Synthesis and Preclinical Characterization of a Cationic Iodinated Imaging Contrast Agent (CA4+) and Its Use for Quantitative Computed Tomography of Ex Vivo Human Hip Cartilage

Article abstract of DOI:10.1021/acs.jmedchem.7b00234

Contrast agents that go beyond qualitative visualization and enable quantitative assessments of functional tissue performance represent the next generation of clinically useful imaging tools. An optimized and efficient large-scale synthesis of a cationic iodinated contrast agent (CA4+) is described for imaging articular cartilage. Contrast-enhanced CT (CECT) using CA4+ reveals significantly greater agent uptake of CA4+ in articular cartilage compared to that of similar anionic or nonionic agents, and CA4+ uptake follows Donnan equilibrium theory. The CA4+ CECT attenuation obtained from imaging ex vivo human hip cartilage correlates with the glycosaminoglycan content, equilibrium modulus, and coefficient of friction, which are key indicators of cartilage functional performance and osteoarthritis stage. Finally, preliminary toxicity studies in a rat model show no adverse events, and a pharmacokinetics study documents a peak plasma concentration 30 min after dosing, with the agent no longer present in vivo at 96 h via excretion in the urine.

Full text of DOI:10.1021/acs.jmedchem.7b00234

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Synthesis and Pre-Clinical Characterization of a Cationic Iodinated
Imaging Contrast Agent (CA4+) and Its Use for Quantitative
Computed Tomography of Ex Vivo Human Hip Cartilage
Rachel C. Stewart, Amit N. Patwa, Hrvoje Lusic, Jonathan D. Freedman,
Michel Wathier, Brian D. Snyder, Ali Guermazi, and Mark W. Grinstaff
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 15 Jun 2017
Downloaded from http://pubs.acs.org on June 15, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis and PreꢀClinical Characterization of a
Cationic Iodinated Imaging Contrast Agent (CA4+)
and Its Use for Quantitative Computed Tomography
of Ex Vivo Human Hip Cartilage
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Rachel C. Stewart,†,‡ Amit N. Patwa,§ Hrvoje Lusic,§ Jonathan D. Freedman,‡,§ Michel Wathier,§
Brian D. Snyder,*,‡ Ali Guermazi,*,¶ and Mark W. Grinstaff*,†,§
†Departments of Biomedical Engineering, Chemistry, and Medicine, Boston University, 590
Commonwealth Ave., Boston, Massachusetts 02215, United States ‡Center for Advanced
Orthopaedic Studies, Beth Israel Deaconess Medical Center and Harvard Medical School, 1
Overland Street, RN 115, Boston, Massachusetts 02215, United States §Ionic Pharmaceuticals,
Boston, Massachusetts 02445, United States ¶Department of Radiology, Boston University
School of Medicine, Boston, Massachusetts 02118, United States
Keywords
Computed tomography, iodinated contrast agent, cationic contrast agent, compressive modulus,
osteoarthritis, glycosaminoglycan
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 47
1
2
3
4
Abstract
5
6
7
8
9
Contrast agents that go beyond qualitative visualization and enable quantitative assessments of
functional tissue performance represent the next generation of clinically useful imaging tools. An
optimized and efficient largeꢀscale synthesis of the cationic iodinated contrast agent (CA4+) is
described for imaging articular cartilage. Contrast enhanced CT (CECT) using CA4+ reveals
significantly greater agent uptake of CA4+ in articular cartilage compared to similar anionic or
nonionic agents, and CA4+ uptake follows Donnan Equilibrium. The CA4+ CECT attenuation
obtained from imaging ex vivo human hip cartilage correlates with the glycosaminoglycan
(GAG) content, equilibrium modulus, and coefficient of friction: key indicators of cartilage
functional performance and osteoarthritis stage. Finally, preliminary toxicity studies in a rat
model show no adverse events and a pharmacokinetics study documents a peak plasma
concentration at 30 minutes after dosing with the agent no longer present in vivo at 96 hours via
excretion in the urine.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Introduction
Xꢀray computed tomography (CT) is a wellꢀestablished imaging modality that captures
virtual slices of biological structures nonꢀinvasively, and is commonly employed to diagnose
disease and trauma, to guide interventional or therapeutic procedures, and to monitor the
effectiveness of therapy. The quality of CT images for diagnostic purposes largely depends on
differences in the Xꢀray attenuation of the tissue of interest relative to its surrounding
environment – known as image contrast. CT provides the best contrast between neighboring
tissues that have different densities and composition. CT is less effective at distinguishing
between contiguous soft tissues (for example, synovial fluid and articular cartilage), both of
ACS Paragon Plus Environment
2

Page 3 of 47
Journal of Medicinal Chemistry
1
2
3
4
which are composed largely of water. When
5
6
7
8
natural image contrast is insufficient for clear
differentiation of tissues, imaging agents are
employed to afford a temporary enhancement
of the image quality.1–9
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Several FDA approved contrast agents
are iodinated waterꢀsoluble anionic or
nonionic (Figure 1) small molecules (<2000
g/mol),10 which are often used for imaging of
the vasculature. Firstꢀgeneration iodinated
contrast agents possessed anionic polar
groups to increase solubility, but ionic
moieties are prone to interactions with cell
membranes and other biological structures as
Figure 1.
Chemical structures of the
cationic (CA4+) as well as commercially
available negatively charged (ioxaglate) and
neutral (iodixanol) contrast agents.
well as possess intrinsically high osmolality.11,12 In order to overcome these issues, secondꢀ
generation contrast agents employed hydroxyl and amide functional groups, that readily
hydrogen bond, to promote water solubility. Additionally, multiple iodinated aromatic rings are
fused together in order to reduce osmolality while maintaining high administered iodine content.
Though several contrast agents (ionic, nonionic and one/two aromatic ring structures) are
approved for clinical use and produced worldwide, new contrast agents with high water
solubility, low viscosity and osmolality, and reduced toxicity continue to be of interest.13–29
Contrast agents that go beyond qualitative visualization and enable quantitative assessments of
functional tissue performance represent the next, thirdꢀgeneration of contrast agents. One clinical
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 47
1
2
3
4
5
6
area where such contrast agents can have a significant impact is in the diagnosis of soft tissueꢀ
based diseases like osteoarthritis (OA).
7
8
9
OA affects over 27 million Americans and this patient population is projected to grow to
over 40 million adults in the US by the year 2030.30–32 Especially prevalent in the knee, hip, and
hand, OA produces pain, lifestyle limitations, and significant loss of productivity in its later
stages. Unfortunately, clinicians are not able to detect the disease in its earliest stages, when the
affected cartilage tissue may still be salvageable. Currently, OA is diagnosed in the clinic based
on symptoms such as pain, swelling, and abnormal joint function including stiffness, locking,
and reduced mobility. To assist in diagnosis, plain radiographs (twoꢀdimensional projection
images) of the affected joints are obtained to evaluate structural changes such as joint space
narrowing, the presence of osteophytes, or that of subchondral bone sclerosis.33–36
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
These symptoms and radiographic findings appear late in the disease, which leaves
clinicians with few treatment options. Unfortunately, there is currently no definitive treatment for
OA, other than joint replacement surgery, but early diagnosis may provide opportunities for
interventions that can ameliorate disease progression rather than treating the lateꢀstage
symptoms. A recognized biomarker for early OA is loss of glycosaminoglycans (GAGs) in the
cartilage extracellular matrix (ECM).37–45 The GAGs impart an anionic fixed charge density,
which contributes to maintaining fluid pressurization, tissue compressive stiffness, and lubricity.
Thus, methods that can detect and quantify GAG concentration, distribution, and changes thereof
in cartilage, are in immediate need.
Today, commercially available anionic contrast agents (e.g., iothalamate,46 ioxaglate47),
originally developed for cardiovascular applications, are being investigated for imaging cartilage
tissue. These anionic contrast agents partition into cartilage in inverse proportion to GAG content
ACS Paragon Plus Environment
4

Page 5 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
due to electrostatic repulsion between the negatively charged GAGs and the contrast agent. The
disadvantages of using an anionic probe to detect negatively charged GAGs include requiring
high concentrations to be applied in order to produce sufficient contrastꢀtoꢀnoise ratios for
quantitative imaging and weak to moderate correlations between CT attenuation and GAG
content. Recently, we reported cationic contrast agents for imaging of ex vivo41,48 and in vivo49
cartilage. These cationic agents perform superior to anionic agents, and show high sensitivity for
quantifying GAG content41,48 as well as equilibrium compressive modulus and coefficients of
friction50 in ex vivo bovine articular cartilage. Our approach belongs to a larger strategy of using
electrostatic interactions to detect proteins, polysaccharides, and nucleic acid,51 to create selfꢀ
healing polymer networks,52,53 or to assemble supramolecular assemblies of nonꢀclinical
interest.54–56
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Herein, we report the synthesis, characterization, and analysis of the hexaꢀiodinated
cationic contrast agent (CA4+) for sensitive and quantitative ex vivo imaging of human hip
cartilage, and key initial data required for subsequent translation of this agent to the clinic.
Specifically, we describe the optimized largeꢀscale synthesis and characterization of CA4+.
Next, the uptake of the cationic contrast agent is evaluated in bovine articular cartilage in
comparison to commercially available anionic and neutral iodinated contrast agents. Donnan
equilibrium modeling is then applied to the cationic contrast agents to describe their mode of
uptake into articular cartilage. Subsequently, CA4+ is evaluated for assessment of GAG content,
stiffness, and frictional performance in human cadaveric hip articular cartilage. Finally, results
from a preliminary toxicity study of CA4+ in a rat model are described along with results from a
pharmacokinetics study using the 14C radio labeled CA4+.
Results & Discussion
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 47
1
2
3
4
Large scale synthesis and chemical characterization of the CA4+ cationic iodinated contrast
5
6
agent
7
8
9
Establishment of an efficient and reliable largeꢀscale synthesis of CA4+ is required for
preꢀclinical development and further evaluation of CA4+ for additional ex vivo and in vivo
imaging studies. Previously, we reported the small scale (<5 g starting material) synthesis of
CA4+,45 but not the large scale (>150 g starting material) synthesis. The procedure for the largeꢀ
scale synthesis of CA4+ is shown in Scheme 1. The contrast agent CA4+ 5 was synthesized
from commercially available 5ꢀaminoꢀ2,4,6ꢀtriiodisophthalic acid 1 in five steps in
approximately 68% overall yield with a straightforward workup involving washing, precipitation
and filtration, without requiring column chromatography purification. Each reaction step was
repeated at least three times and yields are reported as Average ± SD (N ≥ 3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
First, 5ꢀaminoꢀ2,4,6ꢀtriiodisophthaloyl chloride 2 was synthesized by refluxing the
mixture of 1 in thionyl chloride (19 equiv.) and catalytic dimethyl formamide (DMF) for 6ꢀ7
hours. The reaction mixture was then precipitated in iceꢀcold water and filtered to obtain a
yellow solid. This solid was then dissolved in ethyl acetate, washed (basic and brine wash),
concentrated to a minimum amount (not till dryness) and filtered again to obtain 2 as light yellow
powder in 94.2 ± 1.4 % yield. The solid obtained was sufficiently pure (single spot on TLC) to
carry out the next step without further purification. This procedure represents an improvement
over the previously reported procedure.45 Specifically, the quantity of thionyl chloride, used as
solvent and reactant, was reduced from 40 to 19 equivalents. DMF was used as catalyst and the
reaction time was reduced from overnight to 6ꢀ7 h. We also explored an alternative procedure to
minimize the usage of thionyl chloride, where compound 2 was synthesized using only four
ACS Paragon Plus Environment
6

Page 7 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Large scale synthesis of the chloride salt of CA4+ 5.
equivalents of thionyl choride in the solvent ethylene dichloridewith (see SI for details).
Although we successfully synthesized 2 using both procedures, the workup is more convenient
and simple using thionyl chloride as the reagent and solvent.
Next, 2 was coupled with malonyl chloride (exactly 0.5 equiv.) in anhydrous THF (fresh
bottle and transferred through cannula) at 50 °C for overnight. The reaction mixture was cooled
to room temperature and then left for precipitation at −20 °C overnight. The
tetrakis(acylchloride) intermediate 3 was obtained as a white solid by filtration followed by a
quick wash with cold THF (cooled +4 °C for 30ꢀ40 minutes prior to use). The mother liquor was
concentrated, left for reꢀprecipitation at −20 °C and filtered. Use of dry THF, exactly 0.5 equiv.
of malonyl chloride, and maintaining the reaction under an inert atmosphere were key parameters
to obtain a high yield and single spot by TLC (89.4 ± 2.7% yield). The tetrakis(Bocꢀprotected)
precursor 4 was then synthesized by coupling 3 with tertꢀbutylꢀNꢀ(2ꢀaminoethyl)carbamate in
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 47
1
2
3
4
5
6
7
8
9
anhydrous dimethylacetamide as a solvent and diisoproplyethylamine as an HCl scavenger,
generated during the course of the reaction. After overnight stirring at 50 °C, the reaction
mixture was precipitated in 1 mol/L aqueous HCl, filtered and the resulting white solid cake was
washed with water thrice to remove any trace solvents. After airꢀdrying for 2ꢀ3 hours (in the
filtration funnel with continuous vacuum) the cake was further dried by lyophilization to obtain 4
as a white powder in 89.7 ± 2.2% yield. The Bocꢀgroup deprotection was then carried out by
dissolving 4 in dichloromethane:trifluoroacetic acid (1:1). After 2 hours, the reaction mixture
was precipitated in diethyl ether and filtered to afford the trifluoroacetic acid salt of CA4+
(CA4+ TFA salt) 5a as a white powder. After airꢀdrying for 2ꢀ3 hours the white powder was
further dried under high vacuum overnight to remove trace diethyl ether. The trifluoroacetate
counter ion was exchanged with chloride ion (Cl¯) by repeated dissolution in 3 mol/L aqueous
hydrochloric acid (2X) and precipitation in acetone. The crude product was dissolved in
minimum quantity of 0.1 mol/L aqueous HCl and precipitated by drop wise addition into a large
excess of acetone (constant stirring) to yield the chloride salt of CA4+ 5 as a white powder in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
high yield (92.3 ± 3.2% yield, two steps) and characterized by H NMR (Figure S9). The
complete exchange of trifluoroacetate ion with chloride ion was confirmed by 19F NMR
spectrometry. A sharp single peak was observed at −73.73 ppm in the 19F NMR spectra for
CA4+ TFA salt (Figure S7), 5a, corresponding to the trifluoroacetate ion, while no peak was
observed in 19F NMR spectra for chloride salt of CA4+ 5 (Figure S11). All the compounds were
1
characterized by H, 13C NMR (Figure S1-S11) and highꢀresolution mass spectroscopy. The
purity of the CA4+ 5 was determined by analytical reverseꢀphase HPLC with 1 mL/min flow rate
of 95/5 water/ACN (isocratic) and found to be >98%. CA4+ 5 was detected as a single peak at
245 nm with a retention time of 2.5 minutes (Figure S12). HRꢀMS (ESIꢀTOF) for C27H33I6N10O6
ACS Paragon Plus Environment
8

Page 9 of 47
Journal of Medicinal Chemistry
1
2
[M+H]+ was observed at 1354.7031 (calculated 1354.6847 [M+H]+) (Figure S13). The overall
improvements in the synthetic procedure are in time, amounts of reactants, purification, and
yield, and these benefits are summarized in Table S1. Using the above procedure, we have
synthesized more than 1 kg of CA4+ 5 during the last year for use in ex vivo and in vivo cartilage
imaging experiments.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The shelf life of CA4+ 5 as a powder at room temperature and in aqueous solution at 4 ºC
(12 or 24 mgI/mL, pH=7.4, 400 mOsm/kg) was determined by HPLC after one year. The HPLC
trace for both samples after one year was identical, with a single peak at the same retention time,
to that of the HPLC trace after synthesis. Thus, CA4+ 5 as a powder or in aqueous 4 ºC can be
stored for one year, prior to use.
Uptake of the cationic contrast agent (CA4+) into articular cartilage
Diffusion analysis of the positively charged contrast agent (CA4+) over time into bovine
cartilage was carried out and compared with the diffusion of commercially available negatively
charged (ioxaglate) and nonionic (iodixanol) contrast agents (Figure 1). The cationic, anionic
and nonionic contrast agents were evaluated over a period of 48 hours using highꢀresolution
micro computed tomography (µCT). The equilibrium time (time to reach 95% of the maximum
CECT attenuation) was calculated from Equation 1, where α is the CECT attenuation of the
cartilage at diffusion equilibrium and τ is the diffusion time constant (the time required to reach
63.2% of the maximum attenuation, α).
ꢍ
ꢎ
ꢌ
ꢀꢁꢀꢂ ꢃꢄꢄꢅꢆꢇꢈꢄꢉꢊꢆ = ꢋ(1 − e ) (Equation 1).
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 47
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Diffusion trajectories of cationic (CA4+), anionic (ioxaglate), and nonionic
(iodixanol) contrast agents in bovine articular cartilage. (A) Average CECT attenuation
values (reported in Hounsfield Units, HU) over the 48ꢀhour time course of diffusion into
the cartilage of bovine osteochondral plugs. Each point is reported as the average of six
samples (mean ± standard deviation). Each contrast agent was fit to a nonlinear diffusion
model (Equation 1). The arrow marks the time when 95% of the equilibrium attenuation
is reached. (B) Equilibrium CECT attenuation (normalized by concentration) for each
contrast agent in bovine osteochondral plugs (*p<0.0001).
The diffusion time course (Figure 2(A)) and uptake percentage (Figure 2(B)) for each contrast
agent were compared. The time required for CA4+ to reach diffusion equilibrium was
substantially longer than for ioxaglate or iodixanol. CA4+ reached equilibrium (95% of the
maximum CECT attenuation) in 37.5 hours compared to 11.7 hours for iodixanol and 9.1 hours
for ioxaglate (Table S2 & S3). Additionally, CA4+ was more highly taken up by cartilage
(~400% of the exposed concentration relative to < 100% for ioxaglate and iodxanol), as might be
expected, given that CA4+ is electrostatically attracted to the cartilage’s anionic fixed charge
density. For the charged contrast agents, equilibrium distribution corresponded to the location
and density of the anionic GAGs (Figures 3 and S14). Ioxaglate, with an overall −1 charge,
distributed inversely to GAG content. Since the highest concentration of GAGs are present in the
deep cartilage layer, attenuation due to ioxaglate penetration was mainly seen in the superficial
and middle zones. Iodixanol, which is nonionic, also distributed primarily throughout the middle
ACS Paragon Plus Environment
10

Page 11 of 47
Journal of Medicinal Chemistry
1
2
3
4
and superficial zone of cartilage, with more
5
6
7
8
agent in the superficial (likely reflecting its
higher free water content). The equilibrium
contrast agent distribution of CA4+ was
markedly different, distributing directly
proportionally to GAG content—primarily in
the deep zone of cartilage. This partitioning
difference likely explains the longer time
required to reach diffusion equilibrium, since
CA4+ diffusion proceeds throughout the full
depth of the cartilage, including into the deep
layer.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. µCT images of a bovine
osteochondral plug visualized in color maps.
All images are taken of the same plug cross
section. The color map ranges from air in red
(each image bottom), to cartilage in yellow,
green and blue, & bone in dark blue and
purple (each image top). The contrast agents
diffuse into cartilage over time and the color
turns from yellow to green to blue indicating
higher CT attenuation.
Donnan equilibrium modeling of cationic
contrast agent uptake in cartilage
The uptake of cationic molecules in
articular cartilage could be mediated by
binding interactions, nonspecific electrostatic
interactions (characterized by Donnan equilibrium theory), or both. In order to elucidate the
mechanism of CA4+ uptake in articular cartilage, two groups of bovine cartilage plugs (N=3
each) were serially equilibrated in escalating concentrations of CA4+ and imaged using µCT.
Donnan equilibrium theory was used to calculate the partitioning of sodium ion given the
measured partitioning of the cationic contrast agent, and the results were plotted and fit with
nonlinear least squares (Figure 4). Donnan equilibrium theory predicts that the partition
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 47
1
2
coefficients (KCA) for CA4+ will follow the partition coefficient for Na+ (KNa+) via the power
law relation shown in Equations 2 & 5 (see experimental section, where Z is the total charge on
the contrast agent). The nonlinear leastꢀsquares fits for the partition coefficient of cations in
bovine cartilage plugs predicted Z = 3.71 for CA4+ (R2 = 0.55, and R2 = 0.95, respectively
(Figure 4). The results indicate that CA4+ contrast uptake followed closely with the Donnan
equilibrium theory prediction of a molecule
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
bearing +3.7 charge, which is close to the
formal charge of +4.
Cationic
contrast
of
agents
provide
and
quantification
biochemical
biomechanical measures of human cartilage
quality
Figure 4. Donnan equilibrium modelling
showing the partition coefficient of CA4+ as a
function of sodium ion partition coefficient.
Nonlinear leastꢀsquares best fits suggest a
valence (Z) on the CA4+ as Z = 3.71,
compared to the theoretical value (Z = 4).
To assess whether CECT using CA4+
produces CECT attenuations that correlate
with GAG content, compressive stiffness, and coefficients of friction of human ex vivo cartilage,
twelve osteochondral plugs were extracted from the femoral heads of cadaveric human hips. The
plugs were assessed for compressive stiffness using a 4ꢀstep stressꢀrelaxation regimen to
determine equilibrium compressive modulus, and torsional coefficients of friction were
measured using a rotational plugꢀonꢀplaten regimen. Subsequently, each plug was serially
diffused to equilibrium in ioxaglate or CA4+ (24 mgI/mL) at concentrations chosen to produce
similar tissue contrast (i.e., ioxaglate at ~6x the concentration of CA4+; 144 mgI/mL) and
imaged using µCT. Finally, the GAG content of the cartilage plugs was determined following
ACS Paragon Plus Environment
12

Page 13 of 47
Journal of Medicinal Chemistry
1
2
our published procedures.49 Previous reports showed that lower concentrations of ioxaglate
(<100 mgI/mL) provide poorer correlations between CECT attenuations and GAG, hence we
used a high concentration for this study.38,39
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Comparing CECT attenuation to biochemical content, 79% of the variation in µCT
attenuation was attributable to variations in GAG content (R2 = 0.79, p < 0.001) for ioxaglate and
89% of the variation in µCT was attributable to variations in GAG content (R2 = 0.89, p < 0.001)
for CA4+ (Figure 5(A)). Relating the CECT data to tissue stiffness, 82% of the variation in µCT
attenuation was attributable to variations in equilibrium compressive modulus (R2 = 0.82, p <
(A)
(B)
Figure 5. CECT attenuation correlates strongly with GAG content (A) and equilibrium
compressive modulus (B) using both CA4+ and ioxaglate.
0.001) for ioxaglate and 54% of the variation was attributable to variations in modulus (R2 =
0.54, p = 0.01) for CA4+ (Figure 5(B)). CECT attenuation correlated less strongly, and not
always statistically significantly, to coefficients of friction µkinetic, µstatic, and µstatic_equilibrium (R2 =
0.31, 0.44, and 0.19 respectively for ioxaglate, Figure 6(A); R2 = 0.48, 0.54, and 0.31
respectively for CA4+, Figure 6(B)). The coefficients of determination for coefficients of
friction are weaker in this experiment than those previously reported,50 perhaps reflecting the
heterogeneity of the human articular cartilage used compared to young bovine cartilage.
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 47
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(A)
(B)
Figure 6. CECT attenuation correlates moderately with coefficients of friction for ioxaglate
(A) and CA4+ (B).
Given these results, both contrast agents, at the concentrations used here, reflect the GAG
content, equilibrium compressive modulus, and to a more limited extent coefficients of friction,
using high resolution imaging at diffusion equilibrium. Importantly, a lower administered dose of
CA4+, compared to ioxaglate, can be used. These data are similar to a previous study using
bovine osteochondral plugs, and support those findings while extending the proofꢀofꢀconcept to
human cartilage tissue.50 The correlation between attenuation and E is weaker than reported for
young bovine cartilage and this may reflect the tissue type, and additional studies with human
cartilage tissues are needed. These results are important for establishing the feasibility of using
CECT as a clinical tool, since it explores CECT for quantifying cartilage functional performance
in human cartilage, which can be different in several ways compared to bovine.33,57–59
The next step in the development of this CECT technique will be the imaging of an intact
ex vivo human hip in a clinical scanner. CECT imaging of the hip would be valuable in a clinical
setting, since MRI of the hip has unique shortcomings such as magnetic field inhomogeneity in
the pelvic region due to large and variable body habitus.60,61 Beforehand, a number of
considerations will need to be addressed. First, the maximum spatial resolution of current clinical
ACS Paragon Plus Environment
14

Page 15 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CT instruments is relatively low compared to µCT scanners. For the thin cartilage (< 1.5 mm) in
the human femoral head, the cartilage may be difficult to distinguish from the surroundings and
both partialꢀvolume artifacts will likely have a larger impact on obtaining quantitative data.
Second, although femoral head cartilage is thin, it is not clear how much diffusion time will be
sufficient to reach diffusion equilibrium. This needs to be more thoroughly investigated, as does
the impact of incomplete diffusion on contrast agents’ ability to assess cartilage health and
functional performance. Recent evidence suggests that for ioxaglate, preꢀequilibrium contrast
agent partitioning can provide quantitative measures of cartilage quality, but the same has not
been evaluated for CA4+.62 Importantly as well, the human hip contains very tightly conforming
cartilage surfaces, which may impede and bias the penetration of contrast agents injected intraꢀ
articularly in vivo. Thus, optimization of the administration procedure will be required.
Assessment of CA4+ safety, pharmacokinetics, excretion, and tissue distribution in a rat model
A maximum tolerated dose study was performed using N=4 groups of n=6 Spragueꢀ
Dawley rats each (3 male, 3 female). Escalating volumetric doses of 24 mgI/mL 14Cꢀlabeled
CA4+ from 50ꢀ200 µL in 50 µL increments were administered via intraꢀarticular injection. The
animals showed minimal weight loss (from 0.25% to 4.7%, 1.8 ± 1.4% (mean ± standard
deviation)) from the initial body weight. No adverse observations were noted at necropsy, and
terminal complete blood count (CBC) found no doseꢀrelated abnormal changes. At euthanasia
(day 8) nearly all of the animals had high potassium, total protein, albumin, and globulin levels
(Table S4) compared to controls/normal values (Table S6 & S7). Male rats were also found to
have high phosphorus, glucose, and alkaline phosphatase levels (Table S5) compared to normal
levels (Table S6 & S7). No clinical chemistry changes were proportional to dosage given. Also,
no animal died prior to scheduled necropsy.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 47
1
2
3
4
5
6
7
8
9
Nine additional animals were used for toxicokinetic analysis, where animals were
administered 100 µL of the agent at 2X concentration intraꢀarticularly in one randomly chosen
rat stifle joint. Blood was drawn at four postꢀdose time points per animal range from 5 min to
144 hr. Body weights decreased in most of the animals (from 1.85% to 8.99%, avg 3.79 ± 2.93%
body weight loss from initial preꢀdose weight) while two animals gained weight. No adverse
clinical observations nor abnormal findings were noted during gross necropsy.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the pharmacokinetics and tissue distribution study, a radio labeled CA4+ (14Cꢀlabeled
CA4+, 5*) was synthesized by coupling two 5ꢀaminoꢀ2,4,6ꢀtriiodisophthaloyl moieties with 14Cꢀ
labeled malonyl chloride and following the same synthetic steps described in Scheme 1. The
radiochemical purity for the 14Cꢀlabeled CA4+ 5* (>99%) was determined by HPLC with 1
mL/min flow rate of 0.05 % TFA in water (isocratic). 14Cꢀlabeled CA4+ 5* (with specific
activity 5.82 ꢀCi/mg) was detected as a single peak with a retention time of 6.71 minutes (see
SI). Next, twenty additional rats (6 females and 14 males) were dosed intravenously or intraꢀ
articularly with 24 mgI/mL of CA4+. Two groups (Group 1 and Group 2) were dosed
intravenously with 100 ꢁL of 14Cꢀlabeled CA4+ 5* and one group (Group 3) was dosed intraꢀ
14
articularly in one randomly selected stifle joint with 50 ꢁL of Cꢀlabeled CA4+ 5*. There were
Figure 7. (A) IV and IA plasma concentrations (male and female). (B) Kidney, liver, and
lung concentrations (1, 8 and 24 h postꢀdose). (C) Urine and feces concentrations.
three females and three males (n = 6) in Group 1, eight males (n = 8) in Group 2, and three
ACS Paragon Plus Environment
16

Page 17 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
females and three males (n = 6) in Group 3. Approximately 1 hour postꢀdose, one animal from
Group 1 was dead. A gross necropsy was performed on the animal and no macroscopic
abnormalities were noted. Approximately 5 hours postꢀdose, all surviving animals were observed
to be normal. Clinical observations for all animals for the remainder of the study were normal.
Blood draws at preꢀdose, 0.5, 1, 3, 8, 48, and 96 hours postꢀdose revealed that plasma
concentrations for both IV and IA injections peaked at 0.5 h for both males and females, with
females having a higher observed plasma concentration (Figure 7(A)). At the 1 h time point,
plasma concentrations were still relatively high for both groups, with females having a higher
observed plasma concentration. Plasma concentrations were observed to decline rapidly for both
groups, females and males, from the 3 h time point to the 96 h time point. At the 96 h time point,
14C labeled CA4+ was almost completely eliminated from plasma for both groups, females and
males. For both the IV and IA administration, the Tmax was 0.5 h indicating peak plasma
concentrations (Cmax) occurred 0.5 h after dosing. At 96 h postꢀdose, the observed levels were
similar to the levels at the preꢀdose time point. The bioavailability of IA dosing in comparison to
IV bolus (100% bioavailability) was 58.01%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The results from the PK study show >95% plasma clearance in 24h, and yet CA4+ slowly
diffuses into the cartilage, as described above. The rate of plasma clearance and diffusion into
the cartilage are critical parameters for successful use of this technique. However, the uptake
study was done in bovine cartilage plugs while the PK study was done in rats. The uptake of
CA4+ in thin murine cartilage is significantly faster than thick bovine cartilage (1 hr to penetrate
fully, versus 24 hrs – unpublished results).
The concentration of agent was highest in the kidneys compared to the lungs and liver,
with radioactivity levels starting at 100 nCi/g at the 1 h time point and plateauing at
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 47
1
2
3
4
5
6
7
8
9
approximately 350 nCi/g at the 24 h time point (Figure 7(B)). Excreted agent was observed
primarily in urine, with radioactivity levels peaking at approximately 60 nCi/g at the 8 h time
point and declining to approximately 15 nCi/g at the 24 h time point (Figure 7(C)). The
combination of high kidney uptake and high urine activity suggests a primarily renal excretion
route for the agent. However, persistent high activity suggests that CA4+ is, to some extent,
being retained within kidney tissue, likely owing to its high (+4) positive charge. Whether this
has an influence on renal function or whether the CA4+ might be similarly retained in other
tissues with high membrane potential (e.g., myocardium) is the subject of future studies.
Additionally urine was not continuously collected and analyzed, and, thus, we were unable to
determine the amount of injected compound excreted in the urine. In contrast to renal excretion,
radioactivity in the feces was relatively low with approximately 0.08 nCi/g at the 8 h time point
and approximately 2.22 nCi/g at the 24 h time point (Figure 7(C)). The radioactivity was also
low for both liver and lung. Levels for the liver peak at approximately 6 nCi/g at the 1 h time
point, and plateau at approximately 5 nCi/g at the 8 and 24 h time point. Levels for the lung peak
at approximately 5 nCi/g at the 1 h time point, and declined to approximately 1 nCi/g at the 8 and
24 h time point.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusions
The synthesis of CA4+ is accomplished in excellent yield over 5 steps with high purity
(>98 %), and can be scaled to 100 grams in the laboratory. Studies elucidating the binding of
contrast agents in cartilage, including modeling of the mechanism of uptake, reveal that the
cationic agent CA4+ is highly taken up in articular cartilage compared to anionic and nonionic
contrast agents, and partitions throughout cartilage according to electrostatic interactions
ACS Paragon Plus Environment
18

Page 19 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
following Donnan equilibrium theory. However, the time to reach diffusion equilibrium for
CA4+ is slower compared to the nonionic and anionic contrast agents. CECT using the cationic
agent affords attenuations that are indicative of GAG content and compressive stiffness, and
moderately indicative of coefficients of friction. These data, along with promising indicators of
tolerability in vivo in a rat model, offer promise for the continued development of CECT using
CA4+ as a diagnostic imaging tool for early OA. Importantly as well, diagnosis and monitoring
of early OA may provide clinicians and researchers with a heretofore unavailable opportunity to
develop better therapies via an imaging technique that enables quantitative evaluation and serial
monitoring.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
Large scale synthesis and chemical characterization of cationic iodinated contrast agent
General experimental methods
Unless otherwise mentioned, all other chemicals were purchased from commercial sources
(SigmaꢀAldrich, Acros and AK Scientific Inc.) and used without further purification. Anhydrous
THF over molecular sieves and anhydrous DMA were purchased and transferred in the reaction
vessel through cannula. All reactions were performed under a nitrogen atmosphere and at room
temperature unless otherwise noted. All reactions were monitored by analytical thinꢀlayer
chromatography (TLC) on aluminium preꢀcoated plates of silica gel 60G F254 with detection by
viewing under UV light. NMR spectra were recorded by Agilent 500 MHz VNMRS
1
spectrometer with a Varian ultraꢀshielded magnet at 500 MHz for H, 470 MHz for 19F, 125
MHz for 13C. 1H NMR chemical shifts were assigned relative to the signal of a residual
protonated solvent: DMSOꢀd6
δ 2.50, CDCl3 δ
7.26. 13C NMR chemical shifts were assigned
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 47
1
2
3
4
relative to the signal of a residual protonated solvent: DMSO
δ 39.52, CDCl3 δ 77.16. Peak
5
multiplicities were denoted by s (singlet); bs (broad singlet), m (multiplet). High resolution mass
spectra were obtained on a Waters QTOF spectrometer. The yield recorded in each experiment
was the isolated yield. The purity of the final compound was determined to be ≥ 95% by
analytical high pressure liquid chromatography (HPLC).
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5-Amino-2,4,6-triiodoisophthaloyl chloride 2. The commercially available (AK Scientific
Inc. #E824 (previous #69824)) 5ꢀaminoꢀ2,4,6ꢀtriiodoisophthtalic acid (100 g, 178.94 mmol, 1
equiv.) was suspended in thionyl chloride (250 mL, 3.427 mol, 19.1 equvi.) in 1 L round bottom
flask to form grey insoluble mixture. Subsequently, DMF (0.5 mL, as a catalyst) was added. The
reaction mixture was refluxed for 6ꢀ7 h. As the reaction proceeded, the material solubilized to
produce a yellow (deep yellow/brown) soluble, albeit cloudy, solution. The reaction mixture was
cooled to RT, precipitated into iceꢀcold water (1500 mL). The aqueous mixture was stirred for
20ꢀ25 min and filtered and washed with water to obtaine yellow solid. The yellow solid was
transferred into a 2 L erlenmeyer flask, dissolved in 1500 mL ethyl acetate, to which was added
saturated NaHCO3 solution (300 mL). The mixture was stirred for 20ꢀ25 min. The aqueous layer
was discarded and organic layer was washed again with saturated NaHCO3 solution (300 mL),
and brine (300 mL). The organic layer was separated, dried over Na2SO4, filtered and
evaporated to minimum amount (not till dryness) on rotary evaporator (30 °C bath temperature)
and filtered again. The airꢀdried solid was further dried under high vacuum. Product 2 was
isolated as light yellow powder. Yield: 100.5 g (94.2±1.4 %). Rf = 0.28 (1:4 EtOAc:Hexane). 13C
NMR (125 MHz, CDCl3):
δ in ppm 82.8 ((C=O)ꢀC=C(-I)ꢀC(C=O)= aromatic ring carbon), 85.4
(ꢀC(-I)=C(ꢀNH)ꢀC(-I)= aromatic ring carbon), 148.7 (=C(ꢀNH)ꢀ aromatic ring carbon), 156.1
ACS Paragon Plus Environment
20

Page 21 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
(=C(ꢀC=O)ꢀ aromatic ring carbon), 169.6 (ꢀC(=O)ꢀCl). The product 2 was sufficiently pure
(single spot on TLC) and was taken on to the next step without further purification.
7
8
9
5,5'-[Malonylbis(azanediyl)]bis(2,4,6-triiodoisophthaloyl dichloride) 3. 5ꢀaminoꢀ2,4,6ꢀ
triiodoisophthaloyl chloride 2 (80 g, 134.29 mmol, 1 equiv.) was dissolved in anhydrous THF
(640 mL, 8 mL/g of 2), under N2. Malonyl chloride (6.54 mL, 67.14 mmol, 0.5 equiv.) was
added dropwise at room temperature. The reaction was heated to 50 °C and allowed to stir
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
overnight. The solution was cooled to room temperature and then left to precipitate at −20 °C
overnight. White solid was filtered and quickly washed with cold THF twice (2 x 250 mL,
cooled by keeping Erlenmeyer flask, containing THF, either in iceꢀwater bath or refrigerator (+4
°C) for 30ꢀ40 minutes). The mother liquor was concentrated (up to 30 %) on rotary evaporator
and then again allowed to precipitate at −20 °C overnight, and filtered. Product 3 was isolated as
white solid. The reaction was repeated thrice and average overall yield was 75.6 g (89.4±2.7 %).
Rf = 0.52 (1:1 EtOAc:Hexane). 13C NMR (125 MHz, CDCl3):
δ in ppm 41.4 (ꢀC(=O)ꢀCH2ꢀ
C(=O)ꢀ), 88.4 ((C=O)ꢀC=C(-I)ꢀC(C=O)= aromatic ring carbon), 97.7 (ꢀC(-I)=C(ꢀNH)ꢀC(-I)=
aromatic ring carbon), 141.3 (=C(ꢀNH)ꢀ aromatic ring carbon), 149.3 (=C(ꢀC=O)ꢀ aromatic ring
carbon), 164.0 (ꢀNHꢀC(O)ꢀCH2ꢀC(=O)ꢀNH), 169.3 (ꢀC(=O)ꢀCl). The product 3 was sufficiently
pure (single spot on TLC) and was taken on to the next step without further purification.
Tetra-tert-butyl
[({5,5'-[malonylbis(azanediyl)]bis(2,4,6-
4.
triiodoisophthaloyl)}tetrakis(azanediyl))tetrakis(ethane-2,1-diyl)]tetracarbamate
Intermediate 3 (70 g, 55.55 mmol, 1 equiv.) was dissolved in anhydrous dimethyl acetamide (210
mL, 3 mL/g of 3) as a clear, slightly yellowish solution under N2. Next, N,Nꢀ
diisopropylethylamine (67.8 mL, 389.8 mmol, 7 equiv.) was added dropwise at room
temperature. The reaction was cooled to 0 °C, followed by dropwise addition of tertꢀbutylꢀNꢀ(2ꢀ
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 47
1
2
3
4
5
6
7
8
9
aminoetyl)carbamate (38.6 mL, 24.45 mmol, 4.4 equiv.). The iceꢀwater bath was removed after
30 minutes and reaction mixture was heated to 50 °C overnight. The reaction mixture was
precipitated into 1 N HCl in water (1750 mL), forming a white solid. The white solid was filtered
and washed thoroughly with water (3 x 600 mL). The filtration and washing process took several
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
hours. The solid cake was dried to completion by lyophilisation overnight to obtain 4 as a white
1
powder. Yield: 87.4 g (89.7±2.2 %). H NMR (500 MHz, DMSOꢀd6):
δ in ppm 1.38 (s, 36H, ꢀ
NHꢀC(=O)ꢀOꢀ(ꢀCH3)3), 3.15ꢀ3.23 (m, 16H, ꢀNHꢀCH2ꢀCH2ꢀNHꢀ), 3.51 (s, 2H, ꢀC(=O)ꢀCH2ꢀ
C(=O)ꢀ), 6.75 (bs, 4H, ꢀNHꢀC(=O)ꢀOꢀ(ꢀCH3)3), 8.46ꢀ8.75 (m, 4H, ꢀC(=O)ꢀNHꢀCH2ꢀCH2ꢀNHꢀ),
10.20 (s, 2H, ꢀNHꢀC(=O)ꢀCH2ꢀC(=O)ꢀNHꢀ); 13C NMR (125 MHz, DMSO):
δ in ppm 28.2 (ꢀ
+
C(CH3)3), 37.5, 38.8, (ꢀNHꢀCH2ꢀCH2ꢀNH3 ), 42.4 (ꢀC(=O)ꢀCH2ꢀC(=O)ꢀ), 77.8 (ꢀOꢀCꢀ(CH3)3),
90.3 ((C=O)ꢀC=C(-I)ꢀC(C=O)= aromatic ring carbon), 99.1 (ꢀC(-I)=C(ꢀNH)ꢀC(-I)= aromatic
ring carbon), 142.7 (=C(ꢀNH)ꢀ aromatic ring carbon), 150.0 (=C(ꢀC=O)ꢀ aromatic ring carbon),
155.5 (ꢀNHꢀC(=O)ꢀOꢀ(CH3)3), 164.8 (ꢀNHꢀC(O)ꢀCH2ꢀC(=O)ꢀNH), 169.3 (ꢀNHꢀC(=O)ꢀring).
HRꢀMS (ESIꢀTOF): [M+H]+ theoretical m/z = 1776.8770, observed m/z = 1777.2998. The
product 4 was sufficiently pure and was taken on to the next step without further purification.
5,5'-[Malonylbis(azanediyl)]bis[N1,N3-bis(2-aminoethyl)-2,4,6-triiodoisophthalamide],
trifluoroacetate salt (CA4+ TFA salt) 5a. Compound 4 (85 g, 48.44 mmol) was mixed with 1:1
mixture of DCM:TFA (300 mL, 3.5 mL/g of 4). Immediately, bubbling was observed. Over the
course of an hour, 4 was completely dissolved. The reaction was continued for one more hour to
ensure the complete Bocꢀdeprotection. The reaction mixture was then precipitated in diethyl
ether (1200 mL, 4 times the volume of DCM:TFA mixture) to afford a white solid. The solid
product was then filtered, washed with diethyl ether (2 x 350 mL) and dried under vacuum to
1
yield 5a as a white powder. Yield: 86.4 g (98.6±1.3 %) H NMR (500 MHz, DMSO):
δ in ppm
ACS Paragon Plus Environment
22

Page 23 of 47
Journal of Medicinal Chemistry
1
2
+
3
4
3.00 (s, 8H, ꢀNHꢀCH2ꢀCH2ꢀNH3 ), 3.41ꢀ3.53 (m, 10H, ꢀC(=O)ꢀCH2ꢀC(=O)ꢀ, ꢀNHꢀCH2ꢀCH2ꢀ
5
6
+
+
+
NH3 ), 7.99 (bs, 12H, ꢀNH3 ), 8.66ꢀ8.86 (m, 4H, ꢀC(=O)ꢀNHꢀCH2ꢀCH2ꢀNH3 ), 10.19ꢀ10.30 (m,
7
8
9
2H, ꢀNHꢀC(=O)ꢀCH2ꢀC(=O)ꢀNHꢀ); 19F NMR (470 MHz, DMSOꢀd6):
δ
in ppm ꢀ73.72 (s, ꢀCF3ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
+
C(=O)ꢀ); C NMR (125 MHz, DMSO):
δ
in ppm 36.6 (ꢀNHꢀCH2ꢀ), 37.4 (ꢀCH2ꢀNH3 ), 42.5 (ꢀ
C(=O)ꢀCH2ꢀC(=O)ꢀ), 90.3 ((C=O)ꢀC=C(-I)ꢀC(C=O)= aromatic ring carbon), 99.2 (ꢀC(-I)=C(ꢀ
NH)ꢀC(-I)= aromatic ring carbon), 99.4 ꢀC(-I)=C(ꢀNH)ꢀC(-I)= aromatic ring carbon), 109.5,
115.9, 118.2 (ꢀCF3ꢀCOꢀ), 142.9 (=C(ꢀNH)ꢀ aromatic ring carbon), 149.5 (=C(ꢀC=O)ꢀ aromatic
ring carbon), 158.1 (–CF3ꢀC(=O)Oꢀ), 164.8 (ꢀNHꢀC(=O)ꢀCH2ꢀC(=O)ꢀNH), 169.5 (ꢀNHꢀC(=O)ꢀ
ring). HRꢀMS (ESIꢀTOF): [M+Na]+ theoretical m/z = 1354.6847, observed m/z = 1354.6752.
5,5'-[Malonylbis(azanediyl)]bis[N1,N3-bis(2-aminoethyl)-2,4,6-triiodoisophthalamide],
chloride salt (CA4+) 5. Trifluoroacetate ion was exchanged with chloride counter ion (Clꢀ) by
dissolving 5a (150 g, 82.8mmol) in 3 N HCl (450 mL), and subsequent precipitation in acetone
(4500 mL). The process was repeated two times to achieve complete conversion of
trifluoroacetate salt 5a into chloride salt 5. At last, the crude product was dissolved in 0.1 N HCl
(350 mL), treated with activated charcoal (20 g), filtered and washed twice with 50 mL of 0.1 N
HCl. The solution (350 mL + 50 mL + 50 mL = 450 mL) was precipitated by dropwise addition
into acetone (4500 L) with constant vigorous stirring. Dropꢀbyꢀdrop addition and constant
vigorous stirring are necessary for formation of fine precipitates. The solution was filtered,
washed with 14:1 acetone:water mixture and dried under high vacuum to obtain 5 as a white
powder. Yield: 116.7 g (93.8±2.1 %). 1H NMR (500 MHz, DMSO):
δ
in ppm 2.97 (bs, 8H, ꢀNHꢀ
+
+
CH2ꢀCH2ꢀNH3 ), 3.49ꢀ3.66 (m, 10H, ꢀC(=O)ꢀCH2ꢀC(=O)ꢀ, ꢀNHꢀCH2ꢀCH2ꢀNH3 ), 8.30 (bs, 12H,
+
+
ꢀNH3 ), 8.71ꢀ8.81 (m, 4H, ꢀC(=O)ꢀNHꢀCH2ꢀCH2ꢀNH3 ), 10.21ꢀ10.35 (m, 2H, ꢀNHꢀC(=O)ꢀCH2ꢀ
C(=O)ꢀNHꢀ); 19F NMR (470 MHz, DMSOꢀd6): no peak observed; 13C NMR (125 MHz, DMSO):
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 47
1
2
3
4
+
δ
in ppm 36.5 (ꢀNHꢀCH2ꢀ), 37.4 (ꢀCH2ꢀNH3 ), 42.2 (ꢀC(=O)ꢀCH2ꢀC(=O)ꢀ), 90.6 ((C=O)ꢀC=C(-
5
I)ꢀC(C=O)= aromatic ring carbon), 100.0 (ꢀC(-I)=C(ꢀNH)ꢀC(-I)= aromatic ring carbon), 142.8
(=C(ꢀNH)ꢀ aromatic ring carbon), 149.4 (=C(ꢀC=O)ꢀ aromatic ring carbon), 164.8 (ꢀNHꢀC(O)ꢀ
CH2ꢀC(=O)ꢀNH), 169.5 (ꢀNHꢀC(=O)ꢀring). HRꢀMS (ESIꢀTOF): [M+H]+ theoretical m/z =
1354.6847, observed m/z = 1354.7031. The purity of 5 salt was determined by analytical
reverseꢀphase HPLC. Briefly, a Varian ProStar HPLC pump and UVꢀvis detector with Hamilton
C18 HxSil 5ꢀm 250x4.6 mm column was used with 1 mL/min flow rate of 95/5 water/ACN
(isocratic). Product 5 was detected as a single peak at 245 nm with a retention time of 2.3
minutes (Figure S12). The purity of the CA4+ 5 was found to be >98%. Shelf life of CA4+ 5
powder and aqueous solution of CA4+ 5 (12 or 24 mgI/mL, pH=7.4, 400 mOsm/kg), was
determined by recording HPLC profile after one year. Both, CA4+ 5 powder and its aqueous
solution remain unaltered (eluted as single peak at the same retention time) when stored for one
year in refrigerator (+4 ºC). CA4+ 5 powder can be stored unaltered even at room temperature
for a year.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A cationic contrast agent (CA4+) is highly taken up in articular cartilage
Contrast Agent Solutions
Solutions of 80 mgI/mL iodixanol
(Visipaque, Medline Industries, Inc.,
Mundelein, Illinois) and 80 mgI/mL
Table 1. Contrast agent solution parameters.
ioxaglate (Hexabrix, Guerbet, Bloomington,
IN)
were prepared by dilution into Nanopure water (Barnstead Nanopure Thermo Scientific,
Waltham, MA) and balanced to 400 ± 20 mOsm/kg. The solution of 12 mgI/mL CA4+ (Contrast
Agent with 4+ charge) was prepared by adding 1.182 g of dry concentrate to 50 mL of Nanopure
ACS Paragon Plus Environment
24

Page 25 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
water, balanced to pH 7.0 (perpHecT LogR meter, model 310, Thermo Scientific, Waltham,
MA) with 4.0 M sodium hydroxide (Fisher Scientific, St. Louis, MO) and balanced to 400 ± 20
mOsm/kg. The parameters of these solutions are listed in Table 1, where the molecular weight is
listed of the ionized species.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The osmolality of solutions was balanced with sodium chloride (Fisher Scientific, Pittsburgh,
PA) and measured with a freezing point osmometer (The Advanced Osmometer, Advanced
Instruments, Inc., Norwood, MA).
A
preservative cocktail containing 5mM
ethylenediaminetetraacetic acid (EDTA), 5mM benzamidine HCl, 5x AntibioticꢀAntimycotic
(AntiꢀAnti #15240096, Life Technologies, Carlsbad, CA) was used in solutions containing ex
vivo osteochondral samples to prevent nonspecific degradation of the cartilage. The mixture was
added by 100ꢀfold dilutions of 500mM, 500mM and 500x stock solutions, respectively.
Osteochondral plugs were allowed to recover for at least 12 hours in 400 ± 20 mOsmol/kg saline
at 4 °C between each of the mechanical tests, micro computed tomography (ꢁCT) imaging and
the 1,9 dimethylmethylene blue (DMMB) assay for GAG content.
7mm Osteochondral Bovine Plugs
Six osteochondral plugs (7 mm diameter) were cored from the femoral groove of the stifle
joint from a freshly slaughtered, skeletally mature cow (36ꢀ54 months, Research 87, Boylston,
MA) using a diamondꢀtipped cylindrical cutter (Catalog #102080, Starlite Industries, Rosemont,
PA) irrigated with 0.9% saline at room temperature. The samples were frozen at ꢀ20°C in 400 ±
20 mOsmol/kg saline with the inhibitor mixture.
Electrochemical modelling of cationic contrast agent uptake in cartilage emphasizes
electrostatic interactions
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 47
1
2
3
4
5
6
7
8
9
Previous studies using cationic contrast agents have demonstrated high contrast uptake
into cartilage. The mechanism for this high contrast uptake could be purely electrostatic (i.e. high
concentrations of mobile cations evolve in the tissue due to the anionic fixed charge density
contributed by immobilized GAGs), or a stronger interaction (more akin to receptor binding)
could be present between the cationic agent and moieties inside the cartilage ECM. If the latter
occurs, determining a binding constant is an important step for characterizing the interaction and
for further developing the quantitative CECT technique. Receptorꢀligand binding studies are very
commonly conducted by incubating a receptor with a wide range of ligand concentrations to
determine a dissociation constant, Kd, which is an indicator of the binding strength between the
receptorꢀligand pair.63–65 Thus, this technique was used to evaluate the binding affinity between
cationic contrast agents (CA4+, bearing +4 charge) and bovine cartilage ECM. Additionally,
small cations in cartilage are also affected by electrostatic attraction to anionic GAGs, which
could also explain the high partitioning of cationic contrast agents.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The cationic contrast agent CA4+ was prepared as six serial dilutions in deionized water
containing 2, 4, 8, 16, 32, and 64 mgI/mL. Six serial dilutions of ioxaglate were prepared
(Hexabrixꢀ320®, Mallinckrodt, Hazelwood, MO) at the same iodine concentrations. All dilutions
contained a preservative cocktail of protease inhibitors, antibiotics, and antimycotics to prevent
tissue degradation over the course of the experiment (5 mM EDTA, 5 mM Benzamidine HCl,
SigmaꢀAldrich, St. Louis, MO; 1x Gibco® AntibioticꢀAntimycotic, Life Technologies, Carlsbad,
CA). The pH of each solution was adjusted to pH 7.4 with sodium hydroxide, and sodium
chloride was added to each dilution to achieve a solution isoꢀosmolar to synovial fluid (400
mOsmol/kg). Nine bovine osteochondral plugs (7 mm diameter) were harvested from one bovine
femoral condyle to obtain samples with a narrow variation in GAG content. All nine samples
ACS Paragon Plus Environment
26

Page 27 of 47
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were imaged with µCT prior to contrast agent immersions (baseline scan). Three plugs (n=3)
were randomly chosen for CA4+ immersions, and each was immersed in a large reservoir (4 mL)
of 2 mgI/mL CA4+ for 24 hours and then imaged. Following imaging, the plugs were immersed
in the next dilution in order of concentration (4 mgI/mL CA4+) for 24 hours and then imaged
similarly. This protocol of serial immersions was followed repeatedly to collect CECT imaging
data for CA4+ at all six concentrations. The remaining six plugs were assigned to ioxaglate
(n=3) immersions, and an identical protocol was followed for each contrast agent. The contrast
agent solutions were also scanned in a similar geometry as the plugs in order to produce a
standard conversion curve between contrast agent concentration (mgI/mL) and CT attenuation
(Hounsfield Units). After plug immersion and scanning for the highest concentrations (64
mgI/mL), all nine plugs were washed in copious saline for 24 hours and prepared for DMMB
assay. Briefly, the cartilage was removed from the subchondral bone, weighed, and lyophilized
to determine wet and dry cartilage masses. Each cartilage specimen was digested in papain
buffer and the GAG content was assessed using the DMMB assay (see SI).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Images of the cartilage and underlying subchondral bone were acquired using a micro
computed tomographic imaging system (µCT40, Scanco Medical AG, Brüttisellen, Switzerland)
at an isotropic voxel resolution of 36 µm, 70 kVp tube voltage, 113 µA current and 300 ms
integration time. The CT data sets were imported into image processing software (Analyze®,
AnalyzeDirect, Overland Park, KS) and the cartilage was segmented using a semiꢀautomatic
contour based segmentation algorithm. The mean cartilage xꢀray attenuation values using the
Hounsfield Scale were obtained by averaging attenuation values for all cartilage tissue over the
entire segmented volume. Then, the nonꢀcontrast (baseline) attenuation was subtracted from all
contrast immersion attenuations, and the standard curve was used to convert CECT attenuation to
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 47
1
2
3
4
5
6
7
8
9
contrast concentration inside the cartilage tissue. Linear regression analysis was used to express
the intraꢀtissue contrast concentration as a linear function of the contrast reservoir concentration
for plugs immersed in ioxaglate. For CA4+, a modified Langmuir isotherm was fit to the data to
illustrate the nonlinear relationship between reservoir concentration and intraꢀtissue
concentration (MATLAB®, MathWorks, Natick, MA). This model allowed an EC50 value to be
calculated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to determine if the cationic contrast agents partitioned in cartilage according to
the predications of Donnan equilibrium theory, a previously published method was implemented
in which the partition coefficient of contrast agent, KCA4+, was plotted as a function of the
predicted sodium ion partition coefficient, KNa+. The partition coefficients for CA4+ was
calculated as the ratio between the intraꢀtissue concentration at equilibrium and the bath
concentration. The partition coefficient for Na+ was calculated using the law of electroneutrality
and Donnan equilibrium theory using custom MATLAB code. That is, Donnan equilibrium
requires that:66
ꢓ
ꢔ
ꢐ
ꢏꢐꢐꢐꢑꢒ = ꢐ
ꢐꢑ
ꢐ
ꢐ
ꢕꢖꢗ
ꢕꢖꢗ
ꢐꢘꢙ
ꢐꢘꢙ
=
Equation 2
Where ꢀꢚꢛ is the intraꢀtissue concentration of contrast agent, ꢀꢚꢛ is the bath contrast
concentration, ꢀꢜꢝꢞ is the intraꢀtissue sodium concentration, ꢀꢜꢝꢞ is the bath sodium
concentration, ꢀꢚꢟꢌ is the intraꢀtissue chloride concentration, and ꢀꢚꢟꢌ is the bath chloride
concentration. Since charge neutrality requires that the sum of all charges inside the cartilage
tissue are a net of zero,
ꢠꢐꢡ + ꢐꢕꢖꢞ − ꢐꢐꢘꢌ + ꢐꢐꢑ = ꢢ
Equation 3
ACS Paragon Plus Environment
28

Page 29 of 47
Journal of Medicinal Chemistry
1
2
3
4
Where FCD is the intraꢀtissue (negative) fixed charge density contributed primarily by the
5
6
7
GAGs. Combining Equations 1 & 2 and solving for ꢀꢜꢝꢞ allows the partition coefficients to be
8
9
calculated:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢐ
ꢐ
ꢐ
ꢐ
ꢐꢑ
ꢐꢑ
ꢕꢖꢗ
ꢕꢖꢗ
ꢣꢐꢑ
=
ꢣꢕꢖꢞ
=
Equation 4
ꢤ
ꢚꢛꢥꢞwas each plotted as a function of ꢤꢜꢝꢞ according to their relationship shown in
Equation 1,67 where:
ꢐ
ꢐꢑ
ꢘꢦꢧꢨ
ꢩ
ꢐ
ꢐꢑ
ꢔ =
Equation 5
ꢐ
ꢐ
ꢕꢖꢗ
ꢕꢖꢗ
ꢘꢦꢧꢨ
ꢩ
and a nonlinear leastꢀsquares method was used to estimate the net charge (Z) on the CA4+
(MATLAB®, MathWorks).
Cationic contrast agents provide quantification of biochemical and biomechanical
measures of human cartilage quality
Two human hemiꢀpelvis specimens
(ages 40 and 61, female and male
respectively) were obtained and thawed
at 4°C (MedꢀCure, Portland, OR). Six
osteochondral plugs (7 mm diameter)
were harvested from each of the two
femoral heads using a diamond tipped
Figure 8. Osteochondral cores (7 mm diameter)
were harvested from various locations around the
cylindrical cutter under constant
femoral head surfaces.
irrigation (Figure 8). Each plug was
rigidly clamped in a mechanical testing apparatus (Enduratec3230, BOSE, Eden Prairie, MN)
and a compressive preꢀload of 5 N was applied between the cartilage surface and a nonporous
ACS Paragon Plus Environment
29