Analytrca Chunrca Acta,
`
`265 (1992) 5-14 Elsewer Science Pubhshers B V , Amsterdam Amperometric needle-type glucose sensor based on a modified platinum electrode with diminished response to interfering materials Chlen-Yuan Chen
`
`Research Centre for Advanced Science and Technology, Unrversrty of Tokyo, 4-6-1, Komaba, Meguro-Ku, Tokyo (Japan),
`and Department of Agncultural Chemcstry, Natwnal Tarwan lJmverst@, I, Set 4, Roosevelt Road Tatpet (Taiwan)
`
`NOK Cotporatwn, 4-3-l TsuJldo-Shmmachr, FuJtsawa-Shr, Kanugawa-Kin (Japan)
`
`Second Department of Internal Meduxne, Chrba Umversrty School of Medrcute, I-8-1,
`
`Inohuna, Chrba (Japan)
`
`Department of Agncultural Chemutry, Natwnal Tarwan Unrverscty, 1, Set 4, Roosevelt Road, Taipei (Taiwan)
`
`Research Centre for Advanced Scrence and Technology, Unrversrty of Tokyo, 4-6-1, Komaba, Meguro-ky Tokyo (Japan)
`
`(Received 31st July 1991, revised manuscript received 7th February 1992) Abstract A needle-type glucose sensor that can be used to determine glucose in serum and whole blood samples was developed Platmum wre was used as the workmg electrode and a disposable hypodernuc stamless-steel needle electroplated with platinum was used as the counter and reference electrode A method mvolvmg both photocross- linking of PVA-SbQ and cross-Mung with glutaraldehyde was used to unmobdize the enzyme [PVA-SbQ IS a poly(vmyl alcohol) bearing stdbazohum groups] Nafion and cellulose triacetate membranes were used to prevent inaccuracy from mterfenng materials and to increase the dynanuc range of the sensor, respectwely The response, reproducdxhty and long-term stablhty of the sensor and the effects of temperature, pH and metal Ions on the response were investigated Qwmg to the effective method for enzyme unmolxhzat~on, the large surface area of the counter electrode and the relative mactwity of the counter electrode to chemical reactions, the sensor showed good response, statuhty and reproducMlty The sensor did not respond to ascorbate and urate at the concentrations normally found m blood Data obtained from the sensor for glucose m serum and whole blood samples showed a good correlation (r > 0 95) with a chmcal laboratory automated analyser
`
`Keywords
`
`Correspondence to
`
`Cluen-Yuan Chen, Department of Agncultural Chenustry, National Taiwan Umversity, 1, Set 4, Roosevelt Road, Taipei (Taiwan) 0003-2670/92/$05 00 0 1992 - Elsevler Science Pubhshers B V All rights reserved
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 1 of 10
`
`Masao Gotoh
`Hldelchl Malone
`Yuan-Chl Su
`Euchl Tamlya and Isao Karube
`Amperometry, Blosensors, Enzymatic methods, Blood, Glucose, Serum
`

`

`6
`
`EXPERIMENTAL
`
`C -Y Chen et al /Anal Chum Acta 265 (1992) 5-14 Blosensors for glucose have been studied for a long time and many kmds of materials, mcludmg enzymes [1,2] and mlcroblal cells [3] have been used to construct them Enzymes have been used most frequently because of their speclflclty to the substrate [4] Many methods for enzyme nnmobl- hzatlon [5-U] and momtormg of enzymatic reac- tlons [12-141 have been developed for construct- mg such sensors Needle-type glucose sensors seem attractive because of their small size, simple construction and the posslblhty of msertmg them directly into a fruit or the vascular tissue of a hvmg organism Many studies concernmg the construction of nee- dle-type glucose sensors have been reported [15- 171 However, most of them suffer from mstabll- sty, a low response and the inaccuracy induced by interfering materials [181 This paper describes the construction of a nee- dle-type glucose sensor suitable for determining glucose m blood The method combines pho- tocross-lmkmg of PVA-SbQ [19] and cross- lmkmg with glutaraldehyde to unmoblllze the en- zyme PVA-SbQ 1s a poly(vmy1 alcohol) bearing stllbazolmm groups [20] This polymer 1s photo- cross-lmkable with hght of wavelength shorter than 460 nm Blomaterlals mcludmg enzymes [21,22] and organelles [23] have been lmmoblhzed m this polymer Naflon was used to dlmmlsh the responses to ascorbate and urate [23], which are the major interfering materials m blood [181 The responses to ascorbate and urate at then normal concentra- tlons m blood were ehmmated after coating a Nafion membrane on the workmg electrode Nafion 1s a fluorine-contammg material and therefore it 1s difficult ot make another mem- brane adhere to it The high vlscoslty of PVA- SbQ, however, made it possible to encapsulate the tip of the electrode with an mnnoblhzed enzyme membrane Silver 1s most popularly used as a counter electrode [24] However when lengths of silver and platinum wires were immersed m phosphate- buffered saline (PBS) solution at room tempera- ture for 3 days, a dark layer formed on the surface of silver but not platinum Electroplated platinum was used as the counter electrode m this study because of its large surface area and comparative mactlvity to chemical reactions Both of these properties contributed towards a stable response and good stability of the sensor The sensor was used to determine glucose m serum and whole blood A good correlation be- tween these results from the sensor and those obtained with an automated analyser confums the posslblhty of applying this sensor m clmlcal analysis
`Chemicals Glucose oxldase (GOD) (E C 1 13 4) from Aspergzllus nrger and Bls-Trls propane {1,3- b~s[tr~s(hydroxymethyl)methylammolpropane] buffer were obtained from Sigma (St Louis, MO) PVA-SbQ was purchased from Toyo Chemical (Tokyo) Naflon perfluormated ion-exchange powder (5% solution m a mncture of lower ahphatlc alcohols and 10% water) was obtained from Aldrich (Milwaukee, WI) and used as sup- plied Bovine serum albumm (BSA) was obtained from Wako (Tokyo), cellulose trlacetate from Eastman Kodak (Rochester, NY), glutaraldehyde (50% aqueous solution) from Tokyo Kasel (Tokyo) and the electrolyte solution for platinum electro- plating (Platanex 3LS) from Japan Electroplatmg Engineers (Tokyo) A glucose analysis lut based on hexokmase-glucose-6-phosphate dehydroge- nase was supplied by Boehrmger (Mannhelm) and a glucose analysis lut based on glucose oxl- dase by Wako Phosphate-buffered saline (PBS) solution was prepared by dlssolvmg 2 754 g of NaCl, 2 081 g of KH,PO, and 0 477 g of NaOH m 1000 ml of distilled water and adjusting the pH to 7 4 with 0 1 M NaOH solution [25] Glucose solutions were prepared m PBS and allowed to stand for at least 24 h before use to equlhbrate the LY- and p-anomers All other chemicals were of the hlgh- est grade available and were used as received Instrumentatwn and mater&s A potentlostat (BAS LC-4B amperometrlc de- tector, Bloanalytlcal Systems, Lafayette, IN) was used to supply a fured potential to the electrode
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 2 of 10
`
`

`

`C -Y Chen et al /Anal Chm Acta 265 (1992) 5-14 7 A clrculatmg water-bath (Therm0 Mmder Mml- 80, Talyo Science, Tokyo) which incorporated a water-jacketed glass reactor was used to control the temperature of the operating system Mag- netlc stlrrmg was used to mamtam homogeneity of the sample solutions m the reactor The re- sponse current could be read directly from the digital screen of the potentlostat and recorded simultaneously by a chart recorder (Electromc Polyrecorder EPR-lOOA, TOA Denpa Kogyo, Tokyo) A schematic diagram of the batch operat- mg system IS shown m Fig 1 Another potentlostat-galvanostat (HA 501, Hokuto Denko, Tokyo) was connected with a function generator (HB-107A, Hokuto Denko) to perform cychc voltammetry and electroplatmg A multuneter (DigItal Multuneter, TR6840, Take- da-Rlken, Tokyo) and an Ag/AgCl electrode (HS-907, TOA Electric, Tokyo) were used to measure the potential dnft during a determma- tlon The clmlcal analyser used was a Beckman Glucose Analyzer II (Beckman Instrument, Palo Alto, CA) Heat-shrmk FEP (fluormated ethyl- ene-propylene) tubing was obtained from Jun- kosha (Tokyo) A stamless-steel hypodermic nee- dle (0 d 12 mm, 1 d 10 mm) was obtamed from Terumo (Tokyo) Platmum wire of 0 3 mm dlame- ter was obtained from Tokunlu (Tokyo) Preparation of the workuzg electrode Platinum wire (SO mm x 0 3 mm diameter) was soldered to a lead wire (copper wire electroplated with tin and then msulated mth Teflon) Heat- shrink FEP tubing was used to insulate the plat- mum wu-e The tip of the FEP-encapsulated plat- mum wne was cut at an 18” angle and succes- sively polished with water-proof sand-papers (No 320 and 1000, Marumoto, Tokyo) and a &con carblde Paper (No 2400, Struers, Copenhagen) After the electrode had been cleaned by somca- llulose triacetate membrane Glucose oxidase - PVA membrane Naflon membrane Heat shrink FEP tube Working electrode (b = 0 3 mm platinum wire) Counter and reference electrode (stalnless steel hypodermic needle electroplated with platinum) Fig 1 SchematIc diagram of the batch-type operatmg system and the multi-layer membrane system on the tip of the workmg electrode (1) Thermostated circulating water-bath, (2) magnetic stirrer, (3) water-jacketed glass reactor, (4) electrode, (5) potentlostat, (6) chart recorder
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 3 of 10
`
`

`

`Preparahon of the counter electrode
`
`(1 + 1 +
`
`A stainless-steel hypodermic needle was cleaned with methanol and dlchloromethane It was further cleaned w&h an oxldatlve acid solu- tion [concentrated sulphurlc acid-30% hydrogen peroxide (1 + 111 and an oxldatlve alkaline solu- tion [concentrated ammonia solution-30% hydro- gen peromde-water
`
`C -Y Chen et al /Anal Chm Acta 265 (1992) 5-14 tlon for 30 mm, the m&l-layer membrane system was coated on to the metalhc surface of the workmg electrode Three membranes made up the multi-layer system Irutlally the tip of the workmg electrode was dipped mto Nafion solution for 10 s and then dried at room temperature for 30 mm The trp of the electrode was then dtpped mto an enzyme solution composed of 5 mg of GOD (25 U mg-‘1, 10 mg of BSA, 100 mg of dlstllled water and 200 mg of PVA-SbQ for 5 s After dlppmg, the electrode was placed m a sealed, dark box con- tammg glutaraldehyde vapour The box was kept at room temperature for 12 h to complete cross- Imkmg, then the electrode was exposed to a fluo- rescent Iamp for 10 mm to induce photocross- hnkmg of PVA-SbQ FmaIly, the electrode was dipped m a 0 5% (w/v) cellulose trlacetate solu- tion m dlchloromethane for 3 s and dned for 5 mm at room temperature The electrodes were stored dry at 4°C until used The multi-layer membrane system IS shown m Fig 1
`611 It was thor- oughly washed with dlstrlled water after each cleaning step Nickel and platinum were electroplated suc- cessively on to the needle The processes of elec- Stainless steel hypodermic needle electroplated wlth platinum Plafmum wife insulated wlth heat shrink FEP tube and then covered by multi-layer membrane system Epoxy resin shrinking tube as lnsulatOr > Platinum wire as working electrode / PJ~num black as counter and reference electrode
`
`Ftg 2 Procedure for consh-uchng needle-type enzyme electrode
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 4 of 10
`
`

`

`C -Y Chen et al /Anal Chun Acta 265 (1992) 5-14 9 troplatmg were conducted at 80°C The needle and a piece of nickel wire were connected to the potentlostat and unmersed m an electrolyte solu- tion prepared by dlssolvmg 60 g of N&l, 6H,O m 30 ml of concentrated hydrochloric acid and ddutmg v&h dlstllled water to 250 ml A poten- tial of - 4 0 V was apphed between the needle and the nickel wire for 30 s After thorough rmsmg with distilled water, the needle was trans- ferred mto Platanex 3LS platinum electroplatmg solution A current of -20 mA was applied be- tween the needle and a piece of platmum wire for 5 mm The needle was thoroughly rinsed again with dlstllled water and then stored dry at room temperature until used Fabncatron of glucose sensor The workmg electrode was inserted mto the hollow mterlor of the counter electrode Epoxy resm was used to fix the hvo electrodes m posl- tlon The procedure for constructing the glucose electrodes IS shown m Fig 2 Measurement of potentral dnft dunng glucose determtnatwn An Ag/AgCl electrode was inserted m the reactlon cell with the glucose sensor Durmg glu- cose determmatlon the potential between the workmg electrode and the Ag/AgCl electrode was measured with a voltmeter and recorded with a chart recorder Determrnatwn of glucose concentratwn The electrode was Immersed m PBS solution for 1 h to equlhbrate the membrane system The copper wires were then connected to the poten- tlostat and a potential of +650 mV was applied between the workmg and the counter electrodes The baseline current was measured and then glucose solution was inJected mto the PBS solu- tion using a mlcrosyrmge The response current followmg mJectlon was recorded with a chart recorder until the second steady state was achieved Magnetic stirring was used during this operation to ensure homogeneity of the solution The difference between the basehne and the sec- ond steady-state currents was used to calculate the concentration of glucose 111 the sample ac- cordmg to a cahbratron graph Another cahbratlon graph was obtained by adding glucose solution to heparmlzed whole blood that had been incubated at 37°C for 18 h to glycolyse the glucose present [26] The concentra- tion of glucose m serum and whole blood was determined by takmg the baseline and the second steady-state currents m the glycolysed whole blood and the sample, respectively, and usmg them to calculate the glucose concentration ac- cording to the cahbratlon graph obtained with glycolysed whole blood The samples were mixed m the sampling tube by gentle shakmg before determmatlon but were not stirred during deter- mination RESULTS AND DISCUSSION Determmatzon of the appkd potent& The cychc voltammograms of the electrode usmg platinum wire as the workmg electrode and platinum black as the counter and reference elec- trode m PBS with or without hydrogen peroxide are shown m Fig 3 The plateau of this electrode after contactmg with hydrogen peroxide appeared from 450 to 750 mV The potential between the workmg electrode and the Ag/AgCl electrode drifted by -57 mV when determmmg 400 mg dl- ’ (22 2 mM) glucose m PBS A potent& of +650 mV was chosen after takmg the range of the plateau regon and the potential drift durmg glucose determmatlon mto conslderatlon This 69 Cl P !L 200 mV Fig 3 Cychc voltammogram of the electrode usmg platmum wire as the workmg electrode and platmum black as the counter and reference electrode Expenments were done m (a) PBS and (b) 1 mM H,Oz m PBS (pH 7 4) at 37°C
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 5 of 10
`
`

`

`10 C-Y Chen et al /Anal Chm Acta 265 (1992) 5-14 ensured that the polarlzmg voltage of the sensor remained m the plateau region during the deter- mmation of glucose Functwn and effect of each membrane zn the m&-layer membrane system Three layers were mcluded m the multi-layer membrane system The innermost layer was a Nafion membrane Nafion IS a polymer with a negative charge m the pH range of blood and therefore curtalls the passage of interfering an- ions such as ascorbate and urate to the working electrode The middle layer was the unmobdlzed glucose oxldase membrane Enzyme unrnobdlza- tlon was achieved by a double cross-lmkmg method that used a photocross-lmkable poly(vmy1 alcohol) compound (PVA-SbQ) m conJunction with a chemical cross-hnkmg compound (glutaral- dehyde) This led to firm entrapment of the en- zyme and consequently good stablhty The outer- most layer was a cellulose trlacetate membrane This membrane excludes compounds on the basis of molecular size Large molecules were com- pletely excluded and small molecules partly ex- cluded from mteractlon with underlying layers of the sensor This had two advantages The first function was to act as a protective ultrafiltration membrane and prevent large molecules such as protems and other enzymes m the sample from fouling the sensor This can prevent the GOD from being damaged by other enzymes such as proteases m the sample solution The second function was to reduce the rate at which glucose permeates mto the enzyme layer, which extends the dynamic range of the sensor The effects of the successive membranes on the response to ascorbic acid, uric acid, hydrogen peroxide and glucose are shown m Fig 4 The response to 2 mg dl-’ ascorbic acid IS shown in Fig 4A The bare electrode responded strongly to ascorbic acid, but this response was reduced to zero when the metalhc surface of the electrode was covered with a Nafion membrane Figure 4B shows the response to 10 mg dl -’ uric acid The effect IS the same as for ascorbic acid, 1 e , the bare electrode responded to uric acid, but the response was reduced to zero after spreading a layer of Naflon on the worlung electrode The (A) (6) -- ‘---W $ 1 z T a Response time ( l min 1 Fig 4 Effects of each kmd of membrane on the response of the sensor to (A) 2 mg dl-’ ascorbic acid, (B) 10 mg dl-’ uric acid, (C) 0 1 mM hydrogen peroxlde and (D) 50 mg dl-’ glucose Curves (a) are the response of the bare electrode, (b) the electrode coated with Naflon, (c) the electrode coated with Nafion and enzyme-PVA and (d) the electrode coated with Naflon, enzyme-PVA and 0 5% cellulose tnacetate Ex- penments were conducted at 37°C m PBS (pH 7 4) concentrations of ascorbic acid and uric acid nor- mally found m blood are 0 4-l 5 and 15-8 0 mg dl-‘, respectively [27] The response to 0 1 mM hydrogen peroxide 1s shown m the Fig 4C The highest response was from the bare electrode The response was reduced to 40% because of the dlffuslon barrier induced by coating the electrode with Naflon The unmoblhzed enzyme and cellu- lose trlacetate membranes further reduced the response However, hydrogen peroxide 1s pro- duced m the PVA membrane durmg glucose de- termination and therefore the Naflon membrane alone reduces its response Figure 4D shows the response to a 50 mg dl-’ glucose solution The response to glucose was mduced by munoblhzmg glucose oxldase on the workmg electrode The response was decreased after coating the electrode with a cellulose trrace-
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 6 of 10
`
`

`

`C-Y Chen et al /Anal Chun Acta 265 (1992) 5-14
`
`11
`
`7 a- .-.-.
`
`700
`
`2
`E
`
`Cone of glucose (mg/dl) Fig 5 Cahbratlon graphs m (0) PBS and
`
`(01
`
`l
`
`/
`
`glycolysed whole blood for the sensor dip-coated m cellulose tnacetate solution (0 5%, w/v) Expenments were conducted at 37°C tate membrane Higher cellulose trlacetate con- centrations lowered the response intensity of the sensor, but increased the dynamic range The concentration of cellulose tnacetate used m this work was chosen according to the desired dy- namic range of the sensor A decrease m re- sponse intensity is one of the disadvantages of using a multi-layer membrane system However, the remaining response was still high enough to momtor the reaction catalysed by the enzyme Cahbra twn Figure 5 shows the cahbratlon graphs m PBS and glycolysed whole blood for a sensor dip- coated by lmmerslon m cellulose trlacetate solu- tion [O 5% (w/v) m dlchloromethanel The dy- namic range of this sensor was O-360 mg dl-’ glucose, which is sufficient for measurement of the glucose concentration m the blood of non-dl- abetlc and most diabetic patients This sensor was used to determme the glucose concentration m serum and whole blood samples m this study Experiments to determine the response and cah- bratlon graphs using cellulose trlacetate mem- branes of different thickness were also carried out This mdlcated that it was possible to mcrease the dynamic range up to 800 mg dl- ’ glucose by Increasing the concentration of the cellulose tn- acetate solution to 2% (w/v) However, the re- sponse was then low (15 nA for 100 mg dl-’ glucose) and slow (2-3 mm for a 95% response) A 0 5% (w/v) cellulose trlacetate solution was adopted m this study because It covers the range of glucose concentrations that are commonly en- countered m a cluucal laboratory Effects of temperature, pH and metal zom The effect of temperature on the response of the sensor IS shown m Fig 6 The response m- creased with increase m temperature from 5 to 5O”C, the response at lower temperatures re- mained unchanged after testing at 50°C These results suggest that the enzyme, glucose oxldase, 1s stable at fairly high temperatures after being lmmoblhzed by the method employed here The response intensity at 37°C was 81% of that at 50°C Figure 6 also shows the response time of the sensor The time required to reach the sec- ond steady state was dependent on temperature The lower the temperature, the longer was the response tune At 37”C, however, less than 30 s were required to obtam a 95% response The effect of pH IS shown m Fig 7 Four kmds of buffers covering the pH range 4-10 were utl- hzed The response at a given pH was different m drfferent buffers The highest response was ob- tamed at neutral pH m phosphate buffer The optimum pH of GOD m solution IS 5 6 and there is a rapid and permanent loss of activity at pH values lower than 2 or higher than 8 [28] The optimum pH of the munoblhzed GOD was 7 0 Response time Fig 6 Effect of temperature on the response of the sensor Experiments were conducted m PBS (pH 7 4)
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 7 of 10
`
`’
`

`

`12 C -Y Chen et al /Anal Chun Acta 265 (1992) 5-14
`
`120 r
`
`Selectwrty, stabdrty and reproductblhty of the
`sensor
`The
`
`Ca*+ Ca(NO,), 2750 b 990 Mg2+ Mg(NO& 6H2O 1235 b 103 9 Fe3+ Fe(NO,), 30 b 977 CL?+ CuSO, 5H20 25 b 917 Pb2+ Pb(NO,), 25 103 5 -Q+ &NO, 25 00 Hg2+ HdNOA 25 369 a Expenments were conducted m 30 mM Bls-Tns propane buffer (pH 7 4) at 37°C b The concentration IS shghtly higher than the average value m serum [271 TABLE 2 Relative response of the sensor and assay lots to various mono- and disaccharides Sugar Relative response (o/o) Sensor a HK-G6PDH GOD Reference b Glucose 100 100 loo loo Fructose 3 5 07 09 0 Galactose 1 7 04 08 0 14 Maltose 36 18 10 0 19 Mannose 3 1 _c 37 0 98 Lactose - - 0 Sucrose - - 0 a The expenments with the sensor were conducted m PBS contauung 2 775 mM sugar b Data from [30] ’ Dashes mdl- cate undetectable
`’ 3 ’ 8 0 m ’ ’ 3 4 5 6 7 6 6 16 11 PH Fig 7 Effect of pH on response of the sensor Expenments were conducted at 37°C m 0 1 M buffers (0) &rate (pH 4-7), (0) phosphate (pH 6-8), (m) borate (pH 8-l@, (e) NaHCO,-NaOH (pH 9-10) However, the actlvlty loss at pH > 8 was slmllar to that of GOD m solution The effects of some metal ions are shown m Table 1 Calcmm, magnesrum, Iron and copper are the mam metal ions encountered m serum Of these, lron(II1) and copper(U) caused a de- crease of 2 3% and 8 3%, respectively, m the response of the sensor Sliver(1) completely mhlb- lted the response and mercury(I1) reduced the response of the sensor to 37% Inhlbltlon of unmoblhzed GOD by sliver(I) and mercury(I1) has been studied [29] These Ions, however, are present only at very low concentrations m blologl- cal fluids TABLE 1 Effect of metal Ions on response of the sensor a Metal Salt Concentration Relative Ion (ILM) response (%) None
`selectivity of the sensor was tested by determmmg various mono- and disaccharides at the same concentration as glucose The relative responses of the sensor for various sugars are shown m Table 2 The relative responses to glu- cose, fructose, galactose, maltose and mannose were 100, 3 5, 17, 3 6 and 3 1, respecttvely The sensor did not respond to lactose and sucrose The speclflclty of GOD for glucose, galactose, maltose and mannose has been reported as 100, 0 14, 0 19 and 0 98, respectively 1301 The same workers reported that It did not react with fruc- tose, lactose and sucrose In order to make sure that the response of the sensor was not due to contammatlon with glucose, they were assayed for glucose with assay luts based on hexokmase- glucose-6-phosphate dehydrogenase (HK- G6PDH) and glucose oxldase (GOD) The rela- tive responses of the assay kits for various sugars are also shown m Table 2 Roth of the assay luts responded to the fructose sample, but did not respond to lactose and sucrose The GOD kit also responded to mannose These results mdl- cate that response of the sensor to fructose was partly due to contammatlon with glucose GOD reacts, albelt slowly, with galactose, maltose and mannose Accordmg to the operating data sheet provided by the suppher, the GOD used m this study also contained maltase, glycogenase, mver- tase, amylase and galactose oxldase as contami-
`
`1000
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 8 of 10
`
`

`

`a
`
`C-Y Chen et al /Anal Chm Acta 265 (1992) 5-14 13 8 1 .A ._.-0-e !z Q b 9 =
`
`ji -0-e -* \ I 0 5 10 15 20 25 Period of preservation ( days ) Fig 8 Relative response mtenslty of the sensor when stored at room temperature m PBS (pH 74) Experiments were conducted at 37°C m PBS contanung 100 mg dl-’ glucose nants However, sugars other than glucose are present only at very low concentrations m blologl- cal fluids and therefore the resulting errors would be negligible Figure 8 shows the relative response of the sensor over a period of 25 days when stored m buffer solution at room temperature The re- sponse declmed by lo-15% m the first day but then became stable for more than 3 weeks The decrease m response during the first day could have been caused by escape of enzyme molecules that were not firmly lmmoblhzed or by a decrease m the response of the electrode to hydrogen peroxide Because the response of the sensor to 0 1 mM hydrogen peroxide was found to be un- changed after 3 days, it 1s hkely that loss of enzyme was chiefly responsible for the decline Further decrease m response was found on the 25th day The membrane system was found to be folded up when observed under a microscope at that time The reproduclblhty of the sensor was tested on the seventh day after the sensor had been fabn- cated For 24 successive determmatlons of the same glucose solution, the relative error of the responses was less than 1% Determmatlon of glucose m serum and whole blood Serum and whole blood samples were assayed for glucose with both the sensor and the auto- mated clmlcal analyser Lmear correlatrons were achieved m both mstances, 1 e Y = 0 925x + 12 45 (r=O95, II = 14) for serum and y=O991x+ 6 597 (r = 0 993, n = 17) for blood, where y IS the glucose concentration (mg dl-‘1 measured by the sensor and x that by the chmcal analyser Concluswn A needle-type glucose sensor had been fabn- cated Electrodeposltlon of platinum on to a dls- posable hypodernuc needle allowed a cheap but rehable counter electrode with a large surface area to be constructed Either the large surface area or the relative mactlvlty of the platinum to chemical reactlons contributed to the stable re- sponse of the sensor Nafion was coated on to the metallic surface of the workmg electrode This prevented ascorbate and urate from interfering with the response of the sensor The high WCOS- lty of PVA-SbQ made it possible to coat the enzyme membrane on the surface of the Naflon membrane by dlppmg rt mto the enzyme-contam- mg PVA-SbQ solution The enzyme was munobl- hzed by bemg cross-hnked by glutaraldehyde and then entrapped by photocross-hnkmg of PVA- SbQ As a result, GOD was retained firmly m the membrane, and this also contributed to the stabll- lty of the sensor A cellulose trlacetate membrane was used as the outermost layer to protect the enzyme membrane and to vary the dynamic range over which the sensor responded to glucose The dynarmc range of the sensor to glucose could be adjusted from O-30 to O-800 mg dl-’ depending on the concentration of cellulose trlacetate used m constructmg the multi-layer membrane system The needle shape of the sensor means that it can be used m a small volume of sample, implanted m a hvmg orgamsm or inserted m a fruit The construction of this sensor was simple and easy to perform Each step of fabrication can be carried out m a laboratory equipped with a potentlostat- galvanostat, chart recorder and water-bath The response, stab&y and reproducrblhty of the sen- sor were good enough to be apphed to the deter- mination of glucose m blood samples m a chmcal laboratory The batch operatmg system allowed more than fifteen samples to be assayed per hour Faster operation may be possible if the sensor IS connected to an autosampler and flow-mJectlon
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 9 of 10
`
`

`

`14
`
`[31-331
`
`,
`Bloelectron
`5 (1990) 47 B A Gregg and A Heller, Anal
`Chem ,
`
`62 (1990) 258 J Wang, D Leech, M Ozsoz and S Martinez, Anal Chum Acta, 245 (1991) 139 10 G H Hsme, C H Lm and CC Wang, J Appl Polym Scl, 38 (1989) 1591 11 M Demura, T Asakura and T Kuroo, Biosensors, 4 (1989) 361 C -Y Chen et al /Anal Chum Acta 265 (1992) 5-14 12 H Muramatsu, J M Dlcks and I Karube, Anal Glum Acta, 197 (1987) 347 13 M J Muehlbauer, E J Gmlbeau, B C Towe and TA Brandon, Biosensors Bioelectron , 5 (1990) 1 14 H Gunasmgham, C H Tan and J K L Seow, Anal Chem , 62 (1990) 755 15 M Sluclun, R Kawamon, Y Gorlya, Y Yamasalu, M Nomura, N Hakm and H Abe, Diabetologia, 24 (1983) 179 16 M Stuchm, N Asakawa, Y Yamasalu, R Kawamon and H Abe, Diabetes Care, 9 (1986) 298 17 S J Churchouse, CM Battersby, W H Muller and P M Vadgama, Biosensors, 2 (1986) 325 18 D S Bmdra and G S Wilson, Anal Chem, 61 (1989) 2566 19 K Iclumura and S Watanabe, J Polym SCI, Polym Chem Ed, 20 (1982) 1419 20 K Iclumura, J Polym
`It will be necessary to mvestlgate blocompatlblhty [34] and mediated electron trans- fer [35] prior to m vlvo use of the sensor, and this is in progress The Admmlstratlon of Education of Taiwan 1s acknowledged for sponsormg this study Dr Robert Wilson and Mr Mark Buckmgham are thanked for assistance with the Enghsh and for helpful discussions Colleagues m the Depart- ment of Agricultural Chemistry, National Taiwan University, are also acknowledged for helping one of the authors (C -Y C > with his teaching and admmlstratlve responslblhtles during this study m Japan REFERENCES L C Clark, Jr, and C A. Dhggan, Diabetes Care, 5 (1982) 174 W H Mullen, FH Keedy, ST Churchouse and PM Vadgama, Anal Chum Acta, 183 (1986) 59 I Karube, S Mitsuda and S Suzulu, Eur J Appl Micro- tuol Biotechnol , 7 (1979) 343 LC Clark, Jr, m AP F Turner, I Karube and G S Wdson (Eds ), Biosensors, Fundamentals and Apphca- tions, Oxford Science Publications, New York, 1987, pp 3-12 G Kampfrath and R Hmtsche, Anal Lett , 22 (1989) 2423 Y Okawa, H Tsuzukl, S Yoshlda and T Watanabe, Anal SC1 , 5 (1989) 507 C Gahatsatos, Y Ikanyama, J E Mark and W R Heme- man, Blosensors
`
`Polym Chem Ed, 22 (1984) 2817 21 K Matsumoto, H Mlzuguchi and K Iclumura, Kobunshl Ronbunshu, 41 (1984) 221 22 K Iclumura and K Ueno, Nihon Kagaku Kaishi, (1984) 375 23 D J Harrison, RF B Turner and HP Baltes, Anal
`60 (1988) 2002 24 MA Genshaw and E J Jones, J Electrochem Sot, 136 (1989) 414 25 D D Perrm and B Dempsey, Buffers for pH and Metal Ion Control, Kodansha, Tokyo, 1981, p 68 26 D J Claremont, C Penton and J C Pickup, J Blamed Eng , 8 (1986) 272 27 I Kanal and M Kanal, Chmcal Laboratory Methods and Diagnosis, Kobunsha, Seoul, 1983, p 1267 28 B Maruo and N Tanuya, Enzyme Handbook, Asakura, Tokyo, 1982, p 64 29 C C Lm, FM Fryburg and AK Chen, Bloelectrochem
`8 (1981) 703 30 D K&n and E F Hartree, Bmchem J , 50 (1952) 331 31 H Gunasmgham and CH Tan, Anal Glum Acta, 229 (1990) 83 32 R Appelqwst and E H Hansen, Anal Ctum Acta, 235 (1990) 265 33 W Matuszewslo, M TroJanowicz and L Ilcbeva, Electro- analysis, 2 (1990) 147 34 I Hanmng, P M Vadgama, A K Covmgton and K G M M Albert], Anal I&t, 19 (1986) 4Cl 35 J C Pickup, G W Shaw and D J Claremont, Biosensors, 4 (1989) 109
`
`Sci ,
`
`Chem ,
`
`Bloenerg ,
`
`LifeScan Ex. 1041
`LifeScan Global Corporation v. Ikeda Food Research Co., Ltd., et al.
`PGR2019-00032
`
`Page 10 of 10
`
`system
`