Innate immunity in human embryonic stem cells: comparison with adult human endothelial cells.
ABSTRACT Treatment of human disease with human embryonic stem cell (hESC)-derived cells is now close to reality, but little is known of their responses to physiological and pathological insult. The ability of cells to respond via activation of Toll like receptors (TLR) is critical in innate immune sensing in most tissues, but also extends to more general danger sensing, e.g. of oxidative stress, in cardiomyocytes. We used biomarker release and gene-array analysis to compare responses in hESC before and after differentiation, and to those in primary human endothelial cells. The presence of cardiomyocytes and endothelial cells was confirmed in differentiated cultures by immunostaining, FACS-sorting and, for cardiomyocytes, beating activity. Undifferentiated hESC did not respond with CXCL8 release to Gram positive or Gram negative bacteria, or a range of PAMPs (pathogen associated molecular patterns) for TLRs 1-9 (apart from flagellin, an activator of TLR5). Surprisingly, lack of TLR-dependent responses was maintained over 4 months of differentiation of hESC, in cultures which included cardiomyocytes and endothelial cells. In contrast, primary cultures of human aortic endothelial cells (HAEC) demonstrated responses to a broad range of PAMPs. Expression of downstream TLR signalling pathways was demonstrated in hESC, and IL-1beta, TNFalpha and INFgamma, which bypass the TLRs, stimulated CXCL8 release. NFkappaB pathway expression was also present in hESC and NFkappaB was able to translocate to the nucleus. Low expression levels of TLRs were detected in hESC, especially TLRs 1 and 4, explaining the lack of response of hESC to the main TLR signals. TLR5 levels were similar between differentiated hESC and HAEC, and siRNA knockdown of TLR5 abolished the response to flagellin. These findings have potential implications for survival and function of grafted hESC-derived cells.
- SourceAvailable from: Judit Skopál[Show abstract] [Hide abstract]
ABSTRACT: Human embryonic stem cell-derived endothelial cells (hESC-EC), as well as other stem cell derived endothelial cells, have a range of applications in cardiovascular research and disease treatment. Endothelial cells sense Gram-negative bacteria via the pattern recognition receptors (PRR) Toll-like receptor (TLR)-4 and nucleotide-binding oligomerisation domain-containing protein (NOD)-1. These pathways are important in terms of sensing infection, but TLR4 is also associated with vascular inflammation and atherosclerosis. Here, we have compared TLR4 and NOD1 responses in hESC-EC with those of endothelial cells derived from other stem cells and with human umbilical vein endothelial cells (HUVEC). HUVEC, endothelial cells derived from blood progenitors (blood outgrowth endothelial cells; BOEC), and from induced pluripotent stem cells all displayed both a TLR4 and NOD1 response. However, hESC-EC had no TLR4 function, but did have functional NOD1 receptors. In vivo conditioning in nude rats did not confer TLR4 expression in hESC-EC. Despite having no TLR4 function, hESC-EC sensed Gram-negative bacteria, a response that was found to be mediated by NOD1 and the associated RIP2 signalling pathways. Thus, hESC-EC are TLR4 deficient but respond to bacteria via NOD1. This data suggests that hESC-EC may be protected from unwanted TLR4-mediated vascular inflammation, thus offering a potential therapeutic advantage.PLoS ONE 01/2014; 9(4):e91119. · 3.73 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Monomethylation of lysine 4 on histone H3 (H3K4me1) is a well-established feature of enhancers and promoters, although its function is unknown. Here, we uncover roles for H3K4me1 in diverse cell types. Remarkably, we find that MLL3/4 provokes monomethylation of promoter regions and the conditional repression of muscle and inflammatory response genes in myoblasts. During myogenesis, muscle genes are activated, lose MLL3 occupancy, and become H3K4-trimethylated through an alternative COMPASS complex. Monomethylation-mediated repression was not restricted to skeletal muscle. Together with H3K27me3 and H4K20me1, H3K4me1 was associated with transcriptional silencing in embryonic fibroblasts, macrophages, and human embryonic stem cells (ESCs). On promoters of active genes, we find that H3K4me1 spatially demarcates the recruitment of factors that interact with H3K4me3, including ING1, which, in turn, recruits Sin3A. Our findings point to a unique role for H3K4 monomethylation in establishing boundaries that restrict the recruitment of chromatin-modifying enzymes to defined regions within promoters.Molecular cell 03/2014; 53(6):979-92. · 14.61 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: We have recently reported that mouse embryonic stem cells (mESCs) are deficient in expressing type I interferons (IFN) when exposed to viral infection and double-stranded RNA. In this study, we extended our investigation and demonstrated that single-stranded RNA and protein-encoding mRNA can induce strong IFN expression and cytotoxicity in fibroblasts and epithelial cells, but none of the effects associated with these antiviral responses were observed in mESCs. Our results provided additional data to support the conclusion that mESCs are intrinsically deficient in antiviral responses. While our findings represent a novel feature of mESCs that in itself is important for understanding innate immunity development, we exploited this property to develop a novel mRNA-mediated gene expression cell model. Direct introduction of synthetic mRNA to express desired genes has been shown as an effective alternative to DNA/viral vector-based gene expression. However, a major biological challenge is that a synthetic mRNA is detected as a viral RNA analog by the host cell, resulting in a series of adverse effects associated with antiviral responses. We demonstrate that the lack of antiviral responses in mESCs effectively avoids this problem. mESCs can tolerate repeated transfection and effectively express proteins from their synthetic mRNA with expected biological functions, as demonstrated by the expression of green fluorescent protein and the transcription factor Etv2. Therefore, mRNA-based gene expression could be developed into a novel ESC differentiation strategy that avoids safety concerns associated with viral/DNA-based vectors in regenerative medicine.Stem cells and development 11/2013; · 4.15 Impact Factor
Innate Immunity in Human Embryonic Stem Cells:
Comparison with Adult Human Endothelial Cells
Ga ´bor Fo ¨ldes1., Alexander Liu1., Rekha Badiger1, Mark Paul-Clark1, Laura Moreno1, Zsuzsanna
Lendvai2, Jamie S. Wright1, Nadire N. Ali1", Sian E. Harding1*", Jane A. Mitchell1"
1National Heart and Lung Institute, Imperial College, London, United Kingdom, 2Heart Center, Semmelweis University, Budapest, Hungary
Treatment of human disease with human embryonic stem cell (hESC)-derived cells is now close to reality, but little is known
of their responses to physiological and pathological insult. The ability of cells to respond via activation of Toll like receptors
(TLR) is critical in innate immune sensing in most tissues, but also extends to more general danger sensing, e.g. of oxidative
stress, in cardiomyocytes. We used biomarker release and gene-array analysis to compare responses in hESC before and
after differentiation, and to those in primary human endothelial cells. The presence of cardiomyocytes and endothelial cells
was confirmed in differentiated cultures by immunostaining, FACS-sorting and, for cardiomyocytes, beating activity.
Undifferentiated hESC did not respond with CXCL8 release to Gram positive or Gram negative bacteria, or a range of PAMPs
(pathogen associated molecular patterns) for TLRs 1-9 (apart from flagellin, an activator of TLR5). Surprisingly, lack of TLR-
dependent responses was maintained over 4 months of differentiation of hESC, in cultures which included cardiomyocytes
and endothelial cells. In contrast, primary cultures of human aortic endothelial cells (HAEC) demonstrated responses to a
broad range of PAMPs. Expression of downstream TLR signalling pathways was demonstrated in hESC, and IL-1b, TNFa and
INFc, which bypass the TLRs, stimulated CXCL8 release. NFkB pathway expression was also present in hESC and NFkB was
able to translocate to the nucleus. Low expression levels of TLRs were detected in hESC, especially TLRs 1 and 4, explaining
the lack of response of hESC to the main TLR signals. TLR5 levels were similar between differentiated hESC and HAEC, and
siRNA knockdown of TLR5 abolished the response to flagellin. These findings have potential implications for survival and
function of grafted hESC-derived cells.
Citation: Fo ¨ldes G, Liu A, Badiger R, Paul-Clark M, Moreno L, et al. (2010) Innate Immunity in Human Embryonic Stem Cells: Comparison with Adult Human
Endothelial Cells. PLoS ONE 5(5): e10501. doi:10.1371/journal.pone.0010501
Editor: Massimo Federici, University of Tor Vergata, Italy
Received November 2, 2009; Accepted April 14, 2010; Published May 5, 2010
Copyright: ? 2010 Fo ¨ldes et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by: Wellcome Trust (http://www.wellcome.ac.uk/), British Heart Foundation (http://www.bhf.org.uk/), European Community,
NC3Rs (http://www.nc3rs.org.uk/), BBSRC (http://www.bbsrc.ac.uk/), Rosetrees Trust (http://www.rosetreestrust.co.uk/), SROP (http://www.nfu.hu/), and Geron
Corporation. A.L. was a recipient of a Wellcome Trust Vacation Scholarship. M.P.C. is a recipient of a Wellcome Trust University Award (083429/Z/07/Z) and R.B. is a
recipient of BHF Clinical Research Fellowship. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the
Competing Interests: Geron: The authors have a non-financial collaborative agreement, but this does not alter their adherence to all the PLoS ONE policies on
sharing data and materials.
* E-mail: firstname.lastname@example.org
. These authors contributed equally to this work.
" These authors also contributed equally to this work.
Human embryonic stem cells (hESC) are currently being
developed as sources of tissue-specific cells for the treatment of
human disease, including heart failure. It is hoped that hESC-
derived cells can re-seed and repair damaged tissues allowing
recovery of organ function. Although the immune response of the
host to implanted cells has been the subject of a much interest,
little is known about the innate immune response of the grafted
cells themselves. Many cells of the body express a full or partial
innate immune response, and these include both the endothelial
cells  and cardiomyocytes [2,3] which will be required to make
a viable cardiac graft. The innate immune response is often
modeled experimentally by activation of cells with pathogens or
pathogen associated molecular patterns (PAMPs). The best studied
of the PAMPs is lipopolysaccharide (LPS) derived from Gram
negative bacteria. PAMPs are sensed by cells via pattern
recognition receptors (PRRS), which include Toll like receptors
(TLRs) . LPS activates TLR4 which recruits adapter protein
pathways including MyD88, MAL, TRIF and TRAM to initiate
signaling events leading to activation of NFkB and the induction of
inflammatory genes including CXCL8. There are 10 TLRs
expressed in human cells with specific PAMPs for 1–9 identified.
TLR10 remains an orphan receptor at present. In addition to the
sensing of pathogens PRRs are now understood to sense host
ligands as part of a wider role in the surveillance of danger signals
. Where phenotypes of TLR knock-out mice have been studied
directly, TLR4 gene deletion, for example, is associated with
immune suppression, chronic inflammation of the lung ,
vascular compromise and evidence of heart failure . On the
other hand, TLR activation in the heart is involved in the
deleterious responses to oxidative stress , ischemia  and
septic cardiomyopathies , and various TLR knockout mice are
more resistant to these insults as well as to doxorubicin
cardiomyopathy  and hypertrophy . HESC originate
from the inner cell mass of the blastocyst, which if left undisturbed
PLoS ONE | www.plosone.org1May 2010 | Volume 5 | Issue 5 | e10501
would develop in the sterile environment of the womb. TLR
responses in the embryo develop relatively late, and are again
suppressed in the neonatal period . The question of whether
hESC and their derivatives in culture show a similar trend i.e.
whether they retain the immature phenotype or develop the
mature TLR responses, is therefore a vital question to understand
in order to establish their potential in repair of the heart and other
organs. In this paper we directly compare the expression and
activity of TLRs, their downstream signaling components and
NFkB signaling between undifferentiated and differentiated
human hESC as well as with fully differentiated mature human
aortic endothelial cells.
Materials and Methods
Heat inactivated E. coli and S. aureus were prepared as described
previously . Synthetic agonists for TLR1/2 (Pam3CSK4),
TLR2/6 (FSL-1), TLR3 [poly(I:C)], TLR5 (Flagellin), TLR7
(Imiquimod), TLR8 (E. coli K12 ssRNA) were purchased from
InvivoGen Co. (San Diego, USA). TLR4 ligand (LPS) was
obtained from Sigma-Aldrich (Dorset, UK). IL-1 was purchased
from R&D Systems (Abingdon, UK). All other reagents, unless
otherwise stated, were obtained from Invitrogen.
Human Embryonic Stem Cell Culture
Most of the experiments were performed using the H7 line from
Geron Corporation? (Menlo Park, CA, USA). However, gene
array analysis for undifferentiated stem cells was performed in H7,
SHEF2, SHEF4 and SHEF5 lines. H7 and SHEF lines are
ethically derived hESC lines. H7 was imported under a
collaboration agreement with Geron, and with permission from
the UK Stem Cell Bank.
Undifferentiated H7 cells were maintained under feeder-cell
free conditions in mouse embryonic fibroblast-conditioned medi-
um (MEF-CM), supplemented with 8 ng/ml of recombinant
human basic fibroblast growth factor as per Geron’s protocols
described previously . In brief, mouse embryonic fibroblasts
(MEFs) were obtained from pregnant mouse embryos of MF-1
strain. After propagation in 10% FCS-containing medium, MEFs
were mitotically inactivated with 0.01 mg/ml mitomycin C at
passage 4 for 2.5 hours and adhered overnight onto pre-
gelatinized T225 flasks (at a seeding density of 1.886107cells/
flask) in medium containing 10% FCS which was subsequently
replaced with 150 ml of hESC medium, supplemented with
recombinant human 4 ng/ml basic fibroblast growth factor.
Conditioned medium, collected daily for up to 10 days, was
passed through 0.2 mm low protein attachment cellulose acetate
filter units (Corning) prior to feeding H7 cells. The hESC medium
consisted of KnockOut DMEM (KO DMEM) supplemented with
20% KnockOut serum replacement (KOSR), 1 mM L-glutamine,
50 U/ml penicillin, 50 mg/ml streptomycin, 1% non-essential
amino acids (100x stock) and 0.1 mM b-mercaptoethanol. All
undifferentiated cells were cultured on Matrigel (BD Sciences)-
coated 6-well plates (Nunc, Roskilde, Denmark). Before induction
of differentiation, spontaneously differentiated cells were removed
by treatment with collagenase at 37uC for up to 10 min. hESC
colonies were mechanically broken with a 5 ml pipette tip and
were cultured for 4 days in low attachment 6-well plates (Nunc),
suspended in differentiation medium to form embryoid bodies.
The differentiation medium was the same as the hESC medium
except that the KOSR was replaced with 20% non-heat-
inactivated FCS. Embryoid bodies were plated out onto 0.5%
gelatinized dishes and cultured in order to allow continued
Cell Plating and Handling
Undifferentiated H7 cell colonies from 6-well plates were
removed and subcultured in 96-well plates previously coated with
Matrigel (100 ul/well). For this, medium was aspirated and
spontaneously differentiating cells among colonies were removed
by treatment with collagenase at 37uC for up to 10 minutes, after
which time collagenase was aspirated and cells in each well were
washed with 2 ml PBS. Colonies of undifferentiated H7 cells were
broken up mechanically with a 5-ml pipette tip and small clusters
were subcultured to .70% confluence. Differentiated hESC in
T175 flasks or 10-cm culture dishes were removed from the
surface by treatment with Trypsin-EDTA (Sigma-Aldrich),
counted and plated onto 96 well plates coated with 0.5% gelatin.
hESC-Derived Endothelial Cell (hESC-EC) Culture
Undifferentiated H7 hESC were dissociated into clumps and
placed into ultra low-attachment plates in medium containing 2%
FCS (Endothelial Growth Medium-2, Lonza). As described
elsewhere , CD31+ cells were sorted by a sterile cell sorter
(BD FACSAria, BD Biosciences) from cultures 13 days after
differentiation and propagated in endothelial growth medium.
Passages between 3 and 10 were used for experiments.
Matrigel Tubule-Forming Assay
Matrigel was diluted 1:2 with endothelial basal medium on ice
and then 100 ml/well added to 24-well plates and allowed to gel in
a thin layer at 37uC. CD31+ cells (50.000 per well) were seeded
onto the gels and tubes were photographed after 22 hours.
Endothelial Cell Culture
Primary human aortic endothelial cells were purchased from
Promocell (Heidelberg, Germany) and cultured according to
manufacturer’s instructions. The human endothelial cell line
(EAhy-926) was cultured in Dulbecco’s Modified Eagle’s Medium
(DMEM) supplemented with 1% Hypoxanthine-Aminopterin-
Thymidine (all Sigma-Aldrich), 5 mM L-glutamine, 100 U/ml
penicillin, 100 mg/ml streptomycin, and 10% heat-inactivated
To investigate the response to TLR stimulation, cells were
treated in 96 well plates with PAMPS which are agonists to TLR1-
8, or heat killed S. aureus and E. coli while IL-1 was used as a
positive control, for 24 hours. For experiments to measure NFkB
activation cells were treated for 1 hour.
Knockdown of TLR5
For TLR5 siRNA knockdown, ON-TARGETplus SMART-
pool TLR5 siRNA transfection was performed using DharmaFect
reagent (100 nM, final incubation volume 100 ml) per manufac-
turer’s instructions (Dharmacon, Thermo Scientific). Scrambled,
non-targeting siRNA (100 nM; Dharmacon, Thermo Scientific)
was used as negative controls. Fluorescent siGLO Red siRNA
indicator (100 nM; Dharmacon, Thermo Scientific) was used for
optimization and documentation of transfection efficacy (.90%)
after 24–48 hours.
Cell were fixed with 4% paraformaldehyde, permeabilized with
0.2% Triton X-100, and labeled with primary antibodies anti-
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org2 May 2010 | Volume 5 | Issue 5 | e10501
CD31 (PECAM-1, 1:100 dilution, Santa Cruz or Biolegend), anti-
CD34 (Abcam, 1:200), anti-myosin heavy chain (Ab15, 1:200,
Abcam, Cambridge), anti-von Willebrand factor (1:100, Dako).
Primary antibodies were detected with Alexa 488- (Invitrogen) and
Alexa 647- (Invitrogen) conjugated secondary antibodies (all
1:400). DNA was visualised with DAPI (0.5 mg/ml; Sigma). DiI-
labelled acetylated human low density lipoprotein (Ac-LDL) was
purchased from Invitrogen. Images were acquired on Zeiss Axio
Observer Z1 fluorescence microscopy.
Concentrations of CXCL8/IL8 in cell-free supernatants were
measured using sandwich ELISA kits (R&D Systems) and
calculated using 4-parameter-log-fit curves according to manufac-
Measurement of NFkB activation
Undifferentiated hESC were cultured in 6-well plates until 90%
of the well surfaces were covered with H7 colonies before they
were incubated with IL-1b (1 ng/ml) and LPS (1 mg/ml). After
one hour, the cellular nuclear extracts were prepared using a
commercially available nuclear extraction kit (Active Motif,
Carlsbad, CA, USA) according to the manufacturer’s protocols.
In brief, the cells were washed, collected in ice-cold PBS in the
presence of phosphate inhibitors and centrifuged at 5006g for
5 min. The resultant pellets were re-suspended in excess hypotonic
buffer, treated with detergent and centrifuged briefly at 14,0006g
for 30 sec. After the cytoplasmic fraction was collected, the nuclei
were lysed and nuclear proteins were dissolved in a cocktail of lysis
buffer and protease inhibitor. Nuclear protein concentrations were
determined using the Bradford assay before subsequently analyzed
for NFkB activation using the TransAMTMNFkB p65 transcrip-
tion factor assay kit (Active Motif), according to the manufacturer’s
instructions. This assay is based on nuclear NF-kB p65 proteins
binding to consensus NF-kB oligonucleotides fixed in 96-well
plates. In brief, 10 mg of nuclear proteins were added to each well
and incubated for 1 h to allow the binding of P65 to consensus
oligonucleotides. The presence of the resulting complex was
detected by a primary antibody. After the addition of a
horseradish peroxidase conjugated secondary antibody, P65 was
quantified by spectrophotometry.
the RNeasy kit (Qiagen, Hilden, Germany) according to
manufacturer’s protocols. Spin-column samples were centrifuged
at 95006g, 20uC for 15 seconds unless stated otherwise. In brief,
350–600 ml of lysis-Buffer RLT was added to cell-pellets and
homogenized by passing through a 20-gauge needle 5–10 times
before the addition of equal volume of 70% ethanol. The solution
was centrifuged in a spin-column before DNase digestion using
DNase-Buffer RDD solution (Qiagen) at room temperature for 25
minutes. Subsequently, RNA was washed with Buffer RW1 and
RPE under centrifugation. A final wash with Buffer RPE for 2
minutes was applied before the RNA was eluted in 30–60 ml of
determined by spectrophotometry (Spectramax, Switzerland) to
obtain A260/A280ratios and later by agarose gel electrophoresis.
Reverse transcriptase-PCR First-Strand reaction.
first-strand-complimentary-DNA (cDNA) was synthesized from
1 mg of total RNA using the RT2First-Strand Kit (SuperArray
Bioscience Corporation, MD, USA) according to manufacturer’s
protocols. Briefly, 10 ml of rt-Cocktail, containing RNA was mixed
with 10 ml Genomic-DNA-Elimination Mixture and incubated for
Total RNA was isolated from hESC using
15 minutes at 42uC then for 5 minutes at 95uC to inactivate/
degrade RNA and reverse transcriptase enzyme. 91 ml of RNase-
Free water was added to the remaining cDNA.
RT-PCR preparations were performed with RT2-
Profiler-PCR-Array kit (SuperArray) according to manufacturer’s
protocols. In brief, the experimental cocktail was made up
containing102 ml diluted
SuperArray-RT2-SYBR-Green/ROX-MasterMix and 1.173 ml
RNase-Free water. 25 ml of cocktail was added into each well of
the 96-well plate, containing forward and reverse primers for
TLR-related genes, housekeeping genes, human-genomic-DNA-
contamination and PCR controls. Samples were amplified using
ABI 7500 Real-Time-PCR System (Applied Biosystems, Foster
City, USA) for 40 cycles of 15 s at 95uC, 30 s at 55uC, and 30 s at
72uC. Relative levels of gene expression (fold differences) were
calculated according to manufactures instructions. CT values
which were at 35 or higher were considered as indicating
undetectable levels of expression. If 3/3 or 2/3 runs produced
CT.=35 for a given gene, the expression was considered
Quantitative RT-PCR for TLR5.
levels of TLR5 in undifferentiated and differentiated hESC
cultures, real-time PCR analyses were performed with TaqMan
Gene Expression Assay (Hs01920773_s1, Applied Biosystems,
CA). Human GAPDH Endogenous Control (FAM/MGB probe)
was used as a housekeeping control. The PCR was performed with
Rotor-Gene 3000 (Corbett Research) real-time PCR instrument
and the relative expression was determined.
first-strand cDNA,1.275 ml
For quantifying mRNA
Data is reported as the mean6 S.E. mean for n experiments.
Data was analyzed using one-way ANOVA followed by Dunnett’s
Multiple Comparison Test or by one-sample t-test for normalized
data as described in the respective legends.
Phenotype of human hESC differentiated to include
Undifferentiated hESC have a characteristic appearance as
tightly packed cells in colonies as shown in Fig. 1a and b. The H7
line, obtained from Geron Corp., Menlo Park CA, was grown
under feeder-free conditions as described previously  and
differentiated via embryoid body formation in 20% FCS. After
four days of differentiation in suspension cultures, embryoid bodies
were plated out onto gelatinized surfaces and continued to
differentiate in adherent cultures for prolonged periods (over 4
months). Figure 1C-F shows the morphology of cultures at 1 and 3
months after differentiation, demonstrating the emergence of a
variety of features, including clusters of beating cardiomyocytes
(video S1) and vessel-like structures. Immunocytochemical staining
for known markers (Fig. 2) confirms the presence of cardiomyo-
cytes and endothelial cells within the mixed population of cells.
By adjustment of differentiation conditions and using FACS
sorting for CD31 surface antigen, a highly expandable population
of human endothelial-like cells (hESC-EC) was obtained from the
hESC. Cells took on a cobblestone pattern in culture characteristic
of endothelial cells (Fig. 3A). Cells were stained positive for
endothelial-specific CD31 and CD34 markers (Fig. 3C–D) and
acetylated LDL uptake in culture (Fig. 3B). Further indicating
their endothelial phenotype and function, cells formed tube-like
structures on solidified Matrigel and showed migration on
fibronectin surface in wound healing assays (Fig. 3E–F).
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org3May 2010 | Volume 5 | Issue 5 | e10501
Relative expression of TLRs and related genes in hESC
and in endothelial cells
Expression of TLRs and downstream signaling effector genes
was determined in undifferentiated hESC as well as 1–4 months
after differentiation. Fig. 4A shows expression levels of TLRs 1, 3,
4, 5 and 6 in undifferentiated H7 cells and in each of 3 SHEF
lines. TLR 1, 3, 4 and 6 expression was consistently lower in hESC
compared to endothelial cells, with TLRs 1 and 4 particularly low.
Figure 1. Appearance of hESC cultures. Undifferentiated H7 cells (a, b); after 1 month (c, d) and 3 months (e, f) of differentiation. Examples of
clusters of beating cells are seen in c and d, and are shown in the Video S1. Vessel-like structures can be observed (e).
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org4 May 2010 | Volume 5 | Issue 5 | e10501
Figure 2. Presence of cardiomyocytes and endothelial cells in differentiated hESC cultures. Immunocytochemical staining of clusters of 1
month differentiated H7 hESC showing (a) cardiomyocytes myosin heavy chain (a, b) (green) with corresponding brightfield image and (b)
endothelial cells identified with von Willebrand factor (green), CD31 (red) and DAPI (nuclei, blue).
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org5 May 2010 | Volume 5 | Issue 5 | e10501
Levels of TLR8, and TLR10 were undetectable in hESC (Cycle
threshold, CT, values of 35 or lower in the majority of samples
(Table S1)). Levels of TLR2 were similar to or higher than those of
TLR6 in both H7 and SHEF lines, with CTvalues in the range
27–30 (Table S1), but as levels were undetectable in endothelial
cells in the majority of samples a fold-change was not calculated.
Low levels of expression of TLR7 and TLR9 were also present in
hESC but undetectable in endothelial cells. Interestingly, expres-
sion of TLR5 was robust in endothelial cells but even higher in
undifferentiated hESC. Agreement was good between H7 and
SHEF lines for TLR expression. Consistent with the activity of the
downstream signalling pathways, the majority of both NFkB
(Fig. 4B) and TLR signaling genes (Fig. 4C) were expressed at
similar levels between endothelial cells and hESC, though reduced
expression of MyD88 and TICAM1 might also contribute to the
poor functional response. Once again, there was reasonable
agreement between H7 and SHEF lines.
Relative expression of TLRs and related genes following
differentiation of hESC
Expression levels were determined at 1, 3 and 4 months after
differentiation of H7 hESC. Compared to undifferentiated H7
Figure 3. Characteristics of human embryonic stem cell-derived endothelial cells. The hESC-EC cultures showed (A) cobblestone
morphology. (B) Cells showed DiI-labelled acetylated LDL uptake, and were stained positive for human anti-CD31 antibody (green; C), and anti-CD34
(red; D). DAPI (blue) was used for nuclear staining. (E) Cells plated on solidified Matrigel form hollow-like tubes. (F) In a wound healing assay, cells
show migration on fibronectin surface (upper panel showing cell free area at 0 hours; cells migrate into the wound site at 22 hours).
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org6 May 2010 | Volume 5 | Issue 5 | e10501
(Fig. 5A), there was a general increase with time in TLRs 1, 2, 3, 4,
5 and 6, with all above undifferentiated levels by 4 months. Low
levels of TLRs 7, 9 and 10 also became sporadically apparent
(Table S1). However, comparing differentiated hESC with
endothelial cells (Fig. 5B) it is clear that TLR1 and TLR4 were
still at low levels, even after 4 months of differentiation. Only
modest adjustments of expression level of the NFkB (Table 1)
levels were seen during differentiation while some components of
Figure 4. Expression of TLR and TOLL or NFkB signalling genes in undifferentiated hESC. Relative expression levels in undifferentiated H7
hESC (n=3, open bars) and each of 3 SHEF lines (grey bars) compared to human aortic endothelial cells (HAEC, n=3). A: TLR genes. TLRs 2 and 7–10
were undetectable in the majority of HAEC. B: NFkB related genes; C: TOLL signalling-related genes. Statistical significance was calculated using one-
sample t-test for averaged H7 values against 1.0, *P,0.05, **P,0.01, ***P,0.001, n=3.
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org7 May 2010 | Volume 5 | Issue 5 | e10501
the TLR signaling pathways (TICAM1, TICAM2 and IRAK1)
were consistently increased (Table 2). A full list of gene expression
changes during differentiation is found in Table S2.
Effect of PAMPs for TLR1-8 and whole bacteria on CXCL8
release by human hESC and primary human endothelial
Release of CXCL8 was used in our study as a biomarker of cell
activation. In undifferentiated hESC there was no statistically
significant increase in CXCL8 release in cells stimulated for 24 h
with whole E. coli, S.Aureus or with an array of PAMPs for TLRs1,
2, 3, 4, 6, 7 or 8 (Fig. 6A). This was then compared with
differentiated hESC cultures, in which phenotypic evidence of
ESC specialization had been established (Fig. 2). One month after
initiation of differentiation, hESC-derived cells still did not release
increased levels of CXCL8 in response to PAMPs for TLRs or
whole bacteria (Fig. 6B). The exception was flagellin, an activator
of TLR5, which produced an increase in CXCL8 in both
undifferentiated hESC and differentiated hESC-derived cells
(Fig. 6A and B). Remarkably, even by 4 months of differentiation,
TLR5 remained the only PAMP to produce a response in hESC-
derived cells (Fig. 6C). IL-1b, which acts independently of TLRs
but via the TLR adapter protein MyD88, activated undifferen-
tiated or differentiated hESC to release CXCL8 (Fig. 6A–C). In
separate experiments protocols were repeated to compare directly
responses in purified cultures of hESC-EC and mixed differenti-
ated hESC cultures (Fig. 7A–B). Similar results were seen to those
presented in Fig. 6. LPS had no effect on CXCL8 release by mixed
differentiated cultures of hESCs (Fig. 7A) or purified hESC-ECs
(Fig. 7B). Again flagellin and IL-1b increased CXCL-8 release in
these experiments (Fig. 7A–B). Purified hESC-EC also expressed
TLR5, showing a 3-fold increase in mRNA levels as compared
with those in undifferentiated hESC (data not shown). By contrast
to results obtained with hESC and their differentiated derivatives,
primary cultures of adult human aortic endothelial cells released
increased levels of CXCL8 in response to Gram negative bacteria
and LPS (which activate TLR4), FSL-1 (which activates the
TLR2/6 heterodimer) or IL-1b (Fig. 6D). Similar results were seen
when the endothelial cell line, EAhy-926, was used (data not
shown). To ensure that differences in CXCL8 release between
hESC and primary endothelial cells were not due to the different
Figure 5. Expression of TLR genes in differentiated hESC.
Relative expression in differentiated H7 hESC at 1 month (n=2, open
bar), 3 months (grey bar) and 4 months (solid bar) compared to A:
averaged undifferentiated H7 hESC (n=3) and B: human aortic
endothelial cells (HAEC, n=3).
Table 1. NFkB related gene expression in hESC.
Table 1 Differentiated: fold change vs undifferentiated
Gene name 1 month1 month3 month 4 month
NFKBIA1.12 0.680.26 1.09
NFKBIL11.21 0.91 0.471.28
NFRKB1.61 0.830.28 0.91
c-REL0.75 0.41 0.120.76
IKBKB 0.770.41 0.310.60
Fold change of differentiated H7 hESC at 1–4 months compared to average
(n=3) undifferentiated H7, where a value of 1 represents equivalent gene
expression and of less than 1 represents lower expression in differentiated than
Table 2. TOLL signalling related gene expression in hESC.
Table 2Differentiated: fold change vs undifferentiated
Gene Name1 month 1 month3 month 4 month
MYD883.820.89 0.47 0.89
TICAM2 2.843.23 1.18 3.10
TRAF6 2.08 1.76 1.00 1.41
TICAM1 3.655.53 3.123.80
IRAK2 0.750.66 0.261.01
Fold change of differentiated H7 hESC at 1–4 months compared to average
(n=3) undifferentiated H7, where a value of 1 represents equivalent gene
expression and of less than 1 represents lower expression in differentiated than
undifferentiated. N.d.d = not detectable in differentiated cells.
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org8 May 2010 | Volume 5 | Issue 5 | e10501
media used for routine culture, both undifferentiated and
differentiated hESC were re-tested in the DMEM plus 10%
FCS used for EAhy-926 endothelial cell culture (Fig. S1). Use of
the EAhy-926 media altered basal CXCL8 levels to some extent
but did not reveal an effect of bacterial or PAMP stimulation on
CXCL8 release. However, it should be noted that the effects of
long term culture in highly specialized medium may well influence
TLR and other signaling pathways in cells.
Knockdown of TLR5 confirms specificity of flagellin
Silencing of TLR5 using siRNA resulted in a significantly reduced
CXCL8 release in flagellin-treated mixed differentiated hESC cultures
(Fig. 7A) or in purified hESC-EC (Fig. 7B) compared to scrambled
siRNA. Interestingly, TLR5 siRNA knockdown had a general
depressant effect on CXCL8 release in mixed differentiated cultures,
while the reduction was specific to flagellin in the hESC-EC.
Figure 6. Response of hESC and primary cells to bacteria and PAMPs. Release of CXCL8 from undifferentiated (A), 1 month- (B) or 4 month-
(C) differentiated H7 hESC or from primary cultures of human aortic endothelial cells (D). Cells were treated with Gram negative E. coli (C, 108CFU/
ml), Gram positive S. aureus (SA, 108CFU/ml), PAMPs for TLR2/1 (PAM3CSK4; 300 ng/ml), TLR2/6 (FSL-1; 1 mg/ml), TLR3 (Poly:IC; 10 mg/ml), TLR4
(LPS; 1 mg/ml), TLR5 (flagellin; 100 ng/ml), TLR7 (imiquimod; 10 mg/ml) or TLR8 (E. coli 12 ssRNA; 10 mg/ml) or IL-1b (IL-1; 1 ng/ml) for 24 hours
before CXCL8 was measured using ELISA. The data are the mean 6 S.E. mean for n=8211 (A); n=329 (B); n=3 (C) and n=426 (D). Statistical
significance was calculated by one-way analysis of variance followed by Dunnett’s Multiple Comparison Test (to control). A p value of less than 0.05
was considered statistically significant and denoted by *.
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org9 May 2010 | Volume 5 | Issue 5 | e10501
Effect of cytokines and the TLR4 PAMP LPS on cell and
NFkB activation in hESC
To further investigate the underlying mechanism for the lack of
TLR responses in hESC, the function of downstream pathways
was investigated. Effects of IL-1b on CXCL8 release by
undifferentiated hESC were compared with that produced by
TNFa, INFc and IL-6. It was found that cells were activated to
release increased levels of CXCL8 by IL-1b, TNFa, INFc but not
by IL-6 (Fig. 8A). In line with observations above, NFkB was
activated in undifferentiated hESCs following 1 h stimulation with
IL-1b whilst treatment of cells with LPS for the same period had
no effect (Fig. 8B). Clearly, as evidenced by our data, we confirm
the results of others [17,18,19] that NFkB genes are functionally
active in undifferentiated hESC but the TLR4 PAMP is unable to
produce a NFkB translocation in these cells.
The primary findings of the present study are first, that
undifferentiated hESC had no immune response associated with
bacterial TLR activation, or a range of bacterial PAMPs except for
TLR5. More surprisingly, even cultures of hESC differentiated for
up to 4 months had no detectable immune response to TLR2
(Gram positive) or TLR4 (Gram negative) activation. The TLR4
ligand, LPS also failed to activate NFkB in hESCs, consistent with
the notion that these cells do not have functionally active TLR
responses. A population of differentiated hESCs enriched for
Figure 7. Enzyme-linked immunosorbent assay of CXCL8
production by hESC and purified hESC-derived endothelial
cells. Bar graphs showing release of CXCL8 from 1 month-old
differentiated H7 hESC cells (A) or hESC-derived CD31+endothelial
cells (B) transfected with scrambled siRNA (open bars) or TLR5 siRNA
(solid bars) for 48 hours. Transfected cells were treated with TLR4 (LPS;
1 mg/ml), IL-1b (IL-1b; 1 ng/ml), or TLR5 (flagellin; 100 ng/ml) for
24 hours before CXCL8 was measured. The data are the mean 6 S.E.
mean for n=3. # P,0.05 vs control; * P,0.05 and *** P,0.001 vs
scrambled siRNA group.
Figure 8. Activation of NFkB in undifferentiated hESC. (A),
Undifferentiated H7 hESC were activated to release CXCL8 after
24 hours stimulation with TNFa (TNF), INFc (INF) or IL-1b (IL-1) but
not by IL-6 (10 ng/ml for each)(n=6). (B), NFkB was activated in
undifferentiated H7 hESC following 1 hour stimulation with IL-1b (1 ng/
ml) but not by LPS (1 mg/ml) (n=3). Statistical significance was
calculated by one-way analysis of variance followed by Dunnett’s
Multiple Comparison Test (to control). **P,0.01.
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org 10May 2010 | Volume 5 | Issue 5 | e10501
endothelial-like cells (hESC-EC) shared the lack of response
through TLR4 and active response through TLR5.
HESC were directly compared with endothelial cells cultured
from adult tissue, and these were shown to display positive
bacterial-TLR response as assessed by increased CXCL8 release
when cells were stimulated with agonists of TLR2 or TLR4. Gene
expression profiles in both undifferentiated and differentiated
hESC revealed relatively low levels of TLR1 or TLR4 in hESC
compared to levels expressed in endothelial cells. Moreover, TLR5
expression was relatively high in differentiated hESC compared to
endothelial cells. Whilst such gene expression data should be
interpreted with caution, this pattern of TLR expression fits well
with the ability of hESC to respond to PAMPs selective for these
TLRs. It should be noted that TLR2 acts as a dimer with either
TLR1 or TLR6. Loss of TLR1 would therefore indirectly
influence the activation through TLR2.
TLRs are linked via TIR domains to adapter protein pathways
such as MyD88. Our findings show that hESC and purified hESC-
EC responded to IL-1b, which is an agonist that works
independently of TLRs but via a TIR (TLR IL1 receptor) domain
linked receptor, by releasing increased levels of CXCL8 and
activation of NFkB. HESCs were found to express a number of
NFkB related genes. When levels were compared with those in
endothelial cells the relative expression of NFkB2, NFkB1A,
NFKNFRkB, REL and IKBkB were similar in hESC. However,
expression levels of NFkB1 (gene for p50) or RELA (gene for p65)
were lower in hESC than in endothelial cells. The lower gene
expression clearly did not limit NFkB activation or CXCL8
induction. It was interesting to note that, in addition to IL-1b,
hESCs responded by releasing increased levels of CXCL8 when
stimulated with other cytokines including TNFa, INFc, but not
IL-6. These data also suggests that whilst hESC and hESC-EC are
not able to sense and respond to bacterial PAMPs such as LPS,
they clearly have intact active inflammatory signaling pathways.
Gene expression analysis of these adapter proteins and related
genes showed some differences between hESC and the adult
endothelial cells. Whilst levels of TIRAP, TOLLIP and TRAF6
are similar in undifferentiated hESC and adult endothelial cells,
the levels of BTK, MyD88, TICAM1, TICAM2 and IRAK2 were
significantly lower. Interestingly the pattern of gene expression
changed somewhat when hESC were differentiated, with upregu-
lation of TICAM1, TICAM2 and IRAK1. Changes in expression
of NFkB-related genes during differentiation were modest and not
consistent, although RELA was upregulated while REL was
reduced (Table 2). However, functional levels of MyD88 and
associated signaling proteins were clearly sufficient in both
differentiated and undifferentiated hESC to mount a robust
inflammatory response to IL-1b.
NFkB signaling in differentiating hESC has been somewhat
controversial, with a study on the hES-NCL1 line showing
expression of NFkB and pathway components at significant levels
in undifferentiated cells but down-regulating during differentiation
and possibly controlling the differentiation process in this way
. In contrast, Kang et al  found very low NFkB in
undifferentiated SNUhES3 and MizES4 hESC lines compared to
HEK and haematopoetic progenitor lines (although MyD88 and
TRAF2 were equivalent), as well as poor CXCL8 induction by
TNFa. Differentiation was associated with increased NFkB and
IL8 response to TNFa, a finding that was replicated in mouse ESC
. The contradictory results in these studies might suggest
variation between hESC lines, but we have seen robust and similar
expression levels of the key NFkB pathway components in various
undifferentiated hESC (H7 and three of the SHEF lines), with little
change upon differentiation in H7.
The lack of response to bacterial challenge may not be
surprising for the undifferentiated hESC, given the delay in
development of innate immune sensing in the embryo. Full TLR
responses do not generally develop until near or even after full
term. Our results are in agreement with others  who also
reported lack of responsiveness of undifferentiated mouse ESC and
their differentiated derivatives to LPS. These authors further
reported that lack of TLR4 expression in these cells was due to
epigenetic modulation of the TLR4 gene promoter (methylated).
However, most recently a study by Lee and co-workers 
demonstrate positive expression of TLRs in mouse ESCs.
Moreover these authors demonstrated that after long term
exposure (24 days) increased proliferation and differentiation of
mouse ESCs was seen in cells stimulated with LPS (TLR4) or
POLY:IC (TLR3). The apparently conflicting observations in
mouse ESCs are likely influenced by different culture conditions,
time of LPS exposure, epigenetic factors and differing clones of
ESCs used. The likely similarities and differences between human
and mouse ESC cells in this and other respects remain the subject
of investigation. Clearly our data shows that hESC are resistant to
stimulation with PAMPs except for flagellin. We found that hESC
and purified hESC-EC expressed TLR5 and that knockdown of
TLR5 by siRNA inhibited CXCL8 production in response to
flagellin in both cultures. This suggests that TLR5 acts as an active
sensor for bacterial flagellin monomers in hESC. Evidence for the
cytoprotective role of TLR5 comes from studies showing that
TLR5 ligation can block apoptosis by activating downstream
antiapoptotic genes . Activation of TLR5 may protect against
tissue injury in conditions involving high levels of cell death .
The mechanism responsible for the lack of LPS responsiveness
in hESC is still to be determined. However, it has been observed
that mesoderm formation in embryoid bodies (EB) from hESC can
be inhibited (as shown by mesoderm specific gene, Brachury
silencing), by challenge with LPS, the TLR4 PAMP . This
suggests either that low TLR4 levels are able to produce a
sufficient response to affect differentiation processes with pro-
longed stimulation, or that the results were secondary to LPS
release of cytokines from the MEF layer present during EB
formation in that study.
Lack of innate immune and danger-sensing signals has different
implications for different cell lineages depending on whether, like
endothelium, they have a distinct role against pathogens or, like
cardiac myocytes, the pathways involved have been directed
towards more general danger signals. Endothelial cells not only
provide barrier and endocrine functions but are also essential
innate immune surveillance cells. Indeed, endothelial cells are
generally the first cell type that pathogens encounter in the
circulation. For many target cells, including endothelial cells, it is
essential that tissue derived from hESCs express a functional
bacterial innate immune response. The use of hESC-derived
endothelial cells for tissue repair may therefore be compromised
by the lack of innate immune sensing. It remains to be seen
whether the in vivo environment will stimulate maturation of the
hESC-derived cells and, if not, what mix of host and grafted
endothelium will be tolerable to maintain function. It may be
necessary to use pre-treatment strategies to accelerate maturation
in vitro such as mechanical or hormonal stimulation, provision of
extracellular matrix or co-culture with other cell types. For cell
types not directly involved in immune sensing, the consequences of
relative insensitivity to insult may have both positive and negative
aspects. Cell therapy will, in most cases, be directed to areas of
damage and implanted hESC-derived cells will be introduced to
areas of hypoxia or inflammation. Taking the example of cardiac
myocytes, the lack of TLR2 and 4 response would be predicted to
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org11 May 2010 | Volume 5 | Issue 5 | e10501
increase resistance to hypoxia  and so improve survival after
implantation in scar border zone, although the sensitivity to the
inflammatory cytokine milieu of the infarcted heart will be
retained. This might suggest implantation success would be
optimal at a later time period after infarction, when acute
inflammation has subsided and scar is more established.
In summary, we have shown for the first time that hESC do not
sense or respond to bacteria or the bacterial PAMPs that activate
TLR4, but do respond to flagellin which activates TLR5. We show
that this pattern of PAMP sensing is consistent with the relative
expression of TLR genes in hESCs. We show that despite having
no ability to sense LPS, hESC respond in a robust manner to
cytokines linked to MyD88 and NFkB transcription pathways.
These observations are important as they suggest that whilst
endothelial (and other) cells produced from hESC may display
phenotypic markers they do not express a mature immune
function. This has implications for the strategy and timing of
implantation of hESC derived cells for tissue repair.
ferentiated (A) or 1 month differentiated (B) hESC were tested
either in their routine culture media (as detailed in the methods)
(closed bars) or with 10% FCS-containing DMEM 24 hours
before and during stimulation (open bars). PAMPs or IL-1b was
added to cells for 24 hours before CXCL8 was measured by
ELISA (n=3). Statistical significance was calculated by one-way
analysis of variance followed by Dunnett’s Multiple Comparison
Test (to control). **P,0.01.
Found at: doi:10.1371/journal.pone.0010501.s001 (0.50 MB TIF)
Comparison of hESC culture media. Either undif-
n=3), undifferentiated H7 hESC (n=3); undifferentiated SHEF2,
SHEF4 and SHEF5 and differentiated H7 at 1 month (n=2), 3
months and 4 months after differentiation. Values .35 were set to
35: these were considered as undetectable.
Found at: doi:10.1371/journal.pone.0010501.s002 (0.36 MB
CT values for human aortic endothelial cells (HAEC,
(n=3) and differentiated H7 at 1 month (n=2), 3 months and 4
months after differentiation. N.d.d. = not detectable in differen-
tiated; n.d.u. = not detectable in 3/3 or 2/3 undifferentiated and
n.d. = not detectable in either.
Found at: doi:10.1371/journal.pone.0010501.s003 (0.14 MB
Fold changes between undifferentiated H7 hESC
beating cardiomyocytes, as shown in Fig. 1C and D.
Found at: doi:10.1371/journal.pone.0010501.s004 (2.26 MB
1 month differentiated H7 hESC showing clusters of
We would like to thank Prof Peter Andrews and Staff at the Centre for
Stem cell Biology, University of Sheffield for the generous gift of hESC
SHEF lines mRNA, Ms Hime Gashaw for help with CXCL8 ELISAs as
well as aspects of laboratory management and Louise Harrington for
endothelial cell immunocytochemistry. We are grateful to Dr. Joseph Gold,
Geron, for providing the H7 hESC line and invaluable help and advice.
Conceived and designed the experiments: GF AL NNA SEH JAM.
Performed the experiments: GF AL RB MPC LM ZL JW NNA. Analyzed
the data: GF AL RB MPC LM ZL NNA SEH JAM. Contributed
reagents/materials/analysis tools: JAM. Wrote the paper: GF NNA SEH
1. Opitz B, Hippenstiel S, Eitel J, Suttorp N (2007) Extra- and intracellular innate
immune recognition in endothelial cells. Thromb Haemost 98: 319–326.
2. Frantz S, Kobzik L, Kim YD, Fukazawa R, Medzhitov R, et al. (1999) Toll4
(TLR4) expression in cardiac myocytes in normal and failing myocardium. J Clin
Invest 104: 271–280.
3. Patel TA, Belcher E, Warner TD, Harding SE, Mitchell JA (2007) Identification
and characterization of a dysfunctional cardiac myocyte phenotype: role of
bacteria, Toll-like receptors, and endothelin. Shock 28: 434–440.
4. Kawai T, Akira S (2007) TLR signaling. Semin Immunol 19: 24–32.
5. Miyake K (2007) Innate immune sensing of pathogens and danger signals by cell
surface Toll-like receptors. Semin Immunol 19: 3–10.
6. Zhang X, Shan P, Jiang G, Cohn L, Lee PJ (2006) Toll-like receptor 4 deficiency
causes pulmonary emphysema. J Clin Invest 116: 3050–3059.
7. Harrington LS, Belcher E, Moreno L, Carrier MJ, Mitchell JA (2007)
Homeostatic role of Toll-like receptor 4 in the endothelium and heart.
J Cardiovasc Pharmacol Ther 12: 322–326.
8. Frantz S, Kelly RA, Bourcier T (2001) Role of TLR-2 in the activation of
nuclear factor kappaB by oxidative stress in cardiac myocytes. J Biol Chem 276:
9. Kim SC, Ghanem A, Stapel H, Tiemann K, Knuefermann P, et al. (2007) Toll-
like receptor 4 deficiency: smaller infarcts, but no gain in function. BMC Physiol
10. Frantz S, Ertl G, Bauersachs J (2007) Mechanisms of disease: Toll-like receptors
in cardiovascular disease. Nat Clin Pract Cardiovasc Med 4: 444–454.
11. Nozaki N, Shishido T, Takeishi Y, Kubota I (2004) Modulation of doxorubicin-
induced cardiac dysfunction in toll-like receptor-2-knockout mice. Circulation
12. Ha T, Li Y, Hua F, Ma J, Gao X, et al. (2005) Reduced cardiac hypertrophy in
toll-like receptor 4-deficient mice following pressure overload. Cardiovasc Res
13. Levy O (2007) Innate immunity of the newborn: basic mechanisms and clinical
correlates. Nat Rev Immunol 7: 379–390.
14. Cartwright N, Murch O, McMaster SK, Paul-Clark MJ, van Heel DA, et al.
(2007) Selective NOD1 agonists cause shock and organ injury/dysfunction in
vivo. Am J Respir Crit Care Med 175: 595–603.
15. Brito-Martins M, Harding SE, Ali NN (2008) beta(1)- and beta(2)-adrenoceptor
responses in cardiomyocytes derived from human embryonic stem cells:
comparison with failing and non-failing adult human heart. Br J Pharmacol
16. Nourse MB, Halpin DE, Scatena M, Mortisen DJ, Tulloch NL, et al. (2010)
VEGF induces differentiation of functional endothelium from human embryonic
stem cells: implications for tissue engineering. Arterioscler Thromb Vasc Biol 30:
17. Kang HB, Kim YE, Kwon HJ, Sok DE, Lee Y (2007) Enhancement of NF-
kappaB expression and activity upon differentiation of human embryonic stem
cell line SNUhES3. Stem Cells Dev 16: 615–623.
18. Kim YE, Kang HB, Park JA, Nam KH, Kwon HJ, et al. (2008) Upregulation of
NF-kappaB upon differentiation of mouse embryonic stem cells. BMB Rep 41:
19. Armstrong L, Hughes O, Yung S, Hyslop L, Stewart R, et al. (2006) The role of
PI3K/AKT, MAPK/ERK and NFkappabeta signalling in the maintenance of
human embryonic stem cell pluripotency and viability highlighted by
transcriptional profiling and functional analysis. Hum Mol Genet 15:
20. Zampetaki A, Xiao Q, Zeng L, Hu Y, Xu Q (2006) TLR4 expression in mouse
embryonic stem cells and in stem cell-derived vascular cells is regulated by
epigenetic modifications. Biochem Biophys Res Commun 347: 89–99.
21. Lee SH, Hong B, Sharabi A, Huang XF, Chen SY (2009) Embryonic Stem Cells
and Mammary Luminal Progenitors Directly Sense and Respond to Microbial
Products. Stem Cells 27: 1604–1615.
22. Zeng H, Wu H, Sloane V, Jones R, Yu Y, et al. (2006) Flagellin/TLR5
responses in epithelia reveal intertwined activation of inflammatory and
apoptotic pathways. Am J Physiol Gastrointest Liver Physiol 290: G96–G108.
23. Burdelya LG, Krivokrysenko VI, Tallant TC, Strom E, Gleiberman AS, et al.
(2008) An agonist of toll-like receptor 5 has radioprotective activity in mouse and
primate models. Science 320: 226–230.
24. Sivasubramaniyan K, Atluri RR, Sarda K, Arvind M, Balaji V, et al. (2008)
Endotoxin-induced silencing of mesoderm induction and functional differenti-
ation: role of HMGB1 in pluripotency and infection. Regen Med 3: 23–31.
TLR Pathways in Human hESC
PLoS ONE | www.plosone.org12 May 2010 | Volume 5 | Issue 5 | e10501