Circulating microRNA Profiling Identifies a Subset of Metastatic Prostate Cancer Patients with Evidence of Cancer-Associated Hypoxia.
ABSTRACT MicroRNAs (miRNAs) are small (∼22 nucleotide) non-coding RNAs that regulate a myriad of biological processes and are frequently dysregulated in cancer. Cancer-associated microRNAs have been detected in serum and plasma and hold promise as minimally invasive cancer biomarkers, potentially for assessing disease characteristics in patients with metastatic disease that is difficult to biopsy. Here we used miRNA profiling to identify cancer-associated miRNAs that are differentially expressed in sera from patients with metastatic castration resistant prostate cancer (mCRPC) as compared to healthy controls. Of 365 miRNAs profiled, we identified five serum miRNAs (miR-141, miR-200a, miR-200c, miR-210 and miR-375) that were elevated in cases compared to controls across two independent cohorts. One of these, miR-210, is a known transcriptional target of the hypoxia-responsive HIF-1α signaling pathway. Exposure of cultured prostate cancer cells to hypoxia led to induction of miR-210 and its release into the extracellular environment. Moreover, we found that serum miR-210 levels varied widely amongst mCRPC patients undergoing therapy, and correlated with treatment response as assessed by change in PSA. Our results suggest that (i) cancer-associated hypoxia is a frequent, previously under-appreciated characteristic of mCRPC, and (ii) serum miR-210 may be further developed as a predictive biomarker in patients with this distinct disease biology.
- SourceAvailable from: Prasun Mishra[Show abstract] [Hide abstract]
ABSTRACT: Developing more precise diagnostics approaches to predict cancer progression and prognosis is the key to precision medicine. Overwhelming evidence now suggests that small non-coding RNAs such as miRNAs can be useful tools as biomarkers for molecular diagnostics. miRNAs can serve as biomarkers in a variety of diseases, such as neurological disorders, cardiovascular disease, Type II diabetes, cancer and so on. miRNAs can not only be utilized for monitoring treatment but also for patient stratification and hence are promising predictive biomarkers in cancer progression and prognosis, as well as in predicting drug response. This article focuses on some of the recent findings in the field of miRNA biomarkers and discusses its implications for cancer diagnostics and precision medicine.Expert Review of Molecular Diagnostics 11/2014; 14(8):917-9. · 4.09 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Nucleic-acid amplification is a crucial step in nucleic-acid-sequence-detection assays. The use of digital microfluidic devices to miniaturize amplification techniques reduces the required sample volume and the analysis time and offers new possibilities for process automation and integration in a single device. The recently introduced droplet polymerase-chain-reaction (PCR) amplification methods require repeated cycles of two or three temperature-dependent steps during the amplification of the nucleic-acid target sequence. In contrast, low-temperature isothermal-amplification methods have no need for thermal cycling, thus requiring simplified microfluidic-device features. Here, the combined use of digital microfluidics and molecular-beacon (MB)-assisted isothermal circular-strand-displacement polymerization (ICSDP) to detect microRNA-210 sequences is described. MicroRNA-210 has been described as the most consistently and predominantly upregulated hypoxia-inducible factor. The nmol L(-1)-pmol L(-1) detection capabilities of the method were first tested by targeting single-stranded DNA sequences from the genetically modified Roundup Ready soybean. The ability of the droplet-ICSDP method to discriminate between full-matched, single-mismatched, and unrelated sequences was also investigated. The detection of a range of nmol L(-1)-pmol L(-1) microRNA-210 solutions compartmentalized in nanoliter-sized droplets was performed, establishing the ability of the method to detect as little as 10(-18) mol of microRNA target sequences compartmentalized in 20 nL droplets. The suitability of the method for biological samples was tested by detecting microRNA-210 from transfected K562 cells.Analytical and Bioanalytical Chemistry 01/2015; · 3.66 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: In the current study, expression levels of let-7c, miR-30c, miR-141, and miR-375 in plasma from 59 prostate cancer (PC) patients with different clinicopathological characteristics and two groups of controls: 16 benign prostatic hyperplasia (BPH) samples and 11 young asymptomatic men (YAM) were analyzed to evaluate their diagnostic and prognostic value in comparison to prostate-specific antigen (PSA). miR-375 was significantly downregulated in 83.5% of patients compared to BPH controls and showed stronger diagnostic accuracy (area under the curve [AUC]=0.809, 95% CI: 0.697-0.922, p=0.00016) compared with PSA (AUC=0.710, 95% CI: 0.559-0.861, p=0.013). Expression levels of let-7c showed potential to distinguish PC patients from BPH controls with AUC=0.757, but the result did not reach significance. Better discriminating performance was observed when combinations of studied biomarkers were used. Sensitivity of 86.8% and specificity of 81.8% were reached when all biomarkers were combined (AUC=0.877) and YAM were used as calibrators. None of the studied microRNAs (miRNAs) showed correlation with clinicopathological characteristics. PSA levels were significantly correlated with the Gleason score, tumor stage, and lymph node metastasis with Spearman correlation coefficients: 0.612, 0.576, and 0.458. In conclusion, the combination of the studied circulating plasma miRNAs and serum PSA has the potential to be used as a noninvasive diagnostic biomarker for PC screening outperforming the PSA testing alone.DNA and Cell Biology 12/2014; · 1.99 Impact Factor
Circulating microRNA Profiling Identifies a Subset of
Metastatic Prostate Cancer Patients with Evidence of
Heather H. Cheng1,2., Patrick S. Mitchell3.¤a, Evan M. Kroh3, Alexander E. Dowell4, Lisly Che ´ry4,
Javed Siddiqui5,6, Peter S. Nelson1,2,3, Robert L. Vessella4,7, Beatrice S. Knudsen8¤b,
Arul M. Chinnaiyan5,9,10, Kenneth J. Pienta11, Colm Morrissey4, Muneesh Tewari1,2,3,8*
1Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, United States of America, 2Department of Medicine, University of Washington,
Seattle, Washington, United States of America, 3Human Biology Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, United States of America,
4Department of Urology, University of Washington, Seattle, Washington, United States of America, 5Michigan Center for Translational Pathology, University of Michigan
Medical School, Ann Arbor, Michigan, United States of America, 6Department of Pathology, University of Michigan Medical School, Ann Arbor, Michigan, United States of
America, 7Department of Veterans Affairs Medical Center, Seattle, Washington, United States of America, 8Public Health Sciences Division, Fred Hutchinson Cancer
Research Center, Seattle, Washington, United States of America, 9Department of Urology, University of Michigan Medical School, Ann Arbor, Michigan, United States of
America, 10Bioinformatics Program, University of Michigan Medical School, Ann Arbor, Michigan, United States of America, 11Departments of Urology, Oncology and
Pharmacology and Molecular Sciences, Johns Hopkins Hospital, Baltimore, Maryland, United States of America
MicroRNAs (miRNAs) are small (,22 nucleotide) non-coding RNAs that regulate a myriad of biological processes and are
frequently dysregulated in cancer. Cancer-associated microRNAs have been detected in serum and plasma and hold
promise as minimally invasive cancer biomarkers, potentially for assessing disease characteristics in patients with metastatic
disease that is difficult to biopsy. Here we used miRNA profiling to identify cancer-associated miRNAs that are differentially
expressed in sera from patients with metastatic castration resistant prostate cancer (mCRPC) as compared to healthy
controls. Of 365 miRNAs profiled, we identified five serum miRNAs (miR-141, miR-200a, miR-200c, miR-210 and miR-375) that
were elevated in cases compared to controls across two independent cohorts. One of these, miR-210, is a known
transcriptional target of the hypoxia-responsive HIF-1a signaling pathway. Exposure of cultured prostate cancer cells to
hypoxia led to induction of miR-210 and its release into the extracellular environment. Moreover, we found that serum miR-
210 levels varied widely amongst mCRPC patients undergoing therapy, and correlated with treatment response as assessed
by change in PSA. Our results suggest that (i) cancer-associated hypoxia is a frequent, previously under-appreciated
characteristic of mCRPC, and (ii) serum miR-210 may be further developed as a predictive biomarker in patients with this
distinct disease biology.
Citation: Cheng HH, Mitchell PS, Kroh EM, Dowell AE, Che ´ry L, et al. (2013) Circulating microRNA Profiling Identifies a Subset of Metastatic Prostate Cancer
Patients with Evidence of Cancer-Associated Hypoxia. PLoS ONE 8(7): e69239. doi:10.1371/journal.pone.0069239
Editor: Rajvir Dahiya, UCSF/VA Medical Center, United States of America
Received April 23, 2013; Accepted June 6, 2013; Published July 30, 2013
Copyright: ? 2013 Cheng et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Funding was provided by Canary Foundation (to M.T.), Damon Runyon-Rachleff Innovation Award (to M.T.), DOD Prostate Cancer New Investigator
Award (to M.T.), National Institutes of Health (NIH) CA093900 (to K.J.P.), NIH CA143055 (to K.J.P.), NIH PO1 CA085859 (to R.L.V.), NIH SPORE P50 CA69568 (to K.J.P.
and A.M.C), NIH SPORE P50 CA97186 (to P.S.N., M.T., and R.L.V.), NIH TR01 5R01DK085714 (to M.T.), NIH T32CA009515-28 (to H.H.C.), Prostate Cancer Foundation
Creativity Award (to M.T.), and Richard M. Lucas Foundation (to R.L.V.). The funders had no role in study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: Muneesh Tewari is a named inventor on the pending patent application entitled, ‘‘Use of Extracellular RNA to Measure Disease,’’ #12/
993828. H.H.C., P.S.M., E.M.K., A.E.D., L.C., J.S., P.S.N., B.S.K., R.L.V., A.M.C., K.J.P., and C.M. declare no competing interests. The authors confirm that this does not
alter their adherence to all the PLOS ONE policies on sharing data and materials.
* E-mail: email@example.com
¤a Current address: Molecular and Cellular Biology Graduate Program, University of Washington, Seattle, Washington, United States of America; Division of Basic
Sciences, Fred Hutchinson Cancer Research Center, Seattle, Washington, United States of America
¤b Current address: Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, Los Angeles, California, United States of America
. These authors contributed equally to this work.
The natural history of patients with metastatic castration
resistant prostate cancer (mCRPC) varies widely, suggesting a
heterogeneous disease biology in this patient population. However,
little is known about biological heterogeneity in mCRPC and
approaches for clinical stratification of patients are limited, in large
part because metastatic tissue is not routinely available for study.
Minimally invasive (e.g., blood-based) approaches that can help
stratify patients on the basis of distinct tumor biology could help
inform choice of therapy, which is especially relevant with the
recent introduction of multiple new effective treatments for
Circulating microRNAs (miRNAs) are an emerging class of
blood-based biomarkers with the potential to provide information
about distinct tumor biology in individual patients . miRNAs
are small (,22 nucleotide), non-coding RNA molecules that post-
PLOS ONE | www.plosone.org1July 2013 | Volume 8 | Issue 7 | e69239
transcriptionally regulate gene expression by translational repres-
sion or degradation of targeted transcripts . Specific miRNAs
have been found to regulate a variety of critical processes in tumor
physiology, including angiogenesis , epithelial-to-mesenchymal
transition , metastasis  and the tumor response to hypoxia
. miRNAs have been demonstrated to be effective tissue-based
cancer biomarkers–in diagnosing cancers of unknown tissue origin
and in predicting clinical outcomes[7–9]. We and others have
demonstrated that circulating, cell-free, tumor-derived miRNAs
are highly stable and detectable in the serum of cancer patients
. In earlier work, we used a candidate miRNA approach to
identify significant elevation of miR-141 in serum from patients
with metastatic castration-resistant prostate cancer (mCRPC) .
In order to more comprehensively identify prostate cancer-
associated circulating miRNAs, in the current study we profiled
serum miRNAs from patients with mCRPC. We identified five
serum miRNAs that were significantly elevated in cases compared
to healthy controls, including the hypoxia-associated miR-210. In
order to determine whether prostate cancer cells can release miR-
210 in response to hypoxia, we exposed prostate cancer cell lines
to hypoxic conditions and found that miR-210 was induced and
released into the extracellular environment. In analysis of clinical
specimens and outcomes, we found evidence that a subset of
mCRPC patients have increased miR-210 levels and that this
correlates with treatment response, suggesting that increased
hypoxia is a feature of mCRPC that may define a subset of
patients with a distinct disease biology.
Materials and Methods
LNCaP (ATCCH CRL-1740TM) and VCaP  human
prostate cancer cell lines were cultured in RPMI 1640 and
DMEM, respectively, each supplemented with 10% FBS (or under
serum-free conditions, as noted), at 37uC in a 5% CO2incubator.
Hypoxic conditions (1% O2) were established in a Thermo
Scientific 3595 Incubator (ThermoFisher), with cells maintained
under normoxic conditions (20% O2) in parallel.
RNA Isolation from Cultured Cells and Conditioned Media
Conditioned media was removed from cells cultured for 24, 48
or 72 hours under normoxic or hypoxic conditions. Cells were
washed with 5 ml PBS and lysed on ice directly in the culture dish
with 600 ml Lysis/Binding buffer from the mirVana miRNA
isolation kit (Ambion). Lysates were harvested manually with a
sterile cell scraper and transferred to an RNase2/DNase-free
2 ml microcentrifuge tube. RNA was extracted from cell lysates
following the manufacturer’s recommended protocol for total
RNA isolation. Cellular debris was removed from a 500 ml aliquot
of conditioned media (10 ml total volume) by filtration through a
0.2 mm NanoSep filtration unit (Millipore) at 14,0006g, 5 min, at
room temperature. 400 ml filtered sample was combined with
400 ml 2X Denaturing Solution (Ambion) and vortexed. C. elegans
spiked-in oligonucleotides were introduced (as a mixture of
25 fmol of each oligonucleotide in 5 ml total volume per liquid
sample) after denaturation and used for normalization of
variability in RNA isolation across samples as previously described
. RNA was extracted from conditioned media lysates using the
mirVana PARIS kit (Ambion) following the manufacturer’s
recommended protocol for total RNA isolation.
All clinical samples were obtained from subjects who provided
written informed consent. Studies were performed in accordance
with the declaration of Helsinki guidelines and with ethics
approval from the Institutional Review Boards at the University
of Washington, Fred Hutchinson Cancer Research Center and
University of Michigan.
Blood Processing and Isolation of Serum from Clinical
All clinical samples obtained at the University of Washington
and the University of Michigan were collected and processed as
previously described [1,11].
RNA Isolation from Clinical Serum Samples
Total RNA was isolated from 400 ml serum collected at the
University of Washington using the mirVana PARIS kit (Ambion)
as previously described . Equal volumes of each sample type
were pooled to create mCRPC and healthy donor serum pools
(n=25 for each pool) for TLDA profiling. Total RNA was isolated
from serum samples collected at the University of Michigan using
the miRNeasy RNA isolation kit (Qiagen) as follows: 400 ml serum
was divided into four, 100 ml aliquots. Each aliquot was denatured
using 10X volume (1 ml) Qiazol, which was vortexed and
incubated at room temperature for 10 min. C. elegans spiked-in
oligonucleotides were introduced (as a mixture of 25 fmol of each
oligonucleotide in 5 ml total volume per liquid sample) after
denaturation, which were used for normalization of variability in
RNA isolation across samples as previously described . RNA
was extracted using 0.2X volume chloroform (220 ml), and total
RNA was isolated following the manufacturer’s protocol. For a
given sample, RNA isolated from each 100 ml aliquot was pooled
and concentrated to 100 ml volume over Microcon YM-3 filter
units (Millipore) at 14,0006g, 1.5 hour, 4uC, which were loaded
inverted into pre-weighed 1.5 ml microcentrifuge tubes and eluted
at 10006g, 3 min, 4uC. Tubes plus eluate was weighed on an
analytical scale and brought to 100 ml with Elution Buffer. RNA
was stored at 280uC.
Collection and Processing of Clinical Tissue Sections
Laser-capture micro-dissection (LCM) of frozen-tissue
Sections of flash-frozen prostate and lymph node
obtained from radical prostatectomy and rapid autopsy, respec-
tively, were assessed by a pathologist to define regions of tumor
epithelial cells. For laser capture microdissection 5 mm sections of
frozen tissue were made on a LeicaTMCM3050S cryostat at
220uC (Leica, Wetzlar, Germany), placed onto PEN Membrane
Frame Slides (MDS Inc., Ontario, Canada) then stained and fixed
according to the HistoGeneTMLCM Frozen Section Staining Kit
protocol (MDS Inc.). For each sample approximately 2000 cells
were laser captured at 200x magnification (Veritas, Arcturus MDS
Inc., Ontario, Canada). The captured sample was placed in 300 ml
lysis buffer from the RNAqueousH RNA Isolation Kit (Ambion,
Austin, Texas), incubated for 30 minutes at 42uC and stored at
280uC until RNA isolation was performed. miRNA was then
isolated using the RNAqueousH RNA Isolation Kit (Ambion).
RNA isolation from LCM tissue samples.
isolated from LCM tissue samples using the RNAqueous-Micro
RNA isolation kit (Ambion) as follows: Frozen lysates were thawed
on ice, vortexed and centrifuged at 16,1006g, 30 sec, room
temperature. C. elegans spiked-in oligonucleotides were introduced
(as a mixture of 25 fmol of each oligonucleotide in 5 ml total
volume per liquid sample) and used for normalization of variability
in RNA isolation across samples as previously described ,
followed by addition of 3 ml LCM Additive. RNA was precipitated
from the lysate mixture with 1.25 volumes 100% molecular-grade
Total RNA was
Circulating MiRNAs and Hypoxia in Prostate Cancer
PLOS ONE | www.plosone.org2 July 2013 | Volume 8 | Issue 7 | e69239
Figure 1. Serum miRNA profiling and validation. (A) Measurement of circulating miRNAs in sera pooled from patients with advanced prostate
cancer as compared to healthy donors (comprising a Discovery Set) by TLDA profiling. Blue- and brown-filled circles represent serum miRNAs
increased or decreased (with unadjusted p-value ,0.05), respectively, in mCRPC patients compared to healthy controls. Inset: Nine miRNAs
demonstrated .5-fold change (unadjusted P,0.05, Student’s t-test). FC, fold-change. (B) Confirmation of mCRPC-associated serum miRNAs in
individual samples from the Discovery Set from the University of Washington samples. Upper: miRNA biomarker candidates were measured in
individual samples by TaqMan miRNA qRT-PCR (P value assigned by Wilcoxon signed-rank test), where miRNA abundance is given in terms of miRNA
copies/ml serum. Red bars, mean +/2 SEM of miRNA copies/ml serum for each group. Lower: Receiver operating characteristic (ROC) curves plot
Circulating MiRNAs and Hypoxia in Prostate Cancer
PLOS ONE | www.plosone.org3 July 2013 | Volume 8 | Issue 7 | e69239
EtOH, and was subsequently bound to the Micro Filter Cartridge
assembly (prewet with 30 ml Lysis Solution for 5 min) by
centrifugation at 10,0006g, 1 min. The filter was washed
(180 ml Wash Solution 1, 10,0006g, 1 min; 26180 ml Wash
Solution 2/3, 16,1006g, 30 sec; air only, 16,1006g, 30 sec). RNA
was eluted from column twice with 10 ml 95uC Elution Buffer into
pre-weighed tubes. Eluates were weighed on an analytical scale
and brought to 20 ml with Elution Buffer. RNA was stored at
MicroRNA Profiling using TaqMan Low-Density Array
miRNA qRT-PCR and Biomarker Candidate Selection
RNA from pooled serum of mCRPC patients or healthy
controls (comprised of equal RNA volume from each of 25
samples per pool) was reverse-transcribed in duplicate reactions
from 2 ml of pooled RNA using the TaqMan miRNA Reverse
Transcription Kit and the TaqMan miRNA Multiplex RT Assays
(Human Pool Set). A pooled sample approach was chosen for cost-
efficient discovery of miRNA biomarkers. miRNA expression was
profiled from each RT reaction replicate using the TaqMan Low-
Density Array (TLDA, v1.0) as previously described . Multiplex
reverse-transcription TLDA qRT-PCR was carried out on an
Applied BioSystems 7900HT thermocycler using the manufactur-
er’s recommended cycling conditions. Data were analyzed with
SDS Relative Quantification Software version 2.2.2 (Applied
BioSystems), with an automatically assigned minimum threshold,
which was above the baseline of all assays showing measurable
amplification above background. Values that were below the
minimum threshold were arbitrarily assigned a cycle threshold
(CT) value of 40. P-values were assigned by Student’s t-test
evaluating replicate profiling data from each pool to determine
significant differences in miRNA expression between mCRPC
pool and healthy control pool RNA samples. Fold-change (FC)
values were derived by computing 2‘ (AveCTmCRPC
AveCThealthy control pool). MicroRNA candidate biomarkers were
selected by criterion of FC .5 and P,0.05.
Measurement of miRNA and mRNA Levels by Individual
TaqMan Quantitative Reverse-Transcription PCR (qRT-
miRNA-derived from serum samples was reverse-transcribed
using the TaqMan miRNA Reverse Transcription kit (Applied
BioSystems) and quantified by TaqMan miRNA qRT-PCR using
miRNA-specific primer/probe sets (Applied BioSystems) as
previously described . A complete description of TaqMan
assays used in this study is provided in Table S4.
sensitivity vs. (1 - specificity) to assess the ability of each miRNA biomarker to distinguish cases from controls. (C) Validation of mCRPC-associated
serum miRNAs in an independent Validation Set. Upper: Serum concentration (copies/ml) of miR-141, miR-375, miR-200c, miR-200a and miR-210 was
measured by TaqMan miRNA qRT-PCR. Dot-plot associated P values were assigned by Wilcoxon signed-rank test. Dot plots and ROC curves were
generated as described for Fig. 1. Lower: Red, results from the validation sample set obtained from the University of Michigan. Black, results from the
primary sample set obtained from the University of Washington reproduced from Fig. 1B, lower. AUC, area under the curve; mCRPC, prostate cancer
patient sera; FC, fold-change; CTL, control sera (from age-matched male individuals with normal PSA and negative digital rectal exam).
Figure 2. Exposure of prostate cancer cell lines to hypoxia induces production and release of miR-210 into the extracellular
environment. Left column, miR-210 copies/ng RNA in LNCaP and VCaP human prostate cancer cell lines cultured in normoxic (20% O2) (white bars)
or hypoxic (1% O2) (blue bars) conditions for 24, 48 or 72 hours. Right column, miR-210 copies/ml in filtered conditioned media corresponding to
cellular samples. *, P value ,0.05; **, P value ,0.01; ***, P value ,0.001 (Student’s t-test).
Circulating MiRNAs and Hypoxia in Prostate Cancer
PLOS ONE | www.plosone.org4 July 2013 | Volume 8 | Issue 7 | e69239
miRNAs derived from serum samples obtained from the
University of Michigan were quantified by TaqMan miRNA
qRT-PCR using synthetic miRNA standard curves for absolute
quantification identically to that described for the University of
Washington sample set  with the exception that the pre-
amplification step was excluded from all miRNA quantification
other than miR-210 (this was based on empiric evidence that pre-
amplification does not increase the absolute copy number of
miRNA detectable for other TaqMan miRNA assays used here,
data not shown).
Gene-expression was quantified by TaqMan qRT-PCR as
follows: input RNA was reverse transcribed using the TaqMan
Gene-Expression Reverse Transcription Kit (Applied BioSystems)
in a small-scale RT reaction [comprised of 0.5 ml of 10X Reverse-
Transcription Buffer, 0.5 ml 10X Random Primers, 0.2 ml
100 mM dNTPs with dTTP, 0.25 ml Multiscribe Reverse-Tran-
scriptase and 3.55 ml input RNA; components other than the input
RNA were prepared as a larger volume master mix], using a
Tetrad2 Peltier Thermal Cycler (BioRad) at 50uC for 2 min,
followed by 30 cycles of 95uC for 10 min, 95uC for 15 sec and
60uC for 60 sec. Resultant cDNA was combined with 3.75 ml pre-
amplification assay reagents [comprised of 2.5 ml TaqMan
PreAmp Master Mix and 1.25 ml TaqMan Gene-Expression
Assay (GUSB or KRT18) diluted 1:100 in TE; components other
than the input cDNA were prepared as a larger volume master
mix to generate a 5.0 ml total volume PCR reaction], using a
Tetrad2 Peltier Thermal Cycler (BioRad) at 95uC for 10 min
followed by 14 cycles of 95uC for 15 sec and 60uC for 4 min.
GUSB or KRT18 pre-amplification reaction products were
combined with 2.75 ml of PCR assay reagents [comprised of
2.5 ml TaqMan 2X Universal PCR Master Mix, No AmpErase
UNG and 0.25 ml 20X TaqMan Gene-Expression Assay) to
generate a 5.0 ml total volume PCR reaction for GUSB or
KRT18, respectively. Real-time PCR was carried out on an
Applied BioSystems 7900HT thermocycler at 95uC for 10 min,
followed by 40 cycles of 95uC for 15 sec and 60uC for 1 min. Data
were analyzed with SDS Relative Quantification Software version
2.2.2 (Applied BioSystems), with the automatic CT setting for
assigning baseline and threshold for CT determination (See Table
S3 for results of messenger RNA measurements).
Results and Discussion
In order to efficiently discover serum miRNAs that are
differentially abundant between cancer cases vs. controls, we used
real-time PCR-based miRNA TaqMan Low-Density Arrays
(TLDA) to screen for differential abundance of 365 miRNAs in
serum RNA pooled from mCRPC patients (n=25) versus age-
matched controls (n=25; all controls had normal PSA and normal
digital rectal exam findings). After normalization of data using
spike-in control miRNAs and comparison of pooled serum RNA
from mCRPC cases to that of controls (Fig. 1A and inset), we
identified nine miRNAs demonstrating a greater than 5-fold
change in abundance (unadjusted P,0.05, Student’s t-test). We
next individually measured these nine miRNAs from the serum
RNA of individual mCRPC cases and controls using miRNA-
specific Taqman qRT-PCR assays. Serum levels of five miRNAs
Figure 3. Relationship between serum miR-210 levels and PSA
response in patients with metastatic castration resistant
prostate cancer. Upper: miR-210 copies/ml serum versus %PSA
change/day. %PSA change/day represents a measure of response to
treatment and was calculated using available clinical PSA values
measured most recently prior to, and at the time of serum miR-
210 draw. Mean time elapsed between the two blood draws was
30 days. Closed circles represent the subset of patients defined as
"miRNA-high" based on higher abundance of mCRPC-associated serum
miRNAs compared to all control individuals (as described in Results and
Discussion). Open circles represent the patients with mCRPC who did
not meet the definition for ‘‘miRNA-high’’. Solid line represents trend
line of the miRNA-high patient subset, dotted line represents trend line
of all patients. Middle: miR-210 copies/ml serum in patients with either a
PSA Response (R) or No PSA Response (NR). PSA Response is defined as
a decreasing or stable PSA (any change less than a 25% increase), and
No PSA Response is defined as a PSA increase of 25% or more, similar to
the Prostate Cancer Working Group criteria. Lower: Copies/ml serum of
miR-141, miR-200a, miR-200c and miR-375 in patients, R and NR.
Circulating MiRNAs and Hypoxia in Prostate Cancer
PLOS ONE | www.plosone.org5 July 2013 | Volume 8 | Issue 7 | e69239
were confirmed to be significantly elevated in mCRPC cases
P=0.007, miR-200c: P=0.017, miR-375: P=0.009, and miR-
210: P=0.022, Wilcoxon signed-rank analysis). The average fold-
difference between cases and controls ranged from 4.6 (miR-375)
to 27.9 (miR-141) (Fig. 1B, upper and Table S1). In addition,
receiver-operating characteristic (ROC) plots demonstrate the
capacity of these miRNAs to discriminate between the two groups
(miR-141 Area Under the Curve (AUC)=0.899; miR-200a
AUC=0.699; miR-375 AUC=0.773; miR-200c AUC=0.721
and miR-210 AUC=0.678) (Fig. 1B, lower). Importantly, we
verified that control miRNAs were not differentially expressed
between the two populations (Fig. S1).
To validate these findings in an independent specimen set
collected at a different institution, we measured miR-141, miR-
200a, miR-200c, miR-375 and miR-210 from the sera of an
additional 21 mCRPC patients and 20 age-matched healthy
controls collected at the University of Michigan. All five miRNAs
were elevated in sera from mCRPC cases relative to controls in
this independent validation set. MiR-141, miR-375 and miR-
210 were significant at a P-value threshold of ,0.01 in the second
cohort (P=0.001, P=0.021, P=0.022, respectively) and miR-
200a and miR-200c tended toward significance (P=0.073,
P=0.055, respectively) (Fig. 1C, upper). ROC curves were
generally concordant between the specimen sets from the two
institutions (Fig. 1C, lower). Analysis of serum miRNA markers in
various combinations demonstrated that adding
miRNAs to serum miR-141 (which had the best performance
alone) did not improve the ability to distinguish between cases and
controls (data not shown). Consistent with this observation, we
found that among cancer cases in which expression of miR-141,
miR-200a, miR-200c, and miR-375 was higher than all healthy
controls, these miRNAs were also significantly correlated with
each other and with serum PSA (Table S2). In contrast, miR-
210 did not show significant correlation with any of these four
miRNAs (Table S3) nor with serum PSA, suggesting that it
provides distinct information about disease biology.
To determine whether the five serum miRNA markers of
mCRPC are expressed in prostate cancer tissue and could
therefore be plausibly cancer cell-derived, we measured their
expression in epithelial cells that had been laser capture micro-
dissected from primary prostate cancer tissues (n=8) and lymph
node metastases (n=8). We detected miR-141, miR-200a, miR-
200c, miR-375 and miR-210 in all tissue types evaluated (Table
S3), suggesting that these five miRNAs, when found in the
circulation, may originate from prostate cancer, although other
additional sources cannot be excluded.
Three of the serum prostate cancer-associated miRNAs
identified (miR-141, miR-200a and miR-200c) are epithelial-
specific, highly related in sequence and have known roles in
maintaining the epithelial state by suppression of the epithelial-to-
mesenchymal transition . We hypothesize that elevated
circulating levels of miR-141, miR-200a and miR-200c reflect
the epithelial origin of prostate cancer cells.
The presence of elevated circulating miR-141 and miR-375 in
mCRPC patients has also been observed in recent mCRPC
circulating miRNA biomarker studies [12–16]. Interestingly,
elevated miR-210 was not reported in these other studies, despite
the fact that we observed this in independent specimen sets from
two different institutions. This could be due to different
comparison groups used (e.g., localized prostate cancer rather
than healthy controls as the comparator to mCRPC), the use of
plasma rather than serum, differences in the data analytic
approach used to identify differentially expressed miRNAs, as
well as potential differences in the clinical characteristics of the
mCRPC patients across different studies.
The elevated levels of miR-210 in serum from patients with
mCRPC was particularly interesting because this miRNA is well-
known to be transcriptionally activated by the hypoxia-inducible
factor 1 alpha (HIF-1a) [17,18] and may contribute to adaptation
to hypoxia in tumors [19,20]. This raises the possibility that miR-
210 is produced and released by hypoxic cells in the prostate
cancer (and/or by the tumor microenvironment), a potential
explanation for elevated levels of miR-210 we observed in the
serum of a subset of patients with mCRPC.
To test whether hypoxia can stimulate production and release of
miR-210 in prostate cancer cells, we characterized miR-210
abundance in LNCaP and VCaP prostate cancer cell lines (as
well as in filtered conditioned media) under normoxic (20% O2)
and hypoxic (1% O2) conditions over a 72-hour time course
(Fig. 2). miR-210 levels were increased by hypoxia compared to
normoxia with an initial induction in LNCaP cells followed by a
subsequent increased level in the conditioned media (Fig. 2). In
VCaP cells, we did not observe the same increase in miR-210
copies/ng of RNA and the levels dropped at 72 hours. We
speculate that this could be due to cell death or, alternatively, that
the regulation of miR-210 in response to hypoxia in VCaP cells
may be primarily occurring at the level of release. However, we
did observe a stepwise, time-dependent increase in the level of
extracellular miR-210 in the conditioned media of VCaP cells
(Fig. 2). Taken together, the results indicate that elevated levels of
miR-210 detected in serum could reflect tumor hypoxia.
Tumor hypoxia is a well-characterized process that contributes
to cancer progression and metastasis in many human cancers .
Evaluation of tumor hypoxia in mCRPC has been limited to date
due to infrequent sampling of metastases for routine clinical care.
In an immunohistochemistry study of HIF-1a expression that
incorporated a small set of prostate cancer metastases, HIF-1a
expression was observed to vary widely in metastatic lesions .
Here, we show that a subset of patients with metastatic prostate
cancer have increased levels of serum miR-210, providing
evidence for previously under-appreciated hypoxia in mCRPC.
Although non-tumor tissue sources of miR-210 cannot be ruled
out, the fact that systemic hypoxemia is not a typical feature of
mCRPC is consistent with a model in which tumor tissue hypoxia
is the origin of the excess serum miR-210. Notably, elevated
circulating miR-210 has also been observed in patients with
pancreatic adenocarcinoma , a disease in which tumor
hypoxia is well-recognized and is due to high interstitial pressure
due to the host desmoplastic response.
A well-documented phenomenon associated with tumor hyp-
oxia is the association with resistance to treatment with
radiotherapy, chemotherapy and other therapies . To
determine whether observed serum miR-210 levels were associ-
ated with treatment resistance, we retrospectively assessed whether
patients were responding or resistant to ongoing therapy by
calculating %PSA change/day using available clinical PSA values
measured most recently prior to and at the time of serum miR-
210 draw. Therapies varied among patients in this retrospective
population, but typically involved androgen deprivation therapy
using a GnRH agonist in combination with a chemotherapeutic
agent (e.g., docetaxel, mitoxantrone). We found that serum miR-
210 levels were significantly correlated with %PSA change/day
during treatment (Fig. 3A, Pearson r=0.46, P=0.029). To
reduce potential noise from patients who are less informative due
to low levels of cancer-associated serum miRNAs, we also
analyzed a subset of patients with high levels of mCRPC-
associated serum miRNAs (i.e., ‘‘miRNA-high subset’’, defined
Circulating MiRNAs and Hypoxia in Prostate Cancer
PLOS ONE | www.plosone.org6 July 2013 | Volume 8 | Issue 7 | e69239
as patients whose serum miR-141, miR-200a, miR-200c and/or
miR-375 levels were greater than the highest value observed in
any of the 25 healthy controls). In this group, the correlation
between serum miR-210 and %PSA change/day was even
stronger (Fig. 3A, Pearson r=0.61, P=0.029). Furthermore,
serum levels of miR-210 were strikingly lower in patients whose
disease was responding to treatment (PSA stable or decreasing), as
compared to those whose disease was resistant to treatment (PSA
increasing by $25%) (Fig. 3B, P=0.001). Importantly, we did
not observe this association with the other four serum miRNAs
identified in our study (Fig. 3C). Our data suggests a model in
which increased hypoxia response signaling is present in a subset
of mCRPC patients, leading to increased serum miR-210 and
To our knowledge, this is the first report of circulating miR-
210 in association with mCRPC. Our results raise the possibility
that serum miR-210 levels could be used to identify a biologically
distinct, subset of mCRPC patients with tumor-associated hypoxia
for whom the development of alternative therapeutic approaches
could be considered. For example, plasma miR-210 levels have
been reported to be elevated in pancreatic cancer patients and as
an indicator of hypoxia [23,24], as well as correlated with response
to trastuzumab in breast cancer patients . In addition, mTOR
inhibitors are being studied in prostate cancer, and pre-clinical
studies have shown that mTOR inhibition can lead to AKT
activation and HIF-1a transcriptional activation . In this
context, we speculate that elevated serum miR-210 could have
potential utility as a predictive or response biomarker for this class
of therapeutics. In addition, it will be important in future studies to
determine whether miR-210 is not only an indicator of hypoxia
and aggressive biology, but also an active mediator of an
aggressive disease phenotype in mCRPC patients.
Given that the number of new agents effective against mCRPC
is increasing, minimally invasive approaches such as serum miR-
210 analysis may lead to clinical decision aids that can differentiate
and help guide treatment decisions by differentiating between
biologically distinct disease subtypes. This could be particularly
important in settings where PSA is less informative, such as in
neuroendocrine differentiated subtypes, or when cancers progress
to an androgen pathway independent state.
different in abundance between mCRPC patients and
healthy controls. (A) miRNAs were measured in individual
samples by TaqMan miRNA qRT-PCR (P value assigned by
Wilcoxon signed-rank test), where miRNA abundance is given in
terms of miRNA copies/ml serum. Red bars, mean +/2 SEM of
miRNA copies/ml serum for each group. (B) Receiver operating
characteristic (ROC) curves plot sensitivity vs. (1 - specificity) to
assess the ability of each serum miRNA to distinguish mCRPC
and control sera.
Negative control miRNAs are not significantly
ers in serum from mCRPC patients and healthy controls
by single-plex microRNA TaqMan qRT-PCR.
Validation of candidate microRNA biomark-
serum miRNAs with each other and with serum PSA.
Correlation analysis of mCRPC-associated
ed serum microRNA markers and endogenous controls
cancer (‘‘Cancer’’) and lymph node metastases (‘‘LN
Results of measurement of mCRPC-associat-
Single-plex TaqMan assays used in this study.
We are grateful to Jason Bielas and members of his lab for assistance with
hypoxia experiments. We thank Rachael Parkin and Ausra Bendoraite for
technical assistance, Theodore D. Koreckij and Jennifer Noteboom for
help with clinical data retrieval, and Evan Yu for helpful comments on the
manuscript. This material is the result of work supported by resources from
the VA Puget Sound Health Care System, Seattle, Washington (to R.L.V.).
Conceived and designed the experiments: PSM PSN RLV CM MT.
Performed the experiments: PSM EMK AED CM. Analyzed the data:
HHC PSM MT. Contributed reagents/materials/analysis tools: AED LC
JS PSN RLV BSK AMC KJP CM MT. Wrote the paper: HHC PSM MT.
1. Mitchell PS, Parkin RK, Kroh EM, Fritz BR, Wyman SK, et al. (2008)
Circulating microRNAs as stable blood-based markers for cancer detection. Proc
Natl Acad Sci U S A 105: 10513–10518.
2. Flynt AS, Lai EC (2008) Biological principles of microRNA-mediated regulation:
shared themes amid diversity. Nat Rev Genet 9: 831–842.
3. Bonauer A, Carmona G, Iwasaki M, Mione M, Koyanagi M, et al. (2009)
MicroRNA-92a controls angiogenesis and functional recovery of ischemic tissues
in mice. Science 324: 1710–1713.
4. Gregory PA, Bert AG, Paterson EL, Barry SC, Tsykin A, et al. (2008) The miR-
200 family and miR-205 regulate epithelial to mesenchymal transition by
targeting ZEB1 and SIP1. Nat Cell Biol 10: 593–601.
5. Olson P, Lu J, Zhang H, Shai A, Chun MG, et al. (2009) MicroRNA dynamics
in the stages of tumorigenesis correlate with hallmark capabilities of cancer.
Genes Dev 23: 2152–2165.
6. Kulshreshtha R, Ferracin M, Wojcik SE, Garzon R, Alder H, et al. (2007) A
microRNA signature of hypoxia. Mol Cell Biol 27: 1859–1867.
7. Varadhachary GR, Spector Y, Abbruzzese JL, Rosenwald S, Wang H, et al.
(2011) Prospective gene signature study using microRNA to identify the tissue of
origin in patients with carcinoma of unknown primary. Clin Cancer Res 17:
8. Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, et al. (2005) MicroRNA
expression profiles classify human cancers. Nature 435: 834–838.
9. Rosenfeld N, Aharonov R, Meiri E, Rosenwald S, Spector Y, et al. (2008)
MicroRNAs accurately identify cancer tissue origin. Nat Biotechnol 26: 462–
10. Loberg RD, St John LN, Day LL, Neeley CK, Pienta KJ (2006) Development of
the VCaP androgen-independent model of prostate cancer. Urol Oncol 24: 161–
11. Jones ML, Siddiqui J, Pienta KJ, Getzenberg RH (2013) Circulating fibroblast-
like cells in men with metastatic prostate cancer. Prostate 73: 176–181.
12. Brase JC, Johannes M, Schlomm T, Falth M, Haese A, et al. (2011) Circulating
miRNAs are correlated with tumor progression in prostate cancer. Int J Cancer
13. Zhang HL, Qin XJ, Cao DL, Zhu Y, Yao XD, et al. (2013) An elevated serum
miR-141 level in patients with bone-metastatic prostate cancer is correlated with
more bone lesions. Asian J Androl 15: 231–235.
14. Nguyen HC, Xie W, Yang M, Hsieh CL, Drouin S, et al. (2012) Expression
differences of circulating microRNAs in metastatic castration resistant prostate
cancer and low-risk, localized prostate cancer. Prostate.
15. Bryant RJ, Pawlowski T, Catto JW, Marsden G, Vessella RL, et al. (2012)
Changes in circulating microRNA levels associated with prostate cancer.
Br J Cancer 106: 768–774.
16. Watahiki A, Macfarlane RJ, Gleave ME, Crea F, Wang Y, et al. (2013) Plasma
miRNAs as Biomarkers to Identify Patients with Castration-Resistant Metastatic
Prostate Cancer. Int J Mol Sci 14: 7757–7770.
Circulating MiRNAs and Hypoxia in Prostate Cancer
PLOS ONE | www.plosone.org7 July 2013 | Volume 8 | Issue 7 | e69239
17. Ivan M, Harris AL, Martelli F, Kulshreshtha R (2008) Hypoxia response and
microRNAs: no longer two separate worlds. J Cell Mol Med 12: 1426–1431.
18. Huang X, Le QT, Giaccia AJ (2010) MiR-210–micromanager of the hypoxia
pathway. Trends Mol Med 16: 230–237.
19. Huang X, Ding L, Bennewith KL, Tong RT, Welford SM, et al. (2009)
Hypoxia-inducible mir-210 regulates normoxic gene expression involved in
tumor initiation. Mol Cell 35: 856–867.
20. Mathew LK, Simon MC (2009) mir-210: a sensor for hypoxic stress during
tumorigenesis. Mol Cell 35: 737–738.
21. Bertout JA, Patel SA, Simon MC (2008) The impact of O2 availability on
human cancer. Nat Rev Cancer 8: 967–975.
22. Zhong H, De Marzo AM, Laughner E, Lim M, Hilton DA, et al. (1999)
Overexpression of hypoxia-inducible factor 1alpha in common human cancers
and their metastases. Cancer Res 59: 5830–5835.
23. Wang J, Chen J, Chang P, LeBlanc A, Li D, et al. (2009) MicroRNAs in plasma
of pancreatic ductal adenocarcinoma patients as novel blood-based biomarkers
of disease. Cancer Prev Res (Phila) 2: 807–813.
24. Ho AS, Huang X, Cao H, Christman-Skieller C, Bennewith K, et al. (2010)
Circulating miR-210 as a Novel Hypoxia Marker in Pancreatic Cancer. Transl
Oncol 3: 109–113.
25. Jung EJ, Santarpia L, Kim J, Esteva FJ, Moretti E, et al. (2012) Plasma
microRNA 210 levels correlate with sensitivity to trastuzumab and tumor
presence in breast cancer patients. Cancer 118: 2603–2614.
26. Majumder PK, Febbo PG, Bikoff R, Berger R, Xue Q, et al. (2004) mTOR
inhibition reverses Akt-dependent prostate intraepithelial neoplasia through
regulation of apoptotic and HIF-1-dependent pathways. Nat Med 10: 594–601.
Circulating MiRNAs and Hypoxia in Prostate Cancer
PLOS ONE | www.plosone.org8 July 2013 | Volume 8 | Issue 7 | e69239