CA1158720A - Pulsed voltammetric detection of bacteria - Google Patents

Abstract

ABSTRACT

Reliable and rapid detection of microorganism is accomplished in an electroanalytical cell using a pulsed voltammetric detection technique employing the growth medium as the electrolyte and analyte and using simple wire electrodes fabricated from readily available materials. Organism detection occurs as a consequence of the depletion of oxygen in the growth medium/electrolyte caused by aerobic metabolism. Times-to-detection vary with inoculum strength in a predictable fashion, permitting quantification of the organism in question when results are compared to those obtained using known inocula of the same organism. The low duty cycle of the pulsed measurement enables the determination of the relative redox potential in the same cell using the same set of electrodes in order to provide information which may be characteristic of the type of organism being studied.

Description

e~

~lS

PUI,SED VOLI`~MMETRIC DETECTION OF B~CTERIA

The present invention relates to a method for the detection of microorganisms. More particlllnrly, the present invention relates to a simple, efficient and reIiable electrochernical method îor the detection of bacteria by measllring the decrease in polarographic oxygen current passin~ throu~h an electroanalytical cell contnining two dissimilar wire electro~les immersed in a liquicl culture medium.
The determination of whether or not a substance is contaminated with biolo~icaUy active agents such as bacteria is of great importance to the medical field, the pharmaceutical inclustry, the public health Iielcl, the cosmetic industry, the food processing industry, and in the preparation of interplanetary space vehicles. One of the most widely used techniques for making this determination, especia11y in medical applications, has been nutrient agar plating. In this method a microorganism is allowed to grow on an agar nutrient substrate, and the growth of the microorganism is observed, at first visually and thereafter by microscopic examination. This technique, which is most commonly used clinically, requires overni~ht incubation of plates before results are available.
Another technique widely used for the determination of microorganisms involves supplying a microorganism in a growth rnedium with carbon-14 labeled ~lucose or the like. See Waters IJ.S. Patent No. 3,6~6,679 and Waters U.S.
Patent No. 3,935,073. The microorg~nism metabolizes the radioactive gIucose and evolves C1~02, which is sampled and counted. While positive results can be obtained by this radiometric method in a relatively short period of time, this method requires the use of comparatively expensive and complex apparatus and involves the handling of radioactive materials.

~,~

'7~ ~

Th~ pl ior nrt also ciescribes a number o~ detection techniques basecl on electrochemical phenomenn. (1enerally these techn;qlles employ very clelis~ate ancl e~pensive electronic equipment nnd nre extremely di~ficuIt to use in nn on-going detection progrnm. One of these described rnethocls involves the measurem ent of polarographic oxygen current in an electroanalytical cell. CeU current is a function o~ the dissolvecl oxygen content of the electrolyte, and the metabolic activity of any oxygen-consumillg microorganisms present will, therefore, cause the current vaIues to fall off. ~or a general discussion of this electroanalytical t~(hnique see Hitchmall, Measurerrlent of Dissolved O2cy~en (1~77); Fntt, Oxygen Sensors (1~)76)~ and Norris, Methocls in Nlicrobiolo~ (1970). Modern techniques of polarographic oxygen meas1lrement rely almost exclusively on the so called Clark-type electrodes which employ a semi-permeable membrane to prevent the electrocles from contncting the solution; see Clark U.S. Patent No. 2,913,386.
The commercially available membrane polarographic oxygen de-tector (MPOD) is presently used to determine dissolved oxygen in BOD studies, marine ecology, wastewater treatment and the like. The MPOD is usually constructed with an inert cathode material (gold, platinllm) and a silver-silver chloride reference electrode, and uses a relatively concentrated (0.3 - 3.0M) potassium chloride electrolyte. 'I'he electrode areas are relatively large (ca. 1.0cm2) and are prevented from contacting the solution to be analyzed by a semi-permeable membrane, usually of polyethylene or polytetrafluoroethylene. The potential applied to the electrodes is normalIy about 0.8V, cathode negative. This potential must be applied to the MPOD
several minutes prior to any use of the detector, and must remain applied throughout the duration of any measurements to be made. The MPOD îs thus a "steady-statell device, in that all electrode reactions stabillze at new equilibrium values under the influence of an applied potential constant in time. The steady-state cell current detected under these circumstances is a measure of the dissolved oxygen content of the solution. Because proper electrode operation depends upon diffusion of dissolved oxygen through the membrane to reach the cathode, the solution must be stirred or agitated constantly to prevent the depletion of oxygen from the sample solution in the immediate vicinity of the membrane from affecting the results. Some -investi~ntion of !~'00 or~elation under non-steacly state or pulsecl conditions t hns been unclet takell; see T-litchman~ Sllpra~ Cllapter G.
ï he Clark-type polnrograpl- sensors, however, suffer from serious r drnwbacks which malce them undesilnble for the detection of microorgranisms.
These sensors are expensive and cumbersome to use. The relatively high cost of the electrocles precludes the use o~ a separate electrode for each t sample. Thus in order to prevent cross contamination of samples, the electrode Sut faces have to be sterili~ed between samples using a strong bactericicle, and then rinsecl completely with a sterile rinse solution so as not to kill org~nisms in or contaminate the contents o~ the next sflmple cell tested.
The electroanalytical detection of microorganisms by measurement of oxiclation-recluction potentinl has also been described in the prior art; see generally Norris, suprfl, Chapter 4. Tn the redox potential method a platinum electrocle in combination with nny commonly usecl reference electrode such as the calomel electrode will cvidence an eguilibrium potenti~l in ~rowth medium proportional to dissolved oxygen in the medium and to any other oxidation-reduction (electron transport) reactions taking place in the solution.Because this is an equilibrillm measurement, a voltmeter with very high input impedarlce must be used to measul e the potential existing between the electrodes so as not to clisplace the equilibrium as a result of current flow. Microorganisms ~rowing in the medium use oxygen from the solution, and may possibly contribute to other redox reactions which cause the measured potential, usua11y greater than +lOOmV (cathode or platinum positive with reference to the standard calomel electrode) in sterile médium, to shift toward more negative values. l~erobic organisms are able to reduce the solution enougrh to yield measured potentials of -lOûmV to -200mY (vs.
SCE). This is also the range of redox potentials where the voltammetric m ethods cease to function; the cell current due to dissolved oxygen is by this time very small, and is usually swamped by the residual cell current d~le to solution impurities and electrode imperfeetions. Facultative lmaerobes, however, may reduce the solution extensively~ An exhausted culture of P.
mirabilis will have reduced the medium in a sealed container to a value of . .
around -550mV (vs. SCE) before ceasing growth. The redox potential method by itself is not well suited to the detection of bacteria because it is 1, relntively slow ~n(l the response will clepend on the type of organism being cletected.
From the foregoing it is clear thnt a need exists for a simple rapid and reliable method of detecting the presence of microorganisms in a suspect sample.
~ ccordingly, it is an object of the present invention to provide a method for cletecting the presence of microorganisms which empIoys apparatus which is relatively simple in both construction flnd operation Rnd which uses relatively inexpensive non-radiolabeled materials.
It is nlso an object of the present invention to provide a method for the cletection of microortrallisms which facilitates computer controlled automation and which can incorporate disposable components.
Further objects of the invention will be apparent from a consideration of the following description, These and other objects of the invention are achieved by providing an electroanalytical method for detecting the presence of oxygen-consuming microorganisms in a sample comprising the steps of providin~ a mixture of said sample and a fluid culture medium capable of supporting microorganism growth in an electroanalytical cell equipped with two electrodes which are in contact with said mixture; applying a series of voltage pulses-of substantially constant amplitude and duration across said e]ectrodes; and measuring the resulting current prior to the trailing edge of each of said applied voltage pulses; the presence of oxygen-consuming microorganisms being indicated by a decrease in cell current which is a function of the dissolved oxygen content of said mixture.
In a preferred embodiment the present invention also cantemplates a process for the detection of microorganisms as described above and ad~itionally comprising measuring the open-circuit voltage potential across said electrodes during the interval between successive applied voltage pulses.
Determination of the dissolved oxygen content of the cell is accomplished by pulsing the electrodes briefly with a known potential (cathode negative) and measuring the resulting current through the cell prior to the trailing edge of the applied voltage pulse, The growth medium in the cell is used as both the analyte and the electrolyte for the deterrnination. The very low duty cycle of the pulse with r espect to the overall sampling 2l(~
interval obviatcs the necd for constant agil:ation or ;tlrrlng of the sample solukion required by conventional stea~ly~Y~ e methodology, and permits the same electrodes to be uscd to determine the relative oxidation-reduction potential in tlle cell through the rneasurcment of the open-circuit potent:lal existing between the electrodes. Bact:erial dctection is bcst accomplished by measuring -the decrease in pulsed volt.lmmetrlc oxygen current, while inEorrnation characteristic of the ty~e of organism present is best furnished by the relative cell potential determination.
Times-to-detection for all organisms studied vary wlth inoculum strength in a predictable fashion, permittillg accurate quantification of the organism in question when results are compared with times-to-detection obtained using known inocula of the same organism.
The process of the present invention provides numer-ous advantages compared to traditional manual methodology and present automated systems. This process requires a cell of very simple construction, provides ample opportunity or the creation of disposables, promotes automated quality control, prevents any chance of cross-contamination, and can be configurcd - as an instrument very sophisticated in operation, yet extrcmcly simple to operate.
Figure 1 is a front elevation view of an electro-analytical cell useful in the process of the present invelltion.
Figure 2 is a sectional view of the electro-analytical cell of Fig. 1 taken along line 3-3.
Figure 3 is a top plan view of the electroallalytical cell of Fig. 1.
Figure 4 is a simplified schematic diagram o an analog conditioning circuit useful in the process of ~lle present invention.

~t3~'7~,~
Fi~ure 5 is c3raph showing the ccll currcl-lt rc~sponse oE a Eully-c3rown E. c~oli culture at vari.ous pulse wi(lt}ls as a function of applied cell voltage.
Figure 6 is a graph showing the cell cul-rcllt respon.se of a ste~ile cell with constant appl.ied po~cntial as a fullction of elapsed ti.me since pulse applicati.on.
Figure 7 is a graph showing the cell current response as in Figure 6 with the elapsed time extendcd to 300 seconds.
Figure 8 is a graph showing the cell current response as a function -5a-of applie(l potential for n sterile electroanalytical ce~l and for a cell containing fully-grown F. coli cultut e.
~ ;gure 9 is a graph showillg voltammetric cell current response for a sterile ceLl purgre{l ~vitll dry nitrogen; cell current is recorded as a function of elasped time during the purge.
Figure 10 is a graph showing the cell current response for the sterile cell of Fig. 9 purged with dry nitrogen, then purged with roorn air.
Figllre 11 is n schematic representation of the sampling and dikltion scheme used in the preparation of electroflnfllytical cells and pOUI' plates for the e~amples.
Figrure 12 is n graph showinog normaliYed voltammetric cell current response as a functioll of incubation time for varying inoculum strengths of the organism E. coli.
Figure 13 is a ~aph showing normalizecl cell potential response as a function of incllbation time for varying inoculum strengths of the organism E. coli.
Figure 14 is a graph showin~ normalized- voltammetric eell current response as a function of incubation time for varying inoculum strengths oE
the organism E. cloacae.
Figure 15 is a graph showing normalized cell potential response as a function of incubation time for varying inoculum strengths o~ the organism E. cloacae.
. .
Figure 16 is a graph showing normalized voltammetric cell curr ent response as a function of incubation time for varying inoculum strengths of the organism P. mirabilis.
... .. _.
Figure 17 is a graph showing normalized cell potential response as a function of incubation time for varying inoculum strengths of the organism P. mirabilis.
Figure 18 is a graph showing normalizecl voltammetric cell current response as a function of incubation time for varying inoculum strengths of the organism P. aeru~osa.
Figure 19 is a graph showing normalized cell potential rcsponse as a function of incllbation time for varying inoculum strengths of the organism P. aeruFinosa.
Figure 20 is a graph showing normalized voltammetric cell current response as a function o~ incubation time for var!~incr inoculurn strcngths of the orgrnnism S. a~!reus.
Figure 21 is a grnph sho~vingr nol malizeù cell potential response as a function of incubation time for varying inocull1m strengths of the organism S, aureus.
Fiv~ure 22 is a graph showing normalized voltammetric cell current~
response as a function of incubation time for varying inoculum strengths of the organism S. bovi~s.
Figure 23 is a graph showing norma1ized cell potential response as a function of incubation time for vnrying inoculum strengths OI the organism S. bovis.
.__ Figure 2~ is a graph showing normalized cell current response as a unction of incubation time for the orgnnism E. coli with four decades of initial inocu1um concentration.
~ igure 25 is a g,raph showing the logarithm of the initial inoculum dilution ratio as a function of time--to-detection at a 60% detection threshold for the data shown in Pig. 2~.
Figure 26 is a graph showing normalized voltammetric cell current response as a function of incubation times for varying inoculum strengths of the organism E. coli using cells fitted with golcl cathodes.
~ igure 27 is a graph showing normali~ecl voltammetric cell current response as a function of incub~tion times for varying inoculum strengths of the organism P. mirabilis using cells fitted with gold cathodes.
Figure 28 is a graph showing normalized voltammetric cell current response as a function of incubation times for varying inoculum strengths of the organism P. aeruginosa using cel1s fitted with goId cathodes.
The present invention detects the presence of bacteria in a suspect sample primarily by measuring the decrease in voltammetric oxygen currerlt passing through an electroanalytical cell containing the sample and a fluid growth medium. Viable organisms capable of utilizing dissolved oxygen during rnetabolism will cause the deteeted oxygen current to decrease with continued incubation, signifying detection. Sterile inocula will evidence no such current decrease. ~clditional means is provided to measure the open-circuit voltage of the analytical cell in order ~o obtain information as to the type of bacteria (primarily aerobic or facultative anaerobic) present in the cell. Organisrns thflt consume little or no oxygen from the growth - ~ -meclium, yet which have the ability to alter the solution redox potential, rnny be cletected by notirlg the change in the solution reclox potential with incubation, as furnished by the open-circllit cell potential measurernent.
The methocl of the present invention ean be used to deteet the presence of any aerobie or faeultntive organisms ~Jhieh consume oxygen frorn a liquid meclium during metabolism. Specifie exarnples of such organisms include bacteria such as E. coli, E. cloncae, ~. mirabilis, P. aeruginosa? S.
aurells~ S. bovis T~. peneumoniae, S. albus ~. o~ytoca, E. aerogen~ E.
? ~
ag~lomerans~ C. freullclii~ E'. morga~L P. stuartii, S. rmarcescens~ Group B.
Beta strep, C.rp. D. Strep, ~nd yeasts such as C~ albicans.
In the process of the present invention a small portion of a s~speet sample is first introdllced into an electroflnalytical cell containing a liquid growth medium. Tlle growth medium also serves flS the primary electrolyte in the cell. ~ny medium which will support the g~owth of vxygen -eonsuming mieroorganisms may be utiliæed.
Typical growth media generally contain water, a earbon souree, a nitrogen source, caleium, magnesium, potassium, phosphate, sulfate, and trace amounts of other minor elements. The carbon source may be a carbohydrate, amino acid~ mono- or clicarboxylic acid or salt thereof, polyhydroxy alcohol, hydroxy acid or other m etabolizable earbon compouncl. Usually the carbon source will comprise at least one sUcrar such as glucose, sucrose, fructose, xylose, maltose, lactose etc. ~mino acids such as lysine, glycine, alanine, tryrosine, threonine, histidine, leucine, etc. also frequently comprise part of the culture media earbon source.
The nitrogen souree may be nitrate, nitrite, ammonia, urea or any other assimilable organic or inorganic nitrogen source. ~n amino acid might serve as both a carbon and a nitrogen source. Sufficient nitrogen should be present to facilitate cell growth.
A variety of calcium, potassium and magnesium salts may be empIoyed in the growth medium inclucling chlorides, sulfates, phosphates ancl the like.
Similarly, phosphate and sulfate ions can be supplied as a variety of salts.
As such materials are conventional in fermentation media, the selection of specific materials as well as their proportions is within the skill of the art.
The so called minor elements whieh are present in trace amounts are com monly understood to include manganese, iron, zinc, cobalt and possibly others. Due to the ~act that most biologically active species cannot function .' , . ~

'C~~ ~ ~ ~

- ~ -in s~rongly acidic or strong1y alkaline medi~, sui~able buf~ers such ~s potassiurn or ammoniurn phosp}lates rnay be employed, if desired, to maint~in thc p~-3 oî the gro~th medillm neal ncutrality.
Exarnples of well known gro-vth media which may be used in the present invention are peplone broth, tryptic soy broth, nutrient broth or thioglycolate broth. I`ryptic soy broth-based rnedium (6Es Medium, Johnston Laboratories Inc., Cockeysville, Md.) has been found to work welI. The amollnt oî growth mediurn provided in the electroanalytical cell is not overly critical. 5.0cc of 6B medium has proven very e~fective.
The analytical cell useful in the process~of the present invention may be of any convient si~e and shape. The cell can ~ie formed from any materials norrn~lly used in the rnanufncture of e1ectroanalytical cells such as ~lnss, plastic and the llke. ~ny material ~hich does not affect the growth the microorganlsms or the measurement o~ electrochemical phenome)-on in the cell can be employed. In the preferred form, the electronalytical cell useful in the process of the present invention comprise a plastic container of the general configuration shown in ~igs. 1-3. Cell volume may vary accordingr to the cell design and is not critical. A ce~I
of the type shown in Figs. 1-3 has been effectively used at a capacity of about 10-15 mls. In the preferred manner of operation a number of these cells can be utilized in the form of an array to permit testin~ of multiple samples.
The electroanalytic~l cell is a~so equipped with two electrodes in electrical contact with the growth medium. The ~vorkin~
electrode (cathode) is normally a noble metal, for example, gold, silver or platinum. When only voltametric measurements are to be taken, gold ,platinum or silver are preferred ~or the cathode.
When potentiometric (redox) measurements are also taken gold should not be used as the cathode material. The reference e~ectrode is preferably pure silver ~99 95% or better) electrolyzed in place using a basic electrolyte to deposit Ag20 on the silver. Silver chloride may be electrolytically deposited from HCl solution to form the alternative Ag/AgCl reIerence electrode, but this electrode has been found less stable in this application. In actual practice clean, unprepared silver wire will quickly become covered with a mixture OI Ag20 and AgCl due to Cl`ion in the medium and to the very high pH
around the anode when pulsed . Al though pure silver is known to be - ]O -bactericidal, no evidence of such toxicity has been noted using oxidized silver r cference clectrodes.
The electl odes may be used in any convenient form. Preferred are wires o~ the above materials al~hough other forms such as printed circuit traces can be used. Most preferred are l~-shaped staples inserted through the bottom of the cell as best seen in Fig. 3. The elec~rodes, however, can be of other conventional ~orms, including spaced apart vertically disposed hair pin shaped eLectrodes. The electrode wire diameter l is not critical. ~Yires as small as 0.010" rnay be used, provided their ¦
frangibility nnd low sensitivity can ~e tolerated. Wires approaching 0.040"
probably represent a practical upper limit since these materials are quite expensive. Preferre(l electrode diameters nre in the range of from about .015 to .050", with about 0.020" to 0.040" being most preferred. Electrode lengths are lil<ewise noncritical. In practice, lengths of from about 0.5cm to 2.0cm and preferably about l.0 to 1.5 cm are suitable. The wires may be separated by about 0.5 to 2.5cm and preferably about l.0 to 2.0 cm.
It wil} be apparent to those skilled in the art that solid precious metal wires can be replaced with less expensive wire electroplated with the precious metals of choice.
The electrode pair may be covered with a porous gel, preferably a nutrient gel such as tryptic soy agar (TSA). Other gel materials which may be used include gelatin, dextran gel, carrageenan gel and the like.
Best results are obtained when the gel just covers the electrodes. The main benefit of the gel is to reduce measurernent baseline drift sometimes caused by the introduction of biological samples (urine, etc.), presumably by preventing the migration of large charged molecules to the electrodes. The quantity of gel is r,ot critical; l.0cc of TSA has served to just cover the electrodes in the type of ce71 shown in Figs. l-3. It may be appreciated that some ionic conduction is necessary in the gel; hence its equivalent conductance when saturated with growth meclium/electrolyte should approach that of the medium alone.
Ihe electrode pair may also be isolated from the effects of large charged molecules by positioning a layer of porous material such as ordinary filter paper over the electrodes. ~hile this will not prevent contact of the electrodes with tne analyte or even with the microorganisms in the sample, it ~.lill limit migration of lflrge charged molecules to the electrode region.

-In nnother embodiment of the present invention the re!felence electrode can be isvlate(1 Erorn tlle analyte mi.Ytul e by the use of a salt briclge or othel conventionnl means. In this rnnnner it i5 possible to employ a single reference electroc3e in conjunction with a plurality of worlcing electrodes in scparate analytical half cells.
~ fter the cell has been inoculated with a sample to be tested a series of voltage pulses of substanti~lly constant amplitucle àncl cluration areappliecl ncross the electrodes. The voltnge amplitude of the pulses can vary from about -0.35 v. to about -0.90v. Preferred is an applied voltage pulse of abollt -0.70v. Tlle pulse cluration should be at least abollt 600 milliseconds.
The llpper limit of the pulse duratioll is not critieal. ~\s a practical matter,times mllch over nbout 3 seconds reslllt in a redllced current signal but may be used. Preferably the pulse cluration can be from about ~00 to 2000ms. with about 1200 ms. being most preferred. The pulse intervaI is not critical ancl should be short enougll to follow the biolo~,ical changes but long enol1gh to allow the cell to app~oach equilibrium conditions for redox potential measurements in the pulse intervals. Times OI from about 5 min.
to 20 min. are suit~ble. A pulse interval OI about 10 minutes is preferred.
During the testing period the analytical cell and its contents should be held at a constant temperature, preferably 37C ~0.2~C. It is understood, however, that not all biologically active agents exhibit maximurr1 growth within the cited temperature range. I~ it is of interest to determine whether or not a specific microorganism which grows better at some other temperature is present, then the temperature at which the organism in question exhibits maximum growth should be employed.
Current readings from the cell are taken prior to the trailing edge of each of the applied voltage pulses. The pulsed, periodic nature of the measurements involved obviates the need for constant agitation or stirring of the sample as would be required for conventional steady-state polarographic oxygen determination. The requirement of extremely hi~h input impedance for the potential measuring circuitry is similarly relaxed, since the analyticalcell is connected to the external electronics only long enough for appropriate potential and current readings to be taken. The simplicity of the analytical cell coupled with the complexities of sample selectio1l and interrogation suggest that the technique of the present invention be practiced in a fully J ~i nl1tolnate(l fashion. A microcoi-nputel sy~st~m t an be use(l to control all aspects of expel ilnental measllrelrlent an(l to nnaly~e~ tabulate, plot, and store on floppy clislc the information gatllerecl during each experiment.
The constr~ tion of a typicul analytical cell useflll in the process oî
the present invention is shown in ~igs. 1-3. ~ semi-flexible plastic container (1) receives the t-vo wire electrodes (2,3) in the form of U-shaped "staples"
inserted through the hottom of the container. The worlcing electrode (2) is analytical-grflde platinurn, Q.035" diameter. Re~ere~nce ele~trode (3) is 0.0~0" d;ameter ~\g/AgO. The elect~ocle pnir is covered with a nutrient gel (4) such as tryptic soy ngar (rSA). ~ quantity oî liquid growth rnedium (5) is nlso present in the ceLl to serve ns the primary electrolyte and to facilitnte the growth of rnicroorganisms.
l`he per~ormance of a single experimental clata-gathering cycle may be understood by considering 1 single two-electt ode cell as part of the signfll c onversion CirCIIitry, as presented in simplifiecl form in ~igure ~. A
clflta-gathering cycle begins with the selection oE this particular cell. After a short delay to allow the cell selection relays to settle, a potential reading is taken with relay 11 open via operational arnpli~iers OAl and Q~2~ Because the cell redox potentinls on platinum are bipolar with respect to the reference electrode (-~150mv to -550mv) witll incubation of n facultative anaerobe, provision is rnade to offset the potential r eacling using OA2 to present a unipolar, positive sigrnal o~ adjustable gain to the computer A/l~ converter.
Immediately following the potential reading, relay 11 closes, and a positive-going pulse derived from the computer D/A converter is applied to OA3.
The inverted, unity-gain output of 0~3 causes current to flow through the selected cell in proportion to the dissolved oxygen content. The cell current is sensed by OA4, connected as a current-to-voltage converter. The output of OA4 is sampled by a second input of the A/D converter card prior to the termination of the voltage pulse from OA3. ~11 relays are switched off and allowed to settle prior to the selection of the next cell to be tested, at which point the process begins again for the newly selected cell.
After all cells have been tested, all relays are deselected while the programmed time interval between readings, usually ten minutesJ is allowed to elapse.
The voltage signal input resistor (Rinp, OAl in ~igure 4) has the value of 2~2Megollrns. Although this is by no means an electrornetric input resistnnce ns is normnlly employecl to menslll e electroallalytical cell potentinls, it must be remembered tlult the cell is loaded with this resistance for only a few h~lndred milliseconds before the voltage reading is stored, and that the foLlowing voltage pulse clrives the cell far ~rom equilibrium in any cflse. The 2.2M resistor provicles a goocl compromise between ceII
loadin~ and noise pickup in the ceII environment.
The eell current and potent~al values measured respectively durino;
ancl between the successive applied pulses can be cornparecl to the initial values to determine when the thresholcl of detection has been reached. This process is facilitated by normalizntioll of the conected values as clescribed more fully in Example 2. ~Vhen the current level has fallen to a predetermined percenta~e of the initial value, e.~., 60-80%, detection is found to hnve occurred. It is also possible to malce deterrninat;ons by comparing the collectecl datfl to that obtalined in a separate reference well.
The method of the present invention can be used in any application where the detection of microorganisms in a sample is desired. This method finds particular utility in the detection of bacteria in biological fluids such as blood, urine and the lilce.
~ s will be reaclily appreciated by one skillecl in th~ art, bacterial testing can include screening, identi~ication, and antibiotic susceptibility testing. Other ureas of utility include the detection of microorganism contamination in food, pharmaceutical and cosmetic products.
The following examples are included for illustrative purposes only and are not intended to limit the scope of the invention.

E~ PI.E 1 - This example demonstrates the selection of optimum values of the voltage pulse amplitude and duration.
In order to separate the effect of pulse width from that of pulse amplitude in a semi-quantitative fashion, a test was conducted to determine the cell current at various pulse widths for applied voltages ranging from -O.OOV to -1.00V in steps of 0.05V in a cell eontaining a spent culture. As in all examples, the platinum electrode received the pulse~ while the reference electrocle ~as helcl at virtl1al grollnd. r~ e clltlent-to-voltlge conversion gain ~vas held constant.
A sing~le sample cell o~ an at rny consisting of 8 cells as shown in FigLIrc I ~vas prepared using l.0cc of molten TSA (~ûC) transferred to the t cell Vicl sterile syringe. 5.0cc o~ a fully-grown E. coli cul~ure in 6B medium (oxygen depleted~ was similarly transferred to the cell a~ter the agar had solidi~ied. The cell m ray was plnced in a wat m-air incubator at 37 C and allowe(l to equilibr~te Ior 40 minutes, at which time readings were begun under computer control. ~\ voltage scan Oe the cell WflS obtaiTled and plotted for each chosen pulse width. The interval between pulses was nppro~imntely 10 seconcls, depending somewhnt upon th~ selected pulse width.
C~ll culrent tesponses for selectecl pulse widths are presented in Figure 5. It may be noted tha t the desired cell response is achieved with pulse widths greater thnn about fi4(1 milliseconds. ln practice, pulses of about 1200ms. duration 11nve beed used with goocl success. Because cell current begins to fall rapidly from the instant the pulse is appliecl, there is little to be gained ~rom the use oE a pulse width in excess OI that necessary to produce tile desired response. This effect may be noted graphicaUy in Figures 6 and ~ which depict the results obtained when a single cell containing 1.0cc TSA and 5.0cc 6B medium is subjected to a continuous applied voltage oE -0.70V, with current readings obtained every û.5 seconds beginning 0.5 seconds after vol-tage application (Fi~. 6). In Figure 8, one point is plotted for each 20 collected, stored and tabulated.
The cell current has decreased to less than half its value at 0.5 seconds (217) after only 4.0 seconds (106); after 5 minutes, cell current has fallen to 16.6% of the initial measurement value (36), yet steady-state conditions have not been achieved. Once the desired cell response of low residual ~urrent when containing microbiologically reduced med;um has been achieved7 further increases in pulse width merely decrease the desired current signal and displace the cell further from thermodynamic equilibrium (Fig 5).
Conversely, pulses shorter than about 600ms provide unsatisfactory results.
Figure g illustrates a voltage scan overlay for two cells, one containing l.Occ of a fully-grown culture oE E. coli in 6B, the other containing sterile 6B medium, both ineubated at 37~ C for 4 hours prior to measurement. A
pulse width of about 1300ms was employed. The cell responses are observed .' r to scparate nt abollt -0~35V~ r~acll ma~;i1nl1m c1ivergence near ~0~75V~ and converge ngnin nhollt -O.~)OV. Tlle cletection of microorgnnisms, tl1en, requires a pulse potential near -0.7SY for best efficiency under t11ese conditions.
Consicleration of the results presented in Figures 5 and 6, together with similar c1ata obtained during the early phases of experimentation Ied to the selection of -0.70V as the pulse amplitude, with ~ wic1th o~ about 1200ms for use in the remainder of the examples~
It is also observed that several minutes are required for the cell to return to conditions approaching equi1ibril1m fo1lowing the application of a voltage pu1se. Ideally, the time bet~veen pulses shoulc1 be very large compared to the applied pulse wic1th. Ten min~1tes was selected as the sampling interval, since it is a short period of time microbiologically speaking, yet it provides for a pulse duty cycle of only 0.2~,~ while at the same time permitting the acquisition of semi-continuous data for rapidly growing orgcmisms. Even at this low duty cycle, measured ceU potential values are depressed somewhat from their equilibrium values. The measured potentials do, however, adequately represent relative reclox potentiRl changes in the grow th m edium .
Finally, in order to insure that dissolvecl oxygen content is in fact the major parameter determined by the pulse polarographic technique under the chosen experimental conditionsJ a single cell o~ the 8-cell array was prepareà as usual, containing l.Occ TSl~ and lO.Occ 6E~ medium. The cell was incubated at 37 C for 30 minutes~ then read every 60 seconds for lO
minutes. Dry nitrogen gas (Matheson, High-Purity) was then bubbled through the cell by means of a disposable glass capillary pipette. ~e~dings were continued until the cell current dropped to a reasonably constant value. In a separate experiment conducted in the same cell, nitrogen was again bubbled through the cell, and readings begun just prior to reaching the purge plateau.
The gas supply tubing was quickly transferred to a small aquarium pump ~ith flow constrictor, Imd readings continued for twenty additional minutes.
Data from these tests is plotted in I; igs. 9 ancl 10. Cell response is shown to be clearly related to the oxygen content of the solution; response also appears to be totally reversible in the case of this sterile cell In actual use with bacteria, however, cell reversibiIity becomes a strong function of the amount of time the cell is exposed to a fully-grown culture.

I

EX~MPT.E 2 In nll tile following e:calTIples, electrochemical measurements~ unless otherwise notecl~ were carriecl ont in an S-ce11 arrny using platinum (0.035"~
ancl silver/silver oxicIe (0.040") wire electrocle~; of l.O~m length separated hy l.Ocrn. Prior to each run, the cel1s were carefully rinsed with deionized water nnd vigorously shaIcen to clislodge the larger droplets. I.Occ TS~
(Trypic Soy ~gar~ ~O.Og/17 ~BL, Cockeysville, MD) at about 90 C was then transferre(I to each cell usin~ a sterile 3.0cc syringe with 18ga. needle.
The arrRy ~,vas then covered, and the agar ~11owed to solidiey. 5.0cc sterile 6B medium (Johnston LaboIatories, Inc., Cockeysville, MD) was then added to ~ach cell, together with O.lcc of a ~.5g/20ml stcrile glucose solution and/or O.lcc of a 1.5g/20ml sterile glycine solutioIl flS desired. The prepared ce11s were covered ~md set aside while the inoculum dilutions and pOUl- plate dilutions were preplL~ed.
~ resh cultures of the orgaIlisms to be studied were prepared in 6B
medium the clay before each test, ancl al1owed to incubate at room temperature until neecIec]. Previously preparecl sterile 20cc vials fitted with rubber septa and aluminum closures containin~ 9.0cc TSB (27.5~/l,BBL) were used for all inoculum dilutions. Similar lOOcc vials containing 99~9cc 1/2-strengrth TSB were used to prepare pour plate dilutions. The test was initiated by preparing n sterile l.Occ syrin~e containing about 0.5cc of the overnight culture of the desired organisrn. This culture was adcled dropwise to a 9.0cc dilution vial with agitation until visual turbidity was achieved.
After complete mixing, l.Occ was removed from this vial and used to inoculate a second (X O.l) vial containing 9.0cc, achieving 1:10 dilution. l.Occfrom this vial was used to inoculate a third (X 0.01) vial, whieh was in turn used to inoculate a fourth tX 0.001). ~dditionally, O.lcc was removed from the X l.O vial and used to inoculate one 99.9cc vial to obtain l:1000 dilution for pour plate preparation. Similarly, O.lcc was withdrawn ~rom the X 0.01 vial and used to inoculate a second 99.9cc vial to obtain 1:10(),000 dilution. A fifth vial containing 9.0cc TSB was used as the source of sterile, control inocllla.
l.Occ of the X 1.0 vial was then transferred to cell (1) of the array.
Cells (2) and (3) received l.Occ each from the X 0.1 dilution vial; cells (4) .1,~ g and (5) were inoculated with l.Occ flom the X o~nl vial, while cel1s (6) and (7) cach receivecl an inocul~lrn from the ~C 0.001 vinl. Cell (8) was inoculatedWitll l.Occ sterile l`SB to serve ns the control.
l`he inoculated cells were placed in a warm-air incubator held at 37 C ~ 0.2, connected to the analog conditioning electronics, and the measu1ements begurl. No preinoculation incubation of the array was employecl. Cell current and cell potential readings were obta;ned at 10-minute intervals using a pulse nmplitude of -0,70V ancl a pulse duration of 1200 rnilliseconcls.
Dllplicate pour plates were prepared containing 10-15ml TSA t40g/1) usin~ l.Vcc and O,lcc from each of the ~9.~cc clilution vials to yielcl pairs of plates at 1:103, 1:10~, 1:105, ancl l:lOfi dilutions. The plates were allowedto harden at room temperature prior to 24-hour incubation, Details of the sampling and clilution scheme are set out in Figure 11.
All values of current and potential recorded by comp~lter in these examples range from O to 255 as a c~-nsequence of unipolar 8-bit conversion of the input signals. These raw data values are storecl in the appropriate memory array during the experiment. All data manipulation is performed after the experiment is terminated. I~aw data and experimental specifics are storecl on a floppy disk for future retrieval.
Cell current l(T,N) as a function of time (T) and sample number (N) is normalized at a chosen time interval ll for sample N by dividing cell currents observed at all times T for sample N by the cell current value observed at time T=Tl, and then rmultiplying by lOn.O, e.g.: -X(T N) = I(T7N) x 100 0 I(Tl,N) 'rhe same normalization time (Tl) is used for all samples. The normalized current values X(T,N), now ranging nominaUy from O to lOO, are then scaled for plotting $hrough division by a scale factor, herein 2.0, so that aU data together with the machine-generated coordinate time axis will fit on the 80-character CRT/printer line~ Cell currents for each sample are thus easily presented as ~ percentage of the normalized current value, usually taken as the current value observed after 30 minutes experiment time.

- lR -Ccll potclltial res-llts V( r,N) are normali~ecl by simple ~-axis translation alld ulliforrn scnlin~. ~ constant is first derived from the voltageobserved at the normalir~ing time intcr val Tl:
, C = V(Tl,N) - 2n whiclr is in turn llsecl to translate all observed voltage vaIues for a given sample N:

Z(T,N) = V(T,N) - C

The translatecl values Z(TtN) are then scale~l to page width by dividing by the scale factor, herein 3.5? and then printed. Potential norrnalization is usually performed a-t 60-100 minutes after the experiment hns begun. Cell potential readings require about 60 minutes longer, on average, to stabilize than do the current readings.
For the purposes of the fo~lowing examples, detection of the organism is said to occur when the pulse voltammetric cell current hns fallen to 80%
of its value at the normalization time interval. Potential measurements are considered positive when R relative normalized value of 20 is attained.
All values obtained for cell (1) were used as high-inoculum marlcers only, and do not appear in the results.
This example demonstrates the detection and quantifieation of E. coli~
A fresh overnight culture of E. coli in 6B medium was used as the inoculum souree. The sampling and dilution scheme of Figure 11 was employed to prepare sample cells and pour plates. The sample cell medium was enriehed with O.Icc (oî the~ 4.5g/20ml glucose stock solution. Ineubation of alI vials, sample cells and plates was at 37~ C. Cell readings were continued for 8 hours.
Pulsed voltammetric current responses for three deeade dilutions of the organism plus control are shown in Figure 12. The related cell potential results are presented in Figure 13. The Y-axis indicates potential increasing in the negative direetion. The short plateau evidenced at relative potential values Oe 30-40 probably indicates a ehange in the metabolism of the organism triggered by the redueed oxygen tcnsion in the medium. Both eell parameters _ 19 _ give times-to-detection which vary in a predictable manner with inoculum strength. Silver cathodes have also been used without rendering the process inoperative . Pour plate results indlcate that 1. O x lO
cfu/ml E, coli were present in the freshly inoculated X O, 1 cell .
Times-to-detection Eor the duplicate cell current and potential measurements are presented in Table 1.

Initial Inoculum Cell Current Cell Potential in Cell _ B A B
1.0 x 105cfu/ml 120 120 160 180 1.0 x 1o4 " 150 160 220 230 1.0 x 10 " 200 200 260 27U

Table 1 l`imes-to-I)etection in Minutes for Cell Current and Cell Potential for the Organism E. coli EX~MPLE 3 This example demonstrates the detection and quantification of E.
cloacae. A fresh culture of E. cloacae was incubated overnight at room temperature to serve as the inoculum source. The sampling and dilution scheme presented in Figure 11 was used to prepare duplicate sample cells and pour plates. The cell medium was enriched with 0.1cc glucose stock solution. All incubations were performed at 37~ C. The inoculated ce'll array ~as covered with clean aluminum foil~ placed in the incubator, and connected to the analog conditioning electronics moments before the start of the test. The test was continued for 520 minutes (8-2/3 hours).
The cell current response for each of the three decade dilutions of E. cloacae plus control is presented in Figure 14. Cell currents are seen to rise relatively rflpidly from their attained minimum values probably as a consequence of electrode-active metabolic products synthesized by the organism in its latter stages of growth under reduced conditions. The related cell potential data is shown in Figure 15. A short plateau in redox potential values is again noted at relative normali7,ed values between 30 ~ "~

- 2n-and ~0, ag~in attriblltecl to an or~mism metabolic pathway change. Duplicate pour plate counts were used to determine the initial inoculum level in the X 0.1 cells to be 3.7 x 10 cfu/ml. The incubation times required to detect the or~anism are listed in Table 2.

Initial Inoculum Cell Current Cell Potential in Cell ~ B ~ B
.
3.7 x 105c~u/ml 150 150 230 200 3.7 ~c 10~ " 210 210 260 2~0 3.7 x 103 " 270 270 300 310 Table_2 Times-to-Detection in Minutes for Cell Current and Cell Potential for the Organism E. cloacae EX~MPLE ~

This example demonstrates the detection and quantification of P.
mirabilis. ~ fresh overnight culture of P. mirabilis in 6~ medium w~s used as the inoculum source. Sample cells and pour plates were prepared as per the sampling and dilution scheme presented in Figure 11. 5.0cc 6B
medium was enriched with 0.1cc sterile glucose stock solution in each of the cells. All incubations were performed at 37 C. The electroanalytical measurement was begun immediately following cell inoculations without pre-inoculation incubation, and was continued for 540 minutes (9 hrs).
Pulsed voltammetric cell current responses for the three decade dilutions of the organism are presented in l~igure 16. Cell responses appear normal as oxygen is consumed and cell current falls to the residual level, then rapid vertical transitions appear lasting only 20-30 mim~tes. Cell current values seem to stabilize following these events, but do not return to pre-transition levels.-Cell potential responses are shown in l?ig-lre 17. ~gain, a .slight plateau is observed at normalized relative potentials between 30 and 41) urlits. Instcn(l o~ the rapiclly incrensill~ negat;ve potentials noted for E.
coli nnd 1~. cloacae, P. mirnbilis potelltials fall sharply after a sli~ht increase following the plateau. The time intervals noted for this potential decrease correlate well with the observed vertical transitions in cell current previouslynoted. Since P. mirabilis is a facultative org~mism known to efficiently .
reduce growth media, and has been used as a standard organism for medium reduction measurements, these flnomalous results in the long-incubation regime are best explained by the formation of electrode-active nnetabolic by-products, most probably sulfide-containing molecules (H2S, CH3$CH3~
CI13SGH2C~I3 etc.) which certainly would perturb the clectrocle system. The silver/silver oxide electrocle is noticenbly blackened by ea~posure to P
mirabilis for extendecl periods. The electrodes clo not seem to be permanently dnm~ged by such exposure, and may be returned to their initial conàition by careful washing nnd wiping of the electrocle surfaces. The discoloration can be removed only by rnechanical polishing. The X0.1 cells (2) and ~3) contained 2.0 x 105 cfu/ml of r. n!irabilis at inoculation as determined by duplicate pour plate counts. Times-to-detection for cell current and cell potential are listed in Table 3.

Initial Inoculùm Cell Current Cell Potential in Cell A B A B
_

2.2 x 105cfu/ml 190 170 250 230 2.0 x 10a~ " 220 230 270 280 2.0 x 103 " 290 300 3A~0 33 Table 3 Times-to-Detection in Minutes for Cell Current and Cell Potential for the Organism P. mirabllis This example demonstrates the detection and guantification of P.
aeruginosa.
A freshly inoculated vial of 6B medium was incubated overnight at ~ i5~r7;~

- ~2 --roorn ternperature for use ~s the soul ce of inocula. 'I'h~! samplirlg and dillltioll scheme illustrated in Figure ll wa.s emp~oyecl to prepnre sample cells and pour plates. The sampl~ cell medium was enriched with 0.1cc sterile 1.5g/20rnl glycine stock solutiorl in each cell~ All cells and plates were incubated at 37 C. Electroanalytical cell readings were b~gun immediately a~ter inoculation and were continued for a~90 minutes (8 1/6 hours).
Cell current response of the organism with decade dilution and of the control cell containin~ sterile medium nre shown in Figure 18. Normal cell current behaviol is observed. The related cell potential responses are presentecl in ~:igure l9. Becnllse P. aerutrinos_ is a relatively slow growing obligflte aerobe, cell potential response at en~!h inoculum level changes more slowly ancl reaches a limitin~ valtle oî considerably less amplitude than do the fncultative anaerobes. Pour plate COUlltS in d~lplicate were used to determine the initi~l inoculum level in tlle X0.1 cells as 7.7 x 10~ cfu/ml.
Times-to-detection for the detection methods are presented in Table ~.

Initial Inoculum Cell Current Cell Potential in Cell A B A B
7.7 x 10~cfu/ml 190 160 220 170 7.7 x 103It 220 250 2a~0 310 7.7 x 102.~ 300 310 390 390 Table 4 Times-to-Detection in Minutes for Cell Current and Cell Potential for the Organism P. aeru~inosa , This example demonstrates the detection and quarltification of S.
aureus. A freshly inoculated vial of 6B medium was incubated overni~ht at roorn temperature to serve as the source of all inocula. The sampling and dilution scheme of Figure 11 was again employed to prepare sample cells snd pour plates used in the test. The growth mediunn in eacll cell wrs enriclle(l with ().lcc sterile ~1ucose stocl; solution. AlL incubations were carried Otlt at 37 C in a wnrm-air incubator. Cell current and potential readings were reco~clecl every 10 minutes uncler cornputer control. The experiment was continued for ~8~ millutes (8 hours). Cell current response for the organism nt decade inocu1um levels is shown in Figure 20. Norrnal cell current behavior is obtainecl. The related ceLI potential variations are presented in Figure 21. Note thnt very little potential change occurs with continued growtll o~ S. aureus; thresho1d cletection is barely achievecl.
Dup1icate 2~-hour pour plate COUIltS were used to determine the inoculllm level in the X0.1 cc11s to be n.2 ~ lO~ cflltml. Times-to~detection for cell current and cell potenti~l methocls are given in Table 5.

Initial Inoculllrn Cell Currellt Cell Potential in Cell ~ B ~ B
8.2 x 105 c~u/ml l50 130 290 200 8.2 x l0~ " 180 1~0 250 260 8.2 x 1û3 " 250 260 320 3a~0 Table 5 Times-to-Detection in Minutes for Cell Current and Cell Potential for the Organism S. aureus EX~MPLE 7 This example demonstrates the detection and quantifiction of S. bovis.
A fresh overnight culture of S. bovis in 6B medium was used as the inoculum source. Sample cells and pour plates were prepared with reference to ~igure 11. Eflch cell also received 0.1cc glucose stoclc solution as enrichment. All incubations were carried out at 37~ C. The measurements were continued for ~5~ minutes (7 112 hours). Cell current measurements are presented in Fi~lre 22~ Normal current response is observed prior to the current minimum at each inoculum level. Values recorded after each rninimum r ise more rapidly then usual. Cell potential responses are shown in Figure 23r S~

`

- 2~ -bovis causes little change in the potcntial observed as growth progresses, save for the srnaII singularity usually observed in conjunction with the cllrrent minima in Figure 22. Duplicate 2~1-hour pour plate counts indicated 6.5 x I05 cfu/ml to be present in the ~o.i cells initially. Times-to-detection for the detection methods are presented in Table 6.

Initial Inoculum Cell Current Cell Potential in Cell A B A B
6~5 x 105 cfu/ml200 180 280 2~0 6.5 x 104 " 220 220 290 280 6.5 x 103 " 280 290 330 340 Table 6 Times-to-Detection in Minutes for Cell Current and Cell Potential for the Organism S. bovis EXAMPLE_8 This example demonstrates the extended quantification of E. coli using pulsed voltammetric detection. A fresh overnight culture of E. coli in 6B medium was used as the inoculum souree. Dilution vials and pour plates were prepared with reference to Figure 11, except that dilution vials were prepared out to a dilution ratio of 1:106. The growth medium in each of the cells was enriched with 0.1cc sterile glucose stock solution. The first seven cells in the 8-cell array eaeh received 1.0ce from the appropriate dilution vial as inoeulum. Cell (8) reeeived 1.0ec sterile ISB inoculum as the eontrol.
Beeause the cells had previously been meehanically po3ished and reeonditioned, the current sensitivity for this experiment was reduced somewhat to insure that all recorded values would remain within the dynamie range of the A/D
converter. Incubation and testing were carried out at 37 G. No pre-inoeulation incubation period was used.
The normalized cell current responses of cells (3) through (8) are shown in Figure 24. Cell (1) was used as a high-inoculum marker only, since the cell assembl~v rcc3uires nt least 30 minutes to nttnin temperature cquilibriurn. Cell (2) reslllts were nllomnlolls Witil respect to cletection time, prohnhly due to sligilt contaminatiorl of tlle c ell walls or clectrode surfMcesduring reconditionin~, nnd are not shown. Over tlle foul decades of inoculu~n level considered, t;m~to-cletection is seen to vary linearly with the logarithm of inoculum strength. Note that the detection threshold has been reduced to 6û,~ o~ the value observed at tlle time of normali~ation; this provides more clepenclable tirme-to-detectioll vallles in prolonged tests where the slight downwarcl baseline clri~t with time can gellerate cletection times slightly shorter than the correct vnllles. Note that the 60% current level occurs near the point of ma:~imum slope of the ~owth curves. Experience with the system has shown thnt best qllanti~ication when usin~ this technique is obtained when the detection time is taken to be the time at which ma~cimum slope of the ~rowth curve is eviclenced. ~igure 2S illustrates the good quantification achieved for inocula of 1.5 x 105 cfu to 1.8 x 101 cfu in the present exarnple. Repeated trials with the system has shown that inocula greater than about 5 x 105 cfu require sligh-tly longer to detect than predicted. This is pai tly due to the lag induced by the time required for the cell array to reach temperature equilibrium. The remainder of the problem is most likely caused by the finite time required for the orgranisrn in question to acl~pt to the new environmellt imposecl by dilution and sampling. The short time interval between recorded data points (10 minutes3 makes cven slight devia-tions from expected behavior noticeable.

, .
EX~MPI.E 9 The results for the organisms studies in Examples 2-7 using the û-cell array (platinum cathodes) are summarized in Table 7. Data from parallel radiometric tests (BACTEC) are included for comparison. Cell current duplicate resl~ts are shown to differ by a maximum of 30 minutes at any inoculum level for all organisms. Cell potential detection is less reliable for quantification, differing in duplicates by as much as 90 minutes in one case (S. aureus).
.
Pulsed voltammetric detection of the test organisms compares well with detection based upon the B~CTEC system; E coli and E. cloacae are - 2fi -.

ietcct~cl with nppro~cimat~ly ~!qual ~acility by both m~thods. P. mirubilis ~ncl most notably, P. aerllginosa are detec!ted si~ni~icantly faster usintr the cell culrent measurement. ~etection of S. aureus by the cell current method is abo~lt 40 minutes faster then BACTE:C, while S. bovis cletec-tion is accomplished about 1 hour sooner by the B~CTEC system.
Cell potential detection of organism growth compared to either the cell current cletermination or to the B~CTEC system leaves much to be desired. Results can be quite unreliable for organisms such as S. aureus (Fig. 21) and S. bovis (Fig. 23) which produce little chflnge in solution redox potential with ~owth. Thresholds for detection are approached slowly and barely exceeded by such organisms ns compared to the Enterobacteriacae, thlls promoting a test of widely varying sensitivity as a function of the organism being deteeted. The widely àifferin~ reclox potential patterns do, however, provide good clues as to the type of organisrns simultaneously detectecl by other means.
Table 7 _ TIMES-TO-DETECTION, MINUTES
Tested Inoculum Level BACT~C Cell Current CeU Potential Or~anism in Cell or Vial ~ B A
E. coli 1.0 x 105 cfu/ml 120120 120160 180 1.0 x 103 cfu/ml 180150 160220 230 1.0 x 10 cfu/ml 240200 200260 270 F.. _acae 3.7 x 10,~cfu/ml 120 150 150 230 200

- 3.7 x 103 cfu/ml 180210 210260 240 3.7 x 10 cfu/ml 240270 270300 310 P. mirabilis 2.0 x 104 cfu/ml 240I90 170250 230 2.0 x 10 cfu/ml 300220 230270 280 2.0 x 103 cfuiml 420290 300340 330 P. aerll~osa 7.7 x 104 cfu/ml ~80190 160220 170 7.7 x 12 cfu/ml 600220 250240 310 7.7 x 10 cu/ml 660300 310390 390 ~ 27 ~
,1' Table 7 (Contld) r .

TIMES-TO-DETECT~C)N, MINUTES
Tested Inoculurrl Level BACTEC Cell Current Cell Potential Orgnnism in Cell or Vinl A B A B
S. aureus 8.2 x 105 cfu/ml 180 150 130 290 200 8.2 x 103 cEu/ml 240 180 180 250 260 8.2 x 10 c~u/ml 300 250 260 320 3~0 S. bovis 6.5 x 105 cfu/ml 120 200 1~0 280 2~0 6.5 x 10~ cfu/ml 180 220 ~20 290 280 6.5 x 10 cîu/ml 2~10 280 290 330 340 .

Summary OI Results Times-to-Detection for all Organisms by All Methods E~PLE 10 .

This example demonstrntes the use of electrodes of other materials and sizes then those used in the fore~oing examples. A 6-ccll array constructed with 0.020" gold wire as cathode and 0.020" silver wire as anode in each cell was used to test the response of the system toward pulse polaro~raphic detection only. The gold cathode prevented collection of potential data, but gold is somewhat less expensive then platinum to use in cases where detection alone will suffice. The gold and silver wires, arranged in parallel fashion in the bottom of each cell, were each 1.4cm long and were separated by 1.6cm. The observed cell currents were appreciably lower than those observed when using the 8-cell array with larger diameter wires and platinum cathode. No attempt was made to measure true current sensitivity. The cells were prepared with either l.5.cc or 2.0cc TSA and 5.0cc enriched 6B
medium, and were tested with various organisms under the same conditions as for the previously noted experiments. No electrode pretreatment was employed. The pulse arnplitude (-D.70V) and the pulse duration (1200ms) were the same as in the previous examples. Duplicate pOUI plates were also prepared for these experiments. Cell current data normfllization was ~:~

-- 2~ --cnrried ollt as previollsly describetl. I~resh overni~rht cultul es of ~L P-mirabilis and P. fleru~inostl itl 6B med;um were useci as inoculum sources.
All snmple flnCI pour plate preparations were preformed with reference to Figure 1l. Cell current readin, s were obtained at 10-minute intervals. No pre-inocnlation incubation was employecl.
Results obtained for separate tests us;ng E. coli, P. mirabilis and P.
nerutrinosa are shown in Fi~ures 26, 27 nncl 28, respec~ively. The organism inoclllllm level for the most concentrated cell in each experiment is listed below:

E. coli 1.5 x 105 cfu/ml P. mirabilis 1.2 x 10 cfu/ml P. aeruginosa 1.~3 x 105 cfu/ml `

The other tracings in each Eigure are for decade dilutions and a sterile control cell. Tn all cases, detection of the organism was readily accomplished, as indicated by a drop in recorded cell current to 80% of the normalization value. Times-to-cletection for all dilutions in each of the figures are seen to depend upon the inoculum level in a predictable manner, varying essentially as the logarithm o~ inoculum conce~ntration. The slight scatter noted in som e of the data is due to stray pickup by the analog conditioning electronics, a capacitor was addecl across the voltage-to-current conversion operational amplifier to alleviate this problem.` Downward baseline drift noted for the control samples is most likely due to the lack of any preconditioning of the cells. Any increase in times-to detection noted at a given inoculum level of a specific organism when using the 6-cell array is probably a function of the thickness of TSA over the electrodes, particularlyin the case where 2.0cc was used. Once the electrodes are covered, detection times increase as the level of TS~ in the cell increases as a consequence of the increased diffusion path. 2.ûcc TSA was used in the experiment run with P. aeruginosa. The E. coli and P. mirabilis experiments both used 1.5cc TSI~.

OJ`,.~'~'t.~
..
~

~ 9 ~

EXAM PI,E 11 This exa1nple clemonstrates a clinical trial of the pulsed voltammetric detection techniqlle of the present inventor for the detection of si~nificant bacteriuria. The test was conc~lcted in conjunction with Sinai ~Iospital of Baltimore. Over the 33-dfly period o~ the study, 389 urine samples were collected from Sinfli nnd testecl at Johnston Laboratories using the pulsed voltammetric technique in parallel with the BACTEC radiometric system.
TSA pour plates were employed to checlc actual organisrn counts.
The test was carried out as follows. Urine specimens sampled foLIowingr Sinni eollection ancl plnnting were pielced up from the hospital at approx;mately U:OOAM each clay. Prior to sample arrival at JLI, the 16~
eell assembly to be used with the P~llsed Voltammetric instrllmentation was filled with scaldin~ hot water and aUowed to stand for at least ten minutes.
The assembly was then rinsed twice with sterile deionized water, then shaken vigorously to dislod~e any large water droplets. I.Occ sterile Tryptic Soy Agar (~O.Og/l~ BBL or l)IFCO~ at about 95 C was then added by sterile syringe to each cavity of the assembly. The assembly was then covered with a double thicl~ness of aluminum îoil sterilized with isopropanol, and the agar allowed to solidify. 5.0cc of sterile TSB (27.5g/1; BBL or I~IFCC)) containing De~trose (2.5g/1; M~\~LINCKRODl` or J.T. BAKER) was then adcled to each cavity via syringe, and the cover replaced.
Upon arrival at JLI, urine samples were cataloged as to JLI daily and conseeutive sequence numbers, Sinai reference number, and gross physical characteristics. l.Oec of eaeh urine specimen was inoculated via syringe into one cell of the assembly. Similarly, l.Occ was used to inoculate a septum-fitted vial of nominal 50cc capacity containing 5vOec ~ JLI ~A
Urine Sereening Medium (JLI B/N 0379ûlU-1.5uCi/vial) for use in the para}lel BACTEC study. O.lcc of each speeimen was inoculated into previously refrigerated, septum-fitted vials containing 99.9ce 1/2-strength TSB to obtain 1:1000 sample dilution for the pour plate studies. Plates were prepared at 1:103 and 1:104 dilutions for each sample using 10-15 ml TSA ~40.ûg/1) and l.Occ and Olcc from each dilution vial, respectively.
The inoculated cell assembly was placed in a 37C warm air incubator without agitation and pulsed voltammetric testing begun under compu~er ~s~

cont~ ol. TC!St vnllles were r ecor~le~l for all samples at 10-m;n~lte intervals.
Dntn nol mali~ution ~vas bnsed uporl sample dnta v~llues recorclecl after the first 10 minutes for tabulation; all datR values recorded for each sample were ultimately expressed as a percentage of the 10-minute value. A sample was considered to be positive when the normalizecl data value for that sample at any given time interv~l a~ter normalizfltion ~eU below 70 or rose above 140. The latter criterion was employed to permit detection of some highly positive tca. 10~ cfu/ml) samples which produced data minima slightly greater than 70, yet whieh inter~ered with normal operntion suf~iciently to procklce data maxima over 1~0.
Each ~lay of testing concll~cled with the generation of a computer printout which inclllded plots of relative cell potential readincrs~ and a tableof normalized cell current readings for each sample, all as n function of incubation time. The normali~ed, tabulated results were used to determine sample result classifications.
Of the 389 tested samples, 45 were omitted from the s-tudy usually for experimental reasons (Incubator Pailure, 12; CeU Reconditioning Failure, 14; Contaminated Petri Dishes, 12; JLI/Hospital Datfl Discrepancy, 7).
Contaminated samples numbered 30, and were similarly omitted from further consideration. Of the remainin~ 31~L samples, 8a~ were considered significant clinically. Table 8 lists the organisms identiEied by Sinai Hospital found to be present in the sigrnificant samples. The number of samples containing each organism is also noted, as is the percentage of the total containing that organism. Non-integer sample numbers are due to samples containing more than one organism.

Or~anism No. of Samples Percentage E. coli 3a~ 5 41.07 r. aeru~inosa 11).5 12.50 K. pneumoniae ~.0 10~71 .

P. mirabilis 6.0 7.14 ~east, unspecified 3.5 d~.17 S. aureus 3.0 3.57 C. albic ns 2.5 2.98 Grp. B, Beta Strep 2.5 2.98 ~ ~~~

7,4~

~:~n No. of S~lmE~ Percenta~e Grp. 1) Strep 2.0 2.38 S. alblls 2.0 ~.38 ... .
K. o~;ytoca 1.5 I.79 E. aero~enes 1.0 1.19 E. n~g~lon~erans1.0 1.19 E. cloacne 1.0 1.19 C. freunclii 1.0 1.19 .
P. morf~anii 1.0 1.19 Grnm-Negative Rod1.0 1.19 P. stunrtii 0.5 0.60 . . . _ _ _ _ S. marcescens 0.5 0.60 8~.0 100.01 T~BI.E 8 Organisms Contributing to Significant ~amples Results of the test for the 314 samples consiclered for data analysis are set out in l'able 9.

True Positives 80 (25.48%) True ~egatives 215 (68.47%3 Palse Positives 15 ( 4.78,o) False Negatives 4 ( 1~27%) 314 100.û%

T~BLE 9 OI the 84 samples considered clinically significant, the pulsed voltammetric method detected 80 (95.24%3. Of the four samples missed by the pulsed voltammetric technique, two were missed by BACTEC as l,vell.

ZiJI
One of these is known to be from a patient receiving anti-biotic therapy; the other contained S. Aureus. The remaining two false negative samples contained P. Aeruginosa and an unspecified yeast. The BACTEC system detected 76 ~90.48%) of the 84 samples considered clinically significant.
The pulsed voltammetric detection technique properly identified 95.24~ of urines considered significant in the study, yielding a total sample false negative rate of 1.27%, with a false positive rate of 4.78%. The BACTEC system detected 90.48~ of significant urines, with a total sample false negative rate of 2.55~ and a false positive rate of 1.27~. BACTEC
required 3 hours to achieve the reported level of performance;
the pulsed voltammetric technique required 4 hours.
The technique of the present invention is thus shown to provide competent detection of significant bacteriuria, with clinically acceptable levels of false negative and false posi- !
tive results. The rapidity and sensitivity of the method compare favourably with parallel results obtained using the BACTEC
system.
While certain specific embodiments of the invention have been described with particulars herein, it should be recognized that various modifications thereof will occur to those skilled in the art. Therefore, the scope of the invention is to be limited solely by the scope of the claims appended hereto.

; ~ -32-, ~ ................................................. .

Claims (17)

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. An electroanalytical method for detecting the presence of oxygen-consuming microorganisms in a sample com-prising the steps of:
a) providing a mixture of said sample and a fluid culture medium capable of supporting microorganism growth in an electroanalytical cell equipped with two electrodes which are in contact with said mixture;
b) applying a series of voltage pulses of sub-stantially constant amplitude and duration across said elec-trodes; and c) measuring the resulting current through said cell prior to the trailing edge of each of said applied voltage pulses;
the presence of oxygen-consuming microorganisms being indicated by a decrease in cell current which is a function of the dissolved oxygen content of said mixture.

2. The method of claim 1 additionally comprising the measurement of the open-cell oxidation-reduction potential across said electrodes during the interval between successive applied voltage pulses.

3. The method of claim 1 or 2 wherein said voltage pulses have an amplitude of from about -0.35v. to about -0.9Ov.

4. The method of claims 1 or 2 wherein said voltage pulses have an amplitude of from about -0.35v. to about O.90v.and a duration of at least about 600 milliseconds.

5. The method of claim 1 or 2 wherein said voltage pulses have a duration of at least about 600 milliseconds.

6. The method of claims 1 or 2 wherein said voltage pulses have a duration of about 1200 milliseconds.

7. The method of claim 1 or 2 wherein said voltage pulses are separated by an interval of about 5 to 20 minutes.

8. The method of claim 1 wherein the cathode in said electroanalytical cell is made from a noble metal.

9. The method of claim 8 wherein said cathode is made from platinum, gold or silver.

10. The method of claim 2 wherein said cathode in said electroanalytical cell is platinum.

11. The method of claim l or 2 wherein the reference electrode in said electroanalytical cell is silver/silver oxide or silver/silver chloride.

12. The method of claims 1 or 2 wherein said elec-trodes are covered with a conductive porous gel.

13. The method of claim 1 wherein said electrodes are covered with a conductive porous nutrient gel.

14. The method of claim 13 wherein said nutrient gel is tryptic soy agar.

15. The method of claims 1 or 2 wherein said elec-troanalytical cell and its contents are maintained in a con-stant temperature environment during said measuring.

16. The method of claims 1 or 2 wherein said elec-troanalytical cell and its contents are maintained in a con-stant temperature environment of about 37°C during said measuring.

17. The method of claims 1 or 2 wherein said fluid culture medium comprises tryptic soy broth.