Water-soluble prodrugs of an Aurora kinase inhibitor

Article abstract of DOI:10.1016/j.bmcl.2009.01.043

Compound 1 (SNS-314) is a potent and selective Aurora kinase inhibitor that is currently in clinical trials in patients with advanced solid tumors. This communication describes the synthesis of prodrug derivatives of 1 with improved aqueous solubility profiles. In particular, phosphonooxymethyl-derived prodrug 2g has significantly enhanced solubility and is converted to the biologically active parent (1) following iv as well as po administration to rodents.

Full text of DOI:10.1016/j.bmcl.2009.01.043

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1409–1412
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Water-soluble prodrugs of an Aurora kinase inhibitor
*
Johan D. Oslob , Stacey A. Heumann, Chul H. Yu, Darin A. Allen, Subramanian Baskaran, Minna Bui,
Erlie Delarosa, Amy D. Fung, Ahmad Hashash, Jonathan Hau, Sheryl Ivy, Jeffrey W. Jacobs, Willard Lew,
Jack Maung, Robert S. McDowell, Sean Ritchie, Michael J. Romanowski, Jeffrey A. Silverman, Wenjin Yang,
Min Zhong, Tarra Fuchs-Knotts
Sunesis Pharmaceuticals, Inc., 395 Oyster Point Boulevard, South San Francisco, CA 94080, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Compound 1 (SNS-314) is a potent and selective Aurora kinase inhibitor that is currently in clinical trials
in patients with advanced solid tumors. This communication describes the synthesis of prodrug deriva-
tives of 1 with improved aqueous solubility profiles. In particular, phosphonooxymethyl-derived prodrug
2g has significantly enhanced solubility and is converted to the biologically active parent (1) following iv
as well as po administration to rodents.
Received 24 November 2008
Revised 11 January 2009
Accepted 13 January 2009
Available online 19 January 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Aurora
Kinase
Inhibitor
Prodrugs
The Aurora serine/threonine kinases have emerged as promis-
ing anti-cancer targets for small-molecule intervention.1–6 As a re-
sult, a number of Aurora kinase inhibitors have been disclosed and
some have progressed into clinical trials.1–10 Recently, we reported
the discovery of a potent and selective Aurora kinase inhibitor,
compound 1 (SNS-314, Fig. 1).11 This compound has a compelling
pre-clinical profile and is currently in clinical trials in patients with
advanced solid tumors.12 An issue commonly encountered when
developing kinase inhibitors is low aqueous solubility. While we
found that CaptisolÒ can successfully be used to provide more than
adequate aqueous solubility of 1 (Fig. 1) for parenteral administra-
tion, we sought alternative means in a parallel effort to identify
Aurora kinase inhibitors with improved aqueous solubility pro-
files.13 Initially, analogues of compound 1 containing solubilizing
functionalities were explored. Unfortunately, efforts to improve
solubility often compromised other pharmaceutical properties.
For example, while amine-containing derivatives of 1 had signifi-
cantly improved aqueous solubility at lower pH, they also tended
to be heavily effluxed and have less favorable in vivo pharmacoki-
netic profiles.11 Furthermore, of the analogues evaluated in vivo,
compound 1 was found to display a superior activity profile. We
therefore next considered a prodrug approach14,15 where com-
pound 1 would be directly derivatized with a bio-labile moiety de-
signed to improve solubility. In addition to conferring improved
solubility, the prodrug moiety would also have to be chemically
stable in aqueous buffer yet labile in vivo to liberate the biologi-
cally active parent drug (1). In this letter, we describe our efforts
directed towards these goals.
The synthesis of acyl-oxymethylene derived prodrugs of 1 is
outlined in Schemes 1 and 2. Heating a mixture of isobutyryl chlo-
ride (3) and paraformaldehyde in the presence of zinc chloride
yielded chloromethyl ester 4 in 56% yield (Scheme 1). Alkylation
was achieved by treating the free base of compound 1 with sodium
hydride followed by addition of 4 to afford prodrug 2b in low yield
(13%).16 Similarly, carboxylic acid 5 was converted to the corre-
sponding iodomethyl derivative 6 (48%) followed by alkylation of
* Corresponding author. Present address: Ardelyx Inc., 34175 Ardenwood Blvd.,
Fremont, CA 94555, USA. Tel.: +1 510 745 1715.
E-mail address: joslob@ardelyx.com (J.D. Oslob).
Figure 1. Thermodynamic solubility profile of compound 1.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.043

1410
J. D. Oslob et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1409–1412
(Fig. 2). Thus, the use of acyl-oxymethylene derived prodrugs 2a–
c was initially considered a potential approach to improve solubil-
ity. Further it was anticipated that the esters would undergo enzy-
matic hydrolysis in vivo to generate the biologically active parent
(compound 1) along with formaldehyde (cf. Fig. 2).14,15 Although
this first set of prodrugs was shown to be stable in aqueous buffer
and DMSO (Table 1) they only showed modest improvements in
their solubility profiles relative to the parent drug (2a–c vs 1, Table
1). Our earlier studies indicated that incorporation of amines could
significantly improve aqueous solubility of compounds in this ser-
ies.11 Therefore, we next focused on amine-containing acyl-oxy-
methylene derived prodrugs. Our initial attempts to explore
these derivatives were unsuccessful due to instability of the pro-
drug. For example, although we were able to isolate prodrugs 2d
and 2e, they were readily converted back to the parent in aqueous
buffer and organic media (Table 1). However, we noticed that
amine-containing prodrugs where the basic nitrogen was extended
further from the ester-moiety appeared to be more stable (e.g., 2e
is more stable than 2d, Table 1). Consistently, extending the basic
amine in 2e one atom further from the ester functionality did pro-
vide an amine-containing prodrug (2f) that showed no signs of
decomposition over an eight hour time period (Table 1). As ex-
pected, this derivative was also found to have significantly en-
hanced aqueous solubility over the parent compound at a low pH
(2f vs 1, Table 1). Having achieved noteworthy improvements in
solubility with a charged prodrug moiety, we next examined a
phosphate-derived prodrug, compound 2g. The prodrug moiety
in 2g is expected to be doubly charged at near-neutral pH and
has successfully been used to improve the solubility of clinically
used iv drugs, including Fosphenytoin19–21 and Fospropofol.22–24
Presumably, the phosphate moiety is cleaved in vivo by alkaline
phosphatases to release the parent and formaldehyde (cf.
Fig. 2).14,15 In addition, phosphate-derived prodrugs have also been
used to enhance the oral bioavailability (oral %F) of marketed drugs
(e.g., the HIV-1 protease inhibitor LexivaÒ).15,25,26 Not surprisingly,
Scheme 1. Reagents and conditions: (a) CH2O, ZnCl2, 90 °C, 56%; (b) 1 (free base),
NaH, DMF, rt, 13%.
Scheme 2. Reagents and conditions: (a) i—aqueous NaHCO3, Bu4NHSO4, DCM, rt;
ii—ClCH2OSO2Cl, DCM, 0–25 °C, 56%; (b) NaI, acetone, rt, 86%; (c) 1 (free base), NaH,
DMF, rt, 36%; (d) 2 M HCl/dioxane, 91%.
compound 1 (36%, Scheme 2). Removal of the Boc-group under
acidic conditions was achieved to give prodrug 2f in 91% yield.
Phosphonooxymethyl-derived prodrug 2g was prepared according
to Scheme 3. Reaction of compound 1 with di-tert-butyl chloro-
methyl phosphate (7)17 under basic conditions followed by de-pro-
tection employing a mixture of water and acetic acid afforded
phosphoric acid 2g in 16% overall yield.
According to the SAR reported earlier, the 2-aminothiazole-urea
portion in 1 is critical for good biological activity.11 However, bi-
aryl urea motifs of this type are notorious for conferring poor sol-
ubility profiles.18 We therefore hypothesized that we could im-
prove solubility by masking the urea in 1 with a prodrug moiety
that would remove its bi-dentate hydrogen bonding capability
Table 1
In vitro stabilities and aqueous solubilities of prodrugs 2a–g
Compound
DMSOa
(% remaining
at 8 h)
Buffer, pH 7.4a
(% remaining
at 8 h)
Aqueous
solubilityb
[mg/mL]
1
0.014 (pH 3.2)
0.12 (pH 2.5)
0.17 (pH 3.2)
0.28 (pH 3.2)
NT
2a
2b
2c
2d
2c
2f
2g
100
NT
100
0
84
100
100
100
NT
100
0
52
100
100
NT
2.5 (pH 3.5)
4.7 (pH 7.8)
a
In vitro stabilities as determined by LC/MS after 8 h at room temperature.
Thermodynamic solubilities. Values are typically means of two measurements,
b
Scheme 3. Reagents and conditions: (a) 1 (free base), NaH, DMA, rt, 22%; (b) AcOH/
H2O (4:1), 65 °C, 72%.
variation < 30%. See Ref. 13 for protocol.
Figure 2. Prodrug strategy employed to improve the solubility profile of compound 1 involved derivatization of the bi-aryl urea portion of the molecule. Prodrugs 2a–g were
prepared and evaluated. In vivo cleavage of the prodrug moieties (R) was expected to liberate the biologically active parent (compound 1) along with formaldehyde.

J. D. Oslob et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1409–1412
1411
compound 2g has significantly improved aqueous solubility at
near-neutral pH and is also stable in organic media as well as in
buffer (Table 1).
Model studies conducted in vitro indicated that the selected
prodrug moieties are cleaved as hypothesized to release the biolog-
ically active parent compound upon exposure to plasma and S9
fractions (human and rat, data not shown). Thus, having identified
a series of prodrugs with overall favorable in vitro profiles, we
decided to determine their preliminary in vivo pharmacokinetic
(PK) profiles. These studies, conducted in CD1 mice, indicated that
only compound 2g displayed satisfactory in vivo properties (data
not shown) while the other prodrugs yielded low plasma concen-
trations of the parent. For example, the bulky pivaloyl ester deriv-
ative 2c was only slowly converted to 1 following intravenous (iv)
administration and also yielded low levels of the parent drug fol-
lowing oral delivery (po). Although cleaved somewhat faster, ace-
tate-derived ester 2a also resulted in disappointingly low
exposure levels of the parent drug. These results combined with
only modest improvements in solubility led us to discontinue
examination of the acyl-oxymethylene derived prodrugs. While
the amine-containing ester 2f did show quite promising solubility
properties (Table 1), the mouse PK study indicated that neither the
prodrug nor the parent could be detected in systemic circulation
following iv or oral administrations. Thus, only 2g was found to
have preliminary in vitro and in vivo properties warranting further
studies. Therefore, it was decided to focus the ensuing efforts on
this compound.
Figure 3. Rat iv pharmacokinetic profiles of prodrug 2g and compound 1 formed
in vivo from 2g. The profiles were determined in male Sprauge–Dawley rats (n = 3)
and the dose of prodrug 2g was equivalent to 5 mg/kg of compound 1.
First, the thermodynamic solubility profile of 2g was studied in
more detail by examining the effects of the counter-ion. Not sur-
prisingly, the bis-salts were shown to be more soluble than the cor-
responding mono-salts (Table 2). We also found that the solubility
is dependent on the identity of the counter-ion. In particular, the
use of choline achieves an aqueous solubility of 9.4 mg/mL (Table
2). This is a significant (>500-fold) improvement over the parent
drug and provides levels of solubility that could reduce or elimi-
nate the need to use a solubilizing excipient such as CaptisolÒ for
iv-formulations.
Figure 4. Comparison of the rat iv pharmacokinetic profile of compound 1 formed
in vivo from 2g to the profile of an equivalent iv dose of compound 1. The profiles
were determined in male Sprauge–Dawley rats (n = 3).
Having narrowed the selection to one prodrug, we also decided
to conduct additional in vivo pharmacokinetic studies, this time in
rats. First, the iv PK profile of compound 1 derived in vivo from pro-
drug 2g was compared to the profile of 2g itself (Fig. 3). The antic-
ipated15 conversion of the prodrug (t1/2 = 0.5 0.1 h) to the parent
was observed and resulted in significant levels of the liberated par-
ent 1 (t1/2 = 2.1 0.3 h). In fact, the PK profile of compound 1 de-
rived following iv administration of 2g is similar to the profile
from an equivalent iv dose of the parent itself (Fig. 4). The exposure
to 1 generated from 2g is 91% of that achieved from dosing with 1
(AUCINF = 1910 230 ng h/mL vs 2090 430 ng h/mL).
We also studied the rat PK profile following oral administration
of 2g (Fig. 5). While modest levels of the parent (1) were detected
in plasma following oral delivery of 2g (dose equivalent to 10 mg/
kg of 1), the resulting AUCINF value is somewhat lower than the va-
lue obtained when an equivalent dose of the parent (1) is adminis-
Figure 5. Comparison of the rat pharmacokinetic profile of compound 1 formed
in vivo from 2g following oral delivery to the profile of an equivalent oral dose of
compound 1. The profiles were determined in male Sprauge–Dawley rats (n = 3).
Table 2
Effect of counter-ion on solubility of prodrug 2g
tered (2200 1200 ng h/mL vs 3650 1700 ng h/mL). The
corresponding oral bioavailabilities with respect to compound 1
are 44% and 72% following oral dosing of 2g and 1, respectively.
In contrast to iv administration of 2g (see above), following oral
delivery, no prodrug could be detected in the systemic circulation
consistent with a low permeability profile of 2g as determined in
the MDCK cell line (<1.0 Â 10À6 cm/s). Presumably, alkaline phos-
phatases, present in the brush border of the gut, convert the pro-
drug to compound 1 prior to absorption.27
Counter-ion
Salt type
Aqueous solubilitya [mg/mL]
Sodium
Sodium
Mono
Bis
0.51
4.7
0.86
6.0
2.4
9.4
Potassium
Potassium
Choline
Mono
Bis
Mono
Bis
Choline
a
Thermodynamic solubilities. Values are typically means of two measurements,
variation < 20%. See Ref. 13 for protocol.

1412
J. D. Oslob et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1409–1412
Jacobs, J. W.; Lam, J.; Lawrence, C. E.; McDowell, R. S.; Nannini, M. A.; Shen, W.;
In conclusion, we have shown that the aqueous solubility pro-
Silverman, J. A.; Sopko, M. M.; Tangonan, B. T.; Teague, J.; Yoburn, J. C.; Yu, C. H.;
Zhong, M.; Zimmerman, K. M.; O’Brien, T.; Lew, W. Bioorg. Med. Chem. Lett.
2008, 18, 4880.
file of our clinical stage Aurora kinase inhibitor SNS-314 (1) can
be significantly improved by derivatizing the 2-aminothiazole-
urea moiety with bio-labile prodrug appendages. In particular,
phosphonooxymethyl-derived prodrug 2g displays high aqueous
solubility and is stable in buffer as well as organic media. In addi-
tion, it is converted to the biologically active parent (1) upon iv and
oral deliveries to rodents. The identification of 2g also suggests
new avenues for further explorations. For example, substitutions
on the oxymethylene unit of 2g as well phosphate ester modifica-
tions may yield additional enhancements in pharmaceutical
properties.14,15
12. Taverna, P.; Hogan, J.; Kumer, J.; Arbitrario, J.; Hoch, U.; Silverman, J.; Howlett,
A. Ann. Oncol. 2007, 18, 204.
13. General procedure for determination of thermodynamic solubilities: The test
compound was accurately weighed out in a suitable size screw cap glass vial
followed by the addition of the appropriate media. The resulting slurry was
sonicated for 20 min at room temperature, 10 min at 50 °C, and then another
10 min at room temperature. The slurry was then agitated for a minimum of
two days at room temperature at the end of which they were inspected under
polarized light microscope (Olympus BX51) for crystallinity of the residual
solid. Additional time for equilibration was allowed if crystallinity was not
observed. On the day of analysis, the slurry was aliquoted into 1.5 mL
propylene centrifuge tubes and centrifuged at 15,000 rpm at 20 2 °C for
20 min. The supernatant was analyzed for concentration by HPLC (Agilent 1100
binary pump and a Phenomenex Synergi Hydro-RP column (30 Â 2 mm, 4
lm,
Acknowledgment
80 Å particle size)) and for pH by pH meter (where applicable). When mono-
salts were prepared, sufficient volume of the aqueous solution of the base was
added such that the molar ratio of the base to that of the drug is 1.05:1. When
bis-salts were prepared, sufficient volume of the aqueous solution of the base
was added such that the molar ratio of the base to that of the drug is 2.1:1.
14. Ettmayer, P.; Amidon, G. L.; Clement, B.; Testa, B. J. Med. Chem. 2004, 47, 2393.
15. Rautio, J.; Kumpulainen, H.; Heimbach, T.; Oliyai, R.; Oh, D.; Jarvinen, T.;
Savolainen, J. Nat. Rev. Drug Discov. 2008, 7, 255.
We thank Professor Valentino J. Stella for good advice and use-
ful discussions.
References and notes
16. The regio-chemical identity of the prodrugs was verified by NOE studies. For
example, key enhancements observed for prodrug 2g:
1. Gautschi, O.; Heighway, J.; Mack, P. C.; Purnell, P. R.; Lara, P. N., Jr.; Gandara, D.
R. Clin. Cancer Res. 2008, 14, 1639.
2. Carvajal, R. D.; Tse, A.; Schwartz, G. K. Clin. Cancer Res. 2006, 12, 6869.
3. Naruganahalli, K. S.; Lakshmanan, M.; Dastidar, S. G.; Ray, A. Curr. Opin. Investig.
Drugs 2006, 7, 1044.
4. Gautschi, O.; Mack, P. C.; Davies, A. M.; Lara, P. N., Jr.; Gandara, D. R. Clin. Lung
Cancer 2006, 8, 93.
5. Mortlock, A.; Keen, N. J.; Jung, F. H.; Heron, N. M.; Foote, K. M.; Wilkinson, R.;
Green, S. Curr. Top. Med. Chem. 2005, 5, 199.
6. Andrews, P. D. Oncogene 2005, 24, 5005.
7. Wilkinson, R. W.; Odedra, R.; Heaton, S. P.; Wedge, S. R.; Keen, N. J.; Crafter, C.;
Foster, J. R.; Brady, M. C.; Bigley, A.; Brown, E.; Byth, K. F.; Barrass, N. C.; Mundt,
K. E.; Foote, K. M.; Heron, N. M.; Jung, F. H.; Mortlock, A. A.; Boyle, F. T.; Green, S.
Clin. Cancer Res. 2007, 13, 3682.
8. Manfredi, M. G.; Ecsedy, J. A.; Meetze, K. A.; Balani, S. K.; Burenkova, O.; Chen,
W.; Galvin, K. M.; Hoar, K. M.; Huck, J. J.; LeRoy, P. J.; Ray, E. T.; Sells, T. B.;
Stringer, B.; Stroud, S. G.; Vos, T. J.; Weatherhead, G. S.; Wysong, D. R.; Zhang,
M.; Bolen, J. B.; Claiborne, C. F. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 4106.
9. Soncini, C.; Carpinelli, P.; Gianellini, L.; Fancelli, D.; Vianello, P.; Rusconi, L.;
Storici, P.; Zugnoni, P.; Pesenti, E.; Croci, V.; Ceruti, R.; Giorgini, M. L.; Cappella,
P.; Ballinari, D.; Sola, F.; Varasi, M.; Bravo, R.; Moll, J. Clin. Cancer Res. 2006, 12,
4080.
17. Krise, J. P.; Zygmunt, J.; Georg, G. I.; Stella, V. J. J. Med. Chem. 1999, 42, 3094.
18. Dumas, J.; Smith, R. A.; Lowinger, T. B. Curr. Opin. Drug Discov. Devel. 2004, 7,
600.
19. Leppik, I. E.; Boucher, B. A.; Wilder, B. J.; Murthy, V. S.; Watridge, C.; Graves, N.
M.; Rangel, R. J.; Rask, C. A.; Turlapaty, P. Neurology 1990, 40, 456.
20. Luer, M. S. Neurol. Res. 1998, 20, 178.
21. Boucher, B. A. Pharmacotherapy 1996, 16, 777.
22. Yavas, S.; Lizdas, D.; Gravenstein, N.; Lampotang, S. Anesth. Analg. 2008, 106,
880. table of contents.
10. Harrington, E. A.; Bebbington, D.; Moore, J.; Rasmussen, R. K.; Ajose-Adeogun,
A. O.; Nakayama, T.; Graham, J. A.; Demur, C.; Hercend, T.; Diu-Hercend, A.; Su,
M.; Golec, J. M.; Miller, K. M. Nat. Med. 2004, 10, 262.
11. Oslob, J. D.; Romanowski, M. J.; Allen, D. A.; Baskaran, S.; Bui, M.; Elling, R. A.;
Flanagan, W. M.; Fung, A. D.; Hanan, E. J.; Harris, S.; Heumann, S. A.; Hoch, U.;
23. Cohen, L. B. Aliment. Pharmacol. Ther. 2008, 27, 597.
24. Fechner, J.; Schwilden, H.; Schuttler, J. Handb. Exp. Pharmacol. 2008, 253.
25. Hester, E. K.; Chandler, H. V.; Sims, K. M. Ann. Pharmacother. 2006, 40, 1301.
26. Wire, M. B.; Shelton, M. J.; Studenberg, S. Clin. Pharmacokinet. 2006, 45, 137.
27. Cho, A. Ann. Rep. Med. Chem. 2006, 41, 395.