metodo vainillina para proantocianidinas.pdf

8/19/2019 metodo vainillina para proantocianidinas.pdf

http://slidepdf.com/reader/full/metodo-vainillina-para-proantocianidinaspdf 1/8

Critical Factors of Vanillin Assay for Catechins andProanthocyanidins

B a os h a n S u n ,† ,‡ J orge M. Ricard o-da -Silva ,§ and Is abel Spr anger*,†

Inst i tuto Na cional de I nvest iga ca o Agr a r ia , E st a ca o Vi t iv in ıcola Nacional, 2560 Dois P ortos, Portuga l, an dUn iversidade Tecnica de Lisboa, Inst ituto S uperior de Agronomia, 1399 Lisboa C odex, P ortugal

Several par am eters mostly affect ing t he precision a nd a ccuracy of vanillin a ssay were reexaminedan d optimized. The reexamina t ion wa s performed both by vanillin react ion w ith catechins and byvan illin reaction with purified proa nthocyanidins. In ad dition to the acid natu re and concentr a tion,the react ion t ime, the temperat ure, and t he vanillin concentrat ion, other factors such as t he wa tercontent, the presence of interfering substances, and the standard utilized, for both vanillin reactionwith catechins and va nillin react ion with proant hocya nidins, were also important . However, thekinetics of the tw o types of reactions were ma rkedly different. For estimat ing accurat ely ca techinsor proanthocyanidins that exist simultaneously in plant t issues, i t is necessary to preliminarilyseparate them from each other.

Keywords: Ca tech i n s; p r oa n th o cya n i d i n s; va n i l l i n a ssa y

INTRODUCTION

P roant hocya nidins (PA) ar e oligomeric and polymericflavans or f lavans-3-ols. They play a n importa nt rolein determining the nutritional qua lity (But ler, 1989)a ndorgan oleptic properties (Lea a nd Arnold, 1978; H a slamand Lilley, 1988; Cheynier et al., 1997) of food productsderived from plant sources. Recently, catechins a ndsome low molecular w eight P A have received consider-able attention owing to their various biological activities,in p ar t icu lar t h e ir e ffect s on a r t e r ios cleros is (M as -quellier, 1988) and t heir oxygen free ra dical scavengerability (Ricard o-da -Silva et a l., 1991).

Different methods used for PA estimation have beenproposed; th ose most commonly used were vanillina s s a y (G o ld s t ei n a n d S w a i n, 1963; B r o a d h ur s t a n dJ ones, 1978; Price et al., 1978), depolymerization in HCl/BuOH (Porter et al . , 1986), p -dimethylaminocinnam-aldehyde react ion (McMurrough a nd Dowell, 1978;Vivas et al., 1994), and Folin-Ciocalteu (Singleton andRossi, 1965).

The F olin-Ciocalteu method is based on the reducingpower of phenolic hydroxyl groups. It is n ot very specificand detects all phenols with varying sensit ivity .

The meth od of depolymerizat ion wit h H Cl/B uOH isbased on the tra nsforma tion of P A into ant hocya nidinsin hot minera l acid solutions (Porter et a l., 1986). Theformed a nthocya nidins assume a red color w ith a bsor-ban c e max ima a rou n d 550 n m. Th is me t h od is v ery

specific for PA estimation but has many shortcomings.First , the transformation of PA into anthocyanidins isn ot comp le t e; t h e y ields of colore d an t h o cy an idin sdepend on both st ructure a nd polymerizat ion degree ofP A (Sca lbert, 1992). Second, side reactions ar e commonduring the tra nsforma tion and lead t o the forma tion of

red-brown polymers a bsorbing ar ound 450 nm (Sca lbert,1992). This fact ma y undoubtedly result in estima tionerror, a nd th e application of this method for quant itat iveana lysis of P A is l imited.

T h e v a n i l l i n a s s a y f o r P A i s m o r e a t t r a c t i v e a n dpreferred because of i ts sensit ivity , specificity , an dsimplicity (Deshpan de et a l., 1986). It is quite specificto a narrow range of flavanols (monomers and polymers)an d dihydrocha lcones tha t ha ve a single bond a t t he 2,3-position and free meta-oriented hydroxy groups on theB r in g (Sa rk ar a n d H o wart h , 1976). F or ma n y y e a rst h e v an i l lin as s ay wa s e xt e n s iv ely u s ed a s a s t a n da rdcolorimetric method for f lavanols (Swain and Hillis ,1959; Broadhurst and J ones, 1978; Price et al., 1978).

However, lack of reproducibility of this method wasoften report ed (Ma xson et a l., 1972). For this r eason,ma ny investigators a ttempted to improve it (Br oadhursta nd J ones, 1978; P rice et a l., 1978).

Some authors preferred the use of p -dimethylamino-cinna ma ldehyde as an alterna t ive to vanillin (McMur-rou g h an d D o well, 1978; Viv as e t a l . , 1994). Th isreagent offered the same specificity as vanillin, but thecolor formed was not stable.

The purpose of the present work is to study somefactors mostly affecting the vanillin assay, to ensure itsprecision and accuracy.

MATERIALS AND METHODS

Materials. (+)- Cat echin and (-)-epicatechin were pur-chased from Fluka AG (Buchs, Sw itzerla nd). The vanillin wasfur nished by B DH Chem icals Lt d . (Poole, England ). Oligo-meric PA (degree of polymerization ranging from 2 to 12-15)and polymeric PA (degree of polymerization >12-15) wereisolat ed fr om gr ape seed s by colum n chr om at ogr aphy onLichroprep RP-18, as described earlier (Sun et al., 1995, 1998).The purity of both oligomeric and polymeric PA w as >92%,d et er m ined by com binat ion of var ious m et hod s, includ ingformaldehyde-HCl precipitation, HCl degrada tion-PVP , ele-m e n t a l a n a l y s i s , d i o d e a r r a y H P L C , a n d d e g r a d a t i o n w i t htoluene-R-thiol. The details in the determina tion of the purityof these PA will be described in a future paper.

* Author to whom correspondence should be addressed.† I nst i t ut o Na cional d e I nvest iga ca o Agr a ria .‡ Scholarship holder for doctorate (PRAXIS XXI Pr ogramme,

Fund aca o pa ra a Cie ncia e a Tecnologia , P ortu ga l).§ Un iversidade Tecnica de Lisboa.

4267J. Agric. Food Chem. 1998, 46, 4267−4274

10.1021/jf980366j CCC: $15.00 © 1998 American Chemical SocietyPublished on Web 09/24/1998



Preparation of the Standard Solution. The (+)-cat-echin, (-)-epicat echin, purified oligomeric P A, and polymericPA w er e d issolved r espect ively in m et hanol t o g ive st ocksolutions of 120 mg/L. Dilution of each st ock solution wa sm ad e t o est ablish st a nd ar d cur ves.

Reaction of Catechins or PA with Vanillin. The tota lvolume of reaction medium was fixed at 6 mL, comprising 1m L o f s a m p le o r s t a n d a r d , 2 . 5 m L o f r e a g en t a (v a n i ll insolution in methanol), and 2.5 mL of reagent b (HCl or H 2S O4

solution in methanol), as described earlier (Price et al., 1978;

Rev illa et a l., 1991). The van illin concentra tion in rea gent a ,acid concentra tion in reagent b, reaction temperat ure, reactiont im e, a nd w at er cont ent of r eact ion m ed ium w er e var ied t overify the influence of each of these fa ctors on vanillin rea ction.Furthermore, HCl or H 2S O4 concentra tion in reagent b ran gesfrom 1.8 to 9.0 N, vanillin concentration in reagent a from 0.5to 2% (w/v), r eaction t empera ture from 5 to 35 ° C, reactiontime from 0 to 30 min, and wa ter content of reaction mediumfrom 0 to 8%(v/v). When one factor w a s va ried, others w erefixed as follows. For (+)-catechin and (-)-epicatechin: reagenta, 1% (w/v) vanillin in metha nol; reagent b, 9 N H 2S O4 inmethan ol; reaction temperature, 30 °C; reaction t ime, 15 min.For pur if ied oligom er ic pr oant hocyanid ins and polym er icproanthocyanidins: reagent a , 1%(w/v) vanillin in metha nol;r eagent b , 9.0 N H 2S O4 in methanol; reaction temperature,r oom t em per at ur e ; r eact ion t im e, t im e r eaching m axim um

value of absorbance at λ ) 500 nm (A 500). The A 500 v a l u e w a smeas ured by a S hima dzu U V-265 spectrophotometer (J a pan ).

R ES ULTS AND DI S CUS S I ON

The influence of va rious fa ctors, th at is , acid na tureand concentrat ion, react ion t ime, temperature, watercontent, vanillin concentration, diffuse sunlight, refer-ence sta ndar d, a nd presence of interfering substa nces,on the react ion of vanillin with (+)-catechin (R va n/cat ),on the reaction of vanillin w ith (-)-epicatechin (Rva n/epica t),on the reaction of vanillin with oligomeric PA (R va n/oligo),a n d on t h e r ea c t ion of v a n i ll in w i t h p ol y me ri c P A(Rva n/polym ) was respect ively studied. Lit t le differencewas o bs e rv e d be t we e n R va n/cat a n d R va n/epica t . On th e

ot h e r h an d, R va n/oligo was s imilar t o R va n/polym , wh e re asR va n/ca t or R va n/epicat does differ significantly from R va n/oligo

or R va n/polym . For this reason, we present here only theresults obta ined from R va n/cat a n d R va n/oligo .

Influence of Acid Nature and Concentration.The vanillin react ion with catechins or PA should becarried out with acid medium, in which the acid playeda cat alyt ic role. B oth hydrochloric acid and sulfuric acidwe re u s ed a t v ar iou s con ce n t ra t ion s (G o lds t e in an dSwain, 1963; Pompei and Peri, 1971; Broadhurst andJ ones, 1978; P rice et al . , 1978; Scalbert et al . , 1989;Revilla et al., 1991).

First , we examined the influence of these two acidswith different normalities ranging from 1.8 to 9.0 N in

the solut ion of reagent b on R va n/cat . Th e res u lt s h av eshown th a t, for all norma lities tested, both hy drochloricacid a nd sulfuric acid gave good linearity (R 2 > 0.998)(Figures 1 an d 2); increasing the a cid norma lity str ength-ened A 500.

H o w ev er , f or t h e s a m e a c id n or m a l it y , t h e A 500

obtained by using H 2S O4 w a s m u ch h ig h er t h a n t h a tobt a in e d by u s in g H Cl . I n o t h er w ords , wh e n H 2S O4

wa s used, the sensitivity of the method wa s higher thanwh e n H C l wa s u s e d. Th e se re su l t s a g re ed w i t h t h o s ereported previously (Sca lbert et a l., 1989). It is evidenttha t high er a cid concentra tions could increase th e colorintensity. However, as the acid concentra tion increases,slow self-rea ction of vanillin a nd even P A decomposition

may occur (Broadhurst and J ones, 1978; Beart et al . ,1985), so the use of acid with very high concentrationshould be avoided.

Alt h ou g h s imilar re su l t s we re a ls o o bt a in e d f romR va n/oligo, Figures 3 and 4 clearly show tha t t he Rva n/oligo

wa s less sensit ive to both acid na ture a nd a cid concen-t ra t io n t h an R va n/cat .

Because the sensitivity wa s alwa ys higher with H 2S O4

tha n wit h HC l for both types of reactions, the latt er testswere performed only w ith H 2S O4.

Figure 1. Standard curves of (+)-catechin for va nillin a ssayest ablished at var ious HCl concent r at ions. Each point is amean of three replicate measurements ( S D .

Figure 2. Standard curves of (+)-catechin for va nillin a ssayestablished at various H 2S O4 concentra tions. Ea ch point is amean of three replicate measurements ( S D .

4268 J. Agric. Food Chem., Vol. 46, No. 10, 1998 Sun et al.



B utler et a l. (1982) proposed th e use of glacial a ceticac id as s olve n t in s t ea d of me t h an ol t o imp rov e t h ereaction sensitivity a nd color sta bility. Our experimen-tal results suggested that the increase of the react ionsensit ivity and color stability should be part ially duet o t h e i n cr ea s e of a c id it y i n t h e r ea c t ion m ed iu m .Considering that PA could not be dissolved in glacialacet ic acid an d tha t high sensit ivity of the react ion a nd

stability of color products could also be obtained inmet ha nol/H 2S O4 me diu m, g lac ia l ac e t ic ac id was n o tused in this work.

Influence of Reaction Time. The influence ofrea ction time on A 500 R va n/oligo at experimental conditionsis presented in Figures 5 and 6, respectively.

For R va n/cat a t a ll H 2S O4 concentrations tested, the A 500

increa sed very slowly, reached maximum va lue at 5-10min, and, after tha t , sta bilized during >20 min . I n t h ecas e o f P A, t h e A 500 increased slowly until maximumvalue obta ined and, aft er tha t, decreased gra dua lly. TheA 500 obtained at higher acid concentrat ions (>5.4 N),e i t h e r by R va n/cat o r by R va n/oligo , ap p e a re d t o be mo res t a bl e t h a n t h a t a t l ow e r a ci d con cen t r a t ion s. I naddi t ion , t h e max imu m A 500 wa s fo un d t o be l in ea r lyrelated to the concentration of (+)-catechin or oligomericPA. Th e se re su l t s in dica t e d t h a t i t wa s imp ort an t t ou t i l ize mo re c o n c e n t ra t e d ac id as re a g e n t b a n d t h emaximum A 500 should be taken a s the measured va lue.I n f a c t , t h e A 500 p re s e n t e d in F ig u re s 1-4 w a s t h emaximum value obtained.

Furthermore, the f ixat ion of react ion t ime at 15 or25 min as described by previous works (Pompei andP eri, 1971; Sca lbert et a l., 1989) is suita ble for va nillinas s ay for ca t e c hin bu t n ot s u it able fo r p u ri fied P Abecause the va lues measured ma y be not the ma ximumA 500. F or t h is re as on , w e a ls o re comme n d f ixin g t h ereact ion t ime at 15 min for catechin est imat ion. How-ever, for proanthocyanidin est imation, the maximumA 500 should be taken as the measured value.

Influence of Temperature. The influence of reac-t io n t e mp e rat u re o n t h e v an i l l in as s a y fo r bo t h (+)-catechin and oligomeric PA is presented in Figure 7.

I t h a s be en s h own t h at t h e max imu m A 500 obtainedfrom R va n/cat w as re la t e d t o t h e re ac t ion t e mpe ra t u re .On the other hand, when the maximum A 500 of R va n/cat

Figure 3. S t and ar d cur ves of pur if ied oligom er ic PA forvanillin assa y established at various HCl concentra tions. Eachpoint is a mean of th ree replicat e measurements ( S D .

Figure 4. S t and ar d cur ves of pur if ied oligom er ic PA forvanil l in assay est ablished at var ious H 2S O4 concentra tions.Ea ch point is a mean of three replicate measurements ( S D .

Figure 5. Effect of reaction time on A 500 of vanillin rea ctionw i t h (+)-catechin. (+)-Catechin concentration ) 120 mg /L.Ea ch point is a mean of three replicate measurements ( S D .

Vanillin Assay for Catechins and Proanthocyanidins J. Agric. Food Chem., Vol. 46, No. 10, 1998 4269



was obtained at a given temperature, incubation of thisme diu m at d i ffere n t t e mpe rat u re s for 10 min g av edifferent A 500 v a lu es . I n t e res t in g ly , t h is re la t ion s h ipbetween maximum A 500 a n d t h e t e m p e r a t u r e o f t h ereact ion medium wa s identical t o that shown in Figure7. In other words, for R va n/cat the dependence of A 500 on

reaction medium temperature, ranging from 5 to 35 °C,wa s reversible. These results agreed with th ose alrea dyreported (Da lby and Shuma n, 1978). Because increas-ing temperature increased t he A 500 value, 25-35 ° C isrecommended for catechin estimation.

O n t h e con t r a r y, t h e m a xi mu m A 500 of R va n/oligo

depended seldom on the rea ct ion temperatur e. Figure7 s h ow s t h a t t h e R va n/oligo performed at different tem-peratures gives nearly the same A 500.

In other words, for catechin est imation, the r eact iontempera ture should be w ell controlled, w hereas t herewas no problem for PA est imation when the tempera-t u re w as n ot s ign i fican t ly ch an g e d, an d t h e refore P Aestimation could be performed at room temperature.

Influence of Water Content. The influence ofwa ter content on both R va n/cat a n d R va n/oligo is shown inFigure 8. The A 500 de cre as e d ra p idly a lon g w i t h t h eincrease of w at er content , for both types of react ions.Only 3%of wa ter in the sa mple could decrease A 500 b y56-59% in the case of R va n/cat a n d b y 33-38% in thecase of R va n/oligo .

The sh a rp d ecline of A 500 a long with the increase ofwa ter content ma y be part ia lly explained by the th eoryof Levas seur (Ca ba nnes, 1953). According t o Levasseur,the real pH value depends strongly on the wa ter contentin t h e s amp le of o rg an ic s olve n t me dia . B e c au s e t h eA 500 a lso depends on the acid concentrat ion or, morep rop er l y s a i d, t h e p H , a ch a n g e o f w a t e r con t en tmodifies undoubtedly the A 500. I n t h e p r es en t w or k ,a b sol ut e m et h a n ol w a s u sed a s a s ol ven t a n d t h er ea c t ion m ed iu m w a s f r ee o f w a t e r . Th i s n ot on lys t abi l ize d t h e color bu t in cre as ed s ign i fica n t ly t h esensit ivity . As a lready described above, for th e samenormality, H 2S O4 gave much higher A 500 tha n HC l. Thisma y be explained by the fact tha t a considerable amountof wa ter is present in concentra ted H Cl solut ion.

However, t he Levasseur t heory cannot explain com-pletely our experiment a l results. The influence of wa ter

Figure 6. Effect of reaction time on A 500 of vanillin rea ctionwith purified oligomeric P A. P urified oligomeric PA concentra -tion ) 120 m g/L. E ach point is a m ean of t hr ee r eplicat emeasurements ( S D .

Figure 7. E ffect of rea ction t empera ture on A 500 of van illinr eact ion w it h (+)-catechin and purified oligomeric PA. Con-cent r at ions: (+)-catechin ) 120 mg/L; pur ified oligomer ic P A) 120 mg/L. Ea ch point is a mean of three replica te mea sure-ments ( S D .

Figure 8. Effect of water content in reaction medium on A 500

of vanillin reaction with (+)-catechin and purified oligomericPA. Concentrations: (+)-catechin ) 120 mg/L; puri fied oligo-meric PA ) 120 mg/L. Ea ch point is a mea n of thr ee replicat emeasurements ( S D .




content on the color intensity of R va n/cat was differentf rom t h a t of R va n/oligo. I n ot h er w o rd s , t h e c ol or edproducts formed by R va n/oligo were more resistan t or lesssensit ive to wat er tha n those formed by R va n/cat (Figure8).

Influence of Vanillin Concentration. F o r t h ev a n il li n a s s a y , t h e v a n il li n , a s r ea g e nt a , m u st b eexcessive so tha t t he reaction is complete. The influence

of vanillin concentration is presented in Figure 9.I t h a s bee n s h own t h a t t h e A 500, both for R va n/cat a n d

for R va n/oligo, did not cha nge when va nillin concentra tionw a s >10 g/L. In other w ords, the van illin concent ra tionused in the va nillin a ssay should be a t least 10 g/L.Therefore, a vanillin concentration between 10 and 12g/L in rea gent a wa s recommended. Higher van illinconcentrat ion should be avoided as the vanillin mol-ecules might be slowly self-condensed in acidic mediumto yield colored products (Broadhurst and J ones, 1978).

Influenceof ReferenceStandard. (+)-Ca techin iscurrently used as a reference standard of the vanillinassa y (Swa in a nd H illis , 1959; Price and But ler, 1977;P rice et a l., 1978; Revilla et a l., 1991). How ever, thisreference sta nda rd can not express correctly P A contentin the sample because react ivity of vanillin with cat-echin is different from that of vanillin with proantho-cyanidins (Deshpande et al., 1986; Scalbert, 1992).

In fa ct, as can be seen in Figures 2 an d 4, for th e sameconcentra tion of (+)-catechin a nd oligomeric P A and a ta g iv en H 2S O4 concentra t ion, the A 500 obt a in e d byR va n/cat is different from that by R va n/oligo.

To elucidate t he effect of reference stan da rd, w e ha veperformed the vanillin assa y using tw o samples of gra peseed metha nol extra cts with different P A concentra tionss ep ara t e d p rel imin ar i ly f rom cat e c h in s (Su n e t a l . ,1998). The P A cont ents in ea ch sam ple w ere expressed,respectively, as (+)-cat echin a nd purified oligomeric P A.The results are presented in Table 1.

I t shows clear ly th at i f the purified oligomeric PA isused as a reference stand ard , change of H 2S O4 concen-tra t ion seldom modifies the results . On the contrary ,if (+)-cat e c h in is u s e d a s a re fe ren ce s t a n dard, t h emeasured values decrease significant ly a long w ith th eincrease of H 2S O4 concentra t ion. Clearly, the purified

oligomeric PA expresses more correctly P A contents inthe sa mple.As described a bove, th e A 500 depends on not only t he

substrate (catechin or PA) concentrat ion but also theH 2S O4 concentra t ion. So we have

where [C] ) (+)-cat echin or P A concentr a tion (mg/L)and [H+] ) H 2S O4 concentration in reagent b [5-25%(v/v), corres pondin g t o 1.8-9.0 N].

According to th e experimenta l results , w e obtain bylinear regression a na lysis

For the same concentrat ions of (+)-catechin and PA,t h at is t o say [C]ca t ) [C ]pr o ) [C], we ha ve

a nd

F rom e q s 4 a n d 5 , we h a v e

Equa tions 4 a nd 5 give, respect ively, t he sensit ivity ofR va n/ca t a nd of Rva n/oligo to chan ge of H2S O4 concentration.Furt hermore, eq 6 shows tha t R va n/cat is nea rly 3.5 timesas sensitive to cha nge of H2SO4 concentra tion as R va n/oligo.

Figure9. Effect of vanillin concentration in reagent a on A 500

of vanillin reaction with (+)-catechin and purified oligomericPA. Concentrations: (+)-catechin ) 120 mg/L; purif ied olig o-meric P A ) 120 mg/L. Ea ch point is a mea n of thr ee replica temeasurements ( S D .

Table 1. PA Content of Grape Seed E xtracts Expressedby Different Reference Standards

ca lcd concn (mg /L)

sample

H 2S O4

concn (N)in reagent b A 500

a

usingoligomeric PA

a s s t a nd a rd

using(+)-catechina s s t a n d a r d

1 1.8 0.201 ( 0.006 106.6 ( 3. 2 147. 7 ( 4.63.6 0.230 ( 0.002 102.2 ( 0. 9 100. 7 ( 0.95.4 0.248 ( 0. 008 94. 7 ( 3.2 81.0 ( 2.67.2 0.266 ( 0. 009 96. 1 ( 3.5 68.5 ( 2.4

9.0 0.294 ( 0. 007 98. 0 ( 2.5 56.0 ( 1.4

2 1.8 0.415 ( 0.005 222.1 ( 2. 7 310. 2 ( 3.83.6 0.474 ( 0.003 216.8 ( 1. 4 208. 7 ( 1.35.4 0.508 ( 0.006 198.3 ( 2. 4 166. 5 ( 2.07.2 0.553 ( 0.009 208.6 ( 3. 5 143. 7 ( 2.49.0 0.613 ( 0.011 214.1 ( 4. 0 117. 6 ( 2.1

a Mean va lues of triplicat e determinations ( S D .

A 500 ) f ([C][H+]) (1)

for R va n/cat

A 500 ) 0.02[C]ca t[H+] + 0.015

(R 2) 0.989, n ) 20) (2)

for R va n/oligo

A ′500 ) 0.00575[C]pr o([H+] + 0.2625) + 0.014

(R 2) 0.995, n ) 20) (3)

d(A 500)/d([H +]) ) 0.02[C ] (4)

d(A ′500)/d([H +])) 0.00575[C ] (5)

d(A 500)/d[H+]

d(A ′500)/d[H+]) 3.48 (6)




O n t h e o t h e r h an d, f ro m e q s 2 an d 3 , we c an a ls ocalculate the acid concentration [H +]*, in which condi-tion the same concentrations of (+)-catechin and oligo-meric PA give the sa me A 500. Th a t is

where [C]ca t ) [C ]pr o, A 500 ) A ′500,

so

Eq uat ion 9 shows if the H 2S O4 concentra tion of 10.6%is chosen, equivalent concentrations of (+)-cat echin an dPA g iv e t h e s ame A 500. I f t h e H 2S O4 concentration is<10.6%, ut iliza tion of (+)-catechin as reference standardoverest ima tes PA content . I f the H 2S O4 concentra tionis >10.6%, (+)-catechin underestimates PA content.

Hence, for total analysis of catechins and PA in onesam ple containing these tw o types of compounds, with -out preliminarily separ at ing them from each other, weshould use H 2S O4 with a concentra tion of ∼10%; in t hi scase, (+)-catechin could be ta ken a s reference sta nda rd,

and the measured va lue should be approximately trueconcentra tion in sam ple. The results presented in Ta ble1 show tha t for the sa me sample, the P A concentra t ionexpressed as (+)-catechin is nearly equal to that ex-pressed a s pur ified oligomeric P A, at 10%(v/v) (3.6 N)H 2S O4.

H o we v e r , t h e p ro p e rt ie s an d t h e re ac t iv i t y o f c a t -echins a nd t hat of PA are very different , so knowledgeof the proport ion betw een these tw o types of compoundsin a sa mple, especially in gra pes a nd w ine, is importa nt.O n t h e o t h e r h a n d, f ro m t h e p oin t of v iew of colorsta bility or sensitivity of method, preliminarily separa t-ing cat echins from P A and qua ntifying them separa telyusing H 2S O4 of higher concentra t ion a re often neces-s ary .

P rice et a l. (1978) reported tha t th e use of (+)-catechinas re fe ren ce s t an da rd o v ere s t imat e d P A con t e n t . O nthe contra ry, S calbert et a l. (1989) a nd S calbert (1992)reported tha t (+)-catechin underest ima ted P A content .These a ppar ently contradictory confirmat ions ma y beexplained by the fa ct tha t t he extreme acid concentra-t ion s w e re u s ed by t h e t w o a u t h o rs ; t h e fo rmer u s edv ery di lu t e d a cidic me diu m [=3.3% (v/v) HC l fina lconcentra t ion] and the la t ter, very concentra ted a cidicmedium [=46.7%(v/v) H 2S O4 final concentra tion]. More-over, w hen H Cl is used, overest ima tion of PA contentby (+)-cat e c hin s t a n dard may be p a r t ia l ly du e t o t h epresence of wa ter in concentra ted H Cl solution becau sev an i l lin re act ion wi t h ca t e c h in s is more s en s i t iv e t o

w a t e r con t en t of r ea c t ion m ed i um t h a n t h a t w i t hproanth ocya nidins, as described a bove. In other words,our experimental results are in agreement with bothconfirmations.

Effect of Diffuse Sunlight. Table 2 presents t heeffect of diffuse sunlight on the van illin a ssay . Ana lysisof variance showed that , both for (+)-catechin a nd foroligomeric PA, there were no significant differences atthe 5%level between va nillin reaction undergone in thedark and vanillin reaction undergone in diffuse sunlight.For this reason, all of our experiments were performedin diffuse sunlight.

I t h a s b ee n r ep or t e d t h a t t h e s t a b il it y of col orproducts formed by vanillin react ion wa s significant ly

affected by direct sunlight and, to a lesser extent , bydiffuse sunlight (Broa dhurs t a nd J ones, 1978). In fa ct,according to published data by Broadhurst and J ones(1978), the effect of diffuse sunlight wa s obviouslyobserved only aft er 40 min of the reaction, tha t is, a 5%decrease in a bsorba nce ma ximum, wherea s the reactiontimes in our w ork ranged from 0 to 30 min.

Interference Substances. Ascorbic acid or ascor-bat e in t e r fe re d wi t h t h e v an i l l in as s ay in H 2S O4 me-dium due presumably to the oxidat ive nat ure of H 2S O4

(B road hurs t an d J ones, 1978). Obviously, th e a scorba teor ascorbic acid, i f i t exists in the sample, should bepreliminarily separa ted from cat echins and P A. In thecase of wine or grape extract , this problem could be

e as i ly re solv ed by f rac t ion at io n wi t h C18 Se p -P akcartridges as described earlier (Sun et al., 1998).

Chlorophyll, existing in some plant samples such asgrape stem extra ct, can int erfere with the van illin assa y.In this case, after plant t issue extract is tra nsferred intoaq u e ou s s olu t ion , ch lorop h y ll in i t can re a di ly beextracted by hexane without modifying catechins andP A (Sun et a l . , unpublished da ta ).

B roadhurst a nd J ones (1978) reported t ha t a nthocya-nins, a s the m ajority of other n on-flava nols, do not rea ctw i t h v a n il li n a t l ow con ce nt r a t i on . Th i s w a s a l s oconfirmed in the present work by testing the reactivityof va nillin with m alvidin 3-glucoside. How ever, in acidicconditions, an thocyanins a bsorb at 490-540 nm, wh ich

coincides w ith the colored product (500 nm) of thevanillin assa y. This interference can be simply elimi-n a t e d u s in g a s u it a b l e bl a n k t h a t i s t h e s a m e a s t h ereaction medium but in the absence of vanillin (Broad-hurst a nd J ones, 1978).

The rea ctivity of vanillin w ith some other non-flava nolcompounds often encountered in plant tissues, that is,phenolic acids (cinnamic acid, p -hydr oxybenzoic a cid,caffeic acid, gallic acid, p -couma ric a cid, syringic a cid),flavonols (quercetin dihydrate, kaempferol, myricetin,rut in), wa s also reexamined in this work. As expected,none of t hese compounds either reacted with vanillinor interfered w ith t he vanillin react ion w ith cat echinsor PA.

Table 2. Effect of Diffuse Sunlight on Vanillin Assay

A 500a

reaction of vanillin wit h (+)-ca techin rea ct ion of va nillin w it h oligomeric P A

light conditionmax value

(0-10 min) a t 10 min a t 20 min a t 30 minmax value

(0-10 min) a t 10 min a t 20 min a t 30 min

t u be s i n d a r k 0. 647 ( 0.002 0.643 ( 0.003 0.640 ( 0.002 0.642 ( 0.003 0.502 ( 0.003 0.500 ( 0.008 0.497 ( 0.007 0.491 ( 0.002diffuse sunlight 0.652 ( 0.006 0.651 ( 0.010 0.647 ( 0.002 0.649 ( 0.009 0.505 ( 0.003 0.503 ( 0.008 0.499 ( 0.002 0.496 ( 0.006

a Mean va lues of triplicat e determinations ( S D .

A 500

A ′500

)0.020[C]ca t[H+]* + 0.015

0.00575[C]pr o([H+]* + 0.2625) + 0.014(7)

1 )0.02[H+]*

0.00575([H +]* + 0.2625)(8)

[H+]* ) 0.106 ) 10.6% (9)




CONCLUS I ONS

Th e cr i t ica l e valu a t io n of v an i l lin as s a y is s u m-ma rized as follows : (1) separa tion of cat echins from PAin the sample and qua ntifying each of them separ at ely;(2) use of absolute methanol as solvent both for thesa mple an d for t he rea gents; (3) choice of H 2S O4 [7.2-9.0 N or 20-25%(v/v)] as rea gent b w ith strict contr olof i ts concentrat ion, preferably using the same lot ofreagent b both for esta blishing sta nda rd curves a nd for

est imating catechins a nd P A in the sa mple; (4) use ofpurified P A issue of th e source itself and (+)-catechinor (-)-epicatechin as reference standards, for PA esti-ma tion a nd ca techin estima tion, respectively; (5) elimi-nation of interfering substances (chlorophyll, ascorbicacid, . . .) and correct ion of a nthocya nins (in th e casesof red wine or extract of red grape skins) by suitableblank; (6) for catechin estimation the reaction temper-ature (between 25 and 35 °C) should be strict ly con-t r ol le d a n d t h e r e a c t ion t i m e f ix ed a t 15 m i n ; P Aestimation can be performed at room temperature, andt h e max imu m ∆A 500 s h o u ld be t ak e n a s a me as u re dvalue.

H o we ve r , fo r t wo s a mp le is s u es of d i ffere n t p lan torigins, the methods used for separa ting cat echins fromproant hocyan idins may be different . In a ddit ion, ones amp le ma y con t a in s ome u n pre dict able in t er fer in gsubstances but a nother does not . So the procedures oft h e v an i l l in as s a y fo r d i f fe re n t s a mp le s ma y be n o ta lwa y s iden t ica l . Th e fo llowin g e xa mp le is g ive n t oillustrate the application of the modified vanillin assayfor grape and wine samples.-Wine or grape extract (2-4 mL, the volume depend-

in g on t h e r ich n e ss of p h en ol ic comp ou n ds in t h esam ple) wa s passed thr ough the C18 Sep-P ak cartridgesand separ at ed using different organic solvents into threefra ction conta ining, r espectively, monomeric catechins,oligomeric PA, an d polymeric PA, as recently described(Su n e t a l ., 1998). E a ch f rac t ion w a s e va p orat e d t o

dry n es s u n der v a cu u m a t 30 ° C , a n d t h e re s idu e wa sdissolved in methanol to give a desired concentration.One milliliter of this met ha nolic solution wa s mixed firstwit h 2.5 mL of 1%(w/v) va nillin in meth an ol and thenw it h 2.5 mL of 25%(v/v) H 2S O4 in metha nol to undergovanillin react ion. The blank w as simultaneously pre-par ed in the sa me wa y except tha t 1%(w/v) va nillin inmethan ol wa s subst ituted by methanol. For young redwine, methanol-insoluble substances in the catechinf r a ct i on w e r e s om e t im es f ou n d . I n t h i s c a s e, t h emethanolic solution was filtered using a 0.45 µm filterbefore undergoing th e van illin rea ct ion. The va nillinreact ion w ith catechin fract ion w as car ried out in a 30° C w at er bath for 15 min, and th e measurement of A 500

wa s a ls o p erforme d a t 30 ° C. F or o l ig ome ric PA a n dpolymeric PA fract ions, both t he va nillin r eact ion andt h e me as u reme n t of A 500 we re p erforme d a t roomt e mpe rat u re an d t h e ma x imu m A 500 w a s t a k en a s t h eme as u red v a lu e. Th re e s t an dard cu rv es s h ou ld beconducted using (+)-catechin, purified oligomeric P A,and purified polymeric P A, to express, respect ively,catechin content, oligomeric P A content, a nd polymericP A content in gra pe extra cts or w ines.

NOTE ADDED

Attempts have been also made in our laboratory tov eri fy t h e e ffect of d i ffere n t fac t ors on t h e v an i l linreaction w ith individual procyanidinssB 1 an d B 2. H o w -

ever, the am ount of pure procya nidins B 1 a n d B 2 isolat edfrom grape seeds was not sufficient to test all factorsm en t i on ed a b ov e, s o o n ly t w o t h e m os t i m por t a n tfactors, tha t is , H 2S O4 concentration and reaction tem-pera ture, w ere verified. As expected, t he kinet ic be-havior of the va nillin rea ct ion w ith either procyan idinB 1 o r p rocy an idin B 2 i s q u i t e s i m i l a r t o t h a t w i t hpurified proanthocyanidins. In other words, the vanillinreact ion w ith either procya nidin B 1 or procyanidin B 2

was much less sensitive to change of H 2S O4 concentra -t ion and react ion temperatur e, compar ed with R va n/cat .Isolat ion of various individual procyanidins to obtainade q u at e amo u n t s o f e ac h c o mp o u n d fo r t e s t in g a l lfactors mentioned is now being done in our laboratory.Such studies will help in determining the ideal referencesta nda rd of the vanillin assay for plan t P A, that is, usingindividual procyanidin instead of the mixture of oligo-meric or polymeric proanthocyanidins, which may per-mit meaningful comparisons of PA analyses issue fromvarious sources w ithin a nd betw een la borat ories.

LITERATURE CITED

Bea rt , J . E. ; Lilley, T. H.; H asla m, E. P olyphenol interactions.Covalent binding of procyanidins to proteins during acid-

catalysed decomposition; observation on some polymericproanthocyanidins. J. Chem. Soc., Perkin Trans. 2 1985, 2 ,1439-1443.

B roadh urst , R. B.; J ones, W. T. Ana lysis of condensed ta nnin susing a cidified va nillin. J . S ci . Food A gri c. 1978, 29 , 788-794.

But ler, L. G . Effects of condensed ta nnin on anima l nutrit ion.I n C h em i s t r y a n d S i g n i f i ca n ce o f C on d en s ed T a n n i n s ;Hem ingw a y, R . W., Ka r chesy , J . J . , E d s. ; P lenum Pr ess:New York, 1989; pp 391-402.

But ler, L. G.; Pr ice, M. L.; Brotherton, J . E. Vanillin a ssay forproanthocyanidins (condensed tannins): modification of thesolvent for estimation of the degree of polymerization. J .

Agric. Food Chem . 1982, 30 , 1087-1089.Ca ba nn es, J . A. Le pH pondere et ses possibili te s (Weigh ed

pH a nd its possibilit ies). C h i m . A n a l . 1953, 35 , 28-30.

Cheynier, V.; P rieur, C. ; Guy ot, S. ; Rigaud, J . ; Moutounet, M.The structures of Tannins in Grapes and Wines and theirinteractions w ith proteins. In W i n e , N u t r i t i o n a l a n d T h er a -

peutic Benefits ; Watkins, T., Ed.; ACS Symposium Series661; Kenneth L. J ordan Heart Foundation; American Chemi-cal Society: Wash ington, DC, 1997; pp 81-93.

Da lby, A.; Sh uma n, A. C. Tempera ture-induced errors in thecolorimetric determination of tannins. An al. Bi ochem . 1978,85 , 325-327.

D e s h p a n d e , S . S . ; C h e r y a n , M . ; S a l u n k h e , D . K . T a n n i nanalysis of food products. C R C C r i t . R ev . F o od S ci . N u t r .

1986, 24 , 401-449.Gold st ein, J . L . ; S w a in, T. C hanges in t an nins in r ipening

fruits. Phytochemistry 1963, 2 , 371-383.Ha slam, E .; Lilley, T. H. Na tura l ast ringency in foodstuffs. A

molecular interpretation. C R C C r i t . R e v . F o o d S c i . N u t r .

1988, 27 , 1-40.Lea, A. G. H .; Arnold, G . M. The phenolics of ciders: bitterness

and ast r ingency. J. S ci . Food A gri c . 1978, 29 , 478-483.M asquell ier , J . Physiological e ffect s of w ine. His shar e in

alcoholism. B u l l . O . I . V . 1988, 61 , 554-578.Maxson, E. D .; Rooney, L. W. Eva luat ion of methods for tan nin

analysis in sor ghum gr ain. Cereal Chem. 1972, 49 , 719-729.

McMurrough, I . ; McDowell, J . Chromatographic separationand a ut om at ed ana lysis of f lava nols . A nal . B i ochem. 1978,17 , 92-100.

Pompei, C. ; Peri, C. Determinat ion of catechins in wines. V i t i s

1971, 9 , 312-316.Por t er , L . J . ; Hr st ich, L . N. ; Cha n, B . G . The conver sion of

pr oant hocyanid ins and pr od elphenid ins t o cyanid ins anddelphenidins. Phytochemistry 1986, 25 , 223-230.




P r i ce , M . L . ; B u t l er , L . G . R a p id v is u a l e s t i ma t i on a n dspect r ophot om et r ic d et er m inat ion of t a nnin cont ent ofsorghum grain. J . Agri c. Food Chem . 1977, 2 5 , 1268-1273.

Pr ice, M. L. ; Scoyoc, S. V.; B utler, L. G . A crit ical evalua tionof the vanillin reaction as an Assay for tannin in sorghumgr ain. J . A gri c. Food Chem . 1978, 26 , 1214-1218.

Revilla, E. ; Alonso, E. ; B ourzeix, M.; Heredia, N. Dosa ge descatechines et des proant hocya nidols dan s les vins (Deter-m inat ion of cat echins and pr oant hocyanid ins in w ines).F eu i l l . V er t O I V 1991, 82 9 .

Ricardo-da -Silva , J . M.; Dar mon, N.; Ferna ndez, Y.; Mitjavila ,

S. Oxygen free radical scavenger capa city in a queous modelsof different procyan idins from gr ape seeds. J. A gri c. Food

Chem. 1991, 39 , 1549-1552.S ar ka r , S . K. ; How ar t h, R . E . S pecif ici t y of t he va nil lin t est

for flavanols. J . A gri c. Food Chem . 1976, 24 , 317-320.Scalbert, A. Quant ita tive methods for the estima tion of ta nnins

in plant t issues. In Plant Polyphenols : Synt hesis, Pr opert ies,

S i gni f i cance ; Hemingwa y, R. W., Laks, P . S. , Eds. ; PlenumP ress: New York, 1992; pp 259-280.

S calber t A. ; M ont ies, B . ; J anin, G. Tannins in w ood : Com -par ison of d if fer ent est im at ion m et hod s. J . A g r i c . F o o d

Chem. 1989, 37 , 1324-1329.Singlet on, V. L.; Rossi, J r., J . A. Colorimetry of tota l phenolics

with phosphomolybdic-phosphotungstic acid reagent. A m .

J . E n o l . V i t i c . 1965, 16 , 144-158.

Sun, B . S. ; Lean dro, M. C.; Ricardo-da-Silva, J . M.; Spra nger,M . I . S e pa r a t i o n of g r a p e a n d w i n e p r oa n t h o cy a n i d in saccording to t heir degree of polymerisat ion. J . A gri c. Food

Chem . 1998, 46 , 1390-1396.S u n , B . S . ; S p r a n g e r , M . I . ; L e a n d r o , M . C . D o s a g e d e s

catechines et procya nidines du r a sin et d u vin. Optimisa tionde la met hode de rea ction avec la vanilline. Proc., I n t .

Conference on Polyphenols (Polyphenols 94) ; Brouillard, R.,J ay , M., Scalbert , A., Eds.; INRA: P alm a de Mallorca, 1995;pp 455-456.

S w a in, T. ; H il lis , W. E. Phenolic const i t uent s of P r u n u s

domestica. I . Qua ntita tive ana lysis of phenolic constituents.J. S ci . Food A gri c. 1959, 10 , 63-68.

Vivas, N.; Glories, Y.; Lagune, L. ; Saucier, C. ; Augustin, M.Est ima tion d u degre de polymer isa tion d es procyan idinesdu ra isin et du vin pa r la m et hode au p-dimeth yla minocin-nam aldehyde (Estima tion of the degree of polymerizationof pr ocyanid ins in gr apes by a p -dimethylaminocinnam-aldehye method). J . I n t . S c i . V i g n e V i n 1994, 2 8 , 319-336.

Received for r eview April 10, 1998. Revised man uscriptreceived J uly 31, 1998. Accepted J uly 31, 1998. B .S. t ha nkst he Fund aca o Orient e a nd the Fun da ca o pa ra a C ienci a e aTecnologia (P ortugal) for t heir gra nts.

J F980366J


metodo vainillina para proantocianidinas.pdf

Documents