Targeting tumor hypoxia: A third generation 2-nitroimidazole-indocyanine dye-conjugate with improved fluorescent yield

Article abstract of DOI:10.1039/c5ob01460c

Tumor hypoxia is associated with the rapid proliferation and growth of malignant tumors, and the ability to detect tumor hypoxia is important for predicting tumor response to anti-cancer treatments. We have developed a class of dye-conjugates that are related to indocyanine green (ICG, 1) to target tumor hypoxia, based on in vivo infrared fluorescence imaging using nitroimidazole moieties linked to indocyanine fluorescent dyes. We previously reported that linking 2-nitroimidazole to an indocyanine dicarboxylic acid dye derivative (2) using an ethanolamine linker (ethanolamine-2-nitroimidazole-ICG, 3), led to a dye-conjugate that gave promising results for targeting cancer hypoxia in vivo. Structural modification of the dye conjugate replaced the ethanolamine unit with a piperazineacetyl unit and led a second generation dye conjugate, piperzine-2-nitroimidazole-ICG (4). This second generation dye-conjugate showed improved targeting of tumor hypoxia when compared with 3. Based on the hypothesis that molecules with more planar and rigid structures have a higher fluorescence yield, as they could release less absorbed energy through molecular vibration or collision, we have developed a new 2-nitroimidazole ICG conjugate, 12, with two carbon atoms less in the polyene linker. Dye-conjugate 12 was prepared from our new dye (8), and coupled to 2-nitroimidazole using a piperazine linker to produce this third-generation dye-conjugate. Spectral measurements showed that the absorption/emission wavelengths of 657/670 were shifted ～100 nm from the second-generation hypoxia dye of 755/780 nm. Its fluorescence quantum yield was measured to be 0.467, which is about 5 times higher than that of 4 (0.083). In vivo experiments were conducted with balb/c mice and 12 showed more than twice the average in vivo fluorescence intensity in the tumor beyond two hours post retro-orbital injection as compared with 4. These initial results suggest that 12 may significantly improve in vivo tumor hypoxia targeting.

Full text of DOI:10.1039/c5ob01460c

View Article Online
View Journal
Organic &
Biomolecular
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: F. Zhou, S.
Zanganeh, I. Mohammad, C. J. Dietz, A. Abuteen, M. B. Smith and Q. Zhu, Org. Biomol. Chem., 2015, DOI:
10.1039/C5OB01460C.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/obc

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 1 of 11
View Article Online
DOI: 10.1039/C5OB01460C
Journal Name
ARTICLE
Targeting Tumor Hypoxia: A Third
Generation 2-Nitroimidazole–
Indocyanine Dye–Conjugate with
Improved Fluorescent Yield
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
Feifei Zhou,1a Saeid Zanganeh,1a Innus Mohammad,2a
www.rsc.org/
Christopher Dietz,2a Akram Abuteen,1 Michael B. Smith,*2
Quing Zhu*1
Tumor hypoxia is associated with the rapid proliferation and
growth of malignant tumors, and the ability to detect tumor hypoxia is
important for predicting tumor response to anti-cancer treatments. We
have developed a class of dye-conjugates that are related to indocyanine
green (ICG, 1) to target tumor hypoxia, based on in vivo infrared
fluorescence imaging using nitroimidazole moieties linked to
indocyanine fluorescent dyes. We previously reported that linking 2-
nitroimidazole to an indocyanine dicarboxylic acid dye derivative (2)
using an ethanolamine linker (ethanolamine-2-nitroimidazole-ICG, 3),
led to a dye-conjugate that gave promising results for targeting cancer
hypoxia in vivo. Structural modification of the dye conjugate replaced
the ethanolamine unit with a piperazineacetyl unit and led a second
generation dye conjugate, piperzine-2-nitroimidazole-ICG (4). This
second generation dye-conjugate showed improved targeting of tumor
hypoxia when compared with 3. Based on the hypothesis that molecules
with more planar and rigid structures have a higher fluorescence yield,
as they could release less absorbed energy through molecular vibration
or collision, we have developed a new 2-nitroimidazole ICG conjugate,
12, with two carbon atoms less in the polyene linker. Dye-conjugate 12
was prepared from our new dye (8), and coupled to 2-nitroimidazole
using a piperazine linker to produce this third-generation dye-conjugate.
Spectral measurements showed that the absorption/emission
wavelengths of 657/670 were shifted ~100 nm from the second-
generation hypoxia dye of 755/780 nm. Its fluorescence quantum yield
was measured to be 0.467, which is about 5 times higher than that of 4
(0.083). In vivo experiments were conducted with balb/c mice and 12
showed more than twice the average in vivo fluorescence intensity in the
tumor beyond two hours post retro-orbital injection as compared with 4.
These initial results suggest that 12 may significantly improve in vivo
tumor hypoxia targeting.
Introduction
Tumor hypoxia mainly appears as a structural or
functional misbalance between the tumor micro-environmental
a
1
Equal contribution.
Department of Biomedical Engineering and Electrical
2
Engineering, University of Connecticut, Storrs, CT, USA.
Chemistry, University of Connecticut, Storrs CT, USA
Department of
*
Communicating
authors:
MBS,
michael.smith@uconn.edu;
QZ,
zhu@engr.uconn.edu
This journal is © The Royal Society of Chemistry 20xx
Organic and Biomolecular Chemistry 2015, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 2 of 11
ARTICLE
Journal Name
17
oxygen supply and consumption, and is associated with the rapid
1,2,3,4
using a near-infrared fluorescence imaging techVnieiqwuAert.icle OTnlhinee
DOI: 10.1039/C5OB01460C
progression of malignant tumors.
The identification of fluorescence signals were measured to be two-fold higher at the
tumor hypoxia is important for the assessment and prediction of
therapy, because it is known to cause an aggressive phenotype
N
N
N
N
HO2C
CO2H
2
1
SO3-Na+
SO3-Na+
-
-
SO3
SO3
5,6
O
O
N
and is associated with an increased resistance to therapy.
O
O
N
N
N
N
N
O
O
N
N
Positron emission tomography (PET) is an established method
O
O
N
-
-O3
S
SO3-Na+
+Na-O3
S
NH
SO3
O
NO2
N
HN
N
O
N
N
4
3
7
for imaging hypoxia noninvasively. Imidazole compounds are
N
NO2
N
NO2
NO2
N
widely used to prepare PET radionuclides, and it is known that
2-nitroimidazoles show an affinity for hypoxic cells, with the
Figure 1. Indocyanine dyes and first and second-generation dye
conjugates
8,9 ,10,11,12
highest electron affinities of this class of compounds.
19
tumor site, relative to the untargeted dye 2.
Furthermore, we
There are problems associated with PET, including the
requirement that radioactive tracer must be administered to
patients by injection, as well as the high system cost. Such
problems have spurred the development of alternative tumor
hypoxia imaging techniques.
showed that 3 could be detected in the tumor for 5-7 h, post-
injection. We viewed the ester linkage in 3 as too labile and in
an effort to improve in vivo efficacy we replaced the
ethanolamine moiety with a more robust piperazine moiety,
18
generating the second-generation dye conjugate 4.
In vivo
Indocyanine green (ICG, 1 in Figure 1) is the only FDA
approved fluorescence imaging agent for patient use, but its
application is hindered by the problems of a low quantum yield
and quick loss of fluorescence after binding to proteins in
13,14,15
targeting experiments in mice with 4 showed a two-fold higher
fluorescence in hypoxic tumors relative to 3, within three h of
injection, and the fluorescence was 1.6-1.7 times higher beyond
18
three h.
circulating blood.
In previous work, we demonstrated a
16
Further research in our laboratories has focused on
improving the fluorescent yield of the dye, in an effort to
improve in vivo detectability and/or to allow diminished dosing.
It is known that molecules with more planar and rigid structures
have a higher fluorescence yield as they release less absorbed
20,21
noninvasive method for in vivo imaging of hypoxic tumors.
This method is based on fluorescence infrared imaging of dyes in
hypoxic tumors, after injection of a nitroimidazole-indocyanine
dicarboxylic acid conjugate, henceforth referred to as a dye-
conjugate. In that work, dicarboxylic acid 2 was used as the
untargeted dye rather than 1, and 2-nitroimidazole was linked to
energy via molecular vibration or collision.
Based on this
hypothesis, we examined structural changes in the dye (2). We
found that introduction of a ring into the polyene unit of 2 led to
17,16
18
the dye using ethanolamine
or piperazine
linkers as
shown in Figure 1. We thus prepared hypoxia targeting
ethanolamine-dye conjugate 3 and a piperazine-dye conjugate 4,
both linked to 2-nitroimidazole moieties.
22
a lower fluorescent yield, but shortening the polyene chain by
two carbon atoms (see 8 in Scheme 1) led to significant
enhancement in fluorescence (see Table 1).
Our first-generation dye conjugate (3) was shown to be
a successful candidate for noninvasive tumor hypoxia mapping
2 | Organic and Biomolecular Chemistry 2015, 00, 1-
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Journal Name
ARTICLE
O
O
View Article Online
SO2
DOI: 10.1039/C5OB01460C
HO2C
NHNH2
N
HO2C
1,2-dichlorobenzene
AcOH ,
NaOAc
5
HO2C
NaOAc , AcOH
Ac2O
N
N
-
N
CO2H
HO2C
8
6
-
SO3
Ph
Ph
N
H
N
SO3-Na+
•HCl
SO3
7
CH(OMe)2
CH(OMe)2
PhNH2 , aq HCl
Scheme 1. Synthesis of rigid dye 8.
With our new dye, 8, in hand, we synthesized a third
generation 2-nitroimidazole ICG conjugate (12 in Scheme 2), the
hypoxia-targeted rigid dye in subsequent figures. The
bis(carboxylic acids) 2 and 8 are referred to as non-targeted
dyes because they do not have the nitroimidazole moiety
associated with targeting tumor hypoxia. The goal of this study
is to quantify the in vivo hypoxia targeting capability of 12
relative to the non-targeted dyes as well as relative to 4, using a
murine tumor model. A direct comparison of the in vivo
performance of the rigid dye-conjugate 12 with the second-
generation hypoxia-targeting dye (4) showed greatly improved
imaging for 12. The biodistribution of each dye in different
organs was also evaluated and compared with fluorescence
Figure 2. Normalized absorption (solid curves) and fluorescence (dash
curves) spectra of hypoxia-targeted rigid dye 12 (black) and ICG from
Sigma-Aldrich (gray). This figure shows that in measuring the quantum
yield with the standard dye of ICG from Sigma-Aldrich, the excitation was
performed at 640 nm, as labeled with arrows. The fluorescence was
collected from 650-900 nm for all dyes. Six concentrations of each dye
were measured for the calculation of quantum yield. All dyes are
measured with spectrometer gain set at medium and the measurements
are performed in 9.25% sucrose. The spectra of non-targeted rigid dye 8
is similar to that of 12. The spectra of biscarboxylic acid ICG 2 and
18
piperazine-2-nitroimidazole-ICG 4 can be referred from past report.
nm and the emission maximum was at 670 nm, a shift of around
100 nm and 110 nm for absorption and emission, respectively,
when compared with 4. The other measured optical properties
for non-targeted rigid dye 8 and hypoxia-targeted rigid dye 12 in
9.25% sucrose solvent are shown in Table 1.
peak intensity of the dye in the tumor.
N
HN
N-Boc
N
BrCH2CO2Et
N
TFA
NO2
N
N
NO2
N
H
NO2
CO2Me
N
TBAI , K2CO3
MeCN , 80 °C
N-Boc
Table 1. The optical properties of rigid dyes in 9.25% sucrose solution
O
9
10
O
O
N
!"#
!"#
N
N
N
ꢀ
(ꢀꢁ)
Compound
ꢀ
(ꢀꢁ)
Extinction
coefficient
ꢀ (M-1cm-1)
261,971
Quantum
yield (Φ)
!"#
!"#
PyBOP, BtOH , DMF , DIPEA
rt , 48h
N
22%
NO2
N
N
12
N
N
+
NH2 -O2CCF3
O
N
+Na-O3
S
O
NO2
N
-
O
SO3
N
Non-
657
671
0.403
0.467
N
-
N
CO2H
HO2C
N
11
8
targeted
NO2
SO3-Na+
rigid dye 8
Hypoxia-
targeted
SO3
657
669
268,006
Scheme 2. Synthesis of third generation dye-conjugate 12.
rigid dye 12
The in vivo distribution of the Dye-Conjugate as a function of
time
Results
Optical properties of dyes
Figure 3 shows a typical set of fluorescent images
obtained by IVIS Lumina II Imaging System over a 48 hour
period for a mouse injected with 100 µl of 2 (a), 4 (b), 8 (c) and
12 (d) at 25 µM concentration. All dyes were accumulated in
The optical properties of 12 were measured and are
shown in Figure 2. The absorption maximum for 12 was at 657
This journal is © The Royal Society of Chemistry 20xx
Organic and Biomolecular Chemistry 2015, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 4 of 11
ARTICLE
Journal Name
and cleared from the tumor area with visible and different lower average maximum of 1.5×109 approximately at 15 min and
View Article Online
DOI: 10.1039/C5OB01460C
washout rates, reflecting tracer-specific uptake mechanisms. then the signal was decreased below detection level by around 3
The improvement on the fluorescence intensity of dye h post-injection.
accumulated in the tumor region after longer washout times is
visible in Figure 3.
Two-sided student’s t-test was performed between
average values of 12 and 8, 12 and 4, 12 and 2 in the 2-48 h
window and results show statistically significance between each
pair (Figure 4). Within the 2-48 h time window, the average
fluorescence peak intensity of the tumor injected with the
hypoxia-targeted rigid dye (12) is 2.7 times that of the non-
targeted rigid dye (8), 1.9 times that of the piperazine-2-
nitroimidazole-ICG (4), and 7.2 times that of the biscarboxylic
acid ICG (2).
Figure 3. Typical sets of fluorescent images as a function of time. The
first row shows fluorescent images obtained by IVIS Lumina II Imaging
System using a 710±15 nm/810-875 nm excitation/collection filter pair
over a 48 h period for individual mouse injected with 100 µL of (A) ICG
2; (B) piperazine-2-nitroimidazole-ICG 4. The second row shows
fluorescent
images
using
a
640±15
nm/695-770
nm
excitation/collection filter pair over a 48 h period for individual mouse
injected with 100 µL of (C) dye 8 and (D) dye 12. The tumor area for
each case was labelled with red circles when the tumor fluorescence
intensity reaches to the peak at around 15 min post-injection.
Dye washout characteristics
The washout curves of each dye in tumor area were
averaged at each observation point, and the bar plot with
standard deviation (STD) for washout period is shown in Fig. 4.
The number of mice injected with a particular dye is represented
by the parameter n in Figure 4. It is shown that for 12 the
fluorescence intensity reached an average maximum of 8.0×109
(Radiant efficiency, unit: p/sec/cm2/sr/(µW/cm2)) in 5-15 min
post-injection and the signal remained above detection level
(2×108, decided by background fluorescence) in 48 h. For 8, the
fluorescence reached an average maximum of 7.0×109 in 5-15
min post-injection and the signal remained above detection level
in 10 h. The fluorescence for 4 reached its average maximum
intensity of 3.5×109 at 15 min and the signal remained above
detection level in 10 h. For 2, fluorescence intensity reached
Figure 4. Kinetics of tumor uptake and washout characteristics of
hypoxia-targeted rigid dye 12, non-targeted rigid dye 8, piperazine-2-
nitroimidazole-ICG 4 and biscarboxylic acid ICG 2. The parameter n is
the number of mice injected with that dye. All tracers were injected at
25 µM concentration.
Dye residue in tumor after 48 hours
All mice were sacrificed after 48 h post-injection.
Tumor tissue were excised and imaged using the same imaging
conditions (Figure 5). For mice injected with 2, an average
maximum fluorescence intensity of 0.4×108 was detected in the
tumor. For mice injected with 4 and 8, higher averaged
maximum fluorescence intensity of 1.5×108 and 2.2×108,
respectively, were detected in the tumor. The highest averaged
4 | Organic and Biomolecular Chemistry 2015, 00, 1-
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Journal Name
ARTICLE
maximum fluorescence intensity of tumor was in the group
View Article Online
DOI: 10.1039/C5OB01460C
injected with the 12, measured as 5.0×108.
Immunohistochemistry (IHC) and dual labeling results
Typical fluorescence images labelled
with
pimonidazole hydrochloride-FITC (green, channel 2) and with
hypoxia-targeted dye 12 and non-targeted rigid dye 8 (red,
channel 1), are shown in Figure 6 top and middle row
respectively. The targeted rigid dye 12 labelled area is mainly
located in the center region and correlates with the region
labelled with pimonidazole hydrochloride-FITC; while the non-
targeted dye 8 labelled regions are scattered and showed no
significant overlap with the region labelled with pimonidazole
hydrochloride. On average, the correlation coefficient between
channel 1 and channel 2 images of 12 and 8 is statistically
significant (p<0.05) as shown in Figure 6 (g).
Figure 5. Dye residue in excised tumor tissue. (a) Fluorescence images of
excised tumors from tumor bearing mouse injected with 100 µL (from
left to right): dye 12, dye 8, dye 4, and dye 2, all at 25 µM concentration.
(b) Statistical results in different groups of mice tumor.
The fluorescence peak intensity of the excised tumor
injected with hypoxia-targeted rigid dye (12) after 48 hours is
about 2.3 times that of the non-targeted rigid dye (8), 3.4 times
that of the piperazine-2-nitroimidazole-ICG (4), and 14 times
that of the biscarboxylic acid ICG (2). The two-sided student’s t-
test on the excised tumor radiant efficiency data between
hypoxia-targeted (12) and non-targeted rigid dye (8) is
statistically significant (p=0.004).
By comparing the fluorescence radiant efficiency of the
dye in vivo at 48 h post-injection (Figure 4) and radiant
efficiency of the excised tumors (Figure 5), we observed similar
or higher fluorescence intensity in the latter (in vivo: 5.3×108,
1.3×108, 9.7×107, 2.5×107 versus excised tumors: 5.0×108,
2.2×108, 1.5×108, 3.5×107, for each dye respectively). The
imaging was performed on excised tumor without skin, so it is
reasonable to expect that the measured dye fluorescence
intensity is similar or higher.
Figure 6. IHC results. Top row: (a) Typical overlaid fluorescence image
labelled by dye 12 and commercial pimonidazole hydrochloride. (b)
Binary image acquired through microscope channel 1 (Cy5) after
processing with 11-12% threshold. (c) Binary image acquired through
microscope channel 2 (FITC) after processing. Middle row: (d) Typical
overlaid fluorescence image labelled by dye
8 (NT-Rigid) and
pimonidazole hydrochloride. (e) Binary image acquired through
microscope channel 1 (Cy5) after processing. (f) Binary image acquired
through microscope channel 2 (FITC) after processing. The scale bars in
(a) and (d) indicate 1 mm. Bottom row: (g) Correlation coefficients
computed between Ch1 and Ch2 binary images (hypoxia-targeted rigid
dye n=12, non-targeted rigid dye n=8).
This journal is © The Royal Society of Chemistry 20xx
Organic and Biomolecular Chemistry 2015, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 6 of 11
ARTICLE
Journal Name
Stability of dyes 2, 4, 8, and 12
View Article Online
DOI: 10.1039/C5OB01460C
All dyes in this work were examined in vivo over
the course of several hours. After each synthesis, and prior
to the in vivo studies, all dyes were stored in the dark in a
refrigerator until a solution of the dye was prepared in PBS
as required for the mouse studies. We examined the
photostability of dyes 2, 4, 8, and 12 when exposed to light.
A solution of each dye in 9.25% sucrose was examined at
750 nm and 650 nm for the ICG and rigid dyes respectively.
The initial absorbance data was measured and then every
15 min for the first hour and every hour thereafter. Each
sample was irradiated under a 500 W halogen lamp,
Figure 7. Dye stability study for 2, 4, 8, and 12.
Discussion
As anticipated, our results show that higher
fluorescence intensity is achieved with rigid dye 12 as compared
maintained at a distance of 600 mm, for 15 h. These with the second generation dye 4, although the
conditions are rather harsh and certainly more severe than absorption/emission wavelengths shifted ~100 nm from the
encountered under our experimental studies, but we second generation dye: from 755/780 nm to 657/670 nm (see
wanted to understand the relative photostability of these Figure 2). The fluorescence quantum yield of rigid dyes was
dyes. Our results are shown in Figure 7, which shows that measured to be about 0.4, which is around 5 times higher than
both the parent dye 2 and the second generation dye that of the second-generation dyes. We believe that the reason
conjugate 4 decomposed at ambient temperature rather for the spectrum shift and quantum yield increase of the rigid
quickly, although each dye persisted for about 3-4 hours. dye 8 and the dye-conjugate 12 is due to the more rigid
Interestingly, our ‘rigid dye’ 8 and also the dye-conjugate structure of the polyene chain, but the specific structural feature
22
12 showed improved photostability, decomposing much
more slowly. Indeed, there was a detectable concentration
of both dyes after 12 hours. The dye-conjugate 12 showed
slightly improved photostability relative to the parent 8.
that correlates with these changes is unknown at this time.
Florescence images of the sliced, excised tumor tissue
indicate that the dye molecules were not only distributed in
tumor periphery area, but also penetrated throughout the entire
tumor. Although the specific molecules this probe is binding to
are yet to be investigated, from the report of other 2-
23
nitroimidazole conjugated dye studies, we anticipate that this
finding is due to the presence of nitroimidazole moiety. It is
known that 2-nitroimidazole derivatives are trapped in the cells
under hypoxic conditions, and our results are consistent with
similar behavior for 4 and 12.
6 | Organic and Biomolecular Chemistry 2015, 00, 1-
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Journal Name
Despite the high retention of piperazine-2-
ARTICLE
The correlation between the dual-channel labelled
View Article Online
DOI: 10.1039/C5OB01460C
nitroimidazole-linked dyes 4 and 12 in the tumor region, results images was performed after applying a threshold to binary
also show higher fluorescence intensity in the kidney and liver images. By varying the threshold from 8-20%, we found that a
48 h post-injection. In general, we observed that urine contained lower threshold can cause a large variation among tumor
concentrated fluorescence dye. Previous reports indicated that samples due to noise. The best threshold was found to be 11-
ICG was selectively bound to the liver, later excreted in the bile 12%.
due to its binding with serum proteins and then eliminated by
24,25
A cytotoxicity study is beyond the scope of this current
work, but is clearly required if 12 is to be pursued as a useful
imaging agent. However, we found no observable difference for the
mice groups based on post-injection activity or weight loss,
suggesting that no significant short-term toxicity occurred to the
mice injected with 12.
giving bile to the feces.
We therefore expect that this
soluble dye conjugate is mainly washed out through the kidney
and also removed through liver and gall bladder.
Our work suggests that the dye-conjugates are widely
distributed in the mouse after injection. However, washout
occurs quickly, presumably by excretion, but dye conjugate 12 is
retained in the tumor to a significantly greater extent when
compared to precursor 8 (see Figure 5). Further, rigid dye 12 is
retained in the tumor to a greater extent than 4. A direct
comparison of residual fluorescence intensity of 8 with early
versions of hypoxia targeted dye 4 is irrelevant because dye 8
has a 5 times higher quantum yield than that of 4. The hypoxia
targeting related tumor retention (average fluorescence
intensive measured between 120-2880 min/ maximum
fluorescence intensity) of dye 8 was 0.038, which was much
smaller than those of dye 4 and 2 (0.106 and 0.066 respectively).
Our data, presented in Figure 4, indicates that for the
Experimental
Materials
The details of the synthetic procedures used to prepare
2–4, as well as their photophysical/chemical properties in
22,17
various solutions, have been reported previously.
We
prepared the non-targeted rigid dye, (8) with two carbons less
in the polyene linker compared to ICG dicarboxylic acid (2). As
shown in Scheme 1, indole 5 was prepared from commercially
26,22
available 4-hydrazinobenzoic acid,
and 3-methylbutan-2-
one. Subsequent treatment with butanesultone led to indole-
26,22
same concentration, dye 4 is retained in the tumor to a greater
extent than 12 or 8. This conclusion is based on a comparison of
dye intensity in the tumor at > 2 hours post-injection (2-48 h)
relative to the maximum intensity observed for each dye (5-15
min). However, both 8 and 12 have a higher fluorescence yield,
which makes them easier to detect. This observation suggests
that 12 may be used for detection of tumor hypoxia using a
significantly lower dose relative to 4.
sulfonate 6.
The reaction of the bis(dimethyl acetal) of
propanedial with aniline, in aqueous hydrochloric acid (HCl)
22
led to anilide 7 in 90% yield.
Subsequent reaction with indole
derivative 6 in the presence of acetic acid/acetic anhydride
22
buffered with sodium acetate gave a 30% yield of dye (8).
We
synthesized the dye-conjugate, 12, from the non-targeted rigid
dye, 8, by coupling to the piperazine-2-nitroimidazole fragment
(11), as shown in Scheme 2. Fragment 11 was prepared by
This journal is © The Royal Society of Chemistry 20xx
Organic and Biomolecular Chemistry 2015, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 8 of 11
ARTICLE
Journal Name
reaction of commercially available 2-nitroimidazole with ethyl Sodium 4-((Z)-2-((2E,4E)-5-(3,3-dimethyl-5-(4-(2-(2-nitro-
1H-imidazol-1-yl)acetyl)-piperazine-1-DcOarI:b1o0.n10y3lV)9i-/e1Cw-5A(Or4tBic-0le1O46nl0inCe
bromoacetate to give 9, followed by reaction with N-Boc sulfonatobutyl)-3H-indol-1-ium-2-yl)penta-2,4-dien-1-
ylidene)-3,3-dimethyl-5-(4-(2-(2-nitro-1H-imidazol-1-
22
yl)acetyl)piperazine-1-carbonyl)indolin-1-yl)butane-1-
sulfonate, 12.
piperazine to give 10.
The coupling agent Pybop,
(benzotriazol-1-yloxy)
tripyrrolidinophosphonium
A total of 78 mg (0.149 mmol) of PyBOP was added to a
stirring solution of bis-(carboxylic acid) 8 (50 mg, .0068 mmol)
in dry DMF (2 mL) at 0 °C. This solution was subsequently
treated with HOBt (hydroxybenzotriazole, 20 mg, 0.149 mmol),
followed by diisopropylethylamine (DIPEA, 0.026 mL, 0.149
mmol). This mixture was stirred for 15 min before the addition
of 2-nitroimidazole piperazine 11 (54 mg, 0.149 mmol) followed
by stirring at room temperature for 48 h in the dark. The DMF
was removed by air-drying to yield a thick blue oil and further
concentrated in vacuo overnight. The crude mixture was
recrystallized from methanol (15 mL) via the dropwise addition
diethyl ether (50 mL), and the resulting blue solid was washed
with acetonitrile (2x25 mL), ethyl acetate (2x25 mL), and
chloroform (2x25 mL). The crude was purified by C18 reverse
phase column chromatography (H2O:MeOH) to yield 12 (25 mg,
0.021 mmol, 31%) as a blue solid; mp: decomposed to black
residue > 250 oC; 1H NMR (400 MHz, D2O) d 8.17 (m, 3H), 7.71 (s,
2H), 7.63 (d, J = 8 Hz, 2H), 7.55 (s, 2H), 7.50 (d, J = 8, 1H), 7.38 (s,
1H), 6.76 (t, 2H), 6.47 (d, 2H), 5.64 (bs, 4H), 4.27 (bs, 4H), 3.66-
4.05 (bs, 16H), 3.06 (t, 4H), 2.10 (m, 4H), 1.97 (m, 2H), 1.80 (bs,
9H), 1.44 (m, 3H); HRMS (ESI-TOF): [M-Na]- Calc’d for m/z
C53H63N12O14S2 1155.4034. Found, m/z 1155.4028.
hexafluorophosphate, was used in dimethylformamide (DMF)
for the reaction of 8 with 11, and final purification of the crude
product on C18 reversed phase automated flash column
chromatography gave 12 in 31% yield.
Preparation of 12.
Sodium
4-[2-[(1E,3E,5Z)-7-[1,1-dimethyl-3-(4-
sulfonatobutyl)indol-2-ylidene]penta-1,3-dienyl]-1,1-
dimethylindol-3-ium-3-yl]butane-1-sulfonate, 8.
A vigorously stirred solution of 3-(5-carboxy-2,3,3-
trimethyl-3H-indolium-1-yl)propane-1-sulfonate (6, 0.14 g, 0.41
mmol) and 7 (0.05 g, 0.19 mmol) in acetic anhydride (1 mL) and
acetic acid (0.5 mL) was treated with sodium acetate (0.054 g,
0.66 mmol) and heated at reflux (120 °C) for 45 min. The
reaction mixture was cooled to ambient temperature and
anhydrous diethyl ether (5 mL) was added. The resulting
precipitate was isolated by vacuum filtration to give a crude
solid that was recrystallized (methanol: water) to give 8 as a
22
blue solid (0.09 g, 0.12 mmol, 63%); mp: decomposition at
1
287 °C; H NMR (400 MHz, CD3OD) δ 12.92 (bs, 2H), 8.43 (t, J =
13 Hz, 2H), 8.17 (s, 2H), 8.00 (d, J = 8.4 Hz, 2H), 7.53 (d, J = 8.4
Hz, 2H), 6.71 (t, J = 12.4 Hz, 1H), 6.50 (d, J = 13.7 Hz, 2H), 4.15
(bs, 4H), 1.18-1.76 (m, 8H), 1.73 (s, 12H); 13C NMR (100 MHz,
CD3OD) δ 175.7, 169.2, 156.8, 147.4, 142.8, 132.3, 128.7,128.6,
124.6, 112.0, 106.0, 51.7, 50.4, 45.1, 27.9, 27.2, 23.5; 13C NMR
(100 MHz, MeOD) d 175.7, 169.2, 156.8, 147.4, 142.8, 132.3,
128.7128.6, 124.6, 112.0, 106.0, 51.7, 50.4, 45.1, 27.9, 27.2, and
23.5 ppm. HRMS (TOF): [M+H]+ Calc’d for C35H43N2O10S2 m/z
715.2359. Found, m/z 715.2321.
Repeated efforts to obtain the 13C NMR failed to give a
18
spectrum. We observed this problem with dye conjugate 4.
Although the dye (2 and 8) showed reasonable 13C NMR spectra,
4 and 12 did not. We examined the possibility of aggregation,
rotamers, relaxation time, and low concentration due to
solubility issues. All of these issues can lead to poor 13C NMR
spectra. Attempts to obtain 13C NMR spectra at 25 °C failed, and
8 | Organic and Biomolecular Chemistry 2015, 00, 1-
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Journal Name
ARTICLE
we heated the samples to 55 °C to promote deaggregation. We ignored. The excitation/collection wavelengths are shown in
View Article Online
DOI: 10.1039/C5OB01460C
also explored different solvents, including CD3OD, (CD3)2CO, D2O. Figure 2.
We examined extended delay (relaxation) times up to 10 to 15
sec, as well as long acquisition times (up to 12 hours). None of
these experiments led to a 13C NMR. We have attempted indirect
C13 experiments, including HSQC, HMBC, and CIGAR. The
problem is likely due to low solubility of 4 and 12, coupled with
the low isotope percentage of 13C relative to 12C versus the high
percentage of 1H, but we do not have a definitive answer to this
problem. No experiments have been successful, so only 1H NMR
data is provided.
Murine tumor model and imaging
Murine tumor model preparation procedures have
18
been reported previously. The procedures were performed in
accordance with specifications in animal protocol approved by
the Institutional Animal Care and Use Committee of University of
Connecticut. Six to Eight week old balb/c mice with body weight
around 20g were used in this study and transplanted with 4T1
tumor cells. The experiments were performed when the tumor
size had attained a diameter of 5 mm or greater, 10-15 days
post-inoculation. Mice with a tumor bared on the top of their
right legs were imaged with an IVIS® Lumina II Imaging System
Measurement of quantum yield
In order to measure the quantum yield of hypoxia-
29
(Caliper Life Sciences, Hopkinton, MA).
targeted rigid dye 12 and non-targeted rigid dye 8, we followed
Mice were separated into four groups with same
average tumor size. Each group of mice was injected, via
retroorbital injection, with 100 µl of either biscarboxylic acid ICG
(2), piperazine-2-nitroimidazole-ICG (4), non-targeted rigid dye
(8), or hypoxia-targeted rigid dye (12) at a concentration of 25
µM.
27
the procedures described in the reference. Any of these four
unknown samples is measured against a fluorescence standard,
Indocyanine green (ICG) from Sigma-Aldrich with a known
28
quantum yield Φ!=0.012
with Varian Cary® 50 UV-Vis
Spectrophotometer and Varian Cary® Eclipse Fluorescence
Spectrophotometer. The fluorescence quantum yield is then
calculated according to equation 1, by taking into account the
Imaging the tumors began 1 minute post-injection and
were repeated at 15 min, 30 min, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 24 and
48 h. In all cases, the in vivo fluorescence intensity was
quantified. After 48 h, the tumor tissue were excised from each
sacrificed mouse, and immediately frozen in the liquid nitrogen.
After acquiring florescence images, the tissues were stored in
the -80 °C freezer for later histology processing.
absorbance through the absorption factor ꢀ (ꢀ!") and the
!
fluorescence through the integral of fluorescence ꢀ!(ꢀ!") at
specific excitation wavelength ꢀ!" and a range of collection
wavelength band ꢀ!"
.
!!(!!"
!!(!!"
)
! (!!"
)
!
!
!"
Φ!! = Φ!!
(1)
!!(!!"
)
)
!
!"
Here both standard (s) and unknown dye (i) were measured in
9.25% sucrose, thus compared with the reference, the difference
of refractive indices between standard and unknown dye can be
Immunohistochemistry
To visualize the hypoxic area in the tumor samples by
immunohistochemistry (IHC), the Hypoxyprobe™1 plus kit,
This journal is © The Royal Society of Chemistry 20xx
Organic and Biomolecular Chemistry 2015, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 10 of 11
ARTICLE
which contains the 2-nitroimidazole derivative pimonidazole
hydrochloride [1-(2-nitro-1H-imidazol-1-yl)-3-(piperidin-1-
Journal Name
View Article Online
DOI: 10.1039/C5OB01460C
Conclusions
yl)propan-2-ol] was purchased from HPI, Inc. (Burlington,
Massachusetts, USA). A total of 60 mg of pimonidazole
hydrochloride per kg body weight, diluted in 0.9% saline
solution was injected intravenously into mice 30-45 minutes
before sacrificing. Immediately after the animals were
euthanized, the tumor specimens were collected and flash-
frozen in the liquid nitrogen, and subsequently sectioned into 16
μm sections on a Leica CM 3050S cryotome (Leica, Nussloch,
Germany). The prepared tumor sections (sample number: 8 for
hypoxia-targeted rigid dye, 12 for non-targeted rigid dye) were
stored in -80 °C before staining. After thawing, the sections were
fixed in cold acetone for 10 minutes, and later rinsed and
incubated overnight at 4 °C with FITC-Mab1 diluted 1:20 in PBS
containing 0.1% bovine serum albumin and 0.1% Tween 20.
The hypoxic areas were visible with FITC attaching to
.
Hypoxia is an important tumor microenvironment and
targeting tumor hypoxia is significant in predicting tumor
response to treatments. Our in vivo experimental results of
the third generation tumor hypoxia-targeting fluorescence
dye, 12, showed that it exhibited superior fluorescence
detectability for hypoxic tumors in vivo. The fluorescence
quantum yield of 12 was measured to be ~0.467, which is
about 5 times higher than the 0.083 measured for the
second-generation hypoxia dye (4). Experiments on mice
showed that 12 provided more than twice the fluorescence
intensity in tumors 2 hours post-injection as compared with
the second-generation hypoxia dye (4). Our results suggest
that 12 may significantly improve in vivo detection of tumor
hypoxia beyond this initial study
pimonidazole hydrochloride under the fluorescence microscope
(Zeiss Axio M2, Pennsylvania, USA). In order to analyze the
targeting property of 12 in tumor, fluorescence images are
acquired simultaneously with filter sets for channel 1- Cy5 and
channel 2- FITC. The absorption and emission peaks of Cy5
(650/670 nm respectively) are overlapped with that of 12 and 8.
Digital images of the whole tissue slices were scanned and
acquired with 2.5× and 10× objective magnifications. For 10×
magnification, the exposure time is used as 1200 ms for channel
1 and 250 ms for channel 2. Obtained digital images were
transferred to ImageJ software, delineated with the boundaries
of hypoxic areas, split by channels for separate analysis. Each
image is later imported to Matlab, converted into a binary image
with a threshold of 12% of the labelled tissue area, and used for
computing the correlation of the labelled areas between
pimonidazole hydrochloride and the rigid dye (12 or 8).
Acknowledgements
The authors acknowledge Connecticut Public Health
and NIH R01002136 for partial funding of this project. We thank
Prof. Xiuling Lu and Donghui Song, UConn School of pharmacy,
for allowing us to use their laboratory for maintaining 4T1
tumor cell line. The University of Connecticut Animal Care and
Use Program is accredited by AAALAC International.
References
1
2
P. Vaupel, A. Mayer Cancer Metast. Rev., 2007, 26, 225-239.
M.C. Brahimi-Horn, and J. Chiche, J. Pouysségur J. Mol. Med.,
2007, 85, 1301-1307.
3
4
P. Vaupel and F. Kallinowski, P. Okunieff Cancer Res., 1989,
49, 6449-6465.
J.M. Brown and W.R. Wilson Nat. Rev. Cancer, 2004, 4, 437-
447.
10 | Organic and Biomolecular Chemistry 2015, 00, 1-
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 11 of 11
Organic & Biomolecular Chemistry
Please do not adjust margins
Journal Name
ARTICLE
5
6
7
8
9
W.R. Wilson and M.P. Hay Nat. Rev. Cancer, 2011, 11, 393-
View Article Online
DOI: 10.1039/C5OB01460C
410.
M. Höckel and P. Vaupel J. Natl. Cancer Inst., 2001, 93, 266-
276.
E. Hammond, M.-C. Asselin, D. Forster, J. O'Connor, J. Senra,
and K. Williams Clin. Oncol., 2014, 26, 277-288.
A. Nunn, K. Linder, and H.W. Strauss Eur. J. Nuc. Med., 1995,
22, 265-280.
L. Mei, Y. Wang, and T. Chu Eur. J. Med. Chem., 2012, 58, 50-
63.
10 H. Huang, H. Zhou, Z. Li, X. Wang, and T. Chu Bioorg. Med.
Chem. Lett., 2012, 22, 172-177.
11 L.-B.-A. Tran, A. Bol, D. Labar, B. Jordan, J. Magat, L. Mignion,
V. Grégoire, and B. Gallez Radiother Oncol., 2012, 105, 29-35.
12 Z. Li and T. Chu Curr. Pharm. Des., 2012, 18, 1084-1097.
13 Z. Guo, S. Park, J. Yoon, and I. Shin Chem.Soc. Rev., 2014, 43,
16-29.
14 A. Corlu, R. Choe, T. Durduran, M.A. Rosen, M. Schweiger,
S.R. Arridge, M.D. Schnall, and A.G. Yodh Opt. Express, 2007,
15, 6696-6716.
15 M. Aldrich, C. Davies-Venn, B. Angermiller, H. Robinson, W.
Chan, S. Kwon, and E.M. Sevick-Muraca Lymphat Res Biol.,
2012, 10, 20-24.
16 N.C. Biswal, C. Pavlik, M.B. Smith, A. Aquirre, L.T. Kuhn, Y. Xu,
S. Zanganeh, L.T. Kuhn, K.P.; Claffey, and Q. Zhu, J. Biomed.
Opt., 2011, 16(6), 066009-01 - 066009-08.
17 C. Pavlik, N.C. Biswal, F.C. Gaenzler, M.D. Morton, L.T. Kuhn,
K.P. Claffey, Q. Zhu, and M.B. Smith Dyes Pigments, 2011, 89,
9-15.
18 Y. Xu, S. Zanganeh, I. Mohammad, A. Aguirre, T. Wang, Y.
Yang, M.B. Smith, and Q. Zhu J. Biomed. Optics, 2013, 18(6),
066009-1-11.
19 S. Zanganeh, F. Zhou, A. Abuteen, I. Mohammad, M.B. Smith,
and Q. Zhu Biomedical Optics, 2014, p. BT3A. 51. Conference
presentation.
20 N. Nijegorodov and W. Downey J. Phys. Chem., 1994, 98,
5639-5643.
21 H. Suzuki, Bull. Chem. Soc. Jpn., 1959, 32, 1340-1350.
22 I. Mohammad, C. Stanford, M.D. Morton, Q. Zhu, and M.B.
Smith, Dyes Pigments, 2013, 99, 275-283.
23 O. Mazuryk, M. Maciuszek, G. Stochel, F. Suzenet, and M.
Brindell J. Inorg. Biochem., 2014, 134, 83-91.
24 K. Okuda, Y. Okabe, T. Kadonosono, T. Ueno, B.G. Youssif, S.
Kizaka-Kondoh, and H, Nagasawa Bioconjugate Chem., 2012,
23, 324-329.
25 J. Pauli, U. Resch-Genger, W.A. Kaiser, I. Hilger, F.M.
Hamann, R. Brehm, M. Grabolle, W. Frank, and D. Fischer
Mol. Imaging, 2011, 10, 258-269.
26 E. Jung and W.-J. Kim Biorg. Ed. Chem., 2006, 14, 92-97.
27 M. Grabolle, M. Spieles, V. Lesnyak, N. Gaponik, E.
Alexander, and U. Resch-Genger Anal. Chem., 2009, 81,
6285-6294.
28 R. Reisfeld Le Journal de Physique Colloques, 1987, 48(C7),
C7-423-C7-426.
29 S.K. Cool, K. Breyne, E. Meyer, S.C. De Smedt, and N.N.
Sanders J. Fluoresc. 2013, 23, 909-920.
This journal is © The Royal Society of Chemistry 20xx
Organic and Biomolecular Chemistry 2015, 00, 1-3 | 11
Please do not adjust margins