CA2132569A1 - Biosensor and interface membrane - Google Patents

Abstract

A regenerable biosensor probe adapted for positioning in a bioreactor comprises a selectively permeable interface membrane, a porous protein-receiving matrix adjacent to the interface membrane, an indicating electrode, an inlet conduit through which fresh protein conjugate may flow to the protein-receiving matrix, and an outlet conduit through which spent protein conjugate may be removed from the protein-receiving matrix. A
selectively permeable interface membrane, which may be used in a biosensor system to separate biochemical, optical or other processes from an analyte matrix comprises a supporting mesh, a perfluorosulfonic acid polymer impregnated substrate and a homogenous film of perfluorosulfonic acid polymer.
A method of preparing this interface membrane comprises fixing a substrate on to a supporting mesh to form a substrate membrane, casting a perfluorosulfonic acid polymer on the substrate membrane and curing the product to so formed.

Description

2132~B9 ~D OF THE INVENIION
This invention generi311y relates to biosensors and, more ~ " to interface membrane systems which may be used with biosensors.
BACKGROUNl) OF TEIE INv~or~
Since the enzyme elecerode was first conceived, a significant interest has developed in the field of biosensors beca~se of the simplicity and selectivity of these sensors.
This increase in interest began in the 1980's, as evidenced by the L ' ' " of a new i",~ i(",,.l journal Biosensors~. The ral~id growth of hi~ ' Oy in the last decade has now created a demand for more and better on-line sensors that can be interfaced with computers to control and optimize bi~u~ . Numerous enzyme electrode .
have been published, although the amoum of research activity today is too great to be covered in a single review2. Present-day ~ '' ' of biosensors can be found in industrial bioprocess " ", environrnental ' ' ~, the food and drink industry, and clinical and in vivo ~ ' in medicilne3 4.
The increasing commercial imporrance of bioreactors has stimulated research in the area of on-line monitoring of micrcbial and animal cell bioreactors in order to optimize p~.rul.l.dllu~ conditions. The specificity and selectivity provided by the biological component of biosensors offer enormous l~otential, in principle, for ( " , on-line analysis in complex media. Despite the number of biosensor research papers published edch Stoecker, P.W.; Yacynych, A.M Chermcally Modified Electrodes a6 siOsensor3 Sel. E~lec. Rev. 1990;
12, 137-160.

2 Ereitag, R. Applied Bioserlsors. Cni7. Opln. Biotech.: ~nal. siOtech. 1993, 4, 75-79.

3 Schult~, I.S. Sophi~ticated Descendant3 of The carlary in The Coal Mine Are sased on Molec~r Cor~ponents of Plants and Anjmals ~ound to Microscopic Electrodes or Optical Fibers. Scienhfic Amenccn 1991, 264, 64-69.
Reach, G.; Wilson, G.S. ci n Cortirluous Glucose Moritoring be Used r~Or The Treatment of Diabetes?
Anal. Chem. 199~, 64, 381-386.
1 12353\û266518.11P

~2~
year in the scientific literature, relatively few bioprocess sensors are commercially available.
Ul~ruli 'y, many practical problems r~main which have prevented the widespread application of on-line biosensors under rell conditions and these have not been successfully addressed by researchers to date. In palticular, the I~ v, ', and commercialization of biosensors has been slowed by problems of instability such as drift of the sensor signal, narrow measuring ranges for the analyte ~und long response time.
Short term ' " typically result from changes in the enzyme, . t, such as inhibition or deactivation by, , of the analyte medium. Long term ' ' (drive of the sensors signal o~er time) may be due to time-dependent changes in the sensor calibration constants which are caused by membrane fouling or electrode poisoning. Perm- selective ' fclr: UIIIC:lliC biosensrng have been illv~
as a means to address these problems. A literature review of these ' is provided in Wang, 19925.
Although the Irfetime of an enzyme in a reaction system may be prolonged in some cases, the ~' of enzyme~, may be ill~ ;blc and ~,l ..., ' of the enzyme when the activity has degraded to an, - r ' ,y degree may eventually be necessary with virtually all of the ~ .lly available biosensors (whether on-line or not).
The capability to replace only the erlzyme component of the sensor without disrupting the process would not only extend the sensor operating lifetime but would allow for a of other enzymes in order to l~hange the analyte specificity of the sensor.
A few sensor systems have been described in the literature with the capacity for enzyme ll r 1l ' Brooks et al. (1987/88)6 and Bradley and Schmid (1991)~ have Warlg, J. Permselective Coatings For ~mp rometric Biosensing In Biosensors and Chemical Sensors:
Ophlni~ing Performance 7hroL~gh Pc~lymeric Materials; Eddman, P.G., Wang, I., Eds.; ACS: New Merico State University, 199~; Sym,posium Senes 457, pp 1~5-13i.
6 Broo~ts, S.L.; Ashoy, R.E.; Turner, A.P.F; Calder, M.R; Clarke, DJ. Develmpment of an On-Lnne Glucose Sensor For Fermentation Monitormg Biosensors 1987/83, 3, 45-56.
7 Bradley, J.; Schmid, R D. Optimisa ion of a Biosensor Por 7n sihl Fermentation Monitoring of Glucose Concentration. Blosensors & Bioelectronics 1991, 6, 669-674.
112353\0266515.WP - 2 -~escribed the ' ' of enzyme on graphite discs which could be replaced manually.
There is no, l ' of in situ, automatic exchange of the enzyme, that is, without .' ' ~ the sensor or ,, g the r ' " Pieters and Bardeletti (1992)S
described the ' ' of enzymes to magnetic beads which can then be . ' using magnetic fields. This technique ha~i beer~ used in waste-water treatment, affinity separation processes, cell sorting, y~ and drug debvery. Specifically, glucose oxidase has been reversibly ' " ' in an enzyme reactor coupled to a flow injection analysis system using a chain of biospecific reactions based on the binding of ~ r~11 (ie. bi y' ') antibodies to an avidin coated matrLY (de Alwis and Wilson 19899).
In summary, at the present time, only a few reaction parameters can be reliably monitored on-line (eg. ~ , pH, dissolved oxygen tension, stir rate) without the use of highly r~ and costly , on The analysis of ff ~ ' "
substrates, products and ' -' is us~lally acmeved by off-line methods'. However, optimal control of a bioprocess today reqllires that m~ h1f parameters be determined as frequently as possible, which in turn rfquires frequent sampling that increases the risk of f nn r. ~h~ o~ci, off-line mel:hods are usually to slow to be used in a closed-loop control system, and it is often difficult to ensure that samples are not -i, r '~S degraded or changed during the ~ ~ ' g: lS,,;~ procedure. A sensor system based orl an in situ probe which could provide, , resl-time analysis would be extremely valuable, pa~ ly for high-density, fed-batch processes. The advantages and .' ~v ~ of automated sampling systems in contrast with in situ probes have been discussed in the literature (Ogbomo et al, 1990ll; Bradley et al, 1991~2; Filippini et al, 1991i3). The 8 Pieters, B.R.; Bardelett~, C. Erzymi~ I ' ' on a Low-Cost Magnedc Support: Kinetic Studies on Immobilized And r ' ' ' alucose Oxidase And Glucoamylase. Enzyme Microb. Techrlol 1992, 14, 361-370.
de Alwiss, U; Wilson, G.S. Strategies For Ihe Reversible r ~ " of Enzymes by Use of Biotin-Bound Anb-Enzyme Anbbodi~s. Talar~.3 1989, 36, 249-253.
0 Supra at 6.
O~bomo, 1.; Prinzin~, U.; Schmidt, H.L. r'rerequisites For The Control of Microbial Processes by Flow Injection An;!lysis. Jr. Bio~ech~rol. 1990, 14(1), 63-70.
11Z353\0266518.WP ~ 3 ~

~1~2369 practical concerns involved in using an in sifu biosensor probe, such as in situ ' ')il;.y, long-term stability, adequate measuring r~mge, and membrane fouling, have thus far prevented the widespread application and commercialization of this approach. Attempts to address these problems have heretofore been largely, r ~ (l;nfors and Molin, 197814; Cleland and Enfors, 198315; Bradley et al., 1988~6, 1991~7 and Brooks et al., 1987/88'8; Buhler and Ingold, 197619). ~n each of these probe designs, the combined functions of enzyme ~ and r~r llihrn~it~n of the sensor cannot be performed without operator intervention. Thus, an operator must be standing by during a long r, run to manually replace the enzyme componelnt periodically and then recalibrate the sensor.
The outer membrane of a biosensor is very important, as it represents the interface between the sensor amd the anal!~te medium. The purpose of this interface membrane is to allow the diffusion of ana.lytes and (in ~ ' l reactions) electrolytes into the h~ Li6~liv~ or analysts layer while excluding potential interfering species which may be present in the analyte medium, such as cells, proteins, inhibitors or 12 Bradley, ~.; Stocklein, W.; Schmid, R.D. Biochemistry Based Analysis Systems for Bioprocess Monitoring and Control. Proc~ss Contr. Or~al. 1991, 1, 157-153.
13 Filippini, C.; Somlleitner, B.; Fiechter. A.; Bradley, J.; Sohnnd, R. On-Line Determination of Glucose in r ~ _ Processes: t~omparison BetweeD FIA And an In situ Enzyme Electrode. J.
Biotechnol. 1991, 18, 153-160.

4 Enfors, S.O.; Nilsson, H. Design And Responso Char4cterishcs of an enzyme Electrode For Measurement of Pemcirin in Fermenta~ion Broth. r~nzyme Microbiol. TechnoL 1979, 1, 260-264.
Cleland, N.; Enfors, S.O. Monitorhlg Glucose Consumphon in an ~chetichi~t coll Cultiv~ltion With an Enzyme Electrode. Anal. Chim. Acta. 1984b, 163, 231-285.
16 Bradley, 1.; Anderron, P.A.; Dear, A.M.; Ashby, R.E.; Turner, A.P.F. Glucose Biosensors for the Stndy and Control of Brker's Compressed Yest Produchon. In Comp~aer Applicatront~ in Fermentatron Technolor,v: Modellinf~ t~nd Controt of P ' ' ~ ur ProcesseJ; Fish, N.M., Fox, R.l., ThornhiU, N.F., Eds.; Elsevier Applied Science: New York, 1988; pp 47-SI.
17 Supr~ at 7.
18 Supra at 6.
9 Buhler, H.; Ingold, W. Mesuring FIH and Oxygen in Permenters. ProceJJ B~ochem. 1976, 11(3), 19-24.
11Z353~0Z66518.WP ~ 4 ~

~1~2~g In United States Patent No. 5,165,407 to Wilson et al., an ,' ' ' glucose sensor is provided having an enzyme ' ' ' on an indicating electrode and a permeable polyurethane membrane applied over the sensor body to prevent fouling of the electrode and ~If.~ lll of the enzyme. This membrane, as with others currently employed, is integrated with the sensor system and can not be replaced or reused with other systems.
The objects of this inventi~n are to obviate or mitigate the disadvantdges of the current in sit7l biosensor probes and the disadvantages of current membrane systems for use generally with biosensor probes.
SUMMARY OF TEIE INVENTION
In a first aspect of the present invention a lc5~ dlJI~ biosensor probe, adaptcd for operation in an Ull~ . tf..;,.,l by the presence of biological molecules which are substrates for or products produced by enzymes or cells in order to determine the presence of said molecules is provided which comprises a selectively permeab1e interface membrane which separates the ~ ' 1 and optical or el~l-u.,l.~llliudl processes from the analyte matrix CllVil~ when the probe is im place;
an protein-receiving matrix adjacent to the interface membrane; an indicating electrode covered with an electrically insulative material said electrode abutting, at one of its ends, the protein-receiving matrix; an inlet conduit through which fresh protein-conjugate may flow to the protem-receivmg matrix; and an outle.t conduit through which spent protein-conjugate may be removed from the protein-receiving matrix. This biosensor probe is capable of filnf~ti( - in a closed-loop or computer-controlled system.
In a second aspect of the present invention a selectively perrneable interface membrane is provided which may be use~i with any type of biosensor to separate the ' and optical or r l~ u~l~el~ ,1 processes from the analyte matrix which comprises a supporting mesh, a pc r~ acid- ,, " ' substrate and a ,f ---, film of p, r~ u~ulruilic acid polymer.
112353\0266518.WP - S -~. , 21~2~6g In a third aspect of the pre'ient invention a method of preparing a selectively ~u, ' ' interl~ace membrdne is provi~led which method comprises fixing a substrate onto a supporting mesh to form a substrate membrdne, casting a ~.. Iluulu~ulfonic acid polymer film on the substrate membrdne and curirlg the product so formed.
In a preferred form of the Flrst aspect of the present invention, the protein ofinterest is I ' I by a ~ vith the cellulose binding domain (CBD) of cellulases from bacteria of the genus ~ lar~lnq~ These cellulases have a modularstructure consisting of two or more structurally separate domains. The binding domain functions ' . ' 'y of the catalytic domain and can be chemically or genetically conjugated to other proteins (eg. enzymes) which then bind strongly to cellulose.
Accordingly, the porous protein-receiving matrix of the biosensor disclosed herein preferably comprises cellulose. Under the dlJIJlU,U ' ' solution conditions, the binding of the protein-CBD conjugate to the protein-receivirlg r~atrix in the biosensor comprising cellulose can be disrupted and the conjugate protein eluted from the cellulose matrix. The loading and eluting of the protein conjugate is described in gleater detail l~c~ lvw. The exact nature of the binding mechanism has not been rl~tPrm;nrA, however, the binding to cellulose has been reported to be virtually Pleviously published CBD binding studies indicate that adsorption of the CBD to cellulose was complete within the shortest incubation time feasible under the conditions of the experiment (ie., 0.2 minutes) (Gilkes et al, 19922), but the actual adsorption kinetics are likely much faster.
R~,elu~,h~ l ~ have previously used CBD for the ' ' of proteins and as an affinity tag for the l,...ir~ of ~, ' proteins using cellulose columns (Kilburn et al, 19922l). What has not be~,n disclosed or heretofore examined is the use of the reversible binding properties of the C:BD to proteins in a biosensor context, and in Gilkes, N.R.; Jer~is, B.; Henrissat, B.; Tekant, B.; MiDer, R.C. Jr.; Warren, R.AJ.; Kilburn D.G.
The Adsorpdon of a Bacterial Cellulase And Its Two Isolated Domains to CrystaDine CeLulose. .1. Biol.
Cheln. 1992, 267(10), 6743-6749.
21 l~ilburn, D.G.; Turner, R.F.B.; Col~tmho, I.B.; Din, N.; GiD~es, N.R.; Greenwood, J.M.; Hobbs, J.B.;
MiUer, R.C. Ir.; Ong, B.; Phelps, hl.R.; Ranrz, C.; Warren, R.A.J. Cellnlose Binding Domains:
Applicahons in Bintechnology. prGc. Cellu~or~ 92, In Press.
11 Z353\0266~ 18 . WP - 6 -21 32~9 particular, to an in situ, l~ dl/L, bios~ nsor, capable of r ~ in either a closed-loop or open system, as ~ ' in the present invention.
One of the key advantages of the use of a protein conjugate comprisimg CBD
and a cellulose matrix within a biosensor is the ability to calibrate, measure and replace protein during sampling and analysis pro~edures in the biosensor in a simple, reliable way.
The, ' in the present invention of the use of a protein conjugate comprising CBD, the cellulose matrix in the biosensor and the capability of engagmg this biosensor to hardware enabling the creation of an on-line system aUows for efficient and reliable feedback control for the l-r'- '- of reaction conditions in fed-batch processes. One application of the biosensor of the present invention is the I ~,;al production of . ' proteins.
The production of ~ ' proteins can be optimized by ~ the biomass yield on substrate to obtain a high cell density and then ~ cell specific productivity through high rates of gene expression. H!igh biomass yield on a given substrate can be obtained by controlling ' 1 and ~ the excretion of an inhibitory ' ' through the regulation of substrate levels (Smith and Bajpai, 198522). Due to the importance of glucose as the main carbon and energy source for microbial growth in industrial r, applying the biosensor technology of the present invention to glucose monitoring would be beneficial ill order to operate the IJ;O~ CC~ S under optimum conditions.
In general, a closed-loop control system utilizes fJIIIIa~hJ~ encoded in any number of distinct (either ' I ' or related) signals derived from , that may operate either on-line (ie., within th~ controlled bioreactor l,..~i., t) or off-line (ie., external to the bioreactor environmenQ. The .~JIII.d~i(lll iS prrlc.~PA,P.i~hP,r digitally or analog, according to some specified algolithm with the objective of making a decision regarding d~ ; ' ' in some control variable(s) (eg., the rate of glucose infusion in a fed-batch r~ using glucose as the main carbon source). The invention of this disclosure would provide an impol tant input signal to such a control system. The 72 Smith, M.; Ba3p;~i, R. Fcd-Batch Control of Escherichi~ coli Ferrnenttltion to Hlgh Cell Denrity. proc.
15th ~nn. Piochem. Er,g. ~ymp. 1985, IS. 3444.
112353\0266518.WP - 7 ~

21325~9 availability, reliability and quality of this signal are the particular advantages of the invention.
A very simple, llylJU~Ilt~ ll closed-loop control system based on a single inputfrom an on-line glucose sensor might be configured as follows: (1) the sensor is prepared and installed in the bioreactor comprising the biosensor of the present invention; (2) a particular protein: is ~' '/specified as a set-point . at which the control system is intended to maintain in the bioreactor throughout some specifled interval of the rt (3) in operation, the sensor signal is conveyed via a standard analog-to-digital converter (ADC) interfa~e to a standard p.u~u.l;.,~ derivative(PID) controller , ' ' in the softv~are of a personal computer; (4) the output signal from the controller is applied via a standard interface bus (eg. R~-232) to a hardware actuator system that adjusts the rate of glllcose infusion into the bioreactor fee port. The operating prctgram running on the personal computer would be designed to accept auxiliary r, " regarding periodic off-line protein analyses; this ru---,~lliu-- would be used to check amd update the sensor calibration p~lrameters also as described in the thesis by M.R.
Phelps. If the sensor parameters stray beyond certain specified limits, the contrctl system is interrupted (amd placed t~ tt~ ily in sonne safe stand-by mode) and the sensor is guided through a se~uence of steps,usimg computer-controlled pumps and actuators, outlined in Figure 2. This sequence of steps would take less than 10 minutes arnd would be required only rarely (perhaps not at all) during a F~ i.. R- '-' ' could be carried out either by perfusing the sensor with glucose, standards or by utili7ing the auxiliary ~ iUI~ from a series of off-line analyses, obtained during an interval where the bioreactor glucose, is allowed to vary. This is a very simple example - and an overly simplified description - of a hyl~ull~ al system that could be modified in a variety of ways. It is not the intention of the presen~t authors, however, to detail the closed-loop system any further.
The provision of am on-line system in accordance with the present invention allows optimal control (with respect to the analyte) of the bioprocess monitored by the biosensor and ensures that the biological rnolecule of interest is measured as frequently as required without the risk of of the process. Currently used off-line methods 11Z353\026651S.WP - 8 -~132~9 âre usually too slow to be used in a closed-loop (computer) control system, and it is often difflcult to ensure that samples are not -;" ~~ ~Iy degraded or changed during the , '- ' 'ysis proccdure. The biosensor of the present invention usmg an in situ probe provides, , real-time analysis w;hich is particularly useful for highdensity, fed-batch processes. The on-line biosensor of the present invention allows the complete process of diagnosis, ~ . and rPrAlihrA~i-.n to be perforrned in silu and with the capability of the processes being conducted under computer control. The 1" ' of the biosensor are widespread given the possibility of, ~ ~ ~ CBD to other biological molecules, including varied proteins such that the sensor hardware can be used for monitoring a variety of different analytes. The biosensor of the EJresent invention is in contrast to the probe designs heretofore known and avaiEable in which ~ a and ~ ;. . of the sensor cannot generally be performed without operator ~, The third aspect of the present invention which provides a selectively perrneable interface membrane is very im~portant, as `It represents the mterface between a biosensor and the analyte medium. The E~urpose of the interface membrane is to allow diffusion of the biological molecules of inlterest into the biosensor while excluding interfering species which may be present in the anal~te medium, such as cells, proteins, enzymes inhibitors or other ~ I The interface membrane also provides signiflcant mass transfer resistance which increases the linP~arity of response and the working range of the sensor. The interface membrane of the present invention additionally provides an' v~lc aseptic barrier between the biosensor and the external enviromment.
The interface membrane of the present invention comprises a supporting mesh, a IJ~ "ulro.Lc acid polymer ...~.6.~d substrate and a 1- ~, film of acid polymer. The 1~ film comprises the outer-most layer of the membrane and forms, when the membrane is in place, a l.~ interface between the biosensor and the reaction ~
The use of p~.nuulu~ulfonic acid (tra~ename Naflon) polymer as part of a separate l, ' system, formed in accordance with the present invention, has not heretofore been 1~ . ' ' Nafion has, however, been used in various ways in biosensor 112353\026651S.WP - 9 -2132~69 systems. In United States Patent No. s,a~s2,sso to Rishpon et al., 1,, r~ acid polymer was used as an enzyme matrix amd coated on the biosensor electrode. It is not . ' ' in this patent to prepare a detachable, unitary 1,, r~ - acid polymer membrane in accordance with the present invention.
One of the key advantages of the interface membrane of the present invention is the fact that it may be used irl ~ ; with a wide variety of biosensors and bioreactors. Its use is not limited to the l~iosensor disclosed in the present invention.
In summary, all of the practical concerns of using an in situ biosensor probe, such as in situ ciPrjl' ' ''"y, long-term alld short-term stability, adequate measuring range and membrane fouling which has thus far prevented the widespread application and~ ,;d~ iOII of this approach, have been overcome by the ~ ble biosensor probe of the present invention. In addition, an~l equally , i '~/, the selectively permeable interface membrane of the present invention provides:
(1) a ~ c diffusion membrane that permits the analyte (and oxygen) to enter the sensor, while excluding potential interfering species that could foul the electrode or denature, or inhibit the activity of, the biological molecule of interest;
(2) an du~u~ldvdbl~, aseptic barrier that ensures bioprocess ,. , ' lity;
(3) an interfacial surface that is sufficiently l~ lP so that the membrane itself does not become fouled and experience a change in 1....,.. ~r~ which would ultimately result in ~irift of the sensor calibration; and (4) a membrane which is sepalate and distinct from the other biosensor 11Z353~0266518.WP - 10-~RIEF DESCRIPTION OF THE DRA'i~INGS
The ~ , drawings, which are , ' in and form a part of the , illustrate ' " of t~le present invention and, together with the .' p~ serve to explain the principles of the invention. In the drawings:
Figure I is a schematic ~., of a biosensor based on CBD-' ''' ' en2ymes;
Figure 2 is a process flow diagraun .1~ g the principle of operation ofa f~ monitoring system using a l~ .dble biosensor probe;
Figure 3 is a schematic ~ c;llLdlio~ of a pl rl u~dll acid polymedcellulose triacetate interface membrane;
Figure 4 is a schematic ~ of the assembly of an interface membrane onto a biosensol probe;
Figure S is a schematic l~)I~ iUll of a biosensor probe showing a reagent flow system and . ,. - '~',.-1-, Figure 6 is a graph ~ a . . ~ of biosensor ~ r~finn time in Luria broth of an interface membrane system of the present invention and other Figure 7 is a graph ~ , normalized calibration data for a biosensor in PBS and Luria broth (LB) llsing the interface membrdne of the present invention both before and after autoclaving;
Figure 8 is a graph 1~l. ~ the normalized calibration data for the prototype sensor in PBS an,d Luria broth (LB) using Nafion membrane before and after ' ~ ~
11 Z353\02665 1 B . WP

2132~69 Figure 9 is a graph .~..,~c ~ the effect of ' , ti on the biosensor signal at steady state;
Figure 10 is a graph ~ ~ the effect of p~ on the biosensor signal atsteady state;
Figure 11 is a graph ~ v the effect of dissolved oxygen tension on the biosensor signal at steady s'tate; and Figure 12 is a graph showing medium glucose ~ measured by both the biosensor of the present invention and the Beclcman off-line glucose analyzer during fed-batch cultivation of E~. coli in minimal medium.
DETAILED DESCRIPTION OF 'l'llE ][NVENTION
A. RL ' ' Bioserlsor Probe In accordamce with the first aspect of the present mvention, the basic design ofthe biosensor comprises an indicating elec,trode, a porous protein-receiving mat~ix and a protective mterface membrane, all i~- "~ l into a stainless steel probe for insertion into the bioreactor. Inlet and outlet conduits in the probe body allow for perfusion of the protein-rf~ceiving matrix with the protein ,~onjugate solution and the elution buffer as described 1~ ' ... The indicating electrode can be raised to permit complete perfusion of the protein-receiving matrix and then ll:)wered into contact with this matrix to facilitate substrate ~ This is also describ~ d further h~
In a preferred form, the protein-receiving matrix is a porous cellulose matrix.
Further, it is preferred that the protein conjugate comprises the CBD of ~ ln-~ most preferably, C. finli.
The ( .If ' of the prererred form of biosensor in accordance with the present invention are illustrated sfhPn~if~lly in Figure l. The cellulose matrix for 112353\026651 S . WP - 12 -2132~9 li7~fi.7n of the pnotein-CBD conjugate is enclosed im a chamber formed by the pn7be body, the interface membrane and the inclicating electnode unit. The indicating electn.de, as discussed above, can be raised and lowen~d via a threaded shaft. When necessary, the indicating electn.7de unit is raised and the pn7tein is eluted by perfusing the ceilulose matrix with a suitable elution buffer (eg. distilled wate~, sodium hydn.7xide or guanidine l~ydl~ 3, followed by perffision with a soluble protem-CBD conjugate to n~place the ' "' ' pr~tein. In operation, the indicating electn.7de is lowen d such that the ceilulose matrix is ~Wi~il~i between the interfa~e membrane and the indicating electr~de. The protein-receiving matrix is fiiled with electn.71yte which ecluilibrates with the elect~71yte of the external analyte solution.
Figune 2 outlines the pneferred principle of operation of the biosensor system for continuous r ' " " ' g In one application of this invention, the pnotein to be conjugated to CBD is an en~7yme. The entire pnobe, exclusive of the en~7yme, is sterilized in situ during steam f^ili7~fi~7n of the fermenter and contents. Following '~ , the en. yme is loaded into the sensor by perf~sion of the cellulose matrix with soluble en~yme-CBD conjugate, nesulting in ' -' '1i7~fi~7n of the en~7yme (behind the sterile barrier provided by the interface membrane) via attachment of the CBD. The sensor is calibrated during the addition of the en7yme substrate to the fermenter. Periodically during the ' ~ , the sensor calibration can be checked by analysis of samples of the bnoth or, by pumping internal calibration standards thl~ugh the en~7yme chamber.
If the internal ca'iibration c'heck indicates that the biosensor pclro....dl~e has . ;. ., ' ! to an " ,~ " l~il degree du~ to i~ ~iv~.lio.l of the en.7yme or electn7de fouling, then the en_yme can be eluted by perfusion of the ceilulose matrix with elution buffer. The biosensor is then reloaded with fnesh en7~/me-CBD conjugate as befone, the sensor is recalibrated using the internai calibration standands and the ' ' " monitoring continues. The sensor caiibration can be verified, if desined, by, , ' of the sensor/biosensor output with substrate an~llysis of a sample of the f~ ' ;. . broth. Each of these steps of sampling, caiibration an~i adjustment may be conducted using a closed-loop system.
11 2353\0Z665 18 . WP - 13 -21~2~69 The proteins of intere$ may be conjugated to CBD by any known genetic or chemical method. The methods disclosecl in United States Patent 5,340,731 to Kilbunn et al.
are i." ~ 1 herein by reference. In a ptefen-ed fonm of the present invention, the CBD
is chemically conjugated to the enzyme u'sing glutaraldehyde. It is to be clearly I ' -~1, that the present invention is not limited tc~ a particular method of, ; ~
The operation of the biosellsor of the present invention is bascd on the U~ detection of hydrogen peroxide produced as a by-product of the enzymatic oxidation of a substrate by its enzyme catalyst. A p, is used to maintain the necessary bias potential for the ~ u. ll~.l.icdl oxidation of hydrogen peroxide at a noble metsl indicating electrode and also conve]~ts the substrate-dependent electrode cunrent into a usable output voltage. In a prefened fonn, the enzyme is conjugated to CVD, and the conjugate so fonmed is pumped into the porous protein-receivmg matrix, which is preferably a cellulose matrix. The cellulose matrix ~is in close proximity to the surface of the indicating electrode, which is preferably platinum, gold or carbon. Most preferably, the indicating electrode is platinum. The substrate and c~xygen diffuse into the enzyme layer, and react with the enzyme to produce the oxidized fonm of the substrate and hydrogen peroxide. Some of the hydrogen peroxide diffuses to the ~llectrode, where it is oxidized c;lc; llu~ h~ lly, liberating oxygen, protons and electrons. Thus, the cun-ent measured at the indicating electrode is related to the ~ o~ the analyte in the bioreactor solution.
A peristaltic pump was used to pump the ~ JIU,UI' ' reagent solutions through the inlet conduit into the biosensor probe body. In a prefened ~ " t, a, t`il~liof three-way valves was used to select th~ reagent to be pumped. The three reagent reservoirs contain (I) the intennal electrolyte and washing buffer, (2) the substrate (for example, glucose) standard, for intennal calibration, and (3) the elution buffer and, rf necessary, the electrode ~,lr., may be l~ " ' by anodizing/cycling the electrodein an d,u~)lulJI' ' electrolyte. This pump system may be interfaced with a personal computer. Figure 5 is a schematic .~., rm of a redgent flow system, however a reservoir for the protein - CBD conjugate is not shown. If an intemal calibration standard is required, a third multi-port valve can be configured enabling a fourth reagent reservoir to be connected into the inlet conduit stream. l~ltennatively, a single, six-way valve could 112353\0266~18.WP - 14-2~32~6~
~- ' all the necessary solutions to be switched into the flow stream. A range ofelectronicdlly-activated multi-port valves ,Ire, , "y available. Adapting the biosensor of the present invention to a closed-loop ~,ystem is well within the skill of an artisan in computer science, and does not require ally inventive ingenuity beyond this disclosure. The valve systern of the redgent reservoirs carl be actuated by computer-controlled electric or pneumatic actuators (not shown) which allows, with the interface of the pump system with a personal computer, complete automated, on-line, closed-loop control of the redction conditions. In addition, the operdtion of rdising and lowering the internal electrode assemb1y, instedd of being performed mallually by turning a knurled knob on a threaded shaft, may be conducted within the scope of the present invention by a computer-controlled stepper motor. The sensor output is monitored by a personal computer rdther tham a chart recorder and the ddtd used in computer algorithms for sensor ~lihr~fion self-didgnosis and Y g~ as well as feedback control a lgorithms for bioprocess control.
It will also be d~ ' ' ' that the biosensor disclosed herein is not limited to use with any pa~ticular enzyme catalyst. Any other enzyme redcting with specific substrates to produce hydrogen peroxide for generating an electrical signal functionally related to the presence of such specific substances may be used. Suitdble enzymes include, but are not limited to, glucose oxiddse, L-lactdte oxicldse, alcohol oxiddse, galactose oxiddse, cholesterol oxiddse, pyruvate oxiddse, uricase, aldeh~lde oxiddse, xanthine oxiddse, choline oxiddse, acetylcholine esterase, L-glutamate oxida~e, L~ IdD~, d-~-ylO"' ' , imverfase and mutarotase.
In addition, the biosensor d~isclosed herein is not limited to Cl~llu~ ;dl detection ' - ~or example, optical ' may be used, in which case the conjugate may comprise proteins other th~n enzymes.
B. Interface r ~ ~ _ In accordance with a second aspect of the present invention, an interface membrane is provided which comprises a supporting mesh, a ~.lluu-uDuhfonic acid polymer ,, ~ ' substrate and a ~ . - Fllm of ~ 1uu.uDulfonic acid polymer. The 112353\0Z66518.WP - 15-2132a69 roduct so formed is a roughly lamellar structure with a base support, an ilmer f Fullic acid polymer layer and a '~ ln~.u.u~ulfonic acid polymer film that is the outermost layer and forms, when the membrane is in place, an ' v~lc l,i~ interface between the biosensor and the bioreactor Preferably, the substrate is selccted from the group comprising Illl.,rill.,.~;.,,~
such as cellulose acetate(s) (illcluding cellulose triacetate), polysulfone, andAmicon PM series non-cellulose ' , dialysis ' such as those of cellulose, cellulose acetate(s), , " '- , p~ly(vh.yluhlOIid~) and various other porous or continuous (ie., non-porous) polymer ' or films. Most preferablythe substrate is cellulose triacetate.
The supporting mesh may l)e any suitable material able to withstand autoclave t;s. Preferably the supporting mesh is a metallic screen (eg., stainless steel, aluminum, copper, silver, gold, etc.), a polymeric screen (eg., poly(v-.lyl;hlolidt;), poly(~l~lluulucilllyyl~ ), puly~yl1llc~ polyuLd~ , or other fibrous filter material such as cellulose.
A schematic It;~ll ' '' of the preferred interface membrane is plrovided in Figure 3. The assembly of the membran~ onto a biosensor is shown schemically in Figure 4.
There are several importanl features of the interface membrane of the present invention:
1. Selective ~ ily - Nafion membrane/films are essentially non-porous (so there is no "pore size" per se). The perm-selectivity derives primarily from the , ' ' Oy of the insoluble (ie.,cast) membrane/film which, in turn, derives from the casting conditions.
The other property that contributes to the perm-selectivity is the charge; Nafion is negatively charged, and he~ce tends to reject anionic species. The permeability properties of the invention have been found to be almost idleal for r " monitorirlg 112353\0266518.1~lp ~ 16-2132~69 2. P~i~ , ' ' 'y - This is a~ very ambiguous temm, but in the context of bioredctor ,, it basically means that: 1) cells tend not to stick to the material; 2) proteins or other biological , 1 1PS do not tend to stick to the material; 3) the materidl does not reledse chemicals (eg., solvent residues or polymer f"f; ' products) or redct with medium ~ ~ to foml products that are hammful to cells or harmful as a of the final bioprocess product. The interface membrane of the present invention has been proven to be bio, . ' ' 3. S~ y - the interface membrdne must be able to withstand steam i7~fi,~n (ie.: ' v g), since this i~i the ~ .y industry's only acceptdble ~qri1i7 ~ n protocol for bioreactors. Th~ membrane of the present invention is ' vAhl~.
In other d~ ; , however,other -~ i.. protocols may be . .'~ (eg-, ethylene oxide gas, radiation, various sol~ents). It is expected that the interface membrane (of the invention) could also withstand th~,se other treatments (with the ~Q~ of ethylene oxide).
4. Del~ Ldbili~y - In other A~ of Nafion, the membrdne was cast directly over the electrode (and other overlying ' /fiLms) without any intervening structural material to pemnit separdtion of the interface membrdne system from the rest of the sensor Previously, it was not .1 I to achieve this separdtion without degrdding the p~ l r~ of the ensemble system. The present invention su~ >r~ "y provides this advantage.
It is understood that the interface membrane described herein is a separate product from the biosensor described l~ti., b~,ic. The interface membrdne is a separdte unit which may be suitdbly engaged with any biosensor system. It may be produced and sold as a disposable "cartridge" for use with any ensemble biosensor system including the biosensor systems described herein.
112353~3266518.WP - 17 -21 32~69 C. Method of Preparing Interface 1' ~ ~ ~
In accordance with the third and final aspect of the present invention, a method of preparing a selectively permeable interface membrane comprises fixing a substrate onto a supporting mesh to form a substralte membrane, casting a p~ r~ u~ acidlpolymer film on the substrate membrane and curing the product so formed.
As an additiûnal step, an dLl~l-Up ' ' hardware fLxture must be selected or designed which permits the assembled m~mbrane system to be attached to a biosensor. A
preferred form of the present invention tble hardware fLxture comprises stainless steel or brass along with plastic, ' ~/dl l_ . . configured as shown schemically in Figure 4. Asupporting mesh is selected from those disclosed hereinabove. In the most preferred form, the supporting mesh is a stainless steel woven mesh with less than 100 um pores, although the pore size is relatively I . an~i may range up to a few mm. An ~ UIUIJI' ' substrate materi~l is selected from the gr_up disclosed 1~ . In a preferred form, this material is cellulose triacetate, most pref~rably, of the kind and more foliage provided commercially as Gelman Metricel G.A. - 8 membrane ~llters with a maximum pore size of 0.~ um. This pore size ensures that an a~;eptic barrier is maintained even in the event of a breach in the Nafion membr~me. It is to be understood, however, that larger pore sizes may be successfully employed, but may not b~ as successive in the event of a breach of the Nafion membrane.
In order to construct the interface membrane, the supporting mesh must be cut to fit the hardware fixture. The supportillg mesh is then fastened to the hardware fixture such that the mesh work forms a screen covering a portal that, once in place in an assembled probe, locates the mterface membrane at the required position and orientation with respect to the other 1- . The orientation of these I . is best shown m Figure 4. An d,U,UlU,UI srze and shape of substrate m;~terial is then cut to fit over the supporting mesh work and fastened such that the substrate material is held in contact with the supporting mesh work over the entire area of the portal.
112353\026651 S . WP - 18 -213236~
The casting and curing of Ihe membrane may be performed as follows. A
stock dispersion of Nifion polymer is prepared, preferably from a . ~;olly available 5%
solution (aldrich chemicals). In a preferred I ' ~ ' t, this stock dispersion is 0.5%
Nafion polymer diluted from the as-recei~ed stock in a 1:1 solution of isopropyl alcohol and distilled-deionized water. It is to be . ' ~ i, however, that the Nafion: may vary higher than the as-received 5 %, although as the increases the sensor response may be slower and lower in aml~litude.
One or more aliquots of the above dispersion is then applied to the surface of the substrate membrane prepared as described above so that, when the interface membrane system is in place in a biosensor, the disFlersion will be exposed to the analyte medium. In a preferred ' ' t, aliquots ranging firom 15 uL to 300 uL rnay be used and applied in one or two aliquots such that the entire sllrface of the substrate is covered. It is to be lln~110rctnofl however, that larger aliquots may be used depending on the size of the substrate area to be covered and the ~ ~()f the casting dispersion.
The interface membrane is thereby formed by allowing the casting dispersion solvents to evaporate. In a preferred; ' ' t, this is carried out at room i . c and ~ ;f, pressure with the memblane/substrate/fixture remaining stationary in an orientation where the planar surface of the cast membrane is kept level in order to yield a uniform membrane/film over the surface of the substrate. It is to be IPr~tnofl however, that CllVil~ I conditions may be varied over a r-"~.f~ range.
The interface membrane S(l casted is then allowed to cure. Preferably, the curing time is at least 24 hours, most preferably even longer prior to ' Vil.~, or any first exposure to a solution c ..v u~ Preferably curing is allowed to proceed at room t~ lul~ and :~tr'l~ . ' pressure for at least two to three days.
112353\0Z66518.Wp - 19 -2132~69 ~LE 1: Synthesis of Protein Coluugate of Glucose Oxidase and CBD
The enzyme glucose oxidase (GOx) was conjugated with the cellulose binding domain from the cellulase O' from (~71'7 fimi. More specifically, a chemically conjugated GOx-CBD protein was synthesized using the bir, I cross-linking agent glutaraldehyde.
Glucose oxidase (EC 1.1.3.4) was Type X from A. niger (Sigma Chemical Co., St. Louis, MO) and was used without further I ili~lhJl.. The cellulose binding domain from C fimi ~ O (CBDC~,) was harvested from ~ . coli and purified according to methods described elsewherc (Ong et al., 19933). CBDC~X antiserum was produced in rabbits (Whittle et al., 19822"). Grade I glutaraldehyde, 25% aqueous solution (Sigma Chemical Co.) was used as received for the GOx-CBD ~ ~ The syrlthesis, and storage of the GOx=CPD conjugate was carried out in 50 mM potassium phosphate buffer, pH 7. Cellulose powd~r (Avicel, type PH101, FMC T I Food &
r Products, Cork, Ireland), ~vashed with distilled water and phosphate buffer, was used for ~,ulir of the GOx-CB]~ conjugates. All other chemicals were analytical grade and used as received A 10 mg/mL solution of g]ucose oxidase in 50 mM phosphate buffer, pH 7, was first activated with ' ' ' ' y.' ~ a 50-fold excess of glutaraldehyde for each amino group (lysine residue or N-terminus) on the enzyme (Gibson and Woodward, 199225). 40 ,uL of O ' ' ' ~d~ (25% aqueous solution) was added per mL of GOx solution. After incubation overnight at 4C, the excess ~' ' ' ' Jlle removed by dialysis ~3 Ong, E.; Gilxcs, N.R.; Mi!ler, R.C. Ir.; Warren, R.A.J.; Kilburn, D.G. The t~ Domain (CBDC,~) of ~m Exoglucanase From Ce~lulomonas fimi: Produc~on ra Escherichia coli And f" ' ' of the Polypeptide. Biotech. 7~ioeng. 1993, 42, 401-409.
24 Whittle, DJ.; Kilburn, D.G.; Warr~n, R.A.J.; Miller, R.C. k. Molecular Clotrdng of a Cellumonasf mi Cellulase Gene m 7~schenchia coli. Gene 1982, ~7, 139-145.
Glbson, T.D.; Woodward, J.R. Protein StabiDxation in Biosensor Systems. Irl Biosensors & Chemica7 Sensors: Optimizing Perfonnance Through Polymeric Materials; Edelman, P.G., Wang, J., Eds.; ACS:
Washington, DC, 1992; p~ 44-55.
112353\0266518 WP - 20 -2132~6~
ersus phosphate buffer using SpectralPol 2 dialysis tubing, MWCO 12-14 kD (Spectrum Medical Industries Inc.,Los Angeles, CA, U.S.A.). Dialysis was performed for 24 hours versus 0.25 L of 50 mM phosphate buffer per mL of activated GOx solution. The dialysis buffer was changed after 12 hours. CBDC~r was added in a 1:1 molar ratio based on glucose oxidase and incubated overnight at 4C. 0.693 mL of 11.2 mg/mL CBDC~r, stock solution was added per mL of activated GOx solu~ion. The CBD binds to the activated GOx either via the single Iysine residue present or via the N-terminus of the CBD polypeptide (Coutinho et al., 199226; Hansen and Middelsen, 199127) according to the net reaction:
GOx-NH2 + OHC-(CH2)3-CHO + H2~-CBD ~ GOx-N=CH-(CH2)3-CH=N-CBD (3.1) Excess CBD was removed by buffer exchange with 50 mH phosphate buffer in an Amicon stirred ceU using an Amicon l'M30 (non-cellulose) membrane (Amicon Canada Ltd., OakviUe, Ontario). Buffer exchange was performed until the absorbance (A28Q) of the filtrate versus phosphate buffer d~ u,~ll~(l zero. The GOx-CBD conjugate was then purified in a single step by binding to Avicel, usirlg at least 150 mg of Avicel per mL of activated GOx solution. The conjugate was eluted from Avicel by washing twice with 0.1 M NaCI in 50 mM phosphate buffer, and then one wash with de-ionized, distilled water. The final wash was saved and made up to 50 mM in potassium phosphate using 0.5 M phosphate buffer.
The purifled GOx-CBD conjugate was stored in 50 mM phosphate buffer at 4C. For longer term storage, the conjugate was stored bound to Avicel in 50 mM phosphate buffer at 4C
and eluted when required. Batches of 1 rnL and 10 mL volumes of GOX-CBD conjugate were prepared using this protocol.
In the, ; ~ protocol, distilled water was used for eluting the purified GOx-CBD conjugate from Avicel. The use of 8 M guanidine HCl and 1 M NaOH for 26 Cou~o, ~.s.; Gilkes, N.R.; warr. n R.~.; Kilburn, D.G.; Miller, R.C. Jr. Tbe Binding of Cellulo~nonasJlmi r ~ _ c (CenC) to Cellulose and Sephadex is Medi~ted by The N-Terrninal Repeatn. Molec. ~Icrob;ol. 1992, 6, 1~43-1~52 27 Har~sen, rl.H.; Mil~kelseD, H.S. r ~ by The ~ - - -, Procedure. An nve3dgation of The Effects OrReducing The schiff-sases Generated, as s~sed on Studying The of Glucose Oxidase LO Silar,lzed Corltrolled Pore Glass. Anal. Ler~. 1991, 24(S), .
112353\0Z66518.Wp - 21 -elution, as suggested by Kilbum et al. (1!~92)28, caused i~lr~V~ ;bh~ IOSS of en7yme activity.
Using distilled water for elution, the soluble ~ , " product was found to retain greater than 60% of the activity of the original, l ; ~ ' en7yme. In addition, the GOx-CBD
conjugate retained the binding affinity of the CBD for cellulose, and exhibited GOx activity when bound to cellulose . ~ to that of GOx -' 1i7~1 by other techniques (Harrison et al., 1988)29. The conjugate could be adsorbed and desorbed from cellulose repeatedly, and could also be dehydrated for storage at room ~elll~Jr,l~ul~, then lGC, with PBS, pH 7.4, without a significant loss of activity. Non-specific binding of ; ~ ' GOx to Avicel or ~, ' cellulose was found to be ~ ~ The specific GOx activities of various samples of soluble GOx-CBD conjugate are show in Table l to illustrate the typical stability of the conjugate over tirne. In glucose biosensor t;.i using the GOx-CBD conjuga~e, it was found that the conjugate retained sufficient activity to be useful after up to 2 storage (when stored bound to Avicel).
A potential alternative to cl1emical ,~ would be to develop an genetic construct that yields an active GOx-CBD fusion protein. Apropos of this,the gene encoding the glucose oxidase protein of Aspergillus niger has been cloned and expressed in yeast (Frederick et al., 1990i; Dcl3a~ t~lh,l et al., 19913l). The main advantages of a genetically engineered colnjugate would be greater uniformity and y of the conjugate reagent, as well as "",~ of large-scale conjugate production. A possible disadvantage, however, might be that the si7e and GOx-CBD ratio of the resulting conjugates could not be as easily tailored when compared to the chemical ~ ;, method, which may limit the efficacy of the GOx-CBD reagent.
28 Supn~ at 21.
29 Har ison, D.J.; Turner, R.F.B.; Ba tes, H.P. ('' of r~ ~ ~ Acid Polymer Coated Enzyme Electrodes And a Miniaturized Integrated Potenhostat For Glucoso Analysis in Whole Blood.
AnaL Che~n. 1988, 60, 200Z-2007.
Fn derick, K.R.; Tung, J.; Emerick, R.S.; Masiarz, F.R.; Chamberlam, S.H.; Vasavad~, A.;
Rosenberg, 5.; Chakraborty, 5.; Scllopter, L.M.; Massey, V. Glucose Oxidase From ~sperg~ nlger:
Clonmg, Gene Seauenco, Socrehon From ~r . cere~iae and Kmetic Analysis of a Yeast-Derived Enzyme. J. of Biol. Chem. 1990, 26~, 3793-3S02.
31 De Baetselier, A.; Vasa~-ada, A.; Dohet, P.; Ha-Thi, V.; De Beukelaer, M.; Erpicum, T.; De Clerck, L.; Hanoher, J.; Rosenberg, 5. Fennentadon of a Yeast Producnlg A. niger Glucose Oxidase: Scale-Up, Purificadon and (~1 of The Rocombinant Enzymo. Bio/Technology 1991, 9, 559-561.
112353\0266518 WP - 22 -2132~9 ~able 1: Specific activity of various samples of soluble GOx-CBD conjugate. The specific adivity was calculated from the results o~` GOx activity and total protein assays. Tbe relative adivity was determined compared to . ; ~ ' GOx.
Sample Time in Conditions Specific Relative Storage Activity Activity (Ulmg) Ui~; " ' GOx 0 Flreshly prepared in 50 mM 118 100%
phosphate buffer, pH 7 GOx-CBD Conjugate 3 weeks Purified on Avicel; eluted by 68 58 %
Batch #2 distilled water after I week in storage GOx-CBD Conjugate 5 weeks Unpurifled 91 77%
Batch #3a GOx-CBD Conjugate 6 weeks Unpurified 72 61%
Batch #3b GOx-CBD Conjugate 9 weeks Purifled on Avicel; eluted by 30 25%
Batch #3a distilled water after 7 weeks in storage EXAMPLE 2: Reversible ~ ` ' ' of Enzyme in Biosensor Using CBD
Technology In this glucose biosensor c-~nC~n~ te~ from a p1atinum rotating disk electrode (RDE) with a cellulose matrix for enzym~ .m via the CBD, a GOx-CBD
conjugate was used. Using glucose standlrds, the biosensor is calibrated repeatedly in PBF
during multiple cycles of loading and elution of the GOx-CBD conjugate in order to simulate the periodic ~ of a glucose biosensor duri~g a r~. ".. ~ .... period.
El~l~ data were obtained using a Pine AFRDE~ bi-put~ iC~ and a Pine AFMSRX analytical rotator (Pine Instrument Co., Grove City, PA, U.S.A.). A Pine AFMDI1980 0.5 cm diameter platinum rc)tating disk electrode was fabricated in-house from 1.0 mm diameter platimum wire (Aldrich, Milwaukee, WI, U.S.A.) bound in flint glass tubing (5 mm outside diameter) with epo~:y (Chemgrip, Norton Co., Wayne, NJ, U.S.A.).
A 25 mm x 25 mm, 52 mesh platinum wire gauze (Aldrich) was crimped onto the platinum 112353\~266518.WP - 23 -2132~69 ire to increase the surface area of the counter electrode. The reference electrode was a saturated calomel electrode (SCE) (Fishel Scientific, Ottawa, Ontario, Canada, No.
13-620-52). A Kipp & Zonen Model BD~I 12 strip chart recorder (Kipp & Zonen, Delft, Holland) was used for recording sensor output, and a Kipp & Zonen Model BD91 ZYY't recorder was used for recording cyclic vn' ~, The platinum working electrode (RDE) was prepared by polishing with 0.05 ~m alumina and rinsing with distilled water.
Tl ' ~S~ before each w-~, t, the working electrode was cycled in quiescent 0.5 MH2SO4 from -0.26 V to +1.2 V at 100 mV/s for 10 minutes, anodized at +1.8 V for 10 minutes, then cycled again until a stable cyclic ~ ' O was obtained.
The rotating disk electrode was modified using a retainer such that different cellulose matrices could be held in place over the platinum disk. The retainer was either a Teflon ring or a rubber ring (formed by ~:utting a transverse slice off the end of a piece of Tygon tubing) sized to fit tightly over the RDE. Three different cellulose basedwere used during the .,~ tal biosensor trials: I) Whatman No. I qualitative filter paper (Whatman T ' Ltd., Maidstone, England), 2) Spectra/Por 2 ~ el~.~ed cellulose dialysis membrane, MWCO 12-14 kD (Spectrum Medical Industries Inc., Los Angeles, CA, U.S.A.), and 3) nitrocellulose protein transfer membrane with 0.45 ,um pores (Schleicher & Schuell, Keene, NH, U.S..~.). The GOx-CBD conjugate was loaded by immersing the electrode assembly in a so]ution of GOx-CBD conjugate, followed by a single wash with 50 mM phosphate buffer.
The RDE was rotated at 5Cl0 rpm and l ' at +0.7 V versus SCE in a 250 mL beaker with 100 mL of electro]yte. The electrolyte used in all biosensor was phosphate buffered salin~ (PBS), pH 7.4, ,u = 0.2 M (0.1 M NaCI, 5 mM
NaH2PO4, 30 mM Na2HPO4, preserved with 1 mM EDTA, and S mM sodium benzoate).
The sensor was calibrated in PBS by recording the steady-state RDE current in response to a series of aliquots of glucose standard (0.1 M glucose in PBS). The sensor response to decreasing l~ f~nC of glucose was recorded by removing aliquots of the cell electrolyte and replacing with fresh, glucose-free PBS. Elution of the GOx-CBD conjugate was achieved by immersimg the electrode assembly in the elution buffer, followed by a series of washes with 50 mM phosphate buffer only before reloading fresh GOx-CBD conjugate.
112353\0266515.WP - 24 -2132~69 ~ree different elution buffers were used: I) distilled, deionized water, 2) I M NaOH, and 3) 8 M guanidme hydl~ ' ' ' (99+ %(CI) from Sigma Chemical Co., St. Louis, MO, U.S.A.), prepared in 50 mM phosphate buffer and filtered through Whatman qualitative filter paper. All chemicals were analytical grade and used as received.
EXAM~ 3: R~ ' '~ Glucose Biosensor using CBD Technology An Ingold CO2 probe was lJsed as the basis for the glucose biosensor. The Ingold probe was designed to withstand ~ t~ UluD from 20 - 125C and pressures from 0 - 2 bars. In its original ,~,..riC,.. .li.~.~ th~i CO2-permeable membrane was silicone rubber reinforced with a stainless steel mesh and a nylon net, mounted on a sterilizable plastic membrane ca~tridge with a stainless steel sleeve and a silicone rubber washer. The internal electrode was a glass pH electrode. The pH electrode was attached to a threaded shaft and could be raised from, and lowered to, the silicone membrane by turning a knurled knob at the end of the shaft. The probe body conltained inlet and outlet tubes used for mjecting electrolyte and pH calibration standards iulto the electrolyte chamber and was equipped with the necessary fittings for insertion into tn~ 25 mm side-ports of Chemap fermenters.
The basic hardware of the ~ngold C07 probe had many of the features of the proposed glucose biosensor hardware described in the schematic diagram of Figure 1, made available in a convenient, fermenter compatible package. Several ~(ljfi~tjonc were performed to convert the Ingold probe to the prototype glucose biosensor:
1. A cellulose matrix was . ' into the electrolyte chamber.
2. The silicone membrane was replaced with a custom designed electrode assembly with three electrodes for , ~Illlti~, operation.
3. The internal pH electro~e ~as replaced with a custom designed electrode assembly with three electrodes for . ~lllltilliu operation.
11Z353\0266518.WP - 25 -2132~69 4. To fit the new internal electrode unit into the probe body, a custom designedadapter was required to mount the electrode assembly onto the threaded shatt used for raisirlg and lowering the internal pH electrode.

5. The syringes used for injecting electrolyte and pH electrode calibrant into the CO2 probe were replaced with tubing, valves, and a peristsltic pump.

6. A rPr1~P .hlP interface (dialysis) membrane was prepared and attached to the probe body.
The cellulose matrix was a disk of Whatman No. l qualitative filter paper (Whatman T ' " ' Ltd., Maidstone, ~ngland) cut to ~ " ly 7 mm in diameter.
The cellulose matrix was ~ dwi~h~i bet~veen the internal electrode and the glucose-permeable outer membrane.
The internal electrode assembly was designed and built in-house. The three electrode unit had an outside diameter of 6 mm and consisted of a Pt indicating ele~de, Pt counter electrode, and Ag/A.gCl reference electrode. The Pt indicating electrode was a l.0 mm diameter platinurn wire (Aldrich, Milwaukee, WI, U.S.A.) in a glass shroud made from 4 mm O.D. flint glass tubing. One endl of the glass tubing was partially closed by melting the glass over a Bunsen burner, such that the inside diameter was just greater than l mm. The platinum wire was then bonded into the glass usirlg Chemgrip epoxy (Norton Co., Wayne, NJ, U.S.A.) to form a seal. The glass-shrouded platinum wire was ground and sanded with ~u. ~ ;vtily finer grades of sandpaper, followed by polishing to a mirror finish with 0.3 ,um and 0.05 ~4m alumina, leaving exposed a circular platinum disk with a surface area of 0.785 mm2-The counter electrode and leference electrode were l.0 mm diameter Pt andAg wire, respectively (Aldrich, Milwaukee, WI, U.S.A.), tightly coiled around the glassshrouded indicating electrode. The surface area of the counter electrode was much greater than the surface area of the indicating elec.trode to ensure that the area of the counter electrode was not charge transfer limiting. The Ag/AgCl refererlce electrode was fabricated 112353\0266518.WP - 26 -2132~6~
~by anodr~ing the Ag wire in the presence of Cl-. The Ag wire was L ~' at +0.2 Vversus SCE for 8 hours, using a Pt wire counter electrode and an cl~lu~ l cell containing 0.1 M KCI.
Modem cable was used for the electrode leads, because the shielding of the individual wires in the modem cable allo~ved the reference electrode lead to be shielded from the other electrode leads. The electrode leads were attached to the electrode wires using gold crimps and soldered.
T ' ' ~ before use, the bare Pt indicating electrode was cycled in quiescent 0.5 M H2SO4 from -0.26 V to ~1.2 V at 100 mV/s for 10 minutes, anodized at +1.8 V for 10 minutes, then cycled agaill until a stable cyclic v-l ~" was obtained.
For some ~ , described below, the indicating electrode was coated with celluloseacetate usirlg a - ' ~ nn of the method described by Wang and Hutchins (1985)32. A

7.5 ~L drop of 2.5% cellulose acetate (BDH Ltd., Poole, England) in a 1:1 solution of acetone and ~y~ ' ~ (stirred for 12 hours) was applied to the mdicating electrode and allowed to dry overnight. The indicating electrode was cycled and anodized before coating with cellulose acetate A stainless steel adapter wa.s used to mount the internal electrode u nit to the electrode shaft of the probe body The flange on the adapter was designed to meet a shoulder inside the probe body to prevelnt the electrode shift from bemg inserted too far into the probe body. An O-ring was usedl between the adapter and the inside wall of the probe body to maintain a sealed internal chamber. An insulating wrap of black electrical tape was used between the electrodes and the adapter. Silastic adhesive was used to bond the internal electrode unit into the adapter an(l provide a liquid seal. The distance from the tip of the internal electrode unit to the flange on the adapter was set at 35 mm, such that the indicating electrode would corltact the outer membrane when fully lowered. It was also necessary to bore the minimum inside dia neter of the probe body from 8 13 mm to 9 55 mm.
. . , _ ~2 Wang, J; Hutchins, L.D. Thin-Laycr r- Detecto~ Wi~ a Glassy Carbon Electrode Coated wi~ ~ r~ H~ ~d Cellulosic l~ilm. Anal. Chem. 1985, 57, 1536-1541.
112353\02665t8,Wp - 27 -2132~69 ~Reagent flow system:
A Gilson Minipuls 3 perist~ltic pump (Gilson Medical Electronics, Middleton, WI, U.S.A.) was used to pump the ~ ,1,. reagent solutions through silicone tubing into the probe body. A ~ ' of three-way valves, shown in Figure 5, was used to selectthe reagent to be pumped. The three reagent reservoirs contained 1) the internal electrolyte and washing buffer, 2) glucose standard, for mternal calibration, and 3) the elution buffer.
The internal electrolyte and washing buffi3r was PBS (0.1 M NaCI, 5 mM NaH2PO4, and 30 mM NaH2PO4, pH 7.4, ,~ = 0.2 M, preserved with 1 mM EDTA and S mM sodium benzoate). The elution buffer was PBS ~ith 8 M guanidine hydrochloride.
The protocols used for loacling and eluting the GOx-CBD conjugate are listed below. Flow rate was calibrated agaimst ]?ump speed, and the volumes of the different segments of the flow system shown in Fi~ure 5 were determined by measuring the time required for a fluid to pass through the different segments at a given flow rate. The time required to wash all unbound GOx-CBD conjugate out of the enzyme chamber after the enzyme loading cycle was determined by collecting fractions of the effluent from the probe while washing the enzyme chalnber with PBS. The presence of protein in the fractions was .' ' by measuring the absorbance at 280 nm versus fresh PBS. The time required to wash all traces of the guanidine elution bllffer out of the enzyme chamber after the enzyme elution cycle was determined using the Pt indicating electrode as a detector for guanidine at the normal operating potential of +0.7 V versus Ag/AgCI. The enzyme chamber was washed with PBS until the sensor signal r~3turned to baseline.
Enzyme loading protocol:
1. Raise the internal electrode assembly.
2. Pump GOx-CBD conjugate solution for 1.0 minute at 4.0 rpm (ie., 2.2 mls/min).
3. Pump PBS for 4.5 minutes at 4.0 rpm.
4. Stop flow for 1.0 minute. At this point the bolus of GOx-CBD conjugate has filled the enzyme chamber.
112353\026651O.Wp - 28 -2~ 32~9 . Pump PBS for 7.5 minutes at 4.0 rpm to wash unbound conjugate from the enzyme chamber.
6. Lower the internal electrode assenlbly.
The GOx-CBD conjugate was pumped in~o the flow system by removing the inlet tubing from the PBS reservoir and inserting into a reservoir of conjugate. The soluble conjugate solution prepared for use in the enzyme chamber of the prototype biosensor was made 0.1 M
in NaCI to d~ ' ' the ~ )f the PBS electrolyte and satisfy the ~, of the Ag/AgCI reference electrode. Once a bolus of the GOx-CBD conjugate was loaded into the flow system, the tubing inlet was rinsed with distilled water and re-inserted into the PBS reservoir. PBS was pumped througtl the flow system to push the bolus of GOx-CBD
conjugate into the enzyme chamber. The total time required for the loading protocol was 14.0 minutes.
Enzyme elution protocol:
1. Raise the internal electrode assemt)ly.
2. Pump guanidine elution buffer for 2.5 minutes at 4.0 rpm.
3. Pump PBS for 2.0 minutes at 4.0 rpm.
4. Pump PBS for 5.5 minutes at 48.0 rpm (maximum pump rotation speed).
5. Lower the internal electrode assembly.
Separate tubing inlets were used for the guanidine elution buffer and the PBS
wash buffer, controlled by a three-way valve. The total time required for the elution protocol was 10.0 minutes.
Dummy electrode:
In addition to the - "~ rmC described above, a stainless steel plug, or "dummy electrode", was fabricated so that the probe body could be steam sterilized in a fermenter with the internal electrode assembly removed. The internal electrode assembly as ~ U~ U~ could not be autoclaved or th~ Chemgrip epoxy would become brittle and crack.
112353\0Z66~18,1Jp - 29 -2~32~9 A dummy electrode was required as a plulg when the intemal electrode assembly was removed, in case failure of the probe membrane during steam ~ . released ~u~,~. ' steam through the electrode shaft.
EXAMPLE 4: Interface r ~ ~
A custom designed p~ rl ", r ~ acid (Nafon) membrane (Aldrich Chemical Company, Inc., Milwaukee, Wl, U.S.A.) was cast on a 0.2 ,um Metricel GA-8 (cellulose triacetate) membrane filter (Gelman Instrument Co., Ann Arbor, MI, U.S.A.).
One coat of 250 ,uL of 0.5 % Nafion (in 50/50 isopropyl alcohol in water) was solution ca$
on a r~5 mm diameter membrane filter with a 50 ,~4m stainless steel mesh for rigid support.
The membrane filter and stainless steel mesh were secured onto the membrane cartridge with steel wire before casting the Nafion membrane. The Nafion solution was allowed to dry in air for at least 1 hour before the stainless steel sleeve was placed over the membrane cartridge and sealed with Silastic a&esivt~. The stainless steel mesh faced the interior of the membrane cartridge such that the Nafion coated membrane f~ter was on the exterior and would be in direct contact with the f~ f.n broth. It may be ad~i ~ , under some ~, to ~I~I.u, ' lly "clean" the indicating electrode before loading fresh enzyme. Using this ~, t, a smooth Nafion coating could be cast on the outer surface which would likely behave in a more well-defined manner im a stirred solution than the steel mesh. The steel mesh would also provid~ structural support for the membrane during the high pressure steam ' ' of the fe] menter.
In the original prototype, tlle silicone membrane of the CO2 probe was removed by separating the steel sleeve from the plastic membrane cartridge and dissolving the silicone with silicone sealant remover (Dow Coming). The new interface membrane was draped over the membrane cartridge and ~;ecured using steel wire and/or silastic medical adhesive (Dow Coming). The steel sleeve was placed over the membrane and cartridge and sealed in place with silastic adhesive.
112353\0Z66518.WP - 30 -2132~69 rl - - - Al f. . ;`l ;~ ~ of the glucose permeable outer membrane:
Table 2 lists the ~ f~ S of a number of different ' tested as possible glucose-permeAble outer ' for the prototype. ~our properties were determined to be essential for the outer membrane:
1. The membrane must be ' ~IJIG.
2. The membrane must be sufficiently pelmeable to glucose and oxygen so that the sensor response time is fast enough to follow changes in the glucose, in themedium. Por example, a high cell-density cultivation of E. coli with an optical density of 40 (measured at 600 nm versus distilled wat~ r) can consume 2.5 g of glucose from 1 L of medium in ~-~"u~ / 10 minutes (D. HaS~ W ' ' and ~. Jervis, I, ' ' ' ' results).A senor response time of less than five minutes was considered adequate for the present work, although faster response times would certainly be adv 3. The membrane must be i ~ n to ~l~tlUd,~,~iVG species which will contribute to a high b_~;h~;lu~ d signal, a~; well as to medium t I which will inhibit or denature the enzyme or poison the platinum surface of the indicating electrode over the course of a rt run.
4. The membrane itself must l~e resistant to fouling by protein or microbial adsolption (for example) over the course l~f a ~ run.
A number of different ' were ~ tested to the interface membrane of the present invention. Can~lidate ' were chosen on the basis of structure, strength, molecular weight cut-~ff, availability and ease of use. Non-ceLulose were chosen where possible to prevent conjugate binding to surfaces other than the intended cellulose matrix and thereby ~ , ' f~/ g, the system of G~.illl~,.l~l variables:
1 1 2353\0266518 . WP - 31 -` 2132~9 (i) Spectra/Por 2 dialysis tubil~g from ~ O ' cellulose, MWCO 12-14 kD, cut lengthwise to form a fl~t sheet (Spectrum Medical Industries, Inc., Los Angeles, CA, U.S.A.);
(ii) Amicon PM10 (MWCO 10 kD) and PM30 (M~!CO 10 kD) non-cellulose 1, ,. L ~ith suppon backing (Amicon Canada Ltd., Oakville, ON, Canada).
(iii) Filtron Omega 10 (MWCO~ 10 kD), I)ul~ c( ll1tr~filtr~tit~n membrane with support backing (Filtron T~ chnology Corporation, N. b~ u~. h, MA, U.S.A.); and (iv) A custom-designed p~ r~ u~..lrl acid (Nafion) membrane in accordamce with the present invention.

112353\0266518,Wp - 32 -2132~
Table 2~ f~ s of different ' tested as potential outer ' for the glucose biosensor prototype Dialysis Membrane MWCO Steam r, r Response ~cD) sterilizable? rating in time (min) complex mediurn None - - - 5 Spectra/Por 2 12-1~4 Yes - 12 PM303C No Poor 6 PM10la No E7air 25 Omega 10 IC No - c~3 Nafion - No Fair 10 Nafion (autoclaved)2 - No Fair 3 Nafion ( ' ~I)/Cellulose - No Excellent S
Acetate2~
The sensor did not respond to glncose m tbe concentration range from 0-23 mM ~vith this membrame.
2 The membrano caltndge with the Nafion nlembrane was submerged in PBS amd autoclaved before testmg.
3 This notation refers to the combmation of a~ Nafion membrame on the membrame cartridge (the glucose-permeable outer membrane of the serlsor) and a cellulose acetate coatmg on the surface of the mdicatmg electrode.
No off-the-she f membrane tested could satisfy all of the required conditions listed above. A custom membrane was designed using a solution of p~,~lluulu~u fonic acid (Nafion) cast on a 0.2 ,um cellulose triacetate membrane filter with a 50 ,Ibm stain ess steel mesh. Nafion has been used with good results as a dialysis membrane material on GOx/Pt electrodes for the ..~,t~ of glucose in whole blood (Harrison et al., 1988). Thep, ' vily of Nafion is due to the rejection of anionic species by the negatively charted p, rl ~ ionomer membrane, as well as the specific , ' I Oy of the membrane. The ' '-'~/ of the Nafion membrane is discussed below. The 0.2 ,um membrane filter acted as a suppott for casting the Nafion memb!rane, and also ensured that a sterile barfier was maintained in case of failure of the Nafioll coating. The stainless steel mesh provided rigid support.
112353\0266518.WP - 33 -2132~
EXAMPLE 5: E~. r~ of Interface r ~r By far, the most demandin~ conditions in which the camdidate ' will be used are the conditions of a real fermenter run, due to the presence of cells, cellular ' ' , proteins, and other medium ~ . Not only must the membrane be ' ~lc and enable a fast sensor response time, but the membrane must contribute to stable sensor operation over the course of the entire rt ,.... I;.,.~ The potential of each membrane to perform satisfactorily in the r., I~ environment was evaluated in 150 mL
beaker ~,A~r.. using Luria broth as a complex medium for testing. The prototype sensor with a given membrane was first calibrated in PBS and the response time to the addition of aliquots of glucose was noted (Table 2). The electrolyte was then ' 'y changed to Luria broth and the sensor was recalibrated in order to assess the ~lrO~ of the given membrane in complex medium. The time between obtaining the two calibration curves was minimi7ed to lessen the possil)le effect of enzyme d~ ,ti~tiol~ on the sensitivity of the second calibration curve. The addiition of Antifoam C to the Luria broth was found to have no effect on the sensor signal at steady state.
In order to compare the re~ults of a series of t;A~J~ ' ' using different on the same graph, the senso] calibration curve established in PBS for each sensor cnnf~lrPti--n was normalized with respect to the maximum sensor current (i.e, at 23 mM glucose) to give a maximum ," ' sensor current of l. The calibration curve inLuria broth from a given experiment was then expressed relative to the calibration curve in PBS by ' with the same value, since the only ~ I variable changed was the electrolyte/medium.
It can be seen from the results iin Figure 7 that, in all cases, significant signal attenuation and sensitivity loss was obser\~ed in complex medium compared to defined medium (ie., PBS). This l ' has been reported for other en2yme electrodes, as well as ion-selective electrodes, mass ~ , and gas ' ., l' Although not completely ~nrl~rF~ -orl, the observations are attributed to effects of the analyte matrix. A
number of possible ~ have been forwarded, however the l ' may be due to more than one chemical effect, or a symbiosis of several different effects:

.WP 4 2132~6~
1. It is the activity and not th~, ( of the analyte that is measured by chemical sensors. The equality of 1'(1111~ ", and activity is true only at infinite dilution.
The presence of additional molexular species, as in rt media, may reduce the activity of the analyte.
2. Low-molexular weight mol(xules may complex with , ~' ~ ' in biological media,thereby reducing the chemical activity witll respect to . the mass transfer through the sensor membrane, or the affinity of the enzyme for the substrate.
3. Proteims, lipids, and other hydrophobic cc -lr of biological media may occupy a non-aqueous, , i of the solutioll which is not accessible to the analyte. Thus the volume occupied by the analyte is lower than the total volume. This does not necessarily explain the observed signal attenuation and loss of sensitivity, but may explain the poor correlation in analytical results obtained by different texhmiques.
4. Changes in the ., I of the liquid phase may change the properties of mass transfer through ' cignjfi '~,. For example, the 1 v coefficient of the membrane/solutioninterface may differ depending on the sollltion. Protein or microbial adsorption to the membrane may increase the mass transfer resistance of th~ membrane.
5. In the case of en_yme elext:rodes,, , of the medium may inhibit or denature the enzyme or poison the elextrode, causing a d,xrease in the measured sensor signal or a loss of sensitivity.
For example, ~ c inhibition of glucose oxidase by D-glucal (a substrate analog) and/or halide ions (Cl-, Br, I ) has been reported. The ' li7:~fion of the enzyme was thought to prevent the mhibitory effext of Cl- ion.) In addition, other ~ . of the medium, such as ascorbate, are known to adsorb to platinum surfaces, form~ng a monolayer which blocks t lt~.tJt,ll~,...;~l reactions at the surface, thereby attenuating the sensor signal and eventually poisoning the elextrode.
6. In the case of oxidase en~yme-based systems, the analyte matrix may affect the activity or . of dissolved oxygen, which is required as the elextron acceptor for the reduced form of the en~yme cofactor FADH~. The solubil.ity of oxygen, for example, is reduced by high ,~
of ionic spexies.
112353\0266518.WP - 35 -2132~6~
Figure 6 shows the e~ i1ihrPtinn time of the prototype sensor when inserted into Luria broth, before the addition of glucose. All of the I ' tested, except the Nafionlcellulose acetate ' ' ' an initial currr.3nt response peah which then decayed to a stable ba~,h~ Julld.
Although the I ' for this behavioulr is not clear, the irJitial peak in response was aUributed to the oxidation of e~ ua~live species in the medium, and the subsequent decay of the peak was thought to be a result of ~embrane fouling and/or electrode poisoning by adsorption of proteins or other species in the medium. The absence of the initial response peah in the case of the Nafion/cellulose acetate -- ' membrane may have been dlle to the protective cellulose acetate coating on the indicating electrode and the IJ~ rc properties of the Nafion membrane. The Nafion/cellulose acetate ' membrane also established th~e lowest ba~h~l, ' signal compared to the otheru.~ 1, indicating a high rejection o~ interfering species, as expected.
These ~ l ;r ~. in addition to the relatively fast sensor response time and low signal attenuation and sensitivity loss in Luria brt)th (refer to Pigure 7), pointed to the Nafion/cellulose acetate membrane system as the best choice (among the 1~~ ' v, ~ ' here) for use in the biosensor prototype during glucose monitoring of a real f. nn The ' ' ' y of the Nafion membrane was established by r~uto~;lav ~ the memblane cartridge while submerged in PBS. After ' v ~ the membrane and rer11lihn~in~ the sensor in PBS, the sensor response time was found to have decreased from 10 minutes to 3 minutes. This may have been due to the dissolution of a small fraction of the Nafion coating during the high ~el~ c process of autuclav , (Moore and Martin, 1986), resulting in a decrease in mass transfer resistance in the membrane and a faster sensor response time. Most . ~1 '~" however, Figure 8 ~' that the p~lru of the autoclaved Nafion membranein complex medium was not s;".lirl.,alllly affected.
Effect of t~ lalulG and plI:
Using the Chemap fermentl r control unit and the 3.5 L fermenter, a number of fermenter operating parameters could be ~raried to mvestigate the effects on the sensor signal at steady state. The effect of I . ~; was ~. v'r "~ ' ~ over the range of normal operating i , c~ for industrial f, (see Figure 9). ~ direct ., ' ', between the sensor signal and " was expected due to i , ~i dependent increases in the reaction rate constant and the diffusion coefficient, as predicted by Arrhenius' law and the Stokes-Einstein equation, respectively.
112353\0Z66518.WP - 36 -213256~
However, the decrease in medium dissol~ed oxygen .. ,.. 1.. 1;~.. ~ at higher i , ~,~ (as a result of lowered oxygen solubility) may also h;lve reduced the sensor signal due to limitation of the enzyme kinetics. At le~ UI~;S greater than 4a~C the enzyme glucose oxidase is reported to be unstable (Nakamura et al., 1976). Fortier et al. (1990) have ~ ' the effect of ~f ~ .,r greater than 40C for glucose oxidase ' ' ' in polypyrrole on a Pt electrode and reponed a decrease in sensor cunrent, due to i . -indllced ~ of the enzyme. However, the effect wasfound to be reversible up to a i . ,; of 50C.
The effect of medium pH ~n the steady-state sensor signal is shown in Figure 10,which ,' a maximum at pH 5. The pH sensitivity of the sensor is l~rgely due to the pH
--If -~f of the enzyme activity itself.
Effect of medium dissolved oxygen tension, stir rate, and air flow rate:
The effect of medium diss~lved oxygen tension was v. '.i~, -' in the 3.5 L fenmenter by sparging the fenmenter ' Iy with nitrogen and air and adjusting the flow rates of the two gases to control the medium dissolved oxygen at various levels. A direct l~ l.,l~ was observed between dissolved oxygen level and the s~eady-state sensor signal (see Figure 11). The ~
for oxygen as the electron acceptor to tunn over the reduced fonm of the flavin group of glucose oxidase during the oxidation of glucose is known. At high dissolved oxygen . , where the enzyme kinetics are glucose limited, ~ariations in the dissolved oxygen level are not critical. The effect of medium dissolved oxygen on the~ sensor output becomes important in rf-..--- - '-'i..- - where the dissolved oxygen level ~ large rl From Figure 11, it can be seen that a const mt dissolved oxygen 1- would have to be maintained in order to eliminate the dissolved oxygen .Ifl). ---1~ -- ~ of the senscr output.
Dissolved oxygen control during a r ' " is easily - . !. '1~1 using cunrently available technology for feedback control of aeration and stir rate. The effect of variation of stir rate and air flow rate on the steady-state sens~r signal was v. ~ ' in the 3.5 L fenmenter. The sensor prototype was found to be insensitive to stir nate over the range from 300 to 500 rpm . A 3 %
change in the steady-state signal was observed over the range from 25 to 200 rpm, indicating that the sensor signal was limited to a small extent by extennal mass transfer resistance in this n~nge. At zero 11Z353~0Z66518.WP - 37 -2~32~1~9 stir rate, the sensor signal began to increase, which was attributed to Fi2O2 r- 1 in the enzyme chamber. F, with aeration f1~ml ' a 1-2% decrease in the steady-state signal with the ' of air flow, but no further change was observed over the range of air flow rates from 3 to 7 L/min. It should be note~ that these ub..v.v were recorded for the prototype sensor using the Nafion/cellulose acetate membrme system. The results are dependent on the sensor and the mass transfer properties of the membrane system, however, and as part of the process of ~ of the prototype, the sensor response should be rv . ~ . ;,..1 following each oesign change.
As the medium dissolved oxygen was exhausted, a significant decrvase in the sensor response was observed and the signal ~ .vhv I zero (data not shown). This may have been due to depletion of dissolved oxygen in the enzyme chamber by enzymatic l~ , and/or mass transfer into the fermenter medium through the sensor membrane. Glucose oxidase electrodes which depend on oxygen as the electron acceptor cannol: be used in anaerobic vllVil~ unless oxygen is provided by some internal source within the probe body (eg., an v~ ' buffer flow stre~un), or substitute electron acceptors, such as ferrocene derivatives, are used in place of oxygen to mediate the electron transfer from the enzyme to the electrode.
EXAMPI,E 6: Glucose r ~ , Du~ng Fed-Batch Cultivation of E.coli Cultivation of E.coli was performed in a 20 L fermenter for the purpose of monitoring glucose . with the glucose biosensor prototype. The probe body was sterilized in Sitll but the internal electrode assembly was remo~ed to prevent ~ ';.... of the Chemgrip epoxy used to bond the indicating electrode into the glass shroud. If tbis had occurred, liquid leakage into the glass shroud around the indicating electrAvde would have caused ' ' ' variations in the electrode current due to the exposure of vlvv~ lly active internal materials contacting the Pt electrode.
Other ' v' : l v adhesives must be i l v, v ' to rep ace the Chemgrip epoxy in the internal electrode assembly if the electrode is to r~ main inserted in the probe body during ~ . ;li,~li....
(although this may not be essential). The Nafion/cellulose acetate membrane system was used.
112353\0Z66518.Wp - 38 -The GOx-CBD conjugate v~as loaded using the enzyme loading2p1rotocoi described above, and the sensor was calibrated befare inoculation by adding a known amount of glucose to the medium in a series of aliquots. The sensl~r calibration curve was determined by comparing the steady-state sensor signal after each aliquot to the calculated glucose: in the fermenter.
The Mi.,l.~cl;,, '` ~.~ equation was fittecl to the sensor calibration curve and used as a conversion function to calculate the medium glucose, from the measured sensor signal during the fi . . - ~ The sensor response time ~vas five minutes or less.
After 8 hours, the enzyme was eluted and reloaded in situ. The internal electrode unit was lowered imto contact with the cellulo~:e matrix after elution of the enzyme to measure the l,ackæ-l ' signal. During this phase (ie., the second enzyme loading of the prototype) the b~ æ.. signal was used as the baselirle for rpr~ of the sensor after reloading fresh GOx-CBD conjugate. Ideally, the sensor would be " ' ' ~ at this point using internal calibration standards pumped into the enzyme chamber. The sensor was l~ ' ' ' by adding four aliquots of glucose to the fermenter, and the steady-state sensor signal was compared to the results of off-line glucose analysis of medium samples (taken once the sensor signal had reached steady-state). This method is not ideal, as the i.. of the off-line glucose analyzer are I ' into the calibration curve. In addition, the mediu]n glucose l is changing during the calibration due to cellular mP~ ic~ However, th~ sensor response time in this experiment was relatively fast, and it was found that steady-state sensor signals could be obtained within a C-lffl '~/ short period of time to obtain a useful calibration.
The f~.. - '-';.~l- run was carried out for a total of 16.5 c, v~. hours.
r i ," the AYrPrimP-~ was not terminated due to failure or ~ of the probe. The longest experiment reported in the literature (to this author's knowledge) involving glucose monitoring during a f~ with an in situ enzy,me electrode probe is 12 hours (Cleland and Enfors, 1983)33. Most of the ~- . 1 results reported in the literature ranged from 36 minutes to 5 hours of operation. The longevity of this experiment is attributed to the stability of the biosensor prototype provi~ed by the Nafion/cellulos,~ acetate membrane system and the capacity for in situ enzyme ~
33 Supra at 15.
112353~0266518.WP - 39 -Figure 12 shows the OUtpUl~ from the prototype sensor and the results of the omine glucose analyses over the course of the ~ . These results ~' the effect of the analyte matrix on the sensor response ancl the importance of sensor calibration under alu,ululul conditions. From Figure 12, it is obvious that the sensor output correlated more closely with the results of the off-lime arlalyses after .~ . of the sensor in the fermenter broth with cells, compared to the initial calibration of the sensor in fresh medium without cells. After reloading the enzyme and recalibrating the sensor, the ]3rofile of the sensor output followed the profile of the off-line analyses with substantially greatelr fidelity than the preceding phase, correctly indicating the exhaustion of the medium glucose and accurately following the infusion of glucose (and without significant de]ay. The improved correlation was expected, sirlce the of the off-line analyzer and the unresolved effects of the analyte matrix were included in the sensor calibration constants after recalibration. T-- ' " of the fermenter following the initial calibration in fresh medium changed the ~ , of the sample matrix, therefore calibration of the sensor would have best been performed after inoculation and after , of the l~,L~.~ 1 (ie., prior to loading the enzyme). Ideally, on-line calibration could be performed without disturbing the r~ i. . by using a series of irlternal calibration standards in a scheme similar to that proposed by Bradley and Schmid (1992)34. Alternatively, the sens~)r could be calibrated by adding glucose to the medium in aliquots after inoculation and ~' ,, the substrate, after each aliquot by calculation or by using the off-line glucose analyzer ~in the same matmer as calibration was performed following the second enzyme loading in this ~ ;.i t).
The glucose, determined by the prototype sensor were cu.l.i.,t~ ly lower than the results obtained from the off-line glucose arlalyzer. This behaviour may have been due to effects of the analyte matrix and/or systernatic differences between the two analytical methods. These observations are consistent with simila~r ~ 1 published in the literature. An empirical model was formulated for the sensor calibration curve which was used to experiment with corrections to the sensor calibration constants im an attempt to fit the sensor output more closely to the off-line glucose analyzer results. The Michaelis-Menten function used for the conversion of the measured sensor current (,uA) to glucose ,~ (gl],) was obtained by the equation below. An additional 34 Supra at 7.
11Z353\0266518.WP - 40 -21325~;9 parameter, Io was included to represent tlle value of the sensor baseline current which is normally subtracted from the measured sensor curr~nt before conversion, giving S = (I - Io) Km I""", - (I - lo) where S is the glucose ,- - r~n (g/L), I is the measured sensor current (,uA), Io is the sensor baseline (~A), I~ is the maximum sensor current ~A), and K'= is the apparent Michaelis constant (g/L). Using numerical analysis and sev~:ral initial values for the parameters Io~ I,~ and K~m~ the corrected sensor output shown in Figure ]12 was flPtPrrninPA The corrected and ~ ' sensor calibration constants for the frst and seca~nd enzyme loadings are shown in Table 3.
Table 3: Prototype sensor calibratioll constants for the first and second enzyme loadings Loacl #I Load #2 Parameter Uncorrected Corrected Uncorrected Corrected Io (~A) 0 014 0 0068 0.0068 0 0068 L,""~ (~A) 0.047 0 040 0.0127 0.0135 K= (mM) 8 6 9-99 6 44 10 10 Correlation 0 9797 0.9710 0 9856 0.9917 Coefficient The results were found to lle reasonable The calibration constants for the first enzyme loading could be corrected -~ y b~ adjusting the baseline to account for the change in the analyte matrix after; ' r~n The val~le used for the sensor baseline was the b~-uu~d signal ~'~ ' during the r ' '' (measured by lowering the internal electrode unit into contact with the cellulose matrix after elution of the enzyme and recording the sensor current). This value was taken to be relatively constant throughout the PYrprinlpn~ making the .: that the membrane system effectively rejected interfering spe~ ies and resisted fouling during the course of the r, The apparent Michaelis collstant was found to be nearly identical for the frrst and second enzyme loading, which is consistellt with the results achieved for multiple cycles of enzyme 11Z353\0266518.WP - 41 -2132~fi9 loading and elution using the modified rotating disk electrode. The values of I"~ which can be ti~ken to be ~ dlive of the amount of enz yme loaded, were not changed ~i" ~~ ~y by the correction procedure.
It can be seen from Figure 12 that, after applying the corrected sensor calibration consti~nts, the sensor output matches the lesults profle from the off-line analyses much more accurately. The transient r ' " in lhe sensor output oSserved at the beginning of the are not normally observed in the medium glucose on and are presumed to be a result of some initial instability in the local C:llVil~ ' of the probe (eg., due to entrapped bubbles), although the actual cause in this case coulld not be d~i ' In any case, the perturbation was temporary and did not recur. Once the prlll"l.-',."~ subsided, the correlation between the corrected sensor output and the off-line results was excellent.
The approach used above assumes that the off-line glucose analyzer was precise and accurate and that the sensor output was irl error. According to some researchers, the accuracy of the measured value of a single sensor can nolmally be validated by, , with alternative III~DUII ' methods. 1,' ~ 'y, different measuring methods often will produce different results and the most accurate analytical method cammot easily be ~' ' This is especially true in the case of biological systems, which are frequently more difficult to measure accurately than simple physical or chemical systems. Cl I of the off-line glucose analyzer results ' '~, after inoculation (2.17 g/L) with the glucose sensor output (1.98 g/L) and the calculated medium glucose ~ (2.40 g/L, based oll the volume of medium and the amount of glucose added) reveals a significant d;~l~dlll y in both Inethods. The choice of the most accurate and/or reliable analytical method must often be based on P~rr~ i-onre In this CiAIJ~ ' t, the results from the Beckman glucose analyzer were used as tlhe st~ndard for ~ l for reasons of practicality, availability, and relative ease of use.
To determine the useable lifetime of the enzyme component of the sensor, periodic internal calibration checks could be performed to determine at what point the activity of the enzyme had I~ f~ ' -'1 to an ~ - r ' y degree. In this c*,~ . t, the results from off-line glucose analysis of medium samples were used as a reference in an attempt to identify drift in the prototype sensor output which could be attributed to enzyme deactivation over time. This approach makes the 11Z353\~Z66518.WP - 42 -2132~69 that the results from the offlirle glucose analyzer were stable and reliable over time. With respect to this, care was taken to recalibrate the glucose analiyzer before analyzing each medium sample. It was expected that if some sys~ematic ~ 4ll~;y existfd between the off-line glucose analyzer and the glucose sensor output, tlie error would be consistent over time unless some process of membrane or electrode fouling or enz~me d~4~ 4lh).~ modified the results from one (or both) of the sensor(s). The results overaD indicatf~ that continuous glucose monitoring could possibly be performed for at least 6 hours before elution and l~ a of the enzyme would be necessary.
EXAMPLE 7: Glucose r ~ dui~ing Fed-Batch (~ of E. coli Organisms:
A strain of E. coli JM101/1~TUgE07K3 was used for the cultivation. The organism contained the plasmid for the production ~f CBDCex and was stored in 10% DMSO at -70C. The plasmid consisted of the tac promoter andi the leader seqiuence of C.fiff~ ,' (Cex), foDowed by the structurali gene for CBDC,, . The ri,sistance marker was kanamycin and the imducer was IPrG
(E. Ong, ~ d results). The inducer was not added in this ~ Jf ' Media:
Minimal medium M-9 was prepared with the foDiowing ~ . (g/L): Na2HPO4, 11.76; KH2PO4, 5.88; NaCI, 0.5; NH4CI, 1.0; MgSO4, 0.49; CaCI2, 0.01; thiamine, 1.685;
y, , 0.025; and the following trace metals (mg/L): Alz(SO3 7H2O, 0.040; C0CI2 6H2O), 0.032; CuSO4 5H2O), 0.008; H3BO3, 0.004; MnCI2 4H2O), 0 080; NiC12 6H2O), 0.004;Na2MoO4 2H2O), 0.020; ZnSO4 7H2O, 0.0-20. The startimg glucose, nn was 2.40 g/L.Medium for the imoculum was prepared as follows (g/L): Na2HPO4, 6.00; KH2PO4, 3.00; NaCI, 0.5;
NH4CI, 1.0; MgSO4, 0.49; CaCI2, 0.01; I'hiamine, 3.37;1 y~- 0.05; glucose, 2.80 g/L.
112353\C266518.WP - 43 -2132~69 ~C ' ' ' v~liu~.
The cultivation was performed in a 20 L Chemap Type SG fermenter with the standard blade stirrer and three 25 mm side ports. The fermenter working volume was 8 L. A 500 mL
inoculum was prepared in shake flask culture. One side-port was used for the Ingold InFit 764-50 pH
electrode. A second side-port was used for an optical density monitor (Cerex MAX Cellmass Sensor Probe, Cerex Corporation, Ijamsville, MD, U.S.A.). The third side-port was used for the glucose biosensor prototype. A 19 mm diameter Ingold sterili7able 2 probe (No. 40180-03) was inserted through a port in the fermenter head plat~ . For the purpose of monitoring glucose, with the glucose biosensor prototype, it was l - y to use aseptic techniques during the rr. ,... ~
The Chemap 3000 Series base unit and controller were used. Initially, the air flow rate was set at 6 L/min and stir rate was contl~lled between 50 and 700 rpm in order to maintain the dissolved oxygen setpoint at 95% of the air saturation. As the r, reached higher cell density, it was necessary to increase the a!ir flow rate to 7.5 L/min and decrease the dissolved oxygen setpoint to 80%. Temperature was cont~)lled at 37C. Medium pH was l ~ " ' Analyses:
Samples were withdrawn fi om the fer nenter at different intervals using the sampling/harvesting valve. Glucose ~.,....l~.,.~;.-,~ in each sample was analyzed using the Beckman Glucose Analyzer 2 (Beckman I~ Inc., Fullerton, CA, U.S.A.) after ~ ~.bi..b the sample at 14,000 ~pm for 2 minutes. The absorbance of each sample at 600 nm (versus distilled water) was also measured using a Varian DMS200 W-VIS ~11, . ' (Varian Pty. Limited, Mulgrave, Victoria, Australia). In addition, the Genesis Contrvl Series software package (Tconics, I7vAbvlu~.bl., MA, U.S.A.) was used to log i , ~;, pH, dissolved oxygen tension, stir rate, and optical density from the on-line sensors every five minutes during the course of the r, The response of the protot~pe could be calibrated with respect to glucose .
up to at least 23 mM (the maximum, tested) in medium without cells. The sensor signal was relatively stable and noise-free, ~ g less than 5 % variation per hour and signal noise less than 1% of the senor current. The GOx-CBD conjugate could be loaded and eluted successfully 112353\0266518.WP - 44 ~

2~325~9 sing the inlet and outlet tubing of the modified Ingold probate body and the loading and elution protocols described above. After elution of the enzyme, the sensor response to glucose was less than or equal to the baseline signal, confirmin,~ that the enzyme had imdeed been eluted.
The Whatman qualitative f iter paper proved to be a ,.l~i,r.l- ~uly cellulose matrix for the biosensor prototype. The low msss trdnsfer resistance of the filter paper compared to the other cellulose matrices tested was ddv ' ~ in terms of fast sensor response time and low signal attenuation due to the matrix. The filter paper also had a nigh porosity and cellulose surface area for binding the GOx-CBD conjugate, and was anticipated to be more easily perfused with the reagent solutions than the ~ ~ cellulose dialysis membrane or the nitrocellulose protein transfer membrdne. Structurdl stability of the filter paper was not a problem, as in the RDE~ , due to the absence of shear stress from stirring or rotation.

112353\0266518.WP - 45 -

Claims (11)

1. A selectively permeable interface membrane which may be used in a biosensor system to separate biochemical, optical or other processes from an analyte matrix, saidmembrane comprising a supporting mesh, a perfluorosulfonic acid polymer impregnated substrate and a homogenous film of perfluorsulfonic acid polymer.

2. The interface membrane of claim 1 wherein the substrate is selected from those comprising cellulose, cellulose acetate, nitrocellulose, polysulfone, polyvinylchoride, polyurethane and polyvinylalcohol.

3. The interface membrane of claim 1 wherein the substrate is cellulose triacetate.

4. A method of preparing a selective permeable interface membrane which may be used with a biosensor to separate biochemical, optical or other processes from an analyte matrix which comprises:
- fixing a substrate onto a supporting mesh to form a substrate membrane, - casting a perfluorosulfonic acid polymer on the substrate membrane, and - curing the product so formed.

5. The method of claim 4 wherein the substrate is selected from those comprisingcellulose, cellulose actetate, nitrocellulose, polysulfone, polyvinylchoride, polyurethane and polyvinylalcohol.

6. The method of claim 4 wherein the substrate is cellulose triacetate.

7. The method of claim 4 wherein the supporting mesh is a metallic screen selected from the group comprising stainless steel, aluminum, copper, silver and gold.

8. The method of claim 4 wherein the supporting mesh is a polymeric secreen selected from the group comprising poly(vinyl choride), poly(tetrafluoroethylene), polystyrene and polycarbonate.

9. The method of claim 4 wherein the supporting mesh is a fibrous filter material.

10. A regenerable biosensor probe adapted for positioning in an environment characterized by the presence of biological molecules which are substrates for or products produced by enzymes in order to determine the presence of said molecules, said biosensor probe comprising:
- a selectively permeable interface membrane which separates the biochemical and electrochemical processes from the environment when the probe is in place;
- a porous protein-receiving matrix adjacent to the interface membrane;
- an indicating electrode covered with an electrically insulative material, said electrode abutting, at one of its ends, the protein-receiving matrix;
- an inlet conduit through which fresh protein conjugate may flow to the protein-receiving matrix; and - an outlet conduit through which spent protein conjugate may be removed from the protein-receiving matrix.

11. The biosensor probe of claim 10 wherein the protein conjugate comprises the cellulose binding domain extracted from bacteria of the genus Cellulomonas.