Profiling withanolide A for therapeutic targets in neurodegenerative diseases

Article abstract of DOI:10.1016/j.bmc.2019.03.022

To identify new potential therapeutic targets for neurodegenerative diseases, we initiated activity-based protein profiling studies with withanolide A (WitA), a known neuritogenic constituent of Withania somnifera root with unknown mechanism of action. Molecular probes were designed and synthesized, and led to the discovery of the glucocorticoid receptor (GR) as potential target. Molecular modeling calculations using the VirtualToxLab predicted a weak binding affinity of WitA for GR. Neurite outgrowth experiments in human neuroblastoma SH-SY5Y cells further supported a glucocorticoid-dependent mechanism, finding that WitA was able to reverse the outgrowth inhibition mediated by dexamethasone (Dex). However, further GR binding and transactivation assays found no direct interference of WitA. Further molecular modeling analysis suggested that WitA, although forming several contacts with residues in the GR binding pocket, is lacking key stabilizing interactions as observed for Dex. Taken together, the data suggest that WitA-dependent induction of neurite outgrowth is not through a direct effect on GR, but might be mediated through a closely related pathway. Further experiments should evaluate a possible role of GR modulators and/or related signaling pathways such as ERK, Akt, NF-κB, TRα, or Hsp90 as potential targets in the WitA-mediated neuromodulatory effects.

Full text of DOI:10.1016/j.bmc.2019.03.022

Accepted Manuscript
Profiling Withanolide A for Therapeutic Targets in Neurodegenerative Diseases
Erika A. Crane, Wolfgang Heydenreuter, Katharina R. Beck, Petra Strajhar, Jan
Vomacka, Martin Smiesko, Elma Mons, Lydia Barth, Markus Neuburger,
Angelo Vedani, Alex Odermatt, Stephan A. Sieber, Karl Gademann
PII:
S0968-0896(19)30232-9
https://doi.org/10.1016/j.bmc.2019.03.022
BMC 14814
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
8 February 2019
8 March 2019
10 March 2019
Please cite this article as: Crane, E.A., Heydenreuter, W., Beck, K.R., Strajhar, P., Vomacka, J., Smiesko, M., Mons,
E., Barth, L., Neuburger, M., Vedani, A., Odermatt, A., Sieber, S.A., Gademann, K., Profiling Withanolide A for
Therapeutic Targets in Neurodegenerative Diseases, Bioorganic & Medicinal Chemistry (2019), doi: https://doi.org/
10.1016/j.bmc.2019.03.022
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Profiling Withanolide A for Therapeutic Targets in
Neurodegenerative Diseases
Erika A. Crane,† Wolfgang Heydenreuter,⨎ Katharina R. Beck,‡ Petra Strajhar,‡ Jan Vomacka, ⨎ Martin
Smiesko,‡ Elma Mons,† Lydia Barth,† Markus Neuburger,† Angelo Vedani,‡,  Alex Odermatt,‡ Stephan
A. Sieber, ⨎ and Karl Gademann*§
†Department of Chemistry, University of Basel, St. Johanns-Ring 19, CH-4056 Basel, Switzerland
⨎Department of Chemistry, Centre for Integrated Protein Science (CIPSM), Institute for Advanced Studies (IAS), Technical University Munich,
Lichtenbergstrasse 4, D-85747 Garching, Germany
‡ Department of Pharmaceutical Sciences, University of Basel, Klingelbergstrasse 50, CH-4056 Basel, Switzerland
 Deceased June 28, 2016
§Department of Chemistry, University of Zurich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland
KEYWORDS withanolide A, target profiling, glucocorticoid receptor, neurotrophic agents
ABSTRACT: To identify new potential therapeutic targets for neurodegenerative diseases, we initiated activity-based protein
profiling studies with withanolide A (WitA), a known neuritogenic constituent of Withania somnifera root with unknown mechanism
of action. Molecular probes were designed and synthesized, and led to the discovery of the glucocorticoid receptor (GR) as potential
target. Molecular modeling calculations using the VirtualToxLab predicted a weak binding affinity of WitA for GR. Neurite outgrowth
experiments in human neuroblastoma SH-SY5Y cells further supported a glucocorticoid-dependent mechanism, finding that WitA
was able to reverse the outgrowth inhibition mediated by dexamethasone (Dex). However, further GR binding and transactivation
assays found no direct interference of WitA. Further molecular modeling analysis suggested that WitA, although forming several
contacts with residues in the GR binding pocket, is lacking key stabilizing interactions as observed for Dex. Taken together, the data
suggest that WitA-dependent induction of neurite outgrowth is not through a direct effect on GR, but might be mediated through a
closely related pathway. Further experiments should evaluate a possible role of GR modulators and/or related signaling pathways
such as ERK, Akt, NF-, TR, or Hsp90 as potential targets in the WitA-mediated neuromodulatory effects.
1. Introduction
There has been a worldwide escalation over the past
antiproliferative, neurite-outgrowth, and cholinesterase
inhibitory activities.23,25,27,29-33 In particular, WitA possesses
strong neuropharmacological properties, promoting neurite
outgrowth, reversing neuritic atrophy and aiding in synapse
reconstruction.26,34,35 Furthermore, WitA has been suggested to
modulate several targets involved in Alzheimer’s disease such
as downregulation of BACE1, upregulation of ADAM10 in
primary rat cortical neurons, and the activity of several
secretases.36 Some of us completed a synthesis of WitA and
performed synthetic studies to access derivatives, which were
then evaluated for their neurite outgrowth potential.37,38
However, our interest in this natural product extended to a
desire to understand its underlying mechanism of action and
elucidate further biological targets for broader applications,
which we hoped to achieve through target identification of this
natural product via affinity-based protein profiling (ABPP).39-42
century in the prevalence of neurodegenerative diseases, such
as Alzheimer’s, Parkinson’s and Huntington’s disease.1-4
However, there are still no cures for these disorders and current
treatment options are only palliative. So far, drug discovery
efforts have focused on small molecules that mimic
neurotrophic properties of established polypeptidal
therapeutics, as they possess inherently better pharmacological
properties, mainly the ability to cross the blood-brain barrier.1,5-
10 More specifically, natural products or fragments thereof have
become of special interest.11-18 Thus, we became interested in
withanolide A (WitA, 1, Figure 1), one of the most bioactive
constituents in Withania somnifera, also known as
Ashwagandha root, Indian ginseng or winter cherry.19-22 This
ayurvedic herb has been used for centuries to treat a variety of
ailments from arthritis, anxiety, insomnia, tuberculosis, asthma,
to fibromyalgia, among others.21,23,24 Even today, it is sold at
yogi or organic shops in the form of herbal supplements or in
teas as anti-inflammatory treatment, for cognition improvement
and anti-aging effects, increase in sexual desire and fertility,
among others.21,25,26 A major steroidal component of this root,
WitA, was isolated in 1971 from the roots of Withania
somnifera (L.) DUNAL.19 This compound is one of the 900
naturally occurring withanolides, a family of C28 steroids
characterized by an ergostane skeleton and common
functionalization at C1, C22 and C26.27,28 Approximately 550
of these withanolides are considered “classic” or “unmodified”,
with a four-ringed steroidal core (ABCD) and an appended -
lactone ring (E). They are known for their antitumor, anti-
inflammatory, neuroprotective, immunomodulatory and insect
Figure 1. Structures of withanolide
A
(WitA, 1),
dexamethasone (Dex, 2), and mifepristone (Mife, RU-486, 3).
anti-feedent
activities,
due
to
well-documented

2. Results and discussion
2.1. Design of withanolide A molecular probes
WitA exhibits putative electrophilic moieties such as an
Scheme 1. Synthesis of WitA Molecular Probes for Target
Profiling.
O
O
O
1. NH2OH•HCl
pyr,  70 %
N
N
1. Pd/C, H2
EtOH/MeOH, >99%
Me
OH
Me
OH
Me
OH
epoxide and two Michael acceptor systems which could
mediate target binding via covalent addition to corresponding
nucleophilic amino acid residues in protein active sites.
Therefore, the WitA core scaffold was functionalized with an
alkyne handle at different positions and these probes were
applied in gel-based and gel-free ABPP.39-42 In this procedure,
either SH-SY5Y human neuroblastoma or A549 human lung
carcinoma cells were incubated for 1 to 20 hours with the probes
at concentrations of 1–100 M to enable labeling. The
proteomes were then harvested and were bioorthogonally
ligated to either a rhodamine- or biotin azide reporter tag
through a Cu(I)-catalyzed [3+2] cycloaddition, or click
chemistry, for visualization on fluorescent SDS-gels or mass-
spectrometry (MS) based target identification, respectively.43-45
O
2.
Br
2. HCl•H2N
pyr, , 75%
KOH, DMSO
3h, 58%
5
1
4
1. tBuOOH, TBAF, CH2Cl2, 88%
2. N2H4, AcOH, MeOH, CH2Cl2, 65%
3. PDC,CH2Cl2, 87%
Me
OH
Me
OH
Me
OH
1. Pd/C, H2
EtOH/MeOH, 80%
O
HCl•H2N
N
O
N
O
pyr, , 74%
O
O
2. HCl•H2N
6
7
10
pyr, , >99%
(Z)-7
major isomer
Me
Me
OH
Me
Me
OH
Me
Me
Me
Me
For
a successful target profiling, the investigated
Br
O
O
O
O
1
O
O
compound needs to have a balanced reactivity and be
modifiable with a sterically inert alkyne, distal from its reacting
center.41 The wide-ranging biological activities of the
withanolides have been strongly linked to covalent
modification and thorough analysis of structure-activity
relationship (SAR) data revealed that the 2-ene-1-one of the A-
ring was essential for activity.27 Due to several possible
conjugate acceptors existing in the molecule and for
completeness sake, both A- and E-ring probes were targeted.
Our prior SAR investigations of WitA revealed that neurite
outgrowth activity could be retained with simple C1-oximes,
making these favorable targets for the A-ring probe synthesis.38
These synthetic studies also revealed A-ring modification to be
fairly facile, but functionalizing the C1 position was rather
challenging. The -lactone was also found to be particularly
stable, rendering its functionalization difficult. Ideally, the R
group of the C1 oxime would be designed to lack bulk,
otherwise risk disrupting possible binding interactions with the
target(s) in the assay. In addition, the oxygenation pattern in the
B ring could not be removed or modified to prevent important
contacts in the binding pockets of the target(s) from being lost.
The synthesis of inert probes was targeted as well for
elucidation of potential off-target effects, by removing the
conjugate acceptor ability of the molecule.
Me
H
Me
H
O
H
H
Bu4NI, 50% NaOH
CH2Cl2,
10% 8 + 33% 9,
28% recovered 1
H
H
H
H
O
OH
9
8
8
2.2 Synthesis of withanolide A molecular probes
For these synthetic studies, fresh WitA was isolated from
powdered Ashwagandha root (0.038% yield, w/w) according to
our previously published procedure.37,38 In the synthesis of the
A-ring probes, the unsaturated C1 propargyl oxime 4 was
prepared in a two-step procedure with initial formation of the
N-H oxime and then alkylation with propargyl bromide
(Scheme 1). Direct formation of probe 4 was not possible,
complimentary to what was observed in previous synthetic
studies.38 The corresponding saturated probe 5 was synthesized
directly through propargyl oxime formation in a 75% yield,
following a quantitative hydrogenation of WitA (1). The C3
probes were accessed through a Wharton transposition of WitA
(1) and oxidation to form C3 enone 10.38 The unsaturated C3
propargyl oxime probe 6 could then be prepared directly in one
step in a 74% yield. A mixture of oxime isomers was observed
in this reaction after several hours, but the E-isomer could be
exclusively obtained when the reaction time was extended (ca.
3 days). Similarly, saturated probe 7 was accessed in
quantitative yield, following hydrogenation. We obtained a 3:2
ratio of separable Z:E oxime isomers. X-ray crystallography of
the (Z)-7 isomer confirmed the configuration assignment for
each compound.46 Only the Z-isomer of 7 was used in the target
profiling and biological studies.
The E-ring probes were isolated from the same reaction in
an attempt to alkylate the C20 tertiary alcohol under phase-
transfer conditions. Propargyl ester 8 was formed during ring
opening of the -lactone, while unsaturated propargyl ester 9
2

was created through further elimination of the C5 hydroxyl
group. The structure of ester 8 was confirmed by X-ray
crystallography.46
2.3 Affinity-based protein profiling tests
The six molecular probes were then screened in the ABPP
assay. Analytical runs were performed at 1, 20 and 100 M
probe concentrations applied to SH-SY5Y and A549 cells.
Analysis by SDS-PAGE and fluorescence scanning revealed
the most intense reactivity profile for probe 9 (Figure 2).47Also
showing specific labeling, but to a lesser degree, were probes 8,
and the C1- and C3-propargyl oximes 4 and then 6. As proof of
concept, the ‘inert’ saturated probes 5 and 7 did not show any
specific labeling in these assays, serving as a support for the
pharmacological hypothesis for the mode of action of WitA.
These results nicely correlate with the expected chemical
reactivity of each probe, as the unsaturated oxime of probe 4 is
not only less electrophilic than the corresponding enones of
probes 8 and 9, but possible steric congestion within the binding
pocket of the targets with the propargylic oxime group at the
C1-position could very well explain the less specific labeling
observed for probe 4 over probes 8 and 9.
Figure 2. Representative SDS-PAGE showing analytical
labeling of cytosol fraction of SH-SY5Y cells incubated with
all six WitA probes (at a 1, 10 or 100 M concentration) for 6
hours in PBS at 0.1% DMSO. Fluorescent marker (FM) lane
designates protein size in kDa. No bands were found for the
membrane fractions (data not shown)
A. 8
B. 9
6
4
2
C
A
D
V
3.5
3
5
4
3
2
1
0
ATP2A2
RTN4
2.5
2
NR3C1
HNRNPF
3
C
A
D
V
MGST3
1.5
1
GPATCH8
GCLM
PTGES2
0.5
0
-2
-1
0
1
2
3
4
5
6
7
-3
-1
1
3
5
7
9
log2 enrichment (probe/DMSO)
log2 enrichment (probe/DMSO)
Table 1. Targets in Common from Activity-Based Profiling with WitA Probes 4, 8, and 9.
Rank
protein names
gene
names
probe
hit
9
glutathione S-transferase omega-1
mitochondrial carrier homolog 2
glucocorticoid receptor
GSTO1 8, 9
MTCH2 8, 9
10
11
1
sarcoplasmic/endoplasmic reticulum
calcium ATPase 2
ATP2A2
VDAC2
8, 9
NR3C1 4,
8, 9
12
13
14
15
1,25-dihydroxyvitamin D(3) 24-hydroxylase, CYP24 4, 8
mitochondrial A1
2
voltage-dependent anion-selective
channel protein 2
8, 9
voltage-dependent anion-selective channel protein VDAC3 8, 9
3
4
5
6
7
8
reticulon-4
RTN4
CYB5B
CKAP4
MGST3
HM13
4, 8, 9
4, 8, 9
4, 8, 9
4, 8, 9
4, 8
3
cytochrome b5 type B
prostaglandin E synthase 2
PTGES 4, 8
2
cytoskeleton-associated protein 4
microsomal glutathione S-transferase 3
minor histocompatibility antigen H13
histidine-tRNA ligase, cytoplasmic
HARS 4, 8
mitochondrial import receptor subunit TOM40 TOMM 8, 9
homolog 40
Figure 3. SILAC data volcano plots of WitA probes 8 (A), 9 (B) and 4 (C): fold-change, enrichment extent normalized to the DMSO
control (log2 enrichment) versus statistical significance (–log10 p-value); p-value from ANOVA statistical model; top hits designated
in blue and labeled, GR (NR3C1) is labeled in red. Targets in common ranked according to enrichment and p-value are displayed in
Table 1.
3

In order to elucidate the identity of all targets, quantitative,
gel-free ABPP was performed via stable isotope labeling of
amino acids in cell culture (SILAC) with the top three probes
exhibiting the most pronounced labeling effect from the
analytical experiments.48 A549 cells were cultivated in media
containing either heavy or light isotope labeled amino acids.
These cells were then treated with the probes 4, 8 or 9 at 10 µM
concentration (or DMSO control). The cells were then lysed,
pooled with the control, clicked to a biotin-azide linker49 and
enriched using avidin beads. The enriched proteins were
digested using trypsin. The resulting solution was analyzed by
high-resolution liquid chromatography tandem MS (LC-
MS/MS). Log2(isotopic ratios) values of zero indicate no
enrichment as compared to the DMSO control, and ratios
greater than 1-fold with a p-value of <0.05 were regarded as
significant hits. These top hits with our three active probes are
depicted in the upper right hand area of the corresponding
volcano plots (A-C, Figure 3).50 In addition, the hits shared in
common between the three probes were identified and ranked
according to enrichment and p-value to give the 15 hits listed in
Table 1 (Figure 3).51
treatment and restore neurogenesis in cells,56-58 and has become
associated with neuroprotection,59-62 implying that this is a
glucocorticoid-dependent process.63 It is also known that GCs
exacerbate clinical symptoms and neurodegenerative processes
(in vivo and in vitro),64 and that GR-mediated transrepression
has been linked to prevent neurodegeneration in Huntington’s
disease.65 Therefore, the present study investigated whether the
WitA-dependent effect on neurite outgrowth might be mediated
by GR.
2.4 Computational approaches
We initiated our investigation of a potential binding of WitA to
GR with the computational modeling program VirtualToxLab,65
an in silico anti-target screening tool.66-67,68 This technology is
based on an automated, flexible docking calculation combined
with multi-dimensional quantitative structure–activity
relationship (QSAR) protocol that simulates and quantifies the
binding of small molecules towards a series of currently 16
proteins, known or suspected to trigger adverse effects: 10
nuclear receptors (androgen receptor, estrogen receptor  ,
estrogen receptor  , GR, liver X receptor, mineralocorticoid
receptor, peroxisome proliferator-activated receptor ,
progesterone receptor, thyroid receptor  , thyroid receptor  ),
four members of the cytochrome P450 enzyme family
(CYP1A2, CYP2C9, CYP2D6, CYP3A4),
a cytosolic
transcription factor (aryl hydrocarbon receptor) and a potassium
ion channel (hERG). While binding modes and affinities were
predicted towards all 16 target proteins comprising the
VirtualToxLab, the interest was mainly focused on GR.69 GR
agonist Dex was predicted to bind with 1.7 nM affinity to GR
(exp. = 2.0 nM), while the predicted affinity of the known GR
antagonist Mife was 2.6 nM (exp. = 15 nM) (Figure 4). For
WitA, a weaker affinity of 5.7 µM for GR was predicted, and
the ABPP-active WitA probes were predicted to bind with
similar or lower affinity than WitA. Affinities computed with
the VirtualToxLab represent a “thermodynamic” estimate only
and do not account for the kinetic stability of the ligand–protein
complex. To address this issue, WitA was probed further
through iterative inspection in real-time 3D by molecular
dynamics (MD) simulations, which could require correcting the
“thermodynamic affinity” in either direction. Both Dex and
WitA demonstrated stable ligand–protein interactions
throughout the entire 100 ns MD simulation.70 Each complex
was further stabilized by the dynamic relaxation of the
technique, indicating that the “thermodynamic” binding
affinities might be slightly higher than those predicted by the
VirtualToxLab.
Figure 4. Predicted GR binding affinities [–log(Kcalc)] of
compounds of interest calculated using VirtualToxLab.
While several of these hits might be of interest in
understanding neurological pathways, we were particularly
interested in the glucocorticoid receptor (GR, systematic name:
NR3C1). The GR is a 78.7 kDa protein that belongs to the
family of nuclear receptors and regulates genes for
development, metabolism and immune response, through either
transactivation or transrepression mechanisms.52-54 Molecules
that modulate this target, i.e. glucocorticoids, are mediators of
stress response. A 2014 report by Terada and co-workers found
that an antagonist of GR, such as mifepristone (RU486, Mife,
3, Figure 1), can induce neurite outgrowth by phosphorylation
of Akt (protein kinase B, PKB) and extracellular regulated
kinase 1/2 (ERK1/2), which are involved downstream in nerve
growth factor (NGF)-induced neurite outgrowth.55 At the same
time, GR agonists like dexamethasone (Dex, 2, Figure 1) impair
2.5 Neurite outgrowth GR competition assays
We then wanted to clarify if WitA’s neurite outgrowth activity
was linked to its ability to modulate GR. Therefore, we
performed in vitro neurite outgrowth experiments with Dex,
Mife, and WitA. Human neuroblastoma SH-SY5Y cells were
cultured in DMEM/F-12 media supplemented with 10% fetal
bovine serum and 1% penicillin-streptomycin at 37°C in a
humidified atmosphere (5% CO2 content). For neurite
outgrowth experiments the cells were seeded on collagen-
coated 24-well plates and treated with 1 µM of the
corresponding compounds, vehicle control (0.1% DMSO) or
neurite
outgrowth
through
interference
of
activation/phosphorylation of these same targets. In addition,
treatment with Mife reversed the inhibitory activity of Dex on
neurite outgrowth via the same NGF-induced Akt and ERK1/2
phosphorylation pathways. Mife has been previously
demonstrated to reverse effects of glucocorticoid (GC)
4

70%
60%
50%
40%
30%
20%
10%
0%
all-trans retinoic acid (ATRA, RA) as a positive control. After
a 6-day incubation, the media was removed and cells were fixed
with a 4% buffered formaldehyde solution and stained with
modified Giemsa stain to visualize the cell bodies for
subsequent imaging under a phase-contrast microscope. All
assays were performed in triplicate, with 611 images randomly
taken between biological replicates, and at least 350 cells
counted per treatment per plate (three wells/plate per treatment).
Cells having at least one neurite with a length of more than 50
m were counted as differentiated (positive). The ratio of
neurite positive cells to total cells was calculated and error is
represented by the standard error of the mean (SEM).
70%
60%
50%
40%
30%
20%
10%
0%
Figure 7. Neurite outgrowth in GR competition experiment of
WitA (1), probes (4, 8, 9) with Dex (2) and Mife (3) in human
SH-SY5Y cells. 0.1% DMSO and RA are vehicle and positive
controls. Data represents mean ± SEM from three independent
experiments in triplicates For statistical analysis, ANOVA and
post hoc tests (Dunnett’s Method and Tukey Kramer Method)
were used.79
2.6 GR and MR transactivation bioassays
We then performed in vitro assays to investigate whether WitA
might act as an antagonist of GR activity. HEK-293 cells were
transiently transfected with human GR under the control of a
GR sensitive luciferase promoter (Figure 8A). Additionally, the
closely related mineralocorticoid receptor (MR) was tested for
modulation by WitA (Figure 8B). 10 M WitA was not able to
stimulate GR-mediated activation of the luciferase reporter
gene, whereas the MR was weakly activated. However, and
more importantly regarding GR-dependent neuronal outgrowth,
no antagonistic effect on the cortisol-dependent GR activation
could be observed with 1 M WitA. In addition, neither an
agonistic nor an antagonistic activity of WitA on MR function
could be observed. Thus, a direct interference of WitA with the
GR and the MR could not be demonstrated within this cell-
based assays.
DMSO RA Dex Mife WitA
4
5
6
7
8
9
Figure 6. Neurite outgrowth of human SH-SY5Y cells treated
with WitA (1), Dex (2), Mife (3), and WitA probes (49). 0.1%
DMSO and RA are vehicle and positive controls. Data
represents mean ± SEM from three independent experiments in
triplicates For statistical analysis, ANOVA and post hoc tests
(Dunnett’s Method and Tukey Kramer Method) were used.77
From these experiments, we observed that all six
withanolide probes induced neurite outgrowth of SHSY5Y cells
(Figure 6). All four oxime probes (47) induced outgrowth to a
similar extent as WitA and Mife and showed significantly
higher activity than propargyl ester probes 8 and 9.72
Nevertheless, the propargyl esters still showed outgrowth
activity comparable to the RA positive control. Moreover, WitA
was also able to counteract the neurite outgrowth inhibition of
Dex, suggesting that WitA is potentially acting as a GR
antagonist, similar to Mife (Figure 7). The same effect was
observed for both propargyl ester probes 8 and 9. Interestingly,
impaired neurite outgrowth was observed in the simultaneous
treatments of both Dex and Mife with unsaturated C1 propargyl
oxime 4, within the statistical significance of the vehicle DMSO
control (indicating no outgrowth).
To exclude the possibility that either WitA is under the
control of efflux transport or that it was unable to cross the cell
membrane, a GR binding assay was performed under cell-free
conditions. Lysates of SF9 cells expressing recombinant human
GR were incubated in the presence of 10 nM 3H-dexamethasone
as a tracer and the unlabeled competitor (either 10 M Dex or
1 or 10 M WitA) until the binding equilibrium was reached.
3
The GR bound fraction of H-dexamethasone is displayed in
Fig. 9. Unlike with unlabeled Dex, WitA could not displace 3H-
Dex from the GR binding site at concentrations up to 10 M
(Figure 9). This indicates that WitA is not a ligand for GR, as
could be predicted by the calculated binding affinities of WitA,
discussed above. These results are also in line with those from
the transactivation assays, indicating that WitA does not induce
neurite outgrowth by direct antagonism of GR.
5

clearly greater distances and two of five hydrogen bonds
involve sulfur center (Met601 and Cys736). These
a
interactions are generally regarded as weak H-bonds compared
to interactions including oxygen or nitrogen groups.
Furthermore, WitA has less hydrophobic interaction partners
than Dex. Although we were able to dock WitA into the binding
pocket of GR and therefore confirmed the VirtualToxLab
screening results, the interaction pattern shows distinctively less
favorable interactions compared to Dex. This may explain the
inability of WitA to displace Dex from GR in the lysate binding
assay. A weak but covalent binding of WitA to GR might be
missed by the docking calculations; however, in vitro
verification
assessments
excluded
this
possibility.
Nevertheless, regarding its effect on neurite outgrowth, one
would expect an antagonistic action of WitA on GR function.
In contrast to the agonist bound position, active antagonist
conformations can displace the helix 12 and thereby trigger a
molecular shift to alter the coactivator site into the corepressor
site.81 Docking of WitA into a GR X-ray structure with Mife as
co-crystallized ligand (PDB code 1NHZ) showed no
interference of WitA with the helix 12 (data not shown). Thus,
the molecular modeling does not support a direct antagonistic
activity of WitA on GR, in line with the lack of antagonistic
activity of WitA in the transactivation assay.
Figure 8. Effect of WitA on the modulation of GR and MR.
Impact of WitA on GR (A) and MR (B) transactivation in HEK-
293 cells transfected with a corticosteroid receptor sensitive
luciferase reporter gene and human GR or human MR. Cells
were incubated for 24 h in the presence or absence of 100 nM
cortisol (F) or 1 M Mife for GR, and 10 nM aldosterone (Aldo)
or 1 M spironolactone (Spiro) for MR. WitA was added at 1
or 10 M. Data were normalized to vehicle control (0.05%
DMSO) and represent mean ± SD from a representative
experiment performed in triplicates.
Figure 9. Displacement of radiolabeled Dex from GR with
unlabeled competitor (WitA or Dex). Values were normalized
to the maximal binding activity of 10 nM 3H-Dex.
2.7 GR docking study
To computationally rationalize the results obtained from
the transactivation and binding assays, docking studies
comparing WitA and Dex for important binding interactions in
the GR binding pocket were conducted using the same crystal
structure as in the VirtualToxLab screening (PDB code 1M2Z.
Figure 10A illustrates docked Dex (grey) and WitA (purple) in
the GR binding site. The automatically generated
pharmacophore interaction pattern revealed different residues
forming hydrophobic and hydrogen bond (H-bond) interactions
for Dex (Figure 10B) and WitA (Figure 10C). Generally, the
docking calculations displayed different conformations for
WitA, whereas Dex was found to fit in one major pose. Dex is
stabilized by H-bond interactions in the GR binding pocket,
exerted by oxygen and nitrogen atoms of the following residues:
Asn564, Gln570, Arg611, Gln642 and Thr739. These
interactions are known to be important for GR activation.80
However, the H-bond interactions of Wit A with the GR adopt
Figure 10. Docking of Dex and WitA into the binding pocket
of GR. (A) Dex (grey) and WitA (purple) docked into the GR
binding site. Amino acid residues important for binding are
depicted as sticks. Yellow spheres display H-bond features
important for binding. Arrows indicate H-bonds with the
corresponding distance between the ligand and the residue of
the binding pocket (grey numbering belong to Dex and purple
to WitA). The color code of the arrows correspond with binding
6

interactions of part
B
and C. (B) Two-dimensional
absorptions in cm-1. All HRMS-ESI spectra were recorded by
the mass spectrometry service of the chemistry department at
the University of Basel on a QTOF-ESI spectrophotometer
(bruker MaXis 4G). Fragment ions are given in m/z with
relative intensities (%) in parentheses. Melting points (m. p.)
were collected on a Büchi B-545 apparatus in open capillaries.
Optical rotations were measured on a Jasco P-2000 digital
polarimeter with a sodium lamp (D-589 nm) in a 1 dm
cylindrical cell (1 mL cell volume) and are reported as follows:
[훼]푇°퐶
representation of the binding interactions of Dex with GR. (C)
Binding interactions of WitA and GR are shown. Hydrophobic
interactions are depicted in yellow, hydrogen bond acceptors in
red and hydrogen bond donors in green.
3. Conclusion
In this study, we investigated the GR as a target for WitA–
dependent neurite outgrowth activity. While results from target
profiling, VirtualToxLab screening and in vitro GR competition
neurite outgrowth assays supported this GR hypothesis,
additional GR transactivation and binding assays and further
docking studies did not support a direct interaction of WitA as
GR antagonist causing the neurite outgrowth effect. However,
we did confirm that WitA is able to reverse the neurite
outgrowth inhibitory activity of Dex, similar to GR antagonist
Mife. Nevertheless, a GR indirect effect of WitA by interfering
with GR-related targets such as ERK, Akt, NF-, TR, or
Hsp90 might be possible. It is possible that WitA affects
additional targets via a reversible binding mode, which we
could not detect with the current ABPP design and would
require the incorporation of photoreactive groups. Interactions
have been already established between these targets and Mife,
withanolide extract or withaferin A, and it is possible that we
identified GR as a positive hit for WitA in our pulldown assay
when WitA was actually modulating one of these targets instead
of GR directly. 24, 54, 58, 60, 76-86 Ongoing efforts are being made to
assess a potential connection through these targets. In addition,
we are pursuing additional studies to follow-up on the other 14
proteins-of-interest, identified in our ABPP profiling study.
(c = g/100 mL, solvent).
흀
4.2.
Isolation of withanolide A
Powdered Ashwagandha root93 (5 × 100 g + 10 × 250 g,
purchased from Yoga Vidya, Horn-Bad Meinberg, Germany)
was stirred in 500 g batches in a covered 2 L Erlenmeyer flask
with 1.25 L of methanol overnight. The brown slurry was then
transferred to a 10 cm diameter column with an additional 100
mL of methanol to rinse. The MeOH extract was allowed to
percolate out (drip out by gravity) and was collected (~2 h). An
additional 650 mL of methanol was added to the column, and
allowed to percolate further (~3 h). The methanol remaining in
the wet root powder was flushed out using nitrogen flow. The
extract was carefully concentrated in vacuo. This procedure was
repeated five times, until extract was obtained for all 3 kg of the
Ashwangandha root.
The combined methanol extract (combined volume of ~1.5
L) was washed with pentane (3 × 500 mL) to remove grease and
other apolar compounds, and was then carefully concentrated in
vacuo to afford a thick, brown oil. To remove the more polar
substances, this oil was taken up in EtOAc (500 mL), the
organic layer was separated and the aqueous layer was washed
2 more times with EtOAc (200 mL). The combined EtOAc
extracts were then washed with H2O (3 × 500 mL), dried over
Na2SO4, filtered and concentrated in vacuo to afford 17.4 g of a
brown oil. All aqueous extracts were checked for withanolide
A by TLC to make sure none was still present before disposal.
4. Experimental
4.1.
General
Commercial reagents and solvents were purchased from
Sigma-Aldrich, Alfa Aesar, Fisher Scientific, Acros Organics,
or Fluorochem and used as received, unless otherwise noted.
Deuterated solvents were purchased from Cambridge Isotope
Laboratories, Inc. (Andover, MA, USA). Flash chromatography
was carried out on Silicycle 60 Å silica gel (4063 m,
230400 mesh. Eluents and Rf’s are indicated. Analytical thin
layer chromatography was performed using Merck silica gel 60
F254 plates (0.25mm thickness) and was visualized by UV light
and fluorescent indicator, with or without heating (potassium
permanganate, p-anisaldehyde or aqueous ceric ammonium
molybdate stain).
The crude oil was purified by flash column
chromatography (FCC, wet load) with 70% to 100%
EtOAc/pentanes, and then 5% and 20% MeOH/EtOAc, to
afford 5.68 g of a yellow solid (collected during 80-100%
EtOAc/pentane and 5% MeOH/EtOAc eluents). This solid was
purified again by FCC with 40:1 DCM:MeOH to afford 2.1g of
a yellow/white solid, which was then triturated with cold ether
to afford 1.6 g of a white solid. A series of additional columns
with 100X-250X silica gel, with
either 8:1:1
1
EtOAc:CHCl3:Hexane (dry load), or 40:1 DCM:MeOH (dry
load), or 80% EtOAc/Hex (wet load with CHCl3) as eluents. A
total of 1.03 g of pure Withanolide A was obtained as white
solid, as shown in Figure S1 in 0.038% yield (see supporting
information). Also noteworthy, approximately 200 mg of
Withanolide B (S1), or C20-dehydroxywithanolide A, was
All H spectra were recorded on a Bruker Avance III
400MHz (400 MHz for H and 101 MHz for 13C) or Bruker
1
1
Avance III 500MHz spectrometer (500 MHz for H and 126
MHz for 13C). Chemical shifts (-values) are reported in parts
per million (ppm) relative to CDCl3 (7.26 ppm for 1H NMR and
77.0 ppm for 13C NMR). H NMR spectra are tabulated in the
1
obtained as
a white solid; WitB Rf = 0.65 (8:1:1
order: chemical shift, multiplicity, coupling constant (J in Hz),
number of protons. Multiplicity is reported as follows: s =
singlet, d = doublet, t = triplet, q = quartet, m = multiplet or
unresolved; ap = apparent, br = broad.
EtOAc/hexane/CHCl3).94 Analytical data of Withanolide A (1):
[휶]ퟐퟓ°푪
= +87.9° (c = 0.21, CHCl3) [Lit: +86.8 when c = 1.15,
푫
휈
CHCl3]. IR (film) = 3437, 2984, 2937, 1703, 1679, 1446,
1374, 1295, 1220, 1188, 1137, 1099, 1073, 1028, 955, 909, 883,
Infrared spectra were recorded on a Varian 800 FT-IR ATR
spectrophotometer and values are reported as frequency of
1
815, 800, 727, 648, 612 cm-1. H NMR (400 MHz, CDCl3) δ
6.59 (ddd, J = 10.1, 5.2, 2.2 Hz, 1H), 5.85 (ddd, J = 10.1, 3.0,
7

1.0 Hz, 1H), 4.21 (dd, J = 13.4, 3.5 Hz, 1H), 3.32 (dd, J = 4.0,
2.1 Hz, 1H), 3.12 (d, J = 1.5 Hz, 1H), 3.04 (d, J = 3.9 Hz, 1H),
2.79 – 2.65 (m, 2H), 2.52 (ddd, J = 18.8, 5.2, 1.0 Hz, 1H), 2.43
– 2.35 (m, 1H), 2.37 (d, J = 1.3 Hz, 1H), 2.14 – 1.98 (m, 3H),
1.96 (s, 3H), 1.89 (s, 3H), 1.86 – 1.80 (m, 1H), 1.78 (ddd, J =
10.6, 10.6, 2.2 Hz, 1H), 1.61 – 1.34 (m, 7H), 1.32 (s, 3H), 1.18
(s, 3H), 0.96 (s, 3H). 13C NMR (126 MHz, CDCl3) δ 203.3,
166.2, 149.0, 139.8, 129.1, 122.2, 81.1, 75.2, 73.4, 57.5, 56.5,
54.6, 52.1, 51.1, 44.0, 40.5, 36.9, 35.7, 35.2, 31.8, 23.3, 21.9,
21.9, 21.2, 20.7, 14.9, 13.9, 12.6. HRMS (ESI): Exact mass
cald for C28H39O6 [M+H]+, 471.3. Found 471.3. 95
vacuum, and the resulting residue was dissolved in CH2Cl2.
Saturated NH4Cl (aqueous) solution was added and the aqueous
phase was extracted with CH2Cl2 (3×). The combined organic
phases were dried over Na2SO4, filtered and concentrated in
vacuo. The crude product was purified by flash column
chromatography under the indicated conditions to yield the
oxime ethers 57.
(R)-6-((R)-1-hydroxy-1-
((4aS,4bS,6aS,7S,9aS,9bS,9cS,10aS,10bR,E)-10b-hydroxy-
4a,6a-dimethyl-4-((prop-2-yn-1-yloxy)imino)hexadecahydro-
1H-cyclopenta[1,2]phenanthro[9,10-b]oxiren-7-yl)ethyl)-3,4-
dimethyl-5,6-dihydro-2H-pyran-2-one (5): saturated C1
propargyl oxime 5 was made according to general procedure A.
The reaction was performed on 21 μmol scale of saturated
withanolide A (S3, see supporting information for synthetic
4.3.
Syntheses of molecular probes 47
(μ4aR,4bS,6aS,7S,9aS,9bS,9cS,10aS,10bR)-7-((R)-1-((R)-
4,5-dimethyl-6-oxo-3,6-dihydro-2H-pyran-2-yl)-1-
hydroxyethyl)-10b-hydroxy-4a,6a-dimethylhexadecahydro-
details) for
7
hours. Purification by flash column
4H-cyclopenta[1,2]phenanthro[9,10-b]oxiren-4-one
(4):
chromatography was performed with silica gel (150X by
weight) with 120:1 CH2Cl2:MeOH as the eluent, and afforded 5
as a white solid in a 75% yield. Analytical data for 5: TLC Rf =
0.64 (50% EtOAc/pentanes, CAM stain); [Rf (starting material
withanolide A hydroxylamine S2 was prepared according to a
modified procedure (see supporting information for details).96
Withanolide A hydroxylamine S2 (10 mg, 0.021 mmol) was
taken up in DMSO (100 L, 0.2 M) in a round bottom flask with
magnetic stir bar under argon. KOH pellets (4 mg, 0.062 mmol,
3 equiv.) were added and dissolved to give a milky yellow
solution. Then, added propargyl bromide (10 L, 0.062 mmol,
80% by weight in toluene, 3 equiv.) and the solution turned
brown. The reaction stirred for 3 hours and then added 1M HCl
solution. The reaction mixture was extracted with DCM (3×),
and then dried the combined organic layers over Na2SO4,
filtered and concentrated in vacuo to afford a yellow oil. The
crude product was purified by flash column chromatography
with silica gel (×200 by weight) and with 80:1 DCM:MeOH as
the eluent to afford the C1 propargyl oxime as a clear oil (6.3
mg, 58% yield). Analytical data for 4: TLC Rf = 0.67 (8:1:1
[휶]ퟐퟒ°푪
[휶]ퟐퟒ°푪
1) = 0.49 (50% EtOAc/pentanes)]. mp = 200202 °C.
푫
휈
= +93.4° (c = 0.14, CHCl3). IR (film) = 3267, 2944, 2361,
2338, 2166, 1703, 1441, 1384, 1314, 1291, 1178, 1133, 1095,
1
1041, 1002, 896, 761, 686, 631 cm-1. H NMR (500 MHz,
CDCl3) δ 4.66 – 4.57 (m, 2H), 4.21 (dd, J = 13.4, 3.6 Hz, 1H),
3.30 – 3.22 (m, 2H), 2.99 (d, J = 4.0 Hz, 1H), 2.92 (s, 1H), 2.61
(dddd, J = 9.2, 2.9, 2.9, 2.9 Hz, 1H), 2.40 (dd, J = 2.4, 2.4 Hz,
1H), 2.39 – 2.32 (m, 2H), 2.11 (dd, J = 17.4, 2.9 Hz, 1H), 2.08
– 1.96 (m, 3H), 1.95 (s, 3H), 1.95 – 1.90 (m, 1H), 1.89 (s, 3H),
1.87 – 1.82 (m, 1H), 1.79 (ddd, J = 10.5, 10.5, 2.3 Hz, 1H), 1.75
– 1.33 (m, 8H), 1.32 (s, 3H), 1.21 – 1.13 (m, 1H), 1.11 (s, 3H),
0.96 (s, 3H). 13C NMR (101 MHz, CDCl3) δ 166.3, 162.4,
149.0, 122.2, 81.2, 80.7, 75.3, 73.8, 72.5, 61.0, 57.6, 57.3, 54.5,
52.3, 47.3, 44.4, 40.6, 36.9, 35.3, 32.6, 31.9, 23.2, 22.1, 22.1,
21.3, 20.7, 20.4, 20.2, 16.2, 14.1, 12.6. HRMS (ESI): Exact
mass cald for C31H44NO6 [M+H]+, 526.3. Found 526.3.
EtOAc/hexane/CHCl3, CAM stain).
= +171.5° (c =
푫
휈
0.275, CHCl3) IR (film) = 3526, 3279, 2941, 2919, 2359,
2325, 2212, 2050, 1985, 1705, 1446, 1381, 1317, 1290, 1130,
1095, 1039, 1003, 906, 876, 825, 788, 651, 620 cm-1. 1H NMR
(400 MHz, CDCl3) δ 6.67 (ddd, J = 10.3, 2.7, 1.1 Hz, 1H), 6.01
(ddd, J = 10.3, 4.7, 2.5 Hz, 1H), 4.64 (dd, J = 2.4, 1.1 Hz, 2H),
4.21 (dd, J = 13.3, 3.5, 1H), 3.31 (dd, J = 3.7, 2.0 Hz, 1H), 3.01
(d, J = 3.9 Hz, 1H), 2.96 (d, J = 2.6 Hz, 1H), 2.91 – 2.82 (m,
1H), 2.52 – 2.32 (m, 5H), 2.16 – 1.98 (m, 3H), 1.95 (s, 3H), 1.89
(s, 3H), 1.87 – 1.77 (m, 2H), 1.63 – 1.34 (m, 7H), 1.33 (s, 3H),
1.09 (s, 3H), 0.98 (s, 3H). 13C NMR (126 MHz, CDCl3) δ 166.2,
159.2, 148.9, 131.4, 122.2, 118.3 81.2, 80.3, 75.3, 74.0, 71.4,
61.4, 57.5, 56.8, 54.6, 52.4, 45.5, 44.1, 40.6, 37.4 36.7, 35.4,
31.9, 23.4, 22.7, 22.0, 21.3, 20.7, 16.1, 14.0, 12.6. HRMS
(ESI): Exact mass cald for C31H41NO6 [M+H]+, 524.3. Found
524.3.
(4aR,4bS,6aS,7S,9aS,9bS,9cS,10aS,10bR)-7-((R)-1-((R)-4,5-
dimethyl-6-oxo-3,6-dihydro-2H-pyran-2-yl)-1-hydroxyethyl)-
10b-hydroxy-4a,6a-dimethyl-
1,4a,4b,5,6,6a,7,8,9,9a,9b,9c,10a,10b-tetradecahydro-2H-
cyclopenta[1,2]phenanthro[9,10-b]oxiren-2-one (10): C3 enone
(10) was accessed through our previously published
procedure95b from withanolide A (1) on a 0.21 mmol scale:
through the epoxy ketone (S5, 88% yield) intermediate, and a
Wharton transposition to allylic alcohol (S6, 65% yield). Then,
powdered 4Å molecular sieves (200 mg, 40X by weight) were
added to the solution of allylic alcohol S6 (25 mg, 52.7 μmol)
in CH2Cl2 (4 mL, 0.013 M) in an oven dried round bottom flask
with magnetic stir bar. This mixture stirred for 5 minutes at
room temperature under argon, when pyridinium dichromate
(36 mg, 94.1 μmol, 1.8 equiv.) was added and the mixture
stirred at room temperature for 2 hours and 20 minutes. The
mixture was then filtered over a celite pad, washing with
CH2Cl2. The combined organic layer was successively washed
with 1 N HCl (2 × 30 mL,) and brine (60 mL), and then dried
over Na2SO4, filtered and concentrated in vacuo. The crude
material was purified by flash column chromatography (eluent:
General procedure A for oxime ethers 5–7
To a solution of ketone (1.0 equiv.) in pyridine (freshly
distilled over KOH pellets, 1.0 equiv.) was added the O-
propargylhydroxylamine hydrochloride (S4, 3.0 equiv, see
supporting information for synthetic details of S4) in an oven-
dried round-bottom flask with stir bar (equipped with oven-
dried reflux condenser), and the mixture was heated at 70 °C for
the indicated time. The solvent was then removed under high
8

80:1 CH2Cl2/MeOH) to obtain enone 10 (21.6 mg, 45.9 μmol,
87% yield) as a white solid.
1H), 4.68 (dd, J = 2.4, 1.0, 2H), 4.20 (dd, J = 13.4, 3.5 Hz, 1H),
3.33 (dd, J = 3.1, 1H), 3.14 – 3.05 (m, 2H), 2.84 (d, J = 16.2 Hz,
1H), 2.66 (d, J = 16.0 Hz, 1H), 2.46 (dd, J = 2.4 Hz, 1H), 2.44
– 2.33 (m, 2H), 2.20 – 2.10 (m, 3H), 1.95 (s, 3H), 1.89 (s, 3H),
1.87 – 1.75 (m, 2H), 1.69 – 1.55 (m, 2H), 1.52 – 1.35 (m, 6H),
1.32 (s, 3H), 1.02 (s, 3H), 0.96 (s, 3H). 13C NMR (126 MHz,
CDCl3) δ 166.1, 151.4, 148.9, 143.0, 122.2, 115.9, 81.1, 80.1,
75.2, 74.4, 70.1, 61.5, 57.6, 57.0, 54.4, 51.7, 44.3, 42.8, 40.1,
36.9, 35.9, 35.3, 31.8, 29.8, 23.1, 22.2, 21.2, 20.9, 20.7, 20.7,
13.8, 12.6. HRMS (ESI): Exact mass cald for C31H42NO6
[M+H]+, 524.3. Found 524.4.
Analytical data for 10: TLC Rf = 0.5 (EtOAc). mp =
[휶]ퟐퟓ°푪
휈
= +71.1° (c = 0.385, CHCl3). IR (film)
252255 °C.
푫
= 3394, 2984, 2946, 1693, 1658, 1471, 1385, 1321, 1296, 1269,
1187, 1182, 1136, 1116, 1064, 1019, 986, 916, 815, 762, 745
cm-1. 1H NMR (500 MHz, CDCl3) δ 6.97 (d, J = 10.3 Hz, 1H),
5.99 (dd, J = 10.2, 0.9 Hz, 1H), 4.21 (dd, J = 13.4, 3.5 Hz, 1H),
3.36 (dd, J = 3.9, 2.5 Hz, 1H), 3.16 (d, J = 2.0 Hz, 1H), 3.09 (d,
J = 3.8 Hz, 1H), 2.85 (dd, J = 17.0, 2.1 Hz, 1H), 2.66 (dd, J =
17.0, 1.0 Hz, 1H), 2.45 – 2.32 (m, 2H), 2.25 – 1.97 (m, 3H),
1.96 (s, 3H), 1.89 (s, 3H), 1.88 – 1.81 (m, 2H), 1.72 – 1.34 (m,
8H), 1.32 (s, 3H), 1.16 (s, 3H), 0.98 (s, 3H). 13C NMR (126
MHz, CDCl3) δ 196.8, 166.1, 153.1, 148.9, 128.1, 122.2, 81.1,
75.2, 72.1, 57.7, 56.7, 54.4, 51.7, 46.6, 44.3, 42.8, 40.1, 35.7,
35.3, 31.8, 23.1, 22.2, 21.3, 20.8, 20.7, 20.5, 13.9, 12.6. HRMS
(ESI): Exact mass cald for C28H38NaO6 [M+Na]+, 493.3. Found
493.5.
(R)-6-((R)-1-hydroxy-1-
((4aR,4bS,6aS,7S,9aS,9bS,9cS,10aS,10bR,Z)-10b-hydroxy-
4a,6a-dimethyl-2-((prop-2-yn-1-yloxy)imino)hexadecahydro-
1H-cyclopenta[1,2]phenanthro[9,10-b]oxiren-7-yl)ethyl)-3,4-
dimethyl-5,6-dihydro-2H-pyran-2-one (7): Saturated C3
propargyl oxime 7 was made according to general procedure A.
The reaction was performed on 8 μmol scale of ketone S7 (for
details of S7, see supporting information), with a reaction time
of 5 hours. The 3:2 ratio of Z:E isomers could be separated by
flash column chromatography purification with silica gel
(1000X by weight) with 120:1 and then 80:1 CH2Cl2:MeOH as
the eluent. Saturated C3 propargyl oxime 7 (4.1 mg, 7.8 μmol,
>99% yield) was obtained as a white solid. Only the (E)-7
isomer was used in the biological studies. Analytical data for
(Z)-7: TLC Rf = 0.55 (50% EtOAc/pentanes, CAM stain). mp
[휶]ퟐퟒ°푪
(R)-6-((R)-1-hydroxy-1-
((4aR,4bS,6aS,7S,9aS,9bS,9cS,10aS,10bR,E)-10b-hydroxy-
4a,6a-dimethyl-2-((prop-2-yn-1-yloxy)imino)-
2,4a,4b,5,6,6a,7,8,9,9a,9b,9c,10a,10b-tetradecahydro-1H-
cyclopenta[1,2]phenanthro[9,10-b]oxiren-7-yl)ethyl)-3,4-
dimethyl-5,6-dihydro-2H-pyran-2-one (6): C3 propargyl oxime
6 was made according to general procedure A. The reaction was
performed on 12 μmol scale of enone 10, with a reaction time
of 3 days. Purification by flash column chromatography was
performed with silica gel (1000X by weight) with 120:1 and
then 80:1 CH2Cl2:MeOH as the eluent, and afforded (E)-6 (4.6
mg, 8.8 μmol, 74% yield) as a white solid. If the reaction time
was reduced to 4 hours, then both oxime diastereomers were
obtained (86% yield of 1:1 E:Z ratio on 10 μmol scale of enone
10). Only the (E)-6 isomer was used in the biological studies.
Analytical data for (E)-6: TLC Rf = 0.67 (50% EtOAc/pentanes
or 8:1:1 EtOAc/hexane/CHCl3, CAM stain). mp = 185-187 °C
[휶]ퟐퟓ°푪
= 200-204 °C (decomposition).
= +65.6° (c = 0.25,
푫
휈
CHCl3). IR (film) = 3531, 3287, 2943, 2361, 2338, 2111,
2051, 1985, 1705, 1462, 1425, 1335, 1319, 1287, 1181, 1133,
1098, 1047, 1006, 896, 858, 784, 735, 631 ꢀcm-1. 1H NMR (400
MHz, CDCl3) δ 4.72 – 4.58 (m, 2H), 4.20 (dd, J = 13.3, 3.5 Hz,
1H), 3.33 – 3.26 (m, 1H), 3.20 (d, J = 15.6 Hz, 1H), 3.01 (d, J
= 3.8 Hz, 1H), 2.86 (d, J = 2.1 Hz, 1H), 2.54 – 2.41 (m, 1H),
2.45 (dd, J = 2.4, 2.4 Hz, 1H), 2.40 – 2.31 (m, 3H), 2.28 (dd, J
= 15.7, 2.1 Hz, 1H), 2.15 – 1.98 (m, 3H), 1.95 (s, 3H), 1.89 (s,
3H), 1.86 – 1.64 (m, 3H), 1.62 – 1.53 (m, 1H), 1.48 (dd, J = 9.8,
9.8 Hz, 1H), 1.41 – 1.32 (m, 4H), 1.30 (s, 3H), 1.28 – 1.23 (m,
2H), 1.17 – 1.06 (m, 1H), 0.93 (s, 3H), 0.88 (s, 3H).ꢀ 13C NMR
(101 MHz, CDCl3) δ 166.1, 158.7, 148.9, 122.2, 81.1, 80.3,
75.2, 74.2, 70.5, 61.1, 58.0, 57.8, 54.5, 51.8, 44.2, 40.2, 38.5,
38.4, 34.9, 33.8, 31.8, 28.8, 26.4, 23.2, 22.2, 21.2, 20.7, 20.3,
15.2, 13.8, 12.6. HRMS (ESI): Exact mass cald for C31H44NO6
[M+H]+, 526.3. Found 526.3.
(decomposition).
= +81.4° (c = 0.5, CHCl3). IR (film)
푫
휈
= 3518, 3303, 2939, 2360, 1703, 1461, 1386, 1319, 1287,
1132, 1099, 1043, 1007, 914, 731, 682, 648 cm-1. 1H NMR (400
MHz, CDCl3) δ 6.27 (d, J = 10.2 Hz, 1H), 6.18 (d, J = 10.1 Hz,
1H), 4.71 (d, J = 2.4 Hz, 2H), 4.20 (dd, J = 13.3, 3.4 Hz, 1H),
3.33 (dd, J = 2.9, 2.9 Hz, 1H), 3.18 – 3.08 (m, 2H), 3.04 (s, 1H),
2.63 (d, J = 17.7 Hz, 1H), 2.49 (dd, J = 2.4, 2.4 Hz, 1H), 2.43-
2.33 (m, 2H), 2.19 – 1.99 (m, 3H), 1.95 (s, 3H), 1.89 (s, 3H),
1.86 – 1.78 (m, 2H), 1.68 – 1.55 (m, 3H), 1.50 – 1.37 (m, 4H),
1.31 (s, 3H), 0.98 (s, 3H), 0.96 (s, 3H). 13C NMR (101 MHz,
CDCl3) δ 166.1, 154.4, 148.9, 140.1, 122.9, 122.2, 81.1, 77.4,
75.2, 74.7, 69.4, 61.8, 57.7, 57.1, 54.4, 51.8, 44.3, 42.1, 40.2,
35.9, 35.2, 33.6, 31.8, 23.1, 22.2, 21.3, 20.7, 20.7, 20.5, 13.8,
12.6. HRMS (ESI): Exact mass cald for C31H42NO6 [M+H]+,
524.3. Found 524.3.
Analytical data for (E)-7: TLC Rf = 0.46 (50%
EtOAc/pentanes, CAM stain). mp
=
199202 °C
[휶]ퟐퟒ°푪
(decomposition).
= +36.7° (c = 0.15, CHCl3). IR (film)
푫
휈
= 2938, 2361, 2341, 2115, 2048, 1985, 1705, 1458, 1425,
1385, 1319, 1290, 1183, 1099, 1046, 1008, 895, 760, 732, 685,
631 ꢀcm-1. 1H NMR (400 MHz, CDCl3) δ 4.71 – 4.59 (m, 2H),
4.20 (dd, J = 13.3, 3.5 Hz, 1H), 3.29 (dd, J = 3.9, 2.2 Hz, 1H),
3.18 (dd, J = 17.1, 6.1 Hz, 1H), 2.98 (d, J = 3.8 Hz, 1H), 2.86
(d, J = 1.9 Hz, 1H), 2.56 – 2.32 (m, 5H), 2.16 – 1.99 (m, 4H),
1.95 (s, 3H), 1.89 (s, 3H), 1.88 – 1.73 (m, 2H), 1.66 – 1.53 (m,
2H), 1.49 (dd, J = 9.6, 9.6 Hz, 1H), 1.44 – 1.34 (m, 2H), 1.30
(s, 3H), 1.27– 1.19 (m, 2H), 1.16 – 1.07 (m, 1H), 0.93 (s, 3H),
0.89 (s, 3H).ꢀ 13C NMR (101 MHz, CDCl3) δ 166.1, 158.4,
148.9, 122.2, 81.1, 80.3, 75.2, 74.2, 70.4, 61.1, 57.9, 57.6, 54.5,
Analytical data for (Z)-6: Also isolated as a white solid.
TLC Rf
=
0.58 (50% EtOAc/pentanes or 8:1:1
[휶]ퟐퟓ°푪
EtOAc/hexane/CHCl3, CAM stain). mp = 118-120 °C.
푫
휈
= +75.6° (c = 0.34, CHCl3). IR (film) = 3521, 3295, 2929,
2362, 2340, 1704, 1461, 1427, 1383, 1319, 1288, 1133, 1099,
1043, 1007, 972, 917, 816, 763, 731, 647 cm-1. 1H NMR (400
MHz, CDCl3) δ 6.75 (d, J = 10.3 Hz, 1H), 6.30 (d, J = 10.4 Hz,
9

51.8, 44.2, 40.2, 38.9, 38.4, 38.3, 34.9, 31.8, 27.9, 23.2, 22.2,
21.2, 20.7, 20.6, 20.4, 14.9, 13.8, 12.6.ꢀ HRMS (ESI): Exact
mass cald for C31H44NO6 [M+H]+, 526.3. Found 526.3.
2.5, 2.5 Hz, 1H), 2.14 – 2.03 (m, 2H), 1.99 (s, 3H), 1.98 – 1.88
(m, 1H), 1.91 (s, 3H), 1.89 – 1.82 (m, 1H), 1.77 – 1.65 (m, 2H),
1.61 (dd, J = 11.0, 8.6 Hz, 1H), 1.54 – 1.47 (m, 1H), 1.45 – 1.40
(m, 2H), 1.38 (s, 3H), 1.36 – 1.22 (m, 2H), 1.21 (s, 3H), 0.90 (s,
3H). 13C NMR (101 MHz, CDCl3) δ 204.4, 168.7, 151.5,
141.3, 141.0, 138.9, 127.1, 125.9, 124.7, 124.2, 78.1, 75.7, 75.0,
59.7, 54.8, 52.4, 52.0, 51.8, 51.0, 43.5, 40.1, 37.4, 35.1, 28.9,
23.5, 23.4, 21.3, 19.6, 16.7, 16.1, 14.0. HRMS (ESI): Exact
mass cald for C31H38NaO5 [M+Na]+, 513.3. Found 513.3.
4.4.
Syntheses of molecular probes 89
prop-2-yn-1-yl (S,2Z,4E)-6-hydroxy-6-
((4aR,4bS,6aS,7S,9aS,9bS,9cS,10aS,10bR)-10b-hydroxy-4a,6a-
dimethyl-4-oxo-4,4a,4b,5,6,6a,7,8,9,9a,9b,9c,10a,10b-
tetradecahydro-1H-cyclopenta[1,2]phenanthro[9,10-b]oxiren-
7-yl)-2,3-dimethylhepta-2,4-dienoate (8): To a solution of
withanolide A (1, 30 mg, 0.063 mmol) dissolved in CH2Cl2
(anhydrous, 7 mL, 0.009 M) in a round bottom flask with stir
bar at room temperature was added nBu4NI (3 mg, 0.006 mmol,
0.1 equiv.), a 50% NaOH solution (3.5 g in 7 mL H2O, 0.009
M) and then the propargyl bromide (8 μL, 0.070 mmol, 1.1
equiv.). The faint yellow reaction stirred for 3 days. Then,
added 5 mL of sat. NH4Cl solution (aq.) and brought the mixture
to pH of 7 with conc. HCl and 2M HCl. The aqueous layer was
extracted with EtOAc (3×). The combined organic layers were
washed with brine (1×), dried over MgSO4, filtered and
concentrated in vacuo to afford a yellow oil. This was then
purified by flash column chromatography with silica gel (×200)
with 20, 30 and 40% EtOAc/pentanes as the eluent. Collected
the unsaturated propargyl ester 9 (1.4 mg, 5% yield) as a yellow
oil, propargyl ester 8 (10.6 mg, 33% yield) as a white solid, and
starting material (8.4 mg, 28% mass recovery of 1).
4.5.
Single crystal X-ray analyses of (Z)-7 and 8
Selected crystallographic data of (Z)-7 are summarized as
follows: molecular formula C31H43NO6, Mw 525.69,
wavelength Cu Kα, (λ = 1.54178 Å), Monoclinic, P21, a:
6.3281(3) Å, b: 7.4242(4) Å, c: 28.7295(12) Å, α 90°, β
95.892(3)°, γ 90°, V = 1342.61 Å3, Dcacld = 1.300 mg/m-3, F(000
) = 568,  = 0.717 mm-1, T = 123 K.
Selected crystallographic data of 8 are summarized as
follows: molecular formula C31H40O6, Mw 508.65, wavelength
Cu Kα, (λ = 1.54178 Å), Orthorhombic, P212121, a: 6.2944(3)
Å, b: 11.7168(5) Å, c: 35.5369(17) Å, α 90°, β 90°, γ 90°, V =
2620.9(2) Å3, Dcacld = 1.289 mg/m-3, F(000 ) = 1096,  = 0.708
mm-1, T = 123 K.
Data collection was performed by Dr. Markus Neuburger
(University of Basel) on a Bruker Kappa APEX diffractometer.
by placing the crystal in the cold stream (123K) of Oxford
Cryosystems open-flow nitrogen cryostat with nominal stability
of 0.1K.97 The absorption process applied to the intensities was
SADABS.98 Integration of the frames and data reduction was
carried out using APEX2.99 The structures were solved by direct
methods using Superflip.100 All non-hydrogen atoms were
refined anisotropically by full-matrix least-squares on F using
CRYSTALS,101 and computing molecular graphics were done
using CAMERON.102 The data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
Analytical data for 8: TLC Rf
EtOAc/pentanes, CAM stain). mp = 156-158 °C.
=
0.44 (50%
[휶]ퟐퟓ°푪
= –
푫
휈
6.5° (c = 0.265, CHCl3). IR (film) = 3730, 3592, 3485, 3271,
2927, 2874, 2851, 2361, 2340, 2198, 2166, 1687, 1456, 1383,
1
1291, 1262, 1215, 1105, 1033, 975, 906, 785, 630 cm-1. H
NMR (400 MHz, CDCl3) δ 7.01 (d, J = 15.9 Hz, 1H), 6.59 (ddd,
J = 10.1, 5.1, 2.2 Hz, 1H), 6.01 (d, J = 15.8 Hz, 1H), 5.85 (dd,
J = 10.2, 2.1 Hz, 1H), 4.77 (d, J = 2.5 Hz, 1H), 3.30 (dd, J =
3.7, 2.0 Hz, 1H), 3.13 (d, J = 1.5 Hz, 1H), 3.03 (d, J = 3.9 Hz,
1H), 2.76 – 2.64 (m, 2H), 2.56 – 2.47 (m, 1H), 2.48 (dd, J = 2.5,
2.5 Hz, 1H), 2.12 – 2.02 (m, 2H), 1.99 (s, 3H), 1.98 – 1.93 (m,
1H), 1.91 (d, J = 1.1 Hz, 3H), 1.88 – 1.60 (m, 4H), 1.48 – 1.41
(m, 2H), 1.40 (s, 3H), 1.36 – 1.24 (m, 3H), 1.17 (s, 3H), 0.88 (s,
3H). 13C NMR (101 MHz, CDCl3) δ 203.3, 168.7, 141.3, 141.1,
139.7, 129.2, 125.8, 124.7, 78.1, 75.6, 74.9, 73.4, 59.7, 57.5,
56.4, 52.0, 51.9, 51.2, 43.7, 40.4, 36.9, 35.7, 35.2, 28.9, 23.3,
23.3, 21.8, 16.7, 16.1, 14.9, 14.2. HRMS (ESI): Exact mass
cald for C31H40NaO6 [M+Na]+, 531.3. Found 531.3.
4.6.
Affinity-based protein profiling tests
Human cells A549 were grown in DMEM medium (Sigma
Life Sciences) with 2 mM glutamine (PAA) and 10 % fetal
bovine serum (Sigma Life Sciences) for analytical (6 well) and
for preparative (TC dish 150 standard) studies, respectively. For
SILAC studies, SILAC medium (Gibco, Life Technologies)
with 2 mM glutamine (PAA), 10% dialysed fetal bovine serum
(Sigma Life Sciences) and isotope labelled amino acids (K0,
K4, K6, R0, R8, R10; euriso-top) were used. The cells were
grown to 70-80 % confluence. Cells were incubated for 1 h with
varying concentrations of probe (stock solution in DMSO, 0.1%
end concentration of DMSO) in a cell incubator at 37 °C.
Subsequently, the cells were scraped off and pelletized for 3
min at 600 g. The pellets were washed with PBS once and then
resuspended in lysis buffer (1% (v/v) NP40 and 1% (w/v)
sodium deoxycholate) and incubated at 0 °C for 15 min.
Membrane and cytosol were separated by centrifugation (20
min, 4 °C, 21·103 xG). The membrane fraction was washed
once with PBS (1 × 10 sec, 10% maximum intensity) and
resuspended in 200 µL PBS for analytical, 1000 µL for
preparative analysis.
prop-2-yn-1-yl
(S,2Z,4E)-6-
((4aR,4bS,6aS,7S,9aS,9bS,9cS,10aR)-4a,6a-dimethyl-4-oxo-
4a,4b,5,6,6a,7,8,9,9a,9b,9c,10a-dodecahydro-4H-
cyclopenta[1,2]phenanthro[9,10-b]oxiren-7-yl)-6-hydroxy-2,3-
dimethylhepta-2,4-dienoate (9): Analytical data for 9: TLC Rf
[휶]ퟐퟒ°푪
= 0.83 (50% EtOAc/pentanes, CAM stain).
= –59.6° (c
푫
휈
= 0.085, CHCl3). IR (film) = 3733, 3597, 3453, 3302, 2926,
2871, 2480, 2361, 2334, 1711, 1662, 1625, 1567, 1454, 1380,
1292, 1214, 1106, 1076, 973, 889, 767, 630 cm-1. 1H NMR (400
MHz, CDCl3) δ 7.01 (d, J = 15.9 Hz, 1H), 6.93 (dd, J = 9.7, 6.0
Hz, 1H), 6.42 (d, J = 5.9 Hz, 1H), 6.00 (d, J = 16.0 Hz, 1H),
5.96 (dd, J = 9.7, 0.8 Hz, 1H), 4.77 (d, J = 2.5 Hz, 2H), 3.57 (d,
J = 3.9 Hz, 1H), 3.31 (dd, J = 3.8, 1.4 Hz, 1H), 2.49 (dd, J =
10

4.7.
VirtualToxLab studies
slowly down the side of the well. The plates were incubated at
room temperature for 1–2.5 hours and then washed with PBS (1
× 500 L). The plates were then stained with a thin-layer of
Giemsa stain (modified solution, Sigma-Aldrich GS500, 100
L; previously filtered through a VWR 0.2 m PES sterile
syringe filter and diluted with MeOH) to visualize the cell
bodies. After 2 minutes, the stain was diluted with PBS (500
L) and, after gentle swirling for 2 minutes, the solution was
removed by aspiration. The plates were then washed with PBS
(1 × 500 L) and imaged under PBS by phase contrast
microscope (160× magnification, Leica DMI 4000B,
Germany). When it was necessary, the PBS was aspirated off
an additional time and replaced with a fresh layer.
Version 5.8 of VirtualToxLab was used for the data
presented in this paper. The simulations are conducted at the
atomic level and explicitly allow for mechanistic
a
interpretation of the results. The underlying protocol is
independent from any training data and makes the approach
universal with respect to the applicability domain. Overall, a
toxic potential of a compound is derived from the binding
affinities (computed by means of 4D Boltzmann scoring of all
binding modes identified through automated, flexible docking
to the 3D structure of the target protein) towards a series of 16
proteins known to trigger adverse effects and weighted using
the standard deviations of the individual predictions. The
individual target proteins screened include: Androgen receptor
(AR), Aryl hydrocarbon receptor (AhR), Cytochrome P450
1A2 (CYP1A2), Cytochrome P450 2C9 (CYP2C9),
Cytochrome P450 2D6 (CYP2D6), Cytochrome P450 3A4
(CYP3A4), Estrogen receptor  (ER), Estrogen receptor 
(ER), Glucocorticoid receptor (GR), hERG, Liver X receptor
(LXR), Mineralocorticoid receptor (MR), Peroxisome
All assays were performed in triplicate, with 611 images
randomly taken between the three wells, and at least 350 cells
counted per treatment per plate (three wells/plate per treatment).
Those cells having at least one neurite with a length of more
than 50 m were counted as differentiated (positive). The ratio
of neurite positive cells to total cells was evaluated with cell
counting tool in ImageJ, and error is represented by the standard
error of the mean (SEM). Large cell aggregates were not
counted. Plots were drawn using Microsoft Excel. The
significance of the results was examined by ANOVA and post
hoc tests (Dunnett’s Method and Tukey Kramer Method) using
Excel. All treatments were significantly different than the
DMSO control.
Proliferator-Activated receptor
 (PPAR), Progesterone
receptor (PR), Thyroid receptor  (TR), Thyroid receptor 
(TR). Except for AhR and hERG (for which homology models
are employed), all simulations are based on high-resolution
crystal structures from the PDB.
4.8.
Neurite outgrowth GR competition assay
4.9.
GR lysate binding bioassay
Human neuroblastoma SH-SY5Y cells (ATCC-CRL-
2266) were grown in DMEM/F-12 media (Dulbecco’s
Modified Eagle Medium F12 Nutrient Mixture HAM, with high
glucose and phenol red, without glutamine and HEPES; GIBCO
Life Technologies 21331-020) supplemented with 10% fetal
bovine serum (FBS South American CE, 10270106 Gibco Life
Technologies) and 1% Penicillin-Streptomycin (10,000 U/mL,
Life Technologies 15140-122) in canted neck culture flasks
with vented cap (75 cm2, Falcon 353136). Cells were passaged
at least three times before use in experiments; for this, 0.25%
Trypsin-EDTA solution (Sigma-Aldrich T3924) and
Dulbecco’s phosphate-buffered saline (DPBS, 1X, without
Ca/Mg, pH 7.2, Sigma-Aldrich D8537) were used.
Recombinant human glucocorticoid receptor  (GR)
Baculovirus stock was produced using Bac-to-Bac expression
system (Invitrogen). For lysate preparation cells were collected
in ice-cold PBS and centrifuged at 3000 g, 4 °C, 10 min,
resuspended in binding buffer (pH 7.4, 50 mM HEPES, 10 mM
EDTA, 50 mM sodium tungstate, 50% glycerol) and
homogenized using a glass puller with 50 strokes. The lysate
was centrifuged (16,000 g, 10 min, 4 °C) and supernatants
stored at 80 °C.
Recombinant human GR lysate was incubated in the
presence of 10 nM [1,2,4,6,7-3H]-Dexamethasone (American
Radiolabeled Chemicals, St. Louis, MO) and unlabeled
competitor (either 10 M Dex or 10 M respectively 1 M
Withanolide A) for 4 h at 16 °C. Free concentration of ligand
was separated by adding 5% dextran coated charcoal in binding
buffer and incubation on ice for 10 min following centrifugation
at 3200 g, 4 °C for 10 min. The GR bound fraction of
radioligand in supernatants was measured by scintillation
counting.
All experiments were performed under strictly sterile
conditions. The cells were plated at an optimal concentration of
5 × 105 cells/mL onto Collagen I, coated 24-well plates (Gibco
Life Technologies A11428-02) in the presence of the
corresponding compounds (1 M concentration, 0.1% DMSO)
to a total well volume of 1 mL, at 37 °C in a humidified
atmosphere (5% CO2 content). For the control experiments,
cells were treated with DMSO vehicle (0.1%) as the negative or
all-trans Retinoic acid (ATRA, Aldrich R2625) as a positive
control. The glucocorticoid receptor (GR) control compounds
used were dexamethasone (Dex, Sigma-Aldrich D4902) and
mifepristone (Mife, RU-486, Sigma-Aldrich M8046).
4.10. GR and MR transactivation bioassay
HEK-293 cells were seeded in poly-L-lysine-coated 24
well plates. Each well contained 80’000 cells cultivated in 500
μL DMEM supplemented with 10% FBS. After 24 h, cells were
transfected with 281 ng of TAT3 TATA Luc control plasmid,
and 188 ng of plasmid for human GR-α or MR expression using
FuGENE (Promega). 24 hours later, cells were washed twice
with steroid-free medium, followed by incubation for another
24 h in the presence of steroid hormone (10 nM aldosterone for
After a 6-day incubation, the media (1 mL total volume)
was aspirated from each of the 24-well plates and washed with
PBS (DPBS). The PBS was removed by aspiration, and the cells
were fixed with neutral buffered formaldehyde 4% solution
(10% formalin, Sigma-Aldrich HT5012; 500 L) dropped
11

MR, 100 nM cortisol for GR), receptor antagonist (1 µM
spironolactone for MR and 1 µM Mife for GR) or 1 µM of
Withanolide A. Cells were lysed in 60 µL of tropix lysis
solution (Applied Biosystems, Foster City, CA) supplemented
with 0.5 µM dithiothreitol.
K.R.B., and A. O. wrote the manuscript. All authors approved
the manuscript.
Notes
The authors declare no competing financial interest.
Luciferase activity was determined in 20 μl lysate adding
100 μl of D-luciferin-firefly substrate solution at a final
concentration of 0.47 mM D-luciferin, 53 mM ATP, 0.27 mM
coenzyme A, 0.13 mM EDTA, 33.3 mM dithiothreitol, 20 mM
tricine, pH 7.8, and 8 mM MgSO4. Samples were measured
using a SpectraMax-L luminometer (Molecular Devices,
Devon, UK). Luciferase activities were used to calculate fold
activation compared with vehicle control (0.05% DMSO).
ACKNOWLEDGMENT
We thank Dr. Ya-Chu Hsieh for critically proofreading and
formatting the manuscript. This research was supported in part
by a NIBR Postgraduate Fellowship, and a Marie Curie
International Incoming Fellowship within the 7th European
Community Framework Programme (FP7/2007-2013) under
REA grant agreement N° 623920 (to E.C.). Part of this research
was funded through the SNF (200020_163151). A.O., K.R.B.
and P.S. were supported by a grant from the Swiss National
Science Foundation No 31003A-179400. We thank the
Erasmus program, the Nora Baart Foundation (NBS), the Dr.
Saal van Zwanenbergstichting Fellowship, and the KNCV
Section Organic Chemistry (Wageningen University,
Netherlands) for their financial support of E.M. (visiting student
from the University of Amsterdam). We also thank the
Department of Chemistry at the University of Basel for
financial support of L.B. We are grateful to Prof. William Clegg
(University of Cambridge, UK) for X-ray crystallographic
support, Dr. Heinz Nadig for HRMS analysis (University of
Basel), and Dr. Thierry Langer for providing LigandScout
software (Inte:Ligand, Vienna, Austria).
4.11. GR docking studies
Docking was performed using the GOLD 5.2 software
(Chambridge Crystallographic Data Centre, Cambridge, UK).
This program applies a genetic algorithm for the identification
of accurate docking poses for small molecules into the binding
pocket of a protein. The crystal structure of GR with the PDB
entry 1M2Z was chosen because the initial screenings with the
VirtualToxLab were based on this structure. First, the co-
crystallized ligand Dexamethasone was docked to investigate
whether GOLD could restore the original binding orientation
and therefore validate the docking settings. The binding site was
centered on the carbonyl-oxygen of Asn564 (x 27.23; y 3.52; z
13.23) surrounded by a 9 Å sphere. GoldScore was selected as
scoring function. Interactions found for the docking solutions
between the ligands and the GR binding pocket were evaluated
using LigandScout 3.12. This program automatically analyzes
the observed interaction pattern between the docked ligand and
the protein, based on the chemical functionalities, the geometric
distances and the angles between neighboring structures.102
ABBREVIATIONS
GR, glucocorticoid receptor, NR3C1; WitA, withanolide A; Dex,
Dexamethasone; Mife, mifepristone, RU-486, RU38486; ATRA or
RA, all-trans retinoic acid; DMSO, dimethyl sulfoxide; BACE1,
beta-secretase 1; ADAM10, a disintegrin and metalloproteinase
domain-containing protein 10; SAR, structure–activity
relationship; QSAR, quantitative structure–activity relationship;
SILAC, stable isotope labeling of amino acids in cell culture; MR,
mineralocorticoid receptor; Akt, protein kinase B, PKB; ERK1/2,
extracellular regulated kinase 1/2; NGF, nerve growth factor; HD,
Huntington’s disease; AD, Alzheimer’s disease; PD, Parkinson’s
disease; GC, glucocorticoid; ABPP, activity-based protein
profiling; NGF, nerve growth factor; Cort, cortisol; Aldo,
aldosterone; Spiro, spironolactone; H-bonding, hydrogen bonding.
Appendix A. Supplementary Data
Supplementary
data
can
be
found
online
at
http://doi.org/xxx/j.bmc.xxx. The probe synthesis, NMR
spectra, characterization, target profiling, VirtualToxLab and
docking studies, and all bioassays are described in this article.
AUTHOR INFORMATION
Corresponding Author
REFERENCES
* Tel: +41 44 635 42 81. Email: karl.gademann@uzh.ch.
Homepage: http://www.chem.uzh.ch/research/gademann.html.
(1) Joyner, P. M.; Cichewicz, R. H. Bringing natural products
into the fold - exploring the therapeutic lead potential of
secondary metabolites for the treatment of protein-misfolding-
related neurodegenerative diseases. Nat Prod Rep. 2011; 28:
2647.
(2) EU Commission Public Health. Neurodegenerative
Disorders.
http://ec.europa.eu/health/major_chronic_diseases/diseases/bra
in_neurological/index_en.htm (accessed April 5, 2016).
(3) Alzheimer's Association. 2012 Alzheimer's Disease Facts
Author Contributions
E.C. and K.G. contributed the design of probes, withanolide A
isolation, WitA probe synthesis (with E.M.), initial analytical
target profiling experiments, GR competition and neurite
outgrowth experiments (with L.B.), target profiling data
interpretation, and are the main contributors to the manuscript,
and primary coordinators of the project. W.H., J.V., and S.S.
performed and all target profiling experiments and
corresponding data analysis. M.S. and A.V. contributed the
VirtualToxLab studies and data interpretation. K.R. B., P.S. and
A.O. contributed the GR/MR transactivation and binding
experiments, and docking studies. E.C., K. G., W. H., S. S.,
and
Figures.
http://www.alz.org/downloads/facts_figures_2012.pdf
(accessed April 5, 2016).
(4) McGovern Institute. Brain Disorders: By the Numbers.
12

https://mcgovern.mit.edu/brain-disorders/by-the-numbers
(accessed April 5, 2016).
(5) Hefti, F. Pharmacology of neurotrophic factors. Annu Rev
Lactone from Withania Coagulans. Phytochemistry. 1971; 10:
685688.
(20) Sangwan, R. S.; Chaurasiya, N. D.; Misra, L. N.; Lal, P.;
Uniyal, G. C.; Sharma, R.; Sangwan, N. S.; Suri, K. A.; Qazi,
G. N.; Tuli, R. Phytochemical variability in commercial herbal
products and preparations of Withania somnifera
(Ashwagandha). Curr Sci. 2004; 86: 461465.
(21) Kulkarni, S. K.; Dhir, A. Withania somnifera: an Indian
ginseng. Prog. Neuropsychopharmacol Biol Psychiatry. 2008;
32: 10931105.
(22) Chatterjee, S.; Srivastava, S.; Khalid, A.; Singh, N.;
Sangwan, R. S.; Sidhu, O. P.; Roy, R.; Khetrapal, C. L.; Tuli,
R. Comprehensive metabolic fingerprinting of Withania
somnifera leaf and root extracts. Phytochemistry. 2010; 71:
10851094.
(23) Kuboyama, T.; Tohda, C.; Zhao, J.; Nakamura, N.; Hattori,
M.; Komatsu, K. Axon- or dendrite-predominant outgrowth
induced by constituents from Ashwagandha. Neuroreport.
2002; 13: 17151720.
(24) Aggarwal, B. B.; Ichikawa, H.; Garodia, P.; Weerasinghe,
P.; Sethi, G.; Bhatt, I. D.; Pandey, M. K.; Shishodia, S.; Nair,
M. G. From traditional Ayurvedic medicine to modern
medicine: identification of therapeutic targets for suppression
of inflammation and cancer. Expert Opin Ther Targets. 2006;
10: 87118.
(25) Jain, S.; Shukla, S. D.; Sharma, K.; Bhatnagar, M.
Neuroprotective effects of Withania somnifera Dunn. in
hippocampal sub-regions of female albino rat. Phytother Res.
2001; 15: 544548.
(26) Soman, S.; Korah, P. K.; Jayanarayanan, S.; Mathew, J.;
Paulose, C. S. Oxidative stress induced NMDA receptor
alteration leads to spatial memory deficits in temporal lobe
epilepsy: ameliorative effects of Withania somnifera and
Withanolide A. Neurochem Res. 2012; 37: 19151927.
(27) Chen, L. X.; He, H.; Qiu, F. Natural withanolides: an
overview. Nat Prod Rep. 2011; 28: 705740.
Pharmacol Toxicol. 1997; 37: 239267.
(6) Kordower, J. H.; Emborg, M. E.; Bloch, J.; Ma, S. Y.; Chu,
Y.; Leventhal, L.; McBride, J.; Chen, E. Y.; Palfi, S.; Roitberg,
B. Z.; Brown, W. D.; Holden, J. E.; Pyzalski, R.; Taylor, M. D.;
Carvey, P.; Ling, Z.; Trono, D.; Hantraye, P.; Deglon, N.;
Aebischer, P. Neurodegeneration prevented by lentiviral vector
delivery of GDNF in primate models of Parkinson's disease.
Science. 2000; 290: 767773.
(7) Aebischer, P.; Ridet, J. Recombinant proteins for
neurodegenerative diseases: the delivery issue. Trends
Neurosci. 2001; 24: 533540.
(8) Dawbarn, D.; Allen, S. J. Neurotrophins and
neurodegeneration. Neuropathol Appl Neurobiol. 2003; 29:
211230.
(9) Tuszynski, M. H.; Thal, L.; Pay, M.; Salmon, D. P.; U, H.
S.; Bakay, R.; Patel, P.; Blesch, A.; Vahlsing, H. L.; Ho, G.;
Tong, G.; Potkin, S. G.; Fallon, J.; Hansen, L.; Mufson, E. J.;
Kordower, J. H.; Gall, C.; Conner, J. A phase 1 clinical trial of
nerve growth factor gene therapy for Alzheimer disease. Nat
Med. 2005; 11: 551555.
(10) Bulic, B.; Pickhardt, M.; Schmidt, B.; Mandelkow, E. M.;
Waldmann, H.; Mandelkow, E. Development of tau
aggregation inhibitors for Alzheimer's disease. Angew Chem Int
Ed. 2009; 48: 17401752.
(11) Tohda, C.; Kuboyama, T.; Komatsu, K. Search for natural
products related to regeneration of the neuronal network.
Neurosignals. 2005; 14: 3445.
(12) Wilson, R. M.; Danishefsky, S. J. Applications of total
synthesis to problems in neurodegeneration: Fascinating
chemistry along the way. Acc Chem Res. 2006; 39: 539549.
(13) Wilson, R. M.; Danishefsky, S. J. Small molecule natural
products in the discovery of therapeutic agents: The synthesis
connection. J Org Chem. 2006; 71: 83298351.
(14) Jessen, H. J.; Barbaras, D.; Hamburger, M.; Gademann, K.
Total Synthesis and Neuritotrophic Activity of Farinosone C
and Derivatives. Org Lett. 2009; 11: 34463449.
(15) Bonazzi, S.; Eidam, O.; Guttinger, S.; Wach, J. Y.; Zemp,
I.; Kutay, U.; Gademann, K. Anguinomycins and derivatives:
total syntheses, modeling, and biological evaluation of the
inhibition of nucleocytoplasmic transport. J Am Chem Soc.
2010; 132: 14321442.
(16) Wach, J. Y.; Gademann, K. Reduce to the Maximum:
Truncated Natural Products as Powerful Modulators of
Biological Processes. Synlett. 2012; 163170.
(17) Schmid, F.; Jessen, H. J.; Burch, P.; Gademann, K.
Truncated militarinone fragments identified by total chemical
synthesis induce neurite outgrowth. Med Chem Commun. 2013;
4: 135139.
(28) Zhang, H.; Timmermann, B. N. Withanolide Structural
Revisions by C NMR Spectroscopic Analysis Inclusive of the
gamma-Gauche Effect. J Nat Prod. 2016, 79, 732742.
(29) Choudhary, M. I.; Yousuf, S.; Nawaz, S. A.; Ahmed, S.;
Atta ur, R. Cholinesterase inhibiting withanolides from
Withania somnifera. Chem Pharm Bull. 2004; 52: 13581361.
(30) Riaz, N.; Malik, A.; Aziz ur, R.; Nawaz, S. A.;
Muhammad, P.; Choudhary, M. I. Cholinesterase-inhibiting
withanolides from Ajuga bracteosa. Chem Biodiversity. 2004;
1: 12891295.
(31) Machin, R. P.; Veleiro, A. S.; Nicotra, V. E.; Oberti, J. C.;
J, M. P. Antiproliferative activity of withanolides against
human breast cancer cell lines. J Nat Prod. 2010; 73: 966968.
(32) Zhang, H.; Samadi, A. K.; Cohen, M. S.; Timmermann, B.
N. Anti-proliferative withanolides from the Solanaceae: a
structure-activity study. Pure Appl Chem. 2012; 84: 1353
1367.
(18) Crane, E. A.; Gademann, K. Capturing Biological Activity
in Natural Product Fragments by Chemical Synthesis. Angew
Chem Int Ed. 2016; 55: 38823902.
(19) Subramanian, S. S.; Sethi, P. D.; Glotter, E.; Kirson, I.;
Lavie, D. 5,20(alpha)(R)-Dihydroxy-6(alpha),7(alpha)-Epoxy-
1-Oxo-(5(alpha)) Witha-2,24-Dienolide, New Steroidal
(33) Grover, A.; Shandilya, A.; Agrawal, V.; Bisaria, V. S.;
Sundar, D. Computational evidence to inhibition of human
acetyl cholinesterase by withanolide a for Alzheimer treatment.
J Biomol Struct Dyn. 2012; 29: 651662.
13

(34) Zhao, J.; Nakamura, N.; Hattori, M.; Kuboyama, T.;
Tohda, C.; Komatsu, K. Withanolide derivatives from the roots
of Withania somnifera and their neurite outgrowth activities.
Chem Pharm Bull. 2002; 50: 760765.
(35) Kuboyama, T.; Tohda, C.; Komatsu, K. Neuritic
regeneration and synaptic reconstruction induced by
withanolide A. Br J Pharmacol. 2005; 144: 961971.
(36) Patil, S. P.; Maki, S.; Khedkar, S. A.; Rigby, A. C.; Chan,
C. Withanolide A and asiatic acid modulate multiple targets
associated with amyloid-beta precursor protein processing and
amyloid-beta protein clearance. J Nat Prod. 2010; 73: 1196
1202.
(37) Jana, C. K.; Hoecker, J.; Woods, T. M.; Jessen, H. J.;
Neuburger, M.; Gademann, K. Synthesis of withanolide A,
biological evaluation of its neuritogenic properties, and studies
on secretase inhibition. Angew Chem Int Ed. 2011; 50: 8407
8411.
(38) Liffert, R.; Hoecker, J.; Jana, C. K.; Woods, T. M.; Burch,
P.; Jessen, H. J.; Neuburger, M.; Gademann, K. Withanolide A:
synthesis and structural requirements for neurite outgrowth.
Chem Sci. 2013; 4: 28512857.
(39) (a) Gersch, M.; Kreuzer, J.; Sieber, S. A. Electrophilic
natural products and their biological targets. Nat Prod Rep.
2012; 29: 659682, Krysiak, J.; Breinbauer, R. Activity-Based
Protein Profiling for Natural Product Target Discovery. Top
Curr Chem. 2012; 324: 4384, (b) Bottcher, T.; Sieber, S. A.
beta-Lactams and beta-lactones as activity-based probes in
chemical biology. Med Chem Commun. 2012; 3: 408417, (c)
Pichler, C. M.; Krysiak, J. Breinbauer, R. Target identification
of covalently binding drugs by activity-based protein profiling
(ABPP). Bioorg. Med. Chem. 2016, 24, 3291-3303, and
references cited therein.
(40) Selected examples: (a) Yokota, Y.; Bargagna-Mohan, P.;
Ravindranath, P.P.; Kim, K.B.; Mohan, R. Development of
withaferin A analogs as probes of angiogenesis. Bioorg. Med.
Chem. Lett., 2006; 16, p. 2603-2607; (b) Bargagna-Mohan, P.;
Hamza, A.; Kim, Y.; Khuan, Y.; Ho, A.; Mor-Vaknin, N.;
Wendschlag, N.; Liu, J.; Evans, R.M.; Markovitz, D.M.; Zhan,
C.; Kim, K.B.; Mohan, R.. The Tumor Inhibitor and
Antiangiogenic Agent Withaferin A Targets the Intermediate
Filament Protein Vimentin. Chem. Biol., 2007; 14, 623-634; (c)
Falsey, R. R., Marron, M. T., Gunaherath, G. M. K. B.,
Shirahatti, N., Mahadevan, D., Gunatilaka, L., & Whitesell, L.,
Actin microfilament aggregation induced by withaferin A is
mediated by annexin II. Nat. Chem. Biol., 2006; 2, 33-38
(41) Grossman, E. A.; Ward, C. C.; Spradlin, J. N.; Bateman, L.
A.; Huffman, T. R.; Miyamoto, D. K.; Kleinman, J. I.; Nomura,
D. K. Covalent Ligand Discovery against Druggable Hotspots
Targeted by Anti-cancer Natural Products. Cell Chem. Biol.
2017; 24: 1368-1376.
(44) Tornoe, C. W.; Christensen, C.; Meldal, M.
Peptidotriazoles on solid phase: [1,2,3]-triazoles by
regiospecific copper(I)-catalyzed 1,3-dipolar cycloadditions of
terminal alkynes to azides. J Org Chem. 2002; 67: 30573064.
(45) Huisgen, R. Centenary Lecture
-
1,3-Dipolar
Cycloadditions. P Chem Soc London 1961; 0: 357396.
(46) See Supporting Information for X-ray crystallographic
data and analysis.
(47) Also see Supporting Information.
(48) Ong, S. E.; Blagoev, B.; Kratchmarova, I.; Kristensen, D.
B.; Steen, H.; Pandey, A.; Mann, M. Stable isotope labeling by
amino acids in cell culture, SILAC, as a simple and accurate
approach to expression proteomics. Mol Cell Proteomics. 2002;
1: 376386.
(49) Eirich, J.; Burkhart, J. L.; Ullrich, A.; Rudolf, G. C.;
Vollmar, A.; Zahler, S.; Kazmaier, U.; Sieber, S. A.
Pretubulysin derived probes as novel tools for monitoring the
microtubule network via activity-based protein profiling and
fluorescence microscopy. Mol BioSyst. 2012; 8: 20672075.
(50) Also see Supporting Information.
(51) For a complete list of the target profiling hits and original
data, including data processing, please see Supporting
Information.
(52) McEwen, B. S.; De Kloet, E. R.; Rostene, W. Adrenal
steroid receptors and actions in the nervous system. Physiol
Rev. 1986; 66: 11211188.
(53) Hayashi, R.; Wada, H.; Ito, K.; Adcock, I. M. Effects of
glucocorticoids on gene transcription. Eur J Pharmacol. 2004;
500: 5162.
(54) Buckingham, J. C. Glucocorticoids: exemplars of multi-
tasking. Br J Pharmacol. 2006; 147 Suppl 1: S258268.
(55) Terada, K.; Kojima, Y.; Watanabe, T.; Izumo, N.; Chiba,
K.; Karube, Y. Inhibition of Nerve Growth Factor-Induced
Neurite Outgrowth from PC12 Cells by Dexamethasone:
Signaling Pathways through the Glucocorticoid Receptor and
Phosphorylated Akt and ERK1/2. PLoS One 2014; 9: e93223.
(56) Mayer, J. L.; Klumpers, L.; Maslam, S.; de Kloet, E. R.;
Joels, M.; Lucassen, P. J. Brief treatment with the
glucocorticoid receptor antagonist mifepristone normalises the
corticosterone-induced reduction of adult hippocampal
neurogenesis. J Neuroendocrinol. 2006; 18: 629631.
(57) Oomen, C. A.; Mayer, J. L.; de Kloet, E. R.; Joels, M.;
Lucassen, P. J. Brief treatment with the glucocorticoid receptor
antagonist mifepristone normalizes the reduction in
neurogenesis after chronic stress. Eur J Neurosci. 2007; 26:
33953401.
(58) Hu, P.; Oomen, C.; van Dam, A. M.; Wester, J.; Zhou, J.
N.; Joels, M.; Lucassen, P. J. A single-day treatment with
mifepristone is sufficient to normalize chronic glucocorticoid
induced suppression of hippocampal cell proliferation. PLoS
One 2012; 7: e46224.
(59) Behl, C.; Trapp, T.; Skutella, T.; Holsboer, F. Protection
against oxidative stress-induced neuronal cell death--a novel
role for RU486. Eur J Neurosci. 1997; 9: 912920.
(60) McCullers, D. L.; Sullivan, P. G.; Scheff, S. W.; Herman,
J. P. Mifepristone protects CA1 hippocampal neurons following
traumatic brain injury in rat. Neuroscience. 2002; 109: 219
230.
(42) Heydenreuter, W.; Kunold, E.; Sieber, S. A. Alkynol
natural products target ALDH2 in cancer cells by irreversible
binding to the active site. Chem Commun. (Cambridge, U. K.)
2015; 51: 1578415787.
(43) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless,
K. B. A stepwise Huisgen cycloaddition process: Copper(I)-
catalyzed regioselective "ligation" of azides and terminal
alkynes. Angew Chem Int Ed. 2002; 41: 25962599.
14

(61) Ghoumari, A. M.; Dusart, I.; El-Etr, M.; Tronche, F.;
Sotelo, C.; Schumacher, M.; Baulieu, E. E. Mifepristone
(RU486) protects Purkinje cells from cell death in organotypic
slice cultures of postnatal rat and mouse cerebellum. Proc Natl
Acad Sci U. S. A. 2003; 100: 79537958.
(62) Llorens-Martin, M.; Trejo, J. L. Mifepristone prevents
stress-induced apoptosis in newborn neurons and increases
AMPA receptor expression in the dentate gyrus of C57/BL6
mice. PLoS One 2011; 6: e28376.
(78) See Supporting Information for the full statistical analysis,
including ANOVA and post hoc test results.
(79) See Supporting Information for the full statistical analysis.
(80) Necela, B. M.; Cidlowski, J. A. Crystallization of the
human glucocorticoid receptor ligand binding domain: a step
towards selective glucocorticoids. Trends Pharmacol Sci. 2003;
24: 5861.
(81) Schoch, G. A.; D'Arcy, B.; Stihle, M.; Burger, D.; Bar, D.;
Benz, J.; Thoma, R.; Ruf, A. Molecular switch in the
glucocorticoid receptor: active and passive antagonist
conformations. J Mol Biol. 2010; 395: 568577.
(82) Wyttenbach, A.; Arrigo, A. P. The Role of Heat Shock
Proteins during Neurodegeneration in Alzheimer's, Parkinson's
and Huntington's Disease [Online]; In: Madame Curie
Bioscience Database. Austin (TX): Landes Bioscience, 2000-
(63) Haynes, L. E.; Griffiths, M. R.; Hyde, R. E.; Barber, D. J.;
Mitchell, I. J. Dexamethasone induces limited apoptosis and
extensive sublethal damage to specific subregions of the
striatum and hippocampus: implications for mood disorders.
Neuroscience. 2001; 104: 5769.
(64) Vyas, S.; Maatouk, L. Contribution of Glucocorticoids and
Glucocorticoid
Receptors
to
the
Regulation
of
2003;
Available
from:
Neurodegenerative Processes. CNS Neurol Disord Drug
Targets. 2013; 12: 11751193.
http://www.ncbi.nlm.nih.gov/books/NBK6495/
April 29, 2014).
(accessed
(83) Ichikawa, H.; Takada, Y.; Shishodia, S.; Jayaprakasam, B.;
Nair, M. G.; Aggarwal, B. B. Withanolides potentiate
apoptosis, inhibit invasion, and abolish osteoclastogenesis
through suppression of nuclear factor-kappaB (NF-kappaB)
activation and NF-kappaB-regulated gene expression. Mol
Cancer Ther. 2006; 5: 14341445.
(65) Varadarajan, S.; Breda, C.; Smalley, J. L.; Butterworth, M.;
Farrow, S. N.; Giorgini, F.; Cohen, G. M. The transrepression
arm of glucocorticoid receptor signaling is protective in mutant
huntingtin-mediated neurodegeneration. Cell Death Differ.
2015; 22: 13881396.
(66) See Supporting Information for full details.
(84) Pratt, W. B.; Morishima, Y.; Murphy, M.; Harrell, M.
Chaperoning of glucocorticoid receptors. Handb Exp
Pharmacol. 2006; 172: 111138.
(67) Vedani, A.; Dobler, M.; Smiesko, M. VirtualToxLab - a
platform for estimating the toxic potential of drugs, chemicals
and natural products. Toxicol Appl Pharmacol. 2012; 261: 142
153.
(68) Vedani, A.; Dobler, M.; Hu, Z. Q.; Smiesko, M.
OpenVirtualToxLab-A platform for generating and exchanging
in silico toxicity data. Toxicol Lett. 2015; 232: 519532.
(69) Access to the VirtualToxLab is available free of charge for
academic institutions, public hospitals, governmental agencies,
regulatory bodies and non-profit organizations.
(85) Falsey, R. R.; Marron, M. T.; Gunaherath, G. M.;
Shirahatti, N.; Mahadevan, D.; Gunatilaka, A. A.; Whitesell, L.
Actin microfilament aggregation induced by withaferin A is
mediated by annexin II. Nat Chem Biol. 2006; 2: 3338.
(86) Kaileh, M.; Vanden Berghe, W.; Heyerick, A.; Horion, J.;
Piette, J.; Libert, C.; De Keukeleire, D.; Essawi, T.; Haegeman,
G. Withaferin A strongly elicits I kappa B kinase beta
hyperphosphorylation concomitant with potent inhibition of its
kinase activity. J Biol Chem. 2007; 282: 42534264.
(70) All results were obtained with version 5.8 of the
VirtualToxLab (March 2016).
(71) See Supporting Information for MD graphics.
(72) Jessen, H. J.; Schumacher, A.; Shaw, T.; Pfaltz, A.;
Gademann, K. A unified approach for the stereoselective total
synthesis of pyridone alkaloids and their neuritogenic activity.
Angew Chem Int Ed. 2011; 50: 42224226.
(73) Hoecker, J.; Liffert, R.; Burch, P.; Wehlauch, R.;
Gademann, K. Caged retinoids as photoinducible activators:
implications for cell differentiation and neurite outgrowth. Org
Biomol Chem. 2013; 11: 33143321.
(87) Singh, D.; Aggarwal, A.; Maurya, R.; Naik, S. Withania
somnifera inhibits NF-kappaB and AP-1 transcription factors in
human peripheral blood and synovial fluid mononuclear cells.
Phytother Res. 2007; 21: 905913.
(88) Xu, Y. M.; Marron, M. T.; Seddon, E.; McLaughlin, S. P.;
Ray, D. T.; Whitesell, L.; Gunatilaka, A. A. 2,3-
Dihydrowithaferin A-3beta-O-sulfate, a new potential prodrug
of withaferin A from aeroponically grown Withania somnifera.
Bioorg Med Chem. 2009; 17: 22102214.
(74) Burch, P.; Binaghi, M.; Scherer, M.; Wentzel, C.; Bossert,
D.; Eberhardt, L.; Neuburger, M.; Scheiffele, P.; Gademann, K.
Total synthesis of gelsemiol. Chem Eur J. 2013; 19,: 2589
2591.
(75) Burch, P.; Chicca, A.; Gertsch, J.; Gademann, K.
Functionally Optimized Neuritogenic Farinosone C Analogs:
SAR-Study and Investigations on Their Mode of Action. ACS
Med Chem Lett. 2014; 5: 172177.
(89) Ozorowski, G.; Ryan, C. M.; Whitelegge, J. P.; Luecke, H.
Withaferin A binds covalently to the N-terminal domain of
annexin A2. Biol Chem. 2012; 393: 11511163.
(90) Khedgikar, V.; Kushwaha, P.; Gautam, J.; Verma, A.;
Changkija, B.; Kumar, A.; Sharma, S.; Nagar, G. K.; Singh, D.;
Trivedi, P. K.; Sangwan, N. S.; Mishra, P. R.; Trivedi, R.
Withaferin A: a proteasomal inhibitor promotes healing after
injury and exerts anabolic effect on osteoporotic bone. Cell
Death Dis. 2013; 4: e778.
(76) Burch, P.; Schmid, F.; Gademann, K. Neuritogenic
surfaces using natural product analogs. Adv Healthcare Mater.
2014; 3: 14151419.
(91) Wijeratne, E. M.; Xu, Y. M.; Scherz-Shouval, R.; Marron,
M. T.; Rocha, D. D.; Liu, M. X.; Costa-Lotufo, L. V.; Santagata,
S.; Lindquist, S.; Whitesell, L.; Gunatilaka, A. A. Structure-
(77) See Supporting Information for the full statistical analysis.
15

activity relationships for withanolides as inducers of the cellular
heat-shock response. J Med Chem. 2014; 57: 28512863.
(92) Gu, M.; Yu, Y.; Gunaherath, G. M.; Gunatilaka, A. A.; Li,
D.; Sun, D. Structure-activity relationship (SAR) of
withanolides to inhibit Hsp90 for its activity in pancreatic
cancer cells. Invest New Drugs. 2014; 32: 6874.
(93) The quality and content of withanolide A in the
Ashwagandha root may vary. In general, it was found that
smaller, high quality packages contain relatively more
withanolide A.
(94) Structure confirmed by data from Misra, L.; Mishra, P.;
Pandey, A.; Sangwan, R. S.; Sangwan, N. S.; Tuli, R.
Withanolides from Withania somnifera roots. Phytochemistry.
2008; 69: 10001004.
(95) Matches isolation data from: (a) Subramanian, S. S.; Sethi,
P. D.; Glotter, E.; Kirson, I.; Lavie, D. 5,20(R)-dihydroxy-
6,7-epoxy-1-oxo-(5)witha-2,24-dienolide, a new steroidal
lactone from withania coagulans. Phytochemistry. 1971; 10:
685688.; and data in (b) Liffert, R.; Hoecker, J.; Jana, C. K.;
Woods, T. M.; Burch, P.; Jessen, H. J.; Neuburger, M.;
Gademann, K. Withanolide A: synthesis and structural
requirements for neurite outgrowth. Chem Sci. 2013; 4: 2851
2857. and (c) Jana, C. K.; Hoecker, J.; Woods, T. M.; Jessen,
H. J.; Neuburger, M.; Gademann, K. Synthesis of Withanolide
A, Biological Evaluation of Its Neuritogenic Properties, and
Studies on Secretase Inhibition. Angew Chem Int Ed. 2011; 50:
84078411.
(96) Trofimov, B. A.; Tarasova, O. A.; Sigalov, M. V.;
Mikhaleva, A. I. The base-catalysed rearrangement of O-
propargyl ketoximes to N-1-alkenyl acryl amides. Tetrahedron
Lett. 1995; 36: 91819184.
(97) Cosier, J.; Glazer, A. M. A Nitrogen-Gas-Stream Cryostat
for General X-ray Diffraction Studies. J Appl Cryst. 1986, 105
107.
(98) Siemens Industrial Automation, Inc. (1996). SADABS:
Area-Detector Absorption Correction; Madison, WI.
(99) Bruker Analytical X-ray Systems, Inc. Apex2 2006,
Version 2 User Manual, M86-E01078, ꢀMadison, WI.
(100) Palatinus, L.; Chapuis, G. J. SUPERFLIP – a computer
program for the solution of crystal structures by charge flipping
in arbitrary dimensions. J Appl Cryst. 2007; 40: 786790.
(101) Betteridge, P. W.; Carruthers, J. R.; Cooper, R. I.; Prout,
K.; Watkin, D. J. CRYSTALS version 12: software for guided
crystal structure analysis. J Appl Cryst. 2003; 36: 1487.
(102) Watkin, D. J.; Prout, C. K.; Pearce, L. J. (1996).
CAMERON, Chemical Crystallography Laboratory, Oxford,
UK.
(102) Wolber G,; Langer T. T. LigandScout:ꢀ 3-D
Pharmacophores Derived from Protein-Bound Ligands and
Their Use as Virtual Screening Filters. J Chem Inf Comput Sci.
2005; 45: 160169.
16

17