Papers 301
Metabolism of Garlic Constituents in the Isolated Perfused Rat Liver* C. Egen-Schwind'2. R. Eckard'. andF. H. Kemper' * Presented in part at the II. International Garlic Symposium. Berlin. March 1991 Institut für Pharmakologie und Toxikologie der WestIálischen Wilhelms-Universitat Munster. DomagkstraBe 12. D(W)-4400 MUnster, Federal Republic of Cermany
Address for correspondence
Abstract The metabolic and kinetic behaviour of different garlic (Allium sativum L., Alliaceae) constituents were investigated in the isolated perfused rat liver, using aqueous extracts of garlic powder as well as isolated allicin, the main product of the enzymatic degradation of alliin. Allicin (ally! thiosulfinate) showed a
remarkabje first pass effect and passed the liver unmetabolized only at high concentrations which caused considerable cell injuries. Diallyl disulfide and allyl mercaptan were identified as metabolites of allicin, whereby diallyl disulfide probably is the metabolic precursor of ally! mercaptan as shown by perfusion with diallyl disulfide alone. The metabolites diallyl disulfide
and ally! mercaptan could be determined in the
perfusion medium as well as in the bile and the liver tissue. Other degradation products of garlic were also investigated in this model. Ajoenes and vinyldithiins were detected in perfusion medium after liver passage but no metabolites of them could be identified up to now.
constituents after ingestion of fresh garlic were obtained by GC-MS analysis of human breath where allyl mercaptan and diallyl disulfide have been identified (2, 3). Other metabolites of garlic, such as N-acetyl-S-(2-carboxypropyl)cysteine, N-acetyl-S-allylcysteine, and hexahydrohippuric acid, were identified in human urine after consumption of fresh garlic (4).
As the liver is the most important drug metabolizing organ, the isolated perfused rat liver is an acknowledged and convenient model to study hepatic metab-
olism and biotransformation without considering other kinetic parameters such as test substance absorption and distribution. In comparison to isolated hepatocytes, the functional integrity and compartmentation of the organ is maintained during liver perfusion. In the following study the metabolism of different garlic constituents was investigated using the experimental model of the isolated perfused rat liver.
Materials and Methods Chemicals Acetonitrile for HPLC was of Lichrosolv® quality
Key words
A ilium sativum, garlic, allicin, metabolism, liver perfusion, diallyl disulfide, ally! mercaptan.
and obtained from Merck, Darmstadt (FRG). All other chemicals were of analytical grade and purchased from Merck, l)armstadt (FRG), Fluka, Buchs (Switzerland), Roth, Karisruhe (FRG), or Aldrich, Steinheim (FRG).
Animals Male Wistar rats (strain: Bor WISW (SPF TNO); average body weight 350 g] were used for the experiments (animal protection permit: RP MS AZ 37/90).
Introduction
Several garlic (Allium sativum
L.,
Alliaceae) preparations, such as garlic powder, oily preparations, and steam distillates, are used in traditional medicine with the same indications although they contain different constituents. Up to now it is unknown which constituent(s) of garlic represent(s) the active principle(s) be-
cause no reliable data about metabolism and pharmacokinetics of these ingredients are available. Our previous studies concerning pharmacokinetics of garlic compounds succeeded only with vinyldithiins (2-vinyl-4H-1,3-dithiin and 3-vinyl-4H- 1 ,2-dithiin) (1). These transformation products of allicin could be detected in liver, serum, kidney, and
Perfusion medium A Krebs-Henseleit-bufTer (5) containing 11.1 mM
glucose, 3mM MOPS (3-(morpholino)propanesulfonic acid) and 0.5% bovine serum albumin (COHN-fraction V; bioproducts) was employed as perfusion medium (osmolarity: 294 mOsmol/kg). The perfusate was maintained at 37 C and pH 7.4.
Perfusion of the isolated rat liver
The rat was anesthetized with ketamine (10mg/bOg body wt). The operation procedure was carried out
fat tissue after oral administration, but no metabolites
according to Sugano et al. (6). A scheme of the perfusion system is shown in Fig. 1. The perfusate was pumped by a roller pump from
were proven. First indications on the metabolism of garlic
a reservoir after passing a glass fibre prefilter (GF 92, Fa.
Downloaded by: University of British Columbia. Copyrighted material.
Received: July 30. 1991
302 Planta Med. 58(1992)
C. Egen-Schwind eta!.
Furthermore cell injuries were assessed by histo-
Sartorius) into an oxygenator. A permanent flow of 2—31 95 %
the perfusate. Thereafter the perfusion medium passed a pres-
logical examination of stained (Hematoxylmn and Eosin) liver sections (thickness: 5 fsm).
sure/flow regulator before entering the liver through a portal vein catheter. An effective perfusion pressure of 20 cm perfusate (i.e.
the difference in height between pressure/flow regulator and portal vein catheter) was adjusted corresponding to the physiological blood pressure of the rat. The perfusate left the liver through the vena cava inferior. The flow rate through the liver was
determined continuously (average flow rate 2,5—3 mI/min/ g liver). One part of the venous effluent was used to determine P°2' pCO2. pH, and potassium concentration with a blood gas, pH. and
Preparation of test solutions Garlic solution: Chinese garlic powder (supplied by Lichtwer-Pharma, Berlin) containing 0.6% allicin (kindly de-
termined by Woelm Pharma. Eschwege) was suspended in perfusion medium, incubated at room temperature for 20mm, and centrifuged. The supernatant was filtered through a 1 pm filter (Sartorius) and used for perfusion experiments (i.e. garlic
electrolyte analyser (System 2000. Eschweiler & Co.). The
solution).
parameters were measured alternately in front of and behind the liver at intervals of three minutes. Thus the 02 consumption and the K° release of the liver cells could be determined. Another part of the venous effluent was collected in fractions of 8 ml and used
Allicin: Garlic powder was treated as described before. After centrifugation allicin was obtained by extraction of
the supernatant with methylene chloride. The solvent was re-
for measuring the activities of alanine aminotransferase and aspartate aminotransferase as well as the concentration of the metabolites and unmetabolized substances, respectively. The
moved under reduced pressure and the residue was dissolved in perfusion medium.
perfusion system was enclosed in a cabinet which was warmed up to 37°C. Perfusions were carried out as single pass experiments. The test solutions were infused into the flowing perfusion fluid with a constant rate of 1 mI/mm. After perfusion, a small piece of liver tissue was fixed in Bouin's fluid and the remaining part of the
Alum (kindly provided by Lichtwer-Pharma, Berlin) and diallyl disulfide: Both were dissolved directly in
liver was precooled in liquid nitrogen and kept frozen at —20°C until analysed.
Parameters offunction and viability of the perfused liver Gross liver appearance, oxygen consumption, bile formation, flow rate, activities of aspartate and alanine aminotransferase, and potassium release of the liver served as indicators of liver viability. The oxygen consumption of the liver was calculated from the difference between the influent and effluent
oxygen concentration multiplied by the perfusate flow rate
through the liver. The activities of alanine and aspartate aminotransferase were determined by test kits from Boehringer Mannheim.
perfusion medium.
Extraction procedure ofperfusate and bile Ally! mercaptan: After addition of 200 p1 of the internal reference solution ((100 pg CHCI3/ml xylene). a fraction of the perfusate (8 ml) or the bile was extracted with 500 p1 xylene and centrifuged. The supernatant was used for GC-MS analysis.
Dia!lyl disulfide: A fraction of the perfusate (8 ml) or the bile was extracted with 1 ml hexane after addition of
2Opl of the internal reference solution 11 (91.5pg dipropyl disulfide/mI hexane). The hexane extract was analysed by GC-MS. A!licin: 100 p1 of 20% (w/w) trichloroacetic acid were added to I ml perfusate in order to precipitate proteins. After centrifugation the clear supernatant was used for HPLC-analysis.
Fig. 1
perfusale reservoir
infusion pump 1 mt/mm
vena porlae
liver
Design of the isolated perfused rat liver.
Downloaded by: University of British Columbia. Copyrighted material.
02/5% C02/min through the oxygenator served for oxygenation of
Planta Med. 58(1992) 303
Metabolism of Garlic Constituents in the Isolated Perfused Rat Liver
aldehyde/tert. -butyithiol according to Ziegler and Sticher (7). Therefore the studies dealing with the metabolism of alum were carried out with a perfusion medium containing no bovine serum albumin because the protein would interfere with the determination procedure. 200j.d of derivatization reagent (7) were added to 800 il of perfusate. After an incubation time of 15 mm the alliin concentration was determined by HPLC.
GC-MS conditions System: Computer-controlled GC-MS Finnigan MAT 4515; INCUS data system; NOVA/4 with CDC/CMD-Di-ive.
Gas chromatograph: Column: fused silica capillary column (DB-5; 30 m x 0.15mm); carrier gas: helium: flow
rate I mI/mm; injection port temperature: 150°C; split: 1:20; transline temperature: 260CC.
Results Control experiments, i.e. perfusion with perfusion medium only, demonstrated that the liver was viable over a period of 140—180mm as shown by a constant 02-consumption (0.04—0.05 mi/min/g liver), no release of potassium, and no rise of the activities of alanine and aspartate aminotransferase.
Garlic powder
The activities of alanine and aspartate aminotransferase as viability parameters could not be determined during perfusions with solutions prepared from garlic powder because it contains high concentrations of
pyruvate and oxalacetate, the substrates of the aminotransferases. In addition, the potassium concentration of
temperature: 90°C isothermal; MID: m/e 111. 144 (M vinyl-
garlic powder hinders the monitoring of K-release into the perfusate as a consequence of a hepatocyte injury. Thus, the histological examination is required to check the cell integrity in this case. Alter infusion of a garlic solution prepared from 200mg garlic powder/mi for 50 mm (infusion rate 1 mI/mm), considerable cell injuries were observed by light microscopy. Allicin could be detected in the perfusate after liver passage. Fig. 2 shows the allicin concentration in
dithiins), m/e 146 (M° diallyl disulfide). m/e 150 (M dipropyl di-
the perfusate of two livers perfused with a solution pre-
sulfide).
pared from 200mg garlic powder/mi. After the garlic-
Mass Spectrometer: ion source: El-mode; emission: 0.27 mA; ionization energy: 70eV; temperature: 150°C;
detector: multiplier: 1200V; sensitivity: i0 AMPS/V. The special conditions for ally! mercaptan: oven temperature: 70°C isothermal; MID: m/e 74 (M allyl mercaptan), m/e 83 (chloroform) and for diallyl disulfide. vinyldithiins: oven
Internal reference: For quantitative allyl mercaptan determination CHCI3 in xylene (100g/ml, internal reference solution I) and for quantitative diallyl disulfide determination dipropyl disulfide in n-hexane (91.3tg/m1. internal reference solution II) was used as internal reference.
infusion was finished, the allicin concentration decreased rapidly. After infusion of a garlic solution prepared from 150mg garlic powder/mi, allicin could be determined in
the perfusion medium. The resulting allicin concentration-time plot was quite similar to that shown in Fig. 2, only
on a lower level. Alter administration of a dosage lower
External reference: For quantitative de-
than 100mg garlic powder/mi, allicin was completely meta-
termination of allyl mercaptan: 20kg CHCI3 and 25eg allyl mercaptan in 500zl xylene; for quantitative determination of
bolized during the first liver passage. In the perfusate
diallyl disulfide: 1.83 jg dipropyl disulfide and 2.2kg diallyl disulfide in I ml hexane. Injection volume: 1.5 id.
HPLC system for analysing allicin and
alliin
dialiyl disulilde and allyl mercaptan could be identified as metabolites after infusion of garlic solutions. In Fig. 3 the
allicin and diailyi disulfide concentrations are shown during infusion of a garlic solution in two intervals of 12mm each. Apparently diallyl disulfide was slowly formed from allicin. After the second garlic-infusion, the ailicin concentration decreased faster than after the first infusion.
Apparatus: Merck-Hitachi liquid chromatograph 655A-11 with proportioning valve 655A-71 and processor A 65561; detector: variable wavelength UV monitor 655A; column: RP 18 Lichrosorb, 7gm, 250 x 4mm.
Allicin: Detection wavelength: 210 nm; mobile phase: acetonitrile 40%, aqua dest. 60%; flow rate: 1 mI/mm; t5 (allicin): 5.60 mm.
Alum: Detection wavelength: 337 nm; mobile phase: gradient: acetonitrile/0.045 M sodium dihydrogen phos-
phate buffer (pH 7.15) 20%:80% to 30%:70% from 0—8min
(linear) and 30%:70% isocratic from 8—22min; flow rate: 1 mI/mm; tg(ailiin): 17.24 mm.
The quantitative determination of allicin and alliin was carried out by the external standard method. A defined garlic powder and alliin were used as standard preparations.
20
30
60mm
Inftjelon time
Fig. 2 Concentration of allicin in perfusate after liver passage of two perfused livers during infusion of a solution prepared from 200mg garlic powder/mi for 50 mm (infusion rate 1 mI/mm).
Downloaded by: University of British Columbia. Copyrighted material.
Alum: In order to determine aiim by HPLC with
UV-detection it had to be derivatized with o-phthaldi-
304 PIe nta Med. 58 (7992)
C Egen-Schwind et at
— — thee 12 5.0
180 120
0.3 0.6
40 0.4
n
0 61218 24
48
54
80
0
IC
20 30
40
50
60
70
80nn
10
20
40
50
60
70
80 nh
02
N—
0.0
1000
66IT
800
Fig. 3 Concentration of allicin and diallyl disulfide in perfusate after liver passage during infusion of a solution prepared from 100mg garlic powder/mi in 2 intervals of 12 mm (infusion rate 1 mI/mm).
800 400
30
— 150
pQhTio
— Allicin
Diallyl disullido
Fig. 5 Concentration of diallyl disulfide and allyl mercaptan in perfusion medium after liver passage during infusion of 150 pg diallyl disuifide/min for 30 mm.
Allyl
mes •- 1441 14411
10
30
20
40
_________________
60 mm
aIIpcLn Infusion
0
1100 gg/min
+
Fig. 4 Concentration of allicin, diallyl disulfide and allyl mercaptan in perfusion medium after liver passage during infusion of 1 100zg allicin/min for 24 mm.
5
10
15
20
25
30 mm
Balm in$cton 01 3.6 LQIS 1441 arc 1.4 gmil 14411
Fig. 6 Concentration of the vinyldithuns in perfusion medium after bolus iniection of 3.6 pg/mI 1441 = 3-vmnyl-4H-1,2dithin and 1.4pg/ml 14411 2-vinyl4H-1,3-dithUn (after liver passage).
Allicin As
garlic powder prevents the de-
termination of the viability parameters, isolated allicin was used for the following experiments. After infusion of low dosages of allicin (from 100 pg up to 400 pg/mm), no unchanged allicmn could be detected after liver passage —only diallyl disulfide and ally) mercaptan were found. After administration of 1 100 pg/mm, allicin was detected in the perfusion medium as well as its metabolites diallyl disulfide
and allyl mercaptan (Fig.4). At the beginning of the infusion, allicin was metabolized completely until the infused
allicin concentration exceeded the biotransformation capacity of the liver. Then a part of allicin left the liver un-
changed (of Fig. 4). After cessation of the allicin infusion, its
concentration decreased rapidly. The concentrations of the metabolites are shown in Fig. 4, too. Diallyl disulfide was
formed immediately after starting the allicin infusion, whereas the allyl mercaptan formation was observed later. Both metabolites could also be identified in the collected
bile as well as in the liver tissue. This allicin dosage of llOOpg/min caused cell injuries which were detected by an increase of the alanine and aspartate aminotransferase activities and confirmed by histological examination of liver sections.
Downloaded by: University of British Columbia. Copyrighted material.
200
Planta Med. 58(1992) 305
Metabolism of Garlic constituents in the Isolated Perfused Rat Liver
in order to clarify whether allyl mercaptan is the metabolite of diallyl disulfide, a perfusion with diallyl
disulfide was performed (Fig.5). After infusion of l50g/min, diallyl disulfide as well as allyl mercaptan were detected in the perfusion medium after liver passage. The high concentration of allyl mercaptan in comparison to diallyl disulfide in the perfusate can be explained by the
more hydrophilic character of the thiol allyl mercaptan whereas the lipophilic diallyl disulfide is retained in the fatty compartment of the liver tissue. Vinyldithiins To investigate the metabolism of vinyldithiins, which represent very lipophilic substances, a
The vinyldithiins, main ingredients of oily preparations of garlic, do not show a significant first pass effect. They pass the liver in considerable amounts, but they are not eliminated biliary. Metabolites of vinyldithiins could not be detected up to now.
The amino acid ailiin also shows no first pass effect. it could be detected in the perfusate after liver passage. To some extent it seems to be metabolized by liver enzymes, because diallyl disulfide was identified in the perfusion medium after liver passage. These findings are in accordance with the results of Pentz et al. (8) gained in liver slices. Up to now it is unclear whether alhin is first converted to allicin and then reduced to diallyl disulfide, or first to deoxyalliin and then decomposed to diallyl disulfide. Conclusions
bolus injection was performed. Fig. 6 shows the concentra-
tion of the vinyldithiins in perfusion medium after liver
passage. The kinetics of both vinyldithiins are rather similiar. The vinyldithiins could not be detected in the bile. No metabolites of the vinyldithiins were identified either in perfusate or in bile or liver. Alum
Alliin could be detected in the perfusate
Allicin shows a pronounced first pass effect in the isolated perfused rat liver. Aliicin could only be de-
tected in the perfusate when concentrations were used which already cause cell injuries. Therefore, allicin seems not to be the biologically active component of garlic. As metaboiites of ailicin, diallyl disulfide and ailyl mercaptan were
identified from which allyl mercaptan is the metabolite of diallyi disulfide.
after liver passage, but no allicin was found. Additionally, diallyl disulfide was identified in the perfusion medium.
These findings indicate that alliin is converted by liver enzymes.
Discussion
Acknowledgements We thank Miss A. Traupe, Miss A. VoB, and Mr.
D. Hübner for the excellent technical assistance with the liver perfusion. in addition the authors thank the consortium MCM Klosterfrau et al. for the material support of the present study.
References
The isolated perfused rat liver is a wellaccepted system for studying hepatic drug metabolism and
biliary excretion. The formation of metabolites in perfusate, liver, and bile can be determined.
The investigation of solutions of garlic powder in this model was aggravated by some difficulties
concerning the determination of viability parameters (activities of the aminotransferases and potassium release).
Only the histological examination of liver sections was available for checking the integrity of the hepatocytes, when solutions of garlic powder were used in this system. High dosages of garlic powder (solutions prepared from 200mg garlic powder/mi) or isolated allicin caused considerable cell injuries in the porta hepatis zone. After reduction of the dosage, cell injuries were not observed, but allicin could not be detected after liver passage. Consequently in this dose-range allicin is completely metabolized during the first passage when the liver is morphologicafly and functionally intact. This result was confirmed by studies with isolated allicin. In accordance with these results, a fast decomposition of allicin by the liver has
been observed during former studies with a liver
homogenate (1).
In the isolated perfused liver allicin is reduced to diallyl disulfide which, in turn, is reduced to allyl mercaptan. Both metabolites were detected in the exhaled
air of humans after ingestion of fresh garlic (2, 3). This study shows that these metabolites are formed in the liver.
Egen-Schwind. C.. Eckard, El., Jekat, F. W., Winterhoff. H. (1992) Planta Med. 58, 8—13. Laakso, I.. Seppanen-Laakso. T.. ililtunen, R., Muller, B., Jansen, U., Knobloch, K. (1989) Planta Med. 55, 257—261. Minami. 1., Boku, T.. Inada, K., Morita, M.. Okazaki, Y. (1989) J. Food Sri. 54, 763—765. Jandke. J.. Spiteller, G. (1987) J. Chromatogr. Biomed. AppI. 421, 1—8.
Krebs, U. A., Henseleit. K. (1932) Hoppe Seyler's Zeitung f. Physiol. Chemie 210. 33—66. Sugano, T.. Suda. K.. Shirnada. M., Oshino, N. (1978) J. Biochern. 83, 995—1007. Ziegler, S. J.. Sticher. 0. (1989) Planta Med. 55. 372—378.
Pentz, R., Guo, Z.. Siegers, C-P. (1991) Med. Welt 42, 46—47 (Sonderheft 7a).
Downloaded by: University of British Columbia. Copyrighted material.
Diallyl disulfide