An adhesive 19F MRI chemical probe allows signal off-to-on-type molecular sensing in a biological environment

Article abstract of DOI:10.1039/c3cc46471g

We report a new strategy for designing a signal off-to-on-type ¹⁹F MRI chemical probe that operates in biological environments. The present strategy is based on the control of adherence of a ¹⁹F MRI chemical probe to certain blood proteins, accompanied by a change in transverse relaxation time of ¹⁹F nuclei.

Full text of DOI:10.1039/c3cc46471g

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
1
9
An adhesive F MRI chemical probe allows signal
off-to-on-type molecular sensing in a biological
environment†
Cite this: Chem. Commun., 2013,
4
9, 11421
Received 24th August 2013,
Accepted 14th October 2013
a
a
a
b
Tomohiro Doura,z Ryunosuke Hata,z Hiroshi Nonaka, Fuminori Sugihara,
b
a
Yoshichika Yoshioka and Shinsuke Sando*
DOI: 10.1039/c3cc46471g
www.rsc.org/chemcomm
We report a new strategy for designing a signal off-to-on-type to which small compounds, especially hydrophobic compounds
1
9
F MRI chemical probe that operates in biological environments. including hormones, fatty acids, and synthetic chemical probes,
The present strategy is based on the control of adherence of a tend to adhere. The adherent small molecules have a shorter
19
F MRI chemical probe to certain blood proteins, accompanied by
2
T value because of the lower mobility of the complex. Therefore,
19
19
a change in transverse relaxation time of F nuclei.
we hypothesized that if the hydrophobic F chemical probe
adheres to blood components such as serum albumin (short T ),
Magnetic resonance imaging (MRI) is a powerful technique for and produces a less adhesive hydrophilic F product after the
2
1
9
19
2
in situ biomolecular analysis. Recent work has focused on F MRI targeted event (long T ), such a chemical probe may function as a
because of its relatively high sensitivity (83% of H) and the absence turn-on F chemical probe in T -weighted F MRI (Fig. 1a).
of background signals. To date, various F MRI chemical probes
for sensing biomolecules have been reported.
In particular, current attention has been focused on signal
1
19
19
2
1
19
1–12
1
9
19
activatable F MRI probes, which turn on F MRI signals only
when targeted biological events occur. The off-to-on-type
1
9
F MRI signaling allows us to detect the targeted event much
more easily and more precisely.
There are two elegant ways of controlling the signal; the switch-
4
ing of paramagnetic relaxation or precise control of supramolecular
5
19
self-assembly. Both utilize a change in F transverse relaxation
1
9
times (T
2
). These F chemical probes initially have short
1
9
F T
The T
having a relatively longer T
2
values, which are elongated after the targeted events.
1
9
19
2
-weighted F MRI gives a stronger signal for F nuclei
value. Therefore, in the T -weighted
2
2
1
9
19
F MRI, such chemical probes operate to turn on the F probe
after a targeted biological event. Herein, we report a new option
for designing the signal off-to-on-type F MRI chemical probes.
The present approach is based on the control of adherence of
the chemical probes to endogenous biomaterials. The selected
environment in this research is blood. Blood contains abundant
serum albumin (ca. 50% of blood proteins, typically 35–50 g L
1
9
1
2
ꢀ1
)
a
INAMORI Frontier Research Center, Kyushu University, 744 Motooka, Nishi-ku,
Fukuoka, Japan. E-mail: ssando@ifrc.kyushu-u.ac.jp; Fax: +81 92 802 6961;
Tel: +81 92 802 6960
b
Immunology Frontier Research Center, Osaka University, 3-1 Yamadaoka, Suita,
Osaka, Japan
Fig. 1 (a) The proposed mechanism for signal off-to-on sensing of a targeted
biological event in blood. (b) Proposed scheme for production of a fluoroalcohol
†
Electronic supplementary information (ESI) available: Experimental procedures
(product) from fluoroalkoxyaniline (probe) via an ipso-substitution mechanism.
19
19
19
of synthesis, F NMR measurements and F MRI measurements. See DOI:
2
(c) Relationship between log P and F T of compounds 2–5 dissolved in 100 mM
ꢀ
1
1
‡
0.1039/c3cc46471g
sodium phosphate buffer (pH 7.4) containing 150 mM NaCl and HSA (50 g L ).
These authors contributed equally to this work.
Log P values are taken from LOGKOW database.
This journal is c The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 11421--11423 11421

View Article Online
Communication
Along this concept, we designed F chemical probes, shown Table 1
ChemComm
1
9
19
2
T values of F nuclei of 1 and trifluoroethanol 2 in phosphate buffer,
in phosphate buffer containing HSA, or in blood plasma
in Fig. 1b, for detection of hypochlorous acid (HOCl) that is an
important biomarker for inflammatory diseases. We have found
that the fluoroalkoxyaniline scaffold (probe) can react with HOCl
a
b
ꢀ1
Compounds
Conditions
2
T ms
1
1
1
Phosphate buffer
Phosphate buffer, HSA
Blood plasma
Phosphate buffer
Phosphate buffer, HSA
Blood plasma
873.6 ꢁ 184.9
14.7 ꢁ 1.6
expeditiously to produce a relatively hydrophilic fluoroalcohol
8
(
product) via an ipso-substitution mechanism (Fig. 1b). This
15.0 ꢁ 3.2
Trifluoroethanol (2)
Trifluoroethanol (2)
Trifluoroethanol (2)
1744.6 ꢁ 324.0
293.8 ꢁ 18.3
216.0 ꢁ 2.9
hydrophobic-to-hydrophilic conversion is thought to be suitable
for the present concept.
19
Initially, we explored an appropriate F alcohol (product) that
is sufficiently hydrophilic and thus does not suffer from substan-
a
Concentrations of compounds = 5 mM, phosphate buffer; 100 mM
sodium phosphate (pH 7.4) containing 150 mM NaCl with or without
19
19
ꢀ1
b
tial shortening of F T in blood. We measured F T values of HSA (50 g L ). T2 values were measured using a spin-echo method.
2
2
19
four types of F alcohols, 2–5, in the presence of human serum
ꢀ1
19
albumin (HSA, 50 g L ) (Fig. 1c). The observed F T values are
2
in proportion to log P values for each compound, suggesting compared with that in the absence of HSA, while as much as
that lipophilicity is an important factor for determining the 94% was recovered in the case of trifluoroethanol.
1
9
13
19
F T
2
values in blood. Based on the result, we selected trifluoro-
CH
F product for this purpose, because compound 2 has relatively plasma. The F T
Table 1 summarizes the T
2
values of F nuclei of 1 (probe) and
ethanol CF
3
2
OH (compound 2 in Fig. 1c) as an appropriate trifluoroethanol (product) in phosphate buffer and in blood
value for 1 in blood plasma (15.0 ms) was
long F T (293.8 ms) in the presence of HSA and, in a practical much shorter than that in phosphate buffer (873.6 ms). In addition,
19
19
2
1
9
2
19
sense, such a simple chemical structure makes it easier to design the F T2 value for 1 was also shortened to the same range
19
and synthesize F chemical probes.
by adding HSA to phosphate buffer (14.7 ms). By contrast, the
F T value for trifluoroethanol was 216.0 ms in blood plasma and
2
19
Based on the above discussion, compound 1, having a
3 2
CF CH O-leaving moiety, was synthesized as a candidate probe, 293.8 ms in phosphate buffer with HSA, which are shorter than the
which produces trifluoroethanol (compound 2) upon reaction value in phosphate buffer (1744.6 ms), but still >10 times longer
19
with HOCl (Fig. 2a).
than the F T
2
value for 1 in blood plasma.
Binding of 1 and nonbinding of trifluoroethanol to HSA were
From these results, it is feasible that 1 tends to adhere to
demonstrated by an ultrafiltration assay (Fig. S1, ESI†). Only 29% blood plasma proteins, resulting in the remarkable shortening
1
9
of 1 was recovered in the presence of HSA (1.9 equivalent) of F T , whereas the product trifluoroethanol is less adherent
2
1
9
and has a relatively long F T
2
value in blood, suggesting that
probe 1 has appropriate chemical properties for off-to-on
1
9
F MRI sensing.
In fact, probe 1 operates as a signal on-type F NMR/MRI
1
9
1
9
agent. We measured F NMR using a long echo time (TE).
1
9
Probe 1 (5 mM) showed a sharp F NMR signal at ca. ꢀ74.0 ppm
in a phosphate buffer (top-left in Fig. 2b). The signal was highly
attenuated in the presence of HSA (second from the top in the
left column, Fig. 2b) in a HSA concentration-dependent manner
(
T
Fig. S2, ESI†). The signal attenuation reflects the shortening of
2
of F nuclei because of binding to HSA (the same signal
1
9
attenuation of 1 was observed also in blood plasma, data not
1
9
shown). On the other hand, after addition of HOCl, the F signal
at ꢀ76.6 ppm, corresponding to the product trifluoroethanol,
appeared in a HOCl concentration-dependent manner (from top
to bottom in the right column, Fig. 2b).
19
Similar results were observed for T
2
-weighted F MRI (Fig. 2b).
To analyze signal off-to-on behavior of the probe and the product
19
more precisely, chemical-shift-selective F imaging was applied,
where the probe and the product were imaged separately according
1
9
to each chemical shift. Without HOCl, no F MRI signals for 1
and the product (trifluoroethanol) were observed. By contrast,
with HOCl, clear contrast images were observed only for product-
1
9
selective F MRI. These results demonstrated that 1 operates as
1
9
a signal on-type F NMR/MRI agent responsive to HOCl.
Finally, we applied probe 1 in the detection of enzyme
myeloperoxidase (MPO), abundant in neutrophils. MPO is an
important factor in various cardiovascular and neurodegenera-
tive diseases and is also a biomarker for future acute coronary
1
9
Fig. 2 (a) Probe 1 and a proposed reaction scheme via ipso-substitution. (b) F NMR
19
(
long spin echo time, TE = 20 ms) and T
2
-weighted F MRI (chemical shift
selective) of probe 1 solution (5 mM) in 100 mM sodium phosphate buffer
ꢀ
1
(
pH 7.4) containing 150 mM NaCl in the presence or absence of HSA (50 g L ),
incubated with HOCl (0–2 eq.).
1
1422 Chem. Commun., 2013, 49, 11421--11423
This journal is c The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
e.g., simple binding or anchoring to cells and release, could
1
9
operate as a switch to modulate F signals. This approach is
based on a simple concept, and thus could be a strategy for the
1
9
design of a variety of signal off-to-on-type F MRI probes.
This work was supported by the Funding Program for Next
Generation World-Leading Researchers (NEXT), JSPS. RH thanks
JSPS for the fellowship.
1
9
Fig. 3
without 470 nM MPO in 60 mM sodium phosphate buffer (pH 7.4) containing
0 mM NaCl, 4 mM H and 40% blood plasma.
2
T -weighted F MRI images of probe 1 (5 mM) incubated with or
Notes and references
9
2 2
O
19
1
For reviews on F MRI probes, see for example ref. 1 and 2: J.-X. Yu,
V. D. Kodibagkar, W. Cui and R. P. Mason, Curr. Med. Chem., 2005,
1
2, 819.
disease. Accordingly, imaging of MPO activity in blood would
be useful for medical applications. It is known that MPO
2
3
S. L. Cobb and C. D. Murphy, J. Fluorine Chem., 2009, 130, 132.
For recent examples of F MRI probes, see 3–12: B. J. Stockman,
1
4
19
oxidizes the alkoxyaniline group directly or via the produced
J. Am. Chem. Soc., 2008, 130, 5870.
4 S. Mizukami, R. Takikawa, F. Sugihara, Y. Hori, H. Tochio,
M. W ¨a lchli, M. Shirakawa and K. Kikuchi, J. Am. Chem. Soc., 2008,
1
5,16
19
HOCl.
Therefore, probe 1 might also operate as a
F
chemical probe for MPO activity.
130, 794.
1
9
We used T
2
-weighted F MRI to visualize the MPO activity in
5 Y. Takaoka, T. Sakamoto, S. Tsukiji, M. Narazaki, T. Matsuda,
H. Tochio, M. Shirakawa and I. Hamachi, Nat. Chem., 2009, 1, 557.
K. Yamaguchi, R. Ueki, H. Nonaka, F. Sugihara, T. Matsuda and
S. Sando, J. Am. Chem. Soc., 2011, 133, 14208.
blood plasma (Fig. 3). Without MPO (dotted circle, bottom-
6
1
9
right), no F MRI signals of 1 (probe) and trifluoroethanol
product) were observed. By contrast, with MPO (dotted circle,
(
7 K. Tanaka, N. Kitamura and Y. Chujo, Bioconjugate Chem., 2011,
2, 1484.
T. Doura, A. Qi, F. Sugihara, T. Matsuda and S. Sando, Chem. Lett.,
011, 1357.
2
top-left), a detectable signal appeared in the product-selective
imaging, while no signal for 1 was detected. Although the MPO
concentrations used are much higher than that of the physio-
8
2
9 J.-X. Yu, V. D. Kodibagkar, R. R. Hallac, L. Liu and R. P. Mason,
Bioconjugate Chem., 2012, 23, 596.
0 P. Harvey, K. H. Chalmers, E. D. Luca, A. Mishra and D. Parker,
Chem.–Eur. J., 2012, 18, 8748.
11 X. Huang, G. Huang, S. Zhang, K. Sagiyama, O. Togao, X. Ma,
Y. Wang, Y. Li, T. C. Soesbe, B. D. Sumer, M. Takahashi,
1
7
logical MPO level and the sensitivity needs to be improved
1
further, these results demonstrated that probe 1 operates as a
1
9
turn-on F MRI probe for sensing MPO activity in blood.
In conclusion, we showed a proof-of-concept study for
A. D. Sherry and J. Gao, Angew. Chem., Int. Ed., 2013, 52, 8074.
1
9
designing a signal off-to-on-type F MRI probe that operates
19
1
2 Recent attempt at F signal control by molecular recognition, see:
Y. Sun, Y. Takaoka, S. Tsukiji, M. Narazaki, T. Matsuda and
I. Hamachi, Bioorg. Med. Chem. Lett., 2011, 21, 4393.
3 G. Colmenarejo, A. Alvarez-Pedraglio and J.-L. Lavandera, J. Med.
Chem., 2001, 44, 4370.
1
9
in blood. This strategy allows us to develop a turn-on F MRI
1
9
probe, responding to HOCl and MPO, in T
2
-weighted F MRI.
1
1
9
The key concept of this off-to-on F MRI is control of the
adhesive and nonadhesive properties of the F probe and the 14 For a recent review, see for example: J. A. Ronald, Curr. Cardiovasc.
1
9
Imaging Rep., 2011, 4, 24.
5 K.-I. Setsukinai, Y. Urano, K. Kakinuma, H. J. Majima and
T. Nagano, J. Biol. Chem., 2003, 278, 3170.
product to endogenous biomaterials. In blood, hydrophobic-to-
hydrophilic conversion, i.e., adhesive and nonadhesive proper-
1
ties to HSA, was demonstrated to operate as a switch. The 16 J. Shepherd, S. A. Hilderbrand, P. Waterman, J. W. Heinecke,
R. Weissleder and P. Libby, Chem. Biol., 2007, 14, 1221.
adhesive probe keeps the signal off state in blood and turns on
its F signal after targeted events. This concept is not limited
to this case. More generally, other basic biochemical events,
1
7 MPO under physiological conditions, see: S. Baldus, C. Heeschen,
T. Meinertz, A. M. Zeiher, J. P. Eiserich, T. M u¨ nzel, M. L. Simoons
and C. W. Hamm, Circulation, 2003, 108, 1440.
1
9
This journal is c The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 11421--11423 11423