In Vivo Half-Life Extension of BMP1/TLL Metalloproteinase Inhibitors Using Small-Molecule Human Serum Albumin Binders

Article abstract of DOI:10.1021/acs.bioconjchem.0c00662

Reducing the required frequence of drug dosing can improve the adherence of patients to chronic treatments. Hence, drugs with longer in vivo half-lives are highly desirable. One of the most promising approaches to extend the in vivo half-life of drugs is conjugation to human serum albumin (HSA). In this work, we describe the use of AlbuBinder 1, a small-molecule noncovalent HSA binder, to extend the in vivo half-life and pharmacology of small-molecule BMP1/TLL inhibitors in humanized mice (HSA KI/KI). A series of conjugates of AlbuBinder 1 with BMP1/TLL inhibitors were prepared. In particular, conjugate c showed good solubility and a half-life extension of >20-fold versus the parent molecule in the HSA KI/KI mice, reaching half-lives of >48 h with maintained maximal inhibition of plasma BMP1/TLL. The same conjugate showed a half-life of only 3 h in the wild-type mice, suggesting that the half-life extension was principally due to specific interactions with HSA. It is envisioned that conjugation to AlbuBinder 1 should be applicable to a wide range of small molecule or peptide drugs with short half-lives. In this context, AlbuBinders represent a viable alternative to existing half-life extension technologies.

Full text of DOI:10.1021/acs.bioconjchem.0c00662

pubs.acs.org/bc
Article
In Vivo Half-Life Extension of BMP1/TLL Metalloproteinase Inhibitors
Using Small-Molecule Human Serum Albumin Binders
Julien C. Vantourout, Andrew M. Mason,& Josephine Yuen,& Graham L. Simpson, Ghotas Evindar,
Letian Kuai, Michael Hobbs, Emma Edgar, Saul Needle, Xiaopeng Bai, Steve Wilson, Paul Scott-Stevens,
William Traylen, Kim Lambert, Neil Young, Shenaz Bunally, Scott G. Summerfield, Richard Snell,
Rakesh Lad, Eric Shi, Steven Skinner, Lisa Shewchuk, Allan J.B. Watson, Chun-wa Chung, Sandeep Pal,
Dennis A. Holt, Lara S. Kallander, Joanne Prendergast, Katrina Rivera, David G. Washburn,
Mark R. Harpel,* Christopher Arico-Muendel,* and Albert Isidro-Llobet*
Cite This: Bioconjugate Chem. 2021, 32, 279−289
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Reducing the required frequence of drug dosing can
improve the adherence of patients to chronic treatments. Hence,
drugs with longer in vivo half-lives are highly desirable. One of the
most promising approaches to extend the in vivo half-life of drugs is
conjugation to human serum albumin (HSA). In this work, we
describe the use of AlbuBinder 1, a small-molecule noncovalent
HSA binder, to extend the in vivo half-life and pharmacology of
small-molecule BMP1/TLL inhibitors in humanized mice (HSA
KI/KI). A series of conjugates of AlbuBinder 1 with BMP1/TLL
inhibitors were prepared. In particular, conjugate c showed good
solubility and a half-life extension of >20-fold versus the parent
molecule in the HSA KI/KI mice, reaching half-lives of >48 h with
maintained maximal inhibition of plasma BMP1/TLL. The same
conjugate showed a half-life of only 3 h in the wild-type mice, suggesting that the half-life extension was principally due to specific
interactions with HSA. It is envisioned that conjugation to AlbuBinder 1 should be applicable to a wide range of small molecule or
peptide drugs with short half-lives. In this context, AlbuBinders represent a viable alternative to existing half-life extension
technologies.
relevant of these proteins is human serum albumin (HSA),
which is the most abundant plasma protein (40 mg/mL, 600
μM), is highly soluble, and has an extraordinarily long
circulatory half-life as a direct result of its size and an FcRn-
mediated recycling pathway (mouse t1/2 = 40 h; human t1/2 = 19
to 22 days).10 Therefore, in the last two decades, a number of
biopharmaceutical entities have been genetically fused to
albumin to produce biologically active covalent albumin-drug
conjugates with extended half-lives (Table 1, entry 4).5,9,10 An
extension of this strategy entails reengineering albumin with
additional surface thiols for drug conjugation as well as
enhanced FcRn binding.11 In addition, the ability of albumin
to bind to endogenous ligands has been exploited in the
INTRODUCTION
■
Poor adherence to dosing guidelines by patients leads to
increased morbidity and mortality and costs to healthcare
systems of an estimated $100 billion a year.1−4 This lack of
compliance is in part due to the inconvenience and/or
discomfort of frequent administration of some drugs, especially
those administered parenterally to patients with chronic
illnesses.4 To overcome this issue, long-acting parenterals
(LAP), delivered by microneedle injection, have been developed
and are now a well-established approach.5 In addition, several
chemical conjugation strategies, including PEGylation, PASyla-
tion (coupling to peptide sequences containing Pro, Ala, and
Ser), and HESylation (coupling to biodegradable hydroxyethyl
starch) have been used to extend the in vivo residence times of
drugs.6,7 These technologies increase the size/hydrodynamic
volume of the cargo drug, which reduces renal clearance and
enables less frequent dosing (Table 1, entries 1 to 3).8,9
Received: December 2, 2020
Revised: January 20, 2021
Published: February 1, 2021
An excellent alternative for half-life extension of therapeutics
is the so-called “piggy-back” approach, which takes advantage of
the long half-lives of certain plasma proteins.8,9 The most
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
© 2021 American Chemical Society
Bioconjugate Chem. 2021, 32, 279−289
279

Bioconjugate Chemistry
pubs.acs.org/bc
Article
Table 1. Existing Half-Life Extension Technologies
development of noncovalent approaches. Through the years,
antibodies, fatty-acids, small molecules, and peptides binding to
HSA have been used to extend the circulating half-life of several
drugs and biologically active molecules (Table 1, entries 5 to
8).5,9,10,12−15,20−32 However, weak affinity and selectivity for
HSA, low solubility, and synthetic complexity are still significant
280
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
Figure 1. AlbuBinder 1 and its close analogue 2. (a) Structure and properties of AlbuBinder 1. (b) SPR curve of AlbuBinder 1. (c) Structure of
analogue 2. (d) Crystal structure of 2 binding to domain IB of the albumin part of Albiglutide.
limitations that impede the development of a general platform
for half-life extension of small molecules and peptides. Recently,
Heinis and co-workers reported a lipidated heptapeptide with
high affinity for human serum albumin (KD = 39 nM).16 They
demonstrated that this ligand extended the half-life of cyclic
peptides in rats 25-fold (from minutes to over 7 h) when using a
(3xPEG24) linker (Table 1, entry 9). Despite these advances, to
our knowledge, there is no example of a high affinity, specific,
soluble, and easily synthesized small-molecule HSA binder that
is able to induce the half-life extension of small-molecule drugs
while maintaining the biological activity of the original
counterpart.
Herein, the use of a high affinity, ligand-efficient small-
molecule HSA binder (AlbuBinder 1) for the extension of half-
life and pharmacology of bioactive molecules is reported. This
AlbuBinder was discovered using DNA-encoded library (DEL)
technology and easily synthesized in four steps from
commercially available starting materials.17−19 As a proof-of-
concept, this technology was applied to bone morphogenetic
protein 1 (BMP1)/tolloid-like (TLL) metalloproteinase in-
hibitors in rodents.
tractable target for DEL affinity selections.30−32 From several
screenings, AlbuBinder 1 was identified as a high-affinity small-
molecule HSA binder as it displayed a 98.1% HSA binding in a
high-performance liquid chromatography (HPLC) assay and a
87% HSA recovery in an affinity selection mass spectrometry
(AS-MS) assay (Figure 1a). In addition, AlbuBinder 1
presented an SPR profile showing saturable binding with 1:1
stoichiometry and a KD = 49.7 nM using steady-state analysis
(Figure 1b), while analysis of the X-ray crystal structure of 2
(Figure 1c), a close analogue of AlbuBinder 1, binding to
Albiglutide suggested high selectivity and specificity for Domain
IB of the albumin part of Albiglutide (Figure 1d).34,35 Finally,
the acceptable solubility (54 μg/mL) of AlbuBinder 1
differentiates it from other small molecules and lipids reported
in the literature for half-life extension of therapeutics.12,16
In Vivo Considerations. To facilitate preclinical evaluation,
it would be highly desirable that the Albubinders bind to the
serum albumins of relevant animal models (i.e., mice and rat).
However, serum albumin is significantly different across species,
rendering the discovery of cross-species AlbuBinders extremely
challenging. For instance, domain IB has only a 53% sequence
homology between mouse serum albumin (MSA) and HSA.
Hence, AlbuBinder 1 showed 0% recovery in MSA AS-MS and a
KD > 100 μM in MSA SPR. Thus, for proof-of-concept studies, a
genetically modified mouse model from genOway, homozygous
HSA humanized mice, in which MSA has been replaced by HSA
(KI/KI) was used. In brief, the humanization of albumin was
performed by in-frame insertion of the human albumin cDNA
into the mouse albumin gene where the human albumin is under
control of the endogenous promoter. The mouse signal peptide
RESULTS AND DISCUSSION
■
Development of the AlbuBinder Small Molecule. The
GSK DEL technology was first used to identify compounds that
selectively bind to HSA. This technology allows the facile and
small-scale screening of billions of compounds, and the direct
establishment of attachment point for conjugation to the
drug.18,33 Previous reports have demonstrated that HSA is a
281
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
was conserved for enhanced expression and correct maturation
of the human albumin protein.36 The results were compared
against the strain-matched wild-type controls (WT). In vivo PK
studies with AlbuBinder 1 demonstrated the expected human
selectivity with a prolonged half-life of 8.7 h in the KI/KI mice
compared with a very short half-life of 0.2 h in WT mice (Figure
2).
BMP1/TLL inhibition in the serum-based activity assay (pIC50
7.5), but exhibited poor solubility (9 μg/mL) (Table 2, entry 3).
It was hypothesized that this could prevent subcutaneous
administration and having a compound that is too lipophilic may
also lead to increasing nonspecific binding (e.g., with other
plasma proteins or tissue), which would suppress the benefits of
using a specific HSA binder like 1 in comparison to a nonspecific
fatty acid or a lipophilic small molecule. This was confirmed in in
vivo experiments. Although conjugate a had a much longer half-
life than compound 3, the values observed in KI/KI and WT
mice were virtually identical (8.9 and 9.2 h, respectively) (Table
2, entry 3 and Figure 5). Therefore, it was concluded that
nonspecific binding interactions may be largely contributing to
the half-life extension of compound 3. In order to improve the
solubility of the conjugates, two separate strategies were
followed. First, a glutamic acid was incorporated between
AlbuBinder 1 and the PEG12 linker to afford conjugate b
(Figure 4b). Second, compound 3 was replaced by a structurally
related, highly potent BMP1/TLL inhibitor, compound 4, to
afford conjugate c (Figure 4c).37,40 Compound 4 presented
better solubility and potency than compound 3, but still suffered
from very short half-life in rat (0.16 h) (Table 2, entry 2). On the
basis of the crystal structure of compound 4 bound to BMP1,37
the best vector for functionalization (i.e., phenolic position) was
selected (Figure 3b). A straightforward synthetic strategy for the
preparation of conjugate c was developed (Scheme 1), and
conjugates b and c were tested in vitro and in vivo.
Both conjugates b and c showed a dramatic increase in
solubility compared with conjugate a (from 9 μg/mL to 606 μg/
mL and 838 μg/mL, respectively) and reasonable binding to
HSA in the HPLC assay (Table 2, entries 4 and 5). SPR analysis
of conjugate c showed confirmed HSA binding with a pKD = 5.6.
When tested in vivo, conjugate b showed an excellent 22-fold
increase in half-life in the KI/KI mice (from 1 h to 22.7 h).
However, a 6-fold increase was also observed in the WT mice
(from 1.6 to 10.7 h) (Table 2, entry 4 and Figure 5). Hence, we
again hypothesized that the half-life extension was partly due to
nonspecific interactions that were only mitigated in part by
inclusion of the glutamic acid linker. In contrast, using
compound 4 as the BMP1/TLL inhibitor component led to
an in vivo half-life in the KI/KI mice of >24 h but a realtively
short half-life in the wild type (3 h) for conjugate c (Table 2,
entry 4 and Figure 5).
Figure 2. Mean PK profile of AlbuBinder 1 in wild-type (WT) and
heterologous expressing HSA (KI/KI) mice, following a single
intravenous bolus administration of 0.5 mg/kg.
Conjugation of AlbuBinder 1 to Small Molecules. The
extended half-life of AlbuBinder 1 in the KI/KI humanized mice
encouraged its conjugation to bioactive small molecules with
short half-lives for proof-of-concept of pharmacokinetic and
pharmacodynamic enhancement. Compound 3, a highly potent
inhibitor of the BMP1/TLL family of metalloproteases
implicated in fibrotic disease (pIC50 > 9, >6.4, and >6.1 for
recombinant BMP, TLL1 and TLL2 respectively, and 8.2 for
serum BMP1/TLL activity described in this study), was selected
(Figure 3a)37−39 as this molecule also suffers from a short half-
life in vivo (mouse t1/2 = 1.6 h, rat t1/2 < 1 h and KI/KI genOway
mice t1/2 = 1 h), making it a suitable candidate to test the
Albubinder platform.
Thus, a series of conjugates of compound 3 with AlbuBinder
1 were synthesized, aiming to extend its half-life while
maintaining its pharmacological response. The latter was
determined by monitoring BMP1/TLL inhibition in a serum-
based activity assay that measures endogenous BMP1/TLL
activity present in the serum compartment, as influenced by the
presence of serum albumin. Conjugate a was initially
synthesized, in which compound 3 was conjugated to 1 using
a PEG12 linker (Figure 4a). The attachment point on
compound 3 was chosen on the basis of previous SAR from
BMP1/TLL inhibition assays and synthetic accessibility. The
conjugate displayed good HSA binding (97.2% by HPLC) and
To evaluate the influence of the linker-length, AlbuBinder 1
was attached to compound 4 using a shorter PEG linker.
Conjugate d was obtained following an analogous synthetic
route as for conjugate c (Figure 4d). Conjugate d showed high
HSA binding by HPLC (98.3%) and good solubility (536 μg/
mL). Interestingly, serum BMP1/TLL inhibition potency
decreased compared with conjugate c (pIC50 7.8) (Table 2,
Figure 3. BMP1/TLL inhibitors. (a) Compound 3. (b) Compound 4.
282
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
Figure 4. Conjugates of BMP1/TLL inhibitors.
entry 6), possibly due to steric hindrance decreasing the ability
of the HSA-bound conjugate to inhibit BMP1/TLL.
Fatty Acid Control. To compare the ability of AlbuBinder 1
to extend the half-life with other reported technologies,
283
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
a
Table 2. Compound Data
entry
compound
serum BMP1/TLL pIC50 HPLC HSA binding (%) solubility (μg/mL) in vivo half-life KI/KI mice (h) in vivo half-life WT mice (h)
1
2
3
4
5
6
7
3
4
8.2
9.8
7.5
7.5
8.3
7.8
7.8
-
-
-
-
9
606
838
536
386
1
ND
8.9
22.7
>24 h
ND
0.5
1.6
ND (0.16*)
9.2
conjugate a
conjugate b
conjugate c
conjugate d
conjugate e
97.2
96.6
97.4
98.3
98.6
10.7
3.0
ND
1.8
a
ND: not determined; *: Compound 4 half-life in male Sprague Dawley rat.
Figure 5. Mean PK profiles of conjugate a, conjugate b, conjugate c, and conjugate e, in WT and KI/KI mice, following a single intravenous bolus
administration of 0.5 mg/kg.
conjugate e was synthesized (Figure 4e). This conjugate
contains compound 3 as the BMP1/TLL inhibitor, connected
by a PEG12 to fatty acid moiety primarily responsible for the
long half-life of once-weekly type-2 diabetes drug Semaglutide.41
The HSA binding of this conjugate by HPLC was still high
(98.6%), with reasonable solubility (386 μg/mL) and showed a
good inhibition of BMP1/TLL (pIC50 7.8). However, the in vivo
half-life was very short both in the KI/KI mice (0.5 h) and the
WT mice (1.8 h) (Table 2, entry 7 and Figure 5).
Proof-of-Concept In Vivo PK/PD Study. Conjugate c was
further evaluated in a 48 h in vivo PK/ex vivo PD experiment
providing further evidence of extended half-life in the KI/KI
mice (>48 h, Figure 6). Samples of plasma extracted at various
time points were run in the serum-based BMP1/TLL activity
assay and maximal inhibition of plasma BMP1/TLL was
observed through the entire time course (Figure 6).
serum assay. Conjugates a−c showed >8-fold increase for the in
vivo half-life in humanized mice when compared with the parent
small-molecule BMP1/TLL inhibitor. In particular, conjugate c
showed a half-life extension of >20-fold versus the parent
molecule in the HSA KI/KI mice, reaching half-lives of >48 h,
and had good solubility which would facilitate subcutaneous
administration. The same conjugate showed less than 2-fold
half-life extension in the wild-type mice suggesting that the half-
life extension is principally due to specific interactions with HSA.
Additionally, a proof-of-concept study monitoring PK and PD
up to 48 h after a single dose demonstrated that conjugate c
maintained high systemic concentrations over the duration of
the study in KI/KI mice, while also elicitating a sustained
maximal inhibition of plasma BMP1/TLL.
Finally, it is envisioned that conjugation to AlbuBinder 1
should be applicable to a wide range of small molecule or peptide
drugs with short half-lives. This could be an alternative to fatty
acid-mediated half-life extension of drugs showcased by the
recent approval by the FDA of the once-weekly Semaglutide for
type-2 diabetes.41 To our knowledge, the approach has not yet
been applied to extend the half-life and pharmacology of small-
molecule drugs. In this context, AlbuBinders represent a viable
alternative to existing half-life extension technologies. In
particular, drugs which modulate extracellular or cell-surface
targets which have ready access to the plasma compartment
would be well suited to this approach.
CONCLUSION
■
Herein AlbuBinder 1, a small-molecule HSA binder discovered
using DNA encoded library technology, was proven to be
successful for the extension of in vivo half-life and pharmacology
of BMP1/TLL inhibitors in humanized mice (HSA KI/KI). The
high affinity and specificity for HSA of AlbuBinder 1 were
confirmed by SPR, AS-MS, and HPLC. In addition, AlbuBinder
1 was conveniently attached to small-molecule BMP1/TLL
inhibitors. The four synthesized conjugates of AlbuBinder 1
with BMP1/TLL inhibitors were found to bind with the desired
affinity to HSA and to show good BMP1/TLL inhibition in a
284
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
Scheme 1. Synthesis of Conjugate c
Millipore MultiscreenHTS-PCF filter plates (MSSL BPC). The
filtrate was quantified by suitably calibrated flow injection
Chemi-Luminescent Nitrogen Detection.42
HPLC HSA Binding Assay. Chemically bonded HSA HPLC
stationary phase (Chiral Technologies, France) was used for
EXPERIMENTAL SECTION
■
Kinetic Solubility Assay Using CLND Detector. Five
microliters of 10 mM DMSO stock solution of the compound
was diluted to 100 μL with pH 7.4 phosphate buffered saline,
equilibrated for 1 h at room temperature, and filtered through
285
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
▼
●
Figure 6. In vivo PK and PD study on conjugate c. Symbols: ; WT mice, ; KI/KI mice. Plasma activity values were normalized to the predose
(“control”) activities for each strain. Statistics: **p < 0.01 and ****p < 0.001 by two-way ANOVA with Sidak’s multicomponent comparisons test for
WT versus KI/KI.
measuring compounds binding to plasma proteins, applying
2. Standard for compound signal in the presence of protein:
linear gradient elution up to 30% iso-propanol.43 The run time
1 μL of test compounds + 10 μL of protein stock + 5.5 μL
was 6 min including the re-equilibration of the stationary phases
with the 50 mM pH 7.4 ammonium acetate buffer. The gradient
retention times were standardized using a calibration set of
mixtures as described in the reference.44 The average standard
error of the assay depends on the binding strength and kinetics
of the compounds. It ranges from 5% in the medium binding
range which reduces to 0.1% at binding above 99% with fast
kinetics.
Surface Plasmon Resonance (SPR). A Biacore 4000
surface plasmon resonance instrument was used to measure
binding. Human serum albumin (fatty acid depleted, Sigma
A9731) was immobilized to the surface of a CM5 sensor chip
using random amine coupling at pH 5 to a level of typically 2800
RU. Compounds were tested as 11 point concentration titration
using a 3-fold dilution series decreasing from 100 μM in
phosphate buffered saline and 1% DMSO at pH 7.4 at 25 °C.
Concentration−response data from reference-subtracted curves
were fit to a steady-state 1:1 binding model to determine KD with
the BIAevaluation software.
Affinity Selection Mass Spectrometry (AS-MS) by Size
Exclusion Chromatography (SEC). Sample Preparation.
Affinity selections were conducted in PBS, pH 7.4. HSA (Sigma,
cat. A3782-500) and MSA (Sigma, cat. A3139-500) were
prepared as 10 μM stock solutions. Test compounds were
obtained as 10 mM solutions in DMSO, diluted to 1 mM with
45/45/10 MeOH/H2O/DMSO, and tested as mixtures of up to
ten compounds (final concentrations of each at 100 μM). Size
exclusion columns (G25 microspin columns, Part No. SEM
S025, Nest Group) were hydrated with selection buffer, the resin
allowed to swell for at least 3 min, then spun for 40 s at 2400 rpm
on an Eppendorf 5417R centrifuge to remove excess buffer.
SEC. Experiments were performed with samples as follows:
DMSO; no SEC
3. Control for compound breakthrough in absence of
protein: 2 μL of test compounds + 20 μL of PBS buffer;
then SEC
4. Protein binding measurement: 2 μL of test compounds +
20 μL of protein stock; then SEC (done in duplicate)
Samples 3 and 4 were incubated on ice for a minimum of 30
min to minimize sample dissociation during the SEC step.
Sample solutions were applied directly to the surface of the
packed bed of the spin columns, then spun for 1 min at 2400
rpm. The eluants were then diluted with 99/1 DMSO/water (11
μL) and transferred to MS injection vials.
HPLC-MS. Samples were run on a Thermo Fisher UltiMate
3000 HPLC system interfaced with a LTQ XL linear ion trap
mass spectrometer. The LC employed a Kinetix C18 column
(1.7 mm, 2 mm × 50 mm) operating a 3 min gradient for buffer
B from 4 to 90%, where buffer A was 0.1% formic acid and buffer
B was 0.1% formic acid/0.1% water in acetonitrile. Signal
detection was by PDA (200−400 nm). Samples were run in the
following order: standards 1 and 2, blank, control, protein.
Animal Studies. All animal studies were ethically reviewed
and carried out in accordance with Animals (Scientific
Procedures) Act 1986 and the GSK Policy on the Care, Welfare
and Treatment of Animals. Human Serum Albumin (HSA)
humanized mice, generated and maintained on the C57BL/6N
genetic background, were obtained from genOway (Lyon,
France). Colonies of homozygous HSA knock-in mice (KI/KI)
or wild-type mice (WT) were supplied to GlaxoSmithKline
Stevenage (U.K.) In Vivo Science and Delivery animal facility,
where they were multiply housed in autoclaved Techniplast
GM500 cages containing IPS Lignocel BK8/15 bedding,
Datesand Paper Shaving, red Perspex dome home, cardboard
fun tunnel, and wooden chew blocks. Animals were maintained
at an ambient temperature of 20.5 to 23.5 °C and relative
humidity of 39% to 61%, on a 6 am to 8 pm light−dark cycle,
1. Standard for 100% compound signal: 4 μL of test
compounds + 62 μL of blank solution (MeOH/H2O/
DMSO: 45/45/10); no SEC
286
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
with free access to food (Labdiet Irradiated 5LF2Maintenance
or Envigo 2918 Breeder Diet) and animal grade drinking water.
PK Study Design. Dose Administration. Each mouse
received a single intravenous bolus administration of the
selected compound at target dose levels of 0.5 mg free form/
kg and a target dose volume of 2 mL/kg, formulated in DMSO
(5%): PEG400 (20%): 10% Kleptose in water (75%) at
concentrations of 0.25 mg/mL. The resulting solutions were
filtered through Durapore 33 mm, 0.22 μm sterile filter.
Study Design for Each Compound. Post IV dose
administration, blood samples (approximately 50 μL) were
collected via serial tail vein (TV) sampling and/or terminal
cardiac puncture (CP), into tubes containing EDTA as the
anticoagulant. Three groups of mice for each compound were
sampled as follows:
Group 1: 2 WT and KI/KI mice (1 male, 1 female), blood
sampling at 4 min (CP)
Group 2: 3 WT and KI/KI mice (2 males, 1 female), blood
sampling at 40 min, 19 h (TV) and 24 h (CP).
Group 3: 3 WT and KI/KI mice (2 males, 1 female), blood
sampling at 1.5 h, 3 h (TV) and 7 h (CP).
Blood samples were stored on crushed ice prior to
centrifugation (1900gav for 10 min at 4 °C), and the resulting
plasma samples were stored at −20 °C prior to bioanalysis.
PK/PD Study Design. Dose Administration. Each mouse
received a single intravenous bolus administration of conjugate
C at a target dose level of 0.5 mg free form/kg and a target dose
volume of 2 mL/kg (0.25 mg/mL), formulated in DMSO (5%):
PEG400 (20%): 10% Kleptose in water (75%). The resulting
solution was filtered through Durapore 33 mm, 0.22 μm sterile
filter.
Study Design. Following IV dose administration to wild-type
(WT) and knock-in (KI/KI) mice (n = 3 mice, 2 male, 1 female
per time point), terminal blood samples were collected via
cardiac puncture following CO2 asphyxiation at the following
time points: 3, 16, 24, and 48 h. Blood samples were stored on
crushed ice prior to centrifugation (1900gav for 10 min at 4 °C),
and the resulting plasma samples were stored at −20 °C prior to
bioanalysis.
reported BMP1/TLL-targeted cleavage site and adjacent amino
acid residues of prolylsyl oxidase: (5-FAM)-SHVDRMVG/
DDPYPNPY-(5-TAMRA), in which the cleavage site is
designated by the forward slash.47 For assays, serum/plasma
was diluted with assay buffer and inhibitor, then mixed with
peptide substrate in black, round-bottom plates (Dynatech) to
initiate reaction at 37 °C at final reaction concentrations of 50%
serum/plasma, 50 mM HEPES (pH 7.4), 0.02% Brij 35, 10 mM
CaCl2, 2 mM ZnCl2 and 1% DMSO. Fluorescence intensity was
then read every 2 min using a PerkinElmerWallac Envision 2103
Multilabel plate reader (λex/λem = 485 nm/535 nm), and
activity was determined as the slope of fluorescence increase
over the linear reaction phase (usually a period from 5−30 min
after initiation). Activities were measured using commercially
obtained human serum (pooled from 6 healthy donors; BioIVT,
Inc.), to obtain reported pIC50 values, and heparinized murine
plasma samples obtained from blood serially collected at
indicated times, to determine the ex vivo pharmacodynamic
response of conjugate c. Other serum/plasma enzymes also
degrade the FRET peptide and contribute to the observed
fluorescence change. However, the percentage attributable to
BMP1/TLL was obtained by subtracting the portion not
inhibited by 1 mM compound 18 (pIC50 8.2 in this assay),
another potent BMP1/TLL inhibitor described by Kallander et
al.,13 in parallel reactions; percent total inhibition by the
Albubinders approximate that of compound 18 and other
potent, selective BMP1/TLL inhibitors in this assay (not
shown). Concentration−response curves were fit by nonlinear
regression to a four-parameter hyperbolic response curve (Prism
Software) to obtain pIC50 values; statistical comparisons of
pharmacodynamic responses in Figure 6 were obtained using
Prism Software.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.bioconjchem.0c00662.
■
sı
Synthesis protocols and characterization (PDF)
PK Study Sample Analysis. Concentrations of each
compound in plasma were analyzed by UPLC MS/MS methods.
Each compound was extracted from 5 μL of mouse plasma by
protein precipitation using acetonitrile containing an internal
standard. Extracts were analyzed by UPLC MS/MS using a
TurboIonSpray interface. The applicable analytical runs met all
predefined run acceptance criterea. All compound concen-
trations were reported in ng/mL of mouse blood. No statistical
analysis of the data was conducted.
Pharmacokinetic Analysis. Noncompartmental pharma-
cokinetic analysis was performed using Phoenix WinNonlin 6.3.
All computations utilized the nominal sampling times.
Serum/Plasma BMP1/TLL Activity Measurements.
BMP1/TLL inhibition was evaluated for the proteolytic
activities of endogenous BMP1 found in human serum and
mouse plasma. Although BMP1 is the major enzyme of this
family identified in blood,45,46 we refer to the activity as BMP1/
TLL to recognize the potential contribution from uncharac-
terized TLL1 and TLL2 proteins, which are also inhibited by the
parental molecules for which the Albubinder conjugates are
derived.37 The assay monitors the serum/plasma-driven
increase in fluorescence intensity that results from cleavage of
an exogenously added fluorescence resonance energy transfer
(FRET) peptide substrate. The peptide substrate contains the
AUTHOR INFORMATION
Corresponding Authors
■
Albert Isidro-Llobet − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.; orcid.org/0000-0001-
9108-5916; Email: albert.x.isidro-llobet@gsk.com
Christopher Arico-Muendel − GlaxoSmithKline US,
Cambridge, Massachusetts 02140, United States;
Email: christopher.c.arico-muendel@gsk.com
Mark R. Harpel − GlaxoSmithKline US, Collegeville,
Pennsylvania 19426, United States; Email: mark.r.harpel@
gsk.com
Authors
Julien C. Vantourout − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.; Department of Pure and
Applied Chemistry, University of Strathclyde, Glasgow
G11XL, U.K.
Andrew M. Mason − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Josephine Yuen − GlaxoSmithKline US, Cambridge,
Massachusetts 02140, United States
Graham L. Simpson − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
287
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
Ghotas Evindar − GlaxoSmithKline US, Cambridge,
Massachusetts 02140, United States
Letian Kuai − GlaxoSmithKline US, Cambridge, Massachusetts
02140, United States
Michael Hobbs − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Emma Edgar − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Saul Needle − GlaxoSmithKline US, Collegeville, Pennsylvania
19426, United States
Xiaopeng Bai − GlaxoSmithKline US, Cambridge,
Massachusetts 02140, United States
Steve Wilson − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Paul Scott-Stevens − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
William Traylen − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Kim Lambert − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Neil Young − GlaxoSmithKline U.K., Stevenage, Hertfordshire
SG1 2NY, U.K.
Shenaz Bunally − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Scott G. Summerfield − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Richard Snell − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.
Rakesh Lad − GlaxoSmithKline U.K., Stevenage, Hertfordshire
SG1 2NY, U.K.
Eric Shi − GlaxoSmithKline US, Cambridge, Massachusetts
02140, United States
Steven Skinner − GlaxoSmithKline US, Cambridge,
Massachusetts 02140, United States
Lisa Shewchuk − GlaxoSmithKline US, Collegeville,
Pennsylvania 19426, United States
Allan J.B. Watson − University of St Andrews, North Haugh, St
Andrews KY16 9ST, U.K.; orcid.org/0000-0002-1582-
4286
Chun-wa Chung − GlaxoSmithKline U.K., Stevenage,
Hertfordshire SG1 2NY, U.K.; orcid.org/0000-0002-
2480-3110
Sandeep Pal − GlaxoSmithKline U.K., Stevenage, Hertfordshire
SG1 2NY, U.K.
Dennis A. Holt − GlaxoSmithKline US, Collegeville,
Pennsylvania 19426, United States
Lara S. Kallander − GlaxoSmithKline US, Collegeville,
Pennsylvania 19426, United States
Joanne Prendergast − GlaxoSmithKline US, Collegeville,
Pennsylvania 19426, United States
ACKNOWLEDGMENTS
■
We thank the GlaxoSmithKline/University of Strathclyde
Collaborative Ph.D. programme for funding this work (to
J.C.V., Ph.D. studentship); Don Huddler, Liam Bonnette, and
David Dixon for their advice on SPR; Timothy Adams and
Hector Yiannakas for the synthesis of precursors; and Alan
6
Olzinski for advice on statistical treatment of data in Figure .
We also acknowledge Robert Willette, Erding Hu and Pu Qin for
their contributions to development of the serum/plasma
BMP1/TLL activity assay, which will be described in further
detail in a subsequent publication (Needle et al., manuscript in
preparation).
REFERENCES
■
(1) Brown, M. T., and Bussell, J. K. (2011) Medication Adherence:
WHO Cares? Mayo Clin. Proc. 86, 304−314.
(2) Gast, A., and Mathes, T. (2019) Medication adherence influencing
factorsan (updated) overview of systematic reviews. Syst. Rev. 8, 112.
(3) Iuga, A. O., and McGuire, M. J. (2014) Adherence and health care
costs. Risk Manag Healthc Policy 7, 35−44.
(4) Osterberg, L., and Blaschke, T. (2005) Adherence to medication.
N. Engl. J. Med. 353, 487−497.
(5) Dadhaniya, T., Sharma, O., Gohel, M., and Mehta, P. (2015)
Current Approaches for in vitro Drug Release Study of Long Acting
Parenteral Formulations. Curr. Drug Delivery 12, 256−270.
̈
(6) Schlapschy, M., Binder, U., Borger, C., Theobald, I., Wachinger,
K., Kisling, S., Haller, D., and Skerra, A. (2013) PASylation: a biological
alternative to PEGylation for extending the plasma half-life of
pharmaceutically active proteins. Protein Eng., Des. Sel. 26, 489−501.
(7) Hey, T., Knoller, H., and Vorstheim, P. (2012) Half-Life Extension
through HESylation®. In Therapeutic Proteins (Kontermann, R., Ed.),
pp 117−140, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.
(8) Sleep, D., Cameron, J., and Evans, L. R. (2013) Albumin as a
versatile platform for drug half-life extension. Biochim. Biophys. Acta,
Gen. Subj. 1830, 5526−5534.
(9) Pollaro, L., and Heinis, C. (2010) Strategies to prolong the plasma
residence time of peptide drugs. MedChemComm 1, 319−324.
(10) Elsadek, B., and Kratz, F. (2012) Impact of albumin on drug
delivery  New applications on the horizon. J. Controlled Release 157,
4−28.
(11) Schelde, K. K., Nicholls, K., Dagnæs-Hansen, F., Bunting, K.,
Rawsthorne, H., Andersen, B., Finnis, C. J. A., Williamson, M.,
Cameron, J., and Howard, K. A. (2019) A new class of recombinant
human albumin with multiple surface thiols exhibits stable conjugation
and enhanced FcRn binding and blood circulation. J. Biol. Chem. 294,
3735−3743.
(12) Zorzi, A., Linciano, S., and Angelini, A. (2019) Non-covalent
albumin-binding ligands for extending the circulating half-life of small
biotherapeutics. MedChemComm 10, 1068−1081.
(13) Holt, L. J., Basran, A., Jones, K., Chorlton, J., Jespers, L. S., Brewis,
N. D., and Tomlinson, I. M. (2008) Anti-serum albumin domain
antibodies for extending the half-lives of short lived drugs. Protein Eng.,
Des. Sel. 21, 283−288.
(14) Walker, A., Dunlevy, G., Rycroft, D., Topley, P., Holt, L. J.,
Herbert, T., Davies, M., Cook, F., Holmes, S., Jespers, L., et al. (2010)
Anti-serum albumin domain antibodies in the development of highly
potent, efficacious and long-acting interferon. Protein Eng., Des. Sel. 23,
271−278.
(15) Dennis, M. S., Zhang, M., Meng, Y. G., Kadkhodayan, M.,
Kirchhofer, D., Combs, D., and Damico, L. A. (2002) Albumin Binding
as a General Strategy for Improving the Pharmacokinetics of Proteins. J.
Biol. Chem. 277, 35035−35043.
(16) Zorzi, A., Middendorp, S. J., Wilbs, J., Deyle, K., and Heinis, C.
(2017) Acylated heptapeptide binds albumin with high affinity and
application as tag furnishes long-acting peptides. Nat. Commun. 8,
16092.
Katrina Rivera − GlaxoSmithKline US, Collegeville,
Pennsylvania 19426, United States
David G. Washburn − GlaxoSmithKline US, Collegeville,
Pennsylvania 19426, United States
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.bioconjchem.0c00662
Author Contributions
&(A.M.M., J.Y.) These authors contributed equally to the study
Notes
The authors declare no competing financial interest.
288
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289

Bioconjugate Chemistry
pubs.acs.org/bc
Article
(17) Song, M., and Hwang, G. T. (2020) DNA-Encoded Library
Screening as Core Platform Technology in Drug Discovery: Its
Synthetic Method Development and Applications in DEL Synthesis. J.
Med. Chem. 63, 6578−6599.
(18) Clark, M. A., Acharya, R. A., Arico-Muendel, C. C., Belyanskaya,
S. L., Benjamin, D. R., Carlson, N. R., Centrella, P. A., Chiu, C. H.,
Creaser, S. P., Cuozzo, J. W., et al. (2009) Design, synthesis and
selection of DNA-encoded small-molecule libraries. Nat. Chem. Biol. 5,
647−654.
(19) Kumpalume, P., Schon, O., Donahue, C. P., Evindar, G., Israel, D.
I., Paolella, D., Kuai, L., and Prabhu, N. V. (2015) Compounds for
affinity chromatography and for extending the half-life of a therapeutic
agent. Patent No. WO2015049651.
(20) van Witteloostuijn, S. B., Pedersen, S. L., and Jensen, K. J. (2016)
Half-Life Extension of Biopharmaceuticals using Chemical Methods:
Alternatives to PEGylation. ChemMedChem 11, 2474−2495.
(21) Kontermann, R. E. (2016) Half-life extended biotherapeutics.
Expert Opin. Biol. Ther. 16, 903−915.
(22) Koehler, M. F. T., Zobel, K., Beresini, M. H., Caris, L. D., Combs,
D., Paasch, B. D., and Lazarus, R. A. (2002) Albumin affinity tags
increase peptide half-life in vivo. Bioorg. Med. Chem. Lett. 12, 2883−
2886.
(23) Zhang, H. (2004) Trisodium-[(2-(R)-[(4,4-
diphenylcyclohexyl)phosphono-oxymethyl]-diethylenetriaminepenta
acetato)(aquo)gadolinium(III): Gadofosveset. Molecular Imaging and
Contrast Agent Database (MICAD), National Center for Biotechnology
Information (US), Bethesda (MD).
(24) Shechter, Y., Mironchik, M., Saul, A., Gershonov, E., Precido-
Patt, L., Sasson, K., Tsubery, H., Mester, B., Kapitkovsky, A., Rubinraut,
S., et al. (2007) New Technologies to Prolong Life-time of Peptide and
Protein Drugs In vivo. Int. J. Pept. Res. Ther. 13, 105−117.
(25) Sasson, K., Marcus, Y., Lev-Goldman, V., Rubinraut, S., Fridkin,
M., and Shechter, Y. (2010) Engineering prolonged-acting prodrugs
employing an albumin-binding probe that undergoes slow hydrolysis at
physiological conditions. J. Controlled Release 142, 214−220.
(26) Han, J., Sun, L., Chu, Y., Li, Z., Huang, D., Zhu, X., Qian, H., and
Huang, W. (2013) Design, synthesis, and biological activity of novel
dicoumarol glucagon-like peptide 1 conjugates. J. Med. Chem. 56,
9955−9968.
(27) Chen, H., Wang, G., Lang, L., Jacobson, O., Kiesewetter, D. O.,
Liu, Y., Ma, Y., Zhang, X., Wu, H., Zhu, L., et al. (2016) Chemical
Conjugation of Evans Blue Derivative: A Strategy to Develop Long-
Acting Therapeutics through Albumin Binding. Theranostics 6, 243−
253.
(33) Machutta, C. A., Kollmann, C. S., Lind, K. E., Bai, X., Chan, P. F.,
Huang, J., Ballell, L., Belyanskaya, S., Besra, G. S., Barros-Aguirre, D.,
et al. (2017) Prioritizing multiple therapeutic targets in parallel using
automated DNA-encoded library screening. Nat. Commun. 8, 16081.
(34) Matthews, J. E., Stewart, M. W., De Boever, E. H., Dobbins, R. L.,
Hodge, R. J., Walker, S. E., Holland, M. C., and Bush, M. A. (2008)
Pharmacodynamics, Pharmacokinetics, Safety, and Tolerability of
Albiglutide, a Long-Acting Glucagon-Like Peptide-1 Mimetic, in
Patients with Type 2 Diabetes. J. Clin. Endocrinol. Metab. 93, 4810−
4817.
(35) Albiglutide is formed of human serum albumin covalently
attached to two copies of a modified GLP-1 peptide. In this work, it was
used as a human serum albumin surrogate due to ease of crystalization.
(36) Viuff, D., Antunes, F., Evans, L., Cameron, J., Dyrnesli, H., Thue
Ravn, B., Stougaard, M., Thiam, K., Andersen, B., Kjærulff, S., et al.
(2016) Generation of a double transgenic humanized neonatal Fc
receptor (FcRn)/albumin mouse to study the pharmacokinetics of
albumin-linked drugs. J. Controlled Release 223, 22−30.
(37) Kallander, L. S., Washburn, D., Hilfiker, M. A., Eidam, H. S.,
Lawhorn, B. G., Prendergast, J., Fox, R., Dowdell, S., Manns, S., Hoang,
T., et al. (2018) Reverse Hydroxamate Inhibitors of Bone
Morphogenetic Protein 1. ACS Med. Chem. Lett. 9, 736−740.
(38) Dowdell, S. E., Eidam, H. S., Elban, M., Fox, R. M., Hammond,
M., Hilfiker, M. A., Hoang, T. H., Kallander, L., Lawhorn, B. G., and
Manns, S., et al. (2015) Hydroxy formamide derivatives and their use.
Patent No. WO2017006295A1.
(39) Compound 3 was referred to as compound 17 in reference 37.
(40) Compound 4 was referred to as compound 22 in reference 37.
(41) Davies, M., Pieber, T. R., Hartoft-Nielsen, M., Hansen, O. H.,
Jabbour, S., and Rosenstock, J. (2017) Effect of oral semaglutide
compared with placebo and subcutaneous semaglutide on glycemic
control in patients with type 2 diabetes: A randomized clinical trial.
JAMA 318 (15), 1460−1470.
(42) Bhattachar, S. N., Wesley, J. A., and Seadeek, C. (2006)
Evaluation of the chemiluminescent nitrogen detector for solubility
determinations to support drug discovery. J. Pharm. Biomed. Anal. 41,
152−157.
(43) Bunally, S., and Young, R. J. (2018) The role and impact of high
throughput biomimetic measurements in drug discovery. ADMET
DMPK 6, 74−84.
(44) Valko, K., Nunhuck, S., Bevan, C., Abraham, M. H., and
Reynolds, D. P. (2003) Fast Gradient HPLC Method to Determine
Compounds Binding to Human Serum Albumin. Relationships with
Octanol/Water and Immobilized Artificial Membrane Lipophilicity. J.
Pharm. Sci. 92, 2236−2248.
(45) Grgurevic, L., Macek, B., Healy, D. R., Brault, A. L., Erjavec, I.,
Cipcic, A., Grgurevic, I., Rogic, D., Galesic, K., Brkljacic, J., et al. (2011)
Circulating Bone Morphogenetic Protein 1−3 Isoform Increases Renal
Fibrosis. J. Am. Soc. Nephrol. 22, 681−692.
(46) Grgurevic, L., Macek, B., Mercep, M., Jelic, M., Smoljanovic, T.,
Erjavec, I., Dumic-Cule, I., Prgomet, S., Durdevic, D., Vnuk, D., et al.
(2011) Bone morphogenetic protein (BMP)1−3 enhances bone repair.
Biochem. Biophys. Res. Commun. 408, 25−31.
(47) Uzel, M. I., Scott, I. C., Babakhanlou-Chase, H., Palamakumbura,
A. H., Pappano, W. N., Hong, H. H., Greenspan, D. S., and Trackman,
P. C. (2001) Multiple bone morphogenetic protein 1-related
mammalian metalloproteinases process pro-lysyl oxidase at the correct
physiological site and control lysyl oxidase activation in mouse embryo
fibroblast cultures. J. Biol. Chem. 276, 22537−22543.
(28) Bech, E. M., Martos-Maldonado, M. C., Wismann, P., Sørensen,
K. K., van Witteloostuijn, S. B., Thygesen, M. B., Vrang, N., Jelsing, J.,
Pedersen, S. L., and Jensen, K. J. (2017) Peptide Half-Life Extension:
Divalent, Small-Molecule Albumin Interactions Direct the Systemic
Properties of Glucagon-Like Peptide 1 (GLP-1) Analogues. J. Med.
Chem. 60, 7434−7446.
(29) Jonassen, I., Havelund, S., Ribel, U., Plum, A., Loftager, M.,
Hoeg-Jensen, T., Volund, A., and Markussen, J. (2006) Biochemical
and physiological properties of a novel series of long-acting insulin
analogs obtained by acylation with cholic acid derivatives. Pharm. Res.
23, 49−55.
(30) Dumelin, C. E., Truessel, S., Buller, F., Trachsel, E., Bootz, F.,
Zhang, Y., Mannocci, L., Beck, S. C., Drumea-Mirancea, M., Seeliger,
M. W., et al. (2008) A portable albumin binder from a DNA-encoded
chemical library. Angew. Chem., Int. Ed. 47, 3196−3201.
(31) Franzini, R. M., Ekblad, T., Zhong, N., Wichert, M., Decurtins,
W., Nauer, A., Zimmermann, M., Samain, F., Scheuermann, J., Brown,
P. J., et al. (2015) Identification of structure-activity relationships from
screening a structurally compact DNA-encoded chemical library.
Angew. Chem., Int. Ed. 54, 3927−3931.
(32) Franzini, R. M., Nauer, A., Scheuermann, J., and Neri, D. (2015)
Interrogating target-specificity by parallel screening of a DNA-encoded
chemical library against closely related proteins. Chem. Commun. 51,
8014−8016.
289
https://dx.doi.org/10.1021/acs.bioconjchem.0c00662
Bioconjugate Chem. 2021, 32, 279−289