- Email: [email protected]
Ester prodrugs of zidovudine
Ester Prodrugs of Zidovudine TAKEOKAWAGUCHI',
KO20 ISHIKAWA,
TOSHINOBU SEKI, AND KAZUHlKO JUNI
Received April 7, 1989, from the Faculty of Pharmaceutical Sciences, Josai University, 1- 1 Keyakidai, Sakado, Saitama 35002, Japan. Accepted for publication September 18, 1989. ~~.
.
-
_.____
Abstract 0 Ten novel ester prodrugs of zidovudine (azidothymidine; AZT) were synthesized with aliphatic acids (acetio-stearic),and the enzymatic regeneration of AZT from the prodrugs was investigated both
in vitro and in vivo. The enzymatic hydrolysis rates of the AZT esters in the presence of mouse enzyme systems (plasma,liver, and intestine, and kidney) were highly dependent on the lengths of the acyl chains of the prodrugs. The caprate or caprylate of AZT showed the highest reactivity to three of the four enzyme systems;either the decrease or the increase in the acyl chain length resulted in the decrease of the reactivity to the enzymes. Zidovudine (AZT) and three of the prodrugs (acetate, caprate, and stearate)were administered to mice intraperitoneally,and the plasma concentrations of AZT and a corresponding prodrug were measured. The AZT concentrations in plasma following the acetate administration rapidly decreased with a half-life of 14.5 min. This tendency is similar to that shown in direct AZT administration. On the other hand, the concentrations following the caprate or stearate administration decreased slowly and were maintained for as long as 4 h after dosing. The prodrug concentrations in plasma after the prodrug administration were under the detection limit (0.01 pg/mL),except for acetate. The absence of the caprate and stearate in plasma may be attributed to the high hydrophobicity or favorable tissue distribution of the ester derivatives. _______.
Zidovudine (azidothymidine; AZT) has been known to be an inhibitor of the reverse transcriptase of the human immunodeficiency virus (HIV) isolated from patients with acquired immunodeficiency syndrome (AIDS).' Though AZT has clinical activity in patients with AIDS o r AIDS-related complex, the toxicity of AZT, especially on bone marrow, has been reported to be significant, necessitating dose reductions or discontinuation of the treatment.2.3 Since AZT works as a metabolic antagonist against reverse transcriptase and since the antiviral effect can be time dependent, an adequate AZT concentration in a body should be maintained to achieve the anticipated anti-AIDS effect and to avoid the undesirable side e f f e ~ twhich , ~ seems to be attributable to an excessive plasma concentration of the drug. The prodrug approach has been known to be one of the most potent methods to improve the retention of a parent drug in a body or to control the drug concentration in plasma.5 Thus far, many prodrugs of antimetabolic agents have been synthesized and some of them have been used clinically. In this study, several ester prodrugs of AZT were synthesized by introducing aliphatic acyls on the 5'-hydroxy group of the parent drug. The regeneration of AZT from the prodrugs by chemical and enzymatic hydrolysis was measured in vitro. Plasma concentrations of both AZT and the prodrugs were also measured after intraperitoneal administration of the drugs in mouse.
chemical shifts are given in 6 (ppm) with TMS as an internal standard. An HPLC system consisting of a pump (LC-GA, Shimadzu), a variable-wavelength detector (SPD-GA, Shimadzu), and a 20-kL fixed loop injector (model 7125, Rheodyne) were used. The pH values of the buffers were read with a pH meter (model PH51, YokokawaHokushin Electric) a t 40 "C. Chemicals-Zidovudine (azidothymidine; AZT) was synthesized from thymidine according to the procedure of Horwitz et a1.H The esterification of AZT was carried out in dry pyridine with 1.2 equiv. of a corresponding acid anhydride at room temperature. The reaction mixtures were evaporated under reduced pressure a t 50 "C to remove pyridine. The residues were chromatographed over silica gel columns. Elution with CHC1,:EtOH (99:1-95:5) afforded the ester prodrugs. The chemical structure was supported by the 'H NMR and mass spectra of the esters. Stock solutions of all the esters and AZT were prepared in ethanol to give a concentration of 4 x M, and all were stored at 4 "C. All other chemicals were reagent or HPLC grade and were obtained commercially. Buffers-Buffer solutions were prepared with boric acid:sodium hydroxide for the pH values of 9.0; disodium hydrogen phosphate: sodium dihydrogen phosphate for pH 7.0; acetic pho8phate:sodium dihydrogen phosphate for pH 7.0; acetic acid:sodium acetate for pH 4.0; and g1ycine:HCl for pH 2.0. The buffer concentration range was 0.02-0.03 M. All the buffer solutions were adjusted to a constant ionic strength of 0.02 with potassium chloride. Preparation of Enzyme System-Male ddY mice (25-28 g) were obtained from Saitama Laboratory Animals (Saitama, Japan) and were sacrificed to obtain blood, livers, intestines, and kidneys. The blood was centrifuged at 1000 x g for 15 min, and the resulting plasma was stored at -40 "C until use. The tissues were homogenized with pH 7.0 isotonic phosphate buffer (0.1 M) containing 0.19 M sucrose a t 0 "C to give a concentration of 4.0% (w/v). One-milliliter portions of the homogenates were transferred to small glass tubes and stored a t -80 "C until use. Decrease in the enzyme activity of the stored samples, evaluated as the hydrolysis rates of C3-AZT and C10-AZT, was not observed during the experimental period (-8 weeks). Partition Coefficient Measurement-Apparent partition coefficients of the ester prodrugs and AZT were determined in a ch1oroform:O.l M phosphate buffer (pH 7.0) system a t 25 "C. In Vitro Method-The chemical hydrolysis rates of the AZT esters were measured under four pH conditions (pH 2.0,4.0, 7.0, and 9.0) a t 40 "C. Reaction was initiated by adding 10 WLof stock solution (4 x M)to a 2-mL preheated buffer solution in a glass tube. A IO-FL portion of the reacting mixture was periodically injected into a reversed-phase HPLC column (Nucleosil RP-IS), and concentrations of AZT and a prodrug in each sample were determined. The reactivities to chemical hydrolysis were evaluated by pseudo-first-order rate 0
0
.. I
Experimental Section Instrumentation-Melting points were determined on a Yanagimoto M P - S micro melting point apparatus and are uncorrected. Mass spectra were obtained with a JEOL JMS DX-300 mass spectrometer. Proton magnetic resonance spectra were obtained by using a JEOL JNM GX-270 FT NMR spectrometer at 270 MHz, and 0022-3549/90/0600-0531$0 1.oO/O 0 1990, American Pharmaceutical Association
n = O - 1 6
Journal of Pharmaceutical Sciences / 531 Vol. 79, No. 6, June 7990
Table I-Physlcochemlcal Properties of Zldovudine (AZT) Esters Melting Point, "C (Recryst. solvent)
Compound
124-1 25 (chloroform) Amorphous solid Amorphous solid Sticky liquid Oil Oil Oil 34-36 (mhexane) 3 9 4 2 (nhexane) 54-57 (mhexane) 4 1 4 2 (mhexane)
m Acetyl (CBAZT) Propionyl (C3-AZT) B~tyryl(C4-AZT) Caproyl (C6-AiT) Caprylyl (C8-AZT) Capryl (C10-AZT) Lauroyl (Cl2-AZT) Myristoyl (Cl4-AZT) Palmitoyl (C16-AZT) Stearoyl (Cl8-AZT) a Ch1oroform:phosphate
buffer (pH 7.0).
Chemical Hydrolysis Rate,
log PC8
(kbs) h-'
pH 2.0
pH 4.0
pH 7.0
pH 9.0
-0.31
<0.0005b
<0.0005
<0.0005
<0.0005
1.61 2.14 2.66 3.39 4.04 4.87 >5 >5 >5 >5
0.0861 0.0567
<0.0005
<0.0005
<0.0005 <0.0005 <0.0005
0.0366
<0.0005 <0.0005 <0.0005
0.0424 0.0475 0.0533 0.0332 0.0386 0.0421 0.0365
0.0007 0.0009 0.0009 <0.0005 <0.0005 <0.0005 <0.0005
<0.0005 <0.0005 <0.0005 <0.0005 <0.0005
0.138 0.0857 0.061 1 0.0544 0.0432 0.0518 0.0509 0.0495 0.0521 0.0484
No degradation was observed at 40 "C for 100 h.
constants obtained from the slopes of semilogarithmic plots of the ester concentration against time. The enzymatic hydrolysis rates were determined in the presence of one of the mouse enzyme systems diluted with a n isotonic phosphate buffer (pH 7.0) containing 0.19 M sucrose. The experiments were performed at 37 "C. The hydrolysis was initiated by adding the stock solution to the test solution to give an initial concentration of4 x M, the lowest concentration suitable for quantitative analysis of the compounds, and 8 x lO-'M. The changes in concentration of the ester and AZT were followed by HPLC analysis of samples taken periodically from the reaction mixture. The reactivities to enzymatic hydrolysis were evaluated as pseudo-first-order rate constants for 1.0%(w/v)plasma or homogenates. The constants were obtained from the slopes of semilogarithmic plots of 4 x lo-' M minus regenerated AZT concentration, which was consistent with the ester concentration in the same sample, against time. The enzymatic reaction was not saturated at the higher substrate concentration (8 x lo-' M). In Vivo Methods-Each group consisted of five male ddY mice (25-27 g, Saitama Laboratory Animals). One of the prodrugs (acetate, caprate, or stearate) or AZT was dissolved or suspended in saline containing 0.1% Tween 80 to give a concentration of 3.0 x M, and a 0.2-mL portion was administered intraperitoneally. Blood was collected into a heparinized glass tube by decapitation at specified times after administration. The blood was centrifuged at 1000 x g for 15 min and the resulting plasma was stored at -50 "C until HPLC analysis. A 300-& sample of the plasma was mixed with 300 pL of acetonitrile for deproteinization and the mixture was centrifuged at 1000 x g for 10 min. The resulting supernatant fluid was analyzed by HPLC.
Results and Discussion Table I shows the physicochemical properties of AZT prodrugs. The partition coefficients of the esters increase as the acyl chain is lengthened; the log coefficient of C14-AZT was 4.87, and the values for the longer esters were too high to be determined. The chemical hydrolysis rates of the esters were measured under four pH conditions (pH 2.0,4.0,7.0, and 9.0) a t 40 "C. All the esters were hydrolyzed at the acidic (pH 2.0) and the basic (pH 9.0)conditions with the similar rates (kobs and 4.3-8.6 x h-' for pH 2.0 and 9.0, = 3.3-5.7 x respectively) except for C2-AZT. The relatively high hydrolysis rates for C2-AZT (8.6 and 13.8 x lo-* h-l for pH 2.0 and 9.0, respectively) may be attributed to hyperconjugation of a methyl group in the ester. All the esters were quite stable under the more moderate pH conditions (pH 4.0 and 7.0); 4 0 % of the ester was degraded during incubation at 40 "C for 100 h. The effect of varying buffer concentrations (0.02-0.12 M) at constant pH values (pH 2.0-9.0) was evaluated; no buffer catalysis was noted on the hydrolysis a t any pH value. The enzymatic hydrolysis rates of the esters were measured in the presence of mouse enzyme systems (0.024.0%, wlv, preparation of plasma, liver, kidney, and intestine). The relationships between the acyl chain length and the enzymatic reactivity of the esters, standardized to 1.0% (w/v) enzyme preparation, are shown in Figure 1; C8-AZT or C10-AZT showed the highest reactivity. That is, either the
10.0
-
7 1.0 c
-E .A
v)
4-2
.1
c
+-' (D v)
c
.01
:c
L
c a, , @ (DL
,001
2
4
6
8
A c y l Chain Length
10
12
14
16
18
(Number o f Carbon Atoms)
Figure 1-Susceptibility of the AZT esters to mouse enzyme-catalyzed hydrolysis. Key: 1.0% (w/v) liver homogenate (-W-); 1.0% (w/v) plasma (-- 4 --): 1.O% (w/v) intestinal homogenate (-A+; 1.O% (w/v) kidney hornogenate (--V--). 532 I Journal of Pharmaceutical Sciences Vol. 79, No. 6, June 1990
z
,01
I
0
30
60
120
Time a f t e r A d m i n i s t r a t i o n
240
(min)
Figure 2-Plasma concentration of AZT after intraperitonealadministration of AZT (U),C2-AZT (0),C10-A2T (A), and C18-AZT (0);n = 5; +-SEM.
decrease or the increase in the acyl chain length resulted in the decrease in the susceptibility to the enzyme systems except for the intestinal homogenate. The intestinal preparation showed broader substrate specificity to the AZT esters (i.e., the susceptibility of the esters did not change in the range of C6-AZT to C12-AZT). The plasma concentrations of AZT &r intraperitoneal administration of four compounds (AZT, C2-AZT, C10-AZT, and C18AZT) were measured in mice. The plasma concentrationtime profdes are shown in Figure 2. The AZT administered intraperitoneally was rapidly eliminated from plasma, with a half-life of 11.7 min. The AZT concentrationsfollowing C2-AZT administration were lower than those followingthe parent drug administration, and C2-AZT was eliminated a t almost the same rate (tlm= 14.5 min) as that following direct AZT administration. The above result suggests that C2-AZT was rapidly eliminated from plasma and only a portion of C2-AZT was converted to the parent drug, AZT. The AZT levels following C10-AZT or C18-AZT administration were more persistent than those following AZT or C2-AZT administration. The relatively high AZT levels during the initial experimental period (up to 60 min) and the faster decrease in the AZT concentration following C10-AZT administration, when compared with those for Cl&AZT, could be attributed to the higher susceptibilityof C10-AZT to enzymatic hydrolysis. The administration of C18AZT resulted in a stable and persistent plasma concentration of AZT. This plasma concentration-time profile suggests long retention and sustained enzymatic hydrolysis of C18AZT in the body. The plasma concentrations of prodrugs aRer administration of the correspondingprodrug were also studied.Though C2-AZT was detected at 15and 30 min after the administration (0.15 k 0.02 and 0.06 * 0.02 @mL, respectively), other prodrugs were
not detected even immediately after the administration. Since the measurements were made on plasma samples, adsorption into certain tissues such as erythrocytes or fatty tissues may be responsible for the absence of C10-AZT and C18-AZT in plasma. Though further studies are needed to clarify the distribution and the kinetics of the AZT esters, the above results suggest that adequate delivery and sustained plasma levels of AZT can be achieved by selecting an appropriate ester derivative of AZT.
References and Notes 1. Mitauya, H.; Weinhold, K. J.; Furman, P. A.; St. Clair, M. H.; Ba D. W.; Broder, Lehrman, S. N.; Callo, R. C: Bolo esi, 0.; S. Proc. Natl. Acad. Sci. U . d A . l%5,82,7098iOO. 2. Mir, N.; Costello, C. Lancet 1988, zi, 1195-1196. 3. Dournon, E.;Rozenbaum, W.; Michon, C.; Perronne, C.; De h c h i s , P.; Bouvet, E.; Levacher, M.; Matheron, S.; Gharakhanian, S.; Girard, P. M.; Salmon, D.; Leport, C.; Dazza, M. C.; Regnier, B. Lancet 1968, ii, 1297-1302. 4. Klecker, R. W.; Collins, J. M.; Yarchoan, R.; Thomas, R.; Jenkins, J. F.; Broder, S.; Myers, C. E. Clin Pharmacol. Ther. 1967,41, 407-4 12. 5. Stella, V. In Pro-dru s as Novel Drug Delivery Systems; American Chemical Society: dshington, DC, 1975, pp 1-115. 6. Horwitz, J. P.; Urbanski, J. A.; Chua, J. J . Org. Chem. 1962,27, 3301-3302. 7. Horwitz, J. P.; Chua, J.; Urbanski, J. A.; Noel, M. J . Org. Chern. 1963.28.942-944. 6. HoGitz,' J. P.; Chua, J.; Noel, M. J . Org. Chem. 1964, 29, 2076-2078.
Acknowledgments The authors wish to express their sincere thanks to Dr. Mineo Saneyoshi, Faculty of Pharmaceutical Sciences, Hokkaido University, for his advice and encouragement.
Journal of Pharmaceutical Sciences I 533 Vol. 79, No. 6, June 1990
KO20 ISHIKAWA,
TOSHINOBU SEKI, AND KAZUHlKO JUNI
Received April 7, 1989, from the Faculty of Pharmaceutical Sciences, Josai University, 1- 1 Keyakidai, Sakado, Saitama 35002, Japan. Accepted for publication September 18, 1989. ~~.
.
-
_.____
Abstract 0 Ten novel ester prodrugs of zidovudine (azidothymidine; AZT) were synthesized with aliphatic acids (acetio-stearic),and the enzymatic regeneration of AZT from the prodrugs was investigated both
in vitro and in vivo. The enzymatic hydrolysis rates of the AZT esters in the presence of mouse enzyme systems (plasma,liver, and intestine, and kidney) were highly dependent on the lengths of the acyl chains of the prodrugs. The caprate or caprylate of AZT showed the highest reactivity to three of the four enzyme systems;either the decrease or the increase in the acyl chain length resulted in the decrease of the reactivity to the enzymes. Zidovudine (AZT) and three of the prodrugs (acetate, caprate, and stearate)were administered to mice intraperitoneally,and the plasma concentrations of AZT and a corresponding prodrug were measured. The AZT concentrations in plasma following the acetate administration rapidly decreased with a half-life of 14.5 min. This tendency is similar to that shown in direct AZT administration. On the other hand, the concentrations following the caprate or stearate administration decreased slowly and were maintained for as long as 4 h after dosing. The prodrug concentrations in plasma after the prodrug administration were under the detection limit (0.01 pg/mL),except for acetate. The absence of the caprate and stearate in plasma may be attributed to the high hydrophobicity or favorable tissue distribution of the ester derivatives. _______.
Zidovudine (azidothymidine; AZT) has been known to be an inhibitor of the reverse transcriptase of the human immunodeficiency virus (HIV) isolated from patients with acquired immunodeficiency syndrome (AIDS).' Though AZT has clinical activity in patients with AIDS o r AIDS-related complex, the toxicity of AZT, especially on bone marrow, has been reported to be significant, necessitating dose reductions or discontinuation of the treatment.2.3 Since AZT works as a metabolic antagonist against reverse transcriptase and since the antiviral effect can be time dependent, an adequate AZT concentration in a body should be maintained to achieve the anticipated anti-AIDS effect and to avoid the undesirable side e f f e ~ twhich , ~ seems to be attributable to an excessive plasma concentration of the drug. The prodrug approach has been known to be one of the most potent methods to improve the retention of a parent drug in a body or to control the drug concentration in plasma.5 Thus far, many prodrugs of antimetabolic agents have been synthesized and some of them have been used clinically. In this study, several ester prodrugs of AZT were synthesized by introducing aliphatic acyls on the 5'-hydroxy group of the parent drug. The regeneration of AZT from the prodrugs by chemical and enzymatic hydrolysis was measured in vitro. Plasma concentrations of both AZT and the prodrugs were also measured after intraperitoneal administration of the drugs in mouse.
chemical shifts are given in 6 (ppm) with TMS as an internal standard. An HPLC system consisting of a pump (LC-GA, Shimadzu), a variable-wavelength detector (SPD-GA, Shimadzu), and a 20-kL fixed loop injector (model 7125, Rheodyne) were used. The pH values of the buffers were read with a pH meter (model PH51, YokokawaHokushin Electric) a t 40 "C. Chemicals-Zidovudine (azidothymidine; AZT) was synthesized from thymidine according to the procedure of Horwitz et a1.H The esterification of AZT was carried out in dry pyridine with 1.2 equiv. of a corresponding acid anhydride at room temperature. The reaction mixtures were evaporated under reduced pressure a t 50 "C to remove pyridine. The residues were chromatographed over silica gel columns. Elution with CHC1,:EtOH (99:1-95:5) afforded the ester prodrugs. The chemical structure was supported by the 'H NMR and mass spectra of the esters. Stock solutions of all the esters and AZT were prepared in ethanol to give a concentration of 4 x M, and all were stored at 4 "C. All other chemicals were reagent or HPLC grade and were obtained commercially. Buffers-Buffer solutions were prepared with boric acid:sodium hydroxide for the pH values of 9.0; disodium hydrogen phosphate: sodium dihydrogen phosphate for pH 7.0; acetic pho8phate:sodium dihydrogen phosphate for pH 7.0; acetic acid:sodium acetate for pH 4.0; and g1ycine:HCl for pH 2.0. The buffer concentration range was 0.02-0.03 M. All the buffer solutions were adjusted to a constant ionic strength of 0.02 with potassium chloride. Preparation of Enzyme System-Male ddY mice (25-28 g) were obtained from Saitama Laboratory Animals (Saitama, Japan) and were sacrificed to obtain blood, livers, intestines, and kidneys. The blood was centrifuged at 1000 x g for 15 min, and the resulting plasma was stored at -40 "C until use. The tissues were homogenized with pH 7.0 isotonic phosphate buffer (0.1 M) containing 0.19 M sucrose a t 0 "C to give a concentration of 4.0% (w/v). One-milliliter portions of the homogenates were transferred to small glass tubes and stored a t -80 "C until use. Decrease in the enzyme activity of the stored samples, evaluated as the hydrolysis rates of C3-AZT and C10-AZT, was not observed during the experimental period (-8 weeks). Partition Coefficient Measurement-Apparent partition coefficients of the ester prodrugs and AZT were determined in a ch1oroform:O.l M phosphate buffer (pH 7.0) system a t 25 "C. In Vitro Method-The chemical hydrolysis rates of the AZT esters were measured under four pH conditions (pH 2.0,4.0, 7.0, and 9.0) a t 40 "C. Reaction was initiated by adding 10 WLof stock solution (4 x M)to a 2-mL preheated buffer solution in a glass tube. A IO-FL portion of the reacting mixture was periodically injected into a reversed-phase HPLC column (Nucleosil RP-IS), and concentrations of AZT and a prodrug in each sample were determined. The reactivities to chemical hydrolysis were evaluated by pseudo-first-order rate 0
0
.. I
Experimental Section Instrumentation-Melting points were determined on a Yanagimoto M P - S micro melting point apparatus and are uncorrected. Mass spectra were obtained with a JEOL JMS DX-300 mass spectrometer. Proton magnetic resonance spectra were obtained by using a JEOL JNM GX-270 FT NMR spectrometer at 270 MHz, and 0022-3549/90/0600-0531$0 1.oO/O 0 1990, American Pharmaceutical Association
n = O - 1 6
Journal of Pharmaceutical Sciences / 531 Vol. 79, No. 6, June 7990
Table I-Physlcochemlcal Properties of Zldovudine (AZT) Esters Melting Point, "C (Recryst. solvent)
Compound
124-1 25 (chloroform) Amorphous solid Amorphous solid Sticky liquid Oil Oil Oil 34-36 (mhexane) 3 9 4 2 (nhexane) 54-57 (mhexane) 4 1 4 2 (mhexane)
m Acetyl (CBAZT) Propionyl (C3-AZT) B~tyryl(C4-AZT) Caproyl (C6-AiT) Caprylyl (C8-AZT) Capryl (C10-AZT) Lauroyl (Cl2-AZT) Myristoyl (Cl4-AZT) Palmitoyl (C16-AZT) Stearoyl (Cl8-AZT) a Ch1oroform:phosphate
buffer (pH 7.0).
Chemical Hydrolysis Rate,
log PC8
(kbs) h-'
pH 2.0
pH 4.0
pH 7.0
pH 9.0
-0.31
<0.0005b
<0.0005
<0.0005
<0.0005
1.61 2.14 2.66 3.39 4.04 4.87 >5 >5 >5 >5
0.0861 0.0567
<0.0005
<0.0005
<0.0005 <0.0005 <0.0005
0.0366
<0.0005 <0.0005 <0.0005
0.0424 0.0475 0.0533 0.0332 0.0386 0.0421 0.0365
0.0007 0.0009 0.0009 <0.0005 <0.0005 <0.0005 <0.0005
<0.0005 <0.0005 <0.0005 <0.0005 <0.0005
0.138 0.0857 0.061 1 0.0544 0.0432 0.0518 0.0509 0.0495 0.0521 0.0484
No degradation was observed at 40 "C for 100 h.
constants obtained from the slopes of semilogarithmic plots of the ester concentration against time. The enzymatic hydrolysis rates were determined in the presence of one of the mouse enzyme systems diluted with a n isotonic phosphate buffer (pH 7.0) containing 0.19 M sucrose. The experiments were performed at 37 "C. The hydrolysis was initiated by adding the stock solution to the test solution to give an initial concentration of4 x M, the lowest concentration suitable for quantitative analysis of the compounds, and 8 x lO-'M. The changes in concentration of the ester and AZT were followed by HPLC analysis of samples taken periodically from the reaction mixture. The reactivities to enzymatic hydrolysis were evaluated as pseudo-first-order rate constants for 1.0%(w/v)plasma or homogenates. The constants were obtained from the slopes of semilogarithmic plots of 4 x lo-' M minus regenerated AZT concentration, which was consistent with the ester concentration in the same sample, against time. The enzymatic reaction was not saturated at the higher substrate concentration (8 x lo-' M). In Vivo Methods-Each group consisted of five male ddY mice (25-27 g, Saitama Laboratory Animals). One of the prodrugs (acetate, caprate, or stearate) or AZT was dissolved or suspended in saline containing 0.1% Tween 80 to give a concentration of 3.0 x M, and a 0.2-mL portion was administered intraperitoneally. Blood was collected into a heparinized glass tube by decapitation at specified times after administration. The blood was centrifuged at 1000 x g for 15 min and the resulting plasma was stored at -50 "C until HPLC analysis. A 300-& sample of the plasma was mixed with 300 pL of acetonitrile for deproteinization and the mixture was centrifuged at 1000 x g for 10 min. The resulting supernatant fluid was analyzed by HPLC.
Results and Discussion Table I shows the physicochemical properties of AZT prodrugs. The partition coefficients of the esters increase as the acyl chain is lengthened; the log coefficient of C14-AZT was 4.87, and the values for the longer esters were too high to be determined. The chemical hydrolysis rates of the esters were measured under four pH conditions (pH 2.0,4.0,7.0, and 9.0) a t 40 "C. All the esters were hydrolyzed at the acidic (pH 2.0) and the basic (pH 9.0)conditions with the similar rates (kobs and 4.3-8.6 x h-' for pH 2.0 and 9.0, = 3.3-5.7 x respectively) except for C2-AZT. The relatively high hydrolysis rates for C2-AZT (8.6 and 13.8 x lo-* h-l for pH 2.0 and 9.0, respectively) may be attributed to hyperconjugation of a methyl group in the ester. All the esters were quite stable under the more moderate pH conditions (pH 4.0 and 7.0); 4 0 % of the ester was degraded during incubation at 40 "C for 100 h. The effect of varying buffer concentrations (0.02-0.12 M) at constant pH values (pH 2.0-9.0) was evaluated; no buffer catalysis was noted on the hydrolysis a t any pH value. The enzymatic hydrolysis rates of the esters were measured in the presence of mouse enzyme systems (0.024.0%, wlv, preparation of plasma, liver, kidney, and intestine). The relationships between the acyl chain length and the enzymatic reactivity of the esters, standardized to 1.0% (w/v) enzyme preparation, are shown in Figure 1; C8-AZT or C10-AZT showed the highest reactivity. That is, either the
10.0
-
7 1.0 c
-E .A
v)
4-2
.1
c
+-' (D v)
c
.01
:c
L
c a, , @ (DL
,001
2
4
6
8
A c y l Chain Length
10
12
14
16
18
(Number o f Carbon Atoms)
Figure 1-Susceptibility of the AZT esters to mouse enzyme-catalyzed hydrolysis. Key: 1.0% (w/v) liver homogenate (-W-); 1.0% (w/v) plasma (-- 4 --): 1.O% (w/v) intestinal homogenate (-A+; 1.O% (w/v) kidney hornogenate (--V--). 532 I Journal of Pharmaceutical Sciences Vol. 79, No. 6, June 1990
z
,01
I
0
30
60
120
Time a f t e r A d m i n i s t r a t i o n
240
(min)
Figure 2-Plasma concentration of AZT after intraperitonealadministration of AZT (U),C2-AZT (0),C10-A2T (A), and C18-AZT (0);n = 5; +-SEM.
decrease or the increase in the acyl chain length resulted in the decrease in the susceptibility to the enzyme systems except for the intestinal homogenate. The intestinal preparation showed broader substrate specificity to the AZT esters (i.e., the susceptibility of the esters did not change in the range of C6-AZT to C12-AZT). The plasma concentrations of AZT &r intraperitoneal administration of four compounds (AZT, C2-AZT, C10-AZT, and C18AZT) were measured in mice. The plasma concentrationtime profdes are shown in Figure 2. The AZT administered intraperitoneally was rapidly eliminated from plasma, with a half-life of 11.7 min. The AZT concentrationsfollowing C2-AZT administration were lower than those followingthe parent drug administration, and C2-AZT was eliminated a t almost the same rate (tlm= 14.5 min) as that following direct AZT administration. The above result suggests that C2-AZT was rapidly eliminated from plasma and only a portion of C2-AZT was converted to the parent drug, AZT. The AZT levels following C10-AZT or C18-AZT administration were more persistent than those following AZT or C2-AZT administration. The relatively high AZT levels during the initial experimental period (up to 60 min) and the faster decrease in the AZT concentration following C10-AZT administration, when compared with those for Cl&AZT, could be attributed to the higher susceptibilityof C10-AZT to enzymatic hydrolysis. The administration of C18AZT resulted in a stable and persistent plasma concentration of AZT. This plasma concentration-time profile suggests long retention and sustained enzymatic hydrolysis of C18AZT in the body. The plasma concentrations of prodrugs aRer administration of the correspondingprodrug were also studied.Though C2-AZT was detected at 15and 30 min after the administration (0.15 k 0.02 and 0.06 * 0.02 @mL, respectively), other prodrugs were
not detected even immediately after the administration. Since the measurements were made on plasma samples, adsorption into certain tissues such as erythrocytes or fatty tissues may be responsible for the absence of C10-AZT and C18-AZT in plasma. Though further studies are needed to clarify the distribution and the kinetics of the AZT esters, the above results suggest that adequate delivery and sustained plasma levels of AZT can be achieved by selecting an appropriate ester derivative of AZT.
References and Notes 1. Mitauya, H.; Weinhold, K. J.; Furman, P. A.; St. Clair, M. H.; Ba D. W.; Broder, Lehrman, S. N.; Callo, R. C: Bolo esi, 0.; S. Proc. Natl. Acad. Sci. U . d A . l%5,82,7098iOO. 2. Mir, N.; Costello, C. Lancet 1988, zi, 1195-1196. 3. Dournon, E.;Rozenbaum, W.; Michon, C.; Perronne, C.; De h c h i s , P.; Bouvet, E.; Levacher, M.; Matheron, S.; Gharakhanian, S.; Girard, P. M.; Salmon, D.; Leport, C.; Dazza, M. C.; Regnier, B. Lancet 1968, ii, 1297-1302. 4. Klecker, R. W.; Collins, J. M.; Yarchoan, R.; Thomas, R.; Jenkins, J. F.; Broder, S.; Myers, C. E. Clin Pharmacol. Ther. 1967,41, 407-4 12. 5. Stella, V. In Pro-dru s as Novel Drug Delivery Systems; American Chemical Society: dshington, DC, 1975, pp 1-115. 6. Horwitz, J. P.; Urbanski, J. A.; Chua, J. J . Org. Chem. 1962,27, 3301-3302. 7. Horwitz, J. P.; Chua, J.; Urbanski, J. A.; Noel, M. J . Org. Chern. 1963.28.942-944. 6. HoGitz,' J. P.; Chua, J.; Noel, M. J . Org. Chem. 1964, 29, 2076-2078.
Acknowledgments The authors wish to express their sincere thanks to Dr. Mineo Saneyoshi, Faculty of Pharmaceutical Sciences, Hokkaido University, for his advice and encouragement.
Journal of Pharmaceutical Sciences I 533 Vol. 79, No. 6, June 1990