An improved synthesis of n-(4-[18 f]fluorobenzoyl)-interleukin-2 for the preclinical pet imaging of tumour-infiltrating t-cells in ct26 and mc38 colon cancer models

Article abstract of DOI:10.3390/molecules26061728

Positron emission tomography (PET) imaging of activated T-cells with N-(4-[¹⁸ F]fluorobenzoyl)-interleukin-2 ([¹⁸ F]FB-IL-2) may be a promising tool for patient management to aid in the assessment of clinical responses to immune ther

Full text of DOI:10.3390/molecules26061728

molecules
Article
An Improved Synthesis of
18
N-(4-[ F]Fluorobenzoyl)-Interleukin-2 for the Preclinical PET
Imaging of Tumour-Infiltrating T-cells in CT26 and MC38 Colon
Cancer Models
1
,
1
1
1
1
Shivashankar Khanapur * , Fui Fong Yong , Siddesh V. Hartimath , Lingfan Jiang , Boominathan Ramasamy ,
1
2
1
1,3,
Peter Cheng , Pradeep Narayanaswamy , Julian L. Goggi
and Edward George Robins
*
1
Singapore Bioimaging Consortium, Agency for Science, Technology and Research (A* STAR), 11 Biopolis Way,
01-02 Helios, Singapore 138667, Singapore; Yong_Fui_Fong@sbic.a-star.edu.sg (F.F.Y.);
#
S_Hartimath@sbic.a-star.edu.sg (S.V.H.); Jiang_Lingfan@gis.a-star.edu.sg (L.J.);
Boominathan_Ramasamy@sbic.a-star.edu.sg (B.R.); Peter_Cheng@sbic.a-star.edu.sg (P.C.);
Julian_Goggi@sbic.a-star.edu.sg (J.L.G.)
Sciex, R&D, Blk 33, #04-06 Marsiling Industrial Estate Road 3 Woodlands Central Industrial Estate,
Singapore 739256, Singapore; pradeep.naraya@sciex.com
Clinical Imaging Research Centre, 14 Medical Drive, #B01-01 Centre for Translational Medicine,
Yong Loo Lin School of Medicine, National University of Singapore, Singapore 117599, Singapore
Correspondence: shivashankar@sbic.a-star.edu.sg (S.K.); Edward_Robins@sbic.a-star.edu.sg (E.G.R.);
Tel.: +65-6478-7053 (S.K.); +65-6478-7001 (E.G.R.)
2
3
*
ꢀ
ꢁꢂꢀꢃꢄꢅꢆꢇ
ꢀ
ꢁꢂꢃꢄꢅꢆ
Citation: Khanapur, S.; Yong, F.F.;
Hartimath, S.V.; Jiang, L.; Ramasamy,
B.; Cheng, P.; Narayanaswamy, P.;
Goggi, J.L.; Robins, E.G. An
18
Abstract: Positron emission tomography (PET) imaging of activated T-cells with N-(4-[ F]fluorobenzoyl)-
18
interleukin-2 ([ F]FB-IL-2) may be a promising tool for patient management to aid in the assessment of
clinical responses to immune therapeutics. Unfortunately, existing radiosynthetic methods are very low
yielding due to complex and time-consuming chemical processes. Herein, we report an improved
method for the synthesis of [ F]FB-IL-2, which reduces synthesis time and improves radiochem-
ical yield. With this optimized approach, [ F]FB-IL-2 was prepared with a non-decay-corrected
Improved Synthesis of N-(4-
1
8
[18
F]Fluorobenzoyl)-Interleukin-2 for
1
8
the Preclinical PET Imaging of
1
8
radiochemical yield of 3.8
±
0.7% from [ F]fluoride, 3.8 times higher than previously reported
Tumour-Infiltrating T-cells in CT26
and MC38 Colon Cancer Models.
Molecules 2021, 26, 1728. https://
doi.org/10.3390/molecules26061728
methods. In vitro experiments showed that the radiotracer was stable with good radiochemical
purity (>95%), confirmed its identity and showed preferential binding to activated mouse peripheral
blood mononuclear cells. Dynamic PET imaging and ex vivo biodistribution studies in naïve Balb/c
18
mice showed organ distribution and kinetics comparable to earlier published data on [ F]FB-IL-2.
1
8
Academic Editor: Anne Roivainen
Significant improvements in the radiochemical manufacture of [ F]FB-IL-2 facilitates access to this
promising PET imaging radiopharmaceutical, which may, in turn, provide useful insights into differ-
ent tumour phenotypes and a greater understanding of the cellular nature and differential immune
microenvironments that are critical to understand and develop new treatments for cancers.
Received: 14 February 2021
Accepted: 16 March 2021
Published: 19 March 2021
18
Keywords: positron emission tomography (PET) imaging; Interleukin-2 (IL-2); T-cells; [ F]FB-IL-2;
[18
F]SFB; Scintomics GRP module; protein conjugation reaction; murine colon adenocarcinoma
MC38 and CT26) syngeneic models
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
™
(
1. Introduction
Cancer immunotherapy (CIT), the fourth pillar of cancer therapy, utilizes the body’s
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
own immune system to fight cancer. Recently, CIT has garnered significant interest with
the approval of several immune checkpoint inhibitors (Programmed cell death protein 1,
Programmed death-ligand 1 and cytotoxic T-lymphocyte-associated protein 4), cancer vac-
cines and adjuvants (Dendritic cell vaccines and CpG, tumour protein), cytokine therapies
(
Interferon alfa , IL-2) and adaptive cell-based immunotherapies (Chimeric antigen receptor
T-cells, Tumour-infiltrating T-cells, Natural killer cells) [ ]. Despite the success of CIT across
a broad range of human cancers, only a minority of patients show a durable response
1
4
.0/).
Molecules 2021, 26, 1728. https://doi.org/10.3390/molecules26061728
https://www.mdpi.com/journal/molecules

Molecules 2021, 26, 1728
2 of 12
to these therapies and some patients experience severe immune-related adverse effects.
Therefore, there is a tremendous need to stratify patients and optimize immunotherapy
combinations for better management of cancer patient care [2]. Immune responses are com-
monly assessed by measuring lymphocytes in whole blood and/or by biopsies of tumour,
spleen and lymph nodes. These methods are invasive and do not provide spatiotemporal
information about the dynamic immune responses to these therapies in heterogeneous
tumour microenvironments. In this context, non-invasive positron emission tomography
(
PET) imaging of immune cells, especially T cells, might aid in patient selection, response
prediction and treatment evaluation following immunotherapeutic intervention [1].
T cells are crucial regulators of adaptive immune responses in both immune-driven
diseases and tumours [3]. Upon activation by antigen-presenting cells, T-cells (CD4+ and
CD8+ T-cells) secrete the protein interleukin-2 (IL-2) to regulate immune responses. IL-2, a
small monomeric glycoprotein (15.3 kDa) lymphokine, binds with high affinity to its cell
surface interleukin-2 receptors (IL-2Rs), which consists of three receptor subunits:
(CD122) and (CD132). Over-expression of IL-2Rs is found mainly on the activated T
cells [ ]. Radiolabelling of human recombinant IL-2 with short-lived positron-emitting
α (CD25),
β
γ
4,5
18
−
radionuclides such as fluorine-18 ( F ) have already been shown to provide a tool for
non-invasive imaging of activated tumour infiltrating lymphocytes (TILs) in cancer.
Several radiopharmaceuticals for PET imaging of IL-2Rs on activated T-cells have
already been developed [
5
–
8
]. Perhaps the most well-characterized of these is N-(4-
18
18
[
F]fluorobenzoyl)-interleukin-2 ([ F]FB-IL-2), which was developed as a tool for the
in vivo identification and quantification of CD25+ T-cells and is currently being evaluated
in clinical trials (NCT02922283 and NCT03304223) [ 10]. However, due to a complex
production process and low radiochemical yields (RCYs) from protein conjugation reac-
5–7,9,
18
tions, the [ F]FB-IL-2 synthesis requires a high initial amount of radioactivity to provide
18
a suitable quantity of formulated [ F]FB-IL-2 radiopharmaceutical for either clinical or
preclinical use [6,7].
Therefore, we sought to optimize the synthetic steps to improve RCY and reduce
18
the synthesis time described in existing literature radiosynthesis methods for [ F]FB-IL-2.
18
In this article, we describe an improved semi-automated synthesis of [ F]FB-IL-2 via N-
18
18
succinimidyl 4-[ F]fluorobenzoate ([ F]SFB) and compare it with previous manufacturing
procedures. Furthermore, we characterized the synthesized radiotracer in vitro and in vivo
18
and assessed the efficacy of [ F]FB-IL-2 as an imaging biomarker in preclinical PET imaging
of CD25+ TILs in syngeneic murine colon cancer models.
2. Results
2
.1. Radiochemistry
1
8
We have adapted the four-step, two-pot procedure to synthesize [ F]FB-IL-2 from
those previously reported with several modifications as shown in Scheme 1 [ 10 11].
In this work, crude [ F]SFB was synthesised via three-step reactions (Scheme 1) using
a Scintomics GRP module. Manual purification of crude reaction mixture by semi-
preparative reverse phase-high performance liquid chromatography (RP-HPLC) followed
5,7, ,
18
™
18
by solid-phase extraction (SPE) procedure delivered pure [ F]SFB in a final volume of 1
mL diethyl ether which was evaporated to dryness for the downstream application. Semi-
18
automated synthesis produced dried [ F]SFB with a good non-decay corrected RCY of
2.7 3.5% (3.6 0.9 GBq) from aqueous fluoride activity (15.9 n
2
±
±
±
2.8 GBq) in 85 ± 7 mi
18
(
(
n = 15). The radiochemical purity (RCP) of the [ F]SFB after SPE process was > 95%
Figure S2) and molar activities were >100 GBq/µmol (Supplementary Materials).

Molecules 2021, 26, 1728
3 of 12
1
8
Scheme 1. Semi-automated synthesis of [ F]FB-IL-2.
[18
F]SFB was reconstituted in DMSO and incubated with proleukin protein in 0.1 M
◦
18
sodium borate buffer for 10 min at 50 C. The [ F]fluoroacylation reaction produced
1
8
8
[
[
F]FB-IL-2 with a non-decay corrected RCY of 3.8
F]fluoride activity (15.9 2.8 GBq) in 115 5 min (n = 12). Our manual radiosynthesis
procedure mainly differs from previously reported protein conjugation methods in the
sequence of [ F]SFB addition to protein-buffer mixture and the use of co-solvent dimethyl
sulfoxide (DMSO) in a reduced reaction volume [ ]. The RCP of the formulated fraction
was >95%, as determined by RP-HPLC (Figure 1) and TCA precipitation assay. Molar
activities of the formulated fractions were 114 94 GBq/ mol (n = 6) at the end of
the synthesis (Supplementary Materials). The identities of the [ F]SFB and [ F]FB-IL-2
were confirmed by spiking with reference standard SFB and recombinant IL-2 protein,
respectively, in RP-HPLC.
±
0.7% (0.6
±
0.1 GBq) from aqueous
1
±
±
18
5–7
±
µ
18
18
1
8
2
.2. Chemical Characterisation of [ F]FB-IL-2
18
The shelf-life of formulated [ F]FB-IL-2 solution was assessed over 5.5 h at room
temperature, during which no change in RCP was observed. Liquid chromatography–
mass spectrometry (LC-MS) was performed to confirm the molecular mass of the decayed
labelled protein product. It was found that mass of 15,328 (unlabelled IL-2; theoretical
expected mass 15,330 Da), 15,461 (monoconjugation), 15,595 (diconjugation), 15,729 (tri-
conjugation), 15,864 (tetraconjugation)
that each labelled IL-2 was coupled with 1–4 fluorobenzoyl (FB) residues with the most
abundant coupling being between 2–3 FB groups (Figure 2).
±
2 Da for labelled FB-IL-2. LC-MS data suggest
1
8
2
.3. In Vitro Biological Characterisation of [ F]FB-IL-2
18
In order to assess the specificity of [ F]FB-IL-2 binding to IL-2Rs, in vitro exper-
iments were performed using mouse peripheral blood mononuclear cells (mPBMCs).
Phytohemoagglutinin (PHA) activated and non-activated mPBMCs were incubated with
18
◦
18
[
F]FB-IL-2 at 37 C for 30 min. A significant difference in [ F]FB-IL-2 binding between
non-activated (14.78 0.65%) and activated mPBMCs (98.00 2.00% p < 0.0001) was
±
±
observed (Figure 3). In order to determine the specificity of binding, we pre-incubated
with excess recombinant IL-2 protein (20 ng/mL) to saturate the IL-2Rs, which led to a
1
8
threefold reduction in [ F]FB-IL-2 binding (32.12 ± 1.90%, p < 0.001).

Molecules 2021, 26, 1728
4 of 12
1
8
Figure 1. Radio- and UV chromatograms of the PD-10 purified formulated radiotracer, [ F]FB-IL-2.
2
.4. Animal Studies
18
In vivo PET imaging with [ F]FB-IL-2 was performed in naïve mice to study tracer
kinetics and tissue distribution. As shown in Figure 4A, uptake of the radiotracer could be
visualized in major excretory organs, including liver and kidneys. Time–activity curves
reveal a rapid uptake of the radiotracer into the liver, lungs, heart and kidney followed
by exponential washout. Low radiotracer uptake was observed in the brain as the radi-
olabelled protein does not penetrate the blood–brain barrier (Figure 4B). Ex vivo biodis-
18
tribution, Figure 4C, showed a similar overall profile, the [ F]FB-IL-2 radiotracer was
excreted mainly via the kidneys into urinary bladder, similar to the native IL-2 and human
recombinant IL-2 proteins and uptake in bone was observed to be low, indicating the lack
of defluorination in vivo.

Molecules 2021, 26, 1728
5 of 12
Figure 2. The electrospray ionization-mass spectrometry (ESI-MS) profile of unlabeled IL-2 and FB-IL-2. (
ion mode ESI-MS spectra of IL-2 unlabeled, showing different mass/charge (m/z) ratios from +8 to +17 charges; the
deconvoluted mass of unlabeled IL-2 was 15,328 2 Da and that matched the theoretical expected mass 15,330 Da.
) Positive ion mode ESI-MS spectra of decayed labelled FB-IL-2 suggest a different degree of conjugation; magnified
A) The positive
±
(
B
charge state (+9 charge) region suggests IL-2 labelling with monoconjugation, diconjugation, triconjugation, tetraconjugation
and pentaconjugation. Based on the relative intensity ratio, it indicates between 2–3 FB groups per IL-2 molecule are the
most abundant coupling. Note: The labels on the mass spectra peaks +n, indicate the number of protons attached to
the protein.
1
8
Figure 3. In vitro binding assessment of [ F]FB-IL-2 in mouse PBMCs. Selective binding of
18
[
F]FB-IL-2 to PHA activated mPBMCs was observed when compared to non-activated mPBMCs
14.8 ± 0.65%, **** p < 0.0001). The binding was shown to be specific by a significant reduction in the
binding of [18F]FB-IL-2 when pre-treated with excess non-radioactive recombinant IL-2 (32.12
1.90,
** p < 0.001), data shown as % bound ± SD.
(
±
*
18
To investigate the utility of [ F]FB-IL-2 in oncology imaging, CT26 and MC38 tumour-
bearing mice were imaged after 60 min of tracer injection demonstrating homogenous

Molecules 2021, 26, 1728
6 of 12
1
8
tumour uptake in viable tumour tissue (Figure 5A). Significantly higher uptake of [ F]FB-
IL-2 was observed in MC38 tumours compared to CT26 tumours (1.65
0.24 vs. 1.21 ± 0.03
ID/g, p < 0.05, Figure 5B). Fluorescence activated cell sorting (FACS) analysis was
performed on these excised colon tumours and showed that these MC38 tumours had
greater infiltration of CD8+ CD25+ cells compared to CT26 tumours (16.12 4.14 vs.
.14 ± 1.09, p < 0.002, Figure 5C), whereas no significant difference was observed for CD4+
CD25+ cells (52.82 5.31 vs. 50. 82 4.62, p = 0.566, Figure 5D). These data suggest that
FB-IL2 may be able to differentiate tumours based on their levels of CD25+ TILs.
±
,
%
±
5
±
±
18
Figure 4. In vivo PET imaging of [ F]FB-IL-2 in naïve Balb/c mice. (
summed frames from 40–60 min post-injection. ( ) Dynamic PET time–activity curves showing initial uptake followed by
washout of [ F]FB-IL-2 tracer in most of the major organs. ( ) Ex vivo biodistribution data (70 min post-injection) showing
F]FB-IL-2 tracer accumulation in different organs confirming PET-based biodistribution.
A) Whole-body PET-CT fused image represents the
B
1
8
C
1
8
[
1
8
Figure 5. In vivo PET imaging of [ F]FB-IL-2 in MC38 and CT26 tumour-bearing mice. (
A
) PET-CT fused image of mice
1
8
18
with MC38 and CT26 tumours showing the distribution of [ F]FB-IL-2. (
in MC38 tumours shows significantly higher uptake than in CT26 tumours (1.65
SEM % ID/g, * p < 0.05). ( ) FACS analysis showing CD8+CD25+ cells as a % of CD3+ cells in MC38 and CT26 tumours.
CD8+CD25+ cells were significantly higher in MC38 when compared to CT26 (16.12 4.14 vs. 5.14 1.09, *** p < 0.002,
data shown as mean SD as a % of CD3+ cells). ( ) FACS analysis showing CD4+CD25+ as a % of CD3+ cells in MC38
B
) Quantitative estimation of [ F]FB-IL-2 uptake
±
0.24 vs. 1.21
±
0.03, data shown as mean
±
C
±
±
±
D
and CT26 tumours. No significant differences in CD4+CD25+ cells were observed between MC38 and CT26 (data shown as
mean ± SD as a % of CD3+ cells).

Molecules 2021, 26, 1728
7 of 12
3. Discussion
Here we describe an improved method of synthesis for a promising radiopharma-
1
8
18
ceutical, [ F]FB-IL-2, to image-activated T-cells. Semi-automated [ F]SFB synthesis was
performed using a Scintomics GRP module. Crude product obtained from Scintomics
was then purified using semi-preparative RP-HPLC, followed by a SPE cartridge to obtain
™
18
a chemically pure intermediate radiolabelling synthon, [ F]SFB. Purification by RP-HPLC
was adopted as it was reported earlier that inefficient [ F]SFB purification only by SPE
cartridge resulted in low protein-conjugation reaction yields and frequent failures [7].
Dropwise addition of [ F]SFB was chosen to mitigate the effects of hydrolysis. In
our experience, [ F]SFB in DMSO showed better hydrolytic stability, higher reactivity
and better solubility than polar protic solvent, ethanol. In addition, the minimal reaction
18
18
18
volume (300
minute (rpm) provided higher RCYs than existing procedures [6,7].
In contrast to RP-SPE purification [ ], the SEC column method of purification provided
µL) with the use of 1 mL V-vial and stirring the contents at 100 rotations per
7
the labelled protein product with high recovery (90–95%) and without the need for an
additional reformulation step. The use of SEC-PD10 column purification and elution buffer
0
.05% sodium dodecyl sulphate (SDS) in phosphate-buffered saline (PBS) allowed us to
streamline the purification and reformulation of the labelled product for injection into
18
animals. [ F]FB-IL-2 was prepared with 3.8 times higher RCYs and 30 min less reaction
time than published articles procedures (1.0 ± 0.4% in 150 min) [6,7].
18
No change on RCP of [ F]FB-IL-2 formulated product due to chemically or radiochem-
ically induced decay was found at least for 5.5 h at room temperature, as determined by
RP-HPLC in a shelf-life experiment. Additionally, LC-MS data suggest that after labeling,
each labelled IL-2 was coupled with between 1 and 4 FB residues. The degree of labelling
(
DOL) can be determined on the basis of radioactivity incorporated in the protein (GBq),
18
the molar activity of [ F]SFB (GBq/
µmol), and the concentration of protein recovered after
reaction (mg/mL) [ ]. With this calculation, we determined that on average, 1.9–3.9 FB
5
residues to each labelled protein molecule. The calculated value range is in agreement
with our LC-MS data. In vitro radioligand binding assay showed preferential binding
to activated mPBMCs and binding was selective, as determined by receptor saturation
binding experiment.
18
In vivo [ F]FB-IL-2 dynamic PET imaging and ex vivo biodistribution studies showed
that the formulated product has comparable tissue distribution and kinetics to previously
18
published data [5,6,8,9]. Significantly higher uptake of [ F]FB-IL-2 was observed in MC38
tumours compared to CT26 tumours (Figure 5B), potentially due to phenotypic differences
in the tumour types, as MC38 tumours are associated with a microsatellite instability-
high (MSI-high) phenotype compared to the MSI-low CT26 tumours. Typically, high MSI
phenotype tumours have high mutation rates due to DNA mismatch repair and increased
immunogenicity of neoantigens. As a result, MSI-high phenotype tumours tend to be more
immunologically active with higher numbers of CD25+ TILs.
4. Materials and Methods
4.1. General Information
The compounds 4-(ethoxycarbonyl)-N,N,N-trimethylbenzenaminium trifluoromethane-
sulfonate (FB precursor) and 2,5-dioxopyrrolidin-1-yl 4-fluorobenzoate (reference standard)
®
6
were procured from ABX GmbH, Radeberg, Germany. Proleukin (aldesleukin, 18 × 10 IU),
a recombinant interleukin-2 (desalanyl-1, serine-125 human interleukin-2), was acquired
from Novartis, Singapore. N,N,N ,N -Tetramethyl-O-(N-succinimidyl)uronium tetrafluorobo-
0
0
rate (TSTU, 97%), acetonitrile anhydrous (99.8%), potassium carbonate anhydrous (99.99%),
®
4
,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8]hexacosane (Kryptofix 222, 98%), dimethyl
sulfoxide anhydrous (
≥
99.9%), tetrapropylammonium hydroxide solution (TPAOH, 1.0 M in
99.0%) and sodium tetraborate decahydrate ( 99.5%)
H O), sodium dodecyl sulfate (SDS,
≥
≥
2
were procured from Sigma-Aldrich Pte Ltd., Singapore. All other reagents were procured
from Merck (Singapore), Tokyo Chemical Industry and Life Technologies Corporation (Sin-

Molecules 2021, 26, 1728
8 of 12
gapore). All commercially obtained reactants and reagents were used as received without
18
18
any further purification. [ F]SFB and [ F]FB-IL-2 productions were carried out in a closed
type-I glass flat bottom vial (15 mL) and closed Thermo Scientific
™ conical Reacti-vial™
®
(1‘mL), respectively. Sep-Pak light (46 mg) Accell™, QMA carbonate, Oasis HLB and light
cartridges were purchased from Waters Pacific Pte Ltd., Singapore. PD-10 desalting columns
were obtained from GE Healthcare Life Sciences, Singapore.
18
No-carrier-added (nca) aqueous [ F]fluoride ion was produced by the irradiation of
1
8
18
18
O-enriched water via the [ O(p,n) F] nuclear reaction using a GE PETtrace 860 cyclotron
GE healthcare, Singapore). Radiochemical purification and radio-UV HPLC analyses
were performed on a Knaur semi-preparative and UFLC Shimadzu radio-HPLC systems
Shimadzu, Singapore), respectively [ ]. Radioactivity measurements were made with a
CRC-55tPET dose calibrator (Capintec, Florham Park, NJ, USA).
(
(
2
18
18
4
.2. [ F]FB-IL-2 Radiosynthesis via Semi-Automated Synthesis of [ F]SFB
IL-2 was labelled with fluorine-18 by conjugation of the Proleukin protein (Novartis,
1
8
18
Singapore) with N-succinimidyl 4-[ F]-fluorobenzoate ([ F]SFB). Scintomics GRP4V cas-
1
8
sette module was used to perform synthesis of crude [ F]SFB. Details on the setup and
connections of the tubing are described in the supplementary materials.
18
At the beginning of the automated sequence, nca aqueous [ F]fluoride activity (ob-
tained from cyclotron; 2.4 mL) was transferred directly under suction into a V-vial placed
18
in the hot cell. Trapping of nca [ F]fluoride activity onto a pre-conditioned Sep-Pak Light
Waters Accell Plus QMA carbonate cartridge (manual precondition was done by passing
0 mL of deionised water) took place under vacuum along with the separation of enriched
™
1
1
8
water. The trapped [ F]fluoride anion was then eluted with a solution of 3 mg K CO
2
3
®
(21.9
µ
mol) and 15 mg Kryptofix 222 (39.9
µ
mol) in a mixture of 40
µ
L water and 960
µl
+
18
−
acetonitrile into an empty reactor vial. Azeotropic drying of the [K(K222)] [ F] complex
was performed under reduced pressure at 95 C under a gentle stream of nitrogen gas
◦
(
40 mL/min). After the first round of drying, anhydrous acetonitrile (0.5 mL) was added to
+
18
−
the [K(K222)] [ F] complex and the drying step was repeated. Following drying, 5 mg
FB precursor (14 µmol) in 1 mL DMSO was added to the dried complex [K(K222)] [ F] .
+
18
−
◦
The fluorination reaction was allowed to react for 8 min at 110 C to provide ethyl-
18
4
-[ F]fluorobenozoate. It was then deprotected using 20
.5 mL acetonitrile, providing the corresponding tetrapropylammonium salt. Subsequent
azeotropic drying and then a coupling reaction with 25 mg TSTU (83.1 mol) in 1.5 mL
acetonitrile yielded crude [ F]SFB (Scheme 1). The purification of [ F]SFB was achieved
by preparative HPLC followed by the SPE procedure as previously described [ ]. After
reconstitution with DMSO, analytical chromatography to determine RCP and molar activity
was carried out using a Synergi m Fusion-RP 80 Å, LC Column 250 4.6 mm at
µL TPAOH (1.0 M in H O) in
2
1
µ
18
18
2
™
4
µ
×
a flow rate of 1.5 mL/min. Gradient elution was carried out using a mixture of 0.1%
aqueous trifluoroacetic acid (solvent A) and acetonitrile (solvent B). The following gradient
elution profile was used: 0.01–0.30 min 10% B, 0.30–5.00 min 95% B, 5.00–8.00 min 95% B,
1
8
8
.00–12.00 min 10% B,
λ
= 254 nm. The retention time of [ F]SFB was between 7.0–7.1 min.
18
Dried [ F]SFB was reconstituted with 100 l DMSO and added dropwise into pH 8.5
µ
adjusted mixture of 100
µ
L aqueous IL-2 solution (Proleukin; 2 mg/mL) and 100
µ
L 0.1 M
◦
sodium borate buffer in 1 mL V-vial at 50 C for 10 min while stirring the reaction mixture
at 100 rpm (Scheme 1). The crude reaction mixture was then purified by SEC using PD-10
column, which was pre-conditioned with 30 mL elution buffer (0.05% SDS in PBS). The
18
pure labelled [ F]FB-IL-2 protein was eluted in 4 fractions of 0.5 mL each.
18
The RCP, radiochemical identity and molar activity of the [ F]FB-IL-2 was determined
using analytical radio-HPLC. The stationary phase was a Phenomenex Aeris Widepore
C4, 3.6
µ
m, 150 mm
×
2.1 mm column. Water with 0.1% trifluoroacetic acid (solvent A)
and acetonitrile with 0.1% trifluoroacetic acid (solvent B) were used as a mobile phase.
The following gradient elution profile was used: 0.2 min 5% B, 0.2–5 min 95% B, 5–6
min 95% B, 6–14 min 5% B, flow rate of 0.9 mL/min,
λ = 280 nm. The retention time of

Molecules 2021, 26, 1728
9 of 12
1
8
8
[
[
F]FB-IL-2 was between 6.5–7.1 min (Figure 1). The RCP and radiolabelling efficiency of
1
18
F]FB-IL-2 was determined using TCA precipitation assay. In brief, 5
µL of [ F]FB-IL-2
was added to 1 mL of ice-cold PBS and 5
protein was precipitated by adding 1 mL of 20% ice-cold TCA and then centrifuged at
000 rpm for 5 min. A pellet was formed at the bottom and 1 mL of clear supernatant was
µL of 20% human serum albumin (HSA). The
4
transferred to a separate tube. The radioactivity in each fraction was counted on a gamma
counter (Wizard 2470, PerkinElmer, Singapore). The percentage of labelling efficiency was
determined using the formula: percentage of protein-bound activity (%) = [(activity in
pellet − supernatant)/(activity in pellet + supernatant) × 100].
1
8
4
.3. Chemical Characterization of [ F]FB-IL-2
18
The stability of the formulated [ F]FB-IL-2 solution was assessed longitudinally at room
temperature using analytical RP-HPLC (Aeris Widepore C4, 3.6
µm, 150 mm × 2.1 mm). A
gradient elution was carried out using a mixture of 0.1% aqueous trifluoroacetic acid (solvent
A) and 0.1% trifluoroacetic acid in acetonitrile (solvent B), similar to analytical method
parameters. Samples were taken at 0, 1 h, 2 h, 3 h, and 5 h. Experiments were performed in
duplicate for each time point (n = 2). RCP was calculated by integration of the peak areas in
the chromatogram obtained.
A LC-MS 2020 (Shimadzu Asia Pacific Pte Ltd., Singapore) was used for identification
of the labelled protein. While in post-radioactive decay, a sample was subjected to LC-MS
analysis using an Agilent Poroshell 120 EC-18, 2.7
µ
m, 4.6
×
50 mm column at a flow rate
of 0.4 mL/min. Gradient elution was carried out using a mixture of 0.1% aqueous formic
acid (solvent A) and 0.1% formic acid in acetonitrile (solvent B). The following gradient
elution profile was used: 0.00–8.00 min 10% B, 8.00–12.00 min 95% B, 12.00–12.50 min
1
0% B, 12.50–15.00 min 10% B, λ = 254 nm. The mass spectrometer was operated in
electrospray positive ionization mode. The mass spectrometer settings were optimized as
follows: interface voltage, 4.5 kV; nebulizer gas flow, 1.5 L/min; drying gas flow, 15 L/min;
◦
◦
desolvation line (DL) temperature, 250 C; heat block temperature, 250 C. Other mass
spectrometer parameters were tuned automatically. Mass spectral data analysis was done
manually for identification of labelled protein.
1
8
4
.4. In Vitro Biological Characterisation of [ F]FB-IL-2
18
To assess [ F]FB-IL-2 binding and specificity for IL-2Rs, an in vitro binding experi-
ment was carried out using mPBMCs. mPBMCs were isolated from fresh mouse blood by
®
density gradient centrifugation using Lymphoprep separation (STEMCELL Technologies).
Isolated mPBMCs were kept in RPMI-1640 medium supplemented with 0.05% bovine
serum albumin and 1% Penicillin Streptomycin solution. Isolated cells were activated or
non-activated by incubation with or without 10
µ
g/mL of PHA (Sigma-Aldrich, Singapore),
◦
5
respectively, in a 5% CO atmosphere at 37 C for 48 h. Approximately 1 × 10 cells were
2
◦
incubated with 300 kBq tracer solution at 37 C for 30 min. After incubation, unbound
activity was removed by washing with 1 mL ice-cold PBS and cell-bound activity was mea-
sured using a gamma counter. A receptor saturation binding experiment was performed
◦
where activated mPBMCs were pre-incubated with IL-2 (20 ng/mL) at 4 C for 30 min,
prior to addition of the [ F]FB-IL-2.
18
4.5. Animal Studies
All animal procedures were carried out in accordance with the Institute Animal Care
and Use Committee (IACUC No. 181399) and conformed to the US National Institutes
of Health (NIH) guidelines and public law. BALB/c and C57BL/6 mice of 5–7 weeks
were purchased from local supplier (InVivos, Singapore) and housed in an individual
ventilated cage with a 12 h day/night cycle regimen, fed standard laboratory chow and tap
water ad libitum. The murine colon tumour cell lines, CT26 and MC38, were purchased
from ATCC and cultured in RPMI-1640 supplemented with 10% fetal bovine serum (FBS),
◦
100 U/mL penicillin and 100
µ
g/mL streptomycin, and maintained at 37 C in a humidified

Molecules 2021, 26, 1728
10 of 12
incubator (with 5% CO ). Mice were subcutaneously implanted with CT26 or MC38 cells
2
6
(1
×
10 ) prepared in a mixture of matrigel and PBS (1:1 ratio) into the right shoulder
of BALB/c or C57BL/6 mice, respectively. Tumour growth was monitored by caliper
measurement and the tumour volume was calculated by using the modified ellipsoid
1
2
formula (Length × width ).
2
18
To study the normal tissue distribution and kinetics of [ F]FB-IL-2, in vivo PET
imaging was performed in naïve Balb/c mice (n = 5) using an Inveon PET-CT scanner
(Siemens). Mice were anesthetized using a mixture of isoflurane and medical air (5%
induction, 2% maintenance) and kept on electronic heating pads during the scanning
1
8
period. Mice were injected with 15
±
2 MBq of [ F]FB-IL-2 via the lateral tail vein and
immediately a dynamic scan of 60 min was performed. During the scanning, animals
were monitored for body temperature and respiration rate using a Biovet physiological
monitoring system. After the scan, mice were euthanized and different organs were
harvested, weighed, and the tissue-bound radioactivity was measured.
Mice bearing CT26 and MC38 colon tumours were injected with 7
lateral tail vein (n = 5 per group). A static PET acquisition of 10 min was performed
after 40 min of post-tracer injection, followed by a CT scan (40 kV, 500 A; 4 4 binning,
00 m resolution) for co-registration. PET images were corrected for decay, scatter and
iteratively reconstructed to 11 frames (4 60, 1 120, 2 300, 4 600 s) and 2 frames
5 min) for dynamic and static PET scans, respectively. The radioactive uptake in
different organs, including the tumour, was estimated by drawing a volume of interest
VOI) delineated by CT. The VOIs were transferred from the CT template to the PET data
±
1 MBq via the
µ
×
2
µ
×
×
×
×
(2
×
(
and regional time–activity curves were generated. The image analysis of reconstructed
calibrated images was performed using Amide software (version 10.3 Sourceforge) and
data are expressed as a percentage of injected dose per gram (% ID/g) in the VOI.
After scanning, the mice were euthanized and tumours were harvested for flow
cytometry analysis as described previously [2]. In brief, tumours were mechanically
digested by incubating in 20 g/mL DNAse1 (Sigma-Aldrich, Singapore) and 200
µ
µg/mL
collagenase (Sigma-Aldrich, Singapore) in RPMI medium supplemented with 10% heat-
◦
inactivated FBS for 1 h at 37 C. A single-cell suspension was prepared by passing through
a 100
µm cell strainer and samples were then counted for cell viability using Trypan blue
(Sigma-Aldrich, Singapore). Later, cells were stained against CD3 (clone 500A2 BUV563;
BD Biosciences), CD4 (clone RM4–5 BV650; BD Biosciences), CD8 (clone 53-6.7 BV510;
BD Biosciences) and CD25 (clone PC61 BUV395; BD Biosciences). The FACS data was
acquired on a BD FACSymphony and the instrument settings, such as detector voltage
and fluorophore gain, were set up using murine spleen cell suspension. The flow data was
analysed using FlowJo V10.5 software (FlowJo LLC) and expressed as % of cell population.
4.6. Statistical Analysis
Data were analysed using a non-parametric Kruskal–Wallis one-way ANOVA (Graph-
Pad Prism V8.0.0, San Diego, CA, USA). p < 0.05 was considered statistically significant.
Data are expressed as mean ± S.D. unless otherwise indicated.
5. Conclusions
To the best of our knowledge, this is the first study that reports the semi-automated
1
8
synthesis of [ F]FB-IL2 using the commercially available Scintomics GRP
™
±
module. We
0.7% RCY)
and reduce the total synthesis time by ~30 min compared to previously developed methods
18
were able to increase the amount of available [ F]FB-IL-2 3.8-fold (i.e., 3.8
18
(i.e., 1.0
±
0.4% RCY). The improved synthesis of [ F]FB-IL-2 provides significantly greater
end-of-synthesis yields that facilitate the distribution of the radiopharmaceutical to remote
imaging centers without cyclotron facilities. Tracer properties including in vitro binding,
in vivo tissue distribution and kinetics are comparable to earlier published data on [ F]FB-
IL-2. PET imaging in MC38 and CT26 tumour-bearing mice indicates that [ F]FB-IL-2 can
distinguish tumours with higher levels of CD25+ tumour infiltrating lymphocytes.
1
8
1
8

Molecules 2021, 26, 1728
11 of 12
Supplementary Materials: Figure S1: Schematic diagram of Scintomics sequence visualization file
18
of semi-automated [ F]SFB synthesis with positions of reagent vials, reaction vial, transfer lines
and a syringe, Table S1. Scintomics GRP
Table S2. Gradient elution profile for [ F]SFB analysis by RP-HPLC, Figure S2. Radio- and UV
chromatograms of the DMSO reconstituted radiosynthon, [ F]SFB, Figure S3. Calibration curve of
™
time control file (sequence) for semi-automated [18F]SFB,
1
8
1
8
N-succinimidyl 4-fluorobenzoate, SFB, Table S3. Calibration curve calculations of N-succinimidyl 4-
1
8
fluorobenzoate, SFB, Table S4. Gradient elution for [ F]FB-IL-2 analysis by RP-HPLC, Figure S4. Cal-
ibration curve of N-(4-fluorobenzoyl)-interleukin-2, FB-IL2, Table S5. Calibration curve calculations
of N-(4-Fluorobenzoyl)-Interleukin-2, FB-IL2.
Author Contributions: Conceptualization, S.K. and F.F.Y.; methodology, S.K., F.F.Y., S.V.H., L.J., B.R.
and P.C.; formal analysis, S.K., F.F.Y., S.V.H. and P.N.; investigation, S.K., F.F.Y. and S.V.H.; Writing—
Original draft preparation, S.K., F.F.Y. and S.V.H.; Writing—Review and editing, S.K., F.F.Y., S.V.H.,
P.N., J.L.G. and E.G.R.; MS subject expert, P.N.; supervision, J.L.G. and E.G.R.; funding acquisition,
J.L.G. and E.G.R. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by Singapore’s Health and Biomedical Sciences (HBMS) Industry
Alignment Fund Pre-Positioning (IAF-PP) grant H18/01/a0/018, administered by the Agency for
Science, Technology and Research (A*STAR).
Institutional Review Board Statement: All animal procedures were carried out in accordance with
the Institute Animal Care and Use Committee (IACUC No. 181399) and conformed to the US National
Institutes of Health (NIH) guidelines and public law.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: This work was supported by funding from the Agency for Science, Technology
and Research (A*STAR), Singapore Bioimaging Consortium (SBIC), Singapore. The authors gratefully
acknowledge the cyclotron radiochemistry team, especially David J. Green at the Clinical Imaging
1
8
Research Centre (CIRC), National University of Singapore (NUS) for the provision of [ F]fluoride.
Conflicts of Interest: The authors declare no conflict of interest.
Sample Availability: Samples of the compounds are not available from the authors.
References
1
.
Wei, W.; Jiang, D.; Ehlerding, E.B.; Luo, Q.; Cai, W. Noninvasive PET Imaging of T Cells. Trends Cancer 2018, 4, 359–373. [CrossRef]
PubMed]
[
2
.
Goggi, J.L.; Hartimath, S.V.; Hwang, Y.; Tan, Y.X.; Khanapur, S.; Ramasamy, B.; Jiang, L.; Yong, F.F.; Cheng, P.; Tan, P.W.; et al.
Examining Immunotherapy Response Using Multiple Radiotracers. Mol. Imaging Biol. 2020. [CrossRef] [PubMed]
Chapman, N.M.; Chi, H. Hallmarks of T-Cell Exit from Quiescence. Cancer Immunol. Res. 2018, 6, 502–508. [CrossRef] [PubMed]
Capobianco, M.P.; Cassiano, G.C.; da Cruz Furini, A.A.; Storti de Melo, L.M.; Bonini Domingos, C.R.; Dantas Machado, R.L.
Human Interleukin 2 (IL-2) Promotion of Immune Regulation and ClinicalOutcomes: A Review. J. Cytokine Biol. 2016, 1, 14–17.
3
4
.
.
[CrossRef]
5
.
.
Di Gialleonardo, V.; Signore, A.; Glaudemans, A.W.J.M.; Dierckx, R.A.J.O.; De Vries, E.F.J. N-(4-18F-Fluorobenzoyl)Interleukin-2
for PET of Human-Activated T Lymphocytes. J. Nucl. Med. 2012, 53, 679–686. [CrossRef] [PubMed]
6
van der Veen, E.L.; Suurs, F.V.; Cleeren, F.; Bormans, G.; Elsinga, P.H.; Hospers, G.A.P.; Lub-de Hooge, M.N.; de Vries, E.G.E.;
de Vries, E.F.J.; Antunes, I.F. Development and Evaluation of Interleukin-2 Derived Radiotracers for PET Imaging of T-Cells in
Mice. J. Nucl. Med. 2020, 61, 1355–1360. [CrossRef] [PubMed]
7
8
9
.
.
.
van der Veen, E.L.; Antunes, I.F.; Maarsingh, P.; Hessels-Scheper, J.; Zijlma, R.; Boersma, H.H.; Jorritsma-Smit, A.; Hospers, G.A.P.;
de Vries, E.G.E.; Lub-de Hooge, M.N.; et al. Clinical-Grade N-(4-[1 F]Fluorobenzoyl)-Interleukin-2 for PET Imaging of Activated
T-Cells in Humans. EJNMMI Radiopharm. Chem. 2019, 4. [CrossRef] [PubMed]
8
Hartimath, S.V.; Manuelli, V.; Zijlma, R.; Signore, A.; Nayak, T.K.; Freimoser-Grundschober, A.; Klein, C.; Dierckx, R.A.J.O.;
de Vries, E.F.J. Pharmacokinetic Properties of Radiolabeled Mutant Interleukin-2v: A PET Imaging Study. Oncotarget 2018
, 7162–7174. [CrossRef] [PubMed]
Di Gialleonardo, V.; Signore, A.; Willemsen, A.T.M.; Sijbesma, J.W.A.; Dierckx, R.A.J.O.; De Vries, E.F.J. Pharmacokinetic Modelling
,
9
1
8
of N-(4-[ F]Fluorobenzoyl) Interleukin-2 Binding to Activated Lymphocytes in an Xenograft Model of Inflammation. Eur. J.
Nucl. Med. Mol. Imaging 2012, 39, 1551–1560. [CrossRef] [PubMed]

Molecules 2021, 26, 1728
12 of 12
1
1
0. Hartimath, S.V.; Draghiciu, O.; van de Wall, S.; Manuelli, V.; Dierckx, R.A.J.O.; Nijman, H.W.; Daemen, T.; de Vries, E.F.J.
1
8
Noninvasive Monitoring of Cancer Therapy Induced Activated t Cells Using [ F]FB-IL-2 PET Imaging. Oncoimmunology 2017
, 1–8. [CrossRef] [PubMed]
1. Shao, X. Synthesis of N-Succinimidyl 4-[18F]Fluorobenzoate ([18F]SFB). Radiochem. Syntheses 2012, 81–85. [CrossRef]
,
6