Biosmsors t Bioel~mnics 6 (1991)569-573
The development of a catechol enzyme electrode and its possible use for the diagnosis and monitoring of neural crest tumours* C. R. Tillyer & P. T. Gobin Department of Chemical Pathology, Royal Marsden Hospital, Fulham Road, London SW3 6JJ, UK (Received 29 June 1990; received in revised form 12 November 1990, accepted 22 November 1990)
Ah&a& A catechol enzyme electrode is described, in which a Clark-type oxygen electrode is coupled to immobilised polyphenol oxidase in albumin cross linked with glutaraldehyde on a dialysis membrane. Electrode calibration, response time, pH response profile, stability, detection limit and selectivity are evaluated and the feasibility of using the electrode for the measurement of ~~holamines in the urine of patients with neural crest turnours is assessed. Keywords: catechol, polyphenol cytoma, neuroblastoma, urine.
oxidase, enzyme electrode, phaeochromo-
INTRODUCTION Turnouts of the neural crest (phaeochromocytoma and neuroblastoma) in humans secrete abnormal quantities of catecholamines and their metabolites (Boliun, 1966, Stentrom et al., 1983). The detection and measurement of these compounds in serum and urine of affected patients is a well-established aid to the diagnosis and treatment of these cancers (Weinkove, 1987). Current methods of estimation of these metabolites are either spectrophotometric (Nagatsu, 1973)or chromatographic (Moyer et&., 1978; Schleicher et al., 1983; Dower er al., 1984). They generally require expensive ins~men~tion and skilled staff, and are rather laborious and time-consuming resulting in a waiting time for *Paper presented at Biosensors 90, Singapore, 2-4 May 1990. 09.5~5663/91/$03.50 @ 1991 Elsevier Science Publishers Ltd.
the result for a particular patient of weeks or even months, with a con~mi~nt delay in ~atment. Also, the tumours are rather rare but often suspected and the cost of detection in health economic terms is therefore high, particularly when whole populations are screened, as in the case of neuroblastoma (Takeda, 1989; Lemieux et al., 1989).Biosensors are simple, robust, cheap and provide a rapid result; they may therefore provide a more effective method for the detection and monitoring of these metabolites than currently used methods (Turner et al., 1987). We have developed a catechol enzyme electrode using an immobilised catechol oxidase (mushroom polyphenol oxidase EC 1.14.18.1). The enzyme catalyses the conversion of catechol to o-benzoquinone in the presence of oxygen; the rate of oxygen consumption is measured by a Clark electrode and is proportional to the catechol concentration. The feasibility of using 569
C. R. Tillyer, I? T. Gobin
Biosensorsh Bioelectmnics6 (1991) 569-573
this electrode in the detection and monitoring of abnormal amounts of catecholamines in patients with neural crest turnouts is evaluated.
MATERIALS AND METHODS The oxygen electrode was an Orion model 97-08 connected to an Orion 720 electrometer (Orion Research, East Sussex, UK). For the response curves the electrometer was interfaced to a BBC Master series microcomputer programmed to take readings at intervals of 10 s using the authors’ own software. Bovine albumin; glutaraldehyde; catechol; dopamine-, adrenaline-, noradrenaline-, metadrenaline- and normetadrenaline-hydrochloride; dihydroxyphenylalanine (DOPA); dihydroxyphenylacetic acid (DHPA); dihydroxymandelic acid (DHMA); vanillylmandelic acid (VMA); homovanillic acid (I-WA); and mushroom polyphenol oxidase (tyrosinase) were purchased from Sigma (Dorset, UK). The specific activities of the enzyme were 648OOOUmg-’ as catechol oxidase, 82 000 U mg-’ as polyphenol oxidase and 3430 U mg-’ as tyrosinase. All optically active compounds were DL mixtures. The dialysis membrane was an Ezee-mount premount membrane from Elkay (Shrewsbury, MA, USA).
1
24
0
Fig. 1. Schematicdiagmm of the catecholenzymeelectmde. 1. Ckygeneltwmde: 2, O-ring:3, gas-permeablemembrane; 4, dialysis membrane containing immobilised polyphenol oxhiase.
diluted 1:l with O-1M sodium phosphate buffer (pH 68). Solutions were equilibrated with air, stirred during measurement and steady-state readings were obtained.
Electrode construction Solutions ofbovine albumin (3%) glutaraldehyde (3%) and polyphenol oxidase (0.1%) were prepared in O-1M sodium phosphate buffer (pH 68). A lOO-~1aliquot of each solution was mixed gently in a small tube and 20~1 of the mixture was spotted into the centre of a 35 mm X 35 mm dialysis membrane. This was placed in a Petri dish in a desiccator and stored at 4°C for 12 h. The membrane containing the immobilised enzyme was rinsed gently with deionised distilled water to remove any excess glutaraldehyde and bovine albumin. It was attached to the oxygen electrode using an O-ring, with the enzyme-coated side of the membrane facing the electrode enclosing a thin layer of O-1M sodium phosphate buffer between electrode surface and membrane (Fig. 1).When not in use, the electrode was stored in phosphate buffer at room temperature. All measurements were carried out at 25”C, using standards or samples 570
RESULTS AND DISCUSSION Electrode calibration and response cbmcteristics
The electrode was calibrated with solutions of catechol in 2 mM HCl in the concentration range 0.01-l mM. All solutions were prepared fresh daily and five assays at each concentration were performed (Fig. 2). Steady state was achieved by 80 s at substrate concentrations from 0.05 to 1 mM (Fig. 3). The pH response of the electrode is shown in Fig. 4; maximum activity was shown from pH 5.5 to 7.0, which correlates well with the pH optimum of 6.5 for the enzyme in solution. The deterioration of the response of the electrode in use is shown in Fig. 5. There was a rapid initial decline in response followed by a steady decline over 2 weeks, probably as a result of inactivation of the enzyme by polymerisation of its o-quinone product in aqueous solutions (Hall et al., 1988).
Biosensors & Bioelectronics 6 (1991)569-573
0
0.2 0.4 0.6 0.8 1.0 Catechol (mmol/litre)
Development of a catechol enzyme elect&e
1.2
“1
li l’.on’M
20
40
60
60
1
0
I, 0
Fig 2. Calibration of the catechol enzyme electrode with catechol in 2 mhf HCl m). Catechol was added to urine to give indicated final concentration - ‘spiked urine’ (+). Error bars show +2 SD.
0
8
,
,
,
2
4
6
,
,
8 10 Days
,
,
12 14
,
16
Fig. 5. Response profire of catechol enzyme electrode to I rnM catechol in phosphate buffer over 16 days.
The membrane was made in batches and stored at 4°C before use to overcome the relatively short electrode life. Response to catecholamine metabolites found in urine
100
120
140
160
160
SW"tldS
Fig 3. Response of the elecrrode to concentrations of catecholftom 00.5 to 1 mM. The response measured (AO, (ppm)) is the absolute value of the dt~etence between the response at time zero (ia, and the response at time t, i.e. ((0, ppm), - (Oz ppm),&. Sampling was at intervals of 10 s.
81
Neural crest turnouts secrete abnormal quantities of HVA, VMA, normet- and metadrenaline, noradrenaline and adrenaline, and dopamine. The expected ranges of these compounds in normal urines are: HVA < 50 RM, VMA < 50 PM, normet- and metadrenaline < 5 C(M,adrenaline and noradrenaline < 1 FM,and dopamine < 3 FM, and elevations in neural crest turnouts can be from two to 20 times these limits. The response of the electrode to these catecholamine metabolites and related catechols found in urine is shown in Table 1. There was no detectable response to HVA, VMA, normetadrenaline and metadrenaline in the range 0.05-l mM. Dopamine, noradrenaline, adrenaline, DOPA, DHPA and DHPMA all showed similar responses to catechol. Response of the electrode in urine
0479 0
I
3
t
4
5
s
7
8
9
10
Fig. 4. pHptofile of electrode response using 1 mhecatechol
The electrode response to a urine sample ‘spiked’ with catechol in the range 0.01-l mM is shown in Fig. 2. Although there is a considerable background activity in the urine sample, the response is parallel to that of the pure catechol solution. The mean catechol activity in the urine of 571
Biosensors & Bioelectronics 6 (1991) 569413
C. R Tillyer, k? T Gobin TABLE 1
Catechol electrode response (0, ppm) to ~techolamines and their major me~bolites Subsrmte 045
Catechols Dopamine Adrenaline Noradrenaline Dihydroxyphenylalanine Dihydm~heny~cetic acid Dihyd~~henylmandelic acid
7.5 74 74 I.4 7.5 7.5
7.3 7.1 7.1 7.3 7.4 7-4
60 5.7 5.8 6.2 6.3 6.3
5.2 46 49 5-O 5-5 5-8
Me~yl-substi~~ catechols Homovanillic acid Vanillylmandelic acid Metradrenaline Normetadrenaline
7.5 7.5 7.5 7.5
7-5 7.5 7-5 7.5
7.5 7.5 7.5 7.5
7-5 7-5 7-f 7.5
patients with non-neural crest tumours was 0.415 mM (SD O-149mM; n = 40). Analytical
range
limit of detection (3 SD above the mean response of a blank solution) was found to be 0.07 mM for the pure catechol calibrant. The detection limit of the Clark electrode is O*OlO0.025 ppm (eqUiV&nt to 0~~3-0~~8 mM The
catechol).
CONCLUSIONS The catechol enzyme electrode responded in a similar fashion to a broad range of catechols. This response would be expected from the known activity of polyphenol oxidase; the enzyme catalyses the oxidation of a wide range of catechols and is relatively specific for the catechol grouping. The activity towards monophenols is much less and this is reflected in the negligible response of the electrode to the methylsubstituted catechols tested here. Urinary monophenolic compounds are unlikely to interfere at physiological concentrations. The electrode responded well to three of the major catechol derivatives secreted by neural crest noradrenaline and tumours (dopamine, adrenaline), and has some potential for monitoring changes in these metabolites in urine. The use of an enzyme-coupled biosensor means 572
that the response is much more specific and therefore less likely to suffer interference from other major urinary consti~ents (such as urea and creatinine) than, for instance, ~tentiomet~c biosensors (Gobin & Tillyer, 1990).The parallel response of the spiked urine would suggest that the urine matrix has little effect on the sensitivity of the electrode response. The electrode in its present form has some drawbacks. The membrane life is relatively short in use, but unused membranes can be stored for prolonged periods at 4°C (Table 2). The limit of detection, although low (@07mM for catechol solutions) would ideally be lower for measu~ng catechols in some patients, where absolute levels and changes of the order of O*OOl-O*OlOmM would need to be detected to pick up small tumours. A more serious problem when measuring in urine is the relatively high background response, which is probably due to the broad specificity of the enzyme to the catechol grouping. Urine contains a number of catechols which are not secreted by neural crest tumours (dihyd~xybenzoic acids and dihyd~xycinnamic acid}. These will add a certain amount of physiological variation to the background and decrease the sensitivity of the electrode to tumour-derived catechols. Our results on urines of patients with non-neural crest tumours appear to confirm this. A more selective electrode may not necessarily overcome this problem. It may be possible to improve the sensitivity of
Biosensors & Bioelectronics 6 (1991) 569-573
Development of a catechol enzyme electrode
TABLE 2
Comparison of calibrationfor freshlypreparedenzymemembrane and enzyme-membrane stored at 4°C for 8 months Catechol (mM)
02 (ppm) Stored membrane Mean”
1 08 0.6 04 0.2 01
390 488 5.63 6.32 7.16 7.45
Fresh membrane
SD
Mean
SD
0.055
4.42 5.20 566 6.52 7.37 7.98
0.290 0.117 0.281 0083 0.048 0.015
a080 0.238 0.085 O-089 0.0432
Regression: Y = A + BX (Y = Oz ppm; X = catechol mM): Old membrane,4 = 7901 B = -3.890 SE = 0074 8, n = 30 Fresh membrane A = 8.027 B = -3.618 SE = 0.085 6, n = 30 (SD - standard deviation; SE-standard error of estimate). aMean of five determinations at each concentration.
the electrode. Macholln and Schdntl (1977), using an electrode of similar construction, obtained detection limits of 0~007-0~070 mM for phenol, p-cresol and pyrocatechol. Hall et al. (1988) have also described an enzyme electrode that uses polyphenol oxidase in an organic solvent to detect p-cresol. The detection limit of this electrode was low and it is possible that this type of electrode system may be more suitable for measuring very low levels of tumour catecholamines. However, the use of organic solvents seems to make the enzyme less selective and relatively more sensitive to phenols, which would increase the background further in urine samples because of the high content of phenolic compounds. REFERENCES Bohun, C. (ed.) (1966). Recent Results in Cancer Research. Vol. 2. Neuroblastoma - Biochemical Studies, Springer-Verlag, New York
Dower, R. G. H., Bailey, B. A. & Martin, R. J. (1984). Estimation of biogenic amines by HPLC and the electrochemical detector. Chromatogr. Rev., 2, 57. Gobin, P. T. & Tillyer, C. R (1990). The applications of electrochemical biosensom to the monitoring of catecholamines in patients with neural crest turnouts. Proc Biosensors ‘90 The First World Congress on Biosensors - Singapore. Elsevier Science Publishers, Oxford. Hall, G. F., Best, D. J. & Turner, A. P. F. (1988). The determination of p-cresol in chloroform with an enzyme electrode used in the organic phase. Anal. Chim. Acta 213, 113-19. Lemieux, B., Auray-Blais, C., Giguere, R & Striver, C. R (1989). Neuroblastoma screening: the Canadian experience. Med. Pediatr. Oncol, 17, 379-81. Macholln, L. & Schanel, L. (1977). Enzyme electrode with immobilised polyphenol oxidase for determination of phenolic substances. Colln Czech. Chem. Commun., 42,3667-75. Moyer, T. P., Jiang, N. & Machachek, D. (1978). Analysis of urinary and plasma catecholamines by high performance liquid chromatography with amperometric detection. In Biological/Biomedical Applications of Liquid Chromatography, Vol. 2. ed. G. L. Hank Dekker, New York, pp. 75-92. Nagatsu, T. (1973). Biochemistry of Catecholamines. The Biochemical Method. University Park Press, Baltimore, MD, pp. 209-73. Schleicher, E. D., Kee, F. K. & Wieland, 0. H. (1983). Analysis of total urinary catecholamines by liquid chromatography: methodology, routine experience and clinical interpretations of results. Clin. Chim. Acta, 129, 295-305. Stentrom, G., Sjogren, B. & Waldenstrom, J. (1983). Excretion of adrenaline, not-adrenaline, vanillylmandelic acid and metanephrines in 64 patients with phaeochromocytoma. Acta Med. Stand, 214, 145-52. Takeda, T. (1989). Japanese experience of screening. Med. Pediatr. Oncol., 17, 361-3. Turner, k P. F., Karube, I. &Wilson, G. S. (eds) (1987). Biosensors: Fundamentals and Applications. Oxford University Press, Oxford. Weinkove, C. (1987). Adrenaline, noradrenaline and related compounds. In Varleyk Practical Clinical Biochemistry, ed. k Gowenlock, J. R McMurray & D. M. McLaughlin. Heinemann, London, pp. 877-93.
573
The development of a catechol enzyme electrode and its possible use for the diagnosis and monitoring of neural crest tumours* C. R. Tillyer & P. T. Gobin Department of Chemical Pathology, Royal Marsden Hospital, Fulham Road, London SW3 6JJ, UK (Received 29 June 1990; received in revised form 12 November 1990, accepted 22 November 1990)
Ah&a& A catechol enzyme electrode is described, in which a Clark-type oxygen electrode is coupled to immobilised polyphenol oxidase in albumin cross linked with glutaraldehyde on a dialysis membrane. Electrode calibration, response time, pH response profile, stability, detection limit and selectivity are evaluated and the feasibility of using the electrode for the measurement of ~~holamines in the urine of patients with neural crest turnours is assessed. Keywords: catechol, polyphenol cytoma, neuroblastoma, urine.
oxidase, enzyme electrode, phaeochromo-
INTRODUCTION Turnouts of the neural crest (phaeochromocytoma and neuroblastoma) in humans secrete abnormal quantities of catecholamines and their metabolites (Boliun, 1966, Stentrom et al., 1983). The detection and measurement of these compounds in serum and urine of affected patients is a well-established aid to the diagnosis and treatment of these cancers (Weinkove, 1987). Current methods of estimation of these metabolites are either spectrophotometric (Nagatsu, 1973)or chromatographic (Moyer et&., 1978; Schleicher et al., 1983; Dower er al., 1984). They generally require expensive ins~men~tion and skilled staff, and are rather laborious and time-consuming resulting in a waiting time for *Paper presented at Biosensors 90, Singapore, 2-4 May 1990. 09.5~5663/91/$03.50 @ 1991 Elsevier Science Publishers Ltd.
the result for a particular patient of weeks or even months, with a con~mi~nt delay in ~atment. Also, the tumours are rather rare but often suspected and the cost of detection in health economic terms is therefore high, particularly when whole populations are screened, as in the case of neuroblastoma (Takeda, 1989; Lemieux et al., 1989).Biosensors are simple, robust, cheap and provide a rapid result; they may therefore provide a more effective method for the detection and monitoring of these metabolites than currently used methods (Turner et al., 1987). We have developed a catechol enzyme electrode using an immobilised catechol oxidase (mushroom polyphenol oxidase EC 1.14.18.1). The enzyme catalyses the conversion of catechol to o-benzoquinone in the presence of oxygen; the rate of oxygen consumption is measured by a Clark electrode and is proportional to the catechol concentration. The feasibility of using 569
C. R. Tillyer, I? T. Gobin
Biosensorsh Bioelectmnics6 (1991) 569-573
this electrode in the detection and monitoring of abnormal amounts of catecholamines in patients with neural crest turnouts is evaluated.
MATERIALS AND METHODS The oxygen electrode was an Orion model 97-08 connected to an Orion 720 electrometer (Orion Research, East Sussex, UK). For the response curves the electrometer was interfaced to a BBC Master series microcomputer programmed to take readings at intervals of 10 s using the authors’ own software. Bovine albumin; glutaraldehyde; catechol; dopamine-, adrenaline-, noradrenaline-, metadrenaline- and normetadrenaline-hydrochloride; dihydroxyphenylalanine (DOPA); dihydroxyphenylacetic acid (DHPA); dihydroxymandelic acid (DHMA); vanillylmandelic acid (VMA); homovanillic acid (I-WA); and mushroom polyphenol oxidase (tyrosinase) were purchased from Sigma (Dorset, UK). The specific activities of the enzyme were 648OOOUmg-’ as catechol oxidase, 82 000 U mg-’ as polyphenol oxidase and 3430 U mg-’ as tyrosinase. All optically active compounds were DL mixtures. The dialysis membrane was an Ezee-mount premount membrane from Elkay (Shrewsbury, MA, USA).
1
24
0
Fig. 1. Schematicdiagmm of the catecholenzymeelectmde. 1. Ckygeneltwmde: 2, O-ring:3, gas-permeablemembrane; 4, dialysis membrane containing immobilised polyphenol oxhiase.
diluted 1:l with O-1M sodium phosphate buffer (pH 68). Solutions were equilibrated with air, stirred during measurement and steady-state readings were obtained.
Electrode construction Solutions ofbovine albumin (3%) glutaraldehyde (3%) and polyphenol oxidase (0.1%) were prepared in O-1M sodium phosphate buffer (pH 68). A lOO-~1aliquot of each solution was mixed gently in a small tube and 20~1 of the mixture was spotted into the centre of a 35 mm X 35 mm dialysis membrane. This was placed in a Petri dish in a desiccator and stored at 4°C for 12 h. The membrane containing the immobilised enzyme was rinsed gently with deionised distilled water to remove any excess glutaraldehyde and bovine albumin. It was attached to the oxygen electrode using an O-ring, with the enzyme-coated side of the membrane facing the electrode enclosing a thin layer of O-1M sodium phosphate buffer between electrode surface and membrane (Fig. 1).When not in use, the electrode was stored in phosphate buffer at room temperature. All measurements were carried out at 25”C, using standards or samples 570
RESULTS AND DISCUSSION Electrode calibration and response cbmcteristics
The electrode was calibrated with solutions of catechol in 2 mM HCl in the concentration range 0.01-l mM. All solutions were prepared fresh daily and five assays at each concentration were performed (Fig. 2). Steady state was achieved by 80 s at substrate concentrations from 0.05 to 1 mM (Fig. 3). The pH response of the electrode is shown in Fig. 4; maximum activity was shown from pH 5.5 to 7.0, which correlates well with the pH optimum of 6.5 for the enzyme in solution. The deterioration of the response of the electrode in use is shown in Fig. 5. There was a rapid initial decline in response followed by a steady decline over 2 weeks, probably as a result of inactivation of the enzyme by polymerisation of its o-quinone product in aqueous solutions (Hall et al., 1988).
Biosensors & Bioelectronics 6 (1991)569-573
0
0.2 0.4 0.6 0.8 1.0 Catechol (mmol/litre)
Development of a catechol enzyme elect&e
1.2
“1
li l’.on’M
20
40
60
60
1
0
I, 0
Fig 2. Calibration of the catechol enzyme electrode with catechol in 2 mhf HCl m). Catechol was added to urine to give indicated final concentration - ‘spiked urine’ (+). Error bars show +2 SD.
0
8
,
,
,
2
4
6
,
,
8 10 Days
,
,
12 14
,
16
Fig. 5. Response profire of catechol enzyme electrode to I rnM catechol in phosphate buffer over 16 days.
The membrane was made in batches and stored at 4°C before use to overcome the relatively short electrode life. Response to catecholamine metabolites found in urine
100
120
140
160
160
SW"tldS
Fig 3. Response of the elecrrode to concentrations of catecholftom 00.5 to 1 mM. The response measured (AO, (ppm)) is the absolute value of the dt~etence between the response at time zero (ia, and the response at time t, i.e. ((0, ppm), - (Oz ppm),&. Sampling was at intervals of 10 s.
81
Neural crest turnouts secrete abnormal quantities of HVA, VMA, normet- and metadrenaline, noradrenaline and adrenaline, and dopamine. The expected ranges of these compounds in normal urines are: HVA < 50 RM, VMA < 50 PM, normet- and metadrenaline < 5 C(M,adrenaline and noradrenaline < 1 FM,and dopamine < 3 FM, and elevations in neural crest turnouts can be from two to 20 times these limits. The response of the electrode to these catecholamine metabolites and related catechols found in urine is shown in Table 1. There was no detectable response to HVA, VMA, normetadrenaline and metadrenaline in the range 0.05-l mM. Dopamine, noradrenaline, adrenaline, DOPA, DHPA and DHPMA all showed similar responses to catechol. Response of the electrode in urine
0479 0
I
3
t
4
5
s
7
8
9
10
Fig. 4. pHptofile of electrode response using 1 mhecatechol
The electrode response to a urine sample ‘spiked’ with catechol in the range 0.01-l mM is shown in Fig. 2. Although there is a considerable background activity in the urine sample, the response is parallel to that of the pure catechol solution. The mean catechol activity in the urine of 571
Biosensors & Bioelectronics 6 (1991) 569413
C. R Tillyer, k? T Gobin TABLE 1
Catechol electrode response (0, ppm) to ~techolamines and their major me~bolites Subsrmte 045
Catechols Dopamine Adrenaline Noradrenaline Dihydroxyphenylalanine Dihydm~heny~cetic acid Dihyd~~henylmandelic acid
7.5 74 74 I.4 7.5 7.5
7.3 7.1 7.1 7.3 7.4 7-4
60 5.7 5.8 6.2 6.3 6.3
5.2 46 49 5-O 5-5 5-8
Me~yl-substi~~ catechols Homovanillic acid Vanillylmandelic acid Metradrenaline Normetadrenaline
7.5 7.5 7.5 7.5
7-5 7.5 7-5 7.5
7.5 7.5 7.5 7.5
7-5 7-5 7-f 7.5
patients with non-neural crest tumours was 0.415 mM (SD O-149mM; n = 40). Analytical
range
limit of detection (3 SD above the mean response of a blank solution) was found to be 0.07 mM for the pure catechol calibrant. The detection limit of the Clark electrode is O*OlO0.025 ppm (eqUiV&nt to 0~~3-0~~8 mM The
catechol).
CONCLUSIONS The catechol enzyme electrode responded in a similar fashion to a broad range of catechols. This response would be expected from the known activity of polyphenol oxidase; the enzyme catalyses the oxidation of a wide range of catechols and is relatively specific for the catechol grouping. The activity towards monophenols is much less and this is reflected in the negligible response of the electrode to the methylsubstituted catechols tested here. Urinary monophenolic compounds are unlikely to interfere at physiological concentrations. The electrode responded well to three of the major catechol derivatives secreted by neural crest noradrenaline and tumours (dopamine, adrenaline), and has some potential for monitoring changes in these metabolites in urine. The use of an enzyme-coupled biosensor means 572
that the response is much more specific and therefore less likely to suffer interference from other major urinary consti~ents (such as urea and creatinine) than, for instance, ~tentiomet~c biosensors (Gobin & Tillyer, 1990).The parallel response of the spiked urine would suggest that the urine matrix has little effect on the sensitivity of the electrode response. The electrode in its present form has some drawbacks. The membrane life is relatively short in use, but unused membranes can be stored for prolonged periods at 4°C (Table 2). The limit of detection, although low (@07mM for catechol solutions) would ideally be lower for measu~ng catechols in some patients, where absolute levels and changes of the order of O*OOl-O*OlOmM would need to be detected to pick up small tumours. A more serious problem when measuring in urine is the relatively high background response, which is probably due to the broad specificity of the enzyme to the catechol grouping. Urine contains a number of catechols which are not secreted by neural crest tumours (dihyd~xybenzoic acids and dihyd~xycinnamic acid}. These will add a certain amount of physiological variation to the background and decrease the sensitivity of the electrode to tumour-derived catechols. Our results on urines of patients with non-neural crest tumours appear to confirm this. A more selective electrode may not necessarily overcome this problem. It may be possible to improve the sensitivity of
Biosensors & Bioelectronics 6 (1991) 569-573
Development of a catechol enzyme electrode
TABLE 2
Comparison of calibrationfor freshlypreparedenzymemembrane and enzyme-membrane stored at 4°C for 8 months Catechol (mM)
02 (ppm) Stored membrane Mean”
1 08 0.6 04 0.2 01
390 488 5.63 6.32 7.16 7.45
Fresh membrane
SD
Mean
SD
0.055
4.42 5.20 566 6.52 7.37 7.98
0.290 0.117 0.281 0083 0.048 0.015
a080 0.238 0.085 O-089 0.0432
Regression: Y = A + BX (Y = Oz ppm; X = catechol mM): Old membrane,4 = 7901 B = -3.890 SE = 0074 8, n = 30 Fresh membrane A = 8.027 B = -3.618 SE = 0.085 6, n = 30 (SD - standard deviation; SE-standard error of estimate). aMean of five determinations at each concentration.
the electrode. Macholln and Schdntl (1977), using an electrode of similar construction, obtained detection limits of 0~007-0~070 mM for phenol, p-cresol and pyrocatechol. Hall et al. (1988) have also described an enzyme electrode that uses polyphenol oxidase in an organic solvent to detect p-cresol. The detection limit of this electrode was low and it is possible that this type of electrode system may be more suitable for measuring very low levels of tumour catecholamines. However, the use of organic solvents seems to make the enzyme less selective and relatively more sensitive to phenols, which would increase the background further in urine samples because of the high content of phenolic compounds. REFERENCES Bohun, C. (ed.) (1966). Recent Results in Cancer Research. Vol. 2. Neuroblastoma - Biochemical Studies, Springer-Verlag, New York
Dower, R. G. H., Bailey, B. A. & Martin, R. J. (1984). Estimation of biogenic amines by HPLC and the electrochemical detector. Chromatogr. Rev., 2, 57. Gobin, P. T. & Tillyer, C. R (1990). The applications of electrochemical biosensom to the monitoring of catecholamines in patients with neural crest turnouts. Proc Biosensors ‘90 The First World Congress on Biosensors - Singapore. Elsevier Science Publishers, Oxford. Hall, G. F., Best, D. J. & Turner, A. P. F. (1988). The determination of p-cresol in chloroform with an enzyme electrode used in the organic phase. Anal. Chim. Acta 213, 113-19. Lemieux, B., Auray-Blais, C., Giguere, R & Striver, C. R (1989). Neuroblastoma screening: the Canadian experience. Med. Pediatr. Oncol, 17, 379-81. Macholln, L. & Schanel, L. (1977). Enzyme electrode with immobilised polyphenol oxidase for determination of phenolic substances. Colln Czech. Chem. Commun., 42,3667-75. Moyer, T. P., Jiang, N. & Machachek, D. (1978). Analysis of urinary and plasma catecholamines by high performance liquid chromatography with amperometric detection. In Biological/Biomedical Applications of Liquid Chromatography, Vol. 2. ed. G. L. Hank Dekker, New York, pp. 75-92. Nagatsu, T. (1973). Biochemistry of Catecholamines. The Biochemical Method. University Park Press, Baltimore, MD, pp. 209-73. Schleicher, E. D., Kee, F. K. & Wieland, 0. H. (1983). Analysis of total urinary catecholamines by liquid chromatography: methodology, routine experience and clinical interpretations of results. Clin. Chim. Acta, 129, 295-305. Stentrom, G., Sjogren, B. & Waldenstrom, J. (1983). Excretion of adrenaline, not-adrenaline, vanillylmandelic acid and metanephrines in 64 patients with phaeochromocytoma. Acta Med. Stand, 214, 145-52. Takeda, T. (1989). Japanese experience of screening. Med. Pediatr. Oncol., 17, 361-3. Turner, k P. F., Karube, I. &Wilson, G. S. (eds) (1987). Biosensors: Fundamentals and Applications. Oxford University Press, Oxford. Weinkove, C. (1987). Adrenaline, noradrenaline and related compounds. In Varleyk Practical Clinical Biochemistry, ed. k Gowenlock, J. R McMurray & D. M. McLaughlin. Heinemann, London, pp. 877-93.
573
