Discovery of 6-substituted indole-3-glyoxylamides as lead antiprion agents with enhanced cell line activity, improved microsomal stability and low toxicity

Article abstract of DOI:10.1016/j.ejmech.2011.06.013

A series of highly potent indole-3-glyoxylamide based antiprion agents was previously characterized, focusing on optimization of structure-activity relationship (SAR) at positions 1-3 of the indole system. New libraries interrogating the SAR at indole C-4

Full text of DOI:10.1016/j.ejmech.2011.06.013

European Journal of Medicinal Chemistry 46 (2011) 4125e4132
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Discovery of 6-substituted indole-3-glyoxylamides as lead antiprion agents
with enhanced cell line activity, improved microsomal stability and low toxicity
,
a
a,
Mark J. Thompson a , Jennifer C. Louth , Steven Ferrara 1, Matthew P. Jackson a, Fiona J. Sorrell a,
*
Edward J. Cochrane a, Joel Gever b, Sarah Baxendale c, B. Michael Silber b, Henry H. Roehl c,
**
Beining Chen a
,
a Krebs Institute, Department of Chemistry, University of Sheffield, Brook Hill, Sheffield S3 7HF, UK
b University of California San Francisco, School of Medicine, Institute for Neurodegenerative Diseases, Department of Neurology, 513 Parnassus Avenue, HSE-774,
San Francisco, CA 94143-0518, USA
c Department of Biomedical Science, MRC Centre for Developmental and Biomedical Genetics, University of Sheffield, Western Bank, Sheffield S10 2TN, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of highly potent indole-3-glyoxylamide based antiprion agents was previously characterized,
focusing on optimization of structureeactivity relationship (SAR) at positions 1e3 of the indole system.
New libraries interrogating the SAR at indole C-4 to C-7 now demonstrate that introducing electron-
withdrawing substituents at C-6 may improve biological activity by up to an order of magnitude, and
additionally confer higher metabolic stability. For the present screening libraries, both the degree of
potency and trends in SAR were consistent across two cell line models of prion disease, and the large
majority of compounds showed no evidence of toxic effects in zebrafish. The foregoing observations thus
make the indole-3-glyoxylamides an attractive lead series for continuing development as potential
therapeutic agents against prion disease.
Received 12 May 2011
Received in revised form
9 June 2011
Accepted 10 June 2011
Available online 17 June 2011
Keywords:
Drug discovery
Indoles
Prion disease
Structureeactivity relationships
Zebrafish
Ó 2011 Elsevier Masson SAS. All rights reserved.
1. Introduction
encephalopathy (BSE) in cattle, and chronic wasting disease (CWD)
in deer, elk and moose. In recent years, the latter of these has been
Prion diseases, or transmissible spongiform encephalopathies
(TSEs), are a group of rare but invariably fatal neurodegenerative
disorders afflicting both humans and animals. The most common
such condition affecting humans is CreutzfeldteJakob Disease
(CJD), and other examples include familial fatal insomnia (FFI) and
GerstmanneSträusslereScheinker syndrome (GSS). Animal prion
diseases include scrapie in sheep, bovine spongiform
recognized as a significant epidemic amongst wildlife populations
across certain regions of North America [1].
The fundamental molecular event common to all TSEs is post-
translational refolding of normal cellular prion protein, PrPC, into
its disease-associated isoform, PrPSc. This
b-sheet rich form of the
protein readily forms insoluble aggregates, deposition of which is
thought to be central to pathogenesis and disease progression.
Despite in vitro characterization of a large number of antiprion
compounds capable of clearing PrPSc from infected cell lines, little
therapeutic efficacy in vivo has been achieved to date, even in the
face of considerable research effort [2]. An effective treatment to
arrest or reverse progression of clinical prion disease therefore
remains an important goal.
We recently characterized a series of indole-3-glyoxylamides 1
as potent antiprion agents (Fig. 1) [3], many of which demonstrated
low nanomolar EC50 values against a prion-infected cell line
(SMB.s15) [4]. This cellular system is one of several models [5] for
prion disease that persistently produce PrPSc and is therefore used
in screening for potential therapeutics. Previous efforts were
focused around developing the structureeactivity relationship
Abbreviations: BSE, bovine spongiform encephalopathy; CJD, CreutzfeldteJakob
disease; CWD, chronic wasting disease; FFI, familial fatal insomnia; GSS, Gerst-
manneSträusslereScheinker syndrome; hpf, hours post fertilization; MAPK,
mitogen-activated protein kinase; PrPC, normal cellular prion protein (or PrP-sen);
PrPSc, disease-causing isoform (or PrP-res); SAR, structureeactivity relationship;
ScN2a, scrapie-infected neuroblastoma cells; SMB, scrapie-infected mouse brain;
TBST, tris-buffered saline (pH 7.6) containing 0.05% Tween 20; TSE, transmissible
spongiform encephalopathy; vCJD, variant Creutzfeldt-Jakob disease.
* Corresponding author. Tel.: þ44 114 222 9364.
** Corresponding author. Tel.: þ44 114 222 9467; fax: þ44 114 222 9346.
E-mail addresses: m.j.thompson@sheffield.ac.uk (M.J. Thompson), b.chen@
sheffield.ac.uk (B. Chen).
1
Present address: University of Oxford, Chemistry Research Laboratory, 12
Mansfield Road, Oxford OX1 3TA, UK.
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.06.013

4126
M.J. Thompson et al. / European Journal of Medicinal Chemistry 46 (2011) 4125e4132
acceptable physical and pharmacokinetic properties. The structures
in lead series 1 may thus be considered a promising starting point,
with good potential for the development of clinically useful
antiprion agents.
We therefore undertook to establish the effect of introducing
substituents around ‘ring B’ (Fig. 1) upon cell line activity, in order
to complete a basic understanding of the antiprion SAR across the
whole of the central indole-3-glyoxylamide framework. In addition
to this extension to the known SAR, it was also considered impor-
tant to identify compounds with the best preclinical potential by
consideration of other relevant properties. The criteria chosen were
in vitro metabolic stability, and in vivo evaluation in zebrafish as an
initial toxicity screen.
2. Chemistry
Synthesis of the new screening compounds substituted in ‘ring
B’ was carried out using a protocol similar to that reported previ-
ously [3] (Scheme 1). In the presence of electron-withdrawing
substituents (R3 ¼ F, Cl), it was found necessary to extend the
duration of the first step (reaction with oxalyl chloride) to 24 h to
achieve complete conversion to the intermediate glyoxylyl cho-
rides. In terms of final yield, this approach actually proved advan-
tageous for all of the substituted indoles, and was therefore
adopted as a general method during the preparation of screening
libraries. Initially, compounds derived from either methyl- or
chloro-substituted indoles were synthesized. All such products
were obtained successfully, with the exception of those derived
from 4-chloroindole. In this case, it appeared that the extent of
reaction with oxalyl chloride had only been very small.
Fig. 1. Representative members of the indole-3-glyoxylamide lead series previously
identified as potent inhibitors of PrPSc accumulation. In all of the examples 1ae1k,
R2 ¼ R3 ¼ H.
(SAR) at positions 1e3 of the indole ring (ring A, Fig. 1), including
investigation of the most suitable glyoxylamide substituent, which
was found optimally to consist of a para-substituted aniline moiety
(ring C). The effect of substitution at indole C-4 to C-7 (ring B) is one
important aspect of the SAR which was not considered in this
earlier work, and assessment of such modifications is therefore
a focus of the current study.
In addition to their highly potent in vitro antiprion activity,
compounds of general structure 1 constitute an encouraging lead
series given the wide variety of drug candidates containing the
indole-3-glyoxylamide moiety (Fig. 2). Indeed, this substructure
could be thought of as a ‘privileged scaffold’ [6] for drug discovery,
considering the significant number of such compounds in various
stages of clinical or preclinical development, across a range of
biological activities (Table 1). Some attrition has been inevitable,
but generally seems not to be due to adverse events in the case of
this compound class. Thus, given the range of related examples
which have already progressed to clinical studies for other disease
indications (Table 1), we were encouraged that the antiprion
indole-3-glyoxylamides are certainly amenable to tuning toward
3. Results and discussion
3.1. Investigation of antiprion structureeactivity relationship
Methyl- and chloro-substituted series derived from parent
structures 1h, 1j, 1g and 1k all displayed varying antiprion potency
in the SMB cell line (Table 2). This range of R1 groups (OMe, Me,
morpholine and H, respectively, for the parent compounds listed)
was chosen firstly because it represents a range of activities within
the existing lead series (EC50 values between 11 and 320 nM), and
secondly for ease of synthesis. Whereas compounds containing
a five-membered aromatic heterocycle in the R1 position are the
Fig. 2. Several compounds containing an indole-3-glyoxylamide (or closely related) substructure have progressed to preclinical or clinical evaluation.

M.J. Thompson et al. / European Journal of Medicinal Chemistry 46 (2011) 4125e4132
4127
Table 1
Indole-3-glyoxylamides and related compounds undergoing past or present clinical evaluation.
Compound
Action
Indication
Status
Reference(s)
Varespladib (A-002)
Secretory phospholipase (sPLA2) inhibitor
Acute coronary syndrome
Phase II completed
Phase III 2010e2011
Phase II
Phase II
Discontinued
Preclinical
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14,15]
SCIO-469
Selective p38
a
-MAPK inhibitor
Post-surgical dental pain
Multiple myeloma
Rheumatoid arthritis
Fracture healing (as analgesic)
Wound healing (in diabetes)
Solid tumors [14];
Preclinical
Phase I/II
Indibulin (D-24851; ZIO-301)
Tubulin polymerization inhibitor
Breast cancer [15]
BMS-488043
HIV-1 attachment inhibitor
General phospholipase (PLA2) inhibitor
Antibacterial
Secretory phospholipase (sPLA2) inhibitor
Phosphodiesterase-4 (PDE4) inhibitor
HIV infection
Obesity
Systemic infection
Rheumatoid arthritis
Inflammatory disorders
Phase I
[16]
[17]
[18]
[19]
[20]
Methyl indoxam
2 (several examples)
LY333013
Preclinical
Preclinical
Discontinued
Discontinued
AWD 12-281
(GSK 842470)
most effective leads, they possess poor solubility, making their
efficient synthesis and isolation in pure form somewhat problem-
atic. Therefore, more readily accessible analogs were chosen in the
first instance, in order to establish some general trends in the SAR
of derivatives substituted in ‘ring B’.
a given R3 substituent, EC50 values hold to the established order of
R1 ¼ H (least active) > Me ꢀ OMe > morpholine (most active).
Similarly, where R1 is held constant, there is a consistent trend
following the general order 4-Me (not shown) >>5-Me > 5-Cl > 7-
Me > 7-Cl > H ꢀ 6-Cl > 6-Me. In the light of these results, addi-
tional compounds were prepared for cell line evaluation, focusing
on further variations to substitution at indole C-6.
Clear effects were indeed observed on introduction of substit-
uents in these positions (Table 2). Derivatization at C-4 was espe-
cially detrimental to activity (3e5), with EC50 values raised by at
least two orders of magnitude. The 5-substituted analogs (6e13)
similarly showed a considerable worsening of antiprion effect,
albeit not to the same extent, with most EC50 values about one
order of magnitude higher relative to the parent structures.
Compounds modified at C-7 (22e27) were also less potent than
their unsubstituted counterparts, though in most cases the dete-
rioration in biological activity was not as pronounced. However,
introduction of either a chloro-, or in particular a methyl, group at
C-6 produced the desired improvement in EC50 values, indicating
this is the only site on ‘ring B’ which may be usefully modified to
enhance the antiprion activity of lead series 1.
This second series of screening compounds was prepared using
the same synthetic approach as detailed above (Scheme 1),
Table 2
Antiprion activity of indole-3-glyoxylamides substituted with either methyl- or
chloro-groups at positions C-4 to C-7.a
O
O
R1
C
N
H
R3
B A
N
H
The above trends in potency are obvious from inspection of
a graphical presentation of the data (Fig. 3). Within each series for
Ref.
SMB.s15 EC50, m
Mb
R3
R2
1h [3a]
1j [3a]
1g [3a]
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
e
OMe
0.011 ꢂ 0.006
0.13 ꢂ 0.02
e
Me
morpholine
OMe
Me
morpholine
OMe
Me
morpholine
H
OMe
Me
morpholine
H
OMe
Me
morpholine
H
OMe
Me
morpholine
H
OMe
Me
morpholine
OMe
Me
e
0.009 ꢂ 0.002
2.36 ꢂ 1.21
O
O
4-Me
4-Me
4-Me
5-Me
5-Me
5-Me
5-Me
5-Cl
5-Cl
5-Cl
5-Cl
6-Me
6-Me
6-Me
6-Me
6-Cl
6-Cl
6-Cl
6-Cl
7-Me
7-Me
7-Me
7-Cl
7-Cl
7-Cl
>20
a
Cl
0.68 ꢂ 0.10
0.88 ꢂ 0.31
0.90 ꢂ 0.30
0.083 ꢂ 0.005
11.5 ꢂ 3.3
R3
R3
N
N
R2
R2
0.47 ꢂ 0.04
0.56 ꢂ 0.12
0.069 ꢂ 0.004
1.63 ꢂ 0.81
0.012 ꢂ 0.003
0.051 ꢂ 0.008
O
R1
O
R1
N
H2N
H
R3
0.0031 ꢂ 0.0006
0.044 ꢂ 0.028
0.0099 ꢂ 0.0005
0.10 ꢂ 0.03
N
b
R2
3–5
6–9
10–13
14–17
18–21
R
R
R
R
R
2 = H, R3 = 4-Me
2 = H, R3 = 5-Me
2 = H, R3 = 5-Cl
2 = H, R3 = 6-Me
2 = H, R3 = 6-Cl
0.0073 ꢂ 0.0001
0.074 ꢂ 0.010
0.15 ꢂ 0.04
1.40 ꢂ 0.13
0.016 ꢂ 0.003
0.024 ꢂ 0.017
0.47 ꢂ 0.12
22–24 R2 = H, R3 = 7-Me
25–27
morpholine
0.0072 ꢂ 0.0054
R
2 = H, R3 = 7-Cl
a
Synthetic yields for each compound are listed in the Supplementary
Scheme 1. Preparation of indole-3-glyoxylamide screening compounds. Reagents and
conditions: (a) oxalyl chloride, THF, room temp, 24 h; (b) iPr2NEt or 2,6-lutidine, 45 ꢁC,
18 h.
Information.
b
EC50 values in bold type represent an improvement of at least a factor of two
over the unsubstituted analog.

4128
M.J. Thompson et al. / European Journal of Medicinal Chemistry 46 (2011) 4125e4132
Table 3
Second-generation library of substituted indole-3-glyoxylamide antiprion
compounds, focusing on varation of substitution at C-6.a
O
O
R1
C
N
H
R3
B A
N
R2
R3
R1
R2
Ref.
SMB.s15 EC50, m
Mb
1f [3a]
1a [3a]
28
29
30
31
32
33 [3a]
34
35
36
37
38
39
40
41
42
e
H
1H-pyrrol-1-yl
1H-pyrazol-1-yl
1H-pyrrol-1-yl
1H-pyrazol-1-yl
1H-pyrrol-1-yl
OMe
0.006 ꢂ 0.003
e
H
H
H
Me
Me
Me
Me
H
H
H
H
H
H
H
H
H
0.0010 ꢂ 0.0004
0.017 ꢂ 0.007
0.0061 ꢂ 0.0010
0.27 ꢂ 0.10
6-Cl
6-Cl
6-Cl
6-Cl
e
1.63 ꢂ 0.83
1H-pyrrol-1-yl
OMe
1H-pyrrol-1-yl
1H-pyrazol-1-yl
OMe
OMe
Me
OMe
Me
0.17 ꢂ 0.05
e
0.068 ꢂ 0.018
0.0077 ꢂ 0.0016
0.0012 ꢂ 0.0009
0.013 ꢂ 0.004
0.33 ꢂ 0.06
Fig. 3. Graphical representation of EC50 values presented in Table 2, showing clear
trends in activity with variation of R1 and R3. Bar heights correspond to observed EC50
concentrations. An enlarged version of this Figure is included within the Supple-
mentary Information for improved clarity.
6-F
6-F
6-F
6-iPr
6-iPr
6-CN
6-CN
6-NO2
6-NO2
0.61 ꢂ 0.01
0.0022 ꢂ 0.0004
0.0109 ꢂ 0.0014
0.0105 ꢂ 0.0012
0.0350 ꢂ 0.0028
beginning with the appropriate substituted indole. We also
prepared 6-Cl analogs of some 1-methylindole derived structures,
to ascertain whether introduction of the C-6 substituent would
improve the activity of these compounds where R2 ¼ Me. Deriva-
tives substituted with small alkyl groups at R2 are less active than
when R2 ¼ H [3], but are more soluble. This increased solubility
likely originates from a substitution which can disrupt intra-
molecular hydrogen-bonding in the solid state, and also reduces
the marked planarity of the molecules; both of these effects have
been suggested or adopted elsewhere as strategies to improve
aqueous solubility of drug candidates [21]. It was therefore
instructive to compare whether 6-Cl substitution could recover
some of the potency lost by introduction of a methyl group at R2.
Surprisingly, the 6-chloro-1-methyl entities (30e31) performed
more poorly than both the equivalent 1-Me (32e33) and 6-Cl (28
and 18) compounds, suggesting that derivatization at C-6 cannot be
used to compensate for the loss of activity encountered on intro-
ducing an alkyl group at N-1. In contrast to the small but noticeable
improvement seen for earlier examples (18e21, Table 2), 6-Cl
derivatives of 1f and 1a (28 and 29, respectively) actually showed
a small drop in activity compared with their parent compounds,
although both new analogs still proved very potent as antiprion
agents (EC50 < 20 nM; Table 3).
A small set of other groups at C-6 was also investigated
(Table 3). Fluoro-compounds 34e36 showed a little improvement
over their related chloro- (28, 29, 18) analogs, and comparable
activities to the C-6 unsubstituted (1f, 1a, 1h) parent compounds.
In contrast, 6-iPr derivatives 37e38 were greatly compromised in
antiprion potency compared to their unsubstituted counterparts.
The most effective modifications to ‘ring B’ proved to be intro-
duction of a 6-cyano (39e40) or 6-nitro (41e42) group, with the
former producing a reduction in EC50 by essentially an order of
magnitude. The preceding trends (summarized graphically in the
Supplementary Information; Fig. S1) suggest that strongly
electron-withdrawing groups at C-6 offer the best route to an
optimal antiprion effect, although the effectiveness of the 6-Me
compounds (14e15) does provide an intriguing exception to this
general trend.
OMe
Me
a
Synthetic yields for each compound are listed in the Supplementary
Information.
b
EC50 values in bold type represent an improvement of at least a factor of two
over the unsubstituted analog.
offering improved antiprion potency over that of the unsubstituted
parent structures.
3.2. Comparison of activity in SMB and ScN2a cells
To further appropriate the significance of the present indole-
based compounds as an important class of antiprion compounds,
a selection of screening library members, representative of a range
of activities, was also evaluated in scrapie-infected neuroblastoma
cells (ScN2a). This cell line has been widely employed as a common
model of prion disease, including screening for potential thera-
peutics. The present results allowed comparison of PrPSc clearance
in both neuronal (ScN2a) and non-neuronal (SMB) cells of murine
origin, and demonstrated good correlation between the two host
cell types (Table 4). Statistical analysis suggested there is a signifi-
cant correlation of antiprion activity between the two cell models
(P < 0.0001, Pearson r ¼ 0.9731; see Supplementary Information,
Table 4
Observed EC50 values of selected screening compounds in two scrapie-infected cell
lines.a
Ref.
SMB.s15
EC50
ScN2a
EC50, mM
,
m
M
1a
1c
1g
1h
1j
25
26
29
39
40
0.0010 ꢂ 0.0004
0.0013 ꢂ 0.0002
0.009 ꢂ 0.002
0.011 ꢂ 0.006
0.13 ꢂ 0.02
0.014 ꢂ 0.006
0.016 ꢂ 0.006
0.065 ꢂ 0.054
0.091 ꢂ 0.040
0.206 ꢂ 0.046
0.247 ꢂ 0.099
1.26 ꢂ 0.24
0.024 ꢂ 0.017
0.47 ꢂ 0.12
Thus, SAR focused upon positions C-4 to C-7 of the indole ring
had established that substitution at positions 4, 5 and 7 was
detrimental to activity, whereas certain analogs with groups
introduced at C-6 showed much improved activity. The 6-Me, 6-CN
and 6-NO2 series were identified as the most valuable, in terms of
0.0061 ꢂ 0.0010
0.0022 ꢂ 0.0004
0.0109 ꢂ 0.0014
0.175 ꢂ 0.038
0.011 ꢂ 0.000
0.026 ꢂ 0.010
a
An expanded version of this Table is included in the Supplementary Information
(Table S2), which includes structures to aid clear comparison of results.

M.J. Thompson et al. / European Journal of Medicinal Chemistry 46 (2011) 4125e4132
4129
Fig. S2), although this result is not conclusive given the relatively
small number of data pairs (n ¼ 11).
of early-stage assessment of potential drug safety is increasingly
becoming recognized in discovery programs, as a contributing
factor to reduce late-stage attrition. Over recent years, zebrafish
assays for early screening of safety pharmacology have emerged as
an important tool in this respect, due to their relatively low cost and
generally good predictivity of mammalian adverse drug effects [22].
In order to further evaluate the antiprion-active indole-3-
glyoxylamides as a source of potential prion disease therapeutics,
the survival rate of early zebrafish larvae upon exposure to each
compound was determined, to search for any evidence of acute
toxicity. Animals were dosed at 48 h post fertilization (hpf) with
Although actual EC50 values were consistently higher in the
ScN2a model, the compounds still retained excellent antiprion
potency, with the most active analogs maintaining nanomolar level
activities. For example, compounds 1a, 1c and 39 displayed EC50
values of 14 nM, 16 nM and 11 nM respectively, compared with
1.0 nM, 1.3 nM and 2.2 nM in the SMB cell line. The most important
aspect of these results is that structureeactivity relationships
proved essentially identical between the two cell lines. Amongst
the compounds unsubstituted on ‘ring B’, the general trend in EC50
with respect to R1 was consistent, i.e. R1 ¼ Me (1j) >> OMe
(1h) > morpholine (1g) >> 1H-pyrazol-1-yl (1a). Introduction of
a 7-Cl substituent remained detrimental to activity in both cell
models (25 > 1h and 26 > 1j), whereas 6-Cl substitution raised EC50
markedly in the ScN2a model (29 >> 1a) in contrast to the more
marginal differences seen in SMB cells. In both cases, 6-cyano
compounds 39 and 40 proved at least an order of magnitude
more potent than their unsubstituted analogs (1h and 1j respec-
tively), reinforcing the positive result of this modification outlined
above (Fig. 4). Thus, the SAR previously established by SMB cell line
screening was maintained across the ScN2a experiments,
strengthening the significance of the indole-3-glyoxylamides as
a lead series, and suggesting a common mechanism of action in
both cell models. Although the mode of action of these indole-
based compounds is not yet known, it has previously been estab-
lished that they do not display a discernible interaction with PrPC
itself [3].
compounds at 5 mM, and three days later the number of live fish
remaining was recorded and compared to untreated controls (for
a full summary of results see the Supplementary Information
section, Fig. S3). Toxicity screening routinely employs zebrafish
larvae, as opposed to mature fish, since the early larvae are trans-
parent and direct observation of developing organs and structures
is therefore possible.
Itwas pleasing to find thatunderthe assayconditions, overhalf of
the test compounds showed no effect on zebrafish survival,
including the 6-Me and 6-CN series (14e17 and 39e40) displaying
some of the most potent antiprion activity. Structures where
R1 ¼ Me or morpholinewererather often associated with a mortality
rate of at least 20%, suggesting these substitutions at the p-position
of the aniline moiety should be avoided; indeed, compound 1j
(R1 ¼ Me) proved the most toxic of all those evaluated.
Animals treated with three of the most potent antiprion
agentsd1a (SMB EC50 1.0 nM), 16 (3.1 nM) and 39 (2.2 nM)dwere
examined in detail after three days’ treatment and staining with
acridine orange (Fig. 4), and were developmentally normal with no
evidence for apoptosis or obvious deformities. This general absence
of acute toxic effects was encouraging, whilst at the same time, the
zebrafish screen had identified some R1 substitutions (Me, mor-
pholine) which should preferably be avoided. In addition, substi-
tution at indole C-5 correlated with evidence of toxic effects
(mortality ꢀ 20%), although given the considerable drop in antip-
rion potency also associated with such modifications, this obser-
vation was not of major concern.
3.3. Preliminary evaluation of toxicity and metabolic stability
In addition to potent activity at the screening stage, other factors
are important in selecting candidate compounds for progression
into animal studies. To further inform selection of lead compounds
for in vivo evaluation, additional assays were performed to assess
stability and toxicity of the present indole libraries. The importance
It was also considered important to identify any 6-substituted
library members which would be unsuitable leads because of
susceptibility to first pass metabolism. As such, their stability
toward mouse liver microsome preparation was assessed and
compared to that of 6-unsubstitued analogs 1h and 1j. Through this
work it became clear that all of the 6-substituted analogs showed
improved microsomal stability (Fig. 5), suggesting that indole C-6
Fig. 4. Early zebrafish larvae (48 hpf) exposed to three potent lead compounds (at
5
mM for 3 d) proved developmentally normal. Images presented here were obtained
Fig. 5. Introduction of a variety of substituents at indole C-6 imparts improved
after acridine orange staining, and some key anatomical features are labeled for
reference: (1) otic capsule; (2) swim bladder; (3) yolk.
microsomal stability. These data are tabulated in full in the Supplementary Information
(Table S3).

4130
M.J. Thompson et al. / European Journal of Medicinal Chemistry 46 (2011) 4125e4132
might be a locus of metabolism in the unsubstituted compounds. It
was pleasing to note that for the substituents found to enhance
antiprion activity (6-Me, 6-CN, 6-NO2), these compounds also dis-
played a high degree of stability over the incubation period(ꢀ85%
remaining after 30 min), implying an in vitro half-life in the order of
several hours.
The foregoing results point toward the 6-Me and 6-CN series as
a source of lead compounds showing the best balance between
antiprion activity, toxicological safety and metabolic stability. In
addition, the introduction of these substitutions is also beneficial
when considering solubility. Compounds such as 1aef, possessing
a heteroaromatic group at the R1 position, show the best antiprion
potency but suffer from the drawbacks of very low solubility and
a strong tendency to form aggregates in aqueous solutiondproblems
which were documented previously [3b]. It was therefore pleasing to
discover that introduction of a 6-Me or 6-CN group into more soluble
parent compounds (e.g. 1g, 1h or 1j) lowered the antiprion EC50 into
the same range as that for 1aef, whilst still retaining reasonable
medium. The cells were incubated with the compound-containing
medium for 5 days. After this time, cell viability was assessed by
the MTT assay following the standard protocol supplied with the
reagent (Sigma), which indicated that none of the test compounds
were toxic to the cells up to a concentration of 20 mM. For dot blot
(cell blot) analyses, cells were extracted using lysis buffer (10 mM
TriseHCl pH 7.6,100 mM NaCl,10 mM EDTA, 0.5% v/v NP40, and 0.5%
w/v sodium deoxycholate), and the content of the well was loaded
onto a nitrocellulose membrane (0.45
mm) under gentle vacuum at
a total cellular protein concentration of approximately 30e40
mg/
well (determined by the Bradford assay following the protocol
supplied with the reagent; Sigma). The membrane was air-dried and
subjected to 75 m
g mLꢃ1 proteinase K digestion for 1 h at 37 ꢁC. The
reaction was stopped with 1 mM phenylmethylsulfonyl fluoride
(PMSF) in 20 mM TriseHCl-buffered saline (TBS) and the membrane
washed extensively with TBS and immersed in 1.8 M guanidine
thiocyanate in TBS for 10 min at room temperature. After further
washing with TBS, the membrane was blocked using 5% fat-free milk
powder in phosphate-buffered saline (PBS), processed with
solubility (up to at least 10
mM, by observation in the antiprion
screening assays). In contrast, the 6-CN and 6-NO2 derivatives of 1a
could not be isolated from an attempted synthesis, presumably due to
a further step down in solubility for these analogs.
0.2 m
g mLꢃ1 mouse mAb anti-PrP 6H4 (Prionics), and developed
using an ECL kit (Amersham Pharmacia Biotech). Every experiment
was carried out in triplicate and an average value for PrPSc concen-
tration calculated, relative to an untreated control (DMSO only),
together with a standard deviation. Curcumin was employed as
a positive control, and effected essentially complete clearance of
4. Conclusions
The indole-3-glyoxylamide series previously identified as
promising antiprion agents [3] has been extended by considering
modifications at positions C-4 to C-7 of the indole system. Analogs
bearing various substituents at C-6 (R3 ¼ 6-Me, 6-CN or 6-NO2)
were found to display EC50 values up to an order of magnitude
lower than their unsubstituted parents, together with improved
stability toward microsomal metabolism and low toxicity. In
particular, the 6-Me and 6-CN compounds were identified as those
with the best balance between the properties investigated, making
them a priority for further progression toward animal studies.
Significantly, the high degree of antiprion potency observed for the
indole-based compounds in general has been confirmed in
a second prion-infected cell line, coupled with observation of
a common SAR across the two disease models. No evidence of acute
toxicity in zebrafish was seen for the large majority of the screening
set. The present results thus confirm the antiprion-active indole-3-
glyoxylamides as an important lead series in addressing the urgent
need for clinically effective agents against prion disease.
PrPSc at the concentration used (10
initially screened at 1,10 and 20
m
M). Test compounds were
m
M and were considered to be active
if PrPSc levels were reduced to less than 70% of that of the untreated
control after5 days’ exposure. Compoundsshowing activity werere-
screened over a range of concentrations to determine an EC50 value,
such experiments being repeated at least twice (in triplicate) to
validate the results so obtained. Screening in the SMB.s15 line was
previously validated by investigating two well documented antip-
rion agents as positive controls [3a]. EC50 values of 0.95
0.42 M were observed for curcumin and quinacrine, respectively,
compared to reported values of 0.01 M for curcumin (in ScNB cells)
[24], and 0.3e0.5 M for quinacrine (in ScN2a cells) [25].
mM and
m
m
m
5.1.2. Assessment of antiprion activity in ScN2a cells
The methods employed were similar to previously published
protocols [26], with the following modifications. ScN2a cells (N2a
cells infected with the Rocky Mountain Laboratory prion strain)
were seeded into black wall, clear bottom, tissue culture-treated
plates (Greiner) at 4 ꢄ 105 cells per well in 100
mL of assay medium:
Minimal Essential Medium (MEM) supplemented with 10% fetal
5. Experimental section
bovine serum, GlutaMax and 500
mg
mLꢃ1 geneticin. Test
5.1. Biology
compounds were dissolved in 100% DMSO and diluted in assay
medium at 200ꢄ final concentration before addition to the assay
plates (0.5% final DMSO concentration). Compound addition
occurred 4 h after cell seeding. After 5 days’ incubation at 37 ꢁC in
a humidified and 5% CO2-enriched environment, lysates were
generated as previously described [26] and transferred to high-
binding ELISA plates (Greiner) coated with anti-PrP Fab antibody
D18 [27], for overnight incubation at 4 ꢁC. The plates were then
washed three times with TBST before addition of horseradish
peroxidase-conjugated D13 antibody [27] (1:1000 dilution in 1%
5.1.1. Assessment of antiprion activity in SMB cells
Compounds were screened for inhibition of PrPSc formation in
SMB cells of mesodermal origin (SMB.s15 cell line) as described
previously [3], using a procedure based upon that reported by Rudyk
et al. [23]. Cells weregrown in tissue culture-treated plastic dishes in
Medium 199, supplemented with 10% newborn calf serum (heat
inactivated), 5% fetal calf serum (heat inactivated), and pen-
icillinestreptomycin at 10 mg Lꢃ1 at 37 ꢁC in an atmosphere of 5%
CO2 in air at 95% relative humidity. Mediumwas changed every third
or fourth day, and every 7 days confluent cells were passaged using
0.05% trypsin and 0.002% EDTA at a split ratio of 4. To assess the
effects of compounds, cells were distributed into 96-well cluster
plates at 12,000 cells per well and incubated for 24 h to allow for cell
attachment. The compounds were prepared at 200 times the
required concentration in DMSO as stock solutions then transferred,
at a 10-fold dilution, into Hank’s balanced salt solution. This solution
was then transferred at a further 20-fold dilution into the cell
BSA/PBS, 100
treatment, the plates were washed seven times with TBST, then
2,20-azinobis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS; 100
L)
mL), and incubation at rt for 1 h. Following this
m
was added to each well. After 10 min, absorbance at 405 nm was
recorded using a SpectraMax M5 plate reader (Molecular Devices,
Sunnyvale, CA). Doseeresponse curves were generated over
a concentration range from three independent experiments. Cal-
cein cell viability assays were run in separately seeded 96-well
black wall plates as previously described [27], with all screening

M.J. Thompson et al. / European Journal of Medicinal Chemistry 46 (2011) 4125e4132
4131
compounds showing no discernible cytotoxicity up to the
maximum concentration tested (32 mM).
Reactions were performed under an inert atmosphere of N2, and
parallel synthesis was carried out on a Radleys 12-position
carousel. Parallel extractions were carried out with liquideliquid
extraction columns, 20 mL sample loading capacity, in conjunction
with a carousel work-up station. 1H and 13C NMR spectra were
recorded at 400 MHz and 100 MHz, respectively, on a Bruker AV-
1400 spectrometer. Accurate mass and nominal mass measure-
ments were acquired using a Micromass LCT Premier XE mass
spectrometer. Purity of screening compounds was assessed by
5.1.3. Animal husbandry
Zebrafish (Danio rerio) of the AB wild-type strain were reared
under standard conditions on a 14:10 light:dark cycle [28]. Embryos
were collected using breeding traps in aquarium tanks, and sorted
and staged accordingly. Fish were kept in E3 medium [29] (5 mM
NaCl, 0.17 mM KCl, 0.33 mM CaCl2, 0.3 mM MgSO4, 0.00001%
methylene blue) at 28 ꢁC in dark conditions. All procedures were
performed in accordance with the UK Home Office Animals
Scientific Procedures Act (1986), and other relevant legistlation.
HPLC: Genesis 4
m
m C18 column (150 ꢄ 4.6 mm) or Ace 3
mm C18
column (125 ꢄ 4.6 mm); 30e100% MeCN in water over 12 min, hold
at 100% MeCN for 10 min; 15 mL injection; UV detection at 254 nm;
flow rate 1.0 mL minꢃ1. For compounds 3, 4 and 38e42, further
purification by preparative HPLC was necessary, beginning with
5.1.4. Drug administration (zebrafish toxicity assay)
Prior to addition of the drugs, 48 hpf embryos were dechorionated
using pronase (protease from Streptomyces griseus, Sigma;
3mgmLꢃ1)for15min, thenrinsedinmethyleneblue-freeE3medium
three times, and dispensed into 96-well plates at 3 fish per well in
a
solution of the crude product in MeCN/DMSO (1:1) and
m column at flow rate
employing an Alltech Alltima HP C18 HL 5
m
of 20 mL minꢃ1, using the eluent specified in each individual case.
The vast majority of compounds were isolated in >95% purity;
a complete list of HPLC purities and analytical chromatograms,
along with full characterization data for each screening compound,
is provided in the Supplementary Information (Table S1).
a fixed volume of 200 mL. Compounds were dosed at 5 mM, from
10 mM DMSO stocks at a final DMSO concentration of 1% v/v. Plates
were maintained at 28 ꢁCforafurther72h, withsurvivabilityoflarvae
assessed at this point (day 5) and compared to that of vehicle-treated
control groups. No food was provided during the test. All experiments
were conducted twice, each time in triplicate, and full results are
presented in the Supplementary Information (Fig. S2).
5.2.2. Indole-3-glyoxylamides 3e32, 34e42; general procedure
A dry 50 mL capacitycarousel reaction tube was charged with the
relevant substituted indole (1.5 mmol), and this starting material
dissolved in dry THF (12 mL). Oxalyl chloride (144
1.65 mmol) was added and the mixture stirred at room temp. After
24 h, either N,N-diisopropylethylamine (590 L, 438 mg, 3.4 mmol)
or 2,6-lutidine(396 L, 364 mg, 3.4 mmol) was introduced to the
mL, 209 mg,
5.1.5. Acridine orange staining
In order to determine any extent of apoptotic cellular death,
acridine orange was applied to treated sample fish at 1
m
m
g mLꢃ1 in
m
E3 medium for 30 min. Excess dye was removed by successive
reaction, followed by the relevant amine (1.8 mmol). The tempera-
ture was raised to 45 ꢁC, and heating continued for 18 h. The solvent
was evaporated, then the residue resuspended in a mixture of ethyl
acetate (20 mL) and brine (10 mL) and stirred vigorously for 30 min.
After passing through a liquideliquid extraction column, ensuring
washing through of the column with additional ethyl acetate,
a second extraction was carried out with sat. NH4Cl (10 mL) in the
same manner. Product solutions were then evaporated to dryness
giving the crude indole-3-glyoxylamide derivatives, which were
purified as necessary (see Supplementary Information for individual
compounddetails). Typically, twosequential recrystallizations, from
ethyl acetateehexane followed by 2-propanolewater, afforded final
products of good purity. For products derived from poorly nucleo-
philic anilines (p-1H-pyrazol-1-ylaniline,p-pyrrol-1-ylaniline and
p-morpholinoaniline), a further modification was used. Immedi-
ately following introduction of the arylamine to the reaction
mixture, DMAP (18 mg, 0.15 mmol,10 mol%) was added to assist the
final step, as its use was found to enhance the extent of conversion in
the final step in such cases.
washes with the same medium (3 ꢄ 10 min). The animals were
anesthetized using 160
mM tricaine, then mounted in 1% low
melting point agarose (Seakem) containing 160 mM tricaine and
analyzed using a Nikon AZ100M microscope. Images were captured
using a DS-Ri1 digital camera with NIS-Elements software.
5.1.6. Microsomal stability assay
A mixture of test compound (5.0 mM solution in DMSO, 0.8
isocitrate dehydrogenase (31.4 L, 2 units), 0.2 M potassium phos-
phate buffer (pH 7.0, containing 10 mM DL-isocitric acid trisodium
salt and 10 mm MgCl2; 228 L), and microsome preparation from
mouse liver (0.5 mg mLꢃ1 in the same buffer; 100
L) was incubated
in an Eppendorf tube for 5 min at 37 ꢁC. NADPH (10 mM solution;
40 L) was then added to initiate the reaction. Contents of the tube
were mixed using aspiration/dispensation five times, then a t ¼ 0
aliquot (150 L) was drawnimmediatelyand quenched with ice-cold
acetonitrile (300 L), containing 8 M 1-(2-methyl-1H-indol-3-yl)-
mL),
m
m
m
m
m
m
m
2-morpholinoethane-1,2-dione [3a] as internal reference standard.
The remaining reaction mixture was maintained at 37 ꢁC for 30 min,
then a second aliquot removed and diluted as above. The quenched
solutions were each vortexed for 30 s and centrifuged at 4000 rpm
(45ꢁ fixed-angle rotor F-45-12-11) for 10 min, then the supernatant
Acknowledgments
The authors thank Dr Caroline Parkin and Celia Holdsworth for
their generous assistance with zebrafish toxicity assays and
imaging, and Dr Dawn Teare for helpful discussions regarding
statistical analysis. This work was supported by funding from
BBSRC (Grant No. BB/E014119/1), the UK Home Office (Contract No.
HO/04-07-270), MRC Pump-priming translational research initia-
tive (Grant No. G0802527, and the University of Sheffield HEIF
Proof-of-Concept Fund. These funding agencies did not participate
in experiment design.
analyzed by HPLC: Genesis 4
m
m C18 column, 4.6 ꢄ 150 mm; 5e95%
MeCN/H2O over 4 min, hold 6 min at 95% MeCN; flow rate
1 mL minꢃ1; UV detection at 267 nm. Each analysis was performed in
triplicate and metabolism results after 30 min compared with those
at t ¼ 0. The percentage of drug remaining, together with the stan-
dard deviation, is reported for each test sample.
5.2. Chemistry
5.2.1. General procedures
Appendix. Supplementary information
Anhydrous grade THF was obtained from an in-house ‘Grubbs’
solvent purification system; all other solvents and reagents were
purchased from commercial sources and used as supplied.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.ejmech.2011.06.013.

4132
M.J. Thompson et al. / European Journal of Medicinal Chemistry 46 (2011) 4125e4132
[15] ZIOPHARM Commences Oral Indibulin Phase I/II Trial in Breast Cancer at
References
Memorial SloaneKettering Cancer Center, Med. News Today article, URL:http://
www.medicalnewstoday.com/printerfriendlynews.php?newsid¼185007,
(accessed 21.04.10).
[1] (a) S.J. Sigurdson, Vet. Res. 39 (2008) 41. doi:10.1051/vetres:2008018;
(b) M. Prcina, J. Bardon, E. Kontsekova, Acta Virol. 52 (2008) 209.
[2] (a) M.W. Brazier, V.A. Wall, B.W. Brazier, C.L. Masters, S.J. Collins, Expert Rev.
Anti Infect. Ther 7 (2009) 83;
[16] T. Wang, Z. Yin, Z. Zhang, J.A. Bender, Z. Yang, G. Johnson, Z. Yang,
L.M. Zadjura, C.J. D’Arienzo, D. DiGiugno-Parker, C. Gesenberg, G.A. Yamanaka,
Y.-F. Gong, H.-T. Ho, H. Fang, N. Zhou, B.V. McAuliffe, B.J. Eggers, L. Fan,
B. Nowicka-Sans, I.B. Dicker, Q. Gao, R.J. Colonno, P.-F. Lin, N.A. Meanwell,
J.F. Kadow, J. Med. Chem. 52 (2009) 7778.
(b) P. Brown, CNS Neurol Disord Drug Targets 8 (2009) 316;
(c) L.A. Stewart, L.H.M. Rydzewska, G.F. Keogh, R.S.G. Knight, Neurology 70
(2008) 1272.
[17] D.Y. Hui, M.J. Cope, E.D. Labonté, H.-T. Chang, J. Shao, E. Goka, A. Abousalham,
D. Charmot, J. Buysse, Brit. J. Pharmacol. 157 (2009) 1263.
[18] M. Takhi, G. Singh, C. Murugan, N. Thaplyyal, S. Maitra, K.M. Bhaskarreddy,
P.V.S. Amarnath, A. Mallik, T. Harisudan, R.K. Trivedi, K. Sreenivas,
N. Selvakumar, J. Iqbal, Bioorg. Med. Chem. Lett. 18 (2008) 5150.
[19] J.D. Bradley, A.A. Dmitrienko, A.J. Kivitz, O.S. Gluck, A.L. Weaver,
C. Wiesenhutter, S.L. Myers, G.D. Sides, J. Rheumatol. 32 (2005) 417.
[20] M.A. Giembycz, Brit. J. Pharmacol. 155 (2008) 288.
[21] (a) L.-E. Briggner, R. Hendrickx, L. Kloo, J. Rosdahl, P.H. Svensson, Chem-
MedChem 6 (2011) 60;
[3] (a) M.J. Thompson, V. Borsenberger, J.C. Louth, K.E. Judd, B. Chen, J. Med. Chem.
52 (2009) 7503;
(b) M.J. Thompson, J.C. Louth, S. Ferrara, F.J. Sorrell, B.J. Irving, E.J. Cochrane,
A.J.H.M. Meijer, B. Chen, ChemMedChem 6 (2011) 115.
[4] The SMB.s15 Cell Line Used in this Study was Provided by the TSE Resource
Centre at the Roslin Institute, University of Edinburgh, UK (http://www.roslin.
ed.ac.uk/tseresourcecentre/cell_lines.html).
[5] D. Vilette, Vet. Res. 39 (2008) 10. doi:10.1051/vetres:2007049.
[6] (a) F.R.D. Alves, E.J. Barreiro, C.A.M. Fraga, Mini Rev. Med. Chem. 9 (2009) 782;
(b) C.D. Duarte, E.J. Barreiro, C.A.M. Fraga, Mini Rev. Med. Chem. 7 (2007)
1108;
(b) M. Ishikawa, Y. Hashimoto, J. Med. Chem. 54 (2011) 1539;
(c) F. Lovering, J. Bikker, C. Humblet, J. Med. Chem. 52 (2009) 6752.
[22] (a) P.M. Eimon, A.L. Rubinstein, Expert Opin. Drug Metab. Toxicol. 5 (2009)
393;
(c) B.E. Evans, K.E. Rittle, M.G. Bock, R.M. DiPardo, R.M. Freidinger,
W.L. Whitter, F.G. Lundell, D.F. Veber, P.S. Anderson, R.S. Chang, J. Med. Chem.
31 (1988) 2235.
(b) W.S. Redfern, G. Waldron, M.J. Winter, P. Butler, M. Holbrook, R. Wallis,
J.-P. Valentin, J. Pharmacol. Toxicol. Meth. 58 (2008) 110;
(c) S. Berghmans, P. Butler, P. Goldsmith, G. Waldron, I. Gardner, Z. Golder,
F.M. Richards, G. Kimber, A. Roach, W. Alderton, A. Fleming, J. Pharm. Tox-
icol. Meth. 58 (2008) 59.
[7] R.S. Rosenson, C. Hislop, D. McConnell, M. Elliott, Y. Stasiv, N. Wang,
D.D. Waters, Lancet 373 (2009) 649.
[8] (a) R.S. Rosenson, C. Hislop, M. Elliott, Y. Stasiv, M. Goulder, D. Waters, J. Am.
Coll. Cardiol. 56 (2010) 1079;
(b) R.S. Rosenson, H. Fraser, J. Trias, C. Hislop, Expert Opin. Investig. Drugs 19
(2010) 1245;
(c) UKMi New Drugs Online Database, URL:http://www.ukmi.nhs.uk/applicat
ions/ndo/record_view_open.asp?newDrugID¼4619, (accessed 22.04.10).
[9] S.E. Tong, S.E. Daniels, T. Montano, S. Chang, P. Desjardins, Am. Soc. Clin.
Pharmacol. Ther. 75 (2004) PI-1.
[10] D.S. Siegel, A. Krishnan, S. Lonial, G. Chatta, M. Alsina, S. Jaggannath,
P. Richardson, R.J. Hohl, J.A. Lust, W. Bensinger, G. Carrum, J. Moreb, A. Simic,
B. Barlogie, R.T. Maziarz, K.C. Anderson, J. Lin, A. Lowe, S. Vetticaden, J. Zhu,
Blood 108 (3580) (2006).
[11] (a) D.M. Goldstein, A. Kuglstatter, Y. Lou, M.J. Soth, J. Med. Chem. 53 (2010)
2345;
(b) S. Cohen, R. Fleischmann, Curr. Opin. Rheumatol. 22 (2010) 330.
[12] J.A. Cottrell, M. Meyenhofer, S. Medicherla, L. Higgins, J.P. O’Connor, Pain 142
(2009) 116.
[13] S. Medicherla, S. Wadsworth, B. Cullen, D. Silcock, J.Y. Ma, R. Mangadu, I. Kerr,
S. Chakravarty, G.L. Luedtke, S. Dugar, A.A. Protter, L.S. Higgins, Diabetes
Metab. Syndr. Obes. Targets Ther. 2 (2009) 91.
[23] H. Rudyk, S. Vasiljevic, R.M. Hennion, C.R. Birkett, J. Hope, I.H. Gilbert,
J. Gen.Virol 81 (2000) 1155.
[24] B. Caughey, L.D. Raymond, G.J. Raymond, L. Maxson, J. Silveira, G.S. Baron,
J. Virol. 77 (2003) 5499.
[25] (a) K. Doh-Ura, T. Iwaki, B. Caughey, J. Virol. 74 (2000) 4894;
(b) C. Korth, B.C.H. May, F.E. Cohen, S.B. Prusiner, Proc. Nat. Acad. Sci. USA 98
(2001) 9836.
[26] B.C. May, J.A. Zorn, J. Witkop, J. Sherrill, A.C. Wallace, G. Legname, S.B. Prusiner,
F.E. Cohen, J. Med. Chem. 50 (2007) 65.
[27] (a) S. Ghaemmaghami, B.C. May, A.R. Renslo, S.B. Prusiner, J. Virol. 84 (2010)
3408;
(b) A. Gallardo-Godoy, J. Gever, K.L. Fife, B.M. Silber, S.B. Prusiner, A.R. Renslo,
J. Med. Chem. 54 (2011) 1010.
[28] (a) M. Westerfield, in: M. Westerfield (Ed.), The Zebrafish Book: a Guide for
the Laboratory Use of Zebrafish, third ed.. Univ. of Oregon Press, Eugene,
1995;
(b) C.B. Kimmel, W.W. Ballard, S.R. Kimmel, B. Ullmann, T.F. Schilling, Dev.
Dynam 203 (1995) 253.
[29] C. Nusslein-Volhard, R. Dahm (Eds.), Zebrafish: a Practical Approach, first ed.
Oxford University Press, Oxford, 2002.
[14] R.L. Oostendorp, P.O. Witteveen, B. Schwartz, L.D. Vainchtein, M. Schot, A. Nol,
H. Rosing, J.H. Beijnen, E.E. Voest, J.H.M. Schellens, Investig. New Drugs 28
(2010) 163.