- Email: [email protected]
l -lactate enzyme electrode obtained with immobilized respiratory chain From Escherichia coli and oxygen probe for specific determination of l -lactate in yogurt, wine and blood
Biosensors 3 (1987/88) 2743
L-lactate Enzyme Electrode Obtained with Immobilized Respiratory Chain From Escherichia coli and Oxygen Probe for Specific Determination of L-lactate in Yogurt, Wine and Blood
E. Adamowicz and C. Burstein LJniversitC Paris ‘I-Tour 54, Hall de Biotechnologies, 2 Place Jussieu, 75251 Paris Cedex 05 (France) (Received 21 November 1986; revised version received 7 January 1987; accepted 25 February 1987)
ABSTRACT An enzyme electrode for L-lactate measurements in various biological media was prepared with an immobilized bacterial respiratory chain fixed to a Clark electrode. The enzymatic film, which was easy to prepare, contained bacteria immobilized in gelatin, tanned with glutaraldehyde. This electrode was sensitive to 0-I mM L-lactate and could be utilized to IOmM. The apparent Ks,, was 5 mM. Less than 8% of the respiration rate with L-lactate was measured with D-lactate and succinate. The competitive inhibitors D-lactate and pyruvate had a KS0 of 50 mkt.They could be quantitatively measured by inhibition in a range between 5 and 50 mM. It was also possible to discriminate between L-lactate and various metabolites of the respiratory chain: L-malate, succinate, 3-glycero-phosphate or NAD(P)H. Growing E. coli on I% D-L-lactate as the sole carbon source in minimal medium induced L-lactate respiration tenfold. All other respiratory activities remained below 10% of the activity with L-lactate. A computerized instrument allowed successive measurements every 3 min for more than 10 h with the same enzymatic film. Most of the measured samples required dilution but no clarification or purification. This enzyme electrode may have many applications in basic research (metabolism, enzymology) and applied research (blood, yogurt, juices, wine). Key words: Enzyme electrode, L-lactate measurements, respiratory chain, immobilized enzyme.
immobilized
27 Biosensors 0265-928X/86/$03.50 0 Elsevier Applied Science Publishers Ltd, England, 1987. Printed in Great Britain
28
E. Adamowicz,
C. Burstein
INTRODUCTION Immobilized enzymes incorporated into enzyme electrodes are a valuable tool for analysis because of their specificity and stability. Continuous and instantaneous chemical determinations of substrates, products and inhibitors of enzymes are easy to perform with enzyme electrodes. The use of an enzyme as a functional element of an electrochemical device was first reported by Clark & Lyons (1962). Updike & Hicks (1967) described for the first time an electrode incorporating an immobilized enzyme. At least 60 different enzyme probes have already been described (Aizawa, 1983; Guilbault, 1982) using various sensors. With an oxygen probe measurements were facilitated by a polypropylene film which protected the probe. The enzymatic film was embedded in gelatin because gelatin dissolved about 6 times more oxygen than buffer (Belghit, 1985). This last property renders the measurements independent of small variations in the amount of oxygen in the solution tested. The membrane-bound proteins of the respiratory chain have been immobilized with success, as described in our earlier paper (Burstein et al., 1981). The respiratory chain was immobilized in a gelatin film and tightly fixed to an oxygen probe; this provided an enzyme electrode for the substrates and products of the respiratory chain. Two recent reviews on bacterial respiration (Bragg, 1980; Ingledew & Poole, 1984) showed clearly that the initial enzymes of the respiratory chain face the cytoplasm. These FMN or FAD-proteins are usually called dehydrogenases. To avoid confusion with NAD’ or NADP’ dependent
L-Lactate D-lactate
L- malate
3-GLycero-phosphate-
FP,-
q+
Cyt b 4
Cyt
o B
02
AZ cNADPH Pywate
YNADH Succinate
Fig. 1. Metabolic scheme of E. cofi respiratory chain, taken from our experimental
results (Burstein et al., 1986). Enzyme electrodes specific for various substrates of the respiratory chain were prepared by modifications of the flavoproteins. FP = flavoproteins; cyt b = cytochrome b; cyt o = cytochrome oxidase o.
Enzyme electrode for L-lactatemeasurements
29
dehydrogenases and with FAD-oxidases, we propose to call the FMN or FAD-proteins the flavoproteins of the respiratory chain. Our previous results suggest that for each metabolite oxidized by the respiratory chain of E. cofi a specific flavoprotein (Fig. 1) was involved (Burstein et al., 1981). The L-lactate enzyme electrode gave the highest values with the various respiratory substrates. This is due to the possibility of increasing the Llactate flavoprotein tenfold by induction with D, L-lactate during growth in minimal medium.
MATERIALS
AND METHODS
Bacteria (Burstein et al., 1981) Escherichia
cofi
~~~was grown exponentially at 37°C in aerated minimal medium. The culture was supplemented with 1 mg/ml thiamine and 10 g/l of D, L-lactate. Bacteria were concentrated by centrifugation to 10 mg/ml bacterial protein in the same medium without carbon source. They were aerated for 45 min to remove most of the endogenous substrates. Bacteria were washed and resuspended at 50 mg/ml bacterial protein in 50 mM K-phosphate buffer pH 7.6 before freezing at -80°C. Full respiratory activity remained for at least one year. Bacterial protein was estimated according to Lowry et al. (1951) using bovine serum albumin as standard. Immobilization (Romette & Yang, 1983) Immobilization was performed by mixing at 44X, 25 mg/ml (bacterial protein) final concentration with 5% gelatin final concentration (250 blooms-bone gelatin from Rousselot-Chemical Company, France). The mixture was gently agitated at 40°C for 2 min and poured rapidly on to a glass plate covered with a gas selective hydrophobic film (polypropylene from Bollore Inc., Paris, France, with a thickness of 6 pm). One millilitre of mixture was spread on 40 cm* and dried overnight in the cold room, then the film was tanned at 0°C with 1.25% aqueous glutaraldehyde for 6 min. The film was washed three times with 50 mu lysine plus 50 mu K-phosphate buffer pH 7.6 to remove the unreacted glutaraldehyde. The enzymatic film was finally stored at 4°C anaerobically in the same buffer supplemented with O-5 mM azide (to avoid bacterial contamination). O-5mu azide did not affect respiration in E. coli, probably because the cytochrome oxidase is type 0.
30
E. Adamowicz,
C. But-stein
The films remained active for at least 6 months when kept at 4°C. The yield of activity obtained after immobilization was around 75% for the respiratory chain. When the film was mounted on the oxygen probe an amplification of the signal by a factor of 500 was observed, compared with pieces of film introduced directly into the cuvette of the electrode. Respiration measurements (Burstein et al., EM) Oxygen consumption was measured with a Clark electrode at 37°C (Clark & Lyons, 1962). The probe was tightly covered with the polypropylene film, covered by the enzymatic film. The immobilized enzyme faced the electrode compartment (Romette &Yang, 1983). For automatic measurements the enzyme electrode was placed in a cell allowing flow measurements. The probe was sequentially exposed (Romette & Yang, 1983) to: (1) Buffer (50 mM triethanolamine pH 7*6), (2) Air (to saturate the gelatin with oxygen), (3) Sample. The slope of the signal at the inflection point was calculated by microprocessor and compared automatically with the calibration curve obained previously with various concentrations of L-lactate. The entire process took 3 min and could be repeated more than 400 times without loss of response. These experiments were carried out using a commercially available automated instrument, the Enzymat (SERES, France).
RESULTS AND DISCUSSION Growth of E. coli in miniial medium with various carbon sources As shown in Table 1, growth in the presence of 1% D-lactate, L-lactate or D, L-lactate increases by a factor of 10 the flavoprotein of the respiratory chain specific for L-lactate respiration. When glucose was added during growth, no induction was noticed. A diauxy was observed (Fig. 2) by measuring growth at ODm in the presence of O-15g/l glucose and O-15 g/l L-lactate. The flavoprotein of the respiratory chain, specific for L-lactate respiration, was induced only after glucose consumption. This result confirmed the induction of L-lactate respiration, as already suggested (Futai & Kimura, 1977). L-lactate appeared to penetrate freely through the bacterial membrane.
31
Enzyme electrode for L-lactate measurements TABLE 1
Oxidation
of L-lactate or NADH by Intact and Broken Bacteria, Grown on Various Carbon Sources
1% carbon source during growth
Glucose Glycerol D-lactate L-lactate
Glucose + L-lactate
Bacteria
02 consumption nmoles min-’ mg-’ L-lactate oxidation NADH oxidation
Intact Broken Intact Broken Intact Broken Intact Broken Intact Broken Intact
20 30 20 25 200 205 210 220 195 220 20
Broken
30
230
Bacteria were grown on minimum medium in the presence of 1% various carbon sources. The carbon sources after I h of growth were still in excess and were removed by centrifugation. Endogenous substrates were exhausted to reduce non-specific respiration (see MATERIALS AND METHODS). Bacteria were broken by 2-min exposure to sonic oscillations. Assays were performed by measuring oxygen consumption after adding 50mM L-lactate or 10 IIIM NADH.
The enzymatic site faces the cytoplasm (Bragg, 1980; Ingledew & Poole, 1984). In Table 1 no lactate permease was involved because of the similar results obtained with immobilized bacteria and ‘inverted vesicles’. On the other hand (Table 1) NADH oxidation appeared to be constitutive (present even in bacteria grown on glucose). No detectable activity with NADH was observed with intact bacteria; only broken bacteria (bacteria broken by ultrasound or French press containing among other things the inverted vesicles) gave maximal respiration. NADH was unable to cross the plasma membrane, and the site of fixation was also located towards the cytoplasm. Measurement
of L-lactate with the enzyme electrode
After addition of L-lactate to the measurement cell of the enzyme electrode about 10 s lag was observed followed by a decrease of oxygen, allowing the
32
E. Adamowicz,
C. But-stein
0150 0,40
0030
P “0 (D v
6320
Z Z Z : a 5
0,lO
F 0”
0105
I
1
1
2
3
4
5
6 TIME
(hours)
Fig. 2. E. coli growth in minimal medium in the presence of a mixture of glucose and L-lactate. The minimal medium contained 0.15 g/l of each carbon source. Bacterial growth was followed by measuring ODm (-•-). Respiration of L-lactate was measured with the oxygraph (-A-).
determination of a slope at the inflexion point which was considered to be the initial respiration velocity (Fig. 3). Then the cell was successively rinsed with buffer and air to obtain again 100% oxygen saturation of the gelatin at 37°C. Characteristics of the enzyme electrode are represented in Figs 3,4 and 5. Reproducibility of the assay with the same film, during automatic measurements are represented in Fig. 8 (experiment 3) and Fig. 9 (experiment 2). Variation of L-lactate concentration allowed the determination of a titration curve. This titration curve is necessary for each film because of the variations observed (cf Figs 8 and 9). The evolution of the substrate concentration in the film depends on the Michaelis law plus diffusion properties. An approximate Km can nevertheless be determined (KS0 = 5 tIIM). This titration curve showed that the assay can be performed between O-1 and 10 IIIM.
Enzyme electrode for L-lactate measurements
TIME1
TIME
1 1
TIME 2
2
lME3
33
TIME 4
3 TIME (min)
Fig. 3. Kinetics of L-lactate measurement with the enzyme electrode, prepared as described in MATERIALS AND METHODS. At Time 1, SOmM, the saturation concentration of L-lactate was introduced in the cell with 50 mM triethanolamine buffer, pH 7.6. The maximal slope at the inflexion point of oxygen consumption was measured. At Time 2 the cell was rinsed with buffer. At Time 3 air was introduced into the cell to restore 100% oxygen saturation of the gelatin. At Time 4 a new assay was performed by addition of L-lactate.
Effect of the pH
As shown in Fig. 5, the respiration of L-lactate with the enzyme electrode was constant between pH 6-O and 10.0. 50% of the maximal activity can be easily measured at pH 4-O with a titration curve at this pH. This last pH is usually found in yogurt, wine and fruit juices. The flat curve obtained with the intact bacteria suggests that membrane enzymes like those of the respiratory chain may be stabilized by their location on a biological membrane. Effect of heat denaturation
Intact bacteria were heated at 55°C for 210 min and immobilized on the oxygen probe. A loss of activity for L-lactate respiration was observed
34
E. Adamowicz,
1’0 PllOTElN
C. But-stein
2'0 CONCENTRATION
3'0 (mg
ml-‘)
Fig. 4. L-lactate measurement versus various amount of bacteria in the enzymatic film. Assay was performed with the enzyme electrode as described in Fig. 3. Various amounts of bacteria were added to the gelatin. The amount of bacteria was expressed in total mg protein/ml of final volume of the polymer (O-5ml bacteria at various concentrations + O-5ml gelatin 10%). Standard protein concentration of bacteria was 25 mg/ml.
3 h (L-malate, D-lactate, NADH and NADPH respiration were - Cl12 = reduced even more by this heat treatment; results not shown). The heat treatment increased the specificity of the L-lactate enzyme electrode. An enzyme electrode was able to function repetitively 150 times at 50°C without loss of activity. The use of a temperature above 45°C increased the rate of the reaction and avoided most bacterial contamination (most bacteria do not grow above 42°C). Stability Using a microprocessor-linked oxygen probe (see MATERIALS AND METHODS) it was possible to repeat the same L-lactate measurement at approximately the KS0concentration. We showed that 400 assays can be
35
Enzyme electrode for L-lactate measuremenfi
6
4
10
6 PH
Fig. 5. pH effect on E. cofi respiratory chain before and after immobilization. Intact bacterial respiration of L-lactate was measured after growth, on D, L-lactate at different pH in the oxygraph cell with only a polypropylene film. (Closed symbols are for bacterial suspension.) Immobilized bacteria were fixed on the oxygen electrode before measurement of respiration of L-lactate. 50 mM L-lactate was added. (Open symbols are for immobilized bacteria.) The various buffers were utilized at 1OOmM. VD-Na-acetate; l O-Na-maleate; 00 -Kphosphate; A A-Tris-HCl; n 0 -Na-glycine. 100% was O-22 pmoles/min/mg 02 consumed for intact bacteria and 180 pmoles/min/mg 02 consumed for immobilized bacteria. Standard buffer used was 50 mu triethanolamine, pH 7.6. TABLE 2 L-lactate Concentration in Yogurt and in Wine
In yogurt
In wine
Sample
Slope/unit oftime
Concentration in diluted sample mhi
Dilution of sample
Final concentration mM of L-lactate
1 2 3 4 5 1 2 3
0.54 1.02 1.40 1.50 2.24 l-12 1.52 2.12
0.55 1.05 1.70 2.10 44.0 1.65 2.24 3.11
160 80 60 40 20 20 14 10
88 84 85 84 88 32 33 31
Experimental results are presented in Fig. 8 for yogurt and in Fig. 9 for wine. The signals obtained allow conversion (after multiplication by the dilution factor) in L-lactate concentration present in yogurt is 86 k 2 mM and wine is 32 f 2 mM.
36
E. Adamowicz,
C. Burstein
performed, one assay every 3 min, without appreciable variation of activity (< 2%). Optimization may increase the number of reproducible assays. When enzymatic films were kept at 4°C in the presence of 50 mM lysine, 50 mM K-phosphate buffer at pH 7-6 and O-5mM azide in the absence of oxygen, activity remained constant for at least six months. Specificity
The respiratory chain of E. coli is able to metabolize various substrates resulting from its metabolism (Fig. 1). When E. cofi was grown in minimal medium with 1% D-L-lactate as the sole carbon source, the amount of respiration of L-lactate induced was at least 10 times the amount of the other activities. When 100% L-lactate oxidation was measured all the other activities were reduced to less than 8% of that of L-lactate and therefore could be neglected in most of the cases (Fig. 6). Very little L-lactate was metabolized when the bacteria were grown on glycerol (Fig. 6).
1
2
-r 3
a
d
Fig. 6. Specificity of the enzyme electode for L-lactate. Respiration obtained with the enzyme electrode for L-lactate. Respiratory substrates used: a 5 mM L-lactate; b 5 mM o-lactate; ( 50 mM succinate; d 2 mM NADH; e 30 mM pyruvate. Exp. l-bacteria were grown on 1% D L-lactate; exp. 2-bacteria were grown on 1% glycerol; exp. 3-bacteria were grown on 1% D L-lactate and sonicated; 100% was 180 ~moles/min/mg 02 consumed.
Enzyme electrode for L-lactate measurements
37
Intact induced bacteria were more specific than the same bacteria broken because of the contamination by NADH or pyruvate respiration (Fig. 6). Assay of various inhibitors of the L-lactate enzyme electrode The L-lactate enzyme electrode can also be used to measure pyrqate concentration, as shown in Fig. 7. L-lactate was used at Km = 5 mM concentration. An apparent K1 = 50 mM can be determined for pyruvate. In the same way approximate apparent K1 for D-lactate, oxalate and oxamate were respectively 50,100 and 200 mM. The assay of these inhibitors was not very sensitive. When bacteria grown on L-lactate as the sole carbon source were used, no significant respiration of D-lactate or pyruvate was observed with the enzyme electrode prepared with the L-lactate bacteria. D-lactate and
PYRUVATE
CONCENTRATION (mM)
Fig. 7. Titration curve for the measurement of inhibition of L-lactate respiration by pyruvate. L-lactate was added to each sample at 5 mM (Km). Successive additions of pyruvate inhibit L-lactate respiration.
38
E. Adamowicz, C. Burstein
pyruvate were, on the contrary, capable of inhibition of L-lactate respiration (Fig. 7). The kinetics of inhibition are complex: pseudocompetitive inhibition was observed; D-lactate, pyruvate, oxalate and oxamate (inhibitors of L-lactate respiration) are probably interacting with the Llactate llavoprotein. Utilization of the enzyme electrode for L-lactate determination biological media
in various
L-lactate is involved in anaerobic degradation of sugars. In medicine its level in blood is related to muscle diseases, critical care, lymphomas, sarcomas, etc. The L-lactate content in the blood of sportsmen can be a measure of physical exercise. In the food industry L-lactate measurements may be used to follow fermentation in wine, beer, fruit juices and milk products. The commercially available L-lactate assay (Boehringer-Mannheim UV method using NAD’ dependent L-lactate dehydrogenase) took too long and was not easy to perform. With the enzyme electrode the assay took 3 min with an error of less than 3%, and the cost of bacteria was at least 10 times lower than the NAD’ dependent lactate dehydrogenase. Our work was performed with a buffer and pure L-lactate. Although oxygen consumption can be measured in a turbid medium, inhibitors or activators may interfere with the assay in biological media. For each new medium it will be necessary to optimize the assay. Assay of L-lactate in yogurt with the enzyme electrode utilizing immobilized respiratory chain from E. coli The L-lactate concentration in yogurt was determined. In yogurt the Llactate concentration varies with the amount of lactic bacteria used and with the preservation conditions; in the present work values between 70 and 90 mM were found. Our titration curve of L-lactate (Fig. 8) showed that the amount of L-lactate in crude yogurt is too high to give a quantitative assay. After dilution of yogurt (20 to 160 times in 50 IIIM triethanolamine pH 7*6), the final pH was 7.6 and L-lactate concentration measured between O-5 and 5 tIIM. To obtain the actual lactate concentration it was necessary to multiply by the dilution factor. Even though the yogurt dilutions were turbid they could be used directly for L-lactate measurements. Various dilutions gave the same final concentration of L-lactate 86 * 2 mM (Table 2). If the assay was done repetitively 160 times with one dilution, the error was in the order of 4% (Fig. 8). This assay was compared with the use of L-lactate dehydrogenase plus
Enzyme electrode for L-lactate measurements
MEASURE
39
NUMBER
#
I
5
10
L-LACTATE
I
I 5’
CONCENTRATION (mM)
Fig. 8. L-lactate concentration
determined in yogurt. The L-lactate enzyme electrode was first calibrated with L-lactate (-•-). Then various concentrations of yogurt diluted in 50m~ triethanolamine pH 7.6 were measured automatically 10 times for samples 1,2,4,5 and 160 times for sample 3 (-------). This allowed the determination of the L-lactate in yogurt (Table 2).
NAD+ according to Boehringer (Table 3). Instead of 3 min, 30 min were necessary and the soluble enzyme (NAD’ dependent L-lactate dehydrogenase) could not be reused for the next assay. In Table 3 the two assays of L-lactate were compared in a yogurt (different from the one above and with a new L-lactate content); comparable values were obtained, 74 + 1 mM, with the two assays. Measurements of L-lactate in wine Dilution of wine 10 and 20 times in 50 mM triethanolamine pH 7-6 gave good measurements of L-lactate as shown in Fig. 9. The L-lactate concentration
E. Adamowicz, C. Burstein
TABLE 3 Measurements of L-lactate L-lactate (mM) Measurement I 2 3
Means
Enzyme electrode 74.2 75.5 73.8 74.5 r I
L-lactute dehydrogenuse ussuy 72.8 74.6 74.2 74.6 +-2
Concentration in yogurt with an enzyme electrode and with the L-lactate-dehydrogenase NAD+ dependent.
was 32 + 2 mM (Table 2). If the same sample was measured 100 times with the Enzymat the error was in the order of 5%. When the L-lactate assay in wine was performed according to Boehringer with L-lactate dehydrogenase in the presence of NAD’, values were similar to those obtained with the enzyme electrode. It was concluded that the L-malate and ethanol present in the wine sample (dilution of wine was 10 to 20 times) were not disturbing the enzyme electrode assay. Measurements of L-lactate in blood Measurements have been done with freshly collected blood. In normal blood the concentration of L-lactate found was 1 mM. For sportsmen during exercise the concentration of L-lactate increased (depending on the effort), reaching 20 ITIM. The main problem was in the use of small samples of blood (below loo j.&l). If the sample had to be stored at 4”C, an inhibitor of lactic acid metabolism was necessary.
CONCLUSIONS Many L-lactate determinations have been reported using immobilized enzymes coupled to an electrode. Some are potentiometric assays (Shinbo et al., 1979) while others are amperometric assays (Durliat & Comtat, 1980) using ferricyanide or NAD’ as an electron acceptor. Other electrodes utilize oxygen as an electron acceptor. These electrodes utilize purified oxidases from Pediococcus species (Mizutami et al., 1983; Mizutami et al., 19&l), from Mycobacterium smegmatis (Mascini et al., 1983)) or from Acetobacter peroxydans (Cannon et al., 1984).
41
Enzyme electrode for L-lactate measurements NUMBER
MEASURE ,
5P
*
.
.
1
‘V
1 i
2
L-LACTATE
9
1
.
.
8
I 4
CONCENTRATION (mM)
Fig. 9. L-lactate concentration determined in wine. A titration curve was performed with the Enzymat for L-lactate (-•-). Measurements of L-lactate were performed automatically with the specific enzyme electrode. Various dilutions of wine were introduced after dilution in 50 mM triethanolamine pH 7.6 and were measured automatically 10 times for samples 1,3 and 100 times for sample 2 (-------). This allowed the determination of the L-lactate concentration in wine (Table 2).
Our enzyme electrode for measuring L-lactate utilizes a crude multienzymatic membrane system. The cost of the enzymatic material is low. E. coii grown on a minimum medium with D, L-lactate as the sole carbon source was a good source of an L-lactate specific respiratory chain to build an enzyme electrode with an oxygen probe. The enzyme source is very easy to obtain. Intact bacteria (without enzyme purification) gave the best yield after immobilization. Probably because of its biological membrane location, the respiratory chain was already stabilized. The measurement of L-lactate using the respiratory chain was independent of addition of cofactors (NAD’ or NADP’) or coenzymes (FAD or FMN). Only concentrations of oxygen were measured, and
42
E. Adamowicz, C. Burstein
optically clear solutions were not needed (unlike NADH measurement by spectrophotometry or fluorometry). Most of the assays can be performed directly, with the sample diluted if necessary. The assay lasts 3 min and can be repeated automatically. L-lactate is involved in anaerobic degradation of sugars. In medicine, knowing its level in blood is very important for sports, muscle diseases, coma, cancer, etc. In the food industry L-lactate may be a good indicator to improve and control wine, juice and milk fermentation. Each problem, however, may need specialized research to adapt the enzyme electrode to measurement of L-lactate not only in a buffer but in various biological media. The output of the electrode is limited by the rate of respiration of the bacterial protein. As shown in Fig. 4, oxygen consumption is proportional to the protein concentration (activity of the respiratory chain) in the film. The response is determined by the concentration of delivery of the analyte to the electrode as shown by the titration curve of Fig. 8. The fact that ‘intact’ bacteria and broken bacteria gave similar L-lactate oxidation (Table 1) suggests that no active transport is present in the immobilized bacteria. The complete inhibition of L-lactate respiration by 10 mM K-cyanide demonstrates that only the respiratory chain is involved in L-lactate metabolism when measured with immobilized bacteria or inverted vesicles. Stirring or increasing the flow rate in the automatic instrument produces an observable variation in the response of the electrode. Maintaining stirring or flow rate constant during calibration and measurements provides results which are in good correlation with an independent assay using another enzyme in solution (L-lactate dehydrogenase + NAD’) provided by Boehringer (Table 3). In our conditions, transport through the enzymatic film of the electrode does not seem to be the limiting factor of the assay. Nevertheless, in the case of low concentrations of substrates and high concentrations of immobilized enzymes, results).
diffusion
was observed
to be the limiting factor
(unpublished
ACKNOWLEDGEMENTS The authors have benefited from many fruitful discussions with D. Thomas and J. L. Romette of the Universite de Technologie de Compiegne. This work was supported by Le Centre National de la Recherche Scientifique, 1’UniversitC Paris 7, l’ANVAR, I’UniversitC de Technologie de Compiegne.
Enzyme electrode for L-lactate measuremerus
43
REFERENCES Aizawa, M. (1983). Molecular recognition and chemical amplification biosensors. In: Procs of the International Meeting on Chemical Sensors, (Ed. Seiyama, T., Fueki, K., Shiokawa, J. & Suzuki, S.), 683-92. Belghit, H. (1985). Production et extraction-purification d’une alcool oxydase. Realisation et developpement d’un capteur a alcool. These de DocteurIngenieur, Universite de Technologie de Compibgne. Bragg, P. D. (1980). The respiratory system of Escherichia coli. In: Diversity of Bacterial Respiratory Systems Vol. 1 (Ed. Konwels, C. J.). CRC Press Inc., Boca Raton, Florida, 115-36. Burstein, C., Adamowicz, E., Boucherit, K., Rabouille, C. & Romette, J. L. (1986). Immobilized respiratory chain activities from E. coli utilized to measure D- and L-lactate, succinate, C-malate, 3-glycero-phosphate, pyruvate or NAD(P)H, Appl. Biochem. and Biotechnol., 12,1-16. Burstein, C., Ounissi, H., Legoy, M. D., Gellf, G. & Thomas, D. (1981). Recycling of NAD+ using coimmobilized alcohol dehydrogenase and E. coli, Appl.
Biochem.
Biotechnol.,
6,329-38.
Cannon, J. J., Chen, L. F., Flickinger, M. C. & Tsao, G. T. (1984). The development of an immobilized lactate oxidase system for lactic acid analysis, Biotechnol.
Bioeng.,
26,167-73.
Clark, L. C. & Lyons, C. (1962). Electrode system for continuous monitoring in cardiovascular surgery, Ann. NY Acad. Sci., 102,29-45. Durliat, M. & Comtat, M. (1980). Reagentless amperometric lactate electrode, Anal. Chem., U, 2109-12. Futai, M. & Kimura, H. (1977). Inducible membrane bound L-lactate dehydrogenase from E. coli, J. Biol. Chem., 252,5820-70. Guilbault, G. G. (1982). Immobilized enzymes as analytical reagents, Appl. Biochem.
Biotechnol.,
7,85-98.
Ingledew, W. J. & Poole, R. K. (1984). The respiratory chains of Escherichia coli, Microbial.
Revs 48222-71.
Lowry, 0. H., Rosebrough, N. J., Fat-r, A. L. & Randall, R. J. (1951). Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193,265-75. Mascini, M., Palleschi, G. & Moscone, D. (1983). Lactate electrode obtained with lactate oxidase and an oxygen electrode for serum samples. In: Procs of the International Meeting on Chemical Sensors (Ed. Seiyama, T., Fueki, K., Shiokawa, J. & Suzuki, S.), 603-8. Mizutami, F., Sasaki, K. & Shimura, Y. (1983). Sequential determination of L-lactate and lactate dehydrogenase with immobilized enzyme electrode, Anal. Chem.,
55,358.
Mizutami, F., Shimura, Y. & Tsuda, K. (1984). Catalytic assay of L-lactate or pyruvate with an enzyme electrode based on immobilized lactate oxidase and lactate dehydrogenase, Chem. Lett., 2,199-202. Romette, J. L. & Yang, J. S. (1983). Enzyme electrode for specific determination of L-lysine, Biotechnol. Bioeng., 25,255766. Shinbo, T., Sugiura, M. & Kamo, N. (1979). Potentiometric enzyme electrode for lactate, Anal. Chem., 51, 1004. Updike, S. J. & Hicks, G. D. (1967). The enzyme electrode, Nature, 214,986-8.
L-lactate Enzyme Electrode Obtained with Immobilized Respiratory Chain From Escherichia coli and Oxygen Probe for Specific Determination of L-lactate in Yogurt, Wine and Blood
E. Adamowicz and C. Burstein LJniversitC Paris ‘I-Tour 54, Hall de Biotechnologies, 2 Place Jussieu, 75251 Paris Cedex 05 (France) (Received 21 November 1986; revised version received 7 January 1987; accepted 25 February 1987)
ABSTRACT An enzyme electrode for L-lactate measurements in various biological media was prepared with an immobilized bacterial respiratory chain fixed to a Clark electrode. The enzymatic film, which was easy to prepare, contained bacteria immobilized in gelatin, tanned with glutaraldehyde. This electrode was sensitive to 0-I mM L-lactate and could be utilized to IOmM. The apparent Ks,, was 5 mM. Less than 8% of the respiration rate with L-lactate was measured with D-lactate and succinate. The competitive inhibitors D-lactate and pyruvate had a KS0 of 50 mkt.They could be quantitatively measured by inhibition in a range between 5 and 50 mM. It was also possible to discriminate between L-lactate and various metabolites of the respiratory chain: L-malate, succinate, 3-glycero-phosphate or NAD(P)H. Growing E. coli on I% D-L-lactate as the sole carbon source in minimal medium induced L-lactate respiration tenfold. All other respiratory activities remained below 10% of the activity with L-lactate. A computerized instrument allowed successive measurements every 3 min for more than 10 h with the same enzymatic film. Most of the measured samples required dilution but no clarification or purification. This enzyme electrode may have many applications in basic research (metabolism, enzymology) and applied research (blood, yogurt, juices, wine). Key words: Enzyme electrode, L-lactate measurements, respiratory chain, immobilized enzyme.
immobilized
27 Biosensors 0265-928X/86/$03.50 0 Elsevier Applied Science Publishers Ltd, England, 1987. Printed in Great Britain
28
E. Adamowicz,
C. Burstein
INTRODUCTION Immobilized enzymes incorporated into enzyme electrodes are a valuable tool for analysis because of their specificity and stability. Continuous and instantaneous chemical determinations of substrates, products and inhibitors of enzymes are easy to perform with enzyme electrodes. The use of an enzyme as a functional element of an electrochemical device was first reported by Clark & Lyons (1962). Updike & Hicks (1967) described for the first time an electrode incorporating an immobilized enzyme. At least 60 different enzyme probes have already been described (Aizawa, 1983; Guilbault, 1982) using various sensors. With an oxygen probe measurements were facilitated by a polypropylene film which protected the probe. The enzymatic film was embedded in gelatin because gelatin dissolved about 6 times more oxygen than buffer (Belghit, 1985). This last property renders the measurements independent of small variations in the amount of oxygen in the solution tested. The membrane-bound proteins of the respiratory chain have been immobilized with success, as described in our earlier paper (Burstein et al., 1981). The respiratory chain was immobilized in a gelatin film and tightly fixed to an oxygen probe; this provided an enzyme electrode for the substrates and products of the respiratory chain. Two recent reviews on bacterial respiration (Bragg, 1980; Ingledew & Poole, 1984) showed clearly that the initial enzymes of the respiratory chain face the cytoplasm. These FMN or FAD-proteins are usually called dehydrogenases. To avoid confusion with NAD’ or NADP’ dependent
L-Lactate D-lactate
L- malate
3-GLycero-phosphate-
FP,-
q+
Cyt b 4
Cyt
o B
02
AZ cNADPH Pywate
YNADH Succinate
Fig. 1. Metabolic scheme of E. cofi respiratory chain, taken from our experimental
results (Burstein et al., 1986). Enzyme electrodes specific for various substrates of the respiratory chain were prepared by modifications of the flavoproteins. FP = flavoproteins; cyt b = cytochrome b; cyt o = cytochrome oxidase o.
Enzyme electrode for L-lactatemeasurements
29
dehydrogenases and with FAD-oxidases, we propose to call the FMN or FAD-proteins the flavoproteins of the respiratory chain. Our previous results suggest that for each metabolite oxidized by the respiratory chain of E. cofi a specific flavoprotein (Fig. 1) was involved (Burstein et al., 1981). The L-lactate enzyme electrode gave the highest values with the various respiratory substrates. This is due to the possibility of increasing the Llactate flavoprotein tenfold by induction with D, L-lactate during growth in minimal medium.
MATERIALS
AND METHODS
Bacteria (Burstein et al., 1981) Escherichia
cofi
~~~was grown exponentially at 37°C in aerated minimal medium. The culture was supplemented with 1 mg/ml thiamine and 10 g/l of D, L-lactate. Bacteria were concentrated by centrifugation to 10 mg/ml bacterial protein in the same medium without carbon source. They were aerated for 45 min to remove most of the endogenous substrates. Bacteria were washed and resuspended at 50 mg/ml bacterial protein in 50 mM K-phosphate buffer pH 7.6 before freezing at -80°C. Full respiratory activity remained for at least one year. Bacterial protein was estimated according to Lowry et al. (1951) using bovine serum albumin as standard. Immobilization (Romette & Yang, 1983) Immobilization was performed by mixing at 44X, 25 mg/ml (bacterial protein) final concentration with 5% gelatin final concentration (250 blooms-bone gelatin from Rousselot-Chemical Company, France). The mixture was gently agitated at 40°C for 2 min and poured rapidly on to a glass plate covered with a gas selective hydrophobic film (polypropylene from Bollore Inc., Paris, France, with a thickness of 6 pm). One millilitre of mixture was spread on 40 cm* and dried overnight in the cold room, then the film was tanned at 0°C with 1.25% aqueous glutaraldehyde for 6 min. The film was washed three times with 50 mu lysine plus 50 mu K-phosphate buffer pH 7.6 to remove the unreacted glutaraldehyde. The enzymatic film was finally stored at 4°C anaerobically in the same buffer supplemented with O-5 mM azide (to avoid bacterial contamination). O-5mu azide did not affect respiration in E. coli, probably because the cytochrome oxidase is type 0.
30
E. Adamowicz,
C. But-stein
The films remained active for at least 6 months when kept at 4°C. The yield of activity obtained after immobilization was around 75% for the respiratory chain. When the film was mounted on the oxygen probe an amplification of the signal by a factor of 500 was observed, compared with pieces of film introduced directly into the cuvette of the electrode. Respiration measurements (Burstein et al., EM) Oxygen consumption was measured with a Clark electrode at 37°C (Clark & Lyons, 1962). The probe was tightly covered with the polypropylene film, covered by the enzymatic film. The immobilized enzyme faced the electrode compartment (Romette &Yang, 1983). For automatic measurements the enzyme electrode was placed in a cell allowing flow measurements. The probe was sequentially exposed (Romette & Yang, 1983) to: (1) Buffer (50 mM triethanolamine pH 7*6), (2) Air (to saturate the gelatin with oxygen), (3) Sample. The slope of the signal at the inflection point was calculated by microprocessor and compared automatically with the calibration curve obained previously with various concentrations of L-lactate. The entire process took 3 min and could be repeated more than 400 times without loss of response. These experiments were carried out using a commercially available automated instrument, the Enzymat (SERES, France).
RESULTS AND DISCUSSION Growth of E. coli in miniial medium with various carbon sources As shown in Table 1, growth in the presence of 1% D-lactate, L-lactate or D, L-lactate increases by a factor of 10 the flavoprotein of the respiratory chain specific for L-lactate respiration. When glucose was added during growth, no induction was noticed. A diauxy was observed (Fig. 2) by measuring growth at ODm in the presence of O-15g/l glucose and O-15 g/l L-lactate. The flavoprotein of the respiratory chain, specific for L-lactate respiration, was induced only after glucose consumption. This result confirmed the induction of L-lactate respiration, as already suggested (Futai & Kimura, 1977). L-lactate appeared to penetrate freely through the bacterial membrane.
31
Enzyme electrode for L-lactate measurements TABLE 1
Oxidation
of L-lactate or NADH by Intact and Broken Bacteria, Grown on Various Carbon Sources
1% carbon source during growth
Glucose Glycerol D-lactate L-lactate
Glucose + L-lactate
Bacteria
02 consumption nmoles min-’ mg-’ L-lactate oxidation NADH oxidation
Intact Broken Intact Broken Intact Broken Intact Broken Intact Broken Intact
20 30 20 25 200 205 210 220 195 220 20
Broken
30
230
Bacteria were grown on minimum medium in the presence of 1% various carbon sources. The carbon sources after I h of growth were still in excess and were removed by centrifugation. Endogenous substrates were exhausted to reduce non-specific respiration (see MATERIALS AND METHODS). Bacteria were broken by 2-min exposure to sonic oscillations. Assays were performed by measuring oxygen consumption after adding 50mM L-lactate or 10 IIIM NADH.
The enzymatic site faces the cytoplasm (Bragg, 1980; Ingledew & Poole, 1984). In Table 1 no lactate permease was involved because of the similar results obtained with immobilized bacteria and ‘inverted vesicles’. On the other hand (Table 1) NADH oxidation appeared to be constitutive (present even in bacteria grown on glucose). No detectable activity with NADH was observed with intact bacteria; only broken bacteria (bacteria broken by ultrasound or French press containing among other things the inverted vesicles) gave maximal respiration. NADH was unable to cross the plasma membrane, and the site of fixation was also located towards the cytoplasm. Measurement
of L-lactate with the enzyme electrode
After addition of L-lactate to the measurement cell of the enzyme electrode about 10 s lag was observed followed by a decrease of oxygen, allowing the
32
E. Adamowicz,
C. But-stein
0150 0,40
0030
P “0 (D v
6320
Z Z Z : a 5
0,lO
F 0”
0105
I
1
1
2
3
4
5
6 TIME
(hours)
Fig. 2. E. coli growth in minimal medium in the presence of a mixture of glucose and L-lactate. The minimal medium contained 0.15 g/l of each carbon source. Bacterial growth was followed by measuring ODm (-•-). Respiration of L-lactate was measured with the oxygraph (-A-).
determination of a slope at the inflexion point which was considered to be the initial respiration velocity (Fig. 3). Then the cell was successively rinsed with buffer and air to obtain again 100% oxygen saturation of the gelatin at 37°C. Characteristics of the enzyme electrode are represented in Figs 3,4 and 5. Reproducibility of the assay with the same film, during automatic measurements are represented in Fig. 8 (experiment 3) and Fig. 9 (experiment 2). Variation of L-lactate concentration allowed the determination of a titration curve. This titration curve is necessary for each film because of the variations observed (cf Figs 8 and 9). The evolution of the substrate concentration in the film depends on the Michaelis law plus diffusion properties. An approximate Km can nevertheless be determined (KS0 = 5 tIIM). This titration curve showed that the assay can be performed between O-1 and 10 IIIM.
Enzyme electrode for L-lactate measurements
TIME1
TIME
1 1
TIME 2
2
lME3
33
TIME 4
3 TIME (min)
Fig. 3. Kinetics of L-lactate measurement with the enzyme electrode, prepared as described in MATERIALS AND METHODS. At Time 1, SOmM, the saturation concentration of L-lactate was introduced in the cell with 50 mM triethanolamine buffer, pH 7.6. The maximal slope at the inflexion point of oxygen consumption was measured. At Time 2 the cell was rinsed with buffer. At Time 3 air was introduced into the cell to restore 100% oxygen saturation of the gelatin. At Time 4 a new assay was performed by addition of L-lactate.
Effect of the pH
As shown in Fig. 5, the respiration of L-lactate with the enzyme electrode was constant between pH 6-O and 10.0. 50% of the maximal activity can be easily measured at pH 4-O with a titration curve at this pH. This last pH is usually found in yogurt, wine and fruit juices. The flat curve obtained with the intact bacteria suggests that membrane enzymes like those of the respiratory chain may be stabilized by their location on a biological membrane. Effect of heat denaturation
Intact bacteria were heated at 55°C for 210 min and immobilized on the oxygen probe. A loss of activity for L-lactate respiration was observed
34
E. Adamowicz,
1’0 PllOTElN
C. But-stein
2'0 CONCENTRATION
3'0 (mg
ml-‘)
Fig. 4. L-lactate measurement versus various amount of bacteria in the enzymatic film. Assay was performed with the enzyme electrode as described in Fig. 3. Various amounts of bacteria were added to the gelatin. The amount of bacteria was expressed in total mg protein/ml of final volume of the polymer (O-5ml bacteria at various concentrations + O-5ml gelatin 10%). Standard protein concentration of bacteria was 25 mg/ml.
3 h (L-malate, D-lactate, NADH and NADPH respiration were - Cl12 = reduced even more by this heat treatment; results not shown). The heat treatment increased the specificity of the L-lactate enzyme electrode. An enzyme electrode was able to function repetitively 150 times at 50°C without loss of activity. The use of a temperature above 45°C increased the rate of the reaction and avoided most bacterial contamination (most bacteria do not grow above 42°C). Stability Using a microprocessor-linked oxygen probe (see MATERIALS AND METHODS) it was possible to repeat the same L-lactate measurement at approximately the KS0concentration. We showed that 400 assays can be
35
Enzyme electrode for L-lactate measuremenfi
6
4
10
6 PH
Fig. 5. pH effect on E. cofi respiratory chain before and after immobilization. Intact bacterial respiration of L-lactate was measured after growth, on D, L-lactate at different pH in the oxygraph cell with only a polypropylene film. (Closed symbols are for bacterial suspension.) Immobilized bacteria were fixed on the oxygen electrode before measurement of respiration of L-lactate. 50 mM L-lactate was added. (Open symbols are for immobilized bacteria.) The various buffers were utilized at 1OOmM. VD-Na-acetate; l O-Na-maleate; 00 -Kphosphate; A A-Tris-HCl; n 0 -Na-glycine. 100% was O-22 pmoles/min/mg 02 consumed for intact bacteria and 180 pmoles/min/mg 02 consumed for immobilized bacteria. Standard buffer used was 50 mu triethanolamine, pH 7.6. TABLE 2 L-lactate Concentration in Yogurt and in Wine
In yogurt
In wine
Sample
Slope/unit oftime
Concentration in diluted sample mhi
Dilution of sample
Final concentration mM of L-lactate
1 2 3 4 5 1 2 3
0.54 1.02 1.40 1.50 2.24 l-12 1.52 2.12
0.55 1.05 1.70 2.10 44.0 1.65 2.24 3.11
160 80 60 40 20 20 14 10
88 84 85 84 88 32 33 31
Experimental results are presented in Fig. 8 for yogurt and in Fig. 9 for wine. The signals obtained allow conversion (after multiplication by the dilution factor) in L-lactate concentration present in yogurt is 86 k 2 mM and wine is 32 f 2 mM.
36
E. Adamowicz,
C. Burstein
performed, one assay every 3 min, without appreciable variation of activity (< 2%). Optimization may increase the number of reproducible assays. When enzymatic films were kept at 4°C in the presence of 50 mM lysine, 50 mM K-phosphate buffer at pH 7-6 and O-5mM azide in the absence of oxygen, activity remained constant for at least six months. Specificity
The respiratory chain of E. coli is able to metabolize various substrates resulting from its metabolism (Fig. 1). When E. cofi was grown in minimal medium with 1% D-L-lactate as the sole carbon source, the amount of respiration of L-lactate induced was at least 10 times the amount of the other activities. When 100% L-lactate oxidation was measured all the other activities were reduced to less than 8% of that of L-lactate and therefore could be neglected in most of the cases (Fig. 6). Very little L-lactate was metabolized when the bacteria were grown on glycerol (Fig. 6).
1
2
-r 3
a
d
Fig. 6. Specificity of the enzyme electode for L-lactate. Respiration obtained with the enzyme electrode for L-lactate. Respiratory substrates used: a 5 mM L-lactate; b 5 mM o-lactate; ( 50 mM succinate; d 2 mM NADH; e 30 mM pyruvate. Exp. l-bacteria were grown on 1% D L-lactate; exp. 2-bacteria were grown on 1% glycerol; exp. 3-bacteria were grown on 1% D L-lactate and sonicated; 100% was 180 ~moles/min/mg 02 consumed.
Enzyme electrode for L-lactate measurements
37
Intact induced bacteria were more specific than the same bacteria broken because of the contamination by NADH or pyruvate respiration (Fig. 6). Assay of various inhibitors of the L-lactate enzyme electrode The L-lactate enzyme electrode can also be used to measure pyrqate concentration, as shown in Fig. 7. L-lactate was used at Km = 5 mM concentration. An apparent K1 = 50 mM can be determined for pyruvate. In the same way approximate apparent K1 for D-lactate, oxalate and oxamate were respectively 50,100 and 200 mM. The assay of these inhibitors was not very sensitive. When bacteria grown on L-lactate as the sole carbon source were used, no significant respiration of D-lactate or pyruvate was observed with the enzyme electrode prepared with the L-lactate bacteria. D-lactate and
PYRUVATE
CONCENTRATION (mM)
Fig. 7. Titration curve for the measurement of inhibition of L-lactate respiration by pyruvate. L-lactate was added to each sample at 5 mM (Km). Successive additions of pyruvate inhibit L-lactate respiration.
38
E. Adamowicz, C. Burstein
pyruvate were, on the contrary, capable of inhibition of L-lactate respiration (Fig. 7). The kinetics of inhibition are complex: pseudocompetitive inhibition was observed; D-lactate, pyruvate, oxalate and oxamate (inhibitors of L-lactate respiration) are probably interacting with the Llactate llavoprotein. Utilization of the enzyme electrode for L-lactate determination biological media
in various
L-lactate is involved in anaerobic degradation of sugars. In medicine its level in blood is related to muscle diseases, critical care, lymphomas, sarcomas, etc. The L-lactate content in the blood of sportsmen can be a measure of physical exercise. In the food industry L-lactate measurements may be used to follow fermentation in wine, beer, fruit juices and milk products. The commercially available L-lactate assay (Boehringer-Mannheim UV method using NAD’ dependent L-lactate dehydrogenase) took too long and was not easy to perform. With the enzyme electrode the assay took 3 min with an error of less than 3%, and the cost of bacteria was at least 10 times lower than the NAD’ dependent lactate dehydrogenase. Our work was performed with a buffer and pure L-lactate. Although oxygen consumption can be measured in a turbid medium, inhibitors or activators may interfere with the assay in biological media. For each new medium it will be necessary to optimize the assay. Assay of L-lactate in yogurt with the enzyme electrode utilizing immobilized respiratory chain from E. coli The L-lactate concentration in yogurt was determined. In yogurt the Llactate concentration varies with the amount of lactic bacteria used and with the preservation conditions; in the present work values between 70 and 90 mM were found. Our titration curve of L-lactate (Fig. 8) showed that the amount of L-lactate in crude yogurt is too high to give a quantitative assay. After dilution of yogurt (20 to 160 times in 50 IIIM triethanolamine pH 7*6), the final pH was 7.6 and L-lactate concentration measured between O-5 and 5 tIIM. To obtain the actual lactate concentration it was necessary to multiply by the dilution factor. Even though the yogurt dilutions were turbid they could be used directly for L-lactate measurements. Various dilutions gave the same final concentration of L-lactate 86 * 2 mM (Table 2). If the assay was done repetitively 160 times with one dilution, the error was in the order of 4% (Fig. 8). This assay was compared with the use of L-lactate dehydrogenase plus
Enzyme electrode for L-lactate measurements
MEASURE
39
NUMBER
#
I
5
10
L-LACTATE
I
I 5’
CONCENTRATION (mM)
Fig. 8. L-lactate concentration
determined in yogurt. The L-lactate enzyme electrode was first calibrated with L-lactate (-•-). Then various concentrations of yogurt diluted in 50m~ triethanolamine pH 7.6 were measured automatically 10 times for samples 1,2,4,5 and 160 times for sample 3 (-------). This allowed the determination of the L-lactate in yogurt (Table 2).
NAD+ according to Boehringer (Table 3). Instead of 3 min, 30 min were necessary and the soluble enzyme (NAD’ dependent L-lactate dehydrogenase) could not be reused for the next assay. In Table 3 the two assays of L-lactate were compared in a yogurt (different from the one above and with a new L-lactate content); comparable values were obtained, 74 + 1 mM, with the two assays. Measurements of L-lactate in wine Dilution of wine 10 and 20 times in 50 mM triethanolamine pH 7-6 gave good measurements of L-lactate as shown in Fig. 9. The L-lactate concentration
E. Adamowicz, C. Burstein
TABLE 3 Measurements of L-lactate L-lactate (mM) Measurement I 2 3
Means
Enzyme electrode 74.2 75.5 73.8 74.5 r I
L-lactute dehydrogenuse ussuy 72.8 74.6 74.2 74.6 +-2
Concentration in yogurt with an enzyme electrode and with the L-lactate-dehydrogenase NAD+ dependent.
was 32 + 2 mM (Table 2). If the same sample was measured 100 times with the Enzymat the error was in the order of 5%. When the L-lactate assay in wine was performed according to Boehringer with L-lactate dehydrogenase in the presence of NAD’, values were similar to those obtained with the enzyme electrode. It was concluded that the L-malate and ethanol present in the wine sample (dilution of wine was 10 to 20 times) were not disturbing the enzyme electrode assay. Measurements of L-lactate in blood Measurements have been done with freshly collected blood. In normal blood the concentration of L-lactate found was 1 mM. For sportsmen during exercise the concentration of L-lactate increased (depending on the effort), reaching 20 ITIM. The main problem was in the use of small samples of blood (below loo j.&l). If the sample had to be stored at 4”C, an inhibitor of lactic acid metabolism was necessary.
CONCLUSIONS Many L-lactate determinations have been reported using immobilized enzymes coupled to an electrode. Some are potentiometric assays (Shinbo et al., 1979) while others are amperometric assays (Durliat & Comtat, 1980) using ferricyanide or NAD’ as an electron acceptor. Other electrodes utilize oxygen as an electron acceptor. These electrodes utilize purified oxidases from Pediococcus species (Mizutami et al., 1983; Mizutami et al., 19&l), from Mycobacterium smegmatis (Mascini et al., 1983)) or from Acetobacter peroxydans (Cannon et al., 1984).
41
Enzyme electrode for L-lactate measurements NUMBER
MEASURE ,
5P
*
.
.
1
‘V
1 i
2
L-LACTATE
9
1
.
.
8
I 4
CONCENTRATION (mM)
Fig. 9. L-lactate concentration determined in wine. A titration curve was performed with the Enzymat for L-lactate (-•-). Measurements of L-lactate were performed automatically with the specific enzyme electrode. Various dilutions of wine were introduced after dilution in 50 mM triethanolamine pH 7.6 and were measured automatically 10 times for samples 1,3 and 100 times for sample 2 (-------). This allowed the determination of the L-lactate concentration in wine (Table 2).
Our enzyme electrode for measuring L-lactate utilizes a crude multienzymatic membrane system. The cost of the enzymatic material is low. E. coii grown on a minimum medium with D, L-lactate as the sole carbon source was a good source of an L-lactate specific respiratory chain to build an enzyme electrode with an oxygen probe. The enzyme source is very easy to obtain. Intact bacteria (without enzyme purification) gave the best yield after immobilization. Probably because of its biological membrane location, the respiratory chain was already stabilized. The measurement of L-lactate using the respiratory chain was independent of addition of cofactors (NAD’ or NADP’) or coenzymes (FAD or FMN). Only concentrations of oxygen were measured, and
42
E. Adamowicz, C. Burstein
optically clear solutions were not needed (unlike NADH measurement by spectrophotometry or fluorometry). Most of the assays can be performed directly, with the sample diluted if necessary. The assay lasts 3 min and can be repeated automatically. L-lactate is involved in anaerobic degradation of sugars. In medicine, knowing its level in blood is very important for sports, muscle diseases, coma, cancer, etc. In the food industry L-lactate may be a good indicator to improve and control wine, juice and milk fermentation. Each problem, however, may need specialized research to adapt the enzyme electrode to measurement of L-lactate not only in a buffer but in various biological media. The output of the electrode is limited by the rate of respiration of the bacterial protein. As shown in Fig. 4, oxygen consumption is proportional to the protein concentration (activity of the respiratory chain) in the film. The response is determined by the concentration of delivery of the analyte to the electrode as shown by the titration curve of Fig. 8. The fact that ‘intact’ bacteria and broken bacteria gave similar L-lactate oxidation (Table 1) suggests that no active transport is present in the immobilized bacteria. The complete inhibition of L-lactate respiration by 10 mM K-cyanide demonstrates that only the respiratory chain is involved in L-lactate metabolism when measured with immobilized bacteria or inverted vesicles. Stirring or increasing the flow rate in the automatic instrument produces an observable variation in the response of the electrode. Maintaining stirring or flow rate constant during calibration and measurements provides results which are in good correlation with an independent assay using another enzyme in solution (L-lactate dehydrogenase + NAD’) provided by Boehringer (Table 3). In our conditions, transport through the enzymatic film of the electrode does not seem to be the limiting factor of the assay. Nevertheless, in the case of low concentrations of substrates and high concentrations of immobilized enzymes, results).
diffusion
was observed
to be the limiting factor
(unpublished
ACKNOWLEDGEMENTS The authors have benefited from many fruitful discussions with D. Thomas and J. L. Romette of the Universite de Technologie de Compiegne. This work was supported by Le Centre National de la Recherche Scientifique, 1’UniversitC Paris 7, l’ANVAR, I’UniversitC de Technologie de Compiegne.
Enzyme electrode for L-lactate measuremerus
43
REFERENCES Aizawa, M. (1983). Molecular recognition and chemical amplification biosensors. In: Procs of the International Meeting on Chemical Sensors, (Ed. Seiyama, T., Fueki, K., Shiokawa, J. & Suzuki, S.), 683-92. Belghit, H. (1985). Production et extraction-purification d’une alcool oxydase. Realisation et developpement d’un capteur a alcool. These de DocteurIngenieur, Universite de Technologie de Compibgne. Bragg, P. D. (1980). The respiratory system of Escherichia coli. In: Diversity of Bacterial Respiratory Systems Vol. 1 (Ed. Konwels, C. J.). CRC Press Inc., Boca Raton, Florida, 115-36. Burstein, C., Adamowicz, E., Boucherit, K., Rabouille, C. & Romette, J. L. (1986). Immobilized respiratory chain activities from E. coli utilized to measure D- and L-lactate, succinate, C-malate, 3-glycero-phosphate, pyruvate or NAD(P)H, Appl. Biochem. and Biotechnol., 12,1-16. Burstein, C., Ounissi, H., Legoy, M. D., Gellf, G. & Thomas, D. (1981). Recycling of NAD+ using coimmobilized alcohol dehydrogenase and E. coli, Appl.
Biochem.
Biotechnol.,
6,329-38.
Cannon, J. J., Chen, L. F., Flickinger, M. C. & Tsao, G. T. (1984). The development of an immobilized lactate oxidase system for lactic acid analysis, Biotechnol.
Bioeng.,
26,167-73.
Clark, L. C. & Lyons, C. (1962). Electrode system for continuous monitoring in cardiovascular surgery, Ann. NY Acad. Sci., 102,29-45. Durliat, M. & Comtat, M. (1980). Reagentless amperometric lactate electrode, Anal. Chem., U, 2109-12. Futai, M. & Kimura, H. (1977). Inducible membrane bound L-lactate dehydrogenase from E. coli, J. Biol. Chem., 252,5820-70. Guilbault, G. G. (1982). Immobilized enzymes as analytical reagents, Appl. Biochem.
Biotechnol.,
7,85-98.
Ingledew, W. J. & Poole, R. K. (1984). The respiratory chains of Escherichia coli, Microbial.
Revs 48222-71.
Lowry, 0. H., Rosebrough, N. J., Fat-r, A. L. & Randall, R. J. (1951). Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193,265-75. Mascini, M., Palleschi, G. & Moscone, D. (1983). Lactate electrode obtained with lactate oxidase and an oxygen electrode for serum samples. In: Procs of the International Meeting on Chemical Sensors (Ed. Seiyama, T., Fueki, K., Shiokawa, J. & Suzuki, S.), 603-8. Mizutami, F., Sasaki, K. & Shimura, Y. (1983). Sequential determination of L-lactate and lactate dehydrogenase with immobilized enzyme electrode, Anal. Chem.,
55,358.
Mizutami, F., Shimura, Y. & Tsuda, K. (1984). Catalytic assay of L-lactate or pyruvate with an enzyme electrode based on immobilized lactate oxidase and lactate dehydrogenase, Chem. Lett., 2,199-202. Romette, J. L. & Yang, J. S. (1983). Enzyme electrode for specific determination of L-lysine, Biotechnol. Bioeng., 25,255766. Shinbo, T., Sugiura, M. & Kamo, N. (1979). Potentiometric enzyme electrode for lactate, Anal. Chem., 51, 1004. Updike, S. J. & Hicks, G. D. (1967). The enzyme electrode, Nature, 214,986-8.