Allicin, the active principle of garlic, was first isolated and characterized as CH2=CHCH2S(O)SCH2CH=CH2 (diallylthiosulfinate) in 1944 by Cavallito (1). Among the medicinal properties that this substance possesses is an antimicrobial action especially against Gram positive bacteria. It is generally admitted that this activity is based on the reaction of allicin with sulfhydryl groups of certain enzymes to form RSSallyl derivatives, making the enzymes inoperative.
Cavallito's original article included the first appearance of polarography as a technique to follow allicin's decomposition with time, but the studies were not continued perhaps because allicin's chemistry was poorly known at the time and the results were difficult to interpret. Stable sulfur compounds, however, such as cystine and cysteine, were studied using mercury and platinum electrodes as early as 1940 (2), and since then there have been many papers dealing with the polarographic analysis of thiols and disulfides (3,4).
Materials. The equipment used for the cyclic voltammetry was an Autolab potentiostat( Faculty of Pharmacy of Sevilla and Faculty of Chemistry of Badajoz) (Eco Chemie, Holland) and a Metrohm 663 (Herisau,Switzerland) stand with mercury electrode, and for the differential pulse voltammetry was a Polarecord (Metrohm) 626 with a 663 stand.
Synthesis. Allicin was synthesized by oxidation of diallyldisulfide with magnesiummonoperoxyphthalate and tetrabutylammoniumhydrogensulfate as phase-transfer reagent. After TLC separation, the extract was passed through a reversed phase cartridge and eluted with fractions of 30% methanol/water (5) for purification. Allicin standards were quantified colorimetrically with 5,5'dithiobis (2nitrobenzoic acid) (6) and stored at -70oC. Diallyl disulfide (Fluka) was purified by solid-phase extraction (C18) with methanol/water 50% in a similar fashion to the allicin. The purity of the substances was determined by HPLC (Gilson) (acetonitrile/methanol/water, 50/10/40).
Allicin-containing garlic extract. A sliced garlic clove (6 g) and 50 mL of water were homogenized for 5 min in a vortex mixer. The mixture was left for 30 min at room temperature and then centrifuged at 27000 g for 15 min. Ten milliliters of the supernatant were diluted to 250 mL with water and stored until assay as allicin sample
Figure 1 shows a cyclic voltammogram of allicin in water and perchloric acid as support electrolyte. There is a single reduction wave with a 20 mV halfwave potential (vs SCE) and a smaller anodic wave
FIG. 1. Cyclic voltammetry of allicin in water and 0.06 M perchloric acid. Arrows indicate the direction of the potential sweep. Scan rate: 100 mV/s.
For comparison, cyclic voltammograms were made for diallyl disulfide and allicin (Figure 2). The curves are almost identical in the cathodic zone between 0.14 and 0.1 V; but, at difference of allicin, in diallyl disulfide , the anodic and cathodic waves are of similar magnitude,suggesting a reversible process.
FIG. 2. Cyclic voltammograms of diallyl disulfide and allicin. Scan rate: 80 mV/s.
Influence of pH. The allicin reduction was studied in Britton-Robinson buffers in a pH range from 2.36 to 5.76 at 12oC. Figure 3 shows a plot of the halfwave potential vs pH. The slope of the least squares straight line fit is 0.055 V/decade, proving the presence of hydrogen ions in an initial or following steps of the reduction process.
FIG. 3. Variation of the halfwave potential with pH.
When a glassy carbon electrode was used, no reduction waves were observed, which suggests a specific adsorption of allicin onto the mercury electrode similar to that of thiols and disulfides.
FIG. 4. Cyclic voltammograms of allicin with glassy carbon electrode
This adsorption was not fully independent of the support electrolyte concentration, as is shown by the decrease in the response as the perchloric acid concentration was increased.
The similarity of the voltammetric curves for allicin and diallyl disulfide point to allylmercaptan, CH2=CHCH2SH, as being a common reduction product. But since allicin contains a sulfoxide group, the initial products are probably different, and the reduction process would also differ at some stage.
Calibration. The standards were added to a blank solution (water + perchloric acid) following the same procedure as in the assays (see below). The linearity was good in the range 10-6 M to 1.7x10-5 M (r = 0.9997). The detection limit, estimated from confidence bands (7), was 1.3x10-7 M. The standard addition method was followed for the garlic allicin assay as detailed in the next section.
FIG. 4. Calibration curve of allicin (HMDE). The three points upside are off the linear range
Garlic allicin assay. An aliquot of the garlic extract was stored at -4oC in a household freezer and analysed periodically to study the allicin's conservation at this temperature. Nitrogen was bubbled through 20 mL of deionized water and 100 µL of concentrated perchloric acid in a polarographic vessel for 10 min, after which the sample was added and stirred for 15 s for complete dissolution. Then a mercury drop was extruded ( 0.11 mg with a spherical surface of 0.2 mm2 and the solution was stirred for 1 min (1000 rpm) at 100 mV initial potential. After a 0.4 min rest period, a scan was started to -200 mV at -10 mV/s with a 50 mV pulse and the DP polarogram was recorded. The same procedure was followed for each standard. A spectrophotometric determination was made immediately after the voltammetry. Table 1 lists the results of two determinations in a four-day period. The two methods agree reasonably well, and the precision2 is good (see Figure 5). The loss of allicin in these suboptimal conditions is in agreement with published studies of its decomposition in aqueous garlic extracts (8).
FIG. 5. Standard addition in sample 2; r is vs/v0 , where vs is the standard volume added and v0 is the initial volume. The intercept is 1.06 mg/L
|µL of sample||Additions||[allicin] µM*||r||mg/L±ts**||colorimetric|
Interferences. The similarity in the cathodic process of allicin and of diallyl disulfide makes the latter a potential source of interference in the voltammetry. In aqueous extracts of garlic, however, this is not a problem as long as the sample is properly stored. In aged samples or extracts in organic solvents, allicin can form condensation products (9) with as yet poorly known voltammetric responses. Such cases require a prior separation step (e.g., by solid phase extraction).