Biosensors & Bioelectronics 16 (2001) 109– 113

www.elsevier.com/locate/bios

A fiber-optic lactate sensor based on bacterial cytoplasmic

membranes
Sergei G. Ignatov a, Jane A. Ferguson b, David R. Walt b,*
a
State Research Center for Applied Microbiology, Obolensk, Moscow Region 142279, Russia
b
The Max Tishler Laboratory for Organic Chemistry, Department of Chemistry, Tufts Uni6ersity, Medford, MA 02155, USA

Received 6 September 2000; received in revised form 19 September 2000; accepted 21 November 2000

Abstract

A new type of fiber-optic biosensor based on bacterial cytoplasmic membranes (CPM) as the biological recognition element and
an oxygen sensitive dye layer as the transducer is described for the detection of lactate. CPMs from bacteria with an induced
lactate oxidase system are adsorbed onto a cellulose disk. The disk is fixed mechanically over an oxygen sensitive siloxane layer
on the distal end of an optical fiber. This system detects lactate with no interference from glucose, fructose or glutamic acid.
© 2001 Elsevier Science B.V. All rights reserved.

Keywords: Cytoplasmic membrane; Glucose; Bacteria

1. Introduction parameters (Zellers et al., 1996). Recently, biosensors

In recent years, many successful attempts have been
made to utilize the inherent recognition properties of
biological systems to develop highly selective and sensi-
tive biosensors. Biological sensing elements such as
enzymes, antibodies, and antigens are used in conjunc-
tion with electrochemical or optical transducers to cre-
ate such biosensors (Turner et al., 1987; Hu et al., 1993;
Khayyami et al., 1997; Kelly et al., 1998; Minagawa et
al., 1998). A biomolecule’s function and activity are
influenced by its environment and orientation on a
solid surface. It is difficult to control a biomolecule’s
orientation, and hence its reactivity, using direct immo-
bilization methods during sensor fabrication (Lu et al.,
1996; Riedel et al., 1990; Lee et al., 1995; Rainina et al.,
1996; Scott et al., 1997). Formation of Langmuir-Blod-
gett films (Roberts et al., 1983) and self assembled
monolayers (Wink et al., 1997) are the most common
methods for directly depositing recognition elements in
molecular layers on transducers. Biosensors such as
SAW devices based on these thin films are not particu-
larly selective and are sensitive to environmental
* Corresponding author. Tel.: +1-617-6273441; fax: + 1-617-
6273443.
E-mail address: david.walt@tufts.edu (D.R. Walt).
based on novel thin film technologies have been de-
scribed. Examples of this approach include artificial
bilayer lipid membranes (BLM), reversed micelles (Das
et al., 1997) and liposomes (Bennetto et al., 1996).
These types of biosensors simulate a physiological ori-
entation for the recognition elements and offer an
attractive route for the construction of a device useful
for monitoring a large number of biochemical sub-
strates (Tien et al., 1997). Whole bacterial cells have
also been employed as recognition elements for
biomonitoring however, the signal generation of this
type of sensor is quite complex (Ikaryama et al., 1997;
Evans et al., 1998; Jager et al., 1999; Corbisier et al.,
1999).
In order to monitor a biological recognition element,
it is necessary to couple the biological recognition event
to a transduction scheme. For example, dual-analyte
fiber-optic biosensors were used to monitor glucose
with oxygen and penicillin with pH (Li et al., 1995;
Healey et al., 1995). Glucose oxidase was suspended in
a monomer and the polymer matrix was formed di-
rectly over one of two oxygen sensitive fluorescent
sensors on an optical fiber. The microenvironment of
the enzyme-derivitized polymer was monitored while
the macroenvironmental oxygen level, which has a di-

0956-5663/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 9 5 6 - 5 6 6 3 ( 0 0 ) 0 0 1 4 4 - 5
110 S.G. Ignato6 et al. / Biosensors & Bioelectronics 16 (2001) 109–113
rect effect on the glucose oxidase activity, was simulta-
neously monitored. Penicillinase was similarly immobi-
lized over a pH sensitive region and the
macroenvironmental pH was simultaneously moni-
tored. The enzymes used in these sensors were robust
and maintained activity after being suspended in the
polymer. However, many enzymes will not remain vi-
able when suspended in a polymer matrix.
Lactate is an important metabolite for clinical, food,
environmental, and bioprocess analysis. Many lactate
biosensors operate on the basis of the amperometric
detection of hydrogen peroxide generated by lactate
oxidase. Lactate oxidase is a catalyst for oxidation of
lactate to pyruvate with the simultaneous reduction of
O2 to H2O2 (Minagawa et al., 1998). Lactate can be
measured using co-immobilized lactate oxidase and lac-
tate dehydrogenase (Casimiri et al., 1996) or glutamate
pyruvate transaminase and lactate dehydrogenase
(Lobo-Castanon et al., 1997). These schemes require
the addition of NADH and/or DEAE-dextran/lactinol
(Hart et al., 1996) or the use of thermostable soybean
peroxidase (Kenausis et al., 1997). The stability of the
co-enzymes and the support material of these biosen-
sors continues to be improved.
In this work we present the fabrication of a fiber-op-
tic biosensor based on a new type of recognition ele-
ment: bacterial cytoplasmic membranes (CPM). CPMs
are composed primarily of lipid bilayers with embedded
proteins. Bacterial CPMs play an important role in
many biological processes such as respiration and en-
ergy transduction. The ability to control the composi-
tion of enzymes in the lipid bilayer during bacterial
cultivation makes CPMs very attractive for biosensors
(Dong et al., 1993). Isolating the specifically-induced
bacterial CPM and adsorbing it onto a solid surface
maintains its native organization and biological activity
for use in a sensor. We have combined a fiber optic
oxygen sensor as the transducer with the CPM to
fabricate a new type of lactate biosensor.
2. Experimental section
2.1. Materials
All reagents were analytical grade or their equivalent
and were used without further purification. Tris HCl
buffer, tris(2,2%-bipyridyl)ruthenium(II)chloride, ben-
zoin ethyl ether (BEE), sucrose, ethylene diaminoac-
etate (EDTA), lactate, lysozyme, DNAse I, all
inorganic salts, and 3-triethoxysilylpropylmethacrylate
were purchased from Sigma (St. Louis, MO). (80 –
85%)dimethyl(15 –20%) (acryloxypropyl)methylsiloxane
copolymer (UMS-182) was purchased from Gelest (Tul-
lytown, PA). O2, N2 gases and their mixtures were
purchased from Northeast Air Gas (Nashua, NH).
500-mm optical fiber housed in a plastic sheath was
purchased from Spectran Specialty Optics (Avon, CT).
The Blak-Ray Long Wave Ultraviolet Lamp was pur-
chased from UVP (Upland, CA). The PEEK™
(polyetheretherketone) housing was custom made by
Oxford Electrodes (Abingdon, UK). Cellulose discs
type HVLP were obtained from Millipore (MA). Opti-
cal filters were purchased from Omega (Brattleboro,
VT).
2.2. Bacterial cells
Escherichia coli cells (XLI-Blue MRF) were culti-
vated under aerobic conditions at 32°C for 48 h with
continuous shaking at 100 rpm in a medium consisting
of (g/l): Na2HPO4 (6.0), KH2PO4 (3.0), NaCl (0.5),
NH4Cl (1.0), MgSO4 (0.05) (sterilized separately in an
autoclave); CaCl2 (0.005) and 0.4% sodium lactate to
induce lactate oxidase complex. The pH was adjusted
to 7.4.
2.3. Isolation of CPM
Bacterial cells were harvested by centrifugation at
4°C (5000× g, 5 min) washed with distilled water and
suspended in distilled water to a total volume of 50 ml.
The CPM was isolated from bacterial cells as described
previously (Ignatov et al., 1981) using osmotic shock.
The procedure consisted of plasmolysing the cells in
sucrose followed by a rapid dilution in cold water. The
plasmolysis steps were performed in the presence of
EDTA and Tris buffer which were essential to provide
mild conditions for preparing the CPM. Briefly, the
cells were chilled in an ice water bath and ice-cold
reagents were slowly added in the following order: 25
ml of 0.1 M Tris HCl buffer pH 8.3 at 4°C; 20 ml 2 M
sucrose; 5 ml 1% EDTA and 5 ml 0.5% lysozyme
solution. The mixture was warmed to 35°C and kept at
this temperature for 1 h. A total of 0.1 mg DNAse I
was then added to reduce the viscosity, which had
increased due to cell breakdown and DNA release.
Spheroplasts obtained upon disruption of the outer
membrane and cell wall of bacterial cells were lysed by
adding 150 ml distilled water. The CPM fragments were
washed from the cell wall and cytoplasmic fragments by
centrifugation (20 000×g, 45 min) then resuspended in
a small volume of 30 mM Tris HCl buffer (pH 7.4) with
5 mM MgSO4 and stored at −18°C. Immediately
before sensor preparation, the bacterial CPM was ad-
sorbed to a cellulose disk using a Millipore water
vacuum filter system.
2.4. Oxygen-sensiti6e optical fiber
Two methods were used to prepare the oxygen-sensi-
tive fiber optic substrate. In the first method, a thin
 S.G. Ignato6 et al. / Biosensors & Bioelectronics 16 (2001) 109–113
siloxane film containing an oxygen sensitive dye, ruthe-
nium tris(diphenylphenanthroline) (Ru(ph2phen)23 + )
(Ferguson et al., 1997) was prepared and mechanically
held on the distal tip of the optical fiber. The film was
prepared by placing a photo-polymerizable solution
(Ferguson et al., 1997) containing 250 ml siloxane
monomer (UMS 182), 100 ml dye solution (3 mg/ml in
methylene chloride), and 15 mg BEE into a plastic dish.
After a thin layer of the solution spread over the
bottom of the dish, a UV lamp was placed over the dish
for 20 min causing the solution to polymerize and
entrapping the dye in the resulting polymer. The poly-
mer film was lifted out of the dish with a spatula and
placed into a vial of distilled water and kept in the dark
until use.
In the second method, the oxygen sensitive film was
covalently attached to the tip of the optical fiber. The
distal tip of the optical fiber was silanized in 10%
3-triethoxysilylpropylmethacrylate in acetone for 2 h.
The silanization procedure enabled covalent binding of
the monomer directly to the glass surface upon poly-
merization. While the optical fiber was held vertically in
a Servodyne mixer head (Cole Parmer, Chicago, IL), 1
ml of the siloxane/dye mixture (described above) was
dropped onto the distal tip. The fiber was spun at 2000
rpm for 20 s forming a thin layer of siloxane. A UV
lamp was then held over the top of the fiber for 15 min
creating an oxygen sensitive fiber tip.
2.5. CPM lactate sensor preparation
The CPM lactate sensor was prepared using an adap-
tation of a CO2-sensor design (Uttamlal et al., 1995).
The distal tip of the optical fiber was fixed in a PEEK
housing flush with the PEEK surface (Fig. 1). When the
oxygen sensitive film was not covalently attached to the
fiber tip, a small piece of the oxygen sensitive film
Fig. 1. Schematic of fiber-optic CPM-based lactate sensor The sensor
is created at the distal tip of the optical fiber. The oxygen sensitive
film and CPM disc are placed directly over the optical fiber and held
securely with nylon and an o-ring. The proximal end of the fiber is
connected to the analysis system. The optical fiber can be any desired
length for remote sensing.
111
(700-mm diameter) was cut and placed over the fiber’s
end using tweezers. A small part of the cellulose disk
with adsorbed CPM (3-mm diameter) was cut and
placed directly on top of the oxygen sensitive layer. The
disk was held firmly in place using PTFE film or a
nylon net and a PEEK o-ring that snapped into place.
A small hole was cut into the PTFE film to allow direct
contact between the solution and CPM.
3. Analysis
Measurements were performed on custom-built
portable fluorometers (Steve Brown Engineering, Liver-
more, CA). A 470 nm band pass 35 filter was used to
select the excitation light specific for the ruthenium
indicator. Excitation light was sent through the optical
fiber to the oxygen sensitive polymer layer. The fluores-
cence emission was collected by the fiber and sent
through a dichroic beamsplitter having a wavelength
cut off at 505 nm to eliminate stray excitation light and
then filtered using a 610 nm bandpass 30 nm filter.
Data acquisition rate, filter switching, the light source,
and CCD detector were software controlled. All mea-
surements were performed in brown vials containing 2
ml of 20 mM Tris HCl buffer pH 7.4 with 5 mM
MgSO4 while stirring. After measurements were com-
plete the sensor was regenerated by placing it in a fresh
buffer solution for 1 min.
4. Results and discussion
L-lactate oxidase catalyzes the oxidation of L-lactate
with molecular oxygen, producing pyruvate and hydro-
gen peroxide. E. coli cells express L-lactate oxidizing
activity only when grown with L-lactate as the carbon
source (Dong et al., 1993). The advantages of E. coli
are the rapid growth of the organism and its ability to
maintain metabolic activity under both aerobic and
anaerobic conditions with simple nutritional require-
ments. Lactate-oxidizing enzymes of gram-negative
bacteria (including E. coli cells) have been characterized
and shown to be membrane-associated flavoproteins
that are components of the electron transport chain
(Gennis et al., 1996; Gibello et al., 1999) There was no
detectable lactate oxidase activity in the supernatant
from the E. coli membranes.
The lactate detection scheme is based on the reaction
lactate+O2Upyruvate + H2O2
Ru(ph2phen)23 + is sensitive to O2such that its fluores-
cence is inversely proportional to O2 concentration
(Demas et al., 1994). When lactate reacts with the CPM
containing lactate oxidase, the microenvironmental
oxygen concentration near the Ru(ph2phen)23 + /siloxane
112 S.G. Ignato6 et al. / Biosensors & Bioelectronics 16 (2001) 109–113
Fig. 2. Calibration curve of fiber-optic CPM lactate sensor Data for
each point in the calibration curve are obtained from curves similar
to Fig. 3. The initial slope is determined from the straight line
produced upon the addition of lactate. The instrument output is a
ratio of the lamp intensity/signal intensity to reduce background
effects due to lamp fluctuations. The initial rate versus the concentra-
tion of lactate added is shown.
layer decreases causing Ru(ph2phen)23 + fluorescence to
increase. The increased fluorescence is proportional to
the lactate concentration.
The monitoring of lactate is of specific relevance to
the diagnosis of myocardial infarction, congestive heart
failure, pulmonary edema, septacemia and hemorrhage.
These conditions can cause shock, which, due to associ-
ated anaerobic metabolism, may produce life-threaten-
ing lactic acidosis with high mortality rate. Blood
lactate concentrations exceeding 7 – 8 mM usually indi-
cate a fatal outcome, so the progressive decrease or
increase in blood lactate at this level is an indicator of
survival (Kenausis et al., 1997). Normal lactate concen-
tration in the blood is 2 mM (Murray et al., 1988). Fig.
2 shows the calibration curve of a fiber optic CPM
lactate sensor using the initial response rate of the
sensor as a function of concentration. The curvature in
the graph could be due to one of two mechanisms.
First, the enzyme’s activity in the CPM may be ap-
proaching saturation at higher lactate concentrations.
Second, calibration curves from ruthenium based oxy-
gen sensors often produce non-linear curves, which has
been attributed to a two-site quenching mechanism
caused by the heterogeneity of the polymer matrix
(Ferguson et al., 1997; Demas et al., 1994). The data
were fitted to a second order polynomial. A reasonably
good correlation is obtained with a CPM biosensor
over the lactate range 0 – 5 mM. Sensor calibration to
40 mM provided similar correlation data. To date, the
correlation of the calibration curve using CPM as the
recognition element is not as accurate as using the pure
enzyme. The sensitivity of the system, however, is im-
proved with the use of the CPM.
Changing the surface charge of the CPM with the
cationic detergent CTAB (50 mM) or the anionic deter-
gent SDS (50 mM) does not change the sensitivity of the
biosensor. The optical fiber CPM lactate biosensor was
tested for its response to interfering substances found in
biological samples such as glucose, fructose and glu-
tamic acid. Fig. 3 shows that 50-mM concentrations of
these substrates do not interfere with lactate detection.
Measuring the ratio of signal versus lamp intensity
minimizes noise due to lamp fluctuations, however, a
slight upward drift was noted. The slight drift could be
predicted and was mathematically corrected on the
graphs. We found that covalently attaching the oxygen
sensitive layer to the optical fiber improved the sensor’s
mechanical stability. The instrument used for this anal-
ysis did not show any signal drift. Fig. 4 shows the
fluorescence intensity of the sensor in buffer and after
the addition of 500 and 50 mM lactate. The sensor
response time was 3 min. Placing it in a fresh buffer
solution for 1 min (Fig. 4) caused the CPM biosensor
to return to its baseline fluorescence level.
5. Conclusion
We demonstrated the ability to use bacterial cyto-
plasmic membranes (CPM) as a recognition element in
a lactate fiber optic biosensor. CPM was obtained from
bacterial cells with induced lactate oxidase systems. The
biosensor based on bacterial CPM specifically detects
lactate in the presence of 50-mM concentrations of
glucose, fructose and glutamic acid. The unique organi-
zation of bacterial CPMs combined with the high sensi-
Fig. 3. Selectivity of the CPM based lactate sensor Additions of 50
mM (final concentration) of glucose, fructose or glutamic acid did not
affect the sensor. The response of the sensor after adding 2.5 mM
lactate demonstrates the specificity of the system. The ratio of lamp
intensity vs. signal intensity is the measured output from the instru-
ment which eliminates noise due to lamp fluctuations.
 S.G. Ignato6 et al. / Biosensors & Bioelectronics 16 (2001) 109–113
Fig. 4. Response and reversibility of the sensor The fiber optic
CPM-based lactate sensor shows an immediate response upon addi-
tion of lactate to the buffer solution. The concentration of the
solution is determined after 3 min and the sensor is completely
regenerated after 1 min in fresh buffer. Intensity is the output value
of the instrument used in this study.
tivity of the fiber-optic sensor could be beneficial for
specific and sensitive detection of other biologically
active substances.
Acknowledgements
The authors thank NATO Scientific Division (Col-
laborative Grant), ISTC (project c 890) and the Na-
tional Institutes of Health for financial support.
References
Bennetto, P., Halablad, M., Kim, H-J., 1996. Biosensors, 4th World
Congress, Thailand, 57, 1996.
Casimiri, V., Burstain, C., 1996. Biosens. Bioelectron. 11, 783.
Corbisier, P., Vanderlelie, D., Borremans, B., Provoost, A., De-
lorenzo, V., Brown, N.L., Lloyd, J.R., et al., 1999. Anal. Chim.
Acta 387, 235.
Das, N., Prabhakar, P., Kayastha, A.M., Strivastva, R.C., 1997.
Biotechnol. Bioeng. 54, 329.
113
Demas, J.N., DeGraff, B.A., 1994. Anal. Chem. 66, 4133.
Dong, J.M., Taylor, J.S., Latour, D.J., Iuchi, S., Lin, E.C.C., 1993. J.
Bacteriol. 175, 6671.
Evans, M.R., 1998. Pesticide Sci. 54, 447.
Ferguson, J.A., Healey, B.G., Bronk, K.S., Barnard, S.M., Walt,
D.R., 1997. Anal. Chim. Acta 340, 123.
Gennis, R.B., Stewart, G., 1996. Escherichia coli and Salmonella:
Cellular and Molecular Biology, second ed. ASM Press, Washing-
ton DC.
Gibello, A., Collins, M.D., Dominguez, L., Fernandes-Garayzabal,
J.F., Richardson, P.T., 1999. Appl. Environ. Microbiol. 65, 4346.
Hart, A.L., Cox, H., Janssen, D., 1996. Biosens. Bioelectron. 11, 833.
Healey, B.G., Walt, D.R., 1995. Anal. Chem. 67, 4471.
Hu, Y., Zhang, Y., Wilson, G.S., 1993. Anal. Chim. Acta 281, 503.
Ignatov, S.G., Krasilnikov, V.A., Perelygin, V.V., Kaprelyants, A.S.,
Ostrovsky, D.N., 1981. Biochimia 46, 1996.
Ikaryama, Y., Nichiguchi, S., Koyama, T., Kobatake, E., Aizawa,
M., Tsuda, M., Nakazawa, T., 1997. Anal. Chem. 69, 2600.
Jager, C.H., Lindow, S.E., Miller, W., Clark, E., Firestone, M.K.,
1999. Appl. Environ. Microbiol. 65, 2685.
Kelly, S., Compagnone, D., Guilbault, G., 1998. Biosens. Bioelec-
tron. 13, 173.
Kenausis, G., Chen, Q., Heller, A., 1997. Anal. Chem. 69, 1054.
Khayyami, M., Garcia, N.P., Larsson, P., Danielsson, B., Johansson,
G., 1997. Electroanalysis 9, 523.
Lee, J.I., Karube, I., 1995. Anal. Chim. Acta 316, 69.
Li, L., Walt, D.R., 1995. Anal. Chem. 67, 3746.
Lobo-Castanon, M.J., Miranda-Ordiers, A.J., Tunon-Blanco, P.,
1997. Anal. Chim. Acta 346, 165.
Lu, B., Smyth, M.R., O’Kennedy, R., 1996. Analyst 121, 29R.
Minagawa, H., Nakayama, N., Matsumoto, T., Ito, N., 1998.
Biosens. Bioelectron. 13, 313.
Murray, R.K., Graner, D.K., Mayes, P.A., Rodwell, V.W., 1988.
Harper’s Biochemistry, Appleton and Lange, CA.
Rainina, E.I., Efremenko, E.N., Varfolomeev, S.D., Salmonian, A.L.,
Wild, I.J., 1996. Biosens. Bioelectron. 11, 991.
Riedel, K., Renneberg, R., Scheller, F., 1990. Anal. Lett. 23, 757.
Roberts, G.G., Pitt, C.W., 1983. Langmuir-Blodgett Films, Elsevier,
Amsterdam.
Scott, D.L., Ramanathan, S., Shi, W., Rosen, B.P., Dauhert, S., 1997.
Anal. Chem. 69, 16.
Tien, H.T., Wurster, S.H., Ottova, A.L., 1997. Bioelectrochem.
Bioenergetics 42, 77.
Turner, A.P.F., Karube, I., Wilson, G.S. (Eds.), 1995. Biosensor
Fundamentals and Applications, Oxford University Press, Ox-
ford.
Uttamlal, M., Walt, D.R., 1995. Biotechnology 13, 597.
Wink, T.H., van Zuilen, S.T., Bult, A., van Bennekom, W.B., 1997.
Analyst 122, 43R.
Zellers, E.T., Han, M., 1996. Anal. Chem. 68, 2409.