Biosensors and Bioelectronics 21 (2006) 2283–2289

Epinephrine quantification in pharmaceutical formulationsutilizing plant tissue biosensors�

Fabiana S. Felix, Miyuki Yamashita, Lucio Angnes ∗Departamento de Quımica Fundamental, Instituto de Quımica, Universidade de Sao Paulo,

Av. Prof. Lineu Prestes, 748, 05508-900 Sao Paulo, SP, Brazil

Received 9 August 2005; received in revised form 26 October 2005; accepted 31 October 2005Available online 15 December 2005

Abstract

A plant tissue biosensor associated with flow injection analysis is proposed to determine epinephrine in pharmaceutical samples. The polyphenoloxidase enzymes present in the fibers of a palm tree fruits (Livistona chinensis), catalyses the oxidation of epinephrine to epinephrinequinoneas a primary product. This product is then electrochemically reduced (at −0.10 V versus Ag/AgClsat) on the biosensor surface and the resultingcurrent is used for the quantification of epinephrine. The biosensor provides a linear response for epinephrine in the concentration range from5waswitp©

K

1

mpitwa(oes

i

0d

.0 × 10−5 to 3.5 × 10−4 mol l−1. The limit of detection estimated for this interval was 1.5 × 10−5 mol l−1 and the correlation coefficient of 0.998,orking under a flow rate of 2.0 ml min−1 and using a sample loop of 100 �l. The repeatability (R.S.D. for 10 consecutive determinations of3.0 × 10−4 mol l−1 epinephrine solution) was 3.1%. The results obtained by the method here proposed were compared with the official UV

pectrophotometric procedure and also using a plant tissue reactor. The responses obtained with the proposed strategies were in good agreementith both ways of analyses, whereas the values obtained by the official spectrophotometric method was strongly affected by benzoic acid, present

n the formulation of pharmaceutical product utilized for inhalation. Such favorable results obtained with the carbon paste biosensor or utilizinghe bioreactor, joined with the simplicity of its preparation turns these procedures very attractive for epinephrine quantification in pharmaceuticalroducts.

2005 Elsevier B.V. All rights reserved.

eywords: Epinephrine; Biosensor; Polyphenol oxidase; Pharmaceutical products; Amperometric detection; Tissue biosensors

. Introduction

Epinephrine or adrenaline [1-(3,4-dihydroxyphenyl)-2-ethyloamino-ethanol] belongs to the catecholamine family and

lays an important role as neurotransmitters and hormones. Its biosynthesized in the adrenal medulla and sympathetic nerveerminals, as well as is secreted by the suprarenal gland alongith norepinephrine. It was first isolated in 1901 by Takamine

nd Aldrich, and was synthesized in 1904 by Stolz and DalkinHernandez et al., 1998). It is used in medicine in the treatmentf heart attack, bronchial asthma and cardiac surgery (Deftereost al., 1993). Air oxidation is a major problem for epinephrine inamples. The addition of antioxidants in pharmaceutical formu-

� Resume presented in the XVIII International Symposium on Electrochem-stry and Bioenergetics (Coimbra, 19–24 June, 2005).∗ Corresponding author. Tel.: +55 11 3091 3828; fax: +55 11 3818 5579.

E-mail address: luangnes@iq.usp.br (L. Angnes).

lations, for example, sodium metabisulfite, serves to minimizethe analyte oxidation. Several methods have been applied todetermine epinephrine, such as spectrophotometry (Zhu et al.,1997; Sorouraddin et al., 1998a,b; Solich et al., 2000; Nevado etal., 1996), fluorimetry (Canizares and Castro, 1995; Yang et al.,1997, 1998), liquid chromatography (Kawada et al., 1998; He etal., 1997; Sabbioni et al., 2004), capillary electrophoresis (Britz-Mckibbin et al., 1998, 1999; Lin et al., 2004), chemilumines-cence (Ragab et al., 2000; Michalowski and Halaburda, 2001),electrochemiluminescence (Zheng et al., 2001; Li et al., 2002),amperometry (Garrido et al., 1997; Ni et al., 1999), biamperom-etry (Mateo and Kojlo, 1997) and piezoelectric detection (Mo etal., 1999). However, some of these methods are very expensiveor complicated and need extraction or derivatization. Biosen-sors developed using plant tissue materials in combination withtransducers offer a good alternative compared with biosensorsbased on isolated enzymes (Leite et al., 2003; Bezerra et al.,2003; Caruso et al., 1999) as well as others analytical tech-niques for catecholamines determination. These biosensors have

956-5663/$ – see front matter © 2005 Elsevier B.V. All rights reserved.oi:10.1016/j.bios.2005.10.025

2284 F.S. Felix et al. / Biosensors and Bioelectronics 21 (2006) 2283–2289

some advantages, such as low cost, simplicity of constructionand needless of a co-factor for enzyme regeneration (Lima et al.,1997, 1998, 2003; Sidwell and Rechnitz, 1986; Wang and Lin,1989).

In this paper, a biosensor is proposed to determineepinephrine in pharmaceutical samples. Plant tissues from fruitsof a common palm tree in Brazil (Livistona chinensis), with ahigh content of polyphenol oxidase enzymes (PPO), were used.The mesocarp of this fruits was employed in the construction ofcarbon-paste biosensors, which presented very favorable hydro-dynamic characteristics for use in flow injection analysis (FIA).These kinds of sensors were well developed in our group dur-ing the last years for applications in different matrices (Limaet al., 1997, 1998, 2003). In this work, the results obtained ofthe new proposed sensor were compared with standard methodthat uses UV spectrophotometry. Beyond the official proce-dure, a plant tissue reactor associated to flow injection analysis(FIA) was employed for the spectrophotometric determinationof epinephrine in inhalation samples, exploring the formationof a second product (epinephrinechrome), which absorbs at486 nm. Details of these procedures are described in the fol-lowing sections.

2. Experimental

2.1. Apparatus

Cjvptc

eTTs

a6aalB

2

(tassfc

from Du Dyspne-Inhal—S.A.R.L.-France was acquired in alocal drugstore.

Graphite powder from Acheson (grade 38) and the mineral oilpurchased from Aldrich were utilized for the construction of thecarbon paste biosensors. An 1.00 × 10−2 mol l−1 epinephrinestock solution was prepared daily and from there the suitabledilutions were performed.

2.3. Biosensor construction and reactor preparation

The ripe fruits from the palm tree were collected into the Uni-versidade de Sao Paulo campus (USP, Sao Paulo, Brazil) andwere stored in a freezer. The mesocarp tissue located betweenthe thin skin and the woody shell was finely chopped with a stain-less steel knife over a polyethylene base. Next, the carbon pastebiosensor was prepared and homogenized by mixing graphitepowder (50%, w/w), mineral oil (35%, w/w) and tissue of theL. Chinensis (15%, w/w) finely chopped for 15 min. The pastecontaining the palm tree tissue was firmly packet into a plasticinsulin syringe (5.4 mm diameter), which extremity was cuttedoff. The electric contact was established through a stainless steelwire. The paste surface was smoothed on a paper.

For measurements with the tubular reactor containing theplant tissue, a polyethylene tube (2 mm diameter and 2.6 mmlength) was utilized. The preparation of this reactor was quitesimilar to that reported in a previous study, in other words, thefitIu

2

ip

Fp

Amperometric measurements were performed with an ECOhemie Autolab PSTAT 20 potentiostat. The design of the wall-

et electrochemical cell was identical to the one presented pre-iously byLima et al. (1997). The Ag/AgCl reference electrode,latinum auxiliary electrode and the carbon paste-working elec-rode were positioned inside the cell through holes in its teflonover.

A peristaltic pump (Ismatec, Zurich, Switzerland) wasmployed for the propulsion of the fluids during the analyses.he manifold was built with polyethyelene tubes (0.8 mm i.d.).he standard solutions and samples were introduced into thetream through an injection valve operating manually.

A Hewlett-Packard 8452A UV–vis spectrophotometer withquartz flow cell (optical path of 1.00 cm and capacity of the

50 �l) was used for measurements utilizing an in flow biore-ctor. Another conventional quartz cell (optical path of 1.00 cmnd total volume = 4 ml) was utilized for the measurements uti-izing the official spectrophotometric procedure (Farmacopeiarasileira, 1977).

.2. Reagents and solutions

All solutions were prepared with distilled and deionizedMilli-Q Plus system, Millipore), with resistivity not smallerhan 1018 � cm. Epinephrine was obtained from Sigma. Sodiumcetate, acetic acid, sodium monohydrogen phosphate, potas-ium dihydrogen phosphate, benzoic acid, sodium sulfite,odium metabissulfite and hydrochloric acid were purchasedrom Merck. The Epinephrine Dyspne-Inhal medication fabri-ated by Laboratorio Brasileiro de Biologia Ltda, under licence

brous tissue was finely chopped with a stainless steel knife andhe small pieces (∼0.08 g) were softly packed inside the reactor.n each extremity of this reactor a small amount of cotton wastilized to retain the tissue particles.

.4. Enzymatic reactions

The process involving the oxidation of epinephrine was stud-ed previously and in a simplified way, the main reactions areresented in Fig. 1 (Chen and Peng, 2003).

ig. 1. Representation of the oxidation mechanism of epinephrine catalyzed byolyphenol oxidase enzymes.

F.S. Felix et al. / Biosensors and Bioelectronics 21 (2006) 2283–2289 2285

In the case of epinephrine in contact with enzyme PPO, it isoxidized for epinephrinequinone (Chen and Peng, 2003; Kim etal., 2000; Wang et al., 2004) (Eq. (1)). This compound can bereduced electrochemically at −0.10 V (versus Ag/AgClsat) onthe biosensor surface. In this case, the amperometric responseis quantitatively related with the analyte concentration dur-ing the experiments. Alternatively, the deprotonated quinoneis able to cyclize forming leucoepinephrinechrome. This com-pound (Eq. (2)) presents elevated absorbance at 300 nm. Leu-coepinephrinechrome can be oxidized to epinephrinechrome(Eq. (3)), which presents high absorbance at 486 nm, as willbe discussed later. In this way, utilizing a plant tissue reactor,the oxidation product can be monitored both, amperometricallyat −0.10 V (versus Ag/AgClsat) or spectrophotometry at 300 and486 nm, exploring the high absorbance of the colored productformed.

2.5. Flow injection system

The flow system associated with plant tissues developed hereprovides an elevated flexibility, allowing the use of the followingfive different possibilities to measure epinephrine, which threeof these possibilities were explored during this study (a, c ande):

(

(

(

(

(

aT

ds

Fd(ts

manual injector with a sample loop typically of 100 �l, in theabsence of the bioreactor. For amperometric experiments uti-lizing the bioreactor (condition b), it must be positioned inflowing stream, before the amperometric detector. The directspectrophotometric determination of epinephrine (as recom-mended by the official procedure, condition c) is done directlyby spectrophotometry, without the use of the bioreactor. Finally,the spectrophotometric experiments involving the bioreactorcan be performed detecting the first product of the enzymaticreaction at 300 nm (leucoepinephrinechrome, condition d) or asecond product of this reaction at 486 nm (epinephrinechrome,condition e).

2.6. Pharmacopoeia method

To compare the results obtained by the methodology here pro-posed, we tried to explore the spectrophotometric method estab-lished in the Brazilian Pharmacopoeia (Farmacopeia Brasileira,1977). According to this method, solutions of epinephrine inpresence of 0.01 mol l−1 hydrochloric acid solution present max-imum absorption at 280 nm, with a molar absorptivity situatedbetween 2600 and 2800 l cm−1 mol−1. For the real samples hereanalyzed (pharmaceutical inhalation product), a strong interfer-ence from benzoic acid was verified in this wavelength. Thepresence of this interfering agent does not affect the voltammet-ric analyses and also does not interfere with the colored productsfrwe

2

05wifop

3

3

ctAnwdwe

i

a) amperometric, utilizing the biosensor and measuringepinephrinequinone at −0.10 V (versus Ag/AgClsat);

b) amperometric, utilizing a bioreactor positioned previouslyto a bare electrode;

c) spectrophotometric, measuring epinephrine directly at280 nm;

d) spectrophotometric, using the bioreactor and monitoring theleucoepinephrinechrome at 300 nm;

e) spectrophotometric, using the bioreactor and monitoring theepinephrinechrome at 486 nm.

Fig. 2 depicts schematically all these possibilities. The biore-ctor is presented with dashed lines, once can be utilized or not.he other components will be present in all experiments.

During amperometric measurements with biosensor (con-ition a), the standard solutions of epinephrine and also theamples analyzed were introduced in the flow line utilizing a

ig. 2. Schematic diagram of the flow injection system used for amperometricetermination of epinephrine. (a) Carrier solution; (b) sample; (c) sample loop;d) syringe; (e) peristaltic pump; (f) injector; (g) electrochemical cell; (h) detec-or; (i) amperometric signal; (j) bioreactor; (k) quartz flow cell; (l) detector; (m)pectrophotometric signal.

ormed by the enzymatic reaction. The utilization of a tissueeactor is able to transform epinephrine to epinephrinechrome,hich can be measured spectrophotometrically without interfer-

nces of benzoic acid.

.7. Preparation of pharmaceutical samples

Liquid formulations were appropriately diluted with.10 mol l−1 buffer to obtain concentrations in the range from.0 × 10−5 to 3.5 × 10−4 mol l−1 epinephrine during analysesith biosensor. The samples were bubbled with nitrogen dur-

ng 10 min. This time was sufficient to eliminate the inter-erence of sulfite and metabisulfite. The amperometric resultsbtained were also compared with the spectrophotometricrocedure.

. Results and discussion

.1. Preliminary experiments

Preliminary cyclic voltammetric experiments utilizing thearbon paste biosensor (scan rate = 100 mV) showed a reduc-ion wave with a maximum current situated at −0.10 V (versusg/AgClsat). This potential was selected based on the prelimi-ary experiments and later was also verified as the best conditionhen FIA experiments were performed. These experiments wereone in phosphate buffer medium (pH 6.0), and this conditionas selected with base in the previous results obtained by Lima

t al. (1997).Initially, when the analyses involved the real samples a strong

nterference from the other components (sulfite and metabisul-

2286 F.S. Felix et al. / Biosensors and Bioelectronics 21 (2006) 2283–2289

fite) was verified. This interference was minimized working atpH 4.4 followed by nitrogen bubbling. This situation imposed anew optimization of some FIA parameters.

3.2. Flow injection parameters

To determine the best conditions to be adopted in the flowinjection under pH 4.4, the sample volume and the flow ratewere investigated. In all these experiments, the potential wasmaintained at −0.10 V (versus Ag/AgClsat), once a CV of a stan-dard epinephrine solution showed the maximum in this region,utilizing now the new pH adopted.

The sample loop was varied from 50 to 200 �l and the signalwas evaluated by injection of the same epinephrine solution. Asample volume of 100 �l was selected based on the best rela-tionship between sensibility and analytical frequency. The effectof flow rate was evaluated in the range from 0.5 to 5.0 ml min−1.The best reproducibility was found utilizing a flow rate of2.0 ml min−1 and this condition was elected as the most favor-able for our studies.

Under these conditions was done a comparative study utiliz-ing phosphate and acetate buffers at pH 4.4. FIA measurementsin presence of acetate led to the more favorable signals, oncethat peak height was something larger and the reproducibilitywas also better. Although at pH 4.4 was observed a small antic-

Fig. 3. Flow injection experiments using the biosensor system for epinephrinereference solutions: (a) 3.5 × 10−4, (b) 3.0 × 10−4, (c) 2.6 × 10−4, (d)2.3 × 10−4, (e) 2.0 × 10−4, (f) 1.6 × 10−4, (g) 1.3 × 10−4, (h) 1.0 × 10−4, (i)7.0 × 10−5 and (j) 5.0 × 10−5 mol l−1 in 0.1 mol l−1 acetate buffer (pH 4.4).Potential: −0.10 V (vs. Ag/AgClsat). Flow rate: 2.0 ml min−1; volume injected:100 �l.

3.4. Interference studies for biosensor

Preliminary experiments have showed a strong interferencein the measurements. The interference of sulfite, metabisul-fite and also benzoic acid on the amperometric signal wereinvestigated, once these substances were the major compo-nents of epinephrine inhalation pharmaceutical samples. Tostudy the interference on epinephrine determination, threesolutions containing each one 1.3 × 10−4 mol l−1 epinephrineand 5.0 × 10−5 mol l−1 sulfite in the first, 5.0 × 10−5 mol l−1

metabisulfite in the second and 5.0 × 10−5 mol l−1 benzoic acidin the last one were prepared. All three solutions were completedwith acetate buffer 0.10 mol l−1 (pH 4.4). The results indicatethat sulfite and mainly metabisulfite are the greatest interferencesin the measurements involving the biosensor. The injection ofthese solutions showed a decrease of 29% caused by sulfite, 90%decrease in presence of metabisulfite, whereas benzoic acid donot interfere on the peak current of the epinephrine. To over-come this interference, the inhalation samples were bubbledwith nitrogen during 10 min in presence of the buffer solution(0.10 mol l−1, pH 4.4). This procedure eliminated the two inter-fering species in the form H2SO3.

3.5. Determination of epinephrine in inhalation samples bybiosensor and UV spectrophotometry

ig0wpll

ipation on the reduction wave (similarly to the observed by Cuiet al. (2001)), the −0.10 V potential was preferred, because onthis condition a slight increase in the peak reproducibility wasobtained.

3.3. Analytical characteristics

Under the optimized conditions, a series of experimentsin quadruplicate were carried out utilizing standard solu-tions of epinephrine in different concentrations to determinethe analytical working region. A linear response of the car-bon paste biosensor was observed in concentrations rangingbetween 5.0 × 10−5 to 3.5 × 10−4 mol l−1 of epinephrine inacetate buffer 0.10 mol l−1 (pH 4.4). Fig. 3 presents a seriesof peaks obtained in one of this study. The straight linepresented in the insert can be represented by the equationi = −0.0011 + 0.0010 C; r = 0.998, where i is the current (inmicroamperes) and C is the concentration of epinephrine (in�mol l−1). The detection limit estimated for this interval was1.5 × 10−5 mol l−1 (three times the standard deviation of theblank (Miller and Miller, 1988)) and the limit of quantification(10 times the standard deviation of the blank) was calculated as5.0 × 10−5 mol l−1.

Repeatability studies provided a relative standard deviationof 3.1% for an average of ten measurements performed with astandard solution of 3.5 × 10−4 mol l−1 epinephrine. The rela-tive standard deviations obtained for measurements in triplicateof the pharmaceutical product samples utilized for inhalationwere situated between −4.2 and +3.0%, comparatively with thelabel value. The analytical frequency attained was 60 determi-nations per hour.

The UV spectrophotometric method indicated in the Brazil-an Pharmacopoeia method (Farmacopeia Brasileira, 1977) sug-ests the measurement of epinephrine at 280 nm, in presence of.01 mol l−1 HCl. Following this methodology, the values foundere much higher than the values indicated in the label of theharmaceutical product. In addition, the voltammetric study haded to results in close concordance with the ones indicated in theabel of the pharmaceutical product and this made us doubting

F.S. Felix et al. / Biosensors and Bioelectronics 21 (2006) 2283–2289 2287

Fig. 4. Flow injection experiments using the plant tissue reactor system forepinephrine reference solutions: (a) 2.0 × 10−4 mol l−1, (b) 1.5 × 10−4 mol l−1,(c) 1.0 × 10−4 mol l−1, (d) 5.0 × 10−5 mol l−1 and (e) 2.0 × 10−5 mol l−1 in0.10 mol l−1 phosphate buffer (pH 4.4).

about the spectrophotometric results obtained. A spectrophoto-metric search demonstrated that the increase in the signal wasa contribution of benzoic acid, one of the substances present inthe formulation, which adsorbs at 280 nm too. The spectrum ofepinephrine and benzoic acid alone and of a mixture of thesetwo species shows a perfect sum of the individual absorbances.

3.6. Determination of epinephrine in inhalation samples byplant tissue reactor

Our previous experience involving plant tissues (Lima etal., 1997, 1998, 2003) led us to explore here another vari-ety of palm tree fruits (L. chinensis), for the construction ofbioreactors. The enzymatic reaction product can be also fol-lowed by spectrophotometry. Fig. 4 presents the spectrophoto-metric signals recorded at λ = 486 nm for injections of 100 �lof epinephrine in the concentration range situated between2.0 × 10−5 and 2.0 × 10−4 mol l−1, after going through the tis-sue reactor. The insert presents the media of each three peaks andthe linear regression of these points shows an excellent behavior(A = 4.86 × 10−4 + 0.157 C; r = 0.999, where A is the absorbanceand C is the concentration of epinephrine in mmol l−1).

The limit of detection estimated for this interval was3.1 × 10−6 mol l−1 (three times the standard deviation of theblank (Miller and Miller, 1988)) and the limit of quantification

Fig. 5. Absorption spectra recorded by a 1.5 × 10−4 mol l−1 epinephrine solu-tion in 0.10 mol l−1 phosphate buffer (pH 6.0) going through plant tissue reactorduring 20 min.

(10 times the standard deviation of the blank) was calculated as1.0 × 10−5 mol l−1.

Using this system, samples of the pharmaceutical productwere also analyzed. Two methodologies were tested: in the firstone, the data presented in the insert of Fig. 4 were utilized as acalibration curve and solutions and the pharmaceutical productinjected, going through the same reactor. Alternatively, samplesof the pharmaceutical product where prepared without and withthe addition of a standard method. Both led to very close results.

Table 1 shows the results obtained for the analyses of threepharmaceutical formulations using the official UV spectrophoto-metric procedure (Farmacopeia Brasileira, 1977) and comparesthese results with the ones obtained utilizing the biosensor (at−0.10 V versus Ag/AgClsat) and the bioreactor (measuring atλ = 486 nm).

A more detailed spectrophotometric investigation shows avery interesting performance of the tissue of the palm three fruits(L. chinensis). Fig. 5 presents the spectra of a 1.5 × 10−4 mol l−1

epinephrine solution that was circulating through a bioreactor.The experiment was done with 50 ml of solution and the spectrawere recorded at each minute, from 0 to 20 min.

In the bioreactor, epinephrine is oxidized by PPO to formleucoepinephrinechrome, which presents high absorptivity inthe UV region (λ = 300 nm). This product, in presence of thetissue reactor undergoes to epinephrinechrome, that present ablue-greenish color, with absorption at λ = 486 nm. Observingti

TD acope

S

sor −1 0.62 2.53 0.8

E tomete

able 1etermination of epinephrine in formulations by UV spectrophotometric (Farm

ample Epinephrine (mmol l−1)

Label value Pharmacopoeia (λ = 280 nm) (1) Biosen

136.5 168.2 ± 0.4 136.7 ±136.5 163.1 ± 0.5 140.6 ±136.5 167.0 ± 0.2 130.7 ±

rror 1: relative error between label value and pharmacopoeia (UV spectrophorror between label value and reactor (V spectrophotometric).

he behavior of epinephrine in presence of the plant tissue, its clear that the biotransformation can provide a very favor-

ia Brasileira, 1977) biosensor and plant tissue reactor

Error (%)

0.10 V (2) Reactor (λ = 486 nm) (3) 1 2 3

138.9 ± 2.5 +23.2 +0.1 +1.8133.9 ± 4.1 +19.5 +3.0 −1.9131.8 ± 2.9 +22.3 −4.2 −3.4

ric). Error 2: relative error between label value and biosensor. Error 3: relative

2288 F.S. Felix et al. / Biosensors and Bioelectronics 21 (2006) 2283–2289

able condition. There are at least three wavelengths whereepinephrine can be monitored: 280, 300 and 486 nm. Measure-ments at 280 nm can be easily affected by the presence of otheraromatic species (example: benzoic acid). At 300 nm, the molarabsorptivity is bigger and in this wavelength the interferenceof aromatic compounds will be smaller. The last absorptionpeak (maximum at λ = 486 nm was ascribed to the formation ofepinephrinechrome. This specie is dependent of the formation ofleucoepinephrinechrome and the signal measured was smaller.As for other pharmaceutical products some constituents can alsoabsorb at 300 nm, our preference in the spectrophotometric study(shown in Fig. 4) was for work at 486 nm.

Additional aspects can be discussed in relation to Fig. 5.During the enzymatic reaction, epinephrine is consumed andeven so the signal measured at 280 nm still growing. This canbe justified by the rising of the additional peak at 300 nm,which is very close and the absorption of the new specie (leu-coepinephrinechrome) sums to the signal measured at 280 nm.The results suggest also that the molar absorptivity of leu-coepinephrinechrome is much higher than the molar absorptivityof epinephrine, once the absorbance for 1.5 × 10−4 mol l−1 ofepinephrine is around 0.4 whereas for the product (just part wasbio-converted) after 20 min the absorbance surpasses 0.63. Theformation of epinephrinechrome is dependent of the previousstep, the formation of leucoepinephrinechrome. With the datahere obtained, it is not possible to compare the absorptivities oftti

otttmeTmtmb

tvsmatt

4

fbsi

sensor are the simplicity of construction, its low cost and smalltime consuming for it assemble. Plant tissue reactors associatedwith spectrophotometry present also interesting aspects. In caseswhere epinephrine cannot be quantified directly (at 280 nm, asindicated in the Brazilian Pharmacopoeia) the biotransforma-tion of epinephrine in leucoepinephrinechrome (λ = 300 nm) andafter, this product in epinephrinechrome (λ = 486 nm) permit toobtain much more selectivity in the determination. In principle,it is possible to perform a “triple redundant check” to quan-tify epinephrine when plant tissue is utilized and this aspect isnot commonly described. These characteristics turn this biosen-sor an attractive alternative to those procedures that utilize pureenzymes in pharmaceutical, food and environmental applica-tions.

Acknowledgements

The authors acknowledge the fellowships received fromFundacao de Amparo a Pesquisa do Estado de Sao Paulo(FAPESP; Processo 03/08121-0) and Conselho Nacional dePesquisa (CNPq; Processo 141530/2004-9). This work wasalso partially supported with grant from CNPq/CT-Petro (Pro-cesso 502452/2003-0) and CNPq/RHAE-Inovacao (Processo504654/2004-7).

References

B

B

B

C

C

C

C

D

FG

H

hese two species. Another aspect to be considered are reactionshat can follow this step. The formation of dimmers or polymerss well described in the literature for similar systems.

Table 1 also demonstrates that the results obtained by amper-metry and spectrophotometry (utilizing the biosensor and theissue reactor, respectively) are in very good agreement withhe values registered in the label of the pharmaceutical inhala-ion products. The results obtained utilizing the method recom-

ended in the Brazilian Pharmacopoeia leads to a systematicrror in the 20% region, caused by the presence of benzoic acid.hese results demonstrate that the method indicated in the Phar-acopoeia cannot be applied to this kind of sample. Otherwise,

he association of the official method and one of the enzymaticethods can be employed for the indirect quantification of the

enzoic acid present in this type of samples.The paired t-test was applied for results corresponding to

he reactor and biosensor of Table 1, resulting experimental t-alue of 0.40. This result suggests that there is no evidence ofystematic differences between the results obtained by eitherethods. Two degrees of freedom and 95% of confidence as well

s the critical t-value (4.30) were used to support the conclusionshis test. The critical t-value is much larger than the experimental-value (Miller and Miller, 1988).

. Conclusion

The enzymatic material contained in the plant tissue fromruits of a palm tree (L. chinensis) presents very interestingehavior. The carbon paste electrode containing the plant tis-ue shows a good performance for epinephrine quantificationn pharmaceutical formulations. The main advantages of this

ezerra, V.S., De Lima, J.L., Montenegro, M.C.B.S.M., Araujo, A.N.,Da Silva, V.L., 2003. Flow-injection amperometric determination ofdopamine in pharmaceuticals using a polyphenol oxidase biosensorobtained from soursop pulp. J. Pharm. Biomed. Anal. 33, 1025–1031.

ritz-Mckibbin, P., Kranack, A.R., Paprica, A., Chen, D.D.Y., 1998. Quan-titative assay for epinephrine in dental anesthetic solutions by capillaryelectrophoresis. Analyst 123, 1461–1463.

ritz-Mckibbin, P., Wong, J., Chen, D.D.Y., 1999. Analysis of epinephrinefrom fifteen different dental anesthetic formulations by capillary elec-trophoresis. J. Chromatogr. A 853, 535–540.

anizares, P., Castro, M.D.L., 1995. On-line coupling of isolation/in situconcentration integrated with derivative synchronous spectrofluorimetryfor the simultaneous determination of epinephrine and norepinephrine inurine. Anal. Chim. Acta 317, 335–341.

aruso, C.S., Vieira, I.D., Fatibello-Filho, O., 1999. Determination ofepinephrine and dopamine in pharmaceutical formulations using a biosen-sor based on carbon paste modified with crude extract of cara root(Dioscorea bulbifera). Anal. Lett. 32, 39–50.

hen, S.M., Peng, K.T., 2003. The electrochemical properties of dopamine,epinephrine, norepinephrine, and their electrocatalytic reactions on cobalt(II) hexacyanoferrate films. J. Electroanal. Chem. 547, 179–189.

ui, H., Wu, L., Chen, J., Lin, X., 2001. Multi-mode in situ spectroelectro-chemical studies of redox pathways of adrenaline. J. Electroanal. Chem.504, 195–200.

eftereos, T.N., Calokerinos, A.C., Efstathion, C.E., 1993. Flow injec-tion chemiluminometric determination of epinephrine, norepinephrine,dopamine and L-dopa. Analyst 118, 627–632.

armacopeia Brasileira, 1977. Sao Paulo, third ed., pp. 419–420.arrido, E.M., Lima, J.L.F.C., Delerue-Matos, C., 1997. Flow injection

amperometric determination of L-dopa, epinephrine or dopamine in phar-maceutical preparations. J. Pharm. Biomed. Anal. 15, 845–849.

e, H.B., Stein, C.M., Christman, B., Wood, A.J.J., 1997. Determination ofcatecholamines in sheep plasma by high-performance liquid chromatogra-phy with electrochemical detection: comparison of deoxyepinephrine and3,4-dihydroxybenzylamine as internal standard. J. Chromatogr. B 701,115–119.

F.S. Felix et al. / Biosensors and Bioelectronics 21 (2006) 2283–2289 2289

Hernandez, P., Sanchez, I., Paton, F., Hernandez, L., 1998. Cyclic voltamme-try determination of epinephrine with a carbon fiber ultramicroelectrode.Talanta 46, 985–991.

Kawada, T., Yamazaki, T., Akiyama, T., Sato, T., Shishido, M., Sugimachi,M., Inagaki, M., Alexander, J., Sunagawa, K., 1998. Liquid chromato-graphic determination of myocardial interstitial epinephrine. J. Chro-matogr. B 714, 375–378.

Kim, S.H., Lee, J.W., Yeo, I.H., 2000. Spectroelectrochemical and electro-chemical behavior of epinephrine at a gold electrode. Electrochim. Acta45, 2889–2895.

Leite, O.D., Lupetti, K.O., Fatibello-Filho, O., Vieira, I.C., Barbosa, A.M.,2003. Synergic effect studies of the bi-enzymatic system laccase-peroxidase in a voltammetric biosensor for catecholamines. Talanta 59,889–896.

Li, F., Cui, H., Lin, X.Q., 2002. Determination of adrenaline by usinginhibited Ru(bpy)3

2+ electrochemiluminescence. Anal. Chim. Acta 471,187–194.

Lima, A.W.O., Nascimento, V.B., Pedrotti, J.J., Angnes, L., 1997. Coconut-based plant tissue reactor for biosensing of catechol in flow-injectionanalysis. Anal. Chim. Acta 354, 325–331.

Lima, A.W.O., Vidsiunas, E.K., Nascimento, V.B., Angnes, L., 1998. Veg-etable tissue from Latania sp.: no extraordinary source of naturallyimmobilized enzymes for the detection of phenolic compounds. Analyst123, 2377–2382.

Lima, A.W.O., Bora, P.S., Ritcher, E.M., Angnes, L., 2003. Application of anew continuous flow spectrophotometric method for the characterizationof polyphenol oxidase naturally immobilized on coconut fiber. J. FoodBiochem. 27, 237–254.

Lin, C.E., Fang, I.J., Deng, Y.J., Liao, W.S., Cheng, H.T., Huang, W.P., 2004.Capillary electrophoretic studies on the migration behavior of cationicsolutes and the influence of interactions of cationic solutes with sodium

M

M

M

M

Nevado, J.J.B., Gallego, J.M.L., Laguna, P.B., 1996. Flow-injection spec-trophotometric determination of adrenaline and dopamine with sodiumhydroxide. J. Pharm. Biomed. Anal. 14, 571–577.

Ni, J.A.A., Ju, H.X., Chen, H.Y., Leech, D., 1999. Amperometric determi-nation of epinephrine with an osmium complex and nafion double-layermembrane modified electrode. Anal. Chim. Acta 378, 151–157.

Ragab, G.H., Nohta, H., Zaitsu, K., 2000. Chemiluminescence determi-nation of catecholamines in human blood plasma using 1,2-bis(3-chlorophenyl)ethylenediamine as precolumn derivatizing reagent for liq-uid chromatography. Anal. Chim. Acta 403, 155–160.

Sabbioni, C., Saracino, M.A., Mandrioli, R., Pinzauti, S., Furlanetto, S.,Gerra, G., Raggi, M.A., 2004. Simultaneous liquid chromatographic anal-ysis of catecholamines and 4-hydroxy-3-methoxyphenylethylene glycol inhuman plasma. Comparison of amperometric and coulometric detection.J. Chromatogr. A 1032, 65–71.

Sidwell, J.S., Rechnitz, G.A., 1986. Progress and challenges for biosensorsusing plant-tissue materials. Biosensors 2, 221–233.

Solich, P., Polydorou, C.H., Koupparis, M.A., Efstathiou, 2000. Auto-mated flow-injection spectrophotometric determination of catecholamines(epinephrine and isoproterenol) in pharmaceutical formulations based onferrous complex formation. J. Pharm. Biomed. Anal. 22, 781–789.

Sorouraddin, M.H., Manzoori, J.L., Kargarzadeh, E., Shabani, A.M.H., 1998a.Spectrophotometric determination of some catecholamine drugs usingsodium bismuthate. J. Pharm. Biomed. Anal. 18, 877–881.

Sorouraddin, M.H., Manzoori, J.L., Kargarzadeh, E., Shabani, A.M.H., 1998b.Spectrophtometric determination of some vic-diol containing drugs ion-exchange resins. J. Pharm. Biomed. Anal. 18, 883–888.

Wang, H.S., Huang, D.Q., Liu, R.M., 2004. Study on the electrochemicalbehavior of epinephrine at a poly(3-methylthiophene)-modified glassy car-bon electrode. J. Electroanal. Chem. 570, 83–90.

Wang, J., Lin, M.S., 1989. Dual electrode configurations with a plant tissue

Y

Y

Z

Z

dodecyl sulfate on the formation of micelles and critical micelle concen-tration. J. Chromatogr. A 1051, 85–94.

ateo, J.V.G., Kojlo, A., 1997. Flow-injection biamperometric determinationof epinephrine. J. Pharm. Biomed. Anal. 15, 1821–1828.

ichalowski, J., Halaburda, P., 2001. Flow-injection chemilumines-cence determination of epinephrine in pharmaceutical preparationsusing raw apple juice as enzyme source. Talanta 55, 1165–1171.

iller, J.C., Miller, J.N., 1988. Statistics for Analytical Chemistry, seconded. John Wiley & Sons, New York.

o, Z., Long, X., Zhang, M., 1999. Piezoelectric detection of ion pairsbetween sulphonate and catecholamines for flow injection analysis ofpharmaceutical preparations. Talanta 48, 643–648.

as generator reactor. Anal. Chim. Acta 218, 281–290.ang, J., Zhang, G., Cao, X., Sun, L., Ding, Y., 1997. Fluorimetric deter-

mination of epinephrine with 2,3-diaminonaphthalene. Spectrochim. Acta53, 1671–1676.

ang, J., Zhang, G., Wu, X., Huang, F., Lin, C., Cao, X., Sun, L.,Ding, Y., 1998. Fluorimetric determination of epinephrine with o-phenylenediamine. Anal. Chim. Acta 363, 105–110.

heng, X.W., Guo, Z.H., Zhang, Z.J., 2001. Flow-injection electrogeneratedchemiluminescence determination of epinephrine using luminol. Anal.Chim. Acta 41, 81–86.

hu, M., Huang, X., Li, J., Shen, H., 1997. Peroxidase-based spectropho-tometric methods for the determination of ascorbic acid, norepinephrine,epinephrine, dopamine and levodopa. Anal. Chim. Acta 357, 261–267.