Towards identifying potent new hits for glioblastoma

Article abstract of DOI:10.1039/c8md00436f

Glioblastoma is a devastating disease of the brain and is the most common malignant primary brain tumour in adults. The prognosis for patients is very poor with median time of survival after diagnosis measured in months, due in part to the tumours being highly aggressive and often resistant to chemotherapies. Alongside the ongoing research to identify key factors involved in tumour progression in glioblastoma, medicinal chemistry approaches must also be used in order to rapidly establish new and better treatments for brain tumour patients. Using a computational similarity search of the ZINC database, alongside traditional analogue design by medicinal chemistry intuition to improve the breadth of chemical space under consideration, six new hit compounds (14, 16, 18, 19, 20 and 22) were identified possessing low micromolar activity against both established cell lines (U87MG and U251MG) and patient-derived cell cultures (IN1472, IN1528 and IN1760). Each of these scaffolds provides a new platform for future development of a new therapy in this area, with particular promise shown against glioblastoma subtypes that are resistant to conventional chemotherapeutic agents.

Full text of DOI:10.1039/c8md00436f

MedChemComm
View Article Online
View Journal | View Issue
RESEARCH ARTICLE
Towards identifying potent new hits for
glioblastoma†
Cite this: Med. Chem. Commun.,
018, 9, 1850
2
b
a
b
c
e
Chris Sherer,‡ Saurabh Prabhu, David Adams, Joseph Hayes, Farzana Rowther,
d
e
b
Ibrahim Tolaymat,
Tracy Warr and Timothy J. Snape
*
Glioblastoma is a devastating disease of the brain and is the most common malignant primary brain tumour
in adults. The prognosis for patients is very poor with median time of survival after diagnosis measured in
months, due in part to the tumours being highly aggressive and often resistant to chemotherapies. Along-
side the ongoing research to identify key factors involved in tumour progression in glioblastoma, medicinal
chemistry approaches must also be used in order to rapidly establish new and better treatments for brain
tumour patients. Using a computational similarity search of the ZINC database, alongside traditional ana-
logue design by medicinal chemistry intuition to improve the breadth of chemical space under consider-
ation, six new hit compounds (14, 16, 18, 19, 20 and 22) were identified possessing low micromolar activity
against both established cell lines (U87MG and U251MG) and patient-derived cell cultures (IN1472, IN1528
and IN1760). Each of these scaffolds provides a new platform for future development of a new therapy in
this area, with particular promise shown against glioblastoma subtypes that are resistant to conventional
chemotherapeutic agents.
Received 31st August 2018,
Accepted 1st October 2018
DOI: 10.1039/c8md00436f
rsc.li/medchemcomm
highlight that crucial investment is needed in the search for
improved treatments, not only to improve survival times per
Introduction
Glioblastomas are the most common form of malignant brain
tumours in adults and account for 12–15% of all primary
intracranial neoplasms, and as aggressive cancers, are often
resistant to treatment. The median survival time is 6 months
overall with only 28% of glioblastoma patients surviving for
more than one year, and only 3% of patients surviving more
se, but to improve survival along with a better quality of life.
Our initial studies in this area demonstrated that certain
2-arylindoles (e.g. 1 and 2, Fig. 1) have anti-glioblastoma ac-
3
tivity. In the case of 1, the mechanism of this activity is con-
sistent with the generation of reactive oxygen species (ROS)
followed by autophagic cell death, which in some cases gave
low micromolar EC values against patient-derived short-
1
than three years. Whilst the prognosis for patients is gener-
5
,4
0
3
ally very poor, the best standard treatment is currently surgi-
cal resection followed by concomitant radiotherapy to the re-
section site and chemotherapy with temozolomide (Fig. 1), in
term cell cultures. The research described herein expands
on our previous programme of work and leads to the discov-
ery of new and possibly improved scaffolds for further devel-
opment into much needed glioblastoma treatments. Bond
dissociation enthalpies (BDEs) were calculated using density
functional theory (DFT), as a widely used parameter to mea-
sure the ease and stability of radical formation for a species,
in an attempt to rationalise the structure–activity relation-
2
the so-called Stupp protocol. Unfortunately, such a demand-
ing treatment regime only leads to a mean survival of 14.6
months and a two year survival of 26.5%, statistics which
a
School of Pharmacy, University of East Anglia, Norwich Research Park, Norwich
5
–7
NR4 7TJ, UK
ships in terms of a ROS mechanism.
b
School of Pharmacy and Biomedical Sciences, University of Central Lancashire,
The innate heterogeneity of glioblastoma is well-
documented with individual tumours harbouring a wide
Preston, Lancashire, PR1 2HE, UK. E-mail: tjsnape@uclan.ac.uk
c
School of Physical Sciences and Computing, University of Central Lancashire,
8
,9
spectrum of different genetic abnormalities. This molecu-
lar diversity accounts for the differential efficacy of cytotoxic
agents in glioblastoma patients; for example, mutation of the
TP53 tumour suppressor gene and hypermethylation the DNA
repair gene MGMT, are associated with increased sensitivity
Preston, Lancashire, PR1 2HE, UK
d
Faculty of Medical Science, Anglia Ruskin University, Bishop Hall Lane,
Chelmsford, Essex, CM1 1SQ, UK
e
Brain Tumour Research Centre, University of Wolverhampton, Wulfruna Street,
Wolverhampton, WV1 1LY, UK
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c8md00436f
Present address: Syntor Fine Chemicals Ltd., Runcorn, Cheshire, WA7 1SR,
UK.
1
0–12
to drugs such as CCNU, temozolomide and vincristine.
In the present study, the anti-tumour activity of com-
pounds were assessed in glioblastoma cell cultures
‡
5
1850 | Med. Chem. Commun., 2018, 9, 1850–1861
This journal is © The Royal Society of Chemistry 2018

View Article Online
MedChemComm
Research Article
Fig. 1 Structure of the first-line treatment temozolomide for malignant brain tumours of the type glioblastoma multiforme (GBM), and two 2-aryl
3
indoles from our previous work on anti-glioblastoma lead compounds. The work described herein is depicted by the box entitled ‘This work’
showing the features of 1 that were studied and some selected EC50 values.
comprising 2 established cell lines (U87MG and U251MG)
and patient-derived short-term cell cultures (IN1472,
evidence that certain simple phenols do have modest cyto-
toxic activity, and increasing the complexity of the phenols ap-
3
1
9,20
IN1528 and IN1760). U87MG and IN1760 were TP53 wild-type
whilst the remaining 3 cultures harboured various TP53 mu-
tations. Methylation and transcriptional silencing of MGMT
pears to significantly improve the activity.
One of the rea-
sons for the cytotoxic activity of phenolic compounds lies in
2
1
their ability to generate ROS, which in turn, relies on their
ability to lose both a proton and an electron (formally a hydro-
gen radical) from the hydroxyl group, subsequently forming
both reactive chain-propagating radicals, and a phenoxyl radi-
1
3
was present in 3 cultures (U87MG, U251MG and IN1760).
In previous studies, we have determined the differential re-
sponses of these cultures to a number of chemotherapeutic
compounds which act through diverse mechanisms. In all
cases, IN1760 and IN1472 are the most resistant and sensi-
tive cultures respectively, with the remaining 3 cultures dem-
2
2
cal. Gas phase bond dissociation enthalpies (BDEs) can act
as excellent primary indicators of free radical scavenging ac-
tivity and it has been suggested that trends in BDEs are only
1
4
5,23
onstrating an intermediate response.
weakly solvent dependent,
where DFT calculations have
Fig. 1 outlines our approach towards the investigation of
structure–activity relationships between this known class of
compound (1) and its anti-glioblastoma activity. Due to the
inherently non-exhaustive search of chemical space in our ap-
proach, one would not necessarily expect to find a compound
with exceptional activity but one may expect to find an un-
common fragment, or a fragment with development poten-
tial, with reasonable activity for further development. This
further development could then take the shape of a more tra-
ditional structure–activity relationship based approach, thus
allowing the investigation of the more local chemical space
been successfully used to compute O–H BDEs for phenolic
5
,24,25
compounds.
Calculated BDEs provide a useful predictive
measure of the relative ease of ROS formation. We calculated
−
1
a gas phase BDE of 87.7 kcal mol for phenol at the B3P86/6-
311G**+ level of theory, which is consistent with previous
reported calculations of this type when compared with experi-
2
6
ment.
As comparisons between predicted relative BDE
values are more accurate within structurally similar series of
compounds, we follow suit in our analysis herein.
Based on such precedent, and our previous results
3
confirming the involvement of ROS, lead compound 1 can
1
5
of the active drug fragment.
be considered to be an ortho-substituted phenol with the in-
dole ring acting as a conjugated system to increase radical
stabilisation once formed. With this in mind, a small series
of simple, commercially available phenols were screened
Results and discussion
Exploring the 2-phenylindole core of lead compound 1
(
Fig. 2) to see if mesomeric (M) and/or inductive (I) effects al-
While there are many examples of complex phenols, polyphe-
nols and hydroquinones having reasonable anti-proliferative
one could act in the same way as the indole in terms of af-
fecting activities.
The activities of this series of simple phenols 3–10 are
shown in Table 1 alongside 1 for comparison. Reviewing the
1
6–18
activity,
reports on the cytotoxic activities of simple phe-
nols are relatively few in number. Nevertheless, there is some
This journal is © The Royal Society of Chemistry 2018
Med. Chem. Commun., 2018, 9, 1850–1861 | 1851

View Article Online
Research Article
MedChemComm
Fig. 2 The series of phenols tested in order to establish the importance of the indole group. 10 is assumed to hydrolyse to the phenol in the cell.
M and I represent mesomeric and inductive effects, respectively.
Table 1 Comparison of EC50 values (μM) between a series of simple phenols (the values in parentheses represent the 95% confidence interval level,
95% CI), together with calculated O–H BDEs at the B3P86/6-311+G** level of theory
−
1
Compound
U87MG
U251MG
IN1472
IN1528
IN1760
BDE (kcal mol )
a
1
3
4
5
6
7
8
9
126 (101, 154)
834 (717, 968)
760 (656, 881)
888 (759, 1034)
n.d.
130 (112, 215)
1229 (966, 1710)
1205 (889, 1967)
1251 (994, 1630)
17 (15, 20)
>625
99 (85, 191)
747 (543, 1203)
793 (683, 821)
890 (699, 1136)
n.d.
185 (87, 630)
77.1
87.7
91.6
101.1
79.5
82.1
86.8
85.5
a
1731 (1261, 2990)
1389 (1156, 1627)
n.d.
>4000a
>4000a
>4000
n.d.
n.d.
n.d.
n.d.
57 (46, 71)
n.d.
n.d.
a
492 (436, 555)
664 (537, 844)
723 (539, 1020)
1141 (993, 1341)
1565 (1172, 3146)
1341 (988, 2202)
2021 (1296, 5822)
1436 (1255, 1642)
1413 (1221, 1728)
813 (640, 1084)
744 (682, 812)
787 (648, 983)
>4000a
>4000a
>4000
b
10
94.0
a
b
EC50 not reached. BDE for salicylic acid. EC50 values were calculated using a sulphorhodamine B (SRB) assay and represent anti-
proliferative activity. n.d. = not determined in this batch. The best EC50 values are highlighted in bold.
activities of the individual phenols against the individual cell
cultures (U87MG, U251MG, IN1472, IN1528 and IN1760) gen-
erally shows that there isn't much variation in activity, in a
broader sense, with only modest activity observed, with the
exception of the two hydroquinones (6 and 7) against the
U251MG cell lines (EC50s = 17 and 57 μM respectively).
Whilst this data suggests that no clear conclusion can be
drawn for the relative activities of these phenols and hydro-
quinones against GBM, hydroquinones are known to be in-
than the phenols against the other four cultures. The magni-
tude of this difference in activity suggests that, in this context
at least, either the ortho-indole group is an excellent substitu-
ent for the phenol group, or that there is a unique and per-
haps synergistic effect between the indole and phenol moie-
ties. Although there doesn't appear to be a strong correlation
between BDE and EC50 (Table 1), the three most active com-
pounds (1, 6 and 7) do have the lowest BDE values of all
those tested. The almost complete lack of activity in the
IN1760 culture is representative of its extremely resistant na-
ture to chemotherapeutic drugs.
2
7,28
volved in redox cycling in other systems,
and so such
compounds warrant further study in the area of brain tumour
research. Especially, since it is known that such compounds
can be toxic if not-targeted to the tumour site directly, and so
further study would focus on finding analogues that could
demonstrate selectivity over non-cancerous cells. On the con-
trary, with the exception this time of U251MG, in general 1
seems to have significantly higher anti-proliferative activity
The importance of the phenol group in 1
In order to determine the significance of the phenol group in
1, a series of closely related analogues (Fig. 3) were developed
(the synthesis of all compounds prepared can be found in
1852 | Med. Chem. Commun., 2018, 9, 1850–1861
This journal is © The Royal Society of Chemistry 2018

View Article Online
MedChemComm
Research Article
Fig. 3 Analogues used to probe the importance of the phenol group.
the ESI†). These analogues included a compound without the
phenol group (indole), an aniline analogue (11), a compound
without the phenolic –OH group (2-phenylindole, 12), and a
methoxy derivative, 13.
pository or database. The choice of database is an important
one and can strongly influence the results of a similarity
search. Accordingly, in order to discover otherwise non-
obvious structural analogues of our lead compound/fragment
(1), a similarity search was carried out.
2
9
This is an interesting data set since the response is
broadly similar across cultures, implying that the cytotoxic ef-
fects, or lack of, are independent of genetic background. Im-
portantly, upon removal of either the hydroxyl group (ana-
logue 12), the whole phenolic substituent (indole) or
masking it (13), the EC50 value rises to above what can be de-
termined under the assay conditions employed (Table 2). Al-
though the extent of activity loss cannot be determined abso-
lutely, it can be confidently asserted that the phenol is
essential for activity, yet must be sufficiently substituted to
exert its effect (see activity of phenol in Table 1).
On balance, the most suitable database considered for this
3
0
work was the ZINC database. It is of particular interest
since the data is curated into subsets, which can, by judi-
cious selection of the subsets, improve the efficiency of the
search process. In this work, the subset of data that has been
opted for was the fragment-like subset of compounds (molec-
ular weight ≤250, log P ≤ 3.5, number of rotatable bonds ≤5)
which were advertised as currently in stock, a subset that in-
cluded almost 800 000 entries. Similarity screening was
conducted on this database using ShaEP (a software tool for
a rigid-body superimposition and similarity evaluation of
When comparing the activity of 1 with its aniline analogue
3
1
(11), across all four cell cultures that they were compared
ligand-sized molecules) but to account for the fact that
ShaEP handles structures as rigid bodies, up to three low en-
ergy conformations of the compounds were pre-generated
against, 1 has consistently higher activity (with the exception
of IN1472), although both compounds have only modest ac-
tivity, possibly attributed to the fact that only these two com-
pounds have a calculated BDE. Under normal practises, this
result would encourage similar analogues of 1 and 11 to be
studied, including heteroatom analogues such as indolylthio-
phenols or indolylphenylphosphines, however, such an in-
depth approach was not the focus of the study here.
3
2
using BALLOON, producing a total of over 1.4 million struc-
tures for interrogation. From preliminary computational cal-
culations, what a medicinal chemist might call very similar
from a chemistry perspective (e.g. heteroatom analogues or
positional isomers) ShaEP would consider to have a similarity
of around 0.8. The 105 compounds with the highest overall
similarity value were taken forward as a manageable number
of hits to be manually inspected in the next step. The struc-
tures of these 105 candidate compounds can be seen in the
ESI.† Upon inspection it became apparent that these 105
compounds could be grouped based on our lead compound
(1) and its accepted mechanism of action being related to the
A similarity searching approach
One of the key steps in the early stage of drug development is
finding active hits which can serve as molecules that will be
further optimised into potential drug candidates. In order to
identify such alternative scaffolds that could potentially act
through a similar mechanism, we carried out a computa-
tional similarity study, whereby a large database could be
interrogated, thus expanding the area of chemical space be-
ing explored beyond that which may be familiar to a medici-
nal chemists, to those which are documented in a curated re-
3
formation of a phenol radical for ROS generation.
The first and largest group (59 entries) comprised of com-
pounds that had no way of generating a phenolic (or analo-
gous) radical, for example, compounds that had no phenol or
aniline group. Since these compound don't agree with the cur-
3
rent evidence on how our lead compound acts, they were not
Table 2 Comparison of EC50 values (μM) between 1 and analogues probing the significance of the phenol group (95% CI), together with calculated
O–H BDEs at the B3P86/6-311+G** level of theory
−
1
Compound
U87MG
U251MG
IN1472
IN1528
IN1760
BDE (kcal mol )
a
1
126 (101, 154)
>625
130 (112, 215)
>625a
99 (85, 191)
>625
185 (87, 630)
>625
n.d. a
>625
n.d.
77.1
No BDE
84.8
No BDE
No BDE
a
a
a
a
Indole
>625
>625
b
11
12
13
257 (82, 525)
328
155 (82, 828)
153 (59, 256)
a
a
a
a
>625
>625
>625
>625
a
>625
n.d.
n.d.
n.d.
a
b
EC50 not reached. CI not calculable. The best EC50 values are highlighted in bold. n.d. = not determined in this batch.
This journal is © The Royal Society of Chemistry 2018
Med. Chem. Commun., 2018, 9, 1850–1861 | 1853

View Article Online
Research Article
MedChemComm
Fig. 4 The analogues tested with different scaffolds to that of 1. EC50s can be seen in Table 3.
included in the work beyond this point. Excluding this class
of compounds left a further 46 compounds to be considered.
The next group of compounds were 20 isomers or pseudo-
isomers of 1. Unsurprisingly, compounds in this group typi-
cally had the highest similarity score (0.79–0.88). However, this
work was being done with the intention of encountering ana-
logues which would otherwise not have been considered, by
investigating a wider range of chemical space as possible, and
so compounds in this group were also not considered further.
The third group of compounds (26 entries, some of which
are shown in Fig. 4) are those which may be able to form a con-
The analogues of compound 1 identified were screened
for biological activity are shown in Fig. 4. The biological data
for 14–23 can be seen in Table 3.
Benzothiazole 15 precipitated out of solution at concentra-
tions greater than 1250 μM and so its EC50 could not be calcu-
lated. Whereas, comparing benzoxazole 14 and compound 1
reveals that overall, 1 has higher activity, with the exception of
the U87MG cell line. This seems to indicate that the
2-phenylindole core is better than the intuitively highly similar
benzoxazole and benzothiazole cores as a molecular scaffold
for anti-glioblastoma compounds, at least in these cultures.
Comparing 1 against its naphthalene analogue (16) shows
that the naphthalene analogue has higher activity against
U87MG and U251MG cell cultures, with 1 only having signifi-
cantly higher activity against IN1760, indicating that the
naphthalene analogue appears to be a very active compound
(the similarity between 1 and 16 being 0.81). This is signifi-
cant, as it indicates that the weakly hydrogen bonding NH
group of 1 is not essential for activity (a theory supported by
the activity observed with 14 against U87MG cultures), and
that the role of the scaffold may be limited to such effects as
radical stabilisation, sterics and/or π-stacking.
jugated radical (like 1), but have
a core other than
2
-phenylindole (with the exception of 19 which is included here
for comparison). Such a scaffold-hopping approach has been
widely applied by medicinal chemists to discover equipotent
compounds with novel backbones that have improved proper-
15
ties, and using this approach revealed compounds with the
potential for a phenolic-like radical to be formed.
Within this third group, a total of 14 different cores other
than the 2-phenylindole core of lead compound 1 were identi-
fied (see ESI† for structures), with the 2-phenylbenzimidazole
core occurring far more than any other.
Table 3 Comparison of EC50 values (μM) between compound 1 and a series of analogues with different scaffolds as shown in Fig. 4 (95% CI), together
with calculated O–H BDEs at the B3P86/6-311+G** level of theory
−
1
Compound
U87MG
U251MG
IN1472
IN1528
IN1760
BDE (kcal mol )
b
1
126 (101, 154)
130 (112, 215)
>625
n.d.
99 (85, 191)
185 (87, 630)
77.1
98.0
92.5
86.6
87.0
85.1
75.6
82.9
90.0
83.2
85.9
b
14
15
16
17
18
19
20
21
22
23
52 (34, 82)
2257 (1521, 7815)
n.d.
>1250
>2500a
a
a
a
a
>1250
>1250
>1250
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
>1250
>1000
b
48 (37, 62)
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
92 (83, 101)
372 (290, 412)
102 (76, 126)
21 (18, 24)
95 (81, 113)
916 (619, 1925)
23 (19, 27)
104 (87, 123)
361 (130, 818)
85 (66, 111)
29 (25, 35)
120 (38, 429)
744 (459, 1642)
25 (19, 31)
236 (154, 377)
81 (70, 130)
32 (28, 36)
231 (120, 1002)
c
1079
36
c
304 (254, 433)
193 (87, 848)
371 (216, 858)
a
b
c
EC50 not reached as the compound precipitated out of solution >1250 μM. EC50 not reached. CI not calculable. n.d. = not determined in
this batch. The best EC50 values are highlighted in bold.
1854 | Med. Chem. Commun., 2018, 9, 1850–1861
This journal is © The Royal Society of Chemistry 2018

View Article Online
MedChemComm
Research Article
Comparing 1 against 17 on the three cell cultures studied
shows that 1 has superior activity against all. Again, this ef-
fect provides evidence that the ortho-indole group appears to
be an excellent substituent for phenol, as indicated by the
data in Table 1.
Interestingly, comparing the activities of 16 and 17 sug-
gests that the extended naphthalene core is beneficial for ac-
tivity. Similarly, the para-phenyl phenol analogue (18), high-
lights possible areas for further study in a more detailed
approach to SAR.
pear to be able to support a radical and act through a ROS
mechanism (supported by their high activities, Table 3).
Considering all of this data together, it would seem that the
activity of 1 is not highly sensitive to three-dimensional shape,
as seen with the comparison of 1 and 20. Furthermore, activity
does not seem particularly sensitive to the bicyclic ring system
(see compounds 14, 16, 18, 20 and 22) either; features which
we have identified and which has delivered numerous com-
pounds with improved activity compared to 1. However, per-
haps of even greater importance is that new hit compounds
have been identified that are very active against the extremely
resistant IN1760 culture (compounds 19 and 22).
Furthermore, compound 23, found during the similarity
search mentioned above, is both chemically and structurally
dissimilar to 1, yet is considered by the ShaEP software to be
highly similar in terms of both shape and electronics of the
system, with an overall similarity score of 0.785.
However, compound 23 has only modest activity against
the cultures studied. Interestingly, however, 23 is much more
potent than the related simple phenol guaiacol (8) above
(EC50 values for U251MG = 1141 μM; IN1528 = 813 μM; and
IN1760 = >4000 μM).
In general, again there is no strong correlation between
activity and BDE, but the best two compounds identified
from the similarity searching approach (19 and 22) have two
of the lowest BDE values of the new compounds.
Compounds 1 and 20 were chosen for comparison in the
hope that these isomers would give information about the
importance of shape. As previously revealed, 1 behaves as a
ROS generator, the effects of which can be attenuated by the
antioxidant ascorbic acid, a feature that is directly linked to
3
the presence of the hydroxyl group. In this respect, the
electronics of 1 and 20 are expected to be similar, as they
have the same number of canonical forms to stabilise the
proposed hydroxyl radical (Fig. 5). In the event, the com-
pounds were found to have very similar activities against all
three cell cultures on which they were compared, with 1 be-
ing slightly more active against the IN1528 and IN1760 cul-
tures, and 20 being more active against the U251MG culture.
Surprisingly, indole 19, which contains a +I methyl group
para to the hydroxyl, and thus is suitably positioned to stabi-
lise a radical at that position, is extremely active against the
short-term cell cultures. This particular indole is known to
scavenge free radicals in vitro, albeit at lower concentrations
than used here, and so presumably this increased activity
when used at higher concentrations is the result of a more
Reducing the conjugation of compound 1
In order to investigate the necessity of conjugation between
the proposed phenol radical and the indole motif for radical
stabilisation, an analogue with reduced conjugation between
the two ring systems should be less active if a fully conjugated
system is required (24, Fig. 5). The BDE of 24 was calculated
3
3
toxic, longer-lived radical in the cell.
Interestingly, compounds 21 and 22 show drastically differ-
ent activities. Compound 21 has significantly reduced activity
against all three cell cultures, presumably the result of a reduc-
tion in radical stabilisation with the non-conjugated
phthalimide ring, whereas 22 has very potent activity with
values of EC50 ≤ 36 μM for all three cell cultures and is thus
more active than its heteroatom analogues 14 and 15, as well
as compound 1. All these comparisons indicate that specific
scaffolds can have activity against different glioblastoma cell
cultures. Even if the observed activity does not occur via the
same ROS mechanism, the process of similarity searching has
successfully allowed us to identify new hit compounds of
higher activity than the lead compound. Structurally speaking,
the six new lead compounds 14, 16, 18, 19, 20 and 22 would ap-
−1
(90.2 kcal mol ) and follows the expected trend compared to
−1
the more conjugated 1 (77.1 kcal mol ). In addition to its re-
duced conjugation, the three-dimensional shape of 24 is some-
what different to that of 1 due to the loss of planarity inherent
2
3
in reducing the carbon–carbon double bond (sp → sp ). Nev-
ertheless, as has been shown by the analogues depicted in
Fig. 4, activity within this phenolic class of compounds is not
highly sensitive to shape, so this issue may be moot.
In the event, the EC50 values for 24 were: 122 μM
(U251MG); 123 μM (IN1528) and 148 μM (IN1760), with 1 hav-
ing comparable activities against the patient-derived cell
Fig. 5 The possible locations of a delocalised phenol radical, identified by an asterisk (*), on 1 and 20 (left); and an analogue of 1 with reduced
conjugation, 24 (right).
This journal is © The Royal Society of Chemistry 2018
Med. Chem. Commun., 2018, 9, 1850–1861 | 1855

View Article Online
Research Article
MedChemComm
sumably, a similar unmasking of the phenol group can occur
with 13, this time with enzymes that are capable of demethyl-
ation, but based on the activity observed with 13 this either
doesn't take place, or is slow.
Conclusion
Five structural features of compound 1 were investigated
(
–
Fig. 1) where it was found that a phenolic (or analogous
NH ) group was required for activity, as removal of this
Fig. 6 O-Protected analogues.
2
group resulted in complete loss of activity. It was also shown
that simple phenols (with the exception of hydroquinones 6
and 7) had much lower activity than 1, suggesting that the in-
dole moiety has significant importance, befitting of its place
as a privileged structure. Investigating different scaffolds
yielded some interesting findings too, identifying compounds
with improved activities to 1, indicating that the NH group of
the indole is not essential for activity, and that other aro-
matic systems can stabilise the phenoxy radical in addition
to indole. Surprisingly, the O-substituted analogues showed
higher activity than expected.
Overall, this approach has identified six new promising
compounds (14, 16, 18, 19, 20 and 22) as superior fragments
on which to base the design of new bioactives for glioblas-
toma, and future work will utilise these compounds as leads.
A significant breadth of chemical space has been explored so
far within this work, however the local chemical space of any
of the active functional analogues of compound 1 has not
been probed to any appreciable depth. Future work will
therefore also focus on refining and developing the structures
of these newly identified compounds in order to improve
their activities and physicochemical properties via an in-
depth SAR approach.
cultures. As
a result, determining the superior anti-
glioblastoma agent is difficult based solely on this data alone.
What can be inferred is that the full conjugation seen in 1,
which was originally considered to be essential for its activity,
is perhaps not the only factor involved in promoting activity.
3
5
The effect of O-protected analogues
In previous work, compound 1 was observed to have a rapid
onset of activity (<1 h). The compounds shown in Fig. 6
were prepared and tested to further understand and mitigate
this rapid onset of activity of 1 by controlling the release of
the active compound through a, presumably enzyme-driven,
hydrolytic process. The activities of the compounds depicted
in Fig. 6 are shown in Table 4.
Interestingly, all three esters have better activity than com-
pound 1 against the U251MG and IN1528 cell cultures, pre-
sumably the result of improved cell penetration with the
more hydrophobic compounds. However, that the com-
pounds result in similar EC50 values eventually suggests that
3
3
3
4
the esters are hydrolysed, liberating compound 1.
To investigate the possibility of the O-protected analogues
of 1 being used as delayed-release prodrugs, a time course as-
say for both 1 and its O-heptanoyl analogue (27) was carried
out on two glioblastoma cell cultures. The results (see ESI†)
confirmed that compound 1 does have a rapid activity
against both cell cultures studied, whereas 27 reduced cell vi-
ability much more gradually. Importantly, both compounds
tended towards the same activity over 24 h. This agrees with
the hypothesis that the heptanoyl group can be cleaved off to
produce compound 1, as the activity of compound 27 is ulti-
mately equal to that of compound 1, yet it takes a
measureable time for the enzymatic cleavage to occur. Pre-
That said, glioblastoma is a very heterogeneous tumour,
therefore one would expect different compounds to have dif-
ferent activities in different biological samples generated
from it. As such, compounds that have similarly high activi-
ties against all (or most) of the cultures tested could in fact
be highlighting structures that facilitate anti-proliferative ac-
tivity via a universal mechanism which is independent of the
genetic profile of the individual tumour. Such universally act-
ing compounds would be much more useful in a clinical set-
ting and would reduce the need for tumour profiling and pa-
tient stratification.
Experimental section
BDE calculations
Table 4 Comparison of EC50 values (μM) between 1 and a series of
O-substituted derivatives (95% CI)
DFT geometry optimisations of compounds and their corre-
sponding radicals were carried out using the B3P86 func-
Compound
U251MG
IN1528
IN1760
3
6,37
tional,
this type,
previously revealed as effective in calculations of
1
130 (112, 215)
39 (30, 47)
31 (19, 59)
38 (28, 59)
99 (85, 191)
185
n.d.
n.d.
n.d.
2
4,38
a
in conjunction with the 6-311+G(d,p) basis
25
26
27
38
3
9–42
80 (51, 96)
29 (8, 37)
set.
The optimised structures were confirmed as real
minima by vibrational frequency analysis (no imaginary fre-
quencies) and zero-point energy (ZPE) corrections were
obtained. The spin restricted open-shell (RODFT) approach
a
CI not calculable. n.d. = not determined in this batch. The best
EC50 values are highlighted in bold.
1856 | Med. Chem. Commun., 2018, 9, 1850–1861
This journal is © The Royal Society of Chemistry 2018

View Article Online
MedChemComm
Research Article
was used for optimisation of the radical species. While re-
stricted open shell calculations incorrectly prevent spin
polarisation, there is no spin contamination as in
tween 40–60 °C. Assignments of NMR spectra was aided with
the use of DEPT-135, and in some cases HSQC and HMBC.
2
unrestricted open shell calculations and the correct 〈S 〉 for
Compound synthesis
the corresponding wavefunctions are obtained (i.e. 0.75 for
radicals). The total enthalpy of a species X, HIJX), at tempera-
ture T (298.15 K) was calculated from the expression:
4
3
2-(1H-Indol-2-yl)phenol (1) . 3–12, 17, 18, 21–23 and in-
dole were purchased from Sigma-Aldrich, UK.
Compounds 14 and 15 were purchased from TCI
H(X) = ESCF + ZPE + U + RT
(1)
Chemicals, UK.
3
2-(20-Methoxyphenyl)-1H-indole (13)
where ESCF is the electronic self-consistent field energy, ZPE
is the zero-point energy, and U the combined translational,
rotational and vibrational contributions. RT represents the
PV-work term and is added to convert the energy to the en-
thalpy. Vibrational frequencies were used unscaled to obtain
the zero-point vibrational energies (ZPEs) and the vibrational
contributions to the enthalpy. Using the total enthalpies,
BDE values were determined as follows:
2-(Naphthelen-7-yl) phenol (16). To a flask containing water
(5 mL), was added 2-bromonaphthalene (331 mg, 1.6 mmol),
2-hydroxyphenylboronic acid (500 mg, 2.4 mmol),
diisopropylamine (DIPA, 0.45 mL, 3.2 mmol) and
palladiumIJII) acetate (12.3 mg, 0.25 mol%). The reaction was
heated to reflux for 3.5 hours, and was allowed to cool to
room temperature before being diluted with brine (40 mL)
and extracted with ethyl acetate (5 × 20 mL). The reaction
was filtered through Celite®, and the filtrate dried (MgSO4)
and filtered. The crude reaction was purified via flash
column chromatography on silica gel (5 : 1 petroleum ether :
ethyl acetate) to afford the title compound as a pale yellow
BDE = H(RO˙) + H(H˙) − H(ROH)
(2)
All the DFT calculations were performed Jaguar v9.0
Schrodinger LLC, New York, NY, 2017).
(
solid (288 mg, 82% yield). M.p. 93–96 °C; R = 0.37 (5 : 1
f
1
petroleum ether : ethyl acetate); H NMR (300 MHz, CDCl
3
):
δH = 8.03–7.97 (2H, m, Ar), 7.96–7.88 (2H, m, Ar), 7.66–7.54
(3H, m, Ar), 7.43–7.30 (2H, m, Ar), 7.13 (2H, m, Ar), 5.42 (1H,
Generic information
1
3
s, OH); C NMR (75 MHz, CDCl
3 C q
): δ = 152.7 (Ar, C ), 134.6
Reactions were followed by analytical thin layer chromatogra-
phy (TLC) using plastic-backed TLC plates coated in silica G/
UV254, run in a variety of solvent systems and visualised with
a UV light at 254 nm, p-anisaldehyde stain and/or potassium
permanganate stain. Commercially available solvents and re-
agents were purchased from Fisher, Sigma Aldrich, TCI and
Fluorochem and were used without further purification un-
less specified in the syntheses. Flash column chromatogra-
phy was carried out on Davisil silica 60 Å (40–63 μm) under
bellows pressure. High resolution mass spectra were obtained
at the EPSRC UK National Mass Spectrometry Facility in
Swansea University's College of Medicine using a LTQ
Orbitrap XL™ Hybrid Ion Trap-Orbitrap Mass Spectrometer
coupled to a TriVersa NanoMate® ESI source. Low resolution
mass spectra were obtained on a Thermo Finnigan LCQ Ad-
vantage MAX using electrospray ionisation (ESI) or atmo-
(Ar, C ), 133.7 (Ar, C ), 132.8 (Ar, C ), 130.3 (Ar, CH), 129.4
q
q
q
(Ar, CH), 129.2 (Ar, CH), 128.2 (Ar, C ), 128.1 (Ar, CH), 127.9
q
(Ar, CH), 127.9 (Ar, CH), 127.2 (Ar, CH), 126.7 (Ar, CH), 126.5
−
1
(Ar, CH), 121.1 (Ar, CH), 116.0 (Ar, CH); IR (neat, cm ) ν =
3520 (OH stretch), 750 (Aromatic C–C bend); MS (EI): m/z 220
+
+
(M); HRMS found: [M + H] 221.0960, C16
H
12O + H requires
221.0961.
2-(1H-Indol-2-yl)-4-methylphenol (19). Phenylhydrazine (108
mg, 1.00 mmol), 2-hydroxy-5-methylacetophenone (216 mg,
1.44 mmol) and p-toluenesulfonic acid (20 mg, 0.10 mmol)
were heated under microwave irradiation at 200 °C for 20
minutes. The crude reaction was purified via flash column
chromatography on silica gel (5 : 1 petroleum ether : ethyl
acetate) to afford the title compound as a dark solid (68 mg,
1
31% yield). R
f
= 0.09 (5 : 1 petroleum ether : ethyl acetate); H
NMR (300 MHz, CDCl ): δH = 9.28 (1H, br s, NH), 7.65 (1H, d,
3
1
13
spheric pressure chemical ionisation (APCI). H and
C
J = 8.0 Hz, Ar), 7.51 (1H, d, J = 1.5 Hz, Ar), 7.42 (1H, d, J = 8.0
Hz, Ar), 7.20 (1H, td, J = 7.5, 1.0 Hz, Ar), 7.13 (1H, td, J = 7.5,
1.0 Hz, Ar), 7.01 (1H, dd, J = 8.0, 2.0 Hz, Ar), 6.78–6.88 (2H,
NMR were carried out on a Bruker Fourier 300 (300 MHz) or
a Bruker Advance III 400 (400 MHz) with broad band
decoupling, and all chemical shifts (δ) quoted in parts per
1
3
m, Ar), 5.43 (1H, br s, OH), 2.25 (3H, s, CH3); C NMR (100
MHz, CDCl ): δ = 150.1 (C , Ar), 136.6 (C , Ar), 135.3 (C
Ar), 130.9 (C , Ar), 129.6 (CH, Ar), 128.7 (CH, Ar), 128.7 (C ,
million (ppm) relative to the residual solvent peaks of CHCl
δH 7.26, δ 77.16) or d -DMSO (δ 2.50, δ 39.52). J values are
3
3
C
q
q
q
,
(
c
5
H
C
q
q
given in Hertz (Hz). Infrared spectra were recorded on a solid
sample using a Thermo Nicolet IR-200 FT-IR. Melting points
are uncorrected, and were recorded using a Stuart SMP10.
Preparative liquid chromatography was carried out on a
Teledyne Isco CombiFlash® Rf 200. Elemental analysis was
carried out using a Thermo Scientific™ FLASH 2000 CHNS/O
Analyser. Petroleum ether refers to the fraction that boils be-
Ar), 122.3 (CH, Ar), 120.6 (CH, Ar), 120.2 (CH, Ar), 119.0 (Cq,
Ar), 116.6 (CH, Ar), 111.1 (CH, Ar), 100.2 (CH, Ar), 20.7 (CH );
3
+
−
MS (EI): m/z 224 ([M + H] ); HRMS found: [M − H] 222.0924,
+
C H NO − H requires 222.0924.
15
13
2-(1H-Indol-3-yl)phenol (20). 2-Benzyloxyphenylacetic acid
(983 mg, 4.06 mmol) was stirred in methanol (45 mL) and
sulfuric acid (2 drops) at room temperature for 72 hours to
This journal is © The Royal Society of Chemistry 2018
Med. Chem. Commun., 2018, 9, 1850–1861 | 1857

View Article Online
Research Article
MedChemComm
1
yield methyl 2-(2-(benzyloxy)phenyl)acetate. Rf = 0.60 (3 : 1
petroleum ether : ethyl acetate); H NMR (300 MHz, CDCl3):
etate), 0.12 (4 : 1 petroleum ether : ethyl acetate); H NMR
1
(300 MHz, CDCl ): δH = 8.31 (1H, br s, NH), 7.73 (1H, d, J =
3
δ
H
= 7.45–7.29 (5H, m, Ar), 7.29–7.18 (2H, m, Ar), 6.94 (2H,
8 Hz, Ar), 7.50 (1H, dd, J = 7.5, 1.5 Hz, Ar), 7.45–7.31 (3H, m,
Ar), 7.31–7.20 (2H, m, Ar), 7.19–7.08 (2H, m, Ar), 5.69 (1H, br
m, Ar), 5.08 (2H, s, O–C H_ –Ar), 3.70 (2H, s, Ar–C H_ –COOMe),
2
_
2_
1
3
3
.64 (3H, s, CH3).
The crude methyl ester was reduced to the alcohol by stir-
ring in THF (dry, 15 mL) at 0 °C and adding LiAlH (308 mg,
s, OH); C NMR (75 MHz, CDCl ): δ = 153.2 (Ar, C ), 136.3
3
C
q
q
(Ar, C ), 130.9 (Ar, CH), 128.6 (Ar, CH), 126.2 (Ar, C
q
), 123.4
(Ar, CH), 122.9 (Ar, CH), 121.1 (Ar, C ), 120.7 (Ar, CH), 120.5
4
q
8
.11 mmol) under an atmosphere of nitrogen. The ice bath
(Ar, CH), 119.7 (Ar, CH), 115.5 (Ar, CH), 111.7 (Ar, CH); MS
−
+
was then removed, and the reaction was stirred at room tem-
perature for 90 minutes. The reaction was quenched with
NaOH(aq) (10 mL, 0.1 M). The crude product was extracted
with ethyl acetate (3 × 10 mL), and the combined organic
layers were dried over MgSO and filtered before being con-
centrated in vacuo to yield crude 2-(2-(benzyloxy)phenyl)-
(ESI): m/z 208 ([M − H] ); HRMS found: [M + H] 210.0913,
+
C H NO + H requires 210.0913.
14
11
2-(Indolin-2-yl)phenol (24). To a flask containing glacial
acetic acid (10 mL) was added compound 1 (1.0 mmol, 209
mg) and sodium cyanoborohydride (12.0 mmol, 754 mg), and
the reaction was stirred at room temperature for 26 hours.
The crude reaction was then quenched by careful addition of
water (100 mL), before adding solid sodium hydroxide to pH
∼ 12. The crude reaction was extracted with ethyl acetate (3 ×
4
1
ethanol. R
f
= 0.22 (3 : 1 petroleum ether : ethyl acetate);
H
NMR (300 MHz, CDCl ): δH = 7.49–7.32 (5H, m, Ar), 7.28–7.19
3
(2H, m, Ar), 7.00–6.91 (2H, m, Ar), 5.10 (2H, s, Ar–C H_ –OAr),
2
_
3
.87 (2H, t, J = 6.5, C H_ _
2
–OH), 2.98 (2H, t, J = 6.5 Hz, Ar–C H_ _
2
–
4
20 mL), dried over MgSO and filtered. The product was
CH OH), 1.95 (1H, br s, OH).
purified via flash column chromatography on silica gel (19 : 1
2
The crude alcohol was oxidised to the aldehyde by stirring
with pyridinium chlorochromate (1.31 g, 6.08 mmol) in
CH Cl (40 mL) in the presence of crushed molecular sieves
petroleum ether : ethyl acetate) to afford the title compound
as a pale orange solid (92 mg, 43% yield). R
f
= 0.60 (3 : 1
petroleum ether : ethyl acetate), 0.16 (19 : 1 petroleum ether :
2
2
1
(650 mg) at room temperature for 22 hours. The solvent was
3 H
ethyl acetate); H NMR (300 MHz, CDCl ): δ = 9.75 (1H, br s,
removed in vacuo, and the crude reaction was dissolved in
diethyl ether (40 mL), before being filtered through Celite®.
The crude reaction was purified by flash column chromatog-
raphy (19 : 1 petroleum ether : ethyl acetate) to yield 0.70 g of
an 84 : 16 ratio of products, which included 468 mg (51%
yield over 3 steps) of the major product 2-(2-(benzyloxy)-
OH), 7.28–7.10 (3H, m, Ar), 7.06 (1H, d, J = 7.5 Hz, Ar), 6.98–
6.80 (4H, m, Ar), 4.92 (1H, dd, J = 12.5, 8.5 Hz, CH), 4.34 (1H,
1
3
br s, NH), 3.38–3.04 (2H, m, CH
2
); C NMR (75 MHz, CDCl
3
);
δ
C
= 156.88 (C , Ar), 148.7 (C , Ar), 130.8 (C
q
q
q
, Ar), 129.1 (CH,
Ar), 128.5 (CH, Ar), 127.7 (CH, Ar), 124.9 (C , Ar), 124.8 (CH,
q
Ar), 121.5 (CH, Ar), 119.5 (CH, Ar), 117.6 (CH, Ar), 112.0 (CH,
+
phenyl)acetaldehyde. R
etate); H NMR (300 MHz, CDCl ): δ = 9.72 (1H, t, J = 2.0
f
= 0.61 (3 : 1 petroleum ether : ethyl ac-
Ar), 65.6 (CH), 38.4 (CH
2
); MS (ESI): m/z 212 ([M + H] ).
1
2-(1H-Indol-2-yl)phenyl benzoate (25). Compound 1 (0.97
mmol, 202 mg), benzoyl chloride (1.16 mmol, 135 μL),
triethylamine (1.46 mmol, 203 μL) and DMAP (0.1 mmol, 12
mg) was stirred in CH Cl (10 mL) at room temperature for
3
H
Hz, CHO), 7.45–7.24 (6H, m, Ar), 7.20–7.14 (1H, m, Ar), 7.00–
.93 (2H, m, Ar), 5.08 (2H, s, O–CH –Ar), 3.70 (2H, d, J = 2.0
Hz, C H_ –CHO).
6
2
2
_
2
2
The mixture of aldehydes (0.70 g, of which 468 mg, 2.07
mmol was the desired aldehyde) was added to a microwave
vial with phenyl hydrazine (336 mg, 3.11 mmol) and
p-toluenesulfonic acid (40 mg, 0.21 mmol) and heated at 200
18 hours. Upon completion by TLC, the crude reaction was
washed with 1 M HCl (2 × 10 mL), water (10 mL) and brine
(10 mL). The crude reaction was purified via preparative
liquid chromatography on silica gel (19 : 1 to 9 : 1 petroleum
ether : ethyl acetate) to afford the title product as a yellow
°C for 20 minutes. The crude product was purified by flash
column chromatography on silica gel (19 : 1 → 9 : 1 petroleum
solid (141 mg, 47% yield). M.p. 148–168 °C (decomposes); R
= 0.18 (9 : 1 petroleum ether : ethyl acetate); H NMR (300
f
1
ether : ethyl acetate) to yield 3-(2-(benzyloxy)phenyl)-1H-indole
1
(
200 mg, 32% yield). H NMR (300 MHz, CDCl
3
): δ
H
= 8.12
MHz, CDCl
3 H
): δ = 8.60 (1H, s, NH), 8.23 (2H, d, J = 7.5 Hz,
1
3
(
1H, br s, NH), 7.95 (1H, d, J = 7.5 Hz, Ar), 7.79 (1H, dd, J =
.5, 1.5 Hz, Ar), 7.53 (1H, d, J = 2.5 Hz, Ar), 7.42–7.11 (11H,
m, Ar), 5.14 (2H, s, CH ).
Ar), 7.79–7.03 (11H, m, Ar), 6.83 (1H, s, Ar); C NMR (75
7
MHz, CDCl
134.1 (Ar, CH), 133.7 (Ar, C
3
): δ
C
= 165.2 (CO), 147.7 (Ar, C
q
), 136.5 (Ar, C
q
),
),
2
q
), 130.3 (Ar, CH), 129.3 (Ar, C
q
Methanol (10 mL) was added to a two-necked flask
containing 3-(2-(benzyloxy)phenyl)-1H-indole (200 mg, 0.67
mmol) and palladium on carbon (20 mg, 10 wt%). The flask
was evacuated and backfilled with nitrogen twice, before be-
ing evacuated and backfilled with hydrogen twice. The reac-
tion was then stirred at room temperature under an atmo-
sphere of hydrogen for 24 hours, and the crude product was
filtered through Celite® and purified by flash column chro-
matography on silica gel (4 : 1 → 3 : 1 petroleum ether : ethyl
acetate) to yield 2-(1H-indol-3-yl)phenol (119 mg, 85% yield)
129.1 (Ar, CH), 129.0 (Ar, CH), 128.8 (Ar, CH), 128.6 (Ar, Cq),
126.8 (Ar, CH), 125.9 (Ar, C ), 123.8 (Ar, CH), 122.5 (Ar, CH),
120.8 (Ar, CH), 120.1 (Ar, CH), 111.0 (Ar, CH), 102.8 (Ar, CH);
q
−
1
IR (neat, cm ) ν = 1725 (CO stretch), 1260 (C–O stretch);
+
+
MS (ESI): m/z 314 ([M + H] ); HRMS found: [M + H]
314.1184, C21 + H requires 314.1176.
H
15NO
2
2-(1H-Indol-2-yl)phenyl acetate (26). Compound 1 (101.7
mg, 0.59 mmol), acetyl chloride (46 μL, 0.65 mmol),
triethylamine (54 μL) and DMAP (6 mg, 0.05 mmol) were
added to CH Cl (5 mL) and stirred at room temperature for
2
2
as a thick orange oil. R
f
= 0.29 (3 : 1 petroleum ether : ethyl ac-
16 h. The reaction was washed with 1 M HCl (3 × 5 mL),
1858 | Med. Chem. Commun., 2018, 9, 1850–1861
This journal is © The Royal Society of Chemistry 2018

View Article Online
MedChemComm
Research Article
water (1 × 5 mL) and brine (1 × 5 mL), dried (MgSO ), filtered
DMSO [Sigma Aldrich; UK] and cells were treated for 72 hours
with serial dilutions of the test compound. The culture me-
dium was removed and the cells fixed in 10% trichloroacetic
acid [Sigma Aldrich; UK] on ice for 30 min followed by washing
in water and air-drying. Cells were stained with 0.4% sulfo-
rhodamine B [Sigma Aldrich; UK] prepared in 1% acetic acid
for 15-20 min, washed in 1% acetic acid and air-dried. The dye
was solubilized in 100 μL of 10 mM Tris (not buffered) and
read at 560 nm [Dynatech MR5000] for quantification.
4
and the solvent was removed in vacuo. The reaction was
purified by flash column chromatography on silica gel (5 : 1
petroleum ether : ethyl acetate) to yield the title product as an
off-white solid (84.3 mg, 71% yield). M.p. 112–127 °C; R =
f
1
0
.23 (5 : 1 petroleum ether : ethyl acetate); H NMR (300 MHz,
CDCl ): δ = 8.50 (1H, s, NH), 7.70–7.62 (2H, m, Ar), 7.43–
3
H
7
2
.32 (3H, m, Ar), 7.26–7.09 (3H, m, Ar), 6.78–6.82 (1H, m, Ar),
.31 (3H, s, CH ); C NMR (75 MHz, CDCl ): δ = 169.4
3 3 C
1
3
(
(
CO), 147.6 (Ar), 136.6 (Ar), 133.9 (Ar), 129.3 (Ar), 128.9
Ar), 128.7 (Ar), 126.8 (Ar), 125.9 (Ar), 123. 7 (Ar), 122.7 (Ar),
Analysis was performed using Sigmoidal Dose Response
(Variable Slope) (Non-Linear Fit).
1
3
20.9 (Ar), 120.3 (Ar), 111.0 (Ar), 102.7 (Ar), 21.4 (CH ); IR
−
1
(
(
(
neat, cm ) ν = 3357 (N–H stretch), 1735 (CO stretch), 1368 Conflicts of interest
aromatic C–C stretch), 1220 (C–O stretch); MS (ESI): m/z 252
+
+
+
There are no conflicts of interest to declare.
2
[M + H] ); HRMS found: [M + H] 252.1021, C16H13NO + H
requires 252.1019.
-(1H-Indol-2-yl)phenyl heptanoate (27). Compound 1 (0.92
Acknowledgements
2
mmol, 192 mg), heptanoyl chloride (1.10 mmol, 171 μL),
triethylamine (1.38 mmol, 192 μL) and DMAP (0.1 mmol, 12
The authors would like to thank the Sydney Driscoll Neurosci-
ence Foundation and Brain Tumour North West for funding.
Thanks also goes to the EPSRC UK National Mass Spectrome-
try Facility (NMSF), Swansea for accurate mass measurements.
mg) was stirred in CH
8 hours. Upon completion by TLC, the crude reaction was
washed with 1 M HCl (2 × 10 mL), water (10 mL) and brine
10 mL). The crude reaction was purified via preparative
2 2
Cl (10 mL) at room temperature for
1
(
References
liquid chromatography on silica gel (19 : 1 to 9 : 1 petroleum
ether : ethyl acetate) to afford the title product as a yellow
1 A. Brodbelt, D. Greenberg, T. Winters, M. Williams, S.
Vernon, V. P. Collins and N. Natl Canc Information, Eur. J.
Cancer, 2015, 51, 533–542.
solid (225 mg, 76% yield). M.p. 51–54 °C; R
petroleum ether : ethyl acetate); H NMR (300 MHz, CDCl ):
f
= 0.15 (9 : 1
1
3
δH = 8.53 (1H, br s, NH), 7.70–7.61 (2H, m, Ar), 7.41–7.27 (3H,
m, Ar), 7.25–7.10 (3H, m, Ar), 6.81–6.76 (1H, m, Ar), 2.59 (2H,
2 R. Stupp, W. P. Mason, M. J. van den Bent, M. Weller, B.
Fisher, M. J. B. Taphoorn, K. Belanger, A. A. Brandes, C.
Marosi, U. Bogdahn, J. Curschmann, R. C. Janzer, S. K.
Ludwin, T. Gorlia, A. Allgeier, D. Lacombe, J. G. Cairncross,
E. Eisenhauer, R. O. Mirimanoff, D. Van Den Weyngaert, S.
Kaendler, P. Krauseneck, N. Vinolas, S. Villa, R. E. Wurm,
M. H. B. Maillot, F. Spagnolli, G. Kantor, J. P. Malhaire, L.
Renard, O. De Witte, L. Scandolaro, C. J. Vecht, P. Maingon,
J. Lutterbach, A. Kobierska, M. Bolla, R. Souchon, C. Mitine,
T. Tzuk-Shina, A. Kuten, G. Haferkamp, J. de Greve, F. Priou,
J. Menten, I. Rutten, P. Clavere, A. Malmstrom, B. Jancar, E.
Newlands, K. Pigott, A. Twijnstra, O. Chinot, M. Reni, A.
Boiardi, M. Fabbro, M. Campone, J. Bozzino, M. Frenay, J.
Gijtenbeek, A. A. Brandes, J. Y. Delattre, U. Bogdahn, U. De
Paula, M. J. van den Bent, C. Hanzen, G. Pavanato, S.
Schraub, R. Pfeffer, R. Soffietti, M. Weller, R. D. Kortmann,
M. Taphoorn, J. L. Torrecilla, C. Marosi, W. Grisold, P.
Huget, P. Forsyth, D. Fulton, S. Kirby, R. Wong, D. Fenton,
B. Fisher, G. Cairncross, P. Whitlock, K. Belanger, S.
Burdette-Radoux, S. Gertler, S. Saunders, K. Laing, J.
Siddiqui, L. A. Martin, S. Gulavita, J. Perry, W. Mason, B.
Thiessen, H. Pai, Z. Y. Alam, D. Eisenstat, W. Mingrone, S.
Hofer, G. Pesce, J. Curschmann, P. Y. Dietrich, R. Stupp,
R. O. Mirimanoff, P. Thum, B. Baumert, G. Ryan and B.
European Org Res Treatment Canc, N. Engl. J. Med.,
2005, 352, 987–996.
2 2
t, J = 7.5 Hz, CH ), 1.70 (2H, pent, J = 7.5 Hz, CH ), 1.39–1.19
1
3
(
6H, m, 3 × CH ), 0.87 (3H, t, J = 6.5 Hz, CH ); C NMR (75
2
3
MHz, CDCl
3
): δ
C
= 172.4 (CO), 147.6 (Ar, C
q
), 136.5 (Ar, C
q
q
),
),
1
1
1
3
1
34.0 (Ar, C ), 129.4 (Ar, CH), 128.8 (Ar, CH), 128.7 (Ar, C
26.7 (Ar, CH), 125.9 (Ar, C ), 123.7 (Ar, CH), 122.6 (Ar, CH),
20.8 (Ar, CH), 120.2 (Ar, CH), 111.0 (Ar, CH), 102.6 (Ar, CH),
q
q
2 2 2 2 2
4.7 (CH ), 31.5 (CH ), 28.9 (CH ), 24.9 (CH ), 22.5 (CH ),
−
1
4.1 (CH ); IR (neat, cm ) ν = 1145 (C–O stretch), 1746 (CO
3
+
stretch), 2928 (C–H stretch); MS (ESI): m/z 322 ([M + H] );
HRMS found: [M + H] 322.1805, C21
+
2
H23NO + H requires
3
22.1802.
SRB assay procedure
Short-term cell cultures (IN1472, IN1528, IN1760) were pre-
pared from approximately 10 mg of adult GBM biopsy tissue
and maintained in Hams F10 nutrient mix [Invitrogen, Pais-
ley UK] containing 10% foetal calf serum in a 37 °C non-CO
2
4
4
incubator as previously described. Passages of 10 to 14 were
employed for the current study. In addition, we also
employed U251 and U87 which are established GBM cell lines
cultured under similar conditions.
Treated cells were assessed for their capacity to proliferate
following
treatment
with
compounds
using
a
4
5
sulphorhodamine B (SRB) assay. Briefly, 3000 cells were
seeded per well in a 96 well plate and allowed to reach expo-
nential growth (48 hours). Compounds were dissolved in
3 S. Prabhu, Z. Akbar, F. Harris, K. Karakoula, R. Lea, F.
Rowther, T. Warr and T. Snape, Bioorg. Med. Chem.,
2013, 21, 1918–1924.
This journal is © The Royal Society of Chemistry 2018
Med. Chem. Commun., 2018, 9, 1850–1861 | 1859

View Article Online
Research Article
MedChemComm
4
In-house in vitro blood-brain-barrier model (WO2018020274
A1)) showed compound to be highly permeable
compound.
J. S. Wright, E. R. Johnson and G. A. DiLabio, J. Am. Chem.
Soc., 2001, 123, 1173–1183.
M. Leopoldini, T. Marino, N. Russo and M. Toscano, J. Phys.
Chem. A, 2004, 108, 4916–4922.
D. Farmanzadeh and M. Najafi, Struct. Chem., 2015, 26,
Madhavan, K. Buetow, A. Barker, J. Vockley, C. Compton,
J. Vaught, P. Fielding, F. Collins, P. Good, M. Guyer, B.
Ozenberger, J. Peterson and E. Thomson, Nature,
2008, 455, 1061–1068.
(
1
a
5
6
7
8
9 R. G. W. Verhaak, K. A. Hoadley, E. Purdom, V. Wang,
Y. Qi, M. D. Wilkerson, C. R. Miller, L. Ding, T. Golub,
J. P. Mesirov, G. Alexe, M. Lawrence, M. O'Kelly, P.
Tamayo, B. A. Weir, S. Gabriel, W. Winckler, S. Gupta, L.
Jakkula, H. S. Feiler, J. G. Hodgson, C. D. James, J. N.
Sarkaria, C. Brennan, A. Kahn, P. T. Spellman, R. K.
Wilson, T. P. Speed, J. W. Gray, M. Meyerson, G. Getz,
C. M. Perou and D. N. Hayes, Cancer Cell, 2010, 17,
98–110.
10 R. Stupp, M. E. Hegi, W. P. Mason, M. J. van den Bent,
M. J. B. Taphoorn, R. C. Janzer, S. K. Ludwin, A. Allgeier, B.
Fisher, K. Belanger, P. Hau, A. A. Brandes, J. Gijtenbeek, C.
Marosi, C. J. Vecht, K. Mokhtari, P. Wesseling, S. Villa, E.
Eisenhauer, T. Gorlia, M. Weller, D. Lacombe, J. G.
Cairncross and R. O. Mirimanoff, Lancet Oncol., 2009, 10,
459–466.
11 M. D. Blough, D. C. Beauchamp, M. R. Westgate, J. J. Kelly
and J. G. Cairncross, J. Neuro-Oncol., 2011, 102, 1–7.
12 B. Murnyak, M. C. Kouhsari, R. Hershkovitch, B. Kalman, G.
Marko-Varga, A. Klekner and T. Hortobagyi, Oncotarget,
2017, 8, 46348–46362.
13 N. Gaspar, L. Marshall, L. Perryman, D. A. Bax, S. E. Little,
M. Viana-Pereira, S. Y. Sharp, G. Vassal, A. D. J. Pearson,
R. M. Reis, D. Hargrave, P. Workman and C. Jones, Cancer
Res., 2010, 70, 9243–9252.
14 N. Potter, S. Ashmore, K. Karakoula, S. Ward, B. Suarez-
Merino, M. Luxsuwong, D. G. Thomas, J. Darling and T.
Warr, J. Neuro-Oncol., 2010, 12, 67–67.
15 H. M. Sun, G. Tawa and A. Wallqvist, Drug Discovery Today,
2012, 17, 310–324.
16 L. H. Yao, Y. M. Jiang, J. Shi, F. A. Tomas-Barberan, N.
Datta, R. Singanusong and S. S. Chen, Plant Foods Hum.
Nutr., 2004, 59, 113–122.
17 J. Dai and R. J. Mumper, Molecules, 2010, 15, 7313–7352.
18 E. I. Parkinson and P. J. Hergenrother, Acc. Chem. Res.,
2015, 48, 2715–2723.
8
31–844.
L. Chin, M. Meyerson, K. Aldape, D. Bigner, T. Mikkelsen,
S. VandenBerg, A. Kahn, R. Penny, M. L. Ferguson, D. S.
Gerhard, G. Getz, C. Brennan, B. S. Taylor, W. Winckler,
P. Park, M. Ladanyi, K. A. Hoadley, R. G. W. Verhaak,
D. N. Hayes, P. T. Spellman, D. Absher, B. A. Weir, L.
Ding, D. Wheeler, M. S. Lawrence, K. Cibulskis, E.
Mardis, J. H. Zhang, R. K. Wilson, L. Donehower, D. A.
Wheeler, E. Purdom, J. Wallis, P. W. Laird, J. G. Herman,
K. E. Schuebel, D. J. Weisenberger, S. B. Baylin, N.
Schultz, J. Yao, R. Wiedemeyer, J. Weinstein, C. Sander,
R. A. Gibbs, J. Gray, R. Kucherlapati, E. S. Lander, R. M.
Myers, C. M. Perou, R. McLendon, A. Friedman, E. G. Van
Meir, D. J. Brat, G. M. Mastrogianakis, J. J. Olson, N.
Lehman, W. K. A. Yung, O. Bogler, M. Berger, M. Prados,
D. Muzny, M. Morgan, S. Scherer, A. Sabo, L. Nazareth, L.
Lewis, O. Hall, Y. M. Zhu, Y. R. Ren, O. Alvi, J. Q. Yao, A.
Hawes, S. Jhangiani, G. Fowler, A. San Lucas, C. Kovar, A.
Cree, H. Dinh, J. Santibanez, V. Joshi, M. L. Gonzalez-
Garay, C. A. Miller, A. Milosavljevic, C. Sougnez, T.
Fennell, S. Mahan, J. Wilkinson, L. Ziaugra, R. Onofrio, T.
Bloom, R. Nicol, K. Ardlie, J. Baldwin, S. Gabriel, R. S.
Fulton, M. D. McLellan, D. E. Larson, X. Q. Shi, R.
Abbott, L. Fulton, K. Chen, D. C. Koboldt, M. C. Wendl,
R. Meyer, Y. Z. Tang, L. Lin, J. R. Osborne, B. H.
Dunford-Shore, T. L. Miner, K. Delehaunty, C. Markovic,
G. Swift, W. Courtney, C. Pohl, S. Abbott, A. Hawkins, S.
Leong, C. Haipek, H. Schmidt, M. Wiechert, T. Vickery, S.
Scott, D. J. Dooling, A. Chinwalla, G. M. Weinstock, M.
O'Kelly, J. Robinson, G. Alexe, R. Beroukhim, S. Carter, D.
Chiang, J. Gould, S. Gupta, J. Korn, C. Mermel, J.
Mesirov, S. Monti, H. Nguyen, M. Parkin, M. Reich, N.
Stransky, L. Garraway, T. Golub, A. Protopopov, I. Perna,
S. Aronson, N. Sathiamoorthy, G. Ren, H. Kim, S. K.
Kong, Y. H. Xiao, I. S. Kohane, J. Seidman, L. Cope, F.
Pan, D. Van Den Berg, L. Van Neste, J. M. Yi, J. Z. Li, A.
Southwick, S. Brady, A. Aggarwal, T. Chung, G. Sherlock,
J. D. Brooks, L. R. Jakkula, A. V. Lapuk, H. Marr, S.
Dorton, Y. G. Choi, J. Han, A. Ray, V. Wang, S. Durinck,
M. Robinson, N. J. Wang, K. Vranizan, V. Peng, E. Van
Name, G. V. Fontenay, J. Ngai, J. G. Conboy, B. Parvin,
H. S. Feiler, T. P. Speed, N. D. Socci, A. Olshen, A. Lash,
B. Reva, Y. Antipin, A. Stukalov, B. Gross, E. Cerami,
W. Q. Wang, L. X. Qin, V. E. Seshan, L. Villafania, M.
Cavatore, L. Borsu, A. Viale, W. Gerald, M. D. Topal, Y.
Qi, S. Balu, Y. Shi, G. Wu, M. Bittner, T. Shelton, E.
Lenkiewicz, S. Morris, D. Beasley, S. Sanders, R. Sfeir, J.
Chen, D. Nassau, L. Feng, E. Hickey, C. Schaefer, S.
19 R. W. Owen, A. Giacosa, W. E. Hull, R. Haubner, B.
Spiegelhalder and H. Bartsch, Eur. J. Cancer, 2000, 36,
1235–1247.
20 C. A. Gomes, T. G. da Cruz, J. L. Andrade, N. Milhazes, F.
Borges and M. P. M. Marques, J. Med. Chem., 2003, 46,
5395–5401.
21 C. C. Winterbourn, Nat. Chem. Biol., 2008, 4, 278–286.
22 V. E. Kagan, A. I. Kuzmenko, A. A. Shvedova, E. R. Kisin, R.
Li, I. Martin, P. J. Quinn, V. A. Tyurin, Y. Y. Tyurina and J. C.
Yalowich, Biochim. Biophys. Acta, 2003, 1620, 72–84.
23 A. J. Javan and M. J. Javan, Food Chem., 2014, 165,
451–459.
24 P. Trouillas, P. Marsal, D. Siri, R. Lazzaroni and J. L. Duroux,
Food Chem., 2006, 97, 679–688.
25 H. Y. Zhang, Y. M. Sung and X. L. Wang, Chem. – Eur. J.,
2003, 9, 502–508.
1860 | Med. Chem. Commun., 2018, 9, 1850–1861
This journal is © The Royal Society of Chemistry 2018

View Article Online
MedChemComm
Research Article
2
2
2
2
3
3
3
3
3
3
6 R. M. B. dos Santos and J. A. M. Simoes, J. Phys. Chem. Ref.
36 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
37 J. P. Perdew, Phys. Rev. B: Condens. Matter Mater. Phys.,
1986, 33, 8822–8824.
38 P. Kosinova, F. Di Meo, E. H. Anouar, J. L. Duroux and
P. Trouillas, Int. J. Quantum Chem., 2011, 111,
1131–1142.
39 T. Clark, J. Chandrasekhar, G. W. Spitznagel and P. V.
Schleyer, J. Comput. Chem., 1983, 4, 294–301.
40 M. J. Frisch, J. A. Pople and J. S. Binkley, J. Chem. Phys.,
1984, 80, 3265–3269.
41 R. Krishnan, J. S. Binkley, R. Seeger and J. A. Pople, J. Chem.
Phys., 1980, 72, 650–654.
42 A. D. McLean and G. S. Chandler, J. Chem. Phys., 1980, 72,
5639–5648.
43 T. J. Snape, Synlett, 2008, 2689–2691.
44 G. M. Lewandowicz, B. Harding, W. Harkness, R. Hayward,
D. G. T. Thomas and J. L. Darling, Eur. J. Cancer, 2000, 36,
1955–1964.
45 W. Voigt, in Methods in Molecular Medicine, 2005, vol. 110,
pp. 39–48.
Data, 1998, 27, 707–739.
7 A. Rybina, B. Thaler, R. Kramer and D. P. Herten, Phys.
Chem. Chem. Phys., 2014, 16, 19550–19555.
8 C. Jiang, S. Garg and T. D. Waite, Environ. Sci. Technol.,
2
015, 49, 14076–14084.
9 R. P. Sheridan and S. K. Kearsley, Drug Discovery Today,
002, 7, 903–911.
0 J. J. Irwin and B. K. Shoichet, J. Chem. Inf. Model., 2005, 45,
77–182.
2
1
1 M. J. Vainio, J. S. Puranen and M. S. Johnson, J. Chem. Inf.
Model., 2009, 49, 492–502.
2 J. S. Puranen, M. J. Vainio and M. S. Johnson, J. Comput.
Chem., 2010, 31, 1722–1732.
3 S. Suzen, P. Bozkaya, T. Coban and D. Nebioglu, J. Enzyme
Inhib. Med. Chem., 2006, 21, 405–411.
4 T. Satoh and M. Hosokawa, Annu. Rev. Pharmacol. Toxicol.,
1
998, 38, 257–288.
5 F. R. D. Alves, E. J. Barreiro and C. A. M. Fraga, Mini-Rev.
Med. Chem., 2009, 9, 782–793.
This journal is © The Royal Society of Chemistry 2018
Med. Chem. Commun., 2018, 9, 1850–1861 | 1861