Degradation kinetics and mechanism of an oxadiazole derivative, design of a stable drug product for BMS-708163, a γ-secretase inhibitor drug candidate

Article abstract of DOI:10.1002/jps.23050

The stability of a 1,2,4-oxadiazole derivative, BMS-708163, A, was studied in the cosolvent mixture of acetonitrile-water at various pH values, in the solid state and in the presence of various excipients. The objective of this study was to develop a deep understanding of the stability of compound A based on its degradation kinetics and mechanism to enable design of a robust drug product. A series of isotopically ¹³C- and ¹⁵N-labeled compounds were synthesized and their degradation was studied by liquid chromatography-mass spectrometry and nuclear magnetic resonance to prove the degradation mechanism. Compound A exhibited maximum stability at a pH range of 3-5. In forced degradation studies, higher or lower pH resulted in an increase in degradation rate. At low pH, the N-4 atom on the 1,2,4-oxadiazole ring is protonated, followed by nucleophilic attack on the activated methine carbon to cause ring opening to form aryl nitrile degradation product, compound B. At high pH, the nucleophilic attack occurs on the methine carbon to generate an anion on N-4. Subsequent proton capture from a proton donor, such as ambient water, facilitates ring opening to generate compound B. In the absence of a proton donor, such as in dry acetonitrile, anion on N-4 will go back to compound A. Therefore, compound A is stable in absence of proton donor. This study defines the package condition and microenvironmental pH under which compound A can be formulated into a stable product.

Full text of DOI:10.1002/jps.23050

RESEARCH ARTICLES
Degradation Kinetics and Mechanism of an Oxadiazole
Derivative, Design of a Stable Drug Product for BMS-708163,
a γ-Secretase Inhibitor Drug Candidate
RUILING F. HARTLEY, YANDE HUANG, MICHAEL CASSIDY, THOMAS M. RAZLER, FENG QIAN, MUNIR A. HUSSAIN
Drug Product Science and Technology, Bristol–Myers Squibb Company, New Brunswick, New Jersey 08903
Received 17 August 2011; revised 5 December 2011; accepted 15 December 2011
Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.23050
ABSTRACT: The stability of a 1,2,4-oxadiazole derivative, BMS-708163, A, was studied in
the cosolvent mixture of acetonitrile–water at various pH values, in the solid state and in the
presence of various excipients. The objective of this study was to develop a deep understanding
of the stability of compound A based on its degradation kinetics and mechanism to enable
13
15
design of a robust drug product. A series of isotopically C- and N-labeled compounds were
synthesized and their degradation was studied by liquid chromatography–mass spectrometry
and nuclear magnetic resonance to prove the degradation mechanism. Compound A exhibited
maximum stability at a pH range of 3–5. In forced degradation studies, higher or lower pH
resulted in an increase in degradation rate. At low pH, the N-4 atom on the 1,2,4-oxadiazole
ring is protonated, followed by nucleophilic attack on the activated methine carbon to cause
ring opening to form aryl nitrile degradation product, compound B. At high pH, the nucleophilic
attack occurs on the methine carbon to generate an anion on N-4. Subsequent proton capture
from a proton donor, such as ambient water, facilitates ring opening to generate compound
B. In the absence of a proton donor, such as in dry acetonitrile, anion on N-4 will go back to
compound A. Therefore, compound A is stable in absence of proton donor. This study defines
the package condition and microenvironmental pH under which compound A can be formulated
into a stable product. © 2012 Wiley Periodicals, Inc. and the American Pharmacists Association
J Pharm Sci
Keywords: chemical stability; degradation products; mechanism; oxadiazole; kinetics;
formulation; solid dispersion
INTRODUCTION
BMS-708163 (Compound A in Fig. 1) is being de-
veloped for the inhibition of (-secretase to reduce
the amyloid-$ aggregation that is believed to have
The 1,2,4-oxadiazole (isoxadiazole) ring system has
been used extensively in medicinal chemistry. This
structural motif is found in several commercially mar-
1
5,16
a role in the progression of Alzheimer’s disease.
Compound A has a solubility of less than 1 :g/mL,
and is nonionizable under physiologically relevant
pH range. Owing to its low solubility, BMS-708163
is being developed as an amorphous solid disper-
sion in cellulose-based polymers. Preliminary studies
showed that compound A was physically and chemi-
cally stable for at least 3 months in accelerated sta-
1
–4
5
6–8
keted drugs such as Picovir, Proxodolol, Etirac-
9
10
etam, Memolog, Rolipram, Pidotimod, and other
1
1,12
drug products.
The oxadiazole ring serves as a
bioisostere of carboxylic esters and amides, and has
contributed to increased pharmacological activities by
participating in hydrogen-bonding interactions with
4
biological receptors. Derivatized oxadiazole ring sys-
◦
bility studies [40 C/75% relative humidity (RH) open
tems were found to decompose under high tempera-
ture conditions, through ring cleavage, to afford ni-
trile and isocyanate derivatives.1
◦
and 50 C close conditions] in the solid state. How-
ever, it was found that compound A was much less
stable in solution. In solution at both low and high
pH and in drug–excipient compatibility studies, the
main degradant was found to be a nitrile degradation
product (compound B in Fig. 1), which increased with
time during storage at elevated temperature.
3,14
Correspondence to: Ruiling F. Hartley (Telephone: +732-227-
6
226; Fax: +732-227-3782; E-mail: ruiling.hartley@bms.com)
Journal of Pharmaceutical Sciences
2012 Wiley Periodicals, Inc. and the American Pharmacists Association
©
JOURNAL OF PHARMACEUTICAL SCIENCES
1

2
HARTLEY ET AL.
Figure 1. Structures of compound A, compound B (aryl nitrile degradation product of com-
pound A), compound C (amide hydrolysis of compound A), and compound D (amide hydrolysis
15
13
of aryl nitrile degradant compound B), N-labeled compound E, C-labeled compound F, and
methylated compound G.
In the presence of nucleophilic groups, Claisse
MATERIALS AND METHODS
Chemicals
1
7
et al. proposed that the degradation of the oxadi-
azole moiety likely follows a pathway wherein ring
opening occurs through a concerted reaction (Fig.
Compounds A–G (Fig. 1) were prepared by the
Chemical Development group at Bristol–Myers
Squibb, Company, New Brunswick, New Jersey.
Hydroxypropyl methylcellulose acetate succinate
(HPMC-AS) was obtained from Shin-Etsu Chemical
Company Ltd. (Chiyoda-ku, Tokyo, Japan), HPMC
2
), leading to a nitrile and isocyanate ion. The cur-
rent stability study was conducted to investigate the
degradation kinetics and mechanism, both in solution
and in the proposed solid dosage form to mitigate the
degradation risks in drug product.
JOURNAL OF PHARMACEUTICAL SCIENCES
DOI 10.1002/jps

DEGRADATION MECHANISM OF AN OXADIAZOLE DERIVATIVE, BMS-708163
3
lyzed at time zero and stored in stability chambers
under differential conditions. Samples were with-
drawn at predetermined time points, extracted using
methanol and analyzed by a gradient RP-HPLC.
Roughly about 1 mL distilled water was added into
1
g of each solid dispersion formulation, and then
stirred for 1–2 h to prepare a concentrated slurry of
the formulation. The pH of the slurry was determined
using a pH meter and designated as the microenviron-
mental pH of the solid formulation.
Figure 2. Basic degradation mechanism of the 1,2,4-
oxadiazole ring proposed by Claisse et al.17
and polyvinylpyrrolidone–vinyl acetate (PVP–VA)
were obtained from Dow Chemical Company (Mid-
land, Michigan). Acetic acid, sodium acetate, citric
acid, sodium citrate, and deuterated solvents were
purchased from Sigma–Aldrich Chemical Company
RP-HPLC Analysis
A Waters Alliance 2695 HPLC system equipped with
a Waters 2487 ultraviolet–visible dual wavelength de-
tector (Milford, Massachusetts) was used to follow the
stability of compound A. A YMC-Pack Pro C18 column
(
St. Louis, Missouri).
[
150 × 3.0 mm internal diameter (i.d.), 3 :m particle
Compound A Chemical Stability Studies
Chemical Stability in Solution
size; Kyoto, Japan] was used to separate the com-
pounds at 30 C. The sample chamber temperature
was set as 5 C. Mobile phase A was 0.05% trifluo-
◦
◦
The degradation rate of compound A was determined
at various pH values ranging from 0.25 to 9. Com-
pound A solutions were prepared at apparent pH
values of 0.25 (1 N HCl), 1 (0.1 N HCl), 2 (0.01
N HCl), 3 (0.02 M citrate buffer), 4 (0.02 M citrate
buffer), 5 (0.02 M citrate buffer), 7 (0.02 M phos-
phate buffer), and 9 (0.02 M Tris buffer) and stored
roacetic acid (TFA) in water and mobile phase B was
0
.05% TFA in acetonitrile. The gradient elution pro-
file was linear ramp from 75% A initially to 85% B at
5 min. The flow rate was 1.0 mL/min and the detec-
3
tion wavelength was 244 nm.
◦
HPLC–MS Studies
at a temperature of 40 C. For all buffers, the ionic
strength (:) was adjusted to 0.1 M using sodium chlo-
ride. Compound A was practically insoluble in water
The isotopically labeled compounds solution stabil-
ity samples were characterized by HPLC–MS. The
HPLC analysis conditions were the same as the afore-
mentioned. The mass spectra were obtained on Wa-
ters ZQTM 4000 (Single Quadrupole Mass Spectrom-
eter) equipped with an electrospray ion source. The
(
<1 :g/mL); therefore, it was dissolved in acetonitrile
and added to the buffers to make 25:75 acetonitrile–
aqueous buffer solutions. The final concentration of
compound A was 0.05 mg/mL. The samples were filled
into serum vials, sealed with metal caps, tested at pre-
determined time points, and analyzed by a gradient
reversed-phase high-performance liquid chromatog-
raphy (RP-HPLC).
Solution chemical stabilities for the isotopically la-
beled (E and F) and methylated (G) compounds were
followed at apparent pH values of 1.44 (0.1 N HCl),
◦
source temperature was held at 125 C and the desol-
◦
vation temperature was 350 C. The capillary voltage
was 3.20 kV; cone voltage was 25.00 V; cone gas flows
was 25 L/h; desolvation gas flow was 750 L/h. Magnet
scans were performed over the mass–charge range of
18–1000 Da.
5
.40 (0.02 M acetate buffer), and 8.53 (0.02 M ac-
Nuclear Magnetic Resonance Studies
etate solution). Compounds were dissolved in 25:75
acetonitrile–aqueous solutions (v/v) with a final con-
centration of 0.25 mg/mL. Samples were filled into
serum vials, sealed with metal caps, and stored in
a 60 C oven. Samples were tested at predetermined
time points and analyzed by a gradient RP-HPLC and
liquid chromatography–mass spectrometry (LC–MS).
1H NMR spectra were obtained on a Bruker Avance
DRX 500 NMR instrument (Billerica, Massachusetts)
◦
◦
1
13
at 25 C or 60 C using a Nalorac 3-mm H– C dual
◦
13
probe or a Bruker CPTXI 5-mm probe. C NMR spec-
tra were obtained on a Bruker Avance DRX 500 NMR
◦
1
13
instrument at 25 C using a Nalorac 3-mm H– C
1
13
dual probe. H and C chemical shifts were refer-
enced to the residual nuclear magnetic resonance
Solid Dispersions Stability and Microenvironmental pH
Measurement
(NMR) solvent (CD3CN: 1.94 and 1.30 ppm). The
A mixture of compound A and polymer (HPMC,
HPMC-AS, or PVP–VA; 40:60 by weight) was dis-
solved in acetone and spray dried to make amorphous
solid dispersion. Each spray-dried powder was ana-
NMR samples were heated either inside the NMR in-
strument for data acquisition or in an oil bath and
then transferred to the NMR instrument for data
acquisition.
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES

4
HARTLEY ET AL.
A, each of the degradant peaks increased with time
and temperature. Chromatogram peaks from left to
right in Figure 4 were identified as compound B (aryl
nitrile degradation product), compound A, compound
D (amide hydrolysis products of compound B), and
compound C (amide hydrolysis products of compound
A), respectively. Under neutral or fairly basic condi-
tion, no amide hydrolysis was observed and peaks cor-
responding to compounds C and D were not detected
in the chromatogram.
Degradation Mechanism Studies
Isotopically labeled compound E and F and methy-
lated compound G solutions were prepared at pH
◦
values of 1.44, 5.40, and 8.53, and stored at 60 C.
Table 1 shows the HPLC results of the solution sta-
bility. The response factors for the compounds B, C,
and D relative to compound A were found to be 0.85,
Figure 3. pH-rate profile for the degradation of compound
A at 40 C. The observed first-order degradation rate con-
stants were not determined at zero buffer concentration.
1
.00, and 0.83, respectively. The concentrations of the
◦
degradants were determined from the peak areas cor-
1
5
13
rected for the response factors. N- and C-labeled
compound E and F showed similar degradation kinet-
ics to that of compound A. The methylated compound
G showed no degradation over the study period ex-
cept for the hydrolysis of the side chain amide moiety
under strongly acidic conditions, supporting the pro-
posal that the oxadiazole ring hydrogen is involved in
the degradation mechanism.
RESULTS AND DISCUSSION
Degradation Behavior
The stability of compound A in the pH range 0.25–9
◦
was studied at 40 C with different buffers. The pH
versus degradation rate profile of compound A is
shown in Figure 3. The first-order reaction rate was
The stability samples were also subjected to
LC–MS studies. Figure 5a shows the HPLC chro-
obtained using the equation: ln [A] = −kobst + ln[A]0,
t
1
5
where [A] is compound A concentration and kobsis the
observed first-order rate constant. Compound A was
maximally stable in a pH range of 3–5.
matogram for the stability of N-labeled compound
E with the insert showing the mass spectrum corre-
sponding to the nitrile product peak. The mass spec-
+
The molecular weight of the major degradant was
found to be 477 by MS and corresponds to compound
B. Figure 4 shows a typical HPLC chromatogram of
partially degraded compound A under acidic condi-
tions. In addition to the disappearance of compound
trum gives a protonated molecular ion [M + H] at m/
1
5
z 478.75, indicating the isotope N has been retained
in the degradant of compound E. The mass spectrum
1
3
in Figure 5b shows that the degradant of C-labeled
+
compound F corresponds to a molecular ion [M + H]
Figure 4. Typical HPLC chromatogram of compound A stability sample. Peaks from left to
right are compound B (aryl nitrile degradation product), compound A, compound D (amide
hydrolysis of compound B), and compound C (amide hydrolysis product of compound A).
JOURNAL OF PHARMACEUTICAL SCIENCES
DOI 10.1002/jps

DEGRADATION MECHANISM OF AN OXADIAZOLE DERIVATIVE, BMS-708163
5
at m/z 477.76, the same mass as compound B, indi-
cating that the label has been lost during the degra-
1
5
13
dation. From the MS studies of
N and C-labeled
compound E and F degradation products, the 1,2,4-
oxadiazole ring cleavage appears to occur at both the
1
5
N O and C N bond.
Once the position of the atoms involved in ring
cleavage was established, NMR studies were con-
ducted to investigate the mechanism of the degrada-
1
3
tion process. Compound F ( C labeled, 11.4 mg) was
dissolved in 0.5 mL of CD3CN and transferred into a
5
-mm i.d. NMR tube. The resulting anhydrous sample
1
13
was subjected to H and C NMR data acquisition.
Benzylamine (10 :L) was then added to the sample for
1
H NMR data acquisition under three different condi-
tions: time zero at room temperature (RT), overnight
◦
at RT, and 6 h at 60 C. On the basis of the mechanism
proposed in Figure 2, the formed isocyanate could be
trapped as benzylurea and easily observed by various
spectroscopic or chromatographic techniques. How-
ever, no indication of degradation was observed in any
1
of the resulting H NMR spectra. D2O (0.1 mL) was
1
then added to the same NMR sample and H NMR
spectrum was collected after overnight storage at RT.
A small amount of benzylurea was observed that sub-
stantially increased when the sample was heated at
◦
6
0 C for 6.5 h, as shown in Figure 6 (reaction is shown
in Fig. 7).
1
Figure 6 shows the overlay of H NMR spectra at
different conditions. The peak at δ 4.25 is assigned to
the methylene protons on benzylurea. The NMR ex-
periments show that benzylamine does not react with
compound F under anhydrous conditions, but D2O
significantly accelerates the reaction. The sequence
of this experiment clearly demonstrates that without
the presence of a proton donor, a weak base nucle-
ophile alone does not facilitate the oxadiazole ring
opening. In accordance with these findings, a mecha-
nism is proposed, which involves nucleophilic attack
on the methine carbon followed by protonation on the
resulting N-4 negative charge to form an unstable in-
termediate, as shown in Figure 8. The methine proton
in this intermediate becomes very acidic and prone to
facilitate the ring opening process. When the methine
proton was replaced by the methyl group (Compound
G), such ring opening pathway was eliminated and no
ring opening degradation was observed. At neutral or
weakly basic conditions, the degradation is likely pro-
ceeded with the same mechanism shown in Figure 8.
At low pH, the higher pKa N-4 nitrogen of the 1,2,4-
oxadiazole ring could become protonated and thus
facilitate the degradation process. As a result, the
methine carbon is activated and susceptible to the
nucleophilic attack, such as from H2O, to form a simi-
lar unstable intermediate and facilitate the concerted
ring opening degradation process (Fig. 9).
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES

6
HARTLEY ET AL.
Figure 5. LC–MS results of stability samples for compound E (a) and compound F (b) in
◦
15
sodium acetate solution stored at 60 C for 7 days, indicating that N atom is retained on the
nitrile group (m/z 478).
The dependence of the oxadiazole ring opening on
3 2
deuterium (CD CN/D O) in the presence of benzy-
the presence of proton donor under basic conditions
and the observed acid catalysis are in disagreement
with the mechanism previously proposed by Claisse
lamine at ambient temperature overnight. The mech-
anism of H/D exchange is proposed in Figure 11. The
H/D exchange phenomenon implies that the methine
carbon of the oxadiazole ring rapidly undergoes hy-
1
7
et al. (Fig. 2). The concerted mechanism initiated by
proton abstraction by the base would be expected to
result only in reduction in degradation rate as the pH
is lowered and is not consistent with the U-shaped
rate—pH profile shown in Figure 3. The profile is
more consistent with our proposed mechanism where
the nucleophilic attack on the methine carbon (which
is both base and acid catalyzed) is the rate-limiting
step.
During the NMR studies, it was found that the me-
thine proton on the oxadiazole ring is exchangeable
with deuterium under neutral or basic conditions as
evidenced by the disappearance of the doublet peak
2
3
bridization changes from sp to sp like and back to
2
sp like. Although the H/D exchange occurred rapidly
through free base form, it would not likely proceed
under acidic conditions as the protonated N-4 bear-
ing a positive charge would become a highly energetic
doubly charged intermediate. Figure 10c illustrates
that the H/D exchange activity slows down dramat-
ically at pD ∼1, indicating that the protonated N-4
atom hurdle the H/D exchange. Figure 10d revealed
3
the H/D exchange was almost complete in CD CN/
D O without the presence of deuterium chloride.
2
At a pH range of 3–5, the protonation of the N-4
(
δ-9.05, JHC = 230 Hz). Figure 10b showed that the
would be minimal. In addition, the concentration of
the nucleophile is too low, as compared with that at
methine proton almost completely exchanged with
JOURNAL OF PHARMACEUTICAL SCIENCES
DOI 10.1002/jps

DEGRADATION MECHANISM OF AN OXADIAZOLE DERIVATIVE, BMS-708163
7
1
Figure 6. Overlay of partial H NMR spectra (a) compound F + benzylamine in CD
3
CN at
◦
time 0, (b) compound F + benzylamine in CD
3
CN at 60 C after 6 h, no degradation observed,
(
(
c) compound F + benzylamine in CD
3
CN/D
2
O at RT overnight, degradation proceeded, and
◦
d) compound F + benzylamine in CD
3
CN/D
2
O at 60 C after 6.5 h, degradation proceeded
significantly.
Figure 7. The reaction of compound A with benzylamine to form compound B and benzylurea.
13
Figure 8. Proposed mechanism for compound F ( C-labeled compound A) degradation process
in the presence of benzylamine and water.
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES

8
HARTLEY ET AL.
Figure 9. Proposed mechanism for compound A degradation under acidic conditions.
pH > 6, to produce the unstable intermediate that
leads to ring opening. As a result, the degradation
of compound A was the slowest. It was clear that at
higher pH, the degradation of compound A was fa-
cilitated by a strong nucleophile in conjunction with
a proton donor, and at lower pH, the protonation of
the N-4 activated the methine carbon for nucleophilic
attack to form a similar unstable intermediate.
Additional solution stability studies on compound
A in the pH range of 3–5 in acetate buffers were car-
ried out to investigate the degradation process in this
relatively stable pH range. The degradation kinetics
were compared under the same concentration at pH
in acetate buffer at this pH range. One is nucleophilic
attack by the acetate ion, and the other is N-4 proto-
nation. The nucleophilic attack is more predominant
at pH 5 (higher concentration of carboxylate) than at
pH 3.8. On the contrary, the protonation at N-4 is rel-
atively favored at pH 3.8 than at pH 5. Overall, the
degradation rate was higher at pH 3.8 than pH 5 even
though the actual concentration of the ionized acetate
nucleophile is lower at pH 3.8 than at pH 5.
Stability of Drug Product
Although HPMC-AS was selected as the polymer for
solid dispersion of compound A based on its superior
bio-performance, the stability of compound A in solid
dispersions with HPMC and PVP–VA was studied
3
.8 and pH 5 acetate buffers, respectively. The ob-
served first-order rate constant was faster at pH 3.8.
There are two mechanisms operating simultaneously
1
Figure 10. Overlay of partial H NMR spectra (a) compound F + benzylamine in CD
3
CN at
◦
6
0 C for 6 h, (b) compound F + benzylamine in CD
3
CN/D
2
O at RT overnight, (c) Compound
◦
F in CD
3
CN/D
2
O/DCl (pD-1) at RT overnight, then at 60 C for 18 h, and (d) Compound F in
◦
CD CN/D
3
2
O at RT overnight, then at 60 C for 18 h.
JOURNAL OF PHARMACEUTICAL SCIENCES
DOI 10.1002/jps

DEGRADATION MECHANISM OF AN OXADIAZOLE DERIVATIVE, BMS-708163
9
Figure 11. The H/D exchange of the methine proton on oxadiazole ring in basic and acidic
condition.
to help understand the degradation in the dosage
form. The microenvironmental pH values for HPMC-
AS-, PVP–VA-, and HPMC-based solid dispersion es-
timated using a slurry method18 were 4.2, 6.6, and
ing N-4 nitrogen. Overall, understanding the degra-
dation kinetics and mechanism helped designing a
stable drug product.
6
.9, respectively. In the solid dispersion of compound
A and HPMC, there was an increase of 0.8% com-
pound B in the 40 C/75% RH closed condition, and
CONCLUSION
◦
Our studies have shown that nitrogen protonation
and nucleophilic reaction were the key factors in
the degradation process of compound A. The pH-
degradation rate profile showed that compound A was
relatively stable in a pH range from 3 to 5. However,
compound A degraded more rapidly outside of pH
◦
0.5% compound B in 25 C/60% RH open condition af-
ter 4 weeks. Under similar conditions, solid disper-
sion of compound A and HPMC-AS showed relatively
lower degradant level (0.4% and 0.2% compound B in
◦
◦
the 40 C/75% RH closed and 25 C/60% RH open con-
ditions, respectively, after 4 weeks). These data agree
with the estimated microenvironmental pH for the
two solid dispersions. However, there was no degrada-
tion found in the solid dispersion of compound A and
PVP–VA, although the moisture uptake in PVP–VA-
based solid dispersion was the highest among all the
formulations. This finding suggests that in the solid
dispersion, available moisture amount alone is not
critical and the presence/absence of a nucleophilic
group (such as–OH) in the polymer chain in com-
bination with moisture plays a role in the degrada-
tion. This is in accordance with our proposed mecha-
nism (Fig. 8), that is, the combination of nucleophilic
groups contained within cellulose, such as–OH, at-
tacks the methine carbon and moisture protonating
the N-4 nitrogen, leading to compound A degrada-
tion. It is worth noting that all three solid dispersions
remained amorphous and no phase separation during
the period of the stability study was observed. The mi-
croenvironmental pH of compound A and HPMC-AS
solid dispersion is around 4, which places the drug
in the most stable pH range, and the desiccant and
induction sealed package reduces the risk of protonat-
3
–5. Isotopically labeled and methylated compounds
were employed in conjunction with LC–MS to inves-
tigate the bond cleavage positions on the oxadiazole
ring. Once the cleavage positions were established,
NMR studies verified that compound A did not de-
grade in dry nonnucleophilic solvents, and also helped
elucidate the degradation mechanisms. Nucleophilic
attack and protonation of the N-4 nitrogen are two im-
portant factors to initiate the degradation process. Be-
cause of the understanding of the kinetics and mech-
anism of degradation, the nature of the excipients in
terms of their nucleophilic group, microenvironmen-
tal pH, and protection from moisture were all found
to be critical in developing a stable drug product.
ACKNOWLEDGMENTS
The authors acknowledge Dr. Peter Timmins, Dr.
Venkatramana Rao, Dr. Sherif Badway, and Dr. Mark
Bolgar for their helpful discussions and suggestions
in this study and Dr. Wei Ding for identification of
degradant from the hydrolysis of the amide group.
DOI 10.1002/jps
JOURNAL OF PHARMACEUTICAL SCIENCES

1
0
HARTLEY ET AL.
REFERENCES
9. Kharchenko Yu V, Detistov OS, Orlov VD. 2008. Polycyclic sys-
∗
tems containing 1,2,4-oxadiazole ring 2 . 4-(1,2,4-oxadiazol-
1
2
3
4
5
. Jensen LH, Watjen F, Honore T, Hansen JB, Engelstoft M,
Schmiechen R. 1988. Oxadiazolylimidazobenzodiazepines, a
new class of benzodiazepine receptor ligands. Adv Biochem
Psychopharmacol 45:209–217.
5-yl)pyrrolidin-2-ones: Synthesis and prediction of biological
activity. Chem Heterocyclic Compounds 44:600–605.
0. Mulzer J, Zuhse R, Schmiechen R. 1992. Enantioselective syn-
thesis of the antidepressant rolipram by michael addition to a
nitroolefin. Angew ChemInt Ed 31:870–872.
1. Walker C, Boer C, Mazzoni L, Bacher E, Keller TH, Demnitz
Joachim FW, Bray-French K, M u¨ ller T. 1998. N-Arylrolipram
derivatives as potent and selective PDE4 inhibitors. Bioorg
Med Chem Lett 8:3229–3234.
1
1
. Tully WR, Gardner CR, Gillespie RJ, Westwood RJ. 1991. 2-
(
Oxadiazolyl)- and 2-(thiazolyl)imidazo[1,2-a]pyrimidines as
agonists and inverse agonists at benzodiazepine receptors. J
Med Chem 34:2060–2067.
. Watjen F, Baker R, Engelstoft M, Herbert R, MacLeod A,
Knight A, Merchant K, Moseley J, Saunders J. 1989. Novel
benzodiazepine receptor partial agonists: Oxadiazolylimida-
zobenzodiazepines. J Med Chem 32:2282–2291.
. Guimaraes CRW, Boger DL, Jorgensen WL. 2005. Elucida-
tion of fatty acid amide hydrolase inhibition by potent "-
ketoheterocycle derivatives from monte carlo simulations. J
Am Chem Soc 127:17377–17384.
. Diana GD, Rudewicz P, Pevear DC, Nitz TJ, Aldous
SC, Aldous DJ, Robinson DT, Draper T, Dutko FJ. 1995. Pi-
cornavirus inhibitors: Trifluoromethyl substitution provides
a global protective effect against hepatic metabolism. J Med
Chem 38:1355–1371.
1
2. Keller TH, Bray-French K, Joachim Demnitz FW, M u¨ ller T,
Pombo-Villar E., Walker C. 2001. Synthesis and structure–
activity relationship of N-arylrolipram derivatives as in-
hibitors of PDE4 isozymes. Chem Pharm Bull 49:1009–1017.
3. Cotter JL, Knight GJ, Wright WW. 1967. Thermal degradation
studies on oxadiazoles. J Gas Chromatogr 5:86.
1
1
4. Cotter JL, Knight GJ. 1966. Pyrolysis of 3,5-diphenyl-1,2,4-
oxadiazole and 2,5-diphenyl-1,3,4-oxadiazole. Chem Commun
(London) 11:336–337.
15. Citron M. 2010. Alzheimer’s disease: Strategies for disease
modification. Nat Rev Drug Discov 9:387–398.
1
6. Gillman KW, Starrett JE Jr., Parker MF, Xie K, Bronson JJ,
Marcin LR, McElhone KE, Bergstorm CP, Mate RA, Williams
R, Meredith JE Jr., Burton CR, Barten DM, Toyn JH, Roberts
SB, Lentz KA, Houston JG, Zaczek R, Albright CF, Decic-
cot CP, MacOr JE, Olson RE. 2010. Discovery and evalua-
tion of BMS-708163, a potent, selective and orally bioavailable
6
. Suzuki T, Iwaoka K, Imanishi N, Nagakura Y, Miyata K,
Nakahara H, Ohta M, Mase T. 1999. Synthesis of the selec-
tive 5-hydroxytryptamine 4 (5-HT4) receptor agonist (+)-(S)-
2
-chloro-5-methoxy-4-[5-(2-piperidylmethyl)-1, 2, 4-oxadiazol-
-yl]aniline. Chem Pharm Bull 47:120–122.
3
7
. Biftu T, Liang GB, Feng DD. 2000. Synthesis and SAR of ben-
zyl and phenoxymethylene oxadiazole benzenesulfonamides
as selective beta3 adrenergic receptor agonist antiobesity
agents. Bioorg Med Chem Lett 10:1431–1434.
(-secretase inhibitor. Med Chem Lett 10:120–124.
1
7. Claisse JA, Foxton MW, Gregory G1, Sheppard AH, Tiley
EP, Warburton WK, Wilson MJ. 1973. Some 5-unsubstituted
acetylenic and vinylic 1,2,4-oxadiazoles. J Chem Soc Perkin
Trans I 45:2241–2249.
8. Antunes R, Batista H, Srivastava RM, Thomas G, Araujo CC.
1
998. New phthalimide derivatives with potent analgesic ac-
18. Badawy SIF, Hussain MA. 2007. Microenvironmental pH mod-
ulation in solid dosage forms. J Pharm Sci 96:948–959.
tivity: II. Bioorg Med Chem Lett 8:3071–3076.
JOURNAL OF PHARMACEUTICAL SCIENCES
DOI 10.1002/jps