Discovery of Pyrophosphate Diesters as Tunable, Soluble, and Bioorthogonal Linkers for Site-Specific Antibody-Drug Conjugates

Article abstract of DOI:10.1021/jacs.5b12547

As part of an effort to examine the utility of antibody-drug conjugates (ADCs) beyond oncology indications, a novel pyrophosphate ester linker was discovered to enable the targeted delivery of glucocorticoids. As small molecules, these highly soluble phosphate ester drug linkers were found to have ideal orthogonal properties: robust plasma stability coupled with rapid release of payload in a lysosomal environment. Building upon these findings, site-specific ADCs were made between this drug linker combination and an antibody against human CD70, a receptor specifically expressed in immune cells but also found aberrantly expressed in multiple human carcinomas. Full characterization of these ADCs enabled procession to in vitro proof of concept, wherein ADCs 1-22 and 1-37 were demonstrated to afford potent, targeted delivery of glucocorticoids to a representative cell line, as measured by changes in glucocorticoid receptor-mediated gene mRNA levels. These activities were found to be antibody-, linker-, and payload-dependent. Preliminary mechanistic studies support the notion that lysosomal trafficking and enzymatic linker cleavage are required for activity and that the utility for the pyrophosphate linker may be general for internalizing ADCs as well as other targeted delivery platforms.

Full text of DOI:10.1021/jacs.5b12547

Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI
Article
Discovery of pyrophosphate diesters as tunable, soluble and
bioorthogonal linkers for site-specific antibody-drug conjugates
Jeffrey C. Kern, Mark Cancilla, Deborah Dooney, Kristen Kwasnjuk, Rena Zhang, Maribel Beaumont, Isabel
Figueroa, SuChun Hsieh, Linda Liang, Daniela Tomazela, Jeffrey Zhang, Philip E. Brandish, Anthony Palmieri,
Peter Stivers, Mangeng Cheng, Guo Feng, Prasanthi Geda, Sanjiv Shah, Andrew Beck, Damien Bresson, Juhi
Firdos, Dennis Gately, Nick Knudsen, Anthony Manibusan, Peter G Schultz, Ying Sun, and Robert M. Garbaccio
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b12547 • Publication Date (Web): 08 Jan 2016
Downloaded from http://pubs.acs.org on January 11, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
Journal of the American Chemical Society is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society. However, no
copyright claim is made to original U.S. Government works, or works produced by employees of any
Commonwealth realm Crown government in the course of their duties.

Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society. However, no
copyright claim is made to original U.S. Government works, or works produced by employees of any
Commonwealth realm Crown government in the course of their duties.

Page 1 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Title: Discovery of pyrophosphate diesters as tunable, soluble and bioorthogonal linkers
for siteꢀspecific antibodyꢀdrug conjugates
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
2
2
Authors: Jeffrey C. Kern , Mark Cancilla , Deborah Dooney , Kristen Kwasnjuk , Rena
2
3
3
3
3
Zhang , Maribel Beaumont , Isabel Figueroa , SuChun Hsieh , Linda Liang , Daniela
3
3
4
4
4
Tomazela , Jeffrey Zhang , Philip E. Brandish , Anthony Palmieri , Peter Stivers ,
5
5
5
5
6
Mangeng Cheng , Guo Feng , Prasanthi Geda , Sanjiv Shah Andrew Beck , Damien
6
6
6
6
6
Bresson , Juhi Firdos , Dennis Gately , Nick Knudsen , Anthony Manibusan , Peter G.
6
6
1
Schultz, Ying Sun, Robert M. Garbaccio *
1
2
3
Author Affiliations: Merck Global Chemistry; Merck PPDM; Merck Biologics;
4
5
6
Merck Biology, Respiratory and Immunology; Merck In Vitro Pharmacology; Ambrx
Abstract: As part of an effort to examine the utility of antibodyꢀdrug conjugates (ADCs)
beyond oncology indications, a novel pyrophosphateꢀester linker was discovered to
enable the targeted delivery of glucocorticoids. (Figure 1) As small molecules, these
highlyꢀsoluble phosphateꢀester drugꢀlinkers were found to have ideal orthogonal
properties: robust plasma stability coupled with rapid release of payload in a lysosomal
environment. Building upon these findings, siteꢀspecific ADCs were made between this
drugꢀlinker combination and an antibody against human CD70, a receptor specifically
expressed in immune cells but also found aberrantly expressed in multiple human
carcinomas. Full characterization of these ADCs enabled procession to in vitro proof of
concept, wherein ADCs 1-22 and 1-37 were demonstrated to afford potent, targeted
ACS Paragon Plus Environment

Page 3 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
delivery of glucocorticoids to a representative cell line as measured by changes in
glucocorticoid receptor (GR) mediated gene mRNA levels. These activities were found to
be antibody, linker and payload dependent. Preliminary mechanistic studies support the
notion that lysosomal trafficking and enzymatic linkerꢀcleavage is required for activity,
and that the utility for the pyrophosphate linker may be general for internalizing ADCs as
well as other targeted delivery platforms.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1: Phosphate diester linkers for antibodyꢀdrug conjugates
Introduction: The antibodyꢀdrug conjugate (ADC) modality is enjoying a renewal of
interest and success due to advances in expression, conjugation and linkerꢀpayload
technology that have paved the path for the recent approvals of Adectris® and Kadcyla®
1
,2
for Hodgkin lymphoma and Herꢀ2 positive metastatic breast cancer respectively. In all,
nearly thirty ADCs are reported to be in clinical development, all of which are aimed to
3
address unmet medical need in oncology. As a result, while there is significant diversity
in the monoclonal antibodies (mAbs) employed in these candidates, the options among
the varied linkers and payloads are limited to a select few that are used repeatedly due to
the clinical experience and common purpose of targeted cytotoxicity. To fully explore the
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
potential of this modality, particularly in its applicability beyond oncology, there is a
clear need to develop alternative linker designs that address present limitations and
facilitate the use of the larger payload pharmacopeia.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As part of an effort to test the ADC modality beyond oncology, a project was
initiated to achieve the targeted delivery of glucocorticoids to immune cells.
Glucocorticoids are potent steroid hormones that have powerful antiꢀinflammatory
4
a
properties with important utility in a range of immunological diseases. Unfortunately,
glucocorticoids also suffer from an array of sideꢀeffects (e.g. effects on glucose, bone,
mood) due to their pleotropic pharmacology and wide biodistribution as small molecule
4
b
therapeutics. Decades of research in the medicinal chemistry community have been
devoted to the concepts of “dissociated” or “tissueꢀselective” synthetic glucocorticoids to
overcome their limited therapeutic index, and, arguably, the most successful approaches
have come through topical or inhaled delivery methods where biodistribution can be
5
physically limited. Despite this focus, the goal of achieving a systemic glucocorticoid
devoid of these doseꢀlimiting side effects has remained elusive. This problem of
therapeutic index has recently been revisited through the lens of the ADC modality using
6
7
heterogeneous conjugation with Eꢀselectin and antiꢀCD163 using esterꢀbased linkers for
the glucocorticoid. Our project aimed to leverage siteꢀspecific conjugation technology
8
using unnatural amino acid incorporation to enable clear SAR for an optimal drugꢀlinker
combination, and as a preliminary goal, proofꢀofꢀconcept in a human immune cell.
Within the ADC modality, the linker functions to stably attach a potent payload to
a circulating mAb while at the same time, acts as a trigger for payload release following
internalization of that mAb within its target cell. This bioꢀorthogonal property relies on
ACS Paragon Plus Environment

Page 5 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
differences in the physiological environment of the endosomalꢀlysosomal pathway
relative to the extracellular matrix. In taking advantage of these physiological differences,
four distinct linker designs (and physiological triggers) have been successfully advanced
to clinical trials as part of the current wave of oncology ADCs: hydrazones (acid),
disulfides (reducing thiols), cathepsin Bꢀcleavable dipeptides (enzymes) and nonꢀ
cleavable (mAb catabolism). This toolset has performed well for a limited set of
cytotoxic payloads, but we recognized limitations in each design as we looked to leverage
them for the glucocorticoid payloads. Given the nuances of glucocorticoid pharmacology
with the potential for partial agonism/antagonism, we sought to design an ADC that
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
released dexamethasone in its parental form without any residual linker. Early
publications with a similar goal employed an esterꢀbased linker to dexamethasone that we
evaluated and found to be rapidly hydrolyzed in blood suggesting that circulatory
stability for this linker design would be subꢀoptimal. Thus, we aimed to identify a novel
linker for the delivery of glucocorticoids that would provide stable attachment to a
circulating mAb as well as efficient release of the payload once internalized into an
antigenꢀpositive cell. This work led to the discovery of a family of highlyꢀsoluble
phosphateꢀester linkers that have tunable and optimal properties for this specific goal, and
may have wide applicability to the ADC and bioconjugate field. Importantly, in vitro
proofꢀofꢀconcept was established using αꢀhCD70 conjugates 1-22 and 1-37 through siteꢀ
1
0
specific conjugation technology coupled with strainꢀpromoted 3+2 click chemistry. αꢀ
hCD70 was chosen as a model antibody for achieving proofꢀofꢀconcept based on its
precedent in the oncology ADC literature as a targeting vector, and the validated cell
lines to test with that are also responsive to glucocorticoids. The resulting conjugates
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
have robust SAR with clear evidence of antibody, linker and payloadꢀdependent activity
and highlight the potential of the novel class of bioorthogonal linkers for targeted
delivery.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Chemistry:
The desired properties adopted for a linker that would enable the targeted delivery of
glucocorticoids CD70+ cells using an ADC approach were as follows: soluble, stable in
extracellular circulation and rapid payload release in a lysosomal environment. It was
hypothesized that phosphate diester linkers could provide these properties, and be
applicable to a range of glucocorticoids with varying potency (see Results and Discussion
section below for further explanation).
Synthesis of Glucocorticoid-Phosphate Ester Linkers: The synthesis of
dexamethasone phosphateꢀester linkers (4, 7, and 10) was accomplished according to
Scheme 1. Dexamethasone (1) was reacted with pyrophosphate tetrachloride at low
temperature to selectively generate the Cꢀ21 reactive phosphorodichloridate species in
situ with no reaction observed at the moreꢀhindered Cꢀ11 and Cꢀ17 positions. Upon
consumption of the steroid, the corresponding alcohol 2 or 5 was introduced along with
triethylamine at low temperature to produce 3 and 6 respectively, in a one pot manner,
with good yields after aqueous workup. Acidic deprotection of 3 and 6 followed by EDC
1
0
coupling with 2ꢀ(cyclooctꢀ2ꢀynꢀ1ꢀyloxy)acetic acid provided target dexamethasone
11
linkers 4 and 7 in low yields. To obtain the phosphoramidate 10, dexamethasoneꢀ21ꢀ
ACS Paragon Plus Environment

Page 7 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
phosphate (8) was isolated upon aqueous workup from the conditions described above.
EDC coupling of 2ꢀ(cyclooctꢀ2ꢀynꢀ1ꢀyloxy)acetic acid with ethylene diamine yielded
intermediate 9. A coupling reaction of 8 and 9 was carried out in aqueous tertꢀbutanol
using DCC to give the phosphoramidate 10.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
OH
O P O HN R2
OH
O
O
Me
Me
HO
Me
OH
Me
HO
Me
OH
a
H
+
NHBoc
H
Me
HO
F
H
F
H
O
O
3, R = Boc
2
1
2
b, c
O
4, R =
2
O
a. P O Cl , THF, -40°C, 1h, then Et N and 2, 4h, aq. workup, 73%; b. HCl, EtOAc, 1h, rt; 100%
2
3
4
3
c. 2-(cyclooct-2-yn-1-yloxy)acetic acid, EDCI, HOAt, Et N, CH Cl , 37%.
3
2
2
O
OH
O P O
OH
NHR2
O
O
Me
Me
HO
Me
OH
Me
HO
Me
OH
NHBOC
a
H
+
H
Me
F
H
HO
F
H
O
O
6
, R = BOC
2
1
5
b, c
O
7
, R =
2
O
a. P O Cl , THF, -40°C, 75min, then Et N and 5, 30min, aq. workup, 88%; b. HCl, EtOAc, rt, 1h; 100%
2
3
4
3
c. 2-(cyclooct-2-yn-1-yloxy)acetic acid, EDCI, HOAt, Et N, CH Cl , 9%.
3
2
2
O
O
OPO H
O P NH HN
OH
3
2
O
O
O
Me
Me
HO
Me
OH
Me
HO
Me
OH
a
H
+
O
H
Me
H N
2
O
F
H
N
H
F
H
O
O
8
9
10
a. DCC, tBuOH/water, 100°C, 4h, 30%.
Scheme 1: Synthesis of series of dexamethasone phosphate linker molecules 4,7,10.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The synthesis of dexamethasone and budesonide pyrophosphate diester linkers
1
2
(
14, 18 and 22) was accomplished according to Scheme 2. This approach to
pyrophosphate synthesis involved the coupling of two terminal phosphates. First,
alcohols 11 and 15 were phosphorylated to create the necessary intermediates 12 and 16
respectively. Dexamethasone phosphate (8) was first treated with CDI in DMF to form
an active phosphorimidazolidate species. To this was added either 12 or 16 and ZnCl2,
which yielded the dexamethasone pyrophosphates 13 and 17. Deprotection of the Fmoc
group with piperidine followed by EDC coupling with 2ꢀ(cyclooctꢀ2ꢀynꢀ1ꢀyloxy)acetic
acid provided target pyrophosphate linkers 14 and 18. The budesonide phosphate 20 was
activated to its phosphorimidazolidate and coupled with 12 as above to provide the
budesonide pyrophosphate 21. 21 was deprotected in the same manner but instead of
EDC coupling, 21 was alternatively reacted with the preformed phthalate ester of the
cyclooctyne providing a cleaner reaction with superior yield to give pyrophosphate linker
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2.
ACS Paragon Plus Environment

Page 9 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2: Synthesis of dexamethasone and budesonide pyrophosphate diester linker
molecules 14, 18, 22.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The synthesis of a triphosphate diester (28) was accomplished using sequential
couplings according to Scheme 3. To achieve the triphosphate linkage, fluorenylꢀmethyl
phosphate 24 was derived by phosphorylation of fluorenyl methanol 23, and coupled to
dexamethasone phosphate (8) as was described above. Deprotection of the resulting
intermediate pyrophosphate ester 25, yielded the terminal pyrophosphate 26. A second
coupling reaction of 26 with the active phosphorimidazolidate species of 12 provided the
Fmoc protected triphosphate 27. Deprotection of the Fmoc group followed by reaction
with the preformed phthalate ester of the cyclooctyne provided the desired triphosphate
diester 28.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 11 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Scheme 3: Synthesis of dexamethasone triphosphate diester linker molecule 28.
To include fluticasone propionate as a payload alternative, it was necessary to
construct the pyrophosphate linker from the 11ꢀposition since it contains the only
hydroxyl handle in the steroid (Scheme 4). Due to the steric hindrance around the 11ꢀ
position of the steroid from the flanking 18 and 19ꢀposition methyl groups, it was not
possible to directly phosphorylate using the conditions demonstrated for dexamethasone
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
and budesonide. Instead, a stepwise approach was discovered. Fluticasone propionate
13
(
29) was first reacted with PCl at low temperature to install the phosphite giving
3
compound 30 following aqueous workup. Phosphite 30 was then oxidized and trapped as
1
4
the phosphorimidazolidate species 31. Coupling of 31 with phosphate 12 provided the
1
1ꢀposition pyrophosphate ester intermediate 32. Deprotection of 32 using piperidine
followed by HATU coupling with 2ꢀ(cyclooctꢀ2ꢀynꢀ1ꢀyloxy)acetic acid provided the
desired 11ꢀposition pyrophosphate fluticasone propionate linker 33 in good overall yield.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 4: Synthesis of fluticasone propionate pyrophosphate diester linker molecule 33.
As mitigation for the potential for poor phosphate hydrolysis on the hindered
1
5
secondary alcohol of fluticasone propionate, an acetal spacer was synthesized to
facilitate efficient release of the parent drug in the acidic lysosomal environment
(Scheme 5). Due to the steric hindrance around the hydroxyl group, Sn2 alkylations with
3
sp electrophiles were unsuccessful at introducing the acetal. The introduction of thio
ether to give compound 34 was achieved by reaction of fluticasone propionate (29) with
2
the sp sulfonium species generated from the mixture of dimethyl sulfide and benzoyl
16
peroxide. This reaction proceeded in low yield with oxidation of the 11ꢀposition
hydroxyl as the major competing side reaction. Activation of 34 with NIS and reaction
with crystalline phosphoric acid under strict anhydrous conditions provided the key
phosphate acetal intermediate 35. Activation of the Fmoc phosphate 12 to its
ACS Paragon Plus Environment

Page 13 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
phosphorimidazolidate species followed by coupling with 35 yielded the pyrophosphate
diester 36. Fmoc deprotection with DBU, followed by coupling 2ꢀ(cyclooctꢀ2ꢀynꢀ1ꢀ
yloxy)acetic acid using HATU provided the desired acetal pyrophosphate diester linker
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
7.
Scheme 5: Synthesis of acetalꢀspaced fluticasone propionate pyrophosphate diester linker
molecule 37.
Additional details on the synthesis and characterization of the small molecules are
contained within the experimental and supporting information.
Synthesis of α-hCD70 Antibody Glucocorticoid Conjugates
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
7
Drugꢀlinkers were conjugated to an αꢀhuman CD70 antibody (2h5, IgG1) and an αꢀ
RSV antibody (based on Synagis with a H32Y mutation on the heavy chain to further
decrease binding to RSV F protein, IgG1/Kappa isotope) to produce antibodyꢀdrug
conjugates. Specifically, the drugꢀlinker was conjugated to the unnatural amino acid
paraꢀazidoꢀphenylalanine (pAF) replacing the alanine at position 1 of CH1 (HA114) of
the antibody using copperꢀfree 3+2 cycloaddition chemistry as shown in Scheme 6.
based on Synagis with a H32Y mutation on the heavy chain to further decrease binding to
RSV F protein on an IgG1/Kappa isotope (ref: in US 7,632,924 and copperꢀfree 3+2
cycloaddition chemistry has been described in US 7,807,619.)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 6: Representative conjugation of αꢀhCD70 conjugate
All of the characterization data for the ADCs has been included in the supporting
information.
Results and Discussion:
ACS Paragon Plus Environment

Page 15 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Design and Evaluation of Glucocorticoid Phosphate Ester Linkers: The desired
properties adopted for a linker that would enable the targeted delivery of glucocorticoids
CD70+ cells using an ADC approach were as follows: soluble, stable in extracellular
circulation and rapid payload release in a lysosomal environment. Due to the lipophilic
nature of glucocorticoids, we reasoned that a solubilizing linker would not only facilitate
aqueous bioconjugation to an antibody, but would also mitigate against the potential for
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
8
ADC aggregation seen with other lipophilic payloads. Furthermore, to maximize
therapeutic index and provide prolonged halfꢀlife to a potential therapeutic, a wellꢀ
designed linker should minimize nonꢀspecific release of glucocorticoids in compartments
seen by a circulating mAb. Finally, following antigen recognition and internalization into
a target cell, the linker should maximize the delivery of its glucocorticoid payload in its
parental form as efficiently as possible. Given these design criteria, we chose to focus on
strategies that relied on an enzymaticꢀtrigger for release wherein the enzyme family used
had selective expression in the lysosome (as was done with the cathepsinꢀsensitive
1
9
20
linkers). Among the numerous enzyme families present in the lysosomal proteome,
we were inspired by phosphodiesterases as a general family and their substrate phosphate
diesters for their unique property of stably linking together two molecules, and retaining a
2
1
charged, solubilizing state.
The goal then was to synthesize phosphate diesters that connected the payload (e.g.
dexamethasone) with a handle that would enable conjugation to a monoclonal antibody.
22
An expression technology was adopted which facilitates this conjugation by the
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
incorporation of unnatural amino acids (UAAs) with orthogonal reactivity to create siteꢀ
specific ADCs with a drugꢀantibody ratio nearly equal to 2. Among the UAAs
demonstrated with this platform, pꢀazido phenylalanine is particularly powerful when
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
paired with the strainꢀinduced, copperꢀfree cycloaddition of cycloalkynes. Thus, the
design for these drug linkers included a cyclooctyneꢀterminus that could easily be
coupled to an unnatural amino acidꢀcontaining mAb. For the payload end of the
phosphate ester, clinicallyꢀutilized dexamethasone has three potential sites of conjugation
in the C11, C17 and C21ꢀalcohols. We hypothesized that the most sterically accessible
alcohol (C21) would lead to the most efficient cleavage and produced a series of drugꢀ
linker molecules that explored the potential of this concept. The synthesis of
dexamethasone Cꢀ21 phosphate esters 4 and 7, phosphoramidate 10, pyrophosphate esters
1
4 and 18 and triphosphate ester 28 are described in the methods section, and structures
are shown in Figure 2.
ACS Paragon Plus Environment

Page 17 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Figure 2: Series of dexamethasone phosphateꢀlinker molecules.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As part of the medicinal chemistry program, we wanted to also explore the impact of
increasing payload potency within the context of this α−CD70ꢀpyrophosphate ester linker
design. Budesonide and fluticasone propionate represent increasingly potent alternatives
to dexamethasone that have the added theoretical benefit of longer offꢀrates from the GR
2
4
as a mechanism for cellular retention. The linker designs synthesized above were easily
adapted to the free C21 alcohol within the budesonide structure, but extension to
fluticasone propionate and its hindered C11 alcohol proved problematic. (Figure 3) With
the development of unique chemistry to overcome the poor steric accessibility to this
alcohol, both the pyrophosphate of fluticasone propionate and an acetalꢀspaced variant
were produced to mitigate the potential for slow bioconversion of the hindered
phosphate.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3: Series of budesonide and fluticasone propionate pyrophosphateꢀlinker
molecules 22, 33, and 37.
All of the above phosphateꢀlinker molecules were found to be chemically stable,
amenable to reverse phase purification, and to have outstanding aqueous solubility
(>5mg/mL) thus meeting our first design criteria.
Stability Screening of Dexamethasone Phosphate Ester Linkers:
An evaluation of the bioorthogonal stability of this series of drug linkers was conducted
under the following assumption; if, as a small molecule, a drug linker was stable in blood
and reactive in a lysosomal environment, it would retain those properties as an antibody
conjugate. Put simply, if this series of drug linkers did not release dexamethasone in a
ACS Paragon Plus Environment

Page 19 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
lysosomal environment as small molecules, it was assumed that they would not perform
well as antibody conjugates. To test this approach, stability assays that measured for
disappearance of parent and appearance of dexamethasone were conducted in blood
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
mouse, human) and purified lysosomal lysates (rat). These lysosomal lysates have been
2
5
used previously to evaluate ADC stability for cathepsinꢀsensitive linkers.
The results of these tests are summarized in Figure 4. Dexamethasone itself was shown
to be completely stable for the 6h duration of the assay in all matrices examined. Not
surprisingly, for the phosphateꢀdiester series of drugꢀlinkers 4, 7, 10, 14, 18, and 28, all
compounds tested were determined to be completely stable in mouse and human blood up
to 6 hours (data in supporting information). While terminal phosphates are wellꢀknown
prodrugs that convert readily in the blood compartment, phosphate esters are generally
stable molecules. However, to our surprise, these phosphateꢀdiester drug linkers
displayed a wide range of reactivity in the lysosomal lysate stability assay from slow
release of dexamethasone (linkers 4, 7, 10, 18) to rapid release (linkers 14 and 28)
attesting to the tunable and bioorthogonal nature of their design. An important and
unexpected conclusion from this small molecule work was that pyrophosphateꢀ and
triphosphate diesters 14 and 28 had a unique reactivity profile relative with rapid
conversion to payload relative to the monophosphate diesters. Interestingly, the phenolic
pyrophosphate gave an intermediate reactivity profile, and this suggests the rate of
release may be further tuned by substitution proximal to the pyrophosphate.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
00
9
8
7
0
0
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
7
1
1
1
2
60
0
4
8
8
5
4
3
0
0
0
20
1
0
0
Matrix 0m
spiking
5m
10m
15m
30m
1hr
2hr
3hr
6hr
Sampling Time
Figure 4: Rat lysosomal lysate conversion of dexamethasone drug linkers to
dexamethasone
Drug linkers for budesonide and fluticasone were also examined in this assay.
Budesonide pyrophosphate diester 22 demonstrated a similar profile to the
dexamethasone comparator 14 (data in supporting information), but the Cꢀ11 connected
fluticasone propionate pyrophosphate diester 33 did not convert to parent fluticasone
propionate, but instead stalled at intermediate phosphate 38 (Scheme 7). The incubation
in lysosomes revealed rapid conversion to the phosphate intermediate for which further
conversion to payload was not seen. Given the steric hindrance of the Cꢀ11 alcohol
encountered in the synthesis of this compound, it is hypothesized that the intermediate
phosphate is a poor substrate for further conversion. In contrast, the acetalꢀspaced
ACS Paragon Plus Environment

Page 21 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
fluticasone pyrophosphate diester 37 was observed to convert to the intermediate
phosphate and to proceed further with possible acidic hydrolysis to the parent fluticasone
propionate. Thus the introduction of the acetal spacer was required for this linker strategy
to be extended to the Cꢀ11 locus available for the most potent glucocorticoid examined.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 7: Lysosomal lysate conversion of fluticasone propionate linkers 33 and 37
The stable profile in blood coupled to the efficient release in a lysosomal environment
meet the criteria for linker design within the ADC modality. Thus, a representative mAb
was sought out to express with an unnatural amino acid in the CH1 domain, to make siteꢀ
specific conjugates, and to test the functional utility of these novel linker designs in an in
vitro system. At a high level, the goal of the project was to effect targeted delivery of
glucocorticoids using model cell lines. For this purpose, αꢀCD70 was selected given the
expression of CD70 on relevant human Tꢀ cells coupled with the precedent in the
literature as an ADC. αꢀCD70 clones have been demonstrated to be efficient,
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
26
internalizing vectors for the targeted delivery of cytotoxic molecules like the auristatins
2
7
and PBD.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Synthesis and characterization of α-hCD70 (2h5) Conjugates
In order to test the functional performance of these novel linker designs, the team
2
8
identified αꢀhCD70 (2h5)
as a fitꢀforꢀpurpose system. The αꢀhCD70 mAb was
engineered with a mutation in the heavy chain (HA114) to include p-azido Phe.
Conjugates were produced by combining this mAb (denoted as 1-x below) with excess
drugꢀlinker molecules 4, 7, 10, 14, 18, 22, 28, 33, 37 to drive the reaction to completion
and produce homogeneous conjugates with a DAR >1.8 (see supporting information).
These conjugates displayed high monomeric content (>98%) and were found to have
<
1% residual drug linker following purification. An analogous set of conjugates were
produced with αꢀRSV (IgG1ꢀHA114ꢀpAF3) using the same positional mutation to create
the opportunity for a matched negative control (denoted as 2-x below).
In vitro stability of α-hCD70 Conjugates
One of the benefits of working with siteꢀspecific ADCs is the ability to fullyꢀleverage
mass spectrometry to understand their metabolism. To firmly establish the stability of this
linker design in plasma, and to gain insight into its mechanism of cleavage in a lysosomal
environment, the αꢀhCD70 conjugates 1-22 and 1-37 with the pyrophosphate linker were
studied in these matrices (Figure 5).
ACS Paragon Plus Environment

Page 23 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
α−hCD70ꢀADC 1-22 and 1-37 were incubated in C57BL/6 mouse and human plasma for
up to 7 days, affinity purified, and analyzed by LCꢀMS for intact mass with no observed
loss of budesonide (19) or fluticasone propionate (29) respectively. ADCs 1-22 and 1-37
were found intact in vitro up to 7 days and the drugꢀtoꢀantibody ratio (DAR) calculated
for time points through day 7 was unchanged (see supporting information) In contrast,
incubation of this ADC in lysosomal lysates for 6h did result in the release of budesonide
monophosphate (20) followed by conversion to budesonide (19) for 1ꢀ22. (Figure 5) For
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
-37, fluticasone propionate acetal monophosphate (35) was also released as a first step,
followed by rapid conversion to fluticasone propionate (29) itself. As a control, these
lysates were boiled prior to incubation and no payload release was observed consistent
with an enzymatic mechanism of release and not acidic release. The identity of the
enzyme(s) involved in the linker cleavage is unknown, although it is hypothesized that
two different enzymes are involved – one to clip the pyrophosphate ester, and the second
(likely a phosphatase) to hydrolyze the remaining terminal phosphate. Final acetal
breakdown is likely fast and acid mediated. Consistent with the properties observed with
the small molecule drugꢀlinkers 22 and 37, the analogous αꢀhCD70ꢀADCs 1-22 and 1-37
were found to be stable in plasma, and reactive in a lysosomal environment.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Release of payload in lysosome lysates: α-hCD70-Budesonide
(
1-22)
1
.6
1.4
1
.2
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0.8
Bud mono-phosphate
Budesonide
0
0
.6
.4
0.2
0
0
2
4
6
8
Time (hr)
Release of payload in lysosome lysates: α-hCD70-Fluticasone
propionate (1-37)
2
1
.4
.9
1.4
Flu mono-phosphate
Flu propionate
0
0
.9
.4
-
0.1
0
2
4
6
8
Time (hr)
Figure 5: Stability profile of ADC 1-22 and 1-37 in lysosomal lysates (6h)
In vitro potency of α-hCD70(2h5) and α-RSV Conjugates
Cellular assays were developed to test the performance of these siteꢀspecific ADCs
carrying phosphate ester linked glucocorticoid payload. As a proximal readout of target
engagement with the glucocorticoid receptor (GR), a panel of GRꢀtargeted genes was
ACS Paragon Plus Environment

Page 25 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
evaluated for their change in mRNA expression in these cells as a function of
dexamethasone exposure and time. GILZ mRNA was identified as providing the most
optimal signal for GR target engagement and was chosen as the main readout for
glucocorticoid response in these cells. The ADCs were evaluated for their ability to
increase GILZ mRNA expression in targetꢀ, linkerꢀ and payloadꢀdependent manners with
the matched negative controls included.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As shown in Figure 6, there were linkerꢀdependent GILZ activities induced by
2
7
these α−hCD70 ADCs carrying phosphate ester linked dexamethasone in 786ꢀO cells, a
cell line with high CD70 expression on cell surface (flow cytometry, data not shown).
Overall, there was a clear relationship between dexamethasone release rate of the drug
linkers in the lysosomal lysate assay (Figure 4) and the GILZ activity in 786ꢀO cells. For
example, the drug linker 4 that was observed to show slow or negligible dexamethasone
release in the lysosomal lysates was only weakly active or inactive as an ADC in the
GILZ assay in 786ꢀO cells (ADCꢀ1-4). In contrast, the drug linker 14 that demonstrated
rapid dexamethasone release in the lysosomal lysates were among the most active as
ADCs in the GILZ assay in 786ꢀO cells (ADCꢀ1-14). The same drugꢀlinkers as
conjugates with the negative control αꢀRSV were inactive in 786ꢀO cells except at the
highest concentration. While we regard this diminished activity of the negative controls
(~200x) to support our proofꢀofꢀconcept for targeted delivery, it is possible that these αꢀ
RSV conjugates are taken up by the cells but traffic unproductively or are rapidly
recycled to the surface. We interpret the weak activity of the αꢀRSVꢀconjugates to reflect
nonꢀspecific uptake of the ADCs at high concentration as has been generally observed in
2
7
other ADC publications. Taken together, the data support the conclusion that targeted
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
delivery has been achieved in CD70ꢀpositive cells with an αꢀhCD70 mAb using these
novel phosphateꢀester linkers, and that cellular activity of these ADC correlates with the
ability of the drugꢀlinkers to be cleaved in a lysosomal environment.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Representative cellular activity of α−hCD70 (1-x) or αRSVꢀADCs (2-x)
carrying phosphate ester linked dexamethasone in 786ꢀO cells.
While active in GILZ assay in 786ꢀO cells, the Emax for most active αꢀhCD70ꢀ
dexamethasone conjugates was only approximately 50% of that produced by free
dexamethasone (data not shown). This may reflect kinetic differences in the delivery of
the small molecule and ADC modalities and/or the lack of effective accumulation of free
dexamethasone inside the cells by ADC due to limited delivery, slow processing or fast
efflux. αꢀhCD70 ADCs with a glucocorticoid payload of higher potency such as
ACS Paragon Plus Environment

Page 27 of 67
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
budesonide and fluticasone propionate are expected to display increasing potency and
higher Emax in cells. To test this hypothesis, select αꢀhCD70ꢀADCs with the rapidꢀ
cleaving pyrophosphate linker and budesonide or fluticasone propionate payload were
made and tested in this same assay. As shown in Figure 7, the α−hCD70ꢀADCs with the
same pyrophosphate linker but with budesonide (1-22) or fluticasone propionate (1-37)
payload displayed much higher potency and Emax than the dexamethasone counterpart
ADC (1-14) in 786ꢀO cells correlated with the enhanced potency of the payload. Again,
the GILZ activity of ADC also correlated with drugꢀlinker cleavage in lysosome, as
ADCꢀ1-33 was completely inactive in GILZ assay in 786ꢀO cells due to the lack of
conversion to free fluticasone propionate in lysosomal lysates (see Scheme 7 and Table
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
). Remarkably, the Emax of ADCꢀ1-37 equaled or exceeded that of free fluticasone
propionate. Similarly, the same drugꢀlinkers were inactive in 786ꢀO cells except at the
highest concentration as the conjugates with αꢀRSV, further supporting the notion that
targeted delivery has been achieved in CD70ꢀpositive cells with an αꢀhCD70 mAb using
the novel pyrophosphateꢀdiester linkers.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 67
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. The cellular activity of α−hCD70 (1-x) or α−RSVꢀADCs (2-x) carrying
phosphate ester linked budesonide or fluticasone propionate in 786ꢀO cells
Potency in 786-0 cells
Compound
Dexamethasone
Budesonide
EC50 (IP)
2.01 nM
0.58 nM
Emax
10.23
12.03
fluticasone
propionate (29)
α-hCD70
0.25 nM
11.49
4.52
7.93
1.01
15.37
5.60
6.87
1
-14
1.05 µg/mL (14.18 nM)*
0.082 µg/mL (1.11 nM)
Inactive
α-hCD70
1-22
α-hCD70
1
-33
α-hCD70
-37
α-RSV
-14
α-RSV
-22
1
0.029 µg/mL (0.39 nM)
12.28 µg/mL (165.78 nM)
28.08 µg/mL (378.08 nM)
2
2
ACS Paragon Plus Environment