Skip Navigation


Carcinogenesis Advance Access originally published online on March 10, 2008
Carcinogenesis 2008 29(5):977-983; doi:10.1093/carcin/bgn065
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
29/5/977    most recent
bgn065v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by van Leeuwen, D. M.
Right arrow Articles by Kleinjans, J. C.S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by van Leeuwen, D. M.
Right arrow Articles by Kleinjans, J. C.S.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Genomic analysis suggests higher susceptibility of children to air pollution

Danitsja M. van Leeuwen1,*, Marie Pedersen2, Peter J.M. Hendriksen1,3, André Boorsma1,3, Marcel H.M. van Herwijnen1, Ralph W.H. Gottschalk1, Micheline Kirsch-Volders4, Lisbeth E. Knudsen2, Radim J. Srám5, Edyta Bajak6,7, Joost H.M. van Delft1 and Jos C.S. Kleinjans1

1 Department of Health Risk Analysis and Toxicology, Maastricht University, PO Box 616, 6200 MD Maastricht, The Netherlands
2 Institute of Public Health, Department of Environmental and Occupational Health, University of Copenhagen, Øster Farimagsgade 5A, 1014 K Copenhagen, Denmark
3 Business Unit BioSciences, TNO Quality of Life, Physiological Genomics, PO Box 360, 3700 AJ Zeist, The Netherlands
4 Faculty of Science, Laboratory of Cell Genetics, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussel, Belgium
5 Institute of Experimental Medicine AS CR and Health Institute of Central Bohemia, Vídenská 1083, 142 20 Prague 4, Czech Republic
6 Department of Environmental Medicine, Karolinska Institutet, Box 20, 171 77 Stockholm, Sweden
7 Department of Pharmacology and Toxicology, KU Medical Center, Mail Stop 1023, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA

* To whom correspondence should be addressed. Tel: +31 43 3882127; Fax: +31 43 3884146; Email: d.vanleeuwen{at}grat.unimaas.nl


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Funding
 References
 
Differences in biological responses to exposure to hazardous airborne substances between children and adults have been reported, suggesting children to be more susceptible. Aim of this study was to improve our understanding of differences in susceptibility in cancer risk associated with air pollution by comparing genome-wide gene expression profiles in peripheral blood of children and their parents. Gene expression analysis was performed in blood from children and parents living in two different regions in the Czech Republic with different levels of air pollution. Data were analyzed by two different approaches: one method first selected significantly differentially expressed genes and analyzed these gene lists for overrepresented biological processes, whereas the other applied the T-profiler tool to directly perform pathway analyses on the total gene set without preselection of significantly modulated gene expressions. In addition, gene expressions in both children and adults were investigated for associations with micronuclei frequencies. Both analysis approaches returned considerably more genes or gene groups and pathways that significantly differed between children from both regions than between parents. Very little overlap was observed between children and adults. The two most important biological processes or molecular functions significantly modulated in children, but not in adults, are nucleosome and immune response related. Our study suggests differences between children and adults in relation to air pollution exposure at the transcriptome level. The findings underline the necessity of implementing environmental health policy measures specifically for protecting children's health.

Abbreviations: EASE, Expression Analysis Systematic Explorer; GEPAS, Gene Expression Pattern Analysis Suite; GO, Gene Ontology; IFN, interferon; MN, micronuclei; PAH, polycyclic aromatic hydrocarbon


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Funding
 References
 
There is growing evidence of the adverse impact of air pollution on children's health (1–3). In particular, upward trends in childhood cancer have raised concern; an association has been reported between geographical, and therefore related environmental, burdens and childhood cancer incidence (4). In addition, a publication by the Children's Health Study in California (USA) reported on environment-induced early-life risk factors for the development of childhood asthma (5). These studies emphasize the need to investigate the role of environmental exposure in the etiology of such diseases.

It has been suggested that rapid growth and development as well as anatomical and physiological changes in various organs and biological systems during childhood lead to increased susceptibility to hazardous environmental compounds (6,7). For children, relatively high intake of air, activity patterns, immaturity of the lungs and the immune system, as well as of the metabolism have all been reported to increase environmental health risks (3). Furthermore, induction of genetic damage early in life may increase risk of carcinogenesis in later life (8,9). This is underlined by few studies applying the molecular epidemiological approach of assessing biomarkers for environmental health risks among children and adults, which suggest a higher susceptibility of children (10,11).

Biomarkers that are currently in use in molecular (cancer) epidemiology because of their association with disease end points do, however, not necessarily provide mechanistic information on health effects. Furthermore, any classic biomarker of effect provides information only on a single end point of disease, thereby limiting conclusions on integrated environmental health risks. Whole-genome gene expression analysis through toxicogenomics is arising as a promising new tool to overcome these disadvantages (12–14). Toxicogenomic analysis provides a means to gather information both at the generic and at the mechanistic level, by simultaneously studying (changes in) expression of many genes in response to xenobiotic exposure.

Consequently, the main objective of the current study was to investigate whether children and adults, their parents, respond differently to air pollution at the transcriptome level. Therefore, we analyzed genome-wide gene expression in blood using Agilent Human 22k oligonucleotide microarrays. Differential gene expression profiles were compared between individuals from two regions in the Czech Republic with considerable differences in the levels of air pollutants, especially during winter; Teplice and Prachatice (15,16). We analyzed transcriptomic data for significant enrichment of biological pathways, processes or functions by applying two different approaches: the one method first selected significantly differentially expressed genes and fed those into the Expression Analysis Systematic Explorer (EASE) pathway-finding tool, whereas the other applied the recently developed T-profiler tool which performs pathway analyses directly on the total gene set without any preselection of significantly modulated gene expressions (17). In addition, for the purpose of phenotypic anchoring, we investigated correlations of gene expression profiles with lymphocytic frequencies of micronuclei (MN), a validated biomarker for genotoxic risk of environmental exposure (18,19), in order to identify genomic responses indicative for environmental carcinogenesis which is relevant in view of the presence of carcinogenic polycyclic aromatic hydrocarbons (PAHs) in air (Table I).


View this table:
[in this window]
[in a new window]

  		Table I. Study population characteristics and average concentrations of air pollutants measured at the ambient air quality monitoring stations of both regions (mean ± SD)

 

   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Funding
 References
 
Study population
The local ethical committee at the Institute of Experimental Medicine AS CR in Prague approved the study protocols. Table I shows the characteristics of the study population. Twelve families inhabited the polluted area of Teplice, whereas the others were enrolled from the rural and cleaner area of Prachatice (20). Twenty-four parents donated blood for this study. Gene expression profiles of 47 children were available from our previous study (21). To safely collect and store blood for RNA stabilization, the PAXgene Blood RNA System (PreAnalytix, Hilden, Germany) was used as this provides instant ex vivo RNA stabilization.

The populations in Teplice and Prachatice have been followed up for health risk research since 1997. The current study population was recruited via two pediatricians in Teplice and one pediatrician in Prachatice. The requirement was to select twelve families from each district, each family having two children who have lived in the regions for the entire lives. Refusal rates were ~20%. Pediatricians informed all parents on the study aims before they gave written informed consent prior to the study. Personal information was obtained from pediatrician records and checked with the parents upon blood collection. In both districts, the recruited families had similar incomes of average height. Therefore, differences in socioeconomic levels are considered negligible in the two populations. For three families, it was not possible to obtain blood from the mother; therefore, the father donated blood. There were no significant differences in age or gender between the study populations from both regions.

Exposure assessment
Sampling of outdoor air was performed during winter, as the levels of air pollution in Teplice and the difference in concentration with the Prachatice area are known to be highest in that season. During the periods of air sampling in the residential areas, condensation particle counters (TSI, model 3007) determined the concentrations of ultrafine particles (0.01 to >1 µm) in the ambient air at two stationary air quality monitoring stations. Sampling at the stations in each city took place simultaneously to sampling in the residential areas. Time series of the levels of common air pollutants, i.e. particulate matter (PM) PM2.5 and PM10, SO2, NOx, CO, ozone (O3), 13 PAHs and 6 carcinogenic PAHs were measured in Teplice and Prachatice areas according to enacted guidelines by the European communities (1996) and are presented in Table I. Calculations for Prachatice and Teplice regions are averaged over 5 days of 24 h sampling during the period of 27 February–3 March 2004 (except the 28th) and 7 March–11 March 2004, respectively. Particle counts are from 8 a.m. until 24 p.m. (numbers/cm3). All other concentrations are in µg/m3 except PAHs and carcinogenic PAHs (ng/m3) and CO (mg/m3).

Genome-wide gene expression analysis
We analyzed blood RNA for genome-wide gene expression using Human 1A (V2) 22k oligonucleotide microarrays (Agilent, Santa Clara, CA, USA) according to the manufacturer's instructions. Using the Low RNA Input Linear Amplification Kit (Agilent), we generated cyanine-5-labeled cRNA from RNA of each subject and hybridized each of those samples together with cyanine-3-labeled reference cRNA produced from a common reference sample. This common reference sample consisted of RNA from 21 children from Prachatice. This hybridization design resulted in 71 hybridizations, i.e. one for each individual and allows direct comparison of gene expressions of children and parents, as all individual data are relative to the same reference sample.

Data preparations and analysis
A ScanArrayExpress scanner (PerkinElmer Life and Analytical Sciences, Wellesley, MA, USA) scanned the microarray slides after hybridization and washing according to the manufacturer's instructions. We extracted raw data from scan images with ImaGene (BioDiscovery, Marina del Rey, CA, USA) and performed data transformations using GeneSight (BioDiscovery) and the online tool Gene Expression Pattern Analysis Suite (GEPAS v3.1; http://gepas.bioinfo.cipf.es) (22,23), including local background subtraction, log2 transformation, LOWESS normalization and omitting low-abundance genes with <2 signal-to-background ratio and genes with <70% present data.

We conducted our data analysis following two different approaches. The first approach was to primarily capture individual gene expression modifications that statistically significantly differed between individuals from both areas using GEPAS and to use these lists to perform pathway enrichment analysis. Lists of significantly differentially expressed genes (all genes returned from GEPAS with an unadjusted P-value <0.05) were fed into EASE (24) for analysis of overrepresented biological processes or molecular functions.

The second approach comprised T-profiler analyses to identify responsive gene groups or networks within the entire data set. T-profiler uses the unpaired t-test to score changes in average transcriptomic activity of predefined groups of genes within a single expression profile and expresses the significance of the change as a t-value (17). This approach does not preselect significantly modulated gene expressions but is based on the complete gene expression profiles. Thereby, all expression changes, also those that are non-significantly modulated, are incorporated to determine significantly overrepresented and biologically relevant pathways. For T-profiler analysis, we filtered the microarray data set using another, more stringent method to remove low-intensity spots; a spot intensity of <10% of the average intensity of all spots of the same array was taken as threshold. In total, 7199 spots showed a low intensity in at least 83% of the arrays (five of six arrays used for this filtering) and were removed from the data set. To reduce the influence of genes with extreme expression changes, which may result in false positives or false negatives, we discarded the highest and lowest expression values in each gene group. For pathway identification, we used Gene Ontology (GO; ftp://ftp.ebi.ac.uk/pub/databases/GO/goa/HUMAN/gene_association.goa_human.gz) and the Kyoto Encyclopedia of Genes and Genomes (http://www.genome.jp/kegg-bin/show_organism?menu_type= pathway_maps& org=hsa) gene groups and gene groups developed for Gene Set Enrichment Analysis (http://www.broad.mit.edu/cancer/software/gsea_beta/msigdb/cards_ index.html). We have uploaded raw and transformed data files into the Gene Expression Omnibus (series record number GSE7543 [NCBI GEO] ) at http://www.ncbi.nlm.nih.gov/GEO.

Correlations of gene expression with MN frequencies
MN frequencies in peripheral blood lymphocytes from children and their parents have been reported previously (25). To evaluate associations of gene expressions with MN frequencies (Pearson correlation), we applied the GEPAS T-Rex tool. Additionally, we classified children and parents into quartiles according to their MN frequencies and subsequently analyzed the lowest and highest quartiles of the subpopulations for differential genetic pathways using T-profiler.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Funding
 References
 
Pathway finding upon preselection of genes with significant differential expression between regions
The numbers of genes that passed the filters in GeneSight and GEPAS were 20 062 for the children and 19 197 for the parents. Considerable numbers of genes appeared differentially expressed between children as well as adults from the Teplice area as compared with the Prachatice area. In addition, children from Teplice were observed to show a higher number of significantly differentially expressed genes as compared with their parents; most of the genes being upregulated, whereas most of the significantly differentially expressed genes in the parents appeared downregulated. Unadjusted P-values were used as GEPAS returned no genes with adjusted P-values <0.05. Although the numbers of genes found in the parents might in theory all be false positives, the numbers found in the children are well above the theoretically expected number of false positives. At P < 0.01, children returned 471 gene expressions to differ significantly between individuals from both regions (305{uparrow} and 166{downarrow} in Teplice as compared with Prachatice) and 1698 gene expressions (982{uparrow} and 716{downarrow}) at P < 0.05. In comparison, parents returned 140 genes differentially expressed between individuals from both regions at P < 0.01 (42{uparrow} and 98{downarrow} in Teplice as compared with Prachatice) and 753 at P < 0.05 (200{uparrow} and 553{downarrow}). Analysis of these lists of genes by the EASE pathway-finding tool identified more significantly enriched biological processes or molecular functions in children than in parents. In addition, overall EASE scores were lower in the children, showing higher significance of affected pathways. The three most significantly enriched (overrepresented) biological processes or molecular functions identified by EASE in children and in parents are summarized in Table II.


View this table:
[in this window]
[in a new window]

  		Table II. Biological processes or molecular functions significantly overexpressed in lists of genes significantly differentially expressed

 
In view of the larger sample size of children in comparison with their parents, we evaluated 10 random subsamples, existing of 11 or 12 Teplice children statistically tested against 12 Prachatice children. The numbers of genes significantly differing between these subsamples of Teplice and Prachatice children were always higher than in the parents. A substantial overlap (25 genes) was observed between these heterogeneous subsamples. EASE analysis on two of the subsamples was conducted, which delivered a similar result as described above for the total population of children, in terms of overrepresented processes and functions, such as nucleosome-related terms. We therefore conclude that differences in sample size between children and their parents have no impact.

Assessment of genes modulated in the same direction in children and in parents (P < 0.01) delivered two genes upregulated and three genes downregulated in both groups. At P < 0.05, 43 genes were similarly affected in children as well as in their parents; 8 genes upregulated and 35 downregulated. Among the five genes found at P < 0.01, in children and parents, three were functionally associated with immune or inflammatory response.

A possible gender effect was investigated by looking for the formation of specific male or female clusters in hierarchical clustering analysis (data not reported) based on genes that significantly differ in expression between Teplice and Prachatice individuals. No such clusters were observed. In addition, the same hierarchical clustering analyses showed that smokers did not cluster together; therefore, we do not suspect interference by smoking.

For additional analysis and comparison of the degree of gene expression induction or repression (fold changes), expression of each Teplice individual versus the average expression of Prachatice individuals was calculated separately for children and parents (using the data set in which 7199 low-intensity spots were removed). The number of Teplice individuals with upregulated gene expression versus the average of the Prachatice group was calculated as well as the number of Teplice individuals with downregulated gene expression versus the average of the Prachatice group. The numbers of genes according to their fold changes are summarized in Figure 1. Genes were only regarded when e.g. the number of individuals with >1.23-fold upregulation minus the number of individuals with >1.23-fold downregulation for the same gene was at least 11 in children and 6 in parents (half the population size). Overall, these fold changes show no considerable differences between children and adults.


Figure 1
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Numbers of genes differentially expressed in children and parents with >1.23-, >1.41- or >1.62-fold up- or downregulation in Teplice individuals as compared with the average expression of Prachatice individuals. Children and parent numbers are unique for these populations while the intersections represent common gene expressions between children and parents.

 
Network analysis without preselection of significant gene expressions
For T-profiler analyses, the mean expression per gene was compared between the group of Teplice children and the group of Prachatice children or between the group of Teplice parents and the group of Prachatice parents. This returned ratios of Teplice/Prachatice, which were entered into T-profiler. The GO gene groups that were found to be significantly different between children from Teplice and Prachatice were ‘nucleosome’ (upregulated in Teplice children) as well as ‘immune response’ and ‘response to virus’ (downregulated in Teplice children; Table III). Specificity of T-profiler results was assessed by performing the same analysis in the group of Prachatice children only, which was split into two random groups of equal size. The pathways that were observed as significantly different between these groups were ‘deducted’ from those observed to differ between the total groups of Teplice and Prachatice children, leaving nucleosome and immune response as specifically responsive to air pollution as they did not show in the Prachatice subgroup analysis. In the parents, T-profiler identified no GO gene groups differentially expressed between individuals from Teplice versus Prachatice.


View this table:
[in this window]
[in a new window]

  		Table III. Gene groups significantly differentially expressed between children from Teplice and Prachatice as identified by T-profiler and their association with a specific blood cell type

 
T-profiler analyses were also performed using motif/transcription factor-associated gene groups, which also delivered no significant results for the parents. In Teplice children, however, five significant downregulated motif gene groups were identified which were all associated with binding by interferon (IFN) regulation factors, therefore being responsive to IFN (Table III). Figure 2 depicts a heatmap comparative overview of GO and motif gene groups significantly modulated in at least one individual for children, none of these pathways being significantly affected in parents. No similarities or overlap between gene groups in children and parents were demonstrated, indicating a considerable difference in genomic response at pathway level between children and adults.


Figure 2
View larger version (31K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Heatmap comparative overview of T-profiler results for children and parents. Gene groups were included based on significance in children as no significant groups were identified in parents.

 
As peripheral blood was used for analysis of genomic responses, gene groups particularly associated with specific blood cell types were evaluated by T-profiler. Nucleosome-related genes, which appeared to be expressed more abundantly in Teplice children than in Prachatice children, showed a lower expression in (Prachatice) individual samples characterized by a high expression of T-cell-specific genes. In addition, low expression of genes in the response to virus and immune response gene groups in Teplice children relative to Prachatice children was associated with high expression of macrophage-specific genes. Figure 2 also depicts a T-profiler analysis on tissue (blood cell)-specific gene groups.

Association of gene expression with individual MN frequencies
Numbers of gene expressions that significantly correlated with MN frequencies as identified by GEPAS (P < 0.05) were 1250 (678 positively and 572 negatively) in children and 499 (167 positively and 332 negatively) in adults. The most significantly overrepresented EASE pathways within the total sets of genes correlating with MN encompassed messenger RNA-related processes such as processing, splicing and binding and protein metabolism in the children (EASE scores from 0.004 to 0.03) and (homophilic) cell adhesion, protein binding and calcium ion binding in the parents (EASE scores 0.0000009, 0.003 and 0.003, respectively).

Upon stratification according to MN frequencies, overrepresentation analysis of gene groups within gene sets significantly differing in expression between the six children with the highest versus the six children with the lowest MN frequencies identified as most significant cell-surface receptor-linked signal transduction, G-protein-coupled receptor protein signaling pathway and cyclic nucleotide metabolism (EASE scores 0.004, 0.004 and 0.005, respectively) among genes upregulated in individuals with high MN. Among genes downregulated in individuals with high MN, the most significantly overrepresented were RNA binding, metabolism and nucleic acid binding (EASE scores 0.000003, 0.000005 and 0.000009, respectively). For parents, these analyses revealed protein binding, transcription factor activity and development (EASE scores 0.005, 0.006 and 0.01, respectively) among upregulated genes and transferase activity, catalytic activity and metallopeptidase activity (EASE scores 0.02, 0.03 and 0.04 respectively) among downregulated genes.

Correlation analysis of gene group expressions and MN frequencies was also performed using T-profiler; however, no significant correlations for children or parents were identified. T-profiler results of the analysis of high MN versus low MN individuals are shown in Table IV; the most significant results (–20 > t-value > 20) being protein biosynthesis, ribosome and structural constituent of ribosome for the children. In the parents, the gene group with the highest t-value was response to virus (t-value = –7.76).


View this table:
[in this window]
[in a new window]

  		Table IV. Results of T-profiler analysis of highest versus lowest MN quartiles in children (A) and parents (B)

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Funding
 References
 
Over the past decades, trends of increased incidences of childhood cancer and atopy (4,5) have been observed, which warrant investigation of the actual causes. To our knowledge, the present study is the first to assess susceptibility of children to hazardous environmental exposure in comparison with adults at the transcriptome level. For the current study, blood gene expression profiles of adults were analyzed for genomic responses to air pollution and compared for similarities and differences with gene expressions in their children (21) to enable the comparison of the response with air pollution in children and in adults.

Pathway analysis after preselection of genes with significant differential expression between regions
Genome-wide gene expression analysis revealed differences in numbers of genes significantly responsive to air pollution in children and adults. Overall, children exhibited 2- to 3-fold higher numbers of differentially expressed genes than their parents, while more genes appeared upregulated than downregulated. The highest upregulated processes in Teplice children compared with Prachatice children as identified by EASE, e.g. nucleosome and chromatin assembly and disassembly, were not observed among parents and therefore seem specific for children's response to air pollution. In addition, (EASE) analysis yielded very little overlap in responsive genes between children and their parents. This suggests a considerable difference in genomic response to air pollution in children and higher responsiveness in children.

Analysis of the gene expression fold changes in children and adults revealed relatively low numbers of genes with >1.23-fold change (Figure 1). This prohibited analysis of overrepresented processes or pathways in these lists of genes. Small fold changes are common in in vivo transcriptome analyses among human populations exposed to environmental compounds, and the changes are not in the same order as in vitro or experimental animal studies (26–28). Since in a previous study on gene expression in the children's population as well as a transcriptomic study in the blood of smokers, the expressions of selected genes measured by microarray analyses were confirmed by real-time polymerase chain reaction analysis (21,28), we do not consider the small fold changes as an analytical problem. Furthermore, it has been published that the current microarrays are developed so thoroughly that confirmation by independent methods such as polymerase chain reaction are no longer necessary (29). Recent reports showed that intraindividual variation in gene expression is considerably lower than the interindividual variation (30–32). Therefore, we believe that group differences are due to interindividual differences. In future studies, multiple measurements per individual could shed light on intraindividual variability, allowing to compare this with interindividual differences and group differences between differentially exposed individuals.

Network analysis without preselection of significant gene expressions
T-profiler analysis of significantly overrepresented gene sets within the complete data set revealed significantly altered GO gene groups among children (Teplice versus Prachatice children), but not in their parents (Table III and Figure 2). In children but not in their parents, T-profiler analysis, in confirmation of results obtained by EASE, identified gene groups associated with nucleosome and immune response as most significant, again demonstrating child-specific responses to environmental exposure. The relevance of the ‘nucleosome finding’ for children's health is explained by the suggested impact on deregulation of condensation of the DNA, as the nucleosome is the primary structural unit of chromatin. Chromatin is known to play a central role in regulation of gene expression, chromosome replication and cell-cycle progression. In addition, chromatin maintains integrity of the genome by guarding access to the DNA template (33). The importance of affecting immune and inflammatory response pathways is obvious with respect to children's health; deregulated immune function or reduced ability to respond to pathogens increases risk for an array of (infectious) diseases, including cancer.

T-profiler analysis of the data for motif gene groups identified modulated groups of genes based on a shared motif. Motif groups are defined as genes with a match to a particular consensus motif within 600 bp upstream of the open reading frame, allowing no overlap with neighboring open reading frames (17). T-profiler revealed significant motif groups in children but again not in parents. Interestingly, five downregulated motif gene groups, among genes downregulated in Teplice as compared with Prachatice children, referred to ‘IFN-associated motif gene groups’. The IFN family plays critical roles in the host response to pathogens and members have recently been identified to possess antitumor activity (34). Our observation of lower expression of gene motif groups related to IFN may suggest impairment of host defense and antitumor activity in children exposed to higher levels of air pollution as compared with children from a less burdened area of residence. This disadvantage at such an early age might be of profound importance for the development of those children and their cancer risk throughout life.

Enrichment of these IFN-related motif groups among significantly differentially expressed genes that are downregulated in Teplice children as compared with Prachatice children is in line with the observed downregulation of GO groups' immune response and response to virus in Teplice children. This suggests that children inhabiting the polluted area may have a weaker immune or defense response. Related to this may also be the observation of significant expression of gene groups associated with different blood cell types; children from the Teplice area expressing significantly lower abundance of genes in the immune response and response to virus groups, exhibited significantly lower expression of genes specific for macrophages. On the contrary, the finding that nucleosome-related genes were more abundantly expressed in children from Teplice may be associated with a lower expression of these genes in individuals with a high expression of T-cell-specific genes. In this aspect, in general, children from Teplice differed from Prachatice children in expression levels of genes specific to neutrophilic granulocytes, T cells, macrophages and early erythrocytes (reticulocytes). Virtually no literature is available on differential gene expression in blood of human populations induced by environmental factors such as air pollution. Variation in gene expression patterns in human blood has been reported, indicating that specific features of interindividual variation in peripheral blood gene expression patterns were related to variation in the relative proportions of specific blood cell subsets (32). Our observations thus suggest an influence of the exposure on blood cell type distribution and related specific gene expressions, and further investigation of this is warranted. It may be speculated that a shift in cell types itself may represent a severe response to pollution.

Overall, we suggest that the higher numbers of significantly differentially expressed genes and pathways/processes may translate into higher responsiveness to detrimental air pollutants among children.

Association of gene expression with individual MN frequencies
MN frequency data were included in the analysis for the purpose of associating gene expression modifications with this validated biomarker of genotoxic effects, which is indicative of cancer risk (18,19). We hypothesized that by extracting MN-associated biological processes from the processes or pathways found in the analysis of differential gene expression between individuals from both regions, a selection of pathways associated with environmental carcinogenic risk could be generated. Both statistical analyses identified processes associated with messenger RNA fate (splicing, metabolism and binding) to correlate with MN frequencies among children, which suggest a generic effect in association with chromosomal damage on the translation of genetic messages into functional protein. However, in general, the EASE- and T-profiler-derived processes and functions that are most significantly associated with MN frequencies cannot be interpreted in molecular processes leading to chromosome instability and the subsequent formation of MN.

In summary, it appears feasible to generate discriminative profiles of modulated gene pathways at rather low differences in exposure levels of environmental carcinogens. Our main observation concerns the large differences in genomic response to air pollution between children and parents; in terms of individual genes significantly differentially expressed between individuals from both regions as well as in terms of enriched gene groups identified by T-profiler; all being to a larger extent affected in children than in their parents. Furthermore, children predominantly show upregulation and parents mostly downregulation. Specific pathways responsive in children and not in parents are nucleosome, immune response and IFN-associated motif gene groups, which suggest impacting of air pollution on mechanisms of DNA integrity as well as suppressing of immune function. Very little overlap was seen in genomic responses of children and parents. All this suggests a considerably different response to air pollution among exposed children in comparison with adults. Our findings underline the necessity of implementing environmental health policy measures specifically for protecting children's health.


   Funding
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Funding
References
 
European Union FP5 Concerted Action ChildrenGenoNetwork (QLK4-CT-2002-02198).


   Acknowledgments
 
The authors gratefully acknowledge Dr D.Horakova, Dr J.Ruzickova and Dr M.Weigartova for administration of questionnaires and blood sampling and the Laboratory of Genetic Toxicology, National Institute of Public Health, Prague, for processing of MN cultures.

Conflict of interest statement: None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Funding
 References
 

  1. Bobak M. Outdoor air pollution, low birth weight, and prematurity. Environ. Health Perspect. (2000) 108:173–176.[Web of Science][Medline]
  2. Perera FP, et al. Molecular epidemiologic research on the effects of environmental pollutants on the fetus. Environ. Health Perspect. (1999) 107(suppl. 3):451–460.[Web of Science][Medline]
  3. WHO-Europe. Effects of Air Pollution on Children's Health and Development. A Review of the Evidence. In: Special Programme on Health and Environment (2005) World Health Organization, European Centre for Environment and Health, Bonn Office, Bonn, Germany.
  4. Steliarova-Foucher E, et al. Geographical patterns and time trends of cancer incidence and survival among children and adolescents in Europe since the 1970s (the ACCIS project): an epidemiological study. Lancet (2004) 364:2097–2105.[CrossRef][Web of Science][Medline]
  5. Salam MT, et al. Early-life environmental risk factors for asthma: findings from the Children's Health Study. Environ. Health Perspect. (2004) 112:760–765.[Web of Science][Medline]
  6. Armstrong TW, et al. A tiered approach to assessing children's exposure: a review of methods and data. Toxicol. Lett. (2002) 127:111–119.[CrossRef][Web of Science][Medline]
  7. Scheuplein R, et al. Differential sensitivity of children and adults to chemical toxicity. I. Biological basis. Regul. Toxicol. Pharmacol. (2002) 35:429–447.[CrossRef][Web of Science][Medline]
  8. Anderson LM, et al. Critical windows of exposure for children's health: cancer in human epidemiological studies and neoplasms in experimental animal models. Environ. Health Perspect. (2000) 108(suppl. 3):573–594.[CrossRef][Web of Science][Medline]
  9. Wild CP, et al. Children and increased susceptibility to environmental carcinogens: evidence or empathy? Cancer Epidemiol. Biomarkers Prev. (2003) 12:1389–1394.[Free Full Text]
  10. Neri M, et al. Children's exposure to environmental pollutants and biomarkers of genetic damage. I. Overview and critical issues. Mutat. Res. (2006) 612:1–13.[CrossRef][Web of Science][Medline]
  11. Neri M, et al. Children's exposure to environmental pollutants and biomarkers of genetic damage. II. Results of a comprehensive literature search and meta-analysis. Mutat. Res. (2006) 612:14–39.[CrossRef][Web of Science][Medline]
  12. Aardema MJ, et al. Toxicology and genetic toxicology in the new era of "toxicogenomics": impact of "-omics" technologies. Mutat. Res. (2002) 499:13–25.[Web of Science][Medline]
  13. Simmons PT, et al. Toxicogenomics: the new frontier in risk analysis. Carcinogenesis (2002) 23:903–905.[Abstract/Free Full Text]
  14. Watson WP, et al. Role of biomarkers in monitoring exposures to chemicals: present position, future prospects. Biomarkers (2004) 9:211–242.[CrossRef][Web of Science][Medline]
  15. Binkova B, et al. Biological activities of organic compounds adsorbed onto ambient air particles: comparison between the cities of Teplice and Prague during the summer and winter seasons 2000–2001. Mutat. Res. (2003) 525:43–59.[Web of Science][Medline]
  16. Sram RJ, et al. Teplice program—the impact of air pollution on human health. Environ. Health Perspect. (1996) 104(suppl. 4):699–714.[CrossRef][Web of Science][Medline]
  17. Boorsma A, et al. T-profiler: scoring the activity of predefined groups of genes using gene expression data. Nucleic Acids Res. (2005) 33:W592–W595.[Abstract/Free Full Text]
  18. Bonassi S, et al. Human population studies with cytogenetic biomarkers: review of the literature and future prospectives. Environ. Mol. Mutagen. (2005) 45:258–270.[CrossRef][Web of Science][Medline]
  19. Bonassi S, et al. An increased micronucleus frequency in peripheral blood lymphocytes predicts the risk of cancer in humans. Carcinogenesis (2007) 28:625–631.[Abstract/Free Full Text]
  20. Hertz-Picciotto I, et al. Immunity and childhood morbidity in the Teplice program. In: Teplice Program: Impact of Air Pollution on Human Health—Sram RJ, ed. (2001) Prague, Czech Republic: Academia. 207–217.
  21. van Leeuwen DM, et al. Genome-wide differential gene expression in children exposed to air pollution in the Czech Republic. Mutat. Res. (2006) 600:12–22.[Web of Science][Medline]
  22. Herrero J, et al. GEPAS: a web-based resource for microarray gene expression data analysis. Nucleic Acids Res. (2003) 31:3461–3467.[Abstract/Free Full Text]
  23. Montaner D, et al. Next station in microarray data analysis: GEPAS. Nucleic Acids Res. (2006) 34:W486–W491.[Abstract/Free Full Text]
  24. Hosack DA, et al. Identifying biological themes within lists of genes with EASE. Genome Biol. (2003) 4:R70.[CrossRef][Medline]
  25. Pedersen M, et al. Cytogenetic effects in children and mothers exposed to air pollution assessed by the frequency of micronuclei and fluorescence in situ hybridization (FISH): a family pilot study in the Czech Republic. Mutat. Res (2006) 608:112–120.[Web of Science][Medline]
  26. Wang Z, et al. Global gene expression profiling in whole-blood samples from individuals exposed to metal fumes. Environ. Health Perspect. (2005) 113:233–241.[Web of Science][Medline]
  27. Amundson SA, et al. Identification of potential mRNA biomarkers in peripheral blood lymphocytes for human exposure to ionizing radiation. Radiat. Res. (2000) 154:342–346.[CrossRef][Web of Science][Medline]
  28. van Leeuwen DM, et al. Cigarette smoke-induced differential gene expression in blood cells from monozygotic twin pairs. Carcinogenesis (2007) 28:691–697.[Abstract/Free Full Text]
  29. Shi L, et al. The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements. In: Nat. Biotechnol. (2006) 24:1151–1161.[CrossRef][Web of Science][Medline]
  30. Eady JJ, et al. Variation in gene expression profiles of peripheral blood mononuclear cells from healthy volunteers. Physiol Genomics (2005) 22:402–411.[Abstract/Free Full Text]
  31. Radich JP, et al. Individual-specific variation of gene expression in peripheral blood leukocytes. Genomics (2004) 83:980–988.[CrossRef][Web of Science][Medline]
  32. Whitney AR, et al. Individuality and variation in gene expression patterns in human blood. Proc. Natl Acad. Sci. USA (2003) 100:1896–1901.[Abstract/Free Full Text]
  33. Moggs JG, et al. The role of chromatin in molecular mechanisms of toxicity. Toxicol. Sci. (2004) 80:218–224.[Abstract/Free Full Text]
  34. Jonasch E, et al. Interferon in oncological practice: review of interferon biology, clinical applications, and toxicities. Oncologist (2001) 6:34–55.[Abstract/Free Full Text]
Received June 28, 2007; revised February 27, 2008; accepted March 2, 2008.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Toxicol SciHome page
D. G. A. J. Hebels, D. G. J. Jennen, J. C. S. Kleinjans, and T. M. C. M. de Kok
Molecular Signatures of N-nitroso Compounds in Caco-2 Cells: Implications for Colon Carcinogenesis
Toxicol. Sci., April 1, 2009; 108(2): 290 - 300.
[Abstract] [Full Text] [PDF]

Home page
 MutagenesisHome page
C. P. Wild
Environmental exposure measurement in cancer epidemiology
Mutagenesis, March 1, 2009; 24(2): 117 - 125.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
29/5/977    most recent
bgn065v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by van Leeuwen, D. M.
Right arrow Articles by Kleinjans, J. C.S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by van Leeuwen, D. M.
Right arrow Articles by Kleinjans, J. C.S.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?