Novel cyclic phosphate prodrug approach for cytochrome P450-activated drugs containing an alcohol functionality

Article abstract of DOI:10.1007/s11095-006-9187-y

Purpose. A cyclic phosphate prodrug of a descriptive molecule containing an alcohol functionality was designed, synthesized and characterized in vitro as a cytochrome P450 (CYP) -selective prodrug. Materials and Methods. To achieve efficient CYP-oxidation

Full text of DOI:10.1007/s11095-006-9187-y

#
Pharmaceutical Research, Vol. 24, No. 4, April 2007 ( 2007)
DOI: 10.1007/s11095-006-9187-y
Research Paper
:
Novel Cyclic Phosphate Prodrug Approach for Cytochrome P450-activated
Drugs Containing an Alcohol Functionality
Kristiina M. Huttunen,1,4 Niina Ma¨ho¨nen,2 Jukka Leppa¨nen,1 Jouko Vepsa¨la¨inen,3 Risto O. Juvonen,2
Hannu Raunio,2 Hanna Kumpulainen,1 Tomi Ja¨rvinen,1 and Jarkko Rautio1
Received August 15, 2006; accepted November 6, 2006; published online February 15, 2007
Purpose. A cyclic phosphate prodrug of a descriptive molecule containing an alcohol functionality was
designed, synthesized and characterized in vitro as a cytochrome P450 (CYP) -selective prodrug.
Materials and Methods. To achieve efficient CYP-oxidation and prodrug bioconversion, 1,3-cyclic
propyl ester of phosphate was designed to have a C4-aryl substituent and synthesized using
phosphorus(III) chemistry. The two-step bioconversion of the cyclic phosphate prodrug was evaluated
in vitro using human liver microsomes and recombinant CYP enzymes.
Results. This cyclic phosphate prodrug underwent initial CYP-catalyzed oxidation and was mainly
catalyzed by the CYP3A4 form. The hydroxylated product was slowly converted to a ring-opened
intermediate, which subsequently transformed by b-elimination reaction to a free phosphate. The free
phosphate was further dephosphorylated by microsomal phosphatases, releasing the parent molecule
with a free hydroxyl group. The cyclic phosphate was reasonably stable in buffer solutions at the pH
range 1.0j9.0.
Conclusions. Since CYP enzymes reside predominantly in the liver and secondarily in the small
intestine, the results indicate that cyclic phosphate prodrugs represent a very feasible liver- or intestinal-
targeted drug delivery strategy for drug molecules containing an alcohol functionality. This may
potentially improve the efficacy and the safety profile of the alcoholic parent drugs.
KEY WORDS: bioconversion; cyclic phosphate; cytochrome P450 (CYP); prodrug; targeted drug
delivery.
INTRODUCTION
The prodrug approach is a promising strategy to over-
come the pharmaceutical and pharmacokinetic limitations of
After administration of a drug into the body, several the parent drug, since prodrugs are inactive, bioreversible
pharmaceutical and pharmacokinetic barriers have to be derivatives of active drug molecules (1). One of the major
bypassed before the drug reaches its physiological target, objectives of the prodrug design is to achieve site-selective
such as a receptor, and can exert its desired effect. Targeted drug delivery and tissue or cell-specific enzymes offer an
drug delivery is considered as an efficient strategy to produce advantageous way of targeting drugs to the desired tissue.
drugs with increased efficacy and fewer adverse effects by Cytochrome P450 (CYP) enzymes are a superfamily of heme
increasing local drug concentrations (1,2). Different drug- containing proteins that catalyze xenobiotic metabolism
carrier delivery systems, such as liposomes (3), glycosylated Phase I reactions, among which the most common is
macromolecules (4), antibodies (5) and nanoparticles (6), substrate oxidation by insertion of one oxygen atom. Most
have been used to avoid unwanted interactions between the of the CYP-catalyzed reactions lead to the detoxification of
drug and the body, until the drug is released from the carrier xenobiotics but CYPs can also bioactivate chemicals includ-
at the target site. However, not all of these strategies have ing drugs (7). Since CYP enzymes have broad substrate
been resulted in clinically-proven drugs.
specificity and are particularly abundant in the liver and to
lesser extent in the small intestine, they are ideal tools for
liver- or intestinal-targeted prodrug technology.
Kristiina M. Huttunen and Niina Ma¨ho¨nen have contributed equally
to this work.
Previously, cyclic 1,3-propanyl esters of phosphates and
phosphonates have been developed for targeting various
nucleoside monophosphates (NMP) to the liver (8Y12). These
cyclic phosphate prodrugs undergo CYP-catalyzed oxidation
predominantly in hepatocytes, although CYPs are also
expressed to a lesser extent in the small intestine. Oxidation
results in ring opening and b-elimination of an aryl vinyl
ketone. The released monophosphates or phosphonates are
further converted by nucleotide kinases to active nucleoside
1 Department of Pharmaceutical Chemistry, University of Kuopio,
P.O. Box 1627, 70211 Kuopio, Finland.
2 Department of Pharmacology and Toxicology, University of
Kuopio, P.O. Box 1627, 70211 Kuopio, Finland.
3 Department of Chemistry, University of Kuopio, P.O. Box 1627,
70211 Kuopio, Finland.
4 To whom correspondence should be addressed. (e<mail: Kristiina.
Huttunen@uku.fi)
#
679
0724-8741/07/0400-0679/0
2007 Springer Science + Business Media, LLC

680
Huttunen et al.
triphosphates (NTP), which act as inhibitors of either viral 4-(6-Methoxy-naphthalen-2-yl)-butan-2-ol (1)
replication or cell proliferation. In addition to their broad
antiviral activities, NTPs may also be used to treat other liver
diseases, such as hepatocellular carcinoma.
NaBH4 (1.2 g, 32.9 mmol) was added in small propor-
tions to the solution of 4-(6-methoxy-2-naphthyl)-2-butanone
The purpose of this study was to assess whether the (5.0 g, 21.9 mmol) in anhydrous EtOH (60 ml) at j10-C
cyclic phosphates could also serve as potential prodrugs under an Ar-atmosphere. The reaction mixture was stirred
targeting drugs containing an alcohol functionality to the for 1 h at j10-C and 3 h at RT. The solvent was evaporated
liver or intestine. Several drugs used in the treatment of under reduced pressure and the residue was dissolved in
liver or intestinal diseases, such as hypolipidemic and anti- dichloromethane (DCM) (50 ml), washed carefully with cold
cancer drugs, as well as glucocorticoids and other immuno- H2O (2ꢀ30 ml), dried over Na2SO4 and concentrated in
suppressive drugs, contain an alcohol functionality and vacuo to provide the alcohol 1 as white solid (4.8 g, 95%).
3
would greatly benefit from liver- or intestinal-targeted drug 1H NMR (CDCl3): d 1.24 (3H, d, JHH=6.19 Hz), 1.82 (1H,
delivery (13Y15). Once liberated from the cyclic phosphates dddd, 2JHH=j13.63 Hz, 3JHH=9.58, 6.22, 4.82), 1.85 (1H, dddd,
3
by CYP enzymes in the first step of the bioconversion process 2JHH=j13.63 Hz, JHH=9.70, 6.85, 6.51), 2.80 (1H, ddd,
3
2
(Fig. 1), the released charged free phosphates would have 2JHH=j13.93 Hz, JHH=9.58, 6.85), 2.87 (1H, ddd, JHH=j13.93
3
3
only a poor ability to diffuse passively across the cell Hz, JHH=9.70, 6.22), 3.84 (1H, qdd, JHH=6.19, 6.51, 4.82 Hz),
membranes and they could be further dephosphorylated by 3.90 (3H, s), 7.10 (1H, d, 4JHH=2.52 Hz), 7.12 (1H, dd, 3JHH=8.98
3
phosphatases within cells (16,17) to their corresponding Hz, JHH=2.52 Hz), 7.30 (1H, dd, JHH=8.32 Hz, JHH=1.80 Hz),
4
4
4
3
alcohols (18,19).
7.55 (1H, d, JHH=1.80 Hz), 7.656 (1H, d, JHH=8.98 Hz), 7.659
In the present study we designed and synthesized a cyclic (1H, d, JHH=8.32 Hz); 13C NMR (CDCl3): d 23.68 q, 32.07 t,
1,3-propanyl phosphate ester of a parent drug with a free 40.81 t, 55.30 q, 67.52 d, 105.70 d, 118.71 d, 126.24 d, 126.84 d,
hydroxyl group that was aimed to undergo a two-step 127.78 d, 128.89 d, 129.14 s, 133.0 s, 137.21 s, 157.19 s. ESI-MS:
bioconversion process by CYP enzymes and alkaline phos- m/z = 230.1 (M+H)+. Anal. Calcd for (C15H18O2 *0.1 H2O): C,
phatases, respectively (Fig. 1). A simple representative 77.62; H, 7.90; Found: C, 77.62; H, 8.11.
3
alcohol with suitable detection properties and a low amount
of other functional groups was used to evaluate the potential Phosphoric Acid mono-[3-(6-methoxy-naphthalen-2-yl)-1-
of this approach in vitro.
methyl-propyl] Ester (2)
MATERIALS AND METHODS
General Synthetic Procedures
The alcohol 1 (1.2 g, 5.2 mmol) and Et3N (1.1 ml,
7.8 mmol) in cold anhydrous DCM were added dropwise to a
solution of freshly distilled phosphorus oxychloride (0.57 ml,
6.3 mmol) in anhydrous DCM (50 ml) at 0-C under an Ar-
All the reactions were performed with reagents of atmosphere. The reaction mixture was stirred at RT over-
commercial high purity quality without further purification night, H2O (30 ml) was added at 0-C to the reaction
unless otherwise mentioned. Reactions were monitored by mixture and stirred for 1 h. The aqueous phase was
thin-layer chromatography using aluminium sheets coated saturated with NaCl, separated and extracted with DCM
with silica gel 60 F245 (0.24 mm) with suitable visualization. (2ꢀ15 ml). The combined organic phase was evaporated
Purifications by flash chromatography were performed on under reduced pressure. The residue (1.4 g) was dissolved
1
silica gel 60 F245 (50 mm). H, 13C and 31P nuclear magnetic in H2O, treated with 2 eq. 2M NaOH and evaporated in
resonance (NMR) spectra were recorded on a Bruker vacuo. The sodium salt of phosphate 2 was crystallized
1
Avance 500 spectrometer (Bruker Biospin, Switzerland) from acetone as a light yellow solid (1.0 g, 54%). H NMR
3
operating at 500.13, 125.75 and 202.46 MHz, respectively. (D2O): d 1.33 (3H, d, JHH=6.24 Hz), 1.86 (1H, dddd,
3
Tetramethylsilane (TMS) was used as an internal standard 2JHH=j13.34 Hz, JHH=11.48, 5.19, 6.27 Hz), 1.98 (1H,
2
3
for 1H and 13C spectra as well 85% H3PO4 as an external dddd, JHH=j13.34 Hz, JHH=11.56, 6.32, 5.25 Hz), 2.81
standard for 31P spectra. H NMR spectra of the compounds (1H, ddd, JHH=j13.85 Hz, JHH=11.56, 5.19 Hz), 2.91 (1H,
1
2
3
2
3
1, 2, 5 and 8 were analysed by using PERCH NMR Software ddd, JHH=j13.85 Hz, JHH=11.48, 5.25 Hz), 3.93 (3H, s),
(20). Furthermore, the products were characterized by 4.29 (1H, qddd, JHH=6.24, 6.32, 6.27 Hz, JHP=8.14), 7.19
3
3
3
elemental analysis (C, H, N) with a ThermoQuest CE (1H, dd, JHH=9.0, JHH=2.61 Hz), 7.31 (1H, d, JHH=2.61
4
4
3
4
Instruments EA 1110-CHNS-O elemental analyzer (CE Hz), 7.51 (1H, dd, JHH=8.44, JHH=1.75 Hz), 7.75 (1H, d,
Instruments, Milan, Italy) as well by mass spectroscopy with 4JHH=1.75 Hz), 7.80 (1H, d, JHH=8.44 Hz), 7.81 (1H, d,
3
a Finnigan LCQ quadrupole ion trap mass spectrometer 3JHH=9.0 Hz, 1H); 13C NMR: d 21.19 qd (3JCP=2.5 Hz), 31.28
(ESI-MS) equipped with an electrospray ionization source t, 39.56 td (3JCP=5.1 Hz), 55.29 q, 70.80 dd (2JCP=5.3 Hz),
(Finnigan MAT, San Jose, CA, USA).
106.16 d, 118.18 d, 125.87 d, 126.86 d, 128.31 d, 128.95 s,
O
O
O
Ar
1. CYP450
phosphatases
Drug
Drug
Drug
O P
O
O P OH
OH
OH
2. ring-opening,
ß-elimination
Fig. 1. Proposed bioconversion of cyclic phosphate prodrugs to alcohols in hepatocytes.

CYP-activated Cyclic Phosphate Prodrugs
681
1
129.03 d, 132.49 s, 138.81 s, 156.31 s; 31P NMR: d 4.55. ESI- to yield the yellow oily propanediol 5 (2.0 g, 51%). H NMR
MS: anionic m/z = 309.2 (M+H)j. Anal. Calcd for (CDCl3): d 1.92 (1H, dddd, 2JHH=j14.58 Hz, 3JHH=6.31, 3.73,
2
3
(C15H17O5PNa2 *3.5 H2O *1.0 DCM): C, 35.87; H, 5.22; 3.56 Hz), 1.99 (1H, dddd, JHH=j14.58 Hz, JHH=8.99, 8.12,
2
Found: C, 35.91; H, 5.67.
3.98 Hz), 2.89 (bs, 2H), 3.82 (1H, ddd, JHH=j10.85 Hz,
2
3JHH=8.12, 3.73 Hz), 3.83 (1H, ddd, JHH=j10.85 Hz,
3
2-Chloro-1,3,2-dioxaphosphorinane 2-oxide (3)
3JHH=6.31, 3.98 Hz), 4.93 (1H, dd, JHH=8.99, 3.56 Hz),
7.37Y7.24 (1H, m); 13C NMR: d 40.46 t, 61.35 t, 74.22 d, 125.66
Freshly distilled phosphorus oxychloride (6.06 ml, d, 127.55 d, 128.50 d, 144.34 s. ESI-MS: m/z = 151.3 (M-H)j.
65.0 mmol) in cold anhydrous Et2O (80 ml) was added
dropwise to a solution of 1,3-propanediol (4.7 ml, 65.0 mmol)
2-[3-(6-Methoxy-naphthalen-2-yl)-1-methyl-propoxy]-
and anhydrous Et3N (18.0 ml, 130 mmol) in anhydrous Et2O
4-phenyl-1,3,2-dioxaphosphorinane 2-oxide (7)
(150 ml) at 0-C under an Ar-atmosphere. The reaction
mixture was stirred at RT overnight. The precipitate was Phosphorus(V) method. Freshly distilled phosphorus
filtered out of the mixture and washed thoroughly with Et2O. oxychloride (0.60 ml, 6.6 mmol) in cold anhydrous Et2O
The filtrate was concentrated in vacuo to obtain the cyclic (20 ml) was added dropwise to a solution of 1-phenyl-
phosphoryl chloride 3 as a white solid (10.1 g, 99%). 1H propane-1,3-diol 5 (1.0 g, 6.6 mmol) and Et3N (1.8 ml, 13.1
2
NMR (CDCl3): d 1.85 (1H, dm, JHH=15.2 Hz), 2.39Y2.49 mmol) in anhydrous Et2O (50 ml) at 0-C under an Ar-
(1H, m), 4.48Y4.61 (4H, m); 13C NMR: d 25.71 td (3JCP=7.6 atmosphere. The reaction mixture was stirred at RT
Hz), 70.43 td (3JCP=8.2 Hz); 31P NMR: d j1.68. ESI-MS: m/z overnight. The precipitate was filtered out of the mixture
= 155.0 (M-H)j. Anal. Calcd for (C3H6O3ClP *0.01 Et3N): C, and washed thoroughly with Et2O. The filtrate was
23.33; H, 3.94; Found: C, 23.65; H, 4.22.
concentrated in vacuo to obtain the cyclic phosphoryl
chloride 6 as a brownish oily mixture of diastereomers
(55:45, 1.1 g, 72%). Due to partial decomposition, the
2-[3-(6-Methoxy-naphthalen-2-yl)-1-methyl-propoxy]-1,3,2-
dioxaphosphorinane 2-oxide (4)
1
product was not purified further. H NMR (CDCl3): Major
diastereomer: d 2.04Y2.13 (1H, m), 2.47Y2.76 (1H, m),
3
3
4.52Y4.77 (2H, m), 5.57 (1H, td, JHH=11.9 Hz, JHP=2.3
Hz), 7.34Y7.50 (5H, m); Minor diastereomer: d 2.17Y2.47 (2H,
m), 4.52Y4.77 (2H, m), 5.74Y5.80 (1H, m), 7.34Y7.50 (5H, m);
31P NMR: Major diastereomer: d j1.85, Minor diastereomer:
d j2.37.
The alcohol 1 (1.2 g, 5.1 mmol) in anhydrous DCM
(10 ml) was added dropwise to a solution of cyclic phosphoryl
chloride 3 (0.8 g, 5.1 mmol) and 1-methylimidazole (0.41 ml,
5.1 mmol) in anhydrous DCM (20 ml) at 0-C under an Ar-
atmosphere. The reaction mixture was stirred at RT over-
night, the solvent was removed in vacuo and the residue was
redissolved in DCM (20 ml). The organic phase was washed
with 5% NaHCO3 (10 ml) and water (10 ml), dried over
Na2SO4 and evaporated under reduced pressure. The residue
was purified by flash chromatography eluting with petroleum
ether/ethyl acetate (4/1) solution to yield the cyclic phosphate
The alcohol 1 (1.1 g, 4.7 mmol) in anhydrous DCM
(20 ml) was added dropwise to a solution of cyclic phosphoryl
chloride 6 (1.1 g, 4.7 mmol) and 1-methylimidazole (0.38 ml,
4.7 mmol) in anhydrous DCM (20 ml) at 0-C under an Ar-
atmosphere. The reaction mixture was stirred at RT over-
night. The solvent was removed in vacuo and the residue was
purified by flash chromatography eluting with petroleum
ether/ethyl acetate (2/1) solution to yield the cyclic phosphate
7 as a yellowish oily mixture of four diastereomers (85 mg,
1
4 as white solid (1.3 g, 71%). H NMR (CDCl3): d 1.43 (3H,
3
d, JHH=6.2 Hz), 1.70Y1.76 (1H, m), 1.90Y1.99 (1H, m),
2.06Y2.13 (1H, m), 2.16Y2.25 (1H, m), 2.79Y2.92 (2H, m),
3.90 (3H, s), 4.30Y4.42 (4H, m), 4.64Y4.71 (1H, m), 7.09Y7.13
(2H, m), 7.29Y7.32 (1H, m), 7.56Y7.57 (1H, m), 7.65Y7.68 (2H,
m); 13C NMR: d 21.63 qd (3JCP=3.0 Hz), 26.03 td (3JCP=7.0
Hz), 31.46 t, 39.14 td (3JCP=6.3 Hz), 55.28 q, 68.27 td
(2JCP=7.0 Hz), 68.41 td (2JCP=6.9 Hz), 75.90 dd (2JCP=6.1
Hz), 105.66 d, 118.75 d, 126.27 d, 126.91 d, 127.66 d, 128.90 d,
129.10 s, 133.06 s, 136.50 s, 157.25 s; 31P NMR: d j7.14. ESI-
MS: m/z = 350.9 (M+H)+. Anal. Calcd for (C18H23O5P): C,
61.71; H, 6.62; Found: C, 61.59; H, 6.88.
1
3%). H NMR (CDCl3): d 1.42Y1.47 (3H, m), 1.90Y2.16 (3H,
m), 2.21Y2.35 (1H, m), 2.79Y2.96 (2H, m), 3.90 (3H, bs),
4.35Y4.46 (1H, m), 4.66Y4.76 (m, 2H), 5.64Y5.70 (1H, m),
7.08Y7.13 (2H, m), 7.27Y7.41 (5H, m), 7.55 (1H, bs), 7.60Y7.67
(2H, m); 13C NMR: d 21.80 qd (3JCP=2.8 Hz), 31.32 td
(4JCP=15.1 Hz), 33.99 td (3JCP=5.0 Hz), 39.01Y39.19 tm
(3JCP=6.5 Hz), 55.29 q, 66.67Y66.80 tm (2JCP=5.4 Hz), 77.40
dd (2JCP=6.7 Hz), 79.72Y79.91 dm (2JCP=4.6 Hz), 105.62 d,
118.73 d, 125.55 d (2C), 126.36 d, 126.86 d, 127.72 d, 128.61 d,
128.71 d (2C), 128.92 d, 129.08 s, 133.06 s, 136.52 s, 139.25 sd
(5JCP=7.8 Hz), 157.25 s; 31P NMR: d j3.63, j3.73. According
to 31P NMR spectrum, two of four possible diastereomers
were observed in a ratio of 56:44). ESI-MS: m/z = 426.8
(M+H)+. Anal. Calcd for (C24H27O5P): C, 65.93; H, 6.50;
Found: C, 65.71; H, 6.49.
1-Phenyl-propane-1,3-diol (5)
Ethyl benzoylacetate (5.0 g, 26.0 mmol) in anhydrous
Et2O (100 ml) was dropped to a solution of LiAlH4 (2.0 g,
52.0 mmol) in anhydrous Et2O (400 ml) at j5-C under an
Ar-atmosphere. The reaction mixture was allowed to warm Phosphorus(III) method. Freshly distilled phosphorus
to RT and refluxed overnight. The mixture was cooled in an trichloride (0.17 ml, 2.0 mmol) in cold anhydrous Et2O
ice bath and treated with H2O (2 ml), 1M NaOH (8 ml) and (10 ml) was dropped to a solution of 1-phenyl-propane-1,3-
H2O (10 ml). The precipitate was filtered out of the mixture diol 5 (0.3 g, 2.0 mmol) and pyridine (0.32 ml, 4.0 mmol) in
and washed thoroughly with Et2O. The filtrate was dried anhydrous Et2O (10 ml) at j10-C under an Ar-atmosphere.
over Na2SO4 and concentrated in vacuo. The residue was The reaction mixture was stirred at 10-C for 2 h, the
purified by flash chromatography eluting with ethyl acetate precipitate was filtered out of the mixture and washed

682
Huttunen et al.
a) Phosphorus(V) chemistry
O
OH
O
R
O
P
O
R
O
a
P
+
O
Cl
O
72% (R = H-)
3% (R = Ph-)
O
3 (R = H-)
6 (R = Ph-)
4 (R = H-)
7 (R = Ph-)
1
O
b) Phosphorus(III) chemistry
O
OH
O
P
b,c
O
O
P
+
Cl
O
11%
O
O
1
8
O
7
Fig. 2. Synthetic pathways to the cyclic phosphate esters. Reaction conditions: (a) N-MeIm, DCM, 0-C, 30 min; 24-C, 20 h; (b) DIPEA, ACN,
0-C, 30 min; 24-C, 20 h; (c) TBHP, 0Y24-C, 4 h.
thoroughly with Et2O. The filtrate was concentrated in vacuo identical to the spectra and analysis of compound 7 obtained
to obtain the yellowish oily cyclic chlorophosphane 8 (0.38 g, by the phosphorus(V) method.
89%). Due to partial decomposition, the product was not
purified further. 1H NMR (CDCl3): d 2.00 (1H, dddd, Phosphoric Acid 3-hydroxy-3-phenyl-propyl ester 3-(6-
3
2JHH=j14.54 Hz, JHH=2.33, 2.24, 1.70 Hz), 2.51 (1H, dddd, methoxy-naphthalen-2-yl)-1-methyl-propyl Ester (9)
3
2JHH=j14.54 Hz, JHH=11.95, 11.03, 5.36 Hz), 4.19 (1H,
dddd, 2JHH=j11.36 Hz, 3JHH=11.03, 1.70 Hz, 3JHP=4.49), 4.88
The buffer solutions of chemical degradation studies
2
3
(1H, dddd, JHH=j11.36 Hz, JHH=5.36, 2.24 Hz, were combined and evaporated under reduced pressure. The
3
3JHP=12.76), 5.69 (1H, ddd, JHH=11.95, 2.33 Hz, residue was extracted with Et2O (2ꢀ15 ml) and the organic
3JHP=5.16), 7.33 (m, 5H); 13C NMR: d 35.75 td (3JCP=2.8 phase was washed with H2O (10 ml), dried over Na2SO4 and
Hz), 62.45 td (2JCP=3.6 Hz), 73.54 dd (2JCP=3.2 Hz), 125.69 d, concentrated in vacuo to provide the hydrolysis product 9 of
128.47 d, 128.71 d, 140.0 sd (3JCP=2.3 Hz).
the cyclic phosphate 7 as a yellowish oily compound (20 mg).
3
The cyclic chlorophosphane 8 (0.38 g, 1.8 mmol) in 1H NMR (CD3OD): d 1.36 and 1.37 * (3H, d, JHH=6.1, 5.9
anhydrous acetonitrile (ACN) (5 ml) was added dropwise to Hz), 1.87Y1.96 (1H, m), 1.98Y2.10 (3H, m), 2.74Y2.93 (2H, m),
a solution of alcohol 1 (0.4 g, 1.8 mmol) and diisopropylethyl- 3.89 (3H, s), 3.97Y4.05 (1H, m), 4.12Y4.20 (1H, m), 4.42Y4.51
amine (DIPEA) (0.61 ml, 3.5 mmol) in anhydrous ACN (1H, m), 4.78Y4.82 (1H, m), 7.08 (1H, dd, 3JHH=9.0, 4JHH=2.6
(10 ml) at 0-C under an Ar-atmosphere. The reaction Hz), 7.15 (1H, bs), 7.20Y7.24 (1H, m), 7.27Y7.35 (5H, m), 7.57
mixture was stirred at RT overnight, cooled in an ice bath (1H, bs), 7.63Y7.69 (2H, m); 13C NMR (CDCl3): d 21.58 qd
and treated with tert-butyl hydroperoxide (TBHP) (0.59 ml, (3JCP=3.2 Hz), 29.71 t, 31.38 td (3JCP=6.1 Hz), 39.26 td
5.3 mmol) for 4 h. The solvent was removed in vacuo and the (3JCP=7.1 Hz), 55.24 q, 63.88 t, 70.35 dd (2JCP=6.9 Hz), 75.45
residue was purified by flash chromatography eluting with dd and 75.53 dd * (2JCP=6.9, 5.6 Hz), 105.59 d, 118.59 d and
petroleum ether/ethyl acetate (3/1) solution to yield the cyclic 118.67 d *, 125.38 d, 126.31 d, 126.77 d and 126.79 d *, 127.69
phosphate 7 as a brownish oily mixture of four diastereomers d and 127.74 d *, 127.80 d and 127.82 d *, 128.05 d, 128.39 d,
(90 mg, 11%). The chemical shifts of the NMR spectra as 128.66 s, 128.96 d, 129.08 d and 129.11 d *, 133.96 s and 132.99
well as the ESI-MS spectrum and elemental analysis were s *, 136.84 s and 136.88 s *, 144.46 s, 157.13 s and 157.17 s *;
Table I. Half-Lives of the Cyclic Phosphate 7 in 50 mM Buffer Solutions and in Alkaline Phosphatase (ALP) Solution at 37-C (MeanTSD,
n=3 Unless Otherwise Mentioned)
T1/2 (h)
Compound
7
pH 1.0
PH 3.0
pH 5.0
pH 7.4a
pH 7.4b
pH 9.0
ALP
n=2
n=2
68T5.6
73T4.3
73T4.0
78T0.02
74T2.2
78T0.4
Yc
a Phosphate buffer. b Borate buffer. c No degradation was observed during the 24 h incubation.

CYP-activated Cyclic Phosphate Prodrugs
683
O
O
O
OH
P
P
O
O
H2O
O
O
OH
O
O
7
9
Fig. 3. Hydrolysis of the cyclic phosphate 7 in buffer solutions.
31P NMR (CDCl3): d j1.52, j1.39. ESI-MS: m/z = 443.1 (M- surplus from kidney transplantation donors. The collection of
H)j. * Two chemical shifts were detected due to the chiral surplus tissue was approved by the Ethics Committee of the
centres of the molecule 9.
Medical Faculty of the University of Oulu, Oulu, Finland.
The livers were transferred to ice immediately after surgical
excision, cut into pieces, snap-frozen in liquid nitrogen and
stored at j80-C until the microsomes were prepared by
standard differential ultracentrifugation. The metabolic char-
acteristics of these microsomes have been published earlier
(21). A weight-balanced microsomal pool of seven liver
microsomal preparations which have been extensively char-
acterized for primary metabolic screening was employed.
Alkaline phosphatase EC 3.1.3.1 (Type VII-S: from
bovine intestinal mucosa, 2.430 units/mg protein) was pur-
chased from Sigma (St. Louis, MO, USA). Baculovirus-insect
cells expressing human recombinant CYP1A2, CYP3A4,
CYP2A6, CYP2B6, CYP2C9, CYP2C19, CYP2D6 and
CYP2E1 were purchased from Gentest Corp. (Woburn,
MA, USA).
HPLC Analysis
The analysis was performed on the high performance
liquid chromatography (HPLC) system, which consisted of an
Agilent 1100 binary pump (Agilent Technologies Inc.,
Wilmington, DE, USA), a 1100 micro vacuum degasser, a
HP 1050 Autosampler, a HP 1050 variable wavelength
detector (operated at 225 nm) and an Agilent Zorbax SB-
C18 analytical column (4.6 mmꢀ150 mm, 5 mm). The
chromatographic separations were achieved by using iso-
cratic elution of acetonitrile and 50 mM phosphate buffer
(pH 2.0) with a ratio of 55:45 (v/v) at the flow rate 1.0 ml/min
at 25-C. The retention times of analytes were 11.9 min for the
cyclic phosphate prodrug 7, 5.1 min for the parent alcohol
molecule 1 and 2.8 min for the free phosphate 2.
Hydrolysis in Aqueous Solutions
Hydrolysis in Alkaline Phosphatase Solution
The rates of chemical hydrolysis of the cyclic phosphate
7 were determined at 37-C in 50 mM (ionic strength 0.15)
HCl buffer at pH 1.0, citric acid buffer at pH 3.0, acetate
buffer at pH 5.0, phosphate buffer at pH 7.4, borate buffer at
pH 7.4 and at pH 9.0. The incubation mixtures were prepared
by dissolving 12 mM of prodrug 7 in EtOH in preheated
buffer solutions. The EtOH concentration in the incubation
mixtures was 1.5% and the prodrug concentration about 200
mM. The mixtures were incubated at 37-C and the samples
were withdrawn at appropriate intervals. Acetonitrile (ACN)
was added to the samples (1:1, v/v) to hinder further
hydrolysis during the HPLC analyses. The pseudo-first-order
half-lives (t1/2) for the hydrolysis of the prodrug were
calculated from the slope of the linear portion of the plotted
logarithm of remaining prodrug versus time.
Hydrolysis in alkaline phosphatase solution was deter-
mined by incubating mixture (final volume of 1,600 ml) of
50 mM cyclic phosphate 7 in ethanol or 50 mM free phosphate
2 in water, 50 mM Tris-HCl buffer (pH 7.4) and 4 ml (24.3
units) alkaline phosphatase at 37-C. In blank solutions,
alkaline phosphatase was replaced with the same volume of
water. The ethanol concentration in the incubation mixture
was 1% or less. The incubation time was 10 min for the free
phosphate 2 and 24 h for the cyclic phosphate 7 and the
samples were withdrawn at appropriate intervals. The en-
zymatic reaction was terminated by the addition of the same
amount of ice-cold stopping-solution (93% acetonitrile, 7%
orthophosphoric acid) as the sample. The samples were kept
on ice, centrifuged for 15 min at 11000 rpm, and the
supernatant was analyzed by the HPLC. Half-lives (t1/2) for
the hydrolysis of the free phosphate in alkaline phosphatase
solution were calculated from the slope of the linear portion
Biological Material
Human liver samples used in this study were obtained of the plotted logarithm of remaining free phosphate against
from the University Hospital of Oulu (Oulu, Finland) as time.
O
O
O
P
P
OH
O
O
OH
phosphatases
1. CYP450
O
OH
2. ring-opening,
ß-elimination
O
O
O
1
7
2
Fig. 4. The two-step bioconversion of the cyclic phosphate 7 to the free phosphate 2 and further to the alcohol 1 in vitro.

684
Huttunen et al.
6
5
4
3
2
1
0
Half-lives (t1/2) for the enzymatic degradations of the
prodrug 7, the free phosphate 2 and the alcohol 1 at liver
microsomes and recombinant CYP enzymes were calculated
from the slope of the linear portion of the plotted logarithm
of remaining prodrug against time.
RESULTS
Synthesis of Cyclic Phosphate Prodrugs
Two different 1,3-cyclic propyl esters of phosphate,
with and without an aryl ring substituent, were synthe-
sized (Fig. 2). The unsubstituted cyclic phosphate 4 was
obtained at good yields using the phosphorus(V) approach,
but the yields of the cyclic phosphate 7 with the C4-aryl ring
substituent were considerably lower (Fig. 2a). The phospho-
rus(III) approach produced slightly better yields of the cyclic
Enzyme
Fig. 5. The formation of the free phosphate 2 with eight recombinant
human CYP forms.
In vitro Metabolism and Enzyme Kinetic Analyses
The oxidation of the cyclic phosphate 7 was determined phosphate 7 (Fig. 2b). The phosphorus(V) method involved
in isolated human liver microsomes and with eight recombi- the treatment of the alcohol 1 with phosphoryl chlorides 3 or
nant human CYP forms (CYP1A2, CYP2A6, CYP2B6, 6 in the presence of 1-methylimidazole (22). In the phos-
CYP2C9, CYP2C19, CYP2D6, CYP2E1 and CYP3A4). A phorus(III) method, the alcohol 1 was first reacted with
typical incubation mixture, in a final volume of 150Y1,600 ml, chlorophosphane 8 in the presence of diisopropylethylamine
contained 50 mM of cyclic phosphate 7 in ethanol, 50 mM of (DIPEA) and subsequently oxidized with tert-butyl hydro-
Tris-HCl buffer (pH 7.4) and 100 mg microsomal protein or peroxide (TBHP) to the cyclic phosphate 7 (8). The cyclic
5 pmol of recombinant CYP enzyme. Ethanol concentrations phosphoryl chlorides 3 and 6, as well the chlorophosphane 8,
in the incubation mixtures were 1% or less. The reaction was were prepared by treating phosphorus oxychloride or phos-
started by the addition of a NADPH-regenerating system phorus trichloride, respectively, with propane-1,3-diols in the
(1.15 mM NADP, 12.5 mM isocitric acid, 56.25 mM KCl, presence of triethylamine.
187.5 mM Tris-HCl, 12.5 mM MgCl2, 0.0125 mM MnCl2 and
0.77 ml/ml isocitric acid dehydrogenase, pH 7.4) following 2 min Stability of the Cyclic Phosphate Prodrug in Buffer Solutions
of pre-incubation at 37-C. In blank solutions, liver microsomes and in Alkaline Phosphatase Solution
or CYPs were replaced with the same volume of water.
Incubation time ranged from 15 min to 24 h depending on
The chemical stability of the cyclic prodrug 7 in aqueous
the experiment. The samples were withdrawn at appropriate solutions was evaluated at pH 1.0j9.0, with half-lives ranging
intervals and the reactions were terminated by the addition of from 68 to 72 h (Table I). The quantitative hydrolysis
the same amount of ice-cold stopping-solution (93% acetoni- followed first-order kinetics and led to the formation of a
trile, 7% orthophosphoric acid) as the sample. The samples stable ring-opened product 9 (Fig. 3). The enzymatic stability
were kept on ice, centrifuged for 15 min at 11000 rpm, and the of cyclic phosphate 7 was also studied in alkaline phosphatase
supernatant was analyzed by the HPLC.
(ALP) solution, and no degradation was observed during
Enzyme kinetic studies of oxidation of the cyclic 24 h incubation (Table I).
phosphate prodrug 7 by recombinant CYP3A4 and CYP2C19
were conducted in the same conditions as described above Bioconversion of the Cyclic Phosphate in Vitro
using 100 mMY10 mM of cyclic phosphate 7, 1 pmol of
CYP3A4, 5 pmol of CYP2C19 and incubating for 20 min
(CYP3A4) or 240 min (CYP2C19).
The first step in the bioconversion process of the cyclic
phosphate 7 to the corresponding alcohol 1 (Fig. 4), is a CYP-
The bioconversion of the alcohol 1 was determined in catalysed oxidation reaction producing a free phosphate 2 as
human liver microsomes. The incubation mixture contained an intermediate. This reaction was initially studied in isolated
10 mM of alcohol 1 in ethanol, 50 mM of Tris-HCl buffer (pH human liver microsomes and liver microsomes of rats treated
7.4) and 100 mg of microsomal protein. The experiment was with CYP-inducing agents, dexamethasone (CYP3A) and
carried out as described earlier for the cyclic phosphate 7. phenobarbital (CYP2B) or of mice treated with pyrazole
Incubation time was 360 min.
(CYP2A) (23). The most efficient reactions were detected in
Table II. Half-Lives of the Cyclic Phosphate 7, the Free Phosphate 2 and the Alcohol 1 in Human Liver Microsomes (HLM), Human
Recombinant CYP Enzymes (3A4, 2C19) and Alkaline Phosphatase (ALP) Solution (Mean T SD, n=3)
Compound 7
Compound 2
Compound 1
T1/2 (min)
13T0.51
Enzyme
T1/2 (min)
Enzyme
T1/2 (min)
Enzyme
HLM
HLM
CYP3A4
CYP2C19
14T0.57
34T1.7
290T16
ALP
HLM (+NADPH)
2.2T0.0058
37T1.5
43T0.36
HLM (jNADPH)

CYP-activated Cyclic Phosphate Prodrugs
685
Table III. Enzyme Kinetic Parameters for the Cyclic Phosphate 7
cules containing an alcohol functionality. The syntheses of
cyclic phosphates are straightforward, although the variable
yields of synthetic conversion to the cyclic prodrug may offer
challenges depending on the disposition of the parent drug
and the employed phosphorus approach. In this study, the
phosphorus(III) approach produced better yields of the cyclic
phosphate 7 rather than the phosphorus(V) method, al-
though due to steric and inductive effects the overall yields
were quite poor (Fig. 2).
Km
95%
95%
confidence
Recombinant [mM] confidence
ꢀ
enzyme
of 7
intervals VmaxðaÞ intervals Vmax KmðbÞ
3A4
2C19
37
20
17Y70
5Y50
83
0.65
64Y170
0.51Y0.92
2.2
0.033
a mmol free phosphate 2 (pmol CYP)j1 (min)j1
.
b l (pmol CYP)j1
(min)j1
At the beginning, an unsubstituted cyclic phosphate 4
(Fig. 2) was also evaluated as a CYP-selective prodrug.
However, the bioconversion of this cyclic phosphate to the
free phosphate 2 did not occur in the liver microsomes,
probably due to incorrect CYP-catalyzed ring oxidation
(10,13,14) resulting in a rearrangement of phosphate in the
aliphatic chain. This was confirmed by isolating the com-
pound produced by the CYP-catalyzed reaction and charac-
terizing the product by both electrospray ionization quadrupole
ion trap mass spectrometry (ESI-MS) and nuclear magnetic
resonance (NMR) spectrometry. According to MS and MS2
fragmentation spectra, a PO3-group was present in this
compound. Also a naphthalene ring and the O-CH(R)-CH2-
O moiety were detected in 1H and 1H-1H correlated NMR
spectra. To achieve the CYP-oxidation at the desired
carbon followed by efficient prodrug cleavage, we came
to the same conclusion as Erion et al. (2004), i.e. the 1,3-
cyclic propyl ester of phosphate has to possess a C4-aryl
ring substituent (cyclic phosphate 7, Fig. 2) (10). In that case,
the ring of the CYP-oxidized product will open to the form
which subsequently undergoes a b-elimination reaction and
produces the free negatively charged phosphate (Fig. 6).
Due to proper chemical stability of cyclic phosphates in
aqueous solutions and extreme enzymatic stability toward
alkaline phosphatase, cyclic phosphates are potential site-
selective drug carriers and suitable for oral drug delivery.
The hydrolytic rate of the cyclic phosphate 7 in aqueous
buffer solutions at pH 1.0j9.0 was not dependent to any
extent on pH (Table I). Furthermore, the catalytic effect of
phosphate ions on the hydrolysis of the phosphate ester
group at pH 7.4 was not significant.
According to the bioconversion results (Table II, Fig. 4),
the first step of the bioconversion process was catalyzed by
CYP enzymes and the main hepatic CYP form oxidizing the
cyclic phosphate 7 to the free phosphate 2 was CYP3A4
(Table III, Fig. 5). CYP3A4 was also a more efficient enzyme
than CYP2C19 based on Vmax/Km values that were 2.2 and
0.033 l (pmol CYP)j1(min)j1, respectively. These findings
are consistent with the earlier studies on cyclic phosphates of
nucleosides (10) and also with our results obtained with
microsomes from dexamethasone (CYP3A)Ypretreated rats.
The formation of the free phosphate 2 was considerably
affected by the dexamethasone enzyme induction (data not
shown).
isolated human liver microsomes and dexamehasone-pre-
treated rat liver microsomes (data not shown).
CYP enzyme specificity was further evaluated with eight
recombinant human CYP forms (CYP1A2, CYP2A6,
CYP2B6, CYP2C9, CYP2C19, CYP2D6, CYP2E1 and
CYP3A4). The human liver microsomes and recombinant
CYP3A4 and CYP2C19 produced the free phosphate 2 from
the cyclic phosphate 7, while CYP1A2, CYP2A6, CYP2B6,
CYP2C9, CYP2D6 and CYP2E1 did not catalyze this
reaction (Fig. 5).
The oxidation of the cyclic phosphate 7 occurred in
human liver microsomes and recombinant CYP3A4 and
CYP2C19 with half-lives of 14, 34 and 290 min, respectively
(Table II). The formation rates of free phosphate 2 were
5.0 mmol (mmol enzyme)j1 (min)j1 with CYP3A4 and 1.6 mmol
(mmol enzyme)j1 (min)j1 with CYP2C19. Enzyme kinetic
parameters, Km and Vmax, were determined for the cyclic
phosphate 7 using recombinant CYP3A4 and CYP2C19. Four
separate experiments were analyzed by non-linear regression
using GraphPad Prism software (Table III).
The second step of the bioconversion process of the
cyclic phosphate 7 (Fig. 4) is the hydrolysis of the free
phosphate 2 to the alcohol 1 catalyzed by phosphatases. The
reaction was initially studied in alkaline phosphatase solution
with a half-life of 2.2 min (Table II), releasing the alcohol 1
quantitatively. The hydrolysis of the free phosphate 2 was
also evaluated in human liver microsomes with and without
NADPH, an electron source for the CYP-catalyzed reac-
tions, with half-lives of 37 and 43 min, respectively (Table II).
In the human liver microsomes, the bioconversion of cyclic
phosphate prodrug 7 proceeded through two phases catalyzed
by CYP and phosphatase enzymes, respectively, and the
formation of the free phosphate 2 preceded the release of
the alcohol 1 (Fig. 4). However, to detect also the parent
molecule 1, a long incubation time was needed. The parent
molecule, the alcohol 1 was furthermore degraded in the
human liver microsomes with half-life of 13 min (Table II).
DISCUSSION
The cyclic phosphate prodrug approach offers a very
feasible liver-targeted drug delivery strategy for drug mole-
3
O
OH
O
O
O
O
Ar
O
O
CYP 450
ß-elimination
P2
P
Ar
P
O
P
O
4
5
+
O
O
O
O
R
R
R
R
1
O
O
O
Ar
Ar
O
6
O
Fig. 6. Proposed CYP-catalyzed bioconversion mechanism of cyclic phosphate prodrugs.

686
Huttunen et al.
Based on the half-lives of CYP-catalyzed oxidation of ACKNOWLEDGEMENTS
the cyclic phosphate 7 and dephosphorylation of the free
phosphate 2 (Table II, Fig. 4), the formation of the free
We thank Mrs. Katja Ho¨tti and Mrs. Hannele Jaatinen for
phosphate 2 appeared to be the slowest step of the bio- valuable technical assistance, M.Sc. Marko Lehtonen and
conversion process. Since the hydrolysis of the free phos- M.Sc. Pekka Keski-Rahkonen for skillful HPLC guidance and
phate 2 occurred also without NADPH in human liver Dr. Seppo Auriola for providing the excellent MS and MS2
microsomes, the dephosphorylation was apparently catalyzed fragmentation spectra. We also thank Prof. Olavi Pelkonen
by microsomal dephosphorylative enzymes rather than via and Dr. Miia Turpeinen (Department of Pharmacology and
CYP enzymes. The slow formation rate of the free phosphate Toxicology, University of Oulu, Finland) for providing the
2 is probably a consequence of a reversible reaction of ring human liver microsomal samples. The work was financially
opening after oxidation reaction of the cyclic phosphate 7 supported by the National Technology Agency of Finland
(Fig. 6) (10). The slow ring opening reaction, and conse- (Tekes), the Academy of Finland (grants #106313, J.L.;
quently, the long prodrug conversion time can most likely be #204755, J.V.; #204905, J.R.) and The Graduate School of
overcome by using an electron withdrawing group, such as Bioorganic and Medicinal Chemistry.
chlorine, in the C4-aryl ring substituent (10Y12). Further-
more, the prodrug byproduct of the CYP-catalyzed and
followed spontaneous reactions, aryl vinyl ketone (Fig. 6)
associated with toxicity, is rapidly and quantitatively detox-
REFERENCES
ified with intracellular glutathione in the liver (11,12).
In the experiments with human liver microsomes, the
parent alcohol 1 was found to be unstable in the presence of
learned from marketed and investigational prodrugs. J. Med.
1. P. Ettmayer, G. L. Amidon, B. Clement, and B. Testa. Lessons
active CYP enzymes and the formation of the alcohol 1 from
the cyclic phosphate prodrug 7 was not entirely quantitative.
The methoxy group of the alcohol 1 was most likely oxidized
by CYP enzymes (Table II) (24). Therefore, each cyclic
phosphate prodrug of distinct parent drug molecule should
undergo full metabolic and toxicological evaluation during
the phase of prodrug development. Optical purity of cyclic
phosphate prodrugs should also be taken into account when
evaluating the pharmacological properties of the real parent
drugs. In this study unseparated diastereomeric cyclic phos-
phates were used to evaluate the prodrug bioconversion and
it showed to be insensitive to the absolute stereochemistry at
the C4-position.
Chem. 47:2393Y2404 (2004).
2. K. Petrak. Essential properties of drug-targeting delivery
systems. Drug Discov. Today 10:1667Y1673 (2005).
3. P. C. Rensen, R. L. De Vrueh, J. Kuiper, M. K. Bijsterbosch, E.
A. Biessen, and T. J. Van Berkel. Recombinant lipoproteins:
lipoprotein-like lipid particles for drug targeting. Adv. Drug
Deliv. Rev. 47:251Y276 (2001).
4. M. Nishikawa. Development of cell-specific targeting systems
for drugs and genes. Biol. Pharm. Bull. 28:195Y200 (2005).
5. G. Payne. Progress in immunoconjugate cancer therapeutics.
Cancer Cell 3:207Y212 (2003).
6. S. M. Moghimi, A. C. Hunter, and J. C. Murray. Nano-
medicine: current status and future prospects. FASEB J.
19:311Y330 (2005).
7. F. P. Guengerich. Common and uncommon cytochrome P450
reactions related to metabolism and chemical toxicity. Chem.
Res. Toxicol. 14:611Y650 (2001).
8. C. Meier, M. Lorey, and E. De Clercq. cycloSal-2¶,3¶-dideoxy-
2¶,3¶-didehydrothymidine monophosphate (cycloSal-d4TMP):
synthesis and antiviral evaluation of a new d4TMP delivery
system. J. Med. Chem. 41:1417Y1427 (1998).
9. C. Meier, U. Go¨rbig, C. Mu¨ller, and J. Balzarini. cycloSal-
PMEA and cycloAmb-PMEA: potentially new phosphonate
prodrugs based on the cycloSal-pronucleotide approach. J.
Med. Chem. 48:8079Y8086 (2005).
Our results clearly indicate that the quite fast and CYP-
catalyzed cleavage combined with high stability outside of
the liver mean that these kinds of cyclic phosphate prodrugs
could be considered as potential candidates to achieve in-
creased drug levels in the liver.
CONCLUSIONS
10. M. D. Erion, K. R. Reddy, S. H. Boyer, M. C. Matelich, J.
Gomez-Galeno, R. H. Lemus, B. G. Ugarkar, T. J. Colby, J.
Schanzer, and P. D. Van Poelje. Design, synthesis, and
characterization of a series of cytochrome P(450) 3A-activated
prodrugs (HepDirect prodrugs) useful for targeting phos-
ph(on)ate-based drugs to the liver. J. Am. Chem. Soc.
126:5154Y5163 (2004).
11. M. D. Erion, P. D. Van Poelje, D. A. Mackenna, T. J. Colby,
A. C. Montag, J. M. Fujitaki, D. L. Linemeyer, and D. A.
Bullough. Liver-targeted drug delivery using HepDirect pro-
drugs. J. Pharmacol. Exp. Ther. 312:554Y560 (2005).
The 1,3-cyclic propanyl phosphate ester with a C4-aryl
ring substituent was evaluated in vitro as a potential means
for CYP-mediated drug delivery for a representative mole-
cule containing an alcohol functionality. The results con-
firmed that the cyclic phosphate prodrug is oxidized by liver
microsomal CYP enzymes, primarily by CYP3A4 and
CYP2C19, to the free phosphate and dephosphorylated
further by microsomal phosphatases to the parent alcohol. 12. M. D. Erion, D. A. Bullough, C.-C. Lin, and Z. Hong.
HepDirect prodrugs for targeting nucleotide-based antiviral
Consequently, this cyclic phosphate prodrug approach is a
drugs to the liver. Curr. Opin. Investig. Drugs 7:109Y117 (2006).
novel extension to the cyclic phosphate prodrugs of nucleo-
sides and represents an efficient drug delivery strategy for
13. D. Farquhar, N. J. Kuttesch, M. G. Wilkerson, and T. Winkler.
Synthesis and biological evaluation of neutral derivatives of 5-
targeting drug molecules containing the alcohol functionality
to the liver or intestine. Since many drugs possess an alcohol
functionality and exert their action specifically in the liver or
intestine, this site-specific prodrug approach could serve as a
potential method to increase the therapeutic efficacy and/or
to decrease adverse effects of these drugs.
fluoro-2¶-deoxyuridine 5¶-phosphate. J. Med. Chem. 26:1153Y
1158 (1983).
14. D. Farquhar and R. Smith. Synthesis and biological evaluation of 9-
[5¶-(2- oxo-1,3,2-oxazaphosphorinan-2-yl)-b-D-arabinosyl]adenine
and 9-[5¶-(2-oxo-1,3,2- dioxaphosphorinan-2-yl)-b-arabinosyl]ade-
nine: potential neutral precursors of 9-[b-D- arabinofuranosyl]ade-
nine 5¶-monophosphate. J. Med. Chem. 28:1358Y1361 (1985).

CYP-activated Cyclic Phosphate Prodrugs
687
15. F. Kesisoglou and E. M. Zimmermann. Novel drug delivery
strategies for the treatment of inflammatory bowel disease.
Expert Opin. Drug Deliv. 2:451Y463 (2005).
16. D. W. Moss. Alkaline phosphatase isoenzymes. Clin. Chem.
28:2007Y2016 (1982).
spectral analysis of NMR spectra using integral-transform
iterator. J. Magn. Reson. 120:1Y10 (1996).
21. M. Turpeinen, J. Uusitalo, J. Jalonen, and O. Pelkonen. Multiple
P450 substrates in a single run: rapid and comprehensive in vitro
interaction assay. Eur. J. Pharm. Sci. 24:123Y132 (2005).
17. M. Asada-Kubota and S. Kanamura. Intracellular localization of 22. G. Singh and H. Vankayalapati. A new glycosylation strategy for
alkaline phosphatase in freshly isolated foetal rat hepatocytes.
the synthesis of mannopyranosides. Tetrahedron: Asymmetry
11:125Y138 (2000).
18. A. S. Kearney and V. J. Stella. The in vitro enzymic labilities of 23. O. Pelkonen, J. Ma¨enpa¨a¨, P. Taavitsainen, A. Rautio, and H.
Histochem. J. 18:500Y506 (1986).
chemically distinct phosphomonoester prodrugs. Pharm. Res.
Raunio. Inhibition and induction of human cytochrome P450
9:497Y503 (1992). (CYP) enzymes. Xenobiotica 28:1203Y1253 (1998).
19. M. Safadi, R. Oliyai, and V. J. Stella. Phosphoryloxymethyl 24. M. Nobilis, J. Kopecky, J. Kvetina, Z. Svoboda, M. Pour, J.
carbamates and carbonates-novel water-soluble prodrugs for
amines and hindered alcohols. Pharm. Res. 10:1350Y1355
(1993).
Kunes, M. Holcapek, and L. Kolarova. Comparative biotrans-
formation and disposition studies of nabumetone in humans and
minipigs using high-performance liquid chromatography with
ultraviolet, fluorescence and mass spectrometric detection. J.
Pharm. Biomed. Anal. 32:641Y656 (2003).
20. R. Laatikainen, M. Niemitz, U. Weber, J. Sundelin, T. Hassinen,
and J. Vepsa¨la¨inen. General strategies for total-lineshape-type