Download Comparison of the Circulating Metabolite Profile of PF

Supplemental material to this article can be found at:
http://dmd.aspetjournals.org/content/suppl/2014/11/10/dmd.114.061218.DC1
1521-009X/43/2/190–198$25.00
DRUG METABOLISM AND DISPOSITION
Copyright ª 2014 by The American Society for Pharmacology and Experimental Therapeutics
http://dx.doi.org/10.1124/dmd.114.061218
Drug Metab Dispos 43:190–198, February 2015
Comparison of the Circulating Metabolite Profile of PF-04991532,
a Hepatoselective Glucokinase Activator, Across Preclinical Species
and Humans: Potential Implications in Metabolites in Safety
Testing Assessment s
Raman Sharma, John Litchfield, Arthur Bergman, Karen Atkinson, David Kazierad,
Stephanie M. Gustavson, Li Di, Jeffrey A. Pfefferkorn, and Amit S. Kalgutkar
Pfizer Inc., Groton, Connecticut (R.S., A.B., K.A., S.M.G., L.D.); and Pfizer Inc., Cambridge, Massachusetts (J.L., D.K., J.A.P., A.S.K.)
Received November 7, 2014; accepted November 10, 2014
ABSTRACT
[14C]PF-04991532 was attributed to an unchanged parent (>70% in
rats and dogs). In contrast with the human circulatory metabolite
profile, the monohydroxylated metabolites were not detected in
circulation in either rats or dogs. Available mass spectral evidence
suggested that M2a and M2b/M2c were diastereomers derived from
cyclopentyl ring oxidation in PF-04991532. Because cyclopentyl
ring hydroxylation on the C-2 and C-3 positions can generate eight
possible diastereomers, it was possible that additional diastereomers may have also formed and would need to be resolved from
the M2a and M2b/M2c peaks observed in the current chromatography conditions. In conclusion, the human metabolite scouting
study in tandem with the animal mass balance study allowed early
identification of PF-04991532 oxidative metabolites, which were not
predicted by in vitro methods and may require additional scrutiny in
the development phase of PF-04991532.
Introduction
pancreas, which resulted in increased insulin secretion at inappropriately
low glucose levels (Grimsby et al., 2003; Coghlan and Leighton,
2008; Haynes et al., 2010; Meininger et al., 2011; Sarabu et al.,
2012; Pfefferkorn, 2013). To mitigate this hypoglycemia risk,
several approaches have been reported to limit GK activation to
the liver (Pfefferkorn, 2013).
PF-04991532 was designed to have low passive permeability,
thereby minimizing distribution into extrahepatic tissues, particularly
the pancreas (Ghosh et al., 2014). Concurrently, PF-04991532 was
also optimized as a substrate for active liver uptake via liver-specific
organic anion transporting polypeptide isoforms. In preclinical animal
models of diabetes, PF-04991532 was found to robustly lower fasting
and postprandial glucose and was devoid of hypoglycemic risks,
leading to its selection as a clinical candidate for treating T2DM. The
in vivo pharmacokinetics of PF-04991532 were characterized by high
systemic plasma clearance (CLp) values of 51 and 35 ml/min per kg in
rats and dogs, respectively, and oral bioavailability of ;18% in both
rats and dogs (Pfefferkorn et al., 2012). Resistance of PF-04991532 to
metabolic turnover in liver microsomes and hepatocytes from a rat,
dog, and human suggested that metabolic elimination would be
Our laboratory recently identified a glucokinase (GK) activator,
PF-04991532 [(S)-6-(3-cyclopentyl-2-(4-trifluoromethyl)-1H-imidazol1-yl)propanamido)nicotinic acid] (Fig. 1), with preferential GK
activation in hepatocytes versus pancreatic b-cells (Pfefferkorn et al.,
2012). GK is a glycolytic enzyme that catalyzes the phosphorylation
of glucose to glucose-6-phosphate (Mithieux, 1996). GK acts as
a glucostat in the b-cells of the pancreas, establishing the threshold
for glucose-stimulated insulin secretion (Matschinsky et al., 1998).
In the liver, GK is involved in the regulation of hepatic glucose
utilization and output (Matschinsky and Ellerman, 1968). The
therapeutic benefits of GK activators in the treatment of type 2
diabetes mellitus (T2DM) have been reviewed (Matschinsky et al.,
2011; Sarabu et al., 2011). Early clinical experience with GK activators
indicated significant improvements in glycemic control but with an
increased risk of hypoglycemia due to excessive GK activation in the
dx.doi.org/10.1124/dmd.114.061218.
s This article has supplemental material available at dmd.aspetjournals.org.
ABBREVIATIONS: AUC, area under the plasma concentration time curve; CID, collision-induced dissociation; GK, glucokinase; HPLC, highperformance liquid chromatography; LC-MS/MS, liquid chromatography–tandem mass spectrometry; MAD, multiple ascending dose; MH+,
protonated molecular ion; MRP, multidrug resistant protein; NOAEL, no observed adverse effect level; PF-04991532, (S)-6-(3-cyclopentyl-2-(4trifluoromethyl)-1H-imidazol-1-yl)propanamido)nicotinic acid; T2DM, type 2 diabetes mellitus; UPLC, ultra-performance liquid chromatography.
190
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
A previous report from our laboratory disclosed the identification of
PF-04991532 [(S)-6-(3-cyclopentyl-2-(4-trifluoromethyl)-1H-imidazol1-yl)propanamido)nicotinic acid] as a hepatoselective glucokinase
activator for the treatment of type 2 diabetes mellitus. Lack of in vitro
metabolic turnover in microsomes and hepatocytes from preclinical
species and humans suggested that metabolism would be inconsequential as a clearance mechanism of PF-04991532 in vivo. Qualitative
examination of human circulating metabolites using plasma samples
from a 14-day multiple ascending dose clinical study, however,
revealed a glucuronide (M1) and monohydroxylation products
(M2a and M2b/M2c) whose abundances (based on UV integration)
were greater than 10% of the total drug-related material. Based
on this preliminary observation, mass balance/excretion studies
were triggered in animals, which revealed that the majority of
circulating radioactivity following the oral administration of
Circulating Metabolites of PF-04991532
Fig. 1. Structure of hepatoselective glucokinase activator PF-04991532.
Materials and Methods
General Chemicals. Commercially obtained chemicals and solvents were of
high-performance liquid chromatography (HPLC) or analytic grade. [14C]PF04991532 (specific activity, 0.25 mCi/mg) was synthesized by the radiochemistry group at Pfizer Worldwide Research and Development (Groton, CT). It
had a radiochemical purity of . 99%, as determined by HPLC using an in-line
radioactivity detector. The location of the 14C label on the nicotinic acid motif
is illustrated in Fig. 1. 6-Aminonicotinic acid was purchased from SigmaAldrich (St. Louis, MO). A cryopreserved Sprague-Dawley rat (male, pool of
12 donors), beagle dog (male, pool of five donors), and human (pool of
10 donors of mixed gender) were purchased from Celsis (Chicago, IL).
Metabolite Scouting Study in Humans. The study was conducted in
compliance with the International Conference on Harmonization Good Clinical
Practices guidelines, the ethical principles that have their origin in the
Declaration of Helsinki, and the US Food and Drug Administration regulations
for informed consent and protection of subject rights. The clinical study design
with PF-04991532 incorporated a randomized, placebo group, MAD study in
which subjects with T2DM on background metformin therapy received either
placebo or oral doses of PF-04991532 (30–900 mg) administered once daily for
14 days with food (Gustavson et al., 2013). Venous blood samples from each
subject predose (10 ml) and at 2.0, 6.0, and 12 hours postdose (2 ml/time point)
on days 1 and 14 were used for metabolite scouting efforts. Blood samples were
collected in heparinized tubes. Within 30 minutes after collection, the blood
samples were centrifuged at approximately 1700g for ;10 minutes at 4C to
generate plasma. Citric acid monohydrate (500 mg/ml) was added to the plasma
samples to create a plasma:citric acid (92:8, w/v) mixture. The final
concentration of citric acid monohydrate in the sample was 40 mg/ml. All
samples were stored at or below 220C until analysis. Samples from the
highest dose group (900 mg) in the multiple dose study were used for
metabolite scouting. Plasma from each of the nine subjects (five male/four
female, 45–65 years of age) that received PF-04991532 (900 mg) was pooled
individually according to the Hamilton method (Hamilton et al., 1981). An
equal volume of 0.3 ml from each individual patient pool was then combined to
make a multisubject pool. This was done for both day 1 and day 14 samples,
resulting in two 2.7-ml multisubject pools. Additionally, 0.3 ml from the t =
0 time point for each subject was pooled to make a 2.7-ml predose pool. Each
pool was diluted with 12 ml of ice-cold acetonitrile (containing 0.1% formic
acid) and sonicated for 5 minutes. The samples were then centrifuged (2000g)
for 10 minutes. The supernatants were transferred to clean test tubes and
evaporated to dryness in vacuo. The dried residue was then reconstituted in
300 ml of the mobile phase (10% acetonitrile, 0.1% formic acid), and an aliquot
(50 ml) was injected on a liquid chromatography–tandem mass spectrometry
(LC-MS/MS) system.
Metabolite Identification in Hepatocytes. Stock solutions of unlabeled
PF-04991532 were prepared in 10% dimethyl sulfoxide and 90% acetonitrile.
The final concentration of dimethyl sulfoxide and acetonitrile in the incubation
mixtures were 0.1 and 0.9%, respectively. Williams E media was prepared by
adding 26 mM sodium carbonate and 50 mM HEPES, followed by 0.2-mm
filtration and then 30 min of CO2 bubbling at 37C. This media was used for
thawing and suspension of hepatocytes. Upon thawing, the pooled cryopreserved hepatocytes were resuspended in Williams E medium (WEM GIBCOBRL, custom formula #91-5233EC; Life Technologies, Grand Island, NY)
supplemented with 50 mM HEPES and 26 mM Na2CO3. The cells were
counted using the Trypan Blue exclusion method, and a 24-well hepatocyte
plate containing 0.5 million cells/ml was treated with PF-04991532 (final
concentration in 1.0-ml incubation = 10 mM) and incubated at 37C for
4 hours, 75% relative humidity, 5% CO2, and 95% O2. Incubations were
terminated by the addition of ice-cold acetonitrile (4 ml) and centrifuged
(3000g, 15 minutes). The supernatants were dried under a steady stream of
nitrogen, reconstituted with 25% aqueous acetonitrile (250 ml), and analyzed
via LC-MS/MS for metabolite formation.
For metabolite identification via the relay method, the hepatocyte plate was
incubated at 37C with 95% O2/5% CO2, 75% relative humidity for 4 hours
daily for 5 days. At time t = 0, an aliquot of the hepatocyte incubation
(suspension) was quenched with 2 volume of acetonitrile and frozen at 280C
for liquid chromatography–mass spectrometry analysis. At 4, 8, 12, 16, and
20 hours, the hepatocyte incubation suspension plate was centrifuged (3000g,
10 minutes, room temperature) and 300-ml supernatant was transferred to
a clean 24-well plate and stored at 280C until the next relay experiment. The
remaining hepatocyte pellet in the incubation plate was quenched with 2
volume of acetonitrile and frozen at 280C for LC/MS analysis. For the second
relay experiment, the supernatant plate was warmed to 37C for 30 minutes,
and hepatocytes were added to the samples to give a final cell density of
0.5 million cells/ml. The plates were incubated at 37C for 4 hours, sampled,
and processed as described above. Five relays were performed to give a total
incubation time of 20 hours.
Animals, Dosing, and Sample Collection. Animal care and in vivo
procedures were conducted according to the guidelines of the Pfizer Animal
Care and Use Committee. The in-life portion of the studies was conducted at
Covance Laboratories Inc. (Madison, WI) in accordance with applicable
standard operating procedures and the Wisconsin Department of Health
Services Radiation Protection Section (License No. 025-1076-01). Jugular vein
cannulated Sprague-Dawley rats (192–328 g) were purchased from Charles
River Laboratories, Inc. (Wilmington, MA). Purebred beagle dogs (8.5–10.3 kg)
were obtained from Ridglan Farms, Inc. (Mt. Horeb, WI). Animals were
quarantined for a minimum of 3 days before treatment and maintained on a
12-hour light/dark cycle. The animals were fasted overnight before administration
of [14C]PF-04991532 and fed 4 hours postdose. The animals were provided water
ad libitum. The doses selected for the mass balance study reflected the no observed
adverse effect level (NOAEL) in the respective preclinical species.
Rat Study. A group of Sprague-Dawley rats (n = 3/gender) was
administered a single 200-mg/kg oral dose of [ 14 C]PF-04991532 for the
mass balance study. The dose was formulated as a suspension in 0.5%
methylcellulose in reverse osmosis water on the day before dose administration.
Each rat received an approximate dose of 40 mCi/200 g of radiolabeled
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
inconsequential as a clearance mechanism in vivo. Preliminary studies
addressing the in vivo clearance mechanism in rats using unlabeled
PF-04991532 revealed biliary (62%) and renal (11%) excretion of the
unchanged parent as the principal elimination pathways (Pfefferkorn
et al., 2012). At the time of candidate selection, it was presumed that
PF-04991532 would be eliminated in an unchanged form via the
hepatobiliary route involving active uptake by organic anion transporting polypeptide 1B3 into the liver, followed by biliary elimination
through efflux transporters, as evident with related carboxylic acid
drugs, such as fexofenadine (Matsushima et al., 2008).
Contrary to our hypothesis, plasma samples from the clinical
multiple ascending dose (MAD) studies on PF-04991532 revealed
several metabolites in circulation whose abundance was greater than
10% of the total drug-related material and would potentially require
further qualification in preclinical species as per the metabolites in
safety testing guidance from the US Food and Drug Administration
and the later published 2009 International Conference on Harmonisation guidance M3 (R2) (U.S. Food and Drug Administration,
2008; European Medicines Agency, 2009; International Conference
on Harmonisation, 2012). Consequently, metabolism/excretion studies
with [14C]PF-04991532 were triggered in Sprague-Dawley rats and
beagle dogs to examine circulating and excretory metabolites in the
preclinical toxicology species selected for PF-04991532 development.
The collective findings from these studies are summarized herein.
191
192
Sharm et al.
evaporative centrifuge (Genevac Inc., Valley Cottage, NY). The mean recovery
of radioactivity after extraction was ;99%. The samples were reconstituted
with 1 ml of 5 mM ammonium acetate (pH = 3.0):acetonitrile (95:5). An aliquot
(;30 ml) was injected onto the HPLC system for analysis. Fecal homogenates
were combined such that at least 95% of the radioactivity excreted in feces was
represented. Pooled fecal samples were diluted with acetonitrile (3 ml/g
homogenate) and sonicated (15 minutes). After centrifugation at 1700g for
10 minutes, the supernatant was transferred to a 100-ml Erlenmeyer flask, and
the remaining residue was further extracted with acetonitrile (10 ml). The
supernatants were combined, and the mean recovery of radioactivity after
extraction was ;90–99% in all species. The supernatants were concentrated to
dryness at 37C, and the residues were reconstituted in 500-ml 5 mM
ammonium acetate (pH = 3.0):acetonitrile (95:5). Aliquots (;30 ml) were
injected onto the HPLC system for analysis.
Quantitative Assessment of Metabolites. Metabolite quantification from
urine, feces, and plasma was performed by measuring the radioactivity fractions
collected from HPLC effluent at 0.25-minute intervals using an FC 204 fraction
collector (Gilson Inc., Middleton, WI). Fractions were collected in a 96-well
microbeta plate and dried using the Genevac EZ-2 evaporative centrifuge
(Genevac Inc.). Prior to counting, 175 ml of Ultima Gold liquid scintillatant
(PerkinElmer Life and Analytical Sciences) was added to each well. Radioactivity
was then counted using a Trilux 1450 microbeta counter (PerkinElmer Wallac,
Gaithersburg, MD). Data were imported into Laura program version 3.1.1.39
(LabLogic System Ltd., Sheffield, UK) to provide an integrated peak representation
in counts per minute and the percentage of total radioactivity comprised by each
peak within the radiochromatogram.
HPLC Analysis. The HPLC system consisted of an Acela quaternary
solvent delivery pump, an Acela autoinjector, and a Surveyor PDA Plus
photodiode array detector (Thermo Electron North America, West Palm Beach,
FL). Chromatography was performed on a Phenomenex hydro RP column
(4.6 150 mm, 5 m) with a mobile phase comprised of 5 mM ammonium formate,
pH = 3.0 (solvent A), and acetonitrile (solvent B). For metabolic profiling with
unlabeled PF-04991532, the mobile phase gradient was programmed in a linear
fashion as follows: 95% solvent A (0–3 minutes), 40% solvent A (25 minutes),
20% solvent A (35 minutes), and 5% solvent A (38 min), followed by column reequilibration at 95% solvent A (38.01–49 minutes). For animal mass balance
studies with [14C]PF-04991532, the mobile phase gradient was programmed in
a linear fashion as follows: 95% solvent A (0–3 minutes), 70% solvent A
(37 minutes), and 5% solvent A (43–49 minutes), followed by column reequilibration at 95% solvent A (49.01–56 minutes). A flow rate of 1.0 ml/min
was maintained throughout the analysis. Approximately 5% of the postcolumn
flow was split to the mass spectrometer. The remainder of the effluent was split to
the FC 204 fraction collector or waste.
LC-MS/MS. Identification of the metabolites in excreta, plasma, and
hepatocytes (4-hour incubation with PF-04991532) was performed on a Thermo
Orbitrap mass spectrometer operating in a positive ion electrospray mode.
Xcalibur software version 2.0 was used to control the HPLC/MS system. The
electrospray ionization tune settings were as follows: capillary temperature =
325C; sheath gas flow = 6 units; auxiliary gas flow = 5 units; sweep gas flow =
2 units; source voltage = 4 kV; source current = 100 uA; capillary voltage = 29 V;
and tube lens = 90 V. Full scan data were collected at 15,000 resolution. Datadependent product ion scans of the two most intense ions found in the full scan
were obtained at 15,000 resolution. The dynamic exclusion function was used
with a 1-minute exclusion duration after three successive product ion scans, with
early exclusion if the precursor ion fell below a signal-to-noise ratio of 20.
Structural information was generated from the collision-induced dissociation
(CID) spectra of the protonated molecular ions (MH+).
Chromatographic separation of oxidative metabolites M2a, M2b, and M2c
in hepatocyte incubations using the relay method was achieved using an
ACQUITY ultra-performance liquid chromatography (UPLC) system (Waters
Corporation, Milford, MA), with an ACQUITY UPLC BEH (Ethylene Bridged
Hybrid) C18 column 1.7 mM, 2.1 150 mm, maintained at a column
temperature of 40C. The mobile phase consisted of water with 0.1% formic
acid (A) and acetonitrile (B). The flow rate was 0.35 ml/min, and the linear
gradient was as follows: 5% solvent B (0–2 minutes), 35% solvent B
(57 minutes), and 5% solvent B (58–59 minutes) followed by a column
equilibration at starting mobile phase conditions for 5 minutes. The injection
volume was 10 ml. Detection of PF-4991532 and metabolites was performed on
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
material. Urine and feces was collected from intact animals predose (212 to 0)
and at 0–8 and 8–24, 24–48, 48–72, 72–96, 96–120, 120–144, and 144–168
hours postdose. The volume of urine and feces samples was recorded, and
samples were stored at –20C until analysis. For the determination of
pharmacokinetic parameters of total radioactivity and identification of
circulating metabolites, a third group of jugular vein cannulated rats (n = 5/
gender) was given an oral dose of 200 mg/kg [14C]PF-04991532. Blood was
collected at predose and at 0.25, 0.5, 1, 2, 4, 6, 8, 12, 24, and 48 hours postdose
into tubes containing K2EDTA. The blood samples were centrifuged at 1000g
for 10 minutes to obtain plasma. Citric acid monohydrate (500 mg/ml) was
added to the plasma samples to create a plasma:citric acid (92:8, w/v) mixture.
The final concentration of citric acid monohydrate in the sample was 40 mg/ml.
Plasma was transferred to clean tubes and stored at 220C until analysis.
Dog Study. Two male and two female beagle dogs were administered
a single oral 80-mg/kg dose of [14C]PF-04991532 in 0.5% methylcellulose in
reverse osmosis water for the mass balance study. Each dog received
approximately 20 mCi/kg of radiolabeled material. Blood samples were
collected by venipuncture of a jugular vein before dosing and at 0.5, 1, 2, 4, 6,
8, 12, 24, 48, 72, and 96 hours postdose. Blood was collected into tubes
containing K2EDTA anticoagulant and was centrifuged within 1 hour of
collection at 1300g for ;10 minutes to obtain plasma. Citric acid monohydrate
(500 mg/ml) was added to the plasma samples to create a plasma:citric acid
(92:8, w/v) mixture. The final concentration of citric acid monohydrate in the
sample was 40 mg/ml. Urine and feces and wash/cage wipes were quantitatively
collected for at least 18 hours before dosing and then at 0–6, 6–12–24, and over
24-hour intervals through 168 hours postdose.
Determination of Radioactivity. Total radioactivity determination in rat
and dog urine, feces, and plasma was conducted at Covance Laboratories Inc.
(Madison, WI). The radioactivity in urine and plasma was determined by liquid
scintillation counting. Duplicate aliquots of urine (0.1–0.4 g) and plasma (100–
400 ml) were mixed directly with 10 ml of Hionic Fluor scintillation fluid for
radioactivity measurement by liquid scintillation counting using a Packard
2500 liquid scintillation counter (PerkinElmer Life and Analytical Sciences,
Boston, MA). Fecal samples were weighed and homogenized with deionized
water (2–3 times w/v, feces/water). Duplicate weighed aliquots of fecal
homogenates (;300–400 mg) were combusted using a PerkinElmer Model 307
sampler oxidizer. Radioactivity in the combustion products was determined by
trapping the liberated 14CO2 mixed with liquid scintillation fluid for
radioactivity measurement. Acceptance criteria were combustion recoveries
of 95–105%. Pharmacokinetic parameters for total radioactivity were determined using noncompartmental analysis. The amount of radioactivity in
plasma at each time point was calculated using the specific activity of the dose
administered and was expressed as nanogram equivalents per milliliter. Cmax of
total radioactivity in plasma were estimated directly from the experimental data,
with Tmax defined as the time of first occurrence of Cmax. Terminal elimination
rate constants (kel) were estimated by linear regression of the log-linear plasma
concentration time curve. The half-life (t1/2) was calculated as 0.693/kel. Area
under the plasma concentration time curve (AUC) from time t to infinity
[AUC(t–‘)] was extrapolated from AUC(0–t) adding Clast/kel, where Clast is the
observed serum concentration at the last quantifiable time point estimated from
the log-linear regression analysis. AUC from time zero to infinity [AUC(0–‘)]
was estimated as the sum of [AUC(0–t)] and [AUC(t–‘)] values.
Extraction of Metabolites from Biologic Matrices. Pooling of plasma
samples from rats and dogs was performed for each individual animal
according to the method of Hamilton (Hamilton et al., 1981), such that each
sample was representative of total exposures (AUC) to metabolites relative to
each other. Pooled plasma samples were diluted with 10 ml of acetonitrile while
vortexing vigorously for 15–30 minutes. After centrifugation (1700g for
10 minutes), the supernatants were removed. The pellet was re-extracted with
15 ml of acetonitrile, and the two supernatants were combined. Aliquots (100–200 ml)
were counted by liquid scintillation counting. The mean recovery of radioactivity
after extraction was ;90–95%. The supernatants were concentrated to an approximate volume of 50 ml and then diluted with 200-ml 5 mM ammonium
acetate (pH = 3.0):acetonitrile (95:5). Aliquots (;30 ml) were injected onto the
LC-MS/MS for analysis.
Urine samples from all species were pooled such that the sample pool
represented ;90% of the total radioactivity excreted in the urine. Pooled urine
samples were evaporated to dryness at 37C using the Genevac EZ-2
193
Circulating Metabolites of PF-04991532
Fig. 2. HPLC-UV (l = 254 nm) of metabolic profile of PF-04991532
(900 mg once daily for 14 days) in human plasma.
Results
Circulating Metabolites of PF-04991532 in Humans
Circulating metabolites of PF-04991532 in humans were qualitatively examined in the 900-mg dose group from the 14-day MAD
study. Figure 2 shows a representative HPLC-UV (l = 254 nm)
chromatogram of day 14 pooled plasma samples from humans dosed
daily with 900 mg of PF-04991532. PF-04991532 [retention time
(tR) = 24.6 minutes; MH+ = 397.1485] was the major component in
circulation followed by the glucuronide conjugate M1 (tR = 20.2
minutes; MH+ = 573.1803) and possibly three monohydroxylated
metabolites, which included M2a (tR = 17.6 minutes; MH+ =
413.1431) and two poorly resolved peaks that have been designated
as M2b/M2c (tR = 17.9–18.1 minutes; MH+ = 413.1431), respectively.
Based on UV integration, M1 accounted for 11% of circulating drugrelated material, whereas oxidative metabolites M2a and M2b/M2c
accounted for 18 and 20% of circulating drug-related material on day
14, respectively.
0.5 hours, respectively. The mean AUC(0–‘) and terminal phase t1/2 for
total radioactivity was 25,700 ng-equivalent×h/ml and 4.62 hours,
respectively, in female rats. In male rats, elimination t1/2 and AUC(0–‘)
values for total radioactivity were not determined because a definitive
elimination phase was not apparent in the plasma.
Dogs. After administration of a single oral NOAEL dose of
[14C]PF-04991532 (80 mg/kg) to beagle dogs, an overall mean (6 S.D.)
of 82.9 6 3.74% of the total dose was recovered in urine and feces
through 144 hours postdose (Table 1). Most of the excreted
radioactivity was recovered in the first 48 hours. No gender-related
differences were observed in the excretion pattern of radioactivity. The
mean cumulative dose recovered in the feces and urine was 40.1 6
4.5% and 12.0 6 3.1%, respectively. Due to the formation of liquid
feces, an unusually high percentage of radioactivity was recovered in
the cage debris, rinse, wash, and wipe, and this accounted for 30.8 6
3.6% of the administered dose (Table 1). The mean Cmax for total
radioactivity in male and female dogs (n = 2/gender) was 9040 and
6980 ng-equivalent/ml and peaked at 1.5 and 4.0 hours, respectively.
The mean AUC(0–‘) values for total radioactivity were 36,600 and
32,900 ng-equivalent×h/ml in male and female dogs, respectively. The
mean terminal phase t1/2 for total radioactivity was 2.30 and 1.26 hours
for male and female dogs, respectively.
Quantitative Profiles of [14C]PF-04991532 and Metabolites in
Excreta and Circulation
Rat Urine. A representative HPLC radiochromatogram of urinary
metabolites from an intact female rat is shown in Fig. 3A. The
metabolites were quantified by integration of the radiochromatographic peaks. The mean percentages of the metabolites detected in
the urine of male and female rats, expressed as a percentage of
the administered dose, are shown in Table 2. Unchanged parent
Mass Balance and Pharmacokinetics of Total Radioactivity
Rats. After administration of a single oral NOAEL dose of [14C]PF04991532 (200 mg/kg) to Sprague-Dawley rats, an overall mean
(6 S.D.) of 91.5 6 1.3% of the total radioactive dose was recovered in
the urine, feces, and cage wash/rinse over a period of 168 hours postdose
(Table 1). The mean cumulative dose recovered in feces was 77.0 6
8.7%. The mean cumulative excretion in the urine was 11.7 6 5.7%.
The majority of excreted radioactivity recovery occurred in the first
48 hours. No gender-related differences were observed in the excretion
pattern of radioactivity. The mean Cmax for total radioactivity in male
and female rats was 6060 and 10,600 ng-equivalent/ml and peaked at
TABLE 1
Material balance and routes of excretion of [14C]PF-04991532 drug-related material
in rats and dogs
Values represent mean 6 S.D.
Species
Percentage
Percentage
Percentage
Percentage
of
of
of
of
dose in urine
dose in feces
dose in cage debris
total dose recovered
Rat
11.7
77.0
2.8
91.5
6
6
6
6
Dog
5.7
8.7
0.8
1.3
12.0
40.1
30.8
82.9
6
6
6
6
3.1
4.5
3.6a
3.74
a
Due to liquid feces in dogs, an unusually high percentage of radioactivity was recovered in
the cage debris, wash, and wipes.
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
a SYNAPT G2 (Waters MS Technologies, Manchester, UK) orthogonal
acceleration quadrupole time-of-flight MS. The MS was operated in a positive
ion mode using electrospray ionization. The desolvation gas was set to 749 l/h
at a temperature of 300C. The cone gas was set to 20 l/h, and the source
temperature was set to 150C. The capillary voltage was set to 4.3 kV, the cone
voltage was set to 45 V, and the extraction cone was set to 5 V. The SYNAPT
G2 was operated in V optics mode (sensitivity mode), with a resolution greater
than 10,000 at full width at half maximum. The data acquisition rate was 0.10
s/scan; data were collected from 1 to 49 minutes using the dynamically variable
collision energy MS acquisition mode, with a collision energy ramp of 25-45 V
for high energy scans. The MS was calibrated to a mass accuracy under 5 mDa.
Data were collected in the continuum mode from m/z 100 to 900. MassLynx
version 4.1 SCN 803 software (Waters Corporation) was used for data
processing.
194
Sharm et al.
Fig. 3. Representative radiochromatograms of rat urine (A), feces (B),
and plasma (C). DPM, disintegrations per minute.
Metabolite Identification
The structures of metabolites (Table 4) were elucidated by electrospray
LC-MS/MS using a combination of full scan and CID product ion
(MS2/MS3) scanning techniques.
Parent (Unchanged Drug). PF-04991532 had an exact mass of
397.1485 (MH+). Its MS2 product ion spectrum (Supplemental Fig. 1)
yielded a diagnostic fragment ion at m/z at 261.1241, resulting from
the loss of the trifluoromethyl imidazolyl motif. Fragmentation of the
fragment ion at m/z = 261.1241 resulted in diagnostic MS3 fragments
at m/z = 243.1133, 193.0607, 179.0452, 165.0295, 139.0501, and
123.0802. The ions at 123.0802 and 139.0501 are formed via cleavage
across the amide bond.
Metabolite M1. Metabolite M1 had an exact mass of 573.1803
(MH+), which was 176 Da higher than PF-04991532. Its MS2 product
ion spectrum showed a diagnostic fragment ion at m/z 397.1485
resulting from the neutral loss of 176 Da, suggesting that M1 was
a glucuronide conjugate of PF-04991532. The tR and CID spectrum of
M1 was identical to that of a crude preparation of the synthetic acyl
glucuronide standard (data not shown).
Metabolites M2a and M2b/M2c. Metabolites M2a and M2b/M2c
possessed an exact mass of 413.1431 (MH+), which was 16 Da higher
than PF-04991532 and suggestive of monohydroxylated metabolites.
The MS2 product ion spectra were identical for M2a and M2b/M2c
TABLE 2
Relative abundance (percentage of total administered dose) of urinary and fecal
metabolites of PF-04991532 in rats and dogs after oral administration of
[14C]PF-04991532
Rat
Dog
Metabolites
PF-04991532
M1
M2a/M2b
M3
Urine
Feces
Total
Urine
Feces
Total
10.9
0.1
0.4
0.1
66.6
2.2
8.2
77.5
2.3
8.6
0.1
11.5
0.5
38.7
1.4
50.2
1.9
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
PF-04991532 accounted for 10.9% of the dose. Metabolites accounted
for ,1.0% of the administered dose.
Rat Feces. A representative HPLC radiochromatogram of fecal
metabolites from one intact male rat is shown in Fig. 3B. The mean
percentages of fecal metabolites detected in male and female rats,
expressed as percentages of the administered dose, are shown in
Table 2. Unchanged parent PF-04991532 accounted for 66.6% of the
dose. Oxidative metabolites (M2a and M2b) were the major
metabolites that collectively accounted for 8.2% of the dose, whereas
M1 accounted for 2.2% of the administered dose. There was no
evidence for the presence of the putative M2c metabolite in rat feces.
Rat Plasma. A representative HPLC radiochromatogram of circulating
metabolites from one male rat is shown in Fig. 3C. The mean percentages
of total circulating radioactivity are shown in Table 3. The majority of
circulating radioactivity was attributed to the unchanged parent drug
(71.5%) followed by M1 (28.5%).
Dog Urine. A representative HPLC radiochromatogram of urinary
metabolites from one male dog is shown in Fig. 4A. The metabolites
were quantified by integration of the radiochromatographic peaks. The
mean percentages of urinary metabolites detected in the urine of male
and female dogs, expressed as percentages of the administered dose,
are shown in Table 2. PF-04991532 and one metabolite M1 were
detected in the urine of male and female dogs. The unchanged parent
drug accounted for 11.5% of the dose. M1 represented 0.5% of the dose.
Dog Feces. A representative HPLC radiochromatogram of fecal
metabolites from one male dog is shown in Fig. 4B. The mean
percentages of fecal metabolites detected in male and female rats,
expressed as percentages of the administered dose, are shown in
Table 2. PF-04991532 and one metabolite M1 were detected in the
feces of male and female dogs. The unchanged parent drug accounted
for 38.7% of the dose. M1 represented 1.4% of the dose.
Dog Plasma. A representative HPLC radiochromatogram of
circulating metabolites from one male dog is shown in Fig. 4C. The
mean percentages of total circulating radioactivity are depicted in
Table 3. The majority of circulating radioactivity was attributed to the
unchanged parent drug (91.4%) followed by M1 (8.6%).
195
Circulating Metabolites of PF-04991532
TABLE 3
Relative abundance of circulating metabolites of PF-04991532 in rats and dogs after
oral administration of PF-04991532
Metabolite
Rat
Dog
PF-04991532
M1
M2a/M2b
M3
Total
71.5
28.5
N.D.
N.D.
100
91.4
8.6
N.D.
N.D.
100
N.D., not detected.
Oxidative Metabolism of Unlabeled PF-04991532 in Cryopreserved
Rat, Dog, and Human Hepatocytes
Qualitative examination of the metabolic profile in rat and human
hepatocytes incubated for 4 hours with PF-04991532 (10 mM)
revealed trace levels of M2a and M2b (Supplemental Fig. 4), with an
Discussion
The availability of plasma samples from the phase 1 MAD study of
PF-04991532 in T2DM patients provided an initial glimpse of
circulating human metabolites. Metabolite scouting studies used the
steady state plasma samples (day 14) from the highest dose group
(900 mg once daily) of PF-04991532. HPLC-UV analysis indicated
PF-04991532 as the major component in circulation in humans
Fig. 4. Representative radiochromatograms of dog urine (A), feces (B),
and plasma (C). DPM, disintegrations per minute.
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
(Supplemental Fig. 2) and showed diagnostic fragment ions at m/z
277.1186 and m/z 259.1081, which were 16 Da higher than the
primary MS2 fragment ions observed for PF-04991532, implying that
the trifluoromethylimidazolyl portion of the molecule was unaltered.
The MS3 product ion spectrum of m/z at 277 yielded diagnostic
fragment ions at m/z 179.0451 and 139.0483, suggesting that the site
of hydroxylations in the three metabolites were on the cyclopentyl
ring. Additional insights into the structures of M2a and M2b/M2c are
offered in the discussion section of the manuscript.
Metabolite M3. Metabolite M3 had an exact mass of 139.0502
(MH+), which was 258 Da lower than PF-04991532. Its MS2 product
ion spectrum showed a diagnostic fragment ion at m/z 121.0396,
which represented the loss of water (Supplemental Fig. 3). The tR and
CID spectrum of M3 was identical to an authentic standard of
6-aminonicotinic acid, implying that M3 was derived from an amide
bond cleavage in PF-04991532.
increased production of these metabolites (Supplemental Fig. 5) using
the relay method (total incubation time of 20 hours) previously
described in our laboratories (Di et al., 2012, 2013; Ballard et al.,
2014). No metabolites of PF-04991532 were detected in dog
hepatocytes at 4 or 20 hours. Examination of relay incubations also
used a UPLC system with a slower flow rate (0.35 ml/min versus
1.0 ml/min) as a preliminary tactic to resolve M2b from M2c. As
seen in Supplemental Fig. 5, M2a, M2b, and M2c (MH+ =
413.1431) now appeared as three distinct peaks at tR = 35.08,
36.14, and 36.30 minutes, respectively, in the human hepatocyte
incubations. Metabolite M2c was not detected in the rat hepatocyte
relay incubations, which is consistent with the in vivo findings in
the rat mass balance/excretion studies. Interestingly, ketone
metabolites (MH+ = 411.1274; tR = 36.9 and 37.7 minutes) possibly
derived from oxidation of the secondary alcohol metabolites (M2a,
M2b, and/or M2c) were also noted in rat and human hepatocyte
incubations using the relay method. As mentioned earlier, no ketone
metabolites were detected in the course of in vivo rat mass balance
studies and human circulatory metabolic profiling. Furthermore, in
contrast with the lack of metabolic turnover (t1/2 . 240 minutes)
of PF-04991532 in human hepatocyte incubations conducted for
4 hours, the human hepatocyte relay assay yielded a t1/2 value of
;570 minutes, which yielded an unbound intrinsic clearance value of
7.3 ml/min per kg (fraction of PF-04991532 unbound in the incubation =
0.87) using our previously described protocol (Di et al., 2012, 2013).
196
Sharm et al.
TABLE 4
Mass spectral characteristics of metabolites of PF-04991532
Metabolite
MH+
tR
Structure
Fragment Ions
min
397.1485
24.6
261.1241, 243.1133, 179.0452, 165.0295,
139.0501, 123.0802
M1
573.1803
20.2
437.1558, 397.1485, 261.1234
M2a, M2b/M2ca
413.1431
17.6, 17.9–18.1
M3
139.0502
3.0
277.1186, 259.1081, 179.0450, 139.0498
121.0396, 95.0604
a
As shown in Fig. 2, metabolite scouting of FIH plasma samples revealed a well resolved (M1; tR = 17.6 minutes) and two poorly resolved (M2b/M2c; tR = 17.9/18.1 minutes) peaks at MH+ = 413,
with an identical fragmentation pattern. The non-Gaussian peak shape (tR = 17.9–18.1 minutes in Fig. 2) strongly suggests the presence of more than one metabolite, and hence we have invoked the
designation M2b/M2c. M2b and M2c generated in the human hepatocyte relay method were separated into two distinct peaks (tR = 36.1 and 36.3 minutes, respectively) using a UPLC method.
alongside an acyl glucuronide metabolite (M1) and monohydroxylated
metabolites M2a and M2b/M2c, respectively. The relative abundance
of M1, M2a, and M2b/M2c was estimated to be ;11, ;18, and
;20%, respectively, in human plasma. These results, while preliminary in nature, suggest that metabolism plays a significant role in
the in vivo elimination of PF-04991532 in humans. Approximately 38
and 11% of the drug-related material in circulation is attributed to
oxidative and glucuronide metabolites, respectively, with the extent of
oxidative metabolism in humans exceeding the amount observed in
rats and dogs. These findings were somewhat paradoxical with our
preclinical observations on the metabolic resistance of PF-04991532
in liver microsomes and hepatocytes from animals and humans.
Qualitative profiling of metabolites in hepatocytes from animals and
humans with unlabeled PF-04991532 revealed trace levels of M1, M2a,
and M2b in rat and human hepatocytes, with no clear differentiation of
major versus minor metabolites (Supplemental Fig. 2). The formation of
oxidation products (M2a and M2b) in human liver microsomes was
dependent on NADPH (data not shown), implying a role for the
cytochrome P450 enzyme(s).
Collectively, these observations challenge our original hypothesis
regarding biliary elimination of unchanged PF-04991532 as the
principal clearance mechanism in humans. Likely, PF-04991532 falls
into the broad category of slowly metabolized compounds, which
often rely on metabolic elimination as a route of clearance in vivo
(Li et al., 2014; Ramsden et al., 2014). The metabolic (oxidative and/or
conjugation) turnover rates for slowly metabolized compounds cannot
be reliably estimated in standard liver microsomal and/or hepatocyte
assays that are conducted for short incubation periods. Recently, we
described the utility of a novel human hepatocyte suspension relay
assay that extends drug residence time in cryopreserved hepatocytes
and provides reliable estimates of in vitro intrinsic clearance and
metabolic profiling data for slowly metabolized compounds (Di et al.,
2013; Ballard et al., 2014). Using the relay conditions, we were able
to measure a t1/2 value (;570 minutes) for metabolic decline of
PF-04991532 in human hepatocytes that translated into an unbound
intrinsic clearance of 7.3 ml/min per kg. The in vitro intrinsic
clearance for PF-04991532 correlates with the in vivo unbound
hepatic intrinsic clearance estimate of ;10.9 ml/min per kg that has
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
Parent
Circulating Metabolites of PF-04991532
Fig. 5. Cyclopentyl ring hydroxylation of glucokinase activator 3 by cytochrome
P450 (Sarabu et al., 2012).
hepatic systems. Consequently, it would be interesting to re-examine the
compounds with the in vitro–in vivo disconnect using the hepatocyte
relay conditions. A pragmatic first step in this direction was recently
communicated by our group (Ballard et al., 2014).
To examine circulating metabolites in animals, mass balance/
excretion studies were initiated with [14C]PF-04991532. Following the
oral administration of [14C]PF-04991532, mass balance was achieved,
with mean total recoveries of 91.5% (rats) and 82.9% (dogs). Excretion
of radioactivity was rapid and nearly complete within 48 hours after
dosing in both species. Of the total radioactivity excreted in rat feces
(77%) and urine (11.7%), unchanged PF-04991532 constituted 84.6%
of the dose. Likewise, unchanged PF-04991532 accounted for 60.5% of
the total administered dose in dog feces (40.1%) and urine (12.0%),
respectively. Minor metabolites identified in rat and dog excreta included
an acyl glucuronide (M1), two monohydroxylated regioisomers (M2a and
M2b), and 6-aminonicotinic acid (M3). The majority of circulating
radioactivity in animals was attributed to unchanged PF-04991532
(71.5 and 91.4% in rats and dogs, respectively). The glucuronide
conjugate M1 accounted for 28.5 and 8.6% of the circulating radioactivity
in rats and dogs, respectively. However, the monohydroxylated metabolites
M2a and M2b/M2c, which were observed as circulating metabolites in
humans, were not detected in plasma from rats and dogs. Although
oxidative metabolites of PF-04991532 were not detected in circulation
in animals, their presence was collectively noted in the excreta of at
least one of the preclinical species for toxicological evaluation (M2a
and M2b in particular were observed in rat excreta in 8.6%
abundance). While oxidative metabolism of PF-04991532 does not
result in a major alteration in structure, the process yields metabolites
with increased polarity relative to the parent, as assessed from the
calculated logP (PF-04991532 = 3.65; M2a/M2b = 1.57) and
topological polar surface area (PF-04991532 = 94.36 A2; M2a/M2b =
114.59 A2). Like the parent compound, it is possible that the
monohydroxylated metabolites (M2a and M2b/M2c) are also
eliminated in the bile to some degree (Smith and Dalvie, 2012).
Because of the relatively poor passive permeability of these compounds
Fig. 6. Possible diastereomers that can be generated from cyclopentyl
ring hydroxylation in PF-04991532.
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
been estimated from the oral CLp of PF-04991532 in humans (unpublished observations). Qualitative examination of the metabolic profile using the relay method demonstrated greater oxidative metabolism
of PF-04991532 in human (;46% of the total integrated HPLC-UV
peak area) and rat (;26% of the total integrated HPLC-UV peak area)
hepatocytes relative to that noted in the 4-hour hepatocyte incubations (;5% of the total integrated HPLC-UV peak area). Incidentally, no metabolites of PF-04991532 were discerned in dog hepatocyte
incubations conducted for 4 or 20 hours, respectively. Despite the
qualitative nature of this study, the in vitro oxidative metabolic profile
of PF-04991532 in the hepatocyte relay method has much resemblance
to the in vivo oxidative metabolism of PF-04991532 noted in excreta
and/or circulation in preclinical species and humans (human . rat . .
dog). In a retrospective analysis utilizing ;48 compounds, Dalvie et al.
(2009) demonstrated that in vitro metabolism studies in human hepatocytes (under normal incubation periods of ;4 hours) was a reasonable predictor of primary in vivo metabolites for several compounds.
In some cases, compounds failed to generate the human metabolites in
vitro, which was ascribed to the slow metabolic turnover in the in vitro
197
198
Sharm et al.
Authorship Contributions
Participated in research design: Sharma, Bergman, Litchfield, Kazierad,
Pfefferkorn, Kalgutkar.
Conducted experiments: Sharma, Atkinson, Bergman, Kazierad, Gustavson.
Contributed new reagents or analytic tools: Sharma.
Performed data analysis: Sharma, Litchfield Atkinson, Kalgutkar.
Wrote or contributed to the writing of the manuscript: Kalgutkar, Sharma,
Litchfield, Bergman, Di, Pfefferkorn.
References
Ballard TE, Orozco CC, and Obach RS (2014) Generation of major human excretory and circulating drug metabolites using a hepatocyte relay method. Drug Metab Dispos 42:899–902.
Coghlan M and Leighton B (2008) Glucokinase activators in diabetes management. Expert Opin
Investig Drugs 17:145–167.
Dalvie D, Obach RS, Kang P, Prakash C, Loi C-M, Hurst S, Nedderman A, Goulet L,
Smith E, and Bu H-Z, et al. (2009) Assessment of three human in vitro systems in the
generation of major human excretory and circulating metabolites. Chem Res Toxicol 22:
357–368.
Di L, Atkinson K, Orozco CC, Funk C, Zhang H, McDonald TS, Tan B, Lin J, Chang C,
and Obach RS (2013) In vitro-in vivo correlation for low-clearance compounds using hepatocyte relay method. Drug Metab Dispos 41:2018–2023.
Di L, Trapa P, Obach RS, Atkinson K, Bi Y-A, Wolford AC, Tan B, McDonald TS, Lai Y,
and Tremaine LM (2012) A novel relay method for determining low-clearance values. Drug
Metab Dispos 40:1860–1865.
European Medicines Agency (2009) ICH topic M3 (R2) non-clinical safety studies for the conduct
of human clinical trials and marketing authorization for pharmaceuticals, London.
Ghosh A, Maurer TS, Litchfield J, Varma MV, Rotter C, Scialis R, Feng B, Tu M, Guimaraes
CR, and Scott DO (2014) Toward a unified model of passive drug permeation II: the physiochemical determinants of unbound tissue distribution with applications to the design of
hepatoselective glucokinase activators. Drug Metab Dispos 42:1599–1610.
Grimsby J, Sarabu R, Corbett WL, Haynes NE, Bizzarro FT, Coffey JW, Guertin KR, Hilliard
DW, Kester RF, and Mahaney PE, et al. (2003) Allosteric activators of glucokinase: potential
role in diabetes therapy. Science 301:370–373.
Gustavson SM, Pfefferkorn JA, Kazierad DJ, Chen D, Bergman A, Wang X, Rolph TP, and
Rusnak JM (2013) Pharmacokinetics and pharmacodynamics of PF-04991532, a hepatoselective glucokinase activator (GKA), in T2DM patients on metformin therapy, in 73rd
American Diabetes Association Meeting; 2013 June 21–25; Chicago, IL. Abstract 2635-PO,
American Diabetes Association, Alexandria, VA.
Hamilton RA, Garnett WR, and Kline BJ (1981) Determination of mean valproic acid serum level
by assay of a single pooled sample. Clin Pharmacol Ther 29:408–413.
Haynes NE, Corbett WL, Bizzarro FT, Guertin KR, Hilliard DW, Holland GW, Kester RF,
Mahaney PE, Qi L, and Spence CL, et al. (2010) Discovery, structure-activity relationships,
pharmacokinetics, and efficacy of glucokinase activator (2R)-3-cyclopentyl-2-(4methanesulfonylphenyl)-N-thiazol-2-yl-propionamide (RO0281675). J Med Chem 53:3618–3625.
International Conference on Harmonisation (2012) Questions and answers (R2), Geneva,
Switzerland.
Lagas JS, Sparidans RW, Wagenaar E, Beijnen JH, and Schinkel AH (2010) Hepatic clearance of
reactive glucuronide metabolites of diclofenac in the mouse is dependent on multiple ATPbinding cassette efflux transporters. Mol Pharmacol 77:687–694.
Li Y, Zhou J, Ramsden D, Taub ME, O’Brien D, Xu J, Busacca CA, Gonnella N, and Tweedie DJ
(2014) Enzyme-transporter interplay in the formation and clearance of abundant metabolites of
faldaprevir found in excreta but not in circulation. Drug Metab Dispos 42:384–393.
Matsushima S, Maeda K, Hayashi H, Debori Y, Schinkel AH, Schuetz JD, Kusuhara H,
and Sugiyama Y (2008) Involvement of multiple efflux transporters in hepatic disposition of
fexofenadine. Mol Pharmacol 73:1474–1483.
Mithieux G (1996) Role of glucokinase and glucose-6 phosphatase in the nutritional regulation of
endogenous glucose production. Reprod Nutr Dev 36:357–362.
Matschinsky FM and Ellerman JE (1968) Metabolism of glucose in the islets of Langerhans.
J Biol Chem 243:2730–2736.
Matschinsky FM, Glaser B, and Magnuson MA (1998) Pancreatic b-cell glucokinase: closing the
gap between theoretical concepts and experimental realities. Diabetes 47:307–315.
Matschinsky FM, Zelent B, Doliba NM, Kaestner KH, Vanderkooi JM, Grimsby J, Berthel SJ,
and Sarabu R (2011) Research and development of glucokinase activators for diabetes therapy:
theoretical and practical aspects. Handbook Exp Pharmacol 203:357–401.
Meininger GE, Scott R, Alba M, Shentu Y, Luo E, Amin H, Davies MJ, Kaufman KD,
and Goldstein BJ (2011) Effects of MK-0941, a novel glucokinase activator, on glycemic
control in insulin-treated patients with type 2 diabetes. Diabetes Care 34:2560–2566.
Pfefferkorn JA (2013) Strategies for the design of hepatoselective glucokinase activators to treat
type 2 diabetes. Expert Opin Drug Discov 8:319–330.
Pfefferkorn JA, Guzman-Perez A, Litchfield J, Aiello R, Treadway JL, Pettersen J, Minich ML,
Filipski KJ, Jones CS, and Tu M, et al. (2012) Discovery of (S)-6-(3-cyclopentyl-2-(4(trifluoromethyl)-1H-imidazol-1-yl)propanamido)nicotinic acid as a hepatoselective glucokinase
activator clinical candidate for treating type 2 diabetes mellitus. J Med Chem 55:1318–1333.
Ramsden D, Tweedie DJ, Chan TS, Taub ME, and Li Y (2014) Bridging in vitro and in vivo
metabolism and transport of faldaprevir in human using a novel cocultured human hepatocyte
system, HepatoPac. Drug Metab Dispos 42:394–406.
Sarabu R, Berthel SJ, Kester RF, and Tilley JW (2011) Novel glucokinase activators: a patent
review (2008 - 2010). Expert Opin Ther Pat 21:13–33.
Sarabu R, Bizzarro FT, Corbett WL, Dvorozniak MT, Geng W, Grippo JF, Haynes NE,
Hutchings S, Garofalo L, and Guertin KR, et al. (2012) Discovery of piragliatin—first glucokinase activator studied in type 2 diabetic patients. J Med Chem 55:7021–7036.
Smith DA and Dalvie D (2012) Why do metabolites circulate? Xenobiotica 42:107–126.
Smith DA and Obach RS (2009) Metabolites in safety testing (MIST): considerations of mechanisms
of toxicity with dose, abundance, and duration of treatment. Chem Res Toxicol 22:267–279.
U.S. Food and Drug Administration (2008) Guidance for industry: safety testing of drug
metabolites, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, U.S. Department of Health and Human Services, Rockville, MD.
Address correspondence to: Amit S. Kalgutkar, Pharmacokinetics, Dynamics,
and Metabolism–New Chemical Entities, Pfizer Worldwide Research and Development, 620 Memorial Drive, Cambridge, MA 02139. E-mail: [email protected]
Downloaded from dmd.aspetjournals.org at ASPET Journals on May 7, 2017
(Ghosh et al., 2014), the interplay between the export of the parent and
its metabolites from the liver into the blood and/or bile will likely rely on
the individual affinities of these molecules for active transport via influx
[multidrug resistant protein (MRP) isoforms 3, 4, and 6] and efflux
transporters (MRP2, P-glycoprotein and/or breast cancer resistant
protein), respectively, in rats and humans. If so, then the finding that
comparative MRP3 transport activity is higher in humans than in
rodents (Lagas et al., 2010) provides a possible explanation for the greater
percentage of oxidative metabolites in circulation in humans compared
with rats. Examination of the in vitro active transport of PF-04991532
and its oxidative metabolites will shed additional light on this matter.
The mass spectra of M2a and M2b/M2c were identical and indicated
that these metabolites were derived from a single site of oxidation on the
cyclopentyl ring in PF-04991532. With the aid of UPLC, we were
further able to chromatographically resolve M2b and M2c into two
distinct peaks in the human hepatocyte relay assay. Preliminary
insights into the structures of M2a, M2b, and M2c were obtained via
comparison of the known metabolic fate of the structurally related
glucokinase activator, referred to as compound 3 (Fig. 5), in the
publication of Sarabu et al. (2012). In liver microsomes, compound
3 was shown to undergo monohydroxylations on the C-2 and C-3
positions on the cyclopentyl group, generating the corresponding
alcohol regioisomers (Sarabu et al., 2012). It is noteworthy to point out
that hydroxylation at either the C-2 and/or C-3 position of the
cyclopentyl ring in PF-04991532 and compound 3 will result in the
conversion of the prochiral methine at C-1 into a chiral carbon atom and
lead to the formation of diastereomeric products. Our present study as
well as the one from Sarabu et al. (2012) characterized three metabolites
derived from cyclopentyl ring oxidation. In theory, single site oxidations on the C-2 or the C-3 positions on the cyclopentyl ring could
potentially lead to the formation of eight diastereomers, as shown
in Fig. 6. Therefore, in the case of PF-04991532, it is possible that
additional diastereomers are also formed and will need to be resolved
from the distinct M2a, M2b, and M2c peaks detected in our present
studies. Sarabu et al. (2012) demonstrated that the diastereomeric
mixture of C-2 and C-3 hydroxylated metabolites of compound 3,
generated via independent synthesis, were ;2 orders of magnitude less potent as GK activators as compared with the parent
compound. By inference, we conclude that the propensity of M2a,
M2b, and M2c to activate GK would be significantly lower than
PF-04991532.
In conclusion, a human metabolite scouting study in tandem with
the animal mass balance study allowed early identification of oxidative
metabolites in circulation in humans but not animals. A human mass
balance study with a clinically efficacious dose of [14C]PF-04991532
(estimated to be 450 mg q.d. based on phase 2 clinical efficacy data) is
currently in progress to supplement the animal mass balance/excretion
data. Ongoing activities also include the elucidation of the human
cytochrome P450 isoforms responsible for PF-04991532 oxidation and
the isolation and characterization of the diastereomeric oxidative
metabolites. Obtaining authentic metabolite standards will be critical
for the determination of absolute amounts (total/unbound) of systemic
exposures in humans and animals as opposed to percentage values
for metabolites as comparators since toxicity thresholds are ultimately
related to absolute exposure values (Smith and Obach, 2009). These
collective findings will be reported in due course.