Synthesis and applications of 2,4-disubstituted thiazole derivatives as small molecule modulators of cellular development

Article abstract of DOI:10.1039/c3ob00005b

Understanding how the structure of molecules relates to their function and biological activity is essential in the development of new analogues with targeted activity. This is especially relevant in mediating developmental processes in mammalian cells and the regulation of stem cell differentiation. In this study, thiazole-containing small molecules were synthesised and investigated for their ability to induce the differentiation of human pluripotent stem cells and their derivatives. Analyses of cell morphology, cell viability, expression of cell surface markers and ability to induce cell differentiation and regulate neurite formation identified the analogue with the longest and most bulky hydrophobic side chain as possessing comparable or enhanced activity to all-trans-retinoic acid (ATRA). Interestingly, a shorter, less bulky, known thiazole compound reported to be isoform selective for the retinoic acid receptor β2 (RARβ2) agonist did not mediate differentiation under the conditions tested; however, activity could be restored by adjusting the structure to a longer, more bulky molecule. These data provide further insight into the complexity of compound design in terms of developing small molecules with specific biological activities to control the development and differentiation of mammalian cells.

Full text of DOI:10.1039/c3ob00005b

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Synthesis and applications of 2,4-disubstituted thiazole
derivatives as small molecule modulators of cellular
development†
Cite this: Org. Biomol. Chem., 2013, 11,
2323
Garr-Layy Zhou,a Daniel M. Tams,b Todd B. Marder,‡*a,c Roy Valentine,d
Andrew Whiting‡*a and Stefan A. Przyborski§*b,e
Understanding how the structure of molecules relates to their function and biological activity is essential
in the development of new analogues with targeted activity. This is especially relevant in mediating
developmental processes in mammalian cells and the regulation of stem cell differentiation. In this study,
thiazole-containing small molecules were synthesised and investigated for their ability to induce the
differentiation of human pluripotent stem cells and their derivatives. Analyses of cell morphology, cell
viability, expression of cell surface markers and ability to induce cell differentiation and regulate neurite
formation identified the analogue with the longest and most bulky hydrophobic side chain as possessing
comparable or enhanced activity to all-trans-retinoic acid (ATRA). Interestingly, a shorter, less bulky,
known thiazole compound reported to be isoform selective for the retinoic acid receptor β2 (RARβ2)
agonist did not mediate differentiation under the conditions tested; however, activity could be restored
by adjusting the structure to a longer, more bulky molecule. These data provide further insight into the
complexity of compound design in terms of developing small molecules with specific biological activities
to control the development and differentiation of mammalian cells.
Received 2nd January 2013,
Accepted 11th February 2013
DOI: 10.1039/c3ob00005b
www.rsc.org/obc
natural retinoids to isomerise or be removed by degradation
and/or metabolism potentially makes routine handling of these
Introduction
Retinoids are a class of signalling molecules that include small molecules difficult in the laboratory.
vitamin A along with its natural and synthetic analogues. These
Synthetic retinoids have been utilised with great success in
small molecules are involved in regulating important biological probing the cellular effects of retinoids while avoiding these
pathways from embryogenesis through to adult homeostasis, issues.3 In particular, synthetic retinoid EC23, and its sila-ana-
and influence the proliferation and differentiation of a diverse logue, contain a triple bond linker region that prevents isomer-
range of cell types, including those of the nervous system.1 All- isation and has been shown to be more stable and hence,
trans-retinoic acid (ATRA) in particular is the major metabolite more potent compared with ATRA.4 Indeed, enhanced differen-
of vitamin A; however, due to the presence of five conjugated tiation of human pluripotent TERA2.cl.SP12 embryonal carci-
double bonds, it is susceptible to photo-isomerisation into noma (EC) stem cells and ReNcell 197VM neural progenitor
different retinoic acid isomers. In cell culture studies such cells has been observed during treatment with EC23.4c
isomers were found to induce a variety of different cellular
effects compared to the use of ATRA.2 The propensity for
aDepartment of Chemistry, Durham University, Science Laboratories, South Road,
Durham, DH1 3LE, UK. E-mail: andy.whiting@durham.ac.uk
bSchool of Biological and Biomedical Sciences, Durham University, Science Laboratories,
South Road, Durham, DH1 3LE, UK. E-mail: stefan.przyborski@durham.ac.uk
cInstitut für Anorganische Chemie, Julius-Maximilians-Universität Würzburg, Am
Hubland, 97074 Würzburg, Germany. E-mail: todd.marder@uni-wuerzburg.de
dHigh Force Research Limited, Bowburn North Industrial Estate, Bowburn, Durham,
DH6 5PF, UK
eNETPark Incubator, Thomas Wright Way, Sedgefield, Co Durham, TS21 3FD, UK
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3ob00005b
‡Chemistry corresponding authors
§Biology corresponding author
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2323–2334 | 2323

View Article Online
Paper
Organic & Biomolecular Chemistry
The effects of retinoids are mediated primarily through therefore, that RARβ2 is the crucial transducer of the retinoic
binding to, and activation of, retinoic acid receptors (RARs) acid signal in neurons, and ligands selectively targeting this
and retinoid X receptors (RXRs) which are members of the isoform would greatly benefit studies exploring conditions
ligand-dependent transcription factor superfamily of nuclear associated with an absence or lack of function of this receptor.
receptors.5 Three subtypes of RARs exist: α, β and γ; all of However, the design of isoform-selective agonists represents
which have been studied extensively in the F9 murine embryo- an even more challenging prospect since isoforms of a particu-
nal carcinoma stem cell line, and have provided important lar RAR subtype possess identical ligand-binding domains,
insights into the different roles of each subtype.6 While ATRA and differences lie entirely within the ligand independent
and related synthetic retinoids such as EC23 are pan-agonists N-activation domain.11
for the RARs and are useful general inducers of stem cell
Despite these difficulties, agonists for RARβ2 have been ident-
differentiation, more attention is now being focused on the ified using a high throughput screening assay of a chemical
synthesis of subtype-selective ligands that would facilitate library, namely receptor selection and amplification technology
further studies into the function of these receptors. However, (R-SAT).12 Alkoxythiazole 1 was reported to be such a compound,
this is a complicated endeavour due to the highly similar nature displaying 78% of the activity for RARβ2 compared to the refer-
of the RAR ligand binding domains (LBDs): the domain of ence compound, AM-580. Thus, in order to explore any relation-
RARβ differs from that of RARα and RARγ by one and two resi- ship between isoform-selectively and the inducement of cellular
dues respectively. Nevertheless, small molecules exhibiting effects, 1 was synthesised along with two novel derivatives con-
subtype-selective activities have been discovered via high taining more bulky phenyl (2) and 1,1,4,4-tetramethyl-1,2,3,4-
throughput screening assays.7 A 3D comparison of the RAR tetrahydronaphthyl (3) side-chains with the aim of performing
LBDs found that the RARβ binding site was significantly larger in vitro assays on cell differentiation and neural development to
due to the presence of an additional cavity between H5 and H10 observe any changes in cell growth and morphology.
caused by the position of the I263 side chain.8 It may be postu-
lated, therefore, that ligands with larger side-chains are able to
occupy the additional space within the RARβ retinoid binding
site and hence, may acquire selectivity for this subtype.
Results and discussion
Synthesis of small molecules
Synthesis of the reported RARβ2-selective agonist 1 was based
on the literature procedure,12 broadly as outlined in Scheme 1.
Ethyl 4-cyanobenzoate was heated with equimolar equivalents
of 2-mercaptopropionic acid and pyridine under microwave
conditions to provide the desired hydroxythiazole 5 after
recrystallisation. However, despite TLC and MS analyses indi-
cating the presence of only one product, two species were
observed by 1H NMR spectroscopy and were found to be in-
separable. The overlapping ethyl signals and an additional
doublet and quartet peak in the alkyl region of the spectrum
were deduced to be consistent with the keto-tautomer 5b
(Scheme 2), the occurrence of which was not reported pre-
viously and constituted a significant proportion of the mixture
(approximately 15%).
The thiazole moiety is present in naturally occurring mole-
cules possessing important antibiotic,13 antitumour14 and
immunosuppressive15 properties. In particular, 4-hydroxy-1,3-
thiazoles are known as active inhibitors of 5-lipoxygenase16
and CDK517 and exist in different tautomeric forms. The enol-
Differential promoter usage and alternative splicing produce form is favoured in polar solvents and the keto-form in non-
the different isoforms observed with each RAR subtype; the polar solvents.18 The lower energy of the aromatic 4-hydro-
RARβ gene has four isoforms: β1–4. RARβ2 is the most abun- xythiazole structure, as well as the presence of an aromatic
dant isoform and is of particular interest in the area of neuro- substituent at the 2-position, is likely to account for the pre-
science research. In one study, up regulation of RARβ2 was ferred enol form in this case.19 Alkylation of the mixture of 5a
observed in both embryonic and adult mouse dorsal root and 5b would be expected to ‘lock’ the product in the aromatic
ganglia (DRG) neurons exposed to ATRA, along with a corres- form; indeed, after reaction with 2-bromoethyl ethyl ether to
ponding stimulation of neurite outgrowth.9 In models of give 6, loss of signals associated with the keto-tautomer was
nervous system injury, overexpression of RARβ2 by lentiviral observed by 1H NMR analysis of the product. Following a basic
vectors in adult DRG or corticospinal tract neurons resulted in hydrolysis, alkoxythiazole 1 was obtained in 56% yield after
axonal outgrowth and functional recovery.10 It would appear, recrystallisation.
2324 | Org. Biomol. Chem., 2013, 11, 2323–2334
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 1 Synthesis of alkoxythiazole 1.
with retinoid-like structures. Commercially available 4,4,5,5-
tetramethyl-2-phenyl-[1,3,2]-dioxaborolane 8 was reacted with
triflate 7, using K3PO4 and 3 mol% Pd(dppf)Cl2 as catalyst in
DMF/H2O at 80 °C. After 18 hours reaction and purification by
silica gel chromatography, 4-(5-methyl-4-phenyl-thiazol-2-yl)-
benzoic acid ethyl ester 9 was isolated in 69% yield. Sub-
sequent base hydrolysis of ester 9 by stirring with 3 equivalents
of LiOH·H2O in THF proved sluggish and low-yielding, despite
the addition of excess base and applying gentle heating over
several days. However, a microwave-assisted hydrolysis proved
more effective, providing the desired acid 2 in 88% yield after
recrystallisation (Scheme 2). Carboxylic acid 3 was prepared in
an analogous manner using 4,4,5,5-tetramethyl-2-(5,5,8,8-tetra-
methyl-5,6,7,8-tetrahydro-naphthalen-2-yl)-[1,3,2]-dioxaborolane
10 in 85% isolated yield.
Scheme 2 Keto–enol tautomerism of 5a.
In order to enable further functionalisation at the 4-posi-
tion by means of cross-coupling reactions, and hence, access
different analogues with more bulky side-chains and of
different overall lengths, a triflation reaction was performed
with the isolated keto–enol mixture 5. Procedures for the con-
version of hydroxythiazoles to triflates are known20 and have
been employed in the derivatisation of these compounds into
biologically useful agents.21
Initially, the Comins’ reagent was used as the triflating
agent; however, after stirring with 5 for 6 hours in tetrahydro-
furan, TLC analysis showed predominately a mixture of start-
ing materials. The use of triflic anhydride with pyridine also
did not provide the desired triflate; however, when pyridine
was replaced by triethylamine, 7 was obtained in 20% yield
after silica gel chromatography. The yield was improved by use
of a more hindered base, diisopropylethylamine (DIPEA),
which provided 7 in 81% yield after purification (eqn (1)).
Effect of thiazole retinoid-like small molecules on the
differentiation of human pluripotent stem cells
Having prepared alkoxythiazole analogues 1, 2 and 3, they
were assessed for their ability to induce the differentiation of
the pluripotent human EC stem cell line, TERA2.cl.SP12. Cells
of this lineage are proven models of human embryonic devel-
opment22 and have been used to study neural differentiation.23
Cultures of TERA2.cl.SP12 cells were incubated with com-
pounds 1, 2 and 3, supplemented in the culture media to a
final concentration of 10 μM for up to 14 days.
The effects of the test compounds on stem cell differen-
tiation were compared with the effect of the natural retinoid
ATRA, which acted as a positive control. Hence, cultures con-
sisting of cells treated with 10 μM ATRA in the culture media
were also established and incubated alongside each other.
Further cultures were set up and supplemented with the
loading vehicle DMSO to act as a negative differentiation
control. The latter were processed for analysis after 3 days. In
order to minimise ATRA degradation and isomerisation, all
ð1Þ
Suzuki–Miyaura cross-couplings (Scheme 3) were then per- culture flasks were handled under reduced light conditions.
formed with triflate 7 to provide different small molecules For reproducibility of results, cell cultures for each compound
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2323–2334 | 2325

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 3 Preparation of thiazole derivatives 2 and 3 via Suzuki–Miyaura cross coupling reactions.
were set up in triplicate for each time point and incubated for condition. The resulting cell solution from each flask was cen-
over 3, 7 and 14 days. trifuged then re-suspended in 0.1% BSA solution to enable cell
After 7 days, it became apparent that cell proliferation in counting. Cells from cultures exposed to ATRA, 1, 2 and 3 were
cultures exposed to compound 3 had slowed considerably, diluted in a staining solution and the proportion of live (intact
which is consistent with cells committing to differentiate in cytoplasmic membranes excluded the dye) to dead (compro-
response to ATRA. In contrast, cells treated with compounds 1 mised cell membranes stained positive for the dye) cells was
and 2 continued to proliferate, and very high cell numbers determined using a Viacount assay on a Guava EasyCyte cyto-
were observed (Fig. 1), which is a clear indication that these meter. The number of viable cells, percentage viability, and
molecules were unable to induce cell differentiation and arrest total cell counts were then recorded (Fig. 2).
cell proliferation at the concentration used. The appearance of
Continued cell proliferation in cultures treated with com-
the cells after 14 days followed a similar trend, i.e. cell prolifer- pounds 1 and 2 was reflected in the high cell count and
ation slowed in response to ATRA and compound 3, whereas number of viable cells, whereas with ATRA and compound 3
cultures exposed to compounds 1 and 2 became over-confluent these values were lower as cells exit the cell cycle and commit
due to continued cell proliferation.
to differentiation. However, cells exposed to compounds 1 and
After incubating for 7 days in compound-supplemented 2 showed lower percentage viability than with cultures treated
media, cells were prepared for an assay of cell viability by com- with ATRA and compound 3 which is consistent with sub-
bining culture media containing any potentially detached or optimal culture conditions due to high cell numbers and over
dead cells with trypsinised live cells from each experimental cell proliferation. The similarity of cell viability values
obtained for the ATRA and compound 3 treated cultures sup-
ports the earlier observation from cell morphology that these
compounds act in a comparable fashion when exposed to this
cell line.
In order to quantify the effects of the thiazole-containing
compounds on inducing differentiation of the TERA2.cl.SP12
cell line, cells were analysed for expression of known markers
for stem cell and differentiated cell phenotypes. Cells from
each experimental condition were incubated with antibodies
for the stem cell antigens SSEA-3 (globoseries stage specific
embryonic antigen-3) and TRA-1-60 (a keratin-sulphate-associ-
ated glycoprotein stem surface marker). Following this, incu-
bation with a secondary fluorescent antibody allowed the
proportion of cells expressing the marker to be determined
through flow cytometry (Fig. 3).
After 3 days, expression of SSEA-3 and TRA-1-60 in cultures
treated with ATRA and compound 3 began to decrease com-
pared to the DMSO negative control. This was especially clear
for SSEA-3 where levels decreased by approximately 55% and
Fig. 1 Morphological appearance of TERA2.cl.SP12 human pluripotent stem
42% in the ATRA and compound 3 cultures, respectively.
cells exposed to 10 μM of ATRA, 1, 2 and 3 for 7 days. Note the similar appear-
These data provide a strong indication of TERA2.cl.SP12 stem
ance of cells treated with ATRA and 3, and the high cell numbers in cultures
cells committing to differentiate in response to these com-
treated with 1 and 2, indicating cells continuing to proliferate rather than com-
pounds. However, no significant change in expression of these
mitting to differentiation. Scale bar represents 500 μm.
2326 | Org. Biomol. Chem., 2013, 11, 2323–2334
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 2 TERA2.cl.SP12 EC stem cells incubated for 7 days in culture media sup-
plemented with 10 μM of ATRA, 1, 2 or 3. The number of viable cells, percen-
tage viability and total cell counts were determined using a Viacount assay and
represented in graphical format (for reproducibility, results are the average of
triplicate assays SEM for each culture condition, n = 3).
Fig. 3 TERA2.cl.SP12 EC cells exposed to 10 μM of ATRA, 1, 2 and 3 for 14
days and analysed for expression of stem cell markers (SSEA-3 and TRA-1-60)
and a neural marker (A2B5). Expression of SSEA-3 and TRA-1-60 decreased sig-
nificantly with ATRA and 3, with a corresponding increase in A2B5 expression.
Conversely, a high level of stem cell character was observed for 1 and 2, with
minimal A2B5 expression, indicating that these analogues do not induce differ-
entiation. After 14 days, spontaneous differentiation of cell cultures exposed to
1 and 2 may account for the decrease in SSEA-3 and TRA-1-60 levels, along with
the high expression of A2B5. (For reproducibility, results are the average of tripli-
cate assays SEM for each culture condition, n = 3).
stem cell markers was observed for cells exposed to compounds
1 and 2, suggesting that such cells retain a more stem cell-like
phenotype. After 7 days, expression of SSEA-3 and TRA-1-60 in
ATRA and compound 3 treated cultures was greatly reduced
(Fig. 3), while a significant decrease in expression was also
observed for 1 (approximately 40% and 30% reduction in
expression of SSEA-3 and TRA-1-60, respectively). After 14 days,
expression of the stem cell antigens in the ATRA and compound
3 treated cultures remained minimal; however, it was surprising
to note at this point the dramatic reduction in TRA-1-60
expression for cultures treated with compounds 1 and 2, indicat-
ing that differentiating cells may be present.
also monitored by flow cytometry. As expected, A2B5
expression increased in response to ATRA over the course of
the experiment. A very similar pattern of expression was
recorded in cultures treated with compound 3. This suggests
that compound 3 is capable of inducing differentiation of
TERA2.cl.SP12 cells, and has a comparable level of activity to
Since TERA2.cl.SP12 cells are known to form neurons in
response to exposure to ATRA,20 expression of the antigen
A2B5 (ganglioseries antigen marking early-stage neural cells),
which is associated with differentiating neural cell types, was
ATRA. After
7 days, no significant increase in antigen
expression was observed in response to compounds 1 and 2,
yet by day 14, an elevated A2B5 expression was recorded that
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2323–2334 | 2327

View Article Online
Paper
Organic & Biomolecular Chemistry
was similar to the levels observed with ATRA and compound 3.
It would appear that differentiated cell types were present in
these cultures; however, further investigation is required to
determine if this was the result of spontaneous differentiation
in the cultures or sub-optimal growth conditions rather than
the effects induced by these compounds. Nevertheless, given
the length of time required before a change in cell phenotype
was observed, it is unlikely that compounds 1 and 2 behave as
true inducers of differentiation.
Effect of a thiazole retinoid-like small molecule on neurite
outgrowth from differentiated human pluripotent stem cells
As the thiazole retinoid-like small molecule, compound 3 was
shown to induce neural commitment by flow cytometric analy-
sis, it was hypothesised that these differentiated stem cells
would subsequently form neurites. To assess neurite out-
growth a human model of neurite formation was used. Sus-
pension aggregates of the human pluripotent stem cell TERA2.
cl.SP12 were differentiated with 0.1 μM EC23, ATRA or com-
pound 3 for 21 days by media supplementation, to induce
neural commitment. EC23 is a photo stable synthetic retinoid
that has been found to exhibit a higher level of activity towards
the induction of neural differentiation than ATRA.4,25 After 21
days, neurospheres from each treatment group were placed on
laminin and poly-D-lysine coated substrates to induce neurite
formation. Neurospheres were maintained for 10 days prior to
analysis.
To visualise and quantify neurite outgrowth from each cell
aggregate, immunocytochemical analysis of the neuronal
marker β-III-tubulin was performed. β-III-Tubulin staining
showed the formation of many individual neurites projecting
from the central neurosphere differentiated either with com-
pound 3 or EC23. Quantification of neurite number showed a
significant enhancement of neuritogenesis in stem cells differ-
Fig. 4 Neurite outgrowth from differentiated TERA2.cl.SP12 EC stem cells. Cell
aggregates were differentiated with 0.1 μM EC23, ATRA or 3 for 21 days. Differ-
entiated aggregates were placed onto a laminin and poly-D-lysine substrate for
10 days to allow neurites to form. Neurites were stained with the pan-neuronal
marker β-III-tubulin (green) and cell bodies were stained with DAPI (blue).
Results are triplicate. Error bars SEM, *p < 0.05.
entiated by compound 3 or EC23 compared to ATRA or the confluency, experimental conditions were set up involving the
undifferentiated control (Fig. 4).
withdrawal of growth factors from the culture media and the
The ability of EC23 or compound 3 to enhance neurite out- incorporation of either compound 1, ATRA or EC23 at a final
growth in this model is likely due to the stability of these com- concentration of 1 μM. Undifferentiated (growth media sup-
pounds. Increased stability may increase retinoic acid receptor plemented with FGF and EGF) and differentiated (without the
activation and the reduction in metabolite formation can addition of growth factors) control cultures were also esta-
result in loss of heterogeneous biological effects. While no blished to enable comparison of results. Cultures were main-
neurites formed in the undifferentiated group, the stem cells tained for 7 days after which phase contrast micrographs were
stopped proliferating and TUJ-1 negative cells migrated from taken and showed that as expected, the undifferentiated
the neurosphere, indicating that retinoic acid receptor acti- control cultures continued to proliferate and became over-
vation is essential and sufficient for neural differentiation in confluent. In contrast, the control differentiation cultures
this model.
ceased to proliferate and by day 7 appeared as zones of prolif-
erative cells surrounded by less populated areas of neural-type
cells. Cells exposed to compound 1 appeared similar to the
control differentiation cultures, while cells treated with ATRA
Effect of an RARβ2 agonist on the differentiation of human
neuroprogenitor stem cells
As reported above, the RARβ2 agonist, compound 1, was found or EC23 displayed predominately neuronal morphology com-
to be largely ineffective in its ability to induce the differen- pared to the control cultures.
tiation of EC stem cells. We therefore also assessed its ability
Immunocytochemical staining was then performed to
to modulate the differentiation of the neuroprogenitor cell record the expression of the pan-neuronal marker β-III-tubulin
line, ReNcell 197VM. Stock cultures of these progenitor cells and the marker of mature neurons, NF-200 (Fig. 5). Quantifi-
were grown and expanded according to previously described cation of positively stained cells for β-III-tubulin and NF-200
methods.24 Once the cells had reached approximately 75% showed that, compared to the differentiated control cultures,
2328 | Org. Biomol. Chem., 2013, 11, 2323–2334
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 5 Effect of supplementation of ReNcell 197VM culture media with 1 μM compound 1, ATRA or EC23 on the induction of neural differentiation as observed by
immunocytochemical staining for markers for β-III-tubulin (a general neuronal marker protein) and neurofilament protein NF-200 (a 200 kDa protein expressed in
mature neurons) compared to undifferentiated (undiff ) and differentiated (control diff ) control cultures. Scale bar represents 100 μm.
compound 1 did not induce significant neuronal differen-
tiation (Fig. 6). Control differentiation cultures without reti-
noid supplementation stained for the general neuronal marker
β-III-tubulin and showed minimal levels of NF-200 expression,
which is indicative of an immature neuronal phenotype. In
contrast, cultures exposed to ATRA displayed an approximately
1.5-fold increase in cells staining positive for β-III-tubulin
along with a 3-fold increase in those staining positive for
NF-200. Cells treated with EC23 displayed a higher level of
neuronal morphology with a 2.25-fold and 5-fold increase in
β-III-tubulin- and NF-200-positive cells, respectively. The
increase in expression of both neuronal markers in response
to ATRA and EC23 reflects the formation of more mature
neurons and is consistent with previous observations.4b
A likely explanation for the apparent greater potency of
EC23 over ATRA may lie in its superior stability4 and potential
lack of metabolism. The result of this is a higher effective con-
centration of EC23 during incubation with cells under con-
ditions where ATRA would be expected to both degrade and be
metabolised. In addition, EC23 may provide stronger inter-
actions with retinoid receptors than the natural ligands.
Although EC23 is a potent inducer of neuronal differentiation,
the mechanism through which this synthetic retinoid medi-
ates its intracellular effects remains to be fully elucidated.
Fig. 6 Graphical representation of positive cell counts for the neuronal marker
β-III-tubulin (A) and the mature neuronal marker NF-200 (B) in cultures exposed
to 1 μM 1, ATRA or EC23 compared to undifferentiated (undiff ) and differen-
tiated (con diff ) control cultures. Triplicate analyses were performed for reprodu-
Conclusions
cibility with data representing mean SEM, n = 3, *p < 0.05.
The experimental evidence demonstrates clearly that com-
pound 3 is able to induce differentiation of the human TERA2.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2323–2334 | 2329

View Article Online
Paper
Organic & Biomolecular Chemistry
cl.SP12 embryonal carcinoma stem cells on a level comparable were recorded in CDCl3; tetramethylsilane (TMS) was used as
to that of the natural retinoid ATRA. Furthermore, at 0.1 μM, 3 the internal standard and spin multiplicities are indicated by
demonstrates enhanced neural commitment and neurite out- the following symbols: s (singlet), d (doublet), t (triplet), m
growth over ATRA with levels that are comparable to the potent (multiplet). J coupling constants are given in Hz. ES-MS was
stable synthetic retinoid EC23. Further characterisation is performed by the Durham University departmental service
required to determine whether compound 3 demonstrates using an Acquity TQD (Waters UK Ltd) mass spectrometer and
specificity for a particular retinoic acid receptor subtype, and if accurate mass measurements were obtained on a Thermo
this may contribute toward the positive results observed.
LTQ-FT spectrometer. Elemental analyses were carried out
The shorter, less bulky, thiazole containing compound 1 using an Exeter Analytical E440 machine. IR spectra were
has been previously classified as a RARβ2 agonist. However, recorded on a Perkin Elmer FT-IR spectrometer with an ATR
this molecule showed limited ability to modulate the differen- attachment. Melting point values were measured on a Sanyo
tiation of either the pluripotent TERA2.cl.SP12 or the neuro- Gallenkamp apparatus and are uncorrected.
progenitor ReNcell 197VM cell lines. Although RARβ2
4-(4-Hydroxy-5-methyl-thiazol-2-yl)-benzoic acid ethyl ester
signalling is understood to play a role in neurite outgrowth,9,10 (5). Ethyl 4-cyanobenzoate (1.29 g, 7.4 mmol), 2-mercaptopro-
it did not have a positive effect of neuritogenesis like ATRA or pionic acid (0.64 mL, 7.4 mmol) and pyridine (0.59 mL,
EC23 under the conditions tested. It remains possible that 7.4 mmol) were thoroughly mixed in a MW vial and heated at
activation of RARβ2 is a necessary, but not sufficient, factor for 150 °C for 75 minutes (5 × 15 min periods). The resulting
neuronal differentiation by these types of human cells.
yellow solid was dissolved in ethyl acetate (60 mL). Undis-
Overall these results further demonstrate how subtle modi- solved material was removed by filtration, and recrystallisation
fications to the structure of small molecules, specifically the from ethyl acetate produced yellow crystals that were washed
function of the bulky side chain, can have a significant effect with acetonitrile to provide 5 as a yellow crystalline solid (1.50 g,
on their biological activity. Such information about structure 77%); mp 203–205 °C; νmax (neat, cm−1) 2980 (C–H), 1709
activity relationships will advance our ability to design new (CvO), 1580 (CvC), 1510 (CvC), 1470 (CvC), 1271 (C–O), 1100
compounds developed for specific biological applications.
(C–O); [a: δH (400 MHz, CDCl3) 8.11 (2H, unsymmet. d, J 8.4 Hz,
Ar), 7.88 (2H, unsymmet. d, J 8.4 Hz, Ar), 4.40 (2H, q, J 7.2 Hz,
CH2), 2.36 (3H, s, CH3), 1.42 (3H, t, J 7.2 Hz, CH3); δC (176 MHz,
CDCl3) 166.0, 159.1, 158.7, 136.5, 131.1, 130.3, 125.4, 105.6, 61.2,
14.3, 9.4 and b: δH (400 MHz, CDCl3) 8.19–8.13 (4H, m, Ar), 4.41
(2H, q, J 7.2 Hz, CH2), 4.29 (1H, q, J 7.6 Hz, CH), 1.76 (3H, d,
Experimental
General experimental
Reagents were purchased from Sigma-Aldrich, Acros or Alfa- J 7.6 Hz, CH3), 1.43 (3H, t, J 7.2 Hz, CH3); δC (176 MHz, CDCl3)
Aesar and used without further purification unless otherwise 193.8, 165.3, 159.1, 135.9, 135.5, 130.0, 128.6, 61.7, 49.4, 18.1,
stated. Solvents were dried before use with appropriate drying 14.2]; m/z (ESI) 264 (M + H); λmax (EtOH) 244 nm (ε 4950 M−1
agents. Where indicated, reagents were combined in an Inno- cm−1), 356 (5000); Anal. calcd for C13H13NO3S: C, 59.30; H, 4.98;
vative Technology Inc. nitrogen-filled (BOC) glovebox. All glass- N, 5.32; found C, 58.99; H, 4.97; N, 5.34.
ware was oven-dried (130 °C) prior to use. Microwave (MW)-
4-[4-(2-Ethoxy-ethoxy)-5-methyl-thiazol-2-yl]-benzoic
acid
assisted reactions were carried out in an Emrys™ Optimizer ethyl ester (6). Compound 5 (197 mg, 0.75 mmol), 2-bromo-
(Personal Chemistry) in septum-containing, crimp-capped, ethyl ethyl ether (255 μL, 2.26 mmol), Cs2CO3 (268 mg,
sealed vials with automatic wattage adjustment to maintain 0.83 mmol), KI (373 mg, 2.23 mmol) and CH3CN (3.75 mL)
the desired temperature for a specified period of time. Reac- were mixed in a MW vial and irradiated at 150 °C for 10 min.
1
tions were monitored in situ by TLC, GC-MS or H NMR spec- The filtrate was extracted with ethyl acetate (20 mL) and
troscopy to ensure consumption of starting materials before washed with brine (3 × 30 mL). The organic phase was dried
reaction workup. Thin layer chromatography (TLC) was per- with Na2SO4, filtered and concentrated onto Celite. Purifi-
formed on Polygram SIL G/UV254 plastic-backed silica gel cation of the residue by flash chromatography produced 6 as a
plates with visualisation achieved using a UV lamp. Column viscous yellow oil (193 mg, 77%); νmax (neat, cm−1) 2976 (C–H),
chromatography was performed with Davisil Silica gel, 1715 (CvO), 1273 (C–O), 1106 (C–O); δH (400 MHz, CDCl3)
60 mesh. GC-MS was performed using an Agilent Technologies 8.05 (2H, unsymmet. d, J 8.4 Hz, Ar), 7.89 (2H, unsymmet. d, J
6890 N gas chromatograph equipped with a 5973 inert mass 8.4 Hz, Ar), 4.54–4.50 (2H, m, CH2), 4.40 (2H, q, J 7.2 Hz, CH2),
selective detector and a 10 m fused silica capillary column (5% 3.81–3.77 (2H, m, CH2), 3.61 (2H, q, J 6.8 Hz, CH2), 2.33 (3H, s,
cross-linked phenylmethylsilicone) using the following operat- CH3), 1.41 (3H, t, J 7.2 Hz, CH3), 1.24 (3H, t, J 6.8 Hz, CH3); δC
ing conditions: injector temperature 250 °C, detector tempera- (176 MHz, CDCl3) 166.1, 160.1, 157.7, 137.7, 130.7, 130.1,
ture 300 °C, oven temperature was ramped from 70 °C to 125.0, 108.9, 69.8, 69.2, 66.6, 61.1, 15.2, 14.3, 9.4; m/z (ESI) 336
280 °C at 20 °C min−1. UHP helium was used as the carrier (M + H); λmax (EtOH) 244 nm (ε 33 880 M−1 cm−1), 352
gas. All NMR spectra were recorded on either Bruker Avance- (26 460); HRMS (ESI) calcd for C17H22NO4S 335.1191 (M + H),
400, Varian Mercury-400, Varian Inova-500 or Varian VNMRS found 335.1185.
1
700 spectrometers at the following frequencies: H: 200, 400,
4-[4-(2-Ethoxy-ethoxy)-5-methyl-thiazol-2-yl]-benzoic
acid
500 and 700 MHz; 13C: 176 MHz; 19F: 658.4 MHz. NMR spectra (1). Compound 6 (183 mg, 0.55 mmol), LiOH·H2O (68 mg,
2330 | Org. Biomol. Chem., 2013, 11, 2323–2334
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
1.62 mmol), H2O (825 μL) and THF (2.75 mL) were mixed in a 8.4 Hz, Ar), 8.03 (2H, unsymmet. d, J 8.4 Hz, Ar), 7.73 (2H, dd,
MW vial and irradiated at 150 °C for 8 minutes. The reaction J 7.7 and 1.4 Hz, Ar), 7.47 (2H, td, J 7.7 and 1.4 Hz, Ar), 7.38
mixture was acidified with 1 M HCl, extracted with ethyl (1H, tt, J 7.7 and 1.4 Hz, Ar), 4.40 (2H, q, J 7 Hz, CH2), 2.64
acetate (3 × 20 mL) and washed with brine (3 × 60 mL). After (3H, s, CH3), 1.42 (3H, t, J 7 Hz, CH3); δC (176 MHz, CDCl3)
drying with MgSO4 and removal of the solvent in vacuo the 166.5, 162.6, 153.0, 137.9, 135.2, 131.5, 130.5, 130.0, 129.0,
crude product was obtained and recrystallised from ethyl 128.8, 128.1, 126.4, 61.5, 14.7, 13.4; m/z (ESI) 324 (M + H), 325
acetate to give 1 as a pale yellow crystalline solid (95 mg, 56%); (M + 2H), 670 (2M + Na); λmax (EtOH) 254 nm (ε 22 400 M−1
mp 184–185 °C; νmax (neat, cm−1) 2858 (C–H), 2554 (O–H), cm−1), 330 (18 100); Anal. calcd for C19H17NO2S: C, 70.56; H,
1673 (CvO), 1606 (CvC), 1432 (CvC) 1125 (C–O); δH 5.30; N, 4.33; found C, 70.62; H, 5.32; N, 4.44.
(500 MHz, CDCl3) 8.11 (2H, unsymmet. d, J 8.5 Hz, Ar), 7.93
4-[5-Methyl-4-(5,5,8,8-tetramethyl-5,6,7,8-tetrahydro-naphtha-
(2H, unsymmet. d, J 8.5 Hz, Ar), 4.55–4.51 (2H, m, CH2), len-2-yl)-thiazol-2-yl]-benzoic acid ethyl ester (11). Compound
3.82–3.78 (2H, m, CH2), 3.61 (2H, q, J 7 Hz, CH2), 2.34 (3H, s, 7 (500 mg, 1.26 mmol), 1026 (477 mg, 1.52 mmol, 1.2 equiv.),
CH3), 1.25 (3H, t, J 7 Hz, CH3); δC (176 MHz, CDCl3) 171.1, K3PO4 (537 mg, 2.53 mmol) and Pd(dppf)Cl2 (28 mg,
160.5, 157.8, 138.9, 131.1, 129.7, 125.4, 109.7, 70.2, 69.6, 67.1, 0.0383 mmol, 3 mol%) were combined in DMF and H2O
15.6, 9.9; m/z (ESI) 308 (M + H); λmax (EtOH) 240 nm (ε 2760 (10 : 1) in a Young’s tube under an inert nitrogen atmosphere
M−1 cm−1), 348 (2940); Anal. calcd for C15H17NO4S: C, 58.61; and heated at 80 °C for 18 h. The mixture was dissolved in
H, 5.57; N, 4.56; found C, 58.29; H, 5.43; N, 4.19.
diethyl ether and washed twice with water. The organic layer
4-(5-Methyl-4-trifluoromethanesulfonyloxy-thiazol-2-yl)-benzoic was dried with MgSO4 and the solvent removed in vacuo to give
acid ethyl ester (7). DIPEA (2.65 mL, 15.2 mmol) was added to a white residue which was purified by flash chromatography,
a stirred solution of 5 (2.0 g, 7.60 mmol) in CH2Cl2. The flask Rf 0.33 (8% EtOAc/hexanes). Recrystallisation from ethanol
was cooled to −78 °C after which trifluoromethanesulfonic afforded 11 as a fluffy white crystalline solid (410 mg, 75%);
anhydride (4.4 g, 15.2 mmol) was carefully added. The reaction mp 161–162 °C; νmax (neat, cm−1) 1707 (CvO), 1270 (C–O),
was allowed to warm slowly to RT and was stirred overnight 1102 (C–O), 772 (C–H); δH (700 MHz, CDCl3) 8.09 (2H, unsym-
under a nitrogen atmosphere. CH2Cl2 and water were added met. d, J 8.4 Hz, Ar), 8.03 (2H, unsymmet. d, J 8.4 Hz, Ar), 7.61
and the product extracted with three portions of CH2Cl2. The (1H, d, J 1.4 Hz, Ar), 7.49 (1H, dd, J = 8.4 and 1.4 Hz, Ar), 7.40
organic extracts were combined and washed three times with (1H, d, J 8.4 Hz, Ar), 4.40 (2H, q, J 7 Hz, CH2), 2.63 (3H, s,
water. After drying with MgSO4 the solvent was removed CH3), 1.75–1.70 (4H, m, 2 × CH2), 1.42 (3H, t, J 7 Hz, CH3),
in vacuo and the brown residue purified by flash chromato- 1.34 (6H, s, 2 × CH3), 1.32 (6H, s, 2 × CH3); δC (176 MHz,
graphy to give a yellow oil, Rf 0.32 (20% EtOAc/hexanes). CDCl3) 166.5, 162.3, 153.5, 145.1, 144.9, 138.1, 132.3, 131.4,
Recrystallisation from hexane produced 7 as an off-white crys- 130.4, 129.4, 127.2, 127.1, 126.4, 126.2, 61.5, 35.5, 35.4, 34.7,
talline solid (2.1 g, 70%); mp 48–50 °C; νmax (neat, cm−1) 1712 34.6, 32.3, 32.2, 14.7, 13.3; m/z (ESI) 434 (M + H); λmax (EtOH)
(CvO), 1422 (CvC), 1274 (CvC), 1221 (C–O), 1134 (C–F), 262 nm (ε 27 400 M−1 cm−1), 335 (17 600); Anal. calcd for
1093 (C–F), 768 (C–H), 693 (C–H), 602 (C–H); δH (700 MHz, C27H31NO2S: C, 74.79; H, 7.21; N, 3.23; found C, 74.67; H,
CDCl3) 8.10 (2H, unsymmet. d, J 8 Hz, Ar), 7.91 (2H, unsym- 7.22; N, 3.17.
met. d, J 8 Hz, Ar), 4.41 (2H, q, J 7 Hz, CH2), 2.49 (3H, s, CH3),
4-(5-Methyl-4-phenyl-thiazol-2-yl)-benzoic acid (2). Com-
1.42 (3H, t, J 7 Hz, CH3); δC (176 MHz, CDCl3) 166.1, 161.4, pound 9 (100 mg, 0.31 mmol) and LiOH·H2O (39 mg,
148.5, 136.4, 132.6, 130.6, 125.9, 122.6, 61.7, 14.7, 10.4; 0.93 mmol) were mixed in THF–H2O (2 mL, 3 : 1 ratio) in a
δF (658.4 MHz, CDCl3) −72.5; m/z (ESI) 263 (M − CF3SO2), 396 MW vial and irradiated at 160 °C for 15 min after which the
(M + H), 418 (M + Na); λmax (EtOH) 312 nm (ε 23 100 M−1 reaction mixture was acidified with 1 M HCl and ethyl acetate
cm−1); Anal. calcd for C14H12F3NO5S2: C, 42.53; H, 3.06; N, was added. The organic layer was washed twice with brine and
3.54; found C, 42.50; H, 3.07; N, 3.48.
dried with MgSO4. After in vacuo solvent removal the residue
4-(5-Methyl-4-phenyl-thiazol-2-yl)-benzoic acid ethyl ester was recrystallised from ethanol to give 2 as a white crystalline
(9). Compound 7 (500 mg, 1.26 mmol), phenylboronic acid solid (81 mg, 88%); mp 266–269 °C; νmax (neat, cm−1) 2847 (O–
pinacol ester (284 mg, 1.39 mmol, 1.1 equiv.), K3PO4 (537 mg, H), 1674 (CvO), 1284 (C–O), 858 (C–H), 771 (C–H), 700 (C–H),
2.53 mmol) and Pd(dppf)Cl2 (28 mg, 0.0383 mmol, 3 mol%) 681 (C–H); δH (500 MHz, DMSO-d6) 8.09 (4H, s, Ar), 7.80 (2H,
were combined in DMF and H2O (10 : 1) in a Young’s tube dd, J 8 and 1.5 Hz, Ar), 7.55 (2H, td, J 8 and 1.5 Hz, Ar), 7.46
under an inert nitrogen atmosphere and heated at 80 °C for (1H, tt, J 8 and 1.5 Hz, Ar), 2.68 (3H, s, CH3); δC (176 MHz,
18 h. The mixture was dissolved in diethyl ether and washed CDCl3) 167.7, 162.3, 152.5, 137.5, 135.3, 131.2, 131.2, 129.4,
twice with water. The organic layer was dried with MgSO4 and 129.2, 128.8, 126.8, 13.6; m/z (ESI) 294 (M − H); λmax (EtOH)
the solvent removed in vacuo to give an orange/brown residue 255 nm (ε 23 000 M−1 cm−1), 329 (19 000); Anal. calcd for
which was purified by column chromatography (3 → 5% C17H13NO2S: C, 69.13; H, 4.44; N, 4.74; found C, 68.79; H,
EtOAc/hexanes) to give a white solid, Rf 0.28 (10% EtOAc/ 4.46; N, 4.74.
hexanes). Recrystallisation from ethanol produced 9 as a white
crystalline solid (246 mg, 60%); mp 85–87 °C; νmax (neat, cm−1
4-[5-Methyl-4-(5,5,8,8-tetramethyl-5,6,7,8-tetrahydro-naphtha-
len-2-yl)-thiazol-2-yl]-benzoic acid (3). Compound 11 (100 mg,
)
1706 (CvO), 1278 (C–O), 1100 (C–O), 860 (C–H), 768 (C–H), 0.23 mmol) and LiOH·H2O (29 mg, 0.69 mmol) were mixed in
699 (C–H); δH (700 MHz, CDCl3) 8.10 (2H, unsymmet. d, J THF–H2O (2 mL, 3 : 1 ratio) in a MW vial and irradiated at
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2323–2334 | 2331

View Article Online
Paper
Organic & Biomolecular Chemistry
160 °C for 15 min after which the reaction mixture was acidi- creating a single cell suspension and adding 1.5 × 106 cells per
fied with 1 M HCl and ethyl acetate was added. The organic 20 mL of maintenance media to a 90 mm un-treated Petri
layer was washed twice with brine and dried with MgSO4. After dish. The cells were left to aggregate overnight and sub-
in vacuo solvent removal the residue was washed with ethanol sequently treated with the test compound. Aggregates were cul-
to give 3 as a white crystalline solid (80 mg, 85%); mp tured for 21 days in the presence of the test compound to
276–278 °C; νmax (neat, cm−1) 2929 (O–H), 1681 (CvO), 1279 induce neural commitment and the production of neuro-
(C–O), 773 (C–H); δH (500 MHz, DMSO-d6) 8.12–8.05 (4H, m, Ar), spheres prior to analysis of neurite outgrowth. To induce
7.66 (1H, d, J 1.5, Ar), 7.52 (1H, dd, J 8.5 and 1.5, Ar), 7.48 (1H, neurite outgrowth a 48 well tissue culture plate, the well was
unsymmet. d, J 8.5, Ar), 2.66 (3H, s, CH3), 1.75–1.72 (4H, m, 2 × coated with a solution of 10 μg mL−1 laminin (Sigma) and
CH2), 1.35 (6H, s, 2 × CH3), 1.33 (6H, s, 2 × CH3); δC (176 MHz, poly-D-lysine (Sigma) overnight. The tissue culture plastic was
DMSO-d6) 167.8, 162.1, 153.0, 145.3, 145.0, 137.6, 132.5, 131.2, washed 3 times with sterile PBS prior to the addition of the
130.6, 127.5, 127.2, 126.8, 126.7, 35.5, 35.5, 35.0, 34.8, 32.6, 32.5, neurospheres. Neurospheres were added to the permissive sub-
31.7, 13.6; m/z (ESI) 404 (M − H); λmax (EtOH) 261 nm (ε 27 000 strate in the presence of the mitotic inhibitors; 10 μM 5-fluoro-
M−1 cm−1), 333 (17 000); Anal. calcd for C25H27NO2S: C, 74.04; 2-deoxyuridine; 10 μM Uridine and 1 μM cytosine-arabinoside
H, 6.71; N, 3.45; found: C, 73.42; H, 6.75; N, 3.53.
in TERA2.cl.SP12 maintenance media described previously
and incubated at 37 °C 5% CO2 for 10 days. Neurospheres
where then fixed in 4% PFA and visualised by immunocyto-
Tissue culture
Stock solutions of ATRA (Sigma), EC23 (Reinnervate), com- chemical staining of TUJ-1 antibody. Neurites were imaged
pounds 1, 2 and 3 were prepared in dimethyl sulfoxide (DMSO, using the Nikon Diaphot 300 and counted for quantification
Sigma) to concentrations of 10 mM. Aliquots of these stock using Image J.
solutions were stored at −80 °C in the dark and thoroughly
defrosted in a water bath set at 37 °C prior to use. Unless
Cell viability assay
otherwise stated all plastic-ware was purchased from Becton, Cells were assessed for viability by combining a single cell sus-
Dickinson and Company. Phase contrast images of cultures pension (achieved by the addition of 1 mL, 0.25% trypsin/
were obtained using a light microscope (Nikon Diaphot 300) EDTA solution) of live cells with detached dead cells contained
and photomicrographs were captured using digital photo- in the culture medium from each flask. Cell numbers were
graphy (Nikon).
determined using a haemocytometer. Cells were required to be
Human pluripotent TERA2.cl.SP12 embryonal carcinoma diluted 1 : 20 in a staining solution for the Viacount assay and
stem cells were cultured in DMEM (Sigma) supplemented with added to a 96-well plate. The number of viable cells, percen-
10% FCS (Lonza), 2 mM L-glutamine (Lonza) and 100 active tage viabilities and total cell numbers were recorded in tripli-
units each of penicillin and streptomycin (Lonza). Cells were cate for each experimental condition.
maintained in a humidified atmosphere of 5% CO2 in air at
Flow cytometric analysis of pluripotent TERA2.cl.SP12 EC cells
37 °C in a Sanyo CO2 incubator and handled under sterile con-
ditions in a Class 1 microbiological safety cabinet. Cultures At each time point (except for the negative control flasks,
were passaged using sterile acid-washed glass beads (VWR) or which were processed for analysis after 3 d), cultures were
trypsinised using a solution of 0.25% (w/v) trypsin (Life Tech- washed with PBS and treated with trypsin to obtain a single-
nologies) and 2 mM EDTA in PBS to obtain a single-cell sus- cell suspension as described above. Cells were washed three
pension for counting. 96-well plates were used for cell viability times with PBS to ensure complete removal of trypsin. After
studies and cultures intended for flow cytometric analyses centrifugation of the cell mixture and removal of the super-
were set up in T25 flasks.
natant, the cell pellet was re-suspended in 0.1% bovine serum
Human neural progenitor ReNcell 197VM cells (Millipore) albumin (BSA) solution and counted using a haemocytometer.
were maintained under the laboratory conditions described 200 000 cells per well were added to a 96-well plate according
above. Before establishing cell cultures, a 20 μg mL−1 concen- to the experimental plan. After centrifugation of the plate at
tration of laminin solution was applied to all plastic-ware, 1000 rpm for 3 min and removal of the supernatant, cells were
incubated at 37 °C for 6 h and then rinsed once with culture re-suspended in 50 μL of the appropriate primary antibody.
medium. Cells were maintained in serum-free conditions with Each triplicate set of culture conditions (ATRA, compounds 1,
DMEM : F12 (1 : 1, Gibco) supplemented with B27 (Invitrogen), 2 and 3, as well as the DMSO-supplemented flasks [on day 3
2 mM L-glutamine, gentamycin (Gibco) and 50 mg mL−1 only]) was analysed at each time point for expression of the fol-
heparin solution (Sigma). For proliferation 10 ng mL−1 fibro- lowing three antigens: SSEA-3 (antibody diluted 1 : 5 in 0.1%
blast growth factor (FGF) and 20 ng mL−1 epidermal growth BSA), TRA-1-60 (diluted 1 : 10) and A2B5 (diluted 1 : 40). These
factor (EGF) were added to the culture media before applying primary monoclonal antibodies were used as they recognise
to cells. Cells were trypsinised as described above.
specific cell surface antigens associated with globoseries glyco-
lipids, glycoproteins and ganglioseries, displaying highly regu-
lated expression profiles in response to differentiation of
Neurite outgrowth assay
TERA2.cl.SP12 EC cells were maintained as described above. human EC cells.27 One well containing undifferentiated
Aggregates of TERA2.cl.SP12 EC cells were produced by control cells was incubated with the mouse myeloma marker
2332 | Org. Biomol. Chem., 2013, 11, 2323–2334
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
anti-P3X as a negative control. The plate was incubated on ice
for 1 h after which excess unbound antibody was removed by
addition of 100 μL of ice-cold 0.1% BSA, centrifugation as
above and removal of the supernatant. Two further washings
were carried out with 180 μl of 0.1% BSA, after which cells
were re-suspended with the fluorescent secondary antibody
FITC (fluorescein isothiocyanate-conjugated) goat anti-mouse
IgM (ICN Pharmaceuticals, Inc.; Aurora, OH; http://www.
icnpharm.com) diluted 1 : 100 in 0.1% BSA. After incubating
the plate in the dark for 1 h on ice, cells were washed three
times with 0.1% BSA as described above, and re-suspended in
200 μL of 0.1% BSA for analysis.
Cell surface antigen expression on TERA2.cl.SP12 stem cells
and their compound-induced derivatives was observed by
indirect immunofluorescence and quantified using a flow cyto-
meter (Guava Easycyte). A fluorescence threshold was set such
that cells fluorescing with a greater intensity than approxi-
mately 95% of the cells incubated with the negative control
antibody P3X were counted as antigen positive.
2 A. Murayama, T. Suzuki and M. Matsui, J. Nutr. Sci. Vitami-
nol., 1997, 43, 167–176.
3 J. H. Barnard, J. C. Collings, A. Whiting, S. A. Przyborski
and T. B. Marder, Chem.–Eur. J., 2009, 15, 11430–11442.
4 (a) V. B. Christie, J. H. Barnard, A. S. Batsanov,
C. E. Bridgens, E. B. Cartmell, J. C. Collings, D. J. Maltman,
C. P. F. Redfern, T. B. Marder, S. A. Przyborski and
A. Whiting, Org. Biomol. Chem., 2008, 6, 3497–3507;
(b) J. B. G. Gluyas, C. Burschka, S. Dörrich, J. Vallet,
H. Gronemeyer and R. Tacke, Org. Biomol. Chem., 2012, 10,
6914–6929; (c) V. B. Christie, D. J. Maltman,
A. P. Henderson, A. Whiting, T. B. Marder, M. Lako and
S. A. Przyborski, J. Neurosci. Meth., 2010, 193, 239–245.
5 M. Robinson-Rechavi, H. E. Garcia and V. Laudet, J. Cell
Sci., 2003, 116, 585–586.
6 J. Bastien and C. Rochette-Egly, Gene, 2004, 328, 1–16.
7 F. Piu, N. K. Gauthier, R. Olsson, E. A. Currier, B. W. Lund,
G. E. Croston, U. Hacksell and M. R. Brann, Biochem. Phar-
macol., 2005, 71, 156–162.
8 P. Germain, S. Kammerer, E. Pérez, C. Peluso-Iltis,
D. Tortolani, F. C. Zusi, J. Starrett, P. Lapointe, J.-P. Daris,
A. Marinier, A. R. de Lera, N. Rochel and H. Gronemeyer,
EMBO Rep., 2004, 5, 877–882.
9 J. Corcoran, B. Shroot, J. Pizzey and M. Maden, J. Cell Sci.,
2000, 113, 2567–2574.
10 (a) L. F. Wong, P. K. Yip, A. Battaglia, J. Grist, J. Corcoran,
M. Maden, M. Azzouz, S. M. Kingsman, A. J. Kingsman,
N. D. Mazarakis and S. B. McMahon, Nat. Neurosci., 2006,
9, 243–250; (b) P. K. Yip, L. F. Wong, D. Pattinson,
A. Battaglia, J. Grist, E. J. Bradbury, M. Maden,
S. B. McMahon and N. D. Mazarakis, Hum. Mol. Genet.,
2006, 15, 3107–3118.
11 P. Germain, P. Chambon, G. Eichele, R. M. Evans, M. A. Lazar,
M. Leid, A. R. de Lera, R. Lotan, D. J. Mangelsdorf and
H. Gronemeyer, Pharmacol. Rev., 2006, 58, 712–725.
12 (a) B. W. Lund, A. E. Knapp, F. Piu, N. K. Gauthier,
M. Begtrup, U. Hacksell and R. Olsson, J. Med. Chem.,
2009, 52, 1540–1545; (b) F. A. J. Kerdesky, J. H. Holms,
J. L. Moore, R. L. Bell, R. D. Dyer, G. W. Carter and
D. W. Brooks, J. Med. Chem., 1991, 34, 2158–2165.
13 (a) K. Umemura, K. Watanabe, K. Ono, M. Yamaura and
J. Yoshimura, Tetrahedron Lett., 1997, 38, 4811–4814;
(b) R. A. Hughes, S. P. Thompson, L. Alcaraz and
C. J. Moody, J. Am. Chem. Soc., 2005, 127, 15644–15651.
14 (a) J. Mulzer, A. Mantoulidis and E. Öhler, Tetrahedron
Lett., 1997, 38, 7725–7728; (b) C. J. Moody and
J. C. A. Hunt, J. Org. Chem., 1999, 64, 8715–8717;
(c) S. Kehraus, G. M. Konig, A. D. Wright and
G. Woerheide, J. Org. Chem., 2002, 67, 4989–4992.
Immunocytochemistry with neuroprogenitor ReN197 VM cells
and human pluripotent TERA2.cl.SP12 EC cells
Cells were fixed in 4% paraformaldehyde (PFA) in PBS (Sigma)
for 30 min at room temperature (RT) and rinsed with PBS. Cell
membranes were permeabilised by treatment with 1% Triton-
X-100 (Sigma) in PBS for 10 min at RT. Nonspecific labelling
was blocked by incubation on a bench-top shaker (Fischer
Scientific) for 1 h at RT with a solution of 1% goat serum
(Sigma) containing 0.2% Tween-20 (Sigma) in PBS. Primary
antibodies were diluted in blocking solution and incubated
with cells for 1 h at RT [β-III tubulin antibody (TUJ1, Covance,
diluted 1 : 600); NF-200 antibody (AbCam, diluted 1 : 200)].
After washing three times for 15 min with PBS cells were incu-
bated for 1 h in the dark with FITC-conjugated (anti-mouse
Alexafluor 488, Invitrogen, diluted 1 : 600), Cy3-conjugated
(anti-rabbit Cy3, JacksonLabs, diluted 1 : 600) fluorescently-
labelled secondary antibodies and Hoechst 33342 nuclear
staining dye (Molecular Probes, diluted 1 : 1000) in blocking
solution. Cells were washed twice more with PBS and left in
the final wash for immediate imaging. Fluorescence micro-
graphs, including Hoecsht 3342, were acquired using the
appropriate filter sets and an adapted digital camera (Nikon).
Acknowledgements
We are grateful to High Force Research Ltd. and the EPSRC for
an Industrial CASE Award (to G.-L. Zhou).
15 (a) M. J. Remuiñán and G. Pattenden, Tetrahedron Lett.,
2000, 41, 7367–7371; (b) D. Romo, N. S. Choi, S. Li,
I. Buchler, Z. Shi and J. O. Liu, J. Am. Chem. Soc., 2004, 126,
10582–10588; (c) M.-Y. Jang, Y. Lin, S. D. Jonghe, L.-J. Gao,
B. Vanderhoydonck, M. Froeyen, J. Rozenski, J. Herman,
T. Louat, K. V. Belle, M. Waer and P. Herdewijn, J. Med.
Chem., 2011, 54, 655–668.
References
1 (a) M. B. Sporn and A. B. Roberts, Mol. Endocrinol., 1991, 5,
3–7; (b) V. B. Christie, T. B. Marder, A. Whiting and
S. A. Przyborski, Mini-Rev. Med. Chem., 2008, 8, 601–608.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2323–2334 | 2333

View Article Online
Paper
Organic & Biomolecular Chemistry
16 F. A. J. Kedersky, J. H. Holms, J. L. Moore, R. L. Bell,
R. D. Dyer, G. W. Carter and D. W. Brooks, J. Med. Chem.,
1991, 34, 2158–2165.
M. Trudeau, Y. Wei and P. S. Charifson, Bioorg. Med. Chem.
Lett., 2010, 20, 2828–2831.
22 S. A. Przyborski, Stem Cells, 2001, 19, 500–504.
17 R. M. Rzasa, M. R. Kaller, G. Lia, E. T. Magal, T. Nguyen, 23 R. Stewart, V. Christie and S. A. Przyborski, Stem Cells,
T. D. Osslund, D. Powers, V. S. Satora, V. N. Viswanadhan, 2003, 21, 248–256.
H.-L. Wang, X. Xiong, W. Zhong and H. M. Norman, Bioorg. 24 R. Donato, E. A. Miljan, S. J. Hines, S. Aouabdi, K. Pollock,
Med. Chem., 2007, 15, 6574–6595.
18 G. C. Barrett, Tetrahedron, 1980, 36, 2023–2058.
S. Patel, F. A. Edwards and J. D. Sinden, BMC Neurosci.,
2007, 8, 36–46.
19 E. Täuscher, D. Weiß, R. Beckert, J. Fabian, 25 D. J. Maltman, V. B. Christie, J. C. Collings, J. H. Barnard,
A. Assumpção and H. Görls, Tetrahedron Lett., 2011, 52,
2292–2294.
S. Fenyk, T. B. Marder, A. Whiting and S. A. Przyborski,
Mol. BioSyst., 2009, 5, 458–471.
20 A. Arcadi, O. A. Attanasi, B. Guidi, E. Rossi and 26 G. Garipova, A. Gautier and S. R. Piettre, Tetrahedron, 2005,
S. Santeusanio, Eur. J. Org. Chem., 1999, 3117–3126. 61, 4755–4759.
21 S. M. Ronkin, M. Badia, S. Bellon, A.-L. Grillot, C. H. Gross, 27 P. W. Andrews, E. Nudelman, S.-I. Hakomori and
T. H. Grossman, N. Mani, J. D. Parsons, D. Stamos,
B. A. Fenderson, Differentiation, 1990, 43, 131–138.
2334 | Org. Biomol. Chem., 2013, 11, 2323–2334
This journal is © The Royal Society of Chemistry 2013