001). Arsenic inhibition of telomerase transcription leads to genetic instability. J. Clin. Invest.  108, 1541– 1547. Google Scholar CrossRef Search ADS PubMed  Dangleben N. L., Skibola C. F., Smith M. T. ( 2013). Arsenic immunotoxicity: A review. Environ. Health  12, 73. Google Scholar CrossRef Search ADS PubMed  Daniali L., Benetos A., Susser E., Kark J. D., Labat C., Kimura M., Desai K., Granick M., Aviv A. ( 2013). Telomeres shorten at equivalent rates in somatic tissues of adults. Nat. Commun.  4, 1597. Google Scholar CrossRef Search ADS PubMed  de Onis M., Onyango A. W., Borghi E., Siyam A., Nishida C., Siekmann J. ( 2007). Development of a WHO growth reference for school-aged children and adolescents. Bull. World Health Organ.  85, 660– 667. Google Scholar CrossRef Search ADS PubMed  Diaz O. P., Arcos R., Tapia Y., Pastene R., Velez D., Devesa V., Montoro R., Aguilera V., Becerra M. ( 2015). Estimation of arsenic intake from drinking water and food (raw and cooked) in a rural village of northern Chile. Urine as a biomarker of recent exposure. Int J Environ. Res. Public Health  12, 5614– 5633. Google Scholar CrossRef Search ADS PubMed  Douek D. C., Koup R. A., McFarland R. D., Sullivan J. L., Luzuriaga K. ( 2000). Effect of HIV on thymic function before and after antiretroviral therapy in children. J. Infect. Dis.  181, 1479– 1482. Google Scholar CrossRef Search ADS PubMed  Douek D. C., McFarland R. D., Keiser P. H., Gage E. A., Massey J. M., Haynes B. F., Polis M. A., Haase A. T., Feinberg M. B., Sullivan J. L.et al.  , ., ( 1998). Changes in thymic function with age and during the treatment of HIV infection. Nature  396, 690– 695. Google Scholar CrossRef Search ADS PubMed  Farzan S. F., Li Z., Korrick S. A., Spiegelman D., Enelow R., Nadeau K., Baker E., Karagas M. R. ( 2016). Infant infections and respiratory symptoms in relation to in utero arsenic exposure in a U.S. Cohort. Environ. Health Perspect.  124, 840– 847. Google Scholar PubMed  Ferrario D., Collotta A., Carfi M., Bowe G., Vahter M., Hartung T., Gribaldo L. ( 2009). Arsenic induces telomerase expression and maintains telomere length in human cord blood cells. Toxicology  260, 132– 141. Google Scholar CrossRef Search ADS PubMed  Gamboa-Loira B., Cebrian M. E., Franco-Marina F., Lopez-Carrillo L. ( 2017). Arsenic metabolism and cancer risk: A meta-analysis. Environ. Res.  156, 551– 558. Google Scholar CrossRef Search ADS PubMed  Gao J., Roy S., Tong L., Argos M., Jasmine F., Rahaman R., Rakibuz-Zaman M., Parvez F., Ahmed A., Hore S. K.et al.  , ., ( 2015). Arsenic exposure, telomere length, and expression of telomere-related genes among Bangladeshi individuals. Environ. Res.  136, 462– 469. Google Scholar CrossRef Search ADS PubMed  Gianesin K., Noguera-Julian A., Zanchetta M., Del Bianco P., Petrara M. R., Freguja R., Rampon O., Fortuny C., Camós M., Mozzo E.et al.  , ., ( 2016). Premature aging and immune senescence in HIV-infected children. Aids  30, 1363– 1373. Google Scholar CrossRef Search ADS PubMed  Gul K. A., Sonerud T., Fjaerli H. O., Nakstad B., Abrahamsen T. G., Inchley C. S. ( 2017). Thymus activity measured by T-cell receptor excision circles in patients with different severities of respiratory syncytial virus infection. BMC Infect. Dis.  17, 18. Google Scholar CrossRef Search ADS PubMed  Hackett J. A., Greider C. W. ( 2002). Balancing instability: Dual roles for telomerase and telomere dysfunction in tumorigenesis. Oncogene  21, 619– 626. Google Scholar CrossRef Search ADS PubMed  Hernández-Castro B., Doníz-Padilla L. M., Salgado-Bustamante M., Rocha D., Ortiz-Pérez M. D., Jiménez-Capdeville M. E., Portales-Pérez D. P., Quintanar-Stephano A., González-Amaro R. ( 2009). Effect of arsenic on regulatory T cells. J. Clin. Immunol.  29, 461– 469. Google Scholar CrossRef Search ADS PubMed  Kippler M., Skroder H., Rahman S. M., Tofail F., Vahter M. ( 2016). Elevated childhood exposure to arsenic despite reduced drinking water concentrations–A longitudinal cohort study in rural Bangladesh. Environ. Int.  86, 119– 125. Google Scholar CrossRef Search ADS PubMed  Li H., Engstrom K., Vahter M., Broberg K. ( 2012). Arsenic exposure through drinking water is associated with longer telomeres in peripheral blood. Chem. Res. Toxicol.  25, 2333– 2339. Google Scholar CrossRef Search ADS PubMed  Li H., Hedmer M., Wojdacz T., Hossain M. B., Lindh C. H., Tinnerberg H., Albin M., Broberg K. ( 2015). Oxidative stress, telomere shortening, and DNA methylation in relation to low-to-moderate occupational exposure to welding fumes. Environ. Mol. Mutagen.  56, 684– 693. Google Scholar CrossRef Search ADS PubMed  Liu L., Trimarchi J. R., Navarro P., Blasco M. A., Keefe D. L. ( 2003). Oxidative stress contributes to arsenic-induced telomere attrition, chromosome instability, and apoptosis. J. Biol. Chem.  278, 31998– 32004. Google Scholar CrossRef Search ADS PubMed  Mackall C. L., Gress R. E. ( 1997). Thymic aging and T-cell regeneration. Immunol. Rev.  160, 91– 102. Google Scholar CrossRef Search ADS PubMed  Mannan T., Ahmed S., Akhtar E., Roy A. K., Haq M. A., Roy A., Kippler M., Ekström E.-C., Wagatsuma Y., Raqib R.et al.  , ., ( 2016). Maternal micronutrient supplementation and long term health impact in children in Rural Bangladesh. PLoS One  11, e0161294. Google Scholar CrossRef Search ADS PubMed  Moore S. E., Prentice A. M., Wagatsuma Y., Fulford A. J., Collinson A. C., Raqib R., Vahter M., Persson L. A., Arifeen S. E. ( 2009). Early-life nutritional and environmental determinants of thymic size in infants born in rural Bangladesh. Acta Paediatr.  98, 1168– 1175. Google Scholar CrossRef Search ADS PubMed  Nahian M. A. ( 2016): Matlab HDSS Arsenic Screening Survey 2013/14. Seminar presentation, gobeshona monthly seminar. icddr,b. Nermell B., Lindberg A.-L., Rahman M., Berglund M., Åke Persson L., El Arifeen S., Vahter M. ( 2008). Urinary arsenic concentration adjustment factors and malnutrition. Environ. Res.  106, 212– 218. Google Scholar CrossRef Search ADS PubMed  Ngom P. T., Collinson A. C., Pido-Lopez J., Henson S. M., Prentice A. M., Aspinall R. ( 2004). Improved thymic function in exclusively breastfed infants is associated with higher interleukin 7 concentrations in their mothers' breast milk. Am. J. Clin. Nutr.  80, 722– 728. Google Scholar CrossRef Search ADS PubMed  O'Callaghan N. J., Fenech M. ( 2011). A quantitative PCR method for measuring absolute telomere length. Biol. Proced Online  13, 3. Google Scholar CrossRef Search ADS PubMed  Persson L. A., Arifeen S., Ekstrom E. C., Rasmussen K. M., Frongillo E. A., Yunus M. ( 2012). Effects of prenatal micronutrient and early food supplementation on maternal hemoglobin, birth weight, and infant mortality among children in Bangladesh: The MINIMat randomized trial. Jama  307, 2050– 2059. Google Scholar CrossRef Search ADS PubMed  Rahman A., Vahter M., Ekstrom E. C., Persson L. A. ( 2011). Arsenic exposure in pregnancy increases the risk of lower respiratory tract infection and diarrhea during infancy in Bangladesh. Environ. Health Perspect.  119, 719– 724. Google Scholar CrossRef Search ADS PubMed  Rahman M., Vahter M., Wahed M. A., Sohel N., Yunus M., Streatfield P. K., El Arifeen S., Bhuiya A., Zaman K., Chowdhury A. M. ( 2006). Prevalence of arsenic exposure and skin lesions. A population based survey in Matlab, Bangladesh. J Epidemiol. Community Health  60, 242– 248. Google Scholar CrossRef Search ADS PubMed  Raqib R., Ahmed S., Ahsan K. B., Kippler M., Akhtar E., Roy A. K., Lu Y., Arifeen S. E., Wagatsuma Y., Vahter M. ( 2017). Humoral Immunity in Arsenic-Exposed Children in Rural Bangladesh: Total Immunoglobulins and Vaccine-Specific Antibodies. Environ. Health Perspect.  125, 067006. Google Scholar CrossRef Search ADS PubMed  Raqib R., Ahmed S., Sultana R., Wagatsuma Y., Mondal D., Hoque A. M. W., Nermell B., Yunus M., Roy S., Persson L. A.et al.  , ., ( 2009). Effects of in utero arsenic exposure on child immunity and morbidity in rural Bangladesh. Toxicol. Lett.  185, 197– 202. Google Scholar CrossRef Search ADS PubMed  Raqib R., Alam D. S., Sarker P., Ahmad S. M., Ara G., Yunus M., Moore S. E., Fuchs G. ( 2007). Low birth weight is associated with altered immune function in rural Bangladeshi children: A birth cohort study. Am. J. Clin. Nutr.  85, 845– 852. Google Scholar CrossRef Search ADS PubMed  Rocha-Amador D. O., Calderon J., Carrizales L., Costilla-Salazar R., Perez-Maldonado I. N. ( 2011). Apoptosis of peripheral blood mononuclear cells in children exposed to arsenic and fluoride. Environ. Toxicol. Pharmacol.  32, 399– 405. Google Scholar CrossRef Search ADS PubMed  Romano G. H., Harari Y., Yehuda T., Podhorzer A., Rubinstein L., Shamir R., Gottlieb A., Silberberg Y., Pe’er D., Ruppin E.et al.  , ., ( 2013). Environmental stresses disrupt telomere length homeostasis. PLoS Genet.  9, e1003721. Google Scholar CrossRef Search ADS PubMed  Sandgaard K. S., Lewis J., Adams S., Klein N., Callard R. ( 2014). Antiretroviral therapy increases thymic output in children with HIV. Aids  28, 209– 214. Google Scholar CrossRef Search ADS PubMed  Smith A. H., Yunus M., Khan A. F., Ercumen A., Yuan Y., Smith M. H., Liaw J., Balmes J., von Ehrenstein O., Raqib R.et al.  , ., ( 2013). Chronic respiratory symptoms in children following in utero and early life exposure to arsenic in drinking water in Bangladesh. Int. J. Epidemiol.  42, 1077– 1086. Google Scholar CrossRef Search ADS PubMed  Somech R. ( 2011). T-cell receptor excision circles in primary immunodeficiencies and other T- cell immune disorders. Curr. Opin. Allergy Clin. Immunol.  11, 517– 524. Google Scholar CrossRef Search ADS PubMed  Soto-Peña G. A., Luna A. L., Acosta-Saavedra L., Conde P., López-Carrillo L., Cebrián M. E., Bastida M., Calderón-Aranda E. S., Vega L. ( 2006). Assessment of lymphocyte subpopulations and cytokine secretion in children exposed to arsenic. Faseb J.  20, 779– 781. Google Scholar CrossRef Search ADS PubMed  Torrao R. C., Bennett S. J., Brown J. E., Griffiths H. R. ( 2014). Does metabolic reprogramming underpin age-associated changes in T cell phenotype and function? Free Radic. Biol. Med.  71, 26– 35. Google Scholar CrossRef Search ADS PubMed  Vahter M. ( 2002). Mechanisms of arsenic biotransformation. Toxicology  181-182, 211– 217. Google Scholar CrossRef Search ADS PubMed  WHO. ( 2011). Guidelines for Drinking-Water Quality , 4th ed. WHO, Geneva. Yamauchi H., Aminaka Y., Yoshida K., Sun G., Pi J., Waalkes M. P. ( 2004). Evaluation of DNA damage in patients with arsenic poisoning: Urinary 8-hydroxydeoxyguanine. Toxicol. Appl. Pharmacol.  198, 291– 296. Google Scholar CrossRef Search ADS PubMed  Ye P., Kirschner D. E. ( 2002). Reevaluation of T cell receptor excision circles as a measure of human recent thymic emigrants. J Immunol  168, 4968– 4979. Google Scholar CrossRef Search ADS PubMed  Ye P., Kirschner D. E., Kourtis A. P. ( 2004). The thymus during HIV disease: Role in pathogenesis and in immune recovery. Curr. HIV Res.  2, 177– 183. Google Scholar CrossRef Search ADS PubMed  Zhang L., Lewin S. R., Markowitz M., Lin H. H., Skulsky E., Karanicolas R., He Y., Jin X., Tuttleton S., Vesanen Met al.  ., ., ( 1999). Measuring recent thymic emigrants in blood of normal and HIV-1-infected individuals before and after effective therapy. J. Exp. Med.  190, 725– 732. Google Scholar CrossRef Search ADS PubMed  Zhang T. C., Schmitt M. T., Mumford J. L. ( 2003). Effects of arsenic on telomerase and telomeres in relation to cell proliferation and apoptosis in human keratinocytes and leukemia cells in vitro. Carcinogenesis  24, 1811– 1817. Google Scholar CrossRef Search ADS PubMed  Zhang X., Lin S., Funk W. E., Hou L. ( 2013). Environmental and occupational exposure to chemicals and telomere length in human studies. Occup. Environ. Med  70, 743– 749. Google Scholar CrossRef Search ADS PubMed  Zhang X., Zhao Q., Zhu W., Liu T., Xie S. H., Zhong L. X., Cai Y. Y., Li X. N., Liang M., Chen W.et al.  , ., ( 2017). The Association of Telomere Length in Peripheral Blood Cells with Cancer Risk: A Systematic Review and Meta-analysis of Prospective Studies. Cancer Epidemiol. Biomarkers Prevent . 26 9, 1381– 1390. Google Scholar CrossRef Search ADS   Zhou M., Zhu L., Cui X., Feng L., Zhao X., He S., Ping F., Li W., Li Y. ( 2016). Influence of diet on leukocyte telomere length, markers of inflammation and oxidative stress in individuals with varied glucose tolerance: A Chinese population study. Nutr. J.  15, 39. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Toxicological Sciences Oxford University Press

Associations of Arsenic Exposure With Telomere Length and Naïve T Cells in Childhood—A Birth Cohort Study

Loading next page...
 
/lp/ou_press/associations-of-arsenic-exposure-with-telomere-length-and-na-ve-t-dTveSGHjVV
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1096-6080
eISSN
1096-0929
D.O.I.
10.1093/toxsci/kfy105
Publisher site
See Article on Publisher Site

Abstract

Abstract There is limited knowledge of association between arsenic exposure and telomere length (TL) and signal joint T-cell receptor excision circle (sjTREC) that are potential biomarkers of immune senescence and disease susceptibility. We aimed to clarify whether long-term inorganic arsenic exposure influences TL and sjTRECs in childhood. Children born in a longitudinal mother-child cohort were followed-up at 4.5 (n = 275) and 9 years (n = 351) of age. Arsenic exposure was assessed by metabolite concentrations in urine (U-As) from mothers at gestational week 8 (prenatal) and their children at 4.5 and 9 years. TL and sjTRECs were determined in blood cells using quantitative PCR. The oxidative DNA damage marker 8-hydroxy-2ʹ-deoxyguanosine (8-OHdG) in plasma was measured by ELISA. In multivariable-adjusted spline regression analyses, both prenatal and childhood arsenic exposure above U-As of 45 µg/l were significantly inversely associated with TL and sjTRECs at 9 years. Fraction of monomethylarsonic acid (MMA) above spline knot 7% were significantly inversely associated with both TL and sjTRECs reflecting increased toxicity due to less-efficient arsenic metabolism in 9--year-old children. Prenatal and childhood arsenic exposure were positively associated with 8-OHdG at 9 years which in turn was inversely associated with sjTRECs at 9 years. However, adjustment with 8-OHdG did not change the estimates of the association of U-As with sjTRECs reflecting little contribution from 8-OHdG-induced oxidative stress. Our findings suggest that chronic arsenic exposure from early life can result in TL attrition and lower production of naïve T cells potentially leading to immunosenescence and immunodeficiency. telomere length, immunosenescence, sjTRECs, arsenic, birth cohort, school children Chronic exposure to inorganic arsenic (iAs), well-known environmental pollutant and carcinogen has been associated with various forms of cancers and numerous other pathological effects in humans (Gamboa-Loira et al., 2017). Growing evidence suggests that chronic arsenic exposure during pregnancy and early life is associated with increased risk of infectious diseases, in particular respiratory and gastrointestinal infections (Farzan et al., 2016; Rahman et al., 2011; Raqib et al., 2009; Smith et al., 2013). Several cross-sectional studies have indicated that arsenic adversely affects the immune system in both children and adults (Dangleben et al., 2013). However, very few longitudinal studies in human have further indicated that the adverse effects of the toxicant start from in utero (Ahmed et al., 2011, 2012, 2014, 2017; Raqib et al., 2009, 2017). A telomere is a region of nucleotide repeats (TTAGGG) present at each end of the eukaryotic chromosome which is highly regulated to protect chromosomes from recombination or degradation, ensuring integrity during replication. Telomeres shorten after each replication cycle, hence telomere length (TL) is generally considered to reflect cellular senescence (Blackburn et al., 2015). Many genetic and environmental factors can affect TL such as oxidative stress, diet, smoking, infections and different types of cancer and noncommunicable diseases (Barrett et al., 2015; Gianesin et al., 2016; Zhou et al., 2016). Shorter and longer TL both are associated with risk of cancer (Li et al., 2015; Zhang et al., 2017). Limited studies have addressed the association of arsenic with TL of human immune cells (Ameer et al., 2016; Gao et al., 2015; Li et al., 2012). The critical role of the thymus in the generation of a diversified population of peripheral T lymphocytes is well-established. The quantification of signal joint T-cell receptor excision circle (sjTREC) within peripheral T cell populations provides insight into the frequency of recent thymic emigrants or naïve T cells (Ye and Kirschner, 2002). Arsenic is known to produce reactive oxygen species leading to oxidative stress and DNA damage (Yamauchi et al., 2004). This can be assessed by determining 8-hydroxy-2ʹ-deoxyguanosine (8-OHdG) concentration, a known biomarker of oxidative DNA damage. Our earlier studies have shown that arsenic exposure during pregnancy increased 8-OHdG expression in cord blood and placenta, with increased inflammatory responses and reduced T cell counts in the placenta (Ahmed et al., 2011) and reduced sjTRECs in cord blood (Ahmed et al., 2012). Therefore, in the present longitudinal study, we aimed to investigate whether chronic arsenic exposure from in utero to pre-adolescent age influences TL and whether the negative impact of chronic arsenic exposure on sjTRECs at birth persists in 9-year-old children. Additionally, we measured 8-OHdG in plasma to assess its association with arsenic exposure, sjTRECs, and TL. MATERIALS AND METHODS Study design, area, and participants Matlab, a rural area, is situated about 53 km southeast of Dhaka. icddr,b established a health research and training center with a central hospital and 4 smaller subcenters in Matlab and operates a health and demographic surveillance system (HDSS) covering a population of about 220,000. Elevated arsenic concentrations in groundwater are common in Matlab where approximately 95% of the population uses hand-pumped tubewell water as their main source of drinking water. About 13,000 tubewells in this area were tested for arsenic concentration during 2002–2003, about 70% of which exceeded 10 µg/l of arsenic, the WHO guideline value (WHO, 2011), and >60% exceeded 50 μg/l, the national standard (Rahman et al., 2006). Installation of deep tubewells has resulted in somewhat decreased exposure from the time of pregnancy till 9 years but still remained elevated with 43% of tubewells containing arsenic >10 µg/l (Nahian, 2016). However, substantial arsenic exposure through food probably comes from the use of arsenic-rich irrigation water from shallow pumps (Kippler et al., 2016). This study is part of our ongoing research activities on effects of early-life arsenic exposure and immune function outcomes (Ahmed et al., 2012; Moore et al., 2009). It is nested in a large, randomized, food and multiple micronutrient supplementation trial (MINIMat trial) conducted in 2001–2003 in 4436 women recruited in early pregnancy evaluating nutritional impacts on pregnancy outcomes and child health (Persson et al., 2012). The MINIMat children were followed up at 4.5 years of age (n = 2735) to address different objectives. To reduce the burden of examinations and repeated biological sampling from each child, they were divided into 2 groups based on their year of birth (Group A, April 2002–June 2003; Group B, June 2003–June 2004) (Figure 1). For this study, we selected children from Group B, and followed them at 4.5 years (n = 640) and 9 years (n = 564) (Ahmed et al., 2013, 2014, 2017; Raqib et al., 2017; Mannan et al., 2016). Adequate blood cells for studying sjTRECs and TL were available from 275 children at 4.5 years and 351 children at 9 years of age, and among them 213 children were common in both time points (Figure 1). Figure 1. View largeDownload slide Flow diagram depicting the recruitment of children in the current study nested in the MINIMat birth cohort study in rural Bangladesh carried out in 2012–2013. Figure 1. View largeDownload slide Flow diagram depicting the recruitment of children in the current study nested in the MINIMat birth cohort study in rural Bangladesh carried out in 2012–2013. Maternal anthropometry and family socioeconomic status (SES) were collected in early pregnancy in the MINIMat trial (Persson et al., 2012). The family SES score was estimated via an asset index, generated through principal component analysis of household assets and was updated during the follow-up studies. Anthropometry data of children were collected at 4.5 and 9 years (Raqib et al., 2017). The measured height and weight were converted to height-for-age, weight-for-age, and body mass index-for-age z-scores (SD scores), using the WHO growth reference for school-aged children and adolescents (de Onis et al., 2007). Data on illness including cold, cough, fever, diarrhea, and dysentery in the past 6 months were collected based on structured questionnaire. Comparison of baseline features between the children included in this study and those who were not included did not show any significant differences in general characteristics (Supplementary Table 1). Specimen collection Heparinized venous blood was divided into 2 parts, 1 was used to determine specific cell types in blood as complete blood count with differentials using automated hematology analyzer (Sysmex XT-1800i, Kobe, Japan). The other part was centrifuged to separate plasma from buffy coat. Peripheral blood mononuclear cells (PBMCs) was separated from buffy coat by density-gradient centrifugation and stored in RNALater (Qiagen GmbH, Hilden, Germany) in −80°C. Urine samples collected as described earlier (Ahmed et al., 2017) from pregnant mothers during gestation week 8 (GW8) and their children at 4.5 and 9 years of age were stored at −80°C. Assessment of arsenic exposure iAs is metabolized by a series of reduction and methylation reactions producing monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA), the relative fractions of which reflect efficiency of arsenic metabolism. Increased toxicity of arsenic was shown with the excretion of higher MMA% and a lower DMA% in the urine (Vahter, 2002). Arsenic exposure was assessed based on the concentration of the sum of iAs, MMA, and DMA in urine, hereinafter referred to as urinary arsenic (U-As) reflecting exposure to iAs from both water and food (Diaz et al., 2015; Kippler et al., 2016). The arsenic metabolite concentrations in urine were measured using high-performance liquid chromatography online with hydride generation and inductively coupled plasma mass spectrometry. The intra- and interassay coefficients of variation (CV%) were approximately 4%, based on the measurements of a reference urine sample (CRM No.18, National Institute for Environmental Studies, Tsukuba City, Japan (Ahmed et al., 2014). All U-As concentrations were adjusted to the average specific gravity (1.012 g/ml for all urine samples), measured by a digital refractometer (RD712 Clinical Refractometer; EUROMEX, Arnhem, the Netherlands) (Nermell et al., 2008) using the formula U-As (1.012-1)/(measured specific gravity of each urine sample-1). Estimation of sjTRECs sjTRECs are small circles of DNA created in T-lymphocyte during maturation and rearrangement of T-cell receptor genes in the thymus. sjTRECs contents in blood are used as surrogates of recent thymic emigrant T cells (RTEs) and is a well-known biomarker for thymic output of RTEs or thymic function (Douek et al., 1998; Somech, 2011). From the frozen PBMC, DNA was isolated by using the QIAamp DNA Mini Kit (Qiagen GmbH) according to the manufacturer’s instructions. From 200 ng of DNA/sample, quantification of sjTREC was performed by using SYBR Green real-time quantitative PCR and CFX96TM real time system (C1000 Thermal cycler, Bio-Rad Life Science Research, Hercules, California) as described earlier in Ahmed et al. (2012) and Raqib et al. (2007). The following primers were used: forward primer 5ʹAAAGAGGGCAGCCCTCTCCAAGGCAAA3ʹ and reverse primer 5ʹAGGCT-GATCTTGTCTGACATTTGCTCCG3ʹ. A standard (kindly donated by PT Ngom) was prepared by using serial dilutions of a known number of copies of a fragment of the sjTREC gene sequence and included in each run to generate a standard curve (Ngom et al., 2004). The copy number of sjTRECs in the DNA samples was determined automatically using standard curves and expressed as sjTRECs content (copy numbers/100 ng of DNA). The CV% of intra- and interrun variation of determination of sjTREC level were 8.5% and 10.2% respectively. Estimation of TL Absolute TL in PBMC DNA was measured as described earlier (O’Callaghan and Fenech, 2011) by performing 2 RT-PCR reactions using 70 ng of DNA samples for each reaction in a single plate using 2 different oligomer standards. Telomere oligomer standard was used to establish telomere standard curve in telomere PCR reaction whereas single copy gene 36B4 oligomer standard was used for creating 36B4 standard curve in 36B4 PCR reaction. The primer sequences were: telomere forward primer 5’CGGTTTGTTTGGGTTTGGGTTT-GGGTTTGGG-TTTGGGTT3’; and reverse primer 5’GGCTTGCCTTACCCTTACCCTTACCC-TTACCC-TTACCCT3’; 36B4 forward primer 5’CAGCAAGTGGGAA-GGTGTAATCC3’ and reverse primer 5’CCCATTCTATCATC-AACGGGTACAA3’. Plasmid DNA (pBR322) was used with each standard to maintain a constant 20 ng of total DNA per reaction tube. The data obtained from RT-PCR method to measure absolute TL were analyzed as kb/reaction and genome copies/reaction for telomere and 36B4, respectively. The telomere kb per reaction value is divided by diploid genome copy number of 36B4 to give a total TL in kb per human diploid genome. The CT values within the linear range of the standard curves and between 15 and 35 threshold cycles were used for analysis. The CV% of intra- and interrun variation of TL assessment were <5% and <7.5%, respectively. Oxidative stress marker Plasma concentration of 8-OHdG, a biomarker of oxidative DNA damage was analyzed by a competitive ELISA kit (Highly Sensitive 8-OHdG Check ELISA, Japan Institute for the Control of Aging, Fukuroi, Shizuoka, Japan). Samples were measured in duplicates and analysis was repeated if results of duplicate samples differed >10%. The intra- and interassay CV% were 2.85% and 4.6%, respectively. Statistical analysis Statistical analyses were conducted using the software PASW 20.0 (SPSS Inc., Chicago, Illinois) and Stata/IC 13.0 (StataCorp, College Station, Texas). Associations between exposures, outcomes, and covariates were initially evaluated using Spearman’s rank correlation, Mann-Whitney U-test, analysis of variance, or Kruskal-Wallis test, as appropriate. Associations between different exposure biomarkers (U-As at 4.5 and 9 years; MMA%, DMA%, iAs%) and outcomes (sjTRECs and TL) were examined graphically using lowess moving average curves searching for linear or non-linear associations. When linear associations were obtained, multivariable-adjusted linear regression analyses were carried out. When deviation from linearity was found, we applied splines at apparent thresholds. In scatter plots with Lowess curves (Figs. 2A–F), both TL and sjTRECs slightly increased up to 45 µg/l of U-As and then started to decline showing deviation from a linear pattern. Thus, due to the nature of non-linear curve the associations of U-As at all-time points (GW8, 4.5 and 9 years of age) with TL and sjTRECs were assessed by multivariable-adjusted spline regression analyses with a spline knot introduced at 45 µg/l (log2-transformed: equivalent to 5.5). To obtain normally distributed residuals with a homogeneous variance, all exposure variables (U-As at GW8, 4.5 and 9 years of age) were log2-transformed. We chose log2-transformation of TL and sjTRECs values to simplify the interpretation of the beta-coefficients in the regression analyses (average changes in outcome associated with each doubling of exposure). To analyze the influence of arsenic metabolism efficiency on the associations between U-As and sjTRECs or TL, initially we analyzed by multivariable-adjusted linear regression. Again, we observed nonlinear association between % of arsenic metabolites and TL or sjTRECs, thus spline regression was applied. The spline knot values of the metabolites were introduced at 7, 80 and 15 for MMA%, DMA%, iAs%, respectively (Supplementary Figure 2). Linear mixed effects models were applied to evaluate the effect of exposure over time (4.5 and 9 years of age) on outcome changes (TL and sjTRECs) with a spline knot at 45 µg/l. The association between U-As and 8-OHdG was evaluated by linear regression analysis. The models were adjusted for covariates that were significantly associated with both exposure and outcome or changed the effect estimates by 5% or more. The covariates were child sex, age, HAZ, SES, mother’s education and plasma CRP. The p values < .05 were considered statistically significant. Figure 2. View largeDownload slide Associations between U-As concentration (log2-transformed) at GW8, 4.5 and 9 years and TL (A–C) and sjTRECs (log2-transformed) (D–F) in 9-year-old children. In the scatter plots, the solid line represents a Lowess (locally weighted scatter plot smoothing) moving-average curve for the raw data. Figure 2. View largeDownload slide Associations between U-As concentration (log2-transformed) at GW8, 4.5 and 9 years and TL (A–C) and sjTRECs (log2-transformed) (D–F) in 9-year-old children. In the scatter plots, the solid line represents a Lowess (locally weighted scatter plot smoothing) moving-average curve for the raw data. RESULTS Demographic Data and Arsenic Exposure The median age of the children at the different follow-ups were 4.6 years (range: 4.5–5.2) and 8.8 years (range: 8.6–9.6), respectively (Table 1) and were referred to as 4.5 and 9 years of age. At each follow-up, girls (49%) were shorter (p for all ≤ .003) and weighed less (p for all ≤ .049) than the boys (51%). Median U-As in the children at 9 years was lower than that of their mothers during pregnancy (median 88, range 1.9–1576 µg/l) and at 4.5 years of age indicating gradual decline in arsenic exposure over time. However, the mother’s U-As concentrations were strongly correlated with children’s exposures (rs = 0.48 for 4.5 years and rs = 0.40 for 9 years, both p < .001). Family SES did not change significantly over time from GW8 to 9 years of age. The median TL at 9 years decreased significantly from that at 4.5 years, though there was a strong correlation (Supplementary Figure 1). The sjTRECs content did change significantly over time. Mean plasma concentration of 8-OHdG was significantly higher in 9-years old compared with 4.5-years old (p < .001) indicating accumulating oxidative stress among older children (Table 1). Table 1. General Characteristics of the Study Participants Variablesa  4.5 years (n = 213)  9 years (n = 351)  p-Value  Age, months  55.9 ± 1.4  104.4 ± 1.2  <.001  Gender, male, n (%)  99 (46.5%)  178 (50.7%)    Weight (kg)  13.7 ±1.5  22.0± 3.1  <.001  Height (cm)  100.0 ±4.0  123.0 ± 5.2  <.001  Stuntedb  61 (28.6%)  76 (21.7%)  .912  Underweightb  97 (45.5%)  147 (41.9%)  .452  U-As (µg/l)c  57.1 (12.9, 1125.0)  53.9 (8.9, 1268.0)  .007  TL (kb/dg)  188.9 (68.6–464.0)  164.8 (53.3–455.0)  <.001  sjTRECs (copy number/100 ng DNA)  4.8 (0.23–684.0)×104  5.7 (0.18–84.5)×104  .910  8-OHdG (ng/ml)  2.5 ± 1.4  3.3 ± 1.7  <.001  Family SESd         First tertile, n (%)  71 (33.3%)  117 (33.3%)     second tertile, n (%)  65 (30.5%)  117 (33.3%)  .910   Third tertile, n (%)  77 (36.2%)  117 (33.3%)    Variablesa  4.5 years (n = 213)  9 years (n = 351)  p-Value  Age, months  55.9 ± 1.4  104.4 ± 1.2  <.001  Gender, male, n (%)  99 (46.5%)  178 (50.7%)    Weight (kg)  13.7 ±1.5  22.0± 3.1  <.001  Height (cm)  100.0 ±4.0  123.0 ± 5.2  <.001  Stuntedb  61 (28.6%)  76 (21.7%)  .912  Underweightb  97 (45.5%)  147 (41.9%)  .452  U-As (µg/l)c  57.1 (12.9, 1125.0)  53.9 (8.9, 1268.0)  .007  TL (kb/dg)  188.9 (68.6–464.0)  164.8 (53.3–455.0)  <.001  sjTRECs (copy number/100 ng DNA)  4.8 (0.23–684.0)×104  5.7 (0.18–84.5)×104  .910  8-OHdG (ng/ml)  2.5 ± 1.4  3.3 ± 1.7  <.001  Family SESd         First tertile, n (%)  71 (33.3%)  117 (33.3%)     second tertile, n (%)  65 (30.5%)  117 (33.3%)  .910   Third tertile, n (%)  77 (36.2%)  117 (33.3%)    Abbreviations: U-As, Urinary arsenic; kb/dg, kilo base pair/diploid genome; sjTRECs, signal-joint T cell receptor excision circles; 8-OHdG, 8-hydroxy-2'-deoxyguanosine; SES, Socioeconomic status. a Values are given either as mean ± SD or median with range within brackets. b Defined as children with HAZ, height for age or WAZ, weight for age<−2 SDs from the median value of height or weight for age of reference population (according to WHO). c Adjusted to average specific gravity of 1.012. d SES score was estimated via an asset index, generated through principal component analysis of household assets. Table 1. General Characteristics of the Study Participants Variablesa  4.5 years (n = 213)  9 years (n = 351)  p-Value  Age, months  55.9 ± 1.4  104.4 ± 1.2  <.001  Gender, male, n (%)  99 (46.5%)  178 (50.7%)    Weight (kg)  13.7 ±1.5  22.0± 3.1  <.001  Height (cm)  100.0 ±4.0  123.0 ± 5.2  <.001  Stuntedb  61 (28.6%)  76 (21.7%)  .912  Underweightb  97 (45.5%)  147 (41.9%)  .452  U-As (µg/l)c  57.1 (12.9, 1125.0)  53.9 (8.9, 1268.0)  .007  TL (kb/dg)  188.9 (68.6–464.0)  164.8 (53.3–455.0)  <.001  sjTRECs (copy number/100 ng DNA)  4.8 (0.23–684.0)×104  5.7 (0.18–84.5)×104  .910  8-OHdG (ng/ml)  2.5 ± 1.4  3.3 ± 1.7  <.001  Family SESd         First tertile, n (%)  71 (33.3%)  117 (33.3%)     second tertile, n (%)  65 (30.5%)  117 (33.3%)  .910   Third tertile, n (%)  77 (36.2%)  117 (33.3%)    Variablesa  4.5 years (n = 213)  9 years (n = 351)  p-Value  Age, months  55.9 ± 1.4  104.4 ± 1.2  <.001  Gender, male, n (%)  99 (46.5%)  178 (50.7%)    Weight (kg)  13.7 ±1.5  22.0± 3.1  <.001  Height (cm)  100.0 ±4.0  123.0 ± 5.2  <.001  Stuntedb  61 (28.6%)  76 (21.7%)  .912  Underweightb  97 (45.5%)  147 (41.9%)  .452  U-As (µg/l)c  57.1 (12.9, 1125.0)  53.9 (8.9, 1268.0)  .007  TL (kb/dg)  188.9 (68.6–464.0)  164.8 (53.3–455.0)  <.001  sjTRECs (copy number/100 ng DNA)  4.8 (0.23–684.0)×104  5.7 (0.18–84.5)×104  .910  8-OHdG (ng/ml)  2.5 ± 1.4  3.3 ± 1.7  <.001  Family SESd         First tertile, n (%)  71 (33.3%)  117 (33.3%)     second tertile, n (%)  65 (30.5%)  117 (33.3%)  .910   Third tertile, n (%)  77 (36.2%)  117 (33.3%)    Abbreviations: U-As, Urinary arsenic; kb/dg, kilo base pair/diploid genome; sjTRECs, signal-joint T cell receptor excision circles; 8-OHdG, 8-hydroxy-2'-deoxyguanosine; SES, Socioeconomic status. a Values are given either as mean ± SD or median with range within brackets. b Defined as children with HAZ, height for age or WAZ, weight for age<−2 SDs from the median value of height or weight for age of reference population (according to WHO). c Adjusted to average specific gravity of 1.012. d SES score was estimated via an asset index, generated through principal component analysis of household assets. Chronic Arsenic Exposure and TL Adjusted regression models showed significant inverse associations between maternal (GW8) and childhood U-As (4.5 and 9 years of age) and TL at 9 years of age above the spline knot 5.5 (Table 2). In contrast, concurrent U-As below the spline knot was significantly positively associated with TL at 9 years of age. In 4.5-year-old children, U-As at GW8 and concurrent U-As above the spline knot were inversely associated with TL, although the associations were not statistically significant. These associations remained unchanged when total available children (n = 275) at 4.5 years of age were considered (Figure 1, Supplementary Table 2). The estimates of the associations between U-As at GW8 with TL at 4.5 (β = −9.7) and 9 years (β = −10.4) were similar (Table 2). When association between U-As at 9 years and TL were further adjusted with blood cell types, we did not find any significant influence of the various cell types (data not shown). Table 2. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years and Arsenic Metabolites at 9 Years With TL at 4.5 and 9 Years of Age TL  Children at 4.5 Years (n = 213)  Children at 9 Years (n = 351)  β (95% CI)a  β (95% CI)a  U-As at GW8       <5.5b  7.1 (–26.9, 41.0)  10.4 (–8.4, 29.1)   >5.5  –9.7 (–19.4, 0.05)*  –10.4 (–16.9, –3.9)**  U-As at 4.5 years       <5.5b  –8.7 (–42.8, 25.4)  –3.3 (–25.7, 19.2)   >5.5  –4.7 (–16.0, 6.6)  –15.6 (–23.6, –7.6)**  U-As at 9 years       <5.5b  —  23.2 (5.9, 41.6)**   >5.5  —  –33.9 (–41.9, –25.9) **  MMA in urine (%)       <7.0b  0.01 (–0.08, 0.10)  0.02 (–0.05, 0.08)   >7.0  –0.02 (–0.05, 0.01)  –0.05 (–0.07, –0.02)**  DMA in urine (%)       <80.0b  0.01 (–0.02, 0.03)  0.02 (–0.004, 0.04)   >80.0  0.003 (–0.02, 0.03)  0.02 (–0.01, 0.04)  iAs in urine (%)       <15.0b  0.02 (–0.01, 0.04)  –0.001 (–0.02, 0.02)   >15.0  –0.07 (–0.13, –0.003)**  –0.03 (–0.09, 0.02)  TL  Children at 4.5 Years (n = 213)  Children at 9 Years (n = 351)  β (95% CI)a  β (95% CI)a  U-As at GW8       <5.5b  7.1 (–26.9, 41.0)  10.4 (–8.4, 29.1)   >5.5  –9.7 (–19.4, 0.05)*  –10.4 (–16.9, –3.9)**  U-As at 4.5 years       <5.5b  –8.7 (–42.8, 25.4)  –3.3 (–25.7, 19.2)   >5.5  –4.7 (–16.0, 6.6)  –15.6 (–23.6, –7.6)**  U-As at 9 years       <5.5b  —  23.2 (5.9, 41.6)**   >5.5  —  –33.9 (–41.9, –25.9) **  MMA in urine (%)       <7.0b  0.01 (–0.08, 0.10)  0.02 (–0.05, 0.08)   >7.0  –0.02 (–0.05, 0.01)  –0.05 (–0.07, –0.02)**  DMA in urine (%)       <80.0b  0.01 (–0.02, 0.03)  0.02 (–0.004, 0.04)   >80.0  0.003 (–0.02, 0.03)  0.02 (–0.01, 0.04)  iAs in urine (%)       <15.0b  0.02 (–0.01, 0.04)  –0.001 (–0.02, 0.02)   >15.0  –0.07 (–0.13, –0.003)**  –0.03 (–0.09, 0.02)  Abbreviations: TL, Telomere length; β, regression coefficient; CI, confidence interval; U-As, sum of urinary arsenic metabolites; GW, gestation week; MMA, monomethylarsonic acid; DMA, dimethylarsinic acid; iAs, inorganic arsenic. a Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein, and total leukocytes at 9 years, family socioeconomic status at 9 years, mothers’ education. b Spline regression model using spline knot at log2 U-As 5.5 (corresponding to 45 µg/l), fraction of MMA at 7%, fraction of DMA at 80% and fraction of iAs at 15%. * indicates p < .10 and **indicates p < .05. Table 2. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years and Arsenic Metabolites at 9 Years With TL at 4.5 and 9 Years of Age TL  Children at 4.5 Years (n = 213)  Children at 9 Years (n = 351)  β (95% CI)a  β (95% CI)a  U-As at GW8       <5.5b  7.1 (–26.9, 41.0)  10.4 (–8.4, 29.1)   >5.5  –9.7 (–19.4, 0.05)*  –10.4 (–16.9, –3.9)**  U-As at 4.5 years       <5.5b  –8.7 (–42.8, 25.4)  –3.3 (–25.7, 19.2)   >5.5  –4.7 (–16.0, 6.6)  –15.6 (–23.6, –7.6)**  U-As at 9 years       <5.5b  —  23.2 (5.9, 41.6)**   >5.5  —  –33.9 (–41.9, –25.9) **  MMA in urine (%)       <7.0b  0.01 (–0.08, 0.10)  0.02 (–0.05, 0.08)   >7.0  –0.02 (–0.05, 0.01)  –0.05 (–0.07, –0.02)**  DMA in urine (%)       <80.0b  0.01 (–0.02, 0.03)  0.02 (–0.004, 0.04)   >80.0  0.003 (–0.02, 0.03)  0.02 (–0.01, 0.04)  iAs in urine (%)       <15.0b  0.02 (–0.01, 0.04)  –0.001 (–0.02, 0.02)   >15.0  –0.07 (–0.13, –0.003)**  –0.03 (–0.09, 0.02)  TL  Children at 4.5 Years (n = 213)  Children at 9 Years (n = 351)  β (95% CI)a  β (95% CI)a  U-As at GW8       <5.5b  7.1 (–26.9, 41.0)  10.4 (–8.4, 29.1)   >5.5  –9.7 (–19.4, 0.05)*  –10.4 (–16.9, –3.9)**  U-As at 4.5 years       <5.5b  –8.7 (–42.8, 25.4)  –3.3 (–25.7, 19.2)   >5.5  –4.7 (–16.0, 6.6)  –15.6 (–23.6, –7.6)**  U-As at 9 years       <5.5b  —  23.2 (5.9, 41.6)**   >5.5  —  –33.9 (–41.9, –25.9) **  MMA in urine (%)       <7.0b  0.01 (–0.08, 0.10)  0.02 (–0.05, 0.08)   >7.0  –0.02 (–0.05, 0.01)  –0.05 (–0.07, –0.02)**  DMA in urine (%)       <80.0b  0.01 (–0.02, 0.03)  0.02 (–0.004, 0.04)   >80.0  0.003 (–0.02, 0.03)  0.02 (–0.01, 0.04)  iAs in urine (%)       <15.0b  0.02 (–0.01, 0.04)  –0.001 (–0.02, 0.02)   >15.0  –0.07 (–0.13, –0.003)**  –0.03 (–0.09, 0.02)  Abbreviations: TL, Telomere length; β, regression coefficient; CI, confidence interval; U-As, sum of urinary arsenic metabolites; GW, gestation week; MMA, monomethylarsonic acid; DMA, dimethylarsinic acid; iAs, inorganic arsenic. a Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein, and total leukocytes at 9 years, family socioeconomic status at 9 years, mothers’ education. b Spline regression model using spline knot at log2 U-As 5.5 (corresponding to 45 µg/l), fraction of MMA at 7%, fraction of DMA at 80% and fraction of iAs at 15%. * indicates p < .10 and **indicates p < .05. To evaluate long-term changes of arsenic on TL, we evaluated the association of childhood exposure to arsenic in relation to TL at 4.5 and 9 years of age by mixed effect models. A significant inverse association was found between childhood arsenic exposure above the spline knot and childhood TL (β = −15.4, 95% CI = −22.19, −8.64; p < .001). The TL was significantly inversely associated with the fraction of MMA above the spline knot of 7% in 9 years old and above 15% for iAs in 4.5-year-old children (Table 2) likely reflecting increased toxicity due to poor methylation of arsenic in these children. Chronic Arsenic Exposure, sjTRECs, and Immune Cells Significant inverse associations were found between U-As at GW8, 4.5 and 9 years of age and sjTRECs at 9 years of age above the spline knot 5.5 (Table 3). Only in 9-year-old children a significant positive association of sjTRECs was obtained with U-As below the spline knot. To examine whether elevated levels of sjTRECs were linked with illness in these children, we evaluated association between sjTREC concentrations and morbidity outcomes and found no association (data not shown). Again, when association between U-As and sjTRECs in 9-years old were further adjusted with cell types, we did not find any significant influence of the various cell types (data not shown). Table 3. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years and Arsenic Metabolites at 9 Years With sjTRECs at 4.5 and 9 Years of Age sjTRECs  4.5-Year-Old Children (n = 213)  9-Year-Old Children (n = 349)   Model 1 β (95% CI)a  Model 1 β (95% CI)a  Model 2 β (95% CI)b  U-As at GW8   <5.5c  0.72 (–0.02, 1.5)  0.29 (–0.23, 0.81)  0.32 (–0.20, 0.85)   >5.5  –0.12 (–0.33, 0.09)  –0.43 (–0.61, –0.25)*  –0.42 (–0.61, –0.24)*  U-As at 4.5 years   <5.5c  0.36 (–0.39, 1.1)  0.03 (–0.58, 0.65)  0.09 (–0.55, 0.73)   >5.5  –0.11 (–0.35, 0.14)  –0.62 (–0.84, –0.40)*  –0.59 (–0.82, –0.36)*  U-As at 9 years   <5.5c  —  1.3 (0.87, 1.8)*  1.3 (0.80, 1.8)*   >5.5  —  –1.3 (–1.5, –1.1)*  –1.3 (–1.5, –1.1)*  MMA in urine (%)   <7.0c  –0.03 (–0.32, 0.26)  0.19 (–0.04, 0.42)  0.04 (–0.08, 0.16)   >7.0  –0.05 (–0.14, 0.04)  –0.16 (–0.24, –0.08)*  –0.26 (–0.38, –0.14)*  DMA in urine (%)   <80.0c  –0.03 (–0.11, 0.06)  0.03 (–0.04, 0.11)  0.05 (–0.03, 0.14)   >80.0  0.04 (–0.04, 0.13)  0.03 (–0.04, 0.10)  0.02 (–0.05, 0.10)  iAs in urine (%)   <15.0c  –0.02 (–0.10, 0.07)  0.01 (–0.07, 0.08)  –0.01 (–0.09, 0.07)   >15.0  0.15 (–0.06, 0.36)  0.04 (–0.16, 0.23)  0.14 (–0.10, 0.38)  sjTRECs  4.5-Year-Old Children (n = 213)  9-Year-Old Children (n = 349)   Model 1 β (95% CI)a  Model 1 β (95% CI)a  Model 2 β (95% CI)b  U-As at GW8   <5.5c  0.72 (–0.02, 1.5)  0.29 (–0.23, 0.81)  0.32 (–0.20, 0.85)   >5.5  –0.12 (–0.33, 0.09)  –0.43 (–0.61, –0.25)*  –0.42 (–0.61, –0.24)*  U-As at 4.5 years   <5.5c  0.36 (–0.39, 1.1)  0.03 (–0.58, 0.65)  0.09 (–0.55, 0.73)   >5.5  –0.11 (–0.35, 0.14)  –0.62 (–0.84, –0.40)*  –0.59 (–0.82, –0.36)*  U-As at 9 years   <5.5c  —  1.3 (0.87, 1.8)*  1.3 (0.80, 1.8)*   >5.5  —  –1.3 (–1.5, –1.1)*  –1.3 (–1.5, –1.1)*  MMA in urine (%)   <7.0c  –0.03 (–0.32, 0.26)  0.19 (–0.04, 0.42)  0.04 (–0.08, 0.16)   >7.0  –0.05 (–0.14, 0.04)  –0.16 (–0.24, –0.08)*  –0.26 (–0.38, –0.14)*  DMA in urine (%)   <80.0c  –0.03 (–0.11, 0.06)  0.03 (–0.04, 0.11)  0.05 (–0.03, 0.14)   >80.0  0.04 (–0.04, 0.13)  0.03 (–0.04, 0.10)  0.02 (–0.05, 0.10)  iAs in urine (%)   <15.0c  –0.02 (–0.10, 0.07)  0.01 (–0.07, 0.08)  –0.01 (–0.09, 0.07)   >15.0  0.15 (–0.06, 0.36)  0.04 (–0.16, 0.23)  0.14 (–0.10, 0.38)  Abbreviations: sjTRECs, signal-joint T cell receptor excision circles; β, regression coefficient; CI, confidence interval; U-As, sum of urinary arsenic metabolites; GW, gestation week; MMA, monomethylarsonic acid; DMA, dimethylarsinic acid; iAs, inorganic arsenic. sjTRECs were log2-transformed. a Model 1. Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein and lymphocytes at 9 years, socioeconomic status of the family at 9 years of age, mothers’ education. b Model 2. Variables additionally adjusted by 8-OHdG at 9 years. c Spline regression model using spline knot at log2 U-As 5.5 (corresponding to 45 µg/l), fraction of MMA at 7%, fraction of DMA at 80% and fraction of iAs at 15%. * indicates p < .05. Table 3. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years and Arsenic Metabolites at 9 Years With sjTRECs at 4.5 and 9 Years of Age sjTRECs  4.5-Year-Old Children (n = 213)  9-Year-Old Children (n = 349)   Model 1 β (95% CI)a  Model 1 β (95% CI)a  Model 2 β (95% CI)b  U-As at GW8   <5.5c  0.72 (–0.02, 1.5)  0.29 (–0.23, 0.81)  0.32 (–0.20, 0.85)   >5.5  –0.12 (–0.33, 0.09)  –0.43 (–0.61, –0.25)*  –0.42 (–0.61, –0.24)*  U-As at 4.5 years   <5.5c  0.36 (–0.39, 1.1)  0.03 (–0.58, 0.65)  0.09 (–0.55, 0.73)   >5.5  –0.11 (–0.35, 0.14)  –0.62 (–0.84, –0.40)*  –0.59 (–0.82, –0.36)*  U-As at 9 years   <5.5c  —  1.3 (0.87, 1.8)*  1.3 (0.80, 1.8)*   >5.5  —  –1.3 (–1.5, –1.1)*  –1.3 (–1.5, –1.1)*  MMA in urine (%)   <7.0c  –0.03 (–0.32, 0.26)  0.19 (–0.04, 0.42)  0.04 (–0.08, 0.16)   >7.0  –0.05 (–0.14, 0.04)  –0.16 (–0.24, –0.08)*  –0.26 (–0.38, –0.14)*  DMA in urine (%)   <80.0c  –0.03 (–0.11, 0.06)  0.03 (–0.04, 0.11)  0.05 (–0.03, 0.14)   >80.0  0.04 (–0.04, 0.13)  0.03 (–0.04, 0.10)  0.02 (–0.05, 0.10)  iAs in urine (%)   <15.0c  –0.02 (–0.10, 0.07)  0.01 (–0.07, 0.08)  –0.01 (–0.09, 0.07)   >15.0  0.15 (–0.06, 0.36)  0.04 (–0.16, 0.23)  0.14 (–0.10, 0.38)  sjTRECs  4.5-Year-Old Children (n = 213)  9-Year-Old Children (n = 349)   Model 1 β (95% CI)a  Model 1 β (95% CI)a  Model 2 β (95% CI)b  U-As at GW8   <5.5c  0.72 (–0.02, 1.5)  0.29 (–0.23, 0.81)  0.32 (–0.20, 0.85)   >5.5  –0.12 (–0.33, 0.09)  –0.43 (–0.61, –0.25)*  –0.42 (–0.61, –0.24)*  U-As at 4.5 years   <5.5c  0.36 (–0.39, 1.1)  0.03 (–0.58, 0.65)  0.09 (–0.55, 0.73)   >5.5  –0.11 (–0.35, 0.14)  –0.62 (–0.84, –0.40)*  –0.59 (–0.82, –0.36)*  U-As at 9 years   <5.5c  —  1.3 (0.87, 1.8)*  1.3 (0.80, 1.8)*   >5.5  —  –1.3 (–1.5, –1.1)*  –1.3 (–1.5, –1.1)*  MMA in urine (%)   <7.0c  –0.03 (–0.32, 0.26)  0.19 (–0.04, 0.42)  0.04 (–0.08, 0.16)   >7.0  –0.05 (–0.14, 0.04)  –0.16 (–0.24, –0.08)*  –0.26 (–0.38, –0.14)*  DMA in urine (%)   <80.0c  –0.03 (–0.11, 0.06)  0.03 (–0.04, 0.11)  0.05 (–0.03, 0.14)   >80.0  0.04 (–0.04, 0.13)  0.03 (–0.04, 0.10)  0.02 (–0.05, 0.10)  iAs in urine (%)   <15.0c  –0.02 (–0.10, 0.07)  0.01 (–0.07, 0.08)  –0.01 (–0.09, 0.07)   >15.0  0.15 (–0.06, 0.36)  0.04 (–0.16, 0.23)  0.14 (–0.10, 0.38)  Abbreviations: sjTRECs, signal-joint T cell receptor excision circles; β, regression coefficient; CI, confidence interval; U-As, sum of urinary arsenic metabolites; GW, gestation week; MMA, monomethylarsonic acid; DMA, dimethylarsinic acid; iAs, inorganic arsenic. sjTRECs were log2-transformed. a Model 1. Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein and lymphocytes at 9 years, socioeconomic status of the family at 9 years of age, mothers’ education. b Model 2. Variables additionally adjusted by 8-OHdG at 9 years. c Spline regression model using spline knot at log2 U-As 5.5 (corresponding to 45 µg/l), fraction of MMA at 7%, fraction of DMA at 80% and fraction of iAs at 15%. * indicates p < .05. The mixed effect model with spline function showed an inverse association between childhood arsenic exposure and sjTRECs (β = −0.68, 95% CI = −0.85, −0.49; p < .001) above the spline knot at 5.5. However, a positive association was found between childhood arsenic exposure and sjTRECs in all children (β = 0.70, 95% CI = 0.29, 1.10; p < .001) below the spline knot. Further adjustment with U-As at GW8 did not change the estimates. Significant negative correlation of MMA% with sjTRECs levels was found above spline knot of 7% at 9 years of age. No other significant association of sjTRECs was found with other arsenic metabolites at either time points (Table 3). Arsenic Exposure, Oxidative Stress, TL, and sjTRECs Prenatal and childhood U-As were significantly positively associated with plasma 8-OHdG at 9 years of age, but not at 4.5 years of age (Table 4). Again, linear regression analysis showed a significant inverse association between 8-OHdG and sjTRECs at 9 years of age (β = −0.17; 95% CI, −0.30, −0.04; p = .01), but not at 4.5 years of age (β = −0.15; 95% CI, −0.37, 0.08; p = .195). However, we did not observe similar associations between TL and 8-OHdG either at 4.5 or 9 years of age. Table 4. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years With 8-OHdG at 4.5 and 9 Years of Age   8-OHdG at 4.5 Years (n = 213)   8-OHdG at 9 Years (n = 335)     Unadjusted  Adjusteda  Unadjusted  Adjusteda  U-As at GW8  –0.06 (–0.20, 0.07)  –0.06 (–0.19, 0.08)  0.13 (0.02, 0.24)*  0.13 (0.01, 0.24)*  U-As at 4.5 years  –0.07 (–0.24, 0.09)  –0.08 (–0.24, 0.09)  0.23 (0.09, 0.36)*  0.22 (0.08, 0.36)*  U-As at 9 years  —  —  0.30 (0.17, 0.43)*  0.32 (0.19, 0.45)*    8-OHdG at 4.5 Years (n = 213)   8-OHdG at 9 Years (n = 335)     Unadjusted  Adjusteda  Unadjusted  Adjusteda  U-As at GW8  –0.06 (–0.20, 0.07)  –0.06 (–0.19, 0.08)  0.13 (0.02, 0.24)*  0.13 (0.01, 0.24)*  U-As at 4.5 years  –0.07 (–0.24, 0.09)  –0.08 (–0.24, 0.09)  0.23 (0.09, 0.36)*  0.22 (0.08, 0.36)*  U-As at 9 years  —  —  0.30 (0.17, 0.43)*  0.32 (0.19, 0.45)*  Abbreviations: 8-OHdG, 8-hydroxy-2’-deoxyguanosine; U-As, sum of urinary arsenic metabolites; GW, gestation week. Data were expressed as regression coefficient (β) and 95% confidence intervals (CI). a Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein at 9 years, family socioeconomic status at 9 years of age, and mothers’ education. * indicates p < .05. Table 4. Regression Analysis of U-As Concentration at GW 8, 4.5, and 9 Years With 8-OHdG at 4.5 and 9 Years of Age   8-OHdG at 4.5 Years (n = 213)   8-OHdG at 9 Years (n = 335)     Unadjusted  Adjusteda  Unadjusted  Adjusteda  U-As at GW8  –0.06 (–0.20, 0.07)  –0.06 (–0.19, 0.08)  0.13 (0.02, 0.24)*  0.13 (0.01, 0.24)*  U-As at 4.5 years  –0.07 (–0.24, 0.09)  –0.08 (–0.24, 0.09)  0.23 (0.09, 0.36)*  0.22 (0.08, 0.36)*  U-As at 9 years  —  —  0.30 (0.17, 0.43)*  0.32 (0.19, 0.45)*    8-OHdG at 4.5 Years (n = 213)   8-OHdG at 9 Years (n = 335)     Unadjusted  Adjusteda  Unadjusted  Adjusteda  U-As at GW8  –0.06 (–0.20, 0.07)  –0.06 (–0.19, 0.08)  0.13 (0.02, 0.24)*  0.13 (0.01, 0.24)*  U-As at 4.5 years  –0.07 (–0.24, 0.09)  –0.08 (–0.24, 0.09)  0.23 (0.09, 0.36)*  0.22 (0.08, 0.36)*  U-As at 9 years  —  —  0.30 (0.17, 0.43)*  0.32 (0.19, 0.45)*  Abbreviations: 8-OHdG, 8-hydroxy-2’-deoxyguanosine; U-As, sum of urinary arsenic metabolites; GW, gestation week. Data were expressed as regression coefficient (β) and 95% confidence intervals (CI). a Adjusted for child age, height-for-age z-score, gender, plasma C-reactive protein at 9 years, family socioeconomic status at 9 years of age, and mothers’ education. * indicates p < .05. To investigate whether the estimated effects of arsenic on sjTRECs may be mediated via arsenic-induced oxidative stress, we adjusted the associations of prenatal and childhood U-As with sjTRECs by 8-OHdG in children at 9 years. The estimates of the associations of U-As with sjTRECs did not change after being additionally adjusted for 8-OHdG above or below the spline knot of 5.5 (Table 3). At 9 years, concentration of plasma 8-OHdG above the spline knot was significantly higher (mean 3.6 ng/ml) than below (mean 3.0 ng/ml) the spline knot (p = .007), but this was not the case at 4.5 years of age. DISCUSSION The findings suggest that prenatal and childhood (4.5 and 9 years) arsenic exposures negatively influence TL and production of naïve T cells when U-As is higher than 45 μg/l and the strongest associations were observed with the children’s concurrent exposure. On the other hand, with low level of arsenic exposure (<45 μg/l), increases in TL and sjTRECs concentrations were found in 9-year-old children. Furthermore, arsenic-induced 8-OHdG did not seem to impact the associations of arsenic exposure and sjTRECs levels suggesting nonoverlapping pathways of action. There is mounting interest in studies on environmental exposures and telomere attrition which have mostly been carried out in adults in cross-sectional studies. However, longitudinal birth cohort studies are lacking. Here we found that prenatal and persistent childhood exposure to arsenic above 45 μg/l of U-As reduced TL in children. For every doubling of arsenic exposure, there was a decrease in 34 kb/dg of TL in 9-year-old children. The deleterious impact of arsenic exposure on TL seemed to increase with duration of exposure, with increasingly stronger association being evident at 9 years compared with in utero and 4.5 years of age (Table 2). One study showed that adults with chronic exposure to arsenic (average total U-As ≥19.3 μg/l) have shortened TL in leukocytes in presence of hOGG1 Cys polymorphism indicating that arsenic-mediated telomere shortening was influenced by defects in DNA excision repair (Borghini et al., 2016). However, other epidemiological studies in adults have shown longer TL in PBMC in relation to high arsenic exposure. One study (Chatterjee et al., 2015) reported that people with arsenic-induced skin lesions (mean U-As 290 μg/l) exhibited telomerase-independent elongation of TL compared with subjects with lower exposure (mean U-As 30.5 μg/l). Studies conducted in people chronically exposed to high arsenic showed significantly longer telomeres being associated with higher U-As (median U-As 230 µg/l (Li et al., 2012); 80–196 µg/l (Ameer et al., 2016); mean U-As 856.0 μg/g of creatinine (Gao et al., 2015). In the latter 3 studies, poor arsenic methylation efficiency and elongated TL appeared to play an important role in arsenic-related carcinogenesis. We found poor methylation efficiency associated with shortening of TL among these children. The seemingly counterintuitive findings of shortening or elongation of TL by arsenic exposure could depend on various environmental and other factors (such as arsenic dose, duration of exposure, age, DNA repair mechanisms) that can manipulate the TL maintenance machinery and have opposing directions of effects (Romano et al., 2013; Zhang et al., 2013). For example, median U-As concentrations in our study (88, 57, and 54 µg/l at GW8, 4.5 and 9 years, respectively) were much lower than the above studies. Ours was a birth cohort study and the negative association of arsenic exposure (above U-As of 45 µg/l) with TL remained evident at each time point as opposed to the above cross-sectional, case-control studies in adults. It is plausible that far longer duration of arsenic exposure in adults compared with children influenced the outcome. In this study we found that U-As concentrations below 45 µg/l appeared to be associated with longer telomeres in the children. An ex vivo study demonstrated that treatment with lower concentration (0.0001 μM) of iAs markedly increased TL in cord blood leukocytes but at higher concentration (1 μM) significantly decreased the TL; this occurred in parallel to decreased telomerase expression (Ferrario et al., 2009). Similarly, another study (Zhang et al., 2003) indicated that arsenite at low concentrations (<1 μM) promoted telomerase activity and maintained TL in cell lines, while at high concentrations (>1 μM) there was drastic reduction in TL and increased apoptosis. One in vitro study reported that treatment with arsenic (0.75 µM) inhibited telomerase transcription and resulted in TL shortening and chromosomal end lesions with a dominance of chromosomal end-to-end fusions (Chou et al., 2001). Thus, shorter telomeres may promote genomic instability and initiation of carcinogenesis or reduce cell survival through enhanced apoptosis (Hackett and Greider, 2002). There may be a critical threshold of arsenic exposure beyond which exposed cells either undergo attrition or elongation of TL. However, it is difficult to directly compare the in vitro conditions with far more complex in vivo microenvironment. We have earlier shown in the MINIMat cohort that arsenic exposure reduced thymic size in infants (Raqib et al., 2009) and decreased thymic output in neonates reflected by reduced sjTRECs levels (Ahmed et al., 2012). We extend those findings here by demonstrating that both prenatal and childhood arsenic exposure decreased sjTRECs levels at 9 years of age with progressively stronger association being evident with concurrent exposure. Thymic involution begins from an early age of 1 year and with progression of age a shift occurs from efficient thymic lymphopoiesis to T-cell generation through peripheral replication which becomes the dominant mechanism of replenishing the T-cell pool (Mackall and Gress, 1997). Decrease in sjTREC concentrations reflects immunodeficiency, a well-known phenomenon in clinical conditions including HIV-infection (Douek et al., 2000; Zhang et al., 1999), chemotherapy, bone marrow transplantation, severe respiratory syncytial virus infections in neonates (Gul et al., 2017). Several trials of highly active antiretroviral therapy treatment of HIV/AIDS patients, both adult and pediatric, have demonstrated regeneration of sjTRECs containing T lymphocytes and recovery from immunodeficiency (Sandgaard et al., 2014; Ye et al., 2004). Thus, reduction of sjTRECs due to arsenic exposure suggests depletion of T cell pool in the children eventually leading to immunodeficiency. Our findings are in keeping with previous studies demonstrating reduced frequency of T cells in children and adults with chronic arsenic exposure (Hernandez-Castro et al., 2009; Rocha-Amador et al., 2011; Soto-Pena et al., 2006). Immunosuppressive effects of arsenic are mediated through, among other mechanisms, induction of cell apoptosis (Rocha-Amador et al., 2011). We have previously shown that arsenic exposure during pregnancy suppressed T cells in the placenta and cord blood and upregulated apoptosis related genes in neonatal/cord blood (Ahmed et al., 2011,, 2012). Childhood arsenic exposure reduced T cell-mediated function in these MINIMat children at 4.5 years (Ahmed et al., 2014) and impaired mumps-vaccine specific responses at 9 years of age (Raqib et al., 2017). T cells are required for an effective adaptive immune response. Depletion of T cells is likely to hamper adaptive immunity while aging T lymphocytes (with shorter TL) are hyporesponsive to infection or vaccination (Torrao et al., 2014). Thus, our findings of inadequate production of sjTRECs or naïve T cells and cellular exhaustion due to persistent arsenic exposure in childhood are closely linked to accelerated aging of immune cells (immunosenescence) that may subsequently result in mounting of suboptimal immune responses and increased disease susceptibility. Recent reports suggest that having a short or long TL may be largely established early in life and serve as a marker of susceptibility to chronic diseases and cancer in later life (Benetos et al., 2013; Daniali et al., 2013). In vitro and experimental studies have shown that oxidative stress causes telomere attrition (Liu et al., 2003). In support of our earlier studies (Ahmed et al., 2011, 2012), we found that prenatal and childhood arsenic exposure increased oxidative stress at 9 years, and again oxidative stress appeared to reduce sjTRECs levels in children. However, the associations between U-As and sjTRECs remained unaffected after adjustment with plasma 8-OHdG, suggesting that the mechanisms of 8-OHdG-mediated oxidative damage of naïve T cells may be distinct from arsenic-induced oxidative damage with little overlap. The strengths of our study include the longitudinal design, relatively large sample size, availability of temporal arsenic exposure data (from pregnancy to early and late childhood) and biomarkers for immunosenescence and naïve T cells at multiple time points. One limitation was that we measured TL in undifferentiated blood leukocytes that only reflect total PBMC not specific immune cells such as B and T lymphocytes. Another limitation was that the effect of arsenic on telomerase enzyme that compensates telomere shortening was not assessed. In conclusion, chronic high levels of arsenic exposure from fetal life throughout childhood can result in TL attrition and lower production of naïve T cells that could contribute to immunosenescence and immunodeficiency in later life. The mechanisms of these adverse effects of arsenic and 8-OHdG-induced oxidative damage were nonoverlapping. Environmental exposure to arsenic during early life may result in lifelong changes in health trajectories. SUPPLEMENTARY DATA Supplementary data are available at Toxicological Sciences online. FUNDING This work was supported by the Swedish Research Council, and the Swedish International Development Cooperation Agency (Sida/SAREC Agreement support with icddr,b; grant GR00599; GR00933). ACKNOWLEDGMENTS icddr,b acknowledges with gratitude the commitment of Swedish Research Council and Sida to its research efforts. icddr,b is also grateful to the Governments of Bangladesh, Canada, Sweden and the UK for providing core/unrestricted support. We gratefully acknowledge the participation of the women and their children. ETHICAL APPROVAL The study was approved by the Ethical Review Committee at the icddr,b Dhaka, Bangladesh. Written informed consent was obtained from patients before enrollment into the study that included permission to store specimens for future use. REFERENCES Ahmed S., Ahsan K. B., Kippler M., Mily A., Wagatsuma Y., Hoque A. M., Ngom P. T., El Arifeen S., Raqib R., Vahter M. ( 2012). In utero arsenic exposure is associated with impaired thymic function in newborns possibly via oxidative stress and apoptosis. Toxicol Sci.  129, 305– 314. Google Scholar CrossRef Search ADS PubMed  Ahmed S., Akhtar E., Roy A., von Ehrenstein O. S., Vahter M., Wagatsuma Y., Raqib R. ( 2017). Arsenic exposure alters lung function and airway inflammation in children: A cohort study in rural Bangladesh. Environ. Int.  101, 108– 116. Google Scholar CrossRef Search ADS PubMed  Ahmed S., Mahabbat-e Khoda S., Rekha R. S., Gardner R. M., Ameer S. S., Moore S., Ekström E. C., Vahter M., Raqib R. ( 2011). Arsenic-associated oxidative stress, inflammation, and immune disruption in human placenta and cord blood. Environ. Health Perspect.  119, 258– 264. Google Scholar CrossRef Search ADS PubMed  Ahmed S., Moore S. E., Kippler M., Gardner R., Hawlader M. D. H., Wagatsuma Y., Raqib R., Vahter M. ( 2014). Arsenic exposure and cell-mediated immunity in pre-school children in rural Bangladesh. Toxicol. Sci.  141, 166– 175. Google Scholar CrossRef Search ADS PubMed  Ahmed S., Rekha R. S., Ahsan K. B., Doi M., Grandér M., Roy A. K., Ekström E.-C., Wagatsuma Y., Vahter M., Raqib R.et al.  , ., ( 2013). Arsenic exposure affects plasma insulin-like growth factor 1 (IGF-1) in children in rural Bangladesh. PLoS One  8, e81530. Google Scholar CrossRef Search ADS PubMed  Ameer S. S., Xu Y., Engstrom K., Li H., Tallving P., Nermell B., Boemo A., Parada L. A., Peñaloza L. G., Concha Get al.  , ., ( 2016). Exposure to inorganic arsenic is associated with increased mitochondrial DNA copy number and longer telomere length in peripheral blood. Front. Cell Dev. Biol . 4, 87. Google Scholar CrossRef Search ADS PubMed  Barrett J. H., Iles M. M., Dunning A. M., Pooley K. A. ( 2015). Telomere length and common disease: Study design and analytical challenges. Hum. Genet.  134, 679– 689. Google Scholar CrossRef Search ADS PubMed  Benetos A., Kark J. D., Susser E., Kimura M., Sinnreich R., Chen W., Steenstrup T., Christensen K., Herbig U., von Bornemann Hjelmborg J.et al.  , ., ( 2013). Tracking and fixed ranking of leukocyte telomere length across the adult life course. Aging Cell  12, 615– 621. Google Scholar CrossRef Search ADS PubMed  Blackburn E. H., Epel E. S., Lin J. ( 2015). Human telomere biology: A contributory and interactive factor in aging, disease risks, and protection. Science  350, 1193– 1198. Google Scholar CrossRef Search ADS PubMed  Borghini A., Faita F., Mercuri A., Minichilli F., Bustaffa E., Bianchi F., Andreassi M. G. ( 2016). Arsenic exposure, genetic susceptibility and leukocyte telomere length in an Italian young adult population. Mutagenesis  31, 539– 546. Google Scholar CrossRef Search ADS PubMed  Chatterjee D., Bhattacharjee P., Sau T. J., Das J. K., Sarma N., Bandyopadhyay A. K., Roy S. S., Giri A. K. ( 2015). Arsenic exposure through drinking water leads to senescence and alteration of telomere length in humans: A case-control study in West Bengal, India. Mol. Carcinog.  54, 800– 809. Google Scholar CrossRef Search ADS PubMed  Chou W. C., Hawkins A. L., Barrett J. F., Griffin C. A., Dang C. V. ( 2001). Arsenic inhibition of telomerase transcription leads to genetic instability. J. Clin. Invest.  108, 1541– 1547. Google Scholar CrossRef Search ADS PubMed  Dangleben N. L., Skibola C. F., Smith M. T. ( 2013). Arsenic immunotoxicity: A review. Environ. Health  12, 73. Google Scholar CrossRef Search ADS PubMed  Daniali L., Benetos A., Susser E., Kark J. D., Labat C., Kimura M., Desai K., Granick M., Aviv A. ( 2013). Telomeres shorten at equivalent rates in somatic tissues of adults. Nat. Commun.  4, 1597. Google Scholar CrossRef Search ADS PubMed  de Onis M., Onyango A. W., Borghi E., Siyam A., Nishida C., Siekmann J. ( 2007). Development of a WHO growth reference for school-aged children and adolescents. Bull. World Health Organ.  85, 660– 667. Google Scholar CrossRef Search ADS PubMed  Diaz O. P., Arcos R., Tapia Y., Pastene R., Velez D., Devesa V., Montoro R., Aguilera V., Becerra M. ( 2015). Estimation of arsenic intake from drinking water and food (raw and cooked) in a rural village of northern Chile. Urine as a biomarker of recent exposure. Int J Environ. Res. Public Health  12, 5614– 5633. Google Scholar CrossRef Search ADS PubMed  Douek D. C., Koup R. A., McFarland R. D., Sullivan J. L., Luzuriaga K. ( 2000). Effect of HIV on thymic function before and after antiretroviral therapy in children. J. Infect. Dis.  181, 1479– 1482. Google Scholar CrossRef Search ADS PubMed  Douek D. C., McFarland R. D., Keiser P. H., Gage E. A., Massey J. M., Haynes B. F., Polis M. A., Haase A. T., Feinberg M. B., Sullivan J. L.et al.  , ., ( 1998). Changes in thymic function with age and during the treatment of HIV infection. Nature  396, 690– 695. Google Scholar CrossRef Search ADS PubMed  Farzan S. F., Li Z., Korrick S. A., Spiegelman D., Enelow R., Nadeau K., Baker E., Karagas M. R. ( 2016). Infant infections and respiratory symptoms in relation to in utero arsenic exposure in a U.S. Cohort. Environ. Health Perspect.  124, 840– 847. Google Scholar PubMed  Ferrario D., Collotta A., Carfi M., Bowe G., Vahter M., Hartung T., Gribaldo L. ( 2009). Arsenic induces telomerase expression and maintains telomere length in human cord blood cells. Toxicology  260, 132– 141. Google Scholar CrossRef Search ADS PubMed  Gamboa-Loira B., Cebrian M. E., Franco-Marina F., Lopez-Carrillo L. ( 2017). Arsenic metabolism and cancer risk: A meta-analysis. Environ. Res.  156, 551– 558. Google Scholar CrossRef Search ADS PubMed  Gao J., Roy S., Tong L., Argos M., Jasmine F., Rahaman R., Rakibuz-Zaman M., Parvez F., Ahmed A., Hore S. K.et al.  , ., ( 2015). Arsenic exposure, telomere length, and expression of telomere-related genes among Bangladeshi individuals. Environ. Res.  136, 462– 469. Google Scholar CrossRef Search ADS PubMed  Gianesin K., Noguera-Julian A., Zanchetta M., Del Bianco P., Petrara M. R., Freguja R., Rampon O., Fortuny C., Camós M., Mozzo E.et al.  , ., ( 2016). Premature aging and immune senescence in HIV-infected children. Aids  30, 1363– 1373. Google Scholar CrossRef Search ADS PubMed  Gul K. A., Sonerud T., Fjaerli H. O., Nakstad B., Abrahamsen T. G., Inchley C. S. ( 2017). Thymus activity measured by T-cell receptor excision circles in patients with different severities of respiratory syncytial virus infection. BMC Infect. Dis.  17, 18. Google Scholar CrossRef Search ADS PubMed  Hackett J. A., Greider C. W. ( 2002). Balancing instability: Dual roles for telomerase and telomere dysfunction in tumorigenesis. Oncogene  21, 619– 626. Google Scholar CrossRef Search ADS PubMed  Hernández-Castro B., Doníz-Padilla L. M., Salgado-Bustamante M., Rocha D., Ortiz-Pérez M. D., Jiménez-Capdeville M. E., Portales-Pérez D. P., Quintanar-Stephano A., González-Amaro R. ( 2009). Effect of arsenic on regulatory T cells. J. Clin. Immunol.  29, 461– 469. Google Scholar CrossRef Search ADS PubMed  Kippler M., Skroder H., Rahman S. M., Tofail F., Vahter M. ( 2016). Elevated childhood exposure to arsenic despite reduced drinking water concentrations–A longitudinal cohort study in rural Bangladesh. Environ. Int.  86, 119– 125. Google Scholar CrossRef Search ADS PubMed  Li H., Engstrom K., Vahter M., Broberg K. ( 2012). Arsenic exposure through drinking water is associated with longer telomeres in peripheral blood. Chem. Res. Toxicol.  25, 2333– 2339. Google Scholar CrossRef Search ADS PubMed  Li H., Hedmer M., Wojdacz T., Hossain M. B., Lindh C. H., Tinnerberg H., Albin M., Broberg K. ( 2015). Oxidative stress, telomere shortening, and DNA methylation in relation to low-to-moderate occupational exposure to welding fumes. Environ. Mol. Mutagen.  56, 684– 693. Google Scholar CrossRef Search ADS PubMed  Liu L., Trimarchi J. R., Navarro P., Blasco M. A., Keefe D. L. ( 2003). Oxidative stress contributes to arsenic-induced telomere attrition, chromosome instability, and apoptosis. J. Biol. Chem.  278, 31998– 32004. Google Scholar CrossRef Search ADS PubMed  Mackall C. L., Gress R. E. ( 1997). Thymic aging and T-cell regeneration. Immunol. Rev.  160, 91– 102. Google Scholar CrossRef Search ADS PubMed  Mannan T., Ahmed S., Akhtar E., Roy A. K., Haq M. A., Roy A., Kippler M., Ekström E.-C., Wagatsuma Y., Raqib R.et al.  , ., ( 2016). Maternal micronutrient supplementation and long term health impact in children in Rural Bangladesh. PLoS One  11, e0161294. Google Scholar CrossRef Search ADS PubMed  Moore S. E., Prentice A. M., Wagatsuma Y., Fulford A. J., Collinson A. C., Raqib R., Vahter M., Persson L. A., Arifeen S. E. ( 2009). Early-life nutritional and environmental determinants of thymic size in infants born in rural Bangladesh. Acta Paediatr.  98, 1168– 1175. Google Scholar CrossRef Search ADS PubMed  Nahian M. A. ( 2016): Matlab HDSS Arsenic Screening Survey 2013/14. Seminar presentation, gobeshona monthly seminar. icddr,b. Nermell B., Lindberg A.-L., Rahman M., Berglund M., Åke Persson L., El Arifeen S., Vahter M. ( 2008). Urinary arsenic concentration adjustment factors and malnutrition. Environ. Res.  106, 212– 218. Google Scholar CrossRef Search ADS PubMed  Ngom P. T., Collinson A. C., Pido-Lopez J., Henson S. M., Prentice A. M., Aspinall R. ( 2004). Improved thymic function in exclusively breastfed infants is associated with higher interleukin 7 concentrations in their mothers' breast milk. Am. J. Clin. Nutr.  80, 722– 728. Google Scholar CrossRef Search ADS PubMed  O'Callaghan N. J., Fenech M. ( 2011). A quantitative PCR method for measuring absolute telomere length. Biol. Proced Online  13, 3. Google Scholar CrossRef Search ADS PubMed  Persson L. A., Arifeen S., Ekstrom E. C., Rasmussen K. M., Frongillo E. A., Yunus M. ( 2012). Effects of prenatal micronutrient and early food supplementation on maternal hemoglobin, birth weight, and infant mortality among children in Bangladesh: The MINIMat randomized trial. Jama  307, 2050– 2059. Google Scholar CrossRef Search ADS PubMed  Rahman A., Vahter M., Ekstrom E. C., Persson L. A. ( 2011). Arsenic exposure in pregnancy increases the risk of lower respiratory tract infection and diarrhea during infancy in Bangladesh. Environ. Health Perspect.  119, 719– 724. Google Scholar CrossRef Search ADS PubMed  Rahman M., Vahter M., Wahed M. A., Sohel N., Yunus M., Streatfield P. K., El Arifeen S., Bhuiya A., Zaman K., Chowdhury A. M. ( 2006). Prevalence of arsenic exposure and skin lesions. A population based survey in Matlab, Bangladesh. J Epidemiol. Community Health  60, 242– 248. Google Scholar CrossRef Search ADS PubMed  Raqib R., Ahmed S., Ahsan K. B., Kippler M., Akhtar E., Roy A. K., Lu Y., Arifeen S. E., Wagatsuma Y., Vahter M. ( 2017). Humoral Immunity in Arsenic-Exposed Children in Rural Bangladesh: Total Immunoglobulins and Vaccine-Specific Antibodies. Environ. Health Perspect.  125, 067006. Google Scholar CrossRef Search ADS PubMed  Raqib R., Ahmed S., Sultana R., Wagatsuma Y., Mondal D., Hoque A. M. W., Nermell B., Yunus M., Roy S., Persson L. A.et al.  , ., ( 2009). Effects of in utero arsenic exposure on child immunity and morbidity in rural Bangladesh. Toxicol. Lett.  185, 197– 202. Google Scholar CrossRef Search ADS PubMed  Raqib R., Alam D. S., Sarker P., Ahmad S. M., Ara G., Yunus M., Moore S. E., Fuchs G. ( 2007). Low birth weight is associated with altered immune function in rural Bangladeshi children: A birth cohort study. Am. J. Clin. Nutr.  85, 845– 852. Google Scholar CrossRef Search ADS PubMed  Rocha-Amador D. O., Calderon J., Carrizales L., Costilla-Salazar R., Perez-Maldonado I. N. ( 2011). Apoptosis of peripheral blood mononuclear cells in children exposed to arsenic and fluoride. Environ. Toxicol. Pharmacol.  32, 399– 405. Google Scholar CrossRef Search ADS PubMed  Romano G. H., Harari Y., Yehuda T., Podhorzer A., Rubinstein L., Shamir R., Gottlieb A., Silberberg Y., Pe’er D., Ruppin E.et al.  , ., ( 2013). Environmental stresses disrupt telomere length homeostasis. PLoS Genet.  9, e1003721. Google Scholar CrossRef Search ADS PubMed  Sandgaard K. S., Lewis J., Adams S., Klein N., Callard R. ( 2014). Antiretroviral therapy increases thymic output in children with HIV. Aids  28, 209– 214. Google Scholar CrossRef Search ADS PubMed  Smith A. H., Yunus M., Khan A. F., Ercumen A., Yuan Y., Smith M. H., Liaw J., Balmes J., von Ehrenstein O., Raqib R.et al.  , ., ( 2013). Chronic respiratory symptoms in children following in utero and early life exposure to arsenic in drinking water in Bangladesh. Int. J. Epidemiol.  42, 1077– 1086. Google Scholar CrossRef Search ADS PubMed  Somech R. ( 2011). T-cell receptor excision circles in primary immunodeficiencies and other T- cell immune disorders. Curr. Opin. Allergy Clin. Immunol.  11, 517– 524. Google Scholar CrossRef Search ADS PubMed  Soto-Peña G. A., Luna A. L., Acosta-Saavedra L., Conde P., López-Carrillo L., Cebrián M. E., Bastida M., Calderón-Aranda E. S., Vega L. ( 2006). Assessment of lymphocyte subpopulations and cytokine secretion in children exposed to arsenic. Faseb J.  20, 779– 781. Google Scholar CrossRef Search ADS PubMed  Torrao R. C., Bennett S. J., Brown J. E., Griffiths H. R. ( 2014). Does metabolic reprogramming underpin age-associated changes in T cell phenotype and function? Free Radic. Biol. Med.  71, 26– 35. Google Scholar CrossRef Search ADS PubMed  Vahter M. ( 2002). Mechanisms of arsenic biotransformation. Toxicology  181-182, 211– 217. Google Scholar CrossRef Search ADS PubMed  WHO. ( 2011). Guidelines for Drinking-Water Quality , 4th ed. WHO, Geneva. Yamauchi H., Aminaka Y., Yoshida K., Sun G., Pi J., Waalkes M. P. ( 2004). Evaluation of DNA damage in patients with arsenic poisoning: Urinary 8-hydroxydeoxyguanine. Toxicol. Appl. Pharmacol.  198, 291– 296. Google Scholar CrossRef Search ADS PubMed  Ye P., Kirschner D. E. ( 2002). Reevaluation of T cell receptor excision circles as a measure of human recent thymic emigrants. J Immunol  168, 4968– 4979. Google Scholar CrossRef Search ADS PubMed  Ye P., Kirschner D. E., Kourtis A. P. ( 2004). The thymus during HIV disease: Role in pathogenesis and in immune recovery. Curr. HIV Res.  2, 177– 183. Google Scholar CrossRef Search ADS PubMed  Zhang L., Lewin S. R., Markowitz M., Lin H. H., Skulsky E., Karanicolas R., He Y., Jin X., Tuttleton S., Vesanen Met al.  ., ., ( 1999). Measuring recent thymic emigrants in blood of normal and HIV-1-infected individuals before and after effective therapy. J. Exp. Med.  190, 725– 732. Google Scholar CrossRef Search ADS PubMed  Zhang T. C., Schmitt M. T., Mumford J. L. ( 2003). Effects of arsenic on telomerase and telomeres in relation to cell proliferation and apoptosis in human keratinocytes and leukemia cells in vitro. Carcinogenesis  24, 1811– 1817. Google Scholar CrossRef Search ADS PubMed  Zhang X., Lin S., Funk W. E., Hou L. ( 2013). Environmental and occupational exposure to chemicals and telomere length in human studies. Occup. Environ. Med  70, 743– 749. Google Scholar CrossRef Search ADS PubMed  Zhang X., Zhao Q., Zhu W., Liu T., Xie S. H., Zhong L. X., Cai Y. Y., Li X. N., Liang M., Chen W.et al.  , ., ( 2017). The Association of Telomere Length in Peripheral Blood Cells with Cancer Risk: A Systematic Review and Meta-analysis of Prospective Studies. Cancer Epidemiol. Biomarkers Prevent . 26 9, 1381– 1390. Google Scholar CrossRef Search ADS   Zhou M., Zhu L., Cui X., Feng L., Zhao X., He S., Ping F., Li W., Li Y. ( 2016). Influence of diet on leukocyte telomere length, markers of inflammation and oxidative stress in individuals with varied glucose tolerance: A Chinese population study. Nutr. J.  15, 39. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Toxicological SciencesOxford University Press

Published: May 10, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off