Toxicogenomics and Toxicogenetics

1
Toxicogenomics and Toxicogenetics

• Maastricht University
• J. van Delft, D. van Leeuwen, H. Ketelslegers, R.
  Vlietinck, J. Kleinjans

2
General concept
toxicogenomics
toxicogenetics
3
Goals

• Development, validation and application of
• biomarkers of effect as health indicator for
  exposure to carcinogenic compounds
• biomarkers for genetic susceptibility related to
  those indicators
• Based on the newest genomic technologies
• Gene expression profiles as biomarker for effect,
  by Danitsja van Leeuwen
• Multiplex genotyping as biomarker for genetic
  susceptibility, Hans Ketelslegers

4
Phases Toxicogenomics

• Studies to select genes using DNA microarrays
• In vitro studies in human peripheral blood cells
  exposed to carcinogenic compounds
• Small scale field study in monozygotic twins
  disconcordant in smoking
• Application in Environment Health field study
  of Luik III on adults by quantitative RT-PCR

5
Example of a DNA microarray
6
Human 600 Toxarrays of Phase-1 Molecular
Toxicology

• Gene Categories Types of Genes in Category
• Apoptosis Caspases, BAK, Bax, Fas, Cyclins, TNFs
• Cell Cycle Cyclins, DNA Binding Protein, Waf 1
• Cell Proliferation Kinases, Transcription
  Factors, Growth Factors and Receptors, Connexins
• DNA Damage/ Repair DNA Repair Genes, ERCCs,
  GADDs, Helicases, Topoisomerases
• Inflammation Serum Amyloids, Interleukins,
  Adhesion Molecules, Chemokines
• Metabolism P450s, Glucuronidation Enzymes,
  Glutathione Enzymes, Methyltransferases, Redox
  Enzymes
• Oxidative Stress O2 Response Genes, Superoxide
  Dimutase, Redox Enzymes
• Peroxisome Proliferators Peroxisomal Enzymes
• Transport Multi-drug Resistance Proteins, Organic
  Anion and Cation Transporters
• Cell-Environment Connexins, Integrins, Selectins,
  Cadherins

7
In vitro study in human peripheral blood cells

• Model carcinogenic compounds
• Cigarette smoke condensate
• Benzoapyrene
• Tabaco specific nitrosamine (NNK)
• 4-amino biphenyl
• H2O2
• Possible biomarker genes
• Deregulated by all compounds
• Correlating with DNA adducts

8
Deregulated by CSC
9
Deregulated by all compounds
10
Small scale field study

• Monozygotic twins discordant in smoking
• Total peripheral blood cells
• Analysis of
• Gene expression
• DNA adducts (post labelling)
• Plasma cotinin levels
• Data analyses of gene expression
• Smokers vs non-smokers
• Correlations with DNA-adducts
• Validation with RT-PCR

11
Differentially expressed genes in smokers vs
non-smokers
12
Validation with RT-PCR
13
Czech study

• Another relevant field study, though not related
  to current program
• Compared children from polluted versus clean area
  in Czech republic
• Identified
• Differentially expressed genes
• Genes correlating with micronuclei

14
Deregulated and correlating genes
15
Genes selected for field study
16
1) See NCBI at http//www.ncbi.nlm.nih.gov/Gene
17
Field study on elderly people

• aged 50-65 years, n 398
• RNA from total peripheral blood cells
• Quantitative RT-PCR of 8 genes vs 2 house keeping
  genes
• Reference RNA sample pool from 20 randomly
  selected individuals
• Compared data with
• COMET
• MN frequencies
• 8-OH-dGin urine
• Tumor markers in serum (p53, CEA, PSA)

18
Effect of region
Non-smokers All subjects
19
Effect of season
20
Comparison of regions
21
Correlations with effect biomarkers
22
Comparison with classical biomarkers

• Majority of gene expressions differed
  significantly between 2 or more regions
• Classical biomarkers did not always differ and if
  so, with lower significance
• Magnitude of differences
• gene expression 1.2 (DGAT2) 2.0 (ATF4)
• classical biomarkers 1.10 (COMET count) 2.43
  (COMET median)
• Smoking significantly affected
• CYP1B1 and ATF4
• MN, CEA and p53
• Correlations with exposure markers not yet done

23
Conclusions

• Gene expression profiling as possible biomarker
  has been developed and applied
• More in-dept analyses are required in order to
  establish relevance
• Exposure markers
• Effect markers
• Susceptibility markers
• Confounding factors
• ? Gene expression profiling is promising for
  molecular epidemiology on the risks of
  environmental exposures for humans

24
General concept
toxicogenomics
toxicogenetics
25
Phases Toxicogenetics

• Select genes and polymorphism to be included
• Develop and validate methods for multiplex
  genotyping
• Apply in Environment Health field studies of
  Luik III on newborns, adolescents, elderly

26
Selection criteria of genes and polymorphisms

• Genes must be relevant for endpoints / biomarkers
  in filed studies
• Asthma and allergy
• Cancer
• Polymorphisms must be relevant
• Highly frequent (gt5)
• Cause a phenotypic effect (proven or highly
  likely)

27
SNP Database Database 66 SNPs in 41 genes

• Biotransformation (Set 17)
• E.g. CYP1A1, -1A2, -1B1, GSTs, NATs, mEH etc.
• DNA repair (Set 2)
• E.g. XRCC, XPD, BRCA2, OGG1 etc.
• Oxidative stress related (Set 3)
• E.g. CAT, SOD, NQO, GPX etc.
• Inflammation (Set 45)
• E.g. Interleukins, TNFa, PAFAH etc.
• Apoptosis Cell Cycle control (Set 6)
• E.g. p53, p21, Cylin D, CDKs etc.

28
Examples
29
Examples for genotyping by Single Base Extension
30
Validation of genotyping method (1)
31
Validation of genotyping method(2)
32
Adolescent study

• Population /- 450 adolescents (age 16 years
  old)
• Biomarkers
• Effect Comet Analysis (DNA damage)
• Exposure 1-OHP (PAHs), PCBs, DDE, Cd, Pb
• Genotyping Biotransformation, DNA repair and
  oxidative stress related

33
Statistical approaches

• Univariate analyses e.g. Mann Whitney or
  Kruskall Wallis

34
Statistical approaches

• Multivariate analyses e.g. Multiple Linear
  Regression, Discriminant Analyses or Binary
  Logistic Regression

35
Exposure Marker
-
(Confounding) Effect of Smoking?

Remove Smokers from Analysis
In Non-Smokers
Total Population
Relationship Exposure with Effect Marker?
Dose-Response
2 groups based on Regression Line 0 1
Most important predictors?
Logistic Regression Group as Dependent Sex,
Cig/Day, Smoking Y/N, SNPs as Independents
36
Linear Relationships-Adolescents
37
Ethylbenzene CometLinear Regression
38
Ethylbenzene CometLogistic Regression
Catalase (p0.027) GSTT1 (p0.035)
P0.201
P0.131
39
Adult study

• Population /- 400 adolescents (age 65 years
  old)
• Biomarkers
• Effect Comet Analysis (DNA damage), 8-OHdG, PSA,
  CEA, p53
• Exposure 1-OHP (PAHs), PCBs, DDE, Cd, Pb
• Genotyping Biotransformation, DNA repair and
  oxidative stress related

40
Linear Relationships-Adults
41
Cadmium (Urine) 8-OHdG Linear Regression
non-smokers
42
Cadmium (Urine) 8-OHdGLogistic Regression
GSTT1 (p0.041)
P0.224
43
1-OH-pyrene (Urine) 8-OHdGLinear Regression
non-smokers
44
Adults 1-OH-pyrene (Urine) 8-OHdGLogistic
Regression
Gender CYP1A1m4 mEH3 (p0.001)
(p0.05) (p0.023)
P0.003
P0.035
P0.098
45
Comparison with classical biomarkers

• Majority of gene expressions differed
  significantly between 2 or more regions
• Classical biomarkers did not always differ and if
  so, with lower significance
• Magnitude of differences
• gene expression 1.2 (DGAT2) 2.0 (ATF4)
• classical biomarkers 1.10 (COMET count) 2.43
  (COMET median)
• Smoking significantly affected
• CYP1B1 and ATF4
• MN, CEA and p53
• Correlations with exposure markers not yet done

46
Conclusions

• Genetic polymorphisms affect susceptibility for
  effect biomarkers related to exposure
• Sensitive populations can be genotyping for
  relevant polymorphisms
• More in-dept analyses are required on order to
  establish relevance
• Interactions between genotypes
• Univariate analyses
• Effect of / interaction with smoking
• Relations with gene expression
• ? Genotyping enables to identify sensitive
  populations for specific exposure effect
  relations

47
Demonstrated the value for molecular epidemiology
toxicogenomics
toxicogenetics