Preliminary evaluation of 18F-labeled LLP2A-trifluoroborate conjugates as VLA-4 (α4β1 integrin) specific radiotracers for PET imaging of melanoma

Article abstract of DOI:10.1016/j.nucmedbio.2018.02.005

Introduction: The transmembrane α₄β₁ integrin receptor, or very-late antigen 4 (VLA-4), is associated with tumor metastasis and angiogenesis, the development of chemotherapeutic drug resistance, and is overexpressed in multiple myelomas, osteosarcomas, lymphomas, leukemias, and melanomas. The peptidomimetic, LLP2A, is a high-affinity ligand with specificity for the extracellular portion of VLA-4 and several conjugates have been evaluated in vivo by NIR-fluorescence, ¹¹¹In-SPECT and ⁶⁸Ga- and ⁶⁴Cu-PET imaging, but to date, not with ¹⁸F-PET. Methods: Using two highly stable organotrifluoroborate prosthetic groups: ammoniumdimethyl-trifluoroborate (AMBF₃) and a new N-pyridinyl-para-trifluoroborate (N-Pyr-p-BF₃), both capable of facile aqueous ¹⁸F-labeling by isotope exchange (IEX), we present the first PET imaging evaluations of two [¹⁸F]R-BF₃ ^?–PEG₂-LLP2A tracers using VLA-4 overexpressing B16-F10 murine melanoma tumor mouse models. Results: Here, we demonstrate successful one-step ¹⁸F-labeling of both conjugates with wet NCA [¹⁸F]F^? in radiochemical yields of up to 11.6% within 75 min at molar activities of 40–100 GBq/μmol. Average tumor uptake values based on ex vivo biodistribution values were 4.4%ID/g (11) and 2.8%ID/g (12) using ¹⁸F-labeled LLP2A-conjugates with the two prosthetic groups: N-Pyr-p-BF₃ (5) and alkyl-N,N-dimethylammonio-BF₃ (AMBF₃) (7), respectively, and was found to be target-specific as evidenced by in vivo blocking controls. Dynamic PET scanning and biodistribution studies revealed slow clearance of the [¹⁸F]R-BF₃ ^?–PEG₂-LLP2A tracers from the tumors, and also substantial uptake in the intestines, gall bladder, liver and bladder. Observed bone uptake was blockable, consistent with known VLA-4 expression in hematopoietic stem cells found in bone marrow. Conclusions: These studies show that these [¹⁸F]R-BF₃ ^?–PEG₂-LLP2A conjugates (11 and 12) are promising VLA-4 targeting radiotracers, yet, further optimization will be required to reduce uptake in the gastro-intestinal tract.

Full text of DOI:10.1016/j.nucmedbio.2018.02.005

Accepted Manuscript
Preliminary evaluation of 18F-labeled LLP2A-trifluoroborate
conjugates as VLA-4 (α4β1 integrin) specific radiotracers for PET
imaging of melanoma
Áron Roxin, Chengcheng Zhang, Sungjoon Huh, Mathieu L.
Lepage, Zhengxing Zhang, Kuo-Shyan Lin, François Bénard,
David M. Perrin
PII:
S0969-8051(18)30022-2
doi:10.1016/j.nucmedbio.2018.02.005
NMB 8002
DOI:
Reference:
To appear in:
Received date:
Revised date:
Accepted date:
17 January 2018
15 February 2018
26 February 2018
Please cite this article as: Áron Roxin, Chengcheng Zhang, Sungjoon Huh, Mathieu
L. Lepage, Zhengxing Zhang, Kuo-Shyan Lin, François Bénard, David M. Perrin ,
Preliminary evaluation of 18F-labeled LLP2A-trifluoroborate conjugates as VLA-4 (α4β1
integrin) specific radiotracers for PET imaging of melanoma. The address for the
corresponding author was captured as affiliation for all authors. Please check if
appropriate. Nmb(2018), doi:10.1016/j.nucmedbio.2018.02.005
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.

ACCEPTED MANUSCRIPT
18
Preliminary evaluation of F-labeled LLP2A-trifluoroborate conjugates as VLA-4 (α β
4
1
integrin) specific radiotracers for PET imaging of melanoma
a
b
a
a
b
Áron Roxin , Chengcheng Zhang , Sungjoon Huh , Mathieu L. Lepage , Zhengxing Zhang ,
b
b
a,
Kuo-Shyan Lin , François Bénard , David M. Perrin *
a
Chemistry Department, University of British Columbia, 2036 Main Mall, Vancouver, British
Columbia, V6T 1Z1, Canada
b
th
Molecular Oncology, British Columbia Cancer Agency Research Center, 765 West 10
Avenue, Vancouver, British Columbia, V5Z 1L3, Canada.
*
Corresponding author information: Tel.: 1 604 822 0567, Fax: 1 604 822 2847
E-mail: dperrin@chem.ubc.ca (D. M. Perrin)
1
8
Key words: F-trifluoroborates, VLA-4, α β integrin, LLP2A peptidomimetic, Positron
4
1
emission tomography, Melanoma
1
8
Abbreviated title: F-LLP2A-BF for VLA-4 specific PET Imaging
3

ACCEPTED MANUSCRIPT
Abstract
Introduction: The transmembrane α β integrin receptor, or very-late antigen 4 (VLA-4), is
4
1
associated with contributing to tumor metastasis and angiogenesis, the development of
chemotherapeutic drug resistance, and is overexpressed in multiple myelomas, osteosarcomas,
lymphomas, leukemias, and melanomas. The peptidomimetic, LLP2A, is a high-affinity ligand
with specificity for the extracellular portion of VLA-4 and several conjugates have been
1
11
68
64
evaluated in vivo by NIR-fluorescence, In-SPECT and Ga- and Cu-PET imaging, but to
1
8
date, not with F-PET.
Methods: Using two highly stable organotrifluoroborate prosthetic groups: ammoniumdimethyl-
trifluoroborate (AMBF ) and a new N-pyridinyl-para-trifluoroborate (N-Pyr-p-BF ), both
3
3
1
8
capable of facile aqueous F-labeling by isotope exchange (IEX), we present the first PET
18
-
imaging evaluations of two [ F]R-BF –PEG -LLP2A tracers using VLA-4 overexpressing B16-
3
2
F10 murine melanoma tumor mouse models.
1
8
Results: Here, we demonstrate successful one-step F-labeling of both conjugates with wet NCA
18
-
[
F]F in radiochemical yields of up to 11.6% within 75 min at molar activities of 40-100
GBq/µmol. Average tumor uptake values based on ex vivo biodistribution values were 4.4 %ID/g
1
8
(
11) and 2.8 %ID/g (12) using F-labeled LLP2A-conjugates with the two prosthetic groups: N-
Pyr-p-BF (5) and alkyl-N,N-dimethylammonio-BF (AMBF ) (7), respectively, and was found
3
3
3
to be target-specific as evidenced by in vivo blocking controls. Dynamic PET scanning and
18
-
biodistribution studies revealed slow clearance of the [ F]R-BF –PEG -LLP2A tracers from the
3
2
tumors, and also substantial uptake in the intestines, gall bladder, liver and bladder. Observed
bone uptake was blockable, consistent with known VLA-4 expression in haematopoietic stem
cells found in bone marrow.

ACCEPTED MANUSCRIPT
1
8
-
Conclusions: These studies show that these [ F]R-BF –PEG -LLP2A conjugates (11 and 12)
3
2
are promising VLA-4 targeting radiotracers, yet, further optimization will be required to reduce
uptake in the gastro-intestinal tract.

ACCEPTED MANUSCRIPT
1
. Introduction
The α β integrin, or very late antigen-4 (VLA-4), is a heterodimeric transmembrane
4
1
1
-3
protein with extracellular portions and is normally involved with lympho-hemopoiesis,
4
, 5
6
leukocyte trafficking, and cell motility. VLA-4 is associated with conferring drug resistance
7
-17
18-22
23, 24
in multiple myeloma
and melanomas.
VLA-4 can also contribute to angiogenesis
2
5
and tumor metastasis. The peptidomimetic, LLP2A, is a high-affinity ligand with specificity for
2
6
the VLA-4 receptor. Since its initial discovery, LLP2A has been evaluated for in vivo imaging
2
7 111
28, 29 99m
30 64
68
of tumors via NIR-fluorescence imaging,
In-SPECT,
Tc-SPECT,
Cu-/ Ga-PET.
3
1-36
Over the past decade, several reports on radiometallated-LLP2A conjugates have
progressively evaluated the use of improved chelators that afford more facile labeling or
improved images in terms of tumor uptake and contrast ratios.
1
8
Nevertheless, to the best of our knowledge, there has been no report using F-PET to
1
8
image the tumor-specific uptake of an F-labeled LLP2A conjugate. Interest in the use of
18
[
F]fluoride ion derives from its well-known advantages that include a moderate half-life (110
6
8
37, 38
min), higher image resolution compared to [ Ga]gallium,
predominant positron decay
+
18
(
>95% via β -emission), a record of FDA approval for F-labeled tracers, and scalability to
1
8
Curie-level production on-demand in hospital cyclotrons. Yet, F-labeling of peptides and other
large molecules has typically been encumbered by multi-step reactions, which make
radiolabeling tedious compared to radiometal chelation.
1
8
39-42
18
Of several methods that simplify F-labeling,
organoboronates undergo F-labeling
1
8
43-48
via F-B bond formation under aqueous conditions.
Towards these ends, we have developed
several aryl and alkyl boronates and trifluoroborates (R-BF ’s) which are solvolytically stable
3
1
8
and have been validated by F-PET imaging studies involving cancer-specific peptide-RBF
3

ACCEPTED MANUSCRIPT
4
5, 48-50
conjugates.
Hence, here we sought to test this method with LLP2A along with PET
imaging by using two promising trifluoroborate prosthetics: 1) a previously reported
5
1
ammoniumdimethyl-trifluoroborate (AMBF ) that had been used to image octreotate,
3
5
2
53
bradykinin and bombesin, and 2) a new prosthetic that has never been used for imaging
before: N-pyridinyl-para-trifluoroborate (N-Pyr-p-BF ) (Scheme 1).
3
1
8
The new F-prosthetic group, N-Pyr-p-BF , was inspired by our previous observations
3
that N-methylation of pyridinyl-p-BF enhanced resistance to solvolytic defluoridation under
3
5
4
18
aqueous conditions . Therefore, both prosthetics were expected to provide facile aqueous F-
labeling by isotope exchange (IEX), to offer molar activities greater than 74 GBq/μmol, and to
4
5, 46, 48, 50, 51
potentially provide PET images with reasonable contrast.
Hence, the primary focus
18
-
of this report was to investigate the VLA-4 targeting specificity of the described [ F]R-BF –
3
PEG -LLP2A tracers while also evaluating the PEG -LLP2A scaffold in the context of two
2
2
different organotrifluoroborate radioprosthetic groups. Animal studies involved the use of B16-
F10 murine melanoma models for PET imaging and biodistribution evaluations. Our
observations would then provide insights to guide future preclinical studies, with the ultimate
1
8
-
goal of advancing [ F]R-BF –PEG -LLP2A tracers toward clinical translation for cancer
3
2
diagnosis and monitoring via VLA-4 specific PET imaging.
2
. Materials and methods
2
.1. General
Potassium bifluoride (KHF ) was purchased from Acros Organics. 4-Aminophenyl acetic
2
acid, Fmoc-Ach-OH, o-tolyl isocyanate, trans-3-(3-pyridyl) acrylic acid, hexafluoroisopropanol
HFIP) were purchased from Alfa Aesar. Ammonium hydroxide (NH OH) was purchased from
(
4
BDH Chemicals. Fmoc-Aad-OH was purchased from Chem Impex International.

ACCEPTED MANUSCRIPT
Copper(II)sulphate, diethyl ether (Et O), dimethylformamide (DMF), methanol (MeOH),
2
pyridine, and sodium hydroxide (NaOH) were purchased from Fisher Scientific.
Iodomethylboronyl pinacolate ester and 4-trifluoroborate-pyridine potassium salt were purchased
from Frontier Scientific. Fmoc-Lys(Dde)-OH and o-bis-(amino-ethyl)ethylene glycol trityl resin
were purchased from Nova BioChem. Hydroxybenzotriazole (HOBt) hydrate, N,N,N′,N′-
tetramethyl-O-(1H-benzotriazol-1-yl)uronium hexafluorophosphate, and o-(benzotriazol-1-yl)-
N,N,N′,N′-tetramethyluronium hexafluorophosphate (HBTU) were purchased from Oakwood
Chemicals. 1-Acetylimidazole, acetonitrile (CH CN), ammonium formate, ascorbic acid,
3
diisopropylethylamine (DIPEA), ethyl-5-bromovalerate, hydrazine hydrate, conc. HCl, N,N-
dimethylproparylamine, piperidine, proparyl bromide, aqueous pyridazinium-HCl, and
trifluoroacetic acid (TFA) were purchased from Sigma Aldrich. Deuterated solvents were
purchased from either Cambridge Isotope Laboratories or Aldrich Chemicals. All chemicals
were used without further purification.
Mass spectra were acquired using a Waters ZQ spectrometer. All samples were prepared
1
in either MeOH or CH CN and acquired using either MeOH or CH CN as the eluent. H,
3
3
1
3
1
19
1
C{ H}, and F{ H} nuclear magnetic resonant (NMR) experiments were performed on a
Bruker AV‐300 (300 MHz) spectrometer. Peaks for NMR spectra are reported as chemical shifts,
δ (ppm), with multiplicity (s = singlet, d = doublet, t = triplet, m = multiplet). NMR
spectroscopic data were processed using ACD/NMR Processor software. HPLC analyses are
reported in minutes as retention time, t . An Agilent 1100 HPLC furnished with an automatic
R
injection unit and multi-channel photodiode array detector was used to purify the radiotracer
precursors. A semi-preparatory HPLC system equipped with a Chromatographic Specialties Inc.
1
8
Knauer Smartline pump 100 and Bioscan radiation detector was used to purify the F-tracers

ACCEPTED MANUSCRIPT
1
8
after F-labeling reactions. An Agilent Technologies 1200 HPLC furnished with a single-
channel 1200 Series photodiode array detector and a Bioscan radiation detector (linked via an
Agilent Interface 35900E unit) was used to develop standard curves for the radiotracer
precursors, to conduct QC analysis of the blocking agent, and to characterize the final
1
8
formulations of the purified F-tracer samples. HPLC method information is found in
Experimental Supporting Information.
2
2
.2 Chemical synthesis
.2.1. Synthesis of H N-PEG -LLP2A(O Bu) (1)
t
2
2
The LLP2A peptidomimetic was synthesized on the solid phase, as previously described
using a solid phase peptide synthesis (SPPS) vessel (Chemglass Life Sciences, CG-
2
6, 27, 29
,
1
866); o-bis-(aminoethyl)ethyleneglycol trityl resin was chosen as this provides a PEG spacer
2
5
5
and a primary amine for bioconjugation upon resin cleavage. A 250 mg sample of resin (max.
estimated loading = 0.49 mmol/g) was suspended in 5 mL of DCM and swollen for 20 min at
room temperature. The DCM was filtered and the resin was washed with DCM (3 x 7 mL). The
resin was then washed with DMF (2 x 5 mL). 1-(Fmoc-amino)cyclohexylcarboxylic acid (134
mg, 367.5 μmol) and HBTU (140 mg, 367.5 μmol) were dissolved in 4.75 mL of DMF with 250
μL of DIPEA (5%, v/v) and the solution was added to the resin. The suspension was bubbled
with argon for 2 h at room temperature, after which time the resin was filtered and washed with
DMF (4 x 5 mL). 1-Acetylimidazole (40.5 mg, 367.5 μmol) and HBTU (140 mg, 367.5 μmol)
were dissolved in 5 mL of DMF with 5% DIPEA (v/v), the solution was added to the resin, and
coupling proceeded as described for the previous step. The Fmoc protecting group was removed
using 20% piperidine in DMF (v/v) (5 x 6 mL) while bubbling the suspension with argon at room
temperature for 5 min/treatment. The resin was then washed with DMF (6 x 6 mL). Fmoc-

ACCEPTED MANUSCRIPT
t
homoGlu(O Bu)-OH (162 mg, 367.5 μmol) and HBTU (140 mg, 367.5 μmol) were dissolved in
.75 mL of DMF with 250 μL of DIPEA (5%, v/v) and the solution was added to the resin.
Coupling and deprotection were effected as described above. Fmoc-Lys(Dde)-OH (196 mg,
67.5 μmol) and HBTU (140 mg, 367.5 μmol) were dissolved in 4.75 mL of DMF with 250 μL
4
3
of DIPEA (5%, v/v) and the solution was added to the resin. Coupling and deprotection were
effected as described above. 2-(4-(3-(o-Tolyl)ureido)phenyl)acetic acid (synthesized following
2
6, 27, 29
previously published methods)
(104.4 mg, 367.5 μmol) and HBTU (140 mg, 367.5 μmol)
were dissolved in 4.75 mL of DMF with 250 μL of DIPEA (5%, v/v) and the solution was added
to the resin. Coupling and deprotection were effected as described above. The Dde protecting
group of lysine was removed using ~1% hydrazine in DMF (v/v) (i.e. diluted 50-60% hydrazine
hydrate in DMF at a 1:49, v/v, ratio, 3 x 5 mL) while bubbling the solution with argon at room
temperature for 4 min/treatment. The resin was filtered and washed as described. (E)-3-
(Pyridine-3-yl)acrylic acid (54.8 mg, 367.5 μmol) and HBTU (140 mg, 367.5 μmol) were
dissolved in 4.75 mL of DMF with 250 μL of DIPEA (5%, v/v) and the solution was added to the
resin. Coupling and deprotection were effected as described above. The resin was suspended in 2
mL of DCM and transferred to 1.5 mL Eppendorf tubes and dried by speed-vac at room
t
temperature. This gave 292.3 mg of LLP2A(O Bu)-PEG on resin. A sample of the
2
t
LLP2A(O Bu)-PEG (72.5 mg) on resin was cleaved with 1 mL of (3:7) HFIP: DCM (v/v) within
2
3
0 min at room temperature. The resin was then washed with 1.5 mL of DCM and the
suspension was filtered to collect the supernatant. The solutions were combined and concentrated
by speed-vac at room temperature. The dried sample was redissolved in 50 μL of DMF and 1 mL
of cold Et O was used to precipitate the desired product. The suspension was centrifuged at
2
8
,000 rpm at room temperature for 3 min and the supernatant was discarded. After a total of 4

ACCEPTED MANUSCRIPT
rounds of DMF/Et O precipitation, the pellet was dried by speed-vac at room temperature and
2
t
stored at -20 °C. This gave 33.8 mg (33.8 μmol) of H N-PEG -LLP2A(O Bu) (1), for a loading
2
2
of 0.48 μmol of 1 per mg of resin and a near quantitative yield overall. The precursor, 1, was
used for synthesis without further purification. Characterization data are found in Experimental
Supporting Information.
2
.2.2. Synthesis of 5-Azidopentanoic acid (2)
Ethyl-5-bromovalerate (757 μL, 4.77 mmol, 1.0 eq.) was dissolved in 5 mL of DMSO in
a 25 mL round bottom flask, into which was then dissolved NaN (1.24 g, 19.25 mmol, 4.04 eq.),
3
and the clear solution was brought to reflux at 100 °C for 20 h while stirring and then allowed to
cool to room temperature. The desired product was extracted into Et O (4 x 25 mL), which was
2
then washed with brine (4 x 40 mL), dried by addition of ~1g Na SO , then filtered and the
2
4
solvent removed by rotary evaporation. The resulting 640 mg of clear oil was dissolved in 1 mL
of MeOH and 5 mL of aqueous 1M NaOH. The stirred emulsion became a yellow solution
within 3.5 h at room temperature and the hydrolysis reaction was continued for a total of 19 h at
room temperature. The MeOH was removed by rotary evaporation and the remaining solution
was acidified using conc. HCl. The product was extracted into Et O (4 x 25 mL), which was then
2
dried over Na SO , filtered, and the solvent was removed by rotary evaporation. This gave 400
2
4
mg of 2 (280 μmol) for a 59% yield. Characterization data are found in Experimental Supporting
Information.
2
.2.3. Synthesis of N -(CH ) -C(O)NH-PEG -LLP2A (4)
3 2 4 2

ACCEPTED MANUSCRIPT
Cpd 1 (20.4 mg, 20.44 μmol, 1.0 eq.) was dissolved in 336 μL of DMF. HOBt-hydrate
was dissolved in DMF to prepare a 1 mg/10 μL stock solution. EDC-HCl (8.2 mg, 63.8 μmol, 3.1
eq.) was placed in a 1.5 mL Eppendorf tube and suspended in 64 μL of the HOBt-hydrate stock
solution (63.77 μmol, 3.12 eq.). Added 11.9 μL (63.77 μmol, 3.12 eq.) of 2, 14.2 μL (306.6
μmol, 15.0 eq.) of pyridine, and vortexed until all reagents were fully dissolved. The solution of
1
in DMF was combined to the aforementioned solution and the tube was placed on a rotisserie
rotator for 14 h at room temperature. This solution was divided equally into eight aliquots and
placed in 1.5 mL Eppendorf tubes, to which 1 mL of cold Et O was used to precipitate the
2
desired product. The suspensions were centrifuged at 8,000 rpm at room temperature for 3 min
and the supernatant was discarded. The pellets were each redissolved in 50 μL of DMF and the
Et O/centrifugation method was repeated. After a total of 3 rounds of DMF/Et O precipitations,
2
2
t
the pellets containing N -(CH ) -C(O)NH-PEG -LLP2A(O Bu) (3) were dried by speed-vac at
3
2 4
2
room temperature. Each sample was then dissolved in 418 μL of (1:1) TFA: DCM (v/v) and the
solution was allowed to react at room temperature for 1 h. The solution was then concentrated by
speed-vac and precipitated with 1 mL of cold Et O. The suspension was then centrifuged,
2
redissolved in 40 μL of DMF, and precipitated with Et O, as described. After a total of three
2
rounds of DMF/Et O precipitations, each of the pellets were dried by speed-vac at room
2
temperature and were stored at -20 °C. This gave 11.0 mg of N -(CH ) -C(O)NH-PEG -LLP2A
3
2 4
2
(4) (10.31 μmol) for a 50% yield. Characterization data are found in Experimental Supporting
Information.

ACCEPTED MANUSCRIPT
2
.2.4. Synthesis of N-Propargyl-pyridinium-4-trifluoroborate (5)
4
-Trifluoroborate-pyridine potassium salt (200 mg, 1.087 mmol, 1.0 eq.) was dissolved in
5
mL of DMF in an oven-dried 25 mL round bottom flask and sealed with a rubber septum.
Propargyl bromide (80% solution in toluene, v/v, 1.33 mL, 14.13 mmol, 13 eq.) was added and
the solution was stirred at 45 °C for 2 h. The solution was transferred to a 15 mL the falcon tube
and the solvent was removed overnight by speed-vac. The crude reaction contents were
redissolved in CH CN, dry-loaded onto an F60 silica gel column (2 cm x 17 cm), and the product
3
was purified accordingly using (1:10) MeOH: EtOAc (v/v) while monitoring by TLC [(1:10)
MeOH: EtOAc (v/v), R = 0.3, UV]. The solvent was removed from pooled pure fractions by
f
rotary evaporation. This gave 135 mg (446 μmol) of N-propargyl-pyridinium-4-trifluoroborate
(5) in a 41% yield. Characterization data are found in Experimental Supporting Information.
2
.2.5. ((Dimethyl(prop-2-yn-1-yl)ammonio)methyl)trifluoroborate (7)
Into a flame-dried 50 mL rbf was dissolved N,N-dimethylpropargylamine (146 μL, 1.37
mmol, 1.0 eq.) contained in 10 mL of Et O. The vessel was sealed with a septum, into which was
2
injected iodomethylboronyl pinacolate ester (250 μL, 1.37 mmol, 1.0 eq.) and the solution was
stirred for 6 min at room temperature. The white precipitate that formed was filtered, was
washed with 40 mL of Et O, and the solvent was removed in vacuo. This gave 89 mg (397 μmol)
2
of N-proparyl-N,N-dimethylammoniomethyl-boronyl pinacolate (6) in a 19% yield. Cpd 6 (30
mg, 134 μmol, 1.0 eq.) was dissolved in 360 μL of CH CN in a 1.5 mL Eppendorf tube, to which
3
was added aqueous 3M KHF (127.5 μL, 375 μmol, 2.8 eq.) and aqueous 4M HCl (127.5 μL, 509
2
μmol, 3.8 eq.) The tube was placed on a heating block and the contents were allowed to react at

ACCEPTED MANUSCRIPT
4
5 °C for 2 h, after which time, the reaction was quenched with 450 μL of conc. aqueous
NH OH and diluted with 3.0 mL of (1:4) EtOH: CH CN (v/v). The crude solution was forced
4
3
through a silica gel plug using (1:4) EtOH:CH CN (v/v) while monitoring by TLC [(1:1) MeOH:
3
DCM (v/v), R = 0.8, Si-iodine]. The desired fractions were pooled and concentrated by rotary
f
evaporation, re-suspended in CH CN (3 mL), filtered, and the supernatant was removed by
3
rotary evaporation. This gave a ~22 mg (~133 μmol) of ((dimethyl(prop-2-yn-1-
yl)ammonio)methyl)trifluoroborate (7), and some remaining KHF . Characterization data are
2
found in Experimental Supporting Information.
-
2
.2.6. Synthesis of N-Pyr-p-BF -PEG -LLP2A (8)
3
2
Cpd 5 (2.2 mg, 11.8 μmol, 1.0 eq.) was dissolved in 44 μL of (1:1) CH CN: H O (v/v),
3
2
and stock solutions of aqueous 1M Cu(II)SO and an aqueous 1M sodium ascorbate were
4
individually prepared. Cpd 4 (11.95 mg, 11.2 μmol, 0.95 eq.) was dissolved in 556 μL of (1:1)
CH CN: H O (v/v), to which was added 1M sodium ascorbate (49.2 μL, 49.2 μmol, 4.17 eq.) and
3
2
1
1
M Cu(II)SO (19.7 μL, 19.7 μmol, 1.67 eq.), to which was then added (1:1) CH CN: aqueous
4 3
.5M NH OH (v/v) (160 μL). The solutions containing 4 and 5 were combined and placed on a
4
heating block at 45 °C for 2.5 h, and then concentrated by speed-vac at 45 °C for 4 h. The crude
reaction was resuspended in CH CN and eluted through a C18 silica gel plug with CH CN to
3
3
remove salts while monitoring the desired product by TLC [(95:5) EtOH: conc. NH OH (v/v), R
4
f
=
0.3, UV, HBQ, Si-iodine]. The desired fractions were combined and concentrated by speed-
vac. The sample was then redissolved in (3:7) ACN: aqueous 40 mM NH HCO and purified by
4
2
HPLC using method A while monitoring the absorbance of the eluted product at 257 nm (t =
R
1
1.2 min – 12.0 min). The desired fractions were pooled and immediately frozen by placement

ACCEPTED MANUSCRIPT
on dry ice. Once frozen, these were lyophilized. This procedure provided 3.5 mg (2.8 μmol) of
-
N-Pyr-p-BF -PEG -LLP2A (8) in a 25% yield. Characterization data are found in Experimental
3
2
Supporting Information.
-
2
.2.7. Synthesis of AMBF -PEG -LLP2A (9)
3 2
A stock solution of 7 (1.0 mg/20 μL) in (1:1) CH CN: H O (v/v) was prepared. Cpd 4 (11
3
2
mg, μmol, 1.0 eq.) was dissolved in 594 μL of (1:1) CH CN: H O (v/v). A solution of the
3
2
following components was prepared: 1M sodium ascorbate (43 μL, 43 μmol, 4.17 eq.), 1M
Cu(II)SO (17.2 μL, 17.2 μmol, 1.67 eq.) and (1:1) CH CN: 1.5 M NH OH (v/v) (175 μL). This
4
3
4
Cu(I)-containing solution was added to the solution containing 4, along with the solution of 7
106 μL, 31.9 μmol, 3.1 eq.) to initiate the copper-assisted ‘click’ reaction. The reaction contents
(
were aliquoted (~440 μL each) into two 1.5 mL Eppendorf tubes, which were placed on a
heating block at 45 °C for 2 h and then concentrated to dryness by speed-vac at 45 °C for 3.5 h.
H O (30 μL) and then CH CN (600 μL) was added to each tube, which were vortexed, sonicated,
2
3
and centrifuged at 13,000 rpm for 1 min. The supernatants were combined and retained (presence
of the desired product was verified by TLC [(95:5) EtOH: conc. aqueous NH OH (v/v), R =
4
f
0
.55, UV, Si-iodine]) while the pellet was retained in the original tubes. H O and then CH CN
2 3
was again used to extract the desired product from the pellet, as described above, until the entire
product sample was extracted from the precipitate as determined by TLC analysis. The
supernatants were concentrated by speed-vac and then redissolved in (45:55) CH CN: 40 mM
3
NH HCO (v/v) and purified by HPLC using method B while monitoring the absorbance of the
4
2
-
eluted product at 257 nm (t = 11.2 min – 12.0 min). This gave 4.0 mg (3.25 μmol) of AMBF -
R
3

ACCEPTED MANUSCRIPT
PEG -LLP2A (9) for a 32% yield. Characterization data are found in Experimental Supporting
2
Information.
2
.2.8. Synthesis of H N-PEG -LLP2A (10)
2 2
Cpd 1 (10 mg, 10. μmol) was dissolved in 300 μL of (1:1) TFA: DCM (v/v) and allowed
to stand at room temperature for 1 h. The solution was concentrated by speed-vac at room
temperature for 1 h and was precipitated with 1 mL of Et O. The desired product was isolated by
2
centrifugation at 13,000 rpm for 1 min. The precipitate was collected and the supernatant was
discarded. The product was redissolved in DMF (40 μL) and the precipitation was repeated four
time by repeated addition of Et O and centrifugation. The desired sample was concentrated to
2
dryness by speed-vac at room temperature for 1.5 h. This gave 5.2 mg (5.53 μmol) of H N-
2
PEG -LLP2A (10) in a 55% yield. The purity (~95%) of 10 was assessed by HPLC using
2
method C (t = 21.4 min). Characterization data are found in Experimental Supporting
R
Information.
18
-
2
.3. Radiosynthesis of [ F]R-BF -PEG -LLP2A tracers (11 & 12)
3 2
The radiotracer precursors, 8 or 9 (80 nmol), were each dissolved in DMF (15 μL),
aqueous 1M pyridazine-HCl pH = 2 (15 μL), and aqueous 5 mM KHF (4 μL) to bring the final
2
1
8
-
pH to 2.0-2.5. NCA (no-carrier-added) [ F]F was produced using a TR19 cyclotron (Advanced
1
8
18
18
Cyclotron Systems Inc.) via the O(p,n) F reaction. Approximately 37-74 GBq of [ F]fluoride
ion was eluted from a 9 mg QMA cartridge with ~80-100 μL saline (0.9%, w/v) directly into
septa-sealed reaction vessels containing the precursors. The solutions were heated at 82-84 °C
for 5 min and then heated in vacuo for 15 min. After quenching with aqueous 40 mM NH HCO
4
2
1
8
(
pH = 6.8), F-tracers were purified by semi-preparative HPLC with isocratic solvent conditions

ACCEPTED MANUSCRIPT
1
8
-
18
-
using method D for [ F]N-Pyr-p-BF -PEG -LLP2A (11) and method E for [ F]AMBF -PEG -
3
2
3
2
LLP2A (12). Collected fractions were diluted into 50 mL of H O and passed through Sep-Pak
2
1
8
C18 Light cartridges with helium to trap the F-tracers. After eluting with 0.5 mL of (9:1) EtOH
in 0.9% saline, the samples were formulated with 0.9% saline to reduce the concentration of
1
8
EtOH to ~10% (v/v). Samples of these F-tracers were used to assess the radiochemical purity
and UV-vis absorption by analytical HPLC using method C for 11 and method F for 12.
Characterization data are found in Experimental Supporting Information.
2
.4. Cell Culture and Tumor implantation
The B16-F10 cell line (Mus musculus) was obtained from ATTC (CRL-6475). Cells were
cultured in Dulbecco's Modified Eagle's Medium (StemCell Technologies) at 37 °C,
supplemented with 10 % FBS (v/v), 100 U/mL penicillin and 100 μg/mL streptomycin. All
animal experiments were conducted according to the guidelines established by the Canadian
Council on Animal Care and were approved by the Animal Ethics Committee of the University
of British Columbia. For tumor implantation, male C57BL/6J mice were anesthetized by
6
inhalation of 2% isoflurane, and 1×10 B16-F10 cells were injected subcutaneously on the right
shoulder. Mice were imaged or used in biodistribution studies once the tumors reached 7-9 mm
in diameter.
2
.5. Saturation Binding Assays
Cpds 11 and 12 were labeled by IEX at molar activities in excess of 37 GBq/µmol and
purified by HPLC. In vitro saturation binding assays were performed on B16-F10 cells in
5
6
duplicate following published procedures (n = 3). Briefly, cells were grown to near-confluence
on 24 well poly-D-lysine plates. Growth media was removed, and reaction media (RPMI, 1%

ACCEPTED MANUSCRIPT
BSA, 100 U/mL penicillin/streptomycin) was added and allowed to incubate for ~1 hour at 37
°
C. Increasing concentrations (5 pM to 10 nM) of 11 or 12 were added to the cells and incubated
for 1 h at 25 °C with mild agitation. Non-specific binding was determined by repeating the
described incubations with 11 or 12 while simultaneously adding 10 (10 μM). After incubation,
cells were washed twice with ice-cold PBS, harvested following incubation with trypsin, and
measured for radioactivity using a WIZARD 2480 gamma counter (PerkinElmer). Values from
the non-specific binding assays were subtracted from the respective values of the specific
’
binding assays. Representative dissociation constants (K s) were calculated using GraphPad
d
PRISM 7 with a one-site specific binding model, while noting the potential for cell
internalization.
2
.6. Preclinical PET/CT imaging
PET/CT imaging studies were conducted using a microPET/CT scanner (Inveon,
5
1-53, 57
Siemens) as previously reported
Briefly, for static PET scans, each tumor bearing mouse
was injected with 4-6 MBq of 11 or 12 via the tail vein while under 2% isoflurane anesthesia
n=2 for each). After the injection, mice were allowed to recover and roam freely in their cages,
(
and were subsequently sedated with 2% isoflurane and positioned in the scanner. A 10 min
baseline CT scan was obtained, followed by a single 10 min static PET scan at 60 min p.i.. For
static PET images with blocking agents, 200 μg of 10 was co-injected with 11 or 12 (n=2 for
each). For dynamic PET scans, a baseline CT scan was obtained, followed by intravenous
1
8
injection with 4-6 MBq of F-labeled tracer and dynamic PET acquisition of 60 min with the list
mode data rebinned to time intervals at 12×10 sec, 8×60 sec, 7×300 sec and 1×900 sec (n=1).
While still under anesthesia, the mice were euthanized using CO inhalation following imaging.
2

ACCEPTED MANUSCRIPT
2
.7. Biodistribution studies
For biodistribution studies, mice were anesthetized by 2% isoflurane inhalation, and
injected with 1-2 MBq of 11 (n=6) or 12 (n=8). For blocking studies, 200 μg of 10 was co-
injected with 11 or 12 (n=4 for each). After the injection, mice were allowed to recover and roam
freely in their cages, and then euthanized by CO inhalation at 60 min p.i. Blood was promptly
2
withdrawn, and the organs of interest were harvested, rinsed, and blotted dry. The organs were
then weighed and their radioactivity was measured on a Wallac WIZARD2 gamma counter
(PerkinElmer), normalized to the injected dose using a standard curve and expressed as the
percentage of the injected dose per gram of tissue (% ID/g).
2
.8. Statistics
Two-tailed, unpaired t tests (with Welch's correction) were used to evaluate statistical
significance with GraphPad PRISM software.
3
. Results
3
.1. Chemical synthesis
2
6, 27, 29
The LLP2A peptidomimetic was synthesized as previously described.
The
synthesis was conducted on the solid phase with o-bis-(aminoethyl)ethyleneglycol trityl resin to
provide a PEG spacer and a primary amine conjugation handle upon resin cleavage as
2
5
5
previously demonstrated in our lab and known to retain subnanomolar affinity. The modified
t
LLP2A derivative, H N-PEG -LLP2A(O Bu) (1) was synthesized via several high yielding
2
2
reactions and was used as the intermediate for the synthesis of radiotracer precursors appended
with organotrifluoroborate prosthetic groups via copper-catalyzed “click” conjugation (Scheme

ACCEPTED MANUSCRIPT
1
). Alkyne-modified organotrifluoroborates, 5 and 7, were each coupled to the azide-PEG2-
LLP2A derivative, 4, via ‘click’ chemistry to provide the radiotracer analogues 8 and 9,
respectively. Crude samples were purified by HPLC and lyophilized to obtain 8 and 9 in 25%
and 32% chemical yields, respectively. The blocking agent, 10, was obtained by treating the
t
O Bu-protected LLP2A derivative, 1, with 50% TFA for a 55% yield. Complete Characterization
data are found in Experimental Supporting Information.
3
.2. Radiosynthesis
1
8
The radiotracer precursors, 8 and 9, were F-labeled by IEX under acidic aqueous
1
8
conditions (Scheme 2) as previously reported. The radiochemical purity of each F-tracer was
assessed by HPLC (Fig. 1) and the molar activity (A ) of each was determined using HPLC-
m
based standard curves (λabs = 257 nm) generated from 8 and 9 and the corresponding UV-vis
absorption chromatograms of 11 and 12. Complete characterization data are found in
1
8
Experimental Supporting Information. The F-labeling results for 11 and 12 are summarized in
Table 1. The average time from the start of synthesis until formulation were respectively was
just under 75 min for 11 and 12. Nevertheless, the average radiochemical yield for 11 (11.3%)
was ~3 times higher than that for 12 (3.7%), and the average A of 11 (100 GBq/μmol) was ~2.5
m
times higher compared to that for 12 (41 GBq/μmol). The average radiochemical purities were
similar (>95%) for both 11 and 12.
3
.3. Binding affinity
Saturation binding assays were performed to address potential differences in affinity; 8
1
8
and 9 were F-labeled by IEX to produce 11 and 12, and each was used in a saturation binding
assay with B16-F10 cells. Notably, 11 was found to bind ~5-fold more tightly (p < 0.0001) than

ACCEPTED MANUSCRIPT
1
2 with respective K ’s of 0.28 ± 0.02 nM and 1.52 ± 0.03 nM as shown in Fig. 2. Gratifyingly,
d
6
4
the K measured for 11 is similar to the K ’s reported for Cu-labeled NODAGA-PEG -LLP2A
d
d
4
conjugates and may explain, in part, why tumor uptake values were slightly higher for 11
3
4
compared to 12 (vide infra). Yet, the K values we calculated for 11 and 12 may be higher due
d
to potential cell internalization.
3
.4. PET imaging
The static PET images of 11 at 1 h p.i. show distinct uptake in the tumor and in the gut,
bladder and bone marrow (Fig. 3a and 2b). Blocking studies involving co-injections of 10
showed essentially unchanged uptake of 11 in the gut and the bladder, however, with almost no
uptake in the tumor or bone marrow (Fig. 3c and 2d). The static PET imaging results from 11
were qualitatively similar to those from 12 (Fig. 4), although 12 showed slightly lower tumor
uptake (Fig. 4a and 3b). This conclusion is confirmed by a) time-activity-curves obtained from
dynamic PET scan data and b) ex vivo biodistribution data.
Dynamic PET scans showed similar uptake and clearance patterns for 11 (Fig. 5) and 12
1
8
(
Fig. 6). Each F-tracer displayed consistently low uptake in the muscle over 60 min (0.6-1
%
ID/g for 11, and 0.4-0.6 %ID/g for 12). Each tracer also showed fast perfusion through
circulation (6 %ID/g for 11, and 20 %ID/g for 12 at 1 min p.i.) while both tracers cleared after 60
min (1 %ID/g for 11, and 0.5 %ID/g for 12), as indicated by the low residual activity recorded in
the heart. Interestingly, both 11 and 12 showed high gall bladder uptake (1.5-38 %ID/g for 11,
and 4.4 %-16.7 %ID/g for 12), and high intestinal uptake (4.4 %-25 %ID/g for 11, and 4 %-81.6
%
ID/g for 12), both of which increased with time. While gall-bladder and intestinal uptake 11
and 12 remained high and stable over 60 min, liver uptake steadily decreased (22-3.6 %ID/g for

ACCEPTED MANUSCRIPT
1
1, and 10 %-1.8 %ID/g for 12) over the same time period. Uptake values for 11 in the tumor
and the bone marrow (2.6 %ID/g and 3.7 %ID/g, respectively) remained essentially constant
throughout the dynamic PET scan, yet these values were slightly higher compared to those
obtained for 12 (1.7-1.4 %ID/g and 1.4-0.9% %ID/g, respectively) and clearance rates of 11 and
1
2 were similar in these tissues that displayed high expression of VLA-4.
.5. Biodistribution
3
The described animal model was used to evaluate the biodistribution of 11 and 12 (1-2
MBq) at 1 h p.i. (Table 2). These results showed that the average tumor uptake of 11 (4.41
ID/g) was ~50% higher (p < 0.01) compared to 12 (2.84 %ID/g). Blocking studies were
%
performed as described and revealed that both 11 (9.8-fold higher average uptake when
unblocked, p < 0.0001) and 12 (8.4-fold higher average uptake when unblocked, p < 0.001) were
specific to VLA-4 expressed on the B16-F10 tumors. These blocking studies also showed that 11
and 12 specifically targeted VLA-4 that is natively expressed in bone marrow (9.1-fold higher
average uptake for 11 when unblocked, p < 0.001 and 8.4-fold higher average uptake of 12 when
1
8
unblocked, p < 0.0001) and that co-injections of 10 reduced F-tracer retention in the spleen
21-fold higher average uptake for 11 when unblocked, p = 0.0001 and 20-fold higher average
(
uptake of 12 when unblocked, p < 0.0001). Interestingly, considerable average intestinal uptake
was observed at 1 h p.i. of 11 (52 %ID/g) and 12 (51 %ID/g). Yet, the average radioactivity
measured in the intestines was significantly reduced by the blocking agent, 10, for 11 (46%
reduction, p = 0.0395) and 12 (41% reduction, p = 0.0385). The biodistribution measurements
were used to calculate several important tumor-to-tissue ratios for 11 and 12 (Table 3). These
revealed no significant differences for the tumor-to-blood, -muscle, -bone, and –kidney ratios for
1
1 compared to 12. However, blocking of 11 and 12 with 10 reduced the average tumor-to-blood

ACCEPTED MANUSCRIPT
(4.8-fold lower for 11, p = 0.0007 and 6.0-fold lower for 12, p = 0.0031) and tumor-to-kidney
ratios (4.8-fold lower for 11, p < 0.0001 and 9.4-fold lower for 12, p = 0.0037) compared to
1
8
injecting each F-tracer alone.
4
. Discussion
The general synthesis of the LLP2A peptidomimetic ligand was accomplished according
2
6
to previous reports, with a modification of the chosen spacer. In lieu of using the (PEG -
2
2
6, 27, 29-32, 58
succinamide) -lysine(ε-NH )amide linker that has previously been employed,
we
2
2
adopted our SPPS to incorporate a shorter PEG -linker at the C-terminus of LLP2A based on
2
considerations for a more expedient synthesis. This strategy provided a hydrophilic spacer and a
primary amine handle for bioconjugation upon resin cleavage while retaining the VLA-4 binding
5
5
18
affinity of the LLP2A moiety at subnanomolar levels. The desired F-tracer precursors, 8 and
9
, were easily synthesized using a heterobifunctional linker, 2, and copper-assisted ‘click’
couplings of the azide derivative of LLP2A (4), and each of the alkynyl trifluoroborate prosthetic
groups (i.e. 5 and 7).
1
8
The F-labeling reactions proceeded by IEX under acidic aqueous conditions. The time
1
8
18
18
between trapping the [ F]fluoride ion from H O and formulation of the final F-tracers, 11
2
and 12, were similar. Owing to the use of HPLC purification, the radiochemical purities of both
1
1 and 12 were high. Radiochemical yields and molar activity values for 11 were greater
compared to 12. This may be explained by the lower solvolytic stability of 5 in aqueous media
4
9, 54
compared to 7, that is inversely correlated with the rate of IEX.
Nevertheless, the low
radiochemical yields in both cases are surprising compared to yields seen with other peptide

ACCEPTED MANUSCRIPT
1
8
conjugates that have been F-labeled via the radioprosthetic 7. Hence, we speculate that these
lower yields may be due to poor solubility of the LLP2A-conjugates in DMF-water at pH 2.
1
8
Since our main objective in this study was to provide the first preliminary PET images of F-
LLP2A tracers, we did not fully optimize radiolabeling yields.
Notably, 11 provided higher radiochemical yields and consequently higher molar activity
compared to 12. The higher molar activity along with a nearly 6-fold affinity for the target,
assuming similar rates of potential cell internalization, may explain why 11 showed a ~50%
higher tumor uptake in both the dynamic PET scans and biodistribution data compared to 12.
Static PET scans of both 11 and 12 provided images of the B16-F10 murine melanoma tumors
with visible contrast ratios; most notably, the tumor:blood ratio of 14 and the tumor:muscle ratio
6
4
29
of 8 are comparable to those seen with Cu-labeled LLP2A. PET images also illustrated the
3
, 4, 9, 16, 17, 59, 60
well-known, natural expression of VLA-4 in the bone marrow,
while the blocking
studies confirmed that the signals in the bone marrow were due to VLA-4 targeting of 11 and 12
1
8
and were not due to i) the presence of free [ F]fluoride ion, ii) solvolytic defluoridation, nor iii)
metabolic defluoridation. The dynamic PET scans showed slow clearance of 11 and 12 from
both the tumors and bone marrow, consistent with VLA-4 expression in these tissues. There were
also no significant differences in the tumor:bone ratios obtained using 11 or 12 compared to their
respective blocking studies. This, in combination with the dynamic PET scans for 11 and 12,
1
8
suggests that the uptake of the F-tracers was similar in the tumor and in bone marrow. There
are two explanations for this: 1) VLA-4 expression in bone marrow is similar to that in tumor or
2
) VLA-4 expression is very different between bone and tumor yet potential PK differences (ex.
faster uptake in bone marrow with lower expression levels vs. slower uptake in tumor with

ACCEPTED MANUSCRIPT
higher expression levels) may still account for the apparently similar uptake patterns observed
and thus further investigation into this effect may be warranted.
The tumor:muscle ratios of 11 and 12 alone were not significantly different compared to
their respective blocking conditions. This suggests that the blocking agent, 10, also served to
1
8
reduce the non-specific background uptake of the F-tracers. This is also reflected by the general
1
8
reduction of non-specific F-tracer uptakes in almost all tissues during blocking experiments.
1
8
The high static PET scan signal in the intestines suggest that both F-tracers cleared
predominantly via the hepatobiliary pathway, although not exclusively since high bladder uptake
64
also suggests renal clearance. Interestingly, previous PET imaging studies with both Cu- and
6
8
Ga-labeled LLP2A conjugates did not show significant uptake in the digestive tract, aside from
3
1-33, 36
18
the liver.
Yet, the intestinal uptake seen with 11 and 12 cannot be ascribed to the F-
labeled organotrifluoroborate prosthetic groups; indeed if the gut had specific receptors for the
trifluoroborate prosthetic group, then all peptide-BF conjugates previously imaged would have
3
cleared through the gut. By contrast, other peptides that have been similarly labeled with
prosthetic group 7 displayed marginal gut uptake. We, therefore, assert that the gut clearance we
observe herein was due to endogenous transporters that predominantly recognize the relatively
hydrophobic LLP2A construct. Further evidence of this assertion is seen in the blocking controls
whereby LLP2A suppressed gut uptake 2-fold, suggesting that the basis for gut uptake is
predominantly due to the transport of LLP2A and not the prosthetic groups.
Biodistribution analyses and PET imaging also revealed that 11 and 12 exhibited high
uptake in the liver, gall bladder and intestinal tissues which may have been due to the
comparatively hydrophobic properties of both 11 and 12, compared to previously described
radiometallated LLP2A derivatives. Although gradual liver clearance suggested that 11 and 12

ACCEPTED MANUSCRIPT
1
8
18
were metabolized in the liver, metabolism of the F-tracers did not liberate free [ F]fluoride
ion, as this would have shown significant time-dependent and unblockable bone accumulation.
One appreciable difference between 11 and 12 and various radiometallated LLP2A analogs is
that in the case of radiometallated versions, longer PEG linkers are used; while it is known that
the linker does not appreciably affect target affinity, longer PEG linkers are known to favor renal
clearance over hepatobiliary clearance. Other differences, most notably the radioprosthetic group
itself, may also contribute to the lower tumor uptake due to rapid clearance while possibly
promoting higher uptake in liver and transport to the intestines. While this work highlights
success with a new radioprosthetic group (N-Pyr-p-BF ), specific data acquired with LLP2A now
3
suggests a basis for applying alternate linkers with increased hydrophilicity to potentially reduce
gut uptake and further enhance contrast ratios. Such efforts are underway and will be reported in
due time.
5
. Conclusions
The use of F-labeled organotrifluoroborates has been applied to LLP2A to enable the
1
8
1
8
first-ever F-PET images of the LLP2A peptidomimetic. In addition, we have featured the use of
a new organotrifluoroborate radioprosthetic group that heretofore has never been reported.
18
-
Herein, we described the facile radiosynthesis of two [ F]R-BF -PEG -LLP2A tracers, 11 and
3
2
1
2, containing either an aryl (5) or an alkyl (7) trifluoroborate prosthetic group (respectively),
1
8
each of which is highly stable to solvolytic defluoridation. The F-tracer, 11, bearing
radioprosthetic group, 5, gave higher radiochemical yields and higher A values compared to 12
m
that bore radioprosthetic group 7. PET imaging, biodistribution analysis and blocking controls
confirmed that 11 and 12 specifically targeted the VLA-4 receptors expressed in the B16-F10
murine melanoma tumors, as well as in bone marrow. While similar data obtained from both

ACCEPTED MANUSCRIPT
tracers suggest that the radioprosthetic group did not greatly affect uptake values or contrast
ratios, yields and molar activities were higher with the new radioprosthetic group, which
suggests radiosynthetic advantages with all other outcomes being indistinguishable. In summary,
the salient points of this work include: i) the identification of a new radioprosthetic group that
1
8
expands the repertoire of F-labeled organotrifluoroborates for PET imaging, and ii) extension
18
18
of this F-labeling method to LLP2A, thus permitting F-PET imaging for the first time.
Acknowledgements
We acknowledge the Canadian Institutes of Health Research for support. We thank
cyclotron operators Wade English, Baljit Singh, and Milan Vuckovic, and PET imaging
technologists, Nadine Colpo and Navjit Hundal-Jabal, for their technical assistance.
Ethical statement
All animal experiments were conducted according to the guidelines established by the
Canadian Council on Animal Care and were approved by the Animal Ethics Committee of the
University of British Columbia.
Appendix A: Experimental Supporting Information
Experimental Supporting Information to this article can be found online at http://...

ACCEPTED MANUSCRIPT
References
[
1] K Miyake, Weissman I, Greenberger J, and Kincade P. Evidence for a role of the integrin
VLA-4 in lympho-hemopoiesis. J. Exp. Med. 173 (1991); 599-607.
[
2] K Miyake, Hasunuma Y, Yagita H, and Kimoto M. Requirement for VLA-4 and VLA-5
integrins in lymphoma cells binding to and migration beneath stromal cells in culture. J. Cell.
Biol. 119 (1992); 653-62.
[
3] Y Imai, Shimaoka M, and Kurokawa M. Essential roles of VLA-4 in the hematopoietic
system. Int. J. Hematol. 91 (2010); 569-75.
[
1
4] DM Rose, Han J, and Ginsberg MH. α4 integrins and the immune response. Immunol. Rev.
86 (2002); 118-24.
[
5] Y-M Hyun, Chung H-L, McGrath JL, Waugh RE, and Kim M. Activated integrin VLA-4
localizes to the lamellipodia and mediates T cell migration on VCAM-1.J Immunol. 183 (2009);
59-69.
3
[
6] DA Hsia, Lim S-T, Bernard-Trifilo JA, Mitra SK, Tanaka S, Den Hertog J, et al. Integrin
α4β1 promotes focal adhesion kinase-independent cell motility via α4 cytoplasmic domain-
specific activation of c-Src. Mol. Cell. Biol. 25 (2005); 9700-12.
[
7] JS Damiano and Dalton WS. Integrin-mediated drug resistance in multiple myeloma. Leuk.
Lymphoma. 38 (2000); 71-81.
[
8] RA Marco, Díaz‐ Montero CM, Wygant JN, Kleinerman ES, and McIntyre BW. α4 integrin
increases anoikis of human osteosarcoma cells. J. Cell. Biochem. 88 (2003); 1038-47.
[
9] T Michigami, Shimizu N, Williams PJ, Niewolna M, Dallas SL, Mundy GR, et al. Cell–cell
contact between marrow stromal cells and myeloma cells via VCAM-1 and α4β1-integrin
enhances production of osteoclast-stimulating activity. Blood. 96 (2000); 1953-60.
[
10] K Hatano, Kikuchi J, Takatoku M, Shimizu R, Wada T, Ueda M, et al. Bortezomib
overcomes cell adhesion-mediated drug resistance through downregulation of VLA-4 expression
in multiple myeloma. Oncogene. 28 (2009); 231-42.
[
11] DL Olson, Burkly LC, Leone DR, Dolinski BM, and Lobb RR. Anti-α4 integrin monoclonal
antibody inhibits multiple myeloma growth in a murine model. Mol. Cancer Ther. 4 (2005); 91-
9
.
[
12] F Sanz-rodríguez and Teixidó J. VLA-4-dependent myeloma cell adhesion. Leuk.
Lymphoma. 41 (2001); 239-45.
[
(
13] P Drillenburg and Pals ST. Cell adhesion receptors in lymphoma dissemination. Blood. 95
2000); 1900-10.
[
14] L Baldini, Cro L, Calori R, Nobili L, Silvestris I, and Maiolo A. Differential expression of
very late activation antigen-3 (VLA-3)/VLA-4 in B-cell non-Hodgkin lymphoma and B-cell
chronic lymphocytic leukemia. Blood. 79 (1992); 2688-93.

ACCEPTED MANUSCRIPT
[
15] WG Finn, Singleton TP, Schnitzer B, Ross CW, and Stoolman LM. Adhesion molecule
expression in CD5-negative/CD10-negative chronic B-cell leukemias: Comparison with non-
Hodgkin's lymphomas and CD5-positive B-cell chronic lymphocytic leukemia. Hum. Pathol. 32
(2001); 66-73.
[
16] T Matsunaga, Takemoto N, Sato T, Takimoto R, Tanaka I, Fujimi A, et al. Interaction
between leukemic-cell VLA-4 and stromal fibronectin is a decisive factor for minimal residual
disease of acute myelogenous leukemia. Nat. Med. 9 (2003); 1158-65.
[
17] A Vincent, Cawley J, and Burthem J. Integrin function in chronic lymphocytic leukemia.
Blood. 87 (1996); 4780-8.
[
18] IR Hart, Birch M, and Marshall JF. Cell adhesion receptor expression during melanoma
progression and metastasis. Cancer and Metastasis Reviews. 10 (1991); 115-28.
[
19] M Klemke, Weschenfelder T, Konstandin MH, and Samstag Y. High affinity interaction of
integrin α4β1 (VLA‐ 4) and vascular cell adhesion molecule 1 (VCAM‐ 1) enhances migration
of human melanoma cells across activated endothelial cell layers. J. Cell. Physiol. 212 (2007);
3
68-74.
20] S Kuphal, Bauer R, and Bosserhoff A-K. Integrin signaling in malignant melanoma. Cancer
and Metastasis Reviews. 24 (2005); 195-222.
21] D Schadendorf, Gawlik C, Haney U, Ostmeier H, Suter L, and Czarnetzki BM. Tumour
[
[
progression and metastatic behaviour in vivo correlates with integrin expression on melanocytic
tumours. J. Pathol. 170 (1993); 429-34.
[
22] D Schadendorf, Heidel J, Gawlik C, Suter L, and Czarnetzki BM. Association With Clinical
Outcome of Expression of VLA-4 in Primary Cutaneous Malignant Melanoma as Well as P-
selection and E-selectin on Intratumoral Vessels. J. Natl. Cancer Inst. 87 (1995); 366-71.
[
1
23] B Garmy-Susini, Jin H, Zhu Y, Sung R-J, Hwang R, and Varner J. Integrin α4β1–VCAM-
–mediated adhesion between endothelial and mural cells is required for blood vessel
maturation. J. Clin. Invest. 115 (2005); 1542.
[
24] H Jin, Su J, Garmy-Susini B, Kleeman J, and Varner J. Integrin α4β1 promotes monocyte
trafficking and angiogenesis in tumors. Cancer Res. 66 (2006); 2146-52.
[
25] B Garmy-Susini, Avraamides CJ, Schmid MC, Foubert P, Ellies LG, Barnes L, et al.
Integrin α4β1 signaling is required for lymphangiogenesis and tumor metastasis. Cancer Res. 70
(2010); 3042-51.
[
26] L Peng, Liu RW, Marik J, Wang XB, Takada Y, and Lam KS. Combinatorial chemistry
identifies high-affinity peptidomimetics against alpha(4)beta(1) integrin for in vivo tumor
imaging. Nat. Chem. Biol. 2 (2006); 381-9.
[
27] L Peng, Liu R, Andrei M, Xiao W, and Lam KS. In vivo optical imaging of human
lymphoma xenograft using a library-derived peptidomimetic against α4β1 integrin. Mol. Cancer
Ther. 7 (2008); 432-7.
[
28] S DeNardo, Sutcliffe J, Anderson C, Natarajan A, Kukis D, Cherry S, et al. (111)In-DOTA-
and (64)Cu-CB-TE2A-LLP2A targeting alpha 4 beta 1 integrin: Development of imaging
directed (67)Cu therapy of lymphoid malignancies. Cancer Biother. Radiopharm. 23 (2008);
5
15-.

ACCEPTED MANUSCRIPT
[
29] SJ DeNardo, Liu RW, Albrecht H, Natarajan A, Sutcliffe JL, Anderson C, et al. (111)In-
LLP2A-DOTA Polyethylene Glycol-Targeting alpha 4 beta 1 Integrin: Comparative
Pharmacokinetics for Imaging and Therapy of Lymphoid Malignancies. J. Nucl. Med. 50 (2009);
6
25-34.
[
30] AL Zwingenberger, Kent MS, Liu RW, Kukis DL, Wisner ER, DeNardo SJ, et al. In-Vivo
Biodistribution and Safety of Tc-99m-LLP2A-HYNIC in Canine Non-Hodgkin Lymphoma.
PLoS One. 7 (2012);
[
31] M Shokeen, Zheleznyak A, Wilson JM, Jiang MJ, Liu RW, Ferdani R, et al. Molecular
Imaging of Very Late Antigen-4 (alpha(4)beta(1) Integrin) in the Premetastatic Niche. J. Nucl.
Med. 53 (2012); 779-86.
[
32] M Jiang, Ferdani R, Shokeen M, and Anderson CJ. Comparison of two cross-bridged
macrocyclic chelators for the evaluation of (64)Cu-labeled-LLP2A, a peptidomimetic ligand
targeting VLA-4-positive tumors. Nucl. Med. Biol. 40 (2013);
[
33] D Soodgupta, Hurchla MA, Jiang M, Zheleznyak A, Weilbaecher KN, Anderson CJ, et al.
Very Late Antigen-4 (alpha(4)beta(1) Integrin) Targeted PET Imaging of Multiple Myeloma.
PLoS One. 8 (2013);
[
34] W Beaino and Anderson CJ. PET Imaging of Very Late Antigen-4 in Melanoma:
Comparison of 68Ga-and 64Cu-Labeled NODAGA and CB-TE1A1P-LLP2A Conjugates. J.
Nucl. Med. 55 (2014); 1856-63.
[
35] W Beaino, Nedrow JR, and Anderson CJ. Evaluation of Ga-68- and Lu-177-DOTA-
PEG(4)-LLP2A for VLA-4-Targeted PET Imaging and Treatment of Metastatic Melanoma. Mol.
Pharm. 12 (2015); 1929-38.
[
36] D Soodgupta, Zhou H, Beaino W, Lu L, Rettig M, Snee M, et al. Ex Vivo and In Vivo
Evaluation of Overexpressed VLA-4 in Multiple Myeloma Using LLP2A Imaging Agents. J.
Nucl. Med. 57 (2016); 640-5.
[
37] GJ Kemerink, Visser MGW, Franssen R, Beijer E, Zamburlini M, Halders S, et al. Effect of
the positron range of F-18, Ga-68 and I-124 on PET/CT in lung-equivalent materials. Eur. J.
Nucl. Med. Mol. Imag. 38 (2011); 940-8.
[
38] A Sanchez-Crespo. Comparison of Gallium-68 and Fluorine-18 imaging characteristics in
positron emission tomography. Appl. Radiat. Isot. 76 (2013); 55-62.
[
39] WJ McBride, Sharkey RM, Karacay H, D'Souza CA, Rossi EA, Laverman P, et al. A Novel
Method of F-18 Radiolabeling for PET. J. Nucl. Med. 50 (2009); 991-8.
[
40] R Schirrmacher, Bradtmoller G, Schirrmacher E, Thews O, Tillmanns J, Siessmeier T, et al.
F-18-labeling of peptides by means of an organosilicon-based fluoride acceptor. Angew. Chem.-
Int. Edit. 45 (2006); 6047-50.
[
41] P Rosa-Neto, Waengler B, Iovkova L, Boening G, Reader A, Jurkschat K, et al. F-18 SiFA-
isothiocyanate: A New Highly Effective Radioactive Labeling Agent for Lysine-Containing
Proteins. ChemBioChem. 10 (2009); 1321-4.
[
42] J Marik and Sutcliffe JL. Click for PET: rapid preparation of F-18 fluoropeptides using Cu-I
catalyzed 1,3-dipolar cycloaddition. Tetrahedron Lett. 47 (2006); 6681-4.

ACCEPTED MANUSCRIPT
[
43] DM Perrin. F-18 -Organotrifluoroborates as Radioprosthetic Groups for PET Imaging: From
Design Principles to Preclinical Applications. Acc. Chem. Res. 49 (2016); 1333-43.
44] ZB Liu, Lin KS, Benard F, Pourghiasian M, Kiesewetter DO, Perrin DM, et al. One-step F-
8 labeling of biomolecules using organotrifluoroborates. Nat. Protoc. 10 (2015); 1423-32.
45] Z Liu, Hundal-Jabal N, Wong M, Yapp D, Lin KS, Benard F, et al. A new F-18-
[
1
[
heteroaryltrifluoroborate radio-prosthetic with greatly enhanced stability that is labelled by F-18-
F-19-isotope exchange in good yield at high specific activity. Medchemcomm. 5 (2014); 171-9.
[
1
46] Z Liu, Li Y, Lozada J, Wong MQ, Greene J, Lin K-S, et al. Kit-like 18F-labeling of RGD-
9F-Arytrifluroborate in high yield and at extraordinarily high specific activity with preliminary
in vivo tumor imaging. Nucl. Med. Biol. 40 (2013); 841-9.
[
47] Y Li, Ting R, Harwig CW, Keller UAD, Bellac CL, Lange PF, et al. Towards kit-like F-18-
labeling of marimastat, a noncovalent inhibitor drug for in vivo PET imaging cancer associated
matrix metalloproteases. Medchemcomm. 2 (2011); 942-9.
[
48] R Ting, Harwig CW, auf dem Keller U, McCormick S, Austin P, Overall CM, et al.
Towards [18F]-Labeled Aryltrifluoroborate Radiotracers - In Vivo PET Imaging of Stable
Aryltrifluoroborate Clearance in Mice. J. Am. Chem. Soc. 130 (2008); 12045-55.
[
49] ZB Liu, Chao D, Li Y, Ting R, Oh J, and Perrin DM. From Minutes to Years: Predicting
Organotrifluoroborate Solvolysis Rates. Chem.-Eur. J. 21 (2015); 3924-8.
[
50] ZB Liu, Pourghiasian M, Radtke MA, Lau J, Pan JH, Dias GM, et al. An
Organotrifluoroborate for Broadly Applicable One-Step F-18-Labeling. Angew. Chem.-Int. Edit.
3 (2014); 11876-80.
5
[
51] Z Liu, Pourghiasian M, Benard F, Pan J, Lin K-S, and Perrin DM. Preclinical Evaluation of
a High-Affinity 18F-Trifluoroborate Octreotate Derivative for Somatostatin Receptor Imaging. J.
Nucl. Med. 55 (2014); 1499-505.
[
52] ZB Liu, Amouroux G, Zhang ZX, Pan JH, Hundal-Jabal N, Colpo N, et al. F-18-
Trifluoroborate Derivatives of Des-Arg(10) Kallidin for Imaging Bradykinin B1 Receptor
Expression with Positron Emission Tomography. Mol. Pharm. 12 (2015); 974-82.
[
53] M Pourghiasian, Liu ZB, Pan JH, Zhang ZX, Colpo N, Lin KS, et al. F-18-AmBF3-MJ9: A
novel radiofluorinated bombesin derivative for prostate cancer imaging. Biorg. Med. Chem. 23
(2015); 1500-6.
[
54] Y Li, Asadi A, and Perrin DM. Hydrolytic stability of nitrogenous-
heteroaryltrifluoroborates under aqueous conditions at near neutral pH. J. Fluorine Chem. 130
(2009); 377-82.
[
55] D Walker, Li Y, Roxin A, Schaffer P, Adam MJ, and Perrin DM. Facile synthesis and F-18-
radiolabeling of alpha(4)beta(1)-specific LLP2A-aryltrifluoroborate peptidomimetic conjugates.
Bioorg. Med. Chem. Lett. 26 (2016); 5126-31.
[
56] I Dude, Zhang Z, Rousseau J, Hundal-Jabal N, Colpo N, Merkens H, et al. Evaluation of
agonist and antagonist radioligands for somatostatin receptor imaging of breast cancer using
positron emission tomography. EJNMMI Radiopharmacy and Chemistry. 2 (2017); 4.