Biosensors & Bioektronics

6 (199I) 15-20

Microbial Sensor System for Nondestructive Evaluation of Fish Meat Quality Masakazu Taiyo Central

Hoshi, Yasuhiko

Sasamoto,

Michio

R & D Institute. Taiyo Fisheries Co. Ltd. Tsukishima

Nonaka

3. Chuo-ku, Tokyo 104, Japan

Kenzo Toyama & Etsuo Watanabe Department

of Food Science &Technology,

(Received 28 September

Tokyo University of Fisheries, Konan 4. Minato-ku, Tokyo 108. Japan

1989; revised version received 23 February

1990; accepted 26 March 1990)

Abstract: A microbial sensor system consisting of the bacterium (Alteromonas putrefaciens) immobilized within membranes. a flow cell. an oxygen electrode, peristaltic pumps, a buffer tank. a thermostatically controlled bath and a recorder, was constructed for the nondestructive quality evaluation of bluefin tuna. The chemical compounds on fish meat surfaces which are the indicators of fish meat quality were rapidly determined by using the proposed sensor system. Fish meat quality was determined from the rate of current decrease of the sensor. Good correlations were obtained between fish meat quality and sensor response. One assay could be completed within one minute. Keywords: quality.

microbial

sensor

system. nondestructive

evaluation.

fish meat

INTRODUCTION

determinations (Karube et al., 1984; Watanabe et al., 19860) has been proposed.

The sensory qualities of food are important in achieving public acceptance. Rapid nondestructive methods would be useful in monitoring fish meat quality during processing, transportation and marketing. The concept of the sensory qualities of foods especially fish meat include a lot of factors such as freshness, odor, taste and so on. Freshness, odor or taste have been determined on the basis of indicators, such as volatile basic nitrogen and ATP-related compounds, which normally do not exist in the living tissues of fish and are formed by autolysis and/or microbial actions following the death of fish. From this standpoint, an enzyme sensor system forKvalue (fish freshness indicator)

However, various reactions, which contributed to the change of fish meat (e.g. glycolysis, ATP degradation, denaturation and degradation of proteins, oxidation of lipids) proceed at individual reaction rates. As mentioned previously, the original impetus for the microbial sensor system for fish meat quality came from the observation that during the storage of fish meat, large molecular weight compounds such as proteins or glycogen are gradually degraded into smaller molecular weight compounds, which can be utilized more readily by microorganisms (Watanabe et al., 1987). In both of these systems, flow injection analyses 15

Biosenso*s & Bioektmnics

095~5663/91/$03.50

Q 1991

Elsevier Science Publishers Ltd. England. Printed in Great Britain

Masakazu Hoshi, YasuhikoSasamoto. Michio Nonaka Kenzo Toyama,Etsuo Watanabe

16

were employed. Therefore, complicated pretreatments were required for analytical application. The objective of the present study was the development of a microbial sensor system for the nondestructive evaluation of fish meat quality and its application.

MATERIALS

AND METHODS

Fish Bluefin tuna (Thunus thynnus orientalis) and swordfish (Xiphiasgladius Linne) were purchased from a local retail store and kept at several temperatures. Samples were used to compare fish meat quality based on the conventional method with those obtained by a new sensor system. Culture of microorganism A meat spoilage-causing bacterium, Alteromonas putrefaciens, was grown aerobically at 298 K (25°C) for 24 h in a standard medium (0.5% polypeptone, 0.1% KzHPO,, 0.25% glucose and pH 7.2). Preparation of microbial sensor One ml of the A. putrefaciens suspension was filtered through a cellulose nitrate membrane (diameter 25 mm, 0.45 ,um pore size, Toyo Roshi Kaisha, Ltd). Enumeration of bacterial cell numbers was carried out by plating the test organism in the same medium with 2% agar at 298 K (25°C) for 48 h (Watanabe et al., 1987). The filter (3 mm diameter) was fixed to the tip of a polarography-type of oxygen electrode and covered with a cellulose acetate dialysis membrane.

.-w

Fig. I. Schematic diagram of the jlow cell. (4 Oxygen electrode: (2) dialysis membrane.

the flow rate of peristaltic pump 3B (Fig. 2) was kept slightly greater than that of peristaltic pump 3A throughout the experiment. The difference between the Bow *rates was regulated and the output current was kept constant because the

Apparatus and assay procedure Schematic diagrams of the flow cell and the sensor system are shown in Figs 1 and 2. The bottom of the flow cell (Fig. 1) is covered with a dialysis membrane. When fish meat sample is placed in contact with the bottom of the flow cell, the low molecular weight compounds contained in the fish meat surface diffuse through the membrane into the flow cell. To keep the pressure inside, the flowcell- below atmospheric pressure,

waste

Fig. 2. Schematic diagram of the sensorsystem.(I) Bugler tank: (2) water bath: (3A) peristaltic pump: (3B) periktaltic pump; (4)&w cell: (5) AMP: (6) rffotder; (7) sample.

Microbial sensor system for evaluating&h quality

determination of the pressure inside the flow cell was difftcult. The standard culture medium (same medium used for growing the bacterium) was transferred continuously to the sensor system by peristaltic pumps until the output current of the electrode stabilized. A phosphate buffer solution (0.05 M, pH 7.2) was then flowed continuously to the sensor to prevent pH variation that might be caused by diffusion of the sample compounds. After the output current became steady again, fish meat was left to stand in contact with the bottom of the flow cell for 30 s and the current decrease was recorded, followed by washing of the dialysis membrane with distilled water and a more rapid flowing of the buffer solution until the output current became steady again. Fish meat quality was determined in the following manner; the rate of current decrease A and B obtained from fish meat sample and standard culture medium, respectively, were recorded. Fish meat quality was expressed as the ratio A/B. Therefore, the smaller A/B ratio represents better fish meat quality.

17

ATP + ADP + AMP + IMP + HxR + Hx +U where ATP = adenosine-5’-triphosphate; ADP = adenosine-5’-diphosphate; AMP = adenosine-5’-monophosphate; IMP = inosine5’-monophosphate; HxR = inosine; Hx = hypoxanthine; U = uric acid. The relative concentration of these compounds drastically changes following the death of fish: K value

HxR + Hx x 100 ATP + ADP + AMP + IMP ‘+HxR+Hx

is based on the degradation of these compounds in fish meat (Saito et al., 1959). It has been shown that fish meat with K values below 20 (very fresh) are suitable for ‘sashimi’ (the raw fish). On the other hand, those between 20 and 40 (fresh) have to be cooked and those above 40 (not fresh) are not suitable for human consumption. Since there is a good correlation between K value and the sensor evaluation, K value is widely used as the indicator of fish freshness in Japan.

Preparation of standard gelatin gels The gels were prepared from 8% glucose, 2% gelatin and 4-14% sucrose. The moisture levels of these standard gels were adjusted with varying amounts of sucrose. Determination of glucose with the glucose sensor system The glucose contents of standard gels were determined with the glucose sensor system which was similar to the proposed microbial sensor system except for the use of a glucose electrode (Watanabe et al., 19866). This assay procedure was the same as that of the proposed sensor system.

RESULTS AND DISCUSSION Response of the microbial sensor system Typical response curves obtained from the sensor system are shown in Fig. 3. These response curves show the quality change of swordfish meat with storage time at room temperature (298 K (25°C)). The steady state at the beginning of the curves in Fig. 3 indicates the endogenous respiration level 1E.r

Determination of K value, the fish freshness indicator K value was determined by the ion exchange column chromatography method (Uchiyama & Ehira, 1970). The ion exchange column chromatography procedure is the standard method for the determination of K value in Japan. Immediately after the death of fish, ATP decomposition proceeds according to the following sequence:

I 0

I

10

I

20

I

30 Time W

I

I

40

60

I 60

Fig. 3. Response curves of the sensor. Sample (swo@ish) was allowed to stand in contact with the bottom of thepow cell and the output current was recorded. Storage times at 25°C were 0 h (0). I.5 h (o), 3.0 h (A), 5 5 h (0). Thefrow mte. temperature and pH wwe 1-Oml/mitt. 30°C and pH 7-2, respectively.

18

Masakazu Hoshi. Yasuhiko Sasamoto. Michio Nonaka Kenzo Toyama Etsuo Watanabe

of the immobilized microorganism. When fish meat was left to stand in contact with the bottom of the flow cell, the output current began to decrease within 5 s and reached a minimum value within 10 min. These observations indicated that the compounds diffused through the dialysis membrane into the cell and were assimilated by the immobilized microorganism. Oxygen consumption due to the respiratory intensity of the microorganism causes a decrease of dissolved oxygen in the membrane and consequently brings about a marked decrease in the output current of the sensor. The current decrease between the initial and the minimum current did not change with storage time. This phenomenon indicates that enough assimilates for bacterial growth exist in the fish meat sample; therefore, an estimation of fish meat quality from the current decrease was impossible. However, the initial slope of the response curves (the rate of oxygen consumption) varied with storage time or fish meat quality. As a result, the current decreases on the response curves (Fig. 3) were read at 5 s intervals and divided by 5 s. The results are shown in Fig. 4. The minimum value shown in Fig. 4 corresponds to the initial slope depicted in Fig. 3. Therefore, in this study, a peak height of a derivative curve was used for a fish meat quality indicator. Effects of membrane pore size on sensor response The output current of the sensor may be influenced by the pore size of the membrane used on the flow cell. The relationships between pore size and the sensor response (the rate of current decrease of the sensor) were determined by using the test medium (0.5% polypeptone, 0.1% KzHPOd, 0.25% glucose, 1.5% agar and pH 7-2). The response was almost constant regardless of the polycarbonate

membrane pore size. However, the rate of current decrease of the sensor using a polycarbonate membrane was more rapid than that of the sensor with dialysis. Thus, as the initial slope on the response curve was sharp, a correlation between fish meat quality and initial slope was not seen. This seemed to be because assimilates for bacterium passed straight through holes of polycarbonate membrane. Therefore, the dialysis membrane was employed in subsequent experiments. Effect of assay conditions on sensor response The establishment of assay conditions is necessary to enhance the sensitivity of the sensor. The influences of temperature and pH on the bacterium and effect of cell numbers on the output of the sensor have been described previously (Watanabe et al., 1987). Those results were also used for this sensor system. The effects of flow rate on sensor response were determined by using the test medium (0.5% polypeptone, 0.1% KzHPOd, 0.25% glucose, 1.5% agar and pH 7.2). These results are shown in Fig. 5. The output current obtained was almost constant at the flow rate above O-75 ml/min. When the flow rate was below 0.75 ml/min, it was too time-consuming for one assay to be suitable for routine work. Therefore, a flow rate of 1 ml/min was employed in the subsequent experiments. Effect of moisture levels on the surface of the sample on the output current of the sensor It was assumed that moisture on the surface of the

sample

;

-1 L

3

-2

L1

u’ g -3 5- x

I

I

I

20

30 Time (s)

I

I

40

50

I 60

Fig. 4. Derivative curves of the sensor response. Conditions were the same as in Fig. 3.

of

0

Q-

10

affect the rate of diffusion

u

b ::

0

would

I

aJ

-4

5

t -51 0

OA

o_

0

I 0.5

0

I I.0 Flow rate (mllmin)

0

I 1.5

Fig. 5. The eff2ct offlow rate on the output of sensor. Test medium was allowed to stand in contact with the bottom of the flow cell and the output current was recorded. The temperature and pH were 303 K (30°C) and pH 7.2, respectively.

Microbial sensor system for evaluating fish quality

assimilates. Therefore, the effect of moisture levels on the surface of the sample on the output current of the sensor was determined by using some standard gelatin gels which were prepared with 8% glucose, 4-14% sucrose and 2% gelatin. The moisture levels of these standard gels were adjusted to 85,82,78, and 75% with sucrose. The gels, made up with 78% water, and different amounts of glucose (lo-35%) were also determined. The glucose was nondestructively determined with the glucose sensor system which was similar to the proposed sensor system except for the glucose electrode. The output currents obtained from the gelatin gels increased with an increase in the moisture (Fig. 6). When water content was constant, good correlations were obtained between output current and glucose amount (Fig. 7). This indicates that water facilitates the diffusion of glucose to the flow cell. Therefore, moisture levels on the surface of fish meat samples must be kept constant. However, the solution of this problem is very difficult. In this experiment, fish meat sample was wrapped in polyvinylidine film (Saran Wrap, Asahikasei Kogyo KK.) and stored at 278 K (5°C) and 303 K (30°C) throughout the experimental period. A comparison of fish meat quality evaluated by the proposed microbial sensor system with the conventional method (K value)

19 0.3r

0-

40

Concentration

of glucose

bvt-%I

Fig. 7. Calibration curve of the glucose sensor for the gelatin gel. The moisture level of gelatin gels was constant. Measurements were performed under the same conditions described in Fig. 6.

microbial sensor system and the conventional method. The results are summarized in Fig. 8. Good comparative results were obtained between the values determined by the microbial sensor system and the conventional methods.

Reusability The bacterial membrane prepared was very stable. The sensor system can be used for more than 50 assays about 8 h continuously under the optimum conditions (flow rate, 1-Oml/min;

The quality of bluefin tuna stored at 278 K (5°C) or 303 K was evaluated by both the proposed

A iO.‘O K

z

z 0.05 E

I

0

2

Water

80 content

1

5 0: 70

90 P/J

Fig 6. The eject of water content on output of the glucose sensor. The standard gelatin gel was allowed to stand in contact with the bottom of cell and the output current was mcorded. The gels were prepared from 8% glucose, 2% gelatin and 4-14% sucrose (1: 14%. 2: 11%. 3: 7%. 4: 4%). Theflow rate, tempemture and pH were I.05 ml/min. 30°C and pH 7.5, respectively.

I

I

20

25 K value (%I Conventional method

Fig. 8. Correlation between K value and the output of sensor. Samples (blueJn tuna) were stored at 278 K (5°C); l or 303 K (30°C): 0. A is the rate ofcurrent decrease of the sensor obtained using sample: B is the mte of current decrease obtained using the standard culture medium.

20

Masakazu Hoshi, YacruhikoSasamoto. Michio Nonaka, Kenzo Toyama Etsuo Watanabe

temperature 30°C and pH 7.2). But, an exchange of membrane covering the cell was necessary, because plugging of pores in the dialysis membrane trends to decrease the output of the sensor system. In this paper, the dialysis membrane was exchanged every 20 assays. To summarize, the rates of current decrease of the sensor are based upon the increase of the compounds that can be assimilated by the immobilized microorganism in the sensor system. These compounds are associated with fish meat quality. They are formed by the chemical and enzymatic reactions in fish meat during storage and are related to the sensory qualities such as odor, flavor and freshness. The microbial sensor system proposed in this study determined nondestructively the chemical compounds that were proportional to the deterioration of fish meat quality. More detailed studies of the nondestructive evaluation of fish meat qualities by using the improved sensor system are in progress.

REFERENCES Karube, I., Matsuoka, H., Suzuki, S., Watanabe, E. & Toyama, K. (1984). Determination oftish freshness with an enzyme sensor systemJO Agric. FoodChem., 32, 314-9.

Saito, T.. Arai, K. & Matsuyoshi. M. (1959). A new method for estimating the freshness of fish. Nippon Suisan Gakkafihi. 24, 749-50.

Uchiyama, H. & Ehira, S. (1970).The current studies on the freshness of fish with special reference to nucleic acids and their related compounds. Nippon Suisan Gakkaishi, 36,977-92.

Watanabe, E., Endo, H., Takeuchi, N.. Hayashi, T. & Toyama. K. (1986a). Determination of fish freshness with a multielectrode enzyme sensor system. Nippon Suisan Gakkaishi, 52,489-95.

Watanabe, E., Endo, H., Shibamoto, N. & Toyama, K. (19866). Determination of glucose in fish muscle and serum with an enzyme sensor. Nippon Suisan Gakkaishi, 52, 71 l-7.

Watanabe, E., Nagumo. A., Hoshi, M.. Konagaya. S. & Tanaka, M. (1987). Microbial sensors for the detection offish fre.shness.J. FoodSci., 52,592-5.

Microbial sensor system for nondestructive evaluation of fish meat quality.

A microbial sensor system consisting of the bacterium (Alteromonas putrefaciens) immobilized within membranes, a flow cell, an oxygen electrode, peris...
528KB Sizes 0 Downloads 0 Views