Skip Navigation


Carcinogenesis Advance Access originally published online on October 19, 2006
Carcinogenesis 2007 28(3):691-697; doi:10.1093/carcin/bgl199
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
28/3/691    most recent
bgl199v1
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 Leeuwen, D. M.v.
Right arrow Articles by Delft, J. H.M.v.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Leeuwen, D. M.v.
Right arrow Articles by Delft, J. H.M.v.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Cigarette smoke-induced differential gene expression in blood cells from monozygotic twin pairs

Danitsja M.van Leeuwen1, Ebienus van Agen1, Ralph W.H. Gottschalk1, Robert Vlietinck2, Marij Gielen2, Marcel H.M.van Herwijnen1, Lou M. Maas1, Jos C.S. Kleinjans1 and Joost H.M.van Delft1,*

1 Department of Health Risk Analysis and Toxicology, Maastricht University PO Box 616, 6200 MD Maastricht, The Netherlands
2 Department of Population Genetics, Genomics and Bioinformatics Maastricht University, PO Box 616, 6200 MD Maastricht, The Netherlands

*To whom correspondence should be addressed. Tel: +31 43 3881092; Fax: +31 43 3884146; Email: J.vanDelft{at}GRAT.unimaas.nl


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 References
 
Chemical carcinogenesis induced by lifestyle factors like cigarette smoking is a major research area in molecular epidemiology. Gene expression analysis of large numbers of genes simultaneously using microarrays holds the opportunity to study the effects of such an exposure at the genome level yielding more mechanism-based information. Therefore, the aim of our study was to investigate multiple gene expressions in blood, indicative for the effects caused by cigarette smoke. Smoking-discordant monozygotic twin pairs (n = 9) were studied to diminish influences of genetic background. Using a dedicated microarray containing 600 toxicologically relevant genes, we investigated which genes are differentially expressed in smokers compared to non-smokers. We also looked for genes of which the expression changes correlated with DNA adducts, a biomarker of effective dose for exposure to cigarette smoke carcinogens. The mean DNA adduct level in smokers differed significantly from that in non-smokers (mean ± standard error 1.96 ± 0.24 versus 1.17 ± 0.16 adducts per 108 nucleotides, respectively; P = 0.04). The genes of which the expression differed most significantly between smokers and non-smokers are ATF4, MAPK14, SOD2, CYP1B1 and SERPINB2. CYP1B1 and SOD2 can directly be linked to cigarette smoke exposure, whereas the other genes are associated with stress or environmentally induced response. Main functions of the genes influenced by cigarette smoking comprise carcinogen metabolism, oxidative stress response and anti-apoptosis.

Abbreviations: EDTA, ethylenediamine tetraacetic acid; PBMC, peripheral blood mononuclear cells


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 References
 
Chemical carcinogenesis induced by lifestyle factors like cigarette smoking is a major research area in molecular epidemiology. Cigarette smoke is a well known source of chemical carcinogens and comprises a complex mixture of over 60 proven, probable and possible carcinogenic agents (1). It is known to induce a series of biomarkers for genotoxic, pre-carcinogenic effects, both in target tissue as in surrogate target tissue like peripheral blood mononuclear cells (PBMC) (2,3). Many of these early effects, e.g. DNA adducts, micronuclei, chromosome aberrations and mutations, are detectable in blood cells long before health effects appear (4,5).

Gene expression profiling by DNA microarray technology is considered a promising tool to gain more knowledge on the effects of toxicant exposure on the transcriptome level (6–8). The development of tools for analyzing differential gene expression of many genes simultaneously has provided the opportunity for analyzing effects in exposed populations at the genome level, thereby studying the biological response to environmental exposure in an integrative manner.

Differential gene expression in human peripheral blood cells in vivo has been reported in a few studies (9–12). In a study on human leukocyte gene expression in smokers versus non-smokers, particular gene expression modulations have been found to correlate significantly with plasma cotinine levels and these genes accurately distinguished smokers from non-smokers (10). No study has yet associated gene expression in exposed populations with established biomarkers like DNA adducts, which would provide an important opportunity for phenotypical anchoring of gene expression profiles.

The goal of our study therefore, was to investigate which genes are differentially expressed in blood cells from smokers as compared to non-smokers and for which genes the changes in expression correlate with aromatic DNA adduct levels. In order to reduce the potential impact from differences in genetic background, monozygotic twin pairs that were discordant in their smoking behavior were investigated as study subjects, implying that each twin pair consisted of one smoker and one non-smoker. Plasma levels of cotinine (a stable metabolite of nicotine) levels were measured to confirm smoking behavior. We used a toxicologically focused cDNA microarray to analyze gene expression differences as we anticipate that potential biomarker genes for cigarette smoke exposure should be represented within such a typical selection of target genes.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 References
 
Study population and blood collection
Participants were registered members of the East-Flanders Prospective Twin Survey (EFPTS) (13). All twins over age of 20 years were sent questionnaires to collect information on smoking behavior. Twin pairs were included if they reported to be monozygotic and discordant in cigarette-smoking behavior (per twin pair one individual is a smoker whereas the other is a non-smoker). Initially, the study population consisted of 10 pairs of smoking-discordant twins; 5 male–male pairs and 5 female–female pairs. Genotyping for several biotransformation and DNA repair genes indicated that the two individuals of one male–male twin pair differed in 6 out of 19 polymorphisms (data not shown) and thus appeared to be a dizygotic twin pair. This twin pair was therefore excluded from further analysis. Among the non-smokers, there were seven never smokers. Two were former smokers and quit smoking at least 13 months prior to the study. Subjects gave informed consent and the study was approved by the Local Committee of Medical Ethics of the Catholic University of Leuven, Belgium. Venous blood was collected into one tube containing ethylenediamine tetraacetic acid (EDTA) and into five PAXgene Blood RNA tubes (PreAnalytix, Qiagen, Hilden, Germany) for immediate stabilization of intracellular RNA. Tubes were kept at 4°C until DNA and RNA isolation. The PAXgene system was used as it provides a robust standardized and simple method to collect blood for gene expression purposes in the field, where timely isolation of, e.g. lymphocytes is not possible.

Total RNA isolation
Total RNA was isolated within a week after blood collection according to the manufacturer's instructions with minor modifications using the PAXgene Blood RNA kit (PreAnalytix, Qiagen, Hilden, Germany). The entire procedure was carried out at room temperature. Samples that did not show a distinct RNA-debris phase separation were submitted to an extra centrifugation step. RNA purity was determined spectrophotometrically, based on A260/A280 ratios. RNA integrity was determined on the BioAnalyzer and RNA was regarded intact when showing two distinct bands for 28S and 18S RNA.

Gene expression analyses by cDNA microarrays
A common reference RNA pool was made from aliquots of all individuals that yielded sufficient RNA. On all arrays, three RNA samples were hybridized together (14), i.e. the common reference sample and samples from the two individuals of a twin pair: the smoker and the non-smoker (for hybridization design, see Table I). This allowed for analyzing differential gene expression at the level of twin pairs, and also at group levels for non-smokers and smokers. Total RNA was reverse transcribed and labeled for cDNA microarray analysis according to the TIGR protocol for amino-allyl labeling of total eukaryotic RNA for microarrays as described previously (15). For each individual, three aliquots of total RNA were used and labeled with Cy5, Cy3 (Amersham Biosciences, Freiburg, Germany) or Alexa Fluor 594 (ARESTM DNA labeling kit, Molecular Probes, Leiden, The Netherlands), according to the manufacturers' instructions. For each twin pair, three hybridizations were performed, using a design in which each sample was labeled by each dye once (see Table I). Equal amounts of labeled cDNA were combined and hybridized to Phase1 Human Tox 600 cDNA microarrays (Phase-1 Molecular Toxicology, Santa Fe, NM, USA), which contains genes in quadruplicate, in toxicologically relevant categories such as apoptosis, cell cycle, proliferation, biotransformation and metabolism and inflammation.


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

  		Table I Hybridization design for RNA samples from every twin pair using three fluorophores

 
Microarrays were scanned on a ScanArrayExpress scanner (Perkin Elmer, Boston, MA, USA) with a fixed laser power and a variable photomultiplier gain, such that no spots were saturated. TIFF images for the separate fluorophores were imported into Imagene (BioDiscovery, Marina del Rey, CA, USA) and translated into raw data files, which were transformed and analyzed using GeneSight (BioDiscovery). Local background was subtracted and low expression values (<5 pixel intensities) and bad spots were omitted. Thereafter, data were log2 transformed and normalized with Locally Weighed Scatterplot Smoothing (LOWESS). After calculation of the differences, replicate spots were combined omitting outliers greater than two times the standard deviation, so that eventually the mean intensities were based on 12 replicates per spot (3 arrays x 4 replicates per array).

Gene expression analyses by real-time PCR
Real-time PCR was performed to validate the microarray results for a selection of genes. cDNA was generated from 2 µg of total RNA according to the manufacturer's instructions using the Bio-Rad iScript cDNA synthesis kit (Bio-Rad Laboratories, Hercules, CA, USA). Aliquots were used for real-time PCR on the BioRad MyiQ iCycler Single Color real-time PCR detection system using iQTM SYBR® Green Supermix, containing all ingredients; iTaq Polymerase, dNTPs, SYBR Green I and buffers. Reactions were initiated for 3 min at 95°C, then 40 cycles of 15 s at 95°C and 45 s at 60°C were run. Melting curve analysis was performed starting at 60°C with stepwise temperature elevations of 0.5°C every 10 s. Beta-actin (ACTB), cyclophillin A (PPIA) and beta-2-microglobulin (B2M) were included as reference genes (internal controls) and all reactions were performed in duplicate. No template controls (water controls) and a dilution series of a pool of all cDNA samples from the study were included to estimate PCR efficiency. Primer sequences used were as follows:

ß-actin: 5'-CCTGGCACCCAGCACAAT-3' (forward) and 5'-GCCGATCCACACGGAGTACT-3' (reverse); B2M: 5'-TGACTTTGTCACAGCCCAAGATA-3' (forward) and 5'-AATGCGGCATCTTCAAACCT-3' (reverse); PPIA: 5'-TTCCTGCTTTCACAGAATTATTCC-3' (forward) and 5'-GCCACCAGTGCCATTATGG-3' (reverse); ATF4: 5'-CTCCAGCGACAAGGCTAAGG-3' (forward) and 5'-GTTGTTGGAGGGACTGACCAA-3' (reverse); MAPK14: 5'-TGAAGACTGTGAGCTGAAGATTCTG-3' and 5'-CCACGTAGCCTGTCATTTCATC-3' (reverse); SOD2: 5'-ATCAGGATCCACTGCAAGGAA-3' (forward) and 5'-CGTGCTCCCACACATCAATC-3' (reverse); PTEN: 5'-CAGTGGCGGAACTTGCAAT-3' (forward) and 5'-CGTCGTGTGTGGGTCCTGAATT-3' (reverse); APP: 5'-CCGCTCTGCAGGCTGTTC-3' (forward) and 5'-CGCGGACATACTTCTTTAGCATATT-3' (reverse); AXIN1: 5'-GAGAGCCCCAAAGTCTGTAGTGA-3' (forward) and 5'-AAGGTCGGCAGGTATCCAGAT-3' (reverse); BAK1: 5'-GAATGCCTATGAGTACTTCACCAAGA-3' (forward) and 5'-ACACGGCCCCAATTGATG-3' (reverse); CSF1R: 5'-CAGCACCAACAACGCTACCTT-3' (forward) and 5'-CGGGCAGGGTCTTTGACATA-3' (reverse); IL-10: 5'-GGCGCTGTCATCGATTTCTT-3' (forward) and 5'-TGGAGCTTATTAAAGGCATTCTTCA-3' (reverse); OGG1: 5'-AGAGGTGGCTCAGAAATTCCAA-3' (forward) 5'-CAGATAAAAGAGAAAAGGCATTCGA-3' (reverse); PCK2: 5'-CAAGACCAACCTGGCTATGATG-3' (forward) and 5'-GAGTCGACCTTCACTGTCAAACC-3' (reverse), CYP1B1 5'-AGTGCAGGCAGAATTGGATCA-3' (forward) and 5'-GCGCATGGCTTCATAAAGGA-3' (reverse); and SERPINB2: 5'-CGATTTTGCAGGCACAAGCT-3' (forward) and 5'-CCTGTGGATGCATTGATTGC-3' (reverse).

Data analysis
Computational and statistical analyses of the microarray data were performed using software programs GeneSight (BioDiscovery) and SPSS 11.5 (SPSS Inc., Chicago, IL, USA). Gene expressions as modulated by cigarette smoking were identified using several approaches, either by analyzing twin pairs (the expression differences between a smoker and non-smoker directly) or by analyzing each person individually (the expression difference between an individual and the common reference) through pair-wise analyses. Confidence analysis, performed in GeneSight at >99% confidence level and >0.1 regulation level, was used to find differentially expressed genes within the twin pairs. For testing differences between smokers and non-smokers, using the expression differences between each individual and the common reference, the non-parametric 2-related samples Wilcoxon test in SPSS 11.5 was used. Next to these two tests, the signal-to-noise ratio for the smokers versus the non-smokers was calculated: (mean differencesmokers/reference – mean differencenon-smokers/reference)/(stdev differencesmokers/reference + stdev differencenon-smokers/reference). The combination of these approaches was used in order to limit the number of false-positives. In addition, Spearman correlation tests were performed to investigate significant correlations of individual gene expressions with individual aromatic DNA adduct levels.

For the real-time PCR analyses, {Delta}Ct values were calculated for each gene of interest (16). Three different approaches for normalizing these data were used as follows: (i) ß-actin was used as the housekeeping or ‘stably expressed’ gene, (ii) the average of three housekeeping or ‘stably expressed’ genes was used (beta-actin ACTB, beta-2-microglobulin B2M and cyclophilin A PPIA) or (iii) no housekeeping genes were used but the amount of input RNA for a reference and {Delta}Ct values were calculated directly for the gene of interest by subtracting the common reference Ct from the test sample Ct.

Gene ontologies were retrieved from http://www.ncbi.nlm.nih.gov/ with LocusLink or Gene. Analysis of biologically enriched pathways was performed in the Expression Analysis Systematic Explorer (EASE, http://david.niaid.nih.gov/ease) (17).

Cotinine measurements
Plasma was collected from EDTA blood by centrifugation. Plasma cotinine was measured by high performance liquid chromatography (HPLC) according to the protocol described by Van Vunakis et al. (18) with the exception of the presence of EDTA in the standard. The significance of the difference between smokers and non-smokers was analyzed with a Wilcoxon test for non-parametric analysis of 2-related samples in SPSS 11.5.

DNA adduct measurements
Leukocyte fractions were isolated from whole blood (after removal of the plasma layer) through erythrocyte lysis (lysis buffer containing 155 mM NH4Cl, 10 mM KHCO3, 10 mM EDTA:blood = 3:1) for 30 min at 4°C. Leukocytes were stored at –80°C until DNA isolation. DNA was isolated by standard phenol extraction (19). DNA adduct levels were measured by 32P-postlabeling with Nuclease-P1 enrichment as originally described by Reddy and Randerath (20) with minor modifications as described by Godschalk et al. (19). The significance of the difference between smokers and non-smokers was analyzed with a Wilcoxon non-parametric test for 2-related samples in SPSS 11.5.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 References
 
Differential gene expression in blood caused by cigarette smoking was studied in blood cells from monozygotic twins discordant in smoking behavior. The characteristics of the study population are shown in Table II.


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

  		Table II Study population characteristics

 
The average plasma cotinine level in non-smokers was significantly lower than in smokers, namely 5.6 ± 0.2 ng ml–1 versus 247.6 ± 54.7 ng ml–1 (mean ± SE, P = 0.008). Average DNA adduct levels were 1.2 ± 0.2 adducts per 108 nucleotides in the non-smokers group; significantly different (P = 0.04) from the average of 2.0 ± 0.2 adducts per 108 nucleotides in the smokers group.

For differential gene expression analysis, three RNA samples were hybridized on every cDNA microarray, namely of the smoker and non-smoker samples from the same twin pair and a common reference sample. Through a combination of several approaches, genes were identified with altered expression due to cigarette smoking. Either data of twin couples (the expression differences between a smoker and non-smoker) or data from each person individually (the expression difference between an individual and the common reference) were analyzed. Using the data of twin pairs, confidence analysis revealed 34 differentially expressed genes. Using the data from smoking and non-smoking individuals, 76 differentially expressed genes in smokers compared with non-smokers were revealed by a non-parametric Wilcoxon signed ranks test for 2-related samples. In addition, 44 genes showed a signal-to-noise ratio for the smokers versus the non-smokers of >0.5. The data of these tests are summarized in a VENN-diagram (Figure 1). The seven genes in the intersection of the diagram were identified by all three tests and appeared upregulated in the smokers.


Figure 1
View larger version (15K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 VENN diagram representing genes differentially expressed due to smoking as revealed by various statistical approaches. Tests included are the Wilcoxon non-parametric signed ranks test for two paired samples, confidence analysis (confidence level 99%, and expression difference >0.1 on log2 scale) and signal-to-noise ratio test. The seven genes in the intersection were revealed by all three approaches.

 
Second, genes that correlated with DNA adducts were identified by Spearman correlation tests. Seven genes that scored high on correlation coefficient are reported in Table III. For most genes, such as OGG1, gene expression changes appeared to be negatively correlated with DNA adduct levels.


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

  		Table III Correlations of gene expression differences of each individual relative to the common reference with DNA adduct levels (Spearman correlation test)

 
For a selection of genes that might be most interesting for future molecular epidemiology studies based on significance and biological relevance, the expression differences were verified by means of real-time PCR. For the cigarette smoking-related effects, SOD2, PTEN, ATF4 and MAPK14 were selected. OGG1, IL10, APP, AXIN1, PCK2, BAK and CSF1R were selected based on correlations with DNA adduct levels. Furthermore, plasminogen activator inhibitor 2 (SERPINB2) was included based on significant differential expression in previous in vitro studies with PBMC. Cytochrome P450 1B1 (CYP1B1) was included as this gene is known to be affected by, e.g. PAHs found in cigarette smoke and to be upregulated in smokers (21–23) and it also responded strongly to exposure to cigarette smoke condensate in vitro (unpublished data).

For SOD2, MAPK14 and ATF4, the microarray data were reproduced by real-time PCR, especially when applying the {Delta}Ct method (Table IV). For PTEN, real-time PCR results showed no significant difference between smokers and non-smokers, in contrast to the microarray data. For SERPINB2 and CYP1B1, however, real-time PCR showed differential expression in smokers versus non-smokers (P-values 0.021 and 0.008, respectively), where microarray analysis did not report an effect. When real-time PCR gene expressions for OGG1, IL10, APP, AXIN1, PCK2, BAK and CSF1R were correlated with DNA adduct levels, the correlations found for the microarray data with DNA adducts could not be reproduced (data not shown). Ontologies of the reported genes are shown in Table V.


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

  		Table IV Comparison of differential gene expressions between smokers and non-smokers as measured by two independent assays: cDNA microarrays and quantitative real-time PCR

 


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

  		Table V Gene ontologies

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
References
 
In nine monozygotic twin pairs discordant for smoking, we investigated whether cigarette smoking causes differential gene expression of toxicologically relevant genes in peripheral blood cells and whether expression levels correlate with a biomarker for early effects, i.e. DNA adducts. By analyzing cigarette smoke-induced differential gene expression in monozygotic twins, we reduced the impact of interindividual variability due to variation in genetic background. It also provided the opportunity to perform pair-wise analyses, adding statistical power to the study. The analyses revealed several genes to be reproducibly differentially expressed due to cigarette smoking.

Genes which were differentially expressed in smokers compared to non-smokers were identified by a combination of several approaches, namely non-parametric Wilcoxon test for 2-related samples, confidence analysis and signal-to-noise calculation. We choose for this combination of different approaches as it is likely to reduce the number of false-positive calls. The genes that were found by all three approaches (Figure 1) were considered the most discriminatively relevant genes. Resulting genes are activating transcription factor 4 (ATF4), mitogen-activated protein kinase 14 (MAPK14 aka p38), superoxide dismutase 2/Mn (SOD2), alpha 1-antitrypsin (SERPINA1), paxillin (PXN), myeloid cell differentiation protein-1 (MCL1) and phosphate and tensin homolog (PTEN). These genes were all upregulated in smokers as compared to non-smokers.

ATF4, SOD2 and MAPK14 were the genes differentially expressed by cigarette smoking, which were confirmed by real-time PCR. In addition to these genes, two other genes that did not show significance in the microarray results were included in the real-time PCR analyses, i.e. CYP1B1 and SERPINB2. In the real-time PCR analysis, CYP1B1 showed, for every twin pair, a higher expression in the smoker than in the non-smoker, which agrees with other studies (21–23). However, microarray analysis did not demonstrate a significant effect for CYP1B1 due to the relatively high standard error of the mean, which is probably due to the low mRNA abundance of this gene in peripheral blood cells. Also, SERPINB2 showed significance in the real-time PCR analysis, which agrees with previous in vitro experiments on human lymphocytes treated with cigarette smoke condensate (24). Upon microarray analysis, no significant effects were observed for this gene, again possibly due to the low abundance of its mRNA.

Based on the current data, the genes that we would propose to be suitable for biomonitoring in cigarette smoke-exposed humans are ATF4, SOD2, MAPK14, CYP1B1 and SERPINB2 since they showed to be reproducibly differentially expressed as a result of exposure to cigarette smoke. The ontologies for these genes are shown in Table V. CYP1B1 is a well known phase-1 biotransformation enzyme involved in the metabolism of carcinogens. Our data confirm previous studies, where increased CYP1B1 mRNA levels have already been found in smokers and lung cancer patients (21–23). Whether expression changes for this gene reflect exposure rather than genotoxic effect is not clear from the current data. SOD2 is known to be involved in (response to) oxidative stress and superoxide metabolism, scavenging intracellular oxidants, thus functioning as a defense mechanism. SOD2 has been found to be triggered by smoking as it has been found elevated in the lungs of smokers (25). MAPK14 has not directly been linked to exposure to cigarette smoke, but more generally to environmental exposures and stress conditions (26,27). MAPK14 signaling is suggested to play a crucial role in the regulation of cellular proliferation and differentiation and recent studies suggest a (main) role in protecting cells from apoptosis (28,29). Therefore, this gene might be hypothesized to mark an effect rather than an exposure. ATF4 activation has been reported in HepG2 cells after arsenite exposure (30), but no data are available for primary (human) cells or tissues from exposed populations. For SERPINB2, a role has been suggested in (malignancy of) endometrial cancer (31) and this gene is expressed in many tumors (32). The gene has also shown to confer resistance to apoptosis (33). Members of the SERPIN gene family of serine protease inhibitors has also been reported to be involved in COPD, emphysema and associated with lung cancer risk (34–36). Induction of differential gene expression of these five genes by cigarette smoking could therefore be the subject of further investigation in larger populations.

Our second approach was to identify genes for which the expression is associated with an established biomarker for early effect, like DNA adducts (see Table III). Although DNA adduct levels significantly differed between smokers and non-smokers (P-value = 0.04), a clear overlap existed between both groups. Therefore, gene expressions associated with DNA adduct levels, not necessarily have to discriminate smokers from non-smokers. Indeed, none of the genes differentially expressed between smokers and non-smokers were in the list of genes correlating with DNA adducts. Another explanation for lack of overlap between both gene lists might be that cigarette smoke is composed of much more carcinogens and chemicals than those that cause DNA adducts as measurable by 32P-postlabeling. The adduct levels found in the study population are well within range of values found in other studies (37). Unfortunately, when verifying the expression of the correlating genes by real-time PCR and correlating these expression data with DNA adducts no significances were found. Possibly this can be attributed to the small expression range of these genes, which in general is much smaller than that for the genes differentially expressed between smokers and non-smokers. This finding suggests lower specificity of DNA adducts as compared to differential gene expression for the measurement of effects of cigarette smoke exposure.

The genes we found to be differentially expressed in peripheral blood from cigarette smokers compared with non-smokers within monozygotic twin pairs, have not been reported by other similar studies in which differential gene expression has been measured in vivo in peripheral blood (9–11). In a study in leukocytes by Lampe et al. (10) differential gene expression in smokers and non-smokers was investigated and genes have been reported that correlated significantly with plasma cotinine levels. For comparison with this study, we investigated the correlation of CYP1B1 and IL-1ß (the only two genes reported by Lampe et al. that were also present on our microarrays) between the individual gene expression and individual cotinine levels. We did not find significant correlations of these genes with plasma cotinine levels in our data (Spearman test, P > 0.05). Reasons for this might be that Lampe et al. used a different platform and approach for selecting gene expressions and that they investigated a study population with a larger variation in subjects' age. Another recent study has investigated differential gene expression in blood from individuals exposed to metal fumes (11). Upon analyzing the subgroup of smokers in that study, it was suggested that cigarette smoking considerably affected whole-blood gene expression profiles. It was also proven that, with a paired-samples study design, it was possible to measure small changes in gene expression in a population study, which is in line with our findings. However, also Wang et al. used another platform and did not verify their main findings with an independent technique like real-time PCR. A study in benzene-exposed factory workers reported to have found a set of genes as potential biomarkers of benzene exposure (12). Although benzene is a component of cigarette smoke, it does not appear that after exposure to cigarette smoke benzene is an important component to alter gene expression, as no genes were reported that resemble the data of the present study. Again this study used another platform and a smaller study population.

Besides the above mentioned human studies, research on cigarette smoke-induced gene expression changes in animal models has been reported. In SKH-1 mice, cigarette smoke exposure induces MAPK expression (38,39), and expression of a SERPIN (trypsin-chemotrypsin-related serine protease) gene (38). In fetal liver of Swiss albino mice, exposure to environmental cigarette smoke induced expression of CYP genes as well as precursors for two SODs (39). Several MAPKs have been shown to be upregulated by cigarette smoke in the lungs of A/J mice (40).

When performing real-time PCR analyses of gene expression, the data can be normalized by either using the same amount of input RNA or by including housekeeping or stably expressed genes in the analyses (16). When verifying our microarray data by real-time PCR for the differentially expressed genes, consistency between both methods depended on the approach to normalize the real-time PCR data. The best reproducibility of the microarray data was achieved by the {Delta}Ct method (see Table IV), where the amount of input RNA was used for normalization, instead of one or more housekeeping genes.

The present study is one of the few studies that apply transcriptomics within the field of molecular epidemiology. Because of the limited history, the application of toxicogenomic tools in molecular epidemiology may be influenced by some limitations and pitfalls. For example, in toxicogenomic research microarrays are most often used in qualitative manner; to get an overview of expression changes of large numbers of genes. It is common practice to verify (the most relevant) findings of microarray experiments with real-time PCR, which represents a more quantitative assay (9,12,24). As is discussed above, the results of both techniques can deviate. In relation to this, is the issue of (choosing the right) ‘housekeeping’ gene(s) to be included in the real-time PCR analyses for normalization. Another pitfall or limitation of microarray methodologies can be the issue of multicomparison testing, which is prone to deliver false-positive results (12,41). This potential problem was dealt with in the current study by performing different statistical analyses and regarding as the most interesting set of genes, only those that showed significant results in all of the tests. Further developments in the field of toxicogenomics and transcriptomics, the technologies and data analyses, will provide ways to handle these limitations and pitfalls in the future.

Summarizing, this study on cigarette smoking-induced differential gene expression in peripheral blood from monozygotic smoking-discordant twins showed that gene expression analysis of many genes simultaneously yields a gene expression profile as promising biomarker of effect induced by cigarette smoking. Five genes were revealed that we would propose to be suitable for biomonitoring in a human population, namely SOD2, MAPK14, ATF4, CYP1B1 and SERPINB2.


   Acknowledgments
 
The authors thank Dr C. Derom and the nurses of the East-Flanders Prospective Twin Survey (EFPTS) for blood collections and assistance on mailing the questionnaires. This study was financially supported by the ‘Environment and Health’ research program of the Flemish government.

Conflict of Interest Statement. None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 

  1. Smith C.J., Perfetti T.A., Garg R., Hansch C. (2003) IARC carcinogens reported in cigarette mainstream smoke and their calculated log P values. Food Chem. Toxicol. 41:807–817.[CrossRef][Web of Science][Medline]
  2. Van Schooten F.J., Hirvonen A., Maas L.M., De Mol B.A., Kleinjans J.C., Bell D.A., Durrer J.D. (1998) Putative susceptibility markers of coronary artery disease: association between VDR genotype, smoking, and aromatic DNA adduct levels in human right atrial tissue. FASEB J. 12:1409–1417.[Abstract/Free Full Text]
  3. De Flora S., D'Agostini F., Balansky R., et al. (2003) Modulation of cigarette smoke-related end-points in mutagenesis and carcinogenesis. Mutat. Res. 523–524:237–252.
  4. van Delft J.H., Baan R.A., Roza L. (1998) Biological effect markers for exposure to carcinogenic compound and their relevance for risk assessment. Crit. Rev. Toxicol. 28:477–510.[CrossRef][Web of Science][Medline]
  5. Bonassi S. and Au W.W. (2002) Biomarkers in molecular epidemiology studies for health risk prediction. Mutat. Res. 511:73–86.[CrossRef][Web of Science][Medline]
  6. Hamadeh H.K., Amin R.P., Paules R.S., Afshari C.A. (2002) An overview of toxicogenomics. Curr. Issues Mol. Biol. 4:45–56.[Medline]
  7. Toraason M., Albertini R., Bayard S., et al. (2004) Applying new biotechnologies to the study of occupational cancer—a workshop summary. Environ. Health Perspect. 112:413–416.[Web of Science][Medline]
  8. Aardema M.J. and MacGregor J.T. (2002) Toxicology and genetic toxicology in the new era of "toxicogenomics": impact of "-omics" technologies. Mutat. Res. 499:13–25.[Web of Science][Medline]
  9. Wu M.M., Chiou H.Y., Ho I.C., Chen C.J., Lee T.C. (2003) Gene expression of inflammatory molecules in circula ting lymphocytes from arsenic-exposed human subjects. Environ. Health Perspect. 111:1429–1438.[Web of Science][Medline]
  10. Lampe J.W., Stepaniants S.B., Mao M., Radich J.P., Dai H., Linsley P.S., Friend S.H., Potter J.D. (2004) Signatures of environmental exposures using peripheral leukocyte gene expression: tobacco smoke. Cancer Epidemiol. Biomarkers Prev. 13:445–453.[Abstract/Free Full Text]
  11. Wang Z., Neuburg D., Li C., Su L., Kim J.Y., Chen J.C., Christiani D.C. (2005) Global gene expression profiling in whole-blood samples from individuals exposed to metal fumes. Environ. Health Perspect. 113:233–241.[Web of Science][Medline]
  12. Forrest M.S., Lan Q., Hubbard A.E., et al. (2005) Discovery of novel biomarkers by microarray analysis of peripheral blood mononuclear cell gene expression in benzene-exposed workers. Environ. Health Perspect. 113:801–807.[Web of Science][Medline]
  13. Derom C., Vlietinck R., Thiery E., Leroy F., Fryns J.P., Derom R. (2002) The East Flanders Prospective Twin Survey (EFPTS). Twin Res. 5:337–431.[CrossRef][Web of Science][Medline]
  14. Staal Y.C., van Herwijnen M.H., van Schooten F.J., van Delft J.H. (2005) Application of four dyes in gene expression analyses by microarrays. BMC Genomics 6:101.[CrossRef][Medline]
  15. Hasseman J. (2002) Standard Operating Procedure: Aminoallyl labeling of RNA for microarrays (eukaryotic). The Institute for Genomic Research. Available at: http://pga.tigr.org/sop/M004_1a.pdf [effective date 3/4/02].
  16. Livak K.J. and Schmittgen T.D. (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25:402–408.[CrossRef][Web of Science][Medline]
  17. Hosack D.A., Dennis G. Jr, Sherman B.T., Lane H.C., Lempicki R.A. (2003) Identifying biological themes within lists of genes with EASE. Genome Biol 4:R70.[CrossRef][Medline]
  18. Van Vunakis H., Gijka H.B., Langone J.J. (1993) Radioimmunoassay for nicotine and cotinine. IARC Sci. Publ. 293–299.
  19. Godschalk R.W., Maas L.M., Van Zandwijk N., van't Veer L.J., Breedijk A., Borm P.J., Verhaert J., Kleinjans J.C., van Schooten F.J. (1998) Differences in aromatic-DNA adduct levels between alveolar macrophages and subpopulations of white blood cells from smokers. Carcinogenesis 19:819–825.[Abstract/Free Full Text]
  20. Reddy M.V. and Randerath K. (1986) Nuclease P1-mediated enhancement of sensitivity of 32P-postlabeling test for structurally diverse DNA adducts. Carcinogenesis 7:1543–1551.[Abstract/Free Full Text]
  21. Chang T.K., Chen J., Pillay V., Ho J.Y., Bandiera S.M. (2003) Real-time polymerase chain reaction analysis of CYP1B1 gene expression in human liver. Toxicol Sci. 71:11–19.[Abstract/Free Full Text]
  22. Kim J.H., Sherman M.E., Curriero F.C., Guengerich F.P., Strickland P.T., Sutter T.R. (2004) Expression of cytochromes P450 1A1 and 1B1 in human lung from smokers, non-smokers, and ex-smokers. Toxicol. Appl. Pharmacol. 199:210–219.[CrossRef][Web of Science][Medline]
  23. Port J.L., Yamaguchi K., Du B., et al. (2004) Tobacco smoke induces CYP1B1 in the aerodigestive tract. Carcinogenesis 25:2275–2281.[Abstract/Free Full Text]
  24. van Leeuwen D.M., Gottschalk R.W., van Herwijnen M.H., Moonen E.J., Kleinjans J.C., van Delft J.H. (2005) Differential gene expression in human peripheral blood mononuclear cells induced by cigarette smoke and its constituents. Toxicol. Sci. 86:200–210.[Abstract/Free Full Text]
  25. Harju T., Kaarteenaho-Wiik R., Sirvio R., Paakko P., Crapo J.D., Oury T.D., Soini Y., Kinnula V.L. (2004) Manganese superoxide dismutase is increased in the airways of smokers' lungs. Eur. Respir. J. 24:765–771.[Abstract/Free Full Text]
  26. Shim M.J., Kim H.J., Yang S.J., Lee I.S., Choi H.I., Kim T. (2002) Arsenic trioxide induces apoptosis in chronic myelogenous leukemia K562 cells: possible involvement of p38 MAP kinase. J. Biochem. Mol. Biol. 35:377–383.[Web of Science][Medline]
  27. Kim Y.M., Reed W., Lenz A.G., Jaspers I., Silbajoris R., Nick H.S., Samet J.M. (2005) Ultrafine carbon particles induce interleukin-8 gene transcription and p38 MAPK activation in normal human bronchial epithelial cells. Am. J. Physiol. Lung Cell Mol. Physiol. 288:L432–L441.[Abstract/Free Full Text]
  28. Fang J.Y. and Richardson B.C. (2005) The MAPK signalling pathways and colorectal cancer. Lancet Oncol. 6:322–327.[CrossRef][Web of Science][Medline]
  29. Liao Y. and Hung M.C. (2003) Regulation of the activity of p38 mitogen-activated protein kinase by Akt in cancer and adenoviral protein E1A-mediated sensitization to apoptosis. Mol. Cell Biol. 23:6836–6848.[Abstract/Free Full Text]
  30. Ord D. and Ord T. (2005) Characterization of human NIPK (TRB3, SKIP3) gene activation in stressful conditions. Biochem. Biophys. Res. Commun. 330:210–218.[CrossRef][Web of Science][Medline]
  31. Osmak M., Babic D., Abramic M., Milicic D., Vrhovec I., Skrk J. (2001) Plasminogen activator inhibitor type 2: potential prognostic factor for endometrial carcinomas. Neoplasma 48:462–467.[Web of Science][Medline]
  32. Mengele K., Harbeck N., Reuning U., Magdolen V., Schmitt M. (2005) Tumor-associated prognostic factors of the plasminogen activator family: determination and clinical value of u-PA, t-PA, PAI-1, and PAI-2. Hamostaseologie 25:301–310.[Medline]
  33. Dickinson J.L., Bates E.J., Ferrante A., Antalis T.M. (1995) Plasminogen activator inhibitor type 2 inhibits tumor necrosis factor alpha-induced apoptosis. Evidence for an alternate biological function. J. Biol. Chem. 270:27894–27904.[Abstract/Free Full Text]
  34. Senn O., Russi E.W., Imboden M., Probst-Hensch N.M. (2005) a1-Antitrypsin deficiency and lung disease: risk modification by occupational and environmental inhalants. Eur. Respir. J. 26:909–917.[Abstract/Free Full Text]
  35. Kalsheker N.A. (1994) Molecular pathology of alpha 1-antitrypsin deficiency and its significance to clinical medicine. QJM 87:653–658.[Abstract/Free Full Text]
  36. Yang P., Bamlet W.R., Sun Z., et al. (2005) Alpha1-antitrypsin and neutrophil elastase imbalance and lung cancer risk. Chest 128:445–452.
  37. Nia A.B., Maas L.M., Brouwer E.M., Kleinjans J.C., Van Schooten F.J. (2000) Comparison between smoking-related DNA adduct analysis in induced sputum and peripheral blood lymphocytes. Carcinogenesis 21:1335–1340.[Abstract/Free Full Text]
  38. Izzotti A., Cartiglia C., Longobardi M., Balansky R.M., D'Agostini F., Lubet R.A., De Flora S. (2004) Alterations of gene expression in skin and lung of mice exposed to light and cigarette smoke. FASEB J. 18:1559–1561.[Abstract/Free Full Text]
  39. Izzotti A., Balansky R.M., Cartiglia C., Camoirano A., Longobardi M., De Flora S. (2003) Genomic and transcriptional alterations in mouse fetus liver after transplacental exposure to cigarette smoke. FASEB J. 17:1127–1129.[Abstract/Free Full Text]
  40. Izzotti A., Cartiglia C., Longobardi M., Bagnasco M., Merello A., You M., Lubet R.A., De Flora S. (2004) Gene expression in the lung of p53 mutant mice exposed to cigarette smoke. Cancer Res. 64:8566–8572.[Abstract/Free Full Text]
  41. Bretz F., Landgrebe J., Brunner E. (2005) Multiplicity issues in microarray experiments. Methods Inf. Med. 44:431–437.[Web of Science][Medline]
Received July 26, 2006; revised September 27, 2006; accepted October 7, 2006.


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

Home page
 CarcinogenesisHome page
D. M. van Leeuwen, M. Pedersen, P. J.M. Hendriksen, A. Boorsma, M. H.M. van Herwijnen, R. W.H. Gottschalk, M. Kirsch-Volders, L. E. Knudsen, R. J. Sram, E. Bajak, et al.
Genomic analysis suggests higher susceptibility of children to air pollution
Carcinogenesis, May 1, 2008; 29(5): 977 - 983.
[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:
28/3/691    most recent
bgl199v1
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 Leeuwen, D. M.v.
Right arrow Articles by Delft, J. H.M.v.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Leeuwen, D. M.v.
Right arrow Articles by Delft, J. H.M.v.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?