An optimal “Click” formulation strategy for antibody-drug conjugate synthesis

Article abstract of DOI:10.1016/j.bmc.2020.115808

As a versatile reaction for bioconjugation, Cu(I)-catalyzed alkyne-azide cycloaddition (CuAAC) has enormous potential in the synthesis of antibody-drug conjugates (ADCs). In order to optimize CuAAC-based ADC synthesis, we characterized kinetically different formulation processes by mimicking ADC synthesis using small molecules and subsequently revealed unique kinetic behaviors of different combinations of alkyne and azide conditions. Our results indicate that under ADC synthesis conditions, for an alkyne-containing drug, its concentration has minimal impact on the reaction rate when an antibody has a non-metal-chelating azide but is proportional to concentration when an antibody contains a metal-chelating azide; however, for an alkyne-containing antibody, the ADC synthesis rate is proportional to the concentration of a drug with a non-metal-chelating azide but displays almost no dependence on drug concentration with a metal-chelating azide. Based on our results, we designed and tested an optimal “click” formulation strategy that allowed rapid and cost-effective synthesis of a new ADC.

Full text of DOI:10.1016/j.bmc.2020.115808

Journal Pre-proofs
An Optimal “Click” Formulation Strategy for Antibody-Drug Conjugate Syn-
thesis
Erol C. Vatansever, Jeffrey Kang, Alfred Tuley, E. Sally Ward, Wenshe Ray
Liu
PII:
S0968-0896(20)30638-6
https://doi.org/10.1016/j.bmc.2020.115808
BMC 115808
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
7 August 2020
24 September 2020
1 October 2020
Please cite this article as: E.C. Vatansever, J. Kang, A. Tuley, E. Sally Ward, W. Ray Liu, An Optimal “Click”
Formulation Strategy for Antibody-Drug Conjugate Synthesis, Bioorganic & Medicinal Chemistry (2020), doi:
https://doi.org/10.1016/j.bmc.2020.115808
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Ltd.

An Optimal “Click” Formulation Strategy
for Antibody-Drug Conjugate Synthesis
Leave this area blank for abstract info.
Erol C. Vatansevera, Jeffrey Kangb, Alfred Tuleya, E. Sally Wardb,c,d,* and Wenshe Ray Liua,e,f,
*
a The Texas A&M Drug Discovery Laboratory, Department of Chemistry, Texas A&M University, College Station, Texas 77843, United States
b Department of Molecular and Cellular Medicine, Texas A&M University Health Science Center, College Station, TX 77843, USA
c Department of Microbial Pathogenesis and Immunology, Texas A&M University Health Science Center, Bryan, TX 77807, USA
d Centre for Cancer Immunology, Cancer Sciences Unit, Faculty of Medicine, University of Southampton, Southampton SO16 4YD, UK
e Department of Biochemistry and Biophysics, Texas A&M University, College Station, TX 77843, USA
f Department of Molecular and Cellular Medicine, College of Medicine, Texas A&M University, College Station, TX 77843, USA

Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com
An Optimal “Click” Formulation Strategy for Antibody-Drug Conjugate Synthesis
Erol C. Vatansevera, Jeffrey Kangb, Alfred Tuleya, E. Sally Wardb,c,d,* and Wenshe Ray Liua,e,f,
*
a The Texas A&M Drug Discovery Laboratory, Department of Chemistry, Texas A&M University, College Station, Texas 77843, United States
b Department of Molecular and Cellular Medicine, Texas A&M University Health Science Center, College Station, TX 77843, USA
c Department of Microbial Pathogenesis and Immunology, Texas A&M University Health Science Center, Bryan, TX 77807, USA
d Centre for Cancer Immunology, Cancer Sciences Unit, Faculty of Medicine, University of Southampton, Southampton SO16 4YD, UK
e Department of Biochemistry and Biophysics, Texas A&M University, College Station, TX 77843, USA
f Department of Molecular and Cellular Medicine, College of Medicine, Texas A&M University, College Station, TX 77843, USA
ARTICLE INFO
ABSTRACT
Article history:
Received
Received in revised form
Accepted
As a versatile reaction for bioconjugation, Cu(I)-catalyzed alkyne-azide cycloaddition (CuAAC)
has enormous potential in the synthesis of antibody-drug conjugates (ADCs). In order to
optimize CuAAC-based ADC synthesis, we characterized kinetically different formulation
processes by mimicking ADC synthesis using small molecules and subsequently revealed
unique kinetic behaviors of different combinations of alkyne and azide conditions. Our results
indicate that under ADC synthesis conditions, for an alkyne-containing drug, its concentration
has minimal impact on the reaction rate when an antibody has a non-metal-chelating azide but is
proportional to concentration when an antibody contains a metal-chelating azide; however, for
an alkyne-containing antibody, the ADC synthesis rate is proportional to the concentration of a
drug with a non-metal-chelating azide but displays almost no dependence on drug concentration
with a metal-chelating azide. Based on our results, we designed and tested an optimal “click”
formulation strategy that allowed rapid and cost-effective synthesis of a new ADC.
Available online
Keywords:
Copper(I)-catalyzed alkyne-azide cycloaddition
click reaction
antibody-drug conjugate
metal-chelating azide
kinetic formulation
2009 Elsevier Ltd. All rights reserved.
1. Introduction
concentration also substantially reduces the manufacturing cost
of ADCs. However, although efficacious, the cysteine-maleimide
Antibody-drug conjugates (ADCs) consist of cytotoxic drugs
that are covalently conjugated to antibodies. The antibody
component of an ADC directs the binding to antigens expressed
on tumor cells and enables the specific delivery of their
conjugated drugs. Consequently, ADCs combine the anti-tumor
efficacy of chemotherapeutic drugs, whilst minimizing their
significant systemic cytotoxicity.1 With 9 ADCs approved by
FDA and more than 150 in clinical trials,2 the development of
ADCs has revolutionized cancer treatment. This rapidly
expanding use of ADCs has largely overshadowed their
manufacturing challenges. How exactly antibodies and drugs are
conjugated plays a significant role in defining both the cost and
therapeutic index of an ADC.3,4 To date, the most broadly used
reaction for antibody-drug conjugation is the covalent addition of
a cysteine thiolate in an antibody to a maleimide in a drug
(Figure 1A).5 In comparison to random conjugation to lysine
residues of an antibody that are more abundant than cysteine
residues, this reaction typically leads to more homogenous
ADCs. The fast reaction kinetics (734 M-1s-1)6 also allows the use
of a low concentration of a drug (in the µM range) to achieve
efficient coupling to the antibody, which significantly simplifies
the manufacturing process of ADCs. Since most drugs used
during ADC synthesis are complicated natural products and
therefore expensive to produce, the use of a low drug
reaction has pitfalls. Maleimide reacts with lysine at a low level
to form side products.7 The reaction itself is also slowly
reversible, which leads to the release of the conjugated drug that
can contribute to systemic cytotoxicity.8 For these reasons,
substantial efforts to either improve the cysteine-maleimide
chemistry9, 10 or develop novel conjugation strategies have been
made.11-19 As the founding click reaction for bioconjugation,
Cu(I)-catalyzed alkyne-azide cycloaddition (CuAAC) has been
explored for ADC synthesis.20 In this study, we show how this
versatile reaction can be favorably optimized for this purpose.

2. Results & Discussion
the similar reaction in which BTTAA was present, indicating that
During the CuAAC synthesis of ADCs, small molecule drugs
were provided in excess to achieve homogenous loading to
antibody.20 An additional manufacturing reason to use excessive
drugs instead of excessive antibodies is that it is relatively
straightforward to remove residual drugs but very difficult to
separate unconjugated from conjugated antibodies. In order to
use CuAAC to synthesize ADCs, two possible processes might
Figure 2.Reactions of AzCou with excessive alkyne
concentrations. (A) The one-phase exponential increase of product
fluorescence when a limiting concentration of AzCou (50 µM) was
reacted with excess concentrations of 2-propyn-1-ol (0.25-4 mM).
(B-F) The one-phase exponential increase of product fluorescence
when a limiting concentration of AzCou (10 µM) was reacted with
excess concentrations of five phenylacetylene derivatives (0.1-1
mM). (G) Average apparent reaction rate constants for all tested
be considered. The first one involves the reaction between an
Figure 1. ADC conjugation strategies and small molecule mimicking
reactions. (A) The cysteine-maleimide reaction for
ADC
conjugation. (B) The conjugation of an azide-containing antibody to
an alkyne-containing drug. (C) The conjugation of an alkyne-
containing antibody to an azide-containing drug. (D-E) Two
fluorescence reactions that mimic the syntheses shown in B and C.
(F) The soluble Cu(I) ligand used in the current study.
alkynes.
Cu(I) totally lost its catalytic power before AzCou was
azide-containing antibody and an alkyne-containing drug (Figure
1B) and the second one between an alkyne-containing antibody
and an azide-containing drug (Figure 1C). For both processes, we
aimed to identify their most efficient processing conditions. To
facilitate analysis, we chose to characterize two mimicking
reactions as shown in Figure 1D-E, in which two fluorogenic
compounds, 3-Azido-7-hydroxycoumarin (AzCou) and 7-
Ethynylcoumarin (EtCou) become strongly fluorescent following
their reactions with an alkyne and azide, respectively. To
generate Cu(I) for the catalysis, we applied CuSO4 and sodium
ascorbate according to a published procedure21, 22 and preserved
Cu(I) by providing a strong Cu(I)-chelating ligand BTTAA
(Figure 1F).23, 24
consumed. Adding an additional 5 µM AzCou did not lead to any
fluorescence change to the reaction. This preliminary test
demonstrates that free Cu(I) cannot be directly applied for the
CuAAC-based ADC synthesis due to its rapid loss of activity in
water. However, BTTAA as a preserving reagent is able to
maintain the catalytic power of Cu(I) to sustain the time required
for ADC synthesis. We also compared catalytic activities of
different ratios of Cu(I) to BTTAA by varying the BTTAA
concentration and fixing all other conditions. It was evident that
the catalytic activity of Cu(I) improved when the Cu(I)-to-
BTTAA ratio was increased and reached its maximum when the
Cu(I)-to-BTTAA ratio was 1 (SI, Figure S2). To balance both the
preservation and the catalytic
To test the ability of BTTAA to preserve Cu(I), we set up two
parallel reactions in phosphate buffered saline (PBS) (pH 7.4),
with and without 300 µM BTTAA, that also contained 1 mM 2-
propyn-1-ol, 50 µM Cu(I), and 5 µM AzCou. The reaction in
which BTTAA was present was rapidly completed and the
addition of 5 µM AzCou consecutively, 4 times, over a time
period of 3 h led to similar reaction trajectories and fluorescence
intensity increases Supporting Information (SI, Figure S1),
indicating that BTTAA was capable of maintaining the catalytic
power of Cu(I) for a prolonged time. However, the reaction in
which BTTAA was absent was very slow. The final product
displayed fluorescence intensity that was much lower than that of
activity of Cu(I), we maintained the Cu(I)-to-BTTAA ratios in all
following studies at between 1:2 and 1:6.
To mimic the process of conjugating an azide-containing
antibody to an alkyne-containing drug under conditions of
limiting and excess reactants, respectively, during ADC
synthesis, we first characterized the reaction kinetics of limiting
AzCou with excessive 2-propyn-1-ol. We carried out reactions by
varying the concentration of 2-propyn-1-ol and fixed all other
conditions and maintained the 2-propyn-1-ol concentration at
least 5 times greater than that of AzCou for the purpose of
maintaining a pseudo first-order condition between the two
reactants. The strong fluorescence of the product allowed its easy

detection using a fluorimeter or fluorescent plate reader. For all
tested 2-propyn-1-ol concentrations, the collected data of
fluorescence intensity versus time fit perfectly to the one-phase
exponential production formation model, indicating that AzCou
followed first-order kinetics to convert to product under all tested
conditions (Figure 2A). However, unexpectedly the reaction
trajectories for all 2-propyn-1-ol concentrations were essentially
the same and all of the determined apparent reaction rate
constants were almost equal. We originally tested pseudo first
order conditions25 by varying the 2-propyn-1-ol concentration
with the aim of determining a second order rate constant between
an azide-alkyne bimolecular reaction under a fixed Cu(I)-ligand
condition. Apparently, the reaction did not follow a simple
bimolecular reaction model. The involvement of the Cu(I)
catalyst complicated the process. The independence of
determined apparent rate constants from 2-propyn-1-ol
concentrations might have been due to the formation of an
alkyne-Cu(I) resting state before the involvement of an azide, or
two parallel alkyne-Cu(I) and azide-Cu(I) chelating processes in
which the alkyne-Cu(I) chelation was significantly faster than the
other process.
we observed very different reaction trajectories (Figure 3A). For
comparison, a
metal-chelating azide, AzPy27 was also included in the analysis.
Overall and consistent with earlier studies, AzPy reacted more
rapidly than all other non-metal-chelating azides. For non-metal-
chelating azides, azidoethylamine reacted faster than phenylazide
and
phenylazide
reacted
faster
than
azidoethanol.
Azidoethylamine reacted more quickly than the other two
compounds, probably due to the chelating propensity of its amine
toward Cu(I). To confirm that the minimal dependence of the
reaction kinetics on the concentration of an excessive alkyne
when it reacts with a limiting non-metal-chelating azide is a
general phenomenon, we carried out analyses of reactions
between EtCou and different non-metal-chelating azides by
fixing the azide concentration and varying the EtCou
concentration. For azidoethanol, reactions were too slow to
determine k but had similar trajectories (SI, Figure S3). For
azidoethylamine and phenylazide, reactions displayed similar
kinetics at different EtCou concentrations (SI, Figures S4-5).
Figure 3. (A) CuAAC reaction trajectories between limiting
concentrations of different azides (20 µM) and excess
concentrations of EtCou (100 µM). (B) The one-phase exponential
increase of product fluorescence when a limiting concentration (5
µM) of AzPy was reacted with excess concentrations of EtCou
(50-150 M).
Based on all reactions we have analyzed between limiting non-
metal-chelating azides and excess alkyne concentrations, we
conclude that the minimal dependence of the reaction kinetics on
the concentration of an alkyne is a general observation.
AzCou is a non-metal-chelating azide. The independence of
the rate of its CuAAC reaction from the concentration of
excessive 2-propyn-1-ol might represent a general feature of
CuAAC reactions. To explore this possibility, we tested AzCou
reactions with five other alkynes in a setup similar to that for 2-
propyn-1-ol. We chose five phenylacetylenes due to their
significantly different electrochemical features of p-substitutes
that substantially influence their host compounds’ alkyne group.26
Should a simple bimolecular model apply, they would behave
very differently in reaction kinetics with AzCou. However, for all
five phenylacetylenes, their reaction trajectories at different
concentrations were very similar (Figure 2B-F). Although
determined apparent k values at different alkyne concentrations
showed some variations, they were very alike. They deviated far
from predictions from a real bimolecular reaction in which we
expected a linear dependence of a determined apparent k value on
the concentration of an excessively used reactant. The average k
values for all tested alkynes were also very similar, indicating
that the reaction kinetics were minimally dependent on both the
identity and concentration of excessive alkynes when they
reacted with limiting non-metal-chelating AzCou.
Since both the identity and concentration of an excessive
alkyne have minimal impacts on the CuAAC reaction kinetics of
non-metal-chelating AzCou, we reasoned that in a similar setup
the azide identity might significantly influence the reaction
kinetics. In order to compare reactivities of different azides under
the condition of a limiting azide and an excess alkyne
concentration, we used EtCou. When we reacted 100 µM EtCou
in excess with 20 µM concentrations of several different azides,
Figure 4. The one-phase exponential increase of product
fluorescence when a limiting concentration (20 M) of EtCou
reacted with excess concentrations (100-500 M) of azidoethylamine

We have also carried out a similar test to mimic how an
alkyne-containing drug may react with metal-chelating AzPy
conjugated to an antibody. We reacted limiting AzPy with
different concentrations of excessive EtCou and observed that the
EtCou concentration significantly influenced the reaction rate
In order to mimic the process of conjugating an alkyne-
containing antibody to an azide-containing drug that are present
as limiting and excess reactant concentrations, respectively, we
continued testing the reaction kinetics of limiting EtCou
concentration with excess azidoethylamine concentration under
pseudo first order conditions. By fixing the EtCou concentration
and varying the azidoethylamine concentration, we observed that
the azidoethylamine concentration significantly influenced the
reaction kinetics and the determined apparent rate constants were
linearly proportional to the azidoethylamine concentrations
(Figure 4A). Similarly, we tested the reaction of limiting EtCou
concentration with excess concentration of metal-chelating AzPy.
Intriguingly, reactions at all AzPy concentrations displayed
similar reaction trajectories and all determined k values were very
similar (Figure 4B), indicating minimal influence of the azide
concentration on the reaction kinetics under these conditions.
Although we used substantially higher concentrations of Cu(I) in
the EtCou-azidoethylamine reactions than in the EtCou-AzPy
reactions, the reaction at the highest azidoethylamine
concentration was still much slower than an EtCou-AzPy
reaction, indicating that under these conditions a metal-chelating
drug azide remains an optimal choice for favorable kinetics. We
(Figure 3B). The determined k values were linearly proportional
to EtCou trajectories. It is obvious that metal-chelating and non-
metal-chelating azides involve very different CuAAC reaction
kinetics. The AzPy-Cu(I) chelation is apparently able to
outcompete the rapid alkyne-Cu(I) formation, leading to the
observed dependence of the alkyne concentration. Based on all
the data related to limiting azides reacting with excessive
alkynes, we can deduce that for the reaction between an azide-
containing antibody and an alkyne-containing drug to generate an
ADC, the most kinetically favorable option is to react a metal-
chelating antibody azide with a drug alkyne and improve the
kinetics by increasing the drug concentration. When a non-metal-
chelating antibody azide is used, changing the drug concentration
has little impact on the reaction kinetics and more favorable
reaction kinetics can be achieved by changing the azide identity
in the antibody for example, from a standard alkyl azide to an
aromatic azide.
Figure 5. Catch-and-release strategy to synthesize ADCs
might also use the alkyne-Cu(I) resting state theory or two
parallel alkyne-Cu(I) and azide-Cu(I) chelating processes to
explain the observations in Figure 4. When a non-metal-chelating
azide is used, the alkyne-Cu(I) chelation is so fast that the kinetic
dependence on the azide is observable. However, when a metal-
chelating azide is used, the azide-Cu(I) chelation outcompetes the
alkyne-Cu(I) chelation so that the reaction kinetics is solely
determined by the alkyne-Cu(I) chelation. The implication of the
data shown in Figure 4 for ADC generation is obvious. For the
reaction between an alkyne-containing antibody and an azide-
containing drug to generate ADCs, the most kinetically favorable
option is to react an antibody alkyne with a metal-chelating drug
azide although changing the drug concentration does not
influence the overall kinetics. When a non-metal-chelating drug
azide is used, improving the drug concentration will significantly
increase the reaction kinetics and one may also achieve more
favorable reaction kinetics by changing the azide identity in the
drug from a standard alkyl azide to an aromatic azide.
Next, we wanted to verify that the conditions used in small
molecule kinetics can be translated to generate protein conjugates
as well. In order to demonstrate this, we built a model system that
allows us to monitor protein conjugation kinetics by FRET
analysis. We site-specifically incorporated L-propargyl-lysine
(ProcK)28 into superfolder green fluorescent protein (sfGFP) to
generate our model protein (sfGFP134ProcK). We synthesized
compound (16) which is a metal-chelating azide with coumarin

dye moiety. Therefore, chromophore of sfGFP and coumarin dye
can form a FRET pair upon conjugation (SI, Figure S12). After
our first conjugation attempt with 10 µM CuSO4 and 20 µM
BTTAA, we did not detect a FRET signal. This might be due to
the presence of 6xHistag on the sfGFP134ProcK or another metal
binding motif. Next, we tried the reaction with 40 µM CuSO4 and
80 µM BTTAA and the reaction appeared to reach a plateau in 10
minutes (SI, Figure S13). This result showed us that, from a
kinetics point of view, it is reasonable to keep the concentration
of CuSO4 around 50 µM when carrying out click reactions with
proteins.
mutated version of pertuzumab (pertuzumab-2C) which only has
one reducible disulfide bond in the hinge region.30 We first
reduced pertuzumab-2C under mild conditions with TCEP then,
reacted it with 4 mM bromoacetyl alkyne (3) for 30 mins to load
the alkyne group. Next, we incubated the antibody with protein
A-Sepharose and for 2 hrs to capture the alkynylated antibody on
the resin. We subsequently washed out residual alkyne
compound and provided 1.2 eq of AzPy-Payload (Py-Az-Val-Cit-
PAB-MMAE, 14) compound per alkynylated cysteine (there are
2 alkynylated cysteines for each antibody molecule) in the
antibody (see Experimental Methods section for details). The
provided AzPy-payload concentration was deemed to be in
excess. Following 2 hrs reaction, we washed out the residual
AzPy-payload and eluted the conjugated pertuzumab. The overall
preparation time of the ADC is 9 hours. We analyzed the eluted
ADC using electrospray ionization mass spectrometry (SI, Figure
S17) that showed close to quantitative conversion, demonstrating
the high efficiency of this newly designed ADC synthetic route.
The addition of the drug molecule to heavy chain was identified
by the change in mass and the shift in retention time of the heavy
chain. The conditions and chromatogram for the LC-MS analysis
can be found in the SI.
Table 1. Relationships of different alkyne-azide combinations
for the ADC synthesis
Antibody
Drug
Reaction
Effect of drug
concentration on the
reaction rate
Non-metal-
Alkyne
Alkyne
Slow
Fast
Minimal
Linear
chelating azide
Metal-chelating
azide
Alkyne
Alkyne
Non-metal-
Slow
Fast
Linear
chelating azide
Metal-chelating
azide
Minimal
3. Conclusions
In conclusion, we have kinetically characterized CuAAC-
based ADC conjugation processes using small molecule mimics.
Our data reveal unique kinetic behaviors when different azide-
alkyne combinations are used. During an ADC conjugation setup
in which a limiting antibody concentration reacts with an excess
drug concentration, our data indicates that a drug alkyne reacts
slowly with a non-metal-chelating antibody azide and its
concentration has minimal influence on the overall reaction
kinetics, however it reacts rapidly with a metal-chelating
antibody azide and its concentration is linearly proportional to
the reaction rate constant. When a limiting antibody alkyne
concentration is reacted with an excess concentration of azide-
containing drug, our data implies that it reacts slowly when the
drug azide is non-metal-chelating but the drug concentration is
linearly proportional to the reaction rate constant. However, its
reaction with a metal-chelating drug azide is exceedingly fast but
not kinetically related to the drug concentration. Based on these
observations, we designed an efficient and cost effective
CuAAC-based ADC conjugation method and demonstrated that
an ADC can be rapidly synthesized. The CuAAC-based linker is
more stable than a traditional cysteine-maleimide linker leading
to much lower non-targeted drug release.20 Given the broad
applications31 of CuAAC in many basic research and
technological areas, we believe our kinetic observations will also
assist widely in optimizing conditions for these applications.
Based on all of the kinetic observations, we can summarize
kinetic features of the CuAAC-based ADC synthesis as shown in
Table 1. Given the nature of antibodies, an ADC conjugation
procedure that requires a short reaction time and facile
purification is optimal. For achieving a short reaction time, a
metal-chelating azide either on an antibody or a drug is an
apparent choice. When a metal-chelating antibody azide is used,
one may also improve the reaction kinetics by increasing the drug
alkyne concentration. However, loading a metal-chelating azide
onto an antibody either through genetic incorporation or
conjugation to antibody thiolates
will require conditions that may reduce the relatively labile azide
functionality, leading to heterogeneity in the conjugated ADC.
On the contrary, an antibody alkyne is far more stable. A cheaply
produced alkyne such as bromoacetyl alkyne (compound 3 in
Figure 5) can be rapidly coupled to an antibody by reaction with
cysteine thiols. Using a catch-and-release method, exceedingly
excessive bromoacetyl alkyne (3) can be added to achieve rapid
synthesis of an alkyne-containing antibody and the residual
alkyne compound can be quickly washed away. The coupled
antibody alkyne can then react with a metal-chelating drug azide
for the efficient synthesis of the ADC. Although the
independence of the reaction kinetics during these conditions on
the drug concentration does not allow the improvement of
kinetics by increasing the drug concentration, it does provide a
manufacturing benefit. This unique kinetic behavior means that
only a minimally excessive metal-chelating azide is required for
efficient ADC conjugation. Therefore, substantial manufacturing
costs can be reduced with very low amounts of cytotoxic waste.
To demonstrate all of these advantages related to reacting an
antibody alkyne with a metal-chelating drug azide, we designed a
catch-and-release synthetic route as shown in Figure 5.
4. Experimental Methods
Organic synthesis procedures, on-protein kinetics, LC-MS data
for characterization of ADCs can be found in the Supplementary
Information.
4.1 General method for kinetics measurement
Stock solutions for all azide and alkyne compounds were
prepared in 100x working concentrations in DMSO. BTTAA
stock solutions were 250x the working concentration in DMSO.
CuSO4 stock solutions were 100x the working concentration.
250 mM sodium ascorbate stock solution in water was used in all
analysis.
0.15M phosphate buffer, pH 7.0, 25% DMSO is added to the
reaction well (or cuvette). Next, BTTAA, CuSO4 were added
successively followed by azide derivative. After that, ascorbate
In our study, we used the HER2 specific antibody,
pertuzumab29 as a model antibody. Pertuzumab is a monoclonal
antibody used for the treatment of metastatic HER2-positive
breast cancer. Wild-type pertuzumab has two disulfide bonds
bridging the two heavy chains in the hinge region, and two
disulfide bonds bridging each heavy chain hinge region with the
two antibody light chains. These disulphides are reducible under
mild conditions. To reduce the drug to antibody ratio, we used a

was added and the mixture was incubated for 60 seconds. Finally,
alkyne was added to initiate the click reaction. After addition of
each reagent, the reaction mixture was gently mixed by pipetting
up and down. Unless otherwise stated, a plate reader was used for
all measurements. For AzCou excitation at 404 nM and emission
at 480 nM was used. For EtCou excitation at 320 nM and
emission at 400 nM was used. Data were fit to the equation: It =
Io + Ic × (1- exp(-k × t)) where I0 is the background fluorescent
intensity, It the fluorescent intensity at a specified time, Ic the
fluorescent intensity change when limiting reagent is fully
converted to the product, and k the observed apparent rate
constant.
The pertuzumab antibody used in this study (Per-2C) has 2
reducible cysteines in its hinge region.30
First, 100 µL 27 µM of Per-2C in PBS buffer was reduced by
200 µM TCEP at 37oC for 2 hrs. Then, bromoacetyl alkyne
compound (3) (400 mM stock solution in DMSO) was added to
the solution at a final concentration of 4 mM. The reaction was
incubated at room temperature for 30 minutes. Next, 100 µL
Protein A-Sepharose resin slurry in PBS was added to this
reaction mixture. The mixture was incubated on a shaker at room
temperature for 2 hours.
After that, an empty spin column was used for the washing step.
After initial centrifugation, 300 µL PBS were added to the slurry
and the resin washed 5 times with 300 µL PBS. All the
centrifugations were carried out at 1000g for 1 min at room
temperature.
Table 2. AzCou with excess alkyne concentrations (except 2-
propyn-1-ol where 50 µM CuSO4, 300 µM BTTAA and 50 µM
AzCou, 2.5 mM sodium ascorbate used with 0.25, 0.5, 1, 2, 4
mM 2-propyn-1-ol)
After the final centrifugation step, the total slurry volume was
brought to 100 µL.
Reagent
CuSO4
BTTAA
AzCou
Sodium ascorbate
Alkyne Derivative
Final Concentration
50 µM
100 µM
10 µM
2.5 mM
100, 250, 500, 750, 1000 µM
Table 7. The 2x click reaction mixture (200 L reaction mix)
2 x Click Reaction mixture
172 µL PBS
16.7 µL DMSO
1.6 µL BTTAA (25 mM stock solution in DMSO)
4 µL Cu(II)SO4 (5mM stock solution in sterile water)
1.3 µL azide (10 mM stock solution in DMSO)
4 µL sodium ascorbate (250 mM stock solution in sterile water)
Table 3. CuAAC reaction trajectories between different limiting
azide concentrations and excess EtCou concentrations
Reagent
CuSO4
BTTAA
Azide derivative
Sodium ascorbate
7-Ethynylcoumarin 100 µM
Final Concentration
50 µM
100 µM
20 µM
2.5 mM
The above mixture (2x click reaction mixture) was prepared in
the given order and incubated for 1 min after sodium ascorbate
addition. Next, 100 µL of the above solution was added to 100
µL resin suspension. The tube was covered in aluminum foil and
placed on a shaker at room temperature for 2 hours.
Next, the slurry was added to a spin column for the final
washing. After initial centrifugation, the resin was washed with 3
x 300 µL PBS containing 1mM EDTA. For the fourth and fifth
washes, 1 mM phosphate buffer, pH 7.4, with 150 mM NaCl
used reduce the trace amounts of phosphate in the slurry.
ADC was eluted from protein A-Sepharose using cidic 0.1M
glycine, 150 mM NaCl, pH 2.8 buffer for ~30 seconds with
gentle shaking.
Table 4. Limiting AzPy concentration reacted with excess EtCou
concentrations (fluorometer was used)
Reagent
CuSO4
BTTAA
AzPy
Final Concentration
10 µM
20 µM
5 µM
Sodium ascorbate
7-Ethynylcoumarin 50, 75, 100, 125, 150 µM
2.5 mM
Three sequential elutions (90 µL each) were carried out to elute
the ADCs. Each 90 µL eluent was neutralized immediately by
adding 15 µL 2M Tris, 150 mM NaCl, pH 8.0.
The three eluates were combined and EDTA added to a final
concentration of 1 mM, followed by dialysis against PBS. The
ADCs were characterized by LC-MS (SI, Figure S16, S17).
Table 5. EtCou reacted with excess AzPy concentrations
Reagent
CuSO4
Final Concentration
10 µM
BTTAA
AzPy
20 µM
Acknowledgments
100, 200, 300, 400, 500 µM
2.5 mM
Sodium ascorbate
7-Ethynylcoumarin 20 µM
This work was supported in part by Cancer Prevention and
Research Institute of Texas (grants RP170797, RP140141) and
Welch Foundation (grant A-1715). We thank Sunshine Z.
Leeuwon for proofreading the manuscript.
Table 6. EtCou reacted with excess azidoethylamine
concentrations
Reagent
CuSO4
Final Concentration
10 µM
References
BTTAA
20 µM
1. Lambert, J. M.; Berkenblit, A., Antibody-Drug Conjugates for
Cancer Treatment. Annu Rev Med 2018, 69, 191-207.
2. clinicaltrials.gov, US National Library of Medicine. 2020.
3. Khongorzul, P.; Ling, C. J.; Khan, F. U.; Ihsan, A. U.;
2-azidoethylamine 100, 200, 300, 400, 500 µM
Sodium ascorbate 2.5 mM
7-Ethynylcoumarin 20 µM
Zhang, J., Antibody-Drug Conjugates:
A Comprehensive
Review. Mol Cancer Res 2020, 18 (1), 3-19.
4.2 Catch-and-release method

4. Leung, D.; Wurst, J. M.; Liu, T.; Martinez, R. M.; Datta-
Mannan, A.; Feng, Y. Q., Antibody Conjugates-Recent Advances
and Future Innovations. Antibodies 2020, 9 (1).
5. Birrer, M. J.; Moore, K. N.; Betella, I.; Bates, R. C.,
Antibody-Drug Conjugate-Based Therapeutics: State of the
Science. Jnci-J Natl Cancer I 2019, 111 (6), 538-549.
6. Saito, F.; Noda, H.; Bode, J. W., Critical evaluation and rate
constants of chemoselective ligation reactions for stoichiometric
conjugations in water. ACS Chem Biol 2015, 10 (4), 1026-33.
7. Sharpless, N. E.; Flavin, M., The reactions of amines and
amino acids with maleimides. Structure of the reaction products
deduced from infrared and nuclear magnetic resonance
spectroscopy. Biochemistry 1966, 5 (9), 2963-71.
8. Ponte, J. F.; Sun, X. X.; Yoder, N. C.; Fishkin, N.; Laleau, R.;
Coccia, J.; Lanieri, L.; Bogalhas, M.; Wang, L. T.; Wilhelm, S.;
Widdison, W.; Pinkas, J.; Keating, T. A.; Chari, R.; Erickson, H.
K.; Lambert, J. M., Understanding How the Stability of the
Thiol-Maleimide Linkage Impacts the Pharmacokinetics of
Lysine-Linked Antibody-Maytansinoid Conjugates. Bioconjugate
Chem 2016, 27 (7), 1588-1598.
9. Lyon, R. P.; Setter, J. R.; Bovee, T. D.; Doronina, S. O.;
Hunter, J. H.; Anderson, M. E.; Balasubramanian, C. L.; Duniho,
S. M.; Leiske, C. I.; Li, F.; Senter, P. D., Self-hydrolyzing
maleimides improve the stability and pharmacological properties
of antibody-drug conjugates. Nat Biotechnol 2014, 32 (10), 1059-
+.
10. Schumacher, F. F.; Nunes, J. P. M.; Maruani, A.; Chudasama,
V.; Smith, M. E. B.; Chester, K. A.; Baker, J. R.; Caddick, S.,
Next generation maleimides enable the controlled assembly of
antibody-drug conjugates via native disulfide bond bridging. Org
Biomol Chem 2014, 12 (37), 7261-7269.
11. Badescu, G.; Bryant, P.; Bird, M.; Henseleit, K.; Swierkosz,
J.; Parekh, V.; Tommasi, R.; Pawlisz, E.; Jurlewicz, K.; Farys,
M.; Camper, N.; Sheng, X. B.; Fisher, M.; Grygorash, R.; Kyle,
A.; Abhilash, A.; Frigerio, M.; Edwards, J.; Godwin, A.,
Bridging Disulfides for Stable and Defined Antibody Drug
Conjugates. Bioconjugate Chem 2014, 25 (6), 1124-1136.
12. Drake, P. M.; Albers, A. E.; Baker, J.; Banas, S.; Barfield, R.
M.; Bhat, A. S.; de Hart, G. W.; Garofalo, A. W.; Holder, P.;
Jones, L. C.; Kudirka, R.; McFarland, J.; Zmolek, W.; Rabuka,
D., Aldehyde Tag Coupled with HIPS Chemistry Enables the
Production of ADCs Conjugated Site-Specifically to Different
Antibody Regions with Distinct in Vivo Efficacy and PK
Outcomes. Bioconjugate Chem 2014, 25 (7), 1331-1341.
13. Kularatne, S. A.; Deshmukh, V.; Ma, J.; Tardif, V.; Lim, R.
K. V.; Pugh, H. M.; Sun, Y.; Manibusan, A.; Sellers, A. J.;
Barnett, R. S.; Srinagesh, S.; Forsyth, J. S.; Hassenpflug, W.;
Tian, F.; Javahishvili, T.; Felding-Habermann, B.; Lawson, B.
R.; Kazane, S. A.; Schultz, P. G., A CXCR4-Targeted Site-
Specific Antibody-Drug Conjugate. Angew Chem Int Edit 2014,
53 (44), 11863-11867.
18. Yamada, K.; Ito, Y., Recent Chemical Approaches for Site-
Specific Conjugation of Native Antibodies: Technologies toward
Next-Generation Antibody-Drug Conjugates. Chembiochem
2019, 20 (21), 2729-2737.
19. Zimmerman, E. S.; Heibeck, T. H.; Gill, A.; Li, X. F.;
Murray, C. J.; Madlansacay, M. R.; Tran, C.; Uter, N. T.; Yin,
G.; Rivers, P. J.; Yam, A. Y.; Wang, W. D.; Steiner, A. R.;
Bajad, S. U.; Penta, K.; Yang, W. J.; Hallam, T. J.; Thanos, C.
D.; Sato, A. K., Production of Site-Specific Antibody-Drug
Conjugates Using Optimized Non-Natural Amino Acids in a
Cell-Free Expression System. Bioconjugate Chem 2014, 25 (2),
351-361.
20. VanBrunt, M. P.; Shanebeck, K.; Caldwell, Z.; Johnson, J.;
Thompson, P.; Martin, T.; Dong, H. F.; Li, G. R.; Xu, H. Y.;
D'Hooge, F.; Masterson, L.; Bariola, P.; Tiberghien, A.; Ezeadi,
E.; Williams, D. G.; Hartley, J. A.; Howard, P. W.; Grabstein, K.
H.; Bowen, M. K.; Marelli, M., Genetically Encoded Azide
Containing Amino Acid in Mammalian Cells Enables Site-
Specific Antibody-Drug Conjugates Using Click Cycloaddition
Chemistry. Bioconjugate Chem 2015, 26 (11), 2249-2260.
21. Hong, V.; Presolski, S. I.; Ma, C.; Finn, M. G., Analysis and
Optimization of Copper-Catalyzed Azide-Alkyne Cycloaddition
for Bioconjugation. Angew Chem Int Edit 2009, 48 (52), 9879-
9883.
22. Wang, Q.; Chan, T. R.; Hilgraf, R.; Fokin, V. V.; Sharpless,
K. B.; Finn, M. G., Bioconjugation by copper(I)-catalyzed azide-
alkyne [3+2] cycloaddition. J Am Chem Soc 2003, 125 (11),
3192-3193.
23. Besanceney-Webler, C.; Jiang, H.; Zheng, T. Q.; Feng, L.;
del Amo, D. S.; Wang, W.; Klivansky, L. M.; Marlow, F. L.; Liu,
Y.; Wu, P., Increasing the Efficacy of Bioorthogonal Click
Reactions for Bioconjugation: A Comparative Study. Angew
Chem Int Edit 2011, 50 (35), 8051-8056.
24. del Amo, D. S.; Wang, W.; Jiang, H.; Besanceney, C.; Yan,
A. C.; Levy, M.; Liu, Y.; Marlow, F. L.; Wu, P., Biocompatible
Copper(I) Catalysts for in Vivo Imaging of Glycans. J Am Chem
Soc 2010, 132 (47), 16893-16899.
25. Rodionov, V. O.; Fokin, V. V.; Finn, M. G., Mechanism of
the ligand-free CuI-catalyzed azide-alkyne cycloaddition
reaction. Angew Chem Int Ed Engl 2005, 44 (15), 2210-5.
26. Penoni, A.; Palmisano, G.; Zhao, Y. L.; Houk, K. N.;
Volkman, J.; Nicholas, K. M., On the Mechanism of
Nitrosoarene-Alkyne Cycloaddition. J Am Chem Soc 2009, 131
(2), 653-661.
27. Uttamapinant, C.; Tangpeerachaikul, A.; Grecian, S.; Clarke,
S.; Singh, U.; Slade, P.; Gee, K. R.; Ting, A. Y., Fast, Cell-
Compatible Click Chemistry with Copper-Chelating Azides for
Biomolecular Labeling. Angew Chem Int Edit 2012, 51 (24),
5852-5856.
28. Nguyen, D. P.; Lusic, H.; Neumann, H.; Kapadnis, P. B.;
Deiters, A.; Chin, J. W., Genetic Encoding and Labeling of
Aliphatic Azides and Alkynes in Recombinant Proteins via a
Pyrrolysyl-tRNA Synthetase/tRNA(CUA) Pair and Click
Chemistry. J Am Chem Soc 2009, 131 (25), 8720-+
29. Franklin, M. C.; Carey, K. D.; Vajdos, F. F.; Leahy, D. J.;
de Vos, A. M.; Sliwkowski, M. X., Insights into ErbB signaling
from the structure of the ErbB2-pertuzumab complex. Cancer
Cell 2004, 5 (4), 317-328.
14. Maruani, A.; Smith, M. E. B.; Miranda, E.; Chester, K. A.;
Chudasama, V.; Caddick, S., A plug-and-play approach to
antibody-based therapeutics via a chemoselective dual click
strategy. Nat Commun 2015, 6.
15. Ohata, J.; Ball, Z. T., A Hexa-rhodium Metallopeptide
Catalyst for Site-Specific Functionalization of Natural
Antibodies. J Am Chem Soc 2017, 139 (36), 12617-12622.
16. Oller-Salvia, B.; Kym, G.; Chin, J. W., Rapid and Efficient
Generation of Stable Antibody-Drug Conjugates via an Encoded
Cyclopropene and an Inverse-Electron-Demand Diels-Alder
Reaction. Angew Chem Int Edit 2018, 57 (11), 2831-2834.
17. Vinogradova, E. V.; Zhang, C.; Spokoyny, A. M.; Pentelute,
B. L.; Buchwald, S. L., Organometallic palladium reagents for
cysteine bioconjugation. Nature 2015, 526 (7575), 687-691.
30. Kang, J. C.; Sun, W.; Khare, P.; Karimi, M.; Wang, X. L.;
Shen, Y.; Ober, R. J.; Ward, E. S., Engineering a HER2-specific
antibody-drug conjugate to increase lysosomal delivery and
therapeutic efficacy. Nat Biotechnol 2019, 37 (5), 523-+.
31. Jiang, X. Y.; Hao, X.; Jing, L. L.; Wu, G. C.; Kang, D. W.;
Liu, X. Y.; Zhan, P., Recent applications of click chemistry in
drug discovery. Expert Opin Drug Dis 2019, 14 (8), 779-789.

Supplementary Material
Supplementary material is available.

Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests: