Departments of
Metabolism and Pharmacokinetics (S.-J.L., D.C.H.,
J.W.H., R.M.F., K.J.R., R.E.W.) and
Analytical Research and
Development (B.M.W.), Bristol-Myers Squibb Pharmaceutical Research
Institute
 |
Introduction |
Squalene
synthase
(FPP/FPP1 farnesyl
transferase) (EC 2.5.1.21) catalyzes the reductive dimerization of two
molecules of FPP to form squalene, which is subsequently converted to
cholesterol via multiple enzymatic transformations (Poulter
and Rilling, 1981
; Dewar and Ruiz, 1987; Kuswik-Rabiega and Rilling,
1987; Sasiak and Rilling, 1988). This enzyme occupies a strategic site
at the final branch point in the pathway. The carbon flow beyond
squalene is destined exclusively to sterols, whereas the substrate for squalene synthase (FPP) also serves as a precursor for nonsteroidal isoprenes, including coenzyme Q, dolichols, heme A, and prenylated proteins (Goldstein and Brown, 1990). Selective inhibition of squalene
synthase does not suppress the production of the nonsteroidal isoprenes, as might be seen with inhibitors of enzymes earlier in the
pathway. In addition, FPP is the last water-soluble intermediate in the
pathway and has known routes of metabolism, in the event of
intracellular accumulation (Gonzalez-Pacanowska et al.,
1988; Keung, 1991). Thus, inhibition of squalene synthase may have
advantages over intervention earlier in the cholesterol biosynthetic
pathway.
Recently, inhibitors of squalene synthase have received considerable
attention in the search for new antihypercholesterolemic agents
(Bergstrom et al., 1993; Dawson et al., 1992;
Sidebottom et al., 1992; Magnin et al., 1996;
Dickson et al., 1996; Biller et al., 1988, 1996).
An isoprenyl-1,1-bisphosphonate was found to be a potent inhibitor of
squalene synthase and an effective cholesterol-lowering agent in animal
models (Ciosek et al., 1993). However, this inhibitor
exhibited undesirable toxicological and metabolic properties, including
elevated plasma transaminase (alanine transaminase and aspartate
transaminase) levels in mice after an iv dose, poor oral absorption,
and retention of radioactivity in bone after an iv dose of
14C-labeled compound to rats, which prevented its
being developed as an oral cholesterol-lowering agent. I
[(E,E)-6,10,14-trimethyl-1-phosphono-5,9,13-pentadecatriene-1-sulfonic acid] is an analog of the prototypical 1,1-bisphosphonate inhibitor in
which one of the phosphonate groups has been replaced by a sulfonate
group. I exhibited greatly reduced acute effects on plasma
transaminase levels in mice. It exhibited an IC50
of 15 nM for rat liver microsomal squalene synthase and
ED50 values of 0.09 and 10.3 mg/kg for the
inhibition of cholesterol biosynthesis in rats after iv and po
administration, respectively (Magnin et al., 1996). The
purpose of this study was to evaluate the disposition of
[14C]I in rats, to determine whether
it possessed adequate metabolic properties to warrant development as an
oral antihypercholesterolemic agent. In anticipation of the poor
absorption of I, the absorption of
14C-labeled monoester (II)
[(E,E)-1-(ethoxyhydroxyphosphinyl)-6,10,14-trimethyl-5,9,13-pentadecatriene-1sulfonic acid], diester (III)
[(E,E)-1-(diethoxyphosphinyl)-6,10,14-trimethyl-5,9,13-pentadecatriene-1-sulfonic acid], and triester
(IV) [(E,E)-1-(diethoxyphosphinyl)-6,10,14-trimethyl-5,9,13-pentadecatriene-1-sulfonic acid, cyclohexyl ester] prodrugs of I was also studied in
BDC rats, to determine the number of acidic functions of the phosphonate and sulfonate groups that would have to be masked to
achieve good absorption. Fig. 1 depicts
the structures of I and the prodrugs (II,
III, and IV).
| |
Materials and Methods |
Materials.
[14C]I (2.49 µCi/µmol;
radiochemical purity, 95.6%),
[14C]II (2.39 µCi/µmol;
radiochemical purity, 97.3%),
[14C]III (2.57 µCi/µmol;
radiochemical purity, 96.8%), and [14C]IV (2.53 µCi/µmol;
radiochemical purity, 96.4%) were synthesized in the Bristol-Myers
Squibb Pharmaceutical Research Institute. TX-114 and Ecolite(+) liquid
scintillation fluids were purchased from National Diagnostics
(Manville, NJ) and ICN Biochemicals (Cleveland, OH), respectively.
Sprague-Dawley BR outbred albino rats were obtained from Harlan
Sprague-Dawley (Frederick, MD). Triethylamine was purchased from
Mallinckrodt (Paris, KY). TBAS was obtained from Aldrich Chemical Co.
(Milwaukee, WI).
Dosing and Sample Collection. Disposition in Jugular
Vein-Cannulated Rats.
Six male rats (~300 g) with indwelling jugular vein cannulae were
fasted overnight (18-20 hr) before dosing. Single 15-µmol/kg iv
bolus (15 µmol/ml, 1 ml/kg) and po (3.75 µmol/ml, 4 ml/kg) doses of
[14C]I, as aqueous solutions, were
separately given to groups of three rats each. Blood, urine, and feces
were collected periodically for 96 hr. Plasma was prepared from blood
by centrifugation. All samples were stored at 
20°C until analysis.
Disposition in BDC Rats.
Fifteen male rats (~300 g) with indwelling bile duct cannulae were
fasted overnight (18-20 hr) before dosing. Single 15-µmol/kg iv
bolus (15 µmol/ml, 1 ml/kg) and po (3.75 µmol/ml, 4 ml/kg) doses of
[14C]I, as aqueous solutions, were
separately administered to groups of three rats each. Single
15-µmol/kg po (3.75 µmol/ml, 4 ml/kg) doses of
[14C]II,
[14C]III, and
[14C]IV were given to the remaining
three groups of three rats each. II and III were
administered as aqueous solutions. IV was administered as a
solution in Mazola corn oil (GPC International, Inc., Englewood
Cliffs, NJ). Bile was collected from each rat at hourly intervals for
12 hr. Urine and feces were collected for 12 hr. Rats were
exsanguinated under Metofane (Mallinckrodt Veterinary Inc., Mundelin,
IL) anesthesia at 12 hr after dosing. Liver, lungs, kidneys,
heart, gastrointestinal tract, and a piece of femoral bone were
removed. All samples were stored at 20°C until analysis.
Determination of Total Radioactivity.
Total radioactivity in all samples, including the carcass, was
determined by liquid scintillation counting in a Packard TriCarb model
2200CA liquid scintillation counter (Packard Instrument Co., Downers
Grove, IL). Feces, tissues, and carcass were homogenized with an
appropriate volume (3-10 ml/g of feces or tissue) of water. Aliquots
(0.1-0.2 ml) of blood, plasma, or homogenates of feces or tissues were
digested with 1 ml of Soluene-350 tissue stabilizer (Packard
Instrument Co.) for 24 hr. The digested solutions were then bleached
with 1 ml of a 20% solution of benzoyl peroxide in toluene and were
neutralized with 0.1 ml of saturated sodium pyruvate in
methanol/glacial acetic acid/methanol (4:3:1, by volume). After the
sample was mixed with 15 ml of TX-114 liquid scintillation fluid,
radioactivity was determined in a liquid scintillation counter.
Radioactivity in bile (50 µl) and urine (0.2 ml) was directly counted
in a liquid scintillation counter, after mixing with 15 ml of liquid
scintillation fluid. An external standard curve was used for the
correction of counting efficiency.
Preparation of Plasma Samples for HPLC.
The plasma samples collected from three rats at different times up to
12 hr after an iv dose of [14C]I
were pooled accordingly. Aliquots (0.15 ml) of pooled samples were
mixed with 0.15 ml of a 5 mM aqueous solution of dibasic potassium
phosphate containing 5 mM TBAS and 0.6 ml of acetonitrile. The samples
were then vortex-mixed, sonicated for 10 min, and centrifuged to remove
precipitated proteins. The precipitated proteins were extracted once
with 0.15 ml of the phosphate buffer containing TBAS and 0.45 ml of
acetonitrile, as described above. This method extracted 84-95% of the
radioactivity from the plasma. The extracts of each sample were
combined, evaporated to dryness in a SpeedVac centrifugal concentrator
(Savant Instruments, Farmingdale, NY), and reconstituted in 0.2 ml of
the phosphate buffer containing TBAS. The reconstituted solutions were
then analyzed for I and metabolites by a radiometric HPLC
method. The plasma samples collected after the po dose or >12 hr after
the iv dose were not analyzed because of the low concentrations of
radioactivity in these samples.
HPLC.
HPLC was carried out with a Waters (Millipore Corp., Milford, MA) HPLC
system equipped with a Berthold (Gaithersburg, MD) model LB 506 C-1
radio-flow detector and an IBM AT personal computer. Two mobile phase
systems were used. System 1 was first used to analyze plasma samples.
This system did not separate the metabolites in bile and urine; they
were eluted as a peak in the void volume. In addition, system 1 contained potassium phosphate and TBAS, which interfered with LC/MS
analysis. System 2 was subsequently developed and used to analyze and
fractionate bile and urine samples. System 1 used a Shodex RSpak
D8-613 C8 reverse-phase column (6 × 150 mm; Waters). The mobile phase consisted of a 50-min linear gradient
from 100% solvent A to 10% solvent A/90% solvent B. Solvent A was 5 mM dibasic potassium phosphate containing 5 mM TBAS (pH 4.5), and
solvent B was acetonitrile. System 2 used a Hamilton PRP-1
C18 reverse-phase column (4.1 × 150 mm).
The composition of the mobile phase was the same as for system 1, except that solvent A was 0.1 M triethylamine adjusted to pH 9 with
acetic acid. The flow rate was 1 ml/min. UV absorbance was monitored at
200 nm. The effluent was mixed with 3 ml of Ecolite(+) liquid scintillation fluid, and the radioactivity was monitored with a
radio-flow detector. The mixture of effluent and Ecolite(+) was then
collected (4 ml/fraction), and the radioactivity in each collected
fraction was determined by liquid scintillation counting. For
metabolite fractionation, the effluent (1 ml/fraction) was collected
without mixing with the liquid scintillation fluid. Radioactivity in
each fraction was determined by liquid scintillation counting after a
portion (50 µl) of the effluent was mixed with 3 ml of the liquid
scintillation fluid. Fractions corresponding to each radioactive peak
were pooled and evaporated to dryness in a SpeedVac centrifugal
concentrator.
LC/MS and LC/MS/MS.
All LC/MS data were obtained with a Sciex API III (PE Sciex, Concord,
Ontario, Canada) dedicated LC/MS mass spectrometer interfaced to a
Waters 600MS gradient HPLC system. Chromatographic separations were
achieved with a Hamilton PRP-1 column (4.1 × 150 mm, 5 µm). The
mobile phase consisted of a 50-min linear gradient from 100% solvent A
to 10% solvent A/90% solvent B. Solvent A was 20 mM ammonium acetate,
and solvent B was acetonitrile. The flow rate was 1 ml/min.
MS analysis of the chromatographic eluent was carried out with negative
ion-spray ionization. The LC column effluent was split 20:1, so that
the flow into the ion-spray interface was 50 µl/min. The mass
spectrometer was scanned from 150 to 950 Da, at a scan rate of 3.5 sec/scan. For LC/MS/MS analysis, argon was used as the collision gas.
The collision energy was 80 eV.
Data Analysis.
Pharmacokinetic parameters were determined by noncompartmental methods
(Gibaldi and Perrier, 1982). The terminal elimination half-life
was estimated by logarithmic-linear regression of the data obtained
between 6 and 12 hr after dosing. The AUC up to the last detectable
time point was determined by trapezoidal integration and was
extrapolated to infinity with individual terminal slope values. Total
clearance was estimated by dividing the dose by AUC0-
.
| |
Results |
Disposition of [14C]I in Jugular
Vein-Cannulated Rats.
In 96 hr after single 15-µmol/kg iv and po doses of
[14C]I to groups of three rats each,
averages of 18.8 and 2.2%, respectively, of the doses were excreted in
urine (table 1). The corresponding averages for the feces were 72.8 and 96.5%, respectively. Total recovery of radioactivity averaged 91.7% of the iv dose and 98.8% of
the po dose. The presence of a large percentage of the iv administered [14C]I radioactivity in feces
suggested that the drug and its metabolites were excreted mainly
via the biliary route. The urinary data were not used for
the estimation of oral absorption because of the small amount of
radioactivity excreted in the urine after the po dose.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Recoveries of the radioactive doses after single 15-µmol/kg iv and po
doses of [14C]I and po doses of the mono-,
di-, and triester prodrugs to rats
|
|
After the iv dose, the concentration of total radioactivity in plasma
was measurable up to 96 hr after dosing and averaged 0.026 µg-eq/ml
at that time. After the po dose, the maximal plasma concentration of
total radioactivity averaged 0.108 µg-eq/ml at 6 hr after dosing
(fig. 2A), indicating
prolonged absorption of I. Based on the ratio of total
radioactivity AUC values after the iv and po doses, the absorption of
I averaged 2.4%.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2.
Concentrations of radioactivity and
I in plasma after single 15-µmol/kg iv and po doses of
[14C]I to rats.
A, Concentrations of radioactivity after iv and po
doses; B, concentrations of radioactivity and
I after an iv dose.
|
|
HPLC radiochromatograms of plasma extracts showed that the parent drug
accounted for more than one half of the radioactivity in plasma up to
12 hr after the iv dose. At least three radioactive metabolites were
detected (fig. 3). Fig. 2B
shows the concentrations of radioactivity and I in plasma up
to 12 hr after the iv dose. Between 6 and 12 hr, I
disappeared from plasma with an apparent elimination half-life of 4 hr.
The total body clearance, volume of distribution at steady state, and
mean residence time were 2.0 ml/min/kg, 0.64 liter/kg, and 5.3 hr,
respectively. The low clearance value indicated that the drug was not
extensively eliminated by metabolism. The volume of distribution at
steady state equaled the volume of body water (0.66 liter/kg),
suggesting that the levels of binding of I to plasma and
tissue proteins were both low (Davis and Morris, 1993; Lin, 1995). The
oral bioavailability of I was not estimated, because of poor
absorption of the compound (2.4%), but was anticipated to be low.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 3.
HPLC profile of radioactivity in a pooled
6-hr plasma sample from rats after single 15-µmol/kg iv doses of
[14C]I (HPLC system 1).
|
|
Disposition of 14C-Labeled I,
II, III, and IV in BDC Rats.
After single 15-µmol/kg iv doses of
[14C]I to three BDC rats, the
administered radioactivity was slowly excreted in bile (fig.
4). In 12 hr, averages of 41.3 and 14.1%
of the doses were excreted in bile and urine, respectively (table 1).
Total recovery of the radioactive dose averaged 95.1%. Between 0 and
12 hr, I accounted for 50% of the radioactivity in bile. At
12 hr after dosing, liver and kidneys demonstrated higher
concentrations of radioactivity than did plasma. Other tissues,
including heart, lungs, bone, and gastrointestinal tract, exhibited
lower concentrations of radioactivity than did plasma. An average of
31.3% of the dose was recovered in liver at 12 hr after dosing,
indicating that I and/or its metabolites were extensively
extracted by the liver.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 4.
Cumulative percentages of the dose excreted
in bile as total radioactivity, I, and metabolites after
single 15-µmol/kg iv doses of [14C]I to
rats.
|
|
In 12 hr after single 15-µmol/kg po doses of
14C-labeled I, II,
III, and IV, averages of 0.6, 3.2, 43.6, and
29.0%, respectively, of the doses were excreted in bile. The
corresponding values in urine were 0.5, 0.2, 33.4, and 12.5%, respectively (table 1). Based on the sum of radioactivity recovered in
bile, urine, liver, and carcass minus the radioactivity in feces and
the gastrointestinal tract at 12 hr after dosing, averages of 3, 6, 80, and 45% of 14C-labeled I,
II, III, and IV, respectively, were
absorbed.
Biotransformation Profiles in Bile and Urine.
The profiles of radioactivity in bile and urine samples collected
between 0 and 12 hr after administration of the iv dose of
[14C]I and the po doses of
[14C]II,
[14C]III, and
[14C]IV to BDC rats were determined
by radiometric HPLC using system 2 (figs.
5 and 6).
The samples obtained after administration of the po dose of
[14C]I were not analyzed because of
poor absorption of the drug. After iv dosing of
[14C]I (the acid), unchanged
I accounted for 51% of the radioactivity (21% of the dose)
in bile. At least three radioactive metabolites were found in bile
(fig. 5A). No parent drug but at least three radioactive
metabolites were found in urine (fig. 6A). After po dosing
of [14C]II (monoester), the parent
monoester accounted for 28% of the radioactivity in bile (0.9% of the
dose); the remaining radioactivity was attributed to two metabolites
(fig. 5B). Less than 0.3% of the dose was excreted in
urine, predominantly as one metabolite (fig. 6B). After po
dosing of [14C]III (diester),
III accounted for 20% of the radioactivity in bile (8.7%
of the dose). The remaining radioactivity was attributed to at least
five metabolites (fig. 5C). Urine contained no parent drug
but did contain two metabolites (fig. 6C). After po dosing with [14C]IV (triester), the
profiles of radioactivity in bile and urine were virtually identical to
those obtained after dosing with
[14C]III (figs. 5D and
6D). A trace amount (<0.1%) of IV was detected
in bile. III accounted for 12% of the radioactivity in bile
(3.6% of the dose). These results suggested that the triester was
unstable and was readily hydrolyzed to III, either before or
after absorption. No active acid I was found in the bile and
urine samples collected after dosing with II, III, or IV, suggesting a lack of bioactivation of the prodrugs.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 5.
Representative HPLC profiles of
radioactivity in 0-12-hr pooled bile samples after single 15-µmol/kg
iv doses of [14C]I (A) and po
doses of [14C]II (B),
[14C]III (C), and
[14C]IV (D) to BDC rats (HPLC
system 2).
|
|
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 6.
Representative HPLC profiles of
radioactivity in 0-12-hr pooled urine samples after single
15-µmol/kg iv doses of [14C]I
(A) and po doses of [14C]II
(B), [14C]III
(C), and [14C]IV
(D) to BDC rats (HPLC system 2).
|
|
LC/MS/MS Analysis of Metabolites in Bile and Urine.
The metabolites of I, III, and IV in
urine and of III and IV in bile were fractionated by HPLC using system 2. The isolated metabolites were analyzed by LC/MS
and LC/MS/MS. Product-ion spectra obtained for both standards and
metabolites yielded fragmentation characteristic of only the phosphonate, phosphonate diester, and sulfonate moieties. Proposed structures for metabolites were based on analogies with
biotransformation pathways for compounds with similar structures
(Gonzalez-Pacanowska et al., 1988). The structures of two
metabolites each for I (I-M1 and
I-M2), III (III-M1 and
III-M2), and IV (III-M1 and
III-M2) were proposed. The metabolites of II were
not identified because of lack of sufficient material.
Using negative ion-spray LC/MS, I standard showed an
(MH)
ion at m/z 407. The
product-ion spectrum for the m/z 407 ion of I
showed only one fragment ion, at m/z 79, which was characteristic of the metaphosphate anion
(PO3) (fig.
7A). Compound I-M1
yielded an (MH) ion at m/z 369. The indicated molecular mass of 370 Da was 38 Da lower than that
of I (408 Da), suggesting a loss of C5H10 from I and
an addition of two oxygens. The product-ion spectrum of the
m/z 369 ion of I-M1 showed the characteristic fragment ion at m/z 79, corresponding to the metaphosphate
anion. Based on this information, the structure of I-M1 was
proposed as 2,6-dimethyl-11-phosphono-11-sulfo-2,6-undecadienoic acid
(fig. 7B). Compound I-M2 yielded an
(MH) ion at m/z 397. The indicated
molecular mass of 398 Da was 10 Da lower than that of I,
suggesting a loss of C3H6
from I and an addition of two oxygens. The product-ion
spectrum of the m/z 397 ion of I-M2 showed the
characteristic fragment ion of I at m/z 79. Thus,
the structure of I-M2 was proposed as
4,8-dimethyl-13-phosphono-13-sulfo-4,8-tridecadienoic acid (fig.
7C).
The product-ion spectrum of III standard showed fragment
ions at m/z 243, 230, 153, 125, and 80 (fig.
8A). Cleavage of the
carbon-carbon bonds between C2 and C3 and between C1 and C2 yielded the
fragment ions at m/z 243 and 230, respectively. An ion at
m/z 153 was thought to arise from the transfer of oxygen from the sulfonate moiety to the phosphonate diester moiety, yielding the
[PO4(C2H5)2]
anion. Subsequent loss of
C2H4 yielded the ion at
m/z 125. The fragment ion at m/z 80 was
characteristic of the monothionate anion
(SO3). Compound
III-M1 yielded an (MH) ion at
m/z 359. The indicated molecular mass of 360 Da was 104 Da
lower than that of III standard (464 Da), suggesting a loss
of C10H16 from
III and an addition of two oxygens. The product-ion spectrum
of the m/z 359 ion of III-M1 showed the
characteristic fragment ions of the phosphonate diester moiety at
m/z 153 and 125 and the monothionate anion at m/z
80. Based on this information, we proposed the structure of
III-M1 as 2-methyl-7-diethoxyphosphinyl-7-sulfo-2-heptenoic
acid (fig. 8B). Compound III-M2 yielded an
(MH) ion at m/z 385. The indicated
molecular mass of 386 Da was 78 Da lower than that of III
standard, suggesting a loss of C8H14 from III
and an addition of two oxygens. The product-ion spectrum of the
m/z 385 ion of III-M2 showed the three characteristic fragment ions of the phosphonate diester moiety and the
monothionate anion, at m/z 153, 125, and 80. Thus, the structure of III-M2 was proposed as
4-methyl-9-diethoxyphosphinyl-9-sulfo-4-nonenoic acid (fig.
8C). III-M1 and III-M2 were also
observed in the bile and urine samples obtained after a po dose of the triester IV to rats.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 8.
Ion-spray product-ion spectra of
III standard (A), metabolite
III-M1 (B), and metabolite
III-M2 (C).
|
|
| |
Discussion |
I, an 
-phosphonosulfonate, is an analog of
a prototypical 1,1-bisphosphonate squalene synthase inhibitor (Ciosek
et al., 1993) in which one of the phosphonate groups has
been replaced by a sulfonate group. The disposition of the prototypical
bisphosphonate inhibitor in rats was studied (Lan S-J,
unpublished data), and the results showed that it had a disposition
profile similar to that of other bisphosphonates (Michael
et al., 1972; Mizuno et al., 1989; Lin et
al., 1991, 1992; Nakamura et al., 1994; Usui et
al., 1995). In general, the compounds were poorly absorbed upon
oral administration, extensively retained in bone, primarily excreted in urine, and essentially resistant to metabolism of the
bisphosphonate moiety. No information regarding the disposition of
-phosphonosulfonates is available in the literature. It is of
interest to compare the disposition of I with those of
bisphosphonates.
Unlike the bisphosphonates, which were rapidly cleared from plasma
after iv dosing, either taken up and sequestered in bone or excreted by
the kidneys (Lin et al., 1991; Usui et al.,
1995), iv administered I was slowly cleared from plasma,
primarily taken up by the liver, and excreted in bile unchanged or as
metabolites. The lack of accumulation of radioactivity in bone after an
iv dose of [14C]I was an improved
property of I, compared with its bisphosphonate analog (Lan
S-J, unpublished observations). Bisphosphonates are strongly retained
in bone through binding to hydroxyapatite (Fleisch, 1988).
Accumulation of a monophosphonate in bone has also been reported (Bopp
et al., 1977). It is interesting to observe that
I (also a phosphonate) was not retained in bone. The reason
for this difference is unknown. Unlike its bisphosphonate analog and
other bisphosphonates, which were excreted mainly in urine,
I and its metabolites were excreted mainly in bile, probably
because of the higher molecular mass of I (408 Da) (Williams
et al., 1965).
The recovery of a large percentage (31%) of the dose in the liver at
12 hr after iv administration (table 1) and the excretion of one half
of the radioactivity (21% of the dose) in bile as the unchanged drug
demonstrated that the liver (the target organ) was well exposed to the
drug. A preliminary toxicological evaluation of I in mice
showed minimal effects on aspartate transaminase and alanine
transaminase levels after a 150-mg/kg iv dose (Magnin et
al., 1996), indicating that the accumulation of I and metabolites in liver had little or no adverse effect on the liver. The
excretion of a large percentage of unchanged drug in bile and the low
total body clearance (2 ml/min/kg) also suggested that I was
not extensively eliminated through metabolism.
I was a triacid bearing charges intermediate between those
of the diacid monophosphonate and the tetraacid bisphosphonate. Both
mono- and bisphosphonates were poorly absorbed (Michael et al., 1972; Bopp et al., 1977; Mizuno et al.,
1989; Lin et al., 1991, 1992; Nakamura et al.,
1994) because of their highly charged nature. As expected, I
was poorly absorbed after oral administration to rats (3%). The
results of this study showed that masking one acidic function of the
phosphonate group improved the absorption slightly (6% absorption for
the monoester II). Masking both acidic functions of the
phosphonate group increased the absorption substantially (80%
absorption for the diester III). Similar extents of
absorption have also been reported for the mono- and diester prodrugs
of phosphonates (Scarfinoska et al., 1995; De Lombaert
et al., 1994). The smaller extent of absorption of the triester IV (45%), compared with the diester III (80%), might reflect the poor aqueous solubility of the triester.
The cyclohexyl sulfonate ester of IV was chemically unstable
under physiological conditions. Incubation (at 37°C, for 4 hr) of
IV with boiled plasma or boiled homogenates of liver or
intestines of rats resulted in extensive conversion of the triester
IV to the diester III, presumably by solvolysis
(data not shown). Thus, the orally administered triester IV might have been partially hydrolyzed to the diester III before its absorption. Hydrolysis of phosphonate esters has been demonstrated in vitro and in vivo
(Donninger et al., 1972; Menn and McBain, 1974; Appleton and
Nakatsugawa, 1977; De Lombaert et al., 1994;
Scarfinoska et al., 1995; Slatter et al., 1996),
and bioactivation of diaryl phosphonate esters to the active acids has
been reported. However, dialkyl phosphonate esters were found to
be hydrolyzed only to the monoesters, which were not further
hydrolyzed. An NADPH-dependent mixed-function oxidase, rather than
esterases, was believed to be responsible for the hydrolysis of alkyl
phosphonate esters. Our studies showed that the parent drug accounted
for one half of the dose excreted in bile after an iv dose of
I. The systemic bioactivation of prodrugs would have
resulted in the excretion of I in bile. However, no
I or I-derived metabolites were detected in bile
or urine after oral administration of the prodrugs. Thus, systemic
bioactivation of the prodrugs in rats was limited, if it occurred. This
observation was consistent with the results reported in the literature.
The carbon-phosphorus bond of phosphonates and the carbon-sulfur bond
of sulfonates are generally stable in mammals, despite the fact that
cleavage of the carbon-phosphorus bond of phosphonates was reported in
certain microorganisms (Smith, 1983) and (in one case) in rats
(Horigane et al., 1979). No metabolites derived from
cleavage of the phosphonate or sulfonate groups of I or the
prodrugs were identified in this study. The possibility of such
cleavage reactions could not be ruled out, however, because only
approximately one half of the metabolites in bile have been identified.
Metabolism of I and the prodrugs occurred primarily at the
farnesyl moiety, possibly by the same pathways as for FPP (Gonzalez-Pacanowska et al., 1988). A cytochrome P450 enzyme
was believed to catalyze the hydroxylation of the C15-methyl group (Licht and Coscia, 1978; Ichihara et al., 1981). The
resulting 
-alcohol was then oxidized by a dehydrogenase,
monooxygenase, or oxidase, either alone or in combination, to form the
-acid (Lee, 1979; Thyagarajan et al., 1979; Lamboeuf
et al., 1981; Ichihara et al., 1986). Subsequent
loss of three carbons from the -acid of I yielded
metabolite I-M2. The enzyme that catalyzed this
chain-shortening reaction and the reaction mechanism remain unknown.
Alternatively, the C13-C14 double bond of I could undergo
direct oxidative cleavage, resulting in a chain-shortened -alcohol
or -aldehyde that could be further oxidized to metabolite I-M2. Metabolite I-M1 was probably derived from I-M2 through 
-oxidation. The structures of the proposed metabolites of III indicated that III was
metabolized by the same pathways as I.
In summary, after iv administration of I to rats, the drug
was slowly cleared from plasma and was primarily taken up by the liver.
I was slowly excreted in bile, either unchanged or as
metabolites. No accumulation of I and/or metabolites in the
analyzed tissues, other than the liver, was observed. I was
poorly absorbed. Substantially improved absorption was observed for the
diester prodrug III. However, the prodrugs were not
bioactivated to the active acid in rats. The metabolism of I
and the prodrugs occurred primarily at the farnesyl moiety. Based on
this information, two modifications of I were subsequently
made, i.e. 1) the farnesyl moiety was replaced by a
biphenylether surrogate, to enhance metabolic stability, and 2) a
bioreversible bis(pivaloyloxymethyl) ester prodrug was prepared, to
improve oral absorption. The modified compound was found to be a
potent, oral, cholesterol-lowering agent in animal models (Dickson
et al., 1996) and was thus selected for development. The
disposition and pharmacokinetics of the modified compound will be the
subject of a separate report.
Abbreviations used are:
FPP, farnesyl-1-pyrophosphate;
BDC, bile duct-cannulated;
TBAS, tetrabutyl
ammonium hydrogen sulfate.
![[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}
-Phosphonosulfonate Squalene
Synthase Inhibitors
Top
Abstract
Introduction
