A New One-Pot Fluorescence Derivatization Strategy for Highly Sensitive MicroRNA Analysis

Article abstract of DOI:10.1002/chem.201905639

MicroRNAs (miRNAs) modulate the expression of over 30 % of mammalian genes during development and apoptosis, and abnormal expression of miRNAs may lead to a range of human pathologies. Therefore, analysis of miRNAs is valuable for disease diagnostics. In this work, a novel one-pot fluorescence derivatization strategy was developed for miRNA analysis. The mechanism of the derivatization reaction was explored by using instrumental methods, including liquid chromatography, fluorescence spectroscopy, and mass spectrometry. Highly fluorescent N⁶-ethenoadenine (?-adenine) was formed and detached from the miRNA sequence through the reaction of adenine in nucleic acids with 2-chloroacetaldehyde (CAA) at 100 °C. This is the first experimental evidence that the cooperation of formed ?-adenine and water-mediated hydrogen-bond interaction between the proton at the 2′- and the oxyanion at 3′-positions stabilized the oxocarbenium significantly, which makes the depurination and derivatization of miRNA highly effective. Based on this derivatization strategy, a facile and sensitive high-performance liquid chromatography method was developed for quantitative assay of miRNAs. In combination with magnetic solid-phase extraction (MSPE), the HPLC method was shown to be useful for the determination of microRNAs at sub-picomolar level in serum samples.

Full text of DOI:10.1002/chem.201905639

A Journal of
Accepted Article
Title: A Novel One Pot Fluorescence Derivatization Strategy for Highly
Sensitive MicroRNA Analysis
Authors: Shulin Zhao, Li Pan, Huaisheng Zhang, Jingjin Zhao, Ifedayo
Victor Ogungbe, and Yi-Ming Liu
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201905639
Link to VoR: http://dx.doi.org/10.1002/chem.201905639
Supported by

10.1002/chem.201905639
Chemistry - A European Journal
A Novel One Pot Fluorescence Derivatization Strategy for Highly Sensitive
MicroRNA Analysis
Li Pan, [a] Huaisheng Zhang, [a] Jingjin Zhao, [a,b] Ifedayo Victor Ogungbe,[a] Shulin Zhao,*[b] and Yi-Ming
Liu*[a]
Abstract: MicroRNAs (miRNAs) modulate the expression of over
30% of mammalian genes during development and apoptosis,
abnormal expression of miRNAs may lead to a range of human
pathologies. Therefore, analysis of miRNAs is significantly valuable
for disease diagnostics. In this work, a novel one pot fluorescence
derivatization strategy was developed for miRNA analysis. The
mechanism of the derivatization reaction was explored by using
sensitivity. In order to improve the sensitivity, various amplification
strategies have been proposed, such as reverse transcription
quantitative polymerase chain reaction (RT-qPCR),[7,8] rolling circle
amplification,[9] exponential amplification reaction,[10] hybridization
chain reaction,[11] catalytic hairpin assembly,[12] loop-mediated
isothermal amplification,[13] strand-displacement amplification,[14]
and duplex-specific nuclease signal amplification.[15] Among these
amplification methods, labeling the probes with fluorescence dyes is
one of the essential requirement to realize signal amplification.
However, these labeling procedures are complicated, time-
consuming and involves expensive labeling reagent.[16-19] In
addition, fluorescence dye used in these labeling procedures itself is
fluorescent which leads to high background. Therefore, a simple,
fast, zero background, low-cost labeling method for miRNA
detection is highly desired.
As a nucleophilic reagent with no fluorescence itself, 2-
chloroacetaldehyde (CAA) was used to label adenine and its analog
forming the corresponding fluorescence products, which can realize
the quantification of these small molecules,[20,21] and the labeling
reaction offers a high derivative yield. It can be used to detect DNA
lesions involved adenine. However, this labeling reaction is
inseparable from tedious operation and time consuming,[22] such as
multiple step enzymatic hydrolysis or acid hydrolysis to obtain free
ε-adenine. Moreover, CAA was only proposed for labeling nucleic
acids to detect DNA, and detection sensitivity for DNA was µM
level.[23]
Herein, a novel one pot fluorescence derivatization method for
miRNA is established. The one-pot derivatization reaction is a green
and highly efficient approach due to the reduction of work-up
procedures and purification steps. With the aid of DNA, dAMP and
AMP, the mechanism of one-pot fluorescence derivatization
reaction was also investigated and elucidated. The formed ε-adenine
and hydrogen bond significantly stabilized the oxocarbenium of
miRNA, making ε-adenine leave easily. The ε-adenine as a report
reflects the concentration of microRNA. Based on this founding, a
facile and cost-effective method for absolute quantification of
microRNAs was developed. CAA can effectively derivatize and de-
purine miRNAs in one pot reaction, producing quantitatively
fluorescent products, i.e. etheno-adenine (ε-adenine) and etheno-
guanine (ε-guanine). After being separated from other fluorescent
species in the derivatization solution by high performance liquid
chromatography (HPLC), ε-adenine that had a much higher
fluorescence yield comparing to ε-guanine was quantified.
Experimental conditions were studied to achieve optimal analytical
figures of merit. To attest the applicability of the present method in
clinical analysis, it was utilized to determine miRNA-21 in serum
samples. An effective magnetic solid phase extraction (MSPE) of
target miRNA from sample matrix by using a complementary DNA
probe immobilized on magnetic beads was developed to ensure
assay specificity for the miRNA target, and further improve the
sensitivity of analysis.
instrumental
methods,
including liquid
chromatography,
fluorescence, and mass spectrometry. Highly fluorescent N6-
ethenoadenine (ε-adenine) was formed and detached from the
miRNA sequence through the reaction of adenine in nucleic acids
with 2-chloroacetaldehyde (CAA) at 100 ℃ . This is the first
experimental evidence that the cooperation of formed ε-adenine and
water-mediated H-bond interaction between the proton at 2՛ and the
oxyanion at 3՛ positions stabilized the oxocarbenium significantly,
which make the depurination and derivatization of miRNA highly
effective. Based on this derivatization strategy, a facile and sensitive
high-performance liquid chromatography (HPLC) method was
developed for quantitative assay of miRNAs. In combination with
magnetic solid phase extraction (MSPE), the HPLC method was
proven useful for the determination of microRNAs at sub pico-
molar level in serum samples.
Introduction
MicroRNAs (miRNAs) are a class about 22 nucleotides noncoding-
RNAs which play significant roles in many cellular processes such
as cell proliferation, migration, and apoptosis. And miRNAs serve
as RNA silencers and gene post-transcriptional regulators.[1,2]
Abnormal expression or alteration of miRNAs also leads to a range
of human pathologies, including cancer, which suggest that the
miRNAs are reliable disease markers.[3,4] Therefore, the
development of assays to quantify miRNAs for clinical diagnosis
has generated a lot of interest. Current standard methods for miRNA
assay are northern blotting and microarray with high-throughput and
multiplex miRNA gene expression screening capability.[5,6]
However, it’s well documented that these methods have low assay
[a] Dr. L. Pan, Dr. H. Zhang, Prof. J. Zhao, Prof. I. V.
Ogungbe, Prof. Y. M. Liu
Department of Chemistry and Biochemistry, Jackson State
University, 1400 Lynch Street, Jackson, Mississippi
39217, United States
E-mail: yiming_liu@jsums.edu
[b] Prof. J. Zhao, Prof. S. Zhao
State Key Laboratory for the Chemistry and Molecular
Engineering of Medicinal Resources, Guangxi Normal
University, Guilin 541004, China
E-mail: zhaoshulin001@163.com
Supporting information and the ORCID identification
number(s) for the author(s) of this article can be found
under: https://doi.org/10.1002/chem.xxxxxxxxx.
1
This article is protected by copyright. All rights reserved.

10.1002/chem.201905639
Chemistry - A European Journal
Figure 1. Chromatograms of reaction products under different conditions. The mixture of 95 µM cytosine, 3.6 µM guanine and 0.36 µM
adenine was reacted with CAA at 100°C for 15 min (A). Adenine, dAMP, dAdo, AMP and Ado were reacted with CAA at 100 ℃ for 15 min
(curve a) and 37 ℃ for 1 h (curve b), respectively (B-F). The structure of each fraction was detected by MS (Figure S1, Supporting
Information). Reaction condition: 2 µL of adenine (1 mM), dAdo (3 mM), Ado (0.7 mM), dAMP (1.6 mM), or AMP (1.6 mM) were reacted
with CAA (100 µL) in 1xPBS (898 µL) at 37 ° C for 1 h; 5 µL of Adenine (0.1mM), dA (0.3 mM), A (0.07 mM), dAMP (0.16 mM), or
AMP (0.16 mM) were reacted with CAA (50 µL) in 1xPBS (495 µL) at 100 °C, for 15 min.
conditions (37 ℃ and 100 ℃ at atmospheric pressure). Despite the
fact the AMP was not depurinated when reacted with CAA at
100 °C, we found that miRNA is quickly depurinated, just like DNA,
when reacted with CAA (Figure 2A and Figure 2B). Moreover, the
identity of ε-guanine was verified through detecting fraction at 5.1
min by MS. The mass spectrum was shown as Figure S2 (supporting
information)
Results and Discussion
Study of reaction mechanism for the fluorescence derivatization
In the one pot reaction between miRNA and CAA at 100 ℃, the
formed fluorescence derivatives include ε-adenine, ε-cytosine and ε-
guanine. The uracil base can not be derivatized by CAA.[24]
However, the fluorescence produced by ε-cytosine is too weak to be
used for quantification.[20] Therefore, the excitation-emission spectra
of ε-adenine and ε-guanine were investigated. The results indicate
that its maximum excitation and emission wavelengths are 275 nm
and 411 nm, respectively, and these derivatives can be easily
separated and detected by HPLC-FD. As shown in Figure 1A, ε-
cytosine, ε-guanine and ε-adenine have retention time of 3.2 min,
5.1 min, 7.1 min, respectively. The fluorescence yield ratio of ε-
cytosine, ε-guanine and ε-adenine was 1:87:870, which
demonstrated that the ε-adenine provides the highest fluorescence
yield. Thus, ε-adenine can be used as fluorescence signal for
miRNAs analysis in the one-pot reaction, and ε-adenine was further
investigated as a quantification target for the depurination. DNA,
AMP and dAMP were chosen as control to figure out the one pot
reaction mechanism.
The retention times of the etheno-adducts formed when CAA is
reacted with adenine, dAMP, dAdo, AMP and Ado at 37 ℃ for 1 h
were determined using HPLC and their identities were verified by
MS (Figure S1B-F, Supporting Information). The adenine, dAMP,
AMP, dAdo and Ado were converted to ε-adenine, ε-dAMP, ε-AMP,
ε-dAdo and ε-Ado, respectively (Figures 1B-F, curve b). However,
ε-adenine was formed when adenine, dAMP and dAdo were reacted
with CAA at 100 ℃ for 15 min, but the reaction with AMP and Ado
led to the formation of ε-AMP and ε-Ado, respectively (Figures 1B-
F, curve a). Since similar etheno-adducts were obtained at 37 ℃
Figure 2. Chromatograms of reaction products obtained from 20 nM
miRNA (A) (5’-GUA AGG CAU CUG ACC GAA GGC A-3’) and
15 nM DNA (B) (5’-TCA ACA TCA GTC TGA TAA GCT A-3’)
reacted with CAA at 100°C for 15 min, respectively.
and 100 ℃ for AMP and Ado, the 2՛ hydroxyl group may play a
role in the stabilization of ε-AMP and ε-Ado under ambient reaction
2
This article is protected by copyright. All rights reserved.

10.1002/chem.201905639
Chemistry - A European Journal
intermediate would be considered. Because the electronegativity of
2’-OH, the oxocarbonium ion intermediate is even less stable for 2’-
OH nucleosides, which protect them from depurination. The
In order to find out the transition state of this one pot reaction，
the reaction kinetic of dAMP, AMP, DNA and microRNA with
CAA were subsequently investigated, respectively. The reaction
kinetic curves of dAMP with CAA were shown as Figure 3A. As
can be seen, the fluorescence intensity of product (ε-dAMP) arrives
the maximum at 5 mins and depurated completely at 15 mins.
Meanwhile, the fluorescence intensity of the product (ε-adenine)
increased gradually with the reaction time, and reached the
maximum at 15 min, remains almost unchanged within 25 min.
From these two dynamic curves, 5 min was the time required to
form ε-dAMP, and 10 min was the time for the following depuration
step of ε-dAMP. Therefore, it can be concluded that it involved two
step successive reactions, forming ε-dAMP and subsequent
depurination. In other words, modification of adenine in dAMP
occurred firstly, then, the formed ε-adenine part weakens the N-
glycosidic bond to trigger the successive depurination reaction. It is
worth noting that the maximum fluorescence intensity of ε-adenine
was stronger than fluorescence intensity of ε-dAMP, and the reason
was further researched by quantification of the products using
HPLC-ESI-TQ-MS/MS in multiple-reaction monitoring (MRM)
mode. In the MRM mode, the peak only appeared when the parent
and daughter ions were both detected through setting up the
appropriate parent and daughter ions, so the influence caused by
other substances could be avoided. Quantification was performed by
the MRM mode after setting up the parameters of molecule
ion→fragment ion (ε-Adenine m/z 160 → m/z 106, ε-dAMP m/z
356.5 → m/z 160.23, ε-dAdo m/z 276.28 → m/z 160.09, ε-Ado m/z
292.24 → m/z 160.03, ε-AMP m/z 372.63 →m/z 160.04, Adenine
m/z 136 → m/z 119, Adenine m/z 136 → m/z 119, dAMP m/z 332
→m/z 136, AMP m/z 348 → m/z 136). In order to track the
intermediate products in the reaction process of dAMP with CAA,
the reaction was terminated at 3 min. Then the products were
identified and quantified. As shown in Figure S3a (Supporting
Information), from the abundance of each product, dAMP (m/z
348→m/z 136) remains around 15 % with the order of magnitude of
105 which demonstrated that the reaction was stopped in the early
stage of the reaction process. This is an ideal period for detection of
all intermediate products. The major product is ε-dAMP (m/z
356→m/z 161) with the order of magnitude of 106. The abundance
of adenine is 105 which is similar as the blank control without CAA
Figure S3b (Supporting Information), thus it illustrated that the
adenine originated from dAMP depurination spontaneously. Partial
of ε-adenine came from the spontaneously generated adenine. This
also gives a reasonable explanation for why the peak intensity of ε-
adenine is stronger than the peak intensity of ε-dAMP though ε-
adenine and ε-dAMP had the very similar fluorescence yield. The ε-
adenine (m/z 160 → m/z 106) is the smallest abundance, and it’s
hard to say where it exactly originates from, which has two
possibilities, one is from the adenine spontaneously degraded from
dAMP, and the other is from the formed un-stabilized ε-dAMP.
The reaction kinetic curves of AMP with CAA were shown as
Figure 2B. The fluorescence intensity of product ε-AMP
dramatically increased to the maximum within 5 min then kept
almost same. Comparing the AMP and dAMP, cleavage of the
glycosyl bond after reaction of CAA with dAMP occurred and
completed within 15 min at 100 ℃, whereas the loss of ribose
following reaction of AMP was extremely slow, revealing the 2’-
OH group in AMP imparts stability to the glycosidic bond.
Exploring the reason for 2’-OH stabilizing N-glycosidic bond of
AMP to impede the depurination of ε-AMP, the oxocarbonium ion
observations provide
a strong evidence for the inductive
stabilization of the ε-AMP adduct by the presence of a 2՛ hydroxyl
group.[25,26] Moreover, the identity of formed product by the reaction
of AMP and CAA at 100℃ was evaluated by LC-MS shown as
Figure S4 (Supporting Information). The majority product is ε-AMP
with magnitude of 107, and other products including adenine (m/z
136 → m/z 119), ε-adenine (m/z 160 → m/z106), AMP (m/z 348 →
m/z 136) have totally account for 1% which was negligible. It is
understandable that 2’-OH significantly stabilized the N-glycosidic
bond to prevent adenine or ε-adenine releasing from AMP or ε-AMP,
respectively, though the formed ε-adenine from AMP decrease the
stability of N-glycosidic bond. The result is well consistent with the
HPLC results.
The reaction kinetic curves of DNA and miRNA with CAA were
shown as Figure 3C and Figure 3D, respectively. When DNA and
miRNA react with CAA at 100 ℃, both of them produce ε-adenine.
and they present similar kinetic curve. Comparing the dynamic
curves of miRNA and DNA, maximal fluoresce intensity of them all
were achieved at 15 min. This kinetic time is in good agreement
with the reaction time of dAMP with CAA previously investigated.
Moreover, the kinetic curve of miRNA also was validated through
monitoring the ε-adenine concentration (m/z 160→m/z 106) by
using liquid chromatography-electrospray ionization-tandem mass
spectrometry (LC-ESI-MS/MS). As shown in Figure S5 (Supporting
Information), when miRNA react with CAA at 100 ℃, ε-adenine
concentration reached the maximum at 15 min. However, no ε-
adenine was detected when the reaction was carried out at 37 ℃.
Therefore, two consecutive reactions involved in the one-pot
reaction of DNA or miRNA were verified. And it’s well known that
RNA is much less vulnerable to depurination since the
electronegative 2’ oxygen renders the oxocarbonium ion
intermediate less stable.[27,28] However, it’s interesting that when
miRNA react with CAA at 100℃, it did produce ε-adenine. This
suggests that the stabilizing effect of the 2՛ hydroxyl group is absent
or minimal in the etheno-miRNA intermediate. The loss of inductive
stabilization in the etheno-miRNA intermediate is most likely a
water-mediated H-bond interaction between the proton at 2՛ and the
oxyanion at 3՛ positions.[29,30] The formation of hydrogen bond
abated the electronegativity of 2’-OH, contributing to the stability of
oxocarbonium and thus depurination becomes relatively easier.
However, for AMP, hydrogen bonding between the 2’-hydroxyl and
the 3’-oxygen (or vice versa) in nucleosides has not been detected in
water except in organic aprotic solvents has been proved.[31,32] It can
be concluded that there is either no hydrogen bond between the 2’-
and 3’-oxygen in the monoanion or that this does not contribute
toward the stability of the anion in water.[28]
Mechanism pathway for one pot reaction of miRNA with CAA
Mechanism pathway of one pot reaction of microRNA with CAA
was described as follows. CAA is an electrophile reagent whose
carbonyl group attack exocyclic 6-amino group of adenosines of
microRNA to initiate the condensation,[33-35] leading to an unstable
intermediate, and the reaction is reversible. This intermediate then
undergoes intramolecular cyclization to a stable intermediate, as the
result of alkylation of endocyclic N1, and this process is irreversible.
This stable intermediate undergoes the dehydration to the
3
This article is protected by copyright. All rights reserved.

10.1002/chem.201905639
Chemistry - A European Journal
Figure 3. The reaction kinetic curves of 6.5×10-8 M dAMP (A), 1.2×10-7 M AMP (B), DNA (C) and microRNA (D) reacted with CAA at
100℃, respectively.
intermediate N6-etheno, following protonation reaction. Evidence
for adding H at N7 position is that the most stable tautomer in the
gas phase is calculated to be the “N7” tautomer 7 (proton resides on
the N7).[36] Protonation at N-7 can strongly catalyzes the hydrolysis
of the N-glycosidic bond, making the adenine a better leaving group
has been studied using density functional theory (DFT) methods.[37]
Cleavage of the C-N glycosyl bond results in the formation of the
oxocarbenium ion intermediate, and an electrically neutral ε-adenine
electrophile adduct. Formation of the oxocarbenium ion
intermediate in the process of depurination is supported by the
strong resistance of RNA to depurination. The reason is the positive
charge located on the C2 carbon of ribose due to the presence of the
strongly polarized C2-O bond. Thus, formation of the oxocarbenium
ion would require formation of a structure in which the two
neighboring carbon atoms (C1and C2) are positively charged, which
is obviously unfavorable. Cleavage of the glycosyl bond takes place
most likely by participation of one pair of nonbonding electrons of
the deoxyribose ring oxygen (the pair that is in the anti-orientation
to the C-N glycosyl bond due to the anomeric effect). Whether ε-
adenine can leave successfully, in other words, whether the
glycoside bond can be cleaved depending on whether the
oxocarbenium ion intermediate can be formed. Generally, the higher
electronegativity of 2’ position, the less stable of oxocarbonium ion
intermediate.[27,28,38] For dAMP and DNA, both of them without 2-
OH experienced the etheno modification and depurination. Then, the
extremely reactive oxocarbenium ion adds a water molecule, giving
deoxyribose as the final product (Figure S6a-b, Supporting
Information). The attack of an electrophile (CAA) at the exocyclic
6-amino of AMP results in the formation of final product ε-AMP,
which is not favorable to stabilize oxocarbenium ion to break the
glycosyl bond (Figure S6c, Supporting Information). However,
miRNA with 2’-OH, the coordination of water-mediated H-bond
interaction between the proton at 2՛ and the oxyanion at 3՛ positions
and the formed adenine adduct stabilized the oxocarbenium ion
intermediate, leading to modification and depurination occur
successfully (Figure 4).
Figure 4. The mechanism pathway in the depurination of miRNA
Study on one pot reaction for miRNAs with different numbers
of adenire
In order to figure out if the miRNA is able to be fully degraded into
adenine in this one pot reaction, the reaction of two different kinds
of miRNAs with different number of adenine bases with CAA at
100 ℃ was investigated. One is 5’-UAGCUUAUCGGACUGAUG
UUGA-3’ containing 5 adenine bases (5A) and the other is 5’-
GUAAG GCAUCUGACCGAAGGCA-3’ containing 7 adenine
bases (7A). The results were shown in Figure 5, at the same
concentration either 2.0 nM or 20 nM, the more adenine bases the
miRNA contains, higher fluorescence peak (fluorescence intensity).
Comparing miRNA with adenine, miRNA with 5 adenine bases
4
This article is protected by copyright. All rights reserved.

10.1002/chem.201905639
Chemistry - A European Journal
yields 5 times stronger fluorescence intensity as same concentration
of adenine. Similarly, miRNA with 7 adenine bases yields 7 times
higher fluorescence intensity as same concentration of adenine.
When comparing different concentrations of miRNA, we found that
20 nM of miRNA present 10 times stronger fluorescence intensity as
2 nM does. Therefore, we draw the conclusion that the fluorescence
intensity is proportional to the concentration of miRNA. It provides
the basis for the quantification of miRNA.
Scheme1. Schematic diagram for the working flow of the
proposed miRNA assay.
Figure 5. Study on specificity of derivatization reaction for
miRNAs with different numbers of adenine bases. Standard adenine
sample was carried out the same experimental procedure as miRNA.
Application of derivatization reaction for quantification of
miRNA-21.
Based on the above finding, a method for detection of miRNA-21
was developed (Scheme 1). The working flow of the proposed
miRNA assay is described as following: (1) DNA probe-magnetic
beads were added to the sample, allowing the fully complementary
target miRNA-21 be magnetically separated from the sample matrix
by means of the formation of an miRNA-21-DNA probe-MB
complex; (2) after the magnetic separation, miRNA-21 was released
from the complex by denature; (3) miRNA-21 was then reacted with
CAA, resulting in its completely derivatized and de-purined in one
pot reaction of CAA, producing highly detectable fluorescent ε-
adenine; and (4) ε-adenine was separated from other fluorescent
derivatives and quantitatively determined by HPLC-FD.
Under the CAA derivatization and depurination conditions
selected, the dynamic range and detection sensitivity for miRNA-21
assay were evaluated. Figure 6 shows HPLC-FD chromatograms
obtained from CAA derivatization solutions of miRNA-21 at
varying concentrations. As can be seen, ε-adenine was well
separated with other fluorescent compound (ε-guanine) formed in
the reaction, and the fluorescence intensity of ε-adenine was
proportional to miRNA-21 concentration (inset in Figure 6). A
linear calibration curve of fluorescence intensity versus miRNA-21
concentration was obtained in a concentration range from 0.28 nM
to 126 nM. The linear regression equation is: FI= 6.6065 C + 1.4372,
R2=0.9999. Where FI is fluorescence intensity in mV, and C is
miRNA-21 concentration in nM. The limit of detection was
estimated to be 28 pM (S/N=3). To investigate the reproducibility of
this HPLC-FD method, a 15 nM miRNA-21 solution was analyzed
for 5 times. The fluorescence intensity and retention time of ε-
adenine were recorded, respectively. The reproducibility (RSD) of
fluorescence intensity was 1.2% and of retention time was 1.0%
(n=5). These results indicate the proposed HPLC-FD method with
CAA precolumn derivatization offers high sensitivity and good
reproducibility. It can be useful for quantification of miRNAs.
Figure 6. Chromatograms obtained from CAA derivatization
solutions of miRNA-21 at concentrations from 0 to 126 nM. Inset is
a plot of fluorescence intensity of ε-adenine against miRNA-21
concentration.
Selectivity of method for miRNA-21 assay
Multiple miRNAs are present in every biological specimen.
Furthermore, since they chain length are short, they have highly
similar sequences that differ by only a few nucleotides. Therefore,
the methods that capable of performing single nucleotide
polymorphism assays are high desired. In the present method,
miRNA assay specificity is achieved via DNA probe/miRNA target
hybridization and magnetic separation by using a DNA probe-
magnetic bead conjugate. After collecting the magnetic beads easily
by using an external magnet, the targeted miRNA is isolated
/enriched from the sample matrix as a DNA probe/miRNA-magnetic
bead complex. The complex is then denatured at 90 ℃ for 3 min,
releasing miRNA into a clean chemical environment. Preparation of
the conjugate from biotin-ssDNA complementary to miRNA-21 and
streptavidin-magnetic beads was investigated (Figure S7 and Figure
S8, Supporting Information).
The specificity of the proposed method was assessed by
comparing HPLC-FD chromatograms obtained from standard
solutions of miRNA-21, single-base mismatched miRNA, three-base
mismatched miRNA, non-complementary miRNA, miRNA-141 and
miRNA-155 (all at 28 nM). It is well known that hybridization
depends on the temperature and double-stranded DNA/RNA is
susceptible to denaturation at higher temperature.[38,39] Avoiding
interference of base mismatch miRNAs by elevating hybridization
5
This article is protected by copyright. All rights reserved.

10.1002/chem.201905639
Chemistry - A European Journal
temperature was reported.[40] Therefore, hybridization at 50 ℃ was
performed in this work. As shown in Figure 7, the fluorescence
intensity from single-base mismatched microRNA and three-base
mismatched microRNA were < 15% of that from miRNA-21 at the
same concentration. Those from non-complementary miRNA,
miRNA-141 and miRNA-155 were almost the same as that from a
blank. These results indicate that the present method offers good
selectivity for quantitative assay of target miRNAs.
To further improve the sensitivity of the present assay, the
MSPE was utilized to enrich the miRNA-21 in serum samples. 2.0
mL bovine serum containing RNase inhibitor was spiked with
miRNA at 5.0 pM, and then diluted with binding buffer (1+9). 200
μL DNA probe-magnetic beads were added in 20 mL diluted serum
sample. The mixture was then incubated at 50 ℃ for 30 minutes.
After the magnetic beads were collected and washed twice, miRNA-
21 was released into 200 μL buffer after denatured. A portion of the
solution (100 μL) was derivatized with CAA and then analyzed by
HPLC-FD method. As shown in Figure S9 (Supporting Information),
based on the ε-adenine fluorescence intensity of three replicate
analyses, miRNA-21 concentration in serum samples was measured
to be 4.9 ± 0.22 pM. Recovery was 98.0±4.4% (n=3), indicating the
assay was accurate. It should be mentioned that based on the volume
changes in MSPE (i.e. from 20 mL to 200 μL), miRNA-21
concentration increased 100 times. It’s also worth noting that
miRNA-21 concentration in these diluted serum samples was 0.5
pM, i.e. 500 fM. This detection limit is comparable or even better
than some of detection method with complex enzymatic recycling
amplification (Table S2).[44-49] These results indicate that the
proposed assay may be useful for analysis of biological fluid
samples such as urine and cell culture media to quantify miRNA
targets at the fM level.
Figure 7. Specificity of the method for miRNA-21 quantification.
The concentration of miRNA is 28 nM.
Conclusions
In summary, we report herein a one pot fluorescence derivatization
strategy for microRNA with CAA, which has been proven to be a
facile way to modify the adenine base and to depurinate the formed
adenine adducts. This is the first experimental evidence that the
cooperation of the formed ε-adenine and water-mediated H-bond
interaction between the proton at 2՛ and the oxyanion at 3՛ positions
significantly stabilized the oxocarbenium, and thus the microRNA is
completely degraded into ε-adenine. Deploying this fluorescence
derivatization strategy, a novel HPLC-FD method is developed for
quantification of targeted miRNAs. The method involves no target
conversion/cycling-based isothermal nor PCR-based signal
amplification. In combination with a magnetic solid phase extraction
(MSPE) of the miRNA target with its complimentary DNA probe
modified magnetic beads, the proposed method is shown to be
effective for quantification of trace miRNAs spiked into bovine
serum samples at 500 fM without needing total RNA isolation, and
with the capability of single base mismatch discrimination.
Repeatability of the proposed assay (RSD) is < 4.5 %, which is
better than most protocols previously reported. The proposed assay
offers a facile and cost-effective means to quantify trace miRNAs in
biological specimen.
Quantification of miRNA-21 in serum samples
To access the applicability of the present assay in biological and
clinical analysis it was applied to determine miRNA in bovine
serum. In order to avoid degradation of microRNA by endogenous
serum RNases, RNase inhibitor was added,[41] the RNase inhibitor
was added to the 10 times diluted serum with binding buffer then the
serum samples were spiked with miRNA-21 at 0.8 nM, 8.0 nM, and
80.0 nM. Since serum samples were very viscous, they had to be
diluted to obtain a good efficiency of MSPE of miRNA-21 by using
DNA probe-magnetic beads. Five replicate runs were conducted on
100 μL portions of each diluted serum sample to determine their
miRNA-21 contents. Analytical data are summarized in Table 1. As
can be seen, the recovery was found in the range of 94-106% with a
good repeatability (RSD <4.0% in all three cases). These results
showed that there was no interference with the quantification of
miRNA-21 from the endogenous components in this sample matrix.
It’s worth noting that the targeted miRNAs were quantitatively
isolated /enriched from this very “dirty” sample matrix by using
DNA probe-magnetic beads via hybridization and magnetic
separation, and that this isolation procedure was simple and quick.
The biggest advantages of the present assay are its ease of operation
and high repeatability, which is significantly better than PCR based
quantitative assays.[7,8,42,43]
Experimental Section
Table 1. Determination of miRNA-21 in bovine serum samples
Chemicals and reagents
Chloroacetaldehyde dimethyl acetal, sodium chloride (NaCl),
magnesium chloride (MgCl2), adenine, cytosine, guanine,
hydroxymethyl aminomethane (Tris), bovine serum, triethylamine
(TEA), acetic acid, ethylenediaminetetraacetic acid (EDTA), triton
X-100, concentrated H2SO4 were obtained from Sigma-Aldrich.
PBS tablets were purchased from Gibco. All microRNA or DNA
were purchased from Sigma-Aldrich. And the sequences were listed
in the Table S1. RNase inhibitor was purchased from New England
Biolabs Ltd. Diethylpyrocarbonate (DEPC) treated water and
PierceTM Streptavidin magnetic beads (1 μm) were purchased from
Sample
No.
1
Added
(nM)
0
0.80
8.0
Found
(nM)
ND
0.85
7.5
Recovery
RSD
(%, n=5)
-
(%)
-
106
94
2
3
3.8
2.5
4
80.0
78.0
97.5
3.3
ND: not detected.
6
This article is protected by copyright. All rights reserved.

10.1002/chem.201905639
Chemistry - A European Journal
Invitrogen. All miRNA stock solutions were prepared daily with
DEPC-treated water in a RNase-free environment. Binding/wash
buffer was composed of 10 mM Tris, 2 M NaCl, 1 mM EDTA, and
0.0005% Triton X-100 (pH 7.5). To avoid the effect of RNase on
the stability of miRNA solutions, all centrifuge tubes, pipette tips,
and buffers were autoclaved before use.
added. The solution was heated at 100 °C (in a boiling water bath)
for 15 min. The reaction vials were removed from boiling water bath
and placed on ice to cool down. Portions (10 μL) of the
derivatization solution were injected into the system for HPLC-FD
analysis.
Acknowledgements
Financial support from the U.S. National Institutes of Health
(GM089557 to YML) and the National Natural Science Foundation
of China (No. 21874030 to S. Z) are gratefully acknowledged.
HPLC-FD analysis
Samples were analyzed using a Shimadzu 20A HPLC system that
consists of a LC-20 AD liquid chromatograph, a DGU-20A5
degasser, a LC-20AHT auto sampler, an RF-10AXL fluorescence
detector, and a CTO-20A column oven (Shimadzu, Nakagyo-Ku,
Kyoto, Japan). An ODS column (C18, 250 x 4.60 mm, particle size 5
μm) from Phenomenex was used. The mobile phase consisted of 0.1
M triethylammonium acetate buffer (TEAA) at pH 7.0 and
acetonitrile (93:7 v/v). Separation was performed under isocratic
conditions at a flow rate of 1.0 ml min-1. Column temperature was
maintained at 30 °C. For fluorescence detection, the excitation and
emission wavelengths were set at 275 and 411 nm, respectively.
Sample volume injected was 10 μL.
Conflict of interest
The authors declare no conflict of interest.
Keywords: microRNA · etheno-adenine · one pot reaction ·
magnetic solid phase extraction · fluorescence detection.
[1] D. P. Bartel, Cell 2004, 116, 281-297.
[2] V. Ambros, Nature 2004, 431, 350-355.
[3] A. Lujambio, S. W. Lowe, Nature 2012, 482, 347-355.
[4] C. L. Bartels, G. J. Tsongalis, Clin. Chem. 2009, 55, 623-631.
[5] M. Lagos-Quintana, R. Rauhut, W. Lendeckel, T. Tuschl,
Science 2001, 294, 853-858.
[6] J. M. Thomson, J. Parker, C. M. Perou, S. M. Hammond, Nat.
Methods 2004, 1, 47-53.
LC-MS analysis
C18 column (5 cm×2.1 mm, particle size 3μm) was used as
separation column, 5%AcCN in water with 0.1% formic acid was
used as mobile phase. Separation was performed under isocratic
conditions at a flow rate of 0.2 ml min-1.
[7] H. Jia, Z. Li, C. Liu, Y. Cheng, Angew. Chem., Int. Ed. 2010, 49,
5498-5501.
Preparation of CAA [50]
Concentrated H2SO4 (1.0 mL) was diluted 10-times with distilled
H2O and added to 10.0 mL chloroacetaldehyde dimethyl acetal. The
mixture was distilled slowly under a fume hood and the distillate
fraction containing chloroacetaldehyde (at ca. 1.5 M) was collected
at 80-85 °C. The reagent solution was stored in darkness at 4 °C.
[8] C. Chen, D. A. Ridzon, A. J. Broomer, Z. Zhou, D. H. Lee, J. T.
Nguyen, M. Barbisin, N. L. Xu, V. R. Mahuvakar, M. R.
Andersen, Nucleic Acids Res. 2005, 33, e179.
[9] R. Deng, L. Tang, Q. Tian, Y. Wang, L. Lin, J. Li, Angew.
Chem., Int. Ed. 2014, 53, 2389-2393.
[10] H. Jia, Z. Li, C. Liu, Y. Cheng, Angew. Chem., Int. Ed. 2010,
49, 5498-5501.
[11] H. M. Choi, J. Y. Chang, L. A. Trinh, J. E. Padilla, S. E. Fraser,
N. A. Pierce, Nat. Biotechnol. 2010, 28, 1208-1212.
[12] Z. Cheglakov, T. M. Cronin, C. He, Y. Weizmann, J. Am.
Chem. Soc. 2015, 137, 6116-6119.
[13] C. Li, Z. Li, H. Jia, J. Yan, Chem. Commun. 2011, 47, 2595-
2597.
[14] C. Shi, Q. Liu, C. Ma, W. Zhong, Anal. Chem. 2014, 86, 336-
339.
Preparation of DNA probe-magnetic bead conjugate
Streptavidin-magnetic bead suspension was washed twice by
binding/washing buffer, and then resuspended in binding buffer.
Biotin-DNA probe solution (5×10-4 M, 5 μL) was added to the
above solution (100 μL), and the mixture was shaken at 37 ℃ for 30
min. The magnetic beads were collected by using an external
magnet. The beads were washed with 1xPBS solution three times to
remove any unbound biotin-DNA probe. DNA probe-magnetic bead
conjugate was resuspended in 1.0 mL binding/washing buffer.
[15] B. C. Yin, Y.Q. Liu, B. C. Ye, J. Am. Chem. Soc. 2012, 134,
5064-5067.
[16] D. Proudnikov, A Mirzabekov, Nucleic Acids Res. 1996, 24,
4535-4532.
[17] R. R. Hodges, N. E Conway, L. W. McLaughlin, Biochemistry
1989, 28, 261-267.
The miRNA derivatization and depurination with CAA
A portion of sample solution (100 μL) was mixed with 100 μL
1×PBS buffer in a vial. The CAA solution prepared above (20 μL)
was added. The mixture was heated at 100 °C (in a boiling water
bath) for 15 min. The vial was then placed on ice to cool down. The
fluorescence wavelength maxima were set at λex = 275 nm and λem
=
[18] P. J. Oefner, C. G Huber, F. Umlauft, G. N. Berti, E. Stimpfl,
G. K. Bonn, Anal. Biochem. 1994, 223, 39-46.
[19] L. Reyderman, S. Stavchansky, J. chromatogr. A. 1996, 755,
271-280.
[20] Z. Jahnz-Wechmann, G. R. Framski, P. A. Januszczyk, J.
Boryski, Front. Chem. 2016, 4, 19.
[21] G. Avigad, S. Damle, Anal. Chem. 1972, 50, 321-323.
[22] J. Nair, A. Gal, S. Tamir, S. Tannenbaum, G. Wogan, H.
Bartsch, Carcinogenesis 1998, 19, 2081-2084.
[23] N. K. Kochetkov, V. N. Shibaev, A. A. Kost, Tetrahedron Lett.
1971, 22, 1993-1996.
411 nm.
Quantitative of miRNA-21
Sample solution (100 μL) was added to DNA probe-magnetic bead
solution (200 μL) in a vial. The mixture was maintained at 50 °C for
30 min to ensure a complete hybridization.[51] The mixture was
magnetically separated and the decanted supernatant. The
DNA/RNA-magnetic bead complex was washed twice with binding
buffer and then denature at 90 ℃ for 3 min. After a magnetic
separation, the supernatant containing miRNA-21 was transferred to
a new vial. Then, 100 μL of 1×PBS, 20 μL of CAA solution were
7
This article is protected by copyright. All rights reserved.

10.1002/chem.201905639
Chemistry - A European Journal
[24] F. Oesch, G. Doerjer, Carcinogenesis, 1982, 3,663-665.
[25] E. R. Garett, P. J. Mehta, J. Am. Chem. Soc. 1972, 94, 8532-
8541.
[40] N. Xia, L. Zhang, G. Wang, Q. Feng, L. Liu, Biosens.
Bioelectron. 2013, 47, 461-466.
[41] E. M. Kroh, R. K. Parkin, P. S. Mitchell, M. Tewari, Methods,
2010, 50, 298-301.
[26] J. L. York, J. Org. Chem. 1981, 46, 2171-2173.
[27] J. L. York, J. Org. Chem. 1981, 46, 2171-2173.
[28] H. Åström, E. Limén, R. Strömberg, J. Am. Chem. Soc. 2004,
126, 14710-14711.
[42] C. Chen, D. A. Ridzon, A. J. Broomer, Z. Zhou, D. H. Lee, J. T.
Nguyen, M. Barbisin, N. L. Xu, V. R. Mahuvakar, M. R.
Andersen, K. Q. Lao, K. J. Livak, K. J. Guegler, Nucleic Acids
Res. 2005, 33, 179-185.
[29] P. H. Bolton, D.R. Kearns, Biochim. Biophys. Acta 1978, 517,
329-337.
[43] N. Redshaw, T. Wilkes, A. Whale, S. Cowen, J. Huggett, C. A.
Foy, Biotech. 2013, 54, 155-164.
[30] J. Pitha, Biochemistry 1970, 9, 3678-3682.
[31] P. Acharya, J. Chattopadhyaya, J. Org. Chem. 2002, 67, 1852-
1865.
[44] H. Xu, S. Zhang, C. Ouyang, Z. Wang, D. Wu, Y. Liu, Y. Jiang,
Z. Wu, Talanta, 2019, 192, 175-181.
[32] J. Biernat, J. Ciesiolka, P. Gornicki, R. W. Adamiak, W. J.
Krzyzosiak, M. Wiewiorowski, Nucleic Acids Res. 1978, 5,
789-804.
[45] H. Xu, Y. Zhang, S. Zhang, M. Sun, W. Li, Y. Jiang, Z. S. Wu,
Anal. Chim. Acta 2019, 1047, 172-178.
[46] S. J. Zhen, X. Xiao, C. H. Li, C. Z. Huang, Anal. Chem. 2017,
89, 8766-8771.
[33] W. J. Krzyzosiak, J. Biernat, J. Ciesiolka, P. Gornicki, M.
Wiewiorowski, Polish. J. Chem. 1983. 57, 779.
[34] A. A. Kost, M. V. Ivanov, Chem. Heterocyclic Comp. 1980, 16,
209-221.
[47] L. Qi, M. Xiao, X. Wang, C. Wang, L. Wang, S. Song, X. Qu,
L. Li, J. Shi, H. Pei, Anal. Chem. 2017, 89, 18, 9850-9856.
[48] S. Xu, Y. Nie, L. Jiang, J. Wang, G. Xu, W. Wang, X. Luo,
Anal. Chem. 2018, 90, 4039-4045.
[35] M. Liu, M. Xu, J. K. Lee, J. Org. Chem. 2008, 73, 5907-5914.
[36] R. Rios-Font, L. Rodríguez-Santiago, J. Bertran, M. Sodupe, J.
Phys. Chem. B 2007, 111, 6071-6077.
[49] X. Tang, R. Deng, Y. Sun, X. Ren, M. Zhou, J. Li, Anal. Chem.
2018, 90, 10001-10008.
[37] V. E. Marquez, C. K. H. Tseng, H. Mitsuya, S. Aoki, J. A.
Kelley, H. Ford, J. S. Roth, S. Broder, D. G. Johns and J. S.
Driscoll, J. Med. Chem. 1990, 33, 978-985.
[50] B. Levitt, R. J. Head, D. P. Westfall, Anal. Biochem. 1984, 137,
93-100.
[51] X. Qiu, X. Liu, W. Zhang, H. Zhang, T. Jiang, D. Fan, Y. Luo,
Anal. Chem. 2015, 87, 6303-6310.
[38] J. Wang, X. Yi, H. Tang, H. Han, M. Wu, F. Zhou, Anal. Chem.
2012, 84, 6400-6406.
[39] I. Lee, S. S. Ajay, H. Chen, A. Maruyama, N. Wang, M. G.
Mclnnis, B.D. Athey, Nucleic Acids Res. 2008, 36, 27.
8
This article is protected by copyright. All rights reserved.

10.1002/chem.201905639
Chemistry - A European Journal
Graphical Material
A Novel One Pot Fluorescence
Derivatization Strategy for Highly
Sensitive MicroRNA Analysis
Li Pan, Huaisheng Zhang, Jingjin Zhao,
Ifedayo Victor Ogungbe, Shulin Zhao,* and
Yi-Ming Liu*_______ Page – Page
A one pot fluorescence derivatization strategy was developed for highly sensitive miRNA
analysis. Based on this derivatization strategy, a facile and sensitive high-performance
liquid chromatography (HPLC) method was developed for quantitative assay of
microRNAs.
9
This article is protected by copyright. All rights reserved.