Expression Toxicogenomics
Gene Studies:
Pattern of genes expressed in a cell is characteristic of its current state Many differences in cell state or type are correlated with changes in mRNA levels of many genes Expression patterns of many previously uncharacterized genes may provide clues to their possible function by comparison with how known genes act Gene expression data can be combined with metabolic schemas to understand how pathways are changed under varying conditions (i.e., mechanisms of action)
Biology & Functional Genomics:
Having information on the level of expression of all genes that are present in a genome is very good But genetics is the study of the interactions of these individual genes in an organism Want to study complex interplay of all genes simultaneously This requires high-throughput and large-scale technologies to study gene expression of all genes simultaneously
Toxicogenomics is a new scientific field that elucidates how the entire genome is involved in biological responses of organisms exposed to environmental toxicants/stressors
http://www.niehs.nih.gov/nct/home.htm
http://www.niehs.nih.gov/multimedia/qt/ntc/ntcaltcaption.mov http://video.niehs.nih.gov:7075/ramgen/ntc/ntc-eng03.rm
DNA Microarray Technology
DNA complementary to genes of interest is generated and laid out in microscopic quantities on solid surfaces at defined positions cDNAs (from mRNA) from samples are eluted over the surface complementary DNA binds Presence of bound DNA is detected by fluorescence following laser excitation
Has many potential applications:
Studying changes in expression of genes over time, between tissues, and disease states Identification of complex genetic diseases Drug discovery and toxicology studies Mutation/polymorphism detection (SNPs) Pathogen analysis
Microarray Experiments:
analysis of gene expression
Genome-wide (e.g., yeast) Partial selection of known/unknown genes Analyze cell signaling networks (e.g., cell-cycle genes) Determine effects of various exposures/conditions Predict/discover function of unknown genes Compare normal to abnormal (e.g., tumor cells):
Analyze expression patterns Novel gene association/discovery Divide tumors into sub-classes Determine effects of treatment
Scientific Areas in Toxicogenomics
Disease Mechanisms: application of gene expression profiling technologies to define the mechanistic underpinnings of environmentally related diseases; genetic and environmental components of disease, elucidation of disease pathways and networks, and development of disease models. Susceptibility: individual and population susceptibilities to exposure and disease as derived from genetic and environmental analysis and integration; identification of gene targets and factors mediating susceptibility, and gender-, strain- and species-susceptibility. Comparative Genomics: comparative and integrated responses of organisms to environmental stimuli; cross-species comparisons of biological responses to environmental factors at the gene, transcription, and protein level and their integration in model organisms; conserved biological components, pathways and responses to environmental factors; and computational tools to support comparative toxicogenomics. Predictive Toxicology: development and application of gene expression, proteomics and metabolomics technologies in predictive toxicology; development of model systems and research tools, and linkage of predictive responses to disease phenotype.
Classical Microarray Experiments
Normal vs Disease
Example: Analysis of expression patterns in cancer -Pattern of gene expression-networks -Novel gene association/discovery
Molecular Classification
Example: Comparison of Breast Tumors -Samples classified into subtypes
Genome-Wide Analysis
Example: Genome-wide expression in S. cerevisiae
Microarray Technique
Microarray: A substrate with bound capture probes Capture probe: An oligo/cDNA with gene (DNA sequence) of interest Generic experimental steps: 1. Fabrication: Photolithography Affymetrix (one-color array) >40,000 genes Printing Agilent (two-color arrays) ~45,000 genes Spotting In-house (two-color arrays) <24,000 genes 2. Target Generation from a sample of interest: One color (biotin labeled cRNA, phycoerythrin-streptavidin detection) Two color (Cy3 and Cy5 cDNA labeling) 3. Hybridization 4. Analysis Scanning of array Amount of hybridized target is assessed Statistical interrogation of the data
Fabrication of Spotted color oligo/cDNA arrays two- Array Construction
Oligos (20-70 bp) or cDNAs
96-well plate
384-well plate
printed on a glass slide
Two-color oligo/cDNA arrays
200 m
150 m
Oligo/cDNA Arrays Method
Two mRNA sources to be compared are labeled with fluorescent probes: Cy3 (green) used for one sample (e.g., control) Cy5 (red) used for the other (e.g., treatment) Probes are mixed and washed over the microarray (hybridization) Each probe is excited using a laser, and its fluorescence (R and G) at each element detected with a scanning confocal microscope The ratio between the signals in two channels (R:G) is calculated for each array spot Ratios of intensity of Cy5/Cy3 probes is a reliable measure of the abundance of specific mRNAs in each sample compared to control
Two-color oligo/cDNA arrays
Two-color oligo/cDNA arrays
mRNA from Sample 1 mRNA from Sample 2
Scan and quantitate gene expression levels
Affymetrix GeneChip array format
One-color chip (biotin labeled cRNA, phycoerythrin-streptavidin detection) Oligonucleotide probes are synthesized in situ on the chip Semiconductor photolithography technology is used to synthesize oligos in situ on a glass substrate 1 cm2 Masking technology is used to build up oligonuclotides Oligonucleotides corresponding to 5, middle and 3 sections of a gene of interest are used Oligonucleotides corresponding to a perfect match and single mismatch are used to separate signal from noise Hybridization is measured with a laser, quantified and stored as a raw value for comparison to data from another chip
Oligonucleotide one color Arrays (Affymetrix)
Array preparation
mRNA reference sequences Perfect Match/Mismatch probe sets
in situ synthesis by photolitography
One Gene
Target preparation
Biotin-labeled cRNA Double stranded cDNA Total mRNA Cells/Tissue samples
Ratio array1/array2
Affymetrix GeneChip detection principle
16-25 mer oligos, perfect match/mismatch, 8-20 per gene of interest
target mRNA sequence
DNA probe pairs reference sequence perfect match oligo mismatch oligo perfect match probe cells Fluorescence Intensity Image
mismatch probe cells
Affymetrix GeneChips in your lab
Catalog Arrays
Arabidopsis ATH1 Genome Array C. elegans Genome Array Drosophila Genome Array E. coli Antisense Genome Array Human Genome Focus Array Human Genome U133 Set Human Genome U95 Set Mouse Expression Set 430 Murine Genome U74v2 Set P. aeruginosa Genome Array Rat Genome U34 Set Rat Neurobiology U34 Array Rat Toxicology U34 Array Test3 Array Yeast Genome S98 Array
CustomExpress Arrays Made to Order Arrays DNA Analysis Arrays
CYP450 Assay GenFlex Tag Array HuSNP Mapping Assay p53 Assay
CustomExpress Advantage Arrays CustomExpress Premier Arrays
~2
in
~1 in
Cost: $250-$1000 apiece
Fluidics station (stain/wash)
Scanner
Analysis Software
Courtesy of Affymetrix
"Mainstream" arrays:
Oligonucleotide based (two-color):
Home-made vs. Agilent Capacity: up to 25,000 genes/targets
Oligonucleotide based (one-color):
Affymetrix Capacity: up to 400,000 targets
Need for much greater number of targets on a single array:
minimize energy and materials needed for array production/processing faster/cheaper wider dynamic range, increased selectivity and sensitivity
Other platforms that receive attention:
Fiberoptic microarrays Electrically addressable arrays Electrokinetic microarrays
Fiberoptic Microarrays (ILLUMINA/Tufts U.)
Microarray Element Size Packing Density 3.1 m
5 x 104/mm2
Nanoarray 200 nm
4 x 106/mm2
From: www.illumina.com
Electrically Addressable Array (Motorola)
Bioelectronic detection proceeds via a sandwich hybridization assay, wherein three critical components (capture probe, target, and signaling probe) are each present in the cartridge. The signaling probe serves to label the target upon hybridization. Electrons flow to the electrode surface only when the target is present and specifically hybridized to both signaling probe and capture probe. The current generated by this system is measured and interpreted by the eSensor DNA Detection Reader and Software.
From: www.motorola.com/lifesciences
Cy3 Cy5
Cy5 Cy3
Cy5 log2 Cy3
200 10000 50.00 5.64 4800 4800 1.00 0.00 9000 300 0.03 -4.91
Slide courtesy of C.M. Perou
NAME BC/FUMI0 BC/FUMI4 BC/FUMI4 BC601B-A BC601A-B BC/FUMI1 BC/FUMI2 BC/FUMI2 BC/FUMI1 BC/FUMI1 BC102B-B BC/FUMI2 BC/FUMI3 BC/FUMI3 BC/FUMI1 BC/FUMI1 0.242 1.21 -0.253 -0.841 -0.423 -0.363 -0.852 -1.383 -2.642 0.501 -0.25 -0.605 -0.636 0.229 -0.626 adipose prote differentiation-related 0.908 0.485 -0.397 -0.767 -0.886 -0.251 -0.683 0.057 -0.317 -1.2 0.125 -0.536 -0.248 -0.365 plasminogen activator, urokinase re plasminogen activator, urokinase re 0.4635 0.3545 -0.8975 -1.23 -0.8335 0.0175 -1.002 0.1555 -0.4325 -1.008 -0.1785 -0.7445 -0.1485 0.0555 0.2055 0.551 0.151 -0.422 0.007 -0.638 0.087 -0.689 -0.91 -0.853 0.052 -0.492 -0.201 -0.152 -0.368 -0.741 coronin, actin binding protein, 1C A **coatomer protein complex, subun -1.061 -0.8655 -0.1235 -0.9895 0.3815 -0.4955 -0.2775 -0.1465 -1.109 -0.8635 0.2615 -0.0905 -0.3225 -0.6035 0.0195 -0.9345 coactosin-like protein R78490 -0.8835 -0.4545 0.2375 -1.177 0.2155 -0.2975 -0.9385 -0.2815 -1.494 -0.5985 0.4095 -0.3465 0.2185 -0.1345 -0.2895 -0.5525 folylpolyglutamate synthase R44864 0.686 1.583 1.313 0.048 -0.272 -0.143 -0.394 0.423 -0.445 -0.854 0.322 -0.03 -0.412 0.214 -1.098 -0.175 -0.18 1.155 1.575 -1.635 0.355 0.295 -0.805 0.135 -2.145 -0.955 0.575 0.735 -0.435 -0.855 -0.8 -1.705 lysozyme (renal amyloidosis) N639 chemokine (C-C motif) receptor 1 AA036881 0.524 1.233 -1.459 -0.095 -0.122 -0.196 0.101 -0.942 -0.2 -0.133 -0.549 -0.763 -0.059 interferon, gamma-inducible protein -0.181 -0.062 0.37 0.064 0.418 -0.33 -0.098 -0.289 -1.042 -0.332 0.907 1.056 -0.8 -0.193 -0.789 -1.25 cystatin B (stefin B) H22919 -0.188 -0.489 -0.603 0.074 -0.212 -0.295 -0.54 -0.535 -0.453 -0.479 -0.021 0.291 -0.651 -0.536 -0.401 -0.511 cathepsin S AA236164 -0.791 0.334 -0.316 0.723 -0.46 0.39 -0.452 -0.413 1.063 -0.849 -1.088 -0.94 -1.291 small inducible cytokine A2 (monoc 0.2665 0.2955 0.5315 -0.1285 0.4255 -1.099 -0.7265 -0.6035 -1.052 -1.438 0.1355 0.0365 -0.4335 0.0875 -1.218 -0.7785 0.483 0.348 0.575 -0.685 0.971 -0.335 -0.222 -0.116 -1.644 -0.66 -0.322 0.885 -0.08 -0.02 -0.441 -0.51 natural killer cell transcript 4 AA458 superoxide dismutase 2, mitochond 0.431 0.301 -0.836 0.519 -0.492 -0.834 -0.86 0.781 0.005 -1.163 -1.283 -0.969 -0.586 -0.6835 0.4865 0.6925 -0.7895 -0.6005 -0.5815 0.4995 0.0165 0.3755 -0.1225 -1.129 -1.137 -0.6935 superoxide dismutase 2, mitochondrial AA4877 0.3185 transforming growth factor, beta-ind 0.0235 0.6525 -0.3785 -0.5505 -0.3675 -0.4755 -0.1105 0.3435 0.0785 -0.4735 0.7925 1.532 -0.3355 -0.0885 0.2495 -0.1985 -1.122 -1.412 -1.275 -1.764 -0.611 1.259 -1.25 -0.76 -2.159 -1.72 -1.017 -0.972 -0.715 -0.543 -0.658 -0.818 glycine dehydrogenase (decarboxy syndecan 2 (heparan sulfate proteo -1.828 -1.7 -1.409 -1.964 -0.975 1.516 -1.24 -1.75 -2.219 -2.477 -1.08 0.29 -1.641 -2.045 -0.315 -1.356 -1.726 -1.892 -1.568 1.528 -1.346 -2.157 -3.114 -3.146 -0.943 0.236 -1.349 -1.674 -0.416 -1.557 glutathione S-transferase pi R33642 chitinase 3-like 2 AA668821 -0.771 -1.436 -1.454 -0.813 -1.578 0.312 -0.167 0 -0.469 0.129 -0.566 -0.489 nuclear factor I/B W87528 0.464 -1.314 -0.187 -1.429 -0.189 0.551 -1.94 -1.372 -2.152 -1.825 -0.441 -0.928 0.316 -1.188 ras homolog gene family, member -1.382 -0.471 -0.421 0.304 -0.448 -0.805 -0.945 -0.737 -1.222 -0.915 -0.713 -0.167 0.09 1.074 -0.393 ras homolog gene family, member -1.311 -0.763 -0.61 0.198 -0.764 -0.391 -0.867 -1.469 -1.106 -0.486 -0.778 -0.579 0.812 0.348 -0.222 -0.965 -0.571 -0.304 -0.328 -0.417 -0.518 -0.473 -0.973 -0.94 -0.926 -1.153 -0.462 -0.683 0.828 0.347 **zinc finger, DHHC domain containing 5 AA4 keratin 5 (epidermolysis bullosa sim -0.309 -0.485 -0.748 -0.909 -0.403 -0.127 -0.371 -0.778 -1.596 -1.787 -0.782 0.242 -0.559 -0.804 0.79 0.374 keratin 5 (epidermolysis bullosa simplex, Dowl -0.655 -2.421 0.301 0.689 -0.38 -0.131 -1.647 -1.396 0.248 -1.118 -0.389 -1.423 1.963 -0.068 keratin 17 AA026100 -0.593 -2.294 0.181 -0.45 0.457 -1.132 -0.754 -2.708 -0.641 -0.148 -0.201 0.161 2.264 1.758 -0.523 -0.763 -0.726 -0.155 -0.401 -1.8 -1.591 -1.789 -1.076 -0.929 -1.132 -1.051 -0.24 tripartite motif-containing 29 AA055 pleiomorphic adenoma gene-like 1 AA463204 -0.7035 -0.5595 -0.7765 -0.2835 -0.1885 -1.466 -2.035 -0.1475 -0.7075 -0.4025 -1.054 0.3535 -0.5835 secreted frizzled-related protein 1 AA002080 -1.951 -2.022 -1.982 0.069 -0.117 -1.543 -2.996 -2.657 -0.275 -1.187 -0.262 -0.688 3.135 0.295 Homo sapiens cDNA FLJ11796 fis, clone HEM -1.425 -0.74 -0.798 0.243 -0.225 -0.061 -0.957 -0.001 -0.491 -0.28 0.595 -0.721 ESTs AA074677 -0.411 -0.412 -0.879 -0.78 -0.401 -0.135 -0.508 -2.237 0.077 -0.72 -1.057 -1.301 pellino homolog 1 (Drosophila) W86 -0.3805 -1.159 -0.6945 -0.3935 -0.1785 -0.3665 -0.3835 -0.2825 0.1245 0.3185 0.2735 -1.329 -0.9455 -1.313 -0.4235 -0.887 -2.32 0.16 -1.65 -1.54 -1.065 1.453 -1.55 -2.859 -0.04 matrix metalloproteinase 7 (matrilys moesin R22977 0.452 -0.759 -0.433 -0.691 0.148 -0.538 -0.28 -0.478 -0.477 0.019 0.062 -0.001 0.259 -0.24 -0.314 -1.302 -0.5695 -1.843 -0.8355 -0.3325 -0.7305 0.2015 -0.3825 -0.2335 -0.4605 -1.181 -0.6875 -0.3315 0.2825 -0.0605 prion protein (p27-30) (Creutzfeld-J -0.8095 chitinase 3-like 1 (cartilage glycoprotein-39) A 1.474 1.071 0.678 0.987 -1.357 -2.185 -1.619 3.517 -0.465 -1.549 -1.699 -1.262 annexin A8 AA235002 -0.55 -0.832 0.209 0 -0.576 -0.199 -1.046 -0.454 -0.221 0.134 -0.015 0.619 0.519 -0.078 -0.939 -1.002 0.058 -0.058 -0.158 -1.65 -0.794 -1.612 0.17 1.318 0.404 -0.312 -0.039 hypothetical protein FLJ20481 N32 ADP-ribosylation factor-like 7 N353 -0.9415 -0.0585 -0.3685 -0.9365 -0.2155 0.0715 -0.2825 -0.5505 -1.107 -0.5855 0.2285 -0.2475 0.1635 -0.1405 cystatin A (stefin A) W72207 -0.532 -0.941 0.909 1.783 0.164 -0.10...