Photodegradation products of propranolol: The structures and pharmacological studies

Article abstract of DOI:10.1016/j.lfs.2005.04.033

Recently, single-dose drug packaging systems, allowing the administration of multiple drugs in a single pill, have become popular for the convenience of the patient. The quality of drugs and an accurate measurement of their photostabilities within this system, however, have not been carefully addressed. Drugs that are unstable in light should be carefully handled to protect their potency and ensure their safety. Propranolol (1), a β-adrenergic receptor antagonist, is widely used for angina pectoris, arrhythmia, and hypertension. Due to its naphthalene skeleton, this drug may be both light unstable and a photosensitizing agent. In this study, we isolated three photodegraded products of propranolol (1): 1-naphthol (2), N-acetylpropranolol (3), and N-formylpropranolol (4). The structures of these compounds were determined by spectroscopic methods and chemical syntheses. We also examined the acute toxicities of these substances in mice and their binding to β-adrenergic receptors using rat cerebellum cortex membranes. Although the photoproducts isolated in this study did not exhibit any acute toxicity or significant binding to β-adrenergic receptors, these results serve as a warning to single-dose packaging systems, as propranolol (1) must be handled carefully to protect the compound from light-induced degradation.

Full text of DOI:10.1016/j.lfs.2005.04.033

Life Sciences 78 (2005) 357 – 365
www.elsevier.com/locate/lifescie
Photodegradation products of propranolol: The structures and
pharmacological studies
a,
Koji Uwai , Marie Tani , Yosuke Ohtake , Shinya Abe , Akiko Maruko , Takashi Chiba ,
a
a
a
a
b
*
b
Yoshiro Hamaya , Yasuhito Ohkubo , Mitsuhiro Takeshita
a
a,
*
a
Tohoku Pharmaceutical University, 4-4-1 Komatsushima, Aoba-ku, Sendai 981-8558, Japan
b
Department of Pharmacy, Sendai Social Insurance Hospital, 3-16-1 Tsutsumi-cho, Aoba-ku, Sendai 981-8501, Japan
Received 2 November 2004; accepted 25 April 2005
Abstract
Recently, single-dose drug packaging systems, allowing the administration of multiple drugs in a single pill, have become popular for the
convenience of the patient. The quality of drugs and an accurate measurement of their photostabilities within this system, however, have not been
carefully addressed. Drugs that are unstable in light should be carefully handled to protect their potency and ensure their safety. Propranolol (1), a
h-adrenergic receptor antagonist, is widely used for angina pectoris, arrhythmia, and hypertension. Due to its naphthalene skeleton, this drug may
be both light unstable and a photosensitizing agent. In this study, we isolated three photodegraded products of propranolol (1): 1-naphthol (2), N-
acetylpropranolol (3), and N-formylpropranolol (4). The structures of these compounds were determined by spectroscopic methods and chemical
syntheses. We also examined the acute toxicities of these substances in mice and their binding to h-adrenergic receptors using rat cerebellum
cortex membranes. Although the photoproducts isolated in this study did not exhibit any acute toxicity or significant binding to h-adrenergic
receptors, these results serve as a warning to single-dose packaging systems, as propranolol (1) must be handled carefully to protect the compound
from light-induced degradation.
D 2005 Elsevier Inc. All rights reserved.
Keywords: Propranolol; Photodegradation; h-adrenergic receptor; Receptor binding
Introduction
1993), and lomefloxacin (Matsumoto et al., 1992), etc.), are
known to be unstable under light.
Drug stability research is critical in pharmaceutical studies
as the increased degradation decreases the potency of the drug
and can create compounds with undesirable pharmacological
effects (Andrisano et al., 1999). Moisture, temperature, and
light can all alter drug quality. Currently, several classes of
drugs, including dihydropyridines (nifedipine (Berson and
Brown, 1995; Grundy et al., 1994; Pietta et al., 1981;
Shamsipur et al., 2003; Suzuki et al., 1985), nisoldipine
In Japan, drug photostabilities are typically examined
according to the FDrug Approval and Licensing Procedures
in Japan_. If the drug does not meet the criteria outlined in
these procedures, then a form of photoprotection media must
be provided on the surface of the drug or its packaging to
guarantee photostability. A single-dose packaging system, in
which multiple drugs are administered in one dose, has become
increasingly popular for the added convenience of patients.
While this system facilitates patient compliance with the proper
dosages of multiple drugs, neither the quality of drugs in this
format or the resultant photostability has been adequately
tested. Therefore, drugs that are unstable towards light should
be carefully handled to protect their potency and safety.
Propranolol (1, Fig. 1) is a h-adrenergic antagonist that is
widely used for the treatment of angina pectoris, arrhythmia,
and the hypertension (Hoffman and Lefkowitz, 1996). The
(
Marinkovic et al., 2003), nitrendipine (Marciniec and Ogro-
dowczyk, 2003) etc.), azulene sulfonate (Comtet and Mettee,
970; Olmsted, 1969), and new quinolones (norfloxacin
Cordoba-Borrego et al., 1999), levofloxacin (Yoshida et al.,
1
(
*
Corresponding authors. Tel.: +81 22 234 4181; fax: +81 22 275 2013.
E-mail address: uwai@tohoku-pharm.ac.jp (K. Uwai).
0024-3205/$ - see front matter D 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.lfs.2005.04.033

358
K. Uwai et al. / Life Sciences 78 (2005) 357–365
exclude the influence of moisture or oxygen on propranolol
1) stability, both the tablet and ground forms of the drug on
(
Petri dishes without paraffin paper were kept in a closed
cabinet in the same room for the same period. During the
experimental period, the production of photodegradtion
products was monitored by TLC analyses (CHCl –MeOH
3
Fig. 1. Structure of propranolol (1).
(22:3 v/v)).
Isolation and characterization of photoproducts
Propranolol hydrochloride (2.0 g, 6.76 mmol) was treated
with an aqueous solution of 10% NH OH, then extracted with
usual dose, which ranges from 30 to 60 mg daily, can be in
creased to 120 mg daily. In many Japanese hospitals, both the
prescription and single-dose packaging forms of propranolol
4
(
1) are dispensed in small packages or are ground to adjust the
CHCl . The organic phase was concentrated in vacuo. The
3
dose and/or aid in swallowing. The photostability of these
various pharmaceutical forms, including the naked form, has
not been reported. Propranolol (1), however, may be both a
photosensitizing agent and light-unstable, similar to other drugs
that have chromophoric structures containing a naphthalene
skeleton.
resulting solid was recrystalized with cyclohexane to produce
propranolol (1, 1.70 g, 6.53 mmol, 97%).
Propranolol (1, 1.70 g, 6.53 mmol) was irradiated with
95,000 lx (40 cm far light from a halogen lamp (National
JD110V250W/E)) under an Ar atmosphere. After five days
of irradiation, column chromatography of a portion of the
resulting solid (71.1 mg) on silica gel (10 g) [n-hexane-
AcOEt (9:1, 17:3, 7:3, 3:2, 1:1, 2:3 v/v), and MeOH]
yielded five fractions (fraction A to E). Fraction A, eluted
in n-hexane-AcOEt (9:1 v/v), exhibited a single spot on
TLC, which was confirmed to be 1-naphthol (2, 2.1 mg,
6.1%). Fraction D (25.2 mg), eluted in n-hexane-AcOEt (3:2
to 2:3 v/v), was subjected to alumina column chromatog-
raphy using n-hexane-AcOEt (13:7, 3:2, 11:9, 1:1, 2:3, 3:7)
as the eluting solvent, identifying both N-acetylpropranolol
(3, 2.0 mg, 2.8%) and N-formylpropranolol (4, 16.5 mg,
In this study, we examined the structures of the decom-
position products of propranolol (1), determining the LD50 for
each product in mice. In addition, we measured the potency of
3
each compound to inhibit the specific binding of [ H]-
3
dihydroalprenolol hydrochloride ([ H]DHA) to h-adrenergic
receptors.
Materials and methods
General
24%). Propranolol (1) was also recovered in this fraction
(14.9 mg, 21%).
All solvents were solvent grade. Wako gel C-200 (70–
1
50 Am, Wako pure chemicals) and Alumina activated 200
abt. 200 mesh, nacalai tesque) were used for column
N-[2-hydroxy-3-(1-naphthalenyloxy)propyl]-N-(1-methyle-
thyl)acetamide (N-acetylpropranolol (3)): While the properties
of this compound has been reported in previous papers (Chiou
et al., 1997; Dewar et al., 1982; Nelson and Walker, 1978), we
(
chromatography. Precoated Kieselgel 60 F254 plate (0.25
mm, Merck) and precoated aluminium oxide 60 F254 neutral
plate (0.2 mm, Merck) were used for TLC (thin layer
chromatography) analysis and the spots were detected by the
absorbance of ultra violet (UV) light at 254 nm spraying
1
include the spectral data here. Colorless oil. H-NMR (600
MHz, CDCl ) d: 1.22 (d, 3H, J =6.6 Hz, –CH(CH ) ), 1.36 (d,
3
3 2
3H, J =6.6 Hz, –CH(CH ) ), 2.27 (s, 3H, –C(=O)–CH ),
3
2
3
1
13
with Dragendorff’s reagent. H and C NMR (nuclear
magnetic resonance) spectra were measured using JEOL
JNM 400 (400 MHz), and JEOL JNM 600 (600 MHz)
spectrometers. Chemical shifts (d) are reported as ppm
downfield from tetramethylsilane (TMS), and coupling
constants are given in Hz. Multiplicity is indicated as
follows: s (singlet), d (doublet), dd (doublet of doublets), t
3.54 (dd, 1H, J =1.5, 14.6 Hz, CH(OH)–CH –N), 3.75 (dd,
2
1H, J =8.4, 14.6 Hz, CH(OH)–CH –N), 4.04 (t, 1H, J =8.4
2
Hz, O–CH –CH(OH)), 4.10 (quint., 1H, J =6.6 Hz, N–
2
CH(CH ) ), 4.18–4.23 (m, 1H, CH –CH(OH)–CH ), 4.24
3
2
2
2
(dd, 1H, J =4.0, 8.4 Hz, O–CH –CH(OH)), 5.67 (s, 1H, –
2
OH), 6.86 (d, 1H, J =7.7 Hz, aromH ), 7.38 (t, 1H, J =7.7 Hz,
2
aromH ), 7.44–7.49 (m, 3H, aromH , aromH , aromH ), 7.81
6
3
4
7
(
triplet), q (quintet), and m (multiplet), etc. Mass spectra
(d, 1H, J =7.3 Hz, aromH ), 8.21 (d, 1H, J =8.1 Hz, aromH ).
5 8
1
3
were recorded on JEOL JMS DX-303 and JMA-DA 5000
spectrometers.
C-NMR (150 MHz, CDCl ) d: 20.63 (q, –CH(CH ) ), 21.36
3
3 2
(q, –CH(CH ) ), 21.87 (q, –C(=O)–CH ), 46.14 (t,
CH(OH)–CH –N), 50.27 (d, N–CH(CH ) ), 69.81 (t, O–
3
2
3
2
3 2
Photostability test of propranolol
CH –CH(OH)), 72.40 (d, CH –CH(OH)–CH ), 104.84 (d,
2 2 2
aromC ), 120.65 (d, aromC ), 121.55 (d, aromC ), 125.24 (d,
2
4
8
Pharmaceutical preparations of propranolol in variety of
packages and dosages (A: whole tablet covered with
aluminum foil (control); B: whole tablet in a press-through
packaging; C: whole tablet in paraffin paper; D: half tablet in
paraffin paper; E: ground form in paraffin paper) were kept
under scattered light (av. 8400 lx at 1:00 pm) for 28 days. To
aromC ), 125.40 (s, C ), 125.95 (d, C ), 126.39 (d, C ),
3 8a 7 6
127.66 (d, C ), 134.52 (s, C ), 154.07 (s, C ), 173.86 (s, –
4a
5
1
+
C(=O)–CH ). MS m/z: 301 (M ), 256 (M -Ac-2), 158 (M -
+
+
3
+
naphthalene oxide, 100%), 116 (M -naphthalene oxide-Ac), 43
+
(Ac). High-resolution MS calcd for C H O N (M ):
301.1672. Found: 301.1678.
1
8
23
3

K. Uwai et al. / Life Sciences 78 (2005) 357–365
359
1
NMR (100 MHz, CDCl ). As the C-NMR spectra in the
3
N-[2-hydroxy-3-(1-naphthalenyloxy)propyl]-N-(1-methyle-
1
thyl)formamide (N-formylpropranolol (4)): Colorless oil. H-
3
region of aromatic carbons, we could not assign all of the
component forms. d: 20.21 (q, minor), 20.65 (q, minor), 20.98
(q, major and minor), 21.14 (q, major), 21.73 (q, major), 22.12
NMR (600 MHz, CDCl ) d: 1.25 (d, 3H, J =7.0 Hz, –
3
CH(CH ) ), 1.33 (d, 3H, J =7.0 Hz, –CH(CH ) ), 3.59 (dd,
3
2
3 2
1
1
H, J =2.2, 14.3 Hz, CH(OH)–CH –N), 3.65 (dd, 1H, J =7.7,
2
(q, major), 22.76 (q, minor), 41.42 (t, CH(OAc)–CH –N,
2
4.3 Hz, CH(OH)–CH –N), 3.87 (sept., 1H, J =6.6 Hz, N–
2
major), 45.52 (t, CH(OAc)–CH –N, minor), 47.25 (d, N–
2
CH(CH ) ), 4.06 (dd, 1H, J =7.7, 9.5 Hz, O–CH –CH(OH)),
3
CH(CH ) , minor), 49.52 (d, N–CH(CH ) , major), 66.95 (t,
3 2 3 2
2
2
4
.21 (dd, 1H, J =4.8, 9.5 Hz, O–CH –CH(OH)), 4.25–4.28
2
O–CH –CH(OAc), minor), 68.43 (t, O–CH –CH(OAc),
2 2
(
m, 1H, CH –CH(OH)–CH ), 4.90 (d, 1H, J =3.3 Hz, –OH),
2
major), 70.97 (d, CH –CH(OAc)–CH , minor), 71.64
2 2
2
6.85 (d, 1H, J =7.7 Hz, aromH ), 7.38 (t, 1H, J =7.7 Hz,
aromH ), 7.45 (d, 1H, J =7.7 Hz, aromH ), 7.47–7.52 (m, 2H,
(d, CH –CH(OAc)–CH , major), 104.83 (d, minor), 104.93
2 2
2
(d, major), 120.61 (d, major), 121.22 (d, minor), 121.44 (d,
minor), 121.86 (d, major), 125.27 (d, major), 125.38 (s, minor),
125.57 (s, major), 125.71 (d, minor), 125.85 (d, major), 126.38
(d, major), 126.62 (d, minor), 127.46 (d, major), 127.67 (d,
minor), 128.32 (d, minor), 134.48 (s, major), 134.55 (s, minor),
153.74 (s, minor), 154.28 (s, major), 170.19 (s, minor), 170.52
3
4
aromH , aromH ), 7.80 (dd, 1H, J =7.4, 1.8 Hz, aromH ), 8.20
5
6
7
(
1
dd, 1H, J =7.7, 1.5 Hz, aromH ), 8.26 (s, 1H, –C(=O)–H).
8
3
C–NMR (150 MHz, CDCl ) d: 20.80 (q, –CH(CH ) ),
3
3 2
2
N–CH(CH ) ), 69.52 (t, O–CH –CH(OH)), 71.34 (d, CH –
1.82 (q, –CH(CH ) ), 46.09 (t, CH(OH)–CH –N), 51.14 (d,
3 2 2
3
2
2
2
+
CH(OH)–CH ), 104.91 (d, aromC ), 120.78 (d, aromC ),
2
(s, major), 171.28 (s, major and minor). MS m/z: 343 (M ),
+
2
4
+
1
1
1
21.49 (d, aromC ), 125.31 (d, aromC ), 125.37 (s, C ),
8
200 (100%), 158 (M -naphthalene oxide-Ac), 115 (M -
naphthalene oxide-2Ac), 43 (Ac). High-resolution MS calcd
3
8a
25.92 (d, C ), 126.43 (d, C ), 127.66 (d, C ), 134.52 (d, C ),
7
6
5
4a
+
53.94 (s, C ), 165.03 (s, –C(=O)–H). MS m/z: 287 (M ), 256
+
for C H O N (M ): 343.1783. Found: 343.1779.
1
20 25 4
+
M -Ac-2), 144 (M -naphthalene oxide, 100%), 115, 100, 58.
+
(
High-resolution MS calcd for C H O N (M ): 287.1516.
N-[2-hydroxy-3-(1-naphthalenyloxy)propyl]-N-(1-methyle-
thyl)acetamide (N-acetylpropranolol (3)
A mixture of compound 5 (2.1 g, 6.0 mmol) and K CO
+
1
7 21 3
Found: 287.1519. All spectral data are consistent with
previously reported results (Chen et al., 1993).
2
3
(2.5 g, 18.1 mmol) in MeOH (60 ml) was stirred for 2 h at
room temperature. The resulting solution was concentrated in
vacuo. Purification by column chromatography on silica gel
Synthesis of photoproducts
(CHCl ) gave a title compound 3 (1.4 g, 78%) as a colorless
3
N-[2-(acetyloxy)-3-(1-naphthalenyloxy)propyl]-N-(1-meth-
ylethyl)acetamide (N,O-diacetylpropranolol (5))
A mixture of propranolol hydrochloride (3.0 g, 10.1 mmol),
oil. All spectral data were agreed with those recorded for the
photoproduct.
N-[2-hydroxy-3-(1-naphthalenyloxy)propyl]-N-(1-methyle-
thyl)formamide (N-formylpropranolol (4))
triethylamine (Et N) (7.1 ml, 51.0 mmol), and acetic anhydrate
3
(Ac O) (2.9 ml, 30 mmol) in CHCl3 was stirred at room
temperature for 20 h. The resulting solution was washed
N-[2-(formyloxy)-3-(1-naphthalenyloxy)propyl]-N-(1-meth-
ylethyl)formamide (N,O-diformylpropranolol (6))
To a solution of propranolol (2.7 g, 10.4 mmol) in formic
2
sequentially with a saturated solution of aqueous NH Cl, a
4
saturated solution of aqueous NaHCO , and brine. Samples
3
acid (26 ml) was added Ac O (8.7 ml) dropwise at 55 -C and
2
were then dried over MgSO4 and concentrated in vacuo.
Purification by column chromatography on silica gel (CHCl3)
gave a solid. After recrystalization in benzene-n-hexane,
compound 5 (2.99 g, 87%) was obtained as a colorless crystal.
While this compound has already been reported in previous
reports (Chen et al., 1993; Nelson and Walker, 1978), we could
not found its detailed spectral data; thus, we detail the spectra
stirred for 24 h. After the addition of 10 ml of ice/water , the
solution was concentrated in vacuo. Purification by column
chromatography on silica gel (AcOEt-n-hexane, 1:3 v/v) gave
title compound 4 (318 mg, 11%) and title compound 6 (118
mg, 3.7%) as colorless oils. Compound 4: All spectral data
were agreed with those observed for the photoproduct.
Compound 6: While this compound has been previously
described (Chen et al., 1993; Chen, 1994), we could not find
any description of the spectral data, thus, we detail the results
here. This compound was obtained as a 3:1 mixture of
1
unseparable rotamer or isomer. H-NMR (400 MHz, CDCl )
3
d: 1.21–1.26 (m, 3.75H, –CH(CH ) ), 1.30 (d, 2.25H, J =6.8
here. This compound was obtained as a 10:3 mixture of an
1
3
2
Hz, –CH(CH ) ), 2.08 (s, 2.25H, –C(=O)–CH ), 2.11 (s,
0
unseparable rotamer or isomer. H-NMR (400 MHz, CDCl ) d:
3
2
3
3
.75H, –C(=O)–CH ), 2.15 (s, 2.25H, –C(=O)–CH ), 2.23
3
1.27 (d, 0.69H, J =6.8 Hz, –CH(CH ) ), 1.28 (d, 0.69H,
J =6.8 Hz, –CH(CH ) ), 1.31 (d, 2.31H, J =6.9 Hz, –CH
3
3 2
(
CH(OAc)–CH –N), 3.64 (dd, 0.25H, J =3.2, 15.9 Hz,
s, 0.75H, –C(=O)–CH ), 3.41 (dd, 0.75H, J =6.8, 14.4 Hz,
3
3 2
(CH ) ), 1.35 (d, 2.31H, J =6.8 Hz, –CH(CH ) ), 3.59 (dd,
3 2 3 2
2
CH(OAc) –CH – N), 3.82 (dd, 1H, J = 5.4, 14.4 Hz,
2
0.77H, J =7.0, 14.4 Hz, –CH(O-formyl)–CH –N), 3.65 (dd,
2
CH(OAc)–CH –N), 4.01–4.07 (m, 0.75H, N–CH(CH ) ),
3 2
0.23H, J =8.3, 15.4 Hz, –CH(O-formyl)-CH -N), 3.72–3.86
2
2
4.23–4.33 (m, 2H, O–CH –CH(OAc)), 4.46–4.53 (m, 0.25H,
N–CH(CH ) ), 5.45–5.51 (m, 0.25H, CH –CH(OAc)–CH ),
(m, 1.23H, –CH(O-formyl)–CH –N, N–CH(CH ) ), 3.79
2 3 2
2
(dd, 0.77H, J =5.4, 14.4 Hz, –CH(O-formyl)–CH –N), 4.27–
2
3
2
2
2
5.57–5.63 (m, 0.75H, CH –CH(OAc)–CH ), 6.80 (d, 1H,
J =7.6 Hz, aromH ), 7.35 (t, 1H, J =8.3 Hz, aromH ), 7.42 (d,
4.33 (m, 1.77H, O–CH –CH(O-formyl)), 4.46 (sept., 0.23H,
2
2
2
J =6.6 Hz, O–CH –CH(O-formyl)), 5.51–5.59 (m, 0.23H,
2
2
3
1
H, J =8.3 Hz, aromH ), 7.45–7.57 (m, 2H, aromH , aromH ),
4
CH –CH(O-formyl)–CH ), 5.69–5.78 (m, 0.77H, CH –
2
6
7
2
2
1
.77–7.82 (m, 1H, aromH ), 8.18–8.23 (m, 1H, aromH ). C-
3
7
CH(O-formyl)–CH ), 6.79 (d, 1H, J =7.6 Hz, aromH ), 7.36
2 2
5
8

360
K. Uwai et al. / Life Sciences 78 (2005) 357–365
Table 1
groups of 10 per cage (30Â30Â16 cm) in an air conditioned
room (ambient temperature 22T2 -C and 55T5% relative
humidity) under a 12 h light cycle. Animals were allowed to
ingest food (F-2 obtained from Funabashi Farm Co., Funaba-
shi, Japan) and water ad libitum. Samples, suspended in
physiological saline containing 2% arabia gum, were admin-
istered intraperitoneally (10 ml/kg body weight). Ten mice
comprised a group that received the same dose. LD50 values
were estimated according to the Litchfield-Wilcoxon method
(Litchfield and Wilcoxon, 1949).
Color change observations for propranolol tablets in various forms under
a
natural light
Form
Days
3
7
10
14
21
28
A
B
C
D
E
a
À
À
À
À
À
À
À
À
À
À
À
À
À
À
b
c
d
d
T
T
+
++
++
+++
+++
+++
+++
+++
+++
+++
b
c
+
b
T
c
d
+
++
Pharmaceutical preparations of propranolol in various storage conditions
and dosages (A: whole tablet covered with aluminum foil (control); B: whole
tablet in press-through packaging; C: whole tablet in paraffin paper; D: half
tablet in paraffin paper; E: ground form in paraffin paper) were kept under the
scattered light (av. 8400 lx at 1:00 pm) in the room for 28 days.
Receptor-binding assay
Assays measuring the receptor binding of the propranolol
derivatives were performed as previously reported (Malinow-
ska et al., 2003) with the following minor modification.
Cerebral cortex membranes from male Wistar rats (180–200
g) were homogenized using a Potter homogenizer (10 strokes,
1100 rpm) in 25 volumes of ice-cold Tris–HCl buffer (10 mM
Tris, pH 7.5; 0.25 M sucrose; 2 mM EGTA; 2 mM MgCl2).
Homogenates were then centrifuged at 1000Âg for 10 min at 4
-C. Supernatants were centrifuged at 35,000Âg for 20 min at 4
-C, then re-centrifuged (35,000Âg for 20 min at 4 -C) twice in
15 ml of Tris–HCl buffer (50 mM Tris, pH 7.5; 5 mM EDTA).
The resulting pellet was resuspended in Tris–HCl buffer and
stored at À80 -C until use.
b
Slightly colored.
c
Browned.
d
Blackened.
(
t, 0.77H, J =8.3 Hz, aromH ), 7.35–7.39 (m, 0.23H, aromH ),
3 3
7.44–7.52 (m, 3H, aromH , aromH , aromH ), 7.80 (d, 0.77H,
4 6 7
J =9.3, aromH ), 7.82–7.90 (m, 0.23H, aromH ), 8.14–8.21
5
5
(
H). C-NMR (100 MHz, CDCl ) As the C-NMR spectrum
m, 2.23H, aromH , –C(=O)–H), 8.28 (s, 0.77H, –C(=O)–
8
1
3
13
3
of isomers of this compound was overlapped in the region of
aromatic carbons, we could not assign all of the component
forms. d: 20.23 (q, minor), 20.37 (q, minor), 22.16 (q, major),
22.36 (q, major), 41.81 (t, CH(OH)–CH –N, major), 44.56 (t,
2
CH(OH)–CH –N, minor), 45.02 (d, N-CH(CH ) , minor),
Ligand binding assays were performed to determine
receptor specificity. A solution containing of 0.1 ml of 0.4
2
3 2
50.65 (d, N-CH(CH ) , major), 66.31 (t, O–CH –CH(O-
3 2 2
formyl), minor), 67.77 (t, O–CH –CH(O-formyl), major),
3
mg/ml cerebral cortex membrane, 10 Al 8 nM [ H]DHA, and 5
2
70.34 (d, CH –CH(O-formyl)–CH , minor), 70.61 (d, CH –
CH(O-formyl)–CH , major), 104.88 (d), 121.01 (d), 121.43
AL of varying concentrationsof propranolol (1) or its deriva-
tives (2–6) (eight concentrations ranging from 1.0 pM to 10
AM). The mixture was incubated for 30 min at 30 -C. To
terminate the reaction, 0.5 ml of ice-cold Tris–HCl buffer was
added. The solution was then filtrated through 0.3% poly-
ethyleneimine-pretreated Whatman GF/C filters. After washing
three times with 2 ml ice-cold Tris–HCl buffer, the filters were
transferred into a vial. Three milliliters crea-sol I (Nacalai
Tesque, Japan) was added to a vial as a scintillator; the
radioactivity was measured with a liquid scintillation counter
(LS7800, Beckman). To determine non-specific binding, 10
2
2
2
2
(
d), 121.45 (d), 121.76 (d), 125.45 (d), 125.65 (d), 125.67 (s),
1
1
1
25.75 (s), 126.53 (d), 126.69 (d), 127.53 (d), 127.69 (d),
34.52 (s), 134.58 (s), 153.42 (s, minor), 153.91 (s, major),
60.06 (s, minor), 160.44 (s, major), 163.17 (s, major), 163.82
+
(s, minor). MS m/z: 315 (M ), 172 (100%), 144, 115, 58, 43.
+
High-resolution MS calcd for C H O N (M ): 315.1470.
1
8 21 4
Found: 315.1478.
Acute toxicity test
3
AM of propranolol alone was used (7% for 8 nM [ H]DHA).
Protein concentrations were assayed by the Bradford method
using bovine serum albumin as a standard (Bradford, 1976).
Male mice of ddY strain (21–25 g) were purchased from
Nihon SLC Co. (Hamamatsu, Japan). Mice were housed in
Fig. 2. TLC analyses of the photodegration products of propranolol (1). A: whole tablet covered with aluminum foil (control); B: whole tablet in press-through
packaging; C: whole tablet in paraffin paper; D: half tablet in paraffin paper; E: ground form in paraffin paper. Each TLC was developed with CHCl –MeOH (22:3
v/v). spots colored by Dragendorff’s reagent. >: spots absorbed UV light at 254 nm.
3
&

K. Uwai et al. / Life Sciences 78 (2005) 357–365
361
Results
Photostability test of propranolol
The preparations of propranolol used for photostability
testing, with the exception of the control samples, the
pharmaceutical preparations in press-through packages, and
the samples stored in the dark, were colored as summarized in
Table 1. Whole tablets, half tablets, and ground tablets stored in
paraffin paper tended to turn gradually from yellow to brown,
and finally to black. Within 3 weeks, all drug forms kept in
paraffin paper were black. The rate of discoloration for the
ground tablet was much faster than the other forms; the ground
tablet turned black after only 2 weeks, while the other
pharmaceutical preparations stored in paraffin paper blackened
within 3 weeks.
Fig. 4. Correlations between the acetyl or formyl groups and the propranolol
structure from the HMBC spectrum.
compound 3 exhibited an acetyl group bound to propranolol
(1). The chemical shifts of the methylene and methyne groups
next to the amino group were downshifted relative to pro-
pranolol (1), with methylene shifting from d 2.85 to 3.54 and
H
3.00 to 3.75 and methyne moving from d 2.85 to 4.10. These
H
TLC analyses monitored the production of the photo-
degration products (Fig. 2). Whole tablets covered with an
aluminum foil (condition A) served as a control. These results
demonstrated that irradiation with light for approximately 10
days was sufficient to produce a photoproduct (condition E).
This spot was detected by Dragendorff’s reagent after irradiat-
ing for 21 days. The number of photoproducts (spots) increased
with time; two and three spots were detected after 14 and 28
days, respectively (condition E).
findings suggested that acetylation of the propranolol (1) amino
1
group produced compound 3. Correlation between these H
1
and C NMR signals was unambiguously confirmed by the
3
1
HMQC ( H-detected multiple quantum coherence) spectrum.
1
In addition, the HMBC ( H-detected multiple bond hetero-
nuclear multiple quantum coherence) spectrum of compound 3
clarified the connectivities between the functional groups,
allowing us to propose the structure for N-acetylpropranolol (3)
detailed in Fig. 4.
The structure of compound 4 was elucidated in the same
manner as that performed for compound 3. Briefly, the mass
Structure analysis
+
spectrum of 4 exhibited a molecular ion peak [M] at 287,
1
consistent with a molecular formula of C H O N. The H
We used the free base, a relatively unstable form of
propranolol, for the structure analyses of photoproducs,
because the rate of photoproduct formation from the pharma-
ceutical preparations was too slow to isolate sufficient quantity
of photoproducts for a structural determination. After a 5-day
experiment, the photo-irradiated products derived from pro-
pranolol (1) were repeatedly subjected to column chromatog-
raphy to isolate 1-naphthol (2), N-acetylpropranolol (3), and N-
formylpropranolol (4) in pure forms (Fig. 3). 1-Naphthol (2),
N-acetylpropranolol (3), and N-formylpropranolol (4) corre-
sponded to the first, second, and third spots in the TLC
analyses, respectively (see Fig. 2, lane E of 21 day plate
beginning at the top).
1
7 21 3
1
and C NMR signals of compound 4 displayed a formyl group
3
1
bound to propranolol (1). Correlation between these H and
1
3
C NMR signals was unambiguously confirmed by the
HMQC spectrum. The HMBC spectrum of compound 4
clarified the connectivities of the functional groups, leading
to proposal for the structure for N-formylpropranolol (4). Fig. 4
demonstrated the important correlations in HMBC spectrum.
Chemical syntheses
To generate the compounds in substantial quantity for
pharmacological studies and for use as standards in spectro-
graphic analyses, we synthesized N-acetyl and N-formylpro-
pranolol (3 and 4) according to standard methods (Fig. 5).
Briefly, acetylation of propranolol (1), which gave diacetate
(5), was followed by ester-selective solvolysis to give N-
acetylpropranolol (3). Formylation of propranolol (1) using a
Compound 2 was identified by comparison of the acquired
spectroscopic data with those of known samples.
The mass spectrum of compound 3 exhibited a molecular
+
ion peak [M] at 301 which is consistent with a molecular
1 13
formula of C H O N. The H and C NMR signals for
1
8 23 3
Fig. 3. Structures of photoproducts (1-naphthol (2), N-acetylpropranolol (3), and N-formylpropranolol (4)).

362
K. Uwai et al. / Life Sciences 78 (2005) 357–365
Fig. 5. Chemical syntheses of N-acetyl and N-formylpropranolol (3 and 4).
mixed anhydride yielded both N-formyl- and N,O-diformyl-
propranolol (4 and 6).
affinity to the h-adrenergic receptor in low concentrations
(about 47 pM); the photoproducts (2, 3, and 4) and their
analogues (5 and 6) demonstrated moderate activities. The
greater the masking of the N- or O-functional groups, the lower
the affinity to the h-adrenergic receptor.
Acute toxicity test
The acute toxicities (LD ) of propranolol (1), the photo-
0
5
products of propranolol (2, 3, and 4), and their derivatives (5
and 6) in mice were determined using the Litchfield-Wilcoxon
method (Litchfield and Wilcoxon, 1949) (Table 2). The results
demonstrated that both the photoproducts (2, 3, and 4) and their
derivatives (5 and 6) do not display acute toxicity even at
intraperitoneal injections of 1000 mg/kg. In contrast, propra-
nolol (1) is relatively toxic at these concentrations.
Discussion
Pharmacopoeias directs that propranolol must be protected
from light. In a pharmaceutical interview, commercial manu-
facturers described that the powder gradually turned black, but
assured that the quality of the base drug did not decrease when
propranolol was kept under indoor diffused light at room
temperature.
Receptor-binding assay
We confirmed that every pharmaceutical preparation of
propranolol, with the exception of the control sample kept in
We examined h-adrenergic receptor binding for compounds
1
to 6 using rat cerebellum cortex membranes according to the
procedure outlined by reported method (Malinowska et al.,
003). Binding was evaluated by the competition of these
2
3
compounds with [ H]DHA.
Fig. 6 and Table 3 describes the results. According to the
IC50 values, of the tested propranolol derivatives, propranolol
(
1), the parental compound, exhibited the highest binding
Table 2
a
Acute toxicities of the propranolol derivatives
Sample
LD50 (mg/kg)
Propranolol (1) Hydrochloride
Compound 2
Compound 3
Compound 4
Compound 5
Compound 6
a
Samples, suspended in physiological saline containing 2% arabia gum, were
administered intraperitoneally (10 ml/kg body weight) to male mice of the dd
strain (21–25 g). Ten mice comprised a group receiving the same dose. The
LD50 values were estimated according to the Litchfield-Wilcoxon method.
580
2500
>1000
>1000
>1000
>1000
3
Fig. 6. Inhibition of the specific [ H]DHA binding by propranolol analogues
(1–6) to rat cerebral cortex membrane. Displacement curves were obtained
using a variety of concentrations of propranolol analogues (1–6) in the
3
presence of 8 nM of [ H]DHA. Each point represents the mean of three
duplicate experiments.

K. Uwai et al. / Life Sciences 78 (2005) 357–365
363
Table 3
Several photodegradation products of propranolol have been
previously reported, including naphthalene, 1,4-naphthoqui-
none (Salomies, 1987), and 6-hydroxy-1,4-naphthoquinone
Affinities for the h-adrenergic receptor binding site for propranolol, its
photoproducts and their analogues
a
b
Sample
IC50 (M)
(
Sortino et al., 2002). In this study, we could not detect any
À11
Propranolol (1)
Compound 2
Compound 3
Compound 4
Compound 5
Compound 6
a
4.7Â10
4.8Â10
2.2Â10
2.7Â10
1.6Â10
8.3Â10
of these products even after irradiating the drug with light for
five days. Three other photoproducts were isolated, however,
we used spectroscopic methods to characterize their structures
as 1-naphthol (2), N-acetylpropranolol (3), and N-formylpro-
pranolol (4). These photodegradation products are different
from those reported in other studies, because the drug was
irradiated under different conditions. Salomies oxidized pro-
pranolol (1) with NBS, a versatile oxidizing agent, under both
acidic and neutral conditions concurrent with irradiation under
a mercury lamp. In contrast, Sortino et al. irradiated air-
equilibrated solutions in phosphate buffer using phosphor
lamps emitting light in the 310–390 range. The mechanism
whereby these 1,4-naphthoquinones were produced was
proposed as depicted in Fig. 7. The reaction proceeds via a
À4
À5
À7
À6
À3
Ligand binding assays to determine the receptor specificity were performed
3
using a solution containing cerebral cortex membrane, [ H]DHA, and eight
varying concentrations of samples ranging from 1.0 pM to 10 AM. The mixture
was then incubated for 30 min at 30 -C. After the addition of ice-cold Tris–HCl
buffer to terminate the reaction, the solution was filtrated. Filters were washed
with ice-cold Tris–HCl buffer, then transferred into a vial. Crea-sol I was added
to a vials as a scintillator; the radioactivity was measured with the liquid
scintillation counter.
b
Concentration that gives a half-maximal effect.
1
aluminium foil, the pharmaceutical preparations in press-
through packaging, and the samples maintained in the dark,
was discolored by natural light. TLC analyses confirmed the
presence of degradation products. As the only difference
between these conditions is irradiation with light, the color-
change and formation of degradation products depended on
light, not on temperature or moisture. The appearance and
contents of propranolol tablets kept in the press-through
packaging was unchanged demonstrating that the press-through
packaging had been modified to protect the drug from light.
The rate of degradation of the ground tablets was much faster
than that seen for the other forms, likely due to the increased
surface area irradiated.
reversible [2+4] cycloaddition of O2 to the naphthalene
skeleton, leading to the formation of 1,4-endoperoxides (7)
(Sortino et al., 2002). This step is strongly enhanced in water-
based solution (Aubry et al., 1995; Hart and Oku, 1972). The
acetal structure of the intermediate (7) is hydrolized in aqueous
medium (Pierlot et al., 1996) to give 1,4-naphthoquinone (8)
and the remaining side chain (9).
As the propranolol (1) used in this study was in a solid state,
the hydrolysis of the acetal (7) did not appear to proceed. Thus,
the C–O bond of the naphthyl ether was cleaved by a thermally
activated crossing of the bonding S (kk*) and T (kk*) states
1
1
into kj* states (Pohlers et al., 1996), generating 1-naphthol (2)
and the remaining side chain (9). Although the origin of the
Fig. 7. The mechanism proposed for the formation of 1,4-naphthoquinone (8) following exposure of propranolol (1) in the aqueous solution to light.

364
K. Uwai et al. / Life Sciences 78 (2005) 357–365
demonstrated, however, removing the drug from its press—
through packaging or grinding the drug results in the photo-
degradation of propranolol. These administration practices
jeopardize the measures taken to protect the drug from light-
mediated degradation. Although the photoproducts obtained in
this study did not exhibit any significant acute toxicities and or
high binding affinities to the h-adrenergic receptor, the
metabolic and biologic activities of these compounds are still
uncertain. In addition, there are several drugs with structures
similar to propranolol (1) such as naftopidil (10, Fig. 8), a a1-
adrenergic receptor blocker, that may also decompose upon
exposure to light; these photoproducts may be biologically
active compounds. This study serves as a warning to all
medical workers to be more vigilant in managing supplies,
especially the storing of pharmaceutics.
Fig. 8. Structure of naftpidil (10).
acetyl and formyl moieties appears to derive from the side
chain (9), additional experiments (photo irradiation of D-
labeled propranolol (1)) will be required to clarify this
mechanism. In addition, the irradiation of propranolol (1) in
a solution state with light was not suitable for the purpose of
our research.
We synthesized substantial quantities of N-acetyl- and N-
formylpropranolol (3 and 4) for pharmacological studies and
for use as a standard in spectrographic studies. Although the
photoproducts obtained this study are known compounds, their
pharmacological properties had not yet been investigated.
We subjected the photoproducts (2, 3, and 4) and their
synthesized analogues (5 and 6) to biological studies. In an
acute toxicity test, all of the degradation products exhibited low
toxicities, in comparison to the parental compound propranolol
References
Andrisano, V., Gotti, R., Leoni, A., Cavrini, V., 1999. Photodegradation studies
on atenolol by liquid chromatography. Journal of Pharmaceutical and
Biomedical Analysis 21 (4), 851–857.
Aubry, J.M., Mandard-Cazin, B., Rougee, M., Bensasson, R.V., 1995. Kinetic
studies of singlet oxygen [4+2]-cycloadditions with cyclic 1,3-dienes in 28
solvents. Journal of the American Chemical Society 117 (36), 9159–9164.
Berson, J.A., Brown, E., 1995. Dihydropyridines: II. The photochemical
disproportionation of 4-(2V-nitrophenyl)-1,4-dihydropyridines. Journal of
the American Chemical Society 77, 447–450.
(
1). These results may indicate that the free amino and
hydroxyl groups of the propranolol (1) side chain are important
for the toxic activity.
Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye
binding. Analytical Biochemistry 72, 248–254.
Next, we examined the binding of the propranolol derivatives
to h-adrenergic receptors in rat cerebral cortex membranes.
These substances exhibited low affinities to the h-adrenergic
receptor in comparison to propranolol (1). 1-Naphthol (2), which
only shares the naphthalene moiety with propranolol (1),
Chen C.S., 1994. Enzymic preparation of optically active propanolol and h-
adrenergic blockers using esterase. United States Patent No. 5334534.
Chen, C.S., Gou, D.M., Shieh, W.R., Liu, Y.C., 1993. Biocatalytic resolution of
dl-propranolol. A successful example of computeraided substrate design.
Tetrahedron 49 (16), 3281–3290.
À4
Chiou, T.-W., Chang, C.-C., Lai, C.-T., Tai, D.-F., 1997. Kinetic resolution of
propranolol by a lipase-catalyzed N-acetylation. Bioorganic and Medicinal
Chemistry Letters 7 (4), 433–436.
demonstrated a markedly lower affinity (IC : 4.8Â10 M).
5
0
These results suggested that the propranolol (1) side chain is
important for high binding affinity. Both N-acetyl- and N-
formylpropranolol (3 and 4), each of which contains a masked
secondary amino group, exhibited receptor binding affinities of
Comtet, M., Mettee, H.D., 1970. Photoisomerization of azulene. Molecular
Photochemistry 2 (1), 63–66.
Cordoba-Borrego, M., Cordoba-Diaz, M., Cordoba-Diaz, D., 1999. Validation
of a high-performance liquid chromatographic method for the determination
of norfloxacin and its application to stability studies (photo-stability study of
norfloxacin). Journal of Pharmaceutical and Biomedical Analysis 18 (6),
À5
À7
2
.2Â10 and 2.7Â10 M, respectively, values significantly
lower than the parental compound. These h-adrenergic receptor
binding affinities indicated that the free amino group is likely
also critical for the activity. N,O-diacetyl- and diformylpropra-
nolol (5 and 6), both of which have masked amino and hydroxyl
9
19–926.
Dewar, G.H., Kwakye, J.K., Parfitt, R.T., Sibson, R., 1982. Optical
purity determination by NMR: use of chiral lanthanide shift reagents
and a base line technique. Journal of Pharmaceutical Sciences 71 (7),
À6
À3
groups, exhibit IC50 values of 1.6Â10 and 8.3Â10 M,
respectively. As both of these values are of a lower affinity than
propranolol (1), these results may indicate that the region of
propranolol (1) side chain with the free amino and hydroxyl
groups is critical for high affinity binding to the h-adrenergic
receptor. Of the acetyl derivatives, the IC50 value for N,O-
diacetate (5) was lower than that for N-acetate (3), which is
opposite to the results observed for the formyl derivatives; the
IC50 values for N,O-diformate (6) was higher than that seen for
N-formate (4). As the sigmoidal pattern of N,O-diacetate (5)
differed from that observed for other derivatives (2, 3, 4 and 6), it
is possible that compound 5 binds the h-adrenergic receptor in a
different manner.
8
02–806.
Grundy, J.S., Kherani, R., Foster, R.T., 1994. Photostability determination of
commercially available nifedipine oral dosage formulations. Journal of
Pharmaceutical and Biomedical Analysis 12 (12), 1529–1535.
Hart, H., Oku, A., 1972. Octamethylnaphthalene 1,4-endoperoxide. Journal of
the Chemical Society, Chemical Communications (5), 254–255.
Hoffman, B.B., Lefkowitz, R.J., 1996. Catecholamines, sympathomimetic
drugs, and adrenergic receptor antagonists. In: Goodman, L.S., Gilman, A.,
Hardman, J.G., Limbird, L.E. (Eds.), Goodman and Gilman’s the
Pharmacological Basis of Therapeutics, (9th edR)McGraw-Hill, New York,
pp. 199–248.
Litchfield Jr., J.T., Wilcoxon, F., 1949. A simplified method of evaluating dose-
effect experiments. Journal of Pharmacology and Experimental Therapeu-
tics 96, 99–113.
Malinowska, B., Kiec-Kononowicz, K., Flau, K., Godlewski, G., Kozlowska,
H., Kathmann, M., Schlicker, E., 2003. Atypical cardiostimulant beta-
adrenoceptor in the rat heart: stereoselective antagonism by bupranolol but
Currently, single-dose packaging systems and ground
pharmaceutics are becoming increasingly popular; as this study

K. Uwai et al. / Life Sciences 78 (2005) 357–365
365
lack of effect by some bupranolol analogues. British Journal of Pharmacol-
ogy 139 (8), 1548–1554.
Pietta, P., Rava, A., Biondi, P., 1981. High-performance liquid chromatography
of nifedipine, its metabolites and photochemical degradation products.
Journal of Chromatography 210 (3), 516–521.
Marciniec, B., Ogrodowczyk, M., 2003. Photo- and radiostability of some 1,4-
dihydropyridine derivatives in the solid state. Acta Poloniae Pharmaceutica
Pohlers, G., Dreeskamp, H., Grimme, S., 1996. The mechanism of photo-
chemical C–O or C–S bond cleavage in aryl(thio)ethers. Journal of
Photochemistry and Photobiology. A, Chemistry 95 (1), 41–49.
Salomies, H., 1987. Oxidation of h-bloking agents V. Light-influenced
oxidation of propranolol and its glycol with N-bromosuccinimide. Acta
Pharmaceutica Suecica 24 (4), 193–198.
60 (2), 151–155.
Marinkovic, V.D., Agbaba, D., Karljikovic-Rajic, K., Vladimirov, S., Nedelj-
kovic, J.M., 2003. Photochemical degradation of solid-state nisoldipine
monitored by HPLC. Journal of Pharmaceutical and Biomedical Analysis
32 (4-5), 929–935.
Matsumoto, M., Kojima, K., Nagano, H., Matsubara, S., Yokota, T., 1992.
Photostability and biological activity of fluoroquinolones substituted at the
Shamsipur, M., Hemmateenejad, B., Akhond, M., Javidnia, K., Miri, R., 2003.
A study of the photo-degradation kinetics of nifedipine by multivariate
curve resolution analysis. Journal of Pharmaceutical and Biomedical
Analysis 31 (5), 1013–1019.
8
position after UV irradiation. Antimicrobial Agents and Chemotherapy 36
(
8), 1715–1719.
Nelson, W.L., Walker, R.B., 1978. The structure of propranolol hemisuccinate.
Research Communications in Chemical Pathology and Pharmacology 22
Sortino, S., Petralia, S., Bosc a` , F., Miranda, M., 2002. Irreversible photo-
oxidation of propranolol triggered by self-photogenerated singlet molecular
oxygen. Photochemical and Photobiological Sciences 1 (2), 136–140.
Suzuki, H., Fujiwara, S., Kondo, S., Sugimoto, I., 1985. Determination of
nifedipine in human plasma by high-performance liquid chromatogra-
phy with electrochemical detection. Journal of Chromatography 341 (2),
341–347.
(
3), 435–445.
Olmsted III, J., 1969. Photoisomerization of azulene. Molecular Photochemis-
try 1 (3), 331–336.
Pierlot, P., Hajjam, S., Barthelemy, C., Aubry, J.-M., 1996. Water-soluble
1
1
naphthalene derivatives as singlet oxygen ( O
media. Journal of Photochemistry and Photobiology. B, Biology 36 (1),
1–39.
2 g
, D ) carriers for biological
Yoshida, Y., Sato, E., Moroi, R., 1993. Photodegradation products of
levofloxacin in aqueous solution. Arzneimittel-Forschung 43 (5), 601–606.
3