Detection of Sulfatase Enzyme Activity with a CatalyCEST MRI Contrast Agent

Article abstract of DOI:10.1002/chem.201600685

A chemical exchange saturation transfer (CEST) MRI contrast agent has been developed that detects sulfatase enzyme activity. The agent produces a CEST signal at δ=5.0 ppm before enzyme activity, and a second CEST signal appears at δ=9.0 ppm after the enzyme cleaves a sulfate group from the agent. The comparison of the two signals improved detection of sulfatase activity.

Full text of DOI:10.1002/chem.201600685

DOI: 10.1002/chem.201600685
Communication
&
Diagnostic Agents
Detection of Sulfatase Enzyme Activity with a CatalyCEST MRI
Contrast Agent
Sanhita Sinharay,[b] Gabriela Fernµndez-Cuervo,[c] Jasmine P. Acfalle,[b] and Mark D. Pagel*[a]
used to detect enzyme activity, which is known as catalyCEST
Abstract: A chemical exchange saturation transfer (CEST)
MRI.[7,8]
MRI contrast agent has been developed that detects sulfa-
We proposed to develop a catalyCEST MRI contrast agent
tase enzyme activity. The agent produces a CEST signal at
that can detect sulfatase enzyme activity (Figure 1a, left side).
d=5.0 ppm before enzyme activity, and a second CEST
We decided to use a diamagnetic CEST agent to avoid poten-
signal appears at d=9.0 ppm after the enzyme cleaves
tial toxicities that have been encountered with paramagnetic
a sulfate group from the agent. The comparison of the
MRI contrast agents that incorporate lanthanide metals.[9] In
two signals improved detection of sulfatase activity.
particular, salicylic acid has been shown to generate a CEST
signal at a remarkably high d= ~9 ppm MR frequency relative
to the frequency of water (also known as the chemical shift, in
Sulfatases are involved with various physiological conditions
including developmental abnormalities, hormone-dependent
cancer, and bacterial pathogenesis.[1] Although sulfatase activi-
ties have been studied in vitro, the biological roles of sulfatas-
es within in vivo animal models and patients are poorly under-
stood. Fluorescent contrast agents have been developed that
detect sulfatase activity,[2,3] but optical imaging methods suffer
from a limited depth of view and poor spatial resolution when
imaging in vivo tissues. Therefore, a new noninvasive imaging
method is needed to interrogate sulfatase enzyme activity in
vivo.
which the chemical shift of water is defined as d=0 ppm in
MRI studies).[10,11] We hypothesized that derivatizing the salicyl-
ic acid moiety with a sulfate group would eliminate the CEST
signal from the salicylic acid proton, and subsequent cleavage
of the sulfate group by sulfatase would cause the salicylic acid
moiety to appear and produce a CEST signal (Figure 1a, right
side). This “activation” of the CEST signal could then be used to
detect sulfatase enzyme activity. To improve detection
specificity, we included an aryl amide ligand that would be
unresponsive to sulfatase activity and thereby serve as an
internal control.
Magnetic resonance imaging (MRI) may potentially address
this unmet need by imaging soft tissues at high resolution
throughout the body.[4] Chemical exchange saturation transfer
(CEST) MRI is a relatively new method that generates image
contrast by selectively saturating the magnetization of
a proton on a CEST agent, then waiting for the proton to un-
dergo chemical exchange with water, which transfers the satu-
ration to water.[5,6] The chemical exchange rate of the proton
on a CEST agent must be slow to generate CEST contrast, or
else the great majority of the proton population leaves the
agent before it can be saturated. CEST agents have been
developed with protons that modulate chemical exchange rate
after enzyme catalysis, which causes the CEST signal of the
agent to appear or disappear. This change in CEST signal is
We
synthesized
4-acedamido-2-(sulfoxy)benzoic
acid
(Scheme S1, Supporting Information). The carboxylic acid and
hydroxyl groups in 4-amino salicylic acid (compound 1) were
functionalized to benzyl ester and benzyl alcohol to generate
compound 2.[12] Then acylation of the amino group generated
compound 3. Selective debenzylation of the alcohol moiety
with trifluoroacetic acid produced compound 4.[13] Chlorosulfu-
ric acid 2,2,2-trichloroethylester (TCE sulfate) was synthesized
using reported procedures, and was used as the sulfating
agent to generate compound 5.[14] A stepwise deprotection of
the TCE group with Zn dust generated compound 6, followed
by debenzylation using H2/Pd(OH)2,[15] to give the final
compound 7 in 71% overall yield. We refer to compound 7 as
the substrate for the remainder of this report.
A CEST spectrum of 45 mm of the substrate in 200 mL of Tris
buffer at pH 7.2 was acquired using a standard CEST MRI pro-
cedure (Figure 1b). This spectrum showed a single CEST signal
at d=5.0 ppm that was assigned to the amide proton, based
on the chemical shift of similar aryl amide groups.[16] Five units
of sulfatase enzyme from the H.pomatia mollusk (the snail in
escargot) were added to the sample and incubated for 24 h at
378C. The sample dropped to pH 5.4, indicating that a reaction
had occurred. The pH of the sample was adjusted to 7.2 to
simulate buffering capacity in vivo. A CEST spectrum was ac-
quired after enzyme reaction, and the product generated
a new CEST signal at d=9.0 ppm from the salicylic acid moiety
[a] M. D. Pagel
Department of Medical Imaging, University of Arizona
1515 N. Campbell Ave., Tucson, AZ (USA)
E-mail: mpagel@u.arizona.edu
[b] S. Sinharay, J. P. Acfalle
Department of Chemistry and Biochemistry, University of Arizona
Tucson, AZ (USA)
[c] G. Fernµndez-Cuervo
Department of Pharmaceutical Sciences, University of Arizona
Tucson, AZ (USA)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201600685.
Chem. Eur. J. 2016, 22, 6491 – 6495
6491
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Figure 1. CatalyCEST MRI of sulfatase activity. a) The aryl sulfate ligand of the substrate is cleaved by sulfatase, generating a salicylic acid moiety. The aryl
amide ligand is not changed by the enzyme. b) The CEST spectra of the substrate (blue dots and lines) and product (red dots and lines) shows the appearance
of a CEST signal at d=9.0 ppm after the enzyme was added to the substrate, while the CEST signal at d=5.0 ppm was largely unchanged. The results of Lor-
entzian line fitting of the CEST spectra are shown as thick lines, and represent the % CEST signal from the CEST spectra. c) MR images of the substrate and
product show normalized water signals when saturation was applied at d=9.0, 5.0, À5.0, or À9.0 ppm (images shown in blue-to-red scale, labeled with the
saturation frequency). These images were used to produce parametric maps of% CEST signal based on Eq. [1] (blue-to-green scale). The CEST maps were
then used to produce parametric maps of the reaction coordinate for the enzyme based on Eq. [2] (black-to-white scale). The CEST maps and reaction coordi-
nate maps are masked to show the center of the sample tubes, because the edges and exterior of the tube have MR signals that are too low for meaningful
analyses. Unmasked and masked images are shown in Figure S1 (Supporting Information) to provide assurance that the masking is a reasonable representa-
tion.
while the CEST signal at d=5.0 ppm from the amide ligand
was largely unchanged. These results were also represented by
images of the samples before and after incubation with the
enzyme, which were used to create parametric maps of %
CEST signal based on Eq. [1], and parametric maps of the
enzyme reaction coordinate based on Eq. [2] (Figure 1c;
Figure S1, Supporting Information).
strated that catalyCEST MRI with this new agent could detect
multiple isoforms of sulfatase.
Importantly, the ratio of the enzyme-responsive CEST signal
at 9.0 ppm and the unresponsive signal at 5.0 ppm can be
used to detect the sulfatase enzyme activity in a concentra-
tion-independent manner. In addition, we have shown in past
studies that this ratiometric approach eliminates complications
from variable T1 relaxation time and incomplete saturation
during the CEST MRI protocol, and can reduce the influence of
temperature to negligible levels.[17] Therefore, ratiometric cata-
lyCEST MRI can detect sulfatase activity with this new CEST
agent with good specificity for enzyme activity relative to
other effects that can influence MRI contrast.
M
ÀM
À
M
þ
ð1Þ
% CEST ¼
À
M
and M+ are the water signal amplitudes with saturation
À
applied at at a negative and positive chemical shift value,
respectively (e.g, d=À9 and +9 ppm).
Using the substrate and the product generated with sulfa-
tase from H. pomatia, we studied the effects of MRI saturation
conditions on the CEST signal to determine clinical translatabili-
ty of this approach. We used a RL-QUEST analysis method[18] to
evaluate the effect of saturation time, which showed that a sat-
uration for four seconds was sufficient to generate strong CEST
signals from the substrate and product. (Figure 2a, Figure S3a,
Supporting Information). We then used a HW-QUESP[19] analysis
method to evaluate the effect of saturation power (Figure 2b,
Figure S3b, Supporting Information). The amide for both the
substrate and product could generate strong CEST signals with
3 mT that were 80 and 50% of maximum CEST signal ampli-
tude. Although stronger CEST signals could be generated with
higher saturation powers, we performed subsequent studies at
3 mT saturation power because this power level has been used
ð% CESTÞ9:0ppm
Reaction coordinate ¼
ð2Þ
ð% CESTÞ5:0ppm
The same study was performed with sulfatase enzyme from
abalone mollusk (a sea snail) and A. aerogenes (a bacterial coli-
form), which also resulted in the appearance of a CEST signal
at d=9.0 ppm without changing the CEST signal at d=
5.0 ppm (Figure S2, Supporting Information). The CEST spectra
from these additional reactions were identical to Figure 1b, al-
though the reaction with A. aerogenes had a broader CEST
peak at d=0 ppm from the direct saturation of water due to
the addition of glycerol that was packaged with this enzyme
by the commercial source. These additional studies demon-
Chem. Eur. J. 2016, 22, 6491 – 6495
www.chemeurj.org
6492
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Figure 2. Evaluation of CEST MRI. CEST signal amplitudes from the amide of the substrate, the amide of the product, and the salicylic acid moiety of the prod-
uct are shown as circles, squares, and diamonds, respectively. a) A saturation time of 4 s was sufficient to generate a strong CEST signal from the substrate
and product. b) A saturation power of 3 mT generated sufficiently strong CEST signal for detection. c) Extrapolation of the CEST-concentration relationship for
the substrate showed that 1.3 mm of substrate can generate a 1% CEST signal. d) The CEST signal amplitude of the amides are dependent on pH, while the
CEST signal from the salicylic acid moiety is relatively independent of pH.
for clinical studies, and the CEST signal amplitudes were suffi-
ciently strong with 3 mT saturation. The chemical exchange
rates calculated from the QUESP results for the amide proton
of the substrate and product were 560 and 1050 Hz. This accel-
eration in exchange rate was expected due to the electron-
withdrawing potential of the aryl carboxylate on the agent
after enzyme cleavage. This increase in exchange rate was less
than twofold, which only caused a minor change in CEST
signal amplitude of the amide (Figure 1b), so that the amide’s
CEST signal could still be considered largely unresponsive to
sulfatase activity. The chemical exchange rate of the salicylic
acid of the product was 1500 Hz, which matched previous
reports of the exchange rate from similar compounds.[9]
The CEST signal from the salicylic acid proton of the product
was relatively insensitive to pH (Figure 2d). Therefore, the ratio
of the CEST signals of the product is pH-dependent, which
should be taken into consideration during in vivo studies. In
particular, the tumor microenvironment can be acidic and
reach pH values between 6.4 and 7.0,[21,22] which would cause
the ratio of CEST signals from salicylic acid versus amide to be
overestimated.
More recently, paramagnetic CEST agents have been shown
to cause T2 relaxation due to their large chemical shifts and
fast chemical exchange rates that can cause a T2 exchange re-
laxation process.[23] The salicylic acid moiety of the product has
a remarkably high chemical shift relative to other diamagnetic
CEST agents, and also has a fast exchange rate. We performed
T1 and T2 MR studies with 30 mm of the substrate and the
product, which was synthesized to ensure purity rather than
relying on a product of an enzyme reaction (Figure S4, Sup-
porting Information). However the T2 relaxivities for both com-
pounds were negligible, and the T1 relaxivities were too low to
be reliably measured. Therefore, this contrast agent only acts
as a CEST agent.
To determine the detection sensitivity of the substrate, we
used a linear HW-Conc analysis method to correlate CEST
signal amplitude with concentration (Figure 2c, Figure S3c,
Supporting Information). This nonlinear correlation can be ex-
trapolated to lower concentration, which showed that a mini-
mum of 1.3 mm of the reactant could generate a 1% CEST
signal. Therefore, a concentration of CEST agent can detect sul-
fatase activity that is well below the 17 mm concentrations
that cause cell toxicity with these types of compounds.[20]
To investigate if the CEST signals are dependent on pH, we
acquired CEST spectra for 25 mm solutions at pH values rang-
ing from 6.40 to 7.60 units for the substrate and the product
from incubation with sulfatase from H. pomatia. The CEST
signal of the amide proton of the substrate and product in-
creased with increasing pH, which confirmed that the chemical
exchange of this proton was base-catalyzed as expected.[16]
To demonstrate the utility of this agent to detect sulfatase
enzyme activity from mammalian species, we performed cata-
lyCEST studies with HEK293 and BT549 cells. The HEK293
human embryonic cell line is known to express extracellular
sulfatase enzymes,[2] while the BT549 human breast cancer cell
line does not show sulfatase expression.[26] We added 100 mL of
each cell suspension to 200 mL solutions of the substrate in
Tris buffer at pH 7.3 and incubated for 24 h at 378C. The pH of
Chem. Eur. J. 2016, 22, 6491 – 6495
www.chemeurj.org
6493
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Figure 3. Detection of sulfatase activity from mammalian cells. a) The CEST spectra before (blue) and after (red) incubation of the agent with HEK293 cell sus-
pension showed the appearance of a CEST signal at d=9.0 ppm. b) The same CEST MRI study with BT549 cell suspension showed no appearance of a CEST
signal at d=9.0 ppm. In both studies, the CEST signal at d=5.0 ppm was largely unresponsive to enzyme activity. The broader linewidths of the direct satura-
tion of water in these samples were attributed to the biological milleau of the cell suspension that typically shortens the T2 relaxation time of the sample. c,d)
The CEST signals determined from fitting Lorentzian line shapes to the CEST spectra shown in panels (a) and (b) further demonstrate the detection of sulfatase
activity from HEK293 cells and no sulfatase activity was detected from BT529 cells. The Lorentzian line shapes in panel (d) are almost exactly overlapped.
the solution was adjusted to 7.2 and CEST spectra were ac-
quired for the product from incubation with both cell lines. We
observed the appearance of CEST signal at d=9.0 ppm after
incubation with HEK293 suspension, indicating the detection
of extracellular sulfatase activity from this cell line (Figure 3a).
Conversely, no CEST signal appeared after incubation with
BT549 suspension, demonstrating that this cell line did not ex-
press extracellular sulfatase (Figure 3b). For comparison, opti-
cal imaging agents that have been developed to detect sulfa-
tase activity were shown to detect sulfatase activity only in
bacterial cell suspensions, and could not detect sulfatase activi-
ties with HEK293 and BT549 mammalian cell suspensions.[4]
This comparison suggests that the aryl sulfate esters for optical
imaging could be specific for only a subset of sulfatase iso-
forms, while the CEST agent has more ubiquitous detection of
sulfatase activity.
the substrate constitutes a platform technology that may be
easily redesigned to detect other enzyme activities.
In conclusion, we have synthesized and characterized the
first diamagnetic CEST MRI contrast agent that can detect sul-
fatase enzyme activity. This substrate employed the use of
enzyme-responsive and unresponsive CEST signals that could
be detected at potentially nontoxic concentrations and satura-
tion conditions that are clinically translatable. This substrate
can detect sulfatase activity from mammalian cells as well as
from other species. Based on these preliminary studies, in vivo
studies with small animal models appear warranted to facilitate
the translation of catalyCEST MRI to the radiology clinic.
Experimental Section
Synthesis
Our previous studies have conjugated an enzyme-cleavable
ligand to the amine of compound 1 to form an amide.[16,24]
Enzyme cleavage of this ligand converted the amide to an
amine, causing a loss of the CEST signal from the amide, there-
by causing this agent to be “deactivated”. Our CEST agent in
this study improves on this previous design by “activating” the
CEST signal in response to enzyme activity. Furthermore, our
CEST agent in this study demonstrates that the salicylic acid
moiety can be derivatized while an aryl amine can generate an
enzyme-unresponsive CEST signal. This new approach expands
the chemical design of catalyCEST MRI contrast agents, and
the modular approach of derivatizing one or more positions of
Compounds 2, 3, and 4 were synthesized following reported pro-
cedures. Synthesis procedures for compounds 5, 6, and 7, and
characterizations of compounds 2--7, are reported in the
Supporting Information.
CEST MRI acquisition
MRI studies were performed with a Biospec MRI scanner operating
at 7 T (300 MHz) magnetic field strength with a 72 mm volume
transceiver coil (Bruker Biospin, Inc., Billerica, MA). The samples
were maintained at 37.0Æ0.28C. The CEST-FISP acquisition proto-
col used the following parameters TR: 3.196 ms; TE: 1.598 ms; exci-
Chem. Eur. J. 2016, 22, 6491 – 6495
www.chemeurj.org
6494
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
[1] S. R. Hanson, M. D. Best, C. H. Wong, Angew. Chem. Int. Ed. 2004, 43,
5736–5763; Angew. Chem. 2004, 116, 5858–5886.
[2] J. S. Rush, K. E. Beatty, C. R. Bertozzi, ChemBioChem 2010, 11, 2096–
2099.
[3] E. L. Smith, C. R. Bertozzi, K. E. Beatty, ChemBioChem 2014, 15, 1101–
1105.
[4] B. Yoo, M. D. Pagel, Front. Biosci. 2008, 13, 1733–1752.
[5] I. Hancu, W. T. Dixon, M. Woods, E. Vinogradov, D. A. Sherry, E. R. Lenkin-
ski, Acta Radiol. 2010, 51, 910–923.
tation flip angle: 308; centric encoding during acquisition; matrix:
128”128; field of view: 8”8 cm; in-plane spatial resolution: 625”
625 mm; slice thickness: 1 mm; number of slices: 1; number of
averages: 1.[25] We acquired a series of 85 images with selective sat-
uration applied at 3 mT for 5 seconds at saturation frequencies in
0.25 ppm increments from 15 to À3 ppm and 1 ppm increments
from À4 to À15 ppm, for a total scan time of 7:42 min.
[6] G. Liu, X. Song, K. W. Chan, M. T. McMahon, NMR Biomed. 2010, 23, 97.
[7] T. Chauvin, P. Durand, M. Bernier, H. Meudal, B. T. Doan, F. Noury, E.
Tóth, Angew. Chem. Int. Ed. 2008, 47, 4370–4372; Angew. Chem. 2008,
120, 4442–4444.
[8] B. Yoo, V. R. Sheth, C. M. Howison, M. J. Douglas, C. T. Pineda, E. A.
Maine, A. F. Baker, M. D. Pagel, Magn. Reson. Med. 2014, 71, 1221–1230.
[9] P. H. Kuo, E. Kanal, A. K. Abu-Alfa, S. E. Cowper, Radiol. 2007, 242, 647–
649.
CEST MRI analysis
Images were processed using ParaVision v5.1, and image analyses
were performed with Matlab v8.4 (Mathworks, Natick, MA). To mea-
sure amplitudes, each CEST spectrum from a region of interest in
an image of a sample was analyzed by fitting the spectrum to
a sum of three Lorentzian line shapes to account for two CEST
signals and direct saturation of water. The center, width, and
amplitude of each Lorentzian line was allowed to change to
optimize the fit.
[10] X. Yang, N. N. Yadav, X. Song, S. Ray Banerjee, H. Edelman, I. Minn, P. van
Zijl, M. G. Pomper, M. T. McMahon, Chem. Eur. J. 2014, 20, 15824–
15832.
[11] J. Li, X. Feng, W. Zhu, N. Oskolkov, T. Zhou, B. K. Kim, N. Baig, M.T McMa-
hon, E. Oldfield, Chem. Eur. J. 2016, 22, 264–271.
[12] S. Haftchenary, H. A. Luchman, A. O. Jouk, A. J. Veloso, B. D. Page, X. R.
Cheng, S. S. Dawson, N. Grinshtein, V. M. Shahani, K. Kerman, D. R.
Kaplan, ACS Med. Chem. Lett. 2013, 4, 1102–1107.
[13] S. Fletcher, P. T. Gunning, Tetrahedron Lett. 2008, 49, 4817–4819.
[14] H. Mikula, B. Sohr, P. Skrinjar, J. Weber, C. Hametner, F. Berthiller, R.
Krska, G. Adam, J. Frçhlich, Tetrahedron Lett. 2013, 54, 3290–3293.
[15] Y. Liu, I. F. Lien, S. Ruttgaizer, P. Dove, S. D. Taylor, Org. lett. 2004, 6,
209–212.
[16] D. V. Hingorani, L. A. Montano, E. A. Randtke, Y. S. Lee, J. Cµrdenas-Rodrí-
guez, M. D. Pagel, Contrast Media Mol. Imaging 2015.
[17] Y. Li, V. R. Sheth, G. Liu, M. D. Pagel, Contrast Media Mol. Imaging 2011,
6, 370.
T1 and T2 MRI acquisition
T1 and T2 studies were performed using the same MRI scanner. A
25 mm solution of substrate and the chemically synthesized prod-
uct, in PBS at pH 7.2, were used for these studies. All MRI experi-
ments were performed at 37.08C. A spin-echo MRI protocol was
then employed to measure R1 relaxation rates with a 12.7 ms TE;
10 TR times between 18.339 and 10000 ms; 1 mm slice thickness;
782”782 mm in-plane resolution; 5”2.5 cm field of view; and 64”
32 matrix. The same protocol was used to measure R2 relaxation
rates with a 15–960 ms TE in increments of 15 ms; 40 s TR time;
1.07 mm slice thickness; 400”400 mm in-plane resolution; 5.12”
2.56 cm field of view; and 128”64 matrix size. To calculate the R1
and R2 relaxation rate, the signal amplitudes were fitted with
a monoexponential function with a constant to account for a
possible DC offset using ParaVisionꢁ PV5.1 (Bruker Biospin, Inc.,
Billerica, MA).
[18] E. A. Randtke, L. Q. Chen, M. D. Pagel, Contrast Media Mol. Imaging
2014, 9, 252–258.
[19] E. A. Randtke, L. Q. Chen, L. R. Corrales, M. D. Pagel, Magn. Reson. Med.
2014, 71, 1603–1612.
[20] E. Noble, L. Janssen, P. J. Dierickx, Cell Biol. Toxicol. 1997, 13, 445–451.
[21] O. Warburg, Science 1956, 123, 309–314.
[22] M. G. Vander Heiden, L. C. Cantley, C. B. Thompson, Science 2009, 324,
1029–1033.
Acknowledgements
[23] T. E. Soesbe, M. E. Merritt, K. N. Green, F. A. Rojas-Quijano, A. D. Sherry,
Magn. Reson. Med. 2011, 66, 1697–1703.
[24] B. Yoo, M. D. Pagel, J. Am. Chem. Soc. 2006, 128, 14032–14033.
[25] T. Shah, L. Lu, K. M. Dell, M. D. Pagel, M. A. Griswold, C. A. Flask, Magn.
Reson. Med. 2011, 65, 432–437.
The authors thank the Experimental Mouse Shared Resource of
the University of Arizona Cancer Center for assistance with cell
preparation. The research was supported by through the NIH
through grants R01A169774 and P30A023074.
[26] C. Zhu, L. He, X. Zhou, X. Nie, Y. Gu, Oncol. Rep. 2016, 35, 1318–1328.
Keywords: diagnostic agents · enzyme kinetics · kinetics ·
NMR imaging · sulfatases
Received: February 14, 2016
Published online on March 23, 2016
Chem. Eur. J. 2016, 22, 6491 – 6495
www.chemeurj.org
6495
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim