Synthesis and Activity of a Novel Autotaxin Inhibitor-Icodextrin Conjugate

Article abstract of DOI:10.1021/acs.jmedchem.8b00935

Autotaxin is an extracellular phospholipase D that catalyzes the hydrolysis of lysophosphatidyl choline (LPC) to generate the bioactive lipid lysophosphatidic acid (LPA). Autotaxin has been implicated in many pathological processes relevant to cancer. Int

Full text of DOI:10.1021/acs.jmedchem.8b00935

Subscriber access provided by - Access paid by the | UCSB Libraries
Article
Synthesis and Activity of a Novel Autotaxin Inhibitor-Icodextrin Conjugate
Natalie Fisher, Michael G Edwards, Ryan Hemming, Steven Mark Allin, John D. Wallis, Philip C. Bulman Page,
Michael J. McKenzie, Stephanie Jones, Mark R. J. Elsegood, John King-Underwood, and Alan Richardson
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00935 • Publication Date (Web): 30 Jul 2018
Downloaded from http://pubs.acs.org on July 31, 2018
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 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis and Activity of a Novel Autotaxin Inhibitor-Icodextrin Conjugate
a,b
b
c
c
c
f
Natalie Fisher, Michael G. Edwards, Ryan Hemming, Steven M. Allin, John D. Wallis,
d
e
a
Philip C. Bulman Page, Michael J. Mckenzie, Stefanie M Jones, Mark R. J. Elsegood,
g
a
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
John King-Underwood and Alan Richardson *
a
School of Pharmacy, Keele University, Keele, ST5 5BG, UK
b
Keele Molecular Chemistry Group, Lennard-Jones Laboratories, School of Chemical and
Physical Sciences, Keele University, Keele, ST5 5BG, UK
c
School of Science and Technology, Nottingham Trent University, Nottingham NG11 8NS,
UK
d
School of Chemistry, University of East Anglia, Norwich NR4 7TJ
e
Charnwood Molecular Ltd., The Heritage Building, Prince William Road, Loughborough
LE11 5DA, UK
f
Department of Chemistry, Loughborough University, Loughborough, LE11 3TU, UK
g
Computational Chemistry Resource, Old Cottage Hospital, Ledbury, HR8 1ED, UK
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 39
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
ORCHID
Alan Richardson 0000-0003-1825-3375
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
*Corresponding author: a.richardson1@keele.ac.uk
2
ACS Paragon Plus Environment

Page 3 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT
Autotaxin is an extracellular phospholipase D that catalyses the hydrolysis of
lysophosphatidyl choline (LPC) to generate the bioactive lipid lysophosphatidic acid (LPA).
Autotaxin has been implicated in many pathological processes relevant to cancer.
Intraperitoneal administration of an autotaxin inhibitor may benefit patients with ovarian
cancer, however low molecular mass compounds are known to be rapidly cleared from the
peritoneal cavity. Icodextrin is a polymer that is already in clinical use because it is slowly
eliminated from the peritoneal cavity. Herein we report conjugation of the autotaxin inhibitor
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
HA-155 to icodextrin. The conjugate inhibits autotaxin activity (IC50 = 0.86 ± 0.13 µg mL )
and reduces cell migration. Conjugation of the inhibitor increased its solubility, decreased its
membrane permeability and improved its intraperitoneal retention in mice. These
observations demonstrate the first application of icodextrin as a covalently-bonded drug
delivery platform with potential use in the treatment of ovarian cancer.
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 39
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
INTRODUCTION
1
Autotaxin is an extracellular phospholipase D that hydrolyses lysophosphatidyl choline
2
(LPC) to the bioactive lipid lysophosphatidic acid (LPA) and choline. LPA acts via G
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
protein coupled receptors located on the cell surface to activate a variety of signaling
3
pathways. LPA-induced signaling pathways are implicated in a number of biological
4
processes including cell migration, proliferation and survival. Patients with ovarian cancer
often present with an accumulation of ascites fluid in the intraperitoneal cavity which
5
contains LPA at concentrations up to 80 µM. Autotaxin, also found in the ascites fluid of
6
patients with ovarian cancer, is over-expressed in ovarian cancers that are resistant to
7
chemotherapy, and has been shown to delay apoptosis induced by carboplatin in ovarian
8
cancer cells. Autotaxin is known to catalyse the production of LPA in many biological
2b
9
fluids and LPA has a well-established role in cancer cell migration and invasion.
Inhibition of autotaxin has been shown to increase the sensitivity of ovarian cancer cells to
1
0
11
carboplatin and breast cancer cells to paclitaxel. Reducing the high levels of LPA present
in ascites could therefore be expected to delay relapse after chemotherapy and reduce drug
resistance, suggesting that a strategy involving inhibition of autotaxin within the peritoneal
cavity has therapeutic potential for the treatment of ovarian cancer. Several small molecule
1
2
inhibitors of autotaxin have been described which might be used to achieve this. Maintaining a
high local concentration of an autotaxin inhibitor within the peritoneal cavity is likely to be
13
important in order to ensure sufficient and prolonged inhibition of this enzyme. However,
the residence time of small molecule drugs within the peritoneal cavity is usually relatively
short because low molecular mass compounds are quickly absorbed through the peritoneal
capillaries into the systemic circulation. In contrast, high molecular mass compounds are
generally cleared more slowly from the peritoneal cavity, suggesting that coupling of a drug
to a polymer offers a potential strategy to increase the residence time of the therapeutic agent.
4
ACS Paragon Plus Environment

Page 5 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Polymers are becoming an increasingly useful tool in the delivery of drugs and have the
potential to improve the desired therapeutic properties of bioactive small molecules. We have
previously described the successful synthesis and attachment of the known autotaxin inhibitor
rd
S32826 to a 3 generation PAMAM dendrimer whilst retaining pharmacological activity,
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
demonstrating the feasibility of conjugating autotaxin inhibitors to a polymeric delivery
14
system.
In the related studies described in this paper, our attention was drawn to the potential
application of icodextrin as an alternative polymeric support. Icodextrin is the main
constituent of Extraneal™, which has a history of safe clinical use for peritoneal dialysis.
Icodextrin is a water-soluble polymer derived from maltodextrin containing α-1,4-glycosidic
bonds and approximately 10% α-1,6-glycosidic bonds, and an average molecular weight of
13 - 19 kDa. The prolonged retention of icodextrin in the intraperitoneal cavity has been well
characterized in clinical trials and is the basis for its clinical use in peritoneal dialysis, with
approximately 60% of the icodextrin administered to the intraperitoneal cavity remaining
15
after 12 hours. Higher molecular weight icodextrin fractions are anticipated to be eliminated
even more slowly. There have been several studies which have evaluated the co-
administration of non-covalently coupled icodextrin, as a carrier solution to increase the
1
6
retention of drug molecules in the peritoneal cavity but to our knowledge there have been
no reports detailing the use of icodextrin as a support to form a covalently-bound drug
conjugate. Given the many advantages noted above, we reasoned that icodextrin could be
investigated as a novel drug delivery system for intraperitoneal administration and targeting
of the autotaxin-LPC system in the peritoneal cavity. Herein we describe our progress, and
report the synthesis and activity of a novel autotaxin inhibitor-icodextrin conjugate based on
17
the autotaxin inhibitor HA-155.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 39
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
RESULTS AND DISCUSSION
18
The publication of the crystal structure of autotaxin in complex with the small-molecule
inhibitor HA155 enabled us to perform structure-based drug design of a drug-polymer
conjugate. Inspection of the molecular structure suggested that the 3-position on the central
benzylidene moiety was accessible to solvent and might provide an appropriate locus to
cross-link the drug to a polymer without interfering with the drug binding to the enzyme (Fig
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
16
17
1
). Furthermore, structurally related compounds described by Albers contained a methoxy
group at this position, suggesting substitution was tolerated at this position. We chose to
replace the boronic acid moiety in HA155 with a carboxylic acid because this was previously
17
shown to be tolerated and because boronic acid is found in relatively few marketed drugs.
Figure 1. Crystal structure of HA155 bound to autotaxin (adapted from ref 18). The
structure of HA155 is shown in the inset. The proposed attachment point for the polymer is
shown (arrow).
6
ACS Paragon Plus Environment

Page 7 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A convenient and convergent synthetic approach based on the modification of a route
17
previously used to synthesise HA155 was pursued (Scheme 1). Thiazolidine-2,4-dione was
N-benzylated with 4-fluorobenzyl chloride to yield intermediate 1. Selective O-benzylation of
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
3,4-dihydroxybenzaldehyde with methyl-4-(bromomethyl)-benzoate following a method
19
3
described by Plourde and Spaetzel using Finkelstein conditions (NaHCO , NaI, DMF, 40
°C) allowed benzylation of the C4-hydroxyl group; the structure of compound 2 was
confirmed by single crystal X-Ray diffraction (Figure 2).
Our initial studies evaluated the direct attachment of 2 to the polymer via the remaining C3-
phenolic hydroxyl group, however inconsistent coupling at this stage prompted us to
incorporate a spacer group in the form of a bis-ethylene glycol linker. This was accomplished
by alkylation of the C3-hydroxyl group in 2 with 2-(2-chloroethoxy)-ethanol. Saponification
of the resulting methyl ester with sodium hydroxide gave intermediate 3. Finally, a
Knoevenagel-like condensation of the thiazolane-2,4-dione 1 with aldehyde 3 delivered the
novel, functionalized autotaxin inhibitor, 4.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 39
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 1. Synthesis of functionalized autotaxin inhibitor, 4: i. NaH, DMF, 0 °C to rt, 22 h (75%); ii.
NaHCO
3
2 3
, NaI, DMF, 40 °C, 24 h (70%); iii. 2-(2-chloroethoxy)-ethanol, DMF, K CO , KI, 70 °C, 18
h (58%); iv. NaOH, DMSO/H
2
O 70 °C, 2 h (98%); v. piperidine, EtOH, reflux, 16 h (93%).
Figure 2: X-Ray structure of the selectively benzylated intermediate 2. Displacement ellipsoids are at
the 50% probability level. Intramolecular H-bond shown with a dashed line.
In order to investigate its potential for supporting an appropriately derivatised autotaxin
inhibitor molecule, icodextrin solid was isolated from Extraneal™ by dialysis against water
to remove electrolytes, small organic molecule contaminants and icodextrin chains smaller
20
than 20 kDa.
8
ACS Paragon Plus Environment

Page 9 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To facilitate the attachment of the inhibitor derivative 4 to icodextrin it was necessary to
introduce an appropriate electrophilic functional group, capable of allowing alkylation of the
primary alcohol at the C6 position of the glucose units in icodextrin. Thus the carboxylic acid
group in 4 was first protected as the TIPS silyl ester 5, before activation of the terminal
hydroxyl group of the linker moiety within 5 as the tosylate 6. Icodextrin is poorly soluble in
organic solvents due to extensive hydrogen bonding and incorporated water, but can be
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
21
dissolved in N,N-dimethylacetamide (DMAc) with added LiCl to displace water.
Once
icodextrin had been dissolved in the DMAc/LiCl mixture, the coupling of icodextrin and 6
was achieved through deprotonation of icodextrin using sodium hydride, followed by
addition of tosylate 6. The novel icodextrin-inhibitor conjugate 7 was finally isolated by
dialysis of the reaction mixture against methanol. This synthetic procedure was found to lead
to in situ removal of the TIPS ptotecting group from the conjugate. The structure of conjugate
7
was supported by NMR, IR and TGA analysis (Figure 3). An unconjugated analogue of 7
was also prepared (8) in order to allow comparisons to be drawn on the effect of the
icodextrin coupling on the activity of the supported compound. Table 1 summarizes some
chemical and pharmaceutical properties of conjugate 7.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 39
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
3
Scheme 2. Synthesis of icodextrin-inhibitor conjugate: i. TIPS-Cl, Et N, THF, rt, 15 min (76%); ii.
2 2 3
p-TsCl, CH Cl , Et N, rt, 20 h (61%); iii. Icodextrin, NaH, LiCl, DMAc, rt, 4 h (43%).
1
0
ACS Paragon Plus Environment

Page 11 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
6-20-07-MGE-TP.010.001.1r.esp
Water
A
B
ICODEXTRIN
INHIBITOR
ICODEXTRIN & INHIBITOR
8
.5
8.0
7.5
7.0
6.5
6.0
Chemical Shift (ppm)
5.5
5.0
4.5
4.0
3.5
3.0
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
C
1
Figure 3: Analysis of icodextrin-inhibitor conjugate 7: A) 300 MHz H NMR analysis of
conjugate 7; B) IR analysis of free inhibitor 8 and conjugate 7; C) TGA analysis of free
inhibitor 8, icodextrin, and icodextrin-inhibitor conjugate 7.
To establish the amount of inhibitor attached to the conjugate, the inherent fluorescence of
the immobilized inhibitor was measured in a solution of 7 and compared with the
fluorescence of known concentrations of non-conjugated, free inhibitor 8. Knowing the
concentration of dissolved polymer allowed us to determine that approximately 0.09 g of the
−
1
inhibitor molecule was attached per g of polymer (0.19 mmol g ).
The aqueous solubility of the compounds in PBS (pH 7.4) was tested using a miniaturized
shake flask method, and conjugation to icodextrin was observed to increase the solubility of
the pharmacophore compared with the free inhibitor 8 (Table 1).
It was hypothesized that conjugation to icodextrin would increase the retention time of the
autotaxin inhibitor in the intraperitoneal cavity (due to the substantial increase in size slowing
down its absorption). The effect of conjugation on the membrane permeability of the
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 39
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
compounds was tested in two assays: (i) in a PAMPA assay measuring diffusion across an
artificial bilayer and (ii) using a Caco-2 cell monolayer. In the PAMPA assay, the
unconjugated inhibitor 8 was observed to be moderately permeable across the membrane but
once the inhibitor was conjugated to icodextrin there was a marked decrease in permeability
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
(i.e. of conjugate 7). In the Caco-2 assay, conjugate 7 also showed poor permeability (Papp =
−6
−6
7
.4 × 10 ) when compared with a freely permeable drug, paracetamol (Papp = 38 × 10 ).
Degree of incorporation
of inhibitor
Solubility
(g L–1)
Permeability
(Log Pe)
Compound
–
1
mmol g )
(
Free inhibitor 8
Conjugate 7
n/a
0.08 ± 0.02
0.22 ± 0.02
–5.5 ± 0.2
–6.4 ± 0.1
0.19
Table 1: Chemical and Pharmaceutical Properties of Conjugate 7 and Free Inhibitor 8:
Solubility was measured using a miniaturised shake flask method. (Results are expressed as
mean ± S.D. n = 3); Permeability was measured in PAMPA assay. Permeability (mean +S.D.,
n = 2–3) of the free inhibitor 8 and conjugate 7 were determined using furosemide (log Pe =
−
6.4 ± 0.3) and propranolol (log Pe = −5.3 ± 0.1) as controls for low and high permeable
drugs respectively
.
1
2
ACS Paragon Plus Environment

Page 13 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
AUTOTAXIN ACTIVITY ASSAYS
The ability of the novel conjugate 7 to inhibit autotaxin was assessed in two autotaxin
22
enzymatic activity assays measuring the hydrolysis of two different substrates, FS-3 and Bis
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
2, 23
p-NPP.
Using FS-3 as the substrate, both conjugate 7 and the free inhibitor 8 inhibited
the activity of autotaxin. When the potency of conjugate 7 was expressed in terms of the
concentration of pharmacophore attached to the polymer, conjugation appeared have had
minimal effect on the drug potency (Table 2). Using bis-pNPP as the substrate, the free
inhibitor 8 showed a slightly lower potency to that measured in the FS-3 assay (Table 2) but
the potency of the conjugate 7 was comparable in both assays. Unconjugated icodextrin itself
−
1
(
100 µg mL ) had no measureable effect in either assay.
IC50
Compound
FS-3 assay
Bis-pNPP assay
Free inhibitor 8
Conjugate 7
290 ± 160 nM (n=5)
620 ± 160 nM(n=2)
−
1
−1
0
.86 ± 0.13 µg mL (n=4)
0.94± 0.12 µg mL (n=3)
{pharmacophore IC50 = 160 ± 20 nM} {pharmacophore IC50 = 170 ± 20 nM}
Table 2: Inhibition of autotaxin enzymatic activity by the free inhibitor 8 and conjugate 7,
tested in FS-3 and Bis-pNPP assay. IC50 values (mean ± S.D., number (n) of experiments
−
1
shown in parentheses)) are shown as either the concentration of conjugate (µg mL ) or the
concentration of the free inhibitor (nM). To allow comparision, the IC50 of the conjugate is
also expressed in terms of the concentration of the pharmacophore present, determined by
taking into account the amount of drug conjugated to the icodextrin.
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 39
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
INHIBITION OF CELL MIGRATION
Autotaxin and LPA have been shown to regulate the migration of cancer cells. To further
invetigate the biological activity of conjugate 7, its effect on wound healing was evaluated
using 3E3 ovarian cancer cells; these cells were previously derived from OVCAR-3 cells to
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
10
over-express autotaxin. OVCAR-3 cells transfected with the vector 3V5, were used as a
10
control and express low levels of autotaxin. Once an even monolayer of cells had formed a
wound was inflicted and the migration of the cells was measured. Both the free autotaxin
inhibitor 8 and the icodextrin-drug conjugate 7 were observed to reduce wound closure by
approximately 50% compared with cells exposed to vehicle alone (DMSO), (Figure 4).
Neither the free autotaxin inhibitor 8 nor the conjugate 7 had a significant effect on the
percentage wound closure in the control 3V5 cells.
Figure 4: Comparison of autotaxin inhibition by the free inhibitor 8 or conjugate 7 on wound
closure. 3E3 cells were grown to confluence, a wound was created and serum free media
1
4
ACS Paragon Plus Environment

Page 15 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
−
1
containing vehicle (DMSO), the free inhibitor 8 (100 nM), or conjugate 7 (10 µg mL ,
equivalent to 1.9 µM pharmacophore) was added and wound closure was measured after 16
h. The percentage wound closure was measured (mean + S.D., n = 5) and was significantly
different from wound closure in cells treated with vehicle where shown (*, P < 0.001; **, P <
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
0
.005, vs. control (paired t-test)).
INTRAPERITONEAL RETENTION
We next evaluated whether conjugation of the inhibitor to icodextrin promoted retention in
the peritoneal cavity of nu/nu mice. The compounds were separately administered by the i.p.
route to mice. After 24 h the peritoneal cavity was washed and the amount of compound
present in the wash was measured. Whilst 30% of conjugate 7 was recovered from the
peritoneal cavity after 24 h, in comparison only 0.1% of the free inhibitor 8 was recovered
(Figure 5).
Figure 5: Peritoneal Retention of Conjugate 7 and Free Inhibitor 8. 24 h after intraperitoneal
administration of the conjugate (80 µg) or the free drug, the mice were sacrificed, their
peritoneal cavities washed and the amount of compound recovered was determined. The
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 39
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
results show the recovery expressed as a percentage of the administered compound (mean ±
S.D. of n = 3). For comparison, when the drug was collected immediately after
administration, 40% of the drug was recovered. To confirm the specificity of the assay, no
drug was detected in the peritoneal cavity of mice injected with PBS instead of the drug.
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 increased levels of autotaxin and high levels of LPA that are found within the ascites
fluid of patients with ovarian cancer, together with the relatively rapid clearance of low
13
molecular mass compounds from the peritoneal cavity, presents a challenge to provide
continual inhibition of autotaxin in the peritoneal cavity. Our study begins to address these
challenges by preparing a novel conjugated equivalent from a previously described autotaxin
inhibitor series. Our choice of icodextrin as the polymeric support was guided by the
established history of safe and effective clinical use of icodextrin for peritoneal dialysis, and
15
thus its pharmacokinetic properties have been well characterized. The relatively high
molecular mass of icodextrin results in prolonged retention in the peritoneal cavity in patients
and, consequently, it is highly likely that drugs that are covalently conjugated to icodextrin
will be retained within the peritoneal cavity of patients, potentially leading to prolonged
enzyme inhibition when compared with a free drug equivalent. Our new intraperitoneal
delivery approach was shown to improve the pharmacokinetic properties of an autotaxin
inhibitor, as well as increasing its solubility without substantial detrimental effects on its
ability to inhibit autotaxin.
Our previous work evaluating PAMAM dendrimer conjugates of autotaxin inhibitors was not
guided by knowledge of the structure of the autotaxin-drug complex, and thus resulted in an
14
inhibitor conjugate with substantially reduced potency. In this present study, we have made
18
use of the published crystal structure of the complex to guide our design and were able to
1
6
ACS Paragon Plus Environment

Page 17 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
successfully identify an appropriate attachment point on the inhibitor molecule. Conjugate 7
showed minimal reduction in potency compared with the free inhibitor 8 (expressed in terms
of the amount of pharmacophore present). This suggests that conjugation to icodextrin had at
worst a minor affect on the affinity of the pharmacophore for autotaxin. Since we anticipate
administration of an icodextrin-inhibitor conjugate into the peritoneal cavity, such a minor
reduction in potency is not likely to be a substantial limitation on its use.
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
The attachment of the pharmacophore to icodextrin also resulted in two desirable
improvements in physicochemical properties. Firstly, the membrane permeability of the drug
was reduced, potentially increasing the duration of its residence in the peritoneal cavity.
Secondly, the solubility of the drug was also increased. One key goal of this study was to
increase the retention time of the drug in the peritoneal cavity and we have established that
24 hours after intraperitoneal administration, 30% of the inhibitor conjugate was recovered
from the intraperitoneal cavity. In contrast, a negligible amount of the free inhibitor was
recovered. These results provide proof-of-principle that the attachment of autotaxin inhibitors
to icodextrin can improve their retention time in the peritoneal cavity. That the conjugate was
cleared more rapidly from mice than we anticipated, compared with the known clearance of
icodextrin from human subjects, may be explained by the fact that rodents are known to have
an increased amount of
α-amylase present in the peritoneal cavity compared to man, which is
24
known to degrade icodextrin. Indeed, the half-life of icodextrin in rodents is known to be
24
significantly less than in humans.
This presents a potential issue when planning for
oncology preclinical studies, since the relatively rapid clearance of icodextrin from mice
makes it difficult to conduct preclinical pharmacodynamic and xenograft studies with this
species. Consequently, we consider that rodents do not provide a suitable model in which the
pharmacodynamic activity of inhibitor-icodextrin conjugates can be evaluated, and other
model systems, or indeed human clinical trials, may be more appropriate. In contrast, clinical
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 39
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
studies with icodextrin conjugates in human patients seem feasible because of the long dwell
time of icodextrin in the human peritoneal cavity and the existing and safe therapeutic
15
application of this polymer. Indeed, the higher molecular weight fractions of icodextrin that
we have used in these studies are anticipated to be eliminated even more slowly than
unfractionated icodextrin.
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 strategy for intraperitoneal drug delivery that we have described herein should not be
limited to autotaxin inhibitors and may be applied to other drugs where it is desirable to
retain drugs in the peritoneal cavity or to alter their physicochemical properties.
Intraperitoneal administration of chemotherapy is being evaluated for the treatment of
ovarian cancer because it offers the potential for increased efficacy, although toxicity remains
25
an issue with some drugs. Using icodextrin as a basis for drug conjugation may also be
more widely applicable, for example where it is desirable to retain a drug within other body
compartments. The prolonged retention of drugs in the peritoneal cavity, and potentially in
other body compartments, may also allow for a relatively infrequent dosing regime to be
developed.
CONCLUSIONS
In this paper we report the design and synthesis of a novel autotaxin inhibitor-icodextrin
conjugate and its subsequent biological evaluation. To our knowledge, this represents the first
reported use of icodextrin as a suitable polymer for the formation of covalently-bound drug
conjugates. Our novel icodextrin conjugate was found to inhibit autotaxin activity using two
different substrates and two different sources of autotaxin. Furthermore, the conjugate was
found to reduce migration of an ovarian cell line modified to over-express autotaxin.
1
8
ACS Paragon Plus Environment

Page 19 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Conjugation of the inhibitor to icodextrin led to an increase in solubility and a decrease in
membrane permeability when compared with the free drug equivalent. Finally, when the
icodextrin-inhibitor conjugate was administered to the peritoneal cavity of mice, 30% of the
drug was still detected after 24 hours. These observations demonstrate that autotaxin
inhibitor-icodextrin conjugates have significant potential as a new approach for the treatment
of ovarian cancer.
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
EXPERIMENTAL SECTION
Chemistry. Reagents and solvents were obtained from commercial suppliers and were not
further purified before use unless stated. Icodextrin was obtained as EXTRANEAL™
peritoneal dialysis solution (7.5% w/v icodextrin in an electrolyte solution, (Baxter)).
Icodextrin was isolated from Extraneal peritoneal dialysis solution following dialysis against
water for 3 days, changing the solvent 3 times per day, followed by freeze-drying to remove
excess water.
Solvents were removed under reduced pressure with a Büchi rotary evaporator. Any
remaining solvent was removed under high vacuum. The progress of the reactions was
monitored by TLC analysis, using aluminum pre-coated silica gel plates (Merck) visualised
by UV irradiation at a wavelength of 254 nm. Flash column chromatography was carried out
on silica gel 60 (43-60 mesh); columns were slurry packed in the appropriate solvent/solvent
mixture and samples were added as a concentrated solution or pre-absorbed onto silica.
Dialysis was carried out using 20 kDa and 8 kda MWCO (Biotech) dialysis membrane
tubing, where stated.
1
13
H and C nuclear magnetic resonance (NMR) were measured on a Bruker DPX300 Fourier
transform spectrometer, or a Jeol eclipse 400 MHz Fourier transform spectrometer using D6-
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 39
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
DMSO, CDCl
3
, D
2
O, D
6
-Acetone or MeOD solvents. Chemical shifts are quoted in ppm
downfield from TMS; coupling constants (J) are quoted in Hertz (Hz). TMS was defined as 0
1
13
in H NMR and the residual chloroform triplet as 77.10ppm in C NMR. The following
abbreviations were used in analysis; broad (br), singlet (s), doublet (d), triplet (t), quartet (q)
and multiplet (m).
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
Infra-red (IR) spectra were recorded in their natural state on a Thermo Scientific GladiATR
spectrometer or a Perkin Elmer spectrum 100 FT-IR spectrometer and the vibrational
-1
frequencies were recorded in cm .
Mass spectra were obtained from the EPSRC UK National Mass Spectrometry Facility at
Swansea University, using electron spray ionisation (ESI) or atmospheric pressure chemical
ionisation (APCI).
Melting points were measured on a Stuart SMP10 melting point apparatus.
The purity of all compounds was determined by elemental analysis that was performed by the
elemental analysis service at London Metropolitan University, and was greater than 95%.
Thermogravimetric analysis (TGA) was performed using a SDT Q 600 TA instrument, with a
constant nitrogen flow of 100 mL/min, heating rate of 20 °C/min, heating was held stable at
100 °C for 10 min to remove any residual solvent. Results were obtained for greater than 10
mg per sample between 20 and 600 °C.
26
3-(4-Fluorobenzyl)thiazolidine-2,4-dione (1).
A solution of thiazolidine-2,4-dione (2.93 g, 25.00 mmol, 1.50 equiv.) in DMF (anhydrous,
0 mL), under N , was cooled to 0 °C and sodium hydride (60% mineral oil dispersion, 0.92
g, 23.00 mmol, 1.35 equiv.) and a solution of 4-fluorobenzyl chloride (2.04 mL, 17.00 mmol,
5
2
2
0
ACS Paragon Plus Environment

Page 21 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1.00 equiv.) in DMF (anhydrous, 10 mL) were added. The reaction mixture was allowed to
warm to room temperature over 4 h. The reaction mixture was then poured over crushed ice
(100 mL), hexane (50 mL) was then added, and the product allowed to crystallise overnight
at 4 °C. 3-(4-Fluorobenzyl)thiazolidine-2,4-dione was then collected via vacuum filtration 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
1
yield 2.87 g, (12.75 mmol, 75%) as colourless needles; mp 78–80 °C. H NMR (400 MHz,
CDCl
3
)
δ
7.42–7.38 (m, 2H) 7.03–6.99 (m, 2H), 4.73 (s, 2H), 3.95 (s, 2H). HRMS (APCI)
+
m/z calculated for C10
H
O
9 2
NFS [M + H] : 226.0333. Found 226.0333.
Methyl 4-((4-formyl-2-hydroxyphenoxy)methyl)benzoate (2).
3
,4-Dihydroxybenzaldehyde (0.15 g, 1.10 mmol, 1.0 equiv.) was dissolved in DMF (5 mL)
and sodium bicarbonate (0.14 g, 1.65 mmol, 1.5 equiv.), methyl-4-(bromomethyl)-benzoate
0.50 g, 2.20 mmol, 2.0 equiv.) and sodium iodide (0.05 g, 0.33 mmol, 0.3 equiv.) were
(
added. The reaction mixture was heated at 40 °C for 24 h. The reaction was then quenched
with 10% aqueous HCl (10 mL), extracted with ethyl acetate (3 × 10 mL), washed with brine
4
(3 × 10 mL), dried over anhydrous MgSO and the solvent was evaporated under reduced
pressure to obtain a light brown oil. The crude product was purified by column
chromatography eluting with ethyl acetate/hexane (1:4) to yield the target compound as a
1
colourless solid (0.22 g, 0.77 mmol, 70%); mp 151–153 °C. H NMR (300 MHz, [D
6
]
DMSO)
.54 (d, J = 8.2 Hz, 1H), 7.43 (s, 1H), 7.26 (d, J = 8.2 Hz, 1H), 5.31 (s, 2H), 3.85 (s, 3H). C
NMR (75 MHz, [D ]DMSO): 191.21, 166.30, 152.23, 148.40, 145.97, 129.14, 128.91,
δ 13.02 (br. s, 1H), 9.84 (s, 1H), 7.98 (d, J = 8.0 Hz, 2H), 7.57 (d, J = 8.0 Hz, 2H),
13
7
6
ꢀ
27.98, 126.39, 124.64, 115.59, 114.37, 62.43, 52.12.
ꢁ= 3672, 2902, 1709, 1672, 1606,
1
1
−
1
+
15 5
582 cm . HRMS (ESI) m/z calculated for C16H O [M + H] : 287.0914. Found 287.0916.
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 39
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
Elemental analysis calculated (%) for C16
5.05.
14 5
H O requires: C 67.13; H 4.93, found: C 67.04; H
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
X-ray Crystallography for 2.
A suitable crystal was obtained by recrystallization from EtOAc/hexane. Diffraction data
26152 total reflections with Rint = 0.038) were collected on a Bruker APEX 2 CCD
diffractometer at T = 150 K using graphite-monochromated Mo-K radiation. Crystal data:
, Mr = 286.27 g mol , orthorhombic, Pbca, a = 7.6024(6), b = 16.1869(13), c =
(
α
–
1
16 14 5
C H O
3
2
2
21.5409(17) Å, V = 2650.8(4) Å , Z = 8, R[F > 2σ(F )] = 0.069 for 2900 reflections with I >
2
2
σ(I), wR(F ) = 0.143 for all 3422 independent data, GOF = 1.19, 253 refined parameters.
CCDC 1559541 contains the supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. Full details of the refinement and atomic positons are
found in the supplementary information.
Methyl 4-((4-formyl-2-(2-(2-hydroxyethoxy)ethoxy)phenoxy)methyl)benzoate (3 ester).
Methyl 4-((4-formyl-2-hydroxyphenoxy)methyl)benzoate, 2, (0.18 g, 0.63 mmol, 1.0 equiv.)
was stirred in DMF (5 mL) and potassium carbonate (0.17 g, 1.25 mmol, 2.0 equiv.),
potassium iodide (0.10 g, 0.63 mmol, 1.0 equiv.) and 2-(2-chloro-ethoxy)-ethanol (0.2 mL,
1
.88 mmol, 2.9 equiv.) were added and the reaction heated to 70 °C for 18 h. The solution
was then diluted with water (50 mL) and extracted with ethyl acetate (3 × 20 mL) and
washed with water (3 × 20 mL) and brine (3 × 20 mL), and dried over anhydrous MgSO
4
.
2
2
ACS Paragon Plus Environment

Page 23 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The solvent was then removed by evaporation under reduced pressure to yield the crude
product which was purified via column chromatography (hexane/EtOAc (1:3)) to yield the
1
target compound as a pale yellow powder (0.14 g, 0.36 mmol, 58%); mp 56–59 °C. H NMR
(400 MHz, CDCl3)
δ 10.46 (s, 1H), 8.25 (d, J = 7.9 Hz, 2H), 7.57, (d, J = 7.9 Hz, 2H), 7.53
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
(d J = 1.7 Hz, 1H), 7.43 (dd, J = 7.9, 1.7 Hz, 1H), 6.88 (d, J = 7.9 Hz, 1H), 4.77 (s, 2H),
13
3.52–3.49 (m, 2H), 3.12–3.09 (m, 2H), 3.07 (s, 3H), 2.78–2.74 (m, 4H).
C NMR (100
MHz, CDCl
3
) δ
190.97, 166.84, 153.67, 149.36, 141.33, 130.63, 130.08, 130.00, 127.08,
126.93, 112.91, 111.50, 72.79, 70.29, 69.53, 68.84, 61.91, 52.31.
ꢂ̅
ꢁ
= 3671, 3504, 3442, 3407,
−1
+
2
908, 1722, 1685, 1584 cm . HRMS (ESI) m/z calculated for C20
H
23
O
7
[M + H] : 375.1438.
Found 375.1441.
4-((4-Formyl-2-(2-(2-hydroxyethoxy)ethoxy)phenoxy)methyl)benzoic acid (3).
Methyl 4-((4-formyl-2-(2-(2-hydroxyethoxy)ethoxy)phenoxy)methyl)benzoate (0.13 g, 0.34
mmol) was dissolved in DMSO (1 mL) and water (5 mL), and the reaction heated to 70 °C
for 2 h. 1 M NaOH (aq.,1 mL) was added and reaction was heated at 70 °C until the solution
became clear. The reaction mixture was diluted with water (10 mL) and acidified to pH 2
with 1 M HCl (aq.), and the resulting precipitate was collected by filtration to yield the target
1
compound as a pale yellow powder (0.12 g, 0.33 mmol, 98%); mp 116–119 °C. H NMR
3
(400 MHz, CDCl ) δ 10.47 (s, 1H), 8.29 (d, J = 7.9 Hz, 2H), 7.60 (d, J = 7.9 Hz, 2H), 7.55 (d
J = 1.7 Hz, 1H), 7.46 (dd, J = 7.9, 1.7 Hz, 1H), 6.89 (d, J 7.9 Hz, 1H), 4.80 (s, 2H), 3.55–3.52
13
(
m, 2H), 3.15–3.12 (m, 2H), 2.89–2.81 (m, 4H). C NMR (100 MHz, CDCl3)
δ
191.06,
1
7
70.60, 153.64, 149.34, 142.09, 130.68, 130.11, 129.32, 127.02, 126.98, 112.88, 111.45,
−1
2.79, 70.27, 69.54, 68.78, 61.88. ꢂ̅ꢁ = 3347, 2861, 1678, 1584, 1508 cm . HRMS (ESI) m/z
+
calculated for C19
H
21
O
7
[M + H] : 361.1282. Found 361.1286.
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 39
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
(
Z)-4-((4-((3-(4-fluorobenzyl)-2,4-dioxothiazolidin-5-ylidene)methyl)-2-(2-(2-
hydroxyethoxy)ethoxy)phenoxy)methyl)benzoic acid (4).
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
-(4-Fluorobenzyl)-1,3-thiazolane-2,4-dione, 1, (0.51 g, 2.25 mmol, 1.0 equiv.) was
dissolved in ethanol (10 mL). Piperidine (0.22 mL, 2.25 mmol, 1.0 equiv.) and 4-((4-formyl-
2-(2-(2-hydroxyethoxy)ethoxy)phenoxy)methyl)benzoic acid, 3, (0.81 g, 2.25 mmol, 1.0
equiv.) were added and the solution was heated at reflux for 24 h. The mixture was cooled to
room temperature and the resulting precipitate was collected via filtration to yield the target
1
compound as a yellow powder (1.19 g, 2.09 mmol, 93%); mp 188
MHz, [D ]DMSO)
–191
°
C. H NMR (400
δ
7.97 (d, J = 8.2 Hz, 2H), 7.86 (s, 1H), 7.57 (d, J = 8.2 Hz, 2H), 7.38
.33 (m, 2H), 7.25 (s, 1H), 7.19
–
6
7
–7.13 (m, 4H), 5.27 (s, 2H), 4.79 (s, 2H), 4.67 (br. s, 1H),
13
4.18
–
4.16 (m, 2H), 3.80
–
3.78 (m, 2H), 3.53 (br. s, 4H). C NMR (100 MHz, [D
6
]DMSO)
δ
1
1
67.43 (s), 167.18 (s), 165.58 (s), 161.76 (d, J(C,F) = 244 Hz), 150.09 (s), 148.63 (s), 141.78
4
3
(s), 133.79 (s), 131.84 (d, J(C,F) = 3.5 Hz), 130.33 (s), 130.04 (d, J(C,F) = 8.1 Hz), 129.58
2
(s), 127.30 (s), 126.18 (s), 124.07 (s), 118.33 (s), 115.63 (s)*, 115.52 (d, J(C,F) = 21.9 Hz),
1
14.27 (s), 72.69 (s), 69.42 (s), 68.86 (s), 68.46 (s), 60.37 (s), 43.93 (s). *This peak is
obscured at 25 C but can be observed when the temperature is increased to 70 C. = 3672,
8
3448, 2972,2902, 1729, 1678, 1509. HRMS (ESI) m/z (MH ) calcd for C29 26FNO S
°
° ꢂ̅ꢁ
+
H
5
68.1436, found 568.1430. Elemental analysis calculated (%) for C29
1.37; H 4.62; N 2.47%, found: C 61.15; H 4.73; N 2.61.
8
H26FNO S requires: C
6
2
4
ACS Paragon Plus Environment

Page 25 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Triisopropylsilyl-(Z)-4-((4-((3-(4-fluorobenzyl)-2,4-dioxothiazolidin-5-ylidene)methyl)-
2-(2-(2-hydroxyethoxy)ethoxy)phenoxy)methyl)benzoate (5).
(Z)-4-((4-((3-(4-Fluorobenzyl)-2,4-dioxothiazolidin-5-ylidene)methyl)-2-(2-(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
hydroxyethoxy)ethoxy)phenoxy)methyl)benzoic acid, 4, (5.96 g, 10.5 mmol, 1.0 equiv.) and
triisopropylsilyl chloride (3.19 mL, 14.91 mmol, 1.42 equiv.) were stirred under nitrogen in
anhydrous THF (200 mL). Triethylamine (1.61 mL, 11.55 mmol, 1.1 equiv.) was added and
the reaction mixture was stirred for 15 min at room temperature. The resulting mixture was
diluted with diethyl ether (200 mL) and then evaporated to yield the crude product as an oily
yellow solid. The crude material was purified by column chromatography (hexane/EtOAc
(1:1)) to yield the target compound as a bright yellow powder (5.78 g, 7.98 mmol, 76%). Mp
1
1
8
1
07-109 °C. H NMR (400 MHz, CDCl
3
)
δ
8.07 (d, J = 8.3, 2H), 7.80 (s, 1H), 7.51 (d, J =
.3, 2H), 7.46 –7.41 (m, 2H), 7.07–7.06 (m, 2H), 7.03–6.98 (m, 2H), 6.92 (d, J = 9.0 Hz,
H), 5.26 (s, 2H), 4.85 (s, 2H), 4.25–4.22 (m, 2H), 3.94 (dd, J = 4.0, 5.3 Hz, 2H), 3.74–3.69
13
(
m, 4H), 2.29 (br. s, 1H), 1.42 (septet, J = 7.5 Hz, 3H), 1.14 (d, J = 7.5 Hz, 18H). C NMR
1
(100 MHz, CDCl
3
)
δ
167.89 (s), 166.26 (s), 166.03 (s), 162.75 (d, J (C,F) = 247 Hz), 150.66
4
(s), 149.19 (s), 141.52 (s), 134.15 (s), 131.39 (s), 131.19 (d, J(C,F) = 3.0 Hz), 131.02 (d,
3
2
J(C,F) = 8.1 Hz), 130.67 (s), 126.90 (s), 126.79 (s), 125.24 (s), 119.10 (s), 115.76 (d, J(C,F)
=
21.4 Hz), 115.27 (s), 114.14 (s), 72.79 (s), 70.36 (s), 69.57 (s), 69.05 (s), 61.92 (s), 44.59
−1
(s), 17.99 (s), 12.18 (s).
ꢂ̅
= 3536,2944, 2868, 1728, 1669, 1510 cm . HRMS (ESI) m/z
+
calculated for C38
H
47FNO
8
SSi [M + H] : 724.2770. Found 724.2774. Elemental analysis
calculated (%) for C H FNO SSi: C 63.05, H 6.41, N 1.93. Found: C 62.96, H 6.53, N 2.04.
38 46
8
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 39
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
Triisopropylsilyl (Z)-4-((4-((3-(4-fluorobenzyl)-2,4-dioxothiazolidin-5-ylidene)methyl)-2-
(2-(2-(tosyloxy)ethoxy)ethoxy)phenoxy)methyl)benzoate (6).
Triisopropylsilyl (Z)-4-((4-((3-(4-fluorobenzyl)-2,4-dioxothiazolidin-5-ylidene)methyl)-2-(2-
(2-hydroxyethoxy)ethoxy)phenoxy)methyl)benzoate, 5, (0.40 g, 0.55 mmol, 1.0 equiv.), was
stirred in dichloromethane (12 mL). Tosyl chloride (0.21 g, 1.10 mmol, 2.0 equiv.) and
triethylamine (0.15 mL, 1.10 mmol, 2.0 equiv.) were added and the reaction mixture was
stirred at room temperature for 36 h. The reaction mixture was then evaporated to yield a
yellow residue which was purified by column chromatography (50:50 ethyl acetate/petroleum
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
ether (40-60)). Evaporation of the desired fractions afforded the target compound as a yellow
1
crystalline solid (0.30 g, 0.34 mmol, 61%); mp 72–75 °C. H NMR (400 MHz, [D
6
]acetone)
δ
8
7
2
1
1
.09 (d, J = 8.2 Hz, 2H), 7.85 (s, 1H), 7.77 (d, J = 8.2 Hz, 2H), 7.65 (d, J = 8.2 Hz, 2H),
.47–7.43 (m, 2H), 7.38 (d, J = 8.2 Hz, 2H), 7.24–7.21 (m, 3H), 7.14–7.09 (m, 2H), 5.34 (s,
H), 4.88 (s, 2H), 4.22–4.17 (m, 4H), 3.80–3.77 (m, 2H), 3.84–3.82 (m, 2H), 2.37 (s, 3H),
1
.44 (septet, J = 7.4 Hz, 3H), 1.15 (d, J = 7.4 Hz, 18H). H NMR (100 MHz, [D ]acetone)
δ
6
1
68.15 (s), 166.59 (s), 166.39 (s), 163.26 (d, J(C,F) = 247 Hz), 151.52 (s), 150.22 (s),
4
145.66 (s), 143.43 (s), 134.31 (s), 134.2 (s), 132.99 (d, J(C,F) = 2.9 Hz), 131.74 (s), 131.36
3
(
d, J(C,F) = 8.4 Hz), 131.03 (s), 130.73 (s), 128.69 (s), 128.21 (s), 127.67 (s), 125.29 (s),
2
1
19.87 (s), 116.32 (s), 116.17 (d, J(C,F) = 21.3 Hz), 115.41 (s), 70.74 (s), 70.67 (s), 70.21
(s), 69.73 (s), 69.59 (s), 44.93 (s), 21.50 (s), 18.20 (s), 12.76 (s).
ꢂ̅
= 2926, 2867, 2667, 1731,
−1
+
1
681, 1592 cm . HRMS (ESI) m/z calculated for C45
H56FN
2
O
10
S
2
Si [M + NH
4
] : 896.3124.
Found 896.3123. Elemental analysis calculated (%) for C H FNO SSi: C 61.55, H 5.97, N
45 52
8
1.60. Found: C 61.44, H 5.79, N 1.71.
2
6
ACS Paragon Plus Environment

Page 27 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis of conjugate (7).
Icodextrin (0.25 g, approx. 0.16 mmol, 1.00 equiv.) was suspended in DMAc (anhydrous, 5
mL) with LiCl (anhydrous, 0.25 g, 5.90 mmol, 3.85 equiv.) under nitrogen, and heated to 132
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
°C with stirring until dissolved. The solution was then cooled to room temperature and
sodium hydride (60% mineral oil dispersion, 0.04 g, 1.00 mmol, 0.67 equiv.) was added and
the reaction mixture stirred for 4 h. Triisopropylsilyl-(Z)-4-((4-((3-(4-fluorobenzyl)-2,4-
dioxothiazolidin-5-ylidene)methyl)-2-(2-(2-
(tosyloxy)ethoxy)ethoxy)phenoxy)methyl)benzoate, 6, (0.45 g, 0.51 mmol, 0.33 equiv.) was
dissolved in DMAc (anhydrous, 1 mL) and added dropwise to the reaction mixture. The
reaction was monitored by TLC (hexane/EtOAc (4:2)) and stirred at room temperature for 4
h. The reaction mixture was then transferred into a dialysis membrane (8000 da. MWCO) and
dialyzed against methanol (1 L), with the solvent being changed 3 times per day for 3 days.
The contents of the dialysis membrane were then evaporated to yield an off-white solid
which was precipitated from methanol (3 mL) and the liquid carefully decanted, and the
retained solid washed twice more with methanol and once more with dichloromethane (3
mL). The remaining solid was then dried under vacuum to yield conjugate 7 directly, no
additional step was required to remove the TIPS protecting group. (0.40 g, 43%). ꢁꢂ̅ꢁ= 3362,
−
1
1507, 1146, 2361, 2342, 1078, 1019 cm .
Degree of substitution of IDX-conjugate.
The degree of substitution of was determined by measuring the inherent fluorescence of the
inhibitor molecule, ( Ex = 360 nm, Em = 440 nm) using a Synergy 2 microplate reader
BioTek instruments, Inc.). A calibration curve for fluorescence against concentration was
λ
λ
(
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 39
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 by dissolving a known concentration of the molecule in DMSO and creating a
dilution series with PBS (pH 7.4). Fluorescence was then measured for the test compound
and compared with the standard curve and the inhibitor. The amount of inhibitor present
could then be determined as a mol/w or w/w ratio.
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
Biology.
Cell culture.
2
Human ovarian cancer cells were maintained at 37 °C and 5% CO in RPMI supplemented
−
1
with 10% FCS, Penicillin/Streptomycin (50 U mL ) and glutamine (2 mM). 3E3 ovarian
8
cancer cells were previously engineered to express autotaxin while 3V3 cells were
transfected with the empty vector.
Purification of autotaxin.
3
E3 cells were seeded in a T75 flask and when 50% confluent the medium was replaced with
serum free medium (10 mL). The next day the supernatant was collected and centrifuged
3000 rpm, 4 °C, 15 min). The supernatant was applied to a HiTrap Con A sepharose column
(GE healthcare) and recirculated for 1 h. The column was washed with 20 mL tris buffered
saline (TBS; 20mM tris, 0.5 M NaCl, pH 7.4) and autotaxin eluted with 2 mL
(
α-
methylmannoside (0.5M) in TBS overnight. The eluate was dialysed against TBS and used as
the source of autotaxin for the bis pNPP assay.
FS-3 assay.
FS-3 is a fluorescence-quenched analogue of lysophosphatidylcholine (Echelon Biosciences,
Inc. Salt Lake City, UT). Compounds to be tested were prepared as stock solutions in DMSO
2
8
ACS Paragon Plus Environment

Page 29 of 39
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(10
ꢀ
M). The drug (10
ꢀ
L) was incubated with 80
ꢀ
L reaction buffer containing autotaxin for
10 min at 37 °C before addition of 10
ꢀL FS-3 substrate. The rate of increase in fluorescence
was measured at 37 °C every minute for 30 min using a Synergy2 multi-mode microplate
reader (BioTek instruments, Inc.) (
λ=528 nm). Data was analysed using GraphPad Prism
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
software.
Bis-pNPP Assay.
Inhibition of autotaxin activity was measured using the autotaxin substrate bis-para-
nitrophenylphosphate (bis-pNPP). 20 L 5X reaction buffer (250 mM Tris HCl, 25 mM KCl,
mM CaCl , 5 mM MgCl .6H O, NaCl 700 mM, pH 7.8), 10 L bis-pNPP (1 mM), 20
inhibitor was added to each well of an opaque 96 well plate, followed by the addition of 50
L autotaxin purified from 3E3 cells. After incubation (37 °C, 4 h) A405 was determined and
ꢀ
5
ꢀ
ꢀL
2
2
2
ꢀ
data was analysed using graph pad prism software to fit a 4 parameter Hill equation.
Wound healing assay.
3V5 and 3E3 cells (100,000 cells/ml) were plated in 6 well plates in RPMI supplemented
with FCS (10%). When the cells had reached confluence the medium was removed and a
plastic pipette tip used to produce a clean wound area. The wells were washed with PBS and
serum free medium added containing lysophosphatidyl choline (0.5
inhibitor or solvent. The cells were viewed by phase contrast microscopy immediately after
wounding. After 16 h the cells were stained and fixed with methylene blue (200 L, Sigma
M9140; 0.5% w/v in 50% H O, 50% EtOH). Images of the wound were captured by light
ꢀM), and either the
ꢀ
2
microscopy and wound closure measured (µm) using imageJ software. The area of a section
of the wound pre- and post-migration was then measured and the percentage wound closure
14
was calculated as described.
2
9
ACS Paragon Plus Environment