Abstract
Background
Estrogen receptor α (ESR1) plays a critical role in promoting growth of
various cancers. Yet, its role in the development of pancreatic cancer
is not well-defined. A less studied region of ESR1 is the hinge region,
connecting the ligand binding and DNA domains. rs142712646 is a rare
SNP in ESR1, which leads to a substitution of arginine to cysteine at
amino acid 269 (R269C). The mutation is positioned in the hinge region
of ESR1, hence may affect the receptor structure and function. We aimed
to characterize the activity of R269C-ESR1 and study its role in the
development of pancreatic cancer.
Methods
Transcriptional activity was evaluated by E2-response element (ERE) and
AP1 –luciferase reporter assays and qRT-PCR. Proliferation and
migration were assessed using MTT and wound healing assays.
Gene-expression analysis was performed using RNAseq.
Results
We examined the presence of this SNP in various malignancies, using the
entire database of FoundationOne and noted enrichment of it in a subset
of pancreatic non-ductal adenocarcinoma (n = 2800) compared to
pancreatic ductal adenocarcinoma (PDAC) as well as other tumor types
(0.53% vs 0.29%, p = 0.02). Studies in breast and pancreatic cancer
cells indicated cell type-dependent activity of ESR1 harboring R269C.
Thus, expression of R269C-ESR1 enhanced proliferation and migration of
PANC-1 and COLO-357 pancreatic cancer cells but not of MCF-7 breast
cancer cells. Moreover, R269C-ESR1 enhanced E2-response elements (ERE)
and AP1-dependent transcriptional activity and increased mRNA levels of
ERE and AP1-regulated genes in pancreatic cancer cell lines, but had a
modest effect on MCF-7 breast cancer cells. Accordingly, whole
transcriptome analysis indicated alterations of genes associated with
tumorigenicity in pancreatic cancer cells and upregulation of genes
associated with cell metabolism and hormone biosynthesis in breast
cancer cells.
Conclusions
Our study shed new light on the role of the hinge region in regulating
transcriptional activity of the ER and indicates cell-type specific
activity, namely increased activity in pancreatic cancer cells but
reduced activity in breast cancer cells. While rare, the presence of
rs142712646 may serve as a novel genetic risk factor, and a possible
target for therapy in a subset of non-ductal pancreatic cancers.
Keywords: Pancreatic cancer, Estrogen receptor, Activation Function-1,
SNP
Background
The human estrogen receptor α (ERα), encoded by ESR1, is a member of
the steroid/nuclear receptor superfamily and functions as
ligand-activated transcription factors [[41]1]. Upon binding of
estrogen, the ER dimerizes and binds to coactivators. The complex is
then recruited to the estrogen-responsive elements (ERE) on the
promoters of ER target genes. The major functional domains of ERα are
the N-terminal Activation Function-1 (AF-1) which modulates
transcription, the DNA-binding domain (DBD) and the ligand-binding
domain (LBD) that contains Activation Function-2 (AF-2) [[42]2]. A less
characterized domain of ERα is the hinge region, which lies between the
DBD and the LBD. The hinge region contains putative nuclear
localization sequence (NLS) and may play a role in transcriptional
regulation [[43]1, [44]3–[45]6].
Approximately 75% of all breast cancers express ERα, and targeting ERα
signaling, is a key treatment strategy in these tumors. ERα plays a
major role in the development of other malignancies, including
endometrial and ovarian cancers [[46]7] and is expressed in subsets of
additional tumors, including lung [[47]8] gastric [[48]9] and colon
[[49]10] cancers. Pancreatic cancer is the fourth leading cause of
cancer death, with five-year survival of roughly 8% [[50]11]. Known
risk factors for pancreatic cancer include smoking, diabetes, obesity
and pancreatitis [[51]12] and up to 10% of the cases are attributed to
high-risk inherited mutations [[52]13, [53]14]. Yet, initiating genetic
mechanisms leading to the development of pancreatic cancer are mostly
unknown. Pancreatic ductal adenocarcinoma (PDAC) is the most common
histological subtype of pancreatic cancer, representing 85% of all
pancreatic neoplasms while less common subtypes include adenosquamous,
mucinous, anaplastic and signet ring cancers [[54]15]. Interestingly,
ERα is also expressed in a subset of pancreatic adenocarcinoma, most
notably in mucinous tumors [[55]16–[56]19] and in vitro and in vivo
studies indicated growth inhibition of pancreatic cancer cells by
tamoxifen [[57]20, [58]21]. Several clinical trials reported on
activity of hormonal therapy in pancreatic cancer [[59]22–[60]28]
including a prospective randomized trial that reported on median
survival of 5.3 months among 37 tamoxifen-treated patients, compared to
3 months in 39 patients treated with a placebo, with marginal
statistical significance (p = 0.07) [[61]29]. Yet, no benefit was noted
in a smaller trial [[62]30].
A documented rare SNP in the ESR1 gene (rs142712646) leads to a
substitution of arginine at position 269 of the hinge region to
cysteine (R269C). To our knowledge, the activity of this SNP, and its
potential role in tumorigenesis, has not been reported yet. We report
here on enrichment of this functional variant in non-PDAC pancreatic
cancers. In this study, we aimed to characterize the activity of
R269C-ESR1 in pancreatic and breast cancer cells and identify its role
as a potential driver of proliferation of pancreatic cancer cells.
Functional analysis revealed increased AP-1 dependent gene-expression
of this variant in pancreatic but not in breast cancer cells, and
expression of the R269C variant enhanced proliferation and migration of
pancreatic cancer cells. These data indicate unique, cell-type
dependent activity, of R269C and its contribution to tumor
aggressiveness in a small subset of pancreatic cancers.
Methods
Foundation medicine database analysis
We searched for mutations in the ESR1 gene, in the entire database of
the Foundation One clinical database (Foundation One, Foundation
Medicine, Cambridge, MA). The database consists of > 100,000 cases, of
them > 4000 PDAC cases and ~ 2800 non-PDAC pancreatic cancer. The test
has been described previously [[63]31] and consists of deep sequencing
of cancer-related genes on DNA extracted from paraffin embedded tissue
samples.
Computational structure analysis
The secondary structure of ERα was predicted using ConSSert [[64]32],
PsiPred [[65]33], Jnet [[66]34] and Predator [[67]35]. A multiple
sequence alignment of Human ERα and other 38 vertebrates homologs,
collected from SwissProt, was calculated using Mafft [[68]36]. The
alignment and secondary structure annotations were presented in Jalview
[[69]37]. Known domains of ERα were taken from Pfam [[70]38].
Reagents and antibodies
17β-Estradiol (E[2]) and crystal violet were obtained from Sigma (St.
Louis, MO); ICI 182,780 from Tocris Bioscience (Ellisville, MO), G418
from Life Technologies (Waltham, MA); qScript cDNA SuperMix and
PerfeCTa SYBR Green FastMix from Quanta BioSciences (Gaithersburg, MD).
Primers synthesis- IDT (Coralville, IA).
Plasmids and constructs
The ERE-luciferase reporter construct, kindly provided by D. Harris,
(UCLA, CA), consists of 2 repeats of the upstream region of the
vitellogenin ERE promoter. pRL Renilla luciferase control was purchased
from Promega (Cat no E2261, Promega, Madison, WI). The generation of
WT-ER construct (in pcDNA3) was described elsewhere [[71]39]. Arginine
to cysteine mutation (R269C-ER) was inserted using WT-ER as a template
(generated by GeneScript Inc. HK, China).
Cells and transfection
Human kidney cell line HEK293, breast cancer cell line MCF-7 and
pancreatic cancer cell lines COLO-357 and PANC-1 were obtained from
ATCC (Manassas, VA). All cells were maintained in Dulbecco’s modified
Eagle’s medium (DMEM), containing 10% fetal bovine serum and 100 U/ml
penicillin/streptomycin (1%) at 37 °C in a humidified 5% CO2
atmosphere. All experiments were conducted with cells under 15
passages. For estrogen studies, cells were cultured in phenol-free
media using 10% charcoal-treated serum (Beit Haemek, Israel) for 2 days
before treatment. All transfections were conducted with Jet PEI
(Polypus Transfection, Illkirch, France) according to the
manufacturer’s instructions.
Luciferase assays
The assays were conducted essentially as described [[72]40–[73]42]. In
brief, cells grown in phenol-free media using 10% charcoal-treated
serum were plated in 96-well plates, and transiently transfected with
the constructs (WT-ER or R269C-ER), reporter vector (ERE-luciferase or
AP-1 luciferase) and Renilla vector. Twenty-four hours later cells were
treated with 10 nM E2 or a 0.0003% ethanol as a control vehicle for the
ERE-luciferase assay or with ICI 182780 or 0.001% DMSO as a control
vehicle for the Ap-1-luciferase assay [[74]43]. At indicated times
Dual-Glo Luciferase (Promega) reagent was added to the medium then the
cells were incubated at 25 °C for 30 min afterwards the firefly
luminescence was measured by multichannel plate spectrophotometer.
After the first reading, Dual-Glo® Stop & Glo (Promega) reagent was
added to the plate, cells were incubated at 25 °C for 30 min, then
Renilla luminescence was measured similarly. Luciferase activity was
normalized by calculating the ratio of Firefly to Renilla luciferase
units.
Quantitative real time reverse transcription-PCR (qRT-PCR)
Two days after transfection with the different constructs, total RNA
was prepared using the High Pure RNA Isolation Kit Roche (Roche). Total
RNA (1 μg) was reverse transcribed using qScript cDNA synthesis kit
(Quanta Biosciences). Quantitative RT-PCR (qRT-PCR) was used to
determine mRNA level. Primers were designed using Primer Express
(Applied Biosystems, Foster City, CA, USA) and synthesized by IDT
(Coralville, IA). Primers used: GREB1a (human): F
5′-ACGTGTGGTGACTGGAGTAGC, R 5′- CCACGCAAGGTAGAAGGTGA; TGF-α (human): F
5′- CCCGCTGAGTGCAGACC, R 5′-ACGTACCCAGAATGGCAGAC; CyclinD-1 (human): F
5′- TGGAGGTCTGCGAGGAACAG, R 5′- AGCTGCAGGCGGCTCTTT; IGF-1R (human): F
5′- ATGTCCAGGCCAAAACAGGAT, R 5′- CAACCCTCCCACGATCAACA. Equal loading
was determined using β-actin–specific primers. Amplification reactions
were performed with Platinum qPCR SuperMix in triplicate using StepOne
Plus (Applied Biosystems). PCR conditions: 50 °C for 2 min, 95 °C for
2 min, followed by 40 cycles of 95 °C for 15 s, 60 °C for 45 s.
Migration assay
Migration was assessed using the wound-healing (“scratch”) assay.
COLO-357, PANC-1 and MCF-7 cells were grown to confluency in 6-well
plates, with the various constructs (pcDNA3, WT-ER or R269C-ER) and
grown in phenol-free media with 10% charcoal-treated serum for 24 h.
Cells’ monolayer was scraped in a straight line with a 200 μL sterile
pipette tip for 48 h. The cells were photographed at 0, 24, and 48 h
with an inverted phase-contrast microscope (Olympus, Tokyo, Japan).
Calculation of cell migration (d) was determined using the equation
[MATH:
=m1−m22 :MATH]
, when m1 is wound width at time 0 and m2 at 24 or 48 h.
Western blot analysis
Cells were washed twice with PBS, snap frozen in liquid nitrogen and
stored at − 80 °C until the analysis. Cells were harvested and lysed
for total protein extraction in radio-immunoprecipitation assay (RIPA)
buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 0.25%
Na-deoxycholate, 1 mM EDTA, 1 mM NaF) together with a protease
inhibitor cocktail (Sigma). A total of 50 μg protein extracts were
loaded on 10% polyacrylamide gels, separated electrophoretically and
blotted from the gel onto nitrocellulose membrane (Schleicher & Schuell
Bioscience GmbH, Dassel, DE, USA). The membranes were blocked with skim
milk 1% in PBS-T (0.01 M Tris-HCl pH -7.6, 0.15 M NaCl, 0.2% Tween 20)
for an hour, and then immunoblotted overnight with the indicated
antibodies. β-actin antibody was used as loading control. Membranes
were washed 5 times with TBS-T, followed by incubation with horse
raddish peroxidase (HRP, Jackson Immuno Research, West Grove, PA)
conjugated antibodies and were detected by using enhanced
chemiluminescence (ECL) reaction.
Proliferation assay
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)
assay was conducted essentially as described [[75]44]. In brief, cells
were counted and plated in 96-well plates (5000 cells/well), and
transiently transfected with WT-ER, R269C-ER or an empty vector then
cultured in phenol-free media with 10% charcoal-treated serum.
Twenty-four hours later cells were treated with E2 (10 nM) or 0.0003%
ethanol as a control vehicle and at indicated time points MTT reagent
was added to the medium (500 μg/ml). Cells were incubated for 1.5 h,
afterward medium was removed, 100% DMSO was added and absorbance was
determined at a wavelength of 570 nm using a multichannel plate
spectrophotometer.
RNAseq
MCF-7 or COLO-357 cells were seeded in phenol red depleted medium with
charcoal stripped serum. Cells then were transfected with either WT-ER
or R269C-ER in triplicates and then treated with vehicle control or E2
(10nM) for 24 h. Total RNA was extracted using the High Pure RNA
Isolation Kit (Roche, Mannheim, Germany). RNAseq and bioinformatics
were conducted at the Tel-Aviv University Genomics Research Unit and
Bioinformatics Unit (Tel-Aviv, Israel). The libraries were prepared
using NEBNext® Ultra™ II RNA Library Prep Kit for Illumina® (New
England Biolabs®inc., 240 County Road Ipswich, MA). For sequencing:
briefly, 1000 ng of total RNA were fragmented followed by reverse
transcription and second strand cDNA synthesis. The double strand cDNA
was subjected to end repair, A-base addition, adapter ligation and PCR
amplification to create barcoded libraries. Libraries were evaluated by
Qubit and TapeStation. Sequencing was conducted with NextSeq 500/550
v2.5 (Illumina) at 75-cycles, Single Read kit. The output was ~ 21
million reads per sample.
Bioinformatics
Poly-A/T stretches and Illumina adapters were trimmed from the reads
using cutadapt; resulting reads shorter than 30 bp were discarded.
FastQ files were uploaded to Partek Flow [[76]45] for processing. Reads
were mapped to the Homo sapiens GRCm38 reference genome using BWA-MEM
[[77]46] (with default parameters). Quantification to annotation model
was performed using Partek E/M [[78]47] and RPKM normalized expression
levels for each gene were obtained. Differentially expressed genes were
identified using GSA (Genome specific analysis) [[79]48, [80]49].
Pathway and function enrichment were analyzed using specified web-tools
and heatmaps of genes associated with specific pathways were generated
using Partek Genomic Suite [[81]50].
Statistical analysis
Each experiment was performed at least three times. The data were
expressed as the mean ± SD or SE. Statistical significance was assessed
by Student’s t test. A P value of < 0.05 is considered statistically
significant.
Results
Prevalence of R269C substitution in ESR1 in clinical samples of pancreatic
cancer
The substitution of C to T at position 1039, leading to a substitution
of arginine at position 269 to cysteine (R269C), was observed in a
patient with mucinous pancreatic adenocarcinoma conducting tumor
genomic analysis for clinical purposes. This substitution is a known
rare SNP (rs142712646) [[82]51] and its frequency among European
population is estimated to be 0.14% according to the ExAC dataset.
Following this observation, the frequency of this alteration was
examined on the entire clinical database of FoundationOne, containing
> 150,000 tumor samples of various origins. The observed frequency in a
wide range of malignancies, including breast and pancreatic ductal
adenocarcinoma, was 0.29% and was considered to be similar to the
frequency at the ExAC dataset. Yet, a significantly higher frequency
was noted in ~ 2800 non-PDAC pancreatic cancer (0.53%, p = 0.02). This
enrichment suggests that the R269C substitution may play a role in the
development of these tumors.
Structural analysis of ESR1 harboring the R269C alteration
R269 lies within a potential nuclear localization sequences (NLS),
known as p-NLS2 [[83]52, [84]53]. Analysis of Swissprot database
indicated position 269 to be highly conserved among species as either
arginine or lysine (Fig. [85]1). Structural analysis using four
different secondary structure prediction tools (described under
Materials and Methods) suggests that it is located at the end of a
possible α-helix, in a region connecting a zinc finger domain and the
hormone binding domain. Thus, R269 is predicted to be in contact with
residues from the DNA binding domain and the steroid binding domain,
and as arginine (pKa = 12.5) is positively charged at physiological pH,
the substitution to cysteine would result in a change from a positive
charge to a more negative one at position 269. Moreover, it may lead to
the formation of wrong SS bonds in the zinc finger domain, containing
several conserved cysteine residues.
Fig. 1.
[86]Fig. 1
[87]Open in a new tab
Position 269 is conserved as arginine or lysine among species. Multiple
Sequence Alignment (MSA) of ESR1 39 sequences, colored by amino acid
type. The mutated position is marked by asterisk above the alignment.
Four different secondary structure prediction methods are shown below
the alignment with red and green segments (red symbolizes an alpha
helix and green- beta strand). The Pfam domains are shown by blue
lines. The conservation of the residues is shown by brown-yellow bars.
The higher and yellow bars indicate highly conserved positions. The
figure was generated using Jalview. Position 269 is conserved as
arginine (R) or lysine (K) among species as can be seen in the red
rectangle. It is located at the end of a possible alpha helix, in a
region connecting a zinc finger domain and the hormone receptor domain
Increased estrogen-dependent transcriptional activity of R269C-ER compared to
WT-ER in breast and pancreatic cancer cells
In order to study the transcriptional activity of R269C-ER, we
generated ER harboring the R269C substitution. Expression of R269C-ER
in pancreatic cancer cells was similar to that of overexpressed WT-ER
(Supplementary Fig. S1A, B, Additional file [88]1). The
vitellogenin-based ERE-luciferase reporter was then used to study its
transcriptional activity, compared to the WT-ER [[89]40]. For the
analysis, MCF-7 breast cancer cells, COLO-357 and PANC-1 pancreatic
cancer cells and the non-cancerous HEK-293 cells were co-transfected
with the reporter and either WT-ER or R269C-ER, grown in
estrogen-depleted medium and treated with either vehicle control or E2.
R269C-ER demonstrated significantly increased ERE activity compared to
WT-ER in all four cell lines, either with or without E2 treatment
(Fig. [90]2, panels a-d). In COLO-357 cells, R269C-ER increased ERE
activity by 41% compared to WT-ER and with E2 treatment the mutant
increased it by 22% (Fig. [91]2a, p < 0.01). In PANC-1 cells the
mutant-ER activity was increased by 44% compared to WT-ER (Fig. [92]2b,
p < 0.05). In MCF-7 cells, R269C-ER increased the activity, with or
without E2 treatment, by 61 and 114%, respectively (Fig. [93]2c,
p < 0.01). Finally, in HEK-293 the mutation increased the activity by
93% (Fig. [94]2d, p < 0.01) and by only 16% with E2 treatment (Fig.
[95]2d, p < 0.05).
Fig. 2.
[96]Fig. 2
[97]Open in a new tab
Transcriptional activity of R269C-ER in breast and pancreatic cancer
cells. Cells were transiently transfected with either WT-ER or R269C-ER
vectors together with the ERE luciferase reporter and treated with E2
(10nM) or a control vehicle for 24 h. Luciferase activity were analyzed
and normalized to Renilla luciferase units and are shown relative to
the control WT-ER. The results are from a representative experiment of
at least three independent experiments, each performed in hexaplicates.
Each bar represents the mean ± SD. *, P < 0.05, **, P < 0.01
In order to further validate the transcriptional activity of R269C-ER,
we examined its ability to induce transcription of the classic
estrogen-regulated genes GREB1 and TGF-α, which their promoter contains
ERE sequences [[98]40]. Surprisingly, while overexpression of R269C-ER
in MCF-7 cells decreased mRNA levels of GREB1 by 31%, and TGF-α by 25%,
and also significantly reduced the response to E2 treatment
(Fig. [99]3a, b, p < 0.05 for all comparisons), it enhanced GREB and
TGF-α mRNA levels in COLO-357 cells in response to E2 (Fig. [100]3c, d,
p < 0.01). These data indicate complex, cell-type dependent
transcriptional activity of both WT-ER and R269C-ER in pancreatic
cancer cells.
Fig. 3.
[101]Fig. 3
[102]Open in a new tab
Transcriptional regulation of ER-regulated genes by R269C-ER in breast
and pancreatic cancer cells. MCF-7 (a, b) and Colo-357 cells (c, d)
were transiently transfected with either WT-ER or R269C-ER constructs
and treated with E2 (10nM) or a control vehicle for 24 h. mRNA levels
of GREB-1 and TGF-α were determined 48 h after transfection by
quantitative RT-PCR and are shown relative to the control WT-ER. The
results are from a representative experiment of at least three
independent experiments, each performed in triplicates, each bar
represents the mean ± SD. *, P < 0.05, **, P < 0.01
R269C-ER enhances AP-1 dependent transcriptional activity in breast and
pancreatic cancer cells
The hinge region has been suggested to mediate non-classical
transcription through interaction with the AF-1 domain and with
transcription factors regulating AP-1 activity (e.g. c-Fos/c-Jun, Sp1).
In order to analyze the effects of R269C-ER on AP-1 activity, an
AP-1-luciferase reporter was employed [[103]41]. COLO-357, PANC-1,
MCF-7 and HEK-293 cells were co-transfected with the AP-1-reporter and
either WT-ER or R269C-ER. Cells were grown in estrogen-depleted medium
with or without fulvestrant (ICI 182,780), which due to its AP-1
agonist activity, served as a positive control [[104]42, [105]43,
[106]54]. In comparison to WT-ER, in COLO-357 cells, R269C-ER increased
AP-1 transcriptional activity by 48%, in PANC-1 cells by 27%, in MCF-7
cells by 74% and in HEK-293 cells by 49% (Fig. [107]4a-d, p < 0.05 for
all comparisons).
Fig. 4.
[108]Fig. 4
[109]Open in a new tab
AP-1 dependent transcriptional activity of R269C-ER. Cells were
transiently transfected with either WT-ERα or R269C-ERα constructs
together with the AP-1 luciferase reporter and treated with ICI 182780
(100 nM) or a control vehicle for 24 h. Luciferase activities were
analyzed and normalized to Renilla luciferase units and are shown
relative to the control WT-ER. The results are from a representative
experiment of at least three independent experiments, each performed in
hexaplicates, each bar represents the mean ± SD. *, P < 0.05, **,
P < 0.01
Next, the effect of R269C-ER on the transcription of the AP-1-regulated
genes was examined. Expression of R269C-ER in COLO-357 cells increased
mRNA levels of both cyclin D1 and IGF-1R by 60 and 65%, respectively
(Fig. [110]5a-b, p < 0.01). Similarly, in PANC-1 cells we observed an
increase of cyclin D1 and IGF-1R mRNA by 64 and 62%, respectively (Fig.
[111]5c-d, p < 0.01). While statistically significant, the effect of
R269C-ER on MCF-7 cells was less pronounced: it decreased the levels of
cyclin D1 by 25%, and increased levels of IGF-1 by 25% (Fig. [112]5e-f,
p < 0.05). Taken together, these data indicate AP-1 mediated
transcriptional activity of R269C-ER, which is more profound in
pancreatic cancer cells compared to breast cancer cells.
Fig. 5.
[113]Fig. 5
[114]Open in a new tab
Transcriptional regulation of AP-1-regulated genes in breast and
pancreatic cancer cells by R269C-ER. COLO-1 cells (a, b) PANC-1 cells
(c, d) and MCF-7 cells (e, f) were transiently transfected with either
WT-ER or R269C-ER constructs and treated with a control vehicle for
24 h. mRNA levels of cyclin D1 and IGF1-R were determined 48 h after
transfection by quantitative RT-PCR. The results are from a
representative experiment of at least three independent experiments,
each performed in triplicates, each bar represents the mean ± SD. *,
P < 0.05, **, P < 0.01
Global transcriptomic analysis reveals cell-type dependent effect of R269C-ER
in breast and pancreatic cancer cells
In order to further delineate the transcriptional activity of 269C-ER,
a global transcriptomic analysis of MCF-7 and COLO-357 cells expressing
either R269C-ER or the WT-ER was performed. Compared to WT-ER,
expression of R269C-ER in MCF-7 resulted in differential regulation of
135 genes (44 upregulated, 91 downregulated) with most upregulated
genes associated with cell metabolism and hormone biosynthesis
(Fig. [115]6a, b). Interestingly, when treated with E2, an independent
set of 61 genes was differentially regulated, with only 3 genes shared
with untreated cells (Fig. [116]6b). Similar analysis of COLO-357 cells
indicated differential regulation of 15 and 45 genes in untreated and
E2-treated cells, respectively, with eight genes common to the two
groups (Fig. [117]6d, e). Importantly, genes regulated by R269C-ER in
COLO-357 cells were associated with a more aggressive behavior (Fig.
[118]6e). As we focused on the effect of R269C-ER on pancreatic cancer,
we searched for genes known to play a role specifically in this type of
cancer. We observed a 6-fold decrease in VASN gene expression (Fig.
[119]2S), a gene involved in the TGFβ pathway, as evidenced by String
analysis (Fig. [120]7a). Thus, we assessed TGFβ1 mRNA expression and
found it decreased in R269C-ER expressing cells upon E2 treatment
compared to WT-ER (p < 0.01, Fig. [121]7b). As TGFβ1 may play
suppressive role in pancreatic cancer development [[122]55, [123]56],
inhibition of this pathway, may be a possible mechanism by which
R269C-ER mediates aggressive behavior of pancreatic cancer cells.
Fig. 6.
[124]Fig. 6
[125]Open in a new tab
R269C-ER exhibits differential effect on gene expression in pancreatic
cancer cells compared to breast cancer MCF-7 and COLO-357 cells were
seeded in phenol red depleted medium with charcoal stripped serum.
Cells then were transfected with either WT-ER or R269C-ER in
triplicates and then treated with vehicle control or E2 (10nM) for
24 h. Total RNA was extracted and RNAseq was performed. a A heatmap of
differentially expressed genes, between untreated mutated and WT-ER, in
MCF-7 cells was generated. b A Venn-diagram of significantly (p < 0.05)
upregulated or downregulated genes in MCF-7 compared to MCF-7 cells
treated with E2 is depicted. c Pathway enrichment analysis was
conducted in MCF-7 cells treated with E2 (10nM), using Gene Analytics™.
Results show that R269C-ER overexpression is associated with different
metabolic pathways. d A heatmap of differentially expressed genes,
between untreated mutated and WT-ER, in COLO-357 cells was generated e
A Venn-diagram of significantly (p < 0.05) upregulated or downregulated
genes in COLO-357 compared to COLO-357 treated with E2 is depicted
Fig. 7.
[126]Fig. 7
[127]Open in a new tab
R269C-ER decreases TGFβ expression in pancreatic cancer. a A string
analysis of VASN gene is depicted. b MCF-7 and COLO-357 cells were
seeded in phenol red depleted medium with charcoal stripped serum.
Cells then were transfected with either WT-ER or R269C-ER in
triplicates and treated with vehicle control or E2 (10nM) for 24 h.
Total RNA was extracted and expression of TGFβ1 was assessed. Each
graph represents ± SD. *, P < 0.05, **, P < 0.01. Representative
experiment is shown
R269C-ER enhances migration of pancreatic but not of breast cancer cells
The effects of R269C-ER on proliferation of MCF-7, COLO-357 and PANC-1
cells were evaluated using MTT assay. For the analysis, cells were
transfected with WT-ER or R269C-ER, grown in estrogen-depleted medium,
and treated with either a vehicle control or E2. While R269C-ER
enhanced proliferation of COLO-357 in the basal state by 149%
(Fig. [128]8a, p < 0.01) and in PANC-1 by 16% (Fig. [129]8b, p = 0.03).
In contrary, it significantly decreased proliferation of MCF-7 cells by
12% (Fig. [130]8c, p = 0.02).
Fig. 8.
[131]Fig. 8
[132]Open in a new tab
Cell-type dependent effects of R269C-ER on cancer cell proliferation.
COLO-357 (a), PANC-1 (b) and MCF-7 (c) cells were transiently
transfected with either WT-ER or R269C-ER, seeded in 96-well plates and
grown in estrogen-depleted medium and after 24 h were treated with
either E2 (10nM) or a vehicle control for 72 h. Viability was assessed
using MTT assay and are shown relative to the control WT-ER. Figures
show representative results of at least three independent experiments,
each performed in hexaplicates, each bar represents the mean ± SD. *,
P < 0.05, **, P < 0.01
The ability of R269C-ER to enhance migration was assessed by wound
healing assay. Expression of R269C-ER significantly enhanced migration
of COLO-357 cells compared to WT-ER both after 24 h and 48 h by 123 and
90%, respectively, and with E2 treatment by 45 and 58%, respectively
(Fig. [133]9a, p < 0.01). Similar results were observed with
E2-untreated PANC-1 cells, after 24 h the migration was enhanced by 37%
and after 48 h by 25% (Fig. [134]9b, p < 0.01 and p < 0.05,
respectively). However, the mutation did not increase migration of
MCF-7 cells, but rather slowed migration compared to WT-expressing
cells (Fig. [135]9c).
Fig. 9.
[136]Fig. 9
[137]Open in a new tab
R269C-ER enhances migration of pancreatic cancer cells. COLO-357 (a)
PANC-1 (b) and MCF-7 (c) were seeded in 6-well plates and transfected
with either WT-ER, R269C-ER or pcDNA3 constructs. Cells were grown in
estrogen-depleted medium and the monolayer was scraped in a straight
line, then treated with E2 (10nM) or a vehicle control for 48 h. The
results are from a representative experiment of at least three
independent experiments. Mean values of at least 10 measurements for
each time point and condition are shown on the graphs and
representative photos are also presented, *, P < 0.05, **, P < 0.01
Discussion
While early works suggested a possible role for ER-α in the development
of pancreatic cancer [[138]16–[139]19, [140]57, [141]58], its role in
the development of this cancer is still controversial. To our
knowledge, this is the first report to identify a possible link between
this rare variant of the ER and pancreatic cancer development. While
this variant is not common, the possibility of treating even a small
subset of pancreatic cancer patients with hormonal therapy, which is
definitely less toxic and possibly more effective than standard
chemotherapy, may be of great value to patients. In addition, our study
highlights for the first time, mechanistic aspects of ER activation
related to the structure of the hinge region, a much less studied
region of the ER, and suggests a role for this region in mediating AP-1
transcriptional activity.
We sought to examine the prevalence and activity of R269C-ER, a rare
functional variant of ESR1, in breast and pancreatic cancers. Analysis
of a large genomic database indicated enrichment of this variant in a
small subset of non-PDAC pancreatic cancers, and functional analysis
suggested cell-type dependent activity of this variant. Thus, it showed
increased classic and AF-1-mediated transcriptional activity
specifically in pancreatic cancer cells while less in MCF-7 breast
cancer cells and accordingly enhanced proliferation and migration only
of pancreatic cancer cells. A transcriptomic analysis further
corroborated these results.
Pancreatic tumors are diverse in histological, molecular and biological
features. Both COLO-357 and PANC-1 originate from patients with PDAC,
but the phenotype and genotype of those cell lines are different. The
phenotype of COLO-357 cells remains unknown [[142]59] while the
genotype grade (G) is classified as G1-G2 [[143]60]. For PANC-1 cells
the phenotype is depict as ductal/acinar [[144]59], and the genotype is
classified as G3 [[145]61]. The features of those cell lines are useful
in understanding and evaluating the mutation impact on pancreatic
cancer, though should be carefully addressed, as there is much
heterogeneity even within the tumor characteristic. Furthermore, we
used ER-positive breast cancer cells (MCF-7) for deeper understanding
of the mutation and in effort to examine cells that their tumorgenicity
depends on the ER pathway.
R269 lies within the hinge region of ESR1, a region that has not been
studied extensively. Position 269 is highly conserved across species as
arginine or lysine and lies within a putative NLS region. Substitution
of arginine (a conserved negative charge) to cysteine is expected to
alter its interactions with both AF-1 and AF-2. In silico study via
multiple bioinformatics tools (Polyphen-2, SIFT and PROVEAN), predicted
with high probability that R269C substitution may damage the
structure-function of this protein [[146]62].
In agreement with these predictions, our data indicates altered
transcriptional activity of the variant compared to the WT receptor, an
effect that was even more pronounced toward AF-1 than toward AF-2.
These observations point to an important role of the hinge region in
maintaining proper transcriptional activity of the ER and may explain
the low frequency of alterations observed in this area in the general
population and among species. We noted a discrepancy between direct
transcriptional activity of R269C-ER, as determined using the ERE
reporter, and the effect on mRNA of ERE-regulated genes in MCF-7 cells.
This observation suggests a role for additional, possibly non-genomic
mechanisms regulating the activity of R269C-ER specifically in these
cells. The mutation lies within putative NLS2 and possibly affects the
activity of the receptor by modifying its subcellular localization,
however nuclear and cytoplasmic localization were not disrupted
compared to WT-ER (data is not shown).
Surprisingly, the effect of R269C variant was cell type specific. It
inhibited gene expression, proliferation and migration of MCF-7 cells.
The precise upstream regulatory elements leading to these differential
effects are yet to be determined. This observation explains the lack of
enrichment of R269C among breast cancer patients. Thus, expression of
R269C is probably not associated with increased breast cancer risk. On
the other hand, expression of R269C had profound effect on pancreatic
cancer cells. It enhanced transcriptional activity of both classical
and AF-1-regulated genes, enhanced proliferation and promoted migration
of two pancreatic cancer cell lines. These observations may explain the
enrichment of R269C among a subset of pancreatic cancers. It also
highlights the need to study the effects of this mutation in additional
pancreatic cancer subtype models.
A global transcriptome analysis, using RNAseq, indicated differential
gene expression among breast and pancreatic cancer cell lines
expressing the mutated ER. Thus, expression of the R269C-ER was
associated with ligand-independent upregulation of genes associated
with cell metabolism and hormone biosynthesis, and not with growth or
invasion. This observation may explain the lack of association between
the presence of this SNP and increase breast cancer risk. On the one
hand, alterations of genes associated with tumorogenicity were observed
in COLO-357 cells. These included upregulation SERPINB2, a known
oncogene [[147]63–[148]65], and a decrease in TGFβ1 expression, a
cytokine which may function under specific circumstances as a tumor
suppressor in pancreatic cancer [[149]66–[150]68]. Indeed, Hezel et al.
showed that TGF-β or αvβ6 blockade increased pancreatic tumor cell
proliferation and accelerated both early and later disease stages
[[151]55]. Importantly, these observations indicate, for the first
time, a direct role of the hinge region in regulating transcriptional
activity of the ER.
Previous studies, published mostly during the 1980s and 1990s, explored
the role of anti-estrogens as a potential treatment for pancreatic
cancer. While pre-clinical studies were promising [[152]69], clinical
trials showed conflicting results, partly due to under power of the
trials and partly may be attributed to the inclusion of all pancreatic
cancer patients, regardless expression of the ER or the presence of
functional variants. Importantly, while our study indicated increased
activity of this variant, at least part of the increased activity is
mediated through the AF-1 and is E2 independent. Therefore, it is
likely that current hormonal therapies targeting the ER, either by
reducing estrogen levels (aromatase inhibitors) or by inhibition of its
classic activity (tamoxifen) may not be effective and other
manipulations should be examined. The frequency of the variant, even
among non-PDAC patients is low (~ 0.5%). Therefore, the conduction of
clinical trials on this specific patients’ population is likely to be
very challenging. Yet, considering the grim prognosis and limited
treatment options for patients with pancreatic cancer, the ability to
target even a small fraction of patients is of major significance.
Conclusion
Our data suggest a role for R269C functional variant of the ESR1 in the
development of a small subset of non-PDAC pancreatic cancers. Moreover,
our data suggest that this variant may have a protective effect against
breast cancer. The underlying mechanism should be further explored.
Supplementary information
[153]12885_2020_7005_MOESM1_ESM.pdf^ (80.1KB, pdf)
Additional file 1: Figure S1. Expression of R269C-ER in pancreatic
cancer cells. PANC-1 (A) or COLO-357 (B) cells were transfected with
either pCDNA3, WT-ER or R269C-ER grown in estrogen-depleted medium and
treated with E2 (10nM) or a control vehicle for 24 h. Cells were
harvested, lysed and analyzed by Western blotting. The results of (A)
and (B) are from a representative experiment of n = 3.
[154]12885_2020_7005_MOESM2_ESM.pdf^ (40.2KB, pdf)
Additional file 2: Figure S2. R269C-ER exhibits differential effect
following E2 treatment on gene expression in pancreatic cancer cells. A
list of genes down and upregulated by R269C-ER compared to WT-ER
following E2 treatment in COLO-357 cells.
[155]12885_2020_7005_MOESM3_ESM.pdf^ (56.3KB, pdf)
Additional file 3: Figure S3. R269C-ER exhibits differential effect on
gene expression in pancreatic cancer cells compared to breast cancer.
MCF-7 and COLO-357 cells were seeded in phenol red depleted medium with
charcoal stripped serum. Cells then were transfected with either WT-ER
or R269C-ER in triplicates and then treated with vehicle control or E2
(10nM) for 24 h. Total RNA was extracted and RNAseq was performed. (A,
B) A heatmap of differentially expressed genes in MCF-7 cells treated
with E2 (A) and COLO-357 cells treated with E2 (B) was generated.
Acknowledgements