Nuclear Medicine and Biology 33 (2006) 945 – 952

www.elsevier.com/locate/nucmedbio

[99mTc]HYNIC-RGD for imaging integrin avh3 expression

Clemens Decristoforoa,4, Bluma Faintuch-Linkowskib, Ana Reyc, Elisabeth von Guggenberga,
Marco Ruppricha, Ignacio Hernandez-Gonzalesd, Teodoro Rodrigob, Roland Haubnera
a
Department of Nuclear Medicine, Medical University of Innsbruck, A-6020 Innsbruck, Austria
b
Radiopharmacy Center Instituto de Pesquisas Energeticas e Nucleares (IPEN), 05508-900 São Paulo, Brazil
c
Cátedra de Radioquı́mica, Facultad de Quı́mica, Universidad de la República, C.P. 11800 Montevideo, Uruguay
d
Centro de Isotopos, Ciudad de la Habana, Cuba
Received 9 June 2006; received in revised form 4 August 2006; accepted 2 September 2006

Abstract

There has been increasing interest in peptides containing the Arg–Gly–Asp (RGD) sequence for targeting of avh3 integrins to image
angiogenesis. [18F]Galacto-RGD has been successfully used for positron emission tomography applications in patients. Here we report on the
preclinical characterization of a 99mTc-labeled derivative for single-photon emission computed tomography.
c(RGDyK) was derivatized with HYNIC at the q-amino group of the lysine [c(RGDyK(HYNIC)) or HYNIC-RGD]. 99mTc labeling was
performed using coligands (tricine and EDDA), as well as 99mTc(CO)3(H2O)3. Radiolabeled peptides were characterized with regard to
lipophilicity, protein binding and stability in buffer, serum and tissue homogenates. Integrin receptor activity was determined in
internalization assays using avh3-receptor-positive M21 and avh3-receptor-negative M21L melanoma cells. Biodistribution was evaluated in
normal and nude mice bearing M21, M21L and small cell lung tumors.
HYNIC-RGD could be labeled at high specific activities using tricine, tricine–trisodium triphenylphosphine 3,3V,3W-trisulfonate (TPPTS),
tricine–nicotinic acid (NA) or EDDA as coligands. [99mTc]EDDA/HYNIC-RGD, [99mTc]tricine–TPPTS/HYNIC-RGD and [99mTc]tricine–
NA/HYNIC-RGD showed protein binding ( b 5%) considerably lower than [99mTc](CO)3-HYNIC-RGD and [99mTc]tricine/HYNIC-RGD.
[99mTc]EDDA/HYNIC-RGD revealed high in vitro stability accompanied by low lipophilicity with a log P value of 3.56, comparable to
that of [18F]Galacto-RGD. In M21 cells for this compound, the highest level of specific and rapid cell uptake (1.25% mg protein 1) was
determined. In vivo, rapid renal excretion, low blood retention, low liver and muscle uptakes and low intestinal excretion 4 h postinjection
were observed. Tumor uptake values were 2.73% ID/g in M21 avh3-receptor-positive tumors versus 0.85% ID/g in receptor-negative tumors
1 h postinjection. Small cell lung tumors could be visualized using g camera imaging.
[99mTc]EDDA/HYNIC-RGD shows encouraging properties to target avh3 receptors in vivo with high stability and favorable
pharmacokinetics. Tumor uptake studies showed specific targeting of avh3-receptor-positive tumors with tumor-to-organ ratios comparable
to those of [18F]Galacto-RGD.
D 2006 Elsevier Inc. All rights reserved.
Keywords: RGD; Technetium-99m; HYNIC; avh3

1. Introduction patients [1]. This success is based on high affinity for the
target receptor, good internalization and retention in cells
Radiolabeled peptides, such as somatostatin, gastrin-
and rapid predominant renal excretion from nontarget
releasing peptide and gastrin analogues, have been success-
tissues [2]. However, the use of these peptides is limited
fully used to image G-protein-coupled receptors in tumor
by great differences in the expression of those regulatory
peptide receptors in tumor tissues [3].
Recently there has been interest in targeting integrin
4 Corresponding author. Universitaetsklinik fuer Nuklearmedizin,
Anichstrasse 35, A-6020 Innsbruck, Austria. Tel.: +4351250480951; fax:
receptors for in vivo tumor imaging [4]. Especially integrin
+4351250422659. avh3, which mediates binding to the extracellular matrix via
E-mail address: clemens.decristoforo@uibk.ac.at (C. Decristoforo). a variety of different proteins (e.g., vitronectin, fibronectin
0969-8051/$ – see front matter D 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.nucmedbio.2006.09.001
946 C. Decristoforo et al. / Nuclear Medicine and Biology 33 (2006) 945–952

2. Materials and methods

2.1. Materials
Reagents were purchased from Aldrich-Sigma Chemical
Co., except when otherwise stated, and used as received.
HYNIC-RGD (cyclo[Arg–Gly–Asp–d-Tyr–(Lys-
HYNIC)]; HYNIC = 2-hydrazinonicotinic acid) and
c(RGDyK) were synthesized on solid support using
standard F-moc solid-phase chemistry (Biosynthan, Berlin,
Germany) with a purity of N95%, as analyzed by reverse-
phase high-performance liquid chromatography (RP-HPLC)
and mass spectroscopy (MS).
[125I]Echistatin was purchased from Amersham Life
Fig. 1. Structure of HYNIC-RGD (cyclo[Arg–Gly–Asp–d-Tyr–(Lys- Science (Boston, MA).
HYNIC)]) and the proposed 99mTc complex using EDDA as coligand. Na99mTcO4 was obtained from a commercial 99Mo/ 99mTc
generator (ULTRATECHNEKOW, Mallinckrodt,
The Netherlands).
and von Willebrand factor), is a target structure for the
development of this new class of radiopharmaceuticals. This 2.2. Analytical methods
integrin is involved in different pathological processes, such
as tumor metastasis and tumor-induced angiogenesis, 2.2.1. HPLC
restenosis, osteoporosis and inflammatory processes [5]. A Dionex P680 low-pressure gradient pump with a
For example, several studies have shown that there is a Spectra Physics Spectra Chrom 100 variable UV detector
correlation between avh3 expression and the metastatic and a Bioscan radiometric detector was used for RP-HPLC
potential of the corresponding tumor [6]. Moreover, this analysis. A Bischof Nucleosil 100-5 C18 2504.6-mm
integrin is expressed on activated endothelial cells during column, at a flow rate of 1 ml/min and with UV detection
migration through the basement membrane in the angio- at 220 nm, was employed with the following gra-
genic process. Thus, several approaches have focused on the dients: Method 1: ACN/0.1% TFA/H2O: 0–1.5 min, 0%
inhibition of tumor-induced angiogenesis via the targeting ACN; 1.5–18 min, 0–30% ACN; 18–21 min, 30–60%
of this receptor. Development of inhibitors is based on ACN; 21–24 min, 60% ACN.
peptides containing the amino acid sequence Arg–Gly–Asp
99m
(RGD), which is found to be essential in the binding of a 2.3. Tc labeling
variety of extracellular matrix proteins. The cyclic penta-
peptide c(RGDfV) [7] has been proven to be selective for 2.3.1. Tricine as coligand
integrin avh3 and, thus, is used as a lead structure for tracer In a rubber-sealed vial, 5 Ag of HYNIC-RGD was
development [8]. Meanwhile, RGD peptides labeled with incubated with 0.5 ml of tricine solution (70 mg/ml in
111
In, 90Y, 99mTc, 123I, 64Cu and 18F have been introduced water), 0.5 ml of 99mTcO4 solution (N 200 MBq) and 20 Al
[4,5]. Most of them show high avh3 affinity in vitro and of tin(II) solution (10 mg of SnCl2d 2H2O in 10 ml of
allow targeting of receptor-positive tumors in vivo. Major nitrogen-purged 0.1 N HCl) for 20 min at room temperature.
differences are found in the pharmacokinetics of different
tracers. The most intensively studied compound yet is 2.3.2. Tricine ternary coligands
[18F]Galacto-RGD, which allows noninvasive determination Nicotinic acid (NA) and trisodium triphenylphosphine
of avh3 expression not only in murine tumor models but 3,3V,3U-trisulfonate (TPPTS) were used as ternary coligands.
also in tumor patients [9]. In a rubber-sealed vial, 5 Ag of HYNIC-RGD was heated
However, 99mTc labeling is still an attractive approach for with 0.5 ml of tricine solution (70 mg/ml in water), 0.5 ml of
99m
radiolabeling peptides for single-photon emission computed TcO4 solution ( N200 MBq), 20 Al of tin(II) solution
tomography (SPECT) imaging, and a number of labeling (10 mg of SnCl2d 2H2O in 10 ml of nitrogen-purged
technologies have been successfully used in patients [10]. 0.1 N HCl) and 0.1 ml of NA solution (40 mg/ml in water)
Our group has shown that the use of the HYNIC approach is or 0.2 ml of TPPTS solution (5 mg/ml in water) for 15 min
suitable for radiolabeling peptides with high specific 1008C (NA) or for 30 min at 508C (TPPTS).
activity, resulting in suitable pharmacokinetics for imaging
purposes [11,12]. 2.3.3. Tricine–EDDA exchange labeling
Here we describe the radiolabeling of a HYNIC- In a rubber-sealed vial, 10 Ag of HYNIC-RGD was
derivatized monomeric RGD peptide [c(RGDyK(HYNIC)) incubated with 1 ml of EDDA/tricine solution (20 mg/ml
or HYNIC-RGD; Fig. 1] and its in vitro and in vivo tricine, 10 mg/ml EDDA in pH 6–7), 1 ml of 99mTcO4
evaluation for targeting the avh3 receptor in vivo. solution (800 MBq) and 20 Al of tin(II) solution (10 mg of
 C. Decristoforo et al. / Nuclear Medicine and Biology 33 (2006) 945–952 947

SnCl2d 2H2O in 10 ml of nitrogen-purged 0.1 N HCl) for chased from Amersham-Pharmacia Biotech (Vienna, Aus-
10 min at 1008C. tria), and avh3 integrin receptors were from Chemicon
(Temecula, CA). Briefly, 96-well plates were coated with
2.3.4. Tc carbonyl avh3 integrin receptors and incubated with a mixture of
Carbonyl aquaion [99mTc(OH2)3(CO)3]+ was prepared [125I]echistatin and varying concentrations of the peptide.
from a commercial kit (ISOLINK, Mallinckrodt, Petten) Unbound radioligand was removed, and wells were washed.
according to the manufacturer’s instructions (1 ml volume; Bound radioligand was removed with 2 N NaOH. IC50 was
200–1000 MBq 99mTc). Five or 20 Ag of peptide HYNIC- calculated by fitting percent inhibition values using ORIGIN
RGD was incubated with 100 Al of [99mTc(OH2)3(CO)3]+ software (Northampton, MA).
labeling solution (up to 400 MBq) for 30 min at 708C.
2.5. Internalization and binding studies in avb3-positive and
avb3-negative cells
2.4. In vitro evaluation of radiolabeled peptides
avb3-Positive M21 and avb3-negative M21L human
2.4.1. Stability melanoma cells were grown in culture until a sufficient
The stability of radiolabeled peptides in aqueous solution number of cells were available. For internalization ex-
was tested by the incubation of the reaction mixture purified periments, cells were collected up to a concentration of
by solid-phase extraction (SPE) using a SEPPAK C18 2106 cells/ml in RPMI 1640 containing 1% glutamine and
cartridge (Waters, USA) at a concentration of 200– 1% PBS. Each 1 ml was pipetted into a separate tube. After
1000 pmol/ml peptide in phosphate buffer, as well as in a the addition of 99mTc-labeled peptide ( N 100,000 cpm,
solution containing 10,000-fold molar excess of cysteine or 1 nM), cells were incubated at 378C for 90 min in triplicate
histidine compared to the peptide at pH 7.4 at 378C up to with either PBS/0.5% BSA buffer alone (150 ll, total series)
24 h. Stability in fresh human plasma, as well as in kidney or with 10 lM c(RGDyK) in PBS/0.5% BSA buffer (150 ll,
and liver homogenates, was measured in parallel. After nonspecific series). Incubation was interrupted by centrifu-
incubation, plasma samples were precipitated with acetoni- gation, removal of medium and rapid rinsing with ice-cold
trile and centrifuged (1750g, 5 min). Degradation of 99mTc Tris-buffered saline twice. Thereafter, the cells were
complexes was assessed by HPLC. For incubation in kidney incubated twice at room temperature in acid wash buffer
and liver homogenates, kidneys or livers freshly excised (50 mM acetate buffer, pH 4.2) for 15 min at 378C. The
from rats were rapidly rinsed and homogenized in 20 mM supernatant was collected (membrane-bound radioligand
HEPES buffer (pH 7.3) with an Ultra-Turrax T25 homoge- fraction), and the cells were washed with acid wash buffer.
nator for 1 min at room temperature. Radiopeptides were Cells were lysed by treatment in 1 N NaOH, and cell
incubated with fresh 30% homogenate at a concentration of radioactivity was measured (internalized radioligand frac-
250–500 pmol/ml peptide at 378C up to 2 h. Samples were tion). Protein concentration in NaOH fraction was deter-
precipitated with acetonitrile, centrifuged (1750g, 5 min) mined using spectrophotometric determination (Bredford
and analyzed by HPLC. Assay, Sigma Aldrich Chemical Co.). Internalized and
noninternalized fractions were determined by measuring
2.4.2. Protein binding radioactivity, and internalized fraction was expressed as the
For protein-binding assessment, SPE-purified complexes percentage of total activity per milligram of protein.
were incubated at a concentration of 20–100 pmol/ml
2.6. In vivo evaluation of radiolabeled peptides
peptide in fresh human plasma at 378C and analyzed by
size exclusion chromatography (MicroSpin G-50 Columns; All animal experiments were conducted in compliance
Sephadex G-50) up to 6 h. Protein binding of the 99mTc with Austrian animal protection laws and with the approval
complex was determined by measuring columns and eluates of the Austrian Ministry of Science.
in a g counter (Compugamma LKB, Finland).
2.6.1. Human melanoma model
2.4.3. Log P values Tumor uptake studies were performed in male nu/nu
99m
Tc-labeled HYNIC-RGD (0.5 ml) in phosphate-buff- mice (Charles River, Germany). For the induction of tumor
ered saline (PBS) was added to 0.5 ml of octanol in an Eppen-
dorf vial. The tube was vigorously vortexed over a period Table 1
99m
In vitro data as determined for all Tc-labeled HYNIC-RGD conjugates
of 15 min and centrifuged at 5000g for 3 min. A 100 Al
Compound t R (min) Log P Labeling yield
aliquot of both aqueous and octanol layers was collected and
O/W (n = 3) (%)
counted in a g counter, and log P values were calculated.
[99mTc]EDDA/HYNIC-RGD 16.5 3.57 93.9
[99mTc]Tricine/HYNIC-RGD 16.3 3.75 99.5
2.4.4. Binding affinity [99mTc]Tricine–NA/HYNIC-RGD 17.1 3.53 99.8
The in vitro binding affinity of HYNIC-RGD was [99mTc]Tricine–TPPTS/HYNIC-RGD 15.2 4.53 99.7
determined in comparison with that of c(RGDfV) using [99mTc](CO)3/HYNIC-RGD 19.3 1.69 99.6
[125I]Echistatin as radioligand. [125I]Echistatin was pur- O/W, octanol/water.
948 C. Decristoforo et al. / Nuclear Medicine and Biology 33 (2006) 945–952

injected subcutaneously on the right upper back with a

suspension of 6.5106 cells. Tumor-bearing mice were used
in biodistribution and imaging studies. For biodistribution,
mice were injected, sacrificed and dissected as described
above. Results were expressed as the percentage of injected
dose per gram of tissue, and tumor-to-organ ratios were
calculated.
Planar g camera imaging was performed with a mobile
scintillation camera (LEM Siemens) equipped with a 4.0-mm
pinhole collimator. A syringe with a known amount of
radioactivity was scanned, along with the mice, to allow
semiquantification of the results using region-of-interest
(ROI) analysis. All animals were injected intravenously with
0.1 ml (37 MBq, 1 Ag of peptide) of [99mTc]EDDA/HYNIC-
RGD, with and without excess c(RGDyK) (100 Ag) in 0.9%
saline, via the lateral tail vein. Four hours postinjection, the
animals were anesthetized by intraperitoneal injection with
urethane solution. A total of 500,000 counts were obtained
per projection, under a 6464 matrix. Ten-minute images of
60 s were acquired, which were drawn later using magnetic
resonance imaging maximum intensity projection.

3. Results
HYNIC-RGD could be labeled at high specific activities
(100 GBq/Amol) using EDDA, tricine, tricine–NA or
tricine–TPPTS as coligands, resulting in a single major
species as analyzed by HPLC. Corresponding data are
summarized in Table 1. Mean labeling yields (n = 3) were
93.9% for [99mTc]EDDA/HYNIC-RGDyK and N 99% for
Fig. 2. Radiochromatograms: (A) [99mTc]EDDA/HYNIC-RGD; (B)
other coligands. Labeling using carbonyl aquaion
[99mTc]tricine/HYNIC-RGD; (C) [99mTc]tricine–NA/HYNIC-RGD; (D)
[99mTc]tricine–TPPTS/HYNIC-RGD; (E) [99mTc](CO)3/HYNIC-RGD; (F) [99mTc(OH2)3(CO)3] also resulted in high labeling yields
[99mTc]EDDA/HYNIC-RGD in liver homogenates at 2 h. (99.6%; n = 3), but multiple species with a retention time
significantly higher than the main peak (t R = 19.3 min) were
detected (representative chromatograms of all 99mTc-labeled
xenografts, M21 and M21L cells were subcutaneously peptides are shown in Fig. 2).
injected at a concentration of 5106 cells/mouse and
allowed to grow until tumors of 0.3– 0.6 ml were visible.
On the day of the experiment, each of the six mice with M21
tumors and three mice with M21L tumors were injected
with [ 9 9 m Tc]EDDA/HYNIC-RGD (1 MBq/mouse,
corresponding to 1 lg of peptide) intravenously into the
tail vein. They were sacrificed by cervical dislocation 1 h
(M21 and M21L) or 4 h postinjection (M21).
Tumors and normal tissues (blood, lung, heart, stomach,
spleen, liver, pancreas, kidneys, muscle and intestines) were
removed. The amount of radioactivity was determined using
a c counter. Results were expressed as the percentage of
injected dose per gram of tissue (% ID/g), and tumor-to-
organ ratios were calculated.

2.6.2. Non small cell lung cancer model

A549 cells, a human non small cell lung carcinoma cell
line, were grown to confluency and then harvested by
trypsinization. After centrifugation for 5 min at 1000g, the Fig. 3. Plasma protein binding of [99mTc]HYNIC-RGD derivatives
cells were resuspended in PBS. Male nu/nu mice were under study.
 C. Decristoforo et al. / Nuclear Medicine and Biology 33 (2006) 945–952 949

Table 3
Biodistribution of [99mTc]EDDA/HYNIC-RGD in Balb/c nu/nu mice
carrying A549 tumors with tumor-to-organ ratios in comparison with
normal mice 4 h postinjection
Organ A549 Tumor-to-organ ratio Normal
Blood 0.08F0.02 19.25 0.07F0.03
Heart 0.43F0.09 3.58 0.41F0.04
Stomach 1.39F0.06 1.11 1.12F0.13
Liver 1.45F0.09 1.06 1.53F0.25
Kidneys 2.61F0.12 0.59 2.45F0.15
Muscle 0.20F0.01 7.70 0.27F0.02
Small intestine 1.36F0.02 1.13 1.57F0.02
Large intestine 1.83F0.07 0.84 1.48F0.12
Tumor 1.54F0.28 1.00 NA

[ 99m Tc]EDDA/HYNIC-RGD, [ 99m Tc]tricine–TPPTS/

HYNIC-RGD and [ 99mTc]tricine–NA/HYNIC-RGD is
Fig. 4. Cell uptake of [99mTc]HYNIC-RGD using various coligands in shown in Fig. 4. Cellular uptake for [99mTc]HYNIC-RGD
avh3-integrin-receptor-positive M21 melanoma and receptor-negative
declined in the series EDDA (1.25% mg protein 1) N
M21L cells (blocked: coincubation with 10 AM RGD).
tricine–TPPTS (0.94% mg protein 1) Ntricine–NA (0.63%
mg protein 1). For all compounds, uptake could be re-
Stability studies in buffer, cysteine solution, human duced to b0.1% mg protein 1 by incubating with excess
plasma or tissue homogenates revealed a high stability of c(RGDyK). A comparably low uptake ( b 0.1% mg
all complexes prepared, with no significant release of protein 1) was found for receptor-negative M21L cells.
radiolabel or peptide degradation during the observation Based on these results, [99mTc]EDDA/HYNIC-RGD was
period (for [99mTc]EDDA/HYNIC-RGD in liver homoge- chosen for further in vivo evaluation in avh3-receptor-
nates, see also Fig. 2). positive and avh3-receptor-negative tumor-bearing mice.
Protein-binding values are shown in Fig. 3. Very low Biodistribution data in the M21 and M21L tumor models
levels of protein binding ( b5% after 3 h incubation) were are summarized in Table 2. Overall, a rapid almost exclusive
found using EDDA, tricine–TPPTS and tricine–NA as elimination via the renal pathway could be observed with
coligands, whereas high levels above 40% were determined retention in all organs lower than 4% ID/g after 1 h and
for tricine as coligand. HYNIC-RGD labeling using the lower than 3% ID/g after 4 h. The highest values were found
Tc(CO)3 core revealed 25.3% protein binding after 30 min, in the kidneys, with 3.67% versus 2.81% (1 vs. 4 h). Uptake
increasing to 38.9% after 3 h. All 99mTc-labeled HYNIC- in the liver and intestines was low, with 2.62% and 2.04%
RGD conjugates revealed a higher hydrophilicity, with a log after 1 h and with 1.71% and 1.28% after 4 h, respectively;
P value of b 3.5 compared to 3.2 of [18F]Galacto-RGD, blood levels of 0.97% at 1 h and 0.45% after 4 h indicate
except 99mTc(CO)3/HYNIC-RGD; log P values correlated moderate retention in the circulation. Tumor uptake values
with HPLC retention times (see Table 1). A high bind-
ing affinity to avh3 integrin was determined for HYNIC-
RGD with an IC50 value of 6.0 nM, in comparison to
c(RGDfV) with an IC50 value of 3.7 nM in the same assay.
Uptake in avh3-receptor-positive M21 melanoma cells for

Table 2
Biodistribution of [99mTc]EDDA/HYNIC-RGD in Balb/c nude mice
carrying avh3-positive tumors (M21) 1 and 4 h postinjection and avh3-
negative tumors (M21L) 1 h postinjection
Organ M21 M21 M21L
1h 4h 1h
Blood 0.98F0.04 0.45F0.02 1.27F0.16
Fig. 5. Planar g camera images of [99mTc]EDDA/HYNIC-RGD in
Heart 1.01F0.05 0.64F0.09 1.00F0.05
A549 tumor-bearing mice 4 h postinjection, without (left) and with
Stomach 2.50F0.75 1.28F0.22 2.68F0.67
(right) blocking receptors, by coinjection of 100 Ag of c(RGDyK). Tumors
Liver 2.63F0.14 1.71F0.28 2.43F0.11
on the left upper quadrant of the back are clearly delineated. Uptake, as
Kidneys 3.68F0.13 2.81F0.37 3.44F0.10
calculated with ROI technique, was 4.92% (left; unblocked) and 0.88%
Muscle 0.76F0.69 0.34F0.05 0.54F0.10
(right; blocked), respectively, showing receptor-specific uptake in the
Intestine 2.04F0.18 1.29F0.25 1.82F0.01
tumor. The main activity is shown in the bladder, indicating predominant
Tumor 2.73F0.26 2.06F0.37 0.85F0.20
renal excretion.
950 C. Decristoforo et al. / Nuclear Medicine and Biology 33 (2006) 945–952

of 2.72% and 2.06% after 1 and 4 h, respectively, indicate

good uptake and retention. This uptake was shown to be
specific as receptor-negative M21L tumor-bearing mice
showed a significantly ( P b.05) lower value of 0.85%; in
none of the organs was any significant difference between
M21 and M21L tumor observed.
Results of a comparative study in nude mice carrying a
human small cell lung cancer with normal mice are
summarized in Table 3. Both in normal and in A549
tumor-bearing mice, the organ with the highest tracer uptake
was the kidney. Intestinal excretion and liver uptake were
low. Tracer uptake in the tumor reached 1.5% 4 h postin-
jection. Comparative imaging in two A549 tumor-bearing
mice with and without coinjection of excess c(RGDyK) for
receptor blockade is shown in Fig. 5. The predominant renal
excretion pathway is clearly seen with high activity values
in the bladder, whereas liver uptake and kidney retention are
low. The tumor is clearly delineated, and uptake as
calculated from ROI technique was 4.92% (unblocked)
and 0.88% (blocked), respectively, showing receptor-spe-
cific uptake in the tumor.
Fig. 6. Tumor-to-organ ratios of [99mTc]EDDA/HYNIC-RGD in compar-
ison with [18F]galacto-RGD in a avh3-receptor-positive nude mouse M21
4. Discussion and conclusion human melanoma tumor model.

We attempted to label a peptide targeting the avh3

integrin with 99mTc for potential imaging of tumors and resulted in high labeling yields. However, multiple species
tumor-induced angiogenesis. Technetium is still the most with higher lipophilicity, compared to the method using
widely available diagnostic radionuclide with optimal coligands and high plasma protein binding, was observed;
physical characteristics for SPECT. Although 18F-labeled therefore, this radiolabeling route was not further pursued.
or 68Ga-labeled derivatives will be of great value for HYNIC-RGD showed retained binding affinity to the
positron emission tomography with advantages in terms of receptor compared to a standard monomeric RGD peptide
sensitivity and spatial resolution, 99mTc-labeled derivatives [c(RGDfV)] and, when radiolabeled, specific binding and
would be more widely available and clinically applicable. internalization in avh3-receptor-expressing cells. Compar-
We chose c(RGDyK) as lead structure, as it has shown ing various stable coligand systems, EDDA showed cellular
suitable properties when derivatized with DTPA and labeled uptake higher than those of both tricine–TPPTS and tricine–
with 111In [13]. Derivatization with HYNIC at lysine residue NA ternary coligand systems. The differences between these
allowed selective labeling with 99mTc. We have recently coligand systems seem to be not very pronounced in vitro,
used EDDA as coligand in the labeling of HYNIC, showing however, based on these findings [99mTc]EDDA/HYNIC-
advantages over other coligands in terms of peptide labeling RGD was selected for further in vivo evaluation.
[14]. Using a tricine exchange labeling approach [15], high A tumor model [99mTc]EDDA/HYNIC-RGD revealed
and reproducible labeling yields were achieved. Confirming high uptake in avh3-receptor-positive and low uptake in
previous findings, EDDA as coligand showed considerable negative tumor xenografts, proofing specific receptor-
advantages over tricine as a coligand for labeling HYNIC- mediated uptake in vivo. The same model has also been
RGD especially concerning lower values of plasma protein used in the development of glycosylated RGD derivatives
binding. We also tested two ternary coligand systems using [16], of which [18F]Galacto-RGD has been successfully
tricine and a monodentate ligand, a triphenylphosphine used to visualize avh3 integrin expression in patients. Fig. 6
(TPPTS) and a pyridine derivative (NA). Radiolabeling compares the tumor-to-organ ratios (1 h postinjection) of
yields were quantitative ( N99%); however, in some cases, these two compounds. Whereas tumor-to-organ ratios of
residual amounts of [99mTc]tricine/HYNIC-RGD were [99mTc]EDDA–HYNIC-RGD were lower for the blood,
detected. For both compounds, protein binding, compared heart and intestines, comparable values were found for the
to tricine alone, reached levels almost as low as that with liver, kidneys and muscle. Considering the importance of
EDDA as coligand. Especially [99mTc]tricine–TPPTS/ low intestinal excretion and low kidney retention for
HYNIC-RGD showed a very high hydrophilicity, with a radiolabeled peptides for diagnostic purposes, these results
log P of b 4. seem to be promising in relation to the only RGD peptide
Although HYNIC cannot be considered as an optimal evaluated extensively in human studies so far. These data
ligand for the Tc(CO)3 core, the labeling of HYNIC-RGD were confirmed using an alternative small cell lung cancer
 C. Decristoforo et al. / Nuclear Medicine and Biology 33 (2006) 945–952 951

model showing very similar pharmacokinetics with pre- applications seem to be clinically interesting (e.g., inflam-
dominant renal excretion and low retention in the liver, matory processes inducing angiogenesis with concomitant
blood and muscle, as well as comparable lower tumor avh3 overexpression might be a useful application for
uptake values. Imaging studies proved the receptor-mediat- radiolabeled RGD peptides) [27]. Clinical studies with this
ed tumor uptake in this model. compound should reveal the true potential of this new
99m
In addition, other groups have reported on HYNIC- Tc-labeled peptide.
derivatized RGD peptide analogues for labeling with 99mTc.
Early attempts using direct 99mTc labeling approaches were
Acknowledgments
reported without showing specific in vivo uptake [17,18]. An
attempt to use an N3S-type ligand for labeling a cyclic RGD This work was part of an IAEA (International Atomic
monomer resulted in a 99mTc-labeled peptide with receptor- Energy Agency)-coordinated research project bDevelopment
mediated tumor uptake but suboptimal pharmacokinetics of Tc-99m-Based Small Biomolecules Using Novel 99mTc
with especially very high kidney uptake and retention [19]. CoresQ. Special thanks go to Drs. D.V.S. Narasimhan and
Su et al. [20] used a cyclic derivative from a Phage M.R.A. Pillai (Department of Nuclear Sciences and Appli-
display library. They only found moderate tumor uptake, but cations, IAEA) for their considerable effort into the success
could not prove specific avh3-mediated accumulation, of this coordinated research project. We thank Dr. D.A.
possibly related to the low affinity of their peptide [21]. Cheresh (The Scripps Research Institute, La Jolla, CA) for
However, they found improvement in terms of protein supplying the M21 and M21L cells.
binding when using EDDA instead of tricine as coligand We thank Stephan Schwarz and Misley Wambrug
[22]. Janssen et al. [23] used HYNIC-c(RGDfK), an Altanes for their excellent technical assistance, and Prof.
analogue closely related to our tracer. For radiolabeling, Irene Virgolini for her support.
they applied the ternary ligand approach, with TPPTS as
monodentate ternary ligand. No in vitro data concerning cell
uptake or any comparative data with other coligands were References
provided, in contrast to our study where EDDA as a [1] Riccabona G, Decristoforo C. Peptide targeted imaging of cancer.
coligand showed the highest internalization values of Cancer Biother Radiopharm 2003;18:675 – 87.
different coligand systems. They found high tumor uptake [2] Heppeler A, Froidevaux S, Eberle AN, Maecke HR. Receptor
targeting for tumour localisation and therapy with radiopeptides. Curr
in a human ovarian cancer (OVCAR3) tumor model and
Med Chem 2000;7:971 – 94.
compared the monomeric derivative with a dimeric ana- [3] Reubi JC. Peptide receptors as molecular targets for cancer diagnosis
logue. Although tumor-to-organ ratios were superior for the and therapy. Endocr Rev 2003;24:389 – 427.
monomer, they described advantages of the dimer due to [4] Haubner R, Wester HJ. Radiolabeled tracers for imaging of tumour
higher tumor uptake and retention. However, they failed to angiogenesis and evaluation of anti-angiogenic therapies. Curr Pharm
Des 2004;10:1439 – 55.
show any specific tumor uptake in any of their models (i.e.,
[5] Haubner R. aı́h3-Integrin imaging: a new approach to characterize
by using blocking experiments, receptor-negative control angiogenesis? Eur J Nucl Med Mol Imaging 2006;33(Suppl 13):
tumors or control peptides without receptor affinity). In our 54 – 63.
study, we demonstrated specific receptor-mediated uptake of [6] Felding-Habermann B, Mueller BM, Romerdahl CA, Cheresh DA.
[99mT]EDDA/HYNIC-RGD in a murine tumor model with Involvement of integrin alpha V gene expression in human melanoma
tumourigenicity. J Clin Invest 1992;89:2018 – 22.
avh3-receptor-positive and avh3-receptor-negative tumors.
[7] Aumailley M, Gurrath M, Muller G, Calvete J, Timpl R, Kessler H.
Considering the higher kidney retention of the c(RGDyK) Arg–Gly–Asp constrained within cyclic pentapeptides. Strong and
dimer in comparison with the monomer, we see advantages selective inhibitors of cell adhesion to vitronectin and laminin
of [99mT]EDDA/HYNIC-RGDyK. fragment P1. FEBS Lett 1991;291:50 – 4.
Summarizing these promising properties, we consider [8] Haubner R, Wester HJ, Reuning U, Senekowitsch-Schmidtke R,
Diefenbach B, Kessler H, et al. Radiolabeled alpha(v)beta3 integrin
[99mT]EDDA/HYNIC-RGD as a good candidate for imag-
antagonists: a new class of tracers for tumour targeting. J Nucl Med
ing avh3 integrin receptor expression in tumors and tumor- 1999;40:1061 – 71.
induced angiogenesis. Recent data using [18F]Galacto-RGD [9] Haubner R, Weber WA, Beer AJ, Vabuliene E, Reim D, Sarbia M,
showed that avh3 integrin receptor expression can be et al. Noninvasive visualization of the activated alphavbeta3 integrin in
visualized using noninvasive nuclear imaging techniques cancer patients by positron emission tomography and [18F]galacto-
RGD. PLoS Med 2005;2:29.
in vivo [24].
[10] Liu S, Edwards DS. 99mTc-labeled small peptides as diagnostic
However, also other clinical applications may be of radiopharmaceuticals. Chem Rev 1999;99:2235 – 68.
interest. Recently, the imaging of integrin avh3 in coronary [11] Decristoforo C, Melendez-Alafort L, Sosabowski JK, Mather SJ.
99m
arterial and peripheral vascular angiogenesis and the Tc-HYNIC-[Tyr3]-octreotide for imaging SST-receptor positive
monitoring of its treatment using radiolabeled RGD peptides tumours, preclinical evaluation and comparison with 111In-octreotide.
J Nucl Med 2000;41:1114 – 9.
have been suggested to have important clinical potential
[12] von Guggenberg E, Behe M, Behr TM, Saurer M, Seppi T,
[25]. A 99mTc-labeled RGD analogue has been used Decristoforo C. 99mTc-labeling and in vitro and in vivo evaluation
successfully for noninvasive serial bhot-spotQ imaging of of HYNIC- and (N alpha-His)acetic acid-modified [d-Glu1]-mini-
angiogenesis induced by ischemia [26]. In addition, other gastrin. Bioconjug Chem 2004;15:864 – 71.
952 C. Decristoforo et al. / Nuclear Medicine and Biology 33 (2006) 945–952

[13] van Hagen PM, Breeman WA, Bernard HF, Schaar M, Mooij CM, [20] Su ZF, Liu G, Gupta S, Zhu Z, Rusckowski M, Hnatowich DJ. In vitro
Srinivasan A, et al. Evaluation of a radiolabelled cyclic DTPA-RGD and in vivo evaluation of a technetium-99m-labeled cyclic RGD
analogue for tumour imaging and radionuclide therapy. Int J Cancer peptide as a specific marker of alpha(V)beta(3) integrin for tumour
2000;90:186 – 98. imaging. Bioconjug Chem 2002;13:561 – 70.
[14] Decristoforo C, Mather SJ. 99m-Technetium labelled peptide–HYNIC [21] Weber W, Haubner R. Comment on bIn vitro and in vivo evaluation
conjugates. The effects of lipophilicity and stability on biodistribution. of a technetium-99m-labeled cyclic RGD peptide as a specific marker
Nucl Med Biol 1999;26:389 – 96. of alphaVbeta3 integrin for tumour imagingQ. Bioconjug Chem 2003;
[15] Von Guggenberg E, Sarg B, Lindner H, Melendez L, Alafort SJ, 14:274.
Mather R, et al. Preparation via coligand exchange and cha- [22] Su ZF, He J, Rusckowski M, Hnatowich DJ. In vitro cell
racterisation of [99mTc-EDDA-HYNIC-d-Phe1, Tyr3]Octreotide studies of technetium-99m labeled RGD-HYNIC peptide, a comparison
(99mTc-EDDA/HYNIC-TOC). J Lab Comp Radiopharm 2003;46: of tricine and EDDA as co-ligands. Nucl Med Biol 2003;30:141 – 9.
307 – 18. [23] Janssen M, Oyen WJ, Massuger LF, Frielink C, Dijkgraaf I, Edwards
[16] Haubner R, Wester HJ, Weber WA, Mang C, Ziegler SI, Goodman SL, DS, et al. Comparison of a monomeric and dimeric radiolabeled RGD-
et al. Noninvasive imaging of alpha(v)beta3 integrin expression using peptide for tumour targeting. Cancer Biother Radiopharm 2002;17:
18
F-labeled RGD-containing glycopeptide and positron emission 641 – 6.
tomography. Cancer Res 2001;61:1781 – 5. [24] Beer AJ, Haubner R, Goebel M, Luderschmidt S, Spilker ME, Wester
[17] Sivolapenko GB, Skarlos D, Pectasides D, Stathopoulou E, Milonakis HJ, et al. Biodistribution and pharmacokinetics of the alphavbeta3-
A, Sirmalis G, et al. Related imaging of metastatic melanoma utilising selective tracer 18F-galacto-RGD in cancer patients. J Nucl Med 2005;
a technetium-99m labelled RGD-containing synthetic peptide. Eur J 46:1333 – 41.
Nucl Med 1998;25:1383 – 9. [25] Villanueva FS. Molecular images of neovascularization: art for art’s
[18] Noiri E, Goligorsky MS, Wang GJ, Wang J, Cabahug CJ, Sharma S, sake or form with a function? Circulation 2005;111:3188 – 91.
et al. Biodistribution and clearance of 99mTc-labeled Arg–Gly–Asp [26] Hua J, Dobrucki LW, Sadeghi MM, Zhang J, Bourke BN, Cavaliere P,
(RGD) peptide in rats with ischemic acute renal failure. J Am Soc et al. Noninvasive imaging of angiogenesis with a 99mTc-labeled
Nephrol 1996;7:2682 – 8. peptide targeted at avh3 integrin after murine hindlimb ischemia.
[19] Haubner R, Bruchertseifer F, Bock M, Kessler H, Schwaiger M, Circulation 2005;111:3255 – 60.
Wester HJ. Synthesis and biological evaluation of a (99m)Tc-labelled [27] Pichler BJ, Kneilling M, Haubner R, Braumuller H, Schwaiger M,
cyclic RGD peptide for imaging the alphavbeta3 expression. Nuklear- Rocken M, et al. Imaging of delayed-type hypersensitivity reaction by
medizin 2004;43:26 – 32. PET and 18F-galacto-RGD. J Nucl Med 2005;46:184 – 9.