Thalidomide induced early gene expression perturbations indicative of

Prévia do material em texto

Thalidomide induced early gene expression
human embryopathy in mouse embryonic st
Xiugong Gao ⁎, Robert L. Sprando, Jeffrey J. Yourick
Division of Toxicology, Office of Applied Research and Safety Assessment, Center for Food Safety and
ng h
um
pon
tion
mid
ere
M
ssed
ys a
ved in small GTPases-mediated signal transduction, heart development, and inflam-
2010)
oxicolo
her thr
dy, ide
odels
sting h
all somatic cell types
ally developed as an
based on the interfer-
ouse embryonic stem
Toxicology and Applied Pharmacology 287 (2015) 43–51
Contents lists available at ScienceDirect
Toxicology and Appl
j ourna l homepage: www.e l
cells (mESCs) into beating cardiomyocyte foci in culture (Heuer et al.,
1993; Spielmann et al., 1997), and was successfully validated by the
European Center for the Validation of Alternative Methods (ECVAM)
BP_ALL, GO term in biological process at all levels; GSK3b-I, glycogen synthase kinase-3b
inhibitor; HCM, hypertrophic cardiomyopathy; hESCs, human embryonic stem cells; IPA,
Ingenuity Pathway Analysis; IVT, in vitro transcription; KEGG, Kyoto Encyclopedia of
sion and unique potential to differentiate into
(Tandon and Jyoti, 2012). The EST was initi
in vitromodel for the screening of embryotoxicity
ence of chemicals with the differentiation of m
process; cRNA, complimentary RNA; DAVID, Database for Annotation, Visualization, and
Integrated Discovery; DEGs, differentially expressed genes; DMSO, dimethyl sulfoxide;
dUTP, deoxyuridine triphosphate; EB, embryoid body; ECVAM, European Center for the
Validation of Alternative Methods; ESCs, embryonic stem cells; EST, embryonic stem cell
test; FC, fold change; GEO, Gene Expression Omnibus; GO, gene ontology; GOTERM_
culture (WEC) (New et al., 1976), and the mouse embryonic stem cell
test (EST) (Schulpen and Piersma, 2013). Embryonic stem cells (ESCs)
have gained considerable interest for their use in developmental toxic-
ity testing by virtue of their fundamental attribute of unlimited expan-
Abbreviations: AGCC, Affymetrix GeneChip Command Console; ANOVA, analysis of
variance; APE 1, apurinic/apyrimidinic endonuclease 1; ARVC, arrhythmogenic right
ventricular cardiomyopathy; BMP-4, bone morphogenic protein 4; BP, biological
Genes and Genomes; LIF, leukemia inhibitory factor; m
cells; MM, micromass; PBS, phosphate buffered saline; RM
TAC, Transcriptome Analysis Console; TdT, terminal deox
uracil-DNA glycosylase; WEC, whole embryo culture.
⁎ Corresponding author at: 8301 Muirkirk Road, Laurel
E-mail address: xiugong.gao@fda.hhs.gov (X. Gao).
http://dx.doi.org/10.1016/j.taap.2015.05.009
0041-008X/Published by Elsevier Inc.
as traditionally relied on
ng large numbers of ani-
d of organogenesis and
(Xenopus laevis), and zebrafish (Danio rerio) (Lein et al., 2005). Exam-
ples of alternative in vitro test systems include the limb bud micromass
(MM) (Flint and Orton, 1984), the rat postimplantation whole embryo
animal models which typically involve exposi
mals to chemicals during the critical perio
Mouse
Introduction
The Tox21 program (Shukla et al.,
federal agencies calls for transforming t
al in vivo tests to less expensive and hig
to prioritize compounds for further stu
tion and ultimately develop predictivem
in humans. Developmental toxicity te
omide. These results demonstrate that transcriptomics in combination with mouse embryonic stem cell
differentiation is a promising alternative model for developmental toxicity assessment.
Published by Elsevier Inc.
partnered by several US
gy testing from tradition-
oughput in vitromethods
ntify mechanisms of ac-
for adverse health effects
subsequently examining fetuses for visceral and skeletalmalformations,
growth, and viability. These approaches are costly, time consuming and
low throughput (Spielmann, 2009). Over the last few decades, a multi-
tude of alternative test systems have been developed to refine, reduce,
or replace the traditional animal tests for assessing developmental tox-
icity. Examples of in vivo nonmammalian models include nematode
(Caenorhabditis elegans), fruit fly (Drosophila melanogaster), frog
Differentiation
Microarray
m
atory responses, which coincide with clinical evidences and may represent critical embryotoxicities of thalid-
Embryonic stem cell
Developmental toxicity
terms and canonical pathwa
geneswere found to be invol
a b s t r a c ta r t i c l e i n f o
Article history:
Received 4 February 2015
Revised 23 April 2015
Accepted 14 May 2015
Available online 23 May 2015
Keywords:
Thalidomide
Transcriptomics
Developmental toxicity testi
require the sacrifice of large n
ings and animals in their res
causes severe limbmalforma
changes induced by thalido
(mESCs). C57BL/6 mESCs w
72 h after exposure to 0.25 m
dreds of differentially expre
ESCs, mouse embryonic stem
A, robust multi-array average;
ynucleotidyl transferase; UDG,
, MD 20708, United States.
perturbations indicative of
em cells
Applied Nutrition, U.S. Food and Drug Administration, Laurel, MD, United States
as traditionally relied on animal models which are costly, time consuming, and
bers of animals. In addition, there are significant disparities between human be-
ses to chemicals. Thalidomide is a species-specific developmental toxicant that
s in humans but not inmice. Here, we usedmicroarrays to study transcriptomic
e in an in vitro model based on differentiation of mouse embryonic stem cells
allowed to differentiate spontaneously and RNA was collected at 24, 48, and
thalidomide. Global gene expression analysis using microarrays revealed hun-
genes upon thalidomide exposure that were enriched in gene ontology (GO)
ssociated with embryonic development and differentiation. In addition, many
ied Pharmacology
sev ie r .com/ locate /ytaap
(Genschow et al., 2004).
Thalidomide (α-phthalimidoglutarimide) was synthesized in West
Germany in 1953 and launched in 1957 for the treatment of nausea
and vomiting during pregnancy. It was subsequently withdrawn from
market in 1961 after its teratogenic effects in humans were recognized.
cells of endodermal, ectodermal and mesodermal origin were obtained
in the outgrowths. In EST, differentiation was determined by micro-
that contained 100 μl medium. After 3 h incubation at 37 °C, the resul-
tant absorbance was recorded at 490 nm using a SpectraMax i3 plate
and purity (260/280 ratio) were measured with the NanoDrop 2000
UV–Vis spectrophotometer (NanoDrop Products, Wilmington, DE). In-
44 X. Gao et al. / Toxicology and Applied Pharmacology 287 (2015) 43–51
Thousands of congenitally malformed children have been reported
worldwide (Diggle, 2001). Themalformations include limbs and ocular,
respiratory, gastrointestinal, urogenital, cardiovascular, and nervous
systems aswell (Ito et al., 2011; Vargesson, 2009). Limb defects majorly
include phocomelia, amelia, micromelia, oligodactyly, and syndactyly
(Vargesson, 2009). Despite its strong embryotoxicity, thalidomide was
reintroduced in 1998, after more than 35 years disappearing from
the market, as an immunomodulator for the treatment of erythema
nodosum leprosum, and has since found several other applications in-
cluding the treatment of cancer (Okafor, 2003; Teo et al., 2005).
The EST uses cytotoxicity assays to determine embryotoxicity of
chemicals. We argue that developmental effects at physiologically rele-
vant doses may not necessarily result in cytotoxicity, but rather more
subtle effects that can only be detected using more sensitive methods.
Transcriptome profiling of ESC differentiation can probe all possible al-
terations in cellular differentiation causing deviation from the normal
developmental progression and has been proposed as a promising
model for developmental toxicity testing (Winkler et al., 2009; van
Dartel and Piersma, 2011). We have recently demonstrated the useful-
ness ofthis method in a mESC model using a C57BL/6 cell line (Gao
et al., 2014).
It is well known that not all animal species respond equally to
the developmental toxicity of chemical compounds. Notably, thalido-
mide elicited highly variable responses in the many animal species
studied. In primates and a few strains of rabbits severe congenital
malformations (as described above for humans) were reported, but
only moderate effects were found in rats and no significant changes
were observed in mice (Fratta et al., 1965; Schumacher et al., 1968;
Teo et al., 2001, 2004). For this reason, transcriptomic studies on thalid-
omide have mainly used human ESCs (hESCs) (Mayshar et al., 2011;
Meganathan et al., 2012). It has been suggested that one possibility for
thalidomide non-embryopathy inmouse is that it does not pass through
the mouse placenta (Therapontos et al., 2009). If this is the case, then it
is reasonable to hypothesize that thalidomide in direct contact with
mESCs in culture would cause perturbations in the differentiation path-
ways detectable by transcriptome profiling.
Previously, we have demonstrated the usefulness of an in vitro
model for developmental toxicity assessment based on transcriptomic
profiling of mESCs (Gao et al., 2014). In the current study, we used
this model to assess the embryotoxicity of thalidomide. We demon-
strated that thalidomide induced gene expression perturbations indica-
tive of human embryopathy in mESCs that could be detected in as short
as 24 h.
Materials and methods
Materials. (±)-Thalidomide ((RS)-2-(2,6-dioxopiperidin-3-yl)-1H-
isoindole-1,3(2H)-dione) and all other chemicals used in this study
were of molecular biology grade and were obtained from Sigma-
Aldrich (St. Louis, MO) unless otherwise stated.
Pluripotent mouse embryonic stem cell culture. Pluripotent ESGRO Com-
plete Adapted C57BL/6 mESCs, which have been pre-adapted to
serum-free and feeder-free culture condition, were obtained from
EMD Millipore (Billerica, MA) at passage 12 (with 80% normal male
mouse karyotype). The cells were seeded in cell culture flasks (Nunc,
Roskilde, Denmark) coated with 0.1% gelatin solution (EMDMillipore),
and maintained at 37 °C in a 5% CO2 humidified incubator at standard
densities (i.e., between 5 × 104/cm2 and 5 × 105/cm2) in ESGRO Com-
plete Plus Clonal GradeMedium (EMDMillipore). Themedium contains
leukemia inhibitory factor (LIF), bone morphogenic protein 4 (BMP-4),
and a glycogen synthase kinase-3b inhibitor (GSK3b-I) to helpmaintain
pluripotency and self-renewal of the ESCs. Cells were passaged every 2–
3 days (when reaching 60% confluence)with ESGROComplete Accutase
(EMD Millipore) at about 1:6 ratio. C57BL/6 mESCs maintain a stable
tegrity of RNA samples was assessed by the Agilent 2100 Bioanalyzer
(Santa Clara, CA) with the RNA 6000 Nano Reagent Kit from the same
reader (Molecular Devices, Sunnyvale, CA). Each experiment was per-
formed with six replicates and repeated six times.
Thalidomide exposure and RNA isolation. ESC differentiation cultures
were exposed from the EB stage at day 3 onwards to 0.25 mM thalido-
mide or vehicle (0.25% DMSO) for 3 days. Preliminary results showed
that DMSO at 0.25% (v/v) had no significant effect on gene expression
during C57BL/6 ESC differentiation under the condition used in the
study (data not shown). Thalidomide-exposed cultures and vehicle con-
trols were collected at 24 h, 48 h, and 72 h (culture days 4, 5, and 6).
Three biological replicates were used for each condition. Treatment
with thalidomide did not affect EB sizes (data not shown). EBs were
lysed in RLT buffer (Qiagen; Valencia, CA) supplemented with β-
mercaptoethanol, homogenized by QIAshredder (Qiagen), and kept in
a −80 °C freezer until further processing. Total RNA was isolated on
the EZ1 Advanced XL (Qiagen) automated RNA purification instrument
using the EZ1 RNA Cell Mini Kit (Qiagen) following the manufacturer's
protocol, including an on-column DNase digestion. RNA concentration
scopic inspection of contracting cardiomyocytes in the EB outgrowths
on day 10 (Schulpen and Piersma, 2013).
Cytotoxicity assay. Cytotoxicity was measured by MTS assay using the
CellTiter 96 AQueous One Solution Cell Proliferation Assay kit from
Promega (Madison, WI) following instructions from the manufacturer.
Briefly, C57BL/6 mESC colonies cultured in ESGRO Complete Plus Clonal
Grade Medium were dissociated with ESGRO Complete Accutase and a
single-cell suspension at 1.0 × 105 cells/ml was prepared in ESGRO
Complete Basal Medium. The cells were seeded in 96-well cell culture
grade flat bottom plates (Nunc) coated with 0.1% gelatin (EMD
Millipore) at 100 μl/well (1.0 × 104 cells/well) and allowed to adhere
overnight at 37 °C with 5% CO2. After 24 h, 100 μl medium containing
2× final concentrations of thalidomide (1 μM to 1 mM) was added to
the test wells. In control wells, medium containing 0.25% dimethyl sulf-
oxide (DMSO)was added as a vehicle control. The treatment wasmain-
tained for 24 h. At the end of the exposure, 20 μl of CellTiter 96 AQueous
One Solution Cell Proliferation Assay reagent was added to each well
karyotype under the above passaging condition. The cells used in the
current study were at passage 18.
Cell differentiation through embryoid body formation. Induction of differ-
entiationwas achieved through embryoid body (EB) formation viahang-
ing drop culture following a procedure adapted from De Smedt et al.
(2008). In brief, stem cells were thawed and a suspension was prepared
at a concentration of 3.75 × 104 cells/ml in ESGRO Complete Basal Medi-
um (EMD Millipore), which does not contain LIF, BMP-4, or GSK3b-I.
About 50 drops (each of 20 μl) of the cell suspension were placed onto
the inner side of the lid of a 10-cm Petri dish filled with 5 ml phosphate
buffered saline (PBS; EMDMillipore) and incubated at 37 °C and 5% CO2
in a humidified atmosphere. After 3 days, EBs formed in the hanging
drops (Ø330–350 μm)were subsequently transferred into 6-cm bacteri-
ological Petri dishes (Becton Dickinson Labware, Franklin Lakes, NJ) and
were further cultivated for 2 days. On day 5, EBs were plated one per
well into 24-well tissue culture plates (Thermo Scientific Nunc,
Roskilde, Denmark). During further development of the attached EBs,
manufacturer.
dUTP residues and labeled by terminal deoxynucleotidyl transferase
(TdT) using the Affymetrix proprietary DNA Labeling Reagent that is co-
data analysis, all arrays referred to in this study were assessed for data
quality using the Affymetrix Expression Console software v.1.3 and all
Quantitative real-time PCR. Total RNA was isolated as mentioned previ-
ously from samples of an independent experiment. Reverse transcrip-
tion of mRNA was carried out using the High Capacity cDNA Reverse
Transcription Kit from Applied Biosystems (Foster City, CA), using
0.2 μg of total RNA as starting material. Real-time PCR was carried out
on a 7500 Real-Time PCR system of Applied Biosystems using TaqMan
Gene Expression Master Mix, TaqMan Gene Expression Assay primer/
probe sets and the standard thermal cycling conditions for relative
quantification designed by the manufacturer. Results were analyzed
with the 7500 Software v.2.3 on the system using the ΔΔCT method.
Multiple endogenous controls consisting of 18S rRNA, β-actin, and
GAPDH were used simultaneously to correct for variations in input
RNA amount and cDNA amplification of different samples.
Results
Thalidomide cytotoxicity to the differentiating mESCs
Adherent mESCs cultured in differentiation medium were treated
with varying concentrations (1 μM to 1 mM) of thalidomide for 24 h
and cytotoxicity wasmeasured byMTS assay. As shown in Fig. 1, thalid-
45X. Gao et al. / Toxicology and Applied Pharmacology 287 (2015) 43–51
quality assessment metrics (including spike-in controls during target
preparation and hybridization) were found within boundaries.The
data set has been deposited in Gene Expression Omnibus (GEO;
http://www.ncbi.nlm.nih.gov/geo/) of the National Center for Biotech-
nology Information with accession number GSE61306.
Data processing and statistical analysis. The values of individual probes
belonging to one probe set in .CEL files were summarized using the ro-
bustmulti-array average (RMA) algorithm (Irizarry et al., 2003) embed-
ded in the Expression Console software v.1.3 (Affymetrix), which
comprises of convolution background correction, quantile normaliza-
tion, and median polish summarization. Subsequently, differentially
expressed genes (DEGs) were selected by one-way analysis of variance
(ANOVA) using the Affymetrix Transcriptome Analysis Console (TAC)
software v.1.0. The fold change (FC) of every gene, together with their
corresponding p-value, was used for selection of DEGs with cutoff
values indicated in the text.
Gene ontology and pathway analysis. The significantly regulated genes
were subjected to gene ontology (GO) and pathway analysis using the
Database for Annotation, Visualization, and Integrated Discovery
(DAVID) (Dennis et al., 2003; Huang da et al., 2009) to find overrepre-
sentations of GO terms in the biological process (BP) category at all
levels (GOTERM_BP_ALL) and KEGG (Kyoto Encyclopedia of Genes
and Genomes) pathways. As background, the Mus musculus (mouse)
whole genome was used. Statistical enrichment was determined using
default settings in DAVID. The statistically enriched GO terms were
grouped and counted after classification according to GO Slim using
the freely available web tool CateGOrizer (Hu et al., 2008). Functional
and pathway analysis were also conducted with the online Ingenuity
Pathway Analysis (IPA) software (http://www.ingenuity.com/
products/ipa) using default settings to identify biological functions, ca-
nonical pathways, and networks associated with the significantly regu-
valently linked to biotin. Subsequent hybridization, wash, and staining
were carried out using the Affymetrix GeneChip Hybridization, Wash,
and Stain Kit and the manufacturer's protocols were followed. Briefly,
each fragmented and labeled sense-strand cDNA target sample (ap-
proximately 3.5 μg) was individually hybridized to a GeneChip Mouse
Gene2.0 STArray at 45 °C for 16h inAffymetrix GeneChipHybridization
Oven 645. After hybridization, the array chips were stained andwashed
using an Affymetrix Fluidics Station 450. The chips were then scanned
on Affymetrix GeneChip Scanner 3000 7G and the image (.DAT) files
were preprocessed using the Affymetrix GeneChip Command Console
(AGCC) software v.4.0 to generate cell intensity (.CEL) files. Prior to
RNA processing and microarray experiment. The total RNA samples
were preprocessed for hybridization to Mouse Gene 2.0 ST Array
(Affymetrix, Santa Clara, CA) using the GeneChip WT PLUS Reagent Kit
(Affymetrix) following the manufacturer's protocol. In brief, 50 ng of
total RNA was used to generate first strand cDNA using reverse tran-
scriptase and primers containing a T7 promoter sequence. The single-
stranded cDNA was then converted to double-stranded cDNA by using
DNA polymerase and RNase H to simultaneously degrade the RNA and
synthesize second-strand cDNA. Complimentary RNA (cRNA) was syn-
thesized and amplified by in vitro transcription (IVT) of the second-
stranded cDNA template using T7 RNA polymerase. Subsequently,
sense-strand cDNA was synthesized by the reverse transcription of
cRNA with incorporated deoxyuridine triphosphate (dUTP). Purified,
sense-strand cDNA was fragmented by uracil-DNA glycosylase (UDG)
and apurinic/apyrimidinic endonuclease 1 (APE 1) at the unnatural
lated genes.
omide did not cause cell death at concentrations up to 0.1 mM. At
0.5mMand above, significant cell viability losswas observed. Therefore,
in the following transcriptomic study, we chose a concentration be-
tween 0.1 and 0.5 mM (0.25 mM), which approximates the highest
noncytotoxic concentration of thalidomide.
Time-course transcriptome profiling on mESC differentiation after
exposure to thalidomide
Differentiating EBs were treated with 0.25 mM thalidomide for
3 days, and global gene expression analysis was conducted at 24, 48,
and 72 h (Fig. 2A) in thalidomide-exposed samples relative to their
time-matched controls. Using cut-off criteria of p b 0.05 and |FC| N 1.5,
a total of 214, 230, and 56 DEGswere identified at the three time points
respectively (Fig. 2B; Supplementary Table 1 in Gao et al. (2015)). The
total number of DEGs dropped dramatically at 72 h compared with
those at 24 h and 48h. At each timepoint, thenumber of downregulated
genes outweighs that of the upregulated genes, suggesting thalidomide
had an overall suppressing effect on gene expression during early mESC
differentiation. For both the upregulated and downregulated genes, a
small portion of overlap was observed between 24 h and 48 h, but al-
most no overlap was found either between 24 h and 72 h or between
48 h and 72 h (Fig. 2C), indicating the dynamic nature of gene
70
75
80
85
90
95
100
Ce
ll v
ia
bi
lity
 (%
)
0
Thalidomide concentration (µM)
*
***
1 5 10 50 100 500 1000
Fig. 1.Dose response of thalidomide exposure. DifferentiatingmESCswere exposed to dif-
ferent concentrations of thalidomide for 24 h. Cell viability was measured by the MTS
assay. The data are expressed asmean± SD of six repeated experiments (eachwith 6 rep-
licates) in percentages relative to the solvent control (concentration “0”). Significancewas
determined by Student's t-test. *p b 0.05; ***p b 0.001.
46 X. Gao et al. / Toxicology and Applied Pharmacology 287 (2015) 43–51
Day 1 20
Hour
A
B
expression changes induced by thalidomide exposure during ESC
differentiation.
Functional annotation of thalidomide downregulated and upregulated
genes
To unravel the cellular functions and pathways represented in the
DEGs induced by thalidomide exposure, the downregulated and the up-
regulated genes were subjected separately to functional annotation
36
178
0
50
100
150
200
250
24
T
N
um
be
r o
f d
iff
er
en
tia
lly
 e
xp
re
ss
ed
 g
en
es
C
Upregulated
Fig. 2. Time-course of transcriptome profiling on mESC differentiation after exposure to thalido
the embryoid body (EB) formation stage. Hanging drops were set up on day 0 and EBs formed o
the orange arrow, which lasted for 72 h (from day 3 to day 6). The numbers on the top are days
during ESC differentiation (after EB formation) covering thalidomide exposure. B, Histogram of
omide-exposed culture (vs. control) at each timepoint (p b 0.05, |FC|N 1.5).C, Venndiagrams sho
upregulated genes and right panel for the downregulated genes.
63 4 5
0 24 48 72
using DAVID to find overrepresentations of gene ontology (GO) terms
in the biological process (BP) category and KEGG pathways. The down-
regulated genes at 24, 48, and 72 h resulted in 41, 48, and 60 GO terms
respectively in the BP category at all levels (Supplementary Table 2 in
Gao et al. (2015)). Using the CateGOrizer tool, these GO terms were
grouped into 18, 20 and 13 classes, respectively, within the pre-
defined set of parent/ancestor GO terms (Fig. 3). The majority of
the classes fell into the categories of metabolism, biogenesis, biosyn-
thesis, and transport. Two classes of GO terms directly related to ESC
54
17
176
39
48 72
Downregulated
Upregulated
ime (h)
Downregulated
mide. A, Schematic representation of the experimental procedure. The green arrow covers
n day 3. ESC differentiation started from day 3 onwards. Compound exposure is shown by
covering thewhole process, while the numbers at the bottom are the time points in hours
the total number of upregulated genes (red) and downregulated genes (green) in thalid-
wing overlaps ofDEGsbetweendifferent timepoints. The leftpanel shows overlaps for the
.0%
47X. Gao et al. / Toxicology and Applied Pharmacology 287 (2015) 43–51
A
40.0%
25
7.5%
7.5%
7.5%
7.5%
5.0%
5.0%5.0%2.5% 2.5%
differentiation, namely development and cell differentiation, were
enriched at all three time points. Another two directly related GO
terms, morphogenesis and embryonic development, were also enriched
at 24 h and 72 h. Also, cytoskeleton organization and biogenesis was
enriched at 48 h and 72 h. It is interesting to note that two classes relat-
ed to immune responses, response to biotic stimulus and response to
external stimulus, were also enriched by the downregulated DEGs both
at 24 h and 48 h.
Similarly, functional annotation clustering by DAVID also revealed
several clusters (groups of annotations with similar gene members) of
B
C
15.0%
15.0%
10.0%
7.5%
7.5%
7.5%
36.2%
19.1%
17.0%
6.4%
6.4%
6.4%
6.4%
6.4%
6.4%
6.4%
4.3%
4.3% 4.3%
4.3%
4.3%
2.1%
2.1%2.1%2.1%
2.1%
19.3%
15.8%
15.8%
14.0%
12.3%
8.8%
8.8%
8.8%
5.3%
1.8% 1.8%1.8%
1.8%
Fig. 3.Distribution of enriched GO terms according to GO slim for the downregulated DEGs iden
centage indicates the number of GO terms in each class as a percentage of the total number of
metabolism
cell organization and biogenesis
biosynthesis
nucleobase, nucleoside, nucleotide and nucleic acid metabolism
organelle organization and biogenesis
DNA metabolism
cell proliferation
response to biotic stimulus
response to external stimulus
signal transduction
GO terms directly related to embryonic differentiation and immune re-
sponses (Supplementary Table 3 in Gao et al. (2015)). Five clusterswere
identified at 24 hwith thefirst four related to differentiation and the last
one to immune responses. Cluster 1 was closely related to embryonic
differentiation and included such important GO terms as gastrulation,
embryonic morphogenesis, anterior/posterior pattern formation, and
heart development. Cluster 2 included several terms focusing on muscle
tissue development. Multiple terms regulating neural system develop-
mentwere found in cluster 3, such as regulation of neurogenesis and reg-
ulation of nervous system development. Cluster 4 had three GO terms on
cell communication
development
morphogenesis
protein metabolism
embryonic development
cell differentiation
protein biosynthesis
cell organization and biogenesis
organelle organization and biogenesis
metabolism
DNA metabolism
response to biotic stimulus
transport
response to external stimulus
signal transduction
cell communication
nucleobase, nucleoside, nucleotide and nucleic acid metabolism
cell proliferation
carbohydrate metabolism
protein transport
development
cytoskeleton organization and biogenesis
response to stress
biosynthesis
cell differentiation
protein metabolism
protein biosynthesis
transport
cell differentiation
ion transport
development
cell organization and biogenesis
organelle organization and biogenesis
cell homeostasis
cytoskeleton organization and biogenesis
morphogenesis
metabolism
nucleobase, nucleoside, nucleotide and nucleic acid metabolism
embryonic development
catabolism
tified at each time point during ESC differentiation after exposure to thalidomide. The per-
unique GO terms enriched by the DEGs. A, 24 h; B, 48 h; C, 72 h.
small GTPase (Ros and Ras families) mediated signal transduction,
which plays important roles in cell differentiation. More than twenty
GO terms were found in cluster 5, all closely related to immune re-
sponses and included termson inflammatory response, regulation of cy-
tokine production, and regulation of proliferation and/or activation of
immune cells (lymphocytes and leukocytes). Very similar clusters
were identified at 48 h as compared to those at 24 h, except that one
cluster (cluster 4) had several terms on immune system development
(leukocyte differentiation, hemopoietic or lymphoid organ development,
and immune system development). In addition, at 48 h a cluster formed
on regulation of cytokine production (cluster 6) separated from the
cluster on immune responses (cluster 5). Only two clusters were iden-
tified for 72 h, both themed onmuscle development and heart develop-
ment. Cluster 1 also had two terms on blood circulation—circulatory
directly related to ESC differentiation, cell differentiation and develop-
ment, were enriched at 72 h. Still two other classes related to immune
responses, response to biotic stimulus and response to external stimulus,
were enriched at 48 h. Gene functional annotation clustering on upreg-
ulated genes revealed a cluster of GOBP termsboth at 24 h and 48h that
is related to sensory perception of smell in the neurological system pro-
cess (Supplementary Table 5 in Gao et al. (2015)).
KEGG pathways affected by the thalidomide regulated genes are
listed in Table 2. For the downregulated DEGs, two pathways, systemic
lupus erythematosus and regulation of actin cytoskeleton, both appeared
at 24 h and 48 h. The pathways at 72 h were majorly related to cardio-
myopathy. For the upregulated genes, olfactory transductionwas affect-
ed both at 24 h and 48 h. No pathways were identified at 72 h.
Functional and pathway analysis with IPA showed very similar re-
sults (data not shown) with those of DAVID analysis. However, it is
worth noting that several canonical pathways associated with actin ap-
peared on top of the lists at all the three timepoints, which include Actin
cytoskeleton signaling, Regulation of actin-basedmotility by Rho, RhoA sig-
naling, Cdc42 signaling, and Epithelial adherens junction signaling. The
DEGs involved in these canonical pathways are shown in Table 3.
Thalidomide-induced downregulation of small GTPases, dysregulation of
heart development, and perturbation of inflammatory responses
The dysregulation of genes associated with small GTPases-mediated
signal transduction, heart development and inflammatory responses
are listed in Table 4. Most of the genes for small GTPases-mediated sig-
nal transduction and inflammatory responses were downregulated at
both 24 h and 48 h (with a few only at one of the time points) but not
Table 1
GO classes enriched by the thalidomide upregulated DEGs at each time point.
GO class 24 h (8) 48 h (12) 72 h (8)
Cell differentiation __ __ 5
Development __ __ 4
Signal transduction 2 3 1
Cell communication 2 3 1
Response to biotic stimulus __ 2 __
Response to external stimulus __ 1 __
GO terms enriched by the thalidomide upregulatedDEGs at each time point were grouped
into pre-defined set of parent/ancestor GO terms (or classes) using the CateGOrizer tool.
The numbers in the table indicate the numbers of GO terms grouped into each specified
class at the indicated time point. A “__” sign means no GO terms were grouped into that
class. The total number of GO terms by the thalidomide upregulated DEGs at each time
point is included in the parentheses following the time.
48 X. Gao et al. / Toxicology and Applied Pharmacology 287 (2015) 43–51
system process and blood circulation.
The upregulated genes at 24, 48, and 72 h resulted in 8, 12, and 8 GO
terms respectively in the BP category at all levels (Supplementary
Table 4 in Gao et al. (2015)). These GO terms were grouped into 2, 4
and 4 parent/ancestor GO classes, respectively (Table 1). Two
classes of GO terms, signal transduction and cell communication, were
enriched at all three time points. Another two classes of GO terms
Table 2
List of KEGG pathways enriched by thalidomide regulated DEGs at various time points.
Term Pathway
Downregulated, 24 h
mmu05322 Systemic lupus erythematosus
mmu04010 MAPK signaling pathway
mmu00330 Arginine and proline metabolism
mmu04670 Leukocyte transendothelial migration
mmu04510 Focal adhesion
mmu00190 Oxidative phosphorylation
mmu04530 Tight junction
mmu04810 Regulationof actin cytoskeleton
Downregulated, 48 h
mmu05322 Systemic lupus erythematosus
mmu04810 Regulation of actin cytoskeleton
Downregulated, 7 h
mmu05410 Hypertrophic cardiomyopathy (HCM)
mmu05414 Dilated cardiomyopathy
mmu04260 Cardiac muscle contraction
mmu05412 Arrhythmogenic right ventricular cardiomyopathy (ARVC)
mmu04020 Calcium signaling pathway
Upregulated, 24 h
mmu04740 Olfactory transduction
Upregulated, 48 h
mmu04740 Olfactory transduction
DAVID was used for the analysis using the thalidomide regulated DEGs identified during ESC d
a Number of DEGs involved in the pathway.
b Number of DEGs involved in the pathway as a percentage of the total number of regulated g
c Fold enrichment, which is the number of genes (in the list) involved in a particular pathway
particular pathway as a percentage of the whole genome (population background) (Huang da
at 72 h, whereas the genes for heart, muscle, and blood circulation
were downregulated only at 72 h.
Validation of microarray data by real-time PCR
To verify the reliability of the DEGs identified by microarray, six
genes selected from each of the three groups in Table 4 (2 genes/
Counta %b p-Value FEc
5 3.5 0.0090 5.9
6 4.3 0.0593 2.8
3 2.1 0.0669 6.9
4 2.8 0.0694 4.1
5 3.5 0.0729 3.1
4 2.8 0.0853 3.8
4 2.8 0.0931 3.6
5 3.5 0.0946 2.8
10 6.9 4.6E−08 12.7
6 4.2 0.0221 3.6
8 25.0 8.3E−11 42.0
8 25.0 1.6E−10 38.4
6 18.8 3.0E−07 34.0
4 12.5 0.0004 23.5
4 12.5 0.0064 9.2
6 25.0 0.0047 4.0
7 19.4 0.0047 3.5
ifferentiation (see text for details).
enes (up or down) at each time point.
as a percentage of total number genes of the list as compared to all genes involved in that
et al., 2009).
group) were further examined by real-time PCR. The results are sum-
marized in Table 5. As shown, the gene expression changes obtained
by real-time PCRwere in accordance with those of the microarray anal-
ysis, indicating a good reliability and reproducibility of the microarray
data in the present study.
Discussion
An important consideration for toxicogenomics studies is the selec-
tion of the right concentration (dose) of the testing compounds in
order to yield useful information for predicting the hazard of the
chemicals in question. This is especially true for developmental toxicity
testing using ESCs. A recent study by Waldmann et al. (2014) suggests
the use of the highest noncytotoxic concentration for gene array
toxicogenomics studies, as higher concentrations possibly yield wrong
information on themode of action,whereas lower concentrations result
in decreased gene expression changes and thus a reduced power of the
study. The dose–response experiment (Fig. 1) indicates that a concen-
tration of 0.25 mM is close to the highest noncytotoxic concentration
for thalidomide in the differentiating mESCs.
The results from the present study show that at 0.25 mM thalido-
mide significantly altered global gene expression profiles in differentiat-
ing mESCs within 24 h to 72 h of exposure after EB formation. In a
previous study, due to the high |FC| cutoff value (|FC| N 2.0) used, only
59 genes were found differentially expressed after 24 h exposure to
the same concentration of thalidomide (Gao et al., 2014). In the current
study, we used a less stringent cutoff value, i.e. |FC| N 1.5, in selecting the
DEGs. The relaxed criterion yielded adequate numbers of DEGs for
downstream function and pathway analysis (Fig. 2B). Functional analy-
sis of the thalidomide regulated genes revealed a multitude of function
classes associated with these genes, from more general terms such as
metabolism, biogenesis, biosynthesis, and transport, to more specific
terms such as embryonic development, cell differentiation, cytoskeleton
organization and biogenesis, heart development, neurogenesis, and im-
mune responses. This implies that thalidomide induced a broad range of
reactions in the differentiating EBs, which collectively affected a multi-
tude of pathways in the development process and ultimately resulted
in embryopathy characteristic of thalidomide toxicity.
Given the limitations of using mESCs as a developmental toxicity
model due to interspecies differences (Adler et al., 2008), the current
study is among the first to provide transcriptomic information of ESC
Table 3
List of canonical pathways associated with actin identified by IPA and DEGs involved in the pathways at various time points.
Pathway 24 h 48 h 72 h
Actin cytoskeleton signaling Fgd3, Myl7, Myl12b, Ppp1cb, Rhoa, Rras Arpc4, Msn, Myh6, Myl7, Ppp1cb, Rhoa, Rras Actn2, Myh6, Myl3, Mylk, Ttn
Regulation of actin-based motility by Rho Arhgdia, Myl7, Myl12b, Ppp1cb, Rhoa Arpc4, Myl7, Ppp1cb, Rhoa Myl3, Mylk
RhoA signaling Cdc42ep5, Myl7, Myl12b, Ppp1cb, Rhoa Arpc4, Cdc42ep5, Msn, Myl7, Ppp1cb, Rhoa Myl3, Mylk, Ttn
Cdc42 signaling Cdc42ep5, Fgd3, Myl7, Myl12b, Ppp1cb Arpc4, Cdc42ep5, H2-Q1, Myl7, Ppp1cb Myl3, Mylk
Epithelial adherens junction signaling Myl7, Rap1b, Rhoa, Rras, Tuba1b Arpc4, Myh6, Myl7, Rhoa, Rras, Tuba1b Actn2, Myh6, Myl3
Genes with name underlined were upregulated upon thalidomide exposure, otherwise the genes were downregulated.
Table 4
Dysregulation of genes associated with small GTPases-mediated signal transduction, heart development and inflammatory responses.
020
16
AS o
AS o
a on
otein
fam
icted
n 1b
AS o
en 9
atio
17240606 445686 NM_009846 Cd24a CD24a antigen
1
th
47
otei
ype
49X. Gao et al. / Toxicology and Applied Pharmacology 287 (2015) 43–51
17312485 457581 NM_008296 Hsf1 Heat shock factor
17475342 461229 ENSMUST00000002678 Tgfb1 Transforming grow
17491890 463829 NM_024439 H47 Histocompatibility
Heart, muscle, and blood circulation
17373189 447441 NM_008653 Mybpc3 Myosin binding pr
17386778 452967 NM_011652 Ttn Titin
17306532 441641 NM_001164171 Myh6 Myosin, heavy pol
Transcript
cluster ID
DAVID
ID
Gene accession Gene
symbol
Gene description
Small GTPases
17276926 465223 NM_025292 Synj2bp RIKEN cDNA 1810
predicted gene 41
17306274 431165 ENSMUST00000022765 Rab2b RAB2B, member R
17318932 438252 NM_025931 Ift27 RAB, member of R
17477670 456781 NM_009101 Rras Harvey rat sarcom
17485706 426192 ENSMUST00000076831 Cdc42ep5 CDC42 effector pr
17531181 432091 NM_016802 Rhoa Ras homolog gene
homolog A2; pred
17547525 469726 NM_024457 Rap1b RAS related protei
17548683 475068 NM_023126 Rab8a RAB8A, member R
Immune response
17211286 447949 NM_016923 Ly96 Lymphocyte antig
17532137 474639 NM_010851 Myd88 Myeloid differenti
17531685 433070 NM_011562 Tdgf1 Predicted gene 6148;
similar to cripto
17467589 466373 NM_001160127 Smyd1 SET and MYND domai
17347619 436440 NM_011406 Slc8a1 Solute carrier family 8
17470235 421510 NM_009781 Cacna1c Calcium channel, volta
17290457 425477 NM_023868 Ryr2 Ryanodine receptor 2,
17290603 428778 NM_033268 Actn2 Actinin alpha 2
17339395 423375 NM_010867 Myom1 Myomesin 1
17298379 435600 NM_009393 Tnnc1 Troponin C, cardiac/slo
17287827 432610 NM_009369 Tgfbi Transforming growth
17522439 455941 NM_010859 Myl3 Myosin, light polypept
*FC—fold change; the negative signs indicate downregulation.
FC (24 h) FC (48 h) FC (72 h)
G14 gene; synaptojanin 2 binding protein; −1.60 −1.77
ncogene family −1.57
ncogene family-like 4 −1.66 −1.64
cogene, subgroup R −1.50 −1.68
(Rho GTPase binding) 5 −1.53 −1.61
ily, member A; similar to aplysia ras-related
gene 12844
−2.16 −1.67
; similar to GTP-binding protein (smg p21B) −1.61
ncogene family −1.64 −1.67
6 −1.65 −1.55
n primary response gene 88 −1.60 −1.70
−1.53 −1.51
−1.50 −1.51
factor, beta 1 −1.52 −1.61
−1.53
n C, cardiac −1.52
−1.94
ptide 6, cardiac muscle, alpha −2.31
teratocarcinoma-derived growth factor 1; −1.61
n containing 1 −1.50
(sodium/calcium exchanger), member 1 −1.91
ge-dependent, L type, alpha 1C subunit −1.56
cardiac −1.73
−1.60
−1.87
w skeletal −1.62
factor, betainduced −1.58
ide 3 −1.62
50 X. Gao et al. / Toxicology and Applied Pharmacology 287 (2015) 43–51
differentiation following exposure to thalidomide. In a recent study,
Meganathan et al. (2012) used human ESCs to study transcriptomic
(and proteomic) changes during differentiation following thalidomide
exposure. In this in vitromodel the transcriptomic pattern demonstrat-
ed differential expression of transcription factors andbiological process-
es related to limb, heart and embryonic development. In addition, this
study uncovered some novel possible mechanisms of thalidomide
embryopathy, such as the inhibition of nucleocytoplasmic trafficking
and inhibition of glutathione transferases. Nevertheless, the long period
of hESC differentiation (14 days) required by the model to demonstrate
embryonic development renders it less practical as a high throughput
method for developmental toxicity screening. Many of the biological
processes identified in hESCs following thalidomide exposure
(Meganathan et al., 2012) were also discovered in the current study
using mESCs. For example, several GO classes (Fig. 3), annotation clus-
ters (Supplementary Table 3 in Gao et al. (2015)), and KEGG pathways
(Table 2) related to embryonic development and heart development
were identified in our model. In comparison, these were detected in as
early as 24–72 h (1–3 days).
The precisemechanisms underlying the teratogenic effects of thalid-
omide are still unclear, but one possibility is its antiangiogenic activity
(Kim and Scialli, 2011). It has been suggested that limb defects caused
by thalidomide were secondary to inhibition of blood vessel growth in
the developing limb bud (D'Amato et al., 1994; Therapontos et al.,
2009), and correct limb bud formation requires a complex interaction
of both vasculogenesis and angiogenesis during development (Patan,
2004). Although specific GO terms on limb development or vasculature
development, as were found in the hESC study (Meganathan et al.,
2012), were not identified in the current study, several genes and relat-
ed GO terms, annotation clusters and pathways were indirectly linked
to, or are suggestive of, these effects. Several small GTPases in the Rho
Table 5
Expression changes of some representative genes determined by real-time PCR compared
to those detected by microarray.
Rras Rhoa Cd24a Tgfb1 Myh6 Myl3
24 h Microarraya −1.50 −2.16 −1.53 −1.52 −1.10 1.16
Real-time PCRb 0.63 0.48 0.68 0.73 n.a.c 1.25
48 h Microarraya −1.68 −1.67 −1.51 −1.61 1.51 −1.09
Real-time PCRb 0.55 0.62 0.43 0.53 n.a.c 0.85
72 h Microarraya −1.17 1.12 1.15 −1.18 −2.31 −1.62
Real-time PCRb 0.96 1.23 1.26 0.78 0.44 0.53
a The microarray data are expressed as fold changes with positive numbers indicating
upregulation and negative numbers downregulation.
b The real-time PCR data are ratios of expression relative to the control; therefore
numbers N1 mean upregulation whereas numbers b1 mean downregulation. The
values are means of three technical replicates.
c At 24 h and 48 h, the expression level of Myh6 was below the detection limit of real-
time PCR under normal amplification conditions.
subfamily (Synj2bp, Cdc42ep5, Rhoa) were downregulated at 24 h and
48 h (Table 4). Rho proteins play important roles in organelle develop-
ment, cytoskeletal dynamics, cell movement, and other common cellu-
lar functions (Boureux et al., 2007). Similarly, several genes in KEGG
pathway of regulation of actin cytoskeleton were also downregulated at
24 h (Myl7, Myl12a, Rras, Ppp1cb, Rhoa) and 48 h (Arpc4, Myl7, Rras,
Msn, Ppp1cb, Rhoa). In addition, several genes in the focal adhesion
pathway (Myl7, Myl12a, Rap1b, Ppp1cb, Rhoa) and tight junction
pathway (Myl7, Myl12a, Rras, Rhoa) were also downregulated at 24 h.
These genes were also involved in the canonical pathways related to
actin as identified by IPA (Table 3). All these processes may affect
morphogenesis that controls the organized spatial distribution of cells
during embryonic development, potentially leading to deformations in
limb development.
Thalidomide was reintroduced into the market in recent years as an
immunomodulator for the treatment of erythema nodosum leprosum
and several other immune diseases such as Crohn's disease, sarcoidosis,
graft-versus-host disease, rheumatoid arthritis and a number of skin
conditions (Okafor, 2003; Teo et al., 2005). The immunomodulating ef-
fects of thalidomide were clearly demonstrated in this study by the
downregulation of several immune response genes including Ly96,
Myd88, Cd24a, Hsf1, Tgfb1, andH47. These genes are involved in diverse
pathways of immune responses and may thus impact the overall
immune system. Downregulation of transforming growth factor genes
(including Tgfb1) was observed in human embryonic stem cell differ-
entiation upon exposure to thalidomide (Mayshar et al., 2011), and
downregulation of inflammatory pathways were also observed in
thalidomide-exposed monkey embryos (Ema et al., 2010).
Cardiovascular malformations were also reported in some of
the children affected by maternally ingested thalidomide (Jackson,
1968). The nature of the cardiovascular lesions described in these cases
was varying, included aortic hypoplasia and coarctation, patent ductus
arteriosus, pulmonary stenosis, transposition of the great vessels and
anomalous pulmonary venous drainage (Jackson, 1968). About a dozen
genes related to heart, muscle and blood circulationwere downregulated
at 72 h after thalidomide exposure (Table 4). These genes were enriched
in a multitude of GO terms related to heart muscle development, muscle
contraction, and blood circulation (Supplementary Table 2 in Gao et al.
(2015)); and the enriched GO terms formed two functional clusters
both themed on heart development (Supplementary Table 3 in Gao et
al. (2015)). Three KEGGpathways on cardiomyopathy and one on cardiac
muscle contraction were also affected by these genes (Table 2). Together
these results indicate that thalidomide embryopathy on cardiovascular
development was also detected by the current model system using
mouse ESCs.
Thalidomide elicited distinct developmental adverse effects in dif-
ferent species studied. In primates and a few strains of rabbits severe
congenital malformations similar to humans were reported, but only
moderate effects were found in rats and no significant changes were
observed in mice (Fratta et al., 1965; Schumacher et al., 1968; Teo
et al., 2001, 2004). The reason for mouse insensitivity to thalidomide
is not well understood. One possibility is that thalidomide does not
pass through the mouse placenta (Therapontos et al., 2009). Placental
transfer of chemicals depends on a number of factors. Chemicals can
cross the placental barrier via simple diffusion, pumps, plasma
membrane carriers, and biotransforming enzymes (Marin et al., 2004).
Differences in surface receptors or placental structure between different
species may play a role in species-specific placental transfer of thalido-
mide (Vargesson, 2009). Alternatively, the antiangiogenic metabolic
products of thalidomide are not generated in mouse (Therapontos
et al., 2009). The DEGs detected in the current study indicative and/or
suggestive of thalidomide embryopathy (heart and limb development)
support the former supposition, as thalidomide in direct contact with
mESCs in culture perturbed the differentiation process, causing signifi-
cant divergence from the normal developmental track of the EBs. On
the other hand, these results support the notion that in vitro mouse
ESC differentiation in combination with transcriptomic profiling is a
suitable model for identifying differentiation-modulating effects of
developmental toxicants, and highlight the accuracy of global gene
expression analysis in identifying teratogenic potential and assessing
the effect of developmental toxicants on biologicalprocesses in the
course of embryonic differentiation.
Differentiation of ESCs is a highly dynamic process, with thousands
of genes changing expression over time (Gao et al., 2014). Perturbation
of gene expression by thalidomide or other toxicants, and the associated
tissue development, would thus be time-dependent. This is well illus-
trated by the current study where perturbation of heart development
was evident at 72 h but not the earlier time points. These results
also suggest that examining the EBs after a longer period of time may
reveal other effects such as perturbation of limb development, a hall-
mark of thalidomide embryopathy. In the study of Meganathan et al.
(2012), approximately 14 days were necessary to reveal gene expres-
sion changes associated with perturbation of limb formation in hESCs.
Although this study demonstrated the hESCmodel can partially explain
the molecular mechanisms of a specific toxicant, the longer time of
exposure needed by the model renders it impractical as a method for
high-throughput screening.
In summary, this in vitro study demonstrated that the downregula-
tion of small GTPases-mediated signal transduction, dysregulation of
heart development, and perturbation of inflammatory responses may
represent critical embryotoxicities of thalidomide that coincide with
Genschow, E., Spielmann, H., Scholz, G., Pohl, I., Seiler, A., Clemann, N., Bremer, S., Becker,
K., 2004. Validation of the embryonic stem cell test in the international ECVAM vali-
dation study on three in vitro embryotoxicity tests. Altern. Lab. Anim. 32, 209–244.
Heuer, J., Bremer, S., Pohl, I., Spielmann, H., 1993. Development of an in vitro embryo-
toxicity test using murine embryonic stem cell cultures. Toxicol. In Vitro 7, 551–556.
Hu, Z.L., Bao, J., Reecy, J.M., 2008. CateGOrizer: a web-based program to batch analyze
gene ontology classification categories. Online J. Bioinformatics 9, 108–112.
Huang da, W., Sherman, B.T., Lempicki, R.A., 2009. Systematic and integrative analysis of
large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57.
Irizarry, R.A., Hobbs, B., Collin, F., Beazer-Barclay, Y.D., Antonellis, K.J., Scherf, U., Speed,
T.P., 2003. Exploration, normalization, and summaries of high density oligonucleotide
51X. Gao et al. / Toxicology and Applied Pharmacology 287 (2015) 43–51
such as cell biology studies characterizing protein expression of these
genes, are needed in order to confirm these changes on the cellular
level. The application ofmESCs offers the advantage of shorter exposure
time potentially allowing for high-throughput screening of large num-
bers of compounds. Nevertheless, further studies are needed in order
to establish this technique on a wider range of potential developmental
compounds. The results obtained from additional compounds will en-
able further refinement of the assay, possibly allowing the use of fo-
cused microarray or real-time PCR platforms. The findings presented
here will broaden the application of mouse embryonic stem cells in de-
velopmental toxicity assessment, complementing existing in vivo and
in vitromodels.
Conflict of interest
The authors declare that there are no conflicts of interest.
Acknowledgments
This research is funded by internal funds of the U.S. Food and Drug
Administration. The findings and conclusions presented in this article
are those of the authors and do not necessarily represent views, opin-
ions, or policies of the U.S. Food and Drug Administration.
References
Adler, S., Pellizzer, C., Hareng, L., Hartung, T., Bremer, S., 2008. First steps in establishing a
developmental toxicity test method based on human embryonic stem cells. Toxicol.
In Vitro 22, 200–211.
Boureux, A., Vignal, E., Faure, S., Fort, P., 2007. Evolution of the Rho family of ras-like
GTPases in eukaryotes. Mol. Biol. Evol. 24, 203–216.
D'Amato, R.J., Loughnan, M.S., Flynn, E., Folkman, J., 1994. Thalidomide is an inhibitor of
angiogenesis. Proc. Natl. Acad. Sci. U. S. A. 91, 4082–4085.
De Smedt, A., Steemans, M., De Boeck, M., Peters, A.K., van der Leede, B.J., Van Goethem, F.,
Lampo, A., Vanparys, P., 2008. Optimisation of the cell cultivation methods in the em-
bryonic stem cell test results in an increased differentiation potential of the cells into
strong beating myocard cells. Toxicol. In Vitro 22, 1789–1796.
Dennis Jr., G., Sherman, B.T., Hosack, D.A., Yang, J., Gao,W., Lane, H.C., Lempicki, R.A., 2003.
DAVID: database for annotation, visualization, and integrated discovery. Genome
Biol. 4, P3.
Diggle, G.E., 2001. Thalidomide: 40 years on. Int. J. Clin. Pract. 55, 627–631.
Ema, M., Ise, R., Kato, H., Oneda, S., Hirose, A., Hirata-Koizumi, M., Singh, A.V., Knudsen,
T.B., Ihara, T., 2010. Fetal malformations and early embryonic gene expression re-
sponse in cynomolgus monkeys maternally exposed to thalidomide. Reprod. Toxicol.
29, 49–56.
Flint, O.P., Orton, T.C., 1984. An in vitro assay for teratogens with cultures of rat embryo
midbrain and limb bud cells. Toxicol. Appl. Pharmacol. 76, 383–395.
Fratta, I.D., Sigg, E.B., Maiorana, K., 1965. Teratogenic effects of thalidomide in rabbits, rats,
hamsters, and mice. Toxicol. Appl. Pharmacol. 7, 268–286.
Gao, X., Yourick, J.J., Sprando, R.L., 2014. Transcriptomic characterization of C57BL/6
mouse embryonic stem cell differentiation and its modulation by developmental tox-
icants. PLoS One 9, e108510.
Gao, X., Sprando, R.L., Yourick, J.J., 2015. Transcriptomic changes in mouse embryonic
stem cells exposed to thalidomide during spontaneous differentiation. Data in Brief
http://dx.doi.org/10.1016/j.dib.2015.05.014 (in press).
array probe level data. Biostatistics 4, 249–264.
Ito, T., Ando, H., Handa, H., 2011. Teratogenic effects of thalidomide: molecular mecha-
nisms. Cell. Mol. Life Sci. 68, 1569–1579.
Jackson, B.T., 1968. The pathogenesis of congenital cardiovascular anomalies. N. Engl.
J. Med. 279, 80–89.
Kim, J.H., Scialli, A.R., 2011. Thalidomide: the tragedy of birth defects and the effective
treatment of disease. Toxicol. Sci. 122, 1–6.
Lein, P., Silbergeld, E., Locke, P., Goldberg, A.M., 2005. In vitro and other alternative ap-
proaches to developmental neurotoxicity testing (DNT). Environ. Toxicol. Pharmacol.
19, 735–744.
Marin, J.J., Briz, O., Serrano, M.A., 2004. A review on the molecular mechanisms involved
in the placental barrier for drugs. Curr. Drug Deliv. 1, 275–289.
Mayshar, Y., Yanuka, O., Benvenisty, N., 2011. Teratogen screening using transcriptome
profiling of differentiating human embryonic stem cells. J. Cell. Mol. Med. 15,
1393–1401.
Meganathan, K., Jagtap, S., Wagh, V., Winkler, J., Gaspar, J.A., Hildebrand, D., Trusch, M.,
Lehmann, K., Hescheler, J., Schlüter, H., Sachinidis, A., 2012. Identification of
thalidomide-specific transcriptomics and proteomics signatures during differentia-
tion of human embryonic stem cells. PLoS One 7, e44228.
New, D.A., Coppola, P.T., Cockroft, D.L., 1976. Improved development of head-fold rat em-
bryos in culture resulting from low oxygen and modifications of the culture serum.
J. Reprod. Fertil. 48, 219–222.
Okafor, M.C., 2003. Thalidomide for erythema nodosum leprosum and other applications.
Pharmacotherapy 23, 481–493.
Patan, S., 2004. Vasculogenesis and angiogenesis. Cancer Treat. Res. 117, 3–32.
Schulpen, S.H., Piersma, A.H., 2013. The embryonic stem cell test. Methods Mol. Biol. 947,
375–382.
Schumacher, H., Blake, D.A., Gurian, J.M., Gillette, J.R., 1968. A comparison of the teratogen-
ic activity of thalidomide in rabbits and rats. J. Pharmacol. Exp. Ther. 160, 189–200.
Shukla, S.J., Huang, R., Austin, C.P., Xia, M., 2010. The future of toxicity testing: a focus on
in vitro methods using a quantitative high-throughput screening platform. Drug
Discov. Today 15, 997–1007.
Spielmann, H., 2009. The way forward in reproductive/developmental toxicitytesting.
Altern. Lab. Anim. 37, 641–656.
Spielmann, H., Pohl, I., Liebsch, M., Moldenhauer, F., 1997. The embryonic stem cell test,
an in vitro embryotoxicity test using two permanentmouse cell lines: 3 T3 fibroblasts
and embryonic stem cells. Toxicol. In Vitro 10, 119–127.
Tandon, S., Jyoti, S., 2012. Embryonic stem cells: an alternative approach to developmen-
tal toxicity testing. J. Pharm. Bioallied. Sci. 4, 96–100.
Teo, S.K., Evans, M.G., Brockman, M.J., Ehrhart, J., Morgan, J.M., et al., 2001. Safety profile of
thalidomide after 53 weeks of oral administration in beagle dogs. Toxicol. Sci. 59,
160–168.
Teo, S.K., Denny, K.H., Stirling, D.I., Thomas, S.D., Morseth, S., et al., 2004. Effects of thalid-
omide on developmental, peri- and postnatal function in female New Zealand white
rabbits and offspring. Toxicol. Sci. 81, 379–389.
Teo, S.K., Stirling, D.I., Zeldis, J.B., 2005. Thalidomide as a novel therapeutic agent: new
uses for an old product. Drug Discov. Today 10, 107–114.
Therapontos, C., Erskine, L., Gardner, E.R., Figg, W.D., Vargesson, N., 2009. Thalidomide in-
duces limb defects by preventing angiogenic outgrowth during early limb formation.
Proc. Natl. Acad. Sci. U. S. A. 106, 8573–8578.
van Dartel, D.A., Piersma, A.H., 2011. The embryonic stem cell test combined with
toxicogenomics as an alternative testing model for the assessment of developmental
toxicity. Reprod. Toxicol. 32, 235–244.
Vargesson, N., 2009. Thalidomide-induced limb defects: resolving a 50-year-old puzzle.
Bioessays 31, 1327–1336.
Waldmann, T., Rempel, E., Balmer, N.V., König, A., Kolde, R., Gaspar, J.A., Henry, M.,
Hescheler, J., Sachinidis, A., Rahnenführer, J., Hengstler, J.G., Leist, M., 2014. Design
principles of concentration-dependent transcriptome deviations in drug-exposed dif-
ferentiating stem cells. Chem. Res. Toxicol. 27, 408–420.
Winkler, J., Sotiriadou, I., Chen, S., Hescheler, J., Sachinidis, A., 2009. The potential of em-
bryonic stem cells combined with -omics technologies as model systems for toxicol-
ogy. Curr. Med. Chem. 16, 4814–4827.
known clinical evidences. However, it has to be noted thatmore studies,
	Thalidomide induced early gene expression perturbations indicative of human embryopathy in mouse embryonic stem cells
	Introduction
	Materials and methods
	Materials
	Pluripotent mouse embryonic stem cell culture
	Cell differentiation through embryoid body formation
	Cytotoxicity assay
	Thalidomide exposure and RNA isolation
	RNA processing and microarray experiment
	Data processing and statistical analysis
	Gene ontology and pathway analysis
	Quantitative real-time PCR
	Results
	Thalidomide cytotoxicity to the differentiating mESCs
	Time-course transcriptome profiling on mESC differentiation after exposure to thalidomide
	Functional annotation of thalidomide downregulated and upregulated genes
	Thalidomide-induced downregulation of small GTPases, dysregulation of heart development, and perturbation of inflammatory r...
	Validation of microarray data by real-time PCR
	Discussion
	Conflict of interest
	Acknowledgments
	References