Preliminary evaluation of PARACEST MRI agents for the detection of nitric oxide synthase

Article abstract of DOI:10.1139/cjc-2016-0179

Several paramagnetic chemical exchange saturation transfer magnetic resonance imaging (PARACEST MRI) agents for the potential detection of nitric oxide synthase (NOS) have been synthesized and evaluated. These agents are based on an amino acid-or dipeptide-decorated DOTAM (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid amide) chelator possessing either Tm³⁺ or Dy³⁺. The amino acid and dipeptide decorated DOTAMs were designed such that the terminal amino acid pendant group was l-arginine, which may be converted to l-citrulline by NOS. Preliminary evaluation has revealed that some of the l-arginine-decorated complexes are recognized and metabolized by the NOS. Differences in the CEST properties between Dy³⁺-metallated l-arginine-and l-citrulline-modified complexes suggest that these might be suitable for imaging of the NOS enzymatic activity.

Full text of DOI:10.1139/cjc-2016-0179

Page 1 of 27
Preliminary evaluation of PARACEST MRI agents for the detection of nitric
oxide synthase (NOS)
Mojmír Suchý,a.b,1 Alex X. Li,b Yin Liu,c,2 Qingping Feng,c Robert Bartha,b and Robert H. E.
Hudson*a,d
a Department of Chemistry and dThe Centre for Advanced Materials and Biomaterials Research,
The University of Western Ontario, London, Ontario, Canada, N6A 5B7
b Robarts Research Institute, The University of Western Ontario, London, Ontario, Canada, N6A
5K8
c Department of Physiology and Pharmacology, The University of Western Ontario, London,
Ontario, Canada, N6A 5C1
1Current Address: Department of Chemistry and Biomolecular Sciences, University of Ottawa,
Ottawa, ON, Canada, K1N 6N5
2Current Address: Department of Pharmacology, Sichuan University, Chengdu, Sichuan 610041, PR
China
Eꢀmail: rhhudson@uwo.ca Tel: 519ꢀ661ꢀ2111 ext. 86349 Fax: 519ꢀ661ꢀ3022
1

Page 2 of 27
Abstract
Several paramagnetic chemical exchange saturation transfer magnetic resonance imaging
(PARACEST MRI) agents for the potential detection of nitric oxide synthase (NOS) have been
synthesized and evaluated. These agents are based on an amino acid or dipeptide decorated
DOTAM (1,4,7,10ꢀtetraazacyclododecaneꢀ1,4,7,10ꢀtetraacetic acid amide) chelator possessing
either Tm3+ or Dy3+. The amino acid and dipeptide decorated DOTAMs were designed such that
the terminal amino acid pendant group was Lꢀarginine which may be converted to Lꢀcitrulline by
NOS. Preliminary evaluation has revealed that some of the Lꢀarginineꢀdecorated complexes are
recognized and metabolized by the NOS. Differences in the CEST properties between Dy3+ꢀ
metallated Lꢀarginineꢀ and Lꢀcitrullineꢀmodified complexes suggest that these might be suitable
for imaging of the NOS enzymatic activity.
Keywords: Nitric Oxide Synthase, PARACEST, MRI, DOTAM, thulium, dysprosium, citrulline,
arginine, cyclen.
2

Page 3 of 27
Graphical Abstract
3

Page 4 of 27
Introduction
The use of magnetic resonance imaging (MRI) for medical purposes has become widely
practiced because it offers excellent spatial resolution and deep tissue penetration while being
devoid of harmful ionizing radiation. For MR imaging that requires the use of an externally
administered contrast agent, those based on Gd3+ chelates are often used.1 In recent years there
has been interest in developing chelates of nonꢀgadolinium lanthanide ions as contrast agents. In
particular, lanthanide(III) chelates that demonstrate slow exchange of water (bound to bulk) at
the metal centre2 or possess exchangable protons in the ligand (e.g. NꢀH, OꢀH, C(O)NH) can
operate by a chemical exchange saturation transfer (CEST) mechanism as originally described by
Balaban and coworkers for diamagnetic molecules.3 When a paramagnetic metal chelate is used,
they are termed paramagnetic CEST (PARACEST) MRI agents.4
The exchange processes that produce the PARACEST phenomenon are influenced by the
microenvironment of the agent, thus it becomes possible to measure physiological of interest
conditions such as temperature and pH.5 Beyond this, if the ligand is tranformed chemically,
such a change may be manifested in the CEST properties of the chelate.6 This concept has been
exploited for the detection of various enzymatic activities including the proteases caspaseꢀ37 and
cathepsin D,8 pig liver esterase,9 urokinase plasminogen activator,10 βꢀgalactosidase,11 and
transglutaminase.12
Our previous studies have been directed toward the development of synthetic methodologies to
access peptide decorated DOTAMs13,14 and their application as potential PARACEST MRI
4

Page 5 of 27
contrast agents. 15 16 These agents generally rely on a CEST effect generated by exchange of a
,
metal bound water, for example Eu3+ DOTAMꢀGlyꢀPheꢀOH (1, Figure 1), or exchange of the
amide proton proximal to the cyclen ring, for example Tm3+ DOTAMꢀGlyꢀLysꢀOH (2, Figure 1).
The synthetic methodolgy leading to CAs 1 and 2 was found to be robust, versatile and scalable13
and we were therefore interested in expanding its scope. Notably absent in the initially reported
method were arginineꢀdecorated DOTAMs. Thus, the method was extended17 because we were
interested to investigate if arginineꢀdecorated DOTAMs might possess similar or improved
properties as compared to Tm3+ DOTAMꢀGlyꢀLysꢀOH (2).18
Figure 1 A) Generalized structure of an amino acid decorated DOTAM showing the proximal
amide protons responsible for the CEST effect, in bold, and the metal bound water as indicated
5

Page 6 of 27
by the arrows (R = amino acid sidechain). B) Specific structural examples of dipeptideꢀdecorated
DOTAM PARACEST CAs. C) Chemical structures of arginineꢀdecorated CAs 3, 4 (R = amino
acid or dipeptide, as indicated).
We have focussed on lanthanide complexes of metals that are able to induce large hyperfine
shifts of the proximal amide protons, namely Dy3+ and Tm3+ in this work, in order to generate a
CEST signal well separated from the bulk water signal. With some arginineꢀdecorated DOTAM
chelators in hand, we choose to investigate the possibility of using them to detect nitric oxide
synthase activity. Our previous studies,13,14 and those of others,19 indicated that the identity of
the amino acid pendant arms influenced the observed CEST signals presumably by modulation
of the exchange rates and subtle effects on the ligand conformation. Thus, it was hypothesized
that conversion of the charged amino acid arginine to the uncharged citrulline would affect the
CEST signals produced by the lanthanide(III) complexes.
Among the various types of MRI CAs, PARACEST agents20 are well suited for the potential
detection of NOS enzymatic activity.
As noted, these CAs are sensitive to their
microenvironment and have been previously used for the detection of various enzymatic
activities, vide supra. Nitric oxide (NO)21 is a key signaling molecule associated with various
physiological processes, e.g. vasodilatation, immune response to infection or wound healing.
Altered levels of NO can serve as a useful biomarker of certain pathologies, e.g. cardiovascular
diseases or breast and prostate cancers. Production of NO within the body is orchestrated by the
enzyme nitric oxide synthase (NOS), which utilizes an essential amino acid Lꢀarginine as a
substrate.22 Enzymatically, Lꢀarginine (LꢀArgꢀOH) is first oxidized to LꢀNꢀhydroxyarginine, and
6

Page 7 of 27
in a second step, further oxidized to Lꢀcitrulline (LꢀCitꢀOH)along with the simultaneous
production of NO (Scheme 1).22
Scheme 1 Metabolism of Lꢀarginine to Lꢀcitrulline accompanied by release of nitric oxide.
Several agents for the detection of NO have been developed over the course of last two
decades. These involve probes which can be tracked by means of optical (fluorescence) imaging
or other spectroscopic techniques such as surfaceꢀenhanced Raman scattering (SERS).23,24,25,26
The detection of NOS enzymatic activity by radiolabeled imaging agents (11C, 99mTc) compatible
with positron emission tomography (PET)27 or single photon emission computer tomography
(SPECT)28 has also been reported (5, Scheme 2). The sole example of a MRI CA related to the
detection of NOS enzymatic activity was described by Pagel and coworkers in 2007 and relies on
the detection of nitric oxide (produced by NOS).
Their contrast agent operates in the
PARACEST mode and is based on dimerization oꢀphenylenediamineꢀdecorated DOTAꢀderived
Yb3+ complex in the presence of NO and O2 (6, Scheme 2).29 The dimerization of 6 caused an
irreversible loss of both CEST effects: the anticipated loss of the CEST effect due to the aromatic
7

Page 8 of 27
NH2 (at +8 ppm) and the unexpected loss of the CEST effect due to the amide proton (at ꢀ11
ppm). In general, for analytical purposes, the loss of a signal could be a result of factors other
than the dimerization reaction (e.g. dilution); thus, it is desireable to investigate other agents that
provide either a ‘turn on’ of signal or that demonstrate some other observeable change such as a
change in the chemical shift of the CEST effect.
Scheme 2 Molecular probes for the detection of NO. A) Radiometalꢀbased imaging agents, 5.
B) PARACEST MRI CA 6 that undergoes NO mediated dimerization.
It is recognized that NOS is a promiscuos enzyme with a wide substrate scope.22 As shown by
recent work from Correia and colleagues, Re(III) complexes of the general structure 5 (Scheme
2) were found to be recognized by NOS and served as competitive inhibitors of its enzymatic
activity. Moreover, these complexes were found to cross cellular membranes and their 99mTc
counterparts have been used as SPECT probes for the detection of NOS enzymatic activity.28
Armed with this knowledge we have investigated the use of a series of Lꢀarginineꢀdecorated
PARACEST MRI CAs (structures 3 and 4, Figure 1 and Scheme 3) that were recently prepared
8

Page 9 of 27
in our laboratory18 for their ability to be metabolized by NOS. The metabolism of 3 or 4 by NOS
would produce the corresponding Lꢀcitrullineꢀdecorated DOTAMs (Scheme 3) and release NO.
We have independently synthesized the Lꢀcitrullineꢀderived CAs 7 and 8, evaluated their CEST
properties and compared them with those associated with 3 and 4. The metabolism of complexes
3 or 4 was determined by enzymatic mediated NO production using commercially available
bovine endothelial NOS (eNOS). The production of NO was detected electrochemically and
compared to the free amino acid Lꢀarginine, the natural substrate of NOS. The results of these
studies are described below and indicate that CA 4a shows promise for the PARACEST MRIꢀ
based detection of NOS enzymatic activity.
HO-Arg
O
Arg-OH
O
HO-Arg-Gly
Gly-Arg-OH
O
N
N
N
N
O
N
N
N
N
Dy3+
Ln3+
OR
O
O
O
O
O
OH
HO-Arg
HN
HN
HO-Arg-Gly
4a: Ln = Dy
4b: Ln = Tm
3
NH
NH
HN
O
NH2
HO
NH
O
HN
NH2
NOS
-NO
Cit-OH
O
HO-Cit
O
HO-Cit-Gly
O
Gly-Cit-OH
N
N
N
N
N
N
N
O
Dy3+
Ln3+
OR
O
O
O
O
N
O
OH
HO-Cit
HN
HO-Cit-Gly
HN
8a, Ln = Dy
8b, Ln = Tm
7
NH
NH
O
NH2
HO
O
O
O
HN
NH2
Scheme 3 The proposed NOSꢀmediated metabolism of LꢀArg modified CAs 3 and 4 to form the
LꢀCit modified CAs 7 and 8
9

Page 10 of 27
Results and discussion
Chemistry
The synthesis of CAs 7 and 8a,b was accomplished by the “cyclen tetraalkylation” approach
that has been previously employed for the preparation of LꢀArgꢀdecorated CAs 3 and 4.18
Briefly, esterification (SOCl2/MeOH) of commercialy available LꢀCitulline (LꢀCitꢀOH, Scheme
4) afforded the corresponding methyl ester 9 (LꢀCitꢀOMe
●
HCl) in quantitative yield.
A
dicyclohexylcarbodiimide (DCC)/Nꢀhydroxysuccinimide (NHS)ꢀmediated coupling between 9
and chloroacetic acid or Nꢀchloroacetylglycine provided electrophiles 10a (81%) or 10b (30%),
respectively, Scheme 4. The full spectral characterization of 10a and 10b can be found in
Experimental and in Supplemental Information. An aqueous extractive workup, previously used
for the preparation of related electrophiles,13,30 was not used, as we anticipated poor partitioning
of 10a and 10b into organic solvents commonly used. Instead the reaction mixtures were
adsorbed on SiO2 and then subjected to flash column chromatography (FCC) purification. The
relatively low yield of the more polar electrophile 10b is attributed to an increased retention on
the silica gel.
Scheme 4 Preparation of electrophiles 10a,b
10

Page 11 of 27
With the electrophiles 10a and 10b the peralkylation of cyclen (11) was performed (Scheme
5). The free ligands, DOTAMꢀCitꢀOCH3 (10a) and DOTAMꢀGlyꢀCitꢀOCH3 (10b)were purified
by semiꢀpreparative high pressure liquid chromatography (HPLC) and were obtained as the
trifluoroacetate salts in moderate yields (Scheme 5, see Experimental and Supplemental
Information for details). Metallation of the ligands was performed by treatment with
corresponding lanthanide(III) chlorides as shown in Scheme 5. The pH of the reaction was
initially maintained at ca. 10 to ensure complete hydrolysis of ester groups. Once the esters
hydrolyzed, the pH was lowered to ca. 5ꢀ6 to achieve the complete metallation. Resulting
reaction mixtures were purified by size exclusion chromatography (SEC) affording Lꢀcitrullineꢀ
modified CAs 7 and 8 (Scheme 5) in good yields (66ꢀ73%). The identities and purities of CAs 7
and 8 were confirmed by high resolution mass spectrometry (HRMS) as described in the
Experimental and depicted in the Supplemental Information.
11

Page 12 of 27
Scheme 5 Preparation of CAs 7 and 8
CEST spectra of CAs 3, 4, 7 and 8
With the desired CAs 3, 4, 7 and 8 in hand their CEST spectra were collected (10 mM
solutions in a phosphate buffer, pH 7.0, 37 ˚C). Further details related to CEST spectra
acquisition can be found in the Experimental. As depicted in Scheme 3, we hypothesized that Lꢀ
arginineꢀmodified CAs 3 and 4 would be recognized and metabolized by NOS with the
formation of corresponding Lꢀcitrullineꢀdecorated CAs 7 and 8. We were interested in
comparing the differences (e.g. chemical shift, signal intensity) observed in the CEST spectra
associated with CAs 3, 4, 7, and 8. These results are shown in Figure 2; for the sake of clarity
the CEST spectra associated with Lꢀarginineꢀmodified CAs 3 and 4 are depicted in red while the
12

Page 13 of 27
CEST spectra associated with Lꢀcitrullineꢀmodified CAs 7 and 8 are depicted in green.
The Tm3+ complexes 4b (Arg) and 8b (Cit) exhibited a CEST effect (4b, 28%; 8b, 19%) due to
the presence of exchangeable amide protons at ꢀ45 ppm (Figure 2, panels A and B).
Unfortunately, we did not observe a difference in the chemical shift between Lꢀarginineꢀ and Lꢀ
citrullineꢀmodified CAs, implying that the CA 8b would not be suitable for the detection of NOS
enzymatic activity by PARACEST MRI.
The Dy3+ complexes 3, 4a, 7 and 8a appeared more promising in terms of their future use for
the detection of NOS enzymatic activity. Dy3+ complexes 4a (Arg) and 8a (Cit) featuring the
Gly linker exhibited a CEST effect (4a, 13%; 8a, 7%) due to the presence of exchangeable amide
protons at 71 ppm (4a) and 69 ppm (8a, Figure 2, panels C and D). Although the difference in
the chemical shift between the CEST effects associated with Lꢀarginineꢀmodified (4a) and Lꢀ
citrullineꢀmodified (8a) CAs is rather small (ꢁδ 2 ppm), further use of CAs 4a and 8a for the
imaging of NOS enzymatic activity in cell culture studies or in vivo maybe be feasible since
similar small differences in chemical shifts are oftentimes exploited when using endogenous
macromolecules (e. g. proteins, saccharides) as diamagnetic CEST (DIACEST) CAs for in vivo
imaging.31
The absence of the Gly linker in the complexes 3 (Arg) and 7 (Cit) resulted in a more dramatic
difference in chemical shift between the CEST effects associated with 3 (71 ppm, 10%) and 7
(76 ppm, 5%, Figure 2, panels E and F, ꢁδ 5 ppm). The signal intensity decrease of ca. 50% in
all three instances, (Figure 2) may be attributed to the different proton exchange rates in Lꢀ
arginineꢀ (3 and 4) and Lꢀcitrullineꢀmodified (7 and 8) complexes. The Lꢀcitrullineꢀmodified
CAs in this case would possess less favorable proton exchange rates resulting in a CEST signal
intensity decrease. It is also interesting to point out, that the CAs 3 and 7 bearing no Gly linker
13

Page 14 of 27
possess weaker CEST signal (Figure 2) as opposed to CAs 4 and 8 bearing the Gly linker.
Similar results have been previously observed by our group, when comparing the CEST spectra
associated with CAs 1 and 2 (Figure 1) with their counterparts lacking a Gly linker within their
structure.30
B
A
C
D
F
E
14

Page 15 of 27
Figure 2 Comparison of the CEST spectra associated with CAs 4b and 8b (panels A, B); 4a and
8a (panels C, D); 3 and 7 (panels E, F). The CEST spectra associated with Lꢀarginine modified
agents are depicted in red; the CEST spectra associated with Lꢀcitrulline modified agents are
depicted in green. Full view of the CEST spectra (ꢀ100 to 100 ppm) is shown on the left (panels
A, C, E); expanded view of the region of interest is shown on the right (panels B, D, F). CEST
spectra were collected on 10 mM solutions of chelator at pH 7.0 and 37 °C using a 9.4 Tesla
MRI scanner with 14 ꢀT saturation pulse, 3.95 second in duration.
Metabolism of 3, 4a, 7 and 8a with eNOS
Having selected two pairs of Lꢀarginineꢀ and Lꢀcitrullineꢀmodified PARACEST CAs (3 and 7,
4a and 8a) showing promise in detecting NOS enzymatic activity by CEST MRI, we determined
whether or not NOS can recognize and metabolize the Lꢀarginineꢀmodified CAs 3 and 4a. The
NOS enzymatic activity assay was performed with bovine eNOS and relied on the sensitive
detection of NO production by electrochemical means,32 the details are found in the
Experimental and the results are summarized in Table 1. The rates of metabolism were
qualitatively judged by comparison of the amount of NO produced relative to the natural
substrate Lꢀarginine.
Only the Lꢀarginineꢀmodified CA 4a, featuring a Gly linker was found to be metabolized by
eNOS with a rate of metabolism that was similar to the natural substrate Lꢀarginine (Scheme 1).22
The Lꢀarginineꢀmodified CA 3, lacking the Gly linker was, on the other hand, not metabolized by
eNOS. We hypothesize that the linker in 3 between the core of the chelator and the pendant
arginine residue is too short to permit proper interaction with enzyme. Our observation is
15

Page 16 of 27
consistent with the results described by Correia and coworkers, wherein the complexes of
general structure 5 (Scheme 2) possessing longer side chain linkers (5, Scheme 2, n = 4)
exhibited increased competitive inhibitory activity toward NOS when compared to the inhibitory
activity of those possessing a shorter linker (5, Scheme 2, n = 1).28 As expected, the Lꢀcitrullineꢀ
modified CAs 7 and 8a did result in the production of NO in the presence of eNOS.
Overall our results indicate that among the series of CAs herein studied, Dy3+ DOTAMꢀGlyꢀ
ArgꢀOH (4a, Scheme 3) is the probe of choice for the detection of NOS enzymatic activity. It
exhibits different CEST properties (chemical shift, signal intensity) compared to the
corresponding Dy3+ DOTAMꢀGlyꢀCitꢀOH (8a, Scheme 3) counterpart. Moreover 4a was
qualitatively judged to be metabolized by eNOS by comparison to the natural substrate, Lꢀ
arginine (Scheme 1), in terms of the rate and quantity of NO liberated as determined by
electrochemical detection.32
Table 1 Metabolism of Lꢀarginine (1) and CAs 3, 4a, 7 and 8a by bovine eNOS
Compound
Observed result
metabolized
Lꢀarginine
Dy3+ DOTAMꢀArgꢀOH (3)
Dy3+ DOTAMꢀGlyꢀArgꢀOH (4a)
Dy3+ DOTAMꢀCitꢀOH (7)
Dy3+ DOTAMꢀGlyꢀCitꢀOH (8a)
not metabolized
metabolized
not metabolized
not metabolized
Experimental
General experimental procedures
16

Page 17 of 27
All amino acids (naturally occurring L isomers) and reagents were commercially available,
unless otherwise stated. All solvents were HPLC grade and used as such, except for water (18.2
Mꢂ⋅cmꢀ1 Millipore water) and THF (dried over Al2O3 in a solvent purification system). Solvents
were removed under reduced pressure in a rotary evaporator. Aqueous solutions were
lyophilized. Flash column chromatography (FCC) was carried out using silica gel, mesh size
230 – 400 Å. Thin layer chromatography (TLC) was carried out on Al backed silica gel plates,
compounds were visualized by anisaldehyde stain, phosphomolybdic acid stain and I2 vapours.
Specific rotations [α]D were determined by polarimeter at ambient temperature using a 1 mL, 10
cm path length cell; the units are 10ꢀ1 deg cm2 gꢀ1 and the concentrations are reported in g/100
mL. Size exclusion chromatography (SEC) was peformed on BIOꢀGEL P2, 45ꢀ90 ꢃm mesh
resin (15 g, per 0.03 mmol of compound). Ten fractions (10 mL each) were collected and
identified by UV or ninhydrin test, fractions containing complexes 7 and 8 were combined and
lyophilized. Absence of free lanthanide(III) ions was verified by a xylenol orange test.33 HPLC
analysis and purification was done using a DeltaꢀPak C18 300Å column (particle size 15ꢃm; 8 ×
100 mm RadialꢀPak cartridge). Mobile phase: Method A: 90% H2O/10% MeCN – 45%
H2O/55% MeCN over 10 min; linear gradient and flow rate 3 mL/min. Ultra performance liquid
chromatography (UPLC) was performed using a BEH C18 column (particle size 1.7ꢃm; 1.0 id ×
100 mm) and HRꢀESIꢀMS detector. Mobile phase: Method B: 100% H2O – 100% MeCN over 5
min, linear gradient, flow rate 0.25 mL/min. NMR spectra were recorded on 400 MHz
1
spectrometer; for H (400 MHz), δ values were referenced as follows DMSOꢀD6 (2.49 ppm),
13
D2O (4.75 ppm), for C (100 MHz) DMSOꢀD6 (39.5 ppm). Mass spectra (MS) were obtained
using chemical ionization (CI) or electron spray ionization (ESI) TimeꢀofꢀFlight (TOF)
instrument. PARACEST properties of CAs 3, 4, 7, and 8 have been evaluated as follows: NMR
17

Page 18 of 27
tubes containing solutions of 3, 4, 7, or 8 in phosphate buffer (10 mM, pH 7.0) were imaged on a
9.4 Tesla MRI scanner (Agilent, Santa Clara, CA) at 37 °C. The temperature was monitored and
controlled by blowing hot air using a Model 1025 small animal monitoring and gating system
(SA Instruments, Inc., Stony Brook, NY).18 A fast spin echo pulse sequence (FOV: 12.8 × 12.8
mm2, matrix: 32 × 32, TR = 4000 ms, 4 echoes, and effective TE = 10 ms) was used, preceded
by a frequency selective saturation pulse (B1 = 14 ꢃT, saturation range = ꢀ100 to 100 ppm in
steps of 1 ppm, saturation time = 3.95 s). CEST spectra were generated using the total signal
intensity from each tube.18
LꢀCitulline methyl ester hydrochloride, LꢀCitꢀOMe HCl, 9
●
SOCl2 (190 ꢃL, 2.6 mmol) was added dropwise (over ca. 1 min period) to MeOH (1.5 mL)
cooled to 0 ˚C. LꢀCitrulline (350 mg, 2 mmol) was then added, the cooling bath was removed
and the resulting mixture was stirred for 48 h at room temperature (RT). The solvent was
evaporated and the residue was dried in high vacuum overnight to afford Lꢀcitrulline methyl ester
hydrochloride (450 mg, quantitative) as colorless oil of sufficient purity to be used in the next
1
step. [α]D +15.6 (c 0.7, H2O). H NMR (D2O) δ 3.99 (t, J = 6.5 Hz, 1H); 3.65 (s, 3H); 2.98 (t, J
= 6.5 Hz, 2H); 1.78 (m, 2H); 1.44 (m, 2H). 13C NMR (D2O) δ 170.4, 161.1, 53.6, 52.5, 39.1,
26.9, 24.5. HRMS (CI) m/z: found 190.1198 [M + H]+ (calcd 190.1192 for C7H15N3O3).
Preparation of electrophiles 10a and 10b
Nꢀhydroxysuccinimide (219 mg, 1.9 mmol) was added to separate solutions of chloroacetic acid
(10a, 180 mg, 1.9 mmol) or Nꢀchloroacetylglycine (10b, 288 mg, 1.9 mmol) in dry THF (10
mL). The mixtures were cooled to 0 ˚C, followed by the addition of DCC (510 mg, 2.5 mmol)
18

Page 19 of 27
and subsequent stirring for 1 h at 0 ˚C. Separate solutions of LꢀCitꢀOMe HCl (9, 450 mg, 2
●
mmol) in MeOH (2 mL)/THF (7 mL) were added, followed by the addition of Et3N (1.06 mL,
7.6 mmol). The cooling bath was removed and the mixture was stirred for 18 h at RT. The
solvents were evaporated; the residues were resuspended in acetone (ca. 20ꢀ30 mL), SiO2 was
added and the volatiles were removed. Resulting slurries were subjected to FCC on 40 g of
SiO2, CH2Cl2/MeOH/NH4OH (90:9:1). Evaporation of the eluates afforded desired products
containing small amount of Et3N HCl. The impurity was removed by repeated FCC on 40 g of
●
SiO2, CH2Cl2/MeOH/NH4OH (80:19:1, 10a) or 20 g of SiO2, CH2Cl2/MeOH/NH4OH (90:9:1,
10b). Evaporation of eluates afforded the desired products as colorless oils.
1
N–ChloroacetylꢀCitꢀOMe (10a, 409 mg, 81%); [α]D ꢀ13.4 (c 1, MeOH). H NMR (DMSOꢀD6) δ
8.68 (d, D2O exch., J = 7.5 Hz, 1H); 5.95 (t, D2O exch., J = 5.5 Hz, 1H); 5.39 (s, D2O exch.,
1H); 4.25 (m, 1H); 4.11 (dd, J = 14.5, 13.0 Hz, 2H); 3.63 (s, 3H); 2.94 (dd, J = 13.0, 6.5 Hz,
2H); 1.72 (m, 1H); 1.59 (m, 1H); 1.36 (m, 2H). 13C NMR (DMSOꢀD6) δ 172.1, 166.1, 158.7,
52.2, 52.0, 42.2, 38.6, 28.3, 26.5. HRMS (ESI) m/z: found 266.0895 [M + H]+ (calcd 266.0908
for C9H17ClN3O4).
1
N–ChloroacetylꢀGlyꢀCitꢀOMe (10b, 182 mg, 30%); [α]D ꢀ15.3 (c 1, MeOH). H NMR (DMSOꢀ
D6) δ 8.41 (m, D2O exch., 2H); 5.94 (t, D2O exch., J = 5.5 Hz, 1H); 5.39 (s, D2O exch., 1H); 4.24
(m, 1H); 4.13 (s, 2H); 3.79 (m, 2H); 3.62 (s, 3H); 2.94 (dd, J = 13.0, 6.5 Hz, 2H); 1.65 (m, 1H);
1.58 (m, 1H); 1.37 (m, 2H). 13C NMR (DMSOꢀD6) δ 172.5, 168.5, 166.2, 158.8, 51.9, 45.6,
42.5, 41.8, 38.6, 28.4, 26.6. HRMS (ESI) m/z: found 323.1126 [M + H]+ (calcd 323.1122 for
C11H20ClN4O5).
Peralkylation of cyclen (11) with electrophiles 10a and 10b
19

Page 20 of 27
Separate solutions of N–chloroacetylꢀCitꢀOMe (10a, 114 mg, 0.43 mmol) or N–chloroacetylꢀ
GlyꢀCitꢀOMe (10b, 140 mg, 0.43 mmol), cyclen (11, 19 mg, 0.11 mmol) and DIPEA (76 ꢃL) in
MeCN (1.8 mL)/DMF (0.2 mL) were stirred under reflux for 48 h. Solvents were evaporated;
the residues were dissolved in MeOH/H2O (1.5 mL each) and were subjected to semipreparative
HPLC purification as described in General experimental procedures. Concentration of fractions
containing the desired products as afforded colorless semiꢀsolid residues.
1
DOTAMꢀCitꢀOMe
●
4CF3COOꢀ (78 mg, 46%); tR 5.4 min (Method A); H NMR (DMSOꢀD6) δ
8.75 (m, D2O exch., 4H); 7.87 (m, D2O exch., 4H); 4.28 (m, 4H); 3.69 (br m, 8H); 3.64 (s, 12H);
3.26 (br m, 16H); 2.97 (m, 8H); 1.72 (m, 4H); 1.60 (m, 4H); 1.41 (m, 8H). HRMS (ESI) m/z:
found 1089.5997 [M + H]+ (calcd 1089.6016 for C44H81N16O16).
DOTAMꢀGlyꢀCitꢀOMe
4CF3COOꢀ (62 mg, 32%); tR 4.3 min (Method A); 1H NMR (DMSOꢀD6)
●
δ 8.53 (m, D2O exch., 8H); 7.77 (m, D2O exch., 4H); 6.12 (m, D2O exch., 4H); 4.25 (m, 4H);
3.84 (m, 8H); 3.70 (br m, 8H); 3.62 (s, 12H); 3.26 (br m, 16H); 2.95 (m, 8H); 1.68 (m, 4H); 1.57
(m, 4H); 1.39 (m, 8H). HRMS (ESI) m/z: found 1317.6915 [M + H]+ (calcd 1317.6875 for
C52H93N20O20).
Metallation of the ligands DOTAMꢀCitꢀOMe and DOTAMꢀGlyꢀCitꢀOMe
Separate solutions of DOTAMꢀCitꢀOMe
4CF3COOꢀ (51 mg, 2.9 × 10ꢀ5 mol) in water (2 mL) were treated with DyCl3
mg, 2.9 × 10ꢀ5 mol, both ligands) or TmCl3 H2O (8 mg, 2.9 × 10ꢀ5 mol, DOTAMꢀGlyꢀCitꢀOMe ·
●
4CF3COOꢀ (45 mg, 2.9 × 10ꢀ5 mol) and DOTAMꢀGlyꢀ
CitꢀOMe
●
●
6H2O (11
●
4CF3COOꢀ). The pH was adjusted to ca. 9ꢀ10 (1 M NaOH) and the mixtures were stirred for 24
h at RT, complete saponification took place, along with incomplete metallation.
DOTAMꢀCitꢀOH; HRMS (ESI) m/z: found 1033.5345 [M + H]+ (calcd 1033.5390 for
20

Page 21 of 27
C40H73N16O16).
DOTAMꢀGlyꢀCitꢀOH; HRMS (ESI) m/z: found 1261.6304 [M + H]+ (calcd 1261.6249 for
C48H85N20O20).
The pH was then adjusted to ca. 5ꢀ6 (1 M HCl) and the mixtures were stirred for 24 h at RT,
followed by SEC as described in General experimental procedures. Fractions containing the
products were concentrated to leave the desired complexes as colorless solids.
Dy3+ DOTAMꢀCitꢀOH (7, 25 mg, 73%); HRMS (ESI) m/z: found 1192.4462 [M ꢀ 2H]+ (calcd
1192.4424 for C40H70N16O16Dy).
Dy3+ DOTAMꢀGlyꢀCitꢀOH (8a, 28 mg, 69%); HRMS (ESI) m/z: found 1421.5372 [M ꢀ 2H]+
(calcd 1421.5302 for C48H82N20O20Dy).
Tm3+ DOTAMꢀGlyꢀCitꢀOH (8b, 27 mg, 66%); HRMS (ESI) m/z: found 1427.5377 [M ꢀ 2H]+
(calcd 1427.5357 for C48H82N20O20Tm).
LꢀArginineꢀmodified CAs 3, 8a and 8b
These CAs and the synthetic intermediates leading to them were prepared as described
previously.18
NOS enzyme assay with 3, 4, 7, 832
The NOS assay was carried out using a NO sensor with the “inNO” Nitric Oxide Measuring
System (Innovative Instruments Inc., USA) and used bovine eNOS (Sigma) and ancillary
reagents from a Cayman Chemical NOS activity assay kit. The assays were performed
according to manufacturer’s instructions using 100 ꢃL of 100 mM solutions of Lꢀarginineꢀ
modified agents 3 and 4a. LꢀCitrullineꢀmodified agents 7 and 8a were used as a negative
21

Page 22 of 27
control; Lꢀarginine was used as a positive control. The rate of metabolism was judged by the
time to return to baseline reading in the amperogram and the amount was judged by the
maximum peak height.
Conclusions
In this study, we have described PARACEST MRI CAs capable of detecting NOS enzymatic
activity directly. From the chemical structure point of view, the agents are ligands composed of
amino acid decoratedꢀcyclens which have been metallated with Dy3+ and Tm3+. The complexes
feature terminal monoꢀ or dipeptide arms incorporating Lꢀarginine as a substrate for NOS or Lꢀ
citrulline as a product of enzymatic transformation mediated by NOS. The structure of CAs
described in this work resembles other dipeptidic PARACEST MRI CAs13,30 developed in our
laboratory and utilized for in vivo temperature and pH mapping.15,16
A detailed evaluation of CEST spectra associated with the CAs revealed that Tm3+ complexes
are not suitable for the detection of NOS enzymatic activity as there is no difference in chemical
shift of the CEST effect (amide protons), when the Lꢀarginineꢀmodified agents and Lꢀcitrulline
modified counterparts are compared. The Dy3+ꢀmetallated complexes are more favourable as
NOS enzymatic activity sensors because they produced a 2 ppm (8a, Gly linker present) or 5
ppm (7, Gly linker absent) difference of the chemical shift for the proximal amide protons in
their CEST spectra of the Lꢀarginineꢀ vs. Lꢀcitrullineꢀcontaining agents, respectively. The small
change in the frequency and intensity of the CEST signal is attributed to modest changes the
conformation of the ligand (δ shift) and amide proton exchange rates (intensity change) due to
the structural change of the pendant arms and concommittant changes in their solvation.
22

Page 23 of 27
The selected agents which showed promise for the detection of NOS enzymatic activity by
PARACEST MRI (i.e. 3, 4a) were subjected to a qualitative NOS enzymatic activity assay. This
assay revealed only agent 4a (Gly linker) was metabolized by NOS. This observation is
consistent with the results observed by other researchers28 and provides a useful insight for the
design of other oligopeptideꢀbased CAs/enzyme inhibitors capable of detecting/inhibiting NOS
enzymatic activity.
When compared to the previously reported “turn off” NO PARACEST MRI based sensor 6,29
wherein the presence of NO is detected by the disappearence of the CEST effects associated with
6, CA 4a has two notable differences. Firstly, NOS enzymatic activity is detected by a change in
signal, i.e. the difference of the amide proton chemical shift in the CEST spectra associated with
the Lꢀarginineꢀ vs. Lꢀcitrullineꢀmodified agents. Secondly, rather than detect the short lived,
diffusible NO molecule, as in the case of 6,29 4a is directly metabolized by the enzyme and
potentially may respond to the level of expression of NOS. Based on analogy with previously
described cell wall permeable complexes of general structure 5 (Scheme 2),28 the Lꢀarginineꢀ
decorated CA 8a is also anticipated to be internalized within the cell compartment. Taken
altogether, Dy3+ DOTAMꢀGlyꢀArgꢀOH (8a) may find use for imaging NOS enzymatic activity
in cell culture and animal models overexpressing NOS. However, before this can be done, more
quantitative evaluation including the rate and extent of the metabolism of the agent by the
various isoforms of NOS is required.
Acknowledgements
We thank The University of Western Ontario (The Major Academic Development Fund Grant
23

Page 24 of 27
for RHEH) and The Ontario Institute for Cancer Research (OICR) and The Natural Sciences and
Engineering Research Council of Canada (NSERC) for the financial support for this work. We
also gratefully acknowledge Prof. J. M. Chong of the University of Waterloo for assistance with
optical rotation measurements.
1For selected recent reviews on MRI CAs see: (a) Aime, S.; Botta, M.; Terreno, E. Adv. Inorg
Chem., 2005, 57, 173. doi: 10.1016/S0898ꢀ8838(05)57004ꢀ1. (b) Yoo, B.; Pagel, M. D. Front.
Biosci., 2008, 13, 1733. doi: 10.2741/2796. (c) Geraldes, C. F. G. C.; Laurent, S. Contrast Media
Mol. Imaging 2009, 4, 1. doi: 10.1002/cmmi.265.
2 Sherry, A. D.; Wu, Y. Curr. Opin. Chem. Biol., 2013, 17, 167. doi:
10.1016/j.cbpa.2012.12.012.
3
Ward, K. M.; Aletras, A. H.; Balaban, R. S. J. Magn. Reson., 2000, 143, 79. doi:
10.1006/jmre.1999.1956.
4 Zhang, S.; Merrit, M.; Woessner, D. E.; Lenkinski, R. E.; Sherry, A. D. Acc. Chem. Res., 2003,
36, 783. doi: 10.1021/ar020228m.
5 (a) Sherry, A.D.; Woods, M. Annu. Rev. Biomed. Eng., 2008, 10, 391. doi:
10.1146/annurev.bioeng.9.060906.151929. (b) Hancu, I.; Dixon, W.T.; Woods, M.; Vinogradov,
E.; Sherry, A. D.; Lenkinski, R. E. Acta Radiol., 2010, 8, 910. doi:
10.3109/02841851.2010.502126.
6 For an overview of the current state of development of classes of responsive MRI contrast
agents see: Hingorani, D. V.; Bernstein, A.S.; Pagel, M. D. Contrast Media Mol. Imaging,
2015, 10, 245. doi: 10.1002/cmmi.1629.
7 Yoo, B.; Pagel, M. D. J. Am. Chem. Soc., 2006, 128, 14032. doi: 10.1021/ja063874f.
24

Page 25 of 27
8 Suchý, M.; Ta, R.; Li, A. X.; Wojciechowski, F.; Pasternak, S. H.; Bartha, R.; Hudson, R. H. E.
Org. Biomol. Chem., 2010, 8, 2560. doi: 10.1039/b926639a.
9
Li, Y.; Sheth, V. R.; Liu, G.; Pagel, M. D. Contrast Media Mol. Imaging, 2011, 6, 219. doi:
10.1002/cmmi.439.
10 Yoo, B.; Sheth, V. R.; Howison, C. M.; Douglas, M. J. K.; Pineda, C. T.; Maine, A. A.; Baker,
A. F.; Pagel, M. D. Magn. Reson. Med., 2014, 71, 1221. doi: 10.1002/mrm.24763.
11
Chauvin, T.; Durand, P.; Bernier, M.; Meudal, H.; Doan, B. T.; Noury, F.; Badet, B.; Beloeil,
J. C.; Tóth, É. Angew. Chem., Int. Ed., 2008, 47, 4370. doi: 10.1002/anie.200800809.
12 Hingorani, D.V.; Randtke, E. A.; Pagel, M. D. J. Am. Chem. Soc., 2013, 135, 6396. doi:
10.1021/ja400254e.
13
Wojciechowski, F.; Suchý, M.; Li, A. X.; Azab, H. A.; Bartha, R.; Hudson, R. H. E.
Bioconjugate Chem. 2007, 18, 1625. doi: 10.1021/bc0701287.
14 Suchý, M.; Li, A. X.; Bartha, R.; Hudson, R. H. E. Bioorg. Med. Chem. 2008, 16, 6156. doi:
10.1016/j.bmc.2008.04.038.
15
Li, A. X.; Wojciechowski, F.; Suchy, M.; Hudson, R. H. E.; Menon, R.; Bartha, R. Magn.
Reson. Med., 2008, 59, 374. doi: 10.1002/mrm.21482.
16 McVicar, N.; Li, A. X.; Suchý, M.; Hudson, R. H. E.; Menon, R. S.; Bartha, R. Magn. Reson.
Med., 2013, 70, 1016. doi: 10.1002/mrm.24539.
17 Suchý, M.; Hudson, R. H. E. Synlett, 2012, 23, 893. doi: 10.1055/sꢀ0031ꢀ1290616.
18
Suchý, M.; Li, A. X.; Milne, M.; Bartha, R.; Hudson, R. H. E. Contrast Media Mol. Imaging
2012, 7, 441. doi: 10.1002/cmmi.1461.
19 For example: Viswanathan, S.; Ratnakar, S. J.; Green, K. N.; Kovacs, Z.; De LeonꢀRodriguez,
L. M.; Sherry, A. D. Angew. Chem., Int. Ed., 2009, 48, 9330. doi: 10.1002/anie.200904649.
25

Page 26 of 27
20For the recent reviews on PARACEST MRI CAs see (a) Terreno, E.; Delli Castelli, D.; Aime,
S. Contrast Media Mol. Imaging, 2010, 5, 78. doi: 10.1002/cmmi.369. (b) Viswanathan, S.;
Kovacs, Z.; Green, K. N.; Ratnakar, S. J.; Sherry, A. D. Chem. Rev., 2010, 110, 2960. doi:
10.1021/cr900284a.
21For the recent series of tutorial reviews on NO release, measurement and therapeutic
applications see: (a) Riccio, D. A.; Schoenfisch, M. H. Chem. Soc. Rev., 2012, 41, 3731. doi:
10.1039/c2cs15272j. (b) Carpenter, A. W.; Schoenfisch, M. H. Chem. Soc. Rev., 2012, 41, 3742.
doi: 10.1039/c2cs15273h. (c) Coneski, P. N.; Schoenfisch, M. H. Chem. Soc. Rev., 2012, 41,
3753. doi: 10.1039/c2cs15271a.
22
Kervin, Jr., J. F.; Lancaster, Jr., J. R.; Feldman, P. L. J. Med. Chem., 1995, 38, 4343. doi:
10.1021/jm00022a001.
23 Review of the use of fluorescence probes for detection of reactive nitrogen species: Gomes,
A.; Fernandes, E.; Lima, J.L.F.C. J. Fluoresc., 2006, 16, 119. doi: 10.1007/s10895ꢀ005ꢀ0030ꢀ3.
24 Recent examples of near IR fluorescence detection of NO: (a) Liu, W.; Fan, C.; Sun, R.; Xu,
Y.ꢀJ.; Ge, J.ꢀF. Org. Biomol. Chem., 2015, 13, 4532. doi: 10.1039/C5OB00042D. (b) Chen, J.ꢀ
B.; Zhang, H.ꢀX.; Guo, X.ꢀF.; Wang, H.; Zhang, H.ꢀS. Anal. Bioanal. Chem., 2013, 405, 7447.
doi: 10.1007/s00216ꢀ013ꢀ7177ꢀ6.
25(a) Kumar, P.; Kalita, A.; Mondal, B. Dalton Trans. 2012, 41, 10543. doi:
10.1039/c2dt31068f. (b) Yu, H.; Jin, L.; Dai, Y.; Li, H.; Xiao, Y. New J. Chem., 2013, 37, 1688.
doi: 10.1039/c3nj41127c.
26
RiveraꢀGil, P.; VazquezꢀVazquez, C.; Giannini, V.; Pilar Callao, M.; Parak, W. J.; Correaꢀ
Duarte, M. A.; AlvarezꢀPuebla, R. A. Angew. Chem., Int. Ed., 2013, 52, 13694. doi:
10.1002/anie.201306390.
26

Page 27 of 27
27 Zhang, J.; Xu, M.; Dence, C. S.; Sherman, E. L. C.; McCarthy, T.; Welch, M. J. J. Nucl. Med.,
1997, 38, 1273.
28(a) Oliveira, B. L.; Raposinho, P. D.; Mendes, F.; Figueira, F.; Santos, I.; Ferreira, A.;
Cordeiro, C.; Freire, A. P.; Correia, J. D. G. Bioconjugate Chem., 2010, 21, 2168. doi:
10.1021/bc100291e. (b) Oliveira, B. L.; Raposinho, P. D.; Mendes, F.; Santos, I. C.; Santos, I.;
Ferreira, A. Cordeiro, C.; Freire, A. P.; Correia, J. D. G. J. Organometal. Chem., 2011, 696,
1057. doi: 10.1016/j.jorganchem.2010.09.019.
29 Liu, G.; Li, Y.; Pagel, M. D. Magn. Reson. Med., 2007, 58, 1249. doi: 10.1002/mrm.21428.
30
Suchý, M.; Li, A. X.; Bartha, R.; Hudson, R. H. E. Bioorg. Med. Chem., 2008, 16, 6156. doi:
10.1016/j.bmc.2008.04.038.
31For the recent review on DIACEST MRI see (a) Liu, G. ; Song, X.; Chan, K. W. Y.;
McMahon, M. NMR Biomed., 2013, 26, 810. doi: 10.1002/nbm.2899. For the recent example
from our laboratory see (b) McVicar, N.; Li, A. X.; Gonçalves, D. F.; Bellyou, M.; Meakin, S.
O.; Prado, M. A.; Bartha, R. J. Cereb. Blood Flow Metab., 2014, 34, 690. doi:
10.1038/jcbfm.2014.12.
32 Sharma, B. V.; Rowland, N. S.; Clouse, M. M.; Rice, N. A. Adv. Biol. Chem., 2014, 4, 214.
doi: 10.4236/abc.2014.43026.
33
Barge, A.; Cravotto, G.; Gianolio, E.; Fedeli, F. Contrast Media Mol. Imaging, 2006, 1, 184.
doi: 10.1002/cmmi.110
.
27