- Email: [email protected]
Current use of the NTP modified one generation reproduction (MOG) study
Poster Abstracts / Reproductive Toxicology 72 (2017) 31–52
P-13
P-14
Which endpoints of a zebrafish embryotoxicity assay improve its predictivity?
Current use of the NTP modified one generation reproduction (MOG) study
Burkhard Flick 1,∗ , Xiaoqi Jiang 1 , Edward Salinas 1 , Barbara Birk 1 , Martina Dammann 1 , Jens Otte 2 , Werner Mellert 1 , Bjoern Hidding 1 , Robert Landsiedel 1 , Bennard van Ravenzwaay 1
Paul M. Foster ∗ , Mamta V. Behl, Barry S. McIntyre, Vicki L. Sutherland
1 Experimental Toxicology and Ecology, Ludwigshafen, Germany 2 Environmental Protection, Health & Safety, BASF SE, Ludwigshafen, Germany
The zebrafish model is established in various laboratories to predict developmental toxicity, but there is no standard protocol to use the detailed morphological features (scores) to assess the outcome regarding mortality, cytotoxicity, embryotoxicity or teratogenicity. In this context, we are demonstrating the potential use of the different morphological outcomes of more than 40 test compounds in different prediction models. The prediction models are considering (1) different classification of morphological alterations as indicators of general and specific embryotoxic potentials, as well as of lethal potentials of the test substances, (2) different characteristic concentrations (LOEC of general or specific embryotoxicity and lethality) and relationships between them, and (3) different concentrations ranges (up to the maximum solubility, up to 1000 M, or up to 100 M). The data set is based on public known as well as BASF SE in house chemicals with balanced different potentials of prenatal developmental toxicity in vivo. The test compounds were tested in concentration dependency up to their maximum solubility or effective concentrations, partially above 1000 M. In each test group 12 embryos with intact chorion were cultured individually in 24 well plates in 1 ml embryonic water under semi-static culture conditions between 1.5 to 120 hours post fertilization. A morphological score system with 38 parameters assessed in 3 graduations were used to describe the development of zebrafish embryos in the concentration range from the no observed adverse effect concentration up to the concentration causing mortality. The different morphological parameters were grouped as indicators for mortality, general or specific embryotoxicity. The different prediction models (PM) with different decision rules and graduations of alterations of morphological parameters lead to fluctuated accuracy (48–79%), sensitivity (0–86%), and specificity (62–100%) levels. The inclusion of LOEC for lethality in the prediction models increased the predictivity from 48–59% to 71–79%. In our data set changing the decision rule in the prediction models from >3 to >2 in ratio of LOAECs improved the levels of predictivity (up to 15%). By considering the different manifestations of effects above 100 M or 1000 M the accuracy increased by up to 7% or 8%. Based on this data set the best prediction model provides a good predictivity (accuracy 79%) by testing up to the maximum solubility of a test substance, excluding grade 1 findings for general embryotoxicity, and using a decision rule >2.
National Toxicology Program (NTP), National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA The NTP has developed a flexible approach for studying developmental and reproductive toxicity (DART) [1,2]. The MOG commences with time-mated dams exposed to the test article from gestation day (GD) 6 until weaning. At weaning, exposure continues at the same level and different cohorts are identified based on the study objectives, but typically include: (1) a group of offspring exposed for 90 days postweaning for examination of target organ toxicity; (2) a group that will have non-sibling mating of adult offspring and resulting pregnant dams examined for prenatal developmental toxicity on GD 21 and (3) a similar group, in which the dams are able to litter their offspring. NTP has completed 4 studies using this type of design, including one on the sunscreen, oxybenzone, where a comparison to ICH Segment I, II and III studies with colleagues at NCTR has shown comparable toxicity and DART findings. A developmental neurotoxicity cohort of animals has also been developed into the design which includes the measurement of locomotor, motor & sensory, and learning & memory tasks in addition to the 3 groups described above and a study utilizing all 4 cohorts is currently in-life. Future efforts using this approach have already been designed for studies on phosphate flame retardants, HIV combination therapy, herbal remedies as well as industrial agents. NTP has explored other potential uses of the MOG protocol where changes in dosing paradigm may be required for potent developmental toxicants or where repeated mating may be required to look at recovery or reproductive senescence. NTP has also established levels of evidence criteria for DART effects and the MOG protocol readily enables such conclusions to be made. For example, in the MOG study with Bisphenol AF, we detected significant changes in male (delays) and female (advancement) puberty (including some males that did not attain puberty) and two cohorts failed to produce any offspring at the highest dose evaluated (3750 ppm in the diet). Use of the MOG reduces animal usage compared to stand alone 90d, teratology, reproduction and developmental neurotoxicity studies by approximately 50%, while having improved experimental power with multiple measurements on more than 1 pup/litter to detect adverse DART outcomes. References [1] P.M. Foster, Regulatory Forum opinion piece: new testing paradigms for reproductive and developmental toxicity–the NTP modified one generation study and OECD 443, Toxicol. Pathol. 42 (2014) 1165–1167. [2] P.M. Foster, Influence of study design on developmental and reproductive toxicology study outcomes, Toxicol. Pathol. 45 (2017) 107–113.
http://dx.doi.org/10.1016/j.reprotox.2017.06.153 http://dx.doi.org/10.1016/j.reprotox.2017.06.152
37
P-13
P-14
Which endpoints of a zebrafish embryotoxicity assay improve its predictivity?
Current use of the NTP modified one generation reproduction (MOG) study
Burkhard Flick 1,∗ , Xiaoqi Jiang 1 , Edward Salinas 1 , Barbara Birk 1 , Martina Dammann 1 , Jens Otte 2 , Werner Mellert 1 , Bjoern Hidding 1 , Robert Landsiedel 1 , Bennard van Ravenzwaay 1
Paul M. Foster ∗ , Mamta V. Behl, Barry S. McIntyre, Vicki L. Sutherland
1 Experimental Toxicology and Ecology, Ludwigshafen, Germany 2 Environmental Protection, Health & Safety, BASF SE, Ludwigshafen, Germany
The zebrafish model is established in various laboratories to predict developmental toxicity, but there is no standard protocol to use the detailed morphological features (scores) to assess the outcome regarding mortality, cytotoxicity, embryotoxicity or teratogenicity. In this context, we are demonstrating the potential use of the different morphological outcomes of more than 40 test compounds in different prediction models. The prediction models are considering (1) different classification of morphological alterations as indicators of general and specific embryotoxic potentials, as well as of lethal potentials of the test substances, (2) different characteristic concentrations (LOEC of general or specific embryotoxicity and lethality) and relationships between them, and (3) different concentrations ranges (up to the maximum solubility, up to 1000 M, or up to 100 M). The data set is based on public known as well as BASF SE in house chemicals with balanced different potentials of prenatal developmental toxicity in vivo. The test compounds were tested in concentration dependency up to their maximum solubility or effective concentrations, partially above 1000 M. In each test group 12 embryos with intact chorion were cultured individually in 24 well plates in 1 ml embryonic water under semi-static culture conditions between 1.5 to 120 hours post fertilization. A morphological score system with 38 parameters assessed in 3 graduations were used to describe the development of zebrafish embryos in the concentration range from the no observed adverse effect concentration up to the concentration causing mortality. The different morphological parameters were grouped as indicators for mortality, general or specific embryotoxicity. The different prediction models (PM) with different decision rules and graduations of alterations of morphological parameters lead to fluctuated accuracy (48–79%), sensitivity (0–86%), and specificity (62–100%) levels. The inclusion of LOEC for lethality in the prediction models increased the predictivity from 48–59% to 71–79%. In our data set changing the decision rule in the prediction models from >3 to >2 in ratio of LOAECs improved the levels of predictivity (up to 15%). By considering the different manifestations of effects above 100 M or 1000 M the accuracy increased by up to 7% or 8%. Based on this data set the best prediction model provides a good predictivity (accuracy 79%) by testing up to the maximum solubility of a test substance, excluding grade 1 findings for general embryotoxicity, and using a decision rule >2.
National Toxicology Program (NTP), National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA The NTP has developed a flexible approach for studying developmental and reproductive toxicity (DART) [1,2]. The MOG commences with time-mated dams exposed to the test article from gestation day (GD) 6 until weaning. At weaning, exposure continues at the same level and different cohorts are identified based on the study objectives, but typically include: (1) a group of offspring exposed for 90 days postweaning for examination of target organ toxicity; (2) a group that will have non-sibling mating of adult offspring and resulting pregnant dams examined for prenatal developmental toxicity on GD 21 and (3) a similar group, in which the dams are able to litter their offspring. NTP has completed 4 studies using this type of design, including one on the sunscreen, oxybenzone, where a comparison to ICH Segment I, II and III studies with colleagues at NCTR has shown comparable toxicity and DART findings. A developmental neurotoxicity cohort of animals has also been developed into the design which includes the measurement of locomotor, motor & sensory, and learning & memory tasks in addition to the 3 groups described above and a study utilizing all 4 cohorts is currently in-life. Future efforts using this approach have already been designed for studies on phosphate flame retardants, HIV combination therapy, herbal remedies as well as industrial agents. NTP has explored other potential uses of the MOG protocol where changes in dosing paradigm may be required for potent developmental toxicants or where repeated mating may be required to look at recovery or reproductive senescence. NTP has also established levels of evidence criteria for DART effects and the MOG protocol readily enables such conclusions to be made. For example, in the MOG study with Bisphenol AF, we detected significant changes in male (delays) and female (advancement) puberty (including some males that did not attain puberty) and two cohorts failed to produce any offspring at the highest dose evaluated (3750 ppm in the diet). Use of the MOG reduces animal usage compared to stand alone 90d, teratology, reproduction and developmental neurotoxicity studies by approximately 50%, while having improved experimental power with multiple measurements on more than 1 pup/litter to detect adverse DART outcomes. References [1] P.M. Foster, Regulatory Forum opinion piece: new testing paradigms for reproductive and developmental toxicity–the NTP modified one generation study and OECD 443, Toxicol. Pathol. 42 (2014) 1165–1167. [2] P.M. Foster, Influence of study design on developmental and reproductive toxicology study outcomes, Toxicol. Pathol. 45 (2017) 107–113.
http://dx.doi.org/10.1016/j.reprotox.2017.06.153 http://dx.doi.org/10.1016/j.reprotox.2017.06.152
37