Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides

Article abstract of DOI:10.1016/j.jconrel.2015.09.032

Maleimides are often used to covalently attach drugs to cysteine thiols for production of antibody-drug conjugates (ADCs). However, ADCs formed with traditional N-alkyl maleimides have variable stability in the bloodstream leading to loss of drug. Here, w

Full text of DOI:10.1016/j.jconrel.2015.09.032

COREL-07872; No of Pages 11
Journal of Controlled Release xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Journal of Controlled Release
journal homepage: www.elsevier.com/locate/jconrel
Stabilization of cysteine-linked antibody drug conjugates with
N-aryl maleimides
a
a
a
b
b
R. James Christie a, , Ryan Fleming , Binyam Bezabeh , Rob Woods , Shenlan Mao , Jay Harper ,
⁎
Augustine Joseph c, Qianli Wang c, Ze-Qi Xu c, Herren Wu a, Changshou Gao a, Nazzareno Dimasi a,
Antibody Discovery and Protein Engineering, MedImmune, Gaithersburg, MD 20878, USA
⁎
a
b
Oncology Research, MedImmune, Gaithersburg, MD 20878, USA
c
SynChem, Inc., Elk Grove Village, IL 60007, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 18 July 2015
Received in revised form 10 September 2015
Accepted 17 September 2015
Available online xxxx
Maleimides are often used to covalently attach drugs to cysteine thiols for production of antibody–drug conju-
gates (ADCs). However, ADCs formed with traditional N-alkyl maleimides have variable stability in the blood-
stream leading to loss of drug. Here, we report that N-aryl maleimides form stable antibody conjugates under
very mild conditions while also maintaining high conjugation efficiency. Thiol–maleimide coupling and ADC
stabilization via thiosuccinimide hydrolysis were accelerated by addition of N-phenyl or N-fluorophenyl groups
to the ring-head nitrogen. Cysteine-linked ADCs prepared with N-aryl maleimides exhibited less than 20%
deconjugation in both thiol-containing buffer and serum when incubated at 37 °C over a period of 7 days, where-
as the analogous ADCs prepared with N-alkyl maleimides showed 35–67% deconjugation under the same condi-
tions. ADCs prepared with the anticancer drug N-phenyl maleimide monomethyl-auristatin-E (MMAE)
maintained high cytotoxicity following long-term exposure to serum whereas the N-alkyl maleimide MMAE
ADC lost potency over time. These data demonstrate that N-aryl maleimides are a convenient and flexible
platform to improve the stability of ADCs through manipulation of functional groups attached to the maleimide
ring-head nitrogen.
Keywords:
Antibody-drug conjugate
Maleimide
Retro-Michael reaction
Serum stability
Thiol conjugation
Thiosuccinimide hydrolysis
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
[7]. For therapeutic applications such ADCs, stable drug attachment
is critical in order to achieve predictable drug distribution and clinical
Michael addition of thiols to maleimides is a popular conjugation
modality to produce cysteine-linked antibody-drug conjugates (ADCs)
due to fast reaction kinetics, specificity, and ease of incorporation into
drug molecules [1–3]. However, thiol-linked ADCs prepared with
N-alkyl maleimides have variable stability depending on the conjuga-
tion site and drug may be released from the carrier by a retro-Michael
reaction [4–8]. Stability of maleimide–antibody conjugates depends on
factors such as local charge, solvent accessibility and drug properties
efficacy. Deconjugation of drug can result in reduced on-target activity,
increased off-target toxicity and reduced overall performance of an
ADC [7–9].
Several methods can be employed to generate stable thiol con-
jugates using both maleimide-based or alternative chemistries. One
non-maleimide approach is to utilize nucleophilic displacement of
haloacetamides to generate stable thioethers. However, this reaction
requires basic pH [4,10,11], reaction kinetics are much slower than
maleimide-based conjugations [12], and the haloacetamide functional
group may impact the solubility of large non-polar ADC payloads.
Other non-maleimide chemistries such as 3-aryl propiolonitriles [12],
benzyl sulfones [13], and bromopyridazinediones [14] have been devel-
oped to overcome thiol conjugate instability and initial results show
promise for ADC applications.
Maleimide-based conjugate stabilization strategies aim to stop
the retro-Michael reaction by hydrolyzing (and ring-opening) the
thiosuccinimide, which is formed after reaction with a thiol [5,8,
15–18]. This concept has recently been applied to produce stable
thiol–maleimide antibody conjugates by subjecting them to increased
temperature and/or pH after conjugation [8], use of catalysts [18], or
chemically designing maleimides to contain functional groups that
Abbreviations: ADC, antibody-drug conjugate; MMAE, monomethyl-auristatin-E; PAB,
p-aminobenzyloxy; mAb, monoclonal antibody; PEG, poly(ethylene glycol); MWCO,
molecular weight cut-off; HPLC, high performance liquid chromatography; Q-TOF,
Quadrupole-time-of-flight; RP-HPLC, reverse phase high performance liquid chromatog-
raphy; MS, mass spectrometry; NMR, nuclear magnetic resonance; ESI, electrospray ioni-
zation; CHT, ceramic hydroxyapatite; TLC, thin-layer chromatography; PBS, phosphate
buffered saline; BME, β-mercaptoethanol; DTT, dithiothreitol; AcOH, acetic acid; EtOAc,
ethyl acetate; NHS, N-hydroxy succinimide; DCC, dicyclohexylcarbodiimide; DMAc,
dimethylacetamide; TCEP, (tris(2-carboxyethyl)phosphine), HOBt, N-hydroxybenzotriazole;
DMSO, dimethylsulfoxide; A280, absorbance at 280 nm; CTG, CellTiter-Glo.
⁎
Corresponding authors at: MedImmune, 1 MedImmune Way, Gaithersburg, MD
20878, USA.
E-mail addresses: christier@medimmune.com (R.J. Christie),
dimasin@medimmune.com (N. Dimasi).
http://dx.doi.org/10.1016/j.jconrel.2015.09.032
0168-3659/© 2015 Elsevier B.V. All rights reserved.
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

2
R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
accelerate thiosuccinimide hydrolysis [9,16,17]. Stabilization of thiol–
maleimide conjugates through chemically designed maleimides re-
quires careful balance, as chemistry that promotes thiosuccinimide hy-
drolysis after conjugation will also promote maleimide hydrolysis
before conjugation. Maleamic acid derivatives formed after maleimide
hydrolysis are unreactive towards thiols [19] and excessive hydrolysis
before conjugation would require more equivalents of maleimide-
drug to achieve complete conjugation.
2.2. Synthesis of 3-maleimidopropionyl-PEG7-biotin (1)
First, 3-maleimidopropionic acid was synthesized according to
described literature methods with modification [26,27]. A suspen-
sion of maleic anhydride (3.21 g, 32.78 mmol) and β-alanine (3 g,
33.67 mmol) in glacial AcOH (30 mL) was heated to reflux (bath tem-
perature: 170–180 °C) for 90 min. The solution was cooled to room tem-
perature and the solvent was evaporated in vacuo. Residual AcOH was
removed by coevaporation with toluene under vacuum (2 × 50 mL).
The residue was treated with water (100 mL) and EtOAc (200 mL).
The organic layer was separated while the aqueous layer was extracted
with EtOAc (2 × 100 mL). The combined organic extracts were dried
(Na2SO4) and evaporated in vacuo. The crude product was purified by
silica gel column chromatography, eluting with 50% ethyl acetate in
hexanes, to afford 3-maleimidopropionic acid (3.7 g, 21.88 mmol, 67%
yield), the spectroscopic data of which was consistent with reported
data.(27) 1H NMR (CDCl3) δ 8.93 (br s, 1H), 6.71 (s, 2H), 3.80 (t, J =
7.2 Hz, 2H), 3.62–3.66 (m, 24H), 2.67 (t, J = 7.2 Hz, 2H).
Next, 3-maleimidopropionic acid was esterified by adding N-
hydroxysuccinimide (NHS) (23 mg, 0.20 mmol) and N,N′-
dicyclohexylcarbodiimide (DCC) (50 mg, 0.24 mmol) into a stirred
solution of 3-maleimidopropionic acid (34 mg, 0.20 mmol) in
dimethoxyethane (5 mL) at room temperature under N2. After 1 h,
the precipitate formed was removed by filtration. The filtrate was
concentrated under vacuum to afford crude 3-maleimidopropionic
NHS ester, which was used directly for the next step without further
purification.
The biotinylated product was obtained by nucleophilic dis-
placement of NHS with the amine group of biotin-PEG7 amine. 3-
Maleimidopropionic acid NHS ester, prepared above, was added into
a stirred solution of biotin-PEG7 amine (100 mg, 0.168 mmol) in anhy-
drous dichloromethane (7 mL) at room temperature under N2, The reac-
tion continued with stirring at room temperature and was monitored by
TLC, which showed complete reaction after 1 h. Next, 1 N HCl (2 mL)
was added, the organic solution separated, washed with brine, and
then dried over Na2SO4. The crude product obtained after evaporation
under vacuum was purified by flash column chromatography on
silica gel, eluting with step gradients of MeOH in dichloromethane
(v/v = 1:30, 1:20 and 1:10) to afford the target compound
maleimidoproprionyl-PEG7-biotin (1) (65 mg, 0.087 mmol, 52% yield)
as a waxy solid. 1H NMR (CDCl3) δ 7.01 (br s, 1H), 6.96 (br s, 1H), 6.70
(s, 2H), 5.86 (s, 1H), 5.06 (s, 1H), 4.51 (m, 1H), 4.33 (m, 1H), 3.84
(t, J = 7.2 Hz, 2H), 3.62–3.66 (m, 24H), 3.53 (m, 4H), 3.40 (m, 4H), 3.15
(m, 1H), 2.88–2.96 (m, 1H), 2.73 (d, J = 12.6 Hz, 1H), 2.52 (t, J = 7.2 Hz,
2H), 2.24 (t, J = 7.0 Hz, 2H), 1.66 (m, 4H), 1.45 (m, 2H). MS (ESI+) m/z
769 (M + Na). A reaction scheme is provided in the Supplementary
information section (Scheme S1).
We chose to focus on a maleimide-based strategy for ADC stabiliza-
tion in this work, specifically a simple and practical approach based on
known N-aryl maleimide chemistry [20–25]. The goal of our approach
was to produce stable antibody conjugation products while also main-
taining (or increasing) conjugation efficiency. N-aryl maleimides and
their thiosuccinimide reaction products should be more susceptible to
hydrolysis due to resonance delocalization [22] as shown in Fig. 1. For
N-alkyl maleimides, nitrogen lone pair electrons resonate between
the nitrogen and carbonyl groups, thus decreasing the reactivity of
carbonyls towards nucleophilic attack by water (resonance structure
1). A phenyl ring, however, provides additional resonance structures
to nitrogen lone pair electrons through the conjugated pi-system,
which leaves the carbonyl group susceptible to hydrolysis through
decreased electron density on the carbonyl carbon. In this proposed
resonance-based mechanism for thiosuccinimide hydrolysis, addition
of electron withdrawing groups to the phenyl ring should further
increase the contribution of resonance structure 2 and increase the
hydrolysis rate. This reasoning can also be extended to reactivity of
maleimides, as N-aryl functionality makes carbonyl groups in the
maleimide ring available to accept electrons in the 1,4-Michael addition
reaction.
N-phenyl maleimides have been extensively used for protein conju-
gation and crosslinking applications, however, their application towards
producing stable antibody conjugates through thiosuccinimide hydroly-
sis has not been demonstrated. Here, we report that N-aryl maleimides
are well-suited for ADC applications as they chemically stabilize thiol-
linked drug conjugates, maintain high antibody conjugation efficiency
and allow flexibility in maleimide design through phenyl substitution.
2. Materials and methods
2.1. General
All materials were obtained from commercial sources and used
without purification unless noted. Reaction schemes can be found in
the supplemental information section.
2.3. Synthesis of [2-(4-maleimidophenyl)]acetyl-PEG7-biotin (2)
First, [2-(4-maleimidophenyl)] acetic acid was synthesized accord-
ing to the literature method with modifications [27]. A suspension of
maleic anhydride (3.21 g, 32.78 mmol) and [2-(4-aminophenyl)] acetic
acid (5 g, 33.08 mmol) in glacial AcOH (75 mL) was heated to reflux
(bath temperature: 170–180 °C) for 90 min. The solution was cooled
to room temperature and the solvent was evaporated in vacuo. Residual
AcOH was removed by coevaporation with toluene under vacuum
(2 × 50 mL). The residue was treated with water (100 mL) and EtOAc
(200 mL). The organic layer was separated and the aqueous layer
extracted with EtOAc (2 × 100 mL). The combined organic extracts
were dried (Na2SO4) and evaporated in vacuo. The crude product was
purified by silica gel column chromatography, eluting with 50% ethyl
acetate in hexanes, to afford [2-(4-maleimidophenyl)] acetic acid
(6.0 g, 25.95 mmol, 79% yield), the spectroscopic data of which was con-
sistent with reported [38]. 1H NMR (CDCl3) δ 7.40 (d, J = 8.5 Hz, 2H),
7.33 (d, J = 8.5 Hz, 2H), 6.85 (s, 2H), 3.68 (s, 2H).
Fig. 1. Concept of resonance-promoted thiosuccinimide hydrolysis.
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
3
[2-(4-Maleimidophenyl)]acetic acid was then esterified by adding
NHS (90 mg, 0.78 mmol) and DCC (160 mg, 0.78 mmol) into a stirred
solution of 4-maleimidophenylacetic acid (140 mg, 0.61 mmol) in
dimethoxyethane (5 mL) at room temperature under N2. After 1 h, the
precipitate formed was removed by filtration. The filtrate was concen-
trated under vacuum to afford crude 4-maleimidophenylacetic acid
NHS ester, which was used directly for the next step without further
purification.
(2-fluoro-5-maleimido) benzoyl-PEG7-biotin (3) as an oil (28 mg,
0.034 mmol, 20% yield) which contained approximately 10% hydrolyzed
form as shown by HPLC. An additional 90 mg mixture of the desired
product and the hydrolyzed form at a ratio of 3:2 was also obtained.
1H NMR (CDCl3) δ 8.05 (dd, J = 6.7, 2.9 Hz, 1H), 7.47 (dd, J = 4.5,
2.7 Hz, 1H), 7.23 (m, 1H) [8.20 (m, 1H), 7.46 (m, 1H), 7.19 (m, 1H),
minor component], 6.88 (s, 2H) [6.45, 6.24 (AB Type, JAB = 12.9 Hz,
2H), minor component], 6.63 (br s, 1H), 6.05 (s, 1H), 4.49 (m, 1H),
4.31 (m, 1H), 3.66–3.54 (m, 32H), 3.43 (m, 2H), 3.14 (m, 1H), 2.90
(d-AB Type, JAB = 12.7, J = 5.0 Hz, 1H), 2.73 (AB Type, JAB = 12.7 Hz,
1H), 2.22 (m, 2H), 1.73 (m, 4H), 1.44 (m, 2H). MS (ESI+) m/z 835
(M + Na), 829 (M + NH4), 813 (M + 1, base peak). A reaction scheme
is provided in the Supplementary information section (Scheme S3).
The biotinylated product was obtained by nucleophilic displacement
of NHS in [2-(4-maleimidophenyl)] acetic acid NHS ester with the
amine group of biotin-PEG7 amine. A mixture of biotin-PEG7-NH2
(0.25 g, 0.42 mmol) and freshly prepared 4-maleimidophenylacetic
NHS ester in acetonitrile (10 mL) was stirred overnight at ambient tem-
perature under N2. After the reaction, solvent was removed under
vacuum. The residue obtained was purified by flash column chromatog-
raphy on silica gel, eluting with step gradients of dichloromethane to
MeOH in dichloromethane at a ratio of v/v 1:30, 1:20, 1:10:1 and 1:5,
to afford the target product [2-(4-maleimidophenyl)]acetyl-PEG7-
biotin (2) as an oil (70 mg, 0.087 mmol, 21% yield). Additional fractions
obtained were 190 mg and 15 mg mixtures, respectively, of the desired
product and the hydrolyzed form at a ratio of 3:1 and 1:2. 1H NMR
(CDCl3) δ 7.41, 7.30 (AB Type, JAB = 8.3 Hz, 4H) [7.64, 7.24 (AB Type,
JAB = 9.2 Hz, 4H) minor component], 6.85 (s, 2H) [6.23 (d, J = 7.5 Hz,
1H), 6.47 (d, J = 7.5 Hz, 1H), minor component], 6.63 (br s, 1H), 6.55
(br s, 1H), 5.83 (s, 1H), 5.18 (s, 1H), 4.49 (m, 1H), 4.31 (m, 1H), 3.63–
3.40 (m, 34H), 3.15 (m, 1H), 2.90 (d-AB Type, JAB = 12.9, J = 4.7 Hz,
1H), 2.72 (AB Type, JAB = 12.9 Hz, 1H), 2.22 (t, J = 7.1 Hz, 2H), 1.68 (m,
4H), 1.46 (m, 2H). MS (ESI+) m/z 830 (M + Na), 825 (M + NH4, base
peak), 808 (M + 1). A reaction scheme is provided in the Supplementary
information section (Scheme S2).
2.5. Synthesis of maleimidocaproyl Val-Cit-PAB MMAE (4)
Maleimidocaproyl-valine-citrulline-p-aminobenzyloxycarbonyl-
monomethyl-Auristatin-E (mc-Val-Cit-PAB-MMAE) was prepared
according to the literature method with modifications [28,29].
A mixture of freshly prepared mc-Val-Cit-p-aminobenzyl alcohol
p-nitrophenylcarbonate (265 mg, 0.36 mmol, 1.5 eq.), MMAE
(169 mg, 0.24 mmol, 1 eq.) and N-hydroxybenzotriazole (HOBt)
(6.48 mg, 0.048 mmol, 0.2 eq.) in DMF (5 mL) was stirred at r.t. for
2 min., followed by addition of pyridine (38 mg, 0.48 mmol, 2 eq.).
After stirring for 24 h, volatile organics were removed under vacuum.
The residue obtained was successively triturated with ethyl acetate
and methanol (1 L) to afford the target compound mc-Val-Cit-PAB-
MMAE (4) (220 mg, 0.17 mmol, 71% yield) as white powder that
was N95% pure by RP-HPLC analysis. MS (ESI+) m/z 1339 (M + Na,
base peak), 1317 (M + 1). A reaction scheme is provided in the Supple-
mentary information section (Scheme S4).
2.4. Synthesis of (2-fluoro-5-maleimido)benzoyl-PEG7-biotin (3)
2.6. Synthesis of [2-(4-maleimidophenyl)] acetyl-Val-Cit-PAB-MMAE (5)
(2-Fluoro-5-maleimido) benzoic acid was synthesized similar to
described literature methods with modification [27]. A suspension of
maleic anhydride (0.3 g, 3.06 mmol) and 5-amino-2-fluorobenzoic
acid (0.5 g, 3.22 mmol) in glacial AcOH (20 mL) was heated to reflux
(bath temperature: 170–180 °C) for 90 min. The solution was cooled
to room temperature and the solvent was evaporated in vacuo. Residual
AcOH was removed by coevaporation with toluene under vacuum
(2 × 25 mL). The residue was treated with water (50 mL) and EtOAc
(100 mL). The organic layer was separated and the aqueous layer was
extracted with EtOAc (2 × 100 mL). The combined organic extracts
were dried (Na2SO4) and evaporated in vacuo. The crude was purified
by silica gel column chromatography (eluting with 50% ethyl acetate
in hexanes to afford 2-fluoro-5-maleimido benzoic acid (0.21 g,
1.83 mmol, 60% yield). 1H NMR (DMSO-d6) δ 13.43 (br s, 1H), 7.85
(dd, J = 2.7, 6.6 Hz, 1H), 7.62–7.57 (m, 1H), 7.43 (dd, J = 9.0, 10.5 Hz,
1H), 7.19, 7.18 (AB Type, JAB = 3.6 Hz, 2H).
(2-Fluoro-5-maleimido) benzoic acid was then esterified by adding
NHS (32 mg, 0.28 mmol) and DCC (62 mg, 0.30 mmol) into a stirred
solution of 2-fluoro-5-maleimide-benzoic acid (59 mg, 0.25 mmol) in
dimethoxyethane (5 mL) at room temperature under N2. After 1 h, the
precipitate formed was removed by filtration. The filtrate was concen-
trated under vacuum to afford crude (2-fluoro-5-maleimido) benzoic
acid NHS ester, which was used directly for the next step without
further purification.
Val-Cit-PAB-MMAE (100 mg, 0.089 mmol) and [2-(4-
maleimidophenyl)] acetic acid NHS ester (43.8 mg, 0.133 mmol) were
combined in DMF (1 mL) at ambient temperature under N2 followed
by addition of diisopropylethylamine (57.5 mg, 0.445 mmol). After
stirring overnight, MS analysis indicated formation of the desired
product with a small amount of starting material present. After removal
of volatile organics under vacuum, the residue was co-evaporated
several times with dichloromethane, followed by tritulation with
dichloromethane/diethyl ether (v/v 1:1) at room temperature over-
night. The solid product was collected via filtration, washed with
dichloromethane and dried under vacuum to afford the target compound
[2-(4-maleimidophenyl)] acetyl-Val-Cit-PAB-MMAE (5) (25 mg,
0.0187 mmol, 21% yield). RP-HPLC: 93% purity; 1H NMR (DMSO-d6) δ
10.0 (br s, 1H), 8.27 (br s, 1H), 8.16 (dd, J = 7.8, 15.7 Hz, 2H), 8.03 (br s,
1H), 7.87 (d, J = 8.0 Hz, 1H), 7.60 (d, J = 8.0 Hz, 1H), 7.57 (br m, 2H),
7.36, 7.23 (AB type, JAB = 8.5 Hz, 4H), 7.30, 7.27 (AB type, JAB = 7.3 Hz,
4H), 7.17 (m, 1H), 7.16 (s, 2H), 5.95 (m, 1H), 5.39 (m, 2H), 5.12–4.98
(m, 2H), 4.49 (m, 1H), 4.43–4.38 (m, 2H), 4.29–4.22 (m, 2H), 4.01–3.82
(m, 2H), 3.62, 3.53 (AB type, JAB = 14.0 Hz, 2H), 3.38 (q, J = 7.0 Hz,
1H), 3.24 (s, 3H), 3.23 (s, 3H), 3.20 (s, 3H), 3.18 (s, 3H), 3.11 (s, 2H),
3.02–2.90 (m, 3H), 2.88–2.85 (m, 2H), 2.41 (q, J = 15.8 Hz, 1H), 2.29–
2.24 (m, 1H), 2.15–2.05 (m, 2H), 2.05–1.90 (m, 2H), 1.83–1.65 (m, 3H),
1.60–1.20 (m, 6H), 1.09 (t, J = 7.0 Hz, 2H), 1.01 (dt, J = 6.5, 16.3 Hz,
6H), 0.90–0.70 (m, 24H); MS (ESI+) m/z 1338 (M + 1, base peak),
1360 (M + Na). A reaction scheme is provided in the Supplementary
information section (Scheme S5).
The biotinylated product was obtained by nucleophilic displacement
of the NHS ester in (2-fluoro-5-maleimido) benzoic acid NHS ester with
the amine group of biotin-PEG7 amine. Thus, biotin-PEG7-NH2 (0.1 g,
0.168 mmol) was added to freshly prepared 2-fluoro-5-maleimido
benzoic acid NHS ester in dichloromethane (1 mL) and the mixture
was stirred overnight at ambient temperature under N2. After removal
of solvent under vacuum, the residue obtained was purified by size
exclusion chromatography (SEC) on an LH20 column, eluting with
MeOH/dichloromethane (v/v 1:1), to afford the target product
2.7. Monoclonal antibodies
T289C mAb was purified at MedImmune as described by Dimasi
et al. [30]. Standard molecular biology methods were used to generate
the site-specific T289C cysteine antibody mutant.
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

4
R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
2.8. Antibody conjugations
Mass spectrometry analysis was performed using an Agilent
6520B Q-TOF mass spectrometer equipped with a RP-HPLC column
(Agilent Poroshell 300SB-C3; 5um, 2.1 mm × 75 mm). High-performance
liquid chromatography (HPLC) parameters were as follows: flow rate,
0.4 mL/min; mobile phase A was 0.1% (v/v) formic acid in HPLC-grade
H2O, and mobile phase B was 0.1% (v/v) formic acid in acetonitrile.
The column was equilibrated in 90% A/10% B, which was also used to
desalt the ADC samples, followed by elution in 40% A/60% B. Mass
spec data were collected for 100–3000 m/z, positive polarity, a gas
temperature of 350 °C, a nebulizer pressure of 48 lb./in2, and a capillary
voltage of 5000 V. Data were analyzed using vendor-supplied (Agilent
v.B.04.00) MassHunter Qualitative Analysis software and peak intensities
from deconvoluted spectra were used to derive the relative proportion of
species in each sample as previously described [8,31]. Representative
spectra for antibody–biotin conjugates are shown in Fig. S1 and repre-
sentative antibody–MMAE conjugates are shown in Fig. S2, Supplemen-
tary information.
Maleimide–PEG–biotins (compounds 1–3) were conjugated to
T289C mAb in several steps. First, antibodies were mildly reduced to
generate free sulfhydryls by combining 5 mL of 1.6 mg/mL antibody
solution in 10 mM PBS, pH 7.4, 1 mM EDTA (8 mg antibody, 53.3 nM,
1 eq) with 43 μL of 50 mM TCEP solution in water (2.15 μmol, 40 eq.
relative to antibody) followed by gentle mixing at 37 °C for 1 h. Reduced
antibody was transferred to a slide-a-lyzer dialysis cassette (10 K
MWCO) and dialyzed against PBS, 1 mM EDTA, pH 7.4, 4 °C for 24 h
with several buffer changes. Reduced antibody was oxidized to reform
internal disulfides by addition of dehydroascorbic acid (21 μL of
50 mM stock in DMSO, 1.1 μmol, 20 eq) followed by gentle mixing for
4 h at room temperature. Oxidized antibody solution was adjusted to
the desired pH by addition of 1 M sodium phosphate (monobasic for
pH 5.5, dibasic for pH 8.6, or a 1 M stock adjusted to pH 7.4) to a final
phosphate concentration of 100 mM and antibody concentration of
1.3 mg/mL. Next, 1.15 mL of antibody solution (1.5 mg antibody,
20 nmol reactive cysteine thiol, 1 eq) was aliquoted into a vial followed
by addition of maleimide-PEG-biotin stock solution (2 μL of a 10 mM
stock solution in DMAc, 20 nmol, 1 eq. relative to thiol). The reaction
mixture was briefly vortexed and further incubated for the desired
amount of time followed by addition of N-acetyl cysteine (10 μL of a
100 mM solution in water, 1 μmol, 50 eq) and further incubation for
15 min to quench unreacted maleimide. All conjugation reactions
were performed at room temperature (22 °C) under ambient atmo-
sphere. This general procedure was modified as needed to achieve de-
sired reaction stoichiometry (i.e. different maleimide:antibody feed).
For samples subjected to mild thiosuccinimide hydrolysis after conjuga-
tion, 10% v/v sodium phosphate solution (1 M, dibasic, pH 8.6) was
added and the solution was incubated at 37 °C for 1 h.
Antibody conjugation efficiency was calculated from the peak inten-
sities of deconvoluted mass spectra using the Eq. (1):
a þ b
Conjugationefficiency ¼
ꢀ 100:
ð1Þ
a þ b þ c þ d
a = conjugated heavy chain G0 glycoform peak intensity.
b = conjugated heavy chain G0 glycoform + [Na+ and H2O] peak
intensity.
c = unconjugated heavy chain G0 glycoform peak intensity.
d = unconjugated heavy chain G0 glycoform + [Na+ and H2O] peak
intensity.
Note: Na and H2O peaks were not resolved and appeared as a single
peak between +18 and +22 amu relative to the observed antibody
peak.
Maleimide-Val-Cit-PAB-MMAEs (compounds 4 and 5) were conju-
gated to T289C antibody in a similar fashion as described above for
maleimide-PEG-biotins. First, antibodies were mildly reduced to gener-
ate free sulfhydryls by combining 5 mL of 1.6 mg/mL antibody solution
in 10 mM PBS pH 7.4 containing 1 mM EDTA (8 mg antibody, 107 nM
reactive cysteine thiol, 1 eq) with 43 μL of 50 mM TCEP solution in
water (2.15 μmol, 20 eq. relative to engineered cysteine thiols) followed
by gentle mixing at 37 °C for 1 h. Reduced antibody was transferred to a
slide-a-lyzer dialysis cassette (10 K MWCO) and dialyzed against PBS,
1 mM EDTA, pH 7.4, 4 °C for 24 h with several buffer changes. Reduced
antibody was oxidized to reform internal disulfides by addition
of dehydroascorbic acid (21 μL of 50 mM stock in DMSO, 1.1 μmol,
20 eq) followed by gentle mixing for 4 h at room temperature. Oxidized
antibody solution (2.5 mL, 54 nmol engineered cysteine thiols, 1 eq)
was combined with 10% v/v DMSO followed by addition of maleimide-
Val-Cit-PAB-MMAE (27 μL of a 10 mM stock in DMSO, 270 nmol,
5 eq). The reaction proceeded at room temperature with mixing
for 1 h and then N-acetyl cysteine (21 μL of 100 mM stock in water,
2.2 μmol, 40 eq) was added to quench unreacted maleimide. The reac-
tion mixture was then diluted 3-fold with distilled water and subjected
to CHT chromatography to remove free unconjugated drug (Bio-Scale
Mini Cartridge CHT Type II 40 μm media column). ADC was eluted
with a gradient from buffer A (10 mM phosphate, pH 7.0) to buffer B
(10 mM phosphate pH 7.0 containing 2 M NaCl) over 25 min. After
CHT chromatography the sample was buffer exchanged to PBS supple-
mented with 1 mM EDTA, pH 7.4 by dialysis in a slide-a-lyzer cassette
at 4 °C.
Thiosuccinimide hydrolysis of antibody conjugates over time was
calculated from the peak intensities of deconvoluted mass spectra
using Eq. (2):
b þ c−½a ꢁ ðl þ mÞꢂ
%hydrolysis ¼
ꢀ 100:
ð2Þ
a þ ½b−ða ꢁ lÞꢂ þ ½c−ða ꢁ mÞꢂ
a = conjugated heavy chain G0 glycoform peak intensity.
b = conjugated heavy chain G0 glycoform + [H2O and Na] peak
intensity.
c = conjugated heavy chain G0 glycoform + [H2O + 2Na] peak
intensity (i.e. double sodium ion species).
l = b / a at T = 0.
m = c / a at T = 0.
Note: Na and H2O peaks were not resolved and appeared as a single
peak between +18 and +22 amu relative to the observed antibody
peak.
The quantitative nature of the mass spectrometry methods used was
confirmed as described in the Supplementary information section
(Figs. S3–S6 and Table S1).
2.10. Antibody conjugation kinetics
Antibody conjugates were prepared with maleimide-biotin com-
pounds 1, 2 and 3 as described above with minor modification. For all
reactions: the antibody concentration was 1.3 mg/mL, the thiol concen-
tration was 17.3 × 10−6 M, and the reaction stoichiometry was 1 mol
maleimide:1 mol cysteine thiol. Reactions were performed at pH 5.5,
pH 7.4, and pH 8.6, respectively, in 100 mM phosphate buffer at 22 °C.
The pH of each reaction solution was measured before addition of
maleimide. After quenching with N-acetyl cysteine, reactions were
analyzed by reduced glycosylated mass spectrometry (Fig. S7). Percent
conjugation was calculated using peak intensities from deconvoluted
mass spectra and Eq. (1). Conjugation data were further analyzed in
2.9. Mass spectrometry
Antibody conjugates were prepared for HPLC/MS analysis by first
diluting samples to 0.2 mg/mL with PBS pH 7.4 and then combining
50 μL of antibody solution with 5 μL of TCEP (0.5 M in water). This mix-
ture was incubated for 5 min at room temperature to reduce disulfide
bonds prior to injection (15 μL) into the HPLC/MS instrument.
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
5
units of molar concentration to determine kinetic constants. Second
order rate constants were determined from the slopes of curves gene-
rated from plotting 1/[SH] versus time and linear regression analysis
(Fig. S8). Reaction half-lives were calculated from second-order reaction
rate constants using Eq. (3):
aliquots removed at desired times and immediately diluted with
pH 5.5 phosphate buffer as described above and subjected to HPLC/MS
analysis. For each HPLC/MS analysis, samples were sterile filtered,
reduced with TCEP, and analyzed by HPLC/MS. Quantitative analysis of
hydrolysis was performed using peak intensities in deconvoluted mass
spectra and Eq. (2), which includes background subtraction of sodium
ion interference using the T = 0 measurement (Fig. S15). Additional
experiments were performed to determine thiosuccinimide hydrolysis
at 1 h, n = 3 (Fig. S16). Thiosuccinimide hydrolysis kinetic data were
further analyzed in units of molar concentration to determine kinetic
constants. Pseudo first-order rate constants were determined from the
slopes of curves generated from plotting ln·[Thiosuccinimide] versus
time and linear regression analysis (Fig. S17). Hydrolysis reaction half-
lives were calculated from the pseudo first-order rate constants using
Eq. (4).
1
T1=2
¼
:
ð3Þ
k2½SHꢂ0
k2 = second order rate constant.
[SH]0 = thiol concentration at time = 0.
2.11. Maleimide hydrolysis
Hydrolysis kinetics of maleimide compounds 1, 2 and 3 was
performed in phosphate buffer at pH 5.5, 7.4 and 8.6, respectively,
with an initial concentration of approximately 1 nM. Samples were
kept at 22 °C and analyzed by HPLC at each time point defined. Due to
rapid hydrolysis of compound 2 and 3 at pH 8.6, analytical samples at
each time point were quenched with 3 volumes of pH 5.5 and stored
at 0 °C until analysis.
4 μL of each sample was injected via auto injector into the HPLC
(Agilent 1100 Series) with a C-18 reversed phase column (Eclipse
XDB-C18, 4.6 × 150 mm, 5 μm) eluting at 1.0 mL/min with a gradient
of 0.05% phosphoric acid in water (A) and acetonitrile containing
0.05% phosphoric acid (B): 85% A for 2 min and then 85% to 5% A over
8 min. Intact maleimide and its hydrolyzed, ring-opened congener
were separated and detected by UV absorbance at 220 nm wavelength.
The hydrolyzed products eluted faster than the respective intact
maleimides with their structures confirmed by LC/MS under negative
mode to show the M−1 ion. A representative HPLC analysis of com-
pound 3 is shown in Fig. S10 and hydrolysis plots for compounds 1, 2,
and 3 are shown in Fig. S11. Maleimide hydrolysis data were further an-
alyzed in units of molar concentration to determine kinetic constants.
Pseudo first-order rate constants were determined from the slopes of
curves generated from plotting ln·[Maleimide] versus time and linear
regression analysis (Fig. S12).
2.13. Conjugate stability in buffer
Antibody conjugates were incubated in aqueous buffer containing
β-mercaptoethanol (BME) to assess deconjugation. Maleimide-PEG-
biotin compounds 1, 2 and 3 were conjugated to antibody containing
the T289C mutation as described above with minor modification.
Conjugation reactions were performed at stoichiometric equivalents
maleimide:thiol for 15 min at 22 °C followed immediately by dilution
to 0.2 mg/mL (1.33 μM antibody) with 1× PBS pH 7.4 containing
1 mM EDTA. For BME-containing samples, BME was added to a final
concentration of 1% v/v (143 mM). Samples were further incubated at
37 °C under ambient atmosphere without stirring. Aliquots were
removed at various time points, sterile filtered, reduced with DTT and
then analyzed by LC/MS (Fig. S18). Percent conjugated antibody and
thiosuccinimide hydrolysis were determined from peak heights of
mass spectra using Eqs. (1) and (2), respectively.
Conjugate stability was also assayed following a mild hydrolysis
procedure. Antibody conjugates were prepared with compounds 1, 2
and 3 as described above followed by incubation at pH 8.6 (100 mM
sodium phosphate) at 37 °C for 1 h to hydrolyze thiosuccinimides. Sam-
ples were then passed through a Sephadex G10 column with 10 mM
sodium phosphate, 1 mM EDTA, pH 7.4 as the elution buffer. Pre-
hydrolyzed samples were then subjected to incubation with BME at
37 °C as described above and analyzed by mass spectrometry over time.
Deconjugation and thiosuccinimide hydrolysis data were further
analyzed in terms of molar concentration to determine hydrolysis and
deconjugation rate constants in the same sample. Deconjugation and
thiosuccinimide hydrolysis data were plotted as ln[concentration] vs.
time to obtain the pseudo-first order rate constants from the slope of
the best fit line (Fig. S19). Deconjugation was treated as a pseudo
first-order process.
Hydrolysis reaction half-lives were calculated from the pseudo first-
order rate constants using Eq. (4):
0:693
k1;obs
T1=2
¼
:
ð4Þ
k1,obs = pseudo first-order rate constant.
The effect of maleimide hydrolysis on conjugation efficiency follow-
ing incubation in aqueous buffer for various times is shown in Fig. S13.
2.12. Thiosuccinimide hydrolysis
ADCs prepared with T289C antibody and MMAE were subjected to
incubation in buffer containing BME in the same manner as described
above and analyzed by reduced glycosylated mass spectrometry over
time. Percent conjugated antibody was calculated using Eq. (1). ADC
prepared with compound 4 (alkyl maleimide MMAE) showed significant
deconjugation (Fig. S20) and mass spectrometry data was further ana-
lyzed as described above to determine rate constants for deconjugation
and thiosuccinimide hydrolysis processes (Fig. S21).
Thiosuccinimide hydrolysis was determined directly on antibody
conjugates using reduced glycosylated mass spectrometry. T289C anti-
body conjugates prepared with compounds 1, 2 and 3 were incubated
in buffer solutions and monitored by HPLC/MS over time to observe
addition of water (18 amu) due to thiosuccinimide hydrolysis
(Fig. S14). First, T289C antibody (0.75 mg, 5 nmol, 1 eq) was reacted
with-maleimide-PEG-biotin (50 nmol, 5 μL of a 10 mM stock in DMAC,
10 eq) in 577 mL PBS, pH 7.4, 1 mM EDTA. The conjugation reaction
was allowed to proceed for 5 min and then quenched with N-acetyl
cysteine (400 nmol, 4 μL of 100 mM stock in water, 80 eq) and further
incubated for 5 min. The reaction mixture was then combined with
10× PBS (1 M sodium phosphate, 100 mM EDTA) at the desired pH
to achieve final concentrations of 0.65 mg/mL antibody, 100 mM phos-
phate, 135 mM NaCl and 1 mM EDTA. After sample preparation, an
aliquot was removed and diluted 1:3 with 75 mM phosphate buffer
pH 5.5 to stop hydrolysis and obtain an initial time point sample.
Samples were then further incubated at the desired conditions and
2.14. Conjugate stability in serum
ADCs prepared with MMAE compounds 4 and 5 were incubated in
mouse serum to challenge the stability of the thiol linkage. MMAEs
were conjugated to T289C antibody and purified by CHT chromatogra-
phy as described above. ADCs were prepared with compounds 4 and 5
using identical conditions; no additional steps were included to facili-
tate thiosuccinimide hydrolysis. Samples were then subjected to brief
dialysis (slide-a-lyzer cassette, 10 kDa MWCO, 4 °C, 2 h) to exchange
the buffer to PBS, pH 7.4 containing 1 mM EDTA. After dialysis, antibody
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

6
R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
concentrations were determined by A280 measurement (NanoDrop)
and then added to normal mouse serum (Jackson Immunoresearch) to
achieve a final concentration of 0.2 mg/mL (1.33 μM antibody). The
total volume of ADC solution added to serum was less than 10%. The
ADC-serum mixture was sterile filtered and incubated at 37 °C in a
sealed container without stirring. Aliquots were removed at various
time points and frozen. Conjugated and unconjugated human antibody
was recovered from mouse serum by immunoprecipitation using
Fc-specific anti-human IgG-agarose resin (Sigma-Aldrich). Resin was
rinsed twice with PBS, once with IgG elution buffer, and then twice
more with PBS. ADC-mouse serum samples were then combined with
anti-human IgG resin (100 μL of ADC-serum mixture, 50 μL resin slurry)
and gently mixed for 15 min at room temperature. Resin was recovered
by centrifugation and then washed twice with PBS. The resin pellet was
resuspended in 100 μL IgG elution buffer (Thermo Scientific) and
further incubated for 5 min at room temperature. Resin was removed
by centrifugation and then 20 μL of 1 M Tris, pH 8.0 was added to the
supernatant. Recovered human antibody solution was sterile filtered,
reduced with DTT and analyzed by LC/MS. Percent conjugated antibody
and thiosuccinimide hydrolysis were determined from peak heights
of mass spectra using Eqs. (1) and (2), respectively. ADC prepared
with compound 4 (alkyl maleimide MMAE) showed significant
deconjugation as evidenced by appearance of free antibody heavy
chain over time (Fig. S22) and data was further analyzed to determine
rate constants for deconjugation. Deconjugated antibody data was
plotted as ln[concentration] vs. time (s) to obtain the pseudo-first
order rate constants from the slope of the best fit line (Fig. S23).
(Val-Cit-PAB-MMAE) and compared to traditional N-alkyl maleimide
Val-Cit-PAB-MMAE in terms of ADC stability and biological activity.
Synthesis of maleimide compounds was straightforward based on
previous literature reports. Overall product yield generally decreased
based on the type of maleimide substitution (alkyl N phenyl N F-phenyl),
with the major byproduct observed being hydrolyzed maleimides
(maleamic acid derivatives).
3.2. Antibody conjugation
Reaction of maleimide compounds with antibody thiols was evaluat-
ed using a cysteine-engineered antibody. The antibody used (T289C
mAb) comprised a solvent-exposed engineered cysteine at threonine
position 289 on the heavy chain CH2 domain (Fig. 2B). Antibody conju-
gation reactions were monitored by mass spectrometry and the quanti-
tative nature of this method was confirmed for the specific molecules
used for antibody conjugation in this study (Figs. S1–S8).
Compounds 1, 2, and 3 all conjugated very efficiently to T289C mAb
at pH 7.4 and pH 8.6, with complete conjugation occurring in minutes at
stoichiometric equivalents of thiol:maleimide (Table 1). Differences in
maleimide reactivity became apparent at pH 5.5, where conjugation
efficiency was enhanced by N-phenyl and N-fluorophenyl functional
groups (Fig. 3A). Thiol-specificity of all maleimides was confirmed at
pH 7.4 up to 5 equivalents of maleimide, with no other conjugation
products observed on the antibody heavy or light chains (Fig. S1).
Second-order rate constants for antibody–maleimide conjugation
determined by mass spectrometry were similar to values reported
for maleimide-thiol reaction rates with small-molecules. For example,
rate constants for reaction of N-ethyl maleimide with β-mercaptoethanol
and cysteine at pH 7.0, 22 °C were reported to be 0.7 × 103 M−1 s−1
and 1.6×103 M−1 s−1, respectively [32,33]. Reaction of T289C mAb with
N-alkyl maleimide-PEG-biotin (compound 1) at pH 7.4, 22 °C was found
to be 0.5 × 103 M−1 s−1 in this study.
2.15. In vitro cytotoxicity
ADCs were evaluated for in vitro potency following incubation in
mouse serum. Samples were recovered from the serum stability assay
described above and subjected to in vitro toxicity assays in MDA-MB-
361 breast cancer cells. This cell line was chosen because the T289C
antibody used in this study binds an oncofetal protein that is expressed
by MDA-MB-361 cells. Cells were plated in 80 μL of Leibovitz's L-15 with
20% FBS into 96-well flat-bottomed plates at 5000 MDA-MB-361 cells/
well. Cells were allowed to adhere overnight. A 5× concentration of
each ADC was prepared by diluting the test articles in culture medium.
Twenty microliters of each test article was added to cells in duplicate
such that the final dose curve range of 4000 ng/mL down to
0.06 ng/mL in a stepwise 1:4 serial dilution series. The treated cells
were cultured at 37 °C/0% CO2 for 6 days and cell viability was assessed
with the CellTiter-Glo (CTG) Luminescent Viability Assay from Promega.
100 μL of reconstituted CTG reagent was added to each well and the
plate was mildly shaken for 10 min at room temperature. The lumines-
cence of each sample at 560 nm was read using a Perkin Elmer EnVision
luminometer. Percent cell viability was calculated by the following
formula: (average luminescence of treated samples / average lumines-
cence of untreated control samples) × 100. IC50 values were determined
using logistic non-linear regression analysis with GraphPad Prism
software.
3.3. Maleimide hydrolysis
Conjugation efficiency of N-aryl maleimides remained high even at
pH 8.6. At high pH, maleimide hydrolysis before thiol conjugation is ex-
pected, which would decrease conjugation efficiency. Thus, maleimide
hydrolysis rates of compounds 1–3 were measured to understand
why such high conjugation efficiencies were observed (Table 1,
Figs. S9–S12). N-aryl functionality increased maleimide hydrolysis as
expected (Fig. 3B), with the hydrolysis half-life of compound 2
(21 min, pH 8.6, 22 °C) closely matching the hydrolysis half-life report-
ed for N-phenyl maleimide (20 min, pH 8.0, 30 °C) [22]. N-phenyl
maleimides hydrolyzed approximately 4-fold faster than N-alkyl
maleimides and N-fluorophenyl maleimides hydrolyzed approximately
7 fold faster than N-alkyl maleimides at room temperature, pH 8.6.
However, maleimide hydrolysis rate did not increase enough to nega-
tively impact conjugation efficiency. Comparison of thiol conjugation
and maleimide hydrolysis kinetics for N-aryl maleimides showed that
conjugation still occurred much faster than hydrolysis. Specifically, N-
aryl maleimide conjugation half-lives were approximately 70–130
times shorter than maleimide hydrolysis half-lives at pH 7.4. Note that
accurate conjugation rate constants could not be obtained at pH 8.6
due to fast reaction kinetics, with all reactions completed within 30 s
after combining reactants (a direct comparison of antibody conjugation
with maleimide hydrolysis for compound 3 at pH 8.6 is shown in
Fig. S9). Assuming a conjugation reaction half-life of 30 s for N-aryl
maleimides at pH 8.6, conjugation is at least 18–42 times faster than
maleimide hydrolysis. Thus, efficient conjugation was achieved even
for hydrolytically labile maleimides because increased thiol reaction
rate compensated for maleimide instability. A 30-50% decrease in conju-
gation efficiency was observed for N-aryl maleimides that were exposed
to buffer (pH 7.4, 22 °C, 1 h) prior to antibody conjugation, confirming
3. Results and discussion
3.1. Synthesis of N-aryl maleimide compounds
We designed and synthesized several maleimide compounds to
test the hypothesis of resonance-promoted stabilization of antibody
conjugates (Fig. 2). One group of molecules consisted of small hydro-
philic drug surrogates (PEG-biotins) designed for ease of synthesis and
expected high solubility in water. The maleimide functionality on these
molecules included N-alkyl (compound 1), N-phenyl (compound 2),
and N-(4-fluorophenyl) (compound 3) groups. N-phenyl maleimide
functionality was also incorporated into the cytotoxic agent valine-
citrulline-p-aminobenzyloxycarbonyl-monomethyl-auristatin-E
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
7
Fig. 2. A) Structures of maleimide compounds. B) Human IgG1 Fc structure showing the T289 site chosen for cysteine mutagenesis (red spheres) in the T289C mAb. N-linked oligosaccharides
are shown by stick representation in green and the CH2 and CH3 domains of the Fc are schematically labeled. The interchain disulfide bridges in the CH2 and CH3 domains are shown in
yellow.
that maleimide hydrolysis can impact conjugation efficiency under cer-
tain conditions (Fig. S13).
conjugation and subsequent thiosuccinimide hydrolysis can be
achieved in a one-step procedure by simply increasing the reaction
time. Addition of a fluorine to the N-phenyl ring further increased
thiosuccinimide hydrolysis rate, suggesting that pulling ring-head
nitrogen lone pair electrons into ring resonance structures by induction
promotes thiosuccinimide hydrolysis. However, the difference between
N-phenyl and N-fluorophenyl groups was modest (~1.4–2.1 fold
increase for N-fluorophenyl) compared to the large difference between
N-aryl and N-alkyl groups in general. Thiosuccinimide hydrolysis rates
determined for N-aryl maleimides in this study are comparable to
other maleimide modifications intended to increase thiosuccinimide
conjugate hydrolysis. For example, addition of a primary amine in prox-
imity to the maleimide with different carbon spacer lengths results in
thiosuccinimide hydrolysis half-lives of 0.4–13.5 h at pH 7.4, 22 °C [9]
while addition of tertiary amines, amides, ethers and thioethers in
3.4. Thiosuccinimide hydrolysis
Thiosuccinimide hydrolysis was monitored directly on antibody
conjugates by mass spectrometry, which allowed observation of the
addition of water (18 amu) and calculation of pseudo-first-order hydro-
lysis rate constants (Figs. S14–S17). N-aryl functionality increased
thiosuccinimide hydrolysis rates ~15–20 fold for conjugates prepared
with N-aryl maleimide compounds 2 and 3 compared to the conjugate
formed with N-alkyl maleimide compound 1 (Fig. 3C, Table 2). N-aryl
thiosuccinimide hydrolysis occurred spontaneously under very mild
conditions (pH 7.4, 22 °C) in the timeframe of hours. These conditions
are attractive for practical applications because thiol–maleimide
Table 1
Antibody conjugation and maleimide hydrolysis rate constants for maleimide-PEG-biotins.
pH 5.5
pH 7.4
pH 8.6
Compound
1
2
3
1
2
3
1
2
3
Thiol conjugation
13
24
55
500
1300
4500
ND
ND
ND
a,b,d
(k2, M−1 s−1
)
Thiol conjugation
T1/2 (min)a,b,d
Conjugation after 1 h
(%)a,b,d
74
40
17
1.7
0.72
0.21
b0.5
b0.5
b0.5
47
57
78
N90
N90
N90
N90
N90
N90
Maleimide hydrolysis
4.2 × 10−7
2.8 × 104
8.2 × 10−6
1400
1.5 × 10−5
770
3.8 × 10−5
304
2.1 × 10−4
55
4.1 × 10−4
28
1.4 × 10−4
83
5.4 × 10−4
21
9.3 × 10−4
12
a,c
(k1,obs, s−1
)
Maleimide hydrolysis
T
1/2 (min)a,c
ND, not determined.
a
22 °C.
b
Determined by mass spectrometry.
Determined by HPLC.
Reaction performed at 1:1 thiol:maleimide molar reaction stoichiometry, 17.3 μM for each reactant.
c
d
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

8
R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
4 comprised a C6-amide attached to the ring-head nitrogen, thus, linker
structure may have also influenced the thiosuccinimide hydrolysis rate
of N-alkyl maleimides described in this work. Altogether, each ADC conju-
gate will likely exhibit a unique thiosuccinimide hydrolysis rate depend-
ing on its specific chemical structure. It should be noted that differences
in hydrolysis was most noticeable for the slow hydrolyzing N-alkyl
maleimides. Payload hydrophobicity did not have an obvious impact for
the conjugates of N-phenyl maleimide species (compounds 2 and 5),
which were completely hydrolyzed in ~30 h in buffer and serum at 37
°C (data not shown). Complete hydrolysis in this timeframe prevented
significant deconjugation by the retro-Michael reaction as evidenced in
conjugate stability studies.
3.5. Conjugate stability
It is generally accepted that thiosuccinimide deconjugation
occurs via a retro-Michael reaction. Thus, antibody conjugate stabil-
ity can be monitored in the presence of excess free thiols that quench
liberated maleimides formed by the retro-reaction [5,15,19,34].
Here, conjugate stability was assayed in aqueous buffer solution con-
taining β-mercaptoethanol (BME) over time by mass spectrometry
(Figs. S18–S21). Antibody conjugates prepared with N-aryl maleimides
(compounds 2, 3, and 5) were more stable upon incubation at 37 °C
in PBS containing BME compared to N-alkyl maleimide conjugates pre-
pared with compounds 1 and 4. A sequential decrease in maximum
deconjugation was observed as maleimide functionality progressed
from N-alkyl to N-fluorophenyl groups (compounds 1–3) (Fig. 4A).
The maximum deconjugation observed for N-aryl thiosuccinimides
was less than 20% when samples were subjected to incubation in
BME-containing media immediately after antibody conjugation, where-
as the maximum conjugation for N-alkyl thiosuccinimides was ~35%.
The small amount of deconjugation observed for antibody conjugates
prepared with compounds 2 and 3 was completely inhibited by pre-
hydrolyzing thiosuccinimides by brief incubation at pH 8.6 and 37 °C,
but not for compound 1 (Fig. 4B). The conditions used to promote
thiosuccinimide hydrolysis in this case were milder than those
described to achieve a similar affect with N-alkyl thiosuccinimides [8],
thus, N-aryl thiosuccinimides may be better suited for applications
where rapid base-catalyzed thiosuccinimide hydrolysis is desired.
Incomplete deconjugation of compound 1 from antibodies prompted
us to further investigate the limiting factor for deconjugation. Tracking
both deconjugation and thiosuccinimide hydrolysis in the same sample
revealed that maximum hydrolysis (~90%) and maximum deconjugation
(~35%) plateaued at the same time (Fig. 4C). Analysis of reaction
rate data for the antibody conjugate prepared with N-alkyl maleimide
compound 1 showed that thiosuccinimide hydrolysis was faster than
deconjugation (Tables 2 and 3). Thus, thiosuccinimide hydrolysis limits
maximum deconjugation of thiol–maleimide linked compounds from
antibodies.
Fig. 3. Kinetics of antibody conjugation, maleimide hydrolysis and thiosuccinimide hydrolysis.
A) Conjugation of maleimide-PEG-biotins to antibodies at pH 5.5, 22 °C, 1 equivalent
thiol:maleimide, B) hydrolysis of the maleimide group in maleimide-PEG-biotins at pH 7.4,
22 °C, C) thiosuccinimide hydrolysis of antibody–PEG-biotin conjugates at pH 7.4, 22 °C,
D) N-alkyl thiosuccinimide hydrolysis of antibody–PEG-biotin and antibody–MMAE
conjugates at pH 7.4, 37 °C.
proximity to a maleimide results in thiosuccinimide conjugate hydrolysis
half-lives of ~0.4–40
h at pH 7.4, ambient temperature [17].
Thiosuccinimide hydrolysis was also affected by temperature; increasing
the temperature from 22 °C to 37 °C resulted in noticeably faster kinetics.
Hydrolysis rates increased 3.6 fold for N-alkyl thiosuccinimide, 4.3 fold
for the N-phenyl thiosuccinimide and 6.1 fold for the N-fluorophenyl
thiosuccinimide compared to hydrolysis rates at 22 °C. Thiosuccinimide
hydrolysis was slower than maleimide hydrolysis (~5–20 fold decrease)
which reflects lower reactivity of the ring after thiol addition to the
double bond.
Other factors such as payload hydrophobicity and linker structure may
also impact thiosuccinimide hydrolysis rate. The hydrophobic MMAE con-
jugate prepared with compound 4 (cLogP = 5.75, ChemBiodraw Ultra
V.13.0) showed a much slower thiosuccinimide hydrolysis rate than the
analogous conjugate prepared with hydrophilic compound 1 (cLogP =
−2.19, ChemBiodraw Ultra V.13.0) at pH 7.4 and 37 °C (Fig. 3D,
Table 2). Hydrophobicity of the local environment surrounding the conju-
gation site likely affects the ability of water to access the thiosuccinimide
group and complete the hydrolysis reaction. The exact structure of N-alkyl
functionality ahead of a thiosuccinimide should also be considered when
evaluating hydrolysis rate. For example, it was recently shown that amide
groups connected to a thiosuccinimide ring-head nitrogen with various
carbon spacer lengths increased hydrolysis rate, presumably through in-
ductive effects [17]. Compound 1 comprised a C3-amide and compound
Conjugates with slowly hydrolyzing thiosuccinimides exhibited
higher deconjugation, as demonstrated with the ADC prepared with
N-alkyl maleimide MMAE compound 4 (Table 3). Highest deconjugation
(60%) was observed for the antibody conjugate of compound 4, which
also exhibited the slowest thiosuccinimide hydrolysis rate. This data
further corroborates that thiosuccinimide hydrolysis limits
deconjugation as discussed above. The ADC prepared with N-phenyl
Table 2
Thiosuccinimide hydrolysis rate constants and half-lives in buffera,b
.
pH 7.4, 22 °C
pH 7.4, 37 °C
pH 8.6, 22 °C
Compound
1
2
3
1
2
3
4
1
2
3
k1,obs (s−1
)
2.0 × 10−6
96
3.0 × 10−5
6.4
4.6 × 10−5
4.3
1.4 × 10−6
27
1.3 × 10−4
1.5
2.8 × 10−4
0.7
1.4 × 10−6
138
1.9 × 10−5
10
2.0 × 10−4
0.8
5.0 × 10−4
0.4
T1/2 (h)
a
Determined by mass spectrometry.
100 mM sodium phosphate, 135 mM NaCl, 1 mM EDTA.
b
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
9
Fig. 4. Deconjugation of PEG-biotins from antibodies in PBS containing 143 mM β-mercaptoethanol, pH 7.4, 37 °C. A) Samples incubated immediately after conjugation, B) samples
incubated at pH 8.6, 37 °C for 1 h prior to incubation, C) relationship between thiosuccinimide hydrolysis and deconjugation for the antibody conjugate prepared with compound 1.
maleimide MMAE compound 5 showed excellent stability in PBS
containing BME, with less than 10% deconjugation observed after 76 h
and no further increase after 168 h incubation (Fig. 5A).
ADC was less effective than the stable N-phenyl maleimide MMAE
ADC (Fig. 6). Both IC50 and maximum cell killing decreased with in-
creased serum incubation for the ADC prepared with N-alkyl maleimide
MMAE compound 4 whereas the ADC prepared with N-aryl maleimide
MMAE compound 5 showed a minor loss of potency at day 7 and no
change in the maximum cell killing following prolonged serum incuba-
tion (Table 4). This result highlights the impact of conjugate stability on
in vitro activity of ADCs, which has also been shown to impact ADC
efficacy in vivo [4,7–9].
ADC stability was assayed in serum to approximate the conditions en-
countered after intravenous injection. Loss of drug from antibodies in
serum is thought to involve serum albumin, which contains at least one
free thiol (and 35 total cysteines) that can react with maleimide
liberated by the retro-Michael reaction [35]. Other small molecule free
thiols may also scavenge deconjugated maleimide-drugs; such as
reduced cysteine, glutathione and cysteine–glycine, which are present
at a combined concentration of ~15 μM in human plasma [36]. ADC
stability in mouse serum was similar to that observed in PBS containing
BME, yielding similar deconjugation rates and maximum deconjugation
values for the slowly hydrolyzing ADC prepared with N-alkyl maleimide
compound 4 (Table 3, Figs. S22 and S23). In contrast, the N-phenyl
maleimide MMAE (compound 5) ADC was essentially resistant to
deconjugation in serum, with less than 10% deconjugation observed
over a period of ~1 week at 37 °C (Fig. 5B). These results suggest that
ADCs prepared by conjugation with N-aryl maleimides could have supe-
rior in vivo stability compared to those prepared with traditional N-
alkyl maleimides.
4. Conclusions
This work demonstrates the feasibility of preparing stable thiol-
linked ADCs with N-aryl maleimides. The resonance-based strategy
3.6. In vitro potency of serum incubated ADCs
Next, the impact of drug deconjugation on ADC activity towards
cancer cells was evaluated. This experiment was aimed to model the
effect of drug loss from an ADC over time in the bloodstream. Total
antibody (conjugated and unconjugated) was recovered from mouse
serum incubated samples by immunocapture and then subjected to
cytotoxicity assays utilizing target-positive MDA-MB-361 human breast
cancer cells. ADCs prepared with N-alkyl maleimide MMAE compound
4 have reported IC50 values ranging from ~4.5–80 ng/mL and the IC50
values measured for the ADC prepared with compound 4 in this study
are similar to those literature values [37]. It should be noted that anti-
body without conjugated drug does not have cytotoxic activity towards
MDA-MB-361 cells (Fig. S24). Cytotoxicity analysis of the serum-
incubated ADCs revealed that the unstable N-alkyl maleimide MMAE
Table 3
Summary of deconjugation data for N-alkyl maleimide antibody conjugatesa,b
.
PBS + β-mercaptoethanolc
Mouse serum
Compound
1
4
4
Deconjugation
1.9 × 10−6
1.6 × 10−6
1.8 × 10−6
(k1,obs, s−1
)
Deconjugation
1/2 (h)
Maximum deconjugation (%)
101
35
120
60
107
67
T
a
Determined by mass spectrometry.
37 °C incubation.
pH 7.4, 143 mM β-mercaptoethanol.
Fig. 5. Stability of ADCs in thiol-containing buffer and mouse serum. A) Deconjugation of
MMAE from antibodies in PBS containing 143 mM β-mercaptoethanol, pH 7.4, 37 °C.
B) Deconjugation of MMAE from antibodies in mouse serum, 37 °C.
b
c
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

10
R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
ADCs for clinical applications. A.J., Q.W. and Z.X. are employees of
SynChem, Inc., which provides chemical services for ADC development.
Contributions
R.J.C., C.G., and N.D. conceived and designed the experiments. R.J.C.,
R.F., B.B., S.M., R.W., A.J., and Q.W. performed the experiments. R.J.C.
and Z.X. guided compound synthesis. R.J.C., R.F., S.M., R.W. J.H., A.J., Q.W.,
and Z.X. analyzed data. N.D., C.G. and H.W. provided scientific guidance.
R.J.C., Z.X., and N.D. wrote the manuscript.
Acknowledgments
This work was financially supported by MedImmune, a member of
the AstraZeneca group.
Appendix A. Supplementary data
Chemical synthesis schemes, antibody mass spectrometry data,
details of kinetic analysis for antibody thiol–maleimide conjugation,
maleimide hydrolysis, thiosuccinimide hydrolysis and conjugate stability
data are shown. Experiments comparing chromatography methods with
mass-spectrometry methods for quantification of antibody conjugation
are described and associated data is provided. Antibody conjugation
efficiency data following addition of maleimide-PEG-biotins to buffer
for various times is provided. Supplementary data for this article can be
found online at: http://dx.doi.org/10.1016/j.jconrel.2015.09.032.
Fig. 6. In vitro activity of ADCs towards receptor positive MDA-MB-361 breast cancer cells
following incubation in mouse serum at 37 °C for 0, 1 and 7 days. A) ADC prepared with
N-alkyl maleimide MMAE compound 4. B) ADC prepared with N-phenyl maleimide
MMAE compound 5. Data is plotted as the average of two experiments the standard error.
References
[1] L. Ducry, B. Stump, Antibody–drug conjugates: linking cytotoxic payloads to monoclonal
antibodies, Bioconjugate Chem. 21 (2010) 5–13.
[2] H.P. Gerber, F.E. Koehn, R.T. Abraham, The antibody–drug conjugate: an enabling
modality for natural product-based cancer therapeutics, Nat. Prod. Rep. 30 (2013)
625–639.
increased thiosuccinimide hydrolysis and also increased thiol–
maleimide conjugation rates, which facilitated efficient conjugation
over the pH range of 5.5–8.6. Increased maleimide reactivity due to
resonance could be advantageous compared to strategies that only
increase maleimide/thiosuccinimide hydrolysis rate but do not increase
the thiol reaction rate. In this case, maleimide hydrolysis prior to thiol
conjugation could have a negative impact on conjugation efficiency,
likely requiring excess maleimide to achieve complete conjugation.
Furthermore, N-aryl maleimides are a flexible approach to achieve spe-
cific properties of ADC payloads, as phenyl ring substitution allows for
fine-tuning of maleimide reactivity and potentially other molecule
properties such as charge and hydrophilicity. Altogether, the thiol
conjugation chemistries described in this work represent a simple and
practical method to produce bioconjugates with predictable stability
and improved performance.
[3] P.A. Trail, Antibody drug conjugates as cancer therapeutics, Antibodies 2 (2013)
113–129.
[4] S.C. Alley, D.R. Benjamin, S.C. Jeffrey, N.M. Okeley, D.L. Meyer, R.J. Sanderson, P.D.
Senter, Contribution of linker stability to the activities of anticancer immuno-
conjugates, Bioconjugate Chem. 19 (2008) 759–765.
[5] A.D. Baldwin, K.L. Kiick, Tunable degradation of maleimide–thiol adducts in reducing
environments, Bioconjugate Chem. 22 (2011) 1946–1953.
[6] P.M. Cal, G.J. Bernardes, P.M. Gois, Cysteine-selective reactions for antibody conjugation,
Angew. Chem. Int. Ed. 53 (2014) 10585–10587.
[7] B.Q. Shen, K. Xu, L. Liu, H. Raab, S. Bhakta, M. Kenrick, K.L. Parsons-Reponte, J. Tien,
S.F. Yu, E. Mai, D. Li, J. Tibbitts, J. Baudys, O.M. Saad, S.J. Scales, P.J. McDonald, P.E.
Hass, C. Eigenbrot, T. Nguyen, W.A. Solis, R.N. Fuji, K.M. Flagella, D. Patel, S.D.
Spencer, L.A. Khawli, A. Ebens, W.L. Wong, R. Vandlen, S. Kaur, M.X. Sliwkowski,
R.H. Scheller, P. Polakis, J.R. Junutula, Conjugation site modulates the in vivo stability
and therapeutic activity of antibody–drug conjugates, Nat. Biotechnol. 30 (2012)
184–189.
[8] L.N. Tumey, M. Charati, T. He, E. Sousa, D. Ma, X. Han, T. Clark, J. Casavant, F. Loganzo,
F. Barletta, J. Lucas, E.I. Graziani, Mild method for succinimide hydrolysis on ADCs:
impact on ADC potency, stability, exposure, and efficacy, Bioconjugate Chem. 25
(2014) 1871–1880.
[9] R.P. Lyon, J.R. Setter, T.D. Bovee, S.O. Doronina, J.H. Hunter, M.E. Anderson, C.L.
Balasubramanian, S.M. Duniho, C.I. Leiske, F. Li, P.D. Senter, Self-hydrolyzing
maleimides improve the stability and pharmacological properties of antibody–
drug conjugates, Nat. Biotechnol. 32 (2014) 1059–1062.
[10] C.F. Meares, M.J. McCall, D.T. Reardan, D.A. Goodwin, C.I. Diamanti, M. McTigue,
Conjugation of antibodies with bifunctional chelating agents: isothiocyanate and
bromoacetamide reagents, methods of analysis, and subsequent addition of metal
ions, Anal. Biochem. 142 (1984) 68–78.
Conflict of interest
The authors declare the following competing financial interest:
R.J.C, R.F., B.B., S.M., R.W., J.H., H.W., C.G., and N.D. are employees of
MedImmune, a member of the AstraZeneca Group, and work to develop
[11] P. Cuatrecasas, M. Wilchek, C.B. Anfinsen, Affinity labeling of the active site of staph-
ylococcal nuclease. Reactions with bromoacetylated substrate analogues, J. Biol.
Chem. 244 (1969) 4316–4329.
[12] O. Koniev, G. Leriche, M. Nothisen, J.S. Remy, J.M. Strub, C. Schaeffer-Reiss, A. Van
Dorsselaer, R. Baati, A. Wagner, Selective irreversible chemical tagging of cysteine
with 3-arylpropiolonitriles, Bioconjugate Chem. 25 (2014) 202–206.
[13] G. Badescu, P. Bryant, J. Swierkosz, F. Khayrzad, E. Pawlisz, M. Farys, Y. Cong, M.
Muroni, N. Rumpf, S. Brocchini, A. Godwin, A new reagent for stable thiol-specific
conjugation, Bioconjugate Chem. 25 (2014) 460–469.
[14] A. Maruani, S. Alom, P. Canavelli, M.T. Lee, R.E. Morgan, V. Chudasama, S. Caddick, A
mild TCEP-based para-azidobenzyl cleavage strategy to transform reversible cysteine
thiol labelling reagents into irreversible conjugates, Chem. Commun. 51 (2015)
5279–5282.
Table 4
In vitro potency of ADCs following serum incubation. Values are listed as the mean
standard error from two experiments.
ADC drug
Day
IC50 (ng/mL)
IC50 (nM)
Max. kill (%)
75
70 1.7
62 1.7
76 1.4
76 1.2
76 1.4
Compound 4
0
1
7
0
1
7
39
57
71
24
25
36
1
1
1
1
1
1
0.26 0.8
0.38 0.8
0.47 0.8
0.16 0.7
0.17 0.7
0.24 0.7
2
Compound 5
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032

R.J. Christie et al. / Journal of Controlled Release xxx (2015) xxx–xxx
11
[15] D. Lin, S. Saleh, D.C. Liebler, Reversibility of covalent electrophile-protein adducts
and chemical toxicity, Chem. Res. Toxicol. 21 (2008) 2361–2369.
[16] M.S.S. Palanki, A. Bhat, R.W. Lappe, B. Liu, B. Oates, J. Rizzo, N. Stankovic, C. Bradshaw,
Development of novel linkers to conjugate pharmacophores to a carrier antibody,
Bioorg. Med. Chem. Lett. 22 (2012) 4249–4253.
[17] S.D. Fontaine, R. Reid, L. Robinson, G.W. Ashley, D.V. Santi, Long-term stabilization of
maleimide–thiol conjugates, Bioconjugate Chem. 26 (2015) 145–152.
[18] J. Kalia, R.T. Raines, Catalysis of imido group hydrolysis in a maleimide conjugate,
Bioorg. Med. Chem. Lett. 17 (2007) 6286–6289.
[19] C.P. Ryan, M.E.B. Smith, F.F. Schumacher, D. Grohmann, D. Papaioannou, G. Waksman,
F. Werner, J.R. Baker, S. Caddick, Tunablereagents for multi-functional bioconjugation:
reversible or permanent chemical modification of proteins and peptides by control of
maleimide hydrolysis, Chem. Commun. 47 (2011) 5452–5454.
[28] S.O. Doronina, T.D. Bovee, D.W. Meyer, J.B. Miyamoto, M.E. Anderson, C.A. Morris-
Tilden, P.D. Senter, Novel peptide linkers for highly potent antibody–auristatin con-
jugate, Bioconjugate Chem. 19 (2008) 1960–1963.
[29] G.M. Dubowchik, R.A. Firestone, L. Padilla, D. Willner, S.J. Hofstead, K. Mosure, J.O.
Knipe, S.J. Lasch, P.A. Trail, Cathepsin B-labile dipeptide linkers for lysosomal release
of doxorubicin from internalizing immunoconjugates: model studies of enzymatic
drug release and antigen-specific in vitro anticancer activity, Bioconjugate Chem.
13 (2002) 855–869.
[30] N. Dimasi, C. Gao, R. Fleming, R.M. Woods, X.-T. Yao, L. Shirinian, P.A. Kiener, H. Wu,
The design and characterization of oligospecific antibodies for simultaneous
targeting of multiple disease mediators, J. Mol. Biol. 393 (2009) 672–692.
[31] J.F. Valliere-Douglass, W.A. McFee, O. Salas-Solano, Native intact mass determination
of antibodies conjugated with monomethyl auristatin E and F at interchain cysteine
residues, Anal. Chem. 84 (2012) 2843–2849.
[20] T. Kitagawa, T. Aikawa, Enzyme coupled immunoassay of insulin using a novel
coupling reagent, J. Biochem. 79 (1976) 233–236.
[32] J. Li, Q. Xu, D.M. Cortes, E. Perozo, A. Laskey, A. Karlin, Reactions of cysteines
substituted in the amphipathic N-terminal tail of a bacterial potassium channel
with hydrophilic and hydrophobic maleimides, Proc. Natl. Acad. Sci. U. S. A. 99
(2002) 11605–11610.
[21] T. Kitagawa, T. Shimozono, T. Aikawa, T. Yoshida, H. Nishimura, Preparation and
characterization of hetero-bifunctional cross-linking reagents for protein modifica-
tions, Chem. Pharm. Bull. 29 (1981) 1130–1135.
[22] M. Machida, M.I. Machida, Y. Kanaoka, Hydrolysis of N substituted maleimides:
stability of fluorescence thiol reagents in aqueous media, Chem. Pharm. Bull. 25
(1977) 2739–2743.
[23] N. Matuszak, G.G. Muccioli, G. Labar, D.M. Lambert, Synthesis and in vitro evaluation
of N-substituted maleimide derivatives as selective monoglyceride lipase inhibitors,
J. Med. Chem. 52 (2009) 7410–7420.
[33] G. Mooser, N-substituted maleimide inactivation of the response to taste cell stimu-
lation, J. Neurobiol. 7 (1976) 457–468.
[34] P. Moody, M.E. Smith, C.P. Ryan, V. Chudasama, J.R. Baker, J. Molloy, S. Caddick,
Bromomaleimide-linked bioconjugates are cleavable in mammalian cells,
Chembiochem 13 (2012) 39–41.
[35] G.J. Quinlan, G.S. Martin, T.W. Evans, Albumin:biochemical properties and therapeutic
[24] G.T. Hermanson, Bioconjugate Techniques, 2nd Edition Elsevier Academic Press Inc.,
potential, Hepatology 41 (2005) 1211–1219.
San Diego, 2008.
[36] A. Andersson, A. Isaksson, L. Brattstrom, B. Hultberg, Homocysteine and other thiols
determined in plasma by HPLC and thiol-specific postcolumn derivatization, Clin.
Chem. 39 (1993) 1590–1597.
[37] S.O. Doronina, B.E. Toki, M.Y. Torgov, B.A. Mendelsohn, C.G. Cerveny, D.F. Chace, R.L.
DeBlanc, R.P. Gearing, T.D. Bovee, C.B. Siegall, J.A. Francisco, A.F. Wahl, D.L. Meyer,
P.D. Senter, Development of potent monoclonal antibody auristatin conjugates for
cancer therapy, Nat. Biotechnol. 21 (2003) 778–784.
[25] T.O. Sippel, The histochemistry of thiols and disulphides. I. The use of N-(4-
aminophenyl)maleimide for demonstrating thiol groups, Histochem. J. 5 (1973)
413–423.
[26] R.J. Christie, D.J. Anderson, D.W. Grainger, Comparison of hydrazone heterobi-
functional cross-linking agents for reversible conjugation of thiol-containing chemistry,
Bioconjugate Chem. 21 (2010) 1779–1787.
[27] R.M. De Figueiredo, P. Oczipka, R. Frohlich, M. Christmann, Synthesis of
[38] J.W. Sadownik, D. Philp, A simple synthetic replicator amplifies itself from a dynamic
4-maleimidobutyric acid and related maleimides, Synthesis (2008) 1316–1318.
pool, Angew. Chem. Int. Ed. 47 (2008) 9965–9970.
Please cite this article as: R.J. Christie, et al., Stabilization of cysteine-linked antibody drug conjugates with N-aryl maleimides, J. Control. Release
(2015), http://dx.doi.org/10.1016/j.jconrel.2015.09.032