3828–3835
Nucleic Acids Research, 2019, Vol. 47, No. 8
Published online 6 March 2019
doi: 10.1093/nar/gkz153
Sequence specific suppression of androgen
receptor–DNA binding
in vivo
by a Py-Im polyamide
Alexis A. Kurmis and Peter B. Dervan
*
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA, 91125, USA
Received January 24, 2019; Revised February 20, 2019; Editorial Decision February 21, 2019; Accepted February 22, 2019
ABSTRACT
The crucial role of androgen receptor (AR) in prostate
cancer development is well documented, and its
inhibition is a mainstay of prostate cancer treat-
ment. Here, we analyze the perturbations to the AR
cistrome caused by a minor groove binding molecule
that is designed to target a sequence found in a
subset of androgen response elements (ARE). We
find treatment with this pyrrole-imidazole (Py-Im)
polyamide exhibits sequence selectivity in its repres-
sion of AR binding
in vivo
. Differentially changed loci
are enriched for sequences resembling ARE half-
sites that match the Py-Im polyamide binding pref-
erences determined
in vitro
. Comparatively, permu-
tations of the ARE half-site bearing single or dou-
ble mismatches to the Py-Im polyamide binding se-
quence are not enriched. This study confirms that the
in vivo
perturbation pattern caused by a sequence
specific polyamide correlates with its
in vitro
bind-
ing preference genome-wide in an unbiased manner.
INTRODUCTION
Transcription factors regulate cellular gene expression and
the loss of this regulatory balance can lead to a myriad of
genetic diseases including cancer. The role of androgen re-
ceptor (AR) in prostate cancer is one of the most well char-
acterized examples. Early work in 1941 by Charles Hug-
gins and Clarence Hodges showed that the progression of
prostate cancer can be controlled by androgen deprivation
through castration or hormonal therapy with estrogen (
1
).
Later the discovery of the first anti-androgen, cyproterone
acetate, allowed direct inhibition of androgen binding to
the AR (
2
). Since then, the AR has remained the primary
target for systemic therapeutics for prostate cancer patients
(
3
,
4
). In recent years, newer anti-androgens including enza-
lutamide and apalutamide have already been approved and
others are in late-stage clinical development (
5–7
).
Metastatic prostate cancers treated with androgen sup-
pressive therapy will ultimately progress to a disease
state termed castration-resistant prostate cancer (CRPC).
Second-line AR directed therapeutics, such as enzalu-
tamide, are often effective against CRPC, but a second dis-
ease progression is almost inevitable. Two mechanisms that
have been documented to confer resistance to second-line
AR directed therapies are mutations to the AR C-terminal
ligand-binding domain and expression of AR splice vari-
ants lacking the ligand-binding domain (
8–10
). Multiple
approaches have been explored to overcome these resistance
mechanisms, as reviewed recently by Jung
et al
.(
11
). These
include AR transcription activation domain inhibitors such
as EPI-506 and AR DNA-binding domain inhibitors, such
as pyrvinium pamoate (
11
). In addition, our lab has pre-
viously reported the use of DNA binders to allosterically
modulate the binding of AR at the protein–DNA interface
(
12
). We have shown this approach to be efficacious in sev-
eral prostate cancer models, including anti-androgen resis-
tant models (
13
,
14
).
Pyrrole-imidazole (Py-Im) polyamides are DNA minor
groove binding molecules with modular sequence specificity
that bind to target sites with affinities comparable to DNA-
binding proteins (
15
,
16
). Minor groove sequence recogni-
tion is determined by the pairing of N-methylimidazole (Im)
and N-methylpyrrole (Py); the target sequence of a partic-
ular polyamide is dependent on the location of the Im and
Py monomers within the hairpin structure (
17
). An Im
/
Py
pair will recognize a G
•
C pair in the DNA, Py
/
Im will rec-
ognize C
•
GandPy
/
Py will bind to either A
•
TorT
•
A(
18–
20
). Upon binding to the minor groove, Py-Im polyamides
cause an expansion of the minor groove and a correspond-
ing compression in the opposing major groove (
21
). Py-Im
polyamides have been shown to interfere with DNA depen-
dent processes such as gene expression, RNA polymerase
II elongation, DNA polymerase replication and topoiso-
merase activity (
13
,
22–24
). They have also been shown to
activate p53 and induce apoptosis without genotoxicity, and
to have antitumor activity in prostate cancer cell lines and
xenograft models (
13
,
14
,
23
).
ARE-1
is a Py-Im polyamide
designed to target the sequence 5
-WGWWCW-3
, where W
represents either A or T, which is found in a subset of an-
drogen response elements (ARE).
In this study, we evaluate the anti-proliferative effects of
ARE-1
in the setting of enzalutamide resistant LNCaP-95
cells, and in the context of AR signaling. We further exam-
*
To whom correspondence should be addressed. Tel: +1 626 395 6002 ; Fax: +1 626 683 8753; Email: dervan@caltech.edu
C
The Author(s) 2019. Published by Oxford University Press on behalf of Nucleic Acids Research.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http:
//
creativecommons.org
/
licenses
/
by
/
4.0
/
), which
permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
Downloaded from https://academic.oup.com/nar/article-abstract/47/8/3828/5369940 by California Institute of Technology user on 01 May 2019
Nucleic Acids Research, 2019, Vol. 47, No. 8
3829
ine the disruption pattern to the cistrome caused by
ARE-1
treatment. We find that at loci where AR binding is reduced
by
ARE-1
treatment, the consensus ARE motif bears closer
resemblance to the
ARE-1
target sequence, whereas the na-
tive consensus motif has more sequence degeneracy.
MATERIALS AND METHODS
Cell culture
The LNCaP-95 cell line was obtained from the laboratory
of Dr. Jun Luo at Johns Hopkins School of Medicine. The
cells were received at passage 3 and maintained in phenol
red free RPMI 1640 (Gibco 11835-030) with 10% charcoal
treated fetal bovine serum (CTFBS). All experiments were
performed below passage 20, and cells were validated to
parental cell line and confirmed mycoplasma free by ATCC
following experimentation.
Cell uptake
Cell uptake was confirmed by confocal imaging. Briefly,
LNCaP-95 cells were plated in 35-mm optical dishes (Mat-
Tek) at 7.5
×
10
4
cells per dish and allowed to adhere for
24 h. Cells were treated with 2
M
ARE-1-FITC
for 16 h,
washed with phosphate buffered saline (PBS) and imaged at
the Caltech Biological Imaging Facility using a Zeiss LSM
710 inverted laser scanning confocal microscope equipped
with a 63
×
oil immersion lens.
Cytotoxicity assay
LNCaP-95 cells were plated at 7.5
×
10
3
per well in 96
well plates. Cells were allowed to adhere for 24 h, and me-
dia was then replaced with fresh media containing vehi-
cle or polyamide
ARE-1
. After 72 h, an equivalent volume
of CellTiter-Glo (CTG) reagent (Promega) was added to
each well. Luminescence was allowed to stabilize for 10 min
at room temperature, according to manufacturer instruc-
tions, and then measured on a FlexStation3 plate reader
(Molecular Devices). Background subtracted luminescence
of polyamide treated cells was normalized to vehicle treated
cells, and non-linear regression analysis (Prism software,
Graphpad) was performed to determine IC
50
value.
Gene expression analysis by quantitative RT-PCR (qPCR)
LNCaP-95 cells were cultured for 24 h after plating in six
well plates at 7.5
×
10
4
cells
/
ml. Cells were treated with
10
M
ARE-1
with 10 nM dihydrotestosterone (DHT) or
DMSO for 24 h before harvest. RNA extraction (RNEasy
columns, Qiagen), complementary DNA (cDNA) gener-
ation (ProtoScript II First Strand cDNA Synthesis Kit,
NEB), and qRT-PCR (PowerUp SYBR Green Master Mix,
Life Technologies, ABI7300 instrument) were done follow-
ing manufacturer recommendations. Expression was nor-
malized to
-glucuronidase.
Bioavailability in new formulation
All animal experiments were performed at the Califor-
nia Institute of Technology (Pasadena, CA) with prior
IACUC approval. To evaluate a new formulation for
polyamide delivery,
ARE-1
was injected at 10 mg
/
kg in a
1% polyvinylpyrrolidone 10 (PVP), 50 mM Tris, 0.9% saline
vehicle into the right flank of 6 C57BL
/
6J mice. Mice were
anesthetized using isoflurane and blood collected retro-
orbitally at 30 min, 1, 3, 6, 12 and 24 h after injection. Blood
samples were centrifuged at 6000 rpm for 5 min to collect
the serum, which was processed as previously published and
analyzed by HPLC to determine polyamide concentration
(
25
). 9-aminoacridine was used as an internal standard.
Xenograft assay
Male SCID hairless outbred mice (4–6 weeks old) were ob-
tained from Charles River Laboratories. LNCaP-95 cells
(3
×
10
6
) were injected into the flanks of the mice as a
1:1 mixture in Matrigel (BD Biosciences). Mice were mon-
itored for the appearance of tumors and calipered twice
weekly once tumors appeared. When tumors reached 100
mm
3
(using 0.5*l*w*w), animals were castrated by veteri-
nary staff. Following surgery, animals were monitored daily
for 3 days, and allowed to recover for 7–10 days prior to
the start of treatment. After the recovery period, animals
were randomly assigned to treated or vehicle groups, and
injected three times per week with 2.5 mg
/
kg
ARE-1
or ve-
hicle (1% polyvinylpyrrolide 10 (PVP), 50 mM Tris, 0.9%
saline) for 3 weeks. Tumor growth was monitored weekly
by calipers, and growth compared to starting size. Animals
were anesthetized with 2–5% isoflurane
/
air when necessary,
and sterile technique was used for all procedures. Animal
health was monitored daily by veterinary staff, and any an-
imals exhibiting signs of distress were euthanized by admin-
istration of isoflurane followed by carbon dioxide.
Chromatin immunoprecipitation
Genomic occupancy of full-length AR was determined by
chromatin immunoprecipitation (ChIP) with the PG21 AR
antibody (Millipore). LNCaP-95 cells were plated at 20 mil-
lion cells per plate in phenol red-free RPMI 1640 supple-
mented with 10% CTFBS and allowed to adhere for 24 h.
The cells were treated with 10
M
ARE-1
with either 10 nM
DHT or DMSO for 24 h. Crosslinking was performed with
1% formaldehyde in media for 15 min followed by quench-
ing with 0.125 M glycine. The cells were then washed with
ice-cold PBS twice and harvested. Chromatin was sheared
by sonication at
−
20
◦
C at 25% amplitude in 30 s on and 10 s
off cycles for 30 cycles. Next, 1 mg of sheared chromatin was
incubated with PG21 antibody that was previously immobi-
lized on Dynabeads (Invitrogen) overnight at 4
◦
C. Samples
were then washed 5
×
with LiCl buffer (10 mM Tris, 500 mM
LiCl, 1% NP-40, 1% sodium deoxycholate) and once with
TE buffer. DNA was then harvested by phenol chloroform
extraction and purified using the Monarch PCR & DNA
Cleanup kit (NEB). Quantitative polymerase chain reac-
tion (qPCR) was used to validate enrichment at the
KLK3
ARE I site (5
-TGCATCCAGGGTGATCTAGT-3
and 5
-
ACCCAGAGCTGTGGAAGG-3
) compared to a nega-
tive internal locus (5
-TAGAAGGGGGATAGGGGAAC-
3
and 5
-CCAGAAAACTGGCTCCTTCTT-3
) prior to
submission for sequencing. Each sample was immunopre-
cipitated as three technical replicates, which were combined
Downloaded from https://academic.oup.com/nar/article-abstract/47/8/3828/5369940 by California Institute of Technology user on 01 May 2019
3830
Nucleic Acids Research, 2019, Vol. 47, No. 8
for sequencing on an Illumina HiSeq2500. Biological repli-
cates of each treatment condition were acquired. Input
DNA (not immunoprecipitated) was also extracted and pu-
rified using the same methods and submitted for sequenc-
ing.
ChIP-Seq analysis
At least 29.7 million reads were sequenced for each sam-
ple. Reads were mapped to the human genome (hg19) using
Bowtie2 v 2.2.3 and converted to BAM format with SAM-
tools (
26
,
27
). Peak calling was performed using the model-
based analysis of ChIP-Seq (MACS2) program for each
replicate (
28
). Peaks from each replicate of each condition
were compared using irreproducible discovery rate (IDR) to
determine a set of reproducible peaks, which was then sub-
mitted to multiple EM for motif elicitation (MEME)-ChIP
(
http://meme-suite.org/tools/meme-chip
) for motif analysis
(
29–31
). Peaks selected by IDR were converted to bigWig
format for viewing in the UCSC genome browser (
http:
//genome.ucsc.edu
).
Differential analysis between treatment conditions was
conducted using peak-calling prioritization (PePr) with a
P
-value cutoff of 1
×
10
−
5
, sharp peaks and intra-group
normalization (
32
). PePr results were used for all further
analysis. BEDtools was used for overlap analysis and peak
annotation was performed using ChIPseeker (
33
,
34
). Dif-
ferentially changed peaks were submitted to MEME-ChIP
for motif finding as above. Based on the MEME-ChIP re-
sults, Homer was used to examine the density of specific mo-
tifs within peaks (
35
). Data has been deposited and can be
accessed in GEO (GSE125552).
Thermal stabilization assay
Melting temperature analysis of the DNA oligos 5
-
TTGTAGAACACGTT-3
,5
-TTGTAGGACACGTT-
3
,5
-TTGTGGAACACGTT-3
and
5
-
TTGTGGGACACGTT-3
in the presence of
ARE-1
was conducted as previously described (
36
).
Statistical analysis
All statistical analysis was performed in GraphPad Prism.
Gene expression data were normalized to the DHT in-
duced condition and ANOVA analysis was performed on
three biological replicates using the Dunnett’s test for mul-
tiple comparisons. Statistical analysis of tumor percentage
growth between vehicle and
ARE-1
treated groups (
N
=
11
per group) was performed using the unpaired
t
-test. All re-
ported
P
-values are two-sided.
RESULTS
Nuclear uptake and cytotoxicity
Py-Im polyamide
ARE-1
has been previously shown to
exhibit antiproliferative activity toward several models of
prostate cancer including LNCaP, LNCaP-AR, VCaP and
LREX
(
14
,
22
). We further evaluate the activity of
ARE-
1
in LNCaP-95 cells, which derive their resistance from
the expression of AR splice variants (
37
). Nuclear local-
ization of
ARE-1
(Figure
1
A) was confirmed using a flu-
orescein analog,
ARE-1-FITC
(Supplementary Figure S1),
in LNCaP-95 cells (Figure
1
B). Antiproliferative effect of
ARE-1
toward LNCaP-95 cell growth was evaluated us-
ing the CTG assay and compared against the antiandro-
gen enzalutamide and pyrvinium pamoate (pyrvinium), a
molecule that has been reported to bind to the AR DNA-
binding domain to prevent AR–DNA interactions (
38
).
Results from the assay show the 72 h growth inhibition
IC
50
sfor
ARE-1
, enzalutamide and pyrvinium to be 20.1
M,
>
30
M and 44 nM, respectively. A synergistic ef-
fect was observed when a subtoxic concentration of enza-
lutamide (5
M) was combined with polyamide, and the
IC
50
was reduced to 3.4
M. Changes to
KLK3
gene ex-
pression was also evaluated in LNCaP-95 cells treated with
ARE-1
, enzalutamide, pyrvinium, and a combination of
ARE-1
with pyrvinium or enzalutamide (Figure
1
D). Af-
ter 24 h of treatment, the greatest reduction in
KLK3
ex-
pression from treatment with a single agent came from
ARE-1
, and combining either additional agent with
ARE-
1
further reduced gene expression. Based on these cell cul-
ture results, we further evaluated the antitumor effects of
ARE-1
in LNCaP-95 xenografts using an optimized for-
mulation that increased the subcutaneous bioavailability
when compared to the previously used DMSO
/
saline ve-
hicle (Supplementary Figure S2A). Animals were engrafted
with LNCaP-95 cells and monitored until palpable tumors
were observed. Once tumors reached 100 mm
3
, the animals
were castrated, allowed to recover for
∼
1 week and then
randomized before treatment (Figure
1
E). The animals were
treated with either vehicle or 2.5 mg
/
kg
ARE-1
subcuta-
neously Monday
/
Wednesday
/
Friday for 3 weeks. The ve-
hicle treated group grew
∼
380%, while the
ARE-1
treated
group grew 225%, for a 40% reduction in tumor size in
the polyamide treated mice (Figure
1
F). Animal weight was
measured at each injection and was not adversely affected
(Supplementary Figure S2B).
Genomic perturbation of androgen receptor occupancy
The effects of Py-Im polyamide treatment on AR occu-
pancy on chromatin have previously been explored by ChIP
experiments. A related Py-Im polyamide, targeting the same
sequence as
ARE-1
, has previously been shown to decrease
occupancy of AR at the
KLK3
promoter and enhancer in
LNCaP cells (
12
). In LNCaP-95 cells, a similar reduction
at the
KLK3
promoter ARE I is seen after 24 h of co-
treatment with
ARE-1
and 10 nM DHT (Supplementary
Figures S3A). In this study, we explored the genomic effect
ARE-1
treatment has on AR occupancy using ChIP-Seq
analysis. Sequencing results of biological duplicates of non-
treated (NT), 10 nM DHT treated (DHT), and 10 nM DHT
and 10
M
ARE-1
treated (DHT+
ARE-1
) showed
∼
30 mil-
lion reads mapping for all samples (Supplementary Figure
S3B). Sequencing reads were aligned to hg19 and select AR
target genes are shown (Figure
2
A and B). Motif analysis by
MEME discovered the forkhead-binding motif in all sam-
ples, and the complete ARE was discovered in the DHT
and DHT +
ARE-1
samples (Supplementary Figure S3C).
Differential binding of DHT
/
NT and DHT
/
(DHT+
ARE-
Downloaded from https://academic.oup.com/nar/article-abstract/47/8/3828/5369940 by California Institute of Technology user on 01 May 2019
Nucleic Acids Research, 2019, Vol. 47, No. 8
3831
-5
-4
-3
-2
-1
0
1
2
3
0
25
50
75
100
log[
M]
% Viability
ARE-1
+ 5
M Enz (3.4 ± 0.4
M)
Enz
(>30
M)
ARE-1
(20.2 ± 5.8
M)
Pyrvinium
(44 ± 4nM)
N
O
H
N
N
O
N
N
N
H
O
O
H
N
N
H
N
N
H
N
O
N
N
H
N
O
NH
O
N
H
N
O
N
O
N
H
N
NHAc
HN
O
O
OH
IPA
+
NHAc
ARE-1
ARE-1
(10
M)
Enz (5
M)
Pyrvinium (100nM)
-
-
-
-
-
-
-
+
-
-
-
+
+
-
+
+
+
-
0
0.2
0.4
0.6
0.8
1
DHT (10nM)
-+
++++
+
-
-
+
Relative
KLK3
expression
Engraft
Tumor growth
to 100 mm
3
Castrate
Recovery
7-10 days
Begin
treatment
End
experiment
1
2
024 7911 141618
Treatment on
circled days
0
100
200
300
400
500
1234
% starting size
Week
Vehicle, n = 11
2.5 mg/kg
ARE-1
, n = 11
AB
C
E
D
F
*
****
**
*******
n.s.
Figure 1.
(
A
) Structure of Py-Im polyamide
ARE-1
.(
B
) Nuclear localization of
ARE-1-FITC
in LNCaP-95 cells after 16-h treatment. (
C
) Cell viability
as determined by CellTiterGlo assay after 72-h treatment. IC
50
is indicated in parentheses. (
D
) Relative expression of
KLK3
in LNCaP-95 cells after the
indicated treatments. Cells were co-treated with
ARE-1
, enzalutamide, or pyrvinium pamoate and DHT at the indicated concentrations and RNA harvested
at 24 h. (
E
) Schedule of LNCaP-95 xenograft experiment. Animals were engrafted, allowed to grow tumors, castrated, allowed to recover for 7–10 days
and then treated with
ARE-1
or vehicle at indicated times. (
F
) Tumor volumes of LNCaP-95 xenografts in castrated mice treated with vehicle or 2.5 mg
/
kg
ARE-1
asshowninE.
N
=
11 for both groups; *
P
<
0.05, **
P
<
0.01, ****
P
<
0.0001.
1
) was calculated using PePr. Analysis revealed 16,015
peaks increased in DHT over non-treated (DHT
/
NT) and
6343 differentially changed DHT
/
(DHT+
ARE-1
) peaks, of
which 4921 overlapped with DHT inducible peaks (Fig-
ure
2
C). Correlation of peak location to genomic regions,
conducted by ChIPseeker, showed no difference between
the DHT
/
NT, DHT
/
(DHT+
ARE-1
), and overlap peaks,
suggesting that
ARE-1
does not have a regional binding
preference (Figure
2
D). Motif analysis of peaks unique to
DHT
/
NT revealed the canonical ARE where the first half-
site is 5
-RGNACA-3
. In this motif, the first position is se-
lective for A or G (R) and the third position is degenerate
for any base (N) (Figure
2
E). Motif analysis of the overlap-
ping peaks between DHT
/
NT and DHT
/
(DHT+
ARE-1
)
also revealed a complete ARE, however the first half-site
has the sequence 5
-RGWACA-3
, where the third position
shows selectivity for A or T (Figure
2
E); additional motifs
can be found in Supplementary Figure S4A. Comparison of
Downloaded from https://academic.oup.com/nar/article-abstract/47/8/3828/5369940 by California Institute of Technology user on 01 May 2019
3832
Nucleic Acids Research, 2019, Vol. 47, No. 8
A
C
DE
B
Figure 2.
ChIP-seq analysis of the
KLK3
promoter ARE I (
A
) and
PMEPA1
(
B
). (
C
) Overlap of peaks differentially changed in DHT relative to NT, and
peaks changed in DHT relative to DHT and
ARE-1
.(
D
) Identified peaks were annotated by genomic region by ChIPseeker. (
E
) Top ARE motif found by
MEME in the indicated peak sets.
the letter probability matrix between the DHT
/
NT unique
peaks and the overlapping peaks show more A character
in the first position and reduced C and G character in the
third position in the overlapping motif (Supplementary Fig-
ure S4B).
Of the possible permutations of the first ARE half-site,
ARE-1
is expected to have the strongest binding to the se-
quences 5
-AGWACA-3
. Based on Py-Im polyamide pair-
ing rules,
ARE-1
is expected to have lower binding to the se-
quences 5
-GGWACA-3
and 5
-AGGACA-3
, which con-
tain single base mismatches, and to have little binding to the
sequence 5
-GGGACA-3
, which contains two mismatches
(Figure
3
A) (
17–20
). DNA thermal stability experiments
confirmed this trend and showed that
ARE-1
stabilized
match sequences by
∼
9
◦
C; single mismatches reduced ther-
mal stability by
∼
2–4
◦
C.
ARE-1
showed no significant ther-
mal stabilization to a double mismatch sequence (Figure
3
B).
The ARE half-site sequence 5
-RGNACA-3
can be split
into four sequences: 5
-AGWACA-3
,5
-GGWACA-3
,5
-
AGSACA-3
and 5
-GGSACA-3
, where S represents G
or C. Density analysis of these four motifs revealed 5
-
AGWACA-3
to be significantly enriched around the peak
center of DHT
/
NT and DHT
/
(DHT+
ARE-1
)overlap
peaks compared to the other possible motifs. A lesser effect
was found for the DHT
/
NT unique peaks (Figure
3
C–D).
To confirm that the enrichment for 5
-AGWACA-3
was
only present in regions where AR peaks are affected by
ARE-1
, we examined common peaks between DHT
/
NT
and (DHT+
ARE-1
)
/
NT samples (Figure
4
A). Of the 7998
overlapping peaks, 2668 peaks had an absolute change of
<
1.5-fold. Motif density analysis in these unchanged re-
gions showed no enrichment of 5
-AGWACA-3
(Figure
4
B). Comparatively, 5
-AGWACA-3
was significantly en-
riched in 2129 peaks showing
>
2-fold change between
DHT
/
NT and (DHT+
ARE-1
)
/
NT.
DISCUSSION
Py-Im polyamides have been shown to inhibit the signaling
of oncogenic transcription factors and reduce their bind-
ing at select loci in ChIP experiments (
12
,
39
,
40
). Genomic
binding of Py-Im polyamides linked to DNA alkylators
have also been examined (
41
,
42
). In this study, we elucidate
the genome-wide effects of polyamide treatment on the AR
on chromatin. Py-Im polyamide
ARE-1
is a cell permeable
molecule that exerts anti-proliferative effects toward several
prostate cancer models, including the castration and enza-
lutamide resistant models LREX’ and now LNCaP-95.
In this present study, we find that
ARE-1
localizes to
LNCaP-95 nucleus within 16 h of dosing, and is able to
repress ligand-induced gene expression after 24 h of co-
treatment with DHT. In this time frame, our ChIP-Seq re-
sults show
ARE-1
is able to repress
∼
30% of DHT inducible
peaks. Motif analysis of these AR peaks repressed by
ARE-
1
, which is selective for the sequence 5
-WGWWCW-3
, in-
dicate that these loci are enriched for canonical AREs with
5
-RGWACA-3
as the first half-site compared to the com-
mon 5
-RGNACA-3
half-site. Thus, the differential effects
on AR-DNA binding events
in vivo
reflects the DNA target
Downloaded from https://academic.oup.com/nar/article-abstract/47/8/3828/5369940 by California Institute of Technology user on 01 May 2019
Nucleic Acids Research, 2019, Vol. 47, No. 8
3833
A
B
CD
Figure 3.
(
A
) Illustration of match and mismatch
ARE-1
binding sites. Mismatches are shown in red and boxed. (
B
) Melting temperature analysis of match
and mismatch sequences shown in A. C + D) Motif density analysis of possible ARE half sites in peaks unique to DHT
/
NT (
C
) and overlapping between
DHT
/
NT and DHT
/
DHT +
ARE-1
(
D
).
7,998
8,027
1,102
DHT
NT
DHT +
ARE-1
NT
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
-500
-300
-100
100
300
500
Distance from peak (bp)
common peaks with
>2 fold change
(DHT/NT) vs (DHT +
ARE-1
/NT)
(2,129)
AGSACA
AGWACA
GGSACA
GGWACA
Motifs per base pair per peak
Distance from peak (bp)
common peaks with
< 1.5 abs. fold change
(2,668)
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
-500
-300
-100
100
300
500
AB
C
Figure 4.
(
A
) Venn diagram showing the overlap of peaks differentially changed in DHT as compared to NT with peaks changed by DHT and
ARE-1
as compared to NT. (
B
and
C
) Density analysis of overlapping peaks with
<
1.5 absolute fold change between DHT
/
NT and DHT+
ARE-1
/
NT (B) and
peaks with
>
2-fold change when comparing DHT
/
NT to DHT+
ARE-1
/
NT (C).
Downloaded from https://academic.oup.com/nar/article-abstract/47/8/3828/5369940 by California Institute of Technology user on 01 May 2019
3834
Nucleic Acids Research, 2019, Vol. 47, No. 8
sequence binding preference of
ARE-1
in vitro
. These exper-
iments provide evidence of the
in vivo
sequence selectivity of
ARE-1
, and provide a snapshot of how
ARE-1
modulates
the AR cistrome.
DATA AVAILABILITY
Data deposited in GEO: series GSE125552
SUPPLEMENTARY DATA
Supplementary Data
are available at NAR Online.
ACKNOWLEDGEMENTS
The authors thank the veterinary staff at Caltech for all
their assistance, particularly Gloria Martinez for perform-
ing the castration surgeries, and Gwen Williams and John
Papsys for monitoring animal health. Sequencing was per-
formed at the Millard and Muriel Jacobs Genetics and Ge-
nomics Laboratory at California Institute of Technology,
and confocal imaging was performed at the Caltech Biolog-
ical Imaging Facility.
FUNDING
National Institutes of Health [GM027681 to P.B.D.,
T32GM761637 to A.A.K.]; Ralph M. Parsons Foundation
graduate fellowship (to A.A.K.). Funding for open access
charge: National Institutes of Health grant [GM027681].
Conflict of interest statement.
One of the authors in this
study, P.B.D., holds shares in a privately held entity focused
on developing Py-Im polyamides for the treatment of an-
tiandrogen resistant prostate cancer.
REFERENCES
1. Huggins,C. and Hodges,C.V. (1941) Studies on prostatic cancer. I.
The effect of castration, of estrogen and of androgen injection on
serum phosphatases in metastatic carcinoma of the prostate.
Cancer
Res.
,
1
, 293–297.
2. Giorgi,E.P., Shirley,I.M., Grant,J.K. and Stewart,J.C. (1973)
Androgen dynamics in vitro in the human prostate gland. Effect of
cyproterone and cyproterone acetate.
Biochem. J.
,
132
, 465–474.
3. Heinlein,C.A. and Chang,C. (2004) Androgen receptor in prostate
cancer.
Endocr. Rev.
,
25
, 276–308.
4. Sharifi,N. (2005) Androgen deprivation therapy for prostate cancer.
JAMA
,
294
, 238–244.
5. Tran,C., Ouk,S., Clegg,N.J., Chen,Y., Watson,P.A., Arora,V.,
Wongvipat,J., Smith-Jones,P.M., Yoo,D., Kwon,A.
et al.
(2009)
Development of a second-generation antiandrogen for treatment of
advanced prostate cancer.
Science
,
324
, 787–790.
6. Clegg,N.J., Wongvipat,J., Joseph,J.D., Tran,C., Ouk,S., Dilhas,A.,
Chen,Y., Grillot,K., Bischoff,E.D., Cai,L.
et al.
(2012) ARN-509: a
novel antiandrogen for prostate cancer treatment.
Cancer Res.
,
72
,
1494–1503.
7. Shore,N.D., Tammela,T.L., Massard,C., Bono,P., Aspegren,J.,
Mustonen,M. and Fizazi,K. (2018) Safety and antitumour activity of
ODM-201 (BAY-1841788) in chemotherapy-na
̈
ıve and CYP17
inhibitor-na
̈
ıve patients: follow-up from the ARADES and
ARAFOR trials.
Eur. Urol. Focus
,
4
, 547–553.
8. Chen,C.D., Welsbie,D.S., Tran,C., Baek,S.H., Chen,R., Vessella,R.,
Rosenfeld,M.G. and Sawyers,C.L. (2004) Molecular determinants of
resistance to antiandrogen therapy.
Nat. Med.
,
10
, 33–39.
9. Dehm,S.M. and Tindall,D.J. (2011) Alternatively spliced androgen
receptor variants.
Endocr. Relat. Cancer
,
18
, R183–196.
10. Korpal,M., Korn,J.M., Gao,X., Rakiec,D.P., Ruddy,D.A., Doshi,S.,
Yuan,J., Kovats,S.G., Kim,S., Cooke,V.G.
et al.
(2013) An F876L
mutation in androgen receptor confers genetic and phenotypic
resistance to MDV3100 (enzalutamide).
Cancer Discov.
,
3
, 1030–1043.
11. Elshan,N.G.R.D., Rettig,M.B. and Jung,M.E. (2018) Molecules
targeting the androgen receptor (AR) signaling axis beyond the
AR-Ligand binding domain.
Med. Res. Rev.
, doi:10.1002
/
med.21548.
12. Nickols,N.G. and Dervan,P.B. (2007) Suppression of androgen
receptor-mediated gene expression by a sequence-specific
DNA-binding polyamide.
Proc. Natl. Acad. Sci. U.S.A.
,
104
,
10418–10423.
13. Hargrove,A.E., Martinez,T.F., Hare,A.A., Kurmis,A.A.,
Phillips,J.W., Sud,S., Pienta,K.J. and Dervan,P.B. (2015) Tumor
repression of VCaP Xenografts by a pyrrole-imidazole polyamide.
PLoS One
,
10
, e0143161.
14. Kurmis,A.A., Yang,F., Welch,T.R., Nickols,N.G. and Dervan,P.B.
(2017) A pyrrole-imidazole polyamide is active against
Enzalutamide-resistant prostate cancer.
Cancer Res.
,
77
, 2207–2212.
15. Trauger,J.W., Baird,E.E. and Dervan,P.B. (1996) Recognition of
DNA by designed ligands at subnanomolar concentrations.
Nature
,
382
, 559–561.
16. Hsu,C.F., Phillips,J.W., Trauger,J.W., Farkas,M.E., Belitsky,J.M.,
Heckel,A., Olenyuk,B.Z., Puckett,J.W., Wang,C.C.C. and Dervan,P.B.
(2007) Completion of a programmable DNA-Binding small molecule
library.
Tetrahedron
,
63
, 6146–6151.
17. White,S., Baird,E.E. and Dervan,P.B. (1997) On the pairing rules for
recognition in the minor groove of DNA by pyrrole-imidazole
polyamides.
Chem. Biol.
,
4
, 569–578.
18. Kielkopf,C.L., White,S., Szewczyk,J.W., Turner,J.M., Baird,E.E.,
Dervan,P.B. and Rees,D.C. (1998) A structural basis for recognition
of A.T and T.A base pairs in the minor groove of B-DNA.
Science
,
282
, 111–115.
19. Kielkopf,C.L., Baird,E.E., Dervan,P.B. and Rees,D.C. (1998)
Structural basis for G.C recognition in the DNA minor groove.
Nat.
Struct. Biol.
,
5
, 104–109.
20. Foister,S., Marques,M.A., Doss,R.M. and Dervan,P.B. (2003) Shape
selective recognition of T.A base pairs by hairpin polyamides
containing N-terminal 3-methoxy (and 3-chloro) thiophene residues.
Bioorg. Med. Chem.
,
11
, 4333–4340.
21. Chenoweth,D.M. and Dervan,P.B. (2010) Structural basis for Cyclic
Py-Im polyamide allosteric inhibition of nuclear receptor binding.
J.
Am. Chem. Soc.
,
132
, 14521–14529.
22. Yang,F., Nickols,N.G., Li,B.C., Szablowski,J.O., Hamilton,S.R.,
Meier,J.L., Wang,C.-M. and Dervan,P.B. (2013) Animal toxicity of
hairpin pyrrole-imidazole polyamides varies with the turn unit.
J.
Med. Chem.
,
56
, 7449–7457.
23. Yang,F., Nickols,N.G., Li,B.C., Marinov,G.K., Said,J.W. and
Dervan,P.B. (2013) Antitumor activity of a pyrrole-imidazole
polyamide.
Proc. Natl. Acad. Sci. U.S.A.
,
110
, 1863–1868.
24. Martinez,T.F., Phillips,J.W., Karanja,K.K., Polaczek,P., Wang,C.-M.,
Li,B.C., Campbell,J.L. and Dervan,P.B. (2014) Replication stress by
Py-Im polyamides induces a non-canonical ATR-dependent
checkpoint response.
Nucleic Acids Res.
,
42
, 11546–11559.
25. Raskatov,J.A., Hargrove,A.E., So,A.Y. and Dervan,P.B. (2012)
Pharmacokinetics of Py-Im polyamides depend on architecture:
cyclic versus linear.
J. Am. Chem. Soc.
,
134
, 7995–7999.
26. Langmead,B., Trapnell,C., Pop,M. and Salzberg,S.L. (2009) Ultrafast
and memory-efficient alignment of short DNA sequences to the
human genome.
Genome Biol.
,
10
, R25.
27. Li,H., Handsaker,B., Wysoker,A., Fennell,T., Ruan,J., Homer,N.,
Marth,G., Abecasis,G., Durbin,R. and 1000 Genome Project Data
Processing Subgroup (2009) The sequence alignment
/
map format
and SAMtools.
Bioinformatics
,
25
, 2078–2079.
28. Zhang,Y., Liu,T., Meyer,C.A., Eeckhoute,J., Johnson,D.S.,
Bernstein,B.E., Nusbaum,C., Myers,R.M., Brown,M., Li,W.
et al.
(2008) Model-based analysis of ChIP-Seq (MACS).
Genome Biol.
,
9
,
R137.
29. Li,Q., Brown,J.B., Huang,H. and Bickel,P.J. (2011) Measuring
reproducibility of high-throughput experiments.
Ann. Appl. Stat.
,
5
,
1752–1779.
30. Machanick,P. and Bailey,T.L. (2011) MEME-ChIP: motif analysis of
large DNA datasets.
Bioinformatics
,
27
, 1696–1697.
31. Bailey,T.L., Boden,M., Buske,F.A., Frith,M., Grant,C.E.,
Clementi,L., Ren,J., Li,W.W. and Noble,W.S. (2009) MEME SUITE:
Downloaded from https://academic.oup.com/nar/article-abstract/47/8/3828/5369940 by California Institute of Technology user on 01 May 2019
Nucleic Acids Research, 2019, Vol. 47, No. 8
3835
tools for motif discovery and searching.
Nucleic Acids Res.
,
37
,
W202–W208.
32. Zhang,Y., Lin,Y.-H., Johnson,T.D., Rozek,L.S. and Sartor,M.A.
(2014) PePr: a peak-calling prioritization pipeline to identify
consistent or differential peaks from replicated ChIP-Seq data.
Bioinformatics
,
30
, 2568–2575.
33. Quinlan,A.R. and Hall,I.M. (2010) BEDTools: a flexible suite of
utilities for comparing genomic features.
Bioinformatics
,
26
, 841–842.
34. Yu,G., Wang,L.-G. and He,Q.-Y. (2015) ChIPseeker: an
R
/
Bioconductor package for ChIP peak annotation, comparison and
visualization.
Bioinformatics
,
31
, 2382–2383.
35. Heinz,S., Benner,C., Spann,N., Bertolino,E., Lin,Y.C., Laslo,P.,
Cheng,J.X., Murre,C., Singh,H. and Glass,C.K. (2010) Simple
combinations of lineage-determining transcription factors prime
cis-regulatory elements required for macrophage and B cell identities.
Mol. Cell
,
38
, 576–589.
36. Dose,C., Farkas,M.E., Chenoweth,D.M. and Dervan,P.B. (2008)
Next generation hairpin polyamides with (R)-3,4-diaminobutyric
acid turn unit.
J. Am. Chem. Soc.
,
130
, 6859–6866.
37. Liu,L.L., Xie,N., Sun,S., Plymate,S., Mostaghel,E. and Dong,X.
(2014) Mechanisms of the androgen receptor splicing in prostate
cancer cells.
Oncogene
,
33
, 3140–3150.
38. Jones,J.O., Bolton,E.C., Huang,Y., Feau,C., Guy,R.K.,
Yamamoto,K.R., Hann,B. and Diamond,M.I. (2009)
Non-competitive androgen receptor inhibition in vitro and in vivo.
Proc. Natl. Acad. Sci. U.S.A.
,
106
, 7233–7238.
39. Nickols,N.G., Jacobs,C.S., Farkas,M.E. and Dervan,P.B. (2007)
Modulating hypoxia-inducible transcription by disrupting the
HIF-1-DNA interface.
ACS Chem. Biol.
,
2
, 561–571.
40. Muzikar,K.A., Nickols,N.G. and Dervan,P.B. (2009) Repression of
DNA-binding dependent glucocorticoid receptor-mediated gene
expression.
Proc. Natl. Acad. Sci. U.S.A.
,
106
, 16598–16603.
41. Erwin,G.S., Grieshop,M.P., Bhimsaria,D., Do,T.J.,
Rodr
́
ıguez-Mart
́
ınez,J.A., Mehta,C., Khanna,K., Swanson,S.A.,
Stewart,R., Thomson,J.A.
et al.
(2016) Synthetic genome readers
target clustered binding sites across diverse chromatin states.
Proc.
Natl. Acad. Sci. U.S.A.
,
113
, E7418–E7427.
42. Chandran,A., Syed,J., Taylor,R.D., Kashiwazaki,G., Sato,S.,
Hashiya,K., Bando,T. and Sugiyama,H. (2016) Deciphering the
genomic targets of alkylating polyamide conjugates using
high-throughput sequencing.
Nucleic Acids Res.
,
44
, 4014–4024.
Downloaded from https://academic.oup.com/nar/article-abstract/47/8/3828/5369940 by California Institute of Technology user on 01 May 2019