Department of Pharmaceutical Chemistry, The University of Kansas,
Lawrence, Kansas
In vitro stability and in vivo pharmacokinetic studies of a model
opioid peptide, H-Tyr-D-Ala-Gly-Phe-D-Leu-OH
(DADLE), and its cyclic prodrugs (acyloxyalkoxy-based cyclic prodrug of
DADLE, coumarinic acid-based cyclic prodrug of DADLE, and
oxymethyl-modified coumarinic acid-based cyclic prodrug of
DADLE) were conducted. The enzymatic stability of DADLE and its
prodrugs in various biological media was determined at 37°C in the
presence and absence of paraoxon, a known esterase inhibitor. The
prodrugs exhibited metabolic stability to exo- and endopeptidases, and
esterase-catalyzed bioconversion of the prodrugs to DADLE was observed.
For pharmacokinetic studies in rats, various biological samples (blood,
bile, urine, and brain) were collected after i.v. administration of
DADLE and its prodrugs. The samples were analyzed by high-performance
liquid chromatography with tandem mass spectrometric detection, and the
conversion from the prodrugs to intermediates to DADLE was monitored.
The prodrugs exhibited similar pharmacokinetic properties and showed
improved stability compared with DADLE in rat blood. This increased
stability led to higher plasma concentrations of DADLE after i.v.
administration of the prodrugs compared with i.v. administration of
DADLE alone. In terms of elimination pathways, metabolism by
endopeptidases was the major route for DADLE elimination, whereas rapid
biliary excretion was the major route of elimination for the prodrugs. The rapid elimination of the prodrugs by the liver and the formation of
stable intermediates after esterase hydrolysis limited the bioconversion efficiencies of the prodrugs to DADLE after i.v. administration. The substrate activity of the prodrugs for efflux transporters (e.g., P-glycoprotein) in the blood-brain barrier significantly restricted their access to the brain.
 |
Introduction |
The
delivery of opioid peptides to the brain has presented significant
challenges to pharmaceutical scientists due to the poor
biopharmaceutical properties of the peptides. These include their
lability to metabolism by exo- and endopeptidases and their low
permeation across the blood-brain barrier (BBB) (Fricker and Drewe,
1996
; Pauletti et al., 1996a, 1997; Prokai, 1998). The issue of
metabolic lability has, for all practical purposes, been solved by
medicinal chemists through the design of novel peptide bond
bioisosteres to replace metabolically labile peptide bonds (Sawyer,
1995). Until recently, however, medicinal chemists have had less
success in manipulating the structures of opioid peptides to achieve
good BBB permeation while still retaining high affinity and selectivity
for opioid receptors.
Through prodrug strategies, some progress has been made recently in
improving the BBB permeation characteristics of opioid peptides (Greene
et al., 1996; Misicka et al., 1996; Patel et al., 1997; Prokai et al.,
2000). For a prodrug strategy to be successful in delivering opioid
peptides to the brain, the following criteria must be met: 1) the
prodrug should have favorable physiochemical properties (e.g.,
hydrophobicity, low hydrogen-bonding potential, and no charge) for
transcellular permeation across the BBB; 2) the prodrug should exhibit
good "intrinsic" permeability across the BBB but not be a substrate
for efflux transporters (e.g., P-gp); 3) the prodrug should be a good
substrate for enzymes (e.g., esterases) in the brain that would
catalyze its bioconversion to the opioid peptide but have less activity
as a substrate for the same enzymes in other biological media/tissues
or for other enzymes that could lead to nonproductive metabolism of the
prodrug (e.g., endopeptidases); and 4) the prodrug should have a
reasonable plasma half-life and not be rapidly metabolized in the blood
compartment or cleared by the liver or kidney or be extensively
protein-bound.
Having these criteria in mind, our laboratory synthesized cyclic
prodrugs of the opioid peptide
H-Tyr-D-Ala-Gly-Phe-D-Leu-OH (DADLE) using an
acyloxyalkoxy (AOA) linker (Bak et al., 1999b), a coumarinic acid (CA)
linker (Wang et al., 1999), and an oxymethyl-modified coumarinic acid
(OMCA) linker (Ouyang et al., 2002b) (Fig.
1). These prodrugs (AOA-DADLE, CA-DADLE,
and OMCA-DADLE) were designed to undergo bioconversion to DADLE via
mechanisms involving enzyme-catalyzed hydrolysis of the ester bond
linking the C-terminal amino acid of the peptide to the linker. The
resulting "intermediates" (Fig. 1, I-III) were then designed to
degrade chemically to form DADLE. In previous studies (Bak et al.,
1999a; Gudmundsson et al., 1999b; Ouyang et al., 2002a; Tang and
Borchardt, 2002a,b), our laboratory has shown that these prodrugs met
criterion 1 and part of criterion 3.
With respect to criterion 2, our laboratory determined the cell
permeation characteristics of AOA-DADLE, CA-DADLE, and OMCA-DADLE using
various cell culture models (e.g., Caco-2 cells, Madin-Darby canine
kidney cells, and Madin-Darby canine kidney cells transfected with drug efflux transporters; Bak et al., 1999a; Ouyang et al., 2002a;
Tang and Borchardt, 2002a,b). The results of these in vitro cell
culture experiments showed that AOA-DADLE, CA-DADLE, and OMCA-DADLE
exhibited poor cell permeation characteristics. Unlike the poor cell
permeation characteristics of DADLE, which resulted from undesirable
physiochemical properties (e.g., charge and high hydrogen-bonding
potential), the poor cell permeation of the cyclic prodrugs was shown
to result from their substrate activities for efflux transporters
(e.g., P-gp, multidrug resistance-associated protein 2). If these
efflux transporters were inhibited, the intrinsic cell
permeation characteristics of AOA-DADLE, CA-DADLE, and OMCA-DADLE were
significantly better than those of DADLE (Bak et al., 1999a; Ouyang et
al., 2002a; Tang and Borchardt, 2002a,b), thus satisfying in part
criterion 2.
These in vitro chemical/enzymatic studies and cell permeation studies
have provided valuable insights into whether AOA-DADLE, CA-DADLE, and
OMCA-DADLE satisfy criteria 1 and 2 mentioned above. However, to assess
whether these prodrugs satisfied criteria 3 and 4, in vivo experiments
must be conducted. The overall objective of the present work was to
evaluate the potential of the cyclic prodrugs for delivering DADLE to
the brain by investigating their in vitro stability in various
biological media and by in vivo pharmacokinetic studies in rats.
Therefore, experiments were designed to characterize the following
biopharmaceutical properties of AOA-DADLE, CA-DADLE, and OMCA-DADLE: 1)
their in vitro bioconversion rates in biological media (blood)/tissues
(brain/liver), 2) their propensity to bind to plasma proteins, 3) their
bioconversion rates and efficiencies in delivering DADLE after i.v.
administration, 4) their kinetics and routes of elimination after i.v.
administration, and 5) their ability to permeate the BBB after i.v. administration.
| |
Materials and Methods |
Materials.
The opioid peptides (DADLE and its internal
standard [Leu5]-enkephalin), diethyl
p-nitrophenyl phosphate (paraoxon, approx. 90%),
p-nitrophenyl butyrate (PNPB, approx. 98%), guanidine
hydrochloride (>99%), dimethyl sulfoxide (DMSO, >99.5%),
polyethylene glycol (average mol. wt. 300) (PEG300), and Hanks'
balanced salt solution (HBSS) (modified) were purchased from
Sigma-Aldrich (St. Louis, MO). Metofane (methoxyflurane;
Schering-Plough, Kenilworth, NJ) was obtained from the Animal Care Unit
(The University of Kansas, Lawrence, KS). Prodrugs of DADLE and
prodrugs of [Leu5]-enkephalin (internal
standards) were synthesized in our laboratory following procedures
described elsewhere (Bak et al., 1999b; Wang et al., 1999; Ouyang et
al., 2002b). All other chemicals were of the highest purity available
and used as received. All solvents were HPLC grade, including
deionized, ultrafiltered water (Fisher Scientific, Fair Lawn, NJ).
In Vitro Stability and Protein Binding Studies.
Human blood
was obtained from the Watkins Health Center (The University of Kansas).
Rat blood was obtained from male Sprague-Dawley rats (Animal Care Unit,
The University of Kansas). Fresh blood was centrifuged immediately at
1800g (model 59A Micro-Centrifuge; Fisher Scientific,
Pittsburgh, PA) and 4°C for 5 min and plasma was collected. For
stability studies, plasma was diluted to 90% (v/v) with HBSS, pH 7.4, to maintain the pH of the solution during the experiment. Rat livers
and brains were obtained from male Sprague-Dawley rats (Animal Care
Unit, The University of Kansas). The tissues were blotted to dryness
and cut into small pieces after weighing. The tissue pieces were
homogenized immediately on ice with ice-cold HBSS (1 ml/1 g of tissue)
using a glass homogenizer (15 strokes, pestle/wall clearance 0.25-0.76
mm; Wheaton, Philadelphia, PA). Aliquots (approx. 1.5 ml) were frozen
and kept at 
80°C until used. Before each experiment, the homogenate
was quickly thawed and rehomogenized on ice with an equal volume of
ice-cold HBSS. Cell debris and nuclei were removed by centrifugation at
10,000g and 4°C for 10 min using a model 59A
Micro-Centrifuge (Fisher Scientific). The supernatant was collected for
stability studies.
The enzymatic stability of DADLE and its cyclic prodrugs was studied in
various biological media at 37°C in the presence and absence of
paraoxon, a known esterase inhibitor. Each compound (~100 µM, final
concentration) was incubated with the biological matrix for at least
twice its half-life in a temperature-controlled shaking water bath (60 rpm, 37 ± 0.5°C). To test the effects of an esterase inhibitor
on the rates of degradation of the cyclic prodrugs, the biological
media were preincubated with paraoxon (1 mM, final concentration) for
15 min at 37°C before the prodrugs were added. Samples (20 µl) were
taken at various times, and the esterase activity was immediately
quenched by adding 150 µl of a freshly prepared 6 N guanidine
hydrochloride solution in acidified HBSS [0.01% (v/v) phosphoric
acid, pH 
3].
Total esterase activities and total protein concentrations in these
biological media were determined according to methods described
elsewhere (Pauletti et al., 1996b). Briefly, PNPB was used as the
substrate to assess the total esterase activities in the biological
media. p-Nitrophenol, the final product of the enzymatic
reaction, was monitored at 
= 420 nm. Esterase activities were
expressed as units per milligram of protein. One unit represents the
amount of enzyme that catalyzes the formation of 1 µmol of p-nitrophenol per minute in HBSS, pH 7.4, at 25°C. Total
protein concentrations in the biological media were determined using
the Protein Assay (Bio-Rad, Hercules, CA) with bovine serum albumin as standard.
Finally, plasma protein binding for DADLE and its prodrugs was
determined using a filtration method. DADLE and its prodrugs were
spiked into pooled rat plasma at various concentrations
(n = 3). After equilibrating at 4°C for 30 min,
plasma samples were centrifuged at 5000g and 4°C for up to
3 h using Ultrafree MC-5000 NMWL filter units (Millipore
Corporation, Bedford, MA) to separate the protein-bound drugs from free
drugs. For control samples, drugs were spiked into filtered blank
plasma as 100% recovery standards. All samples were analyzed
immediately, and the percentage of protein binding for each compound
was calculated based on free drug concentrations divided by total drug
concentrations in control samples.
In Vivo Pharmacokinetic Studies.
For pharmacokinetic
studies, we used male Sprague-Dawley rats (200-250 g) with a cannula
chronically implanted in the jugular vein or carotid artery (Harlan,
Indianapolis, IN). The rats were housed individually and fasted
overnight before use. Water was allowed ad libitum. For each compound
to be studied, three to six rats were each given a 1 mg/kg i.v. dose of
the drug (200-250 µl). The vehicles used for i.v. dosing were
various combinations of saline, ethanol, PEG300, and DMSO, depending on
the solubility of the compounds. For DADLE, saline was used as the
solvent; 20% (v/v) ethanol and 80% (v/v) saline were used for
AOA-DADLE; 5% (v/v) DMSO, 47.5% (v/v) PEG300, and 47.5% (v/v) saline
were used for CA-DADLE; and 5% (v/v) DMSO, 20% (v/v) PEG300, and 75%
(v/v) saline were used for OMCA-DADLE. For i.v. injections, the rats were anesthetized with metofane, a small incision was made on the
medial surface of the hind leg, and the injection was made into the
femoral vein. The incision was closed with wound clips and the rats
were allowed to recover from the anesthetic (5-10 min). Blood samples
(approx. 0.2 ml) were withdrawn from the jugular vein or carotid artery
cannula at 2, 5, 10, 15, 20, 30, 40, 50, 60, 80, 100, and 120 min after
drug administration and centrifuged immediately at 1800g and
4°C for 5 min (Micromax centrifuge; International Equipment Company,
Needham Heights, MA). Plasma samples were collected and stored at
80°C until analysis.
To determine the routes of elimination for DADLE and its prodrugs, the
experiments described above were conducted using three to eight male
Sprague-Dawley rats with chronically implanted bile duct cannulas
(200-250 g; Harlan). The bile was collected for various time intervals
(5-min intervals for 20 min, 10-min intervals up to 60 min, and three
20-min intervals through 2 h) after drug administration and stored
at 80°C until analysis. Urine samples were also collected for up to
2 h to determine the clearance by the kidneys.
To determine the disposition of DADLE and its prodrugs into the brain
after i.v. administration, male Sprague-Dawley rats (200-250 g) were
used. The rats were fasted overnight before use but were allowed access
to water ad libitum. For each compound studied, three rats were given a
1 mg/kg i.v. dose of the drug. At 10 min after i.v. administration, the
animals were sacrificed by decapitation and the brains removed. The
samples were stored at 80°C until analysis.
Sample Analysis.
For in vitro stability studies,
high-performance liquid chromatography (HPLC) with UV detection was
performed using an LC-10A gradient system (Shimadzu, Tokyo, Japan)
consisting of two LC-10AS pumps, an SCL-10A system controller, and an
SIL-10A autoinjector with a sample cooler. The chromatographic data
were acquired and analyzed using CLASS-VP version 4.2 Chromatography
Data System (Shimadzu). The HPLC analysis was conducted using a C18
reversed-phase column (250 × 4.6 mm i.d., 300 Å; Vydac,
Hesperia, CA) equipped with a C18 guard column (Vydac). Gradient
elution was performed at a flow rate of 1 ml/min from 26 to 58% (v/v)
acetonitrile in water with 0.1% (v/v) trifluoroacetic acid. The
eluents were detected by UV ( = 214 nm). For sample
preparation, aliquots (~150 µl) of the sample mixtures were
transferred to Ultrafree MC-5000 NMWL filter units (Millipore
Corporation) and centrifuged at 5000g and 4°C for up to
3 h (Micromax centrifuge). The filtrate was collected and kept at
4°C for HPLC analysis. The disappearance of the prodrugs, possible
formation of any stable intermediates, and appearance of DADLE were monitored.
For in vivo pharmacokinetic studies, HPLC with tandem mass
spectrometric detection (LC/MS/MS) methods were developed. These methods have been described extensively elsewhere (Yang et al., 2002). Briefly, a Quattro LC or Quattro Micro triple quadrupole mass spectrometer (Micromass, Beverly, MA) was used. The liquid chromatography was conducted using a 2690 HPLC System (Waters, Milford,
MA). Data acquisition and analysis were performed using MassLynx
version 3.5 software (Micromass). A C18 reversed-phase column (50 × 1.0 mm i.d., 300 Å; Vydac) was used as the analytical column, with
the column temperature maintained at 25°C to ensure reproducible
separation. Two mobile phases were used to generate a linear gradient
to allow simultaneous analysis of DADLE and its prodrugs in a single
run. Mobile phase A consisted of water and 0.1% (v/v) formic acid and
mobile phase B was acetonitrile with 0.1% (v/v) formic acid. For
sample recording, multiple reaction monitoring was used so that several
compounds could be monitored simultaneously. Even though accurate
concentrations of the intermediates (Fig. 1, I-III) could not be
determined because the intermediates converted to DADLE during the mass
spectrometric analysis, it was possible to estimate their amounts based
on the levels of DADLE detected by LC/MS/MS. For intermediate and
metabolite identification during in vitro stability studies, mass
spectrometric analysis was conducted using the same LC/MS/MS system
under similar conditions. A full scan covering a broad mass range
instead of multiple reaction monitoring was conducted to identify
possible metabolites and confirm the formation of any stable intermediates.
For in vivo studies, different sample preparation methods were used,
depending on the biological matrix. For plasma samples, protein
precipitation with acetonitrile was used. To a 100-µl plasma sample,
200 µl of acetonitrile containing internal standards was added to
precipitate the plasma proteins. After vortex mixing, the precipitated
proteins were removed by centrifugation at 3000g and 4°C
for 5 min. The supernatant was evaporated to dryness using a Centrivap
concentrator (Labconco, Kansas City, MO), and the residue was
reconstituted in 100 µl of 10% (v/v) acetonitrile. It was
centrifuged again at 10,000g and 4°C for 5 min, and a 50 µl supernatant was injected for LC/MS/MS analysis. For brain samples, solid phase extraction (SPE) had to be used due to the complicated biological matrix. First, the whole brain was quickly homogenized on
ice with 5 ml of ice-cold HBSS using a glass homogenizer (30 strokes;
Wheaton) after weighing. Aliquots (approx. 1.5 ml) were immediately
frozen and kept at 80°C until used. Before sample preparation, the
homogenate was quickly thawed and 3 ml of ice-cold acetonitrile along
with 10 µl of the internal standard solution mixture was added to 1 ml of homogenate to precipitate the proteins. After vortex mixing, the
precipitated proteins were removed by centrifugation at
5000g and 4°C for 5 min. Because the supernatant had high
organic content, it was evaporated and reconstituted with water for
compatibility with SPE. A Centrivap concentrator was used, and the
organic solvent was evaporated under vacuum. After approx. 2 h,
the concentrated supernatant was diluted with 1 ml of deionized water,
and the solution was loaded onto the cartridges for SPE (Yang et al.,
2002). After DADLE and its prodrugs along with their internal
standards were extracted from the brain samples, the final solution was
evaporated to dryness and the residue was reconstituted in 100 µl of
10% (v/v) acetonitrile. It was centrifuged again at 10,000g
and 4°C for 1 min, and 50 µl of the supernatant was injected for
LC/MS/MS analysis. Bile and urine samples were diluted and injected
directly for LC/MS/MS analysis because they had relatively high drug concentrations.
Data Analysis.
For in vitro stability studies, mass balance
was monitored to study the bioconversion mechanisms of the prodrugs.
The apparent half-lives (t1/2) of the
prodrugs were calculated using SigmaPlot (SPSS Science, Chicago, IL)
from the pseudo first order rate constants obtained by linear
regression of plots of log drug concentration remaining versus time.
All results are presented as mean ± S.D.
For in vivo pharmacokinetic studies, the plasma data were fitted to a
two-compartment model, and various pharmacokinetic parameters were
calculated using WinNonlin (Pharsight, Mountain View, CA). To quantify
the bioconversion efficiency of the cyclic prodrugs, the relative
bioavailability of DADLE after i.v. administration of the prodrugs was
calculated. The values were expressed as the ratio of the AUC of DADLE
converted from the prodrug versus AUC of DADLE administrated alone
adjusted by dose. All results are presented as mean ± S.E.
For statistical analysis, analysis of variance and Student's
t test were used where appropriate. A probability of less
than 0.05 (P < 0.05) was considered statistically significant.
| |
Results |
In Vitro Stability and Protein Binding Studies.
The enzymatic
stability of DADLE and its cyclic prodrugs (AOA-DADLE, CA-DADLE, and
OMCA-DADLE) was studied in various biological media, including human
plasma, rat plasma, rat liver homogenate, and rat brain homogenate. The
bioconversion profiles for DADLE and its prodrugs in rat plasma are
presented in Fig. 2. DADLE disappeared
slowly in rat plasma. The metabolites of DADLE that were identified by
LC/MS/MS seem to result from hydrolysis of the pentapeptide catalyzed
by endopeptidases (data not shown). The half-lives of all three
prodrugs in rat plasma were much shorter than that of DADLE due to
rapid ester bond cleavage catalyzed by esterases. Conversion to DADLE
was observed for all these prodrugs, but complete mass balance was not
achieved. Some interesting differences were observed between the three
prodrugs in terms of the formation of intermediates. For example,
whereas intermediate I from AOA-DADLE was not detected, significant
accumulation of intermediate II from CA-DADLE and intermediate II (and
probably III) from OMCA-DADLE was observed (Fig. 2).
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 2.
Stability of DADLE (A), AOA-DADLE (B), CA-DADLE (C),
and OMCA-DADLE (D) in rat plasma in vitro (n = 3).
The prodrugs (  ), DADLE ( ), and intermediates ( )
were analyzed by HPLC. See Materials and Methods for
experimental details.
|
|
Table 1 summarizes the half-lives of
DADLE and its prodrugs in various biological media. DADLE was
relatively stable in human and rat plasma, but its half-lives were much
shorter in rat liver and brain homogenates. For the cyclic prodrugs,
conversion to DADLE was observed in all biological media but with
different half-lives. It is interesting to note that AOA-DADLE and
OMCA-DADLE, which have the same local structure around the ester bond,
had similar half-lives in all of the biological media/tissues tested. In contrast, CA-DADLE, which has a different structure around the ester
bond, exhibited quite different stability in these biological media and
tissues. Using PNPB as a substrate, the specific esterase activities
for the four biological media were determined to be 0.52 U/mg protein
(rat liver homogenate) >0.20 U/mg protein (rat brain homogenate)
>0.05 U/mg protein (rat plasma) >0.02 U/mg protein (human plasma). Of
the three prodrugs, CA-DADLE had half-lives that decreased as the
esterase activities of the biological media increased; no such
correlation was observed for the other two prodrugs. The half-lives of
AOA-DADLE and OMCA-DADLE were generally longer than that of CA-DADLE in
the biological media. With the addition of paraoxon, the esterase
activities for all biological media were reduced to less than 0.02 U/mg
protein. Although paraoxon had no significant effects on the stability
of DADLE, the half-lives of all prodrugs were much longer in the
presence of paraoxon, and different inhibition effects were observed
for the three different prodrugs.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Stability of DADLE and its cyclic prodrugs in various biological media
and tissues
The stability of DADLE and its cyclic prodrugs in human and rat plasma
and various rat tissue homogenates was determined (n = 3) at 37°C in the presence and absence of paraoxon, a known esterase
inhibitor. The disappearance of the prodrugs and their conversion to
DADLE was monitored by HPLC with UV detection. See Materials and
Methods for experimental details.
|
|
In protein binding studies, the prodrugs generally had much higher
bound fractions than did the parent drug DADLE. The protein binding
percentages in rat plasma were 47 ± 11% for AOA-DADLE, 88 ± 6% for CA-DADLE, 60 ± 6% for OMCA-DADLE, and only 21 ± 4% for DADLE.
In Vivo Pharmacokinetic Studies.
The pharmacokinetic
properties and bioconversion efficiencies of the prodrugs were studied
in rats after i.v. administration of DADLE and its prodrugs.
Pharmacokinetic parameters (AUC, clearance, Vss,
ke, and
t1/2) were calculated and are provided
in Table 2. After i.v. administration,
DADLE disappeared rapidly from the plasma with a short elimination
half-life (Fig. 3A). In comparison, the
prodrugs were more stable and had longer elimination half-lives (Fig.
3, B-D). Continuous generation of DADLE from all three prodrugs was
observed during the time course of the experiment. However, the DADLE
concentrations arising from the prodrugs were always significantly
lower than the observed prodrug concentrations (Fig. 3). Assuming the
bioavailability of DADLE after i.v. administration was 100%, the
bioconversion efficiencies to DADLE were 35 ± 8% for AOA-DADLE,
39 ± 13% for CA-DADLE, and 5 ± 1% for OMCA-DADLE. Even
with incomplete conversions, the prodrugs still provided sustained
release of DADLE in the plasma, which led to higher DADLE
concentrations over longer duration compared with administration of
DADLE alone. Interestingly, the intermediates (Fig. 1, II and probably
III) arising from esterase-catalyzed hydrolysis of CA-DADLE and
OMCA-DADLE were also detected in plasma (Fig. 3, C and D). As predicted
from the in vitro results, intermediate I arising from AOA-DADLE was
not detected (Fig. 3B).
View this table:
[in this window]
[in a new window]
|
TABLE 2
Pharmacokinetic parameters of DADLE, AOA-DADLE, CA-DADLE, and
OMCA-DADLE after i.v. administration to rats
Drug solutions containing DADLE or its prodrugs (1 mg/kg) in
appropriate solvent systems were administered i.v. to rats. Blood
samples (approximately 200 µl) were collected through the jugular
vein or carotid artery cannula at 2, 5, 10, 15, 20, 30, 40, 50, 60, 80, 100, and 120 min after drug administration (n = 3-6)
and analyzed by LC/MS/MS. See Materials and Methods for
experimental details.
|
|
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 3.
Time course for disappearance of DADLE (A), AOA-DADLE
(B), CA-DADLE (C), and OMCA-DADLE (D) after i.v. administration of the
respective drugs (n = 3-6). The prodrugs
(), DADLE (), and intermediates () were
analyzed by LC/MS/MS. See Materials and Methods for
experimental details.
|
|
Pharmacokinetic experiments were also conducted using bile
duct-cannulated rats to determine the routes of elimination for DADLE
and its prodrugs. Table 3 summarizes the
results of biliary clearance for DADLE and its prodrugs. Although the
bile recovery amount for DADLE was low, extensive biliary excretion was
observed for the prodrugs, with more AOA-DADLE being recovered in the
bile than the other two prodrugs (Fig.
4). The intermediates of CA-DADLE and
OMCA-DADLE also contributed to the overall clearance, but less than
10% of intermediates II and III were recovered in bile. In renal
clearance studies, less than 5% of DADLE and its prodrugs were
recovered in urine (data not shown).
View this table:
[in this window]
[in a new window]
|
TABLE 3
Bile recovery after i.v. administration of DADLE and its cyclic
prodrugs
Drug solutions containing DADLE or its prodrugs (1 mg/kg) in
appropriate solvent systems were administered i.v. to rats. Bile
samples were collected through the bile duct cannula at 5-min intervals
for 20 min, 10-min intervals up to 60 min, and three 20-min intervals
through 2 h (n = 3-8) and analyzed by LC/MS/MS.
See Materials and Methods for experimental details.
|
|
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 4.
Time course for recovery of the cyclic prodrugs in
bile after i.v. administration of AOA-DADLE (), CA-DADLE
( ), and OMCA-DADLE ( ) (n = 3-8).
See Materials and Methods for experimental details.
|
|
The disposition of DADLE and its cyclic prodrugs into the brain was
determined after i.v. administration of drugs. After sample preparation
and analysis, the amounts of DADLE and its prodrugs in the brain were
determined (Table 4). Brain uptake of
DADLE was very low, and the prodrugs did not show better permeation characteristics than DADLE itself. However, bioconversion of the cyclic
prodrugs to DADLE was observed in the brain (Table 4).
View this table:
[in this window]
[in a new window]
|
TABLE 4
Brain uptake of DADLE, AOA-DADLE, CA-DADLE, and OMCA-DADLE after i.v.
administration to rats
Drug solutions containing DADLE or its prodrugs (1 mg/kg) in
appropriate solvent systems were administered i.v. to rats. Brain
samples were collected at 10 min after drug administration
(n = 3) and analyzed by LC/MS/MS. See Materials
and Methods for experimental details.
|
|
| |
Discussion |
In Vitro Stability and Protein Binding Studies.
In all of the
various biological media and tissues used in these studies, conversion
of the prodrugs to DADLE was observed, but the rates of those
bioconversions differed. These dissimilarities could arise because the
prodrugs have different substrate activities for the same esterases or
they are substrates for different esterases. Differences were also
observed between prodrug half-lives and esterase activities (based on
the rates of PNPB hydrolysis) in the biological media/tissues,
indicating that several types of esterases might be involved in the
hydrolysis of the cyclic prodrugs. It is well known that there
are three predominant types of esterases (type A, B, and C), and PNPB
is a substrate for type B esterase (Walker and Mackness, 1987;
Takahashi et al., 1995; Huang et al., 1996). Because the half-lives of
CA-DADLE correlated with the type B esterase activities of the
biological media, this prodrug is probably a substrate mainly for this
form. For AOA-DADLE and OMCA-DADLE, it is more likely that both
prodrugs were hydrolyzed by type A and/or type C esterases.
For the overall bioconversion process, the prodrugs seem to have
different rate-limiting steps to the formation of DADLE. As shown in
Fig. 2, degradation of AOA-DADLE led to the formation of DADLE only.
The intermediate I was not detectable, suggesting that it very rapidly
degrades to DADLE, CO2, and formaldehyde. Therefore, the ester bond cleavage of AOA-DADLE seems to be the rate-limiting step in the bioconversion to DADLE. In contrast, CA-DADLE
degraded to DADLE with the formation of intermediate II and OMCA-DADLE
converted to DADLE with the formation of intermediates II and III (Fig.
1B). However, in both in vitro (Fig. 2) and in vivo (Fig. 3)
experiments, only intermediate II was detectable because the
lactonization reaction seems to be slow. Therefore, the conversion of
II to DADLE is the rate-limiting step in the overall bioconversion
process for both CA-DADLE and OMCA-DADLE.
A successful prodrug strategy for the delivery of an opioid peptide to
the brain has to include the preferential conversion of the prodrug to
the parent drug via either chemical or enzymatic reactions in the
target tissue (e.g., brain). In these studies, we have shown that the
cyclic prodrugs AOA-DADLE, CA-DADLE, and OMCA-DADLE are substrates for
esterases in all of the biological media/tissues tested (Table 1). The
prodrugs also exhibited increased metabolic stability to exo- and
endopeptidases because DADLE and not fragments of DADLE were detected
during the in vitro bioconversion experiments (Fig. 2). However, two
problems still exist with regard to the bioconversion of these
prodrugs. The first is that AOA-DADLE and OMCA-DADLE tend to undergo
more rapid bioconversion in rat plasma and liver than in brain (Table
1). CA-DADLE seems to have more favorable bioconversion characteristics
because the half-lives of this prodrug in rat plasma and brain
homogenate are approximately equal (Table 1). However, the
bioconversion in liver homogenate is extremely fast (Table 1). In
addition, the CA linker and OMCA linker have the added complication of
generating intermediates (Fig. 1, II and III) that seem to limit their
"bioconversion efficiencies" in vivo. To overcome these problems,
our laboratory is currently synthesizing analogs of the three linkers
in an attempt to alter the rates of esterase-catalyzed bioconversion
(i.e., increase bioconversion in brain and reduce bioconversion in
blood). Because the long half-lives of the intermediates (Fig. 1, II
and III) arising from CA-based and OMCA-based cyclic prodrugs seem to
also be a significant problem with these linkers, modified linkers having more rapid rates of conversion of the intermediate(s) to DADLE
are also being developed in our laboratory.
Plasma protein binding can exert a significant influence on the
pharmacokinetic properties and biological activities of a drug,
especially for brain-targeted compounds (Raub et al., 1993). For DADLE
and its prodrugs, more lipophilic compounds had higher fractions bound
in the order CA-DADLE > OMCA-DADLE > AOA-DADLE > DADLE, but none of the prodrugs had protein-binding values high enough
to substantially affect the distribution of the compounds.
In Vivo Pharmacokinetic Studies.
DADLE and its prodrugs
disappeared rapidly after i.v. administration, but the prodrugs had
improved stability in rat blood compared with DADLE (Fig. 3). At the
same time, continuous but incomplete conversion of the prodrugs to
DADLE was observed. As predicted from the in vitro results,
intermediate I arising from AOA-DADLE was not detected, but
intermediates II and III from CA-DADLE and OMCA-DADLE were present in
significant amounts in blood after i.v. administration of these cyclic
prodrugs. The slow conversions of III to II in OMCA-DADLE and II to
DADLE in CA-DADLE impact negatively on their abilities to delivery this opioid peptide in rat blood as well as to target tissue (i.e., brain).
On the basis of drug clearance results, only a small fraction of DADLE
was recovered in the bile and urine; therefore, metabolism by
endopeptidases seems to be the major route of elimination for DADLE.
The fast elimination of the prodrugs could be attributed mainly to
extensive bile excretion with minor kidney clearance, whereas
metabolism or deep tissue distribution was not a significant factor
(Table 3). The prodrug concentrations in the bile were generally 50 to
100 times higher than their plasma concentrations, which indicates an
active bile clearance mechanism, possibly mediated by efflux systems
(e.g., P-gp). In contrast, DADLE and the intermediates derived from the
prodrugs showed significantly lower bile excretion rates and seemed not
to be substrates for P-gp.
As discussed in the Introduction, for successful opioid peptide
delivery to the brain, a prodrug strategy should meet four criteria. On
the basis of results from previous and present studies, we now have
enough information to evaluate how AOA-DADLE, CA-DADLE, and OMCA-DADLE
meet these criteria. The prodrugs meet criterion 1 and were shown to
have transcellular permeation due to favorable physiochemical
properties (e.g., increased lipophilicity and no charge) and unique
solution structures (e.g., 
-turns) that reduce their
hydrogen-bonding potential (Bak et al., 1999b; Gudmundsson et al.,
1999a,b,c; Ouyang et al., 2002b). With respect to criterion 2, all
three prodrugs exhibited significantly better intrinsic cell permeation
characteristics than DADLE itself (Bak et al., 1999a; Ouyang et al.,
2002b; Tang and Borchardt, 2002a,b). However, in vitro cell culture
experiments suggest that the BBB permeation of the cyclic prodrugs may
be significantly restricted by their substrate activities for efflux
transporters (Bak et al., 1999a; Ouyang et al., 2002a; Tang and
Borchardt, 2002a,b). On the basis of the in vitro stability studies,
the prodrugs partially satisfy criterion 3 with good substrate
activities for esterases and stability to exo- and endopeptidases;
however, the sites of esterase-catalyzed bioconversion and the long
half-lives of the intermediates (Fig. 1, II and III) arising from
CA-based and OMCA-based cyclic prodrugs may limit their bioconversion
efficiencies in the brain. Finally, with respect to criterion 4, although the prodrugs have improved plasma half-lives compared with
DADLE, their biliary excretion rates are still too rapid, and the
problems associated with intermediates II and III generated from
CA-DADLE and OMCA-DADLE also decreased their bioconversion efficiencies
in vivo.
The brain uptake studies of the cyclic prodrugs described herein were
conducted to evaluate their ability to deliver DADLE to the brain after
i.v. administration. As expected based on their substrate activities
for P-gp (Bak et al., 1999a; Ouyang et al., 2002a; Tang and Borchardt,
2002a,b), the amount of DADLE found in the brain after prodrug
administration was very small. To investigate the BBB permeation of the
prodrugs and their substrate activities for P-gp, studies using an in
situ-perfused rat brain model were conducted, and the results are
reported in Chen et al. (2002).
In summary, esterase-catalyzed bioconversion of the prodrugs
(AOA-DADLE, CA-DADLE, and OMCA-DADLE) to DADLE was observed both in
vitro and in vivo. The cyclic prodrugs showed similarities in their
pharmacokinetic parameters and displayed improved stability compared
with the parent drug DADLE in vivo. However, the prodrugs were unable
to deliver significant amounts of DADLE to the brain because of their
rapid biliary excretion, poor BBB permeation, and slow conversion in
the brain. Therefore, to improve their potential to diffuse across the
BBB for better DADLE delivery into the brain, modifications will need
to be made in the linkers used to make these prodrugs to 1) stabilize
them to esterases in the blood compartment while increasing their
bioconversion rates in the brain, 2) decrease the chemical stability of
the intermediates to improve their bioconversion efficiencies to DADLE, 3) reduce their substrate activities for efflux transporters to achieve
better "apparent" permeation properties, and 4) reduce their
substrate activities for the transporter(s) in the liver to limit their
biliary clearance. Studies are currently ongoing in our laboratory to
improve the biopharmaceutical properties of these cyclic prodrugs.
This research was supported by a grant from the U.S. Public
Health Service (DA09315).
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
-glutamyl-dermorphin, a prodrug.
Life Sci
58:
905-911[CrossRef][Medline].
