Highly Efficient Synthesis of a Staphylococcus aureus Targeting Payload to Enable the First Antibody–Antibiotic Conjugate

Article abstract of DOI:10.1002/chem.201705392

A practical synthesis of the complex payload for an anti-Staphylococcus aureus THIOMAB^TM antibody–antibiotic conjugate (TAC) is described. The route takes advantage of a delicate oxidative condensation, achieved using a semi-continuous flow procedure. It allows for the generation of kilogram quantities of a key intermediate to enable a mild nucleophilic aromatic substitution to the tertiary amine free drug. The linker component is introduced as a benzylic chloride, which allows formation of the quaternary ammonium salt linker-drug. This chemical process surmounts numerous synthetic challenges and navigates deeply colored and unstable compounds to support clinical studies to counter S. aureus bacterial infections.

Full text of DOI:10.1002/chem.201705392

A Journal of
Accepted Article
Title: Highly Efficient Synthesis of a Staphylococcus aureus-targeting
Payload, Enabling the First Antibody-Antibiotic Conjugate
Authors: Xin Linghu, Nathaniel Segraves, Ifat Abramovich, Nicholas
Wong, Barbara Müller, Nadja Neubauer, Serena Fantasia,
Sebastian Rieth, Stephan Bachmann, Michael Jansen, C.
Gregory Sowell, David Askin, Stefan Koenig, and Francis
Gosselin
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201705392
Link to VoR: http://dx.doi.org/10.1002/chem.201705392
Supported by

10.1002/chem.201705392
Chemistry - A European Journal
COMMUNICATION
Highly Efficient Synthesis of a Staphylococcus aureus-targeting
Payload, Enabling the First Antibody-Antibiotic Conjugate
Xin Linghu,[a] Nathaniel L. Segraves,[b] Ifat Abramovich,[b] Nicholas Wong,[a] Barbara Müller,[a] Nadja
Neubauer,[a] Serena Fantasia,[c] Sebastian Rieth,[c] Stephan Bachmann,[c] Michael Jansen,[c] C. Gregory
Sowell,[a] David Askin,[a] Stefan G. Koenig,*,[a] and Francis Gosselin,[a]
Abstract: A practical synthesis of the complex payload for an anti-
Staphylococcus aureus THIOMABTM antibody-antibiotic conjugate
(TAC) is described. The route takes advantage of a delicate
oxidative condensation, achieved via
a semi-continuous flow
procedure. It allows for the generation of kilogram quantities of a
key intermediate to enable a mild nucleophilic aromatic substitution
to the tertiary amine free drug. The linker component is introduced
as a benzylic chloride, which allows formation of the quaternary
ammonium salt linker-drug. This chemical process surmounts
numerous synthetic challenges and navigates deeply colored and
unstable compounds to support clinical studies to counter S. aureus
bacterial infections.
Figure 1. Structure of TAC linker-drug vc-dmDNA31 1.
Resistance of common pathogens, including Staphylococcus
aureus, to the arsenal of antibiotics has been on the rise for
decades, compounded by limited development of new
medications.1 This predicament has resulted in increased
morbidity and mortality, as well as higher costs for the
healthcare system.2 In the case of S. aureus, it has led to
alarming rates of methicillin-resistant strains.3 However,
recognition of the challenge of antibiotic impotence has
galvanized efforts to curtail increasingly dangerous infectious
diseases.4 The power of modern synthetic chemistry creates
an imperative to advance medicines to combat resistance
and prevent common infections from becoming epidemic.1c, 5
The linker-drug 1 was designed to include a proprietary
antibiotic as well as utilize established maleimido-caproyl /
valine-citrulline / p-aminobenzyl linking technology (Figure
1).11 The antibiotic was discovered from of a series of novel
rifamycin S derivatives via semi-synthesis. By applying a
benzylic halide instead of the standard chloroformate
method, a quaternary ammonium salt could be generated
from the tertiary amine in the free drug 6.12 The quaternary
salt in 1 was found to be crucial for the desired antibacterial
activity. The complex structure of the linker-drug was also
sufficiently stable for isolation as well as conjugation, though
a variety of challenges needed to be surmounted to move
into the clinic.
Antibody-Drug Conjugates (ADCs) are promising therapeutic
agents against cancer.6 The concept has been extended to
fight bacterial infections, with our laboratories progressing a
molecule against S. aureus into the clinic.8 In general, ADC
therapeutics feature many technical challenges, including the
(a) monoclonal antibody (mAb), (b) drug, and (c) linkage
between them. While mAb production involves biological
processes, the linker and drug components rely on robust
synthetic organic chemistry.9 Due to the high specificity of
mAbs, the quantities of linker-drugs required are lower than
traditional small molecules. However, molecular complexity
and stability concerns compensate for this lower volume with
greater technical hurdles. Herein we describe a chemical
manufacturing process via a conceptually novel approach to
supply vc-dmDNA31 1 for the S. aureus-targeting construct,
termed a THIOMABTM-antibiotic conjugate (TAC).10
[a]
Dr. X. Linghu, N. Wong, B. Müller, N. Neubauer, Dr. C. G. Sowell,
Dr. D. Askin, Dr. S. G. Koenig, Dr. F. Gosselin
Department of Small Molecule Process Chemistry, Genentech, Inc,
1 DNA Way, South San Francisco, CA 94080, United States
E-mail: Koenig.stefan@gene.com
[b]
[c]
Dr. N. Segraves, I. Gavish-Abramovich
Department of Small Molecule Analytical Chemistry and Quality
Control, Genentech, Inc., 1 DNA Way, South San Francisco, CA,
94080, United States
Dr. S. Fantasia, S. Rieth, Dr. S. Bachmann, Dr. M. Jansen
Pharma Technical Development, Process Chemistry & Catalysis, F.
Hoffmann-La Roche Ltd., Grenzacherstrasse 124, 4070 Basel,
Switzerland
Scheme 1. Process chemistry route to TAC vc-dmDNA31 1.
The medicinal chemistry approach took advantage of the
literature for rifamycin S (2) derivatization.14
In this
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.201705392
Chemistry - A European Journal
COMMUNICATION
1,3-benzenediol 3 to investigate a nucleophilic aromatic
substitution (SNAr) path to 6.18
chromatography-heavy
manganese (IV) dioxide oxidation steps (a) furnished a
known benzoxazine derivative by incorporation of 2-amino-3-
sequence,
two
successive
The reaction of the initial protected aminoresorcinol with 2
utilized super-stoichiometric MnO2, generating heavy metal
waste and requiring chromatography to isolate the
((tert-butyldimethylsilyl)oxy)phenol,
followed
by
(b)
functionalization of the phenol in a distal position with 4-
dimethylaminopiperidine 5 to give the novel tertiary amine 6
(see Scheme 1). Conversion of benzylic alcohol 7 to
benzoxazine.19
In order to more efficiently access
benzoxazine 4, we screened further oxidants. While several
metal complexes, such as Mn and Cu salts, were found to
facilitate the transformation at catalytic levels in the presence
of oxygen, they also presented the aforementioned
requirement to remove the metal following the reaction. In
contrast, we uncovered that the organic nitroxyl radical
TEMPO could be used catalytically with O2 as well as
stoichiometrically under inert conditions,20 the latter being
more desirable from a safety standpoint. Initial laboratory-
scale runs with stoichiometric TEMPO delivered 4 in 39%
yield after isolation.
chloride
8
was achieved with thionyl chloride in
dichloromethane. 8 was then telescoped into an alkylation
with 6 in the presence of NaI to give the crude quaternary
ammonium salt. This crude chloride salt was purified by
HPLC to deliver formate salt 1 in 0.3% overall yield for
conjugation to the THIOMABTM. We set out to craft a more
productive route to reproducibly synthesize 1 for use in
clinical trials.
Fundamental challenges included limited stability of the
rifamycin core as well as highly colored intermediates. While
2 is a well-known orange hue,13 the chromophore changes to
dark red once the aminoresorcinol component
3
is
incorporated to form the conjugated benzoxazine 4. Further,
following reaction with amine 5 to form 6 (and downstream
product 1) the electron-rich benzoxazine presents an
exceptionally dark blue, making routine laboratory
manipulations very challenging.
Extinction coefficients
showed that all intermediates had strong relative molar
absorptivities but that 1 and 6 were most pronounced (Figure
2). The absorbance of 1 even starts to resemble the dye
methylene blue in that light is absorbed at a broad range of
wavelengths, rendering routine detection extremely difficult.
0.7
rifamycin S (2)
benzoxazine (4)
0.6
linker-drug (1)
0.5
free drug (6)
0.4
0.3
0.2
0.1
0
190 240 290 340 390 440 490 540 590 640 690 740 790 840 890 940 990
Scheme 2. Synthesis of 4 with recycle of 10.
Wavelength (nm)
Figure 2. Absorbance of TAC intermediates 2, 4, 6, and 1.
When implemented on scale, the TEMPO-mediated
oxidation allowed access to 4. However, process robustness
issues resulted in lower yields, due to instability under
extended reaction times. During preliminary work, we found
that 2 was capable of facilitating the oxidative condensation
with 3 to 4, presumably via imine 9 or a tautomer (m/z
detected by MS analysis), but with limited conversion due to
With 2 commercially available at reasonable cost from
fermentation, we focused on a similar semi-synthetic strategy
to 1. We also envisioned forming the quaternary ammonium
salt as the final synthetic step to avoid handling complex
cationic species over several stages. Our retrosynthetic
approach would thus use the same tertiary amine 6 as well
as an activated form of linker 7.
quinone
2
also being consumed as oxidant.21
The
hydroquinone by-product of the reaction, rifamycin SV (10),
has been shown to interconvert with 2.22 We thus embarked
on an alternative way to access 4 by recycling the by-product
10 (Scheme 2).
By leveraging 6 as intermediate, the construction of this
tertiary amine became a focus.
However, instead of
incorporating the original mono-protected aminoresorcinol,
we sought to utilize a building block that would remain stable
as the aniline free base with no protecting group.15
A
transition metal-catalyzed C–N coupling with
a
5-
functionalized arene16,17 was set aside in favor of a metal-
free system to avoid catalyst scavenging operations, which
can be ruinous to complex and unstable molecules.
Therefore, we embarked on constructing 2-amino-5-fluoro-
This article is protected by copyright. All rights reserved.

10.1002/chem.201705392
Chemistry - A European Journal
COMMUNICATION
reactor with BQ to facilitate oxidative recycling of 10 to 2,
followed by addition of more 3 to the CSTR. This semi-
continuous approach allowed for quicker oxidation due to
superior mixing and temperature control, thus limiting by-
product formation. The process could be run in an iterative
manner over several cycles, ultimately yielding full
conversion of 2 (~85 A% HPLC) and isolation of 4 in 58-63%
yield and >97% purity.
100
90
80
70
60
50
40
30
20
10
0
Rifa S
Total (S+SV)
With 4 in hand, our focus turned to evaluating the
incorporation of piperidine 5. Having the aryl fluoride in
place allowed the SNAr to proceed smoothly under very mild
conditions at 0–25 oC to cleanly give 6 (Scheme 3). Solvent
exchange allowed for aqueous extraction and precipitation to
give 4 in 77-86% yield and >96% purity. Aqueous extraction
was crucial to removing fluoride ion as well as remaining
amine since 5 could compete as a nucleophile with chloride
8 in the next step.25
Figure 3. Oxidation screening (HPLC A%) of 10.Na to 2.
Evaluation of oxidants to convert the commercially-available
sodium salt of 10 to 2 in the absence of 3 (Figure 3),23
uncovered p-benzoquinone (BQ) as the best candidate due
to clean conversion as well as low cost.
Follow-up
experiments with 3 confirmed that combining reagents in a
batch format was problematic since 3 was prone to oxidative
degradation.24 Vital to developing a successful new process
was to separate the cyclization portion (3 and 2 to 4) from
the oxidative recycling element (10 to 2). We had already
investigated a semi-batch procedure – iteratively adding
TEMPO and 3 in portions – but the oxidant and 11 were
always present together. A similar challenge of limited
conversion and by-product build-up persisted with BQ in
semi-batch mode. A more positive outcome emerged when
we examined continuous processing technology.
Scheme 3. Synthesis of 6 by SNAr displacement.
The final reaction involved coupling amine 6 with 8 to form
the crude quaternary ammonium salt. Development work
revealed that NaI was unnecessary and acid-spiking
experiments showed that residual acid from trace hydrolysis
of 8 was inhibiting conversion.
Incorporation of N,N-
T sensor
diisopropylethylamine allowed for a 1:1 stoichiometry of 6
and 8. However, equimolar base was found to cause
formation of multiple rifamycin ring degradation impurities.
Instead, a trace amount was sufficient to enable smooth
conversion. Collectively, these changes facilitated a clean
alkylation without an array of impurities to complicate
downstream purification. Precipitation with ethyl acetate
gave the crude chloride salt as a dark solid in 86-88% yield
T sensor
3 as
solid
and 86-90 A% HPLC.
Chromatographic purification,
Continuous stirred
tank reactor (CSTR)
followed by lyophilization, delivered the formate salt of vc-
dmDNA31 1 as a dark blue solid in 75-76% yield and >95
A% HPLC purity.
T and P
sensors
In conclusion, we have designed and demonstrated a novel
and robust process to the complex and delicate linker–
antibiotic 1, while navigating a number of stability and
chromatic challenges. This synthetic route takes advantage
of an efficient semi-continuous oxidative condensation to 4.
A mild SNAr reaction to 6 as well as isolation of benzylic
chloride 8 set the stage for an expedient formation of the
quaternary ammonium species in the target. The dark blue
linker-drug 1 is obtained in 29% yield over 4 steps, a
throughput improvement of 100-fold, to support the first anti-
S. aureus TAC construct in the clinic.
iPrOAc
solvent
Benzoquinone
solution
Figure 4. Semi-continuous oxidative condensation set up for 4.
We investigated flow chemistry for this challenging reaction,
with a semi-continuous system delivering a higher-yielding
and more robust synthesis of 4. The condensation portion
was conducted in a standard vessel while the oxidative
recycling segment was achieved with efficient mixing in a
loop flow reactor (Figure 4). Thus, partial conversion of 2 to
4 was realized in a continuous stirred tank reactor (CSTR) in
the absence of external oxidant, with concomitant formation
of 10. The mixture was then pumped through the loop
Note
This article is protected by copyright. All rights reserved.

10.1002/chem.201705392
Chemistry - A European Journal
COMMUNICATION
All authors were employees of Genentech and Roche during the
investigation.
Philips, K. Xu, K. R. Kozak, S. Mariathasan, J. A. Flygare, T. H. Pillow,
Nat. Chem. 2016, 8, 1112-1119.
[13] a) M. R. Holdiness, Med. Toxicol. Adverse Drug Exper. 1989, 4,
444.451; b) J. C. Segen, in The Dictionary of Modern Medicine (Ed.: J.
C. Segen), The Parthenon Publishing Group Inc., Park Ridge, NJ,
2012; pp 515.
Acknowledgements
[14] T. Yamane, T. Hashizume, K. Yamashita, E. Konishi, K. Hosoe, T.
Hidaka, K. Watanabe, H. Kawaharada, T. Yamamoto, F. Kuze, Chem.
Pharm. Bull. 1993, 41, 148-155. Please see SI for more details on the
original synthesis.
The authors thank Remy Angelaud, Alan Deese, Jason
Gruenhagen, Christine Gu, Yi Li, Colin Medley, Thomas Pillow,
Leanna Staben, Sarah Stowers, and Ruth Zuniga of Genentech;
and Fritz Bliss, Nicolas Burki, Stephan Lauper, Christian Lautz,
Shaoning Wang, and Andreas Zogg of Roche for support.
[15] 2-Amino-3-((tert-butyldimethylsilyl)oxy)phenol was utilized in the
original work since the unprotected aminoresorcinol is unstable as a
free base.
[16] a) S. Y. Jabri, L. E. Overman J. Am. Chem. Soc. 2013, 135, 4231-4234;
b) P. Ruiz-Castillo, S. L. Buchwald, Chem. Rev. 2016, 116, 12564-
12649.
Keywords: antibiotic • conjugate • semi-continuous • flow
chemistry • oxidation
[17] After our work started, 5-chloro-aminoresorcinol was reported. See: X.
Yang, G. Shan, Y. Rao, Org. Lett. 2013, 15, 2334-2337.
[18] G. Koenig, J.-F. Blain, 2013, Tetrasubstituted benzenes for treatment of
early onset Alzheimer's disease, WO 2013106328 A1. Our synthesis:
[1]
a) A. Fabbretti, C. O. Gualerzi, L. Brandi, FEBS Lett. 2011, 585, 1673-
1681; b) R. Bax, D. Griffin, Handbook of Experimental Pharmacology
2012, 211, 1-12; c) P. M. Wright, I. B. Seiple, A. G. Myers, Angew.
Chem. Int. Ed. Engl. 2014, 53, 8840-8869; Angew. Chem. 2014, 126,
8984-9014; d) C. L. Ventola, Pharmacy and Therapeutics 2015, 40,
277-283; e) K. Talkington, C. Shore, P. Kothari, A Scientific Roadmap
for
Antibiotic
Discovery,
May
2016,
http://www.pewtrusts.org/~/media/assets/2016/05/ascientificroadmapfor
antibioticdiscovery.pdf.
[19] Several degradation products are formed during the reaction.
[20] J. Kaizer, R. Csonka, G. Speier, J. Mol. Catal. A: Chem. 2002, 180, 91-
96.
[2]
N. D. Friedman, E. Temkin, Y. Carmeli, Clin. Microbiol. Infect. 2016, 22,
416-422.
[3]
[4]
H. F. Chambers, F. R. DeLeo, Nat. Rev. Microbiol. 2009, 7, 629-641.
a) M. A. Cooper, D. Shlaes, Nature 2011, 472, 32 ; b) C. Nathan, O.
Cars, N. Engl. J. Med. 2014, 371, 1761-1763; c) M. A. T. Blaskovich, J.
Zuegg, A. G. Elliott, M. A. Cooper, ACS Infect. Dis. 2015, 1, 285-287; d)
National Action Plan for Combating Antibiotic-Resistant Bacteria, March
2015,
[21] Reactions run under N2 inertion with no external oxidants consistently
gave baseline conversion to 4.
[22] a) Y. Kono, J. Biochem. 1982, 91, 381-95; b) Y. Kono, Y. Sugiura, J.
Biochem. 1982, 91, 397-401; c) G. T. Saez, V. Valls, H. Cabedo, A.
Iradi, W. H. Bannister, J. V. Bannister, Biochim. Biophys. Acta, Mol.
Cell Res. 1991, 1092, 326-35.
https://www.whitehouse.gov/sites/default/files/docs/national_action_pla
n_for_combating_antibotic-resistant_bacteria.pdf; e) Declaration by the
Pharmaceutical, Biotechnology and Diagnostics Industries on
Combating Antimicrobial Resistance, January 2016, https://amr-
review.org/sites/default/files/Industry_Declaration_on_Combating_Anti
microbial_Resistance_UPDATED%20SIGNATORIES_MAY_2016.pdf
[23] See SI for conversion data for oxidation of the commercially-available,
stable sodium salt of 10 to 2, supported by cyclic voltammatry (CV).
[24] Degradation of 3 results in insoluble material forming in the reaction.
[25] For details on 8 as a stable intermediate from an optimized reaction /
isolation process, see SI.
[5]
[6]
[8]
J.
Gale,
H.
Fourcade,
June
2016,
https://www.bloomberg.com/news/articles/2016-06-30/superbugs-and-
subsidies-draw-big-pharma-back-to-antibiotics
a) J. Mantaj, P. J. M. Jackson, K. M. Rahman, D. E. Thurston, Angew.
Chem. Int. Ed. 2017, 56, 462-488; b) R. V. J. Chari, ACS Med. Chem.
Lett. 2016, 7, 974-976; c) H. Ledford, Nature 2016, 540, 19-20.
S. M. Lehar, T. Pillow, M. Xu, L. Staben, K. K. Kajihara, R. Vandlen, L.
DePalatis, H. Raab, W. L. Hazenbos, J. H. Morisaki, J. Kim, S. Park, M.
Darwish, B.-C. Lee, H. Hernandez, K. M. Loyet, P. Lupardus, R. Fong,
D. Yan, C. Chalouni, E. Luis, Y. Khalfin, E. Plise, J. Cheong, J. P.
Lyssikatos, M. Strandh, K. Koefoed, P. S. Andersen, J. A. Flygare, M.
W. Tan, E. J. Brown, S. Mariathasan, Nature 2015, 527, 323-328.
A. S. Prashad, B. Nolting, V. Patel, A. Xu, B. Arve, L. Letendre, Org.
Process Res. Dev. 2017, 21, 590-600.
[9]
[10] For more on Genentech-engineered mAbs (THIOMABTM
) with two
defined cysteine thiols for site-specific conjugation, see: J. R. Junutula,
H. Raab, S. Clark, S. Bhakta, D. D. Leipold, S. Weir, Y. Chen, M.
Simpson, S. P. Tsai, M. S. Dennis, Y. Lu, Y. G. Meng, C. Ng, J. Yang,
C. C. Lee, E. Duenas, J. Gorrell, V. Katta, A. Kim, K. McDorman, K.
Flagella, R. Venook, S. Ross, S. D. Spencer, W. Lee Wong, H. B.
Lowman, R. Vandlen, M. X. Sliwkowski, R. H. Scheller, P. Polakis, W.
Mallet, Nat. Biotechnol. 2008, 26, 925-932.
[11] The linker was developed under license from Seattle Genetics.
[12] L. R. Staben, S. G. Koenig, S. M. Lehar, R. Vandlen, D. Zhang, J. Chuh,
S.-F. Yu, C. Ng, J. Guo, Y. Liu, A. Fourie-O’Donohue, M. A. Go, X.
Linghu, N. L. Segraves, T. Wang, J. Chen, B. Q. Wei, G. D. Lewis
This article is protected by copyright. All rights reserved.

10.1002/chem.201705392
Chemistry - A European Journal
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 2:
COMMUNICATION
X. Linghu, N. L. Segraves, I. Gavish-
Abramovich, N. Wong, B. Müller, N.
Neubauer, S. Fantasia, S. Rieth, S.
Bachmann, M. Jansen, C. G. Sowell, D.
Askin, S. G. Koenig,* and F. Gosselin
Page No. – Page No.
Text for Table of Contents
Highly Efficient Synthesis of a
Staphylococcus aureus-targeting
Payload, Enabling the First Antibody-
Antibiotic Conjugate
This article is protected by copyright. All rights reserved.