Preparation of phosphonooxymethyl prodrugs of HIV-1 attachment inhibitors

Article abstract of DOI:10.1021/op400225q

A practical and scalable synthesis of phosphonooxymethyl prodrugs of HIV-1 attachment inhibitors is described. Starting from azaindoles 1 and 2, this two-step sequence features an efficient alkylation using chloromethyl phosphate 5 and an exceptionally mi

Full text of DOI:10.1021/op400225q

Article
pubs.acs.org/OPRD
Preparation of Phosphonooxymethyl Prodrugs of HIV‑1 Attachment
Inhibitors
David K. Leahy* and Shawn K. Pack
Chemical Development, Bristol-Myers Squibb, One Squibb Drive, New Brunswick, New Jersey 08903, United States
ABSTRACT: A practical and scalable synthesis of phosphonooxymethyl prodrugs of HIV-1 attachment inhibitors is described.
Starting from azaindoles 1 and 2, this two-step sequence features an efficient alkylation using chloromethyl phosphate 5 and an
exceptionally mild deprotection for tert-butyl phosphates. After a salt formation, the API is formed in 82% and 70% overall yield
for 3a and 4a, respectively. This chemistry was used to prepare multikilogram quantities of API.
degrades to deliver the parent drug and a molecule of
formaldehyde. This manuscript details the development of an
efficient and scaleable conversion of HIV-1 attachment
inhibitors 1 and 2 to phosphonooxymethyl prodrugs 3 and 4
to support early toxicology studies and phase I clinical studies.
INTRODUCTION
■
HIV/AIDS is a global epidemic infecting over 34 million
people. About 2.7 million new cases are reported each year, and
AIDS is responsible for over 1.8 million deaths annually.1 There
are currently over 25 single antiviral agents approved for HIV-1
spanning seven different classes of therapies, as well as multiple
fixed-dose combinations.2 While the use of combination
antiretroviral therapy (cART) has led to extensive progress in
reducing morbidity and mortality, significant challenges around
drug resistance and tolerability underline the importance of
developing new approaches for additional viral targets.3 One
new class of agents for combating HIV/AIDS is HIV-1
attachment inhibitors, belonging to the general class of HIV-1
entry inhibitors.4 HIV-1 attachment inhibitors prevent entry to
the host cell by binding to the viral envelope gp120 and
interfering with the interaction with the human cellular
receptor CD4.5,6
RESULTS AND DISCUSSION
■
HIV-1 attachment inhibitor 3 is a phosphonooxymethyl
prodrug, derived from its parent compound 1, through an
alkylation employing 5 to afford the di-tert-butylphosphate 6.15
Acid-promoted deprotection of 6 affords the prodrug 3
(Scheme 1). The initial conditions for the alkylation of 1
required 5 equiv of the alkylating agent 5 and relied upon the
use of 3 equiv of sodium hydride and 1 equiv of iodine to
generate NaI in situ and thus effect a Finkelstein reaction to
render 5 more electrophilic. In this manner, 6 was produced in
70−80% yield and 92 HPLC area % purity after column
chromatography. There were a number of issues to be
addressed to make this process more amenable to scale-up.
Foremost was the excess of the expensive and mutagenic
alkylating agent 5. The use of sodium hydride poses a
pyrophoric risk, and its reaction with liberated HI evolves
hydrogen gas. Furthermore, this reaction is highly exothermic,
albeit addition-controlled. Preliminary experiments verified that
the charge of alkylating agent 5 could be reduced to 1.5 equiv
and that sodium iodide was an effective substitute for iodine;
thus the sodium hydride stoichiometry was reduced to 1.5
equiv, and the highly exothermic reaction of sodium hydride
with HI was eliminated.
In a program directed towards targeting viral gp120,7 two
promising HIV-1 attachment inhibitors have been studied (1
and 2, Figure 1).8 An unfortunate limitation of these molecules
To address scale-up issues, a more efficient alkylation
reaction was developed using a weaker inorganic base in a
polar aprotic solvent. Although acetone, acetonitrile, and DMF
all proved to be acceptable solvents for the alkylation, DMSO
was ultimately chosen for its superior reaction profile and ease
of removal during workup. Potassium carbonate was chosen as
the inorganic base among other common inorganic bases that
performed well. The azaindole 1 proved sufficiently nucleo-
philic to react under these conditions without the addition of an
iodine source, and the charge of alkylating agent 5 was reduced
Figure 1. HIV attachment inhibitors.
is their poor aqueous solubility, which limits exposure. In an
effort to improve this characteristic, a water-soluble prodrug
strategy9−11 was examined by the use of a phosphonoox-
ymethyl moiety attached to the azaindole N-1 position.12 Once
in the intestines, phosphonooxymethyl prodrugs13 are hydro-
lyzed by alkaline phosphatase in the brush border membra-
ne,13a,14 leading to a readily labile hemiaminal, which rapidly
Received: August 16, 2013
Published: October 25, 2013
© 2013 American Chemical Society
1440
dx.doi.org/10.1021/op400225q | Org. Process Res. Dev. 2013, 17, 1440−1444

Organic Process Research & Development
Article
Scheme 1
to 1.2 equiv.16 The protected prodrug 6 was obtained in 84%
yield after aqueous workup and crystallization from n-butyl
acetate (eq 1), eliminating the requirement for chromato-
improves the conversion to 32% (entry 2). The conversion
responded well to increasing the number of equivalents of
alkylating agent 5 with the optimum conversion of 89%
achieved using 8 equiv (entry 4). Nonetheless, this excessive
amount of alkylating agent was considered unacceptable.
Increasing the concentration from 0.1 g/mL of DMSO to 0.2
g/mL of DMSO did improve the conversion (48%, entry 5),
with additional concentration to 0.4 g/mL having a negative
effect on conversion (17%, entry 6) as the mixture becomes
exceedingly thick and mixing is not effective.
The choice of counterion for both the carbonate base and the
iodide salt proved key for solving this problem. When the
reaction was run at 0.1 g/mL in DMSO, very little
differentiation was observed, except that lithium proved
completely ineffective (Figure 2). However, a key observation
was made with respect to the rubidium and cesium counter-
ions: the thick emulsion previously observed was now a much more
fluid suspension. Hence, the effect of reaction concentration
could then be re-examined with regard to these counterions.
When the reactions were run at 0.2 g/mL, 55% and 49%
conversion was obtained for rubidium and cesium, respectively,
similar to the conversion achieved using the potassium
counterion (Figure 3). However, when the reactions were
run at 0.4 g/mL, the rubidium and cesium counterions afforded
45% and 72% conversion, respectively. This is considerably
better than the potassium case. A counterion effect is indeed
operative when the reaction is run at higher concentration. As the
counterion effect is likely related to the solubility properties of
the carbonate base, iodide salt, and the corresponding salt of
substrate 2, additional screening of solvents could be expected
to augment conversion.
A detailed solvent screen identified DMPU and NMP as
superior solvents with NMP chosen for its relative cost and ease
of removal downstream. A 93% conversion was achieved using
only 2 equiv of 5 in NMP when employing the cesium
counterion (Figure 4). Lowering the reaction temperature from
40 to 30 °C and utilizing potassium iodide instead of cesium
iodide improved the conversion to 97% while minimizing
impurities. This methodology provide 75−81% isolated yield of
crystalline di-tert-butylphosphate 7 with ∼95 HPLC area %
purity17 after a simple workup and crystallization from n-butyl
acetate.
graphic purification. The HPLC area % purity of material
obtained by this method improved dramatically to >98%,
although 11% of 6 remained in the mother liquor.
The di-tert-butylphosphate 6 was readily deprotected
through the action of TFA in methylene chloride at room
temperature, affording crystalline 3 in 98% yield and >99.5
HPLC area % purity after a solvent switch to methanol. Given
the excellent crystallization properties from methanol for
compound 6 and the ability to use methylene chloride both
in the coupling step and in the deprotection, a telescoped
alkylation/deprotection strategy was explored. Thus, simply
treating the rich methylene chloride organic stream containing
6 with TFA followed by a solvent switch to methanol produced
3 directly in 93−96% yield and in 97.4 HPLC area % purity.
This telescoped approach circumvented the significant loss of 6
to the mother liquor of the previous process. Conversion of the
free acid 3 to the L-lysine salt 3a was achieved in 82−87% yield,
upgrading the purity to 99.7 HPLC area % (Scheme 2). This
process was successfully scaled up to prepare 1.1 kg of 3a.
The alkylation of azaindole 2 proved considerably more
challenging than that of 1 due in part to the electron-
withdrawing fluorine atom and triazole ring, which renders the
azaindole less nucleophilic. Additionally, the triazole ring may
sterically hinder the nucleophilic site. Table 1 highlights the
preliminary studies for the conversion of 2 to 7. Adapting the
above conditions (2 equiv of 5, K2CO3, DMSO) to azaindole 2
led to only 24% conversion to the di-tert-butylphosphate 7
(entry 1). This heterogeneous reaction mixture becomes a thick
emulsion with time. The addition of potassium iodide modestly
Scheme 2
1441
dx.doi.org/10.1021/op400225q | Org. Process Res. Dev. 2013, 17, 1440−1444

Organic Process Research & Development
Article
Scheme 3
Table 1. Preliminary studies for the conversion of 2 to 7
entry
equiv of 5
iodine source
concn (g/mL)
conv (%)
1
2
3
4
5
6
2
2
4
8
2
2
none
KI
0.1
0.1
0.1
0.1
0.2
0.4
24
32
57
86
48
17
Figure 4. Counterion effect on conversion.
KI
5%, and the reaction mixture could be held for over 12 h at
room temperature with only minimal degradation. Isolation as
the L-lysine salt 4a proceeded in 82−88% yield and 98.5 HPLC
area % purity (Scheme 3). This process was scaled to prepare
3.2 kg of 4a.
KI
KI
KI
CONCLUSIONS
■
We have developed a practical and scalable process for the
conversion of HIV-1 attachment inhibitors 1 and 2 to
phosphonooxymethyl prodrugs 3 and 4, generating multiple
kilograms of API to support early toxicology studies and phase I
clinical studies. This research effort has brought to light an
unusual concentration-dependent counterion effect and high-
lights extremely mild deprotection conditions for tert-butyl
phosphates.
Figure 2. Counterion effect on conversion.
EXPERIMENTAL SECTION
■
Analytical Methods. Alkylating agent 5 was prepared
according to the method of Mantyla.15 1H NMR and 13C NMR
̈
̈
spectra were recorded on a Bruker Avance or a Bruker DRX
NMR spectrometer. Chemical shifts in DMSO-d6, CD3OD, and
CDCl3 were reported in ppm relative to residual deuterated
1
solvent for H and the deuterated solvent for 13C. Chemical
shifts in D2O were reported in ppm relative to an external
standard for 13C. Coupling constants (J) are given in hertz.
High-resolution mass spectra were obtained on a Waters LCT
TOF or Waters QTOF TOF mass spectrometer using positive
ion electrospray ionization (ESI). HPLC analyses were
collected on a Shimadzu LC-10AT liquid chromatograph
using a SPD-10AV UV−vis detector and the results are
reported as area percent (area %).
Preparation of 3 from 1. A mixture of 1 (1.40 kg, 3.31
mol), K2CO3 (0.919 kg, 6.65 mol), and DMSO (7.00 L) was
stirred, resulting in a light brown suspension. Di-tert-butyl
chloromethyl phosphate 5 (1.03 kg, 3.98 mol) was added over
15 min, and the reaction mixture was heated to 30 °C for 29 h.
Additional K2CO3 (0.460 kg, 3.33 mol) and tert-butyl
chloromethyl phosphate 5 (0.432 g, 1.67 mol) were added to
drive the reaction to completion. The reaction was stirred for
an additional 14 h followed by cooling to 15 °C. DCM (14.0 L)
Figure 3. Conversion versus concentration in DMSO.
Deprotection of the di-tert-butylphosphate 7 proved
challenging due to the instability of prodrug 4 under acidic
conditions. Treatment of 7 with TFA in methylene chloride led
to rapid deprotection; however, ∼20 HPLC area % of
byproducts was generated necessitating milder deprotection
conditions. Strong acids afforded similar results, while weaker
acids afforded no reaction. Gratifyingly, stirring 7 in aqueous
acetone at 40 °C led to complete deprotection after 24 h.18
Byproduct formation, while still an issue, is typically limited to
1442
dx.doi.org/10.1021/op400225q | Org. Process Res. Dev. 2013, 17, 1440−1444

Organic Process Research & Development
Article
was added followed by a slow quench with water (14.0 L)
maintaining <20 °C, which resulted in a biphasic mixture. The
product-rich bottom layer was separated, washed with water
(14.0 L), and treated with trifluoroacetic acid (3.89 L, 49.1
mol). The resulting mixture was stirred for 1 h and cooled to 0
°C, then methanol (21.0 L) was added while maintaining <20
°C, and the mixture was cooled to 0 °C. The mixture was
concentrated under vacuum to a volume of 14 L (200 mmHg,
< 30 °C). The solution was seeded with 3 (10 g) and then
stirred 7 h at room temperature. The resulting slurry was
filtered, and the wet cake was washed with THF (16.6 L) and
then dried in a vacuum oven at 50 °C, resulting in a pale yellow
to white powder (1.69 kg, 96%). 1H NMR (400 MHz, DMSO-
d6) δ 8.30 (s, 1H), 7.55 (s, 1H), 7.44 (s, 5H), 6.12 (d, J = 10.6
Hz, 2H), 3.97 (s, 3H), 3.85 (s, 3H), 3.80−3.22 (m, 8H); 13C
NMR (100 MHz, DMSO-d6) δ 185.49, 169.26, 166.06, 146.20,
145.60, 140.64, 135.50, 129.68, 128.41, 127.04, 123.46, 121.17,
120.08, 114.32, 72.43, 56.92, 53.32, 45.22, 40.50.
was added over 1 h, and the mixture was heated to 50 °C,
resulting in a thin slurry. The slurry was seeded with 4a (36 g),
and IPA (7.0 L) was added over 2 h with stirring for 16 h,
resulting in a slurry that was heated to 70 °C for 2 h and then
cooled to 50 °C. IPA (11.6 L) was added over 1 h, and then
additional IPA (23.6 L) was added over 2 h followed by aging
for 1 h. The slurry was cooled to 20 °C over 2 h and held
overnight prior to filtration. The wet cake was washed with IPA
(35.2 L) and dried in a vacuum oven at 50 °C, resulting in a
1
white powder (3.23 kg, 88%). H NMR (500 MHz, CD3OD,
50 °C) δ 8.94 (s, 1H), 8.87 (s, 1H), 8.69 (s, 1H), 8.42 (s, 1H),
7.83 (m, 5H), 5.81 (d, J = 12.5 Hz, 2H), 4.30−3.70 (m, 8H),
4.08 (t, J = 6.5 Hz, 1H), 3.67 (t, J = 10 Hz, 2H), 2.26 (m, 2H),
2.07 (m, 2H), 1.88 (m, 2H); 13C NMR (125 MHz, D2O, 30
°C) δ 185.2, 174.9, 173.3, 166.8, 153.1, 146.8, 134.8, 134.3,
131.3, 131.1, 130.3, 129.3, 128.9, 128.7, 128.5, 127.2, 124.2,
112.6, 74.0, 54.9, 47.9, 47.2, 46.4, 45.9, 42.7, 42.2, 42.0, 41.7,
39.5, 30.3, 26.8, 21.8 (some signals doubled due to rotamers).
MS m/z: (M-lysine + H)+ calcd for C23H22FN7O7P 558.1302,
found 558.1293. Anal. Calcd: C, 48.75; H, 5.07; N, 17.63; P,
4.33. Found: C, 49.02; H 4.90; N, 17.90; P, 4.37. Melting point
193 °C.
Preparation of 3a from 3. A mixture of 3 (1.00 kg, 1.88
mol), L-lysine (0.275 kg, 1.88 mol), and water (6.0 L) was
heated to 50 °C, resulting in a hazy solution that was filtered
through a 10 μm polish filter before addition of acetone (20.0
L) while maintaining >40 °C. The solution was seeded with 3a
(5 g) and cooled to room temperature over 5 h, resulting in a
slurry. The slurry was stirred 1 h at room temperature and then
filtered, and the wet cake was washed with acetone (6.3 L) and
dried in a vacuum oven at 25 °C with a bleed of moist air,
AUTHOR INFORMATION
Corresponding Author
*Tel: (732)-227-6541. Fax: (732)-227-3003. E-mail: david.
leahy@bms.com.
■
1
resulting in a fluffy white powder (1.08 kg, 85%). H NMR
Notes
(500 MHz, CD3OD) δ 8.38 (s, 1H), 7.49 (m, 6H), 6.13 (d, J =
10.5 Hz, 2H), 4.06 (s, 3H), 3.92 (s, 3H), 4.00−3.40 (m, 8H),
3.58 (t, J = 6 Hz, 1H), 2.92 (t, J = 7.5 Hz, 2H), 1.90−1.40 (m,
6H); 13C NMR (125 MHz, CD3OD) δ 186.1, 173.2, 171.8,
167.8, 147.4, 146.4, 141.0, 135.4, 130.4, 128.8, 127. 2, 124.6,
122.3, 120.2, 114.6, 73.2, 56.6, 54.7, 53.1, 46.0, 41.6, 39.2, 30.5,
27.0, 22.0. HRMS m/z: (M-lysine + H)+ calcd for
C23H26N4O9P 533.1437, found 533.1437. Anal. Calcd: C,
51.32; H, 5.79; N, 12.38; P, 4.56. Found: C, 48.54; H, 5.32; N,
11.76; P, 4.04. Melting point 170 °C.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We would like to thank Dr. Chiajen Lai for his help developing
the API crystallizations and Dr. Derek Berglund, Dr. Joseph
Bush, and Melissa Marchetti for assistance in scaling up these
reactions. We would also like to thank Dr. Nicholas Meanwell,
Dr. James Sherbine, Dr. Prashant Deshpande, and Dr.
Nachimuthu Soundararajan for helpful discussions and support
of this work.
Preparation of 7 from 2. A mixture of 2 (3.01 kg, 6.73
mol), Cs2CO3 (6.56 kg, 20.1 mol), KI (2.23 kg, 13.4 mol), and
NMP (7.5 L) was stirred, resulting in a light brown suspension.
Di-tert-butyl chloromethyl phosphate 5 (3.64 kg, 14.1 mol) was
added over 15 min, and the reaction mixture was heated to 30
°C for 20 h and then cooled to 5 °C. n-BuOAc (18.0 L) was
added followed by a slow quench with water (30.0 L)
maintaining <15 °C, which resulted in a biphasic mixture that
was separated. The product rich top layer was seeded with 7
(40 g) and stirred for 3.5 h at room temperature, resulting in a
slurry. The slurry was filtered, and the wet cake was washed
with MTBE (6.3 L) and then dried in a vacuum oven at 50 °C,
REFERENCES
■
(1) Statistics as of 2010: WHO, UNAIDS, UNICEF. Global HIV/
AIDS Response: Epidemic Update and Health Sector Progress towards
Universal Access−Progress Report 2011; WHO Press: Geneva, 2011.
(2) De Clercq, E. Int. J. Antimicrob. Agents 2009, 33, 307−320.
(3) (a) Taiwo, B.; Hicks, C.; Eron, J. J. Antimicrob. Chemother. 2010,
65, 1100−1107. (b) Mehellou, Y.; De Clercq, E. J. Med. Chem. 2010,
53, 521−538.
(4) For recent reviews of inhibitors of HIV entry, see: (a) Teissier,
E.; Penin, F.; Pecheur, E.-I. Molecules 2011, 16, 221−250. (b) Caffrey,
́
M. Trends Microbiol. 2011, 19, 191−197. (c) Singh, I. P.; Chauthe, S.
K. Expert Opin. Ther. Patents 2011, 21, 399−416. (d) Hertje, M.;
Zhou, M.; Dietrich, U. ChemMedChem 2010, 5, 1825−1835.
(e) Tilton, J. C.; Doms, R. W. Antiviral Res. 2010, 85, 91−100.
(5) The precise mode of action of HIV-1 attachment inhibitors has
not been resolved, with data suggesting that these compounds bind to
HIV-1 gp120 to stabilize a conformation not recognized by CD4 and/
or form a ternary complex with HIV-1 gp120 and CD4 and inhibit
CD4-mediated conformational changes in gp120 that lead to exposure
of the fusion peptide gp41.
1
resulting in a yellow/white powder (3.55 kg, 79%). H NMR
(400 MHz, CDCl3) δ 8.35 (s, 1H), 8.27 (s, 1H), 8.22 (s, 1H),
7.90 (s, 1H), 7.42, (s, 5H), 5.92, (d, J = 14.9 Hz, 2H), 4.02−
3.40 (m, 8H), 1.24 (s, 18H); 13C NMR (100 MHz, CDCl3) δ
182.75, 170.64, 152.07, 144.03, 134.91, 133.96, 131.82, 130.21,
128.68, 128.27, 128.00, 127.07, 125.81, 124.01, 113.82, 84.00,
83.93, 73.97, 46.12, 41.90, 29.57, 29.53.
Preparation of 4a from 7. A mixture of 7 (3.51 kg, 5.24
mol), IPA (7.0 L) and water (7.0 L) was heated to 40 °C and
stirred for 20 h before being cooled to 20 °C. L-Lysine (0.730
kg, 4.99 mol) was added, resulting in a solution that was filtered
through a 10 μm polish filter. IPA (9.5 L) was added followed
by seeding with 4a (36 g) and stirring for 30 min. IPA (3.5 L)
(6) For recent reviews on the mechanism of HIV entry, see:
(a) Kang, Y.; Guo, J.; Chen, Z. Protein Cell 2013, 4, 86−102.
(b) Wilen, C. B.; Tilton, J. C.; Doms, R. W. Cold Spring Harbor
Perspect. Med. 2012, 2, a006866.
(7) (a) Lin, P.-f.; Blair, W. S.; Wang, T.; Spicer, T. P.; Quo, Q.; Zhou,
N.; Gong, Y.-F.; Wang, H.-G.; Rose, R.; Yamanaka, G. A.; Robinson, B.
1443
dx.doi.org/10.1021/op400225q | Org. Process Res. Dev. 2013, 17, 1440−1444

Organic Process Research & Development
Article
S.; Li, C.-B.; Kimel, B. J.; Fang, H.; Demers, G.; Deminie, C.; Fridell,
R. A.; Yang, Z.; Zadjura, L. M.; Meanwell, N. A.; Colonno, R. J. Proc.
Natl. Acad. Sci. U.S.A. 2003, 100, 11013−11018. (b) Guo, Q.; Ho, H.-
T.; Dicker, I.; Fan, L.; Zhou, N.; Friborg, J.; Wang, T.; McAuliffe, B.
V.; Wang, H.-G.; Rose, R. E.; Fang, H.; Scarnati, H. T.; Langley, D. R.;
Meanwell, N. A.; Abraham, R.; Colonno, R. J.; Lin, P.-f. J. Virol. 2003,
77, 10528−10536. (c) Wang, T.; Zhang, Z.; Wallace, O. B.;
Deshpande, M.; Fang, H.; Yang, Z.; Zadjura, L. M.; Tweedie, D. L.;
Huang, S.; Zhao, F.; Ranadive, S.; Robinson, B. S.; Gong, Y.-F.;
Ricarrdi, K.; Spicer, T. P.; Deminie, C.; Rose, R.; Wang, H.-G.; Blair,
W. S.; Shi, P.-Y.; Lin, P.-f.; Colonno, R. J.; Meanwell, N. A. J. Med.
Chem. 2003, 46, 4236−4239. (d) Ho, H.-T.; Fan, L.; Nowicka-Sans,
B.; McAuliffe, B.; Li, C.-B.; Yamanaka, G.; Zhou, N.; Fang, H.; Dicker,
I.; Dalterio, R.; Gong, Y.-F.; Wang, T.; Yin, Z.; Ueda, Y.; Matiskella, J.;
Kadow, J.; Clapham, P.; Robinson, J.; Colonno, R.; Lin, P.-F. J. Virol.
2006, 80, 4017−4025.
(8) (a) Wang, T.; Yin, Z.; Zhang, Z.; Bender, J. A.; Yang, Z.; Johnson,
G.; Yang, Z.; Zadjura, L. M.; D’Arienzo, C. J.; DiGiugno Parker, D.;
Gesenberg, C.; Yamanaka, G. A.; Gong, Y.-F.; Ho, H.-T.; Fang, H.;
Zhou, N.; McAuliffe, B. V.; Eggers, B. J.; Fan, L.; Nowicka-Sans, B.;
Dicker, I. B.; Gao, Q.; Colonno, R. J.; Lin, P.-F.; Meanwell, N. A.;
Kadow, J. F. J. Med. Chem. 2009, 52, 7778−7787. (b) Regueiro-Ren,
A.; Xue, Q. M.; Swidorski, J. J.; Gong, Y.-F.; Mathew, M.; Parker, D.
D.; Yang, Z.; Eggers, B.; D’Arienzo, C.; Sun, Y.; Malinowski, J.; Gao,
Q.; Wu, D.; Langley, D. R.; Colonno, R. J.; Chien, C.; Grasela, D. M.;
Zheng, M.; Lin, P.-F.; Meanwell, N. A.; Kadow, J. F. J. Med. Chem.
2013, 56, 1656.
(9) Stella, V. J.; Charman, W. N. A.; Naringrekar, V. H. Drugs 1985,
29, 455−473.
(10) Fleisher, D.; Bong, R.; Stewart, B. H. Adv. Drug Delivery Rev.
1996, 19, 115−130.
(11) Krise, J. P.; Stella, V. J. Adv. Drug Delivery Rev. 1996, 19, 287−
310.
(12) Kadow, J. F.; Ueda, Y.; Meanwell, N. A.; Connolly, T. P.; Wang,
T.; Chen, C.-P.; Yeung, K.-S.; Zhu, J.; Bender, J. A.; Yang, Z.; Parker,
D.; Lin, P.-F.; Colonno, R. J.; Mathew, M.; Morgan, D.; Zheng, M.;
Chien, C.; Grasela, D. J. Med. Chem. 2012, 55, 2048.
(13) Phosphonooxymethyl prodrugs have been successfully applied
to poorly water-soluble drugs such as paclitaxel, phenytoin, and
ravuconazole. For paclitaxel see: (a) Golik, J.; Wong, H. S. L.; Chen, S.
H.; Doyle, T. W.; Wright, J. J. K.; Knipe, J.; Rose, W. C.; Casazza, A.
M.; Vyas, D. M. Bioorg. Med. Chem. Lett. 1996, 6, 1837−1842. For
phenytoin see: (b) Stella, V. J. Adv. Drug Delivery Rev. 1996, 19, 311−
330. For ravuconazole see: (c) Ueda, Y.; Matiskella, J. D.; Golik, J.;
Connolly, T. P.; Hudyma, T. W.; Venkatesh, S.; Dali, M.; Kang, S.-H.;
Barbour, N.; Tejwani, R.; Varia, S.; Knipe, J.; Zheng, M.; Mathew, M.;
Mosure, K.; Clark, J.; Lamb, L.; Medin, I.; Gao, Q.; Huang, S.; Chen,
C.-P.; Bronson, J. J. Bioorg. Med. Chem. Lett. 2003, 13, 3669−3672.
(14) Krise, J. P.; Zygmunt, J.; Georg, G. I.; Stella, V. J. J. Med. Chem.
1999, 42, 3094−3100.
(15) For the synthesis of 5 see: Mantyla, A.; Vepsalainen, J.; Jarvinen,
̈
̈
̈
̈
̈
T.; Nevalainen, T. Tetrahedron Lett. 2002, 43, 3793−3794.
(16) Upon scale-up, a kicker charge of 0.5 equiv of 5 and 1 equiv of
K2CO3 was required to reach full conversion.
(17) The purity value includes 0−6% of the monodeprotected
phosphate derivative, which is an intermediate en route to the prodrug
and thus is inconsequential to the purity of the compound.
(18) (a) U.S. Patent Application US 2005/0209246, 2005. (b) U.S.
Patent Application US 2012/0157411, 2012.
1444
dx.doi.org/10.1021/op400225q | Org. Process Res. Dev. 2013, 17, 1440−1444