Synthesis and testing of beta-cell-specific streptozotoein-derived near-infrared imaging probes

Article abstract of DOI:10.1002/anie.200702183

Labeling diabetes: Streptozotocin-derived, cyanine-5.5-labeled imaging probes were synthesized and tested with the beta-cell-mimic INS-1E cell line in cell uptake assays, as well as by flow cytometry and confocal microscopy. Both probes showed superb labe

Full text of DOI:10.1002/anie.200702183

Communications
DOI: 10.1002/anie.200702183
Fluorescent Probes
Synthesis and Testing of Beta-Cell-Specific Streptozotocin-Derived
Near-Infrared Imaging Probes**
Chongzhao Ran, Pamela Pantazopoulos, Zdravka Medarova,* and Anna Moore*
Pancreatic beta-cells are responsible for
the production and release of insulin, a
hormone that controls the level of glucose
in the blood. Beta-cell mass is markedly
reduced in both insulin-dependent and
non-insulin-dependent diabetes melli-
tus.[1–5] The noninvasive estimation of
beta-cell mass by imaging would have a
significant impact on the management of
clinical diabetes, pancreas- and/or islet-
cell transplantation, and the understand-
ing of the pathogenesis of the disease.
Currently, measurements of insulin secre-
tion and c-peptide production serve as
surrogates for the assessment of beta-cell
function.[6] However, these methods are
not suitable for in vivo detection. Previ-
ously several attempts have been made by
Scheme 1. Synthesis of the STZ-derived Cy5.5-labeled probes for beta-cell detection.
a) 1. Anisaldehyde, NaOH; 2. acetic anhydride, pyridine; b) acetone, HCl; c) 4, dichloro-
us and other investigators to image pan-
creatic-islet inflammation, rejection, and
methane; d) 1. TFA; 2. Cy5.5–NHS; 3. NaOMe; e) tBuONO; f) 1. TFA; 2. Cy5.5–NHS;
transplantation.[7–13] Here, we describe the
3. NaOMe. Boc: tert-butoxycarbonyl; Cy5.5–NHS: cyanine-5.5–N-hydroxysuccinimide; TFA:
trifluoroacetic acid.
synthesis and testing of near-infrared
(NIR) probes for imaging beta-cells in
pancreatic islets, based on the beta-cell-
specific ligand streptozotocin (STZ), labeled with the fluo-
rescent dye cyanine-5.5 (Cy5.5).
Our probe design is based on the specificity of STZ for
beta-cells and the following well-known facts. STZ is an FDA-
approved drug for treating metastatic cancer of pancreatic-
islet cells.[14] However, in the diabetes-research community,
STZ is also widely used to establish mouse models of Type 1
diabetes, and this capability is ascribed to its specific toxicity
to murine beta-cells.[15,16] It is believed that the selective
targeting of STZ toward beta-cells is rooted in a specific
interaction with the GLUT2 transporter,[17,18] which is only
expressed in beta-cells (and not in alpha- or gamma-cells)
within the adult pancreas and is essential for its glucose-
sensing function.[19,20] Although the structure of STZ is very
similar to glucose, it is not recognized by other glucose
transporters such as GLUT1, GLUT3, and GLUT4.[17,18] This
selectivity may arise from the nitrosourea moiety of STZ,
which is the only part different from glucose. Therefore, in our
probes A and B (Scheme 1), the 2-nitrosourea-glucose moiety
of STZ was kept intact to maintain the probesꢀ specificity for
GLUT2 and beta-cells. On the other hand, in designing a
high-quality imaging probe, it is mandatory to avoid STZ
toxicity to beta-cells, which primarily arises from the alkyla-
tion of DNA and from the release of nitric oxide (NO).[21,22]
Interestingly, for establishing mouse models of Type 1 dia-
betes, a widely used protocol requires that STZ be freshly
pretreated in pH 4.5 citrate buffer. This requirement is
consistent with the tendency of alkyl-nitrosourea compounds
to generate nitric oxide (NO) and cause alkylation under
acidic conditions.[21,22] These properties of STZ and the fact
that it is an FDA-approved compound prompted us to
speculate that if STZ and its derivatives are used under
neutral conditions, their toxicity may decrease significantly
and they could therefore safely be used as imaging probes.
Moreover, to reduce the alkylation ability of STZ derivatives,
we proposed the use of a longer chain to replace the methyl
[*] Dr. C. Ran, P. Pantazopoulos, Dr. Z. Medarova, Prof. A. Moore
Molecular Imaging Laboratory
Athinoula A. Martinos Center for Biomedical Imaging
Massachusetts GeneralHospita/lHarvard MedicalSchool
13th Street, Building 149, Room 2301, Charlestown MA 02129
(USA)
Fax: (+1)617-726-7422
E-mail: zmedarova@partners.org
amoore@helix.mgh.harvard.edu
[**] Confocalmicroscopy was performed at the ConfocalMicroscopy
Core at Massachusetts GeneralHospitalwith technicalassistance
from Igor A. Bagayev, M.S. We would like to thank Prof. Claes B.
Wollheim (University Medical Center, Geneva, Switzerland) for
providing INS-1E cells.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
8998
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 8998 –9001
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Take advantage of blue reference links
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

Angewandte
Chemie
group of STZ, since a longer chain has a much lower
alkylating ability toward DNA.[23] In our probes A and B, the
butyl chain was chosen as both the replacement of the methyl
group and the linker between the STZ moiety and a
fluorescent dye. Furthermore, to observe the effects of the
=
nitroso (N O) group for specific targeting and toxicity, we
synthesized probe Awithout an NO group, while probe B had
an NO group attached. Since imaging in the NIR region has
minimal tissue autofluorescence, which dramatically
improves target/background ratios, we employed the well-
documented NIR dye Cy5.5–NHS as the fluorescent reporter
for our probes A and B.
To synthesize probes A and B, we started from glucos-
amine (1), which was converted into 2 by protecting the 2-
amino group with anisaldehyde and acetylating the 1,3,4,6-
hydroxy groups with acetic anhydride. After removal of the
anisylidene protecting group with hydrochloride solution,[24]
3
was obtained with a high yield and was treated with 4 to afford
the key intermediate 5. Synthesis of probe A was achieved by
removing the BOC group of 5 with TFA and by subsequent
reaction with Cy5.5–NHS and deacetylation in methanol. To
prepare probe B, 5 was first transformed into 6 with tBuONO
in dichloromethane, and then probe B was obtained from 6 by
a similar procedure to that used for A (Scheme 1).
To test the specificity of the synthesized probes, we chose
the insulinoma INS-1E cell line, which is known to mimic the
behavior of pancreatic beta-cells.[25] For controls, we used the
pancreatic adenocarcinoma CAPAN-2 cell line and the
embryonic kidney HEK 293 cell line, which is nontumori-
genic.
Our flow cytometry studies showed that probes A and B
labeled almost all INS-1E cells (92.5 and 96.0%, respectively)
but only 18.7 and 32.7% of CAPAN-2 cells and 18.3 and
20.1% of HEK293 cells, respectively (Figure 1). Confocal
microscopy corroborated the results of the fluorescence-
activated cell sort (FACS) and showed a significantly higher
uptake of both probes by INS-1E cells compared to that by
CAPAN-2 and HEK 293 cells and relative to the results with
Cy5.5 dye (see the Supporting Information).
Further studies with INS-1E cells demonstrated that both
probes were taken up by cells in a dose-dependent fashion
(Figure 2A) and that this uptake reached a plateau after 8 h
of incubation (Figure 2B). Since the specificity of STZ toward
beta-cells originates from the particular targeting of GLUT2
on the beta-cells,[17,18] it is essential to test whether the
modified-STZ-based probes A and B retain the specific
interaction with the GLUT2 transporter on INS-1E cells. We
conducted these experiments by incubating INS-1E cells with
the GLUT2 inhibitor phloretin[26] for 8 h and then with probe
A or B for 24 h. This was followed by measurement of the
NIR-fluorescence intensity in cell lysates or by flow cytom-
etry. There was a concentration-dependent inhibition of the
uptake of both probes. Flow cytometry revealed a significant
reduction of the mean fluorescence intensity following treat-
ment with phloretin (34.9% with probe A and 38.3% with
probe B; see the Supporting Information). From these results,
we concluded that the cellular uptake of the probes was
mediated by the GLUT2 transporter.
Figure 1. Cell labeling with probes A and B. Note that the probes
labeled almost all of the INS-1E cells.
Since the ultimate goal of our studies is to image
pancreatic islets in the native organ, we tested whether
islets could be labeled with the probes. To that end, we
incubated purified human islets with probe A or B for 24 h at
a concentration of 6.0 mm and then subjected the islets to
confocal microscopy after fixation in 4% formaldehyde.
Untreated human islets served as a control. As anticipated,
we observed a bright fluorescence signal consistent with
intracellular accumulation of both probes (Figure 3). These
results hold promise for further use of these probes in in vivo
animal model studies.
For the successful development of a molecular imaging
probe, low toxicity to the target is also a key factor. Since STZ
Angew. Chem. Int. Ed. 2007, 46, 8998 –9001
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8999
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Take advantage of blue reference links
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

Communications
is taken into account.[27,28] These results are most likely
indicative of the fact that STZ and probes A and B could be
safe if used under neutral conditions (pH 7.4), instead of with
a pH 4.5 citrate buffer pretreatment. Since the NO-releasable
probe B has a very low NO-generation potency under neutral
conditions, it is not surprising that both probes showed similar
results.
Cheng et al. reported that direct conjugation of Cy5.5 with
d-glucosamine at the 2-position produced Cy5.5–2DG con-
jugates, but these products did not show specific interaction
with GLUTs.[29] They ascribed this nonspecificity to the large
size of the Cy5.5–2DG conjugates. By contrast, our probes A
and B have displayed considerable selectivity for the GLUT2
transporter, probably due to the higher similarity of A and B
to the GLUT2-specific STZ molecule and the presence of a
longer linker between Cy5.5 and the STZ moiety in our
probes.
In summary, we have successfully developed two NIR
probes for specific labeling of beta-cells through their
interaction with the GLUT2 transporter. These two probes
have displayed dose- and time-dependent behavior, low
toxicity towards their target, and excellent selectivity for the
beta-cell-derived INS-1E cell line over other cell lines.
Furthermore, we confirmed that cell labeling by these
probes was mediated by GLUT2. In addition, our data have
demonstrated that our probes are not only able to label a
beta-cell-like cell line but also have the capability to label
human islets. We believe that our probes could be useful for
pancreas- and/or islet-cell transplantation, monitoring, and
the understanding of the pathogenesis of diabetes.
Figure 2. A) INS-1E cells take up probes A and B in a dose-dependent
fashion. B) Cellular uptake reached a plateau after eight hours of
incubation; mProbe: amount of probe added.
Experimental Section
Probes A and B: Triethylamine (40.4 mg, 0.4 mmol) and 4 (32.9 mg,
0.1 mmol) were added to a suspension of 3 (38.7 mg, 0.1 mmol) in an
acetonitrile/water (10:1) solvent mixture (5.0 mL), and the resulting
mixture was stirred at room temperature overnight. The solvent was
evaporated to give a semisolid residue, which was subjected to flash
silica gel chromatography with dichloromethane/ethanol (15:1) to
give 5 (54.8 mg, 97.6%). TFA (100 mL) was added to a solution of 5
(4.4 mg, 8.0 mmol) in dichloromethane (100 mL), and the resultant
mixture was stirred at room temperature for 2 h. After evaporation of
the solvent and TFA, an oily residue was obtained. Cy5.5–NHS
(1.0 mg), water (100 mL), acetonitrile (100 mL), and triethylamine
(10.0 mg) were added to this residue. The resulting mixture was
stirred at room temperature for 4.0 h, and this was followed by
addition of sodium methoxide (108.0 mg) in methanol (100 mL). The
resultant mixture was stirred at room temperature overnight, and the
solvent was removed under vacuum to give an oily residue. This oily
residue was subjected to a reversed-phase flash C18 column to give
probe A (0.6 mg, 55.3%); ESI-MS: m/z 1192.9. Probe B (55.8%) was
synthesized via 6 by a similar procedure to that for probe A; ESI-MS:
m/z 1222.3.
INS-1E, CAPAN-2, and HEK293 cells were seeded on a 12-well
plate and treated with increasing concentrations of probe A or B (for
cell uptake studies) or with a fixed concentration of A or B (for
confocal microscopy and flow cytometry (FACS)). For GLUT-2
specificity studies, the cells were pretreated with inhibitor (phloretin)
before addition of the probes. The cells were then washed and fixed
by using conventional procedures for FACS analysis and confocal
microscopy. Islet sample preparation was similar to the cell sample
preparation. Confocal microscopy was performed by using an
Figure 3. Confocalmicroscopy of human pancreatic isel ts treated with
probe A (A) and probe B (B). No signalin the Cy5.5 channelwas
observed in the control of nontreated islets (C). Blue: Cy5.5; red:
propidium iodide. Images were overplayed onto Nomarski optics;
scale bar: 100 mm.
is a known inducer of diabetes in mice, we tested whether our
probes caused toxicity in purified human pancreatic islets.
The results of our TUNEL assay (see the Supporting
Information) revealed that there was no significant cell
death after incubation with either of the probes. There was
no difference in the number of apoptotic cells between
untreated islets and islets treated with probe A or B. Control
islets treated with STZ had similar levels of apoptosis, which
is anticipated when the low STZ toxicity towards human islets
9000 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 8998 –9001
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Take advantage of blue reference links
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

Angewandte
Chemie
Axiovert 200M inverted microscope (Carl Zeiss). FACS analysis was
performed with a FACSCalibur instrument (Becton Dickinson, San
Diego, CA).
Purified human islets treated with probe A, probe B, or ST Z
(2 mm). After washing with phosphate-buffered saline and fixation in
4% formaldehyde, TUNEL staining was performed according to the
manufacturerꢀs protocol (DeadEnd Colorimetric TUNEL Assay,
Promega, Madison, WI). Untreated islets served as a control.
[10] S. J. Kim, D. J. Doudet, A. R. Studenov, C. Nian, T. J. Ruth, S. S.
Gambhir, C. H. McIntosh, Nat. Med. 2007, 12, 1423.
[11] M. Fowler, J. Virostko, Z. Chen, G. Poffenberger, A. Radhika,
M. Brissova, M. Shiota, W. E. Nicholson, Y. Shi, B. Hirshberg,
D. M. Harlan, E. D. Jansen, A. C. Powers, Transplantation 2005,
79, 768.
[12] N. V. Evgenov, Z. Medarova, G. Dai, S. Bonner-Weir, A. Moore,
Nat. Med. 2006, 12, 144.
[13] N. V. Evgenov, Z. Medarova, J. Pratt, P. Pantazopoulos, S.
Leyting, S. Bonner-Weir, A. Moore, Diabetes 2006, 55, 2419.
[14] R. Brentjens, L. Saltz, Surg. Clin. North Am. 2001, 81, 527.
[15] G. L. Wilson, E. H. Leiter, Curr. Top. Microbiol. Immunol. 1990,
156, 27.
Received: May 17, 2007
Revised: July 27, 2007
Published online: October 23, 2007
[16] K. R. Mansford, L. Opie, Lancet 1968, 291, 670.
[17] Z. Wang, H. Gleichmann, Diabetes 1998, 47, 50.
[18] W. J. Schnedl, S. Ferber, J. H. Johnson, C. B. Newgard, Diabetes
1994, 43, 1326.
[19] L. Orci, B. Thorens, M. Ravazzola, H. F. Lodish, Science 1989,
245, 295.
[20] G. W. Gould, G. D. Holman, Biochem. J. 1993, 295 (Part 2), 329.
[21] R. A. Bennett, A. E. Pegg, Cancer Res. 1981, 41, 2786.
[22] N. S. Kwon, S. H. Lee, C. S. Choi, T. Kho, H. S. Lee, Faseb J.
1994, 8, 529.
[23] R. Duncan, G. Davies, Int. J. Quantum Chem. 1989, 35, 665.
[24] A. Kyas, M. Feigel, Helv. Chim. Acta 2005, 88, 2375.
[25] A. Merglen, S. Theander, B. Rubi, G. Chaffard, C. B. Wollheim,
P. Maechler, Endocrinology 2003, 145, 667.
[26] J. M. Duran, M. Cano, M. J. Peral, A. A. Ilundain, Glycobiology
2004, 14, 495.
[27] B. Tuch, J. Turtle, C. Simeonovic,Diabetologia 1989, 32, 678.
[28] M. Elsner, M. Tiedge, S. Lenzen, Diabetologia 2003, 46, 1713.
[29] Z. Cheng, J. Levi, Z. Xiong, O. Gheysens, S. Keren, X. Chen, S. S.
Gambhir, Bioconjugate Chem. 2006, 17, 662.
Keywords: cells · diabetes · fluorescent probes · imaging ·
streptozotocin
.
[1] S. E. Kahn, Am. J. Med. 2000, 108 (Suppl 6a), 2S.
[2] W. Gepts, P. M. Lecompte, Am. J. Med. 1981, 70, 105.
[3] G. Sesti, Ann. Med. 2002, 34, 444.
[4] K. Otani, R. N. Kulkarni, A. C. Baldwin, J. Krutzfeldt, K. Ueki,
M. Stoffel, C. R. Kahn, K. S. Polonsky, Am. J. Physiol. Endo-
crinol. Metab. 2003, 286, 41E.
[5] C. J. Rhodes, Science 2005, 307, 380.
[6] M. O. Larsen, B. Rolin, M. Wilken, R. D. Carr, C. F. Gotfredsen,
Diabetes 2003, 52, 118.
[7] A. Moore, J. Grimm, B. Han, P. Santamaria, Diabetes 2004, 53,
1459.
[8] Y. Lu, H. Dang, B. Middleton, Z. Zhang, L. Washburn, M.
Campbell-Thompson, M. A. Atkinson, S. S. Gambhir, J. Tian,
D. L. Kaufman, Mol. Ther. 2004, 9, 428.
[9] X. Chen, X. Zhang, C. S. Larson, M. S. Baker, D. B. Kaufman,
Transplantation 2006, 81, 1421.
Angew. Chem. Int. Ed. 2007, 46, 8998 –9001
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 9001
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&
&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
&
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
Take advantage of blue reference links
&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&