Caspase-3 probes for PET imaging of apoptotic tumor response to anticancer therapy

Article abstract of DOI:10.1039/c9ob00657e

Apoptosis is a highly regulated process involved in the normal organism development and homeostasis. In the context of anticancer therapy, apoptosis is also studied intensively in an attempt to induce cell death in cancer cells. Caspase activation is a known key event in the apoptotic process. In particular, active caspase-3 and-7 are the common effectors in several apoptotic pathways, therefore effector caspase activation may be a promising biomarker for response evaluation to anticancer therapy. Quantitative imaging of apoptosis in vivo could provide early assessment of therapeutic effectiveness and could also be used in drug development to evaluate the efficacy as well as potential toxicity of novel treatments. Positron Emission Tomography (PET) is a highly sensitive molecular imaging modality that allows non-invasive in vivo imaging of biological processes such as apoptosis by using radiolabeled probes. Here we describe the development and evaluation of fluorine-18-labeled caspase-3 activity-based probes (ABPs) for PET imaging of apoptosis. ABPs were selected by screening of a small library of fluorine-19-labeled DEVD peptides containing different electrophilic warhead groups. An acyloxymethyl ketone was identified with low nanomolar affinity for caspase-3 and was radiolabeled with fluorine-18. The resulting radiotracer, [¹⁸F]MICA-302, showed good labeling of active caspase-3 in vitro and favorable pharmacokinetic properties. A μPET imaging experiment in colorectal tumor xenografts demonstrated an increased tumor accumulation of [¹⁸F]MICA-302 in drug-treated versus control animals. Therefore, our data suggest this radiotracer may be useful for clinical PET imaging of response to anticancer therapy.

Full text of DOI:10.1039/c9ob00657e

View Article Online
View Journal
Organic &
Biomol ec ular
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: F. Elvas, T. Vanden
Berghe, Y. Adriaenssens, P. Vandenabeele, K. Augustyns, S. Staelens, S. Stroobants, P. Van der Veken and
L. wyffels, Org. Biomol. Chem., 2019, DOI: 10.1039/C9OB00657E.
Volume 14 Number
1
7
January 2016 Pages 1–372
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Organic &
Biomolecular
Chemistry
www.rsc.org/obc
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-0520
COMMUNICATION
Takeharu Haino et al.
Solvent-induced emission of organogels based on tris(phenylisoxazolyl)
benzene
rsc.li/obc

Page 1 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
Caspase-3 probes for PET imaging of apoptotic tumor response to
anticancer therapy
Filipe Elvasa,b,c†, Tom Vanden Berghed,e,f, Yves Adriaenssens , Peter Vandenabeeled,e,g, Koen Augustynsb,
b
a
a,c
b
a,c
Steven Staelens , Sigrid Stroobants , Pieter Van der Veken , Leonie wyffels
†Corresponding author:
Filipe Elvas, Molecular Imaging Center Antwerp, University of Antwerp, Universiteitsplein 1, 2610 Wilrijk,
Belgium.
Email: filipe.elvas@uantwerpen.be
Telephone number: +3238213453; Fax number: +3238253308
aMolecular Imaging Center Antwerp, University of Antwerp, 2610 Wilrijk, Belgium;
bLaboratory of Medicinal Chemistry, University of Antwerp, 2610 Wilrijk, Belgium;
cDepartment of Nuclear Medicine, Antwerp University Hospital, 2650 Edegem, Belgium;
dVIB Center for Inflammation Research, 9000 Ghent, Belgium;
eDepartment of Biomedical Molecular Biology; Ghent University, 9000 Ghent, Belgium;
fDepartment of Biomedical Sciences, University of Antwerp, 2610 Wilrijk, Belgium
gMethusalem program; Ghent University, 9000 Ghent, Belgium
Electronic Supplementary Information (ESI) available
1

Organic & Biomolecular Chemistry
Page 2 of 67
View Article Online
DOI: 10.1039/C9OB00657E
Abstract
Apoptosis is a highly regulated process involved in the normal organism development and homeostasis.
In the context of anticancer therapy, apoptosis is also studied intensively in an attempt to induce cell
death in cancer cells. Caspase activation is a known key event in the apoptotic process. In particular, active
caspase-3 and -7 are the common effectors in several apoptotic pathways, therefore effector caspase
activation may be a promising biomarker for response evaluation to anticancer therapy. Quantitative
imaging of apoptosis in vivo could provide early assessment of therapeutic effectiveness and could also
be used in drug development to evaluate the efficacy as well as potential toxicity of novel treatments.
Positron Emission Tomography (PET) is a highly sensitive molecular imaging modality that allows non-
invasive in vivo imaging of biological processes such as apoptosis by using radiolabeled probes. Here we
describe the development and evaluation of fluorine-18-labeled caspase-3 activity-based probes (ABPs)
for PET imaging of apoptosis. ABPs were selected by screening of a small library of fluorine-19-labeled
DEVD peptides containing different electrophilic warhead groups. An acyloxymethyl ketone was identified
with low nanomolar affinity for caspase-3 and was radiolabeled with fluorine-18. The resulting
1
8
radiotracer, [ F]MICA-302, showed good labeling of active caspase-3 in vitro and favorable
pharmacokinetic properties. A µPET imaging experiment in colorectal tumor xenografts demonstrated an
1
8
increased tumor accumulation of [ F]MICA-302 in drug-treated versus control animals. Therefore, our
data suggest this radiotracer may be useful for clinical PET imaging of response to anticancer therapy.
Keywords: Apoptosis, caspase-3 activity, warhead, activity-based probe, small-animal PET imaging.
2

Page 3 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
Introduction
Apoptosis is a highly regulated form of cell death that facilitates the orchestrated removal of damaged
and dead cells from healthy tissues. As such, this type of programmed cell death is an essential biological
process that plays a key role in normal development and in maintaining normal tissue homeostasis.
Disruption of homeostasis, either by excessive or deficient apoptosis is a hallmark of many pathological
processes including cardiovascular diseases, neurodegeneration, sepsis, and cancer.1-5 Because many
anticancer therapeutic strategies aim at promoting apoptosis in tumor cells, early detection of apoptosis
is of particular interest for treatment response evaluation in patients and in drug development.6, 7 The
activation of caspases is a critical component in the signaling cascades that lead to the execution of
apoptosis. Both intrinsic (via mitochondria) and extrinsic (activation of death receptors) pathways of
apoptosis converge in the activation of effector caspase-3, the most studied and characterized member
of this family. Several studies have implicated caspase-3 as an important effector molecule in promoting
therapy-induced cell death.8, 9 It is also known that dysregulated caspase-3 activity can lead to resistance
to drug-induced apoptosis in tumor cells.10-12 Therefore, methods that allow dynamic imaging of caspase-3
activity may have the potential to predict response to anticancer therapy.
Recent advances in non-invasive optical and nuclear imaging techniques have made it possible to image
protease activity, for monitoring disease progression and evaluating therapeutic efficacy.13, 14 Positron
Emission Tomography (PET) is a highly sensitive molecular imaging technique that does not suffer from
the common limitations of optical imaging, such as low sensitivity and low spatial resolution with
increased imaging depth. Several PET probes, or radiotracers, targeting active caspase-3 have been
developed and extensively explored in recent years. Examples of these radiotracers include radiolabeled
1
8
18
18
18
peptide derivatives ([ F]FB-VAD-FMK, [ F]C-SNAT, [ F]CP18, [ F]-TBD) and molecules with a canonical
1
8
18
15-20
18
18
small-molecule architecture ([ F]ML-10, [ F]ICMT-11).
Among these [ F]ML-10 and [ F]ICMT-11
3

Organic & Biomolecular Chemistry
Page 4 of 67
View Article Online
DOI: 10.1039/C9OB00657E
have been used in clinical trials.21-23 The first in human study using [ F]ML-10 - a low molecular weight
18
(
206 Da) molecule based on the alkyl-malonate group of γ-carboxyglutamic acid - indicated high in vivo
stability and favorable biodistribution and dosimetry profiles.21 In patients with brain metastases,
[18
F]ML-10 was able to detect responses to whole-brain radiotherapy several weeks earlier than standard
2
2
18
anatomical imaging techniques. Increased [ F]ML-10 uptake after successful therapy was found to be
associated with radiotherapy-induced tumor apoptosis.23 Notwithstanding these promising results,
[18
F]ML-10 uptake failed to correlate with drug-induced cell death, and the specific target of [ F]ML-10
18
remains undefined.24 Therefore, the utility of this probe for early evaluation of treatment response
warrants further investigation.
[18
F]ICMT-11 is another example of a small-molecule caspase inhibitor which was evaluated in clinical
trials. The probe is a nonpeptidic covalent inhibitor based on the isatin sulfonamide moiety with
1
8
nanomolar affinity for caspase-3 and -7 (0.5 and 2.5 nM, respectively). Isatins are covalent inhibitors
with a dicarbonyl moiety in their structure essential for binding to the nucleophilic active site cysteine.
Yet, the high reactivity of the dicarbonyl functionality, may also result in lower probe selectivity.25, 26 In
1
8
addition, complex voxel-wise analysis of [ F]ICMT-11 PET imaging data is necessary for analysis of
heterogeneous tumours.27 Despite these limitations [ F]ICMT-11 has proceeded into clinical studies.
18
1
8
Initial results demonstrated the safety of [ F]ICMT-11 with a dosimetry profile that lies within the range
1
8
28
18
of other clinically used F PET radiotracers. Recently, [ F]ICMT-11 entered a clinical study for assessing
1
8
therapeutic response in breast and lung cancer patients, where no significant changes in [ F]ICMT-11
2
9
tumor uptake were observed after therapy. This was attributed to the relatively small increase in active
caspase-3/-7 levels induced by chemotherapy.29
In spite of some impressive preclinical results and significant investment in clinical evaluation, so far, no
caspase-3-targeting PET imaging probe has made it into routine clinical practice for treatment response
4

Page 5 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
assessment in cancer patients. Two important reasons are cited: (1) lack of targeting group specificity for
caspase-3, due to the sequence homology among the caspases; and (2) low absolute target uptake, due
to rapid tissue diffusion.15, 30, 31 In order to yield useful probes, they should target individual caspases in
complex samples, and remain bound to the active form of the target caspase. This can potentially be
achieved by using activity-based probes (ABPs). Detection of specific active caspases using fluorescent
ABPs has been described in detail previously.32-35
Caspases must be catalytically active to kill a cell. Therefore, we reasoned that selective caspase-3
targeting ABPs might be useful to better define the individual role of caspase-3 in apoptosis-related
therapy response. ABPs are generally inhibitor-derived compounds capable of forming a covalent bond
with the active site of the target enzyme in an activity-dependent manner. Caspases preferentially
recognize and process short tetrapeptides with appropriate amino acid sequences within protein
substrates, which confers binding specificity to the active caspase.36 It has been demonstrated that
caspase-3 highly efficiently processes the Asp-Glu-Val-Asp (DEVD) sequence, when compared to initiator
3
7
and other effector caspases such as caspase-6 and -7. Therefore, recently a PET substrate containing the
DEVD epitope was developed and successfully used for the detection of active caspase-3 by PET imaging.38
Different warheads such as halomethyl ketones, acyloxymethyl ketones (AOMKs) and aldehydes have also
been incorporated in a series of peptide inhibitors for caspases, including caspase-3.15, 39-43 These
warheads are relatively strong electrophiles that react with the nucleophilic cysteine thiol at the active
site of the enzyme, and they can therefore have a positive impact on inhibitor potency and on selectivity
against other cysteine proteases, such as legumain and cathepsins.34, 44 Recent studies have additionally
identified Michael acceptors as a new class of potent and specific inhibitors for clan CD cysteine proteases,
which includes caspase-3.45, 46 When coupled with the preferred tetrapeptide recognition sequence for a
specific caspase, Michael acceptors did not show cross-reactivity with enzymes that belong to clan CA,
5

Organic & Biomolecular Chemistry
Page 6 of 67
View Article Online
DOI: 10.1039/C9OB00657E
such as calpain and cathepsin B.46, 47 Up to this moment, Hight et al. were the only ones to describe a
1
8
15
radiolabeled ABP, [ F]FB-VAD-FMK, for PET imaging of apoptotic effector caspase activity. The authors
showed that this radiotracer can be used to predict the response to a multidrug treatment regimen
consisting of an inhibitor of mutant BRAF and a phosphoinositide 3-kinase (PI3K) inhibitor. However, this
ABP has demonstrated lack of selectivity, consequently acting as a broad spectrum probe targeting not
1
5
only initiator and effector apoptotic caspases, but also inflammatory caspases at the same time. It is
important to note that this might be the result of a combination of two factors, the intrinsic high reactivity
of the fluoromethyl ketone (FMK) electrophilic warhead, and the presence of the non-specific VAD
peptide recognition sequence, which is efficiently cleaved by inflammatory caspase-1. Therefore, this
probe is limited in its utility when applied to anticancer therapies that induce inflammation such as
radiation therapy.
Since one of the pivotal hallmarks of cancer is the ability of tumor cells to evade apoptosis, cancer drug
4
8
development has predominantly focused on triggering apoptosis. Therefore, in this study we aimed to
develop non-invasive imaging tools that allow the evaluation of early efficacy of anticancer therapy. We
hypothesized that an ABP based on selective caspase-3 inhibitors would provide a starting point for the
design of an apoptosis targeting radiotracer. Here we describe the development and evaluation of
caspase-3 targeting PET radiotracers for apoptosis with an ABP architecture. The general structure of the
compounds that were designed in this study is depicted in Figure 1. All compounds are fluorinated,
containing a galacturonic acid moiety, a typical caspase-3-targeting DEVD sequence and an electrophilic
warhead functionality. To systematically probe for the influence of the warhead on inhibitor potency and
selectivity, several warhead types were included in the series. A careful selection of these groups was
performed, as the degree of reactivity of the electrophilic warhead can have an impact on the selectivity
4
9
of an ABP. Therefore, the inhibitors included in this study can be classified as reversible (aldehydes) or
6

Page 7 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
irreversible (Michael acceptors and activated ketones like AOMKs and fluoromethyl ketones (FMKs)),
4
9
which have distinct modes of binding, thus modulating inhibitor potency. Among the latter group, the
AOMKs are the weakest electrophiles, thus they have shown little cross-reactivity with other
4
0
bionucleophiles. A non-reactive amide functionality was used as control. To separate the fluorine tag
from the caspase active site, we incorporated the galacturonic acid spacer at the N-terminus of the
molecule. This moiety was also used to improve the pharmacokinetics of the ABP, as previously described
1
8
for [ F]CP18, a radiolabeled DEVD-containing substrate-based probe developed for imaging of caspase-3
5
0
activity. The PET radiotracers were first synthesized and biochemically evaluated in a non-radiolabeled
(
i.e., 19F-containing) form. The most promising inhibitor was labeled with 18F via copper(I)-catalyzed
azide-alkyne cycloaddition (CuAAC). The developed 18F-labeled ABP was further characterized by
evaluating its binding efficiency to human recombinant caspases, cell lysates and tumor homogenates.
The preliminary evaluation of the ABP potential for in vivo PET imaging, included the determination of the
metabolic profile and in vivo biodistribution of the 18F-labeled ABP in normal CD1-/- nude mice.
Additionally, a µPET imaging study was performed in colorectal cancer xenografts treated with targeted
therapy to assess the probe’s ability to in vivo image caspase-3 activation and subsequent tumor response
to treatment.
7

Organic & Biomolecular Chemistry
Page 8 of 67
View Article Online
DOI: 10.1039/C9OB00657E
Activity-based probes general structure
O
OH
OH
OH
HO
OH
H H
N
O
O
O
H
H
N
N
F
N
N
O
N
H
N
R
H
N
O
O
O
OH
P1' warhead
fluorine tag
O
spacer
DEVD recognition sequence
O
O
O
1
a
b
R =
1d R =
O
O
O
O
O
1
R =
R =
S
1e R =
1f R =
F
O
H
NH2
1
c
O
Figure 1. General structure of DEVD peptide inhibitors containing different warhead functionalities (R) at the P1’
position.
Results and discussion
Chemistry
The synthetic preparation of target compounds started with the preparation of a general DEV-based,
tripeptide precursor common to all target compounds (Scheme 1, compound 5). The latter was prepared
in gram-scale amounts using solution- or liquid-phase peptide synthesis (LPPS) similar to previously
described methods.51, 52 According to a previously reported procedure, tripeptide 2 was deprotected to
obtain 3, which was equipped with Fmoc-protected β-D-galacturonic acid (Fmoc-SAA) to give 4.53
Hydrogenative O-debenzylation finally led to intermediate 5.
8

Page 9 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
Scheme 1. Preparation of the Fmoc-SAA-DEV peptidea
O
O
O
O
O
O
O
O
H
H
N
H
N
a
O
N
H2N
N
O
N
O
H
H
O
O
O
O
O
O
O
2
3
O
O
O
O
OH
OH
O
HO
OH
H H
N
HO
OH
H H
N
O
O
O
O
H
H
N
H
H
N
b
c
O
N
O
N
O
N
O
N
OH
H
H
O
O
O
O
O
O
O
O
O
O
4
5
aReagents and conditions: (a) TAEA, DCM, rt, 30 min; (b) 1-(Fmoc-aminomethyl)-ß-D-galacturonic acid, HOBt,
EDC.HCl, DIPEA, DCM/DMF, rt, 16 h; (c) Pd/C, H , MeOH, 6 h.
2
The preparation of the Asp-derived, warhead-containing moieties is represented in Scheme 2. The
α,β-unsaturated ethyl ester 6 was obtained using the Wittig reaction, as previously reported.54 Vinyl
sulfone 8 derivatized with N-Boc protected Asp was prepared by reaction with the corresponding
phosphonate by the Horner-Wadsworth-Emmons condensation reaction.55 Acetal protected N-Cbz
aldehyde moiety 10 was prepared as previously described.56 Formation of the acyloxymethyl ketone
(AOMK) 12 was performed from a N-Boc Asp-derived chloromethyl ketone (CMK), using a procedure
described by Krantz et al.57, 58 FMK 14 was prepared according to a procedure described for the
5
9
development of a fluorescent ABP, Bodipy-VAD-FMK. The control amide 16 was synthesized according
6
0
to a known procedure. The deprotection of the N-Boc group in 6, 8, 12 and 14 was performed under
mild acidic conditions for maintaining the tert-Bu side chain protection, resulting in the tosylate salts 7, 9,
6
1
1
3 and 15, respectively. Deprotection of the N-Cbz protected warheads 10 and 16 was conducted via
catalytic hydrogenation to yield the free amines 11 and 17.56
9

Organic & Biomolecular Chemistry
Page 10 of 67
View Article Online
DOI: 10.1039/C9OB00657E
Scheme 2. Preparation of the Asp-derived, warhead-containing moietiesa
O
O
O
O
O
S
O
O
S
O
O
O
O
O
O
a
O
O
O
O
O
O
a
H3N
H3N
O
O
O
N
H
O
O
O
N
O
O
H
O
O
O
6
7
12
14
13
O
O
O
O
O
S
O
O
O
O
O
O
a
b
a
c
O
S
O
O
O
H3N
S
H3N
F
O
O
N
S
O
O
N
F
O
H
H
O
O
8
9
15
O
O
O
O
O
O
O
O
O
O
O
O
O
NH2
NH2
O
N
H2N
O
N
2
H N
H
H
O
O
O
1
0
11
16
17
aReagents and conditions: (a) p-Toluene sulfonic acid, DCM/Et
hydroxide, EtOAc, rt, 24 h; (c) Pd/C, H , MeOH, rt, 1 h.
2
O, rt, 48 h; (b) 1,4-Dihydrobenzene, palladium
2
1
9
The common peptide 5 was subsequently transformed into finalized F-compounds by introducing the
Asp-derived, warhead containing moieties 7, 9, 11, 13, 15, 17 (Scheme 3). The obtained 18a-f were
deprotected and N-terminal extended with azidoacetic acid to give the protected azides 19a-f. Finally,
CuAAC of the azides 19a-f with 5-fluoropent-1-yne, was followed by trifluoroacetic acid (TFA)
deprotection of the obtained triazoles to give the final fully deprotected inhibitors 1a-f.
10

Page 11 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
1
9
a
Scheme 3. Preparation of the F-DEVD inhibitors
O
O
O
O
OH
OH
O
O
O
HO
OH
H H
N
HO
OH
O
O
O
O
H
H
N
H
H
H
N
H
O
N
a
O
N
N
O
N
OH
N
N
R
H
H
H
O
O
O
O
O
O
O
O
O
O
5
18
O
OH
OH
O
O
R
O
O
OH
O
O
HO
OH
O
O
OH
HO
OH
H H
N
H
H
H
N
H
O
O
O
b, c
H
H
N
N
N
d, e
F
N
N
N
N
N
N3
N
H
N
R
N
H
H
O
O
O
H
O
O
O
OH
O
O
O
1
9
1
O
O
O
H
1
, 18, 19a R =
1, 18, 19b R =
1, 18, 19e R =
S
18, 19c R =
1c R =
O
O
O
O
O
NH2
1
, 18, 19d R =
O
F
1, 18, 19f R =
O
O
O
aReagents and conditions: (a) 7, 9, 11, 13, 15, or 17, HOBt, EDC.HCl, DIPEA, DCM/DMF, rt, 16 h; (b) DEA, THF, rt, 8 h;
c) Azidoacetic acid, HOBt, EDC.HCl, DIPEA, DCM/DMF, rt, 16 h; (d) sodium ascorbate, CuSO , TBTA, 5-fluoropent-1-
yne, tert-BuOH/H O, rt, 16 h; (e) TFA, TIS, H O, rt, 30 min.
(
4
2
2
Although we used synthetic procedures available from the literature, and reaction conditions which are
known to minimize the occurrence of racemization (e.g. hydroxybenzotriazole, sterically hindered bases
and short reaction times), in some of the final products a mixture of two epimers was observed. This was
most likely due to P1 Asp racemization during the preparation of the Asp-derived, warhead-containing
moieties. These products were separated by HPLC and only the inhibitory activity of the most abundant
and most active diastereoisomer is reported.
Biochemical evaluation of the inhibitor library
Prototypical caspase inhibitors contain simple warheads attached to a peptide backbone, inhibiting
caspases via covalent modification of the cysteine active site. The most commonly used warheads
11

Organic & Biomolecular Chemistry
Page 12 of 67
View Article Online
DOI: 10.1039/C9OB00657E
encompass aldehydes and ketones with their derivatives, such as FMKs or AOMKs, which tend to be highly
3
7
reactive with cysteine proteases. These small groups bind tightly to the S1’ pocket in the enzyme, which
is in agreement with the observation that caspases recognize small residues in the P1’ position of their
3
7
substrates. Therefore, optimization of the P1’ probe position can be of special interest for increasing
inhibitor affinity and selectivity for individual caspases.37
Inhibitors 1a-f were examined for their ability to inhibit human and mouse recombinant caspases. The
inhibitory potency of these compounds was assessed using a fluorogenic caspase inhibition assay, as
6
2
previously described. Two commercially available inhibitors were included as reference compounds: the
caspase-3 and -7 specific reversible peptide inhibitor, Ac-DEVD-CHO; and a broad spectrum irreversible
pan-caspase inhibitor, zVAD-FMK. All the synthesized inhibitors displayed inhibition potency in the nM
range against caspase-3 and -7 recombinant human and mouse caspases, as shown by the IC values
5
0
listed in Tables 1 and 2, respectively.
12

Page 13 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
1
9
Table 1. Inhibitory potency of F-DEVD peptide inhibitors 1a-f against human caspases expressed as IC
5
0
values.
O
OH
OH
O
OH
HO
H
OH
H H
N
O
O
O
H
N
N
F
N
N
N
R
H
H
N
N
O
O
O
OH
O
Inhibitory potency IC50 (nM)a
Compound
R
caspase-1 caspase-3 caspase-6 caspase-7 caspase-8 caspase-9
455.0 ±
1
a
b
O
>
>
>
1,000
1,000
1,000
20.7 ± 1.4
> 1,000
> 1,000
> 1,000
48.0 ± 1.2
> 1,000
O
2
9.1
1
776.1 ±
O
S
24.4 ± 1.9
> 1,000
> 1,000
O
2
2.7
159.7 ±
.0
1
c
H
19.9 ± 2.1
1.0 ± 0.1
7.6 ± 0.8
> 1,000
69.5 ± 2.3
> 1,000
> 1,000
O
8
O
1
d
e
O
> 1,000
> 1,000
20.0 ± 1.9 13.2 ± 0.5 23.6 ± 2.0
O
1
208.5 ±
432.8 ±
37.7
F
68.7 ± 3.9 35.0 ± 2.5
O
O
1
7.5
1
f
NH2
>
1,000
112.7 ±
.4
> 1,000
> 1,000
> 1,000
> 1,000
zVAD-FMK
154.8 ±
10.2
190.5 ±
10.2
n.a.
n.a.
63.9 ± 2.6 11.5 ± 0.3 19.0 ± 1.3
3
Ac-DEVD-
CHO
> 1,000
18.8 ± 1.6
> 1,000
35.0 ± 2.6 45.5 ± 2.8
> 1,000
aValues are the mean ± SD of two independent duplicate determinations. n.a.: not applicable.
13

Organic & Biomolecular Chemistry
Page 14 of 67
View Article Online
DOI: 10.1039/C9OB00657E
Table 2. Mouse caspase inhibition potency of selected compounds 1a-f.
Inhibitory potency IC50 (nM)a
Compound
caspase-1
> 1,000
> 1,000
> 1,000
10.5 ± 4.4
31.4 ± 4.2
> 1,000
n.d.
caspase-3
1.4 ± 0.1
15.1 ± 0.9
1.4 ± 0.1
0.9 ± 0.1
2.9 ± 0.2
> 1,000
caspase-7
21.7 ± 2.9
129.2 ± 29.6
21.7 ± 2.9
18.1 ± 1.7
57.7 ± 5.0
> 1,000
caspase-8
475.9 ± 45.7
> 1,000
1
a
b
1
1
c
68.4 ± 8.3
4.4 ± 0.1
4.1 ± 0.2
> 1,000
1
d
e
1
1
f
zVAD-FMK
84.1 ± 10.7
8.6 ± 0.5
42.7 ± 3.1
43.7 ± 6.9
4.4 ± 0.5
47.2 ± 7.4
Ac-DEVD-CHO
77.2 ± 6.2
aValues are the mean ± SD of duplicate determinations. n.d.: not determined.
AOMK derivative 1d was the most potent inhibitor for caspase-3 (IC = 1.0 ± 0.1 nM, for human, and IC
5
0
50
=
0.9 ± 0.1 nM, for mouse caspase-3), which is 9 to 19 times more potent than the reference inhibitor
Ac-DEVD-CHO. This is in good agreement with the preference of caspase-3 for bulky and hydrophobic
3
7
warheads at P1’. Furthermore, compound 1d showed some promiscuity against the other caspases, with
2
0- and 13- to 20-fold selectivity over executioner caspase-6 and -7, respectively, and 5- to 24-fold
selectivity over the initiator caspase-8. In contrast, the binding potency toward human caspase-9 was
weak, with IC > 1,000 nM. FMK compound 1e also displayed a good binding potency towards caspase-3
5
0
(
IC = 7.6 ± 0.8 nM, for human caspase). Nevertheless, this inhibitor showed less selectivity for caspase-3
5
0
when compared to 1d. FMK-based inhibitors were among the first to be developed as caspase inhibitors,
yet the high reactivity of the FMK group often leads to off-target binding, thus cross-reactivity with other
14

Page 15 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
cysteine proteases. By comparison, the AOMK-based inhibitors are weaker electrophiles, thus they are
more suitable for investigating caspases because of lower cross-reactivity with other bionucleophiles.37
Although binding potencies toward human caspase-1 were weak (IC > 1,000 nM), compounds 1d and 1e
5
0
inhibited mouse caspase-1. This difference in affinity between human and murine caspase-1 might be
explained by the fact that murine caspase-1 contains 3 additional aspartates between D296 and D314,
which could increase the affinity of these two compounds for murine caspase-1.63
Both Michael acceptors 1a and 1b, showed weaker inhibitory potency against human caspase-3,
compared to the activated ketones (AOMK and FMK) with affinities of 20.7 ± 1.4 nM and 24.4 ± 1.9 nM,
respectively. The increased inhibitory potency of compound 1a (1.4 ± 0.1 nM) against murine caspase-3
compared with the one of inhibitor 1b (15.1 ± 0.9 nM) was surprising as vinyl sulfones are 2 to 10 times
more reactive towards nucleophiles than vinyl esters.55 Nevertheless, the vinyl sulfone containing
inhibitor 1b exhibited a higher degree of selectivity for human caspase-3 over caspase-7 (32-fold)
comparing to ketone counterparts (13- and 9-fold for AOMK and FMK, respectively), in agreement with
5
5
previously reported data. Differences in the size and shape of the S1’ subsite between caspase-3 and
caspase-7, have been explored for the modulation of inhibition selectivity between caspase-3 and
4
3
caspase-7 by changing the functionality that occupies the P1’ position. Another potential advantage of
this α,β-unsaturated scaffold tailored to explore the P1’ region is the lower cross-reactivity towards other
cysteine proteases, as caspases prefer trans stereochemistry in their inhibitors.47 The low activity of
compounds 1a and 1b, taken together with previous studies, yields these Michael acceptor warheads not
suitable for caspase inhibition.55
Compound 1c exhibited an inhibitory potency for human caspase-3 (IC = 19.9 ± 2.1 nM), similar to
5
0
reference inhibitor Ac-DEVD-CHO, with very good selectivity for human caspase-3 over human caspase-1,
6 and -9. However, inhibitor 1c displayed low selectivity for human caspase-3 over caspase-8 (3.5-fold),
-
15

Organic & Biomolecular Chemistry
Page 16 of 67
View Article Online
DOI: 10.1039/C9OB00657E
4
2
in good agreement with the preference of caspase-8 for small and aliphatic groups at P1’. As expected,
control compound 1f showed no caspase inhibition.
Promising work on activity-based probe research has shown that many efforts are directed toward
selectivity for a single caspase. Recently, different research groups have independently developed peptide
libraries containing both natural and unnatural amino acids in order to provide compounds to distinguish
between initiator and effector caspases.41, 64, 65 The inhibitors developed by Wolan’s lab were successfully
converted into fluorescent ABPs and displayed high selectivity for caspase-3 over caspase-7 and -8.34
Despite favorable affinity and selectivity profile obtained with fluorinated DEVD-containing inhibitors, the
development of next-generation derivatives, using the same general ABP structure, but displaying
improved selectivity profiles, are warranted for maximum selective caspase targeting.
Radiosynthesis
Since compound 1d demonstrated the highest inhibitory potency against both human and mouse caspase-
1
8
3
and -7 in the biochemical assays, it was selected as candidate for radiolabeling with F. Compound 1f
which contains an amide in place of the reactive AOMK functional group, was selected as a control
radiotracer. The azide precursors for radiolabeling 20a and 20b were obtained by TFA deprotection of 19f
and 19d, respectively. The radiolabeling of compounds 1f and 1d was performed by CuAAC of
1
8
18
5
-[ F]fluoro-1-pentyne ([ F]22), prepared by nucleophilic substitution of the tosylate precursor 21 with
[18
F]fluoride, with the azide precursors 20a and 20b, yielding the F-labeled triazoles [ F]MICA-301 and
18
18
[18
F]MICA-302, respectively (Scheme 4).
16

Page 17 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
1
8
18
a
Scheme 4. Radiosynthesis of [ F]MICA-301 and [ F]MICA-302
O
OH
OH
O
OH
O
HO
H
OH
H H
N
O
O
H
N
N
NH2
N
20a
18F
N
N
H
H
O
N
N
O
O
O
b
OH
O
[18
F]MICA-301
a
OTs
18F
[18F]22
21
O
OH
OH
O
OH
HO
OH
H H
N
O
O
O
O
H
H
N
N
20b
18F
N
N
N
O
H
H
N
N
O
O
O
O
b
OH
O
[18
F]MICA-302
a
18
Reagents and conditions: (a) K[ F]F/K2.2.2, K
2 3 3 4
CO , CH CN, 110°C, 3 min; (b) CuSO , sodium ascorbate, TBTA, EtOH,
3
CH CN, rt, 20 min.
The optimal conditions for click reaction were 20 min reaction time under stirring at room temperature
in the presence of aqueous copper sulfate, sodium ascorbate, and TBTA as copper-stabilizing ligand.17
1
8
18
Following HPLC purification and sterile formulation, control probe [ F]MICA-301 and ABP [ F]MICA-302
were obtained in 13.4 ± 4.0 (n=4, decay-corrected) and 12.6 ± 2.1% (n=4, decay-corrected) isolated
radiochemical yield (RCY), respectively. By using a biocompatible mobile phase for purification, an
additional purification step was avoided yielding the final products in a total synthesis time of 90 min from
end of bombardment (EOB), including formulation. Both tracers were obtained in a radiochemical purity
>
99%. The molar activities (Am) of both radiotracers were in the range of 50-300 GBq/µmol at EOB. The
identity of the radiolabeled compounds was confirmed by co-injection and co-elution on HPLC of
17

Organic & Biomolecular Chemistry
Page 18 of 67
View Article Online
DOI: 10.1039/C9OB00657E
[18
F]MICA-301 and [ F]MICA-302 with the corresponding non radiolabeled ‘cold’ reference compounds
18
1
f (t = 6.7 min, gradient method A) and 1d (t = 7.2 min, gradient method B), respectively (Supplemental
R
R
Figure 1).
1
8
18
The partition coefficients (logD values) of [ F]MICA-301 and [ F]MICA-302 were determined
1
8
18
experimentally. Both [ F]MICA-301 (logD = - 3.29 ± 0.12) and [ F]MICA-302 (logD = - 3.09 ± 0.05) were
found to be highly hydrophilic, which generally confers poor cell membrane permeability.
Activity-based labeling of recombinant enzymes
To evaluate the affinity and selectivity of the radiolabeled ABPs for caspase-3 we performed a binding
assay using equal concentrations (20 nM) of pre-mixed caspase-3 and caspase-6, -7 or -8, and legumain
and cathepsin B, followed by addition of probe and subsequent incubation with the enzyme mixture for
1
8
1
1
0 and 60 minutes. As expected, only the active ABP [ F]MICA-302 efficiently labeled caspase-3 within
1
8
0 min after incubation (Figures 2a, b and c), whereas [ F]MICA-301 demonstrated no binding to any of
1
8
the enzymes (results not shown). Although [ F]MICA-302 predominantly labeled caspase-3, it also
demonstrated binding to cathepsin B, and to a smaller extent legumain (Figure 2d). This promiscuous
binding profile has been reported before for other caspase-3 ABPs.32,40,44 Moreover, zVAD-FMK also
inhibited legumain (Figure 2d), which although a lysosomal protease, belongs to clan CD proteases similar
to caspases and is not related to cathepsins.66
1
8
Interestingly, [ F]MICA-302 demonstrated a lower degree of cross-reactivity against other caspases
compared to fluorescence-based DEVD-AOMK probes.41, 66 This might be explained by the fact that
[18
F]MICA-302 was used in low nanomolar concentration, below enzyme concentration. As previously
shown, at stoichiometric concentrations a 5(6)-carboxyfluorescein-labeled ABP (FAM-DEVD-AOMK) binds
almost exclusively to caspase-3. However, in the presence of excess probe (micromolar concentrations),
18

Page 19 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
4
1
promiscuous labeling of caspase-6, -7 and -8 occurs. In general, a high amount of probe is required for
fluorescence imaging thereby increasing the risk of off-target binding, while the use of only nmol to fmol
amounts are required for imaging when using radiotracers.
Figure 2. Biochemical evaluation of activity-based probes. Direct labeling of recombinant caspase-3 and caspase-6
1
8
(
a), caspase-3 and caspase-7 (b), caspase-3 and caspase-8 (c), and legumain and cathepsin B (d) with [ F]MICA-302.
The probe (74 kBq) was incubated with recombinant human enzymes for 10 and 60 minutes, the reaction was
stopped, enzymes were separated by SDS-PAGE and the gel was scanned for radioactivity. Samples marked with +
represent pre-treatment with the pan-caspase inhibitor zVAD-FMK (1 µM) which was added 30 min prior to probe.
The location of the caspase-3 p12 small subunit is also indicated. Cropped regions of the gel images containing the
expected protein bands are shown.
Activity-based labeling of apoptotic cell lysates
1
8
The ability of [ F]MICA-302 to detect activated caspase-3 in vitro was evaluated using lysed HeLa cells
treated with staurosporine (STS). Caspase-3-deficient MCF-7 cells treated with STS were used as a
6
7
negative control. First, the optimal time window for detection of caspase-3 activity in apoptotic HeLa
cells was determined in a biochemical caspase activity assay using a fluorogenic substrate (Figure 3a).
19

Organic & Biomolecular Chemistry
Page 20 of 67
View Article Online
DOI: 10.1039/C9OB00657E
Incubation of HeLa cells with a fluorogenic substrate indicated significant caspase-3 activity within 4 h
after treatment with STS (p<0.001), gradually decreasing until 24 h. Pre-incubation with the pan-caspase
inhibitor zVAD-FMK successfully abolished caspase-3 activity (p<0.001; Figure 3a). This was confirmed by
blotting with a caspase-3 antibody (Figure 3b). Following a 60 min incubation with the cell lysates,
1
8
autoradiography of the SDS-PAGE gel demonstrated that [ F]MICA-301 did not label any enzymes (results
1
8
not shown). By contrast, active caspase-3 was detected by [ F]MICA-302 4 to 10 h after STS treatment in
HeLa cells (Figure 3c). Processing of procaspase-3 (32 kDa) involves two different cleavage events. First
initiator caspases cleave within its interdomain linker, found between the p20 and p10 domains, and then
its N-terminal prodomain is removed resulting in the formation of the mature active caspase-3 (17 kDa).
However, if the prodomain of caspase-3 is not efficiently removed a fragment will run at 19 kDa.68
1
8
Detection of active caspase-3 by [ F]MICA-302 was highest 4 h post treatment, and two protein bands
could be visualized at 19 and 17 kDa. In contrast, in MCF-7 cells no labeling of active caspase-3 was
detected (Figure 3c). Although MCF-7 cells do not express caspase-3, it has been demonstrated that
6
7
apoptosis proceeds via activation of caspase-9, -7 and -6. In a recent study, using a rhodamine-containing
ABP Rho-DEVD-AOMK, additional bands for caspase-6 and -7 were observed at 8 and 14 h post STS
3
4
treatment in MCF-7 cells. This suggests again that the use of lower amounts of radiolabeled ABPs might
reduce the risk of off-target binding, when comparing to fluorescent ABPs. Radiotracer specificity was
further assessed by addition of the inhibitor zVAD-FMK, before STS treatment or after radiotracer
incubation in HeLa cells, which successfully inhibited caspase-3 labeling (Figure 3c). Blotting with a
caspase-3 antibody showed that zVAD-FMK indeed inhibited cleaved caspase-3 (Figure 3b). These results
1
8
show that [ F]MICA-302 labels active caspase-3 in apoptotic cell lysates.
20

Page 21 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
Figure 3. Caspase-3/7 activity measurement (a), detection of cleaved caspase-3 (b), and labeling of endogenous
1
8
active caspase-3 with [ F]MICA-302 (c) in HeLa and MCF-7 cells treated with staurosporine. Location of active
caspase-3 is shown based on molecular weight. ***p<0.001, significantly different from control; ###p<0.001,
significantly different from 4 h STS.
Activity-based labeling of tumor homogenates
1
8
Finally, the usefulness of [ F]MICA-302 to monitor activation of caspase-3 in vivo was evaluated using a
model of targeted therapy-induced apoptosis. For this, tumor homogenates were prepared from mice
bearing therapy-resistant HT29 and therapy-sensitive COLO205 tumor xenografts and treated with a
single dose of conatumumab or the corresponding IgG1 isotype control. Incubation of the tumor
homogenates with a fluorogenic substrate indicated significant caspase-3 activity in the COLO205 tumors
2
4 h after treatment with conatumumab (p=0.002), but not after treatment with the control IgG1
antibody. By contrast, very low caspase-3 activity could be detected in the treatment-resistant HT29
tumor homogenates (Figure 4a). Accordingly, blotting with a caspase-3 antibody showed that cleaved
21

Organic & Biomolecular Chemistry
Page 22 of 67
View Article Online
DOI: 10.1039/C9OB00657E
caspase-3 could be detected in COLO205 tumors treated with conatumumab (Figure 4b). After addition
1
8
18
of [ F]MICA-301 or [ F]MICA-302 to the homogenates, the samples were run by SDS-PAGE followed by
autoradiography of the gels for detection of enzyme labeling. Active caspase-3 could be detected with
[18
F]MICA-302 in therapy-sensitive COLO205 tumors treated with conatumumab, but not in the tumors
treated with IgG1 nor in the resistant HT29 tumors (Figure 4c). A doublet of labeled proteins could be seen
as 19 and 17 kDa bands, corresponding to the active caspase-3 as well as to the form of this protease
4
0
18
without its N-terminal prodomain. Addition of [ F]MICA-301 to the tumor homogenates did not result
in any labeling (results not shown).
Figure 4. Caspase-3/7 activity measurement (a), detection of cleaved caspase-3 (b), and labeling of endogenous
1
8
active caspase-3 with [ F]MICA-302 (c) in COLO205 and HT29 tumor homogenates 24 h after treatment with
targeted therapy. Location of active caspase-3 is shown based on molecular weight. **p=0.002, significantly different
from IgG1-treated COLO205.
In vitro stability and in vivo metabolic stability studies
1
8
Because [ F]MICA-302 demonstrated a relative selective binding to caspase-3 when using a mixture of
recombinant enzymes, as well as in the cell lysate and tumor homogenate study, its potential for in vivo
1
8
18
imaging was investigated. First, the in vitro stability of [ F]MICA-301 and [ F]MICA-302 was determined
in PBS and in mouse plasma. In the validation study extraction yields of 91.4 ± 3.0% and 88.4 ± 2.8% for
[18
F]MICA-301 and [ F]MICA-302, respectively, were determined and no degradation of the radiotracers
18
22

Page 23 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
occurred during work-up. As shown in Figure 5, both radiotracers showed good stability in PBS at 37°C up
to 120 min, with only the parent radiotracers being detected by radio-HPLC, without signs of degradation
1
8
18
(
[ F]MICA-301: 93.5±1.2% intact; [ F]MICA-302: 95.8±2.3% intact). After incubation in mouse plasma for
1
8
1
20 min at 37°C, [ F]MICA-301 revealed the formation of one minor polar metabolite (t = 4.5 min,
R
1
8
Supplemental Figure 2; 81.8±2.2% intact), whereas [ F]MICA-302 remained largely intact (95.1±0.3%).
Additionally, stability of both radiotracers was evaluated in vivo following intravenous (i.v.) injection of
the radiotracers in normal CD1 nude mice (Figure 5). Radio-HPLC analysis of the plasma revealed good in
1
8
18
vivo stability of both [ F]MICA-301 (92.6±0.1% intact) and [ F]MICA-302 (87.1±2.5% intact) up to 60- and
1
20-min post-radiotracer injection (p.i.), respectively, in normal CD1 nude mice. Similar to the in vitro
1
8
stability study, for [ F]MICA-301 one main polar metabolite could be detected in the plasma (t = 4.5 min)
R
at 120 min p.i., accounting for 26% of the total plasma activity (Supplemental Figure 2). A previous study
1
8
has reported on the hydrolysis of a structurally related substrate-based probe, [ F]CP18, by plasma
1
8
50
hydrolases with the formation of a [ F]DE radiolabeled fragment. We therefore hypothesize that similar
1
8
enzymatic hydrolysis of [ F]MICA-301 has occurred in our study, with the formation of a polar
radiometabolite. However, this metabolite was not identified experimentally. Nevertheless, since the
formation of this metabolite was observed only after 120 min p.i. we do not expect this metabolite to
represent a confounding factor for radiotracer uptake at the time of imaging (60 min p.i.) in tumor animal
models.
23

Organic & Biomolecular Chemistry
Page 24 of 67
View Article Online
DOI: 10.1039/C9OB00657E
1
8
18
Figure 5. Stability of [ F]MICA-301 (a) and [ F]MICA-302 (b) in PBS and mouse plasma at 37°C, and after i.v. injection
in CD1 nude mice.
In vivo and ex vivo biodistribution study
When used for in vivo imaging, a radiotracer should demonstrate good pharmacokinetic properties with
efficient clearance from the blood and non-target organs and tissues. The biodistribution profile of ABP
[18
F]MICA-302 was therefore evaluated by in vivo µPET imaging and ex vivo biodistribution and compared
1
8
to [ F]MICA-301. Figure 6 shows representative µPET/CT sagittal whole-body images recorded between
0 and 60 minutes and the time-activity curves (TACs) for heart (representative for the blood clearance),
muscle, and major excretory organs liver and kidneys up to 60 min p.i. in normal CD1 nude mice.
3
24

Page 25 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
1
8
Figure 6. In vivo biodistribution of radioactivity in normal CD1-/- nude mice after intravenous injection of [ F]MICA-
1
8
3
01 (a) and [ F]MICA-302 (b). Representative sagittal µPET/CT images (30-60 minutes summed activity) of normal
mice. Time-activity curves for heart, muscle, liver and kidneys are shown. Data are presented as mean ± SD.
For both radiotracers an initial peak uptake of radioactivity was immediately followed by fast blood
1
8
clearance with almost complete elimination from the blood pool at 60 min p.i., with [ F]MICA-302 (4.1 ±
1
8
1
.1 %ID/mL) showing a slower blood clearance profile than [ F]MICA-301 (1.0 ± 0.4 %ID/mL). The uptake
in the liver and kidneys was determined to evaluate the elimination pathway of the radiotracers. The
1
8
radiotracers were mainly eliminated via the kidneys. [ F]MICA-301 displayed a higher initial kidney
1
8
uptake (34.2 ± 8.1 %ID/mL at 6 min p.i.) compared to [ F]MICA-302 (26.1 ± 10.4 %ID/mL at 2 min p.i.),
followed by a washout of activity from the kidneys (Figure 6). As was also shown for the heart/blood pool
1
8
18
activity, slower liver elimination was visible for [ F]MICA-302 compared to [ F]MICA-301. The lack of
significant nonspecific accumulation in other organs, resulted in general low image background at 60 min
p.i. Therefore, apoptosis imaging can be performed within 60 min p.i.
An ex vivo biodistribution was performed at 15, 30, 60 and 120 min p.i. to validate the μPET imaging
1
8
results. The fast blood clearance and renal excretion of [ F]MICA-301 observed by µPET were reflected
in the biodistribution studies (Supplemental Figure 3). This renal excretion profile has also been reported
1
8
50 18
for [ F]CP18, a structurally related substrate-based probe for caspase-3 imaging. [ F]MICA-302 on the
25

Organic & Biomolecular Chemistry
Page 26 of 67
View Article Online
DOI: 10.1039/C9OB00657E
other hand demonstrated a mixed renal-hepatobiliary clearance with increasing uptake of radioactivity in
the small (3.2 ± 0.4 %ID/g at 60 min p.i.) and large intestine (2.7 ± 0.7 %ID/g at 120 min p.i.) (Supplemental
Figure 3). Multiple cysteine proteases, including caspases, cathepsins and legumain, are known to digest
proteins in the intestine as part of a coordinated physiological homeostatic process.69-71 Therefore, the
1
8
uptake of radioactivity in the intestines might also reflect binding of [ F]MICA-302 to cysteine proteases
present in the intestinal mucosa. In addition, the high uptake in the small intestine could also be due in
part to the high physiological apoptosis of enterocytes, as described before for other caspase-3/7 PET
3
1
imaging radiotracers. Both radiotracers demonstrated low accumulation of radioactivity in the bone,
which indicates a lack of defluorination and provides additional evidence for high in vivo stability.
In vivo small-animal PET imaging in tumor-bearing mice
The favorable pharmacokinetic properties and high metabolic stability, together with the ability of
[18
F]MICA-302 to selectively target active caspase-3 in vitro led us to evaluate its potential for in vivo PET
imaging of apoptosis in tumor-bearing mice. The pro-apoptotic effect of anti-DR5 monoclonal antibody
conatumumab has previously been established in colorectal cancer xenografts, where it demonstrated an
increased activation of caspase-3/7 in tumors.72, 73 The tumor targeting properties of ABP [ F]MICA-302
were therefore evaluated in COLO205 human colorectal cancer xenografts before and 24 h after a single
18
1
8
dose of conatumumab or IgG1 control, and compared to [ F]MICA-301, which was used as control tracer.
1
8
Tumor uptake of both radiotracers, represented as TACs, is shown in Figure 7. [ F]MICA-301 and
[18
F]MICA-302 demonstrated a peak tumor uptake of 2.8 ± 0.2 %ID/mL at 2.5 min p.i. and 3.2 ± 0.2 %ID/mL
at 5.0 min p.i., respectively (p=0.132), in conatumumab-treated animals followed by a washout of
1
8
18
radioactivity. However, the washout of [ F]MICA-301 from the tumor was faster than for [ F]MICA-302,
3
5
which is consistent with diffusion of the substrate away from tumor apoptotic cells. Accordingly, there
26

Page 27 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
1
8
was no difference in image-derived tumor uptake values (%ID/ml) of [ F]MICA-301 between IgG1-control
1
8
group and conatumumab-treated COLO205 xenografts (p=0.065) (Figure 7). Additionally, [ F]MICA-301
tumor uptake did not significantly change at 24 h after conatumumab treatment when compared to
baseline (p=0.764).
1
8
In contrast, at 60 min p.i. there was a significant increased tumor uptake of [ F]MICA-302 in
conatumumab-treated mice (1.52±0.03%ID/ml) compared to IgG1-control group (1.08±0.30%ID/ml,
p=0.004) and to baseline (1.22±0.31%ID/ml, p=0.028), as seen by higher tumor TACs following treatment
with conatumumab suggesting increased radiotracer retention in treated tumors (Figure 7). The tumor
1
8
uptake in conatumumab-treated mice for [ F]MICA-302 was in the range of previous caspase-3 PET
tracers.24, 74 For example, in a similar study in COLO205 tumor-bearing mice treated with a DR5-targeted
1
8
antibody, in which a radiolabeled isatin sulfonamide analog ([ F]WC-4-116) was used, PET imaging
7
5
demonstrated a tumor uptake of about 1.2%ID/ml at 30 min post-tracer injection. In another study,
[18
F]ICMT-11 PET imaging performed in cyclophosphamide treated 38C13 tumor xenografts showed a
7
4
tumor activity of approximately 0.5%ID/ml at 60 min p.i. When different caspase-3 imaging agents were
evaluated side-by-side, radiotracer uptake in lymphoma tumors of mice treated with etoposide was
1
8
18
24
approximately 1.5 and 2.3%ID/ml for [ F]ML-10 and [ F]C-SNAT, respectively at 90 min p.i. Ex vivo
gamma counting of tumor radioactivity after imaging revealed a significant 1.86-fold increase in
[18
F]MICA-302 tumor uptake in conatumumab versus IgG1-treated tumor-bearing mice (p=0.025). For
F]MICA-301 there was no difference in tumor uptake 60 minutes after radiotracer injection between
[18
1
8
treated and control mice (Figure 7 and Supplementary Table 1). [ F]MICA-302 tumor uptake compares
favorably to those for other apoptosis imaging radiotracers such as [99mTc]HYNIC-annexin V (1.4 to 1.6-fold
1
11
18
18
increase), [ In]C2Am (1.6-fold increase), [ F]ML-10 (1.3-fold decrease), [ F]C-SNAT (1.9 to 2.1-fold
1
8
18
18
increase), [ F]ICMT-11 (2.0-fold increase), [ F]WC-4-116 (1.3-fold increase) or [ F]CP18 (1.6-fold
27

Organic & Biomolecular Chemistry
Page 28 of 67
View Article Online
DOI: 10.1039/C9OB00657E
increase).16, 24, 30, 74-77 Although these results are promising, absolute tumor uptake of [ F]MICA-302 is still
low. This might be related to a less efficient tracer permeation into the cell. We are currently developing
second-generation caspase-3 radiotracers with improved cell permeability, and higher intracellular
accumulation for enhanced tumor signal after therapy.
18
The ex vivo biodistribution profiles of both radiotracers measured at 60 min p.i. in IgG1 and
conatumumab-treated tumor-bearing mice are summarized in Supplementary Table 1. The increase in
[18
F]MICA-302 uptake in conatumumab-treated mice was restricted to the tumors. Radiotracer clearance
was fast yielding low background from non-target organs. Of note was the lower background uptake of
[18
F]MICA-302 in the abdominal region (Supplementary Table 1). This is in clear contrast with the
1
8
18
suboptimal pharmacokinetic profile of [ F]ICMT-11 and [ F]C-SNAT, displaying high abdominal
background in the PET images, which may limit assessment of treatment response to tumors of the upper
thoracic region.24, 28, 74
28

Page 29 of 67
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C9OB00657E
1
8
18
Figure 7. [ F]MICA-301 (a) and [ F]MICA-302 (b) in vivo µPET imaging analysis, and ex vivo tumor uptake at 60 min
p.i. in tumor-bearing mice. COLO205 xenografts were treated with 1.5 mg/kg conatumumab (n=6) or IgG1 (n=6) for
1
8
18
2
4 h and subsequently subjected to 60 min dynamic [ F]MICA-301 or [ F]MICA-302 µPET imaging. Corresponding
coronal µPET/CT images (30-60 minutes summed activity) of IgG1 and conatumumab-treated COLO205
tumor-bearing mice are shown in the panels. Tumor time-activity curves and ex vivo tumor uptake at 60 min p.i. are
presented as mean ± SD. **p=0.004, significantly different from IgG1-treated; *p<0.05, significantly different from
baseline and from IgG1-treated.
1
8
We further assessed the relationship between [ F]MICA-302 uptake and apoptosis levels in the tumors.
Histological examination of tumor sections showed that conatumumab treatment increased apoptosis
29