Computational Toxicity and QSARsiweb.dss.go.th/reach/Fulltext/Reach_PPT/supappt.pdf · 2006-12-22 · Computational Toxicity and QSAR Supa Hannongbua Department of Chemistry, Faculty

Computational Toxicity and QSAR

Supa HannongbuaDepartment of Chemistry, Faculty of Sciece

Kasetsart University, Bangkok [email protected]

Background and MotivationBackground and MotivationWhat is QSAR?What is QSAR?QSARQSAR’’ss ApplicationsApplicationsResearch Area in ThailandResearch Area in ThailandConcept proposal Concept proposal CompututationalCompututational Toxicity in Dyes and CosmeticsToxicity in Dyes and Cosmetics

The European Chemicals Bureau provides technical and scientific support for implementation of certain EU legislation

on dangerous chemicals and the preparation for REACH

Action 1311Assessment of Chemicals

Action 1313Support to REACH

Action 1314REACH-IT

&Informatics

Action 1321Computational

Toxicology(QSARs)

http://http://ecb.jrc.itecb.jrc.it

Background and MotivationBackground and Motivation

1311 Assessment of ChemicalsNew and Existing Chemicals*Biocidal ProductsClassification and Labeling*Testing MethodsExport/Import

Actions in FP6 year 2006Actions in FP6 year 2006

The Action mission is to provide the scientific and technical support to the conception, development, implementation and monitoring of EU policies on dangerous chemicals.

5

What is in silico / QSAR ?What is in silico / QSAR ?

In silico and REACHWhat can be done in silico ?An example: in silico CYP inhibitionWhat next ?

6

Levels of testingLevels of testing

in cerebro

in silico

in vitro

in vivo

7

(Q)SARs defined by ECB(Q)SARs defined by ECB

Quantitative structure-activity relationships, collectively referred to as (Q)SARs, are theoretical models that can be used to predict the physicochemical and biological properties of molecules.

A structure-activity relationship (SAR) is a (qualitative) association between a chemical substructure and the potential of a chemical containing the substructure to exhibit a certain biological effect.

A quantitative structure-activity relationship (QSAR) is a mathematical model that relates a quantitative measure of chemical structure (e.g. a physicochemical property) to a physical property or to a biological effect (e.g. a toxicological endpoint).

8

In principle, (Q)SARs could be used to supplement experimental data, or to replace testing:

Supplement to testing:1) To support priority setting of chemicals.2) To guide experimental design (e.g. selection of tests

/doses).3) To provide mechanistic information.

Replacement of testing:4) To group chemicals into chemical categories.5) To fill in data gaps for classification and labelling.6) To fill in data gaps for risk assessment.

Possible applications for QSARs in the Regulatory Assessment of Chemicals

Possible applications for QSARs in the Regulatory Assessment of Chemicals

9

Chemical categories and read-acrossChemical categories and read-across

The underlying premise underpinning all structure-activity relationships (SARs) is the expectation that structurally similar chemicals are likely to have similar physicochemical attributes and biological effects. The approaches of read across and chemical categories are based on this similarity principle.

In the read-across or analogue approach, endpoint information for one chemical is used to make a prediction of the endpoint for another chemical, which is considered to be "similar" in some way. In principle, read-across can be used to assess physicochemical properties, toxicity, environmental fate and ecotoxicity, and it may be performed in a qualitative or quantitative manner.

ECB/JRC 2005

10

The chemical category concept and supporting role of (Q)SARs

The chemical category concept and supporting role of (Q)SARs

11

ContentsContents

What is in silico / QSAR ?In silico and REACHWhat can be done in silico ?An example: in silico CYP inhibitionWhat next ?

12

The White PaperThe White Paper

3.2 Research and ValidationDevelopment of alternative methodsOther research priorities

improvement and simplification of risk-assessment proceduresimprovement and development of new toxicological and eco-toxicological methodsparticular reseach efforts need to be made for developing and validating in-vivo and in-vitro test methods as well as modelling (e.g. QSAR) and screening methods for assessing the potential of adverse effects of chemicals on endocrine systems of humans and animals.etc…

13

Use of QSARs under REACH Use of QSARs under REACH

Acceptance of QSAR results - BOTH positive and negative results will be accepted if

Models have been validated Models are adequately documented and meet acceptance criteria for a given application- “fit for purpose” concept

(Q)SARs may support grouping of chemicals Chemical categories and minimise testing

Animal testing is conducted as a last resort

14

RemarksRemarks

Need to use (Q)SARs is explicit in legislative proposal for REACH In silico methods developing fastPredictive power already quite good in some (restricted) areas Future regulatory use of (Q)SARs will be controlled by various acceptance and validity criteria

Alternative approaches can reduce the use of test animals Examples of read-across under 793/93: Flame retardants

TCEP Tris(2-chloroethyl) phosphate 115-96-8

TCCP Tris(2-chloro-1-methylethyl)phosphate 13674-84-5

TDCP Tris[2-chloro-1-(chloromethyl)ethyl]phosphate 13674-87-8

V6 2,2-bis(chloromethyl) trimethylene bis[bis(2-chloroethyl) phosphate 38051-10-4

ENV: UKHH:Ireland

Rapporteur:Germany

Basis for read across

• Structural analogy of chloroalkyl phosphate esters

• lack of data for some substances for some endpoints

• Decision needed for

– Risk Characterisation and

– Classification and Labelling

3

Structural formula:

2.833.692.681.78log Kow

232 mg/l18.1 mg/l1080 mg/l7820 mg/l at 20°CWater Solubility

2.75 x 10-6 Pa at 25°C5.6 x 10-6 Pa at 25°C1.4 x 10-3 Pa1.14 x 10-3 Pa at 20°C

Vapor Pressure252°C (decomp)>200°CCa. 288°C (decomp.)320°C (decomp.)Boiling point

FP < -50.5°C<-20°C<-20°C<-70°CMelting point583430.91327.57285.49Molecular weight:

C13H24Cl6O8P2C9H15Cl6O4PC9H18Cl3O4PC6H12Cl3O4PMolecular formula:

V6TDCPTCPPTCEPSubstance

POO

O O

Cl

Cl

Cl

Comparison phys-chem. Properties

POO

OO

CH3

CH2Cl

CH2ClCH3

ClCH2

CH3ClCH2

CH2ClO POP

O

O

OO

OO

Cl

Cl

Cl

Cl

POO

OO

CH2Cl

CH2Cl

CH2ClClCH2

ClCH2

ClCH2

3

Read across for Sensitization

Guinea pig, M&KGuinea pig, M&KGuinea pig, M&KNo studyStudy:

No classificationNo classificationNo classificationClassification

negativenegativenegativeResults:

V6TDCPTCPPTCEPSubstance

Conclusion:TCEP is regarded as not being sensitizing on basis of read across to structural related substances (TCPP and TDCP) and no observation of sensitizing potential in workers exposed to TCEP.

3

Read across for Carcinogenicity

No studyrat, 2-y-carcinogenitcity study

No study2 y carc. Study in rat and mice

Study:

Carc. Cat. 3 R40 agreedCarc. Cat. 3 R40 proposed

carc. Cat. 3 R40 agreed at TC C&L

Classification:

Renal cortical tumours, testicular interstitial cell tumours, hepatocellular adenomas and adrenal cortical adenomas

Kidney tumours in two species and both sexes;

Results:

V6TDCPTCPPTCEPSubstance

Conclusion:Read across to TCPP from TCEP and TDCP proposed. There are similarities in hydrolysis and one common metabolite. The NOAEL is taken from the study with TDCP. No read across for V6 as it is an alkyl bridged bis-phosphate ester which makes it probably a bulkier and less bioavailablemolecule. (There were proposals from MS to consider V6 as two molecules of TCEP and to take over the same classification).

3

(No) read across for Fertility

No study12 weeks fertility study in rabbits, 2 y carcinogenicity study

No study2 generation in miceStudy:

Repro Cat. 3 R62 agreed at TC C&L

repro. Cat. 2 R60 agreed at C&L

Classification:

No effects in rabbits; significant effects on the male reproductive organs in the carcinogenicity study

High dose 90 day study showed no significant effects

Impairment of fertility for both sexes: reduced sperm counts, reduced litter size

Results:

V6TDCPTCPPTCEPSubstance

Conclusion:Read across was not regarded as appropriate on basis of the available data (V6 see also justification for carcinogenicity). A new 2 generation study will be performed for TCPP and V6.

3

(No) read across for Developmental Effects

No studydevelopmental study in ratsdevelopmental study in rats

developmental study in rats and mice

Study:

No classification/no concernNo classification/noconcern

No classification/no concern

Classification:

No significant effectsNo significant effectsNo significant effectsResults:

V6TDCPTCPPTCEPSubstance

Conclusion:Read across was not regarded as appropriate for V6. Further testing will depend on the results of the 2 generation study.

13141314 -- REACH-IT & Informatics (new in 2006)

Actions in FP6 year 2006Actions in FP6 year 2006

Chemical databases, IT for registration, workflow for dossiers, global portal for sharing data

RIP 2RIP 2

REACH-IT Architecture

Intelligent Testing Strategies (ITS)

Endpointinformation

(Q)SARsRead Across

In-vitro

ExposureScenarios

(Annex VII/VIII)

Existinginformation

TESTING

?

Computational Toxicity (QSARs)QSARs = Quantitative Structure Activity relationships

Actions in FP6 year 2006Actions in FP6 year 2006

Development, validation and implementation of (Q)SARs and other estimation methods for the assessment of chemicals Experimental LogLC50 (mol/l)

Pred

icte

d Lo

gLC5

0 (m

ol/l

)

0-1-2-3-4-5-6-7

0

-1

-2

-3

-4

-5

-6

-7

MOA-SetNPNPN

P86P85

P84P83P82

P81 P80

P79P78P77

P76

P75

P74P73

P72P71

P70P69

P68

P67P66

P65

P64

P63P62P61

P60

P59P58

P57

P56P55

P54P53

P52

P51P50

P49

P48

P47

P46P45 P44P43P42P41

P40P39

P38

P37P36P35 P34

P33P32P31

P30P29

P28

P27

P26P25P24

P23

P22

P21

P20P19

P18P17

P16

P15

P14 P13P12

P11P10

P9P8

P7

P6P5

P4

P3

P2

P1

N58

N57

N56

N55

N54

N53

N52

N51

N50

N49

N48

N47N46

N45

N44

N43N42

N41

N40

N39

N38

N37

N36

N35

N34

N33 N32N31

N30

N29N28

N27

N26

N25

N24

N23

N22

N21

N20

N19

N18

N17

N16

N15

N14N13N12

N11

N10

N9N8

N7

N6

N5

N4

N3

N2

N1

QSAR Action (Computational Toxicity)

((Q)SARsQ)SARs:: theoretical models that can be used to predict the physicochemical and biological properties of molecules. They aresometimes called in silico models.

QSARs can be used, in combination with other types of information, to minimize testing (e.g. animal testing)

This does not imply that a single QSAR replaces a single test

Additional animals Use of (Q)SARs, read-across 3.9 million Minimal use

2.6 million Average use (likely scenario)

2.1 million Maximal use

Animal-saving potential: 1.3-1.9 million animalsVan der Jagt et al. (2004).

Alternative approaches can reduce the use of test animals under REACH.

http://ecb.jrc.it

(Q)SARs and REACH: Use of Animals

(Q)SARs and REACH: Testing Costs

Additional cost Use of (Q)SARs, read-across 2.3 billion Euro Minimal use

1.5 billion Euro Average use (likely scenario)

1.1 billion Euro Maximal use

Cost-saving potential: € 800-1130 million

Pedersen et al. (2003). Assessment of additional testing needs under REACH.http://ecb.jrc.it

Computational Computational NanotoxicologyNanotoxicology

Nanotoxicology

•• Increasing importance of nanotechnologyIncreasing importance of nanotechnology

•• Unique risksUnique risks: adverse effects of nano-particles and materials cannot always be predicted from the known toxicity of the corresponding bulk material.

• Limited understanding of the potential toxicity of nano-sized materials.

•• NanotoxicologyNanotoxicology: addressing the special toxicity that may be associated with nano-sized materials.

NanoparticlesNanoparticles

NoseNose

SkinSkin GutGut

BrainBrain

BloodBlood

Bone Bone marrowmarrow

SpleenSpleen EndotheliumEndothelium LiverLiver HeartHeart Placenta / Placenta / foetusfoetus

LungLung

Background and MotivationBackground and MotivationWhy Computational Toxicity?Why Computational Toxicity?What is QSAR?What is QSAR?QSARQSAR’’ss ApplicationsApplicationsResearch Area in Thailand Research Area in Thailand

Computational Toxicity in Dyes and Cosmetics Computational Toxicity in Dyes and Cosmetics Concept proposalConcept proposal3D3D--Molecular Database for substances in Dyes and Molecular Database for substances in Dyes and CosmeticsCosmetics

QSARQSAR’’ss ApplicationsApplicationsQSAR for Drug Design (Since 1960, C. QSAR for Drug Design (Since 1960, C. HanschHansch) (Most)) (Most)

QSPR for Materials (Inorganic, etc.)QSPR for Materials (Inorganic, etc.)

QSAR for Agriculture and Toxicology (Pesticide, etc.)QSAR for Agriculture and Toxicology (Pesticide, etc.)

QSARQSAR’’ss ApplicationsApplications

(~100 calculated properties)(~100 calculated properties)

Example: QSAR of Example: QSAR of PolyaromaticPolyaromatic Hydrocarbons (Hydrocarbons (PAHsPAHs))

15 published 15 published QSARsQSARs for drug design by for drug design by HannongbuaHannongbua, S. et al. , S. et al.

QSAR for Materials (InorganicQSAR for Materials (Inorganic))• Inorganic cations toxicity - application of QSAR analysis.

Ind. Environ. Xenobiotics, Proc. Int. Conf. 1981, Meeting Date1980, 83-6.

• Enache M.; Dearden J. C.; Walker J. D. QSAR analysis of metal ion toxicity data in sunflower callus cultures (Helianthus annuus Sunspot ). QSAR & Combinatorial Science, 2003, 22:234-240.

• QSAR in toxicology. III. Its use in the determination of thetoxicity of inorganic cations. Ceskoslovenska farmacie. 1981, 30:7-10.

• Use of structure-activity relationships to estimate toxicity ofinorganic cations. Experientia. Supplementum. 1976, 23:83-4.

The aim of the researchTo investigate the toxicological effects of 13 inorganic and 21

organic compounds are evaluated using the pyriformis FDA esterasetest, the conventional pyriformis population growth impairment assayand the luminescent inhibition test (Microtox test).To Study the relationships between the toxic effects of the

substances tested and the ion characteristics of the metal ions or thehydrophobicity (quantified by the 1-octanol/water partition coefficient,log Kow) of the organic compounds.

Methods• Calculations and Statistical Analyses

The relative toxicity of the tested substances was quantified bythe determination of the IC50

For each toxicant and each assay, the concentrations weretransformed into logarithms, and the IC50 was determined byregression analysis.

The relationships between the Tetrahymena 1-h IC50, 9-h IC50,and the Microtox 30-min EC50 were determined by the Spearmanrank correlation coefficient (P<0.05).

The relationship between toxicity results and ioncharacteristics or lipophilicity of organic substances was testedusing mean square root linear regression analysis (P<0.05).

Statistical analyses were computed with Stat View SE 1.03software.

Results

Results

Results

Results

ConclusionToxicology-based QSARs can be used for the prediction of the

toxic potency of chemicals and the interpretation of mechanismsof action (Cronin and Dearden, 1995).

The relative toxicity of metal ions and organic compoundsdetermined with the two Tetrahymena biotests is predictableusing two ion characteristics, the softness index σp and χ2

m r,and the hydrophobicity coefficient log Kow, respectively.

Background and MotivationBackground and MotivationWhy Computational Toxicity?Why Computational Toxicity?What is QSAR?What is QSAR?QSARQSAR’’ss ApplicationsApplicationsResearch Area in Thailand Research Area in Thailand

CompututationalCompututational Toxicity in Dyes and CosmeticsToxicity in Dyes and Cosmetics

QSARQSAR’’ss application in Cosmeticapplication in Cosmetic

The aim

Collect information on all hair dye substances used inpermanent or temporary hair dyes in Europe and then to rankthese substances according to their estimated potency, theperspective being to supplement the current diagnostic work-upfor hair dye allergy with new potential allergens.

Methods• Prediction of sensitization potency

Structures identified were then imported into a molecular spreadsheet TSAR(Version 3.3, Accelrys Ltd., Cambridge, UK).

• Ranking the substances according to their sensitization potentialThe TOPS-MODE QSAR model was used in order to estimate the likely

sensitization potency• Tonnage amount

The European Cosmetic Toiletry and Perfumery Association has generatedtonnage data for use in the EU Commission for prioritizing hair dyes in riskassessment and risk management.

• Cluster analysisThe cluster analysis provided a means of grouping substances according to

their chemical properties such that a representative diverse subset could beselected for further work.

Results

Results

Results

ConclusionsThe sensitization potential of each substance was then

estimated by using a quantitative structure–activity relationship(QSAR) model and the substances were ranked according totheir predicted potency.A cluster analysis by using TOPS-MODE descriptors as inputs

helped us group the hair dye substances according to theirchemical similarity. This would facilitate the selection of potentialsubstances for clinical patch testing and would provide someclinical validation of the QSAR predictions.

QSARQSAR’’ss application in Dyesapplication in Dyes

The purpose of this paperDerive quantitative structure-activity/property-activity

relationships (QSAR/QPARs) for the mutagenicity (rev/nmol)of a variety of 4-aminoazobenzene (AAB), N-methyl-4-aminoazobenzene (MAB) and N,N-dimethyl-4-aminoazobenzene (DAB) derivatives in the S. typhimuriumTA98 bacterial strain with S9 activation (TA98+S9); thisparticular bacterial strain is well known to detect frameshiftmutagens.

Methods• The structures optimized at the semiempirical AM1 computational

level as implemented in AMPAC 5.0.• The CODESSA/AMPAC integrated software package was used

to calculate hundreds (>300) of molecular descriptors.• The entire collection of descriptors was then used in conjunction

with the statistical facilities of CODESSA to develop multilinearregression models for the log of the measured mutagenicity(rev/nmol) in TA98+S9, logTA98.

Results

Results

G1 XY Shadow/XY RectangleaE1 WPSA-2 Surface weighted partial positive surface areabE2 Polarity parameter/(distance)2 =(QmaxQmin)/(distance)2Q1 Average electrophilic reactivity index for a N atomQ2 Maximum electron-nuclear attraction for a C–C bond

Q3 Minimum net atomic charge for a C atomQ4 Maximum electron-nuclear attraction for a C atomQ5 Maximum bond order of an N atomH1 Final heat of formation/number of atomsO1 LogP

Results

Results

Conclusions• Developing QSAR/QPARs that correlate the relative mutagenic activity of

aminoazobenzene derivatives with various molecular descriptors can helpidentify factors that alter their relative mutagenicity.

• QSAR/QPAR studies can be useful in establishing biochemicalmechanisms/interactions, and in developing combinatorial strategies forthe synthesis of environmentally safe chemicals. Such studies involvingaminoazobenzene derivatives are particularly important because of theirwidespread use in the textile industry.

Thank you for your attention