Download risk assessment

Can “omic” data assist in environmental monitoring
and risk assessment of chemicals and particles?
Kevin Chipman
The University of Birmingham, UK
The early days and now
• Too much hype at first regarding the immediate
potential of “omics”...now a rebound
• Early problems around platform compatabilitynow largely resolved
• Insufficient datapoints and complexity of early
work (mainly due to costs)- now largely resolved
• Many early analyses were not sufficiently
objective and interpretation was flawedinformatics and pathway knowledge now starting
to resolve
How omics may help to address the
needs
• Need for improved predictivity for risk assessment
• Related to this we need to improve understanding of
modes of action and to derive diagnostic and predictive
biomarkers
• The omic technologies have the ability to aid both of these
areas
• Contribute to “weight of evidence” in toxicity assessment
– Identify possible mode(s) of action
– Identify and assess impacts on susceptible populations
and life stages
– Improve assessments for mixtures
– Dose-response assessment
– Exposure assessment
– Improving interspecies extrapolations
International Workshop to identify hurdles
Vancouver 2008 :
Article in press. Env. Health Perspect.2009
•
Regulatory bodies are already receptive e.g. USFDA Critical Path Initiative which encourages
innovation
• Reports of the National Research Council (USA)
:
1. “Toxicology testing in the 21st century” shows
the potential and the need to incorporate omics
into safety assessment
2. Committee on application of toxicogenomic
technologies and predictive toxicology and risk
assessment
Transcriptomics / Proteomics /Metabolomics
BIOINFORMATICS
Networks of responses to toxicants
provide a profile of response
reflecting the global status of tissue
Establish fingerprints
characteristic and predictive
of specific toxicities
Identify compensatory,
non-toxicity responses
Define the “systems toxicology”
of individuals and predict health status
Derive focussed (custom/
biomarker) expression arrays,
reporter gene assays etc.
Help to understand
RISK ASSESSMENT
MECHANISM of toxicity
for populations
Relevant to environmental standard setting : can help to validate and monitor
Toxicogenomics in non-model organisms
•Lack of genomic data
•Microarray studies do not have to be limited to a few genetic model organisms
•cDNA clones can be derived from conventional or subtracted EST libraries,
eg. suppressive, subtractive hybridisation (SSH)
•Automatic, practical annotation solutions for cDNA clones are available, eg.
Blast2GO, Partigene
•High throughput DNA sequencing (eg. 454, Solexa) can now allow swift
design of oligonucleotide arrays for non-model species (e.g. Craft and Chipman
Mussel programme)
•Non-pollutant environmental influences and inter-individual variation.
•Gene expression profiling should include laboratory exposures with the aim of
identifying ‘predictive gene sets’
•Clear experimental design and sufficient replication are essential
•Inter-individual variation can inform on the population structure
• Now some examples of the power
of the omics
Note: already successes
e.g. Mamoprint in medicine
e.g. Distinguishing between genotoxic and
nongenotoxic carcinogens
Flounder cDNA Microarrays as
Tools for the Identification of
Expression Changes in Gene Sets
Predictive of Exposure to
Pollutants.
Tim Williams, Steven George, Amer Diab,
Margaret Brown, John Craft, Ioanna
Katsiadaki, Fleur Geoghegan, Brett Lyons,
Victoria Sabine, Fernando Ortega, Francesco
Falciani and Kevin Chipman
Treated fish show many changes in liver gene expression
Which genes and which pathways are altered e.g. by Cd (prooxidant)??
2-fold up
Example scatter
plot
of Cd-treated
flounder
at day 1 vs
saline.
Apparent Induced Genes
2-fold down
1e4
Apparent
Repressed
Genes
1000
HSP30B
clones
1:1 ratio
100
Treatment f Cadmium d01 (control)
100
1000
X-axis: Cadmium Stage 2 (Default Interpretation) : Treatm...
Y-axis: Cadmium Stage 2 (Default Interpretation) : Treatm...
1e4
Colored by: Cadmium Stage 2, Default Interpretation (Trea...
Gene List:
Good Cd 1 (10664)
Cadmium treatment
(Williams et al EST 2006)
Day after Cd treatment
01 02 04 08 16
PfIL295A08 (Chaperonin
subunit
-Chaperonin
containingcontaining
TCP-1, TCP1,
subunit
6A 6A (zeta 1))
PfIL252A10 (chaperonin
containing
-Chaperonin
containing
TCP-1,TCP-1
delta delta)
PfIL009H04 (chaperonin
-Chaperonin
subunit 7subunit 7)
PfH70-g1
(heat
shock protein
-Heat
shock
protein
Hsp7070)
(*not statistically significant)
PfIL294C08
(Hsp70
binding protein)
-Hsp70
binding
protein
Contig417 (stress-induced-phosphoprotein
1 (Hsp70/Hsp90-organizi...
-Hsp70/Hsp90
organising protein SIP1/XST1
PfIL236A07
(DnaJ
(Hsp40)
homolog,
subfamily
C, member
-DnaJ (Hsp40) homolog, subfamily C member
1 1)
PfIL232G03 (Nucleophosmin
1)
-Nucleophosmin
1
PfIL228B06
(DnaJ
(Hsp40) homolog,
subfamily
C, member
-DnaJ
(Hsp40)
homolog,
subfamily
C, member
8 8)
PfIL255E12
shockgp96
protein gp96)
-Heat
shock(heat
protein
Contig204disulfide
(Protein disulfide
isomerase
protein)
-Protein
isomerase
relatedrelated
protein
Contig298 (ER-resident
chaperone
calreticulin)
-ER-resident
chaperone
calreticulin
Contig775
(heat
shock Hsp90
protein 90
beta)
-Heat
shock
protein
beta
Contig768
(Heat
shock
cognate
71
-Heat shock cognate Hsc71 kDa protein)
Contig490
(heat
shock HSP
protein
90 alpha)
-Heat
shock
protein
90HSP
alpha
Contig426
(low molecular
heat shock
protein
Hsp30B)
-Low
molecular
weightweight
heat shock
protein
Hsp30B
Contig1015
(DnaJhomolog
(Hsp40) homolog,
subfamily
B, member
-DnaJ
(Hsp40)
subfamily
B, member
1 1)
PfIL273G12 (Unknown
for MGC:65804))
-MGC65804,
similar(protein
to HSP90
co-chaperone P23
PfIL209A02 (unnamed
protein
product
CAG07414)
-CAG07414,
containing
DnaJ
domain
Contig85 (T-complex
polypeptide
-T-complex
polypeptide
1 1)
Single intraperitoneal injection of
flounder with a low dose of
cadmium (0.05 mg/kg) resulted in
hepatic gene expression changes
related to Chaperones
Protein synthesis
Protein degradation
Apoptosis
Immune
Biomarkers
Day after Cd treatment
01 02 04 08 16
A - Chaperones
PfIL295A08 (Chaperonin
subunit
-Chaperonin
containingcontaining
TCP-1, TCP1,
subunit
6A 6A (zeta 1))
PfIL252A10 (chaperonin
containing
-Chaperonin
containing
TCP-1,TCP-1
delta delta)
PfIL009H04 (chaperonin
-Chaperonin
subunit 7subunit 7)
PfH70-g1
(heat
shock protein
-Heat
shock
protein
Hsp7070)
(*not statistically significant)
PfIL294C08
(Hsp70
binding protein)
-Hsp70
binding
protein
Contig417 (stress-induced-phosphoprotein
1 (Hsp70/Hsp90-organizi...
-Hsp70/Hsp90
organising protein SIP1/XST1
PfIL236A07
(DnaJ
(Hsp40)
homolog,
subfamily
C, member
-DnaJ (Hsp40) homolog, subfamily C member
1 1)
PfIL232G03 (Nucleophosmin
1)
-Nucleophosmin
1
PfIL228B06
(DnaJ
(Hsp40) homolog,
subfamily
C, member
-DnaJ
(Hsp40)
homolog,
subfamily
C, member
8 8)
PfIL255E12
shockgp96
protein gp96)
-Heat
shock(heat
protein
Contig204disulfide
(Protein disulfide
isomerase
protein)
-Protein
isomerase
relatedrelated
protein
Contig298 (ER-resident
chaperone
calreticulin)
-ER-resident
chaperone
calreticulin
Contig775
(heat
shock Hsp90
protein 90
beta)
-Heat
shock
protein
beta
Contig768
(Heat
shock cognate
-Heat
shock
cognate
Hsc71 71 kDa protein)
Contig490
(heat
shock HSP
protein
90 alpha)
-Heat
shock
protein
90HSP
alpha
Contig426
(low molecular
heat shock
protein
Hsp30B)
-Low
molecular
weightweight
heat shock
protein
Hsp30B
Contig1015
(DnaJhomolog
(Hsp40) homolog,
subfamily
B, member
-DnaJ
(Hsp40)
subfamily
B, member
1 1)
PfIL273G12 (Unknown
for MGC:65804))
-MGC65804,
similar(protein
to HSP90
co-chaperone P23
PfIL209A02 (unnamed
protein
product
CAG07414)
-CAG07414,
containing
DnaJ
domain
Contig85 (T-complex
polypeptide
-T-complex
polypeptide
1 1)
Day after Cd treatment
01 02 04 08 16
Day after Cd treatment
01 02 04 08 16
Day after Cd treatment
01 02 04 08 16
E - Protein degradation
F Cytoskeleton
-Alpha-tubulin
Contig305 (alpha tubulin)
-Dynein
light chain
2
Contig626
(Dynein
light chain 2, cytoplasmic)
-Syndecan
2
PfIL316A06
(Syndecan 2)
-Actin
related protein
3 homolog
PfIL240E02
(ARP3
actin-related protein 3 homolog)
Contig475
(cysteine
and 2glycine-rich protein 2)
-Cysteine
and glycine
rich protein
-Thymosin
beta (T
4 hymosin beta-4)
Contig789
Contig98
and protein
spindle
associated protein 1; nucleolar
-Nucleolar
and(nucleolar
spindle associated
1 /ANKT
PfIL207C10
-Annexin
max 3 (annexin max3)
PfIL242A06 (pfBF2D7, beta actin)
-Beta-actin
PfIL242G01
(microtubule-based
motor protein (FKIF2))
-Microtubule
based
motor protein FKIF2
PfIL231A02
(adducin 3 (gamma);)
-Adducin
3 gamma
Day after Cd treatment
01 02 04 08 16
G - Apoptosis
PfIL300G11
(similar
programmed
cell death 6)
-Similar
to Programmed
Cellto
Death
6
Contig465c (Cytochrome c)
-Cytochrome
PfIL255H02
(Reticulon 1)
-Reticulon
1
Contig1023 (Thioredoxin-like
1)
-Thioredoxin-like
1
Contig779 enaphalopathy
(Ethylmalonic
-Ethylmalonic
1 encephalopathy 1)
PfIL236H02
-APG
16L beta (AP G16L beta)
Contig717 protein
(anticoagulant
-Anticoagulant
C precursor protein C precursor (PROC))
Contig196
protein1(candidate of metastasis 1))
-p8
/ Candidate (p8
Of Metastasis
Contig605
(Thymidine
phosphorylase precursor)
-Thymidine
phosphorylase
precursor
Contig376 (COMM domain
-COMM-domain-containing
3 / BUP containing 3; BUP protein;)
PfIL224F06
(Integral
membrane protein 2B)
-Integral
membrane
protein 2B
Contig269
(similar to direct IA P binding protein with low PI)
-Similar
to DIABLO
PfIL209H11
(Survivin 1)
-Survivin
1
C - Protein Synthesis
-PolyA binding protein (P. platessa)
PfIL233E09ribosomal
(40S ribosomal protein
S3)
-40S
protein
S3
Pa003 (Ribosomal
Protein S3A) protein S3a (P. americanus)
-40S
ribosomal
Contig557 ribosomal
(60S ribosomal proteinprotein
L3)
-60S
L3
PfIL288F02 (eukaryotic translation
initiation factor
4E binding4E
proteinbinding
3)
-Translation
initiation
factor
protein 3
Contig800 ribosomal
(40S ribosomal proteinprotein
S14)
-40S
S14
Contig795 ribosomal
(40S ribosomal proteinprotein
S27-2)
-40S
S27-2 -Density regulated protein
Contig770 ribosomal
(40S ribosomal proteinprotein
S18)
-40S
S18
Contig600 ribosomal
(40S ribosomal proteinprotein
S16)
-40S
S16
Contig748 ribosomal
(60S ribosomal proteinprotein
L7a)
-60S
L7a
Contig684 ribosomal
(40S ribosomal proteinprotein
S2)
-40S
S2
Contig642 ribosomal
(40S ribosomal proteinprotein
S5)
-40S
S5
Contig392 ribosomal
(60S ribosomal proteinprotein
L13A)
-60S
L13A
Contig576 ribosomal
(40S ribosomal proteinprotein
S3a)
-40S
S3a
Contig474 (Eukaryotic translation
elongation factor
1-delta) 1-delta
-Translation
initiation
factor
Contig582 ribosomal
(60S ribosomal proteinprotein
L18)
-60S
L18
Contig48 (40S
ribosomal protein protein
Sa)
-40S
ribosomal
Sa
Contig346 (cDNA to
clonetranslation
hab41f08.x1 similar toinitiation
TIF3 subunit 9) factor 3 subunit 9
-Similar
Contig786 ribosomal
(60S Ribosomal proteinprotein
L17)
-60S
L17
Contig2 (40S
ribosomal protein S12)
-40S
ribosomal
protein S12
Contig661 (Eukaryotic translation
elongation factor
1 beta 2)1-beta 2
-Translation
initiation
factor
Contig618 ribosomal
(40S ribosomal proteinprotein
S10)
-40S
S10
Contig577 ribosomal
(40S ribosomal proteinprotein
S21)
-40S
S21
PfIL318H11ribosomal
(60S ribosomal protein
L10)
-60S
protein
L10
Contig602 ribosomal
(60S ribosomal proteinprotein
L5)
-60S
L5
Contig527 ribosomal
(60S ribosomal proteinprotein
L12)
-60S
L12
PfIL211H04 (similartoto S-phase
kinase-associated
protein
1A isoform b)
-Similar
S-phase
kinase
associated
protein 1A b
PfIL277F10
S phase transition
1; hm:zehn1143)
-G1
to(G1Stophase
transition
1
PfIL257A11 (Siah-interacting proteinprotein
(Sip-prov))
-Siah-interacting
Contig317 (NHP2 non-histone
-NHP2-like
1 chromosome protein 2-like 1)
PfIL203B10 (ADP-ribosylation factor-likefactor-like
6 interacting protein) 6 interacting protein
-ADP-ribosylation
PfIL289C07 (Pescadillo)
-Pescadillo
PfCF1H9 (Density-regulated
protein; smooth
muscle cell associated protei...
-Density
regulated
protein
PfIL265G02ribosomal
(60S ribosomal protein
L39)
-60S
protein
L39
Contig542 ribosomal
(60S ribosomal proteinprotein
L19)
-60S
L19
PfIL011D06ribosomal
(60S ribosomal protein
L24)
-60S
protein
L24
PfG6D-l2 (TIF3 / P42) initiation factor 3 /P42
-Translation
Contig292 (Eukaryotic translation
initiation factor
4A)
-Translation
initiation
factor
4A
Oxidative stress
Protein transport
Cytoskeleton
Cell cycle
Inflammation
PfIL315E08 (zgc:56219)
-zgc:56219/Ubiquitin
conjugating enzyme E2Q
(prosome, macropain) subunit, alpha type 4)
(proteasome
PfIL272D08
-Proteasome
subunit
alpha type 4
(proteasome
PfIL314H06
-Proteosome
subunit
alpha type 6 (prosome, macropain) subunit, alpha type 6)
1 subunit isoform 2)
PfIL258H08
-Proteosome
alpha(proteasome
1subunit isoform alpha
2
(prosome, macropain) 26S subunit, non-A TP ase, 12)
Contig443
-Proteasome
26S(proteasome
subunit 12
regulatory subunit S10B )
(26S protease
Contig1004
-26S
proteasome regulatory
subunit S10b
subunit N3)
(proteasome
Contig254
-Proteasome
subunit
N3
PfIL308H05
-Proteasome
26S, (proteasome,
regulatory subunit 626S, non-A TP as e regulatory subunit 6)
beta-subunit C5 (Proteasome (Prosome, macropain) sub...
(Proteasome
PfIL273D04
-Proteasome beta subunit C5
(Proteasome delta)
Contig347
-Proteasome
delta
protease
Contig714
-26S
proteasome(26S
regulatory
subunit 8regulatory subunit 8)
(Proteasome
Contig579
-Proteasome
subunit
beta type 3 subunit beta type 3)
enzyme E 2 variant 2 (Ube2v2))
(ubiquitin-conjugating
PfIL277B08
-Ubiquitin
conjugating
enzyme E2 variant 2
B - Oxidative Stress
-Paraoxonase
2
PfPARA-o2 (Paraoxonase
2)
-Catalase
PfIL265A03 (Catalase)
-MAP
kinase
interacting
serine/threonine
kinase
2 2)
Contig456
(MAP
kinase-interacting
serine/threonine
kinase
-Glutathione
reductase
(*not statistically significant)
PfGR-1 (Glutathione
reductase)
Contig691 (selenoprotein
M)
-Selenoprotein
M
Contig406 (Glutaredoxin)
-Glutaredoxin
PfIL256D11water
(selenide
water 2
dikinase 2)
-Selenide
dikinase
Contig1002 binding
(Seleniumprotein
binding1protein 1)
-Selenium
Contig658glutathione
(Plasma glutathione
peroxidase
precursor)
-Plasma
peroxidase
precursor
Contig419H-1
(Ferritin H-1)
-Ferritin
Contig416M
(Ferritin, middle subunit (Ferritin M))
-Ferritin
Contig123
(copper/zinc
superoxide dismutase)
-Cu/Zn
Superoxide
dismutase
Contig459 (Peroxiredoxin
(Thioredoxin
peroxidase) (NKeF))
-Peroxiredoxin
(Thioredoxin
peroxidase)
Contig444 (Carbonyl
reductase
1 (20 beta hydroxysteroid dehydrogen...
-Carbonyl
reductase
1
Day after Cd treatment
01 02 04 08 16
Day after Cd treatment
01 02 04 08 16
Day after Cd treatment
01 02 04 08 16
H - Cell Cycle
-NHP2-like 1
-G1 to S phase transition 1
-Chaperonin subunit 7
-Ran nuclear GTPase
-GTP-binding protein like 1
PfIL233F02
(centromere/kinetochore
protei n zw10 homolog)
-Centromere/kinetochore
protein zw10 homolog
PfIL235F11
(S eptin 5)
-Septin 5
PfIL282H07
-Cyclin H (Cyclin H)
Contig317 (NHP2 non-histone chromosome protein 2-like 1)
PfIL277F10 (G1 to S phase transition 1; hm:zehn1143)
PfIL009H04 (chaperonin subunit 7)
PfCF 2C3 (Ran protein - member of Ras superfamily, nuclear GTP-ase.)
PfIL306C03 (GTP-binding protein like 1)
Day after Cd treatment
01 02 04 08 16
I - Immune and Inflammation
-Alpha-1-microglobulin
precursor
Contig596
(alpha-1-microglobulin/bikunin
precursor)
Contig405
chain
constant region)
-IgM heavy (IgM
chainheavy
constant
region
-Interleukin (interleukin
8
Contig197
8)
-MHC II invariant
Contig773
(MHC chain
II invariant chain)
PfIL291G12
(melanoma
ubiquitous
-Melanoma ubiquitous
mutated
proteinmutated
MUM1 protein MUM1)
PfIL249C03
(Similar
to Scytokine
mall inducible cytokine)
-Similar to small
inducible
Contig281
(Small
inducible cytokine)
-Small inducible
cytokine
PfIL263G01
antigen
receptor
beta-chain
-T-lymphocyte(T-lymphocyte
antigen receptor
beta chain
constant
region constant region 2)
PfIL248D01
(Similar tocomponent
complement
component C8 gamma)
-Similar to complement
C8 gamma
PfIL263E02
(clonemembrane
WA8-6 sim
to integral
membrane protein 2A)
-Similar to integral
protein
2A
PfIL140C08
(cDNA
-Similar to TNF
13b clone JFConA 894F, S im to TNF 13b)
PfIL295H08
(src family
associated phosphoprotein
2)
-Src family associated
phosphoprotein
2
PfIL273F10
(tumor necrosis
factor
-TNF ligand superfamily
member
14 ligand superfamily, member 14)
Contig707
(Cysteine-rich
protein 1)
-Cysteine rich
protein 1
PfIL259D05
(class
I helical
cytokine
-Class 1 helical
cytokine
receptor
26 receptor number 26)
Contig80
(immunoglobulin
-Immunoglobulin
light chain light
L2 chain L2)
Contig471
(C-type
-C-type lectin
domainlectin
1 1)
Contig529
(complement
component C9)
-Complement
component C9
PfIL267A02
asialoglycoprotein-binding
protein 1)
-Macrophage (Macrophage
asialoglycoprotein
binding protein 1
PfIL264F04
(P lasma
protease
C1 inhibitor precursor)
-Plasma protease
C1 inhibitor
precursor
D - Protein Transport
Contig369 (translocon-associated
gamma)
-Translocon
associated proteinprotein
gamma
PfIL294D10
(ADP-ribosylation
-ADP
ribosylation
factor 5factor 5)
PfIL230G07
(SEC22,
vesicle trafficking
protein-like
-SEC22,
vesicle
trafficking
protein-like
1B 1B)
PfIL245F05
-TMED
7 (transmembrane emp24 protein transport domain contai...
PfIL288A03
-TIMM
23(Translocase
homolog of inner mitochondrial membrane 23 homol...
PfIL312A07heavy
(Clathrin,
heavy polypeptide (Hc))
-Clathrin,
polypeptide
PfIL309B02 (TRAP-like
protein precursor)
-TRAP-like
protein precursor
PfIL223C05
(Protective
protein
for
beta-galactosidase)
-Protective protein for beta galactosidase
PfIL203B10
(ADP-ribosylation
factor-like
6 interacting
protein)
-ADP
ribosylation
factor-like
6 interacting
protein
PfIL256F09 (adaptor-related
protein complex
3, sigma
-Adaptor-related
protein complex
3 sigma
1 1 subunit)
PfCF2C3
(Ran protein
- member of Ras superfamily, nuclear GTP-ase.)
-Ran
nuclear
GTPase
PfIL206D015a
(syntaxin 5a)
-Syntaxin
PfIL306C03 (GTP-binding
protein
-GTP-binding
protein-like
1 like 1)
Day after Cd treatment
01 02 04 08 16
J - Biomarkers
-Glutathione-S-transferase theta 3
-Microsomal glutathione-S-transferase 3
-Cytochrome P450 CYP2K6
-Glutathione-S-transferase A
-Cytochrome P450 CYP2F2
-Metallothionein (* not statistically significant)
-Microsomal glutathione-S-transferase 1
Contig501
(cytochrome
P450 1A CYP1A)
-Cytochrome
P450 CYP1A
Contig458
(Vitellogenin)
-Vitellogenin
A
Contig401
(choriogenin
L)
-Choriogenin
L (* not statistically
significant)
Contig1030
(V A
itellogenin)
-Vitellogenin
PfIL260C08 (Glutathione S -transferase, theta 3)
Contig531 (microsomal glutathione S-transferase 3)
Contig723 (Cytochrome P450 monooxygenase CYP2K6)
Contig367 (glutathione S-transferase)
Contig218 (Cytochrome P450 2F2)
Contig22 (metallothionein)
PfIL254F12 (Microsomal glutathione S -transferase 1)
protein ANKT)
A - Chaperones
Day after Cd treatment
01 02 04 08 16
B - Oxidative Stress
-Paraoxonase
2
PfPARA-o2 (Paraoxonase
2)
-Catalase
PfIL265A03 (Catalase)
-MAP
kinase
interacting
serine/threonine
kinase
2 2)
Contig456
(MAP
kinase-interacting
serine/threonine
kinase
-Glutathione
reductase
(*not statistically significant)
PfGR-1 (Glutathione
reductase)
Contig691 (selenoprotein
M)
-Selenoprotein
M
Contig406 (Glutaredoxin)
-Glutaredoxin
PfIL256D11water
(selenide
water 2
dikinase 2)
-Selenide
dikinase
Contig1002 binding
(Seleniumprotein
binding1protein 1)
-Selenium
Contig658glutathione
(Plasma glutathione
peroxidase
precursor)
-Plasma
peroxidase
precursor
Contig419H-1
(Ferritin H-1)
-Ferritin
Contig416M
(Ferritin, middle subunit (Ferritin M))
-Ferritin
Contig123
(copper/zinc
superoxide dismutase)
-Cu/Zn
Superoxide
dismutase
Contig459 (Peroxiredoxin
(Thioredoxin
peroxidase) (NKeF))
-Peroxiredoxin
(Thioredoxin
peroxidase)
Contig444
(Carbonyl
reductase
1
(20
beta
hydroxysteroid dehydrogen.
-Carbonyl reductase 1
Cu exposure of Stickleback shows similar hepatic expression changes
in cholesterol biosynthesis pathway genes to Wilson’s disease, a
copper accumulation disorder
HMG-CoA synthase
(down 6 fold)
HMG-CoA reductase
(down 5 fold)
Mevalonate kinase
(down 1.5 fold)
Isopentenyl-diphosphate delta
isomerase (down 5 fold)
Farnesyl diphosphate synthase
(down 2.5 fold)
Stickleback Exposure
to 128mg/L Cu
Mouse model of Wilson’s disease
(ATP7B -/-)
Huster et al., 2007 JBC
FLOUNDER FIELD SITES
Q. If fish provided “blind” could genomics identify
sampling location
and if so are the gene patterns reflective of pollutant
exposure e.g. oxidative stress??
Tyne
(Heavy industrial)
Howden, Team
Alde
(rural)
Outer Elbe
(Cuxhaven
Helgoland)
Elbe Harbour
(industrial, harbour, canal
Brunsbuttel)
Predicting Site Membership
by genetic algorithm
GALGO (NERC Project)
Stage 1
Create Initial Population
of Chromosomes
with random genes
Artificial
chromosome
Examples of genes induced at polluted sites
Stage 2
Evaluate all chromosomes
using the fitness function
Population
and
fitness value
attached
Stage 3
Stage 7
if some fitness
>= goal
Stage 4
no
Generate new popultation:
Reproduce chromosomes
proportionally to its
fitness
yes
SELECT
new
population
Stage 5
Random crossover between
chromosomes pairs
crossover
Stage 6
Random mutatations on new
population
mutation
Trevino V. & Falciani F. Bioinformatics. 2006
1;22 :1154-6.
Phase 2
Phase I
UDPGT
Aldehyde
GST
dehydrogenase
Alcohol dehydrogenase
CYP1A
Proliferation marker
CYP2F
PCNA
CYP3A
Protein degradation
CYP8B
Proteasome subunits
Oxidative Stress
Catalase
Superoxide dismutase
Chaperones
Calreticulin
Haem biosynthesis
Coproporphyrinogen oxidase
Could a subset of combined
stress-response genes help to
classify the environmental
samples?
Aroclor
Lindane
Time course chemical
treatments
PFOA
Cadmium
TBHP
3 MC
Set of stress-related genes
up & down regulated.
Use of genetic algorithm analysis using combined stress
responsive genes
Merge all the IDs that were selected in each of representative models
for each treatment: 98 IDs
Class Confusion ( 1 Models)
[project]:knn-3K1LeuclideanD-0,1-loocv
(NA)
0.018
Tyne T
0.046
0
0
0
0
0
0
0
0
0.956
0.037
0.844
0.13
Elbe
Predicted Class
0.013
Elbe H.
0.004
0.866
0.104
Tyne H
Heligolang
NB This does not
necessarily implicate
these pollutants as
being responsible but
it helps to identify
stress response
differences at the sites
0.172
0
0.078
0
0
0
0.016
0
0
0.044
0
0
0
0
0
0
0
Alde
29/29
Samples
Elbe
4/4
Samples
Elbe H.
21/21
Samples
0.897
1
0.75
0.989
1
0.994
0.015
0
1
0
0.75
0.011
0.897
Alde
Sensit
Specif
Heligolang
Tyne H Tyne T
5/5
10/10
8/8
Samples
SamplesSamples
0.956 0.844
0.984 0.971
0.866
0.97
Examples:
X1
X2
X3
X4
X5
X6
X7
X8
X9
X10
X11
X12
X13
X14
X15
X16
X17
X18
X19
X20
X21
X22
X23
X24
X25
X26
X27
X28
X29
X30
X31
X32
X33
X34
X35
X36
X37
X38
X39
X40
X41
X42
X43
X44
X45
X46
X47
X48
X49
X50
X51
X52
X53
X54
X55
X56
X57
X58
X59
X60
X61
X62
X63
X64
X65
X66
X67
X68
X69
X70
X71
X72
X73
X74
X75
X76
X77
Original Class (sorted)
Contig620: Retinol-binding protein II, cellular (CRBP-II)
Conclude: A small number of
stress response genes are
predictive of site of origin !
Contig442: Glutamate carboxypeptidase (Darmin)
Contig665: Ependymin
Modeling
We are using linkage networks (Dr
Francesco Falciani)
to integrate gene expression and
metabolomics (Dr Mark Viant)
with traditional measures.
Linkage shows where data are
related.
This simplified example was
generated using ARACNE and
cytoscape, employing 50 selected
nodes.
Interestingly traditional markers
(in blue) (eg condition factor) are
linked both to transcripts (purple)
and to metabolites (red).
We can focus on particular areas to visualise which genes are linked, in
terms of expression profiles.
Here NF kappa B is centre of an extensive hub and linked to survivin (an anti-apoptotic gene) and vitellogenin.
Survivin
Vitellogenin
NF kappa B
Using class-prediction algorithms (eg
GALGO) we identified the areas of the
network containing genes and metabolites
most predictive of (differentially polluted)
sampling sites (red)
So, starting
to see
connectivity
between
components
of the
network and
the field
These overlap with an area of the
network populated by genes related
to metabolism and energy production
(in green)
Application of “open” technologies
to the study of nanomaterials
• In ecotoxicology, genomics has a major value in
assessing novel agents and also mixtures of
contaminants for which we do not know
appropriate end points or mechanisms. It provides
a non-biased, global approach.
• A highly appropriate application therefore is the
assessment of the effects of nanomaterials, the
products and by-products of which enter the
environment as mixtures with largely unknown
effects.
Omics, monitoring and safety assessment
– Elucidate mechanisms of toxicity (e.g distinguish genotoxic vs
nongenotoxic carcinogens)
– Provide more informative batteries of biomarkers
– Create practical assays e.g. real-time PCR, custom arrays, reporter
assays, ELISAs
– Focus on PROCESSES disturbed rather than single gene products
–
–
–
–
Characterise responses of sentinel species to ‘new’ pollutants
Assess the effects of mixtures
Inform on the basis of population susceptibility to toxicants
Provide detailed case studies of specific sites
 A major challenge will be the ability to distinguish between
adaptive vs toxic responses and the effective use of these markers
in risk assessment. We need to discover patterns of change that are
diagnostic and predictive
Challenges & Recommendations
Research
– Needs:
• Linking genomic changes to adverse outcomes (AOP)
• Interpreting genomic information for risk assessment
• Training risk assessors and managers to interpret and
understand genomics data in the context of a risk assessment
• Development of technical framework for analysis and
acceptance criteria for “omic” information for scientific and
regulatory purposes
Adapted from Bill Benson 2008