Biosensors & Bioelectronics 7 (1992) 213-271
A novel microbial sensor using luminous bacteria Soomi Lee, Koji Sode, Keijiro Nakanishi, Jean-Louis lsao Karubet Research Centre for Advanced
(Received 12 February
Science and Technology, University Tokyo, 153 Japan
Marty,*
of Tokyo, 4-6-l
1990; revised version received 25 February
Eiichi Tamiya &
Komaba, Meguro-ku,
1991: accepted 17 July 1991)
Abstract: A novel microbial sensor system that uses luminous bacteria was developed for the determination of both glucose and toxic compounds. The sensor system consisted of a membrane with luminous bacteria immobilized upon it and a photomultiplier. Measurements were based on the in vivo intensity of the light emitted by the bacteria, as this is affected by their environment. A linear relationship was observed between increased luminescence and concentrations of glucose between 0.05 mM and 0.55 mM. The relative standard deviation was 10% for 0.55 mM glucose (n = 10). Toxic compounds such as benzalkonium chloride, sodium dodecyl sulphate and chromium(W) were also detected by measuring the decrease in luminescence in their presence. Keywords: Photomicrobial sensor, luminous bacteria, flow injection glucose measurement, toxic compound measurement.
INTRODUCTION Conventional microbial sensors contain immobilized microorganisms, which function as recognition elements, and electrochemical devices (Turner et al., 1987). Chemical compounds are detected by measuring the changes they induce in the respiration of the microorganisms, or on their production of metabolites. However, because the sensors measure the level of the compound at the sensing part of the electrode, the response properties are determined not only by the nature of the microorganism but also by the diffusion and reaction properties of the substrate at the electrode. We have investigated the use of luminous *Present address: GERAP, IUT Biologie Applique. Chemin de la Passio Vella, 66025 Perpignan, France. t To whom correspondence should be addressed. 0965-5663/92/$05.00 @ 1992 Elsevier Science Publishers
analysis,
bacteria, which are widespread in the sea. In vivo luminescence of these bacteria occurs by the following enzyme reactions (Lavi et al., 1981): NAD(P)H
+ H+ + FMN
NAD(P)H-FMN oxidoreductase
NAD(P) + FMNH2 E+FMNH,-
E - FMNH2
E - FMNH2 + RCHO + O2 + RCOOH + H20 + hv
FMN
where RCHO = aliphatic aldehyde; RCOOH = fatty acid; E = luciferase; NAD = nicotinamide adenine dinucleotide; FMN = flavin mononucleotide. Therefore, in vivo luminescence is strongly affected by any change in the external conditions that leads to changes in the intracellular concentrations of NAD(P)H, FMNH2, ATP and aldehyde (the enzymatic synthesis of aldehyde requires ATP as a co-factor). Ltd.
273
S. Lee
Biosensors & Bioelectronics
et al.
There have been some reports on the use of luminous bacteria in industrial and analytical applications. For example, Makiguchi et al. (1979,1980a,b) described the screening, characterization and immobilization of luminous bacteria, and their applications in fishing. Seratet al. (1967, 1969) proposed the measurement of air pollutants using luminous bacteria, while Ulitzur etal. (1980) and Naveh etal. (1984) discussed the detection of mutagenic compounds. Curry er al. (1990) investigated the effects of anaesthetics on the bacterial luciferase enzyme. Bulich & Isenberg (1981) and Kamlet et ~2. (1986) have developed and marketed a bioassay system (Microtox”+‘) for the assessment of toxicity in aquatic samples. However, the Microtox system is not suitable for on-line monitoring or continuous measurement of toxic compounds, even though it can determine their levels in routine bioassays using freezedried luminous bacteria. None of these applications involved the use of luminous bacteria as molecular recognition components. We believed that a combination of luminous bacteria and a photomultiplier could be expected to produce a sensor system whose response properties would be determined by the nature of the microorganisms alone. We therefore constructed a sensor of this type, and attempted the rapid measurement of glucose and toxic compounds on the basis of changes in luminescence intensity, using flow injection analysis (FIA).
MATERIALS
AND METHODS
N. Makiguchi (Mitsui Toatsu Chemicals Inc., Yokohama, Japan). The cells were cultured in aerobic conditions at 30°C for 24 h in a medium containing polypeptone (20 g I-‘), yeast extract (1 g I-‘), glycerol (5 g 1-l) and NaCl (30 g 1-l). The pH of the medium was adjusted to 7.0-7.2 using 0.1 N NaOH and it was sterilized at 120°C for 15 min. After 24 h of cultivation. I? phosphoreum cells were centrifuged (8000 X g, 4’C, 5 min). The resulting pellet was resuspended in a 0.1 mM pH 7.0 phosphate buffer containing 3% w/v NaCl and set at the cell concentration of ODm = 1.0. One millilitre of diluted bacterial suspension was dropped on to a cellulose nitrate membrane filter (O-45,um pore size; Advantec Toyo, Japan) under slight suction, and cells were immobilized on to the membrane. Apparatus A diagram of the flow cell is shown in Fig. l(a). The membrane was placed on the surface of a silicone rubber spacer and fixed with two acryl plate covers. Figure l(b) is a schematic diagram of the sensor system. The system consisted of a flow cell (2.5 X 2.5 cm, height 1.2 cm, volume 50~1) a photometer equipped with photomultiplier (CHEM-GLOW photometer 54-7441, American Instrument Co.), a peristaltic pump (A’ITO.SJ1211) a water bath, and a stripchart recorder (TOA Electronics Ltd, EPR 200A). The flow cell surface was placed facing the photomultiplier tube of a photometer.
Chemicals
Procedure
Polypeptone, yeast extract and agar were purchased from Difco Laboratories. Sodium dodecyl sulphate, benzalkonium chloride and chromium(VI) were from Wako Pure Chemical Industries, Kozakai Chemical Co. and Kanto Chemical Co. Inc., respectively. Glucose and amino acids were obtained from Nippon Rikagakuyakuhin Co. Ltd. Other reagents were commercially available or of laboratory grade. Deionized water was used in all procedures.
The temperature of the flow cell was maintained at 30°C. Using the pump, phosphate buffer (0.1 mM, pH 7.0) containing 3% w/v NaCl was continuously transferred to the flow cell at a rate of 1.2 ml min-‘. When the baseline of the luminescence intensity reading reached a constant value, a sample was injected into the flow cell using a microsyringe. The observed changes in luminescence intensity were monitored by the stripchart recorder, and expressed in arbitrary units. When the toxic compounds were being measured, 0.55,~~ glucose was added to the carrier solution to obtain good response stabilization and a high luminescent baseline.
Cultivation and immobilization of microorganisms The strain of bacteria used was Photobacterium MT10204, kindly provided by Dr
phosphoreum 274
A novel microbial sensor
Biosensors & Bioelectronics
3 1
(b)
69
Fig.1.(a) Flow cell: 1, actyl-plate cover: 2, microorganisms immobilized on the membrane: 3, silicon rubber spacer. (b) Schematic diagram of photomicrobial sensor: 1, incubator; 2, buffer solution reservoir: 3. injection port; 4, luminometer; 5, dark box; 6, immobilized bacteria: 7, silicon rubber spacer: 8, photomultiplier tube: 9. electrometer: IO, recorder; 11, peristaltic pump.
RESULTS AND DISCUSSION Detection of glucose
We first investigated the sensor response to glucose, which is known to be assimilated by this bacterium. Figure 2 shows a typical response curve. After the injection of 20~1 of a 0.55 mM glucose solution, the emission intensity increased immediately, reaching a maximum within 1 min. The sensor response returned to the baseline
0
3 Time (min)
6
Fig. 2. Time course of photomicrobial sensor response (glucose). Flow rate: I.2 ml min-‘; pH: 7.0; injection volume: 20 ul; glucose concentration: 0.55 mM.
value within 6 min at a flow rate of 1.2 ml min-‘. The bioluminescence of the bacteria depends on the presence of enzyme co-factors such as NADH, ATP and FMNH2. The increased emission intensity can be ascribed to the various enzyme co-factors concerned with bioluminescence, which were regenerated and activated by the addition of glucose. The recovery from maximum response to baseline values was more rapid than that of a microbial glucose sensor based on the measurement of changes in respiration, which required more than 10 min (Karube et al., 1979). Since the photomultiplier was separated from the flow cell, the response properties of the sensor mainly depended on the nature of the microorganisms. This allows the in vivo reaction of photobacteria to be measured in the absence of other physico-chemical factors. The selectivity of this sensor was examined by injecting samples containing various assimilable acids. The results, and organic sugars summarized in Table 1, show that the addition of sugars other than glucose only produced a slight response. We examined the effect of pH and flow rate on the sensor response and found that the maximum response was observed at pH 7-O and at a flow rate of 1.2 ml min-‘; at higher flow rates, reduced responses were observed. A calibration curve for glucose shows a linear 275
S. Lee et
Biosensors & Bioelectronics
al. TABLE 1
Response of photomicrobial sensor to various substrates Substrates
Change of luminescence (arbitrary units)
Glucose Galactose Maltose Fructose Lactose Saccharose Acetic acid Formic acid Lactic acid Citric acid
100 11 9.3 7.2 1.7 2.2 -1.4 -1.8 0 0
Substrate concentrations: sugars, 10 mM;organic acids, 5 mM; other conditions were the same as for Fig. 2. correlation between increased luminescence and glucose concentrations of 0.05-0.55 mM (Fig. 3). The reproducibility of the sensor response was then examined (Fig. 4). The relative standard deviation was found to be 10% for 0.55 mM glucose in a 20 ,ul sample (n = lo), for continuous detection over 6 h. After this, the response gradually fell. We assume that this decline results from the inactivation or degeneration of the enzyme or its co-factors. It may therefore be possible to extend the working life of the sensor by adding culture medium components at low concentrations to the buffer solution, thus reactivating these substances.
c 25
_
t
0
0,l
0.2
0.3
0.4
0.5
Glucose Concentration
086 103
1.4
(mM)
Fig. 3. Calibration curve for glucose. Optimal conditions were the same as for Fig. 2. c 15 ._ C 2 k
12-
g5
9-
=: E ._ 5 1
6-
x:
3-
E :
o-
~~~0.0000.~
.@.*
I
I 180
I 90
270
Time (min)
31
Fig. 4. The reproducibility of the photomicrobial sensor response. Optimal conditions were the same as for Fig. 2.
Detection of toxic compounds We then constructed a sensor system for the detection of toxic compounds using benzalkonium chloride (BC), which is a bactericide, sulphate (SDS) and sodium dodecyl chromium(VI) compounds. When 10~1 of 0.1% BC solutions (28.2 nM) were injected into the system, at various concentrations, the luminescence intensity decreased within a few seconds and the baseline shifted to negative values. Figure 5 shows that there was a good correlation between the amount of sample injected and the decrease in intensity. Presumably, BC caused a decrease in the number of living cells, or affected regeneration of co-factors. The measurement of SDS, and chromium(VI) was attempted in the same way. A good correlation was observed between decrease in 276
The concentration of BC inJected (nMJ I
10
I
I
20
30
The amount of sample with01
I
40
BC(~LI)
Fig. 5. Effect of benzalkonium chloride on the sensor baseline signal. BC, benzalkonium chloride: flow rate: 1.2ml mint; pH: 7.0; glucose concentration in carrier buffer: 0.55 P.M.
Biosensors & Bioelectronics
intensity and concentration of SDS. ECsO (the concentration that causes a 50% reduction in light output) can be estimated from our results. For BC, the EC% was found to be 23 nM; that of chromium(VI) was found to be 0.85 nM. These concentrations are slightly higher than those for the Microtox test system because in the flow injection systems the toxic compounds were not incubated with the microorganisms. Further optimization of the measurement procedure will also improve the sensitivity for environmental pollutants. The most important merit of this sensor system compared with the Microtox system is the differences in the response curves obtained for various toxic compounds. On the basis of these differences, the in vivo determination of the effects of environmental pollutants may be possible in the near future. Even though the mechanism of luminescence inhibition by these environmental pollutants remains unknown, this system was still useful for the determination of toxicity.
A novel microbial sensor
site with differential
sensitivity. Biochemistry. 29,
4641-52.
Kamlet. M. J., Doherty. R. M., Veith, G. D. &Taft, R. W. (1986). Solubility properties in polymers and biological media. 7. An analysis of toxicant properties that influence inhibition of bioluminescence in Photobacterium Phosphoreum (the Microtox test). Environ. Sci. Technol., 20, 690-5. Karube, I., Mitsuda, S. & Suzuki, S. (1979). Glucose sensor using immobilized whole cells of Pseudomonas jluorescens. Eur. Biotechnol.. 7, 343-50.
J.
Appl.
Microbial.
Lavi, J. T., Lovgren, T. N.-E. & Raunio, R. P. (1981). Comparison of luminous bacteria and their bioluminescence-linked enzyme activities. FEMS Microbial. L&t.. 11, 197-9. Makiguchi, N., Arita, M. & Asai, Y. (1979). Isolation, identification, and several characteristics of luminous bacteria. J. Gen. Appl. Microbial.. 25, 387-96.
Makiguchi, N., Arita, M. & Asai, Y. (1980a). Immobilization of a luminous bacterium and light intensity of luminous materials. J. Ferment. Technol.. 58, 17-21.
Makiguchi, N., Arita, M. & Asai, Y. (1980b). Optimum cultural conditions for strong light production by
ACKNOWLEDGEMENTS The author wishes to acknowledge helpful assistance by Mr S. Abe in the experiments, and MS E. N. Navera for editing this paper.
REFERENCES Bulich, A. A. & Isenberg, D. L. (1981). Use of the luminescent bacterial system for the rapid assessment of aquatic toxicity. ISA Trans., 20, 29-33. Cwry, S., Lieb, W. R. & Franks, N. P. (1990). Effects of
general anaesthetics on the bacterial luciferase enzyme from vibrio harveyi: An anaesthetic target
Photobacterium phosphoreum. Microbial., 26, 75-83.
J
Gen.
Appl.
Naveh, A., Potasman, I., Bassan, H. & Ulitzur, S. (1984). A new rapid and sensitive bioluminescence assay for antibiotics that inhibit protein synthesis. 1 Appl. Bacterial., 56, 457-63. Serat, W. F., Budinger, F. E. & Mueller. P. K (1967). Toxicity evaluation of air pollutants by use of luminescent bacteria. Atmospheric Environment. 1, 21-32. Serat, W. F., Kyono, J. & Mueller, P. K (1969). Measuring the effect of air pollutants on bacterial luminescence: A simplified procedure. Atmospheric Environment, 3, 303-9.
Turner,
A. P. F., Karube,
I. & Wilson,
G. (1987).
Biosensors (Fundamentals andApplications). Oxford
Science, pp. 13-29. Uhtzur, S., Weiser, I. & Yannai, S. (1980). A new, sensitive and simple bioluminescence test for mutagenic compounds. Mutat. Res.. 74, 113-24.
277
A novel microbial sensor using luminous bacteria Soomi Lee, Koji Sode, Keijiro Nakanishi, Jean-Louis lsao Karubet Research Centre for Advanced
(Received 12 February
Science and Technology, University Tokyo, 153 Japan
Marty,*
of Tokyo, 4-6-l
1990; revised version received 25 February
Eiichi Tamiya &
Komaba, Meguro-ku,
1991: accepted 17 July 1991)
Abstract: A novel microbial sensor system that uses luminous bacteria was developed for the determination of both glucose and toxic compounds. The sensor system consisted of a membrane with luminous bacteria immobilized upon it and a photomultiplier. Measurements were based on the in vivo intensity of the light emitted by the bacteria, as this is affected by their environment. A linear relationship was observed between increased luminescence and concentrations of glucose between 0.05 mM and 0.55 mM. The relative standard deviation was 10% for 0.55 mM glucose (n = 10). Toxic compounds such as benzalkonium chloride, sodium dodecyl sulphate and chromium(W) were also detected by measuring the decrease in luminescence in their presence. Keywords: Photomicrobial sensor, luminous bacteria, flow injection glucose measurement, toxic compound measurement.
INTRODUCTION Conventional microbial sensors contain immobilized microorganisms, which function as recognition elements, and electrochemical devices (Turner et al., 1987). Chemical compounds are detected by measuring the changes they induce in the respiration of the microorganisms, or on their production of metabolites. However, because the sensors measure the level of the compound at the sensing part of the electrode, the response properties are determined not only by the nature of the microorganism but also by the diffusion and reaction properties of the substrate at the electrode. We have investigated the use of luminous *Present address: GERAP, IUT Biologie Applique. Chemin de la Passio Vella, 66025 Perpignan, France. t To whom correspondence should be addressed. 0965-5663/92/$05.00 @ 1992 Elsevier Science Publishers
analysis,
bacteria, which are widespread in the sea. In vivo luminescence of these bacteria occurs by the following enzyme reactions (Lavi et al., 1981): NAD(P)H
+ H+ + FMN
NAD(P)H-FMN oxidoreductase
NAD(P) + FMNH2 E+FMNH,-
E - FMNH2
E - FMNH2 + RCHO + O2 + RCOOH + H20 + hv
FMN
where RCHO = aliphatic aldehyde; RCOOH = fatty acid; E = luciferase; NAD = nicotinamide adenine dinucleotide; FMN = flavin mononucleotide. Therefore, in vivo luminescence is strongly affected by any change in the external conditions that leads to changes in the intracellular concentrations of NAD(P)H, FMNH2, ATP and aldehyde (the enzymatic synthesis of aldehyde requires ATP as a co-factor). Ltd.
273
S. Lee
Biosensors & Bioelectronics
et al.
There have been some reports on the use of luminous bacteria in industrial and analytical applications. For example, Makiguchi et al. (1979,1980a,b) described the screening, characterization and immobilization of luminous bacteria, and their applications in fishing. Seratet al. (1967, 1969) proposed the measurement of air pollutants using luminous bacteria, while Ulitzur etal. (1980) and Naveh etal. (1984) discussed the detection of mutagenic compounds. Curry er al. (1990) investigated the effects of anaesthetics on the bacterial luciferase enzyme. Bulich & Isenberg (1981) and Kamlet et ~2. (1986) have developed and marketed a bioassay system (Microtox”+‘) for the assessment of toxicity in aquatic samples. However, the Microtox system is not suitable for on-line monitoring or continuous measurement of toxic compounds, even though it can determine their levels in routine bioassays using freezedried luminous bacteria. None of these applications involved the use of luminous bacteria as molecular recognition components. We believed that a combination of luminous bacteria and a photomultiplier could be expected to produce a sensor system whose response properties would be determined by the nature of the microorganisms alone. We therefore constructed a sensor of this type, and attempted the rapid measurement of glucose and toxic compounds on the basis of changes in luminescence intensity, using flow injection analysis (FIA).
MATERIALS
AND METHODS
N. Makiguchi (Mitsui Toatsu Chemicals Inc., Yokohama, Japan). The cells were cultured in aerobic conditions at 30°C for 24 h in a medium containing polypeptone (20 g I-‘), yeast extract (1 g I-‘), glycerol (5 g 1-l) and NaCl (30 g 1-l). The pH of the medium was adjusted to 7.0-7.2 using 0.1 N NaOH and it was sterilized at 120°C for 15 min. After 24 h of cultivation. I? phosphoreum cells were centrifuged (8000 X g, 4’C, 5 min). The resulting pellet was resuspended in a 0.1 mM pH 7.0 phosphate buffer containing 3% w/v NaCl and set at the cell concentration of ODm = 1.0. One millilitre of diluted bacterial suspension was dropped on to a cellulose nitrate membrane filter (O-45,um pore size; Advantec Toyo, Japan) under slight suction, and cells were immobilized on to the membrane. Apparatus A diagram of the flow cell is shown in Fig. l(a). The membrane was placed on the surface of a silicone rubber spacer and fixed with two acryl plate covers. Figure l(b) is a schematic diagram of the sensor system. The system consisted of a flow cell (2.5 X 2.5 cm, height 1.2 cm, volume 50~1) a photometer equipped with photomultiplier (CHEM-GLOW photometer 54-7441, American Instrument Co.), a peristaltic pump (A’ITO.SJ1211) a water bath, and a stripchart recorder (TOA Electronics Ltd, EPR 200A). The flow cell surface was placed facing the photomultiplier tube of a photometer.
Chemicals
Procedure
Polypeptone, yeast extract and agar were purchased from Difco Laboratories. Sodium dodecyl sulphate, benzalkonium chloride and chromium(VI) were from Wako Pure Chemical Industries, Kozakai Chemical Co. and Kanto Chemical Co. Inc., respectively. Glucose and amino acids were obtained from Nippon Rikagakuyakuhin Co. Ltd. Other reagents were commercially available or of laboratory grade. Deionized water was used in all procedures.
The temperature of the flow cell was maintained at 30°C. Using the pump, phosphate buffer (0.1 mM, pH 7.0) containing 3% w/v NaCl was continuously transferred to the flow cell at a rate of 1.2 ml min-‘. When the baseline of the luminescence intensity reading reached a constant value, a sample was injected into the flow cell using a microsyringe. The observed changes in luminescence intensity were monitored by the stripchart recorder, and expressed in arbitrary units. When the toxic compounds were being measured, 0.55,~~ glucose was added to the carrier solution to obtain good response stabilization and a high luminescent baseline.
Cultivation and immobilization of microorganisms The strain of bacteria used was Photobacterium MT10204, kindly provided by Dr
phosphoreum 274
A novel microbial sensor
Biosensors & Bioelectronics
3 1
(b)
69
Fig.1.(a) Flow cell: 1, actyl-plate cover: 2, microorganisms immobilized on the membrane: 3, silicon rubber spacer. (b) Schematic diagram of photomicrobial sensor: 1, incubator; 2, buffer solution reservoir: 3. injection port; 4, luminometer; 5, dark box; 6, immobilized bacteria: 7, silicon rubber spacer: 8, photomultiplier tube: 9. electrometer: IO, recorder; 11, peristaltic pump.
RESULTS AND DISCUSSION Detection of glucose
We first investigated the sensor response to glucose, which is known to be assimilated by this bacterium. Figure 2 shows a typical response curve. After the injection of 20~1 of a 0.55 mM glucose solution, the emission intensity increased immediately, reaching a maximum within 1 min. The sensor response returned to the baseline
0
3 Time (min)
6
Fig. 2. Time course of photomicrobial sensor response (glucose). Flow rate: I.2 ml min-‘; pH: 7.0; injection volume: 20 ul; glucose concentration: 0.55 mM.
value within 6 min at a flow rate of 1.2 ml min-‘. The bioluminescence of the bacteria depends on the presence of enzyme co-factors such as NADH, ATP and FMNH2. The increased emission intensity can be ascribed to the various enzyme co-factors concerned with bioluminescence, which were regenerated and activated by the addition of glucose. The recovery from maximum response to baseline values was more rapid than that of a microbial glucose sensor based on the measurement of changes in respiration, which required more than 10 min (Karube et al., 1979). Since the photomultiplier was separated from the flow cell, the response properties of the sensor mainly depended on the nature of the microorganisms. This allows the in vivo reaction of photobacteria to be measured in the absence of other physico-chemical factors. The selectivity of this sensor was examined by injecting samples containing various assimilable acids. The results, and organic sugars summarized in Table 1, show that the addition of sugars other than glucose only produced a slight response. We examined the effect of pH and flow rate on the sensor response and found that the maximum response was observed at pH 7-O and at a flow rate of 1.2 ml min-‘; at higher flow rates, reduced responses were observed. A calibration curve for glucose shows a linear 275
S. Lee et
Biosensors & Bioelectronics
al. TABLE 1
Response of photomicrobial sensor to various substrates Substrates
Change of luminescence (arbitrary units)
Glucose Galactose Maltose Fructose Lactose Saccharose Acetic acid Formic acid Lactic acid Citric acid
100 11 9.3 7.2 1.7 2.2 -1.4 -1.8 0 0
Substrate concentrations: sugars, 10 mM;organic acids, 5 mM; other conditions were the same as for Fig. 2. correlation between increased luminescence and glucose concentrations of 0.05-0.55 mM (Fig. 3). The reproducibility of the sensor response was then examined (Fig. 4). The relative standard deviation was found to be 10% for 0.55 mM glucose in a 20 ,ul sample (n = lo), for continuous detection over 6 h. After this, the response gradually fell. We assume that this decline results from the inactivation or degeneration of the enzyme or its co-factors. It may therefore be possible to extend the working life of the sensor by adding culture medium components at low concentrations to the buffer solution, thus reactivating these substances.
c 25
_
t
0
0,l
0.2
0.3
0.4
0.5
Glucose Concentration
086 103
1.4
(mM)
Fig. 3. Calibration curve for glucose. Optimal conditions were the same as for Fig. 2. c 15 ._ C 2 k
12-
g5
9-
=: E ._ 5 1
6-
x:
3-
E :
o-
~~~0.0000.~
.@.*
I
I 180
I 90
270
Time (min)
31
Fig. 4. The reproducibility of the photomicrobial sensor response. Optimal conditions were the same as for Fig. 2.
Detection of toxic compounds We then constructed a sensor system for the detection of toxic compounds using benzalkonium chloride (BC), which is a bactericide, sulphate (SDS) and sodium dodecyl chromium(VI) compounds. When 10~1 of 0.1% BC solutions (28.2 nM) were injected into the system, at various concentrations, the luminescence intensity decreased within a few seconds and the baseline shifted to negative values. Figure 5 shows that there was a good correlation between the amount of sample injected and the decrease in intensity. Presumably, BC caused a decrease in the number of living cells, or affected regeneration of co-factors. The measurement of SDS, and chromium(VI) was attempted in the same way. A good correlation was observed between decrease in 276
The concentration of BC inJected (nMJ I
10
I
I
20
30
The amount of sample with01
I
40
BC(~LI)
Fig. 5. Effect of benzalkonium chloride on the sensor baseline signal. BC, benzalkonium chloride: flow rate: 1.2ml mint; pH: 7.0; glucose concentration in carrier buffer: 0.55 P.M.
Biosensors & Bioelectronics
intensity and concentration of SDS. ECsO (the concentration that causes a 50% reduction in light output) can be estimated from our results. For BC, the EC% was found to be 23 nM; that of chromium(VI) was found to be 0.85 nM. These concentrations are slightly higher than those for the Microtox test system because in the flow injection systems the toxic compounds were not incubated with the microorganisms. Further optimization of the measurement procedure will also improve the sensitivity for environmental pollutants. The most important merit of this sensor system compared with the Microtox system is the differences in the response curves obtained for various toxic compounds. On the basis of these differences, the in vivo determination of the effects of environmental pollutants may be possible in the near future. Even though the mechanism of luminescence inhibition by these environmental pollutants remains unknown, this system was still useful for the determination of toxicity.
A novel microbial sensor
site with differential
sensitivity. Biochemistry. 29,
4641-52.
Kamlet. M. J., Doherty. R. M., Veith, G. D. &Taft, R. W. (1986). Solubility properties in polymers and biological media. 7. An analysis of toxicant properties that influence inhibition of bioluminescence in Photobacterium Phosphoreum (the Microtox test). Environ. Sci. Technol., 20, 690-5. Karube, I., Mitsuda, S. & Suzuki, S. (1979). Glucose sensor using immobilized whole cells of Pseudomonas jluorescens. Eur. Biotechnol.. 7, 343-50.
J.
Appl.
Microbial.
Lavi, J. T., Lovgren, T. N.-E. & Raunio, R. P. (1981). Comparison of luminous bacteria and their bioluminescence-linked enzyme activities. FEMS Microbial. L&t.. 11, 197-9. Makiguchi, N., Arita, M. & Asai, Y. (1979). Isolation, identification, and several characteristics of luminous bacteria. J. Gen. Appl. Microbial.. 25, 387-96.
Makiguchi, N., Arita, M. & Asai, Y. (1980a). Immobilization of a luminous bacterium and light intensity of luminous materials. J. Ferment. Technol.. 58, 17-21.
Makiguchi, N., Arita, M. & Asai, Y. (1980b). Optimum cultural conditions for strong light production by
ACKNOWLEDGEMENTS The author wishes to acknowledge helpful assistance by Mr S. Abe in the experiments, and MS E. N. Navera for editing this paper.
REFERENCES Bulich, A. A. & Isenberg, D. L. (1981). Use of the luminescent bacterial system for the rapid assessment of aquatic toxicity. ISA Trans., 20, 29-33. Cwry, S., Lieb, W. R. & Franks, N. P. (1990). Effects of
general anaesthetics on the bacterial luciferase enzyme from vibrio harveyi: An anaesthetic target
Photobacterium phosphoreum. Microbial., 26, 75-83.
J
Gen.
Appl.
Naveh, A., Potasman, I., Bassan, H. & Ulitzur, S. (1984). A new rapid and sensitive bioluminescence assay for antibiotics that inhibit protein synthesis. 1 Appl. Bacterial., 56, 457-63. Serat, W. F., Budinger, F. E. & Mueller. P. K (1967). Toxicity evaluation of air pollutants by use of luminescent bacteria. Atmospheric Environment. 1, 21-32. Serat, W. F., Kyono, J. & Mueller, P. K (1969). Measuring the effect of air pollutants on bacterial luminescence: A simplified procedure. Atmospheric Environment, 3, 303-9.
Turner,
A. P. F., Karube,
I. & Wilson,
G. (1987).
Biosensors (Fundamentals andApplications). Oxford
Science, pp. 13-29. Uhtzur, S., Weiser, I. & Yannai, S. (1980). A new, sensitive and simple bioluminescence test for mutagenic compounds. Mutat. Res.. 74, 113-24.
277
