The developing brain is vulnerable to chemical-induced injury yet many chemicals remain

inadequately tested for developmental neurotoxicity (DNT). In an effort to develop and characterize

an in vitro model system for DNT screening, we exposed human iPSC-derived neurons to a diverse

set of 80 chemicals (e.g., neurotoxicants, drugs, pesticides, flame retardants (FRs), polycyclic

aromatic hydrocarbons (PAHs)) across a 6-point concentration range (~0.3 to 100 μM) in 384-well

plates. Using HCS imaging, effects on neurite outgrowth parameters (total outgrowth, processes,

branching) and cell viability were monitored after 72 h of exposure. Also, mitochondrial membrane

potential (MMP) was evaluated at 1 h to assess the potential contribution of MMP to altered neurite

outgrowth. The assay-specific noise threshold was calculated based on DMSO control variability and

concentration-response profiles were evaluated using a Hill model to derive benchmark

concentration (BMC) point-of-departure values. Following assay validation with controls and test

replicates, chemicals were ranked by toxicity and selectivity (i.e., effects on neurite outgrowth

parameters independent of cytotoxicity). Neurite total outgrowth and branching were the most

sensitive endpoints; 16 chemicals had an effect on neurite outgrowth independent of cytotoxicity.

The pesticide rotenone was the most selective, while1 FR (triphenyl phosphate) and 3 PAHs

(chrysene, dibenz(a,h)anthracene, acenaphthylene) inhibited neurite outgrowth. Of the 80

compounds, 39 decreased MMP and 9 were active in both assays, which might indicate that

alterations in MMP are linked with neurite outgrowth inhibition for these compounds. These studies

have important implications for moving the DNT field forward from a more traditional assessment in

animals to implementing novel and improved methodologies and should allow for a more rapid

screening and prioritization of potential neurotoxicants.

Abstract

Objective Results: BMC values

Results: Effects on neurite outgrowth &

mitochondrial function

This research was supported by the NIH, National Institute of Environmental Health

Sciences. Drs. Kevin Crofton, William Mundy, Stephanie Padilla, and Timothy Shafer at the

USEPA are acknowledged for their assistance with identifying the compounds to include in

the 80 compound library. We thank Dr. Brad Collins and Stephanie Holmgren (NTP, NIEHS)

for their assistance in assembling the 80 compound library and performing a detailed

literature search on the compounds, respectively.

Acknowledgments

References

Materials & Methods

Future Directions

Results: Assay Variability & Reproducibility

Using Human-Induced Pluripotent Stem Cells (iPSC) in High-Content Screening (HCS) to Prioritize

Chemicals for Developmental Neurotoxicity Testing Raymond Tice1, Fred Parham1, Jui-Hua Hsieh1, Mamta Behl1, Oksana Sirenko2, Evan Cromwell2 , Susan DeLaura3, Kristen Ryan1

1 NIH, NIEHS/DNTP, Research Triangle Park, North Carolina, United States; 2 Molecular Devices, Sunnyvale, California, United States; 3 Cellular Dynamics International, Madison, Wisconsin, United States

1. Characterize the neurite outgrowth assay as a suitable high throughput screen for a NTP

neuro-specific assay battery.

2. Explore the relationship between mitochondrial dysfunction and inhibition of neurite outgrowth.

3. Prioritize NTP chemicals of concern (e.g., FRs, PAHs) for more comprehensive in vivo hazard

characterization studies.

NT/DNT (38) NTP Nominations/Others (33)

1-methyl-4-phenylpyridinium iodide 2- Ethylhexyl diphenyl phosphate (EDHP)*

2-Methoxyethanol 2,2',4,4',5,5'-Hexabromodiphenyl ether*

3,3'-Iminodipropionitrile 2,2',4,4',5-Pentabromodiphenyl ether*

5-Fluorouracil 2,2'4,4'-Tetrabromodiphenyl ether*

6-Hydroxydopamine hydrochloride 3,3’,5,5’-Tetrabromobisphenol A*

6-Propyl-2-thiouracil Isodecyl diphenyl phosphate*

Acetic acid, manganese(2+) salt Phenol, isopropylated, phosphate (3:1)*

Acrylamide Tert-Butylphenyl diphenyl phosphate*

Aldicarb Tricresyl phosphate*

Bis(tributyltin)oxide Triphenyl phosphate*

Bisphenol A Tris(2-chloroethyl) phosphate*

Captan 4-H-Cyclopenta(d,e,f)phenanthrene†

Carbaryl Acenaphthene†

Chlorpyrifos (Dursban) Acenaphthylene†

Colchicine Anthracene†

Deltamethrin Benz(a)anthracene†

Di(2-ethylhexyl) phthalate Benzo(a)pyrene†

Diazepam Benzo(b)fluoranthene†

Dichlorodiphenyltrichloroethane (DDT) Benzo(e)pyrene†

Dieldrin Benzo(k)fluoranthene†

Diethylstilbestrol Benzo[g,h,i]perylene†

Heptachlor Chrysene†

Hexachlorophene Dibenz(a,h)anthracene†

Hydroxyurea Dibenz[a,c]anthracene†

Lindane Fluorene†

Methyl mercuric (II) chloride Naphthalene†

n-Hexane Phenanthrene†

Parathion Pyrene†

Permethrin 1-ethyl-3-methylimidazolium diethylphosphate¥

Phenobarbital sodium salt 2,3,7,8-Tetrachlorodibenzo-p-dioxin¥

Phenobarbitol Berberine chloride¥

Rotenone Carbamic acid, butyl-, 3-iodo-2-propynyl ester¥

Tebuconazole Manganese, tricarbonyl[(1,2,3,4,5-.eta.)-

Tetraethylthiuram disulfide 1-methyl-2,4-cyclopentadien-1-yl]-¥

Thalidomide Negative Controls (5)

Toluene Acetaminophen

Valinomycin Acetylsalicylic acid

Valproic acid sodium salt D-Glucitol

L-Ascorbic acid

Saccharin sodium salt hydrate

NT/DNT = compounds with reported neurotoxicity and developmental

neurotox1,2

Test replicates in italics were used to examine assay reproducibility

NTP Nominations = * Flame retardants (FRs) and † Polycyclic aromatic

hydrocarbons (PAHs) ¥ Others

iPSC-derived iCell® Neurons

Cells were received cryopreserved from Cellular

Dynamics International (CDI, Madison WI)

• Cells were thawed and plated according to CDI’s

protocol.

• iCell neuron methods and use in high-content toxicity

assays were previously evaluated by Sirenko et al.

(2014).

Neurite Outgrowth Assay

• Compounds were tested in duplicate at each

concentration (0.3 -100 μM) for 72 hr on laminin

coated 384–well plates with ~7500 cells/well.

• Cells were stained with Calcein AM & Hoescht dye

(Invitrogen, Carlsbad, CA) for 30 min prior to imaging

on an Image Xpress XL Instrument (Molecular

Devices, Sunnyvale, CA).

• Images were acquired using 10X or 20X objectives

and multiple exposures.

Mitochondrial Membrane Potential (MMP) Assay

• Compounds tested in duplicate at each concentration

(0.1-30 μM) for 1 hr on laminin coated 384–well plates

with ~7500 cells/well.

• Effects were monitored by the addition of the

mitochondria active dye JC-10 (ATT BioQuest Inc.,

Sunnyvale, CA) for 30 mins prior to imaging on an

Image Xpress XL Instrument (Molecular Devices,

Sunnyvale, CA).

• Images were acquired using 10X or 20X objectives

and multiple exposures.

Figure 1. Neurite Outgrowth Endpoints

Viable Cells: Number of viable cell bodies in the

image

Total Outgrowth: Total length of skeletonized

outgrowth in mm (corrected for diagonal lengths)

Total Processes: Number of outgrowths in the

image that are connected to cell bodies

Total Branches: Total number of branching

junctions in the image

Cell Body

Processes (4)

connected to

cell body Branch

Benchmark Concentration (BMC) Analysis

Dose-response information is critical for ranking

chemicals for toxicity or safety. However, robust

calculation of half-maximum activity responses

(i.e., AC50) is limited by the need for two

asymptotes. High concentration response

plateaus are often not present, leading to a need

for an improved analysis method. To overcome

this limitation, we applied the BMC analysis

method developed by the U.S. Environmental

Protection Agency for analysis of toxicity test

data3,4 to the multi-parametric neurite outgrowth

assay. Response data were normalized by

dividing by the mean of the dimethyl sulfoxide

(DMSO) control wells. Concentration-response

was modeled using a Hill model. The model

parameters were optimized by least-squares

fitting using MATLAB.

Model equation: f(x)=v0 + (vmax-v0)* xn/(kn + xn);

v0 is fixed at 1 (mean value of DMSO control data)

Figure 2. (A) Box plot showing the variance within DMSO controls per endpoint for 3 standard deviations (SDs). (B) 3SD values were used as a the benchmark response [BMR] (i.e., noise threshold) for data modelling and

benchmark concentration (BMC) calculations. (C) Concentration response curves and BMCs (μM) calculated for duplicates within the 80 compound library. For each plot, data points are represented by an asterisk (*) for

each concentration [log10 scale] for each endpoint. The Hill model was used to fit the data and estimate the BMC [Filled Circles: Red = Total Outgrowth; Green = Total Processes; Blue = Total Branches; Black = Viable cells].

A C

Ray I am putt

Results: Compound effects on mitochondrial membrane potential

Selective Compounds in the

Neurite Outgrowth Assaya

Category

Neurite

Outgrowth

BMC (μM)

MMP

BMC (μM)

Rotenone NT/DNT 1.37 1.60

Dieldrin NT/DNT 4.91 0.88

Hexachlorophene NT/DNT 8.12 0.06

Dibenz(a,h)anthracene PAH 7.91 30.00

Diethylstilbestrol NT/DNT 20.84 3.97

Aldicarb NT/DNT 21.14 30.00

Chrysene PAH 10.11 30.00

Carbaryl NT/DNT 21.65 1.29

Diazepam NT/DNT 21.82 30.00

Acenaphthylene PAH 24.14 12.59

Di(2-ethylhexyl) phthalate NT/DNT 25.06 30.00

Triphenyl phosphate FR 13.33 6.24

Methyl mercuric (II) chloride NT/DNT 0.55 1.86 Manganese, tricarbonyl[(1,2,3,4,5-.eta.)-1-

methyl-2,4-cyclopentadien-1-yl]- Other 28.52 30.00

Tebuconazole NT/DNT 29.14 30.00

Tetraethylthiuram disulfide NT/DNT 10.29 6.61

a compounds listed in order of selectivity scores (SS) (high to low) for neurite outgrowth by the most sensitive

endpoint

Gray colored box: compounds affected both neurite outgrowth and MMP < 30 μM

NT/DNT = neurotoxicant/developmental neurotoxicant

FR = flame retardant; PAH = polycyclic aromatic phosphate; BMC = benchmark concentration value

Compound BMC (µM)

Valinomycin 0.04

Hexachlorophene 0.06

1-ethyl-3-methylimidazolium diethylphosphate 0.51

Triphenyl phosphate 0.76

Dieldrin 0.88

Carbaryl 1.29

Rotenone 1.60

Methyl mercuric (II) chloride 1.86

Phenol, isopropylated, phosphate (3:1) 2.65

Carbamic acid, butyl-, 3-iodo-2-propynyl ester 2.81

A

B

C

1. Mundy et al. (2009) The Toxicologist

2. Crofton et al. (2011) ALTEX

3. US EPA (2012) Benchmark Dose Technical Guidance

http://www.epa.gov/raf/publications/pdfs/benchmark_dose_guidance.pdf

4. Robin X et al. (2011) BMC Bioinformatics 12: 77

• Automated high throughput, high content imaging can be utilized to evaluate the

effects of chemicals on neurite outgrowth in human iPSCs.

• Based on the variability in the solvent control wells, an assay specific SD was

calculated and 3 SD was used as a conservative measure of an acceptable

benchmark response (BMR) for BMC analysis.

• The endpoint with the least noise in the solvent control data was viable cells

(measure of cytotoxicity), followed by neurite total outgrowth, total processes, and

total branches.

• The majority of actives in this assay were identified by effects on total branching >

total outgrowth > total processes.

• Duplicate chemicals show similar concentration responses & calculated BMC values

indicating good internal assay reproducibility.

• 9/76 unique compounds inhibited both MMP (1 hr) and neurite outgrowth (72 hr) and

belong to diverse structural classes such as pesticides, drugs, metal-containing

compounds, FRs, and PAHs, suggesting that mitochondrial dysfunction may be an

upstream signaling event of neurite outgrowth inhibition.

• Compare the results from this assay with data generated by collaborators and in the

literature.

• In vitro actives in neurite outgrowth assay

• Known in vivo neurotoxicants and developmental neurotoxicants

• Prioritize suspected neurotoxicants or chemical classes of NTP interest (i.e., FRs,

PAHs) for in vivo hazard characterization & targeted/mechanistic follow-up studies.

V(0)

AC50 (half-

maximal activity

concentration)

BMC

Vmax Hill Model

BMR for

each

endpoint

Accurate

estimation

depends on high

quality data curve

Resp

on

se

Dose

B

Total

Outgrowth

Total

Processes

Total

Branches

Viable

Cells

Saccharin sodium salt-1 Saccharin sodium salt-2 Methyl mercuric (II) chloride-1 Methyl mercuric (II) chloride-2

Deltamethrin-1 Deltamethrin-2 Triphenyl phosphate-1 Triphenyl phosphate-2

Re

sp

on

se

3SD=0.09 3SD=0.21 3SD=0.20 3SD=0.11

Figure 3: Stem plot showing a ranking of the 80 compounds by lowest BMC per endpoint.

For all compounds with a BMC value < 100 μM, a comparison is shown between the BMC for the

most sensitive endpoint and the BMC for viability as a marker of cytotoxicity. For several

compounds, the BMC was not estimated because no effect was observed over the concentration

range tested (0.3 – 100 μM). BMC = benchmark concentration.

BMC (μM)

Ranking of all 80 Compounds: Lowest BMC

Total Outgrowth

Total Processes

Total Branches

Viable Cells

KEY

Total Outgrowth

Total Processes

Total Branches

Viable Cells

KEY

• 41/80 (51%) of the compounds had a effect on at least 1 endpoint in

the neurite outgrowth assay

BMC μM

Results: Compound-specific effects

on neurite outgrowth

Control 0.3 μM Rotenone 3 μM Rotenone

Figure 5. High content imaging and dose

response modeling for rotenone. (Left)

Modelled concentration response curves

from duplicate data points (*) and BMCs for

each neurite endpoint demonstrate that

effects on neurites occur at lower

concentrations than effects on cell viability.

BMC = benchmark concentration and is

represented by a symbol (circle or triangle).

(Below) Neuronal bodies and neurites were

visualized (20X) with Calcein AM dye (green)

following a 72 hr exposure using iPSC-

neurons.

Figure 4: Stem plot showing 16 compounds which were identified as selective for inhibition

of neurite outgrowth. Selectivity is the extent to which the BMC for the most sensitive neurite

endpoint is less than the BMC for cytotoxicity. Selectivity was calculated for compounds with BMCs

< 100 μM by subtracting the log10 BMC for each endpoint from the log10 BMC value for viable cells

(marker of cytotoxicity). BMC values of > 100 µM for viable cells are treated as 100 µM. This will

underestimate selectivity. Selectivity calculations of 0.5 (i.e., the log10 difference between two

consecutive concentrations) was required for a compound to be deemed selective. BMC =

benchmark concentration.

16 compounds show selective inhibition of neurite outgrowth

Rotenone

Table 1: 80 compound library

Table 2: Top 10 inhibitors of MMP

Table 3. Evaluating the contribution of mitochondrial function on neurite outgrowth

• 39/80 (49%) of the compounds had a effect on mitochondrial membrane potential (MMP)

BMC (μM)

Disclaimer: This data does not reflect the final opinion of the National Toxicology Program

Conclusions

Figure 6. Heatmap of BMC values for all compounds tested in the 1 hr MMP assay. Blue = decreasing responses in MMP or cell

viability. White = no effect in the concentration range tested (0.1 – 30 μM).

30

μM

0

10

0.1

0.1

5