Fitoterapia 76 (2005) 419–427

www.elsevier.com/locate/fitote

Ilex paraguariensis extracts inhibit AGE formation

more efficiently than green tea

Nicole Lunceford, Alejandro Gugliucci*

Glycation, Oxidation and Disease Laboratory, Division of Basic Medical Sciences, Touro University-California,

Mare Island, Vallejo, CA, USA

Received 24 November 2004; accepted 16 March 2005

Available online 13 May 2005

Abstract

Glycation, the nonenzymatic adduct formation between sugar dicarbonyls and proteins, is one

key molecular basis of diabetic complications due to hyperglycemia. Given the link between

glycation and oxidation, we hypothesized that herbal extracts with a high concentration of

antioxidant phenolics might possess significant in vitro antiglycation activities as well. The aim of

the present study was to address the hypothesis that polyphenol-rich Ilex paraguariensis (IP)

extracts are capable of inhibiting advanced glycation end-products (AGEs) formation and to

compare the potency of these extracts with green tea and with the standard antiglycation agent

aminoguanidine. When we studied the effects of IP extract on AGE fluorescence generated on

bovine serum albumin ( BSA) by glycation with methylglyoxal, a dose-dependent effect that

reaches 40% at 20 Al/ml of extract was demonstrated. Green tea did not display any significant

effect. IP polyphenols are about 2- to 2.5-fold higher in our preparations compared with green tea.

The effect of IP, therefore, may be due not only to the higher concentrations but to the different

composition in phenolics of the two botanical preparations as well. To better discriminate between

an antioxidant or a carbonyl quenching mechanism of action, we explored tryptophan fluorescence

and cross-linking by sodium dodecyl sulfate polyacrylamide gel (SDS-PAGE) electrophoresis.

The conformational changes induced by glycation and substitution of positive charges in arginine

and/or lysine produce a decrease in tryptophan fluorescence. We show that incubation of BSA with

methylglyoxal produces dramatic changes in tryptophan fluorescence that are prevented by

aminoguanidine. This also prevents the downstream effect of AGE formation. Neither green tea

nor IP extracts displayed any significant effect which rules out any significant participation as

0367-326X/$

doi:10.1016/j.

* Correspon

E-mail add

- see front matter D 2005 Elsevier B.V. All rights reserved.

fitote.2005.03.021

ding author. Tel.: +1 707 638 5237; fax: +1 707 638 5255.

ress: agugliuc@touro.edu (A. Gugliucci).

N. Lunceford, A. Gugliucci / Fitoterapia 76 (2005) 419–427420

inhibitors in the first phase of the glycation cascade. The results from the SDS-PAGE serve to

confirm the above-mentioned data. The effect is therefore due mainly to an inhibition of the second

phase of the glycation reactions, namely the free-radical mediated conversion of the Amadori

products to AGE.

Taken together our results demonstrate a significant, dose-dependent effect of water extracts of I.

paraguensis on AGE adducts formation on a protein model in vitro, whereas green tea displays no

significant effect. The inhibition of AGE formation was comparable to that obtained by using

millimolar concentrations of the standard antiglycation agent aminoguanidine.

D 2005 Elsevier B.V. All rights reserved.

Keywords: Ilex paraguariensis; Green tea; Antioxidant; Polyphenols; Methylglyoxal; Glycation; Diabetes

1. Introduction

Glycation, the nonenzymatic adduct formation between sugar aldehydes and

proteins, is one key molecular basis of diabetic complications due to hyperglycemia

[1]. In the glycation reaction, sugars react non-enzymatically with proteins and lipids to

form early glycation (Amadori or fructosamine) products. Alpha-dicarbonyl compounds

such as deoxyglucosone, methylglyoxal, and glyoxal (AGE precursors) form afterwards

by oxidation and are more reactive than the parent sugars vis-a-vis amino groups of

proteins. They form stable end-products called advanced Maillard products or advanced

glycation end-products (AGEs). The AGEs, which are irreversibly formed, accumulate

with aging, atherosclerosis, and diabetes mellitus, and are especially associated with

long-lived proteins such as collagens, lens crystallins, and nerve proteins [2–4].

Glycated proteins accumulated in vivo may provide stable active sites for catalyzing

the formation of free radicals, hence glycation and glycoxidation are intimately

interrelated [3,4].

The glycation inhibitor aminoguanidine attenuates the development of a range of

diabetic vascular complications [5]. However, some problems of toxicity have been

encountered in clinical trials with aminoguanidine, so this drug should be considered a

prototype for many new molecules which are being synthesized and tried in vitro at

present [6,7]. Another mostly unexplored avenue is the search for botanical compounds

that may inhibit the process and be a source of proto-drugs. Particularly interesting in this

regard are some herbal preparations, which have actually passed the test of time in terms of

toxicity or adverse effects. Given the link mentioned above between glycation and

oxidation, we hypothesized that herbal extracts with a high concentration of antioxidant

phenolics might possess significant in vitro anti-glycation activities as well.

In particular, our laboratory has been the first to report strong antioxidant properties in

vitro, cell culture and in vivo of Ilex paraguariensis, known as bmateQ for the Spanish-

speaking or bchimarraoQ for the Portuguese-speaking populations of South America [8–

11]. I. paraguariensis extracts are used in complementary and alternative medicine and as

a very popular folk beverage dating from pre-Colombian times in large regions of South

America. This beverage has gained popularity in the United States in the last few years. I.

paraguariensis St. Hill. (Aquifoliaceae) is a widely distributed tree or shrub in Southern

N. Lunceford, A. Gugliucci / Fitoterapia 76 (2005) 419–427 421

Brazil, North-eastern Argentina, Paraguay, and Uruguay. Its dried and ground leaves

(yerba mate) are used to prepare this traditional beverage and are included in medicinal

preparations as a mild CNS stimulant, diuretic, and in weight reducing preparations [12–

17]. The mate-drinking habit has been popular for centuries, and was adopted from the

native inhabitants of the region (Guaranıes). The Jesuits managed to develop plantations

from the wild species which were used as the economic basis for their system of missions

in Paraguay, North-eastern Argentina, and Rio Grande do Sul (Brazil). The product, which

was even shipped to Europe, was known as bJesuit’s teaQ, bParaguayan teaQ, or bmate teaQ.Nowadays, it is grown in Argentina and is exported to the United States, Europe, and Asia

where it is sold as dried drug material or as extracts in different medicinal, cosmetic

preparations, or as an ingredient in foods. Mate contains purine alkaloids (methyl

xanthines), flavonoids, vitamins such as vitamin A, the B complex, C and E, tannins,

chlorogenic acid and its derivatives, and numerous triterpenic saponins derived from

ursolic acid, known as matesaponins [12–17]. Though the presence of methyl xanthines

account for many of the pharmacological activities of yerba mate, many other very

interesting and important properties have been found to be independent of the presence of

these compounds. Mate extracts polyphenol levels are higher than those of green tea and

parallel those of red wines [18]. Mate also contains saponins that are known to bind bile

salts. There are, to our knowledge, only three reports in the literature showing

antiglycation effects of herbal compounds, one focused on thyme [19], the other on

green tea [20], and the third is a previous study showing that IP extracts protect proteins

from functional loss after early glycation [21]. The aim of the present study was to address

the hypothesis that polyphenol-rich IP extracts are capable of inhibiting AGE formation

and to compare the potency of these extracts with green tea and the standard antiglycation

agent aminoguanidine.

2. Materials and methods

2.1. General

Spectrophotometric measurements were made with a Beckman DU 640 Spectropho-

tometer (Beckman Coulter Inc., Fullerton, CA). Protein concentrations were measured by

the Bradford method [22] (BioRad, Hercules, CA).

2.2. Glycation protocol

BSA (Sigma A-7030, St Louis, MO) in 10 mmol/l sodium phosphate buffer, pH 7.4

containing 150 mmol/l NaCl and 3.7 mmol/l EDTA, to a final protein concentration of 1

mg/ml. After filtering through a 0.45 Am Millipore filter, BSA was incubated in sterile

conditions in the absence or presence of methylglyoxal (5 mmol/l final concentration) at

37 8C for a period of 0–6 days [23]. Samples were incubated in the presence or absence of

extracts of green tea, IP (2–20 Al/ml), or aminoguanidine (1–10 mmol/l). After incubation,

samples were extensively dialyzed against 10 mmol/l sodium phosphate buffer, pH 7.4

containing 150 mmol/l NaCl and kept frozen at �80 8C until analysis [23].

N. Lunceford, A. Gugliucci / Fitoterapia 76 (2005) 419–427422

2.3. Preparation of I. paraguariensis extract

I. paraguariensis powdered dry leaves (dyerba mateT) from commercial sources

(Canarias, Pando Uruguay) was freshly prepared into a mate infusion (50 g/l of herb in

water at 90 8C) in a gourd. The extracts were filtered through a 0.45 mm Millipore filter

and used the same day. The inter-assay coefficient of variation (CV) of the total

polyphenol content of the different preparations was less than 10%. Green tea (in tea bags)

were obtained from commercial sources and infusions were prepared (5 g/200 ml water).

All comparative studies between the herbal preparations were done extemporaneously.

2.4. Determination of polyphenol concentration

Total polyphenol concentrations in IP samples were determined spectrophotometrically

with the Folin–Ciocalteau phosphomolybdic–phosphotungstic acid reagents as modified

by Vinson [24].

2.5. Analysis of protein conformation changes

Tryptophan fluorescence quenching by glycation was determined as fluorescence

spectra at Ex 280 nm using a Shimadzu RS-5301 PC Spectrofluorometer with HyperRS

1.57 software for data analysis. Samples were diluted to 50 Ag/ml protein concentration

[23].

2.6. Analysis of AGE formation

AGE fluorescence spectra were determined at Ex 340 nm using the same equipment

and dilutions as above [23].

2.7. SDS-PAGE

Electrophoresis was run on 10% gels (reducing conditions). Each lane was loaded with

10 Ag protein. Equipment employed was Mini Gel III from BioRad (BioRad, Hercules,

CA). Gels were stained with Coomassie brilliant blue.

2.8. Statistical analysis

Unless otherwise stated, data are expressed as meanFS.D. Comparisons between data

were performed by the Student’s t-test (two-tailed) for unpaired samples. Data was

processed on SPSS 12.0.

3. Results

The polyphenol content in the preparations was 2.6F0.2% for I. paraguariensis and

1.1F0.2% for green tea extracts.

0

50

100

150

200

250

300

350

400

450

500

300 350 400

wavelength, nm

Try

pto

ph

an f

luo

resc

ence

, AU

MG

MG + IP 10µl/ml

MG + IP 20µl/ml

Control

0

50

100

150

200

250

300

350

400

450

500

300 350 400

Control

MG + GT10 µl/ml

MG + GT20 µl/ml

MG

A B C

0

50

100

150

200

250

300

350

400

450

500

300 350 400

wavelength, nm

Try

pt.

Flu

ore

scen

ce, A

U

Control

MG

MG + 10

mmol/l

AG

Fig. 1. Tryptophan fluorescence spectra from control (BSA) incubated in vitro for 6 days at 37 8C in the absence

or presence of 5 mmol/l methyglyoxal (MG) or 5 mmol/l methyglyoxal and two concentrations of IP or two

concentrations of GT or 10 mmol of aminoguanidine as positive control. Spectra were measured after excitation at

280 nm.

0

10

20

30

40

50

60

70

340 360 380 400 420 440 460 480 500

wavelength, nm

AG

E f

luo

resc

ence

, AU

MG

MG + IP 10 µl/ml

MG + IP 20 µl/ml

MG + IP 4 µl/ml

Control

0

10

20

30

40

50

60

70

350 400 450 500

wavelength, nm

AG

E f

luo

resc

ence

, AU

MG

MG + 10mmol/l AG

Control

Fig. 2. AGE fluorescence spectra from control (BSA) incubated in vitro for 6 days at 37 8C in the absence or

presence of 5 mmol/l methyglyoxal or 5 mmol/l methyglyoxal and three concentrations of IP extract. Spectra

were measured after excitation at 340 nm. The inset depicts the AGE fluorescence spectra of BSA incubated

presence of 5 mmol/l methyglyoxal with or without 10 mmol/l aminoguanidine.

N. Lunceford, A. Gugliucci / Fitoterapia 76 (2005) 419–427 423

30

35

40

45

50

55

60

65

70

0 5 10 15 20 25

extract concentration, µl/ml

AG

E fl

uo

resc

ence

(Ex:

340

/ Em

: 420

nm

; AU

)

Ilex paraguariensis

green tea

Fig. 3. Dose dependency of the effect of IP and GT extracts on AGE fluorescence spectra. BSAwas incubated in

vitro for 6 days at 37 8C in the absence or presence of 5 mmol/l methyglyoxal or 5 mmol/l methyglyoxal and five

concentrations of IP or GT extracts. AGE fluorescence at Ex 340/Em 420 (the main peak in Fig. 1) was measured.

Data represent meanFS.D. of two independent experiments in which each condition was analyzed in duplicates.

Differences between activities at 2, 4 and 10 and 20 Al/ml IP versus MG are significant ( P b0.005). Differences

between MG and any of the concentrations of GT are not significant.

N. Lunceford, A. Gugliucci / Fitoterapia 76 (2005) 419–427424

Fig. 1 depicts the changes in tryptophan fluorescence spectra produced in BSA after

incubation with methylglyoxal with or without co-incubation with IP (A), green tea (B), or

aminoguanidine (C) as a standard antiglycation agent. Tryptophan fluorescence is

quenched by more than 90% after glycation with methylglyoxal. No significant effect is

shown for IP (A) or green tea (B) at all the concentrations employed. As expected,

aminoguanidine (C) restored the fluorescence to ca. control values.

Fig. 2 depicts the AGE fluorescence spectra when incubated BSA samples were excited

at 340 nm. Control BSA shows no significant signal. When BSA is incubated with

methylglyoxal, fluorescent products are formed, with a major peak at 420 nm. When BSA

is co-incubated with methylglyoxal, and increasing concentrations of IP extracts inhibition

of the AGE fluorescence are shown. For comparison, as positive control we show, in the

inset, the effect of 10 mmol/l aminoguanidine (supra-pharmacological concentrations),

which reduces fluorescence more than 90%.

Fig. 3 reports the comparative effects of IP and green tea on AGE fluorescence

emission at the main peak of 420 nm shown in Fig. 2. IP produces a dose-dependent

Control MG MG + GT MG + IP

70 kDa

Fig. 4. SDS-PAGE of BSA incubated in vitro for 6 days at 37 8C in the absence or presence of 5 mmol/l

methyglyoxal (MG) or 5 mmol/l MG and IP or GT at 20 Al/ml. Gels (10%) were loaded with 10 Ag protein/lane

and stained with Coommassie.

N. Lunceford, A. Gugliucci / Fitoterapia 76 (2005) 419–427 425

inhibition in AGE accumulation, reaching 40% at 20 Al/ml. By comparison, no significant

effects were found for green tea at all the concentrations explored.

Fig. 4 shows SDS-PAGE profiles for BSA incubated in the absence (control) or

presence of methylglyoxal (MG), and co-incubated with MG and green tea at 20 Al/l(MG+GT) or MG and IP at 20 Al/l (MG+IP). Incubation with MG leads to the appearance

of a second band at 90 kDa that most likely represents cross-linked BSA with minor low

molecular weight protein contaminants that disappear from the gel. No significant effects

were found for green tea or IP at all the concentrations explored.

4. Discussion

Advanced glycation end-products are well-known contributors to the pathophysiology

of aging and diabetic chronic complications [1,2]. We decided to employ methylglyoxal

(MG) as a fast, reproducible way of generating AGEs on proteins. This model produces

very stringent glycation conditions, facilitating its detection [9,23,25]. On the other hand,

if inhibitory effects are detected in preparations under analysis, these effects need to be

strong [23,25]. Endogenously produced dicarbonyls, such as methylglyoxal, are involved

in numerous pathogenic processes in vivo, including advanced glycation end-product

formation. MG is produced in vivo as a by-product of glycolysis, from glycation of

proteins by glucose, as a product of lipid peroxidation, and from the metabolism of

acetone and threonine [9,23,25].

The effect of IP extract on AGE fluorescence in the presence of methylglyoxal was

tested and a dose-dependent effect was observed that reaches 40% with 20 Al/ml. Green

tea, in analogous conditions, did not show a really significant effect. However, a positive

trend suggests a possible mild effect of GT. As previously reported, a mild activity of

green tea vis-a-vis fluorescent AGE adducts formation was shown [20]. IP polyphenols are

about 2- to 2.5-fold higher in our preparations as compared to green tea. Nevertheless, the

effect of IP could be due to the different composition in phenolics of the two botanical

preparations and not only to the different concentrations [8–18,21,26–30]. Glycated

proteins have been shown to provide stable active sites for catalyzing the formation of free

radicals through an enzyme-like mechanism that mimics the characteristics of metal-

catalyzed oxidation systems [3,7]. These free radical reactions lead to the formation of

fluorescent and non-fluorescent AGE adducts. The action of IP could be mediated by an

effect on these latter reactions and/or by a direct nucleophilic scavenger action in the first

phase of glycation. As ethylendiamine tetraacetic acid (EDTA) was present at 3.7 mmol/l

in the incubation buffer, a chelating effect of the polyphenol-rich extracts can be ruled out.

To better discriminate between these two putative mechanisms of action we explored

tryptophan fluorescence and cross-linking by SDS-PAGE.

Tryptophan fluorescence can be quenched by changes in the protein structure induced

by chemicals. The slight conformational changes induced by glycation and substitution of

positive charges in arginine and/or lysine produce a decrease in tryptophan fluorescence

[23]. This approach is extensively used in glycation research [23]. We show here that

incubation of BSA with methylglyoxal produces dramatic changes in tryptophan

fluorescence that are prevented by aminoguanidine. Aminoguanidine acts as a nucleophilic

N. Lunceford, A. Gugliucci / Fitoterapia 76 (2005) 419–427426

scavenger, preventing the first reaction in glycation from occurring. This, of course, also

prevents the downstream effect of AGE formation. Neither green tea nor IP extracts

displayed any significant effect on these changes which rules out any significant

participation as inhibitors in the first phase of the glycation cascade. The results from the

SDS-PAGE serve to confirm the above-mentioned data, IP or green tea do not significantly

affect the cross-linking induced by the dicarbonyl, since they do not alter this first reaction.

The effect is therefore due mainly to an inhibition of the second phase of the glycation

reactions, namely the free-radical mediated conversion of the Amadori products to AGE.

These data suggest that I. paraguariensis extracts exert these effects mainly through their

antioxidant and free radical quenching capacity, which is greater than those of green tea

and red wines as we have previously reported [18]. These data are also in agreement with

those from Baynes [3] who has shown that at the millimolar concentrations of AGE

inhibitors used in many in vitro studies, inhibition of AGE formation results primarily

from antioxidant activity of the AGE inhibitors, rather than their carbonyl trapping activity

[7]. This is also supportive of the hypothesis that fostered our study.

Taken together, our results suggest a significant, dose-dependent effect of water extracts

of I. paraguariensis on AGE adducts formation on a protein model in vitro, whereas green

tea displays no significant effect. In the systems employed, an effect is already significant

at concentrations of the extracts which correspond to a 1:100 dilution of the preparations

usually drunk. The inhibition of AGE formation was comparable to that obtained by using

millimolar concentrations of the standard antiglycation agent aminoguanidine. I.

paraguariensis could be a natural candidate for studies of herbal complement to diabetes

treatment since it combines antioxidant and anti-AGE formation activities.

Acknowledgements

The authors are grateful to Mr. John Schulze for his excellent technical assistance. This

work was funded by an intramural grant from Touro University to AG.

References

[1] Gugliucci A. J Am Osteopath Assoc 2000;100:621.

[2] Bonnefont-Rousselot D. Curr Opin Clin Nutr Metab Care 2002;5:561.

[3] Baynes JW, Thorpe SR. Free Radic Biol Med 2000;28:1708.

[4] Wiernsperger N. Biofactors 2004;19:11.

[5] Sell DR, Nelson JF, Monnier VM. J Gerontol A Biol Sci Med Sci 2001;56:B405.

[6] Forbes JM, Soulis T, Thallas V, Panagiotopoulos S, Long DM, Vasan S, et al. Diabetologia 2001;44:108.

[7] Degenhardt TP, Alderson NL, Arrington DD, Beattie RJ, Basgen JM, Steffes MW, et al. Kidney Int

2002;61:939.

[8] Gugliucci A. Biochem Biophys Res Commun 1996;224:338.

[9] Gugliucci A, Menini T. Life Sci 2002;72:279.

[10] Bracesco N, Dell M, Rocha A, Menini T, Gugliucci A, Nunes E. J Altern Complement Med 2003;9:379.

[11] Gugliucci A, Stahl AJ. Biochem Mol Biol Int 1995;35:47.

[12] Filip R, Ferraro G. Eur J Nutr 2003;42:50.

[13] Chandra S, De Mejia Gonzalez E. J Agric Food Chem 2004;252:3583.

N. Lunceford, A. Gugliucci / Fitoterapia 76 (2005) 419–427 427

[14] Filip R, Lopez P, Giberti G, Coussio J, Ferraro G. Fitoterapia 2001;72:774.

[15] Kraemer KH, Taketa AT, Schenkel EP, Gosmann G, Guillaume D. Phytochemistry 1006; 42:1119.

[16] Muccillo Baisch A, Johnston K, Paganini Stein F. J Ethnopharmacol 1998;60:133.

[17] Pomilio A, Trajtemberg S, Vitale A. Phytochem Anal 2002;13:235.

[18] Bixby M, Spieler L, Menini T, Gugliucci A. Life Sci in press [Available online 9 February 2005].

[19] Morimitsu Y, Yoshida K, Esaki S, Hirota A. Biosci Biotechnol Biochem 1995;59:2018.

[20] Nakagawa T, Yokozawa T, Terasawa K, Shu S, Juneja LR. J Agric Food Chem 2002;50:2418.

[21] Gugliucci A, Menini T. Life Sci 2002;72:279.

[22] Bradford MM. Anal Biochem 1976;72:248.

[23] Nagaraj RH, Oya-Ito T, Padayatti PS, Kumar R, Mehta S, West K, et al. Biochemistry 2003;42:10746.

[24] Vinson JJ, Proch J, Bose P. Methods Enzymol 2001;335:103.

[25] Chaplen FW, Fahl WE, Cameron DC. Proc Natl Acad Sci U S A 1998;95:5533.

[26] Tachibana H, Koga K, Fujimura Y, Yamada K. Nat Struct Mol Biol 2004;11:380.

[27] Young JF, Dragstedt LO, Haraldsdottir J, Daneshvar B, Kal MA, Loft S, et al. Br J Nutr 2002;87:343.

[28] Rababah T, Hettiarachchy N, Horax R. J Agric Food Chem 2004;52:5183.

[29] Schenkel E, Montanha J, Gosmann G. Adv Exp Med Biol 1996;405:47.

[30] Schinella GR, Troiani G, Davila V, de Buschiazzo PM, Tournier HA. Biochem Biophys Res Commun

2000;269:357.