Heat-accelerated degradation of solid-state andrographolide

Article abstract of DOI:10.1248/cpb.51.24

The stability of andrographolide, the major active diterpene lactone from Andrographis paniculata (Burm. f.) Wall. ex Nees., was determined to show that, while crystalline andrographolide was highly stable even at 70°C (75% relative humidity) over a period of 3 months, its amorphous phase degraded promptly. Heat-accelerated conditions revealed second-order kinetics of the decomposition with the rate constant at 25°C (k_25°C) predicted from the Arrhenius plot of 3.8×10^-6 d^-1. The major decomposed product under elevated temperature (70°C, 75% relative humidity) is 14-deoxy-11,12-didehydroandrographolide.

Full text of DOI:10.1248/cpb.51.24

24
Chem. Pharm. Bull. 51(1) 24—26 (2003)
Vol. 51, No. 1
Heat-Accelerated Degradation of Solid-State Andrographolide
a
a
,b
*
Luelak LOMLIM, Noppong JIRAYUPONG, and Anuchit PLUBRUKARN
a Department of Pharmaceutical Chemistry, Faculty of Pharmaceutical Sciences, Prince of Songkla University; and
b Department of Pharmacognosy and Pharmaceutical Botany, Faculty of Pharmaceutical Sciences, Prince of Songkla
University; Hat-Yai, Songkhla 90112, Thailand. Received July 22, 2002; accepted September 20, 2002
The stability of andrographolide, the major active diterpene lactone from Andrographis paniculata (BURM.
f.) WALL. ex NEES., was determined to show that, while crystalline andrographolide was highly stable even at
70 °C (75% relative humidity) over a period of 3 months, its amorphous phase degraded promptly. Heat-acceler-
ated conditions revealed second-order kinetics of the decomposition with the rate constant at 25 °C (k25 °C) pre-
dicted from the Arrhenius plot of 3.8ꢀ10ꢁ6 dꢁ1. The major decomposed product under elevated temperature
(70 °C, 75% relative humidity) is 14-deoxy-11,12-didehydroandrographolide.
Key words andrographolide; Andrographis paniculata; 14-deoxy-11,12-didehydroandrographolide; stability
Andrographis paniculata (BURM. f.) WALL. ex NEES. celerated condition. This result clearly demonstrated the in-
(Acanthaceae) is a medicinal plant widely used in several fluence of crystallinity on the stability of 1.
Asian countries, including China, India, and Thailand. An-
Kinetics of Degradation of 1 The determination of ac-
drographis herb, the dried aerial part of A. paniculata as celerated degradation kinetics was performed at 3 elevated
specified in the Thai Herbal Pharmacopoeia, has a wide temperatures, i.e., 45, 60, and 70 °C. The degradation of 1
range of activities. For example, as recommended in Thai- during the sampling periods from 7 to 91 d fitted second-
land’s National List of Essential Drugs 1999 (List of Herbal order kinetics, in which the linearity was best met when the
Medicinal Products),1) the indications for Andrographis herb reciprocals of the percentages of remaining amounts of 1 at
are for the treatment of noninfectious diarrhea and pharyngo- each temperature were plotted against time t (Fig. 1). The
tonsillitis. The plant is also used in the treatment of diabetes rate constant (k) determined for each temperature is 0.05ꢀ
and hypertension,2) hepatitis,3) malarial infection,4) and also
claimed to be an anti-HIV I agent, even though contrary evi-
dence was observed.5) The active constituents responsible for
its activities are labdane-type diterpene lactones, among
which the major components are andrographolide (1)6) and
some close derivatives such as deoxyandrographolide (2),
neoandrographolide (3),7,8) and 14-deoxy-11,12-didehydroan-
drographolide (4).9,10) The isolation of flavonoids from the
aerial part was also reported.11)
Despite its potential for clinical use, the further develop-
ment of A. paniculata has been hampered due to its stability
problem. Upon 1-year storage of the dried, powdered aerial
part in dry ambient condition, its total lactone contents re-
portedly decreased by 26%, thus restricting the herb’s storage
period.12) However, no thorough investigations of the stability
of either Andrographis herb itself or its constituents have
been systematically performed to our knowledge. Moreover,
Chart 1. Structures of Andrographolide and Its Derivatives
the decomposition paths have not been investigated. The de-
termination of andrographolide stability under elevated tem-
peratures is reported and the possible major degradation
pathway is discussed here.
Results and Discussion
Stability of Crystalline versus Amorphous Phases of
Andrographolide (1) Early in this investigation, a prelimi-
nary study on the degradation of 1 was initiated in order to
optimize sampling schedule for the stability analysis. Crys-
talline form of 1 was allowed to stand at 70 °C for 3 months.
To our surprise, over this period the crystalline sample was
perfectly stable, i.e., remaining totally unchanged both physi-
cally and chemically. Therefore, another investigation was
Fig. 1. Second-Order Plots of Degradation of 1 at 45, 60, and 70 °C
devised using the amorphous form of 1, prepared as a 1 : 2
solid dispersion of 1 in PVP-K30. It was found that 1 decom-
posed promptly during 2-month standing under the same ac- r2ꢁ0.9833, respectively.
The rate equation for each temperature is expressed as 1/cꢁ(0.05ꢀ10ꢂ3)tꢃ0.0096,
r2ꢁ0.9128; 1/cꢁ(0.08ꢀ10ꢂ3)tꢃ0.0108, r2ꢁ0.9248; and 1/cꢁ(0.70ꢀ10ꢂ3)tꢃ0.0130,
∗ To whom correspondence should be addressed. e-mail: planuchi@ratree.psu.ac.th
© 2003 Pharmaceutical Society of Japan

January 2003
25
Here, we propose that the mechanism of the bimolecular
1,4-pyrolytic elimination of 1 possibly involve a paired-com-
plex transition state, in which the two molecules of substrate
are pairing to help each other in withdrawing the leaving
group. This suggestion, nonetheless, needs further and more
systematic examination. While the degradation mechanism
of the diterpene lactones deposited in dried herb are possibly
irrelevant to that of the isolated compound, such study could
still lead to the application of andrographolide and its deriva-
tives for more practical clinical use as pure substances.
Fig. 2. Arrhenius Plot of 1
The Arrhenius relation can be expressed as ln kꢁ23.553ꢂ(10.732/T); r2ꢁ0.7774.
Conclusions
Upon acceleration by elevated temperatures, the degrada-
tion of 1 is significantly influenced by its crystallinity. While
10ꢂ3, 0.08ꢀ10ꢂ3, and 0.70ꢀ10ꢂ3 dꢂ1, respectively. The ex- crystalline andrographolide was completely stable at 70 °C,
trapolation of the Arrhenius plot (Fig. 2) obtained from the its amorphous phase decomposed more promptly. Determi-
relation between ln k and the reciprocal of thermodynamic nation of the degradation of 1 under heat-accelerated condi-
temperature (T) leads to the estimation of k at 25 °C (k25 °C) of tions showed second-order kinetics, and the extrapolation of
3.8ꢀ10ꢂ6 dꢂ1.
the Arrhenius plot revealed the estimated k25 °C of 3.8ꢀ10ꢂ6
As obligated by second-order kinetics, the half-life (t1/2) dꢂ1. The major degradation route of 1 under the above accel-
and shelf-life (t90%) of 1 are dependent on the initial concen- erating conditions is 1,4-elimination-dehydration at the al-
tration of the sample and thus cannot be generalized. For the lylic 14-OH group, yielding 14-deoxy-11,12-didehydroan-
experimental conditions used here, predicted t1/2 and t90% at drographolide (4) as the major decomposed product.
25 °C are 7.83 and 0.87 years, respectively.
Our estimated k25 °C of 3.8ꢀ10ꢂ6 dꢂ1 corresponds well
Degraded Products and Degradation Path of 1 under with the restricted storage period of Andrographis herb rec-
Elevated Temperature Amorphous andrographolide (equiv- ommended in the Thai Herbal Pharmacopoeia. Nonetheless,
alent to 300 mg of 1) was allowed to stand at 70 °C. After 2 the observed degradation route of 1 under these experimental
months, the crude degradation mixture was subjected to iso- conditions failed to explain the decomposition of lactone de-
lation using semipreparative RP C-18 HPLC (55% aqueous rivatives deposited in plant cells. As mentioned earlier,
CH3CN). Along with the peak of 1 at 5.3 min, the chro- degradation during the storage of Andrographis herb led to
matogram also showed one additional detectable peak (tR the loss in total lactone contents, implying a major transfor-
10.6 min). The fraction related to the 10.6-min peak was col- mation at the lactone linkage, not at the allylic moiety as ob-
lected to afford compound 4 as the major degraded product served here. Further investigations regarding the stability of
(5.2 mg from an equivalent of 12 mg of 1; 42% isolated Andrographis herb itself, therefore, have been already
yield).
planned. The results of the stability profiles of both Andro-
The 1H-NMR spectrum of 4 (500 MHz; DMSO-d6) is sim- graphis herb and its active constituents will shed the light on
ilar to that of 1, excepts that some additional signals at d the advanced development of this highly potential medicinal
7.63 (t, Jꢁ1.8 Hz; H-14), 6.73 (dd, Jꢁ15.9, 10.0 Hz; H-11), herb for wider clinical use.
and 6.12 (d, Jꢁ15.9 Hz; H-12) are observed. Also, a triplet
Experimental
signal at d 4.93 and a doublet at d 5.70, originally assigned
General Unless stated otherwise, all chromatographic solvents used for
to H-14 and 14-OH of 1, disappear in the spectrum of 4.
These changes in chemical shifts indicate that compound 4 is
HPLC were of analytical grade and were filtered through a membrane filter
(230 mm; Millipore). UV spectrum was obtained with a Hewlett-Packard
8452A Diode Array spectrometer. IR spectrum was recorded on a Jasco IR-
810 spectrometer. Mass spectrometry was performed in ESI mode on an
LCT Module Mass Spectro detector. NMR spectra were obtained using a
Varian Unity Inova 500 spectrometer (500 MHz for 1H and 125 MHz for
13C). The signals were reported on the d-scale and referred to the signals of
in fact the known 14-deoxy-11,12-didehydroandrographolide
,
which was first isolated from the whole plant of A. panicu-
lata by Balmain and Connelly.9) The structure of 4 was con-
firmed by the comparison with its spectral data later reported
by Matsuda et al.10)
1
solvent (DMSO-d6; 2.49 ppm for H and 39.7 ppm for 13C). Optical rotation
was measured on a Polax-L polarimeter and was reported as specific rota-
tion. Differential scanning calorimetry (DSC) was performed on a Perkin
Elmer DSC7 calorimeter. The elevated temperature experiments were car-
ried out using the Memmert BP600, B50, and Pr402 ovens with the con-
trolled temperatures of 45ꢄ2, 60ꢄ2, and 70ꢄ2 °C, respectively.
Analytical HPLC was performed on a Waters 600 controller-solvent deliv-
ery system, equipped with a Waters 717 plus autosampler and a Waters 486
tunable absorbance detector. Preparative HPLC was performed on a Waters
600E solvent delivery system connected to a Rheodyne 7161 injection valve
and a Waters 484 tunable absorbance detector.
Chemicals Andrographolide (1) was isolated from A. paniculata col-
lected in the vicinity of Songkhla district in June 2000. The specimen was
identified by an author (A.P.), and a voucher was deposited at the Depart-
ment of Pharmacognosy and Pharmaceutical Botany, Faculty of Pharmaceu-
tical Sciences, Prince of Songkla University.
The dehydration of 1, as the major decomposition path
under elevated temperatures, is unexplainable by any avail-
able elimination mechanisms. Presuming that neither humid-
ity nor PVP are involved directly in the rate equation, this
degradation undergoes bimolecular kinetics with only a sin-
gle species of substrate. Generally, the bimolecular elimina-
tion E2 requires a base, whether weak or strong, to abstract
the b-proton (or in this case, the d-proton). On the other
hand, the pyrolytic elimination Ei, which can proceed with a
stand-alone substrate, is naturally unimolecular. Besides the
two contradictory aspects above, 1,4-elimination is generally
rare. More often, 1,4-elimination of allylic alcohol is encoun-
tered when a more easily abstracted group (e.g., silyl group)
is present at the d position (for example, see ref. 13).
The crude MeOH extract of A. paniculata (dried aerial part, 1.5 kg) was
subjected to vacuum chromatography over a SiO2 column using 2% MeOH

26
Vol. 51, No. 1
in CH2Cl2 as eluting solvent. The fractions containing 1 as major component
were pooled, and 1 was repeatedly recrystallized from MeOH as white nee-
dles (2.9 g, isolated yield 0.2%). The authenticity was confirmed on the basis
of NMR spectral analysis, according to data reported by Fujita et al.14)
Solid Dispersion of 1 Compound 1 in the amorphous form was pre-
pared as a 1 : 2 solid dispersion of PVP-K30. An aliquot solution of the 1 : 2
mixture of 1 and PVP-K30 (equivalent to 5.0 mg of 1) was transferred to a
5-ml vial and then concentrated under reduced pressure to the resulting
dried, thin film. The collapse of the crystal structure was detected using
DSC, in which the thermograms of the crystalline 1 and the solid dispersion
were compared. The signal at 240 °C, which was normally found in the ther-
mogram of standard crystalline 1, disappeared in the thermogram of the
solid dispersion, thus suggesting that 1 in the solid dispersion was no longer
in its crystalline form.
Stability Analysis of 1 In all the elevated-temperature systems through-
out this study, closed desiccators each containing a jar of saturated NaCl so-
lution were used to achieve the atmosphere of 75ꢄ5% relative humidity
(adopted from the Handbook of Pharmaceutical Excipients).15) A series of
solid dispersion of 1 in PVP-K30 (equivalent to 5.0 mg of 1 each) were sepa-
rately allowed to stand at 45ꢄ2, 60ꢄ2, and 70ꢄ2 °C (75% relative humid-
ity) according to the following intervals; 7, 21, 49, 63, 77, and 91 d (45 and
60 °C); or 7, 14, 28, 42, 63, and 84 d (70 °C). After each time, the sample
(triplicated) from each temperature was removed and preserved at ꢂ80 °C
until quantification.
19a), 3.84 (1H, dd, Jꢁ11.0, 7.3 Hz; H-19b), 4.13 (1H, dd, Jꢁ7.3, 2.6 Hz;
OH-19), 4.41 (1H, d, Jꢁ1.8 Hz; H-17a), 4.72 (1H, d, Jꢁ1.8 Hz; H-17b),
4.87 (2H, d, Jꢁ1.8 Hz; H-15), 5.02 (1H, d, Jꢁ5.0 Hz; OH-3), 6.12 (1H, d,
Jꢁ15.9 Hz; H-12), 6.73 (1H, dd, Jꢁ15.9, 10.0 Hz; H-11), 7.63 (1H, t,
Jꢁ1.8 Hz; H-14). 13C-NMR (DMSO-d6) d: 13.0 (q; C-20), 22.8 (t, C-6; and
q; C-18), 27.0 (t; C-2), 36.0 (t; C-7), 37.0 (t; C-1), 39.0 (s; C-10), 42.0 (s; C-
4), 54.0 (d; C-5), 60.0 (d; C-9), 64.0 (t; C-19), 70.0 (t; C-15), 79.0 (d; C-3),
106.6 (t; C-17), 121.0 (d; C-12), 127.0 (s; C-13), 133.0 (d; C-11), 146.5 (d;
C-14), 149.0 (s; C-8), 172.0 (s; C-16). IR (thin film) cmꢂ1: 3400, 1750,
1050. UV lmax (CH3OH) nm (log e): 208 (3.09), 250 (2.89). ESI-MS m/z:
355 (MꢃNa)ꢃ, 333 (MꢃH)ꢃ, 315, 253, 210. [a]D25 ꢃ5.33° (cꢁ0.1,
CH3OH).
Acknowledgments This study was financially supported by a research
grant from the Faculty of Pharmaceutical Sciences, Prince of Songkla Uni-
versity. We wish to extend our gratitude to Dr. Sirirat Pinsuwan of the De-
partment of Pharmaceutical Technology, Faculty of Pharmaceutical Sci-
ences, Prince of Songkla University, for her valuable discussion regarding
the reaction kinetics.
References
1) National Drug Committee, “National List of Essential Drugs A.D.
1999 (List of Herbal Medicinal Products),” National Essential Drug
List Committee, Ministry of Public Health, Bangkok, 1999.
2) Zhang C., Kuroyangi M., Tan B. K., Pharmacol. Res., 38, 413—417
(1998).
3) Jain D. C., Gupta M. M., Saxena S., Kumar S., J. Pharm. Biomed.
Anal., 22, 705—709 (2000).
4) A Rahman N. N. N., Furuta T., Kojima S., Takane K., Mohd M. A., J.
Ethnopharmacol., 64, 249—254 (1999).
The HPLC condition for the quantification of 1 was: C18 Thermo-Hyper-
sil column (5 mm; 250ꢀ4.6 mm) at ambient temperature, eluted with iso-
cratic 30% aqueous CH3CN (1.5 ml/min), and detected at l 254 nm. This
procedure was validated with a linearity in the range of 10—200 mg/ml
(r2ꢅ0.9998). The accuracy was evaluated using a sample at 100 mg/ml to re-
veal the recovery percentage of 102.7% (%R.S.D.ꢆ3.0).
The solutions of decomposed mixtures (in 2.0 ml of CH3CN each; equiva-
lent to 5.0 mg of the starting amount of 1) were prepared and then diluted
50-fold with the mobile phase. These were used as sample preparations. Its
HPLC chromatogram showed only two major peaks at tR 5.7 and 24.7 min,
belonging to compounds 1 and 4, respectively. The percentage of the re-
maining amount of 1 was calculated from its peak area in accordance with
the standard curve.
Isolation of Degraded Products of 1 The solid dispersion of 1 (1 : 2 in
PVP-K30; equivalent to 300 mg of 1) was allowed to stand at 70 °C for 2
months. After this period, an aliquot of this mixture was subjected to a semi-
preparative HPLC column (C18 Hamilton PRP-1; 10 mm; 305ꢀ7.0 mm)
using 55% aqueous CH3CN (1.8 ml/min) as eluting solvent and detected at l
254 nm. The fraction eluted at tR 10.6 min was then collected to afford com-
pound 4 as a white solid (5.2 mg from an equivalent of 12 mg of starting 1;
42% isolated yield).
5) Collins R. A., Ng T. B., Fong W. P., Wan C. C., Yeung H. W., Life Sci.,
60, 345—351 (1997).
6) Cava M. P., Chan W. R., Haynes L. J., Johnson L. F., Weinstein B.,
Tetrahedron, 18, 397—403 (1962).
7) Cava M. P., Chan W. R., Stein R. P., Willis C. R., Tetrahedron, 21,
2617—2632 (1965).
8) Chan W. R., Taylor D. R., Willis C. R., Bodden R. L., Fehlhaber H.-W.,
Tetrahedron, 27, 5081—5091 (1972).
9) Balmain A., Connelly J. D., J. Chem. Soc., Perkin Trans. 1, 1973,
1247—1251 (1973).
10) Matsuda T., Kuroyanagi M., Sugiyama S., Umehara K., Ueno A.,
Nishi K., Chem. Pharm. Bull., 42, 1216—1255 (1994).
11) Gupta K. K., Taneja S. C., Dhar K. L., Atal C. K., Phytochemistry, 22,
314—315 (1983).
12) Dechatiwonse Na Ayudhya T., Techadamrongsin Y., Jirawattanapong
W., “Chemical Specification of Thai Herbal Drugs,” Vol. 1, Depart-
ment of Medical Sciences, Ministry of Public Health, Bangkok, 1993.
13) Maeta H., Suzuki K., Tetrahedron Lett., 33, 5969—5972 (1992).
14) Fujita T., Fujitani R., Takeda Y., Takaishi T., Kido M., Miura I., Chem.
Pharm. Bull., 32, 2117—2125 (1984).
15) Wade A., Weller P. J., “Handbook of Pharmaceutical Excipients,” 2nd
ed., American Pharmaceutical Association, Washington, DC, and
Royal Pharmaceutical Society of Great Britain, London, 1994.
14-Deoxy-11,12-didehydroandrographolide (4): White solid (5.2 mg from
1
an equivalent of 12 mg of starting 1; 42%). H-NMR (DMSO-d6) d: 0.74
(3H, s; H-20), 1.08 (3H, s; H-18), 1.11 (1H, ddd, Jꢁ13.5, 13.5, 5.0 Hz; H-
1ax), 1.17 (1H, m; H-5), 1.31 (1H, ddd, Jꢁ13.5, 3.4, 3.4 Hz; H-1eq), 1.38
(1H, dddd, Jꢁ13.5, 13.0, 13.0, 4.1 Hz; H-6ax), 1.55 (1H, m; H-2ax), 1.60
(1H, m; H-2eq), 1.71 (1H, dddd, Jꢁ13.5, 2.7, 2.5, 2.3 Hz; H-6eq), 1.97 (1H,
ddd, Jꢁ12.6, 5.0, 5.0 Hz; H-7eq), 2.35 (2H, m; H-7ax and H-9 overlapped),
3.21 (1H, ddd, Jꢁ10.3, 5.0, 5.0 Hz; H-3), 3.27 (1H, dd, Jꢁ11.0, 2.6 Hz; H-