Synthesis and evaluation of Re/99mTc(I) complexes bearing a somatostatin receptor-targeting antagonist and labeled via a novel [N,S,O] clickable bifunctional chelating agent

Article abstract of DOI:10.1016/j.bmc.2018.12.028

The somatostatin receptor subtype 2 (SSTR2) is often highly expressed on neuroendocrine tumors (NETs), making it a popular in vivo target for diagnostic and therapeutic approaches aimed toward management of NETs. In this work, an antagonist peptide (sst

Full text of DOI:10.1016/j.bmc.2018.12.028

Accepted Manuscript
9
9m
Synthesis and evaluation of Re/ Tc(I) complexes bearing a somatostatin re-
ceptor-targeting antagonist and labeled via a novel [N,S,O] clickable bifunc-
tional chelating agent
Lauren L. Radford, Dionysia Papagiannopoulou, Fabio Gallazzi, Ashley
Berendzen, Lisa Watkinson, Terry Carmack, Michael R. Lewis, Silvia S.
Jurisson, Heather M. Hennkens
PII:
S0968-0896(18)31627-4
https://doi.org/10.1016/j.bmc.2018.12.028
BMC 14677
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
23 September 2018
11 December 2018
20 December 2018
Please cite this article as: Radford, L.L., Papagiannopoulou, D., Gallazzi, F., Berendzen, A., Watkinson, L.,
9
9m
Carmack, T., Lewis, M.R., Jurisson, S.S., Hennkens, H.M., Synthesis and evaluation of Re/ Tc(I) complexes
bearing a somatostatin receptor-targeting antagonist and labeled via a novel [N,S,O] clickable bifunctional chelating
agent, Bioorganic & Medicinal Chemistry (2018), doi: https://doi.org/10.1016/j.bmc.2018.12.028
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis and evaluation of Re/99mTc(I) complexes bearing a somatostatin receptor-targeting
antagonist and labeled via a novel [N,S,O] clickable bifunctional chelating agent
a
b
a,c
d
Lauren L. Radford , Dionysia Papagiannopoulou , Fabio Gallazzi , Ashley Berendzen , Lisa
d
d
d,e
a
a,f,*
Watkinson , Terry Carmack , Michael R. Lewis , Silvia S. Jurisson , Heather M. Hennkens
aDepartment of Chemistry, University of Missouri, 601 S. College Ave. Columbia, Missouri
6
5211, USA
bSchool of Pharmacy, Aristotle University of Thessaloniki, 54124, Thessaloniki, Greece
cMolecular Interaction Core, University of Missouri, 601 S. College Ave. Columbia, Missouri
6
5211, USA
dResearch Service, Harry S. Truman Memorial Veterans’ Hospital, 800 Hospital Dr. Columbia,
Missouri 65212, USA
eDepartment of Veterinary Medicine and Surgery, University of Missouri, 900 E Campus Dr.
Columbia, Missouri 65211, USA
fUniversity of Missouri Research Reactor Center, 1513 Research Park Dr. Columbia, Missouri
6
*
5211, USA
Corresponding author:
Tel: + 1 573 882 5355
Fax: +1 573 882 6360
Email: HennkensH@missouri.edu
1

List of abbreviations
DIEA N,N-diisopropylethylamine
GA gentisic acid
HRMS high resolution mass spectrometry
ICP-MS inductively coupled plasma mass spectrometry
LC-ESI-MS liquid chromatography electrospray ionization mass spectrometry
NET neuroendocrine tumor
PBS phosphate buffered saline
RCY radiochemical yield
SPECT single-photon emission computed tomography
SPPS solid-phase peptide synthesis
SSTR somatostatin receptor
SSTR2 somatostatin receptor subtype 2
Keywords
rhenium; technetium-99m; tricarbonyl; bifunctional chelator; click chemistry; somatostatin
receptor antagonists; biodistribution; imaging
Abstract
The somatostatin receptor subtype 2 (SSTR2) is often highly expressed on neuroendocrine
tumors (NETs), making it a popular in vivo target for diagnostic and therapeutic approaches
aimed toward management of NETs. In this work, an antagonist peptide (sst -ANT) with high
2
affinity for SSTR2 was modified at the N-terminus with a novel [N,S,O] bifunctional chelator (2)
]+
designed for tridentate chelation of rhenium(I) and technetium(I) tricarbonyl cores, [Re(CO)
3
and [99mTc][Tc(CO) ]+. The chelator-peptide conjugation was performed via a Cu(I)-assisted
3
click reaction of the alkyne-bearing chelator (2) with an azide-functionalized sst
2
-ANT peptide
(
3), to yield NSO-sst -ANT (4). Two synthetic methods were used to prepare Re-4 at the
2
macroscopic scale, which differed based on the relative timing of the click conjugation to the
]+ complexation by 2. The resulting products demonstrated the expected molecular
[
Re(CO)
3
1
25
mass and nanomolar in vitro SSTR2 affinity (IC50 values under 30 nM, AR42J cells, [ I]iodo-
2

1
1
Tyr -somatostatin-14 radioligand standard). However, a difference in their HPLC retention
times suggested a difference in metal coordination modes, which was attributed to a competing
N-triazole donor ligand formed during click conjugation. Surprisingly, the radiotracer scale
reaction of [99mTc][Tc(OH
(CO)
]+ (99mTc; t
= 6 h, 141 keV γ) with 4 formed a third product,
2
)
3
3
½
distinct from the Re analogues, making this one of the unusual cases in which Re and Tc
chemistries are not well matched. Nevertheless, the [99mTc]Tc-4 product demonstrated excellent
in vitro stability to challenges by cysteine and histidine (≥98% intact through 24 h), along with
7
5% stability in mouse serum through 4 h. In vivo biodistribution and microSPECT/CT imaging
studies performed in AR42J tumor-bearing mice revealed improved clearance of this radiotracer
in comparison to a similar [99mTc][Tc(CO)
]-labeled sst -ANT derivative previously studied. Yet
3
2
despite having adequate tumor uptake at 1 h (4.9% ID/g), tumor uptake was not blocked by co-
administration of a receptor-saturating dose of SS-14. Aimed toward realignment of the Re and
Tc product structures, future efforts should include distancing the alkyne group from the
intended donor atoms of the chelator, to reduce the coordination options available to the
[
M(CO)
]+ core (M = Re, 99mTc) by disfavoring involvement of the N-triazole.
3
1
Introduction
Several clinical imaging agents are currently available for detecting neuroendocrine tumors
NETs) by targeting of somatostatin receptors (SSTRs) using radiolabeled octreotide derivatives,
(
6
8
3
TM
111
such as [ Ga]Ga-DOTA-Tyr -octreotate (NETSPOT ) and [ In]In-DTPA-octreotide
TM
(
Octreoscan ). Of the FDA-approved NET radiopharmaceuticals available in the United States,
none use 99mTc (99mTc-Tektrotyd is approved in Europe) and all of them act as SSTR agonists.
There is now a rapidly-growing initiative to investigate the properties of radiolabeled antagonist
3

peptides due to their ability to bind selectively, and with high affinity, to in vivo targets that have
traditionally been targeted using agonist peptides.1–7 In this work, the antagonist peptide 4-NO
2
-
Phe-c(
D
Cys-Tyr-
D
Trp-Lys-Thr-Cys)-
D
Tyr-NH
2
(sst -ANT) was used as the targeting vector due
2
1
to its high affinity for somatostatin receptor subtype 2 (SSTR2), which is the most frequently
overexpressed of the five SSTR subtypes found on NETs.8,9 Additionally, the antagonist
[111
In]In-DOTA-sst -ANT was demonstrated to label a higher number of SSTR2 binding sites
2
1
11
3
1
than a similar agonist peptide, [ In]In-DTPA-Tyr -octreotate. And in a small clinical trial,
[111 -ANT outperformed OctreoscanTM by locating a higher number of NET
In]In-DOTA-sst
2
1
0
metastases, further demonstrating the potential of this SSTR-targeting sequence. Very little
work has been performed using SSTR-targeting antagonist peptides labeled with 99mTc, a widely
available and inexpensive radioisotope that is well established in nuclear medicine clinics
worldwide. The decay properties of 99mTc (t1/2 = 6 h) are greatly advantageous for imaging using
single-photon emission computed tomography (SPECT), particularly its high abundance gamma
emission (141 keV, 89%). Development of a 99mTc-labeled SSTR2-targeting
radiopharmaceutical could provide an additional diagnostic tool for imaging NETs that would be
more accessible to a larger percentage of the patient population and thus help lessen medical
disparities.
We recently reported on the chelation of the tricarbonylrhenium (I) and
[99m
Tc]tricarbonyltechnetium (I) cations via an [N,S,N] bifunctional chelator (Figure 1A) that
was conjugated to sst
2
-ANT.11,12 The resulting [99mTc][Tc(CO)
3
]-labeled, monocationic complex
successfully targeted SSTR2s in a tumor-bearing mouse model, but exhibited slow hepatobiliary
and renal clearance that diminished its potential as an imaging agent. Several factors can affect
1
3
the in vivo behavior of a radiometallated bioconjugate, including the bifunctional chelator, the
4

linker,14,15 the metal, and the overall charge.17–19 We aimed here to improve the
16
pharmacokinetic properties of [99mTc][Tc(CO)
]-labeled sst -ANT by using a new tridentate
3
2
[
N,S,O] bifunctional chelator. An [N,S,O] chelator designed to form neutral metal tricarbonyl
complexes may help reduce the charge impact the metal-chelator complex has on the peptide’s in
vivo behavior.17,20 Additionally, [N,S,O] donor systems are preferred over certain [N,S,N]
systems by the [M(CO)
]+ core (M = Re, 99mTc). For example, Karagiorgou et al. demonstrated
3
that in the presence of four possible donor atoms (N-primary amine, S-thioether, N-pyridine
amine, and O-carboxylate), both 99mTc and Re tricarbonyl cores preferentially bound via the
[
N,S,O] binding mode over the expected [N,S,Npyr] mode, despite the occurrence of a larger 7-
2
1
membered ring in the former. Herein, we report the synthesis of a novel [N,S,O] bifunctional
]+ organometallic core (2, Figure 1B) along with its bioconjugation to
chelator for the [M(CO)
3
the sst
2
-ANT peptide using click chemistry to yield NSO-sst -ANT (4). Additionally, we
2
describe the synthesis of the Re-4 and [99mTc]Tc-4 complexes and their in vitro and in vivo
evaluations as potential SSTR-targeting agents.
A
B
C
1
2
Figure 1. Structures of (A) the previously studied [N,S,N] chelator, (B) the current [N,S,O]
chelator, 2, and (C) the previously studied [N,S,O] chelator.2
5
5

2
2
Methods and Materials
.1 General Methods and Instrumentation
All chemicals were reagent grade from Sigma-Aldrich, unless otherwise specified, and used as
1
25
11
received. [ I]Iodo-Tyr -somatostatin-14 was purchased from Perkin Elmer (Waltham, MA).
All solvents used for high performance liquid chromatography (HPLC) were HPLC-grade and
were passed through 0.22 μm membrane filters prior to use. For HPLC mobile phase and
radiotracer experiments, purified 18 MΩ.cm water (Thermo Barnstead) was used unless
otherwise stated. Solid-phase peptide synthesis (SPPS) was performed on an Apex 396 peptide
synthesizer from AAPPTec (Louisville, KY). All reagents for peptide synthesis were purchased
from AAPPTec, Chem-Impex International, Inc. (Wood Dale, IL), and Oakwood Chemicals
(
West Columbia, SC).
Three HPLC systems were used. System 1 was used for semi-preparative HPLC
separations and liquid chromatography electrospray ionization mass spectrometry (LC-ESI-MS)
analyses. This HPLC system was comprised of a Beckmann Coulter Gold HPLC equipped with a
1
68 diode array detector and 507e auto-injector and was coupled to an ion trap mass analyzer
LCQ FLEET instrument with positive ion ionization; Thermo Fisher Scientific, Waltham, MA).
Semi-preparative purifications were performed on a Waters Nova-Pak C18 preparative column
300 mm x 19 mm x 6 μm, 60 Å) using a linear gradient of 20% to 50% acetonitrile (MeCN)
with 0.1% trifluoroacetic acid (TFA) (solvent A) in H O with 0.1% TFA (solvent B) over 60 min
with a flow rate of up to 10 mL/min (Method 1) and UV detection at 280 nm and either 225 or
35 nm. LC-ESI-MS analyses used a Thermo Scientific BetaBasic C18 analytical column (150
mm x 4.6 mm x 5 μm, 150 Å) and a linear gradient of 10% to 50% solvent A in solvent B over
0 min (Method 2) with a flow rate of 1 mL/min and UV detection at 214 nm, 254 nm and/or
(
(
2
2
3
6

2
80 nm. The HPLC was controlled using the 32 KARAT software package (Beckmann Coulter,
Fullerton, CA) and MS data were analyzed using Thermo Fisher Scientific XCalibur software.
System 2 was an Agilent HP 1100 series pump connected to an HP 1100 multiple
wavelength detector. A semi-preparative Agilent Eclipse XDB-C18 column (250 mm x 9.4 mm
x 5 μm, 80 Å) was used with a linear gradient of 0% to 75% MeOH (solvent C) in solvent B over
1
5 min (Method 3) with a flow rate of 2 mL/min and UV detection at 254 nm.
System 3 was a Shimadzu Prominence HPLC with an SPD-20AV multiple wavelength
detector connected in-line to a NaI(Tl) detector, which also used a Thermo Scientific BetaBasic
reversed-phase C18 column (150 mm x 4.6 mm x 5 μm, 150 Å). The linear gradient was 28% to
3
6% solvent A in solvent B over 20 min (Method 4). The flow rate was 1 mL/min with UV
detection at 254 and 280 nm.
NMR analyses were performed using an Agilent DD2 500 MHz spectrometer (Agilent,
Santa Clara, CA) or a Bruker Avance III 500 MHz spectrometer (Bruker, Billerica, MA).
Inductively coupled plasma mass spectrometry (ICP-MS) analyses were performed on a VG
Axiom SC high-resolution ICP-MS (Thermo Fisher, Waltham, MA). High resolution mass
spectrometry (HRMS) analyses were performed on a Perseptive Biosystems Mariner
atmospheric pressure ionization time-of-flight instrument (API-Tof) operated in positive mode.
The sample was introduced into an electrospray ionization source using a 1:1 mixture of
MeCN/H O. The mass was externally calibrated to CsI adducts.
2
2
.2 Synthesis of but-3-yn-1-yl methanesulfonate (1a)
Into a 100 mL round bottom flask containing 40 mL of anhydrous dichloromethane (DCM), 790
μL of 3-butynol (10 mmol) and 2.8 mL of triethylamine (20 mmol) were added. The reaction
7

mixture was brought to -10 °C using a salty ice bath. A pressure-equalizing addition funnel
containing 850 μL of methane sulfonyl chloride (MsCl, 11 mmol) in 10 mL of anhydrous DCM
was placed on the flask, the entire apparatus was flushed with N
2
gas, and a light N gas flow
2
was maintained for the duration of the reaction. Over a period of 25 min, the MsCl solution was
added to the round bottom flask. The reaction was allowed to continue at -10 °C for 1 h and then
at room temperature for an additional 1.5 h. The reaction mixture was washed three times with
equal parts water, and the organic layer was dried over MgSO and then filtered. Removal of the
4
solvent yielded a light yellow oil (1.37 g) in 90% yield. The product was characterized by NMR.
Observed shifts were similar to those previously reported for the product (prepared via a similar
synthetic route).22 1H NMR (300 MHz, CDCl
) δ: 4.32 (t, 2H), 3.06 (s, 3H), 2.68 (t, 2H), 2.08 (s,
3
1
3
1
H). C NMR (75 MHz, CDCl
3
) δ: 78.59, 70.79, 66.95, 37.67, 19.72.
2
.3 Synthesis of S-(but-3-yn-1-yl) ethanethioate (1b)
A 250 mL, three-neck round bottom flask was fitted with a reflux condenser, a pressure-
equalizing addition funnel, and a glass stopper. The entire apparatus was purged with N
a gentle flow was then maintained for the duration of the reaction. In the flask, 2.0 mg of Cs
6.1 mmol) was suspended in 125 mL of anhydrous MeCN and then, directly to this, 544 μL of
2
gas, and
2
CO
3
(
thioacetic acid (7.3 mmol) was added, which caused the suspension to become slightly yellow.
The reaction was held at 50 °C for 30 min, after which time a solution containing 903 mg of 1a
(
6.1 mmol) in 25 mL anhydrous MeCN was added dropwise via the addition funnel, and the
reaction was continued for 7 h at 55 °C. The reaction mixture was filtered to remove solids and
then concentrated via rotary evaporation to yield a volatile oil (0.7 g) in 82% yield. Observed
NMR shifts were similar to those previously reported for the product (prepared via a similar
8

synthetic route).22 1H NMR (300 MHz, CDCl
) δ: 3.07-2.99 (m, 2H), 2.52-2.42 (m, 2H), 2.34 (s,
, 3H), 2.03-1.98 (m, 1H). C NMR (75 MHz, CDCl ) δ: 194.90 (CO), 81.97, 69.55, 30.43,
8.09 (CH ), 19.44.
3
1
3
CH
3
3
2
3
2
.4 Synthesis of 2-(but-3-yn-1-ylthio)-3-(1H-imidazol-4-yl)propanoic acid (2)
Into a 100 mL round bottom flask, 292 mg of 1b (2.3 mmol) was dissolved in 2.7 mL of 1 M
NaOH (2.7 mmol). The flask was then sealed with a rubber septum and flushed with N gas for
.5 h. A solution of 397 mg of 2-chloro-3-(1H-imidazol-4-yl)propanoic acid (2.3 mmol) in 3 mL
O was added dropwise via a syringe, followed by 2.7 mL of 2 M NaOH for a final
2
1
H
2
concentration of 1 M NaOH (8.1 mmol). The reaction was stirred at room temperature for 7
days. The pH of the solution was adjusted to 7 with 3.8 mL of 1 M HCl, and the solvent was
evaporated to dryness. To the residue, 40 mL of an EtOH:H
brown solids were filtered, and the filtrate was evaporated to dryness. The dry residue (420 mg)
was dissolved in 1 mL of H O:MeOH (95:5) and processed in batches by loading onto a Sep-Pak
Vac C-18 cartridge and elution with ~20 mL of H O:MeOH (95:5). The combined yield was 140
2
O mixture (3:1) was added. The
2
2
mg of purified 2 (colorless oil, 27% yield). NMR assignments (Scheme 1) were made using
1
HSQC. H NMR (500 MHz, D
2
O) δ: 8.42 (s, 1H, a), 7.13 (s, 1H, c), 3.46 (dd, J = 7.8, 7.2 Hz,
1
H, f), 3.05 (dd, J = 15.4, 8.1 Hz, 1H, e), 2.93 (dd, J = 15.3, 7.0 Hz, 1H, e), 2.62 (tt, J =13.3, 6.6
1
3
Hz, 2H, h), 2.36 (dd, J = 8.1, 5.7 Hz, 2H, i), 2.24 (s, 1H, k). C NMR (75 MHz, CD
3
OD) δ:
1
1
4
77.95 (g), 134.04 (a), 133.73 (d), 117.76 (c), 82.30 (j), 69.08 (k), 50.24 (f), 30.28 (e), 29.78 (h),
+
9.21 (i). HRMS (m/z): (C10
H
12
N
2
O S) 225.0692 calc. for [M+H] , 225.0849 found and
2
+
49.1312 calc. for [2M+H] , 449.1440 found.
9

Scheme 1. General reaction scheme illustrating the synthesis of 2 and the subsequent pre- and
post-complexation methods used to synthesize Re-4.
2
.5 Syntheses of N
3
CH
2
CO-4NO
Cys-Tyr-
-ANT, was synthesized using conventional 9-
2
-Phe-c(
D
Cys-Tyr-
D
Trp-Lys-Thr-Cys)-
D
Tyr-NH
2
(N
3
-sst -ANT,
2
3
) and 2-4NO -Phe-c(
2
D
D
Trp-Lys-Thr-Cys)-
D
Tyr-NH (NSO-sst
2
2
-ANT, 4)
The antagonist peptide, sst
2
1
2
fluorenylmethoxycarbonyl (Fmoc) SPPS using Sieber amide resin. The peptide was azido-
derivatized at the N-terminus by adding 2-azidoacetic acid via an amide linkage as the final
elongation step. The azidoacetyl-functionalized peptide (3, in linear form) was then click
conjugated to 2 via a Cu(I)-assisted azide-alkyne cycloaddition reaction to yield 4 (Figure 2B).
Here, the protected and resin-bound peptide was placed in a 1:1 dimethylformamide
(
DMF):tetrahydrofuran (THF) solution containing two equivalents of CuI and 30 equivalents of
N,N-diisopropylethylamine (DIEA). To this solution, two equivalents of 2 in 1:1 DMF:THF
were added. The reaction mixture was allowed to stir overnight at room temperature before being
deprotected and cleaved from the resin using a solution of 87.5% TFA, 2.5% thioanisole, 2.5%
phenol, 2.5% water, 2.5% ethanedithiol and 2.5% triisopropylsilane. LC-ESI-MS analysis of the
crude product demonstrated that disulfide cyclization had also occurred, thus no additional
cyclization step was required. Semi-preparative HPLC purification (System 1, Method 1) was
10

used to collect pure fractions of both 4 and the unconjugated, cyclized 3 that remained (Figure
2
A). The purified fractions were combined, lyophilized and stored at -20 °C. HRMS (m/z): 3
+
(
C
56
H
67
N
15
14
O S
2
) 1238.4506 calc. for [M+H] , 1238.4533 found; 4 (C66
H
79
N
17
O
16
S
3
) 1462.5126
+
2+
calc. for [M+H] , 1462.5363 found and 731.7599 calc. for [M+2H] , 731.7752 found. t
R
: 23.3
min for 3 and 18.0 min for 4 (System 1, Method 2).
A
B
Figure 2. Structures of (A) the azide-functionalized sst
2
-ANT peptide, 3, and (B) the NSO-sst -
2
ANT chelator-peptide bioconjugate, 4.
2
.6 Post-complexation Synthesis of fac-[Re(CO) (4)] (Re-4-post)
3
In the post-complexation method, the chelator was labeled with the tricarbonyl metal core
]+ was
following click conjugation to the peptide (Scheme 1). A solution of [Re(OH
2
)
3
(CO)
3
11

prepared by dissolving (NEt
4
)
2
[Re(CO)
3
Br
3
] (13.8 mg, 0.018 mmol) in 500 μL of H O, adding
2
AgNO (10.1 mg, 0.06 mmol), and stirring at room temperature for 15 min before filtering off
3
2
3
the precipitate. The precipitate was washed with an additional 500 μL of H
2
O to yield a filtrate
with a final volume of 1 mL and 18 mM [Re(OH
2
)
3
(CO)
3
]+. In a 1.5 mL centrifuge tube, 1.9 mg
(
[
0.0013 mmol) of purified 4 was dissolved in 190 μL of 20% MeOH (aq) and 110 μL of the
]+ solution (0.002 mmol). The vial was then heated for 3 h at 55 °C in an
Re(OH
2
)
3
(CO)
3
Eppendorf Thermomixer C (Eppendorf, Hauppauge, NY) fitted with a heating block for 1.5 mL
tubes. The reaction was monitored using HPLC, which showed production of two isomers in a
~1:1 ratio with a yield of 84%. The Re-4-post isomers were purified collectively using analytical
HPLC, and fractions of high purity were combined, lyophilized, and stored at -20 °C. HRMS
19187ReS
) 1732.4453 calc. for [M+H] , 1732.5758 found and 866.7263 calc.
+
(
m/z): (C69
H
78
N
17
O
3
2
+
for [M+2H] , 866.7838 found. t : 9.8 and 10.2 min (System 3, Method 4).
R
2
.7 Pre-complexation Synthesis of fac-[Re(CO) (4)] (Re-4-pre)
3
In the pre-complexation labeling method, the chelator was labeled with the tricarbonyl metal
]+ was
core prior to click conjugation to the peptide (Scheme 1). A solution of [Re(OH
2
)
3
(CO)
3
prepared as described above using 38 mg of (NEt
and 25 mg of AgNO (0.15 mmol). To this solution, 11 mg of 2 (0.05 mmol) and 5 mg of
NaHCO (0.05 mmol) were added. After heating the reaction at 65 °C for 2 h, the solvent was
4
)
2
[Re(CO)
3
Br
3
] (0.05 mmol) in 2 mL of H O
2
3
3
evaporated to dryness and THF was added to the residue. The solids formed were filtered off,
and the filtrate was concentrated to dryness. The crude product (25 mg) was purified by semi-
preparative HPLC (System 2, Method 3), yielding 18 mg (after drying) of fac-[Re(CO)
3
(2)] (Re-
-
1
1
2
) as an off-white solid (yield: 73%). IR (KBr, cm ): 2031, 1910, 1636. H NMR (500 MHz,
12

CD
3
OD) δ: 8.12 (s, 1H, a), 7.08 (s, 1H, c), 4.39 (dd, J = 4.1 Hz, 1H, f), 3.64 (dd, J = 17.5, 4.2 Hz,
1
H, e), 3.39-3.32 (m, 2H, e, h), 3.24-3.14 (m, 1H, h), 2.76 (t, J = 6.2 Hz, 2H, i), 2.52 (t, J = 2.5
13
Hz, 1H, k). C NMR (125 MHz, CD
3
OD) δ: 196.37, 195.23, 194.79 (CO), 181.08 (g), 141.00
(
(
[
a), 116.03 (c), 133.96 (d), 79.71 (j), 71.26 (k), 46.57 (f), 39.55 (h), 27.32 (e), 17.59 (i). HRMS
1
87
+
m/z): (C13
H
11
N
2
O
5
ReS) 495.0018 calc. for [M+H] , 494.9956 found and 986.9936 calc. for
+
2M+H] , 986.9713 found.
Purified 3 (1 mg, 0.8 μmol) was weighed into a 5 mL pear-shaped flask containing a
Teflon stir bar. The flask was sealed with a rubber septum and purged with Ar gas for 15 min. In
a separate 5 mL crimp vial, a 2 mg/mL solution of Re-2 dissolved in 75% t-BuOH in water was
prepared and then purged with Ar gas for 15 min. Using a syringe, 395 μL of the Re-2 solution
(
0.79 mg, 0.8 μmol) was added to the flask containing the peptide and was then stirred under a
gentle stream of Ar gas. In a separate tube, an aqueous solution of NaHCO (1.6 μmol in 22.5 μL
O), ascorbic acid (1.6 μmol in 47 μL H O), and CuSO ∙5H O (0.8 μmol in 22.5 μL H O) was
prepared; CuSO was the final reagent added and caused the mixture to quickly become yellow
and cloudy. This entire mixture (92 μL) was immediately transferred via syringe into the reaction
flask along with an additional 100 μL of H O rinsate, and the reaction was heated at 50 °C for
.5 h. Two Re-4-pre isomers formed and were purified collectively using analytical HPLC with
3
H
2
2
4
2
2
4
2
3
an isolated yield of 13% (determined by ICP-MS) and were lyophilized and stored at -20 °C. LC-
19187ReS
) 1732.4 calc. for [M+H] , 1732.4 found and 866.7 calc. for
+
ESI-MS (m/z): (C69
H
78
N
17
O
3
2
+
[
M+2H] , 867.0 found. t : 11.5 and 11.8 min (System 3, Method 4).
R
2
.8 Synthesis of fac-[Re(CO) (2-1-acetic acid-1,2,3-triazole)] (Re-2-model)
3
In a 10 mL round bottom flask, Re-2 (17 mg, 0.034 mmol) was dissolved in 75% t-BuOH in
water (2.3 mL). To this, a solution of 2-azidoacetic acid (3 μL, 0.04 mmol) in 0.5 mL of water
13

was added. The flask was sealed with a rubber septum and flushed with a gentle N
2
gas flow for
1
5 min. In a separate 10 mL vial, ascorbic acid (2.5 mg, 0.014 mmol) was dissolved in 0.5 mL of
water, to which was added an aqueous 1 M NaHCO solution (14 μL, 0.014 mmol) and then
3
CuSO
4
∙5H O (1.9 mg, 0.007 mmol). After addition of the copper reagent, the solution turned
2
yellow and turbid and was transferred immediately via syringe to the round bottom flask. The
reaction was heated at 50 °C for 5 h, at which time the vessel was cooled to room temperature
and the t-BuOH was removed in vacuo to leave behind a white, cloudy precipitate as an aqueous
suspension. The suspension was centrifuged, the liquid decanted off, and the precipitate was
washed once with water before being fully dried under vacuum to afford 13 mg of Re-2-model
-1
1
(
64% yield, Scheme 2). IR (KBr, cm ): 2030, 1918, 1903, 1654, 1618. H NMR (500 MHz,
CD OD) δ: 8.11 (s, 1H, a), 7.94 (s, 1H, k), 7.06 (s, 1H, c), 5.18 (t, 1H, l), 4.22 (t, J = 4.2 Hz, 1H,
f), 3.62 (dd, J = 3.4, 17.5 Hz, 1H, e), 3.48 (dd, J = 7.0, 14.1 Hz, 1H, h), 3.39 (m, 1H, h), 3.32 (m,
3
1
3
1
1
H, e), 3.25 (m, 2H, i). C NMR (125 MHz, CD
3
OD) δ: 162.58 (m), 140.95 (a), 135.15 (j),
34.04 (d), 132.56 (k), 115.98 (c), 54.25 (l), 46.42 (f), 40.27 (h), 27.32 (e), 23.69 (i). HRMS
1
87
+
(
m/z): (C15
H
14
N
5
O
7
ReS) 618.0062 calc. for [M+Na] , 618.0109 found.
Scheme 2. Synthesis of the model complex, Re-2-model.
14

2
.9 Concentration Determinations for Solutions of 4 and Re-4
1
1
ICP-MS analyses were performed for Re-4-pre, Re-4-post, and 4, as previously described. The
purified and lyophilized Re products were dissolved into a 10% methanol in water solution,
while purified and lyophilized 4 was dissolved into a 20% methanol in water solution. All
solutions were prepared to a targeted concentration of approximately 1 mg/mL. Briefly, each
solution was aliquoted (13-15 μL) in duplicate into acid-washed, pre-weighed tubes. The tubes
were reweighed and the solutions were digested with 150 μL of concentrated HNO and then
3
heated for 1 h at 100 °C. The digestate volume was brought to approximately 5 mL with
1
ultrapure water and then weighed a final time. Afterwards, aliquots were removed for further
dilution and fortification with Be/Sc/Y/Tl internal standards of known concentration (undiluted
digestate was also analyzed). Internal standards were likewise added to a series of calibration
standards prepared from MS-D Multielement Standard stock solutions (High Purity Standards,
Charleston, SC), which contained Re and S at ppb levels. Mean results were used, as measured
by signals from S for S and both 185Re and 187Re for Re. Solvent and digestion blanks were
32
analyzed in similar fashion.
2
.10 In Vitro Receptor Binding Affinity
The binding affinities of Re-4-pre and Re-4-post for SSTRs in rat pancreatic cancer (AR42J)
cells were each determined by competition for binding sites with the radioligand standard
[125
I]iodo-Tyr -somatostatin-14. The AR42J cells were cultured by the Cell and Immunobiology
11
Core at the University of Missouri (Columbia, MO) in RPMI 1640 medium with 4.5 g/L glucose
1
Water was passed through a PURELABR Pulse 1 (ELGA LabWater), producing 15.0 MΩ.cm
water. This water was then passed through a PURELABR Flex 2 UV/TOC (ELGA LabWater),
which produced Type I, 18.2 MΩ.cm water via UV oxidation of organics and ion exchange. The
total oxidizable carbon specification (TOC) was <5 ppb.
15

(
GIBCO-Invitrogen, Carlsbad, CA) that was supplemented with 10% fetal bovine serum, 2 mM
L-glutamine, and 50 μg/mL gentamycin. Cells were incubated at 37 °C in a 5% CO atmosphere
2
and harvested using a soybean trypsin inhibitor.
The competition assay was performed similarly to a previously published method.11
Serial dilutions in binding buffer (0.25% bovine serum albumin in GIBCO-Invitrogen RPMI
-5
-12
1
640 media) were performed to make solutions of Re-4 ranging from 3 × 10 M to 3 × 10
M
(
by 10-fold dilution factors). Into 1.5 mL centrifuge tubes were added: 100 μL of Re-4, 100 μL
1
25
11
of [ I]iodo-Tyr -somatostatin-14 in binding buffer (100,000 cpm), and 100 μL of AR42J cells
6
(
2 × 10 cells, pre-washed twice with binding buffer). Each Re-4 concentration was tested in
-5
-12
triplicate, and final concentrations were 1 × 10 M to 1 × 10 M. Additionally, three replicates
contained 100 μL of binding buffer in place of the Re-4 solution, as zero concentration standards.
The tubes were incubated for 1 h in a laminar flow hood at room temperature and were manually
shaken once every 15 min. After incubation was complete, the cells were handled using cold
trays. The cells were centrifuged at 6000 rpm for 30 seconds, and the pellets were washed with
1
25
11
refrigerated binding buffer. This was repeated twice more to remove unbound [ I]iodo-Tyr -
somatostatin-14. Finally, the supernatant was removed and the pellets were counted on a Multi-
Wiper gamma counter (Laboratory Technologies, Inc., Elburn, IL). Data analysis was performed
using GraphPad Prism 6 software (La Jolla, CA).
2
.11 Synthesis of fac-[99mTc][Tc(CO)
The [99mTc]triaquatricarbonyltechnetium (I) precursor, [99mTc][Tc(OH
synthesized by adding 0.8-1 mL of [99mTc][TcO ]– (10-20 mCi, 370-740 MBq) in saline from a
99Mo/99m
3
(4)] ([99mTc]Tc-4)
2
)
3
(CO)
]+, was
3
4
Tc generator (Mallinckrodt Pharmaceuticals, St. Louis, MO) to a kit prepared in-house
16

24
according to a literature procedure. Briefly, a 5 mL crimp vial containing 5.5 mg of NaBH , 4
4
mg of Na
immediately prior to addition of [99mTc][TcO
yielded a solution of [99mTc][Tc(OH ]+, which was then pH adjusted to 6 using 1 M HCl.
2
CO
3
, and 10 mg of Na/K tartrate (the kit) was flushed with CO(g) for 15 min
]–. Subsequent heating at 100 °C for 20 min
4
2
)
3
(CO)
3
The precursor formation was verified by HPLC (System 3, Method 4, t = 5 min) with a
R
radiochemical yield (RCY) of ≥ 98%.
To 18 μL of 4 (1.13 mM in an aqueous solution of 20% MeOH), 186 μL of
]+ (2-5 mCi, 74-185 MBq) was added to give a final ligand concentration
[99m
Tc][Tc(OH
)
3
(CO)
2
3
-4
of 1 × 10 M. The solution was heated at 75 °C for 30 min to give a RCY of 65% ± 5% (n=5) for
the major [99mTc]Tc-4 product, which was purified via analytical HPLC (System 3, Method 4, t
R
=
16.1 min).
To concentrate [99mTc]Tc-4 and remove TFA, the HPLC-collected product fraction was
processed through a Waters Sep-Pak C18 Light cartridge that had been preconditioned with 10
mL of absolute ethanol (EtOH) followed by 10 mL of H O. The product fraction was first diluted
2
to 20 mL with high purity 18 MΩ.cm H
2
O and then loaded onto the pre-conditioned Sep-Pak.
O, and the [99mTc]Tc-4 was eluted with 400 μL of
The Sep-Pak was washed with 10 mL of H
2
absolute EtOH into a 1.5 mL centrifuge tube containing approximately 1 mg of gentisic acid
(
GA) as a radioprotectant. The EtOH was evaporated using gentle heating (≈ 55 °C water bath)
gas, and the 99mTc-labeled complex was reconstituted in a solution of 1%
and a light flow of N
2
Tween 80 in 10 mM phosphate-buffered saline (PBS).
2
.12 In Vitro [99mTc]Tc-4 Stability Studies
Purified and concentrated product was aliquoted (30 μL) into three 1.5 mL centrifuge tubes along
with 30 μL of either 10 mM cysteine or 10 mM histidine in PBS and subsequently diluted with
17

2
40 μL of 10 mM PBS. The final solution pH was 7.4, the final amino acid concentration was 1
mM, and the final Tween 80 concentration was 0.1%. In the case of mouse sera studies, 30 μL
aliquots of [99mTc]Tc-4 were diluted 10-fold using whole mouse sera (Sigma Life Science, St.
Louis, MO).
The solutions were incubated at 37 °C and the percentage of intact complex was
determined at various time points via HPLC (System 3, Method 4). Small aliquots from the
histidine and cysteine solutions were directly injected into the HPLC for analysis at 1 h, 4 h, and
2
4 h. Larger mouse sera incubation aliquots were taken, diluted with three volumes of MeCN,
vortexed and then centrifuged at 2500 rpm for 5 min. The MeCN layer was removed and the
protein pellet was washed once with three volumes of MeCN, vortexed, and centrifuged. After
separation, the combined MeCN layers and the pellet were each counted on a dose calibrator
(
Capintec, Ramsey, NJ) to determine the percentage of protein binding. The MeCN supernatant
was then concentrated to dryness using a stream of N gas and gentle heating (≈ 55 °C water
bath), reconstituted in 1% Tween 80 in 10 mM PBS, and injected for HPLC analysis.
2
2
.13 Distribution Coefficient Determination
A glass test tube was filled with 2 mL of water and 2 mL of n-octanol. While vortexing, 50 μL of
purified and concentrated [99mTc]Tc-4 (approximately 7.4 μCi, 0.27 MBq) was added to the test
tube. Vortexing was continued for 30 additional seconds, and the tube was then centrifuged for
1
0 min. A 50 μL aliquot was removed from the aqueous (pH 7) and organic layers and counted
on a NaI(Tl) well detector. This process was repeated for a total of three replicates. The log D
was calculated using the following equation: log D
7
7
= log (cpmoct/cpmwat).
18

2
.14 In Vivo Biodistribution and Imaging Studies
All animal studies were conducted in compliance with a protocol approved by the Subcommittee
for Animal Studies of the Harry S. Truman Memorial Veterans’ Hospital and the Animal Care
and Use Committee of the University of Missouri-Columbia Animal Care Quality Assurance
Office. Female ICR SCID mice (Taconic, Hudson, NY), age 6 weeks, were inoculated with 5 ×
1
06 AR42J cells via subcutaneous injection in the right hind flank. Imaging studies were
conducted 2 weeks after inoculation, and biodistribution studies were conducted 2.5 weeks after
inoculation.
The mouse dosing solutions were prepared by HPLC purification and Sep-Pak
concentration of the [99mTc]Tc-4 product, as described above, except with reconstitution of the
99mTc-labeled product in 0.1% Tween 80 in sterile saline (vs. 1% Tween 80 in PBS). The
solution pH was adjusted to neutral with 0.1 M NaOH, and the final volume was adjusted with
0
.1% Tween 80 in sterile saline to yield the desired concentration. Finally, the solution was
passed through a 0.22 μm sterile nylon filter.
The final activity concentrations for the biodistribution doses were: (1) regular dose, 4.5
μCi (0.17 MBq) in 100 μL and (2) blocking dose, 4.5 μCi (0.17 MBq) in 150 μL containing 150
μg of SS-14. Mice (n = 4 for each group) were humanely sacrificed by cervical dislocation at 1 h
and 4 h post-injection, and their organs and tissues were collected and then counted on a Wallac
1
480 Wizard 3” gamma counter from Perkin Elmer. Mice that received a blocking dose were
only evaluated at the 4 h time point.
The final activity concentrations for the imaging doses were: (1) regular dose, 70 μCi (2.6
MBq) in 100 μL and (2) blocking dose, 70 μCi (2.6 MBq) in 150 μL containing 150 μg of SS-14.
Studies were performed on live mice and microSPECT/CT images were taken at 1 and 4 h post-
19

injection. Mice were anesthetized using 3% isoflurane at induction and 2.5% isoflurane during
imaging. Mice were kept warm during imaging with a heating coil located on the imaging
platform. Imaging data were collected using a SIEMENS Inveon MicroSPECT/CT scanner
(
Siemens Medical Solutions, Malvern, PA). Spiral SPECT imaging was performed using MWB
.0 mm collimators with a 35 mm radius of rotation and 120 projections. The images were
reconstructed using OSEM3D. The CT images were captured using an 80 kVp x-ray source and
60° data acquisition for 3D volume-rendered data. The images were reconstructed in real-time
1
3
using Fan beam (Feldkamp) filtered back projection algorithms and were processed using Inveon
Research Workplace processing software.
3
3
Results and Discussion
.1 Synthesis of 2 and 4
Previously, highly stable Re- and 99mTc-labeled tricarbonyl complexes were reported using a
tripodal [N,S,O] chelation design, where N-imidazole, S-thioether, and O-carboxylate donor
atoms participated in the coordination sphere.25–29 A chelator employing the same chelation
strategy (Figure 1C) was thus adopted for 2. However, in contrast to previous [N,S,O] chelators,
2
was functionalized with an alkyne group (Figure 1B) rather than a carboxylate group, to allow
for conjugation to the sst -ANT peptide via an azide-alkyne click reaction.
2
The synthesis of 2 is depicted in Scheme 1. The reaction of 3-butynol with methane
sulfonyl chloride in anhydrous DCM led to the mesylated derivative (1a), which was isolated in
9
0% yield. Reaction of 1a with cesium thioacetate (generated in situ) gave the thiolate derivative
1b), which was isolated in 82% yield. As a final step, 1b and 2-chloro-3-(1H-imidazol-4-
yl)propanoic acid were reacted in an aqueous 1 M NaOH solution for 7 days to yield L. Due to
(
20

the selective nature of Cu(I)-assisted azide-alkyne cycloaddition reactions, the addition of
protecting groups to 2 prior to peptide conjugation was not required.
Compound 4 was synthesized using Fmoc SPPS. The sst -ANT peptide was
2
functionalized with an azidoacetyl group at the N-terminus to yield linear 3, to which 2 was
conjugated via a Cu(I)-assisted click reaction. After deprotection and cleavage from the resin,
LC-ESI-MS analysis of the crude reaction mixture confirmed the formation of the desired
3
0
disulfide-cyclized L-sst
2
-ANT product, along with cyclized unconjugated 3. Both the 3 and 4
were successfully purified using semi-preparative HPLC (System 1, Method 1).
3
.2 Synthesis of Rhenium Complexes
Using a post-complexation synthetic method, the [Re(CO)
conjugation between 2 and sst -ANT. This mimicked the preferred approach for radiotracer
]+ complex with 2 was formed after
3
2
syntheses, in which the radioactive component is incorporated in the final synthetic step in order
to minimize loss of radioactivity due to decay, to reduce personnel exposure, etc. Here, the
reaction of [Re(OH
2
)
3
(CO)
]+ with 4 yielded two Re-4-post isomeric products in a ~1:1 ratio
3
with an 84% reaction yield (determined by HPLC), which were collectively purified by
analytical HPLC. The observation of isomers was not surprising, as isomer formation upon
coordination of the [Re(CO)
]+ core with the [N,S,O] chelator can be caused by the combination
3
of two factors: the chiral C center of the chelator, yielding enantiomers, and the two coordination
orientations that are possible with the prochiral S-donor of the chelator, leading to two
diastereromers (each being a pair of enantiomers). The chirality of the sst -ANT peptide further
2
adds to the isomeric products possible. Similar observations were made in our previous work, as
well as with other chelators containing chiral C and prochiral S atom donors.12,20 LC-ESI-MS
21

and HRMS analyses of Re-4-post resulted in an excellent match between the calculated and
observed mass-to-charge ratios (m/z) and demonstrated the expected isotopic pattern for both the
singly- and doubly-charged species.
The click conjugation of the chelator 2 to the peptide sst -ANT results in the formation of
2
a triazole ring, and with it, a possible competing N-triazole (Ntr) donor atom. The use of triazole
]+ (M = Re, 99mTc) has been described
rings to provide a nitrogen donor for chelation of [M(CO)
3
3
1
previously. And similar to the N-imidazole (Nim) donor with its three-carbon distance to the
adjacent S-thioether donor, Ntr coordination would also yield a six-membered metal ring to form
a neutral [N,S,O] complex. Thus, while the findings from the post-complexation method were
consistent with [N,S,O] coordination, distinction could not be made between tripodal [Nim,S,O]
and linear [Ntr,S,O] modes (Figure 3A and 3B, respectively). Further possible was the linear
[
N
tr,S,Nim] coordination to yield a monocationic Re-4 complex (Figure 3C). A pre-complexation
synthetic route was therefore also explored, wherein the Re-2 complex with the intended tripodal
im,S,O] coordination was formed prior to click conjugation, thereby effectively eliminating
competition between Nim and Ntr donor atoms for metal coordination.
[
N
22

A
B
C
Figure 3. (A) The intended tripodal [Nim,S,O] coordination mode of fac-[M(CO) (4)] vs. the
3
alternative (B) linear [Ntr,S,O] and (C) linear [Ntr,S,Nim] coordination modes; M = Re, 9 Tc.
9m
The pre-complexation method used a Cu(I)-assisted click reaction of pre-complexed Re-2
with 3 for synthesizing Re-4-pre. Prior to this reaction, the Re-2 was synthesized, yielding one
major product that was purified by semi-preparative HPLC (System 2, Method 3) and
subsequently confirmed by NMR to correspond to a single diastereomer. The [Nim,S,O]
1
coordination mode of Re-2 was verified by H-NMR and COSY, as well as by comparison
against previous complexes bearing the same chelating unit.25,28,29 The Re-4-pre product that
was obtained by click conjugation of Re-2 with the chiral chelator-peptide 3 was also comprised
of two isomers, similar to Re-4-post. LC-ESI-MS analysis of the Re-4-pre products
demonstrated the expected isotopic pattern with m/z that matched the calculated value. However,
upon HPLC co-injection of the Re-4-pre and Re-4-post products, differences in retention times
were observed, with the Re-4-pre isomers eluting approximately 1.7 min later than the Re-4-
23

post isomers (t
, Method 4).
To further verify the [Nim,S,O] coordination mode of Re-4-pre, a model clicked complex
R
= 9.8 and 10.2 min for Re-4-post and 11.5 and 11.8 min for Re-4-pre; System
3
was prepared through the reaction of Re-2 with azidoacetic acid (Scheme 2). The single
diastereomer product, Re-2-model, was purified by semi-preparative HPLC (System 2, Method
3
) and was characterized by IR, NMR and HRMS. The spectroscopic characterization indicated
that Re-2-model had indeed maintained the original [Nim,S,O] coordination through the click
conjugation. The NMR pattern of the protons that participate in the chelating unit matched those
of Re-2. In particular, the imidazole shifts were very similar, indicating that the imidazole was
coordinated with the metal in both complexes. Furthermore, the shifts of the H-e protons were in
the same range [3.64 (dd, J = 17.5, 4.2 Hz), 3.39-3.32 (m) for Re-2 and 3.62 (dd, J = 17.5, 3.4
Hz), 3.32 (m) for Re-2-model], and the H-i protons appeared as a multiplet, consistent with a
pendant moiety. Conversion to the linear coordination mode [Ntr,S,Nim] would result in different
2
5
H-e proton shifts. It would also result in the splitting of the two H-i protons into two sets of
peaks, as is common for the methylene protons participating in a chelator’s backbone, but not
observed here. Therefore, the Re-4-pre product is well-supported to involve the tripodal
[
Nim,S,O] coordination of the Re-tricarbonyl core by 4 (Figure 3A).
The exact nature of Re-4-post remains unclear. While the observed difference in HPLC
retention times might be explained by [Nim,S,O] coordination with slightly different
conformational preferences due to the additional steric bulk of the peptide-conjugated ligand,
perhaps more likely is that Re-4-post adopted an alternate binding mode, such as with neutral
linear [Ntr,S,O] coordination (Figure 3B) or with monocationic linear [Ntr,S,Nim] coordination
(
Figure 3C). However, spectroscopic data to support which binding mode is present in Re-4-post
24

were not obtained due to the fact that our efforts to synthesize the model clicked chelator (via
conjugation of 2 with azidoacetic acid) were unsuccessful.
3
.3 In Vitro Receptor-Binding Affinity
The SSTR binding affinity of Re-4 was determined for both the Re-4-pre and Re-4-post
products by a competitive binding assay in SSTR2-expressing rat pancreatic cancer (AR42J)
cells. Low nanomolar affinity was retained by sst
2
-ANT after its N-terminal modification,
regardless of the reaction pathway used to synthesize Re-4. The IC50 values of 27 ± 13 nM (Re-
4
-post) and 19 ± 14 nM (Re-4-pre) shown in Figure 4 are comparable to that of our previously
+
11
reported bioconjugate complex [Re(CO)
3
(NSN-sst -ANT)] (15 ± 4 nM). These promising
2
results warranted our further investigation into the preparation of the 99mTc analogue, along with
its stability and biological evaluations.
1
0 0
5
0
0
-
1 2
-1 1
-1 0
-9
-8
-7
-6
-5
lo g [R e(C O ) (L -sst -A N T )]
3
2
log[Re-4] (M)
Figure 4. Binding affinities for Re-4-pre (grey) and Re-4-post (black) in SSTR2-expressing
AR42J cells.
25

3
.4 Synthesis of [99mTc]Tc-4
The post-conjugation labeling approach (i.e., incorporation of the radiotracer in the final step)
was selected for practical reasons, as it is the more advantageous route from the radiosynthetic
standpoint, especially for short-lived radioisotopes. The [99mTc][Tc(OH
)
(CO)
]+ precursor was
2
3
3
2
4
synthesized using a kit prepared in-house according to a literature procedure. The optimized
conditions for synthesizing [99mTc]Tc-4 were found to be adjustment of the precursor solution
pH to 6 prior to heating with a solution of 4 at 75 °C for 30 min, with product isolation by
HPLC.
Initial 99mTc labeling attempts were performed similarly to our previous NSN-sst
-ANT
2
1
1
labeling (75 °C, 30 min). Heating in the pH range of 5-8 led to four radiolabeled products, all
of which exhibited very different HPLC retention times from the Re-4-pre/post standards; one
early-eluting (t
R
= 9.3 min) and three late-eluting (t = 16.1 min, 17.6 min, 18.2 min) products
R
were observed with System 3, Method 4. An initial stability measurement of the HPLC-isolated
early-eluting product and late-eluting products in 10 mM PBS for 4 h at 37 °C eliminated the
early-eluting product from further consideration, as 90% had either oxidized to [99mTc][TcO ]– or
4
had converted into the late-eluting products. This indicated that the early-eluting isomer was not
thermodynamically favored, suggestive of incomplete bidentate coordination.
Attempts to synthetically favor the dominant late-eluting product (t = 16.1 min) over the
R
others revealed that the relative product RCYs depended greatly on both the reaction pH and
temperature. For example, reactions performed at pH 2 favored the early-eluting product,
regardless of the reaction temperature. Only the early-eluting product was formed when the
reaction was performed at pH 5 and room temperature (50% RCY, with the remaining activity as
the unreacted [99mTc][Tc(OH
(CO)
]+ precursor). However, conversion to the late-eluting
2
)
3
3
26

products occurred upon heating, with only 9% of the original early-eluting product remaining.
Reaction conditions of pH 6-7 and heating for 30 min at 75 °C gave the highest observed RCY
(
average 65%) for the dominant late-eluting product, hereafter referred to simply as [99mTc]Tc-4.
All subsequent labeling reactions were therefore performed at a pH of 6, in order to favor
formation of this product, which was purified for use in the remaining in vitro and in vivo
studies.
The observation of distinctly different products from the post-conjugation labeling
approach with Re and 99mTc was initially unexpected; Re and Tc are well known to typically
demonstrate similar chemical behaviors, and we had not observed such differences in our earlier
work. However, it appears that the [99mTc]Tc-4 product does not involve the intended [Nim,S,O]
coordination mode (Figure 3A). The above findings are consistent with, though not definitive
for, a [99mTc]Tc-4 product with a [Ntr,S,O] or [Ntr,S,Nim] linear coordination mode (Figures 3B or
3
C, respectively). The relative hardness of Tc compared to the softer Re may have resulted in a
preference for [Ntr,S,O] coordination.
3
.5 In Vitro Stability Studies
After HPLC purification, the [99mTc]Tc-4 product was concentrated using a Sep-Pak C-18 Light
cartridge, from which it was eluted in absolute EtOH into a vial containing gentisic acid as a
radioprotectant. The EtOH was removed and the product was reconstituted in 1% Tween 80 in
1
0 mM PBS for use in stability studies. This [99mTc]Tc-4 solution was aliquoted to new tubes
and diluted 10-fold with either cysteine or histidine (1 mM final concentrations), or with whole
mouse sera. The percentage of intact complex was monitored by HPLC at 1, 4, and 24 h.
27

The [99mTc]Tc-4 product demonstrated excellent stability in both the cysteine and
histidine challenge studies, with at least 98% of the complex intact through 24 h (Table 1). In the
mouse sera studies, carried out through 4 h, the majority of the radioactivity remained in the
supernatant (73% at 4 h), demonstrating moderately low protein binding. HPLC analysis of the
supernatant radioactivity showed that the large majority of the complex was intact (75% intact at
4
h). The remainder of the radioactivity existed as an unknown degradation product (System 3,
Method 4, t
2.7 min).
R
= 13.1 min), possibly a small protein adduct, but clearly not [99mTc][TcO
4
]– (t
=
R
Table 1. In vitro stability of [99mTc]Tc-4
a
1
h
4 h
24 h
100 ± 0
99.7 ± 0.5
NP
Cysteine
Histidine
100 ± 0
98 ± 2
66 ± 1
100 ± 0
99.7 ± 0.5
75 ± 1
Mouse Serum
a
Average percentage ± SD of intact complex relative to
percent intact complex at 0 h, n = 3; NP = Not performed
3
.6 Distribution Coefficient Measurement
The distribution coefficient for [99mTc]Tc-4 was found by measuring its distribution between n-
octanol and water. The log D
7
value of -0.44 indicates a fairly even distribution between the
water and octanol layers, with the complex being only slightly hydrophilic.
3
.7 In Vivo Biodistribution and Imaging Studies
6
Female ICR SCID mice were inoculated with 5 × 10 AR42J cells in the right hind flank, and
tumors were allowed to grow for 2-2.5 weeks. Purified and concentrated [99mTc]Tc-4 was
reconstituted with 0.1% Tween 80 in sterile saline to the desired dosing concentration, and the
28

pH was adjusted to neutral with 0.1 M NaOH. Blocking doses contained SS-14, the native SSTR
ligand, to saturate tumor receptors. MicroSPECT/CT images were captured at 1 h, 1 h block, 4 h,
and 4 h block. Biodistribution data were collected at 1 h, 4 h, and 4 h block.
The biodistribution data are presented in Table 2. Tumor uptake was high at 1 h with 4.9
%
ID/g, decreasing to 1.6 %ID/g by 4 h. No statistical difference was observed for tumor uptake
between the 4 h and 4 h blocking (1.4 %ID/g) time points. Further, the amount of radioactivity
measured in the tumor at 4 h was roughly equal to that of the blood pool. These findings indicate
that tumor uptake occurred via passive diffusion and non-specific binding rather than by receptor
mediation, presumably due to a loss in SSTR affinity associated with the different coordination
mode in [99mTc]Tc-4 as compared with the SSTR-targeting Re products. Hepatobiliary clearance
was the primary mode of excretion, as seen by high uptake in the small intestine uptake at 1 h
(
31 %ID/g) that shifted to the large intestine by 4 h (67 %ID/g). Secondary to the intestinal
activity, the bulk of the remaining activity at both time points was distributed between the kidney
(
some renal clearance) and liver. For both kidney and liver, the uptake at 4 h was roughly half
that observed at 1 h. Overall, [99mTc]Tc-4 demonstrated significantly faster and more complete
clearance compared to the previously studied [99mTc][Tc(CO) -ANT)]+.11
3
(NSN-sst
2
29