Transcript View PowerPoint Presentation - SUPERB INTERNET CORPORATION

Toxicogenomics: Using
cDNA Microarrays to Detect
Effects of Environmental
Exposures
Research Triangle Park, NC
October 15, 2002
Mary Jane Cunningham, Ph.D.
Director of Discoveries
Molecular Mining Corporation
55 Rideau St.
Kingston, ON
Canada K7K 2Z8
613-547-9752
E-mail: [email protected]
Definition of Toxicogenomics
The use of genomic technologies
for the measurement and analysis
of gene and protein expression
in assessing the risk of
new chemical entities (NCEs)
Three Different Approaches
To study sex differences:
 Indirect gene expression
-Gene sequencing
-Electronic comparison of cDNA library abundances
 Direct gene expression
-Gene expression microarrays
 Protein expression
-2D gels annotated by mass spectroscopy
Gene Sequencing
cDNA Library Technologies
 Oligo dT: Standard method of library production
 Normalized rare clone-biased: Modified Soares* protocol used to
bias library towards low abundance transcripts (75-90+% rare),
increases % unique singletons in rat by 2X
 Subtracted: Modified Soares* protocol used to target differentially
expressed transcripts between 2 tissues, increases % unique
singletons in rat by 2X
 Hybridized: Top 100 highest abundance genes & 1st strand cDNA
probes used to screen out high abundance clones; preferential
enrichment of middle abundance clones; increases %unique
singletons but only 1.5X
*Soares et al, PNAS 91:9228 (1994);
Bonaldo et al, Genome Research 6:791 (1996)
ZooSeqTM Analysis
 Subsetting analyses of hepatotoxin-treated
cDNA libraries:
-APAP timecourse [12h, 24h, d3, d7, d14, d28]
-B(a)P timecourse [12h, 24h, d3, d7, d14, d28]
-CLO timecourse [12h, 24h, d3, d7, d28]
-4-AAF, ANIT, CCl4, Fenofibrate, Hydrazine
(6 h, 24h)
-Female vs. Male [B(a)P, CLO]
-Subtracted CLO library
-Subtracted CLO libraries [female vs. male]
ZooSeqTM Analysis (Continued)
Known Genes (with abundant transcripts present):
APAP:
AP56=Acetaminophen binding protein
Cytochrome P450 IIIA
B(a)P:
Cytochrome P450c (3-MCA induced)
Cytochrome P450d (3-MCA induced)
CLO:
Cytochrome P450 4A (LA-omega)
Cytochrome P450 4A2
Carnitine octanoyltransferase
Acyl-CoA oxidase
Peroxisomal bifunctional enzyme (PBE)
CCl4:
Alpha-1 acid glycoprotein
Fenofibrate:
Peroxisomal bifunctional enzyme (PBE)
QAbun
8
7
16
6
34
6
11
6
21
9
19
Possible Toxicity Markers
252 Incyte Uniques
(QAbun>3)
APAP only
11
4-AAF
ANIT
CCl4
Fenofibrate
Hydrazine
60
CLO
CLO
93only
93
CLO female
1
B(a)P only
35
BP female
3
Possible Sex Markers
Known Genes-female:
Comp.
CLO
femaleGene ID Description
B(a)P fvsm g2224669 KIAA0364
20fvsm g2224669 KIAA0364
CLO
CLO fvsm
testis-specific farnesyl
pyrophosphate synthetase
B(a)P fvsm g57670
ribonuclease inhibitor
CLO fvsm g286245 oligomycin sensitivity
conferring protein (OSCP)
B(a)P fvsm g5007031 transgelin
B(a)P fvsm g203773 CYP2E1
CLO fvsm g4894587 growth hormone receptor
binding protein
QAbun
79
10
g204091
7
6
6
3
3
3
Unique Genes-female:
B(a)P fvsm
B(a)P fvsm
B(a)P fvsm
CLO fvsm
1
2
3
4
Incyte Unique
Incyte Unique
Incyte Unique
Incyte Unique
13
10
7
3
Summary
 RNA transcripts of genes known to be involved in the
response of B(a)P, APAP and CLO were observed with
high abundance in rat liver cDNA libraries.
 Unique genes whose functions are not yet known were
also observed.
 In comparisons of cDNA libraries made from both male
and female rats, transcripts from both known and
unknown genes were detected in higher abundance in
female rats than in males.
cDNA Microarray Technology
 Array Design: cDNAs are deposited onto glass surface
 Sample Prep: PolyA mRNA isolated from cells/tissues is reverse
transcribed to cDNA from control (red label) and treated (green
label) samples
 Hybridization: Fluorescently-labeled cDNA hybridized to cDNAs on
array
Red=down-regulated
Green=up-regulated
Yellow=control and treated probes co-hybridize at same
frequency-no change in expression
 Rat GEMTM has approx. 7400 cDNAs deposited
(predominantly rat liver and kidney genes and genes which are
related to toxic response)
Original References:
Science 270: 467-470 (1995)
Genome Res. 6: 639-645(1996)
cDNA Microarray Technology
+
=
OUTPUT: Ratio of Expression
Rat Hepatocyte Cell Lines
Cell Line
Sex
Strain
Isolated
Morphology From
BRL3A
Female
Epithelial
Buffalo
1968
1968
Year
Isolated
Clone 9
Male
Epithelial
SpragueDawley
H4-II-EC-3
Male
Epithelial
AxC
1961
Epithelial
Buffalo
1974
McA-RH7777 Female
Experimental Plan
Cell Line
State of
Growth
Initial
Passage
Passage
at
Harvest
BRL3A
Preconfluent
Unknown
15
3.8(10)7
BRL3A
Confluent
Unknown
12
5.4(10)7
Clone 9
Preconfluent
16
33
6.0(10)7
Clone 9
Confluent
16
33
2.9(10)8
H4-II-EC-3
Preconfluent
Unknown
15
1.7(10)8
H4-II-EC-3
Confluent
Unknown
15
7.4(10)8
70
79
4.7(10)8
McA-RH7777 Preconfluent
Cell
Number
Photomicrographs
Preconfluent or
Randomly-Proliferating Culture
Confluent or
Contact-Inhibited Culture
Proliferating vs. Confluent
Hierarchical Clustering:
Distance Metric - Euclidean, Complete Linkage
Filter - Balanced Differential Expression >4-Fold
Results from GeneLinker TM Platinum
Female Cell Line vs. Male Cell Lines
Hierarchical Clustering:
Distance Metric - Euclidean, Complete Linkage
Filter - Balanced Differential Expression >9-Fold
Results from GeneLinker TM Platinum
Summary
 Similar gene expression profiles from randomlyproliferating and contact-inhibited cultures were
observed for Clone 9 and H4-II-EC-3 cells.
-BRL3A cells resulted in a slightly different profile.
 In comparisons of expression profiles from Clone9, H4II-EC-3, BRL3A and McA-RH7777 cells, differences
were observed in the female rat-derived cell lines versus
the male rat-derived cell lines.
 McA-RH7777 cells were observed to have the highest differences
followed by BRL3A.
In Vivo 30-Day Study in Rat
*Genotoxic:
Benzo(a)pyrene [BP]
*Nongenotoxic:
Acetaminophen [APAP]
Clofibrate [CLO]
*Control vehicle (DMSO)
APAP
CLO
Days of Dosing-BP
0.5 1 3
7
14
28
Days Post Last Dose
Dosing: BP-3 doses/wk, 2wks
APAP, CLO-single dose
Timepoints: 12h, 24h, d3, d7, d14 and d28
Tissues harvested: liver, kidney, lung, brain, pancreas and spleen
Hepatotoxins
Acetaminophen:
HO-
-NHCOCH3
• metabolized by P450
• detoxification products:
-glucuronides
-sulfates
-glutathione conjugates
• active metabolite: NAPQI
Benzo(a)pyrene: •metabolized by P450 and epoxide hydrolase
to two active metabolites:
-Bay area diol epoxide [BPDE]
-K-region diol epxide [BP-4,5-oxide]
•activation/detoxification
•persistance of DNA adducts in liver out to
56 days
Hepatotoxins (Continued):
Clofibrate:
•Peroxisome Proliferator Activating Receptor (PPAR)
Inducer
-primarily PPAR-a
Cl-stimulates:
b-oxidation of fatty acids
insulin sensitivity
glucose metabolism
-hyperlipidemia treatment
-OC(CH3)2COOCH2CH3
•Induces Hepatocellular Carcinoma in Rodents
-nongenotoxic transformation
stimulates cell growth
suppresses apoptosis
Animal Monitoring
 Physical Signs:
-mild tremors, decreased activity, rigid body tone
-mortality=17% CLO, 6% APAP
 Gross pathology at necropsy:
-Liver: APAP>CLO>BP
-Spleen: BP>CLO>APAP
 Liver to body weight ratios:
-significant increase with BP over time course
 Serum enzyme levels:
-AST and ALT: increases in APAP, BP and CLO treatments
(mostly at 12h, 24h)
GEMTM Results
12 hr. B(a)P-Treated Male Rat Liver
GEM Results for Time Courses
• Total number of genes:
APAP:
269
BP:
146
CLO:
271
• Total number of Incyte uniques:
APAP:
69 (26%)
BP:
37 (25%)
CLO:
58 (21%)
Reproducibility Examples
First Pair of cDNAs:
Day
0.5
1
3
7
14
28
BP1
1.2
11.7
1.1
-1.1
10.8
1.7
BP2
1.0
11.8
1.0
-1.3
12.9
1.8
CLO1
2.9
7.5
-1.1
1.1
----1.2
CLO2
2.9
7.1
-1.2
1.1
----1.2
APAP1
11.1
29.1
6.0
1.1
2.2
1.3
APAP2
3.9
36.8
5.0
1.3
2.4
1.4
Second Pair of cDNAs:
Day
0.5
1
3
7
14
28
28
BP1
1.9
5.2
1.4
2.6
1.9
3.7
BP2
2.0
3.8
1.4
2.9
2.0
4.6
CLO1
2.9
7.5
-1.1
1.1
----1.2
CLO2
2.9
7.1
-1.2
1.1
----1.2
APAP1
3.0
1.7
5.0
1.3
1.8
1.7
APAP2
3.7
2.7
6.0
2.0
2.6
2.8
Top 44 Expressed Genes – 12 Hr.
Balanced Differential Expression
Possible Toxicity Markers
Genes
Distribution of N-Fold Scoring Genes
APAP
B(a)P
186 (59%)
69
(22%)
172 (54%)
89 (28%)
5
(2 %)
9
(3 %)
98 (31%)
10
(3 %)
CLO
60 (19%)
36 (11%)
Expression Patterns

3 compounds have different known modes of action:
-Resulted in different gene expression profiles
-Some common genes induced but at different times
Expression Patterns
 Isozymes were expressed but did not always result
in same pattern of expression
Individual Rats vs. Pooled Sample
Hierarchical Clustering:
Distance Metric - Euclidean, Average Linkage
Filter - Balanced Differential Expression >6-Fold
DMSO-1
DMSO-2
DMSO-3
DMSO-Pool
CLO-1
CLO-2
CLO-3
CLO-Pool
Results from GeneLinker TM Platinum
Sex Differences
Hierarchical Clustering:
Distance Metric - Euclidean, Average Linkage
Filter - Balanced Differential Expression >8-Fold
Results from GeneLinker TM Platinum
Sex Differences-Proteomics
•Approximately 72,000 features curated
•967 features differ from normal by > 2-fold
• 599 from clofibrate treatment
- approximately 50% annotated to date
Peroxidases / peroxisome proteins
- confirm catalase, glutathione peroxidase
- confirm glutathione sulfotransferases
- confirm enol-CoA hydratase
Metabolic enzymes
- confirm cytochrome P450
Adducts
- none cataloged to date
Sex differentials
- confirm estrogen metabolism enzymes
Performed in collaboration with Oxford GlycoSciences
Summary
 Different gene expression profiles were observed from
liver tissues of rat treated with B(a)P, APAP and CLO.
 Several genes whose function is not yet known (unique
genes) were shown to be highly expressed in tissues from
all 3 compounds.
 A wide range of gene expression was seen in arrays
from individual rat samples versus arrays using pooled
samples
 Pooled samples tended to give an average expression
profile-an average of each individual rat’s value
 Differences in gene expression profiles were also found
between female and male rats.
 Similar differences were also detected in proteomic
profiles.
Acknowledgements
 cDNA Library Construction
Laura Stuve
Anne Curtis
Laura Kamigaki
Glenn Fu and group
• GEM Design, Array Manufacture and Hybridization
Gary Zweiger
Scott Panzer
Olga Bandman
Jeff Seilhamer
Microarray Division
 Data Analysis Methods
Roland Somogyi
Shoudan Liang
Stefanie Fuhrman