Discovery of a Novel Cabazitaxel Nanoparticle-Drug Conjugate (CRLX522) with Improved Pharmacokinetic Properties and Anticancer Effects Using a β-Cyclodextrin-PEG Copolymer Based Delivery Platform

Article abstract of DOI:10.1021/acs.jmedchem.9b00892

Novel nanoparticle-drug conjugates (NDCs) containing diverse, clinically relevant anticancer drug payloads (docetaxel, cabazitaxel, and gemcitabine) were successfully generated and tested in drug discovery studies. The NDCs utilized structurally varied linkers that attached the drug payloads to a β-cyclodextrin-PEG copolymer to form self-assembled nanoparticles. In vitro release studies revealed a diversity of release rates driven by linker structure-activity relationships (SARs). Improved in vivo pharmacokinetics (PK) for the cabazitaxel (CBTX) NDCs with glycinate-containing (1c) and hexanoate-containing linkers (2c) were demonstrated, along with high and sustained tumor levels (>168 h of released drug in tumor tissues). This led to potent efficacy and survival in both taxane- and docetaxel-resistant in vivo anticancer mouse efficacy models. Overall, the CBTX-hexanoate NDC 2c (CRLX522), demonstrated optimal and improved in vivo PK (plasma and tumor) and efficacy profile versus those of the parent drug, and the results support the potential therapeutic use of CRLX522 as a new anticancer agent.

Full text of DOI:10.1021/acs.jmedchem.9b00892

Subscriber access provided by Stockholm University Library
Article
Discovery of a Novel Cabazitaxel Nanoparticle-Drug Conjugate (CRLX522)
with Improved Pharmacokinetic Properties and Anticancer Effects
using a #-Cyclodextrin-PEG Copolymer-Based Delivery Platform
Chester Metcalf, Sonke Svenson, Jungyeon Hwang, Snehlata Tripathi, Geeti Gangal, Sujan
Kabir, Douglas Lazarus, Roderic Cole, Beata Sweryda-Krawiec, Pochi Shum, Donna Brown, Roy
Case, Derek van der Poll, Ellen Rohde, Stephanie Harlfinger, Chi-Hse Teng, and Scott Eliasof
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00892 • Publication Date (Web): 08 Oct 2019
Downloaded from pubs.acs.org on October 13, 2019
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 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 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 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.
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 53
Journal of Medicinal Chemistry
1
1
2
3
4
5
6
7
8
9
Discovery of a Novel Cabazitaxel Nanoparticle-Drug Conjugate (CRLX522) with Improved
Pharmacokinetic Properties and Anticancer Effects using a β-Cyclodextrin-PEG
Copolymer-Based Delivery Platform
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
1
2
2
2
2
Chester A. Metcalf III* , Sonke Svenson , Jungyeon Hwang , Snehlata Tripathi , Geeti Gangal ,
2
2
2
2
2
Sujan Kabir , Douglas Lazarus , Roderic Cole , Beata Sweryda-Krawiec , Pochi Shum , Donna
2
1
2
2
1
1
Brown , Roy I. Case , Derek van der Poll , Ellen Rohde , Stephanie Harlfinger , Chi-Hse Teng ,
Scott Eliasof2
1Novartis Institutes for BioMedical Research Inc., 181 Mass Ave, Cambridge, MA 02139 USA
2Cerulean Pharma Inc., 35 Gatehouse Drive, Waltham, MA 02451 USA
Corresponding author:
*
Chester A. Metcalf III, Novartis Institutes for BioMedical Research Inc., 181 Mass Ave,
Cambridge, MA 02139 USA; 617-871-7674; chet.metcalf@novartis.com
Abstract
Novel nanoparticle-drug conjugates (NDCs) containing diverse, clinically-relevant anticancer drug
payloads (docetaxel, cabazitaxel, and gemcitabine) were successfully generated and tested in drug
discovery studies. The NDCs utilized structurally varied linkers that attached the drug payloads to a β-
cyclodextrin-PEG copolymer to form self-assembled nanoparticles. In vitro release studies revealed a
diversity of release rates driven by linker structure-activity relationship (SAR). Improved in vivo
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 53
2
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
pharmacokinetics (PK) for the cabazitaxel (CBTX) NDCs with glycinate- (1c) and hexanoate-containing
linkers (2c) were demonstrated, along with high and sustained tumor levels (>168 hrs of released drug in
tumor tissues). This led to potent efficacy and survival in both taxane- and docetaxel-resistant in vivo
anticancer mouse efficacy models. Overall, the CBTX-hexanoate NDC 2c (CRLX522) demonstrated an
optimal and improved in vivo PK (plasma and tumor) and efficacy profile vs. parent drug, and the results
support the potential therapeutic use of CRLX522 as a new anticancer agent.
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
. Introduction
Cancer therapy still remains one of the most challenging treatment areas despite the efforts and progress
achieved over the past decade.1, 2 Poor survival outcomes of current treatments, especially against multi-
drug resistant tumors, call for improved therapies with innovative approaches to drug delivery. The
treatment of solid tumors with targeted (e.g., tyrosine kinase inhibitors) and cytotoxic drugs face several
challenges: (1) Poor drug solubility due to the highly lipophilic nature of many small molecules and/or poor
membrane permeability; (2) Poor pharmacokinetic (PK) and absorption, distribution, metabolism, and
excretion (ADME) properties of small molecule drugs caused by metabolizable functionality resulting in
high clearance rates; (3) Reduced efficacy due to a lack of sustained drug concentrations in tumor tissue as
well as non-specific binding to cells and tissues, expressed by large volumes of distribution (V ); and (4)
D
Toxicity caused by adverse drug effects on healthy tissue. The encapsulation of drug molecules into
nanosized carriers has become a powerful approach to address these challenges by increasing solubility,
enhancing apparent drug half-life, and targeting tumors and sites of inflammation (vs. healthy tissue) via
the enhanced permeability and retention (EPR) effect (i.e., accumulation of nanosized molecules and
delivery systems selectively to tumor and inflamed tissue via leaky neovasculature coupled with reduced
lymphatic drainage).3-10 The most often studied nanocarriers include liposomes, micelles, and polymeric
nanoparticles. A sizable number of nanomedicines based on the three carriers groups have translated into
the clinic, and some have advanced to marketed products and are benefiting patients.11-15 Liposomes and
ACS Paragon Plus Environment

Page 3 of 53
Journal of Medicinal Chemistry
3
1
2
3
4
5
6
7
8
9
micelles, albeit important contributors to the field of nanotechnology and nanomedicine, have been
extensively described elsewhere16-25 and therefore the remainder of this section will focus on the properties
and characterization of polymeric nanoparticles.
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
Polymeric nanoparticles are mainly prepared through the use of emulsions or precipitation processes
and can involve either physical entrapment of the drug or chemical conjugation (i.e., covalent attachment)
between drug and the constituent polymers.26-30 Nanoparticle sizes depend on the preparation method,
ranging from ten to several hundred nm. Nanoparticles requiring a processing step allow engineering of
polymeric nanoparticles (size, shape, surface composition, and drug loading efficiency) to be tailored to the
respective application needs, but adds to the complexity of manufacturing, particularly at commercial
scale.31-37 Drug encapsulation into polymeric nanoparticles not only improves aqueous solubility, but has
the potential for an improved PK profile of the nanoparticle drug (via slow, sustained release) compared to
dosing of the drug alone. Such improvements in PK are manifested in increased half-life and AUC in
tumors, less frequent dosing (every 2-3 weeks in some cases) and reduced biodistribution into healthy
tissue, which can result in lower toxicity.38-41 In addition, polymeric nanoparticles, with sizes above 10 nm
but preferably below 100 nm, can take advantage of selective tumor accumulation via the EPR effect (in
addition to low V ) to improve target specificity of the treatment, and consequently, the efficacy of the
D
respective drug. 42,43 This combination of reduced toxicity and enhanced efficacy has the potential to
improve the therapeutic index.14, 44
Given the two methods of physical entrapment and chemical conjugation to encapsulate drugs into
polymeric nanoparticles, most approaches rely on physical entrapment of the drug. There are two major
challenges with this approach: (1) The hydrophilic or hydrophobic (in most cases) nature of the drug has to
match that of the polymeric carrier to achieve sufficiently high drug loading; and (2) Release of physically
entrapped drugs generally shows bi-phasic behavior where an initial burst release of the drug is followed
by first-order sustained release.45-47 This bi-phasic behavior is a more than likely a consequence of drug
molecules on or close to the polymeric nanosurface that are released instantaneously under the sink
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 53
4
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
condition of the bloodstream, along with other complexities of the drug-loaded nanoparticle system.
Reducing burst release and improving entrapment efficiency to provide sustained release of drug and
potentially greater efficacy often require more complex formulation approaches.48, 49 Chemical drug
conjugation on the other hand has several advantages over physical entrapment: (1) The initial burst release
of drugs can be greatly reduced or avoided; (2) Conjugation can protect small molecule drugs from being
metabolized until after the payload is released from the nanoparticle carrier, leading to sustained release of
intact drug; (3) Linker molecules between drug and carrier can be designed to release drugs with varying
and differentiating in vivo kinetics in response to a trigger event (e.g., changes in pH, temperature, or
enzyme levels), allowing much greater control over drug release compared to diffusion and matrix
degradation; and (4) Conjugation of drugs to the constituent polymers allows nanoparticle formation
regardless of the drugs’ physicochemical characteristics (e.g., hydrophobicity).50-54 Drug encapsulation
through chemical conjugation linker strategies to form nanoparticle-drug conjugates (NDCs), using a β-
cyclodextrin-PEG (CDP) copolymer, has been the focus of our current efforts (vide infra) to deliver
therapeutic levels of released small molecule drug payloads to the tumor microenvironment, reduce
systemic drug levels (i.e., increase therapeutic indexes - TIs), and provide effective treatments for solid
tumor indications.39-41 The first clinical compound (CRLX101) to evolve from the NDC platform advanced
to Phase 2 clinical development.40, 55 CRLX101 contains camptothecin (CPT) as the payload and glycine
as the linker molecule. The second clinical compound CRLX301 (compound 5c), containing docetaxel as
the payload and glycine as the linker molecule, completed Phase 1 clinical development.
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
Herein we present results that demonstrate the ability to modulate PK parameters and provide sustained
and controlled release kinetics for novel NDCs containing the active pharmaceutical ingredient (API) drug
payloads docetaxel (DTX), cabazitaxel (CBTX), or gemcitabine (GEM) using a proprietary linker
technology platform. From a PK perspective, we are aiming to increase the apparent half-life of the drug as
well as reduce plasma Cmax by making the release rate of the drug payload from the NDC the rate-limiting
step. The polymeric backbone of the nanoparticles employed in this study utilizes the CDP copolymer
ACS Paragon Plus Environment

Page 5 of 53
Journal of Medicinal Chemistry
5
1
2
3
4
5
6
7
8
9
technology. A series of in vivo PK and efficacy studies were conducted in tumor-bearing mice to establish
in vitro to in vivo structure-activity relationship (SAR) correlations. This resulted in the discovery of CBTX-
hexanoate NDC 2c (CRLX522), which has potential therapeutic use as a new anticancer agent.
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
2
. Results and Discussion
2
.1. Synthesis and Formation of Nanoparticle Drug Conjugates (NDCs)
The syntheses of CBTX, DTX, and GEM NDCs are shown in Schemes 1-3. Drug loading for CBTX
1
c-4c and DTX 5c-8c NDCs was high (71-100%) and approaching or at theoretical values, whereas loading
for the GEM NDCs 11c-13c was at about half the theoretical values (Table 1); such lower loaded NDCs,
which leave uncapped negatively charged carboxylic acid groups that have relatively fewer number of
stabilizing cyclodextrin-API inclusion complexes within the NDC, have been shown previously to result in
4
1
higher AUC of in vivo released drug. Nonetheless, all NDCs self-assembled in water to form nanoparticles
between 17-28 nm in average size (Zavg) as measured by dynamic light scattering (DLS). An exception
were the nanoparticles formed by the CBTX-disulfide-carbonate NDC 4c, which gave a Zavg value of 194
nm. This larger particle size, which was only observed for the CBTX-disulfide NDC 4c, could be an
indication of crosslinking between copolymer strands caused by the reversible nature of the disulfide bond.
Similar issues were observed with the GEM disufide linker and attempts to synthesize the NDC in high
purity were unsuccessful. This led to the conclusion that GEM was not a good synthetic candidate for the
disulfide linker conjugation. Due to crosslinking potentially complicating interpretation of release kinetics
as well as the larger particle size of the CBTX-disulfide NDC not allowing direct comparison to the other
smaller nanoparticles, NDC 4c was not tested in subsequent studies. Unexpectedly, these issues were not
observed for the closely related taxane NDC, DTX disulfide analogue 8c. The polydispersity indices (PDIs)
of all self-organized structures were sufficiently narrow with values between 0.2 and 0.5, and the aqueous
NDC solutions all had high drug purities (>95% for all tested NDCs) and high drug concentrations (1.2-4.5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 53
6
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
mg/mL). We have previously reported successful generation of cryo-TEM images of NDC nanoparticles
e.g., CRLX101).41
(
Table 1. Characterization, concentration, loading, and purity data for CBTX, DTX, and GEM NDCs.
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
Theoretical
Drug
Actual Drug
NDC
Drug
Purityb
(%)
99
NDC Drug
Zavga
Loadingb
b
Drug-linker NDCs
PDI
Conc (aq.,
(nm)
Loading
(weight %)/%
mg/mL)
(weight %) of Theoretical
CBTX-glycinate NDC (1c)
CBTX-hexanoate NDC (2c)
24.8
27.8
22.7
194
0.231
0.249
0.251
0.313
0.286
0.367
0.214
0.487
0.223
0.269
0.194
3.4
3.1
3.3
4.3
3.5
1.7
4.5
3.6
1.2
2.1
3.4
25.4
25.0
25.3
24.5
24.8
24.3
24.2
23.9
8.29
8.13
8.08
20.4/80
17.7/71
19.1/75
19.3/79
22.4/90
16.9/70
18.0/74
23.9/100
3.8/46
99
CBTX-alaninate NDC (3c)
100
91
CBTX-disulfide-carbonate NDC (4c)
DTX-glycinate NDC (5c)
26.1
18.4
24.7
26.4
27.4
17.0
18.5
98
DTX-hexanoate NDC (6c)
97
DTX-β-alanine-glycolate NDC (7c)
DTX-disulfide-carbonate NDC (8c)
GEM-glycinate NDC (11c)
95
98
100
100
98
GEM-hexanoate NDC (12c)
GEM-β-alanine-glycolate NDC (13c)
4.3/53
4.2/52
a
b
Size measured by dynamic light scattering (DLS). Drug concentration, loading and purity measured by reverse
phase (RP) high-pressure liquid chromatography (HPLC).
ACS Paragon Plus Environment

Page 7 of 53
Journal of Medicinal Chemistry
7
1
2
3
4
5
6
7
8
9
Scheme 1. Synthesis of cabazitaxel-linkers
O
1
NHCbz
A
HO
a, b
O
O
O
O
O
O
2
NHCbz a, b
O
O
1a: R = CH NHCbz
2
HO
HO
1
b: R = CH NH
2 2
O
O
2a: R = (CH ) NHCbz
2b: R = (CH ) NH
2 5 2
3a: R = CHCH NHCbz
O
2 5
O
NHCbz
a, b
O
HO
O
HO
H
3
N
O
H
O
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
N
O
O
O
O
O
O
O
O
O
H
H
3
O
HO
3b: R = CHCH NH
O
3
2
O
CBTX
4
S
S
NHMMt
c, d
4a:
4
R = O(CH2)2SS(CH2)2NHMMT
b: R = O(CH ) SS(CH ) NH
R
HO
2
2
2 2
2
Reaction conditions: (a) N-Cbz amino acid (1-3), EDC, DMAP; (b) H , Pd/C, MsOH; (c) 4, N-hydroxysuccinimide carbonate, DMAP; (d) Dichloroacetic acid,
2
anisole.
OH
O
B
OH O
O
OH
OH O
OH
OH
OH
OH
OH
O
O
HO
O
OH
O
O
O
O
NH
O
S
OH
OH
OH
S
NH
OH
O
N
n
H
m
O
OH
O
O
O
OH
O
OH
OH
OH
O
n  77, m  14
O
O
OH
OH
1
b-4b, e
OH
O
O
O
OH O
1c: R = HN
2c: R = HN
O
OH O
OH
OCBTX
OCBTX
OH
OH
O
OH
O
OH
OH
O
OCBTX
O
R
O
OH
O
O
O
O
NH
O
S
OH 3c: R = HN
OH
OH
S
NH
OH
O
N
n
H
m
O
OH
O
O
O
OH
O
R
OH
4c:
R = HN
S
S
OH
O
O
OCBTX
O
O
OH
OH
Reaction conditions: (e) EDC, NHS, (1b-4b).
Scheme 1. Syntheses of CBTX NDCs 1c-4c. (A) Reaction of CBTX with linkers. (B) Conjugation of CBTX-
linker molecules to the CDP copolymer to give the corresponding CBTX NDCs following
purification/processing.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 53
8
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
Scheme 2. Synthesis of docetaxel-linkers
O
A
5
6
NHCbz
HO
a, b
O
O
HO
HO
HO
HO
O
O
NHCbz a, b
OH
OH
5a: R = CH NHCbz
2
HO
HO
5
b: R = CH NH
2 2
O
O
6a: R = (CH ) NHCbz
6b: R = (CH ) NH
O
2 5
O
O
S
NHCbz
a, b
7
O
N
H
2 5
2
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
O
N
H
O
O
O
O
O
O
O
O
O
O
O
H
7a: R = CH OCO(CH ) NHCbz
H
O
S
2 2 2
O
HO
7
8a:
8
b: R = CH OCO(CH ) NH
O
2 2 2 2
O
DTX
4
NHMMt
c, d
R = O(CH2)2SS(CH2)2NHCbz
R
HO
b: R = O(CH ) SS(CH ) NH
2 2 2 2 2
Reaction conditions: (a) N-Cbz amino acid (5-7), EDC, DMAP; (b) H , Pd/C, MsOH; (c) 4, N-hydroxysuccinimide carbonate, DMAP; (d) Dichloroacetic
2
acid, anisole.
OH
O
OH O
B
O
OH OHOH O
OH
OH
OH
OH
O
O
HO
O
OH
O
O
O
O
NH
O
S
OH
OH
OH
S
NH
OH
O
N
n
H
m
O
OH
O
O
O
OH
O
OH
OH
OH
O
O
n  77, m  14
O
OH
OH
5
b-8b, e
OH
O
O
OH O
5c: R = HN
6c: R = HN
O
OH OHOH O
OH
ODTX
OH
O
O
O
OH
OH
O
ODTX
R
O
OH
O
O
O
O
NH
O
S
OH 7c: R = HN
O
O
OH
OH
S
NH
OH
O
N
ODTX
O
n
H
m
O
OH
O
O
O
OH
O
R
OH
8c:
R = HN
S
S
OH
O
O
O
ODTX
O
OH
OH
Reaction conditions: (e) EDC, NHS, (5b-8b).
Scheme 2. Synthesis of DTX NDCs 5c-8c. (A) Reaction of DTX with linkers. (B) Conjugation of DTX-linkers
to CDP copolymer to give the corresponding DTX NDCs following purification/processing.
ACS Paragon Plus Environment

Page 9 of 53
Journal of Medicinal Chemistry
9
1
2
3
4
5
6
7
8
9
Scheme 3. Synthesis of gemcitabine-linkers
A
O
1
1
2
NHBoc
N
NHBoc
c, d
NHR1
N
NH2
N
NH2
HO
HO
1
1a: R1 = Boc, R2 = CH NHBoc
2
O
O
N
11b: R1 = H, R2 = CH NH
1
2 2
O
O
R2
2a: R1 = Boc, R2 = (CH ) NHBoc
1
NHBoc c, d
2 5
HO
N
O
a
HO
N
O
b
HO
N
O
N
O
O
O
O
O
12b: R1 = H, R2 = (CH ) NH
2 5 2
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
13a: R1 = Boc, R2 = CH OCO(CH ) NHBoc
2 2 2
3b: R1 = H, R2 = CH OCO(CH ) NH
F
F
F
F
F
1
OBocF
9
OBoc 10
F
O
NHBoc
c, d
OR1
F
2
2 2
2
GEM
OH
13 HO
O
Reaction conditions: (a) Boc2O, Na2CO3, aq. dioxane (b) Boc2O, anhyd. dioxane (c) N-Boc amino acid (11-13), EDC, DMAP; (d) TFA, DCM.
B
OH
O
OH O
O
OH OHOH O
OH
OH
OH
OH
O
O
HO
O
OH
O
O
O
O
NH
O
S
OH
OH
OH
S
O
NH
OH
N
n
H
m
O
OH
O
O
O
OH
OH
O
OH
OH
O
O
n  77, m  14
O
OH
OH
11b-13b, e
OH
O
OH O
O
O
OH OHOH O
OH
OH
O
11c: R = HN
12c: R = HN
OGEM
OH
OH
O
O
R
O
OH
O
O
O
O
OGEM
OGEM
NH
O
S
OH
OH
OH
S
O
O
NH
OH
OH
N
n
H
13c: R =
m
HN
O
O
O
O
O
OH
OH
O
R
OH
O
O
O
O
OH
OH
Reaction conditions: (e) EDC, NHS, (11b-13b).
Scheme 3. Syntheses of GEM NDCs 11c-13c. (A) Reaction of GEM with linkers. (B) Conjugation of GEM-
linkers to the CDP copolymer to give the corresponding GEM NDCs following purification/processing.
2
.2. In vitro Release of Docetaxel, Cabazitaxel, and Gemcitabine NDCs
The in vitro rates of drug release from DTX-, CBTX- and GEM-containing NDCs using a variety of
different linkers under near physiological conditions (i.e., phosphate-buffered saline (PBS) at pH 7.4 and
o
3
7 C) were investigated to demonstrate the influence of linker functionality on drug release kinetics. The
three drugs were conjugated to 5 possible linker molecules (glycinate, alaninate, hexanoate, diester β-
alanine-glycolate, and disulfide-carbonate) with varying degrees of hydrophobicity, length, and ester
functionality/sterics/electronics. Matched sets of linker NDCs (i.e., only API drug payload varied) were
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 53
1
0
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
generated for the purpose of comparing linker release trends between structurally similar (CBTX, DTX)
and dissimilar (GEM) drugs. Glycinate, alaninate, and hexanoate linkers were chosen to determine the
effects of varied linker hydrophobicity, sterics and electronics (e.g., α-amino acid vs. aliphatic ester) on
release kinetics, and all utilized hydrolysis sensitive ester bonds to attach their respective drugs. In addition,
two linkers with dual sites for drug release were studied: a β-alanine-glycolate containing-linker with two
hydrolysable ester bonds, and a disulfide-carbonate linker with potential for additional intracellular
glutathione (GSH) triggered release.56
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
The in vitro release profiles of DTX NDCs 5c-8c with the various linker molecules are shown in Fig.
1. The bis ester-containing β-alanine-glycolate (β-Ala-glyc) linker showed the fastest release, potentially
due to the dual release mechanism and/or increased hydrolytic lability of the API-linked ester bond (driven
by adjacent electron-withdrawing group), with ~90% of DTX released by 24 hr. The glycinate linker
released 50-60% DTX within that time period, while the longer and more hydrophobic hexanoate linker
and the disulfide-carbonate linker both released less than 15% DTX by 72 hrs. The slower release for the
6
c and 8c linkers can be rationalized by the more hydrophobic and/or presumed less hydrolytically labile
ester linker functionality, leading to reduced chemical hydrolysis in the hydrophobic nanoparticle core and
a more stabilized NDC. The observed plateau around 60% release of DTX from the glycinate NDC is due
to the formation of 7-epidocetaxel (data not shown), and appears to be specific to the DTX payload-
glycinate linker combination.
1
00
Glycinate NDC
Hexanoate NDC
Disulfide NDC
-Ala-glyc NDC
8
0
0
6
40
2
0
0
0
20
40
60
ACS Paragon Plus Environment

Page 11 of 53
Journal of Medicinal Chemistry
1
1
1
2
3
4
5
6
7
8
9
o
Figure 1. In vitro release of DTX from DTX NDCs 5c-8c in PBS at pH 7.4 and 37 C.
The release profile trends for the CBTX NDCs 1c-4c were similar to DTX NDCs, with the faster release
rate observed for the glycinate linker (>90% by 72 hr) and slower release measured for the hexanoate linker
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
(
<5% by 72 hrs) as shown in Fig. 2A. Both CBTX NDC linkers show a large difference in release % at the
initial time points as well (<1 vs. ~20%). The alaninate and glycinate linkers showed similar release rates,
which suggest that their release is driven by similar electronic factors associated with the common α-amino
acid ester linker functionality vs. hydrophobicity and/or steric effects. The glycinate and hexanoate linkers
(fast and slow release, respectively) were chosen to examine release kinetics in the presence of plasma
enzymes to determine the extent by which they contribute to the mechanism of drug release from the NDCs.
The CBTX-glycinate 1c and CBTX-hexanoate 2c NDCs were mixed with untreated mouse plasma as well
as plasma that had been heat-treated to inactivate enzymes. The glycinate conjugate showed somewhat
faster release in mouse plasma than PBS (100 vs. ~70% by 24 hr), but no release rate differences were
observed for untreated vs. heat-inactivated plasma (Fig. 2B). The same was true for the hexanoate linker,
in which the release rates in PBS, active plasma, and heat-inactivated plasma were essentially identical
(Fig. 2C). The faster release of the glycinate linker in plasma vs. PBS may be related to the presence of
4
1
lipoprotein complexes that can assist in the disassembly of the NDC. The more stabilized, hydrophobic
hexanoate linker may be less sensitive to these effects leading to the similar release rates in plasma and
PBS. Taken together, the in vitro release data suggest that plasma enzymes should not play a major role in
drug release in vivo, and it is likely to be a pH-driven release mechanism. This is consistent with previous
data for NDCs with ester-containing linkers (e.g., established pH range dependency on hydrolysis rate, IC50
3
9
shifts in MTT assays [nontoxic] potentially influenced by cell/endosomal pH effects on release rate). A
pH-driven mechanism would be predicted to have an advantage over enzyme-mediated cleavage from a PK
translatability/variability standpoint, since there is little dependence on individual factors affecting release
(e.g., species-to-species and patient-to-patient plasma/tumor enzyme level differences). In support of this,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 53
1
2
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
previous NDCs (e.g., CRLX101) have translated well from animals to humans and demonstrated low
patient-to-patient PK variability in the clinic.41
A
100
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
Glycinate NDC
Hexanoate NDC
Alaninate NDC
80
60
40
20
0
0
20
40
60
Time (hr)
B
C
1
00
80
60
40
120
00
1
8
0
60
PBS
4
2
0
0
0
Mouse plasma
Heat-inactivated plasma
2
0
0
0
20
40
Time (hr)
60
0
20
40
60
Time (hr)
o
Figure 2. (A) In vitro release of CBTX from CBTX NDCs 1c-3c in PBS at pH 7.4 and 37 C. (B) CBTX-glycinate
1
c and (C) CBTX-hexanoate 2c NDCs treated with ex vitro mouse plasma and heat-inactivated mouse plasma
o
compared to PBS at 37 C.
The release trends for the GEM NDCs 11c-13c showed some similarities to the DTX and CBTX
nanoparticle conjugates (Fig. 3). The β-alanine-glycolate with its two ester groups released drug the fastest
~90% by 24 hr), followed by the glycinate monoester (~60% by 24 hr) and the established slow releasing
(
hexanoate linker (10-15% by 72 hr). It is possible that the faster releasing β-alanine-glycolate and glycinate
linkers release drug too fast under the in vitro conditions to be differentiated. The collective release data
ACS Paragon Plus Environment

Page 13 of 53
Journal of Medicinal Chemistry
1
3
1
2
3
4
5
6
7
8
9
demonstrate that a range of API payload MW (263-836), lipophilicity (cLogP = -1.5 to 4), and aqueous
.
solubility (1 µM to 100 mM; CBTX and GEM HCL, respectively) can be accommodated and successfully
conjugated into the NDCs while providing sustained and tunable release kinetics. It is conceivable that
linker functionality as well as API local structural environment (e.g., primary vs. secondary hydroxyl
groups) may contribute to overall in vitro drug release kinetics. Given the chemical simplicity and the
differentiated release profiles for the glycinate and hexanoate linkers, both were selected for further in vivo
PK studies to establish in vitro to in vivo SAR correlations for the corresponding NDCs.
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
Glycinate NDC
Hexanoate NDC
1
20
00
80
1

-Ala-glyc NDC
6
0
0
4
20
0
0
20
40
60
Time (hr)
o
Figure 3. In vitro release of GEM from GEM NDCs 11c-13c in PBS at pH 7.4 and 37 C.
2
.3. PK Studies of Cabazitaxel and Gemcitabine NDCs with Glycinate and Hexanoate Linkers
To demonstrate in vivo profiles for NDCs with varying release rates, CBTX and GEM payloads were
selected for the PK studies. As discussed in the previous section, the two linkers chosen for each API
payload were the glycinate linker for fast release and the hexanoate linker for slow release. The
corresponding NDC linker pairs for CBTX and GEM were similar in size (hexanoate - 28/glycinate - 25
nm and hexanoate - 17/glycinate - 27 nm, respectively).
5
7
In the first study, male C57BL/6 mice with syngeneic B16.F10 tumors were given a single intravenous
(
iv) doses of CBTX and the respective CBTX NDCs at an equivalent dose of 13.5 mg/kg of CBTX,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 53
1
4
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
providing dose-matched PK and allowing direct comparison of the NDC and parent drug PK data. Animals
were sacrificed at different time-points, and plasma and tumor tissue samples were collected for the analysis
of total (conjugated + released) drug and released drug levels in each by liquid chromatography with tandem
mass spectrometry (LC-MS/MS). The innate ability to hydrolyze the covalent bonds between API and
polymer backbone allows for the bioanalytical characterization of the total amount of payload present as
well as the quantitation of the released payload from aliquots of the same sample. In plasma, total drug
levels approximate conjugated drug levels (calculated by subtracting measured released drug levels from
total) due to the much lower released drug levels compared to total (~1% or less of total AUC). The
capability to accurately and reliably measure both total and released drug is a key distinguishing feature of
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
the NDC technology, something that is important in a clinical setting and something that is very challenging
to accomplish with entrapped polymeric nanoparticles (non-conjugated drug) or liposomes.41
Table 2. Plasma PK parameters after a single iv dose of 13.5 mg/kg CBTX, CBTX-glycinate NDC 1c or
CBTX-hexanoate NDC 2c was given to male C57BL/6 mice with B16.F10 syngeneic tumors.
C or
0
Fb
AUC∞
tmax
t1/2
VDa
Cl
Test Article
Cmax
[µM*h]
[hr]
[mL/min/kg]
[h]
[L/kg]
[%]
[µM]
CBTX
8.4
1713
20
10
254
2.7
-
-
32
0.16
-
48
23
84
0.11
-
-
-
Total
Released
Total
CBTX-glycinate
NDC 1c
0.5
-
14
240
-
7700
15
382
0.22
0.035
-
31
0.09
-
CBTX-hexanoate
NDC 2c
Released
0.08
180
178
a
b
V
D
is calculated as Vss (estimated volume of distribution at steady state) for nanoparticles dosed iv. F = availability of
administered CBTX dose in systemic circulation.
ACS Paragon Plus Environment

Page 15 of 53
Journal of Medicinal Chemistry
1
5
1
2
3
4
5
6
7
8
9
Table 3 Tumor PK parameters after a single iv dose of 13.5 mg/kg CBTX, CBTX-glycinate NDC 1c or
CBTX-hexanoate NDC 2c was given to male C57BL/6 mice with B16.F10 syngeneic tumors.
Tumor-to-
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
AUC∞
Cmax
tmax
t1/2
Test Article
Plasma AUC
Ratio
[µM*h]
[µM]
[hr]
[h]
CBTX
177
3.7
24
6
2
31
64
21:1
Total
1375
0.8:1
CBTX-glycinate
NDC 1c
Released
Total
407
5454
606
6.3
72
6
44
59
20:1
0.7:1
40:1
CBTX-hexanoate
24
24
NDC 2c
Released
1.4
342
The levels of total drug in plasma for both the CBTX-glycinate 1c and CBTX-hexanoate 2c NDCs were
substantially higher than the plasma levels of separately dosed CBTX parent drug at all time points over
the 168-hour time period studied (Fig. 4A). The total drug AUCs in plasma were ~200-900 fold higher than
that of parent drug (Table 2), and large reductions in steady state volume of distribution for the NDCs
(
~800-900x) compared to parent were also demonstrated. Similar AUC increases for the NDCs were
observed for total drug in tumor tissues (Fig. 4C, Table 3). The total drug AUCs for both the CBTX-
glycinate 1c and CBTX-hexanoate 2c NDCs in tumor were ~8-30 fold higher, respectively, than parent
drug and remained high over the course of the study. The phenomena are a direct consequence of the very
low clearance (0.035-0.16 vs. 32 mL/min/kg for parent) of the NDCs providing long and high drug level
blood circulation times, which have been shown previously to lead to NDC accumulation in and penetration
4
1
through the tumor. In both plasma and tumor tissue, the CBTX-hexanoate nanoparticles provided higher
stability (i.e., slower release) and slower decline of total drug levels compared to the CBTX-glycinate
nanoparticles, as demonstrated by ~4-fold higher AUC and lower Cl. This ~4-fold difference in AUC/Cl of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 53
1
6
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
total drug between the glycinate and hexanoate linkers, driven by the differentiated release rates of the
linkers, provides predictable in vitro to in vivo correlation (IVIVC) on a “rank-order” level, and supports
the notion that release rates are tunable based on linker design.
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
A
B
1000
CBTX
1000
100
10
CBTX-glycinate NDC
CBTX-hexanoate NDC
100
10
1
1
0.1
0.1
0.01
0.01
0
50
100
Time (hr)
150
0
50
100
Time (hr)
150
D
C
1
000
1000
1
00
10
1
100
10
1
0.1
0.1
0.01
0.01
0
50
100
Time (hr)
150
0
50
100
Time (hr)
150
Figure 4. In vivo plasma and tumor exposure profiles after separate doses of CBTX, CBTX-glycinate NDC 1c
and CBTX-hexanoate NDC 2c of 13.5mg/kg each administered iv to male C57BL6/6 mice with syngeneic
B16.F10 tumors. (A) Graph for total CBTX in plasma; (B) Graph for released CBTX in plasma; (C) Graph for
total CBTX in tumor tissue; (D) Graph for released CBTX in tumor tissue.
ACS Paragon Plus Environment

Page 17 of 53
Journal of Medicinal Chemistry
1
7
1
2
3
4
5
6
7
8
9
The released drug levels for the CBTX-glycinate 1c and CBTX-hexanoate 2c NDCs further illustrate
the contribution that the linker plays in affecting PK, particularly in plasma (Fig. 4B, Table 2). The
differences in released drug AUCs vs. parent in plasma are not significant (~2-fold; F ~100%), supportive
that Cl of the API is not affected by CBTX NDC administration. However, the PK profiles (e.g., Cmax, t1/2)
are quite unique and differentiated between the CBTX-glycinate 1c and -hexanoate 2c NDCs as well as
separately dosed CBTX parent drug, highlighting the specificity of the drug-linker combination. This
differentiated PK is a direct result of the interplay between the varying, rate-limiting in vitro release rates
for each NDC and in vivo PK factors (clearance, plasma protein binding, distribution, etc.), all in dynamic
equilibrium, which can have profound effects on the in vivo levels of released drug from each NDC; total
drug levels are much less influenced by such PK factors as shown previously.41 Initially, starting
concentrations of CBTX in plasma released from the CBTX-glycinate NDC 1c were ~4-fold lower than for
CBTX parent drug, but subsequent time points indicated higher levels of released drug over the course of
the study. The initial plasma levels for CBTX released from the CBTX-hexanoate NDC 2c were ~50-fold
lower than parent drug and over 10-fold less than that for CBTX-glycinate NDC 1c. Although the calculated
terminal half-life for released drug from CBTX-hexanoate NDC 2c is longer than the observation period, it
still allows the tentative interpretation that terminal half-life compared to parent drug is increased (~ 4-
fold).
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
PK modulations for the hexanoate linker result in a very flat and sustained released drug PK profile in
plasma. The lower Cmax for both NDCs provides the potential for reduced Cmax-related toxicities compared
to parent drug, while providing sustained levels of released drug over the 168 hour dosing period. The
availability of CBTX in systemic circulation after NDC dosing is high (F = ~100-200%), which indicates
total conversion assuming Cl of the API is not impaired. By avoiding the initial high Cmax, optimal NDC
selection according to PK/linker release profile is possible to stay below toxic and above efficacious
concentrations. The lower released drug plasma PK levels (e.g., Cmax) for the CBTX-hexanoate NDC 2c
compared to the glycinate NDC 1c, again controlled by linker design, indicate that the hexanoate NDC
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 53
1
8
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
releases drug slower in vivo than the glycinate conjugate, and therefore the intact CBTX-hexanoate
nanoparticles stay in systemic circulation longer. Longer NDC blood circulation times have the advantage
of providing more total drug to the tumor microenvironment, creating a local tumor depot from which
sustained levels of drug can be released.
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
In tumor tissue, the high and sustained levels of released CBTX from both the CBTX-glycinate 1c and
CBTX-hexanoate 2c NDCs highlight the strengths of the NDC drug delivery platform to provide long-
lasting, efficacious levels of drug to the tumor (Fig. 4D, Table 3). The tumor levels of released CBTX from
the CBTX-glycinate NDC 1c provided an exposure profile that was consistently above but similarly shaped
to that of CBTX. Released drug tumor levels for the CBTX-hexanoate NDC 2c were initially low, however,
over the course of the 168-hr study the levels of released CBTX from the NDC surpassed those for parent
drug (3-4x AUC). Increases in released drug tumor half-life for the CBTX-hexanoate NDC 2c were realized
over both parent drug (>10x) and glycinate NDC 1c (~8x). Both the CBTX-hexanoate 2c and CBTX-
glycinate 1c NDCs provided high released drug tumor levels as a percentage of total tumor AUC levels
(
11% and 30%, respectively). The CBTX-hexanoate NDC 2c also had high (relative to parent) percent
injected dose (ID) delivered into the tumor (1.1%ID), which exceeds the average of 0.7% reported for
nanoparticle tumor delivery efficiencies in a survey of 232 data sets.58 In contrast, only 0.04% of the
administered amount of CBTX parent drug reached the tumor. CBTX-hexanoate NDC 2c also demonstrated
an impressive released drug tumor-to-plasma AUC ratio of 40:1, demonstrating differentiation and
5
9
advantages for the NDC technology over other delivery technologies. Based on overall differentiated and
improved PK profiles for the CBTX NDCs, we chose to further explore both in efficacy models as
representative of the NDC technology.
The exposure profiles of the GEM-glycinate 11c and GEM-hexanoate 12c NDCs were determined in a
similar fashion to the previous study by measuring the plasma concentrations of total and released GEM.
For the GEM NDCs, our focus was to demonstrate the ability to modulate PK parameters vs. parent drug.
Therefore, studies were performed in non-tumor bearing mice, which would also allow comparisons
ACS Paragon Plus Environment

Page 19 of 53
Journal of Medicinal Chemistry
1
9
1
2
3
4
5
6
7
8
9
between the CBTX and GEM NDCs given the expectation that plasma PK would be similar in tumor- and
non-tumor bearing mice. Non-tumor bearing male C57BL/6 mice were given a single iv dose of the
respective GEM NDCs at an equivalent dose of 10 mg/kg GEM, and plasma levels of total and released
GEM were measured at several time-points as indicated in Figure 5. The GEM parent drug cleared rapidly
from plasma by 3 orders of magnitude within about 12 hours, which is consistent with known extensive
deamination of the drug yielding an inactive metabolite. The clinical dosing regimen for the current
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
2
marketed product recommends high dose weekly administration of 1000 to 1250 mg/m by 30-min iv
infusion to achieve therapeutic drug levels and to maximize formation of the active metabolite gemcitabine
triphosphate.60 Encapsulation of GEM into the NDC could shield it from rapid deamination, and slow
release of GEM from the NDC form the basis for low continuous exposure levels of the cytotoxic agent.
Thus, GEM NDCs may allow for an alternative, improved clinical dosing regimen with fewer toxicities in
the clinic.
Except for the initial time-points for released drug, the levels of total and released GEM from both
conjugates were higher than GEM parent drug and remained high over 72 hrs (Fig. 5). Both GEM NDCs
demonstrated much reduced V values (0.02-0.05 L/kg) and clearance (0.037-0.087 mL/min/kg) for total
D
drug compared to parent (1.5 L/kg and 55 mL/min/kg, respectively), in addition to providing increases in
total drug plasma half-life (8-12x) as indicated in Table 4. Like the CBTX NDCs, the lower Cmax of released
drug in plasma for the GEM NDCs (~5-7x lower) compared to separately dosed parent drug may potentially
reduce Cmax-driven toxicities. Released drug plasma half-life was extended by ~10-14 fold vs. parent.
Interestingly, although the in vivo total drug levels for the GEM-hexanoate NDC 12c were slightly higher
and may indicate greater stability (i.e., slower release of drug) compared to the glycinate NDC 11c, the
released drug PK curves (including Cmax) were quite similar for the respective GEM NDCs. This in contrast
to the more differentiated total and released drug PK data comparisons between the CBTX-hexanoate and
CBTX-glycinate NDCs, which correlated well with in vitro release data. This lack of linker-to-linker SAR
transferability (and IVIVC) between API drug payloads may be due to a “customization” of NDC in vivo
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 53
2
0
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 profiles, driven by both API payload (i.e., local steric and electronic structural environment) and
linker functionality. However, it’s also known that lower loaded NDCs (~50% of theoretical) accelerate the
in vivo release of drug from the nanoparticles, reflected by increases in released drug AUC.41 So it is
possible that the lower loaded, faster releasing GEM NDCs overcome any structural differentiation
imparted by their linkers, resulting in an in vivo released drug normalization effect.
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
An additional difference between the CBTX and GEM NDCs is the higher availability of released GEM
in systemic circulation after NDC dosing (i.e., % AUC of separately dosed parent GEM), which is much
greater than 100% (F = ~500-600%). This would indicate either Cl or distribution of the API is affected by
GEM NDC administration. A possible explanation for the former is that encapsulation of the GEM drug
payload in the NDC protects it from extensive deamination, resulting in greater AUC (and decreased Cl) of
released active drug compared to separately dosed GEM.
GEM
TOTAL_GEM-glycinate NDC
1
000
RELEASED_GEM-glycinate NDC
TOTAL_GEM-hexanoate NDC
RELEASED_GEM-hexanoate NDC
1
00
10
1
0.1
0
.01
0
20
40
60
Time (hr)
ACS Paragon Plus Environment

Page 21 of 53
Journal of Medicinal Chemistry
2
1
1
2
3
4
5
6
7
8
9
Figure 5. In vivo plasma exposure profiles comparing separately dosed GEM to total and released GEM from
GEM-glycinate 11c and GEM-hexanoate 12c NDCs after a single iv dose of 10 mg/kg each given to non-tumor
bearing male C57BL/6 mice.
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
Table 4. Plasma PK parameters after a single iv dose of 10 mg/kg GEM, GEM-glycinate NDC 11c or
GEM-hexanoate NDC 12c was given to male C57BL/6 mice.
C or
0
VDa
Fb
AUC∞
tmax
Cl
t1/2
Test Article
Cmax
[µM*h]
[hr]
[mL/min/kg]
[h]
[L/kg]
[%]
[µM]
GEM
12
7466
75
59
-
-
55
0.087
-
1.3
11
18
16
13
1.5
0.02
-
-
-
Total
Released
Total
820
8.8
830
12
GEM-
glycinate NDC
0.08
-
625
-
16894
62
0.037
-
0.05
GEM-hexanoate
NDC
Released
0.5
-
517
a
b
V
D
is calculated as Vss (estimated volume of distribution at steady state) for nanoparticles dosed iv. F = availability of
administered GEM dose in systemic circulation.
2.4. Efficacy Studies of Cabazitaxel NDCs with Glycinate and Hexanoate Linkers
Definitive in vivo efficacy studies for the CBTX NDCs were conducted in mouse tumor models to
determine the impact of sustained drug release (via NDC) in plasma/tumors on efficacy and survival
compared to parent drug in a multi-day study setting. Given the differentiated and favorable PK profiles
(plasma and tumor) for the CBTX-glycinate 1c and CBTX-hexanoate 2c NDCs, both were selected for the
efficacy studies. Dosing was based on the respective maximum tolerated dose (MTD) of each treatment,
determined as the dose that caused <20% mean body weight loss in tumor-bearing mice after qwx2 dosing,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 53
2
2
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
which was determined a priori to be slightly higher for the more stable CBTX-hexanoate NDC 2c than the
faster releasing CBTX-glycinate NDC 1c and CBTX parent drug. Male mice carrying the taxane-resistant
B16.F10 melanoma syngeneic tumors were given two iv weekly treatments (qwx2) of either the CBTX-
glycinate (24 mg/kg) or CBTX-hexanoate (30 mg/kg) NDCs or CBTX (24 mg/kg); the vehicle solution arm
was used as a negative control. Tumor growth continued exponentially, from the initial average tumor size
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
3
of 60 mm , for the vehicle control as well as CBTX parent drug after 6 and 12 days, respectively (Fig. 6A).
In contrast, tumor growth was inhibited up to Day-30 (18 days post last dose) for CBTX-glycinate NDC 1c
and up to Day-44 (32 days post last dose) for CBTX-hexanoate NDC 2c. The demonstrated efficacy in the
mouse tumor model for the NDCs is supported by the high and sustained levels of both released
(
efficacious) and total drug (driving released drug levels) in tumor, as shown in the B16.F10 PK study. The
corresponding survival curves from the efficacy study showed significant improvement over the 50-day
experiment for the NDCs compared to vehicle and parent drug (Fig. 6B), with nearly 100% survival for the
CBTX-hexanoate NDC 2c. The overall optimal efficacy profile of the CBTX-hexanoate NDC 2c can be
explained by its plasma and tumor PK. This PK profile, providing substantially (>10x) reduced plasma Cmax
and increased tumor half-life for released drug vs. parent, allowed for a more tolerated and higher
efficacious dose (>4x projected tumor AUC vs. parent drug) with greater sustained drug release to be
delivered in the in vivo tumor model.
ACS Paragon Plus Environment

Page 23 of 53
Journal of Medicinal Chemistry
2
3
1
2
3
4
5
6
7
8
9
A
3000
2
000
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
1000
0
5
10 15 20 25 30 35 40 45 50
First Dose
Day Post Implantation
Dosing Period
B
100
75
50
CBTX *
CBTX-Gly NDC * **
CBTX-Hex NDC * **
2
5
0
5
10 15 20 25 30 35 40 45 50
First Dose
Day Post Implantation
Dosing Period
Figure 6. In vivo efficacy and survival graphs. (A) Efficacy graph of CBTX (24 mg/kg) as well as CBTX-
glycinate 1c (24 mg/kg) and CBTX-hexanoate 2c (30 mg/kg) NDCs after two weekly doses (iv, qwx2) in male
mice carrying B16.F10 melanoma syngeneic tumors. (B) Corresponding survival curves from B16.F10 efficacy
study. Log-rank test used for statistical analysis and showed there was a statistically significant time-to-event
difference between vehicle and the CBTX, CBTX-glycinate, CBTX-hexanoate groups with a *p-value < 0.0001,
and a statistically significant time-to-event difference between CBTX and the CBTX-glycinate, CBTX-hexanoate
groups with a **p-value < 0.0001. Black circles represent the vehicle solution as control.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 53
2
4
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
A second efficacy study was conducted in male mice carrying UISO-BCA-1 breast xenograft tumors,
57
a tumor line that has shown resistance to docetaxel treatment. The mice were given two weekly doses (iv,
qwx2) of CBTX (24 mg/kg) as well as CBTX-glycinate 1c (18 mg/kg) and CBTX-hexanoate 2c (30 mg/kg)
NDCs. As in the first study, the dose levels were based on the differing MTDs measured for the compounds;
albeit a slightly lower MTD was obtained for the CBTX-glycinate NDC 1c in this tumor model. Signs of
tumor regrowth occurred for some mice (2 of 8) around Day-90 for the group treated with parent drug
CBTX, in contrast to tumor regression and tumor growth delay out to Day-130 for 100% of mice treated
with the CBTX-glycinate NDC 1c (Fig. 7A). Remarkably, tumors treated with the CBTX-hexanoate NDC
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
3
3
2
c showed growth from average tumor size of 170 mm to about 330 mm during the initial stages of
treatment, before tumor sizes decreased considerably by Day 37 (with no additional treatment). This
resulted in 100% of the mice tumor-free (i.e., tumor free survival, TFS) for the remainder of the 149-day
experiment for the CBTX-hexanoate NDC 2c. Both CBTX NDCs demonstrated greater tumor growth delay
vs. parent drug in this tumor model, and the more stable hexanoate NDC 2c led to 100% survival at Day
1
49 as shown in the corresponding efficacy survival curves (Fig. 7B).
A
1
000
800
6
00
400
2
00
0
1
0
30
50
70
90
110
130
150
First Dose
Day Post Implantation
Dosing Period
ACS Paragon Plus Environment

Page 25 of 53
Journal of Medicinal Chemistry
2
5
1
2
3
4
5
6
7
8
9
B
1
00
7
5
2
5
0
5
0
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
CBTX *
CBTX-Gly NDC *
CBTX-Hex NDC *
1
0
30
50
70
90
110
130
150
First Dose
Day Post Implantation
Dosing Period
Figure 7. In vivo efficacy and survival graphs. (A) Efficacy graph of CBTX (24 mg/kg) as well as CBTX-
glycinate 1c (18 mg/kg) and CBTX-hexanoate 2c (30 mg/kg) NDCs after two weekly doses (iv, qwx2) in male
mice carrying UISO-BCA-1 breast xenograft tumors. (B) Corresponding survival curves from UISO-BCA-1
efficacy study. Log-rank test used for statistical analysis and showed there was a statistically significant time-to-
event difference between vehicle and the CBTX, CBTX-glycinate, CBTX-hexanoate groups with a *p-value <
0.0001. Black circles represent the vehicle solution as control.
3
. Conclusions
We have demonstrated that diverse API payloads (docetaxel, cabazitaxel, and gemcitabine) with a
range of physicochemical properties (e.g., MW, lipophilicity, and aq solubility) can be conjugated via a
variety of linkers to a β-cyclodextrin-PEG (CDP) copolymer backbone. Using this technology, we were
able to fine tune in vitro and in vivo drug release kinetics that led to the generation of two optimized NDCs,
CBTX-hexanoate 2c and CBTX-glycinate 1c. The in vitro release data for the CBTX NDCs provided
predictable IVIVC correlation on a “rank-order” level, driven by the differentiated release rates of the
linkers. Studies in native and heat-inactivated plasma had no influence on the release kinetics, indicating
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 53
2
6
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
that drug release was likely pH- but not enzyme-triggered. By taking advantage of the strengths of the NDC
linker technology (i.e., tunable release rates based on linker design) we were able to provide differentiated
and improved PK profiles for both CBTX NDCs, particularly in plasma, as exemplified in the B16.F10 PK
study. The result was lower Cmax (~4-50x) for released drug in plasma and vastly reduced (~200-900x)
volume of distribution and clearance of total drug for both the CBTX- glycinate 1c and CBTX- hexanoate
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
c NDCs (moreso) compared to separately dosed parent drug. In tumor, both NDCs provided increased (vs.
parent) tumor exposure of total drug (~8-30x); however, only CBTX-hexanoate NDC 2c provided a
substantial increase in released drug AUC (3-4x) and half-life (>10x) vs. CBTX. CBTX-hexanoate NDC
2
c also demonstrated an impressive released drug tumor-to-plasma AUC ratio of 40:1. Increases in released
drug plasma half-life over parent drug (e.g., >10x for the GEM NDCs) support the use of the NDC platform
as a half-life extension technology. The CBTX NDCs demonstrated improved in vivo efficacy compared to
CBTX parent drug. CBTX-hexanoate NDC 2c (CRLX522) provided an optimal overall in vivo PK and
efficacy/safety profile as demonstrated by greater tumor growth delay vs. parent in both efficacy models,
1
00% TFS in the UISO-BCA-1 model, significantly enhanced survival vs. parent in the B16.F10 model (>
8
8% survival in both models), and higher MTD. The results support further evaluation of CRLX522 as a
potential anticancer therapeutic agent. In conclusion, the studies presented here demonstrate the utility and
diversity of the NDC linker technology platform and pave the way for future evolution of the NDC platform,
to include the conjugation of multiple payloads to a single NDC and the development of antibody-
conjugated NDCs.61-63
4
. Experimental Section
4
.1. Materials
Acetonitrile (ACN) HPLC grade or equivalent, N-tert-butoxycarbonyl-alanine (Boc-alanine), N-tert-
butoxycarbonyl-β-alanine-OH (Boc-β-alanine-OH), N-tert-butoxycarbonyl-6-aminohexanoic acid (Boc-
ACS Paragon Plus Environment

Page 27 of 53
Journal of Medicinal Chemistry
2
7
1
2
3
4
5
6
7
8
9
aminohexanoic acid), N-tert-butoxycarbonyl-glycine (Boc-glycine), tert-butyl bromoacetate, cabazitaxel
(
CBTX, Tecoland Co.), N-carbobenzoxy-alanine (Cbz-alanine), N-carbobenzoxy-6-aminohexanoic acid
(
Cbz-aminohexanoic acid), N-carbobenzoxy-glycine (Cbz-glycine), charcoal Darco® 175, β-cyclodextrin-
PEG (CDP, Cerulean Pharma Inc.) copolymer, dichloroacetic acid (DCA), dichloromethane (DCM), N,N-
diisopropylethylamine (DIEA), N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride
EDC·HCl), 4-(dimethylamino)pyridine (DMAP), N,N-dimethylformamide (DMF), docetaxel (DTX,
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
(
LGM Pharma), formic acid (99+%), gemcitabine (GEM, Bosche Scientific), heptanes, N-
hydroxysuccinimide (NHS), magnesium sulfate (MgSO ), methanesulfonic acid (MSA), methanol (MeOH)
4
HPLC grade or equivalent, 4-methoxytriphenylmethyl chloride (MMT-Cl), palladium on carbon (10%
Pd/C), sodium sulfate (Na SO ), trifluoroacetic acid (TFA), water HPLC grade or equivalent. All chemicals
2
4
purchased from Sigma-Aldrich unless noted otherwise.
4
.2. Analytical Methods
Analytical HPLC analyses for (i) NDC drug concentration, loading and purity, and (ii) small molecule
purity levels were conducted using: (i) an Agilent 1200, a photodiode array (PDA) detector, a Zorbax
3
00SB-C18 or YMCpac CN300 column, multiple gradients using water/(acetonitrile/methanol) mobile
phases with 0.1% of formic acid or 0.5% TFA; (ii) a Waters 2695, ultraviolet (UV) detection at 230 nm, a
column Agilent Zorbax Eclipse XTD-C8, a linear gradient from 5% to 99% B (mobile phase A = water +
0.1% TFA and B = ACN + 0.1% TFA) or an Agilent 1200, a PDA detector, a C18 column, a linear gradient
from 5% to 95% B (mobile phase A = water + 0.1% formic acid and B = ACN + 0.1% formic acid). The
concentration of solid material was determined using an oven gravimetric method. Drug loading was
calculated based on concentration of drug vs. total solids. Particle size for NDCs was measured in water
using a Malvern Zetasizer Nano ZS dynamic light scattering instrument and automatic settings. The original
sample was diluted to reach a concentration of solids in the suspension of 5 mg/mL level (mass of a solid
estimated based on gravimetric method) using appropriate viscosity values. Proton magnetic resonance
spectra were recorded on either a Varian Inova 400 MHz or a Brüker Avance 300 MHz spectrometer. MS
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 53
2
8
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
spectra were recorded on either a PE SCIEX, API 150EX (HPLC – SHIMADZU SCL-10A with LC-10AD
pumps) or an Agilent 6410 mass spectrometer. Automated flash chromatography was performed using
either an Agilent 971-FP purification system with RediSep Rf Gold Silica Flash Chromatography columns
(
20-40 µm) or a Combiflash system - Companion XL with Foxy200 fraction collector. Tangential Flow
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
Filtration (TFF) purification of NDCs was preformed using a Pall Minimate System with a Millipore
Pellicon XL 30kDa cartridge. DLS characterization, concentration, loading, and purity data for the final
generated NDCs are provided in Table 1. Purities for final NDC SAR test compounds are > 95.0% and the
analytical methods used to determine purity are as follows: an Agilent 1200 with a PDA detector and a
Zorbax 300SB-C18 (CBTX, DTX NDCs) or YMCpac CN300 column (GEM NDCs), using multiple
gradients and water/(acetonitrile/methanol) mobile phases with 0.1% of formic acid (CBTX, DTX) or 0.5%
TFA (GEM).
4
.3.1. Syntheses of Cabazitaxel NDCs
(
2aR,4S,4aS,6R,9S,11S,12S,12aR,12bS)-12b-Acetoxy-9-(((2R,3S)-2-
(((benzyloxy)carbonyl)glycyl)oxy)-3-((tert-butoxycarbonyl)amino)-3-phenylpropanoyl)oxy)-11-
hydroxy-4,6-dimethoxy-4a,8,13,13-tetramethyl-5-oxo-2a,3,4,4a,5,6,9,10,11,12,12a,12b-dodecahydro-
H-7,11-methanocyclodeca[3,4]benzo[1,2-b]oxet-12-yl benzoate (1a). Cbz-glycine (0.58 g, 2.8 mmol),
CBTX (2.0 g, 2.4 mmol) and DMAP (0.35 g, 2.9 mmol) were dissolved in DCM (20 mL), then EDC·HCl
0.55 g, 2.9 mmol) was added. The reaction stirred at RT overnight. The reaction mixture was diluted with
(
1
(
DCM (50 mL) and washed with water (2 x 50 mL) followed by brine (50 mL). The solvent was removed
under vacuum and the residue was purified by flash SiO chromatography eluting with a gradient of 0 –5
2
1
% DCM:MeOH to furnish a white solid (2.2 g, 90 % yield). H NMR (δ ppm, DMSO-d6): 0.95 (s, 3H),
0
6
.97 (s, 3H), 1.37 (s, 9H), 1.49 (m, 4H), 1.77 (m, 3H), 2.23 (s, 3H), 2.65 (m, 1H), 3.20 (s, 3H), 3.56 (d, J =
.9 Hz, 1H), 3.74 (m, 1H), 3.90 (t, J = 5.4 Hz, 1H), 4.00 (s, 2H), 4.48 (s, 1H), 4.68 (s, 1H), 4.94 (d, J = 9.6
Hz, 1H), 5.05-5.11 (m, 4H), 5.36 (d, J = 6.9 Hz, 1H), 5.88 (t, J = 8.8 Hz, 1H), 7.19 (t, J = 7.1 Hz, 1H), 7.29-
ACS Paragon Plus Environment

Page 29 of 53
Journal of Medicinal Chemistry
2
9
1
2
3
4
5
6
7
8
9
7
.43 (br m, 9H), 7.65 (t, J = 7.6 Hz, 2H), 7.72 (d, J = 7.1 Hz, 1H), 7.76 (t, J = 6 Hz, 1H), 7.90 (d, J = 8.9
Hz, 1H), 7.97 (d, J = 7.6 Hz, 2H).
(
2aR,4S,4aS,6R,9S,11S,12S,12aR,12bS)-12b-Acetoxy-9-(((2R,3S)-3-((tert-
butoxycarbonyl)amino)-2-(glycyloxy)-3-phenylpropanoyl)oxy)-11-hydroxy-4,6-dimethoxy-
a,8,13,13-tetramethyl-5-oxo-2a,3,4,4a,5,6,9,10,11,12,12a,12b-dodecahydro-1H-7,11-
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
4
methanocyclodeca[3,4]benzo[1,2-b]oxet-12-yl benzoate (1b). Compound 1a (2.1 g, 2.0 mmol) was
dissolved in THF (44 mL) and MeOH (1.3 mL) followed by the addition of methane sulfonic acid (0.18 g,
1.8 mmol) and 10 % Pd/C (0.68 g). The solution was placed in a hydrogen reactor (30 PSI) for 1 h. The
solids were removed by filtration through a Celite pad. Precipitation from heptanes at ambient temperature
1
gave the desired CBTX-glycinate as an off-white solid (1.7 g, 85 % yield). H NMR (δ ppm, DMSO-d6):
0
3
4
6
1
.97 (s, 3H), 0.98 (s, 3H), 1.36 (s, 9H), 1.50 (m, 4H), 1.82 (m, 3H), 2.27 (s, 3H), 2.29 (s, 2H), 2.65 (m, 1H),
.20 (s, 3H), 3.59 (d, J = 6.7 Hz, 1H), 3.73 (m, 1H), 3.86 (m, 1H), 4.02 (s, 2H), 4.53 (s, 1H), 4.69 (s, 1H),
.95 (d, J = 9.8 Hz, 1H), 5.13-5.18 (m, 2H), 5.38 (d, J = 7.1 Hz, 1H), 5.87 (t, J = 7.6 Hz, 1H), 7.21 (t, J =
.5 Hz, 1H), 7.36-7.45 (br m, 4H), 7.64 (t, J = 7.6 Hz, 2H), 7.71 (d, J = 7.4 Hz, 1H), 7.88 (d, J = 8.5 Hz,
+
H), 7.99 (d, J = 7.4 Hz, 2H). MS m/z calculated for C H N O , 892.40, found [M+H] , 893.4.
4
7
60
2
15
CBTX-glycinate NDC (1c). CDP copolymer (2.0 g, 0.41 mmol) in dry DMF (20 mL), 1b (0.90 g, 0.91
mmol), EDC·HCl (0.17 g, 0.91 mmol), NHS (0.10 g, 0.91 mmol), and DIEA (0.12 g, 0.91 mmol) were
mixed for 2 hours and the polymer conjugate was precipitated into IPA (300 mL). The IPA was decanted
and the beaker was charged with acetone (150 mL). After stirring for 15 minutes the acetone was decanted
and residual polymer was dissolved in pH 3 water (200 mL). The crude CBTX-glycinate-NDC was purified
by TFF in pH 3 water (30k MWCO regenerated cellulose membrane). The retentate was concentrated (80
mL), filtered through Millipore Steriflip 0.22-µm filter, and the resulting NDC aqueous solution was stored
o
at -20 C. See Table 1 for DLS characterization, concentration, loading, and purity data.
ACS Paragon Plus Environment