Usefulness of enzymatic hydrolysis procedures based on the use of

pronase E as sample pre-treatment for multi-element determination

in biological materials

Pilar Bermejo-Barrera, Susana FernaÂndez-Nocelo, Antonio Moreda-PinÄeiro and

Adela Bermejo-Barrera

Department of Analytical Chemistry, Nutrition and Bromatology, Faculty of Chemistry,University of Santiago de Compostela, Av. das Ciencias s/n, 15706 ÐSantiago de Compostela,Spain

Received 17th May 1999, Accepted 30th September 1999

Several minor (Cu, Fe, Mg and Zn) and trace (Ag, As, Cd and Pb) elements were extracted from biological

materials, such as human hair and mussel, using an enzymatic hydrolysis procedure based on pronase E. The

reaction conditions, viz., pH and temperature, were ®xed at optimum values of 7.4 and 37 ³C, respectively.

Other variables affecting the enzymatic hydrolysis procedure, such as enzymatic hydrolysis time, enzyme

concentration, volume of enzyme solution and sample mass, were studied and optimized. The pH value was

adjusted with a TRIS±HCl solution as buffer system. The minor elements were measured by FAAS while trace

elements were determined by ETAAS under optimum conditions. In order to determine the total element

concentration in samples, a microwave-induced acid digestion procedure in laboratory-made low pressure

PTFE bombs was optimized. The enzymatic hydrolysis was effective for mussel samples (recoveries of about

100% were obtained for As, Cd, Cu and Mg); however, it was poor for human hair (recoveries were lower than

70%).

Introduction

Enzymes are proteins although many are conjugated proteinsand are associated with non-protein groups. Environmentaleffects such as temperature and pH are critical for enzymeactivity;1 thus, an increase in temperature increases the rate ofdenaturation of the enzyme with the loss of secondary andtertiary structure. It is suggested1 that enzymes show anoptimum operating temperature, which is a compromisebetween maximum activity for a short period of time and adecreasing activity due to denaturation for a longer period oftime. The effects of pH are due to changes in the ionic state ofthe amino acid residues of the enzyme and substrate molecules,causing varying ef®ciency in the binding of a substrate. Theenzymatic hydrolysis procedure presents certain advantagesover conventional sample pre-treatments based on acid oralkaline digestions, such as moderate conditions of tempera-ture and pH, which prevent elemental losses by volatilizationand the presence of a high acidity, with a reduction of the riskof contamination because foreign agents are not required toneutralize the excess of acid or alkali. In addition, the mainadvantage of enzymatic hydrolysis is selectivity becauseenzymes act only on certain chemical bonds and thus itshould be possible to distinguish between fractions of elementsassociated with the different components of the sample matrix.Finally, the chemical form of a species is not changed after anenzymatic digestion, so speciation studies can be developed.

The potential applicability of enzymes started to be exploitedin the forensic toxicology ®eld in order to isolate drugs fromforensic samples,2±4 mainly from human hair samples as hasbeen reviewed by Chiarotti.5 The most commonly usedenzymes have been proteinase, proteinase K and pronase,which are able to extract drugs and/or metabolites from keratinmatrices.5 However, there is not much literature on the use ofenzymatic hydrolysis to extract metals, although the ®rst workwas carried out in 1981 by Carpenter,6 who applied anenzymatic hydrolysis procedure with subtilisin Carlsberg

proteolytic enzyme to extract Cd, Cu, Pb and Tl fromhuman liver and kidney tissues. Forsyth and Marshall7,8

have used lipase type III and protease type XIV to liberateorganolead compounds from different biological matrices.More recently, enzymatic hydrolysis has been applied toextract arsenic9,10 and selenium11±14 species. Trypsin,9,10

pronase E,11,12 protease type XIV, lipase and cellulase13 andpronase14 were the enzymes used in these studies. Other worksuch as that of Heninger et al.,15 Olin et al.16 and Schwedt andNeumann17 should also be mentioned.

Pronase E, also called actinase E, is a protease fromStreptomyces griseus and it is known to contain at least tenproteolytic components: ®ve serine-type proteases, two Zn2z-endopeptidases, two Zn2z-leucine aminopeptidases and aZn2z-carboxypeptidase.18 Studies on the thermostability ofpronase E solutions have shown that the activity is maintainedwith respect to certain substrates, such as L-leucine amide, upto a temperature as high as 60±65 ³C. Activity with respect toother substrates and in the presence of certain elements canvary. More details can be obtained from the work ofYamskov et al.18

Although some work using enzymatic hydrolysis has beendeveloped, as can be seen above, the different variablesaffecting the enzymatic hydrolysis procedure such as time ofhydrolysis or enzyme concentration have not been system-atically studied. In addition, there is no information as towhether the enzymatic hydrolysis procedure is able to extractthe total element content in a sample. Moreover, data on theextraction of essential elements such as Cu, Fe, Mg and Zn, andvery toxic metals such as Cd, Pb and Ag, from seafoodproducts such as mussel have not been reported. This fact isvery important in order to evaluate the availability to man ofessential and toxic metals in foods and to assess thebiochemical signi®cance of the total amounts of such elementsin foods. Measurement of the total concentration of an elementdoes not provide information about its bioavailability or aboutits interaction with other constituents of the diet.19 Therefore,

J. Anal. At. Spectrom., 1999, 14, 1893±1900 1893

This journal is # The Royal Society of Chemistry 1999

Publ

ishe

d on

01

Janu

ary

1999

. Dow

nloa

ded

on 3

0/09

/201

3 09

:10:

28.

View Article Online / Journal Homepage / Table of Contents for this issue

the aim of this work was to study the effect of variables such astime, enzyme concentration, volume of enzyme solution andsample mass on the extraction of different essential elements(Cu, Fe, Mg and Zn) and toxic elements (Ag, As, Cd and Pb)from seafood products (mussel) and from human hair samples.In addition, the possibility of developing a sample pre-treatment method based on enzymatic hydrolysis for thedetermination of the total element content is discussed.

Experimental

Apparatus

A Perkin-Elmer (Norwalk, CT, USA) Model 3100 atomicabsorption spectrometer equipped with a conventional Ptnebulizer (glass impact bead, Perkin-Elmer) was used for Cu,Fe, Mg and Zn determination. A reduced air±acetylene ¯amewas also used throughout. A Perkin-Elmer Model 1100Binstrument equipped with a deuterium arc lamp as backgroundcorrection system, an HGA-400 graphite furnace and an AS-40autosampler was used for Ag, Cd and Pb determination. ForAs measurements, a Perkin-Elmer Model 4110ZL instrumentequipped with a Zeeman-effect background correction systemand an AS-71 autosampler was used. Hollow cathode lampswere used in all cases, except for Pb, for which an electrodelessdischarge lamp (connected to an EDL power supply, Perkin-Elmer) was used, and for As, for which an electrodelessdischarge lamp System 2 (connected to an EDL System 2power supply, Perkin-Elmer) was used. In all cases, lamps wereoperated at the lamp currents or lamp voltages, and with slit-widths and wavelengths recommended by the manufacturer.For ETAAS measurements, an injection volume of 20 ml wasapplied in all cases.

A vibrating ball mill (Retsch, Haan, Germany) equippedwith zirconia cups (15 ml in size) and zirconia balls (7 mmdiameter) was used to pulverize hair samples and to reduce theparticle size of the powdered mussel and hair samples.

A laser diffraction spectrometer, Coulter Series LS100,Fraunhofer optical model particle sizer (Coulter Electronics,Hialeah, FL, USA), was used to obtain the particle sizedistribution.

A centrifuge (Orto-Alresa, Barcelona, Spain) was also usedin order to separate the solid biological material from the liquidphase of the enzymatic digest.

A Tectron 473-100 thermostated water-bath (Selecta,Barcelona, Spain) was used to ®x the environmental tempera-ture at 37 ³C.

A Crison 506 pH-meter with a glass±calomel electrode(Crison Instruments, Barcelona, Spain) was also used.

A Panasonic domestic microwave oven (Osaka, Japan),programmable for time and microwave power from 100 to900 W, was used for total sample digestion. The poly(tetra-¯uoroethylene) (PTFE) bombs were laboratory-made withhermetic seals and were suitable for work at low pressures.

Reagents

All chemicals used were of ultrapure grade. Ultrapure water,resistance 18 MV cm21, was obtained from a Milli-Q waterpuri®cation system (Millipore, Bedford, MA, USA). AsCl3,Cd(NO3)2, Cu(NO3)2, Pb(NO3)2, Fe(NO3)3 and Zn(NO3)2

stock standard solutions, 1.000 g l21, were supplied by Merck(Darmstadt, Germany). AgNO3 stock standard solution,1.000 g l21, was obtained from BDH (Poole, Dorset, UK).Mg(NO3)2 stock standard solution, 1.000 g l21, was preparedfrom Mg(NO3)2 (BDH). LaCl3 stock standard solution, 10.000g l-1, used as ionization suppresser in Mg determination, wasprepared from LaCl3 (Aldrich, Milwaukee, WI, USA). Pdstock standard solution, 3.000 g l21, was prepared from Pd(99.999%; Aldrich) according to Welz et al.20 by dissolving

300 mg of Pd in 1 ml of concentrated nitric acid and diluting to100 ml with ultrapure water. If the dissolution was incomplete,10 ml of hydrochloric acid were added to the cold nitric acid andthe mixture was heated to gentle boiling in order to volatilizethe excess of chloride. (NH4)2HPO4 stock standard solution,2.000 g l21, was prepared from AnalaR-grade (NH4)2HPO4

(BDH). TRIS [tris(hydroxymethyl)methylamine] was obtainedfrom Scharlau (Barcelona, Spain). Pronase E, isolated fromStreptomyces griseus, was purchased from Merck (this reagentwas kept at 4 ³C). Lipase type VII, from Candida rugosa, wasobtained from Sigma (Steinheim, Germany) (this reagent waskept at 4 ³C). Nitric acid, 69.0±70.5%, was obtained fromBDH. Hydrogen peroxide, 33%, was supplied by Panreac(Barcelona, Spain). Hydrochloric acid, 37%, was obtainedfrom Scharlau. Acetone, 99.7%, was purchased from CarloErba, Milan, Italy. Reference materials DOLT-1 (Dog®shLiver) and DORM-1 (Dog®sh Muscle) were obtained from theNational Research Council of Canada.

Scalp hair samples

Scalp hair samples were taken from healthy people. Approxi-mately 1.0 g of hair sample was cut with stainless-steel scissorsfrom the nape of the neck near to the scalp region. The hairlength varied between 1 and 3 cm. A previous washing of hairsamples was carried out in order to provide an accurateassessment of endogenous metal content. The washingprocedure was that proposed by the International AtomicEnergy Agency;21 thus, hair samples were ®rst washed withultrapure water, then they were washed three times withacetone, and ®nally, they were again washed with ultrapurewater (three times). The samples were then oven-dried at100 ³C. The hair samples were subjected to a pulverizationprocess in a vibrating zirconia ball mill for 20 min using apower of 75%. In order to guarantee a small particle size,particle size distributions were recorded after differentpulverization periods. The particle sizes were determined bylaser diffraction, and mean particle lengths of about 20 mmwere achieved after this treatment. The study of the particle sizedistribution (diffractograms) for pulverized hair samplesshowed particle diameters from 0.5 to 700 mm, althoughvolumes (%) higher than 2.5 were only reached for a particlesize range of about 10±30 mm. The hair powder was stored inpre-cleaned polyethylene vials.

Mussel samples

Fresh mussel samples were collected from the Galician coast(RõÂa de Pontevedra and RõÂa de Arousa). The samples werelyophilized and ground to reduce them to a particle size lowerthan 250 mm and they were then pulverized by means of avibrating zirconia ball mill to obtain mean particle sizes of20 mm (particle size range 10±30 mm). The pulverized musselsamples were stored in pre-cleaned polyethylene vials.

Microwave acid digestion procedure

In order to obtain a reference method to evaluate theapplicability of the enzymatic hydrolysis procedure, a proce-dure based on an acid digestion induced by microwave energywas optimized. The acid digestion was carried out in a domesticmicrowave oven with laboratory-made low pressure PTFEbombs. The optimum conditions for the digestion of seafoodand human hair samples were as follows: approximately 0.2 gof seafood (mussel) or human hair sample was weighed into thePTFE bombs and 1 ml of ultrapure water and 2 ml ofconcentrated HNO3 were added. The samples were subjectedto two microwave heatings (330 W) for 5 min with a coolingstage (ice-bath for 10 min) between. The longer time and lowmicrowave power were required to avoid losses because lowpressure PTFE bombs were used. Then, 0.5 ml of 33% m/v

1894 J. Anal. At. Spectrom., 1999, 14, 1893±1900

Publ

ishe

d on

01

Janu

ary

1999

. Dow

nloa

ded

on 3

0/09

/201

3 09

:10:

28.

View Article Online

H2O2 was added and the samples were again subjected twice tomicrowave irradiation, now at 370 W for 5 min, also inserting acooling stage. Finally, the acid digests were made up to 5 mlwith ultrapure water and were stored in polyethylene bottles at4 ³C.

TRIS±HCl buffer system and pronase E solutions

The TRIS±HCl buffer system solution was prepared bycombining 50 ml of a 0.1 M TRIS solution and 45 ml of a0.1 M HCl solution, and the pH was adjusted to 7.4 with smallvolumes of the HCl solution using a pH-meter. The buffersystem was kept at 4 ³C.

The pronase E solution was prepared from 0.2 g of pronaseE made up to 25 ml with the fresh TRIS±HCl buffer solution.The pronase E solution was also kept at 4 ³C.

Enzymatic hydrolysis procedure

Amounts of mussel samples between 0.2 and 0.3 g, and humanhair between 0.05 and 0.10 g, were directly weighed intocentrifuge tubes; then, 3 ml of a 4 mg ml21 solution of pronaseE, prepared in TRIS±HCl buffer solution (pH 7.4), were addedand the samples were incubated at 37 ³C for 5 h in athermostated bath. After this time the solution was centrifugedat 3000 rpm for 20 min and the liquid phase was separated witha Pasteur pipette. In order to clean the solid residue, 1 ml ofultrapure water was added, the solid resuspended andcentrifuged again at 3000 rpm for 20 min. The liquid phasewas also separated with a Pasteur pipette and combined withthe enzymatic digest. The solution was ®nally made up to 5 mlwith ultrapure water and stored in polyethylene bottles at 4 ³C.

Analytical determinations

Cu, Fe, Mg and Zn were determined by FAAS under theoptimum conditions shown in Table 1, while Ag, As, Cd andPb were determined by ETAAS under the optimum graphitefurnace programs shown in Table 2. For both seafood andhuman hair samples, portions of about 0.3±0.5 ml of the acid orenzymatic digest were made up to 10 ml and were analyzed forZn against aqueous calibration. For Mg determination inmussel, 0.02 ml of both the acid and enzymatic digest wasdiluted to 10 ml, and for human hair 0.1 ml was made up to10 ml. LaCl3 at 10% m/v was added to samples for Mgdetermination and the determination was carried out usingaqueous calibration. Cu and Fe determinations were per-formed directly against aqueous calibrations for seafood andhuman hair acid and enzymatic digests.

Ag, As and Pb determinations were carried out using Pd at30 mg l21 as chemical modi®er. The dilution was 1z1 for Agand Pb, and 1z9 for As, for their determination in enzymaticdigests, and 1z1 for Ag, 1z9 for As and 1z19 for Pb, fortheir determination in acid digests. Cd was determined by using(NH4)2HPO4 as chemical modi®er at a concentration of20 mg l21, and the dilution was in all cases 1z9. Aqueouscalibration was used in all cases, except for Ag and Asdetermination in enzymatic digests. A fast graphite furnacetemperature program (without a charring stage) was employedfor As determination due to the losses observed when thetemperature was increased from 200 ³C.

Results and discussion

Effect of the buffer solution

Preliminary studies of the effect of the buffer solution on theextraction of the metals from biological samples were carriedout. Two buffer solutions, TRIS±HCl and H2PO4

2±HPO422,

both at pH 7.4, were evaluated. Mussel and human hairsamples were incubated for 2 h with the two buffer solutionsand also with pronase E, prepared with both buffer solutions asdescribed above. In addition, the samples were also subjectedto an enzymatic hydrolysis with lipase, also using both buffersolutions. Results for selected elements are presented inTable 3, where it is observed that neither buffer solution wasable to solubilize appreciable amounts of the analyte elements.The amounts extracted were not signi®cantly different fromthose recovered after an enzymatic hydrolysis with lipase,indicating an inability of this enzyme to hydrolyze proteins. Inorder to verify the effect of the buffer solution on metalextraction from biological samples, the experiments were alsocarried out using ultrapure water (Milli-Q water, resistivity18 MV cm21) which provides a pH of 6.8. Thus, musselsamples were incubated with ultrapure water and the resultscompared with those obtained previously. As can also be seenin Table 3, ultrapure water extracts the different elements to thesame extent as the TRIS±HCl and H2PO4

2±HPO422 buffer

solutions. Therefore, the amount of metals obtained with thebuffer solutions and also with ultrapure water can beconsidered as metals adsorbed onto the solid mussel samples.For the human hair matrix, the amounts of adsorbed metalswere negligible, indicating that the pre-analysis washingprocedures were ef®cient.

Study of the enzymatic hydrolysis procedure with pronase E

To study the variables of this procedure (time of hydrolysis,pronase E concentration, volume of pronase E solution andsample mass), the element concentration obtained wasexpressed as recovery, assuming that the total element contentis related to that obtained by an acid microwave-assisteddigestion. Thus, the samples were digested twice usingmicrowave energy as described above. A recovery of 100%indicates that the metal content from the pronase E hydrolysisprocedure is the same as the metal recovery that results fromthe microwave-assisted digestion procedure. It must be notedthat although pronase E contains a Zn-based enzyme,18 Znabsorbances were lower than 0.020, and were always subtractedfrom the Zn absorbance from the samples.

Effect of the hydrolysis time. Various times, between 1 and24 h, were tested for mussel and human hair samples. A volumeof pronase E solution of 3 ml and a pronase E concentration of2 mg ml21 were ®xed, the sample mass (mussel and humanhair) remaining at 0.2 g. Results expressed as recovery (%), fortwo replicate enzymatic hydrolyses, are shown in Fig. 1 formussel samples. As can be seen, a slight increase in thehydrolysis extraction is observed for times longer than 1 h, andthe percentage recoveries were similar for incubation timesbetween 2 and 4 h. A decrease in the recovery was observed forsome metals (Ag, As and Cd) for times longer than 5 h. Thisfact can be explained through metal losses at longer heatingtimes. In addition, it can be observed that the enzymatic

Table 1 FAAS conditions for the determination of minor elements in seafood products and human hair samples

Wavelength/nm Slit-width/nm Lamp current/mA Air ¯ow rate/l min21 C2H2 ¯ow rate/l min21

Cu 324.8 0.7 15 23.5 2.0Fe 248.3 0.2 30 23.5 2.0Mg 285.2 0.7 15 23.5 2.0Zn 213.9 0.7 15 23.5 2.0

J. Anal. At. Spectrom., 1999, 14, 1893±1900 1895

Publ

ishe

d on

01

Janu

ary

1999

. Dow

nloa

ded

on 3

0/09

/201

3 09

:10:

28.

View Article Online

hydrolysis appears quantitative for As and Cd (recovery closeto 90%), while for metals such as Pb or Cu the extraction ispoor. The recoveries obtained for human hair samples werelow, and thus only recoveries close to 40 and 20% were reachedfor Mg and Cd, respectively, for times longer than 1 h. For theother metals tested, the levels found in the enzymatic digestswere lower than the detection limits in each case. Timesbetween 2 and 5 h are adequate, and a time of 5 h was chosenfor further experiments.

Effect of the pronase E concentration. The concentration ofpronase E was studied in the range 1.0±4.0 mg ml21. Resultsobtained for mussel samples are shown in Fig. 2, where it can

be seen that an increase in the concentration of pronase Eproduces an increase in the extent of the enzymatic hydrolysisprocess. This is because there is an optimum samplemass : enzyme concentration ratio (as was shown above), andpronase E concentrations between 2.0 and 3.0 mg ml21 areadequate to hydrolyze 0.2 g of mussel sample. However, the Cdrecoveries (close to 100%) and Pb recoveries (about 20%) didnot vary signi®cantly in the pronase E concentration intervalstudied. This is because Cd is easily extracted from musseltissue even at low pronase E concentrations, while Pb is notextracted even at high pronase E concentrations. Therefore, apronase E concentration of 2 mg ml21 can be consideredadequate. As was found for the study of the time of enzymatichydrolysis, the recoveries obtained when different pronase Econcentrations were applied to human hair samples were poor,although higher recoveries were obtained for higher enzymeconcentrations. Thus, a concentration of 2.0 mg ml21 was

Table 2 ETAAS operating conditions and graphite furnace programs for the determination of trace elements in seafood products and human hairsamplesa

Operating conditionsÐ

Wavelength/nm Slit-width/nm Lamp current/mA Background correctionb Chemical modi®er/mg ml21

Ag 328.1 0.7 10 DABC Pd/30As 193.7 0.7 180c ZEBC Pd/30Cd 228.8 0.7 8 DABC (NH4)2HPO4/20Pb 283.3 0.7 10d DABC Pd/30

Graphite furnace temperature programsÐ

Step Temperature/³C Ramp rate/s Hold time/s Ar ¯ow rate/ml min21

Ag Drying 150 25 20 300Charring 800 15 10 300Atomization 1500 0 3 0 (Read)Cleaning 2200 2 3 300

As Drying 180 20 20 250Atomization 2100 0 3 0 (Read)Cleaning 2500 1 2 250

Cd Drying 150 25 20 300Charring 500 20 15 300Atomization 1300 0 3 0 (Read)Cleaning 2200 2 2 300

Pb Drying 150 20 25 300Charring 900 20 15 300Atomization 1600 0 3 0 (Read)Cleaning 2200 2 3 300

aInjection volume, 20 ml; measurement mode, peak area; pyrolytic graphite coated graphite tubes and L'vov platforms. bDABC~Deuterium arcbackground correction; ZEBC~Zeeman-effect background correction. cEDL System 2 (mA). dEDL (W).

Table 3 Concentrations of some metals extracted with TRIS±HCl andH2PO4

2±HPO422 buffer solutions, ultrapure water, pronase E and

lipase

Concentration/mg kg21

Mussel sample A Mussel sample B

TRIS±HCl Pronase E TRIS±HCl Lipase

As 0.12 22.45 0.08 0.06Cu 4.55 11.42 8.77 8.00Fe 5.58 18.38 11.35 12.49Mga 13.92 28.63 13.09 13.93

Mussel sample CH2PO4

2±HPO4222 Lipase Ultrapure water Pronase E

As 0.45 0.42 0.46 24.52Cd 0.02 0.03 0.03 0.47Cu 2.33 2.85 2.68 9.34Fe 8.11 8.23 7.72 22.35Mga 7.40 7.32 7.86 25.62

Human hair sampleTRIS±HCl H2PO4

2±HPO422 Ultrapure water Pronase E

Ag vLODb vLOD vLOD 0.17As vLOD vLOD vLOD 1.55Cd vLOD vLOD vLOD 0.79Pb vLOD vLOD vLOD 7.36a% m/m. bBelow the limit of detection.

Fig. 1 Effect of the time of enzymatic hydrolysis on the recovery ofmetals from mussel tissue: (a) major metals (Cu, Fe, Mg and Zn) and(b) trace metals (Ag, As, Cd and Pb).

1896 J. Anal. At. Spectrom., 1999, 14, 1893±1900

Publ

ishe

d on

01

Janu

ary

1999

. Dow

nloa

ded

on 3

0/09

/201

3 09

:10:

28.

View Article Online

selected, although a concentration of 4.0 mg ml21 was alsotaken into account for further studies on human hair samples.

Effect of the pronase E volume. The effect of the pronase Evolume on the enzymatic hydrolysis ef®ciency was studied, andseveral volumes, between 1.0 and 3 ml for mussel samples, andbetween 1.5 and 3.0 ml for human hair samples, were tested.Volumes lower than 1.5 ml were not tested for human hairbecause hair absorbs a large proportion of the enzyme solution,and a volume of 2 ml is the minimum needed to obtain asupernatant after centrifugation. In all cases the enzymeconcentration was the optimum one, 2.0 mg ml21. Resultsobtained for mussel samples are shown in Fig. 3, where it canbe seen that the percentage recovery remains constant in thevolume interval studied. It can also be observed that theenzymatic hydrolysis affords quantitative recoveries of Mg, Asand Cd, while the recovery is almost quantitative for Cu.Results obtained for human hair are in agreement with thoseobtained for the other experiments in that a low recovery of allthe elements was found. Hence, the two variables that controlthe pronase E hydrolysis are the time and the pronase Econcentration, and a volume of 3 ml was chosen for theenzymatic hydrolysis of mussel and human hair samples.

Effect of the sample mass. In order to study the effect of thesample mass on the pronase E hydrolysis of mussel and hair,different amounts of sample (from 0.05 to 0.4 g, for bothmussel and hair) were treated with 3 ml of 4 mg ml21 pronase Eusing a hydrolysis time of 5 h. Results are shown in Fig. 4(a)and (b) for mussel and human hair, respectively, from which itcan be seen that for human hair small amounts provide the bestresults (a human hair mass of 0.05 g can be considered asoptimum), while for mussel samples the effect is dependent onthe analyte. Thus, Cd, Cu and Mg are recovered quantitativelyfrom sample masses between 0.1 and 0.2 g, while As isrecovered from 0.1±0.4 g. The best results for Ag and Zn areobtained with mussel masses between 0.1 and 0.2 g. Finally, therecovery of Fe and Pb is higher as the mussel mass is increasedfrom 0.05 to 0.4 g. A human hair sample mass of 0.05 g wasselected, and a mussel sample of 0.2 g was chosen for all theelements except for Fe and Pb, for which a mussel mass of 0.4 gwas preferred. The explanation of this behaviour is related tothe fact that a certain pronase E amount (concentration) is ableto hydrolyze a certain amount of sample. As was mentionedabove (see under Effect of the pronase E concentration), thereis an optimum pronase E concentration : sample mass ratio. If ahigh sample mass is used, low recoveries will be obtained fora low and ®xed pronase E concentration. For As [Fig. 4(a)], ahigh recovery (around 200.0%) was obtained when 0.05 g ofmussel was used. This might be because the ETAASdetermination is not suf®ciently sensitive for As; a lowcalibration or standard additions slope is commonly obtainedso that a small absorbance value produces incorrect results.

Evaluation of ®gures of merit

The sensitivity of the methods was established as the analyteconcentration which provides an absorbance of 0.004 forFAAS determinations (Cu, Fe, Mg and Zn). In addition, thedetection and quanti®cation limits (LOD and LOQ), de®ned as3s/m and 10s/m, respectively, where s is the standard deviationof eleven replicate measurements of the blank and m is theslope of the calibration graph, were calculated and referred tothe sample mass. For ETAAS determination (Ag, As, Cd andPb), the characteristic mass (m0), de®ned as

m0~0:004CV

Ac{Ab

where C is the concentration expressed in mg l21, V is theinjection volume (20 ml), and Ac and Ab are the absorbance fora concentration C and the absorbance of the blank,respectively, was also evaluated. Results are shown inTable 4, and it can be seen that the detection limits weresuf®ciently low to allow the determination of these elements inreal samples.

The precision of the methods, expressed as the standarddeviation, considering different samples which were measuredfor 4, 3 or 2 replicates, was calculated according to the equation

sp~

�������������������������������������������������������������������PN1

i~1

�xi{�x1�2zP �xj{�x2�2z � � �

N1zN2z � � �{Ns

vuuutwhere xi, xj, etc. are the analyte concentrations for each sample,�x1; �x2; etc. are the average element concentrations for eachsample, N1, N2, etc. are the number of replicates for eachsample, and Ns is the number of samples. The standarddeviations that resulted from replicate determinations of eachelement using different sample preparation procedures (micro-wave acid digestion and pronase E hydrolysis) were notsigni®cantly different. They were about 0.05, 1.33, 0.06, 0.49,6.00, 1.85 and 9.62 mg g21 for Ag, As, Cd, Cu, Fe, Pb and Zn,respectively, and about 0.043% m/m for Mg.

Finally, the proposed enzymatic hydrolysis based on the use

Fig. 2 Effect of the pronase E concentration on the recovery of metalsfrom mussel tissue: (a) major metals (Cu, Fe, Mg and Zn) and (b) tracemetals (Ag, As, Cd and Pb).

Fig. 3 Effect of the volume of the pronase E solution on the recoveryof metals from mussel tissue: (a) major metals (Cu, Fe, Mg and Zn) and(b) trace metals (Ag, As, Cd and Pb).

J. Anal. At. Spectrom., 1999, 14, 1893±1900 1897

Publ

ishe

d on

01

Janu

ary

1999

. Dow

nloa

ded

on 3

0/09

/201

3 09

:10:

28.

View Article Online

of pronase E was applied to two certi®ed reference materials,DORM-1 (Dog®sh Muscle) and DOLT-1 (Dog®sh Liver),which provide certi®ed contents for all elements except for Ag.Analysis of human hair certi®ed reference materials was notconsidered due to the poor metal recovery achieved fromhuman hair after the pronase E pre-treatment. The results areshown in Table 5, where it can be seen that the As and Cdcontents found, after pronase E hydrolysis, are in agreementwith the certi®ed values. Cu and Mg are not in the certi®edintervals but the contents found are very close to them. For theother elements investigated, Fe, Pb and Zn, the contents foundafter the hydrolysis treatment are lower than the certi®ed values.These results are in accordance with those obtained previously,Fig. 1±4, which showed low recoveries for these elements.

On the other hand, excellent agreement with the certi®edvalues was found for all elements when an acid microwave-assisted digestion was performed. The recoveries of metalsfrom DORM-1 were 94.5¡2.8, 60.6¡5.1, 25.4¡3.3,83.4¡4.3, 16.8¡2.3 and 24.0¡3.1% for Cd, Cu, Fe, Mg, Pband Zn, respectively, while for DOLT-1 they were 99.8¡3.1,96.7¡3.0, 57.6¡4.3, 9.7¡4.6, 82.2¡4.3, 12.7¡2.0 and21.3 ¡4.1% for As, Cd, Cu, Fe, Mg, Pb and Zn, respectively.

It must be noted that poor precision (high standarddeviations) was obtained for Cu, Fe, Mg and Zn determina-tions by both the pronase E hydrolysis and microwave aciddigestion procedures. This might be due to matrix effects,which were not compensated for because background correc-tion was not employed.

Application and discussion

The enzymatic hydrolysis method based on the use of pronaseE (TRIS±HCl extraction) was applied to several mussel and

human hair samples. Each sample was subjected twice to anacid microwave-assisted digestion, twice to a pronase Ehydrolysis pre-treatment and also twice to a pre-treatmentwith the TRIS±HCl buffer solution. Table 6 shows the resultsfor mussel. In addition, the recovery of each metal is alsoshown in Table 6. As can be seen, quantitative recoveries areobtained for As, Cd and Cu, and close to quantitative valuesfor Mg. It can also be seen that recoveries of 42.2¡5.8,11.4¡5.2, 46.6¡7.4 and 89.2¡8.4% for As, Cd, Cu and Mg,respectively, are obtained when the extraction is carried out inthe absence of pronase E (TRIS±HCl extraction). Hence, onlyaround 50% of As and Cu, and around 90% of Cd, can berelated to proteins hydrolyzable by pronase E. The extractionwith TRIS±HCl appears quantitative for Mg, and thus wecannot conclude that the Mg obtained after the enzymatichydrolysis procedure is related to proteins hydrolyzed bypronase E (TRIS±HCl extraction).

Mean recoveries of 57.8¡7.7, 46.0¡21.0, 18.1¡8.0 and53.8¡10.1% are obtained for Ag, Fe, Pb and Zn, respectively.However, for Ag and Pb signi®cant differences between theresults using the enzymatic digestion and the TRIS±HClextraction are not evident and the Ag and Pb concentrationsfound in the enzymatic digest are very similar to those obtainedafter pre-treatment with the TRIS±HCl buffer solution.Therefore, recoveries of 57.8¡7.7 and 46.0¡21.0% for Agand Pb, respectively, are attributed to extraction by the buffersolution rather than to hydrolysis by pronase E (TRIS±HClextraction). For Fe, around 17% of the element found in thepronase E digest can be related to Fe extracted by the buffersolution, and thus, only around 29% can be attributed to Feassociated with proteins hydrolyzed by pronase E. The fractionof Zn associated with proteins hydrolyzed by pronase E is

Fig. 4 Effect of the mussel (a) and human hair (b) sample masses on the recovery of metals.

Table 4 Sensitivity, detection and quanti®cation limits, and characteristic mass of the methods. The values in parentheses are for human hairsamples

Sensitivity/mg l21a LOD/mg g21b LOQ/mg g21b Characteristic mass/pgc

Cu 0.12 0.5 (2.0) 1.9 (6.7) ÐFe 0.27 1.7 (6.8) 5.7 (22.7) ÐMg 8.00 18.6 (14.9) 62.1 (49.6) ÐZn 9.00 10.7 (42.8) 35.0 (142.7) ÐAg Ð 3.2 (2.1) 10.7 (7.0) 1.2¡0.2 (1.2¡0.3)As Ð 1.7 5.7 45.0¡2.8Cd Ð 3.4 (1.7) 11.3 (5.6) 0.42¡0.03 (0.33¡0.05)Pb Ð 0.04 (0.03) 0.12 (0.11) 10.0¡0.6 (12.2¡0.5)amg l21 for Mg and Zn. bng g21 for Ag and Cd. cn~4.

1898 J. Anal. At. Spectrom., 1999, 14, 1893±1900

Publ

ishe

d on

01

Janu

ary

1999

. Dow

nloa

ded

on 3

0/09

/201

3 09

:10:

28.

View Article Online

around 25%, and Zn is extracted by the TRIS±HCl solution tothe same extent, about 25%.

Quantitative recoveries were not found for human hairsamples and thus recoveries of 39.5¡7.3, 26.6¡8.0, 51.6¡2.9,56.3¡9.1, 5.6¡1.6 and 68.1¡5.0% were obtained for Cd, Cu,Fe, Mg, Pb and Zn, respectively. It must be noted that theserecoveries are related to elements associated with proteinshydrolyzed by pronase E because the concentrations of theseelements after extraction with TRIS±HCl solution were lowerthan the LOD in each case. This might be due to the previoushair washing step employed. In addition, Ag and As were notconsidered because the concentrations of both elements in thepronase E digests were lower than the LOD reached in eachcase.

Conclusion

The use of enzymatic hydrolysis procedures to extract elementsfrom biological materials appears to be an alternative for

sample pre-treatment. This is only possible for elementsassociated with proteins hydrolyzable by the enzyme. There-fore, the use of enzymes is effective in studying the fraction ofelements associated with different proteins.

The use of pronase E results in a quantitative extraction ofAs, Cd and Cu from mussel samples. Thus, these three ele-ments are associated with proteins hydrolyzable by pronaseE. However, only a fraction of the Ag, Fe, Mg, Pb and Zn isassociated with proteins hydrolyzable by pronase E.

On the other hand, metals can be incorporated in keratinduring hair growth, and thus, due to the small fraction of theelements found after pronase E hydrolysis treatment, pronaseE does not hydrolyze keratin, the main protein in human hair.

These preliminary studies have shown that the use ofenzymatic hydrolysis procedures might play an important rolein speciation studies in order to determine the differentchemical species associated with different proteins. Furtherwork will be performed in order to use enzyme mixtures toextract the elements totally.

Table 5 Results for the determination of elements in DOLT-1 and DORM-1 reference materials using the enzymatic hydrolysis and an acidmicrowave-assisted digestion

DORM-1/mg g21 DOLT-1/mg g21

Certi®edvalue

Value foundusing enzymatichydrolysisa

Value foundusing microwavedigestiona

Certi®edvalue

Value foundusing enzymatichydrolysisa

Value foundusing microwavedigestiona

As 17.7¡2.1 Ð Ð 10.1¡1.4 10.9¡0.4 11.0¡0.7Cd 0.086¡0.012 0.076¡0.003 0.081¡0.003 4.18¡0.28 3.97¡0.06 4.06¡0.10Cu 5.22¡0.33 3.36¡0.37 4.90¡0.20 20.8¡1.2 11.5¡0.6 20.8¡1.2Fe 63.6¡5.3 14.5¡0.8 65.1¡3.5 712.0¡48.0 62.2¡5.3 698.0¡27.4Mgb 0.121¡0.013 0.096¡0.001 0.116¡0.010 0.110¡0.015 0.080¡0.002 0.100¡0.010Pb 0.40¡0.12 0.07¡0.01 0.41¡0.03 1.36¡0.29 0.18¡0.01 1.37¡0.10Zn 21.3¡1.0 4.8¡0.8 20.6¡0.9 92.5¡2.3 19.0¡2.2 92.7¡0.3aExpressed as �x+ ts���

Np for N~11 and t(95.0%)~2.26. b% m/m.

Table 6 Element concentrations in mussel samples

Ag/mg g21 As/mg g21

SampleMicrowavedigestion

TRIS±HClextraction

Pronase Ehydrolysis

Recovery(%)

Microwavedigestion

TRIS±HClextraction

Pronase Ehydrolysis

Recovery(%)

1 0.15¡0.01 0.08¡0.01 0.09¡0.01 60.0 8.81¡0.32 3.52¡0.08 8.13¡0.24 92.32 0.11¡0.01 0.08¡0.01 0.06¡0.01 54.5 13.31¡0.64 4.29¡0.10 12.61¡0.20 94.73 0.10¡0.01 0.05¡0.01 0.07¡0.01 70.0 9.96¡0.24 5.30¡0.06 10.83¡0.51 108.74 0.16¡0.04 0.06¡0.01 0.08¡0.01 50.0 10.09¡0.50 4.32¡0.11 9.43¡0.40 93.55 0.11¡0.01 0.05¡0.01 0.06¡0.01 54.5 11.95¡0.64 5.00¡0.06 12.86¡0.12 107.6

Cd/mg g21 Cu/mg g21

1 0.78¡0.08 0.03¡0.01 0.73¡0.06 93.6 3.64¡0.36 2.88¡0.04 4.00¡0.10 109.92 0.45¡0.08 0.08¡0.01 0.46¡0.03 102.2 4.34¡0.04 Ð Ð Ð3 0.77¡0.01 0.10¡0.01 0.71¡0.03 92.2 5.32¡0.22 2.40¡0.01 5.23¡0.01 98.34 0.75¡0.08 0.09¡0.01 0.68¡0.05 90.1 3.50¡0.26 0.29¡0.060 3.35¡0.03 95.75 0.81¡0.04 0.06¡0.01 0.73¡0.03 90.1 3.28¡0.11 2.14¡0.07 3.57¡0.22 108.7

Fe/mg g21 Mg (%m/m)

1 83.40¡0.27 19.49¡1.15 40.97¡0.67 49.1 6.53¡0.05 4.02¡0.04 4.76¡0.60 72.92 Ð 7.85¡0.05 13.43¡0.01 Ð 8.38¡0.02 7.50¡0.07 7.58¡0.07 90.43 8.76¡0.93 12.03¡0.44 19.97¡0.76 15.5 4.02¡0.14 3.70¡0.03 3.78¡0.10 94.04 51.36¡0.79 10.03¡0.63 30.40¡0.81 59.2 1.76¡0.23 1.31¡0.02 1.60¡0.10 90.95 41.48¡1.46 7.29¡0.54 25.06¡0.47 60.4 5.28¡0.20 3.58¡0.10 4.32¡0.25 81.8

Pb/mg g21 Zn/mg g21

1 7.22¡0.03 0.81¡0.05 0.90¡0.14 18.0 106.08¡13.10 39.80¡0.01 48.51¡6.55 45.72 7.11¡0.10 1.13¡0.13 1.15¡0.06 30.2 174.33¡10.40 44.41¡0.08 103.92¡3.97 59.63 6.41¡0.26 0.89¡0.08 1.05¡0.14 16.4 196.26¡3.09 51.52¡1.76 101.44¡10.87 51.74 12.78¡0.88 0.95¡0.06 0.98¡0.13 7.7 128.97¡8.96 Ð 87.89¡0.10 68.15 4.82¡0.17 0.75¡0.10 0.88¡0.05 18.2 139.71¡4.92 12.25¡0.56 61.36¡3.32 43.9

J. Anal. At. Spectrom., 1999, 14, 1893±1900 1899

Publ

ishe

d on

01

Janu

ary

1999

. Dow

nloa

ded

on 3

0/09

/201

3 09

:10:

28.

View Article Online

References

1 D. J. Holme and H. Peck, Analytical Biochemistry, Wiley, NewYork, 2nd edn., 1993, pp. 261±317.

2 M. D. Osselton, M. D. Hammond and P. J. Twitchett, J. Pharm.Pharmacol., 1977, 29, 460.

3 M. D. Osselton, J. Forensic Sci. Soc., 1977, 17, 189.4 M. D. Osselton, I. C. Shaw and H. M. Stevens, Analyst, 1978, 103,

1160.5 M. Chiarotti, Forensic Sci. Int., 1993, 63, 161.6 R. C. Carpenter, Anal. Chim. Acta, 1981, 125, 209.7 D. S. Forsyth and W. D. Marshall, Anal. Chem., 1983, 55, 2132.8 D. S. Forsyth and W. D. Marshall, Environ. Sci. Technol., 1986,

20, 1033.9 S. Branch, L. Ebdon and P. O'Neill, J. Anal. At. Spectrom., 1994,

9, 33.10 K. J. Lamble and S. J. Hill, Anal. Chim. Acta, 1996, 334, 261.11 N. Gilon, A. Astruc, M. Astruc and M. Potin-Gautier, Appl.

Organomet. Chem., 1995, 9, 623.12 N. Gilon, M. Potin-Gautier and M. Astruc, J. Chromator. A, 1996,

750, 327.

13 Y. Tan and W. D. Marshall, Analyst, 1997, 122, 13.14 F. R. Abou-Shakra, M. P. Rayman, N. I. Ward, V. Hotton and

G. Bastian, J. Anal. At. Spectrom., 1997, 12, 429.15 Y. Heninger, M. Potin-Gautier, M. Astruc, D. Snidaro, V. Vignier

and J. Manem, Int. J. Environ. Anal. Chem., 1997, 67, 1.16 M. Olin, J. Carlsson and L. Bulow, Anal. Lett., 1995, 28, 1159.17 G. Schwedt and K.-D. Neumann, Z. Lebensm. Unters-Forsch.,

1992, 194, 152.18 I. A. Yamskov, T. V. Tichonova and V. A. Davankov, Enzyme

Microb. Technol., 1986, 8, 241.19 Report by the Analytical Methods Committee, Trace Element

Speciation in Foodstuffs Sub-Committee, Anal. Proc., 1985, 22,48.

20 B. Welz, G. Schlemmer and J. R. Mudakavi, J. Anal. At.Spectrom., 1988, 3, 695.

21 Report on the Second Research Co-ordination Meeting of IAEA,Neuherberg, 1985.

Paper 9/03924D

1900 J. Anal. At. Spectrom., 1999, 14, 1893±1900

Publ

ishe

d on

01

Janu

ary

1999

. Dow

nloa

ded

on 3

0/09

/201

3 09

:10:

28.

View Article Online