Pharmacological Research, Vol. 22, No. 6, 1990 709

PHYTOCHEMICAL CHARACTERIZATION AND RADICAL SCAVENGER ACTIVITY OF FLAVONOIDS FROM

HELICHRYSUM ITALICUM G. DON (COMPOSITAE)

R. MAFFEI FACINO*, M. CARINI*, L. FRANZOI~, O. PIROLAt and E. BOSISIOt

*Chemical Pharmaceutical Toxicological Institute, t lnstitute of Pharmacological Sciences, Faculty of Pharmacy, University of Milan and ~ Chemical Research and

Development, Farmitalia-Carlo Erba, Milan, Italy

Received in final form 19 March 1990

SUMMARY The glycosidic fraction of the flavonoids extracted from the flowering tops of the Helichrysum italicum G. Don was isolated, purified and characterized. This fraction was constituted by three compounds, which were assigned the struc- ture of 4,2',4',6'-tetrahydroxychalcone-2'-glucoside, kaempferol-3-glucoside and naringenin-glycoside.

Radical scavenger properties of the single glycosyl-flavonoids and of the in toto glycosidic fraction were tested with in vitro systems where different reactive oxygen species are generated (superoxide ions, hydroxyl radicals) and on lipid peroxidation induced by ADP/Fe 2+ and NADPH or CC14 in rat liver microsomes. The formation of reactive oxygen species was detected by cytochrome c reduction, salicylic acid hydroxylafion and hyaluronic acid depolymerization.

The action of the glycosidic fraction on the release of TXB 2 and 12-HETE in human platelets, after collagen stimulation, was also evaluated.

The glycosidic fraction inhibited in a dose dependent fashion lipid peroxidation in rat liver microsomes treated with ADP/Fe 2+ or CC14. This effect is due to the ability of flavonoids to scavenge free radicals at different stages of the process (superoxide ions, hydroxyl and lipid peroxide radicals). The single glycosyl- flavonoids exhibited a different scavenger activity, depending on the oxygen species and the chemical structure of the compounds. No effect of the fraction was observed on TXB 2 and 12-HETE formation at 100//~M concentration.

KEY WORDS: flavonoids, radical scavenger activity, Helichrysum extract.

INTRODUCTION

Helichrysum italicum G. Don (Compositae) is a shrub widespread in the arid sandy and stony areas of the Mediterranean regions. The decoction of the flowering tops of this species is used in folk medicine as an anti-allergic and anti-inflammatory

1043-6618/90/060709-13/S03.00/0 © 1990 The Italian Pharmacological Society

710 Pharmacological Research, Vol. 22, No. 6, 1990

agent for topical applications. In cosmetics, the drug finds a wide application to treat and prevent skin sunburn and erythema.

Previous phytochemical investigations showed the presence of several components in the crude extract of Helichrysum flowers including essential oils [1, 2] and flavonoids [3]. The flavonoid fraction consisted of three monoglycosides (80% of the total fraction) and of four aglycons. The aglycons were identified as gnafalin, apigenin, naringenin and luteolin [3].

The glycosidic fraction was a mixture of three main components, tentatively identified as glycosides of the aglycons naringenin, 4,2',4',6'-tetrahydroxychalcone [3] and kaempferol [4].

Preliminary pharmacological studies on the activity of the crude extract of Helichrysum itaficum demonstrated an anti-erythematous effect in guinea-pig and man after UVB radiation [3], mainly due to the flavonoidic fraction of the extract.

The aim of the present study was: (1) to further characterize the flavonoid glycoside fraction isolated from the crude extract of the Helichrysum flowering tops, as regards the nature and the position of the sugar moiety; (2) to assess the radical scavenger activity of the flavonoid glycoside fraction and of the single glycosides as mechanism of the anti-erythematous effect.

MATERIALS AND METHODS

Solvents (Merck-Bracco, Milan, Italy) were of analytical grade; kaempferol and naringenin were purchased from Roth (Prodotti Gianni, Milan, Italy); xanthine, ipoxanthine, thiobarbituric acid (TBA) and bovine serum albumin (BSA) from Sigma Chemical Co. (Prodotti Gianni, Milan, Italy); xanthine-oxidase (20 U/ml), catalase (65 000 U/mg), cytochrome c, ADP-Na salt, NADPH and hyaluronic acid were from Boehringer-Mannheim Italia (Milan, Italy). Collagen was purchased from Semmelweiss (Mascia BruneUi, Milan, Italy).

The ethanol extract (70% ethanol in water) of the flowering tops of Helichrysum italicum G. Don was a generous gift from Dr Mariani, Curt Georgy Imes SpA, Milan, Italy.

Apparatus Spectrophotometric measurements were performed with a Beckmann DB-GT

spectrophotometer; electron impact (EI) and field desorption (FD) mass spectra were obtained as previously described [3].

1H and 13C-NMR analyses were performed in DMSO-d 6 solution at probe temperature of 22°C on a VXR-200 Varian spectrometer (200 MHz for proton observation). Chemical shifts (6) are expressed in p.p.m, from 0 TMS.

Isolation of the flavonoid glycoside fraction The crude extract from Helichrysum italicum was obtained as described [3]. The

crude extract (20 g) was suspended with distilled water (10 vol), filtered and dried to obtain a pale yellow cristalline powder (16 g, flavonoid glycosides).

The glycosidic fraction was further fractionated into three components by preparative TLC (solvent system ethyl acetate:formic acid:acetic acid:water,

Pharmacological Research, Vol. 22, No. 6, 1990 711

100:11:11:27). Bands were visualized at 254 and 366nm or by UV light and fuming ammonia. Bands were recovered and further chromatographed on silica gel columns (elution with chloroform:methanol mixtures, with increasing concentration of methanol). Quantities of each component were determined by weighing. Chemical characterization was achieved by UV-visible spectrophotometry, mass spectrometry and NMR spectroscopy.

The in toto glycosidic fraction and the isolated compounds 1, 2 and 3 were tested for their antiradical properties as described below.

Assessment of superoxide ion scavenger activity The inhibitory effect of flavonoids on superoxide ion production was determined

by monitoring the reduction of cytochrome c by superoxide ions generated by the xanthine-xanthine oxidase system [5]. 0.2 ml of 100 mM K-phosphate buffer, pH 7.4 were added with 0.3 ml of cytoehrome c (0.25 raM), 0.25 ml of 26 000 U/ml catalase, 0.05 ml of BSA (6 mg/ml), 0.2 ml of xanthine (25 mM), 0.05 ml of xanthine oxidase (0.4 U/ml) and water to a final volume of 3 ml. Incubations were carried out at 25°C. The reduction rate of cytochrome c, expressed as nmol cytochrome c reduced/min was determined by measuring the increase in absorbance at 550nm, using an extinction coefficient of cytochrome c of 19.6 mM- 1/cm. The reduction rate of cytochrome cwas linear up to 5 min.

The glycosidic fraction and single constituents were dissolved in water and the effect was expressed as % inhibition of the reduction rate of cytochrome c.

Assessment of hydroxyl radical scavenger activity Formation of hydroxyl radicals by the ipoxanthine/xanthine oxidase/Fe 2 + system

was measured by the salicylate hydroxylation method [6]. This method measures the ability of hydroxyl radicals to attack salicylic acid to produce 2,3-hydroxy- benzoate, which is then measured colorimetrically. The reaction mixture contained 0.2 ml of 2.5 mM Na-salicylate, 0.2 ml of 30 mM EDTA, 0.2 ml of 10 mM FeSO4, 0.2 ml of 20 mM ipoxanthine and 0.1 M K-phosphate buffer to a final volume of 2 ml. Reaction was started by the addition of 0.04 ml of xanthine oxidase (4 U/ml). After 90 rain at 25°C, 0.06 ml of 11.6 M HC1 was added and the samples extracted with diethyl ether. The extracts were evaporated to dryness and to the residue were added 0.25 ml distilled water, 0.125 ml TCA (10% in 0.5 M HCI), 0.25 ml Na- tungstate (10% in water) and 0.25 ml NaNO 2 (0.5% in water). Absorbances at 510 nm were read immediately after the addition of 0.5 mi of 0.5 M KOH. The formation of hydroxyl radicals was expressed as nmol of 2,3-dihydroxybenzoic acid formed in 90 min, using a standard curve prepared with 2,3-dihydroxybenzoic acid. The effect of the total glycosidic fraction and the single flavonoids was expressed as % inhibition of 2,3-dihydroxybenzoic acid formation.

Hyaluronic acid degradation assay The degradation of hyaluronic acid induced by reactive oxygen species formed

in the ipoxanthine-xanthine oxidase system (superoxide ions, hydroxyl radicals, singlet oxygen and hydrogen peroxide) was measured by the decrease of viscosity of a reaction mixture containing 0.6 ml hyaluronic acid (1.67 mg/ml in 50 mM

712 Pharmacological Research, Vol. 22, No. 6, 1990

phosphate buffer, pH 7.4), 0.05 ml EDTA (3 n ~ in distilled water), 0.05 ml FeSO 4 (3 mM in distilled water), 0.1 ml ipoxanthine (20 mM in phosphate buffer, pH 7.4) and 0.1 ml distilled water. Assays were initiated by adding 0.1 ml xanthine oxidase (0.5 U/mi in water). The reaction mixture was incubated at 37°C for 60 min under shaking. The in toto fraction and single flavonoids dissolved in water were added to the incubation mixture before the initiation of the reaction with xanthine oxidase. Specific viscosity ~/sp was determined by the method of Betts and Cleland [7]. The inhibitory activity, expressed as % inhibition of hyaluronic acid degradation, was calculated by the following equation: I%=(~?sp-~L; rain)X (~]sp max--~]sp rain)X 100, where ?]sp is specific viscosity of the experimental solutions (containing the enzymes); ~]sp rain is specifiC viscosity of the control solutions (without the enzymes) measured after exposure to oxygen radicals; and t/s p max is specific viscosity of hyaluronic acid solutions not exposed to oxygen radicals.

Lipid peroxidation in rat liver microsomes Microsomal lipid peroxidation was induced by ADP/Fe 2 + complex o r CC14 in

the presence of NADPH. MDA formation was quantified by the thiobarbituric acid (TBA) method [8].

Male Sprague-Dawley rats (140 _+ 20 g body weight, from Morini, Italy) were used. Liver microsomes were prepared in 0.15 M KC1/0.1 M phosphate buffer, pH 7.4 as previously described [9].

A microsomal suspension (0.6 ml) was preincubated for 5 min at 37°C and the reaction was started by addition of 0.2 ml of initiator solution (0.1 mM FeSO4, 2.5 rnM ADP, 5 mM NADPH in distilled water). The mixture (final volume 1 ml) was incubated for 10 rain at 37°C [9]. The amount of MDA-TBA complex formed was calculated using an absorption coefficient of 1.56X10-SMXCm -1 and expressed as nmol/mg protein/10 rain.

The glycosidic fraction and the single compounds were added to the incubation mixture dissolved in distilled water, before the starting of the preincubation, in concentrations ranging from 0.025 to 2 raM.

The antilipoperoxidative effect was expressed as % inhibition of MDA formation, setting the values obtained without inhibitor as 100% activation. Mean values of five determinations were used for calculation.

The effect of the flavonoids was also tested on the kinetics of lipid peroxidation, measuring the production of MDA at various times from 1 to 120 rain after the addition of the initiator solution. For these experiments the concentration of flavonoids was as follows: glycosyl fraction 0.1 rnM, compound 1 0.25 n-~, compound 2 0.5 rnM and compound 3 0.75 mM.

When CCI 4 was used as radical generating system, incubation conditions were as described before except that 1/~1 of CCI 4 was added to 1 ml of incubation mixture, instead of ADP/Fe > . The glycosidic fraction was tested at concentrations from 0.01 to 0.5 mM.

TXB 2 and 12-HETE formation in human platelets Platelet rich plasma (PRP) was prepared from blood of healthy volunteers and

was incubated in the conditions described by Eynard et al. [10]. After 10 rain at

Pharmacological Research, VoL 22, No. 6, 1990 713

room temperature in the presence of the compounds to be tested, PRP was preincubated for 2 rain at 37°C under stirring in an aggregometer and incubated for 7.5 rain after stimulation with 10 #g of collagen. The reaction was stopped by addition of 1 M citric acid to pH 3.0 and samples were extracted with ethyl acetate. T X B 2 and 12-HETE in the lipid extracts were evaluated by multiple selected ion monitoring according to Eynard et al. [10]. The glycosidic fraction, dissolved in saline, was assayed at concentrations of 1-100 #M. Indomethacin 50 ffM was used in these assays as reference compound.

°H HO 4' 5' , OH i 5

, I 6

HO / ~ ' ~ i0 0 H O - - ~ , , / "~"

OH

,G~ 8 \ \ I r l ( B ) l /

"--. &, , ' l c ° '& _o ) HO----~#-,, / " I "

OH

Compound I Compound 2

Fig. 1. Chemical structure of compounds 1 and 2. Compound 1: 4,2',4',6'- tetrahydroxychalcone-2'-glucoside. Compound 2: kaempferol-3-glucos[de.

RESULTS

Characterization of the flavonoid fraction TLC analysis of the glycosidic fraction from the flowering tops of Helichrysum

italicum showed three components with Rf=0.59 (compound 1), Rf=0.50 (compound 2) and Rf=0.42 (compound 3). The compounds, isolated by preparative TLC plates, were identified by FD-MS and NMR analyses.

Compound 1 (Fig. 1). FD-MS spectrum showed three main ions at m/z 434 (M+), m/z 435 (M+H) + base peak and m/z 272 (M-162), indicative of a monosaccharide structure. Results from NMR analysis are shown in Table L The 1H-NMR spectral data evidenced the structure of a chalcone derivative, due to the characteristic pattern of the trans hydrogens H-a and H-fl (Ja_# = 15.6 Hz) [11]. The aromatic ring attached to the/3 carbon appeared to be p-substituted with a free hydroxyl group. The remaining aromatic signals at 5.91 d (H-3 ' ) and 6.17 6 ( H - 5 ' ) belonged to the chalcone nucleus and exhibited a small meta-coupling constant (J=l .8Hz) : their different chemical shifts strongly suggested the unsymmetrical glycosylation as shown in compound 1. 13C-NMR data (Table I), supported by the two-dimensional heteronuclear correlation experiment, agreed well with the above finding and indicated the glycosyl moiety as D-glucose: the configuration at the anomeric carbon C-1" had to be/3, since the corresponding chemical shift was 100.3 p.p.m. [12].

The chemical shifts of the p-hydroxy-substituted benzene ring were as expected and coincident with the values reported for isoliquiritigenin [13] and other analogues [14]. The site of glycosylation on the chalcone nucleus, as suggested by

714

i

e n

v

e,3

I I I I I I I i i 1 ~ 1 ~ I

A

t ~

Pharmacological Research, Vol. 22, No. 6, 1990 715

the 1H-NMR data, was confirmed by the ~3C chemical shifts, since only one aromatic protonated carbon was shifted downfields (97.0 6), compared to the model compound 2,4,6-trihydroxyacetophenone (C-3, C-5:95.0 6; C-l: 104.5 6) [15]. In compound 1, C-1' was found at the expected position (104.9 6) (conjugation of the carbonyl carbon with the double bond would shift C-I' to higher fields, while glycosylation at C-2' would act in the opposite direction). On the basis of FD-MS and NMR data, the structure of 4,2',4',6'- tetrahydroxychalcone-2'-glucoside was assigned to compound 1.

Compound 2 (Fig. 1). FD-MS spectrum showed three main ions at m/z 448 (M+), m/z 449 (M+H) ÷ base peak and m/z 286 (M-162), indicative of a monoglycosyl derivative. Results of NMR analyses are shown in Table I. On the basis of the 1H-and 13C-NMR data, compound 2 was identified as kaempferol-3- glucoside (3,4',5,7-tetrahydroxyflavone-3-glucoside) as already reported in the literature [16, 17].

Compound 3. FD-MS spectrum showed three main ions at m/z 434 (M+), m/z 435 (M+H) + base peak and m/z 272 (M-162), typical of a monosaccharide structure. The aglycon structure was established by EI mass spectrometry: the fragment ions at m/z 272, 255, 179, 166, 153 and 120 were identical with those of naringenin (4',5,7-trihydroxyflavanone). 1H- and 13C-NMR studies are still in progress to establish the position of the glycosyl residue (identified as glucose by TLC analysis after acid hydrolysis) on the aglycon moiety of compound 3, although preliminary spectrophotometric analysis, carried out with suitable reagents (NaOH, Na-acetate, AICI3) suggested that the sugar residue is located on position 4'[18].

HPLC analyses of the flavonoidic mixture indicated that the compounds 1, 2 and 3, are present in the fraction in tffe ratio 1:1:1, approximately (data Submitted for publication). On the basis of these results, mean molecular weight for the glycosidic fraction was assumed to be 438.

Scavenger activity of flavonoid glycosides Table II illustrates the results on the radical scavenger activity of the in toto

glycosidic fraction and of its components in different experimental models. The glycosidic fraction and each single flavonoid exhibited a dose dependent

anti-lipoperoxidant activity in liver microsomes exposed to ADP/Fe 2+ complex and NADPH. For comparison, catechin, a well known antioxidant, was tested (IC50 = 0.01 mM in our conditions). The inhibitory effect of the glycosidic fraction (IC50 = 0.06 mM) was slightly lower than that of catechin. The inhibition of the in toto fraction was greater than that of the single constituents at all the tested concentrations. The reduction of MDA formation was almost complete (86.3%) at 0.1 m~ concentration. The efficiency of flavonoids to reduce lipid peroxidation was in the order glycosyl-kaempferol/> glycosyl-chalcone ,> glycosyl-naringenin.

When the kinetics of lipid peroxidation were studied in the presence of the flavonoids (Fig. 2), the glycoside fraction delayed the initiation of the peroxidative process for 30 min. Chalcone-glycoside behaved similarly whereas the effect was less pronounced in the presence of kaempferol- and naringenin-glycosides since MDA production increased up to 30% in 30 min incubation.

In microsomes intoxicated with CC14, lipid peroxidation was reduced by the glycosidic fraction to a lesser extent than in the case of ADP/Fe 2+. From the

716

~+

===

~'~ e~

.==. ==

,u 0

M-)

cq

-H -I-I +I +I +I +I +I +I

~+I +I+I+I +I+I+I +I+I

~q +I

+I +I -I-I

+I +I +I

+~ ~ +~

+I +I

+I +I -H

+I +I +I

~121 °

t",l

• ~ - ~

~~ e-~

~ ~ ~ . , . ~

+I -H +I

t ~ t ' -I t-xl

+I +I +I

~N +I +i

t ~

Z ,d

t'¢3

_~.~@ , ~ ~ .~

~1 en

+i +i +I

~-i-i +i

+i 4-I

m.

0

0

o~

Pharmacological Research, Vol. 22, No. 6, 1990 717

,oo[ .t t. -I : . ¢ . . . . . . . . . . . . . . . . . . .

! 25 / :':" ,t

I I I 50 60 90

Incubation time (min)

120

Fig. 2. Effect of Helichrysum flavonoids on lipid peroxidation kinetics in liver microsomes treated with ADP/Fe 2÷ and NADPH (mean + s~ of 5 determinations).

Controls ; in toto glycosyl fraction 0.1 n~ . . . . . . ; chalcone-2'-glucoside 0.25 mM • ; naringenin-glycoside 0.75 mM . . . . . ; kaempferol-3-glucoside 0.5 rnu . . . . .

dose-effect curve (data not shown) half maximal effect of the glycosidic complex was reached at 0.27 mM.

Since in aerobic conditions lipid peroxidation in microsomes treated with ADP/ Fe 2÷ and NADPH is mediated by several oxygen species including chiefly superoxide ions and hydroxyl radicals [19], flavonoids from Helichrysum were assayed in vitro for their direct antiradical properties against these species.

The glycosidic fraction inhibited superoxide ion production starting from 0.05 mM concentration in a fashion similar to that of glycosyl-naringenin. The other two compounds, kaempferol and chalcone glycosides, showed an inhibition lower than that of the former ones (Table II).

The inhibitory effect of flavonoids on reactive oxygen species formed by the ipoxanthine-xanthine oxidase system were evaluated using two types of detecting methods: salicylic acid hydroxylation, which is mediated mainly by hydroxyl radicals, and hyaluronic acid degradation induced by all the reactive oxygen forms.

The inhibitory effect of flavonoids on hydroxyl radicals generated by superoxide- assisted Fenton reaction, was evaluated using salicylic acid as detector; all compounds inhibited salicylic acid hydroxylation starting from 0.25 mM. In the lower range of active concentration (0.25-0.75 mM), the flavonoid fraction was stronger than the single constituents and glycosyl-chalcone showed a greater effect than the other two compounds (Table II). These results indicate that flavonoids from Helichrysum italicum can intercept superoxide ions more effectively than hydroxyl radicals (see Table II) a fact which might suggest two different mechanisms: these flavonoids might (1) interact with superoxide ions via a one- electron transfer mechanism and (2) scavenge hydroxyl radicals by chelating iron ions, which are essential for hydroxyl radical production in the superoxide driven Fenton reaction [20].

718 Pharmacological Research, Vol. 22, No. 6, 1990

When oxygen radical scavenging activity was studied using hyaluronic acid depolymerization, which is promoted both by superoxide ions and hydroxyl radicals, the inhibitory effect becomes evident at very low concentrations (0.01 raM) for all the tested compounds, except glycosyl-naringenin (0.1 mM, Fig. 3). This indicates a synergistic scavenging effect of flavonoids from Helichrysum italicum against 02 centred radicals.

.¢:

I00 - ( o ) IO0

50

0.01 0.1 1

50

-(b)

0.01 0 . I I

I O0 c ) I00

0.1 I

50 50

-(d)

I I I r l l l l l I I I I I

0.1 I log conc. (raM)

Fig. 3. Effect of Helichrysum flavonoids on hyaluronic acid depolymerization induced by radicals: dose-response curves: (a) in toto glycosyl fraction; (b) chalcone-2'-glucoside; (c) naringenin-glycoside; (d) kaempferol-3-glucoside.

Effect of the flavonoid-glycoside fraction on TXB 2 and 12-HETE Since flavonoids are claimed to affect arachidonic acid cascade by inhibition

and/or stimulation of cyclooxygenase and lipoxygenase enzymes [21], the effect of Helichrysum flavonoids on these enzymes was investigated. The formation of TXB 2 and 12-HETE was respectively 209+15.9 and 87.1+7.91ng/ml of PRP stimulated with collagen (mean + so of three determinations). The addition of the glycosidic fraction did not influence the synthesis of the arachidonic acid metabolites, at 100 #M concentration. The levels of TXB 2 and 12-HETE found in platelets with added flavonoids were 242 _+ 8.06 and 85.3 + 7.63 ng/ml of PRP, respectively (mean +so of three determinations). Indomethacin (50 pM) caused a complete inhibition of TXB 2 formation and 78% reduction of 12-HETE synthesis.

Pharmacological Research, Vol. 22, No. 6, 1990 719

DISCUSSION

This study gives the definitive phytochemical characterization for two of the constituents of the glycosidic fraction of the flavonoids present in the crude extract of tlelichrysum italicum G. Don. Compounds 1 and 2 were identified as 4,2',4',6'- tetrahydroxychalcone-2'-glucoside and kaempferol-3-glucoside, respectively. For compound 3, naringenin-glycoside, mass spectrometric, chromatographic (after acid hydrolysis) and spectroscopic data indicate glucose as the sugar moiety, probably at the 4' position.

The presence of these flavonoids is quite common in the Helichrysum genus, in particular H. graveolens and arenarium [18], but to our knowledge this is the first report of the unequivocal identification of the flavonoidic components in Helichrysum italicum.

As regards the anti-radical properties of these flavonoids, it appears that the in toto fraction inhibited superoxide ions and hydroxyl radicals to a lesser extent than the lipid peroxidation in microsomes (see Table II). The explanation might be that the fraction is able not only to scavenge oxygen species (initiating stage of peroxidation), but also lipid radicals (R.), alkoxy (RO') and peroxy (ROO') radicals involved in the propagation and the maintenance of the radical chain reaction [22]. The ability of the glycosidic fraction to react with radicals other than superoxide and hydroxyl ones; is confirmed by the fact that CCl4-induced lipid peroxidation (promoted by "CC13 and CC13OO" radicals) is also inhibited, although to a lower degree comparatively with the ferrous-ion-induced lipid peroxidation, A metal chelating property of these flavonoids might be evoked as an additional mechanism to explain the difference in the two systems, since it is known that microsomal NADPH-dependent lipid peroxidation is catalysed by NADPH- cytochrome P-450 reductase and proceeds in the presence of iron ions [20]. CC14 activation instead takes place via cytochrome P-450 and does not require iron ions [20],

By comparison of the inhibitory capacity of the single flavonoid-glycosides with that of the in toto fraction (see Table II), it is evident that single flavonoids act with an additive mechanism, thus giving rise to a synergistic effect in diminishing lipid peroxidation. A combined action was observed also in hyaluronic acid depolymerization experiments. This might be due to the different reactivity of each compound towards C-centred and O-centred radical species, as a consequence of their different chemical structure.

As regards the effect on the arachidonic acid cascade, it is difficult to compare the results reported in this study with those available from the literature. In fact, 5-, 12- and 15-1ipoxygenases are differently affected by flavonoids, and glycosidation was found to decrease the activity of the compounds [21]. On the other hand, cyclooxygenase was shown to be stimulated or inhibited, in relation to the concentration of arachidonic acid used as substrate in the assay [23]. Moreover, as for lipoxygenase, aglycons are more active than the corresponding glycosides.

Despite the lack of effect found in platelets, Helichrysum flavonoid fraction possesses anti-erythematous (UVB) activity in vivo. Hence the last may be due to the scavenger activity of flavonoids, which by inhibiting skin PUFA peroxidation, greatly reduces the development of UVB erythema [24]. Alternatively, with a

720 Pharmacological Research, Vol. 22, No. 6, 1990

mechanism still implying a free radical scavenging process, flavonoids can act as stabilizers of skin mastocyte membranes, thus preventing histamine release [25]. A membrane stabilizing effect by hydrophilic flavonoids has recently been reported [26]: this effect for Helichrysum flavonoids is well documented by the delay of the onset of lipid peroxidation observed in the kinetic experiments (see Fig. 2).

During senescing and pathological degeneration processes of the skin, a dramatic loss in the content and integrity of hyaluronic acid is observed, as a consequence of the depolymerization induced by radicals [27].

The results reported in this study show that Helichrysum flavonoids are highly effective in preventing oxygen-induced hyaluronic acid depolymerization and therefore are proposed as protecting agents in skin degeneration processes.

REFERENCES

1. Tira S, Di Modica G, Casinovi GC, Galeffi C, Pela A. New fl-diketones from Helichrysum italicum G. Don. Tetrahedron Lett 1967; 2: 143-8.

2. Manitto P, Monti D, Colombo E. Two new fl-diketones from Helichrysum italicum. Phytochemistry 1972; 11:2112-14.

3. Maffei Facino R, Carini M, Mariani M, Cipriani C. Anti-erythematous and photoprotective activities in guinea pigs and in man of topically applied flavonoids from Helichrysum italicum G. Don. Acta Ther 1988; 14: 323-45.

4. Maffei Facino R, Carini M, Gioia B, Arlandini E, Franzoi L, Ballabio M. Characterization of Helichrysum italicum G. Don. flavonoids by field desorption and electron impact mass spectrometry. Atti del Congresso Interdivisionale della Societ6 Chimica Italiana, 1989. Perugia, 1989: 545.

5. Shimada O, Yasuda H. Hydroxy radical scavenging action of tinoridine. Agents Actions 1986; 19: 208-14.

6. Richmond R, Halliwell B, Chauhan J, Darbre A. Superoxide dependent formation of hydroxyl radicals: detection of hydroxyl radicals by the hydroxylation of aromatic compounds. Anal Chem 1981; 118: 328-35.

7. Betts WH, Cleland LG. Effect of metal chelators and anti-inflammatory drugs on the degradation of hyaluronic acid. Arthritis Rheum 1982; 25: 1469-76.

8. Ernster L, Nordenbrand K. Mierosomal lipid peroxidation. In: Estabrook RW, Pullmann ME, eds. Methods in enzymology. New York: Academic Press, 1967; 10: 574-80.

9. Jha HC, Von Recklinghausen G, Zilliken F. Inhibition of in vitro microsomal lipid peroxidation by isoflavonoids. Biochem Pharmaco11985; 34: 1367-9.

10. Eynard AR, Tremoli E, Caruso D, Magni F, Sirtofi C, Galli G. Platelet formation of 12- hydroxyeicosatetraenoic acid and thromboxane B 2 is increased in type II A hypercholesterolemic subjects. Atherosclerosis 1986; 60: 61-6.

11. Solcaniova E, Toma S. Investigation of substituent effects on the HNMR spectra of chalcones. OrgMagn Res 1980; 14: 138-40.

12. Breitmaier E, Voelter W. In: Carbon-13 NMR spectroscopy. 3rd ed. Weinheim: VCH, 1987: 379.

13. Markham KR, Ternai B. 13C-NMR of flavonoids. II Flavonoids other than flavone and flavonol aglycons. Tetrahedron 1976; 32: 2607-12.

14. Pelter A, Ward RS, Gray TI. The carbon-13 nuclear magnetic resonance spectra of flavonoids and related compounds. J Chem Soc Perkin 11976; 2475-83.

15. Patra A, Ghosh G, Sengupta PK, Nath S. Carbon-13 NMR spectral studies on chalcones and acetophenones. Magn Res Chem 1987; 25: 734-42.

16. Wartburg A, Kuhn M. Astragalin aus Podophyllum peltatum L. und E emodi Wall. Experientia 1965; 21: 67.

Pharmacological Research, Vol. 22, No. 6, 1990 721

17. Okuyama T, Hosoyama KH, Hiroga Y, Kurono G, Takemoto T. The constituents of Osmunda spp. II. A new flavonol glycoside of Osmunda asiatica. Chem Pharm Bull 1978; 26: 3071-4.

18. ~ubuk~u B, Damadyan B. Flavonoides d'Helichrysum graveolens. Fitoterapia 1986; 57: 124-7.

19. Kappus H. A survey of chemicals inducing lipid peroxidation in biological systems. Chem Phys Lipids 1987; 45: 105-15.

20. Afanas'ev 1B, Dorozhko AI, Brodskii A, Kostyuk A, Potapovitch AL. Chelating and free radical scavenging mechanism of inhibitory action of rutin and quercetin in lipid peroxidation. Biochem Pharmaco11989; 38: 1763-9.

21. Alcaraz MJ, Ferradiz ML. Modification of arachidonic acid metabolism by flavonoids. J Ethnopharmaco11987; 21: 209-29.

22. Husain SR, Cillard J, Cillard R Hydroxyl radical scavenging activity of flavonoids. Phytochemistry 1987; 26: 2489-91.

23. Robak J, Shridi F, Wolbis M, Krolikowska M. Screening of the influence of flavonoids on lipoxygenase and cyclo-oxygenase activity, as well as on non enzymatic lipid oxidation. PolJPharmacoIPharm 1988; 40: 451-8.

24. Roshchupkin DI, Pistsov MY, Potapenko AY. Inhibition of ultraviolet light induced erythema by antioxidants. Arch Dermatol Res 1979; 266: 91-4.

25. Middleton E, Drzewiecki G. Flavonoids inhibition of human basophil histamine release stimulated by various agents. Biochem Pharmaco! 1984; 33: 3333-8.

26. Perrissoud D, Testa B. Inhibiting or potentiating effects of flavonoids on carbon tetrachloride-induced toxicity in isolated rat hepatocytes. Arzneim Forsch 1986; 36: 1249-53.

27. Longas MO, Russel CS, He XY. Evidence for structural changes in dermatan sulfate and hyaluronic acid with aging. Carbohydrate Res 1987; 159: 127-36.