Archives of Pharmacal Research

, Volume 37, Issue 11, pp 1416–1425

In vitro evaluation of 9-(2-phosphonylmethoxyethyl)adenine ester analogues, a series of anti-HBV structures with improved plasma stability and liver release

  • Sha Liao
  • Shi-Yong Fan
  • Qin Liu
  • Chang-Kun Li
  • Jia Chen
  • Jing-Lai Li
  • Zhi-Wei Zhang
  • Zhen-Qing Zhang
  • Bo-Hua Zhong
  • Jian-Wei Xie
Research Article

DOI: 10.1007/s12272-013-0300-6

Cite this article as:
Liao, S., Fan, SY., Liu, Q. et al. Arch. Pharm. Res. (2014) 37: 1416. doi:10.1007/s12272-013-0300-6

Abstract

Chronic hepatitis B virus (HBV) infection may lead to liver cirrhosis and hepatocellular carcinoma, but few drugs are available for its treatment. Acyclic nucleoside phosphonates (ANPs) have remarkable antivirus activities but are not easily absorbed from the gastrointestinal tract and accumulate in the kidneys, resulting in nephrotoxicity. Therefore, there is a need to find effective liver site-specific prodrugs. The dipivaloyloxymethyl ester of 9-(2-phosphonylmethoxyethyl)adenine (PMEA)—adefovir dipivoxil (ADV)—is a first-line therapy drug for chronic hepatitis B with a low therapeutic index because of renal toxicity and low hepatic uptake. In this study, a series of PMEA derivatives were synthesized to enhance plasma stability and liver release. The metabolic stability of ADV (Chemical I) and its two analogues (Chemicals II and III) was evaluated in rat plasma and liver homogenate in vitro. An ion-pair reverse-phase HPLC–UV method and a hybrid ion trap and high-resolution time-of-flight mass spectrometry (LC-IT-TOF-MS) were used to evaluate the degradation rate of the analogues and to identify their intermediate metabolites, respectively. Chemicals I and II were hydrolyzed by cleavage of the C–O bond to give monoesters. Sufficient enzymatic activation in the liver homogenate through a relatively simple metabolic pathway, in addition to a favorable stability profile in rat plasma, made Chemical II an optimal candidate. Next, six analogues based on the structure of Chemical II were synthesized and evaluated in plasma and liver homogenate. Compared to Chemical II, these compounds generated less active PMEA levels in rat liver homogenate. Therefore, chemical modification of Chemical II may lead to new promising PMEA derivatives with enhanced plasma stability and liver activation.

Keywords

Metabolic stability Adefovir dipivoxil Ester prodrug PMEA HPLC–UV ESI-TOF-MS 

Introduction

Chronic hepatitis B is a serious liver disease caused by an infection of parenchymal liver cells with the hepatitis B virus (HBV), which may lead to liver cirrhosis and hepatocellular carcinoma (Sherker and Marion 1991; Kane 1996; Guan and Lui 2011). More than 400 million people worldwide are infected with HBV, and this infection results in 500,000 to 1.2 million deaths each year (Lavanchy 2004).

Finding an effective treatment for HBV infection constitutes one of the current therapeutic challenges in virology. However, only a few drugs are currently available for the clinical treatment of hepatitis B. Administration of interferon α appears to be an effective approach, but only in 10–30 % of treated patients (Hoofnagle 1998). Another approved drug, lamivudine, induces drug resistance (Hagmeyer and Pan 1999; Ling et al. 1996). In the past decade, in addition to lamivudine, several acyclic nucleoside phosphonates (ANPs) were found to have remarkable antivirus activities. Promising candidate drugs included 9-(2-phosphonomethoxyethyl) adenine (PMEA), 9-2-(phosphonomethoxyethyl)-2,6-diaminopurine (PMEDAP), cidofovir, and tenofovir, all of which inhibit the replication of a broad spectrum of viruses, including HBV. However, one drawback of the ANPs is that they are not easily absorbed from the gastrointestinal tract because of a negatively charged phosphonic acid. To improve their oral bioavailability, a range of PMEA prodrugs has been synthesized. Among these, the dipivaloyloxymethyl ester of PMEA (adefovir dipivoxil, ADV, previously called bis(POM)-PMEA) showed acceptable oral bioavailability and was approved by the United States Food and Drug Administration (FDA) for anti-HBV chemotherapy in 2002. Unfortunately, as a first-line therapy for chronic hepatitis B, the orally bioavailable ADV is completely hydrolyzed to free PMEA during transport from the gastrointestinal tract to the circulation, which results in accumulation primarily in the kidneys, whereas only a limited amount is taken up by the liver (Naesens et al. 1992). The high level of uptake of PMEA by the kidneys results in nephrotoxicity, which is thought to develop from active transport of PMEA into the renal proximal tubules via renal organic anion transporters. The low therapeutic index of ADV is considered attributable to its extensive renal clearance and low hepatic uptake. Therefore, the discovery of a novel ester prodrug that would afford selective delivery of PMEA to the target organ, liver tissue, may lower the level of renal uptake and also improve its therapeutic efficacy. An ideal anti-HBV liver-targeting phosphonate prodrug would have the properties of releasing the parent drug in the liver, allowing permeation across the intestine, and exhibiting stability in plasma/blood and most nonhepatic tissues (Erion et al. 2004).

To overcome the drawbacks associated with ADV therapy, research has been directed toward achieving better liver site-specific prodrugs. Starrett et al. have designed and synthesized a much wider range of prodrugs employed for the preparation of phosphonic acid prodrugs (Starrett et al. 1994). This group suggested that prodrug modifications of bis(alkyl) esters, such as diethyl ester and bis(isopropyl) ester, were well absorbed (>40 %), although their diester prototypes and/or monoesters were identified in the circulatory system after oral administration to rats. Further structure–activity relationship (SAR) studies demonstrated that the pharmacophore of purine in ANPs is characterized by the presence of an amino group at the pyrimidine part of the purine system and that the N6-substitution in adenine derivatives preserved or considerably increased the antiviral and/or cytostatic activity of the PMEA series, while other alterations of amino groups generally resulted in complete loss of activity (Sekiya et al. 2002; Balzarini et al. 1993). Investigations into the biochemical mechanism revealed that the N6-substituted analogues were metabolized to the corresponding purine counterparts by liver-located aminohydrolases, and therefore, could be considered to be liver-specific prodrugs (Schinkmanova et al. 2006).

Therefore, in the present study, we synthesized and compared the effects of bis(trifluoroethyl) and bis(isopropyl) as promoieties (Fig. 1; Series 1 in Table 1). Furthermore, a homologous series of 6-N-alkyl carbonates of PMEA was prepared and the phosphonates of the relevant analogues were protected with the bis(2,2,2-trifluoroethyl) group (Fig. 1; Series 2 in Table 1), as were the nucleoside phosphate prodrugs in Series 1. The in vitro stability and biotransformation of the two series of ester derivatives of PMEA were investigated in rat plasma and liver homogenate.
Fig. 1

Chemical structures of PMEA ester analogues used in the work. Chemical I (ADV, bis(POM)-PMEA): R1 = R2 = –CH2OCOC(CH3)3; Chemical II: R1 = R2 = –CH2CF3; Chemical III: R1 = R2 = –CH(CH3)3; PMEA: R1 = R2 = –H; monoester analogue of Chemical I: R1 = H, R2 = –CH2OCOC(CH3)3; monoester analogue of Chemical II: R1 = H, R2 = –CH2CF3. R3 as shown in Table 1

Table 1

Chemical structures of PMEA ester analogues

No.

Compound

R1

R2

R3

Series 1

 1

I (bis(POM)-PMEA)

–CH2OCOC(CH3)3

–CH2OCOC(CH3)3

–H

 2

II

–CH2CF3

–CH2CF3

–H

 3

III

–CH(CH3)2

–CH(CH3)2

–H

Series 2

 4

II-1

–CH2CF3

–CH2CF3

–COOCH3

 5

II-2

–CH2CF3

–CH2CF3

–COOCH2CH3

 6

II-3

–CH2CF3

–CH2CF3

–COO(CH2)2CH3

 7

II-4

–CH2CF3

–CH2CF3

–COO(C6H5)

 8

II-5

–CH2CF3

–CH2CF3

–COOCH(CH3)2

 9

II-6

–CH2CF3

–CH2CF3

–COO(CH2)3CH3

 10

PMEA

–H

–H

–H

Materials and methods

Reagents and chemicals

PMEA and its analogues with HPLC purities greater than 99 % were provided by Prof. Bo-hua Zhong of Beijing Institute of Pharmacology and Toxicology and were stored at −70 °C. The internal standard (IS) lamivudine was extracted from tablets purchased from GlaxoSmithKline (China); purity was determined by HPLC (>99 %). Acetonitrile (ACN, HPLC grade) was obtained from Fisher (Fair Lawn, NJ). Analytical-grade ion-pair reagent tetramethylammonium bisulfate was purchased from Sigma (St. Louis, MO). Distilled water was purified using Milli-Q system (Millipore, Molsheim, France). All other reagents, including trichloroacetic acid, ammonium acetate, and glacial acetic acid, were analytical-grade products obtained from Sigma.

Instrumentation

pH determination was performed using Orion SA 520 pH-meter equipped with a combination pH electrode (Orion Research Inc., Boston, MA, USA). The samples from enzymatic and chemical stability studies of all model derivatives were analyzed by an Agilent 1,100 Series HPLC system (Agilent Technologies, Inc., Torrance, CA, USA) equipped with two Agilent 1,100 pumps, an Agilent autosampler, and an Agilent DAD detector. The system was operated using Agilent Chemstation software. A liquid chromatography-electrospray ionization source in combination with hybrid ion trap and high-resolution time-of-flight mass spectrometry (LC-ESI-IT-TOF-MS, Shimadzu, Kyoto, Japan) was used to identify the post-column fraction of the incubation metabolites.

Incubation of PMEA ester analogues with rat plasma

Because of the poor aqueous solubility of the newly synthesized PMEA ester analogues, an aqueous solution of 10 % polyethylene glycol (PEG) 400 was used as a solvent. PMEA ester analogues were incubated with rat plasma, phosphate buffer, and inactivated plasma controls at 10, 40, and 100 μM, respectively, and were placed in a water bath shaker (shaking incubator SI-600R) at 37 °C. At the predetermined time points (0, 1, 2, 3, 6, 8, 12, and 30 min, and 1, 2, 4, 6, 12, and 24 h), a 100 μL aliquot of the solution was collected and two volumes of 0.1 % (v/v) trichloroacetic acid in ice-cold acetonitrile (ACN) with lamivudine (10 μg/mL, IS) were added. After immediate mixing, the extracts were centrifuged at 14,000 rpm for 20 min at 4 °C and the supernatants (200 μL) were evaporated under vacuum before resuspending in 100 μL of buffer A (2 % (v/v) ACN, 2 mM tetrabutyl ammonium bisulfate, 50 mM ammonium acetate, pH 4.0). Next, 20 μL of the supernatant was taken for each HPLC analysis. Pseudo-first-order half-lives (t½) were calculated from the linear slope of the logarithmic plot of the concentration of the remaining compound at different time points.

In vitro activation of PMEA ester analogues by rat liver homogenates

The enzymatic hydrolysis of PMEA ester analogues was also studied in a rat liver homogenate (RLH) at 37 °C. Chemicals (10, 40, and 100 μM) were incubated with liver homogenates or their inactivated controls at appropriate intervals (0, 0.5, 1, 3, 6, 12, 24, 48, 72, 108, and 144 min). Samples (100 μL) were deproteinated and reconstituted as plasma samples as mentioned above.

Ion-pair reverse-phase HPLC–UV analysis of PMEA and its ester analogues in rat plasma and liver homogenate

For the determination of PMEA and its ester analogues (Series 1 Chemicals I–III and Series 2 Chemicals II-1-6), a 20-μL sample was injected onto a Hypersil ODS C18 column (150 mm × 4.6 mm, 5 μm, Thermo) fitted with a precolumn ODS C18 (5 × 4.6 mm, 40 μm, Agilent) and eluted with a gradient from buffer A to buffer B (which had the same composition but contained 50 % ACN, (v/v)) at a flow rate of 1.0 mL/min and room temperature (Table 2). Detection was performed using a UV detector (G1314A UV Detector) at a wavelength of 262 nm. The retention times of PMEA and IS (lamivudine) were approximately 3.8 and 6.2 min, respectively. Linearity was observed from 1.25 to 60 μg/mL, and calibration curves for PMEA and ester analogues in different matrices were generated. The limits of quantitation (LOQ) for both PMEA and ester derivatives were 1.25 μg/mL. The analytical method met an acceptable percent coefficient of variance (<5%) at the lowest concentration (typically 1.25 μg/mL) and of <1.5 % for higher concentrations. Data were processed using ChemStation software (Agilent technologies, USA).
Table 2

HPLC gradient elution condition

Compound

Time (min)

Buffer A (%)

Buffer B (%)

Flow (ml/min)

Total time (min)

I

0–5

100

0

1.0

25.0

10

0

100

10–15

0

100

15–25

100

0

II

0–3

100

0

1.0

26.0

18

0

100

18–26

100

0

III

0–3

100

0

1.0

23.0

8

20

80

15

20

80

15–23

100

0

II-1,4,5,6

0–3

100

0

1.0

23.0

10

0

100

10–15

0

100

15–23

100

0

II-2

0–3

100

0

1.0

30.0

25

0

100

25–30

100

0

II-3

0–3

100

0

1.0

35.0

30

0

100

30–35

100

0

LC-ESI-IT-TOF mass spectrometry analysis

HPLC–UV analysis of the extracts showed a new peak at 11.2 and 11.6 min (Fig. 4) in rat plasma and liver homogenates samples from incubations performed with both Chemicals I and II, respectively. After HPLC post-column collection of the observed intermediate metabolites, accurate molecular masses were determined using an LC-ESI-IT-TOF mass spectrometer. Because the intermediate metabolites (e.g., PMEA and ester derivatives) could be detected by HPLC–UV, indicating that the concentration level was high enough for MS analysis, no further sample preparation was conducted to enrich the target metabolites.

LC experiments were conducted using Shimadzu (Kyoto, Japan) HPLC system consisting of an LC-30AD binary pump, a DGU-20A5 degasser, a SIL-30AC autosampler, and a CTO-20AC column oven. Chromatographic separation was achieved on an Inertsil C8-3 column (50 mm × 2.1 mm, 5 μm) (GL Sciences) at room temperature. The mobile phase (delivered at 0.3 mL/min) comprised solvent A, ACN-H2O-formic acid solution (90:10:0.1 [v/v/v]), and solvent B, 0.1 % formic acid in H2O. A gradient elution with phase A changing from 20 to 70 % was performed in 0–30 min. MS analyses were conducted on a Shimadzu IT-TOF-MS (Shimadzu, Kyoto, Japan) equipped with an electrospray ionization source and operated in positive and negative mode simultaneously. Some important equipment analytical parameters were as follows: electrospray voltage, +4.5 kV for positive mode and −3.5 kV for negative mode; nebulizing gas (N2) flow, 1.5 L/min; curved desolvation line temperature, 200 °C; heat block temperature, 200 °C; dry gas (N2) flow, 10 L/min; and positive and negative scan range (m/z), 100–700. Shimadzu’s composition formula predictor software was used to calculate the accurate mass of the intermediate metabolites.

Data analysis

Apparent first-order kinetics and rate constants were determined using the initial rates of enzymatic hydrolysis. The apparent first-order degradation rate constants of PMEA derivatives or formation rate constants of PMEA at 37 °C were determined by plotting the logarithm of the concentration of PMEA as a function of time. The relationship between the rate constants, k, and slopes of the plots are explained by the following equation: k (min−1) = −slope (log C vs. time). The degradation half-lives were then calculated by using the following equation: t½ = 0.693/k.

Results

The stability of PMEA ester analogues (Series 1) in RLH and plasma

Chemicals I–III were incubated in three different buffers separately, and samples were collected and analyzed chromatographically for released PMEA at the predetermined times. In order to evaluate whether chemical hydrolysis of analogues by water contributes significantly to the enzymatic bioconversion in plasma or liver homogenate, the stability of all three chemicals in phosphate-buffered saline (PBS), inactivated plasma, or liver S9 was evaluated. The corresponding intermediate metabolites were detected as a single decomposition product for Chemicals I and II; however, the active structure, PMEA, could not be detected during the entire incubation period. In these control systems, Chemicals I and II were degraded in a pseudo-first-order fashion with stoichiometric production of their intermediate metabolite, respectively. For Chemical III, no degradation or metabolite was observed during the incubation period; for practical purposes, Chemical III can be regarded as chemically stable under the experimental conditions when the system was devoid of enzymes.

Furthermore, the degradation rates of the Chemical I and II prototypes in PBS and in the inactivated plasma or liver homogenates negative control were far lower than when in the normal plasma or liver S9, indicating the major role of enzymatic activation for the PMEA ester analogues.

To investigate the potential of Chemicals I–III as liver-target prodrugs, our strategy was to first test the liver activation in RLH by measurement of the concentration levels of PMEA following a 2-h exposure to the prodrug (10, 40, and 100 μM).

The decomposition of Chemical I and II in RLH at 37 °C is shown in Fig. 2. A similar trend was also observed in RLH, which shows that both Chemicals I and II could be hydrolyzed to produce the anti-HBV active structure PMEA. More importantly the newly designed and synthesized Chemical II generated similar PMEA levels (100 % activation) compared to those generated by the reference compound Chemicals I (bis-POM PMEA), indicating an excellent degree of prodrug activation (Table 3). Compound III showed no activation in RLH.
Fig. 2

Degradation profile of Chemical I (a bis(POM)-PMEA)) and Chemical II (b (bis(2,2,2-trifluoroethyl)ester of PMEA) in rat liver homogenate (37 ± 0.5 °C) as monitored by HPLC–UV at different incubation concentrations [100 μM (circle), 40 μM (diamond), and 10 μM (triangle)]. Time courses for disappearance/formation of intermediate metabolite (solid) and PMEA (hollow). Key Prodrug Chemicals I and II were not determined in rat liver homogenate owing to their rapid degradation

Table 3

Stability and activation of ester analogues in rat plasma and liver homogenate

Chemicals

Concentration (μM)

t½a

Yield (%)b

I

10

85.6

100

 

40

82.5

100

 

100

73.7

100

II

10

364.7

100

 

40

377.2

100

 

100

407.6

100

III

10

ND

ND

 

40

ND

ND

 

100

ND

ND

II-1

40

ND

ND

II-2

40

ND

ND

II-3

40

31.6

ND

II-4

40

52.5

12.4

II-5

40

207.5

6.57

II-6

40

128.3

11.8

ND no degradation detected

at½ represents half-lives (minutes) in rat plasma at 37 °C

bYields for the conversion of PMEA ester derivatives to PMEA

A drug designed to act on an HBV-infected liver should survive plasma transport following passage through the liver tissue in order to decrease PMEA accumulation in the kidney tissue. Stability of these three chemicals in plasma was further studied by incubation with rat plasma to observe the disappearance of the prototype and formation of the active metabolite by HPLC. The prototypes of Chemicals I and II were not observed in the beginning of the incubation because they were rapidly converted to their intermediate metabolites after hydrolysis. Therefore, the degradation profile of the intermediate metabolite of Chemicals I and II was used to evaluate the plasma stability of the compounds. The degradation of the intermediate metabolite of Chemicals I and II followed pseudo-first-order kinetics and the half-lives (t½ values) for these three chemicals in rat plasma are presented in Table 3.

Figure 3 shows the bioconversion profile of Chemical I and II in rat plasma at 37 °C. Chemical III was very stable and did not disappear significantly or produce PMEA (Table 3); Chemical II was less stable but did not produce quantifiable amounts of PMEA. The ester bond cleavage rates of Chemical I (half-life of 80 min) in rat plasma was about four-fold faster than those of Chemical II (half-life of 400 min) (Table 3). The stability of Chemicals I–III in rat plasma was also shown to be related to the type of ester group. Some interesting similarities were observed between Chemicals I and II in terms of the formation of intermediates. For example, whereas the prototype of the two ester analogues was not detected almost from the beginning of incubation, significant accumulation of intermediate I from Chemical I and intermediate II from Chemical II was observed (Figs. 2, 3).
Fig. 3

Degradation profile of Chemical I (a bis(POM)-PMEA) and Chemical II (b) in rat plasma (37 ± 0.5 °C), as monitored by HPLC–UV at different incubation concentrations [100 μM (circle), 40 μM (diamond), and 10 μM (triangle)]. Time courses for disappearance/formation of intermediate metabolite (solid) and PMEA (hollow). Key Prodrug Chemical I was not determined in rat plasma owing to its rapid degradation

Chemical III, whose stability in both plasma and liver homogenate was greater than that of Chemicals I or II, possessed the bis(isopropyl) chains, indicating that these more lipophilic groups are more suited for the preservation of prodrug integrity during the distribution phase; unfortunately, the chemicals were not activated in the liver.

Identification of the intermediate metabolites of PMEA ester analogues

As shown in Figs. 4 and 2 min after adding Chemical I in normal rat plasma, two new peaks appeared at 3.8 min and 11.2 min; no prototype Chemical I was observed. Six hours later, the peak of 3.8 min became larger, while the peak of 11.2 min diminished. No other peaks were observed.
Fig. 4

Chromatograms of Chemical I (bis(POM)-PMEA) and its intermediate metabolite in normal rat plasma, observed using a HPLC–UV detector at 262 nm. PMEA (3.8 min), IS (lamivudine, 6.7 min), and Chemical I (13.8 min) in 30 % PEG solution (a); blank rat plasma (b); rat plasma spiked with Chemical I (c); intermediate metabolite (11.2 min) appeared in rat plasma spiked with Chemical I after incubation for 6 h (d)

The intermediate metabolites were identified by comparison of the retention times with the standard and LC-ESI-IT-TOF-MS analysis after post column peak fraction collection. HPLC–UV analysis indicated that the degradation of Chemical I immediately led to the production of its intermediate metabolite (t = 11.2 min), which was further degraded into the parent drug PMEA (t = 3.8 min). Similarly, Chemical II was also immediately degraded into its intermediate metabolite (t = 11.6 min) and PMEA (t = 3.8 min) in the plasma and liver homogenate. The accurate mass spectra of intermediate metabolites for Chemical I (t = 11.2 min) and Chemical II (t = 11.6 min) were obtained using LC-ESI-IT-TOF-MS, and the results are shown in Figs. 5 and 6, respectively; the relevant results are listed in Table 4. From the [M+H]+ at m/z 388.1383 in the positive full-scan mass spectrum (Fig. 5a) and the [M−H] at m/z 386.1236 in the negative full-scan mass spectrum (Fig. 5b, both were 100 % relative signal intensity), the molecular formula was determined as C14H22N5O6P, indicating the cleavage of one ester group of Chemical I.
Fig. 5

ESI–TOF mass spectra of the intermediate metabolite (peak at 11.2 min) generated from the degradation reaction of Chemical I in normal rat plasma in positive (a) and negative (b) ion modes

Fig. 6

ESI–TOF mass spectra of the intermediate metabolite (peak at 11.6 min) from the degradation reaction of Chemical II in rat liver homogenate in positive (a) and negative (b) ion modes

Table 4

Measured accurate mass of the intermediate metabolites of Chemicals I and II formed in rat plasma and liver homogenate by LC-ESI-IT-TOF-MS analysis

Chemical

Positive observed m/z [M+H]+

(ppm error, mDa error)

Positive predicted m/z

Negative observed m/z [M−H]

(ppm error, mDa error)

Negative predicted m/z

Formula

DBE

Intermediate

metabolite 1

388.1383(0.77, 0.3)

388.1380

386.1236(0.26, 0.1)

386.1235

C14H22N5O6P

7.0

Intermediate

metabolite 2

356.0730(0.00, 0.0)

356.0730

354.0582(−0.56, −0.2)

354.0584

C10H13N5O4F3P

6.0

According to the [M+H]+ at m/z 356.0730 in the positive full-scan mass spectrum (Fig. 6a) and the [M−H] at m/z 354.0582 in the negative full-scan mass spectrum (Fig. 6b, both were 100 % relative signal intensity), the molecular formula of intermediate metabolite 2 was determined as C10H13N5O4F3P, indicating the formation of the corresponding monoester of Chemical II as a reaction intermediate.

Therefore, the intermediate metabolites in RLH and plasma were identified as the monoester analogue of Chemical I (peak at 11.2 min) and the monoester analogue of Chemical II (peak at 11.6 min). No significant decomposition of Chemical III was observed in the rat plasma and liver homogenate.

The stability of series 2 (Chemical II-1-6) in rat plasma and liver homogenate

The above results suggested that Chemical II was considerably stable in rat plasma, and it could also be transformed into an active structure (PMEA) in RLH by a high degree of activation and through a relatively simple metabolic pathway, as detected by MS. This indicates that Chemical II may be a promising liver-targeting anti-HBV ester of PMEA. Next, based on the structure of Chemical II, six analogues (Fig. 1; Table 1, series 2; ethyl PMEA-6-N-carbonate, II, n-propyl PMEA-6-N-carbonate, III, phenyl PMEA-6-N-carbonate, IV, isopropyl PMEA-6-N-carbonate, V, and n-butyl PMEA-6-N-carbonate, VI) were synthesized and further evaluated in rat plasma and liver homogenate. However, all six analogues were found to lower degree of prodrug activation in liver homogenate than those shown by Chemical II, which indicated that these six chemicals (Series 2 ester analogues) could not possess a higher therapeutic index than Chemical II in vivo (Table 3).

Discussion

Because of their physicochemical properties, chemicals containing an ionic polar group such as a carboxyl or phosphonate would normally be poorly absorbed by the gastrointestinal tract (Hecker and Erion 2008). With an ester modification of the carboxyl or phosphonate group, the absorption of these drugs can be improved. One effective approach involves a mixed alkyl (acyloxy) ester prodrug, dipivaloyloxymethyl ester of PMEA, which has been used to enhance the absorption of PMEA (Schultz 2003). However, the major barrier to the development of acyloxy ester prodrugs is their low stability. Aliphatic or aryl ester prodrugs are sometimes successful in solving the absorption issues, but most of the time, they show lower aqueous solubility and/or have decreased potency compared to that shown by the parent drug, owing to the formation of stable prodrugs that lead to incomplete bioconversion. For any prodrug, there should be a correct balance between chemical and metabolic stability. There has not been sufficient evidence and research on the stability of the phosphonate ester prodrugs using bis(trifluoroethyl) and bis(isopropyl) groups as promoieties. In this work, we first compared the effect of bis(trifluoroethyl) and bis(isopropyl) groups as promoieties on the stability of the PMEA ester prodrugs. We found that the bis(trifluoroethyl) ester modification of the PMEA had better chemical and metabolic characteristics than the bis(POM)- PMEA.

Previously, electron-releasing or electron-withdrawing properties of an alkyl side chain have been shown to affect the stability of an ester bond. Because both bis(trifluoroethyl) and bis(isopropyl) ester side chains offer electron-withdrawing and electron-donating effects, respectively, they provide good examples for evaluation of the effect of the alkyl group on the stability of the prodrugs. In this study, bis(trifluoroethyl) and bis(isopropyl) groups were chosen as the promoieties because both have been successfully shown to enhance the oral bioavailability of drugs in rats (Starrett et al. 1994). It is known that blood serum and plasma contain various enzymes that catalyze the hydrolysis of esters. Therefore, the hydrolysis of the newly synthesized phosphonate esters in rat plasma and the liver S9 fraction was investigated. To evaluate the electronic effect on stability of prodrugs, we first compared their activation in the liver homogenate fraction.

A prodrug’s ability to be converted quantitatively to the parent drug in the liver by an enzyme-catalyzed and/or chemical reaction is crucial to the success of a liver-targeted prodrug strategy. As expected, simple esterification of bis(isopropyl) (Chemical III) did not produce PMEA with detectable levels, which indicated that this compound may not generate the parent compound in vivo. Conversely, the bis(2,2,2-trifluoroethyl) ester of PMEA was converted to PMEA in RLH at similar levels compared to bis(POM)-PMEA (all about 100 % release) (Table 3; Fig. 2), which indicated that there was an excellent degree of prodrug activation. However, a successful liver-target prodrug strategy depends not only on liver activation but also on good chemical and plasma stability. Next, the degradation profiles of Chemicals I–III were examined in rat plasma, PBS, and heat-inactivated plasma or liver S9. Bis(trifluoroethyl) (Chemical II) showed ~4.75-fold improved half-life in rat plasma than that of the reference compound bis(POM)-PMEA (Chemical I). More importantly, there is no diminution in prodrug activation for Chemical II, judging by the high levels of PMEA generated in rat liver. Further stability of these newly synthesized compounds in PBS solution, inactivated rat plasma, or liver homogenate was examined, which indicated that the prodrugs are more stable in all control conditions as compared to their biological matrices, and no trace of PMEA was observed. For Chemicals I and II, in the present study, they were found to undergo rapid, non-enzymatic breakdown to the monoester by a non-esterase-mediated mechanism. This type of degradation profile for the reference compound (Chemical I) has been previously observed by Meier et al. (2008).

In this study, we found that bis(POM)-PMEA (Chemical I) was the least stable in both rat liver S9 and plasma systems (Table 3). It proved to be more sensitive to enzymatic hydrolysis than the bis(isopropyl) (Chemical III) and bis(trifluoroethyl) (Chemical II) derivatives, This observation is consistent with the previously reported findings that no intact bis(POM)-PMEA or monoester was detected in the plasma of animals or humans after oral or intravenous administration of the prodrug (Cundy et al. 1994; Naesens et al. 1996).

The results from rat plasma incubation also show that the introduction of an electron-donating group was advantageous for increasing compound stability, because the bis(isopropyl) analogue did not show significant degradation in rat plasma. The remarkable stability of the bis(isopropyl) analogue suggested the benefit of electronic-donating groups and this effect on stability is worth further investigation. In contrast, the addition of an electron-withdrawing bis(trifluoroethyl) group led to a significant improvement in the stability in rat plasma and a similar liver-specific release of PMEA compared to the control compound, bis(POM)-PMEA.

Therefore, a series of 6-substituted amide-alkyl carbonate prodrugs was synthesized with an altered methoxy group on the phenyl ring of bis(2,2,2-trifluoroethyl) ester of PMEA to explore if the liver-targeted release and plasma stability could be improved further. Unfortunately, these compounds exhibited lower prodrug activation in liver homogenate than those shown by Chemical II, which indicated that the 6-substituted amide-alkyl carbonate prodrugs might not be appropriate for liver-targeted release.

On the basis of HPLC–UV and LC-ESI-IT-TOF-MS analysis, the degradation pathway and intermediate metabolites are proposed in Supplemental Fig. 1. As summarized in Supplemental Fig. 1, the prodrugs (A) underwent a two-step bioconversion process, resulting in the formation of the active metabolite, PMEA. The first step was ①a fast nonenzymatic chemical breakdown reaction followed by ②rate-determining hydrolysis of the ester bond of the monoester intermediate. The monoester of PMEA was the only intermediate degradation product observed. We found that the efficiency of the compounds in releasing PMEA was hindered by the hydrolysis activation step. The cleavage of the C–O bond is considered to be the first step in the chemical reaction of the compounds, and the prodrug degraded to the monosubstituted PMEA, followed by a repeat of hydrolysis cleavage to release PMEA. It is noteworthy that the decomposition pathway compared favorably with the previously reported result: phosphonodiester derivative → phosphonomonoester → PMEA (Benzaria et al. 1996).

In summary, a variety of bis(alkyl) phosphonate esters of PMEA (Series 1 and 2) were synthesized to evaluate their suitability as prodrugs. Bis(2,2,2-trifluoroethyl) ester of PMEA generated high levels of PMEA in rat liver S9 but was more stable in plasma, thereby demonstrating both enhanced metabolic stability compared to ADV in rat plasma and rapid prodrug cleavage in RLH. Such a prodrug compound may have potential for application in liver-targeting release or controlled-release systems.

Acknowledgments

This work was supported by a grant from the National Natural Science Foundation of China (No. 81001469), National Science and Technology major project of the Ministry of Science and Technology of China (Grant No. 2012ZX09301003-001-010) and China National Science and Technology major special Project New Drug Innovation (2012ZX09301003-001-007).

Supplementary material

12272_2013_300_MOESM1_ESM.tif (214 kb)
Supplementary Fig. 1Degradation pathway of PMEA ester analogues. (TIF 215 kb)

Copyright information

© The Pharmaceutical Society of Korea 2013

Authors and Affiliations

  • Sha Liao
    • 1
  • Shi-Yong Fan
    • 1
  • Qin Liu
    • 1
  • Chang-Kun Li
    • 1
  • Jia Chen
    • 1
  • Jing-Lai Li
    • 1
  • Zhi-Wei Zhang
    • 1
  • Zhen-Qing Zhang
    • 1
  • Bo-Hua Zhong
    • 1
  • Jian-Wei Xie
    • 1
  1. 1.Beijing Institute of Pharmacology and ToxicologyBeijingPeople’s Republic of China

Personalised recommendations