Use of Toxicological Pathways for Hazard Assessment in OECD (Q)SAR Toolbox: McKim Conference...

Use of Toxicological Pathways for Hazard Assessment in OECD (Q)SAR Toolbox:

McKim ConferenceSeptember 2008, Duluth, USA

LMC, Bourgas University, Bulgaria

Chemical Management Center, NITE, Japan

Fraunhofer Institute for Toxicology and Experimental Medicine, Germany

OECD, Environment Directorate, Paris

International QSAR Foundation, USA

Outline Conceptual framework of QSAR

Categorization and QSAR

Predicting human health endpoints in Toolbox

Molecular initiating events and toxicological pathways

Case study with 28d RDT

Mechanism database in Toolbox

MolecularInitiating

Events

Chemical Speciation

Metabolism

MeasurableSystem Effects

Adverse Outcomes

ParentChemical

Conceptual Framework of SAR/QSARConceptual Framework of SAR/QSAR

Rather than developing statistical models of complex endpoints, key molecular initiating events become the

“well-defined” endpoints for QSAR.

Gil Veith; International QSAR Foundation

Adverse Outcomes

ParentChemical

IQF Framework for QSARIQF Framework for QSAR

Black Box Models

Rapid but not mechanistically transparent

MolecularInitiating

Events

Speciation

Metabolism

Adverse Outcomes

ParentChemical

IQF Framework for QSARIQF Framework for QSAR

1. Identify Plausible Molecular Initiating Events 2. Design Database for Abiotic Binding Affinity/Rates 3. Explore Linkages in Pathways to Downstream

Effects 4. Develop QSARs to Predict Initiating Event from

Structure

QSARQSAR

Systems Systems BiologyBiology

Chemistry/Chemistry/BiochemistryBiochemistry

QSARQSAR

The categories concept is part of the historical description of QSARs

QSARs are quantitative models of key mechanistic processes which result in the measured activity

Each QSAR estimate is a result of two predictions:

Qualitative prediction of predominant interaction mechanisms and hazard identification (defined by category)

Quantitative prediction of the intensity (potency) of the specific mechanisms of interaction (predicted by QSAR)

Wrong definition for the mechanism of underlying reaction could result in using of a wrong QSAR for the potency

estimate

Example

Phenols are polar narcotics, uncouplers or electrophilic chemicals.

QSAR models for predicting acute effects for each mechanism have comparable uncertainty

The potency of the electrophilic mechanism can be orders of magnitude greater than polar narcotics

Wrong categorization of chemicals could cause significant errors in defining the potency

The logic for selecting a specific model (category) for a specific chemical is the cornerstone of regulatory acceptance

Basic Assumption for Regulatory Acceptance

OECD QSAR AD-Hoc group meeting, Madrid, April 2007

Modelled human health endpoints in Toolbox

Sensitization Lung Skin

Genotoxicity AMES bacterial mutagenicity Chromosomal aberration

Irritation/corrosion Eye Skin

Commonality between the modelled endpoints

The effects could be characterized by:

Single toxicological pathway

Strong dependency on initiating molecular events (e.g. on molecular structure)

Small impact of subsequent biological processes (“short” toxicological pathways)

MolecularInitiating

Events

Speciation

Metabolism

Adverse Outcomes

ParentChemical

QSAR Framework for modeled endpointQSAR Framework for modeled endpoint

QSARQSAR

Biological processesBiological processesInitiating chemical/BiochemicalInitiating chemical/BiochemicalInteractions Interactions

QSARQSAR

MolecularInitiating

Events

Speciation

Metabolism

Adverse Outcomes

ParentChemical

QSARQSARQSARQSAR

QSAR Framework for modeled endpointQSAR Framework for modeled endpoint

Initiating chemical/BiochemicalInitiating chemical/BiochemicalInteractions Interactions

EpidermisEpidermis

DermisDermisHypodermisHypodermis

VeinVein

Proteinconjugates

Penetration

Proteinconjugates

Metabolism

LymphLymph

Mechanism of skin sensitization

Subject of modeling

Assumptions in the model:1. Chemicals always penetrate stratum corneum2. Formation of protein conjugates is a premise for ultimate effect3. Metabolism may play significant role in skin sensitization

ModelSimulator of skin metabolism ∩ QSAR models

Parent

Metabolism

Phase II

Reactivespecies

S-PrW sensitization

S sensitization

S-PrS sensitization

No sensitization

Conclusion:

The categorization of substances according to chemical mechanisms governing the initiating reaction with protein or DNA is good enough for predicting human health effects resulting from single and “short” toxicological pathways

General characterization by the following grouping schemes:

• Substance information• Predefined• Mechanistic:

•Acute Toxicity MOA (OASIS)•Protein binding (OASIS)•DNA binding (OASIS)•Electron reach fragments (Superfragments) BioBite•Cramer Classification Tree (ToxTree)•Veerhar/Hermens reactivity rules (ToxTree)•Lipinski rules (MultiCase)

Chemicalinput

Profiling CategoryDefinition

Fillingdata gap

ReportEndpoints

Toolbox logical sequence of components usage

Molecular Initiating Events and Toxicological Pathways

General Consideration

Molecular LevelMechanism of chemical interactions

Mechanism of chemical interactions

Mechanism 1Mechanism 2Mechanism 3…

Molecular Level

Distribution in lipid phaseProtein binding Arylcarboxylate aminolysis Michael-type addition Schiff base formation … DNA binding Quinones Hydrazines …

Molecular Level

Distribution in lipid phaseProtein binding Arylcarboxylate aminolysis Michael-type addition Schiff base formation … DNA binding Quinones Hydrazines …

Receptor 1 –Receptor 2 – Receptor 3 – …

Molecular Level

Initiating event/Receptor•Activation of AP-1 、 NF-kB 、 EpRE in hepatocyte →Activation of JNK/AP-1 pathway•Activation of estrogen Signals → Proliferation of bile duct cell and hepatocyte injury 　　•Activation of MAPK Signals - Apoptosis•　　 …

Molecular Level

Initiating event/Receptor•Activation of AP-1 、 NF-kB 、 EpRE in hepatocyte →Activation of JNK/AP-1 pathway•Activation of estrogen Signals → Proliferation of bile duct cell and hepatocyte injury 　　•Activation of MAPK Signals - Apoptosis•　　 …

Molecular Level

Chemistry/Biochemistry

Cell LevelSystem biology/Effect

Molecular Level

System 1 –System 2 –System 3 –...

Molecular Level

System biologyHepatotoxicity mechanism:•　　 Oxidant stress•　　 Mitochondrial damage•　　 Apoptosis•　　 Degradation of membrane phospholipid•　　 Aberration of ion channel•　　 Increase of enzyme activition of drug metabolism•　　 Inflammatory responses•　　…

Molecular Level

Cell Effects•Hepatocyte•Changes in the tubular epithelium•…

Effect 1 Effect 2Effect 3...

Molecular Level

System biology

Molecular Level

System biology

Tissue, Organ and Body Observed Effects

Symptomology

Molecular Level

System biology

Symptomology

TissueEffect 1Effect 2Effect 3...

OrganEffect 1Effect 2Effect 3...

BodyEffect 1Effect 2Effect 3...

Molecular Level

System biology

Symptomology

Molecular initiating event(s) and subsequent downstream effects

Molecular Level

System biology

Symptomology

Molecular Level

System biology

Symptomology

Molecular Level

1. One complex endpoint (e.g., 28days RDT) could be conditioned by more than one toxicological pathway (blood toxicity, liver damage, kidney damage)

Conclusion:

2. (Q)SAR models should be associated with a single toxicological pathway

3. Chemicals which interact by different toxicological pathways should be out of the model mechanistic domain

Conclusion:

4. The categorization of substances according to chemical mechanisms governing the initiating reactions with protein or DNA is not enough for predicting human health effects resulting from multiple and complex toxicological pathways

5. The link between chemical and toxicological mechanisms and respective categorization schemes needs to be identified

Case study:

Twenty-eight day repeat dose oral toxicity test of chemicals (28d RDT)

1. Data produced by:

Safety examination of existing chemicals in NITE- Japan; under Japanese Chemical Substances Control Law;

Fraunhofer Institute for Toxicology and Experimental Medicine, Hanover, Germany

2. Categorization of chemicals for predicting 28d RDT is based on analysis of data by NITE and LMC

No. Structure NOEL* TargetOrgans

1 1 Erythrocyte 6 12ErythrocyteLiverKidney

11 80ErythrocyteLiverKidney

2 <5 Erythrocyte 7 40 Erythrocyte 12 1000

ErythrocyteLiverKidneyThyroid

8 30ErythrocyteLiverHeart

13 300

4 10 ErythrocyteLiver

ErythrocyteLiverKidneyTestes

14 300

5 2 ErythrocyteLiver

10 20 ErythrocyteKidney

* mg/ kg/ day

HO NH2

S NH2HO

28-day RDT tests conducted on male rats that tested 14 aromatic amines

Categorization of Anilines

1. Based on their effects on two organs:

Kidney

Blood Toxicity:

Blood toxicity effects: decrease in erythrocyte count (RBC) hemoglobin level (Hb) Hematocrit (HTC) glutamic-pyruvic transaminase (GPT) increase in the number of reticulocytes hemosiderin pigmentation in the spleen increase in hematopoiesis etc.

Toxicity scale of Intensity

Intensity scale: basis of the number of effects indicative of toxicity strong medium weak non

Toxicity scale of Intensity

Example: determining LOEL of N-ethylaniline:

Test doses - 0, 5, 25, 125 mg/kg/day

Decrease in RBC only has been observed at 5 mg/kg

Decrease in RBC, Hb and HTC–at 25 and 125 mg/kg Hence, LOEL for hemolysis is 5mg/kg/day

★ ★★ ★

★ ★★

★★ ★

★★ ★★

★▲

▲ ▲

0 200 400 600 800 1000Dose (mg/kg/day)

NH2H2N

NH2O NH2

OHO NH2

Cl SO3H

Anemia findings○: 0, ▲: 1, ■: 2, ★: >3

RBC↓, Hb↓, HTC↓, GPT ↑ Reticulocytes↑ hemosiderin pigmentation in the spleen hematopoiesis↑ etc.

HO NH2

Comparison of the intensities of anemia for 14 aromatic amines

-2 -1 0 1 2 3

logP (CLOGP)

LOELs for anemia

RBC↓, Hb↓, HTC↓Reticulocytes↑hemosiderin pigmentation in the spleenhematopoiesis↑ etc.

Relationships between LOEL for anemia and logKow for 14 aromatic amines

Water solubleanilines (logKow<0 )have NO EFFECT

NHOHNH+

CYP450in the Liver

Glucuronide and sulfate conjugation Urine

Adducts with DNA, Blood proteins: Hb, Alb

Hb Met Hb

Mechanism 1

Mechanism 2

These mechanisms of initiating reactions will be used to develop toxicological mechanism based categories

Mechanism underlying the toxic effects exerted by anilines

-2 -1 0 1 2 3

logP (CLOGP)

HO NH2

LOELs for anemia

RBC↓, Hb↓, HTC↓Reticulocytes↑hemosiderin pigmentation in the spleenhematopoiesis↑ etc.

Mechanism 2

Mechanism 1

Relationships between LOEL for anemia and logKow for 14 aromatic amines

Hemoglobin binding index (HBI)

HBI = (mmole compound/mole Hb)/(mmole compound/kg body weight)

Sabbioni, Environ. Health Perspect. 102 (1994) 61-67.

HBI = f (ΔE#)

Nitrenium ionE

Reaction pathway

Validation of Mechanism #1

0 2 4 6 8 10

Caluculated HBI

Calculated HBI (E# [eV]) vs. change in RBC in RDT test

in vitro, female rat liverR. Kato et al., Jpn. J. Pharmacol., 19 (1969) 53 – 62

HN NH2

Rats 28 days>5 mg/kg/day

Hemolysis(Japanese CSCL)

Rats 28 days>6 mg/kg/day

Hemolysis(EU Risk Assessment

Report)

Rats 90 days<150 mg/kg/dayNo Hemolysis

Findings[Johnnsen et al., Toxicol.

Lett., 30 (1986) 1-6.]

Repeated DoseToxicityTest

Metabolite of N-methylanilines

0 5 10 15 20 25

Caluculated HBI

Using HBI of the metabolite(Aniline)

Kidney

Effect on the kidney○： Nothing▲： Weak (Kidney wt↑)■ ： Medium (Other)★： Strong (Necrosis)

0 200 400 600 800 1000Dose (mg/kg/day)

▲★

NH2H2N

NH2O NH2

OHO NH2

HO NH2 NH2

Cl SO3H

Comparison of the effect on the kidney for 14 aromatic amines

-2 -1 0 1 2 3

logP (CLOGP)

LOELs for the kidney

HO NH2

Different effect of the chemical as compared with the anemia. This will be related with interaction mechanism

Relationships between LOEL for the kidney and logKow for 14 aromatic amines

NHOHNH+

CYP450in the Liver

Glucuronide and sulfate conjugation Urine

Adducts with DNA, Blood proteins: Hb, Alb

Hb Met Hb

Mechanism underlying the toxic effects exerted by anilines on kidney

Mechanism 2

Mechanism underlying the toxic effects exerted by anilines

binding with the SH proteins

RDT Effect on the kidney Effect initiating mechanism

Basophilictubule,proximal(100, 500)

Necrosis, Tubularepithelium,proximal(500)

Kidney,wt ↑(720)

Category building based on

the link between chemical and toxicological

mechanisms

Category Building

Category #1. Water soluble aromatic amines – logKow ≤ 0 – no effect

Based on the link between chemical and toxicological mechanisms

Category #2. If 0<logKow ≤ 1; it is eliminated by mechanism #2 and as parent or metabolite has alerting groups interacting with:

Proteins (such as quinone imines; Mechanisms #2) – strong kidney toxicity and weak blood toxicity (Category #2a)

DNA or blood proteins (Mechanisms #1) – week blood toxicity (Category #2b)

Category BuildingBased on the link between chemical and toxicological

mechanisms

Category #3. If logKow > 1 and as parent or metabolite has alerting groups interacting with:

Proteins (Mechanisms #2) – week kidney toxicity (Category #3a)

DNA (Mechanisms #1) – strong blood toxicity (Category #3b)

LogKow<0

Chemical

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

0<LogKow<1

No toxic effect

LogKow<0

Chemical

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

Parent or metabolites

have or form protein binding

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Strong kidney toxicity

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

NoNo No/Weekkidney toxicity

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

have or form DNA binding

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Week blood toxicity

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Week blood toxicity

NoNoNo blood toxicity

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Week blood toxicity

NoNoParent or

metabolites have or form

protein binding alert

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Week blood toxicity

NoNoParent or

Week kidney toxicity

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Week blood toxicity

NoNoParent or

YesYes

NoNoNo kidney

toxicity

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Week blood toxicity

NoNoParent or

YesYes

NoNoNo kidney

toxicity

have or form alert

interacting with DNA

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Week blood toxicity

NoNoParent or

YesYes

NoNoNo kidney

toxicity

have or form alert

YesYes

Strong blood toxicity

YesYes

0<LogKow<1

YesYes

No toxic effect

LogKow<0

Chemical

YesYes

Week blood toxicity

NoNoParent or

YesYes

NoNoNo kidney

toxicity

have or form alert

YesYes

Strong blood toxicity

Demonstrated by using RDT database, Fraunhover ITEM, Hanover, Germany

New Pilot Functionalities in Toolbox

Input chemical

Chemical category based on molecular initiating event

Measured RDT and Genotoxicity data are extracted from the OASIS genotox and RDT databases

OECD Toolbox

Species Organ/System

Effect/Tissue

immune system

intestine

kidney

larynx

lymph node

mammary gland

nervous system

dermal

drinking water

gavage

inhalation

oral unspecified

weight decreased

weight increased

cell proliferation

changed enzyme activity

degeneration

inflammation

metaplasia

Alanine aminotransferase

Alkaline phosphatase

Lactate dehydrogenase

Specification

RDT StructureEndpoint

Target endpoint: LOEL, feed, rat

Damaged organs and effects are unknown before chemical mechanism based categorization

Chemical grouping based on initiating molecular reaction

Target endpoint: LOEL, feed, rat

Damaged organs and effects should be defined based on chemical mechanism based categorization

Subcategorization based on chemical interaction mechanisms and/or structure

based similarity

Subcategorization based on toxicological mechanisms resulting from underlying

chemical interaction mechanisms

Missing information on organs expected to be damaged as a result of initiating reaction

Missing information on effects expected on damaged organs as a result of initiating reaction

Assumed Toxicological Pathway

Binding with DNA

Tumour formation in liver

Chemical Mechanisms Toxicological Mechanisms

Aromatic Amines

Defining toxicological pathway

and building the mechanism data base is critical for the Toolbox project

Chemical Mechanisms Toxicological Mechanisms

Chemical categorization

Grouping by Chemical Mechanisms

Toxicological categorization

Grouping by Toxicological Mechanisms

Predicted initiating reactions for the target chemical

Predicted toxicological outcome for the target chemical

The Mechanism Database in the Toolbox Project

System biology Symptomology Chemistry/Biochemistry

Molecular Level Cell Level

System biology

Tissue, Organ and Body

Symptomology Chemistry/Biochemistry

Cell Level

System biology

Symptomology

Search / Collection of in vitro Test Information

Molecular Level

Cell Level

System biology

Symptomology

• Functions• Pathways• Processes

Molecular Level

Cell Level

System biology

Symptomology

Receptor 1 –Receptor 2 –Receptor 3 –Receptor 4 –...

System 1 –System 2 –System 3 –System 4 –...

Effect 1 Effect 2Effect 3Effect 4...

Molecular Level

Cell Level

System biology

Symptomology

Initiating reaction

(Q)SAR

Categorization

Molecular Level

Cell Level

System biology

Symptomology

Initiating reaction

(Q)SAR

Categorization

Molecular Level

Chemical Mechanism Database

Cell Level

System biology

Symptomology

Initiating reaction

(Q)SAR

Categorization

Tissue

Molecular Level

Expert observation

Cell Level

System biology

Symptomology

Initiating reaction

(Q)SAR

Categorization

Tissue

Molecular Level

Expert observation

Toxicological Mechanism Database

Cell Level

System biology

Symptomology

Initiating reaction

(Q)SAR

Categorization

Mechanism Knowledge Database

Tissue

Expert observation

Molecular Level

Cell Level

System biology

Symptomology

Initiating reaction

(Q)SAR

Categorization

Mechanism Knowledge Database(Toxicological pathways)

Tissue

Molecular Level

Expert observation

Contributors:

O. MekenyanS. DimitrovT. Pavlov

G. ChankovA. Chapkanov

Laboratory of mathematical Chemistry, Bourgas, Bulgaria

Chemical Management Center, NITE, Japan

Case study on RDT of aromatic aminesChemical vs. toxicological mechanisms

(Project of NEDO Japan)

Jun YamadaYuki Sakuratani

Fraunhofer Institute for Toxicology and Experimental Medicine, Department Chemical Risk Assessment, Hanover,

Germany

Data from REPDOSE Database (Project of CEFIC LRI)

Inge MangelsdorfSylvia EscherAnnette Bitsch

Environment Directorate, OECD, Paris

Bob DiderichTerry Schultz

International QSAR Foundation, USA

Gilman Veith

Use of Toxicological Pathways for Hazard Assessment in OECD (Q)SAR Toolbox: McKim Conference...

Documents

LMC powerpoint

Bourgas presentation

BOURGAS MUNICIPALITY NEW SOCIAL SERVICES FOR DISABLED CHILDREN IN BOURGAS MUNICIPALITY

Cathy McKim - copy portfolio Aug 2016

Kathleen mckim creating webquests

McKim Creed

LMC Service Awareness How aware are LMC students of LMC services? How are students being informed of LMCs services?

Wireless Mesh Networks Myungchul Kim mckim@cs.kaist.ac.kr

Myungchul Kim mckim@icu.ac.kr

McKim Map 2 - Greater Sudbury · 2017-09-18 · Broder Map 3 Broder Map 4 Neelon Map Lot 12 Con 3 McKim Map Lot 1 Con 3 McKim Map Lot 2 Con 3 McKim Map Lot 3 Con 3 McKim Map ... CONC

What is SCADA? - McKim & Creed

LMC Stories

PTPI - Bourgas, Bulgaria Student Chapter - "The Vectors"

REGIONAL ASPECTS OF BOURGAS-ALEXANDROUPOLIS PETROL ...anale.feaa.uaic.ro/anale/resurse/ec1kalinkov.pdf · REGIONAL ASPECTS OF BOURGAS-ALEXANDROUPOLIS PETROL PIPELINE Konstantin KALINKOV

2nd McKim Workshop, Baltimore 2012

McKim Design Group

Mckim Hancock 2012 cap1 p1_24.pdf

Lillian Lee McKim Pulitzer Rousseau

Bourgas iclei pptpsept en

LMC Brochure