Transcript Increasing Hardness

Current Status of Reactivity Models
in Aquatic Toxicity
Mark Cronin
Liverpool John Moores University
England
Aquatic Toxicology
Aquatic Toxicity Prediction

Acute toxicity

Chronic toxicity

Endocrine disruption
Modes / Mechanisms of Action

Non-reactive



Reactive


Narcosis – baseline effect
Uncoupling
Toxicity elevated above narcosis
Receptor Mediated
Role of Mechanism of Action in
Predicting Aquatic Toxicity

Mechanism based QSARs

Acute – Chronic Ratio (ACR)

Category formation
Vonk et al (2009) MOA Workshop Report - ATLA
Role of Reactivity in Predicting Aquatic Toxicity

40-70% of industrial chemicals are non-reactive



Toxicity can potentially be predicted accurately
Acute-chronic ratios may be consistent
For reactive chemicals

Grouping may be of assistance
Narcosis and QSARs

Narcosis is reversible; a baseline effect




QSARs need to be robust models


There are a number of narcotic mechanisms
Well accepted and understood, if not defined at the
molecular level
Consistent ACR
Log P models preferred
Domain needs to be defined

Ellison et al for Tetrahymena
Electrophilic Reactivity and QSARs

Mechanisms – see next slide



Acute toxicity greater than narcosis
ACR elevated
Traditionally difficult to model toxicity except
within closely defined classes or mechanisms
Generic Acute Fish Mortality Pathway for
“Respiratory Irritation”









a direct-acting electrophile or can be abiotically or biotically
transformed to an electrophile
molecular sites of action are specific nucleophiles, either thiol
or amino-moieties, reactions are non-selective
molecular initiating event is covalent perturbation of proteins;
biochemical pathways affected are varied and result in general
inhibition of cellular functions
cellular- and organ-level consequences are irreversible
target organ(s) or tissue(s) are the gill
key physiological response is general hypoxia
key target organ-response is sloughing of the gill epithelium
key organism response is a sharp reduction in blood oxygen
level, asphyxiation, quickly leads to death
Schultz (2010); McKim – FATS publications
Applications of Reactivity in
Predicting Aquatic Toxicity

Development of QSARs

Identification of narcotic / reactive / other
mechanistic domains


Definition of reactive domains
Grouping to allow for read across
QSARs Using Reactivity

Acute toxicity to Tetrahymena pyriformis of Michael
Acceptors

Log (IGC50-1) = 1.05 (log RC50-1) + 1.53

n = 20, s = 0.39, r2 = 0.975, q2 = 0.973
F= 699, relationship covers 9 log units

Information from Schultz et al
Calculated Descriptors of Reactivity






LUMO
HOMO
Electrophilicity index (w)
Superdelocalisability
Atomic charges
Limited to categories or do not capture protein
reactivity
Is a Compound Narcotic?

If it is:



If it isn’t


We can predict toxicity
We can extrapolate ACR
More information / testing may be required
How do we determine if a compound is narcotic?
Methods to Determine if a Compound is Narcotic

Mode of action assignment





Domain definition
Excess acute toxicity


Verhaar
Russom (ASTER)
OASIS
Cytotoxicity
Reactivity
Narcosis: Killer Questions

Are physico-chemical properties consistent with
narcosis?


Is your compound in a narcotic domain?





Solubility, volatility
Classes / analogues
Verhaar / Russom / OASIS
MOA definitions
Is the compound “unlikely” to be activated through
metabolism?
Is your compound reactive?
Killer Question:
Is Your Compound Reactive
Mechanistic Basis for Needing to
Understand Reactivity

Need to understand target nucleophile

Possible covalent interactions with nucleophile

How to capture the possibility of interactions
Mapping Toxicity onto the Spectrum of
Soft-Hard Nucleophiles

Nucleophilic sites in amino acids
thiol-group
of cysteine
S-atoms of
methionine
primary
amino-groups
of lysine and
arginine
secondary
amino-group
in histidine
Mapping Toxicity onto the Spectrum of
Soft-Hard Nucleophiles

Nucleophilic sites in amino acids
thiol-group
of cysteine
S-atoms of
methionine
primary
amino-groups
of lysine and
arginine
Increasing Hardness
secondary
amino-group
in histidine
Mapping Toxicity onto the Spectrum of
Soft-Hard Nucleophiles

Nucleophilic sites in amino acids
Aquatic Tox
thiol-group
of cysteine
S-atoms of
methionine
primary
amino-groups
of lysine and
arginine
Increasing Hardness
secondary
amino-group
in histidine
Mapping Toxicity onto the Spectrum of
Soft-Hard Nucleophiles

Nucleophilic sites in amino acids
Aquatic Tox
thiol-group
of cysteine
S-atoms of
methionine
primary
amino-groups
of lysine and
arginine
Increasing Hardness
secondary
amino-group
in histidine
Excess Toxicity: If Seen in Vitro – Extrapolate to in Vivo:
Pre-Michael Acceptors: Oxidised in the Air or Medium of the Assay
2.5
2- or 4-substituted
2.0
Toxicity
1.5
1.0
0.5
0.0
-0.5
3-substituted
-1.0
0
1
2
3
Log P
4
5
Information from Other Species

Is a skin sensitiser a reactive acute toxicant in fish?
Is a non-sensitiser a narcotic?

Need to map toxicity onto electrophilic spectrum

Information from Other Species

Is a skin sensitiser a reactive acute toxicant in fish?
Is a non-sensitiser a narcotic

Need to map toxicity onto electrophilic spectrum

Aquatic Toxicity
Skin Sensitisation
thiol-group
of cysteine
S-atoms of
methionine
primary
amino-groups
of lysine and
arginine
secondary
amino-group
in histidine
Reactive Groupings and Categories




Groups reactive (and hence non-reactive) chemicals
together
Allows for (Q)SAR formation and read-across
Groupings can be formed on the basis of
mechanistic knowledge
QSARs can be developed using in chemico data
O
CN
NO2
O
O
OMe
Decreasing electrophilicity  decreasing reactivity
O
O
O
Increasing steric hindrance  decreasing reactivity
O
O
Transition state effect  decreasing reactivity
O
What About Chemicals with More Than a
Single Mechanism?
O
Michael addition
Schiff base formation
NO2
O
Cl
Aromatic nucleophilic
substitution
NO2
Schiff base formation
Current Status

2-D methods



Molecular Orbital


Verhaar / Russom – type rules are accepted, have
potential to be developed further
ECOSAR classes are accepted, relate to mechanism
indirectly
Little practical use
Reactivity Measurement

Great potential; little acceptance
What is Needed In the European Union ...




Methods that work
Methods that are simple
Methods that can be justified
Methods that will be accepted by EChA, national
regulatory agencies
What is Needed In the European Union ...





Methods that work
Methods that are simple
Methods that can be justified
Methods that will be accepted by EChA, national
regulatory agencies
... science is less important?
Non-Testing Workflow
Start
Step 0: Starting step:
Information collection
Step 1: Preliminary analysis
Step 2: Use of classification
schemes
Step 3: Search for structural
alerts
Step 4: Preliminary
assessment
Step 5: Read-across
Step 6: (Q)SAR predictions
Step 7: Final assessment
End
Where Does the MOA Fit Within an ITS?
1.
2.
3.


To select relevant QSARs
To select relevant analogues for chemical category
development or read-across purposes
To rationalise/resolve disagreements in experimental data
Chemical similarity is context dependent i.e. dependent on
the relevant parameters driving the toxicity
Mechanism of Action provides the frame of reference
Conceptual ITS
Chemical
(Q)SAR, TTC
MODE/MECHANISM
In vitro
Existing data
Read-across/
Chemical Categories
33
MODE/MECHANISM
In vivo
Hazard information
Risk Assessment
Exposure information
Fish Acute Toxicity Workflow 1

Compound
Is this a single organic compound of known structure

Domain of the assay
Inside of the solubility, volatility, stability domain of the assay?

Existing Data
Do satisfactory toxicity data already exist for this compound?


Other QSAR Predictions
Can reliable predictions of toxicity be made with an ad hoc QSAR or an
expert system such as ECOSAR, TOPKAT, MultiCASE, TerraQSAR etc.?
Metabolism
Does the compound have significant and / or relevant metabolites?
Fish Acute Toxicity Workflow 2

Mechanistic Profiler
Tool
Likely Outcome
Source
Verhaar
1 /2 /3 / 4/ 5 (the five
Verhaar classes)
OECD Toolbox / Toxtree
OASIS MOA
1 / 2 /3 / 4 /5 /6 /7 (i.e. the 7 OECD Toolbox
profiles in the Toolbox) + not
known
Protein Reactivity
1 / 2 (1 = no protein alert; 2
= protein alert)
ECOSAR Classification
Not sure how to use – maybe ECOSAR and / or OECD
for categorisation?
Toolbox
OECD Toolbox
Fish Acute Toxicity Workflow 3

Category Formation
Method
Likely Outcome
Source
Formation of a category on
the basis of a structural
group
Formation of a category on
the basis of structural
similarity
Subjective according to strength
of opinion
OECD Toolbox
Subjective according to strength
of opinion and also level of
structural similarity applied
Toxmatch
Fish Acute Toxicity Workflow 3

In vitro information
Method
Likely Outcome
Source
Toxicity value from an in vitro
assay e.g. from Tetrahymena
pyriformis
Is toxicity in excess of baseline
narcosis (1 = No , narcotic; 2 =
Yes, non-narcotic; 3 = special
case where toxicity is less than
baseline – flag for concern e.g.
stability or volatility)
Direct extrapolation to fish
acute toxicity
Experimental or literature
Relationship to a particular
mechanism or mode
Relationship to a particular
mechanism or mode
Experimental or literature
Toxicity value from an in vitro
assay e.g. from Tetrahymena
pyriformis
Mechanistic information from
e.g. cell based assays
Mechanistic information from
e.g. –omics
Experimental or literature
Experimental or literature
Fish Acute Toxicity Workflow 4

In chemico information
Method
Likely Outcome
Source
Is the compound reactive in an in
1 = non-reactive; 2 = reactive [or 3 = Experimental assay e.g. Schultz or
chemico assay with a thio group e.g. low reactivity; 4 = moderate
Gerberick GSH protocol
GSH?
reactivity; 5 = high reactivity ]
Is the compound reactive in an in
chemico assay with a amine group
e.g. lysine?
1 = non-reactive; 2 = reactive [or 3 = Experimental assay e.g. lysine assay
low reactivity; 4 = moderate
(Natsch)
reactivity; 5 = high reactivity ]
Is the compound reactive in another 1 = non-reactive; 2 = reactive [or 3 = Other experimental assay
in chemico assay?
low reactivity; 4 = moderate
reactivity; 5 = high reactivity ]
Is the compound reactive in an in
chemico assay with an oxidising
species?
Does the compound form a free
radical?
1 = non-reactive; 2 = reactive[or 3 = Experimental assay e.g. Natsch
low reactivity; 4 = moderate
reactivity; 5 = high reactivity ]
1 = non-reactive; 2 = reactive [or 3 = Experimental assay
low reactivity; 4 = moderate
reactivity; 5 = high reactivity ]
Conclusions: Current Status of Reactivity in
Fish Acute Toxicity


Predicting narcotic vs non-narcotic mechanisms
Rationale grouping of chemicals



Descriptors for read-across / QSARs
Technology / models are here?
Needs




Expansion of domains
Practical workflows (to enter into ITS)
Case studies
Acceptance and implementation by industry, regulatory,
wider scientific community
Acknowledgements





This project was sponsored by Defra through the
Sustainable Arable Link Programme
European Union 6th Framework OSIRIS Integrated
Project (GOCE-037017-OSIRIS)
CAESAR Specific Targeted Project (SSPI-022674CAESAR)
European Chemicals Agency (EChA) Service Contract
No. ECHA/2008/20/ECA.203
InSilicoTox Marie Curie Project (MTKD-CT-2006-42328)
www.inchemicotox.org