nomiracle: novel methods for integrated risk assessment of

NoMiracle: Novel Methods for Integrated Risk Assessment of Cumulative Stressors in Europe

Society of Toxicology Teleseminar13th June 2008

Dave Spurgeon, Claus Svendsen - Centre for Ecology and Hydrology, Wallingford, UKPete Kille, Cardiff University, Cardiff, UKSteve Stürzenbaum, Kings College, London, UKJan Baas – Vrije Universiteit, Amsterdam, The Netherlands

1

NoMiracle

Outline of the presentation:

• Science and Technology Objectives• Research Pillar on effects : Some results

– Mixture toxicity modelling– Toxicogenomics and “mode of action”– Dynamic energy budget models

• NoMiracle events and information

2

NoMiracle

Science & technology objectives

1. Refine methods for assessing the cumulative risks from combined exposures to mixtures of chemical and physical/biological agents.

2. Integration of the risk analysis of environmental and human health effects. – Open workshop Frankfurt 8/9 Sept 2008

3. To improve understanding of and develop adequate tools for effect assessment.

4. To develop a research framework for the description and interpretation of cumulative exposure and effect.

3

NoMiracle

Science & technology objectives

5. Quantify, characterise and reduce uncertainty in current risk assessment methodologies, e.g. by improving scientific basis for safety factors.

6. To develop assessment methods which take into account geographical, ecological, social and cultural differences in risk concepts and risk perceptions across Europe

7. To improve the provisions for the application of the precautionary principle and to promote its operational integration with evidence-based assessment methodologies

4

NoMiracle

KCL

NERI

ITM

RUAlterra

EKUT

UFZ

JRC

RIVM UJAG

LMCDISAV

URVUAVR CSIC

SYMLOG

SYKE

NERC UCAM

WRc ULANC

USOUTH

LHASA Ltd

VU

NIHP

WU

UALIMCO

USALZ

APINI

ETH

LEMTECRWTHA

ETC

UNIMIB

ENVI

EPFLDIA

The NoMiracle Consortium

KCL

UMFT

5

NoMiracle

RP1: RiskScenarios

RP2: ExposureStudies

RP3: EffectsStudies

RP4: RiskAssessment

Future EU Risk Policy

Consortium overall structure

6

NoMiracle

Research Pillar 3 structure

WP3.1: Chemical mixture

assessment

WP3.3: Toxicokinetics

WP3.4: Molecular mechanisms of mixture toxicity

WP3.2: Chemicals and

natural stressors

RP1: Data & Cases

RP2: Exposure information for relevant scenarios

RP4: Risk analysis based on available data

Ecotoxicology = 75%, Human health toxicology = 25%

7

NoMiracle

Science & technology objective 1

Refine methods for assessing the cumulative risks from combined exposures to mixtures of chemical and physical/biological agents.

Statistical tools for identification of interactions in mixtures

Slides from Claus Svendsen

8

NoMiracle

Models for non-interacting mixtures

• Similar action model: (concentration addition)

∑=

=n

iiTUM

1

][][

ECxi

ciTUi =

∏=

=n

iin chccch

121 )(),...,,(

• Dissimilar action model: (independent action)

9

NoMiracle

1) CA & IA 2) S/A 3A) DL or 3B) DRAdd parameter “a” Add parameter “b”

Extend the reference models by adding “deviation” parameters

The nesting allows statistical testing for best fit (using likelihood theory)

(Jonker et al ET&C 2005)

10

NoMiracle

An example: Cd + diuron in C. elegans

Building a database of binary mixture studies ~150 (collaboration with Nina Cedergreen University of Copenhagen)

Define the distribution of maximum deviations e.g. EC50 gives ECx

Species from bacteria to (in vitro) mammalian systems

Residuals of the mixture data after fitting IA to the complete data set

-80.00-60.00-40.00-20.00

0.0020.0040.0060.0080.00

0 10 20 30

Sum TU50

Res

idua

ls

Residuals of the mixture data after fitting SA to the complete data set

-80.00-60.00-40.00-20.00

0.0020.0040.0060.0080.00

100.00

0 2 4 6

Sum TU50

Res

idua

ls

Predict ECx values by IA and compare to data

11

-40.00

-20.00

0.00

20.00

40.00

60.00

80.00

100.00

120.00

-20.00 0.00 20.00 40.00 60.00 80.00 100.00

IA predicted ECx

ECx

(Dat

a &

IA S

/A M

odel

led)

IA S/A

IA

Data

IA residuals IA + S/A residuals Model fits

NoMiracle

So far binary – what about more complex?

• Build from binary mixtures knowledge

• Predict the level of deviation when mixing 3 chemicals with known binary interactions

• So far only CA and only synergism or antagonism (not dose level or dose ratio)

Ternary mixture modelling - approach

12

NoMiracle

0

2

4

6

8

10

02

46

810

24

68

10

Z D

ata

X Data

Y Data

The Experimental design Singles

13

NoMiracle

0

2

4

6

8

10

02

46

810

24

68

10

Z D

ata

X Data

Y Data

The Experimental designSingles – Binary

14

NoMiracle

0

2

4

6

8

10

02

46

810

24

68

10

Z D

ata

X Data

Y Data

The Experimental designSingles – Binary - Ternary

15

NoMiracle

0

10

20

30

40

100002 0000

30 000

40000 0200 0

400 06000

800010 000

1 20001400 016000

Mes

otrio

n E

C50

(µg

l-1)

G lyphosate EC 50 (µg l -1)

M echlorprop EC50 (

µg l-1 )

G lyM e cM e so 1

CA (reference plain)

Data analysis procedure –2 ways 1) Isobols & two step regression

16

NoMiracle

Data analysis procedure –1) Isobols & two step regression

0

2

4

6

8

10

24

68

10 02

46

810

Z D

ata

X Data

Y Data

17

NoMiracle

Data analysis procedure –1) Isobol & two step regression

0

2

4

6

8

10

24

68

10 02

46

810

Z D

ata

X Data

Y Data

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

X Data

0.1 1 10 100 1000

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

20:20:60

20:60:20

60:20:20

33:33:33

18

a1+2

a2+3

a1+3

a1+2+3

NoMiracle

0

200

400

600

800

1000

1200

24

68

1012

140 20 40 60 80 100 120 140

Z D

ata

X Data

Y Data

Visualisation of response surface

Antagonism

19

NoMiracle

0

200

400

600

800

1000

1200

24

68

1012

140

2040

6080

100120

140

Z D

ata

X Data

Y Data

0

5

10

15

20

25

30

500010000150002000025000

3000035000 0

500010000

1500020000

Mes

otrio

n EC

50 (µ

g L-1

)

Glyphosate EC50 (µg L -1) Mecoprop EC 50 (µ

g L-1 )

Results – Lemna antagonism for three herbicides

Data provided by Dr Nina Cedergreen University of Copenhagen, Denmark

20

?

?

NoMiracle

0

5

10

15

20

25

30

500010000150002000025000

3000035000 0

500010000

1500020000

Mes

otrio

n EC

50 (µ

g L-1

)


g L-1 )

a1+2=0.86

a2+3=1.95

a1+3=0.78

Results - Lemna for three herbicides

Binary interactions

21

Glyphosate – mecoprop - mesotrion

NoMiracle

0

5

10

15

20

25

30

500010000150002000025000

3000035000 0

500010000

1500020000

Mes

otrio

n EC

50 (µ

g L-1

)


g L-1 )

a1+2=0.86

a2+3=1.95

a1+3=0.78

A1+2+3=7.98

Results - Lemna

Ternary interactions

22

Glyphosate – mecoprop - mesotrion

NoMiracle

0

5

10

15

20

25

30

500010000150002000025000

3000035000 0

500010000

1500020000

Mes

otrio

n EC

50 (µ

g L-1

)


g L-1 )

a1+2=0.86

a2+3=1.95

a1+3=0.78

A1+2+3=7.98

Conclusion:The ternary mixture is more antagonistic than the combination of binary antagonisms predict.

Question:How big is the under prediction and what is the deviation from CA??

Results - LemnaGlyphosate – mecoprop - mesotrion

23

NoMiracleResults – Summary for Lemna studies

Lemna (Antagonism): For ternary if: 1) CA SA > IA = antagonism in the ternary mixture2) ΣTU CA SA < ΣTU CA SA+ = magnitude of maximum deviation in the

ternary greater than predicted based on antagonisms from the binaries.

Some ternary deviations greater than predictions for antagonism (some less). Not yet found for synergism

Chemicals

Chemical 1 Chemical 2 Chemical 3

Lemna minor:

Glyphosate Mecoprop Mesotrion

Glyphosate Mecoprop Mesotrion

Glyphosate Mecoprop Terbuthylazine

Glyphosate Mecoprop Terbuthylazine

Terbuthylazine Mesotrion Mecoprop

Terbuthylazine Mesotrion Mecoprop

Diquat Acifluorfen Mesotrion

Diquat Acifluorfen Mesotrion

ΣTU at max. dev.

bin. CA

SA

CA

SA+

1.97 2.89 2.84

1.63 1.63 2.03

1.41 1.41 1.45

1.64 1.64 2.04

2.19 2.19 2.19

1.25 1.25 1.78

2.22 2.22 2.19

1.66 1.87 1.74

24

NoMiracle

To improve understanding of and develop adequate tools for sound exposure assessment.

Toxicogenomic tools in model and non-model species


25

NoMiracle

• Free-living nematode -lives in the soil

• Bacterivorous

• Self-fertilizing, hermaphrodite

• Free-living in the soil

• Detritivorous

• Non-self-fertilizing, hermaphrodite

…the species…

1 mm10 cm

Caenorhabditis elegansLumbricus rubellus

26

NoMiracle

Getting the bits of earthworm DNA - ESTs

27

NoMiracle

Low dose = repro. EC10 = different from no dose

Cd Fluoranthene

Atrazine All exposed / controls

Patterns of effect

CtCt

Ct_Cd

Ct_AZEx_FA

Ex_FAEx_Cd

Ex_FACt_CdCt_FACt_CdCt_AZ

Ct_CdCt_AZCt_FACt_CdCt_AZ

Ct_FA

-80

-60

-40

-20

0

20

40

60

80

-100 -80 -60 -40 -20 0 20 40 60 80 100

Ct

Ct

CtCt

Ct

Ct

CtCt

Cd_43

Cd_43

Cd_43Cd_43

Axes

2

Axes 1

-60

-40

-20

0

20

40

60

-100 -80 -60 -40 -20 0 20 40 60 80 100

Ct

CtCt

CtCt

Ct

Fa_47

Fa_47Fa_47

Fa_47

Fa_47

Axe

s2

Axes 1

-100

-80

-60

-40

-20

0

20

40

60

80

100

-100 -80 -60 -40 -20 0 20 40 60 80 100

CtCt

Ct

Ct

Ct

Ct

CtCt

Az_20

Az_20

Az_20

Az_20

Az_20

Axe

s 2

Axes 1

-60

-40

-20

0

20

40

60

-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60

Ct_AZ

Ct_AZ

Ct_AZCt_AZ

Ex_AZEx_AZ

Ex_AZ

Ex_AZEx_AZ

Ct_Cd

Ct_Cd

Ct_CdEx_Cd

Ex_Cd

Ex_Cd

Ct_FA

Ct_FACt_FA

Ct_FA

Ct_FA

Ex_FA

Ex_FA

Axe

s2

Axes 1PC1 PC1

PC1 PC1

PC2

PC2

PC2

PC2

28

NoMiracle

Thymidine kinase

NAC-alpha

ATP synthase F0 subnit 6Mitochondrial ribosomal protein S24

Hypothetical protein DKFZp313I1240

Acyl CoA:monoacylglycerol acyltransferase 2

FTH1 protein

no significant hit

NADH dehydrogenase 1 alpha subcomplex

Ubiquitin ligase protein MIB1

Regucalcin

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hitno significant hit

no significant hit

Atra

zine

0 μ

g g-

1

Atr

azin

e 20

.7 μ

g g-

1

Cad

miu

m 0

μg

g-1

Cad

miu

m 4

4 μg

g-1

Fluo

rant

hene

0 μ

g g-

1

Fluo

rant

hene

47 μg

g-1

Thymidine kinase

NAC-alpha

ATP synthase F0 subnit 6Mitochondrial ribosomal protein S24

Hypothetical protein DKFZp313I1240

Acyl CoA:monoacylglycerol acyltransferase 2

FTH1 protein

no significant hit

NADH dehydrogenase 1 alpha subcomplex

Ubiquitin ligase protein MIB1

Regucalcin

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hit

no significant hitno significant hit

no significant hit

Atra

zine

0 μ

g g-

1

Atr

azin

e 20

.7 μ

g g-

1

Cad

miu

m 0

μg

g-1

Cad

miu

m 4

4 μg

g-1

Fluo

rant

hene

0 μ

g g-

1

Fluo

rant

hene

47 μg

g-1

Many unknowns

Some hypotheticals

Some other pathways e.g. protein degradation….

….but strong mitochondrial signature

Causes

29

NoMiracle

Resource allocation –dynamic energy budget

30

NoMiracle

Dose effect

aa b

b

c d c d e c d e f gc

c d e f

g h

i

g h i

Cadmium (mg L-1)

Vol

umet

ric le

ngth

(µm

)

0

70

90

110

130

150

0 0.25 0.5 1 2 3 4 6 8 10 12 14 16

aa b

b

c d c d e c d e f gc

c d e f

g h

i

g h i

Cadmium (mg L-1)

Vol

umet

ric le

ngth

(µm

)

0

70

90

110

130

150

0 0.25 0.5 1 2 3 4 6 8 10 12 14 16

Dose effect

Nematodes

Dose effect

0

0.2

0.4

0.6

0.8

1

1.2

0 200 400 600 800

Soil cadmium concentration (mg/kg)Fi

nal a

ttai

ned

wei

ght (

g)

Earthworms

Cadmium DEB effect– assimilation

31

NoMiracle

Cd changes on gene expression - nematodes

0 13 44 148 500Cadmium [mgkg-1]

Biological process Genes in Category

Genes in List in Category

p-Value

GO:31323: regulation of cellular metabolism 620 58 0.00211GO:19222: regulation of metabolism 625 58 0.00254GO:51244: regulation of cellular physiological process 641 59 0.00281GO:50794: regulation of cellular process 641 59 0.00281GO:50791: regulation of physiological process 642 59 0.00291GO:19219: regulation of nucleobase, nucleoside, nucleotide and nucleic acid metabolism 618 57 0.00316GO:6350: transcription 635 58 0.00361GO:45449: regulation of transcription 615 56 0.00451GO:6629: lipid metabolism 109 14 0.00975GO:44238: primary metabolism 2438 181 0.0107GO:8643: carbohydrate transport 22 5 0.0114GO:6355: regulation of transcription, DNA-dependent 570 50 0.0141GO:6351: transcription, DNA-dependent 581 50 0.0195GO:6139: nucleobase, nucleoside, nucleotide and nucleic acid metabolism 1024 81 0.0253GO:7155: cell adhesion 46 7 0.0262GO:50789: regulation of biological process 2085 153 0.03GO:50875: cellular physiological process 3791 265 0.0326GO:8152: metabolism 2770 198 0.0342GO:7582: physiological process 4588 316 0.0343GO:44237: cellular metabolism 2558 183 0.0417GO:6470: protein amino acid dephosphorylation 97 11 0.0461GO:44257: cellular protein catabolism 22 4 0.049GO:51603: proteolysis during cellular protein catabolism 22 4 0.049GO:19941: modification-dependent protein catabolism 22 4 0.049GO:6511: ubiquitin-dependent protein catabolism 22 4 0.049GO:6631: fatty acid metabolism 22 4 0.049GO:16311: dephosphorylation 98 11 0.0491

Condition tree Significant pathways

32

NoMiracle

– Lots of metabolic genes represented on array

– Many change expression

Cd changes on gene expression - earthwormsGlycolosis Electron transport chain

33

NoMiracle

Dose effectDose effect

Nematodes

Dose effect

Earthworms

Fluoranthene (µgL-1)

Vol

umet

ric le

ngth

(µm

)

0

120

130

140

150

0 25 50 75 100 175 250 375 500 750 1000 2000 4000

a

d

c efd e

a b g

d

a b g

c f h c e f hc

a b


Vol

umet

ric le

ngth

(µm

)

0

120

130

140

150

0 25 50 75 100 175 250 375 500 750 1000 2000 4000


Vol

umet

ric le

ngth

(µm

)

0

120

130

140

150

0 25 50 75 100 175 250 375 500 750 1000 2000 4000

a

d

c efd e

a b g

d

a b g

c f h c e f hc

a b

0

0.2

0.4

0.6

0.8

1

1.2

0 500 1000 1500

Soil fluoranthene concentration (mg/kg)

Fina

l att

aine

d w

eigh

t (g)

Fluoranthene DEB effect – somatic and maturity

maintenance

34

NoMiracle

Fluoranthene changes on gene expression


0 14 47 158553Fluoranthene [mgkg-1]



p-Value

GO:9309: amine biosynthesis 20 7 0.000291GO:44271: nitrogen compound biosynthesis 20 7 0.000291GO:6629: lipid metabolism 109 18 0.000573GO:6082: organic acid metabolism 112 17 0.0021GO:19752: carboxylic acid metabolism 112 17 0.0021GO:6397: mRNA processing 29 7 0.00336GO:16071: mRNA metabolism 29 7 0.00336GO:6519: amino acid and derivative metabolism 89 13 0.00929GO:6520: amino acid metabolism 80 12 0.00993GO:6118: electron transport 211 24 0.0136GO:6631: fatty acid metabolism 22 5 0.0166GO:6605: protein targeting 30 6 0.0167GO:6396: RNA processing 52 8 0.0282GO:6807: nitrogen compound metabolism 106 13 0.0354GO:9308: amine metabolism 106 13 0.0354GO:16070: RNA metabolism 88 11 0.0448

35

NoMiracle

Effect on body size? - Atrazine

Dose effect

Nematodes

Dose effect

Earthworms

Dose effectAtrazine DEB effect - assimilation

0

0.2

0.4

0.6

0.8

1

1.2

0 20 40 60 80

Soil atrazine concentration (mg/kg)

Fina

l att

aine

d w

eigh

t (g)

100

110

120

130

140

0 1 2.5 5 10 25 50 75 100 150 200 250 300

Atrazine (mgL-1)

Vol

umet

ric le

ngth

(µm

) a b

a b ca b c d a b c d

c d ec ef

f g f g h f g h i

gI j

g h I

j k

j k

a

100

110

120

130

140

0 1 2.5 5 10 25 50 75 100 150 200 250 300

Atrazine (mgL-1)

Vol

umet

ric le

ngth

(µm

) a b

a b ca b c d a b c d

c d ec ef

f g f g h f g h i

gI j

g h I

j k

j k

a

36

NoMiracle

Atrazine changes on gene expression


0 20 9 59 35Atrazine [mgkg-1]



p-Value

GO:7275: development 2666 186 3.13E-05GO:9791: post-embryonic development 1491 110 0.000348GO:9792: embryonic development (sensu Metazoa) 2072 145 0.000357GO:9790: embryonic development 2074 145 0.000373GO:2164: larval development 1376 102 0.000512GO:2119: larval development (sensu Nematoda) 1376 102 0.000512GO:3: reproduction 1220 91 0.000901GO:42303: molting cycle 97 12 0.00637GO:18988: molting cycle (sensu Protostomia and Nematoda) 97 12 0.00637GO:6259: DNA metabolism 200 20 0.0064GO:31323: regulation of cellular metabolism 620 48 0.00832GO:50789: regulation of biological process 2085 136 0.00942GO:19222: regulation of metabolism 625 48 0.00961GO:45449: regulation of transcription 615 47 0.0113GO:19219: regulation of nucleobase, nucleoside, nucleotide and nucleic acid metabo 618 47 0.0123GO:51244: regulation of cellular physiological process 641 48 0.0149GO:50794: regulation of cellular process 641 48 0.0149GO:50791: regulation of physiological process 642 48 0.0153GO:6355: regulation of transcription, DNA-dependent 570 43 0.0184GO:6350: transcription 635 47 0.0193GO:6351: transcription, DNA-dependent 581 43 0.0246GO:40007: growth 1856 119 0.0261GO:8643: carbohydrate transport 22 4 0.0294GO:9059: macromolecule biosynthesis 268 22 0.0364GO:44249: cellular biosynthesis 408 31 0.0382

37

NoMiracle

– Common effect on expression of metabolic genes (in the mitochondria)

– Toxicity through resource allocation (AKA the energy budget)

– Can use a combination of ecological models and transcript analysis to define exact link between toxicological and physiological “modes of action”for chemicals.

Conclusions

38

NoMiracle

To develop a research framework for the description and interpretation of cumulative exposure and effect.

Dynamic energy budget toxicity (DEBtox) models for mixtures.


Slides from Jan Baas

39

NoMiracle

LC50, NOEC, LOECNo Observed Effect No Observed Effect Concentration is a result Concentration is a result from a statistical test not from a statistical test not an indication of no effect.an indication of no effect.

High variation in the High variation in the control means: control means: NOEC>ECNOEC>EC5050

LC50, NOEC, LOEC all depend on the exposure time.

If no effect is desired why not focus on real no effect??

40

NoMiracle

Alternatives to single time point empirical approaches

Kinetics Dynamics

ExternalConcentration

InternalConcentration

Effect

Consider toxicity as a process in time using biological process based modeling to predict (no) effects

41

NoMiracle

Biological based models

time

inte

rnal

con

cent

ratio

n

time

inte

rnal

con

cent

ratio

n

elimination ra

te

internal concentration

haza

rd ra

te

internal concentration

haza

rd ra

te

NECNEC

blank valueblank value

killing ra

te

Effect depends on combination of internal concentration and hazard rate based on that internal concentration

Toxicokineticmodel

hazard ratehazard rate

toxicokineticsmodel

toxicokineticsmodel

survival probability

Hazard model Process model (DEBtox)

42

NoMiracle

concentration dieldrin ( µg l-1)time 0 3.2 5.6 10 18 32 56 100(d)0 20 20 20 20 20 20 20 201 20 20 20 20 18 18 17 52 20 20 19 17 15 9 6 03 20 20 19 15 9 2 1 04 20 20 19 14 4 1 0 05 20 20 18 12 4 0 0 06 20 19 18 9 3 0 0 07 20 18 18 8 2 0 0 0

Poecilia reticulataExample for dieldrin survival

43

NoMiracle

Poecilia reticulata

Elimination rate 0.73 d-1

Blank hazard rate 0.0064 d-1

NEC 2.8 (2.1-3.1) µg/L Killing rate 0.031 L/(µg d)

Example for dieldrin survival

44

NoMiracle

Applying process models for mixtures

Requirements for experiments for process modelling:

- Design that allows measurements at intermediate times- Design experimental setup that can find interactions

- Experiment with Folsomia candida held on compacted soil.- Exposure to 6 binary mixtures of Cd, Cu, Zn and Pb- 6 concentrations of each metal (36 combinations) in a factorial design

45

NoMiracle

Results using single time point models

Results for the mixture of Copper and Cadmium

Time point CA IA(days)2 – 3 No interaction Synergism4 – 5 Synergism Synergism6 Synergism DL Synergism7 – 9 DL Synergism DL Synergism10 – 15 No interaction DL Synergism16 DR Interaction DL Synergism17 No interaction DR Synergism18 – 21 Syn ld Antag hd DL Synergism

Comparison to Concentration Addition/Independent Action using analysis model of Jonker et al ET&C 2005

46

NoMiracle

Survival for the Cu/Cd mixture

Actually no interaction in this mixture.

Apparent interactions due to different kinetics

47

NoMiracle

DEB for complex mixtureDeveloping relations between parameters

Log NEC = - 1 Kow + 2.0

R2 = 0.91

Log b = +1 log Kow – 2.0

R2 = 0.74

Based on Pymephalus promelas data from Center for Lake Superior Environmental Studies 1985–1990

-8

-6

-4

-2

0

2

4

6

-2 0 2 4 6 8log(K ow )

log(NEC)

-6

-4

-2

0

2

4

6

8

-2 -1 0 1 2 3 4 5 6 7 8

log(Kow)

log(b)

Narcotics hazard model predicts relation between Kowand DEBtox parameters:

Log NEC = - log Kow + constantLog b (kill rate) = + log Kow + constantLog Ke (elimination rate) = -0.5 log Kow + constant

-4

-3

-2

-1

0

1

2

0 2 4 6 8

log(K ow )

log(k e)

Log NEC Log KR Log ER

All this seems to check out pretty much OK

48

NoMiracle

Experimental validation

Predicted mixture effect, based on measurements for phenantherene only (line) and use of DEBtoxparameter QSARs to predict measured mixture effect (x).

• Experiment with fully grown Folsomia candida• Test with phenantherene only• Test with mixture of phenanthrene, fluoranthene and pyrene• Measured survival over 28 days including intermediate times

0

5

10

15

20

25

30

35

0 5 10 15 20 25

time (days)

nr s

urvi

vors

sum PAH < 0.55 nmol/g dry soil

sum PAH = 0.61 nmol/g dry soil



49

NoMiracle

– Process based models such as DEBtox can describe the progression of toxicity in time

– DEBtox can be applied to mixtures and can explain why in some cases interactions are seen for particular timepoints

– For narcotic compounds they can be combined with QSAR to allow predictions of the long-term effects of complex mixtures.

Conclusions DEBtox models

50

NoMiracle51

NoMiracle

Open workshops and conferences

• Workshop: ”Integrated Assessment ofEnvironmental and Human Health”, Frankfurt:Cambridge, September 2008

• 5th NoMiracle Workshop: ”Cumulative RiskAssessment”, the Netherlands, spring 2009

• Final NoMiracle Workshop: September 2009, Aarhus, Denmark

52

NoMiracle

Acknowledgements

CEH Lindsay ListerJulian WrightJo TownsendJodie WrenPeter Hankard

CopenhagenNina Cedergreen

EdinburghMark BlaxterAnn Hedley

Cardiff John MorganJen Chasseley

AmsterdamBas KooijmanTjalling JagerKees Van GestelMieke Broerse

London (Kings/Imperial)Suresh SwainJake Bundy

53

nomiracle: novel methods for integrated risk assessment of

Documents