Amplification-free detection of circulating microRNA biomarkers from body fluids based on fluorogenic oligonucleotide-templated reaction between engineered peptide nucleic acid probes: Application to prostate cancer diagnosis

Article abstract of DOI:10.1021/acs.analchem.6b01594

Highly abundant in cells, microRNAs (or miRs) play a key role as regulators of gene expression. A proportion of them are also detectable in biofluids making them ideal noninvasive biomarkers for pathologies in which miR levels are aberrantly expressed, such as cancer. Peptide nucleic acids (PNAs) are engineered uncharged oligonucleotide analogues capable of hybridizing to complementary nucleic acids with high affinity and high specificity. Herein, novel PNA-based fluorogenic biosensors have been designed and synthesized that target miR biomarkers for prostate cancer (PCa). The sensing strategy is based on oligonucleotide-templated reactions where the only miR of interest serves as a matrix to catalyze an otherwise highly unfavorable fluorogenic reaction. Validated in vitro using synthetic RNAs, these newly developed biosensors were then shown to detect endogenous concentrations of miR in human blood samples without the need for any amplification step and with minimal sample processing. This low-cost, quantitative, and versatile sensing technology has been technically validated using gold-standard RT-qPCR. Compared to RT-qPCR however, this enzyme-free, isothermal blood test is amenable to incorporation into low-cost portable devices and could therefore be suitable for widespread public screening.

Full text of DOI:10.1021/acs.analchem.6b01594

This is an open access article published under an ACS AuthorChoice License, which permits
copying and redistribution of the article or any adaptations for non-commercial purposes.
Article
pubs.acs.org/ac
Amplification-Free Detection of Circulating microRNA Biomarkers
from Body Fluids Based on Fluorogenic Oligonucleotide-Templated
Reaction between Engineered Peptide Nucleic Acid Probes:
Application to Prostate Cancer Diagnosis
†
,‡
‡
†
,†
‡
‡
Gavin A. D. Metcalf, Akifumi Shibakawa, Hinesh Patel, Ailsa Sita-Lumsden, Andrea Zivi,
‡
‡
Nona Rama, Charlotte L. Bevan, and Sylvain Ladame*
†
Department of Bioengineering, Imperial College London, South Kensington Campus, London SW72AZ, U.K.
‡
Department of Surgery and Cancer, Imperial College London, Hammersmith Hospital Campus, London W120NN, U.K.
S Supporting Information
*
ABSTRACT: Highly abundant in cells, microRNAs (or miRs) play a key role
as regulators of gene expression. A proportion of them are also detectable in
biofluids making them ideal noninvasive biomarkers for pathologies in which
miR levels are aberrantly expressed, such as cancer. Peptide nucleic acids (PNAs)
are engineered uncharged oligonucleotide analogues capable of hybridizing to
complementary nucleic acids with high affinity and high specificity. Herein,
novel PNA-based fluorogenic biosensors have been designed and synthesized
that target miR biomarkers for prostate cancer (PCa). The sensing strategy is
based on oligonucleotide-templated reactions where the only miR of interest
serves as a matrix to catalyze an otherwise highly unfavorable fluorogenic
reaction. Validated in vitro using synthetic RNAs, these newly developed bio-
sensors were then shown to detect endogenous concentrations of miR in
human blood samples without the need for any amplification step and with
minimal sample processing. This low-cost, quantitative, and versatile sensing technology has been technically validated using
gold-standard RT-qPCR. Compared to RT-qPCR however, this enzyme-free, isothermal blood test is amenable to incorporation
into low-cost portable devices and could therefore be suitable for widespread public screening.
oor prognosis of pathologies diagnosed at an advanced stage
has been the driver behind research into techniques to
positive PSA tests are in fact false positives, resulting from the
nonspecific nature of a raised PSA outcome since concentrations
can fluctuate as a consequent of infection, inflammation, or
P
diagnose cancers at an early stage, ideally prior to metastasis and
secondary tumor formation. One promising area is minimally
invasive screening using biomarker-based diagnostic tests. Such
screening tools can help identify individuals harboring poten-
tially life-threatening tumors at the earliest stage possible and
minimize overall treatment burden by aiding accurate prognosis,
4
benign prostatic hyperplasia (BPH). This means that a signifi-
cant proportion of men with positive PSA test results will undergo
unnecessary surgery, putting them at risk of infection and heighted
5
prostate inflammation that can actually promote carcinogenesis.
PCa is often diagnosed late, largely due to the presentation of
1
,2
stratification, and personalized medical treatment plans. To be
successful this strategy requires not only biomarkers that are both
highly sensitive and highly specific but also a minimally invasive
screening medium and analytical platforms that are fit-for-purpose
and low cost.
many broadly unspecific symptoms, once primary tumor cells
6,7
have metastasized to secondary sites. Therefore, the use of
biomarkers in asymptomatic individuals, in populations screen-
ing or targeted to those identified as at risk, is a promising area for
early diagnosis and improved prognosis.
Cell-free nucleic acids (cfNAs) are DNA and RNA molecules
circulating in biofluids (including blood, saliva, and urine) that
are therefore accessible via minimally invasive means. cfNAs were
first described in 1948, although this attracted little interest from
the scientific community until the early 1990s, when researchers
detected mutated Rat sarcoma (RAS) gene fragments in blood
Prostate cancer (PCa) is the most commonly diagnosed cancer
in men in the western world with incidence rates of over 400,000
3
in Europe and 220,000 in the USA. The vast majority of PCa is
adenocarcinoma that arises in the secretory epithelial cells of
prostatic ducts. Current diagnosis results from the presentation
of clinical symptoms (e.g., urine hesitancy/frequency and/or
erectile dysfunction) whereupon a blood test may be carried out
to assess prostate specific antigen (PSA) concentration in the
serum. Levels over the nominated threshold of 4 ng/mL will most
likely result in further, highly invasive examination, including digital
rectal examination and tissue biopsy. However, up to 2/3 of
Received: April 22, 2016
Accepted: July 21, 2016
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.analchem.6b01594
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Article
8
samples from cancer patients. Since then cfNAs have received a
growing amount of interest, with many studies suggesting their
Scheme 1. General Sensing Strategy Based on the Concept of
Oligonucleotide-Templated Reaction (OTR)
a
9
value as biomarkers of various diseases, including cancer. A par-
ticular type of cfNA biomarkers are microRNAs (miRs). MiRs
are small, noncoding, RNA molecules typically 18−24 nucleotides
in length that are encoded within the genome. They control,
mostly negatively, the expression of genes involved in multiple
biological processes, including apoptosis, differentiation, pro-
1
0
liferation, and cancer metastasis. Based on genome-wide miR
expression-profiling studies, it is now widely accepted that miRs
are commonly deregulated in various human cancers, acting either
as oncogenes or tumor suppressor genes and making them ideal
cancer biomarkers. However, developing sensing/profiling methods
that are both sensitive and quantitative has proven extremely
challenging for miRs because of their small size, sequence simi-
larity among family members, and low concentration in the
bloodstream.
Quantitative detection of miRs is commonly achieved using
real time reverse transcription quantitative PCR (RT-qPCR)
widely perceived by most as the “gold standard”. Although
RT-qPCR displays high sensitivity, with minute levels of starting
RNA being required, this analytical technique holds many
limitations: (1) risks of contamination and error during each
amplification step (RT and then PCR), (2) high cost and lack of
specificity of custom-made oligonucleotide probes, and (3) high
levels of background fluorescence due to incomplete fluores-
cence quenching and/or differences in quality/purity between
commercial batches of probes. It is also noteworthy that being
nonisothermal and requiring various (labile) enzymes, this sensing
technology is not easily amenable to incorporation in miniaturized
devices and remains exclusively a research tool. Although miR
detection and analysis has progressed steadily since its relatively
recent discovery, the lack of a much-needed standardized tech-
nique that is both sensitive and accurate enough to detect
endogenous miR for diagnostic purposes is clear.
a
Complementary Watson-Crick base-pairing between the RNA target
and two engineered PNA probes catalyzes a fluorogenic Michael addi-
tion reaction and unleashes the (otherwise quenched) fluorescence of
a coumarin derivative.
Microwave-assisted reactions on solid phase were performed on
a Biotage initiator + SPWave Synthesizer. MALDI-TOF spectra
were recorded on a MALDI micro MX instrument using sin-
apinic acid as a matrix.
Synthesis of the Coumarin Probe Head (3). A solution of
,3,6,7-tetrahydro-8-1H,5H-benzo[ij]quinolizine-9-carboxalde-
2
hyde (2 g, 4.6 mmol) and ethyl acetoacetate (2.4 g, 9.2 mmol)
in EtOH (25 mL) was treated with piperidine (0.78 mL). The
mixture was heated to reflux (85 °C) for 3 h, removed from heat,
and allowed to cool to RT. The solution was vacuum filtered and
washed with cold EtOH. The filtrate was discarded, and the solid
was dried to leave coumarin 334 as bright orange crystals
Herein we report a novel sensing technology for the quan-
titative detection of endogenous concentrations of circulating
miR biomarkers in blood samples that does not require any
amplification step and is isothermal and highly cost-effective
(
1
300 mg, 28% yield). A mixture of coumarin 334 (300 mg,
.06 mmol) and 3-formylbenzoic acid (165 mg, 1.1 mmol)
was dissolved in EtOH:ACN (15 mL, 1:1 v/v), and piperidine
115.5 μL) was added. The mixture was heated to reflux (90 °C)
(
<1 pence per test). Three sets of primary probes were designed
to detect three endogenous miRs of interest (miR-375, miR-141,
and miR-132) that were previously reported to be either under-
or overexpressed in the blood of cancer patients compared to that
(
for at least 72 h, removed from heat, and allowed to cool down to
room temperature. The solution was precipitated by adding
diethyl ether. The reaction mixture was then triturated in diethyl
ether, and the solid was collected by vacuum filtration and washed
extensively with dichloromethane (DCM) to afford compound 3
pure as a fine dark red powder (106 mg, 26% yield) (Scheme 2).
11,12
of healthy controls.
Our general sensing strategy exploits the
concept widespread in Nature of oligonucleotide-templated
1
3,14
reaction
(OTR) where the miR biomarker of interest is used
as a template to catalyze an otherwise highly unfavorable fluo-
rogenic reaction between chemically engineered peptide nucleic
acid (PNA) hybridization probes (Scheme 1). Validated using
the gold-standard RT-qPCR, this new technology holds great
promise for noninvasive diagnosis, stratification, and monitoring
of prostate cancer and is also perfectly suited for incorporation in
portable devices.
1
H NMR (400 MHz, DMSO): δ = 8.43 (s, 1H), 8.22 (s, 1H),
8
.02 (d, J = 15.75 Hz, 1H), 7.94 (d, J = 7.70 Hz, 1H), 7.70−7.63
(m, 2H), 7.39 (t, J = 7.70 Hz, 1H), 7.24 (s, 1H), 3.37−3.30
(m, 4H), 2.79−2.66 (m, 4H), 1.92−1.81 (m, 4H).
17
General Procedure for Solid Phase PNA Synthesis.
-mer PNA probes were synthesized on solid support using either
7
H-Rink Amide ChemMatrix (0.50 mmol/g loading) (PCAS Bio-
matrix) or 4-aminobutanethiol-4-methoxytrityl (0.76 mmol/g
loading) (EMD Millipore) resin and microwave-assisted Fmoc
chemistry. The 4-aminobutanethiol-4-methoxytrityl resin was
chosen for the production of PNA-thiol such that, upon cleavage
from the resin, the PNA sequence of interest contained a
free butanethiol group at its C-terminus. Functionalized PNAs
were cleaved from solid support by treatment with a solution
of trifluoroacetic acid (TFA):triisopropylsilane (TIS):double
EXPERIMENTAL SECTION
■
Materials and Methods. Unless otherwise noted, all
solvents and reagents were obtained from commercial sources
and used without further purification. HPLC-purified DNA
and RNA oligonucleotides were purchased from Invitrogen
1
and Eurogentec. H NMR spectra were recorded on a Bruker
AVANCE 400 III HD spectrometer. Chemical shifts are reported
on the δ scale in ppm using the residual peak solvent as the
internal standard. Coupling constants (J) are reported in Hz.
distilled water (ddH O) (95:2.5:2.5, v/v). The desired PNAs
2
B
DOI: 10.1021/acs.analchem.6b01594
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Article
Scheme 2. Coumarin Probe Head (3) Synthesis via an Aldol
Condensation Reaction between Coumarin 334 (1) and
aliquoted (1 mL/cryovial), and frozen at −80 °C. Samples were
thoroughly thawed for 1 h at RT prior to RNA isolation.
Total RNA Isolation. RNA was extracted from 200 μL of
serum using miRCURY RNA Isolation Kit - Biofluids (Exiqon)
3
-Formylbenzoic Acid (2)
9
according to the manufacturer’s instruction. During lysis, 1.25 × 10
copies of cel-miR-39-3p oligonucleotides (Sigma-Aldrich) were
spiked in to evaluate extraction efficiency, and 7.5 μg of Glyco-
Blue (ambion) was added as a carrier. Following on-column
DNase I treatment, RNA was eluted in 50 μL of nuclease-free
water.
General Procedure for Detection of Endogenous miRs
Using PNA Probes. In a typical experiment, to a solution of
5
μM PNA-thiol and 5 μM PNA-coumarin in 10 mM tris-HCl
buffer pH 7.4 (Sigma) was added 10% (v/v) of thoroughly thawed
RNA isolated from biofluid. Non-templated control (NTC) experi-
ments that consisted of the same stoichiometric mixture of PNA
probes - 5 μM each - but with no added RNA (replaced by ddH O)
2
were also run in parallel. Reactions were all carried out in 400 μL
Eppendorf tubes and incubated at 37 °C for 3 h (QBD digital dry
block heater, Grant). Samples were then transferred to 384-well
microtiter black plates, and fluorescence emission spectra were
recorded using a Fluorostar fluorescence plate reader (BMG
LabTech) (λexc = 480 nm, λem = 520 nm). Background fluo-
rescence from the PNA-coumarin probe alone (5 μM in 10 mM
tris·hcl pH 7.4) was also recorded and subtracted from all the
above readings.
Real-Time Reverse Transcription Quantitative Poly-
merase Chain Reaction (RT-qPCR). cDNA synthesis and
qPCR were performed using a miRCURY LNA Universal RT
microRNA PCR system (Exiqon) according to the manufac-
turer’s protocol. Two μL of RNA (equivalent to 8 μL of starting
serum) was reverse transcribed in a 10 μL final reaction volume
using Universal cDNA Synthesis Kit ll (Exiqon). Prior to reverse
were purified using high-performance liquid chromatography
HPLC) and characterized by matrix-assisted laser desorption/
ionization-time-of-flight mass spectrometry (MALDI-TOF MS)
using sinapinic acid as a matrix. Sequences of the PNAs synthe-
sized are summarized in Table 1. (See the Supporting Information
for MALDI-TOF characterization of all probes.)
(
7
transcription, 2.5 × 10 copies of a synthetic RNA (UniSp6, Exiqon)
were spiked in to evaluate cDNA synthesis. The resulting cDNA
was diluted in nuclease-free water containing ROX passive refer-
ence dye (Invitrogen) at a final concentration of 500 nM. Indi-
vidual qPCR was performed on diluted cDNA using the ExiLENT
SYBR Green master mix (Exiqon) and LNA PCR primers (Exiqon)
for cel-miR-39-3p, UniSp6, miR-141, and miR-375. Real-time
PCR amplification was performed in duplicate on an Applied
Biosystems 7900HT system using a cycle condition according
to the template file (available at www.exiqon.com/sds). Data
was analyzed using SDS v2.4 (Applied Biosystems). (See the
Supporting Information for qPCR internal controls.)
Table 1. Synthesized PNA Sequences
a
name
sequence
miR-141_Coum
miR-141_Thiol
miR-132_Coum
miR-132_Thiol
miR-375_Coum
miR-375_Thiol
Arg-Arg-TGTGACA-Coum
SH-CATTTCT-Arg-Arg
Arg-Arg-GTACCGA-Coum
SH-CTGACAA-Arg-Arg
Arg-Arg-ACAAGCA-Coum
SH-CGAGCGC-Arg-Arg
a
Sequences described as from the C-terminus to N-terminus.
Statistical Analysis. Student’s t-test was used to compare the
means of two data sets. One-way ANOVA was used to compare
the means of more than two data sets. Statistical analysis was
completed using SPSS software (v22.0, IBM, USA).
Blood Donation and Serum Separation. Whole blood
samples of prostate cancer patients were collected at Imperial
College Healthcare NHS Trust (London, UK). All samples were
taken from patients in 2014 following written patient consent
and stored as a subcollection in the Imperial College Healthcare
NHS Trust Tissue Bank. Serum instead of plasma was used to
avoid cell contamination and anticoagulant interference. Whole
blood samples were obtained with ethical approval from patients
attending clinics at Imperial College London NHS Healthcare
Trust (London, UK). All samples were taken from patients
undergoing oncological management for biopsy-proven prostate
cancer, in 2014, following written patient consent Serum
separation: Blood was sourced via a standard venipuncture pro-
cedure into red-topped 6 mL vacutainers (silicon-coated, with
clot activator, BD). Samples were left upright at room tempera-
ture for 0.5−1 h to allow blood to clot before centrifuging in a
swing-out rotor at 1100−1300 g for 10 min at RT. Serum, isolated
as the supernatant, was immediately aspirated postcentrifugation,
RESULTS AND DISCUSSION
■
Design Strategy, Synthesis, and Characterization. We,
and others, have recently reported the use of engineered pep-
tide nucleic acid (PNA) oligomers for the detection of nucleic
acids in vitro based on a sequence-specific and fluorogenic
1
3,14
oligonucleotide-templated reaction (OTR).
PNAs are
DNA (or RNA) analogues in which the negatively charged
phosphodiester-sugar backbone has been replaced by a charge-
free 2-aminoethyl glycine (AEG) scaffold holding the nucle-
15
obases through methylene carbonyl linkages. PNAs, exhibiting
higher affinity for complementary DNAs (or RNA) than tradi-
tional oligonucleotides and being more responsive to point
mutations, are perfectly suited for the specific detection of short
C
DOI: 10.1021/acs.analchem.6b01594
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
22−25 nucleobases) RNA molecules and discrimination
Article
(
Upon simultaneous hybridization of both PNA probes to the
same miR target, we anticipated that the thiol groups (from
either cysteine or 4-butanethiol) of one PNA would be able to
reach the α,β-unsaturated ketone of the coumarin present on the
second PNA. PNA length (7-mers) was chosen to be above the
5-mer limit that was previously identified as the minimal length
for full, sequence specific, hybridization of PNA-RNA hetero-
duplexes at 37 °C. The addition of two units was decided to
increase sequence-specificity in order to better discriminate
between miR sequences. Furthermore, a 3-nucleotide gap was
placed between both PNA hybridization domains to provide
sufficient flexibility for both probe-heads to interact with each
other. This gap size was also based upon previous work with the
chosen cleft resulting in an optimal fluorescent signal when
compared to the other gap distances that were tested (1 to
5 nucleotides). To ensure that PNAs were soluble in water at
physiological pH (∼7.4) two hydrophilic arginine amino acids
(Arg) were introduced at either the N-terminus (thiol probe) or
the C-terminus (coumarin probe). Using the above-mentioned
design strategy, probes were synthesized to be directed against
miR-141, a previously identified PCa biomarker chosen here as a
model system (see the Supporting Information for structures of
16
between miRs with high sequence homology. OTRs rely on
sequence-specific Watson−Crick base-pairing to bring together
and promote the (otherwise highly unfavorable) reaction be-
tween two moieties, each covalently attached to a PNA, upon
exposure to a complementary DNA or RNA oligonucleotide. In
the absence of the target of interest, the PNA probes are present
in solution at such low concentrations that their probability of
meeting and reacting with each other is extremely low. In this
case of a fluorogenic OTR, this accounts for the very low back-
ground fluorescence and high signal-to-noise ratio (S/N).
17
In 2011, Jung et al. described a fluorescence-based biosensor
for cysteine based around coumarin 334. Conjugation of cou-
marin 334 to 3-formylbenzoic acid via an aldol condensation
reaction resulted in near total fluorescence quenching, proposed
to occur by photoinduced electron transfer (PET). 1,4-Addition
of the cysteine thiol on the α,β-unsaturated ketone was then
shown to restore the intrinsic fluorescence properties of the dye,
thus acting as a cysteine-responsive “light-up probe”. Herein, we
demonstrate that this reaction of fluorescence unquenching can
be used as a highly sensitive reporter system in fluorogenic OTRs.
Coumarin derivative 3 was synthesized using a protocol adapted
14
1
12
from Jung et al. H NMR was first used to confirm that fluores-
probes).
cence unquenching indeed resulted from a Michael addition. As
shown in Figure 1, at high concentrations of both cysteine and
Stoichiometric mixtures of PNA thiol and PNA coumarin
probes (5 μM each) were incubated in the presence and in the
absence of miR-141 (Figure 2). For both systems (i.e., using
Figure 2. Comparison of cysteine and butyl thiol probes for miR-141
sensing. t-test; butyl thiol (NTC/TC) p = ≤ 0.001, cysteine (NTC/TC)
p = 0.0137. Samples are normalized against a nontemplated control
reaction. Values described are from a minimum of two technical repli-
cates, with miR-141 RNA-templated control (TC) completed in com-
parison to a nontemplated control (NTC).
1
Figure 1. H spectra of coumarin 3 in the absence (black) and in the
presence (red) of 4 equiv of L-cysteine. Spectra were recorded in a
mixture of d6-DMSO−D O (9−1), with a coumarin concentration of
0 mM. Arrows indicate signals corresponding to the α,β-unsaturated
2
1
system that disappear upon addition of cysteine (see the Supporting
Information for full NMR spectra).
cysteine or butanethiol as a source of thiol) a significant increase
in fluorescence intensity was observed upon addition of miR-141.
However, the amplification factor (when compared to the reaction
carried out in the absence of miR-141) was much stronger with
the butanethiol (3-fold for Cys versus 15-fold for butanethiol).
This can most likely be explained by the fact that the butanethiol,
being less sterically hindered and more flexible than the cysteine
thiol, has easieraccess to thecoumarinprobe. Asa result, all further
studies were conducted using PNA thiol probes functionalized
with a 4-butyl-thiol moiety, and further pairs of probes were
synthesized using this set of probe-heads and others constructed in
similar fashion and directed against miR-375 and miR-132.
In Vitro Quantitative and Sequence-Specific miR
Detection by Fluorogenic OTR. For each miR of interest,
sensing reactions were first assessed in tris-HCl buffer
(10 mM, pH 7.4) using synthetic (commercially sourced)
RNA oligonucleotides. Probe concentrations were fixed at 5 μM
coumarin 3, both doublets corresponding to the α,β-unsaturated
ketone disappeared almost completely in less than 1 h. The forma-
tion of the cysteine adduct was also confirmed by MALDI.
For miR sensing applications, PNA probes complementary to
the 3′ and 5′ ends of each miR of interest were functionalized
with either coumarin 3 at the N-terminus or a thiol moiety at the
C-terminus. Compound 3 was immobilized onto the PNA via
amide coupling between the coumarin’s carboxylic acid and the
peptide’s terminal amine. For thiol functionalization, we first
decided to introduce a cysteine residue at the PNA C-terminus
17
as earlier work from Jung et al. showed successful coumarin
unquenching with cysteine. However, a more flexible and less
sterically hindered 4-butyl thiol was also introduced using
4-aminobutanethiol-4-methoxytrityl resin.
D
DOI: 10.1021/acs.analchem.6b01594
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
stoichiometric amount of PNA-coumarin and PNA thiol),
Article
(
and concentrations of nucleic acid target (miR-141, miR-375,
and miR-132 RNA) were varied between 50 nM and 5 μM (see
Figure 3A for miR-141 and Supporting Information for miR-375
Figure 4. Functionality of PNA probes to detect miR-141 RNA (1 μM)
in the presence and absence of (A) heparin (16 IU/mL) and (B) cys-
teine (50 μM). Samples are normalized against a templated reaction
using commercial RNA. Values described are from a minimum of three
technical replicates, completed in comparison to a nontemplated control
(NTC).
20
while keeping optimal levels of RNA intact, which adds time,
variability, and cost to processing prior to completing RT-qPCR.
As shown in Figure 4A, our technology is not affected by the
presence of heparin, with PNA probes operating consistently
both in the presence and absence of heparin (Sigma). Therefore,
the use of our PNA technology would not require additional
enzymatic purification of biofluid samples prior to testing.
Sequence Specificity and SNP Analysis. The ability of
our probes to distinguish between miRs with high sequence
homology was investigated. The effect of single nucleotide
polymorphism (SNP) and SNP position on OTR efficiency was
assessed by exposing probes directed against miR-132, miR-141,
or miR-375 to perfectly matched DNA and RNA sequences and
sequences differing by only one nucleotide, either at the edge or
at the center of the PNA/RNA binding domain (see Figure 5 and
the Supporting Information).
Figure 3. (A) Fluorescence intensity as a function of miR-141
concentration (0.05 to 1 μM) using a fixed stoichiometric concentration
of probes (5 μM each). Samples were normalized against a non-
templated reaction. Values described are from a minimum of three
biological replicates. (B) miR-141 kinetics over a 4.5 h (270 min) time
course at both room temperature (blue) and 37 °C (red). (C) Fluo-
rescence emission spectra of a stoichiometric mixture of miR-141 PNA
probes (5 μM each) in tris-HCl buffer (10 mM, pH 7.4) in the absence
(
black) and presence (red) of a miR-141 DNA. Spectra were recorded
after 3 h at 37 °C (λexc = 480 nm).
and miR-132). For all three miRs tested, near-linear correlation
between fluorescence intensity and RNA concentration was
obtained for concentrations ranging between 50 nM and 1 μM
(Figure 3A), thus demonstrating the quantitative nature of our
sensing technology. Comparable limits of detection (60.77 nM)
were obtained for all three miRs tested. [Limit of detection was
calculated by multiplying the standard deviation of the blank
by three (LOD = 3 × SD blank).] Unsurprisingly, faster reaction
kinetics were observed at 37 °C compared to room temperature,
equilibrium being reached after 120 and 240 min, respectively
(
Figure 3B). However, the same plateau (i.e., same final fluo-
Figure 5. Fluorescence intensity (%) as a result of SNP position (green
arrow − end SNP; red arrow − middle SNP) using the Coum/Thiol
system from miR-141, miR-132, and miR-375. Samples are normalized
against a corresponding templated reaction using commercial RNA.
Values described are from a minimum of three technical replicates for
each RNA tested, completed in comparison to nonmutated miR-141,
miR-132, and miR-375 DNA templated control (TC). t-test; end
SNP/middle SNP (p = 0.007). SNP sequences are specified in the SI.
rescence intensity) was obtained at both temperatures (suggesting
thermodynamic equilibrium), and, as a result, all the following
experiments were carried out at 37 °C, and fluorescence was
measured after 180 min incubation.
To assess specificity, OTRs were performed in the presence of
various competitors, i.e. by spiking reaction mixtures with high
concentrations of cysteine and heparin. Having previously shown
that cysteine can react with coumarin 3, it was important to
show that this reaction is significantly less efficient than the
miR-specific OTR. As shown in Figure 4B, addition of a large
excess (50 mol equiv) of cysteine to a stoichiometric mixture of
PNA probes has very little effect on the fluorescence intensity
measured after 3 h incubation in the absence or presence of miR.
Commonly used vacutainers to collect blood samples contain
heparin, an anticoagulant which can copurify with RNA by inter-
acting with its magnesium cofactor (Mg ) and in turn inhibit
Taq-polymerase (a Mg -dependent enzyme) activity within
RT-qPCR so potentially confounding results.
suggested that heparinized blood samples should be processed
with heparinase along with RNase inhibitors to remove heparin
As shown in Figure 5, when compared to reactions carried
out with the native, fully matched, sequences, OTR carried
out upon incubation of the probes with mutated sequences
all resulted in a loss of fluorescence intensity. Although this
loss was minimal when the mutations were located at the ends
of the DNA/PNA hybridization domain (−14 and −7.4% for
miR-141 and miR-132, respectively), a more drastic reduction in
fluorescence intensity was observed when mutations were located
in the center of those binding domains (−51 and −72% for
miR-141 and miR-132, respectively). This can be explained by
the predominant effect of SNP position on the stability of the
PNA/DNA heteroduplex formed: central mutations resulting in
2
+
2
+
18,19
It has been
E
DOI: 10.1021/acs.analchem.6b01594
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Article
much stronger destabilization (or much weaker hybridization)
than edge/terminal mutations. However, this study highlights
the ability of our technology to discriminate between two similar
DNA (or RNA) sequences differing by only one nucleotide.
In the specific case of our three miRs of interest, a BLAST
search was completed but found no close similarities with miR-
1
32 nor miR-375, i.e. no other miR with a sequence homology
of 75% (with an Expect value ‘E-value′ cutoff of 10 to correct
for random background noise and maximum number of hits
21,22
value of 100). However, miR-200a
shared 90.90% homology
(
20/22bp) with miR-141 (A → C substitution at the 17th bp,
and G → U substitution at the tail of the sequence) (BLASTN,
NCBI). Additional tests performed by incubating miR-141 probes
with synthetic miR-200a RNA and miR-141 templated controls
were completed to investigate the specificity of our system.
As expected based on our SNP analysis, a marginally reduced
fluorescent signal was obtained when using miR-200a as an
oligonucleotide template for miR-141 probes, i.e. a 0.22-fold
reduction in comparison to normalized miR-141 templated
control (see the SI for data and sequences). This result suggests
that our probes may not be able to efficiently discriminate
between miR-141 and miR-200a, but this is also likely to be
the case with RT-qPCR (as suggested by the good correlation
between both techniques). However, our results show that
simultaneous detection of miR141 and miR200a can detect
patients with active forms of the PCa with high affinity and
specificity.
Figure 6. (A) Normalized levels of endogenous miR-141 and miR-375
in PCa cohort calculated by PNA probes. Samples are normalized
against a templated reaction using commercial RNA. One-way ANOVA;
miR-141 (p = ≤ 0.001), miR-375 (P = ≤ 0.001). (B) Levels of miR-141
and miR-375 in PCa cohort measured by qPCR. One-way ANOVA;
miR-141 (p = 0.367), miR-375 (p = 0.231). (C) Correlation between the
expression values measured by qPCR and PNA probe technologies.
MiR-141 (Pearson r = −0.5154; p = 0.041), miR-375 (Pearson r =
Amplification-Free Detection of Endogenous miR-141
and miR-375 in Human Serum: Proof-of-Concept and
Technical Validation by RT-qPCR. As previously described,
12,23
miR-141 and miR-375 are well-reported PCa biomarkers,
having been shown to be up-regulated in patients with metastatic
−
0.5929, p = 0.016). Slope shows line of linear regression. PCa cohort
2
4
tumors. The ability of our sensing technology to detect
endogenous concentrations of miRs without the need for any
amplification step was next investigated by isolating total RNA
from serum sourced from a cohort of PCa patients (n = 16):
remission patients, both with (RY) and without (RN) prostate
glands and patients with active cancers, including those with
tumors that were either localized (AL) or metastatic (AM). For
both miR-141 and miR-375, elevated levels of biomarkers were
detected in patients with active cancers compared to patients in
remission, with the highest levels detected in patients with
metastatic PCa (Figure 6A). The same RNA samples from the
same cohort of PCa patients were analyzed using gold standard
RT-qPCR to technically validate our newly developed sensing
technology. Results displayed comparable trends between our
technology and RT-qPCR (i.e., elevated levels of both miR-141
and miR-375 in cancer patients compared topatients in remission)
despite our technology being free of any amplification step
details - RN: remission, post prostate removal (n = 4); RY: remission,
prostate gland present (n = 4); AL: advanced localized disease (n = 4);
AM: advanced metastatic disease (n = 4).
of the miR-target but also the precursor-miR (pre-miR) sequences
in RNA isolated from biofluids. Research has shown that pre-
25
miRs can be secreted from cells in exosomes and are abundant
26
within biofluids. Furthermore, RNA extraction techniques used
to isolate total RNA from serum would successfully lyse exo-
27
somes present within biofluid samples releasing pre-miRs,
thus allowing their detection. This mechanistic study is currently
under further investigation and will be reported in due course.
However, preliminary studies showed that probes directed
against miR-141 could detect the mature miR-141 and a DNA
hairpin analogue of miR-141 pre-miR with similar sensitivity (see
the Supporting Information). While, under our experimental
conditions, only mature miR were detectable by RT-qPCR, our
PNA probes are also capable of detecting miR-precursors, which
would account for some level of discrepancy in our correlation
between fluorescence and RT-qPCR data (Figure 6C).
(Figure 6B and C).
There remains, however, the question of the natural
abundance of circulating miRs in blood and of the sensitivity
required to detect endogenous miRs in biofluids. Since their
discovery, there has been a lot of discrepancy in the recent
literature regarding the exact form, location, and concentration of
circulating miRs. In an attempt to better understand the ability of
PNAs to detect endogenous miR concentration without the need
for any target or signal amplification, we explored the ability of
PNA probes to detect not only miR in their single-stranded
mature form but also some of their double-stranded, hairpin
precursors (pre- and pri-miRs). It is well-known that PNAs can
invade DNA or RNA duplexes. In the context of miR sensing, this
means that PNA probes could detect not only the mature forms
Although detection of miR precursors (e.g., pre-miRs) is also
possible by RT-qPCR, it requires a separate set of primers and
probes and can only be achieved by running a separate experi-
28
ment, looking solely at miR precursors. Further work is currently
underway in our laboratories to confirm whether PNA probes are
indeed capable of detecting simultaneously mature and precursor
miRs. Since experimental data unambiguously demonstrate that
our fluorogenic PNA probes can clearly distinguish between
patients with active PCa and those in remission (Figure 6A), this
would indicate that the sum of (mature + precursor) miRs can
also be used as a specific biomarker for PCa.
F
DOI: 10.1021/acs.analchem.6b01594
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Article
Specificity toward Cancer Types. Successful detection of
heightened concentrations of endogenous miR-141 and miR-375
in individuals with PCa led to the exploration of miR biomarkers
that are down-regulated in the biofluid of cancer patients. MiR-132
was chosen due to its well-reported nature as an epithelial ovarian
cancer (EOC) biomarker, with reduced expression representative
of heightened tumor aggressiveness. MiR-132 expression was
investigated in individuals with EOC of either stage 3 (typically
indicative of localized tumor spread into abdominal cavity) or
stage 4 (typically indicative of tumor spread forming distant meta-
stases in liver or lung for example), in comparison to healthy controls.
Results, as displayed in Figure 7, demonstrate that PNA probes
were able to detect reduced expression of miR-132 in RNA
not those miRs encapsulated in lysosomes or bound to proteins
such as argonaute (Argo).
29,30
Further investigation is underway to explore the detection of
trapped miRs directly in biofluids via different means, including
in situ protein denaturing, and shall be reported in due course.
The focal finding of these results further endorses our PNA-based
biosensors and their ability to detect miRs directly in blood
samples without any amplification steps, as currently required by
RT-qPCR.
24
CONCLUSION
■
We have engineered and validated a PNA-based biosensor for the
detection of endogenous concentrations of circulating miRs in
serum that does not require any amplification step (i.e., no
enzyme) and involves minimal or no sample processing. Optical
sensing of miRs is achieved via an RNA-templated fluorogenic
reaction between two non- and weakly fluorescent hybridization
probes, using the only miR of interest as a reaction template.
Using a cohort of 16 PCa patients, we were able to detect
elevated levels of miR-141 and miR-375 in samples from patients
with active forms of the disease. Although tests were initially
performed on extracted RNA from serum samples, similar results
were also obtained when using our probes directly in serum: free
from any amplification and from any processing steps. This is
particularly important since it is well documented that the way
blood samples are collected, stored, and processed strongly
Figure 7. Normalized levels of endogenous miR-132 (A) and miR-141
(
B) in healthy and differently graded epithelial ovarian cancer patients
(
stage 3 versus 4 EOC). Values described are from a minimum of three
31
influence the outcome of miR analyses. When compared to
technical replicates, completed in comparison to nontemplated controls
NTC). Samples are normalized against a miR-specific templated
reaction using commercial RNA. One-way ANOVA; miR-132 (p =
.046), miR-141 (p = 0.436).
RT-qPCR (used herein as a gold standard for technical validation),
our technology also offers the advantages of being low-cost
(
(
<1 pence per test in the current format), isothermal, and highly
0
specific. Unlike most other technologies currently available,
it is amenable for incorporation into portable devices and that
could therefore be compatible with widespread public screening.
Work is currently underway in our laboratories to engineer
platforms that could accommodate our probes and be suitable for
such applications.
isolated from serum samples and thus endorse previously
24
reported findings. Furthermore, miR-141 expression remains
fairly constant, supporting that specific miRs can be attributed to
particular cancer types.
Amplification-Free Detection of miR-141 and miR-375
Directly in Serum (in Situ). Successful detection of miR-141
and miR-375 in total RNA isolated from biofluids led to the
question of whether our PNA probes would be able to function
and detect miR-targets directly in blood serum samples with-
out any amplification or other enzymatic reactions prior to
miR-sensing. Five μM concentrations of PNA probes were added
to 20 μL of blood serum, and samples were topped up to 80 μL
with tris·HCl buffer (10 mM, pH 7.4) and subjected to the same
incubation and spectrophotometric procedures as isolated RNA.
Results (Figure 8) showed matching trends to those seen in isolated
RNA tests but presented greatly reduced fluorescent signals. One
reason for this could be a result of detecting only free-miR and
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.analchem.6b01594.
■
*
S
DNA and RNA sequences, PNA probe structure and
characterization via MALDI-TOF, characterization of the
adduct between cysteine and coumarin by MALDI-TOF
1
and H NMR, biocompatibility tests, in vitro quantitative
detection by fluorogenic OTR, sequence specificity and
SNP analysis, detection of pre-miRNA sequences with
PNA probes, and qPCR internal controls (PDF)
AUTHOR INFORMATION
Corresponding Author
Phone: +44 (0)20 7594 5308. E-mail: S.ladame@imperial.ac.uk.
■
*
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
H.P. thanks the Whitaker International Program for his fellow-
ship. Thanks go to Professor Robert Brown, Imperial College
London, for EOC blood serum donation. Research in this publi-
cation was supported by a Scottish Power/Cancer Research UK
grant and a Movember/Prostate Cancer UK Centre of excellence
programme grant ref: CEO013-2-002. This study was also
Figure 8. PNA probe functionality in blood serum (in situ). (A) Nor-
malized levels of endogenous miR-141 and (B) normalized levels of
endogenous miR-375 in a PCa cohort. One-way ANOVA; miR-141
(
p = ≤ 0.001), miR-375 (P = ≤ 0.001). RN: remission, postprostate
removal (n = 4); AM: metastatic advanced tumor (n = 4).
G
DOI: 10.1021/acs.analchem.6b01594
Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Article
supported by Imperial Experimental Cancer Medicine Centre,
Imperial NIHR Biomedical Research Centre, Cancer Research
UK Imperial Centre.
(20) Johnson, M. L.; Navanukraw, C.; Grazul-Bilska, A. T.; Reynolds,
L. P.; Redmer, D. A. BioTechniques 2003, 35, 1140−1144.
(21) Feng, X.; Wang, Z.; Fillmore, R.; Xi, Y. Cancer Lett. 2014, 344,
1
66−173.
22) Barron, N.; Keenan, J.; Gammell, P.; Martinez, V. G.; Masters, J.
R.; Clynes, M.; Freeman, A. Prostate 2012, 72, 1193−1199.
23) Jackson, B. L.; Grabowska, A.; Ratan, H. L. BMC Cancer 2014, 14,
30.
24) Chung, Y.-W.; Bae, H.-S.; Song, J.-Y.; Lee, J. K.; Lee, N. W.; Kim,
T.; Lee, K.-W. I. J. Gynecol. 2013, 23, 673−679.
25) Schwarzenbach, H.; Nishida, N.; Calin, G. A.; Pantel, K. Nat. Rev.
Clin. Oncol. 2014, 11, 145−156.
26) Li, M.; Zeringer, E.; Barta, T.; Schageman, J.; Cheng, A.; Vlassov,
A. V. Philos. Trans. R. Soc., B 2014, 369, 20130502.
27) Cheng, L.; Sharples, R. A.; Scicluna, B. J.; Hill, A. F. J. Extracell.
Vesicles 2014, 3, 23743.
28) Schmittgen, T. D.; Lee, E. J.; Jiang, J.; Sarkar, A.; Yang, L.; Elton,
T. S.; Chen, C. Methods 2008, 44, 31−38.
29) Melo, S. A.; Sugimoto, H.; O’Connell, J. T.; Kato, N.; Villanueva,
(
REFERENCES
■
(
9
(
(
1) Hartwell, L.; Mankoff, D.; Paulovich, A.; Ramsey, S.; Swisher, E.
Nat. Biotechnol. 2006, 24, 905−908.
(
6
(
2) Karayi, M. K.; Markham, A. F. Prostate Cancer Prostatic Dis. 2004, 7,
−20.
3) Ferlay, J.; Steliarova-Foucher, E.; Lortet-Tieulent, J.; Rosso, S.;
Coebergh, J. W.; Comber, H.; Forman, D.; Bray, F. Eur. J. Cancer 2013,
9, 1374−1403.
4) Alberts, A. R.; Schoots, I. G.; Roobol, M. J. Int. J. Urol. 2015, 22,
24−532.
5) Lin, K.; Lipsitz, R.; Miller, T.; Janakiraman, S.; U.S. Preventative
(
(
4
(
5
(
(
(
Services Task Force. Ann. Intern. Med. 2008, 149, 192−199.
(
(
6) Kurman, R. J.; Shih, L.-M. A. J. Surg. Pathol. 2010, 34, 433−443.
7) Heidenreich, A.; Bellmunt, J.; Bolla, M.; Joniau, S.; Mason, M.;
(
A.; Vidal, A.; Qiu, L.; Vitkin, E.; Perelman, L. T.; Melo, C. A.; Lucci, A.;
Matveev, V.; Schmid, H.-P.; van der Kwast, T.; Zattoni, F.; Mottet, N.;
Ivan, C.; Calin, G. A.; Kalluri, R. Cancer Cell 2014, 26, 707−721.
Wiegel, T. Eur. Urol. 2011, 59, 61−71.
(30) Arroyo, J. D.; Chevillet, J. R.; Kroh, E. M.; Ruf, I. K.; Pritchard, C.
(
8) Vasioukhin, V.; Anker, P.; Maurice, P.; Lyautey, J.; Lederrey, C.;
C.; Gibson, D. F.; Mitchell, P. S.; Bennett, C. F.; Pogosova-Agadjanyan,
E. L.; Stirewalt, D. L.; Tait, J. F.; Tewari, M. Proc. Natl. Acad. Sci. U. S. A.
Stroun, M. Br. J. Haematol. 1994, 86, 774−779.
9) (a) Schwarzenbach, H.; Hoon, D. S. B.; Pantel, K. Nat. Rev. Cancer
011, 11, 426−437. (b) Sita-Lumsden, A.; Dart, D. A.; Waxman, J.;
(
2
2
(
011, 108, 5003−5008.
31) (a) McDonald, J. S.; Milosevic, D.; Reddi, H. V.; Grebe, S. K.;
Bevan, C. L. Br. J. Cancer 2013, 108, 1925−1930.
Algeciras-Schimnich, A. Clin. Chem. 2011, 57, 833−840. (b) Kim, D. J.;
Linnstaedt, S.; Palma, J.; Park, J. C.; Ntrivalas, E.; Kwak-Kim, J. Y.;
Gilman-Sachs, A.; Beaman, K.; Hastings, M. L.; Martin, J. N.; Duelli, D.
M. J. Mol. Diagn. 2012, 14, 71−80. (c) Cheng, H. H.; Yi, H. S.; Kim, Y.;
Kroh, E. M.; Chien, J. W.; Eaton, K. D.; Goodman, M. T.; Tait, J. F.;
Tewari, M.; Pritchard, C. C. PLoS One 2013, 8, e64795.
(
(
10) Ambros, V. Nature 2004, 431, 350−355.
11) (a) Nguyen, H. C.; Xie, W.; Yang, M.; Hsieh, C.-L.; Drouin, S.;
Lee, G-S.M.; Kantoff, P. W. Prostate 2013, 73, 346−354. (b) Srivastava,
A.; Suy, S.; Collins, S. P.; Kumar, D. Mol. Cell. Pharmacol. 2011, 3, 115−
124.
(
12) Bryant, R. J.; Pawlowski, T.; Catto, J. W.; Marsden, G.; Vessella, R.
L.; Rhees, B.; Kuslich, C.; Visakorpi, T.; Hamdy, F. C. Br. J. Cancer 2012,
06, 768−774.
13) (a) Wu, T.; Orgel, L. E. J. Am. Chem. Soc. 1992, 114, 7963−7969.
b) Wang, S. H.; Kool, E. T. Nucleic Acids Res. 1994, 22, 2326−2333.
c) Gryaznov, S. M.; Schultz, R.; Chaturvedi, S. K.; Letsinger, R. L.
1
(
(
(
Nucleic Acids Res. 1994, 22, 2366−2369. (d) Grossmann, T. N.; Roglin,
L.; Seitz, O. Angew. Chem., Int. Ed. 2008, 47, 7119−7122. (e) Pianowski,
Z.; Gorska, K.; Oswald, L.; Merten, C. A.; Winssinger, N. J. Am. Chem.
Soc. 2009, 131, 6492−6497. (f) Kleinbaum, D. J.; Kool, E. T. Chem.
Commun. 2009, 46, 8154−8156. (g) Erben, A.; Grossmann, T. N.; Seitz,
O. Angew. Chem., Int. Ed. 2011, 50, 2828−2832. (h) Shibata, A.; Abe, H.;
Ito, Y. Molecules 2012, 17, 2446−2463. (i) Gorska, K.; Winssinger, N.
Angew. Chem., Int. Ed. 2013, 52, 6820−6843. (j) Saneyoshi, H.; Ito, Y.;
Abe, H. J. Am. Chem. Soc. 2013, 135, 13632−13635. (k) Seckute, J.;
Yang, J.; Devaraj, N. K. Nucleic Acids Res. 2013, 41, e148. (l) Sadhu, K.
K.; Winssinger, N. Chem. - Eur. J. 2013, 19, 8182−8189. (m) Wu, H.;
Cisneros, B. T.; Cole, C. M.; Devaraj, N. K. J. Am. Chem. Soc. 2014, 136,
1
2
2
7942−17945. (n) Michaelis, J.; Roloff, A.; Seitz, O. Org. Biomol. Chem.
014, 12, 2821−2833. (o) Gorska, K.; Winssinger, N. Methods Mol. Biol.
014, 1050, 179−192.
(
14) (a) Meguellati, K.; Koripelly, G.; Ladame, S. Angew. Chem., Int. Ed.
2
010, 49, 2738−2742. (b) Percivalle, C.; Bartolo, J. F.; Ladame, S. Org.
Biomol. Chem. 2013, 11, 16−26. (c) Choi, Y.; Metcalf, G.; Haj-Sleiman,
M.; Vair-Turnbull, D.; Ladame, S. Bioorg. Med. Chem. 2014, 22, 4395−
4398. (d) Sleiman, M. H.; Ladame, S. Chem. Commun. 2014, 50, 5288−
5290.
(
15) Egholm, M.; Buchardt, O.; Christensen, L.; Behrens, C.; Freier, S.
M.; Driver, D. A.; Berg, R. H.; Kim, S. K.; Norden, B.; Nielsen, P. E.
Nature 1993, 365, 566−568.
(
16) (a) Geisen, U.; Berding, C.; Geiger, A.; Ørum, H.; Nielsen, P. E.
Nucleic Acids Res. 1998, 26, 5004−5006. (b) Igloi, G. L. Proc. Natl. Acad.
Sci. U. S. A. 1998, 95, 8562−8567.
(
17) Jung, H. S.; Ko, K. C.; Kim, G.-H.; Lee, A.-R.; Na, Y.-C.; Kang, C.;
Lee, J. Y.; Kim, J. S. Org. Lett. 2011, 13, 1498−1501.
18) Satsangi, J.; Jewell, D. P.; Welsh, K.; Bunce, M.; Bell, J. I. Lancet
994, 343, 1509−1510.
19) Al-Soud, W.; Radstrom, P. J. Clin. Microbiol. 2001, 39, 485−493.
(
1
(
H
DOI: 10.1021/acs.analchem.6b01594
Anal. Chem. XXXX, XXX, XXX−XXX