Polymers in Sensors - American Chemical Society

Chapter 6

Micro, Planar-Form Lactate Biosensor for Biomedical Applications 1,4

1,5

1

2

Sayed A. Marzouk , VasileV.Cosofret , RichardP.Buck , Hua Yang , WayneE.Cascio , and Saad S. M. Hassan 2

3

1

Downloaded by FUDAN UNIV on April 22, 2017 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0690.ch006

Department of Chemistry, University of North Carolina, Chapel Hill,NC27599-3290 Department of Medicine, University of North Carolina, Chapel Hill,NC27599-7075 Department of Chemistry, Ain Shams University, Cairo, Egypt 2

3

Lactic acid is an important metabolic product formed during anaerobic glycolysis. There are no continuous methods to measure the extracellular lactate accumulation occurring in the absence of myocardial perfusion. Two forms of biosensors have been constructed. Thefirstminiature version fulfills the operational requirements, such as flexibility, wide linear response range, high selectivity, and fast response time. This biosensor is based on immobilized lactate oxidase with anodic detection of the produced H O . An inner layer, which allows selective detection of hydrogen peroxide, is formed by electropolymerization of m-phenylenediamine. A diffusional barrier of polyurethane greatly increases the linear response range. The second miniature sensor uses an anode made of tetrathiofulvalene– tetracyanoquinodimethane (TTF-TCNQ) charge transfer complex. The former sensor was used successfully under reduced ambient oxygen tension(PO >24mm Hg); while the second can operate in the absence of oxygen. The applications are measurements of lactate in the ischemic rabbit papillary muscle under no-flow conditions. 2

2

2

Connection of Lactate Generation with Ischemic Events Deprivation of molecular 0 fromrespiring cells shifts the production of high energy phosphatesfromoxidative phosphorylation to glycolysis. An end product of anaerobic glycolysis is lactic acid (7,2). This is the reason that blood lactate level is an important metabolic indicator for presence and extent of anaerobic glycolysis. The latter occurs normally during vigorous exercise as the energy demands of skeletal muscle exceed the 2

4

Current address: Department of Chemistry, Ain Shams University, Cairo, Egypt Current address: Instrumentation Laboratories, Lexington, MA, 02713 5

66

©1998 American Chemical Society Akmal and Usmani; Polymers in Sensors ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

67 supply of 0 substrate. Anaerobic glycolysis contributes to the pathophysiology of hypoxia, anoxia or ischemia. Under these conditions excessive production and accumulation of intracellular H , in part from lactic acid production, results in a decrease of intracellular pH where the fall in intracellular acidosis contributes to reduced myocardial contractility and altered impulse propagation. 2


+

Because of the pH dependence of many cellular enzyme systems, the cell has evolved numerous ion channels, pumps, exchangers and transporters to maintain an optimal intracellular pH. Transsarcolemmal transport of lactate-anion and diffusion of lactic acid are believed to be important mechanisms regulating intracellular pH under these conditions. However, the interaction of these various cellular mechanisms for the regulation of intracellular pH under metabolic stress is not completely understood. A complete understanding of these processes is lacking, in part, because an experimental method to measure extracellular lactate, continuously and directly, with sufficient temporal and spatial resolution under no-flow conditions did not exist, prior to our work. Biosensors as a Reliable Tool. The previously described miniature version of a conventional lactate sensor was applied to the measurement of lactate accumulation in the papillary muscle of a rabbit heart suffering hypoxia, a mild ischemic event with low, but non-zero oxygen pressure (5). Three configurations of Au-based amperometric electrodes were fabricated by a photolithographic technique on flexible polyimide (Kapton) foils. All sensors were based on an immobilized lactate oxidase with detection of the hydrogen peroxide produced and oxidized at a platinum-electroplated A u base electrode at 0.5V vs Ag/AgCl. A n inner electropolymeric layer was used to reject interferences by easily oxidized molecules. A diffusion controlling outer layer of cast polyurethane film extended the response range as required by the generation of lactate during ischemic conditions. These sensors have high selectivity, good operational stability, good accuracy and precision (average recovery = 102.3 ± 0.4% for control sera), fast response time (I95 = 20 s) and high upper limit of the linear dynamic range (25-80 mM, with sensitivity of 1.7-0.4 nA/mM at PO2 = 15 mmHg). Fig. 1 shows the cross sectional construction of the anode. Subsequently, fabrication, characterization and application of a miniaturized amperometric lactate biosensor was described that supplements or may replace earlier designs. It is third generation type, based on crosslinked lactate oxidase and tetratWafulvalene-tetracyanoquinodhnethane (TTF-TCNQ) charge transfer complex. The sensor was developed for continuous quantitative monitoring of lactate accumulation in ischemic myocardium under severe depletion of oxygen. The sensor was evaluated in vitro at an applied potential of 0.15 V vs Ag/AgCl; it proved to combine all the performance characteristics desired for the present application, such as proper response in absence of oxygen, good operational stability, good accuracy and precision (103.5 ± 1 . 2 %), adequate response time ( ^ 3 % = 80 s), and wide linear dynamic range up to 27 m M (r=0.9998) in N -saturated solutions and at 37 °C. The prepared sensors (n = 12) showed sensitivity of 380 ± 94 nA/mM, and a background current of240 ± 50 nA. The lower limit of detection is 0.4 ± 0 . 1 5 mM with S/N ratio equal to 3. Results obtained for direct lactate monitoring in ischemic rabbit papillary 2

Akmal and Usmani; Polymers in Sensors ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

68

muscle under no-flow conditions and P0 < 6 mm Hg are presented below. The detailed information of sensor construction are given in a submitted paper (4). Fig. 2 shows the structure of the sensor. 7

Comparison of Lactate Biosensors

Generally, oxidase-based biosensors with amperometric signals generated by either hydrogen peroxide oxidation or oxygen reduction, require molecular oxygen as an electronic mediator. Therefore, the sensor response, for a given substrate concentration, is independent of oxygen tension (P0 ) in the sample, but only, within a certain range. Lower P0 values than this range lead to a substantial decrease in the sensor response and its linear dynamic range, as well. A well known strategy to sustain the sensor's linear dynamic range under low PO^ values is to, deliberately, lower the sensor sensitivity by controlling the diffusion properties of the outer diffusion layer during the sensor construction (5). This is the reason that in our previous effort (5) to monitor lactate under ischemic conditions, the PO2 values were maintained above 2530 mm Hg. Although, this range provided an acceptable compromise between sufficient linearity and reasonable sensitivity, these sensors were limited by low signal to noise ratios especially during in vivo measurements. Moreover, the effect of a small but significant 0 tension may have affected the production of lactate. Since anaerobic conditions are the stimulus to produce lactate in ischemic myocardium, it was necessary and important to develop a more relevant lactate biosensor that can function properly under severe oxygen deprivation. Fabrication, characterization and application of a miniaturized amperometric lactate biosensor are described. The sensor is third generation type, based on crosslinked lactate oxidase and tetratMafulvalene-tetracyanoquinodimethane (TTFTCNQ) charge transfer complex. The sensor was developed for continuous quantitative monitoring of lactate accumulation in ischemic myocardium under severe depletion of oxygen. The orignal sensors (3) were madeflexibleto adhere, or wraparound the papillary muscle of a rabbit under no-flow, and reduced oxygen environment. This condition of mild ischemia was necessary because the sensors required oxygen as the redox agent or co-factor equivalent to a co-enzyme. Thus the response of the sensor to ambient oxygen was carefully measured. Some calibrations of lactate at decreasing oxygen tensions are illustrated in Fig. 3. Clearly, true ischemia cannot be studied without sensors that operate in the absence of oxygen. Thus the third generation electrodes was constructed, tested, and shown to operate at the lowest oxygen level we could achieve in the experimental chamber containing the perfused, beating rabbit papillary muscle preparation. Second and third generations of amperometric biosensors overcome the major problem of oxygen limitations associated with the first generation type. Second generation based on non physiological organic molecules, most commonly ferrocene derivatives and tetrathiafulvalene confined to the sensing layer, which acted as redox mediators and replaced the natural oxygen receptor. The major drawback of the mediated biosensor is the leaching of the redox mediatorfromthe vicinity of the electrode surface. Several approaches were introduced to minimize this effect, such as employing electropolymericfilm,electrostatic immobilization in an anionic polymer, 2


2

2


HD I

Kapton

H

Polyimide

S

Chromium

0

Gold

Ξ

Inner layer

Η

Enzyme layer

•

Platinum H

Outer layer

(1) Layer structure of thefirstgeneration type lactate biosensor. Downloaded by FUDAN UNIV on April 22, 2017 | http://pubs.acs.org Publication Date: April 17, 1998 | doi: 10.1021/bk-1998-0690.ch006

Teflon cap Cavity 1x2.5x1.5 mm

Polycarbonate membrane I Carbon block Bonding -pad

Sputtered gold layer

-Η

Κ

1.5 mm

Conducting line Kapton flexible substrate

(2) Basic structure of the TTF-TCNQ based lactate biosensor.

400-

300-

-•-5.1 - • - 5.1 -·-1.7 -O-17 -A-0.4 -Δ-0.4

nA/mM, P02 = 150 mmHg nA/mM, P02 = 15 mmHg nA/mM, P02 = 150 mmHg nA/mM, P02 = 15 mmHg nA/mM. P02 = 150 mmHg nA/mM, P02 = 15 mmHg

200-

5

100

40

60

Lactate cone, mM

(3) Effect of the sensor sensitivity, in the linear range, on oxygen tension. Akmal and Usmani; Polymers in Sensors ACS Symposium Series; American Chemical Society: Washington, DC, 1998.


70

crosslinking the redox mediator, attaching of the redox mediator to a polymer, and modification the enzyme itself with mediator molecules. All these modifications impose more complicated steps to the sensor fabrication. Third generation of amperometric biosensors utilizes organic charge transfer complexes (OCTCs or promoters), such as tetrathiafulvalenetetracyanoquinodimethane (TTF-TCNQ), as an electrode material, instead of conventional Platinum or carbon electrodes. In this case, a direct electron transfer from the enzyme prosthetic group to the electrode surface is achieved without need for redox mediators. OCTCs have been coupled successfully with the three main classes of redox enzymes, Le., FAD-dependent oxidases, PQQ-dependent dehydrogenases, and NAD-dependent dehydrogenases and with a reductase as well. A literature search shows there are no reliable reports available till now for developing a robust lactate biosensor based on the third generation type. This is the reason that the objective of this paper is to exploit the unique properties of OCTC, taking TTF-TCNQ as an example, to develop a stable, reliable, lactate biosensor of wide linear dynamic range (~ 20-30 mM), reasonably fast responsetimeand with geometry suitable for the experimental setup used for monitoring extracellular lactate accumulation in ischemic rabbit papillary muscle under no-flow ischemia and severe lowering of oxygen tension. The success of the developed sensors in overcoming the limitations encountered in the previous work is discussed. The content emphasizes the development and application of the TTF-TCNQ-based lactate sensors. Experimental Section - Sensor Fabrication Materials, Reagents and Apparatus are comprehensively covered in (3,4). Sensor Fabrication. A cavity of dimensions 1.0 mm (width) χ 2.5 mm (length) χ 1.5 mm (depth) was drilled in a carbon block of dimensions 1.5 mm χ 3.0 mm χ 3.0 mm cutfromultrapure carbon rods, 6 mm diameter (Ultra Carbon Corp., Bay City, MI). The carbon block was glued to a sputtered gold electrodes on aflexiblepolyimide Kapton® foil using silver loaded epoxy (Epoxy Technology, Billerica, MA), see Fig. 2. TTF-TCNQ crystals were well packed into the cavity using a small stainless spatula, The surface was smoothed against a weighing paper. Three aliquots, 1.5 μΐ each, offreshlyprepared enzyme solution of composition: 6% LOx-3% BSA-0.3% GA were added on to the electrode surface. Each aliquot was allowed to dry for 10 minutes before depositing the next one. A polycarbonate membrane (0.015 μιη pore size, Nucleopore, Corning Costar, Livermore, CA) was used as an outer diffusion layer and to protect the contents of the cavity, as well. A Teflon cap (2 mm thick) was used tofixthe polycarbonate membrane. The cap was recessed ~0.5 mm to allow good contact between the sensor tip and the muscle. Silicone rubber coating (Dow Corning, Midland, MI) was used to isolate the electrical connections. The basic sensor construction is shown in Fig.2. Lactate Measurement in Control Serum. The human based control sera level Π (Dade Μοηί-Trol· ES Chemistry Control, Baxter Diagnostics, Inc.) were constituted from the lyophilized preparation according to the manufacturer's recommendation. The


71 manufacturer's mean value of lactate concentration is 4.9 mM. A 3 ml PB aliquot was placed in the cell and bubbled with nitrogen for 10 minutes. After background current stabilization, a 1.0 ml aliquot of the serum sample without any treatment, was pipetted into the cell. When the current reached a steady value, the standard addition procedure was used to evaluate the lactate concentration level. The obtained results were compared with the results obtained with the established Y S I 2700 SELECT Biochemistry analyzer equipped with YSI 2329 lactate membrane. Perfusion of Isolated Papillary Muscle.

New Zealand white rabbits (n = 3) of

either gender weighing 2 to 3 kg were heparinized (200 U/kg, iv) and anesthetized with sodium thiopental (50 mg/kg, iv) in accordance with accepted guidelines for the care and treatment of experimental animals at the University of North Carolina at Chapel Hill. The heart was excised in Tyrode's solution (in mmol/L: N a , 149; K , 4.5; M g , 0.49; C a , 1.8; CI", 133; HCO3-, 25; HPO4 -, 0.4; glucose, 20) for 20 seconds. The atria and left ventricle (LV) free wall were removed. The L V septal surface of the tissue was pinned to a wax plate in contact with Ag/AgCl ground electrode. The septal artery was visualized with a dissection microscope, cannulated and perfused with a perfusate composed of Tyrode's solution plus insulin (1U/L), heparin (400 U/L), albumin (2 g/L), and dextran (Mr 70,000; 40 g/L). The time elapsed between excising the heart and restarting perfusion was < 5 minutes in each experiment. The cannula was fixed to the septal artery with a purse string suture and the septum was secured in a custom made experimental chamber. The non-perfiised portion of the right ventricle (RV) was excised. One of the underlying papillary muscles was attached by its tendon to a piezosensitve element with a fine silk suture. The perfusate was pumped by a peristaltic pump (Digi-Staltic, Masterflex, Barrington, IL) through a custom made membrane gas exchanger in which partial pressures of 0 , N , and CO2 of the perfusate were controlled. The chamber was closed and the preparation was surrounded by humidified gas with the same composition as the perfusate. The pH of the perfusate was continuously monitored by a pH glass electrode positioned in the perfusion line before entering the septal artery. The amount of CO2 was adjusted in the gas exchanger to yield a pH of 7.35 ±0.05. The temperature of the papillary muscle was maintained between 36 and 37.5 °C by passing the perfusate through a thermostated water bath before entering the cannula to the septal artery. The ^02> PC02> HCO3" concentration and pH of the perfusate entering the cannula were confirmed by blood gas analysis (System 1304 pH/Blood Gas Analyzer, Instrumentation Laboratory, Lexington, MA). Intraarterial pressure was measured with a pressure transducer (Millar, Houston, T X ) and continuously monitored on a strip chart recorder. The septal artery pressure ranged between 35 and 50 mmHg and was maintained by adjustment of the perfusion flow rate (approximately 1.0 to 1.5 ml m k r g" tissue).


+

2 +

+

2 +

2

2

2

1

1

Measurement of Extracellular Lactate. The sensor was positioned in direct contact with the suspended muscle as shown in Fig. 4. A separate platinum wire, serving as a counter electrode, and the miniature Ag/AgCl reference electrode were positioned at the base of the papillary muscle on the septal surface. A 0.7 Hz-5 pole Butterworth low pass filter was used to eliminate the noise resulting from the electrical pulses used for


72 the muscle stimulation. While the muscle was perfused the sensor was polarized at 0.15 V and the background current was allowed to reach a stable value. Ischemia was induced by arrestingflowand decreasing the oxygen tension inside the chamber to 3-6 mm Hg. The volumefractionsof 0 and C0 in the recording chamber were measured using the gas analyzer. The magnitude of lactate accumulation was monitored by recording the change of the sensor output against the time after arrest of flow.


2

2

To convert the current response into concentration, the sensor was calibrated in situ, at the end of the experiment, by perfusing the muscle with 7.5 and 15 mM lactate standards prepared in the perfusate solution. The sensor response to these standards underflowconditions was used to construct a two-point in situ calibration curve. Ideally, the sensor should be calibrated under the same experimental conditions, i.e., no-flow conditions. However, such a calibration will not be reliable because under noflow conditions the muscle will generate lactate. In this way the standard value will be completely uncertain. For this reason the sensor calibration was performed under flow conditions. We believe that such in situ calibration is more accurate than the in vitro calibration approach which is accepted when a calibration under identical condition is not possible, e.g., with in vivo measurements. Results and Discussion Sensor Design. In terms of geometry, the planar sensors used in the previous work (3) offered an advantage of having the sensing part on one side of the sensor rather than on the bottom as in the conventional electrodes design. This is of especial importance in positioning the sensor through a small window in the top of the experimental chamber, and allowing simultaneous top view monitoring of the preparation along with the sensors using a camera mounted above the chamber window. To keep the desirable configuration andflexibility,the sputtered gold electrodes designed for the previous work (5) realized onflexibleKapton substrate were adapted in the present work. Simple deposition of TTF-TCNQ on the gold electrode surface using the drop coating technique is not feasible because of the poor adhesion to the electrode surface. More important, use of outer diffusion layer, e.g., polycarbonate membranes on this planar thin substrate becomes impossible. To promote a more convenient geometry for the base electrode, a carbon block with self contained cavity was glued to the electrode surface using silver epoxy. The dimensions of the sensing tip (1.5 mm χ 3.0 mm) was designed tofitthe dimensions of the rabbit papillary muscles, about 1.2 mm in diameter and 4.0 mm length, studied in the present work as shown in Fig. 4. Sensor Fabrication and Characteristics. In most of the previous work, conducting salts were mixed with a polymer binder such as poly (vinyl chloride) or more commonly, an oil to enhance the consistency of the electrode surface This treatment is not appropriate in the present case because the sensor will be in direct contact with a beating muscle. An outer protecting membrane that provides a diffusion barrier as well for lactate is essential for the sensor construction. Microporous polycarbonate (PC) membranes were used successfully for these purposes. The PC membrane, acting as a


73 diffusion barrier, enhances the linear dynamic range by lowering the lactate flux that reaches the enzyme layer.


Sensor Stability. The sensors maintained almost 90% of the initial activity after three weeks, stored at 4 °C when not in use. The short term operational stability of the sensor was assessed by polarizing the sensor at 37 °C in a stirred PB solution for more than 6 hours. No observed change in sensitivity was observed after this treatment. Both storage and operational stability of the sensor are excellent for the present purpose since the entire experiment time is usually in the range of 2 hours The enhanced stability is believed to be due to the combined immobilization, i.e., adsorption and glutaraldehyde crosslinking with BSA. Effect of Applied Potential on the Sensor Response. Generally, TTF-TCNQ electrodes show a potential stability window from -K).5 to -0.2 V vs. SCE (6). Therefore, the applied potential must not exceed these limits. The useful operating potential ranges used for sensors were limited to a narrower range, to achieve a compromise between sensor sensitivity, selectivity, and stability (7). The effect of the applied potential on the sensor response is shown in Fig. 5. Generally, the linear range, sensitivity, and background current increased with increasing the applied potential. Applied potential in the range of 0.1 to 0.15 V offer a good compromise between low background current, high linear dynamic range, selectivity and good stability. Therefore, an applied potential of 0.15 V was used in all subsequent measurements. A similar operating potential range was reported for TTF- TCNQ based glucose sensor

Polymers in Sensors - American Chemical Society

Recommend Documents