Papers 1
Comparative Pharmacological Investigations of A ilium ursinum and A ilium sativum A. Sendi', G. E1b11, B. Steinke', K Redi', W. Breu', andH. Wagner1'2 2
Institute of Pharmaceutical Biology, University of Munich, Karlstr. 29, D(W)-8000 MUnchen 2, Federal Republic of Germany Address for correspondence
Abstract Extracts of wild garlic (A ilium ursinum) and garlic (A. sativum) with defined chemical compositions were investigated for their in vitro inhibitory potential on 5-lipoxygenase (LO), cyclooxygenase (CO), thrombocyte aggregation (TA), and angiotensin I-converting enzyme (ACE). The inhibition rates as IC50 values of both extracts for 5-LO, GO, and TA showed a good correlation with the %-content of the major S-containing compounds (thiosulfinates and ajoenes) of the various extracts. In the 5-LO and CO test the garlic ex-
tracts are slightly superior to the wild garlic extracts whereas, in the TA test, no differences could be found. In the ACE test the water extract of the leaves of wild garlic containing glutamylpeptides showed the highest inhibitory activity followed by that of the garlic leaf and the bulbs of both drugs. The comparative studies underline the usefulness of wild garlic as a substitute of garlic.
are the most prevalent components, while in garlic, allicin is the only major component (see Fig. la and Table 3). The acetone/chloroform extracts of wild garlic and garlic contain, besides thiosulfinates (10—25%), as major components ajoene and its derivatives (10—15%) (see Fig. lb and Table 3). This acetone/chloroform extract of wild garlic distinguishes itself from a garlic extract by having a higher content of methyl- and dimethylajoene and at the same time a lower content of ajoene. The percentage of minor components such as allyl sulfides and vinyldithiins is lower in the acetone/chloroform extract of wild garlic compared to that of garlic.
The quantitative differences of sulfur-containing compounds between identically prepared extracts of wild garlic and garlic are primarily due to taxonomic features and only to a small degree to variation in vegetation, origin, and extraction procedure. Furthermore, it must be pointed out that the total amount of sulfur-containing compounds/dry weight of cloves, is approximately 25 % lower in wild garlic than in garlic.
Key words A ilium ursinum, Allium sativum, wild garlic, garlic, 5 -lipoxygenase, cyclooxygenase, thrombocyte
aggregation, inhibition, ACE, angiotensin I-converting enzyme, pharmacological tests.
Introduction Although wild garlic (Alliurn ursinum L.) is regarded in folk medicine as a substitute drug equivalent in function to garlic (Allium sativum L.) (1, 2), there were no analytical chemical and pharmacological investigations to
support these claims. To address this question, we have
first performed a chemical analysis of different wild garlic clove and leaf extracts and compared them with the corresponding garlic preparations (3, 4). One of the results of this analysis was that the characteristic pattern of secondary sulfur-containing components of the wild garlic clove extracts was largely qualitatively comparable to those of
garlic extracts prepared in the same way (see Fig. 1). However, we found characteristic differences in the quantitative composition. In freshly prepared chloroform extracts of wild garlic, allicin, methylallyl and dimethyl thiosulfinate
This chemical analysis was the basis for the comparative pharmacological investigations of wild garlic and garlic (see Fig. 1). Test models were employed which are relevant to the existing applications of wild garlic and
garlic (hypertension, hyperlipidaemia, hypercholesteraemia, and increased tendency of thrombocyte aggregation) (5—9). Therefore test systems were used which allow us to measure the influence of test compounds on thrombocyte aggregation, arachidonic acid metabolism (cyclooxygenase, CO, and 5-lipoxygenase, LO), as well as a key enzyme involved in blood pressure regulation (angiotensin I-converting enzyme, ACE). We tested the chloroform and the acetone/chloroform extracts for comparison because they differ in their content of thiosulfinates and ajoene derivatives, respectively. The ACE-inhibition was measured with the lyophilized water extracts only, because previous experiments have shown no activities for lipophilic extracts.
Materials and Methods Materials Chemicals and solvents used in the studies were of analytical grade. The labeled [14C]-arachidonic acid was purchased from Amersham Buchler GmbH, Braunschweig FRG, dansyltriglycine synthesized by Hoechst, the other chemicals and sub-
Downloaded by: University of Arizona Library. Copyrighted material.
Received: June 17, 1991
A. Sendletal.
Planta Med. 58 (1 992)
Preparation of the garlic and wild garlic extracts used (a) Chloroform extracts of garlic (GC) and wild garlic (WGC) were prepared according to (3): A. sativum or A. ur-
sinuin cloves were minced in the presence of water in a Warring blendor (wild garlic: 50 g; water: 200 ml), incuhated for 30 minutes at room temperature and then filtered. The filtrate was extracted 3 times, each with ca. 200m1 chloroform p.a. The solvent was quick-
ly evaporated under vacuum (water bath temperature did not exceed 40°C) and the remaining residue was dissolved in a specific volume of ethanol or methanol p.a., as described in the test systems (for concentrations refer to Tables 1 and 2). (b) Acetone/chloroform extracts (GA and WGA)
were prepared as described (3, 4): The cloves (50g) in the presence of water (80 ml) were minced in a Warring blendor and acetone (ratio 1 + 8) was added. After 10 hours with stirring at room
temperature, the extracts were filtered over sodium sulfate and the acetone evaporated. The remaining aqueous residues were extracted 3 times, each with ca. 120 ml chloroform p.a. The extracts were treated further as described under (a).
(c) Water extracts for the ACE assay: The fresh
2.5m1 of a porcine leucocytes suspension, isolated from fresh peripheral blood by dextran sedimentation and centrifugation, in phosphate buffered saline (1.5 x i07 cells/mi) were incubated with the inhibitor dissolved in ethanol or ethanolic control solution (finai ethanol concentration 1.9%). According to the method of Kuhi et al. (35) the lymphocytes were challenged by addition of l2yi of 15MM calcium ionophore A 23187 and 2Syl of 2mM CaC12 in the presence of 0.1 yCi 1-[14C]-arachidonic acid and lOpi of 10MM ETYA, which inhibits the 12-iipoxygenase pathway
of porcine leucocytes. After 5mm the reaction was stopped by adding 525 jzl of 1 % formic acid and its metabolites were extracted with ethyl acetate, separated by reversed-phase HPLC (solvent system see CO test: 50—90% acetonitrile in 20mm, 10mm 90% ace-
tonitrile isocratic elution), detected by radioactivity-monitoring and quantified via peak areas. Effects of the inhibitors were measured by alteration of reduction of 5-HETE biosynthesis (12).
Thrombocyte aggregation test (36) The inhibition of thrombocyte aggregation triggered by PAP or ADP was measured twice as an increase in light transmission and recorded as a percentage of maximum aggregation, which was caused by the solution alone. For the investigation of the inhibitory effects of the preparations, the turbidimetric method of Born and Cross (37) was performed by using a dual channel aggregometer (LAbor. Ahrensburg, Germany).
cloves or the leaves (fresh or dried) were extracted with water (drug:water: 1:10) for six hours at room temperature, then filter-
Fresh blood was collected from the forearm vein of volunteers into plastic tubes containing 3.13% trisodium citrate.
ed and the filtrate lyophilized. The final concentration in the test was 0.33mg/mi buffer (see under ACE assay).
The whole blood was centrifuged at 150g for 10mm at 22°C, in order to produce platelet-rich plasma (Flip) as the upper layer. Platelet-poor plasma (PPP) was obtained by centrifugation of the re-
HPLC analyses were performed with a LiChroCart®-125-4-column with Lichrospher®-RP 18 (Sym) purchased
from Merck; buffer system: acetonitrile/water (20—80% in 30 mm; 210 nm); quantitative determinations were performed as described (3, 4) by using external standards which had been synthesized or isolated from garlic or wild garlic extracts; see Table 3 for contents.
Cyclooxygenase test (33) The inhibitory effect of a substance or extract was measured by the reduced production of prostaglandin E2, which is synthesized by cyclooxygenase from arachidonic acid. Cyclooxygenase activity was tested according to (33); a modified testing procedure has been evaluated and will be published elsewhere (34). 8 yg protein (cyclooxygenase from sheep seminal vesicles) were suspended in 2 ml Tris-HC1 buffer (pH 8.0) contain-
ing 1 mM reduced glutathione, 1 mM adrenalin hitartrate, and
maining blood at 2000g for 10mm. The P1W concentration was adjusted to 255 000/pi with PPP.
The plasma was used within 1.Sh. The aggregometer was calibrated with PPP giving 100% light transmission. 4S,ul of modified Tyrode solution (136.9mM NaC1, 2.68mM KC1, 0.5mMMgCi2 '61120, 1.8mMCaC12' 2H20,0.42mMNaH2PO4'2 1120, 5.55mM glucose ' 1120, 11.9mM NaHCO3) (37) were added to 200pi PEP and stirred at 1000 rpm at 37°C. 25M1 test solution or
solvent were added, before aggregation was induced by adding 2.5 zl PAP-solution (10-6 M final concentration) or 2.5 tzi ADP-solu-
tion (104M final concentration).
The aggregation was recorded as increase in light transmission and observed for at least 3 mm. The samples were dissolved in dimethyl suifoxide or MeOH, and control measurements were performed with dimethyl suifoxide (1 % final concentration) or MeOH. The aggregation of the samples was calculated as a percentage of the maximum aggregation induced by PAP or ADP in PEP containing the solvent in the same concentration.
0.05mM Na2-EDTA. 5mm after incubation (20 °C) with the inhibi-
tor, dissolved in ethanol, or ethanol for control (final concentration: 2.5%) the reaction was started by adding 0.1 yCi I -[14C]-arachidonic acid (37 °C). 20 mm later the reaction was stopped by adding 525 yi of 1 % formic acid. Arachidonic acid and its labeled me-
tabolites were extracted with ethyl acetate, separated by reversed-phase HPLC (acetonitrile/water containing 0.1 % 1 N H3P04, gradient elution: in 12 mm to 36% acetonitrile, in 14 mm to 80%, in 11mm to 10% acetonitrile), detected by radioactivity-monitoring and quantified via peak areas. The quantitation was calibrated against prostaglandin E2.
5-Lipoxygenase test (15) The inhibitory potency of a substance or an extract is measured by the reduced production of 5-HETE, which is enzymatically formed from labeled arachidonic acid by lipoxygenase.
ACE assay (39) Inhibition of angiotensin I-converting enzyme
was measured according to (39) with dansyltriglycine as substrate and purified ACE from rabbit lung as enzymatic source. ACE catalyzes the cleavage of dansyitriglycine into dansyiglycine and digly-
cine. The decrease in concentration of dansyiglycine in the test
reaction, compared with the control reaction, is expressed as percentage inhibition. In the presence of a specific inhibitor, product formation is inhibited partially or totally. 5i.d of inhibitor solution (usually 1 mg/mi buffer or buffer, which may contain up to 10% organic solvent, such as ethanol, methanol or acetone, final concentration: 0.33 mg/nil assay volume) or assay buffer (Hepes-NaOH, 50mM, pH 8.0, containing 300mM NaC1) were placed onto a microtitre plate and mixed
with 25yi enzyme solution (1 Unit angiotensin I-converting enzyme was dissolved in 2.5 ml assay buffer). The microtitre plate
Downloaded by: University of Arizona Library. Copyrighted material.
strates were purchased from Merck, Darmstadt, and Sigma, Deisenhofen FRG.
Planta Med. 58(1992)
Comparative Pharmacological investigations ofAllium ursinum andAllium sativum was placed for 5 mm in a water-bath at 37°C and the reaction was started by the addition of 25 l substrate solution (0.236mM dansyltriglycine). After incubation at 37°C, the enzyme reaction was stopped by the addition of 40 l 0.1 N Na2EDTA. The incubation time depended on the activity of the applied enzyme batch. 10 zl of a 0.354mM dansyl-L-glutamic acid solution in assay buffer (internal standard) were added and the incubation mixtures were transfered to HPLC tubes. The product and unreacted substrate were separated and quantified by reversed phase HPLC with detection
LC
3
of ALCSGASfl.U
215,4
408-
A. ursinUlil
'5
J S
6
at 250 nm. A mixture of acetonitrile and buffer was used as the mobile phase, which separated the reaction product and unreacted substrate within 23 minutes. The decrease in concentration of dansylglycine in the test reaction, compared with the control reaction, was expressed as percentage inhibition, and calculated.
Results and Discussion Inhibition of cyclooxygenase (CO) (see Table 1 and Fig. 2a, 2b)
LC Fl 210,4
of flL7O3F18F1.IJ
form (WGC) and acetone/chloroform (WGA) extracts were significantly less effective than the corresponding garlic extracts (GC and GA), with the acetone/chloroform extracts (WGA and GA) clearly superior to the chloroform extracts. The IC50 values for the wild garlic extracts were 1.15 tg/ml
A. sativurn
S 14
(WGC) and 1.1zg/ml (WGA), while the garlic extracts showed IC50 values of 0.88 jg/ml (GC) and 0.42 g/ml (GA). The total content of thiosulfinates in chloroform extracts of garlic and wild garlic is approximately the same. Allicin is
10
S
LC Fl 210,4
of F1L506F1OFI.IJ
25
20
15
Tine (Olin.)
Fig. lb HPLC profiles of acetone/chloroform extracts of wild garlic and garlic. 2: methylallyl resp. allylmethyl thiosulfinate; 3: dimethylajoene; 4: allicm; 5: methylajoene; 6: ajoene; 14: diallyl trisulfide; garlic and wild garlic: ajoene content 10—15%; thiosulfinate content 10—25%.
loOt Sc
A. ursinum
8c
Table 1 Inhibition of 5-LO- and CO-metabolism by chloroform and acetonef chloroform extracts of wild garlic and garlic. Conc.
GA
[g/ml]
Inhibition %
GC Inhibition %
WGA
WGC
Co
to I
C
20
IS I .
25
12.5 5 1
2 21
0F IRLOOIF102.U
480
A. sativum
0.5 0.25
5.LO 9.6 4.8
3.8 2.4 1.9 0.96 0.57
0.45
100.0
98.3
2.5
65.9 4.9 53.6 6.8 35.8 7.8 —
— —
100.0
4.9 9.3 8.9 6.6
—
50.3
2.3 10
47.2 13 4.6
—
100.0
—
88.1
0 8.3
40.8
94.5 84.6 44.6 26.8 16.3
100.0 0 71.2
13.2
45.7
2.9 7.2
—
—
—
—
15.6 0.8
—
—
—
—
—
83.7 59.9
3.3 6.5 4.6 9.2
7.0
—
—
—
4.1 4.1 5.8
86.4
18.6
100.0
100.0 0 96.7 70.5 20.2
99.3 84.9 44.0 28.7 17.2
5.1
lot
S
iS
15
Tie,, (,ntr,. )
Fig. la HPLC profiles of chloroform extracts of wild garlic and garlic. 1: dimethyl thiosulfinate (DMTS); 2: methylallyl resp. allylmethyl thiosulfinate (MATS); 4: allicin diallyl thiosulfinate; the total content of thiosulfinates in each extract was about 80 to 90%; see Table 3 for contents.
the major compound in garlic chloroform extracts, but in wild garlic chloroform extracts allicin represents only half of the total thiosulfinate content (see Table 3). Based on these results allicin appears to be a stronger inhibitor of cyclooxygenase than methylallyl- and dimethyl thiosulfinates (see Figs. Ia and lb and Table 3 with contents).
Downloaded by: University of Arizona Library. Copyrighted material.
The results of the cyclooxygenase test show
that at every concentration tested the wild garlic chloro-
Planta Med. 58(1992)
A. Sendi et al.
Inhibition (°/) Fig.2a Cyclooxygenase inhibition of freshly prepared chloroform extracts of garlic and wild garlic (GC and WGC); IC50 values of GC and WGC
0.88 and 1.l5tg/mI.
100
conc. hig/mil
120
A mean value
Inhibition (°/.) Fig. 2b Cyclooxygenase inhibition of acetone/chloroform extracts of garlic and wild garlic (GA and WGA); IC50 values of GA and WGA:
0.42 and 1.ljg/ml.
100
conc. hig/mil * mean value
A mean value
In contrast, ajoenes (E,Z-ajoene, E,Zmethyl- and dimethylajoene) are present in the highest percentage in the acetone/chloroform extracts of both drugs. In wild garlic, the concentration of methyl- and dimethylajoene is three times higher than in garlic whose acetone/chloroform extract contains more ajoene. The IC50 values for methyl- and dimethylajoene lay between 7—15MM
(8.2 and 13j.tM), while that for ajoene is 5flM. In previous studies, the following inhibition values were obtained for
Inhibition of5-lipoxygenase (5-LO) (see Table I and Fig. 3) Similar relationships of values were obtained with the same extracts in the 5-lipoxygenase test. The IC50 values for wild garlic and garlic chloroform extracts, WGC and GC, almost exactly corresponded at 2.85 and 2.95 tg/ml, respectively, and those for the acetone extracts were 0.81 tg/ml (WGA) and 0.51 ig/ml (GA). As ajoene has
several garlic components tested at a concentration of a lower IC50 value in the 5-LO test (1.5MM) than methyl50uM (10): ajoene 96.4%, diallyl disulfide 69.1%, diallyl thiosuffinate 67%, and the dithiins 3-vinyl-4H-1,2-dithiin and 2-vinyl-4H-1,3-dithiin up to 5.3%. The higher CO-inhibitory activity of acetone/chloroform extracts of garlic as compared to those of wild garlic might be explained by the higher content of ajoene and diallyl trisulfide in the garlic extracts.
and dimethylajoene (10, 11) (IC50 values of 1.6 and 4.2DM,
respectively) and the wild garlic extracts contain less ajoene than the corresponding garlic extracts, it is conceivable that this difference meets with the lower inhibitory rate of wild garlic extracts. Allicin, the major thiosulfinate in garlic (see Table 3), has an IC50 value of 25 tM (12, 16). However,
diallyl trisulfide, which accounts for approximately 15—29% of the garlic acetone/chloroform extract (GA) but is hardly detectable in WGA extracts, also clearly inhibits lipoxygenase (13). Studies of LU-inhibition by garlic components by us and others showed that several of them inhibit 5-LU in about the same concentration range as known
Downloaded by: University of Arizona Library. Copyrighted material.
* mean value
Planta Mcd. 58(1992)
Comparative Pharmacological investigations ofAllium ursinum and Allium sativum
reference inhibitors, for example nordihydroguaretic acid. Ajoene, methyl- and dimethylajoene, diallyl trisulfide, allyl
methyl trisulfide, and methyl propyl trisulfide, among others, belong to this group of potent garlic inhibitors.
inhibition of thrombocyte aggregation (see Table 2 and Fig. 4)
cible results. The three tested batches of wild garlic and garlic chloroform extracts (different extract batches) inhibited the ADP-induced thrombocyte aggregation complete-
ly, while both acetone/chloroform extracts inhibited it to only 30 to 36%. For the PAF-induced thrombocyte aggregation, the following values were obtained for three different batches: WGC 81—87%, GC 81—94%, WGA 24—29%,
and GA 24—26% (all tested at the same concentration 60zg/ml reaction mixture). As allicin and ajoene are known to be two strong aggregation inhibitors (17— 22), these inhibitors and their homologues methylallyl thiosulfinate and methyl- and dimethylajoene are most likely to be responsible for this effect exhibited by both extracts. Adenosine, which is also known for its strong platelet aggregation inhibitory effect, was not detectable in either extract.
Our results confirm previous studies on the inhibition of thrombocyte aggregation by garlic extracts (27—31), although the composition of the extracts used in these studies was not determined.
Inhibition (0/0) Fig.3 5-Lipoxygenase inhibition of chloroform and acetone/chloroform extracts of garlic and
10o
wild garlic (GA and WGA); IC50 values (GC and WGC; GA and WGA): 2.95 and 2.85; 0.51 and
80
0.81 ig/ml.
60 40 20 0
o,1 o,el .i
0,1
II II II II
2,852,95
10
Conc.(jig/mL)
•GA AGC WGA 0WGC 120
inhibition t%l Fig. 4 Inhibition of platelet aggregation: comparison of garlic and wild garlic chloroform and acetone/chloroform extracts; shown is one batch as representative of two others; aggregation inducer: PAF 10-6 M, ADP iO M; conc. of the extracts: 60zg/mI.
-r
100
80 60 40
J
20
0
WGC
GC
lilt PAP ADP
Downloaded by: University of Arizona Library. Copyrighted material.
In contrast to the results of the CO and LO tests, the chloroform extracts of both plants inhibit thrombocyte aggregation to a greater extent than the acetone / chloroform extracts (GA and WGA). At a concentration of 60 jtg/m1 reaction mixture, the chloroform extracts show an approximately 3-fold stronger inhibition of thrombocyte aggregation than the acetone/chloroform extracts. These results were obtained with PAF- (platelet activating factor) as well as with ADP-induced thrombocyte aggregation. Comparison of the effects of wild garlic and garlic extracts prepared during different seasons showed nearly reprodu-
"U
5
A. Sendl et. al
Planta Med. 58 (1992) Table 2 Inhibition of platelet aggregation: comparison of garlic and wild garlic extracts.
WGC GC WGA
GA
84.17 86.57 26.56 24.70
2.80 6.64 3.33' 1.24*
ADP 10 M Extract
Inhibition [%]
conc. 60 g/ml WGC GC WGA
GA
99.94 0.11* 99.89 0.19* 30.46 35.55
* Mean values of different batches.
Summarizing the comparative studies it becomes evident that the major sulfur-containing com-
Table 3 Range of contents of thiosulfinates (determined in chloroform extracts by HPLC using isolated compounds as external standards) and cysteine
sulfoxides (determined in MeOH/H20 extracts by HPLC using synthesized compounds as external standards) in different batches of garlic and wild garlic bulbs; data refer to dry weight; MATS: methylallyl- resp. allylmethyl thiosulfinate; MCSO: methyl-L-cysteine sulf oxide.
Compound
wild garlic
garlic
alum
0.65—1.10% 0.23—0.59% 0.55—0.81% 0.60—0.90%
0.82—2.50% 0.60—1.37% 0.14—0.23% 0.15—0.42%
allicin MATS MCSO
It is remarkable that the IC50 values of the single extracts are in general in the same order of magnitude as the isolated major compounds of these extracts. This leads us to assume that synergistic effects exist between the single compounds in the extracts.
ACE-inhibition
In the ACE-inhibition test, lyophilized aqueous extracts of A. ursinum and A. sativum cloves were only weakly active. In contrast, using an aqueous, lyophilized wild garlic leaf extract at a concentration of 0.3 mg/ml we obtained an inhibition value of 58 %, in comparison to 30% for a garlic leaf extract. Based on a meanwhile per-
formed fractionation and structure elucidation, y-glutamylcysteine derivatives are responsible for this inhibition (32). The effect remains unchanged if dried, frozen, or lyo-
pounds of the various extracts are responsible for the investigated pharmacological activities and that the individual
results correlate in every respect with the characteristic chemical composition of the single extracts. As far as the ACE inhibitory potential of the lyophilized water extracts is concerned, it could be substantiated that their activity derives from the presence of y-glutamylallylcysteine sulfoxide and other glutamylpeptides. The higher ACE-inhibiting ac-
tivity of wild garlic extract is consistant with the higher content of this type of compounds.
As most studies on the effects of garlic already have been carried out on humans, one can conclude that wild garlic bulbs and the preparations from them can serve as a substitute for garlic preparations, if they are applied in appropriately higher doses.
At the same time, this comparative study shows that some or all applied tests may be used to stand-
ardize extracts of wild garlic and garlic biologically or pharmacologically, if equal test conditions can be guaranteed. The inhibitory effect of wild garlic leaf extracts in the in vitro ACE test has relevance to the described effect of reducing blood pressure, if this effectiveness can also be confirmed in in vivo experiments.
philized leaves are used instead of fresh ones or if the References
aqueous extract is heated. Also, an inhibition of alliinase causes no reduction of the inhibitory effect. Summary and Discussion
2
Due to the nearly identical qualitative chemical composition of the corresponding garlic and wild garlic chloroform and acetone/chloroform extracts, the extracts of wild garlic were expected to show similar pharmacological profiles to that of garlic extracts in the in vitro tests used. The results presented have confirmed this assumption. The CO IC50 values of the chloroform extracts
6
were found to be in the same order of magnitude (0.88, 1.15 ig/ml) with superior activity of the garlic extract. The IC50 values of the corresponding acetone/chloroform extracts lay again in about the same inhibition range (0.42, 1.1 zg/m1), however this time with a distinctively higher inhibition rate of the garlic extract. In contrast the same chloroform extracts showed three times lower IC50 values (2.85, 2.95 jg/ml) in the 5-LO test than the acetone/chloroform extracts (0.5, 0.8 g/ml). This result is consistant with
the observation that ajoene and derivatives are better 5LO inhibitors than the thiosulfinates. Generally the thrombocyte aggregation inhibition test revealed no differences between the wild garlic and garlic extracts, However, the inhibiting potential of the chloroform extracts was three
Hoppe, H. A. (1975) Drogenkunde, 8. Auflage. De Gruyter-Verlag, Berlin. Heinz, U. J. (1984) Das Handbuch der modernen Pflanzenheilkunde, Hermann Bauer Verlag, Freiburg, p. 258. Wagner, H., Send!, A. (1990) Dtsch. Apoth. Ztg. 130, 1809. Wagner, H., Send!, A. (1991) P!anta Med. 57, 361. Harenberg, J., Giese, C., Zimmermann, R. (1988) Atherosclerosis 74, 247. Kleijnen, J., Knipschild, P., Terriet, G. (1989) Br. J. Clin. Pharmac. 28, 535. Ernst, E., Weihmayr, T., Matrai, A. (1985) Br. med. J., 291. Bordia, A., Sharma, T., Parmar, Y., Verma, S. (1982) Indian HeartJ.
34,86. Lau, B. H. S., Adetumbi, M.. Sanchez, A. (1983) Nutrition Research
3, 119. ° Wagner, H., Wierer, M., Fessler, B. (1987) Planta Med. 305. Wagner, H., Breu, W., Send!, A., Burgi, C., RGdi, P. (1990) Planta Med. 56, 665. 12
13 14 15 16
17
Wagner, H., Dorsch, W., Bayer, Th., Breu, W., Willer, F. (1990) Pros-
tag!andins, Leukotrienes and Essential Fatty Acids 39, 59. Belman, S., Solomon, J., Sega!, A., Block, E., Barany, G. (1989) J. Biochem. Toxicology 4, 151. Block, E., !yer, R., Grisoni, S., Saha, C., Belman, S., Lossing, F.
(1988) J. Am. Chem. Soc. 110, 7813. Wagner, H., Fessler, B. (1986) Planta Med. 374. Wagner, H., Breu, W. (1989) Dtsch. Apoth. Ztg. Supplement 15, 21. Mayeux, P. R., Agrawa!, K., Tou, J., King, B., Llppton, L., Hyman, A., Kdowitz, P., McNamara, D. (1988) Agents and Actions 25 1/2, 182.
Downloaded by: University of Arizona Library. Copyrighted material.
PAF 10-6 M Inhibition [%] conc. 60 yg/ml Extract
times greater than that of the acetone/chloroform extracts. This again indicates, that the thiosulfinates possess higher TA inhibitory activities than the ajoenes.
Comparative Pharmacological Investigations ofAlliam arsinam and Alliam sativam
19
J. Aim Chem. Soc. 108, 7045. Apitz-Castro, F., Ledezma, F., Escalante, J., Jorquera, A., Pinate, F.,
28 Srivastava, K. C. (1986) Prostaglandtns Leukotrienes and Medicine 29
Moreno-Rea, J., Carillo, G., Lea!, 0., Jam, M. K. (1988) Drug Re28 31 22
23 24 21 26 27
search 38(11), 901. Fazal Mohammed, S., Woodward, S. (1986) Thromb. Res. 44, 793. Apitz-Castro, F., Escalante, J., Vargas, R., Jam, M. (1986) Thromb. Res. 42, 303. Rendu, F., Dave!oose, D., Debouzy, J., Bourdeau, N., Levy-To!edano, S., Jam, M., Apitz-Castro, R. (1989) Biochem. Pharmaco!ogy 38,
31
8, 1321. Makheja, A., Bailey, J. (1990) Agents and Actions 29 3/4, 360. Apitz-Castro, F., Cabrera, S., Cruz, M., Ledezma, E., Jam, M. (1983)
Thrombosis Res. 32, 155. Okuyama, T., Fujita, K., Shibata, S., Hoson, M., Kawada, T., Masaki, M., Yamate, N. (1989) Planta Med. 55, 242. Liakopou!ou-Kyriakides, M. (1985) Phytochemistry 24,7 1593. Al!, M., Thomson, M., Alnaqeeb. M., Al-Hassan, J., Khater, S., Go-
mes, 5. (1990) Prostag!andins Leukotrienes and Essential Fatty Acids 41,95.
7
31 36
22, 313. All, M., Mohammed Suad, Y. (1986) Prostaglandins Leukotrienes and Medicine 25, 139. Gaflen, J., Tavapes, I., Bennett, A. (1984) J. Pharm. Pharmaco!. 36, 272. Ariga, T., Oshiba, S., Tamada, T. (1981) The Lancet 7, 150. Wagner, H., Fib!, G., Send!, A. (1991) poster presented on the I!. International gar!ic symposium, 7.— 10. 3. 1991, Ber!in. Wagner, H., Wierer, M., Bauer, R. (1986) P!anta Med. 184. Wagner, H., Red!, K., publication in preparation. Kuh!, P., Shi!oh, R., Murwski, U., Zi!!iken, F. (1984) Prostag!andins 28, 783. Wagner, H., Steinke, B., (1991) Munch. Med. Wschr. 133, Supp!. 1, 54—57.
38
Born, G. V. F., Cross, M. J. (1963)J. Phys!o!. 168, 178—195. Heuer, H., Boehringer Inge!heim KG, persona! communication. Wagner, H., Bib!, G. (1991) Pianta Med. 57, 137.
Downloaded by: University of Arizona Library. Copyrighted material.
18 Block, E., Ahmad, S., Catalfamo, J., Jam, M., Apitz-Castro, R. (1986)
Planta Med. 58(1992)
Comparative Pharmacological Investigations of A ilium ursinum and A ilium sativum A. Sendi', G. E1b11, B. Steinke', K Redi', W. Breu', andH. Wagner1'2 2
Institute of Pharmaceutical Biology, University of Munich, Karlstr. 29, D(W)-8000 MUnchen 2, Federal Republic of Germany Address for correspondence
Abstract Extracts of wild garlic (A ilium ursinum) and garlic (A. sativum) with defined chemical compositions were investigated for their in vitro inhibitory potential on 5-lipoxygenase (LO), cyclooxygenase (CO), thrombocyte aggregation (TA), and angiotensin I-converting enzyme (ACE). The inhibition rates as IC50 values of both extracts for 5-LO, GO, and TA showed a good correlation with the %-content of the major S-containing compounds (thiosulfinates and ajoenes) of the various extracts. In the 5-LO and CO test the garlic ex-
tracts are slightly superior to the wild garlic extracts whereas, in the TA test, no differences could be found. In the ACE test the water extract of the leaves of wild garlic containing glutamylpeptides showed the highest inhibitory activity followed by that of the garlic leaf and the bulbs of both drugs. The comparative studies underline the usefulness of wild garlic as a substitute of garlic.
are the most prevalent components, while in garlic, allicin is the only major component (see Fig. la and Table 3). The acetone/chloroform extracts of wild garlic and garlic contain, besides thiosulfinates (10—25%), as major components ajoene and its derivatives (10—15%) (see Fig. lb and Table 3). This acetone/chloroform extract of wild garlic distinguishes itself from a garlic extract by having a higher content of methyl- and dimethylajoene and at the same time a lower content of ajoene. The percentage of minor components such as allyl sulfides and vinyldithiins is lower in the acetone/chloroform extract of wild garlic compared to that of garlic.
The quantitative differences of sulfur-containing compounds between identically prepared extracts of wild garlic and garlic are primarily due to taxonomic features and only to a small degree to variation in vegetation, origin, and extraction procedure. Furthermore, it must be pointed out that the total amount of sulfur-containing compounds/dry weight of cloves, is approximately 25 % lower in wild garlic than in garlic.
Key words A ilium ursinum, Allium sativum, wild garlic, garlic, 5 -lipoxygenase, cyclooxygenase, thrombocyte
aggregation, inhibition, ACE, angiotensin I-converting enzyme, pharmacological tests.
Introduction Although wild garlic (Alliurn ursinum L.) is regarded in folk medicine as a substitute drug equivalent in function to garlic (Allium sativum L.) (1, 2), there were no analytical chemical and pharmacological investigations to
support these claims. To address this question, we have
first performed a chemical analysis of different wild garlic clove and leaf extracts and compared them with the corresponding garlic preparations (3, 4). One of the results of this analysis was that the characteristic pattern of secondary sulfur-containing components of the wild garlic clove extracts was largely qualitatively comparable to those of
garlic extracts prepared in the same way (see Fig. 1). However, we found characteristic differences in the quantitative composition. In freshly prepared chloroform extracts of wild garlic, allicin, methylallyl and dimethyl thiosulfinate
This chemical analysis was the basis for the comparative pharmacological investigations of wild garlic and garlic (see Fig. 1). Test models were employed which are relevant to the existing applications of wild garlic and
garlic (hypertension, hyperlipidaemia, hypercholesteraemia, and increased tendency of thrombocyte aggregation) (5—9). Therefore test systems were used which allow us to measure the influence of test compounds on thrombocyte aggregation, arachidonic acid metabolism (cyclooxygenase, CO, and 5-lipoxygenase, LO), as well as a key enzyme involved in blood pressure regulation (angiotensin I-converting enzyme, ACE). We tested the chloroform and the acetone/chloroform extracts for comparison because they differ in their content of thiosulfinates and ajoene derivatives, respectively. The ACE-inhibition was measured with the lyophilized water extracts only, because previous experiments have shown no activities for lipophilic extracts.
Materials and Methods Materials Chemicals and solvents used in the studies were of analytical grade. The labeled [14C]-arachidonic acid was purchased from Amersham Buchler GmbH, Braunschweig FRG, dansyltriglycine synthesized by Hoechst, the other chemicals and sub-
Downloaded by: University of Arizona Library. Copyrighted material.
Received: June 17, 1991
A. Sendletal.
Planta Med. 58 (1 992)
Preparation of the garlic and wild garlic extracts used (a) Chloroform extracts of garlic (GC) and wild garlic (WGC) were prepared according to (3): A. sativum or A. ur-
sinuin cloves were minced in the presence of water in a Warring blendor (wild garlic: 50 g; water: 200 ml), incuhated for 30 minutes at room temperature and then filtered. The filtrate was extracted 3 times, each with ca. 200m1 chloroform p.a. The solvent was quick-
ly evaporated under vacuum (water bath temperature did not exceed 40°C) and the remaining residue was dissolved in a specific volume of ethanol or methanol p.a., as described in the test systems (for concentrations refer to Tables 1 and 2). (b) Acetone/chloroform extracts (GA and WGA)
were prepared as described (3, 4): The cloves (50g) in the presence of water (80 ml) were minced in a Warring blendor and acetone (ratio 1 + 8) was added. After 10 hours with stirring at room
temperature, the extracts were filtered over sodium sulfate and the acetone evaporated. The remaining aqueous residues were extracted 3 times, each with ca. 120 ml chloroform p.a. The extracts were treated further as described under (a).
(c) Water extracts for the ACE assay: The fresh
2.5m1 of a porcine leucocytes suspension, isolated from fresh peripheral blood by dextran sedimentation and centrifugation, in phosphate buffered saline (1.5 x i07 cells/mi) were incubated with the inhibitor dissolved in ethanol or ethanolic control solution (finai ethanol concentration 1.9%). According to the method of Kuhi et al. (35) the lymphocytes were challenged by addition of l2yi of 15MM calcium ionophore A 23187 and 2Syl of 2mM CaC12 in the presence of 0.1 yCi 1-[14C]-arachidonic acid and lOpi of 10MM ETYA, which inhibits the 12-iipoxygenase pathway
of porcine leucocytes. After 5mm the reaction was stopped by adding 525 jzl of 1 % formic acid and its metabolites were extracted with ethyl acetate, separated by reversed-phase HPLC (solvent system see CO test: 50—90% acetonitrile in 20mm, 10mm 90% ace-
tonitrile isocratic elution), detected by radioactivity-monitoring and quantified via peak areas. Effects of the inhibitors were measured by alteration of reduction of 5-HETE biosynthesis (12).
Thrombocyte aggregation test (36) The inhibition of thrombocyte aggregation triggered by PAP or ADP was measured twice as an increase in light transmission and recorded as a percentage of maximum aggregation, which was caused by the solution alone. For the investigation of the inhibitory effects of the preparations, the turbidimetric method of Born and Cross (37) was performed by using a dual channel aggregometer (LAbor. Ahrensburg, Germany).
cloves or the leaves (fresh or dried) were extracted with water (drug:water: 1:10) for six hours at room temperature, then filter-
Fresh blood was collected from the forearm vein of volunteers into plastic tubes containing 3.13% trisodium citrate.
ed and the filtrate lyophilized. The final concentration in the test was 0.33mg/mi buffer (see under ACE assay).
The whole blood was centrifuged at 150g for 10mm at 22°C, in order to produce platelet-rich plasma (Flip) as the upper layer. Platelet-poor plasma (PPP) was obtained by centrifugation of the re-
HPLC analyses were performed with a LiChroCart®-125-4-column with Lichrospher®-RP 18 (Sym) purchased
from Merck; buffer system: acetonitrile/water (20—80% in 30 mm; 210 nm); quantitative determinations were performed as described (3, 4) by using external standards which had been synthesized or isolated from garlic or wild garlic extracts; see Table 3 for contents.
Cyclooxygenase test (33) The inhibitory effect of a substance or extract was measured by the reduced production of prostaglandin E2, which is synthesized by cyclooxygenase from arachidonic acid. Cyclooxygenase activity was tested according to (33); a modified testing procedure has been evaluated and will be published elsewhere (34). 8 yg protein (cyclooxygenase from sheep seminal vesicles) were suspended in 2 ml Tris-HC1 buffer (pH 8.0) contain-
ing 1 mM reduced glutathione, 1 mM adrenalin hitartrate, and
maining blood at 2000g for 10mm. The P1W concentration was adjusted to 255 000/pi with PPP.
The plasma was used within 1.Sh. The aggregometer was calibrated with PPP giving 100% light transmission. 4S,ul of modified Tyrode solution (136.9mM NaC1, 2.68mM KC1, 0.5mMMgCi2 '61120, 1.8mMCaC12' 2H20,0.42mMNaH2PO4'2 1120, 5.55mM glucose ' 1120, 11.9mM NaHCO3) (37) were added to 200pi PEP and stirred at 1000 rpm at 37°C. 25M1 test solution or
solvent were added, before aggregation was induced by adding 2.5 zl PAP-solution (10-6 M final concentration) or 2.5 tzi ADP-solu-
tion (104M final concentration).
The aggregation was recorded as increase in light transmission and observed for at least 3 mm. The samples were dissolved in dimethyl suifoxide or MeOH, and control measurements were performed with dimethyl suifoxide (1 % final concentration) or MeOH. The aggregation of the samples was calculated as a percentage of the maximum aggregation induced by PAP or ADP in PEP containing the solvent in the same concentration.
0.05mM Na2-EDTA. 5mm after incubation (20 °C) with the inhibi-
tor, dissolved in ethanol, or ethanol for control (final concentration: 2.5%) the reaction was started by adding 0.1 yCi I -[14C]-arachidonic acid (37 °C). 20 mm later the reaction was stopped by adding 525 yi of 1 % formic acid. Arachidonic acid and its labeled me-
tabolites were extracted with ethyl acetate, separated by reversed-phase HPLC (acetonitrile/water containing 0.1 % 1 N H3P04, gradient elution: in 12 mm to 36% acetonitrile, in 14 mm to 80%, in 11mm to 10% acetonitrile), detected by radioactivity-monitoring and quantified via peak areas. The quantitation was calibrated against prostaglandin E2.
5-Lipoxygenase test (15) The inhibitory potency of a substance or an extract is measured by the reduced production of 5-HETE, which is enzymatically formed from labeled arachidonic acid by lipoxygenase.
ACE assay (39) Inhibition of angiotensin I-converting enzyme
was measured according to (39) with dansyltriglycine as substrate and purified ACE from rabbit lung as enzymatic source. ACE catalyzes the cleavage of dansyitriglycine into dansyiglycine and digly-
cine. The decrease in concentration of dansyiglycine in the test
reaction, compared with the control reaction, is expressed as percentage inhibition. In the presence of a specific inhibitor, product formation is inhibited partially or totally. 5i.d of inhibitor solution (usually 1 mg/mi buffer or buffer, which may contain up to 10% organic solvent, such as ethanol, methanol or acetone, final concentration: 0.33 mg/nil assay volume) or assay buffer (Hepes-NaOH, 50mM, pH 8.0, containing 300mM NaC1) were placed onto a microtitre plate and mixed
with 25yi enzyme solution (1 Unit angiotensin I-converting enzyme was dissolved in 2.5 ml assay buffer). The microtitre plate
Downloaded by: University of Arizona Library. Copyrighted material.
strates were purchased from Merck, Darmstadt, and Sigma, Deisenhofen FRG.
Planta Med. 58(1992)
Comparative Pharmacological investigations ofAllium ursinum andAllium sativum was placed for 5 mm in a water-bath at 37°C and the reaction was started by the addition of 25 l substrate solution (0.236mM dansyltriglycine). After incubation at 37°C, the enzyme reaction was stopped by the addition of 40 l 0.1 N Na2EDTA. The incubation time depended on the activity of the applied enzyme batch. 10 zl of a 0.354mM dansyl-L-glutamic acid solution in assay buffer (internal standard) were added and the incubation mixtures were transfered to HPLC tubes. The product and unreacted substrate were separated and quantified by reversed phase HPLC with detection
LC
3
of ALCSGASfl.U
215,4
408-
A. ursinUlil
'5
J S
6
at 250 nm. A mixture of acetonitrile and buffer was used as the mobile phase, which separated the reaction product and unreacted substrate within 23 minutes. The decrease in concentration of dansylglycine in the test reaction, compared with the control reaction, was expressed as percentage inhibition, and calculated.
Results and Discussion Inhibition of cyclooxygenase (CO) (see Table 1 and Fig. 2a, 2b)
LC Fl 210,4
of flL7O3F18F1.IJ
form (WGC) and acetone/chloroform (WGA) extracts were significantly less effective than the corresponding garlic extracts (GC and GA), with the acetone/chloroform extracts (WGA and GA) clearly superior to the chloroform extracts. The IC50 values for the wild garlic extracts were 1.15 tg/ml
A. sativurn
S 14
(WGC) and 1.1zg/ml (WGA), while the garlic extracts showed IC50 values of 0.88 jg/ml (GC) and 0.42 g/ml (GA). The total content of thiosulfinates in chloroform extracts of garlic and wild garlic is approximately the same. Allicin is
10
S
LC Fl 210,4
of F1L506F1OFI.IJ
25
20
15
Tine (Olin.)
Fig. lb HPLC profiles of acetone/chloroform extracts of wild garlic and garlic. 2: methylallyl resp. allylmethyl thiosulfinate; 3: dimethylajoene; 4: allicm; 5: methylajoene; 6: ajoene; 14: diallyl trisulfide; garlic and wild garlic: ajoene content 10—15%; thiosulfinate content 10—25%.
loOt Sc
A. ursinum
8c
Table 1 Inhibition of 5-LO- and CO-metabolism by chloroform and acetonef chloroform extracts of wild garlic and garlic. Conc.
GA
[g/ml]
Inhibition %
GC Inhibition %
WGA
WGC
Co
to I
C
20
IS I .
25
12.5 5 1
2 21
0F IRLOOIF102.U
480
A. sativum
0.5 0.25
5.LO 9.6 4.8
3.8 2.4 1.9 0.96 0.57
0.45
100.0
98.3
2.5
65.9 4.9 53.6 6.8 35.8 7.8 —
— —
100.0
4.9 9.3 8.9 6.6
—
50.3
2.3 10
47.2 13 4.6
—
100.0
—
88.1
0 8.3
40.8
94.5 84.6 44.6 26.8 16.3
100.0 0 71.2
13.2
45.7
2.9 7.2
—
—
—
—
15.6 0.8
—
—
—
—
—
83.7 59.9
3.3 6.5 4.6 9.2
7.0
—
—
—
4.1 4.1 5.8
86.4
18.6
100.0
100.0 0 96.7 70.5 20.2
99.3 84.9 44.0 28.7 17.2
5.1
lot
S
iS
15
Tie,, (,ntr,. )
Fig. la HPLC profiles of chloroform extracts of wild garlic and garlic. 1: dimethyl thiosulfinate (DMTS); 2: methylallyl resp. allylmethyl thiosulfinate (MATS); 4: allicin diallyl thiosulfinate; the total content of thiosulfinates in each extract was about 80 to 90%; see Table 3 for contents.
the major compound in garlic chloroform extracts, but in wild garlic chloroform extracts allicin represents only half of the total thiosulfinate content (see Table 3). Based on these results allicin appears to be a stronger inhibitor of cyclooxygenase than methylallyl- and dimethyl thiosulfinates (see Figs. Ia and lb and Table 3 with contents).
Downloaded by: University of Arizona Library. Copyrighted material.
The results of the cyclooxygenase test show
that at every concentration tested the wild garlic chloro-
Planta Med. 58(1992)
A. Sendi et al.
Inhibition (°/) Fig.2a Cyclooxygenase inhibition of freshly prepared chloroform extracts of garlic and wild garlic (GC and WGC); IC50 values of GC and WGC
0.88 and 1.l5tg/mI.
100
conc. hig/mil
120
A mean value
Inhibition (°/.) Fig. 2b Cyclooxygenase inhibition of acetone/chloroform extracts of garlic and wild garlic (GA and WGA); IC50 values of GA and WGA:
0.42 and 1.ljg/ml.
100
conc. hig/mil * mean value
A mean value
In contrast, ajoenes (E,Z-ajoene, E,Zmethyl- and dimethylajoene) are present in the highest percentage in the acetone/chloroform extracts of both drugs. In wild garlic, the concentration of methyl- and dimethylajoene is three times higher than in garlic whose acetone/chloroform extract contains more ajoene. The IC50 values for methyl- and dimethylajoene lay between 7—15MM
(8.2 and 13j.tM), while that for ajoene is 5flM. In previous studies, the following inhibition values were obtained for
Inhibition of5-lipoxygenase (5-LO) (see Table I and Fig. 3) Similar relationships of values were obtained with the same extracts in the 5-lipoxygenase test. The IC50 values for wild garlic and garlic chloroform extracts, WGC and GC, almost exactly corresponded at 2.85 and 2.95 tg/ml, respectively, and those for the acetone extracts were 0.81 tg/ml (WGA) and 0.51 ig/ml (GA). As ajoene has
several garlic components tested at a concentration of a lower IC50 value in the 5-LO test (1.5MM) than methyl50uM (10): ajoene 96.4%, diallyl disulfide 69.1%, diallyl thiosuffinate 67%, and the dithiins 3-vinyl-4H-1,2-dithiin and 2-vinyl-4H-1,3-dithiin up to 5.3%. The higher CO-inhibitory activity of acetone/chloroform extracts of garlic as compared to those of wild garlic might be explained by the higher content of ajoene and diallyl trisulfide in the garlic extracts.
and dimethylajoene (10, 11) (IC50 values of 1.6 and 4.2DM,
respectively) and the wild garlic extracts contain less ajoene than the corresponding garlic extracts, it is conceivable that this difference meets with the lower inhibitory rate of wild garlic extracts. Allicin, the major thiosulfinate in garlic (see Table 3), has an IC50 value of 25 tM (12, 16). However,
diallyl trisulfide, which accounts for approximately 15—29% of the garlic acetone/chloroform extract (GA) but is hardly detectable in WGA extracts, also clearly inhibits lipoxygenase (13). Studies of LU-inhibition by garlic components by us and others showed that several of them inhibit 5-LU in about the same concentration range as known
Downloaded by: University of Arizona Library. Copyrighted material.
* mean value
Planta Mcd. 58(1992)
Comparative Pharmacological investigations ofAllium ursinum and Allium sativum
reference inhibitors, for example nordihydroguaretic acid. Ajoene, methyl- and dimethylajoene, diallyl trisulfide, allyl
methyl trisulfide, and methyl propyl trisulfide, among others, belong to this group of potent garlic inhibitors.
inhibition of thrombocyte aggregation (see Table 2 and Fig. 4)
cible results. The three tested batches of wild garlic and garlic chloroform extracts (different extract batches) inhibited the ADP-induced thrombocyte aggregation complete-
ly, while both acetone/chloroform extracts inhibited it to only 30 to 36%. For the PAF-induced thrombocyte aggregation, the following values were obtained for three different batches: WGC 81—87%, GC 81—94%, WGA 24—29%,
and GA 24—26% (all tested at the same concentration 60zg/ml reaction mixture). As allicin and ajoene are known to be two strong aggregation inhibitors (17— 22), these inhibitors and their homologues methylallyl thiosulfinate and methyl- and dimethylajoene are most likely to be responsible for this effect exhibited by both extracts. Adenosine, which is also known for its strong platelet aggregation inhibitory effect, was not detectable in either extract.
Our results confirm previous studies on the inhibition of thrombocyte aggregation by garlic extracts (27—31), although the composition of the extracts used in these studies was not determined.
Inhibition (0/0) Fig.3 5-Lipoxygenase inhibition of chloroform and acetone/chloroform extracts of garlic and
10o
wild garlic (GA and WGA); IC50 values (GC and WGC; GA and WGA): 2.95 and 2.85; 0.51 and
80
0.81 ig/ml.
60 40 20 0
o,1 o,el .i
0,1
II II II II
2,852,95
10
Conc.(jig/mL)
•GA AGC WGA 0WGC 120
inhibition t%l Fig. 4 Inhibition of platelet aggregation: comparison of garlic and wild garlic chloroform and acetone/chloroform extracts; shown is one batch as representative of two others; aggregation inducer: PAF 10-6 M, ADP iO M; conc. of the extracts: 60zg/mI.
-r
100
80 60 40
J
20
0
WGC
GC
lilt PAP ADP
Downloaded by: University of Arizona Library. Copyrighted material.
In contrast to the results of the CO and LO tests, the chloroform extracts of both plants inhibit thrombocyte aggregation to a greater extent than the acetone / chloroform extracts (GA and WGA). At a concentration of 60 jtg/m1 reaction mixture, the chloroform extracts show an approximately 3-fold stronger inhibition of thrombocyte aggregation than the acetone/chloroform extracts. These results were obtained with PAF- (platelet activating factor) as well as with ADP-induced thrombocyte aggregation. Comparison of the effects of wild garlic and garlic extracts prepared during different seasons showed nearly reprodu-
"U
5
A. Sendl et. al
Planta Med. 58 (1992) Table 2 Inhibition of platelet aggregation: comparison of garlic and wild garlic extracts.
WGC GC WGA
GA
84.17 86.57 26.56 24.70
2.80 6.64 3.33' 1.24*
ADP 10 M Extract
Inhibition [%]
conc. 60 g/ml WGC GC WGA
GA
99.94 0.11* 99.89 0.19* 30.46 35.55
* Mean values of different batches.
Summarizing the comparative studies it becomes evident that the major sulfur-containing com-
Table 3 Range of contents of thiosulfinates (determined in chloroform extracts by HPLC using isolated compounds as external standards) and cysteine
sulfoxides (determined in MeOH/H20 extracts by HPLC using synthesized compounds as external standards) in different batches of garlic and wild garlic bulbs; data refer to dry weight; MATS: methylallyl- resp. allylmethyl thiosulfinate; MCSO: methyl-L-cysteine sulf oxide.
Compound
wild garlic
garlic
alum
0.65—1.10% 0.23—0.59% 0.55—0.81% 0.60—0.90%
0.82—2.50% 0.60—1.37% 0.14—0.23% 0.15—0.42%
allicin MATS MCSO
It is remarkable that the IC50 values of the single extracts are in general in the same order of magnitude as the isolated major compounds of these extracts. This leads us to assume that synergistic effects exist between the single compounds in the extracts.
ACE-inhibition
In the ACE-inhibition test, lyophilized aqueous extracts of A. ursinum and A. sativum cloves were only weakly active. In contrast, using an aqueous, lyophilized wild garlic leaf extract at a concentration of 0.3 mg/ml we obtained an inhibition value of 58 %, in comparison to 30% for a garlic leaf extract. Based on a meanwhile per-
formed fractionation and structure elucidation, y-glutamylcysteine derivatives are responsible for this inhibition (32). The effect remains unchanged if dried, frozen, or lyo-
pounds of the various extracts are responsible for the investigated pharmacological activities and that the individual
results correlate in every respect with the characteristic chemical composition of the single extracts. As far as the ACE inhibitory potential of the lyophilized water extracts is concerned, it could be substantiated that their activity derives from the presence of y-glutamylallylcysteine sulfoxide and other glutamylpeptides. The higher ACE-inhibiting ac-
tivity of wild garlic extract is consistant with the higher content of this type of compounds.
As most studies on the effects of garlic already have been carried out on humans, one can conclude that wild garlic bulbs and the preparations from them can serve as a substitute for garlic preparations, if they are applied in appropriately higher doses.
At the same time, this comparative study shows that some or all applied tests may be used to stand-
ardize extracts of wild garlic and garlic biologically or pharmacologically, if equal test conditions can be guaranteed. The inhibitory effect of wild garlic leaf extracts in the in vitro ACE test has relevance to the described effect of reducing blood pressure, if this effectiveness can also be confirmed in in vivo experiments.
philized leaves are used instead of fresh ones or if the References
aqueous extract is heated. Also, an inhibition of alliinase causes no reduction of the inhibitory effect. Summary and Discussion
2
Due to the nearly identical qualitative chemical composition of the corresponding garlic and wild garlic chloroform and acetone/chloroform extracts, the extracts of wild garlic were expected to show similar pharmacological profiles to that of garlic extracts in the in vitro tests used. The results presented have confirmed this assumption. The CO IC50 values of the chloroform extracts
6
were found to be in the same order of magnitude (0.88, 1.15 ig/ml) with superior activity of the garlic extract. The IC50 values of the corresponding acetone/chloroform extracts lay again in about the same inhibition range (0.42, 1.1 zg/m1), however this time with a distinctively higher inhibition rate of the garlic extract. In contrast the same chloroform extracts showed three times lower IC50 values (2.85, 2.95 jg/ml) in the 5-LO test than the acetone/chloroform extracts (0.5, 0.8 g/ml). This result is consistant with
the observation that ajoene and derivatives are better 5LO inhibitors than the thiosulfinates. Generally the thrombocyte aggregation inhibition test revealed no differences between the wild garlic and garlic extracts, However, the inhibiting potential of the chloroform extracts was three
Hoppe, H. A. (1975) Drogenkunde, 8. Auflage. De Gruyter-Verlag, Berlin. Heinz, U. J. (1984) Das Handbuch der modernen Pflanzenheilkunde, Hermann Bauer Verlag, Freiburg, p. 258. Wagner, H., Send!, A. (1990) Dtsch. Apoth. Ztg. 130, 1809. Wagner, H., Send!, A. (1991) P!anta Med. 57, 361. Harenberg, J., Giese, C., Zimmermann, R. (1988) Atherosclerosis 74, 247. Kleijnen, J., Knipschild, P., Terriet, G. (1989) Br. J. Clin. Pharmac. 28, 535. Ernst, E., Weihmayr, T., Matrai, A. (1985) Br. med. J., 291. Bordia, A., Sharma, T., Parmar, Y., Verma, S. (1982) Indian HeartJ.
34,86. Lau, B. H. S., Adetumbi, M.. Sanchez, A. (1983) Nutrition Research
3, 119. ° Wagner, H., Wierer, M., Fessler, B. (1987) Planta Med. 305. Wagner, H., Breu, W., Send!, A., Burgi, C., RGdi, P. (1990) Planta Med. 56, 665. 12
13 14 15 16
17
Wagner, H., Dorsch, W., Bayer, Th., Breu, W., Willer, F. (1990) Pros-
tag!andins, Leukotrienes and Essential Fatty Acids 39, 59. Belman, S., Solomon, J., Sega!, A., Block, E., Barany, G. (1989) J. Biochem. Toxicology 4, 151. Block, E., !yer, R., Grisoni, S., Saha, C., Belman, S., Lossing, F.
(1988) J. Am. Chem. Soc. 110, 7813. Wagner, H., Fessler, B. (1986) Planta Med. 374. Wagner, H., Breu, W. (1989) Dtsch. Apoth. Ztg. Supplement 15, 21. Mayeux, P. R., Agrawa!, K., Tou, J., King, B., Llppton, L., Hyman, A., Kdowitz, P., McNamara, D. (1988) Agents and Actions 25 1/2, 182.
Downloaded by: University of Arizona Library. Copyrighted material.
PAF 10-6 M Inhibition [%] conc. 60 yg/ml Extract
times greater than that of the acetone/chloroform extracts. This again indicates, that the thiosulfinates possess higher TA inhibitory activities than the ajoenes.
Comparative Pharmacological Investigations ofAlliam arsinam and Alliam sativam
19
J. Aim Chem. Soc. 108, 7045. Apitz-Castro, F., Ledezma, F., Escalante, J., Jorquera, A., Pinate, F.,
28 Srivastava, K. C. (1986) Prostaglandtns Leukotrienes and Medicine 29
Moreno-Rea, J., Carillo, G., Lea!, 0., Jam, M. K. (1988) Drug Re28 31 22
23 24 21 26 27
search 38(11), 901. Fazal Mohammed, S., Woodward, S. (1986) Thromb. Res. 44, 793. Apitz-Castro, F., Escalante, J., Vargas, R., Jam, M. (1986) Thromb. Res. 42, 303. Rendu, F., Dave!oose, D., Debouzy, J., Bourdeau, N., Levy-To!edano, S., Jam, M., Apitz-Castro, R. (1989) Biochem. Pharmaco!ogy 38,
31
8, 1321. Makheja, A., Bailey, J. (1990) Agents and Actions 29 3/4, 360. Apitz-Castro, F., Cabrera, S., Cruz, M., Ledezma, E., Jam, M. (1983)
Thrombosis Res. 32, 155. Okuyama, T., Fujita, K., Shibata, S., Hoson, M., Kawada, T., Masaki, M., Yamate, N. (1989) Planta Med. 55, 242. Liakopou!ou-Kyriakides, M. (1985) Phytochemistry 24,7 1593. Al!, M., Thomson, M., Alnaqeeb. M., Al-Hassan, J., Khater, S., Go-
mes, 5. (1990) Prostag!andins Leukotrienes and Essential Fatty Acids 41,95.
7
31 36
22, 313. All, M., Mohammed Suad, Y. (1986) Prostaglandins Leukotrienes and Medicine 25, 139. Gaflen, J., Tavapes, I., Bennett, A. (1984) J. Pharm. Pharmaco!. 36, 272. Ariga, T., Oshiba, S., Tamada, T. (1981) The Lancet 7, 150. Wagner, H., Fib!, G., Send!, A. (1991) poster presented on the I!. International gar!ic symposium, 7.— 10. 3. 1991, Ber!in. Wagner, H., Wierer, M., Bauer, R. (1986) P!anta Med. 184. Wagner, H., Red!, K., publication in preparation. Kuh!, P., Shi!oh, R., Murwski, U., Zi!!iken, F. (1984) Prostag!andins 28, 783. Wagner, H., Steinke, B., (1991) Munch. Med. Wschr. 133, Supp!. 1, 54—57.
38
Born, G. V. F., Cross, M. J. (1963)J. Phys!o!. 168, 178—195. Heuer, H., Boehringer Inge!heim KG, persona! communication. Wagner, H., Bib!, G. (1991) Pianta Med. 57, 137.
Downloaded by: University of Arizona Library. Copyrighted material.
18 Block, E., Ahmad, S., Catalfamo, J., Jam, M., Apitz-Castro, R. (1986)
Planta Med. 58(1992)
