109 May 2012 2nd McKim Workshop, Baltimore 2012 Identification of non- genotoxic carcinogens based...

1 09 May 2012

2nd McKim Workshop, Baltimore 2012

Identification of non-genotoxic carcinogens based on mechanisms

Jan van Benthem

National Institute for Public Health and the Environment

Laboratory for Health Protection Research

Bilthoven, The Netherlands

2nd McKim Workshop, Baltimore 2012 | 09 May 2012

2

National Institute for Public Health and the Environment (RIVM)


- no further testingnot genotoxic

In vitro genotoxicity tests

In vivo genotoxicity tests

+consider whether in vivo test is requiredcheck bioavailabilitycheck available dataconsider proper in vivo (follow up) testconsider integration into other toxicity tests

Considered genotoxic in somatic cells

+

- no further testingnot genotoxic

Check for information on a genotoxic hazard to germ cells

Carcinogenicity test


3

Detection of non-genotoxic carcinogens



• Carcinogenicity test

• (Sub)-chronic repeat dose toxicity tests 90-day toxicity test

6-month toxicity test

12-month toxicity test

• Negative in genotoxicity tests


4

REACHREACH

Since 1981 difference is made between new and existing chemicals (marketed before September 1981).

About 99% of the total number of compounds marketed is part of this new policy

In 1981, 100106 (existing) compounds were registered (30000 >> 1 tpa) Since 1993, 141 existing compounds have been tested

Of about 80% of these compounds toxicological data are lacking

Registration, Evaluation, Authorisation and Restriction of CHemicals

Registratie, Evaluatie en Autorisatie van CHemische stoffen

Registrement, Evaluation et Autorisation des produits CHimiques




5

REACH

Major aims:To ensure a high level of protection of human health and the environment• Safety assessment for new and existing chemicals

• Classification and labelling• Risk Assessment

• In less time, money and experimental animals

• Promotion of tests which do not use experimental animals.• Optimal use of in vitro tests• Stimulation of development of new in vitro tests• Minimalization of test strategies

New:• In future it is the responsibility of the industry and down stream users

to obtain toxicological knowledge bout chemicals. Every compound is notified only once (industrial collaboration)




6

REACH data requirements at different tonnage levels

Acute toxicity

In vitro geno-toxicity

Local toxicity

28 day toxicity

90 day toxicity

Repro-ductive toxicity

Develop-mental toxicity

Carcino-genicity

2 genera-tion study

In vivo geno-toxicity

Acute toxicity


Local toxicity

28 day toxicity

90 day toxicity


Develop-mental toxicity

2 genera-tion study

In vivo geno-toxicity

Acute toxicity


Local toxicity

28 day toxicity

90 day toxicity


Acute toxicity


Local toxicity

No studies required< 1 tpa

1 - 10 tpa

10 - 100 tpa

> 1000 tpa

100 – 1000 tpa




7

September 11, 2004: Testing ban on finished cosmetic products in the EU. Marketing ban on cosmetic products and ingredients tested on animals outside of the EU where validated alternative tests exist.

March 11, 2009 Testing ban cosmetic ingredients or formulations in the EU. Marketing ban cosmetic products and ingredients tested on animals with the exception of repeated-dose toxicity, reproductive toxicity and toxicokinetics.

March 11, 2013 Marketing ban cosmetic products or ingredients tested on animals, irrespective of the availability of alternative non-animal tests.

The 7th Amendment requires cosmetics manufacturers and distributors to provide certain product information for the safety of the end user.

The Cosmetics Directive and its Seventh Amendment




8

Risk associated with non-genotoxic carcinogensRisk associated with non-genotoxic carcinogens

Mechanisms of non-genotoxic carcinogens and importance of a weight of evidence approach

Lya G. Hernandez, Harry van Steeg, Mirjam Luijten, Jan van Benthem

Mutation Research 682, 94–109 (2009)

[email protected]




9

Number non-carcinogens in IARC 1, 2A and 2B

IARC group total carc non-carc percentage

1 77 64 13 16.9

2A 57 55 2 3.5

2B 237 207 30 12.7

total 371 326 45 13.8




10

Margin of Exposure



Margin of Exposure (MoE) = Margin of Safety (MoS)

The ratio: Most relevant exposure from animal studies

(Estimated) human exposure

Potential hazard:Carcinogens MoE < 10,000 Non-genotoxic carcinogens MoE < 100 for consumersNon-genotoxic carcinogens MoE < 50 for workersCosmetics in Europe: MoS < 100


11

Approaches



Rodent dose response carcinogenicity study data of the Gold database

using LTD10 calculated from TD50

(Sub)-chronic dose response toxicity studies from literature using NO(A)EL

RfD using EPA/IRIS data

Human exposureusing data from the open literature

vs


12

Reference dose

A dose (with uncertainty factors) of a daily exposure to

human population that is likely to be without appreciable

risk of deleterious effects during a lifetime

Potential hazard:

Average human daily exposure > RfD




13

Risk Assessment

MoE from LTD10 38% (7/18) with MOE < 100

MoE from NOAEL 18% (2/11) with MOE < 50

RfD 30% (3/10) RfD > average human daily exposure

Non-genotoxic carcinogens with potential human hazard: 27% (12/45) among non-genotoxic carcinogens

3.2% (12/371) among carcinogens (1, 2A, 2B)

●




14

MoE cut off’s of EPA



With an acceptable life time cancer risk of 1 in 104

MoE > 100,000 low risk for carcinogenicity

10,000 < MoE < 100,000 moderate risk for carcinogenicity

MoE < 10,000 high risk for carcinogenicity

For the present data:

1/45 (2%) in the low risk group

2/45 (4%) in the moderate risk group

22/45 (49%) in the high risk group

No data were available for 20 remaining compounds


15

Comparison EU vs EPA

Non-genotoxic carcinogens with potential human hazard:

- Europe, MoE < 100

- 12/45 (27%)

- 12/371 (3.2% of all carcinogens)

- U.S.A., MoE < 100,000

- 24/45 (49%) (moderate to high risk)

- 24/371 (6.5% of all carcinogens)




16

Putative methods to detect non-genotoxic carcinogensPutative methods to detect non-genotoxic carcinogens

• Structure-activity relationships (SAR) and quantitative (Q)SARStructure-activity relationships (SAR) and quantitative (Q)SAR

• Read acrossRead across

• Replicative DNA synthesisReplicative DNA synthesis

• In vitro In vitro cell transformation assaycell transformation assay

• ToxicogenomicsToxicogenomics




17

Dissecting modes of action of non-genotoxic carcinogens in primary mouse hepatocytes

Mirjam M Schaap, Edwin P Zwart, Paul FK Wackers, Ilse Hijskens, Bob van de Water, Timo M. Breit, Harry van Steeg, Martijs J. Jonker, and Mirjam Luijten.

Arch. Toxicol. submitted

For more information: [email protected]




18

Primary hepatocytes combined with toxicogenomics

Goal

Identification mechanism of action of non-genotoxic carcinogens

Advantages- relevant cell type: liver major organ for toxicity- metabolic competent cells- requires only small number of animals




19

Experimental design

Primary mouse hepatocytes– Two-step perfusion (based on Seglen, 1976)– Hepatocytes cultured in sandwich configuration

Compounds tested– 16 non-genotoxic carcinogens– Dose ~ 90% cell viability (based on MTT)

Microarray analysis– Cells exposed for 24 hours– Affymetrix HT MG-430 PM Array plate




20

Mode of Action (MoA) Pairs of substances tested (NGTXC)

Immunosuppressant and Calcineurin inhibitor

Cyclosporin A (CsA) Tacrolimus (FK506)

Organochlorine pesticide β-hexachlorocyclohexane (HCH)

Heptachlor epoxide (HCE)

Peroxisome proliferator Wyeth-14.643 (WY) Clofibrate (CF)

AHR ligand TCDD Aroclor 1254 (ARO)

CAR ligand TCPOBOP Phenobarbital (Pb)

Halogenated hydrocarbon Carbon Tetrachloride (CT) 1,1,1-trichloroethane (TCE)

Skin tumor promotor Okadaic acid (OA) Calyculin A (CA)

Metalloid Sodium Arsenite (SAR) Lead Acetate (LAc)

Non-genotoxic carcinogens




21

Data analyses

General– Principal component analysis– Pathway analysis per individual substance

Supervised clustering

Selection discriminative genes specific for mechanism?

Unsupervised clustering

Clustering of substances based on gene expression signature




22

Gene expression analysis - general



Differentially expressed genes:from 40 to >5,600 (FDR<0.05)


23

Results of the gene expression analysis



NGTXC mode of action DEG response PCA

TCE halogenated hydrocarbon 40 3 A

CT halogenated hydrocarbon 56 3 A TCPOBOP constitutive androstane receptor agonist 65 3 A

PB constitutive androstane receptor agonist 238 3 A

TCDD arylhydrocarbon receptor agonist 567 1 A

ARO arylhydrocarbon receptor agonist 762 1 B

CA skin tumor promotor 281 3 A

OA skin tumor promotor 282 3 A

WY peroxisome proliferator 1611 1 B

CF peroxisome proliferator 308 1 A

CSA immunosuppressant 124 1 A

FK506 immunosuppressant 2404 2 B

HCE ligand-independent estrogen receptor 2088 2 B

HCH ligand-independent estrogen receptor 115 3 A

LAC metalloid 5693 2 C

SAR metalloid 4130 2 C


24

Data analyses









25


Lac

Sar

CT

TC

E

OA

CA

FK

506

CsA P

b

TC

PO

BO

P

HC

H

HC

E

TC

DD

Aro

clo

r

Clo

fib

rate

WY

Method to select genes discriminative for MoA

-2 -1 0 1 2




26

• Apply supervised clustering on random pairs to confirm specificity of previous results

• Level of distinction is expressed as C-value; calculated for each individual gene

• Maximum C-value obtained for random pairs ~ 40

Supervised clustering - specificity

Maximal C-value

Distribution random pairs (5000 permutations)




27

C-value p-value

LAC & SAR 181.0 0.0002

CT & TCE 7.1 1.0

CA & OA 15.1 0.314

CSA & FK506 14.2 0.432

PB & TCPOBOP 12.2 0.614

HCE & HCH 11.9 0.635

ARO & TCDD 73.0 0.0002

CF & WY 587.0 0.0002 Maximal C-value

Distribution random pairs (5000 permutations)

Maximal C-values ‘correct’ pairs




28

C

om

po

un

d 1

Co

mp

ou

nd

2

Co

mp

ou

nd

3

Co

mp

ou

nd

4

Co

mp

ou

nd

5

Co

mp

ou

nd

6

Etc.

Compound 1

Compound 2

Compound 3

Compound 4

Compound 5

Compound 6

Etc.


Which substances have a similar gene expression profile and cluster together?




29

Unsupervised clustering – preliminary results

CsA CT

CT CA

Pb HCH

Lac Sar

WY Clofibrate

CA OA

TCPOBOP HCH

Sar Lac

Aroclor HCE

TCDD Aroclor

TCE TCPOBOP

HCE Aroclor

HCH Pb

OA CA

Clofibrate WY

FK506 Lac




30

Data analyses









31

Conclusions

● Primary hepatocytes combined with toxicogenomics useful to identify mechanisms of action of some non-genotoxic carcinogens:– Peroxisome proliferators– Metalloids– Skin tumor promotors– AhR ligands

● Useful tool for risk assessment of chemicals




32



Finding maximal transcriptome differences between reprotoxic and non-reprotoxic phthalate responses in rat testis

Xiaolian Yuan, Martijs J. Jonker, Jillian de Wilde, Aart Verhoef, Floyd R.A. Wittink, Jan van Benthem, Jos G. Bessems, Betty C. Hakkert, Raoul V. Kuiper, Harry van Steeg, Timo M. Breit, and Mirjam Luijten.

J. Appl. Toxicol 31, 421 – 430 (2011)

For more information: [email protected]


3333

Category approach – hazard identification

Category: Phthalates- mainly used as plasticizers to increase their flexibility,

transparency, durability and longevity.- widely applied in children’s toys- some phthalates reprotoxic, others not

Can we separate reprotoxic and non-reprotoxic phthalates by gene expression profiling?

Experiment:- Oral exposure of rats to phthalates, reprotoxic and non-reprotoxic- Sacrifice after 24h of exposure: gene expression profiling and

histopathology of the testis




34

Time (days) 0 1

Treatment

Gene expression profiling

Acclimatization

-21

7-week-old male HsdCpb:WU rats; n = 5

Animals killed 24 h after treatment

Study design




35

Replicates Compound AbbreviationChainlength Description

5 Controle (vehicle) - -

5 2-Methoxyethanol 2-ME - Positive control

5 Dimethyl phthalate DMP C1 Non-reprotoxic

5 Diethyl phthalate DEP C2 Non-reprotoxic

4 Di-n-propyl phthalate DPrP C3 Non-reprotoxic

5 Di-n-octyl phthalate DOP C8 Non-reprotoxic

5 Di-n-butyl phthalate DBP C4 Reprotoxic

5 Di-n-pentyl phthalate DPeP C5 Reprotoxic

5 Di-n-hexyl phthalate DHP C6 Reprotoxic

5 Di(2-ethylhexyl) phthalate DEHP C6 Reprotoxic

4 Mono(2-ethylhexyl) phthalate (MEHP) MEHP - Reprotoxic, metabolite

Experimental design




36

Untreated control Positive control (2-ME)

DHP treatment

No clear histopathological effects 24 hours after exposure to phthalates

Histopathology




37

Data analysis: 3 strategies

1. Identification of differentially expressed genes between treatment with individual compounds and the untreated control

2. Identification of differentially expressed genes between compound classes:

– reprotoxic phthalates (RT)– non reprotoxic phthalates (NRT)– untreated control (NC)

3. Identification of differentially expressed individual probes between compound classes




38

RT NRT

Differentially expressed genes (FDR<0,1) range from 107 (DEHP) to 1,273 (DPeP)

1. Differentially expressed genes-individual compounds (I)




39

• Little overlap between different RT or different NRT phthalates• Only 2 genes (LCMT2 and an unknown gene) are affected by all RT

phthalates, these are however also affected by at least one NRT phthalate

1. Differentially expressed genes-individual compounds (II)




40

2. Differentially expressed genes-compound classes (I)

• Differentially expressed genes (FDR <0,05):

– RT vs NRT phthalates: 469 genes

– RT phthalates vs NC: 454 genes

– NRT phthalates vs NC: 764 genes

• No genes were differentially expressed in all 3 comparisons

• 70 (51 + 19) genes not only differentially expressed in RT vs NRT phthalates, but also between either class of phthalates and NC




41

Discrimination between RT and NRT phthalates based on 70 selected genes

2. Differentially expressed genes-compound classes (II)




42

3. Differentially expressed probes-compound classes (I)

• Differentially expressed probes (FDR <0,05):

– RT vs NRT phthalates: 1897 probes

– RT phthalates vs NC: 4159 probes

– NRT phthalates vs NC: 6189 probes

• No probes were differentially expressed in all 3 comparisons

• 269 (143 + 126) probes were not only differentially expressed in RT vs NRT phthalates but had also altered expression between either class of phthalates and NC




43

Discrimination between RT and NRT phthalates based on 269 selected probes

3. Differentially expressed probes-compound classes (II)




44

Discussion

• Although 24 hours after exposure to phthalates no histopathological effects were observed, gene expression signatures were clearly changed

• Comparison of the different phthalate classes rather than individual compounds results in the identification of 70 genes or 269 probes that can discriminate between RT and NRT phthalates

• We, therefore, offer a proof-of-principle for the possibility to implement toxicogenomics in hazard assessment

• A combination of toxicogenomics and a category approach would allow prioritizing chemicals for toxicity testing and will as a result:

– Be much faster than conventional toxicity testing– Reduce number of laboratory animals used for testing– Lead to a reduction of costs




45 21 October 2010 | Health Canada45

Phtalates:

RIVM: Mirjam Luijten, Jan van Benthem, Harry van Steeg, Jeroen Pennings, Jillian de Wilde, Joke Robinson, Raoul Kuiper, Conny van Oostrom, Jos Bessems, Cees de Heer, ZhiChao Dang, Betty Hakkert

MicroArray Department, UvA, Amsterdam: Xiaolian Yuan, Martijs Jonker, Timo Breit

Acknowledgements

Risk assessment:

Lya Hernandez, Wout Slob, Wim Mennes, André Muller, Jan van Benthem

Hepatocytes:

RIVM: Mirjam Schaap, Mirjam Luijten, Edwin Zwart, Harry van Steeg

LACDR, University Leiden: Ilse Huijskens, Bob van de Water

Merck: Jan Polman, Willem Schoonen

MicroArray Department, University Amsterdam: Paul Wackers, Martijs Jonker, Floyd Wittink, Timo Breit

Fundings:Dutch Technology Foundation, STW 06935NGI Netherlands Toxicogenomics Centre NTC 050-06-510




46




47

Number non carcinogens in IARC

IARC group total carc non-carc percentage

1 77 64 13 16.9

2A 57 55 2 3.5

2B 237 207 30 12.7

3 507 471 36 7.1

total 878 797 81 9.2




48

Phthalates - Toxicity based on side chain length

Reprotoxic (C4-C6) Non-Reprotoxic




49

Results of the gene expression analysis



NGTXC mode of action DEG response PCA

TCE halogenated hydrocarbon 40 3 A

TCPOBOP constitutive androstane receptor agonist 65 3 A

TCDD arylhydrocarbon receptor agonist 567 1 A

ARO arylhydrocarbon receptor agonist 762 1 B

CA skin tumor promotor 281 3 A

CT halogenated hydrocarbon 56 3 A

CF peroxisome proliferator 308 1 A

CSA immunosuppressant 124 1 A

HCE ligand-independent estrogen receptor 2088 2 B

LAC metalloid 5693 2 C

OA skin tumor promotor 282 3 A

PB constitutive androstane receptor agonist 238 3 A

SAR metalloid 4130 2 C

FK506 immunosuppressant 2404 2 B

WY peroxisome proliferator 1611 1 B

HCH ligand-independent estrogen receptor 115 3 A


5050

Risk estimate: many uncertainties and critical data gaps– mode of action– interspecies extrapolation– dose-response analysis

Ethics: high number of test animals involved – reduction & refinement

Time-consuming and costly– gene expression profiles representative of toxicity– prioritization new chemicals

Potential relevance -omics to risk assessment




51

Toxicogenomics

● Gene expression profiles

● Advantage• Reliable• Reduce # of animals, time and cost• Information on mechanism of action

● Disadvantage• Identify multiple pathway-associated gene expression

profiles that encompass all non-genotoxic carcinogens



109 May 2012 2nd McKim Workshop, Baltimore 2012 Identification of non- genotoxic carcinogens based...

Documents

Transcript of 109 May 2012 2nd McKim Workshop, Baltimore 2012 Identification of non- genotoxic carcinogens based...