Photoacoustic Probes for Ratiometric Imaging of Copper(II)

Article abstract of DOI:10.1021/jacs.5b10504

Photoacoustic tomography has emerged as a promising alternative to MRI and X-ray scans in the clinical setting due to its ability to afford high-resolution images at depths in the cm range. However, its utility has not been established in the basic research arena owing to a lack of analyte-specific photoacoustic probes. To this end, we have developed acoustogenic probes for copper(II)-1 and -2 (APC-1 and APC-2, a water-soluble congener) for the chemoselective visualization of Cu(II), a metal ion which plays a crucial role in chronic neurological disorders such as Alzheimer's disease. To detect Cu(II), we have equipped both APCs with a 2-picolinic ester sensing module that is readily hydrolyzed in the presence of Cu(II) but not by other divalent metal ions. Additionally, we designed APC-1 and APC-2 explicitly for ratiometric photoacoustic imaging by using an aza-BODIPY dye scaffold exhibiting two spectrally resolved NIR absorbance bands which correspond to the 2-picolinic ester capped and uncapped phenoxide forms. The normalized ratiometric turn-on responses for APC-1 and APC-2 were 89- and 101-fold, respectively.

Full text of DOI:10.1021/jacs.5b10504

Communication
pubs.acs.org/JACS
Photoacoustic Probes for Ratiometric Imaging of Copper(II)
Hao Li,† Pamela Zhang,† Lukas P. Smaga, Ryan A. Hoffman, and Jefferson Chan*
Roger Adams Laboratory, Department of Chemistry, University of Illinois, 600 South Mathews Avenue, Urbana, Illinois 61801,
United States
S
* Supporting Information
probes that can be activated by a molecular target to elicit a
change in the signal intensity.7−10 Such chemical tools are more
ABSTRACT: Photoacoustic tomography has emerged as
a promising alternative to MRI and X-ray scans in the
clinical setting due to its ability to afford high-resolution
images at depths in the cm range. However, its utility has
not been established in the basic research arena owing to a
lack of analyte-specific photoacoustic probes. To this end,
we have developed acoustogenic probes for copper(II)-1
and -2 (APC-1 and APC-2, a water-soluble congener) for
the chemoselective visualization of Cu(II), a metal ion
which plays a crucial role in chronic neurological disorders
such as Alzheimer’s disease. To detect Cu(II), we have
equipped both APCs with a 2-picolinic ester sensing
module that is readily hydrolyzed in the presence of
Cu(II) but not by other divalent metal ions. Additionally,
we designed APC-1 and APC-2 explicitly for ratiometric
photoacoustic imaging by using an aza-BODIPY dye
scaffold exhibiting two spectrally resolved NIR absorbance
bands which correspond to the 2-picolinic ester capped
and uncapped phenoxide forms. The normalized ratio-
metric turn-on responses for APC-1 and APC-2 were 89-
and 101-fold, respectively.
versatile than contrast agents because they offer the unique
opportunity to interrogate the complex biological role of a
given analyte in its native environment with minimal
perturbation.11 To address the lack of available probes, we
present the development of two PA probes, acoustogenic probe
for copper(II)-1 and -2 (APC-1 and APC-2), for the
chemoselective imaging of Cu(II). In this study, we chose to
detect Cu(II) because misregulation of this important transition
metal ion has been implicated in numerous diseases, including
Alzheimer’s disease,12 an area studied by our group.13 The term
“acoustogenic” is used by us to designate any small-molecule-
based probe that reacts with a specific molecular target to afford
a “turn-on” PA response. APCs are comprised of three
components: a Cu(II)-responsive 2-picolinic ester moiety that
chelates Cu(II) which activates the ester bond for hydrolysis
(Figure 1a, in blue),14,15 sulfated ethylene glycol functional
hotoacoustic tomography (PAT) is a state-of-the-art
Pimaging modality that is based on the detection of
ultrasonic waves generated by thermoelastic expansion resulting
from the absorption of light and subsequent non-radiative
decay.1,2 In contrast to optical methods such as fluorescence
microscopy, in which the typical imaging depth is in the micron
regime, PAT can produce high-resolution images in the cm
range because sound scatters 3 orders of magnitude less than
fluorescence emission in many biological tissues. The
anticipated spatial resolution is ∼1/200 of the imaging depth,
meaning that at 7 cm, the attainable resolution is ∼350 μm,
whereas at 1 cm, the resolution will be in the 50 μm range.1
Moreover, because PAT does not involve ionizing radiation, it
is potentially less hazardous compared to computed tomog-
raphy (CT), positron emission tomography (PET), and X-ray
scans for medical diagnostic applications. Indeed, PAT has been
applied non-invasively to visualize and study various types of
cancer,3 cardiovascular diseases,4 and abnormalities of micro-
circulation.5 To distinguish between physiological and patho-
logical states, many photoacoustic (PA) contrast agents based
on small-molecule dyes, metal-based nanoparticles and carbon
nanotubes have been developed.6 Signal enhancement in a
specific cell- or tissue-type results from the accumulation of an
imaging agent via passive diffusion or receptor-mediated uptake
pathways. In contrast, there have been several reports of PA
Figure 1. (a) Chemical structures of APC-1 and APC-2. (b)
Normalized absorbance spectra of 2 μM APC-1 (dashed line) and 2
μM 6 (solid line), its hydrolyzed product, in PBS + 0.1% CrEL (pH
7.4).
groups to enhance water solubility (Figure 1a, in red), and a
near-infrared absorbing aza-BODIPY dye platform to generate
the PA signal (Figure 1a, in black). Of note, maximum
absorption in the NIR range (650−900 nm), a large extinction
coefficient (>104 M−1 cm−1) and high photostability are
essential properties we considered when developing the APCs.
The aza-BODIPY dye family has been used to construct an
assortment of fluorescent probes for reactive sulfur species such
as cysteine16 and hydrogen sulfide,17 reactive nitrogen species
including nitric oxide,17 nitroxyl,18 and nitrite,19 as well as
carbohydrates like glucose,20 but its application for photo-
Received: October 7, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.5b10504
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
acoustic imaging has never been reported. In this regard, we
selected an aza-BODIPY dye exhibiting dual-wavelength
absorbance bands, one below 700 nm and the other above,
which correspond to the capped 2-picolinic ester probe (Figure
1b, dotted line) and the uncapped phenoxide product (Figure
1b, solid line), respectively.21,22 Ratiometric imaging is
extremely powerful because it can control for artifacts such as
non-uniform dye accumulation and photobleaching. Our
strategy involves irradiating a probe, for example APC-1, at
both the blue-shifted (680 nm) and red-shifted (755 nm)
absorbance maxima to generate two corresponding PA signals
from which a ratio can be determined (Figure 2). In the
TBS-protected tetraarylazadipyrromethene 17 underwent facile
boron chelation which was followed by desilylation using TBAF
to afford aza-BODIPY 18 in 64% yield. Sulfation of 18 with
SO3−NMe3 in DMF gave the PEG-3 sulfate intermediate 19,
which was subjected to standard Tsuji−Trost deallylation
conditions to give 20 in 74% yield over two steps. Lastly, as
before, the 2-picolinic ester was installed using EDC in water to
afford APC-2 in 53% yield (Scheme 1).
Scheme 1. Synthesis of Acoustogenic Probe for Copper(II)-2
(APC-2)
Figure 2. Proposed mechanism for ratiometric PA imaging of Cu(II).
APC-1 (left) is excited at 680 nm (blue) and 755 nm (red) to afford
two PA signals. In the presence of Cu(II), the 2-picolinic ester moiety
is removed to yield 6 (right), and this results in a weaker PA signal at
680 nm and a stronger PA signal at 755 nm.
absence of Cu(II), the 2-picolinic ester moiety remains intact
and thus, the absorbance at 680 nm will be larger than the
signal at 755 nm, resulting in a small 755/680 PA ratio (Figure
2). In contrast, when APC-1 is treated with Cu(II), the
hydrolyzed product 6 will exhibit a decrease in the 680 nm
band with a concomitant increase in the 755 nm absorbance
maximum leading to a large 755/680 PA ratio (Figure 2).
Normalization of the two PA ratios will result in the ratiometric
turn-on response. In order for this approach to be successfully
implemented, the product must be fully deprotonated at
physiological pH to ensure that the phenoxide absorbance band
is maximized after the ester is cleaved. Thus, we tuned the
acidity of the phenol moiety by installing two ortho-chloro
substituents. We measured an apparent pKa value of 4.35,
which is consistent with the pKa for a related aza-BODIPY pH
probe recently reported (Figure S1).21
First, we synthesized APC-1 via an EDC-mediated coupling
reaction between 2-picolinic acid and the dichlorophenol
precursor 6 in 53% yield (Scheme S1). However, because
APC-1 is extremely hydrophobic, organic co-solvents or
surfactants were required to achieve sufficient solubility to
avoid dye aggregation in aqueous buffers which abolishes the
dual-wavelength character requisite for ratiometric imaging. As
such, we sought to develop a water-soluble analogue, first by
employing previously reported strategies for enhancing the
hydrophilicity of aza-BODIPY dyes such as PEGylation23 and
incorporation of amino sulfonic acid moieties.24 Unfortunately,
these approaches were insufficient to adequately solubilize our
probe. Instead, we found that three PEG-3 sulfate groups
endowed excellent water solubility. As such, APC-2 was
synthesized beginning from the dimerization of 4-nitro-1,3-
diphenylbutan-1-one precursors 12 and 15 to afford the
tetraarylazadipyrromethene intermediate 16. To facilitate
purification of 16 from homodimers of 12 and 15, it was
necessary to convert the primary alcohols to the corresponding
tert-butyl dimethylsilyl ether groups using TBS-Cl. The fully
In order to avoid inconsistencies when using different
solvents during the spectroscopic characterization and evalua-
tion of each probe, we performed all subsequent assays in PBS
with 0.1% CrEL, a potent surfactant used as a vehicle to
solubilize poorly water-soluble organic compounds. The
absorbance maximum of APC-2 was centered at 697 nm (ε
= 3.6 × 104 M−1 cm−1), whereas the corresponding maximum
of the phenoxide product 20 was 767 nm (3.7 × 104 M−1 cm−1)
(Table S1). We were pleased to discover that both compounds
were non-fluorescent in PBS, meaning there would be a greater
extent of non-radiative decay correlating to an overall stronger
PA signal.25 However, both APC probes were fluorescent in 2:3
v/v EtOH:PBS, which is consistent with reports from other
studies.21 When APC-2 was irradiated at its absorbance
maximum for 1 h (134 mW/cm2), we observed only ∼15%
loss of fluorescence, demonstrating that APC-2 is exceptionally
photostable (Figure S5). Next, we sought to measure the
acoustic emission of APC-2 as a function of pH when exciting
at both the blue- and red-shifted absorbance maxima (Figure
3a,c). When APC-2 was irradiated at 697 nm, the PA signal was
pH insensitive indicating that the 2-picolinic ester trigger is
stable across the pH range examined (Figure 3a). Likewise,
excitation at 767 nm yielded no observable PA output from pH
3.0 to 8.2, owing to a low extinction coefficient (<2000 M−1
cm−1) at this wavelength for the capped probe (Figure 3c). In
contrast, when 20 was evaluated at 697 nm, a pH-dependent
signal intensity change was noted. In particular, the strength of
the PA signal was more prominent in acidic buffers compared
to at physiological pH (Figure 3a). Likewise, we also observed a
B
DOI: 10.1021/jacs.5b10504
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Figure 3. Photoacoustic imaging of APC-2 and 20 as a function of pH.
(a) Normalized pH profile of APC-2 (red circles) and its hydrolyzed
product 20 (black circles), with excitation at 697 nm. (b)
Photoacoustic images of 10 μM compound 20 (PBS + 0.1% CrEL,
pH 7.4) in FEP tubing overlaid with a 1 cm thick phantom acquired at
pH 3.0 and 8.2, with excitation at 697 nm. (c) Normalized pH profile
of APC-2 (red circles) and its hydrolyzed product 20 (black circles),
with excitation at 767 nm. (d) Photoacoustic images of 10 μM 20
(PBS + 0.1% CrEL, pH 7.4) in FEP tubing overlaid with a 1 cm thick
phantom acquired at pH 3.0 and 8.2, with excitation at 767 nm.
Pseudocoloring represents intensity distribution from highest
intensity, indicated by white, to the lowest intensity, designated by
black. Scale bar represents 2 mm.
Figure 4. Photoacoustic imaging of APC-2 treated with Cu(II).
Photoacoustic images of 10 μM APC-2 (PBS + 0.1% CrEL, pH 7.4) in
FEP tubing overlaid with a 1 cm thick phantom treated with 0 and 10
equiv of Cu(II), with excitation at (a) 697 and (b) 767 nm. Pseudo-
coloring represents intensity distribution from highest intensity,
indicated by white, to the lowest intensity, designated by black.
Scale bar represents 2 mm. (c) Quantification of APC-2 treated with 0,
1, and 10 equiv of Cu(II) for 90 min. Blue bars represent normalized
ratiometric turn-on responses calculated from the PA ratios (PA767/
PA697) for control and Cu(II)-treated conditions. n = 2. (d) PA
responses of APC-2 to various metal ions (2 mM for Ca(II) and
Mg(II), 50 μM for all other cations).
pH-dependent change in the PA signal with 767 nm excitation.
Specifically, at low pHs the PA signal is suppressed, whereas at
high pHs the acoustic strength becomes larger (Figure 3c,d).
Moreover, we were able to calculate an apparent pKa value of
∼4.5 for the phenol moiety which is consistent with the value
obtained from the corresponding absorbance spectrum (Figure
3a,c).
We then turned our attention to evaluating the response of
APC-2 to Cu(II) supplementation using UV−vis spectroscopy.
Upon addition of Cu(II), the major visible absorbance band
shifted from 697 to 767 nm, indicating facile conversion from
APC-2 to 20 with a rate constant of 2.66 × 10−2 min−1.
Employing PA imaging, we observed a slight decrease in the PA
output at 697 nm when Cu(II) was added (Figure 4a). On the
other hand, at 767 nm, a dramatic turn-on response was noted
with the Cu(II)-mediated hydrolysis of the 2-picolinic ester
moiety (Figure 4b). Normalization of the PA ratios resulted in
a ratiometric PA turn-on of 91.3- and 100.5-fold for 1 and 10
equiv of Cu(II), respectively (Figure 4c). Having established
excellent responsiveness to Cu(II), we then evaluated the metal
ion selectivity by treating solutions of APC-2 with various
biologically relevant metal ions. The PA response of APC-2 is
not affected by the presence of physiologically relevant
concentrations of alkaline earth metals such as Mg(II) and
Ca(II) or divalent transition metal ions like Mn(II), Fe(II),
Co(II), Ni(II), and Zn(II). Moreover, APC-2 is selective for
Cu(II) over Cu(I), showing that this probe has excellent metal
and redox specificity (Figure 4d). Of note, we performed
competition assays to demonstrate that the presence of the
above divalent metal ions do not perturb the reactivity of APC-
2 with Cu(II) (Figure S9). We also determined that APC-2
responds to Cu(II) complexes of aspartic acid, bilirubin, and
clioquinol (Figure S10) and that it is stable in human blood
plasma (Figure S11).
To evaluate the PA properties of APC-2 in the cm range, we
developed an agarose-based tissue phantom consisting of milk
and water to mimic the light scattering and absorption
properties of an authentic tissue sample. Of note, all PA
images described in this study were acquired in phantoms cut
to a thickness of 1 cm. While the PA signal at this imaging
depth remained intense and was clearly visible, the correspond-
ing fluorescence signal was completely abolished when APC-2
was overlaid with a 1 cm thick phantom (Figure S13). Lastly,
we imaged APC-2 in chicken breast tissue cut to the same
dimensions to assess whether the PA signal could be detected
in tissue containing lipids, proteins, and blood.26 Gratifyingly,
the PA signal of APC-2 was comparable to what we observed
using the agarose-based phantom (Figure S14).
In closing, we have developed the first small-molecule-based
probe exclusively designed for ratiometric photoacoustic
C
DOI: 10.1021/jacs.5b10504
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(6) Wu, D.; Huang, L.; Jiang, M. S.; Jiang, H. B. Int. J. Mol. Sci. 2014,
15, 23616.
imaging. APC-2 displays an absorbance maximum in the NIR
window, a large extinction coefficient (>104 M−1 cm−1), and
was shown to be extremely photostable. The unique dual-
wavelength absorbance profile and low apparent pKa of the
dichlorophenol group produces a large normalized turn-on
response of 100.5-fold in the presence of Cu(II). Moreover, we
demonstrated that APC-2 is highly selective for Cu(II) over
other biologically relevant metal ions. We also highlight the
strength of PA imaging by performing imaging experiments in
the cm range using a tissue phantom and authentic tissue.
Because this imaging depth is significantly greater than what is
possible with optical methods, we anticipate that APC-2 will
open many new opportunities to study the complex
biochemistry of Cu(II) in vivo. Indeed, discovering and
understanding the influence of an analyte on disease
progression is the crucial first step to developing new potential
treatment options. On this note, the highly modular aza-
BOPIDY platform used in the current study can be easily
adapted to sense other biologically important analytes beyond
Cu(II) by simply installing alternative reactive triggers at the
last step of the synthesis. We anticipate that this initial report
will establish a new standard in the PA probe development
arena and thus, will be of broad interest to the molecular
imaging community. Such probes will be critical for the
continued growth of the PA imaging field in both the basic
research and clinical settings.
(7) Dragulescu-Andrasi, A.; Kothapalli, S. R.; Tikhomirov, G. A.; Rao,
J. H.; Gambhir, S. S. J. Am. Chem. Soc. 2013, 135, 11015.
(8) Pu, K. Y.; Shuhendler, A. J.; Jokerst, J. V.; Mei, J. G.; Gambhir, S.
S.; Bao, Z. N.; Rao, J. H. Nat. Nanotechnol. 2014, 9, 233.
(9) Cash, K. J.; Li, C. Y.; Xia, J.; Wang, L. H. V.; Clark, H. A. ACS
Nano 2015, 9, 1692.
(10) Levi, J.; Kothapalli, S. R.; Ma, T. J.; Hartman, K.; Khuri-Yakub,
B. T.; Gambhir, S. S. J. Am. Chem. Soc. 2010, 132, 11264.
(11) Chan, J.; Dodani, S. C.; Chang, C. J. Nat. Chem. 2012, 4, 973.
(12) Manto, M. Toxics 2014, 2, 327.
(13) Roth, A.; Li, H.; Anorma, C.; Chan, J. J. Am. Chem. Soc. 2015,
137, 10890.
(14) Fife, T. H.; Przystas, T. J. J. Am. Chem. Soc. 1985, 107, 1041.
(15) Kierat, R. M.; Kramer, R. Bioorg. Med. Chem. Lett. 2005, 15,
4824.
(16) Jiang, X. D.; Zhang, J.; Shao, X. M.; Zhao, W. L. Org. Biomol.
Chem. 2012, 10, 1966.
(17) Adarsh, N.; Krishnan, M. S.; Ramaiah, D. Anal. Chem. 2014, 86,
9335.
(18) Liu, P.; Jing, X. T.; Yu, F. B.; Lv, C. J.; Chen, L. X. Analyst 2015,
140, 4576.
(19) Adarsh, N.; Shanmugasundaram, M.; Ramaiah, D. Anal. Chem.
2013, 85, 10008.
(20) Liu, Y. L.; Zhu, J. W.; Xu, Y. M.; Qin, Y.; Jiang, D. C. ACS Appl.
Mater. Interfaces 2015, 7, 11141.
(21) Strobl, M.; Rappitsch, T.; Borisov, S. M.; Mayr, T.; Klimant, I.
Analyst 2015, 140, 7150.
(22) Jokic, T.; Borisov, S. M.; Saf, R.; Nielsen, D. A.; Kuhl, M.;
Klimant, I. Anal. Chem. 2012, 84, 6723.
(23) Collado, D.; Vida, Y.; Najera, F.; Perez-Inestrosa, E. RSC Adv.
2014, 4, 2306.
(24) Kamkaew, A.; Burgess, K. Chem. Commun. 2015, 51, 10664.
(25) Kim, C.; Favazza, C.; Wang, L. H. V. Chem. Rev. 2010, 110,
2756.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.5b10504.
Experimental details, including synthesis of APC-1 and
APC-2, and all data for in vitro characterization and
imaging experiments (PDF)
(26) Ku, G.; Wang, L. H. V. Opt. Lett. 2005, 30, 507.
AUTHOR INFORMATION
■
Corresponding Author
*jeffchan@illinois.edu
Author Contributions
†H.L. and P.Z. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. Michael Thornton (Endra Life Sciences) for
helpful scientific discussions and assistance with imaging
experiments. We thank Ms. Chelsea Anorma for assisting
with the fluorescence imaging experiments. P.Z. acknowledges
the United States Air Force for a graduate fellowship. R.A.H.
thanks the Researchers Initiative Program for research
experience for undergraduates (REU) funding.
REFERENCES
■
(1) Wang, L. H. V.; Hu, S. Science 2012, 335, 1458.
(2) Xia, J.; Yao, J. J.; Wang, L. V. Prog. Electromagn. Res. 2014, 147, 1.
(3) Mallidi, S.; Luke, G. P.; Emelianov, S. Trends Biotechnol. 2011, 29,
213.
(4) Zemp, R. J.; Song, L.; Bitton, R.; Shung, K. K.; Wang, L. V. Opt.
Express 2008, 16, 18551.
(5) Luke, G. P.; Yeager, D.; Emelianov, S. Y. Ann. Biomed. Eng. 2012,
40, 422.
D
DOI: 10.1021/jacs.5b10504
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX