1. T. W. Schultz Presented at the Logan Workshop March 23-24, 2010
Estimating Reactivity From Structure Using the OECD (Q)SAR Application Toolbox
T. W. Schultz
Presented at the Logan Workshop
March 23-24, 2010

2. Topics Background & The Problem Michael Acceptors: An Example
 Toolbox Applications
Pathways Application
Summary

3. Coding Reactivity From Structure
Currently no universal QSAR model
Local models for simple congeneric groups
2D structure
Qualitatively easy
Main drawback false positives
Quantitatively more difficult
Iso-reactive groups (lumping vs. splitting)
Sub-grouping and sub-sub-grouping

4. 2D Coding Michael Acceptors
Qualitatively Predicted
as any Polarized
a,b-Unsaturated Compound
for example, C=CC(=O)

5. 2D Coding Michael Acceptors
Quantitative Predictions Impacted By Substitution at
C1=C2(Pg)C3
Reactivity is extended fragment -based

6. Structural Variation in Reactive Potency of Esters
TYPE STRUCTURE RC50(mM)
Acetylene dicarboxylates ROC(=O)C#CC(=O)OR
Propiolates C#CC(=O)OR
Vinylene dicarboxylates ROC(=O)C=CC(=O)OR
Acrylates C=CC(=O)OR
Alkyl 2-alkynoates RC#CC(=O)OR
Crotonates CC=CC(=O)OR
Methacrylates C=C(C)C(=O)COR >30.0

7. Variations in RC50 Values for Substituted Acrylates, C=CC(=O)OR
Derivative RC50 (mM)
_______________________________________
Ethyl , 0.55
Vinyl , 0.11
2-Hydroxyethyl 0.25, 0.29
n-Propyl 0.80, 0.92
Propargyl 0.19, 0.24
n-Hexyl 0.88, 0.76
Phenyl ,

8. To Quantitatively Predict Reactivity: Must Be Able to Separate
C#CC(=O)
C=CC(=O)
CC=CC(=O)
C=C(C)C(=O)

9. To Quantitatively Predict Reactivity: Must Also be Able To Separate
C=CNO2
C=CC#N
C=CC(=O)C
C=CC(O)OC
C=CC(=O)NH2

10. Toolbox Application
Use reactive to group chemicals into categories and to facilitate the selection of chemical analogues, which allows the integrates of the mechanism of reaction in defining the best category or sub-category
Then do read-across

11. Read Across with GSH & LLNA Data
EC3 = 0.01
RC50 = 0.03
RC50 = 0.05
RC50 = 0.05
RC50 = 0.09
RC50 = 0.02

12. Protein Binding in Toxicity
Mechanisms of Protein
Binding
In Vitro
Measurements
In Chemico
Measurement
Hazard
Assessment
Endpoints
In vitro
effects
Michael
addition
SN2
SNAr
Reactive
Potency
In vivo
effects
In silico
modeling

13. Application Reactivity to Catgorizing an Inventory
≈ 1500 substances on the List of Flavor and Fragrance Related Substances
≈1300 discrete substances of which:
79 Fast- to moderate-reacting Michael-acceptors;
19 Slow-reacting Michael-acceptors;
57 Schiff-base aldehydes;
29 Acetals;
15 Disulfide formers;
11 Cyclic addition diones;
9 Disulfide exchangers;
3 O-heterocyclic ring openers.
>40 pro-electrophiles

16. Pathway Applications Screening Tool Many be used in isolation
Risk Assessment
Must be used as part of an ITS
Represents the molecular initiating event of an adverse outcome pathway

17. Pathway for Allergic Contact Dermatitis
1. Haptenation; 2. Epidermal inflammation & LC activation; 3. LC migration; 4. DC: T cell interaction; 5. T cell proliferation; 6. Increase in hapten-specific T cells; 7. Hapten re-exposure; 8. Acute inflammation; 9. T cell-mediated inflammation
Karlberg et al. Chem. Res. Toxicol. 2008, 21,

18. LLNA-tested Michael Acceptor
SUBSTRUCTURE MESSAGE
C=CC=O Vinyl or vinylene with a carbonyl
[CH2]=C(C)C=O -C-atom alkyl-substituted with
a carbonyl
O=CC=CC=O -C-atom substituted with a second carbonyl
[CH]=C(C(=O))C=O -C-atom substituted with a second carbonyl
C=[CH]c1ccncc1 Para-vinyl azaarene
O=C1[CH]=CC(=O)C=C Para-quinone 18

19. Michael Acceptor Not Tested in LLNA
SUBSTRUCTURE MESSAGE
C#CC=O Ethylnylene or acetylenic with a carbonyl
C=CN(=O)=O Olefinic nitro
C#CS(=O) Ethylnylene or acetylenic with a S=O C=CS=O Vinyl or vinylene with a S=O
C=CC#N Olefinic cyano
C#Cc1ncccc Ortho-ethylnylene azaarene
C=[CH]c1ncccc Ortho-vinyl azaarene
C#Cc1ccncc Para-ethylnylene azaarene
C=[CH]C(=O)[OX1] Vinylene carboxylic acid
O=C1C=C[CH]=CC1=O Ortho-quinone
[CH2]=[CH][CH]=O Acrolein

20. Subcategorization of Michael Acceptors by Reactivity
Extremely fast: quinones, propiolates ,
1-alken-3-ones
Fast: acrylates, 2-alkenals,
3-alken-2-ones
Moderately Fast: alkyl 2-alkynoates
Slow: crotonates
Very Slow: methacryates, tiglates
Non-Reactive: non-,-unsaturated

21. Summary We have a 2D modeling strategy
Quantitative reactivity data is available for QSAR development
We have an application scheme

22. Thank you