245-A highly sensitive glucose electrode using glucose oxidase collagen film

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Industrially reconstituted collagen films have shown excellent prop erties for ~-D-glucose oxidase couplir g. Associated with a platinum anode for amperometric detection of hydrogen peroxide, these er zymatic films form a very simple and easy to handle glucose electrode; this device presents a very high sersitivity (ca. IO-' M) givirg resporses proportional to glucose concentration over 5 orders of magr.itude,

Inlroduction
The association of an erzyme in a soluble form with an electro chemical ser sor was first reported by CLARK and LYONS, in I9621 Since this date, the obtainment of carrier-bound enzymes has permitted the design of er zyme electrodes. In most cases, enzymes are trapped in gels surroundir.g the sensor-leading to systems which are difficult to handle.
In the present work, an er zymatic membrane was prepared from reconstituted calf skin collagen after acyl-azide activation and a couplirg process giving a surface bir-dir g of er zymes of different classes. '.' A glucose electrode usir g ~-D-glucose oxidase (GOD) membrar es was developed, usir g an amperometric method with a platir um electrode to detect hydrogen peroxide, which is a product of the er zymatic oxida tion of glucose according to the reaction: The potential was fixed at +650 mV vs AgIAgCl, KCl 0,1 M and the anodic current was recorded. A stationary response was then obtained allowing the measurement of very low glucose concentrations. In a more sophisticated device, a second electrode involving a non-enzymatic membrane was used to compensate for the detection of other electroactive molecules and to enhance the selectivity of the glucose electrode.','

Instrumentation
The glucose electrode consisted of a modified gas electrode in which the pH detector was replaced by a platinum disk and the usual teflon film by a collagen membrane. In a differential device, electrode I was mounted with a ~-D-glucose oxidase collagen membrane and electrode 2 with a nori-er-zymatic one.
Electronics were made by SOLEA TAcuSSEL: current outputs of both working electrodes were first subtracted (Deltapol) and then twice differenciated (Derivol) with a time-base of one second (GSTP) ; thus different current vs time curves were avaliable and usually recorded after a glucose pulse (SOLEA T.~CUSSEL EPL 2 with TV II GD plug-in unit, and three traces LINEAR 395 recorders).
Unless otherwise mentioned, the temperature of the solutions were carefully thermostated to 30.0 ±O.I oC (COLORA cryothermostat WK5 DS).

Solutions and reagents
Insoluble films of highly polymerized reconstituted collagen (20 em wide) were a gift of the Centre Technique du Cuir, Lyon (France); their thickness is about 0.1 mrn in a dry state and 0.3-0.5 mm when swollen. They do not need to be tanned and can be stored several years without damage. a-s Unless otherwise mentioned, all chemicals were reagent grade. The stock solutions of 0.1 M glucose were allowed to mutarotate at room temperature at least 3 hours before using and were stored at 4°C. Both electrodes were filled with and dipped into 0.2 M acetate buffer, 0.1 M KCl solutions, pH 5.6.

Glucose oxidase binding on collagen membranes
The mild general acyl azide procedure for collagen membranes acti vation was useds followed by the enzyme coupling.
Carboxyls were first esterified by immersion of crude membranes in a methanoljo.z M hydrochloric acid solution for at least 72 hours, then treated overnight by I % hydrazine and soaked at 4°C for 3 mi nutes in 0.5 M NaNO,-0.3 M HCI mixture just before coupling. Thor ough washings were performed between each step and at the end of the activation process avoiding contact between reagents and enzyme so lutions.

Proced..re
Both electrodes were allowed to equilibrate in the buffer solution for IS to 30 minutes after stepping the potential of the platinum disks to +650 mV vs AgIAgCI, 0.1 M RCI. This potential corresponds to a diffusion-limited current for H 2 0 , oxidation. Calibrations were per formed by successive micro additions (10 to 50 mm") of stock solutions of 10-6 to 10-'111 glucose to 10 to 20 ern" of buffer. The stationary re sponse was the variation of the steady state values of 1,-1, when a sample of a glucose containing solution was added and the dynamic response was the height of the peak of the first derivative d(I,-I,) Jdt. When successive additions were performed in the same solution, a current offset was used.

Results and discussion
'When the enzyme electrode is immersed in a medium in which a pulse of glucose is added a steady state takes place after 2-3 minutes, as shown on Fig. I, and the value of the anodic current reaches a pla teau. The variation of intensity is directly dependent on the glucose concentration in the assay: this is the stationary response. On the other hand, the dynamic response is measured by the height of the peak ob tained after 30-50 s by recording the first derivative of this current.
The lowest glucose concentration detected under these conditions is less than 10-8 M (Fig. 2). For values higher than ro-' M, the responses become independent of the glucose concentration. The concentrations, which can be determined, range between ca. 10-6 and 10-2 M i.e. over 6 orders of magnitude and the linearity of the calibration curve is obtained between (3-5) X 10-8 and (3-5) X10-' M i.e. over 5 orders of magnitude ( Fig. 2).
In a typical experiment, the calibration curves remained linear even after 40 hours operation at 30°C and 250 days storage at 4°C, allowing accurate repeated determinations for 160 micro-assays tested. However, a daily calibration was necessary, because of the slight de  [Glucose] (M) 1~-6 1~-5 10'-4' 10'"-3 for 13 successive additior s of IO-' M glucose. the standard deviation from the mean is usually lower than 2 %.
Glucose oxidase itself is very selective for I3-D-glucose; thus en zymatic electrode I presents a high selectivity for glucose compared with usual sugars: selectivity ratios are higher than 2 000 / I for fructose, lactose and sucrose. As other species may diffuse through collagen mem branes and may be oxidized on platinum at +0.94 V (N.H.E.), the use of a compensating non enzymatic electrode 2 eliminates possible inter ferences of species such as ascorbate, urate, tyrosine or hydrogen peroxide; the selectivity ratio for hydrogen peroxide ranges between 80 /1 and 250 /1 depending upon experimental parameters (accuracy of the balance of both electrodes and GOD activity of the film).
The use of nou-enzymati« electrode 2 is specifically of greatest interest for blood glycemia determinatiors, I, usually reachirg ro-yo % of 1,-1,. Fig. 3 presents the typical analysis of blood plasma samples after induced glycemia: in this case, a glucose calibration is necessary for each set of 3-4 blood additions. The dotted line, representing the second derivative of the current 1,-1" has been successfully used for monitoring a printing device of the peak of the first derivative, i.e. the dynamic response." The glucose electrode may be used in a large temperature range, from IS to 40 DC. As both responses are very sensitive to temperature (about 4-5 % / DC at 30 dc) it is necessary to carefully thermostate the solutions in contact with both electrodes.

Conclusions
Industrially reconstituted collageu films were found suitable for glucose oxidase immobilization ; the enzymatic activity was maintained and its stability enhanced. A membrane loading of 50 to 100 mU per membrane (I em diameter) was sufficient to obtain a very sensitive glu cose electrode when associated with amperometric hydrogen peroxide detection. Stationary and dynamic responses of this glucose electrode was proportionnal to glucose concentration from 3 -5 X 10-8 to 3 -5 X 10-' M. Furthermore, this sensor was used for whole blood samples analysis in induced glycemia.