Targeted and modular architectural polymers employing bioorthogonal chemistry for quantitative therapeutic delivery

Article abstract of DOI:10.1039/d0sc00078g

There remain several key challenges to existing therapeutic systems for cancer therapy, such as quantitatively determining the true, tissue-specific drug release profile in vivo, as well as reducing side-effects for an increased standard of care. Hence, it is crucial to engineer new materials that allow for a better understanding of the in vivo pharmacokinetic/pharmacodynamic behaviours of therapeutics. We have expanded on recent "click-to-release" bioorthogonal pro-drug activation of antibody-drug conjugates (ADCs) to develop a modular and controlled theranostic system for quantitatively assessing site-specific drug activation and deposition from a nanocarrier molecule, by employing defined chemistries. The exploitation of quantitative imaging using positron emission tomography (PET) together with pre-targeted bioorthogonal chemistries in our system provided an effective means to assess in real-time the exact amount of active drug administered at precise sites in the animal; our methodology introduces flexibility in both the targeting and therapeutic components that is specific to nanomedicines and offers unique advantages over other technologies. In this approach, the in vivo click reaction facilitates pro-drug activation as well as provides a quantitative means to investigate the dynamic behaviour of the therapeutic agent.

Full text of DOI:10.1039/d0sc00078g

Chemical
Science
View Article Online
View Journal
EDGE ARTICLE
Targeted and modular architectural polymers
employing bioorthogonal chemistry for
Cite this: DOI: 10.1039/d0sc00078g
quantitative therapeutic delivery†
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
abc
Gayathri R. Ediriweera,abc Joshua D. Simpson,
Adrian V. Fuchs,abc
Taracad K. Venkatachalam,abc Matthias Van De Walle,e Christopher B. Howard,
bd
e
abc
Stephen M. Mahler,bd James P. Blinco,
Nicholas L. Fletcher,
abc
abc
abc
*
Zachary H. Houston,
Craig A. Bell
and Kristofer J. Thurecht
There remain several key challenges to existing therapeutic systems for cancer therapy, such as
quantitatively determining the true, tissue-specific drug release profile in vivo, as well as reducing side-
effects for an increased standard of care. Hence, it is crucial to engineer new materials that allow for
a better understanding of the in vivo pharmacokinetic/pharmacodynamic behaviours of therapeutics. We
have expanded on recent “click-to-release” bioorthogonal pro-drug activation of antibody-drug
conjugates (ADCs) to develop a modular and controlled theranostic system for quantitatively assessing
site-specific drug activation and deposition from
a nanocarrier molecule, by employing defined
chemistries. The exploitation of quantitative imaging using positron emission tomography (PET) together
with pre-targeted bioorthogonal chemistries in our system provided an effective means to assess in real-
time the exact amount of active drug administered at precise sites in the animal; our methodology
introduces flexibility in both the targeting and therapeutic components that is specific to nanomedicines
and offers unique advantages over other technologies. In this approach, the in vivo click reaction
facilitates pro-drug activation as well as provides a quantitative means to investigate the dynamic
behaviour of the therapeutic agent.
Received 6th January 2020
Accepted 25th February 2020
DOI: 10.1039/d0sc00078g
rsc.li/chemical-science
bioorthogonal reactions and their ability to provide direct
control over in vivo drug release offers the potential to develop
Introduction
more personalised and effective treatment platforms with
reduced side effects for patients.12–14
Bioorthogonal click chemistry has gained signicant interest in
recent years as a versatile and promising tool for disease diag-
nosis and treatment.1–6 The recent expansion of the eld with
“click-to-release” pro-drug activation has created a surge in
approaches for applications in therapeutic delivery systems to
release drug payloads selectively at the site of interest.7–11 The
minimal biological interference or competition caused by the
The bioorthogonal pro-drug activation was rst explored
with Staudinger ligation as well as 1,3-dipolar cycloaddition
with azide and trans-cyclooctene (k2 z 0.11–0.027 Mꢀ1 sꢀ1).15–17
However, the suboptimal reaction rate continues to be a chal-
lenge when such systems are applied in vivo.18 Alternatively, the
bioorthogonal “click-to-release” pro-drug activation through
the reaction between tetrazine (Tz) and trans-cyclooctene (tCO)
via extremely rapid Inverse Electron-Demand Diels–Alder
(IEDDA) reaction (k2 z 104 Mꢀ1 sꢀ1) offers a more promising
platform for highly selective therapeutic delivery.10 Pioneered
for in vivo systems predominantly by Robillard et al.,8 this
approach generally involves the activation of ADCs bound to the
tumour site through IEDDA reaction to release the chemother-
apeutic. The reaction of a radiolabelled Tz with antibody-bound
tCO containing carbamate-linked doxorubicin (DOX) at the
allylic position affords 1,4-dihydropyridazine, which then
follows an electron cascade mechanism that triggers the liber-
ation of therapeutically active DOX and CO2 (Fig. 5E).8 The
extracellularly-liberated drug can then diffuse into the cancer
cells19 and kill them, resulting in successful pre-targeted
aCentre for Advanced Imaging, The University of Queensland, Brisbane, QLD, 4072,
Australia. E-mail: k.thurecht@uq.edu.au
bAustralian Institute for Bioengineering & Nanotechnology (AIBN), The University of
Queensland, Brisbane, QLD, 4072, Australia
cARC Centre of Excellence in Convergent Bio-Nano Science and Technology, ARC
Training Centre for Innovation in Biomedical Imaging Technology, The University of
Queensland, Brisbane, QLD, 4072, Australia
dARC Training Centre for Biopharmaceutical Innovation, The University of
Queensland, Brisbane, QLD, 4072, Australia
eSchool of Chemistry, Physics and Mechanical Engineering, Queensland University of
Technology, 2 George St, Brisbane, QLD 4000, Australia
† Electronic supplementary information (ESI) available: Synthetic procedures and
compound characterization data; procedures for HBP-tCO and HBP-cCO reactions
with Tz; BsAb production procedure; in vitro and in vivo procedures. See DOI:
10.1039/d0sc00078g
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
therapy. In efforts to further optimise ADC systems, the same synthetic methods. More importantly, a valid way of quantifying
group then developed a rapidly clearing diabody based ADC the amount of drug released by the click reaction or monitoring
system that produced higher tumour accumulation when the in vivo pharmacokinetics of the therapeutic agent would
investigated via single-photon emission computed tomography represent a signicant addition to the currently available bio-
(SPECT).9 This ADC incorporated a PEG24 linker and tCO-linked orthogonal-click toolkit, and offer insight into how to better
drug monomethyl auristatin E (MMAE) that can trigger drug design next generation nanomedicines.
release upon IEDDA with tetrazine.9 Moreover, additional
Herein, we report the development of a novel polymeric
studies have also focussed on developing Tz reaction partners “click-to-release” theranostic system utilising bioorthogonal
that enable rapid decaging,20 as well as investigating mecha- IEDDA reaction between Tz and tCO as the drug release stim-
nistic features of the reaction under varying physiological ulus. Polymeric systems introduce key concepts that offer
conditions to achieve improved efficiency.21 Pre-targeted radi- advances over existing ADC methodology, including the
oimmunotherapy using the bioorthogonal IEDDA reaction potential to incorporate a greater number of drug molecules per
between Tz and tCO has also been attempted using anti- nanocarrier (limited only by the chemistry of attachment), rapid
bodies.22,23 Nevertheless, the longer circulating time of anti- and effective modulation of targeting ligand using bispecic
bodies remains a challenge due to higher systemic radiation antibodies, as well as intrinsic modularity of nanomedicines,
exposure. Despite the progress that has been made so far, the where any combination of drug/imaging agent/diagnostic can
use of this bioorthogonal selective cleavage in living systems for be incorporated in a facile manner. This methodology simul-
cancer therapy is still in its infancy, and several challenges taneously facilitates temporal and spatial quantitative analysis
remain. For instance, there are a limited number of tCO of drug activation and release in vivo using PET-CT (Fig. 1);
attachments that can be conjugated directly to an antibody crucial information that is difficult to accurately obtain using
without interfering with its functionality, and immunogenic conventional theranostic systems and that can signicantly
responses can be triggered in vivo thereby limiting the enhance fundamental understanding of complex drug delivery
bioavailability, and consequently, therapeutic efficacy.3 systems. PET-CT is an extremely sensitive and non-invasive
Furthermore, the synthetic challenges associated with tCO quantitative imaging technique enabling quantitative tracking
derivatives continues to be a drawback for its widespread of biomolecules conjugated with radioisotopes in vivo.24–26 Thus,
adoption, highlighting the need for developments in feasible the use of PET-CT with “click-to-release” pro-drug activation
Fig. 1 (A) The theranostic timeline demonstrating an overview of the new HBP based bioorthogonal approach for quantitative therapeutic
delivery. (B) “Click-to-release” activation of an anti-cancer drug on a tumour cell surface from a HBP-drug conjugate using bioorthogonal IEDDA
reaction. (C) Real-time monitoring of drug release with PET-CT followed by drug release quantification (BsAb ¼ bispecific antibody).
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
enables us determine the amount of activated therapeutic agent infusion (Fig. 2C and E) suggesting predominantly renal clear-
that is delivered to each tissue (or voxel) in vivo via measuring ance mechanism.
retention of the radiolabelled substance in each tissue. Hence,
Sluggish clearance through the gastrointestinal (GI) tract was
this novel approach facilitates previously unattainable under- also observed during the biodistribution study (4.7 ꢂ 0.8%ID gꢀ1 at
standing of the mechanistic behaviour of a polymeric drug 2 h post-administration; Table S3†). In a similar study conducted by
delivery system in vivo with the aid of exogenous chemistry, Zeglis et al., a comparable trend was reported which was suggested
thereby providing a powerful tool to study the in vivo fate of to be caused by the presence of the Tz-NOTA moiety driving
nanomedicine. The use of a stealthy poly(ethylene glycol) (PEG) gastrointestinal clearance.31 To examine this phenomenon and to
based polymeric vector, in particular, a PEGylated hyper- understand the effect of Tz on the clearance behaviour of the
branched polymer (HBP) in the system facilitates high func- molecule, a second biodistribution study was conducted using
tional density while minimizing immunogenic responses.27,28
a
[64Cu]MeO(methoxy)-PEG6-NOTA 10 conjugate without Tz
In this study, we utilise novel bispecic antibodies to non- (Fig. 2B), and the results were compared to the lead compound 8.
covalently bind both target receptor on the cell surface as well as The [64Cu]MeO-PEG6-NOTA (n ¼ 3, 200 mL, ꢁ3 MBq) conjugate
the methoxyPEG (mPEG) that makes up the nanocarrier.29 In exhibited rapid clearance into the kidneys and showed lower
this way, we demonstrate that different targets can be assessed clearance through the GI tract (2.8 ꢂ 1.0%ID gꢀ1) (Table S3†),
for their utility with “click-to-release” technologies in a facile consistent with the clearance of the Tz containing molecule being
manner without requirement for extensive conjugation/puri- caused by the Tz-NOTA moiety (Fig. 2D and F). In the absence of Tz,
cation chemistries.
the [64Cu]MeO-PEG6-NOTA molecule showed fast clearance and
retention in the kidneys, whereas the [64Cu]Tz-PEG4-NOTA mole-
cule is rapidly cleared into the bladder. We assume this change to
be the effect of several factors, including a slight difference in
molecular structure, change in hydrophilicity and charge that are
Results and discussion
Tz small molecule activator
We rst sought a suitable Tz reaction partner capable of suggested in similar studies reported previously.31,32 However, given
remaining in the bloodstream long enough for the click reac- the rapid blood clearance proles of both molecules, these mate-
tion to occur, but which would clear fast enough to avoid rials were ideal for further investigation in the study.
interference from unreacted molecules during imaging and the
subsequent quantitation of drug activation. In brief, a short
dual functionalised PEG linker (COOH-PEG4-NHBOC (5); BOC
The tCO nanocarrier pre-targeted molecule
¼ tert-butyloxycarbonyl) was rst synthesised and attached to The second component of the system, the tCO reaction partner
commercially available (4-(1,2,4,5-tetrazin-3-yl)phenyl)methan- for the “click-to-release” theranostic system, was required to
amine hydrochloride (Tz) via an amidation reaction. The meet two conditions: the molecule must contain two functional
incorporation of a PEG spacer into the Tz carrier enhances the handles to facilitate attachment to the HBP as well as the anti-
biocompatibility of the molecule, prolongs blood circulation, cancer drug; and the drug should be positioned allylic with
and minimizes the non-target tissue uptake of the molecule.30 respect to the alkene functionality of tCO, in order to facilitate
The bifunctional 64Cu chelator, p-SCN-Bn-NOTA, was incorpo- self-immolative drug release upon reaction with Tz. A tCO
rated into the Tz-PEG4-NHBOC molecule upon deprotection of molecule that meets the above requirements was synthesised
the amine group, yielding 25% of pure Tz-PEG4-NOTA 7 as rst by optimising a detailed protocol for the synthesis of 4-
a pink coloured solid upon purication by reversed-phase (RP) cyclooctene-1-carboxylic acid 12, and then by following
C18 HPLC (Fig. 2A and S1†). The serum stability of the Tz a procedure modied from that reported by Rossin et al. to
component in Tz-PEG4-NOTA was tested by measuring the synthesise NHS-tCO-DOX (Fig. 3A).8
change in UV-visible absorbance at 525 nm in a 1 : 1 mixture of
One major highlight in our tCO synthesis is the extremely
human serum : water that showed >85% stability aer 4 h low cost involved in the synthesis of 4-cyclooctene-1-carboxylic
(Fig. S2†), similar to the stability reported in the literature for acid 12, compared to the commercially-available product. In
comparable Tz molecules.31 Aerwards, the in vivo bio- brief, the synthesis of 4-cyclooctene-1-carboxylic acid was
distribution and the pharmacokinetic prole of Tz-PEG4-NOTA carried out in a 3-step reaction procedure starting with the
was investigated using PET-CT. The Tz-PEG4-NOTA molecule electrophilic addition of hydrogen bromide to the starting
was radiolabelled via reaction with 64Cu, and the radiochemical compound 1,5-cyclooctadiene, and followed by Grignard reac-
purity was found to be $99% by radiographic thin-layer chro- tion and subsequent CO2 bubbling to obtain a carboxylic acid
matography (TLC; Fig. S3†). Dynamic PET-CT data was subse- functionality with a 65% yield (Fig. 3A). Employing 30 mL of the
quently acquired over a period of 2 h following administration 1,5-cyclooctadiene precursor compound (20 AUD for 30 mL) we
of [64Cu]Tz-PEG4-NOTA in PBS (200 mL, ꢁ3 MBq) to healthy were able to synthesise ꢁ20 g of compound 12 (<$1 AUD per
female nude mice (n ¼ 3) via intravenous injection in the lateral 0.01 g), which is extremely cost-efficient compared to the
tail vein. Analysis of acquired preclinical PET-CT imaging data commercially-available compound ($157 AUD per 0.01 g). This
(expressed as percentage-injected dose per gram (%ID gꢀ1)), 4-cyclooctene-1-carboxylic acid molecule was then used for the
displayed a rapid clearance prole with a short biological half- synthesis of an activated bis-NHS derivative of tCO adapted
life. Critically, most of the Tz radioligand was cleared from from Rossin et al. with minor modications (Fig. 3A).8 Briey, in
blood circulation into the bladder within 20 min following the rst instance, a methyl group was introduced near the
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Fig. 2 General scheme for the synthesis of (A) Tz-PEG4-NOTA and (B) MeO-PEG6-NOTA. Biodistribution and PET imaging of (C) [64Cu]Tz-PEG4-
NOTA and (D) [64Cu]MeO-PEG6-NOTA in healthy nude mice (H: heart, K: kidney, B: bladder). Pharmacokinetic profile for the clearance of (E)
[
64Cu]Tz-PEG4-NOTA and (F) [64Cu]MeO-PEG6-NOTA from lungs, kidneys, heart, liver and bladder. Reagents and conditions: (a-1) Et3N,
CH3SO2Cl, CH2Cl2 (2) NaN3, DMF, 75% yield (b) 5% HCl, PPh3, diethyl ether, 63% yield (c) Et3N, THF, BOC anhydride, 91% yield (d) PPh3, THF, 31%
yield (e) succinic anhydride, CH2Cl2, 71% yield (f-1) EDC, NHS, CH2Cl2 (2) (4-(1,2,4,5-tetrazin-3-yl)phenyl)methanamine hydrochloride, Et3N, 90%
yield (3) TFA, CH2Cl2, 88% yield (g) p-SCN-Bn-NOTA$3HCl, DMSO, Et3N, 25% yield. (h) p-SCN-Bn-NOTA$3HCl, DMF, Et3N, 12% yield.
carboxylic acid functionality to prevent epimerization during positioned allylic to the alkene group in the tCO ring.8 The
the hydrolysis of the lactone ring in the later stage of the resulting compound was then subjected to iodolactonization
synthesis. This methylation step also permits the regio-selective using potassium iodide and iodine followed by the hydrogen
conjugation of the anti-cancer drug to the functional handle iodide removal to yield (Z)-1-methyl-7-oxabicyclo[4.2.2]dec-4-en-
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
Fig. 3 (A) General scheme for the synthesis of bis-NHS derivatives of tCO and cCO. 1H NMR (CDCl3, 500 MHz) and mass spectrum (inset) of (B)
bis-NHS derivative of tCO and (C) bis-NHS derivate of cCO. [1H NMR: tCO- #-methyl group (1.27 ppm), *-NHS groups (2.84 ppm), a,b-alkene
peaks (6.07 & 5.62 ppm) respectively, c-allylic proton (5.28 ppm)/cCO- #-methyl group (1.25 ppm), *-NHS groups (2.84 ppm), a,b-alkene peaks
(5.92 & 5.75 ppm) respectively, c-allylic proton (5.45 ppm)].
8-one 15 containing a lactone ring with an alkene functional which is more reactive with Tz than the equatorial.8,10 Both the
group. The hydrolysis of this lactone with ring-opening gave the carboxylic and hydroxyl functionalities of the tCO molecule
cyclooctene 16 with cis-methyl ester functionality relative to the were then activated to the bis-NHS derivative 18 using N,N0-
hydroxyl group. The resulting compound was subjected to UV disuccinimidyl carbonate. The cis-cyclooctene (cCO) isomer 19
irradiation at 254 nm in order to obtain tCO with two possible with two activated NHS groups was also synthesised for use as
isomers having hydroxyl functionality axial and equatorial to a control. In both cases, 1H NMR indicated the successful
the methyl ester.8 Photo-isomerization is the most widely used synthesis with the appearance of a singlet peak at 2.84 ppm with
method for the conversion of cis-cyclooctene to trans-cyclo- an integration value of 8 that corresponds to the 8 protons from
octene. The Fox group rst introduced the closed-loop photo- two NHS groups. The alkene peaks in 1H NMR were observed at
chemical reactor that can produce tCO effectively in gram 6.07 ppm and 5.62 ppm for tCO (Fig. 3B) and 5.92 ppm and 5.75
quantities via UV irradiation.33 Despite the effectiveness of the ppm for cCO (Fig. 3C) with expected masses also conrmed by
method, the requirement of an advanced high-cost ow chem- mass spectroscopy.
istry photoreactor is undoubtedly a pitfall associated with this.
Herein, we report an alternative and more accessible “no-ow”
photochemical procedure to synthesise tCO (Fig. S4†). The cis-
The HBP and the polymer–drug conjugate
methyl ester cyclooctene compound 16 was dissolved in
a mixture of heptane : tert-butyl methyl ether (TBME) (4 : 1) and
mixed with methyl benzoate as an initiator to the photo-isom-
erization reaction. The sample was irradiated with 254 nm UV
light in quartz cuvettes for 4 h, resulting in 20% conversion of
The nal part of the system was the design and synthesis of the
polymeric vector molecule and characterization using stan-
dardized reporting methods.34 Compared to the ADC-type
systems commonly reported, the use of a PEGylated nano-
particle as the drug carrier can add stealth properties while
reducing immunogenicity and enhance multifunctionality or
the structural versatility of the theranostic system.30,35 Our
group has previously reported the ability to use PEGylated HBPs
as the polymeric vector in developing targeted nanomedicines
for cancer treatment.36–38 The use of a hyperbranched polymer
as the polymeric drug delivery vector can increase the drug
loading capability due to its branched structure and facile
1
cis to trans as conrmed by H NMR. The tCO product 17 was
then isolated by purication through column chromatography
using silver nitrate impregnated silica gel (10% wt) as the
stationary phase. In this way, the unconverted cis-product 16
was effectively recovered and stored for later usage. The ester
group on the isolated tCO molecule 17 was then base hydro-
lysed, which also facilitated the isolation of the axial isomer,
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
modication chemistries.39 Additionally, the branched struc- spectroscopic analysis, every 12 polymers contained one Cy5
ture provides multifunctionality;27 allowing for the incorpora- molecule (3 ¼ 177 207 Mꢀ1 cmꢀ1, l ¼ 647 nm; Fig. S6†).
tion of therapeutic and targeting agents, as well as imaging
probes within a singular platform.
In order to avoid the reaction of trithiocarbonate RAFT end
groups with amines via aminolysis upon the deprotection of
The HBP 21 was successfully synthesised by following revers- amine groups on the polymer, the trithiocarbonate function-
ible addition–fragmentation chain-transfer (RAFT) polymeriza- ality of the HBP 21 was removed through the Michael addition
tion technique that is well-established to form controlled with PEGMA in the presence of tris(2-carboxyethyl)phosphine
polymers.40,41 Poly(ethylene glycol) monomethyl ether methacry- hydrochloride (TCEP$HCl) and hexylamine, producing HBP 22.
late (PEGMA) and Cy5 methacrylamide were used as monomers The addition of TCEP$HCl aids in preventing disulde forma-
for biocompatibility and uorescence detection respectively, tion, and it simultaneously acts as a catalyst for Michael addi-
ethylene glycol dimethacrylate (EGDMA) as the branching agent tion.42 UV-vis spectroscopic analysis of 22 conrmed an 80%
and a BOC-protected amine functional RAFT agent was employed decrease of the trithiocarbonate group absorbance at 309 nm,
to produce HBPs of controlled size and architecture while facili- while keeping the Cy5 moieties of the polymer intact (Fig. S7†).
tating the incorporation of amine-functional end-groups for post- The subsequent SEC analysis did not display any evidence for
synthesis modication (Fig. 4A).40 The molecular weight of one disulde coupling. The amine groups of the HBP 22 were then
polymer arm was found to be 11.5 kDa by 1H NMR, and one deprotected by reacting with 20% (v/v) TFA : CH2Cl2. The
polymer molecule was found to contain approximately 4 arms via successful deprotection was conrmed by the complete disap-
the determination of the absolute molecular weight of polymer pearance of the BOC group at 1.45 ppm in 1H NMR while
through SEC-MALLS (Table S1 and Fig. S5†). As per the UV-vis generating the amine functionality in 23.
Fig. 4 (A) General synthetic scheme for the base HBP (21). Synthesis, characterization (1D DOSY – 50% gradient, UV-vis spectrum) and selected
physicochemical properties of (B) HBP-tCO-DOX and (C) HBP-cCO-DOX conjugates. aDetermined using SEC-MALLS, bdispersity obtained by
SEC analysis, ccalculated by dividing the absolute molecular weight of HBP from SEC-MALLS from the molecular weight of one polymer arm
from 1H NMR, ddetermined using a calibration curve drawn for free DOX.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
In order to validate the click reaction between HBP bound enables the binding of nanoparticles to the BsAbs through non-
tCO and Tz, a HBP-tCO conjugate (S2) was prepared by reacting covalent interactions that can be formed by simply mixing the
deprotected HBP 23 with commercially-available tCO-NHS in two components. Herein, two BsAbs were developed to contain
the presence of triethylamine (Et3N), along with HBP-cCO an anti-PEG component that can form strong non-covalent
conjugate (S3) as a control. Generally, Tz does not undergo any interactions with PEGylated nanoparticles. The other antigen-
reaction with cCO under physiologically-relevant timeframes, binding site of the BsAb were designed to exhibit specicity for
where the rate of the reaction is extremely slow.43 The conju- the tumour associated glycoprotein 72 (TAG72) receptor or the
gates were reacted with excess (ꢃ10) Tz-PEG4-NHBOC in 3% epidermal growth factor receptor (EGFR). The rst receptor was
ꢄ
acetonitrile in PBS at 37 C for 1.5 h, and the reaction mixture chosen specically for this study as it is reported to be non-
was puried through SEC. The analysis of product fractions internalising, making it suitable for a pre-targeting study due to
1
through H NMR indicated the successful reaction of tetrazine retention at the cell periphery8,50 and the latter was chosen as it
with tCO by the appearance of new peaks in the aromatic region is a commonly reported internalising receptor.29 To the best of
(7–8.5 ppm) from tetrazine and reduction of alkene peak (5.8 our knowledge, this is the rst report of using anti-PEG/anti-
ppm) from tCO (Fig. S8†). Conversely, no signicant change in TAG72 BsAbs, as well as employing them as ligands in a bio-
1H NMR was observed in HBP-cCO conjugate conrming the orthogonal “click-to-release” based theranostic system.
absence of reaction of tetrazine with cCO (Fig. S9†).
The polymer–drug conjugates were investigated with anti-
The nal polymer–drug conjugates for the “click-to-release” PEG/anti-TAG72 and anti-PEG/anti-EGFR BsAbs against
system were prepared by reacting both tCO compound 18 and MCF7 51,52 and MDA-MB-468 20,53 breast cancer cell lines,
cCO compound 19 separately with an equimolar amount of respectively, that are known to over-express these non-inter-
doxorubicin hydrochloride for 18 h, followed by conjugation to nalising TAG72 and internalising EGFR antigens. To determine
HBP 23 under basic conditions. Following purication, the the degree of cellular association of the targeted and untargeted
successful formation of HBP-tCO-DOX 24 and HBP-cCO-DOX 25 (with and without BsAbs, respectively) HBP-tCO-DOX and HBP-
conjugates was conrmed by 1D DOSY NMR, UV-vis spectros- cCO-DOX conjugates, ow cytometry analyses were performed
copy (Fig. 4B and C), and high-performance liquid chromatog- using the MCF7 and MDA-MB-468 breast cancer cell lines
raphy (HPLC) (Fig. S12A and S13A†). The nal HBP-tCO-DOX (Fig. S10†). The results indicated an increase in cellular asso-
conjugate was found to contain 3.9 DOX per polymer molecule, ciation in anti-PEG/anti-TAG72 BsAb targeted polymer–drug
and HBP-cCO-DOX appeared to contain 3.5 DOX per polymer conjugates (>46% Cy5+) compared to the untargeted analogues,
molecule based on a calibration curve drawn for free DOX. which conrmed the increase in the interaction of HBP with the
Thus, the drug conjugation efficiency (efficiency of coupling the MCF7 cells in the presence of anti-PEG/anti-TAG72 BsAbs.
modied tCO or cCO molecule) was found to be 93% and 83% Similarly, the anti-PEG/anti-EGFR targeted polymer–drug
for HBP-tCO-DOX and HBP-cCO-DOX, respectively. The similar conjugates indicated a signicantly higher cellular association
DOX loading in both tCO and cCO polymers was important (>75% Cy5+) compared to the untargeted analogues against
when comparing results since the amount of polymer injected MDA-MB-468 cells (Fig. S10†). Hence, these BsAbs and cell lines
per animal was determined by the DOX content.
were conrmed to be suitable for further in vitro and in vivo
The polymer–drug conjugate system was then subjected to in assessments of the theranostic system.
vitro and in vivo assessments to determine utility in quantitative
The in vitro cytotoxicity of the non-drug conjugated HBP 21,
drug imaging. Pre-targeted imaging and therapeutic systems both BsAb associated HBP-tCO-DOX and HBP-cCO-DOX conju-
usually utilise a non-internalising receptor over-expressed on gates in comparison to free DOX was evaluated using an MTS
cancer cells as the in vivo target, in order to ensure the presence of assay with MCF7 and MDA-MB-468 breast cancer cells with a 48
the initially administered targeting component on the cell surface h incubation time.54 The empty vehicle 21, did not show
to react with the subsequently administered chase molecule, signicant cytotoxicity on MCF7 and MDA-MB-468 cells, even at
despite some pre-targeting tumour imaging studies reported that a high concentration of 5 mg mLꢀ1 either when targeted to
have been conducted using internalising receptors.44–46 In this TAG72 and EGFR, respectively, or in the untargeted cases
study, we examined the polymeric “click-to-release” system tar- (Fig. S11†). Both HBP-tCO-DOX and HBP-cCO-DOX conjugates
geted to both internalising and non-internalising cancer recep- displayed only slight cytotoxicity for both cell lines, even at
tors with the purpose of investigating any difference in results higher concentrations and were signicantly lower when
caused by the internalisation of cellular receptors. To this end compared to the free drug (Fig. 5). This is consistent with free
targeting agents were developed for both types of receptors.
DOX being able to freely cross membranous cellular barriers
and interfere with topoisomerase,55 thereby reducing the cell
viability. Conversely, the conjugated form was unable to
undergo the same process. The lower cytotoxicity from poly-
In vitro validation of the polymeric theranostic system
In recent years, bispecic antibodies (BsAbs) have captured mer–drug conjugates is due to being present in the pro-drug
signicant attention as promising agents for the targeted form, thereby highlighting the stability of carbamate linked
delivery of therapeutics in cancer therapy.47–49 Our group has DOX under in vitro conditions. However, when the MCF7 and
utilised BsAbs for developing protein–polymer conjugates that MDA-MB-468 cells exposed to polymer–drug conjugates were
are independent of post-modication chemistries by using incubated with Tz compound 7, the HBP-tCO-DOX treated cells
a simple and highly effective coupling method.29 This approach displayed a signicant reduction in cell viability as opposed to
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
468 cell lines, respectively. The HBP-tCO-DOX + Tz treatment
group achieved not very differentiable toxicity at 0.59 ꢂ 0.06 mg
mLꢀ1 and 0.06 ꢂ 0.06 mg mLꢀ1 for MCF7 and MDA-MB-468 cells
respectively, suggesting a high degree of DOX liberation from
the HBP (Fig. 5D) upon reaction with Tz. Importantly, in both
cases, the introduction of the Tz group to the cis analogue (HBP-
cCO-DOX + Tz) did not lead to signicant toxicity changes from
the vehicle control. These cellular association and cytotoxicity
assessments indicated the absence of any signicant difference
in behaviour of the polymeric theranostic system when inves-
tigated in vitro using both non-internalising and internalising
receptors as cellular targets.
Prior to the in vivo investigations, the in vitro stability of
carbamate-linked DOX in both HBP-tCO-DOX and HBP-cCO-
DOX conjugates were investigated in both PBS and a 1 : 1
mixture of human serum and water at 37 ꢄC, by monitoring the
drug release via RP-HPLC. Both polymer–drug conjugates were
incubated in PBS and a 1 : 1 mixture of human serum and water
separately at 37 ꢄC for 4 days, and the samples were subse-
quently analysed through RP-HPLC. The peak at 9 min was
detected for free DOX whereas the polymer–drug conjugate
peak was detected at 14 min in the uorescence channel
(doxorubicin excitation: 480 nm emission: 580 nm). The drug
release at the nal time point (96 h) was found to be <3% in PBS
and <6% in serum for both conjugates (Table S2†), conrming
the high stability of carbamate-linked DOX, which is in accor-
dance with the literature.8,10 The slight increase in drug release
in human serum compared to PBS might be due to the presence
of different proteins and enzymes that can affect bond stability,
however, given the time frame, this interaction was minor.
Aerwards, the in vitro drug release of the developed system
upon click reaction with Tz was evaluated by quantifying the
amount of DOX release at various time points through RP-
HPLC. The HBP-tCO-DOX and HBP-cCO-DOX were reacted with
Tz-PEG4-NOTA (1 equivalent to tCO/cCO) in PBS at 37 ꢄC. At 45
min, 3 h, 6 h and 24 h time points, fractions of the reaction
mixtures were analysed through RP-HPLC using uorescence
detection for DOX at 480 nm. Aer the click reaction with Tz,
the HBP-tCO-DOX conjugate showed an instantaneous doxo-
rubicin release similar to previous studies using the ADC
approach8,10 with >92% free DOX appearing at 9 min retention
time when calculated through the peak area ratio of free DOX
and HBP-tCO-DOX conjugate aer 45 min (Fig. S12†). Moreover,
even aer 24 h, only <5% DOX release was detected in HBP-cCO-
DOX (Fig. S13†), highlighting the stability and minimal reac-
tivity of this system in the presence of Tz.
Fig. 5 (A) A diagram displaying the cellular association of BsAb tar-
geted polymer with cancer cells (created using BioRender). Cell
viabilities determined by the MTS assay of (B) MCF7 breast cancer cell
line treated with anti-PEG/anti-TAG72 BsAb (C) MDA-MB-468 breast
cancer cell line treated with anti-PEG/anti-EGFR BsAb targeted HBP-
tCO-DOX, HBP-cCO-DOX, HBP-tCO-DOX + Tz and HBP-cCO-DOX
+ Tz and DOX alone at a range of concentrations. (D) IC50 of free DOX
and HBP-tCO-DOX + Tz on MCF7 and MDA-MB-468 cells. (E) Self-
immolative drug release by the IEDDA reaction between tetrazine and
doxorubicin-functionalized tCO (statistical significance has only
shown for DOX alone, HBP-tCO-DOX + Tz and HBP-cCO-DOX + Tz);
ns – not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
In vivo validation of the polymeric theranostic system and
quantication of in vivo drug release
The system was then further validated in vivo to demonstrate
the utility of the novel polymeric “click-to-release” system to
selectively release the anti-cancer drug upon IEDDA reaction,
while showing the ability to merge this approach with PET
imaging to quantify this in vivo drug release. While few studies
have been previously reported with bioorthogonal click chem-
istry-based ADC activation through a radiolabelled probe and
HBP-cCO-DOX treated cells, which suggested that DOX was
being liberated from the HBP-tCO-DOX conjugate upon reac-
tion with Tz, thereby releasing the active drug (Fig. 5B and C).
The IC50 values obtained for free DOX were consistent with
literature for these cell lines,38,56 yielding the values of 0.29 ꢂ
0.05 mg mLꢀ1 and 0.08 ꢂ 0.03 mg mLꢀ1 for MCF7 and MDA MB-
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
quantication of drug release through RP-HPLC analysis of the intravenously into MCF7 tumour-bearing mice (n ¼ 4) as
tumour extracts,8,9 to date there have been no reports of using described for the pre-targeted system. This pre-reacted control
designed nanocarriers to improve pharmacological properties contained radioligand already reacted with the polymer nano-
as theranostics in order to comprehensively investigate the in carrier prior to injection, which is analogous to conventional
vivo behaviour of nanomedicine. Moreover, to the best of our targeted nanomedicines. A signicantly higher ex vivo liver (2.0 ꢂ
knowledge, no studies have combined the two concepts of 0.3%ID gꢀ1) and spleen (2.4 ꢂ 0.9%ID gꢀ1) accumulation was
directly correlating the amount of drug activated and released observed with this conventional system, as well as a lower GI tract
through the click reaction with the radioactive concentration clearance (0.26 ꢂ 0.02%ID gꢀ1) when compared to either tCO or
observed in each tissue. Thus, here we demonstrate the in vivo cCO conjugates described via pre-targeting approach.
“click-to-release” activation of an HBP-drug conjugate using
Interestingly, despite the absence of any signicant differ-
bioorthogonal IEDDA reaction followed by the direct quanti- ence in in vitro data with MCF7 and MDA-MB-468 cell lines, the
cation of active drug release through PET-CT. The in vivo in vivo results demonstrated a clear difference between two
tumour targeting of HBP-tCO-DOX 24 and HBP-cCO-DOX 25 receptor targets, indicating the advantage of using a non-
conjugates and the DOX release upon reaction with [64Cu]Tz- internalising receptor as the in vivo target in pre-targeting
PEG4-NOTA were investigated in female balb/c nu/nu mice studies. Unlike MCF7, in MDA-MB-468 tumour bearing mice,
bearing MCF7 and MDA-MB-468 tumours as orthotopic xeno- similar tumour accumulation was observed for all constructs.
gras. The HBP-tCO-DOX and HBP-cCO-DOX conjugates were Additionally, low ex vivo tumour radioactivity was observed in
incubated with anti-PEG/anti-TAG72 and anti-PEG/anti-EGFR the HBP-tCO-DOX24
treated mice compared to MCF7
h
BsAbs separately (1 : 1 ratio of HBP : BsAb) for 1 h prior to the comparison mice, and no signicant difference was observed in
injections. A 0.5 mg kgꢀ1 DOX dose of the polymer (1 ꢃ dose ¼ comparison with the control, HBP-cCO-DOX (Fig. 6F). However,
0.0167 mmol tCO and cCO) was injected intravenously to each similar to the case of MCF7, a lower ex vivo radioactivity accu-
mouse (n ¼ 4), and at different time points following polymer mulation in the liver (0.57 ꢂ 0.41%ID gꢀ1) and spleen (0.17 ꢂ
injection (8 h, 24 h), an equimolar amount of [64Cu]Tz-PEG4- 0.01%ID gꢀ1) was observed together with a high kidney (2.21 ꢂ
NOTA (1 ꢃ dose ¼ 0.0164 mmol Tz/ꢁ15 MBq) was intravenously 0.29%ID gꢀ1) and GI tract (3.58 ꢂ 0.29%ID gꢀ1) clearance
administered. The PET-CT images were acquired 2 h post-Tz (Fig. 6E).
injection to allow for clearance of [64Cu]Tz-PEG4-NOTA from
circulation, as shown in Fig. 2C and E.
Given the spleen is one organ that plays a signicant role in
nanomedicine clearance, comparison of tumour tissue accu-
In mice bearing MCF7 breast cancer xenogra, the maximum mulation relative to the spleen allowed further comparisons of
ex vivo tumour uptake was observed for HBP-tCO-DOX24 h, and targeting efficacy to be drawn. In mice bearing MCF7 cancer
the minimum uptake was observed for the HBP-cCO-DOX24 h xenogra, the tumour to spleen ratios were highest in HBP-tCO-
control (Fig. 6C), suggesting the successful occurrence of the DOX24 h (3.2 to 1) compared to HBP-cCO-DOX24 h (1.7 to 1),
reaction between tCO and Tz in vivo. One important observation congruent with the enhanced reactivity of Tz to tCO compared
from this study was that the tumour signal in mice treated with to cCO (Fig. 6A). For the HBP-tCO-DOX8 h, the tumour to spleen
tCO polymer remained relatively low compared to conventional ratio was 1.9 to 1, likely due to the shorter time frame providing
tumour targeting approaches (without pre-targeting), likely as less opportunity for the polymer to accumulate at the tumour
a result of the insufficient tumour penetration of [64Cu]Tz-PEG4- site compared to 24 h, which was further evident through the
NOTA to facilitate reaction with tCO, due to its fast clearance. A higher blood to spleen ratio of HBP-tCO-DOX8 h (3.4 to 1; Table
high ex vivo kidney (2.98 ꢂ 0.36%ID gꢀ1) and GI tract clearance S4†) than HBP-tCO-DOX24 h (2.6 to 1; Table S4†). This indicates
(3.91 ꢂ 0.63%ID gꢀ1) was observed indicating the rapid clearance that the level of HBP circulating in the blood is much higher at 8
of the radiolabelled Tz linker as previously described (Fig. 2). This h than 24 h, possibly reducing the available Tz to react with
result reected previous observations from an ADC system re- tumour bound tCO due to the reaction with tCO in the circu-
ported by Rossin et al., where measured %ID gꢀ1 were much lating HBP. However, due to the higher spleen uptake in the pre-
lower than standard antibody systems.8 This observation can also reacted conventional system compared to the other formula-
be accounted for the administration of a lower dose (0.5 mg kgꢀ1 tions, the tumour to spleen ratio was 0.24 (Fig. 6A), which
DOX dose of the HBP) in the present proof-of-concept study suggests the nanocarrier itself shows signicant accumulation
compared to the generally administered therapeutic dose in in the spleen, however this is not accessible to circulating Tz in
similar therapeutic studies9,38 (ꢁ5 mg kgꢀ1 DOX dose of poly- the pre-targeted samples, reducing delivery of DOX to this
mer). However, the accumulated amount of ex vivo radioactivity tissue. Furthermore, in mice bearing MDA-MB-468 cancer
in the liver (0.9 ꢂ 0.4%ID gꢀ1) and spleen (0.23 ꢂ 0.02%ID gꢀ1
)
xenogra, no signicant difference between tumour to spleen
were commensurately low resulting in less non-specic drug ratio for HBP-tCO-DOX24 (1.7 to 1) and control HBP-cCO-
h
release, thereby indicating a clear benet of this approach by DOX24 h (1.8 to 1) was not observed (Fig. 6D) further indicating
reducing off-target toxicity issues (Fig. 6B). Further validation of the appropriateness of utilising a non-internalising receptor as
the advantage of using a pre-targeting approach was demon- the in vivo target over an internalising receptor. Aerwards, to
strated by administering a mixture of anti-PEG/anti-TAG72 highlight the increase in in vivo tumour accumulation with
associated, pre-reacted HBP-tCO-DOX (0.5 mg kgꢀ1 DOX dose, 1 respect to background signal, a calculated PET-CT image was
ꢃ dose ¼ 0.0167 mmol tCO) and [64Cu]Tz-PEG4-NOTA (1 ꢃ dose ¼ created for MCF7 mice where the activity measured in the
0.0164 mmol Tz/ꢁ15 MBq) as a control. This complex was injected animal was normalised to the signal from spleen as a reection
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Fig. 6 In vivo PET-CT and ex vivo results of the preliminary theranostic study. (A) Ex vivo tumour to spleen ratio for MCF7 mice based on the %ID
gꢀ1 obtained from gamma counter analysis. Ex vivo radioactivity accumulation in the (B) liver, spleen, kidneys and GI tract and (C) tumour
obtained from the gamma counter analysis for MCF7 mice. (D) Ex vivo tumour to spleen ratio for MDA-MB-468 mice based on the %ID gꢀ1
obtained from gamma counter analysis. Ex vivo radioactivity accumulation in the (E) liver, spleen, kidneys and GI tract and (F) tumour obtained
from the gamma counter analysis for MDA-MB-468 mice. (G) Calculated PET-CT images (tumour area zoomed in below) from the in vivo PET-
CT study showing the increase in tumour radioactivity accumulation in 24 h tCO MCF7 mouse due to click reaction (mice in image are supine
viewed from top-down with the tumour in the left, and the scale bar contains arbitrary units due to the normalization of the radioactivity). Left –
HBP-tCO-DOX24 h ([64Cu]Tz-PEG4-NOTA administered after a 24 h pre-targeting interval from BsAb-targeted HBP-tCO-DOX injection); Middle
– HBP-cCO-DOX24 h ([64Cu]Tz-PEG4-NOTA administered after a 24 h pre-targeting interval from BsAb-targeted HBP-cCO-DOX injection);
Right – Pre-reacted (anti-PEG/anti-TAG72 BsAb-targeted HBP-tCO-DOX and [64Cu]Tz-PEG4-NOTA were reacted prior to the administration
and imaged after 28 h). (H) Confocal micrographs demonstrating the distribution of the Cy5 labelled HBP and the diffusion of DOX in HBP-tCO-
DOX24 h MCF7 tumour tissue (marked) and the restriction of DOX to blood vessels in control cCO (marked) (cyan – Cy5 from HBP, magenta –
DOX and green – vascular stain; scale bar – 100 mm).
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
of background signal (Fig. 6G & calculation shown in ESI† as compared to conventional nanomedicines. Our proof-of-
E1), and is plotted from the post-mortem data in Fig. 6A. concept study demonstrates the potential use of a polymer in
Matching the prior observations discussed, the biodistribution a “click-to-release” system in addition to the commonly used
of [64Cu]Tz-PEG4-NOTA demonstrated rapid clearance through ADC systems. Of greater importance to the further development
the bladder. Hence, for the clear visualisation of the tumour of this promising technology, the use of a HBP offers a multi-
signal, the calculated image was masked to remove bladder functional platform, providing more opportunity to increase the
signal. The image with bladder is included in the ESI (Fig. S15†). efficacy of the system and tailor for specic applications. For
This was not required for cCO and pre-reacted groups due to the instance, by employing an alternative modication of the
increased distance from tumour to bladder. A clear increase in residual RAFT end group of the HBP to form an amine func-
64Cu signal intensity in the MCF7 tumour was observed in the in tionality, it would be possible to increase the drug conjugation
vivo PET-CT images for the HBP-tCO-DOX24 h mice compared to by a factor of two or incorporate alternate therapeutics, greatly
the control HBP-cCO-DOX24 h (Fig. 6G) in accordance with the ex enhancing the drug loading of these systems. We also demon-
vivo results (Fig. 6A). Based on the observed accumulation of strated for the rst time the use of anti-PEG/anti-TAG72 BsAbs
[64Cu]Tz-PEG4-NOTA, the amount of DOX release in the tumour as effective targeting ligands to MCF7 human breast cancer
was then quantied using the calculation E2 described in ESI.† cells. The simple mixing of the polymer–drug conjugate with
For the MCF7 tumours in mice treated with HBP-tCO-DOX24 h
,
the BsAbs can facilitate the binding of BsAbs with the polymer
the average released DOX amount was calculated to be 4.4 ꢃ without the need for any conjugation reactions, leaving func-
10ꢀ5 mg of DOX per gram of tumour. Subtracting the amount of tional groups free for conjugation with other agents and
tracer accumulation in the control (HBP-cCO-DOX24 h), it was moieties. The study also draws an interesting comparison
assumed that this calculated DOX amount is due to the self- between using an internalising (EGFR) and a non-internalising
immolative drug release upon click reaction of tCO and Tz.
(TAG72) receptor as the target and demonstrates the advantage
Given the results of our in vivo experiments, our observations of using a non-internalising receptor for pre-targeting systems.
highlight the advantages of the pre-targeting approach as well as The non-internalising receptor targeted nanomedicine was able
the use of a non-internalising receptor as the in vivo target over an to achieve a 1.9-fold better tumour to background tissue ratio
internalising receptor. Compared to the pre-reacted system, our compared to the internalising receptor. Importantly, rapid
“click to release” pre-targeting signicantly reduces the amount switching between these two targets is facilitated using minimal
of DOX that would be released in the clearance organs (such as modulation of the polymer backbone by employing bispecic
liver and spleen), which has the benet of minimising off-target antibody technology. This comparison highlights the impor-
toxicity observed in current chemotherapeutic studies. The tance of selecting suitable targets and their applicability in this
observed lower off-target toxicity of the pre-targeting approach approach in any future translation. The preliminary in vivo
means that there is the option to increase the therapeutic dose studies in MCF7 breast cancer xenogra mouse models dis-
administered to mice to the commonly used therapeutic dose of played the release of DOX from HBP-tCO-DOX conjugate upon
5 mg kgꢀ1 DOX in the future, while still reducing the side effects reaction with [64Cu]Tz-PEG4-NOTA. This study was able to yield
to perform a complete anti-tumour assessment to determine the signicantly higher tumour: spleen ratio for tCO compared to
therapeutic efficacy of the system through a tumour regression controls and the conventional formulation. The observed lower
study. Furthermore, under the conditions of our experiment, the signal accumulation in liver and spleen enables reduced non-
release of DOX within the MCF7 tumour was signicantly higher specic DOX release, thereby minimising toxicity to healthy
in tCO compared to the control cCO, ensuring high specicity tissues and addresses one of the key challenges in modern
and selectivity of drug release that can offer a new route for chemotherapy. Most importantly, the method of quantifying
improving the efficacy of nanomedicines.
the actual in vivo drug release offered by this approach can
To expand upon in vivo observations, the MCF7 tumours were potentially serve as a powerful tool in future nanomedicine
xed in 4% paraformaldehyde, sliced via microtome and were developments by aiding the investigation of in vivo pharmaco-
imaged using confocal microscopy. The distribution of DOX was kinetics/pharmacodynamics of therapeutic agents. Beyond
primarily restricted to blood vessels in the case of control HBP- offering mechanistic insight into how nanomedicine formula-
cCO-DOX24 h (Fig. 6H), further conrming the absence of any tions can modulate specic drug molecule distribution in
signicant drug release in the presence of Tz. Conversely, HBP- tumours, this approach creates numerous new avenues for
tCO-DOX24
exhibited DOX distribution within the tumour further research. We propose that the tumour accumulation
h
tissue away from the blood vessels (Fig. 6H), indicating the could be further increased by using a longer-circulating Tz
effective DOX release from the polymer-bound tCO upon click molecule, providing greater time for the reaction to occur in
reaction. This loco-regional delivery of the DOX is then expected vivo. Furthermore, the system can be rapidly expanded for study
to be highly efficacious for controlled, targeted therapeutics.
in other cancer xenogra models using different BsAbs as tar-
geting agents depending on the upregulation of specic
markers. Finally, considering the current interest in the use of
bioorthogonal reactions in drug delivery systems, we anticipate
Conclusions
In conclusion, we have successfully developed a new HBP-based that this pioneering use of HBPs will facilitate the reinforce-
“click-to-release” theranostic system for quantitative thera- ment of “click-to-release” theranostics, spurring research into
peutic delivery, with improved specicity in drug delivery prole improved site-specic delivery and opening up novel avenues
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
for further applications. Moreover, this technology offers
avenues to further probe the mechanisms that lead to enhanced
efficacy of polymer and nanomaterial therapeutics.
H. M. Janssen, A. H. A. M. van Onzen, P. J. Hudson and
M. S. Robillard, Nat. Commun., 2018, 9, 1484.
10 R. M. Versteegen, R. Rossin, W. ten Hoeve, H. M. Janssen
and M. S. Robillard, Angew. Chem., Int. Ed., 2013, 52,
14112–14116.
Ethical statement
11 J. M. Mejia Oneto, I. Khan, L. Seebald and M. Royzen, ACS
Cent. Sci., 2016, 2, 476–482.
All animal experiments were approved by the University of
Queensland's Animal Ethics Committee (AEC530/15) and con- 12 S. S. Kelkar and T. M. Reineke, Bioconjugate Chem., 2011, 22,
formed to the Animal Care and Protection Act Qld and the Code 1879–1903.
of Practice for the care and use of animals for scientic 13 J. Shi, P. W. Kantoff, R. Wooster and O. C. Farokhzad, Nat.
purposes.
Rev. Cancer, 2016, 17, 20.
14 B. Sumer and J. Gao, Nanomedicine, 2008, 3, 137–140.
15 M. Azoulay, G. Tuffin, W. Sallem and J. C. Florent, Bioorg.
Med. Chem. Lett., 2006, 16, 3147–3149.
Conflicts of interest
There are no conicts to declare.
16 S. S. Matikonda, D. L. Orsi, V. Staudacher, I. A. Jenkins,
F. Fiedler, J. Chen and A. B. Gamble, Chem. Sci., 2015, 6,
1212–1218.
17 S. S. Matikonda, J. M. Fairhall, F. Fiedler, S. Sanhajariya,
R. A. J. Tucker, S. Hook, A. L. Garden and A. B. Gamble,
Bioconjugate Chem., 2018, 29, 324–334.
Acknowledgements
The authors acknowledge the facilities, and the scientic and
technical assistance of the National Imaging Facility at the
Centre for Advanced Imaging, University of Queensland. 18 C. S. McKay and M. Finn, Chem. Biol., 2014, 21, 1075–1101.
Funding for this research is acknowledged from the National 19 P. Polakis, Pharmacol. Rev., 2016, 68, 3–19.
Health and Medical Research Council (APP1099321 and 20 X. Fan, Y. Ge, F. Lin, Y. Yang, G. Zhang, W. S. Ngai, Z. Lin,
APP1148582 (K. J. T.); APP1054569 (C. A. B.)) and the Australian
Research Council (LP150100703 (K. J. T.)). This work was per-
S. Zheng, J. Wang, J. Zhao, J. Li and P. R. Chen, Angew.
Chem., Int. Ed. Engl., 2016, 55, 14046–14050.
formed in part at the Queensland node of the Australian 21 J. C. T. Carlson, H. Mikula and R. Weissleder, J. Am. Chem.
National Fabrication Facility (ANFF), a company established
Soc., 2018, 140, 3603–3612.
under the National Collaborative Research Infrastructure 22 R. Rossin, T. Lappchen, S. M. van den Bosch, R. Laforest and
Strategy to provide nano- and microfabrication facilities for M. S. Robillard, J. Nucl. Med., 2013, 54, 1989–1995.
Australia's researchers. This research was conducted and fun- 23 R. Membreno, B. E. Cook, K. Fung, J. S. Lewis and
ded by the ARC Centre of Excellence in Convergent Bio-Nano B. M. Zeglis, Mol. Pharm., 2018, 15, 1729–1734.
Science and Technology (CE140100036) and in part by the ARC 24 R. Boellaard, J. Nucl. Med., 2009, 50(suppl 1), 11s–20s.
Training Centre for Innovation in Biomedical Imaging Tech- 25 B. Bai, J. Bading and P. S. Conti, Theranostics, 2013, 3, 787.
nologies (IC170100035).
26 M. Wang, D. Svatunek, K. Rohlng, Y. Liu, H. Wang, B. Giglio,
H. Yuan, Z. Wu, Z. Li and J. Fox, Theranostics, 2016, 6, 887.
27 A. Ardana, A. K. Whittaker and K. J. Thurecht,
Macromolecules, 2014, 47, 5211–5219.
References
1 E. M. Sletten and C. R. Bertozzi, Angew. Chem., Int. Ed., 2009, 28 N. Boase, I. Blakey, B. Rolfe, K. Mardon and K. Thurecht,
48, 6974–6998. Polym. Chem., 2014, 5, 4450–4458.
2 B. Oliveira, Z. Guo and G. Bernardes, Chem. Soc. Rev., 2017, 29 C. B. Howard, N. Fletcher, Z. H. Houston, A. V. Fuchs,
46, 4895–4950.
N. R. Boase, J. D. Simpson, L. J. Raery, T. Ruder,
M. L. Jones and C. J. de Bakker, Adv. Healthcare Mater.,
2016, 5, 2055–2068.
3 M. K. Rahim, R. Kota and J. B. Haun, Bioconjugate Chem.,
2015, 26, 352–360.
4 J. P. Meyer, P. Adumeau, J. S. Lewis and B. M. Zeglis, 30 K. Knop, R. Hoogenboom, D. Fischer and U. S. Schubert,
Bioconjugate Chem., 2016, 27, 2791–2807. Angew. Chem., Int. Ed., 2010, 49, 6288–6308.
5 J. C. Knight and B. Cornelissen, Am. J. Nucl. Med. Mol. 31 B. M. Zeglis, C. Brand, D. Abdel-Atti, K. E. Carnazza,
Imaging, 2014, 4, 96–113.
6 J. Li and P. R Chen, Nat. Chem. Biol., 2016, 12, 129–137.
B. E. Cook, S. Carlin, T. Reiner and J. S. Lewis, Mol.
Pharm., 2015, 12, 3575–3587.
7 R. M. Versteegen, W. Ten Hoeve, R. Rossin, M. A. R. de Geus, 32 E. Garcia Garayoa, C. Schweinsberg, V. Maes, L. Brans,
H. M. Janssen and M. S. Robillard, Angew. Chem., Int. Ed.
Engl., 2018, 57, 10494–10499.
P. Blauenstein, D. A. Tourwe, R. Schibli and
P. A. Schubiger, Bioconjugate Chem., 2008, 19, 2409–2416.
8 R. Rossin, S. M. van Duijnhoven, W. ten Hoeve, 33 M. Royzen, G. P. A. Yap and J. M. Fox, J. Am. Chem. Soc., 2008,
H. M. Janssen, L. H. Kleijn, F. J. Hoeben, R. M. Versteegen 130, 3760–3761.
and M. S. Robillard, Bioconjugate Chem., 2016, 27, 1697– 34 M. Faria, M. Bjornmalm, K. J. Thurecht, S. J. Kent,
¨
1706.
R. G. Parton, M. Kavallaris, A. P. R. Johnston, J. J. Gooding,
S. R. Corrie, B. J. Boyd, P. Thordarson, A. K. Whittaker,
M. M. Stevens, C. A. Prestidge, C. J. H. Porter, W. J. Parak,
9 R. Rossin, R. M. Versteegen, J. Wu, A. Khasanov,
H. J. Wessels, E. J. Steenbergen, W. ten Hoeve,
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
T. P. Davis, E. J. Crampin and F. Caruso, Nat. Nanotechnol., 46 H. L. Evans, Q.-D. Nguyen, L. S. Carroll, M. Kaliszczak,
2018, 13, 777–785.
35 S. Salmaso and P. Caliceti, J. Drug Delivery, 2013, 2013, 1–19.
F. J. Twyman, A. C. Spivey and E. O. Aboagye, Chem.
Commun., 2014, 50, 9557–9560.
36 A. K. Pearce, J. D. Simpson, N. L. Fletcher, Z. H. Houston, 47 K. Taylor, C. B. Howard, M. L. Jones, I. Sedliarou,
A. V. Fuchs, P. J. Russell, A. K. Whittaker and
J. MacDiarmid, H. Brahmbhatt, T. P. Munro and
K. J. Thurecht, Biomaterials, 2017, 141, 330–339.
S. M. Mahler, mAbs, 2015, 7, 53–65.
37 Y. Zhao, N. L. Fletcher, T. Liu, A. C. Gemmell, Z. H. Houston, 48 F. S. Mehrabadi, J. Adelmann, S. Gupta, S. Wedepohl,
I. Blakey and K. J. Thurecht, Nanotheranostics, 2018, 2, 360–
370.
M. Calderon, U. Brinkmann and R. Haag, Curr. Cancer
Drug Targets, 2016, 16, 639–649.
38 Y. Zhao, Z. H. Houston, J. D. Simpson, L. Chen, 49 J. Tang, D. Shen, J. Zhang, F. S. Ligler and K. Cheng, Expert
N. L. Fletcher, A. V. Fuchs, I. Blakey and K. J. Thurecht,
Mol. Pharm., 2017, 14, 3539–3549.
39 B. Voit, J. Polym. Sci., Part A: Polym. Chem., 2000, 38, 2505–
2525.
Opin. Biol. Ther., 2015, 15, 1251–1255.
50 J. C. Knight, M. Mosley, H. T. Uyeda, M. Cong, F. Fan,
S. Faulkner and B. Cornelissen, Mol. Pharm., 2017, 14,
2307–2313.
40 D. J. Coles, B. E. Rolfe, N. R. B. Boase, R. N. Veedu and 51 P. Horan Hand, D. Colcher, D. Salomon, J. Ridge, P. Noguchi
K. J. Thurecht, Chem. Commun., 2013, 49, 3836–3838. and J. Schlom, Cancer Res., 1985, 45, 833–840.
41 A. K. Pearce, B. E. Rolfe, P. J. Russell, B. W. C. Tse, 52 K.-S. Kim, Y. Park, H.-J. Hong, K.-P. Kim, K.-H. Lee and
A. K. Whittaker, A. V. Fuchs and K. J. Thurecht, Polym.
Chem., 2014, 5, 6932–6942.
42 A. Lowe, Polym. Chem., 2010, 1, 17–36.
43 M. T. Taylor, M. L. Blackman, O. Dmitrenko and J. M. Fox, J.
Am. Chem. Soc., 2011, 133, 9646–9649.
44 O. Keinanen, K. Fung, J. Pourat, V. Jallinoja, D. Vivier,
D.-E. Kim, Bull. Korean Chem. Soc., 2012, 33, 651–656.
53 A. El Guerrab, M. Bamdad, F. Kwiatkowski, Y.-J. Bignon,
F. Penault-Llorca and C. Aubel, Oncotarget, 2016, 7, 73618–
73637.
54 T. M. Buttke, J. A. McCubrey and T. C. Owen, J. Immunol.
Methods, 1993, 157, 233–240.
N. K. Pillarsetty, A. J. Airaksinen, J. S. Lewis, B. M. Zeglis 55 F. Yang, S. S. Teves, C. J. Kemp and S. Henikoff, Biochim.
and M. Sarparanta, EJNMMI Res., 2017, 7, 95. Biophys. Acta, Rev. Cancer, 2014, 1845, 84–89.
45 X. Shi, K. Gao, H. Huang and R. Gao, Bioconjugate Chem., 56 X. J. Fang, H. Jiang, Y. Q. Zhu, L. Y. Zhang, Q. H. Fan and
2018, 29, 250–254.
Y. Tian, Oncol. Rep., 2014, 31, 2735–2742.
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.