176
Qualitative and Quantitative Analysis of the Phenolic
Constituents from Orthosiphon aristatus
Wahono Sumaryono1'2 Peter Proksch'- , Victor Wray ?, Ludger Witte , and Thomas Hartmann'
Institut für Pharmazeutische Biologie, TU Braunschweig, Mendelssohnstr. 1, D-3300 Braunschweig,
Federal Republic of Germany
2
Permanent address: Directorate of Life Sciences, Agency for the Assessment and Application of Technology (BY.P. Teknologi),
JI. M. H. Thamrin 8, Jakarta Pusat 10340, Indonesia
Gesellschaft für Biotechnologische Forschung mbH, Mascheroder Weg 1, D-3300 Braunschweig, Federal Republic of Germany
Address for correspondence
Received: January 26, 1990
Abstract
Orthosiphon aristatus (Orthosiphonis
folium DAB 9) was studied with regard to its phenolic
constituents. Twenty compounds were isolated and
identified on the basis of their spectral characteristics.
The compounds included nine lipophilic flavones, two
flavonol glycosides, and nine caffeic acid derivatives.
The presence of the recently reported methylripario-
chromene A could not be confirmed. All compounds
identified were quantified by HPLC. The caffeic acid de-
rivatives including the major compounds rosmarinic
acid and 2,3-dicaffeoyltartaric acid (67% of total iden-
tified phenolics) predominated over the flavones (33%)
in an aqueous MeOH extract. The predominance of the
caffeic acid derivatives was even more pronounced in a
hot water extract (94.5% of total identified phenolics)
that was comparable to a herbal tea.
Key words
Orthosiphon aristatus, Orthosiphonis
folium, caffeic acid derivatives, flavonol glycosides,
lipophilic flavones, HPLC.
Introduction
Orthosiphon aristatus (Bl.) Miq. (syn. 0.
grandflorus Bold., syn. 0. spicatus (Thunb.) Bak., syn. 0.
stainineus Benth.) is a member of the Lamiaceae native to
tropical Asia that is currently under cultivation in In-
donesia, the main exporter of this medicinal plant (1—4).
Leaves and stems of 0. aristatus (Orthosiphonis folium DAB
9) are used to prepare a diuretic tea which has also been re-
ported to be active against kidney and bladder inflamma-
tions (1—4). The active plant constituents, however, are still
basically unknown. To gain a better understanding of the
secondary products from 0. aristatus that are present
among the extractives of the herbal tea and hence may also
be involved in the pharmacological effects observed, we
have now analysed the phenolic constituents of 0. aristatus
paying special attention to their quantitative composition.
Materials and Methods
Downloaded by: University of Arizona Library. Copyrighted material.
The dried herbal drug of Orthosiphon aristatus
was commercially available from Alfred Galke GmbH, Gittelde,
Harz, FRG. 250 g of plant material was exhaustively extracted with
80% aqueous MeOH followed by an extraction with 50% aqueous
MeOH. Both extracts were combined and MeOH was evaporated
under reduced pressure. The resulting aqueous extract was par-
titioned against CH2C12. The organic fraction that contained the
lipophilic flavones was taken to dryness, redissolved in CH2C12/
MeOH (95 5, v/v), and chromatographed on a silica gel column in
the same solvent. Similar fractions were combined and chromatog-
raphed on a Sephadex LH-20 column with MeOH as solvent. Final
purification of the flavones was achieved by preparative TLC on
premade silica gel plates (Merck, Darmstadt, FRG) in the above sol-
vent system and by preparative TLC on polyamide (DC 6) with to-
luene/petroleum ether/methyl ethyl ketone/MeOH (30: 60: 5 : 5, v/
v) as eluent. The aqueous fraction was lyophilized, redissolved in a
small volume of H20, and chromatographed on a Polyamide SC-6
column. The elution was started with H20, followed by 50% aque-
ous MeOH and 100% MeOH. The various compounds were isolated
from these fractions by a combination of column chromatography
(Sephadex LH-20 with MeOH or 70% aqueous MeOH as solvents),
preparative TLC [on microcrystalline cellulose with 15% acetic
acid in H20, on silica gel with EtOAc/HCOOH/CH3COOH/H20
(100: 5 : 5: 13, v/v) as solvent system], and by low pressure re-
versed phase column chromatography on an RP-8 column (Lich-
roprep, 40—63 xm. Merck, Darmstadt, FRG) with 70% aqueous
MeOH as solvent.
Plants of 0. aristatus were grown from shoots
that had originally been supplied from the Research Institute for
Spice and Medicinal Plants (BALLITTRO), Boqor, Indonesia. Fresh
leaves were dipped into acetone for 15—3Osec, the extract was
taken to dryness, redissolved in MeOH, and analysed for the
flavones by HPLC as described above.
The HPLC system consisted of a Rheodyne injec-
tion valve (20c1), a 2152 HPLC controller, a 2150 HPLC pump, a
11300 Ultrograd mixer driver, a 2151 HPLC variable wavelength
monitor (all Pharmacia, LKB, Sweden), and a C-R3A Chromatopac
(Shimadzu, Japan). The separation was achieved by applying a
linear gradient from 100% A (10% CH3CN, 90% H20 containing
1 % H3P04 or alternatively 1 % CH3COOH) to 100% B (45% CH3CN,
55% H20 containing 1 % H3P04 or alternatively 1 % CH3COOH) in
20mm followed by 10mm isocratic at 100% B. The flow rate was
1.0 ml/min for the "short" column and 1.3 mI/mm for the long col-
umn, detection was at 340 nm. Two separation columns were
employed: a "short" column (Novapak C-18, 125 x 4mm, 5km,
Waters Millip ore GmbH, Eschborn, FRG) and a "long" column
Qualitative and Quantitative Analysis of the Phenolic Constituents from Orthosiphon aristatus
(Nucleosil C-18, 250 x 4mm, 5 urn, Macherey & Nagel, Dilren,
FRG). For HPLC analysis 0.5 g of powdered plant material was ex-
tracted with 50 ml 80% aqueous MeOH for 24 h in the dark. Prior to
HPLC analysis MeOH and H20 were added resulting in a total vol-
ume of150 ml (MeOH/H20 1: 1, v/v) and 20 il were injected into the
HPLC apparatus. For the hot aqueous extract (herbal tea) 50 ml of
boffing H20 was poured over 0.5 g of powdered plant material and
left at room temperature for 30 mm. Following filtration, cold H20
and MeOH were added to the H20 extract resulting in a total volume
of 150 ml (MeOH/H20 1: 1, v/v) and 20 u1 were injected into the
HPLC apparatus. All extracts were centrifuged (10000 g for 5 mm)
prior to injection. Quantification was achieved by the external stan-
dard method with previously isolated and purified compounds.
1H-NMR and 13C-NMR spectra were recorded in
DMS0-d6 (unless otherwise stated) on Bruker AM-300 or WM-400
spectrometers, respectively. All 1D- and 2D-spectra were obtained
using the standard Bruker software. Mass spectra (direct probe. El,
70eV; negative ion FAB mass spectra, glycerol or triethanolamine
as matrix) were measured on a Finnigan MAT 8430 mass spec-
trometer. The FAB mass spectrum of compound 17 (triethanol-
amine as matrix) revealed the signal of[M + Na] at m/z = 741. No
satisfactory FAR mass spectra could be obtained for compounds
18—20 due to the presence of an excess of salt. The GC-MS system
consisted of a Carlo Erba 5160 GC [equipped with a 15 m x
0.23mm i.d. fused silica glass capillary column coated with DB-1
(J&W scientific); conditions: temperature program 100—300°C;
6°C/mini that was coupled with the quadrupole mass spectrometer
Finnigan MAT 4515. Spectra (El) were recorded at 70 eV in combi-
nation with the Incos data system.
Results and Discussion
Twenty phenolic constituents including
nine flavone aglycones (1—9), two flavonol glycosides (10,
11), one coumarin (12), caffeic acid (13) as well as seven caf-
feic acid depsides (14—20) were isolated from 0. aristatus
(Fig. 1). The flavones were methylated derivatives of scutel-
larein (5, 6, 7, 4'-tetrahydroxyflavone) as well as of 6-hy-
droxyluteolin (5,6,7,3' ,4'-pentahydroxyflavone). Several
flavones including the well known major flavone sinensetin
(9) were characterized by the presence of a methoxy group
at C-5, a structural feature usually rare in flavonoids (5).
From the flavones isolated, compounds 3 and 4 had
hitherto not been reported for 0. aristatus (6—9). As the
flavones 1—5 and 6—9 were isomers differing only in the
number and position of methoxy substituents their identifi-
cation especially by their 1H-NMR spectra was difficult. The
structural assignments were thus based mainly on 1H-nOe
studies and on comparison of their UV and mass spectral
data. For an overview, the 1H-NMR data of all flavones iso-
lated are summarized (Table 1).
The flavones 1—9 were present exclusively
as aglycones since hydrolysis of the aqueous fraction of the
crude extract failed to detect even traces of liberated
flavones. Instead of flavones the ubiquituous flavonols
kaempferol and quercetin were present in the aqueous frac-
tion as 3-0-/3-glucosides (10, 11). Since the flavonols were
not present as aglycones a clear separation between
flavones accumulating exclusively as aglycones and
flavonols present only as glycosides is observed. The
flavones were recovered by a brief solvent wash of fresh
leaves (15—3Osec in Me2CO) suggesting their presence in
epidermal cells or in glandular trichomes found on the leaf
surface.
Planta Med. 57(7991) _!L?_
In addition to the flavonoids, seven caffeic
acid depsides (14—20) as well as free caffeic acid (13) and
the coumarin esculetin (12) were isolated from 0. aristatus.
None of these polar phenolic constituents except rosmarinic
acid (14) (10) have previously been reported for this
species. The depsides included caffeoyl tartrate (15) as well
as dicaffeoyl tartrate (16). Both compounds are rare natural
products that were reported previously, for example, in
species from the genus Echinacea (Asteraceae) (11, 12).
The stereochemistry at the asymmetric carbons of 15 and
16 was not elucidated.
The structural elucidation of the new com-
pounds 17—20, deduced from the 1H- and 13C-NMR data
(Tables 2 and 3, respectively), was somewhat more com-
plex and will be considered here in more detail. After H/D
exchange, 17 gave well resolved 1D- and 2D-1H-COSY
spectra in DMSO-d6 from which the nature of the individual
aromatic and side-chain spin systems in the molecule could
be determined. 1H-nOe difference spectra allowed the as-
Downloaded by: University of Arizona Library. Copyrighted material.
signment of the side chains to the appropriate aromatic sys-
tems to give four fragments corresponding to 1—9, 1'—9',
1"— 9", and 1 "—9" in the final structure. A strong nOe to H-
7 and H-8 upon irradiation of H-8' suggested the relative ar-
rangement of two aromatic moieties. Comparison of the
data for 17 with literature data (13, 14) indicated that 17 is
a lithospermic acid derivative with an additional dihy-
drocaffeic acid substituent. The position of this moiety was
deduced from the '3C-NMR spectrum in CD3OD (Table 3).
The presence of trace amounts of
paramagnetic material in the sample preparation caused
selective broadening of several signals in the spectrum. Of
particular significance was the observation of broad signals
for C-8" and C-8" and the almost complete absence of carb-
oxylic acid carbonyl signals. This can only be rationalised if
both carboxylic acid residues of the two dihydrocaffeic acid
substituents corresponding to systems 1"—9" and 1"—9"
are available for complexation with paramagnetic ions.
That such selective broadening occurred under these cir-
cumstances was tested with rosmarinic acid and the
phenomenon has been observed previously for lithosper-
mic acid isolated from natural sources (14). It should be
noted that C-8 and C-8', readily asigned from their charac-
teristic chemical shifts, were not affected.
Comparison of the 1H-NMR data of 19 and
20 with 17 indicated the sequential replacement of the di-
hydrocaffeic acid moieties by dihydro-p-hydroxycinnamic
acid moieties. A second compound, 18, isomeric to 17 was
isolated. Significant differences in the 1H-NMR chemical
shifts to those of 17 were observed in DMSO-d6 (Table 2).
The exact nature of 18, however, could not be deduced from
the available data. It could either arise by isomerisation at
one of the four asymmetric centres in 17 or through posi-
tional change of one of the dihydrocaffeic acid moieties.
A recent report claimed the occurrence of
methylripariochromene A in 0. aristatus comprising up to
4% of the dry weight of the plant (15). However, we were
not able to detect this chromene derivative even as a trace
constituent in any of the collections of 0. aristatus available
to us.
178 Planta Med. 57 (1991) Wahono Sumaryono et aL

OH
OH OH C
J 'OH
HO.y.Ori1
R20'
R1' OH 0
OH

OH 0
H
HO 4L
HO'O'O fOH
OH OH OH
R' R2 R3 R'
1H H3C H H 10 11 12 13
2 H H3C H CH3
3H3C H H CH3
4 HC H3C H H 0
5 H3C
6 H
H3C
H3C
H
OH
CH3
Cl-I3
c°
7 H
8 H3C
H3C
H3C
OCH3
OH
CH3
CH3
OO_C*O
9 H3C H3C OCH3 CH3

OHOH

14 15 16 OH

Downloaded by: University of Arizona Library. Copyrighted material.

17/18 R3 R R3 r R4 OH
19 R3 OH, R3 H
or
R3 H R' R3 OH
20 R3 R3 H, R R' OH Fig. 1 Structures of phenolic constituents from Orthosiphon ar/status.

Table 1 1H-NMR data of lipophilic flavones (1—9) from Orthosiphon ar/status.

1 2 3 4 5 6 7 8 9

H-3 6.67 (s) 6.97(s) 6.63(s) 6.63 (s) 6.71c(s) 6.76 (s) 7.00(s) 6.57 (s) 6.78(s)
H-8 6.88 (s) 6.99(s) 7.08(s) 7.20 (s) 7.20 (s) 6.84 (s) 7.06(s) 7.145 (s) 7.18(s)
H-2' 7.83 ('d) 8.09 ('d") 7.99 ("d") 7.91 ('d") 8.01 ('d") 7.45 (d) 7.61 (d) 7.45 (d) 7.52 (s')
H-3' 6.65 ("d") 7.l3rd") 7.09rd') 6.92 ('d") 7.09 (d") — — — —

H-5' 6.65 ('d") 7.l3rd") 7.09rd") 6.92 (d') 7.09 ('d") 7.05 (d 7.15(d) 7.06 (d) 7.08(d)
H-6' 7.83 ("dO) 8.09 ('d") 7.99 ('d") 7.91 ('d") 8.01 ('d) 7.53 (dd) 7.74 (dd) 7.49 (dd) 7.63 (d')
5-OH — — — —
13.23(s) 12.88(s) 12.88 (s) 12.91 (s) —

3'-OH — — — — — — — —
9.44 (s)
5-OMe — — — — 3.76
3.72(s) 3.76 (s) 3.76 (s) (s) 3.76(s)
6-OMe — 3.79 (s) 3.79
3.72 (s) 3.73(s) 3.80 (5) 3.72 (s) 3.74(s) (s) 3.78(s)
7-OMe 3.91a(s) 3.93(s) 3.89(s) 394b(s) 3.g4d(s) 3.91(s) 3.95(s) 394h (s) 3.94(s)
3'-OMe — — — — — — —
3.90(s) 3.88(s)
4'-OMe — — 3.85
3.86(s) 3.85(s) 3.85e(s) 3869s) 3.86(s) (s) 3.83(s)
1: 'd = M' BB' system; J(2._3') + J(2'_5') = 8.9.
Irradiation at 7-OMe gave a flOe at H-8. 2: J2_ + J(2_5) = 9.0.3: J(2..3)+ i)2 = 8.7;
4: J(2._3)+J(2_5) = 8.8.
Irradiation at 7-OMe gave a n0e at H-8. 5: J(2_3)+ J2_5) = 8.9.
Irradiation at H-3 gave a flOe at H-2' and H-6'.
Irradiation at 7-OMe gave a n0e at H-8.
Irradiation at 4-OMe gave a n0e at H-3' and H-5'. 6: J(5_6') = 8.5; J(2'_6') = 0.8.
Irradiation at 7-OMe gave a n0e at H-8.
Irradiation at 4'-OMegave a n0e at H-5'. 7:J(5_6) = 8.6;J._6.) = 2.1.8: J(5' ..6') = 8.6;J(2_6.)= 2.1.
Irradiation at 7-OMe gave a n0e at H-8. 9: J (5'-6) = 8.6;
1)2_6') not observed.
Chemical shifts (ppm) are relative to TMS, coupling constants are in Hz.
Qualitative and Quantitative Analysis of the Phenolic Constituents from Orthosiphon aristatus Planta Med. 57(1991) 179

Table 2 'H-NMR data for compounds 17 to 20. Table 3 '3C-NMR data for compound 17 in CD3OD.

1P 17 18 l9 2o Shift' Assignment Shift Assignment

(DMSO-d6) (CD3OD) (DMSO-d6) (CD3OD) (CD3OD)
176.8 s" 9",9" 122.11 d 6
Chemical Shifts ó (ppm) 172.80 s 9' 121.91 d 6"
H-5 6.78 6.79 6.83 6.85 168.86 s 9 121.35 d 6"
H-6 7.12 7.11 7.17 7.12 7.10 148.95 s 3 118.36 d 6'
H-i 7.11 7.31 7.36 7.27 7.27 146.41 s J 3" 117.94 d 2"
H-8 6.16 6.16 6.19 6.14 6.14 146.31 s 1 3" 117.54 d 2"
H-2' 6.70 6.71 687 687g 145.74 s J 4" 117.13 dx 2 5,5'
H-5' 6.70 6.72 678g 679g 145.58 s 4" 116.51 d 5"
H-6' 6.55 678 67g 144.62 sx2 5 3',4 116.29 dx 2 5",8
H-7' 5.79 5.97 5.58 6.00 6.06 144.46 s (4' 113.53 d 2'
H-8 4.20 4.32 4.31 4.30 4.31 142.50 d 7 88.28 d 7'
H-2" 6.83 6.80 7.15 7.16 133.92 s 1 78.81 d" 5 8"
H-3" ——— ——— 6.76 676' 131.01 s 1' 77.88 d" (8"
H-5"
H-6"
6.55
6.42 "
3.07
6.44
2.92
6.76
7.15
3.12
6.76
7.16
3.19
130.45
126.64
124.69
s
s
s
1"
1"
2
58.49
38.65
38.19
d
t
t
8'
J 7"
1 7"
H-7"A 2.95
H-7"B 2.74 2.97 2.84 3.00 2.98
H-8" 4.84 —c 4.90 C
'The multiplicities were determined from a 135' DEPT spectrum.
H-2" 6.40 6.48 6.45 6.71 "Very broad signal.
H-3'" ——— ——— ——— ——— 6.51 These signals are broad and Ca. 20% as high as the other doublets.

Downloaded by: University of Arizona Library. Copyrighted material.

H-5" 6.52 6.53 6.50 6.51
H-6" 6.22 6.28 6.27 6.25 6.71
H-7"A 3.01 3.07 2.92 3.12 3.19
H-7"B 2.54 2.75 2.73 2.71 2.79
H-8" 4.81 —c 4.96 C

Coupling Constants J(Hz)

(5—6) 8.3 8.4 8.5 8.5
14
(7—8) 15.9 16.1 15.9 16.0
(2'—6') 2.6 2.4
(5'—6') 8.1 8.1
(7'—8') 4.8 4.7 5.6 5.3
2.0 —2 ——— ———
(2"—6'
(2"—3")+
——— 8.6 8.5 A(340flfi
(2"-5
8.0 7.9 ——— ———
(5—6'
(7"A—7"B) 14.1 —14 14.0 14.6
(7"A—8' 2.8 4.5
(7"B—8") 10.0 —9 9.7 9.9
2.2 2.1 1.9 ———
(2"—6")
——— ——— 8.6
(2"—3")+
(lt 5t) ———
(5"—6") 8.1 8.2 8.0
(7"A—7"B) 14.0 —14 14.0 14.7
2.6 3.5 2
(7"A—8")
(7"B—8") 11.7 —10 11.1 11.2

aThe assignments in systems 1"—9"and 1"—9"' are interchangeable.

b6,856,65 (multiplet).
I I I I I
0 4 8 12 16 20 24frlifl
I
Under residual HDO signal.
dme broad signals preclude accurate determination of coupling constants.
a 6.58—6.54 (multiplet).
'The positions of the systems 1—9' and 1" —9" are interchangeable. Fig. 2 HPLC separation of a crude 80% aqueous MeOH extract from
Complex, three-spin system. Orthosiphon aristatus.

The aqueous MeOH extract of leaves and
stems from 0. an status was subjected to HPLC analysis. All
compounds isolated except flavones I and 9 were separated
within 25mm employing a RP-18 column (125 x 4mm)
(Fig.2). The critical pair I and 9 was only resolved by
switching to a longer HPLC column (250 x 4mm) which in-
creased the analysis time. As flavone I was only a minor
component in all samples of 0. aristatus analysed the shor-
ter separation column was favoured for routine analysis
due to the increased speed of analysis. Several unknown
peaks in the 'flavone region" of the chromatogram may cor-
respond to further lipophilic flavones (e.g. cirsimaritin or
rhamnazin) reported previously for 0. aristatus (6—9) that
were not isolated in this study.
Rosmarinic acid (14) was the predominant
phenolic constituent in the aqueous MeOH extract amount-
ing to 22.2 tmo1Jg dry weight followed by 2,3-dicafTeoyl tar-
trate (16) with 7.4tmo1Ig dry weight (Table 4). The flavones
were by comparison only minor constituents. The major
flavone sinsensetin (9) yielded only 5.3 (kmol!g dry weight.
The predominance of the caffeic acid depsides, compared to
the flavones, was even more pronounced when a hot water
extract comparable to the herbal tea was analysed (Table
4). In the latter extract the caffeic acid derivatives com-
prised almost 95 % of all soluble phenolic constituents
whereas the concentrations of flavones declined sharply
compared to the aqueous MeOH extract.
180 Planta Med. 57(1991) Wahono Sumaryono et aL

Table 4 Absolute amounts (in mol/g dry weight) and concentrations (in % of References
total soluble phenolics) of phenolic constituents from Orthosiphon aristatus.

80 % aqueous Me0H extract hot water extract

Hegnauer, R. (1966) Chemotaxonomie der Pflanzen, Vol. 4, p. 314,
Compound ymol/g dry weight % mol/gdry weight % Birkhäuser Verlag, Basel, Stuttgart.
2
Wichtl, M. (1984) in: Teedrogen (Wichtl, M., ed), p. 244—246, Wis-
1 + + + + senschaftliche Verlagsgeseilschaft mbH, Stuttgart.
2 1.0 1.7 + + DAB-9 Kommentar (1986) (Hartke, K., Mutschler, E., eds.), Vol. 3, p.
3 + + + + 2607—2609, Wissenschaftliche Verlagsgesellschaft mbH, Stuttgart,
4 + + + + Govi-Verlag GmbH, Frankfurt.
5 2.5 4.2 0.3 + Wagner, H. (1982) Pharmazeutische Biologie: Drogen und ihre In-
6 5.5 9.2 0.5 0.1 haltsstoffe, 2nd Bdn., p. 49, Gustav Fischer Verlag, Stuttgart.
7 3.2 5.4 + + Wollenweber, E. (1982) in: The Flavonoids: Advances in Research,
8 1.6 2.7 0.5 0.1 (Harhorne, J. B., Mahry, T. J., eds.), p. 189—259, Chapman and Ball,
9 5.3 8.9 1.3 2.5 London.
6
10 + + + + Bombardelli, E., Bonati, A., Gabetta, B., Mustich, G. (1972)
11 0.5 0.8 0.3 + Fitoterapia 43, 35—40.
12 + + + + Schneider, F., Tan, H. 5. (1973) Dtsch. Apth. Ztg. 113, 201—202.
13 1.5 2.5 1.9 3.6 Wollenweber, E., Mann, K. (1985) Planta Med. 459—460.
14 22.2 37.3 23.8 45.1 Malterud, K. E., Hanche-Olsen, I. M., Smith-Kielland, I. (1989) Planta
15 3.1 5.2 6.6 12.5 Med. 55, 559—560.
16 7.4 12.4 13.3 25.2 10
Nikonov, G. K., Savina, A. A. (1971) Farmatsiya (Moscow) 20, 29—
17 2.2 3.7 0.6 1.1 31.
18 2.1 3.5 2.9 5.5 Soicke, H., Al-Hassan, G., GOrier, K. (1988) Planta Med. 54, 175—

Downloaded by: University of Arizona Library. Copyrighted material.

19 0.7 1.2 0.8 1.5 176.
20 0.7 1.2 + + 12
Bauer, R., Remiger, P., Wagner, H. (1988) Dtsch. Apth. Ztg. 128,
174—180.
Numbers of compounds follow from Fig. 1. + = 0.1. 13
Fukui, H., Yazaki, K., Tabata, M. (1984) Phytochemistry 10, 2398—
2399.
14
Kelley, C. J., Mahajan, J. R., Brooks, L. C., Neubert, L. A., Bre-
nemann, W. R., Carmack, M. (1975) J. Org. Chem. 40, 1804—1815.
Previously it has been proposed that the 11
Guerin, J. C., Reveillere, H-P., Ducrey, P., Toupet, L. (1989) J. Nat.
flavones are responsible for the observed resistance of Prod. 52, 171—173.
aqueous extracts of this plant to rotting caused by microor-
ganisms and hence can be expected to be among the active
constituents of a Orthosiphon tea against bacterial infec-
tions of the bladder or kidney (7, 9). The minute quantities
of flavones in hot aqueous extracts of 0. aristatus, however,
cast doubt on the importance of these compounds for the
antibacterial activity of the tea. In the search for phar-
macologically active compounds in 0. aristatus that are re-
sponsible for the diuretic properties and for antibacterial
activities the predominating caffeic acid derivatives should
be taken into consideration.

Acknowledgements
Financial support of this project was by a grant of
the DFG to P. P. and by a scholarship of the Overseas Fellowship
Program/World Bank (Ministry of Research and Technology, In-
donesia) to W. S. which are gratefully acknowledged. We further
thank C. Kakoschke (GBF) for technical assistance.