used for the larval mobility assay or collected
for gene expression analysis.
RNA extraction and qPCR analysis in D. rerio
Pooled zebrafish larvae were used for collec-
tion of RNA using a TissueLyser (Qiagen) for
homogenization and the RNeasy Plus Mini Kit
(Qiagen) for isolation of RNA. After quanti-
fication with NanoDrop, cDNA was synthe-
sized with an iScript cDNA synthesis kit
(BioRad) and RT-qPCR was carried out with the
SsoAdvanced Universal SYBR-Green Supermix
(BioRad). The primers and genes that were
tested are listed in table S6. More detailed in-
formation about the protocols used, as well as
for the analysis of gene expression in zebrafish,
is provided in the the supplementary materials.
D. reriolarvae mobility
Mobility of larval zebrafish was recorded with
the ViewPoint automatic behavior tracking
system (ViewPoint Life Science) and an in-
frared camera. Larvae were transferred to indi-
vidual wells in 48-well plates, and a protocol
of alternating dark-light cycles (5 minÐ5 min)
was used to induce locomotion ( 106 ). Distance
traveled was recorded for a total of 40 min
and analyzed with the ViewPoint Zebralab
software (ViewPoint Life Science). Three
independent experiments with three techni-
cal replicates each (n= 6 to 8 per replicate)
were performed, and results are presented
as a pool of the three experiments. Further
information on the mobility assay and analysis
is available in the supplementary materials.
SMACH
We conducted SMACH using the SELMA
pregnancy cohort linking human exposures
with estimates of adverse outcomes in zebra-
fish andXenopus. In short, we modeled the
total distance each zebrafish swam using four
connected 10-min flexible curves fit using trun-
cated power cubic splines (fig. S7A). Charac-
teristics of these curves were evaluated across
zebrafish as a function of exposure concen-
trations. Herein, the average maximum dis-
tance across the four cycles was modeled to
reflect the change in swimming patterns
due to exposure to MIX N. We used a piece-
wise nonlinear model to approximate the
concentration-response relationship splines
(fig. S7B). For the XETA, results from three
individual experiments were combined with
the T3 condition as control. The read out was
a normalized level of fluorescence (where the
control mean is 100%). A piecewise model
was used to approximate the concentration-
response relationship (fig. S7C).
The SMACH was conducted by first testing
whether the concentrations in SELMA preg-
nant women were sufficiently similar to the
domain-specific reference mixture ( 80 ). For
thosedeterminedtobesufficientlysimilar,the
SMRI was calculated (the sum of concentra-
tions relative to the estimated BMDL from ex-
perimental studies). The BMR was based on
either established guideline values (XETA) or a
one-SD decrease from the control mean (zebra-
fish average maximum distance). The percent-
agesofwomenwithandSMRIexceeding1were
tabulated to provide a level of concern. Finally,
adjusted models were used to determine if
children associated with the highest and low-
est decile of SMRI were significantly (P< 0.05)
different in language delay for MIX N.Ethics issues
The EDC-MixRisk project maintains both ex-
ternal and internal reviews of the ethically
relevant aspects of the project. Each partner
is subjected to H2020 self-assessment pro-
cesses and to reviews by an internal ethics
advisory board (EAB). The EAB is composed of
two leading scholars with expertise in bio-
ethics and research policy. The project as a
whole and each of the participating partners
comply with Horizon 2020 ethics policies as
well as with national legislations. Each part-
ner follows internal guidelines concerning the
collection, handling, and securing of informed
consent; health-related information and data
in the epidemiological studies; human bio-
logical samples in the epidemiological studies;
data processing, dissemination, and privacy;
risk to research subjects; and human embry-
onic stem cells and human fetal tissues. In
addition, depending on the various national
legislations and the mandates of the local
institutions, local institutional review boards
(IRBs) from the various partners were also
involved in the relevant aspects of the project.
Finally, the project has also allocated one team
member to the coordination of ethics issues.
Regarding human embryonic stem cells,
following the regulation establishing Horizon
2020 ( 107 ) and the Horizon 2020 ethics self-
assessment guide ( 108 ), only established cell
lines registered on either the European ( 109 )
or National Institutes of Health (NIH) ( 110 )
registries have been used. The use of stem
cell lines is also subject to the laws of the
countries of the participating partners: Italy,
France, and Sweden. The procurement and
use of fetal tissues upon informed consent
from women undergoing elective abortions
was approved by the relevant local ethics
committee (University of Edinburgh) and,
where necessary, subjected to additional review
bytheEABandapprovalbyrelevantcommis-
sion officers. Procurement of human samples
followed informed consent in all cases, and
samples were codified (pseudo-anonymized).REFERENCESANDNOTES- A. C. Goreet al., EDC-2: The Endocrine Society’s Second
Scientific Statement on endocrine-disrupting chemicals.
Endocr. Rev. 36 , E1–E150 (2015). doi:10.1210/er.2015-1010;
pmid: 26544531
2. A. Kortenkamp, Low dose mixture effects of endocrine
disrupters and their implications for regulatory thresholds in
chemical risk assessment.Curr. Opin. Pharmacol. 19 , 105– 111
(2014). doi:10.1016/j.coph.2014.08.006; pmid: 25244397
3. Å. Bergman, J. J. Heindel, S. Jobling, K. A. Kidd, R. T. Zoeller,
Eds.,“State of the science of endocrine disrupting chemicals -
2012 ”(Technical Report, World Health Organization, 2013).
4. C. Cheroni, N. Caporale, G. Testa, Autism spectrum disorder
at the crossroad between genes and environment:
Contributions, convergences, and interactions in ASD
developmental pathophysiology.Mol. Autism 11 , 69 (2020).
doi:10.1186/s13229-020-00370-1; pmid: 32912338
5. H. Peyre, M.-L. Charkaluk, A. Forhan, B. Heude,
F. Ramus, EDEN Mother–Child Cohort Study Group, Do
developmental milestones at 4, 8, 12 and 24 months predict
IQ at 5-6 years old? Results of the EDEN mother-child cohort.
Eur. J. Paediatr. Neurol. 21 , 272–279 (2017). doi:10.1016/
j.ejpn.2016.11.001; pmid: 27889381
6. E. Vlasblomet al., Predictive validity of developmental
milestones for detecting limited intellectual functioning.PLOS
ONE 14 , e0214475 (2019). doi:10.1371/journal.pone.0214475;
pmid: 30921424
7. T. Flensborg-Madsen, E. L. Mortensen, Associations of early
developmental milestones with adult intelligence.Child Dev. 89 ,
638 – 648 (2018). doi:10.1111/cdev.12760; pmid: 28198006
8. M. Marinopoulou, E. Billstedt, P.-I. Lin, M. Hallerbäck,
C.-G. Bornehag, Number of words at age 2.5 years is
associated with intellectual functioning at age 7 years in the
SELMA study.Acta Paediatr. 110 , 2134–2141 (2021).
doi:10.1111/apa.15835; pmid: 33686710
9. C.-G. Bornehaget al., Association of prenatal phthalate
exposure with language development in early childhood.
JAMA Pediatr. 172 , 1169–1176 (2018). doi:10.1001/
jamapediatrics.2018.3115; pmid: 30383084
10. M. Westerlund, C. Sundelin, Screening for developmental
language disability in 3-year-old children. Experiences from a
field study in a Swedish municipality.Child Care Health Dev.
26 , 91–110 (2000). doi:10.1046/j.1365-2214.2000.00171.x;
pmid: 10759750
11. J. R. Rochester, Bisphenol A and human health: A review of
the literature.Reprod. Toxicol. 42 , 132–155 (2013).
doi:10.1016/j.reprotox.2013.08.008; pmid: 23994667
12. A. C. Goreet al., EDC-2: The Endocrine Society’s second
Scientific Statement on endocrine-disrupting chemicals.
Endocr. Rev. 36 , E1–E150 (2015). doi:10.1210/er.2015-1010;
pmid: 26544531
13. C.-G. Bornehaget al., The SELMA study: A birth cohort study
in Sweden following more than 2000 mother-child pairs.
Paediatr. Perinat. Epidemiol. 26 , 456–467 (2012).
doi:10.1111/j.1365-3016.2012.01314.x; pmid: 22882790
14. European Chemicals Agency (ECHA), Candidate List of
substances of very high concern for authorisation;www.echa.
europa.eu/en/web/guest/candidate-list-table.
15. L. L. McIntyre, W. E. Pelham III, M. H. Kim, T. J. Dishion,
D. S. Shaw, M. N. Wilson, A Brief Measure of Language Skills
at 3 Years of Age and Special Education Use in Middle
Childhood.J. Pediatr. 181 , 189–194 (2017). doi:10.1016/
j.jpeds.2016.10.035; pmid: 27908645
16. C. Carrico, C. Gennings, D. C. Wheeler, P. Factor-Litvak,
Characterization of weighted quantile sum regression for
highly correlated data in a risk analysis setting.J. Agric. Biol.
Environ. Stat. 20 , 100–120 (2015). doi:10.1007/s13253-014-
0180-3; pmid: 30505142
17. O. S. Andersonet al., Epigenetic responses following
maternal dietary exposure to physiologically relevant levels of
bisphenol A.Environ. Mol. Mutagen. 53 , 334–342 (2012).
doi:10.1002/em.21692; pmid: 22467340
18. L. N. Vandenberg, T. Colborn, T. B. Hayes, J. J. Heindel,
D. R. Jacobs Jr., D.-H. Leeet al., Hormones and endocrine-
disrupting chemicals: Low-dose effects and nonmonotonic
dose responses.Endocr. Rev. 33 , 378–455 (2012).
doi:10.1210/er.2011-1050; pmid: 22419778
19. B. S. Abrahamset al., SFARI Gene 2.0: A community-driven
knowledgebase for the autism spectrum disorders (ASDs).
Mol. Autism 4 , 36 (2013). doi:10.1186/2040-2392-4-36;
pmid: 24090431
20. T. Wanget al., De novo genic mutations among a Chinese
autism spectrum disorder cohort.Nat. Commun. 7 , 13316
(2016). doi:10.1038/ncomms13316; pmid: 27824329
21. S. J. Sanderset al., De novo mutations revealed by
whole-exome sequencing are strongly associated with
autism.Nature 485 , 237–241 (2012). doi:10.1038/
nature10945; pmid: 22495306
Caporaleet al.,Science 375 , eabe8244 (2022) 18 February 2022 13 of 15
RESEARCH | RESEARCH ARTICLE