European Journal of Pharmaceutical Sciences 22 (2004) 25–31
`
`Stereospecific chemical and enzymatic stability of phosphoramidate
`triester prodrugs of d4T in vitro
`Dario Siccardi a,b, Mark Gumbleton b, Yadollah Omidi b, Chistopher McGuigan a,∗
`
`a Medicinal Chemistry Group, Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff CF10 3XF, UK
`b Pharmaceutical Cell Biology Group, Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff CF10 3XF, UK
`
`Received 2 July 2003; received in revised form 10 February 2004; accepted 17 February 2004
`
`Abstract
`
`The phosphoramidate triester prodrug approach is widely used to deliver nucleotide forms of nucleoside analogues into target cells. We
`(cid:2)
`(cid:2)
`(cid:2)
`(cid:2)
`,3
`-didehydro-2
`,3
`-dideoxythymidine (d4T). Chemical
`investigated the stereoselective stability of a series of prodrugs of the anti-HIV agent 2
`stability was evaluated in phosphate buffer at pH values of biological relevance (i.e. pH 2.0, 4.6, 7.4). Enzymatic stability was tested in
`human plasma, in Caco-2 cell homogenates and monolayers and in rat liver. The compounds were relatively stable to chemical hydrolysis.
`Between 50 and 70% of unchanged prodrug was recovered after 16 h incubation in human plasma, with no stereoselective preference for
`phosphate diastereoisomers. The p-OMe phenyl derivative, however, was an exception and only 5% of one diastereoisomer was recovered. In
`Caco-2 cells the stability and stereoselectivity largely depended on the experimental conditions: high enzymatic activity and stereoselectivity
`was observed in cell homogenates, but not in monolayers. In rat liver S9 fractions the stability profile was similar to that in Caco-2 cells
`and carboxyl ester cleavage appeared to be the sole mechanism of degradation in both media. The large and unpredictable differences in
`stereoselective metabolic rate of the pronucleotide series here presented suggest that in vivo circulating levels of intact prodrug could exert
`profoundly different activity or toxicity due to preferential body distribution of one diastereoisomeric form.
`© 2004 Elsevier B.V. All rights reserved.
`
`Keywords: Anti-HIV pronucleotides; d4T; Ester prodrugs; Caco-2; Antivirals
`
`1. Introduction
`
`The phosphoramidate triester approach is a prodrug
`methodology (“pronucleotide”) increasingly used to im-
`prove the pharmacological activity of existing nucleoside
`analogues (Wagner et al., 2000). The principal objective
`is to boost the activity of those nucleoside analogues that
`exhibit inefficient intracellular phosphorylation by host cell
`or virally-encoded nucleoside kinases to produce the active
`triphosphorylated form. The pronucleotide approach is based
`on the ability to deliver to the target cell the monophos-
`phorylated form of a variety of nucleoside analogues. The
`phosphate group is masked with neutral lipophilic groups
`to obtain a lipophilic membrane-permeable derivative able
`to access intracellular target sites, wherein hydrolysis yields
`the free nucleoside monophosphate.
`
`Corresponding author. Tel.: +44-29-2087-4537;
`∗
`fax: +44-29-2087-4537.
`E-mail address: mcguigan@cf.ac.uk (C. McGuigan).
`
`0928-0987/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
`doi:10.1016/j.ejps.2004.02.006
`
`In the case of the phosphoramidate derivatives of the
`(cid:2)
`(cid:2)
`(cid:2)
`(cid:2)
`2
`,3
`-dideoxy-2
`,3
`-didehydro analogue of thymidine, d4T
`(stavudine), the ability to release the free mononucleotide
`has been proposed to depend on carboxylesterase-mediated
`hydrolysis of the carboxylic ester function in the amino acid
`moiety (Balzarini et al., 1996). Spontaneous elimination of
`the phenol and an enzymatic cleavage of the P–N bond will
`then release d4TMP (Fig. 1) (Saboulard et al., 1999). For
`optimal pharmacological action the activation to d4TMP
`must occur within the target cell, but not during the process
`of absorption and distribution. Previous studies investigat-
`ing the activation pathway of a series of d4T and AZT
`pronucleotides in different biological media (i.e. CEM cell
`extracts, human serum, mouse serum) have shown different
`degrees of stability with variation in the amino acid moi-
`ety, and the carboxyl ester group (Saboulard et al., 1999).
`The alaninyl phosphoramidate triester derivatives were effi-
`ciently converted to the free nucleoside monophosphate in
`intact cells (Balzarini et al., 1996). In this current work we
`have investigated the stability to chemical and enzymatic
`
`IPR2018-00126
`
`Page 1 of 7
`
`I-MAK 1021
`
`

`

`26
`
`D. Siccardi et al. / European Journal of Pharmaceutical Sciences 22 (2004) 25–31
`
`O
`
`_
`O P
`
`ONuc
`
`NH
`
`CH
`
`Me
`
`COOH
`
`AAM
`
`Phosphoramidase
`
`ONuc
`
`PO
`
`NH
`
`O
`
`O
`
`Me
`
`Esterase-
`mediated
`activation
`
`ONuc
`
`O P O
`
`_
`
`_
`
`O
`
`X
`
`O
`
`O P
`
`ONuc
`
`NH
`
`CH
`
`C
`
`Me
`
`O
`
`O Me
`
`X
`
`logPa
`
`1.04
`1.09
`1.19
`1.43
`1.60
`1.93
`
`Hp
`
`-Cl
`p-Me
`p-OMe
`p-Br
`p-I
`
`So324
`Cf1523
`Cf1525
`Cf1526
`Cf1517
`Cf1572
`
`Fig. 1. The aryloxy phosphoramidate prodrugs (pronucleotides) of d4T examined in this study and the proposed activation pathway. The cleavage of the
`methoxy group is considered the first step in the enzymatic hydrolysis of the compounds and is followed by spontaneous cyclization and displacement
`of the aryloxy group and formation of the amino acyl metabolite (AAM); the last activation step is catalyzed by intracellular phosphoramidase. Chirality
`at the P atom of the pronucleotide generates two diastereoisomers. Taken from Siddiqui et al. (1999).
`
`ester hydrolysis of a series of pharmacologically active
`alaninyl phosphoramidate analogues of d4T (Fig. 1) in hu-
`man plasma, rat liver and in different Caco-2 cell preparates,
`since esterase mediated degradation in the GI tract can be
`particularly critical in limiting oral absorption of ester pro-
`drugs (Van Gelder et al., 2000a,b, Annaert et al., 1998). The
`phosphoramidates are chiral at phosphorus and prepared
`as roughly 50:50 mixtures of phosphate diastereoisomers.
`The aspect of phosphate stereochemistry on the process of
`activation was also, for the first time, herein considered.
`
`2. Materials and methods
`
`2.1. Materials
`
`All reagents used were of analytical grade. Caco-2 cells
`were obtained from the European Collection of Animal
`Cell Cultures (ECACC). Phosphate buffered saline pH 7.4
`(PBS), Dulbecco’s modified Eagle’s medium (DMEM) and
`penicillin/streptomycin were from Life Technologies (Pais-
`ley, UK). Tissue culture plastics were from Corning–Costar
`(Bucks, UK). D4T phosphoramidate prodrugs were syn-
`thesised according to previously published procedures
`(McGuigan et al., 1996). Pronucleotides solutions were
`prepared by spiking DMEM with a concentrated stock of
`prodrug dissolved in DMSO. The final concentration of
`DMSO within solutions used for experiments was always
`adjusted to 1%.
`
`2.2. Chemical stability
`
`Pronucleotide stability to chemical hydrolysis was studied
`in 0.05 M NaH2PO4 at pH 2.0, 4.6, and 7.4. An accurately
`
`weighted amount of the investigated compounds was dis-
`solved in DMSO, and then diluted with the respective phos-
`phate solution to final concentration of 0.2 mM of pronu-
`◦
`cleotide. Samples were incubated in closed vials at 37
`C
`for 20 h in the dark. Solutions were withdrawn at the end of
`the time period and stored at −80
`◦
`C or immediately ana-
`lyzed by HPLC. The storage at −80
`◦
`C for up to 1 month
`was found not to alter pronucleotide stability.
`
`2.3. Degradation of pronucleotides in plasma
`
`Pronucleotide stability was tested in plasma prepara-
`tions obtained from healthy volunteers. The incubation
`mixture contained 160 ␮l of plasma, 20 ␮l of drug stock
`solution at the concentration of 5 mM and 20 ␮l of MgCl2
`(25 mM) in phosphate buffer (pH 7.4). At
`the end of
`◦
`C) 200 ␮l of
`each incubation period (0, 3 or 16 h at 37
`ice-cold methanol/acetonitrile 1/1 (v/v) was added,
`the
`tubes were vortex-mixed for 20 s to precipitate proteins, left
`on ice for 30 min and centrifuged for 5 min at 15,000 × g.
`The supernatant was immediately analyzed by HPLC or
`stored at −80
`C until analysis. The storage at −80
`◦
`◦
`C
`for up to 1 month was found not to alter pronucleotide
`stability.
`
`2.4. Cell culture
`
`Caco-2 cells were maintained in DMEM supplemented
`with 10% fetal bovine serum, 100 U/ml penicillin and
`100 ␮g/ml streptomycin. The medium was changed every
`48 h. Cells were seeded at 50,000/cm2 in tissue culture
`◦
`flasks and cultured until confluent at 37
`C in 90% relative
`humidity and 5% CO2. Cells between passage numbers
`30 and 40 were used for degradation experiments. The
`
`IPR2018-00126
`
`Page 2 of 7
`
`I-MAK 1021
`
`

`

`D. Siccardi et al. / European Journal of Pharmaceutical Sciences 22 (2004) 25–31
`
`27
`
`cells were detached from the flasks by washing with PBS,
`followed by 0.5% trypsin and 0.2% EDTA in PBS.
`
`2.5. Degradation in Caco-2 cell homogenates, suspensions
`and monolayers
`
`Caco-2 cells were grown on culture flasks until conflu-
`ence, detached, resuspended in DMEM and counted. Sus-
`pended cells were pelleted by centrifugation, resuspended in
`PBS and washed two more times. Cell pellets were stored at
`−80
`◦
`C. On the day of the experiment cells were diluted in
`PBS containing MgCl2 (10 mM) to a density of 2×106 cells
`per 100 ␮l and sonicated in iced water for 60 s. The incuba-
`tion mixture contained 100 ␮l of cell homogenate and 100 ␮l
`of drug stock solution 1 mM in PBS containing MgCl2
`◦
`(10 mM). Reaction mixtures were incubated at 37
`C and at
`fixed time-points 200 ␮l of ice-cold methanol/acetonitrile
`1/1 (v/v) was added, the tubes were vortex-mixed for 20 s,
`left on ice for 30 min and centrifuged for 5 min at 15,000×g.
`The supernatant was immediately analyzed by HPLC or
`stored at −80
`C until analysis. The storage at −80
`◦
`◦
`C for
`up to 1 month was found not to alter pronucleotide stability.
`For inhibition studies homogenates were preincubated for
`◦
`30 min at 37
`C in the presence of different concentrations of
`the carboxylesterase inhibitor phenylmethylsulfonyl fluoride
`(PMSF) before the addition of the pronucleotide solution.
`For the preparation of cell suspensions Caco-2 cells were
`grown on culture flasks until confluence, harvested, washed
`twice in PBS, counted and aliquoted at the concentration
`of 2 × 106 cells per 100 ␮l media. After the addition of
`100 ␮l of 1 mM pronucleotide stock solution the vials were
`◦
`placed on an orbital shaker and left at 37
`C. The incu-
`bation reaction was stopped by adding 200 ␮l of ice-cold
`methanol/acetonitrile 1/1 (v/v), followed by sonication and
`centrifugation at 15,000×g for 10 min. The supernatant was
`analyzed by HPLC. When not analyzed immediately, the
`supernatant was stored at −80
`◦
`C. Under these conditions
`pronucleotides were found to be stable for up to 1 month.
`Uptake and degradation was examined in Caco-2 cells
`cultured for 17 days in plastic wells. Before the experiments,
`the culture media was removed, the monolayers were rinsed
`and preincubated with 1 ml warm DMEM. After 15 min the
`media was removed and replaced with 1 ml DMEM con-
`taining the pronucleotide at a concentration of 0.5 mM. At
`0, 1, 4, or 16 h time point media samples were collected
`and stored. Monolayers were washed twice with ice-cold
`PBS and then exposed to ice-cold methanol/acetonitrile 1/1
`(v/v), the cell lysate harvested, sonicated and centrifuged
`for 10 min at 15,000 × g. The supernatant was analyzed by
`HPLC.
`
`nizing vessel on ice and homogenized with a motor-driven
`Teflon pestle in thee volumes of cold phosphate buffer
`50 mM pH 7.6. The homogenate was centrifuged at 9000×g
`◦
`C) Sorvall RC-5B centrifuge
`for 30 min in a refrigerated (2
`(Dupont, Newtown, CT). The supernatant (S9), containing
`the cytosolic and microsomal fractions was isolated. Pro-
`tein concentration in the S9 fractions was measured using
`a Bio-Rad RC DC colorimetric assay (Bio-Rad, Hercules,
`CA) using absorbance at 750 nm. Each incubation mixture
`contained 1 mg protein diluted in phosphate buffer 50 mM
`pH 7.6 containing MgCl2 10 mM and NADPH 1.0 mM.
`Standard solutions in DMSO were then added at the final
`concentration of 0.25 mM of pronucleotide and less than
`1% DMSO. Controls were prepared with no NADPH added
`and immediately extracted with acetonitrile. For inhibition
`◦
`studies samples were preincubated for 30 min at 37
`C in
`the presence of different concentrations of PMSF before the
`addition of the pronucleotide solution. The samples were in-
`◦
`cubated at 37
`C for different time-points. The reaction was
`terminated by placing the flasks on ice followed by the ad-
`dition of an equal volume of ice-cold methanol/acetonitrile
`1/1 (v/v). Samples were spun at 5000 × g for 10 min at
`C and the supernatant was stored at −20
`◦
`◦
`20
`C until HPLC
`analysis.
`
`2.7. HPLC Analysis of d4T and phosphoramidate prodrugs
`
`The concentration of phosphoramidate prodrugs in the
`incubation media was determined by HPLC. A Thermo
`Separations Products HPLC system was used, consisting
`of P4000 Quaternary Pump, autosampler, column oven and
`a diode array detector. The column used was a Merck Li-
`ChoCart Lichospher C18 (250 mm × 4 mm). Elution was
`◦
`C isocratically using a mobile phase wa-
`performed at 30
`ter/methanol/acetonitrile 60/25/15 (v/v/v) and at flow rate
`of 1 ml/min; the injection volume was 50 ␮l. Detection
`wavelength was 266 nm. The compounds were identified
`according to peak retention times and UV spectra. The
`concentrations of test compounds were determined using
`calibration graph of each compound: standard curves, ob-
`tained by linear regression, were linear (r2 > 0.99). The
`relative standard deviation for within-day and between-day
`precision was assessed by analyzing each day for 5 days
`(n = 5) a set of two concentration levels, and was less than
`5% for each compound. The accuracy was determined by
`subtracting the measured concentration from its theoretical
`value; the mean relative error (M.R.E.) of the difference
`from theoretical was less than 10%.
`
`2.8. Statistical analysis
`
`2.6. Preparation of S9 fractions
`
`Male WISTAR rats weighing 200–250 g were sacrificed.
`Their livers were removed and washed in PBS. Tissue was
`finely minced with scissors, transferred to glass homoge-
`
`Comparisons between two groups were made by
`non-paired student t-test with level of significance at P <
`0.05. Comparisons for more than two groups were made by
`ANOVA and Duncan’s multiple range test with P <0 .05
`significance level.
`
`IPR2018-00126
`
`Page 3 of 7
`
`I-MAK 1021
`
`

`

`28
`
`D. Siccardi et al. / European Journal of Pharmaceutical Sciences 22 (2004) 25–31
`
`Table 1
`Stability of pronucleotides in phosphate buffer: recovery (%) of drug after 20 h incubation at 37
`
`◦
`
`C
`
`pH 2.0
`
`pH 4.6
`
`pH 7.4
`
`SEa
`FEa
`99.9 ± 2.9
`99.8 ± 2.4
`H
`87.5 ± 3.7
`93.1 ± 1.0
`p-OMe
`93.8 ± 3.1
`94.7 ± 3.0
`p-Me
`96.5 ± 2.6
`95.1 ± 5.0
`p-Cl
`86.1 ± 3.8
`85.9 ± 3.8
`p-Br
`80.3 ± 3.0
`80.2 ± 3.0
`p-I
`Values are mean ± S.D. of triplicate assays.
`a FE and SE refer to the two diastereoisomers with stereochemical variation on the phosphorus atom (fast eluting, FE and slow eluting, SE).
`
`FE
`99.7 ± 8.0
`91.2 ± 5.6
`93.8 ± 3.9
`87.9 ± 2.9
`80.7 ± 6.0
`79.1 ± 5.0
`
`SE
`99.7 ± 12.7
`90.3 ± 5.3
`94.6 ± 1.8
`83.3 ± 3.1
`77.3 ± 5.8
`73.3 ± 5.5
`
`FE
`99.6 ± 7.3
`91.1 ± 3.7
`98.3 ± 2.9
`99.6 ± 2.1
`99.8 ± 11.1
`86.0 ± 3.0
`
`SE
`99.7 ± 3.1
`89.0 ± 2.0
`98.3 ± 3.9
`99.1 ± 2.0
`99.1 ± 10.9
`87.1 ± 3.8
`
`SE diastereoisomer of CF1526 was significantly (P <0 .05)
`less stable than any other compound analyzed. The addition
`of PMSF (1.0 mM) inhibited, although not completely, the
`degradation of this pronucleotide: after 20 h incubation the
`percent recovery of the FE diastereoisomer was 58.9±0.8%
`(mean ± S.D.), while no SE diastereoisomer could be de-
`tected in the incubation media. In the treatments containing
`PMSF, 82.3 ± 1.7% and 78.4 ± 1.7% of FE and SE di-
`astereoisomers, respectively, were recovered. The degrada-
`tion profile for this compound after 3,16, and 20 h incuba-
`tion is shown in Fig. 2.
`In Caco-2 cell homogenates (Fig. 3A) all of the phos-
`phoramidate prodrugs underwent substantial degradation
`over the 1 h incubation period with significant differences
`between the phosphate diastereoisomers. Both diastereoiso-
`mers of the p-I derivative were completely degraded, as
`were the SE diastereoisomers of the p-Br and p-Cl deriva-
`tives. When incubating for a shorter time (i.e. 30 min), the
`FE and SE recovered were (mean ± S.D.) 35.13 ± 7.1%,
`10.90 ± 2.1% for the p-I derivative, and 32.30 ± 0.7,
`11.4 ± 1.5% for the p-Br derivative, respectively.
`In order to assess the dependence of the degradation in
`Caco-2 upon esterase activity, the prototype prodrug So324
`(final concentration 500 ␮M) was incubated for up to 4 h
`with Caco-2 homogenates. In this media So324 was ex-
`tensively metabolized, half-life was 1.5 and 0.26 h for the
`FE and SE diastereoisomer, respectively; 16% of FE di-
`astereoisomer could be recovered unchanged at the end of
`the incubation period, while no detectable SE diastereoiso-
`
`FE
`SE
`
`3 hr
`
`16 hr
`
`20 hr
`
`20 hr + PMSF
`
`100
`90
`80
`70
`60
`50
`40
`30
`20
`10
`0
`
`% Recovery
`
`Fig. 2. The p-OMe derivative’s fast eluting (FE) and slow eluting (SE)
`◦
`diastereomers recovered after incubation in human plasma at 37
`C for
`3, 16, and 20 h in absence and presence of PMSF. Bars are mean of two
`experiments performed in triplicate, with C.V. <5%.
`
`3. Results
`
`The pronucleotides were quantified by reverse phase
`HPLC-UV detection. The phosphate diastereoisomers were
`efficiently separated and were assigned as fast eluting (FE,
`less lipophilic) and slow eluting (SE, more lipophilic),
`according to the retention times.
`Chemical stability was studied at experimental condi-
`tions of biological relevance, i.e. at pH 2.0, 4.6, and 7.4
`(Table 1). Stability was only slightly affected by the pH over
`the 20 h incubation period. The p-Cl, p-Br and p-I deriva-
`tives were more stable at pH 4.6 as compared to pH 2.0 and
`7.4 (P <0 .05). Differences in stability between the FE and
`SE diastereoisomers were not statistically significant for any
`pronucleotide examined.
`The stability of the d4T pronucleotides was also studied in
`human plasma (Table 2). The amount of unchanged prodrug
`remaining after incubation in plasma samples for up to 20 h
`was calculated against a set of controls (n = 3) extracted
`from plasma samples at t0 in order to account for any pos-
`sible loss of sample during analysis due to protein binding.
`The mean extraction efficiency of the pronucleotides from
`plasma samples was approximately 95% with a C.V. never
`greater than 13%. Statistically significant (P <0 .05) differ-
`ences in stability in plasma between diastereoisomers were
`only observed for the p-OMe (CF1526) derivative, with a
`half-life of 25.9 and 3.8 h for FE and SE, respectively. The
`
`Table 2
`Stability of d4T and phosphoramidate prodrugs in plasma: percentage
`◦
`recovery of drug after 3 and 16 h incubation at 37
`C
`3 h (100.19 ± 0.72)a
`16 h (85.52 ± 3.72)a
`FEb
`SEc
`FE
`SE
`85.6 ± 5.3
`90.2 ± 6.1
`73.1 ± 3.8
`71.1 ± 5.8
`H
`97.2 ± 3.7
`58.6 ± 2.3
`70.2 ± 2.3
`5.5 ± 0.8
`p-OMe
`94.8 ± 4.9
`94.1 ± 5.0
`58.6 ± 3.9
`55.5 ± 5.0
`p-Me
`92.1 ± 1.3
`81.3 ± 5.1
`65.2 ± 2.1
`57.5 ± 4.0
`p-Cl
`93.0 ± 8.5
`89.8 ± 8.0
`61.9 ± 2.1
`68.1 ± 14.2
`p-Br
`93.8 ± 1.4
`91.8 ± 1.9
`53.1 ± 4.0
`53.5 ± 4.4
`p-I
`Values are mean ± S.D. of triplicate assays.
`a d4T.
`b Fast eluting diastereoisomer (FE).
`c Slow eluting diastereoisomer (SE).
`
`IPR2018-00126
`
`Page 4 of 7
`
`I-MAK 1021
`
`

`

`D. Siccardi et al. / European Journal of Pharmaceutical Sciences 22 (2004) 25–31
`
`29
`
`series was then compared in a 4 h incubation experiment
`(Fig. 3B); no statistical difference between either SE di-
`astereoisomers within the group, or the FE diastereoisomers
`within the group were noticed. Comparisons of the respec-
`tive FE and SE diastereoisomers showed that for only the
`p-OMe and p-I derivatives was there any statistically signif-
`icant difference, with the FE diastereoisomer less stable in
`both cases.
`The pronucleotides were also exposed to Caco-2 mono-
`layers at 17 days post seeding to assess if differences in en-
`zymatic expression of differentiated monolayers would af-
`fect the decomposition profile. In a 4 h incubation exper-
`iment differences between FE and SE diastereoisomers of
`the respective p-Me, p-Cl, p-Br, p-I derivatives were statisti-
`cally significant. The half-life of So324 calculated from the
`concentration of pronucleotide recovered in the incubation
`media and intracellularly was 14.9 and 16.5 h, for the FE
`and SE diastereoisomer, respectively. After 4 h the amount
`recovered was 91.6 ± 7.0 and 91.3 ± 4.8% (Fig. 3C). At all
`time-points the fraction of So324 recovered intracellularly
`was less than 2% of the total amount recovered.
`Comparative stability data between the cell suspensions
`and the cell monolayers showed no clear trend for the stabil-
`ity of the pronucleotides. Interestingly, opposite stereoselec-
`tive degradation of the p-Cl, p-Br and p-I derivatives resulted
`in the monolayer and suspension incubations as compared
`with the homogenates (Fig. 3).
`The stability of the prototype prodrug So324 was further
`evaluated in rat hepatic S9 fraction to assess the substrate
`specificity for CYP450 and hepatic hydrolases. The degra-
`dation profile in rat preparations was similar to that ob-
`served in Caco-2. The FE diastereoisomer was more stable
`than the SE diastereoisomer (Fig. 4). Half-lives were calcu-
`lated at 17 and 0.6 min for FE and SE, respectively. Incu-
`bation in the presence of 1 mM PMSF significantly inhib-
`ited the degradation. For example, at the 90 min time-point
`the amount of So324 recovered in samples treated with
`PMSF was 92.9 ± 7.6 and 87.4 ± 3.3% of the initial con-
`centration of the FE and SE diastereoisomer, respectively,
`
`Fig. 4. Enzymatic degradation profile of the FE and SE diastereoisomers
`◦
`C. Disappearance
`(0.125 mM) of So324 in rat liver S9 fractions at 37
`of starting material corresponded with the appearance of the amino acyl
`metabolite (AAM). Points are average (±S.D.) of the determinations.
`
`FE
`SE
`
`FE
`SE
`
`FE
`SE
`
`80
`
`60
`
`40
`
`20
`
`0
`
`% Recovery
`
`(A)
`
`H p-OMe p-Me
`
`p-Cl
`
`p-Br
`
`p-I
`
`100
`80
`60
`40
`20
`0
`
`% Recovery
`
`(B)
`
`H p-OMe p-Me p-Cl p-Br
`
`p-I
`
`100
`80
`60
`40
`20
`0
`
`% Recovery
`
`H p-OMe p-Me p-Cl
`p-Br
`p-I
`(C)
`Fig. 3. Percent recovery of pronucleotides incubated with (A) 2 × 106
`C. (B) 2×106 Caco-2
`◦
`Caco-2 cell homogenate, 1 h incubation time at 37
`cell suspension over 4 h. Between groups differences of averages were
`not statistically significant (P >0 .05). (C) Caco-2 cell monolayers, 4 h
`◦
`C in six-well plates. Bars represent the average
`incubation time at 37
`cumulative recovery of pronucleotides in the free incubation media and
`the intracellular fraction. The standard deviations (error bars) are means
`of two to four experiments performed in triplicate.
`
`mer could be recovered. In contrast, when the cell extract
`was incubated in the presence of the carboxylesterase in-
`hibitor PMSF, the amount of unchanged prodrug recov-
`ered after 4 h was approximately 85% for each diastereoiso-
`mer. PMSF concentrations in the range from 1.0 to 10 mM
`were used, and provided similar results. For example, when
`10 mM PMSF concentration was used, 85.1±0.8 and 87.5±
`0.6% of the FE and the SE diastereoisomer, respectively,
`were recovered. This is taken as direct and clear support for
`the notion that carboxyesterase-mediated cleavage is a key
`step in the activation of the nucleotide analogue. However,
`the fact that it was not possible to reach 100% inhibition
`suggests that other PMSF-insensitive hydrolases may be in-
`volved in the hydrolysis of the methyl ester bond.
`The stability of the pronucleotides was also evaluated
`following exposure to intact cells in suspension (4 days
`post-seeding), and as such probed the accessibility of the
`pronucleotides to metabolizing enzymes in the degradation
`process. Both diastereoisomers of So324 incubated with
`Caco-2 cell suspensions had a half-life of approximately
`10 h; the fraction of unchanged pronucleotide measured af-
`ter 16 h was 35.61 ± 4.7 and 37.30 ± 5.7% for the FE and
`SE diastereomers, respectively. The process was not stere-
`ospecific at any time-point (P >0 .05). The pronucleotide
`
`IPR2018-00126
`
`Page 5 of 7
`
`I-MAK 1021
`
`

`

`30
`
`D. Siccardi et al. / European Journal of Pharmaceutical Sciences 22 (2004) 25–31
`
`while no intact prodrug was found in the treatments with-
`out inhibitor. The metabolic profile of So324 was identi-
`cal in the incubation mixtures containing NADPH (1 mM)
`and in the controls without coenzyme, therefore excluding
`the involvement of CYP450 in the degradation of So324.
`These results suggest that esterase is the sole enzymatic sys-
`tem involved in the initial metabolic reaction of this pro-
`nucleotide.
`Using reverse phase chomatography, a number of peaks
`could be resolved after incubating the pronucleotides with
`plasma, Caco-2 or rat liver samples. Characterization of
`these compounds by HPLC retention time, UV spectra and
`LC/MS confirmed the previously reported notion that the
`main metabolite formed is the alaninyl d4TMP (Saboulard
`et al., 1999) (Fig. 4). Significant amounts of d4TMP were
`also found in all the cell-based experiments. Traces of other
`metabolites were observed in every in vitro preparation, al-
`though sufficient material for unambiguous structural deter-
`mination was not obtained.
`
`4. Discussion
`
`Previous work carried out by Saboulard et al. strongly
`suggested that carboxyl ester cleavage is a necessary prereq-
`uisite for activation of the prodrug to d4TMP and eventual
`antiviral activity (Saboulard et al., 1999). The role of esterase
`enzymes in this activation step was confirmed by incubation
`of So324 with pig liver carboxylesterase and mouse serum
`in the presence of the serine protease inhibitor PMSF which
`is also known to be an inhibitor of carboxylesterase (Shao
`and Mitra, 1994). To obtain optimal delivery of the ddNMP
`inside the target cells (i.e. the HIV-infected lymphocytes),
`an efficient conversion rate by carboxylesterase may be
`considered to be essential. However, in vivo the prodrugs
`can reach the target cells only if they are resistant to hy-
`drolysis by extracellular and intracellular carboxylesterases
`during the absorption and distribution process, as par-
`tial conversion of the prodrugs to intermediate metabo-
`lites or active nucleotide would almost certainly result
`in a lower cell penetration and, hence, reduced antiviral
`response.
`We here describe the in vitro stability of the prodrugs
`in aqueous buffer, in the extracellular environment (i.e.
`plasma), and in different cellular preparations prepara-
`tions. Chemical stability within the phosphoramidate series
`was studied under experimental conditions of biological
`relevance, i.e. at pH 2, 4.6, and 7.4, and temperature of
`◦
`37
`C. All the compounds in the series expressed satisfying
`stability both at acidic and neutral media; decomposition
`by hydrolysis was less than 30% over a 20 h period and
`difference between diastereomers were recorded for the
`p-Cl, p-Br and p-I derivatives (P <0 .05). The reason
`for these stereoselective differences is not readily explain-
`able, though not surprising, in consideration of the fact
`that diastereoisomers are distinct molecules with different
`
`physical and chemical properties. Previous work with di-
`aryl phosphate derivatives of AZT (McGuigan et al., 1994)
`demonstrated elevated antiviral activity for derivatives con-
`taining electron-withdrawing substituents on the phenyl
`ring. This finding was associated with the possibility that
`electron-withdrawing substituents may increase the rate at
`which the aryloxy group leaves after enzymatic hydroly-
`sis of the carboxy ester. In our study, however, no clear
`correlation between electron-donating (p-OMe, p-Me) or
`electron-withdrawing (p-Cl, p-Br, p-I) p-aryl substituents
`and chemical stability was evident. Since the unsubstituted
`lead prodrug So324 was the most stable compound under
`every pH condition, it is possible that both electron-donating
`and electron-withdrawing effects can influence the degra-
`dation rate of these pronucleotides. These electronic effects
`could also be responsible for the higher stability of the
`para-halo compounds at pH 4.6 as compared to 2.0 or 7.0.
`Prodrugs were significantly less stable in the presence of
`◦
`C
`plasma or cell extracts, than in buffer at pH 7.4 and 37
`suggesting that enzymatic metabolism drives the activation
`to the parent drug in vivo. The pronucleotides were bet-
`ter substrates for cellular esterases than for the plasma es-
`terases; this is not surprising since a large number of differ-
`ent hydrolytic enzymes with different substrate specificities
`are present in those tissues (Satoh et al., 2002). The degrada-
`tion of the So324 pronucleotide in Caco-2 cell homogenate
`and in liver S9 fraction was blocked with PMSF, thus con-
`firming an esterase-mediated activation step in these media,
`while excluding the contribution of P450 enzymes to the
`degradation of the pronucleotide.
`The metabolic profiles of this pronucleotide series ap-
`peared to be strongly stereoselective in biological media. In
`plasma the FE diastereoisomer of p-OMe derivative, for ex-
`ample appeared to be significantly more stable than the SE;
`the p-Cl derivative also showed significant stereoselectivity,
`but not the other pronucleotides. In Caco-2 cells both extent
`of metabolism and stereoselectivity were highly influenced
`by the experimental conditions. In cell homogenates all the
`pronucleotides were rapidly degraded and stereoselectivity
`was more marked than when exposed to suspensions or
`monolayers. In rat liver S9 fractions the prototype prodrug
`So324 was rapidly and extensively metabolized, showing a
`stereospecific pattern similar to that of the Caco-2 cell ex-
`tracts.
`Degradation of the ester prodrugs of this nucleotide ana-
`logue series should be expected in the process of absorbtion
`and distribution. We suggest that the in vivo circulating lev-
`els of intact prodrug may largely depend on the stereospe-
`cific affinity for metabolic enzymes, which can modulate
`the activity or toxicity of a particular pronucleotide due to
`enhanced distribution of one preferential diastereoisomer.
`Our findings also suggest that different experimental con-
`ditions used within the same cell line may express differing
`extents of metabolism and stereoselectivity, which should
`be taken into account when developing in vitro assays for
`the evaluation of metabolism.
`
`IPR2018-00126
`
`Page 6 of 7
`
`I-MAK 1021
`
`

`

`D. Siccardi et al. / European Journal of Pharmaceutical Sciences 22 (2004) 25–31
`
`31
`
`Acknowledgements
`
`We thank the Biotechnology and Biological Sciences
`Research Council (BBSRC) for financial support, and ac-
`knowledge the Wellcome Trust for equipment grants for the
`LC–MS.
`
`References
`
`Annaert, P., Gosselin, G., Pompon, A., Benzaria, S., Valette, G., Imbach,
`J.L., Naesens, L., Hatse, S., de Clercq, E., Van den Mooter, G.,
`Kinget, R., Augustijns, P., 1998. Comparison of the disposition of ester
`prodrugs of the antiviral agent 9-(2-phosphonylmethoxyethyl)adenine
`[PMEA] in Caco-2 monolayers. Pharm. Res. 15, 239–245.
`Balzarini, J., Karlsson, A., Aquaro, S., Perno, C.-F., Cahard, D., Naesens,
`L., De Clercq, E., 1996. Mechanism of anti-HIV action of masked
`alaninyl d4T-MP derivatives. Proc. Natl. Acad. Sci. U.S.A. 93, 7295–
`7299.
`McGuigan, C., Cahard, D., Sheeka, H.M., DeClercq, E., Balzarini,
`J., 1996. Aryl phosphoramidate derivatives of d4T have improved
`anti-HIV efficacy in tissue culture and may act by the generation of
`a novel intracellular metabolite. J. Med. Chem. 39, 1748–1753.
`McGuigan, C., Davies, M., Pathirana, R., Mahmood, N., Hay, A.J., 1994.
`Synthesis and anti-HIV activity of some novel diaryl phosphate deriva-
`tives of AZT. Antiviral Res. 24, 69–77.
`
`Saboulard, D., Naesens, L., Cahard, D., Salgado, A., Pathirana, R., Ve-
`lazquez, S., McGuigan, C., De Clercq, E., Balzarini, J., 1999. Char-
`acterization of the activation pathway of phosphoramidate triester
`prodrugs of stavudine and zidovudine. Mol. Pharmacol. 56, 693–
`704.
`Satoh, T., Taylor, P., Bosron, W.F., Sanghani, S.P., Hosokawa, M., La Du,
`B.N., 2002. Current progress on esterases: from molecular structure
`to function. Drug Metab. Dispos. 30, 488–493.
`Shao, Z., Mitra, A.K., 1994. Bile salt-fatty acid mixed micelles as nasal
`absorption promoters. III. Effects on nasal transport and enzymatic
`degradation of acyclovir prodrugs. Pharm. Res. 11, 243–250.
`Siddiqui, A.Q., Ballatore, C., McGuigan, C., De Clercq, E., Balzarini,
`J., 1999. The presence of substituents on the aryl moiety of the
`aryl phosphoramidate derivative of d4T enhances anti-HIV efficacy in
`cell culture: a structure-activity relationship. J. Med. Chem. 42, 393–
`399.
`Van Gelder, J., Deferme, S., Annaert, P., Naesens, L., De Clercq, E.,
`Van den Mooter, G., Kinget, R., Augustijns, P., 2000a. Increased
`absorption of the antiviral ester prodrug tenofovir disoproxil in rat
`ileum by inhibiting its intestinal metabolism. Drug Metab. Dispos. 28,
`1394–1396.
`Van Gelder,